From 14c2b0e5a24cf2318beb0c997a78aad942876aef Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Tue, 25 Feb 2025 17:54:56 +0000
Subject: [PATCH 01/33] icon for mac

---
 .gitignore                                  |   1 +
 ico_design/icon_v2.iconset/icon_128x128.png | Bin 0 -> 31444 bytes
 ico_design/icon_v2.iconset/icon_16x16.png   | Bin 0 -> 849 bytes
 ico_design/icon_v2.iconset/icon_256x256.png | Bin 0 -> 119441 bytes
 ico_design/icon_v2.iconset/icon_32x32.png   | Bin 0 -> 2551 bytes
 ico_design/icon_v2.iconset/icon_512x512.png | Bin 0 -> 195130 bytes
 ico_design/icon_v2.iconset/icon_64x64.png   | Bin 0 -> 8619 bytes
 7 files changed, 1 insertion(+)
 create mode 100644 ico_design/icon_v2.iconset/icon_128x128.png
 create mode 100644 ico_design/icon_v2.iconset/icon_16x16.png
 create mode 100644 ico_design/icon_v2.iconset/icon_256x256.png
 create mode 100644 ico_design/icon_v2.iconset/icon_32x32.png
 create mode 100644 ico_design/icon_v2.iconset/icon_512x512.png
 create mode 100644 ico_design/icon_v2.iconset/icon_64x64.png

diff --git a/.gitignore b/.gitignore
index 5e25991..608b4a4 100644
--- a/.gitignore
+++ b/.gitignore
@@ -36,6 +36,7 @@ logs/
 .gitignore
 
 ico_design/icon_design.ai
+ico_design/icon_v1.iconset/
 include/GLFW/glfw3.h
 include/GLFW/glfw3native.h
 include/libdeflate.h
diff --git a/ico_design/icon_v2.iconset/icon_128x128.png b/ico_design/icon_v2.iconset/icon_128x128.png
new file mode 100644
index 0000000000000000000000000000000000000000..9d99c859cc4b682599e7cc6f0605a52fc9b770ed
GIT binary patch
literal 31444
zcmV)JK)b(*P)<h;3K|Lk000e1NJLTq004jh004jp1^@s6!#-il00009a7bBm0004E
z0004E0beK3Qvd)U07*naRCt`6y?Kmf*?r&lIm=z%R(n@>S9eeEGd(jLvWLr%LyF``
zq!dx22$7VPKuc^OkrTrS<QRgHAPx{Hh!h(!5Cd@>3vnbxjvPv)Wy#t|nG(q%DN-EH
z3^|-0POsBF-PK#|uikQ(b58y^=iK+Ix=A~^K%w4p-(Ajc|NfRM{_kJ;nEM}IJx^!(
zoy;l0vrnDC)fI2eZ*k!06KHF&l|eg8T^np|k=Ej@rLGKBWiYiR!C`8HbPi`M&N{Fb
z;T*>L?~48gS!WL690I6o0}g6qG1eiRLkqw;aNwN7Iful5=JtFJ)p!@3wh@4_0Ab&-
zIqT2koC6{Jv4lUo4c`gj_pQniNz$UW4kaO8h7gn2{tQ#^#pGN9sEQFf&G+ml1XwrO
zCw?Kq%lO_okOJyzjMRB^T<aV%T;%jA<L5$5t}UL!)CEds{(U%qTmvz+pA;a!b+g4U
z9DbR>)}OQM1o!)6Zm$nO3EH_vry7wcPzoa?QcHxE2r1EtqI3e3L@0@q5-S8w%5Zc^
z4M-{c_rf8h|1IEO3lKt}r2h*a_(c)|DZ?cTfe^wkLZk`;A^qp^IFX`4g!6{qMaFQ5
z<`nT55UCwc?Y|UZ!Q2$1G&B#FAf`SOQatckv)J)<L;1<+BQfG-O#e=ZsUtZTf7&nw
zZt`1WC*SYiBJ;?2-QhR#ckO%G)MI{O=@MgmmmAw7HpUfBfYb`B6e`iEM1fW)tq@8z
zOpv9TL@RWnND_tA3KhN+QecI|;E9bCSUh2n(i2rlloDv=zn76<QsDW-6FIJ)40jN@
zUj(X6m>@2WXb6g+VVpqPhA}1#<zLI`1!p2PZV&bSOZ%|Zj?CM<oSdV%tSL3+z6Zso
zZow1wArkzs%}l8qH}zTkc>jIKVS0?;-26!{n>jwBUgU%CIKk5LEJg~fQaB+&NsLmA
zoxn<oRMInnk|-@PLg0j;LLildu#eXOf*L{Pz&JokoCuN|B~*HytE2!)Ah-ww_p2cu
zq6wJnhMSh=aPy6t4*SQy=`9}CI{zo8jyXBrM6#z9M*Q7F$T!(9n5AgGn!m}RAHKQC
zIg||7?;JL00XNw#T)S(I7w^UIAC&R526L0&>kmA6nk)C0xRO83+q2imzWU$u>h1wP
z@b32^6NPr5t)Q+AT5Gg&U<?wlz|PQ+35k=ABmryeo|7r%PzbECC?y*yw#Fi*4EEk4
zq{LY}MVW8{0Rrb$xAb=1GX(T2$6_9By8z)LCm_Rb9L_JY6v&1vCaN|11xkRL*3S>n
z)SqV$jopNR57Kepu^vX~aEym-4S}7bG>$}t`<mjIX+J|asGVQy3A09X=&!xe9#hw%
zPaT})jkP{ms#(pCpv_&L9<`AgvOHnzz}27yB$)+*BNc6zmS_P=IE;ft3NW5_N<iWq
z(mHDA{BMHfd-~e201|60LP(skdmJ105GJ^|KIS_cW3CP>KDadE9m_4i;apI&VQB=O
zA-oE9Q=d+p$;O{(R%JrC{n9DIg}JaxLx3rQ#h>gsSM%NEeN^U_DZ)&hD+*EVpfZ+3
zspdREDegI^5PPqC+P|1q|NHq+A0EHPbF;fBB}h!oo7E3<w09G?d5LSUD~dv5Dnq6;
z<8e)<1gVe+<ESdb`u2!iNpdL}7ZnPS2p<ht?`=!xCATj6j09Bv&apP=IUCfKwf<(p
z6<g=LP>267-K)5~Yq?!oT2hb;`1a<Qr)F(K&}Jd0-1f%Xo|fR|bI$~(E}PHyd)5zZ
zQ(sQ$EUc-hm|}=<ENk6fXYTtQ+Sb^*XpU>0Lpz77YiwYf_}v4Cw4OPgtsi>12C&n9
zj1Y*gy*YH<L{eKQjiZ%#8~e&{{StS#H~8PLpFs}0EY7#-W*Ob~BF{Z@mc>l*{EOFU
zrx_b-I|yq@oW<3KYG;d9Z-!PX=ylrERn1Z_XGv`_J9`M6!y4-s#@WCG4r47?uejs)
zWo3Ec)()?46>L<7Tnf(3WGv^3&)pgDuDP6NSK7#FJ4g)ky#0+fQckIC8yu^+9_4sQ
znzuc}hMlan-Y*HnZ0xUd$l%;r=TJgmox@aPlC;$*lo&?Zh65*`#p{gk<1V(|8wNJ!
zEfPL3>_nO;ui~XvWr0evJq#Y1X?jyapmQxLjQ=7sKy4hK-#9^CZ1S<?Z}I6jm*`h>
zlq1c`(jks4cF@k!N+i335l$&ut(0mwrqxXt?>MTmAQ6(adt*#dVu~4-4-0A`NCZeE
z4ue8^3;>Hl;)DeYcXumZxVgi%e!<8(T2k=rQk!?KbjX$9Gp}y~aCSaJN#RXN_!Bs!
z6fv?xHDWVSk+PxEB+T>fg+NKat5k9#KEi9ZC`nNjM+Z?NeB`u8!e|I1@Juolsq*q>
z#q--$v;X;y=Ivc*>B%u<(B~#FW|V&CTy&ABo#WHD3pR~qC6WBdJhZfT0KJKtFp-+V
z6r=zx1!(1^KGXgStsra7@VTqY{P;?X|NP8t{xrG9O!<>sxpS9a`y)dpVE4@{{P;&c
zz?C<*Siil_$s_Zu-`&8FAQFu;W0qEy8I5X=Ezh9E21-lH%By;$C%%$kk>D(^-5&Do
zJD%_WEh%{JaF3@Bb;zW5X1=>N<m#~Ev9{tsH)%i-DT1l7(M}2~>rf4?0|BjT?6C}s
zD}+GHAj#7%b|f4gu#!`u7{Mv6oL|VPZywkj)6SS+yEJ=<5AgD!rXvLBGs1N^5-E{P
zoWBPAi9UpL>{gadkA$&WS+0)-Pv-s(jhpgI9GJii&V}o)!_5olsEnhXNJ?WdRRvD-
z$HO82>2l3~Cg%A+%1vB0!@J-4D08{uv!C7K{?>@C-7z(=wz*5`B$cTtD}~gWqI3*L
zCF{E-b6tlOk}MTqEp-i2IGnY7^~xF-?+;iVdRx5GN}0<P7q&(qpp`0*>{+<7<(cEm
zjG9u{d(~T%=u`=hL<Z53&WBarzi3=bYz(p7L_<U=k19S<(yP%J(YeV!{`L2B9O88e
zjB{Kan1_CUVFc%c6A`5q0v-_6|Gs<Ga_=#wBhF|dqWYgk6_pg|R7pzXiC-FvR^Gm}
zW;!@IggWO_w-)%DkF@!KxWnhtE`^1{f|9!!YAlwdol}|uE5Kqf)>2hN$}(rv-=v;t
zfe>gdgIxj*nA$R`Ej#0y)nUo0nVh#C?(z7+E;vW_+C9Fo+6Um7V>8_Cmt5E>SV$D7
z=2WBVwGcs)C*HaWMmDPT81lshjRX+_Oh6&E2+rqpjOHd@whXSd4V<FF!Th()i|@Nh
zA$WC)0p8t9_<G-KgH7YOH8dP<DS{=6esbg7Mr{(QZ60#I$jsB4!%f|L{9Ie-sI5aQ
zs0{3kYVt&3ETk(RX4dV1srWzF7x?ky5BTW9edZ6;bcN!Xcf5n=|MF$BY>v~%4{`nK
zDhGQB*&~mTU};may*;A6bcpS7jT4SE)fi(C#)mjcLt7;L_|r!?etnZCkIdj~hzK0~
z=;O;U#RNNrU%0m}wm=vxA`sQv{l<&8VX;Jm#qAB7@QWHFhbSQ?EKcJvMVoE?8-Z(l
z0D@_V+G0;U_kEb))~IHqp2QF5dW!AZ^4*cYw`)Vg(YEr}WrM8`U>nddMhyS-wBx}f
z40Mt3LA&jbL<FFf5Ofm1IC)z(KF^~kmgqhHBw~AuZ{Jv>OmpU+{QK;F=U?*VxBf@|
ztXuIHUq8<O_KSaydmBT3?@L$s*^j-G7hb%^?14EtTCj0<n`h6TL;$miVEfKBq>>CP
zi*SNGk=Qz{n{_;Od>(6p0Ve=M0zUlM0j#^v7gqZ|{F8#W9`54CJ59hVfO~_AgB^`A
zu5r$a+LB0t5Dpy(8$Ep)ma(=DEoI{aSO>LrBvP<5HeQ=ZL82t76euRgch0M3wcq>)
z3Op?KzPdY+{*^@1O(aKh*)YKCBf~S+kxGG9;-Snm9bwe_%^xARGR+Y@hzAtLdT^|{
z-l-iHx(R0w^;npnVR>$jb0=3gdvuZY#bvrn9X>H^(VNeyr6S1^YNhG+GSW_m#rY0W
zYw}i(N)&VL#KXr@AhaY&lusZ@pG1;cg7SF*sXSDyq(Ft=C?)xka|bxt%NSY5<8v+Y
zRDlE?atbfp9`ak4H~Gfxfd`xfP2NF9vp{sB#1w?ao<F&7KP3fqSmb7{rfqZca<9GK
z_ffNz)-{9%I1?oOba&$VkFDeCZXGU3JoGs6vwu%`A>7l0pXuq(8npuKj0`Wny2`vA
zaO2(v<Gc5HeEtb;t&aJXVw0cfX?|2+<L>U5H}4MF*xcZajU#OD?yz>LWWJ@Tw>BwO
zI@~HOhdQ3<Qc4EKpqxM?a_`BV^(dk9(l0`oRvUwmf~}zbpE}%a2xSDkaC4V0-QA@n
z1&_?;4S|I4(y2oZIQj+#fePx|$MTa^b;1R;A8e!)$W%@tcOk-H3Q^oglxd;bZw@Z^
zYsOO&`}k}E2vP~>TB>1=D?`Jhy#%7yF5E*kigVu|w3@~ptecoF4{d`U*S3iXt%E^f
zNR{C7>NXedSyD4(EZRJh_W6z9|6>l%pW=gW{SmHjuahTV;nJV}BYtluV~NN4{m*`x
zR7r}FWv<g=u~qYh7q3#*g2ir&cOJHk5wtUfbQ0?<R!F4sv3lu(*Ny+&8jDntT)<Z@
zuCZCxoa$v9nadlS_RU+K__-8(^7K5*-GrugKzcLNc*P+*q@{<AO$Z+o1x4wZ#YOL1
z`Tg_YVo!cQjsC?@QZ$RbpW=}TB9X2RCn(+HEk)XhrQqn48F-_xRMsJtzs?CoC)Xl^
zNmKiy09=y{a`EDs_zaOFQYDe4aop=ytPc!#`xTob!{*3P8_S7Dj<K|Ik`srQdE#Bq
zvdtoAjvnN%{m?1?`bXbEe&*xc9ly*c^J~~daC>8w?(95shgRqwUZJ)42t(bWw=~E7
zK}n4z(TYwh@hV*T)UmH4n5rdE(&zG}^bU}eyuLmJ;IX9+Aw|ebcPqYl+xrEdIJ3ZV
zFAb^ZNqRWe6L{L}t0!=i&uv(u=+KGAM2Jb=BbWv6AUYpH2K-^Knx339dhrO@sVvuu
z3Huz))t+qP>0YvjkgxY^oP#>J`8FauQ=5`vPq?_35onOiM*63&P>Emy9+^vVLQvF}
zOiPUM4%3$|-Q>}^HW#->eDUI&yyw(WzI?0C>%a9fb#1wL;cL9@Ex$x@{SWyY2mXYA
zXb$t8%a_q2XK{X!eoJxq&@#hO#gpe&NaSs#RupB8RubWnUzG6aVy&IeA=FSfUmawH
z;I%imSRIzkr<%7Mo5cZNyS~ly_e)wz@`<yH9PTAeyi^CE=l-DL&7Fd`AL`)3B1#zv
zC{T(9;#Hykz&bHW-GV2+6Ifi6UY+Fl!5FtU3TPNB?lobKtGjzK`jb7q$1bN*@K`Q*
zee8dCd1!cQHfb_}k(kq|@9$*-Jdb-+R|P*PLfHbZ?W_YWMR0PAHwB~8q7{_m65%YH
zHArVUxG>AL``aAqw)hX8KE|M^Y3aA~(i`_k`W2mS!H;G4_=D6zb$~)FFczB4QAIHr
zaj)Ovk$Hhk6sZ7=!5D`U5^Eq$pca6!2n9Oi2T}n~9iQj==7>y5v{t-!Ynw0L+NPxh
zKXq=2gS`xEZDa4065QD-_{`NU#@2DHm$J~(!O@M7sy71=vGr&yeXJk%-GkfX0kdd+
z1R&%?e6+?2_!luJe`U(~d%UH{q-eg2+mUR-0Jkg4dSN(_dq;7P1c!4>@xq=h!r)R)
z@&?nX<cOR~DWP9@;#akgIs0P+5}sH&z)DN;#C)4a4=+#*`^+)s*nG;F11-)h&tf-L
z8I(`(hnqjl=I)SBEDTvXY#8j`BXvX8w|7ZeIospXr#_WzDu;wtjpy%+M`TlEq7oGf
zPkbcs(eo?3=gcBXNxpsCV*>9xm~m*X6{?4%_u7@@-k{<$S2h`2#|IB~IoL}>xf$H(
zm;67MxB2Sokiu9W&P8?4RB$jIeFdjbiWpT)d;F7~zgKi11>7GSR?CU{Kh>E6{HL}P
zEg$B_?bg^%p?dqae~$=2fOmYYYp{*!V@WLq^DTvzf=ml$GRZ<)<E-Pg3m3S#dWQ?w
zZt|rUUggBn0vGy*Pd$H?fAhjke*1G@=lwtMe!5G?xPN<tKkPq6HaFyxoo{k*`Bkpo
zeubHC#&|T~&{B`Qogx#>&<ScGNmEbcS_n{*cBar;;bPsOk{CIuHOw>Z!JBaB2fE6G
zGQp-xExFw<`Q3}_jIHDS2fI9dxWnB+&2L`V;Ke&b{^aH^o0Z}8d`6Zi5K!AFNdat$
zK7ojpI3eD46MG(O>wFa9f}^(&Y?)$$Yg4g*E(J$I_!o;*<CaM{kxjh1S9i?=HeO8q
zcjCb;U)-hvASK0~-$%Sd4od4QVZ--z07=p;r7A0SMkB^)&d&B4+v9@S*;$@^bcM%H
zEu*!>*_xql(V06+tM?fH>vwWyW);8K8?q}ea%X#!dz)MA76z20M4%H*k|;_fwU9`q
zD4d{h9t#j!Q43!qsI}JuT1#{YBhMe10pP+~iBg_9gp^$081v~s{O1n$c;;9Sr6lt$
zO{N52x<AALA3rh6VoNt^Ri#9O3GBrSqN?=%ffBwTC3@0LxXGl((DTiMOz^G2#QvYm
zB)Jj~kLAT)?XYPq_r~?!&~M-W?^SEy@U}kI6i)9KO@QvCl2K(t(!~GFO1~&)&mLzq
zDmmTl(mu7o>(}q{_QhFFEw#wB<VWB4HeS53jRfBI)<=2$;uYo>-^M?C<F9yItK_Fo
z9_5ej-KLWs;7-3!dC>UOaUu~$QrY08*TzeN5e)>Kr(kLj0S&aqCvp(B@d;kMv%~hw
zw>i{F*&bEAwlxCagDW$<>sYUG!M6&-W@V;s-e*8<EH4u=k>e4=NSTQo`oxix4r>L3
znA}X~7N)QV*Eot^SKS=#js3ru`j6YOwj!6J`OUR~Wie}Ty$6uyhmk*iucd6l!TtM*
zc#o|_QdE|rvLwnUM6xX5?%F0|xW=uu4Q^b%%(1RweS5?U7w+=qmv3?D+u!8<Cl`5Q
zWs&=L?{j$R0PWTSZ{9z}pWZveP{2o)ZZS5`Bh5{$@P(#DC@xJ>jntAPRV1lGC*J-?
zzd!~bA(oQI(sL&SKlZj2K6<i4ODnE#j~n9u;L%y0IWZTij(uIhr!Q>)@Dq>BvXCfV
zyf;Dygi!@YY)`4V*t>{PJ2WNb0iDxQPNZ?48WusDIN-{Z_n%9_(YD&>U+kq^A_d2%
z%)q6Q*+ahRGa2g>dHc7x_jDF;`ZO`^t)=iz;HV1rV(dSh>13QZv`Bw<$P-6aSky=O
z(yKT4k#mRn`462$2*IB(<^1b!-XQ6;Isf=+F1)$MvBlHW`7Bz(AFtoxr<Oj;hq}A`
z$<43wLuWolJ5NZQ#MZ`3iUg_}MFgmVMAkG|wiUjH(1kF|g=pdI!Hn}~miX<zdJ}->
zj?MAxsd=3BDds!<ir;;G4S=71bddv{gyVA=E$vgPaoLoF67NgM5WqFjL4%}-rqFac
z$cx1&4v$JUTEtC&m)D0A?K8HHe|BT|0KxWdgFWvyYRmm`&2p}ZR~`Ay=O`Z30kNJy
z?IV2T^T-eqygKY;9(3tsicTs?wdDTDvRjsv#`4mwJ}S*9s|xFUxVSm;5bgTTh~2`X
z(-f^WquMgGild7&{MxmeqNDi9`3jX-h89$+z4RwO5>RR4?R?@T9u;6?B|&?;FSJ60
za&xIBV3Aak=guzigU1(m@7X0BzUuf+zvNS|t$`5y^tmMt&1Og`Xr&79>a~bTBRQ2R
zjNaXpN^kJ^sT{&y?eKugXfT1xJLMkr|NnP;V`wHCYTs%1Z!tj?Nx%Q>)9>}_>tAEH
zFb%iwgxI%Ltn%|u9-}T5d3KkjGa28!u+7Pt+uXUhgLRH~FThTAlh$I3$B$Ng*X^*f
z(Bemq=2TS;^mrnB4wrQ3@h#ZA#LdeCHdGhkG*wwaWl5yBiLSDk+K@?!sf{mj4n^+H
z`j#_mEI5HRj$*V0IOY;Kf5<V~T?gm5yHoRfm%S8y{A7zcJ!U)@f^%G0t9f;^=5`3l
zPUey)<}^pLF?As@0)*YC(kFU;@-K!1Q~Oy{QWgaI`^CT9e|X`?|JUu((D3%&coO$V
ztjam+{ikpiX+n`}xJ1K)3?q<;FJBwDL<tJxniQz9j{X~8=g_(L2C&9EYenst3*n2i
zP=(mCHr{EB)i>5TDg%iUj4I2ZwB%aw<=_6hn4PO!KKfs<)IP@ZfBqGg4y`cLZqsYG
zDEb330ml!`F&Y;1GR?7toaMHntSwu+C3RWT)`FY&chRGJoIQRVSNj4|Yb_G^M}K~q
zk+FRI(F2@2*a?oM<Fnsg=cNs=ephnMY^u0Bs@bRv@0-^=b8NopTO_27<Ll;M46O)6
zm|}ps7-CiKE0+Wm#_=z24jT>dXiM?jfh^=7{92W@B~N^5oJ9D-(kurDFK?E6%Etb~
zgBb^M-K2pdXZ9=0|Ft@9_WPk3$zyYQ158c`zE@dEKzDMVkmvy8+D8Kt=rh@|c2JBB
zGcCoq3aGgd+}bQSzK|gW>=qW4Nvwmdv1KNc3`;|vNK&n6DZ#3mC(obY<elH+>ve+R
zEcaHwj6L{4wzdW=bW*HyeB*^zv4eHymyfga$U(l{-(@_id^}zoWL+{l*JGxta6(cG
zN1{Lo$=c4Ck+H0_Qyx3q!&e4NzHo7kmo|nhB$^MOnB|e#)Gv;8{MpqlzOp&y<U++#
zTQ{~_#FAn$k#yHAr3=x62=xVEtYvHrMQLbd>VeGPl7g*%K`T!vh3}M-QZOnlWhf9y
zu_WQNkIf`|7~qYeWx3^F%b*Hu2=n%DF*>nss*hkgevIdj*(d?DbVw`+UqY2DL9YD`
zE_D+!B^j2M`@1!(Tcr=<gs({HrqXw}8pm?m*Xs3hO{OKS;G{fu>Vs@8e3Yg6ZGL{a
zVCLjWL~DfuhvsSb+8jT=%)&y3v3*}!CKb1DZ*b|ASGe)o3ycOs#)A<%_wLhfrL=k(
zD2dYEXFv#+W^$fB*kg5A@`X$5loqb7kNC^GyR?+#=iau=BMZ6rBLeXA_QM?jE^Lf_
zjzah%5g`Pn^Ev#=Mm4$@3)r6s8W**%XmYSR-m9CBv}CAua%^vp7!FHPCFt)K)U}})
zl~h&TNNTPGCm)Cg_7nQyj(`e|v)A7HG=YfVmP41JMXRVRtt5K5P}Gh=<wzA+8!Bs*
zpq(n_TN-c-$`FDGUz#i}q*}03*ruON3J)%|wcyhAbw2;fo4mCAE`C7X<ZtmC{93Qe
zOD`!h)#CP*OSF_FOM5uE$!>p_!^c-R{NDFa7A5`70m4GoYjf*<pTl~HI1#G<4H)73
z0DkQJ5yZ>4c=pU9nU;L}_6`6aJ~2-#)uAfL24XrosrNx{kLo7L6AM7IP{SCb2A?Pr
z7<{#HEFlY6urEOqFsv+BrWoK<rVz$?bko87LWkN|`lFg&FZ1i4X_OOFkp>)XX|5IC
zp8IHEk{gJd-j@%K=%kzeVEX&;UWX1XjI}7~ZF*!8DLnU72HKgAcT3|@>DZsIl_&1F
z_U}vMRZc6J5Z_u)6yv;&l*Zsn%`5b{HJ|c>9fRz>!nk~cLGdbs@jZ%agrUZ0e}U~Z
zVYZ#2(gc|(+L@*{7NI>>pp?Xgynzi~{f|F=gifXrN;ZWgCl7Rdk6_3h#B5G6r9vYF
zjF`gw!!e_22#`?&uEdxhyh=C5a&w=w9!fpR36Ldn($Z~g(;6?Db4K$5r#gFyeQWBt
z->VR0zfNd!2~!AC6Atc-4Q1_u#QF=VLUnMieX)eGzPuWI5kV|a)N+z5R}#`tL=kJz
zo%2nNF|um)W+|4Q;P9a*7`i9;)xTP%$_M<|>>bV>zRJ;)o46#!q&X&?rxYnWwWUyk
zLinbV(mE=oJgDTrO8G#?vIu2jQTJt9O;%<a>z7^{Us&A*;L*8!Lj2gJEFaDULQakq
zVd3~zD#@Lxyg$IvmO=>7iE3&PRA`}(T1PhV{3HaONhK@kM8ac6G5+RzZDQYF^O>9w
zA3%;OuL_IWw^9u&8>&qGO{GebY7vMmD2>+uN_f?mDS?&~Z~?r@C~HfmBuWUcMuDaI
zHociyHcXG(We?YVhJX6v3I`ngVm6{*!!SEWk#s4OE@T<Caa_MQU~6b7k)Sn|6Ii8j
zN|Gh2NgN3&NFtUH7FS5g@p<n#efRF5QPWju8vV-kZC+d-awwBLb8My=Ug4soi%Dy`
z^T~mx#!xi>r||*5e|w5JCsXO2>xNN86GbW&vzVd@@uHU>0lp`v>IGLz%e}GLBjq!R
z<op!8oXRPx@88^Bi=d73aAKbLVq-1+#nm=Sav1SKtUmEKZ9+&?S+60S1QARFK&mC<
zdeU3x&56U*6$`qgbVJ@4-zDj+vw1k-CF}U<@qmB4{u)Vn4s4APZA6+d*G_yZc93YT
zpbVKjD<w$+<I*9ucUWZTaB>2Kb^O4&Wo~SZc>ek(Wo?<wHKWpSX??)8enCq~o?A(S
zXt<_reSbT;Z@v^Ewho3scw22PjS8==<Ei<Cyzhy<q$6|q#j(y$N|AP|9?^NmiMCrd
zV3XtAY=Xo0Vyv_z%h_JD0Py|;8Sg*f$HJJZ*iZ9^{uhAmrhZI|4Ni^~u90Bt9A)j$
z(wl(FI^W|T2wH^w1R$L^2Tss*7va2L(boQbe^jwu2o61Q7TwCZas3sj&+w^NN(Ob#
zUw`Cv{<~Ax_@8d9Lv{vHbZK>#IeYd1!%@Y~>#H~vX*XwU^EOs{rwaTulhhb-hyVZ}
z07*naRJt-glNlT>Z}L2F^s%Rp@VQrR^MxCmdlu(pFXJc9Eiv5P<oP$Yc;E3^FNMxC
zwhV1Qu{Ax`=_?Uy^`;y^998YnQDYpB&Lp%F#m;UKOrl?YYrO+jIY(7gsCGhW3{uE_
zwFWfDe|$FCQ?)(${=sdM`gzbv*b@#4fu3#Y(AP%j)AO9y%@n@rz%i;EsS+e1apEQ0
zCoN_&wWq<t1aKsko?+U`JCKPK99wMj?0Am9_}m|0jiGnuS%z=Cgh`(uX{CI6>j)p#
z&-1rWuJMP(7H?kf@akpD8y7F2ZH;glX{*EiYr8D2ER%Lhl=cvI8qQ&BU(YC&cj&Bh
zASK;a!YAH&jJMwDbM5{v0yx-BdE!u)+F1VbUd1=oE5^0sL#O8&lU4_S1{_80c;()Z
zOFLz=pl5ptPcCF+%9oOv+9z(L@*&^$&X~NFP*w(OE!H@UmW)D*H)|zq^a~$yT1Vy#
zX*NY&M`JR?{%I25(+P||#AGl;WQXbA0Ea`5gYEC6(kBnXAxh(XVnj*W!Zs*jX)G-r
z94YIl+yo6wwXZ=dLnP2l;lN8%?clp>JJ6X!l{JHVcQ6Qsr9p@>rnt}7j}Ljgt$A*g
z@Z0*^^apQ2&2;Ex^Njl?b~GZ(G6p*X2BN?SiEAWNAe8d8gtb@SQ8U=E1g8%5IJMkq
z5=K~uR+9G|&8S7fOKTn}{NXe6SZjSaXdT~J9q{%0g9mW^)uQIFHpcwGQj1f$NB6Kc
zFoVyt6agD(hF@6k3|eC-3qzV|OfZ*Pc)@eA0Q7r^c`Sc>7!&LnN)edA?fE_)TA&j-
z@$i%2{5gWX4nBaE-sTs^c}@>{#CNay89?CJEsW3g<8cHfeKacpsXT)m?X`I2Y9B2@
z%^ty(JIMA7$~x@QySaU1mS27C0{_)B>-@cW%WvHK61T4}G7wMD5+K_1P!^~(1t}2P
zSJn#U3t3%lD6J&bFsgm$ixUE!D4(vgmfD3*Atf=wA*{zje)RDrc<r8Ng3F!>oU?rX
z+BPq3`gHT-y^MD(x0uVkH-Bk;%s19Y{Mo%B@0}4GXq<m<)9c!zw890lIO7n?1?bv)
z{?<Bb<Le%@^sH*d)L6ScvK@N^n~xq~g8f~}avu}KZ>@94i2Uv04(E_T#YeR-fLzN4
zc2v^a+S<X+*q{}(L%8De%##yGjy*J!F*iTU`S*N~&gt{aEgz&_c^uR3Fz$BA^A_UZ
zqu9Ck@xQ#XjO;4@PR~)d4W=_s)t;kFGsMh1mCzyf_Zh{QK{O#GESxVPQ;DxSa-p7}
zbRIadO*}LS7ljo3@VNt=Uu+`;3@Xd>*LQenb3`r$Kk>*cA2~C}!Cp$1NOCQC%W{jK
zIXlPUO!L(p$By$6h&7J7tdUw_YTMK|>fASi`1%J$tT9kZk|rutC+&@@V)W|i^Po!j
ze&JxGDMCE3pA-U_1nj?wZ^O<lE#o9N4;fC?*Bk2?RhFI5L^6{r5-G^F2RPcP^nG+{
zZvkbN2>zq@KEmZ|*AV@6{^R$Z<t+#2c;AVGy!Fr`KlkhzK63sz($xs5`6rhuzFABD
z!P|zU*}q{}Y*37L*;%{InMYTcKQxO-{OGzY(_~>$g_g84KfbO)(rIb3M52_Wm1$~x
zXevWN$n?}QB_;1ay}*YaU1WV&@wIycT2k;+k1lfdK)VSGr^~`pE%||iKGeKcI+Ru<
zt%O!Lrz&c?GZ{FCP8C85(o8WPmKf`p?WUAv?Q0Uk;LS)ZkDp9Ag^jFEZ2W}4-=7J>
z5%ypL);VNlA7VtNSSO->u`RhaFf1*Vot!hOsn~qbsF;1nmO`-IFL7J9sMl{&)t0h0
zw6wyQnyR)`m7&#25%MG#Z@-J*ynUKGspXf>NmT1L!{QA_!<*FBF!VKLq?z`WL|T!h
z6I=l0iLTV7+PLXIjN|H*Y`SmbY{OQju|n{TTRtWBu7h2cdubD%MfDy>*hwKcHk&c8
z1nb7_xo>BDs+}<Z$c522COC!Bjyu3qrqP%Y2}Dc?y`PxzfAs4AU77&)9?JnTk$X8T
zZYJOWv2H#7Fx4IbRD!1MO0c)Tu8JY$zLSqi!<(C9obKR~E*EcZFc^-ww>zeEf}3l7
z?%v;|Dl5h&1>IpdI>@iSIKyk(hMztD29KQnF2nRDH*ejhEJ}2$sf5I7jg^w25grc+
zDnCzEV**?-4{4(O#M1!6q=^a>%tGmz5UlT3T-`00Pc-j7G1tJgHmrpoY7;PlDIMLn
zjv|E6Rp_-CmxkIo>e~9TxW?lN&N{}Wm-wo-p{aFBMyKd8E$xjt2w~;KJn-+-1hKIt
zZn8AC$pK8K6`i9QO=8_-K#`U{?=R(~)Oc9AkQ4CUZ&|x$vc~kLlwe%exP=wW(ot4N
zHM>r-SsDhl<;v!W)j}YW9(j9_B%2}Yt}v+H$*;Y>#PQ`ee{&^gVKL$6_C1E<5~&qw
z>c`j?zFpY+0inFyDM{r7rSlG>5t2$w23q>Dwi8I!3C}>=g|Fj#Y@y{lYgF9FGl7i=
zU@!;w1{Iseu_!z;S{p}Q8jzB*FgWY60c&X-Lm?&O(t705(R{{~Fzt;59;WgiM)B{X
z)rqFW|4vTVj8E5&$8=$2j+h=D*OuD(&@r}5w-eEfV{t&HMDyEv>RNUSgOrl`{wiYg
zJ{=`+#fa1zG8`*2Io(8oDJhB(wA%h!ifvRm;+J1tW5X%_)?<D0`7x!faY|7MMV@Qg
zt%M{^$TN*j6cqv^BsvM@X40F1ILS1cgoy4bJ(*%^r|oQb?}S9ENinIA5z_UgX2)~M
z;e=-3rPK;r)g-C%fRq4h3_^mb4b`|rD}kx}kc7H27~@O7{DhB5SyzOJ_oRRKF+rm}
z9^&v#zqO5K7icX!TIY#2p?(!iKpa4<g!czBB^r!cg|3D{<;Zm$wBY|c>)VcE$-Q-6
zBWw)|jvhb3XjE`}vyXAGS=9^-%=HKCj0#M<OIJ&3W3as*Cl?;4zct6!+9tm|y2O8Z
zCg=UdH*pK6(V1?V9E@-znSvVX3S^>D%21ZzoNd~=^fZy;;z$8Pg7o9$+L_1Jw~GoX
zeJ8VDJ&AKRAb<WIt+mYLnvWb&+`R86sXOP}+EvRtc|wLEn9_%RDp3e0DNBP?K1`I-
zx4y+}Y&4^i35*<_^$769p_Sejh!NM6AF&@j+-q6T+uNHs2O~M{u{+9gOflT?kZ)xW
zsh;G}!vU=n?2c;^DJZ13V_D*FrnY`f3*(tX0Gm6zT)cLj^C#xm?vHrs+FcfNO*z_P
zUTaowZ=j^$#9YF-?yx&9dGt`5$Br+tJwD3I7jIIGPVwt+uJU(Iyu@rez*5mlEq6Bt
z2;-}1atQ)MDr$s;(t3}35G1UtC6)D~?_C&sYn-7p6=iMF!f|RT<76-6!sdu)HU}J<
z$*|61f-`4r7>Vz~EJ7Q)g|)z<J#eNV&lARj0<9&scENN)ua_|%*Nk?@%+9qL7d2U;
z(86G?X{hF$&%;b8L*r>rm!mm1{leEfOmFdbQ5X@S(dCOTeACqyrbprd+<x;SGY8N3
z8bRk%y&f8c(io2cSQ^Q<4l*qmmKLSFlh`jTb1fBOT3BDb!otdF%1}R$YSD16mL5N_
zHVi~+G@oEl8fxR{<tpSC{JB;yf0OaW&+|vyf?n%KNvuT+$@ZO#^ty*UUOyO+_ZCSd
zBsxJ0$6#}ZZm);JBYlIdZO*K;Se#p?G?uFJp8AcoF@Jb@hl9E1!zVk;<PsU7RR>jV
zNu<w*BvMe12cX;h{xy##e(ZRQOolvyo%H4jJQ~=fdt8i|9JVS*(%chiYX7~Ir0MjD
z&r~4b2{?zXE2K_nrb03KO$4N5uqnEI_z7BC2K3K1d4c=u8zcwLF`Fxfp>wE}NXE6}
zP$sFZV|Q%ng*w5Q91)cdZIl$G7P`4Y;t|8qV3*k=PcjTeqH%QFw1b%?lG@l%T;g*R
zbK!V#7EV!FTB+jx{j0pLcloJ>lwY~@8g?6&mLH?s8j<hzNo7u|I^-nCszOUknx)ur
zhs|M0nnK?0QV!cxA|an$q8e3Xxrc2}wJaYf6<@ls$!9M1IXl-110x)lH^$6mijO|F
zK$3vOCwBh$`k33L<@{Vux80r)(&4+DT<ix7v0p5%352PfMe2<BpbQuGkD+4Zm?+T_
zsI5f@yYGV5P&y4mBEq0d;T=;!tP%9gp{v?~^W|RS%F#;wB+pa{4s;WQ@a16b)DKY-
zVluU|oyw+iwlbD1kyz&epm;H<1R+9MRVZNWgkzP)+e$5bmrYqa`eTC%6|>{o(oGeK
z_PGfiDwea5x-P1W-?{oIKYw(apE>Xff7WiZv1ZBBIc#Sci>5o<VbmXyW`3%O5t6i%
zV@3re36+z~A6i7~!Q?pyVetj<-gC<=c2d4@b%R&d_cm<IXNtmjCwN>s{^&-@ozilC
zzQuFLds8aeA56xin~76J95%qx(MO7Xh7)j((0CHejS$$dxbZvXQ<nRDe}N3+ZsKJK
z>6=($RkIMFl@x4`OjGV%)(#_l;wH|))lyIy-)xX6UpN{8jGeJzR9hC>nyrz~1(YVt
z0TkYRl$DowM`*DQ4$kJ#u+Y(rO5gMAr(#QlgEWzokW&aU1!Lo-rKJV;2Q}y%D^BxI
zUVELt|D@wTnHlq|gB#T9{r-_D7MNY28r+4dKn5RXVdVhpH}2rZC0beX!wVP&>{PzK
zKsY}zb2P5WwBV`Z^E`fZmaBJnS=}1ptmW8Tiv!&h;e4;czkYq4JEi66g%&?}Vz#l*
zQhM-Ai%Efp47C#ofpOk^$e`VYqyeUIli32!g%P!}UVw?2l!4&;+tXtIQm`)mv*^Hy
z80G_W?O>;+z)x-wBofAz4dGyT>V%yXDp=<a)=niYEh%e1J2*y?T8L(LO(*r>P%I6z
z&TEWzDk*A5q6LW-^uov#6$&v*W7!y+rjKo7P%|tHqsq}rBsfE5Q?#1pf4<aZ)e8RW
zvm-j4ud=hXN--MH@Anyxh7{!(BQ+*T*x2Y}aiA5t*I{#KgcZ<fc`zx_@j(}Ug`Ds;
zghv+JJacN1r%x<!a4sj&l2PsW-B;JRGp=}QM)RT5^9?qD@ZqDBg0I}#r3?|338AAO
zf8*;YOazsf3dh9+{uOenqG)0Xh_to)QJ|@{j_;d@GQf-ukV4W(S{feRA_$p3?JHwP
z)ughgs4b=OpVdLq6D>k=$j8M>PKZBoBE8m#<zXI86wL#g;(mD|ng=8}2c5*1qK|7A
zTG$jKWH3`{(V`|Jp)lw9=eK6L>uP>%W<)x74<m2kir1-JNu?zw%pwpNi~yNR>bk;~
zBczff+CTPY?EJ_Yl%jA@Bf_Y4Nojq!CZ&HA&Sx*Kb9+?s)IysNJd%ZBjuXmBDfo-)
zJA7?@#9zGG4`c5pa|nFGY*ObRiHexCs7Jw;;jl3WF+K39VU&G1L2Pet!oR7oP!OPC
zDPnzp=oSfy9TS|0M8-x)7?j@0OO*HcGcD07)X`6IUTGY|()wY9Q(S9;CaA5?I7a4k
z;k;7#$YNOe!c-F+L9a<v2wy_=ZEYCz?=#q4g+otJRqx^-Upj}Far}JuBF{ekEv~Ko
z8N=aC%IYSiDKJ>Jc864^_UgBlv376Wk6{bbOR`jvB-)R)4G+hP<=j#$?}G@zt*tR{
z^v4`*rF{6Yr2r*Nz#;_v+iTmryftPe(|qXoY$Jh|$-^o{q-=N{E03$d3DecdE~@QD
z3ry7hWVfmHZg@dVG1i0$WYiABAlVUt8dd&LZea}wYfmec%`k~Pkqk<Uv)&YxVYpOf
ze5~BdRg)49lUo}^F$oTl4S-X05UbF{*3#bLOVng?N#hs=2Fvmm-T7lAD?fnV9k5+W
zhUGE->4kYRS#kb)K_#E2bds@|XIwcdp=cjCK;;C(TC+1OSv++L3m737887iV@lbLr
z_l%Hg#Y{W#j2F;B6T+<dT!lhY&zMd~YUlX;)h)iYIpT1x`S@cCB--a0whH4NHvyJd
z5*{fLq+Ny)uwru4-ff7cWg6U>Fo_6a9@EA#0%(rz_L>D5t9x5|Z{_MtuE?fBH6OQ2
zMup}2M#;6cF<YaWUM78!rtwTLtb8<(1UNX6{wAZ-l1Rb!sHR^`e1=3orhI-~OYb!H
z$6j+d>+yuu?aKEq7{}Vc&`BiKXa{F1cDL?x^3jtVI(3%E-*TR`-RDzgi3{>D|KO>M
zurMZ@ljH}NY0u4Y`;Dt)i}Nfkby-?j;_i)$%ye7ixn{AKQd+pZTj8YS>UzPQp<!)o
zxU*ZKl%G_Rr;2kcJx<Kzys=ZTwp)4v2V(xk)h%A$7;z}q{M6%1B%w6!_MqaoFKzN~
zukTQY+Q2ekA0`M&RHgM4!DTQfjpG+z2Z3wq6XW+HB8HP_A~HyDL40juVSM+`^yct&
ze^U4h%M;bau6ZJ9XTErGb5u78BoR*9P9(<q%3X&CZF~^yk6l?uH&Ynv+p`lD#?uFB
zPn2L>JJLk5HLQ`sYm%joW~VSH>4!h=mX^Y$7?IM>4Y$^}81?&%s*-x=5w5IlbGw(a
zyioDWk6q_q+%{afo3fktP+GIOzE0igBDcrny=4Yvg%y&1ZBU7#Ed-Ud9GGvB3P)9%
zP&)vXwT<+>@AML%`SvQGe)TSAdx}oJ$;I^{8&%EWR>H^6Ej7Y$uV3+dVRBhZOTKk)
z$WsSezO3GP+ZMBpu?E9OViS+vH6!kW<l{}(q;p{67r3aA8fbVjQ9Uw&jOrkI`u>OE
z$(+VJTjR=5TlNs!ImdiUQ5whA$Y88vdt{m_+BlI#1)JS~5+Mf6v>#nkTh};2&UvOt
z!vNb}uKZu7uy_WkjrSc!g#oFsNlKb^7>)~6ZLoET6?14c$JWjYf3Ug4u1t7~x=gBH
zXLqzkMM*`)*c85WDbI0Agn?APlBm<th)_(LB#Kt9$ub>!3MIAk8N!1LZT`l4j`OY~
zv)mh7UcSE@h@bMwCzt8u+RsoJRQ#*U8%<5&*VjjU>Hd&^b!B@`amoHU#v(YCf@WZO
z<0H5r(lM;wvqe7G7afSWb69gdDh<YZi5^#$#f}VvqP)7@E&L>pab-i8C}^b;YhhH`
z;IaE201tH8rc+iKUz^v~z69Di7}ee^ct{zvoIK&90&g<H-NeWNJlK&a-%70n>{v*&
zMr199${@>{VSj^mZ;`Zh3O6p$b<Mhbiog5(=lO5nYx&vyI=_AeM(!wPe3N?Ep`JTU
z)|zElRgA6A9!lxQ=2#ctYUyVfcF>H9x?$Kd;0p7dlxI#a(v@TW<c$$4?UYZx^$;!X
zAD(q*r)nmbz2`uOcdm3PD#KT9_PMZAaA9rCTMo1WHb7tohcOd%9-X<ErjOyI*Sy|A
z@Xr+Ri&uy5>=cCPteXtLh#DmJq$x>g4hfAO_>gSAt^7z)C7Q+$XD8r<wT^LR>1GNk
zy=0fhp`@ghc<!w1<S`}Q>z9qjiH@KzzxGTQz0}BrweeF!8weH8U`_}+SxVmaY2!|J
zfl6tt&ag?2nwe*Qc8+BJY5w8Y53_FIZ$EmSrRA47diDZix5a2U#3UIyQOLv>lK2_}
zpHfyyoL1tCNTd$)3#G3#BK*%1#pwen&mNiMljm2+Q-zX})qctEes|5!EIi!f1E=Q5
z63JXU;k`$DeDv51Z#~%FBlR(+n@lT-Qt#i#0u(8lwuq@zu?TDD{RSmOvz~~_BXpY)
zF<BzXv=94ycd+MvJq0`@+S}${%g^Z>)s~|6L#5_girL)z2s63Q3-og3?Ou@LRE1{?
zhCabz>23Z@?tOt&Pr}Toi4y7S2f!C+j%$l`9;w_M)^sw3s|q%Dhtz9poIG}jxs}5_
z`p99-?ttFX5^0)n{o)0*Qs~Yp{`sY2*p}kIdBl>~gSdPa+3m2gc8_*DXQrFd>7*Db
zxjQt}2sR7L$O<YU7#JS`I4Q|8<?9Z#FEmB^ku}dfe!!aztytYI_?=hpF*cSD9+}~p
z6LTKflO9i)$u*BHwb&RNE^QQz?T$q9k7W)5D%e94DI#Db8dESa8Iv@(3z8mT{Foz*
z8bmt_LV78!jia525W9QOCq@P)bOCizNgTJDXkRyAfx7n1BdHctwPSl!vpka`gr7bV
z6C2L?r;|t_C<4GV2;*A3r|d|C&op*2|KyTP`<tnQ4>8lyRK}rog0zltvCE}9Ye*ru
z3%IIc_uhSMT_6bc_v1c`a~phX&M~(&<P)uuU(r|C8Xlo7pfnk!6Bq<pp3q7q#ki)Y
z71CKu?LF#I;b>*TSNaO;%gjVTN(E@GC@aIg?SkL=&K*zuqqDsG#2m(0KU>lGw7E5w
zZ`|DFi+6l;YJRKFyXF+n9GyJNNKS=yvEwIV`EH7sQE^0*@1}NW`~z>zn&%G+0eW?-
zppz+j;aO$dqZ%Rn#E*kYP+6Z7XlW1I+S-z-Nwetv-I{KuJeU(ywr?2qB~VU=x`5i(
zzSq{Hdya`Zx7MS4{leo1u>x7(0jF4Et^<A^oysiLoLHKnT<mgtV;4K<qgx#gojS@!
ze~hHV+s_=~3$MLNPuFa3wD_YhN<P;;!p|LjiI2_p`OL-zmR6o+TpNZ~V3a~*{nSz^
zNN@;aybm5$vaAegB7D(__6%E<exj#CkZ8qhCuKg<yl1JyyHCzf^-a2_Ye*`|QYYn9
zCuO;%{gcEt3i`Is4?fZhGYKXSw=yAZPiVvQkFK4}Hr_u9))M`INiCodG`@xa^r-{6
z&%j9^7f$!C`wSJh@Ie5;{Ea%F;$7+{R3S+dGkv9TbaRCX(SQ#R1I#QY!wZVqYZ9gW
zIkgNEyRCQ5?UXu<&o;8o1+C<ZGS<f>sjDcG4Be5qR?g1W76mCX&A5JR4OLYLnIhWf
zshvR>!>_;oZr+u@$S-uh!@n3FMkXhj>7;BLpP{R9?ADNHUbP!n`+-%)VX)Ldl6a^X
zY?xsfs*H`Lm23XmJCD)Q(mPEKu$};ek>SSXh}leWd?Dx5LgsgIj>R|D`0~b>TnV0C
z=>%fCrZ&^N2l2s`UiN*uI0T1oUx-M2Os{Kl10o2J??l4e?&jeGfw-9#e)>l^@Srfv
zwlo`qnnF9ahBcLOoH>v;oxiE{IDX_*V<AZ=jU^H8UE8f`XNpdy{FquvKnJ}k@((N*
zO{LAw*k=hd?XRr}d4akdQV+M-*{%54``^i(tv+dz@Rwh|!orag<czp`W0jQ?k1-w(
zm{~kPJsz@p>pJt@qf|#~ULV2FsBeQ_rqxQ2z(&7dU@YBs%AMVU#5sz>km3toyP5Xg
zM9z`|gOR1B{Wx10YAck0nRbFP7U6u_Tx(d}9Pz2|++`dPw$t5==T6Mh%luQqo?cE^
z9Sgp;KH|wG-+>ezd9A>P@$b<Ld-T+UW;);njWg*cS|HNZMa>YLbA98|iyl>Suv65`
zxBL{UjkQ&F%{P79Sdh+}#?eXq48cS=oN-)TpPVlY-}Y!=C<fdX-q}zR`s+90iXYBY
z3fLBtL<dYvg$My`Y*_;s;R_mkce2k+js{yiGhd<)<;bmz%#X%slklOlvsC3>vMi(f
z^dYwTcd?nFF77h3G>bZPl#P39n4K90y@bBpWNv(qTQ|OqtsJg)WI}L%STmc1p^~M+
z)D~A6w!(icpa?MD2uD5GL8*lQx2-pgwKY5MI)7`eciQur?>YC@y|?PtJh(iN#BuDz
z0Xhk>qeKY-0ufEpN_VRjq$cpGTXYN2bgL1xkN|Z@K_C$b0YWo4gba2P+aZZ#yIihv
zmCH5Xy7Rqf-1B~iwN`(4)_V7`=_6THt~2ek-!(kX|M?HNU_leE&nTF1&hf8bIb&rV
z@88S-&l{7PU;6qXAKS}#W+%foD+a;P;_i4xF*NIk=V=Kf>-gHW4xtWDy{gMLC^hy4
z>-TwmO(>;FMh8!01NPsuo<T3Av5xM^Df`!+V_8{7g$x4jPb#`a%1%EKl+Suj=QXu;
z><n@y%Z6dcu&5)ts3{u}JIqT<v6`c|9%p};VV!UZ`njR8PGVcEXP`7iU>VDCCf9;T
zPSWVlUs+GRn(^2E+wUMIu&o-40Y=@FMdi?1<ZQJS<BPJk<f);bY0j3mg;IN&;pKn-
zw|VD3{{p{w|NVU6p-1`d4?WN4-@eJu{P9Oo*0NYu+&EnDwHrrNlR3>wjH)|{=JMW<
zlkt*KqqulwckO2wd4o=bsIT6du(Xb6HamRcxjn4&d~8+m>#v{inR_M2mF6H%czxDT
zXvMXiA%XjXb)H+}l0Q0J@leO`f!%^!D}wcaXbZzbHr9g4&(Zp2zLTtxA^AEK2}wpj
z)ntLCqJVx!;Z%Z7a|WqrT3Y%=!sT7@uWOXNu@#QdAfs#?jaCf%nGAc3!-so4H!?uk
z8Yu3crVM&rN)%bD=qjn4EuJJ%G*0FNRrsUg!&Z9T%T|HbgaM~hW{nA4LeHWO$mK#|
zoy--IG~w>N#wbl`6*Y>}l_xc(oq~Fm@LG8VW3Tg<5B@X%ael-<{!EUa&-phmj(Gm^
zkoWG+dFT2etJ5)lRgoG^f7s(}x@55$3yGxgkfwNL#aOzrcpMH1#&~#dK<CXP7S^MT
z!6;9WXZ-l{`~1=MGro9l%8Ow)f6s+JNn&^~ule<N$2_;0^Tp#O3i$ZN9^EWy2>~9D
zS;wEqd!HC$c%*Da)lvi_!PyvvURJ@#F(7On*0_AAz&e=`6;LQ5ibPX8LG<d7Iw}MX
zA$N2$=?c=&7WUE@St2yDZur5G+MFc{7kr2TkUDS0%Eq-+avKX7coP5sAOJ~3K~&sF
zqM>tS)ChgewNNs^gK0%=U8q5X)VZh}rwhyJLcI2-E9r1{`YB~lKQC*CHj1Jc@Eh|z
z=B4Mk@d-<RoyquJTsh-tvXYk1)imCVC?rd%s#*vmTFjo%K?|U@9LI_178?XV=yei4
zus`JbWW^u6bJ`M3v{76bMlJz-@4<j?dq}MAM}?tC#OhuFfBKOT`@OX7?^W#mRRD|w
z7^Jldq#V^2*D8EX8$#tFIT%V%kx~j3?%ep02f3lPVkE7APNKOxu6XOPCQB3t!;GzN
zLS+R$?xvbqDNLW%pFXs160js&`)tt&uvEw_K`YS7nr<Gsf)VAqbykrF#!wP)gF#_<
z5Lil!y5aG?ZhKX{cf8`_DCca^@Vl>{^3g{}^mD_(C}X;^ynD3d$^9;6BQ5Loqb2Wq
zd>41=XZb&W`^)^)=I8iZo1f?Js0F|O%IB%>TqE;ca+9)o`FS??wt4tq#K9=%#aC`K
zKRKh@?~&vgwU;)qa$uBPJ)O+3ti%HMPd>NLs<C|b#wpiNmmKsn7PaN2BjFo<bT{L}
zSB5Q0I;|Y1)!G?M#i?Mp-+$-E4Cg$5;&LB#o@1a8W|Cib3bKg=r8P=M^T{go?Tr(-
z)hqd?d1aYZg2+VxVb#by;Hmvii~pD5NKUj6M6CBrR+c;wQesT7Y&mS+i+VWG%xkHh
z8yC_%!;B!f(MfCRJzL#`Jdr73B%x|0<LTphO_6HuPiyi-aT*raymBmR%Lg9nb267S
zv`BItbyLd5vnVa6jpM=~<M-~(_`RFA(94ql*wkDZSblad=WmRkr2E*nla>qIc+7Ns
zM!i~6E|%OYQfAXRN-5@(Ig4b0G8$_^r%*Rmj3B&>$Z;O;9e?Wm7kTb@!OJ(#xISKS
zHw0#X??WTrx7mo~9n0Nu$-jB?1O@!?l@VWhFz1B_3x43?AzCTkJX!JLxa4A<i2lNy
zuc`x35Vn;ISVDAji7MyBxv=}xxr$K<1-zfiOgi;K3h5}#VO{Z!yCuVJ$|yGsI;O?!
zlb{;PQq__mztzdoXvJ)0DLk#sd7{@XibBSpX`(3`hj&VRf)~<;M#5xmJ$2*Q>Lsii
z&)y&<OBADS!o9PSZl;-+VrP40yTeAu&<{{`G`$=YhE0?3U|Mr^r^62)Jj!QM!<%n^
ziGQ)2^6}*^fA#XK{POq#ckU_7a=>uwX-4~d9BdD`yw&9!Z{A~mbV4!gl5}dBDHtVT
z7e-N44T+bTe`0*-$2Gef9X_$y;Xi!kHg^{l-}BIj?|5XB#pGBD?pe*hdE*#`Sl#dS
z#p=EYR`;jNhR3%GexPw&9c3+DJt91E_=YnzU;;5`*@u(0ah@c$THpMbWT>OHYU9?T
zdstaJKK0T)p1C;U+piAU>L>A(|E9JLgbDktKR433l!}fd>!U5Ml`7y5nNg?^YTeC*
ziWQ^Vitze+nPH>5o@3-j(aRHtxhB;)N3)tow$4erQ%!CZ-QZ**z>*ASu3i&DXLIsF
zUVjyC?mW!jz5HeVr)<XG8BBQhg(tc4)RR1MWrwGBb8-dK+H=!6_V>5g(VjHb5Lm!X
z<H%Bja*o7%R5-m7<;Ve?<0Fr5vuG?IePWlUu{hv<7$7O7__I&$us=xg-qXnp-g$oK
zy)mziOaAn=4W8TUiXm6H7NST572QPQ#TrOqp$>>>s{a2A8;GkTR@`+Ufr6l4yQ2>C
zvZ0KgH{n<jN?H3vC1i<j=B<Go-;>aW#b6_dMXl5VwG&Wk)d&g2glj6YlD(JfU{VSf
zA*z#VD?n0dMW*dN94#!1+RG5DcJzyc7^nbpwXLjYQATGpO<fQ0b)Wz1#(j=Pp1-!a
z%SXTK9&f+<HNJ84m>1rgFsVI9XLHU@Pnj)C#;aO53R(!L7%{OlW}T3xf*{o*{QgVZ
zT|W8z#Wo=GXkq!6uRg#7KmODX`x|*X9YiZW^X{0hoUYi6mc60qJ5Lw;FLqs=BF1HK
z4l>r~AK>}N6yMCU32*MX&pmaa$KEg(e8b5bScQ{fQCaa93b3v7zRj+SAJz+#I9BlX
z+$RlE6sB5;p-RaFP_Z6E&r3m^j8>uAx0Fp#)-;aOS=}z$vT-f5rV|RyMJ+Rl#=*Rn
ztE6rmw~m*5_1$Am=L;&+qpsk+ukZ3Vzfxj`89$a-(#{*abL$@0@1Ah`bj4fm++Z;u
z^WbzYn1@;W5JaI#%uJTnSheb8hiXYDC>cb3>YXJ^>-hd_TkLOi<oFKN@MqpW<Hge@
z7mI`+dtyt3IIM{xz&Rsp&P@o{U*N;$w4r@W7PX_BiiTeK7_=;tz-8^It><K3b1=%$
z2n+c6i+wzBIICq6Xgw#3hCyy916@1`)-&DQNUL@3co(kWT4fv)tc8N^DN@C}_ANOi
zdh07KVv@`#rllO0lSRX-c06&RLs>f<GSJ%SrZPD!4AYgEK(8t}NfxDmQ$gg5(ux~L
z3x4~J8@zt!5Wj*>Z--^MWVrJ^%vHi)dEpQF%U4T&v~z=he)|u2WAX$ivy%C}x6#cU
zpLAKM1<pveL5j>htnvZ~g%-sbsB9yPLkX!w2bYqw*@~6*Vs*bi057n*R-RA4b;^s!
z3l55uAA4#$Q1Csqb5x;)_DWVmOcr(F3W&)@)~{y*9wFHzxlye>Y?N8Yd)7G-U<eK%
zlfhBf&<!}iuwz(Mj`31Rnu%7_T82B`dy-%(6$d?yTkHL@fc({Q!gX%iHCc)Hw06Qq
z&H``H`@mHQ96PPV&+xIm4%3xoR$5Z6xUwxgyJhWobg#?3aYdfW8@{L<+r6aiSTr(T
zZW>Ej)vP9ST>k<|w+pwg^VoAA<dI98{D&_;#Q*x@AMxK`evO~Le4W4d_7Htkpt@Uh
zvrAOk;I%M`lQf~K8ZxaZa#)rGoIbF0wIS6WrxnYUoJ<xt*y!;7?TnX>D?b0$F&}+w
z8?E6tzi}iFzmxIf&uo)v#mjeRymGYQ?n;a-u60wM+wSoArnrTo6{;1W$|Khyi6!y`
zQ>{n@GL<d3YJxp{sNO5<nN<z*(z2=(W@UqQaCIvu)v^nt9m=Zl98N0^HgYLGlwwxN
zgG`h416$UfZWs{FSC+&mI$DeNzzaQ0oId5Gi1hI!44R^9Nm^qm98c#>>sS)$(sAG+
z6U&5UEx?cuCzerAI#1(0qttLRS@Qbz+a%txv$e@n5AE{V#f*<Xdx?vqocpH-yflA;
ze|me#C-)!lS3mGFKmU#EP+GG5y8$7JbUG>JO35N^EZQkD56a1-XgqXsP{AK4Nev2*
zHsGD(k3Zf+Cpq8w*bZ95r(b=*OGk4KIvGFm%nn&<_{<x}ycCj8_KSqg)Np^*@Tr><
zK3;ZtcIeyFRUq_m0yr<hoL>HZ1tfW@TkfA6&xTdyg8R3$4VjUAfKFfs$KyIn6qS`q
zJWCXv)KFT_FgJlWr`ztKa_ep)_VL}!$kZ^LHnCPT8VVyy<T$@h0y5Z1<?HJ(KHcaf
zbTbnaOO8Qi+92r62u-aD^y>!Cph()bGyd&bE3(8e*xJNe%Y|OS<?TM*L7!nJbh(c_
zyhSyAn6JO1`P74#`04E_f9c>BznnZy(%nFlU|EE|-=nM?$*!TT{&ga+V^xX1(OKaW
zXne>WkPhhwKX3_e9l!DNU0ynxbFnLMu~G1Q@0{?`$%4%^;lH{%<jO`aou2o6=AAP>
z_h7+6o^i3O)*@7+Xo4-h-0-nr0DTgu*F_?{{LE;2xnWjX7NG(UbD94lg|e~2Bb=-(
z(^8JqBGH`8Yk3pY`hmt#ZWbVGxsRj}Ypf?#s-5>o>X-B3b6PrzJSdo*_mojN<UGsT
zax|;K3w>-E=ws8B#X4A&qA!e0rN(*YwPPAww`G@D8V`%blGl$XymI@L)ojkqNyWHy
zynVdn^}B~u(_>t%_=Q*Ita{u0*L(L_lrM99d>`i<O0Pj1lFU#kgEt0eOq*+<q5|1y
zFsAh{j0It&fq5l~EBoDyKlh<abn}E;XC+^}Kcg_3pM3uwmp7$IQ~|*I_KST0zCM=I
zMQasMSX$smW>c|SQzRRmq-ESFCB<5%Y0<aH{+*SU%Nse9l|>ef;{LSa=?h&R+RQkd
z))YpIOgv2B=9OiCn6-tT4wdeF)u2%H3*iG4sT`RjW2i&>X}zF_G0@6}W2;CN%i1&Q
z80N?|bULqj=3-x{SzbC3$tuu1nAQw)85r#jQj}6mRzhFzW>VkWpVnO3-{9lD3%qbc
zadPiIU%TnCCgC$*ev?6w@$S3VSv`0SWfD4DkD$vA|Lnm&|IOB?`SHE4@Cys5-8Mxh
z!Bv1&<hjSzkSS;?z<6x!=@f>l6nC&rp0HXqI4#N|AG(CfJ3jR2E+2Y$tDTs>dQXy1
ze&CTUI$;=8)mCKvNsPpU!&OW}5vpFI#0#jgQY7opJ=HEw2*Z)y3AS$@*)>VyNi$)V
zZuZ0)wUcWyEpjbSMs33bb8@847nWYGvDVYg<a-+D*CEGVt`Zyi_^3q$>HTBTsKb$8
zH3AWX$gFECf9$N17SCwK{wSvovbM7E!LUMpZ?7-`9@cb&tSvKA0edfXL1P{FP8Lj*
z-~ld8sq2P(HeqmKi~gX;g)3J%@ik}jDU~WnyH{z79sa@V5A)-XyvaX!`c2Lr>7k6F
zR*uAgwMHOeCl5d3@nvJhrp^kx*%_&UloO|MlWKY-)Qz}{c_no2(?F8Fy3>`N?d4AI
z1<+?vt@~-OgvS{TQC&pqRG^BhkR8039Lsg!*D8|KC}ztAW^zng7@Adsa|sI-qT#f7
z0%gu~cPu7K-a78D6nM|sT(o_O(NwiSxv5duYC$zSp-2<76Q-|GqRERA@GOi1BcN2<
zAO#Oaod)U}wDl;vpq?G0jApjdtQyO*a%^@}$vd!)rZL!MgGn_@E5j|Ng`txf`8@-p
zs1{SEtGgtNGj<BasGo2&n{sta^Td9_YbN2<>3||>aJmAeQD?VFlMB4H%K72#cbUF>
z7yH2y+f+E)psW~EdDoCAPvaVtw<u>Z9>7sI4Vi|zvS{zn%2L-A-iq1Ova;wzVrcRJ
z+F4s+(ZZMTQbTy>`Qj}(As+23ylbQb@NlxUjB@##<0ixl)jz4Hhb><3Lm^UokFiU3
z)EK+;Y#C6c+$Ad~ty7{s*2_%4uH=|AVVI*87%OAPGLR4|;B-1=dixH){=!X4>)9K1
zSp;ORo2NMMxqmd{`1ni&mUT^9<dP!oEvsO$?rp`^-UY68j`_atz62JWg-xv(rcjrT
zj#hL5U&2UviK6vl{!lwlFVQr$Ck;i>w_WS=Ll5m!*$aYHmX=<23GYR`@*M{Q)E|4Q
z^*)UJNbACc&1lRw{*X8S)z9%OXG8q_5wruUJM=f6MXkmpDnVK3bbHvc#Arj7X85X-
z2dOkyHa*umLu?}ik6ekxJkE1<lyH4o^5%m%@7pid>>%g({r47p^>oESZur2i#<>Q(
z=d(ALyfJO~iATB&QbodA8s;1%8<(CE%wDOMgz16flVgULo@|xCux6L&cCpAOOY6hb
zP_-)G$YnQLwUV+sN5{Ng9&pdsEbuJpas716)!hy1(o#2$MVWG=)HF?lHwoQ3XTO^<
zts4&A9MjYo<xyuH_Ol)vm!53FrN)L7uga2zDPtOF?Gb-)J`hgB@fhv#=96Q#E<Z(C
ziw*85kW!<3-g!X~BcxdbT%im~C+or9ewBB!l0UWafPbMT9KSOl%M^pLA#Dbvu0eY!
znv^2zp;i@orAd<l-B^q<SY6ZK40C}d5F)&cNI(AEh?~>*_`O?mDxa{?Nmx~u*N@i8
zC*S|*7Dg}0Iz2x9<|!|YYc3WE$CY8Dm&@~u-4>ptY&<5D!N|FJlE8$qYIKe)4(CHf
z=>wK-0|vjW<!)`fZ`&-Tb&DxDC)NBYPq?(dNuDK~%`5gd`rJ95u+b|>Q$to1jCwUk
zixRwNRoARk2a_nWqQKW0m87B!LbVXI_fgOqjVEMr8l^Z{)Qmb36VnNyP0o9?Q%r;Z
z5lAb#tUaZ5t;jXagD%3$Y@!LfE^Y*okcp;ZvCR1SmoD<RKmLHf?B@KfEBD~+W2~AQ
z)g6V}MizIiQ<6NvR~1Q`LmF5@d4PVMM@h^v-pX~M(ER9!F7fLx|Ch-p*Fy41C)F&c
z%N8~UvDW$2%?ZVA#fNt~eA|9ct^=h=vP4)o(J4%^?%5$ANnxj-=b5Xp5Na=^3=<f=
zSt3+L$6CM;i;p}+$$a?Q7WXC<?>(4sX|op`z#M}-L3>9r?65oR@y>%&t_})5_~bs?
zd%pbkA+xgN;`Wf2Zyrh*bTFuRFssRomL!u@ge6tu$O}WF1%vPBCY%5wJ5P<gt<l*;
zOsE{gz*bf&vXP_%>1CQ_Ep8>&fi`PrHSZxwGcvaLyI&de$qT3a&7C3tk1n`%drr~s
zL6VVndiceX-OGDavjyYZhwSeTsg@Pnn>~8jvACNi2wlz+C%RbY=y$R<`Q*mY9Pd3B
z29kUdLn%Icv*e|-nu|rs_dPsfua|OfTJqcPP59h{1;Z@iiLG4ll{8R>&VR)J79Vn)
zn(44OT)?QinQCohs!*&gDj`1@<y(Pb91AN1oBz~TZZTbyeE6{oeEHoY_J$ogc}g!c
zOjE_H$72p>OGcfXcgJ(?eCZpwx~3^BOk+8|d5g!cT?uh9o>}FExQfggTm*@_2}iSr
zA{7s#EZ1UJ8~6^jl|pb8V)N>tMgqYf@<cOT*_N^v-Mo64oRun^Jkd_Kt`^w(jKk@Y
zPj6h{l5%{c`+5HD#&<EVUZSZESe|-{bj0;n@1vYzbLRrLuHPV4hS})}g9A&N8tOV8
z)1mNd1V~cOp}gYpg8`52^}@VB9=z7@>#rTl!|!DLnP+z>!pvZQnDfa;H~2?iKjMYM
z1y5`jcp!;_3$0=}gjC~|m10Q|y=8-CtX8B-1+AWMH35lI;RH~vh(l~<*6q)#cC4Ib
zBhiwhW1-&_oZx7>M31_7Cpmw{7%DV-!#<nZQ#Unl%qnbMk-3Ix+0f+fo07BV!Lr7M
zqkmO9iog>nGO>8cf=h?iL3hyZE^$Mdm-5<M6g0S_do3+#9Y*Alu<F7h7k{OMCVv=f
zFZ1rLoWJF7@n;`d@{_}_@P%H7x8IzSxdTi##N;^{3Cr4&W+@(xGKeirxB`5*+Y|<a
z6uK_%MCXVq8l~lG`L(az<He&n`-S1pyl<EMocpKJu+d4`N)30yC_guvB;L`4dr|Ug
zT+2yIwZeEg(qn{)R<ykT^J0TKB#fm-jF8%b(Z|{W4(B!B^Xxu{(~2*=cS_<t^Qs|D
z6W)6;!F$K$tpRCbI9b+gcQQWw*dC3w%wD}iq6ZxGQ~uz+Lz=u2r&8~!te}~V_iej`
zS9B9W59`LaT~2fw$&4)YQK4IuEEbMgqB&bO?DUeBDMeV^W#cF^EvH1Psl&psI-@(-
zrS=JRIfv#Uer;m-xxJV9u9@bYuEN&?ND@XDFOxPEM>p<r_0cO-(>c2vU3$ZbpoU>{
zS-@e@%CypO!Adk*Ih@iM=lHdk?(yP-85g@5Kk`^XH&4Y6*~>lBc=+<|DJQk%k!~uM
z`FN7Z06rg2k{K;h@o@aFYEMU!PN*GWj=!$_t@CK(TRB%Wbqw#BJd3aZ4hAW|__=Q|
zUzU8=(^q)&aLU!4A(i)B9B#0fFZt5D4>+F9*cy#EoR@s+_g|r^OUAk&U(7hh@yx{o
z($0)oC9rRi$}yb;5=k%B!fOoLgMOh~lZr*@7<S~I&y9p;5*GKMFicmLrm=K06Fdx~
zxFvWNh9OlF5Vgt*ojs$XSsdeZhh%G?d^RRazQX_U<`sVW;W<CLHRDVECGOpSlKQRp
zFmA}!g)7{C_bs&bOb#1HR|}pq8rR5);9c+-v<{PqXMpi?707~?lX_UBT<m82xer~0
z<*A^3XfdwVnip=K@wt053a$9AhlXnv@p&#_tOO#~zN+QTO-zV`gEk5_(TYVK2Rv)7
zSse<b=rZo&$pS1Z$Ad{l$5}F^dFyDw(PYliykfR$n67HBpU#Q?x6V0=JY#cvi{0&A
z@}i*X4nS#c&Q`d(X^-wi3-oCN&b<nYuwRJNXe5k8s(Rh{*5)o2zm1blAWt+!rb)FB
zSF8_t$n7<_maf-UzFWExLrp>Bgj-Rk1@prV5`CF}dQ;KctN9ayDXaW-nruOBC!C(1
zVY8mFr;<J-l7-~7a5ONgjl=c9o=%5MgvcyHYd-eW9)ICGF3ZkWYa2SR6kohC=CgOE
z6gp7MhFM5PRcPzi54X0yg>GFK$I6wX+TPxV`_U*xHx;|rvXR284sU4e|Kp;HJ%!fd
zM)vemS9$L7OE_nFWOsuL8yzl;I@~>5;FKoKQu>1-)2e2pNcr&7SNLPkTp>>solb$;
z*<-aTabfO|C5q|FG7mY({Y(Wqx^Np8m0;t!kvgSF<$J`YX3d6*&~TPWZgTBphir5W
zSrEE3)^j)$Xn9dd+nE_5sWvJl%W{&eD+wTVgXw=8%c_g3pXcv?b;y}Z_+K8IF{#E(
zZOvkKk19!7DueFjWc?n>XtGR9DwGk5nKGJ&7=0syB}~XdNW`R~$c0&?1LFv-_~P|5
zK6`sY$7ufS)4S{qGd$}7Q4*!ziZx3&3ockKY14b~QZy~C(7ubnXz68|RQonku56^;
zQ_xbrqPA_A8-D1qJ^scof0?<1zw*iV@yg8!7dJbg;JY8+;^u=H|MpAQQANVDmv$J>
zmi+V2zf3xt!^ViJs!(;!CqDE959PPQd>;n6#J+eh!>c?I0?N(P6<4<k!NZ-rmhqxt
zx1Wa7Nqh>ll9QpEX{Jltf=kD9%Y}i^$ek7Qs=a{_Ve?Q>&Vq1OH6?k!3yX%d$mtDx
z9Dn0;Jo(3dki)lrAM*F{E4P1>{9wrc{=PB4d_3iz|4Alqe}SSi;_&z$<H3kDPg{>h
z1Cq+Bp{u|-@g^_{5P+{58JC+x<CGK%N^8FG)-k_(YfQ&z{@k;B?Dn&@fF&$^-QEM|
zc4|Bl2_1oBeZ+b%3uRu~Ru(P~IRHNAt`Qv&#`trUQY>pnx5(+HLfM*D4Ot>ZS00!~
zy)=<Esui6qq2KE<@AcX2cj<P!Bvly<lY<|Tk5`durO&NYPN|R|B^-gOiFbNT<nZm^
z?BYBm{I|qNGU_t$23O&GMQI~D-#HmbHC9fUG%cu$f=-^%PZKKDCr=DM8IdF-#`hlM
zpWWY~(>Hv4IAd<_Qu;$GyP&ZRUI~FT%@Ry%0?k}T%tUJ%8Q->{SkblyExTVS#oLE-
ze)qjoI!5yo&+l_-vm>OL@H4cQ@pA&SAy1i>_)>-3;fva%wS2Bf6jdXWxW>aEOITUy
z^;fkIZ@dfes1p>;2#}NokG!&ReE#c4JaJ{0$+F^~e*PWCtBR9p$)KCFT&#HK?kP!P
zn5-&pAC9@QHQ*D^U!_`=eC5V5O`_R*=z!mT{Wim`lVC3^PUEN3N*u+U=oe?D<<cl?
ziLE+ZO}#K4E<?7F(gKE+jVy{HHS{uJ?~FQv5;ji2p0lN8S;;9^Hgd9cQt1dVDbv$K
zG=_2G83WrFALjM9Z;;VrKAw=~J<4&$|M2x5Ke6?IpYHu9{>kn`EKct*>_1LA+{9^3
zqa_H-BYX`NHkFG9vI+3Lha?fSPzU@24UZoT_~WNrT-)ohH^{Npg~cI@*f~w2d~5?l
z5#$NOkzmLa+Px5J8LCF937YVIvIA%}>#HCQ>jZ^wq<Y1x<(UgT-aDMr%{uH13MQqc
zpUJB1>4JK1#OZ9sQx~^c)Rx9JxWx*uHAR-8Q50J7ut_Mo8HJIBxvCw#JPD&tc|HBW
zOlq8riMM-(dFfE8%n2InC=&4<h&Hv$u$Utgxz^G2Qck#0M~JVfiBdL^x(Q<F@Og*Y
zWZ}&QsyU|UW4$IfhJ`N3RZZ5wuiQE1&tK^AW8G7}(C=_OpP=*(CQV2Ii{uAsYXkfY
z+6H$pFLX~Afc;bnSQJlAl;&ek@8N?FfdiZkmYvGAeoR`cwWCz3qd-H1ot;nt*Irtk
zR0}y{vb4;~xXU$3VDn^J-k|6!6x;FGayUGfwmbaZYj^nkYq$B(!y~%6;js&SE^ZWj
z^r;=5yS#~Zj^$#-Mv*eF9sl9g8~nR3zrm`iIb4>!dhdv5AK7DTbHHR}vEI|oHKXF2
z)3>ej9L{P|quA~xq(;*VV^O05CCw=AO==0NNEAgX1#V)*4)4LV=GJM&?bDL8rI=!L
zLf%1Y<Qk}~$lb7(PM(r#cx-o*M|Zb4IzHj~$F7i7Q#N}FXKQqlGr#jys-p?Nc<Xtd
zeN^)&k~#UtD@@DRNHXy#bXwyRgH984Dn`~i7+UMpU=oSItG)0MV+fWCm3T}>L(wLe
z)}bF)S_<N1eP2r@8~b-5R`}paLu=gZfnL5I40$ia%Eo$D@#OHdB#ty5UqD%VUb{Zw
z(%z6&<M^GgA5wa_KTR2R5?;PJ<+Yn9=uS>KTQQxlxVYKp`O8}zPL{lU>zIuqW2e~U
zD{tPRI~vPOTgy#8Udc|0RI$W}fVFl)KA8rMfL5x7tP{llc<H2ljYVWXmuqWpkhV3#
zY-PjudbIL%QfV)vrEZO8xhiQ)%JuP_exljjzr@$yy-kz$IXRo*ngz}o(u4P7r-!7x
z&R=~Y=Py2V#`_LC3=giL5`#|?Ve>dS?(s%+k%6@9mB$%*y+xjYwIL9~i=|Hpvl7LS
z*w&yzOX)-xktUjCesqIx{qT5HHKjhg$tr8Mp7#I%8e&OAK~y~4Dv=}dc;2)sp;0FZ
zcdn>zdby#j*R7yCR~&lpQPr5|w@&%o=e|a&6`%Oz`*`EQl8Z&jqPAQdC0x)oUw!+8
zjm)yM)nm1q@&{kIgI&z|;4_bK_raLybjBw?@+8A1VRCpKtwiS+E8C+*!?2LoGG2<N
zVHyChMo9)yFEcDdy;C_wvl=ryzJYCogfR`f`gq<@HMW(bZTB-tNq3%AsBKOr6<fU&
z)695o(^KNOwAZ8G&3NJW-{AXy;Cp%L#aBqTy4*X=*xbyy|Mpi<=?)J)^P}9o{_ps$
z2b=t*gPi^JK8wW>lgW}a2>B~3RjqFvwr=E~jR)sw8pn7hwNK*3lfc=Ms+^;pu$`3j
zxYTvvOgk~GSUFIi{_-n+UON#9D1jKQb4>5O&i>VBSO(9&G|^0!mVPePx(_GEcxhYr
z&{gf%ccYhd?pYP?FXveuy-8>HakhFHN_m!neJrqc2v+o-G-MEK6$K2^sK(AhfEUNn
z&2O-E@ks%otUvel;qPR94S7PyQy}O&PaENu8pWG$yu#&c&!W5!eui?x8@+Z1lU6ch
zDF*x`Gpgko+<WgeE?s*Dqg3$i5$c(wt%c}0CF0H`7&@$KOOeVc;}pE}!hhn;m%hk5
z+ds~BH{<&C`}Dg7&1_CnHf#?2bn}dAHb*JLX0PDVc8|OFPw_e-#q;Q}Chu+2G(wD7
zR92#8louwF7Y32`kPI$7%5DIB7NMWD&U15dh}w9VjuV2!xU?Kx&H~+A-jG{o6<06k
zEs=986n?_!m0DqILn73Ax#8~enElI-GYQc$9rexDu<;|Nm_=+NM$|+v90wG@<oE&k
z*264o$3}OZAJ{k;WMalx227x`kShW{;F+!*olFY6+IhA%?lQV?O=9DWEF6u@G(ZF;
zF&ELu-A6YQM)b;hcKa#g<9qbBAI5pekOH`N>%0RS7EjDxRD|B1IFwSX`j2oYKjizX
zf5olM4Zg5^hQIlfKgd7%x4%tqZ;!M4hn$uMU1scz3f{eS$ZKDFgR7VK_@VFpc2;G>
z?C1u)jct5mS*$EY>d3W3_hw2MM9z7UuaB1vX)k#Ggg01+RqulP)akquD_AAg?Y&&L
zL6WuefdTB$O1uQi#tGTegoxRIw|Zb29)9CR&)x(Cab8+Bdx==p1uuXICuf?FRvO@1
zFJDs!#Im%~1;heyQCT^0g3+a|<+Uz@+$>~Hx9wq+o2MG))-w;Yf>kXi-a1qR=9LZc
zIEEq>6_e4TuBsZ5(%FEO)PXD$-^Vg+j3Uw4pmaiueqd2ry1BuValN?6&+Oe|eB%z&
z>Ke1!lXXViJ6)1^Lt+wa(=aJL*5pw1Anj2pCF*1)Zlu<L(~iQ(HH~sCtRvToDo~>I
zAk4o7&Xh$kuL5Dyh(^Dc8~VAlP))$f6$0LP<S2}C>&z$-Yvd+4g^2%u#OSkBGcE0U
z7bq#hHhPH|Oq%E^Ov~c6fn_9?r{XzSyST#y1&pfU#(8>~33K}xSRoHPO2`L=p&tYa
zc_Khj5~OM|prR9?;@U=^0|A%B2vf{4R*k%_L@TPGOKhR=uwW#GMSYVLzG~KX^Dzx9
z3O0@MG&99t_`@!5RegSH<0W3ZeV4axy+%_XGpj7Mb6D?i-qCo6%Vx~#6YfoF&T5ZE
zypSvkV+>A7B(6zxo38G)=p>WzvLOj{ugoZZ>uX1JbHlcEY<3K;5ZjkgVYqo#a%q%_
z`;<{!-OPFWXoXUWhc?odSdk?nhw=!Azh4+;QH0?=2ctC5+@b-Zyzz@*H=Afp7EYur
zfIKGL$y%=+MVi1s8f1!jC5uL(=%k7!37kZw15ZHgXa<FDBWh9;I08oU2jT+m1&Wvl
zhPf<anGXmF#(P=?^r$e>Mby46c#RgBU6IJ5U4;}+S!`mYS)7C5<<u46Y&MzF8$H8M
zefJM?=k|S0um2MK=HJ1ZZ)0J5Jn`5=Z1!{fe21!bT;09K)$Q-(&37K~8~^zYF6<8Y
z$Yn*E8oU-Cgk{x;Ldige$5}^L1ts=&PuRGN%CfA5{nE*dsBjv`cwxhvAVQ0HP0+q@
zMK2tai)sy!M3dvj3V%NqjY`SbHhw5*O`+nhiL&yTEf9B)<m%CS5*sZBR30TJilP0C
zy10lxbW`csIagH5*QlUN2-*Y{GkN5-_tJF%7+-U$<=P?;XylZVQzHolVH_FL{(c>5
z8JXv2dd<I#?`9;;Xsz(hvfayh?CL)M{^U8vd6)0M@D7vZyUeF^rjrGWRmHNZnUyl@
zudE{*^hk;ftH2vg74QM4G&Zz=I+$6SL?(;LGWhhw5U3A6ya9N|i{K3moGwFjn-lBT
zaBEOmM?Vwe$fXtc@7N-06*zfmu$#{m(|`&#;iv^cAFcB=R(8Ou_Vh9pbb(&*@32sO
zI4bL?8H6_HsBz-r+0ArY50uC*T!yiB**FZEC62R|ZC$joU`Aaxj#cF-6UQ)kC;PZd
zy_ej=aLv@ASOkq%9ve=C_&QsEKIFkB0cA8=UV9#{x++l5Y`9vJtl*Wq*SYtbhm_3;
zXS2J!@yuVO`^NvxpYEOTtEUzI<~AKAD(#~a&HXz=oL%zV(@(QMOvz?9sJ!GJHO_?v
zAaf6=99Ai`(m3tV){F8c)rvu`gI92@lr5{qwQe0T`7%DZ#;;Yl&IS8%9jIka+CUAQ
zE*h#ZIGQXQbRdu}YA;o{mK_`iM`hsZWr5NZe`gis<ypG+4Q#A{Rgte>hqg|TOWEx)
zygE|AQY~cCQAcF&%V1w`ooCQ76qyN;tn1&cRNURNxTAf&B&EwkZQL4CZ@mw~_qcG(
zD<xM?JpCdcA#O0OpwUCdE5mfPg&#c2Q`au?qW=(fv(NkHuhQuqv$V%dRx8G56V6tO
zan*w?5$k#*>1VZ4EIm{xl0@<j8YQFRhTzd0so%y(s}x^Uq7@t6gd!0FWtxcR|9Zq5
z7ENJVLx{-#m-|e-cswxdBw`k=6dS#?eKU=gK~9|4pAVOegwpfJWE${mYkeSMtf6b`
zeee*HVl7LA0i;6ppJ?$fM8T{Q6N}?{Lm5tuEYZQvO%{%aHuff-1mc9|!_jZrQ!G#9
zJ<LqdRw&vjqjjEH;0-t@V3G>=iVlUJh-O3eo}}o!!=(%RT-ZN=UQZ<5qeuDqx1VKy
z!|>NHy+kp3o5kvgiUKp*BpVLtbTWaB6LBN)N|UCEbOwQBB9O2lNi`}lXltb{iAmUr
zUZ;Mv_kI|Rly<H7jUTh6mE)Rq+ZG>A?PQ0mu=8rSP9lw$>pBISQ1MG47iCPU{hDy<
zxHUOdhCZ~!)Kw*gW@1|PQ%gmY!psHDq9zML8PSSH*uwI)vr<ebjag^&sIW^xAm`f3
z-xjff_HXgxaS-c|I9CB7X0941>>t0r#&lTx%0pGxRI4Qq7A3bQb5z}sBni5z&?e{2
z-jFYp3w}`FLk-_xTD{A3HD|u67*~#^g<2ueH>sh>4N01S)+8Yb1+BptA&zuH5@#<r
zBw@H=!a|sqj*~@$^I`$h%fwZD6&A<QtYJ_X37KfZfF+GsyF|VhnUE_XAah#y?TrsK
z^Kf@IR<QWo$j~VI|EjP{tdn*ro(8d?ssaYl5ANfg!2XrYULUG+#jJEK-K%PR8&;7R
zSxk8%wL>S>fg--HEn-o+2*_nO07%`SA2?guz$_9f8puM3S4655c;E5}BVrj%teS8=
zHXf2BWv837y!{@@{kO<`#Zwo$?DibPy?y@hwIToLwa1uj7=CiYG1`5PJGcJ`R);)s
zameM}t_W=M6rCEzwU~7bx(UiCN-uDDjr=o-(=3g!dILX_)My5|W)V~gQlL3{T08_~
zZZ5S#9&Kd9kR<{#tH$HJu!n3IWIHD<WTNCwH#&q-vL>EHZ-F>Ni0xGjk&5F`=hk<u
zK6gU6pj=T(IDLr{N0A~m?chk*Epp5&B`st;H4-hYVcgx(k~X$@jf;K<w6X1-Ovb;_
zMmPQ~ga@4}!8DYY2cn56sBe>3wE}}B*Pd&ST;OEBVr&N7KU+d)15;YG>5+8~_~rKw
z`Kg13@5OWL0~eUwSKOUc?Dh>BMV2R2){~_kgQIqiEKBf>#cOa{yo0m~yEaP{Nh0@l
zG-aBWBEMSIj(I7V|Dtl-Iw?7tH{3og+dtB|NEnqPwO1IHWkVIH;!#<qfIFvU>n0l0
zMdIBVEB5u#FAH?l2$&S-_|XY#QCUvrO-uJ~Z>X@4k7r^K)yd3SLLN{%t)&{jH>prc
zk%h2~RqX<@DTY(ym1Vke9L{RWMvnd}T$xJ2jgykPksy}Jt`}@=U5iF~M7OZC(m^bP
zHJ$TPlT?kMp=ixU0<Yh?!@cA8=wNx`(va<};gQ`T-R>5u*rdxf{`G^46c;;OzWW8p
z54n1A$krei@=#(hnPK4-l~;6%l*&Qr6)rfLO=^!GFPdQEDvfB&olOFve^EAV;TOI6
z$`JiBUN(Z2dog0J17A?#L&|rMh5L0ORRw}V86I>M7{pOSP$X*~z<B@0zB{I7tQy}^
zx<Gp8_!_lREGtWD<Y;Z2Ymf2x{Vtqv3fh<(t%8x0m-nKqT^m~MJuE^+=)9+&3vE8p
z(qayZMBazc-bI{0O2)Jbj;o;-NQ`tI0!GPcutrJ^yMrF5ci&+?d5;}4<Kkw)W(Iqk
zeTv;jxOnMdx*Jb&b?>|R$KMz-+|2oCew%xfh5S7T4xoZRVNqlfQxm-kC6NP*A}LbM
ztM`|5QZaS-()BTW!<>PIi-QzML;kGda8|S3Pq>tbn&;AR#5e9Q+3lz74aB4W!XRly
zB!j}R3KjUQlmcy77`D37W0%&G>lhF%y1~q7at)G@DR1N;@DC!9V^kPcjXbnbCm~CM
zdw5tlJN*>AXS0{EqvUr@V}3xWb6jXy%f|84UWc#VbgfC2bOj1$#R<(j+Slx$O*xs@
zVo@85*Ld}%>)tWI)*<#zL!!m|XuF@_ootA1rUP$7@zU9p-Tf!{mw)$vrnH`a_^U7T
z=l|Ys@{wx?Tt6=P&Zn=iJ?e3Eyh+{EeBZ}D$iYUz8=w6}y1(;t{Ezqc`QLu{N6GS(
zL@TzthWVoAZ0YG0DXEtbE!S8EnI`FHn$1pv^^(pt=%x%i3GOUfyZLq^HY!Zpu7tmj
zA`|SOjdllNnJ7W%8Y?vG!<oZ`<2*4U&TOmafXJoi?ul3-pZ^d^m;8Iy4w~fuCjs5l
z;cKj2Upo-2WkWp^V|<!`ylJCIjc$!7(g1YnbJctN&`!9M5!Fn>=E&m+S<$+##YA4i
z`c##_oA&2soV&L|8+d7JgPmbUZ5_`%e1RA5o$~Od3w&g|OXCn;|J6;b^Xv{XwtER*
zyLg4Kz4SSL?$T}2!lF_WIx9zBg!j%ljJN0}kawLYdH1zHpbmyim33@rg`X>yCpUQK
z)jyClk3^2FEWn*I3Q}!L_5jcQF!1nR>|i|LLfa@&@)ke61<D8bS0r6A#lrar!A9xR
z+B^7LKjh^1j1RA4oh7Dd5q@qPfg2wX>jE#l&n=T!{#`LYewQ1sNhzA@s3HpPUOGCK
zoU_6W-y1iGLUDHg9aMXb$k(4=#_}3sh-Ub2tXWh5dUcP?zRX)+D*(l1x8i4>uGm<8
zozAJi_T^NrsBWHc|E^=6Ck!h`x%V-)MpGtt-e5c%W6K=f%yA6Jk^~xuavp=CUR9*p
zkX(KGyIRuZWZAIUOGz}m^EJojwGXqXL|WyX=fSMzU?Zm&`~()kd$(=`L0eRkD>z+Q
z(f}_{mX>l^(^CrlAZ6##4>JfSMH!^kMuim_u>qvTi#sWP$RY`Eu!gr^QCxoNJLQPg
z3Wt0$T*s7lZ7;KG#5#7PkhV9Df{haR`R3^oeefYJ4AS7d9ZW6HEyN}apeiecVmH(6
zVvUO7D^Gu<U2Na-^1qo_5?a-m-r)&aPY*kff4DW4>@TfA>|xFl1v);=N)$4v+PG#J
zb%YW9hyVO%sb2q8{@*%de)ofPk}bBkA7$^t7MFJh=*BW0&-ut>TcQk#Lx^6Uv_?@y
zDt<au=nH3+V_DiDq6n)stnUc0R6%|fA>pH0%`y~IMbPX|SK{CmCxLaW@B>*Q8dOJc
z==}d8pcLP+_$D=0WM|g8_GeUX?Fdo@<j@7~g0(>wMsWB@6#Yy~B)we&Q1SV-1EKIb
zVlASSw1ZkpnyLz)kK9Su{(t#r?S#7!yByA)M74{?6JlTI2x%z+@2CUzvIyZ3v1^cy
zBnEIRymx$gsd1x>pV%_2vb!u6$DB;Y934-%f3}c~>m+1Kt5Eeejcb3n*t$5*<_(Q?
zOjefJs$pl42Gy?|k7r9ELv&Kj@uH!V$s71!QgQpN3>^HmJ$q_o7cFbwI(Npwhzosv
zYQhLL%pkt?LlL_P-}e1?$0b>?Z1HYwa=mJN>mbquQlt$7nA3&0p7t`$=|Y5{tH#qb
zzSR+)|Ls*61jT<Y8&MojSC*rhw7W4@K29wav?p7nvz<Q`oFfEohYl+|#M~%WmE&Yy
z3k+|hm^507U2C0;yshVC-Y^MKIKpKxu-xY#zR}~gS;1d_%404}7;SG-Rx|D$Pq}q8
z;qZ9O3vVBbk5N1_yO|u#wG(VQ&x8<C+BK~y-t#(;AZoi7UBm&&u$z!2n#r<hN4ee1
z;E*AV^Fh#Qgf9>Txb5PI67BOOigO!3z^C$BR<-BSX5LOJo4~$FwQ8a5cq%O_A$QIz
zIlficS>067!y*xb<v2XD&bPZc&M1UJFNmelYgVvx>wxS!oM!D+rq)<Vbm@u@y@|C*
z{2DJm_h=>#5q0Qx7FED5l#pQNRUD!zCQBjes&M@ovxA#o;?uVeI7-j>$&FLy_BPY;
zdrTHfxyHO^Wj%VjYIx^x#m&=_dD$?poQN!)X!$N}<aATbrH!0r<*0+@sQ2*DX2$(#
zCGFoZYn{TgP@O-#Q!uYA4{a69Dl6gzFASa-30nk<n)cq*!BjHto^w0@d^GBLYFMJ=
zL7Xicme!LR!l;&3pO6TL4ew#IBj!7&%SOm99l7piq3^H4u1t;orrq2!e3W8dI%bun
zlNmO<32CSsYPTMgxpT+6po79DYQp`0UXmX-VPgd(dTh5~f0WULVODBHfRYEh-l$vH
z>?S;Y&||BcaCh1W>Nwbh-gA`x3I6`uKgMC*@MBxQf!bKHd2yG~sK@5PHlO&&BT@ro
zsfa;~)NnMbnO7o{vR>}i#;#Yb{+3-C@1e5stST2EV$D`BrL3jD2cc(q1P5rsc3#$g
zeNQ@hsAav*CO98#5Fe)0EOJ0sjUaZ7lO)q<D_dI$q;TQht*sQe_op>?Cl$+DTukFs
zF^+;`d*;H6I$1P954o1hMeAAXJcrYovX+03rW3J!U4=oBM|Q76T+FKQyR7vgT6fJc
z@Zlur=SF-F0<8UDTC;3n`hcv4Lb8+T*3ZZVV~^2BkG+Q<#cv$&+@n|cm6M0KIW~N+
ze!z0|HV^LJVzrzxuPw<=FM*!yxW>ayPXLeCZ%L&KVq`q49apz<j^+)WRMaOoj!Qmx
zxzGJc&3M^xWjklQXvkABQz}x;PM%21lFCk0T8J0U%l%tB&mh;7*0Yi7R%=N31B+m=
zybRN!RUPbG!WA-J)Eqn5?xi$dcFOIulAV4k_O+SDIaz?aeYv0SPiyuD5|HUU94{J1
zMZ*3l=V)3nEiH}HEGo-%B_PeBln||01Z$;uY_G$3(eUokiZl^A{jze@=LS9z>rWKa
zT8g)L)fK6950P<PTNjM46!#{QgRnbDIT)st)=@U5MKbX|Jbbu*ykxVRa&aR^BeeR5
zFKp4vGhTV^4SezKe7651{K@iH_=Nt~{F8E*^2^&C-F;p#8{%=>R=aZ33J=bRCqOES
z(!wb2j4O&%w<~y~n}k|`_9#5>!BYhCs8&h>CnG){?yR`O8!rUXs5>Ma^~Z~bP9i?U
zS(w@lI^q!S8{q|3;d9pc*67J2i#RnpXamJTv<VWgSlzDziz!;;OjnkhtA>7_FkY^h
zmyNi5H;(OIMv)mhsb*DK7Nr!zsZo@kun^?lwjSybxmN}FT3R_6uNo8vtwcNF9CUN}
zUZWISy+qR2Vg25vQftV`>q7W8bW+2pBg`x3ppz#&yS>fvFz5E{D&KxEVt;qU?|)^N
zQSu(2>@@uDLsOd78HuoO1SPAzCk-&%va;-_LZCdHHf#su#XK`?hf^YY@f^--VcA+y
z2xUn^$UK&I%K$D^;Ef0}%dlXIBzh6bBTEy-aw&JK)*{Ec&`+h;Zv;H*=7IL56@#t>
zD#?{4MXAKbS@IzKJX>1n0)SO5@o}?&UwH4iFwFRR*@U}aTGJ;kbb0q^#Z&toKz8)m
zqG7tSEGkQpC{E{=L@SC^a}r$0#U0ImlQwcuuQ>zO!L+no-pE+go>}SG=p<n{r9vS}
z`*6#UJD6z2@w{oh7knrdt#urXvKBy7-cweV*YBNiGM=+jWW0EH%>H=Ey}Nt-{EaPs
z_PLV(ZhOuzjYlNCF#lG_$XACSv^(tJoyP>zgvb)A8pr8Uj=DwTc=|%8<rDUEBb4vJ
z$X(V##g6juUY^rSHTP$A;MJ>8Xvt9@@BO4F6T;d-w=XA3Wj(z>2UkkEiFuLJ%Qdr=
z<;rHpvKE&zVF62_Iw>u?{RHRZOd;WHY1!>1?YQ@7-mt&XXOL?&3ERDdRqePq%D6wN
z1I^rXG?&+Xd6c0QZ1+>j#&I;Oxj4$1jXUH<qt)7|YT0;(d1Ng~k+<EGQEewtY;;oL
z6A(NM!koog&&kTNH%Q3?#ck9vA<b2_>g$D5+&EcrFv{t63@-RBzJGhjY@^_}UU-8)
z`Q#Oze`JTxHurh+*8BOj!zDj(aEqV0@Kydl7TN|7hRO?U00000NkvXXu0mjfelh_!

literal 0
HcmV?d00001

diff --git a/ico_design/icon_v2.iconset/icon_16x16.png b/ico_design/icon_v2.iconset/icon_16x16.png
new file mode 100644
index 0000000000000000000000000000000000000000..5ad1cc5955275261e28b84d334d72f9603d7dfc2
GIT binary patch
literal 849
zcmV-X1FrmuP)<h;3K|Lk000e1NJLTq000mG000mO1^@s6AM^iV00009a7bBm0000i
z0000i0i!y&M*si=14%?dR5*<-lFLtAWgI}y_qgA^^C&Yg4j`QrJ4Eb@hTz!Jt;X0i
z(XBiG0^|SS&W#H<Ze5vZ(uIj8teT+GR2vsGk;<bYZD}bEAj|-BpWn|#dzL3TyQ4n;
z`g@r>J5BQ95Zym!?a2{;?DUzdJKAM~FA5Y2#Iv-Q;S_X+0~U4@ZjFcNxlV_Lhj+QR
zb4naVm{LeBvb@=>&=-X=8V`UX_Ch!FD5a2BPRF_BKYNM4k5)J>SD3xpVysr;hpj%}
zZ1u^MMq9&nT+l0o!yHl(6fd#$s^R_F7QN{PgWd`ccIyxr{y9!)R~^q{PrGX9#-1xx
zL*|9mUj8Z<Y<;Lgv-ui@7jFO5WqIumtuwQH`oU$sU0kQ<6*_cG*K(31rxGZlz%c6t
z0kqM82u8thqe-{kq31W4xG=$IAI%VzL%!QM;+MUIBrC`TPzn?%rBL@Cbwyi8;UOSJ
zOh}px_~Og2sQ4)^HilT+K4PmE^Wj95^JRq;0=_^zXk#tPICS7p#(@sd*0E-$8Mw!c
zPK|N%dYfTq`SJNNx1SE!&p~MeTBGi+zYt7_QlN_z?IMaaX6er^J{w^A8w~dAgvT+<
zn+NP1CrpHf>kUUE&}bJ%47|kz5!OWnE=1Xo&deLkcP{cYZu0nKm7r1Klbe^h-m0-V
z$oYLgMF$33qzP3UQe-(hdp*{kos!2V+?Z=40#mI~4l3_pi*=TtkI@(@bN(<!3eRud
z7#)NmQ4~=smH4+;aDV%Rg@;|L<r1Z+#I>uF+<LFWlfeXwo4;}rz{l^t#dxhmIda%6
zNl0@+DHy51jrI_|5y-Rn<tllC7vWrMoYRwQj5LqfS54+;MhVjbb!T;7DwQ%Kf+|wN
zC_?gt9ESjs$5<EPvlthaSh~N;&r6HUpS#3`dKIgEN)l_VhAfS7Rv~!;8Z?k*30f&6
zPw_@GJ{&TA<!x%!T@K<i*f5GHlp_}*j1(A$cu2ejAjjb)Kr10Ij<c<CUZ0#~<*!F9
bJXqp?K@m~RthB5100000NkvXXu0mjft5|?V

literal 0
HcmV?d00001

diff --git a/ico_design/icon_v2.iconset/icon_256x256.png b/ico_design/icon_v2.iconset/icon_256x256.png
new file mode 100644
index 0000000000000000000000000000000000000000..53f07684e8dcbecdfd88c1569e9dc603874e4107
GIT binary patch
literal 119441
zcmV)IK)k<+P)<h;3K|Lk000e1NJLTq00961009691^@s6Tym&p00009a7bBm0008S
z0008S0b7xs3jhEh07*naRCt`6y=jmo$$i-O`)5|Yv+th!*tvEWdtbmp1VDfwaq<==
zLo!WMA8gYWN%{C7M>xWA*pBc=g-zQb({hAFT5`~&Y!W0RKoa+ozyjFC9<v89v%7Qc
z&e`45clW!hGSeS2vnt=4?giwGnCbWGWo2b${_iX0tH1Xjx<A!-dG0Uo!>C7S%pYN)
zlhIX<&0figV~2S1)I9xB$@0n;k8qx+KKN;NOhzXYtg$qeV^mnIb+i;X>oCSrl!jqp
znaTxc9nLv$4y^yX)hj8CWwxo=8kDp%K|6;MC<?>KIGPFsz&V_C2o7r;rFCclaeiFu
zz*!4GSz5O5US($CKD5^Unz3jcB7QC=^p)^$P#79oFxFv=Lj<gKWJ&-~;ThmiBL11y
zfdl7!KNseS1LwaH|5)b`SAE4<!(exvR(ApG@Xv9M!h{~;@6D7SGB)R&|0#^6LRahj
zyGi{>jCJU+e#V6%okQ^RH?Dp^LLW2OM&~WEEXP;}A{54wDIrV|6&~UMk+472{OZqf
zE-v}**P`yDB=m{nhIw0SaHAfL&NMd6T?9%CTKrry6~`4!c(8T;cNIRn<hhbgG+C24
zcI=OHXyw<kG$p1Sq8nX=@i}TG6aiLLSbHRbxQCO;W~}e;04-9D7vHgh57KpY6oVZa
zt?BAF73LfF#5ji*zYa<S8@?w?J+|g?Kh)-*{eVwjaWFk(ZF?KCjGfZ4v}>4KnC0B@
zIcB>Jy3LG3i&NxU=$DQ>*EDj4(u&fF|EvTPAVO<JrWM_$1`$LF9(|>nLA$AGH#MT5
z+tiGV!-`NiP+E~`!Nr#d*L$UZ>s<TYcOsDV0Z>X1{Pm%A*Z}Afes=<*RJHJ>^~h5g
zM`2)OT>wa_0IGUMK%NQ#l~l-I0lRphj7Z|A=GP@Ien4ZCNc9{GXzk~zLVr84gDSj-
zgt0^fgP*Ih4vPoA(pp4?ah*qq2s|TWu~4sJct-V}@!bNUi@+^_R`nXDz#5w*z~!g`
zI})u3w0`dO`=y_?^6wuBLrORz!O#hAS?4N3SRquk594!;W5)MQkxnIRI^Luj$5>Yp
zM(6+d^>VKIY{LVbtJr`i(8>Bl<c~bDUN2XzS31Wy%Sdp?2(9?>Z1SvlN4h5IKE?Z0
z_V`JbE^YD)vuix_*bmsf`*r%u*C>=?SXu_gvf3-TyVYlRP@uG;+05uPJYwgWBF_{e
z!qA3I(u&eSfnWn7q6Do46?a}ML<w3eHiwqNdV=6LSg<0LPADCS3KU+1Oe-u<xOx*1
z@f2J9cp`$sbH6<F_j6Fy0!SbN5{MOnkZ3Km0x`>Vz4R5g3-gokOe7GYWZ*~y2#90j
zhK7DoqKOb1Z;ymPl;5;W<Q`y*KoqY>LSXU4fk63HfX0gGQ9mQo0ieS~)%YWDxSAV|
z&AsBK6=0<ZRuj8%q#B!i(l369BggSPeo2ME>A$#xBIEa0&x`95A0gvNGl|;i`p2eZ
z5?NeL*uws*s&6F5QW3C7)CmMMUW;>$pH1?RILA1Cio_`sh(u=(8kNkyCfGG>)oT;S
z08szwrD?u-;w+!Ke2w3jT4UzKOY9DpC<c8-&hgqiOI*0M$(z?!xqN$*wf>Msu4v{8
zr95(FxyC6)yQ#@Ck4#D{oCqpYp4$dUmB@($@hB`5)^m0zLY8SVEoi0CN&<3u#BOMz
zk$GCLl_w^mgq@M~i!DN79c5Tt8&)L?i0<E5RU6<rxbkWQgs}s7iyuP=X(AG)xR~+&
zxOt2%da^?z9$3N#o51O%f?(e-5)&P`c4=e3Fm@O(B4*^!alR@%r{bI`Y$Osx1frBP
z*@VMY>k|6wa4!jmBdAjGJyQg&pLq{l-h+IV6f^PcF?vcxrz#OUB6B4nL`IOjcS?X#
z0+G(G>X;&8dOhvtY9bZaB<`MwXHZ@hb(KmMrKc2`5lQx;=IZgUmMFowTH2_l<SK$t
zNew9x43cSSUA2zsy43Tn-%D-Tr+9t4!=;}+gMDd*|6I2@eCjHr{w})Jq`)yW1`Bkj
z+q4=P5}`vWj3^2RS*9tRM|{Z?BZpT92-o%?0#-b?o$uyMH8oivgIX)J0vGz~EFf(q
zehLPU@S+6KYD_`^L8&0_+6Yt$q$LRD2<k=<<8$q$01#YUt<c{(SIylISC1o;D+{6T
z_;rye1^I1@+&ltv6qzGu8smT!GLAE35wN&QEU(E*4g4vJ`>m5o?3M%y(>JhDND1#>
zi=9yoO62#Xr4rMA2w9E+_IMk^Pf2DjjM-7fIAP>y$yW*ey>zq(*WBm%laolr`u*k!
zHDda?1)?leL?CkEO8l0+b@cSzqBIcqP$kzVMf$j=u3}a7Z<7Q%HZ_0lQFeAu^WT4K
zi3=Y;#MzscFSOSfb^eBp-66WwVuT<HY+2AR3J`@;3eld6XPL%{=gumK*|}D9b48we
zE0h(vhX|-Z1p3DRUiSf~6nU^C80UWnf3H^ISt11HkCVvd1A+w%%9KaMxZuP^MJgb*
z5C}&tHAFpB1?AF-(9o)q4kY~MInn+^fP`3(aJ<Q`kq{{H2Ej}v>O*`T;yAUQUIBhZ
zi2Qv~4TwfV4;T@EB)ltu*wq9E0+1#E7CE0J|Ly^&l*m=jtH11ndp}-F#&%MRvh~Gi
zQSram*X#cKO=6N$hT-0KzL5)8>%qkR)cSvl?2~F$N~p(P7f>WN5Q)a!V2GYj9U_s~
zi@;Tc#Ifh^YPY|OQi~MBU*UJ(IKsa?+vWpz`~1WC+vLq9KJ)Z_{QNT~IdEu}yxXSL
z%oqekwX`s>0dTbhi9nDnW2>+XoiH*WS_P5)UP&Me#=auRol&gZEgUW&dlu9Q9n=d6
z{apf;BkLF$57ZG53hNN1`~sy25-6|Ms-~3zBSpMgpu+b`=ZdQtH@PA<<A}c>5g;J~
zmI=E-(hU;bFy(%6j>%>_H-0@f<G|dz*8zxXga^<}=;`=%8v|C^hf5-3fWH?JNS7m3
zJt{poimUs~ch7GVog>m$h&Zt>mV2*_4Ydz)q@7Z3opRGiEOda?ajxt=Eo16d^4(4B
zV>Ae+-V3GbmGlxuy<R5nldC0;YAt1aZ{l+%1sO$QC=VT{dHexf=jZwIwP*Oh<_>Y}
zy$#Nlw^%VkuXJ>$rkI)OfCxk5P+35D2ckXa9$AlQ{nF8BWbBMAWk66B{836NL_F3Q
zCm0uW?|_&ko(s2g!B{|fE4PzqAd_mr1#hvpfsAYI=T^Ev^(9)@D>msDBeC2^qhYm5
zVWn)~#u6?3DMTkk!v-<Fx<CHehX}-bE2Kz<b3xLvL6`HuGNIE?Q1tj)RC8Q_PA}z*
z6PlDW#W@%Pu@L{f=uR4w943G!mHh8PoNyr({o^~vYbJPR_UfMw^P`3Cx{bNq;`Ss9
z_kN=GUf1@J_<<TZjq`A461`s-C#LaQT=U8}n9QddFNp6IaVL19!u@e?s{5*Wy0PbF
zL2?zw-=ngbFm4yfjvYDRZ#I@pD+cznyeA*!&pur6$Xow6_T}4rcm4s^x8G%V^$<so
zL!*(gHG+1gXu1sJELsUO?F?hA=U{;dH1n_s)*@QrjP;z}DNqLIgq^-&R607jLW@W0
zR;J0Ep|Flz`S^g+fktpjVM_1j%C+E>k4Gqj6wy+OzViCK6~S6ZBU3nQJwgOs-c|7g
z(SqxJyUqy``vw<b+zn^FIwUp%i<dJ+T;-9d3}*Oe>j=yu2o8;~J~Z6z3-`}Rwei&^
zY8AxVDEbMK_?!ff=uxj6<6y3Kk*V<Rar~$PO=3eWB1cb3kh%iYI07=IM6lvK+}rlT
zAXodYQ~OKGO;bRgbbpTh8rW`X0JFGC=bu!MB9~95GWN{<h(K&(M&HD<k{b7-{w-1?
zvr5%CYF%DSZ5XWgWt=dmFhk=!Oj?0l)vHx=c{jE#leUV<_i)L}3y^*CkrOnv;`xg=
zS?diErTF+04{>y%&AaM4AAjzj^UP0P<M&jLi?@D?;StSLE2GiOaiu{i!5P6?i`F3G
z#r=kYJDY~-ruIl89{EKp6c(J&$c44ukgbv9P$xqaG?bSLOyI%}$h3eGl(C3|!dUW5
zp>T8>nvGt`R8ynz!8^T?@rWHHlvb|%LMw%`!DC<@c>r>8ptMjLTOpMw<%5O7swoxv
zs(=&$n!Q-RKM_z$#;EjCujI|u0p~Y|fa0^+fJYWufiu+c1<{pR>7&s%6^KXz27$Y}
zdXppJsF1A|7b-QxIlM)~RR|rHIpGlgon!(Tg&+~_i=@;<bE2D4;}=qnnMw`&o5#ZX
zGOqK+_x*8f4(WZ){s{4(BtAjnzIY^#b2Zj6C@l_XW!|71C4)=};-YjQ)jLO{>ZJOW
zi}z&p3qz(=@H&k5i44WL0AKN4!m5?_UVA_)<fbh<Fx6nTsnL1Pk@=kU?E&qEW-v6o
zGn(Vr=g#uz*KYFJ(<iyGcns(6_+S+!NUrG*4J|EXN@I-i+z}A1Xg0Lhy_F!&g06{x
zGoFREn;HGW2IC$ewJR(`(a6BYs6A0Yy<>vlltpFAZ@%(KQxQ}ph?UBS;yJ&<nwkhi
zL=)$qTiWPX_rFC|f~b2%;0r{+T8~KYM+6+gJ8J{pTpMtuSB!1ISC)DR@aSTz0&HVR
zTzZL)9FZ<k{salfVgM^nOThgQiJ4$Y8YxzdT>(=4$&-jgZP){-_HC8KlBiJOfhpqc
zh3f8oQ8>;dndyGg$oOOLy*?>Hy-(lxTRPt;!WYI;mX>Cg-*b0dYau~IYBeYA9bqg+
zUP3J~#ou9Eq@SF(!(pvwoc2eoOZLvv4lQN5yt2V`u9#B7^Kai`s@tU~Zt^eRTIaKm
zevrd&y~XLdchK21WE#md))+eN9F4=4hEX9jwHJC!>Ggh{SmBAtZ!LoIqIvFU<cboZ
z6v$ODd{`J*hY?3hdy!vhz$%nA-hgK<$|*8sy&VEamMLsuD+*h>K(PY{ch>qC1?%sL
zD^vz*t;15PM9i*&YsP?}CK}GwOCO1V4aU(omU?`DrO#Tqms!rYFE8~t*vUB5&cf!8
zM;cmNTe+zKRw_|Ais{wQwYE_G9wm`z4~g7JSI{vIDPC-i_Kj5PsGsv=wFz|~a??1q
z7{M2>#amKDnQZmDMlL*VOiS$nV>XY8(W1NT9&WkM1IG-)$xqHfE7y?g>RMzvG2A$P
z|0rc8d}3@4*1F0~8zTlI^u!Qnq;@GP3GpRi0Tfb6274iYN_4Uh%xl(rhT|RK$ZU&p
zR4~&NPA%rVa{T~1H$TlEuO6m5_W}O?^*`Xq(T`H#5Nol<Fe*w~8l1KCM+JFPgS9wk
z5N8lk7-MiDCe<3_XJ9SP!C+Lf-5=4O$|*~54TW`{;}44h1vHx)hodMh-u5VzrSS`C
z9V#e(L+i)@!D5^RXR1YyIxDWuN$}j8O4qgFn&5K>T+sJ4u98Ar94m<Dm6)5DScj$K
zjpZIM-0gdh$wWKb&H=}jehI)rra0JAIOoyMCile9-bp~D+C-OJ3&0+x;=(+mODWD%
z0+5L!#Ew%%4tkZydYc0^jj7&o4<OpJO}ut$y^h>J_37VptF`rBO|&YqGU*?KU_|cK
zgX6coeEE@!?n`J~bQsrbJK4#F1vl1_1x6E}rIl2B7)(mzx4U|r*pxtFoVRZ!V@2*>
zKR4?AajXz{gsX_i<Qk-MlL*M~{j1+~2=s>~S}R6{@jiN`u%*EogTix;R<X-)v`JfX
zvgtOrw{n~j+MN!jG}z&QJa5s=GDf9kG%CqNnCZ0014s4=JVL*)IP2&S3Pyz?R|*X@
z8abvgSnEY91>~7GE@oPJ$-!8xv9xl3-8x6EG|qV|sI!jIV4HSp3WutYyb_aLvJo~$
zYmpd>TopfGsR`BnH6mE&-Dq(WcSeraHWY8{?u{Dr8lIZA9Bn)P;I^)=e{9<F_*_6Z
z>nb3wpEtpIT;yK$?{RZ$z#13Z$GD}F3W=ih5={)&j1bWvY7F^nPEzyhdIrfa2|M$1
zh>?PB40nXu5S9ud@sB4&22t7gB92S`)wO!ADbP<kJ%5~55AKtutS8J$PmJZpdj|%h
z)xXWN_G$^5Cz63R1xn{*eJ1buQEigYX_b*Z_=#U9*Fr1Tm9b)LWN^+=8pm`iqY=aj
z1HAF&e@6Q+{)D*?9p`_&c7%<MN0{rju!W`E9x{F4AagTa%+3Z2$4;<*Yl&kg4)Nr<
z!zkxCG1p|fH{#}Uj~h4F!IT_7IL*@XCd1tU?Ws0puTRk*GJjx}-qtS83FXc%jdl~A
zWi%QYomS4MSE2=uE>5vi8cb<9Fw-CxK`mco?&yPLI>T5`NN}F>mDUC>=#7dx%FSU7
zW2T_B4jVGzM8K4`TD+q4o)2p+w>C$-bZ3VvyQ4i9em>JYd$7%;2U<7>-@LvJ3;}2>
z;RA=}$#NZxJ+4A%>l~Q~rM2E*7XUrZFITmg=@B3XinvNOsfkw&1vZS4>XtFXZq(mG
zvL<=n^iq<GUXa>=7u|T7NF}P)$m=TPt~jPpNmg1?1PumVttt>MtVJVOTjGj7I%L4b
zOt;kd=HhpWQFxy7RXG}NTxwB)$~nfxaOYfQv5)&6RUc~#tQnCtx?_7P)K-0{<L~Qt
ztM@U@(o9`~ah!DBs`1A(C(qDyvtA2F?LnA)F4Xt#DPcJ;sQ_dKMO%kB@*r|dH#Ln+
zvDr61aYTjR9ayF5fARt5-oC>Z&XxSm>#JlNvkc0kG>oM&yuqFK)^WpanzIKP3=4Fg
zkvDR3!MpHu#_nLm)}0&7A3IK~)52vvN=6L&Bdih1!jiXJ=%Hb_)1w${VG2dwm_g5W
zuogC#mYG^Sz|abJmbOqhws)J%O*NUuCnCAva&)*P!6vv@yi?J~UmNQaN1XQ_c;yN*
zrICOLS*C-iXKCb$H*W3l;+<XAigJ%-hGVUa#~0c>cBoxR7Q@o<+UBt8^T1StX0C&{
z=Ey{aY%?H)F>9r&=sd6c1)*MIVv6rCh!?NJ^exe*;a!(8=te6aj}fJuD%4cs^Xd?0
z6@gg!&pw316D6G>&De)<j}mW3iK6KwK>CC&sa##g)`?8cdL+&hD|eO<4?jIg5HQSF
z#_b=HAjm!Xu*B<9&9@?WVaymuF=5K6(r#laIE*8mLn^|TN%G#omgIme5Z#zsoT?+?
zbpoNOiABsihyiV;B35;@p$=+5O@QiodljK3g4t|K<D6q)91T1rEsSGjcf@Q<d+Vih
z6_@D`g@5|J!#s9oi@&`*<Uc-XcqJR~>h~3`e8_MYtnSljX?i<rh$=a>I76$U$?)#9
zhE{Cu3>lRL))*R@B5UU8Mn;Cg>Ku^{N*t1Fn$sOdo0@Voz?Iu%4MnbUoGln^th3h4
zDR#FQ^$N<_8ICS?Xp}xo**Pc?ay$pKHl!SvmR8_AIG>1QZKSB49_zSxPJ-9M(JKtE
z-`e5DJKOa4aaN!0HhA`Mm;2_L)XJK}IpK|!-LbxpEw(BNBXT#TLhN|tSgtm52_e40
zj{A_~dsOocgl%&3qk!q-%3TmM>$G1jDYu$%T~*^iL#sV?>IkWdIyYT{!e8m=woW}Z
zwz_5P*^wCSA?;V!+{E|zj?rUt&pneeu$%lEBqgcun{4-p(UL@qN*%=+32UT>HX#`C
zy99ed95>oACTlc|sT--y2~UWG5H}zb5sTu-O2JeKe$B&vW`Zw5p%m<nEREoFR6xIU
zG&ATGt^xq-EIL2OcYd(L{D;=~+{LS0=st=*)nRb=8r{X8LSq<iUSs9@YjkGLa%{27
z>4P2eu%+#W;xiw-k8kIWYwz5mFv6%mq}j|ldh7tpt2-1$$@0<$rZhZw?g;O!ZPLsa
z$Ql_K%g)w-Bj--Ba`P_h%j;x1M>QOy;fPyzw>YK>loI+z&=LZF0)l4>bVE}bp9dFm
ze^6Rb!9rz?r?R7<K5Pw3UU+wdSMT=r01MxqUTE{|u_@+SUT<|(HjPLAL1}sMZole(
zx(P?88<f@sE-pznS@8D;&h8_>>fe+SByq+^T+~V^P4FTlKKIB`oPaT)rT)Q(X-E~o
zQdmcx08MEf4V7Spi%CS2=$DAJ;)AwgvjnL@JaIAzI?g#DPo-e(Pgm~wY0vKoyVl8?
zPGYo*+>~*bYvhjc23NgrJ&#nP5o&e9gm;WJ8lv<#`CJz`zfw}?{6~o?3_K~ZR5q0$
z!8%tH;+U+KW*?@utyF^1iHkAw0W;9d6b)62Ki#HcRQgD{Ru;kw1)9w!s@tYqU*z?b
ztGwR%3I2(Ci+}%t9sbEXmgdZtup>uaoZ_>e|7|W@zQ*tW(O)og@D#a%nb~P}@2*p>
z-ePC%I-mT7U*;zsI?BuMta9tt8n3;0k%i+2$#TU{K6RSwceZ)+mA7!lkmnh@E4!R{
z@Dy`XbG-aF-{kZ|5AoO|5AgPtB?db^4jfwK+~FyPckUu0baF+%v<!_U3tZkgZ@kO2
zq9`qOcApDg3tMgMyQ>3Uy0O94?IHWMg*@X64=pg)$+6bo>tl%ORd>Ma%e^tF;Y`Oe
zG>%+(s;peBEEFxk9<oG1X<W6z(R^s}*VD*9AS5YC0Mr<RUIRc3Ix3wHJV9VsnxIlR
z8cBAfbFLD@OXocbX)KSU=D1!R@vX4lDxN367LhDQI7W;bC2md~<3XHHHBW4c`_Y9x
zn*>R&4mqaE43w~TuWxvp&9^qkV>1@C;?XJ1>8@r<f+t8k5wIbi!nwK%Pea%76uDN3
zB<WkUQ$)fMs|e*v7?}tk-sUDEytPyEgI&v75RVqL;)#~xk*3_k;R|bnYtfO%M7$78
zH!`;RtOKEm5=x9*%$izJ8q2V>L0kx>g^UX~yNXg`!I=`*@8Q%8SATGp#sRqW@dG^i
z!fpP&Gn(6{=lS|8D@bpKTN}IFS-MN3)1f`pLRqlIkVd<Ko1UYiJFNE%8@+;^!H9lg
zXw1(tD*QIB?Tom)*~6(Er5osO8{M5@<TR4!%p5q2Ra2~NjmSG4ByW;88}+6uZwF8U
zq7}o!pp}qIsA{A<5lA=8ItNN|{=Icxys^ppXm9;KZEHozi#K=p<k>loEJ5VQMK_+;
zU*B9GRQI%%@W8YuzHv95a}?H*GFwuE8orMaL8^KILG&Z){Yms7wHO^afePG11r;C$
z2W8sZNvzb7i}#B7QeF1|AJ8@;m-F|>=TvbN<L0eua*@-8@UF!8SxW;I^6$YB_iGW!
zQwgOKcmneN#hn3Hhj!2V*No*`o0e~F7X0#J#*utnpN_~cLa<@@-=dT}7V4Z*uRuhi
zr=W@n0C2ug%Z<L_TU+BWw`MHQ?J87W^5|?{sb&#Zb@GThCZ#qO#^=RYf!IX>Sz#R;
zeZzE1voo?ZwJ-?bwDWE4^-brgdB1R!#xUHziz#8MJ)|7o;-5ah&fk5q&Ck8wXCZ6g
z&YtAC;Z<II`Mb#05>I~QGko}w)66y$b4|s{)`(XxEpzpUSGaj|mFJ9b_r?lh4b5hQ
z(O`(_4|x48L$Tc>pPOO0wu;*tk~bUdY!nQ4`!wh0=#Lz?SGEvqDSCbSI|D|e4)dde
zL*fgLScFa<iiue3t=L7a6N|vmSe&uEd~J;vZfw#wdjrROBjdxzrn!Hy&36)&;MFz%
z_p=Yo2mRZN)X@k3`qD0IMI8+E>|z@w+qLm4gp3v@a%)!uT(sIyi>*<2Zv;uf2l-G*
zI7)dUE&%|tsy1M-(&Oi5Z0#X1HhM`QqmF4O0J4u4fJl@?>X>!J8;L9KgwH9c44MIH
zBBW+f3=ZRDQFL`)dcamGTziV|lPyNP2}y24vR;PTxVBx`eUX2%y|`ugrA4ZU!RTk#
zNM)-@4UMst2_bjMyBSx<3OH8>l0}1RuGqKT99SNj3Hvx<fAVt*#EBS_Fg~kV1*1wg
zSDC<pZbqgRGfhpa;j8`3v^1rISpF(k!c0?<WjU>54={h|5ZZ0iJ@R?xP8Qsq8uHS!
z^L*fqZGQXU0UkPcl4t+cCupDkFjsD^b7^UZm#?ky()Ar~Zx>v8`#m;pyu;H^oaFC*
z>=X|_ahmDHS#~$JXwFU1S(xF;hmUjg>`{tg0a=c*mcjNe?b#`g9GPbA>SfwZ%ZDF3
zMzf(A^hb22+nhN(%~ZR=$XHsz!ag*Civ%%grlo1<Ao?p`m}F2G-d@?=6Zz*G8J|D1
zz;Ar`I8U9JVXBq$xd#vM^kS##^XgiUuU=ZO3|~qJ<-LSAR{LY)J+RQM05)2}9fUdu
z-q4q8FY+})a*k3GN|nVjY<?^WRLo~&hkgRhOvp2-YzgtLl+*^t7;<W<s}f~_rNk6h
zKaT*+0nzrd>PL4W2_c9z6SWqaT2<;r6!mqur!cjVu$f6!pCJyJw!LUUYNSk?*XQeB
zoP#U7`$uI?w%di{_AtD^j3b^)IB<$<5y5mcfQC6J@oNxU9tfm!btdNfsX4=tSQd$I
zgqujrwRrf*L_(g+_`ByyRqAfEYp4jQd8Jr}hgctzF1X6|1HOR7k(nlg^_z_H70TW+
zgTCRp*SC4;t}yc-J<H9%zR1USmU(CWJ+{|xvfaDQ`pyvDXwYh8ENu@lZbTu=kYzqT
z-#SWXDT<MIZxlne3rBBcDT)G96b$>@ltm9?9K%sTZ@5Zn3O4&CgZ?JvXp2!{Xt#2d
z5*oSobqE9U#0otwq$4{K28IeC=ck(d>W5EqwA-u}?d(j8U-`g6e(PiR@$|`Al=wm<
zDzL-PJv2@PURm`-06a1x!h4%TZVhU&?&<lqHxx>!naDV>=m0zw!wkTemLD4`7XSbt
z07*naRQnRflSHB$OiK-Hm3KgD16sHokhmruu?R_8JQazlrBaPKHit15O?3|mLR3G6
zNe8nAu@sqBBt$07xld&NiLv%WlJ^r5Qfcq|+do>ZOWU=6t;ED<#~PH;7?;j1t(G)i
zkcF{p>3+oY;{CHA<Gr=21YX~-Vnpz+MV0Jvd~ON{sb@$3_J8@`{C6Yc*cn<%>n;9k
zyCufK=Foayy7L|a>)`s^FY)}X2BY0Ivb>Gmxr1(;Vw#Wg-s(-hc_&AmI>e#tcQ~rw
zWpwxkM-G3KXYN14$@w;iW}7T_a%Sdc+1TD;X=$CemP&43yFob|a^iveunsnsSJ~a(
zrnj}r%={FkF*KV^@|k(c(TLrxEe@TykN(aUcQ*~ajcvqPoNlmkcY~(eV0z&g1M4|x
z1b`~wLu$*d5*QlKjWVrx=<pQ7LCH^_JHSWIEplk4H5M%5T&;g0!u^YFcKZc)LZZ*@
zenEd=I5XcyiSYHSo2-v&uJcpJrkQFgioq_;_Eg2CqF-B-PfBt`B(LMT<7P(vGA$<t
z5Mu+Zb+xxZ>V&9hf-^B0Cjd#b4HU%yB*Q6P>$Fg*3X<FuB^Z)%BB6?(q1MUkS_5@%
zLIhBi%)`21Qz%F1yjeYK?A;lA#QV0Yj|f+`OD^<&d<g(xse~t{vx>u}B7Y=?P*>%h
z#NpT{a`*^0sf|T~t!k%qytU8#X0>px6Uy9M@3V*;J!SwJ=j%zvHJ;41NY_h=SM@?<
zSP((A(CB2`*&H$7&A72TV5XZh6_gX>EK{>{=&4UIG9_2fe~mBv#IG~eZS#kJ-s04K
zpWy3TJuV;lV}9%HZ*t<{hq?RoV|?n|K{6LYc=Z&+bF*|ipW}~y|MzL#*+#1d_nke#
z(+?fxz2zNx{erviy-Taz<`W-3%jFwuDA9DAn%nQ)VtZ+aPk!<lp8wuOHg2ylckD28
z^HXfE@3678%hBA>$~8`Wh%wfKp|cjPgQZ_V=KQ9rplAz1x0UmAPaF-a=1Hzo%I5=!
zpfzZxD!oob`1}VB0Cy{!z$>f%@3Y6IxV%$)5EgRHeG4sYS$c{bjBYhT)b-;?08NYd
zvk3wwjOBq#Nb8Lo?R=%8)OCfc^BVj;{&L#K1wf3hu_!9WK%Ern?vVn*54Eu^+$?x^
zpZF~(;=57<sTQhOhqxZqgBTw_<+jP?_mv7<^7+d9AphSv)Z|Z=_xAQ&*erNzzEQ^+
zNVGN7HZsC`WJ-c-J&9`|2{a00R>T0Rg4ORAN*wpZ_al+Y{5WY%KqYZqlRa|vT;jDX
z4m{P;-bXIbD=mj-8Z<KDv16U!((%NhnP;r7uCcRY$=ZsO=bq$yZ=R>$>(MYvyz<)j
zC<{YxewA;3;^*kT_yVu&{0rW?{t2|nn9titsVR#-umA7{#~yf``%WI`_N`TVBg0?6
zbcKF@gc%H3IC_ZT&N?r>c8Q(sf<_}JZ{~>A%p6+e{10znO2^4ZPBZA2Y;Nq*%!L#8
zAEVvhu5@PSd{$slTJQD>t5R4?Gn~B9$P{I1kpRI?BvcEoU4(>1Zm7s{y+J5)#?L%>
z0Nh<(U9E}0h4pdW|HFs7W6)|`!W|rSy*rWGvLC2>>1xYy1g0^BsE{v6Dfo;<Pc!P&
zW@~H1RWwG~0E}f29TQcxvKmKDUKk$*7`+@o1#C9GA$*Ks*eXh0;Uz>uSh3VVStUZH
znClp?$ds&KiuijGg7o`lVY$%Xn*coA)|_Z-j^&D5#l1G5%YDm}^R=<n;oqwa#<sdX
zWNqmeqIaWqucvdW&pxOtGw&N*#klxhsfVu4@RR@-nz}9k5b8eE`xl9?5C}jsQ?+>I
z9G%SDt|P5#XaOC5Yh--<sR!6`ZH`Pg`217H`OW|7PZ(_v`P66rT`pc<VtIXyg@wPz
znP)%3A8dSuU%L1|(`;Q}^T`uDf9VFdR<>Ban$uRl$bbEtzr@3bT5JzWzV*XpzW%+-
zoI5_t11Ar1|G_r5*LwWfpM8h>AAgX8b6r+9b~$%shQ*HNi(h)3kAC<J&zza(_QsH5
zVK_G3Ksn3pON|gr<FgAb!k~mg98K*BUAvi8bqO2}rM-Ji08xO8$sl!Jb*AF!E`BLh
z<NyKqKXGmW3%nMHK<f8zDdB;GooZ}|R^g~pP;FS;Wg%og8e38Gh-^xPerdewqCER`
zE=JjfKpYpuMd!wJzu0sjVnbtTYH#1r!BFQB+*RiYl*TcNxe)=7dLw`OQ!5`gFlFK;
zN`Lr>UxO@><0U|BBpRi$LB#e_h@I4VCNUu_Yh2#<VVqUFva`SLf2N}`z}b%G*5<tg
z;C5lT-8UR*YadV+q?lYMl95ClsyVC=pROktgMi|zxgbGW=O~?fe;a@#s#0PAa^b&8
ztN+MS!IpC&f;1rh(D;+QvSe()bD1nPz$7i|mQS8sVAPn#m6nB0#(~2pxuR!y;?w~o
zZ*c$cFb8MbJafFu`3IijR`xJI^~xm{4j$x@pLm2XzWp*^xl__;FS0P*z=$y2%4oN8
zq|sz*c7|i~E#|s8#<t1l7dUogmZLLG7H66q>1GtAW$MTxi*p?^?MsW}G9Of9LhOI#
zw^CqCRnl84k7Ul-P}wJl(auh?G8Bp;L<gz>c-FJ3P_o+jGUm>oDwJi4Pd%`R0Owc7
zqXr+IYtd=wajvU;8MIi3&u^wrC&t0!huT%$0nZ@_(SfmWXL8U-#EgiA3DCr9G574M
z__c`E=@kCPbX4&KCK8(}$tYO(hapj_DvlWnt46N1mss!O^8+#4Q~Z@Ne6K%;52>-y
zG2WNtngDO^WBqR_;p|j~v5vD{&9^54gJQd~YdF@{KEI?|m-J;OuCXT?RR|jmb<wcr
z_C~uO*@Q$wN?IOD2TANiEoo^92lW(o-1U&BR^WYFQw!aO#yIF?UM+CWG1Jz}HWZC!
z&SJOesaoM!ooRC9*FVCQzdFzC3%B^C2cBYnUeg;5*)2Q*SrJ;zoIzn&-5H=886|-m
zwVEKBk#$_SxkazAG;&Q@loZxe*qBh`&kFEw>H-1K0?phXI8s_y6)2VHGK&REwW`E#
zsZ6PL1e?F^0b7M~O|kGu;p$T2nO1!K%=|s;_Mzi5W5zorLFY@{fJ)$v<vw4&wZoTh
z@9@TQpTNj{O+=~1zUaD3j}ugT>|K?ymqI}N7p_HBWDKxWoTu++T{PH9^(-wi7JDG#
zRV^1fSBNQTO@d@932@%)26M^(HSI|6iyYH2{+Ee+ZuCpm%X>LHz?0pqvOi2`iZji<
zo$(j@W-MxP(*E)Pr^WeN+>a&2?05g3w^r1z&Id?&gK=K!{kNLCh%i6Z;Lvo=(b)#W
z(lI;JrkN|+xu)IBIX>Uz;7k(%PRvg6(v?-dldbd6Z+)KI-@3@`t4ln3W}cDRWni2y
zt%a~#Sh7}w`I!!lT$5*t-H~Nvq21CPoN00D@HDgC2Cb$hYc@hf+mQ7aN`O05a)5_`
z^55gZB^Ii904M&tlA0*Qm6D2}h<6>v1rIwP&oGE*91uic>)`E`aqfS1s>ysOtF6X?
zxJ0p0hv&Y&)Z@#ywyXQTe0zs?H%DHFmtdPnMD0j4?l}=?k`ToEs_)}JI~Xe<5xc};
zBI3sjj{4b+%$Ihr!j-jz^B1WCXvTcflT!=-MT1sEMv&@nyeWN0=lnB%oRFR?JNp~|
z&UAIneFaXm^q&0&)^T~KtY%Z^L!>V<-u6@J$0P@diBR_j6#c07e*GhCdHAZ=59$Cc
zwHL>T9g95`6351F!R3`cmzMkVOUs?5yXfsLF0b@h-s*E{d7G;%yEHRzDZjf_@}(cX
z!&mHej{GMdrT^-6K0PWa8kbnzEkSFB7B&WkjjbNb8#}!C-ZtAKht^Px3f6W@-nlzq
zr*Nze3|j-k&UPO}$Z}0-VWnrtGF^q=+DHJz`}>=@j|>F75h@_C#RIHL5IQ1h)ZerU
zX9z-l-XYZI4&Y<u9bUBheuBgRk00uW^9E|mt#y8lV*<#V%YD9lbNs$%=31Pdi>@#k
z`|I#=^kq~FM8-K_vbT|Y$28TNb<{;z$9FqHIH|iw`$8Pq1|7DUYntK6xEPb}i~6`R
zcjKfnY>a5#(*orPz3$=e;XNiHxC$c<e@E7FaUbK~v4-McOV@|c0uRq*d-etIMM&pd
zh1OP+nw(^6u=9WK;q)%f`~3_6N%tSGTeau$m@=Ve8gU!;eN%-CWr*t*2l(OA4!5?q
zdH&+NJaqKq>@Hnluv75E_m+9#mA4r6@AB9qA7JgnM_Jh(uzjn;bIU`X|Na`Mj<51_
zPaoi8msa`0!rNR}y2*#m&GG%~yPSXbF5Ao3C{NBXEG#8(ZM8>l<vQO#ugC<3y*}kL
z=Qz30V(r}vTt7R{qsOPYxm~cl)~A_unQJO<&<hAGz-qtrIwLSJFjW;GRfw}yAq`Q=
z7vKoR*PH}TLK1r)y#xNb#o#mDG7*fkyuH*Ls}Zn}XFPnQOT{k&yOBV2taF^dz022c
zZH=w>Q*$kT=FDuBR$3irU2#hnYB0jcIB%>CEZ0?!@emk8a-&xyTI5wg&KQJ}B5OTH
z*njZ^#3uJggP@2v_=y0j7B!6po%z8`RAHz&Q*<+$KnSDWTE_^~>Puxu2m?gPp=x1$
z(%Oy2IxcOOdzNMcxWBCv`NFSVLkSPH^`1oI>d>-2G)(6zyth<GyYHokGqHXGA>X5}
z!sG~RV19%VFlmv7r~Rw)^P#1+0*rHH;Vr$}(lfPeWGtl<x(%r)wpK!K<UkyqhF{LI
zbXB~-t5=qJ^Tr0OUgKYXeU{~=n=o~f=ij`-=F(MW^DW*w|1SUh;6c9q!>ddm{yayI
zonW&!WU0T)?_IdT`MKBmT<<EcUivHk4>RKYoA1ER3Z{3R8}H3<^WAHdy$88+^%e%h
z#?7l}GeQI|{=l++a0Z*LaP8{bys|jUYj0j*YkilihZp(uQzs~m^(O;4j~q%V+L;i~
z0xIwsfxCsJok^%d<Llx1n)v?2VsXL9hpo~#g(NUWDYUaRni<Yotkb-)G|t7JINYr)
zy)pr{BJjrT-o3~_-{xmd&s3eXqK-QV8!Ad@1U8te1IBM%g=oh4@Y;%V$hZqp5LzKw
zz&dXfEM1a%9DukkeBlRDfuLksD*u0s_pj!d0I9`Q23P#7vD5`IBEwJBj=kELkuB*r
zVf)14Mc*sE{jL8G&E~19T-|xLvo`?%xVme2a!%Fm(YPLXTTvMatMk8Qi4N}9FTQ)7
zf-U$cTQhnedC1iSBOU;1!!xw!3Ke8cj^gtV>b;9OBpHA)mRlQ!>6W*I=822p#cMkp
z?E2%thoxm}P=GT$`^fzqn|Xvk|IVBI#y|K6=w^#Q{PqR@i+}iQOij1>+V?N;um9-(
z<kWqS@jL(c?{j>viL;jNp`|x6T)*4nYp*=dPe1=ZGALi><!pmN@ew}th2P}b)f=qd
zzQc20{U%wf&Ch)9GhDy9MsKIbnd6JxxqX-0>qCCyKln|ad*LEq{pJribYhXiC%bHI
z4)}}jUFEUqKBwCmJ0tJrJTcqwi)TC{J8Q8Px(&Z=g>j)2x5A0Vy0iq9^L7MhE6aCP
zXD}d!3(~{J_K+)EgR#Yb^7stIGYRl6n5ehEd1sfe+}IrJ|3u*Yv5ua#-h~#qZgfLN
zhXp|euz|}vS0|Ym7j<_J7PSG0s#1-LYEJ-YV;y<s2g;O>(~p&uJg`dT?~m?KUrlMe
z<X22Pj>%1l_p@4Ch4Zl(3DBAZQ1{YMCnfH&E;8QX{Stp3ZflyVmT2l*g>zFGt@UD0
z$>Czp^2|cyrZp0#9*G!hG1+30jz}Ft;HV9&O?~h4#$GKD?I?|A;`hTiR}tJ{X=#Mh
zv};wRCWvt?0M*JAJ44HrP^e;iXjs}V*c=$HuMX*t41?0K(Jxuu8PP8tGu<|)7N<CK
zV2Xu9hd6a;hDVN0b7WzT`PmMKr<;sao6ey}kcGn>=;pLDp_wZVOl6#yZSdr&S>_fW
z<h7?yaQwn8zK|;p9PV;>x=Ckxn)b{bWj;k~?hvOBPcu8;<-v1Dn4j-*`1nDZ(=%im
z7LOjGu`o+#s?Fg8v*fKNqp}RTtD@V|UN;Uq9P<I<@#QA5h-;-7Sy#pIEA7u5j48JS
zqo9)76;{TemZ0~(aBY2T)1EmvMW-2zN`VSTm)^w<`118}<bP_u%_q;yRX`D|w-Cf{
z7`FjP&AO^XQ7eucxnaUl<L8mFbwnPCIO!Z|BqEg|F45O3NQ__<e%}*W>p55V6Dk~~
z@-gGc9Jz9IkELU!C{oSeO-KUa^ZWV!&vwT(;4#z_uLkVZ){Nyw&qR<w0?{mzgJt5p
zv`Qo|>er3__*%WG-($b#lE1YlrdA2#dPO2^lQmJvbJDT08yh2L+Zjh^nhXldbW^iA
zFcik;sy}$7<BJ|-!l=kt&xPx^*15b6bJJ~Bc6xmG$~vWWT)OlwUz(fZ$V`ismAe#0
z$<E3u>$?R9x*Bo9u&|7b<E^D0Ke%>>x36Epd}NNl_m*S0_<er&j~y*DqB*=tVUMC$
zE#G-(h07POkvqqMV}}`RZes_1ez0_icW-atL@1o%z2z--cl*pQ%z?U%!!b0_)Im54
zZ>{iQTO}bhR(voFiVy<lb${ce^4t6R{FdMhpfs*JvoMa^FD$Pt?~HBQg9p3B!$%?p
zDuMGidVKTlIL17)*x^&>=ID*A2ddED1+wqI)^cttx;ht9Pze%=P!+E5k!eA+4{54R
ztwP?tUjspq7&b^aSd}^j*aBndP#Gj6R|*w%f+XMp&{f_EU;ECdyT%xTC>i8Il!-JY
z5&{bqTaKLX9u}V>hpUm_;^*<=K8Zik_rI6UZSkORc1m-8cklbJ^$aIFKCrqlt~yFB
z<*?3qI7+ANg0vR*vVwp;i057{u4$EKO&vj4t)B{YAp+1uDK9NZ8XuRnLi*`_^9_GC
zU}-r#orQxfyxXdweB~0!e43#(hLzziTf5slet3@8-?<K1PSI}h>gD&C%@sSlLw@F?
zPjm71I)Cu(ODxV!(bSs4ICh5xS8uGabnQiE)!*UW(JB7Lfv@nB>zDZ-9(kLq2fO^?
zOP5(}K7-j1-ah{-<*>v~&9lC|LaUK6JwL_8H!qU4r#XA}FvI?kt&M`i#}9CLrb%}@
z$2dj1p&1pHk%K0lo0m9T2yHH{@&=MBV$D&69czZO3azz1w_XeWSc8Z#{tSRD^E&m*
z?{183*3nLbBlGQmuR`f8rE~n?<~CoM;QUW6bolt0Ih^xHnGrY3dR?(_F6i|hk|Jlh
zwmIa!nMTE7OXq6UfSU4$RER1lCkr}YuE078)qZ@_+6Uz*JkmP6u`{{}EAAZOx<XnB
zS)Fyfdl7RmTqu4bb&(nH#=p{rs=v7U7_v#^aLIGW1&3M<NjWWWd3S%~-;>>JtcF1P
zUhx^n+Zio2+FK3i8$J|@=J-G^LSR@B783)yP^)N>K*fI5Nzg8Y4ex9D@GaU(D&t;a
z=&OiNMZ|4YWHprvV})iVvFNMRH@6##R_^_}BjXS3%rc?VRLr(C&4%XS;ykmRHvPef
zr|&<?o#i#&UD@L3>=c9LHSSzrVR3$*Pk-=!jviRx%dcGHPhY>q|Nh(s{>`iJ@^8L-
ziSNF4k=s|UbLQbs^ZB3nX})yz!(2LaKY#1y9UfWhqZisNoH&Q=%y9S4MRtcfjCOZf
zUAn{U;ymY0AK}`in-s=z?!+wZrpD@=V@KwA^wd1fM#jimvdnt~28G3nqIBMHM+g8C
zk;;#ailwmX;GiNPU_3?G;A<tKe2s)=2!YL%;)AE>Io4^^Kz4M7W~MNB#LTtg!mVw-
z^6t2r@bqGb&px<N1qDR_(&9O}FaPb46zYh%Z@jz3|9fq-3Kohr1ZuI*2PMTUHs=yg
zhSaeH70?86vH}`>l1i?^$x0ERB{8@qoF#fIV!VJ&#P^uu9Gi*&s65e##={DrT~by&
zz-S~oX-1+Kn_YNc-!(q->OR~Xyzd`s<WE1RYS9y&y;<|XI&St%70e|<m3bGnwrB{9
z*C*E?eQsR8_frAtx2_X`AZ7yAbqN&l-}q_7RV~0oLd!b2Ff3g#KtNH3T<t({*ZYQC
zdFrNt!Jy>k@?Gwl4#yW}`RljtaOTWOc7`Q)R#rK5-wBS*cet^#$xD})xqf$@pZnk$
zP9I%lZM(;Ox5;~}JM@RgxpwUUZ@zt%L2pFc-sQc<CSQK|ex_c3i+}X!JpXF>c{c72
znOZo4DRNw^$?V}H^adkt+}-BjgAdT1pXK`MfIFK*a3vc%Lw;Bm%oYWWM#jj3(wf;$
z$g;z8^gNUZ%QMBuSgaKq%0~${l%q7>?>};urV?-#7gCG0=-^Ed^5At=&e3jV{OYqO
z`Gc2kurVlj`qZ4?Fl#XuZf*?t>j}>P^nnhad0?^HVP7X-LOO5A5myn&>MF+Hcz28Q
z8{XLX)jK<gz@rD-HK<etL|tV(tg-<mNCXM=MZq<S>pK4Az{sRN>uSygH9JZvN}NVS
zR2S3=sY`ta4F{d6s-##v7cx$*#`@~D5*|6O4#aaEMZD){KP$dJiAe_}7?%Q~K`>r8
z)mFT?v$qlPYTs~nDy!6<_&zGEp#(}0XAvcVDCNZ0f@p+HA~yGFQTnL~w=eV&pP9Ba
zv|6fntxJdjWQ|-g8jXh_Z1xO|+;jXv;p^qMVn9r$$d!<dh75K_^ox?Vvb36Q8lw@r
zyIb5XciFyqhp+YfsO=r*XCL6{)3Yd<BiD-Qn=RIMMqIge6|;VeTiq^tu)?=4uJP^Z
z9{<i0NBG>GZ64bA0q>quEUy^EZ7}LTN}nFZc8gxSK{Ly^c4eFG{!#4iHnz09x4gyi
zg)TEdBNOsmQx*m*ip*IIibkdwj!JCs!RK00;1h<FvM485=f(bJrhS|M${$4P6r<80
z4qrqydIUleQKyyjYacqnOV`%D#9*xt0K)REtLtOch8~}W&pmXI(pX9p&L+$?WoasR
zTo%r1@POrYzi(XK<h6})PeUaVMx8H!R?N1GVrpr@hGz->Bw9U&NA;H|C3yE|VzhE0
z{Wx;Ci2p7GUy-=~B%Gd<jBcu1qx&+ROW@bCbmJ%<?H}<%Jgq6tg)mlS&jo(4Pdq=s
zW8HD%UaC?=174&7hgzDWnc;T1w;J#<>uBkkpoNU_s?K4G&;IxN41}}*%8!uzt1b@4
zT!)dvn@S3*IX;ZXM2sB>8|IklAgZnP4DHMV+;m&BJ1{Ko_yC#twq|Lk^d5m>$--2J
zQ|C@_-%OML=f$`AN5A+CN@x7emv;EIk3E7>885%t=DF{`$<g^4e){2qoH@|)S!0=S
ze0G8D!hw65|LyIIEMI&LWS%+wE8KVAvpj$ECC+cXz;AC}#x0)UzdiCSvUHc}L-(Vz
zjAG>uYj@ZA^)Gyezj^5$ZePE{snbU}eqf5#jXu|xHaO58;FOR2)0sk9hdA%gx7O?Y
zy}}=S;sg9bjQ~+tEQRyBuXmMs1a4&hShDGsPe#JI>I^+mf#T!nv~oWFfdg0@vf-R#
zePG7igY&uO!-q?+K4?X*T<x|~Dx60MN}@5xS7v(d%Du?{xnt8jcF5ZrGM$`NAAryp
z$Dd!_qN#+F-HZe6jFZ#(7}txg%os;7X=`(iD*9*aXmF{&MB`c1g>#iiU2Z(q-9N%-
z|D~dH#o1!0c6>|<)KNW25Z4o*T1B|B5BCQ+GgVvvV?<^uXx!7k>A1a*Hu3g$$rE#V
zFv|J$sSoJzbBU7ASgpdS%csk@&-2q3-><r_x;NGXa8+@TMvQEY_f~eGdc4d8t8llW
zd}$`-4L>t&O}}*XN0xR&u>cGkikq3@&c+rOwr+8IuFGOG<BKm`qCc{X*4Fv6?_a@X
z8H4p@K5*_N{ncgu_*)lv{LC@tx*2zNO1c^@-dyFoZ@kS^tIOdBPO)=$ncnUWZ@>0^
zWOR%6!Uh*U_ypba7x==7ucAJh^T*$Nfx*g2W@j{=?jm3R?(2+7;rN*&>}>Dy&ejfd
zvt1rMdx*~Z6_i$#D3tTo`%&S`ek1;b;7${;mLlkzS*VaKqEIRn*~ABSw3NbFOA$=#
zI+(`2MC?5Rk?MOq@kYOxTy<sQK#(U7cgge?rE?Ychgt<FO{`@f?d!tVF0b?IM16tJ
z9Gm8eBOULxiO&tjpU64?^y(J3haQ5j^b6Wb_@$Fm9PQ?nu`dR*SVuUQJ^(IBGF0?F
zxDXDF^MuO9e1rg?!PZdMF^DG;NyYJF1teoERsvEd#=|{<KZ@%CnOtOv0Tp80#LA%L
z)@X0B{Y+CalT_Y}>V|Q(x1&NBynW$J2fnjAIg)RedWI+F@}$yKo$ecBC87!z!?ol1
z5oRJ0iqG7yN$8)_XBXFn8nM%K?b>_7SRs};UZ$0=ov(ztlCAPVBDsROnY^-?w{t}^
z%V;zj6uTqtY{4%+b(X*U+V@!P4VXKAjGJqlh=XpU!EgWkNBFB(uke@Od4s>YaEH9n
zM6!&cG`N+UY;WD+XFvN3{Or>w_~NU#xb)_EHs5*~J$(Sxe}Lb+_ApPK$oa39-(+z1
z6klE3U~8pMzW9?ITR6uXfAbAK@R85*k;m`j%8hk;yM69EGRNabyKD|}M10obsPK8~
zh(dx@x}m(a-CDtxh7w_SWM~G1PAm7vfH~{X)=~tvsFgqDBW@^o-8Mt8KI^#3>&~<a
zm6jZn4CL@k6EPlOEUpUW%|fBbJRJC7f;#|wK!d;TtC!YzeF6(92e1GDAOJ~3K~(ua
zeQcT!9BBK9#eneNxVFi1zvOS<KZ|q17cXyC$Pd08`LCRs;YcS7!8jiHMkXR<jfDP?
z5ug7b>mEb^)#8aj3_gkCZxrWoJ_g^P3lPJnBa@8Q^7=@uv|P1_CmyjzlcK)Yt46-x
zq-#>d+fFWT@6Y|ucCwi1T@fvaReZ-I>Eh?<49+$cSM~@fx?NaShlWGV`e^VdAw^w1
z5-sBUhrfC7h{Ql5=fLbuw~}}CW2b~72319(AAgLJK&8^ic$H?Y#7h99($dOQReUn~
zxJ^YaJiw1iM`<jBon6ez23w;I{$zcf;qEq_*@JA|x<M=RuD(J!{`~w+&R<#PnP;Bl
zQ%~H77O!4d<M>~`(qj9KO<sEE8hNw9YcIb>ueXcLodjth!z<{)5Lw*hz0aSc^{4Of
zkDoolg~#EYn=kV6+ZRyN8P~5~=Ih;Ac5l8*e`AL`$Bxs`r&+KT(Ov}@Sf2^l&7gs1
zYfxgupLuKn8>9lAB?oB-AFjop6{y5wjPI+hkEOO@Q=8tGhB4mmfOQzLb)=Vr<MVCO
zw$UqDY)5@fXa}iOls7!ageDQ;>+jr){GT~K!y|_}G_+5@Md11C+m&JPi<i8yaT57|
z`P2-DyIF;#)_KvtnJEh6Dg$85oE4=))!90t&{;1PWr3qb!=O$$jKx)_OyLMcO}x#d
zA_jPA#X(XGB++rB*qm6s(ulok_z2U^&++y5vHssbomGay$}7UXrMRT}&$qLkj4Q*v
z4}ZO~Ybv8+<gPW+jdSeS{G4mmNeG^wy)%)kF=OezI7ZaW;_q=SV%T|>>`D5b5{RD~
z8CPY<mDVvT9o<~f$`qr*GAMnunCL>tGEH}8n(ovbB8v7*mu9XRwc2!AT?*?c2b+|p
z;J{Ri*-neYGc9J@84AZt!=FDOIFuQnaoo4iCY$P#&n_bQ6y3RlnEXlF3#WPY>Ouaq
zKfTJkpF6|s_m=qOci>kadX%013PvxJ=MJ(At8$#yG&*hC4NW80-hYnJ8(Cj?!`Gpv
zk!uFVVM8jg6(P%f=%f{4XdUaLP**MUF2mCK#G=AN&w47KXFg(3#77nCSbRS~SPaNf
zBAi@kkGTk6ytPFv>X<2^2prQuRT3K!zHw=d^K0X&#h*Jl%Tvd?RfU)6GJJ5MRT*n<
z56dy+Zz<tdPS0?-n+N2UN@{Qc2(wuJTfF-(y5p)W!-})Xm>xN4?N{NQ<7qS~s?%vA
z5Kg+M;wx8Hd8w0(q6Z^Io;aP@C$^jY{eAxrw>3>&i^tXbBx8-WsuP@>-akX|QqNZM
zc*^OmOAVfTx9ZtyT*BI~#rh0!PonBpSc`G>$=7M=ut+Gz2vGQ^M%K|Ue1M7)&%sLL
z4GGgtMKe=XwU{hKm~>lB2JIO}u0tcs>Gk{QQ3;(Ez5aj_P1fkrYUIpMb-22`i!qkh
zudnmQt!=h@BX;@&jIkVDnBmAwi^B&G<C-o0KjPjr#<C>4@B2l>UEZ2oWmVQ*)z$mn
zJ!=m+L(Wj*28t3i<dBpl*^(^VvSkCdYzT%d3;gB}f<Fi{VA!xBh_GatGZd-CB01!6
zMk5YqhO_iCy?1r3T}$rkTkaAO{%|7heVJ9&q%A<cKzC;4yYIex<3*e}=YRg^e~O}@
z==JIM`V9L$s$v<pev)tR-R8+Z^E`VuR=M@sHTo+{bcSQr4@UI58H0|;E6c=s=6We0
z0-AZMSsbL~nWprzz(q1#7Hf@QAC*^(ER1XETi4Arl}C)D7M_6Ckt3l|T2olrIhCgA
ztv9O?V6>L}xl*7spSaM9glw0V-@h%tr-{H|NvNi)orDBmy}a4z3&!Io&&~7v=|SUu
z>p-qiT5)2K@iQ0Z@0m0mt@!-Kd5-tf(0HUp%%cIh|3ss~2x4q2lP=-{9Nyv%P3utV
z%#2$dPMs}1jvkWp?e5mP;;IF$ycn6;oq2XrN6-1nhe&=$>hJ6hdNl2J??}kmu|n{E
z$Ch>6n%G141&4R$_%nJcQTuR~$n!@D@f@+aEy>H%=>D0Q9ijCEj_M3XtVIFlzy4c)
z=a)01C@qW%aYo|Aa9+N|!g@+4ZH}+M^IhJao?xy&&$*QaZf=eE+{d5ciHj@T+SujO
zFFwIz4?n{0c+7j3ud`EB{Iwr{hNsUAdFJdq&z~K#GMDr8+2fE8dHvcdo0aD3`ZmVZ
zeD)L1FxTy}u{Xl*Y+@&OIQ#4mv$=hnalgmw{bjQ2S9oEmV&mLjU~6?uS=LPG(Cwrg
zTVCSM_Jm>e0T&;7lFdTU=x%06X0{9KWT9&%bpZxp{>6dmq?+2roV9oZwfBdLa-vp1
zM08DPuccIASgA<15#zct-(~-x;Lb=~$+ZG*Z%^qan&ZQ~(e%H*HQ`^qyUFDvMMXb#
zeu3vs59CX9v4KZR?kKF|%hz_<KcakdB1?F2rArdz#~{#GwQk+%u=Q8P1f8TLCOK3h
z?0tJqoLr4Q!OW~|mb*I4OjB{9su{p<W|xn>CnKrGIghsmN!B^Cg|^FJcIQ?~8fScE
zv*bP*J33nN<0tY%?+y0lfrl(skFyq&WGLXh2dD~XC@utQaHu42@QkKK7i~#&6FRBZ
z`!QOvUwPhsz-KHN#bbknC^8yTp^Zda7;884I>BRz;Gu>fU?b0$IP00#0tTh2rjx3s
zOYr{Gvd}Y)n<~;zo^ArJ8sW=rhSW^9?$Axnk|zeYzs|vQiYX76>-VT!&Fa=3m4k(j
zVbD!k+bg)awneqS%Bh7ebEzW7k|zo~9pfuUl4@M7=**oaEgW|@7WjKN-=s6_@t2=G
z$MLr}c=7mK@Z=6xuOuw>2Y786Z%kQu=xJ0bl-6VYgh^qgMAK*z@0-5t%1YIgvyNJ6
zwD<J$gtD^2`!^yeYMmm*hjz}Q1&D?e0km+ERq!F_@2qGID(9QxA%(=nf9i!Z@ak>8
ze^<(XukKB``sS35(VXgK+!>cl?}G&W<b?&EIyn%gpmPnCS5%A@6YKcWJ6qhD-n#?d
zE^2=L@-Cl$WPvOZZ^6vOlGe4-P(Dly8{BRYO>SwzDy*DFmi*9`-1@t5ca5#cA`n;w
zu+!5{RAzRmrtjd<ojK0nIOeqXAC17({rgw%0X)$)ZCuWbJ!wAR-aj_i4-5IyAfdB$
ze`@eM6U&c;XQ`b(<W<=)ezWoV>_=K3$wU4_ML{CQ=MXP38uM5QILk7O;~Cq7D@Cgi
z?OPzy{I&n)Z~w9adZ}r|C87{Wo@ly(4<9Z5=8f05rOxu5ch|Ue^*w&#v!CJjzWgO#
zyRycK)93m2cV6Yzd#}<T4){-g@lP>znqU9w4|sjo@~z8j{Ofnt`0Y3E@caMfTYT`=
zm-%b|-Y@YNf8-II>Eov(zV?-Gvfmlvo#!w7<V)D3$9KQ+ud!3;77p*WxM}_*`mu-j
z{N-O~FF!%?=+Cfq`!#B}$JvKI%FlfCJZASQonsG^rYSpPD}*Lmw-*916?o{eeO1ik
zMvAJ@(rPJgsyf845wR9w1=YmLaZ)!byk62h4dA5U(GvrfQq7&62@zPXy=SMa50PNv
z@zWO<dG^elka|2#hM*3gF{wPi_V&i1#oy72A3Htd_Nb!vux}k3ql(9uI?cqWeH)Kk
z7bZ<TB{3OFC(Yz)bpTqAe|8ebW7gG;C(80~x5KEo7R!sH9D40q7NCr$!_}pD0G`8(
zJ?@}K?%~^8CENGg{(+Cqr!40B{tLfx!S8*abr!sM0M1Lb?Am=kqxNt-(JbYrNd-rD
zKc+{c2NW55MoCqs_Y(3IwPUyPT%O)P6%ajuI7UXbYEH4M6VQ;n{LJVyMMNR%K=c@1
zzdQBij1tNA#0sh$j-^3LFID)kn|<fE|2Mw0{v<w|XYKY4*6(ic%oAtGx(j^o)-Ins
z(<e9Z*8A_Xd!X<t;XnJi&+z=kC92vree2%6y~j6Rd4=!2zr)8~c$Rl>Z{zTsKh~$;
z$@%(Q*HKl8_wdr==Xm9{w@Hnqm`)i_#>}0(!Dr8|@YKyeq|7>eYq8>cZ>=&&9^&+|
zALge|O?mFqKZjQdxq(cH+Oc<@x^_5g@y_F%ClekjUKCNNDodgj)6&Xrs|X^Oq2-g0
z=X>!KWQ67Y5UwMXV>~z;&v7Bu<FQr6-r5zm^(pS`PQb(Nq~hiE!_|a8b#al8oRhfT
zg~q1MJay7u{EM%zb9>q{-#S|H^N%cXa*%RozvS26m)e6bJiN$>UIepz$O1@H$(VJR
z9&BRL9$DIH_t+RK$CO{Q6Z^x9W;VDc5^Vg;`81O;Q4(6ktW>DMY><3wE!A|7PJf}j
zygzJk;`Ko^|Nr65!+6Q;c>G5f`plZHrU6N`O|y^T<2g`QHO>~8tPjAIgNonWyni;}
ziLT*Koai)RAGG%6LD@Z-PHo4a4<<GLZ2kV%znE!$=47V@|AHa#glqU2R;E(-80T3h
zM<M4X&dW}}KC0QB)S?`NSmbW1+1xMLO*Gbe7J3O)4bNXZ4(S|k=LvU4j*ma{4E<il
zwf)Qd=^uWIy-CgL*l_2Kx4CfX9G`i5g{4kgR;v|UI^Jbr_$j`A^=o|Xy;T--%?l5$
z@S{(k=JxslyHm%j*KcvG-{aHIUSzXQNtCD8>#(}I&W*Pdu3!8Ny_0MF<<&Ro+j+j7
zbs3eL+`RHK<?=-a-Gss=^b@FS8{#5It~3RJa%4tQo7Q`>(EUdlBZMHWF&eyeC`o*`
z<R2wnSd-W>NGlW)1}Hrf4$OoHXpM3nRB#a$JiIdC;gtb*b|(De@83G~h5z(JOT2Jy
zzMV`VUW}f}xVm@oceUd4j~?UHAjRW3F-ZBs!;2^dD}4b+WzvaHsA$=Gt@|!)KkFb>
zGvpW*J)L%v#S|bj?~mt+W29zWS2-`n#IqR#?_<-l_7ZJ)5ud+ZG!Q4d-2aDxe@ltt
z)$RK)s-sOiW?qn8(pm|JOnIoy*8PsRCzc<o9et^JZirBhHvW#OVK|HjxG*mFD;AnP
z;!f_uxEqA6>Il`8Dt3H6?72nUs==?q$0XK!y!EUM68gEuIoO%hEc6og#x>^_^M+?^
zl%ki3sKr-byUzOiKcG8zj4%D}@AKFbPjS4f`Fp?dhq&n~o2%PA{=|!{mzF<xcZ*Lw
zx=1e(VB(;#ym@=X-}|Ru=b;meeD?W=c;)&g>jx$O^qZI2-W{_uo$!f^CwS-97Qg-S
zHEzB27OuABiDjD2aq84be(=pNvib14_|qqN;e$K;mydS%Vlk#O`V9Ty9#^*t9yyUy
z)>f+G@YG&0wE-m2nkp<*(XA7vV1mOcjd4;f@066odO|GfkhWq<^8az+f0z-7|3)hk
z6xGZk0?$|9-8fVy@ROI0@xu9ep&Bdkz=*((cZ`dgFTSyH&*J}u$B(fx%<$e*)l&C=
zav0N(vEgeQ3pO!*vlu{W-GEnu!uy#uq1qMLT<1OkbU3ed3fgOUM4vzrRf|F!^_lZ$
zk?S^$Xg@|pV}%%yH$Fu8<96w|UEbdS{PB+m*A~nk*T<G8=32>k8;S$s&fL`UgBXbz
zpA#E|J{SXl7)pBLK-5mOGZg`+gXZTw6z3zMjt8{Xk_+D}gzc6mEn-^*nsA~;Gj3W7
zM}EHu3o0F=%ZjX%Vsygx?ufPRUG~Zx=QRuST_#m23^$;(Vga|uma4Y&djqz|6As2x
zHn#S-ySdAxu%v0qe(6Z_4kk@m=w%c-qgEMX)x(<<{D|W#7r1@L^D8f}@zpbikM0}(
z`tdW=`X<hm%=c4vO2>X3(sNp4jI=<v>aZG+mNvB#mxi4vN{H)_%Cm%G7u_PP=<x3t
z+|K+SQ-ZD9aO4T-=5vJ}s`=B;pKiJm|MbJleDtEE485B9*21r)^|JUU?WFE%#TT9s
zEreJu+mr^+AW!kWXXC@@r^d5l72+Xyf7bmzvP-Gi*){vCQZtjU{j9bKXy)T*V{s1I
znDFQJ#o5;tur{sPdVuEtzw6`5xQ@dWVt={shzUge0;R(4-)7w4KH%h$BYNi$Q-5u7
zPaNwQg4|<F{JrCvD1@$ASW%d*J#=%!xRBkd2<)&<0{zU;F`DO254pMv<++Qj-MPb4
zmmcP^a|_(uI^g-0Wd>Qo8#j*e>WvLj1&^H?u-H#1YB4C6j!9pES06jiZ@qejSFUey
zcUtqv$z^`%u@!a?3cABSYr9hhdB*b>Pq4Z(VqAH&HkdSFSfrd<Im1_9exGXk1HN@f
zaq{6gK7M(J7whj(*UvHR=j;`Vm3~SmQP|3Yau{V8XhmH+Oe_Mj4(&aqhs0QvF(H;0
z`f!>^Hl(g>(C~Mt$p46ya)d0w13ZBk26*e5J^<Qil=sYaQ~vzN&+_K2Jw9@AQ9KSU
zwFw%j(eNI|Rn4!ywt9H+8_h31vCPUaJ4C5%;@dDAeJGc`zEfbm=h8yn0&GnHDMckF
z(2=!@yLYQhH0$O!lSx5LA<WWQqTSL;t{Zg&qx;+Jj*VL%LK3a8_c+|)#2I4+AFlcT
z@A|k}h!|}e9^9B#KQbQrNLVmyPt`roeE5hUbId9vItbc^1uvnNI#drZo0U0wB;{z2
z+IhA{;^IW`U{YHKdCINTT8dRt4Ry`9DA3a#9$QTK$wyD|d#_FT@yF*_9>O~dmd`!f
z=eY~V_(#8YmoI<!7SB9+iGiB1vo)255xy-j`YSITXKn2kZ(M$#OJ`1UsXyU!7gAoo
zvB&QAgxh=DTv%S@sbx?8{0WqTi%T88^qm`2-37kz$usP2zQilr6Bg7*czLttmC0B5
ztD|)e>;&rz=Cc&#YKp1DR<$fj1emjR=#A#2YjN#_zU!SwIjI^fYex#wHq1-)UsZ|8
zb`cny{TZ$&CO!Nuq))vkgc1r@7kDhTu4El~Pi8#NogLunDgbm&_ypcjp_mpmzxso9
z?o4Lfexv!qV~d;^7@V^Ktfkpf;yt~7cft=gC)_R$yUM3?!=rN<&n$O>IHyBPrv$vS
zO$I|LFAMxo3T$Q@kAH~<XvW8I$p4g_Aojzzy#4^4Hb}XI85X?9xpr}j8JeBN@)@u0
z|MA@Oe^<v_I~C6@CPIByEe6t#i|pIc%W5%=nNa6GhXRahH3!v<S7vCNmJ&R;L*ZJ$
zB;?l_Z|WcalmFd2@0$jaM#mcQmZSlq2vb<gbn64M;R&41=_F8B!lBn%F%9T>Y7A3R
z7D0!O!bcAu5cSG=Y;8dWCY)AHy|)X0Pg4`_O?204C)O^}in{hF=cz;4t+zE@ZQ0+v
zMP3(NRR)TRG@0YX(phkhy-`W3HN|)$JAazfOH;guex5QZ9jSt8DU(p6=yg&O4@D)0
zA=9c6Vu&d}XIqrKwvGfL4_PZaU#i91a31QqM%xM9?p(W(`VfaV>Y8EEg@yKO@55@*
zPLPgP{OnoB@vdqXV%)7`#^Ux!@%44Z&L6Q!b`*Se$#bq#gqCT^kx9!vi~Z2-{V?D4
z(7iS9Y)phYe42PH-eGNpGFjuvG?j2O9>cY<;<tA{gh2oQ!*MEspIvI1fz9zp`S-w`
zQt>Z#?*BX&Q_oK<{(lQ~j*c;aF$u}zpZ;@nQ{7aAIS*T7sZ-Yq<~l|esr959-upMd
z!o#2VIdpf30=lW7kRSylqsq}Sii6V8&xJbckKD5h3*UR`GpLkeQaOwkoInA+R5kb1
zONBU6IU$mC6H)9Pm5$x90G98sPwD4|L8i%4&AY$*KeO}Q-()veTv_(qS^G4<^z@J8
zUCnpjUggfUn;2Ko(T?u?33_?LTtDYze~u3h_84{(Mz+KjB}>Dci^t~JoL21Z7tD1G
z-BhzOmx(Zlfw@%N^+bu0*qzkOXQH+=p4Q|>Cib|j*_+hV#e})(2P~g`QpO~t3Y~K(
z-;zfZ;x+u<TkEVAt(s8RXnygT6;90scAIkzIY&~Bj_=&u<7;;h_b7PKG4=3Un~EP>
z>hjcy0eRL5*NTNZwB#RucrnY&^7r|tKeYJOeO_-;d$XGr;H;&X?2>dAnz4X(_lhy~
z-Tf&K0_Z+AO!??i)_m@46BF-c9!d^fOiKjlW=S-*s&KX>?exQq%a6i&{?{A#PdnbJ
zp-2bJg>^Mc30B0)LmI3t9!i1sHShqyQXA54pW6DSu449kh~?Jt`3_?OhkwR#PTpIa
zaCWH^yjI^7g(P7xHuft99cj$DJGN9_T((}KsI2GRwF&1JJ7@)K`xVCqX~R+*myT|1
zJ+5gCfiV)JOlrrS{hCw5gmEp(DuY}zk@{8vJVI=0YtR1F(#s7W?3OI{QjCH@Ct+MT
z9y!rve`*;QmVR#7DJ|ca)BLAvQ@+z(;QyLG$<{%|W2gGK=>gjt>#%=|(~rG?Pjgn+
zHt;LU+*!NL;^Kg-SJ$w`1nnVBk5f9&yI0nw5ytX>jXlfNJG(3mIy`rN9tRW-YCPlG
zvp20dn1Z*CNohH=kTaT8Z0r}@Ufm}*j($fZ$m&{Bh0coJK&sjtxAU+&uK0u7hZp}Z
zJhQ^7`A#@TzSXGpZ@uSt-rK&Xh^V6#PtJFEWU13EytPrq54I+3RpKK2-j>Jdf@e?W
z4IwN}v6jLM-bacw?Y4ZM?0}Cjm}(V?<7dYxHO9h6FQX9RRaLvDzkx#Z6Kq&n?|ewp
zkK>tqLRm|icdnBhT6C53;^iqvKQov>^{=d85NWuMbsp_P<|5o*rXJiq2;lvLn&;-z
zhOiVNmZncbG=8(-pr>sOf1e``bz-94qC7}Ut5z;<8m&0#;qOC!Ka>W1{8XQ^_H;XK
z(MM^e3b3m+#|8<t^DJnE?Ivt1*_A6N?4`2{Ifa#2ci1tk?v)s=SnL|+I;Lr2QIw*n
zqqLM2Yvkv0UyRE_t|_f&XJW~XqOuO96#Y~LJ*Ty2cPa`#Cx$7Nh4oQQKh?~2<#WcR
z<@9_eLY4;B(-B{M^kIJb?oECo{Vjg=tG8if>F6skJxh|Otlj<{@{MO1ES%=n2dnf4
z18!ftMV2U%Zl6Kdu)e;--D`I-Ns1ruv2mrKGgx47Yyqt`@2-yM=LwrTW9r&5omNz3
zO`aJh`xDaCu)3)!i&~g>#xUC5!}J3%W)x*Db@yqQC<!tlYVR6+CLT{5>v3|vgGFk}
zB|Mjq$U4vOyt{eN;(v0!!)MOVF${I<Qu6D$xRCSgN|)ci(UuZ_ZKLG$e9eiT6c2Sm
zkH=Yxt5UvMzz+fj<Kg__T_?Kp0G?JN+<M7tM?|}FUa%Uk7*8tl%rv`cY7}qnKA8I3
z(TZb<Vs~8Oy(di#lS%|yM0rS<cc`{_W?D;lHm$5I(#&9OWZ4C@UTzxaJ*RX1Km%}P
zVtIZcJG2YO=a2k_sGZ;u)h1}U5)V2wQm!96VH3U=pb;gBhQ|Ar9voAG;r1mJ3nZyl
zm^5~YkyAQ>5-dUxG}6aW6F*NxJy#iNIkk2m1fqqGk?J_rcCRUF$NscNYsE0YpX17L
zbz{orLCKFjwjcwdMTK~KTyuPwiuF}8NJT9bi*wyXtgn`^!zk!wqRN{miVKUm?1(6|
z?lB$x3~vvf<4*6_`5StRQx{+5_Nkn;ea%<Cw#YNje}Y@@{R*zS%b80rasB;UbQg#0
zY_D_S^l|7d$^egd*j{^w#k0>+)tgLrS5Z7de|dp^Kj+rQ0ec%e*s(13x~MV6kaRmJ
zXBqES*zuSo%UC#dj8n@4?$lG1DltK3OlwV1i^oKCc|9b`aO#Mbz_pzzMW|yC2{F~M
zbzZ%_$16uF2wytU=hGJ#IGjswi6Y&^@FN%ID4gf5y#j!5-5K#`A6*E_KVkP*F(hag
zjd!y8TN}_WXazwYOkO|KS}aW7m@2D+@JKV!wS+OFQAw5<s-i|IshpIeEyeW*6wLrU
znhKqJv@<3jh&MBu3b<;bhM)7~c_Ocs7V8|&M$JCQ-hsv0lBM3Pf{<^t8ZQnK*tviB
zal3RJ)b5b_kc{K(@2U~)cz_+d=Mbf58d@((H6GRMwN>~o4%uYYf|1U}Iu)V|(Z1WG
z=or-qKIrhe4;P6se=ig1sJV{eps<W9=_S}nL}e&WxQ$WG{!}^+eXw1yu_xSsvrF9|
zQzR*>PNF!wkddjVOe+(qsy)kH!%|0+8c9tiiDJH+kQha(Jdx!czu0~!<MTiM3;cy2
z{UqPp{WSm3z;l=P`HR&&oxyR^(@*lyL#N5IPm%O5^3=s+B>kMTC+A5PmziHa#;2b-
z$>SH6==T(5wL^btj<h>OnJ(mg>e{ifJLS%m57_(Q2D_{8uyOr0c30oRRU?#E)YB5@
zEX8D-YBI(<$I@KR@#Ax#2%3ckrD#f&WB6mk^9(vEFP&*AvD;<MufBYTN#&S?(ruJt
zXHxO?BdWunJ2S_RJiH{Go@25-enyyhu}t^VXNS%Eu8b?z#x+^sMhHJ5;zYi=$8jw^
z6ie%w*3$Gvnt()zFGx{EpH#KA>W8+)`{by+_td4h_}=*-*(O?vh4tJlA8Y`YlTc=@
zMbbS@G<9hy4hkj*B4}IJvO}guQxp}`{er5fNWvyo9u(NJUEr!&5AyG>(GM&U2Do}3
za?;VbKgRO2dHfGMJbngvZY9O^5!^7|W6U%h4GFB5EY>}6qtY}*Cl8QS!#V-)#zEMl
zcH$-`u>+P;fuk&gzdT48c8rLErn+%KI!Q}g1P(QdjJt_eGmR>g^rDXs*Ga`S74M-{
zMvV`+v3o$4Xec+>?~fRK;Tg7Gzs_TKx1fBLS1-TA$i9PP%=N7?<+SALt!>KDE~{Jn
zynb_!yE_G=sRiGmC`+muyh=q_q?=KcHA&Im*v$X{AOJ~3K~%4cPjk=%l<9%$;!FlQ
zL1igkX|(C#(Ae6tH>ug1l!A-I-8A$hh%CI=v%rVN_d|~^9}*O~Jucas*3wlcEXHqN
z-#x@8d~vzQCoat6k;R~Bi+`kWDSCC%*I>SraIw=)lbd@1j0)FJv<m7-s!<5)q(Ya@
z65W*Y3g#2y)OKdkMyK2}(_;?=Pn#D|O2}JHzX2`6EzW!1-FtBAFH^9bN+!m;D7h6r
zHU#oIjk4e&qm&p_(0Q-{pjM_O|Elm{alr&i2WLOT09+2?!O_Jx3kt?-$1(k*O#qD)
zo%b^|?fY1d*+0e4isO)I-AaN1(%r^-CPiQas@QnYQ^vaZ-~#N7OVYk$S~)UfBu2^&
z&cP()(o>`8CBkj)2d`kUmq?Cskg``;Y}h$cEgbi<ZuQZ6sV1uAW=63ws?kc3CIXUq
zLxU4|C-I{3kfa)|B)<ED%OCKy*REj4m$|jEz~6Z_=gjG+c<I*b{H?`JwiZ6%jbnHC
z#@BE2-EZ_MrW2-<34Z%7SN6BL`tB8SEUD4VojlL>2RHELIwo^W_tx1|hI}|+@yrT?
z3n$p#+-I(zva>Ux9#62gB*`;jJ)T^lo|Y7~=lz@8RD0XF^Ww?G*gC2J+YlFPqp`Kf
zviI_YU;4y(e*OEmxVksta~GC4z1R)idmPi!^8J}YqNOz9hb}Gx(nj1d0~CcFAvO?I
znnv?)IfH9KhT!^s$%`lZhgy2ahLaH~PlKVy`_OL8H<L5=CYa3>HrWl|#-ec|r%DAq
zMc?2CN(tC(vP9-vBF$ai{}Ax+@mv=9WLlGSQmV3Mq#*00O=C_F12HM24PKIH(qS$=
z{wj+xnz_X;u8yTr9CD%K%DOa2VC(*<{q54RHnp7Sv_`_u{EUm!`FkbF9&|+d^r5i}
zZ`CsfcP9OR<o&INAPN$PEW8W>g=!D`2Q>>_(+FN9vlsE6PHGxU7)2o`rDOe|<or^H
z`L3Y~;+T`#HL1RMM~U(*^%BOx#fZd)m<sA9ioL>Ol%kV}dQZnFcBTT(6wXtXPMTL}
z(KFCW)O_+JQYAH;+q+!5wNAG@#v5-xM*pE_cx~qz%V)3gkuSf^$A5l>Tiqe!$&{kF
z$zX6+3`x?Xs0)%B%3_n@(pmb)&QY%3#iYkD-365QsBQ<9XjYbbq*^iP_wd$|^aku~
z?2zY~kbqG1203+Usf&{FK|!JfxJ-<p@|HSq3$#+y)>1goK7c(TpzdUbKmXEszI$bh
z7cVYIMV)gbT61mVaJuiKr-$UJ5s(X-ngSK%bmQdlNS(fPMvLFm%)T8Kp=xHV4=J?Z
z&Z-FP$@>~kziN$21B8VM?C@>lk?__uPGU79d98$^9YYAMWYKO;YSyX;xTnCWULvZ_
z){~?O9;pD$@`P?CjLukDX=}k2Ols(5hG}J~YtdWiW`;^7;L1Y_uE~7(Rw?K~VBoEZ
z7=V}#j38pHEt`$qlRv@;M8kV@v!C7N)y&V?bf8u<UQkS>A{#&6&~yh<z+yi^d0AM!
zL^|YVMl~*WWgYud%OE!tRvxCCY0fQm$U+;g80(akBR7Vbgb=Z5p$>U}>pYcn4cIuU
zB-K+mN9i<+9fL(MhcckDQMtIXpo4+UwCL*94qRPg?37`(g{$|t`25Ft;jwdExn)?Z
zPw~Qkdx<yy$+yTJ9nwFuOtpO(t53sZm(m@OmUpSmF2?q7>I7MC*sm+nBqyKiP)|!#
zlAw&??C}9}nPfV2c*<9ur3&1JEHgZ`GRK{bJ@&RMN?SqVWrx&&(PYYn#V;+)z)Jjz
zr|?L3A}wFq$1W{3xil4ujdqR*fjoI?7(`7&m6>jVSrCH|QRHaoQE+o|7>&|OF%20C
z!G;CY(;=UCXvzj7pit;AS46`<WPL&lcBR!Jk&LDSF2uEgScwMmOhQK$#-@`e+<rj2
zZ;)}iG@mgpEt9dt^R;#8L^H^frvHJ|g~)hRrKR6VNR$*$6``G)Q8MP%wJEnl*^F?U
z`49teWny`8v0dbm`RHlG-Pr|vG?qU!yhlD)#&_0SM<Wpoi=R2?$nQ-Wa2(vKek$%^
zW;A<+rJsdn8d{NukBxcZUY;=MC3xq_H0)0uy<C$hnASqBs67la=`N&|^waAmGMKfO
z&)Jz+=5iA<3__YQN;n3oG0je816WaqohQ-I3&lxFLzIIifR|Q@u%^+TOCS9)ZoK{*
zTzmb0;@TUZMA-`c!W;aJH|Oa-dxF2T3!i<o=I^^r-hAyh(c3T5S@<r_tf1@wpY_?_
zy2ioXYjlUF+1<Iq!FEED9D~!3QCF6){NYtPos4=~g0mv@Fq)F2iKOxlMr_{M#7;}n
zJR=+SFx8kW6>*1DD^Lb&Eh^-nJBc(O@y?S+9<7!5PicukC2c0fMzgNo>0U-J3%&1r
z3)-Ow659dTb*_nRJ<<aHZrM(!3&R|T#P>40d=mz$)+K0WSHx}N9@UVDLS{f~)zDX)
zdUfBb26<X;ya`64sALSPqQWGGqO{!n5a#b=xn?V5GL%tFMg=HEe~?i*$F!32Z><zX
zY4IM?Oq8sxlW-skbWmDwo@~bOH49C(-oV3ygs*%^7_c$5EO**ogz>A2ReOIt=bvVb
zo3Wn#-M2tu4DsUMXL%LT<BIQ3wnsHfgOr*8%6ZQql>F+M2~$NxM?ojiqzZH*h<rcO
z-~}}F-m};-j4ENKnIL$UnUM2Ug5p-rvlsxBNMNayG$@%?k|7;rk{YYS#H(g3rjUii
z5EOx4o0>r4P*M<K`pazBj%4s0C$fvIEU&OvXfD6IPqDp^>rMI8kK{c4)*3%`sl%0p
zik<aa^pEdT9@Lo0S@QZ6y?#p8U7*~bp!1Vt`2srYi^f1Fp+DCpH;V3jPC1_7fqn1L
zI>A*nNjE3!<=E7qbwX8I@GbzDvV&JvN|CeRA68a!)V23$twqcvunDdAGYcq|-n7#w
zHYAO_j@UN50sav6VHJyw6ezFw#*Mv0lkIpvZGQLRT8DIa6iu1(AresYax41b0~Qg(
zyx5OG7JZwpiXMs&;fe~bK-x+1-m_PGc0WWs|9GZI(*Wc-kI4)%`%wi?CQpsfwT+Wz
zvEeyPs;O%6UX#RN%q)K%X3L)AGvXy}{-C}A03S?R0}xfA;sTuA@MDPaVaIHMro#0?
z^w!?j?D@gXn%(thkED|sw#PM@QS244&^3*#UOLDTciRV=utP{IUBp$7n}g9Xo;rrP
z#)U6eI^M|Si$Ic53X_OCm!+CfDU(08>%z-T+rajUg(LvoaK1?bKOyFG!^1^|S%54}
zSXmnK=GvGlA8_*0Q+(>N6>jZ}`Pu}=(`yXWWq#$(JACeiAus-*;LoUx^rvc)-6iUW
zJG?vE;EnGjkUmAWaFT=Fcj*kzV0wLgZSh7>>fWHk(xAgau9;Rt@SYoLhq|cIot$E%
zSU$19w5S>Fj<C~$8jtlBV-)+fA~#|LCGF{&SUF3DC*XV#a#Ub^NUK??kFd_MGT&*w
z)T?{bAV((NSQhFegn}JNK%w+I%2U00=YThNTM^Qw)NpPw4^2c{T5{aYn+etIHf>1Y
zVrUTzi$)f{bu^(vT>Pz^S_lPZ7Jx?4$x|lN8Xdfzt>VGpU#=7@9Ye2^P?VPOq^8#s
zBWA5<*cFdV2U3sL2?yhfs&aJu3AxeeTxwdR)3U~SjkmKzntP3a2hMgM+&AFz*z(*W
zO*f;sOUI+xnF*oZha7SI<9l>_JHTw5A9^zS_B%Rso$xc!N@BAUL+Nj3Jq;^um`bWI
zF*FFTLez7GbtIqG4T$|xU#OOK&`cX-nn~qDb5Tj>H6Lt5d?eM1%!uXfq?*bKXqaoo
zr1o@D%|RhDWrN&M)s8GN(#Rw{NH?IA9tFX}P{8tBk4MuvUVHBbuiskZt<^1@clfOh
z((W8C`7|4sYyOW9?sD^ZpTGG24NfJBrDrbkclYn`{ckQXIQ}X2tYiGa2jp`rR7Hs&
zAJFOc@g}91mfTw1<NA#)s4Tr+D!X)9k*0<OPhC0guI}P0i>n>|`5|sP!Z8J<8Ri1S
zS}!fV(nN=59}>H3CD{Py6*`m~TPx+dKDcTRofyzHx=91R`oSJgoEkKpcnJmtps<4G
zJ1^hE4{q=C<s<xoPoEirl3lx1>S<B^Xn6I^VsG63FbSi(NdN;Q(jCb`cp;=HJ$sFI
zhStt8o`~!3thnp$Xbaq#<FSsUP7fwxB#hA%A;dLCF)nPwmh(v6g)x%8w$4*p*VL$Z
zFIbneo|JY2iz!9#{S0WQcycb`5BKiROuSP$HYPR8oumN*<IW8n3ImRJ3`>drP<?~O
zTSB}nJ7bV$IAN`~5smeIJORD7Uy&Kb=BTE2;9bl7E3BZybpYzDlSNk6GDzK2SXtH!
zHF#V)%DQbL;(<;oVj{C(ha&WY_mTpP8iWUh#Rg)F(TZ;H6o#2*v1{<bSQS-U_#*WS
ziWo(D#BEmIv%EN9?Vup-_ju~uNq*v)3%qdd1bJ_fEFVzS6Go#Gl<Q}4UB}~}dXnp}
zeZbpyZ?Tr_QFQh>m{urbn7eq1S{cZ5bbp@OdrY2?^l~1)c!F~mj&c0#F)lrHf(sXx
znLEBf(&?bGoZ+z{Cr&NV9d=1NDWiia*2=DC0>!d)o-C2fOX*}`W=7!<WZoo3EV&8c
zS!@O$a`Z|oK6$n!3;pYBTOq8F_gjKUY1p;376ba)?R~y{W9QHWdto8tiIskX&BTcx
zzdfZ@gUqWqBjXRHLMP%Gz5yKiY-o3{#Du=~dd8gy=cI`t&l8*zDz7$*bA#l8yX2CQ
z>-O?QL^H#@i`ANWs+&36O?5Ng&I`^VV=k{7CE}}zQO*6vMHX=o-~3%b@=wgA_uf=|
zpHhc_z?KSq@9zAwD>+Xd;TN1r;8OSA?nAR0|A+1|zHuf(yPru(4oXWm)68`f0&diF
zv@HW_+)bl9DVTtDu8}5AgGDq-QQ#Yu()eKBBb{$)2XZ!plNbpf@>DY^o%DD|!T{~t
z!Kpn=Dj^&xrI=Vxt_4t3gfeF93H%P6^SpC=hvN%F)^6Y7&0E)aXW=-;JN8!JA?XZ-
zR=n_S`lqI><U`!b8T$Xh3z*-(!=HO%p4DeIx$){XY_ULPD@;{m%6+PANSfq;FcaT+
zf1L!!boYSOEThv+smBF*o{%Jl{mp$g(PWuu@(mXH*n=W~cM7zplbI&AcOJY}I4!bd
z9+4-{QX?WJiIf$$X+1H1AAM+vcQ!`c85K;d<JVuk!_Pi@lKD;o$b!ui1BgfUt!q1c
zYxQvY@MNa=v5WmUYnv3L35z-E3wVX}LyVb#8BZXF1I+{v>)Hj{uz-=lGnvUGC`AZ4
z4#&dYdnV&rTGJcNk1b`qy<c;+Xp#4)JBIV~QW-fa>L#R&Ieo2t#Pq^_ss?055!#}O
z{(}l@z*$FGIl4M5sNne|0ibMBf?=+|up~sSukW;>!1Mit7Zy^6NgM6QN4p>3?~%&V
zKe^K3PZHOVr<m-Kb$ZP>MJ>l@NN2hJYUWxV8qUS+i%r;kVwf@4F^oz}sU+v_aa19_
zr?qEXNeS+#uoR)8C4_$SUaU~<X}a1*?RN<$Je}ll&$CEh^uX>!a{K#J$;b9G2@UqA
zLNuzK=oNMn&8Vp9<x+z`wW9CeNi^e9Jc7h%k_d-tc>Lr7|L6B^((fjG^oghW<P+z)
zzB%SMzdB}pZ=c@e9aQ}!lbc&?pI+r}-@46JyUm|_=>*SwZ<oJ%ev!X><`t@qkC84v
zjlJ_8I_+@$p@*1G3g-Jglu~^3u@fwH6Aq>oxq?w)S>HThQd$gf?AU-~{R~%I*0x5B
zc1EOL;apAS!8%1Anu1PBD;0uH#@wp5RzS1`!@vnb1_+C<c9J6dnWs<kkG}r_liG57
zT=Kts`vxzb7;tg1E5%qIu5L|ub^YL;{sX7G8Gq`WjQOo?!4Xc+Nj<&=OqgvdihG35
zZ<K!@U<6bEA_e#pBj6zDJ_O)z=G>;@5b+QS@9FeXshZS^c?B;mrF<rlDpai$lcHu?
zgqo9HN>N$L(lYF4Ov{>SVObhxGy<dwl+u*7c%7yro<*u93^girNlrB`zz(w%iI}>S
zTlb*_nCJP$l;;=IX064AH*3)1#^_XA1v$HLdBERPNU16>g0$+$$$sXa-pBf$!v^8d
zO@#3cO&mq}H&L?qx}jrmS<CJ^3hAL-!|tR+<(B!54z<bA!iQRY)vCR8(!)xpik9`X
zS6Cq-8Q7cFEDr>SNP}CJry7EuzjL1DfsltRLcp-rvlT>0V(o%MHR+x=RyU}tnggA1
zbDZ!}n$zn`0gbO~w6-X>kF8c%-yL!J&Erh$asK}5E&kFI^F009CjaA;$NB9i$NbB0
zzKyBA0qI$G4la;Tv$nd%U^q{fr}X<7${428ib3CS^KQlHpg<up{lc)r?zEsTOGf)+
z20f|fN{ps*78}0YJQGFYabbm}g-By*yiik}!&&i~v{uxW7zORc%|E%=<(FPO!#{cX
z7L!_b$d}g+czNxx-1yPs#pNEKzPNxJt+6_?{Ijcq3H;n+$5<I;0^|jo<F#t!&;^8&
zwBY@WK$PBA1S1kWmX`fEqi_=@sr3j?!F$TeVXY(e23tFHV#M<d`F*1mNuZZI>u4(0
z5WJwMELBMNMg4p?gH=?~i|`@!XdR<zO=1*t-B_I4f^N~SHH#|zaki`S?6br_#J@#C
zP(yu~`3`2^cmE*AH{ftQ;`wn$Z!*3!;ReB|#`7Dpmj|ttVs9#osF!O}4f9<yL&uc1
z;5tk=!2&#U9RXHqFTDgJs8oAV0UA}V5%4&A#7uEFQS@_((@xH140D6gio!`~&`CnG
zj+ucsnkMs*C7MB|MfX1ffnG>NnUMvi>Y72GvM}tkIjSkVC+}sr@fy>KXX;N;=a+cp
zyQgpqFHsLa!}jW*;O~E9!of=?7<}h84_$ea(!ImN`E}CM>+EjtvbVX8%5%n5&Cb-Z
zF|llnYxZl;dT2!A0F@=Aos@B583(;ML5_vu>lAb|)AaZ&tbF~Mk$#1-Hr*tq6`9nE
zI)n<^gcMqcxh-&Fq04{x(s>@5>)tc=M~|-2eD>@-KlS7Z^3<?4^8E8_Bc{$Xb)H{)
zcaznFQpPg)@(Qp%K!ef0r={3(#Nf})=IE9-($>vRs+t%z)vCcLte142bsn8ar(#z-
zLIfn6iINS%)zWt%cCm_T%z~e&ChUldVOrUymqTo{;t(Sc(et?I-R#XDIpXuown}Tk
z!Z!Af*P1=zKO@F+^zyTwKodI5d`<(c#05X&;sc@Jt2It;=J`<y^~P%9r?m@6es~yU
z;a1i+MIP}@uYBbn@Z_hyKyNSv1=G?oDr}>mJ4genA4o!#7dG3tl0}~<a;@Cx##nn7
zC6OgPl_j#97uHi*F$75>oPULvg;58VT|mNHZFd1HzxzA?kng&WFi}1J)vH$+PPXXj
z9crs74yL%dGnm}4|Ngfy`H<nMhge-7Ggw$;vUwMW<HaZTdCtAb*=sjhdiV_gX1V10
z&WOA3E@5|`VtDFFj80hWSO!UrtrLv)P>h)t3Z)XV!4OrI*wF~QrZ?y^zq|mGT|T+I
z$@wRK7!!iDvbK0B<Q8<?iNVN~;4Rj9jCXBrI-F+$yjxT*^riE{8}RV{-F@C&-RI4X
z(IJ<AF-v&vM4#u+&eID;Rn~cSH*WKP-{77uLx1+sW1Jdha<6eFDJUu%QhBr$nZX_Q
zKHx+;=F@#c+G-GR@7rE_%v{4-wszLpFMDXMv1Lu(&qb8gXwp<uma>42R-grsan4g!
z4(}{VX<SvK)1=8-&CGqF{HCUnbJGHSHF3UD;B3ht^<<qvyCP>6adhY7HRbnq86QR6
zIs9_FgL98<n7O`S0C{iz5aq&o9|Tt03+kEc&0atLKKp*Hlt|mxuI<bkOM@ermS`Fj
zx9#619D#9ZnGZbYjZwvNKW$J~XbC>!T5c)PLMTcgh%*RoV7<g5b-)XhQcVp(o+!qZ
zgIXq2W`t&%1x2FRC;_B7W+$W?^HlTbu|?{n&u_hUh4s}t<kcFL>d?u$xaB@)=ZEYj
zi+J6qNEgA6$rg@bDo35=ym#vcuW$GHk3W8iC*Hitvu25>9y`T9xVpu}RrG9$b&gvf
z+yz_GPf`-V6=U+wJW19SVMtrUXdjd2*nZw{<&~DU;EB<K7buO_q9s`2NRtFrTb%RM
zgrX^Bu(i}lOe%Y5k)`5aV?yYlyyx*#b6hyy=W|Y2g9np}K_?+07?dy}Vp=OVia-1K
zfZup`)F=;6o#!{++vE$6FS9aCh2Ih0mnAyvmZD|YN>s@LN)4<YRNUArS)bGtGh5xc
ze!_`f!r4KKo;N8mrSRTiD~s-m`){onKW*BO)0K|eJF;#@mM2uTWiqYMiN<7#x@hUg
zRbd;w$?y>p=WE5Zs;DaoU6L#ju)1`j=p2-VnwDNO3wFkdX%}td-~0Rf^fR%-eH@Ah
zA4kqNGn|0<U(C$JI*jJ6;T}Spg62rJgQ&u&5-q@U^Jx&>loL`UADWt2tf<%-SDfs7
z2C2ZKy<D^C<xAAiNQ`^npmdU_*Q#;d(PD8+>+wN~%mxmB9#Vf2CyAhlQNr<WI;s!^
z4Q67oXa%QMB$>U3?WyITzj~Rfbfg!a=Hl@mWn>*2@4ilP_a>KD6Oy!tN3(V9UHWq?
zOxM<#jQ6mc2lU)GnLmGv?<76ymrnB2-&p15R52|V@Wl;wHgh`ZBAtcP^oK)oJWI=S
zjCMv$_jaWaC`^F-<T(<fD90miU%5jkttgLKl+n~k&bkhM?S5&=wS-X-Kgmq!KqMdv
z(>i=7+GD-YT8&nC>zj@~MUYuDM$<_|@n*iC(GYPIR9*~2^xDwEB>aVEkMpZPSZ|EL
z#Cg8>&IZ5m_%g=_sUXlwG65*@;#}kmgl+K$Yh%8%HesuN5Kmq>y<M&N_)>>wj&;a%
zU_W|~RvK*#<)p-93DvX~jfJA7nn)c*-cQk5Gv1%FJ*`kmk@f^|EyopUFQ+al>Y}F8
z&js|7>BwkQF+LF5aMn#3hUYriFGw@P(tJ)`Dez?oA=(WvF1D~P1?M>Ae$QmWRJ%Z3
zYab7Rip>JP??}bZi7HGppF7K`X#N;KJ$m#KYQ|a+MVo;zf91dW?|!+TNE3;Lj%jN7
zjS|lwHBti}Z*KMKD;zubG~G_#lx~(*c9L$YrPHt$Lt(V2{6s!Ml4y!h@*8(^>!nYe
z4e|ZnB=kM;VlawYCT?OB8zUid^itgvxWwIGE7hcDj8fdX`U-DPS1@VL){UF!weK<B
zxXyTejdXeooer2EF5tIc$J>hG@*~t$h3fXO<86lB9!WlrHGQ-hbMkb>sDGT3yCXjJ
zM2E>^JKVaviXR=vnCsZ-9g5v`ii0uL-YV7J9;6*yUEs!hxakx-IY4>EsS774iz$yR
zTNY0|Bwv<N0RWXTZ{mG~$!Eg9r=fgWYvE%E7$%f!6<q8z7|-~f=`g+uAH-e4ELFN`
zNgm_z$#{<}%jtHyJhIf`#&*F$SirT1cef^-ALh(;lD3^#n1}?Ua&x!f*Dmky`fkDG
zKE#-#NA2PIxaRfkf`vr0+)Kbais_hXm7o)YE$dLrFFUkJ4YqRBm8B{xOp?&+XVkUB
zmbIk!w8Bm+P>Q6JV6CIBJ#}TJ=&F=*;jEKjYnjW%v~C)^RMt~lOJ+PK$z|~!d7QZD
z?{V{qLLV~=);Gg+C~m*^o%jB3>(DCwKF6#=Q{xTH-0CcLFpC)|6nd#^8drUGH?3T&
z#v3mZ<)O-;{GtdsQ5O(YC+WaE)f@z}&M?>HMztQrOyYD}%Y;p|0D>l1f1?x|`=w|N
zD#^)aS_P$G&$MzB5vVzGL^(5s=n0-%98ep>!G6u|LCK`7sEPug44Cx#c$MI?;}lsA
z3hE?9C4JQK)7X<AXMgvHxc%-$zW6UH{)ewm`SwF|JoMT+|3x{^M?da3$iK<K^!rS9
zZZJOBWq)^{{evmBtEkGUxLo;w%2w31#U?4^$_r|)1lzDmH`d%KQLoWPlO;;>_5n+=
z$nG3F6CG5JNf<*LkZ}MX0JJPMWLn16h40T3yr}>x#+jpa=#Z>i^`~x@@N-We=XAFf
zBAGbPm)_mv&YtXwv%sLzitpYU@vB#M8j;bX$LT!bQZMCFFQt2g&@*wKe|cxbx9^Mv
zJPQC|{7-3xpG{pxW}y?0unyEW?~#6vNN6MP53@qI9t%ogD~EToIe9+=CL)bv%AiQp
zEUGYGv7#@&#>~GP#C^s^Cl0l53!d5wb7(F<&}1IL4i5j`wxOTBv_A6hxGBW(O;-0R
zmP5;72hu5*!<LIMkdl9d-Eqlq4u%~`9p;H5OBCZm%3*P`tH)e#!whr*?G2^BHKE(C
z(oBK{p34oHQ4G5Y)=6<tRXe7&qo0Wwg!MEoax6q@Xr?m{dv2-E_iq=}gE^MYJx8rm
zs%{6nHXtzxXhYpSN7h-CNi&$oX+@TGaLV9pMPjB@_95!x3~F+V-@Ne_AN$OCmcG5p
zzyGlnF8%C?^_`UOeSe>7w?nPZlc}1bIzSgA++Y#wJ?Z>1R0Xb_uy@cWlkOeD<@W+?
zdFMGOEDO0*_e~wxT0*E%?dT;E<GY}%D7EbZoK_C%(lg8qRc-M>>Kdb&mNs~HGA=o~
zVeLW8@%y%|wF>q3TG7qL2z=>{wL=X_q6GD9G0N{=-r?o#$vyLSDKUI>xyxh6ayp4_
zIvgwD?m@+yTT@=yEt=WJ#7Y1FAOJ~3K~(d<xjkiSEzf2yEMQ6h$qp*C=(v!T)|gl?
z5sZXR4LEsNlchN4q)DsRq9|lVR=$jxg|PaxQsP<E5^9M`P$DBVvq&Rv;poDT?tYUz
z^x>}nfk1x0ss;1nn`T#+hX#QBw>kWge|Qep2F#ul@xEl(5f4DP+B&=di8{>Qi=-d#
zJ%dg{KhvSrlq~L|a`Zy=-qd;qne6(pf=}3j^5J<?At*c7sB%n8CxVYgCRaB|O;xqz
z{Y%%TIIQzHBVr-<i<{!ZT{=GCo7cxI3<exuImSUz)9<9b>%iqH!%oij%A+`2gVEI5
zU=oAv_L%GU+1=S;c=9CU$uaVx;vny_`gX<NoLuIwf94Dy|K<jNe!kCJA3x4lzV`!C
z_XsBKqIx}&dPzF#D1|Z!rrUuu#TEq?Ntb%yz)J4lS!q9(Y5_nqE&PEjl+7g}1vV<g
zAS8(<Q$j$gJS2&r+S5=Vq{9TAmZI6$O(blv)=A1R_@hdJ4n;b(bq%N$6^509Zl3V@
zr%&)t-&kX(w0!>YWfuEsXhbOk{yUeq?^*mEt@z~eJ};i^wLK2-jf}MdRt71j2Pu#4
zX8g+!4w{hR<-H2JIp@>ElvotF(B55XMVcp~oK#uLNreiPgRXM8+LClqslf;&Dr02q
z04g2%0OHAXGs8HDk7}*xWQL+Rq6z3@SVhxTJ>1rz^Nynbv0;n6HUrMx=Y4;i<AJBm
zcpX|DrvEmf!0hi?=v|g5LH>*2V%4E!SK+CwYvPPFOqBY_9i&>ZUx<EtZge<b!h=|2
ztu(~alsE{9mK#cpVr{QvTuD)eS2WlFGhkmY)$}uoU2SmvW>}Vo4(l9Gon7Mm@*K~c
zUuJ&LVf|prv86>$&ChXaen79E<Fb^cg(12w1%%2wbdrRH#U)0hSZ##pk#rgB$Jx7n
zlJ9J<bLOQ-x%>z3QN4MKlPfFeyrOUwwW}#7`&g4er%UY}MQy18ZxWI`#pE3bldPKx
z#kKYxZvsy}HPk3}rVbZynbc^q%%F|%)}xM~Q{a_^24X109j`;!(9d)b{}30pbY6B&
z1t7#2=LxZ|64$udy~G=kaNvYAART56dB+34`~D6uZyjFzr#cBg|L`IoJKdKQoND9F
z#_ffMyKsI!=dV1z$mzW8BK7jbuxSbX6@BB*1L#DE|IYfxBhZNsI(~8Y6It=tyfu=4
z>V^WWFh)`eF7$8k;h9x<e=<X1jf*<M;ZfM;eldG=v*Uk97k>=1j=}&jM*p`tVpw(*
zV05mn!fZgm|NL+M_rEN=xI9cm+O>)dK6S{wzIEjlE<F4!-EMx!48_VLts+sU^>s=s
z0O7a_Tz>^S6Dvg+sb~x2iHL`!p=n5zeonN!zn@8=Mtne}6q%8tmZL-FJ*-`OmBGo!
z`N%^{WJ$)i-e2dfmtVsk9B|>>DW1Q0oL~Ipd9K_&09)}Vp1Z{R+Y^5F`HS4#Dfr~Y
zV|@PO7kT{DGNu~w&YQ2%U%o(^+@`L7K$d50uUPgTp5yqp*7!3|9>+bu#~bgxirQYM
zwhCi<;wIM>Zf}#iwk)4K#z&sJNWY(Q!fY}){;*6yFV7XTCy5eIC>Bwr!F|e%2(LtG
z>rpLUK%zx^QH-g?vqlLfqO_>`D5J<Vjk-@1?}!+IQ7|3vk)$2ctV1lZ(h!JXnqenF
zDe>NvR=oVdKHpe9%rZRPN%;B4kA-t%5sV^=U&eOe@oUqnRW7<POnG~|pbB$qyY@Uj
zpNogo$%Jj%&*=7ZX%J#1RhRcNdVT2$G1;4<jVA93DA?^J<at6_)Ko={_reuOQp2#H
z(#sPjMMYg&oOKNPImRdsMio_6(#sS{=#&Hm4aK;_-+#F_22b4|?2B{wQ20Qbu4Cr!
z5J14R^Z&v+#2x`{(UO_5F46ctVDJ2YSbMXW%dYFr@3)6D+~K?5P;(VooJ3Kxs3~g8
zcG|XVtDVG#&~bN=4m$l3pwS7EMjrBz1OWm#LGsZ3*g+nGB=AGfaodjLxM8=fq1|rD
zkqwigWl<tUkt~umSAEkR&bar^!`l1Y^OZ=oo#X&y)%VrC=bm%!UTd%QU;p)A?|tZq
zB%cK&LafP0V$SF)t=a++DoA9}_9EI@#Ia$ilW!S0{U+fBNQEJbI_RzQQL%OFMc?ZP
zz@^=CXfC;bc+6^b%<;o_XsZ=%*E6w-y=l(8EV#N?QZ+3%9<FKHmPuimmo^TO@TlNu
z%Mxqn<jqBHe0Puk@i%Vs55L}_zxZMNx9;%oR1<#j7j~$2Ux7l?DNEl~__m^TEnX>f
zmNB1*lvwA4cVA`(WhCUcan7PLT59{&V2!4B0t5y?8!aWMn6edCQ&^!D+c?oTi|V&N
zu0Rtj2|oT^<ifUTUmJHs$*bcI7#C0K)C+`|zKFwrxas)%?X%59SIUepK6j2R(<3B$
zt7Sl1h{zJvj~K{KuKCG})8Xf3fVXQmAX>&6w3hUcNbd2@qv9r_wU(KZ1es73`7K2z
zf$xx)4_uYVf3aqWE)t;adyjvlEmv~+vmtl_SN|f%c2dv~8USbWb$Q)i@Pt3&2+_f(
zbElm@B~?BIv`qN<%iO}dAL~v&p?7|$>!)MWIu18I0m!t}>T{!o!Imx-={n4`it<c?
zLF7hBIuZwA^k>KoKl+;cKG220z0;Pa?Xa7hsP>ej!{gXM19wg<ybDaDSfn+IuCC}?
zQSdFT24(SC&gFRlKBv=9&~X_b{&iga8UE30_o)8*b5t+h;;%Pze*C8w_~HRgQ_-|5
z>UKq6Jz%{#CLo6V@N^^Nr!s?Aiad*Lkw6})s~DkM8zU(JAee&QR%fKwT^4@t6TwK2
z97TG)qzQqMCLL->7gP%0(h(hr9z}m1HMAIm;)Qok&&ZG;WR_q4(0Qq1z_8<|oiYsJ
zzlh9;gBVG`M@RGF3lk1TdEMRGNgf|}{K0<XX5f1#({hwt>6ifHnr<jN2OkK|;d?ir
z;3h^W-pPb$;xqAXH0^nOjCRoUpZKEfvrE5!w?lBiR6W_gXWSz&L>m7;IDvsTFe1#2
zsmWT@NfqyQuYa)*klp{F6zOjp_3YDH(vj)s194!4++Y*=0b98WLATpc)gWDDY7hd!
z7o^55?aJ;woPZBPOtDIFx@o9-kMa)NZOD}4wCPz^EqkS<s@5E?JEmp9#oY<zY>G#5
zyl%PiU`5wd(6t0>G4lo9%xJTx(9_4r=hu1p#hm~4rI)$&g=;i_dXvA`T;W6SKc;D4
zp?7cM%?)~ekG?sgS{-rPc#O3KEkh>JX&!*0S0d-7jHZs|DQiR>*I2O{rKMD;wC1#x
zxq``249L9QMgl7C`Y0hBCAD1}`2?v2*@tNG@mZ%C1wMwWMRYcqQJgkCFFrDX=oj9%
zFHwwEQ8{WW6tXRwZ5_fHkSs08_}S0yY}tWtG%_VxkMnUDb~hF8YtR%JtqEo1S)sP_
zUf%hkNhz&M6SV<C919;nRG-4QoEe{sI$O(6_c87Tc-P#1^xwFmknH7Mm-7DyConwI
z!MTn~l?hAHN|EXJ_E_(Bq`@I_LOHn)48Wi~nCNXh>=Y(NaJ0^oTgAcDL`kph9p*Nc
zmO}JQY?~||bx^$@;}k!9Xa!Z%vuYii)=@Vt${2#lIjJ4TRgW>6!pe@E6nU3*&t91i
zY{p4$QLaV1hLfg2tv7Tk!~2HGWJX>T7=M9RUkdz#@4n9I&pd|v<G1-Rx0yYDpRo5P
zy}m_Tz71x@(a8x-??jm>NSKxxP2;7~;gC{P7bK)@z3lp}3qw;&h&N~q>`${8wo96z
zOf*`trWiVl$VD=Q)IbyTHaeK(dT=1HT`-<4p{0*>VWkw`ym3}R=&xQ~aAmg`xCL6t
zFcP^|w$~k3$Pn_zC?FYC(nLsnj&z*#A@0x&jY1)UC*Dic+4Q1S5s=bh00@D$bA!b?
zmxyNIdPkbD)5k`l;6*g_(Oi6csG$mwj&YLkX!-ozjx#m?5W!8}>+|@36$*^skpSJX
zdStR$tB^2gJiT{&l5oaAU996TYlkX5^C;%tM*3-Kw_e73>0)c0)atcjkhdD5gVBwp
zDMZ$Y#$DS6CRXUF0f`t?r2kH{e>HUB1Kd6BSo#fbpH>88m_PPOx;&@#o*TFB^6P)%
z`I9&A;;M$PeCH;<?|J9`lDnrX-npM~yj*kp;gZ#=BTUcZj!)QZ99HGfoY1sV>TA>`
z^6WA0eCr|q<vVxy<cE6x&f$X3e|W*G*Y^0||K$7Jcnx%&vDvKo-Wzvl9^T?#`(WT;
zmJNLYMQ+Fx)Sb|Lo6fUp19%@*j-ijCdfhoxP*@AWK^28UeB@OGLe%@$gkDO2DLRl{
zND-WiBfBC2$V-l&5c*`uDwYpZxEz3&M<u_~Xg+>*J}iEXkep1kqTUI^k)Q^05Apm7
zUGTA`SjFpnbVKTaz7s`{suFge?|Z!W*s`FjI*wf+^bV8fI1kmP1EZ;y4PEQd+Mp(e
z;5}Jx>6)H(-Lp~Py(gdMco(QQtwav<5*_&j<6~g-DYjLB_;<Ojts?T)mbM_@_;inw
zh0-g-=%@6$zwDtN`E%?9Aa-3I)e;VE-bK;0ac6HLAizbER0=V#E3IalsbM~xA`a&W
zVW{~}qb@*hWt>!+)t-*rN*S^4rBcv)p#UeT@TT{8zt#UUT5VZ{?{a7bPhXreFD-ky
z;_-9)>|VOWrAwE|7Yi=#FF5TyS9kZ((-}o(I6s@vd3bVffis4~25OztnT+Xd4~1f~
zdx6Q~0(SRtyz3~Z3+DURG38}k@qUi)Kg+*-@j6GJc!vC&Z}Rc)SN!BBF4L=$T2HCk
z4R@C<ofd{)6DM96qnQ+zpd@8#ykgx+33fisBz;nfSz)L{l08?TBzowAv}QY{Shju$
z{VNyfb)*6-6~q2Gw>|=e#N9yal$*ygV8jm+O<2by^?SF^iitdbVa6=#-KD9uf;mVT
z*v5)QkL=q0D3nfd^cYuG3YOhj7m&peU1vr@^!WM8m<WajRFJi-jghn}Vk0_6PCf`x
z$QGG6y-XU}WXzW)%i=Jc7<JC3*$<^1vbN;Hli`Q-kK1&6_#y4~$?%7JW*iBq(f@x`
zjsLy=;fIeD{tw{aBWVgDPHCEEqrB#kB;-+VW%ah7T0e4`n=a9IW3Fzuxq2N5Ds?Yy
z!ABc6kc8xO0?vg&%4uxmvrKlKS+0lhFumtTJ*Ped-g>y9b)J(=%Ucf)IlOnDd-v|r
zbdnZanit&PtZ~lqU|G|8Pvd&-Zt94RD-NawyYn4vCUR4*UIT%Avcq(?OK+o?s_XIf
z19EjppTEN&Zr()y;)l3*;~qbKf5PAX*=v~n7jgM*To5qJA;O-047t~xrwfw5qZPB#
zu)9;BwPIFUsy+<4yRn*1Ni`tP1Sd#X&SGLEs)(9)g-AQGaZyx63llCGX<7h2Z2Rjo
zld1y@iEElaa9p1i=709W3=*&vzc(BC^g<rnR&11F=w%RmXS*<mo&*#()@Rj!=OK1Q
zw%rh63dPss+Ae-ZobA`P__oJ6hwJ-+B}ja&>v~BA68Npa_l~;iaJ?Il`qi?b>$cZ1
z+T=+SkO#ugXuI8R)?4oV8AlP$Vh3m67>4KZu50@*;{@WxCgV>{Ok?`CE@A_ZR{RjT
zCx!H9FH_;1=<|Vi5iv}mq;9%!<~(Vbhx0HmG*(Ngr|LXyFFgmTapkdHnWw($q}?=4
zCvIYjG0UW5u8l;bs`FIup`IE7EUTVdM-643%a!*%IFG6-4%ZuwH!Z%YF?EFwj-#Vf
z&hJcEFV`3q*eNVZYfd&5O;Z7$rmX-)a2>tU=%$IyQ^G}8N@GtB_`Mfy^N)_cPyf@;
zP=50rKHqda_pzF~yiKnZ!8ItQ=n%!9pfuJRicB-hjevhvvuynU-0hT>KFEl$6rJ^P
zdT0`7?GYdrC4?YrP2A0+@JI~l)1b`|LB)g>B%uwI7_rqt)X|YEfKr@1GKJ{MZV{^q
zQt<5pkxCqU^rLey1L(K~zr+b`&t<gY?bGH=2jS(Y`7ms23QS=n%yteU!6z?Lqpqxa
z8GaH5+X`A7gm)g_Ir^^0b$x7Uj^)I7kAz5^iSs8%D@%+i*@HS8$wnhdfz?^@4z>pT
z-rLvcH2(@G@Ls;)(ch=e+4Q-BJ4Y;|jmDqHy7TyWw>o<upj8+YctGm<g;C5hNd?ka
zsDPvbjSH+>3Ga;(%wVT5%!}=3Dy67<87gX(gpEq6SfZ4}YC(I`wHp1S=cygcGCuv}
zE{npT4j-_3@EXfIZ()_^!eY)(ec%%NvnfSk`Skm)l21x5EK108&P@!TeCh(1&+U-)
z4bAc)%x6r?IqTJ-^!D#8Xch}x-_le|^6rSR-r({hHh1><{Vx^#Pv1J?_Agz>zIx2x
z-8s)se&{K@HMCk$dkMqK+^})b2Ba}a3oSLz4Cm*W7^%@HlZ;z2Evz_OqH<6W%5HAO
zp(qszLCC^y5)Kx3=2UesG34#zmXM|e6LVap)d2jtAlG>Gh&F#(s4j-sAqe}kO=?D(
zGe)1bMTL}I6B3^)5{(OI8ig(vTBaBog|Y^l8S>ncYKh)Cbf%e3GHhmK-k%m`pzAu}
zmSl#IN_|cu0Th(8QpSO6$;DmUlNX{u?Av~DJgGkdMAIUb{>}#PgOU8u^GW%?{QRyj
zlTzg<k~;I^M~-pI^ysloRkpnysn6rllUS?3Yc(kN{3vow%6!^R2sufvt8ucE@1$lO
zADG02lxqbUs7wvvUgGdqofnng2sSoeW(nYZ)D@JeIP+X;n43-z-qOmh;-Yley8)Pq
z%YM9WIj$PQ{s9>Z?LDFMoNOAt@yc!P9-U(Qme=kaQ#;3>y>&=tgob<jaLI$i4QmW~
zvMb_|!J%wHRc$af!y8MMP03A3n@y>!6Y_SKzVm36@zQtR;y+uz%zyE@XZX;UZu4`O
z9e;TLcd4r*toku2?P-aWCYeUZa;Q>JJCWWht%BYYf~M^~I^}OhQ)I&C3QkH`E@(0f
z!ON~@(S#r%Vo(yD*+h#~8m*&@d;Eg60wNC36X=0FhNhAh1fJZR@aE}uh=|rAx79nL
zf*TV%Tbzh^c0CxiRRK!OMXv%nwte?L45h+n&rP^CH+<ub`@CFPu4pI2IpVl)=K_7#
ziJX_#Qrqtw$IAxaI+QlD!|NW9D4=WIU>ZTG*jyAO?9XEZ5DJW;sX9p+wCE8elSqQ1
zV~KmYpEPk<o>PE<S-A~Jgm{(GfHO`YzAja?B=sJ>NR45^GxqzQy#p#<LkJAO?O?-d
z%R8{?3v7C68kt#f0&OJTgb;?Afvpq0y)O9e0Fg*94O+44#5gB$WLOp#Qq%*H<{q~d
zN=UTVrVpS3r&Z6?3KzVNDTYF#jicHz&$kcAYA~@H(D%|0kU+`a%c!tT>-olO5AZ?p
zGar2#<$7`r-@Uu$Vb$=~o%>8CIYqI{-J=t7tvFg&OmfTXcTZ4VL)AH^x#jBnp62F*
zWAfRA;$WYCv%$FzZww#(@RRIJ3rtX)ZdTlTaG&+bn)ULS5DMy(1;6_3tNg?h6E45~
z8vp(U$Mm_&<i4dWEKS?9H_wFF5|R1SMu5C5g_Bww115Pp&HmI1eK#rzA6C6g*;NV)
zs36Z#E5#%iKE0P2<T7AG0ZLPM4jb_ruM{o_gD^!`jg!g6L4{$A*Z8=7(r|6BL>rCr
z5~nNTcy}LXAE0nvCOVH7q?62jJG>N$JW8v1e&2I#RuY;PT}Vr^(hAo}zrVId`YQB<
zuI>jNfO2;_g!#qJR3-s89mTvD(gJHmQp~lzs8r>avb4BZ1u#}K-!166niTN+h(T%6
zgyiw0-cgxTZKv`m5(dInRZE@047V7;$e6dP%I_Lu#Miv{F}!n>E~K(+BvEc1fXs{+
zK7iIT9JJ{Jd!^i9=OG{i712=lQPDPve54hzZljWqi8JmHVox3yl7K%WZ>bc=anAc<
zV!6Cf!$Zm~JY4n63N6D+gJzOzmbD}L2z&^vJ6SY4dCYm0l*(Ly-YH(WyX5Z4iqAfE
zfe&0ePt`cSeP_dI-Ez|REDFPM<M69BSFc{?#_^JAuGu)n`oRjP4c9Iou)nv%&3h-5
zd;84Fk~f@185uqO#I-AY@ba7+4=b+cmpHs~jaP2m<@lt=H66xkUac&zy#8gr_}n%A
z&YL$0=cc^zm_b`ZsTBoYRF0RP)7r7A9gi=h<F1LFf@OTBUDwk&$E>h;=P7dNI>+8D
zLj}>cUw5MPBXr)t)Ee5}p_KgD^f1pvicD+i4RDcaoKSGDG%k4Bm_rs(KkVs?bG~-t
zXjp8o-dpk5eu+U5;^rShD(QwTy?Z2Wo|vA2SRRhE4`KikD+Y3o7L6%0(Mu%YJJI_y
zg&~i4N!NP1x<eVoWM0s<JzdpdONl7G59GPfyZgvA?RwD(=)GrWK4E8KS=YTx2hI#-
zZZOJ1Xk+<wy8#8Zwe(cWPrWmVpfq*}W5j=O6{83tfpTYebT!)L)w_Fv(XSK8_e0KQ
zbUa!>w|8&8!K!s=1*Me|3kqh1;p8+S@6s~abv>bPG3%Su4=Ps9p^zE(ZP(LzaHWL+
zP1i@!3B?8mrJh74Nr&@6blbI|aZaYP#O^Ocz*mOj6H#nEi4xeg*3_-E_SS8W)|w`=
zxU4Y*?+9JRYj3>F6+D~eJ#O82z>Yr$1d8J~cqS;6b`++-8qG<&p<GvdELU`$2k&rm
z*^w7r&0fF5<C9}(mY7cS3>~iXWJO6)6&&8lxq0h^omolO^%To{JnbE1imW}Qp`^<z
zraKeV&wP}Z|L9fJ!G@#b`{*uE8YyE{%MHHu?5M!%B+di~0i9WvQ82WrTS}`~)>2w^
z&QUi#UFRvRW>a@etn|&dUC$)9G`*~(wv+E#b&k$yw3qQ$&S;v}(RU4bx1z4C<kH0S
zLo0g0+t@EXmnW7xaRk>3cUOGsYDqa}l!dTuFAlQP`?$lWY0qgvwqdJH=!5*MTnBiR
ztFXqSoFlk6$}5v}fF3-Cwo_<&vRs~zYb!$Mu$f1N1{EAx4#Blljf-Nc!VFcqLwS$u
z6phXZ&Y^sZ@^aqLw|M2GWjtY9foSAo4McHTgVp&kIJyM5B^rAI`ya&w(lJyPqA`2#
z3Bluf;XkC&7u)uYt;$ov8hnhb(j3GT0ff=#DWy=q{(t^j0C2$%7n*W?fG`>>5(pgK
z`aX*bPohl@D##!O?`is|-m8C@wb%85W!IBy;k<XwOUS9E1v!j_3Y&=Pmo`lf_N;2p
z&O|2XWRc|JJ*-+MW68A=K}up<QScnyevQS&Cowh?Z;fKv_E;sV#kpa~S7n5<>~spS
zaWpyUp@kr<xe&YK^p;kdX1qV#bmWnRSND$HNsbPZGD$O!yf-&sGU)n_!|%R;Ila%F
z%TrE|8=B>u`QFn^raM9isyq6+W^ZSQ`*&Uk*Q3>h#rbPw8jkP0h3{IlDadCBXr=L8
z$6_(3Z5_EaOpA=;qf<7k6(%bviwV=xGRrKt?w@e+U_o8i96h|tVsVj)Zdgp`Xe-@-
z-i4t#=CqP>+a%MxeH{4Rx3*y5C-xQ3Uy1-@B!TI;SgldROh27=v}6o%t#mSXj?|gz
zeUI-NY%!IU7!=<<5{L4^J>vJB3?`+ETLVf4LHzCq^gg`H?l=nP(?p>VY39m1z+-GN
zd{zis?ZZKl`bWj%j~s)6hcruJ<P3P!A;>@WT_dt&k9dpW-iP0BFJzdqgtW%zR*#%P
zz#5DF@aMiLVSen{&mxChhANJT907t4yzu4!g~j`So{&x0n~E^V{Z-F2_UIQ@(MBon
zgg&kt5v-hMn#{nemE8d((nuZPZy&c@o><OJ1hp=$;<$1{%~eOQH_bE;);)6<cx*2p
z6pWphHr3t*zV+2#<+;!N60=!3=n!^ZLO-h*B(S7JDc)Lzz$A9Dt(#a6R`UBKhSHD7
zNUf9Sr9$9x6hA#_JXXO@A%LhRc9$X5l8MwdIa<~nzxHR8H(%p3PrQ$Rn7z%*-<xsq
z;-{FNzldJ1px%Jl<9xl~!JGH-?E@zH4zu+fYcuM5EAnDO=mL|yeTlHvD;DlNUDHb1
zkQu7=47XmRP-K&m$s}iIo^$uevMly!S1anZ;n2^xvUAMi7oTJ@ThRB8P3N)E;nMUX
z>+S;A7mmk|Zu0h0z`z%e0tb70eB|09j*ko55#3sFqTVHTqNE&nk(Nt5kl+PF3P9I(
zg!O$UiwkJrr8^tGdMKTTc`@fB7bn77AWF)Q`bAPBS_lFxst|{J#&wj8Y$z>`j7f%)
zNLteSfbU!AYKrMD5QZ}8(2*M7uhm%xl#JPj)TG3)JCENpW(=mwv5GV4f!LlSpUlrV
zfHT4(VT-LK>><sR0MOAXr|0kEGqJ6coG}r2+NksHL!jy60}J8+Ts-@O(+#`(rOX3R
za4<1!S|3$zG^LTQvCfMKqc)12Ty_WVq}rf@>?%#`spI{o>cOV<99NEuGs(%?UK)QE
zr5Ox-r_i*0VADFLxgm=HRqteIgnZdDd8lfAlxh;gTsNK&0ZkJ@wsqr~W@$)@=oJd`
zt(fMT%%?`CI5Q#PCQ3lEc`lCN!qf}`o;t|R*2Tyok4B(1ENtL-0+WC1gZQuC=9dr7
z^ZlPy{LVMO%IT|5lh?<DY{sIzz|uGo{KzxBYN?KH6P%*RHFoa;?Sq@NbtOAp-?4mn
z2wg{`G*y-fr(jaz+m7b+l<pL&)db%k(zG+0^%_$YG<8Rt*AYZbA?((mCb?mvp+<NZ
zzx4by{^xHr9Q;uL03ZNKL_t)%F?0le<&B4Qz30cD-ltU|Vh&PjtU4E;fqd^bj%u#&
z7ek0#7$N{E2IKs;{QJ_~4PUyoZ9D#Fw^u9*!{dv5J9tzlW$0~$CR&v-j1COMv~Bpk
zwX%?+4+3d~q16hf6g^v=KiH<TjDD}S|2!CGDohQb_x2U>bxBlYD{ntUNM~L;!W70v
zAOCm8@u+RMKXz2f38fhh>CSc2`C*n{8r5x-qA)VrJE4WCM5L7B%5KR#mu8-cRh(2k
z>rPULb>lgz<jX#oSf;s_K@CB;0G$sk8&B1G?kqcIxn`%-%nL!at1gK-%7Z?bS*%e!
zwwqH}9cKy3mvvJ0_#oiarVE_ZXN4%cE|A#IO)KEqvX&BN=flvY_;4ewrd=O6-1H(q
zl4)j{9stZ$>uF+byNj<i8cu4*x)U%m4Lv=sJ$F|fU6eNu9uC(%N1L9;1*Va?*Hw=H
z=9@3`I~Oe<`1Vcy#$>`{*LP^_KK*QuysYp=!EAq@cKtS8{SMXoCVqB+-8;u-`H=O!
zn>6(rbimH1_~5aV0yUeG@64F&F9=#wot)CFm!M@1-rmJsnoUD7Ezw2E=I|JkXA~0=
zo$xAl^f?(^YGYr<VIu(JqBLaI@Wqc`qa4v?|Ha!!{NopI@ts>s0Wf<h^)3oSnQ30S
zwdDW!&V7F4`wzHv(#R-uC*r257O9kqjZ5(7cPqYh^Q`8;C-+L8*eeF37;E~0OqWg*
zs$4XLv_mISR4Rfbz%vb|NzBNg#DLxC+{0(ZXmU`sis@{Kc9NXUSg|@8sVpbib;9tw
z2wHo&_DGXBn+_;-mST)-NOixJaT}aMyx#$T2-~P2MP{jG-zpVn0|r^Z8Y!`LQS8%1
z51@v=0I3IOnP!#?2(_44=~mN<B9mbdGM-6<8m5Jzb6e@IiB+r`&*hzrggP|6j217f
zG#f2S>Cbmw7GQ1^RqHvJS#qOTwVo@x`Ea9o6edkaKxoNboTFD}nu!uhant)Dr3e9L
znGkP!7kF$pr%A#i3YL{)cWOnV+(|Ss$>bclO`)uW_TEcc;Dba2(@YwAG6Rb!xt<y6
zi1YxiWcA8({P_oO^TogQG1OPy<gZ;k;;n<X(91bpRS=q&xqAzo@~E(jwg&{Q>Ab>B
z4^WdCN@<*Pv{i*F3PJuobogPg2*wI~vF#{#X854VCVNy>heu;(3+!})*2^J($0#O+
zh?P`LM`;Zq^q@qip~Z7)cf#L%{u=+}rCY-+!Z%K9-Z-t{2XH->-W2f0a?2e6`2E+9
z`P-kkJd_W$k`Bl83|~1>eC^@snfE-uSMpO&E(T&z3iIvA0(2Z&YSdONHi6qFMgc}A
zYawbI%IkVBo8GX)E0J77$u6(rqOlQRQz4}iDuQt_)Q;C{Mos`msVv*|m)5HH;fHek
zbf{4(kzhZo!9-$L!WGhO$vJi8(x?F-Oe#>xIhlX&Kl;!A>(&W;_Hg3{X4vgzPY7}B
zcOQlcFgIU&k)6vQWM@`V=A!doW<nu-=dk9&+zzF@HX2`UrBP*SWyXMxIkM5xxUz0#
zSDoYrV-!^vC9VUk8W)*lk+npUIFjf?rs|5*8kcEqzy2-GU;hww5TQ*eCyF<eMj{b0
z(MqJW+8CNAuUtDxDO`{;sL}E;3!`E`yiDgQtdMbv*e4(2mz~Auwdtkmkr{!aa-%rf
zI8K%u-g)~6yqy08_uo9`_utv&<Wj{)?w|0*h33j93;w+N0e-VjyINA7dzQ($4=}s@
z6z!$~*E75P81MhUbte0JRIRkS+H!*LT--4NzVFG(g2$e@iqjc=+oQ9b>SReao71-)
zHqZI=C!b{9I&^!;wF~FTCsTaHJnk=BCb?x|6=i0CC|qhZ(_FJV$#`mi!rh~a<s)+7
zC#~bS?Ky7GvIhYw?=dd$<iUhVW`+`^3xV(5-|*FYXL}Alu|MJSPw(<5`&-BDy>rp9
z$2+dxrmtgW+;|6eev~V#?HW@GsRp?oo0VH&90}D2ACk805ulJF2Nl!lL~=@8g7FTo
zqG8H^3~M-@8o>1}CL7NMNL!~G07?1WEufixY1Ni(X<|wi9|3xvYvzUFXwyYXun=?l
z-c#l>6!f@u)V-(bB8em}jCJba4Km^P+$`h4+EMpmK*am_q2;Lee=I4zZXC<H=jKsE
z)B6EaSk{h{+F^|tPwE};5vcJ4b4;RyWgDb7Ao2oGvLnwjQP%Zd?n`OefsblW7n>H8
z5+-Et0-M%zZ`ITHo=x2C3!|7?i9nQ+A*xU87buAJ^R1YwR)UgOy-=W2ZkcQ0K{PIK
z(m2|_XJ=Bdzq3Q#?{j!^mEV4~!#%d3{o|W_hJuegxkuYKbbwcLdS&I_GmF=XzIPlq
zfvQ(HG~PQrn%+5__joUS_n<{}D4Uh&$%MA<XzPyNIr_H8xgM-RStCtGKCo&757#2a
zRpyr3#X*}|8j%vXX|#mDR}ZHAt<PTPr=B=qGD->F>nM%pC$28|z@_={+|s)E-n|W9
zxwSs?`sep2{Pa`1!_Ji)d5Ts-KxoWL4dd-n?oB**(jJVzl5>r7?sRUFqkvQobF#^A
z^K@kW_OV(J;ugFu%`t@a<6|2UjBo)pat0&kHRS!nvHi&pcd<=#-RKP=5J^`mI=>&*
z|I-SrLzG)4nPJ^JI%mi<T-lSk@VQpZ3n?E+HDHVI&N3}sfJQES-8#<CGZwip_W&kC
zx8b<N7-%9$GcU~G{RF&&RpVK=jw`!4o7!OvgU8CW;;?cOa)&_GiT0c8J;ul^L5W=C
zCTkb40g6b?3SJE?!K!ualsQ$*qc&Y&R%p&yOYh-u)eS@@tza5E5;v{j6gH*+XPlc-
zRGnvPppuPPfIox-(L~@unsA`rETK9@hn7rF_?_;EU;d9jMg6Pa<?sF2WBhlYIpTXS
z|A5V=r1VdsSkN8aCCf_I>lN#2m%grPn;M(tv~7)FABpCBK9$eyd%;$s(@@?MyijR_
z_jGl`-J=bwbxZDAvdn;$vD=-O_V2nAmY|OzwlvlxD+$ptT7KfW3q1eCKCj$9<;KGe
zM|H>Tjf@w2Vy9#`H(Z+MeBjECgwjgI`IBNfFW+79<#*1CjXb|U;b))Oi!J8coI069
zTKuUVo*Ed#$au#FxYd-DbVH8}5G7;hU5IErTM*6rVF3@d_>?-R6wVKSZm%$<Dk;*?
zZ1w-fMy`h6D|yC`n0|6hA&dz<kEnJiV&obI=3lzD<d~9^g2)6k1e8^r%S49BdC~Ek
zm6E&9tTdaXTuUoBtz{&*<TGkexand?VBLAjOb>bhAq2W0GwVVOb<(_c8+TYW++bi{
z7@9uNIL`z_)dgmS8p60qOgns#*>dw-cBjU98PFNz%h9n!n1G%-Mj5k0fX6<FfrcQO
z1tL-^CCF>L84oLmF+!znHjZ`c*<EDZTXq<&IO+v|*eiuIUbQl{xDW7P)lp^|l}i-T
z^b$E;SmcDJ=i7D1>3Tz3>@%}9+LiqA>ucUvzRrL83r}Ldaf`qIiA(&8pPccvS6}An
z^);b?oVIV7njGI$xZtUlYntT|J}>3oU5_rNV6&)#q&a+WN@&^vEjL*~*LIYP1<JLY
ztXjIZLt6>&OQo3>aJ23yjbJ1oJ83O20G$`d8e)A`bp2pnT0VAtj}Je-C*62nsN+ez
zQ%ZxfP&cikbyC`W@#ZOCe(U&*@qhB%l+Qo27xQCTn+dHKFmLOCG+|iS5E1?OQy-IW
z6RkCs-v%sUWPr)gw$>gw4RI6kk#oWjRx>yWg>e4y+fKmYOmP=})Ud~Sd*ZTqwt2fA
zy>=X|B>Z3~ZIWc&<llX~#w0n`$MoQAiljxlsf-|6X5>rFGm)S6h`|;{k;NODm6j~l
zYaIem^bU4PGa&4{rD0L%Av9J>8iUfsYL&);N!#v_d-sC0PcqFcQ(W1}P{{6E7;zkX
zrHNy*L`bv`>EVar{o}Gn=cF7rW}BtZDEO`BV3C=jbFFiMHi)hPz^3u!MzM(ZXti9^
zs&TZPr|u+WDC6tzE<4)JG0Qa%mL02B6m?XP(qh$lT;JoAVZOM`?A&t%dztX?{k-$?
zbNp8?-s8^apP>DNxA=4gU-;OJYW6x+wZ@jaG_DuGFL>IvhRF;EbX|qny9m==ytQ~`
z(AH8+3qh%^1#2mG=Va4T24+qtLeb81>9$Mj!DupT@Q4-wS`Cw#Y9~AF5OT-wwvVf`
zjip$vMa!>`Q4a_pn;RW_6byxx>cFGM|FIqT{QLHYkzXnTZ#t$OA&5bm7)7MK{_PWq
zI8v2ooO6d1dXK<4#a8K9h+iW-X<UMf>pg@~d-S7+Vo=yh0a9wC9*6yprUKh^AZ~<l
z?I!1vewU6R`WZztDN0W3-bTlz$N#Rk&jLW#tu*u`q6_ii44p}&<^cih>OOGusA3X9
zE|qdIC4k-T#0Vch5ZQ0r(c>!aJ}RQ2Z4?RdNI!eyB(2z=+M$QJ^WweJXEaVg#T2$#
zB@U={avpJHTe*^?moE=fCJvQ0cHKECyJSYowZ&QUxsLf;tRYP^LA}?lXA%WUCb=Of
zxG>L{ms0yT@ykr$;DwzWqu{CYWvr-(CPMFo@G_qkTs?n|S(!6C*dt#&LvI%JdXLpZ
z&u_kQm({=fEb7bO<;wSKvb_>>@qlIZ5brk}oIl{%$1hUMX9T10zJckSu5EEn^8c=F
zP({Y&iwmZU8LoGDpjy`y^9e_H4{?3Z6Ib^s3XRf6N?b}rHjB)#?mVZROm51p7U?o0
z_3@-8Y?M3zg&1dUHDxBwGHoKV041GloJ@r3J$&c(k}tiv?JWTK=w8W>Uyk2mT*xs-
z$0Nint!2`WM3eIK*KRENCvTqco%<EmNOY;yb_}>tY5)zB-ee%oIfxO=fVQixp{2;c
z#Mh{}wv1Nd+@iA?l$l{W>im(NcNoyh2)&QDm%1`jC*LTf8JX><u_xN9iRsx;*+?HA
zuRDaT_{vy+!A^5okcAPHJ+;-Q0WB#k2H2k%%1m)#UXUA=RIGWh7A2pm^DJBG0$jI_
zOLI%-!vN+f1c<XrE+P#BK@p7+ZZA7_3d3$`1WdF_!fm5CH?!%fP#VQ)<Jq4YO!RnF
z@0nVW5kLFb<f+IoQn^=9?398#<gt6v`#54NhIMJ!X440%wx`UEu){nYRgT^V_9jLI
zQk;lVPV&G>B_b2+))9h1Y2jK}CCxI@$!Iw}zQ>IhZV)EdIDY+Kl4qXc+_U71r|FiL
z`P0|G&dzMfwZHOyK6ih~fB)h0{95zRcxiQw=5)r}@BH7qee(dl?kMf2$ZSr3^bU)w
z&x*YIx~ACO<@Dh(fA-c*wD%ZeneEOwe`$}mUVMqmANnwxyZ8C(KmF(IT)NJ=^MP4m
zDXgOI<vYH#$T?bf+*|h&>W4s)8GHyV`#{r34@Z*e(nu;Kb#2ew>Y-1-C>i%wSaGU1
zk81wl?Xx-m#}20a^iy-1WgU%1h+t_aON_rnbxp?gwKq@r*0GctzjS*=2=IaPrD#W}
zSOZU+9NWmx#Z4xa=!}sy--ZxHQNv*9AjWRhwxu>3)f5eCh#*u%{2wJFxk&mEquq*i
zcm&+ej5^awJg!)zU#AT))m${=!5?8@GA5K$AyIrE^M;VB9!Z1)gsRHugqhwNKQQo~
z)AZm?7h=6%9_;SK()EIr%WOXhv)7TWXS8Bbib8Gg0uR<bC?dMgKtWC;$1@{0F)Iw`
zr;>vnt$R8bShb#$O;6K#T#!K{)BI7PGP&lak7dS8#oRRvDnotr+DV0W(}kg=muSNZ
zn3aayh!d$@;KJO-@#gaHWv-bO!VkDO&zR+!3-b)6mC&knYI&7bSET^2QGG|&-=~;7
z1g)^s$LPulS-y)}KE*%(qY3}(8yT;B;xd<BJLd0(DepUX%>D;n;II6c!)LG3@Q}8D
zK-aEm)lKTv1KQ0oUA4sZE#TrV?CF~oUP0G&I2AC~;?X$o=&Ci&d)Wx0B-%+MJO8<z
zoIb$bEMqnit-K~qiJniSL^vtzFp$I;>CV(zcHZ2|Ooyfy4upxX`}&(lXY>U>c5cSc
zf8fFh4OfFGNfL=k1SK_u&x${|amouvXL$xFqL|4R7)+m}V+Em)44L%vAnmQUcvsqm
zQ;sibZbJFfxnqi=)PO0al4gibBN^y8y-4M`WZX#<b@;`0C7-!3S=*zM-}W*|Tb`UZ
zLZLD4Pk&EUIJ-YF5!gtSPc?YS#`yunOCLzQ5l?InMW(J)op9kNxugcIlZUjb96O~r
zhDq$-w<^%*2Bmxyzfc34>%!bpW}0Q~*_)a;`8LpaQKmiE$p}G+6Z68bZoNztwt8?@
z<I0QFeABQ*MQOyy%S^>IM#4#jR5?tLK`AcC;z_E+S<LOaG^IH1NL}ZdP7U6R!?cA6
zO|BbHIWrtg%z*w@(YYxlo?d4qYUc^O_Vo4paMr?nK{+W2-IT1qLcN}|pIzl2fA810
z^xQt@zI21nz3*|{`SV=7c)(Y?H70=5r*ytWo0gyqE}P+sj@~OwIioB}Lhoshj_7-j
z(>bPC;0a9j7r=Ew*Wp}G<G@CMc2-)f(NtZ}`FSp>Q!hL2`FW0u)b50G8>6vW%C!^X
zY=Ri7PIDu)b0r|!QPuP2h#a(+S$_Pf{XtnMM*O6@I}tyT?^26KV1Pe(^Msck8BF?x
zYdd`8!sHAk7`Eag84<ry+Z|G?t<Ikw3}$O|Qg07rY#ZJZA7E>yQk6PmGkV)NboNkj
zDsUKVAbsvo&SPZyqsl@$wu4u;<;)c0WO~~_GyJ;(9kyD9cH6EfU`o~dtxzMv3t+fF
zx!9vk$3*u$aZpglaLyRR(Yj-|lpL~*`hEFmm}lMjt&CTYsNrxejYK-;@<lG_zScsY
zofU@tiJ|w9=PF8y$%LW3sZ=Rk5Nd82yPdq>?kGND6zL%-rN}Kq%945uOe`z*PDoFM
z7W!{y#PKLA_q97Q<Qh&IaZqwC!eARe?MY@isU4M$)L)|9nTF%WW1M4Wcb7}gJjL6u
zeUHhLpCvdj%O!2y(4Ib|YZ~&&MIJo7NAd8J{L1}T_}tUi_{A5#$JMLn_-(sjQ~w<M
z`HH@7C}w*YeTCI>N#S#h)@)X5E?vLI-eSVN+lT1QhPLgv`_>)GgY(?K`4D3*&wlV}
z?%h4c`=xL#kRhU-hnvfeP2F>SFGt1HX<0ie7uYRKEUyN%mO-5wm|!qbY3HQwDJ{da
zAg{pnfj94;t@VHG(oFjGeN>l@VX{KjXw`)wluv2jAHIJ4uEzi2i_?Jxm*!ZAv8J+P
z0*GTsZ-9_`0pvZoR=CkFEcO@W)>K+wVIUbPb++|38K@XvjGq13vEnMG06}dd3}S^N
z;RD+v%BGKYIDO`1@@J}WXeHt}Y1bD?cF<;vq3I^h=}R|}%6FL-uh~Z0Y`R$=y^t}Q
zs`n!D5xrp73$T+S1FK?pUd*+Wk`Oo`!$KyRh<vyp^X|(SJ>*96!$~d!MXJv8&S3>X
zK)_`brA*gV#Mi6%fRYe}Qqt&D$DKFCFG3^WqBXL^C$B#yOe|YZX_HV!x(|6Cn=Wv)
z?m4L(Rp*EI9&g;R5|*(dFpmPLn=YjyarZZxuI*S>6|V2_r$^{!gAWSTc4(br@;%zN
z!`c~LQ?qk<!&6UN-v49I@!<G`g9k^<caEvEcjz#jtR2g$CS)_(rp9$G&KTC4h6fK$
zXx3}`b`wc0ini@3_IA)#vuZl(s*!PI){q&EjXev6HO$L6hC7I8tBXKh)hB>ArbAjd
z5so<hDaf!Bhm?|&WSp^@vz$RVm=)Bq2A}A_M6^9o%@bJ&rTEIbOeOyLYYRSdX?A9x
zOkJsIH{J41!eFF9Y<oK?IsGo4BUKa<FxdN49Z2&HWzDBBKjrB|-9Dl3(NP4%W>RXB
zw&&E)6c`zQx}JCqVe}c}NFn`w7y*{H{81#Mjg&%tOaXKe7kL)*pS@DR9UC8XVfDbL
z&Ww1-C(ccId@mbVck9-(ZX5^mjJ6kK-x@()3!_*!j$6x)rVp$d$E}l=w~rf6YsV~;
z8vZQOiW5TIc`5Tb;Qq4X(rzqcMXVvyidEwVPpp)ZhLRxHBnw4|L0T!+wd4L$fUMR9
zc5*@c)B8gROs!@bDayNrDD+Cipt*h8VPkdR`}Zqmxn^D(b_#)nT@WzwaNV<P#LK0g
zg{~JM^!~EN8aO|-7^B&o+~?%YKWG2qRW?UASsmYH=jtUUSFfTkU1oPSV|w8-&wcze
z%=Xq?-l^#pJ-=dd{<jZ5&9nC&@(<qs2EY8*mZ;skcr(NHC!Bx$3TiS%@9uK$$;WBi
zhI?<mNp<{?FxjEpU+~<=pP^ZwGTWUqnHGHSkG_WT6|*vzDMFd_U(6?#{h4KNnsMW(
z;jnTnTSp%ROF6819&Q{*wd3BV=eY5#y1>0v$1*mW%?dNjahT^a?9?j7^@HsYP=Mn~
zB)^MHvxrlO9a1?+I6$fxeeDNFyzuB~vd0&E^wLz=dC{@gv4^E|fv>!|WYhcT^utg&
zFtNPZdl4=%aT93$mP|AB6hvcm(FvryJubrJat44rKn(Nc?^cZnaie@+uF(%E=ULUk
z6hX*G7_}{5rqbo8ND^POWhI7?dK|XIC?D;_CIN(WD5Z$fd*r<L(e!?(1cU&`RnO!5
z1!ZYO2}VgRwCw`7PFm?rRxm9rw~uQMW*P6hFd+p*g_VgwiIRJ6ZXtdFlfp2|H6BG4
zMKBdY1{#kSlevW3v1RFgrWxeca(CGahM^g~N8#{MzFMjgI)<H98^gm`^S27JO!MTq
zasYy?g8QqURpU6ATHZctxVXr<IJY!?>`}MSG=ZJcMh${k*7S}%E6EQoP@FD78$Nh#
zpH3HSsvYVJ2OO<7JbP`Q+po>}v+p$g>9=m7!q20$;qyOpgAaf8`+Q+~jhEh^@q_Q)
z$C)F1v0(qHJqm5v-6?2KPjK}bJzJosGxpE#@!0tp-^~lo?M-M3!`sd?+1qDlNAwqz
zmdV6EILVJaoNjtf5hBx0p=o-}jr(i7^IYC7q$OO-vvVj`wPP`{EMoJNK@B;w3UK}0
zl&jO6TahXFtvgFT^VB{T=%PWHl=o^$DUE_Jzk0}vhi8rd3y<&e(JQk!qaktw;?CST
z&u_hU%&n&9Zr$^5Jh>3wN5m5ln208Z%bD4h{dyk;y^3VGz#t4d#t4ROJ8YY660s+R
z^&p>>a`*HsQY3nIV9xe^sZU@G11PP~KJbXI7~Yy95Y$$XB}E@TlIU#hKzq-u*n((^
zZC>-k=T#i8I}VD755!1ef0~K7sPmLY%4)%jgSfKDnHPpSmK!h3EH{svi1Z7H<h_(%
z(y)+D)U>clz`B#lQRGIt6_qp&^{uB-V$h0M))@gXN`+x6tTlQ76`cpC!<GOPyQZYa
zLH?WOq8gMu^``AP?mg4O(DuUpu43a$*L!v+7Go42ygVTwVw^4nn$ELR8rDdJuxUjp
zH*_9vC$y^I^u`ZpPVeCJ1#iCdfa%VHHxExy>l57R34d5sG{<ZDu!D7AruR{O$M3D)
zV&{w3x$@=P{0GnWeD9Y!?%mD!-LJpEJTIBd4>&ERw5O+dYw+tsT(#s@-_SWv+iZCK
ztq0VHN9cTy<>{J3vl@nn1TT`~&Pf`zyHiBAo#H|5$Q4}JE3rYQE9XWe%yOeRsXA<Q
z&`xdB10B-(mH&VG`aZw?y}JM$w>`iA@;!d~x$`&}nrPBD+WWv)e{jgR9@+VS_S%9E
zj7Ey9v~U$Po#VH@f5h!5WqzmW`Tk+eCoa!K3j&5@JJYeFPf28811O`14P{*OPdlt=
z128xe#(8_h3=)}WFx*jD4|$1j0Q45n8(Z9{9yt6qVHzQBR-<Tt_^bhQP*J}?C5f_a
zv|#7C44-g)U^yyR`4H=^L0E6CPehkEkirEf(HQbfQyAgbgY*xC2uAc?RDY5}aA~Dn
zW@2vVyx<DURu;}A)6`1HLRDn-&2r(SPpo`N-iM*wXtk7tirDbuMv!a>(&wPH&`?u7
z*s4_170)X*??@&Jf+^IUN5`^VVT8bQS_?p0W}5pek&d=n1ViTeRsl$9#j=)EU>-XS
z1(A(qAs;whH|S!<<gq8wrz_0)D`eG*dq=0V>lV{Av?oUtc}aWpfOhs-boT+Qe?X{?
zsls)>`OTa^y=?doKNWcYH{a&p-g%C%?dSYoo0r+N@297vt~zq1FngDztku-i4j$Y;
zL6uY1Cu?k;^VFw4!ST_Ws_CeD&je9<7R*JA{rpZ&rr>bhky%Zq6jKGeg@)j9nRH3c
zV^7Aa?kO_OvhL}9V4lhMwXD62iT8oWFYfUC;f8PBlb-)?-CuE7_k8@)glk#Kg%xid
zR{X&mN8G8}Gv@JgkMHsP)!7g_dt}|FQtxlQwl#i$Pwba`_S*bR2`yCutkJ`|b!q*_
zb(#P^qor2>QrmpYrALOJAZU};A=_@mM}fU0CnhQP*e#5Gz$57Z3_S<w97(5O91SFA
z5W-d_KSclnt}1L5QQM$kQHUsLqT(iRA;u(Eu292hr#|W<Wm=H=P3zg4$WGtK93CaR
zV5gLXjn30FV$4d5fd(T$-)Sv~d0{oH#z(5Jl-!EQ{#$jPC@w@AYgR;5GdGchkseg^
za*4vMVmPFfqKooeX~NK^3qpD#cKCUTN=hqYBkA*<mp+AqspalzE91)w!!*|-&Y(o=
zKr6-G)Y5gH!%a_aG$nB7xZ(0{A+raZBUCjyH~|8KgfdxE<Fky#Y)0F*<QkfoqELVU
z03ZNKL_t(;h4LGyPsr!jXxof#bAfPs#hV}N_?v&@v%K@`FVLPh^jAyFtsmgJ8F^mN
zwiQJ%bl%A>Eg%UVg_#x@giaicMQ2t<d>ffe)cb(d8WRV1q9sSy2pD^QYQ;#LqK*}m
z9_WJ9`dxrLH>l2u(yx;ZBU948td!>GKYE4mZFng<fwz`5Z!K#|t+;aY4!76s8LmJ&
ze)jqv&tKaSP?ZtzrX@xF%~u|Bdo;zUj2qK$zIw#1y60+X`2e3lV87r&$uoP!&_EVQ
zeI&A0MufW2&Kiw8I)rgVgOW<amQ^diRIwE8N9Pz&dljuzac2EKdZcth#WlJ`{Syum
zWz|Q~KvMU5R6Z=F6T0fcpsCl!K@M6ADlcRU$+2^z=^`?_Y<lWWlwm>$Okx?YYF*6r
zBsX?GM4doAfQky)LDMW!ts?m*4h#het4?aivKCD`f(UaQu6tIslP^068a?K4Cser`
z-Rvwkv~f3RqjOVQssL3VhO}XtNpzx>c=a;25vP`CqhbiF;AGR2Tg8*-OU}<T(a8%)
zD&YgWg_eGW7$xmbE%Q<;8t3Qv03H>&VV*0xuEU=kl7$wP7pU2Uysqflmac1XI-^h(
zg*n9TG}K*(M=_uF?C;Kb?X3m>!`nxE>A5|2Tf?usujJ=`;fTKdHm65#(fbum{eW(L
zpY_q(;3}-Ds5TEkY1(Q@wb{_DD=-F6pz_2KUz)OzNjx?>hRn)nvfYWOH7PCs5rMkQ
zC}|xwGLqb*utqrlUP;6jl#qo2(m(L|k6hsk&t71!$cFE(_JKD}>vuK&pMP?n=dbMy
z2)i-c(PecP_{~=i&lrC?-?xryUf*=oKJfac<Db5><U0?lXb{_J$kCRg;cYqfshW`L
z^*>Ze7z|i|!bI{()0KugKU-i>#p_OgxY6gF^#&1Fh|$PEvreyN+XC%;YDG`?u>H&w
z1pwH#_n5*C6xTX<4%a;@?(9YE(raA0QJ7>V)@41%8_98f45?PFl*POcY>emw&T>6W
z@6m`ynrDi8>t4`SC#pYPtj%Xe(e_?21fyv~p!INmW;m)Gx0hWsR)s<sXS=1Qh=!hS
zI0VrfXd+ofA&CAuMi&Vgm3an=<Hj+|WKf3I#B@kSu7M)BfX+15N_0@Qj@*i4lxatj
z5>zfw#C)+xPN?_72heb|5xoJeEqezSs8%h0zF?z;PMfVO>dlhAK4!M}ILh~Q=ikpx
zcM*EU+MN(ofjzj!&Xp&4{Rgk`r+@Mi+5DKl|KsQRH*cKs%ttQpjjQ*0@M=j}z7I2*
z<1&w1-etX9QLm24vcSp7ifLI;?j5jNKcKEQGHIA?+9xoe<oUsQ3F{kqmPIC7h+1o=
zxzMJy5)5V2Zo_P2)X*DZ;%1U25ocPeOV2;C&j+vW^47yOcaPSD=9Eg$dGX%rtnojw
z&&RJVWXZ=IT;?)L&i|X=eZcKU2ABdoJrh{_cGXdhkm9f0S@Xnh!G38ZEKa#R$ymbR
zToS!Eh0{ZX5P?09fY?gO8W|ID5)UHv#%Q*Pe<~%8nSEz07r71>iTpC4cdJgFj9t#}
zhlaHwI(Q@kNMNI_d(ZuKKXe`KO)ckUmea~n=3?kw1j`yHH<wyNXL*|bB4P2ekzI3M
zXqIhYY85B7qcC72UrNOX;R8%_BcWD`o`NQyh>X@wVer7ws^iikCj==kPIH4*QVJU6
zn~>>1W5Z9mT<e9DG$sj$AiH>GHOffp(0NG}Hl1f)YK}IJGSlo&3{@xSv$gUp%82?b
zVmomHgM5gMN?s(+F7L>0);d4v7i30JJGeBpXsuZ{9ZgeXjlmm>c8*@lI4_ki=&}se
zgYgcpEX`y>;SFuG;`n&UV^3MK!jSD;!=Gq=_x25*{oG~D@4do*^4x;oT|D50KVMT-
z*O^#Px%X|FlgG(ILEkH;_5xL-z!+3!2*I=I13ShL=m&hJ>wT1IGX$k5bH}W-sNe<Y
z3J|=KQEkyJ&kITA=7nL?xQJ#8$!gs?@*unbON_3N{)=ZX&3R(a^3KgAzw_2v!0+dt
z*yj`1ceh{UDD-Wd=QrNP_&;$l;nP>=ED9N6zG-^C@y?PLN7UgL?r!+}6AOsIv1XgY
zGb-H;A$7E=EE35?c1x~A>4@}ehHFb3`PeCp>-Ay8&XW*-YqVQmHD(~5jZ)I<G_nrO
z2xyelCiWmaf;H%!TnITP=6qcwTok!JGaa~fGSV6?yR%24gM*o6Qb-%^$)=-?;M}_E
zIjLRrlxMRVtz<~Tx|N{)q;{+t$ML48YF(UkD;It+&&Z5`b*EL&;iiw$U}+A4JcVWL
zBJWyCM}3S0Oaux;%GaV($gJkL>W7h3Mv3y#s&;rUl;k3&0BO9{ruFo_^t6{o4dLui
zsY1$#im3k6d%;QKLKp18v&c0kb&oZgbMq3fEWRvp#^CZC>pj!SoY{PDK#r92G*yTC
ze<*viAj`7rJn!3kpXtsyS7i;|RXvaB1_2PnV3HsyLJ}o$0td_T^kU0)*b#PE_JblE
z(t{ryp*KHRR#*zjvLsRxERrHA3M4>+6M!awMxz1rRNY<kkeQX4cRthJ+Yf7>b8l5s
zmUJ0lbya5Gy!YI*)?Vvh|N8&=9K0uvQ~JXJ)<M<VAnBb!?>xhmcb?+E|Hp@X>%-eT
z^Xi1ZvIqMw+#~U~iL-Zj=J_Sn=Id1M7)OuVIf~g76UXF*m}+)H^6g|X%UTltAd~sM
ztc;73abZ2in9v%j)@fMUNZBP=R1T*US!!B>mJOAd*HXnb-h#$xsW`*B-7EKK6^8*n
z^XMioKe8dEv|fpBqBOJ0@%ir_a%Z^;{!A-=>ak7!%oCdovZ(EgyNTgPFK@7)MeX+<
zESq*=ZS!y~$eP)IL)0O(9t|&mwX%98GV0qFVa5Lk)JpwVsWrA>`*k_(+r*<*ano$Y
zHtj^?qo4xz(Y%Ico3PPJTzfmMBp86M0+JnR5yyl^5-cc`2C__KRG7s|mL4=@0pBq3
zH2qAhm#p*b4r0#qBQ`ow7>%x-z#(sKb_~;{4V-&X2ihCT{=avg+oz>0Rsf?;#IO^w
z(J^6uyV~TctY@PuGj`#f2hDRGgh6XzH%3M-Xg3y_^3874vj41yQ6^xX^N4bBB}&3d
z0nUfbMHR$GJ(GY-h{35ON%{PwGFjH1+_LHwy{B++G%HyYHA!UH?8U@c#>Q~K-o`L!
zPkAcqu-f3frZy4rXu$S`i$uG7C=(N<efBS0<jnQ}og~CjOqyiqbd$Wi%&)$Z@x@OM
z7`}O*zd7jhqd&3W%%xLa+<z3brAo$BDxvfquRMj<M8gqr))5)DNTlAgG$KnxwW&hk
zw19kBtg#`Y8lsLdf!(Kr(25el*w~2A!?Je7u?hS9QZ1ID9MzF-u>r5d_WZ{#DfW9}
z$nj?%-{A)y+mdVPn-z0m7cNT6zj!G4J4W+!Pw(*bUQf0=^?Ls;ulvRA4wt(zKXYk=
zUwn3Jh2po7yc*dxNxfMmZl2GteP0{Y(8{MpT8AlsUwv;Ik?qTEbtsyEZcq4!56`uK
zYZwFV)KjUIOdFD*Y>3f@4D>*I0YqC}S-b5ex>fFtw4xK6FhdW0ngSikny?Ebn1(?(
zEya6nvlF#1rcuSQUiA8eD<E@tJk(Z7pMGXqmEF^&Wm>o(Bc{2qDbmkZk4w%DW3ot5
z*syo)g|3?##i(Oi4ntdB1(?(Txt9q=5RYju_9&I|ew<rInL#N?BExu5Gsq0f8U~T{
z?e`a!PAucTSSwBzHK&W3a~mnvim|HmtJl`_h=~adGrvlLXGST$bi8^x=9TwuP^B5C
zNB4>An$7J^bh633tf=amtu*1>){rZ2zRRhtsL~EkKl=oES#$m74es2%1Ir~<Rg)ws
z*2ip}eUf`ehQIU8Q=b3yW2C?H9X|K+C9eG7(|qOa_bFyivAOv%@`G=X?Owu83$DHS
z7NfK0i5fwWM6*3iSeBM<6j7#z!$~2N9p_PA%GqfJsTL0j#cH1<x6suh3{nKUW3M$y
zzzmW|CNbT_;IyKstmrZX##<6;YUBCY51;4Vds9AiaRcx~krAtKq!OLvmS6wg!2{#L
zj?w(wGdr9aNbA^u%g(Q)w-c>+ey_*#d)+|gUiI`^3s%+C<9aPtn?Ai6`8KC-6Vyhp
zm;8-fS2*GQ>TvKrfQ`QC?bm+??Q}po<TZd%t(K3Piov$kZTMQKj@EIC#?=@vE6$#l
zQ7C0hTlyP8R2v`!q=#)^7}7I0)j|{rf}Yo&ZmQdnTpCND#NmZSTFU7<@W*eQ6hul?
zffT~(>xw@)WpMkAjV%3aW~QZUiZzp{0wqBVqDk57n^7uUus9Oa34mqc0!@^58W3wa
z`6DCmVG%$~rC=kKslYUM+?|x7U#Jy_3yap0pqz)><BC~fnHH8wVcYSk5MpF9ay+Xz
znpgB9#b(C{pjOw^r^jOWF`Y1<&)C}>^7#HXAG*AcQkq3sac;ARt1BkUrC3!N&Ah6~
zPfs|xb&bjWJ7^P8n~2)xDAUEoTjaMN<L}(LP5xs~u>9r?KK$)dKGxfyyBSfK1RGVT
z-T;z}(|b2SX`<NBOC#dQV2tKqCVq5IDZJ9crVO(-t%d4*xUeGk7AgP}n5EN7#-eF#
zh_s@g%4VSpArcu87!?WjpukL+i6b{|Cyn^%<;@3Hb)!JesjTB)J{0_!(friYJM0cq
zKoU@cj5op3+}cX5L_I3!`S!t*Z{C^n%{vS3&Z{<IxbV#$IsAM`66>e!din3xo_}4b
zT3&njdQ(8F)dZnQQcZBLZy7evwJzonV!i!2-6XJ825Kh_4d&2lB8C?fiGr^v*eHbm
zY->?!iM3>`#AsT!QOGvOxd3B5<W#PmrxOWXb6$ID@5J>s%<>y)Fxl&dt!@NLwGKR5
zNx8l(oHz*uA<{+VNE1!RNTsnTEU^*gpC&suEk&b1T-6QsAj?3_h3bU_mDloK^2)V0
zL>aUKJ>p}4U-fcvq}bm`ISuVVHwcI%T5QyJ1~HL>!byUPjqD{fAd*swy+I60mYH(v
zP|l&En7B8D*^Cl!)snYv+@n9r8N`M%&(YTM`g^yr&dI#P!y7m6q3aUo9K+Er^YLAr
zbC{|^+mhM*luk5b<HAk4s^*<-&0qSDzMt>?ufM_LKl=hNJ?HtOFTX<KG~>w(RW31E
z#_{nf%OnS*m{yKRYm!J5g)NG1CuZ1<s4CgPm@FOJ-AK$+g1~5^bnAy`RPcJ#if(4W
zI@z!cg8C{{msOYyWwEBLT@Z}*t+J8?nCu@U0fftI$3K7VF83Zf`9JsE9{a;o9HP7r
zQwrG-)mo8;<JAWO+Uc_9>$hjT`p`xp2XnR}!zXqVE_IrK3ca@XD?FeHDprxJ*8bk~
z{DRuAl2H@Y4T=wI$;qz{+Xrg)P-%E2EpjVs*ZfRK6zhDVhN`c_b(5u}pb*?t%HGQk
z{4^}rG*4jts&t){g2MX{Xc`IFMJ3$vUaXlGmRThqTrh71Xr~Bia2+yA$TMNBT*^TQ
zCkspE#rErTVJU0ZuJYcUifsAetP;C&r2;Pjl$Lg2S*`kjM(5pV@hU{E`X{9V_d2vO
zT1(G=yr?;tl<hux?P-;3`e7G9`IXjx>3lH$P@#H~@7FYHGm_{=sVo`cS@n$`(k#Od
zH?YeY3Wts$nlG52oN)K<A)Un>?HxylC+*3woStItAL5P=Ns^RexQ&Xt)OCdxk~w8Q
z1NSZ)7w+@Yg&n^B=579u@4Z9!zx^2gJNNm|mKBeE^etREW;vfzlqIga#d3B^UY0B>
z>3T~K{Z1?lNTnEdBL<zQ-3;95MxqxGi4Q{4-)FJB-*$@R6{opl=|msULwkKuK|(AR
z-gO)Ywc)_;ZnXDw@MWmOq43XMz1{8*w88)U8P1L}!GeNqeS_IG1h7UPyxGnxt^Z)~
zhsSYk`HeeEzIZ1H$b`UeY~fZQ<XW$;Xcv*1N}%-}SUbHN`?|KpXj+N2?`bdt`R}&I
zXp>9>GB%ZObJ904tMuCQhI`+yJ&o329{Y3u_21a-CmhZy(kKvgl+ZY{u$$k$CI!dW
zUL)Rml1{fH(qDxOy^oxhod#?R<Bn+_7H!%x%D4SV5r(%Xi&_?YB26dk^jj}|(KxpZ
z5&`F?xl9Nq%evM2YXW=K=LtS>boDjPJn}qqYuZ*M3g#SZ>jG`RCCtB-m9q_;gn8+R
zj1IyitC3+RR_%Ll$S*Er;W!fhgAOQnL-&l0B8i0`P!%PUyKm!0Pjd723B#QYws+2u
z#|hD53B?lUL_EWlC0qM@?C$N-9rP*YOLUa6dw!oEeDM-V)?r*KT$B(Gcd+?_I*+(N
zJK#GvZ*l$l^IV-h%^yzh@qHiLX7Q~Xys$Imm4)MUJfas*dFq9inM{_P%f>wQ#N(K3
z*k13ulM<r@q)m+kG~ScNrmc8rZZK$%FgcABSu6md)&dwe$axVK#F~|Ihp<Z>3hgTD
z#hC73r$w}5qgj@gU;g^dR%*Nn{-1y0EPJC=Q13QqRC^+-FllI7f#&sgVohY>`dprG
zD>j_XB4M+JZ#=5u=D6gEjkM*4h{(uFk5EqTRj*#VMyYt6hPyJ3X@i~>k!>B(Z<9bs
zU{$fiL|J=3wV)y^{mv$7xHhq<u%{sZ=hiBV<}yVup6}C%HH%W@uN6}I_rvB|QG0Hm
zl$;wS98F7>wc~hJu(2h-Ad4latcOJ|K$-mAvsj8CND{JrtXczavPj{)$U@&7mn=%h
z`Hcjl6_cf9qiX_5r<8fBhmB4YdM}aGGD=2IbpVyhLM};BSzH)1>P!MsE`)FQy+Nh{
zx%DlREsG6lB#C8Fg?;o|u1BObi?G(a*Nf!C?MU$JKyva(Ti~30a78V^V=s~A3Xf{n
zrDarE6XPM#?2l3`F-Mab%GOw0qP4-LeN+^&m4Yz^-6Z2U>!Qn&bEBAz6b8K%Iw`)c
z>Bo-RI3iu(Panmq6XK0sR6OF~?H&Gl{dIov#Z6xN$`PO1O8B+(>vXryvpJk|&CF0*
zv9ylNXi5*GNV79cv9*w%YA1AG=VZ69(Zf?lNKDQ64{NEsA__fzVV#UfqwtxzV4GfA
z`f)f-oyRGIbB<*h?oh3$f=ruto-{SQbZ*F<D@Sec|KdyMI6F)vm_ibl-GG%7u6HX(
z9zuoJl;ba->!TH%>BZtX=;gEg#;rMDJDjt0p4&yuuf99xryd{CiF6PUk^86-)RER(
zP9y;c<!g0Q0RYxE<2vcZjZ$duF;a@R#GUolfRMnN0~7)uAx!WZHLdlLyKn)*^eII9
zIwsK|hUTyTFMsoIXd^lRQ8*=?3n;!76nk$N6Kfb|5%;gX#_-&;m^c*>N5f=k(MIOO
zvI8JVAPZUo%^q^Yf@&`Ovi78r=JB13EYiV*LzH;3NV41@HRPqEa8eF>Cjq6DB#I<z
z#2vzE5WEPEu6>*RM_wd~<JSDN(FXKNN<SIfg;pU@S{v$S8QK=DSK<jaoOT*%0~)Jn
z?d5w)yP_i(wQFtF8z9L+5N`xS&FkO(I=}R$A}hySzj1^4ET>NU)V)4Yr$gkR)Ct4Q
zA=lr2gQLR(4(?AeS%<3E=k)d+-n@FqwVQYG^BIvgoZh=fnx+DD#yi-~)37+8Ufe{H
z!s!Ozxsz~h+wp_>lwOkYXP>J0l{df6^nS_XTO&UF+*9-h8}0H-CkcRGAsmm>rRXv?
z?7OHfUsbD>7aLIx3WXouhuJvN$2a7xDx9><h$;{FL|&HI{FLr+R}gLoi&~b|E^Kz`
zD#gul&gVXKj=f<jK6q{w*uvZkTc!czTL%}_S~Xx4oExT$I#G*|TkE+vN;#K!u1qSK
zI$6i9NyYO!o%TLy?w|)e00Pjq?h)94bj|v@U;mi~m}{OdoZ3xmp}eOqmqbbDgTDuY
z;_|BPl@>?0)y1mfSqZ6{@R>zbm`~3t&)sp!uoH29Bj)~82shp;^4c;?4M+1zm~u|m
zVw(!#%pk&}xHYNR>4{ahRl=z+YbR>Hinc9Kp)?t#n%v5%Ixj5Sy+}rbsbR9T6iz$@
zH#&xS<pUXMl^HjCQ8-<EFoF!hTRTeUF~+m7o?)OOM|xFy)z-==Gu(71i@F^J7C~N2
zA<IciE9}1_7yZG)2C+~jE=Ju%(=m_*tpcsUhK!9DB8;pt%kQ37;)%DjIp7n|U*@at
zb@;I#_!Kvf#$3IBN_08~Yncy6Ji4>Ny=yml{Kbzk8g!|g<-70QKu3m4FMNnkJhsD^
zzk7o#w-0fV;*qn@bA0Croo=JboU$s(&U_5Bv&VEi!%Zht<&<yV_}8Qtp5Q-!?Q2|D
zPm^B$A7M~H^tuEZGQ&TZ6x^MZY{ZI3HxqEu=DmNgU~dr9jX(ud**i~`i2D#uSf$pA
zAc~_#hovH~1as*|nk0ryjEJaeCtzyrg<B9RJOQhqixIs>;H67jJacYDKNSHR=RCQC
zjt;86R+4xYHSL$?_BiL-nu5^9QOe#RX*D03lhJ#H(FWx`XZsO<X<zeCZ%O3bE^5AV
zYsM$e_k+(wbEx^1M<oPfkJV=`RzfSuut*D8Ex;3NbANdaHdF%Zh~NqnSb>TSGZEh~
z3PZwij@amSlAs6`LKty+<j=@^IqAAFJieQ?nJ7z|Q(S!K6k-Y87;NUXXK6)8U=eoT
z6Akm+;etAEV#K-SWNCx8z9z3dlcHu>3f*_3V?-lARzYQ0=#0J8ka`#-nn@|jGL7)4
zSX?<4m183$5$6Rc@?K65SsPX)i6kj)SF=`8D-t6V8*Y-le%x9!cVbNvRy}8xZ$HF-
zDuyJzSh1*A%N82XqLL3c3mb+?i9}j#<nKl8Nm5~~#YTbCOr|HidF8uwa>qBndYdBZ
za_;O|)UeBZF()cY=6TL@PoCq-n{RNdcBsJyq&<d_;o5swF!Mw1y?=)^?UE#Y7Pqgn
zxwXy7c!|cd+0jg=$GGt!ikiIg>~>4C$3Dh~FTc(g)eo>yO!@KYFL9;Up)P7LF<Q%V
zR+dCOb2g?ZEl2a3K_<9R5^Lf2`@nv5ax>?J#ab`IueIlNQ4=fZCkF4tb5h282Al-5
zMU;~Tlofr&B+`^B1ammbY8CYTLnA5s`yRYhYFVTMULfFc8gBnZRr71#J>dOm!2^%m
zuoXxAnad*{*~kJ}D<J1?waByEQT))EK3~2kI|*OCzvQ`{j<`XEAXFhqxzIjDp;8n>
z8r5P{jY6=32WYts?xn-O8{Jd{N@(F~<?R>F71_cJLGQyC8p?a5)}Pxqxh?Pld=O`W
zs41PYKm%5xhwX<l+c_7;Fp9*;6QAGgf>D%SPRKvGK4Wi~u-%QgGb!n$3B5$4QN+gc
z+RYgkHd7X5CCoZJ)578UF=vKx+q-DxMH?`r(QXh3xph+T=vInSl9c+f!Y2`TCKX00
zMwtoPg%XiXKUO%$W@1>BmfMqxyt14fCin_=dQl5TIp;`Xkr2~iw3i35&!Lg=-odP5
zrzcBOi%JqmtQAQlmC5O{#-q6JJe%E!L86JZ=m07&?<+OIVM#;fH1qM8`B6gm;xpWM
z`z^Ar=Eq-pluPF~c=zs{Yj=;idVkK3f8Qm(^V+wWpWLSFFY(zQd4bp%e(Tp?;Tzw(
z58Yj!`N+q)d}hSgW)q%1e}>&X*yzMWX~rLY=L*xSZ=kg%>uuA!c!3W;vCH56(jTz>
z_+R5=`kVaKxBd~oed%ScN4x9}5{CVlMP9QtNXQJ#mWuD56dcSecDoUe>~vU^j<;^l
zNQ`2XYG%tyJ_GCM#e%zZ(@69s1-ho6s8&!VF{VA`?k*~Zsf>m@v7&amy{SPaN|^~k
z>7<pi%2Nk+mpC%ZXX#w%uZ?27sQLWs_Z|QZ*B{4K%`d-wpC8!k@%<NuI4@q3tq5-T
zzL(B++XMjc?n%x^_IoXu=8^ZI!wD^aU3mdEWP$Y&p;ol7zqGD3z{%9AjZoqFnjlv+
zhnQ+P37+O43hjVojxeo|cB0YKaG_GjgFZ%dJg?~_I`reB*iu@@tgu)wblt{qlK{BA
ze>g*TBhK_=j^;Hx{g})s-aA~9B{D;Q|9(MYWOVl2zNocCMsr~!5z)t5<i01l4HDc^
zx;Mc;%WWv{ga&+PT+&J8R9b4&X8GEASwnVER-$P?X#4@3&_n_D`Z0+~Fk1Tec_rI<
zWhI+|9@$T*O1TcPR^&nVzNj3%)Q|_y!$Bqlp~4DjD2p^Z{a8x>kjxgfBh_IKK`Dl7
zlai%XOcy2n&9jUyJjTi0dkoG!!sh0PFTHVx&f9mH<Ru5Q1;i=8{`L3hoO_gbw1r>h
z{K0pw;FKmi`xGDh^rKw4bHeqL8ON0)@AP@~yYFL^!xc4ddz<r{8$9#KkMNzF2l#`#
z9N)gl@87)!i|05wI^~^s!vFrrr})0{yL|sy#hWK3&u+*~y|5wZwC0(!T{0~S%m=dy
zXB{88G^7l?fU1ho-Z9NBb?t~W6omk?D$<PS6*DgtS5<p-te9D-O3Ows;dE989SY5S
zUawBTKu7l;d@X{svQ?u|)`jgt8?cCg+c*h;KYs7{1Hs=hn$0BQ_8N!as|RzgPfC8`
z$xV{L97|(OS&K4ovL>4caD7(s5mqO#3AKFN%ljtCL+p6pstHAUjp)<#<w~`h6HQRB
zKRj~K7=+!rZg8?nLTfbUwbQ>5&S@$d#R{iW|L~vuJ>T>uTG3KD7iCRsG`*nAE29s`
z^yoUicZR}6VFgx7mNXI-9gy;EQ96=H`kYY~D8g$ONGU~L+Ldn_-L^(n001BWNkl<Z
zYZoQF;bt@vX-{5*^+F<%a$nerVoD;&YIA`T1NrGKI-_$~ZOF8624d4F6R$=n&iVFr
zHuy?2ifMMLVI{oS{Mm|+tS}iNuY^)>SxJVCLuJ*B%Nh_AUvYA9ixaiQV!5P~rf3u6
zoH!L1MHyztiZn^^PAZ(buJO*1WwP2j=ye#+gbM7H63{a*uvXldvLr<rg;p9{*DT8l
zg>T6^D2=a=DZSybUJ3I8yRlEV-$zvnk0GG}HDl7C%~zJP%+!g*^FgBoX~$!&YJZ2b
z;)qn$vSd})Ze=S6)Yehdo^BLT)wY$|a&<}Ua=N`O0r$d9E&m<ZcU~qt-Zz_YwEeIG
zPILIjS02<7_|%T#r5*Y2aqfBL5Z+l+C(kte(6;CDFx-dE;%tr8DSz+k>chU2!k^uD
zZBMUQo!SD4H1xC}*~RK>O~5MF6UJ&J-DpNOS-|7P;6)`ZKq&9uujJ3d-`D8^%CB&R
z_DgGtNt=l1$&da7_h(foKcJVyyn8fbs0@3fgl-y9R*uSAvPg6F>;IhO;wR~jwmCaY
zc=uq=g{=;0q`7snWV4shNh0n|mK;wjQmuIG+<=D?C;)j4_r?p^#b4H(9VYDbQl^Ea
zs4e%#1?RRhdTGp}s7Q>VaxRqWki~{&Y01k#S9LHgY9buP8^6iL=YE)bdCZw%D%)WA
z*8R6~a5690>ZfQ0r}L6gSK5(oEXyp1(}FlMbd!iqOt6q*)n^61gZfk68@`!8ul`8^
z%Q{pE<pk^(_|zvp$D$Vcb!{DkM7;R2NO5<P2k9o*?j@vQJh?0^9?y0+4n3t*94gd^
z&ck?7qK%ATvqaO6#mlf+UOHZsoE>H?DsenMKgzgzQt`$&f0IYQ`paa`4)|Nwx0%g8
zP8KVAam3I6*oU~V)#25vr+n+y9WLw+_=)Gw3E$p3ipnycm-J%8bWw6VuX+3WQP2xe
zoF1N_bwrj%6nQ~n3^va>dv2G_-5%F(9pkNHKFu+9#&bJ{;n^ppv_?QiyrZ_l#%lmG
z*VO6(j8w>BaQ|=t9JJT}htCiA_@xc`8Sm)>e__}1_Q8@b-JG(tj@-e&K2&_%clrJc
zeRREKv6%8lhY7H%7%mRGbcbEQhYeUsKxHiiFz<!({+rk5WJYnm6S3RToascuaSAWl
zX!*dJWun@Jty|W#1iZ()Q&c*@`A}6hZ3L?IJ^&v^(7vTOH{--cjt;;2U0NxkxI1Jc
z^$fxmT<tuMCVfil*zAegO`}>Wr+viv{c{XAhRkwHZ<I0|rKFLj(;vzrjP>lsJ#?og
zi!~b?X`7hVw*t%DvK#m4r<%b+fWkP@Z1?32T{zp&ETNjqRHADX9ize+&eO?Y5X!#F
zdXlAOm>Ke`DgDtdjD)7!>qpWH1yE*EST@r>8y!LQ_xnxOULmgMx>9QOf;@P`Vi4Ih
zM;40$a>FcpSktYjUF$AP$lRHeYT$Gd;dMgN8xbKOVMj}8cJH*rD$QuK#~_tCdM^>u
zO(*f_)C>0^fKW;)lJNOV3d^!^bVoh1u&bV>Lb95bj=1BP)t>IKOBHF-M4>w#pBkRN
zs2RW4=j8i#xb*&geqq<~|D<2$^wxRCzTlmc7unqEkPhH*evi>^pRC^_(=um|D@)dO
z7(BTuxqmvPjt3|&*Jx2r@l}P>87hj!doPMOUcjDDu<?Ll7UANOYGzR8gibeWJqK+q
zLoE|}TNhTX#att(0RzrEA`~imu=L;G=n_S;0^9^lY$BdJpKx)j%NO1`=DnE!g<m^b
z@Xn;<BRdKAri!<wYs4e1cxJbYiDk+W#o|v{SVxx19ki$%zj<TKozQ~5zpUwM#ZNyv
zU^j_EEJy{@$cD>E6;OKR3asSXx<U~}aXU(0J;DmSY5v=<3`{*3v#J1^<>%%OS4t73
zk*0Q{3s^Ym?-ZV4*zwOR8RKNJCJvO{X<lH0x@|XX5ti0+^Q2&JkdQ>e;l40Rg3^$f
zh8h*jMiJ43whVO5o0YwSBG_uz)^k$Ge@EdBDKD-(jb4COFv+FQ9Cjjh`yxFO0g?9X
zbt3LgN+!7obVf!o$wSZvodLXWi9f@jVSjI2%94<lz5dD)z$tM^?!}s_4*#7fmZj_{
zXw_IeN#bf+p(Y`8!hctFEkLF9L4-A^4k7P<me=g|V-}_S-jii5>$$OJQCK>WVpfRw
zN>Mw;OVLA!m5{4EP(+%_O6fjcR`gSmN3+&3U0UXmW~&#GMVh*H9G8~Vc>d3C@ALFN
z{JEPq`0LMVzHs3?ym^1l?|t({-oEn$Cx`dRPp4eJ6Z0=$la=qCLBeKgc=z6n$a}8b
znQ(eKqnhVXRYXxtzE}{Mh{beH*6RwLJD2M1+i%@N;hDt|`E*I53)E2BJCtg<`x@b|
z`><hHTG!$O6-{|9m_Q?3(>!De=Rs8;;c39I6Z3N)+T&|i$GmcH#=<&|%bMT5{h+KE
zzz^;Z7^W-dW2N9rS0~(C*8KDnBdmvCdvE-J7->f<e&*33JKacj6T%tp){wpC{Al<C
zO-bLB?#;BJ3E;9hd{_embt77;+Qj1g>fmj#f~K;JZk-fieMtIr;l4*<;XuY(BD@h;
zf{|jDJCef6K7A~9;8pk~mGcZE;ruVH03XX*f<Fmqt#Tf3Jynp*1{h|B$+BjnC-T~5
zC|$BxGcT+NPHIJE<pe%6h`2keiM8TPUo;2uS^%cndK|u`OXju5=Tc#4M{8f6XzA6%
zz6v1Zq@oft(;@&>c_~2CIF~jEBnf3f<);x=bp(=7<;jfnzDq9~hRwR>yz-0}HCady
zvQ}IX8{L{`0BrSQt{g5gM$?NG=QdIfCncM`i1D)Kz<M6p%+N|mO}EDdk&!j-rnEem
zltSQ24YyB43S5+Z6@$#M-HUnma7iyU?2iOfP#%2zEWdue;uAkuv-PXj_{hs=_|WBj
z{?=P>QMnO%KEW^N93OhVm&Kgj>`^+1HdF4LEUBxSdq)#&UgFCVRn^o*g{>_nO(BX&
zvV^*-u&%~d3hNa{Ys$O?G`6a63QCVG@MsyUHbknthE&PuS0u3H&aQwY{QK@GV>698
zDMTmW&4W3Q?e<zMDmI$ZO8G8{0Y3KVh^O{?ymEcaw;u8y0Qk{!L!Lj=6<i0oC;#yJ
zlvjf*@xQn-#={4K|1Vw~u-%P9p!(GUky^)+d_&M{x&3WX83I-XuuLIi^#+4zMtw@r
zGJ5OK4m2jcI{ZH8R&s25(1{o?D(0o-cu}|PwmK|7OmaE>CWV}KwezGw-dQU+nOEdi
zxC8<kYC%jxsu?c@AvSvDt`cc^@q^58IIlUF$?vZMF{0U;T9&fWHS8F+yAfHU8D{cx
z<HE7kjYuLb4nZNntdsJtVX2k1<B`pT{b3v~x~wmcQqA3QiTBV?Bp93Ejw9JmH%6u&
z%Sy;Zjc)r!C!&`GPzTyRJ5rL|2AOX64w_!eXpt5#Ysah<cO9)fJH3d08gX`%urr7`
zn3h~SF1UG;b73=OZxAyt1t@#zLXS&Z8D%Zw#eOCVI!zCFvasS_WF7Zsa?^`qvEjw@
zJzl!dr<WM69_QR1m$KIogkbxrA&ay)GE-8`F7T^2?sE4hF0%QBH|hS-A?4^28#`O5
ztzC+`q{i^v<7fHo%lo|i=!ngJ!Vf&N$0wdU%j1uoWn*uPq~C=!5!@q5KpBW4#;0>;
zr!&xs&9mDKc18?$hIB?foOLK8;BzlF*nsYf(y(G)hw)lSnrSR++PPavgDWNX<7Xb<
zZUgbP!x`VbGm~*zxEGDmK`;cgk=>0(#OC40XD^QU$c4VR>PA8x_rMcd8J+Naw~P9L
z;Lo(;XD^R9H%MAW?8=BowSP1qB5EDYm%q2kK&>doDuC+J>#tmpC5^}<G0pOmB+W)I
zQ(`yp0FLGrk8XDWT0;!U7EZRbjAW%wtcb%!9d(*AU-9^E#<G$NztxKaz~(t#)NFMl
z_D8V<Lx^gJ=md5mdBNvK2`*&tS#DVrj?y|t-3af+LaM=dVlDoBjdnpJ<u&LEBT)1N
zMD8or3Y=6gc`apjq8lY(xlFr5IrS<lUyF>Qs$_=Ui$$R(36$16&``&@ZL@1*jaFy@
zq?VOr*1C|bv8FU_rVP%*M#rFyLKQVfUWvZ|(kdKHOSbzlz0@!*q<0*qhGlAa_i#a8
zWeif`JfxB0XjZeVoNP@h#UPPPztM}Cl?AaD$+F{6Vbp;RduEi9MvCjFB^$j2t>EsY
zlChVAM|Va{l;+0C7B@CB?!EK^qsblq@`YFU)#=+T?o5b$#PaYiU;O=J-q?MN{-8%s
zTjqI<s!PWAC*<Q(Ow>W+h~pGxA?<gl7YjCaMl8k)%0<rYcW<?&eWbM{iE#f}R{Qdi
z(ReLWp2V!m+RDoJc48AG$C}O<o;p9``|i*A<~@-d|H73c%G&Y#xq&E3Avlc<KqFf5
zg*T2qU_SEMiz8k>*B6j(Ey}EiGyRyqbZLWMe(&UgpUJf1bC*Z#_2PEo;R8n2i~?0K
zBwbn`Di=991EqzA1JHnsO$*a>4h^kYDH#zuFS{!)RO2$ul5uJSGDe~LFesu+TdADt
zO*s^9TouV_nS{j!Xp~4L&^nov{#q$cDiOoj>Br){>P0d-(NH_Z+XqW(>zEWWdru-c
z#drIWEC?A*r92ll6KX3bWtM1qk*EPrmzK?LkfjRDJ!51WPj*o}l@|k#W$D}PC=o%_
zRL<g^<Mym(Zy?G-jr_T@o_QskZB^~)$GYWIceG#?+g)joj8+UYVFMa1c41AAb$42`
zD22{kTER@xSacO<l?w?#%FRI{^X2ifrntZ4rALO`xSxv?axWH97A@KV)_I;e)8X2Q
z;qIg)+Dv(7zej2mZ`@t*<tryVxzpu^3w?a(X@{BN`TZ`D5uD=v`#DFmoO7du^BXCn
zpu%)>Tyka*Gb=2wT%GdC$44md`HP=;f{*C){Hw1%!{7eBuaIqA;)VUM@IQ@T<+D$p
z<Bvb3_`)CN+_?Q+a0TD}_64H+qHx{4!zhFCH3r9E??udL8)q$BXLq=KX28AsGhV#3
z&CxjL-S_TM&KB6^5~U-YHe|gnX%t~OksXk5Od2DGE1Ml7%tNJET1#Pt(%p%*e4eGH
z6GpzydH%x7=TWcS<{Nh><ks<9?;de$ob$r|fJ@sQaI)$7*1b7jyE*30L*^qtx@Y*%
zxqf@HHiDf}=xWN^@&~u4Jp9;-jV!Q*gTMq*kQ~Im)u{Z^^*ML*nm>QB&oHb|uOvcf
z3x%c{AhatDKq%7?NdiqP+b%?T*-!D#(}~QAW8kI2j~cx{=Y89M%F9y^{JH*WmX%|m
z!(xS(k~a-lL2ZSo5gP)X(s8n=NFvQJ6FcuLxaCAzwK67=f;@nM@`7oTQB&5fm{vd-
z*GU;35>;44YDOP%C~-TnZa1#(j7wn^hUbv)liy<k!QXoUcZ`xTUngiCbRsFAmO)5l
zm}u%ScFMw<@^~qJcnS$#v6+UVmfkeds}rsEp1oeoAQ3Z^Jn#jw*pLTY!6Nv|$)e)o
zX4=xR#|zoc+wLWScI?>hC*+lDXXHjfoM<+>F}F?%I*DmXS9vXbgWCGQT3nTm{aEw;
z7l(Z3+7@r#e1b<V?DO`ucRBp@Www6p3V-H@p5X4I5%;URJhtEG?c*`3%rNv&bRp`X
zl|d&1e9|HAX1Kg$W7MVJ&sda}ZYN@7V|p8Xk}PF*<WNR%>U7XSl{FxXLfxQm9czvE
zk(_!Ou&yPy^s@+W9VPPF)j*<DJBIxWFP)>C82;e8oYJo!&3OGtn0=28Gv1r56oS_u
zKXR_iv!hA`O&i|~)#(3uD(Cq8TgTkaKL{4yE^7YYcTV})CpSdLvZ>`+O`CrA#*A-G
zWb*K@Zp`?}O9Oeo4Rqheqr~2dQW*YdvWo&AScOf9Gz&k@t~nru>cWN!BPzo}O!KQ6
z?>Fy#yA9Tq8_i3Y<`vz-G773a-pkmdQR=CzV`-&iJD67V(ul`)GXW8U?%SlWENV~U
zAq}WUqQPl-VR^F{Vx@#G>eb=|G&53p)tzYdw%0SK_XJ6@a3jT<?XCz|Omi24QG(Yy
z0q8o>ie;q(-I8$nmoAKLlnOwUFax`ZB%PgJL@zTWS`>k5Cz5BUOA$@+#<aR@wS&!$
zv@X|9O72by&W<u>h0NtQI}y{uax|;h>BlG~tw>ROHhU3Ti$~*!&-a-YmK!Gp%EP>J
zcoI^xa)Hfj$5UszjF&Z|PQ;>=iOA8UWFs>a7M?oW3q&c!SFWC5#z%bX-J38UbMrXE
zot*Gn^IQB^|LKQW{-ba6sh76d|IB0j>hI=M(GN4ayn!zjYJP%=6MVkFJIiu<3P6;m
zy#3a7CbK=3vn8=nj3;yM-#Z~mVzG5AmuRIqJ)DsGDSCs}M%*ZA-&9yX&MTp>``VKx
zhSEW-<>1k&B2toky@k??!RIHR+U4o9L%x3XxV0h!{*A%^*-M)|y{T|{B_(tNI3mft
ztbLFnf3WBOq5VF;e`nUN5&y<}V}9z%jTOKb;_aJv7rZ)N?I)bpj%g`NQgSmtlmr4$
zi01)}q8%iJB;eM5rrEMoN|dNpo11|LAQDZgwb`0go^C{YqgpQ)GBrWv$2t~;<L<O#
zm}caG$(2T$B7lJo>5&U7lY-4|OrlmwB}OX-Swv+;Omtd<QBqO~1y^p`W&+U949ilS
zg9oXR>cX!;#oDbl0Uvm1ILWO{F_dDMYT{V8O2eL@KNl*G#0a9EM5{%PdFAORA|*B{
z#rRY!5w;jFEz3&E{zh?Fa0<Z!@={1kNvu~Nq!)*p+B(??^YF%<Ip;?Sn_a^=x17#O
zs#+9|YU_CWV8LELp_>{ykz%r}DQm}O$MEP*#`{M(8y&-}w9HCJUgz{v!!UCJ*pv6u
zO$~WviHv4iIJO4~Z`_^}$C~H&yWAR=6wdSJ^+Ue(jwU}A3G(^1_vn-rfAWnU*RQ|B
zPyWyjFaGg;hLz|4&NrxHhv{wOI+nDTktH$H+jlV9;B3iqa-YaV7!<dQ3ZoU@I-U>#
zKA+=y1ENlbpB8wf$j4)xAA{=2tsRhWr(}0t<i{JiHjP211LNG1Xh`*n^RA5oD>M4k
z$YkKhUO3AS*L&PPS#mI5a%Wob*2#lC|7R|Y`0&LM%f+-k{lqj>c*cv`2LqA^(f-dp
zvBlYb%w9L<pS>&9#;48?0)b5KNfl;(&+TNqJt?_e2uI*29~p3_7e5g6Edt2uo|oys
zT1}wV8MdvkG)6T=t`Amzp5Q)5^Z%oRX~pHOM68uUIapeub~bFpB*-`!El;%Bi#flY
zN_lG<&pZi=^`!%lV4midQ|B3VBUJ5KI2ns2W+hzEWVBf&!O;<vhkj<rD@P~QZ0MCq
zNofV-NJGMTaN--TnS)NY3ch*iNW8cX$9i?KnzrHruq9mBt*#iAI2D$Ei~v)m6^}b*
zMEkzUv~wpQV=oc%&7!mnQUh9Xvak$Nk;6`6&7`nomE+@&4Y_rkb26{#riM#99lD8T
zmRpoo%nFOQp8M01-C;r+NxM;dc;{eAFA*}+qO>6?bb<x7OhIOaZRyj^P0|g9tB>q-
z*y<W4xoa1kUVLPWXGWKK_4<hKy?>KlWUzNj#>-cz3dh@hMgL>F?0(~r|KWMdZ`d#K
z`s=UaX4|CJlMc_r;*_$=N#jjS(#51*I$PWH`W;+ZaAtb|9`0Pdi%w#SlM@sUYERPb
z5c@*NE3w9SkFzqRsh!6t@hkMsgZ9A#!iz(d^UMlMW<&_a+U9*}mX*+{ol<lY!xLxv
zJigcCZ~wvj58OL1@AUZi<t>zIf-2k;fkiqi9l!FB0ZB(Ie*UR#&h%rv_naLh{QMJ}
zcn@a>u`KR{Kz88>&mzrFKR)7LU7hfWGhMbj(%v-KKzk#rRW4Zlxhm(IB+%T%%|xsb
zBJm*^no!ZI0HAB5h@Lr<2|d=Wo+J(<tY-GO{!j`=ojCMCa+l?mXIfZx`_XFV7=Vn#
zD8Bs8382{SM(hj|rg=>-F+6*=D;r6H9OD8h#w1ZI;GmRZT3Gr56|}un8m$O74+np9
zc|Z(2td}=_^?uIoFy;IxM%iXlON>Gy9Rw(~xD1s}oQ;%+(}kszXbEKGUTDlk`&!x@
zt)=zJE61X=Y;<FG`w=$y4H_*%r#jqxc`dTm=R2}FxY3Q7=9b+=vk_W@?VxwCnHdge
z6>+ZU#+u!J%rG;&f3)QHublG2evfDO`}B6R0O-mokw!w)N+Qj*<AUwXu+xjB|My5W
z6-hhujcXHr<e5zt8h-rwea>Az&u@RP!*rhWGv9xSfAV{ZTkpqg_bp%kYR2d9pXIN9
z8b0~EH~F9cgD3dvFM59EtAE1HYbEw{kJNTi>MVM5m%-K^qs;;Pn;jnC?Q#45oKHWw
z&EYuboo>RNg9$EQkoLN0rP<pZQ7?KpbAUt9i4<1Jp`>buK|Z&_%k?(l`bJ(>ixcBQ
zbtt+5&N__NSfphera_m`d0xAI%1OC8#b4U(^XH!5!#m%~f+=KNXq97G*8Iww_dgi?
zPw#MM(45{16rAZNG*K*OBc4CGHRrkAER^}uGXM1BBg(*;Y30>ucmk~Y^M~TF>G>bX
z6QONTYE{M8p(4m4g$u!tl4+W*Do5T>@g54Ef77Em^*vHPN5Ni8BjdD{(}`&1ElSIq
zcgB4AokPBNXU;sg%uAWwN0D&&i%NE$%UbY(Bm}ziK|4NVG3TY6Q%D8yZ+kQsMr*}Z
zUl8`DlxiT8Sg#VD4+=8fFa`0U3TaG2L0W?>QmqYigZvvMj=<afgh7xpkAeo?q>y#s
zPPAGIX)K|&5^cJ(qlDBbmX)P)LLe%FFx}CtV$_Z42WoIP*6a)t&J0qLNVC~BbYn@3
zQU=R=5&S|x!iEo?MOuVA6CKD^p6!0pTE96bqu8xZ#Ky48<GUSRymS%0c#ij8Qyd(9
zlC=M6{^=Wz>=$3=_7~schu?<3_Suhf@!1Y`b3#p@I_(n=w|M08KEq*`Gh2P0+#fLP
zXLzMZ62r!@!_HP8l_X^S9-F%x^!gcF8(k8k>Bh3-Z;`S;jT&otnSu9Gvs%nDT6X!3
zZinR#!4W(`dOkH2VN+0R#dGI}oD1e78%e|uKeZDoQbmKs3sFcgpMU<{qX(u0nO6M5
zGdm(l7MklY&R><5At8M6>V)4rnDZ-F#^j;&^dV>)JiHBJMGF)*=zsJ4ZJf6M)=UT1
zzQ77RUsnrSdpRNw5*f|<*Z!Rz>(8Y{7_^zyKm5>I@zj}K*r3x~yFX*l5q)^A*3ka&
zBC=T1iFI2p86}ykRTU0@dX#wd{d+u|7n0APy1WAqfG6wI52a{Ne;Znmb=zeJoE=1M
zpVd!I5Fu6cVj&3i6Zw0C@Az<SI>E0l4zg@Xv|<EyVjVURJ-dS>C?*LI+1v;F8!5Yk
zR5S)%LoHn;FFZ_g$7VmKpFOA$BmWajX0sc$BT%KJ)!7~<eDaB1I*DO_vxD_JB=d80
zJ6-hN4&#Go_@(J>p8hXB!<B#i8l&I4#%G?}<?LQeUY?Q@<BZ|avqScG`kdLw$Wp_(
ztqxu(`dw)~Yi)4akR(P<{4`>#-x0d3)|kkMkE77nM6&|qbVp%=nTbb%3UVPz*05bj
z9NLJ0Ng6mFQE-RprZIo{qZc^WP58+9k*E^~%}KOsY3g}p`KRB$_rMBq$7p`>*<H>I
z(>CbZAblta{LcFmz8Sm-Zx=Pc`QEs_x0;`80<}pB4<v!+Lwo44Y6l+r9;E`8V5J`1
ze9@?twVH1J;os+x*P-VP0b{h3>DD(cIF0JrinKH8B<v14;9ScKDA!JKt(yG+QT|c*
z*8XW(xzMk#iz5qvzwusM2S)z1hjUUr+&C%1s4V=P7t8^qXL|V25GW!;`<k0?%K~ND
zd*70%OlY5KPiwCxBAT?50eQc&8*pn}Ft6kZr=??3*awz9JVIBV7M9vIHs|uQi_&pt
zTrexdhw-!!m!qO~6xOw1Dqg@-=Nguo|G@h@m{rUwCyPpjWmbs8Y}W+DyK<DahA1Yl
z4A_#gHvGm{Z}R#2I-~#iquhV}IzMx=!!LaD3HDyT&Bo|$?%#czSKhqEo%>V1arJ~(
zu1<LCV8J4|`gjz1QBf`zOr~?D(*+K=bvnlqR=q>8md;Zvg;Qca;=`ZIT2`C22B!#?
z<nlEgR-T*rd1YnXZ2d|%u$#sF^vmaXbhl3xD8*i>7B?vz{L*V54E~>edWSQkbk%>W
zwHee}68N3>#}7!4ceLWu=LhYb++a~n(1w?_dX9(V^G}V}hyKk6`LFQv&Ar{=0nPs;
z0q9l9@bB_aUij`dv;h4?Ge~7)ZdOUzT3C4jRS*~1=w~cy$D=y~c6u?f5tkXYb|Oxf
zvczJPVp{re0VU|ICy08w)&z7Nz$3H=AGlcI5ovJ%>L*PgtTJ{(&HU36q0)htTnTuR
z8hMXR`!QKsnU45SVVIya97;wX#&BT-OC3JpGAIP?^kU|DO;Jg}rNMM$QUtTp@SeJf
zAoq<iM$_N*QbR8>O!8X5(^v=RBSN1i!WLDSd=&M{8fu)|VBWFffEcYPY8hL;f4Jbc
zzIVXsqUPr50z*w@M^wuRWu6m95%c3N|L`kU`3LqrF8?>5;m#kv!N;%7`M>^?pXAvW
z?ocn@<krmt-ne?e&7&!A-I;KIl2a=IL9g7Ka_!nZX7`VnO&3(wa(KGrbXv&gse{ay
z)<JQY);zRyQVArnj0%-nRWKO+P>C#+K@{2n-T~^G001BWNkl<Zt>rp5LAKFP>19&I
zHSI%FrT*Hx^#dLRnbCai*<JQGvhW(!1NU8XnxhoI{od&Zg8%29*yKz<UZd{H;b+}S
zw(fyM^1u!G;J-JA(F2u0s9u`qgBAU_8UN`pU3j1vZY;q5ZI3h#F0LWVEb6fR)>Z^T
zLB*}O1dnXWPW~b|0Tp$ljVC{&wFHs$+MZ^S%OmPBW!R)k&C1@$c>(Jh0|cX2uD}mo
zGXZo;D`TT(9O>bK3SwD*@Mpb~sX=O17jzgl6(@xwON5v+FC5dnmN~B%^3RIx1t&|}
za<S$8%7;)?VxOHR(#zsyqSBD6#H$hJxy9Cw!Uoo6P{z?-thRUhG2LLt)`?|-s2j=X
zv#MpZDJNyn-VgBctddH>I>+h428~36MTJR>Wl;(Vs&a-h?xWK_l~?G_Il8lbCIF{E
zSie_BeC^9A|J$qIrT-US=I~G6!+zm)9^D?|wht&(#k{O>TC>qhSQHiAM0_Hriwfr~
z&PW?jIR#o1rLpwLTK?{7dVHw{wAN(ts?=|eJFkRAsg?Yk?4W4Sa;nS0B>a^$5fRvM
z=p%ajLcY>kDvm$CHsfxwii^x>e&P8&&TnR|H5jatK&#XX{LZ_lyn6h>;NJ<9d60>c
z3rLl3`?9rU(%ix8mB3mOXeM>=P!gaCbix5DpSNYDKj?f|5pFi!GqWlc8Y$&<b|A>c
z+u^jNOHCFVI>6yH*n)@N)j6n?OaPqs!Fx~BPX|)QR>HsT1nzmPMM%MhIc4Jz+Ycr7
zy-C6O&5T)C+3m-=?ZrXZY|*L;dti*({A0t#PffGIM=&8>UDHVOfrmoo&fr@<y$=`B
z1a3rG3s(Sy;TW67@=U6V#0WpY){Z2LTBReDj1W(jGW$;x!ypx(!fq0=tR44GOE$U@
z<7L&J+^LZp`CweIs2vBBlAXRV0;kKG^BZaV-j<;qn3TfRsA5r4s;sO&JMTr;Au$*w
z*K#^5vBx>T`{p5U-aO)Whh1L1a+|cC;^RKPUZC$El25MDLlZ6c7*7;m=6n44mmcAx
z|KTUO^7ntApZclqqyE$#KL1<4N80%m?~RWr<`>9MC%m#Z;^^p<>sPMgCU?<x!SX~=
z<QwFRg1(tz20>jHPl*x_1*3`acomR+@9-9U;Ktfol7q93+DnxXMc~nRB?;SE$D)=)
zCpMx~430F`!N@}vXHFLt-#S{6ggQp^)6bpZ>?jS+QUd%&Dv;E)WE1N>U%GO_H`n$F
zI$H7bPj0p1vkkZ?x0bxe^$v7>w!e-AG%7-EN8gl|Ymeq@4G1XltXD}V5MPW@M2)Up
z%cfKR|BZn5LUQROhND@<nL*5*aSlqxQ#I0uDMf3+V}j#PH`c)<R5t(KJzVngg<e<`
zk_&jctb=wsL;>_Vo>dIeh}~h#^^=Orn~BJ4)n2>{G?ZcWQH7qic{soF{oCpqrn$v=
zIVD9#OipaA)I(HgAYABW<=V-|yp(djpJ?(*Zh{i|9sRJxlSP_-DnRDZycU5}C9AEc
z!EGlmEwK@W;dc)g5|Eh>K7*2^#*3OL09V#4=0%NGaB(xiE7%((+&nFajW{_eC9gFP
zm4;S=c6Sm(wWygFmNd8>I}h(3Ex5d$wf+Sr61?Tkq~K4c$J{!eQ4~3MPnXQbQ<O5;
zYzL(TtX$}rxxRqwZh`KSPPX{FU;GZA-}L-<|K-Oy{Rgk|V;8&p&c5RA_r6Ee>2s(e
z;%=AMuOFl8n!3vI#WB{-n7f>ghbZZ?T+G00oC2>jd1Y~t!qpa&$Ox^pGSgqymNXGs
zxG{=lDaIzkYL!Kwd1<MFs<c;%UMAX=rIRsY<>mA)17;=si3jPu&un)&znQH?si#@#
zZB8Af;EPvIc=e&+|GB5O*&D=R+}>26&BVY{g?r63qktwue1HxdS_XWe$FGuwYJWtv
z%RtSPM=Bi^Mug$>R*I;$JkW<%*3$_kZu6uonyr)&r}K)eXBl-OVxxF$x6A3G;>JnN
z?jT{*jab&Q{4&ce`-6m0$F%a=%gS*utpvnTN`x+q;$&VCY0(z!!~*Up;Dw6=7KLTJ
zw456y6pq%3_tDKnN*X6CvMA3WZu%X+D)j;5VGvw}$GN2w3;8LIb&IUag%#dF9m=FK
z7^r$Lv-M_kahMqzJHk#Z%SsnVal3Kof$_3N>zGcW=_Um66u~lLt!1q@HCXGJ<~6rY
z3NCD>oGe7-W4jk|FDM%K62nF}W>&h^bmPVIeUeC+mE)x-FfH<$+-Q!cCFeI&Cb<=(
zmr7*1jfTx`#70+6qjwLN5ZaMWYIyDK18$oxC*vt`S@8bBA*QbAp8p8>V1zrk%d%MF
zRmy1pGX4FtYz=x`duNB&zY`O8XUvvg=dXNpi@h5KKew;=YVQ`e4*QHxDl&VXakoP~
zU$9&p6WbD_6HtBXvZBayqIyZ~x+ICH?(FnqIUM4^#V{IM%V#$1L?YT!3k$H5MiixO
zy&t`UG&XcYndqGuk~A8GvO%h_R^yOyX)iUL-|9Y)0Ir_qBK#5si!>+8R!RvC8SnV5
zH}5|X{Fzq#%v0NJ_hYJ{JJF!?KHPKH$0g3e`9U1E`*<)3uwt$@z+w~R&U+#Ru#D8o
zcc76Gue}(gTIX937-$`>qAVzpHs*)TXw4{*{T8u|@B+Lr?z*W#@&D2FX2F(S=Y8LA
z4SP8Ay|?eZJ)yhN7#aXU00c>q)IdujRSYdDR;1WcJXNZaR3+uAyyYoF<t3Fkc}S&F
z<w_;0$WI=WGG#}WOi4D)P@*`ABqo9Y(P#`k-uc`!?`f@-hi~n3yD9p-0DZgf8TQ$0
zt?&DP|3TrMr))$^E?R!aR##}TRqZi}T8DQe`YhEovx@86xd4QfC($w|PhuFyAQiP*
zFV*XK?QW`Le7A&@kZ887Vo?XWNi-`lLX@c*A3;M%IAa@jQ{nvk5TX`70*<x#xM^S>
z>CCO;kWw<|ubpgK+kMMvB@DgOINP_elx!k-B+-gR>F8$Sf;=xJ&%UKf8wK8o?bAjt
z0pN+94hPeUN#W2+aWt>U5_!%%(OlllqhEp8q|Zvnle=9h^#goxT#&_dr9>+>Q^mDy
z(O=}LAv18UmvA_%8D(aDYB(R*><Y1HQOAskEaCA{k2{A)5MZ$^$(3f+8?rMTa^!)7
zdjy-Z>W}!;wTnD?VZ`~34qyGjfF{6eZwz@gyUS-EU-EZfyUQ<q^df&>YyR{;Y#2vb
z6u4rEG8ty;X~Bc)8Xr7fYn(C?QLYu0Q_LDqH<x7Mh4W+zlEh$x7df!exW?g77!@~l
zFY47*EtK=1gjJ}u+!tCI%CeFBE^Y#ek||fB6<eKz&s^y7tvzuK9ajxsdVP;y`{)IH
z9Pe7A)@}D6|JdU18qHsS<{ak-X*8XU2daV}>@Ru!X!!`07~oQ7d3u!c!fuzWB?Sd}
zjkdIxRzqaHgf?+XMTVjLn<j30W$kFz?7*-FMO*TmwN`fi5aP&AK~)O~r-Hcgrcsht
z)Dn+<x@t&zB3dZo?`Br9C<G*PwI{W*U|mF}L3X_~id4!{POw$wP^noPlkN^vn#MC(
zHTWQcm|h|bx-@Qar-fr9H-uQSX(d^0%CGmybsNz5HnUCa%~WVbGOXWAh%?eUf^Q~P
zA$2@K=9Mtcw1U%R!=PgyarzY}T3I}+I#7f&7T>J$;sKx)TKRRKCrmSG8{T-Zq#F%6
zRA`tMqA^H~;&|SOX-UV@O$FS`6HUHsNR7PSvXT*@i8_WsYB((#Mwwx}YABn)MsDbj
zjNldlj;EEN;Yz?htrUY?Gp+>)9G4Bo$V|MmT*@eoz;so{$%(<(A<taiArP3XDteuq
zUO&e*4&VPMi~WorWgoEpg=e|*&G&d_7v8vVL_O8a774m(NKApZL+A|&%LODU4FS^b
zy02f>o>6XDm5q>@5Jp>@VP^qF%q3JJ0WTYeLi{6D6IeD<L0PS0RCv5ojAq$*28oWF
zgo-xfVln@TYXd$wU9lIXst=YG^U9H@hQ!Fk#3;odzkS5_kIw9cJ4W-@pWSA6kcx|u
zkwtpzXvJ4=PdKR^Kk0G5Y`A~W@a>}&U${Kr##Yusglin2?d=1^O`xq_)w*>hgf)2z
z-$secA#YG^EI4{udRCR(n0Z9lwTg(iiiE@vQ(Rle90f(?Sk~eqbo->Bae>1{&8&17
zEj{VG`%6ANUQs#W1+*S!X`D|UFKSwi|K7ABjkMVZlaj+p#r<(X64_vrWkYUc<WsxA
z$;t^0R7*KuIw`xSr6kYzc+sbYTlei&D|wBGtlyv1987Dp5`?@PWw~Xv^!7nUWJbd%
z*Q^@Pr1X&p1Vp?ceF!_SQHLHvh|>{~^ZIx)x4Ux_1VEY=4HpJ73S3qWZA3PGaU|}@
zKFHK1wVIPf&1n&%wt}=5T`Lspi^G&W(u7xy)Os@J6H9I#Q<>V*6NDW&S=JCjlo&TG
zD$mi;DSz>UTa@?j6P9x{9<Kt=UfkwaUV5B<C#CV8WOtWq+XH^)$!$J&V~bHI<FR3%
zk3P18P86ooW3t%dzxr;$?`&M=rzaEsn@=3^^mE^3cKWYz^LJ?YHjDl5p!^h%$2BEQ
zSx~#0pcUF`M!keI(RiilWQI;=1#q@9_H#=9E=zTs6bNqO5!@vxMV49-yeZ)an26dZ
zMw2Gu@t}>0sXUR8qZD~+_|>OVt`EgD=jX3(a%HQtUVs6`AH8+R_YcoPLvo|}jc2ww
z*Ov^8padBI=Itq8x^?pJT>K9|3Lp6WJ5$~`SgjYl@D|jWm@TkwQ)?Gx&|dyq2jQ(4
ztx2p|^9kD0U#37xrT`uRi<}ZynKpy3400k?4lxc@GOD{Y%ptepLX)er1rSm8RSeSD
z8D^Z!<)lFiSG&a&GOOq&hT2O%{QTt}b+o$bWtQ5?hfK6+yN6xb=Nli`?psb44L%5H
zxYe=IvRhi1EYgw%5LA&qizu-)R258D5{P1z!XV{eJDN<5v=*aGGmox6S)$qOCM=@M
zt1uU%Rdt-P%KH^Cv84{DMy*>37t6H7$UM5;Wb$Zhw$V#DU6q_H#RM}k@`Y`7Er}IB
z{>iFgdyqtb0@=(KrCe_i+wZCouf@6vSgYCSS~`hlymHiyXS17-Sk2L_rk@+ei-zsK
zWz@CY@t|yuHx}3F;Iw5Smb0AS>~r;8kEbr~@b=Ld(u~WSeXefhj5?O*FZH=U?lH&`
z?(XiOQOw7IH*a^ylZK!At2Li`d(MA+Wx$s{>v-*(`=C6gn=mg|q^gIq2BmV)H3mqG
zAvKyhTC|C}&#m?OMk`V)^7Bq&Sk(>{Pd6Jk^G>3psVD^HV@s;h-mQ~#tWku>U0JKb
zA}F3G@V9>YD&M<x%uClcqw9`TB47E@A#Xf%^0yp;^TUi)<%G8ZeC5`d*M7`nFxQIb
zMmdjfWNVYogGJ37C#&@)@y8G51OnHSmc(;5B~<E+2OzCdh^{{|-YmpPLi{W)g!L4x
z9VuGLyH{(FYhL)MwU-m3pITA_vsw~dyq8v>-&M>D;jCL=qi4zD?7H@Wc_n~b@QEyz
z$YL|GPnnl8cF7Us8f1JjF9hgox$v_>d;@xkrkg}xfFRC1X&hT>#k!3tVU3)&8Exa0
zqn8<~cyd}LAXqE?ZM+DP5-kf~BO}5^%mrN50i$I@+KvIHQdP)CR#rZ;1yyuOk`dFg
zaM(C%(@L$0H&%md1i;zqSQbs7hitYNMp*>vL`QLcm@+LKm2;G}=VEkP9_BK(s~mJv
z&An5ppop(;ltQP%oex12a)aEkD1~QG)t=o!LfMFlu6A<SHhY#lfhW$N<Gp<c+K{Cw
zemP_B=#*RCl+|)YlNhY?+&i9fyexS6>L%~ppOR`AFIFs8ONu0?H@Lt;_Xs-hJKubb
z{Aag#;pT*2P81*BIAC>{QXM!Fl~OxT&?zc}sJoBsP!A1ls{+<SrUH$U>5GYfCw;b5
zls+QtsTI%M(gzYn^lB7OUTqA@xpf<7tz=HtLE_tlQy>I4h2rJw=LBy6l-B&4H}@V{
z{2inD&F9avGf1gipqmPI^l#od{!5Ghxlzu~UL7*b#O_^5)y~z8j2F&#`1<V`uSTQM
zKY6g=Z(Pt^ij|2954X0_0vws?xBQOQAxHqnR*Kx3_(ZOd2CKD-G||BEqM>lI*o>CV
zpo}V|WhE@ML_;?fam>Y0!gMA5kWp}HBV*K!Zn{cjtfm#fpw^by0Oz{$v%8au+$ffn
zEaa<OGGlFHrnA&=VPL76c3}#MV}Dlj#8$>m-%`53sAGg<(1?krd|o**dWo2Tw7UNm
zEv=jcg$z^0w2F~U(f+&kK}L0nW{}E>Vzex*B323tDee2YrfH;q)JitG5TKu%XzwMz
zzuhyGwP#k!4H$qlHPpdzd6)?pRXR30V*k}i4Mr&zh2zRr#)V-*>EwILt$?pKZh{*f
zyS~;ZOUHOwlcfgdCEiWN6J(UN0h-Xy3`JeD(Xpgfaau@)FkV)yvWj7^%k|w&Ub}UV
zjlqEb;4@G2#7>9zFOK+M563)nVT1qd7oS|KDz9#4G#<8kmT%m;z*pZn<mTN2I^zt}
zG}QaM{NCFO=DV5~zkS4i`sq!+-W&0i<7av1+2@%Y9<#l(&D~pfVxMl%*5DdXH&d)i
z@%+!CCA*HKC+|EyC>BLcCo!Z5y}Lxler+JBiQWMLEw<u1T8*zY`W|f*c@%_epi%<b
zHL>bTVl3SszP87!KUVht=0`7Zag;$2FN{FohkNtC<S+P{-5#I4zOha`YMpvoh(0r#
z&s`r;-U_@G-GcsNtoilb_T(4fxo%yx+6li_*q|c*5yMjBVW^SMAR;Q_0SWf@xRi08
zRt!3ZbCDvcRn)X=@#-g&@B;_z&q_ucz4hsnnaF)#^V}6D^ID{}ajx+&si^jHajg+o
zoH_)G!qFh~*)np{GXrU2sN!6^bZ3sJ7lx^@01=bLR-7re<|@)?Lp0q8(8%jrqt`(v
zotRiWs{|!3QS=g7P~)YG3*;=+vy5H^(tEY`UT`#P*ob3JF-MS6s|kUkIYWPa2&^j4
zvWc!QDvqgwY#PhPaWJdE3+c!;p0W`g!11DHkXu3s^inOZzKLNsH_S^}@F>N_p%8)w
zorLMKBJZV43h^Xpxd@ABI65s|jB#^-5zFz!(6cZx^6!tvQ?jyPR#$xGt$pr|2fTfM
zjGfJyE{1&Ntz$O32|K+c-WwIu!tu2??r^#)DCSfA;*`oevUJ3U@2vQTvpxRK7ccPZ
z->J#26ufrjJKX-z66!4$(I3!d9&LmNvUC#r?Uf?4k}~W<ph2YI#z@9vzmw234rQd|
z*GhB)+9*`SK*Ss-2)oe9W-_lleM=+|ISMcKZ6+f8K1A2uwz3S0ok0fCXf!vP-+J*P
zmqyu|ywgcECyR<dyE}Pg@&A>J13vTkhTt;lA@9E+?S^Vce<AR5m;37t;DZwO7Y!Hs
zNqi=?uJ)WiYX#bX(X*pgvMdN~%8*BCjaG7ftn;yNeMnTlD8*T3vuoMNHJwzEc{%Mg
zFkUp2w#J1(*J_R$A<9fw4fCSmshtk}Tohb~(~4Q?8KgqUFp+RGuRKSSigUe$u2sxJ
zU~gKptUQU5(Ul^ws^!MNds5(nxCp0KOf^mx4ZX}rV25irpSJIopCut$h3m6X<g)fM
zrnF*yQdF|oCxs)?vXF<lY&<R7Pg%uuRYS*$RaxaQnNUgdL_d@X*cv~`n5OcwE84h;
zghW$^K#IbkSQbrm_>u8cSqm|AyKi|gE=W_uR@V}|R6ScAvo<HK8qeOeqK*J)CspiE
z%1D(}pfvYROD+skmX)LSL>8cYU-!o)7uq}k1@|TulcL7eHAnT5vT0Zxo?!Z2%zVLH
zAKqtG?Q-|{l&V~D_ud{q+!=CpcgXQFrDHW;d+R=p58Qe87SqKXKi>zNk#@QSKEPXt
zUkv%$TN~uh!N=dZ&;N3z%kNCS!?)jAk@5*Fma3@;>Kxu^YA-wyYc$p>R#ijc105q4
zUBL?xOdGjxs@h?rAgFA7RFQ^ABL{-u2SG{RWgfvng0uxRjo5#MK%3HIj8qz{Mlgiu
z9uq9!&ptTeKYZa5mo{?*o{(z8`M`H?PbeOu*S~bG%g3)sa7arTt~?PZkP4uVt*F(C
zQ77SvZo+#@Io|HhDlYcZHGjZp0sGqTuf5NFU>$GXO6cSCPsaCg9XcPd7e^w@X+7<=
zg3VsSAopie052t8n|uG_V?&OeC9|5%p5^4QVB$Q3+_E<*sl70-F6D+{ZYUek&jTnE
z1t}qLx~RD{lw{#68yOqjg!c|tY;R-~jch=*mk*F9n%!X<`}jcVU{*LbI+AR&jP-pl
zP6-sU5hPYITl#2uCRb|~lZS`75m%m6N@>Y75G<sZDSEPD3LX*Pk3w39EH!kKc2kR+
zIIOFV5L$yzp_eXdkM}ZC-Ayd>xm?R};n?gZ+&e8On!vLcdo;CU)G>sROJ&x#94#DO
zEosHW)UZ8BnXVcMV5tQ3%Z#Ej!a^k8gu^^%S_lL1WZ6&#@kShW469mBi+*Neh^t}W
z!>3++lI!X+-@199d&?ysef%=7zx^J}W)xc+0$6Qq(rCl}>557l28nFY-+%u;x~XYS
z?_(DCNt%Lsbe*h!p7Qhw(!9jaG(W08#``DV;eYzn79aW6gkO4b!n@6a@rN}gJrC*-
zrN#N@TtAU8QY)uqW(~G0Jd`HVfHyjV<7fib$|j(VrmCb3=tYOAwt7HooOcrI6~|^G
zF&5_J!!dE~bf9)BnyM-q;JL>(c<%Cuerlw(2%^sjja0C&K7<c!rk0<7a!br#qc=mQ
z6-993FxAA2L(WV=v{sydXal%CulZE$?;$e!{F%2zOEOBdqG?HRo;7YDhnZBctxtf3
zXcl3GGH9(-1GOWTJSyJYYE2V%A+Xa=gjyPjFw??wb<6T#QZOwXTV09E?q-IYhXu>Z
zujPlvXzrY>r1vy>%~!Tc`o}c7{R+cM@&0kasGCG{5y|8;$R^Y7e0dxpKCA_*T|h@}
z{{7REfP_8-+L$k)N`_cUwb5R&UYE2Rt=6i(mNHwrfUm_=G}VgI2dc`iK}i>6dQm#a
z6IqCDJaf?m2C0lA%f>U0y?)UIETRV7$P!SBQO??xCbtTu45!OxO%gkr$?uk4*qd4_
zri+@)BqTNhfCzAUvXHdl)TlTH)1+20$_>-f^ZMNx7dK_p_1H$vwe2l_c=v?cSuV0|
z1<i87y$|jP9emhF&*t2F_mrD?7q1OjRZt!6VXT2_j&U=*>XD{7P8)hpd>UQMs9eGe
z&pgGwqY?l6>wm%*KeOZuZ&v*6?UeuH;#Zm8`#giRqVYp~>*$k+e_T<!wLgACT$ll>
zAS^x;V7zJ=WYXt{IL-@7%Hzfdss=1U*qcEX-s!RyhF~MawA4r<lz@yx4?=1EK!{)S
z%Cpn4F)UW0f<%@%rFie~(Z2svmj?7QIY6>_P$(@pSnY%pLJ*3%7GQwTN(19Vt&jS+
zIYq#)b(w}U{)<+ejeUcU;P)8_*fjBHMI)&<MnNyr>qzHrDunSzGXbREF3Oq{SCJUa
zr1Wd2mcv=am95TN@ME;_;V+L;dG1((idwL{Hoa^&fTKl?_b{H<oa;-d=g~@L+1;eg
z4HUD`Is~5F>8wk)CW>-WtJd+~YX_m)vpdWLefKcPWh`j4T((u|$+~8ZQ3NDCIMs4;
z8+9hBt%G1AEpMWXLL_Cx-YKA>Pk|7RT;!Pd64^XvmFIZjQME*)cT>sw+Zm*E6D`3g
zv0XS^G&mpF=tz5FwNQ|)sVQquH#77yOVW52rK4(uYq2ODT?~1yI9)b$VvEvC#2{l<
zG?YffU(-^&9UkBAa%Wsn)Q)ewewTW)Cq1bGSJpzB>h@^*L)3DPIo?O9lylb}<E8BZ
zojm1eTCl&hL6+wn-?`1|*t2RfOt~Ok&M1|SXm`Wj@d?X?WqA<zzrO*K=M|sdcl`Gk
zMyx*hfXW*P7X(<<;`gt$B28ils1_Du;}m%!WVSlU0g$#WnbKJ8se*h@sS<m*ZW{gh
zU1Vr_4_cEa0*uy;$2l)fP^HF2^gK`X+G@V8#kskOm4YE!)@zLop7D<{{GPcmAP_g9
zfcOgrslWtkFfse@Y|$$XA3hXH+Wvmq_g9g)w=VD1SsHK%GVyR;h*529Upv~gEhk`@
z%Ec;NU=d9qv{ErfJ1&Uu(6Oo&C-D>-<t9d686Mlnutu}hwe0jPN=I$Eml=Di1&tMu
zPW<(D`w2}D@Fg>fel8a@iS%R_1k@YGIk<P8Je6^v){5z}iB@59DmM?6;~*C|)e<z~
z<0c`gMQ%NMOpRjLiI!ss<Em9ePOE72k%;`Zh#7~8)$8XOFXfANK@3e-by$yRx~XI&
zwvNSh^s(<{XR^b`x4V%7EvL+yQ>26~m?y#!C@LrAf5*~`7HMg;z}oCu4(Fn+*ytvB
z1<Obx8s>UkQXbA5IWgi5y{yE)u&h0qhOKVGC{IXJ$@}Z(DWMi`fJ`KHcrRd!Py4tK
z*qg67sTBJQCpUJQ5p2pL8DKYWU<W(cv`><E(I&<82B_YEo$d3GY3AXh)WsM1$3N)u
z`qd$qzIVXS&uU&i-zV5WmPzc_!pY{j@^YfD8dx;4;FK1klu}~()z2+TYaEIswbF(u
z!4SHc#>I-D@<hLfGg43htkr9#Uu&wVv@lDOM95-!ly7^aB5w~nAEJHPLyP}nR~(&*
z%)FKdvX&DoAvM`(Xd<1uOl2O<8)kg$b^rh%07*naRCs)liVL!q@n~lCy610qjWvBC
zGXFd+CZ*O>hL!{s02`DTT(o_*<;-g@vtp{SEakLmpP&wgtyGk6Nu${8h$!apxZvU_
z#b`|(Bne0>iNkUp&JB|Fhw0~1Z5%8d1S#c>QtS>=$|i7cT1h}>E!NhvQu?aRuBGw<
z=xh&@^=a%v5HpMrXW&X>6H5`x_@NaC8LNcm>k$@J=>o%C%Ot@=6?xZ&n7^0Gm()ol
zVW>r3hnZ$vhycZEG3=;fZ`TAV`%`<ipxX;Rj3%55r6uSh{@qsJ@?cWYNrVab{&5v8
zy*2M1uDH6D%gD0{?9XakfSs<e+YaZBQ8z(H@TjZ<n^Any%Cz%T`2W+Di{1-?^P`km
z{El{d7N;8~T5<PiL6*&U_rVyFlpj7g!Bo)SInPHfZ1TpvW6HaC=%xuTJ$Z$zTRpCx
z>k^#jor4(~#l6!xi!>!S9h7e<O$J4QO%iAtZoYL39~>bxc-yC3{2~wP*LnUwe~z#I
zkH1Iu>}9ggKZ#eGyen<us`g|mP*x3Ds>u?u2~QL(N(frqoAX>#mJQwueY|oWWt1oe
zy~b)MBe+zKRbypSPgNio@vv+SX^k=H7}U`cjFc87YMN+Vq>Lw41WGCdf|6-j`^b%6
z)5=@AeTW;riT7&@%snh%`}5mV9)9ff()MOrONoU5=lQeyb3SvaM<*U8KFGwY9W{a|
zXG4(F-$b&UJ-qXywM4K$L_w{Od=pwEJCGYqz*84XmisrUrUMcq29v%koPAq5K3GQY
zH6yzdz+P<B8z0sQ!KIV_dr>yDBaK8WPMW~HkN^v3HHi`ltuYZ;OAYhF(TQe|Cr-Qt
zY9AOo(Yz}HRk>h%_bp7?#fLyAu_&WZL8#9%IuRyWq2#%x%8-66w~DF`oM=UvYK$>7
zjl-juMkSxMimY_Z3Mm^CE9M(XOeZR|@V~3Zlf){*JF!kv&4N%bx&Pjq94#D|hN<+L
ziDDY0Pb^*Fokh)NZYlN@_l%-rHLH2evTDd@8He!(m=_M8M$(FcgA=JTmX#zOhY-i5
zAu!4;GbgfWg;ZG~HQXInOee=&GBu0od)%-EM}9`Pms5l}TXsb_IpE29j!XB2Pk)%8
z?FL%s+&eAUzdvTs%Xl%HF`gZwLqYI0!JdQ83T;pE22^kaZ9uKS)HF%WrqR6R9sOT^
znbVucFq<$v-9uLm21QYnIOVYc7EVU88mTUwYfzPA;R8Ogl0+O^z^ZC!8qYkAfSbmz
z<K~xDjSg}uE<6xC^HoEp!5WLNrqsoP+3Xa2G*XGz(iK9ABf!eYX79bcuk#NH3E!U;
z_;SAHU$j$#hG@$b(Rv$8cm=$9Sn}rN%!^>dz{O5OUHbKGKX|!E{^bWt9#oFKWzAo^
z){%gn5J%)u{3aTRc5Kd?t)~9j@Bdy1#4<GysgY;B|078|1jcuN2*az4`U5GCykFZ4
zxj1iVPrVSy6>|DL`g{RFaod<;q2X#8?@6PtUi->L<Ij{)>df#`tg-Y@e4oen-olI?
zXVmNR?%|UA$5T#c3;Mk-iPnsI8NJkUvZ%0DbAFid&ix6GZ4by&jf<RvRpr<_nKPN3
zG8hgBLDAGDzN)}lk~E`O6;c`pg=;DyAX$sj(Q&4!@y?;NVYE4--^rQHXMAL|VB^v=
zq*mtpm6NPPgNmO^u38-zx!ARbHRux+<&-hzD^cn!cp_da6{Iy<HJ*+YKZf?ZaNe&?
zQi{ftC6?3ucgedWI{j_LNl8dgR*5KQRyodfY}-n*a^g@Vj?Uuo*WPf60q+ETj)&*^
zAV`dAN8jt8RZf6bWi75MIC%Z*bWV?Hp5Nh(+b0yWi)f#c_BIJkiC-;HevHCmlJj&1
zBYd@@I^D;l9g=KBxt!7M_Q^XvrsE@2(}0$wC8IP+)+IHD)pCJvYBY)@%?aM&8&7H+
z*ETzJ`aA2%4uLcMeZ-sOXK{tcAN*gh6>D#X&tFRT$hL_KH-d(14#3&-g#d3Id%kk;
zD2DKba}`gYi@Qpk0#pwE;I89Ab$0Ln+=k|*a}l@mEkqV=$vJzS#s|_Q!9M?)Ux}m^
z#g9MQ6WgCXKi{7o60ZLOH?E&w+g2?bFJl(Y3^|&}tDY1R2r9sZl<rzff6~MgyK*9<
z&5dkC?azxSrfH22T2l^?Uh8Z|arS6m=P&;CF})j~=fb%mll>*{-8|y8n|o}UF85BS
zY}H+!y>gDjrsQC};1lPzxs}g2);T$b8b?vo94~5)R&y%1;P{>s6`F#ytSHl*ZZAjI
zCEY}F_qv(pw5+L4R+y>)BXy<;HFdSZ*O%Ds_wdVOUb=k9jZgim5SQXbE*e4fRV)dY
zjmH|%0knDVz0}BQ;=>w9TGn!%Cxxe<Yw^;LBcN67sUxtPN0QXC7Hq>ikBJ~f4cz?J
zzodWRS*~8cu?}?I>`3Zv8F8X2rZ(k?THBW;kwDgpWSTgNP2&4#m3IB;NYtX~^B7as
z<|uYjP32&@Sg?5e&oS@6&ADeaf7m(VJKyS(*(;nImAEFOt{iDyV9kgu>(Sjlk1EDA
zrv+KwMTHDsTT<)kjE49!M|)8@CV7Wqd`g;USZncRMw(~%vcw0$Zfs)d+6m8&Mr>bt
z8t()nYZ{M^Ou@!^(QRnCDMK7765yw=9Prf-P5}7EzT@Smu85ty4-b)w!g}LSz$<qq
zeC6Quk$d{%TOFR<@N`EzvXH&V>i_xM$B!)jk8k9B`tjj<(fiOwk)62>wHrVd|IW6E
zdELmD_pmFDlqqKqDLn!m8&4}0Bc~O3C%gfTXHo?c11Xx^fcH`YhbY(y(RIbfjY0#N
ziX7$mZ;TT@coXD3X)PN2wcjA(VVK5Kv2>B#5KAeo0!`iU%Iz6PvkGG^AGy3q(KG}z
zWV4@hce3K>WJ)(r`QT_lH-SafaQkRVSvwgu))IHUHRy3VJHcc*nTDme80E-uoZA|5
zb)(DavOpU{cTsY?anxCYu1Zv=!|vc5`{PseV$P9o82FOlg!HtmW%RVHJiSy%D$7bd
z5XvS{HsWSujbd6lY~;I_jYy4WrSKfa3vmou76FxMaloxT{Y*28m`fE6TZ=}fJyt`4
zq!$OJsA>nU7Oq9(>DtH~RAPX-)io4Guf;1ZbFv9C1@Ub;i-2rRBRYhDj7&`?*@P{~
zJ1-=~r!`PnRD$_f*Ra@1*!a62!~Wa5{L)TN@u|D~=>6MF79Eo1Hc4-Y>y;!72-;FE
zXCjI2p9Al3Wku5I(iC%+rziBcw#hmv+EwIvht=^B)v_k!T^WB5`>4vYIGRG1Vmm3F
ztV`<*B~?Q}Sz!?>qP1HaRUS+D5Es|SuW#|A!v*&jB`fFoN8i26Z@hSsespL~bWEl~
z3~s)Ad&0N&A59s)lqdYsQ=3eW7qoTmstJ7Q?c+xP!j2Xh_?K@@*y$KXndR|~G#(Vh
z$}zs*IB8HSjsdlnk=Dar`0e8Q@l!Sg*y`nL=bk)TEEP?_M(?}ap2bk5M*8!j_G`0|
zl?&^(Vf}h>M%zoIr>lLCg&qkSMO-8fX$f-MdYnn!(P9glvr$ND6<IL+Xm84&-@F@5
zIxL%=jKmnWyD6{F7F^MaoeLx0Kb-N@`5qr0PS_pw*y?3WRwV~V6W+gj2)zz0mo!Cz
z?sTx-0pmr%xM;XPT~aJo1Z&ZCMOsz3ULR1TrxWfU9ig=$baHN7y1-=o0Ar_XUBRqy
zWQmY@P7C>bWt=*sMloG_Hd4z*$I`Ki*+SBa%IJy{0x^uFwMUbQM=u(wl+MNRnUv^*
z+)~C(#Ur%kUaH8V-+)okdoaqOotJ5ke&~>J9`SHQI3q0!(i$0s&PvC2*NU7uPK6}n
z5ixNIu-!8(>o$^3fY-VaOXK~+g5Q4o1Ag(b4Sw^^lIz#|eDl{EzV?kS`)_%I`w>Pv
zs?`x8?T~I=Ak0r$o*t6egtW84YI;c48KFCU=HoHTiNPw(%occ8kq*z1^g7g&Dbs^7
z$~kN*GTzmq#4Ts2OPfL~juV66y)gJp6bymP?Y$Qp@>mJ<GRqe~a-M(k$~{(%<H54x
zAAIXWUf%6XTVVyjzCB*?>cO1VkD2mb>ZJUQ=eL1YFCZkkA7A_cD<AmUK_ni<J!I>F
z;d7V!TplJE)Y>_#4ctkU!p=)!ovk8Ix_z{#Y#l=^&(eMZCySEdD6EY>KwN`lg9wk<
zjEnV^LbZrDXO24|cCm}10rij~vGxpOMq$(TF+_aA%Q!F428sIUloQ_vQk|&y_ls4{
zOV_sOSdo@q-0boCy)h`m!J_2Ja~t$?%k9G%%T>wEyT@$gDXD=w2NSZ?a{2rg7q&-y
z=iLXWZWl8g;;U+1^;|nQWNXl4x-97Ra(r3w-tmM@V^~^4XVhny=FH22ax&xobb(sd
zG<AbV<kr1R7DAhbe7bC;kG7iAm0PFy7!CKPf-`iYR68|rIIpRjK%N>lJ3?bVp4aU3
z6R|$?Pz0D24d;3Z-pdHE@&bHrb%hPsnx_`ER0TyNplK6?-)@w+4y|H^0<vM?#c{_+
z>a!TDSdNzs+da#&4yfn{VYHB6>Uav~@jkLf?u*I=n!3jOP2PLwXP8c3<>j9_$NBFa
z@f&WJjgK1sVeux_aTjZEpz}}St2yOtj7f5O*PcarM?GJX>|8|~gRd*>Xp<yK&_JFi
zER4Z>$7;49@AomrkmU)*bb(PYI5#3|B35CfZ)rmG6Nrg6iG{imuS{ZaXwXqwtQ5PO
z9sbtGFY}MTFPi|sD+hC4IXDv%{iMgs+g<+3^-a2|!8!T8!g;>*!(;A06ie1Vo`@5I
z)5@)-#3zm8pM5yt3m5ymu-ifVur^thbS1EZ7!N&2TlU?fU3I$=z!{f#cbJRxG%ol>
z6LrQuJYuaYjyXDR0_|)vx|zh1$q+*;ln^W=7EH5><C4s(wU$6BqS0s@49a@(x9>f0
z=7A21%bOXm-<y+X8T*qJE-cv{<ebb_ym)zo_m5_jjfbY<!;={=J+?z9lX2z7u*Z15
z;_zU?@oYhsrxayHu_`cLMoz{#?%q3Kd~Sn9SrHCT<a5>ys+&<21)-^_bJ?+#wJ4<+
z^gCi?;oE=)jq;vVqv<D_VQ$4BB(nF)NcI`UwB_0h2;_W-Q)A)5_fi>mdKcJ<W7lY$
zLyn?5kSOlb!LVaQdtp_q%EYExE4j?pDhek&j5>ylCK1hRv7t#JHgGMd6*%i_tJ<&Q
z$QQL|w{O?gWC+ktTFnBiqG^j!vLRGf5rV>M!*bH)kHa4S<!5iO_kaF5`DZUO*h~p|
zNp*6->hu_+8!TNyQ?r;gD5c2~u@Ey!N<dRDmpE$$u&xsMj;zI4nG__c!c7-=C)1wU
zWCf{1sc4}VxdfVcpT}#jwU_qOiO>cw7G-!j{4Q^G`A<H5l`sEbkM|~LHsL?%(J`9O
zTpaM3>zne$A?v2{@c;gV#sAYgJwAD<FNb$LR6jf|`O2NydL#Ig2eR3Gbf>dsEw<~=
zM&NjzaNCRI6^QD&NGhnEh-g}wY!y#ltynw!6kf(x3UR|p<IU2VC)l`&#LZM@=3eBj
zquj7+%^Q&3u$yX$a!Q3Hv5d-x)QE#mn(FoY1aT^AA4WkhH@td#MJF+Q_Nh(EX2ha!
z+?$lVb^nwf+&$s1zIc_|!}s1h;xkWNV6v*X-0d=}8Yc6C@uFa-pY!t7E&lD>_jvaF
z78f?Vymf!f(cv*(8UDSOuXAa8B;~pmn(y~MJm$>@rwq^KTpH#KdpVVZ_xFz3J3b}r
zc?>?<i4*nza4@SmnTwf4ViduNBCw4Gvs!UdH9R;idFFf<9T%d~;xs&pMj*zD2BD4^
zPii!^3TqukY6Q4zx#Ko+9a=M!t+)}STJc~~(a#JQhN-yCIvH(t5`}swXU)X<a)?cE
zClOn;!$QuTDl$2h7UP*FAWprlWE$c1pDr8v3T!@L_tFN_J8$xhuRcca<q6xrc8Mpx
zH{*YJYL~Cy(7bZ<2|h575l(M2AK#%r++pL18ziA&eCIZoo_va8GG$p7oV#+7oy{)Z
zi$G}a!I;(YgmP80b9slyE^gsn!w2slvbEXa`lT)0e4oIA!pG<?Bg6OB5h?^9)q`@n
z+Hh84`*=N^^KA4}{`RM@^6vheAMH(f>uB*v$-kLeK6bv(#~vG!wTEaMMP7UU@UCL7
zdNh#q3)ePyW~U=<awIjW09QvD|4}dDU%WHsesnhe)BOdP2MJrbY{EW%zt+YRrjE48
zxRJ<$dI%6qjH0W~(pK6xS+ttuK%~3M&eCfCTYF+38Gcb6yiTx9v|^_(djD?ZLHhtZ
z{e)#KKaCdCj?~IY=Ynwg17{Rvs|81BnSk<dEZe=n>8fG8YB-uz<ca2ZQE_Ff$M&Gh
z!K~oHw4~q9=x1WPb#J`j%DFznPRh02AywVbOD#>~m=+byvGDzU?DrgsRpl7Z3p%M`
zwyHRstWZj`Kc1jmO_CTM9M9RCE=c-atSKR!89PW^x}*eGPLhX}iq(vAC?z_G(m6Vb
z=E`O+)sB`8d<yIgQ^{CxVZEqq!~<GeG|_pS8Ns&qnY2^#=azuvy?7yRcg5@lqz7GA
zB1+0rP0vL9CgM6Oj`mWkDPze$$i=m{H8>rlf)NCQv^T?qCJM$17ue_;C^WuaF`4#o
z{vyp`!M}R*eVT5;|NP0vnf=*azUVJ7&nCPveH$BcI=yXH<$~$yDcU)_NjW*0;a3H&
zu9?nOq**FPp*U8H60bEnNm&&&vsJ}vRZ$fcMblwcIFtjGiSbJa0)}`Kbp}fODEn!g
zO2su4LM*i9l#|44=@xnNe4i(Fdi=7IDb9FNvDKB9ti@6kGF{LSTl><RN9>(3f+b_%
z*=;Tir0frfh^bw;9ua)~$`d1g`>kVHr#|rN!HUma9jpOO5l$)Ad3ULO6bN*YcigrG
zMn!Q<U9Xu-?abbB=0pDQLq}ptt0r7jo?fCKX*GneET>Hq)^P&#qU-PJLiF~XRQlES
zJ!>WGuwD{lpR`mTt%gDY=lUskPb=O#nX*4#Fe_?OYw0D5EHT_Vm{U|Wk6+m0tp_JO
zu{(^@2E}Aql32s>qT<yLj~FC|2Pb2aG{aiSol|wqpWobLv)5r+*Wg4-I-Slib%U}A
z<Gc4c-ap28kI7Q5>~2v{7x>73utu|3HS2b!o5-eII6=_c&)KnR?fN_Fi4J30NNL?k
zHM7Do_7YO!0t6>`$*T64$o;Px8UKY4m==yS(d!w1+ly-@B(U1af^ARrR70ahitK|p
z(A=L&HUx3jzlfr!7MMI;H6mTsl2p`qspPDYs!~SDo}dIAo)u0^SL=YQ3Yx_pdDr0l
zJ>1-o=ALhK3;xdMZ?O2oH+lB60k`!AX76LD!AqFJvH$)L=_Fn9ETdU2W8WWGPNwV^
zE0lAnpioVP)d^mC7N>LWIS)-kQ!bgFTHakXjM6#RyAn(0ya!|OweTIJ$``J^j(wJE
zS32Ta9NPZNdpx3Nkg>Kf3eWXZ(XMEX4t~vXu|~5fo1Yj+`diO#bAFJni52ZyYU4tp
zc{HPr<+)+D_7r$)TJj4FWRxzu0J@boX7;S~KU=);<Cs7Qft^7f!324<lHG8&2|e<d
zEE}AMSt+ETStYvawtBSsj5}RK&VA+NVq``!?AY}<qmE=F8-b)jX0TSK29HcD;`<tP
zCA;9({*+sH54pR4%w)FUW7oF%!bf(wwwaR}!^bWSxV|&wXP?^P(jcXqT0VJglN-AO
zE_M{h<1sH>+2K;ZgD+NOz$YF*&*0pMs;L+sAHwkin#nP4?;h!LhS!R$si^Cca&|&f
zmz*E<`S`VSYz%v(Msoc+Q6JG7v9xhwWgWvJB0$;%u~IWizPvmJdE7iIX&TQ~*N{bp
zUlLDak3@?f&xJnRswHhC@2n%*VI33Se&uWaM?97G<_)KX<0vLUHSzf>Cl+bTTFT$j
z1tx`KRyf&Il%y+XR@ym`lg?<tcxo>b1r=}FwZDQ|(XD@ku?rfrB1wM=KY5;4-+Y0;
ze{;y!KmR!D{S*G{YnybQx{uBWsNpGFPk)7c_ddPNltuAv@HOdRNWHvEar`Fr^ghMO
z2h__kql+W9ubihTPAHD<uo~}S@*GXY^v;hs-M@#{fIe$kF?yXi{BTk+3Y3w=VkHV6
zlos}(w(-zY(W!Uc@~HI%_sD%5lw$esNF)BO=eD`HkxQHL(3jQ5lC>3j`~2rSXY~8R
z2M(9Qymdih^>lzUYa+r=jLvRJ8m*|H_G0&4$5k($cdSbb7l{h%Hy(u(gDA?;v667{
zZdCCm*G9xU-MAp)%dJThmGkTV$~@F&l(mmpe5{|(A)D2?uGmSJWy6Dci9=DB6%S@5
z##(AL)@riUVzuVMX~F(NVz{zAW7x~cJF?-;n!sUMlaw_kNjO=0I!Q{NrIa>fz6v;R
znAe^v>r>`kd;nuJv`JYk7cA<AqLGu)c_G61ASbSKay{E3s&Zn(X0@;ZD;F4Ka=I@o
z&&g6O)BBl;;SVBHwN}8Y)M$EEu{)G#@v89*GPzN!=!T<}JQqmmAA*Y5kBC-0g1NMN
zf%DR;^->YPj5?MOL>m!8h(2>NHL68)j`valPD{7GFIqEB;&@6S1h>S<1R*mDLrSU6
zQ|cS2#j}+DaTe7lxIcN0@9sU#|MrbzUcZs>)cu<OB8S3$8Jpi^_k2d>KHy~h9*vq%
zR0Tz_c%@l2N+usBCn(KyT2MH^CM?UEx^BdpylE&a32jvnEyrci*Iso0wU3`e$4P;f
z>uHRRHt=$-Lp<Qx4Wqs5TNbN{l0Fw8wT7a0{O+rJj|P%z#c#$y()CTQb)4zP=Un$-
zK*WU(R*!;)o#<hqlzc}Pq8Vo?vo)pU3-6>F*R+70%;Z-wNOFH#a=9~DpFl~hIf^Eb
zX)#<#%vrWzo~U(o(VlkVQ4U7Qf2XA*X@tR6`@neBu-Uay^(G&x@nH=PDkJZ$jIZ(N
z`7fu*-n3vof&FR0FWflKufA}BUN_^9ez4EmcgE`@Zl|B}M{nF?Su}ibFr|}e-aS;@
z+@G>ql^CP=;>U0B<<~#p%BV*e^f_3RoG!#K|C3K#=82si_s0vmc}m@QzI*E)i=rgi
z+#oo|)nSiWS#dC3@%yjbV#lm-K~P#b`4z*CY><mOu&PAfYdfs%$+}j6y2C|-)v((a
zh2FUIETWTAFEdOFk5v*#nkCYrwcskSCJxCMQLcYC5tpXY3DGD-ET*hI(<0!#r<+;M
z_f72QrLt8aI?y66N+Hggn-JI?B#ale%>AS4w2YGNWhLX~RqYw%vYD*vfU@HBgmb8L
zNZ5D+V-0Tbl-PZR1?BjNU@V#YDSr0{AF{Zx<d=?)`ENT59z4IzZ-09eb!suq0XF>w
zN?%g$-Q@h08+7{vd{xlz^_c7(QmhJ!$&`Gs&GqM=qb^F0Zts)jDKC8XXX$f7(8YQo
z69Rc+s3U_?1y9*XJ7hE{G&%&lkNyN&Q#olF@Ud=IvO=pklh>h@1#4Di!|%TG;E~1O
zF`8e!>}2sP1;MYsOBJQQ?cLuVyI#S;tm22U{0G>yMnZLv4Z>*gML;P<E^0-|+A~7f
z=vYPR1Tn261v0g*epxDpJc3nixM)zKB3xBIqR+C+v~Rc_EBc5&i$!tN+eaJg=~CQt
z=A~cj*xTP#StOtY*?m(jH=KEB{Rd>z>gNgHdvC#W7Y96hxzDw&oW_Zb*4J+ya&J86
z!l;Au8Sm^r;PDGv9E=w{cX^BZ;{~tWKgOuQCm!46r=Q;8Pv5!6W7{J}y^Mod$?G5P
zVYKGP#Szb(@A5M@w&^B{n+GfQPgk7G7bs(SVt2&HAK&7;?;SF}wTC)6!F54-&ql6U
z)Pa69BApk4l6Ml#AX6+VIrV#SObcf*(#A-%*>2A=k8HiFiETrqRck_&vPyMO)SkNW
z1Zz+l4i^ohj=>nsmeo8+<hzzCPAt)?QZ}H?T(3`&_l`>jnPIDI*1)Y0%p%J(-i&D?
z<x5ffn7bgL*lE?Q%}6t&*Mz9T1&Yc;QKPaBoyV@Bjt?m}&NG~!Qa1yt`92;^XZJc*
zz08;Qzr{0;)x7+T2PAz%u$QPjbVoJ4t!F8Y?&23mY(93GK_|hLHBVj`@XeDMez~Mt
z&bavG6}ERba7D@K$(+%^^3shRX2%VoEKu5@gTp98a2}(LjQxUm4hY>@HeQAJ@TrK1
z8l$5sQ#J=BeD$c?kk*zUWVwIxy?cLY@&DHIyL8G2XE%6@IS|M8zLg&c22lAxReS#M
zy+<j<AKj8lRwI=ffUP6nAFbkPO*HdRk1*`4R^+LCfVM0`D^%QaG~#B{mZxjhU$pD)
zCbCILZ%^!b#cf5(RfWQf(M6)rsikq!C(bH)y<x|Yq6u*m`*9M`dIN}b;M8c&4^v+G
z@PwQD6Ltq3W<|r{Y>C0M(aHJ#dk0vhc=_5nRyADN9&o&<$*txi7q<YIH}LIu_UNZ6
z-}zuq++?zhJTX|4@XD<R>>uUK7OQot@_4f3!E}MLDJQ36-oE)BD-XW*xIv$xo1l%6
z2&XvKY|Sk4ctWNcoROZoSTr$~9k#oM!ojSNYgw;q7SY#`c%rPTKsUNIyCCR&(FnLX
z>WB@vQi02(gjL-VdrTzf$RZu4ib*LF?4p()zjETCCk{kJ251{Q6^UR<=&At$g>5S4
zs=`DQh2mycvK%sd)?SpgXV@|HQ_b92T(Kb8e_y`;<r3HFqjezdoue$K)W^3_!+oyq
z9&mT3!|<;?&9{b39{>O#07*naRK+(A_<QH~`O}LDue?^^%P*r=8e=YU`>nU=_IfDb
zWG|zbkMVU$y0yc}$&{jOaK(yhK4(IY|L^;^$(loc=1Qw1jj6*Zbns_h0Y+n$#~`3z
zAdp4?Fa(7$o=oYeLKOXsHWE`7e2C*aMJI!gT^@1&=HYtr|IHUKaDI?-dJwUL_JoTg
zLXg&}_4*5OZ}}JRJQ_f9F}Hl`a$h!TrJ~XoY!dsm7WfJA1|as@Edx-|{v0lDbRrAv
zY{4Ba8d;!5ul3IDsT`Dy3tN?#WVM||hDrL6Y%*13Bx(g|CbeZ5i9`hHU2S{e<MUhn
zDEQcZo?fc?+;dwv=lSQ~{g8ik>kXW9>|Qy~&pvyFU;Wqx{^6?PV7la2Ub@Ub`tE%`
zePx5!?woL8qsQkS8!^l+Uwh*&fAD8N<b{`Ra5$Tz<_p+Pc=F;7I>77iywCSeC*T_z
z=Wz9k?emxLc^7-|fO%7MSV}4O#FN+g{3k!c58k;6X35^H;>u>uq^Pl0yctuY(3!}G
zGcA7}FFn==p*kx?VgiSAF%;DVHhO84LPI~Z5|9zU!=m&oD)Aod#f6(Cg4k;m9V_h1
z+{;LH{k@{r69uirf~;&jqX73n2)}N^R?jkBIZjuNq$xGBfmDsqi*u`4loI>aOQhe;
zqrXIIMW?S3Cg;4A^rqT*2~kZny*xoxGZse^`j?*P`0l&7?N4HsD;|IHapqOS2d}?E
zH~Bifb2b0Z_b;M~7x?#|`jEeO^bYy<<NV;t1{ac7c<}yDG1z^U+1?*8c`!k_U98gd
zyO-E~_M`L%eeS>d3dg4pFxE1<@&d)`m>+)YTWoC%`MK-Q$;PS`&MSh7D%21lwn8W!
zM|dLK(y<+o|Cd%FwMHc9O86ZD2oeoEg%$hawG9Bi`ra{r^JAB|w2|YQW__<~b!H%{
zwTx=ppJ}ZrYsc@sb;N^*2t!+m;n$wt5}LB}>xx||u>13do&MQrn;0$Xz(-r^^~M82
z0$DX@yH}zmZrOXN@N0re>7^XhakaMoYyF4}w~tj5Vu0ouE%a#7&`C`skVt8q3{;d3
z3k|uf15UNN0(q`hjG}<sCoAsVsd)Rrgnq91%!}6mc<ufv-+p(G`=<+zCrh%#@K3+_
zK8;trvp3;%QQ;cT(PSYw!hFG#Ph8;E{R50^FdfTmx#Hf@DUJ7Bzr4%E=N_Z1YCM{{
zuDN+IA<+q8W0Nj{r#1#0PNs}!OTPNbP2B1PcgeBQNw|Aji6YQ>&=MfhG!k&Rj4Vr#
zKH6%SlzuIux-dv7owyuViBwYKRTCG#jO5ylb{0iBR>?JWK2Rv(9E>u<tnyKoEYCe}
zq_oY9RC2pRnLFnXO<lJ8Lao(G=*LP4nWzac=osdOr}2uq@sw5oM9Ez|+w2n~lMAK9
zIA&3MXdEUVl5PyBmMd&;3tbdw0{0FNadm@DEJ^bVyz};Zm}vpL!}q5%{^_-kGy1{z
z`0t+FX8e)o`3HyZFx&TZd)Mh+ILFW=xL}wbPC2>rKGU>|9c^&=scS5c515_ollM{{
zfAP~~egedtJtZ+@LjRSzyKDk!1hw%(WD0TPjVo51pkkRXYeRO_;9KUc1eau5?9X1l
zw#l=XhIA7h$AxiQUQguIdNt~nejH#?JAUuY!$;VHxmNu0lR_zOEyJ6(WmXE^a=^7_
zt?;t3$A4I?gS>xSkRZsi@dD&YD-(e@2{kBPh&O|Zx^Srq@~B5|-^8kx=P+8)&9uz<
zk=Gb?%({|DwfM0aB}OGE)+J|avJm3$wc`+oWS~@h^EWcf6Bjnw-sp2_G{P7oWF0NZ
zG5gb!S<x`BJ==o}2pp{%7G+Hppqm($bt5lk6RiFJ`FgWpOSA00>$ms*#xveI@6F1g
zGHdMWs_I#-ZnfHy!pJtVvE{HeU;}~(hye2jyzt;)cp>14ARHzjjKL3JOF{-DN6YF)
zQcH5TTCE;y9<nOOn|bFmpYsiSZywhEzLVM2!HLMos?5Cio?#Db{nvkph4ENa<XJ|P
zB&10~Ka1HJ<m6e#W;dtT>k%hHK5;t2nuybj3D%0$mJO(XlZM<0L7h!9aAjreH^>a5
zrKBXOfY9fuWWH_W5toKBWsv=rwP(DnMMc@J83}-?YS*#@Yuh-{%KMi^s!ftzR8Gt{
zy^MG}v2IB!^U|@ZJewWq#jNvOtQ;0`WloKxLxVW#8m;JMhB|DR#&RxcM}WS~b`+JP
zl&2_z*EvOPiPJ8=tf=FdxLnBOC`#}uMwc7-qD$P{=IC&PfAYayI-k1+_l}vo@sR#O
z$*Mf1RyE34st|l@-HcT+r>unf>r4t+AL|_?T}(biM^UJ_6@*r`a*CxTj8Jt*+zAML
zm=+X{Fda}tk%Ys>k+fJ<upsjRy?0^bJJrN?D>{i;e?J7&tDSGtecI8Xm&2*39l!O)
zDTm7^H~%j^zujg<1O$>M=?EaoBF&-_ps(R_H-2J|^8(5s>WP$UDX$LkoNFYexoJQ{
z^~B+^3aD`!1&;wmn<&&Vs%<+ag}CJmg3Xr+EVmEOE4eu$VMcXg6<k%MeMn-Fy2|_F
z+w;+TxVo8ASjV5fejB$au+9<pGM;(n1~2Y*`N|L8Cy5My`wJiEz2gbnorE{;U-0>7
z_Nc98QT2H9&LiG>?~v!7zs_6tk5K0m>TZv#*Y?>fYvx7C>1@f-qf=aLP#%=C#JwJF
zJV80jTX*l^yrXyd3NPKf#>vAEN#Z5rMa}LY;b>fPX%Letj|!3B+JbihOR0UEe593Z
zw8N<cn-<p6i9<Lh;z=;bg%jW$U|_te=%*2RWLnoJfQ?K{OHO7r{Y(;yr4kI|Ed*z&
ztfdZ2(FWuzZ7^STh>LO!wesyjA^?{s2@20hi)7gW`~6tLSiEOZ3NSWWS@9n55}j5T
zOJ>s%!>wJ;kMA?wyMe16yVsv)URjQ3E2d{}gZqFY{vvMhNwVs@=-EH#bnpTne*7B0
z@?gx>{UQI&&2w!16&~Kt(epj91M=9B?S2N8^;wM1`0)G!UCl}JKC`nUrpG4?@&*3+
zXVH;nRak1T@LK5J6`++{SGismUL47VrvV*UfoZ@5R4m}3Mxm5|-E|cVQXEv_05)DK
zIOhm-Yu|?Rwu!%5zGhk0{O0Q?PjCi0M)Ql$ZF6auw9~yP>}$muw2elEW01&VC|ixP
z4-UIh6*T>I^MN-co^F^H0Q4foH00_vJcl~;?3s~1)1yK{uH5j1F%!tmPzQSHtXxZ6
zCxvCJ8@1BZej40)t8fqmeuDK)OwxMuJLmYpy$Q?8@u`=d<(bWtm4olxI^tq7<F>M7
zal*l{$HTJ;2OAw$tCH*6UEVpGaQAe?Zk`e)8P}e>&b{MPl1Nhy2WV@#{pgf5&G`Jy
zJ@z-c9FKQ-YO}}jWX1pet=BQivC1=y(R2rW%BqI4;Pnsh(<>)<=h)~(EKAEyFUDxa
zq?FrlS&0T94+fx}Sh0~tENag*l=%JB5Njcj*g(w5Vj-R+VR4+#Yx<dH<FSgW`FxVV
z!fQMan~F}XvpreJa42dkZC(;0!V8HvJLrqbFN+l2M2ukTS}MBzo&cL!teGq=D=X#i
zAeXjo9LBjIo{>f?E)ai$#meCdN0j&27!J7@pA+?Vsg0qIQ|_Fe<I57IJn`Tq953U?
z4{=BDvs{iz^1sIa{MC0!vb&@&?eXQ?7yQ@z9sc;zCH}9Yh&;PS)IFd+|8rL7?@=ZD
zWNA+C+70UQh~nf9c_(7ynOEq#Q{l!axup%l+$0FJ9KlP_%WIDUz*abj14^&L4peB$
zN?fiTLL=5V^3+g=P*Y_?g^S<`TD!p8mn&Rpg|be3I=o!2zxn#{M+ifI^MxG_21#2f
zD^y#}HJ$)k2|&xd@^~+bJNdq1sL_;wm6Od?2BJ!=1O3;p5o3i!c`YhWu7br>q)FRN
z8GiR`aSE)1L7QHv4l{A1gzmeUOPLu3;Gzr_KqCom=AMyO?TACZpOnTkAMxzo0HamQ
zpRb+g{YMk_1|6O~*rK1tj1~nE(9aU~H@bXqG-9ut6D1L=*^+)QXIjJ4TN`Mjc{rId
zuN+IaVz#R2rJ6Gn^Wb8EHkzj|?Q%FT*o+ertr>Jv?qAF(stQw9B%MCqozYFiUp|i%
zi`p|@R76JFf!WG4U6t$(5;kJZqVgmLHbSsz^S(JJLakNhNaLiX<1Q*k5(gS<?b@W6
z$OzxSIamTdxHVh9jxBfIv#McMSefsK_t_%{PZ6rVPQ1P#yrUKM)D8|u1zWuYlwz_d
z+3U+~Gz%FK%OF6z)Q_>&1)V{SR*D!v85YBRI{pl=jDVnvDRrJ>mW6DFsKYWJ5*2f_
za=2s<J-9|PF;qu6|Ja`M?3Z?U{;Lo9%U{0ApFh3i&KnatbqAyOP$tJ1FJ*X~;Nk?I
z4bYvGwBIMG=V&cub`&X;)xo;k6M4|avnZ?x(X_^T;l)U*lz@NU1u~l!`ne14C*BKQ
z;H{_!n|~wu2V!>WY1)f$^9M3ggvhyn{RfXeGB*5MFYd5ENL%}G?|th-=mWmzwd}Dd
zH;Os{nza{2U8KYsxp_3Bz~F4mrC}zBE>Av;R)n^+@f-vXqe2GviP7Q4k|!k*;g9(R
z28EdO()Ty^VT;m*sHZi%P%9zOGz(5Q_(1#cZ1obJzd9h&ia&VsA^+@8-zHJ;!qrRs
z)C&jv)Uz9$&Psm&jYs^_=dbb2_eVUl*Wv!fil6z|K7&l?t6zKV0l)d@AMm9YuXFcg
z%%Z5c)a`J2tHas6;s<w+`NKDFW1NLWf$<i-F=XKtrY@-+oIP6Mv}S*Io4@fB&vEbQ
z7;namR*ua~(<rHTdT}6mDK@%M8xg*$<si;tc@K@2aqH2vrgol#LBb%HL$ROfHVRs4
z0d2=C%SI}8)*4xONv2u^da|-?hKxuMLXit?T{jkN=E1lONVsRG8?me$_bw`S`w9I-
zO9mtBMc1U&bQ8_B?YxbLvtHASg%Tcx*VphJVj~X5vyq`%Oh_&suyySwAH4G(o&75$
ziv@ez8&qg!4-Q$J-6E;aiLd`OU;M;N{P3;!8NK}*Bpa9MT=@*E`XBMN$v*L4f0mu!
zd7uB})jhucHz#~>{|@hbXOsTHUt{m7%RKn*A5*^fXGC^My88;HpYp+X|6jIq4bOax
zER85ji*r_RfJ#gw6%v*!30e-k2*i3J6&f#WDepW*C6!cUbZZ?}*RuJm$_2YGnG$3n
z(p(`n<ISz6l-%Si>-qKXf0QuvOCR6k@*r)&TztUxyh!uwFn#cWa;$4Fy-H&Ak465U
z+)`3XOpChhxj}%N+Jzsa<W@>UPos@kd3g^f^NLwelASN=CjcnxJ<D3W>gJ_5u2f#K
z-A2p0rLGD^ShX@(@t*TP=Glsw#<Fx=+s=q0g_+D)E(+4v@QJH^wtF#m&X)Ak2()He
zSvsj9&%}EC=58)-v8xIyi!p{gO^6lj?+m%RGvI~G8z`mm#$c;j5@@Q5(m8sm2sVs@
zT4@||LmX`7Z1z%OCE*l>l`(4Vgei%lExDx!ln-WJSkC9Q+|*Y16Qf1Ns*unS6Ts0}
ztB^+^<F?9r&SzySLUExSTojHXXfCR{Ni}}*z1-?bwNY6w{saLao~@eEov1CBLBlaw
zT2AL8L5+iJlMidmhMO%7W8MZZR7fj03&tajO&v;FlEg$rCZaMiR!3}hI(+KdE;h~S
zBnf3%p_7bhY3X-zsxm=E19}^~EFKN{`qw-B=f6GW)=yuh^X3V^HjelYU%iRje#oj?
zh*DBp>ct0C_8g)Jr)Mnf``G%tO{CRYrU0b_uXLEF%UD*0e1di)rB%!QN6SH+#&SD)
zwf;<_NNB@o!Q$#*OW$e+e&m{!7Tohse+*&hS6|uZa?lAhv}BZc2~<v6fh0C9i0A;_
z&^G1Ip8R2r&i+UC)!NM$|B-*Y<|+h~(#dV26l|o@>rGd3iZ;<m%|nNmJ51#;YTAn=
zmOhx!>o+5oNPkqp_}J5IrLmJ`O)t$TdmX$n3==`5t6H?yd7iS>G3@q|U}B;;=*Ntg
zwK&yA5jK`vF4l^|!})Yh?HpBQ(ONUebK=2(I!>_t0RWdbdR)J>D*-1}h4P+68_s7X
ztFn^mM~L*w0%f-`M{WAGW^6YONxt3-fU2}eMoa72>cwrv)o{E^D`nrJXpB5`JDKq=
z2to9k(U}LTZnOW}-&xbkw|a^AEovpLg;Lzu7JE39SY?;t`Ogd2eh!h5Uq(w5w`}mW
zYE4R0&md$J3ZoD_5oyvT(UvTUiDH8dt)o()Y(+sv74Kndp+}cMskew-LN88GtIKGA
zo%!r}e(TX0*Z$V0IsEDy4DL<YdlpK&Kq*gMS5&%>r%zohaHT@q9UNJE^D3k1XGZuM
zMy6Y7ELswjjO3aCkwzo%qlbXjA-q%3j`x((#F1(<Dz(w=A<!H?&4J*7HxDNt+5Erq
zszj4(C2U-f@82W^E3I3j93L)ftra?pG^u&Acy1n#`vEi$?_t<Uf-(w}ruC(Bq0DGE
zp%zQ9A{c!*rP^Dha$-B>+eM<<m?RPWt$~U<SW|m18(RKs?RpAA7W&c0tPq~UbX5o7
zMU(eBESzVytT-GM6h+0+#f*zZ&Cyr_E(tW)@uC(6n{(jB1VY=A*>c4oOF5s{#70Wu
zSPAe(o>4~$Pi+mEElWnL5*5W1<qA_4WP#y0nXRPqaUuem7q(4kUe#WVCs>=Oq((?N
zmj*E#xoKZ_62Q^ILKE<;@WH-IHt=RQVy73kiAJT9Jcp)Iz>7F3HX&UWd7mri1IQ;j
zkn>qdVZ|)edAQVz+KEC@%jQpvZ0xgnB@u451X472Nfrgxr=}G_Ed%GlheN-t8>@Mb
zHwF`BcnpKs;IvFTCd(49G*ux!cu|sYd6=`2#h^X4&N-W}P*&lhoFwmYaRy%-onrsa
zCpdiTF287V;>`~@I=xL5-KLK2Qt5}()tuT`YO}=aMEY+d>(?S1)M-uqH~~cm$0{XQ
z6yX+bZ~NvD2nUI47QT{yBO^fp8hzNXZ6(m0z)Er*KYL@7fBm`L_2&P?We$ef`nj+s
z_BFP0AnVA5HPx^RquL8V(R@Kz|0McJ;I2_Cfq3P2!wgwYLJ1RfP9_IY(4G50DA~w`
z^4d!jgCq*lQ}Kk;feBUDo>}DtNzF~WVrnP0)E*VYNDUKEZVLHla_f|?nFPo)UsaZO
z?#=ktha)cLOJ2RaL+u<z>G;|kk2u)q@`L;5l*^LW?@c&gRvb-AK0KK-T2#bF@#g(=
z-oAg%jol&Nx^>LycuI}pjl&Z<NldIYPi^%WW)YKc>peU<<KFQ(8yg#PI4q{9txa4}
zb7!*PcfWrRwL0hLx&o9gYR_&jlGkg+_ItdvYjYDTJ^a*LdU-@Ikt*uZxMY8r(o1AA
zlE(&DTc#x_91g))riE+GE4>dnidGJ}VCTn9FJ@ke?e`YqGdNxOR=(TM<j_fD&2%N7
zbtCrb<TA6ifgofQT-(Y}3MMN{xe_4pd?~F;VpMxrI4{*>KQ&U(gwHTik`GW<1#i6m
z2DVsWjuK+$`Qg12EQ++MNxR#mc8NW|;Mc$T7Ws5UKD@?kX;~hAh)TP}^S4;(OPnpA
z=D&J)m*mPBf8)kw{?5G>#nun_7v_lbtrOJw0j`F{^t%|>$K<a-PbOX}Y`((U@csc_
zs;M;CgAqItRTKn4wZ$vlGF??TC?X{plS;|@Hd<3vwhfe#z3y9kvBwKs6%LKh+}H%*
zV3=_*Oxr4|nL0QpK)xejbCCdj4b|IO*}GAL$~WV*<`4iCN@!vI<-yEBkb3Dw_j(b>
z(~52;Gv`?;2xuL2?ahgk2d9@P-0*U9iRovPx!0JF3_C&Cf)G<KmX_^q)K&{X%R-Bc
z5Q-cSC0gv5Sn~j^SA6&Oh*eeb>1*3u{Zs}XzVg~5UcYz7Ylj!~VnaVq`9J^s4wtuj
zeB;&`%6r~DnsEF0jPZEROII)PiJO=Bjc>if(|cPuFRv+zB1-G{tFJyKWqx8vG)z|e
zeB-@i9*j%!NHfbKUcI)<`$rSrdT`9UcONovW6G;lsAy=7GK<<%Im`P;3pTnj!%XJl
z*EbRgVpVXykUYO<_wp7P)V1emQnv4RWyK(LT1bU4%rvE!_uYBNXi<}d!zT}hqmxox
zZpMYB8^*h%Wi6~uE5M&okoF?RE6E;gMuQ#|q~wjMOIQzDCPIhff{i@FD9v~!pKrq$
zO=D^KENiEz@v=h4DQaU2jvkWlT}Rs*XEe!d3MjnRT)z1%4#lH;M<k;&rmF?lKlTY`
zCx=`-x<mi+PclEhg{f!g=YE#NK{^}pd-qP5?S6${ez)T1zX<=u?Hjx{`z+n~f@=8(
z47P^sJoRz<{t=aji_+7N6=`CqRxr$Ev)8o(FVSP(iv4|EOHe6Ff^IS`6|3?xfpSh<
zxUCcZgLR%X5p6&fD8YEq*9fsl^aU-OM|r$geDeC{`qKCEdD(!+MG7LVgJ+~v<2H=(
zP2glG@6p~jn=b@Ogi^K+C%jRLUNC47N~&r_E0yzXb))uX86+Bm4|LoRF&;{bCOt-y
zW7a5f1135r*3inYk^ScR%vQly-UuU3z%D6o5+hDCE|jdz=J&u>Ct|Ccu-nV&cVe<g
zv)N1d_~jutuk3JTv&+s#pBMKB3_B^iy_Cv|Gfo<523gM5)&@x~sH!%aL7wo^Q=1gt
zun555qNupKk&p+A^eh&m`^bB0Tl4JxCV85$HORO+=-^F+NyT5U6RFnF(RsNs&u1l<
zh6&wN<h4c#{kP#cGytKzCsKY(uP%#eRyv7U*Q&Jw^GY}vi&`=V8UW9_wWZoH6D>y)
zKI<y@FU%hs)mEWBI-gbT!V=JFeMQv)#7iT=IJUbI10M=WvRJd3N!UmfiQQIVg<Ehs
zubCAxzEpU;G9*a{)$3BX{jIDcrJz_X>GgUjU^<`B$vdJAQ5v5NDYGsln^bhkV#EH{
zCcXWAqUb4p|MouhFFeb`KY5=k)gi;JN06Q2^AfKWok54epo>w$D3yv;(FrEDT5H*$
zTDPUW4=n;n8J|U(%0rEqtR_Z>&qKGpePhU~j3$ocmr?>QHXF7v5pik_NH%69-qYd>
zI^br)s9oFhTPJ*xqyc7;uZbGuKGBebkW8%<`H$)UgJBWd-IO?_!sM}x-%96OGEN*e
zy^d-50<+T64c_n0$^3t_V}#7(p^n5B+h}npuAGdqQX@_@n~@?4ph+4GE}O`s2B?Z3
z69p**B>R~cw}_4658l4ufBD)Q^anj&zP!PgZ)|dN_cCukT5#uV!LNS)I>+;hFAWmT
z<`o)ve*Y#Po~`)$8;AVC>v#CV(|b%S&%KK!8@-r3G8|7<{G+eG#h}w+wpg$|=(8#-
zR@SmN=yEt(@YaJ9`OvJT+s*j5e)bin<5P_JP7vBCP8VXGwLcJUy^Qa~S?8c1b2hIT
zFKYJsF`Zl})psun;fo`;Xd|}T3?>~(q}c04oXu-aW;LDEkQf-|hK<}XDI6QYuWwR%
zHgg>)!IHMyON9q;We~HfJ!kWZS&#;Abi`?9UN}ZeOFuKDMu;iSOAE5B<v`A3&E<`R
z+CCZIzT1n%TfmC-*~Ow}rx%mO8e1!>)e6o}FvAV1<45S7U8-fpQ~Nt4NzASLM?CoO
zeGcFI0ok=z`O+tz=kRRId*Ayr_O^C7c<O+Ax9_sq>l1J6aQ@DBnAR8A%9EG}sGSG=
zXMgfCHxE9}-+lNFzxcv;8T}hq`EUP^119q^Z+_>0;vhaiKl2&7c|=_jRE>@EkqgyZ
z;Otw6j)Zd&#|G^j-UT~%)!<c56s=CA?{)3O627p(s8zSF!pe!qpnc38SSz{g#9kd~
zL%gFIFDQiA<!n%2wweS)k?2B7kQ1Cq!@4i@?xXzwk3QrS?nIhiFmh-{J<hM4dtvPo
zyC@ygmF0L^1@HU7LHCd0450<^K1hf|Ph*VcsX<IPG4lP?`W7@l@>D-!X&!5?Z>3CD
zj(g_|?ww6|{q7ORlNm2R)#I6+j83B1$YWl<(qp?5G3>-dO7Zk|&W){vS1xxs*vPmy
zUGdV@ZT2@Z%AzDS5icEV@XTJHC^FnXIcGANap%qhP98iGgO7fP%}z?5CB#LEGltJT
zy~j^Ix5f3{4L}@s<4CcYnbsP+QQ!3vscx(lHJ^FRWJPQ7UATQ-Feya&H_Rh;d!n&#
zWYVz~;!Zy^tV&CwXh)094ZgCT2Nwm4(#lv)t)~y&Q0X->aCt1;h@D<U<>9^KC9BeM
zHZA$#g9RU)EO|68Ia^p3rDIt+5rKJ`I!Ny?n?h-+2J%?^0Jgg^X(XmFA~ANPk>R<g
zuVPHXAWN`m%F8z{@v-Zdhz5PSS(kEEa&32$k3W5htwA5PIHjLiUU}vzY+X|gcVPcI
zQQ9S5-6d8tlwY9EuQHi_mN(8nNB*;~5`X_8mw$MSH9g{9&d#+DvDtT9B;FgTgyTri
zf2B2XYy=q8Le5Z15yf&qE2G=Li;Mxiwh@a%;&3wtjAIc~Ri1W)*GvOsnjp8iQerRG
zJSPXOrU13x15Fj;yj;$eTZe>t6ji`pqITmv1_Cyk?sjvs#vvNx!DXQ0X)QBp>jhM>
zPBi0b0C<W@%s{uh5%VIX`#9-4wW8G=#70o;I+U2bVDwqlE<~2e6V4Oe`W4hxm~lV4
z@tfy-F|Qcgk`GT8oXrdJPD+%-luq$`Z=CSbL7#UXEqFAV@x%K$-@Y^EbJvHwekifb
zlVzBbtq=eJAOJ~3K~%-jv|<$4dAE<p+&P<b`}Bg?X#VJ}IeDJaOH=4OI$6eiwPG@z
zQ`9vdJUU}ETarhb8i)3d`)5;r@3kXVlM{Axn3q<@b6R@lCX(HR(tBw=stO+ePAY=N
zP;*|0uD)@3k@rPrYfT(=ZCxvp+U!Uf+c-AGMzjJF=_P1<JBITOj<Q|=!(4#MH00Y^
z>)7chY-EyE*i3SYTEN3e^IXG#Cf1GQTYxPaf;O^9iSg@hjAJbbQfTHy!R>nwD64`z
zGQ@7l$#}`Qm{QFbcx%yVkBjMyyGLV2;|V?=Fgicu?e{+<wsZ8^Lsn&pT}`R8J)BL@
zY8U56#Kk*U3~vrBU-;EeGyc7I_?62QuXSoZ7=iJL(5H<?Igc`)D%j&MtYBOXz@`y&
zzpShz@c<@z{6>em7EOg#k||I-QBXQ>g-f7(AQ*}BuuKD7>mBH!tpJn~WKlP5b&C@O
z_McK11GSf%-k3&#7z|^8$gxKLTL8_AfuPZfSoxq4kQ>9HIG?Q;Mmc$`sY8mc(K7q(
zrlQ&_LSIwa^*n#8V_GIzV}My%DMjYhdcNN*eyx<W7~xQe4a`>JEK^!J7@X&cN}wst
z0LIIb!wdNRcTR~6{Nhi%fYyp{zkkFZPmX!-bP?2rhCjPKCN-MVS;clYWwNUI{kM;J
z=i!L0Ud~^7=_-G6>k(U7!d9=#-O-GVJfmKe{Kc2Aa$~Q<#iGVL&j%-Se)z#dJ{Zlp
zvMCyX&wuV?ym|M8dv_mke_D{uj``B3EtiHVlclsr;`b*T@MKm=IUSn3juH-i8bH>0
z;kdGyh~M60hw%Bl=Aa*Q9C8BFsIgiXWVm$lwGGkJ1yE{#kcb&;xH0PRJi;`Vvw6)R
zmFTfp2eqJ9Y<EPz?>uzkBxFp;XJv9j5@|+@Iw&(GRAX66&cVt#M)R88e#~w^X10<x
zWmdxZv>-}i)W#0e(-C>z;qYXH0{-N~BO+@_%N2TikM!EJTpT{+k0wj>d`fohRTk$*
zoZLR7J3L@MzKfrp&^vgEGW#6aRK8Dn{7vlY`#9zKhyTwu|6_HHf9<(_{{35bNe&|Z
z@cox~s(VFlCaqA$U@OPmdWzD~3FA>6K@#u@uLPXLJ65HPhttSVmR5Wc(I_R#$>x%^
z-qT4%EVQbe<Yzo~2oR=TF6<50Ob-GRrzULbKs0NpaV}7e8-EN5lks8)9)Q3o4ja<Z
zqyfwP*an83N418V4v9eOXHu#qTG(jwQmTPL5S##LFDJf^R2xRwOdJ}@^CJAN5l;Ea
z${4H(C?PzjDojEe`{hRO-%w4TxcR+r<;;UDA<a{s-|KO7G2@*N4>1uuwbAFr{XT<E
z%KIlXvN+;1*N2?UDz0s&98F8Mx-odiColJTG%Yz@l>EfgI~+|`^mWA5Sh3elC>)ID
zOaADsDQDvu&RPmDD!S*kholjlREqE1I%K*kxU#>^OV@W;&0<g!low!gQ8_mAh-EE2
z{>msi&a`yuA|SYVtV4jN3btvMZf2OQ#6!`9GI6IHwUOI{)U+ei=5sSzu~=Ess9~om
z2C3$eTl){x-nV`L3h@)@CYq`-bEzfgZ&ElCBOZiFr0K>g;0_g6H&Zr(qfp}qn1z(5
z%qT7mVqBO6I46p=eyXVR7*|{3)e1H2;}<JBTSJzU6+dxxi|s*|KX~T>`QjXXcAKiV
z#ZNxB$I*Deoe$n7Hjc|LJkRa7-(Y*AOVrz@E>6jc?@=ukDl%AS=<R-nVSj_O<JU=d
zPdVw#`1R*{3?3Ey+SL*MMlR*GF}SKCO(pAaxvIg5#d@qfb?xP*4<MnnioO<Ew8u+(
z;N^X*Yg*J^hf1uf9i>yjtr_Xl!XeO%0v+Plk>Kz~fMzR0h$nb69$|YXvDlP*Ezi+8
zF<q64K^&-}$Vk8SqmRg#kNlYp{YjQ4p?8xpjt(Vh5{2}X$2X*ghocom?b^7jVIE;U
zJT)*hapDGuv}Yc`y~f1CdXFy!nJ>N2OGUV{<`Fz;XRc6M4z6K0WoM(y#j@sm_eOm8
zwc98KU;4t!eCnBPp597{-;a2BzT}y`oJZ4&P8Ly5E0k9Jl~;E;TUfsF_8I^Dwfp@2
zUwDN#?oasLw;u56t2?}Ob(@8CeD(Vec<t~M<H3wa)VAj43(xSm7cTSdcMdt8t%xr!
zC^}u9eQ}px`0O>_xZOpK-sK`}V5JqM7yE3WB?UDH+{t;tFf%M_&%rRJcAkfmijzsn
z)hJ`HA7hk!4`W2Gn=69)3w$jlyw)l>`O053BkZ!a><$u4WYBR;5*so0uoCDp@M79|
zwy+!wV!W5d|8QDShbdB73p4T2#ImX!SB6Q;Mr#aS@m`X2B1Ib0j+P%xN;dMi^@Ql8
z63~g#9PI7!;PxRm_x5;4Pq}%p!@*XU?;lPXJ-Wl<(Fc6yuly2!_vc^acmC`iKm6v`
zc<Sk!Jb&eYcV2&sdiP`8xN(IyzWH_h`MXr?U?^F3UqNlY#3x@k;Nj62RehEBE)u@}
z`pf+F7gGM-=sT1)W|YO4G$K<lU)6Xt-8`l!YfNaDjdpma<d=15WyC~OO9iX71Fvv(
zu-kL8u}h>)J6k!F7f{w{4?4()wZeP(+|nc#&W@3m!znq0P0OBJ9|XpWxol?EEn#XD
zCQSXDg9=ad{ty3T%`<3bXwBw*?2qOD9jp0>qj$bRlu2x|Qo>Dk&T$e(42=n>*3v6J
z9G4gaRT8l%YbHuSk9n*}8pabs+^N0r51Mg_wVtK*A(<t#8)3t*DvLH^D%pUY3UJ=B
znxAw3tv}{C>o7T8a4<RIi<`$d@7bSJ9KU@?ksFdlMLsDxez%W4UT``~7@aS1X+qtL
zsGUL|kN7Eb%K3W}_H50k+!_0e0Y~p8^wWqN#R;2cDlD$e0&gvsq6&NW4$s;-dwPZ+
zR-~Pbt<f1L@7yOEjlo|~-2GD?#u1ZJ+;|(i?dDO11$Hs7m?aUhfr$>{pxSdjFR{5{
z(1}nerYlD$S(~D01q#Ki6t9Ji;M@zas4QtPX(_Ge-MbOX(ILmTEwhUwtg{p$NMu=B
zN-I^tA`;?JYGAaK?^~2s<hhY%Q9Am$;pkX$5XXt2&mj0uV?+1&`WnqEmXYDigv^3+
zMlV|N`Hg#QrwjH6XKc);T<=G0jgQbp$`|t~=UZnK8C<Vl<I$UE^e>M2^7V*8|BT`I
zJN(@KDS2u*s74HSE^zrSkfL-SpU+6Tx42TCqxKRicMVlIp1-lj?rz2M6QAYXKf6u$
z@*#KM`vJZz8O>K1FMbKOvSg7F7iH&z2cX6~E1nP@LGYcB;+hw1D|ydi1>VZ^tS&6x
zIz~oDhhApoG9Gr8s#-8Qy$i_C5~Xe+=`}db<3yey2CDQLuc^X8VH-fu#Hio;y|4JP
zmgG@sV#M!EBBV8|QW#~W^K^}8bpHozJ@qk?w1WoD7LwZ2P<E3@2hoS)VrALuMKTwM
zU?C<i043#qX&s9I`qeH_Shc{=XG_aQrdv+}kDOfA`!<oNCP*j4NTakK<sIV(uW{+>
zOPs7CHgY9`pMZQ@7o2l~zhD}PO;sG}_6@DV2O}#~v5|?9rdQ(22SUcETm#%Sr)R)Q
zyl(+c6=v52G~5AovEu07n_Rl_$)LiN*H}3p{(J4mSUy&k`>-yWqNh1zA3wwEFp^D-
z#^N!^O&$f|&11Np#HneBklX6`?prL)klnolG7YE@?H*c%W>WKb(e*e8`*^^tcOKWU
z-~xw0v-XM*wVUt&nsp%#QHI5Q!t~@8SD*c~@CEeRW~+8Ws*ygq$=|c>NVSnWd*?*P
ztK@59qXTgvJa6x)D&aarM&kpSS}7~;L`qXE#+-cPcW}{!@A+qV<>u$OKUxxbxOr*7
z>mQzQa&!XLl6MD`(=pes>~nlN#W}~$=1_2eSnz}6voTlpwy0{$XuM#5tA}=uZWgnd
zr!0z!w{G3T#2wN!p{g9c1h(~*&7CVW^6#?tLBuCxLGLAt!ytceR3Lp>FJNIfI38a_
zNt_a0dhut-V#TtOS!^1qK+BX9WSASukkpeH#p@UM*nalQLXgmkZa4tE^Q<f(#u^GI
z1RiG{2g9UYAo4*dKrt&Uv%&^CvHV$Me_cAb<V1zmP+c3aQMmE4+^2<QQHg}Is2#b7
zeyW+CKj899KSN#_F7+Z=<e{PVUizt7A-&ZwHC!xf`l;N8(^3G)EI1G;C3wShWf`Qh
z*?&ye^pS^yC)y#9HfmlvbX_vKIOfKyKijGRUD%Zu3o-oI$^~5?uN;F^co&h8K$VqU
zM|?HUM|c>LtXx)BVz9$VN`AJdvVte1MwEii$$Q<66_r;kW>b>hO`d=G<E=G(X+7OU
zFq$+Fe=7TAtJb_W#o7~~d9OhZ=0)XNeMu2GBr$7Vy1a&vW^{5+y_#_4<6joa`D0+u
zw3I;6AJ@9BpNrgdjTokIB5i*Bp_Hsa3-l8Gk<UsQ&Do<vK6vxD+5O(TtPVE$Uq9?}
zehh<5@fZ46`N3(${7kX;)IL`Zwz>W2O->%<loN*`;nWVGEJ*T%JkL3QSW<U)scXyh
zbVS|VrdpJ2n=ai;LvGz2Gmj&-c7|kWOzs^;Q4#qM8SLE<f!E41Fp9Fa3}PYQ6}6*Y
zRP<tzdpCf7L((IFy1vOKXmGSh1v}<M(=jHfmBfv9Q96tdbJ;M~h=EL77KKDt=aFJo
zIPpr4G}DDhL7f-s=&Ei24&TO9TPtoc$CHwS4UyMgEG1_8Y+jSaDws=1nY1Xy7w;r&
zjQw83s`MPsD!PebCpVnUYqS>&Ypo#Gp1P{p@1@ME5Ut$=SNP!M3K$P%Ef(B$?IHNZ
zDMU$U1I{q5WD|(S+C70gXhmCHw2O!J^rVs4Y866209%VX5a&_ObH0$?Y?z5Hc2P?T
zvsa3ap$+@cq5en|Ojm1@&&V{aN)at3N{n6-<FQI!TkR!`GY+h&st$o6)-kost!&xv
zY3j;it*2s5IMLo?t%!77V2Kv>`n_lrmd?{M1HBWvaFa(6hu7AP#Va4Jm@aB&rDMMz
z369}BiLP-@gk%jCVMFC;LjBUPNE_3d$D`~)4&P<v0)IhKRN?EDtnu1;^3cxMpq_*f
zb>}=;;&IiIV|SC^ytKi;xw^%tlYh+nFZI}sQ*PfaIef3?<3I6l^Y8uK^K>HkEgWxs
z=eOCt@(Nk9#lyE>WB=y!EM_YvCr9Kv8=Ty|kG3nG{p82_@cXZkq#Z8qjd}08QrUd&
z3!mf@H+Cd<0MBe$GP^IhR~i{?EIEy_HY`d@q+q%fgIQ6{h(lBrs=LQq%I5G^2%xV8
zaP+~DLf)7tl%rx6Dm7mksWzd<D>1E{t;FEqWKuF(3Gj0~EqV8NA$;RNZ9JS*6t#Sh
z$_i$1G%dMzu@b_B604|H?HDg>kpi1W#2_ymC9>k8wA5an@9PidOqX@|S&FT06e7!|
zG;D5eyoiscfxiy4j6~<fbHF1uRToQ3ZDn(%LFj~Em%e3C-pvZjGPDVe&VC*GB*dy_
z!s3yQFUmRHmh+|aJQ$bED$nt>!iVahDb+0^{Vl_jo|jHG0rK7$EtOE?tzdL`USYnj
zl)TQOauneJhIKS_6*wD82^iPBMu$*evq*8as9Q$j>Aa@)o^jz?eZLERs7G3>RWOU4
zEG_dwL`<`g31GBhRS9;1f;`fUmzH@c7s{#@YOqu`;ZU(_>|jw#MfkW%wX`lI4a=d^
zyryYk#izphR$=OpRHVl4zO=$z6di_0!>#vi)<J3E(PY83{ToEPf16vEukgux5BLvi
z$AA3EEBO3~(rrPOqmAa$Fy+$T0B1JY-r8ok(W6!wS1)hT+Z<rhjLplJsHznt9WGzm
zptIG(>l%+^KEBU<ame<DjM$ru#562I8+tHpga{;*e5qcO5Wm-uf~=BCww7Iw_44mM
zQbA8kM};DK>(PRnJDEsO33~fLsf_~ZzVb5XY)}kqJ=zKCpBTkPCt?v6=XN)y_gF;K
zybz*Gqd*fsc<7~uVIHxnJ-KO4DZvpIm8S|6$#LmIi4TLk8Ewg}_sW$Xo4HU~vshDE
zS%i^!%*U{Hl(l2K6SdOXCf_brPkJ6W597Hw+;mgTv~<i0$8Ikoi-fCSy(~aw!p-3%
z7q7A|s3CPQ>=2ffR_jq_Q8{+Iio!K+Euw;Jcv#J1A1$m@VSyB+ye#C0lakuS>~<sO
zm2mpIsiB)1&KI?~{V3^qm$iKVm6y+JUO9S+0CYRSKR}X$*8-76;RqpClGqC}nhX^b
zT1Y*S7W=Y!>Db5&RtcFW4tPqU!6+GFj+d6vqDCQ#LJ6E%&%aCS>BPbpxR{HRaTICh
zs|uy$(2PRdT*FBq;0cX>!g`N(s#TC$C+lmxvaR2uR*gzk$Ttm{uCg8zJpIx@Q=Tkq
zauZ>Uq8l5ICQIHunUHT5{71k1lf3f}|B&Tu%2|<OSB9h05qC%XT-i+N<Ox1G;Dh^9
z^1Mc69UhKLri&7t=8VU4tWK#a&#l7=vxTPD-@*Hgv`$gZv8dpzaCEfLj1!~rTH0hC
z7K;ij!ewFU<*`V7jRI#;T2WVyIJB^7xVWtkEqJ8R%3$#5{b529iOp0u7+yp{t~n2?
zt;Zw3J#wOoLyVLeC-aI2qZJpkinCx?*%&@5B|Z5d$Vr_O5A@S%$smhD`6i3;Y*w|P
zmI)wV)5B^7w~iN-wUoHxj#DvSSt<Rz+~AQGhLZ}n{G#?e8Wl3C((Bu*Q6Zi$#1cBz
zf+xh04oSLlxNUY!Tf)ai=G3doaj~d^aYv9r8%=EF&``dSx;7TwmRO5bS7qfSmw-vN
zj!O^kRX5DYyNQ6QjWDE_8n(N!92{Y-&sH_FRgLnV-F_k}w9wO!7B#0cv4|coE4&vE
zzO8P=$)sded*-F3wrkBr8vNQOOEIpDwBm4Fkj0u|C!!On)^lK1xc0f4(B_m@DupH;
zX}q-bGWm*C<#=#WaIvgeln#v#hNF(dQNc3UzAY=EJ&QY$3Q=O>AK{!N-_#8X9*+4z
zYT9Hajf@yql@r5SwH{432DaAA!9NWqD_V&mZmTOnZb+lSXhSD4T;IrfezQls+hymM
zUghZ4317;VY~MWO;_Q^tO2sysmsG`VoGGcjqAF)hXEX4MV!32hR@CJM&MoD66btHd
zLQzgB{SnqIaY_tgoDr+=#`>d1Q6q<fSF%UF5>{?QigQ{npiU~LyOk5a4zFZA$B|}Q
zI96dDMhE@4Eiq@MW01;+5bKcM5+(wz2_@3^#hHL)Mk~&z;uW_Y7GZ1@^U~5u#W`hO
zN-s9ZgvRTwr;|pE<`sK`ggn;4AR>}L%%Jd#qlWeiy(FTSMNER<eLpkk>0@dswe}%w
zry))N#O*=CgK>fOaC0YX4Mjo7OGWK4v1YFq2ge%7v=EP)oV?I}N|K9`$nsvIQ4LE8
zS(yD`&Fzq?$tk&wnpCBE>Db5(DmCDRvtL-xAdlG0r5b1`)mf}Lo7W_{p|UbTu%Vjh
z#R}DzAkxx$bmi!!hIuIta)Nglx{2ltnZp;A!&=Y5P<-P3N&-fgwQI{>6KS@4LMU1W
zYH>Fa3GW~m+wh_euOS?=wsH(Ip&54*c|FTYFpJ63;;d&sgo;cHi41R8o;$smRqZ&R
zR~&4lA}WeBbQ7lVm;w^KR{`HqZ7iI^Q&=Y;rWMVCN3!|0x?)}`(r3|;)ZUXRAyGA~
z$|P2#S`MjMVaXHW=`SmfvzCoc$}fESI=}z5Ilp%IdmLT#`O@=4{+*MWYtJtDFYn$Z
zi4C1B;^Fy>!rx)1bDb>jaApljw}<vHDdr4!Hd)+vcmoG}eNI16l;x7;>O&Ux5LH3!
zzepCGf}K;s`S)ugQ0W~-<w;{hZR=n>BA=bI4ou@RZ0JX{#6A;<NvJ@*H;U+VA^v__
zok-C8z&JZ!RMbx7sd6|7c(kfLlT}4O6HDvbdY<3!k_8>~&M;<C`EdIv?q95UZm&Zx
z)p!?f8SmKaM5K`xk&6~p*~z@(@-U&O9OI=Uk7beNv2OdeES3W*jTF!AWz<%lV^K>w
z&Uk6*#wsKo38gsJiW^&5n`onzOg!<zS%`xk-3Mb6qh!2R1gF|gB0Wm&A=QdBlMiGT
zau4w`Miu`$TC41(lfjY+%sT<G6r*JgK}l8tvw7WyaYR~iJgpdZBC24=UAT~ykc*j0
zQ9FuIF?51t_;^;a)s3hfj20z(g9Ph5N7IVUJQ8tIZr1KRUPegMP{j=r&9oAS+sa9G
zaIqB9&%6}>0Gmf-#<#NNt&R~1ayYcQKv9dDu%F7gc)qC7MsYf;<l-=j$+Biv7>S81
zEqmRV!A{1qavV)7#*3QC#gv{F?53FxbYm%jm$k&ETi3J(>kFXrLSNSnc&sIKCJK1a
ztdxm@#S;gdL1OC&LAOS#n^dZgekbPL`=|V~@BRRPuTN?I67|yaFVAkX^OGZf?#*NV
z-sT_iweM{3t@ppk_rLchcBZ-a{twCe29)N}{fD45k%xyL-ou#zj+F1caSs=7<CX<m
zSAL3(<&1JU;fMDo+?{%QY0Tb6igto~DWy4@*K{IHjU0q&YKV-+I#FM)f+Ms-s&EA?
zD@$w?(^5=kBz(rByeGPUQLxpGIh$9~TUpQLVa!%1=6F_dGM5h|F$(X*L^3go?QTqj
z^k`Q%63!PjivV&BjV#h-WqILJN08?BRu&S)W}@jQnn%-$*~)U~ydaN5nW7E7)U>?y
zex|WbO03P?Ft5aBs+VdO-f=ds#nHAWW3h{s0}XKiv4o2d4j`lOI(P>(6OTx>Z(K7e
zYV68mE#+iC)6}JuQnB(Z0<oi)h^13e_@LlYWT7uLVUsqgN%KlJz_NlWeE-BKiVB_@
z%1xMPDfgZ8404HEyI5K>BcnpCU|a-BZRiP|_iT6NFz%;9#~v?hv=*Lzvk)SsNE2y0
z0Cu~PjMQQ=Stal(?wzm5_QXU)%p?`ty_jyST7Ul}lKh74T#F_?l+c}2COEy|0Nhj_
z-L7t}<5uC&iM3+kAx{N2Sa^aFszGHjWe_we1ZP>44y`33ek;(P1)!Bs8Kae;;6ap>
zDtW*1M038dbmH*+z4XwXSblHD%GL?)VjXk>H^CUgY;;a}+~reW_+_r_4tVcq#JdmA
z_}?F#Q(XC7e&OABxbOaufBN2-`0PGe|3y}-6|qumKK(S>d-6D9W0-U2?K?dG>6e);
zDki5Bu04GjT+P8&muGi7+&^3L#+x57I$wZw>|fm>j}=$FBaRgtNsM<6ZKZXIB8`bW
zi$WUoUS{Nd@{k!ZEy;rubZs<=lGaIkq4-CedBlyal#^LSVLj7T&7u?0jpbDR)~ylW
z{_u<&yFH%Y?NFAEt>9>@w8C2f>|!Gn)25dlWQHH!TXOj^4nXq|W6E&TD)`=mIj>yq
zN_@0g8&jUnWz3ftjWW{Xc`qN%cx8hdjd04l!9$>|JymU6kxMU;o5s8K&zDv~%Eve?
zUK5_UR-#Q<)&dxAHj-z@dJ2NXsjE<$;b~{`jW>QH)KHJT?mU)R^TIn~EemmwXnKkC
zEm;snMP`kmXan0_)4D17aI)hX#U>$kO+l>NIIzaeSJSBd2%}0*7OUXBAkQ&Y0xI+p
zO%f~O#4t=7wx5`^NC2y9Cm79y#Ch6;=;nZL9!4qp3G@>aphcM;m{ca!v4L(Hj8stc
z6OD=F0G#=ZyLaT)_sE*af^BstF_K@vS{^QPWogm0R%Ch&1lQKnjm5-M!DHDG^0kc{
zcEexjBs!3<<g?W#rgLz>zxx+%^1?xv4^H-&uPg_{g#YaiFYynGf5~6H^RM{(d#C(^
zD?h>KKk;S0`~D;Lh8=$P)vLUF@0{zqefp`Plg9kzSN6Djvf$Q(4xfK!6OeX#ziZeV
zB;2~)VRCj#*2y^h-~rceUTTc|f&r@`(XOp>ah@cWOIIYrK44yo$e7@1DF>blV;&h@
zgMQ&s&S+(+94vx;x3$s}mX#PAY~~SHHgfVrQ&x3|e-{6~Rpo*WmGA``|Fg1QlZbx2
zZMUG7gDfI5ipk0bUbs9eCntSm<REIm`o`g;m+H3vFRW)#3!}>`MK=}MY7|Cg&8Vtf
zq^KZyUT@DhjNY0*NsPGUG!#}?3zzEPl;ix9uj@x;$xr^Dn69jWKHnae&Bpd&Av6GD
z)9)4m;F9kdMB3qNn%C24>>HoP!pg`j(khGu15Y80;vVDgV8iz?TD<tuSW~0I%>1M7
z(4qrq+--jJRetm#(*|jmn*2cP<$!K7Cm^sIZ9^6I?X;kMpIb&~qf1%;48au$<*z(P
z=S2C2Z^Uiu&*}f~c{g7#4@VToT;J+ZS2dARytv=t#mimxh6%&oHU8+{K6jqKf;k-X
z>Et0*d_iR=xT+wE3`K33t!lwXDj9b=CGBYxOBL43A}TLj8(e``8l7dty&km#>m9~O
zwG~H-(udDAQ0cXi$|Z~l+F%mGWk>kzwN~OM5T*&y#Zu<?v6g-^jx<NpicVsPjpAdM
zI(+h}9+Oo~H#Nk=j3!!eQdl023gYdYZXDh&FAlv9*ytuKD@PtjKYn3YL1LfUNa<yU
zPT>1TMoC<CxY0`M+W^xjEJ)$dNZ&nP)>L5xIV&u6?E;8n(MCp5gG@iM*cvM=twI^-
zM2ytT!<(O5S&uRzgKj<Tn@#0GDUbiiRu4bpM-CaZ!q&2|WW@LQp(H4GGrL~|Vw5-X
z+{)Uc!&N0&j_Qdhu7)4b(3gu^=Kodg>7<%Ym`OS-NV)?SwPP4SJELY?a&!OyAOJ~3
zK~$iLdO4j|?3PY|F)tET(Bz3!QWD<c!-2MbeLmnM(PNcK>DHTRQph3SNp#z4)z)LR
zB2v<7%qveC06iBbHno!jOEo2ZaG`QmeqJ0ZyvJFI&GYd{)?pew`F9OzNpcL=4%prY
z?Md?lm8L2deE-eY`R`X1pSZD4trW9mg|;ic{p~l=^N0M#@$+1J@fH5l;v0PB)_=$0
z-IV+Dm$`HM2%{Bm<T?K;(kzaS7>#CB%Yw=InB}6REGwQq7;^JcpIdj&SX74L{uX&B
zXRDu*Qo~XXjKpXbP69<5RbS;5b?qp_tgmvOULJ?DKpcvc7Q&a(D*V4(cTr;z)BrlQ
z_0(R}UUyDc+}O>7y|aWiOtdVr+R7N`@@B^QqDFNhA+9WB^i)~Tb}yFRw{)$%w|OWe
z-@dFImp4S$o&?%y9m?U|z9_=xO93L&STina`njOpXS14NC*s9RIg3iFk*$uQs1&EN
zv(ZX!lU3k+Hy?Ckb*)qb{=mHUYX)K{xdk+poIDfDVwr?dS2xyeo+8jX;_6UpTPvHk
z*&_2IL~w;gjhBU9c@fi?KyS@sDf6a<Wuv1Rrh>fR8&zy{B08}UswS%jr;vkdR#-0e
zA|B1^b`!>typF<ps!~WxXY)#=!%9&rp%p8w$c@l-_qLNjx%Q0a6`fdOyBCGzK|C(z
zsok2Gqm&}T2Qm(UOeWZhRzfkiUMS*O0Q9^9Wj*84(FuGC1<Z@O<*c{dfdB++<gl==
zF);DLOjNb3K<^R5Q}5ctC<&j03+vbDH7&Y1xLj!WqHt$fYdJW$%x-nS@%#URuinbh
z$#Xb+h|}-kS1Xd<C;9ZtKh4giP5y@;zRAzud6%Erc#hM{hLh1Z$v3`$ot-nAo)UL=
zx%bX@Q0|PUU;K-lJOI8xLBD#PVt&Es?tnMrh+_GO&wl!oJag?C(Wz)qwq|iBVvt9H
zyj0W8B33IIVb4lQnjSVji#pK79T;420v3D<@2LS5_yEzev~*L$z4Jm!Fyv+fLN!nd
zf}9#orUmoB<XU;hUN7QsTCv*`lFgly75jsPG}0W;Dn?83ITQKl6Z;>JVn2j$JRHfS
zbGIL3i~w>?h{3Fo@zW|0je41p8_C1-dl}P8Ofa)plNsI8Z_7aTXaYxbBezHMG{X7N
z9P+FJh*tr1<?tc=VO3ZnC7fg%7;S^hd_>&_YrqPmVrW_j6G~}ujgfz$1KCO|g>oXi
zN+TKjEgqw#&Pz*GIl7tQ$}q+&#dvA4UUdF>Fwb065=XTPY`KkI#NkB&-Vo_X0w<KG
ztfdq`9G4sn5{_pzJKc!f(Aq}fVO2`v@J=V<;iTkj4%fF+QC_wiTYtolK@M4IsY)v%
zBCTMwax99PGzfLZE5~kEb2_PbdOH<Yph8A_<HFHRqyLYyH;c6_JI}+uHSFok=Z-aY
zRaZ~d6KRn)DKRl+(V_!OiXcQ`EZMN*2RR0kAOVuMJO%+Gzz@lb5C;O1AV?e=h7|-#
zEL)<iNHR%@q(wHH%^s?|Yv{UFb*FRhnfA1%JpBLK=T@^R2{{0X>Ke{HXYak%`v33$
z2Ar$S3}1py`?%41-WXCH#Xi8YC9OuLnDxe*4&x|T<<gM3Dd5;SMd7xKhDcHnRNW<t
zU5-I5=Hcl%mQNI3c;&C*)3>hSn|Dr7o6ESd|5JGP&2M9NG{+yj^AXa6IqL5v_-OC5
z_|Ff&gX>>t@!vgj4R3z)5QDrxFW<-f^cc5Z|BJBm6Fj_kfNL+j1Yr%H*`4CX{s;%h
zbG-N70S=ajc=g5<+k+IsdU&Rm3rTU~N&;D?phS3Tlqx*6s1k%<Ng>gbL|9u^hHxIf
z2LPaH?x~O_62qK$^rvf!UMevXgoM_{qsSyiF2$(;*2Y3A;9y~3ERVAsQFRjrhX$$a
z2tl>=Plb;dW)YrO4)ew%Pl;@!6~#zfN$CCcaRPwiw)>exZ6ae!!peY^p%bFD6e?LH
zL_MCd8;*cL**F|79A4bZ5z&Z-jVvOwOd9$QiDD~sA-mC|j|he*3a}I`M;0uku;@mV
z7yT41KcwCRI-y0=&lMcAB15>qUEu4NC1lunaMq&a%18;o2->+Us7**Dkh5PyY%@}|
z5&J|nW}Ye}ic4VbcchRV7)E`3(rFYde~+`$Vk|^wGO|C)(E8|YPtd0GfW>B`{`e-~
zd|;V2O^ODofP}*R(;8RC8467j<AQ}=09$4ABw)UD2tjnM(T3<pr+;fb25JM00t9-w
zfQ}F}$mn6Wp4g8UKOaEtlk=n#h#Og88uPR#x^Nbu6m;6dM<+G%bcto%005pGmnhXf
z20OcW_U3h58yEPUyFZ3=a{&1>F9F{>#O=WmzE<`S#Ix{ijUY!@o-fgs0U}SZn4Lg~
zJ`R%#I!W;UgE^Ga7!Gq}nI`uv>(Q4ap^qn+2jGH(V^-v{r3qY`2sG9qlL96XV>d7`
zn5GJCp!=Bf=BoF2_b4=t<sg7{Bha)STRi$PDf#pF6GGebM4|MX9_`?)#W1H9;$&qo
z%oI*ai%X*vW$Q7>h}9OAd@&==SdS0_w%IJgd0<vs3^EeL1W@U+s2#;7NrA(q!TwM~
zOF_NrFhRd3akMmecvj=uR*qpVq4@zG4084<Bw1vpBrm|w`TMV1k8!R#Hx(NQw2~5C
z5^-s$Kkua!-Vje{q6Yu)A^>Pn=Za3R0MR+O^b=47I)JiuShMEdG8t!kkYMQ9sy$Gt
zmTQ5PAv^AAUn7k(M&3}33muhpf0*z>pxnVhAZoDlM0PPgV5>(PNIwxcT05d8GgWt3
zXlxY{hfBwbI$)rqtvkBS`r{yIxA}4TJxoZDlqjMcLkL`+WE(9qDDCL+l4H^|@!T8Z
zQMM#16F_HL(s0?HClaZYxCk#RDKO4NcW)}Tn6sWXcOR8#{Azh}B^eO(ywNB@cJkH-
zs6b@0-RThfxQgZ7H}Kx^M@Z5Eye+Zt3a#}hZa#-Ezj_6?_6PXJyD!2$e28DYdkru4
zF5|W3H}T7t=D4%7#MQpXcRx@#K6nGlCg=@c#^K3B$k1Z`<QhIIrbxWSGq*0`)f-z7
z!DCctl)|$+Q-Bln{2EKA{IYZqptUsQ`F;wj3>>o)*+iXy?jX>5*NNA(Wk!^-9<z#B
zP=w{<ys}V2<7j1&X)<}NtcMl?I|B_7fTOj=rGchCt0lm%;NI9;3VfV${yV^ZmLNvc
zxDXiTbPFps#M~PtBtuOko-8dgC9u^aq%_e2m&OS)rqJ$;67*7uRZTu<`=bnr6394S
zY1&T^w7{sxDZAVP6l`hkJUT7^S?MtCDJUf{5LERL&6gfRa<nA?u9r$Ao;eNjX%*nb
zSEdEKu!0|E?oG``-kK*0UJxBvYcli@h_1jC7&)&TDKaInH%j0<;VQjELZ>vwv@W3H
z`0!atY`g8AM%f8*=sjyoB&=mkqL^sow%w<gFU^$EwF%f8Bxv~LC*|DJAZo5ZdE?Up
z7&Gg$Y8{-GnC23-3s8dCcg6=y3Wd`(r4q*xpA7-aI$&I6s6b+$L`irAZg=QeQBojf
z=O-VC@~uF4j#dK7(lw4F#t4$eR~lWDH8CE0dV>!^5dA&=XYB-z4iAy-Kaaou+D+U!
zS)*K+*xwrB+wa^%V=dnN=nQvHR<P%D^ltnVX7_)9f3Yf2UAlq)c=0xV=9LlBSFWPJ
z^*+A(d#^#Lfc5M<7*B6Qhy+)jxq%xuZ(_Ar;o%1lad4%@Bv)8AEv6YqlWPUIfGpMU
zp2TB=UII9Wq^D7{J9KF%foYu42siLFq-+VJ$x;mw2nf`T_2?<GjXqghVn5{y#uM>E
za}N(e39=7CV<Lc9UafI(&?G=HwBWaU8rFJ@2T8ZPUNjg>n1EwEJ+uX%T=Uv>8hQ`f
z=_4vWWVE_$99&8h1LJ{{r3IwGw4i>ty^s-TZGcL+)Cb`L(UrKiR5G@XJOr~u;*1v#
zuqmxwpzsbFk#VG!@m(()OZ`lV&a@Grkzeb<J_;1e9KT!Z$mWb%5CRKNjK%&SL2DcY
z_twWNgDjCWw{J*poiGU7Pw6CRcrg$KSRm6xq5|0e%@GHns!6FgE;No7ZI>5!va%Q!
zn*HZUsH7b`JOg-i-eOW{Ona)U;*2L*aX;f$@4~%}1fRG$78&6;CHqY@OxIOn6GW56
z`xd(C)P-@Z<o1K^Q@2J#Kxt`H&6qm9Y(2IMMFyRSj7<8T5s!&eAAk-6!jl4O3k-6a
zlK22Nz#%_|Tnp^)@8R(92*3aRyIs2kYk}5aI-TN`>s#0vr1-{r`zW7$h_pArR(2JK
z`5etpzmDYB-@r%}_}<wPM%9phg5ITn3&yojLgRR0kj|E9mTO$Tbsd+t3S1o~Sk(pt
z1<an%_x3_UiGY5QU{M+bprh4G5U9G<Pf!_$wWUc(CTT-UIEO%iLT(TOWWnT{)&=Yh
zxU^nV>Pu_L5=vbd#A^^R<vcJcH1<X*6fbHisiZTWLL@x!iKY!cbQWevx$h%WNE*vf
z*-e6)l>55UlK6<C$H8Wlj{NcYXHfP$FMdz5&_YY(T0FHeMA~PeXxF!aPisXNFx64#
zz~P<z?<jGGHnZr$6BTguQ1ZhGLe{m|2xZ6(8Bj`MS}3Fmm(zmqjcKuo_VkW85ViH)
zNh_qlDCNy1_NE{{Hw2w%gIqF30oHFsjsdkU6Sno9^eH82tEaIu(8v`%D2vLWY6!kh
z8SS4I8krWDt!>8#u%6g;$tM&5-9!M<6*_df^xDyPA&D!ARuV}kaLyBu0J<KD-T=0q
z+acpaXQHX3*jxue&q*(pRPsu~0df_8O_f4SOpY~~1dWz^sYEZiKo;XxkKP-o>Xv2G
z0k&<>FA7KiPv%PqB{9-$g`R6@+d_B)<t?0VG0-iN-2(mp=yN!@bAZ3u6c|qLVO{?T
zs_HG2^<6Aa?*i>P8XK^jFVL=+ShpVY*24>lgYyd32UtOCp|)h<KJ6!jRe|t<BBcOR
zFMvE%$P=obvxKt;xbHXMlqM-5XKRb&rJ;FjAb_mL3aLC*Fw9x^AmrEhfOSp%*|Kps
zTU$8Lj(NOnQ6yw<ePx<qURf+_%ejMe(F4$-w+O(=vL!!(iZ+$j61Z2j9*f#x**He?
z2}_vt6jm+u>3s&dj+ZUwwL@vB*N-!PAxLju85%35TxMp93`v%)XH=ZL0um*n12huG
z3dlc=PPRFRf7;lO^2rk;wBktxzEO=Mtb?-hs0?qdM`=9)T-;(9hwfyp_<c!<Na{Zq
zsBNj3N_$>8JXw;3widied0N0FtJu~pc`KByN8LIU2|eTybUaxZBwC_%z_Rw;;?@#4
zFAZ9cPU0OMnzW$BlWUp^3Gp;h>LOvNb~IAdf<V2~wMA81lns5xM`sOM<FVusmX@?x
zAW&6uUfPZ)QE+))TQ;+FRKbBUoS3o|H-X?OUCMfL&so&u8WowC6idLLHIT9~P8e(;
z&w#R_8y^6yss?BGA3$1zD_5?fwFa#**xK4g@E#xDxsP`rpW&@XODv905abkPqtF_I
z`-fBf4}bU`{*!NhAOCt=<JtR5{PLd1XTK;R``>{p<`9D#=;cWB48`^?PL>Vcyf?$?
z+R|cou`H6%i9(){;pn{a$TE#oOBAU><>{GgJzY!dDRxXsYRQ*o6U8Q&B&2tT30g~>
z?b>)u3exribV0ZJnrNrIYv&1hur-!sn|p%<56)U_7Yes_3Y?Y}%f{i-D8*J!Vo^E%
zP(<8GC7rI1&kY_gTD-DP<_8rMSC;JDGcCyAAWD;^Bu}%C76gK23OHO6RrtA`9QV#z
ztmh4G?d0%*PRaAiVP~p3%J9*`U{oli%y<ie`1#{pAy25~>8I2R6iGa3$Tz_8d#*V9
zAXOarDCmN&ThGKG(Oo|&iOO3Vk0c{SU!s7%mM9yVKvXRSN<LW_TpA|W?h#(%G*t}!
zlmO8sXZaPH=IK%bkIoxF5HL&GdqmoqmE_I1#o%h5usMYD)N-W~w{~-GWdI-OWJna1
zlQJW4D_;Leq0u^vMY9Pm5s344q7cWlUx7fQ*_t*T=Se>?$Tjwc3C>H0ERophlhbXQ
zP!@v{B*)IQz}_%{6)DafRdJC_G|pr<nu{Bb2Wj6UO}ldt;``-!f)p#ZI6-JVy6Q51
zSBa<x5OGF}rg6BsH^q<de+J+Aqi^B+-}nPS7Z7fRM|b}SmEOf){M_g9cfN2FS0@>c
z&OU?R`>lV9t1o>XLG|(E%{%z+_rCzC3;a|0HO!vd!Y{2$eCo<I{Fm}7wszmc>TC~>
z9v$K6?jORnL;Q_@|8L;4&t1YOm&ioG;Y?sY7f2F`L8@?8TGCu-iAC$Msw@%#j0@7|
z1n1y9iLQbWtO5lFY$+e*>Zt&9q)sjx;?F~HFa%R;7q!PI-GC_82lP_S<_aEnjw<Z*
z35>gYQeo0hus2L`URoe9sFupEgwqU~ZTCqH*m|I<9MAhT3NHPQmxfBSRAG<`G(w;a
zgqAx`B$$3mWzVdl(>qm~LNo>K<dcFhh}uxkyw%fypx!Wsi=;|oZ9?Y<7pc0HrIW7V
z(@BBk(*2BnzEJ3dINnoKT71Qa(Rt2UVDVAZ_<Kgdm85pVg@AEEK&Jo!Xq56wpR=-?
z)(C-Ai0}fvRAH2D7MX`c!5^b+JywERr9>%8<q;CAmXwF*6?q00n(`S*Ye3aZ3M#py
z5Np+X3^Q8z6wMyTP)x~YK~*FN^);Z`ClX3_i|nW&{A9|c9b;(2B7zWI7b%GrUauPn
zK>)2|xo>5OpPzDxFS)<Tm_Q~3jkn?qU-RGfpl2uMT#(BYVKKQD=xNeTkjoMUr~=f|
zc>v3@#dlARkR}2*Zr#GkszD~F(BcY;CgAS*8h?0ahRfp|?>;<5mgkrs9-vWM(CG-O
z{s=+ef+{C?w?4w%*RNyux8B8ndV7kucdy_#-&tU5@(M29_zPGbpWuVXa}0+g+}tS$
ztmARKkTeNu9gRg><KbDY>Kse<m4JgpYb}aQBSAn7AY(1fAfxI)!$V66U8E9C<*=%W
z5M&%J{>BA3@7W`8W62$*m*ZTa*e!79u*CIg3I~Bfp|RZ~MAuk}^U7h6O4x+@_B2wh
z02GNL>1su%?X;KB?#{jO?kIt=0gKuq*9me3I0&Rlb;b_kLgBomUODilVm)COrST|u
z@kHYat*OuJrBuNjE)4dDNf(VZQj(1@Xd@tvJWs8}riCajuo~k;D?VJYj1Hhm2_RrF
z3Q)9+v|p+Kii9G%1&CjtC_?_ZlOO|8vEjHwFn2)$X>ws!Ed->des|S4Y!4Dra0Ya9
ze-{Gwd1BKV`W|PM!!RS`iGHSdI<fJMyZAY_2UHCPUX)sLDb3$6@O2PMAZMV>c`|@W
zwS<(!R<pdhMcL{oQ_wsy(Gtj%!m`m=h?s3i2Z-e+J+@>i_cKxg!LsRPcg?8{N%_Eu
z1SWtsZY5}Qt_+p`J(jBnFs8(df0%2m)+OqrQ#|*=v-tVX-^RB-Ji+5djn}VE@r}3c
z<NbMwcMoQ$n*r`UKEVqweGzvLkB~Mii1AIVW(O#yyFhvyj}9K-zda224?lMmfBlc|
z;luMOcAgvJ>h5j4dgC?BW=p*F_WPKJ9JSxUG}p-CAU!?*S}TnEWKf$Xz_3uA+2^9N
zFou$a(nKL=46ER0v}!4SEE66I3R(o{mMw#N!=yW%CPU%|kd|Cg5)YTbaZGelI?UHC
zp4~0*_F;uyrqE9n-aD$$&uB+_Zm+<)b*K!L%ASEZ?*Zd2-ao4F!hR2<ly>c9<6tap
z6qZZMtfz@X?gENb;do_`DuI5c@b$Y(?DaKnU+UpxWpP$I><l!NP{@_QY~5ly$<caX
zwlEkM3Ym;fD!}z=inSp`GWsF5j)@X%5iSW)>rQ%r`0?GJi>Mw9G=pU}qGmBmB&x(Y
zvk0Nv7#bH)wvO`SK;^1&L>CS|U~Qv3mY9C0YlE$xrhDOuO1;%jpk+K2>HcL}c5!UP
zcc4dq!5D@Tz@$emIvJCcW@`(@1{s(88iPz?SrK<XQ8bZhxlvXk>gOe!$%O71K?qo`
zU03bJ_N&kWi+01G$k<(I)dV~!8#Fdx+Sll%)Q*v9%ckm#ZqluzabVO!G{!@Fnpj*M
zQ%XsjX3budHV}@KD8Z(nGy#yzF(?~O?O{$u!fkHVT0D1UiZ5S&3IFQ%ejj(<`VJs+
z%qxex$qJ^piXZ>X>-fbVyMkA)3^1<@{`imnC7!wV0=(bHhi`lp!%Mdz^Bn%XL6UwE
zTU%epf#1QeUHu9C-12+4{oDX|FMS=q`zQYj=l28t-M{j;@ukndfN>&VT8Gm`iCG|P
zy?!Rq7&i5k0*7l0=RCH08d?+orI18*E{&tS%^-l_ktJl#Dj~2o4x>~;D*#O(w_ig0
z$Um;Oln)@l0vtq#TPmH_mVzSI4UrWzkD@Y7X!E*p%&M~3=_j3-M}QW%Hv#*jl$>t^
zEBk~%BG8#t*gyo75P;*x!htg7Vo!W!l%h6d2GUC<%C#YPS5K5%Ey)NXj?I+dV2jpc
zR48_m<(aP%C>uJ7J=lme0Chv7P{|K1nMwMNNJ3QrF$m+s9qWU@Clw@zGqB!43Ur)$
z<A7CbG0HWGj`&`zJ@srcVmnoI8UcW@#9kYuigf|eyp;q!y-eY>v>4_J175f}`v(ND
zR|C(@m7>p^Xi2J2M~*`w(2QJwM@d?CLL&#!;Q%RDRX)%}!aHo*B)OSFtc}N_B;eAx
zfKg5$;<QJehQ~`Xj#Y3p5#UOPBDksj5ug*j^>l9xle1DKFpwPA&6N__j?uH4X}~*b
z4V9pc171MMjk!r|LkK)(I(eYYXTGZN-J?SklP&ZziNSOq+f|L#>KN@x;Nhbq96h&#
zRF^oIRj|%twp>9bIiLn;R;NhQ9V{yWT^?g?UchgE@BsOq!H<98GIqXqjIchyiMs)r
zHF)db950?Vcp?Qxn(mF(3TdJsm4Gt>y<FpT(I80`5-%|;E&2(iX3=zs<-kI~da}(k
z@jm*1BvA+=KnH;?5?X=`2D}e25*&dH1TVTW8G^vN=}0400>fP5@skEF1U$dr!)#@6
z@1#PJN~9{oDA)Mlq$XF_)}fba3^Rp0M<o&o{MDBx;QNWz(R_KkPhNoq^T|)v28EW$
z656?2M?oUL^I(A&_WK=MEmJh|t1PWT%b^`7Yl9DuOQc$2XOQC3D1jD0?YOiS^bnIM
zNNx-#LEMHM)Dy719U6zOYM~8)`2p-O8~8L*v3Em&RuX0OWCzidTkkokSK`e_Yg`?t
z*dHa#79`Julo9aOh7PEj<Wx~C8jm+0l_)fUdhxXH^H_@%j=;(UtV@!5r-{U>ak!Ez
z)Fz;(2^jkDq~bE2HVP%_^GiwiM&p4uAC=e{B-k4y$Taoj2WK^;B2R-ybD9EWTH>@c
zXbeSuU!7zS046;;C|c<|bCb#jJUFdz=cLB1tsJB6oIzjN@c|k~bniT&#T^Y+JXbtt
z6-jO51{XOc1y>Rc_w=;`&Q~p|ECb!!K_;=+*X(LdZNqL~p>-a2j%tK8nRR^Sg_rTa
z{NuOr<nBjM)8}BFfvUfT!IhuE7hZo3uU#AB=2nKE|Lk@A+IxL`=GIM2hGYEZqbqpn
zrDt*4Nc09N_IAI7J8%ADjQTI*`4@f`Z?66yJo9{x>;L8ezk2%||LlsvOIP>t*_&e+
zW1#|I+6HF=5L!Y?i997{<8+u|S&?bab}ljLYf9zv#5p-$v?w%P$JPeIiyWEBmJLnv
z<f10m)`W>JlB!YU3K*uFp8cYxf6TMt=pFdsNrh$WJ4>-auF=oPhN_oJ)W+d>*=|lO
z>yaz4@(-Ozhfs0XjU*AM?U<KD4_-AM=aub9HxTroj&nt}=8AyEf@Q@C&zu|Uu|G;_
z)MP#GpVrVyBGu$>yJ(0`*|^Xt9i^bkXIdyq(hZyZyF_mM3BV6x+4|0sJZ=J7Z}b7t
zaD~2S0GX=_%qxpwPJV>Zin?V*TrZ(}l4_zg8=hm^K$~@Kd{^GbSa0Wf<j7b?&E;$_
zRZMyk80MPU^Yp!<_d=m%$ABx6&8XIcNMNI!9Qluy28+hyyrSF!1gaYMh9vr0HdILr
zb6W6OB5C@fm|e%$#0T1-5=8**q)#l*NJ$1CPEp9^17i|Q>LE+{i$Ns3LHmGlrZ7w;
zNz6f;yA;gz=RAd8N-<ngiViQLdrhOlg!Ld<qVee)V<i0v+T{_-atYg<p~VrJv&XnL
z&Tx4`K8shz8Iml;{?-t?;{lZJ;nJncxP5&Wqn*pxy}XYkOdyNr@!6N2$9Es4_}|_(
zs6YQQUU+hjpA!z}<_HTD5EPC4k~D!2q`OE{1;uS<mPjZiQDh2Ps*$G}LXwYXswGlQ
znUhLNG)#gDQX)@jBe9f>Mg2f&sni0`l;^d>Y-RA;l|DRZo~@YkZXA_!%i5we32yEb
zXbiOk2-K&XEDfp=**b0a6WrSEJ*B{#^fkWz;i5~rMWC{+Y^i^23`KS2n(U`4L%|Zx
z2P_(5(G4>a5mna1dU`OWOp!@i6lHD6&Kg9Nxv`b7%@vIhM>**qC=MH)V$nsAiP?S>
z$;@M~SRVR~9zK@8N&<x<@l#~O8TJYsq{PLI<1)`ZZnO~}6*O6)3`A1x&1(m#i20Q$
z3eqSvF&<Y9jdN~p<>;rfi(U7iX+_JeLoZ`yqbKUE13K3w&#OrZGUx2)8Y>ZSTv`;B
zdkO#)0G^BX@m!u4S}`GrHycKzEdz+L+Gw4JWj0kr>H%7nM{aBY03ZNKL_t)zr%Max
z175w<N9%mY292BmGAaR7*np_cg^o>HTVjbufHx+8H6GAmvuV&_LDQB*5#U(19&_)|
z*X*(s2w$1k4#QmG#x%w0I>9?lgWvwzJ(OjQ{&)hHNg%lbQC@}fHGc5u46j}3V}Fw2
zt8YC<efj|&9$v-4`2clW<Nb$6$g&I%4<5q?i3St6#T@TFc#QM=^z8kg56-cEeiyfY
z@Bzy5Pw)@d+xW3(uHyM=f#qV3+6O#Zv`7<8*%dax@rj*kg?ZyJ?rGE(=t=(mnz98l
z5t!H%AcIHa$opfMDQFVG04HnmP|uie-5MIp{mDlw<eFG|gIuGrU`JPp+xrEq2X5~d
z-5mYuB;!y6g@@-Ywg(ArZe^X5PkgA5N&Ld=Tc~Wn{aJ&4rm@vicww)A5@bDPJUt{Y
z?HA-9e*t|r9{AGjDU1tPHZ-Oh<mB)h7%bWyWgV5dGm8kEUC)+W0AT<%kJ@M+9-gA_
zrb>3mek7?%MQp`{H-t&9coMJyqhgVeaXS#eMx%~0^HM2seVU=M9=k*8?=4f5YfIuB
zfGVZ!zV4*D;4(ko%BaoY2}B|VZ0ozKN=m|1gb)z?d80Lpk^<v`q}7=gcxkUd5CV@q
zRbjOy4*8@|Xqa9pg+QM@|5`_;BzZzB&bt7eZYrK=zn&+YRwHPGYI)Ay7>_29EAuc{
z4E)MYhY(q77hjqI)mTAELDiq&a>4~bI<kz5Q`zW;*g6t|Wy}sFvIv2mLT(I2grw<6
z@U*F$UWOmMb06QX$M}_h@2BzUXMPctC6nCU;ur9LfA;`?`1S{Q^Ug;oq{F%m_{uMQ
z1%L3)G0qlq{OlM10>1VBW8Av3jhA*_#P9sz0Kf7#{{yU6C4S?Z@8WO${4e40yv2)`
z#%P4aKdIluU;fswV{rW~e6^h7AAWy<9sL~s?(6&59u*K$U_3|>tRwT&LP3oZ8VNdz
zIa^skf&4VQ_Y^|P2YV_6wtBRKrvlKm^*C8uJes$V2$=Q~I8PIRmoD}3wfE03$TXha
zEpTw&K=A37=c;qCiQr%?vmglwYMo-s?v2l1+yuZt()&jxj#dWGZUH`*=w%A)md1Lk
z#z6_iK@&jA8$fF>aKkkz1XGn@Z$QMCgR=&eajY;5jM&GYP7hV8$X%9BeEJ@ePv%65
z?lZ)!LVWCDjJ@c^*F}pi%BJ(&OK<)xkwo>?a<fnX5*9^rO=D?XcDynSRx<5Zu-{ui
zM{~aTa|LK|c@r=w$b6&A1>}-hGGozrQoEHVpr4RrS%4PwinRbErYAdgD-J-1bkcvv
z-VGbev3Pyre+RBg8b=!}s4cOe>!Ot;Z$VERB>-faq|fx-@RZ>P{cIlbsm;%{p6KF<
zjQC5G1=GDF@f6vZ(KM07M3O1OO7KKv!x-lc+(Y-rc<#!x_`>ZyJiC+O(ZZl^J)W7S
z_=6A5aoA||XXj|L9BD7drQK`TouwE}9bR~5AJ?z!VO(eoGKKeN7PqfXuxf^Q?#SRX
zw{OA*f&Ec}VXkplW~e^*4SePs-$LCp_<uzSV?GavfZb6_KyL_WCH4AR0>Kf&f3Pyh
zwZwU4Fh~UkTEaQHenM`NzM{Q)5CS@p0@ue0yz@vCiPm@w^8~4uSkxBJ@Aj}gOwn48
zXD|2h!G}~HXAF{s5Y^bClZ&FGjiTF$f7EN**CbNuY0};cTsVoT2ljE-=bOWA8f`X#
zJn|Fn6U&AaXf#m}T{(YovBthR{*DBlZjGfAU0|c0i-sOgFZDmiE5!N#W`Tl#jue}T
zK?s{3)<f_pk-l&A&(HOSZ{Py$=AC8LQ$LdC+e&b_3cqjTp7vu<f;Lg1X`I%XrZT~1
zkZT^H(WE1qls;{&`sceMv1W>*p*TeXe?RX%IuARZ+N4^b`PlcncnvR_lWsl_Nyp<x
z!4!ec;HIC-&iwP@`)Em1A0_bA_lqwd$XjoxpLR|{<BSYu)~&<M-2pBQGo1MpTcZK4
zY>&}P6~1t5hsM=~#@*uzFYeKr`tql);_4)W@xV{s-ov$Njt8d|O(+oX<rlW_Ywrpi
zeQttR-#Np7y)(vFx4#A>KaIvwWwSOOnT!#50f7PUNuQj&MuEo7K}tyKe-Ip*=6L7H
znl+GAS|^&2!O4L9_zeOkDaCfJT8Fbj!-asG+XWg&H%m$m@^Hj8CIwx1>cXXvxH_Rz
zC4B79HV=ouRzD%YMYB6D(2-j#*Rq+4z*_7BSsn>N{`!jMx0$Bim;ew8OQ=l1IOh|R
z*MMR__*~FJgQVjkl>#QvL&S1<Qcp#)&7bmJKK_ax0YCKj12~PDCJRDrMpJ^Xn^y$J
z3L!dvLR`!-;6W+*iJwJ7qF~SWz;x##QBVC{=@>wgfMx40$-IF?l4EejBpr`UOC$b_
z&HR&I@r+h+6AwU-F^Gn-0wIaP7(XLj57PV7W&E=@8wm;3vF0|v?;PPL&T{7B4}Ax+
z$Nt36E+|6Fd5=YHAUnD;u{&J=rh^ne_1aDR#@FA(qA~c)xQ8o~484?sJGc7@j#q}j
zQUJYNLn#U4J*GX%Tp0IL<OzkY>`ZbzJh%AP4=ViqSNHMLZy(}s-FODeQJ)AI0_ZD2
zV|GU@UBzE21z{!R6ij1JF9m{g5JDn2j-0n-C6-D7y*NL`>7s^G66?mod0^E@+*`HS
z9TnJ-hCI^R8n!XW;831Ff$-a~j?!BfmF!C7crD3@L0s_3>!#z~V<9DpHF^NnGY3&<
zXt^oZeIWPL7NnXJ0bQU6Jt%ZS2ODrUw<vNAyEb@qvVt}Z&QBg8>rb$5h#X2iW&AZb
zX=hU=$Dh-N_%MnKpP7KE^hYebKi?~|`2@<<8N98rUYz1$Ftv@n8Vgs}75#q2pCg))
z#!cymyw><<SOzRLuo<&-FB+?*o%9)~1mL02G;2Ve<M`wZsr0BV&5+Xz$LA+GAZ)J5
z$NucXB{bf%OAvvhkv~~+wHFWI)(MpJ#|XB@^6d0uZx%K<(kFcL=BY0#VEiVM+~VV#
zIBiNFf3Nucgj870PcYCPvYz3ihZ)pvA2yL#))tvoF!L46YKG6>9Ac1ZWOjx}k13q4
zYAx=a1#G7ZeOu$vagF0;jSu_)qe8=%4Gy0yxXsJ3QUa&5H7?;Y(%~(9yZ#zJ{j<+w
z{Iz44m)G!i4M$!c@U?*i#d|p$fZ%aDuh7pVs@fqDfaH;W@E-7<(uU)R+k0rOApX%m
z`JY0(w6QdflYvB%K!mKB=pU|5A7Z$583trFk*a{!g1PvlP+HXnS`onIIApRbw}m7(
z-$1)qKi%Zl`w(zmJ0it^7N%vvN!$R851o0$thB@&)0BQXuN}sj1mIDhJwh_Rj8tnL
zpKZMSjiYR|LQ6;yy3NUX-+kAF(^j<Jl)jCl5up74a}@IX1N0f-EtY4G(cij8mSBEk
zE~*siwoRaLhxl)8Aiy%-r`C(7*lz*x#u6(jKIX0CsRaK{X*_zGy!8X9U2`B89lLyj
zyci+x4T0b}UixW3uJyi?K6lPC9o9e_zvrSo;ul(y>alf1&6a|ywipnq1aT#ZEiPb9
z4YNGKaOZmWbA9j^u08iGf5!C|;_3TExQy)H&j-<c_V}`%u1l&UOufcxb&me@GUXXM
zdPY)BQyC$MxJ6YOm|dy5wlSKGT2R}SGSevKbv3UoAgC_~h)dE>1O^(ga7dIwunx<^
z`>2`{XMF*)$e>U_$qa3IOdFWmMP3Y{0F3e!a1O`w3WHt_$NqPdJi)3msHz&yd0gGU
z1o6Emb#M%pX+hvsWr*N$d6IRuNt%1&H@^BmWBTgPV{2y%&Jpi?QfQ=F;K_M|UaqmK
zEl$=Jm&Yl_IidSeHX6Ys?*k4NhI>9zgzfYb;$(wVa+U$SVaD!2nzHm3jl+b&F>5_$
zl||D!^mC1~wZXMXhP&VR-|_tCzJk0*YD4Qn$82+fK&j|doUnJD4{TEnDtqH;9QFQd
zw&VWz1M?#vzyMrqRpZn_Nr98u0`Gt8H}IE!`tJaOBC?3eNx+!(;VHHuKtNf-@n!L|
z#Yq7`3qB?qHIBw(Qf&MP5D1N5w!X7L3rzn`6N$SgH6DKVHz6j^<K}a(5H*+SvAtwt
z{1H6|HB*j(jZsJ7pAkZ2RD%0X!e>h3&`%Xd#1S)dUfM2ZZp^^r@@$5?Z~Sw7=1YIO
z`)(1)j849P)+^>r%8%WH?z3W3K#2=DPo!k`Qia0@cktxl+qnLjpF=7Fry~>JLI4~o
zPN1@sz^e8b77E5V^fDTWw$`JUQBL8y@fhbcfvVYED<D8?s`$=0g8k|^CF2oK&*Ha!
z^B<u8<KISqFUPOz2FtlbQr?1DT*qj#511bMg95^PY;TXz>ksJH;~bfinD$egEh?PP
zE2!Y`=^In@iWE}u&5CR~fPCt<2MJmpjYXb!1gy9fVKRz?^#T1vV%krz)7Q8<$!I(i
zfK@};gPJ+mAy8jte89C)f__H5&oI~M^UPX;Dhi^gGO0%r_+_~Q@Jwt_ioiZ4Db=^O
z983wIG6pAWi?fQ1L%=g(2?81&^S*22$U{Jqz@~_dLmM^^30EMN2g`xq>ree3FRu9C
z5fF<DCUVkgbJ|E&mo?=jdZ6<+5Fk2moQn%Vod5XV;1OY=ncKfK-TYn(Y5{~K@>Zb)
z#u-tgQ$_bvhJgK1LZvwKK;~-;V*^A$XOR~F*Gm=Y%K1BWp97-%u4PNy4bQ7hDsrV!
zJd%`**zXS$j50-`6^PRg&l^hoe(31IlAohL>wjrN-{s={5`0bLdrE|kcY(k`u%0FY
z9;6=-A^KOMqmBz~2wDgv<mzb}t<F{k0-{UM4uS4z**I7iqwHz(E>ejiQ7AOsldVD0
z#h#^-o<-?B)^Kp2-oozP8o#o&ji0!!p!)AX$j7M4JE&HFg!S3GsEvn}G%`+7fnJ{B
z(srLheAgDEQ30tH27?^C!xS>Qw@QI--a9Nb`k7+-t;Dp@@%aMf#L?_}QCsw>N9Y`P
z9xn_Imj;y~<BM2sU8K$tVk>aU^weAnHV%z9IOfOEA~c?MUVs>CtqV9>Sscz=R4q*}
zMukRgiRdxO6(A{OGt&Zx^9Hpco<9j@IPLcWBdsz4W#hX9-S~GRfyFtx$)qa#^Hl(G
zitzDQ2mwlM`dG;$nwSC<rx_y9sb5<X1J#zUb7hDq6(=k%kQgccx-m2sBA-CtNx3~s
zW{szcz<PSm$f6X20Hw?04Euuwqe5Zb&^Yzt6eTk1qB!c4UXhCstGHN&MG{Y;Bfv?|
zRWvq@u05nf*|dhO^N2)h>-<v=Pw|>we7O)hqZ1QEr(*!{R0Ze4X2eR<1Hsnn^yh?`
z5sm${v}kO=S>-x{)PNO~A<%m?hMu9dp?GrRXcRn1$uxJAE1a$^&MS+m@g4cCG68EA
ze1))aN(!FtjWP*-?So@{?XOO8<69^AWn9MaQU;fu!1x-@0#;98ZHw7#jwgpFIGdOF
z;PE*=nw2OWVFI=FSd=Y}SLE0wOB*@twD_W9Zeu(=f1wqR)r=+TY1KNS$@UXGS+-cU
z#Kc?H4n7dOZCx2S>v6JbNy2(z(Fmr?_EL$n(qhp#+&iuD{&58XvL&~k7Hq$p{j=+i
zBL-k&$lP+>I6Rouc<1q|i{4r^<ePtEn!|fwXON(m2r5Z=F(@e@;tW18z$&=3<C8g#
zlGZJy1V>)MMOyTq_PQ8E^6^(R6BXS5uUiseK~U0BEGzSbDyJk?`hhmVK~BODae=x@
ziv^tYBs%g_j8CsEExxnTa@r5^<X1I$K`bhAUs^Ly+`51)m0fS!ZCE1Oy8ZBVxOi;}
zP2gT_JkINlqO)X=z;)~4JxyOy$%#-6wI{9h7pf%z@MzQUhLUo*N|Oa57zJ!L0wp%u
z4P!3=BG^=6LT>2jaWjobiD#TcE8<N^K@{a)sxay4iw47}`Qpzj$Fy^*gvYspu>rHv
zqG3w*(b6DKB>Jhs+K{PE!uO|Q2PP>AJST6?fO%ygk^;hQ<J%{59R0OtvAT1NUl0Yt
zu1AXjY)H^lC-AOBRhC#c7Bb7=q(o~eh~;cv5;r7Gp%aBg;~{&Ryz?VspDN(ttVU@_
zFhHjQUA07SSXC|nhx3+0F@dwv5;e22$dy2zD(nnXj0)W?=8L<u55WD?ny8^#;@ULF
z_8`G&X;Gw#RbC`iX<fj&v3Tdn8V_bQj#dWSeGSD-v|TnKi3vujl3f6Y3uJ$FRyhKt
z9B@`TvZ!LK?C3?8arwMz97S;n`ZFy#c8XU(>nR)Y<EPM{_KFtjKG3N=uWVQTTNiM;
zGH4kvr8zuRWUGeKds99w*M_Xj8XIu3q=g(9qY<D*90NN{AaYG@>7+?DfwU1nXe=@P
z2D!q~+L9NflwCIA;iAQC&6|PeGMKNa3!hvE)m%MkC9vXHam5tp#?gt`@aI`I4)fY!
z(Rv&#ERL3R4YU$iH?~{QhS`G=hl<u?0ULWZ1o~W5@zbA^*CS0ZKAvybJ%pev%&xi!
zY}BI2Uoa{-y{D|q<7JCg<8c49p$$??><-9irbq~*n%52|Yl}7GScwuiUNlIQz}_H1
z*?OETTO>?yd;hq?vZgjQWtn;W4q8#wAiaZHJ%bP4FYw>I{Q<uIQXk2q5`VAP!_RzC
z!ff4z*K|-$_O_7a0*kX5s#S&Ayu`i3Ihw|xsvDf2E%4!!IUdYP$_d=+lL9SMB;2?<
z%`xp!3{k4dq=K^G=pyAQMMd2|t#EU<z`<F?<354=XDw!>!Id%L0E3jU0PwCxWD0Vq
zb%BIMy+m}f*26`MRc(<a3Zp`GWnwQ?-RTi{XMX!}iGHTAJxsc}>y7OkJALAG6W}Df
z>^$Si;3MJklR!mQ+DS6CD75G#m9YwlIEeRDYRA8yd_pM3C%wp558Xp6rNAKD<RzwB
z;Be7&>bySl>tk<fJWWStrDg3u^WzmI;#w9DwH}xhgzlF%bd0%{Tf5Q(EJ_;nj&n{N
zW+iB<1g=do^t8nGAVroaYz<NbPrzp)iJFZIC#(y!U_M?E5%3o}7AOLFG8T#KK*~v>
zaDAGx$cs+SajuXlib2zgvlF?B@;-v9o2p8HDyf_`BU(_k)5Z;gtD#S_FgKoT>T*p%
zDY5@|+>=y>#%9(B6iVX8R>pK<`q@P)v1}ZcwT0xh)5{emg~E1E<N7$kAd_87xiiWj
z1^F-}lJ0qrpXWhNld7YI!J1W!(bHj&5WZ#23dW5Qs@Jf%zl(qLCh*4iWqfIVj<2XA
zy!>hfKb!)WZ{peKZ(w(4g2RUgSgcE&E^6F+vP3>CaC%l^akj(<_h-0wxPatm0}5oc
zU@i4&LXeeLV6-z+<f;Qf7rdVm_kK}1JXu*95w?WR1EdJ+C1L|+vGIr$q654R7Y&~(
zqI1_hSyM<v;J#l;Qqtv#!kEo4$P`1MY#nB6i#?9QUa)~h?7O9;5_*skpWX5t-}%s)
zP8_c-A;3c5y!IR^#$y`Ld!08_vTC&fe=efNKk5}vuV^(TK%t#EBgO)>K%Ecx%#9Hy
zg`%83{`Xb1e_~C<#c)&4z-nZ_#XoBuo$LVd7II#^tA=dXc83Y}he>B=-rB&v4%9Z;
zK$T6#wC+U9?o^Dw7oQnofrbv=F$`8(Pnbob=-NBRF%(ymL-qn#)*iF9#j2);+d7YV
z<KQfvtOBIuq+l2F@vE)>#MUETHRCslM6qad{%MS$<>)Mu1n>qKc`TB|S?rlfq9IGJ
zOyi!$_a3cra8~1NZAmj>Har8;NCezJt+6sOPl5b0lpymG0TLUv?CH=l(d@(H67$;P
zc-7Lq3_r{^QRx&i-Ga2c@TWKMf!xP?Ke>(M-5LI~@pUMD4`zNu@@LtiT;D@kHz@0d
z2wg=AnP}pD)P|UV3l6MQ+=HxG3FSPE%_5ETbZse7G?rIBV1Jw;Qv%Oz=OmnP0sTy4
zQfRz$SYp}c8wT4`^A~zJp4~2Ry0*v@i6;xnApD72V`K{W@U+3mQG(6YNs8J8JefD#
ze+C?!HMlfN5kkPMB=eD%uk<^cae!5QF5sLilE@#3aak<`QpQss7|AY`A+WIKFp@kW
zu0j+f#iXAf^=jB_@Z<b=`q`XY7pMm{E_6}qjSUbk;N19*Hy=-~R0;GGO@8!z@(we}
zF1enah0!?>w~kVTY_3tH5+lxiI9VDLDdir<3<LpS(j#Y~BO&l)VURK&U@Z0OLpD?i
z5J)v)7Fy7T*5mgfu7N;?(3R8;s6QWP3dd{9Dn9y*tJWb+B(?`hM;HtIT~8MgAZ8u}
zbN_vypP4E;SYo9VK<}vqU4Iwo*SRK7L)sYl^K<29JORJ)zAH&!VXCMFu(8!)T<QKv
zg%|hx-Sk5%f!2_grk_cqk}UHFg~q(H9JEWD=6PiiJW0-t^<B`+^{pHqoi(^RPH}as
zz&lUMPTXd_?{X%Do&eb{?8&zv;Ss7D@7_7Wf7RT>fAOhXc=el)@Qd3|@Vm<s{LvrR
zQ04^LU<wO?$)tzLu)zBtJjUMM6uq4Z_V>2%?BzZY#pad6x^Y;y4vq2H8K&?|vCR{5
zicw0StjPjABIH*_DGuf>UfA!EmtgL(Jxoz(f!q5%eDC23lS1M4rQTD5m;iw+5xB9H
z;oYMO$4i5<_1GPznC1%SwZkA&I9nUaE9eu?V7tFb8@{nqV3<n0aj?eWqQ#|giWp56
zt(~SljpH>1UrYy@P;d}9WtjIMd3;mzqFWh{z80vxK+dJOaTE^H8B<(P`+d?jkV?^w
z^sJ{D@8-{d)7l{k^ai7x9C8GYlHzw*)*ez42jXN+90?@_6v!k+ND8)0xor#ZJR>J6
zv~vzgLL)Z;biHb6=-xe<H@Lo?W41KtWeQUUoQw^a6dIZZIz=iwm*YuJ_HfPunT+22
zPk;AV)ntq#^i#rdoDUdg<PtrvTI_IW3_|EI7BcCSI9gjo)XoR8iSM%+h!O(5gr1>z
zjpNG&+V~3gd5C$7{Y1cl53>vY|60(7oL>+XNq{p33CvbPU{N`|f70OcFvZ>|0es-^
zC~?L(P5@PtmGRgcBzQ7!prpX1af(@KApz_R6C5oKu1+%m<U}L|fz-o{u$b4c^k6-3
zZ`PtnCB}t9S+;OGK{hPVHZP)S4Rjs=UqMtqgMaV`_wbWfD*V!&kMNg%avT5s>HmfI
zzn|d7vwshHI>6rk2+!_~@Zm=@Kx^z@-ockYwU4XA1Ts@hLg7)yFq1f28|(~{PU9W_
zPNJE>(2{i61~TH*5*?~MTa#yB)l%uRJ51Rv_G3%M)5(x(iP_p9QJbDuNphhXWQs)$
zbhA&_EgI{I0_;30>v3(8VP}}|sUR@z=?<%aCpqXa=fy9mTWx*lT#)#$;jFS)n@wK4
z1oh0oCWu4Ir&)_X3Ivpbd1~TmpE~1-*fB^q+6mA2K=dQ1c!ZQFipnlq-)ROS9uPh3
z6Xw*%Hcbh#fXcK$FC#6!4;#*dl5`pww%KYd^D6k-N51;9GI%(v(ON>}qZWZidrXA-
zQ~IAoFyQAyhvpL|LzQ1YBarf}Gz4PuV)UL)+FS`_3CpAT<OZL{HM2S!>n}Jt(|uj@
zsMLB+!cky9N16mEA)XTa#4+%q@eFDPCg{*cFFHU_V5g_C)zjo;?1<0<0H-U%LZZ+)
zT{j$wCmF98<_gn3RRyt?i<OI(0+T{fB)lhl>tton&xjP3D}iyLQ6vg=V<3`k=yZyt
z*oIRBlwttadktTIcp2~fJ2#QuKgVBgGhE&qBh3{;l3=!O0i_|b40)bFBxI>qy8vlC
z&Ccf)InK;DY;aauXeDv)xWW&flsGRf9-r4dw-uQ76C{#ceFureR*$^ouZ)r|+UsoH
zG8Oa_5Lf{HMB>edD?DDbc(Q0Ip~w?5pC=M;JzBAHj9Q2vyFSFr`vnq35BD%5$CEpU
zC5D+IwwfSbdR&|Rj4p2UBM>0Hcx?j?SGEgT%#>hV0|%6NLUkJ#y7bx@k^*oEFGRN(
zFTMgc0+WC!^w~HKovv(Wu-UkPC69t?OCYG^jbOK*&?&&44GFhdhx3+*F?<Tg#l39k
zL^)X+e00`ewl+AJ)nqp=XoEQC@Q`_JJC@?ScEk&C9zg`m8;fgOIS$Ti6p6&~f~4Z4
z@FXzwV9rFN3)ot`&q~HBh(2Q^b9L1YfHx0TSTz(%T4-*U1pPa$__H9gG-*s1-IZ4@
z;Q_`6T$Grct|=-jjy;*U(v8(5e|}o%w1FERqQroB0Fu5%B0=<Xun|itnV&#7R-O>f
zFfBAzt;bpA@a|!Sqh*WAkmp0?XhV5)-k>&|P#icNnx`4tJ+-MM1%cAzLg5K>Tq+Z2
z<4_8(J$nVuJ^u`Pm-gUQiqVyu__u%TMSS_kUq&_f68^_GPVj&9fG?hu_`8EO9yc}4
zYlp3IpT^6xGgt^@MTSe0oGQIX=PmYz33f**Cb>do`~|R2V3_q-xIv}PodC?M7NsP@
zLe)ClIj%4+XoQ!|4YnACzqVBXAf0wEdD;_z^xGakerpULs9&Ae)Dv3nYp+Z)q?$bX
zXl&G=NI1m9_zu9kwOgPv9!JX-&unFAK)bf_0fit&T4Mu7>>xC+shZj1DT56p`JCjE
zf-@3<)!IUfO+5OD4L^)yi)(hqO=M)AMRF0(t))P)lp)u&WJ1zQ>0#a<CAj~1&cPZ4
zQY{<GB}}v=#+VOyFmt#($#A;1nD$i17#w6|d_snon~c4j9e7TbEwWT0K;U?7aBVxs
z$;v|V<ips|X)wHJqZR)AiITW9%+O2NXFo##01=x>L_t&}UfSzHN`ad@IZjJQBVWFM
zTC#Y@lc7drcS;fE)l9iI9<8CzdbDhTKt?V5qYS-FVObeGUK)%TsEuBR%c><1H9ARy
ztjHCw?AYH2znQ}ixgViXD_MUH^SC+DRL2HvCR%|ObW9YcDr5r+Cj9ezsY0S9*2cqG
zT9DVKDHd&X6%K5{Ca^oC18|T^3=)cbGmgS{%EqHeX~EAcqR=ncymZNU)T*Ip^O>y-
z-V;fzpDH-pV%<7eBnZyJW(k_r3isarKKhA8dis4V{4Rd`%wYee9qj(#F8<23e~oW@
zf$!b9iU;?e!+Q219()*Jw^IC@47fbbp>OR_KeEm4vD>MJVbV?DehY#<(?h^REimb&
zSa1)0ZJOcR4^|je4%ep{mfXJ&a*dsWLOZ_u(F&iwJ|Nz|qVnSbmzbZrLNvuo!vx=d
zyvCxoxHick0la#-ht>i2PipjYg<e7n*al#yuW_=r*y&SwMC1~h{>M;JfO_9`L#6%k
z$^f7RI*cS0OQ`vFK^squ)!OOn4uB~w*+=;R)Si%X7pS*vn0Mg8VHuA6@dn8ze8T$x
zEvXj<f4*^%;V1;W_hgNoVTvCdkVU$Zz%yHPPo69daxHLUE5)OEi=#!0tJ56gToF~b
zq<F79p>lE3BevP|`#qdh7NcAtPcqDwEoPO&PG6&!DU1q@%cB&_*5jFJhNba%aNbhN
zjt#gzB2L99Q+Vs~8n0aLqp=2eXDu#|5?ZLKMA<rcPbXyS2y^-Hw8E&+XsyTTnrN~&
zb_z%lo&xlq+s%<Gfs#{+-#?~*_h4qRKTJ_%8Y^BX0I(9i)07m7^%9QmNi4gkJu*S5
znN2w9Da<Q}QK67Xa?W)i7GecjaQ&28t=O&@52g<boUR?tYl~q<HuEM3Jgy8@t;6>Y
z*2uNQl}UyxY~*@lD}xIGv$e&|?Hmgp-EQ?0)=mf{5I9=4nB>G*jm9(YAC)AI%M@PN
z>tUEHj*|zRRH$l$Q9s9MG=_vnl4`&ieCNAw!>#TkOEO%3<?mzKPw@L57+(K<{Hpsy
zygrop7yWne;d{ReJ@|PXAH0W?4;FayjUV9kPkjnn2_Efm-;?m{wsiri211~+vm12)
zn5|pb0agyF@g&9byTr7+y+>9>#swneu=zs6BdJ%e^if<GVGUD>XSZ`aK5q#4VRqW2
zmvo|&agX@St??N3G_H(OqE|K)fbh&D#iUT3icw2oZ=g5x_CUs_tCrGjD7;j0_(U8p
z(Z3yM3Jn<~0-#dcx&Xy|0P6_&TU8E9N%RtlCrg76NNkg;4ZSo-0UvxfLi9ZS5Q3!7
zLa@O`CZVLnt=%5ZD~nHEAE4pcePujY!HHEv_FPr#G3gOK_i$mbZfK$OGujLe<_(0%
zFe(%d&J2tTxHiow;8jUn9xG&u7=&69N$0GzuufswP{iA^B|pB>cuWKlh(?9RlSPYB
zPGpx<Q+Vv-GXfDaEs!U)@Fux}+se^P=>5|~!C2bt%huuL{T?wMnWDUEiS^iU_(|DV
z<}(B=OA4sn>g!I^K)MJH`?QXyBuvubYqj0in|36i6BnJAG_N;~bSqj4)Sj@bULxRa
zK<(-M)(v3^lU&00G#;PO#3LvI^{Uom!6Q`!U|xIdaD^7DBLZHDx!ThLXQe0ZM<Cu!
zuIU-9nYO(<OtCxUwvH!qY1Genk0MJjQ8}_K1+>Ou-J&!GVgCyB&OY?I#7W&??Hz1T
zI9?e1vllab_N^7Z^85t<>vQM0`|fW+s2%9u6v^Zkk}QP=k6vO-RgB9@H`V&);pYDv
zWE!eRWzDEixIZKE!p?wDdgCbkB1<JkIdHgWF&EU|PYR8)p)?<|4yT!TKcg72noISX
z+lC^QSeL{e%cQ_M1Z;8ceXQ*+g26G?TM9r3iBTpydTji20dy)W#l2*rC@M^gSSHe#
z&RN=otfQR7hQ&rXbKCnFakk5r`o4Z9sdsA~hPg&Bl~^|p=N1@cs-uunWVXlQqCuW%
zoUJX7>k=#*x}N*P1Rg*?LkAdD4YebKOyY3SV#_CxlJfhE2M*^3(?QaitE5UnX^Gj|
zl9IA^80Cu4Zw3o1<M7d0gWXX|BVk1r+vhC-r`!C!AD=ZinAOB-FX$e<@o+_P)UC&C
zP2nr0@kq78`MSlmX^wHBNf!{v3^r1mPuC8uAvdUbX#h!1JMnpI9C1WkAfCcjKLJ2s
zZ)Eqb4WXdK`Jm~~lBZU*InXH@gH6iT;Xz9qMXF^tS~MP5lon@v0Q3@x^V(xp+U|Lx
zwSexPCK6cJzH5!*5$8cu4XwzE%<(geK&`?5SJs*ANOII+{EIC!D>Ex=>*}Rux@Q;$
z3&}RJ50*d%2`nL*8xjXD9FR{u1kb>QE6>0kIAFwqMKTw_>)12n=^6KQSJ#?VxkN?;
zhks;tTf9IktE(dNzx=-Mn=dQWl4Iy`?z*7>S8W`|sq*Z^tsgZbXYDW)c<oMx*&xC)
zFHqMSfBfhIjnTM!IK&TMJH>okV4U>u?=K4c;qvb|dHV==|9XXAyncXRV+^a7h;0Ua
z{SnN`cfg$`thew!dk*wts|^7`xb?xRhM*M(x}i!hd^i~dm}Ux_vgvNV%F<=gzW9~&
zUh{(Nrt5~-_xZk|Bv$`VO9I{rTDFz#2*Yjd@O)JhIO912W9iM<evvdN<2T|Lpi4Y%
z_1etu^IKp0+MrW_Nk7zU(`2@##mXNzZ7GRoUlYYBJ-B48pWyp9ew$F61Oi#ayXf<c
z=ljjA@6|esiw$|{w}{1yHKpb*b|k2(wZ-K|cg7{;CS<!x=c`Q(Z5&d6xde)1TT*H>
z4Wme-VGLDHz)id~=gd#KIp7G*-B+aZXnXJVUgLaO<H@}ANkgpyln!Mpl~#&=q>3{r
zpD`GRg8FmkaCaC&Fkn+OSo`OwjU^KrM;-{q(5Wu9K|hovQu9-XQBRWJpa1-%r+1x1
zJm9Slqw5rn`<lS6+cxcjGb9mJJuM-*EJ(v*0i<38rz9sSHbWO1-@Y8Qb(r>5J3I7Y
zFSPb_j<|@tws6*U_I8rfL3**(s2hU28cS3DKyeb_1p<<_UJ75IE(^|)B|Ut4kz<}$
zuohV474}-A7lbIZ#`DD%N(ET#O6+ss^6>z_d-Mz+y*0)1)0g<wVU3^tIK%L8j?!6Z
z$064xrQ|pVfKy^<We8Xp#o}vg6DgPv9H?-1XW)0egq2pt;@Pso>oe&+=7?D@Ym0m1
zu=6hiXJ9>n!#=?ujX`&)s>b2LB%~~QV{y5q{wWX)X~4ii(&5UNkb{WenMxa<?1~OS
zvA$esEwtMHvkf}QJt7(i>Ioay+JRFKSo+H@mh_pkP~c^*F^Xk}gBlAwTUNL~36Us)
zBcZx1Ho;h@8Lt|L`;(|M8(r=j*ov|a?%z$3*A^$k5X+nneOX)dV~NXc4aON3d5u?R
zsgD?^)K?`r&L%Wub<W{(t5Io#o)V<e3<b0m#GDI;hlf3%qv%j*f>b7{!u3w$de`99
zSps7nt_8>Wszf%Uvic|s@#w0+`KrRZuZ}T{1Rh<L`21>*VXW}hSxWhl*5UP&xVxC@
zhSr&@g0@Q6yM``le>fk{D^%J*364oD5M~NrZ8WOJ;&Q7Y_cSoH_9m^jYFx)O5V}s6
zVhi--n%>6KRAR9w7Jrt}iuig*z5FaAY`knJZFyT$BK7GoKpN1>dhN-?EaE65iG_3B
zcgY-4twJ#*N?=`De6^}^Iu5-FDLvEkT@A?@CNVurJ4dJ+i?L5et_`hw1)~i~oKb?7
zwVoLDD`aae(q4dlS>e;A!BMJkKF{%Vxy5YQ$G1;L`16NP@$`eo7>zS*t`?}~Pa(p0
zF@Nzl{PwBFI}hK)d!PRk_V@&=Z{EeV{Vwi&>kP6Ndc%kC#`z`?bUUpJgDj*T%Ax&m
z=-%LoBG=!SImJ4i3_>gmgK;c6#aJL|^+co;^;&K2z^Z8QU=kq-DF5&3s=#Mg7X3(K
zoJbsa;l!h>0{6x=H7+$}xJk)zzN+D@>%a=<?Vo7hPJ=q`ao3<HP&W?Sf>LSVs0H}i
z8`+X$i}WSkAQF@aEE(p$gubo3<IICH+F=lS(Tye>=2m&wG6c&VMRgfR7i(b&kzMT@
z91TLR0CYGTM=*x;=qpa8=;<)T`Lcpzz|_aA4P&Y*;(*|qI1n8d5Gm44jAD{?@BAF0
z9|?pW_6a1z7ngg?h7pdkfJ$1)u&p#^F-@`+2iBFrYqJEAWLWJRB$Dv%htnAQ+Tb#8
zNG#*4pwMq$=35O0hh1fm_7n<j5D1B*l!l0723%}vFvjp=ReN0AdOm|8jX3s<`ON?g
zj%#g^M#39^(%ZPPMC!#MN!KfFv8gOF?=!KkNLk21mEfo+P-uh75YGYt(^O$oS)bnQ
z1xy@Bg0Q>LIA}HjMq3BxNT0z#fKS^x`x_Z|1_3q|MX(*j5<6{iT^SrC3Kv`2lqI3S
zwxHaEK~Ez0^5sEzqXl7%JUH{B#>eX`%#Mz*E-J*4z#Dgmc>l8tyjX5A=tVd>93zVr
zl6zBp{3OAXfBgkufbK1D&T=e&`T)m&n&StrT;P58dz^pv6UfAu>-IJe&B!fP@SEO7
z`!d-%T4fmPaPC)#k}>QJp_`R+7)2bpb{Iq)&N4{<H`bNy9B)^pxsh~&n1I79#MQRO
z-4XSdk)(Crs3*ZdeP8PtP}CL-4ueFJbdxdP7SQu6!J!ygH(5uBbEzHr5y!ePFwR3v
zjP}2R0iwNe-Hmvs=Uas|Bu+mKxOZxy{q=Dcpl)b0Vf}#*{SeW343vXf593J6ys3<L
zs&TX%2Cyp)>W11W6>u-+x$!OxeVb5dgR=vgpl?dr=?@bces1;zQS~FyMU?lw7;2uk
zGE}<o6&Mb(n@e@EZ!k$D`k{nkG-WqFadw(2oQ@(W#j)5M1d@X}APPA0+JINUVMgYc
z%tKf69j*2pBa@GDOcC<A4{G6xd<WM45LFGmiHD;Ivp`_KA^{p_%rgQsz-;*o>pU6^
zpW@8g#ANP+NNqRp4FZ8I5<b4|CaPUB(UtGcS!_!~bDV%yt$9sSc<XtCbxyOZK~Lh@
zqQqX)=A+fWwAI+UG+o8XnRhZ041!UcMxrPO$=b_H;LAmYB;-Ud3PqWVeiqeOi?S#&
zj3j!B<6>EZaS7BVdh0c;wOC9hc=PN4fl^2#iPLF{M~y}m4#Dd_p8WS3zu64&_A4Ck
zyu89MM+WA>Cy*8lS@8cM3~dNUdk`o~z~OLuj0eXVj0j*Y9gIzBz&OJs5uI;96nN`q
z27K^kjwlc~>MJ~*7jU3e%_I$I-QfprN^%^f5|Ou%=Khv18ii5Bk%YqY`vBk(T}8tb
z*L?wWV5XBpBq9B+L@{hii_hnK9NPdVnF1HAvp^&8gLThQ)r8u%Z|~TU-@({_sj;K-
z7>)!~g?EdIBY}&p#=5k)H;{fZZE-sEhuoL2clr`V5nwQyVpki?GJ$ETFp5P7ER8&0
zKj{fLM%9S+tJq;ghgLThnlmhNjd7yza2jF0Cvj8@)&KxXgI>UKHVke^M(c3kf4D&^
z@!?DQUhW+qqCXiR4JltR4iY3GhhPL5tqnu$VVJ8OfmB&fdfJW#a7r@F1|ha3_4>16
zfKiY1@6I~(lpq~M<1kG~-*B<kIPNF3f=v}zG{EEKVV5v$Tq_Ww65IGodRJ5400+Dk
z7eNv25Uy+}pq7Io#vG%8vtq#3SoFOrvoSOe%0e2jS_iPkVlqfknF-QRz^1_SRfSM6
zjMD(80kt9%FH*a`fs&j=YXChb=Pd{0kop&B7;voi4X$?v+8F4DVVVUP^u*2J-2ow&
zI5|GXyT_;a#_1H#mxo}EqWI!CLXqc~9!&7_@7)K7!$Cj5ql*D{&)<NGN08+un&Jv(
q@_qbg_YWlZKEaP3S9s-ghW`Oiu&(zfz2v?C0000<MNUMnLSTZ`O=v&>

literal 0
HcmV?d00001

diff --git a/ico_design/icon_v2.iconset/icon_32x32.png b/ico_design/icon_v2.iconset/icon_32x32.png
new file mode 100644
index 0000000000000000000000000000000000000000..65f2eaf50411c8feba7f480bdea1f576e9f70d81
GIT binary patch
literal 2551
zcmV<T2?+LyP)<h;3K|Lk000e1NJLTq001BW001Be1^@s6b9#F800009a7bBm00013
z000130eIR=VgLXMsYygZR9J;gm&=bG$92Yk_2}n)&6~q{YDkeB$`A>~3K=IxoH$;@
z5F8~7!+?=hHbJ~HkbSbuCYvCCfH&d7JKKN_#fF_2ibN`rOokv$N){t>#GyFD8FKFX
z_M_@#QFlfvpn>j27wVkveBXCYtH1vKKgiXqf5GMDHqsw4>Wz`%gvnrl@lZ`Fa1JVm
zt184R@E*J%B6#P}N}~`&z>6n5fg&vLJ)-o<vxrA2jaHz~UVVdj_&?l#{Di}*;%R@(
zv*DPD_h1ypDwKw9<@nDhBOde$P=Y7{!sW$>V7y1XM-ULtZ@VS``e?-cUWNA#5zpHP
z1-FkXL_A(R_RX(d;+MbJ=hn?@v={ygMxN@~GtSKAT;5z{chKj(k9&O5n~-VE%QF^b
zqkxJM5J4*qA_#(bP)dV>Cq>1a(d-t&cE=!71W+I-6bc;~&Fa=R?X%xyIlaNk<|4~y
zH`v%(#56Oyh2!$ZEZcKAD0n<_^qruLK^u)S8c~YE2}UWjR-mBt!o9KM4`wV6OQ8}$
z14@O1jZ%mRdQ^Iz9uE24K}m1s4L<tt$K2W7<^GF;sFiW~mD9YuK8w<t=apmPh5d=+
zS>fo`f|?$hMh4Lek&xo!kz*y+oNYzS8u)bNP7n|TC)2$~1Fq)9XiO!Rd-Zu{^!xOE
zjcw=XMuu&symod0fKC>3_j$qH=LOgI#xO+ztw1T%6oLO6)~qK6fVI@{Wa2+(<rkJ6
z^p0pGo^u_|_EN;*<_6<_%A&iDi_fD|h181ml_r1tl~sOz_mGWtOm|$dlw0=4HOsk0
zDM2Yg@POu%!ZR5=-s@KYq+0PuK9E%e6oLZOZ=Q^$9Ck5T2QLs?h4aF*ClC3_kABGR
z=nS12);letqTuSwr}*c$?y;OloXRXa-GXm!wrH6kPghmkJt(-|tNG4Sj8+Px;om+h
z_{*h)d;OZ0fy*;73ec4c=ot-p6b=W_Hs{vxG>h(MJnZ-B*^HiObTj39Uq8oD=~-&0
zOuUdK7GpJNjnbNKC9Fg`kW@huX|@v0?%0K_r%F%?Y-FICvQr4b0K~DS?=s#Qe3{>G
z-R99j%#z74c}i2@&DU1>aOVZDF6O+mJ0#I?<6y*<nV79qp_O7>IJAZ|(tKk%3%()2
zh?u&FPK?HAh1H76308;A&D65JzQ)G(SNO<ZWKP{;p!&o!4NN0tb7P)Adu@$Byf9BE
zkC=Gj?|<(U{aV<sLY-@o;E2;~L)rm><AiNh*VJ{5scQ^SRf1MfRyF+^4ku%d%nBbX
z&FhVa=vIz(9`8LyJieN+H>x;4m!phktr;Q1g35WUQq;$oYX$D}%#UBI8xIbo7<GyA
z4jQPv-~}Q#K&NfQGv&1~8oc1z5B`On?iP#5JWUiA7E>aV@YbEDAi}pVbolSbBmQJH
z2j}t5BXx}m_7~vy9zi1_*d%Y_U4@PtL>g2G5>p8#ZQ#A&e2Ch_CD)HuxZ3<3ANOE6
zTc(OrURqt?@0QyXqmpYk9&u?Q=jOA5FU=>!R!@6_LcE~lxKRcB)*^z@3I&wz#I2`#
z4Q}P9aMey2<XhYrud}`Q83$#T!fGT-sN#hEQAOc}LSSc9^TB>FD$&8L)UjCwLQFZ}
zz<a>}iB*)-=VQ|<UGQ+DA+j2xn31`_yWK~8b^ba-PLa+uNHc>qim&!Z+}Z8ZErR53
z?vJ@RYe{v;SOFp-5beR1qb}8`2Q|c2<3;dZD9Ztn2TkCBXoW9&{P|`Z?;PH3Q<e*?
z&pqXEvdLm(I2=z{JKdmk!s$lDd!M{uXISx3cfxnhw&_)#mH{8g5E|nLyJMLQkFePc
zUV;n(r08MN_6bx1l*2APp94@iA&oQ-KD^D%U;l)@+9s-6w0*%FSI)4rbI8BnenL@s
zHky{5LB*>x2`#HQ+teTm=R7*Gn#5`%1NV0aL`GpvXqNzq)mW`?K4`a-8k&ir5o=~L
z%jWt94_IeGenn@kgKniH?TnW;=ecrbfsNS=wC2)6&aa=1=+}Za!6ij0Y*BkgrKj?a
zS5})Gjp`796JnsKJsyY+j78{9+zCp9((&7eFL><NSY6qmdiWOo$vUd>D%uH`UpdWe
zGv>#?dCaY{Ce?~=<#~5+!qrnb+Gwn}8t-b*HO_042z4zOAKq&<bx@^199n9;n+_IA
z5nHI0a8RT?8lUDHvpYnc9HwL1`r0htsw?)MPuS~Cz<W}oxUo0p%2JB@;L!^yhDT(L
z4uT4U0#J^6T>IdNr`-vk?!6#0inJQ=kAL&0I3MI`HTX#nAGJak6`1r6x%Pv9;$iO+
zOUW!vt@-AqB@!{beRmI>=l1@PjYdSG1RW(ACeAQPj*c^Il4FvbD9P{=q2rt+$(eY=
z*lBE>lcWuFl%b;VZP=S*lN@ZuyS?))H1F`V?9y`{-E0tdn*8ylRkAGRTjyudM$^e+
z^sy;0wc*cA2=Jbb*__B4jEzVmi&8<BYTCk)P>$M#fmDD&eulf_bF5`Ic|1BqQj2M3
zM9q}fFPvg;>{x6iTv}~mkM-z;sYy)rsUXb~jL{@%LNiP7!xHD^<Xi8<KJx{sihwr}
zlX#Wuy@D^d?qNG$CysJTzzV#6ZUJ#nj!V>k-TzEn(ZfY86c8Cj?SZmSzp!L^%xGLu
zPAZfevAl8$riW_{)wqXE+IR%56q8|>?%3e#l43j}jVlfwyw9h-mvFtBIpa9D)PhQ&
ztg&^}Vlb|m$vZR>!?1AVMl%^HRu?<yX^D(hj0Z=kq>Y+(*5l%(jTs`N7!;mF;M~eA
z(`up>{KxHGdXH~WBwxZeQ#R8Ejm)4X1=<=#JF}sN60-0o4GEkNL&dR=dkNWxvdPK4
zQVK7jB*)9B@cpMpG5gs?O#c@=Ee1iRd5USIfrQQn(hVjat)O-u=Y-ma%<UJ(X*K1C
zRuCI?a^SFVB$1j<yi-yOco9xF9P^7y?6+QLYhe#^jqT*9Mv5(mU8->(Ezn2}qoEIQ
z4vxrUO;v)Mw(EM*M}Hm*j=%vQYa*>E3c=I^%CX^#>&>rSWN&DhX>H*~dmKLaIq&ap
zv9-L0RY?nrFi}lR8d12KN-5IZL(&EX#wa2kkg9A8GrjjB5Z92ji6hN$SW{b1l(q2^
zO5g<2rekZqMeQM)o8zswZ&MxpFHh?^?0myA@P;(D3=2;u4L^}qTZ{p#!+&4xg~Gwa
z``~~;q!p3TJReq}zXL_-s9fO9@z0Re0bS<<Z_Zr2!aH{sSa|+Z{vUW={S5KVQbzy)
N002ovPDHLkV1k4A>{I{%

literal 0
HcmV?d00001

diff --git a/ico_design/icon_v2.iconset/icon_512x512.png b/ico_design/icon_v2.iconset/icon_512x512.png
new file mode 100644
index 0000000000000000000000000000000000000000..8b8580adf500fb16975ff30ff05a721d037e5b19
GIT binary patch
literal 195130
zcmXteWmFtpv}_OV?(QT&fDgAIK_ghO5In)%T?Tgx5Zpb;;0{B8;O_2jgTtVY`|i3w
zdYxX~e|nwXXIJg23Rh8*$Ht(*0002{qk_y=06=)XMF7xHUY9HP>HF6O-CjZ42>_78
z|K|ePm}kB|qys+6NU6JL9JgB}G7q@W+^qE9TEb5MthN@LWz%4x@CN=hq({S`lZgZR
z8RPlha)y#zch@i{h>7<}Tq!^2zg#fTUMh&((7!$Sw)+DdVx5M7PQZ-}228m>k1ej$
zUURhD&1mQid2#VRZLLxcjYN5V*24Ad?2tIrFqK<w!f%%GcEh~CTzd-T2ObUn$>y+l
zZ%M4818-Kk0dQ$yAOK!*q6RwK%0R%76(Z0X-~9jrXgj>nfvVN5%RMe#L_jRc3<Sjb
zU9p0IGUm(FBRn97jUDK$J3G0CH~qPD?*#!WLk@3u9s&t~s<%6&Es2R`N8i~7?UHB=
z2;mQ1c5}{u_XO`A%lowyJo-ZSZ<|;2wIi3k^*Ozs*-&$T`(D5#M@hf>AFAtfnQG7%
z{RPKradv=!`Q``CT{R*L;R9v*6N196uM{1|zJdjQY7B-4eR^;Q;dTk?Jue)<zK9ka
zLtFTCMJ?;74-Y;p(rpT;%jsv{DYb2nYT10Y(S;%aDXD9j{YMXs-uHzYIV;yYYX7~n
zfO_*YCi7aiKl11o2`hUVU|c?yHs!vzk)VF5v4#3&LC@X_YBpt_VkN_z6K<~k$8Qzj
zsYX|Q+)s0RrFs#MO^XiI|2Uv*-F-tAi?!W(PZitnQZK95eQl;j_j?f^khoa9zisCm
zQ;gGE6`V5KDZ9=ThV@NcMLHt>d8)oh^ZX6P`LZ#&=csaVTSH}iaa{`E_Ozu+QeN}c
z%J4Wf%E}+~iw$)iuan;uCruPD{&XWQ&_P@JEt>r)S_9~u9*&vy9F2nbC|^C7c^4ko
znN*t59HA`u-al+17f|F-6zU!f!2OA+`yCEqWhE<MDM{J(kC2;b=3jB2)C;crfAbm$
zRqWO2x;y?l+J<r*T=!N>#9n<HSh$%IOt$vnc^c{`y-W8sTeN7KXeRBvS?XU{PcO^f
zknh$PGZ@V#t6s+QcvRl~XgfW`leqR+w!GNE^?oswm0e-$bGwUIe$7a-6_=2&L_>J5
zwd?do(*oaBvmfMugm6!wU81!1F^LoeNL=6f0{qX8Mu^mSnqwD(FY6tQ05}@t$E}4-
zPqSQ02%N1`M*`q~+HyCx(aAWdu9u(e580H3j^0<!WjacNoG}69|2Da(?lq9X<L@kU
z?7dg<8BH1Fej^Ftso)*91hG?zaDKRvk<<&MLjS?6{T-jtK^{z(w7@wu6aSNeQ=Ds0
zP-gWYfa~znBtF+C5j;g%>@`k07&E?S&6U=+xEkI%T4vgV9y$SJ>t2SYwC`!cRVvZ9
zDg3q8WkCTrg^!7B<Py_uXe!v!=A20eTDv@AWD4-N8>E^A-pkSP-0|{vEm>O6^%i6J
z{w3k_?=I+`)P`2neMN|T74n$*cktA{AqhpqrYpqHWg}^1;5j@bS-N8VsM;>@MfmTn
zoWA$E*7IbTWkiOhOq;3fm?)Sb+P`7XE2JQc!C6*6Gl16J%fR_g?8kLldLm@iFH+?L
zwC<(!u^vYr#iBl%(MENZ;ca12+ipv#-Q37;NQYkuJP3h-J^3r*#sEC8A~5W<EFR*o
zvK6@Bu3EG8EgkdbdC{Rp=uKO!wASHApTfxwm0gd#5UK+w$c&4(w;L9s%^ZVqt8H!n
znq^AbnAYIfSEDS_$bkXwVavL!jxeO8=GA>@ykK#2_17f_yUxEDlDPc$j&1ZD0^9Re
z2pp4hq<VO!Hs7oI^_j;hZTLYX8#;<$_C+|dO!QS7)RrG#4n19Zn_OJV$tYxT&?K&W
zi0)BW6uakm!DNv<<jRR6Bm||KbiPI+T7HBm5@|GX%-hiMgrAT2_fr)v$y5njo$0OI
zR`#A?!E?c}({;_silIa=iSY+p!Oz409W=143p{bF7GJ(udK~NUqyf$@L&>(@*TQ6p
z_&0iViXEssu}-^1z@u5M6R6z#$z;Ghk<R0ehCYvOt1Go+)rshj9*G$jfU0HH{5weI
z4OKqDA2xv~4dO7gqb@f&@d@fpS@1pD74imY(&GnwE=C_M(3_qc#Q^6k&kf1IPlM;@
zHS^EK#+&H}1iWw9!PCE4`(qX2-ouj?h@|*%Yg`GLmQ|4~tk)v04(H`$kyszuXT0&i
zUAZ)8?!m-w$v|O;ANVmwc@dQ_W_M{J3i(G$cZriRG0Bzoq8#@7r03O^7|uIc<zo5G
ztt9)!dzgQuL1c<y%nbRtXj^6CL^KYv$&_j2YbruN=ZOEJ44a`0JXN!sM&M~ae9dcK
zRYf8?msL~S_%?%Q4RPK4uLEKAwkIt2$xxBpBx)B)-sRGHc^NNS{-LW34vqQtb9t$!
zhUz;_Y!s$+paUYlNu0s5rnk5&GJU(?yKBcE`#R?MxjR1}VA+jd{M7RateY=>;@{wb
zLa)0mk*qx)#6oZ2K`*xacY`il?8d~Xv%7Zux->!pUtCM6K6y&K`%-h;mtMD1cP&+w
z#`O6{eEt%9Pz-H0lr1!nA}zZ2iF=3lM~+<duK<=DqHxdWkpDCg>NOBm_~Xpk!Y)@i
z$b%qCHYDL?Cs2hW53Z8`d=N6E&Sq!V)?sC8UczJ^BiB@4?aXm__9(@Xv5_)7f^8(Q
z`F*gXsNh`f7J`h*F|j7;UU`(r)|ZStqczm$m?=N)mdKkwC}kP3%2f#IKS*#Ji<-7J
zQUFks{hOxcBJ`5WwYI_4i0l?G5O9X8EQSl*_4MI@faisHt%>07!nye(iN=m>egLx;
z0PiVhfPmw*ZtqH6T@>I8=H!~_*lMnt<FZRn!;=)U&|W(n?)_K)X4L3+X7mE)T62Gq
zUN><}dowY;8&%1sW4$?vN-YaO{D3sAfgKTFQKl*snuKi&g$Q`nm%5?ReIPZwvx@VP
z;%~ziNJFBomdjQ69VGR$^SYn$`qQ?~im@<BQHBIA!+CUA+xZNocJi|)WB86nP*OQ}
zCStg$tX7rhf)oj;S{VUsq55A5SKuv!8Kd`h1%`#zC-h(H(de@5vt)nk>Cye(f4D*a
zXYj&yh#DU&eujxCTqn@aJ@Idg2uY2ynL+<O4bxWEVhDQt^5BX&3kQ!SLz4ZMpeq!D
z!Z2KOZ`fP)FO})*$tb~l{;)>gG%Dtow)tzwqZY+}Ddy~W<-|9pF0)g&kj=d}h$%ID
zKWQAAb>WMNV+&F5x>%j^-?-5ZU#k40VC93kNlP2vX%)34^EOJ(f7+SsGE0G#w{sa~
zsL_o_rN!;_%Z>Cyt|PQ+R?kUNy*kYnZSL@l&mt{3uAUlIbzN+!h9&uXWrtE>flt2s
z*HXurA-k96*~xbgS=<7AJu1V0QpMTq9$;mbME67L5iFAYKw*1u@6jsbVg^fRR%8F-
z{3)gWDWfHV<Ldp!$jt6yK7PP*Q5O8GDgGm#jIy!%8|l?!s=VtC3}GgpzzI+^E3X=-
zZUZEsRGer5H=fH)yiXusgsT3jI{u*Qo=|!P|HvfNYSNX}nww9IrZJs)x5;@3jlnrG
z+$r7G=1{kc1{I*v5yk=BuZv;G!12RWBM5le4vaeXd&rto1>2u5`|N{KvW&2$nyUvW
z{H3x<5nZA&I!CFg1jQH~x{;V7!7ZC>H5?$veJfpI{PDL{9sdE6RPQ>;Ft@2Z1Li*m
zwuQSBRQ_ReZ??-tSJ@WV!&8Zi_0od))H_Jyae>}oNDU%c|K^ZDr<F%?Msp#DqbRgW
zhLMzbt0T}R%s+bgOHN=Uuk&Nu6F|8ra5321ItJuBtu?0nr%m5%N#O_m$-gEi#>s5K
z*a0~cHkt&yv2J4li>=j3Xi$4(68`rAzZpS=xj9%WOJR6Fo5e%adN~#igc>f(c_628
zpV0AcZG+5$J+DTASKmO@SwBRy+-?|QzI^(90?~dkTK_6E1uCqh^*G{ycJ%$zU0GyU
zL3{tb(tPXqSr%&Hj*gwq+0L2SCTL;Pw-Gecu<4pn9;}F;G2+Ml<t2_1l^PO)BHhOv
zI9z*#;<PG<Jbzug16d25XxHI=bV_`Sw@ND8uj529tJGa|>L7EUA<Rp+=-X$W<Jj>Z
z<E_pRc?MwzY3c5;rP#3>bCO7${G*0Q%pcW$VcI?G$8Pfylukh~1+yg#aeNq+4!Ya2
z3#tk;z;}`+G>j7?CRYTXxe16!9$`-)33;*d!hv$ZHk&nuQ(blcopcv{OF9R}(3_=u
zv=mWTlrG3$%?chO*OO&8&HJbb=eQUxSx!rfy1B%*q`SRw4doNOdhF-h41Mp`p=z!E
zH<V}8Mab#iCl9^J1P@kGgx;jsXn}OdvGPI>bRJqX%)HdT^oRq{Ym3pd+l9UPvKK8_
z8wPMkTDvBR9GY}Q^}b4IOaY3dd{12+J(UxR(J40~7tm4!uF60h#8Af*;Rh+T4eI?x
zOAFf?%tV~QlNlSYl?E_E-xdqF<k%}6(hpE_e<55C8z=2)S)Ic2?CDp)qhIR(c-QZ-
zH~y~a={ueeIi(I_5Hb^6Zk)9InWGSbz!w&d%H6hivy3Lw0}PToM?q@nHK^|RIp;6+
z=hP}eG}b>SB6tVYa&~5VE6>Ll?IMMU@D(i{TUk9g848q)1Bn7~17ilb6~QD03IGB;
zxL5>Q#+Rr$RDH>OE&k2UFo{)4KbKO~+f_aG$I{O$!f0Cj@E?rd&(+M8edIYVst6lq
zHDSI!`gK`@72-Gw{|JH=)))G@z;$ifYo-Gay4J|5I;U<NT8~Fu6mUPf={*0h(ebs2
zrf>5d1{PADwruax`)o^CIwOhw{KmZnS!7chGQVw+CCnKcZ`Y5tV*HZ6iM~SIor5+?
zZ#P`)J&Mb}@bS7u$mJ-F3EmY!3=+|uWtwfz#G}hWDt~}0Rij$}E0gs?qib|?x-}yK
z&k@x1_UCI)77_5I=X3Y_FNAb`ByX1<5wMrqC-Jy^V1$OTicct$*FTE-u(yx<W|%`Y
zIkR(2Qi>3Ex%27KLQk$89r-ZHJa|rGD$vg)bZ&?K<T)KDQWS<pjm`6_{%o?1TWcb&
zH7NPGlvyX$E@s*-F`g<>q$7t{r1K?X$=OPE?_PxQQMeHm7&mws1AvzccrplpthrGD
zXOU(?yz6%FGLeDD^J#5o%k@ZFI3HkW@tbwuuH%jqTvd}&DlEMovN(&pVTofg!aeW8
zQ#xHvUh)}%s@e|tZ<2NHZ$1C4`fymXr~IKZaU0s3&VDL>E`)|Fbh>~t4~;o|rjECc
zN{G6Zn$6#0j`H57u;InM@6oBF8*O5_$Yb;X6^Ikl+h+08Oe!N3&_nyc&g_F#r1VM{
z4Sq9;OsNNN=v-wT>*t>}2-hz$6=;IVJiKIx>8)Zuvt~<Gsmr9AqvS|Jk(*IH5I&Fu
zQG{qM78_eg#ISTyZbh@u4aCXv{fZE^t>oXLl5JR<EkRy8>HjSJE-MT_sqv8y?H}Z4
z=s=}4*3YFQCeBtH1sSaCiR#onZ;e1>xpX&ks~(;J7Vz-D-V^R4B}$(SUT?N_?#Iht
zxUFgaIplv7`yP#yYK@fKX#*SlE%1x~?xNs#Aae?1$1^WMyMRzBIp+1-I+bS>|D|CM
z0ky)p3~~cd8M<l>`Hh(O#q|^o>RN7Mb%TWYQ%;cY-QM`BpJno#uI)XAv2Eu(1yD%P
z3C1%)FxEUoT<344aTt_f!IeA%#l_$bkeTDR^D20$XM;DIZ<pXmJ!yk0LUu<|OE%D-
zHz<O=BMd*w9LVy}tGv4pT%V=3;x-cE86kuX-jjY5xceiMNLX*z+S;zKZRU{p?(9Z<
zFSu^ZPlBhGmPbpRbqK58|1XkZA4)nt;NGl(fRxkf6X@v*DsXq0zJ>~v(;7y1Dnnf`
zfV1)U<>XD@@Jq`7d)IvFHtN&?u1sL=<VQf8kHhj6MaS|(1v0vG#?k-=`<RzcZBMA)
z`cCKU7rS2qaLSG{@)WjqixWS-^FZif?X*G|#E_v^?lkP|*TM%gSjes)j#Qo}M<lu=
zcV<amec%I$iyLg8wXT$&=<_ln(0c-skS|-X*S2y*K`j_tmt<wLXd!)}N`!#E7V}2I
zn+huFHWSD6$KicUT*}Ru>`7ehHJ&E;BVT|&0}-4oOj)wVaRo&YTrVtWV=8y5U@eJ<
z5<h00x-EUyrwWaFr-h4ffyRJf4)De@{$MIpq#KO3y|;qGq~cXr`&_9_1VyKFD19ou
zEB($7!fx#q7AXVh<Yk#Eh*n_B?u9$){U6tqW7l~w*K+$MoYr^><vB}@w-yOCRl65f
zI2F%(k=b)^%)(L{mP>BO5bmn>q!!1(q+ff5vw!to+9C~|keIhF#6dn~+gT;tdhp7a
z<gZy|vOfCoPT;XckC4V+ZcH%Fgl)T~4_k0jH1b}qq<%xpQTJCNt;_|XwSqo%ll&EI
zdx<_-*}bRb_mkXeYWJblmtYcKGl_FW`dFAjJ}Ad0*~%4AoWYV452Z6%iu3E|48X<?
zlE}V9=YASJvoXiAMv2H^0W6x+3(@--E>)q%#yeK97ZR*qYda=~Ri`u|k#)J9_6qcQ
znwEb(&59D=4ynYO1cjwqAdO#gC};LSfZaqf02siL7-6BYFSG#o@Z%`b-N}l)>U8H$
zk{SR$O8%B$A4ZJ1@5tl9Vb^c#9RtXxRDwF;RQJdr`ygsH&I~RzlM(1O7_qrnhLC#&
z<USA^QjyJ|D5)oHeNREHf`C)PoaN_|$&b@A*Qu)o=Ruf_a_6N+(DhdualC3r$tB4K
z#AjD7o924LCWJ@pap76Q7Y;WO>SZ+uf9E@>jJ~ssyGRA<>5&h*=rt?h_cYM{yOc+Z
z=!^t^_rcyA)+D+h+Nbz01C=k|uxrkoj6mk36HY@UD7qOHJ)spw@TU7DCLg`z2i!=r
zqriiLPL+%Z&#o;mhJ`e;?dnq>tDzy>xw&Z0MeI#2)7-`P(R-x2MZ<0z^J9i$FJyI;
zF20-3f{#wqk_e_!T48_)PQ5<`jfW|{`C@i>+hC}LFM|{x8uK=LPQ#DrYv%?F_=ra-
zZiS6HfiM`-A@+Qedj7`-*()U;;^0>vv{eyEY3vYe#}a7o)Vidd4xLj^y<BA~f2?cU
zfYcL(FS(|}K0i2ZF7YE~TYe2B^cF;P=jg0$ezIu}Vd?)(=_pao6}98n{n~seBtOzK
zoA2mw_m6wQmxNxTSUSytuG4^Neep7M72h<<H3rA-J~2Fjgq)qD(7K5%k17grkfxL5
z$Md4-#j~Nhsk2uAv{+L5&KY@i49|c_6}fjGFkg;wD|W_BjUf2C%RNnEZ?I0gzYjed
z)2B9LSpKux&ueLml3KTsu4D7V6S6ckf&0lMA^(41>?$yK)QndARWke?exw7bB7HGA
zZ<q_;o{D{s1PIlo-vUBALTc;9&~pd~P=2fs1Od>k%I+B)NZSKB(77ws+j%+lD9Wz~
zydfJPMb^~@QHQ_C)1UAC0zv5FjIpMd@6~t=vh+?tfGZI7rO;dNUw?KRODFv2X+i>)
zS8{T5_8e39`m^wPBuf?tUsij$$E~DZ?(deN9J=o-2ATO-1qN|g?yAm!6$amEHun+;
zjG#654Y8W;?bEcV4*qJ|-8c6U1bniIS;v>Kty@L|(8xJ07XlwfCxh@xM(mzhSI(DP
z9JG#@Nn5H;7(&<i1Va3+#J_&!AuS_|PoDph84b9K(Ze*(y*(q791Gq`Vz)V$hW<>(
z^!9WM7N$0njrq^9nB{_jy01h!Tq#g}RKXOewv&s{`ym`?=l>OLO4$lo@}ncqAJTeS
zf!s;oki9@i?==D?tjsDtr~z`#KQV`Ae7qxY67|{jwp%@mWkHO%Vos-@1ozhO!(UJj
z*9TXhiW6)pTSESo$!~f#4n9~lY4PjnCu_FCYaL+CAjrfIPhYv4xQSuk`}SJ$mwpM@
z7hwYgF6GRuF;%U>vLZJ<?_G6iTc0H-WB$YrpX8X>UmY+_UY<WYyl&POOMJj5eoJ+A
zxE@k^Dgd7wh}MAkP`=ReAtWXao#0E>L>l7c>STlQGz{_bWG52`SL&J@W8HZN!8~)(
zR0?oYe{+`6j_QG7yd@uLjpImFD>%3y^p0Zwz9AL=5(#SxiI0z&e38RS%*Ph{Wp|tB
z`S_7y6bJnB5YX+R?QRn9{QIOz(7~$N?i+DjtHFl_B%Yra)Ze|kQnx}t=S^6)b?r9!
z|3|8>R==9Tx`+V7WV)-oq2>o%&U08&5;o9zOk}VYr+1`uLjnRqp`8;5wD%NNI34b|
zW@z3<W7H`hu+_~}^M5y88e+;}H`)4Il|^lv@m10w3H8%}eyF)nP}B7Vch@2HA7ue9
zBmEQsx<@T~ggN4%G<v%bLCQ`~iSHgF7Y%lD96<;ip}Ti`$0$7lZk=YMk@Sr5WPe=>
zE6_Cg@wZh1cNpbM1${<#C<XqyGasO-|086+bVCtIkdX=^lp@SY!&wmpZzlbx+Fwq_
zs1PvU^;S<x?u2Evy<<(m-*>we4~t~W0^wJ;oRpzQZO`pk#YR#R%`Enpg%W?R_WA&W
zHY?PdF_%SFn|Pg2u4`)SG_Yo;+ywGQ^8Dkmi-Cj}{WXffzy9U9@26_Rox7G}gEl4M
zUf06R6O7XpI}K?D4JP=8)+N!w%F9wdxH^k8#B*~xH|FY=&Co#uyS1PZV<J?Y<k+5p
zdQnaT9WKERfsPI?uC&tuo_AgYSXK+-bHciYW4K3JD%%d|GeUe02BM)THG)qg(UD1?
z*SMf=^0~0uhHN9gpa#8HP(<9PXZk@Oc-7t$u}oZgKkP@I4b-F|sOc|}=b@j?@d~_H
zU-`FJN<_Dz9C+y9ba<uigADX(_iscq#tyPCsC5YAq|TyUYeb5Nn*`D<xCKuBVP>>F
zY7(B!*3&#+MO)piDe3;@9ZHWLFtBL<b6(h}8DfRNv#o3D{q_7Zv0Z$VGLyWc8qTMI
z!N@OI{-{{Ilm<S@{r`P!Rb_9x`PDXv>N$bB`a3<t@PgXuXq)UZ*PC9D!!KdC<mWG=
z=j;G@%x^y*03-H!X#LHPWCehgQKg0i0#GIb9fAvndcfIU&k|G;yEB&}5zJ_9TZ8S>
z0w_l$&99^W&|yU&29+YBJuG5Q&SCkg%S?ba8CB40PDuDw@LH%t&=9o=aFJA9L9liu
z%T|_)bGT(f+FLaBl4K~|E<(*MII!#!*w`m;0t2TU-2sQ1R2G#e#wMDH{k#0PqflQ&
zBjyECu7{r9A__dxJWv>3?_XSUI5U%*;GK}Ou`cETguy8rQ-_b-_qNL&S!vsciq5;<
zyWHZeG3Lt^*qmpP52O*b{+$mrmTk~?nJj4U=>%5&ND6}5&*4y|zzV_?$lRC~JkyoI
zk$mJv|FT?J={SOET#l~oXU}-qeMEIvA2Jm}l<}my3N08u+0@|YH&usu@A9{DC)@@{
zE&sM16r0L}tJWNer_<6AJ}yohr?`|$NaTf`q~S+Y7N?0XWm@Zcu}E}P^(fXiD-+Sy
zjgR~8_ImW+&odgq<@sy&Nh)@%d!dud%AW-`^tYb)_neMCLQ46R7Vn?myi86{?Xf$0
zd>Yj;Ksr%oOvqKQ6GzOi^m9s!NiPDOVGkv*qEBiqK^qq+DWN0()~cQaH{ohQFy5CP
zSA4z+1W<0;&w`ypEZXdes$6pJK3SCbJw9S`uIKgXqz>(c<;QM5kX-W2-T2)`zZjcv
zv@~LF@zx?zMl@kjMyh?DV|pNka%OYPn#Va$ho`Aa`Tv&SP6Z#Rlx$?`rkQ-!@YFYY
zaXs!w0&G`M+w14tn(e)m1)XpI{_Y8aN4l--p|@c;>n6KZn``sW7+P>(t9vq3Ds~0#
zs&ol&y{%H*J!lm6>pX0gKnLz73lRY@{2wj=`0kwejd|lZJcIT1ax2+YVkN4QjPI8c
zW9wOP%6o2VJN}-ye$))0@Y5fpC`*hEW9Dq7viG2s|C0*;1oihBl#iJcY+fBqyAvVO
z!vpYflNg|%5U2`QcUc}dh0a|wNQ7rlmSFO&qzG*~$&6Si-m4G=;&?4)zi#8X?&uCW
z2mEx)aD+Gm^}qsy)t+nw#}{RrI8RHjSiF2~|L?}4XQ`OP)n`e0yXJvxE2~_;r2Q<N
zIs|rXPz@`)hw9r@H!Z4Rc3((xGRHkNel7-kb;)$J3rzlU*zLRP8_aaHA;H=qS}1%x
zKHI<U<uu7M@O*iEN~DDGKEbaoF~UD2hKUtNKbSluG+4b2m60o=Zxl|3k1swurCx;y
zSb6(|*s$%Ck$X5kwm+i8<3yr1YK11*)s<aDYb&Gv#wu=6wYQS1_vqwj%VF$f@Xr1k
zBL1u&lb-qz!nfvbv^aF|uyJd64%b=S!HD?pfnFi~9A>iRl?yAp8S`%B-M~CaD{7NF
zDzXB8d@Ok*cF_n2(#Qdu-VWCuJAQ=dOJPND#7?XRDPvk#NV2?TdZUPx6=nlD(tB-T
zY)5Al_#W@=4xXdvuhAWTw0}H(+QwO;cKo1lcg30Qk`uYB8%JCNpDhaCy&ht*);8hU
z?U%bz+nYsH(w0?nfpI+Q{6Zs<v0jg*moI<O@!dxMZ5B<9eE8Kp-=U0u{GfKYbmhx>
zzvB16MHp7a8P*mTUdyU5TXtv?@CRvtt!sAqjKW2HYEKVE&}#?)Yj5J(ne9=jAevSF
z*HH19Q$n;4f1&VV#aL$bq=w~3OZN^xqk%FXI<?1!+o*$Z^wKnPYOFl*ts(3L;*gnV
zG4AWCM-eK)|IabmQ4hb;v-Hq;ju`$a;iKx)_6(+hpIM8Ahn|$H$<ogxgLWm2)m0SO
zppa;7=U#P4@4)|M0fM2j7shkGr9YTd7;pIte820xoD1;3KiX28BCgKdfA{sPcres<
zyT)p}r8nH9>vI{7*+avzeTx^y-TaSm!ZMsONWGu!%V`#Y+i>CDhZprz-IusTt2)Ct
zrOJ-4ByUY^I_|CREWKPJ5+A3G@{m5Pf{OkaE0UJ7#1O^T4=rT)QMQ2KpDVN|>v?(?
zxN+Xnet5V;>bTFK?3v{>-QqTvxoVyN`#UkvFx7FyVT8Wjv%pbM=_KIcnA#=jvCHCg
zDns|0M1CXa_K=pO#Bb_tyT4|wQFi*D?brl%j&ZUZx~Th{D7kJvn_#g<>8#as$ux%$
zz!j%lia!+J;cEg#$=1==f1r|ey`e|0U}wlk<Qb0nV$8+u%clulADE%>&OjpWiT(RE
zuD9DyNd*$vs4NER6jM_j-$Z;p#iI<^?B100h4fTU#7aNVb@HHX{T_t%E!{2+W;>$Y
z;<Kb;HO`fRw}yg9wWun;<EG=Ahk$8UG&c5QFv<`Im7XJ!?cD`@ZMwpkzLJOjd-pi@
zMb;W_t0R6=A-F`_4)PM7>zXBYI@xQEUL19uPRrO->0Kh4irYIBXor+uuABM3I)D}>
zq(vj;&Wbz}2SXBgI{h)^`qE#>VKPse5-&4rD8TSD)BpwG@Qo5gy~|z`1GES$IV6Vn
z_Q?Ng1&pJzXXM)Tt@?QTu*%aWmQfW}J#d?Gw>jQZ0qbY@t6uBfx3ydWgu<~LN{E2w
ztY-<R&C3@XI2&!%l2MsuTCTQ31=g26X1`hMlV*dFN4<r<>7W8vbJ~@K?X3V{H8f|U
z+w>BsA)b}C<BB<f^+oX<w)};ZE~k+#jf8w%+KX%g5`iLfi#6YxaLHF<+MVea+_$y<
z8U#_IPYYt3QMgO0qx&{OS8}VyPwtuR!@_>V!mduS<h$d;A+u%F_EW?6vyU*$#;vQ-
zDraxzT0Bd{i5kLTXDof!kf$wwcbBE@Z4SdK!VLX`((y2g0UnvXz=fsxr>9>#1`q4k
zTnIAUeKDTg@TYxormD-LmN<b3<gY?Opuszo{Q<`VFH>)BN8x`JyTw<kbSG0~H#$29
z6$V=u)V^7*D+e-JgZ)Mv7SHpr-xj?io*IJjjW_JFrzGJPeHCL7BVuM{*OU~ZPAjm3
z+%a}&qf0gv_1|zz5Ts$xsYhod_<8Y&%`a2`)=Q5p!N6uv+i}-t@~g0&yPeZ06J1#p
z%a_k@HZ>Gg%`7`Kvg$O4e6B9)8zIwY^j|xQvQD)2{RW#$vlX%%Lmo{ZH`64?$4RD6
z#)36#z3{M(AQHB1wE6Kd(tRnNYhKu3JbB2>tMh^|&YGix=mY4TA>;wvbRUW<RsE%h
z9{<DF)q8Lx4$seIutQ(DL`$YBsq}vT4$@a|2#%oA)syXW)Z@bjRj2)YDf{-pZ%4vF
zA7bxuqBU+@fVOI9RBlYNUGVW2M?k#e$Jv#z&5HPi_(|(0t0fCM2^+}|T<?OCLMX4J
zBpnDCraMc^(%m<@T-ux~eDY}Icw^2Oy-d$@eQHxUjVUTTKo84~qGQ_f?&h~&CM!Zv
znB}Dhcko-e#w$S-nzkQ&u{~Ij-L-|6#&D@at0^g>)6xpl7nd6*wzpgo%#AhXY&<1Y
zi;0GX*gyNhoXpy&+{`)s(6@B-=Q8?h0vw7u7;vl6EBtQW0EXv1+CV2{b&UPs<$2yD
zk<PjDxQ7KC{g>Xhc-y&4jnXb3$^QAB-0?LFq^iiR+B65cuqPWxcTzb&+v*GYlHoAw
z(BbbQbyjA!3IaH&O{05`sTe8QW>xYf!)XA7fh@GPGgd4a>Hqk<cSL`YJ5&v(1^-2{
z!W$Id)7n=$LBexYEBhDaaI)+Hbav6i6d598ZoEbc$mMA|laAVI^mW&7T-|jlpApiu
z|I<h6Cqc<cYsVAg(dy9z<#ch$^401wrxIEj7{FEU@z-ZaAas#kT>2kkY*{^Unlc#1
z-Y(SAl}l?^?I}Sv=Y#n(R#Q4VakK6un%!_>kF%3}-c82NIFZiQy2NsRG5W5iMi{Qp
z_Qt5-laU|~uF+lJLz=>*2cq}C2pN0vx!j9zJWxDzZ84;=+7sqNBTtJ=KNs2btEQ07
zE-uzDGsTI(0LPMn2XX&Bf+8MW@a=N8hB0JCJzL8o=TV=eS$(cTC5$4WhJm&Imo~&g
z+2z5z5*9ufPv?q35s9}~amy&Xsh1B|Kl~>ALHRCu@raqlE>BLq+F{|<Nhmp-HLlc&
zRyjkD;A2xIhkP^_I;RlEPTCbmHcq=3px4U|SM^or&7IJUHobl(7yqu$uU9xxH(eyQ
zu%?cs&VpRnKK9&dl~FSkVj4VfZwz0-s-?iI`x?fe@Q~mAgg6DfnE~!g-e2HC`CI4`
zOH^f`#CN7RX#3M`h(`GI+_KL!<mvAM>UGhTU}b0$^T)hs1~D3?<3bPxnu#HT>^T<4
zG{l*2GFt=&xDB4w73>$9?SV6|LpOTFo2#JrV+C6~{wj#7YzEzHtkgVspK#mT`~~ib
zvqWnnev-~PG)wp%iv!1}B<|L?STqZQp$9SUI%F~(uRJZaj8dBh_no4kvT5I#e&beP
z!9(<gN0=CrQQ@tP(85F9g<aKxZBX)2(Tw^^(Yu4$DWvt^<8jMHC%L1Q<kup5tu6GQ
z^WMJAar)IOp^qo?=i2+!q}0wP^(fu@Dcdj9<}!A9NZ?<##905mAj}x)N!<()8BJN@
zfE%b9-#H3A>I5V;%4_8eL!3}<G&%E8bEFZ$;3u#7U7oU5`P>^X+&HHcM={}dkP8_V
z9M9uoZnB&JIyJnODS`^R4m`bcQ~MuQtZ50pL!UGk(^vj>AherQ3dX$&@$7s$=@$sm
zV_nVFpydaTl$9?uF`V2$oZzbzj2jDj<x)a*p-Y)GZ)ol1q^<|zqQH9Ebgu__)cwtE
zFq+T#e<^P>z)j8sWkdrH`#|Mhds(nD{SFdcsH|ivCjtS)bZ`BTLTm2fZdH`s9X@z3
zbK?UDiB0{@f9*8$Acv(}>&<q=yDj*Wla;pk?Wl``d!PDGW*V~;qN>Q=)35^ezos!S
z1#58+$kbYv^FY0WpL+2{$){dMtw5}IrE+~*w-*-w7THAtmUC&@l1C~|<Kr#yDEcRA
zl~szyc<EpZB@TyKb>d%Tc5F(B1gWB-@)2>*bj>OhTEy)cM~yR4SVzGXxK0_ZG;_uK
zPs@LF9ajVlOKDu2l^}h{Zw<JnzBIHoMyEdz^ZmAf$7yq&NZO3*U<bA<u|UB|bGtvQ
zC@>0Z{#)PezYy#*$&9JJgt1QgrntIzt5+_YO`DPNZAdJx;_AK%0D&iP#443lRLL0j
zX?r@X61n1{UKcxUEhdP7%L;)OUd`)wE37eIf>qJZ(gM8UMCbDUC>5$GH*HDsvG2)8
zWTuKL1>B9k_FHF5m5d6zwoP4-(}HtUUoDoiKJz&~8Rx@E5q%*)YA*eevaQgGr%ScF
zd4G@Pdb0c*GV)6awpCN@t(|v|=rnk&LmT~>1M#VC!$LUHlk_(BBn-vHl#=?BVzCie
zp=YfJVVrX#=|R?^mSnvru!;b8JATT?3YIaOv@nr~W1@|+g!OBE<JxEi)?yc>v{(!6
z59Qa4#z9qGB|Tx<&M;KK`$4uLn#<>y2Uy31w1EJ3k7sEBe5&V^0?PF}P`m;iFI;ZD
z(r&5E^dY-EhV_f3F%Yopx*7@siz3ml)h5Tg&y`P@34Z(c7#zEb4BVL}k3l^y>aMbx
zPs3w7>*y{RLl@-yo(YCl4D@9cgD}m=o2^y@P+=1NTlT3ocT+1X0650|<%r<;p2rFW
zXT?ttSnp~jRRuM9X2D$|IJpptzICjk3_0yGyuKD4kbHyKK02Ot`*l3Y1K&0H=laKe
zCa-tI;`y|Fp;k}a;rIX>n!47regK*VIiuc(F&SF-ZH`DV27i{18}h4W(0y=C!%`T_
zF|xVk#$l+Mr*!zogyeFq;LO2u<MF$I>NplC_9UCoS9fG-bICdfEtMtVsr@59WO4SR
z_073gCp6}6wN$aqpDL=vGGVra^2Dt${cXwx%DN|<kKQz1+xf19wmv-lzp1D3)>ONW
zm?3q_uMyi(dT-5x7G@mP)T}0&7#<6M-}0ZKMJ&AymYN_lvry(oA&1Ks8uim8=-Q<^
zM-DrQVCAvbSktNS>K|y;9pta?ST{q%_d}o%vr2Z__XcSXB@-t4mxNZFjX62^Djgzg
z1DA%o?_l{q`LK2`;|k@nFl0C7YPp$5g$2@jEqN%s5Xor$d%f(kVZ)pujZr9)rA&tT
zraEq@LE`u=E@wBm;-Jl^Snh#;{WFa3{PYFabD2){_%V|ho7m=$Pd?k(JPI6kCNoC!
za8Dqk<-SDSO__sixXaMPScD<Av2krTO9^XDXgBP6#9*ctI%)|&kwuj^l{7duSD?!U
zGIjTxOe^=($HktTr&oFzkBI4RKZZqOCTby3vw?jU=ACb_(Mo!`7!rG0PC4Uf;FbPj
zR3i9<7eti)mItp{qkB?l!rbgbe~Vt4zi#{OH$v{j-s83T2_(Vr2B)@FJ<9bfx!@^E
z=HH*!|I~+~y(7VSXbEap!?_(Lo#%a8WsYl~*&nF~DV6^gs-F1r$<eu}Gm|<Lb$7Av
zT(clXp3hNvhSGDeke#5=qZ$5w0x*9{jj80GBp$osE3;f_>zOiGKm>$b{o`M%o0SoO
zXVJb^wCAV%sJfSfkE0HP(90JPp#I*Xc&s0OY4G?C=p2Hyzp5YYYcGUA=ctdjo&ayp
zsNON9{SyP@@n2@S$Fk(t5ySfi_V_w5B;Yz+{k>JzM*uR!M<^opI{F-LTVd-ZftB3P
zI<aerfcpc7JagL_-w?!$*x9SCfkJn^4UnfXZYCimY0iY+iva6cp)?4R?%Nxv0Klrh
z*M>q#pi$)}5j*#UDx}=GyAZkjpDN<mz|(iys{E~@Uj9f2ef(BmULueq|K8RG@}*Y?
zkq`BNaCY_hVuSV(p7<wHO)vMu6B8<EP0aN$stMW2sq5MNs<!vtI6_g+h~C~Oy~v)(
zj+9N}=kw1x4`j==Dn@EfUbUZPBvI;P5*SeRzx+#O`MFen|KZJ7TydfR?k`K0qUwrK
zs&yav5M=w}mKb5lk24NuUV>yn@n>7?%ayvYUYH;#J#XRyO<|Dz>-kphvL!W?$OxP0
zZ&<-Brp`xxmf7AD{NlHeG_-c*6|9<Sg;|(B8EFKLXX*-7LK_}Z4oOeZ4o}#bX(jRI
zZLPmI@c<ixvhvo-H36$8s=G@*g>dlAOikAefl#yRCIoU_{Y}W*TMHHUbRTJrAfgxD
z$fgn|6*jwWkj^T_KcdMvt`TIt%8J=(BaU3=pn}~<GPPG^mVtHUL7r#<y$3oRszcS<
z9|cdg<-!x}-HkHOoKTyijRd?$c=7py?sbcmI|h#~of0*UJktAm?p3gc-$Qp^U5omh
zT^hD8e{%~cWEKDJa<MzsPvFY@%u!f*{vPd5!(LL`fk-pZJ-@%xmaNux`(7ed(&pi>
z8+fO|d*(N-r0L-lUR=eaWrmMbNq(y}KO#MifxdWbMj^4z>iw%bkHHw&xz<+apFpa4
ze*Eu<aP6|2SI^u1g^EVBZhcI{<zxiC8l@-UKv_&l*#`gqc3!M1-$n<l`w|^1qfaPV
z+V(7#X5gP<w}oQ%CUja}_N<!^bE#0Z-_`v!wNjP(A4Z$9#iOs@p~l?>eBi>W$GN`z
zY%keMs8)4z>ht;JsM3%G)zz$LIP^PBRGXoQ0YQgnZYb0#hq+~+B!9H7yoxB?8fS9Y
zHc7iQiRQn9K}Nk)_|pcz(s!PdnWHFkt3Le`yEb=Pz%WAb*!_CyaV)gM3k5j4g_#UH
z@2;-oM4Z)|aRPw34eKYstny68X@e>|JADgF{gQwS(u$b}^~dNu<`DIo5he47y_vu9
z(r}e(W1GO9`-f_%c0-Y(y&2{A<l7>>yraSC_i@`krX453Vqlq-B6h*!)DHSf=n`j4
zB+?6(aQ;madp#hiRPl8rVIrSvsUh`Wvl;#3TUbD-<jsW<s<_aaW8c%Clq+I|0XWNc
zPtkK`ALA(02!+L&jASBx%mQp);A%rc4i?xDgQ3uc=ufD7t?t|72y8*b?pnXdm<})Y
zm-cuG*nW_v!!P${?M=qjr{1o-*UJ4v(aOgw`*iBHO+QC;K7DS)r(}9Dz4aBpL`-gA
ziiyWdCRS)wLiXy0P-pG{QjhM*W$2^)Z%;a2j!ctWLY4mw!JjuLp_AtNk5Y?rS^g-q
zyPGu4f8@Gf$)z9X&am9ki5eb1=kzWuOs}$7@m4IaKI7e5l-#=900Vq{NnORND9om-
z?>FmEb$mYQ>?aPPb;i~%hMe5EemcMR24qNO#y$h(EG?rMjciD?@^<Pn)yhh`inJUf
zc4{;Hl{ZJ#Zj?S{S{uW1;fzk3wcATp*;*ASDmt}ZE^iiw0~-7aJz_a1vms{NG*93X
zx#-0ivVQiQ+3oMNRJ=K9DL-;>!TQObh)%ll3FOIp^b@!ZYo-KpN%=-h&z_kVBae%y
zv#s;0_f~}ZaTa66_k6AN_cDa~!N!B}Hp+0pRiV<~gP&unX1UJiAbP#f`x~9RSUcU-
z-vhKjR!3b3B(v?ucK-Hw0xatwdAv-g=ZVV&gDQ>J4ypcD;+FZ=L*nr)&2jp!_`1r$
zXga+6H>FoEx#sF%`v~v%m=)JXUbg6k<B9+mr?;8!uK!R#Y^|X-_Y3udT)(t)Z0t6D
zKAYKfx!Vc4T<z*54f~2Rb@*vmQ%}|KO%RW#9ivzAu{&H$d_qTSXm$z5APZCUb_%wc
z%fsvJD(`{Y)hIq^`$bATLhp>^MkSo0W*S>X1D0V>N_E1H4mb+1KEMja26C*dd0)<A
zpNW8VLflt$m}hpljohrY-=JW;G1K3{^0sTcdPRgQQI0u5CXDq>(L9TiZ>JVgSAp0A
zv{j?fKK<RJC*oP_FI+W6hL;!2>VPHs`y(agTon5!Hb{rE-kz@_S2fEIwO)ACv6SQw
z!o#6e#G_x0IdeZpI={xk6Kgh9B&vlI?tb72-Q}DO1PuN`!T7Kc+jbL9<nGvSur1B4
z#{D8>d@^`VljU{9mh~PgQtgxd8u~OolyLIEzG6FTM*Y3+kjLN{!Kb>pUJ{d1_~!)F
z=HfyKf$U6-_T7wj=Jq`@iAxMT%AKWRKizp2*x)5IplJxyF_2@V{TGYZoY&}sL3As;
zUf9u?%N8&m3?Vze_9{z7<6o2>L*=s58~;-j`pb6-)-bjL@%q>=<UH*1S)9<Y<@D-Z
zLF!Ik(_Uiu7?Ziqgs-{jh+K$VAm96eT+@9k*#e?UtzBREN6tkx#@}L%tCY#!Zo{73
zq0?w$yNj_rBD60Z9_>_=|6)E+YDBA9@t<yuxfZBy{%p-|m$MVfHySnJtK+Mp5Y82G
zT65MFVL4rUkV$Rlw+<Pk;5&~Ag4t@`g?WfM!J@yCu`)USR@j@<)ki2aCBpjU`TgPd
zHK<BG()P0t$eIQiK`Rk{X>as1RR&hnI=!~aPG|{;=@KGeg5A%L3K+MSWXB5#=r2Gj
zIueaUdFaFXAhhQzB<t!qu*b(X#WcT)Ip3+s?5~uK;Uwm@7EoVb;=ZxU_0&stL1*0T
zdC^$b)8*jn5m1Ns;}2K-^Q%JsY5ubW`Q*)*4<iH?G)RC(TsoSN2XpMR8P&%AsfD_g
zvGw(Y7N&XBk;;Yj7mp)C!!(PF>J#kJID9X2QQ@2Jm{<DRfg)X|yniBI+LkcswW9yK
zfn<BWet^M2F8AjgI}{*Oea#m}s>nMrz%rBdc8opa?X9whbiNGI7?qGZ$XBR#$+yE=
zr<ky-u)TJ<VaQc=g9?ozFJge<{W>WKP!oN52dr-*@VXyvJ)XZDzdB?ueSSAIKxgWM
zYu8LrYV&JaQD4jh1F-+(db`vfy0xP*Q)Lw+=rM7(Jr@H-2fq0{AOY*eej8>0kS#RW
zivBz^bo}}3bm=vec7bC>_M3^e^NPg^Nt{C!5vIRg5<F!Wv70mWRN+PW29(Ht%vkjz
zzS^$hcI$P%G=>12-gWu_mfoBAyKcwCH~!z6dnkgQyuPW~uYqDr$ujot50OO}Jx~NZ
z5zY50@oqlQZOQ^N`PFKC3{15_rWj1s2}+m@MJ>IzYfS-Kqma$(odLqzVIrJtu{y+9
z>Q(=R(f|nzC%wlD)Z0y5gHbMotmZb93?%Uvx;-`fiOJA6^Ht$o0mk_DF72^qmD@W$
z=6igmtvFWWzDI-E&F~Vkw-9oM8Tnol>+sTBC*GfiEpxhr=u8b?^Y9=L0eA{+S7ezk
z0R;A&KT>wnQ_qLRTkd0Elrbij9ic`|Nk8)wF7~vAAVt{PP%I06>iF#M3r`6SI<`Y+
z@A46Tq4<LSiF@x7W}EN5tsMVkgu=GTjH1oIk4!chs=?vyVo`K1*t(15_5!t{d#>Zn
zQqk-_U8lC0;;2O4y=ff;2Tx?C1O4>hme{Q#rY>gtB>CrjoN_Ijr3DFGOCiy|S^@u3
zUnXEp5)cZ`8S}f#i!utg*3SjGmH&~mrt0QjHn!n{3n}>7{-}QHdzgI55_k7q-$u+q
zq@Sy>G0>1#YX3Xj`G6ZXJ#~K=n6p!XGGAtoO*^)hEJ!P3)ynEb*5Q5Nl00Z<b8lSP
zeTVgpnq^?nZgRlS;!ofh%|fu>)j8u`&-+L*xcmu~e#Sp8KYK1yo%XYo*Iy7KA>kpM
zv=nK9BIBDSQtLw0C6CLkUTX7nEP0gs&a7W^?P~<S`8?@=M^4;E4_k|?to1W{e@w>K
z&J<!6)+_$nezm+rcYj{I5qYVfh-{uaB_U5<npVm*^omb{c*pw0kkzvPu8ec}P=Hk0
zIp*#g%*y^dP;{Z}sjzWv$5Bsi_}kBP@0p#qSCoC<M%o%;<6&dZ$+W0I+k<>mir*t0
z<Hn267Ey~!bo3DH2co1|gqx><{>B$SFSl?~-%vM@n3L?q-?{Xt;RCTxfcm~Zn%Kv~
z=Nk=;YC%sZ>(kz9-dMk{OTtjafGt=9I+oy>ZP4(y<rot?>zJR$&nW?JW9``CU(-ei
zRJD!9{kUw|X5X(aDwx|618RSzo{2a-`>bj}(8#ks0l}C$7=ff9CWU0JZ31@%`T`q0
zJ`^YFU&%?yrzYB4;a+a}C}+j2JNnU*t;HgD@r<o!ck!9Nx*pqKUF`MdgzAJM4W8G<
zA2)_b8~9LxPG#zq%#tL2CH)7flIs6T&IcX!pZs1P6UUhP8+iwpS<Ns9Zt~}nIDR#r
z*VXe7Z%dDhxmh$b{rrJhh{k4)aOG2*I^SQi@Q%l_92Q5TGa}uWI2Y_%R!bhdg8L-G
zSBqEu(A3(r(qvCLpJ)H~A$>0Q(+4@(@vvWLN@M_9@UG|tmVXG8xbJk=oD+o>mvOX-
zY;u;iSU)5C$0dvKi?*i=;hl_Z3eT^l2V=LTOE@rmY13D6;FU1SE3mp6qf?_L@4T`q
zcf>o@YeiKvmTjY)e^3q!opVpQ5N@Un(_j1f(osm<7Gf3l*vNqmT!nk9gWZ1NZ6ERj
z!Il70sh1)cudrXp91H8~m^q5~6)<V=gq{X$z{ci^ydMzDcbxR^jrs6Ly{n|1!8`fN
zlA;FhpYMFU&ZgvVjmW68(z}$`8d0I=KixLMTiy8a9rCG58+#(bujaq9!Jrkw<_Ukc
zL~D!5X14szcWIN_r4h^`pl~ald2QuOa|lXaW(IYm!`tHKl$5ogYq^S)cLEOG2?FkF
zI_(Ecw@;?8JFr;puQk_TzE@=7Xgd`h=Y}vHWqwri{YpU6!@Ux7QUkKX{~yT>n?A1<
zOQ)hRggqTQoT7;>^<?^Y$R<4?>2b|O>zWM!$3?9mU@YsN6IiG9NwPW4_KO=@yJcHC
z_47Ud41i(O6&scx+;)=zKzX@a;O9z(y>FTyuEK2cidrlJkX7#S!Q-&B7zD5?Rqq<^
zh~?{L#+-U3@?O~tieJw`93=F8g7eW(2u`ZJxL6_D?OlCSQ&G$OS$<CxR!FnS3k`m2
zjI=vj#B%pz?ft3fUU`mO#zr-qTQ`|zYXGn;+&$+u;In|I*z-k!I5aQp*DRP6BwLaC
zXxV9TK_}HFJ5Q_6E9Ux7KzFtr9a!WC-JWLr?Sp;qLVQ?o%lq?H$K#ziN)lP|bxY&>
z*W&@b`0H~QsLBVzt+G>~>MO(fx%NI=>jm~IZrhKQ$9@Xe+D<%+6jvvrM-9<jH6^o7
zQ8hv;_<Yf~Yp%T6oG%N$R?_^c2xg)Ey5^susQ&Txm4U=e&Nsa~@bH02>*HEziNsAA
zetSOnP>O!8v(Td;qWvNh<KKdpEX}`j=ylXyIMTJa`@PYq6&pK5jJD8Dz@{mep4+j0
z)op4<Kzu3hKf0FEQNQ0Y6@)arxb~Z(AEx`&4EaI)!%rW)JdTbG^Tn^e`6pk8jCpL8
zVT*Uf3f4GRr*&2B#rR50-qf|-FD|UOXpoq#KJ*Ih@EI39K4>E56o`Y8l!OGH-$aoa
zB*)Lrbul=P5U$%ip)6Q=%rI@^`C8C4#ThdB&{?XDm5ui63-(Vv<5a`FkScMTR-SsS
zaDM%GzRkZqQF<+1s(F6{uVZfCou6)-t3H>x(k#J+GjI5Oa8hsv7z|5T)-^^5<Mc&c
zckh4izf`|Rd3tyz_u$|j^Qo?d=*NU4^3n!aKcUA@bCZTxdo@ct=_<3#o^d6<Z(r#^
z6)e540Nl0i6NF!0Ri)U`Xj-}kT?DVk3i<!b0+`KS4m1L7u^Ma=8QZWLU(t76Dm@a+
zxIGId4Mtoar#nTH=XYD2vhZ&{2tI+65y(kgs}-T&3}H|I4*)^`zMi<zzVYnW006)A
z$zSJkl0onH{{wsP-+dRC{|iChuN`FM|C=Yi4gfgv!4CoeE_~!sP*ve`_p`TtN7}98
zxOM=W_jA1d{M;7+0Dtg%zsJ`lr+?wt%l<C$+JE99-sSJD^Q?(jTW3wcmA6iDy;+v`
zGatT}<$H%U3yx9^#F@7^-@(c`=JVOlvc}_E-{HLmmQORfUOaq&f0tOJuyTdVFE2B?
zJ=B#9g`IdoLu|Wr<l~fCh>+&jfjAEUrw(K<Qpcf?Ajw&+yZ;dY^^&g{0ug|Z-M0rb
z9fw0pv)DZ^+vsY{S&U$H%V8y=fEWdlYFxiZdUcJ%i!-BZ0*>yM`pn>aXSa6#d-T9O
z;3#B>WHCGv_q!*mW<o*=A^rDW{ZdaMe!311D|%-Ti1ypsA|p{e(r=qQG*5sdzF)ar
z*L<czMLrQGBtqEGNGQk0=Fo4;7Z9168(zu7kZywTwpawTZNL-TxV3JPFbuCjU+<e_
z#Psc}#=zuGPc%LV<5<?lU3bVtCbw%S60_{K1^LVQ7beX7_wamy2X_>Ber<g1!SU4r
z4$KsOJiSIO)|gO?JOx;39)nDDjB2_9f=-_^t1++{_nOZ)Q(;~<i(l%v4-=5Bs`y5E
zhTl}4rjJpblrPd@ZGJQA<fnmml-~CmfS79MKr+JA;fa^Ri1Ux?R^(3WE28UGi}bBm
zzW;#4-}P|0f#@sqMEUv0`W`4>&p$mM>Z$k!5gMDnFK*g#*84w;*!>%vwEO5?OJVYl
z+qFnv+c9!C+ewc{)&xX80mm<Fz2o3VmS!>2O~)a~(OxI}&X4K~kU8HdpBtAPnd^-S
zW7PVie$>VjXAHWr_HUjhVDgXsd%<oTdSPt{8qd&!3k7y|91eEIpB`RZE7Qz?Il2{8
zB)2JO0Q7b&i0xcg8fvm%6IIdyiDRFV3y&P;40U0bE-tT{gXu0eap@9&uvb^b20Y6T
z*PZu&7y$5{7ud^xeT@;`RLSx8Pe1n^0Kh;J5v(n-_xv|M!}0s`-(z2aXIB4L0Kkd2
zJ_D)%S8=;bT}J3MD1;B&(2u?J4glcpqwEuK*PV9&0M49d{_4B%SKoV;<&wSZ%MG^S
z?>WeM12$J#CH>U1TyAxRmlRWy!@n!u`y{>+i?X}lvoF5I{KO0YF3DiHA<+QW1o=-(
zK>v{k4g&yY*F!QM=Sb#n@?P8n$<?pB{B;##`HXUmAbXN0f4U%E1jtg-$Gs0&Q$Vp1
zj5GsZ@IF^$3Pk4#ek*2uE~W%E*U3sJ#4S_wduVY6fA68a_&;CSI$03Fnt)^H*ILUc
z9o;kAgnmNGzF~_rHs#cm9d_a4;UtfMtu4>y9OtbKqP+v8{lMNV-A(&C-s(h3H5awl
zovB#G2%bWegY`Yo(e_A0Dk*=vF0)?F9)YKa=<BYB#{J*kU5jk|b=!cADnN3&5oJ)D
z8=?L284ugnqtRXit<4A0cG~7OYAK&#Gpw0W>%VfXKaum_-BS)bWU>6vLV>ea$2TFI
zS}XD34$Xm|-e;9aHDtl&tpG*Jnb*MP^w;sRrq$fS(2hhSYQW3c)GSPz27u7i9#~$s
zU0>H|P&7(Ht6L3$L;JYc=cwUJFuWV@FRDSPU3j6@H=t_x02qX=A4;vKABYAL=i^4)
zzxnX_y#w?^X88CIr9ZEl1FLB;4^T820|C(oGMWI3^c$Lj^w_hv8U`bGE_$Hhd(fzF
zGzA$=z~RMN{Jn>^&J*zTr8WG~3m5SEYJP{a!!upnvoxKV$(!vq<+BCb`HTx_@@hpp
z9$S_g9bx9gAG!WbQw3N?H?eeFaahf5?(cgHc#mJ74E=v#A=v)!nYWu*k4EpM<Kb<(
zG4gkNT5Za5>{kvlgXzBWKBJI%q80ILAg!+_nd(0MIA;P}tr?LY+`*cJFFwx5`#1i_
zF9QI2EBrD4+)w=~0N|CkSrc&n3TqH{@BU=~z`Y;;Q2@ZdzxrnYfS-NkUvNKWUIqXx
zKYj;LIehCi_O3sDfinc2e}fVEtoyqFfPeZMKMMf(;C}WwSlQr*_dmVF`3e5wJFfu%
z?!SY52aev!j&Kj`<9?jE%;@tgU*Ww89(eRY0KmRQc8t4pm2K1aALO|BJ##!jfBYxM
z005u*#JvE3C+=nM_A^&yH$u^P9F}o%0u|4^%ExNsWCZF^sm}udfUqtaj6Ug}4giSI
z5v0yh0}i$L1o-s8iYy7BLeeM`rSp*@8b_$>Dj<7OFY~aKaUs8DjF8oaON?M+1nm)7
zU1R^Q8T{OXd-0VMc}>7plRp4Fx^E7CYh_7;5P!Ur+}aT`FQlKT%`$rCt!<GwKZ9lr
zw6-~mYyn;btcxZ_5x4Hfr#={3zDRCBKm?k1U=V9~&=U7+#8=r0^>L8<u6se~E}<TQ
z_`L?Ab`k1z<iXG;bngJobY`O!bQpARkxkKrsi`J}cS?Dw(1~|31n(Dbmex`(mD8@P
zl)2S3=>79HY@csS079MC&$_F(k-j$(S&%+q+WCZ52Nh0jj=unyapWic62N@N;XudX
z+?bAp=LQw7^eXI`cEO8XeQZn~0L{o}yKQkn&epAI*(oz{9_gF<4Mc2w1_s{eCurLy
zAuE>4gwNFGpJBQFWeb_cuD#lyutyIcE4NLb=SKVY!%76eRj353d+&W2VZTQ8J_D(5
zK&|-%^u;$w+V0L5GzXNl*YNW9n^MSi!>@rzEpqi@=sy<GW6SeaW6+EiFno%PT)z2s
z9|EgEXuO~eq;1*}Iv|UHnJqK{Iht{V#}3V>?V0JKsWP;hfOI0t$sbO)pND-8EFaUn
z{zG=$Ee%8RqqaA(O`Xi)O@_KT--<~g_Rwm>Gy&V)-i_1o2@&|w8HZ`}o_=6LEx_~3
zW6JOzjLU6mQX|&G-iMI$KeGhOke?W3rSaiAIhMRP#hLtSSyk?q=z_fmI3wQc1taCh
z?q+X(>N4VeV3WP$_eqAoC+^@6`?E(M2LPPza5TV=Kh3uI-TN4+f9O+>pe_l2eDeDM
zfU{?LW8t~kJxTk6l9B3;Db5@?H(dZIP)u?AJ9o0&hj;Puu35fbRxHo-BKPme0ha%e
zWIo&@zXv61qCcpBqQH(r-2UF3ta&I1K;f`7*OdX}aqF@;C`%x=*#hYzpss3OTu7Wd
z$$5Y$^?3et&M|U_tQ`G@K%qR<(W<|29<lp!kr%%shkjh1WP2m0<=AVT3i7IpTFa>>
zfCWY&{R?|f)VO152EX`$z4+4ci`iWS+vxE<bBR}fV)KS7F}QKBz!D_u%cD8&CogT_
z&RtWOE*vz|V#v6q{fCS%lKwjNk<jy^V(7$3C|U&gXauR>YLFyBLKK;j^pY*UWqtC<
z-xqI|mVV_?5m6YBRI?El2;OF^UiccNW6-ABd_x8~zF%z?C_;Llk^+WmQ^xuGm$tVh
zljyQXs?AHS?O9)iT=^VdWXchMmnKyIKe*6oe78TavxC2yAe!LhMv2eXQRc>6bch}1
zOh>cp-w_uM7GYb9pY>(&kCA(C)+YWwkJwn(s}$4v4M3v5{`;1`Z$G>m03`ybsQqq&
z1DU7>ezKE+P)&fz9O!?|?jU4-1L99L2{{{lW%!dt?7<5}&WHY_@)!hW_B613JiyH%
ze}BCJZPW7&O+b5j8)eYYFys)VVRvwm7XjmF0v<gW@)P(ex#4plzqv25k@NZU>-fss
z*YM?&S6Q8w@nINSLohOg&lRKiCwo1AJX$lk`WE>&5c$p43}oaOz2Xfw{NA>gp*OD&
zaCL&Mx)%$F!?R7aKjjEVXU5k8Y*aOluS}qccyAt)+a%QK8!t>d{$2ZFoRFqv#nF{b
zVV2_@^a#-cx`N*+596*W{@`B&BjY{ujHtb>Ml$hTQl$Os>>KdPZ~r9#;K=%i002M#
z$;VNbgeQ+Z0{~dNwt*f(PJaSW6prKMi7+d^3!6QDh%aB041Kb<KwoU#jsO%6Q!~@T
zhX5+Pm;KA{aHhYV_i`SDdFg-E<Glk0(QZ`a$mjU)4@!6&r69kp76FQ!og=^%l4l`A
z12}RFZzV4Px98w0nZ#;dJXCU>K%j!cR>FxKqH$<w0?6lmfFLzup|~BrQn&_RMI0Kp
zT?lg5l@$TFV`&z@G$cAe>4UpwusGL^RQ=K?IZ&IpZ%4+lEW`64r@WlU&adOEr#TAa
z^yUCxK6w?tcyy8dXJi~|<IB(`W)}8q=M&!7!0%QT+tTfNV)#3#nun?~c?4nz*|rFX
z5i^CTh+(fi5&Uzb&BT*$8EIMo_^636u7X75chmbY@O4Ef%2S_XZ`(KC{4o-aqbz*)
zE%Gz{sm)21U*&8LM+#LNx#~^nTzi!n?}oyr(eBoKUzudm1;_r!nhc5t_MTaXlN;k}
z6^^eC@PQqvlG#5GTRmTd;T_*5vdyTy|3))#kK4Y*BV$*Te{9r>)o>{J=X=!lei(>~
z8UvY8)Vm)xPk?FwG!?Sct<8rA1_=iZfHtOk5~LahMbQ3!WNhU&USDnGUQ<BnZ_rFA
zpR`|o46;4|%EoE}+I;|Wnt*0r0WFr1!#?@8Uxv@&C5{dlMi1_8=KW9jns&o`)c)x{
zc7DxI{z|7e2l&#-tE{%pA%C(R%ur^WS~@V~@@Va3`DHZ}Bk~&}&*T{cGZnL^Cv513
zzxd-Zn0P!|s8`iEF~LL)u&O^t5ANuS)9`q7s#jrU(rt)u_RM>m07R*pZ`OG02ViM|
zZRds9xaZ8?0lRjJE&T#d46|KE%DM~Qj(N6~@7t+Jo9W0d&Y1T{&vN|!x9{Z_;8VZ(
zMF7CbfA!}8fc;`xf9$?Ja5dqZPp_gxi1)7nMu2_xiq~EKW@cFK-8)#mslow;Xht0Q
zJreXepOegg`^0<y-u=9USd@ONuVHtZuP2$WNIXsjVUr8q{k&kyF9JBnzw-#aSH>C&
z@(UCPo_rO7>n?$M?8K&DiFZG0DIX*57EOQx)w-<VcO8UTc^bs&<skWB6X2@U_D*8D
zQKVD=IKR5JO>%d)gAeTI{1mN-Dw^bCfXzcif~3_~p~my)*6`JLSI|bf2>6Agi<mB)
zAVK0>{1n<Fb=OxpO#T7YNLU2x$0|9X=91p@FDyG!B<`8%3BxCeBG}@iK=vT^bwEJQ
z1M#JZ1+0^D{5vHthQ1G)@lEo1g?tx&VH)=r-q*oueC*x<qxQ#UX!DoMf-d+pSZ*!j
z9<}517Z9KHX$rO$b4xlt;kLi`Ocz+tC@_-=QR(Q6!}Aj)7CNy~;)x~K7=9nuujAJS
z`A5(MWHaGqs8P-!>B8JkBCj{RDYRYdVLdF&18aw`H23#!lm-9*YT~^E^i}alw9qe&
zJOFau?;hAd=#-WZkXe3}kJK}o1{>e+ylp#0^a0t6{Ko5pOgn+Uiqnovn{ofzo;EwC
zPk=vf_?hxUeF6-TeNqWHae3>+KacI3Yic?g|Cr+pGTv(PS312p!1J1n$RK|*_bE&E
z{@k!8!2g%ZH6rDUkBG#F9d9?~#JXv|2ivQo-VP~WnKbr4^7PM**Mke=?+pM<?vUu|
zwkF`FQfJ&e$NJc7-0^T90WNp-dS8ComcAl+rtgz@`m^UQ0sx8?j@N(f0)J4?i|t)^
zAM8uY*DrifcZK8g7bW!NPhR5P1y}Fq%zZ!lk3R(fxc-OF0RTS#Fx%q4-hCAS;o?d!
zh*}ib++;bI*Z6m3ndP{2g(rdUzR4PhAG?RW{dqS%V9;Yemsk0>SF+sK#mC^H^j{Jn
z`dtO}OU+sR3S94Txr>5w@7U1+T!Bs1RMd>=s>onR6yo(aG@e>k^s32d+y0KCoc*O}
z47^u9coCpJ0jvq2Mn*ZW2`JTe4?$1>)TaVrS30MN&}ag5&xS^3FR8~-62O}m*YL(l
zKK6gMaCr3a0wZbX5a-_afFeR%iZr_gcRC)X=g+MVo%{jdiNzUw;xKEX%E)KI`*3JK
zDUCt^SZ%BA0U`eYr%!0wvJ-021hglAQ-)}Pq$F~Sk3uFhSt~!wE*8fly!5(7>R}&r
zE;ucD0Gw>vpeoARXp+6{+@%_ew(iH{uCzVVcfGYHW(#Vw&2BZ~u?8<LtJN=UmN?rV
zU)?_A2=_#jdnLlzJu42+7FaCY*batQ%L=F0OYdOVu(?}JSKA<Lr3o<OXvfuBZnI?g
zdtZRM<>CZ1uBKubRfb=494RRm&ZozJ<1_$DB>|hFZU-U+Zjidq_-B#6x3!b!mz2xc
zRu!3gzScHgSmdoAM1v?Zz6fOGviSgr+RyuLnf8Mw!0>G#-K{kZp@SJsK*RRmj)+fO
z*m~2ShZbkpe3V1d4WEd14aW0l*YMSMuWi{D06eiYgU{W&v(fLIW+5L1kfZI@1PuQG
zneoZ=W7G~bKAn?(0Y}s*jIFu2QQB?|!-VX{j=L8+I#Xb(Wp->+d*44h*>rJYbsI7R
z-qr-%6dGHvqx*~xL;ik+Dqnz?WEZ^CD~zbV^~xy#z(a>V4FI@)?sWjbh9KXU-r`Jr
z-+lTx0HC+d8iGe2{r~{r@+S@f04}WXZiFkR+2;F=^NbF^^(1Ewy!#IJ4fvUlu;bW|
zy~@YGw)+GC;N`RLf~p!HyMGY?;mJ2yzGvUy-TBtezX<>+?_|W^>#NP5JNAZ5NSo(5
z<KR<I^Gl-M+?#8x2^c(aKLFs)-R#S-e2Oy`zH=XIGCp(%=NWiMG69~y%=I)w<Ki@-
za0GJlY6?Om0B}JAaPgXGG`vlb|5@*AMmB|Dj0jx~K;!-?<odS=UH1lX8s8u98Y#AV
zqTmzYSDun#(C!`J!0oE<!9dRUC657ndlO2zUetuEYXdxUaq9)ZV+R*xju$nCw4Wpw
zCyM549#dUG_sbeDo?F9ToE|>;KfXAFpSWii+q5&@^ki1*4t(F#M_}4y2J4ANVk(LA
zZ6fXjQAGXoJ|)Ul7o^^{j|8dv*VrHl<C>E4L@-6=kWD<kF+bkTFx0crM_R9|VYya=
zU=W^;bI~#9jRqp@ZkTo@luyU6_HxAc0ZmPU+#lQjMm_>#Yc}h$lp@6IZ=)%q>{Q-K
zyz~L!!5u}?jvp83y$jFXg9{yee}c*U)r|^2GN_U{BiwKPFJzcZ$KO*9gb#iGm^9kW
z`gRzFRd^0`*TKktYC9+%Um&Nki=W$eDvjpQpVw$2^I(PD7KMLm@%@`i6&ST!4S+s?
z`ndbu0u9nt1n*Ocl9t!=byChp2S|C%PhgRMZIE`zJZQfR8Sj6KfUPE=;VY0LeP`nB
zhi3M(`3wAcQ{H}QhD0UAZ}PDri2S~VX&hRdNl#kMge5ahFP<NL^8ehuyIPRxh=_b-
zucb`+#zbOI*y33zKcX00^KYq;9u7Gkm9PCEX1(Zh<FH*Ch?`x1jPfT}1`}=kv#W48
zJi9dpjK^!wG^Yk0mo};zZ{DacfEQ?c*xV+jj^^qf$eT1jB@>s=HO~*0dKa$sbOPev
zxvBJl7z|Pbp(~YVUS~xA!aG;_!lgF>0RQeUILH3^bMNx`1<nZgtru7$uzc<f0Kl#;
zf0U0udkO&XuO-vpH(%oGI}Whd|DnT2xZWBg>2vFh7=QO=-u>^{Mb5$hg*C4C%*j7x
ze&r^p0-k=6eF(718UR{*8zq1<Z?Oj8-BT?0V2yMBzy8K)9w(G20j$2me2bEixV}Hf
zU*G|~vXA+oa{(2AQ?GM--`~UI@SODf+7*`jjRTx9@FySVe(}n*_O@%1`_~(U_;qp)
zv!a4<G4d%$Y|ey2ey>YX>DdlApMaFEB(tDq1T2VvLUaEU`UAlT{`ufPh2&At`_4#T
zqZ*<;0pfLxb10-8M!s>B0{3I1H^8$OSc{aSqkHGP?LD!bE0AgV`0Mo=toHwl5!C*V
zFU{g7?%ByKJ+fooy7xg9Ya9^1TZ4GzPJ}5hQ;MjStwslU?_G1<B0<$CGyo{^ByMwt
zhh+i(r@kXWBOt$W(#wzZy3ffj7dq-|qV3mdegf&fH*INrzs+1&nvRe;ulLW+e}4Q`
z6GQR1ggq~OitsA^&IQCD^(*cV#v}h6ww-U`=by`Z>+*#ooS1Oi--jdTKck>X?-d12
zz(WfK&P}jG;K{WT4=i*VyEIzOlTz5>fzNA3T0h&=X+3>@a_gY*Ug>@3@@3Wjd}$tv
zvaToI4X{S@JsrfF(@0*Q7sLX#7M@#C$EKP$!cUp+!{2xffRgSa7`*=F!P5suou#zA
zNB$(`AjeYQJ5wPk7j_Ly`FeilK@NEYgpo(=29cXw`ng^CtVX~h;1FVH`J~6z7r{TM
z*4M!N=4|fHoVQU5&%S+a%j7?^g~LaV>}<BL{7gUXh3C&)$Dc{krmg5BOS6ppjr@&<
z8W7n2la6x>Bg0M>bR%WxX<_Ftr`H=Fl%ec1{KrJ%+3VYl@P>`U*!*r#z;WvjDZbL5
zY~vpQe#_o*1i5!^vb_P`>Qz{sG$%k_oZEDR<evcb);@ieAEq5Me931t%j{zGy*C+^
z?UP*hbCT&!b9!Ifkbt|I`F{N22LJ$vcRmaN_z&NH0RZq{{eyo104Szeh5P&8WE=T^
z`(OX(0DuLF!~fcoeE#44!M_0j+<n(00D%ALpZp2{;EqN1(pLnzB3}5q>)6{D+57%$
zPk##l@VCG9KL7wW2E13mcZ(}<AZ+wFqu>{Q={ErYZ@$4AgiB}60s#KMKjTb;of($v
zC%*7g%>P~XJy=`icJ94{5&YRR?89*GGHWLO@qfX5zwzBySf0P+aoB%^%OARn)7V~H
zW*>(?{cF}NJhFp*1MZQ01FJnnegNQbWGCkrkoBaG-}kmn$^2J|t$tp2UlZi2HfR?u
zU5T9E^-p*EOZKCtfT*~#1IjA#zE;Fbq6w%Cf^#)bZt;qhoC7LeIz4x8YxMskhjxUI
zf^(p%N<Q-XJgV(Iwgny=aq@p`aRz_;-d%|@K>(%Kn02I|j(D7ctva^ZQMBiPoRj<k
zy3o^Ti%O7Z9gTN9PHO8;qIuLOpantt{nsNzTO;-*O%Z~Yi8VU0PlYNBeB8cYXhw}v
z&@lKZz~6t8HX>w1K@(79Cd|;^RO9G<9pbAI3tgD=Zi?hJWYb2qZ5KviSG#l)8+AVS
z{4yGiE!hMB-k_5cX8yZpy1=fkYamf;sk74V=sG+&Q%sa_=#8}sPb{Sih=73gw@%~;
zEz5QP9*s8e$!?zr_iJFzdyShuzj9nhWyU4{T1WhRo>cXH@Z&82)4$iLrcJ+JQ(~I+
zNZz!W>^*{JUI5#Z+Zq5v6>9WRd!(=N|0Sh)0s0w@fJO49IS#uDrsb8dY6?6*mCGQn
z5VvoTej|!N+q1{!*C2J92Ot*>kku3vIZc3(L&^FuAn<+q)G{vjw*JsPbYQ{oGrk0#
zVS3`&=~aC7?a?OxAHOFc|3<U|>|x79*doVppMdsua+wk{&9t0PFN~`R*djb&BAPs7
zpdE)?hqDcj+R&IZbO6WWmMQ+?HpKpqE66<)Y5{IIC%|pmcIv3j?~=r+zkZPs=7*24
zSNe*mmvn`%`*mNFIDXxCenSu`=b#7HHs!h)Z+6ZzQa-$gKh#HLSH0c)_X7a#-p{u2
zopX}Ye>cbLFYRLQ{@wFjzUP?F+<w*|&~Dx(aPK^4+Ed$iS+krPS6r2a@ZoY#onjmQ
zb01;b`cFK`y9oZ;hgiNpIXI7^1`bOG!}AM_$QQ0NBHvuTgaLrXMMmuR?q85`VT7-{
z0xrx7B9{s7{5iHUcO-wn-h(Xv<u$&(u)rFS{X2LZ);D?lHm|c!Kpn4`$%)TKfl|`Z
z`rZ8m0c!j{m9>Z4R8U%;h<AHbTOaXyv~bSvX)xso&V{`ah-y$ler4Kq2#8-1s|%V2
zj&Bb+_{)mD^9NP5U%Vrc*Cj~n_z=(khnHrtG_SiTMi-(ChJGbn2BdlXD)9A-!zcg8
zmuB%9nfzm5OhmyK!ysO@HG8D%k#(K?5&~ck{{&UP^3jZt7Li9yCUZ(EXLxSZ=Vz)Y
zK#{M81(|;MeN;e-%&J%of)+3M^e|kv0n8q`H>Ic}O+b#U^d!8!uY<O4GN6U101D52
zEbo@>MF`Qzmv&1tKsM{aoX<93c*pIXUY~5^9~=AMMq3U0y$c=Sa&L<g?W8#Y4$T%G
zd0KCL^}g3_+x5Lw6h1J570G+jB>P<2AQ^wZ>hw2jA{&G1mj7+uodyn4O_}l4v6`^T
z$f;M8_cN+*Ubj$wSSS9bQUylkl{!_(2=?SRL<giu+S>GCj>Db+Dc=wkpnQ<<^RKx9
zUVhI2oBQ7)``j*r_PUZD<}?aO`vKFhaNRti4WEF_-T`v{we#2U`n3&gMUNfY8D1<4
z2yH)wPFw}PesMSo`^fHDeD40mhI~QL(GX`7Fk3{gzgx@H@CC`XYvjv`Y1(V@+ghHn
z6#=$AJ5Gkt<hk<4Mc5NJw5@WC!|wRXWQh9cOo8bM)!G1qfk)>i%M7?JIRV<}X5GFY
za_YQul~;1RBpgKdvDb`pI|SJ;Z={+4ofjTF$Ov6mC{u#Kdwu42rOU@}pS}VBc#VIs
zC^r2IVk`ga*ZG3lW?p^u6ae5)78zL|+{wuJ!UeXaYevA;3mn&fMKl<DMYTop<=B*d
zsBQcBId1R8H`$hd;%)Z2FMpIj^1pms5ci%Z0RaEYAJypAz*NQKG`-2u3jK=n5EPe~
z|F@5Gl*4PUya@nQ<D`=S03ZNKL_t*O%F=7^VMh2H%bZ`J+F%XK3ujpK@s^AW5o3jc
z7=7OnO~Ja1&yr|F;QoL1-ZaRL>pBzsZf3r<UacrB1%TK{fD}o904dUxEK9Uj)3U6E
zlHGPs_t-s-n3(SQ<(TN`nThD0KNB%O$1@$%5z}K&v?pw<ZEcp@rYMQptX&ida0LO7
zSPKQ9P^i7$nwj(CoO3hp$(QwBy#i#KtQ!Hm%FLU0%glT4_nmD9K!PCGxPK)NFcP0$
zWW-{xZw&d^_dSTvxx-rgzI94Z=cDw#)dR^meD>{c7#SMFGAcvX1MnI7Q|tj7RgjTc
zdK1tQorE+JHYU4+=ku>$U*8kocrn1C?Nye`Ck%vxEDa$s@>5`N>gp=K)yLxh@V2U5
z{6mHwf=C_hbd&WtpX|JO<EMmk+Af`ou`<q1O6Cr@a>!VaU<wf?zSZfxXm;3g_<V^!
zTx+Gn$lFh00|USqw||n_mN76+`-It%jq182ahFYi>qf!VDe1-pB#Z=)FVg)<=`Pv*
zsUSsRG~gVQC&;8B8ae5GPG&={5trM;R14QnvHC+Jx3mY#m)8bcOb-tEneuui?TPT<
zkdK!JsQ+9e!h56GZa}2^B=&<FlORuqZJD(6yVPZvEwa}LSjqY`72Kwz-_wkr4!G^e
zSm{_m#Mdoz<aQ`w82DH*V0lq3w-|sPDF)Cs^bbfLfLI44x&p3;KSA);AAr&U$Y}lB
z`_@0N5wb<NB6F9R1{&!r8N2;XwEe$E@OenRhhG3sD~;M0P$Oqmeu^o$wc5h-H`Z7F
zdwlPxbDVK)t&sJpnbp2z{ts`f;=}ihLyv>*k1jtKny31uZp<k+r(V4j^^L$c#suWH
zKezZz&YV4%fWgbn>u_H7I|C(q_h)1Eo?jho@gMU&92go5y{FTGp<tj0fw$Esz*gFH
zvN{0rEccC*%vyG04VLllWRnq*4H`qovXF>7NkQ6%nE>(V6|sl6*(S+zmsSbc9o<O?
z>&3H|xkV()ZC42KzwpK-0Kh13<yTfo#{R+go&W%xo*@MN=iet8{x)U|O1u!t&T^G~
z{GY!<_Wjo{{FKV|J_7)_*xC;O_~Xg10|4GP{{{fyf4iSN242`v1!50>@!Xq;0PHM1
z3;+lhsGo1VKvo2uW(OSrg&KMITe?BaK)Dzo1~58Cx&*JkMAi&Jg>)_MIYdlBv)%+?
zu*A&F)E@F+xRaTgp=BB$63u&7C!@tt4CS|EEw=#}gEk}Yn4hQEi+9-GUl6@#Vs-*V
z$1yVm;y*y1c*!?_pJ=m3QABqBVX_4`zCmb15v0zGY}6z=1c6}&B4zLX!lm^W>ay=)
z|6~<wofs8IE7qXH?Y7#!FgSH(8DH;FCICD%QN^zu978jV!2}jcqBfu2Sqrmb5}9~A
zW3TUS(Fvgo<KfES?XO{viZB*&ThCy298(NI>SimuxBEXA`36d6DhOHH-%(ohk_3s#
zuUr=qwDiJ8l~jv^mOaMmS1uaoFo``O?b1t5TXi@=j_HC-68}6c^fd`06ALwBGL4l%
zXLJyHAxDMCG_uRxj~L1ABEY<AUK~*S|LzKD{%chX-7MYy1bHVX4wnaG0nV?5xQEx1
zIQ?Cueq%Mx@|$M{YuhJ4O?k@8>rbf1Ss1m{;h=s1fUI8${rvMh`em38l|kNnVczAv
z&icdPUp3EIE2zGDv6XnjkL<)CDlnEA;NOG+;EBZ|dE?}>On{VcV-LuA<L1){Qftl2
z#B6l{l+53F>OO_sd+7{h&cnSw!vu6cmM=VGPp%fMwN{9e*VcE+bKgW2<$@c<QIbaF
zr>-pF>lf$OEf1i@|ATjrCnqxYNO_1|e)No*GhrDr?fLelWi-PmJ2T}>%&8LAZ%?-h
zyR<i3>IRXj&aUg(Y|1Nt!#DndQqO8P->&6$gzEzY{sDkvV}(uEFX`y$U`)c<Mu_DB
z+uwJ(V>%sguN2q~gUrur!mB7xe=U}Uif{SRvQW-kpcpptN>^gP0Nc0Gq`J0D=jZ2X
zGCpw+$<qJgYd--1Gz(%aK*)IZGReH_yo1?)H)-vE_!06au=ry_>d!oX0RT{)oS^a-
zi8>yxQF*Hz)vt4l%8S@{|9e02-v9u9?HHB&@BVM{RJiL1we$C<FVZ;pyJ>v3RZxx%
zc8!y7|Ld)rhydKW{5$}lP~AoCFvO(P=SfH5{^Lgh00-DQ;M@P@yVM_!ZE(+SI)DBW
z<&S-04EF9Miu=Cp#E9HF!`2q8onP?)cpf@Y2>1pob=r{^iXLDRZF&MWtQF7k1jHJ^
z7=xk**4ifqA&#lsAPNZ?QkXEZZ}+)e<9Z+jBx+w$j>!DmoUh^=z%vL$_8+7i{xAac
zjKPh?)pZ^J0KnsiCanFo^>ZLHT^4!=CdSppW?zf{iEUMU`0xamm%1Ra7VJ10yL0QX
ze}n=zz7xeTNUR$KQCdVAC*RLk&Mn|-vx6J;4nB3yC`!I($DyE*x@HO_-ht2kO#2Cd
zNtgvI1LX5YI+5cifVyGqoidLmNP{vtRskl>2pFVwBJ{phjxCeOH^vH*>evd2XE@3e
z+2`$+FY}lDue)a-gB5v6dtx_(?T^T&1<ASN+Er`cAmg#oMR3XZCgr#-&G@QMU4D77
zjX~1Rf`^*&caXPC>hs+|@I=YSm~XI<fziFw#nli;N4<>lt3B38R<%p=*{G1l%_mHo
zM)Y|}%9U!<viP`b?|UcHrCcqd6lTgs*X4;q#{$sbseBK*JjrI)aq%aj-qR?D>Dx*@
zQV5`7;*@^?I<IvCvM~hs56_JSV5MQ=@lVO<V~zAJ^6svEZ<S)Sb%3pgCsOs{X&nHQ
zsRvHJjQ@d5>;f0cW^@QLegd5AxO>+yo;<uAdFkLt8RJ96gz?Bho->P0{N*b*a?6o)
zVp|m-IkGM3k46yM<DvdTwLe#T%osE~v0eND;A*pjubf*XCP1ft5p@PB_c}J+&)e+1
z5LB=B<dvUSp;`C%w&A(l6|eh+I~O}(@E*NU9}GR;H>mCZdSwn6A@J17FGvsIw!JuS
z&lFTy$k`MvXTHG7|6pbNOq4KLpyuoJAVroa3y{2OjRqAkeS@_3Z-!(wuxEmh?RQ@%
zZTy1=DF(m_Pm0%AyMM<%Lgu^2>AowoB*T5~4Vu(1-y~%B>yI1+034ViyZl?rtUW(O
zvhz!G#0b=zdjJ3zE|6^f<um80oohtuS0ZZXr6Sey%)5wbsJ%#03IFf`Vlw{Ijqf9X
z!SXDbsMRKj$q3s-vH9f^VgM66i4kbjsr)N9sN9`L4g&y2#z(my)ZXj!WF=r?8iz$T
zfqSJxyZ+G@{T2cbfS2f87z5+`7^~3uM=>FF6GkweXyvn~Hs7GhQ4IrD8yEZVC?wf)
z(PQ831~9&dqU=;dLhe4O{1*EYC~&zDkAWRfIX;y~%z&?A6L<*NuYjENL^d18c;pbF
z@u$y{-bOmzyS--n?fZtHfFNI%v9Gm5{PpSCo=n1t$tpf_cw5pgL9i*^i2^ehq1{od
zUmh!;nE=Bwej|l)zGf%Fm(I@H#b46ZW(QwBw}{W&Gm28+WsxD5C(y_u>GV$Cq_xlu
zx5u6UOiD)}jcO7=st`+pplx(Ldy9*}Z8&Vc?&vz0<cjv~@1$;<0FAiQZH66To22-*
zlC5e+B|+xgUZyexc{(zCV=}VU%to+JCLmi%_bVNRY^9wSRtB^E2Y~xV68c};i|6gv
zw1+^Q;c(5vj|W)H7sCj1tq9wSuB-DbbHerWvAl;yQ3ARJ&SObE2c*v<+h0$;SFzc|
zCJ<SMD$j33LBhWC`Cg=oX6kKu{F^*iUXOFSu*Cojm_kFSUyy%?hhE4zk>$7eR~Q4U
z6JTZgd<^$k|7H;-{gLt5=ex+<B5jT2m3_Z86H$yqq9f3q3CJ)E8C`=sodLM{k^+#<
z@a{v~GEBgDsep$Mj63HvBLdIh8>eS6-@}98!;?e!=#j~Mgq@2_x!sxT?r-{XqZ8qa
zT^4^yGrS3CrhTSoX6Brkcpb)}Yy0|yMoIs2W+Yuq?r`L_qepqYQlI&I`a2l3{!ONy
zOpCfas0RI!vTqf&Z)X|~JnRexEBEZ3@(Iu_{eqwCKb8~}MT?xH!sTr|7CGXS<QLh$
z_zRF7oHh;dS>_}=@SsIn{EZbt#C4_^Djt<nE|JW>*rMyJtE32avqRPc%hSXde6vnS
z*;^)LKR!x``UHDy^9)%Rh)iX8nq}x*@0%A0`G@lq&F|_EY3KV5LYseontb|yf0&T@
zXO9t*|JXIMUU+c+6>9fA9=zD#`fP|cfFNE#3?QsOfDk}{HPivLm#Kd73|I<C#((h*
zDsQ<?3_`@#5fLu}m#-6JuxEnWH^L8yR4WV$fdSy7(+&yY#e~>{m}Kx_NJyQC9uG}k
z5Th7?=ULgO?-}?Y#A|#*J5b}ni%GlRll+44#iN{O0MEBk1I0sO4AwP>V;Ijvt0jIU
zxcy!V$%;+_SQc!=gPH+i+5{K?f5%L~L%T>nggn`?*10i&=fPtp;LX`pyvl4!UOGAq
zeDpqIf}}qku}2oIcb?ojtaIUtibb~HEw+K7apD^qod{n!JCB)0?@6x}$@5L?Hn{bI
ze=Q@y5J&s~@ath^i&45?#%^CSRtDYesIwOxJL*(FMJFJ=&oF$ewf;z5lS60PNFlb!
z)}xsZH@`3wdj699I*MVe7z}%?`X}4lP7ZUS_fvUHMQj-f**L=1pWXym{}Ac@9FJi$
zj0e)X2Y`Ef1pd)TRg%_4=(fMSuj=9Y(#AI?udIc5Y=Y~NcBtRP>deijbpmus?oZ}7
z_dPXEzRZns7{>#f0mP}|@}dsr=>#C>x^9U^@V4J#4Ws(0$8n1R*kp<jBQmkI4v=4~
z6JR7?lAnGJ=jR6`3;`b}W+-x<j_t&tkbZvDKDpi<vFqp``i0Y023qSsQ=i<YWa+s&
z47MUwzCBMrj~?2FUuP!Z#DVc7zflYXm;19Gw*I?{z+*e(Y<v2Z1@}Boew%$R{<}*7
zo;os)(UPB!tn+^J+L71(yp)T~&HC5#^6PT*`qiTySp?r>{BzsYU$&m(u+g*$8d>hY
zFre*!*)zAr_CGHj86Hfxpb^EmIAEU!PMduWZ?hE04na03w;1V*PkqsT9j?$rUS*Gf
zH&!C&zHz>OeMQ9xV7hanMlyWS;@`)1?SJ)Cva|p1KS6f&o=-38|Nggx?EmR6lh6Oq
zqw_z1nHYla{EVXKO^uWG{I9-?7KekAgtWy2V43@|XOy;@%Wc{6lOKWq>kLKXyK(vy
zoi{`+kNKabcJ6x=0Put9=KugFm&jV-UoDdLLA;In^KW-O1`j}dovZ@3?I7I&KcIHE
zXRjdyuyB*C4F1{YD0<-co;wQwxOR!G6ZRh<CScDb$++hh2*qBxM)yrtNGCuLeA+Xk
z@jS9i2gvui&qy4J)-=iDJ*-(B005qG?52Gfi--l$3-T8$5J83k{n&hlw>BO_(TJl0
z)`|vzXRQIi%z#Pncv<FaJd?=yQrhb8KC}%}V`bDzzV%2Lm>7WP?9#6;x7KwZT=qSD
zd|!xCK$w&eBr_!*F%LT0o-u~>4P*ny^?N2Y9x0i%=ud?07+*PitMA3X=nMBj?Fx8q
z?L^6D#iDAHAS9WBUzx4rM!kcPqL00`0(P>?))4FoF)6}H@*N99V591BMhC&9kh@Nf
zFLf=$H|U~{ouUZ3v?q-WjY*b$!_14T1EH)F)@h&hWGMUFRI?$?G+FRxjByr`JjS6b
zQZpv;sF#Dd6XwG$#{f9BG}vN!xSYR0XLJFOt+!VydIt9tfpZ%V4}i0)A;lKplz7TZ
zQ6$LPxeH?@8$+%0k%w$O`(E<>>UxF&f!Ul6fGyW_uP;+C13~+}#m=$hD{4Tc>&^HB
z>E6COw=H~!r7-nd07VNd#4$RcKszC5@o4jPaH!kR82E8)f5q)-Y=}OW-^dtvsp}Fj
z$M|I0CvI@%*e<4iV)}=Oocu(o<IFEl?w(Ahd~5^sq!#HoN`5o>IC-LgXs^2_Yx;h>
zXyhF4ofxv4k@WR+`!4hd9`ydbBPa$7?n!B1N&zrE``hs~DYw5sHeafLjdqMLp1y^v
zy%zs+fX^NoN6GhSfr#T&v9_b`uk_mvgTX7ab$sJ`s_(Nv#WOfqEnu?jV^6Jsoz()m
zuHLF2dC9&Pc}Z^$bh-9<_E+C}n~iS84_t0uJsUm0*_6}%<yMGm1KQ3HYOlYMw67H7
zTx;VC<=IAvC!+X{=qGHY-YIZocL0W1D^s)#MY~@VPe&N(OO(Gl>kf#&02b+>$;Z>I
z18{ADWcjZxk<5LDJ>qRE(aZVepPX~bYuzI4{=fbi`Rb1uNjK(6`~Tj%$!efJN5^0M
zd-AMz^dQ;mk1^$SjTs5CfH^%))&)O$iI|64ncA`Q@O}Wm%1tV7ZIu{^Gp8sffN4>?
zD-*PEy!Oa3D)%JS`#XETLF4wz0T3H};m5=jG-j!tqa!{#0EUK0+x*!doTPqsXq@*Q
zAWwj+%QS(XTcY~MMu!jsxbGk_97FRo9>xI1_b57+N3wbgSv>IwC`i$>TJk2QO+Zx?
zf+N}t2$LXlPv9ANNbLAyM);l~zW@<0{@RmT+0Z6HWb<vdDv+(a4_YK)i2!#oupZk4
z2`Mf2kUivyoqZvN$P7m-ys(b!f1=>y#9bp;n5E|^vRzm9FYR%y0E|h-!<nv3BHI=w
zpwW)-h0`~CBLAs!fKMGBPy87OqhsJ~0(>L=c9B7Wq0oyr>iFjM?8c`a#dvk41*`zL
zS;s@eMZAAc%|ax{vtO6R-7*6yipVo26%!yiN3D|0np;k6pC{iXGD=}o1o11xBH1l6
zmUbybB+pT5)luR^wr-VO4KWXv>9X|-(scWmNj(;SEmPOXG<F}OO4%raFD(x4`gdE|
z&uy=^ZV1(L$dk3(=5t4m)C_*O0yGBL6rEq`;9j;+5VPI{5WCOOsCFd`fpeHKbJpz_
zz-DFPiU~=lLHBtkRla6a^y5#MDYre60yQpq_erwRL{+|Y+mLGCR=Q0oHULaw9UxW#
z^7+bh&{hCi7a&qEdSVU0C$wx`+0|!c?lOTZt$t6*zK!BU6Dt6zcYWQ1tV~^%Blr2H
z`*lefKLP6gyn1`kb^Axl{Fk1g-FzAU1jz;ZUB^RUqZ8u`r|0@w{6C#q{GIyjLPdwo
zj9pMN9xu;!U;K4CHd4a-_a^3(-HOW$Xs-5u&iLr-neyebt|upB&go2)z+O5Td1;ft
z^A{CG@Z8#9+W+?txsKZIWV*XF*nKYz7}fB$q<1*+5Xers%1~eCHPH>Q3<1wkV(}r&
zz!>k`MCMTBiM_%ZN{rl#Mbdt6uaclhyu5zm_#ptm)8G6t0AOy7WVd5`2+2?1B4sW?
z07G2Pzxw!l003Wmfe`W2KYWRfUn47lLXqA^BQki_;*Mu2n&0vg<$Lm@zd^qzNS6QQ
z7YON}dzldW;_Kw&-ybGzaj#C+2LJDccL4y7@1_3#n;WE)&^Sa)!ZVW#@M8~4vvsrp
zOm2H00ASAqc@TW*d*1>8-20x70s!83l$e4yuMqRGwn`{=-!8IdI66g4#bSf%BV=zt
zejI$X+9BY3SZ#^(eE`OT5jvicbCLLcl-Z+S6vcq=BkCj`|BRQEQ}VzF7l20yAMg{|
zZWIHa58pxjun5{Wu?2j_OX&3|668;v5hpkSQV7h(7D$k<AU`sG@G>3%qj-I<U!!H8
zWyz^8KZSh7uR@U_xa_=>=yJpmke>zZ%imz7*}-3*nZvbSi+?G=C-0enXK2i<!zU{f
zr*_3rjBj0Ba=Zls@|_yIe4~!9U+ca2A0I8@iM_RyopHi6015IDe-Q3|EzZo7V&^Z6
zNMM-+7eU+YL!7z~^o)#2eDo*V4SCM)u`WlRlUNstY*%zjtW}0ZhAx5<gzYmEU^nI|
z4{+%lEb^51(B`L1o@%F2<7&5fYGa=#oj3e^qbSCO)xmB5?->pd#;J8KhiQ?NM33x_
zoHde7r^8j?mGvWccTLwj5mws~Y6UM>t35XE38f;CVOk{FeCoI$>l=#~t<QY*{A0wf
zm(2|(quQNUQt~|-=QnR3u5|+Rb1`cD*5&A!9$V=)rx-wJBKu&)OiMmv2^u5T0M<jG
zdXud`fE)xZ$l2Qe_dwpBLe`c6$h$Ak)1I~-=;<*~PP_MK+t>XI&M*Ox;@3x-xhW5k
z>hrnQ0jgafZU4R&|8js&9+@B|#2yIpTNjsba-os(7MRQgFWso)nVI#YB?7=hy!fY@
zos4Ub0n0A^Ds2=2XC7DganENKdX+yQ6R^p2$Cv<PaB)D}|D6R7wSqgvXPA@Qn0A#&
z<S<A&^NaEfn3vw6On}t^5TEW69WV(#Lq8%TwaRW8;ev4vMO$1DLSuzcO31R;r562u
ziDLpZ+5WvcO0w>yt5>)^ZdAWE3x~gYl92q#H%Vsy@WT%Q06uhVAAn>BCDsdK2mOD1
zohabtm)N7=*$V)GLW!7#Q$HtbfVDMhuNk|GyGPeelNblTLH-QPHY(@RuMx5@e}%OA
z|MjEfH{kS90I`QN)1*uA^E0F~P#Pg-V){JkK0G@_{tH$)_CS4sCiZJL$a7)QKL7xj
zknnmiFuqB}W+0?YGHU~|`YG~rS>i>h%~1s7IBgvu3V$BNlb`1qw$G2<O+b{^2}oHN
z0PsBz9;@$%>@iS068fy@7e=H=Ck%iv0;QBWfuD*ZDE<zF!3ZVBfQ|A;yV<lXZ2x!_
z7jmos78)U<I3|RX`sfA8RTLy6BLk0+u`wy+mmpVT$g|=XU%lCL@!utjKTnimrDClk
zQbYah`9+*uNcutE0vGVfyGL<$u7PL!p#2{jDd7WqYc5hSw!{1#CyI?Cj*#!o+7okm
z9R);1D_+c76Pc>JunUUf*VQtKf^?ZMk``u=d7?u=?@>{26b0EbV<q#I(aeCfn;Ry&
zBy!%yJW%`Dj2*Y^W7E}lQQ*^luabIH^EFW8<8~C|<pFK~kB<a)oSj)Z)l4np&+{+`
zl`n6>yu0Gz=PMg$D_&g<abmoX*`R3Ibr$J%Kc9!RlW}%)XrP<`3zHR3_L*WJ6O%)?
zL(c0e^m{LtbyP_k-7_bDHmaao-xdRKhf{1~z(n+mL>EAQwdcUZPXHyWE1-P+XQKXT
zM4gf4=ja4T(y;>+17OYV^K=I6gswaY!p&#b1=U`Sv@=XV9uuJ%0ssfM4<RpI<c)z1
z@L}`^q7c7oZzdUoMmxk8U%QE!#`-~8rpf_6bz~b#fv2m@AjqFhui}OI>?UTW*};E4
zv(VS#e`2(R58PGDZ8!bG{e?Q#Ggsa)s0*11xCcT`F84Xv>IpN`{sEbxK2siZ(mRI2
z#?pnA!2<sPSl{;l4k6w_qYW;E5$2kMi8*d-`UNpYZ<Q2?EKMv-L`6yvmh7a(I>7o1
zU<VEjJ8?6DkcCAM8YB3j5%S(P#0uK0^P~h&8R5ygPK-dSOm_a&Hl1&_=!Q;%khagB
z0w-(K($OmY?i}Ic3cd7ejKoIRrP`H6(mjx!pBK{PDq6_<CaBz?M)ej7q(cxb&<inC
zBZS^sA*Nwu;s5~Ph4beD06#iQIs^an#|{Dj?*F?B0Dw;(A?^O@M;-(s;Jc?tr=aZn
z2muTi?xS{y!M9f=`KV7VD_5vrRrdHN2u-5%iB`Pyx5Pf{#oix?0iU(SeIE^$?U!Ky
z;~|20xFaN9FfrhHFr6roaT{Y1s#s}ElnD97oL@``DIdv1jc3s27puWYt7rfij!EEg
zxvUFdtiJ)8M*Sd(FOL|s31DU+K$<}ys{v!kF8{8H3dTzX%r~=>=X@)~sp%CQD$^!_
z5RrlC;k5)2rei07k)e%p2s=@<Iw8LF+8k!;y%zrhKKrgolzC2ewc@EbV~pLAAKO{M
z*~Laz6T<mUZ@TpOXbJDXYuFxJ#>DCx#Lgx_ba-S6mV6y$AnzDSAVK!T79~uA-7lpP
ziAiMTIteo&F%<0iq<FU>@~0K0(XowLz{p>cW_+A_ksu|l6QRx0(?Z>J70z>nG8Ikr
zBTwRyx8ZtmK->Q#W#3^*x)yI>A9&t=DGx~}7u*|bfrniMZy=M!mzUc3)sm}8m?!Jk
zF==%4ynd73{|aLirI5RRmK!Qx=~75ZI+lnqX6>|R)6{jb``70+lhCWBI^Pxp@C!>Z
z;3di)Hfo@}(82)NV_^)00T8Okoe?v#dWHCnFIvDVzft)VqR;x)UthOCBlN87Jclvp
zp-Z4W4BCF@=p=ZI;Kf~VxAfELMKrZ*Ma@{`AmsG<Mmxk8Ub%^x`uZMFrpf_6d;cWL
z1(QG1s`AT$kEf1|W2&@q8}8$yWqe@Yuv2``l#?+#q4YxXZijkKoqZ$qoC_HSXZ_*_
zPkHs{oW2beue1lZ{Xf`m<i9%rcn6I-c&_mlU;re&LzsY|#XFrkzi1VX4k!qqD8A4|
zCdfAqmps1SeGj4IXAhNWXV>6pb*)=fzMm+gIri;elHe_j?AI1aN5Eq#<WP|idbvpV
zjkAZr^EZeRef$Wu@Wsn?-_2FZFGy~sPRL*Ef2StMli)695O!`S#D1Qm5jLA7>o4q}
za%vI%4i8bi^G&i2h;EV0|0nF{;Hgg%Q*hxK(kYk@h>9Jant&f0_zUFuZ=UV`4^`Nj
zqQuW*iM9akFhU!^C__h?RtfkU+U5HO<0f$W)nx5?!}kB4VT8~5#kr2?1QZPAYw>+b
zK-gmd03ZNKL_t(__7oV2mN!8-YuzEtks!bV>kI^ONUe-%@h%6%2vD?11Av_NNWhy4
zGZr!H9$5bZeBJkq<6Ouv6X0?GwN(KjK@Z3GkK@#}<y;ov2UiR}v_FBe(tcqu#K*tk
zaZ39MFvj5eLLJY(xzNYrU%*rMPug)4#wLnW+VchoBSxr(7LPcJQ3`x~>hL)JdYu45
zIvpP^<B5I4)*ioQqkM|&(4E*sxm*ykMmu^M6q&s(PuV9&oGD!np{P_nNNMRyp=lwk
zLC(%0xgdTB=(|qZW09dkjv8r6du5)ph@a0}>ki8R(Ot<Bl^Pk-#lD|gCGE|~w!3vN
zRlf(|#Q|;q%bvlJ5toY3lR+E%U-#6L3Fva=z)%1zY+Q8Fh+>>y?Enrz@xI{TP>^Yp
zB2W?!l?nmt{PJAvdm%b4%DP7@6WjN4Gw6SvbK37fVlCpnpQ4kJVL~AHB}@h7>;5ox
zYyjJi-rd+rw==~6LKCBH5k94mk*;M1?8LznR7QIv2XU(?e%en!uIxVNdM*>7ka!~7
zSK3p#Q3PFtt!3{$V+Ulc2;lq<l>>a>&;;_*>nkyyy$Ccy^+jcrX7qRi7v|RRh0`~2
zwa4PWs~q5|`zKKjyv$^fv4hV+uBE`kJ)`{@SRWrP<B9!>`B3BDZ|aF)d&=~?_bZiS
z-NTKyyUPUbHmJ{XZf_cx2{6DL1KR$t=lXZO`fgVmE_j2r0dE0o0G;0EOh7QkYR}^P
zzscjj!ao1Yj9d!*Jd>Ri{jRm$m;mx3Rjqb2l~Z7Q{KLBlaTm9f%>A|Nr2YTlCw`gQ
zQKj?We~EnUf9#200s#0vY2m+ombCD{{r5iw0JO@4#Q(=nKMnwR7u(Ap*~!uK_(fc-
z(^5M%`Un8vnJJnme|nY>`&Un11^`^1CtZUA@5P_|2(@dTbso;WLE8C?)0b!fBWnMi
zM}HjvaOTqA0RaAlH{h%HJ^%pN^(J{9{DYm;zi;gMJ@|2qm7kL}!=*E?0{}dqn26zV
zTD(Sw>G%hy$=cz;m>7wBeGl+~Ii}5OtSwIrjz@O=KD)1#D0#jIL}c$zMgtyvFVQ|X
zfsaB+2%bF5d1$sm06sAR#<149Po4&i;V6R!AtPE3v{+jYdH~>gD8aQ3@I8>7f65-+
z7!QW)34D-$0>_Q82R}|}ZF{ssH&!<wK{P%Lj_n)ArCV!wp5HN@USEl^{K8ed|Ih^X
zZDU3%ZT%n>83RNS8c~ezzj+Hkz24jOZg-`C&poi+%3-ZTCy`NGVLCNf{3TcqiQJFZ
zX4mk&8@(0&%AUb5@2#a|^oILG*F~2@qIZfeiH&7|bi0@!GA1Kql<Ck$Axz8MQ~fhu
z>VA`y;I8Y*<k}U98d#<W&V7j$fD6zP9fMG{OFSLwXmEPpXRQ}Jb*%jnXv~=dmBo{z
zzL2-9NINd9v<J5RKTz{?%QyOu9N}><+?zbdZqETYRP}MTvGK+HOe4e-F(6xB`(E;I
zs9wa3-x_VTqTf#|L$@7p%5lF>Bo;#hr1X#^ekr6r`V|6QN9~JL{)~SHm0zjqI7||N
ztcOK)eJlNf6EOiMVFLJpHu({W>_0T_ua=!_gsQCr5?uhr1mtMnCrCSo3D93&F9ffx
z5;U??h}-%zu(z$RGmx<YP~-F)$M;}rC?7#zU2Ea<FJHr7y>Jz$XO_Dn&$V`lQ&*Pp
z|GhBN=i*<)=N{OOO2N;MzL>3suiGnl>Dn5eneM)M(Ww!|`0}|07PX`+P!;t+=6xgQ
z936>Xw`UfX%%b0e8R(gM+%T|`K%e<?vu&&+=LfX?FM9?DYbpAFql?#V0=x%mgU25@
zztVZzBk#9L3dBCX(MgaP02CPcmsuxZm9@WR=e^9kyRj<mwT#`-@XkqjXk+DkL~TMd
zaf@X1!%TwLZ_RK%^2k?Wt@vn}V!*F7SZlvU6YvPj;M)=1fBn`HA5%SJjQndXTfebH
z$ocY(Ta<74Isjm5j1c(<_p`(#e2E#1u(3w%<$?nqmx#6Z$|K~lu)roP*KU&MzCXJB
zDqSC@cK-gm$e+OW(>DPC@7_+=&AaB|!Qjm|X@Dk%NLRr3D6nvImOUdLBqqTy>+Bf-
zzQF{Kf4xb_US#uh-Gk3|_~i57L)eOdz$c{b85G4MU`R-tMBpAghHh166v)${Z?)i!
zF{BmGvU~B^S4d6q#`COfKNin_j8G{bBShaMuHQhLfH;NxJ<qec2oZPAr0r`BA$A)r
z&Ln}1Obq<f$L_+P{q!<sR-4F6=a(8dztq52&&=ZBaM_-3gb`-e`VVuqyHdnc4@|=M
zJ+z{jkc5|bWDNMZC5S0@&~O~bfCs!Zvx;Y?m-?<3Nc;pmb$Bd6$g)8arXfm=dn#FE
ztXwpUE`dzPj>N>6#Lir-F}k!%_s^07HUr!<TK1X(&{`Wj@g$f=a*5r4JMGyK2_kn8
zR;s>iECelM?b1Qr*Hn8YX(m8#gwp${W)$P)0mI}zQ1u=C3uDIzUF+?dy0l-XO#9XT
zvgjEcDf>7x01NQqatHT~6rAVM$&uAdvN1$d`*N*!RX!U7L6sN9gR23^s{ka3T>7D{
zoUL-WlGW>^-%qW>VxNOL&M<ibu=?r9R=RU30%`nf{{U7QNgkk3vzh@=GXF$2uRQ$e
ze7@JU9LpqFkAYmDubBY2$67}q=X#zG`22G9b!o3cp!yg|_G&S}AH44%9%k`VuXKL7
zVgGGp@t>*`@f0I}oia>79)khk#hFz+b7f<Tzoe_p4!(GHp-b!lw_c;_@0Ido*;&RQ
z@BUsD^td6nz5P-dy~@AMh%<K_de?Jdd9cMF0FDgjAny(9fDGIOo9@kPOOJHD9bvhh
z0vcQCj-@~r8rdP2l~%v(IEF-vnU6<VcffiMv^x{taq@&bR3-#I#z^?*r^v%!;~II)
z`{XA+0RZ@$FZ@RUz*>P2_2HxM1^|5Rhh#S&tdNX;&!Hm#fPeY9CjkI|{|cS|kAL$V
zwQE0l0witp<OQ1JiG2BuSLpgJiaqc@{p<e~0Pw_v<Z<t>o+D3$XU>ui!neNueE>l1
z-unRneoX!TuRs2N0Kmy})SvG@{nr41Vu4=zDk$n-6cW?$)fw{n|Hx5l*FQh|X8?fC
z*a-kY`0mG1h+|B@_)Rnc6g<kecjt!z03UsjtPECOBXqjNvj1usfbp=*_TsCo#34Ez
zWuKo1BR|9T`Cc3YfyW*o$*$gb9*S{nW#GOip@kW_1|B_U-$%?=0>U%|9IZ~Y=pCfv
zF%toUPKQGmGY(8gqtu~<CAuHDMB{mg!!UK-xE}3N+T@O}b!H%b{n!-tjg|1s**Vk)
zl&|Kc-IXGK>)1{de9|7UcD}q2mwn?zNlZK$8fP3{npwry`&j&skCkz~(dim+=xVcr
zFP&Y$r|%idc+3;TAWTtU0Mg}P#>+A*j6_nh!}Me<U%P1Ii%))G0(5jg?M~eyk(_->
zt{pw7z3%nY`(}Oq6Kx^xH6*ufC;LIL(G~f}#(WUn4J8xK1%nD<rCSELNYt41Ln^<%
zGI->_Nu_*eJ^kg?x6b<7j}m-dKcX1eR}E0|T4)a7OaH=Z2gk;Wtb&tJP`bI9_e`Ca
z*I(=U%6ajCYZ;|Tev>x0qQT~hbgyNYl9(Bk<ol6vl<OMV>*F*-kxp92Mc!{~g_F|p
zNw;Gw{R5}a0BvI|;ukA`P`~*`z42DdNg?o!S_fpb>iIZFXTZxt<hmTmr8&ZkwSa1u
z+?Q)r(7nC391V5(_klel_;25H7v8g-<fXSG$>Lur`q}o_j;Q{q^T5Uy{|_A+!)FeU
zV|Teb%DUR@;N{u1d}M8#nKQHM6Fs&1RleNI?)A*l(4%~l`J>m3IZT2*yiEwR&+2<!
zSZ+tS&>T$ue|ObiufoC7CW|{%9Bhd>sUie!@c?%wnU}9o1m{-ReMd20rw7s2PQW=$
zu);}|W$0oTF0#j;J@fwn0KPCw(evUWdDOdcm9*$*uaQjr!TZ?feT8J`Kc6OT{9oEb
z+WaNn*uM4#A@ju*u8-~C|K(G^0swgEz=YFY+40D{eC8@mus=9Sp8J0O26-TS_wl2A
zopck#5zoGDRKNH!_}*(50RW~=jDqZ>9>14l{3l;L4FD+kG;U4SU0AD=F2Tqy>h}|S
z$=|^Jm)V-3NXY(26K#00!E+qduofHy0PNmI)(B5dQF+JS_mtCr8J{BeGY~{19vyi+
zB1X`1UUW5zHsHG4C+g@dOJ&3_J5li{=*YHH<=SY0v3s19FZCw~zia(~do0E1<4~8A
zidleUXB@W4K6&W|S(D6rQ@FO+z=ip`)MV?d1Rh?!wTAW5?rIUAJGK);iuRZGxzPfh
z40N8qlT*{n_}b;gb?X9v<Kq>4@ZhN3x!2nvzI1jz!vvfdE8{~4M!Fz47piJs{~AFx
zR>&WQUtlYVLYZ&x4(Su6c_n++*_BW8Ow#Sn$mY2|i~QHko**o%J?+kSc7JKBSI9{i
z68?VL4Vp*FLu1{WugcX)w_nn8b4~nkaWMD4zq`APZPsyAchsF|?Q5a%`t8-@DbKCb
zruQxiNx#M~x^B2Htak9XH}iFG)=R%LRmMb7bj+k3ZbOXv-6&R1rl%BRi9`k;#W7x3
zXyeBNh-LN_J$z!9d@r!&rB7&P$+kn`JW1NKpib&>_GF|6SJm7~cO=D#5g8yv3?Bw9
z0~BLo0(@4;5|g+%Vmh5o{o*L1kh|<mheGudAM5#s=xC1=0_D2L3V_CEx}WpgW99*M
zeMvb?+sl2bew$zTM4|1aU7nXL=0K{PtT-T}80eS?8aHgGw9nR0`D{D*`0kxG+`Y4g
z_W*Q0rjM}R3h|fEPp|u|c2|q|t%r7@5_qZdxjnkyDy92(>gvW8e;_Hp6nOZ|;c;gZ
zaALfS4;>hF`k8Hj9*^$r%xLV{^70Cnd^zRW%R=M#C?}`9WVj77n~{E!E-%yEKHJrI
z`@3gL#(~nS1KPpwDtM{Ie*^96Hw{?afuR89TP-vO5Cfc9?%?rlTA$`N_v;;=0(mZC
z7d*?_r(!oR+VC>uB2y60W^>Dggcte#keP(BD!q)nLBvA`NVbjwDPo=Gmvv^9Cc#G^
zd=LOIR3&@=3(FKG?>B#$WbaGdp6hG$Ld?8ENPGVwLf*gf03qrLmVsHH<ca|}#Qq1y
zhTj7K_}=;N006#omSpwg?BVeEK`QS950Fm6)mgUxXW9Fyi?ldwV{3%>9X&wf(IG}8
zU_@PIkA~N;lQqMo)70;a2Z*t#?jrmC-@i^)0Ye>{;D21Kz%w4&%{jCITz&I-06=S!
z>KQ6|06a7!>MsqqhfXIX8rlHHXHS5>fgjNzG}{s2`-DJ!-c)!VMQt>R2fiX`k&4)p
zAeHA^`+h@%3lC&n5X*V47kQfiwgwQ{K1K7}{3gJFq9eMo11Ls;NiF(hC!d8oVkd44
z#XI=R=cjRX9R<MM)gpfT_%0i*GJ~i<le^EdW*aq8cKqDGlhaH1+NC}h|3hOIsf%`g
z9FsTBr;d!{OK0Y>s}kTt2S@GVZc_QEohfW!KoX^tOgd$Sn50-CB$JYrbH1p{ZRu6D
z8`4kw40M^)yFnOLo&oIr`jJMeJZ)#KV-{#nh=Q|pGrH@rX4{`Je^lM;rV}vN2nV+P
zzt)cMf1dAcqI&03Z=V(YrgQZOk1CxE8u#RTGV1-4VV3t;$H0)J{X__|wz@-n?8D@0
z<oA=b2g+n06x;HFgh3UT@S*6^Dcu}ljC6lAV~}^OyRxMN@D@!GfOZ_&7r{<w@&J`x
zzo)eHt&BdS-EY)+_x?mDfUDDX`RZ8f7-VGe+Ft?x=K1`$sZX7N+;SAxl^1tj%fRcc
z5P$x&%Y7{VQ?9nZUV*fBzxwiwVcNfs#s9<vE&kLWTY^)5Iq>nBBjdFAr`uuk^{6DL
zotf)yvH9oo?#sQLSMU1e-FD@Dt)tL)`I~Eqfx$}yQvYw)v=~OXmSaY^WgT#PQXtxd
zF>i@>vIB=9M#2K978rpGvR~pD)pw1PY&?v+@3^Z)u@a#3-+u8T0ATiwmjD2j5z?0b
z!Z%6V{>XiV__tT+zCZd3$;hJyA@sTHB%{CgKGL3_Wm)tO&yaopdk&9Ny@5MHuCQXw
zt5<1<`bS^-9suCr4$A+;Lk9r>CohuBb9RLmfG_=sn27mBk_9g`NLKvVfgJ#V*Qcr8
zXI{7f0GNL5HEM51{0Y!+qe#rh)D&s+zx1QOqW;`X9s=XtRNnC`<PYJ0I84mJ=bNi2
z#2$)kWOXo9BiZBmS!%~UJ4zJgPh|9l7H!hvHx}79S}9-phV7(1kE2hKRe<q4_)$de
zsn{MJF^K~L@#Gi9@B+>UVFK7wpyx9}H~c;&RtJ^|U_|W2ME^?^!ET6uaB+Zp9Udcb
z7>a$+brTSy-tN8lPgRTf?C~j-i&CNk`CTzYdK19UK@jGt>18~9X(6{BNe@p{@R55a
zQZi|JE^YSPQKGPSh@%*lAWfIX4pP_abp{dedl5@N6ml1Pc{%oD;}q(p)%jD5gONhk
z6$c=n3Fs=mNXpAt7vxwE$k}B7YMlIu9XXF@it$L7n5XqFqpW568Dc|G3TbBHZ2c|f
z2KOzH&aHN^ry69m{pocz)~dz=3yp9Ug0@K^a=m}I>tdXJ$2O9z>Qll!IoCU46(r3P
zMnJUw`Mf^Xx-0fQg8qXKH6z8pTMWQkJVgN7CPG^%OzYdxMcQ7V?DUlfKTqe^)&m-u
zTR#8_iAz3p-Ba~t5O@v~khy-xm;l94==zO^O3}x$-TA%(c2|q|+=<;N7gDZPH6qQi
zEJz%f7@VA5#?u%3Sp2tD@sYz5Tul$;o9po;@4i8idk&M*3jzP4n1Ei4M6dI?_4UY~
zm$A?L-dt`EBqx8nr=Af4x0rxUr9dW5k-;vpM?2X8Raw?A!z}Np*e@pQhr4rGX2jj*
z{Bj)d!r>P<Xi1A`ui_XXUf&~zVSbU2`VIEYU-U~{KixkzO0xV*S8wrkuAke<!~lGc
z8Zf^m)(rGJRITxOLeeX1q+NfLJqoVzdpFK9{6c|b`^!tzu0o01%jd^fccJMOc|52d
zap7xxcF!5p#J_##egMGL8L~$B)AOVYFn^p--0v;X?~Xm>8Sy_{x=3iGQ$-A5lwB(>
zG(@YP8a%=F<_wMDMKR#}Fd*o=)*>V-(esRW=o1+|NHGv`Oq+l(gfZe#&A`OT2kH5V
zoxTTJggYU5B(#W%ktHz<>njE(3WRrx3CL~&L_yGiHUTFNj3Wl{%(<ILr`^>ee*57)
zsIXSDv_lZHNg)Oq$0P%NAHHvJ^710S*5~5?;d{od%-wJqi9NAXyfaSeWDz(g7r`pT
zotjIHDo4jJH!8n9tF-^!#%zKK=yp2IRK%&vx*n6#VaVb1yD}uH`h?)<&zwFxuKIoh
z_xJTpIRAM&r$!Xx(pm@mhG@RC$dw0<rk`Z{Y4^bynYVH7x3K~ESR-nqt`C~z{!r;G
z=<r~|5O}KIOncHEP?7I3up7gaPIkO{bzm#KHBtzmZ4w3`<Ky4B$BCVPcLbk7)?OZ>
z_q*rMFauEcxqbrPGE9KxU={agJP!{a7{`-`x7#{)SBv=Vcki`}f4bZZ6JWcOMam~H
zFXHJ7^XQdMOwi)rHT6QKDW;+O{kmN_`FrUEaMJ5(MmF0wv!E}#6Lx-iIft2gdW=WU
z_N-q_mpl8r&yefiGxe<vKwj!6F2a^hz+fp5nU#3R5*MN+UN|JSNQDi}*O(!I$}d`#
z;+JHIC(AM~NXK>);y=$iPp!KS003^xkj(r3Lw5rJj@(7cPM2?yM(aa6CjkJpA{G3~
zCDIu<b%_v~Xdxe&B1AvVJ4FeKVzvLb5i;K~Oi29H5%L)LwNqpde)1wQ3D?+^>YiOx
z?|YAsHNfHu*|JuK$<Fo0A~6QF5|#T9>jYe1AckO#sSl4AP@g>n4mYX(sh#Ba;M*^d
zUxRjobRvE{Lk-?_H&M}#o+GsQScg0vc3Q`%UXF3F)Fh~~v+4ox$!eqEA#8`V8DMRH
z69Mpi)HwQIrz2JX0KiDz_YsF=I%i_?1lVHPeh^c1K?6{}w5(rz-5c7G`kqg^2fmLE
z>n8Zd0G_d?icm~Ite60c_$B&Zj14gXavTd2z)XeDvTT`=iJE5wZ-5{&fbZko2ghLm
zoV>b(-+gp1$^{qsN6aX+Qw#{e(GeMOdMSqC`AduVT0e{bwkke$WSd2-QJPt>9fm@}
za6Rf-lSl*40F6$B>np^x2{UtOti<I|K8@g+a<FRu)`z*(%4Q}n)n95TlINP+-rh`r
z)SDnN1!+=63qwNp^1LP1A)Mdmj9U9!D%8j^DWBTlXmgvCgRXT`7bc)fuJg-V%75#n
zm)AOYZ&XClOU1C@f%N%7iWQI60g<FV1!@_1-u}T$riy`*j_Q~AnV=+$xd;`zGHpn#
z10;Xub=3@0pKadq1MrV6@tk6OpWNd59`Rtl4>0|XIEEMah&o{xJALIZz%G(}to#88
zt#0e%8<IuWSVh6Eego9voM{h#v_@?LlIyeyuxm!2n}BYLfL7r!Mek<p(Pa~1JUnu6
z9QW@TLA4kpAEd4~waBN-O<#Zh(n5cW|6~=vcJHJsCQ;8_{xGRUKHY8rXtX1|K3B)(
zr6$g=v{28HEl(7D9H<p=SFMP{+se{lm;ct|l(%@>7fHrJ<=fyUAWsU=e|gz)*-Xkd
zSH`I4ojn_HYD@dSXF9*q!M!7ep7}E0%~o1B1+7(*dCLy})~fq5YR^M;mZ8^5^f0Ec
z6JqX&Z*kTcK!?t+H3>O;4MMU<#%a&-vBOgUfN#A>Ncto965^d?)qvN>sGeUvOo;o~
zenS3#_AT;|_x0z<4*$V-9isLalI1VlqVhz^O<smNYr$98D&b!|P89Er8&u9K=Lm@(
z+Cy6UhlZ%UU)@jreEAZM(;|C}yShXZ`u<6xYVV(9=H#IrJZ{vF0~0i{e&J{20kF-U
z4nOxx!~iVKJq`dk$>oljhXDXD%+dIrtUn6?_=8oBirAriou~ms5k~z2-~)yoE;Q?`
z!@<Y0SlOK<Bi4`^61$WMvJJ&oz3a1HbUwuMpV;AwFKziWt&Tuqcn$-g-Qr{Afl%^g
zJP7JVOBP*wUgfhsuU*8P$m)5A-EQ)AdcQi#qhq?=0QqxhwOfg-%vP6JRhU%!+L1sF
z%6Rh15}vu(pQ+zR?jA+hX{+{gf9wNv$0a>Z3-u5`zqX3gi_N|i%5Jqo+?o&Z{CoqI
z>18~&t&9gJ%P0mZ38n2PVP1K@al|?<lFcFyX}CKM+0-4;XL4m25ch`6ute^8`#wp2
z_jR$)HQCa+`4cU1p(4Umwir>=ot{pBYThP+a;&r>Tx&-;dA5?Ii)$Sm8BRoSvfkMB
ziHBJLd-i>ibxaVpbGnP4hGgB#jib;r#qf<VN2)dQUa32LYxeEt*H$md`=}`^p69eK
zP>yJBO9$W|Z8CuG!T9VdQ)K^2yWc3;zP29FzZo3@E9+Ok>b$&3p1Lp52}pgR&aufk
z);a*LKZDGDxw-|SEf>4_C8-HS%f2(^*+YGt)mgZ<)WmnMETdOCF;T%sj<_;^qwBF3
z)Xm1A(TVWf^eX=NwK<%eZw#bOUXNmYceai{d3^!TUt8-^QBHeu+m*C$-5Yv<Xg$IA
zzM=d5{U<kn_V(`Y&&Ctox262oUpmu>th@TY(x8v!9hS_WfBuhSAdxwA6p|9pWr6_x
z)XoUVzW+W?xJ~w%-MB^3@`}TRj3bY<`9)Wt!@FNUaLv^k%J+PM7rKVH78EN4UhqnV
z{-hZRqHvLI$jpXxDDd5dOOTfJ7^U(T3a&>u8HZSntB(5A=?Jpse7?J2#Ub^#!qEYj
z=ZLW=bqL9x_k2QIO+vDTfS8mW6J$Lg$aHCyugmjc>j#mIdj)zfRkmu7jpYzCJ;I=f
z&V|GRkc@&Es~r6?;QO0xmoX^H^Az($NHMZg=JAyszP(Pr9qU1m$4~6IB|qnr^2JI)
z*2E4oQ(j2@uT(~y@@=w2fRw9e6oqsZqMcm>{^k;TF2+YkfcH(M5v!_23q95QJJ&Gy
z#x;Y5+iXI9pagtyM+85j_a-qX4brX2MdrO}_Z$Q>IM{Lzb5NUlTJ}uN(@k<&hKL!o
zkeC|d7u_2ava+YM6$1c1f5XS(mid2Q>4U=&?y2NJXPs`N6yVf~ho7u%d`EFtA;y#A
zzgTvs*Ly6<_vt&{(gFBKo?;N~8k-m<Hi=}?#I!iE@G?dW14E}IlVpig5>0hp%eL*G
zT%Q&Fs$aQZi<GU7fIUx{wd~H8!?z~+r6RYyME1|JaorDFFz1u<Y<=s?+@-%6S-;8$
z11vTBUHlDxWj8HWdi?c3$n#5|f9WRhgM~W-buKi4FJJZW!3l$jqSH|)Wo6^ZIIKSj
zbD6z#1^P|uzTD@TX|Nsn?(OThT?4dxu7!aY|4|>fuN+y?R?>{j)}{YK<pb8|`np$S
zFTF|SpR5f=JYQ;>WCPG|8uWd7hos=%_x%ol#J9i9Svor4E%y9pwHf(68DZc17D@Ac
zq6%>K@Bb74aQ9;rz0RwUkY8T23PZ|3y>}9PaFr2cnUVeqBT?~GSmui29vdw|-fhcp
z(0Q?Tkbc#Lk>K_V(3R#%)I&W^HX0&a#-z>HE8Ib0ELOQ6i)*fj#W!Xf+&>!E8tW_+
zxu0jg`5yoP3(q|b09Yzfd2dWc@L+KD#`^#O|Kx#>^0-j>KR->L6lcy|L5nf#001BW
zNkl<Zrv62QV9IP2P#GnLrphrIMuvzH*i|KjeC0anKn%10fz=Kn!;qPWiDAlrV1n|m
zG&t(sGBFkMp2&Mv8)1J5JdR>vH^SE~(&vE}CP89UjB?mRvG!PF>kWyDDCMrT=)A;|
z=rjrWk2Hvxn%qlKDkbTtf;L(qmzQ{qgOe|<82FVRT|y(u_a+Dc4{xvGWAB>GPEIz8
zVRr^ZJfA&#ZXVAstRJaxvuQpA{`O`Gzd7Y$#}G&5<fxTNyZC$JV|`vT8$A`YdR)&i
z4mrp_Z`srLYrFhBV#2-;%xpI6q}LIYNvB7@7xq52pr$2Z=T{p522%h$HeAI0qlr?S
z?yuGs*CJz?52f9K=W-U>gP!()DCJ9Z%_wAsvFt_z>_!Y-n6Bgc0D2GSS_bbc)G%B~
zMp=6Z)YZyqVPKUcGFL@K#w&^euinNW&&e1cg$j*_=q&i^x#WEhHfD&tpX&MP{^{#^
z$F^<!yQKs0Hc1g;L?*Vcsf1&Wg$n2M9e(s;#?%qYU6kbPF_+`H=cQbmPedK|jdI^K
zx!)oc@mC;7K~Bqgt3sgOSgk{#ViF{8baxMcvC$cMp->-WqDvszjdR}1`I3Yp|G~Mr
zbzKPqwD>=H|8}=Iwtl^}a1R@u82{z;9G;)+fAOz)2KSGY@Wief-oKOnePCw|C&tT|
zP+^R9YQ(@-XCmBOQ#)bXuAHT|m+pkk*JpW|%j!7|cz<p?atm0$zPLyI>GaCV;PU_d
zL+d}c!P9MKl)4mmY<O_u`TXi&WA>Kw6Yv&kPH^@HA)>vkt)3`b^UhB)il;rbGrY+*
ztTje>Ek3^}p7&(O$>oTTenG}(xc+_PRKLvgH@MtwwSrT>>}b|RtDkiT?2br8=xT#0
z_M)<Ltuw_i!%8?){8Fwlq7=rV#!Q1S1M<O3eIkRFhKqLnGJEK&G9<mSNQ=UFmHH#P
zdNo#t5dRAI?5H^RudxS3VFVVKlK8<0wd=pTL5#uA*t*~^MyC)U#ym%9JTgT)P}5jJ
z1Yr3lc^t$t=@RTdavTu=Qz#+^FngW+E5z)7VdfReH$GlN1mMD(#3Y1_q}kCT3I$rw
z#46y-n`E^hIvv3z`DwVh#JUSJgrFvB^xTi_8-+0*Bp6zQ-;WNDv^;M?@+*w^MTeoz
z<FaS01b7}+>-7HIT&D54dYzQ<N<K44WdI;^JL4PjivV8WePfw`HhUT?0DuvD_{8VC
zd@nv?G&&K!eQ|wrzZ2VQ_{h7qGxA8H^*M3Ejd)TtmI0CP?F+BYVy5078ks2gcy#9w
zj_w+=FRYHDhB(Hh#TK5swu19(-OFvnz*nx+@w*4BFo7_9F2v9{M(nb+=fH@^Q|$UZ
zx!6VQk(ZLos{B+2jrcUuPquw76ObpnPpm}TawNKByCc>SE`pZ#L=d1n$7{`SVAsDp
z3m%HbpdH561FC#?g6K!euC9izN*GG3%VYSvU@Fd6`wN+`<$Kl!dD1UWzK8Ta+_>X+
za3}z7t-ndmOQ%;uJUE(SdenTZreo{(!PvZt`B993Jx*O8bnlRUt!4eP#!l*i=xCMJ
z)2=N`F*<p5>h~4yaRyhd+k>t2k1mCX(SotaUw#F#hx`Jx`Bf7pK+aoaubBYj{#wQW
zaDMFuNxjgKaW`tQ92<?;UF4`e`NbK8pYsJv|J5fG>n}*M`})gJ2vai<dgky9UY%aT
zdTDpHh>zaC!#S5Alf0QI$@|SSbA2uT70=*fca7p-JhX!rf2UlxcmUWpTEwUBoxmsd
zji8bfQ=k#Wcy_vxy<kOx|4>W-rJhTvu1=40ZbSNPfMx=E+<vJK?{<0V@KA!F|G|*$
zM(a&wFM0+?%7ZZ&*V_?RIvW=sbZy96A^ih06T!WV5L=8~D`ozmB?mXK#fYdO`FMie
z#?b?1$2}%Um}SND+&__(%g#oQ_l}Y5SUd{K&S;olkn2mdQxuuVc!kc3OgZtfOxK%?
z6pMjNDNFyPUvfM-<T4jyjQFK}3ml<KbOqWr?mZn3vFyLmcE@M2A+ZnWx_BUD;|9z!
zl{LijtdWY_pEcf@?;a&txvYD_*v=Nn<KSN%CJ%~FUL)E5#|vKr0Q~v4uA&eFG1pV}
z$m5|`--{SPnYH`pu6~{Rck?I!pf<J}u>oc-lfQ+b8X@<Y^V2-8i~uTC%Fpedn<YOF
z*Dla~e!$iQYqTide3R-g*NCa8Z5!i$(V}wZ8ue$0br9xQzAebS!8!$Pmg$!msnnN=
zi4ZG{>vQBmvel$~^4>`BqO}ETk3U4`6MJI7_{JiszyR<JSu+@ZZ#;t_qGQ9(i#k?j
zZxWU$Cp~u8sNIUo^zVV*CBuk6W=0_3M#=l|jWcsN*$3@^s8+(qk4#vd6DJCqNf<oq
z*)OG@eq^$Oky3y!zA@hwonCB&ICXsuk4z0EF+uoyC<g{!3h`S6?iptjARZS17m-Q(
z`eg#TAsi-8oMhBN8u6qROSR`kLDJ%(%_1wf{kDEJnAuq44Vt{4dOub04VqDmSJwuo
z|92FC7EdH~e$S=-@*T*#C`^sr$aB7?ETYEkpQJ{|wVn~HEWObaoiAZlB}RpH&mkG*
znCstP@o^?6)NikJVYPz?d0xzug+rb9)Qe!FYr4&|{Yj_12$20GdEYvUL9wDD^}X2p
zlxR{j8rIrOby&wVkfi%5+pjGK;BA{?v<=V}r9^qr_-dH*o9GBok`Gvd8d=0|-0Si}
z>-)4r3#&_@?t^6n_`K3B(2PQcX|TUo_da`GicvE6Obi1XOCgK+>9_w+s1$hl$T7nN
zB=w$}Ug~S{e|*Ofo_b&s<y1Ynx)hmX`rTP8;HkUEbD4mj&ehQjQyZPU1y|K;^dELO
zH^&5YV@lS&U$-l__&i3Vzvir$E^k8iKN)x!(rgt5X99XHZG$g;If<LO86Aexn_y3D
zGq!XB004pP5G7ch$Q-qKNm;#gvg4EK&z9rP=QfpiYiPAoeK}9ZH~6EF<1t3~a-Te}
z4C3dV>N3lOW#=@koIkJej`;jdLY#9e#4LR92qRH7u?qt*&pY+)jNq;8%PuUkqo!Zk
zd5<syBAZ9Cf0yx<ajF(wM?KR2{o}-pSXnba56>gy-B|rN0N`|uVgt-nzYGBQLyuw_
zOzl4f0JyrR2org@zD)4y+3%3UnFk+w3;-~FmaH^P`x+tu`wx+4!i#SZ;vX6(tBvOT
zEII)8?`Gw-Dm$)`jm6pmV1W4>uLA(a_Y$KJcIFTQXe^TMz%cHiawo?S1E_Hn#pzi>
z=1Vt;NeI_mWZw})&%()|!rJKUi7i@O;rnQz42o3l*wi+}#$Y$=Ak1{wQ>1C}c+!7q
z?SJEW_PweEh8|i>?sq+A01(FsL%^{S#;WVPDqU)XR%R^kahR4>8^c@pWD_v6(7-c&
zL=XJfU86X*yO!~2NHR?03Q0}!9ZKaM*V~LDY^w(N@ZMqk&84hJqY=e;b+&;=rszG4
z)I4DfE$I=Pl*N|uNFkckCO~uotWu^lGs&V5I<omv%z~VE`e(FBph?C@4rQY()+aem
z;N`SfW2~}&jbgw|VneK@D1U+Huab?bYH)6C6ZC%pXf``2v%ig1X39!6nlEG?US-Bi
z22ymO>}$XSSz^Y`-p9;iXgz~6(iCP_{EDpbcgNcQ6V>h;VH-n%$1?p~4wrqL9e@S6
z-j1*qMyLh7H{Ut?SgRAziIVSx+N<v0WH10>Oak9M7KX{LBmCKGe9^7lK(b!s$^5lM
z(OQ34&!X=7y_MdkDMpOgix3-^0njr31j)12fNZ!lvN!Ji@``24Qxi_+*Eqj6zj2Rs
z`=Va==hUN>H#H>Hx&|5R08N#4k!c^?Q9~ujM>gNNaLbB_a*?Y-Cf_`FE4MsJPwpGT
zLwiToEhp3NTqc0hk!=;cYoye*zMtG&vm(2sg1qW;FF2QbBC<ZO>*wEZc^e0Wy8rzZ
zu-uMtxjFa-V5+o!JsYt?*htEWqC9wc0QOY|X93P{f<;NEcRCYb&F{4qqE0fAYTXSv
zuXpY?dY4fZaT4S!I}?%V&v7h*BFk)(e7tjsr9QW%H(Ay#cIJZkMLXAeqT`+L-U-sS
zw=x;aQPWSE8@zXnWle(E1ksCjugF40r+{B-*)iAoyiAx9i(r%)0?`pD7`H#MUeLJy
zp)$4SGavaa0Km5KQ2@aI@~5PA|DT4w3IO=8&XLyphbBo!;NRCuFtI^pH`%@4eSl&y
z%wBvCwYNy+Kd_&W{q!}m2H3xo?fU1*DqwsY#bh{oh?Lo8S*!oG#mfK;I&DI$!xO|D
z#7$-pmI+Oo66u_X=f$~2Li%&(Nrxd^q6M$h=7leG{ZEW-KZpoGVW>j&tu<wU5CLe_
zNejOia!`pVL=0eJxX9)5LfKgaOrn!g^bO$oh}t1wNLC9!krj%7S>7Y7ea@FXV^Av>
z@YtSFeD`AaMY|qG_|MNz<I_iXp;`_AW2`j_tDdBy6~_4SoAbC@Ute4L$^BzEzIOy6
z2S^gxaM3pcZMpf0?A;<_gOs%lA0z3{qq~Oi8b4pBA5o06w;DLQqiVk=>x_w$s6&Jn
zQM>(<?|Dths;!QMl5tu>FLGth4q{>*gcOHjla9YG<KS2^2p!NzqT6D0*fXgx0P>xR
zCsC38itf(k0nOL|V4t7Nqrxo6!1!EWk;5}CGeQrUIkPUpc^rI>u3BZCFNy9bQBOM!
z_rBUNF>W3HZd;ts_}rhWyID;3x80GZZwwCA0z5lExHZ7hvEKQl*P;DR-Rtt+#3{xM
zaAQSmW{==lyC7M|Y&+!pOyd-joY%=y?Mq?<@NbR|fF$d$BGtdG^bSZd$eUeTA$!XR
zNIs=KuVwn$4?v>LFGxJ@dVdD_%Y1L#`m~OLd_cN>s0~u0<fM+FzmP0)cTy$>g8PZs
zSP4>q(hWFqU;^W%d{MyFwHE&5xhuGMQ_1vII~tt`Kb~1Xj=+h@Dvm4UpRZjn3#L1M
zxx#q>J8MOZcab*(z~vQXjgU7>=FXn_J?ND`x0&j4&kTJPpUv0j#vsECxQAypu>qL!
zv#UaWu%SUy4~C%^wX%WLax>Nd*EVep;HGyf6VTHmgUIs5XT8jKf_w{oZXrZ(NVMt2
zu3J3li7cipx&rFG@YKX@wHdj*N%j{Y?GxF(FbJzmyUBfJr6p~Fj9Q4~x*(Lc(n0XJ
zL}`cBI_L7l8X=0@msLC%icC!wo)$Y!@I1F&l4J*WlV#ph+gTRhBE-HtMB4PV&pikL
zxb^gD0Kk2_NOu2p^Arq#m(S8|&G2<J0o44xhyYx=N#(X1w8OtNO^CX)N_P7<=2$nO
zK#0KlWnlY!FD7QC)uQ{O4v(cN5Hj-tz@WrTMoVHCR2gB^ONaq@VGS^V2lhoTJ7tf`
zfyY|bMSAY0MTU)jQ00;G5CMpIl2~FJ_a$Zwc9Z}ZB=-2}arQg}h}}(qJg1m80bv-y
zAVJcGib1X9<5S0`@cHMjpdNOgpz2|SGdEUnaK{ia4U$a4Yn{Bbn0r3}?5-5?v3n;i
z6Cw8ni~x;A8q~NX<S{K8^4^JvGiF(Qz=ru!-aR$oXI+9Bo?mHYmGo0c%F23~IZ-jA
zc)>T;kB2Y_K0~0E^2lg)0=Pf&IP|<7J9nAX{b?P5RDmkVJYyLX2Pv6UI|PA?2URsM
zc*$7H^;@k7H#QOZZ%2?INb!^@^NL1<lC;AtPfe^k{(QjCBV>OW#>)Pag5x+<u3D++
zv6{pmg6e~s?r?BuaBF}U7utC5r07byNk85^Pl4%VQtxTjMdapECS*wL9A=`8q6ZU@
zxv$IZK=yqQkB@SFivf6tr5NPYvZELP%>>AGs{<g^lg#U$dp(JQD-()ao-hMi)}KM_
zO1V;-FIgtR==?4t0NdW~NZ01GQ4YmyR3U%4U3(r>6MULs*fU<i$L@9Q{nKf@7~uU!
zCON;$I7o4)Zmj0sU-3Qs=KYgy{Wf)HF7!%i7brbGwj1<EDQAYLM==_mxEH|hiP-bn
z*Av<Hy4-8R?>VvS_ol~<JrJZZW>a$OyZo&4opj0t=66abpjU=It35C@xL3e8>bH*p
zko1;g0`71MLbXK*7#WN9XT3?kqMa@S6xi>Bt}~#wNI6zB@dCR?Eex?;H}8atqSY;0
z^6c?Wc8YZsmtGK)coY;F;SeKzLH6V9Ti?b}W`<y?DO#>{xn*M)2!iMH9kq~gIaP^0
zU}crye$f4*iy%m%Aa?VzVDScA`x6jZtc|+J<px}z#}HVw(|>r5eA$2JMY69C*UkX|
zrf-s9?BAZEo$;RC_X7YPpMDtt@bAXw0048Nmx0*e)rV&Q06+NNRRF;AKOh->vq7@?
zMxCPu&Xaxrg=Jy{&Yz)rOHA^M3~_8~KLB8H`V!sOBHe^SF$NH$wM3o>Z%wn6$1rK<
zPwd$V0NA^q7S8!OqQ6Im=(#N{P&+zo9T0IWf&w!DaVG>|P}@c2bQ*NM&F|S8m#E$C
zCHg(IkG2?r2f#}_xf#$VAY`q0kC7nBa*gZh%{K@;WEMa~0z+bB@ZjE2REhz<{L)o4
z6q5h+u_^2CfB>0Kh@3U(-@LhsTdn-X;;~)B7%B#6NAevS`+i6-L<`?~6zj6i`<{}0
z8o-MPVGGk#TW8g|w$jF4j_HsvXk5O07uFRtm~nCEn?m|J5=ud2**d@D%ikpZ(bTkj
zw`m{!ajIR3NXormo>YOyRY&oR(&L5X`<3@cWb1Z*V?^B!V_awqzF^%|_ED^OR;QxQ
zGWvk60XqD?ta7`G5|vXswBjTTLY?PB@$@GQK!Ia>$okM`-4f9?7#%03FX{-R(4}wP
zo{wrr!7~F}11y9QW}6`<OZj31y-~?&h8t7hHf{oBBbxy*@+<G1s?t;MsiaJdiaz(r
zTBpV`LhIkYmEN|=0N^NsK;qKaB7e?i8G=}jyCZyI2(-CYq7&fe*RciS%y~H1Q_z(2
zE9QZ7Wj-)fkIrY2vPJ&JIrcI-0Sc+EgQV}?F@%3|VvnN}@T>bLuzS3cK^keqe(6>n
zd1<^D;IRW^8Kj<P@19`-^mUWhe)^-SXQuAOr23P`DCm5uoJ@f^`Az1Ju7^o+?uCsa
z*9^n6;Y2k~uXV7rj!36h+EwXus?HnVw=l9jP#xSF;Oz44u?EoTtvp`0GX?TW3gYGH
z)^_JB6aO;PS7WT@t|vmR<!=)&2xV4-)vvMaOyqWnJ3q^I`ujLW!6?@&p7dnFvRc$C
zs-I1kv7r1D%nXPQf~2&_GTk8AafxR&VGv{|C5(ea>l1$o3k}zER#SG;T(7+HhPQeR
zrDGv62^c7*JVjAGbMqvay)aD(yV4@<{8vsN0RRjgBprYk7cKw*Cil|s!(S(zfJZ+|
zNO`6*0z?L_Mjb5x%_brDT5T^v0A54^hWtVmF@UhsL<mwIi2$^k#3&3;>_udNp~<@d
z0IlofS0E@5RUL4w0+CJoWkMJpBfVV{B)b<xSR10_n4_-+RbmPj*iS&I#BZwDwekZS
zVteQg+cS5XEyMuo91|cgK4K)=<i0V0@1xxb2{AJgDjGPOfR03gljkJJQWyl~aZh#{
zdnc;+{rBw0*It@IqZ8rr!`l$Xv9&XoD1v@Wg@sXsYpeO1(MPurAvOtPB+P?hoduJI
z2rQE~UV^Oc#wbc38<?~%jDeNix4L*nbNUH)aTkw<$@LhCu0=BL@|8x$DJm&YbSq*i
z*FjJwk;z*fh}bPBjXVvE-H7`7da6o*Y4%iONaes-*>-ZC-wj5@iHA43Zgr(XV!keI
zLiulh(L;?Vf+lPA3qu<e*x!g)C5aUX=dZ9wLTRt)3|K!D9oKJx&+~MZ?-LJz;*nFV
zmZaRa;R%YPU*;iPt@WPXi5;vB9vk3XJ;LL0(taHsP<tZu)p==_))DZT2?#ah-av|h
zjE+v;GO5q&{Q6k>o6#ju<6z6zB0%-KmENH#MhtkdoghWGTRumbw&Z-6bF6=@%s+O|
zTaSSX`Nuj|fO46cSP!@pZ4$cx>t0V{7`PN}G6QIjfbPoVWpo16C#!V=QZWsxMIWDe
zWG_B>e0P#RgW!{PoL}nVbDz_ZUBhnvUQ9q9g3Vzvw4Z>Rz2@ay9RhW|M{|0W+ta#1
zKhI2Y?)%myM_<huB()>pQhV@)WOp&CJde`uI}O;8jWQ=2Nt>|-SPUb~ZNjfYuk@B<
z0&a720_*vY$6pYR{+`BJSwj$pAXbsZi!XhVh3lP@$h;-WU5({7^6X?j6+8L682L+i
zB2%--fs1dog#qB5t};WDWRGQsD6)R8-^#io*DhJwwaOo`*!4C!Un`=5#w9iZ*C*Bo
zE#)g)>XmU9M#EF@e^J^eObl;XMfUB7v`7rIy?oRmJNp9<JqiFgeuR+tn=EU;(ILD4
z<NxLm0N~Zn|Cs7IObg1;o^2?=pfOLn1irT$AppO=h&F&YT0sXup>>Uz0e_LMhgF0C
zjM;$*K!u-c-02_$P%aQ6uZ*ykKg;F=eqJ8`-Ze>xe}pGHiRmC#7V<PtvFAUL?TgMv
zkrBy3_FFJ>lRXR0w*eTmR>?{r2msplFv3hDUf)B}kWAl33FP;-69X_F5VJK&48Sw+
zxqag2q2%!<fIS70Qa$+(7%By>5RYIbj4_1F3@j4wmZqvjOtQDZCL#w}3qtjo2}nF<
z^3~X72*d||x;@=C0T=7}?HMlmwtw<{%6smoim?m@*dE_d1B~f0pQV@pFcaV@$m4PU
zA})`O80>dvw{?ccId;!W?|d02Ny=W^MsbsRIh*+WZ_50{otF2eTncbyrGwTckmT-)
zZxOkqI$O1fKLDxE<6~hA<as2X5b34&eR)6VdYdQ0A&w#_^HYgGuqOWs^J!(tJP_=z
zBNN{~ci*PcflXKgoL%W)d%1h;1$j<d^k=QVw8Jlz&vd!gZ|djX4D+1R4($QcQ#U9#
z8oB2)IG{)Sw$eK!MW6_QRwTr~8KfWU<JjI0BuLxp4%io*k-C-L^RGhMBD1z;GwQlU
z_>S^eR@R?!q@>zs&uHYV%w#k|*P^dXEP#woK==OH)DyucX{_XTLD;GsrL&OKV{+<u
z+nYu_XJ*&1p5EF>ImjY~+-_*5pie5G7ZZ>vU)7Vzr7vu3dhg97WYd-U;0SepkwX9K
z)OY>4<1k?AP2CK1`K@LR@Rm=vi3teSl(s4>U8g*x(Rq=5i?+Qd2!kiQl4?;wO_WvM
zk=9rSFEU6$y30J%RN3xGcA&yE$U-AJ2{rMkr*?kgd5<0KS#5vO8Ib-+l)n*1tVs+5
zMxdS8IWJ5{K>}qf*^c$;4yliRW#=a5&Q<<HqHQHsV7wqk>Ah#7Qy@xq;$h6^E#RR?
zJ_rD~^zv5#02f~V_w@V9FHtEWE#cnrR{#Ki_$ql0tQ;k6``=?5?<Wq^c>Ga(6Mk&)
z%8O5<0l=F-fewK3*mG1~uoDpgKdcdGE)5|DuzKqP0AS@B{Z@vFL1@mCwZO_eBa47^
z6^3_{u7b!cpZ&>s>M#2T5T*c;=v>fYYn5gB1!1J<d}oEs=H_NdCn4nf3#=<p943Y#
zXp;k;Vu4<C8SjDtBTfLmk1&)t>GW(wjjzmh)K2_&&qFD#1??HcVl@CjwEJPWzLady
z%A^^g#!>!cwv+W6V3PM(eBe7;>J~X0%b*D2H=s>`)FaFR-Ja6AkAW9w^OX?~4HZ%F
zl71BZ)W*W3w4R-nN^})qbqu)k8R+7C1t#&gAopplYGFDOPisUkWIUr8Br^SEU{bnB
z?gF1M2z<=%iHwsVf0@tPP5Rz^eot1LA<l1tqTW-YCwNPFWR&pL_0^`+z@Zww@4{p@
zEORM-A9#LirI;)L1(E6V4nGv8O`^1N#uj@5wA880<IZ%SFg`l*YnIUsD6tq|5Efv8
z#Q>8fKU1sTpGp#DMC$_F?xfq}&hA~#D_mmVL$$A!akN$~sv;*}m+~yWTEF>Q>7AA$
z#0X7{(CP&6gutJQoVV5h_7{8}&au93tp+R;fbPFm<`4eWh+Z)X83w>jmP)hb(Qjhk
zuf8y00yJ{gOn}O7lPIOiu_|AORm~vdoRp{Js>{jH`Mv5<2>arE18=Tm-2hA4H(Yd2
zWPr6v|8%KHHRE<M0ey1LS)GB+F$0qHdp9#6tu1M)y#5rOea|=8?g3Z6-VDUX3*L-1
zz?nN_4e*vvo7}ks^Q@gZ!3MuFjMh`0*aIuy{fTEU*@@E$s|~HT8sgzk?BP8wUu37k
zEXaMLq%S*9k+HAJj$06v?|vX6Dod2ccmYy&_TwrLg;3j8HqXD(F4;+owylhZ*yVQ=
z;%Vo|m~6~=%Pep?;@L}f<knw+AR0QBh0z}(?fQAq&gH0pJBkMY06Qm1%YUg!<)1r4
zcJ<A9`i-iDq~G&?LX{7^M)vu?afp!en`3PRFt{~)6>R|9w-dr{t#nv7U@uw#N}VZ0
z0RDgW-YiIy>pBzrZf5>#EnQu`H+lh@0Em^~zG$&1ii<`e%amqhg}j8k#7xA59{k|2
zBfPvh{NhoBA0`~3ml=f>V>@JPCStU4mqd;vl0y;%2?7un0%$DVUA@#^|DJjMaL#u#
z^W>lPS5<cdEwS!HS7+wUdvE5SdGmhXIp;g`N*y)T+kdpxmneXi_#U)cnsu=$Is<+(
z5ubu2LFRyk1&J$WG{><cGOm%v{f7xh-QS}~7=Ua%0pQS7=AfmT<n(4{Jw^pIG-rU%
z<d+m%eiXc;nbSdh10u=!5L~3A;LAcJIw}4>R1kzdk5p{E001BWNkl<ZpRq6s4!Po^
zQyaM0taB)o+vpsYTMeuZjnpoHmF;18&Pq?c%j4B)8zgeh6yXw)rG<WE{AEQzcwybX
zdbyejuvi{kZp$JnUy&96gmF1PO)Mz?nERs7!xSbU9J0O>UalbsMwO6Iu8R=*(aqom
zZje(Dc;*7pDEqPaupl(!qA!ZZ-}C${I|rwy?{CGDh!P%+pd6~t3v&)|!gIxWCfBtf
z(rxMcz~N3*>2D&hPk!B;x3s^aJ_Z0J%4{$iATvHI`TgTr=D7V&0rB2$G{7gO><<Wk
zp9r`PsqeQ~ByRU}d}W2qYQ5U_3-2vZZ|koshYY}BnOZzNCh!HNKr#Vk>Xo0VTn|7O
zjzs7*-I#in`2B$3*}p;dDDST_0hqquo@d`|gYbQgPeKKGTHOFP7q(43YXfJoJ_Qaq
zJ!jChmi1v;i+pW*(8_qtdCnd$^#50{Z(wy?kEkA9Y+<R>unkSGT+0NM$2p4usfz-r
zR}9?S3~0NxXU#&ovlI)rcMGGOQO?cC{I`4A@40t|@z{5}g~P4*;L!l<X@;A3F&f~W
zF9&>7lD4iOshv-+dzzdW_iw2sB6QmRIZyhuydo!J9WlSUo=>lOo6?zpzNbC`3+fY)
zOU=sPN${>OBO<Fv7Q9v!+7t0GZhuEFgh)GN8;i?~1XuiNu#uIiZ7YP;G3~Q|j+2Up
z;e%yXZn^hv0M9@$zq#F}+P+U|w!P=Ci%omnk~sO(mqhvR6Audm@V4d)|4TLJ_6Dy5
z0Op-|`~Uvc7Xbi2|9)xj1JBA2|7%V@LXBS>13p5sDYpB`d&Hanr5(vV@Z5J2i~!6x
zq~FdF>34hMC5Zx9JO==1b&34tC3UfIM+~o<n-k=|JTL7pwaC0Ilx|S!L<eA^@z71p
zW;Rp_v=b+VX&C77oeg0w(!mfO`2vPE&RE^Xw}#^V-*ST7$poe9%^#4EA{#OW%o>=b
zjIWmCf<X225Yc;`J^=_&Jca|JT$?}5fRhU?Jil6%0>0WCVK7e7X~g;c(7%l!>hA-?
z0GE5i(>lucrj-cVofri7{MBuIf7K)^GUdsWb0*9)huBsjo=k-G2`GKf9qZ-33anT{
zN_Ib)0OY<7JR$hx7&mB3P!^99O*s_2+-tsY%r4kc-zF;aaULY&c<X=(YnsmCnDfC@
z)K%|%+U}WqquVV@HFADh=#+hA!nne=9N*Sx0xJD=)EST7|B)_2s0YA#i^`v|*M)Q0
zCBS?wc3}c^gGX;CH^7Tq6D+sNdNXx#^0ru{e!Iv&v`YYEp6AMVYM!%dJiq+qGN}Y;
z-0X2z@W+h$hvg@{c;^9^<ueIQDnWX-ybz+ZVwD-tZG-452;F$GX9i(bm;uhy6$YXF
zH~18o{k7Ttj19e(2?&h9EFS?^{8O}{RoZ{_>;me_58v7>_gm)^U~T4g(E;H!hi9*B
z<BM-j$^26&j~!{_;?Y)>QLs}>4YIE_2ULc?JEKs6?0fW$s4Z2efKr*s1k7MC_FK*;
zR$Ogw?N-KFb6VZ?+@ax5Yk5z#?w?E0uKP;QD0eFwp!s*{E4j~G7m{`nDV$u86=hJ0
z(kEqPY)T7i%%`FrlT2~t7&No>QEO4tK2W0NN3dfw_Lr$UpvRewkcBq*K5g?_kv=oA
zb;9BWbA0-&Fj|3L%tU(3)<>H!ZPP8S&u9bI+oq<ro==?vZZlsveIe{hvty(b1LI<R
z7+CB0qS)#mYDnDs4__7qzaPCVw)_`XR3AWX=6+QWd!r-uv*c$005@L|B=N7_60QF?
zPY6=~&wniKozZa}fAS(SAK{;ENzQ<0{zkG8&Y$=xxqiJ*y*DLCL3>e{fOI%Q3c&A3
zmce#sPOc)BtbSwl*1xhMAD+u^$hawxK!n5fM>^8@XSHe4=1v>`$~9>x)AN}PaPEjO
zB-ucawZ|a$0dNsi`dkz-qX-i!^Z|f#!ZbyZ{GJ7*O2x3q%kfl?2PQz<@c<CzJ_8LW
z%mKaIy)q3x^;nu~;nI-~UfrmA;s4;;7M{AW6yBdk<OE@}0vU>uRA4dp29L{rz7Nk{
z*}_-e+Wq6-jvRja!Xk3$o!i>^w&j4vmytbf*Tz;Eg$numod+>gnFC&spTYzbz4Q6I
z<Yze@PZ<&hWkS7=fnlgyKj?q>eDwOc*S`ahrDgB>Q?bL!0n7mBItd!BC@h4w`(>XH
z`aTU!6c=stJ)Mv8`Bq9;5cBztd7fUEj+NOM$2Gqj=S3<Kl83Mitj(oT$rc%k&aW}>
zL^sBF50E%%Bh7GYkm6V?&TZ=i9w`5{-}K$00q&nmfc2_R&VEYSH=?*cmj9*ZduVZj
z9bd(_p~$T9eX#qdl)p=Bo5S*M76(KwL+13wXOsXZM13?Z^3QGm`k?cHHu?cUAAs9h
zkG=%8$es#<h6$+n4!Cmr7NJ_Dzp!m=`W8t?CM{|;T3~wIeHFR#eEIv)i$^i7tc@o4
z_N%L)O{y-NU$r0Zt}+2Xyt0Kazp*~Gzhe2BQ|;XL&tjR9&~2sTshgq9MEsQlQ+LnK
zB3Rd+9mkA%v)b2k;sD|8P3Lf?!{2xR=gWx)JddZ}H7a2C@AIMM=7Cv&+l{)IRu&o&
zJ~V%Doq~J0B-giO%3e_Ci(48MKm^O5p<yUYdhc`oV|XYv_j6mDoWNWI9o<jF@51@Y
zI53_zWa*RRT;WoGfD?PW!HHh}($-m>7!)ZId6?_!n>^h_^g$_6_mknE&Hz`!3yhM`
z@r=lv&{sj*@0gqgHjj8d{qO_nblkK-%l_j>1=0TChM=~FYUjV_l-TMo9up*gZCw!W
z^BY1PGroN^RImOW(F4e~)T{o2FahuV#G?R!XTKu)0oRYq#D8OH3$c$dZtD-;A7cn$
zUcK;_@4Ey5*txl?Q2+w1la_dY_;id3fG)F@=@0zS{U-nby@B{34D>}lbwqpuXk)*r
zSpmlbd2XE|nZ<c&|A_h!Y^(|6FzgG%(A$#pnogn^XfkJ%Mw)uSMHs21*+L@%qF9hp
z<bvX4!vTmR?bn4YfwcKcZh#n*Ns3^Xi=)6qM3k!IAJql$K7amP<Z$WaJRVr=;I+*v
z3-GNg8(3&1c<O>^Iyjf>3=o-0H$>`eSavy=>i~S`jSYO`%I+Tj0Py~$CZ0IiR)paY
zXNC7X%CjUgNWvAp@Zn_JIp}Q?-vyTs&2zj4fd{6AIshe8PX5k>tbdM+0E4;%BJzd%
zlgHDSipzBoO82i&8VZ?PYD*8#2bKBs_kMV>iI;AU_N4@Pq2Vy0vY`5CQAcH{%vL2j
zfy{?4IBBa7zC^q<{X5o`_&|;KMDm<xV|_vRJ7(6`O6S3pD{q|+>{W)}*O&KJ>*Y*C
zGj{m7qYZrN*8U;_p6n)g-_mpk+S~Q^|L}4XTN8g^{{;K<2Vi1+1M0@*YW5eSD2n~;
z^|}rjfWxw{;^C3HEQF&G@rS?qBog_jVSdmL1d+ap76>-}kQ0F0<^2_XfH1D|M3*xF
zb^F;ZAwRdv2;S-e<X$*sgl3Vy%`H&fE-wTtHucEilMf$9UHQsuxA22EHgYGnaJog@
z)7u|RJpS~Bn|u2BpKdyQ^8UH<_-cI*YL1)X)lva9(F(P`3)6e4jT)$@AnJy3dq_Bm
zfS*6wz%lFXyh}OLs@C;0#`|}#?0!A(U2NdPdqmys$l;SmoA~*YtvfNWsihe?eC~7`
z|L{}`54Yo5hT>!lJl68~%+Urudvf|><Q*+D9-CxaZKu@WUsPyoq^E<|Czlx1sOO8;
z2?_6Yq7O)$$bi(b6nY;wf{jT2())I3;+DC>w=0{VWEJtaSsCkJ1Mowk?TeR3Uj=3*
zq!VwLzR3(5TMi2u2Kpw@=Fj`oFKEU~eXanxf|q)M0WrS=ZI&-y5z6+&k~oZAzj74-
z@ci4a%kd>a)UhJV%@y(XPufzS_pQ0~Z2-XB8PNqe)so}csrzLAZ@pjI`73GvKkW!L
zef51e(eM#o|H11R0!W8TI&QUzjwQ}N9*OQi(v)@^NlO<CIZ{ahJok$DE+qPd3^oM8
zuQcT)YAf|T=?OzX?`y^-Z{1o00Nm1cn@v*Ha-s7}@&QX#+$Jti^C2ET0$Kv`b7eSH
z)Ju!DeA=3$2yvFdIemUzt|!op9T3Ng1f3{1MtMGW+N=T914tDi6We;=zU3}Hd2tzk
zd3hDn%2!^yg{{FDpM2y*V9xZ4!Z_)>A{3oN5;;8c<_7-a%?%9pke>E*)8Usd!~rq0
zu>CxaUyyQ4Tf4Kzi|0u;Nx&6+g^I(^nSj!!*$-50p=>A1xK*g{@_7FW<1ppgovX&V
z*8zDVxxPkI#fA6&NW<ZCCtJ9@Gr_C96j!HAMhlU{sb+*TiNonugm$a^@#`t$Wuuhn
zy-!3Q>BMA&`U=d7&M!qJ%fbx^N`OK6udD;W?*JJ&G9S!V9Bz+6qGTe%JYHaIiX~rn
zIsDYgHtwIB;D;MyyfMl!Fu%pAMubb97*8%W(2C_Z^?`vZw`+g4y)MV>_vUOT!P!p!
z+vMLNsX6S+f|_~|h0D?U3V_V_%;uc1&tn}!5XNWjciqXAXtQbc56fYhRWf*tU52r9
zXn08(+<+ke1i}c$5Fdet&-aJpO6z8j5H8ja+fGvCW{{WF?q_}PL~KTx<$T+DIPc(*
zr^<!*;dq$)Bl&te7YPlo2ZCI#-3~=RE^#EhH`!v+;o&@f1odNY+X?;CA<)qS4)7j7
z{oWJU9E|bw%GT81;agWX@#3u>K6>s5-g|l;tt8497lp6YVB+!ewQYRw+7{MEdvmZl
z+iBvL&SfY9QjW7SUfuxtu`1U$?PFXVXXV?zk3~Sa!SZI&k#JhI`u57#d(grw2Y`!n
z3C?R+wf!&$2CT9uE4FYO#a^eq4>R`5TXgXIu*<+|1|sF`T!ORRT;ex`G3s{Z%-x%N
z$I9(KH}6zQPO7BmB9-=Z={V8{lJW9P6T|Fv1QuLz_7w5bn|()Z=s}yoEXe{e2@HTQ
zZPSKD-++*(J!Bh{dh{)za$#JvwBv4@eQ85;0x;foYf#RQ&8&W$vzfDiUe#nWPAOeQ
z#J{6)=*(@v{s;64(DqrM(G<Z;mfpEm99kQC?`c8Y|Ihc{1OUwGs`NvTJPrW(*drGJ
z0GDq`O!pW5LjK;^ln+U_FN}bjh^@c>rl8EVvnv3A|M_k4_J1uEWc`~T7hi!Vz9&8h
z|MY_R8vJkfeIJo`7~T930HAYR`cJRE3=g1{h)w>~0}sgc_XU9;QJ)1Ok`FyB?Jw(f
zY%AiS77CxI8@GfR=<9Pzg5=mS@rgL5B08BmrCir|`W^L!Xn0|XQnmF5#G|(SmO|Y$
zlaDp~aSm{C6n<|Rre@uXqkxDaIJhX+2>@hKz>lOX4x;js-R02)6J2!hW0yGK7al*2
zYuh7S*_kfQyEdBOORwI-mtMVvON(uso^OVzmf<AB%FYO{?vi47S~=Tk;y-@k1k!DJ
z9(dh}-VEBJlalwmgWg8OMd%bc&2m_D(evd+0B&y)AdgY5h^WC1xlRFnYIxgA*%I^T
zaN1&^m_^W;`Ga2fLp|QssFg|Z`B21X`fF3ElCL!xLY+6LW5M$mm<3Yr`$(V51%2KJ
zBc3PwmvvreyIy~v{Ceg2pte8t17g)1V6;;sDn>)OFmH2yiuC)CnKfmY?#=Ct&3Aym
z8|S^CR9WN8Rk3(YD^m{UNL~KY$+&JIQP=)n>jC=Q6c{}qH-%Qt^Sf>qd<NvN^-Zf|
z2&#iBZyquLhvl}4hezgeWKYCTq)uiag`}z_a<_;-NA8-k)<uBN;3ILkBM!=(gxq--
z!TEfDe!jeHUOzhayv&LDhRi@j8Q=<nx7+&BA)*b)RS|Ju4D`G>5;SL-fY7eg51%G7
z0YqBi;|O${G5(W}p2z?F@87Il1hBliIoS8d-j>IfI{55E%jh&?Z0U${CPNm#&VGb>
zo#A>=!vusAP@bl4e-ThDd=3IJQTT|PhOo<YS*UIiT^X6r#OX7}JFDCw9$@e7)-nVK
zypEZ-`3^((dxZ(OYs|@=DO|t?WX&~EYLF_CFTMAJegL!P>5iRI`C>^QZm2i5Ysrf9
zhz9IWOf&&Ukp-mXO8r?1h*1D`Wy!+5u`xC=yCKI<pNCCl0FKX>^$-{>!2angK<9!$
zQE7$a$_NnIlL%=Rz0Luw7LgS4?0>xPbb7Pdy`Ec<xc-Sswr;C8@|!*R`|*Z2#B}HA
ztt;*M36X^&+e{dNbS!cEH&i+AmFLBWpr6X$f4=fI0N}rPL_E(Q|Dj~v``;gx<6nNX
z2Y81+{^s9e2p}H)HBtc0hU5&`+P)~)1@R5IwIVkE`2}gO)nv9vK`t&<L^Bw}M==)W
z<Oh25nlKW*E%^`<Q_}l-X`u}`U_)*F=}4Zt{^pLh??CgghePBfly97g*ZxF5F8m#v
zrF`oy#8)S!*lzj=6!O)>iKvhw`b?p*ApHWP5c0@Fkg}IW0KMeN2)7zBe&ZwO@%0y0
zYRv(6qI~kcMST4JC1jaLnt6CmUwU8q;}n1R+)bSBHt^|(j)v!$*BlkSa17Cb0y7e3
zn-Wso{~{oKlml{<lKT{eW)ZnU&IDy#vf_5$3`)a7&VVoq$Lmd@4V(;6!^?A;k^xX3
z6-H+?O3zb1FC~F#MX9|I?ul}t*FG76vB~a7TYH}MuE+%B^(Iz6H{3TT&!3KCN9RR;
zFcv#@TxDb%el0_+RFG=vbCE|K)FIj$Vj=7Z^@#NOy$H<5^mc2StMs?`MOa}TK+Xi%
zV%xVidb@4M9|&*9?^C%S7Y>`A%y;y#9G2TEf?9ou1_)ULO|(GJ17JNq==y*%4&LY7
zDD*(-e4G8Rq9};+@$j9nU!JgSbO5X^$;P`^Y)yLHu4IGPD9<au!If>BJD|{Sw;AJ?
zo;r(9UR=hVE}b~SZ#;SupSZMK?pKS4q_F<O=WpWbV2tmt_VCr0*9tRJdHnf~(x1Xa
zG`$}GQ(4bXrH*rod#^KdrJC)i21w+GZT;Ols3dop(f<C+!P>Lqs>i4`!0MovG2^_M
zIu83R_B!9KQpx`LZWkR-!hvmSd?VxkTM9wx+<!!!S7`grJYQ7#OJ-mYw6Ard5AnhB
zcTtf)mHH-%XgAdse?k=i=~KWD2yNwCTCZs)&N!Lz;?f4F<07*_Ujv$KPcJs7Ue8dE
zlPTbRn`F}`1b%SrcY+K1pxg!4C*sr(R#x6WF+LFCgQ<vqUh~~AEr_=}^&DuMY$>wN
zR4I@iwXP}I`g~iw{-61-#GQA~epLD$2@}#=xd8z9=ePa>0Ps7X5u5$luL?8pz_~5G
zAL0dj`hqYNV<${R@4D(0Wa5i3e^f%;hHAT{FGifmHWk+BFvqz65uH%Hc%>w2OKs^7
zLmO&ekqwa+zr#?o0TuYj2FUXe2O57e5t^Ub9;u4aR38Chj$C+f65Ss=C!r;5r|19|
zIZ;+~4oSiL+{HP<6{10Dcep&~g#K+u4nOzENj!RX0bhLX`c$2O{gw|NpTo~wI*N|o
zLl;HChk^(yO+Ehj$2+(-$c??s8Bk;bGDYSkg+IM~2NGsZ=M|^#OS#@&n1HaLE1Dg`
zfAV5KY}8Dsm;?A<-L^<7Fo2m8bAax*+4*=MA}W5WgDQ0(CLvB=&X3JPl)p3UbD>Ej
z-w)3NpKAxeAtuv6<{>PEwOteKz?=d;v=f!;ao#0glLp3uxnZO+i}b_nKMPvd?9O!@
zbiWyQMH|1S*%D&k<<nr#w~Rsk!f5(&+g`Ejq8ZB7oKQjdW*pu$9KxYZyWx6R-T1I@
z<B$P3EO)d>7QvhWAoZ1J6?CQTJg(pTWp=)R@XGoCmI<)>05;yAf+gM{MyrhHEMlh)
z0EMtD!nYdx&YTx^8w6XYFldg~A<pH6Hh%M?=ke=Lp2Z`}<rw|FmiHf@!$1A-IehNX
z6XD~Z_n(hvIPv(SSGG$Zf0n=m*sO?_vowIe8vkXHyesXlt*6_O?avm|f2CI4X|sUa
z0g;_I&*#iCa~!7q{SN4W<-m-_I}M2M5deOt8IId6?zxqr2XdsFYw}BlurkyCjjdOH
zNja)&gLvj^zXt$#^b@}h0O-zZq7%Ij1GSOU%bOD{S1$8PK?JRS+Gc5j*NL5hB4iFY
z3NjfS7nS^+S^5xh({^UF4icG@Xc&h3Alr;WNDP7}R^T`pf6MO(ud6QqIIo1v!FRv%
zf6CwNg8+cO6XgGAKX?@Yu+SH8{%)jM^HTK&-xkNPY*CPYtW3|=kH4kc^1CdYc>n;g
zazl~&lEnS@Z;3)+roYSw9udU<zU+Abz`55Y8e#c8(*8G2j1YO?b@jf#_L5M`>H1^Z
z&!c)DgabLM%7XLC0Hl*dkLwRHl<^FP{4l6qg+5PiEVll1N3-Z@ma;At0*_171p2~6
zeC)XRFkE=#r==egNK5Z%ADk!3NSq3Ud>$FCFfoi5br>pn(@lFJ(24QH<AS0e^5>Vi
z55oJY=l4b_{`SfSUR&+q`u1pQsNd0Mf-`dsTsYdrh2;*Gx{c6~?T6#vxOoMC@_LGE
z#<$0o_b+$xsRxhdNKv2T*g`=gYEBz=Y@kkLK4G~HxqSUWPsDiJ2b;7j@3)92t6YEi
zcx4=_q`A5-jM&yl-akr5r+VL8n%#0@;`AFjzm5z757%#${5bA0eT_0krRY4VK0uLX
z3AMRatiK+o7cvD&h(^%<$Y7LZU&38T!on0<U*5ZU6?AE~=gS>VgTlq}uv*PH^ZnDG
zKWiheEj5p?Ma(mXTYo;^Th#$*b*a)>9w$qlvqCDQwt^NLr7(+eu>0%!=kI+O1#npI
zN|ERQhY36qIE>)YgsbQTgm`}^-ua9Q2wwY;m%!Gq*9j1P0S|czLexOm)+hj{G^NJz
zyEvA~-9gxTMhDP4-{=A$${qa5(E)OP_-pD_;^$4Ae5Ied^GtL>w-w`q_b=n|^Gkpi
z=eM<;5w?0`beb_vFSSF|0iaHyyS1;Sl{h@M<Z$i!?hAltR(rrJz|UP;DxIHu&-2Ej
z934=JCgA;ZrX6l-1=h1g_iqj}Yz!wDr5+oj4BktknIo+jt=Qp|hSgQ015CT-e&6|0
zcOiQ#yt4;AzVA{q+`D#rW|e*3zISb$m2un~Isr-0;wJ(&TKim~kb1SsbX2(HI~`V%
z=Hj@rR=|(mh|vILVT_wQV(UFVCod>h#`b0rm=TfpLbgCfN>tP<#xEn+{HGIt!{`do
z%XMP(1%i^F-l&irP>=Vh?xnzs$b>d~dWVxq7zdlFuDHz-cBn{-Q2{*4=!1uj2{NA~
zLjQj8hr$5dSP{j()>v%)V<&O=?WQ2%?6}lFcSMl$Mk4L_NcImKnuNv+;%+WUynjsH
zgQg&xm*0K^0Pv$VIq#o*<Prejk>`a8cq$ge|M(+k;Iat+@-;!+<7|vEfNm;T0+D4%
z0bG4k7zQY!m`gNQfL?c|3_z<bE9<1$!5FzWzSFpV>}-hwp+-UYxt8?f^I)9J>HT&&
zBa`@Y+<y!~S>DScAO?Uql6Yw66mdpys<4--Z9gDVZ9jGZq8RBU1)TZ_K)mUJ9Ki=K
zarGtO-!O$U^bi1?3krpi2M{|9CIxS9t#)*-iBlR6KGG-zG6-ZKs6QahxqwKM0wRC(
zSd666!I$6O!EWW5)gB@TeEPvWkr5XalVVg+VG?etn|=xDadB3dBrD$HdYzn)*83B@
zvNpsUJ7c`o+gs+gOPvNT%q2Lt(7*|*yVW2y-Y@<J%-xXc?-+H;-v4Am0`m}a<S6aW
z^m${hk7h@Hucj&ij`er#s}Br)3*x5EQ$u;KbEQ(9Uxwx9t73xE%@EJ8=MBcv4}G&j
zJ&n@gZ5UIxggyjK(HJN#FbgH*uV?>ud&VlI!i?4}+UnXV5=@2qckL5AqkVb3^XEGt
zcJr4eO&89a{(TMIn)?v4$kyDKLk8fm+~wloqYNoL;sOJZ>jS7_pxL%eK?aX#3WHE#
z0JxtF0ztNm5ET;c1mtZHzKz$<k-Bu2E8&p{ppF0$d07u2{I!aK260=3z$nf`2LT^+
z1#-f`jLRwvTBd@#YZ!*g^_u-po{@#*tVO`nDm?X>2ag&cou@vA2W1{mnvGG|u8har
zPsk$q<y!+hyE?#?;r`RdzS<w-)&3Z`*~fAc;fbXd9$#*WnK1Mdez0AkV>qnU_lvwk
z{euz$cL>@B9S&|6)}7^gnCj^Ft|8m~-nZK$UhZ6Zmnn&4tyhi%RRE4v)wDi;M3t?I
z*}n>fipR`~*D!AHbb46VRr%>TX<(iZT-&2!-ceuHP-bhixf6MF7~8rMrB^?liz>M{
zXd?{yt`%A6+-Pl|WJHMY=}n$TWyybnm$>d{js%<Uns=Ei|J<foF4h)A<-kTj9<qN;
zqc;BN2d@CgSLBV0qQrOqiJt=i>@bVoOFxqFUK51=%8F{~H^heN^-iz8{t^ISZc%gr
z2G_;b+3#sKK~>`GX5Rz=9JwG-0N+WZoyo^f0|5TPUyFj=n*$m5Xtaa|BE-Wt-~sI1
z5`^Av9K{$wXX&i8w;;?!OMl_z<BM_~>Ps-#6dQgVkd02BI_Y*!pO@<*lhaclgde}H
z{b@A8F)BU^Odl%^AL;faEq_e66j8O*>zdD9%voD!|AzCT49G=*9_Au|zBUQd5$buV
z_dpy6@9bC=_9zX@Tm+2D;d4iZD00%i3b{!?j!^_I_?)CVqjT{=hKvZUspDd_0(6*}
z2OdA&M%-xP3$L$XxAOF@9=w<6fS^RD+03gh=>PyA07*naQ~>k5@X{`G0tAGn<IM~c
zkbfLK2C2u7uJ`cmo4w!_f4gOMlHnUS`}of403TUy<B4N!(U;dnK+q2ejGZFqSnk}g
zC(Ko>w)hDZMI$10Ri5nYbF05cOPNmk5cIahS7(jcF7^I3x@CpSk<O=$uKF4b$EAMS
z>Vpu3Udq!Wc66feZI^|mnw<s!hXvI^;qSGrx*H937Cp8|9|!($a{Y3<T*Y_7YTVa`
zA(n}j{J5zwl}RWt5P@m2k4N!#Oj!g}5I7E2$T!wG{gv8vc@7Z%$);A!+qH*-4~uq}
z8s^8DdZ^hxEQjT;lnlTaS@8ZRNy~(v)B_mnqdK8;3@NW*it?pxt1FO&zdR3mUU}gd
zN+sT(-uq#tCx2|Vzp&rTtn^B^+5f@}Y@!29e_@+v)p*P2<HOYB`S$evHp;*{(dQ<q
z@_`SWKZ1X7e@^k&m5NS)mZ)+)=ENEu0sejVdLRGd`J4F5>pOSs<6kWO%;Ot3`}n<=
z*YVt~L1lb1<I2PHHoa8NuY7=NI@mLvU>mGFAbV|Sw^nDC1ADvAJnew(zC+C&yu3RD
zN%+Ig(blamAdx>Of)IbMOR>SE{75v-L~0+HHe-x+drZVlEq47f5V<q%pEi3o9~8dI
z5@OgMmzjV>CsKM-J0g8!n-4w$&TMOc>&lQcm}`LElj@z$D1Z@d_R4Hr-<AuakHbQX
z-v0VwipncAmf$^2jd7D4gto$!4T=B0a6%CC+U3{f`1*4IfNR%;vYkM@a+@PT@~$OF
zyE`vl@%^j942&bGw{-GJ>31kA{-~{*{fZQh%u79o4@kYS=mFf?5TAhFDM8%dc}k)I
z9$b-Vg#YD3;)C!fSViO=zV+>=<(J$3C`JI1#@n(u7&b8i5RKJ#y&+12ZY;<s?JE*q
zIF1xRlqpi`ZzBbe4WxfQFO5We0M6bojK;<dxeruAj68rSLRK;lM95OH?Kc#;#ti_7
z!vo0pXh|RhG@Vl=KT!mnq#`p+pOJR3?F%hGR0g4)04|O~R<}g62qMS?JGj^(ab-p#
z3Cw})ip}#I)+GvyvWg>{<Sd9Z0g{tu1JV`X?eeUMq~#%tfXoA5y1Y9R@cq>u+{<Ni
zfP*dqeBMt5i4}4OP!@T!Kf%{7Z{pQG+V^j_^fQkyuk`TB<`}<tVLqS#Qk3Ru&WA+)
z>5#U1#>?w`qBis}&#2DDpvNV!wv$M`j`}3@^?Jvlo)-OqM;hG|^a3;s-B_ao#sQ(1
znFmG>uzo^?LY*(Er@<(O5S^vZ$EJ?Ud5zvF6Sr~<Rdc{Bq#uy^uBDJQ)0MA}B{S`Z
z?Wxp3wbT#?dtM0bH#{=%1tuD)(u%nrmdsSM%D1laA1Q|nz+rhuOXfT>=YviFr}MFH
z+x=t)a)j?n`z_)Ri~*(}4~l_-6H~+<m%X$CLbh@KMz|%{4Zzg;6(+$j2l-(O!KX75
zq0C|eD)r6yP_td*EXD@7{ekm~_}s<WOu*B(dicsKRVE<co<9DfV&JnYJ^ZT|SKqOZ
zf3dvUALI94UdN3dxvM=+?B;syhhyf4BrslkEi;h*jPcZ!eUICgeSQe{(eP~@zU}&V
zt-P}XNtV@nm74YKn*LK`Bkwl&fKGp~GP9p~rD=QTJW5;nNWD&Ja}Tz7ZLq7zdR-Oc
z7F7s<$hsdC3d`eWeDK7q{K#O?Mn=S+nEFA9P$&AETe1>Ac|;H@KP2>hpw>U{4{c~v
zP-Oo}-NbDQZM;DmTu+$O_Rk*`FZ}Q62W)#wC#t55Giqv9Jw^0>P7r**JE!9pq>(iR
zX%E%QA0w$hQEz;w(FX2>ApSId007W?>!Q?KmFxfiUKF2%QBRP`Klw$m;s4Upg1rBm
zN2H(eN45}qkAL^gZ(<0bI~ZaJpfwhT!ub;z14zfhWH^m~ZaUHOPlf|z02)y&+btPi
zHr$czED{I7b^|H$`yUq|fOq=4M-Ib@2V8_mZS7ud&*Fy{p&2RSPCXDiG!>CcbFoGX
zASeWqQJ|MQQxbbk{kx?Ii@p~Ub&xBrwG#(KQP7(AK*^zx`Ws-vA-SZ2Z;MBfh~@rH
z5=m4-%IhpM1J21f^7WB02t+6$o1O>We{K=pdwltgb?jE2zSTqK#V4Soy&w9?9Rvda
z(s6wCl{GxGe(>Z*m%0rsHDfF^@`cWuJ7Ww|zh}q6e&+Fiy}pUho$BJfNA!L1#rupf
z@KTmR=Yq|?@_9DNO@c2H=RHQZ7~dcM{$v<}H#nc$5)^c~w6^-B1bq^k*aE|<3_?&a
zM7SPb*bz5Gv4T1Rp44rFzhvk-zr4cpWn>^Ke_8B8!$wOvr3&g4@B)*dJHoHO+d8Rq
ze|8_w%wm}avsuS5a4RRcr!y$_zWJKt_PpHHK)-*LfghH`@(vUaiP=xRN9-KN$k0F(
zq6O&PAF>NVvI&m8N5i>-Zh+KdbO7_|Sv>)IH#15gIF}o*_&kdr<Ylo@0_Hq=vxktu
zyaEqks%}8={@1935S3u{1!OzSp~kB}AD_uuXxld>it^CKX;BE=FUyGH4b%Igf1f%p
zVVScQ0ne=VfR|+vP#jMsTbi~z7-#s?SJv?A&ZyehE~OngJi6Gz`GqDPT5e(5Jo-W1
z7)>&~yf(xu8zYGZm{Gp8(nFfH@z~KOcoXWp;VT!Uf5QFbxKN#}r<FaQSNP~nAC4>E
zj9K;f+Rm)@4(RPZ`?MJsu;1%D=>7ORT9TDbjk}-c%2p0=DrlOgdzNh@tjP2<Qnwpw
z!&(qPhy~3FKpy}if5yr8RD3^plNwunpmOQOOe)uLe|fxFh_hCNPjAxTj;~CDLVvC}
z7gaHjHoyh7%@L8$DHFkXa(a1lKbZvDYMEzyXI!qw%e1zex|QO2m4AKuiX!!Sxv|UV
z-zVFV+5$U*pwsJu{F7MMd>WWM>4?{Sv#ojnTp`ZO#GdJRnoVgZbBa9Dv9>4ZY@7-s
zFd2&N-}S}o{tIus008*#&t8!6eOsaf{>u-GP5!~p@4#ga*Vkpd-+Nl348|L6wL@RP
z!~snwNFy6>V+f!#5TqZ^AC<rQac>+E-2#l&yLs3{24G9Qn44Zg2^q&70YDUmc<$Ii
z76BRxxUJC^BAAh*561B`?z^J_?mN1_n>c-@<RhCYBB%aEr+`RZlt&%#IhyG6ztD_;
zIL3IAmKOm<BWmn|egTnn%+NJ$zlE~%A#)CJu`toUu`fUWB`0#{g3Uho7Gxezoj(HS
zfG=I1y$Fb25ubp-@ac6F70(6}kAMB*EnJx*JzQ7Xk;6xick#rDHd=8MqC;46Uxxg8
zT1kY*j<xaFu{Pe?8sX_1gP99~uihBI0XXBmj8+k{H^@Af4Up%8NZT0`5}ie3eI7ES
zAHe!C1rM^aGiuyaZ+-2b=SK$7XKFKx^ao|MQn2ltuPl)^`{i>U7)&avay|(&z0|QE
z?b0GLZzhzwd9L)l73+Ib+nMw1BB%P_sr{FNbKb>L+r2LzujEFU!CY{gK0}8-0EgvW
zEgl{|^6*Nj^Meqw1?pe=GKKA$dZ}3;+U>0P1pq=ezKS=0(EKlNhwOco=m0Jb%I(zq
z5Y)62nemtpMIJq1-gKKgz-BM3@xGsxlc1VA!OXN;K5*^`K6h~mv&u<T;H%38FqnA!
z>%YCZuaE!7j&<-~KXx1+Ix~k>!hl9Q)tmca?*IA4CjO)MEaI2X&ZA?;X3AG?4Di-K
zg>lTWgMK$=M;GjcC#WAsrTzUicgM=R)a|~@U2|R7Z@J3@Nfy*god`9o{F#hJD{|gT
zXzQx6{j(mCFIQso8lX(XON7pqCf_-F?RT{v<DD4~Kd%Z{T=|R*15nBXUMMXSrO*tv
zbvc2EoVf{@x4UJsb`iO;IU+WGfJWiE%Ke9|ZrX297}VplCJwuDyL3Vj{WmX*@XC^U
z=btzw2zXoVn{(>r`x0%8E!kdZi>-g{)I+kLwmiMCi82JtGM08E8enlwn1_vZL3<0Q
zCGNl1mwsF71-;#p`YV4c+y8xEd;@;_XN4j7$TuZh;D2^R+I!)nXMili*S_~d06-&q
z5GjC8qXQ41)s^;I6HypUGIbYp4k?7$$N;qUy5fc+qv0@@5e5)tA}`Dp{?@2$Q_raM
z=Kv&eV2tQv($qd1wC#J35FOy0MsMVK=6CvXU-Re`2SiSpHG1(ojK>*V6d^$dL@wA^
z>8m1&f)SjG5CKB(^LZW9rzMh)o9F_#f~{Z19R+AZ)Fkah^cheFB<LWN%mBG^V5$sB
z(aZX&a}pi!l^S!vXD%$^kyCRRW<(w?*X!^egB19~7jEL}9zXtP+6g{=ei5e@8knR7
zMZBVK0SfmIE<7{xC-=?Zvf<%lZJeHO;Lk2^?dl`&H#<#yrjaB5+QR!YAwpN50DZ0+
zs`y7FP9|!kOay&Q8~Pq+^mS6-f+XzM?2Tk1$Q;ne$q1v2>Dv)>8OjSICWq<`sh5FL
zyNv1@4yQ9FVSQx~t9BdixAk7H``ocTBgj}OzW98X&YM2N?#EhkC(41Eb5&o5;6k#;
zexe!s56fYB_lqh90yz_4qX9T2=S+Yu<mlAIZF64k{m+?<6l)owhirc)Ig9nqpM~r#
zqPOih%P`b11}>WZH%xFw=q0oRR;xkqp%_MjduyT&>gsd%Q|FH0GZ(5%z-KQk;eBW3
z>rIHx;ZI&#-Sfx)W5>Gq%}0;n)Iy`wf7PpOI%5G_txy(QF@E*ZA|9P@*7iT}z`x&Y
zVEhh$AVT`rJ)SqKy}K8$Kcn73k?qXWW>nZ`zxN6gaNw+qGfL7lP&s4?(j^+br-`6x
z3Syd9ny~3*Og#X`GX|TxA__7CjhgFW{jFg+{=buy8HwDMS+-e*dDwc{>Vsjk?r~vE
zCw)c*T;CL*g<~PWRgqL+LgYU~o8`@152F{XjZKB+f{z1p66|$8p)%ZK=L8Y2>BRXS
z)dhIrs-Uo|s-SmlS#12&S6EQwK%IfUD)t>;5Tt!`RnBiVWxL<foB|rB?2c-`JHkL@
zkuoPv?wd;_?!Onk7XUE1EVljcyjP+DKL2sa2Ke;9_&NY!JSV;gXHTC-<Q-NOIcM1z
zLjXxS5!-!#3nKun?jlkE&9;oo4+oM1U?TM>Cmcnhe&1B)BZ)N10Yf07?@6rL{2WH(
z3~?NxuUXyV$N_PLj+6L)*^aR@oB(hE;V-r#L6WK)Ak9XBS))=Pk+xR{zeH{C!-+T^
zW*Gn%!8y$xpo{^Ld<3Y^fV9K+YEGZCRFQr_c93mG;WUy6&N;-i87l*ksjfg#238gU
zg(uqOx&XmATSQn+_CIy*2qKsJ1bpt|5*|N2Pq(=evSM#vdu0u;8dbV!r5!o^?3o3;
z?_@{&<^7ehI7Sob@WMFhhzFF;W5ffmi_4y7?C^{CFQE12HlEp-KCYG2;kA*+!<{I6
zNQultUq+EF_z2L}&!`PwM()-Gln-1k$tyFH+x)rkqZhv~)el)3tzZ?ik?JqO_WC29
zZWgw@Ec#>xX0@y+TT(i8ITH}lcJp^BM4wD6mCNS?v+G3ZYLUM$J`aWVXE6cR_Wo`Z
zx6456ed2d9XS-z6BK|`L;IQ0NC4=Y$aG^NT2bW9$@u5`;v<v_znu%cmxNyzqHQM&e
zHvS3{4~QM5<F11Ir!xYWwcjAs+W7oBUjnOC$m^&<>Q#i#{nP6g1~PSAwx9Q%JA$9P
zZxO$EaS89gZ=rI2vHlOQY=;~IyOnn2@DCqc#*?SIyL>UseX&eH&8AJp#qv{U=5V%E
zA1w94FvI4g{<Bbv%-!w_{@Dn>h6*@fsW1TtM(F#XT5hYKgPnHZa*r?pcX}X+^*U#v
zjo#YMS>I-NvP=RC-;EOsEv;8>-xls}lyzBXy{q^h5V6x%I<J((j@m|9oBeL6jzCM1
z@VrJNFc}yTG7%NW!^;b~2=oAT6MD)8QkNh$54N3P=nGXVQCnNQj$h9E<@bs0eNivq
zPk-<h0N}-E#Fn0D9REX)KLP-F_^dDipL*mt0N{7O`XT^eWlhl9r=GY70Qk4h$OZoL
zlj6Pqh3BpS0G>K8=YQ-W(F3@)AyDD_FT4x@`1&_QiE!bJ^yizBm2k8rSqk60BFODu
zE=x2(_t!500N(SLa{PBb{r~{rkDj;!_z3^z3ol><AYM6z6o8NK!vsJ#PB4b7@FxIV
zIuRwntyKU(q|C*c^P(&8{>SCIw$??xE_Pcot_%PoI7QS9jT7V@E5oG-JdT91arEJc
z^A)rr`DP3t&O`gUL}po*KSZ?6I~Sf)MvDg}OfmZB(J$H#&r9G%k02^AOi_dZsU;`O
zK~r%z(|+pZltUnq=bA`){WdgDiy&=M6@Z3ngzNXQns8F!BNvY%Qf$d-BvD96--ox>
z2Kf5ho0wHPk;6ZF_!v&9c0O|&@cxrYqpt(+0k4<Y7NhXK`62*?(`(C_I;q!gIQ;6R
zMf~22Yv|QT55LgM@UtgWxsvx&8K<$@$lLnDH&iijN25|w>I9@^4gZ!dIwq>vx3ezF
zpNY2Hp40w@6TkyqegEk!x}$7LU;i?T;7Bha(Ft}z$Hn<-byNBo^~>Rx3kz-ipER(w
ztxTWh0BFw@z7sB2>MPmoouu$CmGFD*Z3ZlTuBLo)0`XDPTs{qxiI`q75DjMh_eO|d
zet-G9Wktph`Fs&fpZb?P4Zq)q48UQzCyOMH$$bK<%7IeI2xNAf0EtM;{^ct{;2f2i
zfS?d)5q?+@^SFLktxew|(7h_~8C!es{;#Q5^+hnRw9QE{%O}7x1hve7Wh`_&W>&6V
zSGoMf<#h~ZX!`Txe|n*@3t%+&+x#sH$kdl<gh4Bb@aZ#iwdY*RJl2bXP1yT)g5B)3
zzO9#!_f8?hy9{+}iJQCC@11S;UhxUIQv<o{LO@)Z2LxQVGUEiFmRB;?_HRo_0vOuS
zzk0&NEaezgK*XFg5uCtug{#Pnihg81g1&&(W0pcP7Q9a81_<$%X2P{h#$F3A=hP96
z%QsBQ4Yl2GZ>uMEDhOcIKPP|3LdVi8eMz&7?eqnCoL!Q3`$JjTU%n+i1jCVRcNCd0
zsv}bdfHq8OgOAiU-<e2s0*1mETz)}r)bGA1h~KwGzu<R2FIfz;FTM%@_>E5p^Dz3Q
z6fSf4<F87V!uD8v1>l4MXpH;8_MZVjI>HD*6LZL1uC2~!m+dX}Wf^+_9us{)7GsIG
z?~f*Mu|p#R5IIDVECT#EEiD2X5TAf70^&%t`=dAtasK>lH;Z2AaTI(v7|-u@=|aW8
zM56;jw!q9IidA$7T#%^kpEmgd0*^HtAr%jq13-i#bFrY{Ev|^3<DxGQz5Vqv<WH1;
z-c!hN>Tu9Ef!POnU3~va3V;w_h3~wvfvYp(?%R>WuRnYY$L1Q)1cxZN9+<3}E-*A<
zXW@A4T%kS}1m&3}tVIHnl>a=u+{UG~VXYkT8~qf=J8>{s(R-g!G$GzUP!{>8-$88e
zMe?E*qEhrPz5ez6BlAfgB#yJC%;k<j;Nhki>NWU2XveYkn~uuS3Y`DwOTwrI7uuVS
zCGGEI_vhn~y+PxDYHRy1P8p}nPy==CbC@=Jp>6Y?&<Q9oCzWy6d>4bHy775AEQjTu
zEYu0G(E)i@L8pJE1PN!|&cd6Z3rS;pwKn^p?H`<3rz36a6Htj32>T6UcQpu|3q_la
zuo4|mGj4;>T@CWj35FUbz%mRLY1r(E6=I;~Ji~0P_s96o$~I<|&s|)?sl{e(dp4V4
zbwIYh9nSRYo<;?H_(Z36oUaeksrzef_cg`a>#S|Zw7SppW`DnG?&E$--FbCPz+Gb?
z{+==cw|QO3*sdUhH~PDCg*$3X?<zu_7%x|@ysYm6kypr*t}9^*u2JESQ3|<@v)m4^
zyQ9dTQ3>Oq&0X%Fz6Je2H4xFMj)9K9X`UDxjj-QB<eF+Y8>1G6BUR|@i4tGx1mUkK
zGmzXbarlhpxTZEgdLwt4wNb}Kq%j^yy@@gbL+zhF6JSn(hV*xOUKoXqP1znt7v;M3
z)0}zL9~g)Y^8a|{y8wXy_SYn8!2Q{e0RaE#Lt?{U{nQb}-r?I{za@8ga8i{2oa7LQ
zCgKB-4AvxC01iU{&R>8Be?N!S4Pi7Moa6OHd6GZpH0xt5`VMgvn2m<|1Y|1H>%AiF
zAmT3DO)fyF6v%cJQ44a)SMJRSKokYW0FW$#<G>)vZs+BrPA_feBugND0vuHK7x{34
z95D(}DEfzptN0o?=foGnDa3Yex{WlB5NSNSCmc5lf^47O=Sk#Bi-4!!+C<-Ey0qm#
zIJ1C<Pjt0C{T)!a4Cmy&C=$v6kt<x5&u=&HQ%;yvz0ujOZcxE718{D!iL=cZS4P#3
z{m^^7xjn&A#^bB4pTAcC;Jlw^-X+4P;$A+z<@Kh0w3&S{Ra<_@=13@*%F;O(z~Vfr
zz8Qw03`&Ih`^J$`h)v7w?`irfFfrj8%yw9=I^_YZ_x_Y7_fjlFR2i@7hu1mF7nWmv
zw|xR!__xe(5J-gQ%baKL`(ZgO_jHlm0XY*u7o}VWfZI~64ge8x&Io7)B7LKz7ZzsB
z63BH<ygL&xGYTLu3bT$+XB2|3LJe}Si&h9b>J$TE_M?27oe?Wh1Qn(OrR^fqu-+fz
z`!l@$A3f5>htDrofoNqMW}p?P=IMyO>Y8B3X^%^JXkkj;TP^eh*sWM)#Y&m+MYZ3X
zDr4ZxU)I}Of0gf<Id7(7Vcm)QWdaTejNWl%ey=bA2fT0sJdBv9-q;ZCcCohq<@R*v
zC~xYsP}`dE)*%N$c|jC@C)(`!m)`$Atfh3tr0?=5Ahqc?gri~HWydIP*(i$OI}p6P
zWxtIYs4HP5t`|x@g3X?K{n9q4+4GW2CfbIR?M+qo+EOpyCAH<hu_oU0t2^=u?W&^Q
z$s<C0w&w(iF}!YFleApEsoMQRVF)l*p8&PJ$C{%cTN5X$QA6|tBF!z3#M006-<PO`
z|J6Sa0Ql|CNp!%mFG*CxZ$BkWz|O-Ph~V(_cVEO1Ks(VyKv61*fg;fZ@WTWX0HZDO
z`rqh_??5AFo^}8cFjT}g2gt`G(`PPC;hYmhRmc(;K|Zq7-fu^cACX%A9&oX0>PI=$
zbZ|iAg73h1B8Yucb7U-c5+HVn^|)6?f$`#D#U9DW*Qb&oh&};iCL(xU)ClsW{y-@^
zVC+H`L8sU4RR<s~=>z2P_2B&}6jw-0L<Yr$Xb7o0RK}wb)-w*@d2@HK|4tm?Q}2;z
z6TZhG3L=BvKV|w7S3>@oaVQMPugC--XAE*58X|3!uHQMQOo8k;-i?7<)5pC!_IN;*
zO0BYAD2lvGIt|F@+gN|k$Sln2Y}fq#8%c5Dm-~2xZYSzv!1;*RnH3ouAW@%%kVVks
zCV-byy)qU);q^^(C(Dtk4l94IOtt9<yQ+Y<m$KVbYLx`-LL)NI!G0NzcdFR#B-+JQ
zAib>@d~xXHd^ZMnSPskGDjpsga-9I{8<2%<=(bK|oJTA1??lvw39z>Jz4`=b3AX$?
z@Bbp_3FH9Q>I>9F7g(Qw$T9{7`B!}wc4s0gUjKSxRKo;Vz#opA9qYXjp1!$*-O8sg
zE}`8hOYB-<oQhD6xu4aEdv_*a)+dJ7(Y8Jig>ox+DNbu(U%O|Hr`pf%0A*&WyI(c^
z?frYV$4~biW&&=z?6=!{#3^9EJ`@SPKO2!*k>1Hsp{UqbZr_KA1*PLb51{himPf{g
z?t+T;Q@N0@DZTOO9oyCwZexb81G5s6naIKlS<lmc^YEfFgHj`l5Nz+k>vo?H4(Gl8
zuq=v8ABcf^?>j}hnSR(d)JbVwmFr@i$f$8L)-NQ-Mc<1pRTx}qi!Gnb34KL6O;rX=
zrQb-S4mO7Jx1mabW2f=Pne3nIh<<_B?1E`_0s!#R4@5WMcVdYK_#ZwcOhD%guLA&{
zd{D0A#S1B7IE?$^@9ef?`RKRh<I)%l^RTfYz7HcsVm_TH0+Z{OcyPwwN7&Xl`bp5>
z7i8Rs07P+UhqilqMKk*zpSNc05J$1ZdAlNSCCotVJrDy&NkJ(>^^lKhPsh=YA&Y>F
zN`VDL6czS+*={B3W1zYKME;R75}7LHNfbi<v8U~qSq&rSM9-i!)8~r-+Je)-EO_E-
zI0uZgq7OrEdwpSL_mBUj`4%2IH3#p#yi-A|UfPLt{LHCPLWmJ$;&8=y{_^;9?R#I`
z9ee>86_7Ild=4C(b2vFyA2>DizF<QyBk;&Y;r9jdd55<2kj>9TD}|^LBGV8>Ql>aZ
zsu$2P3&WrkNNHCtoQdFrH@LuX=Af|cPeOfdeqYm8o|B;v{7@=WnE{?%=ZmtBQg`~U
zz6ST;BB18><qWv~YduE7I+^+ayWH=?a#-&9;+>%XiB8yK!vxqYfwcYeFByTE$Uo=-
zaC^Gfeu2X+Z92;YgaxBP@D+q_koYX00F<}GR8Z$zz+!zK>U;ny$hllUvJ6ATrY`^a
z!S&r20Kf3yvKd#6uK|0hQ&Oz>IN0kk_obEzC~q@NvbKch>oyU&X|L}5N_%_#2JLkq
zdm|e5&`izmFsncAzH1-XJKOHN##mI!-o}wwMIpX#bxM7C^JEOBcNCmowcRa6psp<k
z<{Q&rES|?~c`SThtqp^Si&+LOa~iqwewI4v&yl%qYF&poX_eBVtv^IjXni7N7rcCB
zf46aG)>>S-t!#>|eoJlqZ>l%HSL8o`?t?PEW?7-^@)c3w`@`?Z_IF;F?X)k*{A<ri
z^guQdMDf;*E$wexx6nB)asU7z07*naR0YYeZ^=Y{^_Cpp-j?zBt|$zq>l*-oUQc}y
zG#UXz**_e~MDNvCVdzc)0GbWaE4cQI7+U|Qx2^#IKKPi7=eJe`+5f_0M*sjXo;wPc
zIsCzQMJM3$E3z194H`%R40|%pmtGa!f#EGtH2mnJf+UZ2V*qimVFU6}5mRm)7%0gG
z=+MwO`=M@cGkJ~P-z0Y8B`-ia#Or$ih$0M>L2znw55O$-h@1;q2P5VHP=u}0eX6;i
zYxc*Gk2<}_$!O4~pA`^#aS`By7k}aTbuQ1EXKcfAVzqwc^LC>+(x`+y`(6~z1Fmik
zu{Nqt_Va<`T`YGCUgydHIQllQUot2hR>+N@Q9n*Ax&q@dO>Brh5TTzuHv*Xe$NRze
zuD5%#BDR<EI<?<E^$xfYoiNvhJ~!%fGuC?R^tM;z9qW5c#<Q<_1T^jRRatOM9SY6n
z={oB3LWM$Q?8eF<?&$N|BV(ZTnyL>%rf1M2v#8Nk^Wwu3Wp?2=y^z`9wCj0}VT57!
zmv$GJfV%aIMab3%tlFj=t@kx3w`lvF*6us82-w@YtjYwiFuLuK0XQsw$BXC$kO{C^
z1evWZ&k|T(=+<TJt5XIH+h)H-e9lG}SnqqwAXMW0YY@Ce>U@!{k3xk>sL?Ipcx%R2
zLDV%&MJRPy5&^tE!}I@%Q&X}9+Uqv^>lg`Vj?ZK^W-$S}xiQ=w8Py^5S%7=b_4Y`G
zS3gv4Pwm&8zn2f+-paeg1k60`4u7=o86sf!fh21Kd7zJ|hJAmi2zLfz%@y$U#WBaj
z%EKsBv0eK>5%Kc@r?x+lG~>+~=c>-O6)$@#_tDlKtI3Yx4nzudw}rO<TRUlKo4pVb
z4YY&Mk2Xl#9d!V@hUu`m6!u#<pI$mCUim*(!M`QFA?J^Y_w;yPy!Yd#AdB@KLGUM!
z3WDBLq<Qv;)PGyG@muPA0M9@$zjb_0&SzxP{Y!FQstm)bPK1L<<NRY`2D)>S3YHU2
z=ZM(${ZM@Y)Ha``@`>(f{f6c&psz?e5g&o|8=_qJSL@Q=uY6wiU;Luv9C&~43;^Kj
z2T#Lg5q|t*(E*5~E>Zw)+`|OGO+`p^QG^V66u?0B0UCO9sia3O{;?_mGM;@fmik0a
zV@32wRimE!C#twd=BGC<#KVg*UlA$*7lmj9X8DVYh}AjtwB4ia7(gEXPsYGG89yns
z7&2MX_U|BzfS{}4Ad8ryO*}9Tx$U0Ag|^4z!Ry#dG(+wsF4rBAaZn`gyvG}x_1XT;
zcA7Z5)PZwq`#0Cm91UgLeeR18E7ic`lM<qHx~(}BxWBjvC@~HN#9F8V2i&X?{9%bS
z{@<B}na^ZKv{6j-)yYjWTF>Y=L^a{4ku~Ph1mVRUmi<TiUh{fl-3~rkY(~e?)Z-&n
z_H<c!VbxXLfnnc@aq}ur>W_>o@2V>zYF9w&#S;rmK-Cwe%&-P#!EBc%`YG*|zmIu^
z3X^~Xl)JD9u;VfE9u?m+d2U^$BAurX8Gys`_o9emV1`sDBAYW{tS<%`fDq4YY~#81
zzy7IH+#~Y0x&YQkz~m0F3`3<KGp=g8DD_{N>5LCTK;(h(n9dM{V%s&%+bDq9OhsPr
zwao$Q%EQMxT8%ypdt?G?#*=TO=J@XQoV{@UdwHkc1aP`0RJhp8UgPNNh1i^_x6e{@
z`FlTqj|hIx13755LndH9rH(mCEJDp!QncAjKXI0Ag>k!lKy&S3yEZHrX4NA_YqJct
zOq@Mel2qd3aG92y>(X>s_7&iGXbTC7NqXLn+Q<@B!s2<`V<HYYPlFEvBf@%t!7xrt
z2UP%MZ6cdDlujK%Y;hZEqkH3~#PhfMV*B>ZhW?d$j23ukSv2}zy?O%xkg84i=9Wy-
z^cjdWOW9ga_TS%?cF4pmcZFftc7oK$>%uhfS55~q4$h6uj@0A*8!Ph920a~z+C~RA
z<vN^P&!j1O0ROHhOvMBL;bQ=R>tB~_hM$@jg~RkS^MH5wgRg&I`W;<F3LwhfFAjle
zDp><dzrRyo4nY7>lxOLqt(`V`pJqVh!tdNs8z>QPSHc4V6VOQ<qBug5rU2B&9z_zZ
z;#5h{DU%kvK)KNNPh>4wCN;O7(+^baka%gAS@wn#4@41mlu?*(xE$$6fXh5w?1Hy<
z(bgV*IG6VuJ6HN>#Lgj(U0@OzB_qKDE-(PH>BRQzfVbD{KmGu4{%A*Z92BX;c@!1l
zLT&;bAG7<B(Ti#jB9S}abNWK?_s<N2m%o?Ne~}Sc*_mK<QqKT%BOp_6d?*v(^?6S8
zJw&QH070=&e{X%>;s&D(B(8$f$D}VlK3oW<^yXGyf-uhnNk46uz8;BT8ps6Y7cQsi
zb2Fg+gvk}a?QLb$;PrZ2c3#%QcDu%gzw319iszRP1YZpv=l6?MOh?WYD(}C;@O<pm
z6vOB5PUjA=?b)^GG(|(_a-V>J5AN)+4$EPAcZ+-=eI<(^$>-GCwrRpLUjEhXGJ>xl
z^Sb>tOhDB(Z;>=g`)#{n_l)R(3d3NS2$os826;tR+kKpN90MzeW6J3xt#bb}koqhp
zpjOckru|unxZoS`+|7RN_yJ%s-h&O_exU4e?0&p3FsWCAr>f^0=Ah<!YsWRCeS5$6
zt{Aw-Qq$hO<8?oy+#|!Z&hH6r^F2k#eD5i=(x@4Du-VV(kW`v3MA26SO|AX3vcl5-
zO;iBuo8W*v+n9NC=38akYgd_roiXE+wIN+_6Cv=8I(I~D+nc&FC-P_i{L2S0X#Q(E
zb9!Hil*mM|J`orn467hmS9rU&)b@65AfL9jdI^UKPLTnl0-Pd<RgKGE?+N9a4F#FS
z1C8VNg3#C2M6K<~2gNJ@t!skF;}hb2zn}sFoY*&d(*95t0h@|Ah9lYTw8Uq@tL=Xh
zs~55A1%wZj{*a8upzlYrAyETgd|H@*t@q1#f9>brBlZ6H6#&4eKO#N>Prq+~#5?@(
zxoa2$m~##j0MTTiwpdx^3nAhNg;mKeLL7M@ijg1#;wacA2jeu4zlXyujea=ZiA8Zw
z#bD-@2_RB!#tw+%5H&!A%^U^P5opE$Vi)umpvKT{Q2dkeF)AgF96-GR<fpVXr$I40
zo{Q*BueR_g%6($Im*@aR?4lD86{2|xjk_F?du1^A9+L?OdIib|>^4uF?KT5a$_j|n
zIb~enFfO$(zn^msjhISx(m(*T$e9R05qAJ~F6aydB+g4S1LbuNCmFs!gT*-0PJoeW
zhuSYW=MA;3Pli&D(HLO?=H--@Qj5dBSvb+RKyO&;Jp)8!`~lkOFCCao?I$g7^MX6d
z<n@PTh9U~giQYUCMe5U#ODV>sk8yq<%Qt|vbC-qii(qCc0Q2H!U{}S!^0YpKf8Kiu
z*LpxEAUv0Q?Qcf?Jlm#Oi1BZM5r_&5*dYUOSpHs?%t>SbqXU?xHP->q9kE6HMC=AB
zSM&j<Bj!rfK*cLQE}vJ+1Q?rt9m0p%uJauT&6JslFu~U_2la>_rSl6N?EZI3X(dsi
zeS_p{?CSY=XCQr;aT_Ki{{Y$RE4O=n7u!DooQ@+D{if>?aLxV7&(!nF0PFS{1J<jd
zwD)|zDfaiOT+d!VwR^1hjxzx>+TE`y_a0#aLbA8fgewO*(5$2WOxbI;z+3@J{x&Uy
z*TQj7VXhUIUq~00PUQ!dwrJYmCnc}o+_p!Z0V)O>z#j^4485yEl!Mk|Gy&^ThrnJH
z0ez-xEUO=OyV(Gj)$9D+X+h-Ay(zYGw<<_|d*gKgKpG38KYB*A@J}p??f;XHNZkFE
z6{)|eUdJDPP&D_iZx6J-Aay9ia7^QhFP)Wm|5tB`jeq5;#MckEB!0hnO4{!#!r$JK
zeioOd{n3I(6)@}A4bdMM_Ql5E9E*lOqX*JC*&bh$sD-b5SrE(Wqq6@W|H=gb!11q(
zPs49NdJ+Ke2alaa<Q<;9DJuIDBD%D&0*)h9M2xYf6MYg8EC3e?LXRBUPU@v;2E;Lr
zONxWwr4H}m;z+!>nG=9Ec;_I_(@DtQ*wl1?SxN>%HSR@vFt_C@L?xn+b=+x@VTvLQ
zgMwiGU<YI`O~5%FpE4eC<Va{}7DA1RAfgP&C4VPrsu%R>kQ9-P%cCYjcEB0BB63;|
zk<tP}agl@9`GL9{Wc(VDE&_@OaLOWp<Av%1IK7@Yk^^aoVvs*cB-g?D;lff2|M<~k
z_}b-7yxOB{X<qJYM~IR*+~?exRNfaN|Azh@23x#DnGlgDip1wMM+cdmfiilG<BxQL
z$;O)fGZw@j2ZlizK;D<Y1n7LF=eM-KFwX_w0{SLM|6R=qLgsR!45U*ArnoY2&IL(s
zQY+(vF9lGi!&0CaY}pGzP(Ze<Z`YMI#7w13N{HH3b4-;x-j|VopVa|WFTY~*Hw;T`
zu6JsC&K)uUhvn~g@enp($ORj1ry=giB6y3S%?wh*1k@?_)nx~)B6!0%aDQMFP(EJq
z38)$G9vK7cBM@08AXKkuzs4rNmj&*yprKz;{<Da@V%x8|pSBa*PE<ZnVK}Coztm3f
zs}C;Xv!}ag73wtsoL2!OUp!$CWuM=HJ-0inpSmcEgEIlM+TUZtx61?^l-$2vWWL|?
z?-3?|()P&cmCZQ&V3Io6Nb~7LC*G2GGY}i^ej+mVzn}vMNne<4&%C%HOP@N~I{FqE
z#6h10_DY0L8?#k5q~6CoZv<ENykEgPJkok)8w`C9$bhhauGBa_D#7hlficgDChamS
z-?}RK`J-YhZ*U@~cDRmbL$e0{&5NRpx1zTGWFqana!rH<28u|2_=YeG%%V5ZcCW4q
zQop$>2z^JD4Wf=P5-FJ&#t|oKW9&=)nBM;?)X->%k3ds>14g<%=!<e-b0qCfsFbI<
z1%B|{8vua+Zbz~R{*ON;J^_z?S+4JMr=|aIT>n!H;Lz%?>HYZx;Jn(<U9fqN6=`)8
znL8BY-1BUA%m$b$a&O0wIQmH@h}C;R<eCjH%7`@cxP2E#XzBABqQKJ(h;k&ErefeF
zArP_z>NsNE9#1kj=ft-qR&QVDf=)%^lnDsl@6v80=?~>sjht+&u{DaLpyXE^q-04{
zB&hFa6a0wEjz)1DaN)U~R<86$sy9GHlIs=lJz<u-0OgB%22Qdo0;=PnLoW817Vbmv
zQE-eN(DNL#Kjuh^A38=EWZvWP6Kxz#Z{fx52A=6Bcrfu84@RLckKD!{6cx!(Dq`pF
z%Y_Sl0qBLlu3q-ph&}=ux35tKlc?w;lJ`R;z-jL99tb%Bg5HWM52j6RpAQW0UQAyE
zsTX`Z7&c3Ng2E}<9m-K0j*r#U5*VMb4r5JU__u&?*<bZC;gB;B;jm!@%<g>ry2IvZ
z$!+&F4e9xH%6n6)=8y9}3VRLDXBo##S?KBC@ZBocKV$$7%TGkfAS!{WeSw3(3b)HX
z0d<OgGns&jkAbNlkm&UI{vaS@_E+Ng4KopJ^G2KBJ2U>!y@?jEN+LBn0hKuU;GAZz
zNB?u_)I4@8S2xNcK{i)H;XwIsxtm&sAm6mtQ*NLd|3AI`M&$9t;s_sWX1Fij=PW+8
z;pyYGl(qd^*IphSFulF{)2Dx;5A4-{r;Ofp&D?1wVCFbxwY$%L-(@Bsn5Jp7qLLjI
z;F$c1$T-h-rW2GeBX{l}8Jqmr7j2mcS=4A#XBes1anS3Y$ho6`hegeA<S#Ut3^R|1
zIAv?|p^pLg4+&)yL2RmxtDBUq_k(fy72=AJz71=t6gJ;9FJ95M2p~2}i_AX9;~`s^
z{w3n#3a+Pa@_+NnP2Df`KK0}!8JET%lfYmWwp+>+ym3uz<(-b$_FFAM*eh#_xb;S^
z?Wh;O+8$qBk>hSG?erC~boKbz2c`eFZwiypR7Jtove@n?J;wVhqR<cUlaJpI09a6)
zWUSY-*%MuX>o;Y*I~#HxJ6qDe_fmhbAwC7a|E%~J{LDq^_w<_*9q?bB&~ZL;3eGz;
zH7j8|aX^w_sM_sI9mx%V%mYy@NRHm(&V{UsjYtq~l4U>?DT0pRvNSLT!KNKY;w>LL
z_*9sIqcS|T)x{W4!(EZ{>b4-t5QU>==b&Uq1UD)&Q!ZE9%sk>a5^Z;<^H#w&FOB5B
z1B~WK0A&P}De!up{vO<>?Hr2m`Cc{^Nqe=mlV1Zo&UTx)(#w%@vAlk(hcip<d}xt}
zLnKO|4&a-r3Ip(76&gzn0Z-1Eme?Rmn<X$;Hk5WF;}cNAp+vCGAyMY)^pY@D<Gx&1
z(qc3YuR~f$6}boRdPU+%OS5D~s=PUv$oNNjeGi~>MEVKY2FJ3LZ7t}2GDeyWv^C$-
z>n<|{iB2#hGFaLV=WqHdwCRIT$OYg5#Cra`hHZ|=jEpE#m${OnQ)Wl!VP8IO=jE_~
z!Ia{11=Wf-y>17_hD?B|SnbbjvtG?gDfq(F#9!_ynxQa`{n~&kz1FN#GgC2G5Y2b;
zkO4R>KT*X)?Ee7~2XE1`Ehego4yd!`SE2)Id;+Ro{$|C>=4~zlaKAyob+-N*9|7yj
z5YDUfai{^@f$^xBN-IuY&f((GF6zol%o9*K%_4J59iJ_;atBnm%j2KryO7VXHHf`3
zvF{Bzdb3@(Ig<&n$bUxnmE*UI+-ET_dv1S^_3kheFnf42D!j|cey=6KTdn~4;x@v3
zCjZ$~?_wh5rFMA*N5;S-E49NV#$cikI1#(<=Ze(c@tqQ~$tS|5dE`gIXQ6z4$aWOf
z*p{@vb(7s|QS-B(UMx>!TuCa>#jR~QZ%(O;re4RKm%kl_-pqn%h=kDI4wVmeKYbs#
zo+h&~(Q%MLnWVD);)-VL+YqGuu?IwZ|B?H|OTVwb+q2hXg56YG<%Zh27d0wkUF&7r
zg0Rn?Jq-Z3ye{o`bzD8|cju<`m#S^@zSH7+aCAwc4&J^lOhJ1|kpEm;(A{gPYNf|?
znvx5;c%Q`YKYCiI;>)*$dAQJ&XoA%<qV)ITWyxN+z9GmX9SSriGI^;l?Y{WxR{;Q@
zf3KYPTdzx$fxjUiq1Vnw5c9W#OJcO63}{4Xi5{IGpGJh$tq}l+O=eBgzvmYl0GuGQ
zC<?hD8XA}1HqYsJJP{uQz0Y%M)X&uGz6FUYklX{n5oIC-=|_Pv+oob15jv0~Zm5q)
z6k{|IgbWY3IL2^Net$(FF!yDM9Mp#ZK<wl>i=qOua>^v=`=CAq6OAU|d&qILVtscC
z-uaK6UBEZrUdOa@r8mMxe<Ilc9T0~PV$LW8Z+<9q5=Vm69Z*sX)bTp!0x}K^NGVHa
z?xUbCAAt$TQt_$ad&y;3akh}fq%sm(qtRC7&5%Y>e<x;v)CFd&?_E=E@0<t6`umOt
z%=RZE=x9!lF%i4k<db&UM~OCiYWdGAlFxNjl#$ivfIdgOPYaquWI^XaDq0S8{$-Rw
zTgM$cQp7T^lspIubGk+R^xeo)NR*$SvJF00u-kRv<Ejg<oQLq0x>Fk}?4Rwe@ABvZ
zpYN-*wR;m&tA-p6W|+0iMNOH7@DCY)!}60^9DoSb_<idWU>2y<5vcVEFt+{NJKr)5
zIt~+c5S*%PI~K9mFcP_~x+a>yG6A)@3{Y-Av}2|mqa$E#<-x}Qz}cf6997DDTKU#%
zt9b*aojM(YSxDVR2h>L|)a|dd2w;kHS>5s7DD5ew?t2uDS3YpFTHd|?nYR79`*1*n
zdr)Fw&q%&*xCig}_R5{|fw)U0S$X?40Kj$u2>?J_ZB&_Aip=TCbz@UrEP8KJliBMN
zHKxr=z02brL0|)w)7+*ql|Gt@PJE*=ZBo)-+73<#0Dz$$PZGxSOMgr>nwk~lq=<~B
z6t0K|y0RW9q8;k^sFWDfTbEG`T-o;(L2vj{e;drwru_~miK_j~DN^7zC(5DLr<XAA
zi`OItek}LzmDj}U{fw@;dt28509P)HPryQUOs`N8*}DAB@9~`I_4?wL_9OjojO0$X
z)%O3sCVx*t1+^W{>+h}iX;E!~OA}!L&UX~)4TV97$I||L)R>)8XQ3s`K&!8D{nx~X
z`t}uJ5Ekaz+P~bVmR{#2CT&n7d1EBe3`5P0FmB5Dy9=6R=2K4s0G|KmOR~Kz*SWTK
z13t|VDI#M0Jmc~ETT(AmWOhWK^Q|>yYUqtlTeE7pZxI3N{-H82^l8}|$X`CE)I%5y
znPpJgW$uffB0?%grpgHLeW8!Wa8iaOqjCAYcDg-I)zm@cNS_^j4u(1(OjIedITpsJ
zzr*-+jRWTgM#n=R9tsokgGJwi$B(!0S8rG2=mFsA>)ZIy*^c^QXt7@|ZIDp{0U5C$
zzL&;^U&>-wD3R6~Sq354s^u?_JM~?0`rf%rp0ilv;ayPNQ>I4mrv|-*=Y>>42F`o>
zJb>~3<>x0=MNLfP3X=w(yOCbM`lhH0f=^{X8Ni?eKt?3vb1@cXhdu;4j$x!s#)R1b
zwVzDKJL2!I$loc0sJ=&brIIsAMDlum4aBa<F7by9Q3(3p80MO-O>F-LN>}w?rVgVr
z{<<gv-d9gwx5>D4f-~o5X1kKE*)S&Ic?6$hF0f5|C~t@N!;G5?qC*DYu>52f=L8)S
z_}1F;&2fW}tqnag$hzW%Z*B9H`2R9euP_b<A=`QtrhwHewsU)4nF(;_I&0f8#DaxR
zoS>gPH;?u%Vq;(Z@us;hQ=t+CU>Jy+QaN77B<z!En2GGUpr{Ol&qYmp2Sw(!)$JFc
zQe*e_`5f<+%Bb|e`}V;b+GD+0^>4S#d;s^`-hoH5zi$5U-~PX)pEUJl8@{zYRGeO%
zS!!^<?pV9>G5}!l#QoB4tV>iqub~skSXWRAfC5q?a?fTwc(iG4jp_X^=g}56)|Gc+
zCQ3aVY+?3y9EP}9jw>#0^L}j))P|<$g-k?4zZ^aw6FHBJOILjec%9C>QvI+p)PB~k
zidXyM$@>B1@$|vAruDa!S>W|DIzvr&!IyvtCH#!q&&VJsG7)wLsMzqrb(F`?D3i?m
zkU<whkKa<2HIA3R$3T(D5sg}i0um&Hqzp@1?w8jaD*_wq%6#>Or)B@9L>F8;(}nX6
z8>;s{+!9}e`6Y>J=yZkIpjZAt5&2}G%*TQpzj5tVy)b2VoG=iJ$Hf<czBTLD-jMq-
z)*KAV#5CuY^*m)5G;VxR6&&fk&0GszrPS9|&!M-yt=B8<wOi7ESDCaqWhk!P5QgRC
z@}gdkT*un2741)y6g!bbODuM!|0ofLB}56N<=>CL=cd}I`M5ATp%q1V?)n6Oex*J+
z(8rEPc<i`Ya2RcR&QDgMGfT_QS4HvD<{mNg3g-#FQ0yo80%e-ZL8D+2&2bS%%lFrN
zytd_Wwkwf1VO<g4?{cw7*v|jf%iUf_6wa9be%7sj@(x(Pj6C_i^8E_B27JHC#3>^Z
ztKLUGOkeuz%z~|=h<d;U#;j67@L^oKZJ(3!?p?m`<r63RT<|{3bSAaGd!_#L8<J(y
zYccwTK512EfWN0IO71$k!*W=DLW+l%#jGp8F*Xk0BEuZv8OpB4>Bu{KT+cHL?m8rx
zhlfsYw^F}`DF_QpyT5|ueff0G7?yFUbTb{nd*|@@$pq)R^%LW_Z)SLTt&YR2p>}Px
zx-tuj)*;Nw?@;rl(t3Zp;qiqV9)EfzPi$2@UIA7!#%=4@^_$n9Uh|IMpZ%6fbMFw;
zt&HPt^?RpRFn9O<B#-~p=K<uy4sr8Du)31qdp;a<g~1iSHpIAkzPGycc>1fq2LM=n
z__MOVA&6v2r8AjV{f2tg)7zX3XiNK{z*SK6<BD9zy`#1+B9X2l4kG;RQQ3CLLZ1e!
z)&J(YAhXlz4SP%_2#01I>zb`AD69qKfYLU{b5phE=k-3HTWXfZ$--ea^+pisKJ(|l
z4*+=YM}9@duQqaece6i6br6vSorGvQvyvLy8WB8^8{Ze+=a59imFrVZf%k(1O2_1Y
zpiTszFP?u|XG_-K7?Qb>^A?mbVL$vH1q90WwL&LP+Xw(y+Y+R^{>l#l0PVMhQTgb_
z2LS;8EO`SB@9^vkBJzLo<c9zN-Q$9Yqa8um7^v4tyaWI^J`f~*^L1eqvXS^=beh5p
z%<Tw*PL#13T$gs5iqJi?WNOqz@L|#Z7E_7p7;Bt&Lm7a!J~;gy8E?8HQ5-~CjgCAY
zouNEu3-j`KWlen*S~C8TG6#cgX>T=?_RcKGxXvqs*O`;+oG2qh#Y^fF1Z1LA9^YRg
z^PaxQAA4|s8_!)0Su}0=tCb9`?lC-ZpY+E#^3*)n-Y`h?*(_!`EFKR&j5^SfneQ0)
z&US6q)A2}|jUCm|SUjQQlahMiM=L%2`7QB%xi<Fr!nFjy_RtbqaTHqOdl!fEc>l<Z
zmfx4uFlS-D(QU3<b03CsKX#$vE0jXHg!{qs**@<tb26w80MrMl)jh%%blf%lhJMQZ
z+I4naJ&S0qi<c#Q>iP)`MNNO)AEvMA>(Ih?PZ_$7yzjN9qZ(UEowA?RJ>YlRCgX~1
z7Dw~^XXZIMWB?A!VJVap9%JY8BsTmIgMX!Dj`i(dzG!}G{qR@!2kH*x{h3+dhAf6W
zFGMAj+X?jpQ(zf`x^2r~=mb(@W4GS*R#7l447tPj7@S<{;A7{HVz=_O*KXlQZ*Q{4
zsefw`<UykYYTC*BsbKd~y&tXg@Z~F8r6#TnC-^rnuj^uGI>QjcruWD|%tVGWu77$m
zo?Y7CZ~yx(2W_s_xnqyJ?-~=Z&-;73%^k2LTmduAxM9M4=tZ2bAf*0qY*v`T_9fe?
zS-RZZ6vV*%c1r*NAOJ~3K~!|)s2~y|NzMz)YCEUG-6X8c<V9`kxVZA6&j1zln7);9
z^ejXk+|6-vMLpF1sAt8ix48#6E+YPubLBW$B2L=Q7uy==YY;pUEb9{yb=6qi(4{tQ
zdR7xXU)iUncKSL#Qlh*I>~BYrG~<&))<iS!uwGY<-J8@g2zm}>lA6y8nVSZqH@NcG
zdin3h6}uu-GD~`#xfi%{J<(;nF?*H&P@A1LTei#JuW67V=lexXTeq%Ce4Sr@6abL@
zwd4-?&1FIQFMrm-d51s#_SfV-yd=H=&4DTqHUuFrwohOJpcm&#ejX07D#Ar>gbcuN
zRoeHeqtN30S`q}A-FzD!z+@oa-od6j86yR-bMuy7uzI00p1dK90^`Vu__NU%fWv50
zZ2e>Xj`p{eahR;a0~ikFI)*#)ekLukU5+$rVW`m*ywACkp5#d2@7$l1ZTe)q<o$7?
zlMp*hhI0OqB*KR;9m95igr`@^AIxI;+H1G4H5lU)4;)3e5o09j8Aay+2=V=#pQuck
zKL`2yJ{YI?*5wT>wqktv{9@P-@OgAlLNF}2^5}sg^Wfo7eh)J5@$8h3KMNNDzxLn~
zT1oCnRYd4A1`jwF7y-V}k#m5{8G_t*rZ@`YUG56;?*<9yZ_9K<PGw16cV?<5`n%=}
zWcEKD+1ltVmP+=%8H_|Y&m7Oa-pljA|Iglg1>2HbS7Ixxj=kPJ+vdI3-DvO700A0s
zNU*^nL6Jicj^ZdZp`fHt#7Gkge(6IY{me)oQs_epg+BNpDHs}w0=r?zp@u|&Acq9O
zfj14b_ubE%>+f-Os`Rn<%B<XV>YaOEH|XZ%$33^IveH#mX0C14E)`%7%Nt?muLp?&
z$9>}MrGnyr!dp=zdYF7$y~Km_vmRUT0Xr^M5|SEGg5Ac>BmlECOB0f?1r{D*({YUx
z9hW<v&;4~ym^xAGFZWFr`CExV<=038e5e<X+>NB5a%}1~<)PB9Is~dnxGsud&2c?X
zf&6^*vC}xW)Wyxx*DvnicfNTZ-+pVm+OD{Hl?%=}^s_O(_0~3i=lhrOy&DJki;KIS
z(%ITD)bj+~{Q5JO5AenF^&1&LHUZCEoy>U16p*>;hyBM&4r-Gb_sZoXN)o0C29BC4
z06so(%P|O!+>TqUXzFd##`oQ>?`Ww(oj!Fs)X7g|HbRKv#}BB>Xl9y?6I8Gjk_)Gn
z1(_s@bUBYuM|(qc<20C=Z^p>Ar{ymzvR+f&{GLh@$hcQ1;yq9aMa$R#P`7tOx&Gg?
zp;paou=YF4rF=#WoHOIX9qpP*0<NhI!|_EyWJI)N8a%N`ZvY)XwUS<6ko=5Ius5J?
zQvJ?5llN#N#!4`ZjFf2Dxan5c6)09EQL;hAr0&4j(s5ns(u(}%d&WJ(;5Z6fb2-*=
znQO=Ik@rf_;@jqTOT8?}xJL;K>mdN$>}ApUuD$^43hn48KhpmGi?0cv!js1ZiU0a{
zgs<Z9pMMVk;M;dDAcezoFRx+%pgS*{fNmm?I&BMg;Nq%?vv=P3w-^F&$ri={#<}Q3
z=TE#JIe?bF_`S<wdyuO{ARmhEeqMEltvNyF!|Mmg0d%@*Ros&IrKMgOEk*pp{R2Vt
z+qy`qwR%&n{+oK=>M=pS7oL@VZjfi;Kz@5y)OKM_Y!2ET^79vf)A!zKYy>!O6SPFz
zjJyC6=Ou@|o=To58EcbYL`v;t1wQ)NX?*ruZ{bo+oB#mW9A^0ASFhu1Z*Akj<MX(4
zZ4OJ_rjJj*qTUdc{EYGfFJ9iqo0~m+|HeV3ygz$&6Q6qYxEQG<&c|Zld)RIW;w}{W
za(!qjWnd<!0RTU^+?%@i1Hcn29sJno1?|5@q?3?b;Ij8L97L%6PALg6l0T-bC|!3Z
zDkAi=bU`AKc_R*;A5@YM`K7xsk63?uZb-%kQ8SHbg>Imz_s-J#LM#EO<U{fi(FO@h
zjM!~`eqHj`ty--C=80Aw=FK9s2UVMlD7i9XcTxF{9WW9CJ_kzHd{kO9E~Z#81PABQ
zrcRjFM#T5KLI@C-Z*AXZX_jWGl!)X<kciX#%^)Ao7!D~4jGaST6vzsh+{P%7Lc)?J
z5Nh4ji7i0`g%S?ShB?)G5;%W8cMeG*5{RS-$NK$;V<+N4CCWwNBxEu3+21gk%US04
z{gw6Gw}UyIV}KvI-ybK?%r5rTO%j>4O(V3G(fLj@#V6i(2cDj?2>^htQHHOc-^N$Z
zZ)3UHz(Om<>BTl))fMH+-Vj@ZF)s9OUe2p~13Z6uA9roc$#}56KKTZ#RNh64G>@OT
ze1Jdpb_(_BiPa9CJS&^8qWn-ofpWqARx`%Oc3f{Uh3Z^&b(~OdUGpbs73yk?%QMwh
z`g7=V51DGpd$+4Obje4pxb`_6Epg-Jb=54CJ5KH>sRQQDgVw!X$H1MLiM|(f#j2xJ
z1Uat^bJPiugMVoJipd;GD|v6dUgr1)FxRHPe-X8piIN5)?30SrgDy|5)9;@)5pB=h
zri^!rJc*#W!y-2}$B{aDN(gwK2>H}9b;O9up?5{gSj{uBwM}mmA;(0{9qL~7c1m>E
zFYBeicW7QEIYi9Vo!_qQ9C%v<eZG4Ct|IAW_3GfbQv$+qpf>^A!TtrSY{}2(#JQO7
z19#-K5ulEl2$b(fkPNu!JS8fAiXRKI-Fx9a0Kg9(5FP!;{_&##0QV=a0|0*cxb*+c
z^*7Lf19k+#6}>h_0E)p6kpZ|SfEq5zF$Um@6^sEi8|v4esx~zTMsR4($@xZiUdKTt
zMB}kY(KKpe+6sQvxs#Q-05HxZADR7%T(&BV8;|iw@~PE&XK`M~RiI$AE0T%+0Xf|D
z`&^Jksx>EK`7{w^>zusvi9Y}Uhi<d<J){mjjm^+$BtVj)7-c%S@p(x7!*L!VyM<1Q
zU;nW?@O$5U8yEKD(7<(Rdz4{&l;Qj#^V{?nZ|>maQX8FSio_K_ngqT9?H}Jy`yLhr
z@WQskpI)E7_&>hV!Oxyq@|!^CAht-ZY}2Fj7En?%gjF?(l;fdYDfld8(X`C1oL@Q5
zc@8GCc2M%sAmZ1QMdL9+hhI`Dm9-b%o7PM;4Q*JWHUvSpO{5yiFK+^917T-H+n%5j
z;kfcVd^-Umerpp^;*+3F(GgVx+)Pgv7zCRP9~HK2j9O}sNBmK5O7p)2Q>aLUEQ6r;
z)+&!<Da~vEW@(lVkqU@@-|F_Q^?qh19&a<i4Wr}v_?Y=~e_ns7)8}=5=WLg;>aVpW
z&~geh5f-*Qw!dQaF3B?l`V;YlV*US}r%;UZS9t;)I;|A{$%oG2iIYqC>qs|78Gd+a
zFK(Bexoe^(zOW7a>GkQ00~h~jbwp*@N)?#1&RjParbh8|Q_fdzKiYyfp@5q54_R)d
znYY@W!}a%;Z6j(@G4;MF*N<B5JA7OXTH`b2g>u^~;-$`=i0#s@`s;fnuMBLeljd>Y
z?AD2rI>088z3M^@Ld=zk?yVbQmw)#$)zKToO<Mu(XlaYUo%)>j-{piV=Z`OFbS%A}
ze&#&p4vqW(e4YnHS~^bb7i|!@b7cJ;)roo}Yy2~*^Y?zsdAu_W##r$L%$+G0qW*|F
zB+0*EYzD}1=$ZBOKK7fstwrrGt^Bz$wbu1y+6&PZ<(HC;b|c=i@_lF+oh7{&Xj?%^
zOUno4iR-71tYN$=dK_=PpAr!3PkrM+b<+Fl@6nN13G?@V5CHJnYhMHa{1;n-@PFp?
zru?c-{D0B-=ih&ypqu~e`w~+izxa??>0f^V0I;wjUIp#0=+Ha!QeOYY7IHYG;|yZ}
zi?@qc!990MJNEX(+Wy)_sdwD(BXdA&PTv3VNbYlR$Q?XVEe_<qc1PNG`&|-M@%m-)
zR_I-q^76tVhr{MY=}&uE{Qn19Z&M`=KmOPm0Klu~#g<^8I{ef>Z}~&-6epqqT-s5_
zKxG`HHN7xrre0FR@2I1a0DcF^DL)2&`LWYDzR<$IdF=*zH9P_DR{GYJecX3^9?P8u
z0O#MAQMq2KH=K;a53lqOwfMhxwTquSx1{9}nHk8bDOE^X$d6|>Z7Pc(zd~B&7anOk
zm89rAOPxQPuI0(?q0OSs3GIk@KuY3eez!!X7ExQ<FJ?>#*CW1yU`tUUkS%XyOZVB7
z`SP`X!!T;^%Kdog#>-rHzCJdil+T3=euEqfSSIhQ)T5+61<6KrgHuGeT3!`jf_DRu
zs`wdY=6UjcPA7{bW)gr|nx&(p0yzp-U<|4hbfc@+y(GYkMqPB?OAI`Mi+?Lt``p0#
z^P0~{4(!BNnELA^1y;1k89hwQ{B_Z<_4}_Nze+U3$q3e;HAwsYcdg?;`oJA{aD5&>
zs`P>53;4$mpYWT@a{YDI|3AFa!=GH(pIY{NSG#!f+%jsGoCzz)&5^sSTrfw6M9zl=
z=81V8(+iq>{Z=Ik_P$Cnhc5S!`D(J=x&-1DEBtFtjRo%@FAciBukqClcRZQqXm#6B
z-8Sce8~fx7kYge=e%PKEMQ5IUT8pzKdY$#H&x?-!nwGn$k_h_w_Y3j`NdA_#zeCD;
zGr+0e$(DDdfscU5eOIl>Z(9{aM1)G7f@=pce*uvzZ5-$|5ID=(zj&u=iCW0vtkokm
z4zX?HuusXtyh<84?mSQA%=QHYrbvNEUTrb>!I57fr3hQg+0}8PWM!o7_TCZNPF}yH
z$Z?@f&U(prMmg%~^}%lOb)l_6TRY^f`4!pk=zHU>nEh*}5(FIwiq2_^uuVyqMzLE|
ztNK;t@ZZeDiodrl-T=S9Cy4$-j|zet{_(2-fDhdLH~`?gXIGIF4&Qi7x%{ypE`e?V
zIe-)?G6yt^LKg34L(Uhnb;w5<asb=v-5~D$5?6i<0J-E7k8_7&JVvJd90N#-T;Aba
z3<_Iwb0`ot7Ogpr@~GDRBQ=UE)Fz<SXv+PJ7U<2$6CeQ+Wg6r>Ktw!N$r1T4x=lfP
zWF(}Q#sU%`Nd&PvC)Ue}Lqc8=2RUCxaEbHCG=amhr7nK$iQDk#m3@5ewHtVCM<e3B
zgVaeAeBk6F?pmM2a<_qDmPbD2M9hh{FDpsBp1HVxsKx*B)h<4KZdsnUftW=ZSvw+e
zPq!41co{^}_5wt}!5}eO2VHC>@5(yjIGg|8a~*FBxJ4inF7UFShT=cIsnN9W1nIj8
zNZN0d^SB?rohn(%OEw731<3)&$zR9hl$P4)+l7)OL(fz4Flj^OP#uTK@`-X{j(8n1
zJ{CToGuQK)iujy3|9QR&GY{4=4Ki^YW?lfZG)uRT3W$aN*vV$edI|7%=p0(Q2}n`$
z0&rs8Pb++uL+k#EcYx-nw?VGr;7|@?Z-9cL<6ug~>83pScx!*}4WPMP0u*`h0!R`G
zQC$*;OEx|43)jg~^ZDyc-aA0^(W?Mvu4DPm;qK$}ICpFw7q@zN{^BlPxVDeY!Q_y_
zlTs&5aPLYNw=Z|_;Hd=^MIk_BqSZa?*LKm`|Kc9L^!9db9ZZj}cJXuPmi@8y9+*~7
zbpU|!!<%lcAK#uhS8<J-n6dHzO#AhEcz>A88C#BN|MX-sR?4&Eci3_dnWl}yyHws0
zGW<wWgGkA9r|V9Bf5^Du%220UrBf>mI4wW)8(*N#n>!RA8$c1KVX&e<IT5Lv9{8Op
z>)%j6`6WfFoBh~kfx3UDVD+fS{#VWnd^@1+j;;EGt$}t^3!?V`+eIt>18q<0+aLdW
z9kg1@dn<P>xTpQ+crng9`4))a*$zY0=ymc+aOZWee#e@#J~>TB!`_gR73oKh@!a)x
zJ{L+D)*G?y#Em_ffN9&%G7=9;2B?!{y=;ENcma^<u_C$Af%JR~+CLpmKhB=<Pw%I$
zkP-k6KnTk#r~3!$zugmSZpK${CSnEuuSTLn|KvZDm;n9Xc@6;ZpZ~<|0D!Ic-#`k7
zXP<oqLjcW#4n_bv`B@AAbg+azfbr&K@d{{mkOAl^_x{0wEEbF^NG8CB26oSeqFc}R
zHvM>v0W?~3x_FE0;8>$z_6?#RtHi^pmq3w^^j8Nk7X+O3r2fmBf^hft<hZ4A<r}G!
zmplQGI5fsGfWcdawm@S+Zn{VfRku$o_5uKsB>0y@ek;L=N{fzjvWYpp+QsdwU3>(<
zjlBV0ySk5q;TTu;2G}2!aXfJ6avSYNisSPw9A9qx_sqW!`hIxZ5{3SGwD>qQQs8^%
z55M?-<ZiL8AhHcs^d+S1g71-Kp!3}O`q%2-^>)7ru*ln|QaR~_#fwMuj-lkEBy=d%
z=e)$w&+pM^8p>ulhEBl$@4W+vKqm&13jHQBkB>`>rY~>udgCRrTE2}cTLIe9_*)g5
zd5<R~Tb}hcm@KI%+XPfD>3u}{@pA~O7OmHeukDA`oXjKuvouTZAQcd_K1Pld3368;
zDI7))XcmDlfX9>!1Ren45%9?0NCG@Bfd3WbPsCn1?-4(F0c6q*d;+oL!TS4KUI2>_
zol77gqveFp?<E5=+a-yLZ@oky_Onlt1bLo^#Z0feL`jA`GH1Op5?xzpV{N{LESE{Y
zC;tdaE%K7ekALJw%V$t2-ypK@UD(ARzqvK_J$QVri=Vq|r9OkHuOczAl7Lzy&3rc(
z0Pwk2`&ZXK)2?aluORhl$lQuBj+iQffm`G~c!v^%Tb2Y&FZpN<d6$aw8-SEM5Z&6X
zE5|q!5iz;z=R2|A{XBKU+_AC(8V%{rgWU4;ix2*pM5>BJwH%gDMCz^H8DBh&_f&j1
zYzGAd9{K4pcT(Je5`j~P&5e?c8nxgV<%lrVi$U#fdmcH*{j<7kjwg3!T}o0I4N;F-
zJ|C3dfpceTq6m5khPIbFS=LWPzhI*35kYg@84G}31bmlhZGYRcydT$llz_<jE_nyk
zR)V%3%S)2~%I+xY-=;QzUM4Fv;?VZeOT^ojXhW{41cL8@Ytd#!$Bi~C+Wuopf~X6P
zQuj<>?%z=k{9#*c0OlKld=upf`1-4Yh+o?k8-U;X=ve?j<Et0s_`N3p0Dt}u7T|J+
zFTW%bg3VX=F#s^1-GLE+(P)SffVs4V5rA}{`s_md`3E~95y*BnkpXD7r2Yo<TN;ba
zK%*_Xey8!gQx!eARBQz{uWf5T1-Vm~T#Qt$pmEAGz1~?88waOIe`i-D4vc!2h-12!
zz;pn>B~TLsY^u#rCv`yLf;Uf@q0R-SJ$`@WgJ`FmlVua({jW81K$NmLdRqh{Gg>=q
zJ#s!p?!6S`50N;n?dcQ4`Hay<#k<KN%X1O?A7b%;Vto!zD)JBh<8qDPIr$S3Zl<(f
zay?#@Tq&Q`@pBX4bQ55GB!V}Q+?W>S`@`Z@gj2rGax*FjxNJ+{+vLgz(mylL1__7c
zw^0Z^1E@i!!3L4*f{q)#wOqJXPJU>)?ywomU?m|2nOg*Jt@X#;{FrgH$i5<3@a>B0
z<@Jd95Y_QF4BSCd<&k~7-ei7#iJ`w8v6&aZEX~rpkO~wik^-5E2&sag!iru3)LaK1
z0iEFNMp!@C;9Z~-DdKz~2#dMj<g-x$?F47<<AmWERIRAi&PZOhl9*#>^4aU0g7lc~
zs+Ie=lQvV9-KpDS(;r`tE!St0<D)G8AGv3>TCP8i+gb63@Rx7i-l@sTc0<7?o6To4
zOsi&EhAHiGcASowrZx9nDtBsmw_49TPDiYFN;@0;o)TBqg@p$c?fd{_y2JNYh-L>u
z?sn_GJFou11F!4V`}Mcwc?0B_R{zxH(5juh1m3#cSdFtCEx!XVArX|2cq2y2wEp)y
zWBsLHyyq8C2>_AsvU04`G{O6|r||?pWXlDiYnc9sY=f085quH#r!LwPc^EjN-nu?7
zpAEW5kBD;S(|U9q$<vVeC~%T*6e}r9e*cmV)K-E>JT=de*(oWwq!NLJLUqVw;?nZE
zdINQpmo-1R=GPgELC+UzGeQKEsR)B!8rzCwsT&_VY2U)qNdUl{LfifgZRmCRefz5X
z{x>_q;s4)0c?JNm`Z?hP__3ds<MyKqK;iJeUw#cE07*|6|2iGv1<12ZBm>G1;M6N1
z8>r-=J16~6M!;m$!w5iYMI;T*Q9@Fvw?QJ30p<K}D|KdXI7ES9%}<8T&b;)m&=0Mv
zu>i(ds^g~LyGj5P6)~lWAbh^}MP2}D;>R^t@p%ee)D?O2ec|(`M2b3hisMt2SC3T^
z;C0R@BhL_Hfd26GgGT`%br?|6p!xY7NxOr!JtYd1Sa=%|h}i!yi~pmGz?1i^qJUw{
zb0vdHCrI7<#Y3Vf<^fvkI4hb?N{G2+8(|Tu`3x)lb>;jn+$hZpNSus^%unQK=l9rd
zF-$_0FwzF{jt$gJ5^1O9uCGm675(^0BV8U3S86B6Au;U=`@(qMw~VM3(HB~tv*WGL
z*-KQ+^*Us2dtGOnQRjt836^J2M4YcG+0j3yeM~b6z%0$uyO;`>10|jS236uO8U6w<
z0Fk^-OnN?Cj}f|O4m6Si&nsZg7iOaH+eCx#gG3<SpjZ5{E%L8J+BFh_Dq^?g#K+DY
zS0o9h-or{NJR&#cfA9PbzWDkkrlm(0fuA~2UKLG-3iLM_Ya|erae~7jEAZtvcD-6F
zx^9cPRY^ddWS|xqPpNNO{nN^ya)0G=?f4(9;M%+DhV@4#37FR5cc-{UUv;btiqyOg
z$?R3wQLC%5J0|Y1wsa>>1bco{be?ydR4#S1Bk@dQy0sz0-fil2KpnMbMxyn*N4n*m
zbeMmLB&-gd<@xAO+LE^FzG?Np(xn%KoTo07@uUg7gN~8?kn;ntTXcFH@6bJilV><o
zB(BJph^yfz_&5>iGPZ)<akq?@#8{t8&hgaqO)q-YeB=Y5jRd_Gv_C}<{Yd{B?2jC8
zYQ|MH+voeiSOZ*mjwxdI1b}8<V1HRIB@SdRWd3s+)seboCfYi9hDI_bD$+l_AxwNX
zdg2W*x2C$>>~86=GE!#x2JKy!@<039WdOjh|As^j?EatM0|1<QT<&YFbkRV9b}kZt
z-sS7)1DH!Cdf>*!T>yal?-1UBAM8r~<DFeWHKTzb`#Iq$7!L=k!*4}A07)^z7{Gi>
zbor-_Oa9CH{PwN~B%1^1D-+`Ky5!%wbOQh|zbul6v!_MU@ba5-9%HfCkD^PuMC1<6
ziR3K>HuseeL7&H|cA|!T#AuTEe7yfR5Rg>jkZS%h3VFC=F=upIcuN@s<uRt-o5kx!
z6*(uRFU2K^N8FTX4ZS2nZtON3F76HvwfH}>(!oz1vz!D@kSXV*$h9!I;zfSkB}oBC
zi1SjaJORDo7@vQ63l|0%F84Bg?4D(`)5N#ie<!&}8X77|@U}<1u9-c5Kp}Iwr<N`#
z0W8lm!4_en-J7c-SZi&b`gyD15NZFQ<09qoxp?M5l*W<g9V*DzTlafOr?%X=()gDA
z<4qSP@hLcz_NnC?TuCQi7A?3*12lQXtuQIp<g@cQvxTULdX6?){yFOXv_0{>IFp++
zlK{-pEd2;l0U$$O(ftP-07*F~yeT^WaNb4dD>{A>>u^Fah~7&AIHA-a(SpH=$`dLO
zfT3XmtRQb1B~>H_H8ubh1Roy<K<p;}BD(D+5(i97mv#pDy&qnge6f@sS?S;(oLj86
z-`4M&;U-^T|0<G#c(cUC9{?^4GJO8!ZLjt;71`S-Cu`d=CA(?vgyPIr{gCxnioHd|
zUMoR)r{x|o+q-4caMV<Z1#!gs8q`^H=jl1>y{<35p=}%ot+hDQsv_r(UO%=Vh|BAS
z&3AK6nc=uoC(_;>$}v@S^us@)oc=^6j4MxTe)f}MGtV67*bgai*;t)>WUT$UsJdk$
zPUi0$>veD4uJuzl9d?Ftp8oDTDq$c(rZGPeH~sSIl@MYTOis@1ZP`BO;^)c|7CZ9G
zxuoT7Ds<tYE>OE+Z4{bD_s?<e4dpp9IwFxC%cEp~=P6+s>xxKjf0oy6>GM?^4N6A%
z9u!I-;Sv?Od&v+NMC}J945};N*^<frjc>mp`MS3O0JdKJiqz9o+mdxb_^l;D?4w<=
z75LVxe+d9soEPMN+mGK10QkwPa{NC&Bs%}!9sCi}BEfg}PRls#%wquH!rT8A0Pxm%
zjTT62`o{aj3!u@G#d`B?0Kn2|X@}>kS0vwE79_u`od3_fDPai{N^sO$V18L7ZyP5>
z!tv&JL}GFB!G{0<2bZq`0KWW(e<uApaVG#^eFFdr%<1#wd(_WF0>C))$5oO;uYk+y
zr9sF;B-U0wfm{*tjy}hMippqvQt*Y)#J8<Ch~FQ^uV3&IGtFt<*HI>}i)=sVp)4=Z
zALaP{XRl6;JMie*96o$b#&G}y$wf)`PDztT-bH!-3SVy?S^pcd9w*)opM7c5FaAsy
z1{ps0@-}`=B>|lOno5<&YCP|XL8feh<%(z5v$CWIVc+E4DmCBGSod?AnW{|&%i)9Y
zk`GkNcfK7CF!Pw#{YR38fb5-~O#Ne*=V5IE#zqoA$wZLM7)b_yIZj-+Sf~N~B6za1
za7jXg(BJZT%iqBFCbb(_P%<P(VSKetKOS-e-}B5!5(=|XpGg2_X_kJ3sQ@r`x!2`W
z0uZeCb)wc||H}<2^Lga26Sg@H`SiD96JX2J_EDtjug3_nqBEb0;0@wuviU56uhZRI
zWNc<oPaIXTMc{3={iZzU@Fy=^#pWnpXl;6QeI6gaXAPwp`KIl!N+K-Muet67`9JmI
zCNA_Rx9Q>}NkFw*aXmKK^Y0WW9~A-D0`<cp@tX0jYtFlk&})mEq$_zh(zI)=BP;*_
zAOJ~3K~(nNTK^hEiav^;rlF5UHF?KicZ4<t-HJ|F^v~Cw3w7g5K5AaPV_ohl!e{(=
zMc}mB_15pYg$fLPZEe5Yw`7nmZ3}jFXFcC;Mg0x74Dw43eU1-z_>42pofxABF0?rD
zO4?M(Ml)?j_irCCsv;#Kx?`ia0Xg=kSj%#p*?w<5Z)^_2xW_wTBKw3<_lzGs`4<?2
zVWbiTA_(p@hgvWBA&9*Aez6<JRC04&$Hn>qcNje|jU_=unCwSO{omQXfscwvZxAVu
zk`wYP_&GJ#e3>lCjZ6?Qy4L{!h3f7bbHey{&=F?9=0KVIw$1?n{_L*=p^r3z?fOrP
zgy1Khk*ItB^mYkm^FM8DAt@ZLot78_kDrknzWS|~gvk#}!u6kB!x%s-7cTk238EkF
z2~R=0A>)wuMiK*Hz9TjPAs(B02aHiTU_4N!Lo5h&@kne2B%o379zY@91Vt`U2eSj=
zKNyc=$yyEoNupi^4qH7LKat2etj$ZlOWVRHu%R(M<h`bUzQexqFi-_Tehnf-S=bYe
zJ)&0Q&IvL~gEv4*oi`vh0xHR(e}85e0u#MI8;b<v^(%XLVe95G1|D6X$3MJJBsy6>
zLcm$@%2kH~NStTn^#0i6G9q<e*WFa;>!a9}u7Ba>`o+JTdRc*$HW>}%PZ?1<XLtw@
zY^2ELltoB7%nMbCKId~e+0Y$gR^C{a7TI83ulZC`W5eZEZ3eVfbDs0Lk9FV>BYU>e
zkH6DmD1nOGR~ShIht1j`IW=6c40}G)#OZm$NwYo&hR1lF^x0jUkA<+^Vcskwe^+j}
z{1n{K$C4Os_-7J;S(>H4hExE^T@kGMow_4BZroI)t@(pDK)kV|7lE}2VE)3N*I!Q1
zoaidLemC(x8>J9d=Wmd2#h;(i04*}E^Y5Q#6JU|NMY!Jh)?DxlMtnaP!f&77!OhZ%
zZVNwgc4dMD#AK}4FvRD>VcmEL^8f6MH>NKBr`rvD?4A{IEuI4W>i7<Ba>wD$9~l|f
zTz{v{nv(Bbi{^LK{=Q4q{-_IXP*)!_HlgIF86$Wm>wj;P&+d0BB?N)8PFjx@*%R^6
zPoDnbEN|!$tsL+5$Y~!-47`6YJ4%9<7P<e`zuhxSY5Pn+iCD=vx2w_HR#bOyy$J?7
z<BoNF))t#l`>9JJ;vrMsk{(mXZ6zypG@WHbRN>mThwcVR>25{3yIVq9N|5em071G-
zx?4)=8juF*6p*2Z5C#}v=y}<Dzt0a?AJ<y<eO||Loe^Q@;n}{Im?R9EG%*>D>9UAG
zH9~f*ttW+G4i?<6;fc##|Ah0Dt5BreqzKe>dO$4)bg8?Rqv<nW(yG2D;XHsE+eium
zL}%C&mvJJYx~m`MCs%?sFR0w(KkfCb5!yZll}k`bb7Sg3y;*>7h@I~2eX#+vVDf_V
z&-&hvpohm#007$=$xz?(urL5X*9F1n@UuZ_a2N5_V8*qiJ;M>OV#t5yed!Jgn~b*J
zFT$Sp)9MtX_VVoIQIVRlBbhynDKA-H_C4+(5Ef>+-!kq&_gq1xNolLFj6#3RF*P7U
zKU)mtlrzG@V&TRcJQoIYss_)aN|IGH)|tkXfxG>{sSoUw#o1bQLpWRWLrl*@s@r&F
zT+l=6q3fSVTid2P`Ps#IKWS8GayMpm_Vr@6!6)4zixpbATe}3Wm8&u{4&;vr!fTBb
z9YCv=q34TFqYtD;mug#XNi)6PDCAW<&jsjG9?;t4-ffmDyA$^&k9muZ{z#QjBUg&D
zRh;LB{s7C5$nP-ycRcsxk{!P+mUGt1c&ArMS0C{aH6~JSHNu)R5*5u2s-&A!F&|90
znx~!a#Z}?S$BvN$odM#Uzhk)@D`~T)_NN4n)*de|6U$>xE}*-bro!qk87}lQry0Xe
zM`VnpFem9LrD~X;=@MVtFOsdD-3`b>>~AXZoX0G-VrwQV+X36-x}U%Nz8>+Z;#s#5
zM-9;gL6HRBm=uPInIdmWPl*HisaNpbe_riYWzT-}=3)A94s$A43gFpku%NRQm4fdu
zyh8Y*6!rKSESixxW3U?RjyiT#(|0wBh1x&X^T=y@WIje~W<Df;`>b#bZOLcy+g09w
z&0oJEZ@7LwHXe`MIcUUOq4JAr?8jqPtQ5Jgjsy?FnHDqcko7XH=!5=<Af2WLQPr`p
zqozh(5aU$R$7-jeH$;^ylF8Kq^GQ?_Y-;Z)vI}Dd)b|`(e^>i1RyYRqBzFPT^s2pF
zYj^<b@(mtm8|<>@8jdZgI^_!HZB_nLQz)}TCyR-H3W`4|IXBUa%S8UGFF67_my%sd
zu#ngWZs?Dru~23BcPKsfWcipmSpU!rMLPvjkV^qw{HK=N>cpG^L$16T3Yg?8`asU)
ziND%gHsAZK#ZQ(_N_)UX0sW?Y(tP_xPxOG|S>KA|BI$qeT0tF5Lda#|H0@oZNPv}w
zcq7?|dlEnxFx0^NNjf0gpZsy6z2b49D#~vQPOa`9HQ0o}p+GivQRNORm6k!b3bput
zy`aE?CqAF#yEzXly{7vY*YG@4pf@v-w#sSN$sYQci1K0Yr6JL1vxE7`jHKYE6=%Na
zJqs-~K&Iz)gW0R+Fh+8V2#RJgKBAq6Ad6lrd~c#|uL?9?amQRM-6DvqWhvU_lc{SE
zCTklSHc@=3Os+_5(ox18pU)wuy{aHPN7*M=r{33pyZoaRK3pEkg<@<?od*6ze69<p
z6Yt((d6lEad`}sGVOBkg#s_$d^aB{4st@Ahkv353u{PTxuvn8Iy+qvh);W}8jYP@r
zB?%Fa1cHyG7VdzLBS|o{BDzVZSx`@kAU#K2v6<C**^vG3f&7$w<=<K?NYe^L+>`@Z
z>r;&8su2*_D11FpEpku&v{mzpT4ZphAJ20GALYqBQL}%KYDm0TDbke4IU_Zw#)Nv1
zVeWaZ?H4guz&h~^Mt`D5h6^jww~7z<c`9fM#QcmqM!HAU65G}heqpjVH-Wkq=Mx~o
zJCykqOZg!^*)Fob5~Z+h;n*|SU3!=2?>F%Uo|48cn$Pu1ohvaJInyt_jx<A{m}D*?
z?C1N8`|VjeGXML|la5)sZTqA5RpH6t?pJ0sZfnK4a>r}>+5fz*qxFw#xT?->o0L$Z
z(dUI{jE|n{*OV4r16|vw?{&-bvc(hGie;VLj)U0)ceN5#a>aZZCADy~bmPHHey)WM
zk!fgAdl}q1+57dPtyYX5NA4=g?vZk^TUUQb^AR$m%DdI`-mzm9OG~eV&##?;M?g1c
z0%Pq4rsEt~3(eUrU)?n<$4Ay0WsUO?AE<eU^zUnm5K;K9cvkg0JvByui_qw=9}>HQ
zptN@1ZU4}!nK6EC*8>hFZ79N__%_sdmCck@e-m~aPZvCe5{IJbl-h|#)mE1rA~HzN
z!%u6;G1%lyLKd<Z)&Gg8VL$gAPZiId`36NGkJg?{`wbNp>cTvSos;G+h!wYZWDn&f
z!Ok)Gp)GuAVY_T-0Y%;Nmw(rL2>|E!a0?{Bcc;a^ieZ;`0NX+JF+llKIt9|$&8SJk
z{X=`!aE1FU317I9{5Pa^^8}<Tg1z3!uktR@D(IKn;r!UF@oC`kJVDyHR>VuDjd8%a
zy$T(F`cg>KEt;8d)G=o%uFTsNkX$WTl+>;=;V?`<a|ol%_TJow*@@w2i`ecp&)TS=
z$N017U_YOwts?@Tu?qaIhV3=J46uV7->=;n#tn2-^JfL}75_L@-wD2w<x|zS)ry>k
zWw7J^q|ui4JDrUOb~70FR69?#gkb0JSl_J3vx<=S5Pf5PyWiPwq|=s|W{jL}qWg5l
z9(xW`%Q>0_(>ddR*OW6kGpeC+s;5()@*?AlP?G|0GERIdd0K6k7b7l3VWi#r5mx+8
zY7V`dnd5RPg9und@d<r-gg?e*$;_^3sE+27T491CSOb|~{pXc%j%Q4tw15Os)|*+{
zAU7l}TwW>su4jrnQz^hSMwYI-^QR1|AzJS{33IPr`~g$G2&%t*>bo!l)5zTW0*<1i
zS4bbwkb+cY@DNC~PiLP*L!p0;;t}KXJJKaAx*^k7h>1|vmPsgS%`(z?%~)S$pQ+NA
zAYM(k3zZeHof%2_x9_9{7W93Xju-HYN{YChlgl`}Zy96XS?5w-I-GEEKrh<c-u!zN
zWU=U>i?PRdonC~;PZKr=>MbJN{yZ3RB`pf7(eMy|qF^9qfWvoR^I49GziW_WlR7Rx
z0!=^TC%}<9;Q`OJ$1>XN1mWbp{^lL@z~io(L5p{fIL_oO%%{X}k=4XM)rW1|bWXE8
zn4I}aC9TQweHk`<45pmyk59Q4otY~t<TN>jdx|Gc6G$?Z=67^3;V}%-jlwdWB3iWh
zup(G3!`$%5`PGP9EQKYK<$j+lsiL?L#ZVBCGNds*w~VA2>tQX+YO|=W@4!J%UGx`A
zB&83V!KS4Tot5<W=>&a76DyKe9c*KP0jA7oPgyZ>&@N5yb2q66lukF@lm`dZHN;Hg
zSWv1I^Y!CBSt{w{47?8=aC`ikso4E?gWPAd&ktEN%$4;{>u(6C1zOwEF33h)nVA0T
zFVOv3Yy{8-e)SsE9`(&s0m<s4;`c|W7?Ti=^y8WUAk4;Z;^_mU5GLQ_QZ~}d@r%@6
z{?fKZFxNu96A`bA-apVv0>B~;j6cM<{5qmj+>Yg9?2xEEzJS+y?MTo9zW;VK3a(Vp
z)3P~{VJ6M@b6C=HLuD7RECm&cdgP-x5__K~qJCQG@-G{~vFa}RXch@izpm(f8;%bU
z+4U=8cJDmLX`)qQHsJ_cERm+#g+<FeSBc{VABW@d9hFBm@?!#x56EU=Mg(sqo{nD`
z=5d5xpV_s{Gaue+21xm0?Z7PH8eGAy+`UFR+@iMoo3A(;E&QefV(EQ$S_DZHPA76&
zkv!097NV`0vj%eMbgjDu1Y9>k;C0LlW-)d}*(ICf*+KC5PTApI$Gnutv0_T~t--ju
z8POTv(yz<0gfZyox3=MC617-lI;UOX|AkRJDp6bjqubcHA+<u$*<sM8dA1nv%Re6N
zFoIXsp<bTPD=O&N$^16VPW4YG-j8Io;Z$|NRnLJ*6D{kBys+<VnMQmAfN2-*V&2|f
z@fc{MDp{qh5fhd>{*!J}6+!r;_R$X5k=bAqb#+(7{7?2UsEE8nWb8c3sO^cBq3eY|
z!LN&xey3OXHH^SKNg!Ie7r5pz_29J_adK~Jv{Ao?+M}-;#!<YT3QUv+wKEEnwYxS4
zDxv&*gVgcmj~+8{w3hxwB_o}`=?9Lb0jmjbezhG-jb2PUXni3!+ej;T&b_Atdub59
z|A-#BBXst@bD`AMIwbty>H@#LZtUA8X1;&klz-Lm+L-=qi&&7R;jcG|7*<*(I_mYE
ze9&WJnSvaQ**02H`&*N?QR-yM1DUgoph4qoMGh992L&q4Q^+5mMDe20G}43aF1Aw^
z{oUOAR8T7i$%RKI4b930opGgF2=|xp{HiHzzh4_3K|^K4X<3&@Ez8WDmZtTmlu9Sz
z>qAArjIjs**_TU)<<qS&Tww)4JLPv__u+jXLf#SH-WMc0QhytLS+_5B=j8?f(B5CF
z1Hvr7kzoDuJ9A>>ps@&jt`)M;|MqO6K88di&LxC)s&L=lR{c9$oRO0K{(%C;VOst9
zCSPT9d3nE2d|C0wg_-`DcqMUNcpu0u(W@0N<(Zqu;c6_6JWf5oPnhE})%aQ*!0SW&
zh2CdX*A`S}nrfrYL$BI1q+VzlC)`k59}x;>Yn8;<tYlJhuw5+EoYHGY5{a&0o#bIS
zfA7<l1g#p<^^289juu~mG_AQ>d+C^`w11_v=xK=_g_61vsnG{dzm4$JVj*Bu_2RvN
z8Jw$8?bdv1+_rY3x^&wiA#g7K%RDL`HuRAufUJX1Gv(Bs7`T08Bs|Bb`HCEaQVKC4
z6ngZJc&!FD)?2_S!J{g=wdFH|Vq?Cku^(vB79t}W=$!xk6qUs1O5ri{`nR0rl+gi)
z`<iRV53S7&W8RL%Muk-A74T&c(tDRu=f&aAHt~nEe-g<M!m##*mujD7CZ2z;5j<Ss
zVOYFdXdHpgECT<VaezyiZ6%}N*u^e@ME06Ig0Uh<$=xm(8>N^LYQ;r}K*y5lc~gPE
zdHj#0&a_?SFMWZDMVLvsuxJt9XX`RdbZZLzzEo6+Ku;o{BWg%{ude%R<c6PdeCujg
zN;VSLgH9c*)I0V^wpHQU5>4S+0^I*JLQNVe8z+O9B%aiIZUxM#Y;Olw{Cc|Hx>i4;
z%H?u!r9#iP3C};e5OVV5#ceq8E(!rnK(1sabeDa8=EUdtibxxPnu%3qf;=~E`_f(f
zje94pM&-A9D^+KRaOwV&mVI}F67F82F_FU;sHr`7lsHm9j&Qtbg%9eDM5BN505d(F
zITd{}h_2ti^`8s`j|J;6-Pd~ou4Q#rI_rN&x}4HiDZfaElq}jrNwrcRA&PYMTm=%d
z8Myv`i&OWEE9KOXMzNA@P&JiCH;i+>Ry%#BE=E~;+lKX6Oz@6s>s2quRluxanAi_m
z3_TJ`6i2Rs3*Z(ri(uM4Q(yF@__+n=#P{kuWVW9BD3~OZyp{aKYq0<ZzfHoE$AP57
zSy<NLZ%d%5$P?t%_3c-k(XJ5L4I9NxV96%x7am7h!?9oU^hN#o^noL^w`IM;6@TeS
z4tn)1yI%SpW;U79q96!ebW*GaGc6?2+ZEs)51>C{jmDdt89Xg0dpr2r#fXWFv<rx0
z1K2H8dB<yixtTn6Se+K};yn6ZyTxeA%lA^>_ztmJ3D|KXk-H#!Dz2gI-&FbOSo*-Z
za4se1(b*}gR1qAy<*22y>m(t~Cc$*$@Dd+4cw{_S{af&&0ATv#cam&Ks1f>r98jJ4
zA0TXW=KIz|wy#qKz%uX*J9K|ee-i;V`544?soDiu-2P=qU>WFhq7#o8Jp;?6jcq8b
z(^ZyBY-b50DTZ@BFa@{SU+`svX0fA`46n5M)T6hc;CIyQP`XCU++8)B-x4Sj@%hOv
z#KU6j%G>GlUov|m?mf%Y$Nmgha9a|j0VZWmeu^FB<+F2OKkRX)HENU4_qjMfuA%+y
zHMH(U*(xp434BM}utS!Wi)X!c7pwIk_doCu6s5{qZnkl%bA!wu1Ez=nR>vePfRAlL
zgh8ypB^()GhyHb1*f?hQ+g<;5zQXo26fVlGyF7K|h^erm&RtKQAA+F2z9iBiu6RC2
zoiE<V7pUd;V#tIywDuJEbdMAYy&I%QuuAzNzE;yYt$k3fz*}k=M$GNwhu-|2-;1|5
zGvAL8ZM)+VyIQkfT%^kuH~68G+P}Erw|Gww6uj#B8MgE{wC)tty8ewSAx#N!*38&y
z{&sjSIR*0rSfKHxMo|Tcf>`$9b<!N=QS_qVSV}_Z#vSnsAe&I$@3p$XY3Ww3kBpJA
z3CM8v=Q7|~dBAZBO+q2U@VO3|X>G(bT-Q`Z_CaJ&%xV|zD*VMsZGc0CEyyyA_h_4K
ztV_wmlC8%eHs2%eXnF*{XFBlL7PYmz-4Fkh!hej3F<Do(Kniii5t|mL5(r;$R17iL
z9WresQG0bOb<ZBPrp5cNyY}I)f!r4~>WaBW?JmMJS@o0kKjh5$Et%qoo1dQFV}$<_
z9`xF4Cp9`_|DZjaWG>`H+0y9+ti_E*=P`0nOyi_my(O-DQsRZg7xJH6W6M5TKIGEx
zxfNuDKU6`}Z~i{CwgUh~W~7^8(&kRcQG;|e67jhGR_x1nlwxaIjGsRSI5os%0`t1I
zWi3aQNA~Z@9F=S5@PgLoXO#5Uyx<Q*Cm)^knNcMLjbU%ZGO$wGs8BM;&N~VJq@NU}
z{h)qe>KC;uXBGy=iT2V`AYJ#)QTw$NqJ<!C3*+zrWr&Kt&^|DZW~e0~p9Ebl5?<Cw
zx5_{1F?yjW5xNTi_sTgc=NEQ+gi(WgW9X!R#x2Cx-WPMq!T9k*kW*csmQM0+NA`PT
z_cIr2TuJDFR)Yc`k}^D+_|wp#Q#$vd1V8^n?tWpL;mu}dh?z+AzOD!y9vL(^Fg^Q{
zE7nk(0f%wC%vaf8=i!mA#HU`fR^RSB1rb>aqqc;rtus2`G`eo@D^4k|$FXPicQ;|h
zpprBk*VDIW-@Zy<wzBTlwIJAEfvxRbi2Uf-@+ko;yTpHb5kijKq*7<G(ey+T=o{Q9
z_jX?oGAM@sMoN@pM&VE(o|R!hiqa8lyn775D34l2EmsWRh{Z9-uLw7I^HZIe^mylM
zDWL-Y^p`6E)i4R=?flug`-jLOHb1$`eY2gH&I}p*&lSyzM_tm5<Yt^LoeUrtFOVHu
zM8FZ_l3o)ljD~@3y&jJERaWJGSBGLI@hxGQ5B~M`7cAQYB1a8(^Y$<@S~q;FbCT%U
z;~zj^DGvHFwSDE@SJkB%{5W7Uum;kAo9kH?U(o~y-6;7VS5OO^KhjwqZB?Ba55=zp
ztx_g5OH$aF`bD(Q2zxx};jkU#U1)yC|Ja<%E-ZXnGif-kljmWAB@A*cLTpnTrn{|O
zh-*gEtkC&<o(1QtOx`Dw)DV3nA~OAptpBmiY%?xnd1Vogh)T>NdYK870hley7Yktw
zWkk|SSwu^2=}}Z9ZDA<(5lIy#s#wuH$x2iCRK*AB6sYp+FJ#nSp1ByqO*`oL7oe&p
zq2)E@zZKIIK*vJpv)oenNd}l_4?iMEtWLs>r@>%!xyjQ-`P0|rm0>9*G;OfQuocpV
z2-bGUlRbH!6pyPGWMUD?K%KpfV{}RqPhE-28gz&d#wQK0tyYUQ^H6n>a(Fp(I8v8L
zU`fi3#ik*Gl=>N89lb31foSINXCEj=@V`HhlmZW^=oC!~&T3db-H^L3JbEGyWqR=A
z;?oe<$`za1Ks9DeH|cAaaZm|(`v{*{q@3I$bvXDPUqV>D5Fk2ub&yb|TZJXiN@_%9
zv@EQF4?tDX)GHdMW#n@n*Zd=V^=-%dQUX@Bo;ei&b1Wln^w-I7C1TTXw%FCK4y%20
z8P2N2mpW_-Yp{b*xknu9=x2QyRekk445iKtWh|50)AKg&``2QvJQup0l#_CIIJwxm
zA^)%?d9k<Bp&%6P5Lo@1CIHaWDcqkIG;t$}!3-JpEi2j=?wzUB|H~SB8g8?Bz_2ov
zR21!Vt(hS&&Cuf8pZS-lF!1PJlN<<mS#xv_3Pl_bIW^KIgw<h)h_4LIZ1#n*xG3j1
z&SWNiF$k_nA--H$HVD9G!jh*va+3T?)KuG(O&8LYso0m`USRV(oZoWo{#Or05kX$T
z+e^MFkg{2Ny&rO2yhfMW-xoIOGA5(ZR8u1cNj`RZ@2iPwO#FyX*2@g>*|3TR?Ja5{
zU<CL4cli4Yo$C^aq<rzp9n%E)Enw$lS-gLS;%oTKhrhkFrj>oz`WA}WbZ)=vgwSIq
zYxYU_C6EP$kjVCS!dwN0rSHan`v3Fq{UjsW>!GvNS|DYyenjVS;kLa(EAcA5dTEHg
zNdm*@2EYA9P>UMy<`_D`V0t{8H0>y*KCDfmy9mg8ME;!!^fB9waJ8`TT*W?VSpK86
z@z~~xBrGoBEec%}+F5o%PV%)Zs~U3|5V<Gly{<po@YxZin3H}_OvG=g%xHA~;PvT8
z9iNSj0>XmPfk&BvbhkHsyuMDwyN^=)om5vAZF@*?kOIj0)93@%WK>d)vP6U(Fo7D?
zXWaIjDWdWe*@d*)iTg3+vb)6aPGWQ6UJU&WOi62N(I@e<udl5PlUJ_Yh;_j0kUl05
z@(?kNiT=!y-I^9|@S?D&cBIxgAtlY-ufUTX=ChSHEq*V~9MSHQE*N(f4Qk$(G)v)#
zq;F5%C7Z8WSUxT)fl?f#g%+s_R5Bq)6$9{4?fL~+A3a=4*#@Zv^~h`6i9@?TCjWsc
z5qtF&0@214k>wC<oN*G8NSVU*Wn?EmLD^6BEa*=2&HO!B$=Y<bQ~Ub`w3Gex4hkhI
zcv_(aBYQ?m-=)*hqwrG|$`Z)5uam8SHdR<qi5>jkW%=^f$DhioCoU;H7`;K}Z~uWT
zLNA^Yh*6VpqC-hSi9^2v4Nof_vs`02JoKN3vnV3HBW_*xb6`eGDHoXy2-B51isqU3
zCW1|NGwnQQfbO@HY;={&qt5Z(8N6jU)q{q;ySCtzzcw73c%yE%g+;+hhsyGCES!x_
zDk|O0=loNn1ssa#&eZZTy<KD)<s$AIIw~p!%%}MXFOjml2YiaDvXI!=0cI@D-~?}9
zy9D2#k$t@PW??oVeW5MJZ^b!LZX8HtsCkn6*;EqDoAB6!M43svw=b>^gIAp=i$^JN
zqh82Foj6Yh8N~_OOt&TK2GZUBmeDoNq5Wv=00$S&?-zC-UB@)#z4!u>Ax6|8D>n;k
z$oy0P#dv|ydf~__as33v5R{#j#6rnNP|Hn~Lcaixo}5f-V@|*ck)-yhOxOz7+!Obu
zFs791@$s9nX@AQ1j~D8-Ng~pvyS08{j~hXOl!!ePza4qMhEjd`G~Y$<uJvD4w36Xk
zT=i;N-<*N(%r~jG1*9W}i)ymTZ-pUn&Rv~UR2o|36&gANV)~l36UVf$T;Ih__YP-e
zZD`Z;3ztiC=K00!uIB9l5r^*R`bvSc=$|`l;uMn3W|<S-L1@GJd*@;}iK_GKq0o&`
z<oSA>&dpw4H31r(l{#@plWo(p`yp*-&Lf@JUu35Q(k^uC#bl#DKZ(bth*}gz%dEd)
zj;dSke#5u#%&FH~-HaUb9pz^sBe%=wk2uyf6JWsu(Wp()^6H|`eSjhhX`=CuUy#h4
z0_IeS<n|UbHH&*r<gFgE9(nHp25{I!Moo}<ST_9sSOCqdXLKym*0PNArjQ=cE|^WU
z(!eR%f_MaE?8wr?961x+M!opKwH7r(g<qGom&^|E?-WaKnjAHj&OsR`(BL1=4=7y>
zD*Z{k;M3#8>-kNy2|AaU%;z1I?-!lF&XO~0v;#1xoUBQj{EuHYt}hu`006-!)Zp~{
zSLKh0%PJuN0I^BEljsf~d|A()-}jXi-^lWx6=v~5Ck9%Wukw*g-mgR9&0cQeaPwxt
zL#S%!RHd4G+uoyD6$@@y-41bkJ5aL>m-evZ6(HY8R_ARxGNWv<Wyfb}#{$a)ua2RB
z@mvKj@3^pQVNF5a?u`Owtb^Q$LyQkd%&~{R3iL<R*@Wach|*xh*nP<%d^P3evq&{C
zA%455Ft2H4QD$pr)|Ur(dJr38FXh;;-7LVPnFUgtYwy?bGruz@3Uwu-&TNYGPtDi3
z;?M#2A{DN8H#t=ot=_6k|40ijTL4dX%S$}{YL(}EE*RQ=-p9`}{Y3+kq}fgEWpJf9
z!^^Xfrr5r7l9VXtu5p(|9)SFNLOiXG^Z9J1_C%tWU=pJ_6}ZNpPs9j%@EXIDr$X|<
znxYzWH;wPzwi`!B8)AqAS4104l{ITS`!AX{)VXIG)hoAC6=F`^8NMPv3<U5bWXEf3
zWT*_iLH_@036M=c9N8v@M}}k?f$jz%Dq3&~vjDV$w0Zyz!`ggEIcG%Qp-XGvc-4U4
z<)bYVH=8I2QiXipsU=IH`%;=@e5jU5CJ%YAcTJZTXo1eXIL^;;WNio1LC&<FgbyzI
zoQc@66WTV>0WF5w+X7%I;KC=L&ka%9Qi1bsu9N9r6;ID{#*al}dpn?Ijf-aznE&l<
zK3IZz?HRdWARB%GA&1X)t9}6YOyB`V*<7I|^@~sbYY$a4MsW-U0w;pPtApc%A|-_!
zWON)WJmp;xPSY!aqCJC)#jV!{ibGlcNz2P86Xoq?F<B}(F{v>*%g+ksn^4e@92)A-
z*^_!=>+PA0Ps1AheKwR6<0(H>V{v?ep!Z`7mBmgKC>ZV-a#C+HwJO$%?B-)Yt~nwa
zsO-?EdF>=!q7nIz;OQttQcTNbS8&V6$iyOR7_T`7e4O&t1mA@wEq5gou1gNKZIQ13
zwyQyn6i7CEU?a;yl1DY1dR<?et;Dy(i9wZk+k5E03%L96dGz^Nsv{8&T=tz)l4Idj
zC}bnIfwQcA_wMBh?@Qj}(>eE&1AmIkg=ZqU78^d}T@#P#PgQLqQ?el2=IU4Wrv{>l
zWDZx=>Q5Jpa35mles&`5+?FAPFK4L`yGmE-GLG9cv!{gzn;t)`kb_@NMG!V^wL2hl
zS_HVu4NE2*;dVzw%zV2n@IZmA66kj<f*CdY7^m<w*#?zCFaFRntS-Ys;+1GKhJ@=A
z#-6K{m(ajQkzgSjuSFQ7Yx^<30QN}PGH-cYZk=fkP*ETbd#6~%K@4bSQQh+P;Pv*f
z4+y9fdzx?APE(*zFZlA!SL6cUV;Uw)gR92AQdW&xPPtmO7c2Iev6lFf^lx){W!ogv
z$oKR8o6&v<D1?^W)f(XcWBu1CozymqUZ44hU241w<p;|(%AF(mxF_JvA|tJ&D4nyT
zr?1l5%gZhi9E?8-$@M?ZoLbZDPjpV;!l2Ze77m|h?VmBq<!cq5cgTpo*FG-SCxXg2
zrXB>QelP`0<LKoY!(ky2T&nKBJ2JcQP#altbB8p78EhPixct~M#zRfZPoa&tV?(f5
zj#cjV&;`<t``zm|g1f&-{r&dnrppSETJ2%{KmR+*6wk*7h<dt_?w4VB+TcBKInro0
zy-deQYrZb&hdRD{DV)pAJ-K*q@3RqZ;`8HFhMy{NvM$TU`(flBbbh^hVqaSy`%!0<
z!eRZ**z4zW!42;VX~C6WY_#GWRyEdjRwI)NL_IB%{rY@n@@{cN*1{gQSCUS*Td&od
zjl#`8v}};X^8{a>E#+n1UD^X9rh%6o@6&ljGAA1j_|eg?yPSmg?=x&&s4EYOe9cyh
z$|`fIouc8dxN18an^htws@)ayi@H8Aa+sI~d1R?BRwO?yt6%C0*U64OYGuxy#7H>0
z{S2Yk|G=ahSrPNuGG|8VYqC}DIw}arXwQgb!la+pUqD)a=a?shh5n#CKCl+?w(uxe
zWvIJM7thgIQ3{3C>1oE}Q*TQaBx(8tGf^*ATy6$1d=tR6Ypf@;H&Y#7mA%LpLIdS@
z?CvjcRZoSzNgPo=c?z*9ORY7KBr3yr#@-+EE7l_1?!`1UrX)YLlu9aU%;vZAxa&zK
zMw@A>4xC{Lacge{C*(Bs(QjgAUq4NVdo0TFde`@vJoqV&T2$i=%<3eGUTU#_DOA52
zUZ!lWp_9>PL3z|rf>~dK=h<5ZD-}wD?pVeVrhI)b+1nJ`3s*+>G<h=rOc0m-Ygn0=
ziGz5<-*9c|C@yllBwFFeMFAv~UL$<vj94n{N()OOgX@nwRAwxS>QV~V;O))<Km_9A
z|KrC{O^%hi3)>-N25u68Rx1n$`^b9nS}&1N#f_%<jjHZmCwVlRmgtUpyj!VC<JhoL
zSTIt8<eUnzOV!$5wNbxH5kiEVW`#2dXRZx!-^W75LpUuzvZyNhfmR6ViF*_jC^%{t
zVwnCPA`0|4i}m}c18d8hFtY(>tCX7bhh;484KEb;{116UiL~8?^pfMw0)cDsf(EBX
z2x`Zp1aZ`eB12x-sFB!hgOUnQRjX1QiI*mdI|?0A3*DSsM`=~W^j{$P(43v^G4iD^
z`iRl@s}Bv7Br|3XLhgzIW<(>>x<1L*S*2QijE;%X3^!1|*!==T+5R0v+&W25#>6R(
z@IlPv&)u!jNfn)75~rI_S4mBEd;c4c!s+(`p4mZrGqI7)s3R;&vZ7fJV|4wFf)J&7
zlaEG(&9ck1+au4d4XpQ;!_wf}C+%*F>)m#hkhh-g`-aw(O|<&14!{U4fq^gA+bHfY
zzZ}buy>jQY;}n-EjgSDpn5v?pq?arpZ;;LVC3Dn=0%<LQ;F>%3_0CRKsQ_0r(A^mN
znK@Jg?SA?X@Ux!XBdc`#aftpzUT*Wjr7jDSG7n<*D9bIQ+s%O3O3%jDd#q*cpTCpG
z_}FW4pGC^DLr1COS<LabOC_HY3aiASii5)aJg-t2{NCYE%@lw(bZPbfHrf^A)aHy6
zBhf`FaFw40n)<9|r5Ckfj*>AzuN|NGR`~v5%B|^O$&_rON#&a3I)otPNc)&;lV`$(
z$g}G5#!prq(bN<)PvC|U+q#T0_Mopt?A~U*S_w0nNgR|PuS8rA#|PXf8MEI>3tWZ<
zdJt~w&o`Crr|3T?e2v|^#sA!cEMh6}3Z;lzQ==#I`@zCzcg@t#O!^}-nJVAYep;j=
zF4I!_bu7HDyNsbQS$7HcIeHtbN+Fa}yHrAMmzZxppUS%~PbLP6)Fi81rSSkF&Aiv<
zt1ls3C8K*^{jS->5i}nK5WIatT;Fuf&<T~FiDBNd%ufw2LS)jKZNDJ*^Jk-4<00UH
zZBbdX`q!y|VwS6(fFy>nN)l%QDjYp0r>jAG<q0=VM%UwrP#fhfx9Czn!POIjw<1-)
zLyGZaA~e_{lHNYpy&KvMHWQgMsAHz*z^2#iaZyI`{L6HY9Oh*_4^6}c1{WQ+&bu(n
z<Z>cs!3SBea_^;FYBc8}7|WSc11XkL^5NwK_T6-=Dq)PHMHX@yS~IQZHEgahF@(lZ
zn&gF}E!r9XyoXaKGI=}33rSL^pNa@tft>G2a*ja!#CcVRIJ4Y~V~`R8+vi<&!YhQ<
zVh1Ii%=4E61dVG(hc%L{$y0nHTZj%k)I6&n;3^rgbPtqs%s<I~W_<NKz4iK?mY7{o
zhZ};%5C0(2AMy4>1<$k(i-aj*Q@(MFQwlDZ5PM1N{SjC$9sZOXyt6agU!8V_+=ltp
zW?Kbo?spgPi>Af>B&mXiP1?YIcX9IdZ-=8UnhhmNNQ+E-2KQwH|ILRzm(*)p+d_&+
z+6fz9qrUnfE~}I6quhILYq9T|GKv_qKP23_2b95Hzv+r{owgUC-R8JJu2cR~nDph$
z@F|iMY>+G3<WbradXHLe>E-M!62T1czx?<1MVhAc<8J>9^{RaExnANEXft2ScZ6%p
z9+PcGn`WTCkJ{0n$kE$OUHHgCr7Rv26S!a1T#4(`VglDaO=!qdnLI{5a)bJ`1ot0X
zt=>F?6<(O0L@d9!7DoF%gTH^)p&8=u7!A&D4`25ut{&5ff_LkQScnVB!^A>oNM8m@
zwJ2;mr-5go<9-jHyUn;QKn$3$Jw5Icirf1g-=CliT<$uBUxb8)zF%X(bK`uewO@J9
zyZqcD%OH5k@%*6-#?S@yh_QmZn2aMynLJE`3Y%u4hjmM7`Y&bq#Xdqs0#6t1SMaO2
z1SX%LU}<E!$j5t#%VSLJV=U?8u8{a82^CK9?8V)57X4l$h;t6E<RW`ux<YzUTW;lJ
zo?<wk1`e^iqCUEAZPO^MT0wmSdVJ*4n>L2EDy{&*C@&YV``VwOFqazm4$F3$vbI0E
z66YJjB%qbq*h7o&r6eG1a;1C_^yHGtE^$@T4FH_-JplmcN24=p;-Lpp^nSN>*%s-Z
zH)n$s{l~}2s^Sg7>OH7Qs?Qe<^!LI=L9~FTV7(rqL$Wy?TMblDLmJ<u=q%T#>QSTj
za$7dO{`TU3W7seMvPvtKrqP{%2y#ph6L)%@-ZKtBGeMpLHiXPlS%3g^_GTJ8w3Lqi
z;l?`*CPrXXogxy3=Wp?uguvp{pu;3G+kkL7v$bKj9)$0^1qE|<v^yc1Asog}pglep
z%li*Wb=@eB&o~RIh~2K+XxFS#IMJPxwKLK*Fu&216MGycS0SoU9JeZIQz=#T_R>$*
zXxf7}4^8AfIrH(EDERWt2c_=VlU{w+G2#fnxqbQ^kAu%Lhq9%*%QfOjMlb)WJT*eK
zQ-O$RV@92{NMm%V_!fDoWv<a0p~ZFf4C<}gnOMC2yunL7zS`{(u;R7n-*5lL^`E6!
zG3bdL0fgQNCh3odk-xDN-OA32IDZm)t5q4FF`p3LykiAXsmewn%`RmO+PV;EBmnq+
z30k1M5;gpWeZovEsOABop%wob2lkgvMPzsmCIZ~chY{oO$YGGLnMN)Beq3=ex}zyp
ze@W^EL#ls)?RLz*=2Db4%Ib6TCgWuza~XDrT})`py<2PUS38Nz<*mi~%cr+Giyd`F
z))=c7<wbWrjE-k}M|@IM0#mIa9MPDY!O`J$mLT#GQVRdrD4GSazg|a9&+Wp3X)Cq#
z?MWZsUKtK4dGcS0eVU?hYig<;`?foBfwf%i8L9|F!w=z^5l=Xd6!Dm(TswMlpw2({
zE@fLNWUQ-%FT3job$yPPrDmL&DjeAzjXO>9t<a2j$n3999NITejg;^-Nl3Wy*=nRs
zQRUAnCL}UjrFHCH?EbRb@WWAxybmL|sLeoc&^Ns*0=C1k<T8(5(ded7f+@95{I^@2
zBO&yzQbBv&Xvp`bP~ff9XX<|R;0HSfpr}r4|B=BFG@eS+9mE$C+EBA&@WgvVwheEO
z3<9o_I6b{DBbsh3O!KYLlofiV?vjnoWVK0p^79!@_8Am%fd}XZ-V>$rzk6x;M)Yz<
zOfDOR|JtUmTfAuF2k5+e`y$ChhU^)Z*u%QBH#43KyHLX9^;u+Kc2e<5anAWjjuBkK
z<sqMp*2DCbq#%1Ce)!pK`ynFDtLHH|EWpsf-hWp%v=|KY8a-2Jz1Zx()(z%z0jzAH
z&yH~(StzlC*4;#Zh4n1L<w~eQyUlB~3BC34^nRhr3Xz(M3G&I+a-P$74p~?w?nKe-
zN$U(TPxI5~?(v8biO`!f!PcM@uh<(K9Zv0NWElhUe~-bWkTR`8R?6^X`3Y@Zg}gAU
zsI$^bCLml76nhnrlZJYUf@(_(iMQu5+8^RW*|-!jHRypSaSG379_Z3KiRPpN&f8kv
z)9EDVf0?LlA@ZqCWOZQaX_Ima(hbF!I33pvZIE|{tl}rMxs9NTybx({(M)Fmm3!e`
zqvt;wD#h?O*yR;4vPvn0y1mq!uVo)IKi#U+m{}ZJ$8e@w;t5z;;b$zd^WSy)>t4et
zStZ{-Ira(_wIg;ya_1^yt3uEL*<uwxmm`9MaGCNz`Tk<oqI_g5u`%nX7Zn9a&tN(;
zvBez(4DLctO`eW7^-ts*EcDT7?eqcq%lh*2T8bB%?A|K;%gU@+nzw0Qt!Yusqi4PU
zRGsI0wMSe1Xj0HH{hRmBUAik$dCg}%j-xHqF5RD@b*c-Oc_)uao2~Elfp#5L<>*8-
zL}X>h9xq7!Mf2Ia*i|`qsm&sV(LN%3WzRCY2e+9on6xH+jEhR+aj*KNdI_$WPMDqD
z5(r(cc+UXi>?8hH8%WnbJ4!Yj=ROyx<aWUo75g5StMizp<&0puquTawvmkdQ9qDlK
zoB1|;*$5fXqplLr{%^_g`n%sd&p&zsT-aQ2SGA?cJu8$$2P-ad8ORS*S`UdL&%Vc|
zv{+WXuL3FBi}HM$UROAL*y-gh@556(UJfiuj{wrH5lkwK+|aPe(@)O9<4Eeucn#}M
z^I)ajqlNresmm;MSd5|D^^<_GC_CA%_|GT;|0dt$AQuH(M<hgN6%jHW`%A*xGSY3S
zykuZJJn=+-o^L7+gT<&E@6v#7mfp{QzqWYo`<9dz=wtuNUfvHtlp(fI0AZgH-O{Ll
zji=aXB!ISr@yn`&zDtND2B77VR5YyjHOCAdK0oU0am>`O8@!rE@Jr1RzRYn`8Oj<!
zVYDIi$@&VKf`#__fGnKv2rfjGxSdA$Iz@u=<nzP)rH+c=VHcJ|9Ci5NkAj5&=37Y6
zUl8MmzE6?on>Z{UKUBobnZ4gsxfT&4zkualGGDp^$$N@>6yzX98Nshprs6pDSHEYG
z6v5v{^tK4}e=5v*S@hNXCYcRNGOn#}!`2iU>gVFNOFt70C3t#XPi21NOzin<D~);k
z;U!LL<$55=!Em#+*|I`)M(V(n=vW;Yq<)=$NFTgsx7v^gsLBH<rC+;NE0M~YCNr9p
z>*lys!#>5C9Gf-q%<2~=72CJAm_2IvTC}8*toOx^t59KMNIzR-TSlJfN|waV1r$AY
zeQprHXgc4d8Tlso20ruO#2SIme1VHOW|{SY{);?HR5F^C4F)M8lSuOFQ)9d}%KH>?
zB1g0%>K7qwunwd2@;IgUfa&o2o;JxJ)Wc?FD~mKj+;-N{S7B<r)>vOMo!<jVhO&A*
z!x>!poC(fzpqBZ^5X4Y3+T*V$YQ{7v)E^Bh_B+RQcm8WgmcSr#OVAJC0tJ$gWVd+#
z&u*0b)5_Z%{ny;L*c#pcL{@|}@F&p&gf)G1{A!u|4}_<h(qfo4EnSi_ooR?m=?vv`
z^*290%wcxnDQ4VT>p?WRT<c(vX_kK(dqsbBdr)>=dWs;ihxkefHhh!^Wz`=`9*ZaE
zE)NTe_x~89E>A_qc8Rq`r{N~^w?2v6qA#9M2ZFO8zg7zBmPhaT92cccCM}nYkw#w$
z8Ul!(M6$SN4`RD?Sd*j~0z_SXJ44&@T>r%W5e<q(%6UC-<0NbH??$#s3mqBM?~&*;
ztF{9VC#H}hcS9+;_E*I!yL(g1`1E8zJjbF85~`{{<@r6GNld7w$3Ayq;TYVA9^hSg
zT0zlJABq@G<3Haw;yva+47s-!ykVK3c0s=3D51i7OV?^^)?ofO;eK*)-C(al?53G9
z(BzIFYp7}RwMq*}!v(WC2(FPEB72<pz6aOPmE0|A)K&ULzC3_2hKmq1JKi2jlQ<!E
z;u$$^r!ZK=wxi&nFs0P3YiK<q2LBrP#l{z-qDPr6G)-RxPzkyh3*l$j`n-3hh1r#F
zsx#|iTq-8MX3z+Wh2vA$i<%uvfQNfVP|po7yByi$kX7`z3Ec5t160(z9z2lfuZ7g*
z$yVTj33!ZSA{;qDuh$V?6p3mC6KD;0==pKkwnUEQT4k6AxiguEJYS}zjayONv6K=R
zWXli2;!RP44}!@5qbCP|zix8vbIlqlP^=A#7Ct|1e0R@kT82isdX98qQ6%6<c=h^~
zzvyEY(8P=Tx`*4@xIr5ynry<}>2=bS8HQZYZ<^ES))s%=Vw=HZ-TCjzp^qfwZpa#!
zsI97Zcca88IRf#jJnvdhuIG20h51@Uu@Yir{!|XUU{DGgyJk)cgv*{hRnB$CrRcYG
zMuE~3A!uqaJ59vPEY$u8`jJYU-m4RWMsb|bPl5b$8vSL9-|zL$Yt(`AE?-*uk2t;m
zOG`%)tJAQy&76rFAn3Pk01`J{cn(tioMFezzXY!5N9Xp=<ZSbehIrGE$1}efZOg}G
zzO6C>+Pv9N{m=@_ivXu3+Acbj*q6|taijOxpte1??}xz_0Uv3zxcz&?-i;x3b@URF
zq+U>E5XQ!1U?v^CndrpP+KKYNnQbmH6UjtANw%16luwG}YJ0s6xQ?b@{2^^C*#*x(
z&bDf(nbO*b_+p3P8H2#ivN3(w8!bsol{n@e9zZ=Pw&kaPbW24{z~<pWtU{Up7ABY^
zm5pJza)g#XosLN2LB4z-APjRXIXg*N893ews?;TIZwY3!Vq^w=VX-NSEfsEgzOb*-
zDn!mwuLSjzhaHOP7if(uLw4Y`B!1dHH8x-V-Q|zvi(HDSb)LD<j$zjHXZ?<Q)Zutz
z6<YVj256{}fSSi1vL7@U@XZ3R#ltv<Wbb(aj`{X^!bf(|_WXpCuRk$dX?%mo>FGkH
zN2NpN{zC8M`FeqE?>Q@2`b!D@bG`bG`K&{1llN))p)WWH`m)t%W%K3w-xDm$jO9J!
zR1`=ENU<c_4C_;R2-;@hfKCYyyGgGDr2)l1*>S&Ec3yA;wm|nydw!Myc!2Xf=m&I~
zb7}76lwTjA$BMXs@9i(QCB&Cx0ASO9KjKE+1{ol;BN6?a)Hn0`DRi@sml$6uz|5H?
z%&Sn=Jp<yNNP5JMqE{t`P;fzCT#xzcAB%$h2X$|l0zWp|6}gEd=5-$lb>us|V8X&V
z6(7V)F)J@vw{Q2+Pf;3{rG|U^F-C@Az}UsjQ_)jlhm>BY;A2>!+KAp>K=`J60s<lq
zsY-vOvFx7#193kPO*mDrMqo<+lQ!dsvl<rzkF$muo|2v1&PNwjGh5k|#1a&k3}sJ4
zH(Ej#=1-oG5tnzGkMlW-&cDYWifYGru7z;p{GYzcCy6G{wndNqVLE1<9(|}+_%pM>
zczhAF2E$VrlU`*#AJ3>idu@jGa)uC|=G%|j-j{g2%n;!Jffj_qAOtA2G<CYW%X+YJ
z8T$r?T`eB5t<H=a{g>o{dU4XfLc9ws_sD{v(O<T!w0}ivdPyM8QJF8$3s)qUY80<n
zePygCu*TN+ye?2<;`WM>2d+Hx`QF5I8|n?bA0^pZht9fG+~#-*W6=UWJOVEUb`4ac
zj*0{zCwExCeS*TLR&Jz&t?`rVeYsALneqLm1TO^GUY*~nVbp0lg;G~h|Dn+CZA;|p
zl2x!XV^x3m4SV59%%DMApHDmVY}-AKx!<x+aYzd94qQvW)pWh|esa-&Pe_(oK?9?<
zPQ4~?IYHre?*M|IKpcjX_ck-%&}7`{m|#<X$j3%IWg#vUBe+QlJ$zYy&u=!Fa{rgl
zxcjXBUhxZ&_Rr@*DOy7nKxE(GM)mN}cF&cr3T-1h!tisy;<L=*x_dxgHHw5l(CUFv
z8eapYt{F8YZ%DY;d4$tyoJC^%Ty5ERwNmDA)01wDQjK-AJwHu8D-GYR0z#D5VI*Qj
z@k-D#OQGHRWOk%tOqr7^RF#K-VuA?VFQW#;J8aJlYTP|zMrUjhkm<~sOzKaD>bgv9
zxNQUhFU^yaa(tLcK|)zy(@6>=$*VHnxO1-(tvIkW-qj3h6WgUYO~QBI)mD6cCzCtz
zAhn{WDyK{s`gYHcaldW>^%dm0sNwA^PVE6X{a!acwVm<}>yw0*(b%ElPR32tm+!ts
zfG#sG!skX4&<{eys0h(M)H%HfZcVIEIK&hzb4>GUCGrlsT6!g{sjwsaTG+)$tQMsT
z)maoNg;}KXcUZ(X63S`ji;cTPRDk);V$ZDTg@E?Bin6?f8Zzk?t16?4xMuhTg?n(r
z^of@8rjV@9jFkNsT&D&Vu>%!?d<2eWXQ7BWsUr^*P+rsC0L7~)bSTx#SE9ph!+6I`
zQxm@XFw-!AbX%6)I8e!!#J5FGC&i&y!@%@G$CHZY_@~uB0Ao;q)HICTKQ>ZIxJ#=v
zfxZBTSfaInf*eiC|AFzPsjp*o=6rLBck}1+0A+~mWE;AhE8fNF<b%H*(35>H)MD#r
zqswkOB+J9Ok7KtsQ<(U*v;bXZfOf=y{Sg&B{NtGV6G6<%nu?v0XI+`j&L{U&W_mm&
z4_%OBkS;`UtCuX{^9a>Wz$O~qPlYwN0Q}?9jTafSAJcdO!s-Sn;PM$ufc!Up6aVIk
zk+67QG%DjBy@$SZd>P{1AEz1CjK)t=Iy3s9CoF0?Y~N4s%W5<AUUo>I0+s8jdX?(D
zZ(IoQ>JRIUnA`){&zC3eH2|QUZ#={&>G@+(YGD)Gsvx1-i3Leeh`cQrLbH~LN+;SK
zfjEE03*IXv%l~i`d6Rwb1yRGH0|D_s{_Bigf^8^IJ%p<LQY=V)D4Cb8%MI?v;q&Rx
zI(4#fc{?k&(0|%=KQ_A{zC-twD_$)R&f-*%Wwa7Gtx%2sV*zG4AEsZ`6YXMGKd{V~
z@fMzK!)c7zYxg%wG~8Cma(KPb7~Dq-+Q#$Q+TU=}JspkDY3{WZcPY*o71Qj?Dl>wR
zVw}32e4Q6-{8|eM3W+M)Lza#fEJHTm0Ese{nQBRn{j|6okH5y|do6_s93Z>cHb7tO
zv@CYQ=F9HV$jB~xC7Tn@ojt>p(5EY^tqwGx&iGjw0o91|h+>tWd=^YtY4(MKgBddM
zW)ssat7GAxQ2RG{s77squi)d2iGO5H+>Xy{V6flOgR95j8H5Re3l}ySrSn<b^n*Gd
zh~e}%>YWq}<>PH$9tkweS(ozmz4y{@JGnBye)?=GL&M>(2c3}6>QkLqO-|hKdb{Qu
zI&RgzPnIcxMIov^1c4JYDh-bZJu0twExQ;`kK#}noYX|iDp#qB8SXB7SN0#n&#^=t
zo>yp*0gqsaqUU*{4+p|VviXT}TK!=&2u}8*51oE`CUBg=`>4dd?5d0u&64@(LyhlH
zZcHWUErG<0m1R)M1on$_x(`kN<7VU%79&FyxVi*cnKKk;!WvULE^5TIlv*KO58()V
zh<@)HeS0IG;B`S;ER2)5Ap(pk|Ml@3lKV7BI%)a%DpxmlC3peJQE%>UqH*CH+w@HO
z9>!6%Lci>*-4|@TlpXtVZgS;(^+GZDHk{-zanQ@H3SWiTO%O`T;JL{FMjox9`8g~K
zadBU>pl2mUO(TlBYyOS{f$#S)N7JY-#}%A@uEZxO`ZlU2FiO5D@$FD?b_Pe{X-(*P
z7h_pM`}b-vB7cDL?@XO%0#^~Pd<?RR!t@YyGl=P^xhUEISpf=|$SgC=cHt;p9)9Ti
z)<~a4F|_>raiBE~aQP0k5c$w6Q^B;OCuVe5+quR3+j}>Tyu^aELpae$$bWb_dn~jq
z>;#_^?MzY|EL0~TnbyiGvhkiJ<aHR8s!S8ow*)hJNmtK4jx+$U%TSKU6tTi^_Vx!K
zo!3jlP7#anCcMymkd00>u?To;E3-wmj$$PBU3qBaL&@ip{*y}cgWw8K=42ZpEor9e
zc1avETk}EnJ|KLiw@32Mx^vqnCYUo#2(d^??)jR?asA&l!Ja#OeNx$ZI&~q#!y)0)
zAEUWS+``jeVDrX>>xV#3AD4KL=PmhXpLL_Lw(_y(+_=yV5(}NbWR&~g@djdFNYS+>
zn7Ak6Yf3~=fmC9!0m~85_Pqq<bBp*(5Dl$=94y#=+xgE2L(}}a4<3<n8N4XtBJ6@R
z_Fmb(pUm#v`pN|?reg=h>1Z$s!A~T{NOt&*F&wI=m;M>xaKFOWYF+2oXe#_Pvnc&X
z*>3SN_8bAc+2mGB-xVWxWQ18dS$5ZU&vDJ-V1@xy2+b7Z2vuV<ICzV>$8G2mM+Mp-
zzc2UZL}70m1-wSI1^9L`Ol=m@exsF-$*XTD5SMZwGpqcQLHT>A<ksrFVsX6!vOkd6
zcu=%gc>Z93r$?&_Q!ZNgkOQ<;u(kbvG+kv>6ye_ArI8X4=~}u$q+7ZL1ZhxEQhI5Y
zl+Fc28U+NTyF(gLx?8%JUSQef-FxqO&zTSNd1jvZ$FFRa0^tz2rOMG|A;yKwtj6hD
zru%rd?fDUAL;BA6xwVam@8xi!_jdAi5$8Nt%o!MxLUMksEF8#pnbEQLQ7i2AvXC$r
zSw4Kt=Lz6!<^!{Q(6$T=7qDWSqJH$(ur^A-;Y!~AQDjhQ9`as=6?^pUz#8flw;u%w
zeDl=zt+Kp;)66gV6NK*Cqj%Y3gBn#{pb4bAf$11!%(!ve!F1cGJV}phcuhM&im#%I
zB%RqUT60sWplK3Q{!Z#jP~W~xnu{Ws@5Tm{uCdi(qRuaMF#%PXp88`4!J!^dK%ssW
zz7<xo#|Q2drVG8G>iGNS2DUqnt4^UgoJ}mw&qickJ73gv_G{fsu?n6&!QjtK!z_q4
zd{8Kr0~C3m(&eW#l`NBUh}U{>aN^_5cIBUjAK2NH+5*U<m6)*_#OcKxjzgf=;$J66
zXa?)W`0H3-5fL3ql8va>T!xHs*zLZ=2_IqV6r3fR-ukfW5iP#{+mB;RjVkRGu<(S$
zAmaC3ezXA=ZW*yTL-^j{Vd*FRel0uMFSB_HKQ39=9_Ic5tgPbywF&_NybS@jclwwx
zX=!JekKu{R-mi+Fj;<tmnvv&1Dxt8EF%7J`n@9F^&Lzleb=+1U3iZSF>-+FG7-V*r
z+0X9mXz`OatqDe${F{dM0KQ~z)=lHgpJ{|4gECG^&_ul&Jvs~E7iwP&x*`RdqL%}J
zw!wl=gg@28`lF@hXEM5YpDS?5uV~8L2w9~J^)S7RE{x}p@h4>cH-5RIYV~lpfSIZN
z>RDWwHJ#o2=Yov;MA*a9s%wQz<E_5?lijE5Lxt?J`F|N{Oe!Bd7EwKy(WTh;d3c{X
zTZ`@zU=BK`W=n?~jmt#ltIjsx4xraMNpIZkOXG#7!F2k#A5fR~LX^f;LC)<5DK9N>
zAwm2-^Uo34lCpkIltU^!ubv!x2Kyg^D>4M;6N8|^?T?D*uz<J8H{toP0=~=9%T*>y
zkz>9eVIR{gs==$7Wq0hSd4{?UmMo4+S<)q|O(llZ(|q88LrXTf6n6Ue6OjC4kGco-
zLzn$8FXjlGiB2Fahh7w!sjE7<<VVZ;0WOF4(bmx`cXTZNa?6UV!eHGiR^!>oc*pjp
z2J^%1tSh%6`|t++*Q~lD!x0=YHRd>3Z}r5CJz4>^hx<G<vq^YtXI|%lx?a4htNH`}
zXDoFYmx@0zkNw&T+ZvlfWJ~C{gf2+M*w_h$M>&(*=aRx@IegZZx*HhY%hg!SrRMK)
z;C?m-pvEAj_T8sm7Gr=ncHND9$9Du7N_JNB?l+eu{ise0GOS=*OFYw$T4S{BM1E^c
zT0e;r@Xwlr-v<XD^p*ocvhVZ$ma5^srm#SK?TRkX5C8&tBbiV;JiXYZM#Xut_XG<T
zEjWO$Ib_@2U5j4#PWsH7H%<FjTiRUXrG6@!>*L4C8c%fMiZ=ckgdIQjcHja8C!gd0
z;%Gu)0yZ|?=oK)O?pIlVU`24%5hpWoMwNbvxDJNVAsL2gt!f#B4B<`^ia2-_4v{J;
zRSH`4(lQtt)!$}+Tw@Pq(%sQdc1IEm{OgJMW?>!>yB)4n6bJq|qm3~nm-4}E`Q1qi
z1!<JZbMP1s{U=tFtr!=NmgpDWK=4uMDUFxtYq1gb*;h_T6)~X55jFACM*{9&QzSor
zdvqAue`=0{S(*=|pNPgz?c|E`!y3+KD(UIU($9O2pE5|k0FmMGRqv_uzEa+77(6c=
z#TIOFqPm?}!OJsWJ1~PKJ*P!eeGdJEOUNl%ojKJf?_y8>ZSF5WmP~>YE~Us@vGQyb
zcu3D}2#m)+TWSGKDI=)mFhEp-rm_;WXpg2$gUw$^(mnU}t$ZJYjmo8~%gyRhNO0R6
z!2#?W2R@Y>0LLL2hbx1OHCo$|Vxt~@DH#ZBv!h6!9tda)EDsRa0NgqUq4KC;J^}gj
zi5NNnyT=?mnSAn%g9*<$3<(`T&VAe>A`h&fAR)mYQOzQTGj6%wzn5>d<arLCepMdl
zYO&s*VP#3W_rdb=(iseEgCmyrBA@VKRvml0YZ}IXJ|IrtbWKkW-ax<)k{W2A+#e8+
zj_!5Wqgtl#HD}oRpX0Yxe6k@eS&{!D!6<&?eLVB;ZH12ZSdz|@x^EXz?u})Nr~{Ej
zk*PgtnIpcRF142FM-A{FI)xQ&XM3*0NVG$h_{r6Qs|`izn1oT0l-Q4((Y#eQlk+cs
z%{=60nEmsn5H3}tsy?rXe%-yRx8gKW;|}Dn`?@0u&YHO%Vo<sP`+quttDo2(+EX^H
ze0g={cQUS&avi@T+T&;0HS}b;3`%htST^cBS(tChNU+Etv9`qQqqsFW#9vo}6$XE+
z5%Wbr)&KRjxall{11$Y|Q-3bEW%e)oDdPKH-_J_jL)cHOf?6&~<Y4U8_(bg`AmpgJ
z8wXHIq~4Vz<BI`syC`MK5c{Hp;RaTJ3ePV0K-K!>U;#R=9tiYL+tkP+$y!lIN$PL9
z>Y)2wcSJG2;q`C(F_|O#rH>z^*3o1Q5e~ZW5+9|g;M39?l%4WG5`N=PuCk$C#!gnU
z&nIHa7z+6MpNv^?`-b|rHu~UdGLQEGXk0*R9VrI{tWiAB0nWy)5G;TmA01iNaq2OU
z4&lc)UQ$Ms#dj{{xLAXO?@PI)f>qboNAG1O2%U{C>GL<?)O``m{Jg_Hn%=Ul5?6GN
z-IyAL*UC%w-)S^MU}6nMSWt(e?^~K|M9-Yw5{3vy2vOsbhyH?k&<y@e#01Wy?+DBZ
z;VeIIbqI<*J4h<{$d(3@u)b0t#iLS+QDbV0^~GPrP)=sUyv|P6V6EGP9M(x#Ne2fJ
ztV0ZnOkLSJYo`Z$HJi^)??|Qno&P|Z=bmtgS=r{vr5^48NVI3qx6)U;t~=_c`X6Fy
zy<{(Mgtlo~kSQYythwh-)`(fI;ZM~`Xrhys6}Q4aX*(J6DFF@dZ}1uZGH(v@>|d$k
zXU0!DP@dV@my`MS*i8pnDMvcL>A~}b{&V-hgMcDKt;N(L!YL-ADW)BLj9RQ3g>rlB
zjm1Y?Z0BH**ZSnMq<A4~(hdQKVP<bjZ`ft!Q+<+1B&qbny+`!=+_okBISA4I;j!M*
z^$>21l3+b3{K#ms%S!ApIRC}_>_lw-Ebhn0oDE!p<=er}hMG%!Y0kG?m(q46j-MAT
zSHFAp9x-m+Sdm^_Q_euH;eTsLpj}%FKne981?ES{Jjp_yl|Fanq~+Tbip7$K{G#9C
zQib#a8tb;J*Hjr$&`!>-(p#5N9l6)6f+cu4H@(9CI?wNRmcJiW=%p3xSem2*l`QDM
z<E~QD+p_40$_)Q2n*gai87Vd_hZ%+4qofC)mvZ2hm3KmsJ>N}rYvfwDoPO$$&hNOF
zEFQPaVPEm8IK}BDVJfH+DWSB`C;FBDi<!Xt727_8i}x~L&yfxifv&5GQVcDg+d9zU
z+**p^l!DVt6n-heuJlym{-oB};y4{{nME~E=HkiZn8}#t8pewssy71QbZrsF#kF*}
z@C&Xs<axUeCSWe(OL4AhA|@Kd#SRUI8BYG4x1vPcSAJOuc2XtPF?o6-me<OAml)k&
zAk7cKg#K{Df|TOP%wgAx2BV!6t4P#PnsH%kRh6SYbw7}EKM5!|7Bi}cJ-OZ|7>Y6I
z67<SRa*@Se4l$xvpk9g(xU9`Zt4;*(i_bDN-xja7li64YDOb?(e?eQ12GdjWMEf{v
z%=|)YPrEzYmVldBdmL#A)ePI5?lE2)?0~0O;&L2D6e79u!;8j7_@pw|G9Km?LXzCI
zb+uww(i5-`X5CVUs5_IU<@e*d&m}N0$s-o@NBVTKq1wu7{=o(293IAP^o@{7n6Hg$
zw+&Kf809iin^0J9s6<5-VbV`Q#oGA?&XD!;N(f6mtAbeSrD&`*SG|0)C(x!sMsBa#
z<=K{0M)3K7T(;$+^Sgs|iolCMti%cX57a|r>;=E?6vHExe*cLhw>E+uWSF{#4X99F
z_I&_w9p&c`aoEP_x+wvD0Mr=YP%c6L0xtb^;3t2D&6|(!_tfc}Pb6}sf_>Psn0;sb
zI}i^@;_<s236z~eup6X7&$AO@>@sWWu1=%&-?HTJXbZp^%o4hX0@hFEYUVBQLkt~E
ztgS-&I}t%iA+3RVr*g289@Q*a3^*HcfaVikx9ihO8`@hG?s0470?+s!)-yH!1y@~5
z(^auvrs=XtZk;3}_ae&<dErWM!0qN|ijEq!S5}XZQT12EA?C}SMO4X%)xg*EL3)r>
zgbi%dgDkp}jo|ais5(<Dx3qP@69GCw3iUU#j~tJKdi-{U%)qk5ELo%vxJ2u;$BpgJ
zS{_JX*5Xq&UYrE=bNY``M!C$3=l+biFMTYP616B+TPNPinH<}eC(f_5oEc#qz*?>(
zWytOsNa$AyE~Hpb&!!-oV{wwtTN#9CG9hPN?N*od-Sp2rB-ihxjGw$RrQIy6lI5?s
zRAVE!@=`ffStM6i$s3E28NWxOu7=`>R5mmAPF@Dh&!8m0owAy<x+x{$Hs8UYujBXK
zjVcV|uAl9n6r~f^LNxE)SMl`MKgIT!SyCu5MjVpzjWL-4PDE{hEy9M4-^~eVaV=7!
zmx?d_D|dCe3j4y+#C(cSE*TX8^c)`?RGZnL)zNr*`PD{xW9n-CujXVbEvf$ME-Q+5
zT8JCg%@~Y_#Kf0O_c=<=P55&VV*(;oR0jfomc#UFiSlEGXf&3B4e2LjEQ%y0gKHrE
z4GqINydRE)bh?v<#rFae7ff-V{yXxiWE5q?xhAW0ysB;Nbf#=&Rcw+utWA@21tIdT
zG$o#oj7^-S&FyIwHhsTplUgr6aqDJTFD!Fq%mXPEtYV_h8apW;O|8$yxHcYa4;v`}
zS*uvpm6&qt@>a`8XMgLNjg$kbbir;*w!=NpoH6ygve?<@XT_g4KOd$>k<rn6bZZ8;
zUd!wkw}Gqk|0NR8QjE6?3}s$jEbg7}6;~+XX5I~F!&~^3BxS9+Ejv+&`903pm8@$4
z%`rP{r$H4QK4zi)ta#wgV>85m3IUORx6+Vgj#~x4?D^O~b6wJUDR(xb#?xvnB&pcL
z-8|^iqioaA-|>60ZU*$;w*PAUBZqnL2Y&feL}~J68DQii>0UZfFxbt2Pw=*Kt0%mE
z5QC@v9Z_iWI3QmFCx?2iolC%l<F+0<eD-6L1;gg}OSncbk9_t|<_nleMqp2`<0`G?
zy6A1WVR%N<tBNM=ZmAnB+-A_(;HT3j6zpisG_`ZJTp;6L2pjA&uvj?hfgNCN2=?Sc
z=)E9xGQSiv+OABbUQ}V2Dh*cr{iT>cQ<C@4Ydvrm%}n@R3s;9+$LM(9(>6ya2hFE3
zF&}ceR1tu(=)MGkKa2~wj$Dg03jg50y-Lw5Sw*mcu6ML2cRcKro=Yx^73wWDQM1UW
z{JQwTRLG*?OI`JXJ*1)hJ80^qf5vG?Lr{E^>S}QLR^&Kq*rJhztNI6uledqA0Jq~$
z-P|YklUWK&5~hW&9EWptc4fFdU}bLumRMhKj<bD4^*QsrWLOSxX?fkqps+4BBI*js
zDY1H}MU7yx8+SY@xvnK?qi9sKh9Sg{v(tv3#(-b$`dy*@auMqYw4Gel;R}a(ZCi2E
zZ&9iqueRaEd<r2-(G{+yLr$}Cc?&b=eb_dp2(enc345uTpOLFO>iVMTee$GIHZLaR
zhN2UBbY5@Yo&}gsBld<Imt?P<hdIA)c_9CSsgtTv=7Y;IC0$HAD`*ekM|Us{Ghq`=
z(r<7h12zAs_$`8wR5$9ODnH1e1!ugIKPR^RV{y*<n@4;lX*T+&@sQTCEqCMmIJ!iO
zpsM-mO8ejuWZEC+70Y?gbn!#!S(g;WA_(_PQr`n<ZQ|FQl-aW-?UZO%?gM-K<7h#s
z-`OW&kH;1d30|jFh!0)qZ|=80b*#ZnZm0EnyJYY-*w(5_ry_>)Cjh5+J%YA`!`;zk
zp>l|)dM)k@P7D;_7d<N+Q28eLt`#%ThO4KYgy>&NA&bK?^vj?1_V!KH_eA8rADs$o
zv>SUihW|9AFBsp*40Sf=v)IKMF~kHBZ}d9SuCLpF&>(qDLZ~yN`L!%kT$1j6-R@wH
zkfoTJQhq%~>gX|S_=plFy4<bKJMQwc@V;H+LfWaSVZfZyJciBetBQhlj>m$FoFwzQ
z=5`VzthIs_5dKmPeffNS&(bfT|C(C6W~+X#b3gQoC%FzTQ*>-X&Uz5frYe=%heXS{
z=btd818(ug(%;+~Vc!hzeGVAw4m6yAR3LMLx~rY|lbLK?CKN-QvIlNZQAZzgvPb0+
zAo`jn7^<)Db~T{H{j1ciY#9ZIuSZ#EwYf2RL*P2T9uv}y79&5%sg8%Y4qYtITfOQD
zq@0tyUDF_dI1b|TzihkJb>a*B=Z`68>UhH^nsHMrDeilb+$uobHl23#1<6Y4^F?Ea
zIYvkHTh+<1XtUnpRTBAOX8ar=)sa$blFY}*LYH!o{&upjZkjPW`F{8Z<J|3$jeikV
zfh_1QTKoXtxrsMivXrN8vhp9+mQeu7{*mujNmIE|n}%PL7W0MHCI}n<{?%`*#mD;-
z*_>@eVn!H9>={K+M?Y6d)pOy(FDr#R`<F&DeatdEO>1d;bky@T-)Y|Fk#+qeJ#clR
zJoNj$q|Yi@=6-yDpYl;IYR5xme$MI{SSf+OWx6P+DOc^^+ITMa5D6QT(vr98@7xPn
z7r`>8UEkV~G1}O8nSh-Hz61%<VWl_J&zt0S*guWqN{w2rJ>8~0zz~+@+KQ}Yx|JP?
zc^3q3R3y1Vd<QERf6*HToixY7k!L|7)WYw~4=YX7KW|q%NDap<=qXvvxdlkgxR<N;
zRqdAl<QOTaLKektdJb*F|Bk+3np<ZUBG9oU?Z$qzHu{Zxm?L@GYmMC1bMOJhO${{S
z0?-c@ZZ+-R*v2yVd@Dv;6|A<@XZ=EHEDbOd>Gg2v>sXfl_b`%ajFXO(<a3NvV%FiJ
z;K9#Xg@SV&&EI8A&^H$Qj*4R>1HX|IoN@7CR2ju<r^F*kpm(KnUTf!1-z!a0nr!5)
zidHgJ;4r%nIoO(PXIbSZ%>g^b$38jM9X$LI)JXS8R5@kQT7q-9hh@Gvv+62L<PfO*
zSN){p4%m~fg8PU{H|(&3V39EDtAD{iuN*)NtUBqD(*9o0W<5rT&xXfz-SV302TCJ>
zVS)lcKX7pjZ1v{(jqJg{b8gwd43$ydscTP`JuAC_j)V{U;8sKI6sH-ZBLRB5xpS*`
zJY|OkSX1o`Y$57<u%1-g+gihl+3>fTOU}BW*9%qXRUP^YB4Y!l0R)ZTzUhff3f%r=
zf;z>S@<m}jc#uJzWQtr3*;jl$LNAm3Hnlq>bbu>}sSnc>7JEqC7AWDm<S(Nw1nrp&
zjz7b@Fj)r7Bh;=~>aHQ%*ZL~Xy0GD^@u<_xURYkfau7+7qp;2y^7E&Mf1268=SBqU
zEetkg_3$4^8DYh*rGA<QV^1({*Jk+wcQzwMD8^8z-x}oWhIZxw#Zg1$zQ+q{T@2OI
zCGYpzGEUQitd#DDDLi{T2(Vav@>XaymBc&0@|vs1FY6#Ij<&iF58n*W*~!v&=Xp2J
zd?-KVvt98XWGHRnJx%n$b1CHz!YOCc%suJUl)#&_p{OMGdg@W4!a7cpv96a$bX?YT
ztH24UkB-R0i|LqKKD~xED{i?PkkT!poEf`y?^!l<()Fn9RLz^%<rq!r>81kX2M6u#
zt`^iRq^e7JUvmv5;i%^}eu@fO_gJq33HtME6-q)eK?9R}l0y_WFJ9}*&oy^esioya
z*L`U81nK5@>WFk1;u$R0b*~;-sKnXGg7{9XX%l$ekNzRn^(HyCew1r0$->&8#0&fo
zl6Fa*g}mG*IxfHW0b<!#n%~-VgDmd@YEEkBNcfK+1~U7Ty2?Y-?_8Yjh&qEEzZywM
zRyEQ)nOYf=E_C}VQ~<_CfQ;SDlLU_$0Xto;i~*$V#Ak<DOlXB33t7lKCEMY~{ihVV
zhG(r-R>?OIq)A!utD*CF2rP98CH5@lOM@Ig-;*dC8^tOgT5cSpBHiuTcYh*lA~=Cr
zX3%a?UNYRoLYKITPZGiZ!fv%U0-FnSBIlZz#x$5Nq12cmtjaToq!S>8#`zLt7m&3l
zWd~~neAp$S<=35F#10@#3!pYmGz5ez78-Wc%>@fNd)v^q;|&_)hv(2Z6%UL$cuA+<
z>dc<4OG_ugEKP@aXb0>AIw}825g<EJAEy7A;=?2*p_f50H!{LAN>`VECbLJuJDWJN
zp5x*_ivfB$8#~LmW)F-s)th}oD#=XB8!t#``z6W=-hz4uZ!y#ciEEbSl3ZNtr%gld
z>W|`upwmUri8qw_;yY*dt-w5+A53n`nZ%0C|JJ+>+3*iM3IA82t_Vu7@S#j9MTxQg
zk^@iAJwfbtOgmUqdn>+O$dgHE>59#9dj9E_-zu_GZ76?>(-#`Lpp(e}9X)uEQ95!5
z1j+%GYq5@>ov;jP`CV=eFu=cfui%AnZZg~^eVDpGN`IOkxJ(GyTOs`M+f~NR8QnAf
zZRf!5pg6dHcZU44NuoyeuiA1+shx8D-=<s}?xuhS?x@v5p!Nvl`aUPAH;iD)tfR8@
zD_PJ&31)v|Nl9Y&_qG?BKPLrZE9@CGy#bovS*=w`?LSM}U*l1g@h|uTF1=r@;3uV*
zhU*oTN`a2ltM@>Uu_YDnBx2U)8yODGEqK8vX0hAs&zMu_*Ur5hm3)bU_*Jf1q}t10
z=pFX#sZi9yO69Qsim9%4ie?YD>@Mzk;%Y_=&j|)kc}Nh3c5&PYNKnKds<!)&&cSh>
zA=2@D)xayF_XTs)`u()rX=~1{b4*%k5ofI}n69Q$Zu_1zdpXdUERcR(bat9~4P=x#
z8A0Q}{sV`(%#ZxB&9Yf<rvb@Zatt0$^RVJCBZiKT+DECT#d(;qy-1aWh=Qe8r{KPU
z`s}XHb;sWkXdHjv{-I;kJvZ*XI(0FU_@n6KJH^uy+mPriC{`j`MVu&hC9(hO{H*1u
zq^Iu3C>))j;I4gt(8-v#mwCTr2<TxUiOEmG)nWLoPKerZqAa4Xb>j_-WP%qh$`e{6
zzdQ^4#(whkt&VckmSMi9gZ9J+T_nqjDd}85(Xh<H@T6BjG=QqH_ZwqGjj1vQ)9Vp=
z!%DFKD!+Az_*@|UYR$gT_A5<%AV$AnbejhQX4v*sdlmPlCRxm`DoK{oQcfbLP-WJ$
zyDX-fJKm|w-R@XHZluBga{&Ni=YMFh_Em%u$B7FWaDp-sltwhSJ(<*L^$k8Y(xR=T
ztd9Cv{Nn?Gn`N-uH}sF!Z|HTcR{M3o_z)la;w29}PW$M-tDphtO!>uge~oS0%FLGX
zbz#vxG_!}Faa4b@PWbpNpeSmv{I)0MomZjwo*ZlKaL#&ndG9w4g&|3n@S6JiuR@s#
zg&bTpWEonPoM+=G{^?S8BokHY7Pm1!2C{7*Eh0Jb^kV>E!tPp7MF?$&EqM&V|1^Bv
z4Ypcg{j5wB@dlFPvrA+tQ5$|^3&0vl2xG117M_fqAC%W@y`|es9axQkVpcQA<&O)g
zjl4&;K0(`tbd9ZU0k*|~*hbi>YoB$IP>4g}bl-Hwd~6}!{oZG?u&eua-T1A1^j`!{
zNApWTAAOu=WDLIZ^Td8G7((ntx)m*1{dTJv=$>P|6Q+=9?tGoxUC?5kzYb3q?=alA
zG+xUUmDhK819lag+c79UO_#h=N9IaSRqfffEJ8i+|J)STl7372Vu24F>TA27<4<tp
z1q^mJ=i$t*nT)gO!e8Z<HCk>Yju7u3CyG&2Cj7iQZmPR!O$B{#o_yMvHY1!tcYw%~
zeH=>Dma7nc;&ZChw(^o3C>%?Za+SzIUF&S!nbf9o`XCM|Rg{+H8vg5l?3u^#y53Xj
zthyBcZv^NU>Fh6Zp-Y)FumMz29oOb%^6c6SB9Su)Vh~hQpF49dY6wjR0oBvKPWXe)
z?^9~Ad{<BHPn2}GaY~To_Yj{qC98(xAL4K93!i&*97!yL-PND&M{2xepSh3F<5-ka
z+78SoSu397UINN>a<{pHkfcEj8dMdHak7+!Mu|E<`K^YQ2dm<`qF1;w{WkV%48|fM
zjMW08o^FARMsk#C&n8TxR)C5;7!=9Z^3OYzC?1xs_s}->kAH0dAuPe1dQW{$Ww`Fx
zW${??6aPk+bbPw5p|Y@0#VUvZ$N=E^&qsc>CysvC-$lGpdg$_efAWgB&r|W@-<p0m
zYXVUUub0mUX^y|5Ab>(<-oiEzqxh|BbsUw19;>eU=v7Wb{m@sew2iX8p+Qqc`Gzm#
zRGH0@8aVQt5h6BE^_#0t=wAR{rJ8<NAOc$V0#qojzX6{{_J#@_`1EpmRaLEU&JS;-
z7^M2%@3LnFJs+`qF6!?85$}>*<R}=RvIjV7?g0!MKeAJ_lcWSs?A7Vv0nL|BZw+_!
zWdv_e>Zj-OV|OI&gUMqqJB>7fvYy^#R-G-0pJmx2(LK+gldl%K*QVu49tUlw6~42<
z^=%8bTio`vV+Hcd&m*)UT_4UBM4i~#xKq@*G?u#YQ##tm+6Y75^yc21C`cwI4?3Q+
zUCKr|G6AXn)8|?F#{#_)6yBCp@T>c5<Jfn-d{W>9&o{;W+BHNMHebj$&$FI^wSX@Z
zbw$8RyoM|@4AZHA{0YaPf9%Q$tznYufK2qcRmLaf8>U(dh!j$-Lrz4qayc251&0jJ
z4nuubsb02|c0F7?d7c&E3UCbl<(i-UYsI*jB;%Lr+JRr=vyJomRPjx-C&wuaG}ZrV
z$+<2B-qapVW^5nx;4kd?zkc(r)76zV50yi%aE!L}Z<0#mO*48Hn*X&*O73ZEYH|MI
z^BJ2|;&_d(t9r%G_J?#a+e1gP*;a`@<G&O0@Su#`;7Vx@GIN7mKGLgqOTWx;LH?MF
zElY|s>ZzaqLL|RS>Y9>1b)VM#RRt0{-5=t6_4bPl2S}IIA%#B0Ic+uOW8kWjl*&lV
z*?lg(Cc>pW)kqK5$9EgxhC=^yj(p^4fAHnddxAS{VVAO}Gb8Ct0r{#RlT#(?$^<E0
zkpGQT6dNKOmOZ~5$|JM^6BP~9cqabKY_7q(g>f_l77ihH7B{=DX!$Pu>yt{!w8sdW
zv0j>RB^o~X<R4-C{j^hk?*5a)Qe{>AkzU<-<vK>0kFo^Qw)L40sBrR0N{DO}ghYz|
zx~CL53a{C9r@BRY?8(_7ormv4uAKyhv_?{9>Klp~MDOw%ZFK%b^9O6n)f^p#eXZ6V
z_T-oIl>=A@YDy!KC<`>pp{p$j$VPR{8PJ`UHi0?O%IA!L$#5~iErYIK0<3M{uu}Df
zn(}bcygalC@-+PI@F{kvv4-{X`Cf6IirvfmEXTg7uC-HsGC=A*UHc)hp%5BApefcs
zr%orp#*;-tiit-?KdyorH3eeL{G+i+u_pv5glJG<s)4&3FywOpp|wyd&M2mYSDbCW
z5M@yn47L=#uzJ;@^Tb&kNLL4u9N($;S+O1ESsCuGkc4@y9_8qg>>zI$77kv#a$e%z
zSdVq{Y#+HmwY7WxKsN)`!_6Hv9Rs+sf%iM#nP!h)jnNY~?~cgNbF!$2Z*h<6=PhWw
zAVm-v$C?Sj=-1$3{@#oMsGVzRTvjC9n0P&MQxCq{_QZH`<edCZYaxhv>f?t>!9dn=
zn3)|1RZwkwP2VYKJ_^d@#0IJ+tun(;y|f($$F4+8Y#IM&4>-kyoH8IRh*Ihu0Ik|z
zcd#IF`}0oM=l_<pLHNd7cJ7UU5)mYDI&AodWdV3Ae7USi(DWOFhuXUMJOL8#)_LWX
zP6(YWYJ)U~##4cAdea5>@g|eDC9qs?SY()aQ|pn}rRjWVTmaU7JCCu^JU(s>X{am7
z4@F&%o`2r4c(5knP;#)qfk`ZJL@WIA+!1amY4#5A-CPpw^8ZAOz8!$Z2C-tr_~f;B
zmr^&~N-3Tx@6gXY`R=7J19mTYO%!wXeoywueRwe8Ej|F9#-Vt8!Wv_LB$9EhOZvO*
zR4b;2in43_B#^2)`}*b?em6vO;m^w5r%BH&ms~3CWM2EF`Bb*-eNUh32maYIgGQBr
zPxt1GMwF&XY7NzSbW@$Q%6SZwA<kne*%jau!P8?eEFq7#)Au57i22i*>+l~g*=^I<
z%3q^+n_u|^8rgE~2xeu?BFm>Tyzg0-f<wiaXHOzH_NEo9uJPFD8-8x||J~d1Jq!Ic
zC3>>IljtxzLVlnhr5*8mZ|R#zcsgMJwm3rh$mYSL?qa$8g$si;e;!A5^%D^5W;268
zS^Lb5!BcrGPlyitdijwDBl1OB8jl<53|*kgfAjfXKz-AsgjsG5;N^QS-oDMe*$Rqc
zV`Q@o)P38q$|L1bWP4O<$~NJOnX*H4y(d_*6;79y-@(Os5#&Rl&+c)z9Waphm>cT*
zfVeCf!vvsUj1N?R*V*D+Cvv=d!RSs6urWU#gFC?(3=nh=aYd=!Bkp@gU$DPVv9FL5
zSP9eD6vw7h-HCvbUaO{XI$34xPAZZxVZ5HCVu=Rz<9yAc9RFOhi^a@e{rmexkNQY+
z9T)Oy@WZ**!eR=XR4n_wc0mpvmcD^@q3GO8Q)?_gS3alC-cUloU}CT-x%QkQKn08%
zJiI77;wXt`AUR^gj9^E~UI&y$slYNk-j6a1`aa~Uo)5{jCVo~~Q+5-MJ^!5X{JiH6
z-kJ}*p^=QNDqaIbxW{AD-Tj{dt8Cy?!_7O?JVv-KBY~ipDJddA?n$m}wG+oUw(is0
zxKZuEqmP!clLC#KSXRL{<oPnfp^j9>+kYG#qQ(zIimO_o3#{tLo9i>Kr{(ia=kUZt
z>mLl?f09REV-`R|x<-8Es9otryQ~1V-z`B~>`ducw}_q|v$wY$bG{J_9}mK$5HswZ
z%;{JE!)K8q97i1TB}IAZcrB_{FE$u$bfG;L9%L9+9vk(mL@=L}pyTvCO~SzSqp~Q4
z{8<=_CF8+m9lRZ;An29P0;3iR%KN5)|2R-M<OcT*R!4j8wA6(MUtbvp!Wt<$x*AgX
zjWaz=SMdY{PB$*)tolh@oKMjQvk#Cj$?)EL_}P%`u)639{H*tazst|lY25#zVdAuy
z0TZ6W`^+7~Ma18C^YDws1MuAAHX&Qe*g*xsa@4ymvj<u9*u$+TIfKZomQX>fD#tRU
z#MV)nT$FEY)+2wzX26N2Q4DT}ETip*cN9$D3p+(jMr^|kq8h6tUxR``?1yVHD`*sd
zJ4imSvxAXk85%d-8<j1WEC03jkvQ>n1{9yRc`2jRUg0WmUo}XhQe=iER4rG^85yLP
zvWB02wXnPW+cTfzDo1OX(>!Z^Y5EX*uKSbBnuA{`0@Z3jU#O}qAW2Osn?hbGy-%HT
z7v14!+g%pRNRjBo^9?*ie<V|cht~hyh#@74=9p>XeqFw`4*zHWi!N&D*_|gHw6LbE
zmR($jS|*v1n-qB!ZC<h}p1-Uhu4ZGXRho&5^q-mK4nh8fSR+2k+2hD|C$!;+>5*OS
z3#FuWco~WVC|#m$H*PI9L1)+HXm4_t;EiVh>%*fAbjOHH-Zqmo!66Hut#SkCFg}k!
z-Dhc`6jZ%#45Ry5Bh;izj|8Ywk|QWu1xg8-fv?LngNGcO<%&1Y^b&R@zD0Yhu}N*E
zT+J^;%oyn9qojE5A&Nd7vnn&L;n!eTq8#Z1_6>F>mb|?*eg-MMvvtMP7wFrMbo{vT
z63Gh~gGDeN(v!+-0zgsp%re$PA3Ucgzq=(mpZA&SlR(rBab9<4_@l1_wrV7|L&LBU
z2`%FGa{Pxs)kbPbq>!150wXmZ=IE(dva(!3)hYn*9e9_StfOY8r`cs?adW(1yJyRe
z{WTx<^KS;duUr7f*Io?9E?(ypuIF}}>>JZVs+G-#*)O1ka;<Ocv(a#Bzkv;%?U-~|
zdRTY7xSpCe_lS1g&MVQi#a&Z=#*FSRU7`OBSm9Um$z-e{>QBpu<PXmZb69&i$BZ${
zSmnoMs^s59tOU`wb1>JCVdgq|(qZt^tw}ujeT@WcC+9|8p+n0BL@@%BLchV=hhz~w
zou0RUJ)>Sr$p>D2UM1#tZFKW1zkdGwk(7>Rj5k^02)AT0)7j(S=VFV%OF-OXUuxqk
zIUvCaLtaY@^JxQt*E$NLZDoIx$JOOs^Z4fYOgy=C=gP5F4`jlBeH}h#^#`Am(}!`}
zY9@yayl@q@=YtD2y9s9|V<kI(8l>mhZ00Cdj^}O;K3v0DxF==mEoKc-dkg$F8@jgi
z#Ul#rmSa?oF{f$t5}X-dHCP%*Q$b{i2x?W6zNttt&4ZyWN@C8IWPFBpeL5(_eWlb~
zUIfL&j&zd#Ey)-SMOfPF>fmF7FP~jIm)jR^%TT6>k;+H_oikiBG`=*XYc2R`qxOx}
z@!6O?WEu_9kCa-rv1+83_DIMT44?(@$*d3~`Uan7vh<0BcRj>gL07H{SUzZbiLI$G
zO!jRi#V#G+Ncs_+2gZ6l7#>Eqxf9%BTcIfI;bLQhqwJ`pMDZPqw^+Is?+P{h`N@DD
ziY=*o;+rkfQXJ8lbhsWL?N@uKNO^wieJvHDDN6J4wtOnoEhjskwEisDr==AG;Cq@t
zyCkoWDK=Nsy$K$H{~bhYN`esx5==l=Fe?OO?Xh5ggCzLYg^BpAL@pU$lpo3c9<Wu?
zU#tC{;Z95Ocll&hH=8qx=}V=8kJG9PF`(d0B?!aqyb&!(smxW>f7F@wtNP1#j$6~j
zy7QLX5&Vo`N=Fi!uiz`{I_haeKB*T)KEsBIhY`dWX8$fecPL&+a-W={dRo@ZMaQ4i
zn(*h()VB(sZx!({(#UEVo_13JHE8IV^pAiFkU@V|Td(dq;HjKwXlgx+@4qEUh`(pB
z9xAs8eR^Q)K6qR228BWT+m6XC48$?=BUQL(oBEjRZI-;n*8aYAiB?h~kwyZ8Z!OJ+
zUP*PUo1hmLy_FUn2Cbq&GPleYB`&a=8}kR?hFfvOy149n@MUX52`kg2L<)hqzvj)#
zE}Zq7%ee_j=xKsX5<$;Jws-D}vFkUUtik`k_RE)v;gnV>lA^bOO#)M-*#OX%o{6Lo
zS2#=)qe4}F&W==WM9-ybVeSacTbb_e0C_{KNxSGUtjKB@5$hTl9r)5w70CP5qiaX}
zq&1m@;QOxB+l_N|Eu*zI-C%Zp6ng(ntsXnin_s@s2-kW3IHZ$+DOj405qPs=Yo*@P
zM=m1m;fv+Rc|12djr;NF_WDlN2rG;Dt>*3ah@NKbc>11<F6j%Bj=lAa9V;e$aY4AX
zRcL6yx9xyZkW`w3YDxjd2;Fzki|w&QUON}jk>4dUNGRy0_9qk0sYXm%X?XD}`|<~6
z$i1qRhs>V#PvW$;^_gN(pTmp3QpwB$rO;@Dkv%+zs4`gYwOPLgiKZoC2&uBZ_KEJ%
zFCFIW9e!UZIlJsUkkuRda=q=kRba<`NNn)-Y)NPRk|Akkl7m6~HHC*GOqT}(nTu3a
z2d1#CKRz$~w^#~sgP;ZFy!Lg-5#@IX{op6Dbfu%Guf=|n(Uy1IG|y$_=RKpA8Ncj-
zAAV=*E&QEVB7%#ga<+G)INbk4h|LGPHtPk+6}1H%W^sLEDha(or5@&Q^v4I{&IuP<
zmh8QP<X7fB=;#L7h=9~RY*yfX_hwhll+eO&beN6P5U@gC_Ev-uj##=>Z)pK#)hZ4^
z+DA`<EdEsqa<#?ZB(Ire9gcoQsYi`Ylt4iKeR9nO<1J6j6Y;rREWpFRQ}N$|g2hx8
z0E50>K0(J>{kKx6EYQ`-XXSptZ<?_wC3<(shSo6t5Pe0Nj6<d7(puncr%bguCO@;f
zW?rErZ%&wpHe2*EC`LVgxMu0lhe^kF@&(nHyao$F!K?^vOqfIb&bHFexF2H~mmk#G
z@On{i=lt!DTMAk04p%Zjz4AosTb!XngOL-JKXZeha>zgJr3zPtW&T2pk6C{z2`3;h
zTX4RvSv=hQZM23PtoJht2-)jWbYKFpWRnyI3lIGCEGbd2W&JNSoJ)(Vgeq}J08zM+
zC#PdcOPlmn=RS3LB4OoCBeJA&^2nM}X`Tjf%`RgBo_!J+o|tsR?*@B3OMiWQul+M^
zU6m}^J0v$U{vLn$M)StVEMhDm2qKnwzUdjF&#a;{<=Anzb~_63zln3j#YK}?9g^4H
z#}Z9wlX(HtfL`$mZ0ZwX3PcCj41^kso-CWER4*2Xo*KN^jZ0>VP_XTn5eU}392J@z
z_{G3HZDUIo5mUP-Dkm-UsbI{{i?n91b^Us`r*y=^d-AEf-}h`B*1E4|b_(jCZxMyg
zt)U`!*!O9p_gD70&wyWP$~I)|KFZ6!C*xRf2A@aCwRTl!ZTP_Q`yPxb7DY(qmn^-j
zkfA5Zw3g-z5(}GhhC`}4y7u(y_knHrw|^GZ1Fx^ae~8l07StD6Tab}n*NRXl-1dVm
z6XrM%a<>N4-V)s;e0euULDd;gQOt_LQ6n!zF1y~(6$K#n$>p-acVgsB8WTp{kWCW3
z^A=^FvKMdt<|2J0(5Dodax<e^GRq;pSi$gW`<6wY70M3P=;ou$G}vqVM!jD<z7K1U
z|NTLRs+TKCirZ@H6+T`H!{TgUwi0UQWR}8D7s~|^H2$f>@|u;-#kn`IE=iZ(w^Ht+
zcr1$!_lY-N+Fgz;&&=G`K^on-k(y71fg*S90d(d=1H(<dM#PmYkKRP&#2<%w6*ez6
z3Zl`Q49~&H0z0K89)qb~T$A;k>mCQ(BR0ySAJ<)o2Bzq2r?AL5kw`$w)d+Pa=dfWx
ziCV1W2*n0#=g&N>1NRgtp9Th}UrQ;9CzASu9&FFE7>ZH0?hA_}!$#YxA0s|(1lS@Z
zP>D{h2HfWhAXQec^)1XH?96;dIc$LMO_o+$wK*w&kDGcW@s}AJt4j|3Bn>)C#uN{g
zH8vL~yI^**&vGrgOUQ)%06emAN=vdw;Q81~8o<vb##y0rxsX+&2H;m0f=__i3)c65
zcOURb3Aqgwz=a0AB2)}1dvUYjn=4*}o^&AtVgoeHZ}1gWB?xh1o|27F&_F%7V>EEY
zT_bStrl|mWY5e*j@pj}iJSIgXk4lun--c)voK>2s293K}YIj*wBYtUeNiZ#2z9OSS
zb|s|0cAun#W0e82yl>+zZ?~Oz5d0`c8HvxWwOOY;P}wQuAKP!rwfX+;VHu~^=a2iV
zz1Twca+F{VjL6^2402*0DhK0btzs%Yy&WynWE%6BfRCYzl7t}{bIZ3gWkPRW6<#Kc
z7|Z<9csk{gY-Wq}n?A|Eit#P{>w0;WAFY^}trd)M;FD*XE!bVfb0xgtS3v(i^7w(P
zmPRs@XxWf_mct8wrWlB{sZNOdI7`3`{+tcmOs^u(>@;ME@2rscM-D*yAf7J@gn)eH
znYqqBX1?Y$?%^9Y!Q2jf2#^Z2x?LXX+ekMf>Zsd|{e`@mOf~RT)0Rhkam_xM7dRLw
z#*9Ul9NrGErXC<OmFYsl-#onDTI^rWptUG}evqw@MDl0uJiw0QIP!bV3Mlax<gUY!
zz?HY_m&~=Rz$V@^V`ahgvGExGC`oIlz0+0u#VgBS2pOC>bbJl*@~N~m=d^y&-&}r~
z>zz%yWq%9&-c-u@Tn<)5kP7KS)BQ7!sWu-=uUsW@HrRjx#;e@G((TNu+0|A?&SjZR
zZ8gIrZHq=Ip$|tB<AvP?S}=0M>FyKg`qQpuQ7EN_ZuToayN6F=!^OVuxN`q*SnIs!
zC6W7HCG3MDsaPS4sZ!O52d;mWk{zpCrNjci2O*~kGw4UB)R1b?$(77gvb*|VYQE}X
zo+l}ndLKrT{BN$(IR9R>*@?)HVT^N7WC=CNw-#Uj?0;ueiM3C;@jI63vAct{Z3mIw
z83=n<`-IBK>C22_NqdMp9qHtBb<eDbLg}>Wc)Z4%et?JlBd6GB2JXg3h_|8Fc2X`S
zJ+i8+C25tz8>Om)Pxd&*6BgY}`(jKMv9Av3$$yZ4(&ndpn9Gm?3B2I)%i}QG0|Fi&
zF!fInT>@kgk9TkTb$blDahXJKr%UpIfR&ad1xooEcRUNgti77_b*7&Z;E)`oN0m+1
z^mtpu^w5_&)zxlSY(;7GAr#R=okF-`q#IJG_(G0%4Z4?T*1J<mJo+Hd6QWoTI^RFh
z@N;+5OpH|<WdQ9tJJdl~&|W-ZHrjKhl2wb3n>tJ^deK0ta1Ew8z_%|pJBXN4lJ}Kj
zFP9QoF|2uUBXSCiN5!a#dv(#-q{sP?oho-}5B*BvE7?SjqoMFiVZc}UhkDK+RBQI~
z?P+6z%n=~OM~mcNA$+)H^rzgt3q}R545eO4Q9IUX0(u_l2p#c=)?F7)bfj8<{E(El
zw2D0*8QAo(AVyL*CyB3RU;zSRkLr&hk19OxPJ+W_h{_h4PrTWlh^W5Li!u53d9Q`+
ze}tvSe2!FrLVVsm{{DfH<WwO{NKgQOA_2<30=bf1mzBH5-1sSYX@te!+x++12DngQ
zD;pM-Dq!))Tq$4i=Ao5y7qGJK3HY1AwBS8@x1@ucA(qXlK{A%Ok&<fADzeZdY!?LB
zyZ01pKCdkIX}X1fYVIc9ldfpJ`$8)dc#r^oxrK!uu~1i2AA^%Z9X<z$O|)0PCUcXI
z$WOh3Sp=Yv6ck?azc7w?gqF<4PwDPq_{7F8?=HYdTURrZz;78zL`AX7PQdYzDvnnj
zKctX(hYv6rkkmZac(%caB5Hg3sULK@NBr30uYCSt*Vtz~VUf6!)4$>;WqvD>j`{Sv
zx#-X|HFPq|=6#Joooo%#xQ3!gDH=|62kYB<<=iNp+rfTQzM6Rm-UWTPoTI{N`1tj;
zI&mwdQirqd$d-rTU~e>(p(0&z^E4TGq(ppWKWq#*mNVsXdw-~W$}!6M^@>2S3W{~;
zQ8^^e7p!7(>sbi;#vE31tuQto0f$r+C6aN=OlgFUEAr4o=21$0lM@%xPoZ(vHaz>c
zY?RUK)BRtWcG}amu6xwace-ycmH~i_{9yINN_1l8YepGj!0l$a6&67Hrkw-`c#c{f
z2LQU4QW0aQ!GM&0oX3kObP+OduL;MKZ}Gqr0vOPv69K?lE%-k9(iKS1Np^mkcWyRi
zOl)6@Ao~eyGHcrPFA=E$Ny0F)Pw@hxGGk+7oN`2w`|Xf9M6d=qZ9puKXJR`W5Xy!D
z9k9}0?Z&fzrF=$>@s24wMlg2ZSAoJkG(3%In21yY&Dn{K;7fQi-xRGoq-br85nh(^
zMArS|^I&g1aq7jFdR+hBj*k_X5+4WYfreABhF|UyC^3Cg?b*1aHN!Q1V9cZG9U6Oj
zS2AC(q;?c;6|CF9XrA-~`mS|yAcMbBiUHMvuV#a5A0XUy`z}6R%hT^LuD_8U$qI~A
zg(r0p_CRt%Rvb5^ExGa4ko>Go)gf0h+dQ86L%xX7{}GG2B}P~U6~d}^u##uk<;)Pl
zhlgd{TFj+%(nb0A@?a#8^-KKd!-P$6d%)P}+?rg}1c4W-<IjB%vBH>7mm&N92NFL@
zU%ad3Y^mlQEZ@B{^}x+Pybo8url(7%T|>~g6Xj7)C%hFMz;h`EH{t`--3D)K#Xek0
z!cf-><_rZUYa;?EEuCY4S!hO?+IuJ>H_D2z=VTOrK~?{q2YsEcpC$LCtT>9zKjtty
zHcxlR&}JUl1E+|w2W`qoakRA_X?$c3>;9?Nqmz^AeqpALm6G=5?FG)W6QpKd;%BY|
zn-dNGz#{E-+ch0gm_q)gPHF;+wuMXz0%%n~u4m01i|lchy=ZQGsmppebo8-)SM@`$
ziGdD~l=EbLaox6z#K}+8|J{HxO15N3XXQSi^H4>oXiNZ3XFKn2M65J+Ty3I!GgtqO
zgY-gfqmQOC30@m`cy%m2_&I7lYPRz?xDL)?cp?0Kju}NPCzc;#PI>nkBJ#kQ9O#-h
zq0Q60`;F1X^#8d4LGa;T1$-hfOb@hfR?IvJP*mRDO*UWe;GD#I>}yiPLq!GOzXSA?
zKkhXH0hbB&fRLolpy%ww6DDSu0A60a?=Kr|dWrBb`vE%iWUIku7K>vEn%yrPw<kwT
zmx2eam5`xnqRRk65Z07QXxdvogQ*Or(Aq8AqVGG$csFrmMcxkzc;o|P>>;v92M`rp
zzyHrVIn^~MRV?QV!LmegbpScj^WE?uN+C7?w}c=yo(mdtHRLo@&YZauUdn%7gu93Z
zsDXx%7mBO|+4m<R$hIZ4pGjo{Xv+r@Jva7nEgnju&S>`{QvdZ!jvRDo=aDB<;GsQS
z{7MX3ZiV^K64=QE`iG?vq2;v-+fcN;_Zr*VW*^m6!rJpu+IT6BJOq-V&CT^BenuC+
zeD*r(G(prmZk$@iSNMP92<nnug;c8*0JSHw4tgZlTP}?2@NB!!VGX%%B@2B+wxM0{
zd;_r)Fb9Bm$*r>#xK?E<E_I9zELCPuVsx(s+2D%Hu}o<9xb8=uGO+3H79#G*aNxJc
z349NC>EV1RMk%q}#fWEW+{XuN&=AP`N63L%V+7o3_@py$D^UW7`?ni5j*H%X!@EV)
zBMjpa=exm+F=}Pdb8Mm_5Fse}uf@Hf5BQ`g^T#Co8Cp=%HHonI;e%I;MQmKOEt#zs
z9v|w260ez3+haNq$+s_Qq~CP99Tq2kWjF_sD3okzR>6MGVy-^96{2u{Q!5q7R~tol
zVAf@-pXB&OTSuv^O{OlaZw6wr)a-+4OboU=EcgowR2eB02<8!*l%oPCH7*$AEG1Q1
zKCU~Y_vA`|+X8iS@D5E$S}}|`ehQrJ{-{G+%L#un(+$@jE3J-b9w#ZWG!38d1O@AE
zB~AZSNQ^U;L()1oX`FgO4PAyhSv`!xdSbDPa=BYKVCFS}Rf<axIC52=xG%?FX*-BE
z{@<i`%HUHQ@_h*j;oQ;^WPuT$#dSPYOn!)t?%3m&!Fz1C?JW?qFHFuK{q5X?QwXIY
z^HKl+IV>Ls?I#d{<{GRTD%{e>f!uEoctAk$m;5^wE>E(9KbVb)(dfg%X#zr5IF`1W
zXKk#TxybrYv<D|9AYZ?fh7)kV>rl--y_DjjHL=~Qy1n6e8x}33h9!YJ2>3=i7NR#|
zZ9h>Y7gwh)fsM1zbtZs~MlP=_t@L$+QwE<T58+ui<QeB;m?V|V{Ta;-uH0ZERqtun
z(fz&}=l1?1Rwbw2yt+fhGakR;L=H@*7cHw=XUOpV2fN5KC%YfBbv(!}zA3xkd2r9)
zg#h@R4y9FFi+91mBN}2aP1pJ%J~pmz`Rs#>po47idfJu4P^c&ndo5&TG>8=#lbu2r
zd0b%|Uu!PhZLYx?s2hQa-}7G#^RF&Wh&mgPwfn9XgOjvo5-A<wVgvN;HX{bWqI~kE
zbKA+V0vPncQ;ZU3Heqz{;Y&x$)1IFeIR@n!?%=9(pQD$sZhID~-^6HnKnZwWOkjYw
zT4=X)tGy;yRMQ{M{s_wYd$aq<*opj;R(5aC@M*i<JSs<5Be?+YDd;;-Cvd2(t(6|C
zD-M;rvu;J^_F8=eLX?uG0q{6JlgV};it+xg)MaZ{{_{6z&t)6(`YJ{zw;YwDGJe>W
zH^KhW1%8;wR8B4&MyTQxrnEn9ZIb(#|C%J)JG|m#)uGM%BrVz5X94wROvsaQ<X@Ur
z8phjKzdZEuL+ABu{TwCOV2aBEla66?PjZNPL4*jBuTURk0U0w7mtH&XpjQDs#PE_-
zFBMx?zOxvqMc*w#JUPS{0L579HhA%WJe>ts9MQI|n}%S)-Q6`1f<xmT+}(p~aF^ij
zE&&20xO+o_1&83)xVuBsa5?AQ`$p|QP@`(g+-uHniLqj_GMhHu`q-8mh6))oJr?Kt
zOTwfs?ovGIARqN+r97ohm?e_@(+sEfUo`XQsmU<IuiTXF;h)ME?XU$J1NE^le3}9*
zVaKR)MqVp-v96N>mxtHW`eCQrUV3+kbVH7v`!WVyZU$I)d?wg_5A7(AXHE{ox2ov3
z9m$K6Ub*I9C|mTGDYAV@>_ewOuiJcZfLaLae$WLZfCy0ca9#iebU_2_IUX)RFKAeB
zuPa_X?*I>nK|jAEnO;Ub_jGXFtEeE)I5RT_;~Ir?^&%`f{e8X|Oldk^#=k^f)WatC
zE;=Qpc#jRIYm~WfR|$Exp`p!21wus_qw{EEun{Uf0Kv1y5Ws1T9EAeIb&o`*mb20u
zZ6~s~zlsg+cQ4I+Kg}c?PT4;TTAbDt@y^^!K2s%<Ii3#~DW9wHn21xJ^^An#>I;zI
zE$ZfCE}B_INS#ZVM0h^NCFbQ8Bf$goYOt$TcB<%l^&+`*gX!e!x&nF!z%o0OmScG8
zUcGnSeZ547kp=0(3i&ss%dI~|lDZxJ@Aj1fIkSG=yYI<xN=jij@qLbDA--akHLQN{
z87iO){Le)b;-RxYHvkJ*f5{KGs(UO@4Hs2GygRut_t9iMmFh;29#B9`&$tep4^qbI
z9T?R|FFFtC;Q$TwM4C`?LU(=TrtbMRTv6Rl?t!ICocfzy>x=7NmqFgvM3f?#;xF$y
zUTNfm3(p)b3nk(O#VE}PwNg_NZYFKIz*Uc?HLuU4-7n{_*A(b%4pDwRDj)(xc$%5N
z-z5EGqmk*GD0aR^N}F7e>&PC;N}GQiK8q|rW>6OpWzK3WiL%O!(gYbD1y)CvFD_Xu
zpNGrEir!|gEkqk|07Z6*2@ijC7~HqFUC(W)Z4-R_Q7X2AeJawM%@l2;S9(MP0twO}
zA$o<1A{AFoN@@qFec$OvOm3Q{iXPteANzOs@7m&XGN;ELGt>l?LZgLu=zTi^^SxxD
z%IKOY-X~543Jtq=--os7fwZ_Y7(HG?7DVa@%Jzu`=@)IV>8YpaqGgtb996Y<?LcZm
zWTA<j9$H+FT7Txq6u|NEO0JEzCzC*udk@Gr*dj<jxsJlKNyMLXmg`?5Erga)l_Sem
z2@N*U1pC@L9j|qzMnFfQt#v<f5bsfsLDWHi2Q|n<hz12HE<?e!U^KwPIfywB=oF;(
z45xW3pggYJ+YeVz(DRK*EEGc!F$(We1rd=eE*{5M(c|8nlG5BmGU@rO#swZL_7^p7
zgd5rp(ZJOhunL&=%gWOqwFyYG1Bhlxwkn}X2~H}2#?@VZ#wv;+LJpBDd3cMDQN^PG
z?4_}C@Ge(sMUb%uvD0MUbS6JTh{8@>{q~AYpaR`b&ka-@<&tqLI?};KAKDQG2A~DK
zEvUOCY?ok-`&I>W0&RQYye!-7P2%iOgHej;f|)xh*c?4B$`-USnjV7AR`Gt8QQd}2
z3f{a9i`I5x(4|$BW}+qs#{9#qL%E=Xi<5UTl@I<gTT+u}fEi(21r);cJ5OFh#0mb-
z3H-?*v&tiQ!l5Is#9~!GYgl9m{o?DH;co{tfJpfc)#b?e;i^cu^ZrZRHO0Q@x$n9~
zk+aB#BJkON>}*cQhJ0ha(2c|&MI5e)XUqD5adf~K0Z#83<QvGVk+6U%J={_BvU?iF
z`I1<eD^q=vb4_q~hv_hBg1iKCi+QRo>S|>{Ve$ptRTcS*aF*@Qwhwb!+TM31FsR>@
ztm4zhOgY(qZK8(FKA0cnoVPrMmM3~*05Y6KZV!=E=#-_T^7-^mXy??*VXG)QB5N$z
z{336xo^T?SY&JV*Akp-b?dQ5KdO_)B(vLAzD}3f2?J(##jcAam8*o_#t2a8XV1*}W
z?wAo>=8MNlyZKF<Y#ur%_%4+^YSy-69>JmKExjZklIK;`-`CvOkb>q0=ez!vVMUQ+
z*9q5VWD&2!m>0%oEWxT_B{Z{;9b}eQwVX?(<tj~3%h7+iir6V1`aS+5&H|Fo27zx!
zXibBGf9<X=r?^?q;!@_eS96|JJiWhwkgkaDGZ?VQZ}Xqe(0*T+`TJLQZkORh2X#Ag
z4}zV+8==c48Um*3xO)GV#4{8Diu&1#ADugnUjIU%I*i5;(M^Zy8|ajY227Uyc({VB
zro<HXvjT!ugqaH9DIY~H>AVkL9~A+;Uqsp8f+zQ!y>y5sFMF7D+F~%64SZVz^Qj%*
zS;`Zy)bEQi>P@<;Dr5uL(*WxXronGGWE$Hqent=-Md9cc&)tCd@GHYjCT?U_=?KSL
z1l@_iI<sJI94el9jbeG6B0Q?2n6;p)`?BGW*z-XuHApT2FV`Or4$jNOa>P+2zV|#>
zJw)F180{*!QmW3UQprDviJ4k<MGOw|Z20$H=F*{XaGaXfezmL!9hw-YQJ|_vPr7s#
z|64M%^!J+!mWvDy7ClNvtKh>`Nb)O>-7eJ-eH~C0HkogtQEH_5j-IwBTyVuZRr(}4
z?|zqO5z2(}m+-LhVNH*B@`w?M%Jr&jp4O=+1i|z_%lQjIc|V<;l6f_d38g4fOa}L4
z3BAh+%RLW{0w2%|Tk9~sB~_(DL6#u&`JoN)qxEZ~q?z=2x<1}-@0?~&L(uKDyAwtX
z5T;2X(odrh;VobVFFu(%%lBS+-UknZ<0jXH!gAi>=yReLID<HX9ixH*dr6;P*jbz*
zGpJI8NNk*pA=8?6eiTa;e;O$uW9#HSCa4ucGxy?a)B!&x<71Q^)k@rx$9^B!@*YIK
zuKUjmCU;X*hLvNa3u@C8@j1mF92IvST!iA+6T3koXg<)W-pSZ5N@v(V36~tb1tOG|
zgB*=&3UjLXcU8ZPJ4bD&<3wu`4q;*i;2xjrQeL=CJe<YZ5&k;(?j+JBFRI9Jd(t0E
zC*;ZFEu??U*cbaKhVc$UiY+PtkOFFlD<%fUs5+9};Lr<kKm*JAm=ZbmB|_@nbbHwE
z1N$^I3|5IiNxf47>V+NtlI_%bl(y|3U{nP}UC|P5-er+;k)a!yM(9^L7{=<NjlJJa
zB)LAVjp&{I=yqtGp~BXKkw>!;`YE3RFlvL}cwIMnU*waKd%>T<II2vsV-ro31(ntT
zTFCe(@u1Td|5ioPE0I$NCwpl6$G0A?6Mx4S(Py7%u)--YKOcrg(-n8n@4r5A_EYxa
zrq0y8UKVzf0_rZE3*Q0!Z?bk??>5oHKj`}J;-3JS=A^GwxvoSqBBkygX6CePRaBms
z4_hn~Y-|6TQ=TH1Q)TS`WaLIlXTkYa*ZiwOM>>pagX&ycmR_uBasgAg3PIIM2w*WJ
z;PHAxtBwxbM9>XK8&VNt=Ml?#+AALhm+xM5|C@v>ahge`s8&t_yg4{-QBS>7khn+m
z4s>fiZ#8w|l^S2gs8{2)6==oR2NDMBzGEc*>kcCKOEI<m70+mFu#{=zU8+5}y<e`z
z&NX6!xRl#3z2&<`WQ%Em{<Yg@ygp$R;3uk5Y!sAnoK!uT`9YOt(-ra@0AKRt`m*We
zETC3^k4K*rAj0xn_wV8V8s~$KHGtyhH(0~8d>I*I*2!R}gBvdeEcI7lbS|Xr(Y?bL
z@`tJspDK`45^#3gb+(>~chG;j0wvrTeLR6A;;IHuIgSWZN_G*Jja%1ZJ?@#4;K>0A
z_*Nc%f5SgH+cw#dtBxBw&#$*$`F3{!ojL3F+$A~t0x733nQ~E87TcWkViEiRQ@Pcp
z0p8LJukvAG*IW1zldJw9dl237k)fWJkre&AT<43Yp$4Bkd8l{0*v$G^Q)@5mC#|q8
zIos`qRCf0obAj>EWG(*3AO>xU*{aW*WjFjFJAO6N!;#agE>2MeVeLkEb}dMWA8EGU
z@u%`*l&_{nqf4w!>y5WWaxuD_!~`(ipCyF&_V+FEb`Mey=N(^wdm;;8=&-*AG`7Ri
zSCH-B8-G8)lyWX<IQ;Im5`o6`Xmc>t6Z#zZ*fZE<WOUwH!^*SoC&bH<s`Rg6c2oHL
zhk^}II*qKHWN$6OgofBxh83x1O_&~IC(wZv2V7ZzcdmY%;t{XccW+-p<n-6u?AhaJ
zhB571Psw&H4spvX+^v%A-HT1UA<)eVkIUXqzr&@9xqPppCDs#WsN>6Gwxk*jsQtG9
zT9M9Yqc``z+!>o~HYo@)I)M>NJSs2XA7%>8fOMRqoj@n@AT;uT@+L3MuV<I+)Af8S
z>MG`?Yuk}3Z~z&MNqE2y8EG1?+wXE1#{Tj!FEo^^bz4z*z(ZtWRQSe}N9C&*<|zD&
z_X8pjq0QA`WXbUSJA4rVc{-GDcI11pC#;iZnn~8EAMm><20`ra@O@sB&;i_s@gH(P
z-_QL3-(;D0X&?Nf(yBq#uwyxlKU2$vq^Z5(a2=<^1@JO3YlL})Xk=pqW0~r3V3QV(
zn9z&rKHQM|7b5eDw8y&Ncs^bw2p`}&^uPdO3`xC!;aILPaWn-2D?_wYA^ZV;5x$t^
zGW;e~f_+LuqoK9P2(@3~+6{WmWZaK~?E8xQxed39SStfTpnZu*dn^B^XC8&(`eFWK
zR?xNg(c_2PwR6?XxIs%m2{1<qEf{TvGX)`wZJ~F(c3?ND>lX9(WBNYEmDfGtAW`h9
z0H%&eLiR0m%N-#OPHCC*|I~24d_9~wdOkcrBtNB6p{XNqyTCxJWW3OOVa)<#IvGa!
z#<7q-5j_SE2L(q?1#tPTV?a6~nM!!nj<tpm_u}qau}pHJFnxhu+G`KIGHbl;Xu=G(
z3ilCaZ;Bsf)fhb&Bd74}FNK}XgNV)w*j+kRj7!5u{VUx>3%o0BElY#()fi3=9#4Dj
z3GM$*lxYgx=4k^XV06gs&ro@syGOlq&Lt*KUwFL&-URQE$hHJ44~aK#W5pc((PI(~
ziYK04rK8m$RtR=IF^s4DtL$fT`coB2>AKXIUVl#!FZESs+~|Uw=?92DD47`UE<@sW
z);)*DMfxkRv2Z{uE6&cu+P#?y)TW_EZ#u1|yjae0*q30|!7|@{sCri;t&+uB75U53
ztv+ByTqxiU30qKPSPwID9yiZ$|7#s?JB-BV)`biQrO|}hHk(*YsuX70j_<fYOX)G&
zb1OFbt59%)dKb9*iCQ(N{IlN|>ALlR`O~nB$?Iq{oj4!;?d8;Tm1q>~Db#9&5ho#g
z<Joro!3h#MdXzN4T7Vpu<(fO!-QG{Ro{G+M?`>ao7<b;3>)5445GtAVzbC_{8sgfH
z6nqdO!s#w^wyk#oT;%KvY^P0Tt*iTOVJ>ES%!$Ft{e$D^=&6*znm6RUmbNlG6zDER
zhmcr+6$J2_9DU4tl6-hkv>PwE6@LAmc5kR<j{H0#FMjoU;3M!l3{#(YB|$(ob@32i
z24Tw_z*V$L?Nfg8!iHn|_Qr-kiW9EYickc<ldnpFi^j<!x<u2}PN5Xp{o7?(WjX1_
z-4S}z>hY}8OsQZ72p40H4xvgqAFPm~Lc^-G$q@9Rfu%HD6R&ZK0TDvv_&=_~!R3TB
zoWGkUu{iM;S2r<WOmQ>CLJZjjs^O?=6Ncx>Mo+bK2>UxAa6+6cy8l54d;eo1e_49x
zzjcC^pn|E#N=IYpMn*OFS0wi>slIFQR&SO)vJ!r2$sTJRJO~F)H`?3G*MQ0C+uCOO
zn@@RbPQIeHzth;oVvHE&=hg4auGC)^ivhTm4=P8=?=e#0ex%NTR+Vdw%W=qW$OCne
zM>l;(A44c8s)zpd!g^rwlnWMaIH=@s|7(zU{2VQh#i+0m<^sn8==i`yD~=GRDgbX)
z`sh1YlvWBJ!ouxB=JpvZA`-yXE#ET9siH;>=5GmS#YHfG8uBj=2mF}jM;>UJ+-v-W
z@=>m=-OYw(VJLtjcnmh!aqh8f8<cssY$q8;r^v7MMJAzAAGlKU=He|B28D3;bmh?g
z<Zk{>l&S6M>hH@05#^Z2s(;?nh6cw5y-S4fh4~6J{YZz6j2AfbzM&GgFgP4q?&$G=
zz50B=mu14g@ab1TRtHNydI-DN$mMU>)9P}n=(XN={^;-Bog>6SfkFy7Cl@~(Kz0bX
z^D`AX@Jl8Vt@KpvJKPVvr;p>L7u}5I|J?k2;|IFHn|ql5iAhtm4C}nbspQj7lN!E_
zV!KX>tt%Ox8~=!1U*An2&}u}^w$v>1v4$nLaQG}~P~`l!0-C-bg}Wq#J&kTK88mo`
z$sXx&6HOAZ{!5DXdyT@<@`*_ygI(dMK=bL%JFSz2S@^Vnd4^!R7SVL913qaKwF6h`
z2caP(dgkZ+(^3rICS{Zm8oXGZE^^M=6{c9hiJPY;#!Iuf5R~1((t?-ahKN)gRy~($
zbYnw2I(hkYbdnS~Xr%?CL2gQ#)@|?5UgCkfI#7}oUIE!|nX;>JQdS-t@}&Z1*$(&g
zoA)M+_Pfc#r5p$R(86%r^7X$Xsg>5BDd?W>hgh*$CuQwt)^BT$v0aR?E#yA(;gc_T
zmA%R_eBW{S#hentvlrg-i%#xkbz09caR=g{0F5}N*(91@q>$>r9EJgc2OFsF7k)Ye
z0oJ0P6aW|4P$nGUo+5?+nscnS2w2pPUxZXN$5ir;>fqM$64RieAbk)EzT)Q&&7vc+
zT90A<=`kEKi=?gJqbspRwP-)3d@w_`SXuz(Aq^hA9H9}@)qD7GwJOmUD_p85H9`;R
zU<Nq>KpJ62bdeu+9M|Hh6+aWh?GokjUf)A5fcEwu<)Do~Fba1Hf+NUdJMBXLs|0av
zqXoXOXHz(+GXwkf%Zz1NIA%e%I5S$c>_`Jkv8@6Tk*RHC*xR4k1zrSmn7{hcr@7k3
zNw+j<Z!n$6a4oMPaB`k+{0(>}I7!cO4P|sb-8xKXpvbS;Hj<6VhmgMlGq5w2=j9np
zQUo~RCK!6(oe=MZ@PShL6;SXP1p~op8`qua*`O(}g3A;B|7Z90;vT@dQ3(R{oPmGs
z-;i#&zj+A3&Hk+oMuqv#tH<xE0wH|hTPj`CNfW%cSl1Qb`uV#@e?R~2H9|ZoY&OC5
zkGNZnUYTl)p?^yu9YLbE!<qjcM=2Ez9Q`+gV_%)~3+0W}{`gUFt}P#KsTjAz-m||f
z&Bow9@17d@-)==XbhtYNfkj{|R;k?qVE>Nh#f6wY*0znb%BEi(8b?{tb(P!occC3C
z^rX2R1`5Ec36twg5gj*8j~R>=--(&4gUh%mXiLLIWeI%)H4t2R_m51aiH3;9zI{F{
zkH+2+R3;tgr#vsPRH7J@lu)t%DPywj)Tp)__YZ!x&QY<^Q`1pTrMnZ;m4QWCU&@L~
z?tG*ORO@JM%a2+;W0VR&%e_m=;Kusy3|az4%IK_H8ee^<l-TCrch}my@7$?I{D#vE
zA+!IPOR?~j*~%=R3@wi^Duu*y3+a}6-UYJgO@y)vn8qR5W7bY;-F@B5fYqKB)x^Pm
zANXCDIBE7~Md?ByS*Mu9grjdEemfffK#;v)VCRk0!9dLhKZAW&&Gi1m+BC~IvP=(h
zP&fh^r+-4tax+q~u<WmWn$Z4KIuGXQ8QN;M#g`S9r6Nis-wEi_JSiafYT`%QFG@{S
zHTi4~x_!*|v9IeEZqH|mI;EoHCPHiluSC_?<c#m?DRXi9U;o6uc_2Cm`xVSJ#|4d-
z(~ixI^j^^G5g!OA+1{oUL-^_IB=6gd+FxfreSE&s3Xkp^<&28#!zRPW5ZmSU1*Gf>
zG{Gqq`fC=*)_X5H{{2ol>SK7uLRiEmxB?Kv(q%1E2T<|MI5Pm48N^j~_hR}Le|BPO
zTCN__2dkOD7yyzO{jNaCM1MF{6kIB_s?zvvAvl0VS}u-QF0D4Z47RC_1FU|KiV8<1
z^v!agY=OLlkR&Pf2}lpD!t{;9-5A!5g80}_nE1J(96J)fQVNPjKSp=l8cZFw;Xn|s
z*aL3UcO~qBKmOdJie9%6&yExaI5E(e>3@<26tvfWqvUWp*V{Aroaxp1u)Q#U8J%j8
zMar>n-nAIW17PE$KIIk-nlu0jsJN^*A{UU(En3g|r+IvM16mQ=^Z!kgF`0e5XBU0_
zF9K=?l9NUyZiyq+@x4qN?l5%vT#!8JQ5(R#!7y`wdIwbm#ykT{cky@LuU(V<TC=#s
zora3>Fq$#34f!QBnVyTv-o=+_NIt;4D2E;^&GN{8ZNHfhpIS0gQg*BFLI__Wfg4yy
zM>f?{Jw~Ijb<*P?$5_xmea_#wnS6Dt2WRhZ?yau<aKgg~UCJL<A?zqgYbta~(+f5O
z(&|mAPRD-b30b{8vQoX|v3o+?+S}*=^&M{gREq&h8()*TBZV?U--aA=T)LK)?IRvN
z*T3YM+N_cZeMynel`9?R6;$tO-fBb<9vdEt+2nlc^TLb?Tu$B1qGyZ45hnIYl2qVS
zQZB7>C(?-ksqmA?n`^qz3|f($rSst!wB1FuL?#1$%G2HGF3EuJ+yd0bvr-CF!x)qc
zPej+jFeOi6z1^GM;WS^we}PJW1Oy*vUOr7EN2OJo(`!9-`6`ijtKyVyhpGXqo1=Eh
zCxT?@$<t08Dlq&U7x+(`ca!sL&+~D}j|8J{4%L$Kuzjm`{8wx`w{xSnzu^czd2BrF
zTz?>dYyS}EC{{Xo-*;DC5pFFtYOvtTuN-YtYwwkt;tzJM*lyxrgE1vtN9;d&I;kf?
zFKr(S&IH&xUf#2YZm1=n-!ok^UzO8s_0YNnm>Hc>YRUO;C#6ZY+)Q-b{CVr|DAL>@
zJk%-5F?oOR4jcBOGbGjeT81fujI;4rp79Od6z3+1YWBspvM@A;m;eP(0-uxP@(!qv
zG;EO;2lrbwNdP$dgo)Ol*%N|u$YV#jBdnenObtMQ8xe%(^B|j@EdVdYuYMMz*{8QU
zGzda3P8+>swtfy-yv&VDPfL~-oni|8En3bfJa>x}f&+&R*hJ%(`eeb~bG_*VMEhXK
zx%u&&-|i9kzRmsryZ~CE!n=tLG`CJ@Y>pIfRf$@iiw9VwD)TqFCXqBc_q8H1(^lL~
zq|61^kOC{rq@$3&=uK4XBYL4i5<u=LMa`!U-D8<=_q^K(Du(O5+vS5?eCo;zq`=z~
z!8E-sew6_|HDsc>Y0EY<w>7sgU`FkAGCif`!X(c~G3EO~Z**c^xc?@#UY*rLr2^j3
zA)8`RizA0o;G~(V+XJGq3g!V#7#ji6@}@Wzde?B#?gI1O4fpIT#x<$1LZJnyMn=#^
z-70*<4ZG>%iu-5gP0W$?1&D|zczj42q7UzayP;2AdTZ0i25Pje(E0r}<od{D*<o{(
zlY_$M^7Z8HRewbgGor~=QwS5l7C5R*>z4}S4$3RtqP)h@w@<JCeCJli!R|iZrjk!0
z3Ogose0H!!%JhO}uDs<^T&-|lch)Q4Y8A$QH@s?xslA7wl|)!0{+Ejes5bSx`}8U!
z?-Lr4_0E58*djv&$pcX`vb2nnA@4dmB)1pKw0eh=&Q=YY+_mUR&~sGtzs$vUPWYof
zu;VD0-_D(VN+JW$7&!Uba4MJnbVFe^Gu^_7IW3h1wOA}%pUC9%RKwXUsGx*Sldoj$
zXD12?9Y^*x#@IB9uWuA9XLiugh|&xv+-5O7R=%fEKlNsFKRT-ZB1YpR{m{SPWU%j%
z(YdmVF_)e+wwQ(DMjwy9p<ua8?v;1huG_<@wl){Y>7#bq7N+8{qhGL2xD=R7OV5ek
z19Ie}H@Myo)znDzWl5e2-A`+j;E<8s?z<MDn0NhiKX&h2ZI5TDO5?6k@?M<~^Zb#W
zvHbUoMi3@hEsNR4{qz)^WeJB}Cq|`I==MJa0>hYuU8-44#sFXCueX_*pDc=Akka6>
z&SMrm<OSfJE6#e_S%*8yQ(9Aa5B7da9U83R_upMyzIhdo-@D~=anF;w1(j4%JdfDX
ze_)|NAPwniml}eAA1|+G&hFpKZiK}C)_W*TfbdOnl%RzTI>Kt42Mz#4y!hhJYT<8w
zd#ZVOhUjmsSd39g6$qTu10SUqMLr$W7)<7bwke9EAyEH3!l_w^4N*6x7b>23@n997
znWMr^m0-=Dlg{jUVucG1RzuK;7F$a2TPn?xql0T=LB>h`kR?Zts;^u%k4g_;RC*dd
zodt(zm0F4euyD9}>o9@ZKN;t0x;?GvZfaWBwI({X74&u;KG_{WaqCv(@cPwX|BQf-
z3(y07Itn5959YAA#KPM&kjV8rLxJ6{>FJe6JvxA_uwtuU>RoMZ)5R&NLBMC^T-Q)g
zXT!I|aRu<yS)r_g?vsHd@o{z{Apf}d9Fyu;4Mz^FAq29%&Y6wrRRG+q+WojYt^Qw&
zmIbp+C`XFsLhVAQ=Gvjsq3pY+Gy`YZk?UF1;4Rnf=d<IhH)uR8|Mp=X&dEo*!+%uZ
z18n8`wc*}cJi~I|XVHuMT;ESi=j}QFCn7>>jrbAXRE&*VED0d3UMgkBq@Qt$B=j3Y
zmg#xtqU1NWaVV#6(k_ys_{S6ER6KDFZg~Jg1RXL56~N-I>`O;^=3#fo-EB?wVdN`R
zd0qH^{fr_tMgFKDvSy+yn^C1enDN>l@(Oz`TmV%SY<Q)cVZ+CoG`&3OYJ25A9ObrB
z9&Nv3iuUZxy+}5`yYqiGf$(=Z=L1}<zsJ0uEYS;h^H9u^-lZPh>->IvBnf;uVY8D=
zp)wLU*{DA4eP06Z!+a2V5Yilf4fcKBMFe@Jryy2c6)1jd6VELQ6+&LJ^N%6MnmI!p
z)@>nP8XkI}`Uc1(-VVqdZYJ2F&O{B$(L2=-`Kc!-8#p0)j32XNCcv#Ne72WBU1Jzv
zB(=l1Uqx=r+Uv@in|d|K`58y=SAJJPQZPZh`Vi>1FQJXQ7KkM?<VoY6J;)WmcFVGJ
zXjGH>7iH)p{$R%4kr@S_w3NJSr|ej^iEF0>=%LED3uG(ZL^sEeZSIJe0~8^TZ$<IN
z{OSXC^Gd##FNh=)Hcmb#tJUdxdFCHNr(C4SrsQHg8jqoAxs*v%GuWaPSHSJvr{PTQ
zpscl&8V2NTxEq=C$~7BY%a4U_VWt5YY$D>M_-ptut6$h>l1RLK6DJx1Y77q$Yv-n1
zE?`5)e8iVdgMM&}w|K1*N|5>DHLnu8pz^`-<Gr!()OUOw2I1grd5zoGH5CI(5oztN
z4hQU)m%|)e0EIyg9N>B1MEHFXV<3FtnEW39aJO7l0HoU?>UD*X{l2#>H>0Ge$7*Qk
zCgPwfC$rDF+Y10#G-(R--h&eKU4Z?kmtl;!8d2Pl5J(+Au+abcZF!i;u|m%?`GCDL
zKFXH_aIsZT2|yyOUsH=KwZ937N0TZ^$5OlZzU+?)nAE+VC>}8t=0++d4rcl-u3@Zz
zjWxhT3s3VnEp&YpO?GtB>ZG%usuo3trT5nH@#mvO=D&y1d|aAGAdvnV@FlWiv3L(+
zL)=YR)ue|RjamE1g8djHKh}U3?->g5zwR3j@~H|1zn<W*(EHX3+0o}ovZ|30awjGK
zA#J^*c>7Zi2VDJM;^I4-kfMe5eI2&51HyenT1ROYDDs)M_a`T|UOT!zC82m*Eo{~^
z7frm39oNO<6h?OCNPJEsp_eI>>aX=i5nSd`tQp3ldOh>zVx8o4?W#;+;}6u7vC@>;
zg><IC$r%+$tNwiDHOVM`1#cnWR6a#gy_j)5x`SVf3a}pS4c_nbcVT=M<7g{T;5iZe
zLj;c|T=Uv-FS`R%5)1dGA0(U{Ll;)DvZ@Nr6zz%gm9~2|yIpzXL9%x}Z_WoCBLvsx
z<@N19T%L<SoMp_y!cct^K!R$6iIjz8-uc#uU0_ml8*kOgu&iLS{cli8-4JEWEc7!@
zqc3r8h50IR0PF#5r&gfWck|6m*m+g*utbjOsqxyjhjpAUien)T$o}0DIzhBVSMCiY
zvtx~N$P=YF;_;-}<e%gvi#2|+ax!|fe^Y1*$ipdj67{rtiU9wq>?cTK&oP>J+tc)Y
zikGxaUrM}T4xBCSEJiieuFB5{KgVyzZ)aUFR=_tf>-PffLG~y;CYo7xnrSj|bq#E9
z)}~=yIrKtkcjzg{-Q*L+HVeWM84ND^soNXuE7}n4V;utK^EQbbnZsI~zctnfvdd%r
zzAR$BhU@XY4F}~owKaSeMtbwAqYaw-_2`xxmpV93la0ouFY!R~VU&!M1eB`O*P$q7
zT+$lKgK2&>>qyt+Plkj)P?b5qG395!_}BnDughQUS2nB)3K(^`oy%CdxEGwS0kxS1
zKD4%+NI$MS`pIYBDW!w`J`8N5*^Vljg(O@*`u#ad9o!OfUIPFe_)dJs13CkJO>@FL
zN3x*!7G$DbHp?L<ON|cUuLq2A;tzfxs#Le%>x#9752EJ$t6hWp^Vxr>gTn$OGvrIm
z9DnM6r>=a)kJco~5+nvtAKRf(B?Rnrb95tgpg~Fh{h<7aB-XicidXXMYxM_%K!sWW
zHMIZNeI@#VzDN!sFB-Xm9!21yG(iH9OU0F!^i|$L^0!W=hGXi&FZdIUM9(&EWO#1L
z-nBr;*B%|R0+CN)bfXvJnDf{fyY;BXzD@&+<^!Y$)kx-3Zt;ERChI4^7}-F9Zitf<
zjjoHA$w=|*N$IO5kx_bRf9vYH6UP@^eLBB+c^M#&vUXxjGBE#i#f!Oi7pE#LEJdP`
zmQe5e^*UGkRP2&>z3wHA4lg`l|6VYjw1I+%3sAz%@;2nmdS1;C#Mr|-SLgS2m)0cc
zAe5L?QX3M>f3}c>qfn4};E*RtLq9-ia4}@k%(TS_ccRY|`LJfUS4dlEo%gTHzg(s!
za^NW&Q2g512la$<y_s*8_TcGG^8R4fIBj!C%3`wq_YK%oZg@%()rlGe9y+?sBMB1-
z{=og>`M5D-JxZz5CHBAvE(|HF6~QzG{(cI>dAF|^mkTwxmvuu(qD8L83ov$Fb~;w4
zx2MOYa(X0gRcN^)NWrHuz@%CKYZ6eHyfiVL!L(BA&}85AcVtjVzBFCd-0d4w1G*61
zpZ3WyF3CM{f6Xe;%a)|{XS%jH>yS)&)Ps`~uTOxG@jASOeh28$OR<?#YMmYU38dm4
zbid__=6|4*B(G^H79Ng&yz)guvqhkQDYqx78_P<7g~qLNalrdQ_j=vLa57OY11)w!
zhwXatOL-&gQ!N)WhN>;4n+9PvdcmS{tfm^XDGm2LYef0)?Bd16NHP-=XX8J8A;;O2
z^PHeZA%nZTGHTU%1zVQV+br6WV(1lY=#t}eh(qX+?0C7BOwrAEE*0j#_Pv3GjdXOJ
z)iF;-t{glGPN|OuTMwGk*KN%u4pDivWc$bDlKt$<%b4?MAp^KSP@?X~_X0QntN?;#
zRva$*_@br`zXV-pP+xfw#-Z@H+XV4pREwzp3EB^xqow=04IeAj=4YxT(E%PI{e&)7
zR-+4BIJw8^P9Em>-w&PcZDu@a+UL}Vz21c9-;WAGWW$jmoi_J+eeZ1mOvT|C!GE1{
zV}ogj$p{f#!+D3);)k20YzQ@Ir%YSAWOM?d7!%TRHmRBK3lifg;a$uD7KOLtvya#_
zPN;;k1tC8N0pF49KSmTCc?ZCcscsdPW+uB&)Fc|_rqDEICTMOahe@Jc_y{e^u6|6$
zLyeMui=(vv5WhdZo^{Uyx%!jj^ZXgjT&U;K?wb_xVpBO}{S3@Y%%n<`YzF`Pd>Phx
zk9l)_lFTLC%wu-GHT5m&^?qGdv;Yva2~n=qNz<?taeovR)ax@8pde!0*b78bVV@09
z2HFz5F{n7qE&mt-SbQh{3elD<(oCMF;J*mYD}KR!IO!}9;7yFDdk{?=sfZh?c=j`S
z+g#<ky@Hk8jA@GqeM-Up1IO=)n0qy_lhL)VS<Ouj(6MMvKj4`5vh65|MYTb(+wwu-
z0tXg&s%^#)AC|t~-L>5>NbHuh!tQQ}Ccc&wp!-!W%&I5HYjL3mC@G)U87qTfjm|}x
zY$95BAP5(W!ZS#!8Z!CsGe4US=5lu?sj3?Cbfnl_oiJb%<lteP#`GzW;v)p(g4f+?
z5{Fr39^rN|S@b(8C0nc2u@;iBdmihvRNxEtu~m52CrM2q{|bp&FZ{5=Q#H$Wd>UqI
zmg;c(G8d{)FrkjZO4D48YBf7&Fz$_LR`KVnksRs_pEKvDVY@PkPJbx@8s%Lc8>EGF
zX6;M((jINPjz3~OoagVWr98YO++tQ+z2rz+D`Kb)!fF`9YXei+b#YaZ^?8p!aP)hp
zgO6DpY7!H%ZW8rN(pzSaNAgwUvhxX@f1@^GWSjh*8LVWB6MiY8We;;yoa>R}j1JgC
zm~LA&2M;3#*ypxbiM9@9o+@`5?v5(bj*elj*;@~zRZ%XwTBzx_oinfBKhtd^Cb?pj
z&{qGp+*q61>(VTh#joBdx0}aU8vazvOWgX#yPkt+I0XF}a&)OGTxYzJF%!~WN-Xpp
zoPGL|H?aJ=PA@spn)U9&qTKN{_uQx+_YN)p)4nz8{h_r|0EE7=sl)O5*ZApQw`6g@
z+n+{^N8GLbEsjG-cde_<+d}`ooV+k5EC)D)fPjkG6Ik-u5)2G^<OggV?O_4{um8%G
z((ZQd*_3-T3%Ne`m7T$DjhS8hBE2QfM$`uDxz6Ex3E*OO_;mKK?Yu-no27%73|TFK
zXww6l=|XmvY@vm*v||?m=?vDR3s9a3K$>}zkk6fPKQ_i+wprWJdlJG#qDJ7%<f5^a
zP8_l8hrLTdm1W@2-NIifOH^R^7EQ0LXc|M6#eMm)Z#7Q;*TomYiA8TFDLM}yEYh_v
z413`u!F=cBrn;3w*nxY|J{HmP<Oki3>lg`9kYf3B%6Pul9*l<i3=5ryik$G?VgjiT
zTDCuk?}V7Gf7bIo(h?GT*TPFXp+Ve3G*W{mzxpM5JwkP09OI#T%|U18XtRO5=KX-T
zS@bp#=&jCz^WZk#>vHaiji19FB`wb+b_}I1B)buxwh-&61C;tq`wJGz|41T}E*?R7
z;0Ao54hNeaSPn>00Tm14CV3&$XmXRTI%njjzt#rouC}g9kz6FZ@qhE9ydXhSznx+(
zm(u|FDyk^I`VvKM05`26u9!>#IpKYCpeWRf>P)b+RR0bPFTFW|RU=y{k8gy4GwCr|
zO{z%3_x+e4<L8j11cI8*I8GP0gcj|=uLQrdx5MV#RbTEDEwL;c?~^yZVOhBQgFmP-
z&hN<xVWnD9bi=Q05(w$<T1cwLYMd5t%kkYshL3)H&5BWCVxaX8GzS7N{Y<8>qf-j-
zS%O!ZP>ux-M72JPsNRm=mBG~>|LkGPy7(@Lmtw-IO>GTMLC=D3<~v2kE^>io9BGk7
z_ku((S<RrV4AR&~;ZbuG<h7p6${|fe5>(I-6L11&{m{-gu96#vlSJEyB+WcDE@Sf|
zvPwp2EHt0&+wLt7kJqc1>6pMwC9O}|Y(YbwuEGu0T!y)&6dpYb3H>v=NlVO!qWuVZ
zL0vxolAu=(<lVauacdWqK9XbS>pUy3w?9uabz+^G{A`d<=Rawj%H|}Foe@rxZp<V7
zKyTChNb3J{IMgv$<<#RJYQNix?TU}9`KeMq7f*UjGwtT%c4C*!g6*>FY5&vt;n2jA
z$HKVcMq&3OJD^v^e>p3T^Q{xeec=S@_Wfk$E0z}j$@&ux34npsu?dbYnkz>hkt(Ju
zs0Zs))_@AO>bD&$zlF;fq$&B6+FRN`u;0{YB%xZo!+-7ivtIot;d~H4<;Vf;dy;X=
zLPQG^Z*<90Tf}~hl^Y3c?l~{<Yvt|AO16fX8!Ao?w*Z09U4-b#_oASU)CpqusZm{-
ziKx_q55q=xxue{MyF&h|+0|(mTZh>aN<sY?G)I1awmXKYPnXqvLD9^r?hGuNllML4
zwBO{<1Omk_MPEeL(R#GR!@*HQ-;4p0^Rnwug>^K&ALkUc<Iv1IF=*%7mMbly>3<cC
zm7IKX@wc%~ji+D*U#{2&^w~!aQ-mkSgFjipNMw+5Y~-5rW@IxM@&s!O`1~HpQ04UU
z!-uA|NZxL&O~fEc=P>`Ehj(!<_oi)Euh|2ABhPiOPb>JqrLA+AQZTF?0_>f5&v_4s
z0-f|kHt&lMfui8Bwz6rirrb%($JJQnD*1-+@(fy{Hh|Pu89{!jkHvhre%>==Crvx$
zx>CIMQgV!YXweVD-s<(N0K~#suAkWfo$&ZyPju=cc<`Vmfk&GbhKyG!>^`xVNC`=O
zs@=!&=Ek0d@WIGhr`7N5ZW?dwhKtR#ZCbTYB<!NkomtzT8ol$SWg4#Q&-Ea5-JhDO
zvtTSE653Wf10NDX6EZps|A_H=#pl`;>8S!FZBe_lB`$&eJhoKs;H0L(%;&1P!0O=C
zD*h1sR<Uh~OLSfg{)wpJo32j<n2)^#HYJ(k&#H2a7aY1JgpmcYTZF@Y60llBoU+Bi
z#Ddf;R;z5b>%Lgxf7h9r5qF8$_H$X%8fcOt7vEJur6*vKpR)!>CO#_{FFau7$(83e
zaZV)a#q%B?T)<TJ?t5)l!hTKHK*iDYI5i<IGr1*loOb7`E8*jP+VZ#dJq7uEpi|%R
zyV~>RTb)Yk3iabU*KI*SQN7uz(^q1+I;%LzZ|%tX9H)*94P$4oXAxY4kx29Bk@qLT
zCNDMQt=DO5Q<+)E8(<S|(%Vf&=Rai_r*Z47OE|~P$}YAMOfgAxV~BFrMWNwKuiP4k
z8dil(xNS^dfIGF2qby(Jmm|?wN?Mc=O0sNQM1|yh2%SdJrPih^3N7YQcfZ=!+)YBu
zVMZ6c=_vyj!nXY|LjR*r&ScHNy#nbwkAU=3{M5zGfRXpM86>rNn~is|)LNDjKE!Da
ztGOFTXBSFAiIn-_?#Z$}KS7o>{RlUAjADsMJJ)ht&yKV~jDntf-P9<d0KcIFXyroX
zdhB|v-PC>8en|mhnwT2vRLPzfMmI{Z;J53%LS%Equ6a|NW7#)N|276P*+!lSL@D&V
z8G_%3_p%moVu&5!e^9cb{D3g6I!ILko}5?i-AN9_7Wb`|(5-zHu>{6IwVqZ8#b8uh
zrpAfNq;a~EueaiFntp5GD1haq>w6y*x@@p|o}FmQ<>C><|5nn82La77E0dw;6xyp{
z7-)eUK0BNS96H|74c(VMsP+3aw*YDkB@blCraJusI+piJ$zG!7@JB;pg?!Qdxt+d|
z=vxI3j+F|Bkt=)3T2{C0)520NIqZM_mV&&cV<@71o(?G}STtyk*A3Vz?;barXR7Oi
znt|mHq(V|j?qef+jcmTT+>Ra&UT3!YU9U0c;EE}bk*}4_?k|3nhRs87$IPRVpOF&D
z;~(TQqSicJ=|!~CJ61998Ya8(Pkdv%TYZ?)yI>*`)!&{xK*JIhP%I$jd!lbi;%Vrb
zx6#C?geMYRLjE5_WxR_9Z@;ZHB*!dG2RsVz8hbe3e^Rz#YX)a=#0U9&Hf*zy^Axy{
zxx6IPzuGf}jTBSuMe%jDIN!U6Z?mE@8Q5z0zrotSw8-k`X3@kd47^ZK!xxbOiHU)f
z)QwV>Ynnj>j{UaZvKk!q(>tNoljOM#gp%_0>vK~vx02f2)*D3*O5#>`UE~m+m4Lhy
zz0wS-WA*oHb8QV!lPo^YknRNA_jiqnOONAkaW(MMUF&HZ;mO0~OSBIuSz(Je;_;XA
z5VqgD9&5Gm3S5G5Fz_f;yF!B$al-sjXdh~{T_p1wAen~EqK7Jlaq0xzr&+X!N|w|3
zgV$${p^!0%`RNPqv+~Cf_lYm9tf|yC8WyS&5o5c_g7)+P3^^f*Uf~Jq`~65Bk+_RU
zezBtou<S>dWU;+sygv@l2|0+A@~XTy1wLnPxt=WXVGawHED=#>aDabvi+kl^_nBU+
z8Sa`HdL^_=J9<GWcdK4$_BeuL5)}uWdw6DnFv0~=8?<4hB5siJUgY|2GP<6LG;}}{
zk0>PA)}<XcMD!whA^7bWx4Zvve48e$^LC&2|K&QS2^k=BPc9(}q`6;$xfvPdzu1x>
zd-)t}dyqH>S~Q?)t@b-GsTx9eGBN3Xc|rP!Q0xer1+-y3M)E$2o`)udFc}A_>g84J
zAumEEHnt;u;My);o-qf&69Qe$yoPaa%Q*cau67Yf@{F;~gzyYg;Vw9kWUwl2XQBbZ
zT+<frpwP`)TtI_R;%mj4?)cF~QZ}-sQGkYeK$-Jjt<2I`IL|`GPaRQ@rDvg1D$CAU
z{uFd=UUTZgx`!%vL)SGSRp8Bc7872W^m2GvLyz>^7bo>41wsFO$=^m*8m=LyTJ#yD
zM0VW1%Bb)IJ(~N`C@<o`mUBZ$f)tKi)*~H68?xUI$C_cO6Q<k<poB9uZUxVc7{R#2
zJ-W5C?7SF1qwXIHe8x6+s^83as)tG3Ii9yFGL(3gB^Z)UPcGsopW!9`gB}FKbO`84
z^7zW-g5tu<p=zt)zewZyCIbmR-ocq5GVYIQpEe3NY{sBf?6>#2Qd`?N7P=x1#zYyx
zX!>~-(PM%{*S<&mAShMQYWdtfd9#}6W_<eY+D2SS$r95#60~Z_XZ>CO6DbtQD=WCn
z;kL^!5*zMslrkI9OkfJ{O{<gM*z5e~Yg8iIl8cii+!H<v0%Z#7(0!>M@5vh3xHyBA
zAN2eciyrp%wzh^W@$k$&Ke|B*Ry{~Wf6r###nNS2o0AM{rp0C3ko#i;T*|K*02`IB
zza<a@p+V2=ov%ae>*BXV6Ma?^3D|a8fGF5AHv?P4;-hOEf!4X!Q@xH=>i#vBgoq8j
z7#_7Gs?-QG7O^^~tW9%AhE|e(2x_r=Y+Uey`ZOtZiH<lkx1+>}<4|4%@o^%p1;x#L
z)rU003_U2mS%XwR=9k#8EMvkoU6o<%#T-H@P8^Tq>m^Zkj!?<pgkt3eCy`ONqbqFh
z2cbnNR*!!3vdfAwB=jJNPqJJ)9>0k?7@aDih>UjqynxXJ;;?lU7(@X`$Hm3B>@M_@
z@Y(NgTI`W(rwYAIaMD1FeE@Q$nOFHWk}QMfl>;DM3pqWsS7%!uGHAX35e;!8vg{KC
zOAU{*)VNp=myul0lB7Wu*awY=kxF7VjTl*xJ^XNX5Ujx>72FSW=9{XyZFd;g6c-ER
zanF%l@L}fnewnL9t3pIK;uPwMwaT#h`#)3HIr(^yj+T#gqUo}UvoW`5LLbCVejyog
z4wJ2}Z$XOjMIXcfVz#G)&^ypMM7e($J{)VbZ$w;<Ch_lEL%SxR;d?Ke5+TEG%b`={
z2GOOWwiId_oR6wo5c|lcbYFa4Y!@-EHFT$0Xgku7&k7IHvbZ7~eGIM`%fsgyrSw!W
zJzQLBF^byfL%Z@sU0ErHPu}@x(UT}ic#gl#^;_@+RFq<Qur%K;K)XyAeoX}}e?*#9
zVhXU)ZW7qjE8qFtw|^RQnlWy?PB{}Hmk(i}S;Lb^&j{xvvEoWa4d6=#aTOLmFFoen
z=U4`J{Ro3iqc8t?vGx6tb~^22#f40~ViI?=+c6(RY_SaKxxxq*4uzo7@{d@bJ_LJN
z(e4Mvm&DFS$<3{HDVC1B=qN!4d)yR1SU(rcHSNL5h%<w?>C#68y~Wp8*m^3%$NJ?^
zTznWJg?`w+hgbxaUKN;%%rw|^t*@FB^pH!QvEG>TFLna~{3Z>2{l0d{0dI|<yvN6|
zo(%sIB|BVG05%+k?ilf}fm2@82DIC1Ih%pMWU+~LvZY<zR-cZExv>oV2$DHRPTJVY
z7>d94v()hE)(#NM-NNMZe9QjN3xIo-*c8UYab9b%cfI4c2h6UDF64>ty`w+IJXu6C
z{c_CPKqo0h_lpd7{6|0>jxTLmsbjKC<hA(r!q9gWBEHZ`kD$cutm|dji*5H?>=XW)
zo&9de#zQiCJ~V(^=?vts0vlnLI}d|DKd*QiHobUOAG>R+q7?j`tJ5iD8PSCfmL&`J
zH)MG39Uhee^DRqTt6Uqwv|`9Ij<8#h{Eu66$xGSyqExRvjGSV{Pkdh=s$5=eFN|2J
z9A5SlYspO;KMG?_Kl(;+El8BlUBx+)Ebh3o=QrVD++*NA=N?C9aY#QLYItaF3#%8d
zkBaVr!G%R@(;J?QN5Q?bJm!G^oT;odU%&)kY<(k_AK-UZ&e5@`?t9vRVc3CDi%=~8
zLTf6A%%9{9B53k5eW)~2XUH}G$_prG)&VHcnMligK8<g?P9?Ea#~ze76TgC{JYphk
z?}hI5y#t^x#K@GnN>4|WB(6!hPY--EXkIHQf%ktRLgI={8tOtL+5ZqFL7q<`BUR=?
zz4oF_F4uU7*pA*7#r2RvWjeoWqZKX>9KZ(=GqhkGdiSy{ziJoylijred^2OPuj+p1
z)(iFNzjO80Jv98}sw7J`kLKBHiLIaIx_DUMNn%VwJ-@NG^JC=-SAh7Oupd<l5dlhx
zr0^GgB-p1B1Kg1HM*K&FAi}Exh7^+8;n6-s_von&rfp=yE=N}A_RVAs%P0+FR)K7o
z-0s<r>Mk!&G}q6{pRWQ>dz~V=mR9nFHDfa?ubjBSotoP?Ty*O0_GA~rg!qjpKjQE?
z`)eWPU)Z-+q}Gyl*L$dk8Oy7KthFiN0h+qC1&Lj_Je1}jPz<h=N+&MWV*@4UBPs~Q
z8dT7LW4C&E?<V#+!kl^%jVTb}^~u5Azd$rn!X*%=qUw_qjGtj^)fff<sFF4YR;i2;
zo+*wc5lK)pS{1HtKk$D~s`M>Jy^KDLiH(!Zm0Orx0K6sJDkEaE_?(Z?e<Tx$H-CJ;
zj7`DljLH?CuuW?g{@c&_rSf*Dg$tOUIR3{9!`X<BUCP~=7=TqahA8NyA5}H#GgWuF
zKf~@BJ4c!MSxBw0=|y>EEWOTgOx}=dT(aE`+)8wB>M+KG*4$ys?AbbBZEQ!0g}=6K
zZfo}QF0dl0-Tp&p$zCc^3^y&i^~uHT7w6_vjW{yB!a5$X*#yR`2ntr!`>ouaq2h<a
z0(j3H;+K0bq#@wL>NHM)35Opic~Ay~BXIHx9<fy=(+k7fOJsm@l{DdvT<I^;M@S&V
zq{A2k@=iB?^;jS!4umR)%Je>S>W~H6(YGii_W8u9q_#Fb6o}WE_=$W}b{u;nDy6Ft
zr>i~25+GS}NDktETW#dV1G$Lu`~oF{?Kg3#{<we@4&w6{D9QMC?(1H7o?XxcIG!n>
z$LP{G19vy}h%qE^;18tQiSAi70(m5qd95$`@d(vUyl>=NaDg~DGVq1v+)4Oh+axx4
z|I6+dLqnIFk8IlWxnZD}w_Q4a)4mnAfrR=aF3&YwxY)c`X!?vs;R`cB<mDnDVKPk7
z#)5JW<MF6w>8RnO2>_4>vS2k_N-5J$1A(8*+{3jj0bjd#n1wV`?;j75e)@4?s4scJ
zllrZ!bez`Gl%c>y4EeokeWqIb02PPfczz*KDnh#T^aj07jz;hgP6nX>>b~|rmwRc)
z!?n5n)qZg{<Oc2Q$%u9}JC5}l4z*B%%vxqX(WLkpv|I&kMp1W-ev%|0ZTdTJgMU7i
zGjCb&fWF+jKspF$!u|DSn;==@!n~phZMrpz$x<j`7%L8Y`EQGWOvyGc?RYaWk}v~~
zB%2p0H%m780=1ih&|8Re07X*fu=p0zIOJeh@H(qv(76@tV?OP!9T!OLR2lKfsca;@
zdL|NY|B_^(-L^R@$=@(m*;Uxyf7Gw+j8;-#gk&KGT*Qxa^<#>EzD-9;dXAHU4P&+Y
zhNSbR8&ZGF`Gzmz_u2E`6y1&gc--w$6=+L`fhVuFB<QoI-|kK6ODX49avczKSzGwJ
z%7Fx<UMbHSjc&+7DWL2szpIST#`;q>{8iiu5r>#xxb8CGV*QQ$we{;94v--HLI>Ee
zZ~i2u7k(Zzxu@4`X^bKJqw;@<ddsjT1HOHDgn)D?Eh!)%-O?qfNJ__Oq)R$RgER`#
zA>Cb41Eiz{kr+KfsR0uPto}UD^WOJ+9J{_>U(VzD?aYUN)&gPAtnkOfcqP?<4QtX*
z?D}w*#p6oeyj9EcCbw{$HeIu6{6w8w>>R>cE-=d<tf852(8!?g>A_%NTqEN1j3Kq%
zP~b$zeAF$G5pk))$j@MKBh8R>o279YIJ$FRaAGBIJ5n!;mR<<eosAT>1$JNFSoFkG
z0Ku`%Zhm5Y9fqUXMJD$B_C5AA6z&&AHRrM?-9=kK@V6vIlZ~@pBNI%t-C|CI7e>@0
zMq>|;<8EUEw%z6FNshXylcp<^6oGd5$V9t=54DR%(L3nB;k83S{~@3Dg=*{p`q`y#
z?1fytrEcmT7*Px3-U0>Pf{z3+R^mS@BR{@6tlEK5cRje;*HUT*O2k!8msqk){0rh?
zTj3Ta?!$?5e85c2v%S9XlqFY?yI%kHtl>$YkGt#nt#W@D6<+jpDD1wm|6=DJB6F+^
z-`l2wSfVh@1f!=*X!3zzXZS#5j8QEYYQ3zD<pV7fW9C;E>~2`x1@3E`<e-{H3BYj9
zIp4QDAFm&VX%?@t?ae;}E8R}oJDj(jd7C1Kznru?d~Yk}rySl1YVGH9E$;epiHRbL
zVxf9wBxY_-M(ZY6&GRL(L+Z`r%W!HAgOh7HLS^rUituXuPFr@?hW6NXd-IeV+rRp;
zcWZvCN0WkdzFgQnkv17gXS``!Rf0B}I8u)@dKIfJaYyTWpM2BLt^Bh;(j`4NcRf#*
zkDj$MQL#XF-q#i9x4&xc*8qDtCVdXR5w2)hG?kYa*iD@_MQYE|53eeSk8mPtMS5;5
zs@%I(jY|@!_ANH72$DXJxz$tes1c57)zdtDTRDBNa!M3;rFT&#aCD}GoT{3w2*n5}
zV!y0axWy~{9za#On_IE8-tvdXqAniOStn(1NTA8{Eq=6daupec5~>pvkndVdFsVRm
z|MqiAHwJ<oX2oEN&&~S;Gj;7U9z%2vpD1o#+3KMpWe&c)A5+J#>6#2oFKnHX%A}8!
zi;;y)!823z!}k+Ms=HF(U5WIw{#zf`EV}HO4^eAkUGgnS2wQlkTLBVwf_fe(V!BVU
zs#96#`wxOuLoaD}r+0=sj{Q+M?udA}Cg9f?1GQgY3ceCUWv+LIy47Fiwa_FQcvY>|
zN_Xn}=sB~o(X8WtxGra$59U`ogQP^9w5&Qj;4f^+zcBjpiA(;@!1*I8Udi#yW!)&Z
zqh+BJBfE_XFsy5u4j!*&r?-a(%;t7QK?x;%Bw+=rU)Eq$oD{eWl)a^ngECn8b4d?{
zZ1Jh{*l}8~eZpmt>QHpF-&5x}oub?a1#q}Abb$GBF3Zxf-Bnj_;O#Dfpc@6nby=%E
zKXiO{_QugXB+{xPa+in=NyLhUw-pUdeI)5L*NV8l`)h1pnG)%x^yStB*Xx{$TH`<e
zPJli?(%HRpHEwWU*HpROavbunbNZx>Bx&{^XUGvFk-0ehuXNKC8`E$m2+rcS1Tkb1
zc%~QUEZ|`Fzt!g<Apc>Gr>9B6^a-jVDVzLleyBfxIq8iCf*kNbfstpWVT`fqrn<e6
zQXZ$fz&;=3jsoUN)aq^}YKwfsU$rUme*xtU#js-Y!cfK(FS-X7$F2(fg+{4S0v^{v
z-g{*Hn0=T6Q_0;%lXRRe8M{wdx_~)<w#-^^7$%Taxk>ZwZRm+b28rq7%nMNZbC>w^
ztR^X4=o@$DV5WZ?z;yZ7MCu7D%NGd>CLbm6p7Ijw9JzQ-J#4JQkV9!TW0@Rd^`G|f
zaXudaEgSs-`Ts~ym}Akv7dMy6x5`Zj4xctU_gpaF#)o;@{BR30UQRZZ4){fhj(2rD
zHZSNrl?kRf5wB)9>Dp@AY=0Q<z2F;P`Sy>fsJU9X6{FZlo1ZnZi{5GHy79@DTnVcT
zLrS$BQKgDdl)VL?`8A}|5_{cq-|`!bWB(3Y3~V`Zmo>jPS9$-qtNYupP+ug^G0a!P
z?X20EKYGZs*EC*n*g3^aDE6;4=j)wiCsKZlshW^9Dz5Bc%08W6HNhC{%AEN1eQ^0)
zHz5wRf@1b5k3PRO-@mAAsUc}jIX>G$H-y4*S*uJftN7-s_Wm{HFj1~rTDwF_H&Alz
zW!3w9L)i)&-M;i7eacCdnsxdf&rxuCb{suz4fc?i+o1pR=7CV<cMP9wF({*HwHuuc
zv%~|0UUah$0=sRXh-n6<Vh{%Yq+5KLHEcQk@+g*+SJeKmdZ(B&ZV9B>+q`lTWIH>?
zf8CXV^WL-RD_Mz!tw&ssWz(w^!+QvIzUw-cluW@*^FP77wzuFM^PHfzW8pKn#%b^w
z@b~x6z+JJF;2!dp%cB3LcLe`Eu=IJJ1c$$kn3CwQAf+o|0wOclefzws#m^xR*J~0t
zM+mo7T%tVBn`+Pm&A@r3E{732JZxF|v)!>|Ny@j?-ycnKOn7fXBV_>={e#b^H^Wyk
zjD`5Hjg4`H=c#R77^tOq@^ybyc2S;2KXV9-^BfaWK`^<{)q~hhQv|Gg_+QLznH@tE
zjT60tOOt)u44L!mOAvMWqPW&0$k#uV5_WX8al<tD%5e!~-=+SqhL}7F?6#kXc&5H9
zI+CkO$MCvYsgLSp^a=Y`u{7K4Hb}>sb$o>0=V>#WA$oij^UE|4(udjIN@iVp6g12P
z=)^shxgV8uxHGv!QkOO0t8^}o&gPqk9IG?0)(ZImXc!&-Mz}B0xE+r3ekDKlBUii|
z;V?p&cgqV?vo$ZB?IHx5p#Sq^)G(%MzY87gag-X}h`tUpAf@2rzaqXzChiy?Y1j01
zY*#~T&sKPK8Wuh2C^dB8A7$K?pTj?TbFv^kJ7btccrlE*vr$EP14K)nLRQ7yGsqg}
zW2V1TueHj`Q{_eAgeNfD=Z8ugEOuH7(H-l3#2sJP<tkh@?+ceHDe!UR;72vK<n|t}
zVDKC+eAJ&d4j44X@TO5Fx_EZXLPFIZ1ik!v)4N6^if8ll>C)AQCxxJSVW!QtssjMP
zr;Bz0=y=^J<td_)-G(O454uiZf9Rmtr@aT#v<LmItT}M1rb_=jmJ$-zU8mMFp1O9G
zeL(gRwT-8`<}sbFGwtS?h)3xnul+M$nK8=J1u*u5x#Ay7cdN~E;XeAwC%#%8rpw&Y
zfT8fIS<llI^!jGSmV|rJf-ceR>wS4E0S=Yi&e3i+vAPO`#-T67;~7VBdNsQ7Zz(f%
z9;*sRPvf0m3Kd?(sGb|B<|I~2DsUaw7w^}UZDKvkYl7x+?=X7{B5gB*UN6}!sfa?h
z{Qh>pT^-^JePiSDqm|C|PH_>xIeM>w=BSq?U*{8Ozi%&XC6-h&+5QaSO?6v$KM$G~
zlTI$mHz=UE?N1sa@v;2V+OZw>E-?v4Qta^AlydQxxTLXQI02x!7x@CP{cVoERN*d>
z{jOL?i%bFDLNDlhvz13nSJ5vL_eiP1-s*AI8c)N9d)t1%wcEhhK4TAnTPL3`u8g3h
z4V3Um$#3T=WD4h(GpC=D!`hlLIjY}v`Ydn|`Z%Dh#nmi#EoO!@i0oZZ=x4YF>HHp5
zu)v29Eu5mxfN-|J(|bF)Mnao-*Ojc(sxjyI#TjML#pC2J2OLVKsMbwHiSM6E-pMR_
z?70EYd7Wan$OoztABV1w{HzyttY4}ek8WBt4nzOFrSo?94No~bP3!G-54}QzSH=~e
z=WpPHeSPV4&fcAuRaK&pV-X5;Z_GYx+C^HuoIhf4{tu(zPB%vMhfl3%q|b_T@BuLe
z;;ZXE*?iUIGEnNKi3D?WGZ66(m2a~hbn88O*Ds}H4WBjqr;ZKBN}NnNP^vvYnc|?w
ze+|!z*u5INzxuxG6b{1Q7~FK34g`iWD#Cw7VQ-NBw-~<gJ!6FIO$uTetFMUZ-@l*Q
z48PqVTbCueEFiRn%xg9^d~!=sdC?jfgdO^N_P|KxC0kzv>V~e0G}jQfyzZFUYxEhU
zv@3{0c}&5vUmQ^MQ0eQ2*-+I`&=FhBD$hrh__^%noDA=}6X##kpB!ml1s@~qx#&y6
zl}Lil6=LP${%PbaHU3(jyGFK{mPKs#jXn-t>NKvo)DdlaC8%wN^*md}IHfXNf_prO
zOlv)KXRte!r=aWZhZaZ5xT(3ksz>J2hXXHPT1V$a@i9bmXY_KQM=z%gv5Wrx@Gqmu
zTO6F;gYedl85Pj8kQ}Bqq?>_xx`>Zyl~MZ^Z&dzlae=V8(=JsdWg|MJo7>;bsmv8Z
zqrhMCZhx}X*gpAJ+Jd##;!Ani%ID#>AL_$4U&imoJig3M*-C1X#?VH&W#djVa!dRc
zmt<J+2-@HkCDtFGDppvYYzCWx2UFtOd}PFHB=dK}__CvO)SdbN#Xf|_orv!Xu(_7s
zDU%jz3%TH~#hg#hd__ucW7Hk(NVlwB055P<ymRP@>DHU*qdLXkgtBqDyd@7v{(I->
zcp-(kI3!!$LIVY?luqHa3|=msi9QFydTkdQo8so4t|3pJ0K9*xlPEp3BjD?G|9~SN
zP&0hkZCE~5ZpRp9CUg1b<0JCbAslMJtKJDYu4Htv4@=hT13z6`=@dbzd!}=xW&+9q
zn!241SlMbpOxml%1+EzZZOZwlTUMUmW>}mZVIOp$5P5ZBU!^@jrq7bXpn3#B^D=(9
zSE`kqd{RVlup`4_hv-0<J02*vf<@zI6%Lmc^PFSA5p(D|z1+|6?e1M3Gu{fhNyA#m
zjM<49zAK$bgB*tlSje2LBl=$E?MOA9Z6D5R*#(sz_HR$iE%p+fd*eTv+xcoH`9E&U
z`yXiM6Nl<J4zGUZef}qop8N^GVXqS<c(VW89K4Ksj<6uHZuI}|j=jD(h0H-6rsle4
zYKx1P;TLBO@s6HC8=O<mA^mPwVUi)dgsAXB<Mq_iiD^aU{%_n$NfGM>0AdQofN&O_
zC&gpSTRqYIvJ^&_iq&#p;YhHbKkbzd#SzRh^B-*|v@8@gr~U3?ctKkG?Nm{55F?f*
zN2x7;d-o-cVqZYQjJ9ZVs$<?xSL4|>tT=b<r!|#z9v}?m>kCNl<q0v5Kz;2ewldws
z<yO*ssgHA*;cHYqUI%yZC3V|0i<o*^?Y{rbIGc!yIry<Nm1!W8Vj_P&Wx~|Uzwqg{
zFs6UKMckaF6E3X|wNrai&zUNJjUVyg1g~-!{<3m@&X>he3d6^%usHDmeqe>V9Ud<;
zxmCG?Fa@g^Gt0K0Pn!F?{HG=TcRFv!!C{N(S6kr+Pn$fi27E@{vWvrGYH#D%6YSde
zVS4Pns*juBp-7U2&evZp;-Yp!Xe*Aro$(yU1H;<f8=F>%<tw;%_~#070QI`6+e=p}
zD<><)3hnUCZ*~Zg(e`|P`<%*Z2Fl@q;N?_sug$pe`02~6Z($sP<gFRbja!hmCB%Wk
zcyH_dJcUX*#~vuA^H0HGD^zmTGazT|q^0aPt6oHcdXUS&d%4#%l&+i2{fttK@|*l|
zkJ{dyI7bXh0Amt58m?W|LmD1tTrfA$JeWJ2O--Sx!Di<e>Mft9^UkBP9fzuN$r3lE
zZ2aGDbhlb1Dx12}N{9wyF8$McSD*GX{bSKxkLyW8nHB^RpYorM$Iz$e1+_eHq_^t`
z`QG|E903`}0YK*%I3=-dv>z`sjFa_{6~2`ZZx+X&STOD^y^%a?8uhp8bPqqBX<+j2
zuX#`opk)<#`L)1Qw$sBxgKltg)z=@DDuN&1z)cO%(58nr4+ilF#QF%X_uIjnTUw$?
zl~gjnc}>%Up;Oo824nKw{!2NDGGi#kHd`<c@=F=pR5I`T`7c6#VfYJouZRcPQ%CRz
z-R^S3e(hiB@v8OsJ`0MQR<M0AHY+nSyV6!JWfdVdcB&YB@l?TNW1Q)}?)xmR16G!K
zNXLD?Tf;D&nw?R3Yx^U+NL>jA3F&_jdU_P?j=7rH87-r^`StO~dj9i{UcY9Hw`p9k
z%-CX*UEqhtIK|$U_~kA_SNc5XxF%3{JTTOaAgg-ZHGVqAH`Swfc5aIbPxyafdI{^E
zkY)2@D4qU#%>3Fkj$AZzM-FeEKv!7^&pJ7>Gj@Fx|2_B}QXl=Mp$Yx%{&u4i)+!^?
z%R4x{4Ze}89xpz!Jkqtt#YgqOwxVQJT8ln0RZ)ZW6SoUg4g#Mkb<1l37D%14xxM_o
znSstc-CnMQ%Uz7!y<20Bb;U--4VleydP6n2FSO^<Y&_~MgnwKUz%Nd(PP<rt=QD&K
z8YLgF`Il5pXT7=k>nmnbR7(qKfb2T>?0J;jCRJ6ncr_(wd&0=-1MUG0Sw@^^i7oaw
zGRkk<AFB_QY96f!%UU_SQe!K5V)5Q2Mv1*l#JD{+iM^UY5>}a}Y1EgFf~n?*$)9VV
z6LB{0Pgd59m$a3O%%lG((7qSTEUu*-NV}S&DeqAr{WYgTE?Qw6jDDNpN);3O0lj`r
zdL50<a(OX7^6!`3X(20*ip{{80LIW>X;n}v=L$AbuzGgLL&mUvkDa|jOVvsmnz3)s
zTGzQpxw>-rw@0XIPZ=Jb{3`wEvDVJZH)Wk2zS#S=*CT+Fzg;n_#jIs{)nHPvTRrZp
zw)qW9Zi0k<uu5v3d0aa)=Tg?g4VYUF_pYFO&i+I77iF(=PX7eQ*l=CLhmMNBo?7vF
z!8Gb0x6?_8kg$W&-<pNrs{G<{bnVYZ%$d&tH>XL|<`>#;jtMn89*q!66S=L}ipC5|
z6q&#Be)+~VE2QcXs43{6ta>)vI07+FaZp81bYGZUHg;Q4AgF+ON!j!r>)<<r-+`M2
zPV<}>5Y1u4s$gsnp8mevWHPue=q>tcuh=Xv_AgcGGh(a7ySvR5&ijhYJAi?8`xs`?
zaVi4AROl2C08Y>!NB`bHYDC<SOHbH*16{{kL%*~#QrlOjnLh=W`$Z5$M7!Tz@k1Qk
zT9~?W(PHk;5b8|45%Ob;)xti-k8sIqUzpjSiPEwADydN^9nT+(W4_LG5gb~^TPZg*
ziU5$IbKRRKCA0gm<>zh6S>^bs*<V<cN#a?WHS5oZE|j5-kH#pxgbCQoDhGRUuX}(p
zPMdbA;V?msc~QmWEW19hVDe(=?{|x&`GoFntnGSDBPO|(U3wIy%TE0WOwUj^9t5@^
zcy`ixa{*So1lD9-EPRHCyIl)f?dj)MDQyR&9F9?)pDfU?y^4d?pHb9fhpm`cA@0z<
z4@<_6#<+-lMgPwQnmh@H+G+)SIV9e`--X4uzaaC)dp}e^5s`?ZV-d<4Fus@~tDloC
zG11|s$pGj!a~x0U(*yeDcFo@1JW;k+f~l9F2M6J$^L5tm`*$Ywt_U?JWu{Y%y*!+$
z`NcZ=&^EHTi(OKt50-LgK+t=>L}w->!_&%hWndRHyrmfA#;l|LKm-sx$&VmUx{3+3
z<WLe=fv_k;Iw?<@6CxDid~UW)`^;7}F#KcG*(pM4v2($A9ia|HizTI5CXEhkpNJ@b
z2FlR0eLstzGhQ~7XQyU$!ZK6pxf!UItehqOJFsgcMAt(;{kC3rCA+akN%M0Yo(Hk)
zEZiPQqW2gepsa+jG*e_I%xXH3W5aU5scwRMX{aiDx-(!W)fu0KeR$57Y2{W{jo9IC
zC&?xdt-465Mf_2W!}MEey`8<$zjYZwjMgd%L&ixuQkkjT3eem8GHmpcz}mjOA-}8P
zNu47+yGvOM|AXz%vpP;gv2kVoiGu(odzxUUXHf+?qU@?Q8{ZHAqETUA#2`g1UBLl=
zBw8WQYky04Jj(D&Iz+npv9Nm=RE-`_IyP$N6#ZPd*+oGd(2viu={461=Y=+%p#LP%
z9K+=9SD%bN1noZRAhlu|3igDp`6}==5r{DN6&*C)ZGfar(sG2BwhpewGTk%LoO{4a
zt*ZBiEIdrDMVz2z-<<L;<@l?o(=S)$@5v1&14*^Tfv3n$%No=EEiRh#v;q50RCH6W
z<Mzincz?*A0D%0WFpU#Ww=k3zpu2yy+_-je)hc4z;j+X*B)Z`E&Eh59c@+2{RmASq
z=j7ZRyv-7yfXq_SnuA}nkXFQyd(C<7o2Tw9CNYBo$MM|AupZm-Vuq=$rdgS(tX)f9
z+ww<EuM%iwaf8=>kK(hvg$rpjKj;iQoFUp@Il3BFVk!0`?Fw<snvm!6-EKpBgQp2&
z3#9e#?JEAya<o~}2ml-|lqWH=vTspiv3}yga&!b)S-}iK@*UQ&FE~R%!Qi)>>(=0$
zF2;WD`a%YU5I0=d8F^}m*M4*KQZza3#Pm;wGR=L1GLir9ti&pQdBm_b5o!TozYun{
z)*umigguxj23a<rBIq(eyTJ;f@?fO92;sMA^NVhP*LOTWx)Ta289LG;=Vxb#g!NuU
zctEi_<!G>O-)!^qsG?UyX%rET^dPqTub3R{r1Sh3wjPp@26}M+j=KMz((A(R(`|Aq
z>EA!}$MBqe#|>>>I#E=t9(bz8Ufu{sjBk9;0GYJ=n3b^h5a`d7h4I~Nmg`b17Cou`
z0N3<ThoD$8V%#!YFZf+c&U%^JIoo3$MX1OLxrO3+Iq-`s#RN99-KucEw5IT3BF6G5
zA;KxKW@W)u&ShV&cSY@Nn>NMpw+<$;L~S8q*WVTa5j0}{n^Ie?31P;|4MVIz5kKGI
zEUeR9F5k`RvC$U2J1R|(o;o8Kv1W-xLu9Yc{^JFh5l}@HhpjjI-hQcxH~!dHN8>1l
zUU8lZ2>+Nt7(hW*<Mz`ZSdr54W>Ub>B7#^$CEj$RR!!9Xzz`;R`r`4}JH^*~SQKy4
z{52*XCI6*Z9=XN&W;N%kK*kDhDfEpM3$BcQ16*dZ&bC++q%3LA!n60y+Zqfs>$;xH
zP}WydxcxE3#Mb`s2=;;#k9g;GENjl4{Nhf|!NMGpjXZ(_RTHz=m-d(RGIeVMMw|M1
zEj$WAm6eqI5(1RsZ+`=$lc#XzGh;UxB5_st=)HKXNBWVU?Hekr<Zyp0UJmF@Fp^o0
z8FhmC*4s^guA6BN>jWFMOoabUt=R*D#@}J2IGJF@s0C-h-~~beGpe|AgL0zj(wk*2
zMbDYa@)-jFByT|EfC=(@Ju+D%4F~WW7D^`o5W7h7-*9GvrG&7u;iR83sp=+HrUNv8
z)Mvf+dkXl%5|)W@g6?l3>D;tBl3yA0MXzN#Q`_otdZF{s3PXqCmufat+teZc?w)l4
zN+-_;;(TevB8Y&)16JH6l(Z{iIF7ykE&Rj>)e>IT7KATJPb~4J0bb`a#Gxav*ilh+
zrSvvQ;_uN;nd+X>drKRV;po$)c;pQthRlEV)-62auCi)#SW3bgH~FjACUX0Ie=>=x
zNyIVoNhm?1*FfuvOF<fa#Q%uF3|U|fuGnv0{(F_vXj^~}830S|tSMWP5E(YXYP#d;
z@m+}HE?q|VN7Q3$6Rfa#*du&lCdVlRQ7BO>MH|6Ob^}V>O7@8XSSUDz$X7T<q`Pf#
zxT)U{!5M}Mfxy|~hS=Y00JfKA%b=qX*8Thbbr7CPvXlv~23gUXS$Vt?F5HYcG;9%~
zdzWOiVzFBHJw{G^yLERGh_|Kj^i9%Q#QaPCj&90)aZ;o*i`P<<o3oYbV6d&Cm#NNn
z#de1E#WiYDmu71GY<YyZo#JW`#~^|GpFWWwlxHBl1U1VfaNI(R-CXT5+|eqe`jb!s
zJ6Xjn!lyB#5+!c%#8codpkK9k(J&Y<IgR17L4!!Ib?jW3m3`oK!rt#krpX|vsq5Uq
zSJV($P<vhStHukJ$G$QNQ`_9nwUzRNX3FsFNO$$Kma7D?^XV9z4&2vQZ22v5U&rFz
zPFlTrhNxHT!sTt3g47B&LsY>@JG>m0H9N5<6=>@&WH+@-u~TaKk5_1|PIS2!-NzdV
z&)B|9sH*zuZ$ELNUxJT|J48tin#O!B)`eg(^Sw%NqvZTu2O@f68xp|V7O*`syOV8F
zJZ<o}wa_K)E-7d;z+C+_BC#q@ewXj}G0MAE2R(#{qi5HZfF=L%`N@myed8H3=?qZn
zwVS|Rr?SIc1?(m;Cu8>!IqaT>n*wUuFrO70xYmh199vH37`C^EhTK-CAedM!RB=l>
z?8Y5_vh)CY8Y=t67FVVT1Y1Y(jR_20?h~-7kCaFau&|$Xj<#`|xFU8~U(B;4>lCwm
zkMRSCcy9r|Cs$Q4TQxoViQEavM9%hVoa#|>Z|2*CjMPqd9y&!1D;|581dctWs0-YH
z&4-GU{PI?P_Rz}Po_X&@4((V)m=*NdWrB2NM&FxkRc8vm47}3rbDX|5B-Q=Z0*4B{
z%ttsiW&|<gl&J+<{8KZ<7mONU_+>_lJUR$)sS5Gr$=(A0=VMWJX<S=ATlygQZG;ce
z`^clwS=1E#t{<n>TDA!sz{HN1{ugHb`7<WY3c${RV2lLZ9wsQN{ZrXgFE0?oQ3tOw
z|Dc;Yj-13hK82eRy)LHxN=SL{(PZ6$0-vPLlPoh_hfLt($2TdkX?c;OzlXOdvcykv
zv*%ZQp?v}5cdp6gojv>5Y}}qg((<FS>$U7?OV4UfN9k{GhqOOQor8SF|6G7FN-V8e
z%he0yoV2e~8vJWm1c>a6*NslFqIQGJZ$^)Q%!T{c(_b`DCm6Kv5_1xYo)^NpjJ#OC
z_X)mhUl^E5HbQb=kMebay*4R-=qS2}k3o}ju_2!M0+;<v)5P?`k|h7QZ~D8mMng3x
zLQjjFiI3DAS?f)af_kp=C7|Z?H61xmzC+KxATofL?V1~AwcMF9Gu8{K<iTeI<6)5@
z00i0(rGnoI#fcm|qO`r)ad6#I{%l3m<`^>bo=M>c^Meq3xwBKo;oH+?K}guxWQHzj
z%_=ed*P0jZ1Cxfz`JnF_?N2!*IUbnDcLFoc%T=<m*tN)3qO}Wsv}=*O_vXV}KI(FD
zflFFSyH0hiAA3+rp0H0vqL1N3KKtBD6K3pPm|}t*+a7FVxQ|w4lN6;(`7_z@6qBkg
zyqtiT>w<^g#ME~;`0E7CM7t-wnh$Zzu);I9uN)zIU-Yr&^Y53*B{}3G!NFLq-y?w$
z$d5UJo7}~$<&{l20a}vq1j{rc+1&Pt(8TB+diJIchfNXF>A;5{o$)bC5LR3O`ImdV
ztn{pIVqa~{_QG3&x_|8eQG*<zJ%!Y3fZeluV!Jj=s1;;3M!NnDA0Qn(e6X_;)9+~z
zETEMPL9KjqT`Q3E{qXFQwfnFe+Nn5j_$~|XhVJQCtnDDC2F!(H+KgDWke|Of?m$1`
z#!b+uHZHB;c%MM0M`ni}{*)a@^s9c<bG0pMdN(w2GLuu8+rm^!*h!TB`$thR{E^v?
z`&xAQej&75@p6qaT;tJr$v6SXsSB;kimf0%Usy~k%8E;p`@kBrV)<wvCq8Oxq2NWc
z`!S26ry%)9Vj@7tvwv1z3eR*cEHqO7BPaELVowV0!@EyYIJo7g&m*r0i!(7>-%b&|
zyFl0jN@lto75Vv8dmm!T(CsIb9Ijb}YnOO&LtC5a-GH#GYo^hl8?*#YS6BW`$Fis5
zuW@XZw&U%%B6x7{IQZ?LFt@E{$bG5YG1+>CyRVXrnAKmfmF`<!jXSl4(Zl%0Ipf*K
zzP*1A>k}*zMFh(06U@wl2mz}QK38RX33lsS5fblbQ6-H{kyZX5Cy9(rY&uog>DWEk
z<t%{-uUnQKn6@fr%($6ExXeUS35Sqjp?H(&5*~d4(YtJFcb1Geq|bYV-l85J{JmHA
z*<*RWxL06rL{!(R4yGNSVGnwzHLz9LZu_7l%)pz)dj`f#Rj4t~IARxPos$~0o8wNB
z3--^)Tw5d#c&$@TrpdWj;ekSbrr&vwT^F---NE#t&ky1c6rcFo763b;k2tYnhh!g#
zGOT)>Cc^w67>_e^S!Svw&Ln>fysDv4vCkIq&2P04DoIxJRO_I^4J4)(u5vR;y?jAR
zo}os`32JqL5T2GcfaE4BYf2A~b9wdw+o|d<GLMXvXw<c$U6T5Zqulw=)NeSl{FO59
zMoR3tK4zRqi_k0(kH~qZLNVIkq?`O$ZU6i#@~@kSnEs8MLVo<M)n*Rp5*8#ILnY;F
z@SZByx#(Epx!=Guz(&*D(pLn<+?EHw72y&ZWeF1sQcQ4u*B4(YPMuu0Q!D9tH13St
zvNi%U*lkx{3n0|0T;Z)VIkyYdzv$*x(B1b@hAk8Hy@$00YJ~OO>IgF7j^2IEpJ2oI
zSumo7LDs|N%-i*Mk2-xa`MtA2+Gm=+<CRMUwTsuRplQfG`Khq6!>$iFRfSk7gx~_2
zyRG58?0C)n&akA_SI1q4v#Y3eNJ*ClzecHGB9+5KfRm;~qChJ!Q!ze<3U}t3_3NXL
z?4k*L+%<dKR879KAmf<zp^2cDd(p@WxeK_1AHo8PKMxikKJ*tQ@lYxNz~P{rH0-3w
z1I3mFm77yUtj!nnRTxMH_GBAD%W?C~!r=FJeAM4`ufYL%n#aWkLtq{YZz|#cx=RzY
zM;zXlEg#Iqe(Za2(*=bRbq@n!U30OesdVAL;(16?-&`5zMF27mCbC%RAGi@yb$8vq
z*NAY#Q&ZHUi6|-3HJ^_Uyzjsh*N)(~%HF^B#{=ITHN@ZDfd5VFfwxcWo}sQ3IcO+7
zF1wp7*BJ8{D24B^HKV~GP+4H%r?hT5I@uRp4UjXVZDXNwmRdIlbAx)%hQUxH!6(&B
z<adVrkNxbciy9v+e<ssOi1fG^2s4w}PmV<O`7%+bBQ=l-b|dsIUFq-3O%Sh|+_}=k
zN`CWh1;@J9KR0<S8C{=1!KaehI(np#s*gsm#m9LGHRL_B_FkZzWZ^CC=vUno$`CR{
zHd%c$u13h*aY{|_7Uf8dG%K6ets(bXFNaOPLZk*jKPeGvqia{uuAKt4HC7VbC1Vlv
zT(zmdUbh4%zm2ty0BcPkbjDRUCmux(C6*o&-m<U%hY+-Ex%RcF2n~NIL1ggpN@qkG
z(E4f$sj|u_aTX)&8cDV)CKY+t|4L1O`(vfoG|gXq>ZpE4ZwvpG9l}$NPuD_~T8#Ke
z{w|KUgdN6`FvqTYou3E@Bh@%Kpf@}uYx1UyI=Ha11gb3<BsYK;TjAv&sc?<=<YyF}
zKa{gCgv=w+gc<nqoPYYA+<JEys0O-5NSph61{6$!Jy~0UOWLQ)A>8)l5-^FheXiEx
z^U$&UXH8ZQ=EblKxk5EM=L)N{-ucE2m8)yAOqH*^4aP_dEZTzteTNQCf6e89Fqh_?
zSg?W%2;63&f*)_Z^}dvh%{}Z18K9BsD+${@{R`;6I_m{Qh^(Qmtk(Km;Z40=9|a@Y
z7gd;){AP1{8_4om-v-n0ezMNqlniOU!~DsjXu$T2g@zueo&DakCHz3PI%VVud*ehz
z9~7g&il>N2DVw=agzzcX1bqvLcz64n%|1psMp_gG;5_iDWQ$vZoG#srTprG$vA&-8
zc3g|k)yoREje#N3dsvb=7WIB+bD~3qubhc^T1VKTw0;hf#J`kc0<WUyvhcPk8t)S=
z80#%i79HJ~UaG}xb&I8$(ovY)f5DQMNy<&iL`%npAn?LNOqJKd`;WFB)2S$E0}{2O
z`baNzY$N7X6_NA)LW_SWN0`)yU|~e8v1pk=fndBCQWbh-uN)~A!JahTrjeWb_~!h>
zse9yNx4$`f!YOb<AN{Hm6^|%lx|Hq0ZeP+`ZMm&?y`AcyR!}zZ0$29k6udTDjTgBx
zYz5#124p<vFQ795lCz*LlN;Q|`J$JZ&Hhet|D_TqOfD+!ij6zn>vZ$CvA_{|&?>ML
zZF1N;knn0gsC9@K=cto!snpFhIjZRkGB)BJFnxVKTfe1efLeidH4pUf;lciaALfd}
z-2rno<>?LX$EDTDaldR!%TIw-DSBApf)~=wHq)nc<Xt<zW8v|%J16%|7Ofy?oCxU~
zu{c2ZzwT9{s+`zUS<hUu%K+?1`uO5p9&wQSTzYib3=2F?dQOmOmrZFnpN{NRf9u<>
zxL;$+*{3@63|^S(ePf+_Ej?f^CmP+z{-?vp+W|GPDh7xU3fsM}QG(FD<w{vOLkPBU
z^S#;9Hpp1`18O=H0DVwEbRzV5F$3xi370t<)Or^~Sq!gyQZj!C_P-tpok?aYW1mTi
z`+;8*^Rg<(PZ!$vx1T8HLqqX6-iScc(;%kmu*5elO++Do(Ns|iq&BIzGZYSGnRh@c
z{=;g&l?uA(UC}?`etq-_opVb{jv3>;Qn66)g>6lPQfMFD=e|hswAFAC?**%BV7U>U
z-B$}@Xt|mt<fMu4Skj8e=zikLZ5=guXNRHr&GNEaF6_G9GM+KU)pKAh82jr;-WGZ$
zWz|t1)p#oo*oGCf-J*aEz;<L_SMcTwkt9*(FOx=dg>}Vh#MZ$-#c|>TKy-j+UyW5!
zx(3$UuLbnx%7lZvP}VkgQTr}n*PY~MSI}dz_=$fHU3Ct<@SXr*VRPanzG4St7vIEm
zR=)G#66&jEs4hs@7ydM2#d(>3U4HF18^rzC)bYh5jG$0NBL~h4!~4qkYZ<y*X`Bio
zrAK1g5p4*q9~}61D*0d?k&|E_+;sN9gI}bdo`088G|@0I8sod607}ONb9GMC44J8L
zQJ2DOa5Kf&uaY(ALj5D6(PzQ*EuNOa=i^{nnb7^|WOcPnshDG%Un{1Gg#Z8NC@<>~
z!;6V@>LmV3&@$G8D<DJ=qwri%iS}$R${i?_MnRdRb1(9CQHBi?#Y}|!kxYa<#|>5G
zq<FalLy0}rR>FoqVcq$**mm<OG2ifi-by_jMS`fJb|Fy^rO~z59Po3dY4TtuSNKvv
zD}(B9yElnvT~3jv;3*MUXKC8o*fzVHyV%B#2$4B^ngold0xq-qx^{k07%XSxS($#9
zURcv40V7RLl=p_8Mf|oBsBv>qo<47`w$w))Hw*Kxy<i}|$1@P~OFt!>Cy7G0(^P~a
zk*nz=Q))<-Fk5{1z--ob15;f4CiFEZ>gZTU(YIE2j;ZTy779%<%?%2HP~H-_A5D_q
z9D_VG`=+FZM|+K6l5J?N(9~mD|A90Or5MmPl^w0%mSn!ebOe<M`vD8p&xR{^tMMaQ
z-)YLpOB!04hS8H4{kH-1-%$(lt~xTdPum&b3iQwfSt+}`Scy9cPd()M=`!gp4_Po?
z{R+Qs26fh3Jx#0RM3w{&gkM%&<lKd>mXgw|kAJD#%1%I=v_GI<xEb_PmtAQhyB@Ra
zPDfofg_Nh$hEp&!$a9#@s21P5O|klLU(}B}r@S@1)%MZpj%a_FFLR*z3oa7dwMV(S
z6|^IZx}GySnIu;)Ni)A5e_UHaEr-1SMEdn0f2^+Wi5ZtO7p`Z*2jMoJC9sd6bk#|?
ze?Q(w_VR{mhRX)8z7CnjH_)*O^%-v1RHLu!jNAT0PJ=I#lB|;(LI8l^fYVx5ps`D^
zDym5ox!+VAyt#(BzIU@@&+2Q=rHX?Y1(VLf^_D6GB`vQB9M*H1a=8{+wDep*>n6AN
zb|yU<ye7+s^**-yHMux;Qz-8QOZ!fiKOT}_X;__NoQ@-ntlwv;uCzbNtWW3+U;QP}
zY8CV8P2{sQKs8-{J8fp>N~m`nD#RU2aN-M*Ze3@Nng|->erUw)@Vg3@BA*!wiSU{w
z)toZm%4ZjfJ=C7zm=Q&75*}Y)U`M=P-W=a9keT8CyMukdYRDPE{}WwRypM&ok_rLW
zM_ls}*c-gTY#nX8T(~y;x&QA}4;xueU=29$|F|)|+haXZdqMSG{Sng`eD2U=z%#sX
zbC7|PW7nh3Kd-H89BJ$mc*9byHvxiEsLt7`8*c>ib{TY{1{tTKx%3Kn#Qt`%NKLVH
z>stdnI3_YWd@dUAFQ>MZ<?ss(i+bwzTsW-amgIo2gl9+_6Y^}`H{&}=)!}x*cN%^J
z3={@Aq==#Dv|CNF5c#u~en&DqSfoBYmCbSn#b(hg27VYQDn2?sM1P+azN^#!#M-FA
zZ87q!w1}2Oh4OZH2lFl>ZF>}ZGXML&#Nvg)mAASTSq@<T_zTEGm0v@I>Jm#a!ehp6
zMAZK&g7;kW%_c=;@wtgLL?;TFo@X{OSU0LJ!%J-{Xm{prRXk~`s?m8ax4>9DPU#-k
z$j_}JInA(^sH5Kxev^Ma8Cf{LZEXq3m@Kj)?z2&M%RFt;&Obzz*TkE*+cYl5g(zWd
z6Bk9yjU-=k7;S)7k9faP_kO=-y+qv=I1|4nPE6Z{r%bS|+aq63vJXN(cb?MizUDmR
z5w+YBeQ900o}VoTYGiz=vM{;EORPuK`r5p0EG>PU9)0|H$fK8C0+c~**iW3p@stDn
zVFvu<C2eMwHeC4#BeQ>d4OQQ8l;!R>ik#q>5TTk4xatb06TPmrvC=e^^fO!LWS(VJ
z=O(3>S^d2<YdR2sLhVw>+K=r?2(zfYkh2HBW-2duk_TSO8)Kz%?2=9^cdiY5iQ89P
z-+a2I>MRt6H-<irn-dasH%NjuSvc&y5jC+4h9q~KauvN`>JsTU7S$jqcbJfMX4q|F
zZn^CZhBU%K;*EF!3`D1>{TS>`258P+_2bU&skR?n8+-8rzz6mHgG-;*0AA>JlL5T2
zwMGcYS}(6&G~VyB2uLb^eJ*;xV8b1RVy6aZYbh7PI~=Bq4-D)J-67sFKEIFO?$4LU
zIke)peM7ulhu2;eC{Nsk{O(9A8}k)-X+x#!BMghs##JL1@q19{A);bEYJV-c8B8CW
zd73&RuG_ZFY3=j#V)(D=&2DC)5D>rIpFt5k?K;FxQ#K0sab)9f7qUez+}BOfNu96Z
zy$1~{VCJ4fx@n+=&Qzs!FOWBN0fsLb|HCdHa4E-i9wUe+IPn+S_b;K5Rx=Uz=amco
zr|s_bHxXRzr^j|s_e=H9@BdH~*J1Qlt}Apnb>E?p3v@gsD~&rDgh^M3@om=70Z%L7
z??Lvwh=Jp{>@I9y=-+$65_ood19A9{e3-AduJ_68dfusQF)^S2<F*=mrSQSPYe?PB
z$L+uZN-)Y_qPfyV-LS(=ILKe5BT3V1!Y^OB-uo6fb<NK7`PaAv$Ds&&ne~2C0YK%M
zf!ehSuFdoPx_rr{zL^Xn)$rLw&TXDdx~A4JwHIbGwG%+|fCk`$X6<R}+_qs@f7nc;
z@5-*<jT|@W$AStzjQ7|&c8sagtdVkP;2&qD<G=_rKx2%wrq6c~bpn*@o$xoqGW%=E
zqXapTmfiN7cxgq74iKps|38uq183b&K)dD64`_cpV;1(u-YIZmHYAqZUr*GpEDJg`
z#yza#sb~Nqg)Ff1*YHT~`kcz`9~z`G4IAHF5ayl7C(QbIs*u&A)DGyMO!l_gJ75OO
zYHEV#h_GyN`%}gM8QhMo*!P|=ziuG3K!)>?Vu`-mfNka86Hct(oWdv5bh__Pd6Ww0
z4B@rR?;kt%%|JT$3x$86gBADkN_I(Pp$WgepKNLvlae2kao+wIYq~grMWe@TA_b9(
z_D@lv^QZ&(g*VPkKvLclL+}VQuF7kH-vOG$Zx9`CXU5&+?``6t=91yp2(2m6w~BW&
zAqQk)Gnyb?pe9L`HXeXHAp&-q-vIK2%5Qc#1I+g4Rt|oRpAN`2;Q~t0AToeSaWE;s
z>x!3Y4jee;cFyzVzSod-)}ypQrUXg$6!6=VQI+b+^GgD)UR|N8Ke*KbeE>j$t7Ng6
zEbR{bu5t!(Z;S^|G7H8tNzNh8aw~0WJy>f_P%?uuNPs%|+X2OWZ1GPcU?tU>f)o{p
zZUt-`4Y!P0CsD%ZP_ozY1ORCN&UH$#)V{;!=SDJq%E>EQN%hskKiqrlxcrxoCuLRa
zJ&TvTsqpEYp5yWoQXTrco|fv@wolA3$DqvK8bQRV7p9l9;oi=Ej3(>r#B8Tw@0&Bn
z+W0>pXrc>p8AdBf7-+NSvdaghlm9Eyh}^>{b*e96`jC>S^=@CYxH6T^BVyf1wJ8b%
z9?Dj$McB05A}ND@j{Lp5j-QZW3|wcs3b*M3dt4wCDpmGzGHyfHBe#2Y$e%1D+9adz
zz_*((tSl7=qbJjfbA}eDXu?hz2c8HsuRnJ;D6F5b;&V?+IBg=(g*Yv@qHS>#yE{O!
zLQDg9?s5|XC89%F7^|xB%9!xDS1#79&RQ<-i|uDx(4`*|B|~B0EnX{N!dp!XDl$YY
zB|d)8Ko0J$R8ozueCz0B9ON&4vXYra23O>aA(tBFk~IuU&~|yDP!zJzno-Clrj#LW
z_03H))O8tk*(4*%-d-W31F`ODv$+zO%V}s@6;}&AN)RD{iIHsk)!xt@`4{Ss(#L&H
zW4r`=&r$ZZ)1Nx4vQSa&N^qcDmS#gPs8pXNig|gCQ!cfbI#QLtxWi%w_Vv=fMR_ia
z)zV)>4c<8zL<80ww+37$+)CX~vu8ZwIzKqve&(}G@ax)$2Bag}P)ij<lw#$Fd;)QO
z_}93P!L~cD;X%YkTC~S?LHRxIBs;`1&)yWauFgEG1Z6b;$yaQRoJuo8@RP*^uztRg
z7nlX@lm3+U{Q3K+KS(ax1`NBDFRPEacZCU=_fuk|fzN|=CXQ83&GWWHR8Or#Mo)Oj
z6V}+r)Y_bRCT44Pk#DDQ#3@b@s@mn5r$QnNThr~U!}lh@-g0ppXlI?~XNP=N)<!W`
zE6Bm;oc_zq(JD<F%*|@*?N(LY*t7&=+&TXh_<jj2TPQz3slqWY;^?!aC+4?=>It~r
z9)6vc{bgHpd#nCSI-q*l-Riy^bylhKRu$RpX{AP}WMQg8qSRofL12A3;_Qkg_hqbS
zXfrwtmGSw#i<?{}nf8S8Ij*Rh-X3S#mwS_B69$%NE%eG}VvRw6rusp8m(q{b0xEb}
zkg;}1Uq^vB-a2#}im&NQ375WI)o4~%!ndPzp^DH?ZrpVPC}~0Zr`X#yj?_?N*69zu
z@__P5fL2OVTQLiG51a6ZvG<|J!)vNxC~&CH7$;VRRyo_3J$IRYMMjSm&!z?N-zJ2u
z=1<>9bEg<0Y%tg&cFJbra0&@LcHa`H97kG3(PI?~H?(p=#4~BZWg`3pzZ?QKlw}3%
zt{W>k?9Q_Nc5cnmo|%I#R0WUwyOfiO2j2Zk>C;Q><qrlqU$<7m9~1)vTckvvyy8{f
zBi-IJd>yrKR-r9TwV=v%t=MtrN-SAEOfxN4`R7Y-XOW;BKB_|GCuz|cu1;Qe&AxBl
zD^)3a*M=vFOTeb(p_jm(!fg8Lfgsb@#PrfRb;|buTqlA@O+|=eN}ESs{4PcBpvYXb
zeAoN?A<yaw{}<*&XVJZJ5%*SZ-U2%<?5ns0NTt70=h}}v5qy;cV<R%E=c)PbdbMb7
zcf@C!Y2`twAAe=q-tsApKM|$8tn(;O=2U`oyn{3SGau;iL%kFq{*M>HL>l`~Pt~N0
zqQ0=rUYL&=XfXXE@bfCUc`_@dL6#-M=h{`bLTJ$rd#Uj>NTI7x^(Dch?<LTDwQGCL
za`qg(P(Z5rjF1S~sX~fI#={q+t!susApM_HFyQzr*;2Y>dVI?<jgjQC;ZC>gME)<g
zLHopq0eCZMX>t>X{4kmz#Tc!UgGl-`qmrwieSc)rPLrZ<8*Y6=AVt`9d?;CSSa1Bb
z94!M-`7ta+2p{q_s42)ha}6*G(v_v)eL0U3s$>6$D_tNdFlFKh<%Nc`KKx{Q-Q$-}
z=ly-{R56DuKR!jtkoCt)OOY7Xx4ARm)+uJPI-{+or(@3xkOiX;q_M3cM2a$YREnU$
zo=Ez^r2x0hG{5m_E@3k@;u2<vf2@b5o_%A)Wvh2Z;<vk{$_$)mX{#tkyS+iPY#ABr
z&JgvJ`z_VI&m-yJW53fR&XHWVTLwy5*438hwMkwtc9fp!na|(vNy{+N>78(5dGm`A
z3N6}d4*<I>KGX>|=>a&wogtjWwf(oon-(ON3PSx)HDw+IgJt`ZB8p(I9%(!iVX_$u
zF1u$2HZ8x`xgrB~ZzpRQ<l034=KngLq2$0kwf^$vsfcJ%RUqKM8CK8tKMHPdNPYy!
zaq6o5dlDwCi^>l!Z-U-_eHf9Ls4@TbPSIpye-$SQjL<kEN%u_{jNd0Gnw(FFRP7^H
zAya-YJ7*e>109<0)%i{dnD?{@TuH+|zc2oj7(|ED(@-j851sPAa^K2A_5yDqz=jVo
zZf{gumt#$FVz-y{G^)rlen-(CeosUkXKjfiG<M*&vK0>VWJG@tC5_NpagkD)oarUs
zt3T>0zIWiV9SPx@P|PDzMSs4wLE7GvXeu#fk>a5aj)xC#3lwc8fYU==TN3!^O#@Cx
zwy$Ja3<M*_rwMLWedDH=5o99h9ofn=bSxd2GwPNw?%$jsqsq5$4w`5Z9r|6WT^brV
z(sKhBfI6UfJA?W*K64NlHsh9aiBNpuan&%r3Upuyh4!d3Bg+DsOu@QK6r$ALj*y?p
z>=mka?4y>h;Pjm_OV_G}96<q-r|n=Q?^L6|-BRP{(PS@8NaN-5v5afZ8=CELR%B3I
zUYZYOnu-;x?a2KO0wGL2CEHM+OQke%Pr?B($7#2E<B;D(;1YVPF*PQy%B{XBX!|Mi
ztV18#?F6pbv}_7LL_dJ~7SM>2_4(~>l<0wyHXj`cv46Vt-WPA?^mJQ=1KDip;47!<
zJ`){83jCD{^ln8txt}$TVuLSryF2BFZW}e@ZR9fWTolSTe!>kE&pc@&Ss6<k-B!Dc
zWp9Z9+mlFIZy&cGeL8&8<lTWI2j=l&A=ocLQ_Q0L))lUr1EEK7K*XG!>so}AP!6p%
zvcGA~TkH#o-AQ*1B7Xx&06<|HXQnRa@(A8z0#5s|2Ml(Un43fY$myNS844T#3@n<2
zKe*TGOhTJx+TUe`rU57+7bxn1)X;=l5}h_yL^`1R`ui3kfVsw3sW@ynRqR%F?Q|qP
zV4Sjd^dR8Rhu~c=wQpe128d3SZMF@q?2Vi7fu~Q(XjWKbfp+8_OoPg-u#|h0Dw3q-
zSZ<2(kJe`@jX%$3?+L_7?LVY5$6-#&u_Go-i{4;$rKX0>QCg{OCfNG-It0ru6ckK5
z$67pzocX2@C+}a0p~JI9oc4{4d&s6m{DG3ooge*QVxjDExyMAbh_q&oLCHX?1OvNd
zqIIOwLSuF>SPn?rRtY=Nls`2hx588Z#IH=N`h^!eMu4J%+nOz(hj|0;+8=yfQ>;J4
z(Ur6pHDS4iVrlLxqt`#yd`(sfFC0QF)0r#$0ub%g2I(6(4H2WtzN~-x{#XkBMu+@D
zgVy%)^qeA(pU2o9KgQzGLoLatw|Y%frrL&C0Y#^5!?&Q-kT#PyttT(s3KK$L`CC&6
zQQ)kpX|9As-C^!VDX+v^Js17Xr{<IbuI)Itss#2-={t&$W<|F@-y>|M+HEwOaUjj=
z<^)=~<(g6UzAh5^Cmz2G470M}H=@9Spl;O|#cMO!%~dH+ho5=UBRI8IV;^=$KKkw!
zbs*MUi*3)MMW2;BJN&H=V=uSHJ>O)%)HiDTMEqfr5awN^8xq92uXZ0aDnvYoBs_un
zaMeCbJWS1xt>AP}y<0=n4DvK!b%4UG3*H=%tpOWN6K`l90Ais$*NSz_xzbsWpPDug
zFV^(TRtV)P5PC(Bmn&*7#mFG47&c6aihbOR8TnL*Z9tGZ|LXG?sSjtlQf3`6+Ji6p
zqcIV*xUmA?c(sVq>wPypot8s$EV;=(6breRr$iNZ*hO+@P6!HF*cCw+$J^Xp<q~I}
zlLLNSo_*e0X$#M=igSwHi}=|`8PEOsP7OZbpl4F0?v(NKaicWtu<<p;KD=W?KZY5f
z72P&~ml&S0uj&L7skP1cf{)fcVcv5Kn-QBS)@GjxhBU1c*`isiqDscVXLQg!`Q=C-
zi5RycUv{a4&7#sM`^a>bDCE1VZIs{-(f>!(cRx}c|NkB<L`GK09$7~yWn_<tY{ju>
zNC?@SV@70VXJtfLIrch-%8p}|?Z`Uzv5#|(bKLX!e!lnKzu^7L>-~JbUXM8;v9Iv>
z^06qJfz!fB<I5vOowzZ3uuBbLu4pus+t1+3ErZ5UqIGVwZ|7f&Y+cRhlYy$ohwnx!
zeQH1ixT0r4%+(fkGR5JroFs0eK4KUchEX{qHx|Ae-Sx>|wTEy}0~m0}Ua9mK-jGQC
zNa%w&vPdx%3r?>x@^Rk0xPP;Uu~OGk?vdZz8yJL=vDcY7MYc=Vf%MjOh20vG3hh<3
z9|b>SupGn<`VLd(qT!;FR^nt;$*dk4VK)S%h9s-B*Ld^#)E85E(%%l9N9p3+TYfNF
zQ@TvW`@vNSI6~^_+SQr9#09-%*Hapvzr<*E({=}?ofM=Fh$c*XVvDb;d|HcIsgcdg
zL*IISGF$bQ@E<i3xf(Dr@F{i4d-jzd=?j|aYVEdoUt-2H4dNJ8LP6wrLT3}@hi%Gd
z-zl}UuAZI4EinHkDAq8a9LQ)rV~W9Ry*k9Bt3`kPkBu~jD~iw7d48cdBFBec+%39k
zRL_@vEiYV_MYoSlcPiD&M0M%r3W%t2b`;*8xVb3nFvI)FG{2t<i|Nn&U9?POqK~fl
zI{t4gv1QawE|uEHk5(4OTe6Kz5v*tQ(E%#z1D;5zZe$gU9Pc9T2rzYqy;pe#3Xgr(
zcemJ43GSy^li^&F1qD8G8_Zr_C%h$Hq^Uo2v`b-Ric=$5b{G0NRf2vWh!mIHy%d<6
zq^?1yp+p3PP4vfj9<U$387snw2T;8n#G%99M%wbbgsbUCqrhA5wS5+*)Whs_pDVJA
zmFCc;pxz6fBn-lXrY#cg8b-WZ(;ar+fqxvR>E=XdSQ#{LMiH`#-T|$`)?o?B^?QZi
zf>PM&wQtjx&+)+B!ZU7fK*!tYuSGU^*%seX;|zqsq`Hbr*>S_<Pe=YZi{JmIZo0S|
zPedv!MZ2Ul!9P;3$s$uVNU>UXRD$F2ecT?blY8~Rc&M@T?6pzn$;p|<tz0|)I1?-B
z3*}o>lR1!hokt9ejtJGQqILg~n@*^+5L@_n_d;)1f}&J5F$#N-1X=^^9n5y&s4<eB
zD@K<JBLczu=x_?a_i335gqali5~pS7E6ywB{zhl{9wxQtz*2JQII*w_>dsr_TQAs!
zjZ-ptf;xOD%70y4^jE6bPph?RL8PJTI@+)f=vpFq)wg+`dFIE*lo}uIF|8Wuu;9WI
zgi9QvOa2+dA{L5w`X_rh5;L#A)n|N7X&jn8m-#xq)IGW^8_=lB712KG^LxTvT9q)b
z9FFRJRP-DM#>%VctowL*pVHkYM3$;;n<$gUc4CIwv9JA+WyXDnR{;RhVQy*KV02A)
zDq(^%!K?v01rp<1WLxq>uHTH|5zI$A_Mr(32{b-8MI7jdbjErq^&|ZmhXLDnw<)#q
zuO0xY)-a6tkm72_Al2j4E4N~6PIL}!WcXhDCyIX|)5P!vf3i#Okq7CjNiWp&8#Uy)
z+$V3W4CB5_m3u5?2_h<6jo3M{G#i~COLN>P%A^cNvyAMI3L3j@#A;ZCriUXhI9<$9
zh%zpk)2!Y0IEP<;S#^;+Cv8sIcQL*3@EmT99goeZ*_RG21D^g~awA-Tu&#e&F#gm|
z0kZVMtKfz3L!*RWJg@V8aFeu?6EQ~3n<4Xc;lc)gYc}_*5O}Ky1aE6rt;8(z=T{Zq
zs#xM~bDM4(NnD^_uv_vnD>IwzSQ_!1Z=0d$au0?uhC+rn&gZ`!;n|h$AqnCN0=S-S
zg$jBxv&45vK#AjS54)is(r*o)GJJMTf3>>i1xd?sXX=<eU;TD?aX;dyXw-QA@_9<6
z?Y7h}+A!H0l8;Ttr^1=;CGI&lTkq`!q$CCSz1jRy4l#Dca8T>KzVzOr=Jz>E#bL?(
zMGNelmi-}&8kd8K%P}~8<PEC4!5Je>){fPKLojYkY`7E~;dqtaXc}1F*;|u#ndSvm
zAKw4o=WE@_AM?}ZvO-gTW={B1kaa3HF0C(mj+z%eFK>A1TBq0o*RhQfxvL*(KPx>l
zdjdHkX|_Az-X@BWB3c(><p$$8*<o-pi2m`648S-@PKMIhVcC=pBb7&0BKLzyZAqc2
zPov>tT03&~>Nvg9<>>0CBNb~Lw%kmEQDa;OmZTk}r61|GUq+>7LNR1V{%xX4{nV&|
z(!qtDe4BCbrmd;MxC`a4r^fzS_gjY)-pIyz|1?l@rtO*HAk)@U%OknU6fU>)t1q@O
zf2F^M6XIB&T$sHll)A?W_N`(+Y3n2XSJ&BV%TzseILK<LN+ka?q={+S*I!Z@AK&E%
z8<~ypa<vQCI9$gfbNdhxTg>mNwNzW|&-&yE?h4}}VyZriH%q8&y=jAAct;);P#K{}
zW1IY$s2JPDWrgd1K5e8-6tS0&Lf9!iI#}QhTduU`><0oy*+TG|$~;tQDHn<<?6q=E
zT$e(9%Fmf75jLI?hBTJ!fO`h9m(4e4d^fnpLkK-ZGtRWwkxDrum-g8a(4av<i&wS%
zufI9x7b&wniwD2kJOsk(_|xlI$^J+s`e*w&idZt(Od+dBA_|=r<P^v*z^7b$RT8Zq
zjE`$Q*;}t%X3{Pz$BG8$gLKQ!=ssXnHC^GzcCVv>2!W4sxz-Ma+0oP_t;T#lf7T6X
zR^06J>KXbFUGdym7_$4VZ$Wjw{3fS0anshC3^8dT9*A#eJK2_+h=#I-cVqt2PcDYK
zV-x~EV5eLaN>;6`8p4;@$_%|<gSgH~%uqJB>i1oUdrCID>(BKt_G63N9a4n}DX>eX
zU#mdtzJESutqw)N;A9<N6INy4P>;lc_cX3DH!EWuDHbltMN(WEAQjI1$q^A-E^6b!
zQ(a8aGfHm?=~Fbfd30}^t6<pvin#vVr#RCmIx>6Og|XdejOf89Br(ZrHRjvnz-i|H
z3i!@3{t{C<M*d6DMD;q|wlRzlMJ14y-~bx;-&HXMW@<}~X=-~4<=>0{pb>JCELTek
z-d-N2ez)G)MQpZybinpRuFr<e%>Rp-OkXgbx$%%d@PXX`<T|$W<_S9=p%2^7(RPi2
z&*qd44(74caoZ6}%cKa*!{Vh*4@B7VCZf3Ar67s~^x6&n8}<Z57K@|Hw0^8*Jv=$8
zf8f?L4paKMkG3>dbN$lj*Z-C*5&N=8{I;Fm`^BUmlHc=nptItR-4e#xm+6X$HY4PJ
zR&)!%mD0uGdEs{ET@9JnBtTp?L-p4Z$!)zE-J`ybaC)oJc3!Pifz^Dq9F5P!QEIj4
zFL*m><T=C3v;mR7Zyug;7_FReq=qA+y(2+NE<=soR0}5D+819Z#Ksw)JwUiT{TC^P
zE$aWYNE-STl3v)K=C2|XqbOAtky;efSaADJbPp@JpAQt%N5Ep8VYn(5Kznd06<J7x
z^8(p~k<+%(Ebneei>afXy?<#SDm$_s-js+?g)6;V9{`d$1Bt;lco9oVC)Mw7OJ9J0
z8&>-Lq99)prBhCw%F4!lAFeki24lfSH$9ZgjeN$kHuge2f`Veui)%k(#Pl$1p4JlQ
z3Dcot8JpK3xq<>1+LI39exD@Q+R{B&qc#E6y5KE!zJc%V!e2HAOZn}Si^?>${v5G6
zfD3oHC!S&pckgu7pRJ*a9%%XgYn+;+Qa$P5zmoI5eKVM(DoG9z7^Qkbc4-d}pa8td
zPPmk-o5w<#x<_6|&hX9?M-3Y3yYwo0jqF?Cs@Pf*hm11f3}l_03^SdMT>e>%n^sWC
zDU{Q$5{qmPZ~&r^!3ZFoX+BG<Q+#%ek+K`hZbZw;{^LMmr<?)5Z^Qki7$VgZx>#S~
zl=!qZeEGI4_$rFO%h~^3W)7>!SfIVoC&-C3=5{@L?rim2SFqo^v&myY1IV&bc8)yk
z&pojzua9*1SBb*n*UnuVPPhQm)g`Gg4j`}7mBhlz3w!pse1s;Q-Qcp=SR{;Nk}++y
zcoWHr>I4dLGRk<C0NqD1$tPMp!g1&r3F-OsB?Ha%;VcuVz>-D9F1tcYDxuz;p-|E=
zesk%rq%&3HWrcH7GAUN=q5p9yH!lcozIU~pLgP@A%6rk6W6*BIUWji6db;`i+?+X&
zN^VU;x~W_iJ~U$r_d{-O|CmEl1NQ1Zl{zTMCOI5U&eZV0>b`R04KKMQnIBiHlEB@g
zp#2c=dUfE{)p|dj=uYfRph~JHz<){P7FpH#f5PwDNFk#vC}6(B*A{dl%o)dX?=4xm
z7M-^Tj`1si%CYc)OF`sh02#oVa(bO6Q#%>z&WdMX3$mY46VHF0{XW62gDVSY-;~?H
z`#My}$48HL>|rF-<t_?*<6Tztok*b<5yx8qfI;RD=Jm1RGe)(9!Wqcs>U~2=gSExu
zXJFz)sOitups|97scTJ#YwxeB9Sf*ILVq|xG>VI)AoOOyfBqNJUIENTH|jOxFR3+f
z0-F9;gG2p5W*H;3OJ);@xxU);cq&))&;(gKLRH(aM@zqFN`c;aK*W;oZZm4jzgQt~
zeE=zy#2#OH!B!LqUc6!sASWq((wNfH(!)=zLPkWsloBRqUr@^qE)e-+%jF&TdBxt-
z{tM%(TWxezr%AK2Ap_Byle#swiE`tGuKe=!Y$K#}pRzVE&NBK-du-MzxctrCq`2^k
zMq6K&<6T9^9{0xUua6ScPJ`mI-)AXk7C3cE{pP&SMgoC=d#S~v4^QP<>Dz+_i{QWZ
zqAc=qodTAb#waYZ%^t~B2c_$U3$r0YoA`p?QX+q@(t}1vHmk!?F#gXf_ABN&A^&bC
zAFe)qY_#{LyU62)R$vKGNK>Ge!$|5RWFIV%bgsnFlFSy71ftUKH?n9}M%4gcIHbG2
zhio{rJ1J<Nw`X3;*2pf$U7ms2426Hq(duQ$TvyX(F_*w{Yv3Lv*^+SF*530WjS|LB
zU)Ffi8#0$OpBtaekkJmtiH^Jj(m?6mmAu7#It2Z&izke~BMD!8;V@@378t-+@bfw9
z<g;VW5kXfc7Bkbu@uT=7EE=K%9a;KOx=013*$OMivBn7+_Bjr!%=;J8<iX~Ql<&BC
zEcc_|cuI7_NzO)4Lw7v&Cf;7fH^VKD%E5D{&OX<oj`0x8rK<c15fMjr_WDvnT*HKh
z&OeY4IiVKgbq5X)UIM<l!jYiYPK;T<V_Wd>ys!U;isHjJZLL28ZxpE8QCM1j$3kja
z@Hw{tU$F=D@@ipm$~8L0O#j&TV)@rjck^M`j`zDpu2FhcgzsK)KDPk(r28VTdPUvf
z2;&Y!Y1nnHn?rX1QotmKQXX}^vAO(xp)iFZc5Yj9o!^^7P_Hj8CyG=;wnD9<kqOo&
zRI8$_?r9O?rYwb~c*(Ss0{iC=69p71bx^yQrWI_!Ma0*^kge^`5~(p}$LMT|@&E3I
z+Y)R*sYo|E`Z+*z+E?<@OrgLJ_*5DtE6V<{RGLCp`jC^$+<hOeZKJ~_qwki_PWz}w
zG`oC+$PLYb(`U{z22lB51Z)InlI9|J9UUr2Pqp`a1T}AVE`v#*r+vUHLxlSwn+eKe
zO8ZN7X-$Q4KfRi**{|`7@F#i)1N7d7sk&<m2?-#bkYx;eZ7)Jc%)3~UtP*@?hXN>F
zLNcJXTAMfdy5P>Y{6H!P(r;YyZi#~Bw`eZN(-?(PwAXKLg~TNgX9ud7`p+*XA8#&P
zg~1&paB<ov8<5qShG(RVwCR!R{%c^;X<oaNx{tQMfMTV2@2B9Ya|;bY_)~MKVi-LM
z0-T-wg$$kw{|G<rzX9m6D3TLu7zKBo{Yo2?f&O-zMKHGR;o$fi&N5P_4<h10(<xG%
zyO!6%)aw}qHnQu!K|jaB@C=~ld`pt9T`vlP_3XD=UT$jJg34Ajy%hGEQ7%ojd3`*k
zwP>h)-1Y|7|6Zsf?IM>e$KbDj2Hq|2p}`t<ljY>OL{QwYdBVvIc6h9ks_Vu1>f{#8
zlPiAsAD@0E&?+(8B(qP(PbBtjtGkC~(U2sYM&K&LxZ3(+?#Z5ruZOJNTbp*al+n#$
zj>P-M4A-=o-R*{2(Hd@p3Of0W8o6#my=lT~1lHzD3k8O=MMeAlES`6z0xs&eWUR+l
zA{}q+J~|i{eu)orS2<dZ^)<P+PG9iY%S1u>3EZaJOZRH*;nXx7zl@%ZD8jP>R;X17
z(0$13f-DToa}H}*XQ}Kqc3c1IZ-6sa3bCNH&C|HHYw<eVhjBkldR1vJE4y2|@H_nK
zUws5Q;0^Wz_Vj8}dF|kG&H_dOn7x=nMSbf*5|R614Vojr7T^R`%f}smQ%--N($j1#
zTxWU7MTg$&@j#?p!qn*0O=fi8!+%zQ0LJV+RZ!kq^50@=BUZE}K!bmbUz+mto{(h<
z56eb_=ug+YFzEZ7YPxe!7QGz4UQuP?^YZnfqZxz1BCuM<Jm~qTKnYO#xPWRU<kB{-
z8JYD4^z^BI<Pe|G41Z+ZHxF8eZeg{hn4JXI&)CA>x?xn(+WHe|>?!G`0{2s_J;~hv
z2P$fLCQhqn^3WqonUkd3uPq4)b=fwWB&J@6XsaE61d^q*NPmUNtY4jm-UTs0QL0dT
z_{7=-Kp`y!hR@yQv!YL=tsxF>@|p=~f$caSm;3~my@pdO%N_kVtyM=sWPho)GyQ|+
zgEk`Ae>}IWY86DX-^{5{>yzZDar;E0;rrM0?Z+qZy1-Ev!VyYPgT@5#DiugwS5@vi
z%n_DCvjtLMP1&qe!4))=kJnkQC)wCai=K~BXbQFdy}^tOeJyeJ>Ss7XBN4Lczlr}S
zh22z4$A0jFtagFu!KiHPPENvXtUFtLghyQddh9FoasxpX(l>upA2C$Q>|OJT@rcm*
z&Si7EG5qWUR+{5WZA5D89=LS|XlM`IDHJ*9vO1C~lo|Q48V$OBjyDyOEq^G~>Zo?F
zhr|8-g`iYk#rpo8)AM@kt7T`|-aeMUL{|ahhTj|3=QTdzl8L*#Ow7X;@CA!UT0D%`
z!#V4Z#q90n^dsA(MQD|?6E5PkWEmfVow&MJXEqdPE{ZEF^x37gi!U&7`zEKqyKn5K
ze)UU)oG^ZRQdTABA2n6Bm3yYF+*%s8r?l;wAmQufrRoW9<4+CrHwgDtW6dfvy$Fg!
ztK&m#%~ww^Ue(K?{FF*JR^`i-roB_|-xR52m+P57TRh2*Ko9SR7kS))cs`3he6P>W
zZI<CL7?C)ZG8%v5C3SenAK+VF$nuzl`NWxWqm5)i^O_Xg!Q{aV@(KWuw!FWNbT!v-
zn*y>daW?{4xjXOu2}Cw|?=O9W0RX`taXb&~%PP(*;dB6b{SmJvg?F4ivT#9(k@^Lq
z2xb;#54bKL)_vFX-V?#Fr_KC_8S__R@q*>0()w;7#Svs=`vyS%6|JbgB}0_s&kQ5i
z-U7~kZzHM*ysL+a)M-T}^w5jnqq$M^RVQzHGmxG$K3}Xm_AaT-Iq^6_fUgKfBaib;
zQ;g$3q$C=QqSLJ@Q!HC<<^v;TMAW~32x*Fuf2!@R-n8|3J=?jJ^1sh<63oOX#`fw1
z;6vs$h37C{cYxhLm8;6DzSuRB5GgD2yZpG^#||H2nCd>Qe=N@BPILXB&xH6W5v0+b
z9H43`kitao^}vte?vh+9-v`ijjYmIULY84;7g3kd*Yc)B4s`jH&$<#IyKl7SelPm{
z3QYX72fMnDSrW4g7Z(lYp48yG*}QtRVV#-ur+J*sHtiTcRJvlq_`2D$;@E49PDYK!
zn1yVPaP1#B@#=+a6kgO8uwMqUn*L;zVE*W9Gh{!#c8?d$Sz3_SoD~+YGN2;)SQYby
zKgI%)i#WWl0Kmx6e}xAw%yyHwf&@Ym$BnOW-yjLkf$U83uWNj9Y7J}ks!h8-N{_U_
zg)}<t+~5tbg?eD@iP>9mskG$BUJ2nju)mTosg%mkuSn>kVW{%}Qx3{%<&$W_xVP$~
zkspX|Z<U!C$uQ#cCo41!bYicZhRFFSt7gi4q18%f9Vu&XvQ!{9jD0C(i`WLFVqwp;
zszvFtUUs>!E(F{(J>jA(;3x{n*k>8gMuo{nOZfQlh^cgxg}9=I22}dLF|sRKT+;Ib
zK+@*Fpd72|ee1uX`8S-W+Nm@N71&ZyTHml1C-~+ur3zzQWO;hsx!6L9PTu#3J55Sl
z*84B_C&E?U9z-oUV8%BW&a5}h&}O%dxpKT+*ZjSvh~7N&!CZUuUDZR+#lHMLK6hEz
z{5MMd2=eHW*YckS{hLV3R^O{J{7eloeKcvcVVjBTl(ZKVOUO}8pq_8(ehs(eK7UIV
za3=+<^8ub}shb;cl#mwD6*D~8N75<`cFX}xTG>R*9l?YDZelzeHCY3G{2BehO0g@p
zTh?)#oN5mDRWwpoL!6F2VVXMe$?dJqp^zsV-FG{enO*EyQwH^M&^|w-uTSz^nV;#p
zYK#J-HHqKpQ~J>?pvq}JC(U`PZygg!y4qwO=*nML&jX6I(@<hy-a!tstl3MQvnQsn
zSSdZ2$M35vlHKs+O?F*L5O(-QAadWDEd3m03{d+S;$CwzusGwfHJ#dj*aH&aDwxIj
zFo4{TK~1ZP>Z1@_uLW;AnSo8{IwfT}a70~<P8|&2a`tJ$#R&85kW;SQNe^I7!w6{|
z;GH3D6#{_0+%P^XhJJan&)xINFO0x*mTxt_tG#VXnb@gyT=+W|paGzs+zbp<VOErQ
z6~0)LdS&It-Xi?6Hp9q#P*uEvws(YDVqLgGCiTG!b}6x|qx$1{&yqp%H@7m8v4f}4
z<xzIruFsvgup#ELrMIFh)Usm>DFm!i^#&!c9W>@S0Cb6q3cu#c168RAhe;N{b*j3y
zAm{q^E{Lmr%nh=uoounSZk?nk#s~cmt~LWSc@nLHn*@ui_)jHYa$oiZ)xab$^iOMk
zFN6_T>hpb$X6iWagvP<f+4^;m<rV*Op_lt}{+EX$dmZ{Bd7~c+C~6&49qvWXs7RNe
zkhR}H<&0)bRJP!9;1>~q$X0L8NqE52yl>b-+T3Z1@)cwF*tIdSS4Teks0l&3V?zAm
z4bMGBw@N%El;x$_m_puly==}IR8H93*+-zG>6m$Duco)><u~w5`r+k@7VJ3~nR{uL
zV6!hjTE&>`EOz}{k>eI4>e?IY2^y&TR%Z>iy%PI#f2}d_M~#kZ<bqbs>&iAY@6Uoc
zY9_R%@|oIBcefJuPAE1K3&S1w!65I`AQWHfmlCdPe-%<(K>99AKB}cJuRbIrJ3U)!
zeGffnUc=rdTvX;ZL*>0`B6bwaD8v0``Z4vd^ZH6=9}3lPMRo^NyX|*qFa4IhO$P}g
zRmzksxA>|;HlG93Ah@JJLf~={@u!a!ggcxu^8FR5l<!ite+i4GMvwuB7yA9Tal00^
z#O2$&aQiNt@7q@^dy65RKHpMr4h)V-$=sp`2<;tf)-D1Xxiv3WU$9*YoZn>r&V)Ac
zx{L<l6JwRd)S*je+8R5VTM~GQh#f-5HDk+DT|F-N5NScs%2<D4BK9(;L}BcY-s%Ic
zLr+Y?A~fb>(whge=s=&K3=B&cq$Q=Pcz)B!!+fxM8L#mOv=^-Zdc8*yb$m7j_5MM2
z{utE69{E40l*b4fDyv9#ym73SClbKfNGUH65t`YP>$5nAYZrE)|Ey!EmToU+5+z+A
zrx3bwJ3Es0-sulXW*U^Ib)<q7dDCt25BoC*4O@}4GjTY@rO34AqI9&io`Xi!Bvr!v
zQg;II!Xkf&Zc3s01!}|%7<WHeWJ54@Fn9e4=7_CsB}?!ycTJ5(o6y41(yWnjFS7h+
z+~PE^3_kCCgnG_0I6<wTFwfe2{Zn{$O>|KiOaF{O`Wb7)mQ=AU?`n+5yYKZj4`Pk|
zwnGnc`8QGvTi{9SG+jm;a1Zh0h$0mT{}L+uE6w2Bw-PQo65(qP+-hNXC1OGXQ6F*V
zerS*URossqcn?weDV3%$VZnj@C#lem39)A2X`%h=t!}o-X!_Ztuct^HL&KBNNs=%S
z(H-HSyMO7SDO+#Xbk4P>s}`;2#pyFO2`Mq8jF?@m(QWYQfj4FWpE*Egd6PAhAM;;N
z_Q5otw6g`(+1!GDPq@REl)YIidmir?C-sZF(YPP`1$Qpr)8{nPlX2yj&Q6La!f0qG
zXutdVwTtsL?<PB&82LBVOrX}ulMc$&U%F;?G?0|FFB6ZyAB9A%oy@*2O45XdZUdfS
zju{)E-bGHbAApQ?KTE1;?Y>L9LgGJUn>^~cry%Yzw5i_hw$M2J8;QEOnxt|@@m7NS
zq-<rzdVcei+40!75_4NmU{}J8^h&d`$g5<=<pzA#-|vm}xrRMwLC?@@@{M7En_gXv
zjH3eb5$>;vy($CDNKC?75crHF52>J<k_o;%ys;bqyc6v8ck>?-G6a)K!g$#c&Q4yM
zR7oKK&yU^j0s4Qqd1rWwkQFj~=~(~$CfqtcCwG4m`tfy`JS%l15bXaKkLOGlFA;6v
z^BySCFkvMY)&NReWYyB8sQ|)_#H=Bh^?SIT_VHe{FoLDB+URe1yox`7hiYXbL0D~{
z$_Ue#A*t7~Xx0e%@W3HkWeqs3o3g?ZL3~peEl+03zIR{TX`PGDvlu9_dUMcelPji5
zDB-`q0*T0gWO6otzK<cCRDL;mtKSoR3I0x#i6zSXy%LJ+O_SiWB7fU;RBNPxqvT9k
zCl@Gopmk3pJ^<2?VM+a?k<$J9sr=exF|$Pu=``7KRV?GoNi7eCXu0)t?iRVR)=HnQ
zOnn#$H1Mw5;?ASaZhCQ0*CP8*C2;`W+I#uerdQJR-|JKr{)UV{b^b}mKi0YHD|g4I
zzu(Ekzl`?#ginY&47P!pZCqobvqI5fFM^A1QI(KoSleW8eN^qux?*|hia(IF&V!)N
zb_2Pq2~qXAt21u)(y@n^*ViJDmF)6o06T~54!jEn0p*Qf{HsR5RemtI%mp2Q*SMWU
zu8{Rel9p#c_2qDw&+?;kLZi=}RqP^_=4!p&_rN`0D*OquajhO%VOreA{OMe2@)CYI
zOh@t!!kmU(kT9o&D}IH$OCnDa<qMj@AxF-C$9VD4SXpG3H-4(W^REwCROciw8qPF}
zU6Q_B>#}jSfj6CId7*#be0WLw5bXtm)XD53<5GMD%DJ}M<)#_)E4vW%Xo>Zm4mL_l
zPV{}Rd+iwx;3#>rZ<R(aoqG4wh8RO8T*s&bBS}<qT7UwK1B6hCIOpV@EF33noWKWs
zE5OXHUyAu<R*2bu1Y;pZ3b#)q(p@#PKi=)i&sN`-|6XV&OW9Xb5p<zap2UAI%J9=h
z(C;$d3o;e|NhUluZd1x1ujj~KJK|Qg((1H^4htgLj~a3NORd2~1cZPfc{e6g!|4Hs
zjF1y{z)G_ZksPqHc#=Z~kSdl#US-KWQ}=IswV0!y^IPO<Qx5KE)a<<$4(Nmo$*SW0
zAE{13eW_GV68{9HPEdOsl71sh(d-F$?FJrvxoxr6d4%@PopgZc7N&iz`#ftH1vPUJ
zzY*mfwG{K2-}}`u;{0ah&3M9Rejg~&D>z3~LUU)8n48*t`Z}yC3M%RaF(Ic?pnQJg
ztsZL}aOh!TX;W~>QY8>npBAnGG0!Ukz%^6l?N|&r2fjT1tCCmb3FuH=DToYp$d?wI
z^Lc@wq_V*>4ZEtj*#Eiw%Qc~@=0W{hbRpp7t<Tyeuc54oNskOjbSCYDpNw$Jl}U=G
zDf_SW$+7>-3eF!Jy~rSx2<zaSJUcS72)_Jn1!qXhYj=GqN;mjyL`syzk{r}7NO4F(
zISW%|*Y8@9`;|oe;^1P4-7PJ3_^mQK&Eb=VPMx4TY8UaPQ$)khJ?98(6b2dJk~-JU
zj{orNaYmDpz~uJcO*JZg9-FUdb9S}0{^-ZC%W^!?l9p`w+w|>J@SRf^GHKnW3VDb?
zuB{bcC`&mT<DeDyB)iZG(%&q@D~aWE@Ed-@ph5zT{aILGW16j@wuM+RzPAob?*k`P
z10v)15Ry<F;`MF)i|P@^kizG|`3Wcg9=|hY=qt@8zm<;o88n&$Qfk8S?B61vEr}};
zAm<z#;TU~S7+*f&6}!<}4|f~95=zIrNi;GQEI%mTjKS0tQ7IltadU=+w9C4bzc-P(
zQ~ILo^lM6U(A%lBsiY4vDxlnYs2z08tnpoyMXp7}ej@0u;+=PIjHgWBobXZr4$>(a
zDry45?T;M?Fs1itaJMGiX8&D$>~lz>Q0HQE^|D5~E<QpI2r$<G%@%mHGa$Fj^Fw^D
zjPKF&oA=EOoTJ6=2%_=^MoxYN`D9QusMC!EJcgj&C)YDKLWgGVX~EGQ*8_Qdsm?Qn
z4NqbTrCfkl^_HuN@N6oB8Z_1UR7`wfjp`$8(YxowlGR56OTUJ(c|q}r#v3qrgp~?U
zW8tcJ1uVG{n*u6;g_D}gj>JAmRdRTBaSMiBt&(S*JgP+|YHE(#7dEWzN(2%=->|UZ
z%Kx!UXnlS4gXPnJYH)feGhGYBV~Xj5v9JaE=qH2ngYO}2!9V(g9ySh(n4N3;vAx_=
z2=%uh1EBEdwogU8>=?uwbA|mL*uh-blm9$FM^qZ9A1)!g7y!ACJ=weNSitDaAMU6Z
z6Fdk)Nkyk3Pq-jcdv2qxTv&<#$rrFR1t?q)yB<(@%`~{TqnJV8l5}Oc|F~TH`xkA8
z7_@D>^fXKqWZe)!fF;KyH-GB#C)ya^K7OSU7PB#|KU*K7NCS^qrA(+U2M;w;k{8}b
z`2Oep@gFW@L>2V(`*Tur*^DzVtcaJG;yOx)ADBgp=^QvGSnEFy@|_7${>XZ^dVb4|
zYz`c|kSum1MT~Ow>j1$GQd2~`Pv@%i%HfAm*EcM+D(1mM0F&sNnm$znU+kOfecSWr
z<c(C~h_+P%o0QUkFtXoks?<)wP37BiCjX0vyE1D*SryMD-eO&Tk^j}RQ|S?Lk<niR
zE6Tgsl0BvP&MfahuzCuL2vgrioahDl^Oh}-I}7AEPbFpy+Lw?lA~!!-GX)oAyy6Nj
z@E^tvu<f{$&2~A{{F|maS*qopJ>e}gvrOpVk2oa09^_MMO)E|-!L{CLnmTic`R;W7
z15fL#SNz$~iUh5G-@Lf_xXWZQ=}YCM0w;TQT99-WKW{4!2cqn!rlt_1cQ+M4K4AMn
z3o>1u$Ybk|#GqiW*pl&s*{MSvtCt_LYb?VHVu5*_I@f=Se-0^Q126k9JGjH=C6gQn
zyuO}e{u!LepmX|P*(-a%Rz0pFH%|N%wb)p(ySKu?kTF_n&KsxW^*qhxN0~DZXcF}I
z$HUIi6tct<#&B+2Cvb4v;ZfKAL(MS`JJ~*^h+^5|(v*bdV!>G^;&a6jWu0%FZcsJZ
zBcu!DBvhMrTg6t6Om*qUSM*RMS~3nlJqWK9Yp^ZQrT81x=EmOKjJ$~d3rjge1nrFI
z>Rzn7PVu*OJF@>`DBe95YovtvjshW`jJcy95vmW|2Jj1{EEJSM>dHUC#)3i9aHqNc
z3MK+8g!)N%UHtDa!NhWH3;IFdx!Ll*mwlyoSZf~+xHr2#vDoL%?_1liez`5yd5(Xm
ztzU&Bt#03fg47Kb)SM6Y@&&s|vc5LQQ!(Q;TN9=HnGU;Rj!}rMND`<?KA4ub)GkQe
zQ|G@A<qF9wbG<@VeustUug)nBsh(Ew>aH&YB+gX4iGXoq@BuU^oxcZ^T7??jRu?l3
z%Uhng<3&en0fDm+m`H)Dx5prv7x3dLh0}HcZk9Bnp83-XpVzH^Y`)Kar$c@0+#;mE
z8OjWu9^OnAIZf}rhyg@u_xD~mK~T{}zPOk8Gbg%eHJ42e_LC*k4NDK|ZIG2%4PW2J
zo?0Y`70p~kV#B+x#QLrpc{kb6T$$Hw@x|zB{;UOXpFTRIMSY>({oQRl5!a6WfRpCJ
zh_lqWqp0V>uk3u+0>L7bRi0#3?_0`JTps7lN^7%hfi+|o!r$=U>}wpUtUcO5oe)He
zY$%QeuYFQ#Y#aOaV2mZhlnX3V?vnxgqhE~UjC^-PWi9`rCUQUP{^ufR*Sj-y$ntt{
z`q8R}UeT0>BHQKTh0%fMd%TGEy%^Dy_p%dbi;0Tcv5cQ|KE2`L?m#b%?NWJxlRf(u
z7%!+67G$)hK8fHjS+eEqhC_FxA{aLZpO0S_UcbAjuBIsS?Td$M{Tj(`x#(z9UB#Vm
zwDhL%U3Y2PEqdvfR^H1ogg;^6U$3)%QfPLdRK1bS;+6oXiOW2!?GxL+RgaK4zM1zc
zN1C8}#U}9Oco@XAr#OhJk5Nh>PZ{RJmCm?~69e?@1WE3cK>DGK#BRpTiLahY(Cv49
z9-U}jTj>o167(Rp8$WwqtmLL^kK>kuduDe$+!WpRp}oC;Ax;beskZJ|RLMEx-Cpip
zo`pR^#{08VeO?5ZdUK7<MO>SfQm~sxPc24+*j3s**khIbVqa9L-v23`aoft2!9#3o
zdIkW{ug(F;u083vlN2tM*aAw6%Uc5>3H<G#lO}iF%DdvHJE(F78l+t66A3PbmEG(A
zOA5~)BT7C1=*=|j>D2P7HpKsh%uB7StYQUQnV58s!U=z{-s)=CA;mEJvkp(7CFf!W
zUykvcW4Yb2w$Yn&m+Ipq%s4IWUdB|0gmKjlp@A}v9}$*{$J}E0Spkd4#*8=gDw4HZ
zeZ!JNldLzlVpC?Whjs8xZX}wW?I`k_Hc08?Vp}n%-_n~k6WytQ4;)-KR9~=32$n64
zi%B@+2cg|k<`PX*gv8Qn(~>1OWI(z*#`Se4#FcX^wGWl=#~E+U64LZu`e&*<R@eH9
zf2wZcJ4()O+h4};H3Aa49IlW^_dL73qKC+yK^uB|Z0z}BGf+}954b*gLt;P~CdJR3
zT)nTJLWv(69+l*3&?5Wr^g>%=B{!^;!PxjL<be8s_XhRc3C)fO@i|Wz_U)fkjVDgs
zylJ&{jQusU#3o92JGQ!B*SMus_M6S5XT<H{$uD{3m@m6ojimz|ga7u?276%*dwq_^
z@^tY=Wf5Z>?fZKXc!~3~?GQibUtieiKq{~0<mX!Y(S!aG?smgDdDb1PZQ}%LessvX
ztcS|5smQ(Py4`n}M<Rs`dEqV%tKrYt2d7-nojfs_CQ1CoF@sW;oTtlAqWsOTook<f
zv0QbGr5N15*=ITxmsL2IIpgf<cN960fx<P`(g&w@sas2~`;+)j2fkqrFhYD5D1nGD
z`yzVb!*fe#)xR4~4QRvdC>e_sIHbhy&g|pEu9SBXsgUh@0D!p9dp}MSe16C{o6!nc
z0Md&fAQ>3BVq$mD2bKUF)V8(NKk|By!>C{6@4}KF{|p+nI0V-K-KV!3U2O41wo=~n
zAFa`3fMrrQ+1$md#RrMgZGDkhBaOdJxew$<#`ZJ0Tes?BO+XZ`$7s(S-e&}!S45vF
za!;FYlLH3ikphmgHEC&=5f|u%T!cNuvpn;Xa2JuC58+t&-<tu}&Ggmuxu+hWY+ci8
z>p9uy;l34i0O*>f{m<Kj_4RkT8S!WMDqN|LY7uYQ&AXcnhoUi&D`n8)MC)6}`$xfG
z9*%aozY@J&NLO}sNu~T6hu2RBB4$p#D#EQL#s;0dS<!Dr@Pg$+wfiDk9U8|$a=ETa
zjJls6r;aSG3~yzbH7X%?wxPQRHz%Yhf88~VnY7TGBA&BY)7q5?6yekrXny@zT#rEI
z5k{DO59fvlP>a{F)9+8?A4c@=<;WdaUgD}N_d_x(o*|^X<7T_Y1fHk>7Eyzt8y}?-
zq}{?4K3OCvK~ZtPnrfBKL*W|o@u#+ujS(@Kcp<^vg$P&|yooSUVuGfS-_jgB{gTxZ
zY5m*r^9epG=HF}gzOriG;3e)Io!@3h#&e9WWH*CI9rd=S|7Q7r6crAO6xFeCJXtHn
zR%0Ac_5155>Z8k<cU4fM<`;mAlh78G1$4OsJL=h?q7GBpWp~m8<MpM@8N@{}l!e!q
zXT-Gh=rhXV#aM(DY_s!Ux0s5(pw<_ASLZvMC$D@|k>^Jhw0qh27Cvur>MWLDr^S9F
z;3+_&A@(qI#ci-?NQ@?&85+74Y;K+|${?etrClI))=6!0*a05&6438&ax)tpFrkbd
zCKtPxA|ZK=?jw-lRqE!P`3g?dBJnNHfzRBB(1TN9&#NtxfySoxiUZJ79I;T^1KwKK
z2fQk{TBR~)J8UfqXL%U;4NObm6#epyhF=&z@mRr!)VEOm#rp7N!-wlVLu@B=L5iEz
z@Hh{1+BV&>Uhd6Db7dPl7kOqRVQDI_6Qx69hZxHJYjBTfWG}ebd2m(lTrSx&OWrU=
zDuN0iZN{!PB-)%YM=*tk*kzY+zLz+dqu73&514+pUJ*P(;=e}9VKd4EqG>OqPCKud
zV-gbOhof;59J)VVAcbG!8}SxpF#UAn`~OIGNYlAm$l69v)SJ(RmkG*B*eQQwA;BYD
z^8Kw+#N#4W+6g*`XHa@`M@Pk?5k8%*bWQ1}Q9d?CE88&KdsIBGbfYBrFBVwXie9*W
za7=3*#T_fJ-C!uePi|W!Tp4a*<&fIv+gKE7{!4>h!}O69cCtA`Wo*y?WP)@G$10Sv
z1+~0(RYHrfVzqmH{3TK~zj>&W>r)iT!qu?saffnv#Ug(IyPMPBM+;IPM???{!)72<
zY)><kLxAa~DI}f>T$VQmAIQ>9!Q6gFG&PTFql=-rZP^s|kv;cM98dA5MS7Liwma36
z&MOqdI<-fFvE{VuDmtXOoJkgBu3~sW<`AYdR=9t9agdf}0-O7zAkdnoY$RnhAt}Qt
zigTB%Vm`3BH6yta(0e~HmHn$@pVEoe`RXh80#B}Ov|_}>RJRR>=Gse0ZtImAItR0K
z*0wItiJ%qu_62T<<ZB(5pJX2p{GRfV_btrLfS!wWNf)h|Eza)@qA9<_&k)m}+CO6X
z=x3&zV`iL-X2G2n{cJVm)K0B43TD<nFNa$G27in*-QRk1bwH*dB-@bST)N0#=HBZ^
zx;JW~5kz;h88+`m2Bgt>8O$C!{xn1ehD>Q3P33zk+s?aqbDf<M#Q7R-?H97dF8TSv
zGb!<<Xod5I7|AvAuFjVe%n!^DFYl-l34llm)uWNWc2BaDR0`CiVx8c>XnM*Xn%(Pa
z_s*U>88K56m)04vnRpz-*>h%H1L>bPGkg$4E7{Q*7Kn6ub?sr~{tc&Nr|W`%zLI{L
zo<EO2DIwFamp^D#-~@&-29Miyvi7ySJiY~^xZoz<-LgbzfL_z3kj|#!%TGhVaP1!|
zsHE<h45U{hXE{`@mX~~1z(I737H@xYPal+#@4Lw^zCBW|v{Ul`q-MnSY|e+`(_@>w
z!8Q<wpYiTw1g_9bK<m=|5}a}(0DESY%YU`NW$`%JUT2GJg3j@4JHHj*D{N}xC52z=
znNeHaKhAYjs+g$ZAz>g?vGTSm(%<0CW25;2D@S+juUYsy@`(4B>x~K8<kxqkYUSge
zU7Kp@zH6^~#j`eCuc-wrf3z?%0}NH^tCgq^V@q3)M5zi;tId($Ij>ENDr_~hf3q&1
zsXl~gyiypMB-&ME!_rcLIy<caDwRyjpqKh)5YITN^(v~OgKGS9EpyD!XZcPiiP^eR
z_I1tafMdr-fh8UGTkX&aBR=h(fjvgo_a7yKLtbR(Et=nT?EiIV2-76-OJgL3DN_(E
zay8omyy4&LH2ub8PbJZf{dm`aTA#v?^2NKO@I=rB4}Tb`w4=JvDSd=v;YN032uX})
z;q6O9>V*!Oa1K-gS!J5_u>F%)iF+IOD}cC3j6`H17T8jFUAdJE%d0TbBj=uW57+&P
zG4}nE0{ai3x`&4W0`(8{-{qbb-fGLpV?#X~hf1qvVLl0-MgD%M!lo^_(M<8@C#DuW
zA<yDeSUfa$36&i>78zMs^V+dLy!5^x4*7lc<rCKDcRjmvgObQKeD-gSp<C527A{IR
z4z7cjp12N8pFEA{2XFON9IkZ~(=)z~-^A!8+@EyqKuTLi%JN!l%G#aUw!k@3Zwo*|
z=70rD&rkjoUz`+^1C%aYlNc^~Irc9`M}mw698u|FWUm$n1)%E4i><>B>><itjjE*X
zrHq!wAEY{!L(l?{?mfB!?#5MZ=yRO{@cfEa9ALHtGI!tncQ_84`ePn<z31=^hbACq
zxRge*Pw9!Gs<2hnV1?92G4U_`KdR91+kZ2x`{@9p*E9G(?uJa)U?NSsa13qHHFawY
zNynlSC`_-U9v4Tq;eA7m`}frP0ux5QV4n4B|359vWs(@%YnkaSiZ_nBk6hNuP|?LR
z$vPC}du@0kMTUiPRz@<_FR32RT^}gxH2`Ph0hTQO@UnVIkt7i)-B=U!4#VHIB)8n5
z-QO+XKi;M;{<!Mc?6k~yRCi#_s6dYE4HEy`@nYUqT-^XaJV&f!$)oe-6@lK~QPg>@
zu-nn_jL#D#z+4jEJB*XnuMG!r8s3}`=dnUpf7@5$f4SuLLS@*{4&L9Z)Y$X>`8}ah
zFpGv{i~q+giBRnkV+ZGIcVQ`ew6DrQp#iQZ7}mypjD3;{pWWw49R)MdSzvM)O*s!k
z(N<Bz7Po#BcdI@G^d%d2ek}Oq7=B$wa!cu8CU!6Gy(H3IF3o;%<~Qn@U-lo)c!eJp
zc-D=T33k&{GcH;D(~cX;+S-GqCMZJb04tq}Z+Is-PX7M!zs5pNbSUu&GHQ|R9N}EV
zzeJO)Px2CFZo@n%4vO@hA)<dHRF>IleHXLtdR@U%@dBRlOr8B%7_@X3wMuGH*!oG!
zZ=8gc!=k{@b(g%fySVK>8)kD|3$#AFT;Iq@P@OvkXmgC~LDIX;>Nc9(MqDchVouz}
zg>PVFJRrjdWng9}rFBYw;U_Aou#Y%?t`{0B`K?;RO_5ujz@&Jf>r}qf*a4!9CbX*_
z7zvdfggz1+@tnk?<avjp-3e#T#ho;Oi#~g4O5%(+m;-RQ^1cBJS#dR+#8-R%T}Ii-
z9`ry$Pp&CGEE67{Jats$UbJs5;+d8s1Bkq=C2rq#K2hb#ssdc?m)87WEx<EGeud+F
zI^kYq?8aLEIkI!qJ!dlXa*_7&)dV#lO}a#QWo4y989B(xz#^%*G+|K-N$9DFH+f%y
zfK6cP{SH_FtXos2CE(|>i-#*TK@}y0eg$%>j0s}xB(&W}@2^p2wR@(7PQXtYH4XcA
z9?N=QqvEpFe|=LVbV5U)>ItPhD7S90XO(y8xloRu8ZJN{{{H$PW@J)Y%u3C8=%ax`
zNzSe~Zu2BCBoIyuqE4cPUCGe>EZ=8MMV%S^7Gru#u9&)5>CPx#oaTv+mp=%2#Y$%)
z&SRpD9js8gtJR>5n*AZ04nUXTVd41_^0gm=V_5WCG&tz&edj|3`cB|)Ctt)KOX~5z
z9?n^B(WH4Q;lK~XOiZ<k8vXzYi12?Kjleqj2?^3gzW~?e)hlgVVSVF>lTQWw1B-Dc
zx?j<lMuNsQ9?S?w_Z{GaoIx@szrz}L4e%>z`>NSfTaV#gr8-t^4auK9Y)(hvhXPcC
za#D{PW*3_-+Zb)K^LqMC9_fL&%xjn7MV|JsaJxE8CD<c#_O-R*&2O`y@h)(@?(*D@
zM*eVOtC!n7-%HWz++9J@{3w>nQekyo*6tZim(d@4<~&ORak|)phlod2&+jE=6`{!%
zgdc(bWwUP_eBZ4eN35C-R#y9#p6gGT=(#G^VFmhrKS)|-HGf@swwAD;ch7$$t-t-0
z?DX?Zk-h-6*+os*BmPaQof!9960bC1ff2k6F&oC&ZH1bf#W}(;2>*<NWNoY8vi5)A
zpSlo#Sv1Q}G=_f`tO-$6ZB{SG1A*PrDxa#NDy;!oA3uLO4(Z}}tFtKZ4ZKd|hR!rg
z9;nN0-RAmmj42Pvz7@cK(9h*>8=R60+HDQ;xf8Z|b&=9Xm}#Xy!i=br0d5^w_g!rr
z1ps>9qva%tyrCl*l$WC$_GEy|OCN4c6AEM~k_m%>oqTBy8z@4Ps==>L$&rFLcQOwT
zv-txFJz~2nF`h3KbT`ORtE}PxWEkO{vJa*Wdx7zKStNg@0{{%1ONN%feN3q9kxg*9
zR*w+ZdyyZTSlP0m!mmn&Nzc4a{<>rxS2DHqOUze6?-$R>#6p@<S#=)TOU{cbSoH9;
zzu`{u&Lce#lTKc&nVQ`DG6)KuHvRvzyGZom?^=;ZFwZOfHy;+PEOoe}p6uGsXeUJy
z|6*Hqj@H_DD-@kJsf)PL_K&3z7(u>snZBL|0?Pr5O}}$NB<+6US&+ACLl)w?MNjJc
zaAof-iJy&Hs}eRxv|nk&JmYGz5-pSB`5%|ePH<x)=8$yL>obX9#2QEKXZ_evt2xg;
zL07dIS>BVZn6p_z&igm()-VKHK9p|RQq?6w%rhwGwFUNNUtU_DL)4frzr_t9Y(|i6
zI(^N+7}A+yL!UikcWgIguR~-ycP68DgzdulJSV&zD&Iz>g)U2z%1;p73}zaU7oVll
z&qzy8#6O}AQKiS<FWjJciP#?8qjt4kk6>W*FBfzXK^CFhP=tptWf1_T)AiSytJf7@
zlRaX{#ZJC86CWEldrBI_e6f<GLM|EG{-QMvM8gm&BQ<b-&F=)mr>zSLWhIniR$ki~
zI`Bu$^<<^_p;$js?_GP*R;sV0U&vG)zw1t%JD2QtGb(N(=qEzo2tOI~!A-|S#sTjf
zq}htNgogJ&6$a${R)g~j_2$6I9LX_*A-SJM%4B6#7@{jQ_Z@erf4@-~KK@Er<uhlK
zWS?4nHs>3VgR?CV<3Ei)@g9)mRmjnxc!Ag*G?>tI`bVuj^ffb83cEG))uI>ib_y()
zP_ft6;@o1&4+80(*|JdUAyYHH;0kYyWXwX`TRtJc8>I3&l!c@<GK`V_2p)!(0jxke
zNi2TR6$sFi+8ss(c+-h<1@v6u4%j=YOjKaWnt&-fQl`Q!E-nW^qjV~O3x3gc7UP!F
zL{x5ZnV^wdZjVye3&#2ve-E6D?mly;Q%gn_sd_TLnCnt(blQjb{sCK%^l&LZKKVod
z0JD}zGiow`p}l-B-U?z<l@hM1zhY?w@>bw^^3a$y+=sT%Wx-X4T@HKcSuXGp&cbQ5
z`j(Yel;ztu4L#g)CJ)tpE@8m?cFl~Bcf929i>QNI{=WvL)>P)cnoxh#kKJG!fIf}A
z3V4Sn*^T0Z2^maP^&WS$BHw;PXxq~FUH@OUl$5**6wWZnQtROEokmLc$KuR8-(;2m
z#Itk$$IHyCU0Md053%EIe#pD}zNr!xwR?Yxd~BPumRjrQh7Tsc^##Fxvzgobmi@L8
zp;8nv5UsX6oO}9Js%ZxFqkajRl^1@B5H}pAkqA1&1brR78ft7IXD#)k<Ob5E2A;eI
zC)yhgkKHX9fm<mg2`h6jePa52PWM&xy1I%RVs0ft6yxz%Vf<uLi*0>x&0EdRb?Xh>
z@Tjk8h*kz@gv#dlGetM9`$d+2eSMHakxVVd-4*K&mAcNnp{sgJv};hHe4Afizjcr~
zJ$zXsJCv!L5#+6OU1-aATsn5m9~&n6(;pvDbVDlaH8LCh{H-7V^KNQi3Z+*7Y(O%d
zDN8Zp=X;TR+UA2NucP=owDu0~4a{XoJi5vMYEi;1p(7o^{qV3^@)1;Uigqn<GMIg&
zu$R@6Y4ZA!z)rF0BHucu_>&W>vBScMhGn21-8e(ms`v8B+0v!rC<P=HxH?F`6kXw)
zQCzv}A~ubF*Csb56GEQ^eDcajZYzuRU!u&@S?Tmejeq;CI2Y{O4F!<2-UG4UMG7QW
zQY&$bbUG#6_QG(z^G^ZE42h+tM3nubS|Z&4A{79D?DW@@==771{bCnBQsJ~zq&#9=
zDmegl0_#H++dz>y^$|NHc74hErAgc1B3a!XO?yJ|W%<(8W<EbDYH$xH1^+)xy>(R6
zf%i8)LIF`}P>GS!A|cX^AP7haLrPi!>5hSblrWH1I;6XEAe|$mV@Qo2Aq*JXe)~Mn
z_w#*z|Lm{#IXh?P?!NDR-B;99un<iC!+o5vcl%sr-hAH7^1l6-y))BFyFVk~WMy0b
zP(o}9P+AkL6^${-$Ip$x`2DX(5F1^y(ryC~fo*<~wBP4&u}1{Ft=Pyr44mj^$wl_X
zD$cPXi}7X_vyC7~I!CSp3!S<%?SILID2LLkoP=6RED!q)S=X8WeMy9RJB#ud@k38%
zvBO|C+-<ec?U(p&UzwNLOI;<~$S7G8dL?PA)NU&Qsr9}2!4*K3k#<J4O`M!Vt9n1y
z>nNy7=3xsboo44~QG);(qLi2`=g&hL2sx>Wctzl#toPM7bmlHv@6lJ<Y*N`YC2#TT
zhW+UD=R#+tuc_Wf=?K_f^I21gr0Ku}=?LEv)T(ePLiE<%DjRIw7WHwyBKJ-1{_2}=
zW-kP7cTXGLxJAHOlQZd<Vg;U4eJJr2k&r`7QaPJj=*tQ|ntmgc!y-sEYWzr32AikG
zXi-&Pr|doC0Qv;bQn9(llud&Dj|JToV?KW-n7o?3&{~=*hUS<njQt>~Q*g)>ixRSD
z7)0eVxVf^DlV+RP=n00uDGC+ROgP@ONMHFr<**@-85fI*JS9Tu6pKkJoRRT}q~E-5
zmDX4aE@uB{d+Nt1aX&)r+%G$+Z^L(vLMbZ;`8&bS^*e)`?`8bWcQ>0`o|ohYd*Rn`
zn~c!|4EGVz7x(MA$uCFXdhEF!0n1r?_362=oh%)LDyB;e#{^incgea72t3=yQ!N<Q
z+U^3bHat)OK-dILVc9F$mEQn`{r9)ft|X#0mq?qO8FDiKSrFf0w3fx}*NzMHoFI9w
zfT`b%C)+)>^^yMXuLeS2^X9!$vR@KdvK^G8+((s_5d6uMzpYju2<1!GXO*-<a>de@
zsBy4nl**-$DK{;%1w3HoY9?mDD<(=xg+i9-31+8EnN`{C*N50G4LoMB|G%cV$Ic%@
zpIb#jnc&78xYgBlFG~IxsJz2sBrs3C%$AE-Zi7BUC<(J4*R1Z6$pi&3)gtI)gqcHX
zk8T4kuzsE22q^=uEfYA1`?)nwwNGo#Y{~FUTF_OFq&+`W(EV=mYDGKe?B75x{-y)j
zPYXVfpW7jdgys!(hGL;o*hh4&eHskt^8$J{9^Qy7YnkGP{_XntFGi$qP7h{!)RrlR
z-XIkL%Be)KYIDTtv6at(jH;%3v}PM=95sW{m!hg-LdKvos@G+tx}Ua<!b8iKdD=6Y
z3i30ujfUigi^)OWee4t5Dq}`DOuCq3${01GH%j6$0W3RFMZ+5?PsT_cPy6QOrfeS<
zd3%p}+clRubvgo&hLM7tD_eKo#!x}9qOUex?Q*|AaES2L=(Ll15)W}XMe+_rKuc5c
zWA(7x_huXw0#`=Kv<&Wc3tXR~^S1mtMWWQA3`D7vV6$#qV?(Pvg6^>xN%s#xh78~g
zoeVH9%Fo5uQtq*8pA~6HZ1{No84@f~$hNV9pV{y6k9<KKbnafu#1u_m1LMb^j1uO=
z7VBu$n3nAZ6B%)iA#7h7yL#pXAp*l6ZYiXGUqh|=69G;v^ikV0Y4;anE4ahWDRsf<
zWtYXvDmF|`;FSnIDIF3-4LE6j2*y7tio`hsx;q(VJ@rGE+kqs2l@>gBq7ysu4HJJW
zh5mpHWyS|<bQ5f1n&$)=*X1r=+ypt!pWDAFJvrmuq&DsRxYC3n5L_*O2LOoHd*{&I
z(|;puS-VD(@Eun*q{Ma{*R^@4lBkMSN~91UDN>aLePsJO%X&`X<hiB%8HagViy2Ut
zax9+PCN^Y_@PE{4*Fn0U0AX!s4glEGU9iWN&|ePNz4V40aF`49VbogvJ^!G;Elb<2
zcUm|;wbn~0`4H2CFjEl>`G9;bg^Rn|3imh<*#%aJ6{JjB@A;}kROMA0JV^Yc>bsd(
zx&EwjNn_}I!J{aTZaWoi<q<Bx>EXR7Kpcb934gwSQYqVM!fz7Fzh*3?&{fpB7Fqpq
zE6HrR79Bn6nkq$n_Ltv$IWfpGlAru%pnp5Ml|7#}Tbv@rR$$&6JciZ;9b0c%ITZ1z
z6rtSI`(xF5Yz(j9w^E(ZmV5|^<wez#Z6P^G9jO#K5ER7n?D#9$1qCe4J#U~#kp(W0
zQ$cA;EECq9I`6TM8Ad!UWOgh3_gJP{66M(8`0JAluj71PGMk?@DA-OWu%G2E|E=|#
z`_H^PJo!F5KQHHkBP%%6FVz0%4+%c2yUon+>cuc$8!QHz%@4yP6K+G?F56=N$<n?=
znw&m)S+{k98GHUiH8nPA(77Y+^Kk9vNGV&9FsR|OFPe1o71E7KGLj<scXA(_uE7`&
z82e%Ju`b3c)G;`wUX+>kX2rAoFpMdop5<tgl^?K>B9<4~xkb-}QzXp{85*ZLn?b!r
z1pLumKU5*#VBwy76V!)@wPX7C`&-Fxl+pS1b5^v2+vy(rsO9s9_;m91vB$B+7R<)9
z-MuKB1)v*d(TFse3K-rm7zj4(u)kngK&rc7Me(L!K^qo?fJ_P7!&}_Wi+al2p*SA`
zfP9D3MMW0BE}?JDsTQF7wzk?hz?-GuZcVeZ6J~9N53qhL4SfJmPhq=3$Y1sXLM{TT
z-^(A$5#TNS#a6#GJLt--G?#=41x_rXetQTR%MWej!dIyZwm&a;jT|ok33$kv?tgrY
zqyyYzVq%Kh*RWnxOgXnUm{kCYYs7zJd0x~VPTk5?6JQgRIATNe^hN5hr%hK;{`P-`
z#lbl#Qj7WK<ff`5|N5!cv3%GR{!d?ua-)%I3&B$PKv-KE(&rt{SO%=}Z*OnV&}^N!
zZ-qGSe(1~fQdC?kk<|3>R2Y^%<s|b?`Apbo!{58lV}UNYDX=H!5~Gp7=wbrD>Q%M#
z!}IInYB4Y8oNP-D&Ci@)hq_wJ&#R|Cy?*ri1#M-8jwpG4AQB~hv*=eY^Nf>+t5!RM
z0>Jiyw|7sb@}!^F70A1c67e8V_b^*zuzj8Jj6>;{s#1mPbwbLEZ>f;SYD$Yo9er-U
z*pns|-)?y5|7<|}S+i{t$${1AC_nJF{~TPd2Y#IzlqDc|74d>j%M^^Y<aO0db2LM|
z@by-8cC_L_vr5uFanW-~{_^3zW}2$ny)ci!iE{(mKat8z4?w@gbY#f6!doOi``&g3
zg<PW2cOKBmq%`E|FJR-LV0OD7v|>WXpO+-$d;KmFQX^>oFr@1~B$ad6R-ea4X$f|I
zrD@<<S>CUQ9(KeI?lBsQk)z8M5AQd}2uUOle79ZRV2M7GUriinWrpY4+CtQW_*Al-
z&aU|Ah=9&hsL{(2>L$068{6R|iqu7w{%zNA3ha0gnvbZtVTYaw^dzFimSD+FUK8||
zboD~D@P3;6==4*>=3ay)`;NsdN79MhfiCT6pb)NS5~<Kki(qYT*^^|E5tmIrLSM?6
z`(NNU`J67N3M@PjMtMyjNX(!Ugikr!zJL<~D&)m-ZmT;yMsK#`oq-*8=eu*m|Dk(!
zlBAlQ{5Ep#z!)*W@=9pT(OTyhFddjBz|wnZ<E$B2O#pZXKEu;u|Nd&+Jobu(%j2f;
z1_##4j3ltiXZ1mLWG(e}N4}_4>84I+(+G$~!+AVRbadw}TpCnnb6N{T)~PG#E}?i|
zNCHM|)Ugr>-7$@f6}{vG69F9t`WFdfxAL0g8zptZty)zPojQCtU54&)d58b87b+5j
z>S*4{)p&7Znn-%B*56G<pi#h@4g~5G-yd2uW2@pw0J^vn)hLk(_g`V{c~Nm}SN6vD
z1`_1XHd$sVo5{9*1i}m&Hb-nem@M`mO0lukeq>M%3wJFl?UblXnX)#AO23_Zw%pa5
z(kwv~FLC`!CWe(K(vQV&&EjZk<YQUj&5u|7u9YpeijyP3CK^^T=Q}|~M$fk_J|ZoQ
z^;!reX&RTtFYmkOWFg?G;w_mICc@Mrhr(F>p|`22UYB5NvN|wN_C-%BCdIZVD9WiQ
za3&5W*WQW9^J#)26l#v-xZjTX8%LlWe${}?EgnBrA7q0(BbS>YJ3RTIpts@VS;#tc
z9^ThaM_KnhyDwc^?Gc!k(xWq4TIaso;hl*n54H~F`oS{?!i!DM3=)Syhr9hQj?g6z
z&FG!Q2DMFMdS6l`&(&JK$VvsGZPyM4ek38}PoX1moK_g1kW(aar<neMA}Xv}>Y9=5
zi=jct#vu3Dym3BqMUcF4fKiZJmd*VjhY>RV-Pe0~i93r5Bv)j_P98hrsj#v^@BCR4
z$OS}IO_l<wtDm7TqwJzUcV{u(=Z7eEC=^@PT*%F?Ig&v0w=a;iz$$u*K1NO#?4PWZ
z?!6Be3>HeW_aOf=4|7}H0~ss`y7&DGNR@R%uRoW@$<bY3YAElYbB`H;_x$uPrF8^x
zInYZyHZagl>3dr6b>Z{}8x{g%{Ecf|mHbEWXO9|SSp~Mtcp0nu4R9zc%n7&u9%)q;
zd$0TWhT{^#OJC(xX*T`B7xxt;-o_~}AaXM%&=_&B$>g3tQAEsXA{IMsZl3`#P{=rx
z{Y6+{qRVH8UjmZ22D+fbVDQ7oxaz|)B3|^zF1n_yM{u}DJRM|I&y=vsfoGZLf3+v)
zdOh^3!cSN~LvDqtXTE4<hLl_}(2-!`@eCLSSIOr<<ttT{lSGWX_v%bDqghx-7yH$f
zegiQvxe8({FdOrRe=EW~xOY{#O2u?p9(MBr?`8RV$f7#sT#W73T*1|Zhx~)!8-y<D
z2vYDJl7pSfjVPJxHNRD*17nj&9foA$b%M*ErtL5{Rl%-3FLvf@8`%KnWkDP6RJ7;h
zPFnEq(g^J)_-5Jgi*$~krq=~7^0jpTu2Vi<l52?*gpR{>>|yg)+3tmZSS>H0&YDDK
zy0lg~;!Z~^#ihFhaU5)~XN7=G16cxM@d*hGY^$F+=Vb~=AM@`kbQv`VwiJbJCUmQV
zcEuo<QN%uGy5BJ`#sn2(q(BMrUnonLss+He{NymHL8M5~Vbvr(Wrl!3Rc-a|on!fa
zw`o=-b_zC&PN151lVEJ<zL7z6%+6k~+0;RQgeARF{QMn0^Qy5ZbiplR-S5(8&5PbW
zoJr7ewth$|&*3&2GjZ%1pC}Nn=Wl?^&g&ZVJtIX&?wrd`NXu4xD%_Jx_hhii8+Cka
zv?YDtDZnkj|1Sy0<(Fe<pFot`+U|Acti^V!^n7@ZB`EPu{())YZRGD^_t5totK^84
zp^)LDS=-p3@(~_@kRwMZ5y01HJn&{Ud#}$xS$W}oO0wy7`*}z00q$y0@W=WYuQcwb
z(HjO!&*i{N-1du2+71~<(=R$sxTk=1p&apWu&oz47xl!1NhiGh6Kb)=`iM7j3{-Ox
z^oYPXz~+L!5rcJFYrb(APlrH91wvtFzx4j9&Jd_)g7v-}J<o`^QA-`93c5Y}OL3!j
zR=>HtwLYZx3tkjqPDWeg;nxO`Bv#wk8BTeS^j_sCcwkYz;SKjck~-7Vz>uT=k(J|3
z+Z0Mae8c{bPgah2Kt6QRF%=FK{$%+`06^si!EHy-ETFtpHmvDG@%r&v|3+rY;F?E;
zuaBrm6o$0qK2<8rp*(;YKV=pY8PU^M&2$M1*JWx^#!EP*D()exRM;(B&~kCB1^>qM
zd2&(~)5btiI_Kr>*_(J{1DWg;@HV;G54O6DiTQPOm7y~t%f_>(t9kGe4yPYo*q^!w
zx}J?7)k)&dJr8ftsWE5U$|CANU-COE{Zu5kJ=8NI!sa}V7C6I7$q&Si4s!R2-$R-a
z9Svs(1|90my@YKlC09A{6po`EGI`D4%X!G!8>L23u2PypU;;`**>{AQT-?=5n7zEa
zVg&T*j83MxDb+rpSs_I8+e-2Na<#Yi9cJFy1CX2dNQ>03@d6r8=2adl`PV(-fwd<T
z3cL$oTgtx=rvqJXmCv7(D)RZOsv{)?>8g&9E5z{@c!EN2FWyLr@JB{&tx)X1KT+BP
zD_t+m8A<juA#mSKa+Y4(T`oz=eIt-k{Anw2<}A<jsc>vg3*|D&rO(HKoozV?{}V-c
z%DuBvL?(HVd?v<Xi=H13n@QLd>jv0J@<v`u#Dne^e76x09*pkNuR9z*qbhL!$100R
z-HN0rX*f?93D7S#Ck6y!Xz*|NwE&yHYys_d1QLx_^kHy+gcM*s!R@RXUgvx32zk@$
z((+*__;LeJU_5yFhGE^_ry?7JH%|kAtq<_48>H*gA*NI(Q`L+v{k?E!sKr&0jwNZV
zT?v4x;=NeU^x5Fy>GwE9Jq#4YkkjmqbOK3#7CnHM)(_Z9ydqx|FK|mO1{7`~d83JP
zSMD=!-Zvm2V&ziXIM)yt(fdpoY)_%K$gYq&@(@}+&HVb$O||SYa8MUYm|&`A9fQ*G
zN)aS9`Cs8Hg>h+OOj#}Qm4$R!gh_oCE>7~Ur@6{>(#9#{ovx7G(OZ8M1_HX@An&5q
z!20L()*&$p1z`fz_Sz3`BqNbjiLD8#mlWy2pJAIDdyX8_Q*m@C@K2xYdwMQ<*d!M#
z;%hy;T}YwQm6D7SR#Sy_%7^Z|rQcM{JAP+20=CMGb&u6#J(_BxgGND4`o;~0FHLR6
zHwEU|)(TFyYRe-xHfY;n*I2co&%ZXFocT3NQ&Xb#SC5%UKhBb9fiK$EW|BfA4A~vc
zmi?=q#+t4PkyJ(Gs8OlBg9>nA<-)2>@A87a+^|&zK5~Cma`7p0qZlvYUa?0-P4N{=
zbcxC7N4>X5>ep<0&`_gJP|Z1ghD8MRMJH|_J!e(kyEmlyvR?vqT^eXJiZJ^oR^=R!
zLuxT(AvY5FZ3cOl_{u>|BSnS%b7C!2>53V)GlZHrK<deFKB`XJ=5cougVP3ys?B$(
zU99ifu5YFTyDcSuB$^YVbyHbXUw&k{_Zr<&SvPor+C3L6rtU+wrDY#@4N-rG6$`T-
zsasrZZ<=GCYHqj%Q}~DK=OMnOCYL~{Jc3F}vR~z>z)6-fb9P$JR!D}jJ>sDZSlKP2
z94Rk6Ot9Xg4CsEj)^gWUrj1zTHuo;>tnPZnqSI3DlIB}AKAn`FtHuXMbPx&Jyj$_+
z@%(yE1Bmn?c)^{-6Q_25A(Tq5deRxk?0FF0Cea7zHt@P%Ks=U>B?vgg`{t)}1rV<b
z7tHuWnkv?u>rG8tY^KOIk~2BTY$GKpc5F-ADHu6FQxPTHOSVD0f0p(AUh{A(ef>AS
zgt}jL5d^odxVzoEIdNtNm-4DOo;9du6D|StK=U#{W4Oz<D(Qd45WsG|yaCo(HTSy?
zY`-^68T?XvDDO~i-WIT13KBG&X#V&jRF&RM_UIfVuS_sWw>%I}q8XAyDL}y2VWB1$
zT&WO|T=<8(kR1qpccb(wQki`&B-fCWS<psQtFO$zFqu1W^T*$+*V!DT^J3ge2jRYd
zUKb^XG~}22v#lAauO~1??rDF{NKi5!o2fk5?v=j(<Cf6X#MQ;GiRWg;ip7^r|JwV-
z+pHs)!~JL1nm0A8+>h^m$(Co&i<*3L>GyfVJ<k22fU-w{lW}ekO%}}*lU(7wt2DP9
z_lv*M;KjGVTbjI~;-|6kPk+5@@MrmGeIE3Tk+nNMJ9hTqx)A<c?uR>P0|PBo3>)4j
z2Bv>Z$APzRn<%LiIZT7h;hi9`XD$WNIdvHSY!IYvSdwFK=y76^GD3B<C|kev-le2A
z!dMpu)vq%aZVNq1F<1S@KIx+Z|7p+JelC*A!K&c#@gm@ubPj*LEqQvBb{w*kvY*!P
zCbv~H88~zGh?h>m!Tv?h#$VXzx%;cxM5iAcCcA!qSptGL$K|Ur!KW49C;TRTGU2>#
zq)N%(0{i`Obz`MLbq&5Be0w4`{>ufx$2<QeQKw%H3--29WqW|$;mG;fED*39u+t8O
ze!KzP<eAz?r*PQCcb!g5?rYH%hz*U%S2*DUBU1%;3Nz3?UBrNvb4Mh8rP}|pM}6zV
zKmT@)?7)ZbA3@=9JR2?s`ZnOwr_r7%q$fe+%N8q7j05uI5y0|X;V|eu@e1grrlQf5
zOv?3@S{0IgG~1bxq^QrqA*$;Gl0kv&-vXGZ7V}Z@xu$jWkFD~(jgsw|d;-11+x9#M
z0(9%Mw=G-%g(>&{5(~iHnh(f_V}EPJP&*_8{JMe_dcFTGLU;lD)$iW*P_OF}yEndm
zBybG*Fl8MR3LMX$bbU-YQ7_M2xzI>W&kjs6NsMOxfmWbH9{a`YC3_xS)ri2iV?Hhs
zr^)>M5%|_hb9As_fQd-R&9*q@aMAX>N4?i#$mJKuk=KQC{J)i#m2{#nkib(prKT+H
ziIG$#IKctcPa^@#i%8@Cc+RLt;fF<vRY$Gn$1}6%8i~3ZQB}-UXhoPk8u@nFOwQ_K
zBspkbGJL)HrWG3+h<ZXPt@rA_z>eL@GS&>{2WXnXJ|vu1AO%Hfxb<c`YYBLh^2RDF
z!r!I))_of)E<rm6{`=jbpIMub%-=HnJ+FlHGr!4iN%n#jANb~lXy-$;;^FG8KO1;l
z&5bYo>RHtr+XoKWk)(n0V2V0Y^#k)L(NEjee44+`T4K+8yk&B^kP%7Z)neu;bo%}@
zd_RO3$3}RMBLX2Ds2~pMwfp5OB=Sk_7U)8!kpHkkC~sb)rqGF2dtG(qUgr+<*G@(0
z+~XKJ$XboR7|hGZjf)y8&C974?R0AfdAf1_?JW4a>QDjKZ`AwLkVQ%^_V`2zjey(o
zU|j=9{+myTr4Sf{6T`}%4wM5kh{<myXAo-98u8H{o`6ka#M4{2sBRY#Ie2K*;-xnj
zY>5He%_iemRvHgOfE_+~x5At7+AESwhWCz;1^EH01-M1C|EF@q0MRmz26`h1=x%gI
z>*0pnHt*f?BQuKr3U#V$_iTsPY5SsNV+l*X3Rx59>Mu~}3tPAG>FbNjBaAAtAkE)s
zPL@bgwQ4-BTbS+T50khS?mT&@z$Xy%i)SkD=-Tpa-m~v&514ZVL$yg}YevR!(qp#5
zBLiAWd7<4$7!#(Da0A*Zb)N0+3~Gf$k<XbRv$6ci|7FI2c{Jjd1n`UgffRwvv+uja
zce?i<AKpf#KdrsBz73`lxD$VplSe91MS0~p+MPa~>j@}S;Vvxq{lMmNz2ARhG<DA_
z!mXE7u4X9yS(vKg(CgE_m!-r!a|hcrU*8sX3dM%f0*p#Bx#K>u@u=~bJ}%8tVnwAF
zn=Ou-%i5x-4ua*cPP{C5x1x)x=>oZ3xwpdTr#l-{_<N$s@rEoQLUV}^{IG+=$klP>
z%U-b@lEt9@Uvjl?Vihg*<+p<95S78ZS0Rn(-q)G_a%|e2F0=f^{NenXuRvQ-A840a
zH$i`5oN&*t<RV-n#E9`!;&^$)SdBa#1w8A!Z4Bx6K}6KI%VZVE1^HeyziSTWms8@P
zy@$o+kR)|Qa~i)&OsA|Xupvu`OS2JuJdnqda8UvdUwP7ppWy<|_Lfdla!pJ9exXMR
z5xUPRaA*6T@2Ur^w<W3n1v}#e*qriicX~643QmzLtWv{8AJ=x@oVXSF(i@-e%a0d)
zwYFZ)esQj_nIPvrn`G0|;GMaJUS4mRMX7@*Ug#8MJnPrJp2P<LKQn|sUqm6IyIvmJ
zKhS|8SR}Y~bNcFu&hFbTdsdO+cBMHrhn|lKb%Mu*5DPVZOvIn;VS=BA;>*q1dUe<D
zGQTePcw{Y~CH+zT7OfNX5OA^zzQvE6jPP4JX{a@PQLL`zrhOhO46uY+lcLEjZt%y2
z>c9)Xa)=W;Qm3NJ;wA&nZ>V1I^ZFQt!5+{faGfu^K8<e#1=n5S$uc+O^oA!cZ^7&O
z_F8*Ld7g;^qV=tPe70~FH+`ooGG0|krsJ~w@-}FO|LRm8wTVZ6p_VZru_I4LC}zXF
zO#EL$0)eC9e!;-(!v~=Tc$4VU{UzKSSh_)APx=#lv+nQbZ-pHoBA!|<a^E|E;eu!z
z;dCS?QyhJQ`v=j_PVN7XjQcj_-iw<@Mx@F7b$>}~-b^0@&l?d#j8vgIDkp!XQU^p%
z@P<aT`9zfa8Oee*G>eBOi7HJ^f#)&3^CNmrA8NC$X+r`7>pNvETHj?NWyM@>qFYTL
zBbPXqg)YyA@4WYO!9x%aHN31|#4-y2KfL{sG@X$U{jLwK@+Ssp&Y7?4+mQV!ZpiE?
z(~EeBuJ+x^S*^65;+%z2ypo-sIc?o>inp~b|GzCir=i*mQxZkVhvA$Nlo#tUjM-j=
zDT6T0zk*eyFOCr_^C@Ugii3`vZO7@6nZ=nv>{cd2rAl<TZ=m-VB~z7GPwYQ|L0dz5
zN3(u%0)Cx4DJ?`B^gN$~rLRJY6pC+txK#$1)&=Xv6QgSRZVSG<yhJ#uaK1m)8Vf&8
z;IfIFEnj*ztg=Dqd8N3J=iRx6X9Smdf0ZKoR1H<LXY<I6dNFZ%#Iln~(22amX3+#o
zdD7py9&`~(XY^{QH9C0ROUB~sz(YZkx*E2cn+r2P^>~N?^XNt&&+9SP=9(d;0=L|K
zTi%4ay`E#X939y3>82^q@|;X!gz5?{Y@gxx%YQ3s=R-tV@01$fZ`RW6r*J94nT>vX
z%0`D=g~|PPDo#2|>sJEPVe-HKMz5@ij&7q^9)CHm1wJ#!`~CSgSYyOhErPi-c5wIM
zGmQ_dr*QiLJDc+LHmPO&&Ep8U7tYH+DFpb}KYSeOnh~8)@UxU8*^ghXUJ-4=rVu&8
z8g)#VaQiS*JhKZ8toQElI%CqyI?+cyS_W?~;Fa4m4D!}6)rY@l+f(UWKlB`<WCZ^x
z&6PH_zUMsIe#{8J;OX<yqlGl}*UljON?!5Z15>=ChEm+Q`NOL#{<|vifkZs{n-n$r
z=cU~;uMNU<A0%(FbB>Sx`G1X9LUX)>AdU;~HutO#e+YPZ2(}<?vjyC~w*hlVgTEuY
z=7ikep77`2=@ANOv+Yi!psBy_-Xlcp&qEx@T2!@?#<vBpNhz=RU}$Nic85<E8aOU$
zgMy;E14S89AN(!fL`3%$h7_jkTRI%keWyADy7%w#f7T(STxj$B{7JFge{0!1`-tU8
zD0!7Bx<bT6(QRjB<mEUIaX(^T(qF|6-4j_UbAG0)(W-??EK)j3KK>H8DoWrdg-0ed
zD{R;tK|H=#qMNquWdmH>tB!=myBd!MAB+M)?XuaD4rQ=^*<Ew@y94jHmh57efSZWf
z3Brr;gNmW)TsJ<8x_^wGI1obW#CIq(H<;4Cr?)vwRSS3G$2J;vR;?BUkd5U0?U$Dj
zUPrwzS7vl%;M{g>FjTAf_XQX@vc|hr^_*uM_tfV%6{p*lT++?vRp7*fG9tapQ`(K2
zpB3@fr<|P;xenZ;(X{0Ki98Ne(<r+~0x!n{>Y%ayYYEm$N0}QkYjO>UB`eXbpGLT0
zcj5b?|AaLLHZf+XKJkOR7OE8%M#SaI{SUhW{pX1K+>dDj(sh}<_f%3<RS`Be9%6^!
zqsbY;1dm!}Ai76c`B13ClMHZD9lY3Z3B`o6?FGCan8s~MDzC!>&+h`(F<R}|5$1XT
z6VsDe+>LDeagk9lCBQZAURbB{CUs#E;s|?S_c$FSu1(<Y4R7}m;QI^pwS0gpw-CA0
zQw=`ALkVFI90&KwV^dXKFVFx05PIc?@4RiQR2|n%d-t5??4qySy$As{V~Ctz^|`sB
z{zENT?Zku_YRyCFQ5e-*+TDD@5o+=}huQ9_tKjP^61|l>!16Z=Af+Cug0fe-^y!Ls
zfaA7%k|Na9Z#m!Sa{nS0;0p!adr#htn_v6cZ`@>o2y(>BvBwQp&1RJ^T+tJ09a}{z
ziZn;*;#&Q6+muSgs!D{9U*&eYCICW#7$#x@;;Wu$LJ~4+XkQJ18X2qfSKfhdoUW84
z(OX<X$Vs2w(d_n}*8Tg^6bI<a8JwZnp=QPp6ydRK(8QuW64&5^XJ&R?R1Q3Gtf(Ep
zD!n(81X@-Qb$>NxY#VxL8D>~_sZ5oyFho-a_)R0PB|JIp=JsQdgH*>Kh|rQE%G!)F
z<fT_xJ*Aa4`-AW;t6BKazbRR#$Q_<o9$h)Ixw>?oJ~g96*cAY`#?HtZA7O8PNf7#S
z)-VlQKmEDRx<5IZr1j*+H#ms*_3ETw4s=~Z*Y(lK^A804ThuH^thLDjN<8b@C$iVK
z4g(%Cn5&KUu6h|E9b8fD+QYNB53BmJwUz6V!aZJb@C;RWBH{Cj(Rz{(yIkYoTmec3
zD_p7s&6xeIc4VF}!$^Sd-Pbc<gNIP&xq?Al_8q=?tITF|f4<GzvuR}}>8GEX<sdSC
zp0wY&=^CFj1TSTs4Uj_~$bL~UxFs+8Npkl4V%;Ryn2dqrnB4Q37|g!k?eE#9v6~<6
zBEL!D@z2Tw_Z4B|t4FIA2Je*nN|LNt+gQJm99(~&;}u3TvyreG*gQy3{q12Jh^dpu
zhnAC?Eka!Xl@1h_1!XKJ)msd;Ey&GXSBUKhxA4U%LaYB#Chra{g_bKq@BV$9h8ZO#
zGG#diCOplqzwKldUzxICnV!-v|MiHJ+Hf;nEnDZ)9BUcv#&z1klVc*prpG7H8&xnS
z*6p2EG$%`x_1R%ez_h%_o~o!Q+9wC+EGimsAf4^vl45mb@)GI28iv$qRm187eFSk9
zSGtJ1fxfoy2GK$P0xYo2c6Q!JBj;;tiToFx=i{)9jKvjI)Ybcqd*706XR=(k)(;nj
zeq<~<`yE4ruZzdW4qcFeqhn*{{_moCi-gi=Q-Y2Ua1&!=;x7JBTmkfk-_Gu`G$(Lw
zY>a-AeCpiVIRHMk)Ko=+cf!rc4p>*4KB;E;>tImHmtdSA3^q4V%ICvioEDQQD|>31
zyvv;fl)^!S!8z|v*Dt;kRi4Tmlm#eVA-GNESOIY_=8Uc5UOX)?U#chzBc}ubKy{f3
z%h9)(tyXRbniKD90w?2AFxNJ?C}%-m?=2EuE?iRoZsf3vZteAzsj(|nn&<Obb28Q?
z{I+i<B_#b<mX}snh(%nuBEb97_tIBx>oIb8X5gIdof16(i{_5;dJkkP&nzLk`>!y9
zUzW{#ETQ$9{cPbPSX{;gjHA56ea+H$=8=7MZ=PcNX{}<ZD0Faw5@Lc8u}b>lO6Ycq
zJdMit2Fwn|KejKE+W1gzQJZ_~g)8|M-7CWK%5~b};jo_ib#KNv^YwJtv}n)C`iENl
z;E6Q&co$>YDWVNxd(Pojq4NM-vyGse;duWc`8ZZK%SU|gRiD%hGAho>_0{^11mOMo
zT;M^|w?+{-tM#r`ZlnB9na?du)|V~seVt0z0(lbmWoHRrQfO}fHV}KZKQ!a=iPRPP
zkcORr`{@C}8_p_?1oW`c+->I2QvyZ7Nd9d7<tN2^eYQX5Tc#eCF?<<I-1eZ|EPV2G
z7oVwWM%k}1CwcxDLlnxGg3^+w)C9vkY*TOjCw@wBtqhLMyM1eil3W*N9bGBs0>3FF
zXsCP@h|GLI)sj9j802w(TH)f4nVJFgbiFGhLrYLf6R6#D_!zYSW0oMX^g-q2sO<?f
zAM@a_Ak-wc)C5^e=P`m=-NGtpFfr_%j~=rp4C650N<u?}MQ0I>&v#kaS0jgd?BU+p
z-CKS+h@6`~E>Y3<H>p4)<Z9#6nD|#6(EON5CkJE3qE8n><Ev0b02=v`X-69JO|!hN
z3ES;WSMFkHH&<17{x&*brmoSMjg~c_V5wrJ&Wmk8)C@HlU}!r>?`~`uFJsunQkEKU
zQt3T<w;&J1MO-^O?O*MbZ}_v(<fgNQxUva{sd&O1@_*ay6MiR~#*6U-lJty=#n4*o
zr_!tlGzm#=in;tF>bJ$4Iy}TXWue8HNMgdo=}&)&J>46d9Uu=?7=f@d3V!-Z--Yre
zke#?>A*q#_eS@vlpeD=dYCso<72-P6>ijTjE$Lae0ne{iJzAFK<?=33jsXoP1};(W
zCKuV0EUxSw3$UB+QKMDJ0w{?qgd5mM!ajh&n4}D9U(}j6e`|G<lWD?}@f(YBpwA1u
z(Qaceoy-T$0RkFuidRL`Pv*ghJQRhvm{+3^UjPA2fe|ar_~TB#x2xH->CMNGlgSTM
zI!}c?>e!wWyyR}7H)io^c}jIKhw6JZn60&k|7K^f!Op35O&8=s_TT|MrrbvcczyN-
zvK(=(=MpO8(y(x9*(<6si)+CvF{`UkeY+N&_0_@Gi98qE3vxpbS$2$<S1o+Db&rq2
z9Ig>OtBI;L`O%6m30Lndppm<BjP-k(Dr{%HAC3-gGD-FcFiIZ4ewcZMJrsNN0P5m5
z8JfJR8y6dub#Uug_PYCgD>Fepc{3f^zelZ$_fl`lqGVe-#r{B{1!dK;B@fP@gSDA{
zDmZ9tCC$HOzY|Ty*<JZ_?v%$(M%*(@6!y61Z);OZG9`#ajaMfJ9pEuPNcBvE$MV+?
zF4shbU;`1hx(BKDg~XC7(Eh<vdi$+u5I7#QaXyic_kO)?$G2Fd84Tu<bGmSC$<yDj
zY_$K=?D{>vQAR^<P{7uWk$`za{!LN&ooh4{`zr^x2{U~uTHh~O{p{+ePp0U}My$@y
zjK!tIV;Wi1gb;XS8NDx$f8hGD3)};S%go5=Y(rP$Tl12K9LxxMyhsea{C+@qJDSqf
zxIbdvvOzikG4txn6LL3ZR8cdecv(?bHVs!TJ8Cq!%PR3CZp@@|vUyQ*_orcsY_iBv
z>u@SqsIbY}Ztk78LV4YkrHx^`Ek%k}k4G2+NO}-h?G6L)LOCv>Lg7oN-|iAP;D?L<
zOc;YE&W&+pQpQ+1>CbEghG%033y+_YaCwks=hC<mN;;$s&Ms~i2f5PdNq-?qNC<gY
z3$#Kw<VKR@80Y(XCq-kJKaz8cyj^{3?C40+)pj7Aa<1V!a4x~$FTXH(STS|=HI7mE
zC9}E!b5S?B6ovGuoQCr!d1<iKHRz*D1zq)NG~o*N_XNZ#QqaujHR!>XQN4BWC%1eF
zd73bM^pwaVao?YUDv8?ZsGg=6_$EQO^8(w~a+piib^Iv@e^*j=g3vuj<|99`F^aH@
za<|tb9}bydoZ1A!J@kg~sOw^{<mp!HPO42C^T_S9CqkwPzoN9zFKx_1Oz#9(z<bvQ
z!ivha4E>mV%Ql}-Oeu*n+H-|Z=Qy#?HCjN6Ll2%#sIxxj2`w`Ab2aufb$zVl5=s)0
z@V#&1xQS(orMVVMlu2yDY*ZP2isqMMO=B>01n}s|hxaNNsxFbMg<rg?$U2aD>lDN`
z`y`s3?-Pk|;6L7y3o4Lg8w;kVmn6}q^SD?bs$<<aQU<0Na*}*HnSaA<O33+$#(iXy
zvXHfe*VO#z3%OGJW0WaBE29(%`6rTtCsj(XhQ7L4p!s2bWnbOFTHYE!%J7u3m{pZJ
zOJ;rPP0BliRgmSuY<F0P^ybk1X}bj~2W9FAEs`ozK2OI@4Va0Fw%4tr;9RoY8C7&a
zYe!^>(g&nHw+#UsNAVo}3_|__U=T(4PZQG7G*gE4P8YublbQLkvBN?0K>XN%y-34!
zKnpIeEpUY##4zY-d`3o)U4LR|^(FGXA*BA1X&IEcO*B$m<m}3kk^ks4lZEBM_)-e<
z2vH8LCK%`IF!#B?MS6O6x}(8K;?d3cOZSLwvTZcVBn4D{v}q7MZqa!7*7Rx6;a;qv
zZIYR+&<#g<86WK(M&jK6TebsTklCHJAm0U6K!?04+4BQ}%ymHFyc7_~6ymvNO2jII
z^)#2$9*lUX;1EzC_L2uwNnf!qa#z)f1@fK!ULXB(DZ!S0@5E~2m$hKH&yoYbK1@PJ
z$SU{jjZve%^joLYTUV=4n1D_>#<RoWt03RpMZb^CU7bUZI{jnMQ<cBeHTsZ!(=3r!
zI%CNTPqN?s@kvut3Rvg-cF?4msEO&Z{?r&4xuogUfqD^aqiWf(Sz0g0;;_)W@d_IH
z>D?a&hzSK5LC;Cw;3Cn+*93duwx}kAT3x_uD=*yc3FJHfRJ?ILG-q1=A<=7|)_a%|
zO(Zg!tsda+)v9p+ON2T5gT%Nm6Us`I3Suo3QD#Yqc<y00ha(+T0rLX|M0=I7Xsw0S
zabEA=@q3k+&15gY=#x#KK1vX?Z&}_jkcoK-38$&gQT(fg@h?Gp@8@nkvbm$Mr0O-=
zJLy%k;UoO)Pmaz=z-_$DZ3#>1x&3D%_2t|SW^IRkYu^<|>5-+Wmd-%8UlPFcF_{nr
z<BsHm#;ngwH_GA^7@ONYZq$c&Cu1<5kT~tJf+aUX5Uo}2#W!w>v+i?e&5!h=e~8zy
zrJ9&~1EGgW#f2PoO^Ac-3l2yq_>vrXeT`yv%LKW?0*0iH{KvYMP`9|hLy&-@c(tPI
z0C&aAcv#RN)tBDJW_;_M3&O1)X{iAoLNKV8pC$7r53Z4)S1;iYl2^TKJ`FbEZ~lRB
zC^swzv4*?oa>OEK5kozi)z@AWm(0UgIhPZl_NpW6`3`1S)+zM8w3LhMd+|;)8l_j_
z;U)JskiR)d4a;q5DicADrrRzbSPSo~(5eKUILsMxZI#m*l38ObydQp{pEkG2DE%6L
z7V_kOt5Yr{j<5cMVV31V&XBbH-#z(gngm!<ye3l!%?koOZ^HQDVG{yd(sJd)446?-
zi|V2D0=A9v$3I^UC{7kOn)%I%pjLBRk!Oo4iZFXplOYfhJ~)atg|K)FL4O)JA?AJ=
zC!HBv#x>~)2nO)|%G!m2|C}2v)~=eOm=C|z-JRf*zw+aKxWRV8E{-hjP4d+}8r;m3
zQ$=UPgEVGKPI$aas$Q*C>_jkIFbjXIbw_kUeaq6qqn>N~0ixo*j1^LoC02hcmYpDc
z>(pxIkL8K0g>0W!Eav!KZM}%JGEpL>AD+?vaRPcQD5~YpD;;u>q!UD2`-zg&Wp`Fo
zF8<s3nP~=8zB?@Ll<ltBFUjXZHnHK6TFNv(!FmTT>K^Mo2rt1^p#c=Ox0#b}r%VJQ
z0)x$m;~beuymbUaPp72j{JF)!Ml@sRA=r7cj}lK(MSo-DX82`IWE_t<j(t+e{0Xz7
z%Oohocihv1kq2E`;{eY0rERJEi(%_|0bicoJ(qlU`_DU+DKC9d3nW^+_~@+w6=wqX
zO(W%j+1-tiK=P^d_$PO^y?!N8TkR6=26MiWog<jU$IBAdeNSjCP2Rqv*NbBH!hEzC
ztnLh+85`^J-en50z(C}k?d*W<m*jTh6IM&JQYP6qe)t8>2<|$EK?cvu9vQ*UDnfa^
z0(HI+L9foC4t93X^ESkR%g`pa6}n-F?(*-Iy|I5cf#2C;HHFeM7F{}b5V%>~Kl#%G
zq`&uwOJILFA=W<U+7{oV%PW!?p82C8d@d(YiNINW=MAgI3fb9-x1*Wmk=T6vq<|f6
zW`1qH?@Qkc$CUhgXcuJqCI(DiCah1MzdvH_@|dM{j(;UL4{_lE3}dP@N%fo9MLIW&
z*ph#vLcc3^uFT^gx2GX~+2hgoe*lO7d{8)X*pl2gw22fr%Dm&s{yyaGA2MPB7)FdU
z*XBCs2NE1(wf=F?>*VSkT>&eO-r2isld7Ke>utW3twqb<+Mgi52?sPhmEI%8syS)t
z^y|BTGr-`Fb6>!(uV`}9kj*@AZO|@j``5Zt1QQbZS)E}$RN{e-^m9v&{vlmzj)CEi
zQL_0{kY3oW7AJAYf?zg)MM)E9Y0AB{rM3C`#;y5fpbSNL_lacogt4_es^g!*iNiy^
zwddC-q0fubCC`%v2_Q{!1dTFlS53<|*uktnWF;})bu5E^dk~yipXsT94y(#`i0a|2
z#xo)(M;Y<D2LmH9wz%K*)gzBMSWXPwx82)Y<+AzjrPlXddUdS;=ZJD5sW(dwDL?`?
zp;1cKSqirr6hUdHGYT`vmS;aMdEv5}Qt?m-8NWHvaLwRyG;+9j{{8W&Ld`$PMQS4c
zc}?!*<v*rbC25r0$(MKEEj?QAIoHM{-DzfX({q<^6)e>zQxKVho}fZ03qQrI<k*yk
zjK#&$?4Q%KI7AElwvB#5^%3h1eSAIfEnosxC4(=*9N%FOPDiMfJ5cQhs;{9sTKPoU
zC14EGRVFY|fkqc(d#FGCg}K-`qr>l3<|z1h&Z%E8_D8^qk<qQcE6b2~OYO1f<+LpV
zaLD3E-&0S~AqxcWwPj>H_?O1GN>TUb7#V=32;R^PzD?`e$E07v!E2Wl%b!527Ij^F
zU6o_0Dg9QB8I96&UzGp6tZ(u&2e)q@<ePyi12I(@pz;_5#->=LoNqaKVzj~(1#$y?
zRhEbFqx~vF%M2HDVYfHe`+wCJ2^$Z{{)o@<{&yf92!76(0fMr`Q30^|UkAMR-0UyG
z2P)!#p`J~uFZzigGYO&l$p`fW1Ile}xwl|Ho*Jj<*JD~`4_lk+maX#Sx#uvF+pE37
z!8hMIh+P7tY<*c|7s{V-drusOADm1Ak*{ln5m!t!XvMfP3s$x*-lX~ag&rbL&FPM!
zH={3#vd%eH-heGqfMedH=V!9lV||u`FYcX0VI&4nVlXMhPeg$OkcETA<z<|Yd0Jn|
z(q4<ai2h5Z=x1`lw<&u%$xTDWn}Hu!(qV3fzuklDS?*|9C=SdUGd%w<7hujYVE}|X
zq7YZ=c~<?%f->W#_DD8n;bzk%YDk|u7LXO*_YmSm9*}h*3F7m8_3kdT_{VM^3dQ8{
z1_z%n`zQ6&YNM4ty6BqY>qwZ0%wGdXqR-t^G&{%!uOWoa5AZKKm^{b5J)~lVJ*|V{
zmNxHG-+Wo-PlTK{TMgHv^dJQjA5Br3euV4&Nd`s56ceHOw5)Nk+l1^<Tp|v8*a7Zx
zpGds3aFFBss*!NgK4B1dtUS-)`#12~XW!o3tx3*QB%w{`US5m)0M*@O)`G8Jj<<6B
z*4X}9fP107Cp7V#9k*>9aNO$mW@K|}mt|#JU1+D(oQSm4WnhrYfN^y7yh!c2BV=*D
z(a><z%h*KFmds#7sd3X!6b2t|6L)MLU<s5Jzz%Khp_Zhyk!l=Z7kriEaxVWboOFr&
zkoTwrW!>McFXQHIqRVn42Eoxn<Pz@0ywCbVT!><QX!F2RsZ^b%FJCPA!|3H~k?K)*
zRBiXp%4C35{XC2}2W)&?p{)G==<x&0dDw?m4*6l^GiRlM)X`Ym3Xyq4@b^>c{aQ6M
zkGMgnn7_@_6F~_Mi+QAT{(%7}HBwYRIn#@?9v$9T3i4m_5r;4fR@XF8-1Pk#y)^uY
z591m|a)*iPUJ|82Dgaq>H!riqu)~hd#i!T36QfBLKc_9SeS;;q%m(;ZS>1r{)_aba
z%nNbCso0#fY!6>0$*1xH@`L&}xtr5d7sX|2G5|)H{EXDBsvq(c5uv5+l!hSmI}d%+
zIO26bpV)}?(^rRZS6YwMSLR@_!8<NUdsd=^my^^UD-)L)0ZWa3V!vLzwX|aS-$XTe
zu*A?^-9_)X&+4tu>La>Bo_ZQ@MS~OkPByh)n@C}dLMx(^*dgxkRf!e+-?|<~xR*-E
zJXsUH2{@mZb6*`B3*L;KXoJh?eGBfP+M)A1VI|nD&&3gP7i*KwG$imJw|6a(H9wf>
z`+a|;HaJ+qP%}23a%IW04_#g@sds`h_df0^h91mPfVX*#9<6}Vg&9kkSDPqYik(GS
zfojU4<rVMHh~*Sed)+;X)lBl|d6&(x2n}g<zM~P`%QDAu6Xx<r<(#Cprm~qj|0`1N
z+?h+p&a&HJ=jICidAls{Fo4+q0UYNRDEj$C9cB0T4jT3Ch`!YmX~3Wt#GI@icTC90
z9^u1E{9!r%nE1)n%U|=l=K5Z2-Wmaa8t3W3Muw?t`d+2meyD$GCQE))1!*EP3eqjv
z4MUeNk=NU5*~PC}B-MK_756ee_(f9Jk~7ci2=C5C3%WeUr$&C;?6MM_xe0Vxg&Xnw
zuHDp$H#9hB_a2&kglY@SVHRDA9kd|VU0C?O@x!}{s8$|X(YB^kj69qipOv<=n@ifc
zT%%%nknQaE_Y`4^&VY|Sg<<lrK3M}EPPQcI9&c=q%o;E|<y>nJ<s<9;^2wFd>hk3U
z_zBSaMBTh{F`cn@Du87BbhDWYopX_p_+5YIo%Wtu1w#;}A2tyYu;kaL#FC8ZJ{fw*
z-c)gmEs%Y&vbacL6m<L4zzq48lJH!{P2)r|V86Io#AlrhDsB6CW3MOLGF>;mA^a^f
z=VzUVp+NH575EuG6!5^_-uj5_0e@l?g+{1lN4$2{+ib9HDxG*#fwuC7M&Z_LJW6d;
zy3;o?SC)}6?)}eH?mrj_a`pQ}f1O9qT@x3c?#q89wa{%?Ben{6%v~+5nNv}>)-4b9
zVA$1F&lW?%mrpKpXhfbD__dx-S^$r)4wg9(w7C|dODM!HCSds{(}~3eRW;YM)K-m)
z-D$l0bSGO_w7$CZ$KS!=3CDW^fjt*wH^8$^HH<92a2T0VV5@&aDXFt5BC-9?IgNvu
z=9_L`<H=|}Y&QhK#{7nRm4EBrZW|nQj2N0;U~*@82*m$ePjEie$a`Nb8I=Tx?s)_&
zM<+7`t}WShU#lO*T}J+EQ{gUl5qbrWsboWUj6H$9AX;$yF0biTv+!}>*5@b{{|jVY
zFdJCNjdsc4qQ$eL_X1;H?|EqopP2Qh4lS$p#AQu<4CamMR+ldSfN>Dmp_+5D3oXa}
z8O+w1{UaeKQ`h!ytn5Uk&rLmh%HP1zkpVs<*U-$MbYR9Z|1|)+dgn-8daWv$2Oq9c
z;LD~SJrFHpb2j;*)qX2H-8_3f>cxIQ*|F{1bEbsP6;S!LPP@EDXUGaA$87#vy1I<|
zY>Tmfz0?z5Vpy$RS=L{@X#&cBcG~5pRs%@0Tq7jC<K6!8#EDjYyhOxD7ICM`${ja<
zlufw%S{Gu+T13%oQ?og=y7i|6stqe8*nPh1oRwJoNA99P>Tgsk>ELU#Fls!DHh(+y
zg5l9q*<V(s`EyE!?#}Dawq<;?MH9D`FM7KIk}=GvKtyVlToY8rMY{@}Y<C(dHeg%9
ziM~3IyAT;cegtAhHnRn@J^U#)X5>8~K0fnQ`5B^E)8D<@lHE5Uy(&${yWJ}7QGV!v
zWRudWjQTij1CLomsOO!taX20SQazUn`^e$bp;V`|$E|Dic_ewzurF)6SIe#7ndP9Z
z8<C&&cH(7+zG_iXuR%@QNC*l@ziH`Qt@vE>Me*k94tM{(M>k$9_t?@?w0}SUOe{Z@
zUw7#pA%Pdx?IVI$w#gaBD-;3$GfE#q`aX%)zjJ;e4UbYhKu4{lz#(*ZMvcnum|R?K
zT7$E|X_r1u5@jx{(iP(a!rgDFV`V)0hqkCz)=skexN1cP{SFQ}XXVY-f99a!_=R_D
zGzVYY^_tm6WC3*BRjU1VQEY2T-&qAF%#QxN73gSKm(GtE)@#qn;!8~#j>=d`6`7TQ
z%`<AtqPS``OYtV86q9y3_W~WR;LB_ShIDWjHF?c#37HYEXj@=3v&M_22X25WhYmPq
z8u!7@4u%PLwARr^$5S7DKKtDMnk?>mRcnZ3F3z|#ftVR~eRBWQ*u$wKjreiNLH79=
z*!UG#TeEzk@u6j7bydI^a=-}Oi+^L>&VUCgSj&|*A%o*LyM+RAo5>{LjT=_7KcVx0
z9k<aJcWR7XFx0~#opKB;&W}gqOde4!U!bH-ntS}`yr#HGqkSt2e2g%1(2F?uzjFcq
ziu0R#pm=?4@PKDCQ`(+^G;sm+4~u1>$Dx!^lGFvqoIQ{7jaz}iPeRN$jNHL0T9dx7
zrK=NXtr@kF@2S8JmgQwRcJh65v+9DC$0NF@Iv9<`wLcbfvgenIUaZ+RzAah4SUIf3
z&P89#*^lF-t6tX;KLu*VXGkax|36-BzxWS=c5@2F&*3z>bbr7MhEz0GL?Zn%Uv3GF
z4_P!KzaoC^&_ul3iAE>Fri2%ztvbVq(SE0g*V2PP%Q-Hc`jU;}(6$Pw^ceS|wG0FF
z<|h=t26n>mSGzGk)Fkfp|JT%c#<ShGVV@d7YZX;BOVzAdii8-gd6BBUi9Kpm#g<w%
zYPM$WRW)kws!)5>tXTv>O6(N`d2)T8`+4p=ul}#{>inGJdmQKa!;ZK43^z4JWbXsh
z^AWoglUId~OvcE0C&~I7^Xd~W*(i>mBW@QLVR@WY!z}iaZf-#38iC?`3tUU^jzG1N
zq`Z{wN&SCl*#jI;P4e_iH34i~m5$*Ke{6Zbb~_EZUm*HpEu(Y{Q~nRcp0EE9)v$U5
z35mKgoIBka51<<2Ul$wwu74^IGxwHHIXTi<g3s!hheo_3m}lOS>@JS}?Ua;slF))H
zR`;|iDZ)~}iHXvJW~aWe?+nZjS|o2u=F+3yk)Ql?bmi<(6QR3|Nx7y5gkyMVh!EZ0
z;Y<m2NJTTG)I<87$A_dg!JH@8IE!hND(-WRdm;2WF5EQ<gece!L)DH~#f-#{w{WPX
zf`l$PAKe30DyBG%(e{&2jWNB^y1Z$qcz%!OnR<<FCPapYx2J|jr%yOHV)R)8J1(#<
z(b$&1vgFE<?h^xtF|@vOdJZ7)>Q{7HLmxV@gif6+<_R4&o}J2#{TpWBvvayS=jRQl
zLS!jbNgs*3<|1tXhgg;0`6o*H*q|%Q$-EQ)wS;WLVeXl9g%VSzs&Y;ImXXmyWu@Z`
zPwpn8X2Ju=roH5;<k4Y8T~{fuS5E*fZTn?MKR#c&AmfA!8O7(K*;1<m;Sw>d8V*qw
z{P+&a#eWF6-PB6+H$2W~2uB3LB^?5`n4nb3lI&DjH{Y8;@FbHN;4UZ{-OM=_OeN8Y
zv761yHUr!Qb9t1(3kT-x^n-WqU9a!Yla2+RD4kq$z}vu^>ewS@cJWIbc&qkW)MBB|
zaH$RnNBMxqs>(O_fD+DXfDVDb)qa6r@)|ob<4s?49#0IYI&40?8FJ{QHqOP)b{%$(
z2%63MJsFi%_RN1<BKwdO9r0F*LXjZm{#y{Q#WRG+6WomHO`ovqNgLXdc3J@jkNb-_
zzpR=1aFIw1A&fS3SO5hK;)^|dqnak(deJ!*1v`w5wpo)2jPZ_fIg+QzzKtp%u6y!2
zJYx0YuFT&njOBNG7`~8r68+>p$DPk{=bkMn2oso;3%;gbW;_HQnHrF>T_43)=|FeP
z-80(a#&*t}wGg`IHR`0MwZlaFR$F{fa-r;rOMhE=0w{^ErEX_`-$T~1aQ2KnkF=@J
z|F5&(f`H)&+Ed>2^5L~YS4s3-#a7!5RdcwDU9H$|lxyp~E{2NR?;TnZV|G5Euhz7#
z7Ia!8jVwogXPBdw0q*<<Tx>*7DlZsu{UrVMERK=qfhQ6HF&Hnp`KMge0Yskoa(TO=
zPI~J_#kU!bEke_Z=uZ_P&RzRR$)odDEAA}S+tx=Z_pyGh7e1H?RB{%d$=#K>kde@c
zzbzIeOp&iFA++rSVS>qGab8f41cp*_>EO{1gvUr{c$}>oSQEk&GMvE6*rDtwx2{3C
z2-|$5fjRB-@~dlSHa1XD$A)j~nWATa^%Tax-`+Tps~2$hu1(H9UD0efy$IESJ&tkx
zT%AH31E-}mC!yrCxkai(QpYrJ7LY5LMhhsvn<lrZY)3kN^8K#=iY&<_5tNj<XbFo$
zaUyniZmiBRmfu&V^01Z3rT8N~LgJi>{4-M~^ap8PUy#{&$y{lb)kuERkH|G?2+j1l
z?3Wg+-<2Ax%W&E!aIT-(-ehMq6DUvj&M+v>lZD2Ib7$`Y85IRo5O@49I=wl6?bJwJ
zzChkZ$=^zRAvIJGMcQo1z-_%HzxURY?Bh>sM|0n{lk+*^cw5B<dRIPG%M)Fwmue;7
z_6$RXZrnWk@90B>Kk}G2-9VcT62^XS>vqivHf2^l&c$FTf_ZapLO!GvI8$|o-Fbdk
ztk!6S4J_{-(K(4c9__OOWHSA*o<eDK#{VtjB;xktuhAS6+H9$Xz`~ntA@ekz^er?d
zq<%zW<*})3ugpF=zDKhQJ0A^v>I+N_OYXAag0U|WDTvB{*?GBBlHumbvAte)l_W!?
zn|cTUu><`=`p(NQCKBmu)f-vUWDE|u;b188M)Qt1K^~_pm+EY7BK5t@C++O3+z81;
zgHt1@pq4(SNo=4?oGE(IYDrP%H&XwbCyq!bI*-<VT=FS1=}Qb4{d(+FGtqsgc51CZ
zuzdc&>A212>NF$zP*$@RSvH<98%&21DOTHkUr!gEp76nx&E=W+-s1MxH(#0iKJ7W^
zkQ$ca4rII--$(;cQuE|;B7c|=TGS)Ek>OrJBi>fIJ{k`L2wb089M1GfgN?q=#!y`_
zJJES^E;L&+-!I#vnw+k}G5DuO7wQYxxB05&G_)Gx^1Y_Rzr%_#FmIK7!upCiObdnV
zescfu`ke?08sf1JDgsDDi&|_{l^?Eg`U%1OvH=cS=-<e)1g<cPwgB*$VIEf#`*5H)
z%*`R^a-7`3kiBwYeB{ex`7?NnXtB-DJe*_N@4WIl8v_%>{x!B=&ps=wYOJO~=Cz7s
zR))2?sl)E!_0oTd0iSiz(2lXwyZJek+JVam@rw5px}?a+2#G%>1rQZfIPVrWr$p&g
z;$H%#U<(96zTFy>53?ga*C}Q{v4wAa$$3!QkU(Sh=gk=V3?EPu1O=%bt)cZr16~1#
zcpsAQu!X<Q1<kQ(p^md$mpPAA=L_^WKO5hx^@AF4vGg3&p+A&HSpLi$XE;FSWfjP!
z!j87_bqQHEpx?`L^AP%a0UXIuodDslFY2fGNiVTnf;-lZuQ)EAAI;4>GuS?c!Ibwy
zGrT8BC=lf96ieZ16(;jrwd3H*_>cXJ1F7b`GyXHOoszt`H4sM>7txG-r2|MC+WS{g
z3~1N8Lf_1AJJ1_3A{c#n{G{dC&5%wo_tw+==Q2Wep}2Rr=!@qXk{H@?_*;P%Fy~Hc
zG~;A}Us~LLu$T9EMaxr&fp(*4`SY}kSuW!oKEIneo472*{-cO69}U>YcYr^PXmV1%
zquwZcuyIuX(Mu(&uTVAEL<6WpdkHBEi{o2n)dJkh!!j5}B9G;de|~-L#z9I(_Pj1~
zEg;ULmBZU$h#{?omcJ+7ls7#L<4JgzFT9u>EnqJj%DC^!rUd11!+CQ;a?V6`#6?3o
z6h|uKwwhBSlTZ0p`ap{?Qf!!v5V}z$V^*rSGd-en@`jZ*hZn9JAOzwpx^&S`7e<Gk
z%J96=)hhmawkJghHeBByLi$3wUt%0BPlE8S)&8(fc<Wy`JQlC{r@XBWewyn0&BI~0
zDDlsL`KpJL1G8wLp%Q-lPiym@6#wG@4jghk?oKK@Ea>kwDCO2l(+Om`wc=3bZE@j$
zyMM^Ii)2NGS~Ah=xzlSg86r7_ha};}By{~`%3>V}nJ@0}KkA!L?j3yvWQ7@~Z?M$Y
znlGK^9QdJ2jp-^!8c2G*`5c11fB8z}`7Kw>Uw3IE3(V#DzS-^8xVxY0kVBv^fWWuJ
za{WCGJ`bT%>%S`CJO3PJ&)JCnQ{ZIS_p+Jh>Yb0jvUC*Ren0B4>n<(t(@WWROhS+B
z3*$0_u~pujmTiA)^G%Fbb&eXM)}-6tI;*B*F-xy-%ByaM{%dOXaUFSd$s(U`er6o#
z-**HV_7NwNkVfy2%TVw9?$?ZM3n}kT0F1c2V;v9p7<EKq>fl}4aNfdU&{sEcgFE8-
z`T*vzeHcx?V=KcM9M5|EOQ=>E1O`nGXmP$-kZ8IC4U6g!d=%@$OL=gR1faB;n^i_G
zi^?1Bv38WQ8^p>Nf&8tWjfYKYj@yBO6+S#^UhDA;LWw(JZ!C>H3RX(TzE7iUuZ{hB
z2D{lYoabn}hn>kX>Wh?p3^}w(q@aI5PJJ=x`?XGt4=a+O-+qUj@@*wH@W8CjqMCK}
zcv}SQM{<sf%En~8Lt1G4{+4=C*Mzh|9@p=Y87MPARl>P3j|9SW%7`I3r#>X<g0pQI
zA04TGDtkLV4hjxEY3L@Y{($+l?M>J6J&U*B*?GBo3}tUgRqNRz1L0?(N+w#hz2T%-
zg*K!`E;?M(EC3%+CI78xTOCiuZKDzi1@Y_`D<KBQu3T|hj$k_n2RIz_B{Or^;uMDT
zyXcfF8d`4ino)I_$@$?<7{%n#(Lej-tA8ENG@JW-triMVhE=+ME#JS&#<N=>Jw-Zi
z+VD31YVguk;ivy?(gX-?zqHddO88ynRcSmmRah5&%~6Yu!Fv~ORxVjpc340``}65%
zP~jt)fzsC)Zy1!{zjh2$V$2IPhJ99~@oV4d+^5uS*)E{@BQMf;EDlRB8tWwH2dY;-
zgAMW5I-`Ayk*F<;Bu6XO+|QRvuZbQK3Du}~cb}o={ELjlDsnc$fM-ntv~d|A654)i
z3*Z#3g=hNS-lmthNY5AVcoLB5QSB<f-My!5;lchcpXh<d$Ca)`gR{Lw%9QcPiN`y`
z8M3<zyfg8I+~S8yt>Iv8*DF9HTcEr)fG9EZRPGV2IZ`doWFLH!V)Uci7KIx$hPc=4
z#N;F%Z}pbRd{ZSd&v^MxA2)SaAX$Lg0^D090dWUw|C8OF-q^wOg_Vt{5v~Zn3Ou^}
zZc$E7wUq>DdxtJXtzKrwG2K4o>Z>~JLPal-*$-SrxGT9oLsKTc#3<CIscy*G8Ka@W
zI2`*1#2e3^J#La@8I#*^9H0<q@|Rbl^ODubWL?9o?$`{nqF_fF;WaR3kgI1)6B<Gr
z^~fUXV%UVamS*z!7IQQ2fMC*6@!wc<{43t$)_nBR?6S9=uYXfqYyTx$%HY@fKGkkm
z{ps(Ifla#CkY~$RDdoFc-gQ$gU`0RUea}1Gu3GGY=mMU4I~#F?sQMIg@+wz4YnO5?
zorQYCdB5%MiqkLKGY?T&<%3N=_PZ@2y`Aub$$#57<lE(X%8E*atVggU>q+%O=ZNWa
zt%>~%GH~qDVMbnNpSe;HC|$obm>cgF=ft*~SikVz0rvJ?Rl+7F<?Px2<zmLR@LSRn
z_JHRdh=E*zKJZX3LP705)fqgSuLFzCc}=#_FKz>V2O3!Feg({AmmV@k&FPo5M7*-P
zhj8Gd)nb5L&wF=tb3cxWk}S%2GJVS`-Mw?PHUX>)1}Qzb4|L4IZ+Y8?fPu)sSjt*G
z{FdXr-#Xx5K-XCMieUYR?d|-}m2(NE(BpV{HvCl+)Rq}NsKqp^Sn0V8cj~pp`_Xj2
zxesyj_1Fv7=Qrc-U4HqvTV;4*<+{f-R~^xh;(Z*=&5e0?*v+$Y_4i3b<)SP;AI8W*
zAwIW{An9*&zIyX3>}8N%f7#9Ze(e!|$)|h6n9U5KV=129L#5aPi4LwuTmVn1fdwZ%
zTG6>}7g|#!$y>^&qg=B3zswCdo+45H+K<_ZFY9G?a~~`P<a>4SMYwWrk(`vZ!OR29
zll!V9i`e?TgsWGc7N5n<AKffW5hYB<ak31&+k`yZRyl?5n3zt8(5W)Y408Vkgel^d
zGG-fJ@9E(|V_@|y8P~!cbc^cqZAZT}d!TLhdlQe6_M56cpSb(uA#`PRIc!mN_8mz7
zsyB!!_fducm729{`2gN$Mmea{pJeca)5&0lOX9X_<7!XIaUMkz7>w!(+yG6oRjF0%
ztcMK$t=dFG@dAXuI^8zG18+Hoy1>)NE|F2Q`&;|J=uJGhz9y>amnV2HRH8ZX)~6LA
z7Qg(S7~ZgY$cstV!~ITjGc&Kljs9h_>4-P|1RAuJ-G2#gJ{~tUoHy#&LAYWqbUgzM
zDZ(%0YwK2aE%bNryR843JFfh}?v~Nrk;gQ(U><jq|8o62WmbbVvcK{KBeDuJ*A{VH
zV)wBq!jX?L#VI8{nz-(mJ5706-ly4YfpSfQPO>i?Y2+rGA6qPo;HEuqIw8tUm$=I@
zZ#`M<tuq^`efcqgaHudbr<4AjcAh+~Fo?1@#JZy6u+z7kc0-_RSvB+E&ILk0{Pqh(
zohj14_W8U<URJ2t%6qiRo!gMDtmd2&N>-*`L6CX5wRTCnQA76O+i?#WNlbMCSv4kc
z{wh|fDDLZ_BJG-(f7|x}a5om3{7?CHDA&p%KtexYh9%kBT!Ej1h&3>!g{}baNrhv9
zK(&u&H_!n&U|svWiTWH`p8?SR9@RsNl~`ft<-`01a&GQ4c==>h>D=w>sE^ZZ0|2+(
zhvEVL3pMA{PoTetO#wC6EYeA370QcWMWS`cF!E0irQlqi1@#4dv;()W)T+Bwd$@=5
zd>YAb3X9LP#V*O1i<om8>3$YJp}i%OZgO+CR83sck+vqNWsLiAmFT$IfWZ@|{;EFz
zFJ_|vj>31Tel#0C+0>9cm?yq)k?Pg+X#&sS3{!L%EEO%~ql@?&`EiE6&tiHgnrJ9y
zn-djeJ3TwgD-WKSvL;SIPrwnx64Ju-Nl`IbN(x|1`f2y^8w+!x1+84L=DYN<lh}~g
zQC05}W-oNzB8$!@nDF0kp|^hcSgBP8(sf@4<AdgsJ3>K+s3K;9C++XtpNUL;TD2SB
zQn_<8LRxc092250_)RYuztP;`=->dp(J*4XQgg*O`&8CFL+qmA+m~<x-~7ZdD$w$_
zt5pk<hn@B9Xm%&a*(DIqt&e$Vg3ievHvAy(zfMVTueAQsy8<>+=w<i-&A4hJo%R}O
z54K0~D_LhkfNgap0222h=PAd2z1E&A9p)iCoBfe^hC9Fhg+EQ%-RUrks<|wXBW%Ub
z_`d%+WZ1Ufn;XKK=KcFm6Q6NS%y5qIY!|}=p@CHYCjG1j+cI>?UlZQ1{m$K|SQQ%?
z^r3>)vQZs<izz{A==M@U;3*nJ4<18z3MlRwtPCCIBSxl_Vzni=SpoKjM7e?bHd9`H
zpxg4G)=A$^l~E~xa;<P}R-Rn}Zyq$?Pt||9=rU(n?Zy*2g^HON84$Eya=F32$_Z4M
z?Sl)~YE*{247l2=TlMdKTI}`!B$V_ycH4EEJ~nfGL+w-%I`@}`-|Cv*c)>uNCcHs}
z`(6$-3kap%uMQM`qtQlm7)TfM8>#akUH7^Ncv7p%G*+bl<Y~jvX&UBTkl1LY+6=wH
z@&KlKW3aZ$@6KViZIJ2GD>Y~mbWiyCQ2WOxJ>F?P$rZ<52nmG-u#8c`G;Pmtycsu0
zmp$>~pYGIZweZ!jg4Hn@yGOT3S_e%ywE`?7?+P?RPrl2`I8Lp_QJQ=Z7VwEc+@|%N
zK+y4NkM&Ci-1N8P|H+4SG7Xr>dLcruzLJ~(b8W*~WH?Ca5^jw#&Ac$$&78cJsiI!V
z_3clHJ>%dA@IhP;$)BG!V90_F)B11V>mKL+3P<lbv1OEMjAvtGCVvF8W{^?kOSJ^c
zy6^QB#B8{7sjliRlf&akb(o-Jg`3d;NZsRq5*IP6L8Bn$<yYqJIwGR`a~jW;^aB}*
zVor<BeTz~3ZAA7vU>D*Qfr4mXEjM*&;m+JZ)}xuQI%MGFONS!LtY&cjgM7Ho&h6Jh
z>WGT(A&cG%#D+t8h}OqrR?#HD)B}C2ruNoU`THCPr#*zAh?0|>s}N*_a*U;)1cDyk
z4@fO_OfTKaPdJ8O{hgMTWc9wgx1V<Ub?`%^1%4(tg*YdN1;zh*YheiX99*c=i16Sg
zjhf(+PIOcXyf-e#tQ=Z;3#-C9|8i7WBl%RZGB1RDmdNhWLlRs6=pBItY@@Ygvt}aS
zk(X`gKz;oEDmdO+P19Q7OcRHlM%sT<N%{tur0M4w6W1`lF(q5Ishd)xq)+H4^Rkkd
zOq0jfzA`OA;-0%X0}}Z(q*c|hgKp)AHDz<cXf&3Y5!gHMDOYtW&&B!ZX>)NXKGt~3
z1AK5AU@`yHsXb34_igIT?(0)jc3-9`#k}4xtYof33E!ek&GztxBS@)crbzWY{AT+#
z0;xndk)Own$b(fN8OkcRJLG-}a({QIWORD*J=D!NaoDH-dXcy*_DUE|T(HtQ(ndiV
ziCR5MO7-G4D`^)=2>R`IzXU*rHIVbrhLP;shV*8>#K*)>wYQJ@hpdI722t}unS;`)
z2pyom<r@S?e(6KS%^Sl^i$=!mP}`jT*qiZvCVMaT$D_ZE5q-$J7&!d;s}r^>Vxb5Z
zn}Jiyg~j7@IykXEsPTbb4V%@+LbHTom`V%waB!$srcuDzd;G;;{lZU-j3^Y~cK7*N
zSJw)CS2#I2OlL<P7hmJ!WQbC4d|Ntfq8ysugiMed&abUM7gxB=@I=z}fAFozzu*Kp
zLgOvfY1m54+BAu6eXznVFs7RqVge*JZubMxOT^|RBSufH_tVV?i#W{<$KjFRkstf&
z6iVBc&6Mxn`_j&?zlwP+pjw<kx7$(((slzM#SF7MO2fkVjjmIUuI5cp;w$vpk2XM;
zt%Y;)Y+(ZxFTktnrgJ?`y6aPZKi`yq$Akmn_dP&ku4@WJTPug2KLnt;Z6dWx7d`g*
zU{A@>wK85ZNf!~}AxV;gP88DMJ5Q6=KNYmTp)SyxRIMs9ac8!~N1%JVzlJ=&9fGf9
zpmVxXA2X^s$)ANCzk*zCS|CHQdOLX=AJD&o#4Yuux_SKJOGv9pIYr3Y$W!ygw)0?q
zDVK-zdb=-nI{sRkbM`nHV2FDqKqFB{jMdT8ka(ITu2I@7&Kb}#4FDQYm>r$L`7P>0
zc^{?3>DtBCn%r!pVx`6ycX^PhvzLW};$h;PwErJnD@zn~<9ucCV~MhYzrNIOzll71
z=N_wZA#=_aYTw>_d63x8AV>?|_k`_3dE8bj^G{*E6zqDf1j1<ugZ9*{X-(niI*!44
z0Z&X~yMKnqyQC{Ib%s}SydJ%8@=hxssivzUqt|`(e&@?gI9=WH>c|~Erct3+o=ab1
zT`JtHJgt6LZS)udRzFW)-tyX}jhuTGPPQ($$4oFcXY8izj$?ruoRPT1kV6m$?{gc~
zSr(e0;G`GNlKPy!J<SN6lum%L#{}$uN1h>lBVSfs;C7QS6hW}=*^Nl%^dOLBAi)i;
z5WL|_h)p4$`UHcs?W}ad_cb<GO0quu)R$0~&#2}3#x7g%9UmWxJV_Y{_>XWPb>=wW
zB54oc|NSN2tcpqO9||IMC3I$q1o()}zh*%r=Tii;rlx@${lz!Z)wzi}m5!2c>uU|L
z5CPij*ds~ASqIb`;B<{c0|oFL@hDycXS3Nu<asXs#@5KuC&U9VWB^>)`XPExQoO>b
zxP{rjWWg&0SA|A<2sM$I<*`o<r!SHy>uztB=lQB~9K3Bl>70)dr5seeh{bY|>1tP1
zXB1aOsb!S(*x<XTi8B3L{)i@Yhv!aVYl>n@!q~F21nFs^rFPH8mn+{jiW*YwChjD@
zyf2(OGGNxn<B)rxc<X1_2kt~@T99LoV`Xo(&$Qt2Pw_qvrO&IhPenbJkNs8AYmI*r
zo)>5-;pZ=GvFqXz%Mj<NU^Q_y=$igSaT}ZZNLpBq-@St3uQnFhBCe94(zEP=$-e>L
zXEqMsfW;twZ9VOxl&h$amcHOGhrw3ST)D!HidEab4GfSAU>ih1*uJ(X6pOLg(6ah0
zYTG<jz{-^{{x$BQWS$+Oy{NpPArGZ!esC}`OJpObU9dlIpGTY25FF32TO<OS4TM`{
zQ=X`&Ocxcpuj)d-mm4K2|7|hNh0|iP-MID`_H6e6Nsne^#a#%mC5L0iAYEjW-d>#S
z_he|+ijB8c!zd!}@^XL;M0~o&pS(PPYn#|+1t@$3)1phOQF~N2O8$ptlkzqpim!||
zwSvjtnvS#QJ9PUzwBX`8VGy~m{%A3_s*CvOwB;*;lWP&gr%2%*{U?lrhOGh}=~ocV
zY683peN8D2X8iaRwVR#i8^OZcDe^I+oc;iSJSZoeH_Rd!X{N4&wz}bCBYoMe$eGj4
zs+%UlAc9p<BT|q<qs~1XmTL?x=KpwHBM5F<DI5!Do|y+hd)@wH`Du09qhk{hxWUv|
z@&DyhNmrf<9+ghFIHHKKd`p~HyBPo?0(I2~u%-ZCz#RJNr%q9r2Q$9@puSNTi^^-I
zS>MSqXFeU_jgsSzaK3v37o)D|_@hN**6%0gU|~qec8Z`I|Gs>1LV2Lt3;1Ini=&E(
z=X_~{CS`A!3>u<?G2Z3mYqcB$lE^ePOk0ZNr5IcP&VlvywZA~Cy@*PU^M>w@>XLtN
zWzk*`3yRiWc6W!;;>}WIbg+T~gr(=*91OUDbey@~BOXc`lE|O=NyC=(@te+oq*XJ`
znRlyRVmYkrOl3WESjBqj<A(SWkJXY@p6)%sB~vwp64XMR(^FD)mHq<bKH05U8b){q
zgZu-0|LY0=p_n4M`v<~c`(q8;%yfIbgGd&b_GmG=?_yFp^H*pg3u-99CWrQYY-9ze
z-LoBn8T(oI_d4THHs_^by)W<kKn>1xGT#6xclX^>Gzq4)$KOl(KSW70-u`S1+RYn3
z<od)F$aS0ZE`be30e*q@V!0LFE~AabFlxs=E{|B9)CQ>Yy0zEjyxOan4X7rJUc&Ub
z0&^tD;n_1x7YD|KvkY}f<t49zQ*eSV<;AA#f7w>VB|CV&*>;o$2MRtLl#AkdI_bT?
zSDwrI@=n9e%WYJLqEYCa!u8xFS^C{5@}i|URt!05ns8}+@X4KXk_}NV(vv7O*R@&<
z1KHe0>rq>)Dh<V5IM`L5?h7Nz%FQ(xoASEoZRG1~_C@AEM90eftsGvXznC4_TeSgy
zv4hLe+qval1?U0%MHmtfK4b4M47rf4Ryu&>E!F7!8-h2itQjVyyNjXA4YQ;fra@nv
zjQ*!*Bq8*S)F*lK#z|D9?}C}K-|3`$n;|9$slZAJRWk?L#N7yV+D%tNrMyL*)S;()
ztU>&}Xa3pQQ8~g3=pSE0Hj$qke~`T@V3%?myFU0m6g2&HiaJY=!-UXK#rSi4aU&V6
zZ>GI%1=Ni1&xPIaTdZnCb~XKa>pU-3Ys<1(&VBEXU<f>8lykc6BzI0hyi0nXnm@rg
zHxv8yJG&dy!V1L{om>T^!;p?qKpu>JxGw>`ViBM`3bvE%!49d8qwV5fR%c~Mh=13%
zu^chF_W*P!Ari5SJj*#No<YBd#-p|aj-nHXPW_cQ?i9)M#47T~6_9x9JE-N$^aaQR
z?+%;v*IQ=x;fBIwi*UsEo=O+AqhS1|nx!iHF{y8^$3ym0q6b2gNXrY8sTmEF%dSJ$
zO6x=vjYmr^`lb+1#<Ec@$kxOT0d^YD(@lT*qm|zHiFg0-PlV{@{$`T7_n^7*erpCn
z1&N{Sc32AA>H4)2N>rGU=b=^rc=_g)anwl^-~BO`J+M1gK0?aH_>ha+(<=g0Y${~C
z-U+8UlFF1Z=i=VyOpm^0BHQZn5aVv9{`399U>ot^pOk)`I||@DF}jAJypPxQz7)4~
z9pdfS(s6M(+UKG!ip6NRtECAaqehNIpq=c6vJ-ElDx>Q~pm!Rs%E+)=-X?&HV_RW+
z>pao??COfCcJCQ1$M#L-0#V*|l@Ok~hqC3(vit?&GPz=Tj~#ro6UVb&-GSQ2?}m#K
zRu4-M>jPb>$Y8<>0vua}CRAt6gi%v_hklE^uC0mo#ML%$nw?1@>4yKAagC(QU4YVK
zbYg+eKg*>~@8$mhX^P*Zh{(<KL54hbUuFwZL*B77plzZc>r}L;kI)xRYaYAZusz0p
zcKYK9O7DR@q+W0wyJU1<6D-@4+OSopI<BO4%1z5Zfqm2Kq+;=~j$}F6v4QW*wfFQC
zH*(g?@P*@ozt-z;04D&EF+QqU1=Pe|ETLLi<9Mii+?;^hd_d*8r%wI%DRm8dZMMo#
zHtqv@O5H%Hlbraq&DEjx!Q7j;y!H*MNMJDJ8|e&*I0)o=^kz53=T_;vxzi}aplk6_
z&nJ8Qk0HOp;uJj>PI(wvmmc#pJxMgljy9KE-IJDM3ke2MZ#GZcw{K*=EB+^m=@k?5
z6TBmsi`2gEq$OOUt3^WL;f0}~VOaK)ih0?6_CyxQ16tATLA5l>YpfUR-xf<KMsI@-
z66^N9>cInlbm%~g#WycrGeCHt%0EFGix22TWLP9WLy~CrN~>C)HaP02>mq&y2+RMh
zRsFFQ6p-s2V!b5ExtPe$1~mzNDkD?ytPs~|5U{z3u35<hfIO2bTUVC(VIeQM?~F)b
zl1;Y1T&$6I@|3c2MFAF({FlFcIIm6OIK~_|ol}yuE}7Zrq<9!<RlxaIf!=D_MW~*w
zRWGVSsoz2d2CESVY-}vMt5+AIxvbcUb`ZhVnk-s~#I#{1_o8msh?SW>3xEd(6JVV~
zY0hw62VcSm9qfw5n8>-6nWfZ0ddMPu_`Jsw0qNnl+r2aQ;Lt9DWF~sk^_gIe3dEOF
ze`~U_m=}A?#`^CGi+{avU?W4b*QY#$+xoYxi^y>w=gezH%L<E0!hJaX+w2wWnsgR4
z6!a@k7wnO(C$Dy5K+aIt9&T`(a--s0T5L?$-n5W^(GFMmlMkguHwAvtjzay9pkscF
zT=&Pa^>Q1b6@FY!!weW(KKTk_R~CbX1N~uRzY}I!4Gwak1nlN_lp{ixMrlxmDET2Q
z1NOHdxr#+imfG$;tediv0bnxV-x)<Y9~wLi3NKzvry;&r{IPy9N!aXd>+@GtTDZOI
z?#2F=Q_1vP-qD_Ikl`^aRJ3+k?dbf#t;*^!M<SPOD1_AoeB3LSr3W7nA|Etn{&uy&
zWPrWJGKu|Drn%YZUGkBag7tLRg!ZpWz8bMw3HZo}@zx<-o}Tot&5$9TzulLj^n6dz
zgB5U-`KNKX&!8o<ti8~YY6{A!^P)<%o2w`GvsXLAj=$X4Tz?#n!|`V-j<TP44^gv0
z?WUm=7Z+f`ocL_Aa7Udq>H#V&@S4IM@*T8KXx+xn!_;%8jw?%pPQF9U!heg4=3KwB
zNjg3YyK%p?*y=iy<Asu}40u0DdEU!1PYvss+fo*ZZmq8x%tHs9<h2KNsFfw6S;gIR
zi?~RPQcIpzxV!;Dl*Gj7T2vnLVSHkY$35sCgASVL#P_^!ZWdE-dZq{PSR@&DnOEDB
zJ-~jMKv#>M{H>YW*^g#eDCmyIuBkq<NQo;u_8=Uyl{jjWX)<z-y#6qM!2hD=Y`)!w
zbpO_N8}8jLf0g<I^2~4RWBmyjzo3o3!R!+nw9NX^hjSq<d$U%j`+t7q5r&Ovpp^t6
z5?8-ek3r<IMFjqrhqz|)nC;baaA>;NVRD^L{T3fGG%`}9;QOn;`#;u9LQFPGOQ6Y4
z9)1fW`sY-$NuZK&G=t$4FYGRrz5tAaD9qAJou@*P&l(UN712LHgP9BiiCX~dpFsEg
zer7X6h-PGJQMyOQ^MfC;*Kwde7E0En(#w||zszViE}Tn;bv_D{C6R!*lA@z^UHCY6
zrfE1*pY~2y4Lc6a@%oo6tv|e@T4Q<LVz~DBABwG&>x3{*y+M`LV9jP~))#k{k-m4e
zm`O>-{2w3v(7GoFR%d*lp*DA>FJS(X0QjhO_C=0IZ|P8wlYijs6=`ARG6e9<wTFQG
zr<fa-2X3Q%XH%kRCU4DJOZHk;Q_1Fkeq1}gkvaLvSfBkZ9Wk{2k13*{EO$Ho6JKJJ
z>;P4Cr{yG1K3=}OqgLT5WMBQz&3*!>mip7$8Pmd^8g8B`A~hf5<v}P!6a<F>)ZCH#
z-flmbkFeLjMcczvV1W`$bmvm5#80yy9bgK7)a&Wb3b&FozhY{w|IFON>9_8yOgs@!
z@Je_&f3<9}>e86>W~h3<2auuvE9RNI($W+Buvl?-?P6H2NDwVO|LDhElf_ANL;3~f
zi!K3|Hg!?u1dy9=oJ*iNpz7z#G1kpb{@Ixar@8bGs%WE#`B%hRq;4)A%dyLaXbiZG
z%H5vY7x^Hh3S5<VvK2zaVhRjXDJ*{nwAaqe{7NP+T%{#FxMwriTj(M<W_#Jy<|J^2
zs4%6k_K+2VZ-mvK5QOtqv2UjbJVx2Oc>IqI6C&RT3gLgYLO}Zz3+;Q~GfUg0J2Cqg
z!Dq|+_@>0$i`fEd6z&EIVs2OMXq?s6Hv_^(;QH;5oUvK_q>JnP(ld4?C%XT4Vgh&D
zT1Ap|&MA(lMg`10lU1%@^R1^YVt0T%sc2q&RcR*d?t%HiqI-#8M9mXy6jYBvt2=yZ
zG>{o0z8zg*QASws<XhNCh|h6cwwa`5tREF>y6}t&LVeVZUtvK<U7HmYR2{|{cZhh>
z#oBB~!HRp5$_I=@Y?r>Bn*()?+u7d|F#`4RG&sMoBgU_{t6@pBAae5cGWzeM;vy*{
zlL>G^dnP(FHUD^!IjCe@)#!T}gc?d8Kr-0fAO7<b$4t$8DFqjZe(IEeG-cRMr?!$L
z$0NX>sq<Cc7ceUI7d0`5cZLtUyo)USEN9a+sRNFVm68UG9D|xmgt{bpi)edPMET9b
zioa25@Nlw061>z{gPm^<)tPGbX8@JHboU!}4(}*Q9B@?&igHTh+gd0p^`JlP#PjKq
zuUdzXV}HCFE}V?-^CpDGS6B>S6Wm<S@Au}WXmX0T;#P#0kbj}yXipFGbRLsD;_O)~
zvR%GQR<qjgX2dadZkHj3r`J33PeSpX^5i#px<5pOf^@zH_tXrMue5{GiyqjOoTkiS
zLDhqTl=3v?2dA3r9?x9l26)I>7M}PR9aJKspthWLIty1C6LiDV|48-=FuIkI-K7dH
zp`oA<o@N9Zs9Ii~V`qaOhCRX#O5Ej(IfPmkZ5QrEb(8FKN&#B-eo)J(Ty}fy^lQKN
z_j#3z$DsdR1x>W{CPTzjy`^~DQ87H4*;??KRZU*)u4(@$uirj!^bD`|=jOsHpsN~R
zhD%Vt+Ge?C_uHhdP^G>#{_k}if)WcasBBu;PBV$h*uHw+8ZQ5fu-R&IMp1gmf7*Mo
z@qv&QM_xEe()`)et+lmNi&2XJ_`0hCWsW8Q>)>>bhi2I7iq9*>u%G;mOXl(~qn2pG
z$#!=$hJiju(~VQ@TxYM>Eij<CTO63K&T7CbrMQ?Jzi02YuQh5_<|btyftQ^x!D(YT
zvlqiKT?^Wr@Ma_VzCf?_>taFHN+-JXasJuUKJltIpn(teA#ZkqO<$4?^v^fw<-M<X
zH(Z3Rm4sYidxoNyy6%Oz6zF9;8mW&>_pT4}-8<Wm#V>V*CieeiNi5TC4zypcj^exR
zNHik``(k#6VVjqfyW{>GoW$9C!>0wwI$cZ@z6DJQ33jp1*YmRExR{%2n0$-xEX&_~
zZdilrFw^9ijhMM`#SY^(%06@~Z;tw((P+Nsx6y(4{c|_5%wn);*1kDHN=L6qnHsKc
zs9i2V7yH)Jm_|$I=g=(!WdZNCrox-M$B*yb>3{D=H6w|(nrWeV3c(h)`+1Gxm<+@k
zManfeqkNwQncqi7XY&1^RYgo`VzVv>FT;Ud+n<X@gTY*irz#vluLP-okd)&~+Jkb{
zxMYECR!V+ibkh8{xSrH&cr{G6=a{|Rja!vv=VQN(pL6#pELvq&s6g*c7uynN<@p4g
zQZsT06`VV6_%%%)f7o|qxE23Vc#%BiAx=3TSMDCm{Wvy-kI};?Cl`d%V!qyA8nZlU
zzbdW1^xkZm(K1MvBcGZI$>ZF)L{@L`a`szYZef-=8}nRBp6y*+{`r2*LtrSWeQ@+a
zC}AU%W^sx$7dK1@qTL@zNy+-m(q)3Wi#V$q>%VOzg1P$7k*AqCD;XS~_~4hv;<wk%
R1w@3O#tWV2Rm#?3{|Dc4x=sK9

literal 0
HcmV?d00001

diff --git a/ico_design/icon_v2.iconset/icon_64x64.png b/ico_design/icon_v2.iconset/icon_64x64.png
new file mode 100644
index 0000000000000000000000000000000000000000..9749f19920ff6c085b99968fa43a253cff5b7b2a
GIT binary patch
literal 8619
zcmV;cAynRpP)<h;3K|Lk000e1NJLTq002M$002M;1^@s6s%dfF00009a7bBm00027
z000270mdy!jQ{{307*naRCt_?n^}xx$#tH;SniTrWo0d0UA;~3(>*(9B}GzPM2eC%
zv;fhTAX2bk*f0#+Fkrv<%`blPvtJAYf&l@dWXrTfaY!v(#Nm)b4u`|p_nzssy1J{X
zt19=qMT8$BZe}%Uz)ql0nN^uLBTk%i{{Q^vT=5_N?H{_o7r(~GFWkrHPE>T7OHh>9
zyu=k1qj7<@23J(L%HWJ48x4t~1ZON&X)wm3oCE8?T5t|$9aUK%rN$bEzyW}DIOm@`
z>ktA=RUxE83jdb+tqu@?F%?oO?s_;7V)j<7sSr}nK2v{Q0v5nlNU6gT98w5`0R3}!
zk&nhN(l-}TNrX&eD(64w1nA14jKB(ip~5<xl8CS%4M-utSP&K`B+gibgxoq>THu^R
zBCysWq`=v*Na<JU042q2Q{ZrJZUF%ZA?9xu;l~}=Y*_PM8^piQS%1B|>>FO_=DnY1
z-|et?<-fs$|B9%QU>#^B{D!BNrV^;|8SNZKDMX|Y6}WIECjbLr>1URoZH~DzE;-q(
z_|%am;QW3aND0n41Og%4?Bqqg!g{m6^bjE2T`%jm)tfCV>y3HPS$^SCA{a*i`uFZr
zl;slT%5Ssv#$WNw?IX0)Hpf<1Xei0;tr4OqanqbcE0RdFpbC@(TUmrLSYvV4aqUjd
z&u>rID;)PUC8>m^1dJ*DLahx8atP;72Ef|N@56x<e!;FjZ8!h_{Ivb&Uhf;88yWz9
zV^MOt9lXHK&F2LC?N-UH3jXQ(+ykm$XMNwr3s?t{e31Y2pMOHP{64R-#{Ks5#HK>(
z7WsHY*`E@Nl4cZfb}gmX>_BA@*7DMayS%nJWUr`bNXZ}EzryWd!Mj)XIls~(lHUF6
z+~5u1Ut0$Phg4BG+}i!UZ`H1?l)n2Io#TV7A`Peg?gUPC6k3Y;=AFX<nX8yuOO7Ox
z0!bZGK7^@^kC}eiU`IPXzm)M|Vv$DUD$BSu7-Mi$#7RUY1S*L@2wY|8=Z5nuO&&hd
z;nnMXzI&xlA_WgE#-x$LT89(f<)xf|pFh?4*$N55dB3)Z*-3M^M+2~(TW*$?iw&QP
z-kutUm8Bc0xt9nrGvwTG-Ye?AX(0rWav1Bl_u^w*yYm1yS{Y9@-{AeCOUs=jje5vl
zi&P4_TF_2%bgZ$##Xk4oVO(V}mEogjdOZ8y9bVZU^X2mkNG0hPB@2-R>#!D-lHeRN
zfQAtMM67d^l|>4H7GQ;#$N4X$h09~_`lp(T+&bQ#8m>+(4|P?@2f?UfzI`dgUHQQ|
zhtvXMB}uj9+=*q5udQ?V<Re_l_F36|k^3J#N2}W=U1*WEQqG<|#Ok3APHUW2s8}N-
zh0qEWX^;wsV7(dBZD_6y3jXB1J^F>kNslH{ih9>LA&^S?^$J1dpmGi&#DU~kKbafg
z{i&g+BxgDi7ZxH?DR^UKkkVgTh97$%10MW)a0YZgGZ^d0O3TmR-(o9ou(7?#6X(86
z>*RnRee*x^#_(e-HWpcJH@UGh;B3e6$h}>Z6xhNiIN?K@62YrGW4>^DiEG1>XD$zT
z`s5;u%?M{L8tL;vfK@4>tUybF63|MNPkcTTBJlJ=9l*VfNyTnud8DOhCwp%ycy(gA
zHm*3@&@+drLyd49W8LiUN(gkKJs`9bNq?BJQN<*pL31agm}Cs9i#)pcBffj}0{QA$
z)|Y#<mINw^v7(&8iKhaNMjG+du`b<a%>R62m#5Dzb5A$M8OuwX6TtDvYJ<{298OjW
zaCN*2feFb`hOn67cE}GOj)MEQ=H%#fOY!Q&a%EyU)(j{93n}EFQECsUQ3K;ZQIV7x
zktu0Q%WAvHR&|6Q-8#;19KXUezr-LbC<UlALd6j>*2p+Q#TuPN$Vl;t3v0Z-Ipq9e
z%&FBDLMdL_9P|2K#@XctFK<oxvrGG2>K8M=6rv7`xvSRhKlA(m_U}v$OImWQsb{CV
zo=SRJaxJs-IQhRM&)TT<L)&15NC|{<jEzAT6I$I<P)taBhuM=UZ;zhk(bn7iw7*6v
zPGaL2377zFGFZG0?~j~Z;#+Ur<kgKKTcez7{hTjdSf)QI`RSc0>#62UFPR0Wkiwr>
z%?5p*MBs4J8^Srq+x>#fI+nEHr+13kMx3*x0yZnl#>8;6seZ`_F05D15~L@545o&5
zqFCrOShkA9RJ5faP6g|WZQ8Ns#+{EbGPilK^(@;1i%KsdL);oAuzBf|j`SS+*jk&<
zmrFeN!GQg&<o6yrgfousT-)WTLoM>kqP4<W>pdVCpvL_g9sF+*lDc4H>fQflVYyix
zIPC+E%Lfc#L(F?4xw}{(4s?I(8q?BZl;Bub&`2Yy%yD*UN~;m^&W%m(In=@}tx*<T
z`nRuhpWI|Hyn!|fo0U*nTv0(~5tYFf6{a+-wk=N{Z_!N^#+3Z8H+OkxIpMMOgg0(Y
z`OcMn4yBTFo#66zUL@6oNjhKXR6uDQmvc)`3BGbHIdHTNV+>L%{%E7%T4r$89N++f
zaT)^%oR}l4?yQ`k5v=~H$Dy%w0)S>L$x6q+J97@_e4cO?D*IX9efb+a_v}}BeeVp`
zwpnZ+qT?#!D(Bu4YZT*(Dl^E^asSCRTxIyyD;uo0V}9k_A^zsAEpCrX&MhW<Fv!V_
z<ni^yv%|m`4>CsqOiTD+XvnPNOj9#1A<}Y&7?A*^=0d~p%D7@Phoj9Xe7*u4I?{6w
zsEeT_lEfh{Cc!F=STZRcd1+XRG?lGOHbJZfjm~+l<sb4Zoy&Z8_Xv}5gCZMXCk1(F
zu+s_3D2}hhh}c8bz3W|Gy|%~Ox5r!=7W~0Ohe)O1+_>Ot@9y&CQHeDc8|s~UfiC<W
zXB8hzDgX{^nC6C}FeoiC#-bA+CM&Uou_sE$n%bia^K>8&5Gv<DPYO$<1d$4FP=eAp
zT9Kd`X{MFCi&QvZGM@6*c!d$WJkWfeSGJch#W~v10t<&$k!49Xnlj1;$XH{5Ge;IU
zv)1NE@7(5OH>H_ISYz1^n?edC5@~&{Y)n|RC+;{Q$StH&u-5R}L828^Wk_O;lolc2
z;$p<lhZS3;tuIR=x$99=VJSdQ3xiG-#(IVp$paWE1WE+a!g>>GDQI_FB;Av|eN&L;
zS9$vAHebE*2s-I-C#y&i2o+(Shf3@4$uh4vwYI?Dys^pU-Hg3)!4Gfr`NUcSq(le{
zK~Ip%ha4dpN-6m4a)hcgxD-g~S)@`vtyu&T$5)Ow?xM|4uQza}@>;~*wm`R1Z%v0M
zGDYP`BmX5eUJ5)+I#(Ad(4ZjJh8G4G*#bV-dzK$fAI7>a(>%vjlBRaZSR-oo<>1I_
zn@^oT#P{FZ0^rwAEpg?}kZb*%lidU>EqMj6-=30Hmg6nW;Y3hX7OAAK@SMXKkj_zA
zhcN;bg*@QqjPaLtd!RUS;iVS%Qx$<S##+?L7%99`Q&bKq9ZL9<Ok&Mezt1#-Qm$~d
zN>SM+pXt8N-(4S&+EX+|LVHPKwMNDcCp-pQ>o|R6fzxYkY-PY0RvV6QUK;uYWi8*n
zHlQ0xx{>7h?Tj-`!E$Mk+F=1}txt?KbHve%{lSR;hY@N%G0Op|mh;k3`f5plx2*Hn
znMQ)jI&UbT*NS-J!g+4(kGV6gSZbx*y0yrwqi^!tt5^8TtuBall3oXu#waPs0fi4G
zEd^CxqQY8x?T9a4T;a;jgdg1K^VFdhkFU49erPR!`pzz81V;m9mO(!ha94FFr3hFL
z?*G#Wvi6L*hxmV##=BRnWJnkXvZ;Wwzk|*PbQ>{?os>qR`PuFRJUd$F*N?o#r1&W=
z5~xN@lt$<{B1$6vJekvGp%hXpmOCkf+;E}S;PK-N2q_UtAf)6_D%q((YUPj3TAYI-
z_X)HrjaQe%ymG9G69<0%k`Vx1SvOC70fuK35uXQ2<Bg)VPp6e}Ohn4IRgA|Ol~QaK
zf~dL5TjMs;{(#@T*yrcFo1tg|SNSlomGiHy4=3vtQG`~!yEo;@BORnxxXK``#7Uoh
zQyfKQ5yk=zVZ6Xnn!xm=PlQ4^TrH;jVmKUl_;3;EGS4u<09lrk<{3`SaU(N!rpo34
znzAe-YD7enDXu6O4k}t2vR2H!hu68Y+v3@s_xarN>&WCwWCg?xg%yU#2uv;zT40?(
zYJn5pI2Tr0403~zc<mr;;*-Zy-nl-(Ig3;RXMFt~ry6T4N{N~GS`H{c2a2AV?i8S0
z5ctMV-o55(ZhZ6B2A!pKVkK%qhVo>KJm6|4RZRD9p*w3cSCTjolIo!5?_jcfnV)_8
zzwpDYkCIf5#6~o9LK>w+))5O!Bm|G1UO?$4JNr4`etU=0iwTJmjPr^o)|(8*Q=Yk+
z^Z0Vi{Y$a;8XJ^P=QTuE6(F@|lVQisG0zF1Z%`Y+T2KD8O;^xTpcQn+W7>;rXz5|k
zIvC{!B?OTGB|H<X@{Cq*4HMwPgtbJ@(TJocOb&*lA%B1C44+(jj=#G80G-BpmK!a`
z;|UAR2BYzWrEUumHxNmO!%>4TH5)wt-ZpRDnR33DG9}_0SNC~%rNM*iE!=@7jBsAc
zlg?qBoi{$h|LRkA1R+)`z{_${1tJt{5tyF9%oG#hWpC%u0eiu#H+9xk!nXk&pe0nH
zYRf7|Wh`1rw1lhKA;j)3U%clXzPZsN$yQND5j%ga%sE7?5Rqq$y#>LqKYWBduPBO=
zKYi&Y_b<gfetZFE9XI+p@9$=O`dEAJ+%qk~<Gd7L5di?a3R>g7bE3A2-&3qS2hdW?
zLdI{%1<0BsKxOA8rO!Jea0vgtkk3dXNfbU;S<8p^S9o^&IKO}Fd2a5%#<&_W1~O~O
z!<WcdqhpOsqM$D--rpMI1U!1Yjns<k{hV)H>2q!=Wl#o+sASL+d>o4)NXP?>R>%YB
z05jobrU>bPsfqSHKuf{2unaOotOOzX6W1Dp@CE`;j<odQQdE|_^r$f@EY4Ioy+Gs8
zI$O~?&up*rmGeX51;KDShGv6OSmH*6)0(l9?3XaI9`f!#(IeH8pWmGD(`y5sxxCL8
zPxol2n!os<&-bp6LY+X(m+#tqZ45$xO@xgf6h-B<bYzvA!Cg^VvdYm)B)z6)m|I4<
zPk?z9>gri?6%@u%7@ufjEosJrSP7CyQaFWe_ZZ*VCC`_*G`_%#gJV3_dV#&kEpF{y
zV?@N%fs8e+R?KRzL2n_&MVcsyc>3WZ3`@%kw}yQ2+zN~Bgs;E9M=wzfOT)9*M_yGv
zpfuZ{t2qKWI)DUT6$<WOQ{g((YbA)3kN>(oB?Tmr!~`Z-cXGhPl(k+fsF9>J-cyVX
z@|cQ{jTWYCq83(pIh(MxKjkauU*=o4mKYUBiP8ueM+hmf&iLo0|Gwp};FITiTpYn-
zD*5wQZt>u1izknD*c}yMZP@(GzdelB4OkZ%vPR71^}v}0AhXh992Aw$s5LEALGH3a
zED?AYCs<I-Tt8&kkR&7JT`ksvBvLelp`1)`ro@#6NfNVRBVL)bdAxIrL3xKtfodis
z&6u<qqXT2~f~zJ;6=#ohc=2kVhYq)S>U0mOCCi<dH+QByzcFS|I$u#LF;i7%p)m{o
zx(E_AZ$h_IzdfyLF=o!N^3u{u6=mh_N-9c-(BAScR>zdAnO*4b31@M(;?(L2d&543
z)X1_RT6~zF?Ch}~U*lu#x6tWhs5HjrUimS?V6`R=iCIM&p*6qxkt5j3;wr<WwCs;c
zx{ZjeH2lr=DZhC#rW;G7gsikAQev#{G}Qutu=V;pMu_je{H~jpmQJGB-rgkb9Kr$Z
zM4^>n7!qOa9M(Ff!)>}tYe;Ib((Cvt6jNTvbnNfm=Dke^S|gpsRE9W-7>+Vptp>yW
zJ$lV4$^JWBDbI6s<uK=#Q}+8AA70(VObZ&3LgqtGt}f3uzc(!S^VfC&SZgRgezeJp
z8zWM!IMz~JA6DF$7`}8*BXprHrYaGED9vUG1n8u9jD^S-Jmb8gwQ`(Btk5-dWX4m?
z0DDqG9!na_$LG+<TwzfG{)}}bh2^2gjxv~5q>+abDFxHg(g=B_6-)l?Pu}G4`0M=9
z+n?s?Cm&^ZtK`N91KM#!OvSRcoIBg|VryP<dutD*<X?XDI4@n@<1a1^anAAhTEg4?
zoVN!B|L8;q9ml+UYs^Ph8{P|~o{6n~Xan>@s%AK!Xh9=VL`twTENR4&ZbM<5FMuND
zA+o4>fD|->lvX(pjnmS3<Kel%epa#6(hM>$z1`~P99>K)jHR0@CWU3?#HaZFi<;j*
z@(h3c{dZ}MmRMRo&OL{h*xuNt(%_T^?HS+$Cwshqd%*YJze5^g{lVpg(mHNWD?YW>
zMo7uVsH7jl*GYj?J`}8l$_vCZPY66fq9nvpaQXI_Tm53DqAKaxU!7NKXz{Z)zWPdn
zfvP!9(oHoB4NW&ybP`D?(JZwT?O4%F6pdK&(5Xdw>!<kZTZj4QXRebcm+2i|;ne9v
zq>CL|y%tF`B1vPCG(sj3zw^<PEH-0a*c|ZCp*9G?Ydcf^@r7lcI^N;U-Hc29f=?e0
zntROyf|MqtpMa6I#^&AGQCcXiLnz6eQ88;M8t1eBeB)472M|Ei5E&Gv5JHvpDo|ZF
z*pM&ka=S7frj%ObjXP)fyRDpm|M-}V!BuWgDs~2aas)&XGMTsGRIGXU>?&Wlca8H$
z7Wl#DnBTgvLa&qZ%=>!){MzX*E(oeMPY~gC{m`X#Gj(O08?s47uc>+PSewc^CWRp{
zEfz?%B#pwblMpEBMH2!ube0*KRM!7(kQqOvQCgZ#5i7~=xbiZ5;jjC>*KhN~>wTiM
z$DM^g<bPawfhWYX{PpvWX#Wn2EMv6x^99<Md6jX*DcpCmhpP-1dnwb(^7Xg25kl|}
z&M)!$=7`&)f<q0>#YN8&YlYtl5JDm=;}{o~X<1ixj;!>Be=C+uOG{zAdpn>o%!(%K
zJn^w%3QJ0V`cgouedu%>%6E+;UmPvBqnY*?DTpO(Zx2{m=yLzP=lK24KS6Zjw|H)J
zichcpEoLF7jA9z?gt!?Kr7<dw&`Cs+dSyvRnx`LHVOV(f^PBgr@|_R&d2MIHLZZ3U
z&v|(}#{}AJ0I8Ix{Z2rv1hJO>l$D3E$~wy2(To+7!cZEAlF*0*Rlw)Ec#>fT$XYNV
ztC%nYI4vyA#1|39K^e+xYrVi3DaqPOn-@OF$@fOQ`0gz(UHyQaaf>)Q&SzII@$RsL
zOky8G0m77XxJn?yT7{N8cxs79Pxg50`aahN1z&#n5Xw2uXAQpg?k<lX6&NQ_3QSOv
zG-U?DpJbk8=yFODDU|h0Fq=B8u~g27K!u<xGU6)pr!GRzAQ+=G0TQLalsQgj*iw>}
zb_SQR5~v`?W`!fK3YwzAOoo(hg{0LX9rm~>j`C9fHU9O5OSnkjDv*)D7;s7;t;0%?
zS|F6~Y1qoLpOu{GCM1yrYuOr?gu<9XVOWTT*9ZiN7v9@&bt!%4_6-(SkIs7N^>4w9
zJ|m^%&ejIsdEanlZ-PM3N@H$p?yxtS(rP5+c}||?Jh&F~FMj6}ehS3Qs?Zrbt!H+|
zyEjP|4`ZyaBTD1#nP|^Ahg+9;@jL&O?`<3-F&$cTNn}JrE40u=*7DJN7wL6;KDc>j
z!r#8V!zWKHus<nzZezqJ4#g~{G2gg4<a6sy&MasI0v)xNNK|nD(U7#WG9NW_zS8q`
zzV`?z1dZ{C#luZHa>T8@F{4sYHCNd!Ml7_NG&_>Xc*>yc)99{{SB^MjU1I~EP;(tW
z&PKRoiAefMmCE_h(OQs5!QSo;&u{hk%-ZXG^X6yClNDMjLN^o4G$U>##O(%2r-Lmk
z?nzQUV-??j*Ee6jaC(Wzl|1v|fYa@mxAqE#rQ@-+Cc57292n*cb<CwP8{SkaEBpN^
zWnK_z#bP7oPE{hE<@RX8WICmqPHCKIVl71BJ7ho$DFk`tNVS-`gYWb@@;Wv{52CCb
z63d(W9pFBmTzZY4jW#I<Cotobg|vg|Pcbg>RiyCve)edWGoktN!RCOkzq89fJio%p
zZp_}O;E&(l;lb4=y5{(Td7W?Fe0;Ue__dzmhZlmBZVU>@E8~e!q`9Y)FpVWUN>FC=
z2@dKof|)T68s3I4uU{K{BgN|(65Tw<dzFIn$2`^E<wduQ&>adPh$D$g6e{tun#9dG
z3BV67@A1X6T~4pIaK`Z7ZbmOwB#A=Ti*N_@|GKa-fy98vbt%0TV60<roU=2^Ns<Tz
z>`n?sd5K=_B8_8Urxaxsgj+W&rfT6v&Fa(niOahF2smGF!8l1_1+LNI?fzND=`}vN
z`YM0-;Xdu;EV-1FN@27{IImA=zyxdmvtK<;J{}^J<j0rycx`9O?_5~&hOJ#q2tgy3
zv)0%!Gyyt^3T_*uGa*PL#o?Busk<yKw>h@d;5lb$HWF^`joBQGsLF~94;}&L!xjMJ
z($Y%2<km?QrLi6sonxBY8Le(5e(bfh4lOKEXMt9ygJqSis>usmWB$qgSNZCPhp47g
zN^^v$5knAktYCJSq5`QUKmKr^SGUIe?t_PD#|mjJ+BXxtJC>F83zCvnEHMu9S_HA)
zL1XFVGnd4Sr&G3W4%pnAa$_*1C@l{hUFM$U4ja2eu5Ik1CG;mIaB**pN#Q5DCxt~z
zD2$^aJQt~}#A#`XmGlx85?fT<8BRz^QBLx~<RV|&+~E`5*ZHwHjBa<3jl@%Q7h3C%
z*|3FBk|Rq^E*|O7jTP1yoC&kSzPV917rmQjr6)3x@}W@6=OWO_>9j;R%RQ@Y4ll$E
zMpJGLbM__~+x;<ndBx*LT0Yq&lA?06rfVyf{`LX^Wvx5|BIpmLb+lrIRvry-mMV@p
z(p#VyRurzpMGIWp-lns>$P>K}AsQi~2!i_y0hRDAbSXe+#o426OzCg>sKOwo-r2l7
zOFr3f8*4q|%fn<yO>Ju^tgK_&-{#(CjA{!y3tlj3A1}#D$7GyQ_6$|!IJPRtcCTWT
z|C|V(GzzgE!kyv%7R}J`%mWSAi6#rVgQCyBd7{U{Vw+;<D9r+SVafUt&0d#9w}sk&
zkC(Cz#c+z7WHh4~la(~%2wRji$Y`a0h|3s<u-5B|Ke;}20?+$5H!kDaM`yagq%gD+
zMPYn0%qvSfQI!1+bob~iU-+cyH`fTm6_s_2cCI7ZM`<L|Q|SOvCgc=j9Z4h^Z(l<%
z9w*U)L<=URSEn1Xq(3p}@dgWr?m<gt#f%M9TT3tB{qMi}GcJt3!|TzPxpAi?ACxq0
zMY*@duRd~)qrI3UX^~9|q%lw#s?s2f(o$5GcQ?m4R~Ic_BWx#%{?wqQAkmVARD%$d
zl}}btP?2LT!h6Kbo3!+urZ|#*5FwPvGN6JIUOJaTP#R0D=GtPSJ?Y9zOCu5##?Sv8
zkQrvBZ5_h7>GFf(V|4eP<HqhS;)M>n)#Uh@qg)&0Y>f(>(nM(lTB9P3juaB%`+xPs
zs1C$O3PBplnFun+@<5QJCpb#XX9^(5X5mK*U-i{0vc$hu!N8T7kE{3!QThR9;h)#Y
z6a}7NPpH&CQJ8Frm7pj<EZ)x#?yPX0R~Tm742mhT)uxORT&!@BMnutU$~{iBSA<%L
z;%G0Wa$X5`)HS{HCrAw`E_93qm4)0m%0Lk->t=n!gY%mvXmG}cChBYe!^gjI^HRDN
zO*|ArWh|v()^&Dx>CL#gI#mV6TH1&gZe+yApW=h+3>VAqb9MJE-n@Pr*=*5XXyCM>
zkO~n;Snd0fGEyR(wy3+!zElmJ*ZMj5FRuFlzGpc9+iWr6|5Xr8YU9<-m&%13%$PGl
zKlJ3k$A;?^;jWU}D3zN_cj6b;;*IJUj0SWQ&2_oPYgxp{mtSYMyo6JJxLYJKQQ|vJ
zB8qU4MkiYOxn1e!=Zynr{rv-#akNrj%U8xxls4S)1zF`tl!wW>AD|9sh4l#)rKJ(8
z*+~fq#Sl3B0G3i=Hc1Gw()#r%i9j%zRwyCv3d5Oq1x=zfj%KP@(K+2#%KB=LOV=-w
z<M`F}+t8R2H(JEagt2inwZurUIp|?lQCZj<l_UZ859XRIw?w)YMm=3E!e)KU&o@6C
zLklNo@pq%RfW=D8%W9|#EjR0Mn_5{Cf}o<)h$Uv~lPt`6ejSoEcj+!Pcw$YHPevT>
zbvV5I6y@Y1qI-?C=}UaDd6p+1^0$q-ydVn7kS_OcFfI&vWf69tNK32*tp$<u39^-V
z<}UFqwEpfb5)GB`(_i&N{K7F!`&&pHdC2;85o8^d2Xe)9uuUUYcWDJvH=k3?23ri&
z-4{P-NDwSHN<R68$N0;iT;qWwi*y<hk!rBiI?A&@{2Gr`Z!s92gsj4r5^W47H;9)v
zGdIo+%~-N^^D@!m3C6kM=%OEQ+n-cKLA|lgu|28~qYc)No}v-eh*7)!oZJBHZ(XCg
za(t#7*K-J?!qSc<S!J;nu3mYcGiS~dD>2IrlMsSh3bMUxEUldkTb+m5!SA}YahdMg
zJ>;dEC0_7o;(&Kv{Bth7@>QPs@FSd9|0wsLI?QqusBrCerGYG~X5*+?>HBKa!eX7z
z3$YSJkqR|}7dk8I$q_fJ@?1FKNp46ZNnySFqCLJ#y!<W&B(b8XEK%t1S3yQMHq04R
z?yj(`Jww(V>hRAyqmt!T#I*FEi<Dr0Sa7HlGc8v5_RXXG_L&WSxPOOtZbbAVg}gl|
z7-WW>QAMPn6{}ertDPur4-2I9I^I%C(~2ch_}HvRV=5b%*+CbDP(w)WjJ4a4vH$=9
zQAtEWRI6E84KOZDkla0?Dk?XdOBBrd81yqkE0G9lDD|*gn)#-UW5z01+Yy7@Fw6~+
zl0-^!td~$4OVn9mH~B0-8{f}^jpvD4j<aXh&_}xQ%zdxy5o9UFC^zg+4JVe;8N3N0
z=%kYAM9z|69N1r3ImXtZq{n<4KC`i(v)Ir~OB=9U(2C_Ow@eDdaw}p|81^TXuM$hY
zfyIVqdstFshEC~bP+EUqSs4;7nOH{}`D^wX3YRKM>y5HMF)TI}3$2)sKmHJ}yt_be
zag}>U|DA6=`;fnJGKeljN|9<2=6n=cX=d5mB8UQuQ&P?nQKG_KUV#~U*K{~3sHCKu
zY7cSFA%zUY$=4p>+i|%Att3)7nu#K<Bt_**<yN8?D@m+8oi0qx3;Yy|br2yyDSul;
xuu!a^LSVI+(d<M^R=ZtFV%{8_<csIV{2#<e_`s)!Q`P_g002ovPDHLkV1gZpjuQX?

literal 0
HcmV?d00001


From d330cbf09a6959fa2cb825eb1a604f41166d5733 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Thu, 6 Mar 2025 18:30:19 +0000
Subject: [PATCH 02/33] delete opencv dependency

---
 .gitmodules | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/.gitmodules b/.gitmodules
index 3e82001..a18765f 100644
--- a/.gitmodules
+++ b/.gitmodules
@@ -14,3 +14,6 @@
 [submodule "subprojects/pytorch_scatter"]
 	path = subprojects/pytorch_scatter
 	url = https://github.com/guanshaoheng/pytorch_scatter.git 
+[submodule "subprojects/opencv"]
+	path = subprojects/opencv
+	url = https://github.com/opencv/opencv.git

From 5414a7e71ea59e473a3e3e1fe81c68f031d0979e Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Thu, 6 Mar 2025 18:30:42 +0000
Subject: [PATCH 03/33] add nlohmann dependency

---
 include/nlohmann/json.hpp          | 25510 +++++++++++++++++++++++++++
 include/nlohmann/json_original.hpp |  5306 ++++++
 2 files changed, 30816 insertions(+)
 create mode 100644 include/nlohmann/json.hpp
 create mode 100644 include/nlohmann/json_original.hpp

diff --git a/include/nlohmann/json.hpp b/include/nlohmann/json.hpp
new file mode 100644
index 0000000..07dea3c
--- /dev/null
+++ b/include/nlohmann/json.hpp
@@ -0,0 +1,25510 @@
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+/****************************************************************************\
+ * Note on documentation: The source files contain links to the online      *
+ * documentation of the public API at https://json.nlohmann.me. This URL    *
+ * contains the most recent documentation and should also be applicable to  *
+ * previous versions; documentation for deprecated functions is not         *
+ * removed, but marked deprecated. See "Generate documentation" section in  *
+ * file docs/README.md.                                                     *
+\****************************************************************************/
+
+#ifndef INCLUDE_NLOHMANN_JSON_HPP_
+#define INCLUDE_NLOHMANN_JSON_HPP_
+
+#include <algorithm> // all_of, find, for_each
+#include <cstddef> // nullptr_t, ptrdiff_t, size_t
+#include <functional> // hash, less
+#include <initializer_list> // initializer_list
+#ifndef JSON_NO_IO
+    #include <iosfwd> // istream, ostream
+#endif  // JSON_NO_IO
+#include <iterator> // random_access_iterator_tag
+#include <memory> // unique_ptr
+#include <string> // string, stoi, to_string
+#include <utility> // declval, forward, move, pair, swap
+#include <vector> // vector
+
+// #include <nlohmann/adl_serializer.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <utility>
+
+// #include <nlohmann/detail/abi_macros.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// This file contains all macro definitions affecting or depending on the ABI
+
+#ifndef JSON_SKIP_LIBRARY_VERSION_CHECK
+    #if defined(NLOHMANN_JSON_VERSION_MAJOR) && defined(NLOHMANN_JSON_VERSION_MINOR) && defined(NLOHMANN_JSON_VERSION_PATCH)
+        #if NLOHMANN_JSON_VERSION_MAJOR != 3 || NLOHMANN_JSON_VERSION_MINOR != 11 || NLOHMANN_JSON_VERSION_PATCH != 3
+            #warning "Already included a different version of the library!"
+        #endif
+    #endif
+#endif
+
+#define NLOHMANN_JSON_VERSION_MAJOR 3   // NOLINT(modernize-macro-to-enum)
+#define NLOHMANN_JSON_VERSION_MINOR 11  // NOLINT(modernize-macro-to-enum)
+#define NLOHMANN_JSON_VERSION_PATCH 3   // NOLINT(modernize-macro-to-enum)
+
+#ifndef JSON_DIAGNOSTICS
+    #define JSON_DIAGNOSTICS 0
+#endif
+
+#ifndef JSON_DIAGNOSTIC_POSITIONS
+    #define JSON_DIAGNOSTIC_POSITIONS 0
+#endif
+
+#ifndef JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON
+    #define JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON 0
+#endif
+
+#if JSON_DIAGNOSTICS
+    #define NLOHMANN_JSON_ABI_TAG_DIAGNOSTICS _diag
+#else
+    #define NLOHMANN_JSON_ABI_TAG_DIAGNOSTICS
+#endif
+
+#if JSON_DIAGNOSTIC_POSITIONS
+    #define NLOHMANN_JSON_ABI_TAG_DIAGNOSTIC_POSITIONS _dp
+#else
+    #define NLOHMANN_JSON_ABI_TAG_DIAGNOSTIC_POSITIONS
+#endif
+
+#if JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON
+    #define NLOHMANN_JSON_ABI_TAG_LEGACY_DISCARDED_VALUE_COMPARISON _ldvcmp
+#else
+    #define NLOHMANN_JSON_ABI_TAG_LEGACY_DISCARDED_VALUE_COMPARISON
+#endif
+
+#ifndef NLOHMANN_JSON_NAMESPACE_NO_VERSION
+    #define NLOHMANN_JSON_NAMESPACE_NO_VERSION 0
+#endif
+
+// Construct the namespace ABI tags component
+#define NLOHMANN_JSON_ABI_TAGS_CONCAT_EX(a, b, c) json_abi ## a ## b ## c
+#define NLOHMANN_JSON_ABI_TAGS_CONCAT(a, b, c) \
+    NLOHMANN_JSON_ABI_TAGS_CONCAT_EX(a, b, c)
+
+#define NLOHMANN_JSON_ABI_TAGS                                       \
+    NLOHMANN_JSON_ABI_TAGS_CONCAT(                                   \
+            NLOHMANN_JSON_ABI_TAG_DIAGNOSTICS,                       \
+            NLOHMANN_JSON_ABI_TAG_LEGACY_DISCARDED_VALUE_COMPARISON, \
+            NLOHMANN_JSON_ABI_TAG_DIAGNOSTIC_POSITIONS)
+
+// Construct the namespace version component
+#define NLOHMANN_JSON_NAMESPACE_VERSION_CONCAT_EX(major, minor, patch) \
+    _v ## major ## _ ## minor ## _ ## patch
+#define NLOHMANN_JSON_NAMESPACE_VERSION_CONCAT(major, minor, patch) \
+    NLOHMANN_JSON_NAMESPACE_VERSION_CONCAT_EX(major, minor, patch)
+
+#if NLOHMANN_JSON_NAMESPACE_NO_VERSION
+#define NLOHMANN_JSON_NAMESPACE_VERSION
+#else
+#define NLOHMANN_JSON_NAMESPACE_VERSION                                 \
+    NLOHMANN_JSON_NAMESPACE_VERSION_CONCAT(NLOHMANN_JSON_VERSION_MAJOR, \
+                                           NLOHMANN_JSON_VERSION_MINOR, \
+                                           NLOHMANN_JSON_VERSION_PATCH)
+#endif
+
+// Combine namespace components
+#define NLOHMANN_JSON_NAMESPACE_CONCAT_EX(a, b) a ## b
+#define NLOHMANN_JSON_NAMESPACE_CONCAT(a, b) \
+    NLOHMANN_JSON_NAMESPACE_CONCAT_EX(a, b)
+
+#ifndef NLOHMANN_JSON_NAMESPACE
+#define NLOHMANN_JSON_NAMESPACE               \
+    nlohmann::NLOHMANN_JSON_NAMESPACE_CONCAT( \
+            NLOHMANN_JSON_ABI_TAGS,           \
+            NLOHMANN_JSON_NAMESPACE_VERSION)
+#endif
+
+#ifndef NLOHMANN_JSON_NAMESPACE_BEGIN
+#define NLOHMANN_JSON_NAMESPACE_BEGIN                \
+    namespace nlohmann                               \
+    {                                                \
+    inline namespace NLOHMANN_JSON_NAMESPACE_CONCAT( \
+                NLOHMANN_JSON_ABI_TAGS,              \
+                NLOHMANN_JSON_NAMESPACE_VERSION)     \
+    {
+#endif
+
+#ifndef NLOHMANN_JSON_NAMESPACE_END
+#define NLOHMANN_JSON_NAMESPACE_END                                     \
+    }  /* namespace (inline namespace) NOLINT(readability/namespace) */ \
+    }  // namespace nlohmann
+#endif
+
+// #include <nlohmann/detail/conversions/from_json.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <algorithm> // transform
+#include <array> // array
+#include <forward_list> // forward_list
+#include <iterator> // inserter, front_inserter, end
+#include <map> // map
+#ifdef JSON_HAS_CPP_17
+    #include <optional> // optional
+#endif
+#include <string> // string
+#include <tuple> // tuple, make_tuple
+#include <type_traits> // is_arithmetic, is_same, is_enum, underlying_type, is_convertible
+#include <unordered_map> // unordered_map
+#include <utility> // pair, declval
+#include <valarray> // valarray
+
+// #include <nlohmann/detail/exceptions.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstddef> // nullptr_t
+#include <exception> // exception
+#if JSON_DIAGNOSTICS
+    #include <numeric> // accumulate
+#endif
+#include <stdexcept> // runtime_error
+#include <string> // to_string
+#include <vector> // vector
+
+// #include <nlohmann/detail/value_t.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <array> // array
+#include <cstddef> // size_t
+#include <cstdint> // uint8_t
+#include <string> // string
+
+// #include <nlohmann/detail/macro_scope.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <utility> // declval, pair
+// #include <nlohmann/detail/meta/detected.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <type_traits>
+
+// #include <nlohmann/detail/meta/void_t.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+template<typename ...Ts> struct make_void
+{
+    using type = void;
+};
+template<typename ...Ts> using void_t = typename make_void<Ts...>::type;
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+// https://en.cppreference.com/w/cpp/experimental/is_detected
+struct nonesuch
+{
+    nonesuch() = delete;
+    ~nonesuch() = delete;
+    nonesuch(nonesuch const&) = delete;
+    nonesuch(nonesuch const&&) = delete;
+    void operator=(nonesuch const&) = delete;
+    void operator=(nonesuch&&) = delete;
+};
+
+template<class Default,
+         class AlwaysVoid,
+         template<class...> class Op,
+         class... Args>
+struct detector
+{
+    using value_t = std::false_type;
+    using type = Default;
+};
+
+template<class Default, template<class...> class Op, class... Args>
+struct detector<Default, void_t<Op<Args...>>, Op, Args...>
+{
+    using value_t = std::true_type;
+    using type = Op<Args...>;
+};
+
+template<template<class...> class Op, class... Args>
+using is_detected = typename detector<nonesuch, void, Op, Args...>::value_t;
+
+template<template<class...> class Op, class... Args>
+struct is_detected_lazy : is_detected<Op, Args...> { };
+
+template<template<class...> class Op, class... Args>
+using detected_t = typename detector<nonesuch, void, Op, Args...>::type;
+
+template<class Default, template<class...> class Op, class... Args>
+using detected_or = detector<Default, void, Op, Args...>;
+
+template<class Default, template<class...> class Op, class... Args>
+using detected_or_t = typename detected_or<Default, Op, Args...>::type;
+
+template<class Expected, template<class...> class Op, class... Args>
+using is_detected_exact = std::is_same<Expected, detected_t<Op, Args...>>;
+
+template<class To, template<class...> class Op, class... Args>
+using is_detected_convertible =
+    std::is_convertible<detected_t<Op, Args...>, To>;
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/thirdparty/hedley/hedley.hpp>
+
+
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-FileCopyrightText: 2016 - 2021 Evan Nemerson <evan@nemerson.com>
+// SPDX-License-Identifier: MIT
+
+/* Hedley - https://nemequ.github.io/hedley
+ * Created by Evan Nemerson <evan@nemerson.com>
+ */
+
+#if !defined(JSON_HEDLEY_VERSION) || (JSON_HEDLEY_VERSION < 15)
+#if defined(JSON_HEDLEY_VERSION)
+    #undef JSON_HEDLEY_VERSION
+#endif
+#define JSON_HEDLEY_VERSION 15
+
+#if defined(JSON_HEDLEY_STRINGIFY_EX)
+    #undef JSON_HEDLEY_STRINGIFY_EX
+#endif
+#define JSON_HEDLEY_STRINGIFY_EX(x) #x
+
+#if defined(JSON_HEDLEY_STRINGIFY)
+    #undef JSON_HEDLEY_STRINGIFY
+#endif
+#define JSON_HEDLEY_STRINGIFY(x) JSON_HEDLEY_STRINGIFY_EX(x)
+
+#if defined(JSON_HEDLEY_CONCAT_EX)
+    #undef JSON_HEDLEY_CONCAT_EX
+#endif
+#define JSON_HEDLEY_CONCAT_EX(a,b) a##b
+
+#if defined(JSON_HEDLEY_CONCAT)
+    #undef JSON_HEDLEY_CONCAT
+#endif
+#define JSON_HEDLEY_CONCAT(a,b) JSON_HEDLEY_CONCAT_EX(a,b)
+
+#if defined(JSON_HEDLEY_CONCAT3_EX)
+    #undef JSON_HEDLEY_CONCAT3_EX
+#endif
+#define JSON_HEDLEY_CONCAT3_EX(a,b,c) a##b##c
+
+#if defined(JSON_HEDLEY_CONCAT3)
+    #undef JSON_HEDLEY_CONCAT3
+#endif
+#define JSON_HEDLEY_CONCAT3(a,b,c) JSON_HEDLEY_CONCAT3_EX(a,b,c)
+
+#if defined(JSON_HEDLEY_VERSION_ENCODE)
+    #undef JSON_HEDLEY_VERSION_ENCODE
+#endif
+#define JSON_HEDLEY_VERSION_ENCODE(major,minor,revision) (((major) * 1000000) + ((minor) * 1000) + (revision))
+
+#if defined(JSON_HEDLEY_VERSION_DECODE_MAJOR)
+    #undef JSON_HEDLEY_VERSION_DECODE_MAJOR
+#endif
+#define JSON_HEDLEY_VERSION_DECODE_MAJOR(version) ((version) / 1000000)
+
+#if defined(JSON_HEDLEY_VERSION_DECODE_MINOR)
+    #undef JSON_HEDLEY_VERSION_DECODE_MINOR
+#endif
+#define JSON_HEDLEY_VERSION_DECODE_MINOR(version) (((version) % 1000000) / 1000)
+
+#if defined(JSON_HEDLEY_VERSION_DECODE_REVISION)
+    #undef JSON_HEDLEY_VERSION_DECODE_REVISION
+#endif
+#define JSON_HEDLEY_VERSION_DECODE_REVISION(version) ((version) % 1000)
+
+#if defined(JSON_HEDLEY_GNUC_VERSION)
+    #undef JSON_HEDLEY_GNUC_VERSION
+#endif
+#if defined(__GNUC__) && defined(__GNUC_PATCHLEVEL__)
+    #define JSON_HEDLEY_GNUC_VERSION JSON_HEDLEY_VERSION_ENCODE(__GNUC__, __GNUC_MINOR__, __GNUC_PATCHLEVEL__)
+#elif defined(__GNUC__)
+    #define JSON_HEDLEY_GNUC_VERSION JSON_HEDLEY_VERSION_ENCODE(__GNUC__, __GNUC_MINOR__, 0)
+#endif
+
+#if defined(JSON_HEDLEY_GNUC_VERSION_CHECK)
+    #undef JSON_HEDLEY_GNUC_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_GNUC_VERSION)
+    #define JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_GNUC_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_MSVC_VERSION)
+    #undef JSON_HEDLEY_MSVC_VERSION
+#endif
+#if defined(_MSC_FULL_VER) && (_MSC_FULL_VER >= 140000000) && !defined(__ICL)
+    #define JSON_HEDLEY_MSVC_VERSION JSON_HEDLEY_VERSION_ENCODE(_MSC_FULL_VER / 10000000, (_MSC_FULL_VER % 10000000) / 100000, (_MSC_FULL_VER % 100000) / 100)
+#elif defined(_MSC_FULL_VER) && !defined(__ICL)
+    #define JSON_HEDLEY_MSVC_VERSION JSON_HEDLEY_VERSION_ENCODE(_MSC_FULL_VER / 1000000, (_MSC_FULL_VER % 1000000) / 10000, (_MSC_FULL_VER % 10000) / 10)
+#elif defined(_MSC_VER) && !defined(__ICL)
+    #define JSON_HEDLEY_MSVC_VERSION JSON_HEDLEY_VERSION_ENCODE(_MSC_VER / 100, _MSC_VER % 100, 0)
+#endif
+
+#if defined(JSON_HEDLEY_MSVC_VERSION_CHECK)
+    #undef JSON_HEDLEY_MSVC_VERSION_CHECK
+#endif
+#if !defined(JSON_HEDLEY_MSVC_VERSION)
+    #define JSON_HEDLEY_MSVC_VERSION_CHECK(major,minor,patch) (0)
+#elif defined(_MSC_VER) && (_MSC_VER >= 1400)
+    #define JSON_HEDLEY_MSVC_VERSION_CHECK(major,minor,patch) (_MSC_FULL_VER >= ((major * 10000000) + (minor * 100000) + (patch)))
+#elif defined(_MSC_VER) && (_MSC_VER >= 1200)
+    #define JSON_HEDLEY_MSVC_VERSION_CHECK(major,minor,patch) (_MSC_FULL_VER >= ((major * 1000000) + (minor * 10000) + (patch)))
+#else
+    #define JSON_HEDLEY_MSVC_VERSION_CHECK(major,minor,patch) (_MSC_VER >= ((major * 100) + (minor)))
+#endif
+
+#if defined(JSON_HEDLEY_INTEL_VERSION)
+    #undef JSON_HEDLEY_INTEL_VERSION
+#endif
+#if defined(__INTEL_COMPILER) && defined(__INTEL_COMPILER_UPDATE) && !defined(__ICL)
+    #define JSON_HEDLEY_INTEL_VERSION JSON_HEDLEY_VERSION_ENCODE(__INTEL_COMPILER / 100, __INTEL_COMPILER % 100, __INTEL_COMPILER_UPDATE)
+#elif defined(__INTEL_COMPILER) && !defined(__ICL)
+    #define JSON_HEDLEY_INTEL_VERSION JSON_HEDLEY_VERSION_ENCODE(__INTEL_COMPILER / 100, __INTEL_COMPILER % 100, 0)
+#endif
+
+#if defined(JSON_HEDLEY_INTEL_VERSION_CHECK)
+    #undef JSON_HEDLEY_INTEL_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_INTEL_VERSION)
+    #define JSON_HEDLEY_INTEL_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_INTEL_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_INTEL_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_INTEL_CL_VERSION)
+    #undef JSON_HEDLEY_INTEL_CL_VERSION
+#endif
+#if defined(__INTEL_COMPILER) && defined(__INTEL_COMPILER_UPDATE) && defined(__ICL)
+    #define JSON_HEDLEY_INTEL_CL_VERSION JSON_HEDLEY_VERSION_ENCODE(__INTEL_COMPILER, __INTEL_COMPILER_UPDATE, 0)
+#endif
+
+#if defined(JSON_HEDLEY_INTEL_CL_VERSION_CHECK)
+    #undef JSON_HEDLEY_INTEL_CL_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_INTEL_CL_VERSION)
+    #define JSON_HEDLEY_INTEL_CL_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_INTEL_CL_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_INTEL_CL_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_PGI_VERSION)
+    #undef JSON_HEDLEY_PGI_VERSION
+#endif
+#if defined(__PGI) && defined(__PGIC__) && defined(__PGIC_MINOR__) && defined(__PGIC_PATCHLEVEL__)
+    #define JSON_HEDLEY_PGI_VERSION JSON_HEDLEY_VERSION_ENCODE(__PGIC__, __PGIC_MINOR__, __PGIC_PATCHLEVEL__)
+#endif
+
+#if defined(JSON_HEDLEY_PGI_VERSION_CHECK)
+    #undef JSON_HEDLEY_PGI_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_PGI_VERSION)
+    #define JSON_HEDLEY_PGI_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_PGI_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_PGI_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_SUNPRO_VERSION)
+    #undef JSON_HEDLEY_SUNPRO_VERSION
+#endif
+#if defined(__SUNPRO_C) && (__SUNPRO_C > 0x1000)
+    #define JSON_HEDLEY_SUNPRO_VERSION JSON_HEDLEY_VERSION_ENCODE((((__SUNPRO_C >> 16) & 0xf) * 10) + ((__SUNPRO_C >> 12) & 0xf), (((__SUNPRO_C >> 8) & 0xf) * 10) + ((__SUNPRO_C >> 4) & 0xf), (__SUNPRO_C & 0xf) * 10)
+#elif defined(__SUNPRO_C)
+    #define JSON_HEDLEY_SUNPRO_VERSION JSON_HEDLEY_VERSION_ENCODE((__SUNPRO_C >> 8) & 0xf, (__SUNPRO_C >> 4) & 0xf, (__SUNPRO_C) & 0xf)
+#elif defined(__SUNPRO_CC) && (__SUNPRO_CC > 0x1000)
+    #define JSON_HEDLEY_SUNPRO_VERSION JSON_HEDLEY_VERSION_ENCODE((((__SUNPRO_CC >> 16) & 0xf) * 10) + ((__SUNPRO_CC >> 12) & 0xf), (((__SUNPRO_CC >> 8) & 0xf) * 10) + ((__SUNPRO_CC >> 4) & 0xf), (__SUNPRO_CC & 0xf) * 10)
+#elif defined(__SUNPRO_CC)
+    #define JSON_HEDLEY_SUNPRO_VERSION JSON_HEDLEY_VERSION_ENCODE((__SUNPRO_CC >> 8) & 0xf, (__SUNPRO_CC >> 4) & 0xf, (__SUNPRO_CC) & 0xf)
+#endif
+
+#if defined(JSON_HEDLEY_SUNPRO_VERSION_CHECK)
+    #undef JSON_HEDLEY_SUNPRO_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_SUNPRO_VERSION)
+    #define JSON_HEDLEY_SUNPRO_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_SUNPRO_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_SUNPRO_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_EMSCRIPTEN_VERSION)
+    #undef JSON_HEDLEY_EMSCRIPTEN_VERSION
+#endif
+#if defined(__EMSCRIPTEN__)
+    #define JSON_HEDLEY_EMSCRIPTEN_VERSION JSON_HEDLEY_VERSION_ENCODE(__EMSCRIPTEN_major__, __EMSCRIPTEN_minor__, __EMSCRIPTEN_tiny__)
+#endif
+
+#if defined(JSON_HEDLEY_EMSCRIPTEN_VERSION_CHECK)
+    #undef JSON_HEDLEY_EMSCRIPTEN_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_EMSCRIPTEN_VERSION)
+    #define JSON_HEDLEY_EMSCRIPTEN_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_EMSCRIPTEN_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_EMSCRIPTEN_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_ARM_VERSION)
+    #undef JSON_HEDLEY_ARM_VERSION
+#endif
+#if defined(__CC_ARM) && defined(__ARMCOMPILER_VERSION)
+    #define JSON_HEDLEY_ARM_VERSION JSON_HEDLEY_VERSION_ENCODE(__ARMCOMPILER_VERSION / 1000000, (__ARMCOMPILER_VERSION % 1000000) / 10000, (__ARMCOMPILER_VERSION % 10000) / 100)
+#elif defined(__CC_ARM) && defined(__ARMCC_VERSION)
+    #define JSON_HEDLEY_ARM_VERSION JSON_HEDLEY_VERSION_ENCODE(__ARMCC_VERSION / 1000000, (__ARMCC_VERSION % 1000000) / 10000, (__ARMCC_VERSION % 10000) / 100)
+#endif
+
+#if defined(JSON_HEDLEY_ARM_VERSION_CHECK)
+    #undef JSON_HEDLEY_ARM_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_ARM_VERSION)
+    #define JSON_HEDLEY_ARM_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_ARM_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_ARM_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_IBM_VERSION)
+    #undef JSON_HEDLEY_IBM_VERSION
+#endif
+#if defined(__ibmxl__)
+    #define JSON_HEDLEY_IBM_VERSION JSON_HEDLEY_VERSION_ENCODE(__ibmxl_version__, __ibmxl_release__, __ibmxl_modification__)
+#elif defined(__xlC__) && defined(__xlC_ver__)
+    #define JSON_HEDLEY_IBM_VERSION JSON_HEDLEY_VERSION_ENCODE(__xlC__ >> 8, __xlC__ & 0xff, (__xlC_ver__ >> 8) & 0xff)
+#elif defined(__xlC__)
+    #define JSON_HEDLEY_IBM_VERSION JSON_HEDLEY_VERSION_ENCODE(__xlC__ >> 8, __xlC__ & 0xff, 0)
+#endif
+
+#if defined(JSON_HEDLEY_IBM_VERSION_CHECK)
+    #undef JSON_HEDLEY_IBM_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_IBM_VERSION)
+    #define JSON_HEDLEY_IBM_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_IBM_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_IBM_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_TI_VERSION)
+    #undef JSON_HEDLEY_TI_VERSION
+#endif
+#if \
+    defined(__TI_COMPILER_VERSION__) && \
+    ( \
+      defined(__TMS470__) || defined(__TI_ARM__) || \
+      defined(__MSP430__) || \
+      defined(__TMS320C2000__) \
+    )
+#if (__TI_COMPILER_VERSION__ >= 16000000)
+    #define JSON_HEDLEY_TI_VERSION JSON_HEDLEY_VERSION_ENCODE(__TI_COMPILER_VERSION__ / 1000000, (__TI_COMPILER_VERSION__ % 1000000) / 1000, (__TI_COMPILER_VERSION__ % 1000))
+#endif
+#endif
+
+#if defined(JSON_HEDLEY_TI_VERSION_CHECK)
+    #undef JSON_HEDLEY_TI_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_TI_VERSION)
+    #define JSON_HEDLEY_TI_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_TI_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_TI_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_TI_CL2000_VERSION)
+    #undef JSON_HEDLEY_TI_CL2000_VERSION
+#endif
+#if defined(__TI_COMPILER_VERSION__) && defined(__TMS320C2000__)
+    #define JSON_HEDLEY_TI_CL2000_VERSION JSON_HEDLEY_VERSION_ENCODE(__TI_COMPILER_VERSION__ / 1000000, (__TI_COMPILER_VERSION__ % 1000000) / 1000, (__TI_COMPILER_VERSION__ % 1000))
+#endif
+
+#if defined(JSON_HEDLEY_TI_CL2000_VERSION_CHECK)
+    #undef JSON_HEDLEY_TI_CL2000_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_TI_CL2000_VERSION)
+    #define JSON_HEDLEY_TI_CL2000_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_TI_CL2000_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_TI_CL2000_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_TI_CL430_VERSION)
+    #undef JSON_HEDLEY_TI_CL430_VERSION
+#endif
+#if defined(__TI_COMPILER_VERSION__) && defined(__MSP430__)
+    #define JSON_HEDLEY_TI_CL430_VERSION JSON_HEDLEY_VERSION_ENCODE(__TI_COMPILER_VERSION__ / 1000000, (__TI_COMPILER_VERSION__ % 1000000) / 1000, (__TI_COMPILER_VERSION__ % 1000))
+#endif
+
+#if defined(JSON_HEDLEY_TI_CL430_VERSION_CHECK)
+    #undef JSON_HEDLEY_TI_CL430_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_TI_CL430_VERSION)
+    #define JSON_HEDLEY_TI_CL430_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_TI_CL430_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_TI_CL430_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_TI_ARMCL_VERSION)
+    #undef JSON_HEDLEY_TI_ARMCL_VERSION
+#endif
+#if defined(__TI_COMPILER_VERSION__) && (defined(__TMS470__) || defined(__TI_ARM__))
+    #define JSON_HEDLEY_TI_ARMCL_VERSION JSON_HEDLEY_VERSION_ENCODE(__TI_COMPILER_VERSION__ / 1000000, (__TI_COMPILER_VERSION__ % 1000000) / 1000, (__TI_COMPILER_VERSION__ % 1000))
+#endif
+
+#if defined(JSON_HEDLEY_TI_ARMCL_VERSION_CHECK)
+    #undef JSON_HEDLEY_TI_ARMCL_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_TI_ARMCL_VERSION)
+    #define JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_TI_ARMCL_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_TI_CL6X_VERSION)
+    #undef JSON_HEDLEY_TI_CL6X_VERSION
+#endif
+#if defined(__TI_COMPILER_VERSION__) && defined(__TMS320C6X__)
+    #define JSON_HEDLEY_TI_CL6X_VERSION JSON_HEDLEY_VERSION_ENCODE(__TI_COMPILER_VERSION__ / 1000000, (__TI_COMPILER_VERSION__ % 1000000) / 1000, (__TI_COMPILER_VERSION__ % 1000))
+#endif
+
+#if defined(JSON_HEDLEY_TI_CL6X_VERSION_CHECK)
+    #undef JSON_HEDLEY_TI_CL6X_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_TI_CL6X_VERSION)
+    #define JSON_HEDLEY_TI_CL6X_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_TI_CL6X_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_TI_CL6X_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_TI_CL7X_VERSION)
+    #undef JSON_HEDLEY_TI_CL7X_VERSION
+#endif
+#if defined(__TI_COMPILER_VERSION__) && defined(__C7000__)
+    #define JSON_HEDLEY_TI_CL7X_VERSION JSON_HEDLEY_VERSION_ENCODE(__TI_COMPILER_VERSION__ / 1000000, (__TI_COMPILER_VERSION__ % 1000000) / 1000, (__TI_COMPILER_VERSION__ % 1000))
+#endif
+
+#if defined(JSON_HEDLEY_TI_CL7X_VERSION_CHECK)
+    #undef JSON_HEDLEY_TI_CL7X_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_TI_CL7X_VERSION)
+    #define JSON_HEDLEY_TI_CL7X_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_TI_CL7X_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_TI_CL7X_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_TI_CLPRU_VERSION)
+    #undef JSON_HEDLEY_TI_CLPRU_VERSION
+#endif
+#if defined(__TI_COMPILER_VERSION__) && defined(__PRU__)
+    #define JSON_HEDLEY_TI_CLPRU_VERSION JSON_HEDLEY_VERSION_ENCODE(__TI_COMPILER_VERSION__ / 1000000, (__TI_COMPILER_VERSION__ % 1000000) / 1000, (__TI_COMPILER_VERSION__ % 1000))
+#endif
+
+#if defined(JSON_HEDLEY_TI_CLPRU_VERSION_CHECK)
+    #undef JSON_HEDLEY_TI_CLPRU_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_TI_CLPRU_VERSION)
+    #define JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_TI_CLPRU_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_CRAY_VERSION)
+    #undef JSON_HEDLEY_CRAY_VERSION
+#endif
+#if defined(_CRAYC)
+    #if defined(_RELEASE_PATCHLEVEL)
+        #define JSON_HEDLEY_CRAY_VERSION JSON_HEDLEY_VERSION_ENCODE(_RELEASE_MAJOR, _RELEASE_MINOR, _RELEASE_PATCHLEVEL)
+    #else
+        #define JSON_HEDLEY_CRAY_VERSION JSON_HEDLEY_VERSION_ENCODE(_RELEASE_MAJOR, _RELEASE_MINOR, 0)
+    #endif
+#endif
+
+#if defined(JSON_HEDLEY_CRAY_VERSION_CHECK)
+    #undef JSON_HEDLEY_CRAY_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_CRAY_VERSION)
+    #define JSON_HEDLEY_CRAY_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_CRAY_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_CRAY_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_IAR_VERSION)
+    #undef JSON_HEDLEY_IAR_VERSION
+#endif
+#if defined(__IAR_SYSTEMS_ICC__)
+    #if __VER__ > 1000
+        #define JSON_HEDLEY_IAR_VERSION JSON_HEDLEY_VERSION_ENCODE((__VER__ / 1000000), ((__VER__ / 1000) % 1000), (__VER__ % 1000))
+    #else
+        #define JSON_HEDLEY_IAR_VERSION JSON_HEDLEY_VERSION_ENCODE(__VER__ / 100, __VER__ % 100, 0)
+    #endif
+#endif
+
+#if defined(JSON_HEDLEY_IAR_VERSION_CHECK)
+    #undef JSON_HEDLEY_IAR_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_IAR_VERSION)
+    #define JSON_HEDLEY_IAR_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_IAR_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_IAR_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_TINYC_VERSION)
+    #undef JSON_HEDLEY_TINYC_VERSION
+#endif
+#if defined(__TINYC__)
+    #define JSON_HEDLEY_TINYC_VERSION JSON_HEDLEY_VERSION_ENCODE(__TINYC__ / 1000, (__TINYC__ / 100) % 10, __TINYC__ % 100)
+#endif
+
+#if defined(JSON_HEDLEY_TINYC_VERSION_CHECK)
+    #undef JSON_HEDLEY_TINYC_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_TINYC_VERSION)
+    #define JSON_HEDLEY_TINYC_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_TINYC_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_TINYC_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_DMC_VERSION)
+    #undef JSON_HEDLEY_DMC_VERSION
+#endif
+#if defined(__DMC__)
+    #define JSON_HEDLEY_DMC_VERSION JSON_HEDLEY_VERSION_ENCODE(__DMC__ >> 8, (__DMC__ >> 4) & 0xf, __DMC__ & 0xf)
+#endif
+
+#if defined(JSON_HEDLEY_DMC_VERSION_CHECK)
+    #undef JSON_HEDLEY_DMC_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_DMC_VERSION)
+    #define JSON_HEDLEY_DMC_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_DMC_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_DMC_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_COMPCERT_VERSION)
+    #undef JSON_HEDLEY_COMPCERT_VERSION
+#endif
+#if defined(__COMPCERT_VERSION__)
+    #define JSON_HEDLEY_COMPCERT_VERSION JSON_HEDLEY_VERSION_ENCODE(__COMPCERT_VERSION__ / 10000, (__COMPCERT_VERSION__ / 100) % 100, __COMPCERT_VERSION__ % 100)
+#endif
+
+#if defined(JSON_HEDLEY_COMPCERT_VERSION_CHECK)
+    #undef JSON_HEDLEY_COMPCERT_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_COMPCERT_VERSION)
+    #define JSON_HEDLEY_COMPCERT_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_COMPCERT_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_COMPCERT_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_PELLES_VERSION)
+    #undef JSON_HEDLEY_PELLES_VERSION
+#endif
+#if defined(__POCC__)
+    #define JSON_HEDLEY_PELLES_VERSION JSON_HEDLEY_VERSION_ENCODE(__POCC__ / 100, __POCC__ % 100, 0)
+#endif
+
+#if defined(JSON_HEDLEY_PELLES_VERSION_CHECK)
+    #undef JSON_HEDLEY_PELLES_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_PELLES_VERSION)
+    #define JSON_HEDLEY_PELLES_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_PELLES_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_PELLES_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_MCST_LCC_VERSION)
+    #undef JSON_HEDLEY_MCST_LCC_VERSION
+#endif
+#if defined(__LCC__) && defined(__LCC_MINOR__)
+    #define JSON_HEDLEY_MCST_LCC_VERSION JSON_HEDLEY_VERSION_ENCODE(__LCC__ / 100, __LCC__ % 100, __LCC_MINOR__)
+#endif
+
+#if defined(JSON_HEDLEY_MCST_LCC_VERSION_CHECK)
+    #undef JSON_HEDLEY_MCST_LCC_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_MCST_LCC_VERSION)
+    #define JSON_HEDLEY_MCST_LCC_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_MCST_LCC_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_MCST_LCC_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_GCC_VERSION)
+    #undef JSON_HEDLEY_GCC_VERSION
+#endif
+#if \
+    defined(JSON_HEDLEY_GNUC_VERSION) && \
+    !defined(__clang__) && \
+    !defined(JSON_HEDLEY_INTEL_VERSION) && \
+    !defined(JSON_HEDLEY_PGI_VERSION) && \
+    !defined(JSON_HEDLEY_ARM_VERSION) && \
+    !defined(JSON_HEDLEY_CRAY_VERSION) && \
+    !defined(JSON_HEDLEY_TI_VERSION) && \
+    !defined(JSON_HEDLEY_TI_ARMCL_VERSION) && \
+    !defined(JSON_HEDLEY_TI_CL430_VERSION) && \
+    !defined(JSON_HEDLEY_TI_CL2000_VERSION) && \
+    !defined(JSON_HEDLEY_TI_CL6X_VERSION) && \
+    !defined(JSON_HEDLEY_TI_CL7X_VERSION) && \
+    !defined(JSON_HEDLEY_TI_CLPRU_VERSION) && \
+    !defined(__COMPCERT__) && \
+    !defined(JSON_HEDLEY_MCST_LCC_VERSION)
+    #define JSON_HEDLEY_GCC_VERSION JSON_HEDLEY_GNUC_VERSION
+#endif
+
+#if defined(JSON_HEDLEY_GCC_VERSION_CHECK)
+    #undef JSON_HEDLEY_GCC_VERSION_CHECK
+#endif
+#if defined(JSON_HEDLEY_GCC_VERSION)
+    #define JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch) (JSON_HEDLEY_GCC_VERSION >= JSON_HEDLEY_VERSION_ENCODE(major, minor, patch))
+#else
+    #define JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch) (0)
+#endif
+
+#if defined(JSON_HEDLEY_HAS_ATTRIBUTE)
+    #undef JSON_HEDLEY_HAS_ATTRIBUTE
+#endif
+#if \
+  defined(__has_attribute) && \
+  ( \
+    (!defined(JSON_HEDLEY_IAR_VERSION) || JSON_HEDLEY_IAR_VERSION_CHECK(8,5,9)) \
+  )
+#  define JSON_HEDLEY_HAS_ATTRIBUTE(attribute) __has_attribute(attribute)
+#else
+#  define JSON_HEDLEY_HAS_ATTRIBUTE(attribute) (0)
+#endif
+
+#if defined(JSON_HEDLEY_GNUC_HAS_ATTRIBUTE)
+    #undef JSON_HEDLEY_GNUC_HAS_ATTRIBUTE
+#endif
+#if defined(__has_attribute)
+    #define JSON_HEDLEY_GNUC_HAS_ATTRIBUTE(attribute,major,minor,patch) JSON_HEDLEY_HAS_ATTRIBUTE(attribute)
+#else
+    #define JSON_HEDLEY_GNUC_HAS_ATTRIBUTE(attribute,major,minor,patch) JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_GCC_HAS_ATTRIBUTE)
+    #undef JSON_HEDLEY_GCC_HAS_ATTRIBUTE
+#endif
+#if defined(__has_attribute)
+    #define JSON_HEDLEY_GCC_HAS_ATTRIBUTE(attribute,major,minor,patch) JSON_HEDLEY_HAS_ATTRIBUTE(attribute)
+#else
+    #define JSON_HEDLEY_GCC_HAS_ATTRIBUTE(attribute,major,minor,patch) JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_HAS_CPP_ATTRIBUTE)
+    #undef JSON_HEDLEY_HAS_CPP_ATTRIBUTE
+#endif
+#if \
+    defined(__has_cpp_attribute) && \
+    defined(__cplusplus) && \
+    (!defined(JSON_HEDLEY_SUNPRO_VERSION) || JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,15,0))
+    #define JSON_HEDLEY_HAS_CPP_ATTRIBUTE(attribute) __has_cpp_attribute(attribute)
+#else
+    #define JSON_HEDLEY_HAS_CPP_ATTRIBUTE(attribute) (0)
+#endif
+
+#if defined(JSON_HEDLEY_HAS_CPP_ATTRIBUTE_NS)
+    #undef JSON_HEDLEY_HAS_CPP_ATTRIBUTE_NS
+#endif
+#if !defined(__cplusplus) || !defined(__has_cpp_attribute)
+    #define JSON_HEDLEY_HAS_CPP_ATTRIBUTE_NS(ns,attribute) (0)
+#elif \
+    !defined(JSON_HEDLEY_PGI_VERSION) && \
+    !defined(JSON_HEDLEY_IAR_VERSION) && \
+    (!defined(JSON_HEDLEY_SUNPRO_VERSION) || JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,15,0)) && \
+    (!defined(JSON_HEDLEY_MSVC_VERSION) || JSON_HEDLEY_MSVC_VERSION_CHECK(19,20,0))
+    #define JSON_HEDLEY_HAS_CPP_ATTRIBUTE_NS(ns,attribute) JSON_HEDLEY_HAS_CPP_ATTRIBUTE(ns::attribute)
+#else
+    #define JSON_HEDLEY_HAS_CPP_ATTRIBUTE_NS(ns,attribute) (0)
+#endif
+
+#if defined(JSON_HEDLEY_GNUC_HAS_CPP_ATTRIBUTE)
+    #undef JSON_HEDLEY_GNUC_HAS_CPP_ATTRIBUTE
+#endif
+#if defined(__has_cpp_attribute) && defined(__cplusplus)
+    #define JSON_HEDLEY_GNUC_HAS_CPP_ATTRIBUTE(attribute,major,minor,patch) __has_cpp_attribute(attribute)
+#else
+    #define JSON_HEDLEY_GNUC_HAS_CPP_ATTRIBUTE(attribute,major,minor,patch) JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_GCC_HAS_CPP_ATTRIBUTE)
+    #undef JSON_HEDLEY_GCC_HAS_CPP_ATTRIBUTE
+#endif
+#if defined(__has_cpp_attribute) && defined(__cplusplus)
+    #define JSON_HEDLEY_GCC_HAS_CPP_ATTRIBUTE(attribute,major,minor,patch) __has_cpp_attribute(attribute)
+#else
+    #define JSON_HEDLEY_GCC_HAS_CPP_ATTRIBUTE(attribute,major,minor,patch) JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_HAS_BUILTIN)
+    #undef JSON_HEDLEY_HAS_BUILTIN
+#endif
+#if defined(__has_builtin)
+    #define JSON_HEDLEY_HAS_BUILTIN(builtin) __has_builtin(builtin)
+#else
+    #define JSON_HEDLEY_HAS_BUILTIN(builtin) (0)
+#endif
+
+#if defined(JSON_HEDLEY_GNUC_HAS_BUILTIN)
+    #undef JSON_HEDLEY_GNUC_HAS_BUILTIN
+#endif
+#if defined(__has_builtin)
+    #define JSON_HEDLEY_GNUC_HAS_BUILTIN(builtin,major,minor,patch) __has_builtin(builtin)
+#else
+    #define JSON_HEDLEY_GNUC_HAS_BUILTIN(builtin,major,minor,patch) JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_GCC_HAS_BUILTIN)
+    #undef JSON_HEDLEY_GCC_HAS_BUILTIN
+#endif
+#if defined(__has_builtin)
+    #define JSON_HEDLEY_GCC_HAS_BUILTIN(builtin,major,minor,patch) __has_builtin(builtin)
+#else
+    #define JSON_HEDLEY_GCC_HAS_BUILTIN(builtin,major,minor,patch) JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_HAS_FEATURE)
+    #undef JSON_HEDLEY_HAS_FEATURE
+#endif
+#if defined(__has_feature)
+    #define JSON_HEDLEY_HAS_FEATURE(feature) __has_feature(feature)
+#else
+    #define JSON_HEDLEY_HAS_FEATURE(feature) (0)
+#endif
+
+#if defined(JSON_HEDLEY_GNUC_HAS_FEATURE)
+    #undef JSON_HEDLEY_GNUC_HAS_FEATURE
+#endif
+#if defined(__has_feature)
+    #define JSON_HEDLEY_GNUC_HAS_FEATURE(feature,major,minor,patch) __has_feature(feature)
+#else
+    #define JSON_HEDLEY_GNUC_HAS_FEATURE(feature,major,minor,patch) JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_GCC_HAS_FEATURE)
+    #undef JSON_HEDLEY_GCC_HAS_FEATURE
+#endif
+#if defined(__has_feature)
+    #define JSON_HEDLEY_GCC_HAS_FEATURE(feature,major,minor,patch) __has_feature(feature)
+#else
+    #define JSON_HEDLEY_GCC_HAS_FEATURE(feature,major,minor,patch) JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_HAS_EXTENSION)
+    #undef JSON_HEDLEY_HAS_EXTENSION
+#endif
+#if defined(__has_extension)
+    #define JSON_HEDLEY_HAS_EXTENSION(extension) __has_extension(extension)
+#else
+    #define JSON_HEDLEY_HAS_EXTENSION(extension) (0)
+#endif
+
+#if defined(JSON_HEDLEY_GNUC_HAS_EXTENSION)
+    #undef JSON_HEDLEY_GNUC_HAS_EXTENSION
+#endif
+#if defined(__has_extension)
+    #define JSON_HEDLEY_GNUC_HAS_EXTENSION(extension,major,minor,patch) __has_extension(extension)
+#else
+    #define JSON_HEDLEY_GNUC_HAS_EXTENSION(extension,major,minor,patch) JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_GCC_HAS_EXTENSION)
+    #undef JSON_HEDLEY_GCC_HAS_EXTENSION
+#endif
+#if defined(__has_extension)
+    #define JSON_HEDLEY_GCC_HAS_EXTENSION(extension,major,minor,patch) __has_extension(extension)
+#else
+    #define JSON_HEDLEY_GCC_HAS_EXTENSION(extension,major,minor,patch) JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_HAS_DECLSPEC_ATTRIBUTE)
+    #undef JSON_HEDLEY_HAS_DECLSPEC_ATTRIBUTE
+#endif
+#if defined(__has_declspec_attribute)
+    #define JSON_HEDLEY_HAS_DECLSPEC_ATTRIBUTE(attribute) __has_declspec_attribute(attribute)
+#else
+    #define JSON_HEDLEY_HAS_DECLSPEC_ATTRIBUTE(attribute) (0)
+#endif
+
+#if defined(JSON_HEDLEY_GNUC_HAS_DECLSPEC_ATTRIBUTE)
+    #undef JSON_HEDLEY_GNUC_HAS_DECLSPEC_ATTRIBUTE
+#endif
+#if defined(__has_declspec_attribute)
+    #define JSON_HEDLEY_GNUC_HAS_DECLSPEC_ATTRIBUTE(attribute,major,minor,patch) __has_declspec_attribute(attribute)
+#else
+    #define JSON_HEDLEY_GNUC_HAS_DECLSPEC_ATTRIBUTE(attribute,major,minor,patch) JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_GCC_HAS_DECLSPEC_ATTRIBUTE)
+    #undef JSON_HEDLEY_GCC_HAS_DECLSPEC_ATTRIBUTE
+#endif
+#if defined(__has_declspec_attribute)
+    #define JSON_HEDLEY_GCC_HAS_DECLSPEC_ATTRIBUTE(attribute,major,minor,patch) __has_declspec_attribute(attribute)
+#else
+    #define JSON_HEDLEY_GCC_HAS_DECLSPEC_ATTRIBUTE(attribute,major,minor,patch) JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_HAS_WARNING)
+    #undef JSON_HEDLEY_HAS_WARNING
+#endif
+#if defined(__has_warning)
+    #define JSON_HEDLEY_HAS_WARNING(warning) __has_warning(warning)
+#else
+    #define JSON_HEDLEY_HAS_WARNING(warning) (0)
+#endif
+
+#if defined(JSON_HEDLEY_GNUC_HAS_WARNING)
+    #undef JSON_HEDLEY_GNUC_HAS_WARNING
+#endif
+#if defined(__has_warning)
+    #define JSON_HEDLEY_GNUC_HAS_WARNING(warning,major,minor,patch) __has_warning(warning)
+#else
+    #define JSON_HEDLEY_GNUC_HAS_WARNING(warning,major,minor,patch) JSON_HEDLEY_GNUC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_GCC_HAS_WARNING)
+    #undef JSON_HEDLEY_GCC_HAS_WARNING
+#endif
+#if defined(__has_warning)
+    #define JSON_HEDLEY_GCC_HAS_WARNING(warning,major,minor,patch) __has_warning(warning)
+#else
+    #define JSON_HEDLEY_GCC_HAS_WARNING(warning,major,minor,patch) JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if \
+    (defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L)) || \
+    defined(__clang__) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,0,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0) || \
+    JSON_HEDLEY_PGI_VERSION_CHECK(18,4,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,7,0) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(2,0,1) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,1,0) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,0,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    JSON_HEDLEY_CRAY_VERSION_CHECK(5,0,0) || \
+    JSON_HEDLEY_TINYC_VERSION_CHECK(0,9,17) || \
+    JSON_HEDLEY_SUNPRO_VERSION_CHECK(8,0,0) || \
+    (JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) && defined(__C99_PRAGMA_OPERATOR))
+    #define JSON_HEDLEY_PRAGMA(value) _Pragma(#value)
+#elif JSON_HEDLEY_MSVC_VERSION_CHECK(15,0,0)
+    #define JSON_HEDLEY_PRAGMA(value) __pragma(value)
+#else
+    #define JSON_HEDLEY_PRAGMA(value)
+#endif
+
+#if defined(JSON_HEDLEY_DIAGNOSTIC_PUSH)
+    #undef JSON_HEDLEY_DIAGNOSTIC_PUSH
+#endif
+#if defined(JSON_HEDLEY_DIAGNOSTIC_POP)
+    #undef JSON_HEDLEY_DIAGNOSTIC_POP
+#endif
+#if defined(__clang__)
+    #define JSON_HEDLEY_DIAGNOSTIC_PUSH _Pragma("clang diagnostic push")
+    #define JSON_HEDLEY_DIAGNOSTIC_POP _Pragma("clang diagnostic pop")
+#elif JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_PUSH _Pragma("warning(push)")
+    #define JSON_HEDLEY_DIAGNOSTIC_POP _Pragma("warning(pop)")
+#elif JSON_HEDLEY_GCC_VERSION_CHECK(4,6,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_PUSH _Pragma("GCC diagnostic push")
+    #define JSON_HEDLEY_DIAGNOSTIC_POP _Pragma("GCC diagnostic pop")
+#elif \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(15,0,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_PUSH __pragma(warning(push))
+    #define JSON_HEDLEY_DIAGNOSTIC_POP __pragma(warning(pop))
+#elif JSON_HEDLEY_ARM_VERSION_CHECK(5,6,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_PUSH _Pragma("push")
+    #define JSON_HEDLEY_DIAGNOSTIC_POP _Pragma("pop")
+#elif \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,4,0) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(8,1,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_PUSH _Pragma("diag_push")
+    #define JSON_HEDLEY_DIAGNOSTIC_POP _Pragma("diag_pop")
+#elif JSON_HEDLEY_PELLES_VERSION_CHECK(2,90,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_PUSH _Pragma("warning(push)")
+    #define JSON_HEDLEY_DIAGNOSTIC_POP _Pragma("warning(pop)")
+#else
+    #define JSON_HEDLEY_DIAGNOSTIC_PUSH
+    #define JSON_HEDLEY_DIAGNOSTIC_POP
+#endif
+
+/* JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_ is for
+   HEDLEY INTERNAL USE ONLY.  API subject to change without notice. */
+#if defined(JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_)
+    #undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_
+#endif
+#if defined(__cplusplus)
+#  if JSON_HEDLEY_HAS_WARNING("-Wc++98-compat")
+#    if JSON_HEDLEY_HAS_WARNING("-Wc++17-extensions")
+#      if JSON_HEDLEY_HAS_WARNING("-Wc++1z-extensions")
+#        define JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_(xpr) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    _Pragma("clang diagnostic ignored \"-Wc++98-compat\"") \
+    _Pragma("clang diagnostic ignored \"-Wc++17-extensions\"") \
+    _Pragma("clang diagnostic ignored \"-Wc++1z-extensions\"") \
+    xpr \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#      else
+#        define JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_(xpr) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    _Pragma("clang diagnostic ignored \"-Wc++98-compat\"") \
+    _Pragma("clang diagnostic ignored \"-Wc++17-extensions\"") \
+    xpr \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#      endif
+#    else
+#      define JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_(xpr) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    _Pragma("clang diagnostic ignored \"-Wc++98-compat\"") \
+    xpr \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#    endif
+#  endif
+#endif
+#if !defined(JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_(x) x
+#endif
+
+#if defined(JSON_HEDLEY_CONST_CAST)
+    #undef JSON_HEDLEY_CONST_CAST
+#endif
+#if defined(__cplusplus)
+#  define JSON_HEDLEY_CONST_CAST(T, expr) (const_cast<T>(expr))
+#elif \
+  JSON_HEDLEY_HAS_WARNING("-Wcast-qual") || \
+  JSON_HEDLEY_GCC_VERSION_CHECK(4,6,0) || \
+  JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0)
+#  define JSON_HEDLEY_CONST_CAST(T, expr) (__extension__ ({ \
+        JSON_HEDLEY_DIAGNOSTIC_PUSH \
+        JSON_HEDLEY_DIAGNOSTIC_DISABLE_CAST_QUAL \
+        ((T) (expr)); \
+        JSON_HEDLEY_DIAGNOSTIC_POP \
+    }))
+#else
+#  define JSON_HEDLEY_CONST_CAST(T, expr) ((T) (expr))
+#endif
+
+#if defined(JSON_HEDLEY_REINTERPRET_CAST)
+    #undef JSON_HEDLEY_REINTERPRET_CAST
+#endif
+#if defined(__cplusplus)
+    #define JSON_HEDLEY_REINTERPRET_CAST(T, expr) (reinterpret_cast<T>(expr))
+#else
+    #define JSON_HEDLEY_REINTERPRET_CAST(T, expr) ((T) (expr))
+#endif
+
+#if defined(JSON_HEDLEY_STATIC_CAST)
+    #undef JSON_HEDLEY_STATIC_CAST
+#endif
+#if defined(__cplusplus)
+    #define JSON_HEDLEY_STATIC_CAST(T, expr) (static_cast<T>(expr))
+#else
+    #define JSON_HEDLEY_STATIC_CAST(T, expr) ((T) (expr))
+#endif
+
+#if defined(JSON_HEDLEY_CPP_CAST)
+    #undef JSON_HEDLEY_CPP_CAST
+#endif
+#if defined(__cplusplus)
+#  if JSON_HEDLEY_HAS_WARNING("-Wold-style-cast")
+#    define JSON_HEDLEY_CPP_CAST(T, expr) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    _Pragma("clang diagnostic ignored \"-Wold-style-cast\"") \
+    ((T) (expr)) \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#  elif JSON_HEDLEY_IAR_VERSION_CHECK(8,3,0)
+#    define JSON_HEDLEY_CPP_CAST(T, expr) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    _Pragma("diag_suppress=Pe137") \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#  else
+#    define JSON_HEDLEY_CPP_CAST(T, expr) ((T) (expr))
+#  endif
+#else
+#  define JSON_HEDLEY_CPP_CAST(T, expr) (expr)
+#endif
+
+#if defined(JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED)
+    #undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED
+#endif
+#if JSON_HEDLEY_HAS_WARNING("-Wdeprecated-declarations")
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("clang diagnostic ignored \"-Wdeprecated-declarations\"")
+#elif JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("warning(disable:1478 1786)")
+#elif JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED __pragma(warning(disable:1478 1786))
+#elif JSON_HEDLEY_PGI_VERSION_CHECK(20,7,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("diag_suppress 1215,1216,1444,1445")
+#elif JSON_HEDLEY_PGI_VERSION_CHECK(17,10,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("diag_suppress 1215,1444")
+#elif JSON_HEDLEY_GCC_VERSION_CHECK(4,3,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("GCC diagnostic ignored \"-Wdeprecated-declarations\"")
+#elif JSON_HEDLEY_MSVC_VERSION_CHECK(15,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED __pragma(warning(disable:4996))
+#elif JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("diag_suppress 1215,1444")
+#elif \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+    (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("diag_suppress 1291,1718")
+#elif JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,13,0) && !defined(__cplusplus)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("error_messages(off,E_DEPRECATED_ATT,E_DEPRECATED_ATT_MESS)")
+#elif JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,13,0) && defined(__cplusplus)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("error_messages(off,symdeprecated,symdeprecated2)")
+#elif JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("diag_suppress=Pe1444,Pe1215")
+#elif JSON_HEDLEY_PELLES_VERSION_CHECK(2,90,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED _Pragma("warn(disable:2241)")
+#else
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED
+#endif
+
+#if defined(JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS)
+    #undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS
+#endif
+#if JSON_HEDLEY_HAS_WARNING("-Wunknown-pragmas")
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS _Pragma("clang diagnostic ignored \"-Wunknown-pragmas\"")
+#elif JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS _Pragma("warning(disable:161)")
+#elif JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS __pragma(warning(disable:161))
+#elif JSON_HEDLEY_PGI_VERSION_CHECK(17,10,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS _Pragma("diag_suppress 1675")
+#elif JSON_HEDLEY_GCC_VERSION_CHECK(4,3,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS _Pragma("GCC diagnostic ignored \"-Wunknown-pragmas\"")
+#elif JSON_HEDLEY_MSVC_VERSION_CHECK(15,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS __pragma(warning(disable:4068))
+#elif \
+    JSON_HEDLEY_TI_VERSION_CHECK(16,9,0) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(8,0,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,3,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS _Pragma("diag_suppress 163")
+#elif JSON_HEDLEY_TI_CL6X_VERSION_CHECK(8,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS _Pragma("diag_suppress 163")
+#elif JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS _Pragma("diag_suppress=Pe161")
+#elif JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS _Pragma("diag_suppress 161")
+#else
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS
+#endif
+
+#if defined(JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES)
+    #undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES
+#endif
+#if JSON_HEDLEY_HAS_WARNING("-Wunknown-attributes")
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("clang diagnostic ignored \"-Wunknown-attributes\"")
+#elif JSON_HEDLEY_GCC_VERSION_CHECK(4,6,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("GCC diagnostic ignored \"-Wdeprecated-declarations\"")
+#elif JSON_HEDLEY_INTEL_VERSION_CHECK(17,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("warning(disable:1292)")
+#elif JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES __pragma(warning(disable:1292))
+#elif JSON_HEDLEY_MSVC_VERSION_CHECK(19,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES __pragma(warning(disable:5030))
+#elif JSON_HEDLEY_PGI_VERSION_CHECK(20,7,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("diag_suppress 1097,1098")
+#elif JSON_HEDLEY_PGI_VERSION_CHECK(17,10,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("diag_suppress 1097")
+#elif JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,14,0) && defined(__cplusplus)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("error_messages(off,attrskipunsup)")
+#elif \
+    JSON_HEDLEY_TI_VERSION_CHECK(18,1,0) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(8,3,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("diag_suppress 1173")
+#elif JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("diag_suppress=Pe1097")
+#elif JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES _Pragma("diag_suppress 1097")
+#else
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES
+#endif
+
+#if defined(JSON_HEDLEY_DIAGNOSTIC_DISABLE_CAST_QUAL)
+    #undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_CAST_QUAL
+#endif
+#if JSON_HEDLEY_HAS_WARNING("-Wcast-qual")
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_CAST_QUAL _Pragma("clang diagnostic ignored \"-Wcast-qual\"")
+#elif JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_CAST_QUAL _Pragma("warning(disable:2203 2331)")
+#elif JSON_HEDLEY_GCC_VERSION_CHECK(3,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_CAST_QUAL _Pragma("GCC diagnostic ignored \"-Wcast-qual\"")
+#else
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_CAST_QUAL
+#endif
+
+#if defined(JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNUSED_FUNCTION)
+    #undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNUSED_FUNCTION
+#endif
+#if JSON_HEDLEY_HAS_WARNING("-Wunused-function")
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNUSED_FUNCTION _Pragma("clang diagnostic ignored \"-Wunused-function\"")
+#elif JSON_HEDLEY_GCC_VERSION_CHECK(3,4,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNUSED_FUNCTION _Pragma("GCC diagnostic ignored \"-Wunused-function\"")
+#elif JSON_HEDLEY_MSVC_VERSION_CHECK(1,0,0)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNUSED_FUNCTION __pragma(warning(disable:4505))
+#elif JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNUSED_FUNCTION _Pragma("diag_suppress 3142")
+#else
+    #define JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNUSED_FUNCTION
+#endif
+
+#if defined(JSON_HEDLEY_DEPRECATED)
+    #undef JSON_HEDLEY_DEPRECATED
+#endif
+#if defined(JSON_HEDLEY_DEPRECATED_FOR)
+    #undef JSON_HEDLEY_DEPRECATED_FOR
+#endif
+#if \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(14,0,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_DEPRECATED(since) __declspec(deprecated("Since " # since))
+    #define JSON_HEDLEY_DEPRECATED_FOR(since, replacement) __declspec(deprecated("Since " #since "; use " #replacement))
+#elif \
+    (JSON_HEDLEY_HAS_EXTENSION(attribute_deprecated_with_message) && !defined(JSON_HEDLEY_IAR_VERSION)) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(4,5,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(5,6,0) || \
+    JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,13,0) || \
+    JSON_HEDLEY_PGI_VERSION_CHECK(17,10,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(18,1,0) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(18,1,0) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(8,3,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,3,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_DEPRECATED(since) __attribute__((__deprecated__("Since " #since)))
+    #define JSON_HEDLEY_DEPRECATED_FOR(since, replacement) __attribute__((__deprecated__("Since " #since "; use " #replacement)))
+#elif defined(__cplusplus) && (__cplusplus >= 201402L)
+    #define JSON_HEDLEY_DEPRECATED(since) JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[deprecated("Since " #since)]])
+    #define JSON_HEDLEY_DEPRECATED_FOR(since, replacement) JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[deprecated("Since " #since "; use " #replacement)]])
+#elif \
+    JSON_HEDLEY_HAS_ATTRIBUTE(deprecated) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,1,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+    (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10) || \
+    JSON_HEDLEY_IAR_VERSION_CHECK(8,10,0)
+    #define JSON_HEDLEY_DEPRECATED(since) __attribute__((__deprecated__))
+    #define JSON_HEDLEY_DEPRECATED_FOR(since, replacement) __attribute__((__deprecated__))
+#elif \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(13,10,0) || \
+    JSON_HEDLEY_PELLES_VERSION_CHECK(6,50,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_DEPRECATED(since) __declspec(deprecated)
+    #define JSON_HEDLEY_DEPRECATED_FOR(since, replacement) __declspec(deprecated)
+#elif JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0)
+    #define JSON_HEDLEY_DEPRECATED(since) _Pragma("deprecated")
+    #define JSON_HEDLEY_DEPRECATED_FOR(since, replacement) _Pragma("deprecated")
+#else
+    #define JSON_HEDLEY_DEPRECATED(since)
+    #define JSON_HEDLEY_DEPRECATED_FOR(since, replacement)
+#endif
+
+#if defined(JSON_HEDLEY_UNAVAILABLE)
+    #undef JSON_HEDLEY_UNAVAILABLE
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(warning) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(4,3,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_UNAVAILABLE(available_since) __attribute__((__warning__("Not available until " #available_since)))
+#else
+    #define JSON_HEDLEY_UNAVAILABLE(available_since)
+#endif
+
+#if defined(JSON_HEDLEY_WARN_UNUSED_RESULT)
+    #undef JSON_HEDLEY_WARN_UNUSED_RESULT
+#endif
+#if defined(JSON_HEDLEY_WARN_UNUSED_RESULT_MSG)
+    #undef JSON_HEDLEY_WARN_UNUSED_RESULT_MSG
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(warn_unused_result) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,4,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+    (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    (JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,15,0) && defined(__cplusplus)) || \
+    JSON_HEDLEY_PGI_VERSION_CHECK(17,10,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT __attribute__((__warn_unused_result__))
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT_MSG(msg) __attribute__((__warn_unused_result__))
+#elif (JSON_HEDLEY_HAS_CPP_ATTRIBUTE(nodiscard) >= 201907L)
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[nodiscard]])
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT_MSG(msg) JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[nodiscard(msg)]])
+#elif JSON_HEDLEY_HAS_CPP_ATTRIBUTE(nodiscard)
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[nodiscard]])
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT_MSG(msg) JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[nodiscard]])
+#elif defined(_Check_return_) /* SAL */
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT _Check_return_
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT_MSG(msg) _Check_return_
+#else
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT
+    #define JSON_HEDLEY_WARN_UNUSED_RESULT_MSG(msg)
+#endif
+
+#if defined(JSON_HEDLEY_SENTINEL)
+    #undef JSON_HEDLEY_SENTINEL
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(sentinel) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(4,0,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(5,4,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_SENTINEL(position) __attribute__((__sentinel__(position)))
+#else
+    #define JSON_HEDLEY_SENTINEL(position)
+#endif
+
+#if defined(JSON_HEDLEY_NO_RETURN)
+    #undef JSON_HEDLEY_NO_RETURN
+#endif
+#if JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0)
+    #define JSON_HEDLEY_NO_RETURN __noreturn
+#elif \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_NO_RETURN __attribute__((__noreturn__))
+#elif defined(__STDC_VERSION__) && __STDC_VERSION__ >= 201112L
+    #define JSON_HEDLEY_NO_RETURN _Noreturn
+#elif defined(__cplusplus) && (__cplusplus >= 201103L)
+    #define JSON_HEDLEY_NO_RETURN JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[noreturn]])
+#elif \
+    JSON_HEDLEY_HAS_ATTRIBUTE(noreturn) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,2,0) || \
+    JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,11,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+    (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    JSON_HEDLEY_IAR_VERSION_CHECK(8,10,0)
+    #define JSON_HEDLEY_NO_RETURN __attribute__((__noreturn__))
+#elif JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,10,0)
+    #define JSON_HEDLEY_NO_RETURN _Pragma("does_not_return")
+#elif \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(13,10,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_NO_RETURN __declspec(noreturn)
+#elif JSON_HEDLEY_TI_CL6X_VERSION_CHECK(6,0,0) && defined(__cplusplus)
+    #define JSON_HEDLEY_NO_RETURN _Pragma("FUNC_NEVER_RETURNS;")
+#elif JSON_HEDLEY_COMPCERT_VERSION_CHECK(3,2,0)
+    #define JSON_HEDLEY_NO_RETURN __attribute((noreturn))
+#elif JSON_HEDLEY_PELLES_VERSION_CHECK(9,0,0)
+    #define JSON_HEDLEY_NO_RETURN __declspec(noreturn)
+#else
+    #define JSON_HEDLEY_NO_RETURN
+#endif
+
+#if defined(JSON_HEDLEY_NO_ESCAPE)
+    #undef JSON_HEDLEY_NO_ESCAPE
+#endif
+#if JSON_HEDLEY_HAS_ATTRIBUTE(noescape)
+    #define JSON_HEDLEY_NO_ESCAPE __attribute__((__noescape__))
+#else
+    #define JSON_HEDLEY_NO_ESCAPE
+#endif
+
+#if defined(JSON_HEDLEY_UNREACHABLE)
+    #undef JSON_HEDLEY_UNREACHABLE
+#endif
+#if defined(JSON_HEDLEY_UNREACHABLE_RETURN)
+    #undef JSON_HEDLEY_UNREACHABLE_RETURN
+#endif
+#if defined(JSON_HEDLEY_ASSUME)
+    #undef JSON_HEDLEY_ASSUME
+#endif
+#if \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(13,10,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_ASSUME(expr) __assume(expr)
+#elif JSON_HEDLEY_HAS_BUILTIN(__builtin_assume)
+    #define JSON_HEDLEY_ASSUME(expr) __builtin_assume(expr)
+#elif \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,2,0) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(4,0,0)
+    #if defined(__cplusplus)
+        #define JSON_HEDLEY_ASSUME(expr) std::_nassert(expr)
+    #else
+        #define JSON_HEDLEY_ASSUME(expr) _nassert(expr)
+    #endif
+#endif
+#if \
+    (JSON_HEDLEY_HAS_BUILTIN(__builtin_unreachable) && (!defined(JSON_HEDLEY_ARM_VERSION))) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(4,5,0) || \
+    JSON_HEDLEY_PGI_VERSION_CHECK(18,10,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(13,1,5) || \
+    JSON_HEDLEY_CRAY_VERSION_CHECK(10,0,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_UNREACHABLE() __builtin_unreachable()
+#elif defined(JSON_HEDLEY_ASSUME)
+    #define JSON_HEDLEY_UNREACHABLE() JSON_HEDLEY_ASSUME(0)
+#endif
+#if !defined(JSON_HEDLEY_ASSUME)
+    #if defined(JSON_HEDLEY_UNREACHABLE)
+        #define JSON_HEDLEY_ASSUME(expr) JSON_HEDLEY_STATIC_CAST(void, ((expr) ? 1 : (JSON_HEDLEY_UNREACHABLE(), 1)))
+    #else
+        #define JSON_HEDLEY_ASSUME(expr) JSON_HEDLEY_STATIC_CAST(void, expr)
+    #endif
+#endif
+#if defined(JSON_HEDLEY_UNREACHABLE)
+    #if  \
+        JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,2,0) || \
+        JSON_HEDLEY_TI_CL6X_VERSION_CHECK(4,0,0)
+        #define JSON_HEDLEY_UNREACHABLE_RETURN(value) return (JSON_HEDLEY_STATIC_CAST(void, JSON_HEDLEY_ASSUME(0)), (value))
+    #else
+        #define JSON_HEDLEY_UNREACHABLE_RETURN(value) JSON_HEDLEY_UNREACHABLE()
+    #endif
+#else
+    #define JSON_HEDLEY_UNREACHABLE_RETURN(value) return (value)
+#endif
+#if !defined(JSON_HEDLEY_UNREACHABLE)
+    #define JSON_HEDLEY_UNREACHABLE() JSON_HEDLEY_ASSUME(0)
+#endif
+
+JSON_HEDLEY_DIAGNOSTIC_PUSH
+#if JSON_HEDLEY_HAS_WARNING("-Wpedantic")
+    #pragma clang diagnostic ignored "-Wpedantic"
+#endif
+#if JSON_HEDLEY_HAS_WARNING("-Wc++98-compat-pedantic") && defined(__cplusplus)
+    #pragma clang diagnostic ignored "-Wc++98-compat-pedantic"
+#endif
+#if JSON_HEDLEY_GCC_HAS_WARNING("-Wvariadic-macros",4,0,0)
+    #if defined(__clang__)
+        #pragma clang diagnostic ignored "-Wvariadic-macros"
+    #elif defined(JSON_HEDLEY_GCC_VERSION)
+        #pragma GCC diagnostic ignored "-Wvariadic-macros"
+    #endif
+#endif
+#if defined(JSON_HEDLEY_NON_NULL)
+    #undef JSON_HEDLEY_NON_NULL
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(nonnull) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,3,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0)
+    #define JSON_HEDLEY_NON_NULL(...) __attribute__((__nonnull__(__VA_ARGS__)))
+#else
+    #define JSON_HEDLEY_NON_NULL(...)
+#endif
+JSON_HEDLEY_DIAGNOSTIC_POP
+
+#if defined(JSON_HEDLEY_PRINTF_FORMAT)
+    #undef JSON_HEDLEY_PRINTF_FORMAT
+#endif
+#if defined(__MINGW32__) && JSON_HEDLEY_GCC_HAS_ATTRIBUTE(format,4,4,0) && !defined(__USE_MINGW_ANSI_STDIO)
+    #define JSON_HEDLEY_PRINTF_FORMAT(string_idx,first_to_check) __attribute__((__format__(ms_printf, string_idx, first_to_check)))
+#elif defined(__MINGW32__) && JSON_HEDLEY_GCC_HAS_ATTRIBUTE(format,4,4,0) && defined(__USE_MINGW_ANSI_STDIO)
+    #define JSON_HEDLEY_PRINTF_FORMAT(string_idx,first_to_check) __attribute__((__format__(gnu_printf, string_idx, first_to_check)))
+#elif \
+    JSON_HEDLEY_HAS_ATTRIBUTE(format) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,1,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(5,6,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+    (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_PRINTF_FORMAT(string_idx,first_to_check) __attribute__((__format__(__printf__, string_idx, first_to_check)))
+#elif JSON_HEDLEY_PELLES_VERSION_CHECK(6,0,0)
+    #define JSON_HEDLEY_PRINTF_FORMAT(string_idx,first_to_check) __declspec(vaformat(printf,string_idx,first_to_check))
+#else
+    #define JSON_HEDLEY_PRINTF_FORMAT(string_idx,first_to_check)
+#endif
+
+#if defined(JSON_HEDLEY_CONSTEXPR)
+    #undef JSON_HEDLEY_CONSTEXPR
+#endif
+#if defined(__cplusplus)
+    #if __cplusplus >= 201103L
+        #define JSON_HEDLEY_CONSTEXPR JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_(constexpr)
+    #endif
+#endif
+#if !defined(JSON_HEDLEY_CONSTEXPR)
+    #define JSON_HEDLEY_CONSTEXPR
+#endif
+
+#if defined(JSON_HEDLEY_PREDICT)
+    #undef JSON_HEDLEY_PREDICT
+#endif
+#if defined(JSON_HEDLEY_LIKELY)
+    #undef JSON_HEDLEY_LIKELY
+#endif
+#if defined(JSON_HEDLEY_UNLIKELY)
+    #undef JSON_HEDLEY_UNLIKELY
+#endif
+#if defined(JSON_HEDLEY_UNPREDICTABLE)
+    #undef JSON_HEDLEY_UNPREDICTABLE
+#endif
+#if JSON_HEDLEY_HAS_BUILTIN(__builtin_unpredictable)
+    #define JSON_HEDLEY_UNPREDICTABLE(expr) __builtin_unpredictable((expr))
+#endif
+#if \
+  (JSON_HEDLEY_HAS_BUILTIN(__builtin_expect_with_probability) && !defined(JSON_HEDLEY_PGI_VERSION)) || \
+  JSON_HEDLEY_GCC_VERSION_CHECK(9,0,0) || \
+  JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+#  define JSON_HEDLEY_PREDICT(expr, value, probability) __builtin_expect_with_probability(  (expr), (value), (probability))
+#  define JSON_HEDLEY_PREDICT_TRUE(expr, probability)   __builtin_expect_with_probability(!!(expr),    1   , (probability))
+#  define JSON_HEDLEY_PREDICT_FALSE(expr, probability)  __builtin_expect_with_probability(!!(expr),    0   , (probability))
+#  define JSON_HEDLEY_LIKELY(expr)                      __builtin_expect                 (!!(expr),    1                  )
+#  define JSON_HEDLEY_UNLIKELY(expr)                    __builtin_expect                 (!!(expr),    0                  )
+#elif \
+  (JSON_HEDLEY_HAS_BUILTIN(__builtin_expect) && !defined(JSON_HEDLEY_INTEL_CL_VERSION)) || \
+  JSON_HEDLEY_GCC_VERSION_CHECK(3,0,0) || \
+  JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+  (JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,15,0) && defined(__cplusplus)) || \
+  JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+  JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) || \
+  JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+  JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,7,0) || \
+  JSON_HEDLEY_TI_CL430_VERSION_CHECK(3,1,0) || \
+  JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,1,0) || \
+  JSON_HEDLEY_TI_CL6X_VERSION_CHECK(6,1,0) || \
+  JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+  JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+  JSON_HEDLEY_TINYC_VERSION_CHECK(0,9,27) || \
+  JSON_HEDLEY_CRAY_VERSION_CHECK(8,1,0) || \
+  JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+#  define JSON_HEDLEY_PREDICT(expr, expected, probability) \
+    (((probability) >= 0.9) ? __builtin_expect((expr), (expected)) : (JSON_HEDLEY_STATIC_CAST(void, expected), (expr)))
+#  define JSON_HEDLEY_PREDICT_TRUE(expr, probability) \
+    (__extension__ ({ \
+        double hedley_probability_ = (probability); \
+        ((hedley_probability_ >= 0.9) ? __builtin_expect(!!(expr), 1) : ((hedley_probability_ <= 0.1) ? __builtin_expect(!!(expr), 0) : !!(expr))); \
+    }))
+#  define JSON_HEDLEY_PREDICT_FALSE(expr, probability) \
+    (__extension__ ({ \
+        double hedley_probability_ = (probability); \
+        ((hedley_probability_ >= 0.9) ? __builtin_expect(!!(expr), 0) : ((hedley_probability_ <= 0.1) ? __builtin_expect(!!(expr), 1) : !!(expr))); \
+    }))
+#  define JSON_HEDLEY_LIKELY(expr)   __builtin_expect(!!(expr), 1)
+#  define JSON_HEDLEY_UNLIKELY(expr) __builtin_expect(!!(expr), 0)
+#else
+#  define JSON_HEDLEY_PREDICT(expr, expected, probability) (JSON_HEDLEY_STATIC_CAST(void, expected), (expr))
+#  define JSON_HEDLEY_PREDICT_TRUE(expr, probability) (!!(expr))
+#  define JSON_HEDLEY_PREDICT_FALSE(expr, probability) (!!(expr))
+#  define JSON_HEDLEY_LIKELY(expr) (!!(expr))
+#  define JSON_HEDLEY_UNLIKELY(expr) (!!(expr))
+#endif
+#if !defined(JSON_HEDLEY_UNPREDICTABLE)
+    #define JSON_HEDLEY_UNPREDICTABLE(expr) JSON_HEDLEY_PREDICT(expr, 1, 0.5)
+#endif
+
+#if defined(JSON_HEDLEY_MALLOC)
+    #undef JSON_HEDLEY_MALLOC
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(malloc) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,1,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,11,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(12,1,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+    (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_MALLOC __attribute__((__malloc__))
+#elif JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,10,0)
+    #define JSON_HEDLEY_MALLOC _Pragma("returns_new_memory")
+#elif \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(14,0,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_MALLOC __declspec(restrict)
+#else
+    #define JSON_HEDLEY_MALLOC
+#endif
+
+#if defined(JSON_HEDLEY_PURE)
+    #undef JSON_HEDLEY_PURE
+#endif
+#if \
+  JSON_HEDLEY_HAS_ATTRIBUTE(pure) || \
+  JSON_HEDLEY_GCC_VERSION_CHECK(2,96,0) || \
+  JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+  JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,11,0) || \
+  JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+  JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) || \
+  JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+  (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+  JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+  (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+  JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+  (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+  JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+  (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+  JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+  JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+  JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+  JSON_HEDLEY_PGI_VERSION_CHECK(17,10,0) || \
+  JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+#  define JSON_HEDLEY_PURE __attribute__((__pure__))
+#elif JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,10,0)
+#  define JSON_HEDLEY_PURE _Pragma("does_not_write_global_data")
+#elif defined(__cplusplus) && \
+    ( \
+      JSON_HEDLEY_TI_CL430_VERSION_CHECK(2,0,1) || \
+      JSON_HEDLEY_TI_CL6X_VERSION_CHECK(4,0,0) || \
+      JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) \
+    )
+#  define JSON_HEDLEY_PURE _Pragma("FUNC_IS_PURE;")
+#else
+#  define JSON_HEDLEY_PURE
+#endif
+
+#if defined(JSON_HEDLEY_CONST)
+    #undef JSON_HEDLEY_CONST
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(const) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(2,5,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,11,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+    (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    JSON_HEDLEY_PGI_VERSION_CHECK(17,10,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_CONST __attribute__((__const__))
+#elif \
+    JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,10,0)
+    #define JSON_HEDLEY_CONST _Pragma("no_side_effect")
+#else
+    #define JSON_HEDLEY_CONST JSON_HEDLEY_PURE
+#endif
+
+#if defined(JSON_HEDLEY_RESTRICT)
+    #undef JSON_HEDLEY_RESTRICT
+#endif
+#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) && !defined(__cplusplus)
+    #define JSON_HEDLEY_RESTRICT restrict
+#elif \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,1,0) || \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(14,0,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) || \
+    JSON_HEDLEY_PGI_VERSION_CHECK(17,10,0) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,2,4) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(8,1,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    (JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,14,0) && defined(__cplusplus)) || \
+    JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0) || \
+    defined(__clang__) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_RESTRICT __restrict
+#elif JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,3,0) && !defined(__cplusplus)
+    #define JSON_HEDLEY_RESTRICT _Restrict
+#else
+    #define JSON_HEDLEY_RESTRICT
+#endif
+
+#if defined(JSON_HEDLEY_INLINE)
+    #undef JSON_HEDLEY_INLINE
+#endif
+#if \
+    (defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L)) || \
+    (defined(__cplusplus) && (__cplusplus >= 199711L))
+    #define JSON_HEDLEY_INLINE inline
+#elif \
+    defined(JSON_HEDLEY_GCC_VERSION) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(6,2,0)
+    #define JSON_HEDLEY_INLINE __inline__
+#elif \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(12,0,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,1,0) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(3,1,0) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,2,0) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(8,0,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_INLINE __inline
+#else
+    #define JSON_HEDLEY_INLINE
+#endif
+
+#if defined(JSON_HEDLEY_ALWAYS_INLINE)
+    #undef JSON_HEDLEY_ALWAYS_INLINE
+#endif
+#if \
+  JSON_HEDLEY_HAS_ATTRIBUTE(always_inline) || \
+  JSON_HEDLEY_GCC_VERSION_CHECK(4,0,0) || \
+  JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+  JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,11,0) || \
+  JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+  JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) || \
+  JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+  (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+  JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+  (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+  JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+  (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+  JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+  (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+  JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+  JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+  JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+  JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10) || \
+  JSON_HEDLEY_IAR_VERSION_CHECK(8,10,0)
+#  define JSON_HEDLEY_ALWAYS_INLINE __attribute__((__always_inline__)) JSON_HEDLEY_INLINE
+#elif \
+  JSON_HEDLEY_MSVC_VERSION_CHECK(12,0,0) || \
+  JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+#  define JSON_HEDLEY_ALWAYS_INLINE __forceinline
+#elif defined(__cplusplus) && \
+    ( \
+      JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+      JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+      JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+      JSON_HEDLEY_TI_CL6X_VERSION_CHECK(6,1,0) || \
+      JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+      JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) \
+    )
+#  define JSON_HEDLEY_ALWAYS_INLINE _Pragma("FUNC_ALWAYS_INLINE;")
+#elif JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0)
+#  define JSON_HEDLEY_ALWAYS_INLINE _Pragma("inline=forced")
+#else
+#  define JSON_HEDLEY_ALWAYS_INLINE JSON_HEDLEY_INLINE
+#endif
+
+#if defined(JSON_HEDLEY_NEVER_INLINE)
+    #undef JSON_HEDLEY_NEVER_INLINE
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(noinline) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(4,0,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,11,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(10,1,0) || \
+    JSON_HEDLEY_TI_VERSION_CHECK(15,12,0) || \
+    (JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(4,8,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_ARMCL_VERSION_CHECK(5,2,0) || \
+    (JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL2000_VERSION_CHECK(6,4,0) || \
+    (JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,0,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL430_VERSION_CHECK(4,3,0) || \
+    (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) || \
+    JSON_HEDLEY_TI_CL7X_VERSION_CHECK(1,2,0) || \
+    JSON_HEDLEY_TI_CLPRU_VERSION_CHECK(2,1,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10) || \
+    JSON_HEDLEY_IAR_VERSION_CHECK(8,10,0)
+    #define JSON_HEDLEY_NEVER_INLINE __attribute__((__noinline__))
+#elif \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(13,10,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_NEVER_INLINE __declspec(noinline)
+#elif JSON_HEDLEY_PGI_VERSION_CHECK(10,2,0)
+    #define JSON_HEDLEY_NEVER_INLINE _Pragma("noinline")
+#elif JSON_HEDLEY_TI_CL6X_VERSION_CHECK(6,0,0) && defined(__cplusplus)
+    #define JSON_HEDLEY_NEVER_INLINE _Pragma("FUNC_CANNOT_INLINE;")
+#elif JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0)
+    #define JSON_HEDLEY_NEVER_INLINE _Pragma("inline=never")
+#elif JSON_HEDLEY_COMPCERT_VERSION_CHECK(3,2,0)
+    #define JSON_HEDLEY_NEVER_INLINE __attribute((noinline))
+#elif JSON_HEDLEY_PELLES_VERSION_CHECK(9,0,0)
+    #define JSON_HEDLEY_NEVER_INLINE __declspec(noinline)
+#else
+    #define JSON_HEDLEY_NEVER_INLINE
+#endif
+
+#if defined(JSON_HEDLEY_PRIVATE)
+    #undef JSON_HEDLEY_PRIVATE
+#endif
+#if defined(JSON_HEDLEY_PUBLIC)
+    #undef JSON_HEDLEY_PUBLIC
+#endif
+#if defined(JSON_HEDLEY_IMPORT)
+    #undef JSON_HEDLEY_IMPORT
+#endif
+#if defined(_WIN32) || defined(__CYGWIN__)
+#  define JSON_HEDLEY_PRIVATE
+#  define JSON_HEDLEY_PUBLIC   __declspec(dllexport)
+#  define JSON_HEDLEY_IMPORT   __declspec(dllimport)
+#else
+#  if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(visibility) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,3,0) || \
+    JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,11,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(13,1,0) || \
+    ( \
+      defined(__TI_EABI__) && \
+      ( \
+        (JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,2,0) && defined(__TI_GNU_ATTRIBUTE_SUPPORT__)) || \
+        JSON_HEDLEY_TI_CL6X_VERSION_CHECK(7,5,0) \
+      ) \
+    ) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+#    define JSON_HEDLEY_PRIVATE __attribute__((__visibility__("hidden")))
+#    define JSON_HEDLEY_PUBLIC  __attribute__((__visibility__("default")))
+#  else
+#    define JSON_HEDLEY_PRIVATE
+#    define JSON_HEDLEY_PUBLIC
+#  endif
+#  define JSON_HEDLEY_IMPORT    extern
+#endif
+
+#if defined(JSON_HEDLEY_NO_THROW)
+    #undef JSON_HEDLEY_NO_THROW
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(nothrow) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,3,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_NO_THROW __attribute__((__nothrow__))
+#elif \
+    JSON_HEDLEY_MSVC_VERSION_CHECK(13,1,0) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0)
+    #define JSON_HEDLEY_NO_THROW __declspec(nothrow)
+#else
+    #define JSON_HEDLEY_NO_THROW
+#endif
+
+#if defined(JSON_HEDLEY_FALL_THROUGH)
+    #undef JSON_HEDLEY_FALL_THROUGH
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(fallthrough) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(7,0,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_FALL_THROUGH __attribute__((__fallthrough__))
+#elif JSON_HEDLEY_HAS_CPP_ATTRIBUTE_NS(clang,fallthrough)
+    #define JSON_HEDLEY_FALL_THROUGH JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[clang::fallthrough]])
+#elif JSON_HEDLEY_HAS_CPP_ATTRIBUTE(fallthrough)
+    #define JSON_HEDLEY_FALL_THROUGH JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_([[fallthrough]])
+#elif defined(__fallthrough) /* SAL */
+    #define JSON_HEDLEY_FALL_THROUGH __fallthrough
+#else
+    #define JSON_HEDLEY_FALL_THROUGH
+#endif
+
+#if defined(JSON_HEDLEY_RETURNS_NON_NULL)
+    #undef JSON_HEDLEY_RETURNS_NON_NULL
+#endif
+#if \
+    JSON_HEDLEY_HAS_ATTRIBUTE(returns_nonnull) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(4,9,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_RETURNS_NON_NULL __attribute__((__returns_nonnull__))
+#elif defined(_Ret_notnull_) /* SAL */
+    #define JSON_HEDLEY_RETURNS_NON_NULL _Ret_notnull_
+#else
+    #define JSON_HEDLEY_RETURNS_NON_NULL
+#endif
+
+#if defined(JSON_HEDLEY_ARRAY_PARAM)
+    #undef JSON_HEDLEY_ARRAY_PARAM
+#endif
+#if \
+    defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) && \
+    !defined(__STDC_NO_VLA__) && \
+    !defined(__cplusplus) && \
+    !defined(JSON_HEDLEY_PGI_VERSION) && \
+    !defined(JSON_HEDLEY_TINYC_VERSION)
+    #define JSON_HEDLEY_ARRAY_PARAM(name) (name)
+#else
+    #define JSON_HEDLEY_ARRAY_PARAM(name)
+#endif
+
+#if defined(JSON_HEDLEY_IS_CONSTANT)
+    #undef JSON_HEDLEY_IS_CONSTANT
+#endif
+#if defined(JSON_HEDLEY_REQUIRE_CONSTEXPR)
+    #undef JSON_HEDLEY_REQUIRE_CONSTEXPR
+#endif
+/* JSON_HEDLEY_IS_CONSTEXPR_ is for
+   HEDLEY INTERNAL USE ONLY.  API subject to change without notice. */
+#if defined(JSON_HEDLEY_IS_CONSTEXPR_)
+    #undef JSON_HEDLEY_IS_CONSTEXPR_
+#endif
+#if \
+    JSON_HEDLEY_HAS_BUILTIN(__builtin_constant_p) || \
+    JSON_HEDLEY_GCC_VERSION_CHECK(3,4,0) || \
+    JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+    JSON_HEDLEY_TINYC_VERSION_CHECK(0,9,19) || \
+    JSON_HEDLEY_ARM_VERSION_CHECK(4,1,0) || \
+    JSON_HEDLEY_IBM_VERSION_CHECK(13,1,0) || \
+    JSON_HEDLEY_TI_CL6X_VERSION_CHECK(6,1,0) || \
+    (JSON_HEDLEY_SUNPRO_VERSION_CHECK(5,10,0) && !defined(__cplusplus)) || \
+    JSON_HEDLEY_CRAY_VERSION_CHECK(8,1,0) || \
+    JSON_HEDLEY_MCST_LCC_VERSION_CHECK(1,25,10)
+    #define JSON_HEDLEY_IS_CONSTANT(expr) __builtin_constant_p(expr)
+#endif
+#if !defined(__cplusplus)
+#  if \
+       JSON_HEDLEY_HAS_BUILTIN(__builtin_types_compatible_p) || \
+       JSON_HEDLEY_GCC_VERSION_CHECK(3,4,0) || \
+       JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+       JSON_HEDLEY_IBM_VERSION_CHECK(13,1,0) || \
+       JSON_HEDLEY_CRAY_VERSION_CHECK(8,1,0) || \
+       JSON_HEDLEY_ARM_VERSION_CHECK(5,4,0) || \
+       JSON_HEDLEY_TINYC_VERSION_CHECK(0,9,24)
+#if defined(__INTPTR_TYPE__)
+    #define JSON_HEDLEY_IS_CONSTEXPR_(expr) __builtin_types_compatible_p(__typeof__((1 ? (void*) ((__INTPTR_TYPE__) ((expr) * 0)) : (int*) 0)), int*)
+#else
+    #include <stdint.h>
+    #define JSON_HEDLEY_IS_CONSTEXPR_(expr) __builtin_types_compatible_p(__typeof__((1 ? (void*) ((intptr_t) ((expr) * 0)) : (int*) 0)), int*)
+#endif
+#  elif \
+       ( \
+          defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 201112L) && \
+          !defined(JSON_HEDLEY_SUNPRO_VERSION) && \
+          !defined(JSON_HEDLEY_PGI_VERSION) && \
+          !defined(JSON_HEDLEY_IAR_VERSION)) || \
+       (JSON_HEDLEY_HAS_EXTENSION(c_generic_selections) && !defined(JSON_HEDLEY_IAR_VERSION)) || \
+       JSON_HEDLEY_GCC_VERSION_CHECK(4,9,0) || \
+       JSON_HEDLEY_INTEL_VERSION_CHECK(17,0,0) || \
+       JSON_HEDLEY_IBM_VERSION_CHECK(12,1,0) || \
+       JSON_HEDLEY_ARM_VERSION_CHECK(5,3,0)
+#if defined(__INTPTR_TYPE__)
+    #define JSON_HEDLEY_IS_CONSTEXPR_(expr) _Generic((1 ? (void*) ((__INTPTR_TYPE__) ((expr) * 0)) : (int*) 0), int*: 1, void*: 0)
+#else
+    #include <stdint.h>
+    #define JSON_HEDLEY_IS_CONSTEXPR_(expr) _Generic((1 ? (void*) ((intptr_t) * 0) : (int*) 0), int*: 1, void*: 0)
+#endif
+#  elif \
+       defined(JSON_HEDLEY_GCC_VERSION) || \
+       defined(JSON_HEDLEY_INTEL_VERSION) || \
+       defined(JSON_HEDLEY_TINYC_VERSION) || \
+       defined(JSON_HEDLEY_TI_ARMCL_VERSION) || \
+       JSON_HEDLEY_TI_CL430_VERSION_CHECK(18,12,0) || \
+       defined(JSON_HEDLEY_TI_CL2000_VERSION) || \
+       defined(JSON_HEDLEY_TI_CL6X_VERSION) || \
+       defined(JSON_HEDLEY_TI_CL7X_VERSION) || \
+       defined(JSON_HEDLEY_TI_CLPRU_VERSION) || \
+       defined(__clang__)
+#    define JSON_HEDLEY_IS_CONSTEXPR_(expr) ( \
+        sizeof(void) != \
+        sizeof(*( \
+                  1 ? \
+                  ((void*) ((expr) * 0L) ) : \
+((struct { char v[sizeof(void) * 2]; } *) 1) \
+                ) \
+              ) \
+                                            )
+#  endif
+#endif
+#if defined(JSON_HEDLEY_IS_CONSTEXPR_)
+    #if !defined(JSON_HEDLEY_IS_CONSTANT)
+        #define JSON_HEDLEY_IS_CONSTANT(expr) JSON_HEDLEY_IS_CONSTEXPR_(expr)
+    #endif
+    #define JSON_HEDLEY_REQUIRE_CONSTEXPR(expr) (JSON_HEDLEY_IS_CONSTEXPR_(expr) ? (expr) : (-1))
+#else
+    #if !defined(JSON_HEDLEY_IS_CONSTANT)
+        #define JSON_HEDLEY_IS_CONSTANT(expr) (0)
+    #endif
+    #define JSON_HEDLEY_REQUIRE_CONSTEXPR(expr) (expr)
+#endif
+
+#if defined(JSON_HEDLEY_BEGIN_C_DECLS)
+    #undef JSON_HEDLEY_BEGIN_C_DECLS
+#endif
+#if defined(JSON_HEDLEY_END_C_DECLS)
+    #undef JSON_HEDLEY_END_C_DECLS
+#endif
+#if defined(JSON_HEDLEY_C_DECL)
+    #undef JSON_HEDLEY_C_DECL
+#endif
+#if defined(__cplusplus)
+    #define JSON_HEDLEY_BEGIN_C_DECLS extern "C" {
+    #define JSON_HEDLEY_END_C_DECLS }
+    #define JSON_HEDLEY_C_DECL extern "C"
+#else
+    #define JSON_HEDLEY_BEGIN_C_DECLS
+    #define JSON_HEDLEY_END_C_DECLS
+    #define JSON_HEDLEY_C_DECL
+#endif
+
+#if defined(JSON_HEDLEY_STATIC_ASSERT)
+    #undef JSON_HEDLEY_STATIC_ASSERT
+#endif
+#if \
+  !defined(__cplusplus) && ( \
+      (defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 201112L)) || \
+      (JSON_HEDLEY_HAS_FEATURE(c_static_assert) && !defined(JSON_HEDLEY_INTEL_CL_VERSION)) || \
+      JSON_HEDLEY_GCC_VERSION_CHECK(6,0,0) || \
+      JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0) || \
+      defined(_Static_assert) \
+    )
+#  define JSON_HEDLEY_STATIC_ASSERT(expr, message) _Static_assert(expr, message)
+#elif \
+  (defined(__cplusplus) && (__cplusplus >= 201103L)) || \
+  JSON_HEDLEY_MSVC_VERSION_CHECK(16,0,0) || \
+  JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+#  define JSON_HEDLEY_STATIC_ASSERT(expr, message) JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_(static_assert(expr, message))
+#else
+#  define JSON_HEDLEY_STATIC_ASSERT(expr, message)
+#endif
+
+#if defined(JSON_HEDLEY_NULL)
+    #undef JSON_HEDLEY_NULL
+#endif
+#if defined(__cplusplus)
+    #if __cplusplus >= 201103L
+        #define JSON_HEDLEY_NULL JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_(nullptr)
+    #elif defined(NULL)
+        #define JSON_HEDLEY_NULL NULL
+    #else
+        #define JSON_HEDLEY_NULL JSON_HEDLEY_STATIC_CAST(void*, 0)
+    #endif
+#elif defined(NULL)
+    #define JSON_HEDLEY_NULL NULL
+#else
+    #define JSON_HEDLEY_NULL ((void*) 0)
+#endif
+
+#if defined(JSON_HEDLEY_MESSAGE)
+    #undef JSON_HEDLEY_MESSAGE
+#endif
+#if JSON_HEDLEY_HAS_WARNING("-Wunknown-pragmas")
+#  define JSON_HEDLEY_MESSAGE(msg) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS \
+    JSON_HEDLEY_PRAGMA(message msg) \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#elif \
+  JSON_HEDLEY_GCC_VERSION_CHECK(4,4,0) || \
+  JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0)
+#  define JSON_HEDLEY_MESSAGE(msg) JSON_HEDLEY_PRAGMA(message msg)
+#elif JSON_HEDLEY_CRAY_VERSION_CHECK(5,0,0)
+#  define JSON_HEDLEY_MESSAGE(msg) JSON_HEDLEY_PRAGMA(_CRI message msg)
+#elif JSON_HEDLEY_IAR_VERSION_CHECK(8,0,0)
+#  define JSON_HEDLEY_MESSAGE(msg) JSON_HEDLEY_PRAGMA(message(msg))
+#elif JSON_HEDLEY_PELLES_VERSION_CHECK(2,0,0)
+#  define JSON_HEDLEY_MESSAGE(msg) JSON_HEDLEY_PRAGMA(message(msg))
+#else
+#  define JSON_HEDLEY_MESSAGE(msg)
+#endif
+
+#if defined(JSON_HEDLEY_WARNING)
+    #undef JSON_HEDLEY_WARNING
+#endif
+#if JSON_HEDLEY_HAS_WARNING("-Wunknown-pragmas")
+#  define JSON_HEDLEY_WARNING(msg) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS \
+    JSON_HEDLEY_PRAGMA(clang warning msg) \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#elif \
+  JSON_HEDLEY_GCC_VERSION_CHECK(4,8,0) || \
+  JSON_HEDLEY_PGI_VERSION_CHECK(18,4,0) || \
+  JSON_HEDLEY_INTEL_VERSION_CHECK(13,0,0)
+#  define JSON_HEDLEY_WARNING(msg) JSON_HEDLEY_PRAGMA(GCC warning msg)
+#elif \
+  JSON_HEDLEY_MSVC_VERSION_CHECK(15,0,0) || \
+  JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+#  define JSON_HEDLEY_WARNING(msg) JSON_HEDLEY_PRAGMA(message(msg))
+#else
+#  define JSON_HEDLEY_WARNING(msg) JSON_HEDLEY_MESSAGE(msg)
+#endif
+
+#if defined(JSON_HEDLEY_REQUIRE)
+    #undef JSON_HEDLEY_REQUIRE
+#endif
+#if defined(JSON_HEDLEY_REQUIRE_MSG)
+    #undef JSON_HEDLEY_REQUIRE_MSG
+#endif
+#if JSON_HEDLEY_HAS_ATTRIBUTE(diagnose_if)
+#  if JSON_HEDLEY_HAS_WARNING("-Wgcc-compat")
+#    define JSON_HEDLEY_REQUIRE(expr) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    _Pragma("clang diagnostic ignored \"-Wgcc-compat\"") \
+    __attribute__((diagnose_if(!(expr), #expr, "error"))) \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#    define JSON_HEDLEY_REQUIRE_MSG(expr,msg) \
+    JSON_HEDLEY_DIAGNOSTIC_PUSH \
+    _Pragma("clang diagnostic ignored \"-Wgcc-compat\"") \
+    __attribute__((diagnose_if(!(expr), msg, "error"))) \
+    JSON_HEDLEY_DIAGNOSTIC_POP
+#  else
+#    define JSON_HEDLEY_REQUIRE(expr) __attribute__((diagnose_if(!(expr), #expr, "error")))
+#    define JSON_HEDLEY_REQUIRE_MSG(expr,msg) __attribute__((diagnose_if(!(expr), msg, "error")))
+#  endif
+#else
+#  define JSON_HEDLEY_REQUIRE(expr)
+#  define JSON_HEDLEY_REQUIRE_MSG(expr,msg)
+#endif
+
+#if defined(JSON_HEDLEY_FLAGS)
+    #undef JSON_HEDLEY_FLAGS
+#endif
+#if JSON_HEDLEY_HAS_ATTRIBUTE(flag_enum) && (!defined(__cplusplus) || JSON_HEDLEY_HAS_WARNING("-Wbitfield-enum-conversion"))
+    #define JSON_HEDLEY_FLAGS __attribute__((__flag_enum__))
+#else
+    #define JSON_HEDLEY_FLAGS
+#endif
+
+#if defined(JSON_HEDLEY_FLAGS_CAST)
+    #undef JSON_HEDLEY_FLAGS_CAST
+#endif
+#if JSON_HEDLEY_INTEL_VERSION_CHECK(19,0,0)
+#  define JSON_HEDLEY_FLAGS_CAST(T, expr) (__extension__ ({ \
+        JSON_HEDLEY_DIAGNOSTIC_PUSH \
+        _Pragma("warning(disable:188)") \
+        ((T) (expr)); \
+        JSON_HEDLEY_DIAGNOSTIC_POP \
+    }))
+#else
+#  define JSON_HEDLEY_FLAGS_CAST(T, expr) JSON_HEDLEY_STATIC_CAST(T, expr)
+#endif
+
+#if defined(JSON_HEDLEY_EMPTY_BASES)
+    #undef JSON_HEDLEY_EMPTY_BASES
+#endif
+#if \
+    (JSON_HEDLEY_MSVC_VERSION_CHECK(19,0,23918) && !JSON_HEDLEY_MSVC_VERSION_CHECK(20,0,0)) || \
+    JSON_HEDLEY_INTEL_CL_VERSION_CHECK(2021,1,0)
+    #define JSON_HEDLEY_EMPTY_BASES __declspec(empty_bases)
+#else
+    #define JSON_HEDLEY_EMPTY_BASES
+#endif
+
+/* Remaining macros are deprecated. */
+
+#if defined(JSON_HEDLEY_GCC_NOT_CLANG_VERSION_CHECK)
+    #undef JSON_HEDLEY_GCC_NOT_CLANG_VERSION_CHECK
+#endif
+#if defined(__clang__)
+    #define JSON_HEDLEY_GCC_NOT_CLANG_VERSION_CHECK(major,minor,patch) (0)
+#else
+    #define JSON_HEDLEY_GCC_NOT_CLANG_VERSION_CHECK(major,minor,patch) JSON_HEDLEY_GCC_VERSION_CHECK(major,minor,patch)
+#endif
+
+#if defined(JSON_HEDLEY_CLANG_HAS_ATTRIBUTE)
+    #undef JSON_HEDLEY_CLANG_HAS_ATTRIBUTE
+#endif
+#define JSON_HEDLEY_CLANG_HAS_ATTRIBUTE(attribute) JSON_HEDLEY_HAS_ATTRIBUTE(attribute)
+
+#if defined(JSON_HEDLEY_CLANG_HAS_CPP_ATTRIBUTE)
+    #undef JSON_HEDLEY_CLANG_HAS_CPP_ATTRIBUTE
+#endif
+#define JSON_HEDLEY_CLANG_HAS_CPP_ATTRIBUTE(attribute) JSON_HEDLEY_HAS_CPP_ATTRIBUTE(attribute)
+
+#if defined(JSON_HEDLEY_CLANG_HAS_BUILTIN)
+    #undef JSON_HEDLEY_CLANG_HAS_BUILTIN
+#endif
+#define JSON_HEDLEY_CLANG_HAS_BUILTIN(builtin) JSON_HEDLEY_HAS_BUILTIN(builtin)
+
+#if defined(JSON_HEDLEY_CLANG_HAS_FEATURE)
+    #undef JSON_HEDLEY_CLANG_HAS_FEATURE
+#endif
+#define JSON_HEDLEY_CLANG_HAS_FEATURE(feature) JSON_HEDLEY_HAS_FEATURE(feature)
+
+#if defined(JSON_HEDLEY_CLANG_HAS_EXTENSION)
+    #undef JSON_HEDLEY_CLANG_HAS_EXTENSION
+#endif
+#define JSON_HEDLEY_CLANG_HAS_EXTENSION(extension) JSON_HEDLEY_HAS_EXTENSION(extension)
+
+#if defined(JSON_HEDLEY_CLANG_HAS_DECLSPEC_DECLSPEC_ATTRIBUTE)
+    #undef JSON_HEDLEY_CLANG_HAS_DECLSPEC_DECLSPEC_ATTRIBUTE
+#endif
+#define JSON_HEDLEY_CLANG_HAS_DECLSPEC_ATTRIBUTE(attribute) JSON_HEDLEY_HAS_DECLSPEC_ATTRIBUTE(attribute)
+
+#if defined(JSON_HEDLEY_CLANG_HAS_WARNING)
+    #undef JSON_HEDLEY_CLANG_HAS_WARNING
+#endif
+#define JSON_HEDLEY_CLANG_HAS_WARNING(warning) JSON_HEDLEY_HAS_WARNING(warning)
+
+#endif /* !defined(JSON_HEDLEY_VERSION) || (JSON_HEDLEY_VERSION < X) */
+
+
+// This file contains all internal macro definitions (except those affecting ABI)
+// You MUST include macro_unscope.hpp at the end of json.hpp to undef all of them
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+// exclude unsupported compilers
+#if !defined(JSON_SKIP_UNSUPPORTED_COMPILER_CHECK)
+    #if defined(__clang__)
+        #if (__clang_major__ * 10000 + __clang_minor__ * 100 + __clang_patchlevel__) < 30400
+            #error "unsupported Clang version - see https://github.com/nlohmann/json#supported-compilers"
+        #endif
+    #elif defined(__GNUC__) && !(defined(__ICC) || defined(__INTEL_COMPILER))
+        #if (__GNUC__ * 10000 + __GNUC_MINOR__ * 100 + __GNUC_PATCHLEVEL__) < 40800
+            #error "unsupported GCC version - see https://github.com/nlohmann/json#supported-compilers"
+        #endif
+    #endif
+#endif
+
+// C++ language standard detection
+// if the user manually specified the used c++ version this is skipped
+#if !defined(JSON_HAS_CPP_20) && !defined(JSON_HAS_CPP_17) && !defined(JSON_HAS_CPP_14) && !defined(JSON_HAS_CPP_11)
+    #if (defined(__cplusplus) && __cplusplus >= 202002L) || (defined(_MSVC_LANG) && _MSVC_LANG >= 202002L)
+        #define JSON_HAS_CPP_20
+        #define JSON_HAS_CPP_17
+        #define JSON_HAS_CPP_14
+    #elif (defined(__cplusplus) && __cplusplus >= 201703L) || (defined(_HAS_CXX17) && _HAS_CXX17 == 1) // fix for issue #464
+        #define JSON_HAS_CPP_17
+        #define JSON_HAS_CPP_14
+    #elif (defined(__cplusplus) && __cplusplus >= 201402L) || (defined(_HAS_CXX14) && _HAS_CXX14 == 1)
+        #define JSON_HAS_CPP_14
+    #endif
+    // the cpp 11 flag is always specified because it is the minimal required version
+    #define JSON_HAS_CPP_11
+#endif
+
+#ifdef __has_include
+    #if __has_include(<version>)
+        #include <version>
+    #endif
+#endif
+
+#if !defined(JSON_HAS_FILESYSTEM) && !defined(JSON_HAS_EXPERIMENTAL_FILESYSTEM)
+    #ifdef JSON_HAS_CPP_17
+        #if defined(__cpp_lib_filesystem)
+            #define JSON_HAS_FILESYSTEM 1
+        #elif defined(__cpp_lib_experimental_filesystem)
+            #define JSON_HAS_EXPERIMENTAL_FILESYSTEM 1
+        #elif !defined(__has_include)
+            #define JSON_HAS_EXPERIMENTAL_FILESYSTEM 1
+        #elif __has_include(<filesystem>)
+            #define JSON_HAS_FILESYSTEM 1
+        #elif __has_include(<experimental/filesystem>)
+            #define JSON_HAS_EXPERIMENTAL_FILESYSTEM 1
+        #endif
+
+        // std::filesystem does not work on MinGW GCC 8: https://sourceforge.net/p/mingw-w64/bugs/737/
+        #if defined(__MINGW32__) && defined(__GNUC__) && __GNUC__ == 8
+            #undef JSON_HAS_FILESYSTEM
+            #undef JSON_HAS_EXPERIMENTAL_FILESYSTEM
+        #endif
+
+        // no filesystem support before GCC 8: https://en.cppreference.com/w/cpp/compiler_support
+        #if defined(__GNUC__) && !defined(__clang__) && __GNUC__ < 8
+            #undef JSON_HAS_FILESYSTEM
+            #undef JSON_HAS_EXPERIMENTAL_FILESYSTEM
+        #endif
+
+        // no filesystem support before Clang 7: https://en.cppreference.com/w/cpp/compiler_support
+        #if defined(__clang_major__) && __clang_major__ < 7
+            #undef JSON_HAS_FILESYSTEM
+            #undef JSON_HAS_EXPERIMENTAL_FILESYSTEM
+        #endif
+
+        // no filesystem support before MSVC 19.14: https://en.cppreference.com/w/cpp/compiler_support
+        #if defined(_MSC_VER) && _MSC_VER < 1914
+            #undef JSON_HAS_FILESYSTEM
+            #undef JSON_HAS_EXPERIMENTAL_FILESYSTEM
+        #endif
+
+        // no filesystem support before iOS 13
+        #if defined(__IPHONE_OS_VERSION_MIN_REQUIRED) && __IPHONE_OS_VERSION_MIN_REQUIRED < 130000
+            #undef JSON_HAS_FILESYSTEM
+            #undef JSON_HAS_EXPERIMENTAL_FILESYSTEM
+        #endif
+
+        // no filesystem support before macOS Catalina
+        #if defined(__MAC_OS_X_VERSION_MIN_REQUIRED) && __MAC_OS_X_VERSION_MIN_REQUIRED < 101500
+            #undef JSON_HAS_FILESYSTEM
+            #undef JSON_HAS_EXPERIMENTAL_FILESYSTEM
+        #endif
+    #endif
+#endif
+
+#ifndef JSON_HAS_EXPERIMENTAL_FILESYSTEM
+    #define JSON_HAS_EXPERIMENTAL_FILESYSTEM 0
+#endif
+
+#ifndef JSON_HAS_FILESYSTEM
+    #define JSON_HAS_FILESYSTEM 0
+#endif
+
+#ifndef JSON_HAS_THREE_WAY_COMPARISON
+    #if defined(__cpp_impl_three_way_comparison) && __cpp_impl_three_way_comparison >= 201907L \
+        && defined(__cpp_lib_three_way_comparison) && __cpp_lib_three_way_comparison >= 201907L
+        #define JSON_HAS_THREE_WAY_COMPARISON 1
+    #else
+        #define JSON_HAS_THREE_WAY_COMPARISON 0
+    #endif
+#endif
+
+#ifndef JSON_HAS_RANGES
+    // ranges header shipping in GCC 11.1.0 (released 2021-04-27) has syntax error
+    #if defined(__GLIBCXX__) && __GLIBCXX__ == 20210427
+        #define JSON_HAS_RANGES 0
+    #elif defined(__cpp_lib_ranges)
+        #define JSON_HAS_RANGES 1
+    #else
+        #define JSON_HAS_RANGES 0
+    #endif
+#endif
+
+#ifndef JSON_HAS_STATIC_RTTI
+    #if !defined(_HAS_STATIC_RTTI) || _HAS_STATIC_RTTI != 0
+        #define JSON_HAS_STATIC_RTTI 1
+    #else
+        #define JSON_HAS_STATIC_RTTI 0
+    #endif
+#endif
+
+#ifdef JSON_HAS_CPP_17
+    #define JSON_INLINE_VARIABLE inline
+#else
+    #define JSON_INLINE_VARIABLE
+#endif
+
+#if JSON_HEDLEY_HAS_ATTRIBUTE(no_unique_address)
+    #define JSON_NO_UNIQUE_ADDRESS [[no_unique_address]]
+#else
+    #define JSON_NO_UNIQUE_ADDRESS
+#endif
+
+// disable documentation warnings on clang
+#if defined(__clang__)
+    #pragma clang diagnostic push
+    #pragma clang diagnostic ignored "-Wdocumentation"
+    #pragma clang diagnostic ignored "-Wdocumentation-unknown-command"
+#endif
+
+// allow disabling exceptions
+#if (defined(__cpp_exceptions) || defined(__EXCEPTIONS) || defined(_CPPUNWIND)) && !defined(JSON_NOEXCEPTION)
+    #define JSON_THROW(exception) throw exception
+    #define JSON_TRY try
+    #define JSON_CATCH(exception) catch(exception)
+    #define JSON_INTERNAL_CATCH(exception) catch(exception)
+#else
+    #include <cstdlib>
+    #define JSON_THROW(exception) std::abort()
+    #define JSON_TRY if(true)
+    #define JSON_CATCH(exception) if(false)
+    #define JSON_INTERNAL_CATCH(exception) if(false)
+#endif
+
+// override exception macros
+#if defined(JSON_THROW_USER)
+    #undef JSON_THROW
+    #define JSON_THROW JSON_THROW_USER
+#endif
+#if defined(JSON_TRY_USER)
+    #undef JSON_TRY
+    #define JSON_TRY JSON_TRY_USER
+#endif
+#if defined(JSON_CATCH_USER)
+    #undef JSON_CATCH
+    #define JSON_CATCH JSON_CATCH_USER
+    #undef JSON_INTERNAL_CATCH
+    #define JSON_INTERNAL_CATCH JSON_CATCH_USER
+#endif
+#if defined(JSON_INTERNAL_CATCH_USER)
+    #undef JSON_INTERNAL_CATCH
+    #define JSON_INTERNAL_CATCH JSON_INTERNAL_CATCH_USER
+#endif
+
+// allow overriding assert
+#if !defined(JSON_ASSERT)
+    #include <cassert> // assert
+    #define JSON_ASSERT(x) assert(x)
+#endif
+
+// allow to access some private functions (needed by the test suite)
+#if defined(JSON_TESTS_PRIVATE)
+    #define JSON_PRIVATE_UNLESS_TESTED public
+#else
+    #define JSON_PRIVATE_UNLESS_TESTED private
+#endif
+
+/*!
+@brief macro to briefly define a mapping between an enum and JSON
+@def NLOHMANN_JSON_SERIALIZE_ENUM
+@since version 3.4.0
+*/
+#define NLOHMANN_JSON_SERIALIZE_ENUM(ENUM_TYPE, ...)                                            \
+    template<typename BasicJsonType>                                                            \
+    inline void to_json(BasicJsonType& j, const ENUM_TYPE& e)                                   \
+    {                                                                                           \
+        /* NOLINTNEXTLINE(modernize-type-traits) we use C++11 */                                \
+        static_assert(std::is_enum<ENUM_TYPE>::value, #ENUM_TYPE " must be an enum!");          \
+        /* NOLINTNEXTLINE(modernize-avoid-c-arrays) we don't want to depend on <array> */       \
+        static const std::pair<ENUM_TYPE, BasicJsonType> m[] = __VA_ARGS__;                     \
+        auto it = std::find_if(std::begin(m), std::end(m),                                      \
+                               [e](const std::pair<ENUM_TYPE, BasicJsonType>& ej_pair) -> bool  \
+        {                                                                                       \
+            return ej_pair.first == e;                                                          \
+        });                                                                                     \
+        j = ((it != std::end(m)) ? it : std::begin(m))->second;                                 \
+    }                                                                                           \
+    template<typename BasicJsonType>                                                            \
+    inline void from_json(const BasicJsonType& j, ENUM_TYPE& e)                                 \
+    {                                                                                           \
+        /* NOLINTNEXTLINE(modernize-type-traits) we use C++11 */                                \
+        static_assert(std::is_enum<ENUM_TYPE>::value, #ENUM_TYPE " must be an enum!");          \
+        /* NOLINTNEXTLINE(modernize-avoid-c-arrays) we don't want to depend on <array> */       \
+        static const std::pair<ENUM_TYPE, BasicJsonType> m[] = __VA_ARGS__;                     \
+        auto it = std::find_if(std::begin(m), std::end(m),                                      \
+                               [&j](const std::pair<ENUM_TYPE, BasicJsonType>& ej_pair) -> bool \
+        {                                                                                       \
+            return ej_pair.second == j;                                                         \
+        });                                                                                     \
+        e = ((it != std::end(m)) ? it : std::begin(m))->first;                                  \
+    }
+
+// Ugly macros to avoid uglier copy-paste when specializing basic_json. They
+// may be removed in the future once the class is split.
+
+#define NLOHMANN_BASIC_JSON_TPL_DECLARATION                                \
+    template<template<typename, typename, typename...> class ObjectType,   \
+             template<typename, typename...> class ArrayType,              \
+             class StringType, class BooleanType, class NumberIntegerType, \
+             class NumberUnsignedType, class NumberFloatType,              \
+             template<typename> class AllocatorType,                       \
+             template<typename, typename = void> class JSONSerializer,     \
+             class BinaryType,                                             \
+             class CustomBaseClass>
+
+#define NLOHMANN_BASIC_JSON_TPL                                            \
+    basic_json<ObjectType, ArrayType, StringType, BooleanType,             \
+    NumberIntegerType, NumberUnsignedType, NumberFloatType,                \
+    AllocatorType, JSONSerializer, BinaryType, CustomBaseClass>
+
+// Macros to simplify conversion from/to types
+
+#define NLOHMANN_JSON_EXPAND( x ) x
+#define NLOHMANN_JSON_GET_MACRO(_1, _2, _3, _4, _5, _6, _7, _8, _9, _10, _11, _12, _13, _14, _15, _16, _17, _18, _19, _20, _21, _22, _23, _24, _25, _26, _27, _28, _29, _30, _31, _32, _33, _34, _35, _36, _37, _38, _39, _40, _41, _42, _43, _44, _45, _46, _47, _48, _49, _50, _51, _52, _53, _54, _55, _56, _57, _58, _59, _60, _61, _62, _63, _64, NAME,...) NAME
+#define NLOHMANN_JSON_PASTE(...) NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_GET_MACRO(__VA_ARGS__, \
+        NLOHMANN_JSON_PASTE64, \
+        NLOHMANN_JSON_PASTE63, \
+        NLOHMANN_JSON_PASTE62, \
+        NLOHMANN_JSON_PASTE61, \
+        NLOHMANN_JSON_PASTE60, \
+        NLOHMANN_JSON_PASTE59, \
+        NLOHMANN_JSON_PASTE58, \
+        NLOHMANN_JSON_PASTE57, \
+        NLOHMANN_JSON_PASTE56, \
+        NLOHMANN_JSON_PASTE55, \
+        NLOHMANN_JSON_PASTE54, \
+        NLOHMANN_JSON_PASTE53, \
+        NLOHMANN_JSON_PASTE52, \
+        NLOHMANN_JSON_PASTE51, \
+        NLOHMANN_JSON_PASTE50, \
+        NLOHMANN_JSON_PASTE49, \
+        NLOHMANN_JSON_PASTE48, \
+        NLOHMANN_JSON_PASTE47, \
+        NLOHMANN_JSON_PASTE46, \
+        NLOHMANN_JSON_PASTE45, \
+        NLOHMANN_JSON_PASTE44, \
+        NLOHMANN_JSON_PASTE43, \
+        NLOHMANN_JSON_PASTE42, \
+        NLOHMANN_JSON_PASTE41, \
+        NLOHMANN_JSON_PASTE40, \
+        NLOHMANN_JSON_PASTE39, \
+        NLOHMANN_JSON_PASTE38, \
+        NLOHMANN_JSON_PASTE37, \
+        NLOHMANN_JSON_PASTE36, \
+        NLOHMANN_JSON_PASTE35, \
+        NLOHMANN_JSON_PASTE34, \
+        NLOHMANN_JSON_PASTE33, \
+        NLOHMANN_JSON_PASTE32, \
+        NLOHMANN_JSON_PASTE31, \
+        NLOHMANN_JSON_PASTE30, \
+        NLOHMANN_JSON_PASTE29, \
+        NLOHMANN_JSON_PASTE28, \
+        NLOHMANN_JSON_PASTE27, \
+        NLOHMANN_JSON_PASTE26, \
+        NLOHMANN_JSON_PASTE25, \
+        NLOHMANN_JSON_PASTE24, \
+        NLOHMANN_JSON_PASTE23, \
+        NLOHMANN_JSON_PASTE22, \
+        NLOHMANN_JSON_PASTE21, \
+        NLOHMANN_JSON_PASTE20, \
+        NLOHMANN_JSON_PASTE19, \
+        NLOHMANN_JSON_PASTE18, \
+        NLOHMANN_JSON_PASTE17, \
+        NLOHMANN_JSON_PASTE16, \
+        NLOHMANN_JSON_PASTE15, \
+        NLOHMANN_JSON_PASTE14, \
+        NLOHMANN_JSON_PASTE13, \
+        NLOHMANN_JSON_PASTE12, \
+        NLOHMANN_JSON_PASTE11, \
+        NLOHMANN_JSON_PASTE10, \
+        NLOHMANN_JSON_PASTE9, \
+        NLOHMANN_JSON_PASTE8, \
+        NLOHMANN_JSON_PASTE7, \
+        NLOHMANN_JSON_PASTE6, \
+        NLOHMANN_JSON_PASTE5, \
+        NLOHMANN_JSON_PASTE4, \
+        NLOHMANN_JSON_PASTE3, \
+        NLOHMANN_JSON_PASTE2, \
+        NLOHMANN_JSON_PASTE1)(__VA_ARGS__))
+#define NLOHMANN_JSON_PASTE2(func, v1) func(v1)
+#define NLOHMANN_JSON_PASTE3(func, v1, v2) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE2(func, v2)
+#define NLOHMANN_JSON_PASTE4(func, v1, v2, v3) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE3(func, v2, v3)
+#define NLOHMANN_JSON_PASTE5(func, v1, v2, v3, v4) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE4(func, v2, v3, v4)
+#define NLOHMANN_JSON_PASTE6(func, v1, v2, v3, v4, v5) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE5(func, v2, v3, v4, v5)
+#define NLOHMANN_JSON_PASTE7(func, v1, v2, v3, v4, v5, v6) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE6(func, v2, v3, v4, v5, v6)
+#define NLOHMANN_JSON_PASTE8(func, v1, v2, v3, v4, v5, v6, v7) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE7(func, v2, v3, v4, v5, v6, v7)
+#define NLOHMANN_JSON_PASTE9(func, v1, v2, v3, v4, v5, v6, v7, v8) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE8(func, v2, v3, v4, v5, v6, v7, v8)
+#define NLOHMANN_JSON_PASTE10(func, v1, v2, v3, v4, v5, v6, v7, v8, v9) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE9(func, v2, v3, v4, v5, v6, v7, v8, v9)
+#define NLOHMANN_JSON_PASTE11(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE10(func, v2, v3, v4, v5, v6, v7, v8, v9, v10)
+#define NLOHMANN_JSON_PASTE12(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE11(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11)
+#define NLOHMANN_JSON_PASTE13(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE12(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12)
+#define NLOHMANN_JSON_PASTE14(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE13(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13)
+#define NLOHMANN_JSON_PASTE15(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE14(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14)
+#define NLOHMANN_JSON_PASTE16(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE15(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15)
+#define NLOHMANN_JSON_PASTE17(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE16(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16)
+#define NLOHMANN_JSON_PASTE18(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE17(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17)
+#define NLOHMANN_JSON_PASTE19(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE18(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18)
+#define NLOHMANN_JSON_PASTE20(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE19(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19)
+#define NLOHMANN_JSON_PASTE21(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE20(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20)
+#define NLOHMANN_JSON_PASTE22(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE21(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21)
+#define NLOHMANN_JSON_PASTE23(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE22(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22)
+#define NLOHMANN_JSON_PASTE24(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE23(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23)
+#define NLOHMANN_JSON_PASTE25(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE24(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24)
+#define NLOHMANN_JSON_PASTE26(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE25(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25)
+#define NLOHMANN_JSON_PASTE27(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE26(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26)
+#define NLOHMANN_JSON_PASTE28(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE27(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27)
+#define NLOHMANN_JSON_PASTE29(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE28(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28)
+#define NLOHMANN_JSON_PASTE30(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE29(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29)
+#define NLOHMANN_JSON_PASTE31(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE30(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30)
+#define NLOHMANN_JSON_PASTE32(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE31(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31)
+#define NLOHMANN_JSON_PASTE33(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE32(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32)
+#define NLOHMANN_JSON_PASTE34(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE33(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33)
+#define NLOHMANN_JSON_PASTE35(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE34(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34)
+#define NLOHMANN_JSON_PASTE36(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE35(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35)
+#define NLOHMANN_JSON_PASTE37(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE36(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36)
+#define NLOHMANN_JSON_PASTE38(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE37(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37)
+#define NLOHMANN_JSON_PASTE39(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE38(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38)
+#define NLOHMANN_JSON_PASTE40(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE39(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39)
+#define NLOHMANN_JSON_PASTE41(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE40(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40)
+#define NLOHMANN_JSON_PASTE42(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE41(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41)
+#define NLOHMANN_JSON_PASTE43(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE42(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42)
+#define NLOHMANN_JSON_PASTE44(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE43(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43)
+#define NLOHMANN_JSON_PASTE45(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE44(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44)
+#define NLOHMANN_JSON_PASTE46(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE45(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45)
+#define NLOHMANN_JSON_PASTE47(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE46(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46)
+#define NLOHMANN_JSON_PASTE48(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE47(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47)
+#define NLOHMANN_JSON_PASTE49(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE48(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48)
+#define NLOHMANN_JSON_PASTE50(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE49(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49)
+#define NLOHMANN_JSON_PASTE51(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE50(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50)
+#define NLOHMANN_JSON_PASTE52(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE51(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51)
+#define NLOHMANN_JSON_PASTE53(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE52(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52)
+#define NLOHMANN_JSON_PASTE54(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE53(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53)
+#define NLOHMANN_JSON_PASTE55(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE54(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54)
+#define NLOHMANN_JSON_PASTE56(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE55(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55)
+#define NLOHMANN_JSON_PASTE57(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE56(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56)
+#define NLOHMANN_JSON_PASTE58(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE57(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57)
+#define NLOHMANN_JSON_PASTE59(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE58(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58)
+#define NLOHMANN_JSON_PASTE60(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE59(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59)
+#define NLOHMANN_JSON_PASTE61(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59, v60) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE60(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59, v60)
+#define NLOHMANN_JSON_PASTE62(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59, v60, v61) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE61(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59, v60, v61)
+#define NLOHMANN_JSON_PASTE63(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59, v60, v61, v62) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE62(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59, v60, v61, v62)
+#define NLOHMANN_JSON_PASTE64(func, v1, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59, v60, v61, v62, v63) NLOHMANN_JSON_PASTE2(func, v1) NLOHMANN_JSON_PASTE63(func, v2, v3, v4, v5, v6, v7, v8, v9, v10, v11, v12, v13, v14, v15, v16, v17, v18, v19, v20, v21, v22, v23, v24, v25, v26, v27, v28, v29, v30, v31, v32, v33, v34, v35, v36, v37, v38, v39, v40, v41, v42, v43, v44, v45, v46, v47, v48, v49, v50, v51, v52, v53, v54, v55, v56, v57, v58, v59, v60, v61, v62, v63)
+
+#define NLOHMANN_JSON_TO(v1) nlohmann_json_j[#v1] = nlohmann_json_t.v1;
+#define NLOHMANN_JSON_FROM(v1) nlohmann_json_j.at(#v1).get_to(nlohmann_json_t.v1);
+#define NLOHMANN_JSON_FROM_WITH_DEFAULT(v1) nlohmann_json_t.v1 = !nlohmann_json_j.is_null() ? nlohmann_json_j.value(#v1, nlohmann_json_default_obj.v1) : nlohmann_json_default_obj.v1;
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_TYPE_INTRUSIVE
+@since version 3.9.0
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_type_intrusive/
+*/
+#define NLOHMANN_DEFINE_TYPE_INTRUSIVE(Type, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) } \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void from_json(const BasicJsonType& nlohmann_json_j, Type& nlohmann_json_t) { NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_FROM, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_TYPE_INTRUSIVE_WITH_DEFAULT
+@since version 3.11.0
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_type_intrusive/
+*/
+#define NLOHMANN_DEFINE_TYPE_INTRUSIVE_WITH_DEFAULT(Type, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) } \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void from_json(const BasicJsonType& nlohmann_json_j, Type& nlohmann_json_t) { const Type nlohmann_json_default_obj{}; NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_FROM_WITH_DEFAULT, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_TYPE_INTRUSIVE_ONLY_SERIALIZE
+@since version 3.11.3
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_type_intrusive/
+*/
+#define NLOHMANN_DEFINE_TYPE_INTRUSIVE_ONLY_SERIALIZE(Type, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_TYPE_NON_INTRUSIVE
+@since version 3.9.0
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_type_non_intrusive/
+*/
+#define NLOHMANN_DEFINE_TYPE_NON_INTRUSIVE(Type, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) } \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void from_json(const BasicJsonType& nlohmann_json_j, Type& nlohmann_json_t) { NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_FROM, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_TYPE_NON_INTRUSIVE_WITH_DEFAULT
+@since version 3.11.0
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_type_non_intrusive/
+*/
+#define NLOHMANN_DEFINE_TYPE_NON_INTRUSIVE_WITH_DEFAULT(Type, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) } \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void from_json(const BasicJsonType& nlohmann_json_j, Type& nlohmann_json_t) { const Type nlohmann_json_default_obj{}; NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_FROM_WITH_DEFAULT, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_TYPE_NON_INTRUSIVE_ONLY_SERIALIZE
+@since version 3.11.3
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_type_non_intrusive/
+*/
+#define NLOHMANN_DEFINE_TYPE_NON_INTRUSIVE_ONLY_SERIALIZE(Type, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_DERIVED_TYPE_INTRUSIVE
+@since version 3.11.x
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_derived_type/
+*/
+#define NLOHMANN_DEFINE_DERIVED_TYPE_INTRUSIVE(Type, BaseType, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { nlohmann::to_json(nlohmann_json_j, static_cast<const BaseType &>(nlohmann_json_t)); NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) } \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void from_json(const BasicJsonType& nlohmann_json_j, Type& nlohmann_json_t) { nlohmann::from_json(nlohmann_json_j, static_cast<BaseType&>(nlohmann_json_t)); NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_FROM, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_DERIVED_TYPE_INTRUSIVE_WITH_DEFAULT
+@since version 3.11.x
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_derived_type/
+*/
+#define NLOHMANN_DEFINE_DERIVED_TYPE_INTRUSIVE_WITH_DEFAULT(Type, BaseType, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { nlohmann::to_json(nlohmann_json_j, static_cast<const BaseType&>(nlohmann_json_t)); NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) } \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void from_json(const BasicJsonType& nlohmann_json_j, Type& nlohmann_json_t) { nlohmann::from_json(nlohmann_json_j, static_cast<BaseType&>(nlohmann_json_t)); const Type nlohmann_json_default_obj{}; NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_FROM_WITH_DEFAULT, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_DERIVED_TYPE_INTRUSIVE_ONLY_SERIALIZE
+@since version 3.11.x
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_derived_type/
+*/
+#define NLOHMANN_DEFINE_DERIVED_TYPE_INTRUSIVE_ONLY_SERIALIZE(Type, BaseType, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    friend void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { nlohmann::to_json(nlohmann_json_j, static_cast<const BaseType &>(nlohmann_json_t)); NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_DERIVED_TYPE_NON_INTRUSIVE
+@since version 3.11.x
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_derived_type/
+*/
+#define NLOHMANN_DEFINE_DERIVED_TYPE_NON_INTRUSIVE(Type, BaseType, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { nlohmann::to_json(nlohmann_json_j, static_cast<const BaseType &>(nlohmann_json_t)); NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) } \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void from_json(const BasicJsonType& nlohmann_json_j, Type& nlohmann_json_t) { nlohmann::from_json(nlohmann_json_j, static_cast<BaseType&>(nlohmann_json_t)); NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_FROM, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_DERIVED_TYPE_NON_INTRUSIVE_WITH_DEFAULT
+@since version 3.11.x
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_derived_type/
+*/
+#define NLOHMANN_DEFINE_DERIVED_TYPE_NON_INTRUSIVE_WITH_DEFAULT(Type, BaseType, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { nlohmann::to_json(nlohmann_json_j, static_cast<const BaseType &>(nlohmann_json_t)); NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) } \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void from_json(const BasicJsonType& nlohmann_json_j, Type& nlohmann_json_t) { nlohmann::from_json(nlohmann_json_j, static_cast<BaseType&>(nlohmann_json_t)); const Type nlohmann_json_default_obj{}; NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_FROM_WITH_DEFAULT, __VA_ARGS__)) }
+
+/*!
+@brief macro
+@def NLOHMANN_DEFINE_DERIVED_TYPE_NON_INTRUSIVE_ONLY_SERIALIZE
+@since version 3.11.x
+@sa https://json.nlohmann.me/api/macros/nlohmann_define_derived_type/
+*/
+#define NLOHMANN_DEFINE_DERIVED_TYPE_NON_INTRUSIVE_ONLY_SERIALIZE(Type, BaseType, ...)  \
+    template<typename BasicJsonType, nlohmann::detail::enable_if_t<nlohmann::detail::is_basic_json<BasicJsonType>::value, int> = 0> \
+    void to_json(BasicJsonType& nlohmann_json_j, const Type& nlohmann_json_t) { nlohmann::to_json(nlohmann_json_j, static_cast<const BaseType &>(nlohmann_json_t)); NLOHMANN_JSON_EXPAND(NLOHMANN_JSON_PASTE(NLOHMANN_JSON_TO, __VA_ARGS__)) }
+
+// inspired from https://stackoverflow.com/a/26745591
+// allows to call any std function as if (e.g. with begin):
+// using std::begin; begin(x);
+//
+// it allows using the detected idiom to retrieve the return type
+// of such an expression
+#define NLOHMANN_CAN_CALL_STD_FUNC_IMPL(std_name)                                 \
+    namespace detail {                                                            \
+    using std::std_name;                                                          \
+    \
+    template<typename... T>                                                       \
+    using result_of_##std_name = decltype(std_name(std::declval<T>()...));        \
+    }                                                                             \
+    \
+    namespace detail2 {                                                           \
+    struct std_name##_tag                                                         \
+    {                                                                             \
+    };                                                                            \
+    \
+    template<typename... T>                                                       \
+    std_name##_tag std_name(T&&...);                                              \
+    \
+    template<typename... T>                                                       \
+    using result_of_##std_name = decltype(std_name(std::declval<T>()...));        \
+    \
+    template<typename... T>                                                       \
+    struct would_call_std_##std_name                                              \
+    {                                                                             \
+        static constexpr auto const value = ::nlohmann::detail::                  \
+                                            is_detected_exact<std_name##_tag, result_of_##std_name, T...>::value; \
+    };                                                                            \
+    } /* namespace detail2 */ \
+    \
+    template<typename... T>                                                       \
+    struct would_call_std_##std_name : detail2::would_call_std_##std_name<T...>   \
+    {                                                                             \
+    }
+
+#ifndef JSON_USE_IMPLICIT_CONVERSIONS
+    #define JSON_USE_IMPLICIT_CONVERSIONS 1
+#endif
+
+#if JSON_USE_IMPLICIT_CONVERSIONS
+    #define JSON_EXPLICIT
+#else
+    #define JSON_EXPLICIT explicit
+#endif
+
+#ifndef JSON_DISABLE_ENUM_SERIALIZATION
+    #define JSON_DISABLE_ENUM_SERIALIZATION 0
+#endif
+
+#ifndef JSON_USE_GLOBAL_UDLS
+    #define JSON_USE_GLOBAL_UDLS 1
+#endif
+
+#if JSON_HAS_THREE_WAY_COMPARISON
+    #include <compare> // partial_ordering
+#endif
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+///////////////////////////
+// JSON type enumeration //
+///////////////////////////
+
+/*!
+@brief the JSON type enumeration
+
+This enumeration collects the different JSON types. It is internally used to
+distinguish the stored values, and the functions @ref basic_json::is_null(),
+@ref basic_json::is_object(), @ref basic_json::is_array(),
+@ref basic_json::is_string(), @ref basic_json::is_boolean(),
+@ref basic_json::is_number() (with @ref basic_json::is_number_integer(),
+@ref basic_json::is_number_unsigned(), and @ref basic_json::is_number_float()),
+@ref basic_json::is_discarded(), @ref basic_json::is_primitive(), and
+@ref basic_json::is_structured() rely on it.
+
+@note There are three enumeration entries (number_integer, number_unsigned, and
+number_float), because the library distinguishes these three types for numbers:
+@ref basic_json::number_unsigned_t is used for unsigned integers,
+@ref basic_json::number_integer_t is used for signed integers, and
+@ref basic_json::number_float_t is used for floating-point numbers or to
+approximate integers which do not fit in the limits of their respective type.
+
+@sa see @ref basic_json::basic_json(const value_t value_type) -- create a JSON
+value with the default value for a given type
+
+@since version 1.0.0
+*/
+enum class value_t : std::uint8_t
+{
+    null,             ///< null value
+    object,           ///< object (unordered set of name/value pairs)
+    array,            ///< array (ordered collection of values)
+    string,           ///< string value
+    boolean,          ///< boolean value
+    number_integer,   ///< number value (signed integer)
+    number_unsigned,  ///< number value (unsigned integer)
+    number_float,     ///< number value (floating-point)
+    binary,           ///< binary array (ordered collection of bytes)
+    discarded         ///< discarded by the parser callback function
+};
+
+/*!
+@brief comparison operator for JSON types
+
+Returns an ordering that is similar to Python:
+- order: null < boolean < number < object < array < string < binary
+- furthermore, each type is not smaller than itself
+- discarded values are not comparable
+- binary is represented as a b"" string in python and directly comparable to a
+  string; however, making a binary array directly comparable with a string would
+  be surprising behavior in a JSON file.
+
+@since version 1.0.0
+*/
+#if JSON_HAS_THREE_WAY_COMPARISON
+    inline std::partial_ordering operator<=>(const value_t lhs, const value_t rhs) noexcept // *NOPAD*
+#else
+    inline bool operator<(const value_t lhs, const value_t rhs) noexcept
+#endif
+{
+    static constexpr std::array<std::uint8_t, 9> order = {{
+            0 /* null */, 3 /* object */, 4 /* array */, 5 /* string */,
+            1 /* boolean */, 2 /* integer */, 2 /* unsigned */, 2 /* float */,
+            6 /* binary */
+        }
+    };
+
+    const auto l_index = static_cast<std::size_t>(lhs);
+    const auto r_index = static_cast<std::size_t>(rhs);
+#if JSON_HAS_THREE_WAY_COMPARISON
+    if (l_index < order.size() && r_index < order.size())
+    {
+        return order[l_index] <=> order[r_index]; // *NOPAD*
+    }
+    return std::partial_ordering::unordered;
+#else
+    return l_index < order.size() && r_index < order.size() && order[l_index] < order[r_index];
+#endif
+}
+
+// GCC selects the built-in operator< over an operator rewritten from
+// a user-defined spaceship operator
+// Clang, MSVC, and ICC select the rewritten candidate
+// (see GCC bug https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105200)
+#if JSON_HAS_THREE_WAY_COMPARISON && defined(__GNUC__)
+inline bool operator<(const value_t lhs, const value_t rhs) noexcept
+{
+    return std::is_lt(lhs <=> rhs); // *NOPAD*
+}
+#endif
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/string_escape.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/*!
+@brief replace all occurrences of a substring by another string
+
+@param[in,out] s  the string to manipulate; changed so that all
+               occurrences of @a f are replaced with @a t
+@param[in]     f  the substring to replace with @a t
+@param[in]     t  the string to replace @a f
+
+@pre The search string @a f must not be empty. **This precondition is
+enforced with an assertion.**
+
+@since version 2.0.0
+*/
+template<typename StringType>
+inline void replace_substring(StringType& s, const StringType& f,
+                              const StringType& t)
+{
+    JSON_ASSERT(!f.empty());
+    for (auto pos = s.find(f);                // find first occurrence of f
+            pos != StringType::npos;          // make sure f was found
+            s.replace(pos, f.size(), t),      // replace with t, and
+            pos = s.find(f, pos + t.size()))  // find next occurrence of f
+    {}
+}
+
+/*!
+ * @brief string escaping as described in RFC 6901 (Sect. 4)
+ * @param[in] s string to escape
+ * @return    escaped string
+ *
+ * Note the order of escaping "~" to "~0" and "/" to "~1" is important.
+ */
+template<typename StringType>
+inline StringType escape(StringType s)
+{
+    replace_substring(s, StringType{"~"}, StringType{"~0"});
+    replace_substring(s, StringType{"/"}, StringType{"~1"});
+    return s;
+}
+
+/*!
+ * @brief string unescaping as described in RFC 6901 (Sect. 4)
+ * @param[in] s string to unescape
+ * @return    unescaped string
+ *
+ * Note the order of escaping "~1" to "/" and "~0" to "~" is important.
+ */
+template<typename StringType>
+static void unescape(StringType& s)
+{
+    replace_substring(s, StringType{"~1"}, StringType{"/"});
+    replace_substring(s, StringType{"~0"}, StringType{"~"});
+}
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/input/position_t.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstddef> // size_t
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/// struct to capture the start position of the current token
+struct position_t
+{
+    /// the total number of characters read
+    std::size_t chars_read_total = 0;
+    /// the number of characters read in the current line
+    std::size_t chars_read_current_line = 0;
+    /// the number of lines read
+    std::size_t lines_read = 0;
+
+    /// conversion to size_t to preserve SAX interface
+    constexpr operator size_t() const
+    {
+        return chars_read_total;
+    }
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-FileCopyrightText: 2018 The Abseil Authors
+// SPDX-License-Identifier: MIT
+
+
+
+#include <array> // array
+#include <cstddef> // size_t
+#include <type_traits> // conditional, enable_if, false_type, integral_constant, is_constructible, is_integral, is_same, remove_cv, remove_reference, true_type
+#include <utility> // index_sequence, make_index_sequence, index_sequence_for
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+template<typename T>
+using uncvref_t = typename std::remove_cv<typename std::remove_reference<T>::type>::type;
+
+#ifdef JSON_HAS_CPP_14
+
+// the following utilities are natively available in C++14
+using std::enable_if_t;
+using std::index_sequence;
+using std::make_index_sequence;
+using std::index_sequence_for;
+
+#else
+
+// alias templates to reduce boilerplate
+template<bool B, typename T = void>
+using enable_if_t = typename std::enable_if<B, T>::type;
+
+// The following code is taken from https://github.com/abseil/abseil-cpp/blob/10cb35e459f5ecca5b2ff107635da0bfa41011b4/absl/utility/utility.h
+// which is part of Google Abseil (https://github.com/abseil/abseil-cpp), licensed under the Apache License 2.0.
+
+//// START OF CODE FROM GOOGLE ABSEIL
+
+// integer_sequence
+//
+// Class template representing a compile-time integer sequence. An instantiation
+// of `integer_sequence<T, Ints...>` has a sequence of integers encoded in its
+// type through its template arguments (which is a common need when
+// working with C++11 variadic templates). `absl::integer_sequence` is designed
+// to be a drop-in replacement for C++14's `std::integer_sequence`.
+//
+// Example:
+//
+//   template< class T, T... Ints >
+//   void user_function(integer_sequence<T, Ints...>);
+//
+//   int main()
+//   {
+//     // user_function's `T` will be deduced to `int` and `Ints...`
+//     // will be deduced to `0, 1, 2, 3, 4`.
+//     user_function(make_integer_sequence<int, 5>());
+//   }
+template <typename T, T... Ints>
+struct integer_sequence
+{
+    using value_type = T;
+    static constexpr std::size_t size() noexcept
+    {
+        return sizeof...(Ints);
+    }
+};
+
+// index_sequence
+//
+// A helper template for an `integer_sequence` of `size_t`,
+// `absl::index_sequence` is designed to be a drop-in replacement for C++14's
+// `std::index_sequence`.
+template <size_t... Ints>
+using index_sequence = integer_sequence<size_t, Ints...>;
+
+namespace utility_internal
+{
+
+template <typename Seq, size_t SeqSize, size_t Rem>
+struct Extend;
+
+// Note that SeqSize == sizeof...(Ints). It's passed explicitly for efficiency.
+template <typename T, T... Ints, size_t SeqSize>
+struct Extend<integer_sequence<T, Ints...>, SeqSize, 0>
+{
+    using type = integer_sequence < T, Ints..., (Ints + SeqSize)... >;
+};
+
+template <typename T, T... Ints, size_t SeqSize>
+struct Extend<integer_sequence<T, Ints...>, SeqSize, 1>
+{
+    using type = integer_sequence < T, Ints..., (Ints + SeqSize)..., 2 * SeqSize >;
+};
+
+// Recursion helper for 'make_integer_sequence<T, N>'.
+// 'Gen<T, N>::type' is an alias for 'integer_sequence<T, 0, 1, ... N-1>'.
+template <typename T, size_t N>
+struct Gen
+{
+    using type =
+        typename Extend < typename Gen < T, N / 2 >::type, N / 2, N % 2 >::type;
+};
+
+template <typename T>
+struct Gen<T, 0>
+{
+    using type = integer_sequence<T>;
+};
+
+}  // namespace utility_internal
+
+// Compile-time sequences of integers
+
+// make_integer_sequence
+//
+// This template alias is equivalent to
+// `integer_sequence<int, 0, 1, ..., N-1>`, and is designed to be a drop-in
+// replacement for C++14's `std::make_integer_sequence`.
+template <typename T, T N>
+using make_integer_sequence = typename utility_internal::Gen<T, N>::type;
+
+// make_index_sequence
+//
+// This template alias is equivalent to `index_sequence<0, 1, ..., N-1>`,
+// and is designed to be a drop-in replacement for C++14's
+// `std::make_index_sequence`.
+template <size_t N>
+using make_index_sequence = make_integer_sequence<size_t, N>;
+
+// index_sequence_for
+//
+// Converts a typename pack into an index sequence of the same length, and
+// is designed to be a drop-in replacement for C++14's
+// `std::index_sequence_for()`
+template <typename... Ts>
+using index_sequence_for = make_index_sequence<sizeof...(Ts)>;
+
+//// END OF CODE FROM GOOGLE ABSEIL
+
+#endif
+
+// dispatch utility (taken from ranges-v3)
+template<unsigned N> struct priority_tag : priority_tag < N - 1 > {};
+template<> struct priority_tag<0> {};
+
+// taken from ranges-v3
+template<typename T>
+struct static_const
+{
+    static JSON_INLINE_VARIABLE constexpr T value{};
+};
+
+#ifndef JSON_HAS_CPP_17
+    template<typename T>
+    constexpr T static_const<T>::value;
+#endif
+
+template<typename T, typename... Args>
+constexpr std::array<T, sizeof...(Args)> make_array(Args&& ... args)
+{
+    return std::array<T, sizeof...(Args)> {{static_cast<T>(std::forward<Args>(args))...}};
+}
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <limits> // numeric_limits
+#include <string> // char_traits
+#include <tuple> // tuple
+#include <type_traits> // false_type, is_constructible, is_integral, is_same, true_type
+#include <utility> // declval
+
+// #include <nlohmann/detail/iterators/iterator_traits.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <iterator> // random_access_iterator_tag
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+// #include <nlohmann/detail/meta/void_t.hpp>
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+template<typename It, typename = void>
+struct iterator_types {};
+
+template<typename It>
+struct iterator_types <
+    It,
+    void_t<typename It::difference_type, typename It::value_type, typename It::pointer,
+    typename It::reference, typename It::iterator_category >>
+{
+    using difference_type = typename It::difference_type;
+    using value_type = typename It::value_type;
+    using pointer = typename It::pointer;
+    using reference = typename It::reference;
+    using iterator_category = typename It::iterator_category;
+};
+
+// This is required as some compilers implement std::iterator_traits in a way that
+// doesn't work with SFINAE. See https://github.com/nlohmann/json/issues/1341.
+template<typename T, typename = void>
+struct iterator_traits
+{
+};
+
+template<typename T>
+struct iterator_traits < T, enable_if_t < !std::is_pointer<T>::value >>
+    : iterator_types<T>
+{
+};
+
+template<typename T>
+struct iterator_traits<T*, enable_if_t<std::is_object<T>::value>>
+{
+    using iterator_category = std::random_access_iterator_tag;
+    using value_type = T;
+    using difference_type = ptrdiff_t;
+    using pointer = T*;
+    using reference = T&;
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/call_std/begin.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+NLOHMANN_CAN_CALL_STD_FUNC_IMPL(begin);
+
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/meta/call_std/end.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+NLOHMANN_CAN_CALL_STD_FUNC_IMPL(end);
+
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
+// #include <nlohmann/detail/meta/detected.hpp>
+
+// #include <nlohmann/json_fwd.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+#ifndef INCLUDE_NLOHMANN_JSON_FWD_HPP_
+    #define INCLUDE_NLOHMANN_JSON_FWD_HPP_
+
+    #include <cstdint> // int64_t, uint64_t
+    #include <map> // map
+    #include <memory> // allocator
+    #include <string> // string
+    #include <vector> // vector
+
+    // #include <nlohmann/detail/abi_macros.hpp>
+
+
+    /*!
+    @brief namespace for Niels Lohmann
+    @see https://github.com/nlohmann
+    @since version 1.0.0
+    */
+    NLOHMANN_JSON_NAMESPACE_BEGIN
+
+    /*!
+    @brief default JSONSerializer template argument
+
+    This serializer ignores the template arguments and uses ADL
+    ([argument-dependent lookup](https://en.cppreference.com/w/cpp/language/adl))
+    for serialization.
+    */
+    template<typename T = void, typename SFINAE = void>
+    struct adl_serializer;
+
+    /// a class to store JSON values
+    /// @sa https://json.nlohmann.me/api/basic_json/
+    template<template<typename U, typename V, typename... Args> class ObjectType =
+    std::map,
+    template<typename U, typename... Args> class ArrayType = std::vector,
+    class StringType = std::string, class BooleanType = bool,
+    class NumberIntegerType = std::int64_t,
+    class NumberUnsignedType = std::uint64_t,
+    class NumberFloatType = double,
+    template<typename U> class AllocatorType = std::allocator,
+    template<typename T, typename SFINAE = void> class JSONSerializer =
+    adl_serializer,
+    class BinaryType = std::vector<std::uint8_t>, // cppcheck-suppress syntaxError
+    class CustomBaseClass = void>
+    class basic_json;
+
+    /// @brief JSON Pointer defines a string syntax for identifying a specific value within a JSON document
+    /// @sa https://json.nlohmann.me/api/json_pointer/
+    template<typename RefStringType>
+    class json_pointer;
+
+    /*!
+    @brief default specialization
+    @sa https://json.nlohmann.me/api/json/
+    */
+    using json = basic_json<>;
+
+    /// @brief a minimal map-like container that preserves insertion order
+    /// @sa https://json.nlohmann.me/api/ordered_map/
+    template<class Key, class T, class IgnoredLess, class Allocator>
+    struct ordered_map;
+
+    /// @brief specialization that maintains the insertion order of object keys
+    /// @sa https://json.nlohmann.me/api/ordered_json/
+    using ordered_json = basic_json<nlohmann::ordered_map>;
+
+    NLOHMANN_JSON_NAMESPACE_END
+
+#endif  // INCLUDE_NLOHMANN_JSON_FWD_HPP_
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+/*!
+@brief detail namespace with internal helper functions
+
+This namespace collects functions that should not be exposed,
+implementations of some @ref basic_json methods, and meta-programming helpers.
+
+@since version 2.1.0
+*/
+namespace detail
+{
+
+/////////////
+// helpers //
+/////////////
+
+// Note to maintainers:
+//
+// Every trait in this file expects a non CV-qualified type.
+// The only exceptions are in the 'aliases for detected' section
+// (i.e. those of the form: decltype(T::member_function(std::declval<T>())))
+//
+// In this case, T has to be properly CV-qualified to constraint the function arguments
+// (e.g. to_json(BasicJsonType&, const T&))
+
+template<typename> struct is_basic_json : std::false_type {};
+
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+struct is_basic_json<NLOHMANN_BASIC_JSON_TPL> : std::true_type {};
+
+// used by exceptions create() member functions
+// true_type for pointer to possibly cv-qualified basic_json or std::nullptr_t
+// false_type otherwise
+template<typename BasicJsonContext>
+struct is_basic_json_context :
+    std::integral_constant < bool,
+    is_basic_json<typename std::remove_cv<typename std::remove_pointer<BasicJsonContext>::type>::type>::value
+    || std::is_same<BasicJsonContext, std::nullptr_t>::value >
+{};
+
+//////////////////////
+// json_ref helpers //
+//////////////////////
+
+template<typename>
+class json_ref;
+
+template<typename>
+struct is_json_ref : std::false_type {};
+
+template<typename T>
+struct is_json_ref<json_ref<T>> : std::true_type {};
+
+//////////////////////////
+// aliases for detected //
+//////////////////////////
+
+template<typename T>
+using mapped_type_t = typename T::mapped_type;
+
+template<typename T>
+using key_type_t = typename T::key_type;
+
+template<typename T>
+using value_type_t = typename T::value_type;
+
+template<typename T>
+using difference_type_t = typename T::difference_type;
+
+template<typename T>
+using pointer_t = typename T::pointer;
+
+template<typename T>
+using reference_t = typename T::reference;
+
+template<typename T>
+using iterator_category_t = typename T::iterator_category;
+
+template<typename T, typename... Args>
+using to_json_function = decltype(T::to_json(std::declval<Args>()...));
+
+template<typename T, typename... Args>
+using from_json_function = decltype(T::from_json(std::declval<Args>()...));
+
+template<typename T, typename U>
+using get_template_function = decltype(std::declval<T>().template get<U>());
+
+// trait checking if JSONSerializer<T>::from_json(json const&, udt&) exists
+template<typename BasicJsonType, typename T, typename = void>
+struct has_from_json : std::false_type {};
+
+// trait checking if j.get<T> is valid
+// use this trait instead of std::is_constructible or std::is_convertible,
+// both rely on, or make use of implicit conversions, and thus fail when T
+// has several constructors/operator= (see https://github.com/nlohmann/json/issues/958)
+template <typename BasicJsonType, typename T>
+struct is_getable
+{
+    static constexpr bool value = is_detected<get_template_function, const BasicJsonType&, T>::value;
+};
+
+template<typename BasicJsonType, typename T>
+struct has_from_json < BasicJsonType, T, enable_if_t < !is_basic_json<T>::value >>
+{
+    using serializer = typename BasicJsonType::template json_serializer<T, void>;
+
+    static constexpr bool value =
+        is_detected_exact<void, from_json_function, serializer,
+        const BasicJsonType&, T&>::value;
+};
+
+// This trait checks if JSONSerializer<T>::from_json(json const&) exists
+// this overload is used for non-default-constructible user-defined-types
+template<typename BasicJsonType, typename T, typename = void>
+struct has_non_default_from_json : std::false_type {};
+
+template<typename BasicJsonType, typename T>
+struct has_non_default_from_json < BasicJsonType, T, enable_if_t < !is_basic_json<T>::value >>
+{
+    using serializer = typename BasicJsonType::template json_serializer<T, void>;
+
+    static constexpr bool value =
+        is_detected_exact<T, from_json_function, serializer,
+        const BasicJsonType&>::value;
+};
+
+// This trait checks if BasicJsonType::json_serializer<T>::to_json exists
+// Do not evaluate the trait when T is a basic_json type, to avoid template instantiation infinite recursion.
+template<typename BasicJsonType, typename T, typename = void>
+struct has_to_json : std::false_type {};
+
+template<typename BasicJsonType, typename T>
+struct has_to_json < BasicJsonType, T, enable_if_t < !is_basic_json<T>::value >>
+{
+    using serializer = typename BasicJsonType::template json_serializer<T, void>;
+
+    static constexpr bool value =
+        is_detected_exact<void, to_json_function, serializer, BasicJsonType&,
+        T>::value;
+};
+
+template<typename T>
+using detect_key_compare = typename T::key_compare;
+
+template<typename T>
+struct has_key_compare : std::integral_constant<bool, is_detected<detect_key_compare, T>::value> {};
+
+// obtains the actual object key comparator
+template<typename BasicJsonType>
+struct actual_object_comparator
+{
+    using object_t = typename BasicJsonType::object_t;
+    using object_comparator_t = typename BasicJsonType::default_object_comparator_t;
+    using type = typename std::conditional < has_key_compare<object_t>::value,
+          typename object_t::key_compare, object_comparator_t>::type;
+};
+
+template<typename BasicJsonType>
+using actual_object_comparator_t = typename actual_object_comparator<BasicJsonType>::type;
+
+/////////////////
+// char_traits //
+/////////////////
+
+// Primary template of char_traits calls std char_traits
+template<typename T>
+struct char_traits : std::char_traits<T>
+{};
+
+// Explicitly define char traits for unsigned char since it is not standard
+template<>
+struct char_traits<unsigned char> : std::char_traits<char>
+{
+    using char_type = unsigned char;
+    using int_type = uint64_t;
+
+    // Redefine to_int_type function
+    static int_type to_int_type(char_type c) noexcept
+    {
+        return static_cast<int_type>(c);
+    }
+
+    static char_type to_char_type(int_type i) noexcept
+    {
+        return static_cast<char_type>(i);
+    }
+
+    static constexpr int_type eof() noexcept
+    {
+        return static_cast<int_type>(std::char_traits<char>::eof());
+    }
+};
+
+// Explicitly define char traits for signed char since it is not standard
+template<>
+struct char_traits<signed char> : std::char_traits<char>
+{
+    using char_type = signed char;
+    using int_type = uint64_t;
+
+    // Redefine to_int_type function
+    static int_type to_int_type(char_type c) noexcept
+    {
+        return static_cast<int_type>(c);
+    }
+
+    static char_type to_char_type(int_type i) noexcept
+    {
+        return static_cast<char_type>(i);
+    }
+
+    static constexpr int_type eof() noexcept
+    {
+        return static_cast<int_type>(std::char_traits<char>::eof());
+    }
+};
+
+///////////////////
+// is_ functions //
+///////////////////
+
+// https://en.cppreference.com/w/cpp/types/conjunction
+template<class...> struct conjunction : std::true_type { };
+template<class B> struct conjunction<B> : B { };
+template<class B, class... Bn>
+struct conjunction<B, Bn...>
+: std::conditional<static_cast<bool>(B::value), conjunction<Bn...>, B>::type {};
+
+// https://en.cppreference.com/w/cpp/types/negation
+template<class B> struct negation : std::integral_constant < bool, !B::value > { };
+
+// Reimplementation of is_constructible and is_default_constructible, due to them being broken for
+// std::pair and std::tuple until LWG 2367 fix (see https://cplusplus.github.io/LWG/lwg-defects.html#2367).
+// This causes compile errors in e.g. clang 3.5 or gcc 4.9.
+template <typename T>
+struct is_default_constructible : std::is_default_constructible<T> {};
+
+template <typename T1, typename T2>
+struct is_default_constructible<std::pair<T1, T2>>
+    : conjunction<is_default_constructible<T1>, is_default_constructible<T2>> {};
+
+template <typename T1, typename T2>
+struct is_default_constructible<const std::pair<T1, T2>>
+    : conjunction<is_default_constructible<T1>, is_default_constructible<T2>> {};
+
+template <typename... Ts>
+struct is_default_constructible<std::tuple<Ts...>>
+    : conjunction<is_default_constructible<Ts>...> {};
+
+template <typename... Ts>
+struct is_default_constructible<const std::tuple<Ts...>>
+    : conjunction<is_default_constructible<Ts>...> {};
+
+template <typename T, typename... Args>
+struct is_constructible : std::is_constructible<T, Args...> {};
+
+template <typename T1, typename T2>
+struct is_constructible<std::pair<T1, T2>> : is_default_constructible<std::pair<T1, T2>> {};
+
+template <typename T1, typename T2>
+struct is_constructible<const std::pair<T1, T2>> : is_default_constructible<const std::pair<T1, T2>> {};
+
+template <typename... Ts>
+struct is_constructible<std::tuple<Ts...>> : is_default_constructible<std::tuple<Ts...>> {};
+
+template <typename... Ts>
+struct is_constructible<const std::tuple<Ts...>> : is_default_constructible<const std::tuple<Ts...>> {};
+
+template<typename T, typename = void>
+struct is_iterator_traits : std::false_type {};
+
+template<typename T>
+struct is_iterator_traits<iterator_traits<T>>
+{
+  private:
+    using traits = iterator_traits<T>;
+
+  public:
+    static constexpr auto value =
+        is_detected<value_type_t, traits>::value &&
+        is_detected<difference_type_t, traits>::value &&
+        is_detected<pointer_t, traits>::value &&
+        is_detected<iterator_category_t, traits>::value &&
+        is_detected<reference_t, traits>::value;
+};
+
+template<typename T>
+struct is_range
+{
+  private:
+    using t_ref = typename std::add_lvalue_reference<T>::type;
+
+    using iterator = detected_t<result_of_begin, t_ref>;
+    using sentinel = detected_t<result_of_end, t_ref>;
+
+    // to be 100% correct, it should use https://en.cppreference.com/w/cpp/iterator/input_or_output_iterator
+    // and https://en.cppreference.com/w/cpp/iterator/sentinel_for
+    // but reimplementing these would be too much work, as a lot of other concepts are used underneath
+    static constexpr auto is_iterator_begin =
+        is_iterator_traits<iterator_traits<iterator>>::value;
+
+  public:
+    static constexpr bool value = !std::is_same<iterator, nonesuch>::value && !std::is_same<sentinel, nonesuch>::value && is_iterator_begin;
+};
+
+template<typename R>
+using iterator_t = enable_if_t<is_range<R>::value, result_of_begin<decltype(std::declval<R&>())>>;
+
+template<typename T>
+using range_value_t = value_type_t<iterator_traits<iterator_t<T>>>;
+
+// The following implementation of is_complete_type is taken from
+// https://blogs.msdn.microsoft.com/vcblog/2015/12/02/partial-support-for-expression-sfinae-in-vs-2015-update-1/
+// and is written by Xiang Fan who agreed to using it in this library.
+
+template<typename T, typename = void>
+struct is_complete_type : std::false_type {};
+
+template<typename T>
+struct is_complete_type<T, decltype(void(sizeof(T)))> : std::true_type {};
+
+template<typename BasicJsonType, typename CompatibleObjectType,
+         typename = void>
+struct is_compatible_object_type_impl : std::false_type {};
+
+template<typename BasicJsonType, typename CompatibleObjectType>
+struct is_compatible_object_type_impl <
+    BasicJsonType, CompatibleObjectType,
+    enable_if_t < is_detected<mapped_type_t, CompatibleObjectType>::value&&
+    is_detected<key_type_t, CompatibleObjectType>::value >>
+{
+    using object_t = typename BasicJsonType::object_t;
+
+    // macOS's is_constructible does not play well with nonesuch...
+    static constexpr bool value =
+        is_constructible<typename object_t::key_type,
+        typename CompatibleObjectType::key_type>::value &&
+        is_constructible<typename object_t::mapped_type,
+        typename CompatibleObjectType::mapped_type>::value;
+};
+
+template<typename BasicJsonType, typename CompatibleObjectType>
+struct is_compatible_object_type
+    : is_compatible_object_type_impl<BasicJsonType, CompatibleObjectType> {};
+
+template<typename BasicJsonType, typename ConstructibleObjectType,
+         typename = void>
+struct is_constructible_object_type_impl : std::false_type {};
+
+template<typename BasicJsonType, typename ConstructibleObjectType>
+struct is_constructible_object_type_impl <
+    BasicJsonType, ConstructibleObjectType,
+    enable_if_t < is_detected<mapped_type_t, ConstructibleObjectType>::value&&
+    is_detected<key_type_t, ConstructibleObjectType>::value >>
+{
+    using object_t = typename BasicJsonType::object_t;
+
+    static constexpr bool value =
+        (is_default_constructible<ConstructibleObjectType>::value &&
+         (std::is_move_assignable<ConstructibleObjectType>::value ||
+          std::is_copy_assignable<ConstructibleObjectType>::value) &&
+         (is_constructible<typename ConstructibleObjectType::key_type,
+          typename object_t::key_type>::value &&
+          std::is_same <
+          typename object_t::mapped_type,
+          typename ConstructibleObjectType::mapped_type >::value)) ||
+        (has_from_json<BasicJsonType,
+         typename ConstructibleObjectType::mapped_type>::value ||
+         has_non_default_from_json <
+         BasicJsonType,
+         typename ConstructibleObjectType::mapped_type >::value);
+};
+
+template<typename BasicJsonType, typename ConstructibleObjectType>
+struct is_constructible_object_type
+    : is_constructible_object_type_impl<BasicJsonType,
+      ConstructibleObjectType> {};
+
+template<typename BasicJsonType, typename CompatibleStringType>
+struct is_compatible_string_type
+{
+    static constexpr auto value =
+        is_constructible<typename BasicJsonType::string_t, CompatibleStringType>::value;
+};
+
+template<typename BasicJsonType, typename ConstructibleStringType>
+struct is_constructible_string_type
+{
+    // launder type through decltype() to fix compilation failure on ICPC
+#ifdef __INTEL_COMPILER
+    using laundered_type = decltype(std::declval<ConstructibleStringType>());
+#else
+    using laundered_type = ConstructibleStringType;
+#endif
+
+    static constexpr auto value =
+        conjunction <
+        is_constructible<laundered_type, typename BasicJsonType::string_t>,
+        is_detected_exact<typename BasicJsonType::string_t::value_type,
+        value_type_t, laundered_type >>::value;
+};
+
+template<typename BasicJsonType, typename CompatibleArrayType, typename = void>
+struct is_compatible_array_type_impl : std::false_type {};
+
+template<typename BasicJsonType, typename CompatibleArrayType>
+struct is_compatible_array_type_impl <
+    BasicJsonType, CompatibleArrayType,
+    enable_if_t <
+    is_detected<iterator_t, CompatibleArrayType>::value&&
+    is_iterator_traits<iterator_traits<detected_t<iterator_t, CompatibleArrayType>>>::value&&
+// special case for types like std::filesystem::path whose iterator's value_type are themselves
+// c.f. https://github.com/nlohmann/json/pull/3073
+    !std::is_same<CompatibleArrayType, detected_t<range_value_t, CompatibleArrayType>>::value >>
+{
+    static constexpr bool value =
+        is_constructible<BasicJsonType,
+        range_value_t<CompatibleArrayType>>::value;
+};
+
+template<typename BasicJsonType, typename CompatibleArrayType>
+struct is_compatible_array_type
+    : is_compatible_array_type_impl<BasicJsonType, CompatibleArrayType> {};
+
+template<typename BasicJsonType, typename ConstructibleArrayType, typename = void>
+struct is_constructible_array_type_impl : std::false_type {};
+
+template<typename BasicJsonType, typename ConstructibleArrayType>
+struct is_constructible_array_type_impl <
+    BasicJsonType, ConstructibleArrayType,
+    enable_if_t<std::is_same<ConstructibleArrayType,
+    typename BasicJsonType::value_type>::value >>
+            : std::true_type {};
+
+template<typename BasicJsonType, typename ConstructibleArrayType>
+struct is_constructible_array_type_impl <
+    BasicJsonType, ConstructibleArrayType,
+    enable_if_t < !std::is_same<ConstructibleArrayType,
+    typename BasicJsonType::value_type>::value&&
+    !is_compatible_string_type<BasicJsonType, ConstructibleArrayType>::value&&
+    is_default_constructible<ConstructibleArrayType>::value&&
+(std::is_move_assignable<ConstructibleArrayType>::value ||
+ std::is_copy_assignable<ConstructibleArrayType>::value)&&
+is_detected<iterator_t, ConstructibleArrayType>::value&&
+is_iterator_traits<iterator_traits<detected_t<iterator_t, ConstructibleArrayType>>>::value&&
+is_detected<range_value_t, ConstructibleArrayType>::value&&
+// special case for types like std::filesystem::path whose iterator's value_type are themselves
+// c.f. https://github.com/nlohmann/json/pull/3073
+!std::is_same<ConstructibleArrayType, detected_t<range_value_t, ConstructibleArrayType>>::value&&
+is_complete_type <
+detected_t<range_value_t, ConstructibleArrayType >>::value >>
+{
+    using value_type = range_value_t<ConstructibleArrayType>;
+
+    static constexpr bool value =
+        std::is_same<value_type,
+        typename BasicJsonType::array_t::value_type>::value ||
+        has_from_json<BasicJsonType,
+        value_type>::value ||
+        has_non_default_from_json <
+        BasicJsonType,
+        value_type >::value;
+};
+
+template<typename BasicJsonType, typename ConstructibleArrayType>
+struct is_constructible_array_type
+    : is_constructible_array_type_impl<BasicJsonType, ConstructibleArrayType> {};
+
+template<typename RealIntegerType, typename CompatibleNumberIntegerType,
+         typename = void>
+struct is_compatible_integer_type_impl : std::false_type {};
+
+template<typename RealIntegerType, typename CompatibleNumberIntegerType>
+struct is_compatible_integer_type_impl <
+    RealIntegerType, CompatibleNumberIntegerType,
+    enable_if_t < std::is_integral<RealIntegerType>::value&&
+    std::is_integral<CompatibleNumberIntegerType>::value&&
+    !std::is_same<bool, CompatibleNumberIntegerType>::value >>
+{
+    // is there an assert somewhere on overflows?
+    using RealLimits = std::numeric_limits<RealIntegerType>;
+    using CompatibleLimits = std::numeric_limits<CompatibleNumberIntegerType>;
+
+    static constexpr auto value =
+        is_constructible<RealIntegerType,
+        CompatibleNumberIntegerType>::value &&
+        CompatibleLimits::is_integer &&
+        RealLimits::is_signed == CompatibleLimits::is_signed;
+};
+
+template<typename RealIntegerType, typename CompatibleNumberIntegerType>
+struct is_compatible_integer_type
+    : is_compatible_integer_type_impl<RealIntegerType,
+      CompatibleNumberIntegerType> {};
+
+template<typename BasicJsonType, typename CompatibleType, typename = void>
+struct is_compatible_type_impl: std::false_type {};
+
+template<typename BasicJsonType, typename CompatibleType>
+struct is_compatible_type_impl <
+    BasicJsonType, CompatibleType,
+    enable_if_t<is_complete_type<CompatibleType>::value >>
+{
+    static constexpr bool value =
+        has_to_json<BasicJsonType, CompatibleType>::value;
+};
+
+template<typename BasicJsonType, typename CompatibleType>
+struct is_compatible_type
+    : is_compatible_type_impl<BasicJsonType, CompatibleType> {};
+
+template<typename T1, typename T2>
+struct is_constructible_tuple : std::false_type {};
+
+template<typename T1, typename... Args>
+struct is_constructible_tuple<T1, std::tuple<Args...>> : conjunction<is_constructible<T1, Args>...> {};
+
+template<typename BasicJsonType, typename T>
+struct is_json_iterator_of : std::false_type {};
+
+template<typename BasicJsonType>
+struct is_json_iterator_of<BasicJsonType, typename BasicJsonType::iterator> : std::true_type {};
+
+template<typename BasicJsonType>
+struct is_json_iterator_of<BasicJsonType, typename BasicJsonType::const_iterator> : std::true_type
+{};
+
+// checks if a given type T is a template specialization of Primary
+template<template <typename...> class Primary, typename T>
+struct is_specialization_of : std::false_type {};
+
+template<template <typename...> class Primary, typename... Args>
+struct is_specialization_of<Primary, Primary<Args...>> : std::true_type {};
+
+template<typename T>
+using is_json_pointer = is_specialization_of<::nlohmann::json_pointer, uncvref_t<T>>;
+
+// checks if A and B are comparable using Compare functor
+template<typename Compare, typename A, typename B, typename = void>
+struct is_comparable : std::false_type {};
+
+template<typename Compare, typename A, typename B>
+struct is_comparable<Compare, A, B, void_t<
+decltype(std::declval<Compare>()(std::declval<A>(), std::declval<B>())),
+decltype(std::declval<Compare>()(std::declval<B>(), std::declval<A>()))
+>> : std::true_type {};
+
+template<typename T>
+using detect_is_transparent = typename T::is_transparent;
+
+// type trait to check if KeyType can be used as object key (without a BasicJsonType)
+// see is_usable_as_basic_json_key_type below
+template<typename Comparator, typename ObjectKeyType, typename KeyTypeCVRef, bool RequireTransparentComparator = true,
+         bool ExcludeObjectKeyType = RequireTransparentComparator, typename KeyType = uncvref_t<KeyTypeCVRef>>
+using is_usable_as_key_type = typename std::conditional <
+                              is_comparable<Comparator, ObjectKeyType, KeyTypeCVRef>::value
+                              && !(ExcludeObjectKeyType && std::is_same<KeyType,
+                                   ObjectKeyType>::value)
+                              && (!RequireTransparentComparator
+                                  || is_detected <detect_is_transparent, Comparator>::value)
+                              && !is_json_pointer<KeyType>::value,
+                              std::true_type,
+                              std::false_type >::type;
+
+// type trait to check if KeyType can be used as object key
+// true if:
+//   - KeyType is comparable with BasicJsonType::object_t::key_type
+//   - if ExcludeObjectKeyType is true, KeyType is not BasicJsonType::object_t::key_type
+//   - the comparator is transparent or RequireTransparentComparator is false
+//   - KeyType is not a JSON iterator or json_pointer
+template<typename BasicJsonType, typename KeyTypeCVRef, bool RequireTransparentComparator = true,
+         bool ExcludeObjectKeyType = RequireTransparentComparator, typename KeyType = uncvref_t<KeyTypeCVRef>>
+using is_usable_as_basic_json_key_type = typename std::conditional <
+    is_usable_as_key_type<typename BasicJsonType::object_comparator_t,
+    typename BasicJsonType::object_t::key_type, KeyTypeCVRef,
+    RequireTransparentComparator, ExcludeObjectKeyType>::value
+    && !is_json_iterator_of<BasicJsonType, KeyType>::value,
+    std::true_type,
+    std::false_type >::type;
+
+template<typename ObjectType, typename KeyType>
+using detect_erase_with_key_type = decltype(std::declval<ObjectType&>().erase(std::declval<KeyType>()));
+
+// type trait to check if object_t has an erase() member functions accepting KeyType
+template<typename BasicJsonType, typename KeyType>
+using has_erase_with_key_type = typename std::conditional <
+                                is_detected <
+                                detect_erase_with_key_type,
+                                typename BasicJsonType::object_t, KeyType >::value,
+                                std::true_type,
+                                std::false_type >::type;
+
+// a naive helper to check if a type is an ordered_map (exploits the fact that
+// ordered_map inherits capacity() from std::vector)
+template <typename T>
+struct is_ordered_map
+{
+    using one = char;
+
+    struct two
+    {
+        char x[2]; // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+    };
+
+    template <typename C> static one test( decltype(&C::capacity) ) ;
+    template <typename C> static two test(...);
+
+    enum { value = sizeof(test<T>(nullptr)) == sizeof(char) }; // NOLINT(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+};
+
+// to avoid useless casts (see https://github.com/nlohmann/json/issues/2893#issuecomment-889152324)
+template < typename T, typename U, enable_if_t < !std::is_same<T, U>::value, int > = 0 >
+T conditional_static_cast(U value)
+{
+    return static_cast<T>(value);
+}
+
+template<typename T, typename U, enable_if_t<std::is_same<T, U>::value, int> = 0>
+T conditional_static_cast(U value)
+{
+    return value;
+}
+
+template<typename... Types>
+using all_integral = conjunction<std::is_integral<Types>...>;
+
+template<typename... Types>
+using all_signed = conjunction<std::is_signed<Types>...>;
+
+template<typename... Types>
+using all_unsigned = conjunction<std::is_unsigned<Types>...>;
+
+// there's a disjunction trait in another PR; replace when merged
+template<typename... Types>
+using same_sign = std::integral_constant < bool,
+      all_signed<Types...>::value || all_unsigned<Types...>::value >;
+
+template<typename OfType, typename T>
+using never_out_of_range = std::integral_constant < bool,
+      (std::is_signed<OfType>::value && (sizeof(T) < sizeof(OfType)))
+      || (same_sign<OfType, T>::value && sizeof(OfType) == sizeof(T)) >;
+
+template<typename OfType, typename T,
+         bool OfTypeSigned = std::is_signed<OfType>::value,
+         bool TSigned = std::is_signed<T>::value>
+struct value_in_range_of_impl2;
+
+template<typename OfType, typename T>
+struct value_in_range_of_impl2<OfType, T, false, false>
+{
+    static constexpr bool test(T val)
+    {
+        using CommonType = typename std::common_type<OfType, T>::type;
+        return static_cast<CommonType>(val) <= static_cast<CommonType>((std::numeric_limits<OfType>::max)());
+    }
+};
+
+template<typename OfType, typename T>
+struct value_in_range_of_impl2<OfType, T, true, false>
+{
+    static constexpr bool test(T val)
+    {
+        using CommonType = typename std::common_type<OfType, T>::type;
+        return static_cast<CommonType>(val) <= static_cast<CommonType>((std::numeric_limits<OfType>::max)());
+    }
+};
+
+template<typename OfType, typename T>
+struct value_in_range_of_impl2<OfType, T, false, true>
+{
+    static constexpr bool test(T val)
+    {
+        using CommonType = typename std::common_type<OfType, T>::type;
+        return val >= 0 && static_cast<CommonType>(val) <= static_cast<CommonType>((std::numeric_limits<OfType>::max)());
+    }
+};
+
+template<typename OfType, typename T>
+struct value_in_range_of_impl2<OfType, T, true, true>
+{
+    static constexpr bool test(T val)
+    {
+        using CommonType = typename std::common_type<OfType, T>::type;
+        return static_cast<CommonType>(val) >= static_cast<CommonType>((std::numeric_limits<OfType>::min)())
+               && static_cast<CommonType>(val) <= static_cast<CommonType>((std::numeric_limits<OfType>::max)());
+    }
+};
+
+template<typename OfType, typename T,
+         bool NeverOutOfRange = never_out_of_range<OfType, T>::value,
+         typename = detail::enable_if_t<all_integral<OfType, T>::value>>
+struct value_in_range_of_impl1;
+
+template<typename OfType, typename T>
+struct value_in_range_of_impl1<OfType, T, false>
+{
+    static constexpr bool test(T val)
+    {
+        return value_in_range_of_impl2<OfType, T>::test(val);
+    }
+};
+
+template<typename OfType, typename T>
+struct value_in_range_of_impl1<OfType, T, true>
+{
+    static constexpr bool test(T /*val*/)
+    {
+        return true;
+    }
+};
+
+template<typename OfType, typename T>
+constexpr bool value_in_range_of(T val)
+{
+    return value_in_range_of_impl1<OfType, T>::test(val);
+}
+
+template<bool Value>
+using bool_constant = std::integral_constant<bool, Value>;
+
+///////////////////////////////////////////////////////////////////////////////
+// is_c_string
+///////////////////////////////////////////////////////////////////////////////
+
+namespace impl
+{
+
+template<typename T>
+constexpr bool is_c_string()
+{
+    using TUnExt = typename std::remove_extent<T>::type;
+    using TUnCVExt = typename std::remove_cv<TUnExt>::type;
+    using TUnPtr = typename std::remove_pointer<T>::type;
+    using TUnCVPtr = typename std::remove_cv<TUnPtr>::type;
+    return
+        (std::is_array<T>::value && std::is_same<TUnCVExt, char>::value)
+        || (std::is_pointer<T>::value && std::is_same<TUnCVPtr, char>::value);
+}
+
+}  // namespace impl
+
+// checks whether T is a [cv] char */[cv] char[] C string
+template<typename T>
+struct is_c_string : bool_constant<impl::is_c_string<T>()> {};
+
+template<typename T>
+using is_c_string_uncvref = is_c_string<uncvref_t<T>>;
+
+///////////////////////////////////////////////////////////////////////////////
+// is_transparent
+///////////////////////////////////////////////////////////////////////////////
+
+namespace impl
+{
+
+template<typename T>
+constexpr bool is_transparent()
+{
+    return is_detected<detect_is_transparent, T>::value;
+}
+
+}  // namespace impl
+
+// checks whether T has a member named is_transparent
+template<typename T>
+struct is_transparent : bool_constant<impl::is_transparent<T>()> {};
+
+///////////////////////////////////////////////////////////////////////////////
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/string_concat.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstring> // strlen
+#include <string> // string
+#include <utility> // forward
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
+// #include <nlohmann/detail/meta/detected.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+inline std::size_t concat_length()
+{
+    return 0;
+}
+
+template<typename... Args>
+inline std::size_t concat_length(const char* cstr, const Args& ... rest);
+
+template<typename StringType, typename... Args>
+inline std::size_t concat_length(const StringType& str, const Args& ... rest);
+
+template<typename... Args>
+inline std::size_t concat_length(const char /*c*/, const Args& ... rest)
+{
+    return 1 + concat_length(rest...);
+}
+
+template<typename... Args>
+inline std::size_t concat_length(const char* cstr, const Args& ... rest)
+{
+    // cppcheck-suppress ignoredReturnValue
+    return ::strlen(cstr) + concat_length(rest...);
+}
+
+template<typename StringType, typename... Args>
+inline std::size_t concat_length(const StringType& str, const Args& ... rest)
+{
+    return str.size() + concat_length(rest...);
+}
+
+template<typename OutStringType>
+inline void concat_into(OutStringType& /*out*/)
+{}
+
+template<typename StringType, typename Arg>
+using string_can_append = decltype(std::declval<StringType&>().append(std::declval < Arg && > ()));
+
+template<typename StringType, typename Arg>
+using detect_string_can_append = is_detected<string_can_append, StringType, Arg>;
+
+template<typename StringType, typename Arg>
+using string_can_append_op = decltype(std::declval<StringType&>() += std::declval < Arg && > ());
+
+template<typename StringType, typename Arg>
+using detect_string_can_append_op = is_detected<string_can_append_op, StringType, Arg>;
+
+template<typename StringType, typename Arg>
+using string_can_append_iter = decltype(std::declval<StringType&>().append(std::declval<const Arg&>().begin(), std::declval<const Arg&>().end()));
+
+template<typename StringType, typename Arg>
+using detect_string_can_append_iter = is_detected<string_can_append_iter, StringType, Arg>;
+
+template<typename StringType, typename Arg>
+using string_can_append_data = decltype(std::declval<StringType&>().append(std::declval<const Arg&>().data(), std::declval<const Arg&>().size()));
+
+template<typename StringType, typename Arg>
+using detect_string_can_append_data = is_detected<string_can_append_data, StringType, Arg>;
+
+template < typename OutStringType, typename Arg, typename... Args,
+           enable_if_t < !detect_string_can_append<OutStringType, Arg>::value
+                         && detect_string_can_append_op<OutStringType, Arg>::value, int > = 0 >
+inline void concat_into(OutStringType& out, Arg && arg, Args && ... rest);
+
+template < typename OutStringType, typename Arg, typename... Args,
+           enable_if_t < !detect_string_can_append<OutStringType, Arg>::value
+                         && !detect_string_can_append_op<OutStringType, Arg>::value
+                         && detect_string_can_append_iter<OutStringType, Arg>::value, int > = 0 >
+inline void concat_into(OutStringType& out, const Arg& arg, Args && ... rest);
+
+template < typename OutStringType, typename Arg, typename... Args,
+           enable_if_t < !detect_string_can_append<OutStringType, Arg>::value
+                         && !detect_string_can_append_op<OutStringType, Arg>::value
+                         && !detect_string_can_append_iter<OutStringType, Arg>::value
+                         && detect_string_can_append_data<OutStringType, Arg>::value, int > = 0 >
+inline void concat_into(OutStringType& out, const Arg& arg, Args && ... rest);
+
+template<typename OutStringType, typename Arg, typename... Args,
+         enable_if_t<detect_string_can_append<OutStringType, Arg>::value, int> = 0>
+inline void concat_into(OutStringType& out, Arg && arg, Args && ... rest)
+{
+    out.append(std::forward<Arg>(arg));
+    concat_into(out, std::forward<Args>(rest)...);
+}
+
+template < typename OutStringType, typename Arg, typename... Args,
+           enable_if_t < !detect_string_can_append<OutStringType, Arg>::value
+                         && detect_string_can_append_op<OutStringType, Arg>::value, int > >
+inline void concat_into(OutStringType& out, Arg&& arg, Args&& ... rest)
+{
+    out += std::forward<Arg>(arg);
+    concat_into(out, std::forward<Args>(rest)...);
+}
+
+template < typename OutStringType, typename Arg, typename... Args,
+           enable_if_t < !detect_string_can_append<OutStringType, Arg>::value
+                         && !detect_string_can_append_op<OutStringType, Arg>::value
+                         && detect_string_can_append_iter<OutStringType, Arg>::value, int > >
+inline void concat_into(OutStringType& out, const Arg& arg, Args&& ... rest)
+{
+    out.append(arg.begin(), arg.end());
+    concat_into(out, std::forward<Args>(rest)...);
+}
+
+template < typename OutStringType, typename Arg, typename... Args,
+           enable_if_t < !detect_string_can_append<OutStringType, Arg>::value
+                         && !detect_string_can_append_op<OutStringType, Arg>::value
+                         && !detect_string_can_append_iter<OutStringType, Arg>::value
+                         && detect_string_can_append_data<OutStringType, Arg>::value, int > >
+inline void concat_into(OutStringType& out, const Arg& arg, Args&& ... rest)
+{
+    out.append(arg.data(), arg.size());
+    concat_into(out, std::forward<Args>(rest)...);
+}
+
+template<typename OutStringType = std::string, typename... Args>
+inline OutStringType concat(Args && ... args)
+{
+    OutStringType str;
+    str.reserve(concat_length(args...));
+    concat_into(str, std::forward<Args>(args)...);
+    return str;
+}
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+
+// With -Wweak-vtables, Clang will complain about the exception classes as they
+// have no out-of-line virtual method definitions and their vtable will be
+// emitted in every translation unit. This issue cannot be fixed with a
+// header-only library as there is no implementation file to move these
+// functions to. As a result, we suppress this warning here to avoid client
+// code to stumble over this. See https://github.com/nlohmann/json/issues/4087
+// for a discussion.
+#if defined(__clang__)
+    #pragma clang diagnostic push
+    #pragma clang diagnostic ignored "-Wweak-vtables"
+#endif
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+////////////////
+// exceptions //
+////////////////
+
+/// @brief general exception of the @ref basic_json class
+/// @sa https://json.nlohmann.me/api/basic_json/exception/
+class exception : public std::exception
+{
+  public:
+    /// returns the explanatory string
+    const char* what() const noexcept override
+    {
+        return m.what();
+    }
+
+    /// the id of the exception
+    const int id; // NOLINT(cppcoreguidelines-non-private-member-variables-in-classes)
+
+  protected:
+    JSON_HEDLEY_NON_NULL(3)
+    exception(int id_, const char* what_arg) : id(id_), m(what_arg) {} // NOLINT(bugprone-throw-keyword-missing)
+
+    static std::string name(const std::string& ename, int id_)
+    {
+        return concat("[json.exception.", ename, '.', std::to_string(id_), "] ");
+    }
+
+    static std::string diagnostics(std::nullptr_t /*leaf_element*/)
+    {
+        return "";
+    }
+
+    template<typename BasicJsonType>
+    static std::string diagnostics(const BasicJsonType* leaf_element)
+    {
+#if JSON_DIAGNOSTICS
+        std::vector<std::string> tokens;
+        for (const auto* current = leaf_element; current != nullptr && current->m_parent != nullptr; current = current->m_parent)
+        {
+            switch (current->m_parent->type())
+            {
+                case value_t::array:
+                {
+                    for (std::size_t i = 0; i < current->m_parent->m_data.m_value.array->size(); ++i)
+                    {
+                        if (&current->m_parent->m_data.m_value.array->operator[](i) == current)
+                        {
+                            tokens.emplace_back(std::to_string(i));
+                            break;
+                        }
+                    }
+                    break;
+                }
+
+                case value_t::object:
+                {
+                    for (const auto& element : *current->m_parent->m_data.m_value.object)
+                    {
+                        if (&element.second == current)
+                        {
+                            tokens.emplace_back(element.first.c_str());
+                            break;
+                        }
+                    }
+                    break;
+                }
+
+                case value_t::null: // LCOV_EXCL_LINE
+                case value_t::string: // LCOV_EXCL_LINE
+                case value_t::boolean: // LCOV_EXCL_LINE
+                case value_t::number_integer: // LCOV_EXCL_LINE
+                case value_t::number_unsigned: // LCOV_EXCL_LINE
+                case value_t::number_float: // LCOV_EXCL_LINE
+                case value_t::binary: // LCOV_EXCL_LINE
+                case value_t::discarded: // LCOV_EXCL_LINE
+                default:   // LCOV_EXCL_LINE
+                    break; // LCOV_EXCL_LINE
+            }
+        }
+
+        if (tokens.empty())
+        {
+            return "";
+        }
+
+        auto str = std::accumulate(tokens.rbegin(), tokens.rend(), std::string{},
+                                   [](const std::string & a, const std::string & b)
+        {
+            return concat(a, '/', detail::escape(b));
+        });
+
+        return concat('(', str, ") ", get_byte_positions(leaf_element));
+#else
+        return get_byte_positions(leaf_element);
+#endif
+    }
+
+  private:
+    /// an exception object as storage for error messages
+    std::runtime_error m;
+#if JSON_DIAGNOSTIC_POSITIONS
+    template<typename BasicJsonType>
+    static std::string get_byte_positions(const BasicJsonType* leaf_element)
+    {
+        if ((leaf_element->start_pos() != std::string::npos) && (leaf_element->end_pos() != std::string::npos))
+        {
+            return concat("(bytes ", std::to_string(leaf_element->start_pos()), "-", std::to_string(leaf_element->end_pos()), ") ");
+        }
+        return "";
+    }
+#else
+    template<typename BasicJsonType>
+    static std::string get_byte_positions(const BasicJsonType* leaf_element)
+    {
+        static_cast<void>(leaf_element);
+        return "";
+    }
+#endif
+};
+
+/// @brief exception indicating a parse error
+/// @sa https://json.nlohmann.me/api/basic_json/parse_error/
+class parse_error : public exception
+{
+  public:
+    /*!
+    @brief create a parse error exception
+    @param[in] id_       the id of the exception
+    @param[in] pos       the position where the error occurred (or with
+                         chars_read_total=0 if the position cannot be
+                         determined)
+    @param[in] what_arg  the explanatory string
+    @return parse_error object
+    */
+    template<typename BasicJsonContext, enable_if_t<is_basic_json_context<BasicJsonContext>::value, int> = 0>
+    static parse_error create(int id_, const position_t& pos, const std::string& what_arg, BasicJsonContext context)
+    {
+        const std::string w = concat(exception::name("parse_error", id_), "parse error",
+                                     position_string(pos), ": ", exception::diagnostics(context), what_arg);
+        return {id_, pos.chars_read_total, w.c_str()};
+    }
+
+    template<typename BasicJsonContext, enable_if_t<is_basic_json_context<BasicJsonContext>::value, int> = 0>
+    static parse_error create(int id_, std::size_t byte_, const std::string& what_arg, BasicJsonContext context)
+    {
+        const std::string w = concat(exception::name("parse_error", id_), "parse error",
+                                     (byte_ != 0 ? (concat(" at byte ", std::to_string(byte_))) : ""),
+                                     ": ", exception::diagnostics(context), what_arg);
+        return {id_, byte_, w.c_str()};
+    }
+
+    /*!
+    @brief byte index of the parse error
+
+    The byte index of the last read character in the input file.
+
+    @note For an input with n bytes, 1 is the index of the first character and
+          n+1 is the index of the terminating null byte or the end of file.
+          This also holds true when reading a byte vector (CBOR or MessagePack).
+    */
+    const std::size_t byte;
+
+  private:
+    parse_error(int id_, std::size_t byte_, const char* what_arg)
+        : exception(id_, what_arg), byte(byte_) {}
+
+    static std::string position_string(const position_t& pos)
+    {
+        return concat(" at line ", std::to_string(pos.lines_read + 1),
+                      ", column ", std::to_string(pos.chars_read_current_line));
+    }
+};
+
+/// @brief exception indicating errors with iterators
+/// @sa https://json.nlohmann.me/api/basic_json/invalid_iterator/
+class invalid_iterator : public exception
+{
+  public:
+    template<typename BasicJsonContext, enable_if_t<is_basic_json_context<BasicJsonContext>::value, int> = 0>
+    static invalid_iterator create(int id_, const std::string& what_arg, BasicJsonContext context)
+    {
+        const std::string w = concat(exception::name("invalid_iterator", id_), exception::diagnostics(context), what_arg);
+        return {id_, w.c_str()};
+    }
+
+  private:
+    JSON_HEDLEY_NON_NULL(3)
+    invalid_iterator(int id_, const char* what_arg)
+        : exception(id_, what_arg) {}
+};
+
+/// @brief exception indicating executing a member function with a wrong type
+/// @sa https://json.nlohmann.me/api/basic_json/type_error/
+class type_error : public exception
+{
+  public:
+    template<typename BasicJsonContext, enable_if_t<is_basic_json_context<BasicJsonContext>::value, int> = 0>
+    static type_error create(int id_, const std::string& what_arg, BasicJsonContext context)
+    {
+        const std::string w = concat(exception::name("type_error", id_), exception::diagnostics(context), what_arg);
+        return {id_, w.c_str()};
+    }
+
+  private:
+    JSON_HEDLEY_NON_NULL(3)
+    type_error(int id_, const char* what_arg) : exception(id_, what_arg) {}
+};
+
+/// @brief exception indicating access out of the defined range
+/// @sa https://json.nlohmann.me/api/basic_json/out_of_range/
+class out_of_range : public exception
+{
+  public:
+    template<typename BasicJsonContext, enable_if_t<is_basic_json_context<BasicJsonContext>::value, int> = 0>
+    static out_of_range create(int id_, const std::string& what_arg, BasicJsonContext context)
+    {
+        const std::string w = concat(exception::name("out_of_range", id_), exception::diagnostics(context), what_arg);
+        return {id_, w.c_str()};
+    }
+
+  private:
+    JSON_HEDLEY_NON_NULL(3)
+    out_of_range(int id_, const char* what_arg) : exception(id_, what_arg) {}
+};
+
+/// @brief exception indicating other library errors
+/// @sa https://json.nlohmann.me/api/basic_json/other_error/
+class other_error : public exception
+{
+  public:
+    template<typename BasicJsonContext, enable_if_t<is_basic_json_context<BasicJsonContext>::value, int> = 0>
+    static other_error create(int id_, const std::string& what_arg, BasicJsonContext context)
+    {
+        const std::string w = concat(exception::name("other_error", id_), exception::diagnostics(context), what_arg);
+        return {id_, w.c_str()};
+    }
+
+  private:
+    JSON_HEDLEY_NON_NULL(3)
+    other_error(int id_, const char* what_arg) : exception(id_, what_arg) {}
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+#if defined(__clang__)
+    #pragma clang diagnostic pop
+#endif
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
+// #include <nlohmann/detail/meta/identity_tag.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+// dispatching helper struct
+template <class T> struct identity_tag {};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/meta/std_fs.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+
+#if JSON_HAS_EXPERIMENTAL_FILESYSTEM
+#include <experimental/filesystem>
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+namespace std_fs = std::experimental::filesystem;
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+#elif JSON_HAS_FILESYSTEM
+#include <filesystem> // NOLINT(build/c++17)
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+namespace std_fs = std::filesystem;
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+#endif
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+// #include <nlohmann/detail/string_concat.hpp>
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+template<typename BasicJsonType>
+inline void from_json(const BasicJsonType& j, typename std::nullptr_t& n)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_null()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be null, but is ", j.type_name()), &j));
+    }
+    n = nullptr;
+}
+
+#ifdef JSON_HAS_CPP_17
+#ifndef JSON_USE_IMPLICIT_CONVERSIONS
+template<typename BasicJsonType, typename T>
+void from_json(const BasicJsonType& j, std::optional<T>& opt)
+{
+    if (j.is_null())
+    {
+        opt = std::nullopt;
+    }
+    else
+    {
+        opt.emplace(j.template get<T>());
+    }
+}
+
+#endif // JSON_USE_IMPLICIT_CONVERSIONS
+#endif // JSON_HAS_CPP_17
+
+// overloads for basic_json template parameters
+template < typename BasicJsonType, typename ArithmeticType,
+           enable_if_t < std::is_arithmetic<ArithmeticType>::value&&
+                         !std::is_same<ArithmeticType, typename BasicJsonType::boolean_t>::value,
+                         int > = 0 >
+void get_arithmetic_value(const BasicJsonType& j, ArithmeticType& val)
+{
+    switch (static_cast<value_t>(j))
+    {
+        case value_t::number_unsigned:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_unsigned_t*>());
+            break;
+        }
+        case value_t::number_integer:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_integer_t*>());
+            break;
+        }
+        case value_t::number_float:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_float_t*>());
+            break;
+        }
+
+        case value_t::null:
+        case value_t::object:
+        case value_t::array:
+        case value_t::string:
+        case value_t::boolean:
+        case value_t::binary:
+        case value_t::discarded:
+        default:
+            JSON_THROW(type_error::create(302, concat("type must be number, but is ", j.type_name()), &j));
+    }
+}
+
+template<typename BasicJsonType>
+inline void from_json(const BasicJsonType& j, typename BasicJsonType::boolean_t& b)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_boolean()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be boolean, but is ", j.type_name()), &j));
+    }
+    b = *j.template get_ptr<const typename BasicJsonType::boolean_t*>();
+}
+
+template<typename BasicJsonType>
+inline void from_json(const BasicJsonType& j, typename BasicJsonType::string_t& s)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_string()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be string, but is ", j.type_name()), &j));
+    }
+    s = *j.template get_ptr<const typename BasicJsonType::string_t*>();
+}
+
+template <
+    typename BasicJsonType, typename StringType,
+    enable_if_t <
+        std::is_assignable<StringType&, const typename BasicJsonType::string_t>::value
+        && is_detected_exact<typename BasicJsonType::string_t::value_type, value_type_t, StringType>::value
+        && !std::is_same<typename BasicJsonType::string_t, StringType>::value
+        && !is_json_ref<StringType>::value, int > = 0 >
+inline void from_json(const BasicJsonType& j, StringType& s)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_string()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be string, but is ", j.type_name()), &j));
+    }
+
+    s = *j.template get_ptr<const typename BasicJsonType::string_t*>();
+}
+
+template<typename BasicJsonType>
+inline void from_json(const BasicJsonType& j, typename BasicJsonType::number_float_t& val)
+{
+    get_arithmetic_value(j, val);
+}
+
+template<typename BasicJsonType>
+inline void from_json(const BasicJsonType& j, typename BasicJsonType::number_unsigned_t& val)
+{
+    get_arithmetic_value(j, val);
+}
+
+template<typename BasicJsonType>
+inline void from_json(const BasicJsonType& j, typename BasicJsonType::number_integer_t& val)
+{
+    get_arithmetic_value(j, val);
+}
+
+#if !JSON_DISABLE_ENUM_SERIALIZATION
+template<typename BasicJsonType, typename EnumType,
+         enable_if_t<std::is_enum<EnumType>::value, int> = 0>
+inline void from_json(const BasicJsonType& j, EnumType& e)
+{
+    typename std::underlying_type<EnumType>::type val;
+    get_arithmetic_value(j, val);
+    e = static_cast<EnumType>(val);
+}
+#endif  // JSON_DISABLE_ENUM_SERIALIZATION
+
+// forward_list doesn't have an insert method
+template<typename BasicJsonType, typename T, typename Allocator,
+         enable_if_t<is_getable<BasicJsonType, T>::value, int> = 0>
+inline void from_json(const BasicJsonType& j, std::forward_list<T, Allocator>& l)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be array, but is ", j.type_name()), &j));
+    }
+    l.clear();
+    std::transform(j.rbegin(), j.rend(),
+                   std::front_inserter(l), [](const BasicJsonType & i)
+    {
+        return i.template get<T>();
+    });
+}
+
+// valarray doesn't have an insert method
+template<typename BasicJsonType, typename T,
+         enable_if_t<is_getable<BasicJsonType, T>::value, int> = 0>
+inline void from_json(const BasicJsonType& j, std::valarray<T>& l)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be array, but is ", j.type_name()), &j));
+    }
+    l.resize(j.size());
+    std::transform(j.begin(), j.end(), std::begin(l),
+                   [](const BasicJsonType & elem)
+    {
+        return elem.template get<T>();
+    });
+}
+
+template<typename BasicJsonType, typename T, std::size_t N>
+auto from_json(const BasicJsonType& j, T (&arr)[N])  // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+-> decltype(j.template get<T>(), void())
+{
+    for (std::size_t i = 0; i < N; ++i)
+    {
+        arr[i] = j.at(i).template get<T>();
+    }
+}
+
+template<typename BasicJsonType, typename T, std::size_t N1, std::size_t N2>
+auto from_json(const BasicJsonType& j, T (&arr)[N1][N2])  // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+-> decltype(j.template get<T>(), void())
+{
+    for (std::size_t i1 = 0; i1 < N1; ++i1)
+    {
+        for (std::size_t i2 = 0; i2 < N2; ++i2)
+        {
+            arr[i1][i2] = j.at(i1).at(i2).template get<T>();
+        }
+    }
+}
+
+template<typename BasicJsonType, typename T, std::size_t N1, std::size_t N2, std::size_t N3>
+auto from_json(const BasicJsonType& j, T (&arr)[N1][N2][N3])  // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+-> decltype(j.template get<T>(), void())
+{
+    for (std::size_t i1 = 0; i1 < N1; ++i1)
+    {
+        for (std::size_t i2 = 0; i2 < N2; ++i2)
+        {
+            for (std::size_t i3 = 0; i3 < N3; ++i3)
+            {
+                arr[i1][i2][i3] = j.at(i1).at(i2).at(i3).template get<T>();
+            }
+        }
+    }
+}
+
+template<typename BasicJsonType, typename T, std::size_t N1, std::size_t N2, std::size_t N3, std::size_t N4>
+auto from_json(const BasicJsonType& j, T (&arr)[N1][N2][N3][N4])  // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+-> decltype(j.template get<T>(), void())
+{
+    for (std::size_t i1 = 0; i1 < N1; ++i1)
+    {
+        for (std::size_t i2 = 0; i2 < N2; ++i2)
+        {
+            for (std::size_t i3 = 0; i3 < N3; ++i3)
+            {
+                for (std::size_t i4 = 0; i4 < N4; ++i4)
+                {
+                    arr[i1][i2][i3][i4] = j.at(i1).at(i2).at(i3).at(i4).template get<T>();
+                }
+            }
+        }
+    }
+}
+
+template<typename BasicJsonType>
+inline void from_json_array_impl(const BasicJsonType& j, typename BasicJsonType::array_t& arr, priority_tag<3> /*unused*/)
+{
+    arr = *j.template get_ptr<const typename BasicJsonType::array_t*>();
+}
+
+template<typename BasicJsonType, typename T, std::size_t N>
+auto from_json_array_impl(const BasicJsonType& j, std::array<T, N>& arr,
+                          priority_tag<2> /*unused*/)
+-> decltype(j.template get<T>(), void())
+{
+    for (std::size_t i = 0; i < N; ++i)
+    {
+        arr[i] = j.at(i).template get<T>();
+    }
+}
+
+template<typename BasicJsonType, typename ConstructibleArrayType,
+         enable_if_t<
+             std::is_assignable<ConstructibleArrayType&, ConstructibleArrayType>::value,
+             int> = 0>
+auto from_json_array_impl(const BasicJsonType& j, ConstructibleArrayType& arr, priority_tag<1> /*unused*/)
+-> decltype(
+    arr.reserve(std::declval<typename ConstructibleArrayType::size_type>()),
+    j.template get<typename ConstructibleArrayType::value_type>(),
+    void())
+{
+    using std::end;
+
+    ConstructibleArrayType ret;
+    ret.reserve(j.size());
+    std::transform(j.begin(), j.end(),
+                   std::inserter(ret, end(ret)), [](const BasicJsonType & i)
+    {
+        // get<BasicJsonType>() returns *this, this won't call a from_json
+        // method when value_type is BasicJsonType
+        return i.template get<typename ConstructibleArrayType::value_type>();
+    });
+    arr = std::move(ret);
+}
+
+template<typename BasicJsonType, typename ConstructibleArrayType,
+         enable_if_t<
+             std::is_assignable<ConstructibleArrayType&, ConstructibleArrayType>::value,
+             int> = 0>
+inline void from_json_array_impl(const BasicJsonType& j, ConstructibleArrayType& arr,
+                                 priority_tag<0> /*unused*/)
+{
+    using std::end;
+
+    ConstructibleArrayType ret;
+    std::transform(
+        j.begin(), j.end(), std::inserter(ret, end(ret)),
+        [](const BasicJsonType & i)
+    {
+        // get<BasicJsonType>() returns *this, this won't call a from_json
+        // method when value_type is BasicJsonType
+        return i.template get<typename ConstructibleArrayType::value_type>();
+    });
+    arr = std::move(ret);
+}
+
+template < typename BasicJsonType, typename ConstructibleArrayType,
+           enable_if_t <
+               is_constructible_array_type<BasicJsonType, ConstructibleArrayType>::value&&
+               !is_constructible_object_type<BasicJsonType, ConstructibleArrayType>::value&&
+               !is_constructible_string_type<BasicJsonType, ConstructibleArrayType>::value&&
+               !std::is_same<ConstructibleArrayType, typename BasicJsonType::binary_t>::value&&
+               !is_basic_json<ConstructibleArrayType>::value,
+               int > = 0 >
+auto from_json(const BasicJsonType& j, ConstructibleArrayType& arr)
+-> decltype(from_json_array_impl(j, arr, priority_tag<3> {}),
+j.template get<typename ConstructibleArrayType::value_type>(),
+void())
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be array, but is ", j.type_name()), &j));
+    }
+
+    from_json_array_impl(j, arr, priority_tag<3> {});
+}
+
+template < typename BasicJsonType, typename T, std::size_t... Idx >
+std::array<T, sizeof...(Idx)> from_json_inplace_array_impl(BasicJsonType&& j,
+                     identity_tag<std::array<T, sizeof...(Idx)>> /*unused*/, index_sequence<Idx...> /*unused*/)
+{
+    return { { std::forward<BasicJsonType>(j).at(Idx).template get<T>()... } };
+}
+
+template < typename BasicJsonType, typename T, std::size_t N >
+auto from_json(BasicJsonType&& j, identity_tag<std::array<T, N>> tag)
+-> decltype(from_json_inplace_array_impl(std::forward<BasicJsonType>(j), tag, make_index_sequence<N> {}))
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be array, but is ", j.type_name()), &j));
+    }
+
+    return from_json_inplace_array_impl(std::forward<BasicJsonType>(j), tag, make_index_sequence<N> {});
+}
+
+template<typename BasicJsonType>
+inline void from_json(const BasicJsonType& j, typename BasicJsonType::binary_t& bin)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_binary()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be binary, but is ", j.type_name()), &j));
+    }
+
+    bin = *j.template get_ptr<const typename BasicJsonType::binary_t*>();
+}
+
+template<typename BasicJsonType, typename ConstructibleObjectType,
+         enable_if_t<is_constructible_object_type<BasicJsonType, ConstructibleObjectType>::value, int> = 0>
+inline void from_json(const BasicJsonType& j, ConstructibleObjectType& obj)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_object()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be object, but is ", j.type_name()), &j));
+    }
+
+    ConstructibleObjectType ret;
+    const auto* inner_object = j.template get_ptr<const typename BasicJsonType::object_t*>();
+    using value_type = typename ConstructibleObjectType::value_type;
+    std::transform(
+        inner_object->begin(), inner_object->end(),
+        std::inserter(ret, ret.begin()),
+        [](typename BasicJsonType::object_t::value_type const & p)
+    {
+        return value_type(p.first, p.second.template get<typename ConstructibleObjectType::mapped_type>());
+    });
+    obj = std::move(ret);
+}
+
+// overload for arithmetic types, not chosen for basic_json template arguments
+// (BooleanType, etc..); note: Is it really necessary to provide explicit
+// overloads for boolean_t etc. in case of a custom BooleanType which is not
+// an arithmetic type?
+template < typename BasicJsonType, typename ArithmeticType,
+           enable_if_t <
+               std::is_arithmetic<ArithmeticType>::value&&
+               !std::is_same<ArithmeticType, typename BasicJsonType::number_unsigned_t>::value&&
+               !std::is_same<ArithmeticType, typename BasicJsonType::number_integer_t>::value&&
+               !std::is_same<ArithmeticType, typename BasicJsonType::number_float_t>::value&&
+               !std::is_same<ArithmeticType, typename BasicJsonType::boolean_t>::value,
+               int > = 0 >
+inline void from_json(const BasicJsonType& j, ArithmeticType& val)
+{
+    switch (static_cast<value_t>(j))
+    {
+        case value_t::number_unsigned:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_unsigned_t*>());
+            break;
+        }
+        case value_t::number_integer:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_integer_t*>());
+            break;
+        }
+        case value_t::number_float:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::number_float_t*>());
+            break;
+        }
+        case value_t::boolean:
+        {
+            val = static_cast<ArithmeticType>(*j.template get_ptr<const typename BasicJsonType::boolean_t*>());
+            break;
+        }
+
+        case value_t::null:
+        case value_t::object:
+        case value_t::array:
+        case value_t::string:
+        case value_t::binary:
+        case value_t::discarded:
+        default:
+            JSON_THROW(type_error::create(302, concat("type must be number, but is ", j.type_name()), &j));
+    }
+}
+
+template<typename BasicJsonType, typename... Args, std::size_t... Idx>
+std::tuple<Args...> from_json_tuple_impl_base(BasicJsonType&& j, index_sequence<Idx...> /*unused*/)
+{
+    return std::make_tuple(std::forward<BasicJsonType>(j).at(Idx).template get<Args>()...);
+}
+
+template<typename BasicJsonType>
+std::tuple<> from_json_tuple_impl_base(BasicJsonType& /*unused*/, index_sequence<> /*unused*/)
+{
+    return {};
+}
+
+template < typename BasicJsonType, class A1, class A2 >
+std::pair<A1, A2> from_json_tuple_impl(BasicJsonType&& j, identity_tag<std::pair<A1, A2>> /*unused*/, priority_tag<0> /*unused*/)
+{
+    return {std::forward<BasicJsonType>(j).at(0).template get<A1>(),
+            std::forward<BasicJsonType>(j).at(1).template get<A2>()};
+}
+
+template<typename BasicJsonType, typename A1, typename A2>
+inline void from_json_tuple_impl(BasicJsonType&& j, std::pair<A1, A2>& p, priority_tag<1> /*unused*/)
+{
+    p = from_json_tuple_impl(std::forward<BasicJsonType>(j), identity_tag<std::pair<A1, A2>> {}, priority_tag<0> {});
+}
+
+template<typename BasicJsonType, typename... Args>
+std::tuple<Args...> from_json_tuple_impl(BasicJsonType&& j, identity_tag<std::tuple<Args...>> /*unused*/, priority_tag<2> /*unused*/)
+{
+    return from_json_tuple_impl_base<BasicJsonType, Args...>(std::forward<BasicJsonType>(j), index_sequence_for<Args...> {});
+}
+
+template<typename BasicJsonType, typename... Args>
+inline void from_json_tuple_impl(BasicJsonType&& j, std::tuple<Args...>& t, priority_tag<3> /*unused*/)
+{
+    t = from_json_tuple_impl_base<BasicJsonType, Args...>(std::forward<BasicJsonType>(j), index_sequence_for<Args...> {});
+}
+
+template<typename BasicJsonType, typename TupleRelated>
+auto from_json(BasicJsonType&& j, TupleRelated&& t)
+-> decltype(from_json_tuple_impl(std::forward<BasicJsonType>(j), std::forward<TupleRelated>(t), priority_tag<3> {}))
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be array, but is ", j.type_name()), &j));
+    }
+
+    return from_json_tuple_impl(std::forward<BasicJsonType>(j), std::forward<TupleRelated>(t), priority_tag<3> {});
+}
+
+template < typename BasicJsonType, typename Key, typename Value, typename Compare, typename Allocator,
+           typename = enable_if_t < !std::is_constructible <
+                                        typename BasicJsonType::string_t, Key >::value >>
+inline void from_json(const BasicJsonType& j, std::map<Key, Value, Compare, Allocator>& m)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be array, but is ", j.type_name()), &j));
+    }
+    m.clear();
+    for (const auto& p : j)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!p.is_array()))
+        {
+            JSON_THROW(type_error::create(302, concat("type must be array, but is ", p.type_name()), &j));
+        }
+        m.emplace(p.at(0).template get<Key>(), p.at(1).template get<Value>());
+    }
+}
+
+template < typename BasicJsonType, typename Key, typename Value, typename Hash, typename KeyEqual, typename Allocator,
+           typename = enable_if_t < !std::is_constructible <
+                                        typename BasicJsonType::string_t, Key >::value >>
+inline void from_json(const BasicJsonType& j, std::unordered_map<Key, Value, Hash, KeyEqual, Allocator>& m)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_array()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be array, but is ", j.type_name()), &j));
+    }
+    m.clear();
+    for (const auto& p : j)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!p.is_array()))
+        {
+            JSON_THROW(type_error::create(302, concat("type must be array, but is ", p.type_name()), &j));
+        }
+        m.emplace(p.at(0).template get<Key>(), p.at(1).template get<Value>());
+    }
+}
+
+#if JSON_HAS_FILESYSTEM || JSON_HAS_EXPERIMENTAL_FILESYSTEM
+template<typename BasicJsonType>
+inline void from_json(const BasicJsonType& j, std_fs::path& p)
+{
+    if (JSON_HEDLEY_UNLIKELY(!j.is_string()))
+    {
+        JSON_THROW(type_error::create(302, concat("type must be string, but is ", j.type_name()), &j));
+    }
+    p = *j.template get_ptr<const typename BasicJsonType::string_t*>();
+}
+#endif
+
+struct from_json_fn
+{
+    template<typename BasicJsonType, typename T>
+    auto operator()(const BasicJsonType& j, T&& val) const
+    noexcept(noexcept(from_json(j, std::forward<T>(val))))
+    -> decltype(from_json(j, std::forward<T>(val)))
+    {
+        return from_json(j, std::forward<T>(val));
+    }
+};
+
+}  // namespace detail
+
+#ifndef JSON_HAS_CPP_17
+/// namespace to hold default `from_json` function
+/// to see why this is required:
+/// http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2015/n4381.html
+namespace // NOLINT(cert-dcl59-cpp,fuchsia-header-anon-namespaces,google-build-namespaces)
+{
+#endif
+JSON_INLINE_VARIABLE constexpr const auto& from_json = // NOLINT(misc-definitions-in-headers)
+    detail::static_const<detail::from_json_fn>::value;
+#ifndef JSON_HAS_CPP_17
+}  // namespace
+#endif
+
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/conversions/to_json.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// #include <nlohmann/detail/macro_scope.hpp>
+// JSON_HAS_CPP_17
+#ifdef JSON_HAS_CPP_17
+    #include <optional> // optional
+#endif
+
+#include <algorithm> // copy
+#include <iterator> // begin, end
+#include <string> // string
+#include <tuple> // tuple, get
+#include <type_traits> // is_same, is_constructible, is_floating_point, is_enum, underlying_type
+#include <utility> // move, forward, declval, pair
+#include <valarray> // valarray
+#include <vector> // vector
+
+// #include <nlohmann/detail/iterators/iteration_proxy.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstddef> // size_t
+#include <iterator> // forward_iterator_tag
+#include <tuple> // tuple_size, get, tuple_element
+#include <utility> // move
+
+#if JSON_HAS_RANGES
+    #include <ranges> // enable_borrowed_range
+#endif
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+// #include <nlohmann/detail/string_utils.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstddef> // size_t
+#include <string> // string, to_string
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+template<typename StringType>
+void int_to_string(StringType& target, std::size_t value)
+{
+    // For ADL
+    using std::to_string;
+    target = to_string(value);
+}
+
+template<typename StringType>
+StringType to_string(std::size_t value)
+{
+    StringType result;
+    int_to_string(result, value);
+    return result;
+}
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+template<typename IteratorType> class iteration_proxy_value
+{
+  public:
+    using difference_type = std::ptrdiff_t;
+    using value_type = iteration_proxy_value;
+    using pointer = value_type *;
+    using reference = value_type &;
+    using iterator_category = std::forward_iterator_tag;
+    using string_type = typename std::remove_cv< typename std::remove_reference<decltype( std::declval<IteratorType>().key() ) >::type >::type;
+
+  private:
+    /// the iterator
+    IteratorType anchor{};
+    /// an index for arrays (used to create key names)
+    std::size_t array_index = 0;
+    /// last stringified array index
+    mutable std::size_t array_index_last = 0;
+    /// a string representation of the array index
+    mutable string_type array_index_str = "0";
+    /// an empty string (to return a reference for primitive values)
+    string_type empty_str{};
+
+  public:
+    explicit iteration_proxy_value() = default;
+    explicit iteration_proxy_value(IteratorType it, std::size_t array_index_ = 0)
+    noexcept(std::is_nothrow_move_constructible<IteratorType>::value
+             && std::is_nothrow_default_constructible<string_type>::value)
+        : anchor(std::move(it))
+        , array_index(array_index_)
+    {}
+
+    iteration_proxy_value(iteration_proxy_value const&) = default;
+    iteration_proxy_value& operator=(iteration_proxy_value const&) = default;
+    // older GCCs are a bit fussy and require explicit noexcept specifiers on defaulted functions
+    iteration_proxy_value(iteration_proxy_value&&)
+    noexcept(std::is_nothrow_move_constructible<IteratorType>::value
+             && std::is_nothrow_move_constructible<string_type>::value) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor,cppcoreguidelines-noexcept-move-operations)
+    iteration_proxy_value& operator=(iteration_proxy_value&&)
+    noexcept(std::is_nothrow_move_assignable<IteratorType>::value
+             && std::is_nothrow_move_assignable<string_type>::value) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor,cppcoreguidelines-noexcept-move-operations)
+    ~iteration_proxy_value() = default;
+
+    /// dereference operator (needed for range-based for)
+    const iteration_proxy_value& operator*() const
+    {
+        return *this;
+    }
+
+    /// increment operator (needed for range-based for)
+    iteration_proxy_value& operator++()
+    {
+        ++anchor;
+        ++array_index;
+
+        return *this;
+    }
+
+    iteration_proxy_value operator++(int)& // NOLINT(cert-dcl21-cpp)
+    {
+        auto tmp = iteration_proxy_value(anchor, array_index);
+        ++anchor;
+        ++array_index;
+        return tmp;
+    }
+
+    /// equality operator (needed for InputIterator)
+    bool operator==(const iteration_proxy_value& o) const
+    {
+        return anchor == o.anchor;
+    }
+
+    /// inequality operator (needed for range-based for)
+    bool operator!=(const iteration_proxy_value& o) const
+    {
+        return anchor != o.anchor;
+    }
+
+    /// return key of the iterator
+    const string_type& key() const
+    {
+        JSON_ASSERT(anchor.m_object != nullptr);
+
+        switch (anchor.m_object->type())
+        {
+            // use integer array index as key
+            case value_t::array:
+            {
+                if (array_index != array_index_last)
+                {
+                    int_to_string( array_index_str, array_index );
+                    array_index_last = array_index;
+                }
+                return array_index_str;
+            }
+
+            // use key from the object
+            case value_t::object:
+                return anchor.key();
+
+            // use an empty key for all primitive types
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+                return empty_str;
+        }
+    }
+
+    /// return value of the iterator
+    typename IteratorType::reference value() const
+    {
+        return anchor.value();
+    }
+};
+
+/// proxy class for the items() function
+template<typename IteratorType> class iteration_proxy
+{
+  private:
+    /// the container to iterate
+    typename IteratorType::pointer container = nullptr;
+
+  public:
+    explicit iteration_proxy() = default;
+
+    /// construct iteration proxy from a container
+    explicit iteration_proxy(typename IteratorType::reference cont) noexcept
+        : container(&cont) {}
+
+    iteration_proxy(iteration_proxy const&) = default;
+    iteration_proxy& operator=(iteration_proxy const&) = default;
+    iteration_proxy(iteration_proxy&&) noexcept = default;
+    iteration_proxy& operator=(iteration_proxy&&) noexcept = default;
+    ~iteration_proxy() = default;
+
+    /// return iterator begin (needed for range-based for)
+    iteration_proxy_value<IteratorType> begin() const noexcept
+    {
+        return iteration_proxy_value<IteratorType>(container->begin());
+    }
+
+    /// return iterator end (needed for range-based for)
+    iteration_proxy_value<IteratorType> end() const noexcept
+    {
+        return iteration_proxy_value<IteratorType>(container->end());
+    }
+};
+
+// Structured Bindings Support
+// For further reference see https://blog.tartanllama.xyz/structured-bindings/
+// And see https://github.com/nlohmann/json/pull/1391
+template<std::size_t N, typename IteratorType, enable_if_t<N == 0, int> = 0>
+auto get(const nlohmann::detail::iteration_proxy_value<IteratorType>& i) -> decltype(i.key())
+{
+    return i.key();
+}
+// Structured Bindings Support
+// For further reference see https://blog.tartanllama.xyz/structured-bindings/
+// And see https://github.com/nlohmann/json/pull/1391
+template<std::size_t N, typename IteratorType, enable_if_t<N == 1, int> = 0>
+auto get(const nlohmann::detail::iteration_proxy_value<IteratorType>& i) -> decltype(i.value())
+{
+    return i.value();
+}
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// The Addition to the STD Namespace is required to add
+// Structured Bindings Support to the iteration_proxy_value class
+// For further reference see https://blog.tartanllama.xyz/structured-bindings/
+// And see https://github.com/nlohmann/json/pull/1391
+namespace std
+{
+
+#if defined(__clang__)
+    // Fix: https://github.com/nlohmann/json/issues/1401
+    #pragma clang diagnostic push
+    #pragma clang diagnostic ignored "-Wmismatched-tags"
+#endif
+template<typename IteratorType>
+class tuple_size<::nlohmann::detail::iteration_proxy_value<IteratorType>> // NOLINT(cert-dcl58-cpp)
+    : public std::integral_constant<std::size_t, 2> {};
+
+template<std::size_t N, typename IteratorType>
+class tuple_element<N, ::nlohmann::detail::iteration_proxy_value<IteratorType >> // NOLINT(cert-dcl58-cpp)
+{
+  public:
+    using type = decltype(
+                     get<N>(std::declval <
+                            ::nlohmann::detail::iteration_proxy_value<IteratorType >> ()));
+};
+#if defined(__clang__)
+    #pragma clang diagnostic pop
+#endif
+
+}  // namespace std
+
+#if JSON_HAS_RANGES
+    template <typename IteratorType>
+    inline constexpr bool ::std::ranges::enable_borrowed_range<::nlohmann::detail::iteration_proxy<IteratorType>> = true;
+#endif
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
+// #include <nlohmann/detail/meta/std_fs.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+//////////////////
+// constructors //
+//////////////////
+
+/*
+ * Note all external_constructor<>::construct functions need to call
+ * j.m_data.m_value.destroy(j.m_data.m_type) to avoid a memory leak in case j contains an
+ * allocated value (e.g., a string). See bug issue
+ * https://github.com/nlohmann/json/issues/2865 for more information.
+ */
+
+template<value_t> struct external_constructor;
+
+template<>
+struct external_constructor<value_t::boolean>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::boolean_t b) noexcept
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::boolean;
+        j.m_data.m_value = b;
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::string>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const typename BasicJsonType::string_t& s)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::string;
+        j.m_data.m_value = s;
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::string_t&& s)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::string;
+        j.m_data.m_value = std::move(s);
+        j.assert_invariant();
+    }
+
+    template < typename BasicJsonType, typename CompatibleStringType,
+               enable_if_t < !std::is_same<CompatibleStringType, typename BasicJsonType::string_t>::value,
+                             int > = 0 >
+    static void construct(BasicJsonType& j, const CompatibleStringType& str)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::string;
+        j.m_data.m_value.string = j.template create<typename BasicJsonType::string_t>(str);
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::binary>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const typename BasicJsonType::binary_t& b)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::binary;
+        j.m_data.m_value = typename BasicJsonType::binary_t(b);
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::binary_t&& b)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::binary;
+        j.m_data.m_value = typename BasicJsonType::binary_t(std::move(b));
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::number_float>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::number_float_t val) noexcept
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::number_float;
+        j.m_data.m_value = val;
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::number_unsigned>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::number_unsigned_t val) noexcept
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::number_unsigned;
+        j.m_data.m_value = val;
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::number_integer>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::number_integer_t val) noexcept
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::number_integer;
+        j.m_data.m_value = val;
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::array>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const typename BasicJsonType::array_t& arr)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::array;
+        j.m_data.m_value = arr;
+        j.set_parents();
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::array_t&& arr)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::array;
+        j.m_data.m_value = std::move(arr);
+        j.set_parents();
+        j.assert_invariant();
+    }
+
+    template < typename BasicJsonType, typename CompatibleArrayType,
+               enable_if_t < !std::is_same<CompatibleArrayType, typename BasicJsonType::array_t>::value,
+                             int > = 0 >
+    static void construct(BasicJsonType& j, const CompatibleArrayType& arr)
+    {
+        using std::begin;
+        using std::end;
+
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::array;
+        j.m_data.m_value.array = j.template create<typename BasicJsonType::array_t>(begin(arr), end(arr));
+        j.set_parents();
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const std::vector<bool>& arr)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::array;
+        j.m_data.m_value = value_t::array;
+        j.m_data.m_value.array->reserve(arr.size());
+        for (const bool x : arr)
+        {
+            j.m_data.m_value.array->push_back(x);
+            j.set_parent(j.m_data.m_value.array->back());
+        }
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType, typename T,
+             enable_if_t<std::is_convertible<T, BasicJsonType>::value, int> = 0>
+    static void construct(BasicJsonType& j, const std::valarray<T>& arr)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::array;
+        j.m_data.m_value = value_t::array;
+        j.m_data.m_value.array->resize(arr.size());
+        if (arr.size() > 0)
+        {
+            std::copy(std::begin(arr), std::end(arr), j.m_data.m_value.array->begin());
+        }
+        j.set_parents();
+        j.assert_invariant();
+    }
+};
+
+template<>
+struct external_constructor<value_t::object>
+{
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, const typename BasicJsonType::object_t& obj)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::object;
+        j.m_data.m_value = obj;
+        j.set_parents();
+        j.assert_invariant();
+    }
+
+    template<typename BasicJsonType>
+    static void construct(BasicJsonType& j, typename BasicJsonType::object_t&& obj)
+    {
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::object;
+        j.m_data.m_value = std::move(obj);
+        j.set_parents();
+        j.assert_invariant();
+    }
+
+    template < typename BasicJsonType, typename CompatibleObjectType,
+               enable_if_t < !std::is_same<CompatibleObjectType, typename BasicJsonType::object_t>::value, int > = 0 >
+    static void construct(BasicJsonType& j, const CompatibleObjectType& obj)
+    {
+        using std::begin;
+        using std::end;
+
+        j.m_data.m_value.destroy(j.m_data.m_type);
+        j.m_data.m_type = value_t::object;
+        j.m_data.m_value.object = j.template create<typename BasicJsonType::object_t>(begin(obj), end(obj));
+        j.set_parents();
+        j.assert_invariant();
+    }
+};
+
+/////////////
+// to_json //
+/////////////
+
+#ifdef JSON_HAS_CPP_17
+template<typename BasicJsonType, typename T,
+         enable_if_t<std::is_constructible<BasicJsonType, T>::value, int> = 0>
+void to_json(BasicJsonType& j, const std::optional<T>& opt)
+{
+    if (opt.has_value())
+    {
+        j = *opt;
+    }
+    else
+    {
+        j = nullptr;
+    }
+}
+#endif
+
+template<typename BasicJsonType, typename T,
+         enable_if_t<std::is_same<T, typename BasicJsonType::boolean_t>::value, int> = 0>
+inline void to_json(BasicJsonType& j, T b) noexcept
+{
+    external_constructor<value_t::boolean>::construct(j, b);
+}
+
+template < typename BasicJsonType, typename BoolRef,
+           enable_if_t <
+               ((std::is_same<std::vector<bool>::reference, BoolRef>::value
+                 && !std::is_same <std::vector<bool>::reference, typename BasicJsonType::boolean_t&>::value)
+                || (std::is_same<std::vector<bool>::const_reference, BoolRef>::value
+                    && !std::is_same <detail::uncvref_t<std::vector<bool>::const_reference>,
+                                      typename BasicJsonType::boolean_t >::value))
+               && std::is_convertible<const BoolRef&, typename BasicJsonType::boolean_t>::value, int > = 0 >
+inline void to_json(BasicJsonType& j, const BoolRef& b) noexcept
+{
+    external_constructor<value_t::boolean>::construct(j, static_cast<typename BasicJsonType::boolean_t>(b));
+}
+
+template<typename BasicJsonType, typename CompatibleString,
+         enable_if_t<std::is_constructible<typename BasicJsonType::string_t, CompatibleString>::value, int> = 0>
+inline void to_json(BasicJsonType& j, const CompatibleString& s)
+{
+    external_constructor<value_t::string>::construct(j, s);
+}
+
+template<typename BasicJsonType>
+inline void to_json(BasicJsonType& j, typename BasicJsonType::string_t&& s)
+{
+    external_constructor<value_t::string>::construct(j, std::move(s));
+}
+
+template<typename BasicJsonType, typename FloatType,
+         enable_if_t<std::is_floating_point<FloatType>::value, int> = 0>
+inline void to_json(BasicJsonType& j, FloatType val) noexcept
+{
+    external_constructor<value_t::number_float>::construct(j, static_cast<typename BasicJsonType::number_float_t>(val));
+}
+
+template<typename BasicJsonType, typename CompatibleNumberUnsignedType,
+         enable_if_t<is_compatible_integer_type<typename BasicJsonType::number_unsigned_t, CompatibleNumberUnsignedType>::value, int> = 0>
+inline void to_json(BasicJsonType& j, CompatibleNumberUnsignedType val) noexcept
+{
+    external_constructor<value_t::number_unsigned>::construct(j, static_cast<typename BasicJsonType::number_unsigned_t>(val));
+}
+
+template<typename BasicJsonType, typename CompatibleNumberIntegerType,
+         enable_if_t<is_compatible_integer_type<typename BasicJsonType::number_integer_t, CompatibleNumberIntegerType>::value, int> = 0>
+inline void to_json(BasicJsonType& j, CompatibleNumberIntegerType val) noexcept
+{
+    external_constructor<value_t::number_integer>::construct(j, static_cast<typename BasicJsonType::number_integer_t>(val));
+}
+
+#if !JSON_DISABLE_ENUM_SERIALIZATION
+template<typename BasicJsonType, typename EnumType,
+         enable_if_t<std::is_enum<EnumType>::value, int> = 0>
+inline void to_json(BasicJsonType& j, EnumType e) noexcept
+{
+    using underlying_type = typename std::underlying_type<EnumType>::type;
+    static constexpr value_t integral_value_t = std::is_unsigned<underlying_type>::value ? value_t::number_unsigned : value_t::number_integer;
+    external_constructor<integral_value_t>::construct(j, static_cast<underlying_type>(e));
+}
+#endif  // JSON_DISABLE_ENUM_SERIALIZATION
+
+template<typename BasicJsonType>
+inline void to_json(BasicJsonType& j, const std::vector<bool>& e)
+{
+    external_constructor<value_t::array>::construct(j, e);
+}
+
+template < typename BasicJsonType, typename CompatibleArrayType,
+           enable_if_t < is_compatible_array_type<BasicJsonType,
+                         CompatibleArrayType>::value&&
+                         !is_compatible_object_type<BasicJsonType, CompatibleArrayType>::value&&
+                         !is_compatible_string_type<BasicJsonType, CompatibleArrayType>::value&&
+                         !std::is_same<typename BasicJsonType::binary_t, CompatibleArrayType>::value&&
+                         !is_basic_json<CompatibleArrayType>::value,
+                         int > = 0 >
+inline void to_json(BasicJsonType& j, const CompatibleArrayType& arr)
+{
+    external_constructor<value_t::array>::construct(j, arr);
+}
+
+template<typename BasicJsonType>
+inline void to_json(BasicJsonType& j, const typename BasicJsonType::binary_t& bin)
+{
+    external_constructor<value_t::binary>::construct(j, bin);
+}
+
+template<typename BasicJsonType, typename T,
+         enable_if_t<std::is_convertible<T, BasicJsonType>::value, int> = 0>
+inline void to_json(BasicJsonType& j, const std::valarray<T>& arr)
+{
+    external_constructor<value_t::array>::construct(j, std::move(arr));
+}
+
+template<typename BasicJsonType>
+inline void to_json(BasicJsonType& j, typename BasicJsonType::array_t&& arr)
+{
+    external_constructor<value_t::array>::construct(j, std::move(arr));
+}
+
+template < typename BasicJsonType, typename CompatibleObjectType,
+           enable_if_t < is_compatible_object_type<BasicJsonType, CompatibleObjectType>::value&& !is_basic_json<CompatibleObjectType>::value, int > = 0 >
+inline void to_json(BasicJsonType& j, const CompatibleObjectType& obj)
+{
+    external_constructor<value_t::object>::construct(j, obj);
+}
+
+template<typename BasicJsonType>
+inline void to_json(BasicJsonType& j, typename BasicJsonType::object_t&& obj)
+{
+    external_constructor<value_t::object>::construct(j, std::move(obj));
+}
+
+template <
+    typename BasicJsonType, typename T, std::size_t N,
+    enable_if_t < !std::is_constructible<typename BasicJsonType::string_t,
+                  const T(&)[N]>::value, // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+                  int > = 0 >
+inline void to_json(BasicJsonType& j, const T(&arr)[N]) // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+{
+    external_constructor<value_t::array>::construct(j, arr);
+}
+
+template < typename BasicJsonType, typename T1, typename T2, enable_if_t < std::is_constructible<BasicJsonType, T1>::value&& std::is_constructible<BasicJsonType, T2>::value, int > = 0 >
+inline void to_json(BasicJsonType& j, const std::pair<T1, T2>& p)
+{
+    j = { p.first, p.second };
+}
+
+// for https://github.com/nlohmann/json/pull/1134
+template<typename BasicJsonType, typename T,
+         enable_if_t<std::is_same<T, iteration_proxy_value<typename BasicJsonType::iterator>>::value, int> = 0>
+inline void to_json(BasicJsonType& j, const T& b)
+{
+    j = { {b.key(), b.value()} };
+}
+
+template<typename BasicJsonType, typename Tuple, std::size_t... Idx>
+inline void to_json_tuple_impl(BasicJsonType& j, const Tuple& t, index_sequence<Idx...> /*unused*/)
+{
+    j = { std::get<Idx>(t)... };
+}
+
+template<typename BasicJsonType, typename Tuple>
+inline void to_json_tuple_impl(BasicJsonType& j, const Tuple& /*unused*/, index_sequence<> /*unused*/)
+{
+    using array_t = typename BasicJsonType::array_t;
+    j = array_t();
+}
+
+template<typename BasicJsonType, typename T, enable_if_t<is_constructible_tuple<BasicJsonType, T>::value, int > = 0>
+inline void to_json(BasicJsonType& j, const T& t)
+{
+    to_json_tuple_impl(j, t, make_index_sequence<std::tuple_size<T>::value> {});
+}
+
+#if JSON_HAS_FILESYSTEM || JSON_HAS_EXPERIMENTAL_FILESYSTEM
+template<typename BasicJsonType>
+inline void to_json(BasicJsonType& j, const std_fs::path& p)
+{
+    j = p.string();
+}
+#endif
+
+struct to_json_fn
+{
+    template<typename BasicJsonType, typename T>
+    auto operator()(BasicJsonType& j, T&& val) const noexcept(noexcept(to_json(j, std::forward<T>(val))))
+    -> decltype(to_json(j, std::forward<T>(val)), void())
+    {
+        return to_json(j, std::forward<T>(val));
+    }
+};
+}  // namespace detail
+
+#ifndef JSON_HAS_CPP_17
+/// namespace to hold default `to_json` function
+/// to see why this is required:
+/// http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2015/n4381.html
+namespace // NOLINT(cert-dcl59-cpp,fuchsia-header-anon-namespaces,google-build-namespaces)
+{
+#endif
+JSON_INLINE_VARIABLE constexpr const auto& to_json = // NOLINT(misc-definitions-in-headers)
+    detail::static_const<detail::to_json_fn>::value;
+#ifndef JSON_HAS_CPP_17
+}  // namespace
+#endif
+
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/meta/identity_tag.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+/// @sa https://json.nlohmann.me/api/adl_serializer/
+template<typename ValueType, typename>
+struct adl_serializer
+{
+    /// @brief convert a JSON value to any value type
+    /// @sa https://json.nlohmann.me/api/adl_serializer/from_json/
+    template<typename BasicJsonType, typename TargetType = ValueType>
+    static auto from_json(BasicJsonType && j, TargetType& val) noexcept(
+        noexcept(::nlohmann::from_json(std::forward<BasicJsonType>(j), val)))
+    -> decltype(::nlohmann::from_json(std::forward<BasicJsonType>(j), val), void())
+    {
+        ::nlohmann::from_json(std::forward<BasicJsonType>(j), val);
+    }
+
+    /// @brief convert a JSON value to any value type
+    /// @sa https://json.nlohmann.me/api/adl_serializer/from_json/
+    template<typename BasicJsonType, typename TargetType = ValueType>
+    static auto from_json(BasicJsonType && j) noexcept(
+    noexcept(::nlohmann::from_json(std::forward<BasicJsonType>(j), detail::identity_tag<TargetType> {})))
+    -> decltype(::nlohmann::from_json(std::forward<BasicJsonType>(j), detail::identity_tag<TargetType> {}))
+    {
+        return ::nlohmann::from_json(std::forward<BasicJsonType>(j), detail::identity_tag<TargetType> {});
+    }
+
+    /// @brief convert any value type to a JSON value
+    /// @sa https://json.nlohmann.me/api/adl_serializer/to_json/
+    template<typename BasicJsonType, typename TargetType = ValueType>
+    static auto to_json(BasicJsonType& j, TargetType && val) noexcept(
+        noexcept(::nlohmann::to_json(j, std::forward<TargetType>(val))))
+    -> decltype(::nlohmann::to_json(j, std::forward<TargetType>(val)), void())
+    {
+        ::nlohmann::to_json(j, std::forward<TargetType>(val));
+    }
+};
+
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/byte_container_with_subtype.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstdint> // uint8_t, uint64_t
+#include <tuple> // tie
+#include <utility> // move
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+/// @brief an internal type for a backed binary type
+/// @sa https://json.nlohmann.me/api/byte_container_with_subtype/
+template<typename BinaryType>
+class byte_container_with_subtype : public BinaryType
+{
+  public:
+    using container_type = BinaryType;
+    using subtype_type = std::uint64_t;
+
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/byte_container_with_subtype/
+    byte_container_with_subtype() noexcept(noexcept(container_type()))
+        : container_type()
+    {}
+
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/byte_container_with_subtype/
+    byte_container_with_subtype(const container_type& b) noexcept(noexcept(container_type(b)))
+        : container_type(b)
+    {}
+
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/byte_container_with_subtype/
+    byte_container_with_subtype(container_type&& b) noexcept(noexcept(container_type(std::move(b))))
+        : container_type(std::move(b))
+    {}
+
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/byte_container_with_subtype/
+    byte_container_with_subtype(const container_type& b, subtype_type subtype_) noexcept(noexcept(container_type(b)))
+        : container_type(b)
+        , m_subtype(subtype_)
+        , m_has_subtype(true)
+    {}
+
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/byte_container_with_subtype/
+    byte_container_with_subtype(container_type&& b, subtype_type subtype_) noexcept(noexcept(container_type(std::move(b))))
+        : container_type(std::move(b))
+        , m_subtype(subtype_)
+        , m_has_subtype(true)
+    {}
+
+    bool operator==(const byte_container_with_subtype& rhs) const
+    {
+        return std::tie(static_cast<const BinaryType&>(*this), m_subtype, m_has_subtype) ==
+               std::tie(static_cast<const BinaryType&>(rhs), rhs.m_subtype, rhs.m_has_subtype);
+    }
+
+    bool operator!=(const byte_container_with_subtype& rhs) const
+    {
+        return !(rhs == *this);
+    }
+
+    /// @brief sets the binary subtype
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/set_subtype/
+    void set_subtype(subtype_type subtype_) noexcept
+    {
+        m_subtype = subtype_;
+        m_has_subtype = true;
+    }
+
+    /// @brief return the binary subtype
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/subtype/
+    constexpr subtype_type subtype() const noexcept
+    {
+        return m_has_subtype ? m_subtype : static_cast<subtype_type>(-1);
+    }
+
+    /// @brief return whether the value has a subtype
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/has_subtype/
+    constexpr bool has_subtype() const noexcept
+    {
+        return m_has_subtype;
+    }
+
+    /// @brief clears the binary subtype
+    /// @sa https://json.nlohmann.me/api/byte_container_with_subtype/clear_subtype/
+    void clear_subtype() noexcept
+    {
+        m_subtype = 0;
+        m_has_subtype = false;
+    }
+
+  private:
+    subtype_type m_subtype = 0;
+    bool m_has_subtype = false;
+};
+
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/conversions/from_json.hpp>
+
+// #include <nlohmann/detail/conversions/to_json.hpp>
+
+// #include <nlohmann/detail/exceptions.hpp>
+
+// #include <nlohmann/detail/hash.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstdint> // uint8_t
+#include <cstddef> // size_t
+#include <functional> // hash
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+// boost::hash_combine
+inline std::size_t combine(std::size_t seed, std::size_t h) noexcept
+{
+    seed ^= h + 0x9e3779b9 + (seed << 6U) + (seed >> 2U);
+    return seed;
+}
+
+/*!
+@brief hash a JSON value
+
+The hash function tries to rely on std::hash where possible. Furthermore, the
+type of the JSON value is taken into account to have different hash values for
+null, 0, 0U, and false, etc.
+
+@tparam BasicJsonType basic_json specialization
+@param j JSON value to hash
+@return hash value of j
+*/
+template<typename BasicJsonType>
+std::size_t hash(const BasicJsonType& j)
+{
+    using string_t = typename BasicJsonType::string_t;
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+
+    const auto type = static_cast<std::size_t>(j.type());
+    switch (j.type())
+    {
+        case BasicJsonType::value_t::null:
+        case BasicJsonType::value_t::discarded:
+        {
+            return combine(type, 0);
+        }
+
+        case BasicJsonType::value_t::object:
+        {
+            auto seed = combine(type, j.size());
+            for (const auto& element : j.items())
+            {
+                const auto h = std::hash<string_t> {}(element.key());
+                seed = combine(seed, h);
+                seed = combine(seed, hash(element.value()));
+            }
+            return seed;
+        }
+
+        case BasicJsonType::value_t::array:
+        {
+            auto seed = combine(type, j.size());
+            for (const auto& element : j)
+            {
+                seed = combine(seed, hash(element));
+            }
+            return seed;
+        }
+
+        case BasicJsonType::value_t::string:
+        {
+            const auto h = std::hash<string_t> {}(j.template get_ref<const string_t&>());
+            return combine(type, h);
+        }
+
+        case BasicJsonType::value_t::boolean:
+        {
+            const auto h = std::hash<bool> {}(j.template get<bool>());
+            return combine(type, h);
+        }
+
+        case BasicJsonType::value_t::number_integer:
+        {
+            const auto h = std::hash<number_integer_t> {}(j.template get<number_integer_t>());
+            return combine(type, h);
+        }
+
+        case BasicJsonType::value_t::number_unsigned:
+        {
+            const auto h = std::hash<number_unsigned_t> {}(j.template get<number_unsigned_t>());
+            return combine(type, h);
+        }
+
+        case BasicJsonType::value_t::number_float:
+        {
+            const auto h = std::hash<number_float_t> {}(j.template get<number_float_t>());
+            return combine(type, h);
+        }
+
+        case BasicJsonType::value_t::binary:
+        {
+            auto seed = combine(type, j.get_binary().size());
+            const auto h = std::hash<bool> {}(j.get_binary().has_subtype());
+            seed = combine(seed, h);
+            seed = combine(seed, static_cast<std::size_t>(j.get_binary().subtype()));
+            for (const auto byte : j.get_binary())
+            {
+                seed = combine(seed, std::hash<std::uint8_t> {}(byte));
+            }
+            return seed;
+        }
+
+        default:                   // LCOV_EXCL_LINE
+            JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+            return 0;              // LCOV_EXCL_LINE
+    }
+}
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/input/binary_reader.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <algorithm> // generate_n
+#include <array> // array
+#include <cmath> // ldexp
+#include <cstddef> // size_t
+#include <cstdint> // uint8_t, uint16_t, uint32_t, uint64_t
+#include <cstdio> // snprintf
+#include <cstring> // memcpy
+#include <iterator> // back_inserter
+#include <limits> // numeric_limits
+#include <string> // char_traits, string
+#include <utility> // make_pair, move
+#include <vector> // vector
+#ifdef __cpp_lib_byteswap
+    #include <bit>  //byteswap
+#endif
+
+// #include <nlohmann/detail/exceptions.hpp>
+
+// #include <nlohmann/detail/input/input_adapters.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <array> // array
+#include <cstddef> // size_t
+#include <cstring> // strlen
+#include <iterator> // begin, end, iterator_traits, random_access_iterator_tag, distance, next
+#include <memory> // shared_ptr, make_shared, addressof
+#include <numeric> // accumulate
+#include <string> // string, char_traits
+#include <type_traits> // enable_if, is_base_of, is_pointer, is_integral, remove_pointer
+#include <utility> // pair, declval
+
+#ifndef JSON_NO_IO
+    #include <cstdio>   // FILE *
+    #include <istream>  // istream
+#endif                  // JSON_NO_IO
+
+// #include <nlohmann/detail/exceptions.hpp>
+
+// #include <nlohmann/detail/iterators/iterator_traits.hpp>
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/// the supported input formats
+enum class input_format_t { json, cbor, msgpack, ubjson, bson, bjdata };
+
+////////////////////
+// input adapters //
+////////////////////
+
+#ifndef JSON_NO_IO
+/*!
+Input adapter for stdio file access. This adapter read only 1 byte and do not use any
+ buffer. This adapter is a very low level adapter.
+*/
+class file_input_adapter
+{
+  public:
+    using char_type = char;
+
+    JSON_HEDLEY_NON_NULL(2)
+    explicit file_input_adapter(std::FILE* f) noexcept
+        : m_file(f)
+    {
+        JSON_ASSERT(m_file != nullptr);
+    }
+
+    // make class move-only
+    file_input_adapter(const file_input_adapter&) = delete;
+    file_input_adapter(file_input_adapter&&) noexcept = default;
+    file_input_adapter& operator=(const file_input_adapter&) = delete;
+    file_input_adapter& operator=(file_input_adapter&&) = delete;
+    ~file_input_adapter() = default;
+
+    std::char_traits<char>::int_type get_character() noexcept
+    {
+        return std::fgetc(m_file);
+    }
+
+    // returns the number of characters successfully read
+    template<class T>
+    std::size_t get_elements(T* dest, std::size_t count = 1)
+    {
+        return fread(dest, 1, sizeof(T) * count, m_file);
+    }
+
+  private:
+    /// the file pointer to read from
+    std::FILE* m_file;
+};
+
+/*!
+Input adapter for a (caching) istream. Ignores a UFT Byte Order Mark at
+beginning of input. Does not support changing the underlying std::streambuf
+in mid-input. Maintains underlying std::istream and std::streambuf to support
+subsequent use of standard std::istream operations to process any input
+characters following those used in parsing the JSON input.  Clears the
+std::istream flags; any input errors (e.g., EOF) will be detected by the first
+subsequent call for input from the std::istream.
+*/
+class input_stream_adapter
+{
+  public:
+    using char_type = char;
+
+    ~input_stream_adapter()
+    {
+        // clear stream flags; we use underlying streambuf I/O, do not
+        // maintain ifstream flags, except eof
+        if (is != nullptr)
+        {
+            is->clear(is->rdstate() & std::ios::eofbit);
+        }
+    }
+
+    explicit input_stream_adapter(std::istream& i)
+        : is(&i), sb(i.rdbuf())
+    {}
+
+    // delete because of pointer members
+    input_stream_adapter(const input_stream_adapter&) = delete;
+    input_stream_adapter& operator=(input_stream_adapter&) = delete;
+    input_stream_adapter& operator=(input_stream_adapter&&) = delete;
+
+    input_stream_adapter(input_stream_adapter&& rhs) noexcept
+        : is(rhs.is), sb(rhs.sb)
+    {
+        rhs.is = nullptr;
+        rhs.sb = nullptr;
+    }
+
+    // std::istream/std::streambuf use std::char_traits<char>::to_int_type, to
+    // ensure that std::char_traits<char>::eof() and the character 0xFF do not
+    // end up as the same value, e.g. 0xFFFFFFFF.
+    std::char_traits<char>::int_type get_character()
+    {
+        auto res = sb->sbumpc();
+        // set eof manually, as we don't use the istream interface.
+        if (JSON_HEDLEY_UNLIKELY(res == std::char_traits<char>::eof()))
+        {
+            is->clear(is->rdstate() | std::ios::eofbit);
+        }
+        return res;
+    }
+
+    template<class T>
+    std::size_t get_elements(T* dest, std::size_t count = 1)
+    {
+        auto res = static_cast<std::size_t>(sb->sgetn(reinterpret_cast<char*>(dest), static_cast<std::streamsize>(count * sizeof(T))));
+        if (JSON_HEDLEY_UNLIKELY(res < count * sizeof(T)))
+        {
+            is->clear(is->rdstate() | std::ios::eofbit);
+        }
+        return res;
+    }
+
+  private:
+    /// the associated input stream
+    std::istream* is = nullptr;
+    std::streambuf* sb = nullptr;
+};
+#endif  // JSON_NO_IO
+
+// General-purpose iterator-based adapter. It might not be as fast as
+// theoretically possible for some containers, but it is extremely versatile.
+template<typename IteratorType>
+class iterator_input_adapter
+{
+  public:
+    using char_type = typename std::iterator_traits<IteratorType>::value_type;
+
+    iterator_input_adapter(IteratorType first, IteratorType last)
+        : current(std::move(first)), end(std::move(last))
+    {}
+
+    typename char_traits<char_type>::int_type get_character()
+    {
+        if (JSON_HEDLEY_LIKELY(current != end))
+        {
+            auto result = char_traits<char_type>::to_int_type(*current);
+            std::advance(current, 1);
+            return result;
+        }
+
+        return char_traits<char_type>::eof();
+    }
+
+    // for general iterators, we cannot really do something better than falling back to processing the range one-by-one
+    template<class T>
+    std::size_t get_elements(T* dest, std::size_t count = 1)
+    {
+        auto* ptr = reinterpret_cast<char*>(dest);
+        for (std::size_t read_index = 0; read_index < count * sizeof(T); ++read_index)
+        {
+            if (JSON_HEDLEY_LIKELY(current != end))
+            {
+                ptr[read_index] = static_cast<char>(*current);
+                std::advance(current, 1);
+            }
+            else
+            {
+                return read_index;
+            }
+        }
+        return count * sizeof(T);
+    }
+
+  private:
+    IteratorType current;
+    IteratorType end;
+
+    template<typename BaseInputAdapter, size_t T>
+    friend struct wide_string_input_helper;
+
+    bool empty() const
+    {
+        return current == end;
+    }
+};
+
+template<typename BaseInputAdapter, size_t T>
+struct wide_string_input_helper;
+
+template<typename BaseInputAdapter>
+struct wide_string_input_helper<BaseInputAdapter, 4>
+{
+    // UTF-32
+    static void fill_buffer(BaseInputAdapter& input,
+                            std::array<std::char_traits<char>::int_type, 4>& utf8_bytes,
+                            size_t& utf8_bytes_index,
+                            size_t& utf8_bytes_filled)
+    {
+        utf8_bytes_index = 0;
+
+        if (JSON_HEDLEY_UNLIKELY(input.empty()))
+        {
+            utf8_bytes[0] = std::char_traits<char>::eof();
+            utf8_bytes_filled = 1;
+        }
+        else
+        {
+            // get the current character
+            const auto wc = input.get_character();
+
+            // UTF-32 to UTF-8 encoding
+            if (wc < 0x80)
+            {
+                utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(wc);
+                utf8_bytes_filled = 1;
+            }
+            else if (wc <= 0x7FF)
+            {
+                utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(0xC0u | ((static_cast<unsigned int>(wc) >> 6u) & 0x1Fu));
+                utf8_bytes[1] = static_cast<std::char_traits<char>::int_type>(0x80u | (static_cast<unsigned int>(wc) & 0x3Fu));
+                utf8_bytes_filled = 2;
+            }
+            else if (wc <= 0xFFFF)
+            {
+                utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(0xE0u | ((static_cast<unsigned int>(wc) >> 12u) & 0x0Fu));
+                utf8_bytes[1] = static_cast<std::char_traits<char>::int_type>(0x80u | ((static_cast<unsigned int>(wc) >> 6u) & 0x3Fu));
+                utf8_bytes[2] = static_cast<std::char_traits<char>::int_type>(0x80u | (static_cast<unsigned int>(wc) & 0x3Fu));
+                utf8_bytes_filled = 3;
+            }
+            else if (wc <= 0x10FFFF)
+            {
+                utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(0xF0u | ((static_cast<unsigned int>(wc) >> 18u) & 0x07u));
+                utf8_bytes[1] = static_cast<std::char_traits<char>::int_type>(0x80u | ((static_cast<unsigned int>(wc) >> 12u) & 0x3Fu));
+                utf8_bytes[2] = static_cast<std::char_traits<char>::int_type>(0x80u | ((static_cast<unsigned int>(wc) >> 6u) & 0x3Fu));
+                utf8_bytes[3] = static_cast<std::char_traits<char>::int_type>(0x80u | (static_cast<unsigned int>(wc) & 0x3Fu));
+                utf8_bytes_filled = 4;
+            }
+            else
+            {
+                // unknown character
+                utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(wc);
+                utf8_bytes_filled = 1;
+            }
+        }
+    }
+};
+
+template<typename BaseInputAdapter>
+struct wide_string_input_helper<BaseInputAdapter, 2>
+{
+    // UTF-16
+    static void fill_buffer(BaseInputAdapter& input,
+                            std::array<std::char_traits<char>::int_type, 4>& utf8_bytes,
+                            size_t& utf8_bytes_index,
+                            size_t& utf8_bytes_filled)
+    {
+        utf8_bytes_index = 0;
+
+        if (JSON_HEDLEY_UNLIKELY(input.empty()))
+        {
+            utf8_bytes[0] = std::char_traits<char>::eof();
+            utf8_bytes_filled = 1;
+        }
+        else
+        {
+            // get the current character
+            const auto wc = input.get_character();
+
+            // UTF-16 to UTF-8 encoding
+            if (wc < 0x80)
+            {
+                utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(wc);
+                utf8_bytes_filled = 1;
+            }
+            else if (wc <= 0x7FF)
+            {
+                utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(0xC0u | ((static_cast<unsigned int>(wc) >> 6u)));
+                utf8_bytes[1] = static_cast<std::char_traits<char>::int_type>(0x80u | (static_cast<unsigned int>(wc) & 0x3Fu));
+                utf8_bytes_filled = 2;
+            }
+            else if (0xD800 > wc || wc >= 0xE000)
+            {
+                utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(0xE0u | ((static_cast<unsigned int>(wc) >> 12u)));
+                utf8_bytes[1] = static_cast<std::char_traits<char>::int_type>(0x80u | ((static_cast<unsigned int>(wc) >> 6u) & 0x3Fu));
+                utf8_bytes[2] = static_cast<std::char_traits<char>::int_type>(0x80u | (static_cast<unsigned int>(wc) & 0x3Fu));
+                utf8_bytes_filled = 3;
+            }
+            else
+            {
+                if (JSON_HEDLEY_UNLIKELY(!input.empty()))
+                {
+                    const auto wc2 = static_cast<unsigned int>(input.get_character());
+                    const auto charcode = 0x10000u + (((static_cast<unsigned int>(wc) & 0x3FFu) << 10u) | (wc2 & 0x3FFu));
+                    utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(0xF0u | (charcode >> 18u));
+                    utf8_bytes[1] = static_cast<std::char_traits<char>::int_type>(0x80u | ((charcode >> 12u) & 0x3Fu));
+                    utf8_bytes[2] = static_cast<std::char_traits<char>::int_type>(0x80u | ((charcode >> 6u) & 0x3Fu));
+                    utf8_bytes[3] = static_cast<std::char_traits<char>::int_type>(0x80u | (charcode & 0x3Fu));
+                    utf8_bytes_filled = 4;
+                }
+                else
+                {
+                    utf8_bytes[0] = static_cast<std::char_traits<char>::int_type>(wc);
+                    utf8_bytes_filled = 1;
+                }
+            }
+        }
+    }
+};
+
+// Wraps another input adapter to convert wide character types into individual bytes.
+template<typename BaseInputAdapter, typename WideCharType>
+class wide_string_input_adapter
+{
+  public:
+    using char_type = char;
+
+    wide_string_input_adapter(BaseInputAdapter base)
+        : base_adapter(base) {}
+
+    typename std::char_traits<char>::int_type get_character() noexcept
+    {
+        // check if buffer needs to be filled
+        if (utf8_bytes_index == utf8_bytes_filled)
+        {
+            fill_buffer<sizeof(WideCharType)>();
+
+            JSON_ASSERT(utf8_bytes_filled > 0);
+            JSON_ASSERT(utf8_bytes_index == 0);
+        }
+
+        // use buffer
+        JSON_ASSERT(utf8_bytes_filled > 0);
+        JSON_ASSERT(utf8_bytes_index < utf8_bytes_filled);
+        return utf8_bytes[utf8_bytes_index++];
+    }
+
+    // parsing binary with wchar doesn't make sense, but since the parsing mode can be runtime, we need something here
+    template<class T>
+    std::size_t get_elements(T* /*dest*/, std::size_t /*count*/ = 1)
+    {
+        JSON_THROW(parse_error::create(112, 1, "wide string type cannot be interpreted as binary data", nullptr));
+    }
+
+  private:
+    BaseInputAdapter base_adapter;
+
+    template<size_t T>
+    void fill_buffer()
+    {
+        wide_string_input_helper<BaseInputAdapter, T>::fill_buffer(base_adapter, utf8_bytes, utf8_bytes_index, utf8_bytes_filled);
+    }
+
+    /// a buffer for UTF-8 bytes
+    std::array<std::char_traits<char>::int_type, 4> utf8_bytes = {{0, 0, 0, 0}};
+
+    /// index to the utf8_codes array for the next valid byte
+    std::size_t utf8_bytes_index = 0;
+    /// number of valid bytes in the utf8_codes array
+    std::size_t utf8_bytes_filled = 0;
+};
+
+template<typename IteratorType, typename Enable = void>
+struct iterator_input_adapter_factory
+{
+    using iterator_type = IteratorType;
+    using char_type = typename std::iterator_traits<iterator_type>::value_type;
+    using adapter_type = iterator_input_adapter<iterator_type>;
+
+    static adapter_type create(IteratorType first, IteratorType last)
+    {
+        return adapter_type(std::move(first), std::move(last));
+    }
+};
+
+template<typename T>
+struct is_iterator_of_multibyte
+{
+    using value_type = typename std::iterator_traits<T>::value_type;
+    enum
+    {
+        value = sizeof(value_type) > 1
+    };
+};
+
+template<typename IteratorType>
+struct iterator_input_adapter_factory<IteratorType, enable_if_t<is_iterator_of_multibyte<IteratorType>::value>>
+{
+    using iterator_type = IteratorType;
+    using char_type = typename std::iterator_traits<iterator_type>::value_type;
+    using base_adapter_type = iterator_input_adapter<iterator_type>;
+    using adapter_type = wide_string_input_adapter<base_adapter_type, char_type>;
+
+    static adapter_type create(IteratorType first, IteratorType last)
+    {
+        return adapter_type(base_adapter_type(std::move(first), std::move(last)));
+    }
+};
+
+// General purpose iterator-based input
+template<typename IteratorType>
+typename iterator_input_adapter_factory<IteratorType>::adapter_type input_adapter(IteratorType first, IteratorType last)
+{
+    using factory_type = iterator_input_adapter_factory<IteratorType>;
+    return factory_type::create(first, last);
+}
+
+// Convenience shorthand from container to iterator
+// Enables ADL on begin(container) and end(container)
+// Encloses the using declarations in namespace for not to leak them to outside scope
+
+namespace container_input_adapter_factory_impl
+{
+
+using std::begin;
+using std::end;
+
+template<typename ContainerType, typename Enable = void>
+struct container_input_adapter_factory {};
+
+template<typename ContainerType>
+struct container_input_adapter_factory< ContainerType,
+       void_t<decltype(begin(std::declval<ContainerType>()), end(std::declval<ContainerType>()))>>
+       {
+           using adapter_type = decltype(input_adapter(begin(std::declval<ContainerType>()), end(std::declval<ContainerType>())));
+
+           static adapter_type create(const ContainerType& container)
+{
+    return input_adapter(begin(container), end(container));
+}
+       };
+
+}  // namespace container_input_adapter_factory_impl
+
+template<typename ContainerType>
+typename container_input_adapter_factory_impl::container_input_adapter_factory<ContainerType>::adapter_type input_adapter(const ContainerType& container)
+{
+    return container_input_adapter_factory_impl::container_input_adapter_factory<ContainerType>::create(container);
+}
+
+// specialization for std::string
+using string_input_adapter_type = decltype(input_adapter(std::declval<std::string>()));
+
+#ifndef JSON_NO_IO
+// Special cases with fast paths
+inline file_input_adapter input_adapter(std::FILE* file)
+{
+    if (file == nullptr)
+    {
+        JSON_THROW(parse_error::create(101, 0, "attempting to parse an empty input; check that your input string or stream contains the expected JSON", nullptr));
+    }
+    return file_input_adapter(file);
+}
+
+inline input_stream_adapter input_adapter(std::istream& stream)
+{
+    return input_stream_adapter(stream);
+}
+
+inline input_stream_adapter input_adapter(std::istream&& stream)
+{
+    return input_stream_adapter(stream);
+}
+#endif  // JSON_NO_IO
+
+using contiguous_bytes_input_adapter = decltype(input_adapter(std::declval<const char*>(), std::declval<const char*>()));
+
+// Null-delimited strings, and the like.
+template < typename CharT,
+           typename std::enable_if <
+               std::is_pointer<CharT>::value&&
+               !std::is_array<CharT>::value&&
+               std::is_integral<typename std::remove_pointer<CharT>::type>::value&&
+               sizeof(typename std::remove_pointer<CharT>::type) == 1,
+               int >::type = 0 >
+contiguous_bytes_input_adapter input_adapter(CharT b)
+{
+    if (b == nullptr)
+    {
+        JSON_THROW(parse_error::create(101, 0, "attempting to parse an empty input; check that your input string or stream contains the expected JSON", nullptr));
+    }
+    auto length = std::strlen(reinterpret_cast<const char*>(b));
+    const auto* ptr = reinterpret_cast<const char*>(b);
+    return input_adapter(ptr, ptr + length); // cppcheck-suppress[nullPointerArithmeticRedundantCheck]
+}
+
+template<typename T, std::size_t N>
+auto input_adapter(T (&array)[N]) -> decltype(input_adapter(array, array + N)) // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+{
+    return input_adapter(array, array + N);
+}
+
+// This class only handles inputs of input_buffer_adapter type.
+// It's required so that expressions like {ptr, len} can be implicitly cast
+// to the correct adapter.
+class span_input_adapter
+{
+  public:
+    template < typename CharT,
+               typename std::enable_if <
+                   std::is_pointer<CharT>::value&&
+                   std::is_integral<typename std::remove_pointer<CharT>::type>::value&&
+                   sizeof(typename std::remove_pointer<CharT>::type) == 1,
+                   int >::type = 0 >
+    span_input_adapter(CharT b, std::size_t l)
+        : ia(reinterpret_cast<const char*>(b), reinterpret_cast<const char*>(b) + l) {}
+
+    template<class IteratorType,
+             typename std::enable_if<
+                 std::is_same<typename iterator_traits<IteratorType>::iterator_category, std::random_access_iterator_tag>::value,
+                 int>::type = 0>
+    span_input_adapter(IteratorType first, IteratorType last)
+        : ia(input_adapter(first, last)) {}
+
+    contiguous_bytes_input_adapter&& get()
+    {
+        return std::move(ia); // NOLINT(hicpp-move-const-arg,performance-move-const-arg)
+    }
+
+  private:
+    contiguous_bytes_input_adapter ia;
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/input/json_sax.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstddef>
+#include <string> // string
+#include <type_traits> // enable_if_t
+#include <utility> // move
+#include <vector> // vector
+
+// #include <nlohmann/detail/exceptions.hpp>
+
+// #include <nlohmann/detail/input/lexer.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <array> // array
+#include <clocale> // localeconv
+#include <cstddef> // size_t
+#include <cstdio> // snprintf
+#include <cstdlib> // strtof, strtod, strtold, strtoll, strtoull
+#include <initializer_list> // initializer_list
+#include <string> // char_traits, string
+#include <utility> // move
+#include <vector> // vector
+
+// #include <nlohmann/detail/input/input_adapters.hpp>
+
+// #include <nlohmann/detail/input/position_t.hpp>
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+///////////
+// lexer //
+///////////
+
+template<typename BasicJsonType>
+class lexer_base
+{
+  public:
+    /// token types for the parser
+    enum class token_type
+    {
+        uninitialized,    ///< indicating the scanner is uninitialized
+        literal_true,     ///< the `true` literal
+        literal_false,    ///< the `false` literal
+        literal_null,     ///< the `null` literal
+        value_string,     ///< a string -- use get_string() for actual value
+        value_unsigned,   ///< an unsigned integer -- use get_number_unsigned() for actual value
+        value_integer,    ///< a signed integer -- use get_number_integer() for actual value
+        value_float,      ///< an floating point number -- use get_number_float() for actual value
+        begin_array,      ///< the character for array begin `[`
+        begin_object,     ///< the character for object begin `{`
+        end_array,        ///< the character for array end `]`
+        end_object,       ///< the character for object end `}`
+        name_separator,   ///< the name separator `:`
+        value_separator,  ///< the value separator `,`
+        parse_error,      ///< indicating a parse error
+        end_of_input,     ///< indicating the end of the input buffer
+        literal_or_value  ///< a literal or the begin of a value (only for diagnostics)
+    };
+
+    /// return name of values of type token_type (only used for errors)
+    JSON_HEDLEY_RETURNS_NON_NULL
+    JSON_HEDLEY_CONST
+    static const char* token_type_name(const token_type t) noexcept
+    {
+        switch (t)
+        {
+            case token_type::uninitialized:
+                return "<uninitialized>";
+            case token_type::literal_true:
+                return "true literal";
+            case token_type::literal_false:
+                return "false literal";
+            case token_type::literal_null:
+                return "null literal";
+            case token_type::value_string:
+                return "string literal";
+            case token_type::value_unsigned:
+            case token_type::value_integer:
+            case token_type::value_float:
+                return "number literal";
+            case token_type::begin_array:
+                return "'['";
+            case token_type::begin_object:
+                return "'{'";
+            case token_type::end_array:
+                return "']'";
+            case token_type::end_object:
+                return "'}'";
+            case token_type::name_separator:
+                return "':'";
+            case token_type::value_separator:
+                return "','";
+            case token_type::parse_error:
+                return "<parse error>";
+            case token_type::end_of_input:
+                return "end of input";
+            case token_type::literal_or_value:
+                return "'[', '{', or a literal";
+            // LCOV_EXCL_START
+            default: // catch non-enum values
+                return "unknown token";
+                // LCOV_EXCL_STOP
+        }
+    }
+};
+/*!
+@brief lexical analysis
+
+This class organizes the lexical analysis during JSON deserialization.
+*/
+template<typename BasicJsonType, typename InputAdapterType>
+class lexer : public lexer_base<BasicJsonType>
+{
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using char_type = typename InputAdapterType::char_type;
+    using char_int_type = typename char_traits<char_type>::int_type;
+
+  public:
+    using token_type = typename lexer_base<BasicJsonType>::token_type;
+
+    explicit lexer(InputAdapterType&& adapter, bool ignore_comments_ = false) noexcept
+        : ia(std::move(adapter))
+        , ignore_comments(ignore_comments_)
+        , decimal_point_char(static_cast<char_int_type>(get_decimal_point()))
+    {}
+
+    // delete because of pointer members
+    lexer(const lexer&) = delete;
+    lexer(lexer&&) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor)
+    lexer& operator=(lexer&) = delete;
+    lexer& operator=(lexer&&) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor)
+    ~lexer() = default;
+
+  private:
+    /////////////////////
+    // locales
+    /////////////////////
+
+    /// return the locale-dependent decimal point
+    JSON_HEDLEY_PURE
+    static char get_decimal_point() noexcept
+    {
+        const auto* loc = localeconv();
+        JSON_ASSERT(loc != nullptr);
+        return (loc->decimal_point == nullptr) ? '.' : *(loc->decimal_point);
+    }
+
+    /////////////////////
+    // scan functions
+    /////////////////////
+
+    /*!
+    @brief get codepoint from 4 hex characters following `\u`
+
+    For input "\u c1 c2 c3 c4" the codepoint is:
+      (c1 * 0x1000) + (c2 * 0x0100) + (c3 * 0x0010) + c4
+    = (c1 << 12) + (c2 << 8) + (c3 << 4) + (c4 << 0)
+
+    Furthermore, the possible characters '0'..'9', 'A'..'F', and 'a'..'f'
+    must be converted to the integers 0x0..0x9, 0xA..0xF, 0xA..0xF, resp. The
+    conversion is done by subtracting the offset (0x30, 0x37, and 0x57)
+    between the ASCII value of the character and the desired integer value.
+
+    @return codepoint (0x0000..0xFFFF) or -1 in case of an error (e.g. EOF or
+            non-hex character)
+    */
+    int get_codepoint()
+    {
+        // this function only makes sense after reading `\u`
+        JSON_ASSERT(current == 'u');
+        int codepoint = 0;
+
+        const auto factors = { 12u, 8u, 4u, 0u };
+        for (const auto factor : factors)
+        {
+            get();
+
+            if (current >= '0' && current <= '9')
+            {
+                codepoint += static_cast<int>((static_cast<unsigned int>(current) - 0x30u) << factor);
+            }
+            else if (current >= 'A' && current <= 'F')
+            {
+                codepoint += static_cast<int>((static_cast<unsigned int>(current) - 0x37u) << factor);
+            }
+            else if (current >= 'a' && current <= 'f')
+            {
+                codepoint += static_cast<int>((static_cast<unsigned int>(current) - 0x57u) << factor);
+            }
+            else
+            {
+                return -1;
+            }
+        }
+
+        JSON_ASSERT(0x0000 <= codepoint && codepoint <= 0xFFFF);
+        return codepoint;
+    }
+
+    /*!
+    @brief check if the next byte(s) are inside a given range
+
+    Adds the current byte and, for each passed range, reads a new byte and
+    checks if it is inside the range. If a violation was detected, set up an
+    error message and return false. Otherwise, return true.
+
+    @param[in] ranges  list of integers; interpreted as list of pairs of
+                       inclusive lower and upper bound, respectively
+
+    @pre The passed list @a ranges must have 2, 4, or 6 elements; that is,
+         1, 2, or 3 pairs. This precondition is enforced by an assertion.
+
+    @return true if and only if no range violation was detected
+    */
+    bool next_byte_in_range(std::initializer_list<char_int_type> ranges)
+    {
+        JSON_ASSERT(ranges.size() == 2 || ranges.size() == 4 || ranges.size() == 6);
+        add(current);
+
+        for (auto range = ranges.begin(); range != ranges.end(); ++range)
+        {
+            get();
+            if (JSON_HEDLEY_LIKELY(*range <= current && current <= *(++range))) // NOLINT(bugprone-inc-dec-in-conditions)
+            {
+                add(current);
+            }
+            else
+            {
+                error_message = "invalid string: ill-formed UTF-8 byte";
+                return false;
+            }
+        }
+
+        return true;
+    }
+
+    /*!
+    @brief scan a string literal
+
+    This function scans a string according to Sect. 7 of RFC 8259. While
+    scanning, bytes are escaped and copied into buffer token_buffer. Then the
+    function returns successfully, token_buffer is *not* null-terminated (as it
+    may contain \0 bytes), and token_buffer.size() is the number of bytes in the
+    string.
+
+    @return token_type::value_string if string could be successfully scanned,
+            token_type::parse_error otherwise
+
+    @note In case of errors, variable error_message contains a textual
+          description.
+    */
+    token_type scan_string()
+    {
+        // reset token_buffer (ignore opening quote)
+        reset();
+
+        // we entered the function by reading an open quote
+        JSON_ASSERT(current == '\"');
+
+        while (true)
+        {
+            // get next character
+            switch (get())
+            {
+                // end of file while parsing string
+                case char_traits<char_type>::eof():
+                {
+                    error_message = "invalid string: missing closing quote";
+                    return token_type::parse_error;
+                }
+
+                // closing quote
+                case '\"':
+                {
+                    return token_type::value_string;
+                }
+
+                // escapes
+                case '\\':
+                {
+                    switch (get())
+                    {
+                        // quotation mark
+                        case '\"':
+                            add('\"');
+                            break;
+                        // reverse solidus
+                        case '\\':
+                            add('\\');
+                            break;
+                        // solidus
+                        case '/':
+                            add('/');
+                            break;
+                        // backspace
+                        case 'b':
+                            add('\b');
+                            break;
+                        // form feed
+                        case 'f':
+                            add('\f');
+                            break;
+                        // line feed
+                        case 'n':
+                            add('\n');
+                            break;
+                        // carriage return
+                        case 'r':
+                            add('\r');
+                            break;
+                        // tab
+                        case 't':
+                            add('\t');
+                            break;
+
+                        // unicode escapes
+                        case 'u':
+                        {
+                            const int codepoint1 = get_codepoint();
+                            int codepoint = codepoint1; // start with codepoint1
+
+                            if (JSON_HEDLEY_UNLIKELY(codepoint1 == -1))
+                            {
+                                error_message = "invalid string: '\\u' must be followed by 4 hex digits";
+                                return token_type::parse_error;
+                            }
+
+                            // check if code point is a high surrogate
+                            if (0xD800 <= codepoint1 && codepoint1 <= 0xDBFF)
+                            {
+                                // expect next \uxxxx entry
+                                if (JSON_HEDLEY_LIKELY(get() == '\\' && get() == 'u'))
+                                {
+                                    const int codepoint2 = get_codepoint();
+
+                                    if (JSON_HEDLEY_UNLIKELY(codepoint2 == -1))
+                                    {
+                                        error_message = "invalid string: '\\u' must be followed by 4 hex digits";
+                                        return token_type::parse_error;
+                                    }
+
+                                    // check if codepoint2 is a low surrogate
+                                    if (JSON_HEDLEY_LIKELY(0xDC00 <= codepoint2 && codepoint2 <= 0xDFFF))
+                                    {
+                                        // overwrite codepoint
+                                        codepoint = static_cast<int>(
+                                                        // high surrogate occupies the most significant 22 bits
+                                                        (static_cast<unsigned int>(codepoint1) << 10u)
+                                                        // low surrogate occupies the least significant 15 bits
+                                                        + static_cast<unsigned int>(codepoint2)
+                                                        // there is still the 0xD800, 0xDC00 and 0x10000 noise
+                                                        // in the result, so we have to subtract with:
+                                                        // (0xD800 << 10) + DC00 - 0x10000 = 0x35FDC00
+                                                        - 0x35FDC00u);
+                                    }
+                                    else
+                                    {
+                                        error_message = "invalid string: surrogate U+D800..U+DBFF must be followed by U+DC00..U+DFFF";
+                                        return token_type::parse_error;
+                                    }
+                                }
+                                else
+                                {
+                                    error_message = "invalid string: surrogate U+D800..U+DBFF must be followed by U+DC00..U+DFFF";
+                                    return token_type::parse_error;
+                                }
+                            }
+                            else
+                            {
+                                if (JSON_HEDLEY_UNLIKELY(0xDC00 <= codepoint1 && codepoint1 <= 0xDFFF))
+                                {
+                                    error_message = "invalid string: surrogate U+DC00..U+DFFF must follow U+D800..U+DBFF";
+                                    return token_type::parse_error;
+                                }
+                            }
+
+                            // result of the above calculation yields a proper codepoint
+                            JSON_ASSERT(0x00 <= codepoint && codepoint <= 0x10FFFF);
+
+                            // translate codepoint into bytes
+                            if (codepoint < 0x80)
+                            {
+                                // 1-byte characters: 0xxxxxxx (ASCII)
+                                add(static_cast<char_int_type>(codepoint));
+                            }
+                            else if (codepoint <= 0x7FF)
+                            {
+                                // 2-byte characters: 110xxxxx 10xxxxxx
+                                add(static_cast<char_int_type>(0xC0u | (static_cast<unsigned int>(codepoint) >> 6u)));
+                                add(static_cast<char_int_type>(0x80u | (static_cast<unsigned int>(codepoint) & 0x3Fu)));
+                            }
+                            else if (codepoint <= 0xFFFF)
+                            {
+                                // 3-byte characters: 1110xxxx 10xxxxxx 10xxxxxx
+                                add(static_cast<char_int_type>(0xE0u | (static_cast<unsigned int>(codepoint) >> 12u)));
+                                add(static_cast<char_int_type>(0x80u | ((static_cast<unsigned int>(codepoint) >> 6u) & 0x3Fu)));
+                                add(static_cast<char_int_type>(0x80u | (static_cast<unsigned int>(codepoint) & 0x3Fu)));
+                            }
+                            else
+                            {
+                                // 4-byte characters: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
+                                add(static_cast<char_int_type>(0xF0u | (static_cast<unsigned int>(codepoint) >> 18u)));
+                                add(static_cast<char_int_type>(0x80u | ((static_cast<unsigned int>(codepoint) >> 12u) & 0x3Fu)));
+                                add(static_cast<char_int_type>(0x80u | ((static_cast<unsigned int>(codepoint) >> 6u) & 0x3Fu)));
+                                add(static_cast<char_int_type>(0x80u | (static_cast<unsigned int>(codepoint) & 0x3Fu)));
+                            }
+
+                            break;
+                        }
+
+                        // other characters after escape
+                        default:
+                            error_message = "invalid string: forbidden character after backslash";
+                            return token_type::parse_error;
+                    }
+
+                    break;
+                }
+
+                // invalid control characters
+                case 0x00:
+                {
+                    error_message = "invalid string: control character U+0000 (NUL) must be escaped to \\u0000";
+                    return token_type::parse_error;
+                }
+
+                case 0x01:
+                {
+                    error_message = "invalid string: control character U+0001 (SOH) must be escaped to \\u0001";
+                    return token_type::parse_error;
+                }
+
+                case 0x02:
+                {
+                    error_message = "invalid string: control character U+0002 (STX) must be escaped to \\u0002";
+                    return token_type::parse_error;
+                }
+
+                case 0x03:
+                {
+                    error_message = "invalid string: control character U+0003 (ETX) must be escaped to \\u0003";
+                    return token_type::parse_error;
+                }
+
+                case 0x04:
+                {
+                    error_message = "invalid string: control character U+0004 (EOT) must be escaped to \\u0004";
+                    return token_type::parse_error;
+                }
+
+                case 0x05:
+                {
+                    error_message = "invalid string: control character U+0005 (ENQ) must be escaped to \\u0005";
+                    return token_type::parse_error;
+                }
+
+                case 0x06:
+                {
+                    error_message = "invalid string: control character U+0006 (ACK) must be escaped to \\u0006";
+                    return token_type::parse_error;
+                }
+
+                case 0x07:
+                {
+                    error_message = "invalid string: control character U+0007 (BEL) must be escaped to \\u0007";
+                    return token_type::parse_error;
+                }
+
+                case 0x08:
+                {
+                    error_message = "invalid string: control character U+0008 (BS) must be escaped to \\u0008 or \\b";
+                    return token_type::parse_error;
+                }
+
+                case 0x09:
+                {
+                    error_message = "invalid string: control character U+0009 (HT) must be escaped to \\u0009 or \\t";
+                    return token_type::parse_error;
+                }
+
+                case 0x0A:
+                {
+                    error_message = "invalid string: control character U+000A (LF) must be escaped to \\u000A or \\n";
+                    return token_type::parse_error;
+                }
+
+                case 0x0B:
+                {
+                    error_message = "invalid string: control character U+000B (VT) must be escaped to \\u000B";
+                    return token_type::parse_error;
+                }
+
+                case 0x0C:
+                {
+                    error_message = "invalid string: control character U+000C (FF) must be escaped to \\u000C or \\f";
+                    return token_type::parse_error;
+                }
+
+                case 0x0D:
+                {
+                    error_message = "invalid string: control character U+000D (CR) must be escaped to \\u000D or \\r";
+                    return token_type::parse_error;
+                }
+
+                case 0x0E:
+                {
+                    error_message = "invalid string: control character U+000E (SO) must be escaped to \\u000E";
+                    return token_type::parse_error;
+                }
+
+                case 0x0F:
+                {
+                    error_message = "invalid string: control character U+000F (SI) must be escaped to \\u000F";
+                    return token_type::parse_error;
+                }
+
+                case 0x10:
+                {
+                    error_message = "invalid string: control character U+0010 (DLE) must be escaped to \\u0010";
+                    return token_type::parse_error;
+                }
+
+                case 0x11:
+                {
+                    error_message = "invalid string: control character U+0011 (DC1) must be escaped to \\u0011";
+                    return token_type::parse_error;
+                }
+
+                case 0x12:
+                {
+                    error_message = "invalid string: control character U+0012 (DC2) must be escaped to \\u0012";
+                    return token_type::parse_error;
+                }
+
+                case 0x13:
+                {
+                    error_message = "invalid string: control character U+0013 (DC3) must be escaped to \\u0013";
+                    return token_type::parse_error;
+                }
+
+                case 0x14:
+                {
+                    error_message = "invalid string: control character U+0014 (DC4) must be escaped to \\u0014";
+                    return token_type::parse_error;
+                }
+
+                case 0x15:
+                {
+                    error_message = "invalid string: control character U+0015 (NAK) must be escaped to \\u0015";
+                    return token_type::parse_error;
+                }
+
+                case 0x16:
+                {
+                    error_message = "invalid string: control character U+0016 (SYN) must be escaped to \\u0016";
+                    return token_type::parse_error;
+                }
+
+                case 0x17:
+                {
+                    error_message = "invalid string: control character U+0017 (ETB) must be escaped to \\u0017";
+                    return token_type::parse_error;
+                }
+
+                case 0x18:
+                {
+                    error_message = "invalid string: control character U+0018 (CAN) must be escaped to \\u0018";
+                    return token_type::parse_error;
+                }
+
+                case 0x19:
+                {
+                    error_message = "invalid string: control character U+0019 (EM) must be escaped to \\u0019";
+                    return token_type::parse_error;
+                }
+
+                case 0x1A:
+                {
+                    error_message = "invalid string: control character U+001A (SUB) must be escaped to \\u001A";
+                    return token_type::parse_error;
+                }
+
+                case 0x1B:
+                {
+                    error_message = "invalid string: control character U+001B (ESC) must be escaped to \\u001B";
+                    return token_type::parse_error;
+                }
+
+                case 0x1C:
+                {
+                    error_message = "invalid string: control character U+001C (FS) must be escaped to \\u001C";
+                    return token_type::parse_error;
+                }
+
+                case 0x1D:
+                {
+                    error_message = "invalid string: control character U+001D (GS) must be escaped to \\u001D";
+                    return token_type::parse_error;
+                }
+
+                case 0x1E:
+                {
+                    error_message = "invalid string: control character U+001E (RS) must be escaped to \\u001E";
+                    return token_type::parse_error;
+                }
+
+                case 0x1F:
+                {
+                    error_message = "invalid string: control character U+001F (US) must be escaped to \\u001F";
+                    return token_type::parse_error;
+                }
+
+                // U+0020..U+007F (except U+0022 (quote) and U+005C (backspace))
+                case 0x20:
+                case 0x21:
+                case 0x23:
+                case 0x24:
+                case 0x25:
+                case 0x26:
+                case 0x27:
+                case 0x28:
+                case 0x29:
+                case 0x2A:
+                case 0x2B:
+                case 0x2C:
+                case 0x2D:
+                case 0x2E:
+                case 0x2F:
+                case 0x30:
+                case 0x31:
+                case 0x32:
+                case 0x33:
+                case 0x34:
+                case 0x35:
+                case 0x36:
+                case 0x37:
+                case 0x38:
+                case 0x39:
+                case 0x3A:
+                case 0x3B:
+                case 0x3C:
+                case 0x3D:
+                case 0x3E:
+                case 0x3F:
+                case 0x40:
+                case 0x41:
+                case 0x42:
+                case 0x43:
+                case 0x44:
+                case 0x45:
+                case 0x46:
+                case 0x47:
+                case 0x48:
+                case 0x49:
+                case 0x4A:
+                case 0x4B:
+                case 0x4C:
+                case 0x4D:
+                case 0x4E:
+                case 0x4F:
+                case 0x50:
+                case 0x51:
+                case 0x52:
+                case 0x53:
+                case 0x54:
+                case 0x55:
+                case 0x56:
+                case 0x57:
+                case 0x58:
+                case 0x59:
+                case 0x5A:
+                case 0x5B:
+                case 0x5D:
+                case 0x5E:
+                case 0x5F:
+                case 0x60:
+                case 0x61:
+                case 0x62:
+                case 0x63:
+                case 0x64:
+                case 0x65:
+                case 0x66:
+                case 0x67:
+                case 0x68:
+                case 0x69:
+                case 0x6A:
+                case 0x6B:
+                case 0x6C:
+                case 0x6D:
+                case 0x6E:
+                case 0x6F:
+                case 0x70:
+                case 0x71:
+                case 0x72:
+                case 0x73:
+                case 0x74:
+                case 0x75:
+                case 0x76:
+                case 0x77:
+                case 0x78:
+                case 0x79:
+                case 0x7A:
+                case 0x7B:
+                case 0x7C:
+                case 0x7D:
+                case 0x7E:
+                case 0x7F:
+                {
+                    add(current);
+                    break;
+                }
+
+                // U+0080..U+07FF: bytes C2..DF 80..BF
+                case 0xC2:
+                case 0xC3:
+                case 0xC4:
+                case 0xC5:
+                case 0xC6:
+                case 0xC7:
+                case 0xC8:
+                case 0xC9:
+                case 0xCA:
+                case 0xCB:
+                case 0xCC:
+                case 0xCD:
+                case 0xCE:
+                case 0xCF:
+                case 0xD0:
+                case 0xD1:
+                case 0xD2:
+                case 0xD3:
+                case 0xD4:
+                case 0xD5:
+                case 0xD6:
+                case 0xD7:
+                case 0xD8:
+                case 0xD9:
+                case 0xDA:
+                case 0xDB:
+                case 0xDC:
+                case 0xDD:
+                case 0xDE:
+                case 0xDF:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!next_byte_in_range({0x80, 0xBF})))
+                    {
+                        return token_type::parse_error;
+                    }
+                    break;
+                }
+
+                // U+0800..U+0FFF: bytes E0 A0..BF 80..BF
+                case 0xE0:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0xA0, 0xBF, 0x80, 0xBF}))))
+                    {
+                        return token_type::parse_error;
+                    }
+                    break;
+                }
+
+                // U+1000..U+CFFF: bytes E1..EC 80..BF 80..BF
+                // U+E000..U+FFFF: bytes EE..EF 80..BF 80..BF
+                case 0xE1:
+                case 0xE2:
+                case 0xE3:
+                case 0xE4:
+                case 0xE5:
+                case 0xE6:
+                case 0xE7:
+                case 0xE8:
+                case 0xE9:
+                case 0xEA:
+                case 0xEB:
+                case 0xEC:
+                case 0xEE:
+                case 0xEF:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0xBF, 0x80, 0xBF}))))
+                    {
+                        return token_type::parse_error;
+                    }
+                    break;
+                }
+
+                // U+D000..U+D7FF: bytes ED 80..9F 80..BF
+                case 0xED:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0x9F, 0x80, 0xBF}))))
+                    {
+                        return token_type::parse_error;
+                    }
+                    break;
+                }
+
+                // U+10000..U+3FFFF F0 90..BF 80..BF 80..BF
+                case 0xF0:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x90, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
+                    {
+                        return token_type::parse_error;
+                    }
+                    break;
+                }
+
+                // U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
+                case 0xF1:
+                case 0xF2:
+                case 0xF3:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0xBF, 0x80, 0xBF, 0x80, 0xBF}))))
+                    {
+                        return token_type::parse_error;
+                    }
+                    break;
+                }
+
+                // U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
+                case 0xF4:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!(next_byte_in_range({0x80, 0x8F, 0x80, 0xBF, 0x80, 0xBF}))))
+                    {
+                        return token_type::parse_error;
+                    }
+                    break;
+                }
+
+                // remaining bytes (80..C1 and F5..FF) are ill-formed
+                default:
+                {
+                    error_message = "invalid string: ill-formed UTF-8 byte";
+                    return token_type::parse_error;
+                }
+            }
+        }
+    }
+
+    /*!
+     * @brief scan a comment
+     * @return whether comment could be scanned successfully
+     */
+    bool scan_comment()
+    {
+        switch (get())
+        {
+            // single-line comments skip input until a newline or EOF is read
+            case '/':
+            {
+                while (true)
+                {
+                    switch (get())
+                    {
+                        case '\n':
+                        case '\r':
+                        case char_traits<char_type>::eof():
+                        case '\0':
+                            return true;
+
+                        default:
+                            break;
+                    }
+                }
+            }
+
+            // multi-line comments skip input until */ is read
+            case '*':
+            {
+                while (true)
+                {
+                    switch (get())
+                    {
+                        case char_traits<char_type>::eof():
+                        case '\0':
+                        {
+                            error_message = "invalid comment; missing closing '*/'";
+                            return false;
+                        }
+
+                        case '*':
+                        {
+                            switch (get())
+                            {
+                                case '/':
+                                    return true;
+
+                                default:
+                                {
+                                    unget();
+                                    continue;
+                                }
+                            }
+                        }
+
+                        default:
+                            continue;
+                    }
+                }
+            }
+
+            // unexpected character after reading '/'
+            default:
+            {
+                error_message = "invalid comment; expecting '/' or '*' after '/'";
+                return false;
+            }
+        }
+    }
+
+    JSON_HEDLEY_NON_NULL(2)
+    static void strtof(float& f, const char* str, char** endptr) noexcept
+    {
+        f = std::strtof(str, endptr);
+    }
+
+    JSON_HEDLEY_NON_NULL(2)
+    static void strtof(double& f, const char* str, char** endptr) noexcept
+    {
+        f = std::strtod(str, endptr);
+    }
+
+    JSON_HEDLEY_NON_NULL(2)
+    static void strtof(long double& f, const char* str, char** endptr) noexcept
+    {
+        f = std::strtold(str, endptr);
+    }
+
+    /*!
+    @brief scan a number literal
+
+    This function scans a string according to Sect. 6 of RFC 8259.
+
+    The function is realized with a deterministic finite state machine derived
+    from the grammar described in RFC 8259. Starting in state "init", the
+    input is read and used to determined the next state. Only state "done"
+    accepts the number. State "error" is a trap state to model errors. In the
+    table below, "anything" means any character but the ones listed before.
+
+    state    | 0        | 1-9      | e E      | +       | -       | .        | anything
+    ---------|----------|----------|----------|---------|---------|----------|-----------
+    init     | zero     | any1     | [error]  | [error] | minus   | [error]  | [error]
+    minus    | zero     | any1     | [error]  | [error] | [error] | [error]  | [error]
+    zero     | done     | done     | exponent | done    | done    | decimal1 | done
+    any1     | any1     | any1     | exponent | done    | done    | decimal1 | done
+    decimal1 | decimal2 | decimal2 | [error]  | [error] | [error] | [error]  | [error]
+    decimal2 | decimal2 | decimal2 | exponent | done    | done    | done     | done
+    exponent | any2     | any2     | [error]  | sign    | sign    | [error]  | [error]
+    sign     | any2     | any2     | [error]  | [error] | [error] | [error]  | [error]
+    any2     | any2     | any2     | done     | done    | done    | done     | done
+
+    The state machine is realized with one label per state (prefixed with
+    "scan_number_") and `goto` statements between them. The state machine
+    contains cycles, but any cycle can be left when EOF is read. Therefore,
+    the function is guaranteed to terminate.
+
+    During scanning, the read bytes are stored in token_buffer. This string is
+    then converted to a signed integer, an unsigned integer, or a
+    floating-point number.
+
+    @return token_type::value_unsigned, token_type::value_integer, or
+            token_type::value_float if number could be successfully scanned,
+            token_type::parse_error otherwise
+
+    @note The scanner is independent of the current locale. Internally, the
+          locale's decimal point is used instead of `.` to work with the
+          locale-dependent converters.
+    */
+    token_type scan_number()  // lgtm [cpp/use-of-goto] `goto` is used in this function to implement the number-parsing state machine described above. By design, any finite input will eventually reach the "done" state or return token_type::parse_error. In each intermediate state, 1 byte of the input is appended to the token_buffer vector, and only the already initialized variables token_buffer, number_type, and error_message are manipulated.
+    {
+        // reset token_buffer to store the number's bytes
+        reset();
+
+        // the type of the parsed number; initially set to unsigned; will be
+        // changed if minus sign, decimal point or exponent is read
+        token_type number_type = token_type::value_unsigned;
+
+        // state (init): we just found out we need to scan a number
+        switch (current)
+        {
+            case '-':
+            {
+                add(current);
+                goto scan_number_minus;
+            }
+
+            case '0':
+            {
+                add(current);
+                goto scan_number_zero;
+            }
+
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+            {
+                add(current);
+                goto scan_number_any1;
+            }
+
+            // all other characters are rejected outside scan_number()
+            default:            // LCOV_EXCL_LINE
+                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+        }
+
+scan_number_minus:
+        // state: we just parsed a leading minus sign
+        number_type = token_type::value_integer;
+        switch (get())
+        {
+            case '0':
+            {
+                add(current);
+                goto scan_number_zero;
+            }
+
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+            {
+                add(current);
+                goto scan_number_any1;
+            }
+
+            default:
+            {
+                error_message = "invalid number; expected digit after '-'";
+                return token_type::parse_error;
+            }
+        }
+
+scan_number_zero:
+        // state: we just parse a zero (maybe with a leading minus sign)
+        switch (get())
+        {
+            case '.':
+            {
+                add(decimal_point_char);
+                decimal_point_position = token_buffer.size() - 1;
+                goto scan_number_decimal1;
+            }
+
+            case 'e':
+            case 'E':
+            {
+                add(current);
+                goto scan_number_exponent;
+            }
+
+            default:
+                goto scan_number_done;
+        }
+
+scan_number_any1:
+        // state: we just parsed a number 0-9 (maybe with a leading minus sign)
+        switch (get())
+        {
+            case '0':
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+            {
+                add(current);
+                goto scan_number_any1;
+            }
+
+            case '.':
+            {
+                add(decimal_point_char);
+                decimal_point_position = token_buffer.size() - 1;
+                goto scan_number_decimal1;
+            }
+
+            case 'e':
+            case 'E':
+            {
+                add(current);
+                goto scan_number_exponent;
+            }
+
+            default:
+                goto scan_number_done;
+        }
+
+scan_number_decimal1:
+        // state: we just parsed a decimal point
+        number_type = token_type::value_float;
+        switch (get())
+        {
+            case '0':
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+            {
+                add(current);
+                goto scan_number_decimal2;
+            }
+
+            default:
+            {
+                error_message = "invalid number; expected digit after '.'";
+                return token_type::parse_error;
+            }
+        }
+
+scan_number_decimal2:
+        // we just parsed at least one number after a decimal point
+        switch (get())
+        {
+            case '0':
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+            {
+                add(current);
+                goto scan_number_decimal2;
+            }
+
+            case 'e':
+            case 'E':
+            {
+                add(current);
+                goto scan_number_exponent;
+            }
+
+            default:
+                goto scan_number_done;
+        }
+
+scan_number_exponent:
+        // we just parsed an exponent
+        number_type = token_type::value_float;
+        switch (get())
+        {
+            case '+':
+            case '-':
+            {
+                add(current);
+                goto scan_number_sign;
+            }
+
+            case '0':
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+            {
+                add(current);
+                goto scan_number_any2;
+            }
+
+            default:
+            {
+                error_message =
+                    "invalid number; expected '+', '-', or digit after exponent";
+                return token_type::parse_error;
+            }
+        }
+
+scan_number_sign:
+        // we just parsed an exponent sign
+        switch (get())
+        {
+            case '0':
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+            {
+                add(current);
+                goto scan_number_any2;
+            }
+
+            default:
+            {
+                error_message = "invalid number; expected digit after exponent sign";
+                return token_type::parse_error;
+            }
+        }
+
+scan_number_any2:
+        // we just parsed a number after the exponent or exponent sign
+        switch (get())
+        {
+            case '0':
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+            {
+                add(current);
+                goto scan_number_any2;
+            }
+
+            default:
+                goto scan_number_done;
+        }
+
+scan_number_done:
+        // unget the character after the number (we only read it to know that
+        // we are done scanning a number)
+        unget();
+
+        char* endptr = nullptr; // NOLINT(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+        errno = 0;
+
+        // try to parse integers first and fall back to floats
+        if (number_type == token_type::value_unsigned)
+        {
+            const auto x = std::strtoull(token_buffer.data(), &endptr, 10);
+
+            // we checked the number format before
+            JSON_ASSERT(endptr == token_buffer.data() + token_buffer.size());
+
+            if (errno != ERANGE)
+            {
+                value_unsigned = static_cast<number_unsigned_t>(x);
+                if (value_unsigned == x)
+                {
+                    return token_type::value_unsigned;
+                }
+            }
+        }
+        else if (number_type == token_type::value_integer)
+        {
+            const auto x = std::strtoll(token_buffer.data(), &endptr, 10);
+
+            // we checked the number format before
+            JSON_ASSERT(endptr == token_buffer.data() + token_buffer.size());
+
+            if (errno != ERANGE)
+            {
+                value_integer = static_cast<number_integer_t>(x);
+                if (value_integer == x)
+                {
+                    return token_type::value_integer;
+                }
+            }
+        }
+
+        // this code is reached if we parse a floating-point number or if an
+        // integer conversion above failed
+        strtof(value_float, token_buffer.data(), &endptr);
+
+        // we checked the number format before
+        JSON_ASSERT(endptr == token_buffer.data() + token_buffer.size());
+
+        return token_type::value_float;
+    }
+
+    /*!
+    @param[in] literal_text  the literal text to expect
+    @param[in] length        the length of the passed literal text
+    @param[in] return_type   the token type to return on success
+    */
+    JSON_HEDLEY_NON_NULL(2)
+    token_type scan_literal(const char_type* literal_text, const std::size_t length,
+                            token_type return_type)
+    {
+        JSON_ASSERT(char_traits<char_type>::to_char_type(current) == literal_text[0]);
+        for (std::size_t i = 1; i < length; ++i)
+        {
+            if (JSON_HEDLEY_UNLIKELY(char_traits<char_type>::to_char_type(get()) != literal_text[i]))
+            {
+                error_message = "invalid literal";
+                return token_type::parse_error;
+            }
+        }
+        return return_type;
+    }
+
+    /////////////////////
+    // input management
+    /////////////////////
+
+    /// reset token_buffer; current character is beginning of token
+    void reset() noexcept
+    {
+        token_buffer.clear();
+        token_string.clear();
+        decimal_point_position = std::string::npos;
+        token_string.push_back(char_traits<char_type>::to_char_type(current));
+    }
+
+    /*
+    @brief get next character from the input
+
+    This function provides the interface to the used input adapter. It does
+    not throw in case the input reached EOF, but returns a
+    `char_traits<char>::eof()` in that case.  Stores the scanned characters
+    for use in error messages.
+
+    @return character read from the input
+    */
+    char_int_type get()
+    {
+        ++position.chars_read_total;
+        ++position.chars_read_current_line;
+
+        if (next_unget)
+        {
+            // just reset the next_unget variable and work with current
+            next_unget = false;
+        }
+        else
+        {
+            current = ia.get_character();
+        }
+
+        if (JSON_HEDLEY_LIKELY(current != char_traits<char_type>::eof()))
+        {
+            token_string.push_back(char_traits<char_type>::to_char_type(current));
+        }
+
+        if (current == '\n')
+        {
+            ++position.lines_read;
+            position.chars_read_current_line = 0;
+        }
+
+        return current;
+    }
+
+    /*!
+    @brief unget current character (read it again on next get)
+
+    We implement unget by setting variable next_unget to true. The input is not
+    changed - we just simulate ungetting by modifying chars_read_total,
+    chars_read_current_line, and token_string. The next call to get() will
+    behave as if the unget character is read again.
+    */
+    void unget()
+    {
+        next_unget = true;
+
+        --position.chars_read_total;
+
+        // in case we "unget" a newline, we have to also decrement the lines_read
+        if (position.chars_read_current_line == 0)
+        {
+            if (position.lines_read > 0)
+            {
+                --position.lines_read;
+            }
+        }
+        else
+        {
+            --position.chars_read_current_line;
+        }
+
+        if (JSON_HEDLEY_LIKELY(current != char_traits<char_type>::eof()))
+        {
+            JSON_ASSERT(!token_string.empty());
+            token_string.pop_back();
+        }
+    }
+
+    /// add a character to token_buffer
+    void add(char_int_type c)
+    {
+        token_buffer.push_back(static_cast<typename string_t::value_type>(c));
+    }
+
+  public:
+    /////////////////////
+    // value getters
+    /////////////////////
+
+    /// return integer value
+    constexpr number_integer_t get_number_integer() const noexcept
+    {
+        return value_integer;
+    }
+
+    /// return unsigned integer value
+    constexpr number_unsigned_t get_number_unsigned() const noexcept
+    {
+        return value_unsigned;
+    }
+
+    /// return floating-point value
+    constexpr number_float_t get_number_float() const noexcept
+    {
+        return value_float;
+    }
+
+    /// return current string value (implicitly resets the token; useful only once)
+    string_t& get_string()
+    {
+        // translate decimal points from locale back to '.' (#4084)
+        if (decimal_point_char != '.' && decimal_point_position != std::string::npos)
+        {
+            token_buffer[decimal_point_position] = '.';
+        }
+        return token_buffer;
+    }
+
+    /////////////////////
+    // diagnostics
+    /////////////////////
+
+    /// return position of last read token
+    constexpr position_t get_position() const noexcept
+    {
+        return position;
+    }
+
+    /// return the last read token (for errors only).  Will never contain EOF
+    /// (an arbitrary value that is not a valid char value, often -1), because
+    /// 255 may legitimately occur.  May contain NUL, which should be escaped.
+    std::string get_token_string() const
+    {
+        // escape control characters
+        std::string result;
+        for (const auto c : token_string)
+        {
+            if (static_cast<unsigned char>(c) <= '\x1F')
+            {
+                // escape control characters
+                std::array<char, 9> cs{{}};
+                static_cast<void>((std::snprintf)(cs.data(), cs.size(), "<U+%.4X>", static_cast<unsigned char>(c))); // NOLINT(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+                result += cs.data();
+            }
+            else
+            {
+                // add character as is
+                result.push_back(static_cast<std::string::value_type>(c));
+            }
+        }
+
+        return result;
+    }
+
+    /// return syntax error message
+    JSON_HEDLEY_RETURNS_NON_NULL
+    constexpr const char* get_error_message() const noexcept
+    {
+        return error_message;
+    }
+
+    /////////////////////
+    // actual scanner
+    /////////////////////
+
+    /*!
+    @brief skip the UTF-8 byte order mark
+    @return true iff there is no BOM or the correct BOM has been skipped
+    */
+    bool skip_bom()
+    {
+        if (get() == 0xEF)
+        {
+            // check if we completely parse the BOM
+            return get() == 0xBB && get() == 0xBF;
+        }
+
+        // the first character is not the beginning of the BOM; unget it to
+        // process is later
+        unget();
+        return true;
+    }
+
+    void skip_whitespace()
+    {
+        do
+        {
+            get();
+        }
+        while (current == ' ' || current == '\t' || current == '\n' || current == '\r');
+    }
+
+    token_type scan()
+    {
+        // initially, skip the BOM
+        if (position.chars_read_total == 0 && !skip_bom())
+        {
+            error_message = "invalid BOM; must be 0xEF 0xBB 0xBF if given";
+            return token_type::parse_error;
+        }
+
+        // read next character and ignore whitespace
+        skip_whitespace();
+
+        // ignore comments
+        while (ignore_comments && current == '/')
+        {
+            if (!scan_comment())
+            {
+                return token_type::parse_error;
+            }
+
+            // skip following whitespace
+            skip_whitespace();
+        }
+
+        switch (current)
+        {
+            // structural characters
+            case '[':
+                return token_type::begin_array;
+            case ']':
+                return token_type::end_array;
+            case '{':
+                return token_type::begin_object;
+            case '}':
+                return token_type::end_object;
+            case ':':
+                return token_type::name_separator;
+            case ',':
+                return token_type::value_separator;
+
+            // literals
+            case 't':
+            {
+                std::array<char_type, 4> true_literal = {{static_cast<char_type>('t'), static_cast<char_type>('r'), static_cast<char_type>('u'), static_cast<char_type>('e')}};
+                return scan_literal(true_literal.data(), true_literal.size(), token_type::literal_true);
+            }
+            case 'f':
+            {
+                std::array<char_type, 5> false_literal = {{static_cast<char_type>('f'), static_cast<char_type>('a'), static_cast<char_type>('l'), static_cast<char_type>('s'), static_cast<char_type>('e')}};
+                return scan_literal(false_literal.data(), false_literal.size(), token_type::literal_false);
+            }
+            case 'n':
+            {
+                std::array<char_type, 4> null_literal = {{static_cast<char_type>('n'), static_cast<char_type>('u'), static_cast<char_type>('l'), static_cast<char_type>('l')}};
+                return scan_literal(null_literal.data(), null_literal.size(), token_type::literal_null);
+            }
+
+            // string
+            case '\"':
+                return scan_string();
+
+            // number
+            case '-':
+            case '0':
+            case '1':
+            case '2':
+            case '3':
+            case '4':
+            case '5':
+            case '6':
+            case '7':
+            case '8':
+            case '9':
+                return scan_number();
+
+            // end of input (the null byte is needed when parsing from
+            // string literals)
+            case '\0':
+            case char_traits<char_type>::eof():
+                return token_type::end_of_input;
+
+            // error
+            default:
+                error_message = "invalid literal";
+                return token_type::parse_error;
+        }
+    }
+
+  private:
+    /// input adapter
+    InputAdapterType ia;
+
+    /// whether comments should be ignored (true) or signaled as errors (false)
+    const bool ignore_comments = false;
+
+    /// the current character
+    char_int_type current = char_traits<char_type>::eof();
+
+    /// whether the next get() call should just return current
+    bool next_unget = false;
+
+    /// the start position of the current token
+    position_t position {};
+
+    /// raw input token string (for error messages)
+    std::vector<char_type> token_string {};
+
+    /// buffer for variable-length tokens (numbers, strings)
+    string_t token_buffer {};
+
+    /// a description of occurred lexer errors
+    const char* error_message = "";
+
+    // number values
+    number_integer_t value_integer = 0;
+    number_unsigned_t value_unsigned = 0;
+    number_float_t value_float = 0;
+
+    /// the decimal point
+    const char_int_type decimal_point_char = '.';
+    /// the position of the decimal point in the input
+    std::size_t decimal_point_position = std::string::npos;
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/string_concat.hpp>
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+/*!
+@brief SAX interface
+
+This class describes the SAX interface used by @ref nlohmann::json::sax_parse.
+Each function is called in different situations while the input is parsed. The
+boolean return value informs the parser whether to continue processing the
+input.
+*/
+template<typename BasicJsonType>
+struct json_sax
+{
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using binary_t = typename BasicJsonType::binary_t;
+
+    /*!
+    @brief a null value was read
+    @return whether parsing should proceed
+    */
+    virtual bool null() = 0;
+
+    /*!
+    @brief a boolean value was read
+    @param[in] val  boolean value
+    @return whether parsing should proceed
+    */
+    virtual bool boolean(bool val) = 0;
+
+    /*!
+    @brief an integer number was read
+    @param[in] val  integer value
+    @return whether parsing should proceed
+    */
+    virtual bool number_integer(number_integer_t val) = 0;
+
+    /*!
+    @brief an unsigned integer number was read
+    @param[in] val  unsigned integer value
+    @return whether parsing should proceed
+    */
+    virtual bool number_unsigned(number_unsigned_t val) = 0;
+
+    /*!
+    @brief a floating-point number was read
+    @param[in] val  floating-point value
+    @param[in] s    raw token value
+    @return whether parsing should proceed
+    */
+    virtual bool number_float(number_float_t val, const string_t& s) = 0;
+
+    /*!
+    @brief a string value was read
+    @param[in] val  string value
+    @return whether parsing should proceed
+    @note It is safe to move the passed string value.
+    */
+    virtual bool string(string_t& val) = 0;
+
+    /*!
+    @brief a binary value was read
+    @param[in] val  binary value
+    @return whether parsing should proceed
+    @note It is safe to move the passed binary value.
+    */
+    virtual bool binary(binary_t& val) = 0;
+
+    /*!
+    @brief the beginning of an object was read
+    @param[in] elements  number of object elements or -1 if unknown
+    @return whether parsing should proceed
+    @note binary formats may report the number of elements
+    */
+    virtual bool start_object(std::size_t elements) = 0;
+
+    /*!
+    @brief an object key was read
+    @param[in] val  object key
+    @return whether parsing should proceed
+    @note It is safe to move the passed string.
+    */
+    virtual bool key(string_t& val) = 0;
+
+    /*!
+    @brief the end of an object was read
+    @return whether parsing should proceed
+    */
+    virtual bool end_object() = 0;
+
+    /*!
+    @brief the beginning of an array was read
+    @param[in] elements  number of array elements or -1 if unknown
+    @return whether parsing should proceed
+    @note binary formats may report the number of elements
+    */
+    virtual bool start_array(std::size_t elements) = 0;
+
+    /*!
+    @brief the end of an array was read
+    @return whether parsing should proceed
+    */
+    virtual bool end_array() = 0;
+
+    /*!
+    @brief a parse error occurred
+    @param[in] position    the position in the input where the error occurs
+    @param[in] last_token  the last read token
+    @param[in] ex          an exception object describing the error
+    @return whether parsing should proceed (must return false)
+    */
+    virtual bool parse_error(std::size_t position,
+                             const std::string& last_token,
+                             const detail::exception& ex) = 0;
+
+    json_sax() = default;
+    json_sax(const json_sax&) = default;
+    json_sax(json_sax&&) noexcept = default;
+    json_sax& operator=(const json_sax&) = default;
+    json_sax& operator=(json_sax&&) noexcept = default;
+    virtual ~json_sax() = default;
+};
+
+namespace detail
+{
+constexpr std::size_t unknown_size()
+{
+    return (std::numeric_limits<std::size_t>::max)();
+}
+
+/*!
+@brief SAX implementation to create a JSON value from SAX events
+
+This class implements the @ref json_sax interface and processes the SAX events
+to create a JSON value which makes it basically a DOM parser. The structure or
+hierarchy of the JSON value is managed by the stack `ref_stack` which contains
+a pointer to the respective array or object for each recursion depth.
+
+After successful parsing, the value that is passed by reference to the
+constructor contains the parsed value.
+
+@tparam BasicJsonType  the JSON type
+*/
+template<typename BasicJsonType, typename InputAdapterType>
+class json_sax_dom_parser
+{
+  public:
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using binary_t = typename BasicJsonType::binary_t;
+    using lexer_t = lexer<BasicJsonType, InputAdapterType>;
+
+    /*!
+    @param[in,out] r  reference to a JSON value that is manipulated while
+                       parsing
+    @param[in] allow_exceptions_  whether parse errors yield exceptions
+    */
+    explicit json_sax_dom_parser(BasicJsonType& r, const bool allow_exceptions_ = true, lexer_t* lexer_ = nullptr)
+        : root(r), allow_exceptions(allow_exceptions_), m_lexer_ref(lexer_)
+    {}
+
+    // make class move-only
+    json_sax_dom_parser(const json_sax_dom_parser&) = delete;
+    json_sax_dom_parser(json_sax_dom_parser&&) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor)
+    json_sax_dom_parser& operator=(const json_sax_dom_parser&) = delete;
+    json_sax_dom_parser& operator=(json_sax_dom_parser&&) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor)
+    ~json_sax_dom_parser() = default;
+
+    bool null()
+    {
+        handle_value(nullptr);
+        return true;
+    }
+
+    bool boolean(bool val)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool number_integer(number_integer_t val)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool number_unsigned(number_unsigned_t val)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool number_float(number_float_t val, const string_t& /*unused*/)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool string(string_t& val)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool binary(binary_t& val)
+    {
+        handle_value(std::move(val));
+        return true;
+    }
+
+    bool start_object(std::size_t len)
+    {
+        ref_stack.push_back(handle_value(BasicJsonType::value_t::object));
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        // Manually set the start position of the object here.
+        // Ensure this is after the call to handle_value to ensure correct start position.
+        if (m_lexer_ref)
+        {
+            // Lexer has read the first character of the object, so
+            // subtract 1 from the position to get the correct start position.
+            ref_stack.back()->start_position = m_lexer_ref->get_position() - 1;
+        }
+#endif
+
+        if (JSON_HEDLEY_UNLIKELY(len != detail::unknown_size() && len > ref_stack.back()->max_size()))
+        {
+            JSON_THROW(out_of_range::create(408, concat("excessive object size: ", std::to_string(len)), ref_stack.back()));
+        }
+
+        return true;
+    }
+
+    bool key(string_t& val)
+    {
+        JSON_ASSERT(!ref_stack.empty());
+        JSON_ASSERT(ref_stack.back()->is_object());
+
+        // add null at given key and store the reference for later
+        object_element = &(ref_stack.back()->m_data.m_value.object->operator[](val));
+        return true;
+    }
+
+    bool end_object()
+    {
+        JSON_ASSERT(!ref_stack.empty());
+        JSON_ASSERT(ref_stack.back()->is_object());
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        if (m_lexer_ref)
+        {
+            // Lexer's position is past the closing brace, so set that as the end position.
+            ref_stack.back()->end_position = m_lexer_ref->get_position();
+        }
+#endif
+
+        ref_stack.back()->set_parents();
+        ref_stack.pop_back();
+        return true;
+    }
+
+    bool start_array(std::size_t len)
+    {
+        ref_stack.push_back(handle_value(BasicJsonType::value_t::array));
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        // Manually set the start position of the array here.
+        // Ensure this is after the call to handle_value to ensure correct start position.
+        if (m_lexer_ref)
+        {
+            ref_stack.back()->start_position = m_lexer_ref->get_position() - 1;
+        }
+#endif
+
+        if (JSON_HEDLEY_UNLIKELY(len != detail::unknown_size() && len > ref_stack.back()->max_size()))
+        {
+            JSON_THROW(out_of_range::create(408, concat("excessive array size: ", std::to_string(len)), ref_stack.back()));
+        }
+
+        return true;
+    }
+
+    bool end_array()
+    {
+        JSON_ASSERT(!ref_stack.empty());
+        JSON_ASSERT(ref_stack.back()->is_array());
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        if (m_lexer_ref)
+        {
+            // Lexer's position is past the closing bracket, so set that as the end position.
+            ref_stack.back()->end_position = m_lexer_ref->get_position();
+        }
+#endif
+
+        ref_stack.back()->set_parents();
+        ref_stack.pop_back();
+        return true;
+    }
+
+    template<class Exception>
+    bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/,
+                     const Exception& ex)
+    {
+        errored = true;
+        static_cast<void>(ex);
+        if (allow_exceptions)
+        {
+            JSON_THROW(ex);
+        }
+        return false;
+    }
+
+    constexpr bool is_errored() const
+    {
+        return errored;
+    }
+
+  private:
+
+#if JSON_DIAGNOSTIC_POSITIONS
+    void handle_diagnostic_positions_for_json_value(BasicJsonType& v)
+    {
+        if (m_lexer_ref)
+        {
+            // Lexer has read past the current field value, so set the end position to the current position.
+            // The start position will be set below based on the length of the string representation
+            // of the value.
+            v.end_position = m_lexer_ref->get_position();
+
+            switch (v.type())
+            {
+                case value_t::boolean:
+                {
+                    // 4 and 5 are the string length of "true" and "false"
+                    v.start_position = v.end_position - (v.m_data.m_value.boolean ? 4 : 5);
+                    break;
+                }
+
+                case value_t::null:
+                {
+                    // 4 is the string length of "null"
+                    v.start_position = v.end_position - 4;
+                    break;
+                }
+
+                case value_t::string:
+                {
+                    // include the length of the quotes, which is 2
+                    v.start_position = v.end_position - v.m_data.m_value.string->size() - 2;
+                    break;
+                }
+
+                // As we handle the start and end positions for values created during parsing,
+                // we do not expect the following value type to be called. Regardless, set the positions
+                // in case this is created manually or through a different constructor. Exclude from lcov
+                // since the exact condition of this switch is esoteric.
+                // LCOV_EXCL_START
+                case value_t::discarded:
+                {
+                    v.end_position = std::string::npos;
+                    v.start_position = v.end_position;
+                    break;
+                }
+                // LCOV_EXCL_STOP
+                case value_t::binary:
+                case value_t::number_integer:
+                case value_t::number_unsigned:
+                case value_t::number_float:
+                {
+                    v.start_position = v.end_position - m_lexer_ref->get_string().size();
+                    break;
+                }
+                case value_t::object:
+                case value_t::array:
+                {
+                    // object and array are handled in start_object() and start_array() handlers
+                    // skip setting the values here.
+                    break;
+                }
+                default: // LCOV_EXCL_LINE
+                    // Handle all possible types discretely, default handler should never be reached.
+                    JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert,-warnings-as-errors) LCOV_EXCL_LINE
+            }
+        }
+    }
+#endif
+
+    /*!
+    @invariant If the ref stack is empty, then the passed value will be the new
+               root.
+    @invariant If the ref stack contains a value, then it is an array or an
+               object to which we can add elements
+    */
+    template<typename Value>
+    JSON_HEDLEY_RETURNS_NON_NULL
+    BasicJsonType* handle_value(Value&& v)
+    {
+        if (ref_stack.empty())
+        {
+            root = BasicJsonType(std::forward<Value>(v));
+
+#if JSON_DIAGNOSTIC_POSITIONS
+            handle_diagnostic_positions_for_json_value(root);
+#endif
+
+            return &root;
+        }
+
+        JSON_ASSERT(ref_stack.back()->is_array() || ref_stack.back()->is_object());
+
+        if (ref_stack.back()->is_array())
+        {
+            ref_stack.back()->m_data.m_value.array->emplace_back(std::forward<Value>(v));
+
+#if JSON_DIAGNOSTIC_POSITIONS
+            handle_diagnostic_positions_for_json_value(ref_stack.back()->m_data.m_value.array->back());
+#endif
+
+            return &(ref_stack.back()->m_data.m_value.array->back());
+        }
+
+        JSON_ASSERT(ref_stack.back()->is_object());
+        JSON_ASSERT(object_element);
+        *object_element = BasicJsonType(std::forward<Value>(v));
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        handle_diagnostic_positions_for_json_value(*object_element);
+#endif
+
+        return object_element;
+    }
+
+    /// the parsed JSON value
+    BasicJsonType& root;
+    /// stack to model hierarchy of values
+    std::vector<BasicJsonType*> ref_stack {};
+    /// helper to hold the reference for the next object element
+    BasicJsonType* object_element = nullptr;
+    /// whether a syntax error occurred
+    bool errored = false;
+    /// whether to throw exceptions in case of errors
+    const bool allow_exceptions = true;
+    /// the lexer reference to obtain the current position
+    lexer_t* m_lexer_ref = nullptr;
+};
+
+template<typename BasicJsonType, typename InputAdapterType>
+class json_sax_dom_callback_parser
+{
+  public:
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using binary_t = typename BasicJsonType::binary_t;
+    using parser_callback_t = typename BasicJsonType::parser_callback_t;
+    using parse_event_t = typename BasicJsonType::parse_event_t;
+    using lexer_t = lexer<BasicJsonType, InputAdapterType>;
+
+    json_sax_dom_callback_parser(BasicJsonType& r,
+                                 parser_callback_t cb,
+                                 const bool allow_exceptions_ = true,
+                                 lexer_t* lexer_ = nullptr)
+        : root(r), callback(std::move(cb)), allow_exceptions(allow_exceptions_), m_lexer_ref(lexer_)
+    {
+        keep_stack.push_back(true);
+    }
+
+    // make class move-only
+    json_sax_dom_callback_parser(const json_sax_dom_callback_parser&) = delete;
+    json_sax_dom_callback_parser(json_sax_dom_callback_parser&&) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor)
+    json_sax_dom_callback_parser& operator=(const json_sax_dom_callback_parser&) = delete;
+    json_sax_dom_callback_parser& operator=(json_sax_dom_callback_parser&&) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor)
+    ~json_sax_dom_callback_parser() = default;
+
+    bool null()
+    {
+        handle_value(nullptr);
+        return true;
+    }
+
+    bool boolean(bool val)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool number_integer(number_integer_t val)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool number_unsigned(number_unsigned_t val)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool number_float(number_float_t val, const string_t& /*unused*/)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool string(string_t& val)
+    {
+        handle_value(val);
+        return true;
+    }
+
+    bool binary(binary_t& val)
+    {
+        handle_value(std::move(val));
+        return true;
+    }
+
+    bool start_object(std::size_t len)
+    {
+        // check callback for object start
+        const bool keep = callback(static_cast<int>(ref_stack.size()), parse_event_t::object_start, discarded);
+        keep_stack.push_back(keep);
+
+        auto val = handle_value(BasicJsonType::value_t::object, true);
+        ref_stack.push_back(val.second);
+
+        if (ref_stack.back())
+        {
+
+#if JSON_DIAGNOSTIC_POSITIONS
+            // Manually set the start position of the object here.
+            // Ensure this is after the call to handle_value to ensure correct start position.
+            if (m_lexer_ref)
+            {
+                // Lexer has read the first character of the object, so
+                // subtract 1 from the position to get the correct start position.
+                ref_stack.back()->start_position = m_lexer_ref->get_position() - 1;
+            }
+#endif
+
+            // check object limit
+            if (JSON_HEDLEY_UNLIKELY(len != detail::unknown_size() && len > ref_stack.back()->max_size()))
+            {
+                JSON_THROW(out_of_range::create(408, concat("excessive object size: ", std::to_string(len)), ref_stack.back()));
+            }
+        }
+        return true;
+    }
+
+    bool key(string_t& val)
+    {
+        BasicJsonType k = BasicJsonType(val);
+
+        // check callback for key
+        const bool keep = callback(static_cast<int>(ref_stack.size()), parse_event_t::key, k);
+        key_keep_stack.push_back(keep);
+
+        // add discarded value at given key and store the reference for later
+        if (keep && ref_stack.back())
+        {
+            object_element = &(ref_stack.back()->m_data.m_value.object->operator[](val) = discarded);
+        }
+
+        return true;
+    }
+
+    bool end_object()
+    {
+        if (ref_stack.back())
+        {
+            if (!callback(static_cast<int>(ref_stack.size()) - 1, parse_event_t::object_end, *ref_stack.back()))
+            {
+                // discard object
+                *ref_stack.back() = discarded;
+
+#if JSON_DIAGNOSTIC_POSITIONS
+                // Set start/end positions for discarded object.
+                handle_diagnostic_positions_for_json_value(*ref_stack.back());
+#endif
+            }
+            else
+            {
+
+#if JSON_DIAGNOSTIC_POSITIONS
+                if (m_lexer_ref)
+                {
+                    // Lexer's position is past the closing brace, so set that as the end position.
+                    ref_stack.back()->end_position = m_lexer_ref->get_position();
+                }
+#endif
+
+                ref_stack.back()->set_parents();
+            }
+        }
+
+        JSON_ASSERT(!ref_stack.empty());
+        JSON_ASSERT(!keep_stack.empty());
+        ref_stack.pop_back();
+        keep_stack.pop_back();
+
+        if (!ref_stack.empty() && ref_stack.back() && ref_stack.back()->is_structured())
+        {
+            // remove discarded value
+            for (auto it = ref_stack.back()->begin(); it != ref_stack.back()->end(); ++it)
+            {
+                if (it->is_discarded())
+                {
+                    ref_stack.back()->erase(it);
+                    break;
+                }
+            }
+        }
+
+        return true;
+    }
+
+    bool start_array(std::size_t len)
+    {
+        const bool keep = callback(static_cast<int>(ref_stack.size()), parse_event_t::array_start, discarded);
+        keep_stack.push_back(keep);
+
+        auto val = handle_value(BasicJsonType::value_t::array, true);
+        ref_stack.push_back(val.second);
+
+        if (ref_stack.back())
+        {
+
+#if JSON_DIAGNOSTIC_POSITIONS
+            // Manually set the start position of the array here.
+            // Ensure this is after the call to handle_value to ensure correct start position.
+            if (m_lexer_ref)
+            {
+                // Lexer has read the first character of the array, so
+                // subtract 1 from the position to get the correct start position.
+                ref_stack.back()->start_position = m_lexer_ref->get_position() - 1;
+            }
+#endif
+
+            // check array limit
+            if (JSON_HEDLEY_UNLIKELY(len != detail::unknown_size() && len > ref_stack.back()->max_size()))
+            {
+                JSON_THROW(out_of_range::create(408, concat("excessive array size: ", std::to_string(len)), ref_stack.back()));
+            }
+        }
+
+        return true;
+    }
+
+    bool end_array()
+    {
+        bool keep = true;
+
+        if (ref_stack.back())
+        {
+            keep = callback(static_cast<int>(ref_stack.size()) - 1, parse_event_t::array_end, *ref_stack.back());
+            if (keep)
+            {
+
+#if JSON_DIAGNOSTIC_POSITIONS
+                if (m_lexer_ref)
+                {
+                    // Lexer's position is past the closing bracket, so set that as the end position.
+                    ref_stack.back()->end_position = m_lexer_ref->get_position();
+                }
+#endif
+
+                ref_stack.back()->set_parents();
+            }
+            else
+            {
+                // discard array
+                *ref_stack.back() = discarded;
+
+#if JSON_DIAGNOSTIC_POSITIONS
+                // Set start/end positions for discarded array.
+                handle_diagnostic_positions_for_json_value(*ref_stack.back());
+#endif
+            }
+        }
+
+        JSON_ASSERT(!ref_stack.empty());
+        JSON_ASSERT(!keep_stack.empty());
+        ref_stack.pop_back();
+        keep_stack.pop_back();
+
+        // remove discarded value
+        if (!keep && !ref_stack.empty() && ref_stack.back()->is_array())
+        {
+            ref_stack.back()->m_data.m_value.array->pop_back();
+        }
+
+        return true;
+    }
+
+    template<class Exception>
+    bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/,
+                     const Exception& ex)
+    {
+        errored = true;
+        static_cast<void>(ex);
+        if (allow_exceptions)
+        {
+            JSON_THROW(ex);
+        }
+        return false;
+    }
+
+    constexpr bool is_errored() const
+    {
+        return errored;
+    }
+
+  private:
+
+#if JSON_DIAGNOSTIC_POSITIONS
+    void handle_diagnostic_positions_for_json_value(BasicJsonType& v)
+    {
+        if (m_lexer_ref)
+        {
+            // Lexer has read past the current field value, so set the end position to the current position.
+            // The start position will be set below based on the length of the string representation
+            // of the value.
+            v.end_position = m_lexer_ref->get_position();
+
+            switch (v.type())
+            {
+                case value_t::boolean:
+                {
+                    // 4 and 5 are the string length of "true" and "false"
+                    v.start_position = v.end_position - (v.m_data.m_value.boolean ? 4 : 5);
+                    break;
+                }
+
+                case value_t::null:
+                {
+                    // 4 is the string length of "null"
+                    v.start_position = v.end_position - 4;
+                    break;
+                }
+
+                case value_t::string:
+                {
+                    // include the length of the quotes, which is 2
+                    v.start_position = v.end_position - v.m_data.m_value.string->size() - 2;
+                    break;
+                }
+
+                case value_t::discarded:
+                {
+                    v.end_position = std::string::npos;
+                    v.start_position = v.end_position;
+                    break;
+                }
+
+                case value_t::binary:
+                case value_t::number_integer:
+                case value_t::number_unsigned:
+                case value_t::number_float:
+                {
+                    v.start_position = v.end_position - m_lexer_ref->get_string().size();
+                    break;
+                }
+
+                case value_t::object:
+                case value_t::array:
+                {
+                    // object and array are handled in start_object() and start_array() handlers
+                    // skip setting the values here.
+                    break;
+                }
+                default: // LCOV_EXCL_LINE
+                    // Handle all possible types discretely, default handler should never be reached.
+                    JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert,-warnings-as-errors) LCOV_EXCL_LINE
+            }
+        }
+    }
+#endif
+
+    /*!
+    @param[in] v  value to add to the JSON value we build during parsing
+    @param[in] skip_callback  whether we should skip calling the callback
+               function; this is required after start_array() and
+               start_object() SAX events, because otherwise we would call the
+               callback function with an empty array or object, respectively.
+
+    @invariant If the ref stack is empty, then the passed value will be the new
+               root.
+    @invariant If the ref stack contains a value, then it is an array or an
+               object to which we can add elements
+
+    @return pair of boolean (whether value should be kept) and pointer (to the
+            passed value in the ref_stack hierarchy; nullptr if not kept)
+    */
+    template<typename Value>
+    std::pair<bool, BasicJsonType*> handle_value(Value&& v, const bool skip_callback = false)
+    {
+        JSON_ASSERT(!keep_stack.empty());
+
+        // do not handle this value if we know it would be added to a discarded
+        // container
+        if (!keep_stack.back())
+        {
+            return {false, nullptr};
+        }
+
+        // create value
+        auto value = BasicJsonType(std::forward<Value>(v));
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        handle_diagnostic_positions_for_json_value(value);
+#endif
+
+        // check callback
+        const bool keep = skip_callback || callback(static_cast<int>(ref_stack.size()), parse_event_t::value, value);
+
+        // do not handle this value if we just learnt it shall be discarded
+        if (!keep)
+        {
+            return {false, nullptr};
+        }
+
+        if (ref_stack.empty())
+        {
+            root = std::move(value);
+            return {true, & root};
+        }
+
+        // skip this value if we already decided to skip the parent
+        // (https://github.com/nlohmann/json/issues/971#issuecomment-413678360)
+        if (!ref_stack.back())
+        {
+            return {false, nullptr};
+        }
+
+        // we now only expect arrays and objects
+        JSON_ASSERT(ref_stack.back()->is_array() || ref_stack.back()->is_object());
+
+        // array
+        if (ref_stack.back()->is_array())
+        {
+            ref_stack.back()->m_data.m_value.array->emplace_back(std::move(value));
+            return {true, & (ref_stack.back()->m_data.m_value.array->back())};
+        }
+
+        // object
+        JSON_ASSERT(ref_stack.back()->is_object());
+        // check if we should store an element for the current key
+        JSON_ASSERT(!key_keep_stack.empty());
+        const bool store_element = key_keep_stack.back();
+        key_keep_stack.pop_back();
+
+        if (!store_element)
+        {
+            return {false, nullptr};
+        }
+
+        JSON_ASSERT(object_element);
+        *object_element = std::move(value);
+        return {true, object_element};
+    }
+
+    /// the parsed JSON value
+    BasicJsonType& root;
+    /// stack to model hierarchy of values
+    std::vector<BasicJsonType*> ref_stack {};
+    /// stack to manage which values to keep
+    std::vector<bool> keep_stack {}; // NOLINT(readability-redundant-member-init)
+    /// stack to manage which object keys to keep
+    std::vector<bool> key_keep_stack {}; // NOLINT(readability-redundant-member-init)
+    /// helper to hold the reference for the next object element
+    BasicJsonType* object_element = nullptr;
+    /// whether a syntax error occurred
+    bool errored = false;
+    /// callback function
+    const parser_callback_t callback = nullptr;
+    /// whether to throw exceptions in case of errors
+    const bool allow_exceptions = true;
+    /// a discarded value for the callback
+    BasicJsonType discarded = BasicJsonType::value_t::discarded;
+    /// the lexer reference to obtain the current position
+    lexer_t* m_lexer_ref = nullptr;
+};
+
+template<typename BasicJsonType>
+class json_sax_acceptor
+{
+  public:
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using binary_t = typename BasicJsonType::binary_t;
+
+    bool null()
+    {
+        return true;
+    }
+
+    bool boolean(bool /*unused*/)
+    {
+        return true;
+    }
+
+    bool number_integer(number_integer_t /*unused*/)
+    {
+        return true;
+    }
+
+    bool number_unsigned(number_unsigned_t /*unused*/)
+    {
+        return true;
+    }
+
+    bool number_float(number_float_t /*unused*/, const string_t& /*unused*/)
+    {
+        return true;
+    }
+
+    bool string(string_t& /*unused*/)
+    {
+        return true;
+    }
+
+    bool binary(binary_t& /*unused*/)
+    {
+        return true;
+    }
+
+    bool start_object(std::size_t /*unused*/ = detail::unknown_size())
+    {
+        return true;
+    }
+
+    bool key(string_t& /*unused*/)
+    {
+        return true;
+    }
+
+    bool end_object()
+    {
+        return true;
+    }
+
+    bool start_array(std::size_t /*unused*/ = detail::unknown_size())
+    {
+        return true;
+    }
+
+    bool end_array()
+    {
+        return true;
+    }
+
+    bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/, const detail::exception& /*unused*/)
+    {
+        return false;
+    }
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/input/lexer.hpp>
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/is_sax.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstdint> // size_t
+#include <utility> // declval
+#include <string> // string
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+// #include <nlohmann/detail/meta/detected.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+template<typename T>
+using null_function_t = decltype(std::declval<T&>().null());
+
+template<typename T>
+using boolean_function_t =
+    decltype(std::declval<T&>().boolean(std::declval<bool>()));
+
+template<typename T, typename Integer>
+using number_integer_function_t =
+    decltype(std::declval<T&>().number_integer(std::declval<Integer>()));
+
+template<typename T, typename Unsigned>
+using number_unsigned_function_t =
+    decltype(std::declval<T&>().number_unsigned(std::declval<Unsigned>()));
+
+template<typename T, typename Float, typename String>
+using number_float_function_t = decltype(std::declval<T&>().number_float(
+                                    std::declval<Float>(), std::declval<const String&>()));
+
+template<typename T, typename String>
+using string_function_t =
+    decltype(std::declval<T&>().string(std::declval<String&>()));
+
+template<typename T, typename Binary>
+using binary_function_t =
+    decltype(std::declval<T&>().binary(std::declval<Binary&>()));
+
+template<typename T>
+using start_object_function_t =
+    decltype(std::declval<T&>().start_object(std::declval<std::size_t>()));
+
+template<typename T, typename String>
+using key_function_t =
+    decltype(std::declval<T&>().key(std::declval<String&>()));
+
+template<typename T>
+using end_object_function_t = decltype(std::declval<T&>().end_object());
+
+template<typename T>
+using start_array_function_t =
+    decltype(std::declval<T&>().start_array(std::declval<std::size_t>()));
+
+template<typename T>
+using end_array_function_t = decltype(std::declval<T&>().end_array());
+
+template<typename T, typename Exception>
+using parse_error_function_t = decltype(std::declval<T&>().parse_error(
+        std::declval<std::size_t>(), std::declval<const std::string&>(),
+        std::declval<const Exception&>()));
+
+template<typename SAX, typename BasicJsonType>
+struct is_sax
+{
+  private:
+    static_assert(is_basic_json<BasicJsonType>::value,
+                  "BasicJsonType must be of type basic_json<...>");
+
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using binary_t = typename BasicJsonType::binary_t;
+    using exception_t = typename BasicJsonType::exception;
+
+  public:
+    static constexpr bool value =
+        is_detected_exact<bool, null_function_t, SAX>::value &&
+        is_detected_exact<bool, boolean_function_t, SAX>::value &&
+        is_detected_exact<bool, number_integer_function_t, SAX, number_integer_t>::value &&
+        is_detected_exact<bool, number_unsigned_function_t, SAX, number_unsigned_t>::value &&
+        is_detected_exact<bool, number_float_function_t, SAX, number_float_t, string_t>::value &&
+        is_detected_exact<bool, string_function_t, SAX, string_t>::value &&
+        is_detected_exact<bool, binary_function_t, SAX, binary_t>::value &&
+        is_detected_exact<bool, start_object_function_t, SAX>::value &&
+        is_detected_exact<bool, key_function_t, SAX, string_t>::value &&
+        is_detected_exact<bool, end_object_function_t, SAX>::value &&
+        is_detected_exact<bool, start_array_function_t, SAX>::value &&
+        is_detected_exact<bool, end_array_function_t, SAX>::value &&
+        is_detected_exact<bool, parse_error_function_t, SAX, exception_t>::value;
+};
+
+template<typename SAX, typename BasicJsonType>
+struct is_sax_static_asserts
+{
+  private:
+    static_assert(is_basic_json<BasicJsonType>::value,
+                  "BasicJsonType must be of type basic_json<...>");
+
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using binary_t = typename BasicJsonType::binary_t;
+    using exception_t = typename BasicJsonType::exception;
+
+  public:
+    static_assert(is_detected_exact<bool, null_function_t, SAX>::value,
+                  "Missing/invalid function: bool null()");
+    static_assert(is_detected_exact<bool, boolean_function_t, SAX>::value,
+                  "Missing/invalid function: bool boolean(bool)");
+    static_assert(is_detected_exact<bool, boolean_function_t, SAX>::value,
+                  "Missing/invalid function: bool boolean(bool)");
+    static_assert(
+        is_detected_exact<bool, number_integer_function_t, SAX,
+        number_integer_t>::value,
+        "Missing/invalid function: bool number_integer(number_integer_t)");
+    static_assert(
+        is_detected_exact<bool, number_unsigned_function_t, SAX,
+        number_unsigned_t>::value,
+        "Missing/invalid function: bool number_unsigned(number_unsigned_t)");
+    static_assert(is_detected_exact<bool, number_float_function_t, SAX,
+                  number_float_t, string_t>::value,
+                  "Missing/invalid function: bool number_float(number_float_t, const string_t&)");
+    static_assert(
+        is_detected_exact<bool, string_function_t, SAX, string_t>::value,
+        "Missing/invalid function: bool string(string_t&)");
+    static_assert(
+        is_detected_exact<bool, binary_function_t, SAX, binary_t>::value,
+        "Missing/invalid function: bool binary(binary_t&)");
+    static_assert(is_detected_exact<bool, start_object_function_t, SAX>::value,
+                  "Missing/invalid function: bool start_object(std::size_t)");
+    static_assert(is_detected_exact<bool, key_function_t, SAX, string_t>::value,
+                  "Missing/invalid function: bool key(string_t&)");
+    static_assert(is_detected_exact<bool, end_object_function_t, SAX>::value,
+                  "Missing/invalid function: bool end_object()");
+    static_assert(is_detected_exact<bool, start_array_function_t, SAX>::value,
+                  "Missing/invalid function: bool start_array(std::size_t)");
+    static_assert(is_detected_exact<bool, end_array_function_t, SAX>::value,
+                  "Missing/invalid function: bool end_array()");
+    static_assert(
+        is_detected_exact<bool, parse_error_function_t, SAX, exception_t>::value,
+        "Missing/invalid function: bool parse_error(std::size_t, const "
+        "std::string&, const exception&)");
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+// #include <nlohmann/detail/string_concat.hpp>
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/// how to treat CBOR tags
+enum class cbor_tag_handler_t
+{
+    error,   ///< throw a parse_error exception in case of a tag
+    ignore,  ///< ignore tags
+    store    ///< store tags as binary type
+};
+
+/*!
+@brief determine system byte order
+
+@return true if and only if system's byte order is little endian
+
+@note from https://stackoverflow.com/a/1001328/266378
+*/
+static inline bool little_endianness(int num = 1) noexcept
+{
+    return *reinterpret_cast<char*>(&num) == 1;
+}
+
+///////////////////
+// binary reader //
+///////////////////
+
+/*!
+@brief deserialization of CBOR, MessagePack, and UBJSON values
+*/
+template<typename BasicJsonType, typename InputAdapterType, typename SAX = json_sax_dom_parser<BasicJsonType, InputAdapterType>>
+class binary_reader
+{
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using binary_t = typename BasicJsonType::binary_t;
+    using json_sax_t = SAX;
+    using char_type = typename InputAdapterType::char_type;
+    using char_int_type = typename char_traits<char_type>::int_type;
+
+  public:
+    /*!
+    @brief create a binary reader
+
+    @param[in] adapter  input adapter to read from
+    */
+    explicit binary_reader(InputAdapterType&& adapter, const input_format_t format = input_format_t::json) noexcept : ia(std::move(adapter)), input_format(format)
+    {
+        (void)detail::is_sax_static_asserts<SAX, BasicJsonType> {};
+    }
+
+    // make class move-only
+    binary_reader(const binary_reader&) = delete;
+    binary_reader(binary_reader&&) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor)
+    binary_reader& operator=(const binary_reader&) = delete;
+    binary_reader& operator=(binary_reader&&) = default; // NOLINT(hicpp-noexcept-move,performance-noexcept-move-constructor)
+    ~binary_reader() = default;
+
+    /*!
+    @param[in] format  the binary format to parse
+    @param[in] sax_    a SAX event processor
+    @param[in] strict  whether to expect the input to be consumed completed
+    @param[in] tag_handler  how to treat CBOR tags
+
+    @return whether parsing was successful
+    */
+    JSON_HEDLEY_NON_NULL(3)
+    bool sax_parse(const input_format_t format,
+                   json_sax_t* sax_,
+                   const bool strict = true,
+                   const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        sax = sax_;
+        bool result = false;
+
+        switch (format)
+        {
+            case input_format_t::bson:
+                result = parse_bson_internal();
+                break;
+
+            case input_format_t::cbor:
+                result = parse_cbor_internal(true, tag_handler);
+                break;
+
+            case input_format_t::msgpack:
+                result = parse_msgpack_internal();
+                break;
+
+            case input_format_t::ubjson:
+            case input_format_t::bjdata:
+                result = parse_ubjson_internal();
+                break;
+
+            case input_format_t::json: // LCOV_EXCL_LINE
+            default:            // LCOV_EXCL_LINE
+                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+        }
+
+        // strict mode: next byte must be EOF
+        if (result && strict)
+        {
+            if (input_format == input_format_t::ubjson || input_format == input_format_t::bjdata)
+            {
+                get_ignore_noop();
+            }
+            else
+            {
+                get();
+            }
+
+            if (JSON_HEDLEY_UNLIKELY(current != char_traits<char_type>::eof()))
+            {
+                return sax->parse_error(chars_read, get_token_string(), parse_error::create(110, chars_read,
+                                        exception_message(input_format, concat("expected end of input; last byte: 0x", get_token_string()), "value"), nullptr));
+            }
+        }
+
+        return result;
+    }
+
+  private:
+    //////////
+    // BSON //
+    //////////
+
+    /*!
+    @brief Reads in a BSON-object and passes it to the SAX-parser.
+    @return whether a valid BSON-value was passed to the SAX parser
+    */
+    bool parse_bson_internal()
+    {
+        std::int32_t document_size{};
+        get_number<std::int32_t, true>(input_format_t::bson, document_size);
+
+        if (JSON_HEDLEY_UNLIKELY(!sax->start_object(detail::unknown_size())))
+        {
+            return false;
+        }
+
+        if (JSON_HEDLEY_UNLIKELY(!parse_bson_element_list(/*is_array*/false)))
+        {
+            return false;
+        }
+
+        return sax->end_object();
+    }
+
+    /*!
+    @brief Parses a C-style string from the BSON input.
+    @param[in,out] result  A reference to the string variable where the read
+                            string is to be stored.
+    @return `true` if the \x00-byte indicating the end of the string was
+             encountered before the EOF; false` indicates an unexpected EOF.
+    */
+    bool get_bson_cstr(string_t& result)
+    {
+        auto out = std::back_inserter(result);
+        while (true)
+        {
+            get();
+            if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format_t::bson, "cstring")))
+            {
+                return false;
+            }
+            if (current == 0x00)
+            {
+                return true;
+            }
+            *out++ = static_cast<typename string_t::value_type>(current);
+        }
+    }
+
+    /*!
+    @brief Parses a zero-terminated string of length @a len from the BSON
+           input.
+    @param[in] len  The length (including the zero-byte at the end) of the
+                    string to be read.
+    @param[in,out] result  A reference to the string variable where the read
+                            string is to be stored.
+    @tparam NumberType The type of the length @a len
+    @pre len >= 1
+    @return `true` if the string was successfully parsed
+    */
+    template<typename NumberType>
+    bool get_bson_string(const NumberType len, string_t& result)
+    {
+        if (JSON_HEDLEY_UNLIKELY(len < 1))
+        {
+            auto last_token = get_token_string();
+            return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                    exception_message(input_format_t::bson, concat("string length must be at least 1, is ", std::to_string(len)), "string"), nullptr));
+        }
+
+        return get_string(input_format_t::bson, len - static_cast<NumberType>(1), result) && get() != char_traits<char_type>::eof();
+    }
+
+    /*!
+    @brief Parses a byte array input of length @a len from the BSON input.
+    @param[in] len  The length of the byte array to be read.
+    @param[in,out] result  A reference to the binary variable where the read
+                            array is to be stored.
+    @tparam NumberType The type of the length @a len
+    @pre len >= 0
+    @return `true` if the byte array was successfully parsed
+    */
+    template<typename NumberType>
+    bool get_bson_binary(const NumberType len, binary_t& result)
+    {
+        if (JSON_HEDLEY_UNLIKELY(len < 0))
+        {
+            auto last_token = get_token_string();
+            return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                    exception_message(input_format_t::bson, concat("byte array length cannot be negative, is ", std::to_string(len)), "binary"), nullptr));
+        }
+
+        // All BSON binary values have a subtype
+        std::uint8_t subtype{};
+        get_number<std::uint8_t>(input_format_t::bson, subtype);
+        result.set_subtype(subtype);
+
+        return get_binary(input_format_t::bson, len, result);
+    }
+
+    /*!
+    @brief Read a BSON document element of the given @a element_type.
+    @param[in] element_type The BSON element type, c.f. http://bsonspec.org/spec.html
+    @param[in] element_type_parse_position The position in the input stream,
+               where the `element_type` was read.
+    @warning Not all BSON element types are supported yet. An unsupported
+             @a element_type will give rise to a parse_error.114:
+             Unsupported BSON record type 0x...
+    @return whether a valid BSON-object/array was passed to the SAX parser
+    */
+    bool parse_bson_element_internal(const char_int_type element_type,
+                                     const std::size_t element_type_parse_position)
+    {
+        switch (element_type)
+        {
+            case 0x01: // double
+            {
+                double number{};
+                return get_number<double, true>(input_format_t::bson, number) && sax->number_float(static_cast<number_float_t>(number), "");
+            }
+
+            case 0x02: // string
+            {
+                std::int32_t len{};
+                string_t value;
+                return get_number<std::int32_t, true>(input_format_t::bson, len) && get_bson_string(len, value) && sax->string(value);
+            }
+
+            case 0x03: // object
+            {
+                return parse_bson_internal();
+            }
+
+            case 0x04: // array
+            {
+                return parse_bson_array();
+            }
+
+            case 0x05: // binary
+            {
+                std::int32_t len{};
+                binary_t value;
+                return get_number<std::int32_t, true>(input_format_t::bson, len) && get_bson_binary(len, value) && sax->binary(value);
+            }
+
+            case 0x08: // boolean
+            {
+                return sax->boolean(get() != 0);
+            }
+
+            case 0x0A: // null
+            {
+                return sax->null();
+            }
+
+            case 0x10: // int32
+            {
+                std::int32_t value{};
+                return get_number<std::int32_t, true>(input_format_t::bson, value) && sax->number_integer(value);
+            }
+
+            case 0x12: // int64
+            {
+                std::int64_t value{};
+                return get_number<std::int64_t, true>(input_format_t::bson, value) && sax->number_integer(value);
+            }
+
+            case 0x11: // uint64
+            {
+                std::uint64_t value{};
+                return get_number<std::uint64_t, true>(input_format_t::bson, value) && sax->number_unsigned(value);
+            }
+
+            default: // anything else not supported (yet)
+            {
+                std::array<char, 3> cr{{}};
+                static_cast<void>((std::snprintf)(cr.data(), cr.size(), "%.2hhX", static_cast<unsigned char>(element_type))); // NOLINT(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+                const std::string cr_str{cr.data()};
+                return sax->parse_error(element_type_parse_position, cr_str,
+                                        parse_error::create(114, element_type_parse_position, concat("Unsupported BSON record type 0x", cr_str), nullptr));
+            }
+        }
+    }
+
+    /*!
+    @brief Read a BSON element list (as specified in the BSON-spec)
+
+    The same binary layout is used for objects and arrays, hence it must be
+    indicated with the argument @a is_array which one is expected
+    (true --> array, false --> object).
+
+    @param[in] is_array Determines if the element list being read is to be
+                        treated as an object (@a is_array == false), or as an
+                        array (@a is_array == true).
+    @return whether a valid BSON-object/array was passed to the SAX parser
+    */
+    bool parse_bson_element_list(const bool is_array)
+    {
+        string_t key;
+
+        while (auto element_type = get())
+        {
+            if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format_t::bson, "element list")))
+            {
+                return false;
+            }
+
+            const std::size_t element_type_parse_position = chars_read;
+            if (JSON_HEDLEY_UNLIKELY(!get_bson_cstr(key)))
+            {
+                return false;
+            }
+
+            if (!is_array && !sax->key(key))
+            {
+                return false;
+            }
+
+            if (JSON_HEDLEY_UNLIKELY(!parse_bson_element_internal(element_type, element_type_parse_position)))
+            {
+                return false;
+            }
+
+            // get_bson_cstr only appends
+            key.clear();
+        }
+
+        return true;
+    }
+
+    /*!
+    @brief Reads an array from the BSON input and passes it to the SAX-parser.
+    @return whether a valid BSON-array was passed to the SAX parser
+    */
+    bool parse_bson_array()
+    {
+        std::int32_t document_size{};
+        get_number<std::int32_t, true>(input_format_t::bson, document_size);
+
+        if (JSON_HEDLEY_UNLIKELY(!sax->start_array(detail::unknown_size())))
+        {
+            return false;
+        }
+
+        if (JSON_HEDLEY_UNLIKELY(!parse_bson_element_list(/*is_array*/true)))
+        {
+            return false;
+        }
+
+        return sax->end_array();
+    }
+
+    //////////
+    // CBOR //
+    //////////
+
+    /*!
+    @param[in] get_char  whether a new character should be retrieved from the
+                         input (true) or whether the last read character should
+                         be considered instead (false)
+    @param[in] tag_handler how CBOR tags should be treated
+
+    @return whether a valid CBOR value was passed to the SAX parser
+    */
+    bool parse_cbor_internal(const bool get_char,
+                             const cbor_tag_handler_t tag_handler)
+    {
+        switch (get_char ? get() : current)
+        {
+            // EOF
+            case char_traits<char_type>::eof():
+                return unexpect_eof(input_format_t::cbor, "value");
+
+            // Integer 0x00..0x17 (0..23)
+            case 0x00:
+            case 0x01:
+            case 0x02:
+            case 0x03:
+            case 0x04:
+            case 0x05:
+            case 0x06:
+            case 0x07:
+            case 0x08:
+            case 0x09:
+            case 0x0A:
+            case 0x0B:
+            case 0x0C:
+            case 0x0D:
+            case 0x0E:
+            case 0x0F:
+            case 0x10:
+            case 0x11:
+            case 0x12:
+            case 0x13:
+            case 0x14:
+            case 0x15:
+            case 0x16:
+            case 0x17:
+                return sax->number_unsigned(static_cast<number_unsigned_t>(current));
+
+            case 0x18: // Unsigned integer (one-byte uint8_t follows)
+            {
+                std::uint8_t number{};
+                return get_number(input_format_t::cbor, number) && sax->number_unsigned(number);
+            }
+
+            case 0x19: // Unsigned integer (two-byte uint16_t follows)
+            {
+                std::uint16_t number{};
+                return get_number(input_format_t::cbor, number) && sax->number_unsigned(number);
+            }
+
+            case 0x1A: // Unsigned integer (four-byte uint32_t follows)
+            {
+                std::uint32_t number{};
+                return get_number(input_format_t::cbor, number) && sax->number_unsigned(number);
+            }
+
+            case 0x1B: // Unsigned integer (eight-byte uint64_t follows)
+            {
+                std::uint64_t number{};
+                return get_number(input_format_t::cbor, number) && sax->number_unsigned(number);
+            }
+
+            // Negative integer -1-0x00..-1-0x17 (-1..-24)
+            case 0x20:
+            case 0x21:
+            case 0x22:
+            case 0x23:
+            case 0x24:
+            case 0x25:
+            case 0x26:
+            case 0x27:
+            case 0x28:
+            case 0x29:
+            case 0x2A:
+            case 0x2B:
+            case 0x2C:
+            case 0x2D:
+            case 0x2E:
+            case 0x2F:
+            case 0x30:
+            case 0x31:
+            case 0x32:
+            case 0x33:
+            case 0x34:
+            case 0x35:
+            case 0x36:
+            case 0x37:
+                return sax->number_integer(static_cast<std::int8_t>(0x20 - 1 - current));
+
+            case 0x38: // Negative integer (one-byte uint8_t follows)
+            {
+                std::uint8_t number{};
+                return get_number(input_format_t::cbor, number) && sax->number_integer(static_cast<number_integer_t>(-1) - number);
+            }
+
+            case 0x39: // Negative integer -1-n (two-byte uint16_t follows)
+            {
+                std::uint16_t number{};
+                return get_number(input_format_t::cbor, number) && sax->number_integer(static_cast<number_integer_t>(-1) - number);
+            }
+
+            case 0x3A: // Negative integer -1-n (four-byte uint32_t follows)
+            {
+                std::uint32_t number{};
+                return get_number(input_format_t::cbor, number) && sax->number_integer(static_cast<number_integer_t>(-1) - number);
+            }
+
+            case 0x3B: // Negative integer -1-n (eight-byte uint64_t follows)
+            {
+                std::uint64_t number{};
+                return get_number(input_format_t::cbor, number) && sax->number_integer(static_cast<number_integer_t>(-1)
+                        - static_cast<number_integer_t>(number));
+            }
+
+            // Binary data (0x00..0x17 bytes follow)
+            case 0x40:
+            case 0x41:
+            case 0x42:
+            case 0x43:
+            case 0x44:
+            case 0x45:
+            case 0x46:
+            case 0x47:
+            case 0x48:
+            case 0x49:
+            case 0x4A:
+            case 0x4B:
+            case 0x4C:
+            case 0x4D:
+            case 0x4E:
+            case 0x4F:
+            case 0x50:
+            case 0x51:
+            case 0x52:
+            case 0x53:
+            case 0x54:
+            case 0x55:
+            case 0x56:
+            case 0x57:
+            case 0x58: // Binary data (one-byte uint8_t for n follows)
+            case 0x59: // Binary data (two-byte uint16_t for n follow)
+            case 0x5A: // Binary data (four-byte uint32_t for n follow)
+            case 0x5B: // Binary data (eight-byte uint64_t for n follow)
+            case 0x5F: // Binary data (indefinite length)
+            {
+                binary_t b;
+                return get_cbor_binary(b) && sax->binary(b);
+            }
+
+            // UTF-8 string (0x00..0x17 bytes follow)
+            case 0x60:
+            case 0x61:
+            case 0x62:
+            case 0x63:
+            case 0x64:
+            case 0x65:
+            case 0x66:
+            case 0x67:
+            case 0x68:
+            case 0x69:
+            case 0x6A:
+            case 0x6B:
+            case 0x6C:
+            case 0x6D:
+            case 0x6E:
+            case 0x6F:
+            case 0x70:
+            case 0x71:
+            case 0x72:
+            case 0x73:
+            case 0x74:
+            case 0x75:
+            case 0x76:
+            case 0x77:
+            case 0x78: // UTF-8 string (one-byte uint8_t for n follows)
+            case 0x79: // UTF-8 string (two-byte uint16_t for n follow)
+            case 0x7A: // UTF-8 string (four-byte uint32_t for n follow)
+            case 0x7B: // UTF-8 string (eight-byte uint64_t for n follow)
+            case 0x7F: // UTF-8 string (indefinite length)
+            {
+                string_t s;
+                return get_cbor_string(s) && sax->string(s);
+            }
+
+            // array (0x00..0x17 data items follow)
+            case 0x80:
+            case 0x81:
+            case 0x82:
+            case 0x83:
+            case 0x84:
+            case 0x85:
+            case 0x86:
+            case 0x87:
+            case 0x88:
+            case 0x89:
+            case 0x8A:
+            case 0x8B:
+            case 0x8C:
+            case 0x8D:
+            case 0x8E:
+            case 0x8F:
+            case 0x90:
+            case 0x91:
+            case 0x92:
+            case 0x93:
+            case 0x94:
+            case 0x95:
+            case 0x96:
+            case 0x97:
+                return get_cbor_array(
+                           conditional_static_cast<std::size_t>(static_cast<unsigned int>(current) & 0x1Fu), tag_handler);
+
+            case 0x98: // array (one-byte uint8_t for n follows)
+            {
+                std::uint8_t len{};
+                return get_number(input_format_t::cbor, len) && get_cbor_array(static_cast<std::size_t>(len), tag_handler);
+            }
+
+            case 0x99: // array (two-byte uint16_t for n follow)
+            {
+                std::uint16_t len{};
+                return get_number(input_format_t::cbor, len) && get_cbor_array(static_cast<std::size_t>(len), tag_handler);
+            }
+
+            case 0x9A: // array (four-byte uint32_t for n follow)
+            {
+                std::uint32_t len{};
+                return get_number(input_format_t::cbor, len) && get_cbor_array(conditional_static_cast<std::size_t>(len), tag_handler);
+            }
+
+            case 0x9B: // array (eight-byte uint64_t for n follow)
+            {
+                std::uint64_t len{};
+                return get_number(input_format_t::cbor, len) && get_cbor_array(conditional_static_cast<std::size_t>(len), tag_handler);
+            }
+
+            case 0x9F: // array (indefinite length)
+                return get_cbor_array(detail::unknown_size(), tag_handler);
+
+            // map (0x00..0x17 pairs of data items follow)
+            case 0xA0:
+            case 0xA1:
+            case 0xA2:
+            case 0xA3:
+            case 0xA4:
+            case 0xA5:
+            case 0xA6:
+            case 0xA7:
+            case 0xA8:
+            case 0xA9:
+            case 0xAA:
+            case 0xAB:
+            case 0xAC:
+            case 0xAD:
+            case 0xAE:
+            case 0xAF:
+            case 0xB0:
+            case 0xB1:
+            case 0xB2:
+            case 0xB3:
+            case 0xB4:
+            case 0xB5:
+            case 0xB6:
+            case 0xB7:
+                return get_cbor_object(conditional_static_cast<std::size_t>(static_cast<unsigned int>(current) & 0x1Fu), tag_handler);
+
+            case 0xB8: // map (one-byte uint8_t for n follows)
+            {
+                std::uint8_t len{};
+                return get_number(input_format_t::cbor, len) && get_cbor_object(static_cast<std::size_t>(len), tag_handler);
+            }
+
+            case 0xB9: // map (two-byte uint16_t for n follow)
+            {
+                std::uint16_t len{};
+                return get_number(input_format_t::cbor, len) && get_cbor_object(static_cast<std::size_t>(len), tag_handler);
+            }
+
+            case 0xBA: // map (four-byte uint32_t for n follow)
+            {
+                std::uint32_t len{};
+                return get_number(input_format_t::cbor, len) && get_cbor_object(conditional_static_cast<std::size_t>(len), tag_handler);
+            }
+
+            case 0xBB: // map (eight-byte uint64_t for n follow)
+            {
+                std::uint64_t len{};
+                return get_number(input_format_t::cbor, len) && get_cbor_object(conditional_static_cast<std::size_t>(len), tag_handler);
+            }
+
+            case 0xBF: // map (indefinite length)
+                return get_cbor_object(detail::unknown_size(), tag_handler);
+
+            case 0xC6: // tagged item
+            case 0xC7:
+            case 0xC8:
+            case 0xC9:
+            case 0xCA:
+            case 0xCB:
+            case 0xCC:
+            case 0xCD:
+            case 0xCE:
+            case 0xCF:
+            case 0xD0:
+            case 0xD1:
+            case 0xD2:
+            case 0xD3:
+            case 0xD4:
+            case 0xD8: // tagged item (1 bytes follow)
+            case 0xD9: // tagged item (2 bytes follow)
+            case 0xDA: // tagged item (4 bytes follow)
+            case 0xDB: // tagged item (8 bytes follow)
+            {
+                switch (tag_handler)
+                {
+                    case cbor_tag_handler_t::error:
+                    {
+                        auto last_token = get_token_string();
+                        return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                                exception_message(input_format_t::cbor, concat("invalid byte: 0x", last_token), "value"), nullptr));
+                    }
+
+                    case cbor_tag_handler_t::ignore:
+                    {
+                        // ignore binary subtype
+                        switch (current)
+                        {
+                            case 0xD8:
+                            {
+                                std::uint8_t subtype_to_ignore{};
+                                get_number(input_format_t::cbor, subtype_to_ignore);
+                                break;
+                            }
+                            case 0xD9:
+                            {
+                                std::uint16_t subtype_to_ignore{};
+                                get_number(input_format_t::cbor, subtype_to_ignore);
+                                break;
+                            }
+                            case 0xDA:
+                            {
+                                std::uint32_t subtype_to_ignore{};
+                                get_number(input_format_t::cbor, subtype_to_ignore);
+                                break;
+                            }
+                            case 0xDB:
+                            {
+                                std::uint64_t subtype_to_ignore{};
+                                get_number(input_format_t::cbor, subtype_to_ignore);
+                                break;
+                            }
+                            default:
+                                break;
+                        }
+                        return parse_cbor_internal(true, tag_handler);
+                    }
+
+                    case cbor_tag_handler_t::store:
+                    {
+                        binary_t b;
+                        // use binary subtype and store in binary container
+                        switch (current)
+                        {
+                            case 0xD8:
+                            {
+                                std::uint8_t subtype{};
+                                get_number(input_format_t::cbor, subtype);
+                                b.set_subtype(detail::conditional_static_cast<typename binary_t::subtype_type>(subtype));
+                                break;
+                            }
+                            case 0xD9:
+                            {
+                                std::uint16_t subtype{};
+                                get_number(input_format_t::cbor, subtype);
+                                b.set_subtype(detail::conditional_static_cast<typename binary_t::subtype_type>(subtype));
+                                break;
+                            }
+                            case 0xDA:
+                            {
+                                std::uint32_t subtype{};
+                                get_number(input_format_t::cbor, subtype);
+                                b.set_subtype(detail::conditional_static_cast<typename binary_t::subtype_type>(subtype));
+                                break;
+                            }
+                            case 0xDB:
+                            {
+                                std::uint64_t subtype{};
+                                get_number(input_format_t::cbor, subtype);
+                                b.set_subtype(detail::conditional_static_cast<typename binary_t::subtype_type>(subtype));
+                                break;
+                            }
+                            default:
+                                return parse_cbor_internal(true, tag_handler);
+                        }
+                        get();
+                        return get_cbor_binary(b) && sax->binary(b);
+                    }
+
+                    default:                 // LCOV_EXCL_LINE
+                        JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+                        return false;        // LCOV_EXCL_LINE
+                }
+            }
+
+            case 0xF4: // false
+                return sax->boolean(false);
+
+            case 0xF5: // true
+                return sax->boolean(true);
+
+            case 0xF6: // null
+                return sax->null();
+
+            case 0xF9: // Half-Precision Float (two-byte IEEE 754)
+            {
+                const auto byte1_raw = get();
+                if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format_t::cbor, "number")))
+                {
+                    return false;
+                }
+                const auto byte2_raw = get();
+                if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format_t::cbor, "number")))
+                {
+                    return false;
+                }
+
+                const auto byte1 = static_cast<unsigned char>(byte1_raw);
+                const auto byte2 = static_cast<unsigned char>(byte2_raw);
+
+                // code from RFC 7049, Appendix D, Figure 3:
+                // As half-precision floating-point numbers were only added
+                // to IEEE 754 in 2008, today's programming platforms often
+                // still only have limited support for them. It is very
+                // easy to include at least decoding support for them even
+                // without such support. An example of a small decoder for
+                // half-precision floating-point numbers in the C language
+                // is shown in Fig. 3.
+                const auto half = static_cast<unsigned int>((byte1 << 8u) + byte2);
+                const double val = [&half]
+                {
+                    const int exp = (half >> 10u) & 0x1Fu;
+                    const unsigned int mant = half & 0x3FFu;
+                    JSON_ASSERT(0 <= exp&& exp <= 32);
+                    JSON_ASSERT(mant <= 1024);
+                    switch (exp)
+                    {
+                        case 0:
+                            return std::ldexp(mant, -24);
+                        case 31:
+                            return (mant == 0)
+                            ? std::numeric_limits<double>::infinity()
+                            : std::numeric_limits<double>::quiet_NaN();
+                        default:
+                            return std::ldexp(mant + 1024, exp - 25);
+                    }
+                }();
+                return sax->number_float((half & 0x8000u) != 0
+                                         ? static_cast<number_float_t>(-val)
+                                         : static_cast<number_float_t>(val), "");
+            }
+
+            case 0xFA: // Single-Precision Float (four-byte IEEE 754)
+            {
+                float number{};
+                return get_number(input_format_t::cbor, number) && sax->number_float(static_cast<number_float_t>(number), "");
+            }
+
+            case 0xFB: // Double-Precision Float (eight-byte IEEE 754)
+            {
+                double number{};
+                return get_number(input_format_t::cbor, number) && sax->number_float(static_cast<number_float_t>(number), "");
+            }
+
+            default: // anything else (0xFF is handled inside the other types)
+            {
+                auto last_token = get_token_string();
+                return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                        exception_message(input_format_t::cbor, concat("invalid byte: 0x", last_token), "value"), nullptr));
+            }
+        }
+    }
+
+    /*!
+    @brief reads a CBOR string
+
+    This function first reads starting bytes to determine the expected
+    string length and then copies this number of bytes into a string.
+    Additionally, CBOR's strings with indefinite lengths are supported.
+
+    @param[out] result  created string
+
+    @return whether string creation completed
+    */
+    bool get_cbor_string(string_t& result)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format_t::cbor, "string")))
+        {
+            return false;
+        }
+
+        switch (current)
+        {
+            // UTF-8 string (0x00..0x17 bytes follow)
+            case 0x60:
+            case 0x61:
+            case 0x62:
+            case 0x63:
+            case 0x64:
+            case 0x65:
+            case 0x66:
+            case 0x67:
+            case 0x68:
+            case 0x69:
+            case 0x6A:
+            case 0x6B:
+            case 0x6C:
+            case 0x6D:
+            case 0x6E:
+            case 0x6F:
+            case 0x70:
+            case 0x71:
+            case 0x72:
+            case 0x73:
+            case 0x74:
+            case 0x75:
+            case 0x76:
+            case 0x77:
+            {
+                return get_string(input_format_t::cbor, static_cast<unsigned int>(current) & 0x1Fu, result);
+            }
+
+            case 0x78: // UTF-8 string (one-byte uint8_t for n follows)
+            {
+                std::uint8_t len{};
+                return get_number(input_format_t::cbor, len) && get_string(input_format_t::cbor, len, result);
+            }
+
+            case 0x79: // UTF-8 string (two-byte uint16_t for n follow)
+            {
+                std::uint16_t len{};
+                return get_number(input_format_t::cbor, len) && get_string(input_format_t::cbor, len, result);
+            }
+
+            case 0x7A: // UTF-8 string (four-byte uint32_t for n follow)
+            {
+                std::uint32_t len{};
+                return get_number(input_format_t::cbor, len) && get_string(input_format_t::cbor, len, result);
+            }
+
+            case 0x7B: // UTF-8 string (eight-byte uint64_t for n follow)
+            {
+                std::uint64_t len{};
+                return get_number(input_format_t::cbor, len) && get_string(input_format_t::cbor, len, result);
+            }
+
+            case 0x7F: // UTF-8 string (indefinite length)
+            {
+                while (get() != 0xFF)
+                {
+                    string_t chunk;
+                    if (!get_cbor_string(chunk))
+                    {
+                        return false;
+                    }
+                    result.append(chunk);
+                }
+                return true;
+            }
+
+            default:
+            {
+                auto last_token = get_token_string();
+                return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read,
+                                        exception_message(input_format_t::cbor, concat("expected length specification (0x60-0x7B) or indefinite string type (0x7F); last byte: 0x", last_token), "string"), nullptr));
+            }
+        }
+    }
+
+    /*!
+    @brief reads a CBOR byte array
+
+    This function first reads starting bytes to determine the expected
+    byte array length and then copies this number of bytes into the byte array.
+    Additionally, CBOR's byte arrays with indefinite lengths are supported.
+
+    @param[out] result  created byte array
+
+    @return whether byte array creation completed
+    */
+    bool get_cbor_binary(binary_t& result)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format_t::cbor, "binary")))
+        {
+            return false;
+        }
+
+        switch (current)
+        {
+            // Binary data (0x00..0x17 bytes follow)
+            case 0x40:
+            case 0x41:
+            case 0x42:
+            case 0x43:
+            case 0x44:
+            case 0x45:
+            case 0x46:
+            case 0x47:
+            case 0x48:
+            case 0x49:
+            case 0x4A:
+            case 0x4B:
+            case 0x4C:
+            case 0x4D:
+            case 0x4E:
+            case 0x4F:
+            case 0x50:
+            case 0x51:
+            case 0x52:
+            case 0x53:
+            case 0x54:
+            case 0x55:
+            case 0x56:
+            case 0x57:
+            {
+                return get_binary(input_format_t::cbor, static_cast<unsigned int>(current) & 0x1Fu, result);
+            }
+
+            case 0x58: // Binary data (one-byte uint8_t for n follows)
+            {
+                std::uint8_t len{};
+                return get_number(input_format_t::cbor, len) &&
+                       get_binary(input_format_t::cbor, len, result);
+            }
+
+            case 0x59: // Binary data (two-byte uint16_t for n follow)
+            {
+                std::uint16_t len{};
+                return get_number(input_format_t::cbor, len) &&
+                       get_binary(input_format_t::cbor, len, result);
+            }
+
+            case 0x5A: // Binary data (four-byte uint32_t for n follow)
+            {
+                std::uint32_t len{};
+                return get_number(input_format_t::cbor, len) &&
+                       get_binary(input_format_t::cbor, len, result);
+            }
+
+            case 0x5B: // Binary data (eight-byte uint64_t for n follow)
+            {
+                std::uint64_t len{};
+                return get_number(input_format_t::cbor, len) &&
+                       get_binary(input_format_t::cbor, len, result);
+            }
+
+            case 0x5F: // Binary data (indefinite length)
+            {
+                while (get() != 0xFF)
+                {
+                    binary_t chunk;
+                    if (!get_cbor_binary(chunk))
+                    {
+                        return false;
+                    }
+                    result.insert(result.end(), chunk.begin(), chunk.end());
+                }
+                return true;
+            }
+
+            default:
+            {
+                auto last_token = get_token_string();
+                return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read,
+                                        exception_message(input_format_t::cbor, concat("expected length specification (0x40-0x5B) or indefinite binary array type (0x5F); last byte: 0x", last_token), "binary"), nullptr));
+            }
+        }
+    }
+
+    /*!
+    @param[in] len  the length of the array or detail::unknown_size() for an
+                    array of indefinite size
+    @param[in] tag_handler how CBOR tags should be treated
+    @return whether array creation completed
+    */
+    bool get_cbor_array(const std::size_t len,
+                        const cbor_tag_handler_t tag_handler)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!sax->start_array(len)))
+        {
+            return false;
+        }
+
+        if (len != detail::unknown_size())
+        {
+            for (std::size_t i = 0; i < len; ++i)
+            {
+                if (JSON_HEDLEY_UNLIKELY(!parse_cbor_internal(true, tag_handler)))
+                {
+                    return false;
+                }
+            }
+        }
+        else
+        {
+            while (get() != 0xFF)
+            {
+                if (JSON_HEDLEY_UNLIKELY(!parse_cbor_internal(false, tag_handler)))
+                {
+                    return false;
+                }
+            }
+        }
+
+        return sax->end_array();
+    }
+
+    /*!
+    @param[in] len  the length of the object or detail::unknown_size() for an
+                    object of indefinite size
+    @param[in] tag_handler how CBOR tags should be treated
+    @return whether object creation completed
+    */
+    bool get_cbor_object(const std::size_t len,
+                         const cbor_tag_handler_t tag_handler)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!sax->start_object(len)))
+        {
+            return false;
+        }
+
+        if (len != 0)
+        {
+            string_t key;
+            if (len != detail::unknown_size())
+            {
+                for (std::size_t i = 0; i < len; ++i)
+                {
+                    get();
+                    if (JSON_HEDLEY_UNLIKELY(!get_cbor_string(key) || !sax->key(key)))
+                    {
+                        return false;
+                    }
+
+                    if (JSON_HEDLEY_UNLIKELY(!parse_cbor_internal(true, tag_handler)))
+                    {
+                        return false;
+                    }
+                    key.clear();
+                }
+            }
+            else
+            {
+                while (get() != 0xFF)
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!get_cbor_string(key) || !sax->key(key)))
+                    {
+                        return false;
+                    }
+
+                    if (JSON_HEDLEY_UNLIKELY(!parse_cbor_internal(true, tag_handler)))
+                    {
+                        return false;
+                    }
+                    key.clear();
+                }
+            }
+        }
+
+        return sax->end_object();
+    }
+
+    /////////////
+    // MsgPack //
+    /////////////
+
+    /*!
+    @return whether a valid MessagePack value was passed to the SAX parser
+    */
+    bool parse_msgpack_internal()
+    {
+        switch (get())
+        {
+            // EOF
+            case char_traits<char_type>::eof():
+                return unexpect_eof(input_format_t::msgpack, "value");
+
+            // positive fixint
+            case 0x00:
+            case 0x01:
+            case 0x02:
+            case 0x03:
+            case 0x04:
+            case 0x05:
+            case 0x06:
+            case 0x07:
+            case 0x08:
+            case 0x09:
+            case 0x0A:
+            case 0x0B:
+            case 0x0C:
+            case 0x0D:
+            case 0x0E:
+            case 0x0F:
+            case 0x10:
+            case 0x11:
+            case 0x12:
+            case 0x13:
+            case 0x14:
+            case 0x15:
+            case 0x16:
+            case 0x17:
+            case 0x18:
+            case 0x19:
+            case 0x1A:
+            case 0x1B:
+            case 0x1C:
+            case 0x1D:
+            case 0x1E:
+            case 0x1F:
+            case 0x20:
+            case 0x21:
+            case 0x22:
+            case 0x23:
+            case 0x24:
+            case 0x25:
+            case 0x26:
+            case 0x27:
+            case 0x28:
+            case 0x29:
+            case 0x2A:
+            case 0x2B:
+            case 0x2C:
+            case 0x2D:
+            case 0x2E:
+            case 0x2F:
+            case 0x30:
+            case 0x31:
+            case 0x32:
+            case 0x33:
+            case 0x34:
+            case 0x35:
+            case 0x36:
+            case 0x37:
+            case 0x38:
+            case 0x39:
+            case 0x3A:
+            case 0x3B:
+            case 0x3C:
+            case 0x3D:
+            case 0x3E:
+            case 0x3F:
+            case 0x40:
+            case 0x41:
+            case 0x42:
+            case 0x43:
+            case 0x44:
+            case 0x45:
+            case 0x46:
+            case 0x47:
+            case 0x48:
+            case 0x49:
+            case 0x4A:
+            case 0x4B:
+            case 0x4C:
+            case 0x4D:
+            case 0x4E:
+            case 0x4F:
+            case 0x50:
+            case 0x51:
+            case 0x52:
+            case 0x53:
+            case 0x54:
+            case 0x55:
+            case 0x56:
+            case 0x57:
+            case 0x58:
+            case 0x59:
+            case 0x5A:
+            case 0x5B:
+            case 0x5C:
+            case 0x5D:
+            case 0x5E:
+            case 0x5F:
+            case 0x60:
+            case 0x61:
+            case 0x62:
+            case 0x63:
+            case 0x64:
+            case 0x65:
+            case 0x66:
+            case 0x67:
+            case 0x68:
+            case 0x69:
+            case 0x6A:
+            case 0x6B:
+            case 0x6C:
+            case 0x6D:
+            case 0x6E:
+            case 0x6F:
+            case 0x70:
+            case 0x71:
+            case 0x72:
+            case 0x73:
+            case 0x74:
+            case 0x75:
+            case 0x76:
+            case 0x77:
+            case 0x78:
+            case 0x79:
+            case 0x7A:
+            case 0x7B:
+            case 0x7C:
+            case 0x7D:
+            case 0x7E:
+            case 0x7F:
+                return sax->number_unsigned(static_cast<number_unsigned_t>(current));
+
+            // fixmap
+            case 0x80:
+            case 0x81:
+            case 0x82:
+            case 0x83:
+            case 0x84:
+            case 0x85:
+            case 0x86:
+            case 0x87:
+            case 0x88:
+            case 0x89:
+            case 0x8A:
+            case 0x8B:
+            case 0x8C:
+            case 0x8D:
+            case 0x8E:
+            case 0x8F:
+                return get_msgpack_object(conditional_static_cast<std::size_t>(static_cast<unsigned int>(current) & 0x0Fu));
+
+            // fixarray
+            case 0x90:
+            case 0x91:
+            case 0x92:
+            case 0x93:
+            case 0x94:
+            case 0x95:
+            case 0x96:
+            case 0x97:
+            case 0x98:
+            case 0x99:
+            case 0x9A:
+            case 0x9B:
+            case 0x9C:
+            case 0x9D:
+            case 0x9E:
+            case 0x9F:
+                return get_msgpack_array(conditional_static_cast<std::size_t>(static_cast<unsigned int>(current) & 0x0Fu));
+
+            // fixstr
+            case 0xA0:
+            case 0xA1:
+            case 0xA2:
+            case 0xA3:
+            case 0xA4:
+            case 0xA5:
+            case 0xA6:
+            case 0xA7:
+            case 0xA8:
+            case 0xA9:
+            case 0xAA:
+            case 0xAB:
+            case 0xAC:
+            case 0xAD:
+            case 0xAE:
+            case 0xAF:
+            case 0xB0:
+            case 0xB1:
+            case 0xB2:
+            case 0xB3:
+            case 0xB4:
+            case 0xB5:
+            case 0xB6:
+            case 0xB7:
+            case 0xB8:
+            case 0xB9:
+            case 0xBA:
+            case 0xBB:
+            case 0xBC:
+            case 0xBD:
+            case 0xBE:
+            case 0xBF:
+            case 0xD9: // str 8
+            case 0xDA: // str 16
+            case 0xDB: // str 32
+            {
+                string_t s;
+                return get_msgpack_string(s) && sax->string(s);
+            }
+
+            case 0xC0: // nil
+                return sax->null();
+
+            case 0xC2: // false
+                return sax->boolean(false);
+
+            case 0xC3: // true
+                return sax->boolean(true);
+
+            case 0xC4: // bin 8
+            case 0xC5: // bin 16
+            case 0xC6: // bin 32
+            case 0xC7: // ext 8
+            case 0xC8: // ext 16
+            case 0xC9: // ext 32
+            case 0xD4: // fixext 1
+            case 0xD5: // fixext 2
+            case 0xD6: // fixext 4
+            case 0xD7: // fixext 8
+            case 0xD8: // fixext 16
+            {
+                binary_t b;
+                return get_msgpack_binary(b) && sax->binary(b);
+            }
+
+            case 0xCA: // float 32
+            {
+                float number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_float(static_cast<number_float_t>(number), "");
+            }
+
+            case 0xCB: // float 64
+            {
+                double number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_float(static_cast<number_float_t>(number), "");
+            }
+
+            case 0xCC: // uint 8
+            {
+                std::uint8_t number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_unsigned(number);
+            }
+
+            case 0xCD: // uint 16
+            {
+                std::uint16_t number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_unsigned(number);
+            }
+
+            case 0xCE: // uint 32
+            {
+                std::uint32_t number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_unsigned(number);
+            }
+
+            case 0xCF: // uint 64
+            {
+                std::uint64_t number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_unsigned(number);
+            }
+
+            case 0xD0: // int 8
+            {
+                std::int8_t number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_integer(number);
+            }
+
+            case 0xD1: // int 16
+            {
+                std::int16_t number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_integer(number);
+            }
+
+            case 0xD2: // int 32
+            {
+                std::int32_t number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_integer(number);
+            }
+
+            case 0xD3: // int 64
+            {
+                std::int64_t number{};
+                return get_number(input_format_t::msgpack, number) && sax->number_integer(number);
+            }
+
+            case 0xDC: // array 16
+            {
+                std::uint16_t len{};
+                return get_number(input_format_t::msgpack, len) && get_msgpack_array(static_cast<std::size_t>(len));
+            }
+
+            case 0xDD: // array 32
+            {
+                std::uint32_t len{};
+                return get_number(input_format_t::msgpack, len) && get_msgpack_array(conditional_static_cast<std::size_t>(len));
+            }
+
+            case 0xDE: // map 16
+            {
+                std::uint16_t len{};
+                return get_number(input_format_t::msgpack, len) && get_msgpack_object(static_cast<std::size_t>(len));
+            }
+
+            case 0xDF: // map 32
+            {
+                std::uint32_t len{};
+                return get_number(input_format_t::msgpack, len) && get_msgpack_object(conditional_static_cast<std::size_t>(len));
+            }
+
+            // negative fixint
+            case 0xE0:
+            case 0xE1:
+            case 0xE2:
+            case 0xE3:
+            case 0xE4:
+            case 0xE5:
+            case 0xE6:
+            case 0xE7:
+            case 0xE8:
+            case 0xE9:
+            case 0xEA:
+            case 0xEB:
+            case 0xEC:
+            case 0xED:
+            case 0xEE:
+            case 0xEF:
+            case 0xF0:
+            case 0xF1:
+            case 0xF2:
+            case 0xF3:
+            case 0xF4:
+            case 0xF5:
+            case 0xF6:
+            case 0xF7:
+            case 0xF8:
+            case 0xF9:
+            case 0xFA:
+            case 0xFB:
+            case 0xFC:
+            case 0xFD:
+            case 0xFE:
+            case 0xFF:
+                return sax->number_integer(static_cast<std::int8_t>(current));
+
+            default: // anything else
+            {
+                auto last_token = get_token_string();
+                return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                        exception_message(input_format_t::msgpack, concat("invalid byte: 0x", last_token), "value"), nullptr));
+            }
+        }
+    }
+
+    /*!
+    @brief reads a MessagePack string
+
+    This function first reads starting bytes to determine the expected
+    string length and then copies this number of bytes into a string.
+
+    @param[out] result  created string
+
+    @return whether string creation completed
+    */
+    bool get_msgpack_string(string_t& result)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format_t::msgpack, "string")))
+        {
+            return false;
+        }
+
+        switch (current)
+        {
+            // fixstr
+            case 0xA0:
+            case 0xA1:
+            case 0xA2:
+            case 0xA3:
+            case 0xA4:
+            case 0xA5:
+            case 0xA6:
+            case 0xA7:
+            case 0xA8:
+            case 0xA9:
+            case 0xAA:
+            case 0xAB:
+            case 0xAC:
+            case 0xAD:
+            case 0xAE:
+            case 0xAF:
+            case 0xB0:
+            case 0xB1:
+            case 0xB2:
+            case 0xB3:
+            case 0xB4:
+            case 0xB5:
+            case 0xB6:
+            case 0xB7:
+            case 0xB8:
+            case 0xB9:
+            case 0xBA:
+            case 0xBB:
+            case 0xBC:
+            case 0xBD:
+            case 0xBE:
+            case 0xBF:
+            {
+                return get_string(input_format_t::msgpack, static_cast<unsigned int>(current) & 0x1Fu, result);
+            }
+
+            case 0xD9: // str 8
+            {
+                std::uint8_t len{};
+                return get_number(input_format_t::msgpack, len) && get_string(input_format_t::msgpack, len, result);
+            }
+
+            case 0xDA: // str 16
+            {
+                std::uint16_t len{};
+                return get_number(input_format_t::msgpack, len) && get_string(input_format_t::msgpack, len, result);
+            }
+
+            case 0xDB: // str 32
+            {
+                std::uint32_t len{};
+                return get_number(input_format_t::msgpack, len) && get_string(input_format_t::msgpack, len, result);
+            }
+
+            default:
+            {
+                auto last_token = get_token_string();
+                return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read,
+                                        exception_message(input_format_t::msgpack, concat("expected length specification (0xA0-0xBF, 0xD9-0xDB); last byte: 0x", last_token), "string"), nullptr));
+            }
+        }
+    }
+
+    /*!
+    @brief reads a MessagePack byte array
+
+    This function first reads starting bytes to determine the expected
+    byte array length and then copies this number of bytes into a byte array.
+
+    @param[out] result  created byte array
+
+    @return whether byte array creation completed
+    */
+    bool get_msgpack_binary(binary_t& result)
+    {
+        // helper function to set the subtype
+        auto assign_and_return_true = [&result](std::int8_t subtype)
+        {
+            result.set_subtype(static_cast<std::uint8_t>(subtype));
+            return true;
+        };
+
+        switch (current)
+        {
+            case 0xC4: // bin 8
+            {
+                std::uint8_t len{};
+                return get_number(input_format_t::msgpack, len) &&
+                       get_binary(input_format_t::msgpack, len, result);
+            }
+
+            case 0xC5: // bin 16
+            {
+                std::uint16_t len{};
+                return get_number(input_format_t::msgpack, len) &&
+                       get_binary(input_format_t::msgpack, len, result);
+            }
+
+            case 0xC6: // bin 32
+            {
+                std::uint32_t len{};
+                return get_number(input_format_t::msgpack, len) &&
+                       get_binary(input_format_t::msgpack, len, result);
+            }
+
+            case 0xC7: // ext 8
+            {
+                std::uint8_t len{};
+                std::int8_t subtype{};
+                return get_number(input_format_t::msgpack, len) &&
+                       get_number(input_format_t::msgpack, subtype) &&
+                       get_binary(input_format_t::msgpack, len, result) &&
+                       assign_and_return_true(subtype);
+            }
+
+            case 0xC8: // ext 16
+            {
+                std::uint16_t len{};
+                std::int8_t subtype{};
+                return get_number(input_format_t::msgpack, len) &&
+                       get_number(input_format_t::msgpack, subtype) &&
+                       get_binary(input_format_t::msgpack, len, result) &&
+                       assign_and_return_true(subtype);
+            }
+
+            case 0xC9: // ext 32
+            {
+                std::uint32_t len{};
+                std::int8_t subtype{};
+                return get_number(input_format_t::msgpack, len) &&
+                       get_number(input_format_t::msgpack, subtype) &&
+                       get_binary(input_format_t::msgpack, len, result) &&
+                       assign_and_return_true(subtype);
+            }
+
+            case 0xD4: // fixext 1
+            {
+                std::int8_t subtype{};
+                return get_number(input_format_t::msgpack, subtype) &&
+                       get_binary(input_format_t::msgpack, 1, result) &&
+                       assign_and_return_true(subtype);
+            }
+
+            case 0xD5: // fixext 2
+            {
+                std::int8_t subtype{};
+                return get_number(input_format_t::msgpack, subtype) &&
+                       get_binary(input_format_t::msgpack, 2, result) &&
+                       assign_and_return_true(subtype);
+            }
+
+            case 0xD6: // fixext 4
+            {
+                std::int8_t subtype{};
+                return get_number(input_format_t::msgpack, subtype) &&
+                       get_binary(input_format_t::msgpack, 4, result) &&
+                       assign_and_return_true(subtype);
+            }
+
+            case 0xD7: // fixext 8
+            {
+                std::int8_t subtype{};
+                return get_number(input_format_t::msgpack, subtype) &&
+                       get_binary(input_format_t::msgpack, 8, result) &&
+                       assign_and_return_true(subtype);
+            }
+
+            case 0xD8: // fixext 16
+            {
+                std::int8_t subtype{};
+                return get_number(input_format_t::msgpack, subtype) &&
+                       get_binary(input_format_t::msgpack, 16, result) &&
+                       assign_and_return_true(subtype);
+            }
+
+            default:           // LCOV_EXCL_LINE
+                return false;  // LCOV_EXCL_LINE
+        }
+    }
+
+    /*!
+    @param[in] len  the length of the array
+    @return whether array creation completed
+    */
+    bool get_msgpack_array(const std::size_t len)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!sax->start_array(len)))
+        {
+            return false;
+        }
+
+        for (std::size_t i = 0; i < len; ++i)
+        {
+            if (JSON_HEDLEY_UNLIKELY(!parse_msgpack_internal()))
+            {
+                return false;
+            }
+        }
+
+        return sax->end_array();
+    }
+
+    /*!
+    @param[in] len  the length of the object
+    @return whether object creation completed
+    */
+    bool get_msgpack_object(const std::size_t len)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!sax->start_object(len)))
+        {
+            return false;
+        }
+
+        string_t key;
+        for (std::size_t i = 0; i < len; ++i)
+        {
+            get();
+            if (JSON_HEDLEY_UNLIKELY(!get_msgpack_string(key) || !sax->key(key)))
+            {
+                return false;
+            }
+
+            if (JSON_HEDLEY_UNLIKELY(!parse_msgpack_internal()))
+            {
+                return false;
+            }
+            key.clear();
+        }
+
+        return sax->end_object();
+    }
+
+    ////////////
+    // UBJSON //
+    ////////////
+
+    /*!
+    @param[in] get_char  whether a new character should be retrieved from the
+                         input (true, default) or whether the last read
+                         character should be considered instead
+
+    @return whether a valid UBJSON value was passed to the SAX parser
+    */
+    bool parse_ubjson_internal(const bool get_char = true)
+    {
+        return get_ubjson_value(get_char ? get_ignore_noop() : current);
+    }
+
+    /*!
+    @brief reads a UBJSON string
+
+    This function is either called after reading the 'S' byte explicitly
+    indicating a string, or in case of an object key where the 'S' byte can be
+    left out.
+
+    @param[out] result   created string
+    @param[in] get_char  whether a new character should be retrieved from the
+                         input (true, default) or whether the last read
+                         character should be considered instead
+
+    @return whether string creation completed
+    */
+    bool get_ubjson_string(string_t& result, const bool get_char = true)
+    {
+        if (get_char)
+        {
+            get();  // TODO(niels): may we ignore N here?
+        }
+
+        if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format, "value")))
+        {
+            return false;
+        }
+
+        switch (current)
+        {
+            case 'U':
+            {
+                std::uint8_t len{};
+                return get_number(input_format, len) && get_string(input_format, len, result);
+            }
+
+            case 'i':
+            {
+                std::int8_t len{};
+                return get_number(input_format, len) && get_string(input_format, len, result);
+            }
+
+            case 'I':
+            {
+                std::int16_t len{};
+                return get_number(input_format, len) && get_string(input_format, len, result);
+            }
+
+            case 'l':
+            {
+                std::int32_t len{};
+                return get_number(input_format, len) && get_string(input_format, len, result);
+            }
+
+            case 'L':
+            {
+                std::int64_t len{};
+                return get_number(input_format, len) && get_string(input_format, len, result);
+            }
+
+            case 'u':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint16_t len{};
+                return get_number(input_format, len) && get_string(input_format, len, result);
+            }
+
+            case 'm':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint32_t len{};
+                return get_number(input_format, len) && get_string(input_format, len, result);
+            }
+
+            case 'M':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint64_t len{};
+                return get_number(input_format, len) && get_string(input_format, len, result);
+            }
+
+            default:
+                break;
+        }
+        auto last_token = get_token_string();
+        std::string message;
+
+        if (input_format != input_format_t::bjdata)
+        {
+            message = "expected length type specification (U, i, I, l, L); last byte: 0x" + last_token;
+        }
+        else
+        {
+            message = "expected length type specification (U, i, u, I, m, l, M, L); last byte: 0x" + last_token;
+        }
+        return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read, exception_message(input_format, message, "string"), nullptr));
+    }
+
+    /*!
+    @param[out] dim  an integer vector storing the ND array dimensions
+    @return whether reading ND array size vector is successful
+    */
+    bool get_ubjson_ndarray_size(std::vector<size_t>& dim)
+    {
+        std::pair<std::size_t, char_int_type> size_and_type;
+        size_t dimlen = 0;
+        bool no_ndarray = true;
+
+        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_type(size_and_type, no_ndarray)))
+        {
+            return false;
+        }
+
+        if (size_and_type.first != npos)
+        {
+            if (size_and_type.second != 0)
+            {
+                if (size_and_type.second != 'N')
+                {
+                    for (std::size_t i = 0; i < size_and_type.first; ++i)
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, size_and_type.second)))
+                        {
+                            return false;
+                        }
+                        dim.push_back(dimlen);
+                    }
+                }
+            }
+            else
+            {
+                for (std::size_t i = 0; i < size_and_type.first; ++i)
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray)))
+                    {
+                        return false;
+                    }
+                    dim.push_back(dimlen);
+                }
+            }
+        }
+        else
+        {
+            while (current != ']')
+            {
+                if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_value(dimlen, no_ndarray, current)))
+                {
+                    return false;
+                }
+                dim.push_back(dimlen);
+                get_ignore_noop();
+            }
+        }
+        return true;
+    }
+
+    /*!
+    @param[out] result  determined size
+    @param[in,out] is_ndarray  for input, `true` means already inside an ndarray vector
+                               or ndarray dimension is not allowed; `false` means ndarray
+                               is allowed; for output, `true` means an ndarray is found;
+                               is_ndarray can only return `true` when its initial value
+                               is `false`
+    @param[in] prefix  type marker if already read, otherwise set to 0
+
+    @return whether size determination completed
+    */
+    bool get_ubjson_size_value(std::size_t& result, bool& is_ndarray, char_int_type prefix = 0)
+    {
+        if (prefix == 0)
+        {
+            prefix = get_ignore_noop();
+        }
+
+        switch (prefix)
+        {
+            case 'U':
+            {
+                std::uint8_t number{};
+                if (JSON_HEDLEY_UNLIKELY(!get_number(input_format, number)))
+                {
+                    return false;
+                }
+                result = static_cast<std::size_t>(number);
+                return true;
+            }
+
+            case 'i':
+            {
+                std::int8_t number{};
+                if (JSON_HEDLEY_UNLIKELY(!get_number(input_format, number)))
+                {
+                    return false;
+                }
+                if (number < 0)
+                {
+                    return sax->parse_error(chars_read, get_token_string(), parse_error::create(113, chars_read,
+                                            exception_message(input_format, "count in an optimized container must be positive", "size"), nullptr));
+                }
+                result = static_cast<std::size_t>(number); // NOLINT(bugprone-signed-char-misuse,cert-str34-c): number is not a char
+                return true;
+            }
+
+            case 'I':
+            {
+                std::int16_t number{};
+                if (JSON_HEDLEY_UNLIKELY(!get_number(input_format, number)))
+                {
+                    return false;
+                }
+                if (number < 0)
+                {
+                    return sax->parse_error(chars_read, get_token_string(), parse_error::create(113, chars_read,
+                                            exception_message(input_format, "count in an optimized container must be positive", "size"), nullptr));
+                }
+                result = static_cast<std::size_t>(number);
+                return true;
+            }
+
+            case 'l':
+            {
+                std::int32_t number{};
+                if (JSON_HEDLEY_UNLIKELY(!get_number(input_format, number)))
+                {
+                    return false;
+                }
+                if (number < 0)
+                {
+                    return sax->parse_error(chars_read, get_token_string(), parse_error::create(113, chars_read,
+                                            exception_message(input_format, "count in an optimized container must be positive", "size"), nullptr));
+                }
+                result = static_cast<std::size_t>(number);
+                return true;
+            }
+
+            case 'L':
+            {
+                std::int64_t number{};
+                if (JSON_HEDLEY_UNLIKELY(!get_number(input_format, number)))
+                {
+                    return false;
+                }
+                if (number < 0)
+                {
+                    return sax->parse_error(chars_read, get_token_string(), parse_error::create(113, chars_read,
+                                            exception_message(input_format, "count in an optimized container must be positive", "size"), nullptr));
+                }
+                if (!value_in_range_of<std::size_t>(number))
+                {
+                    return sax->parse_error(chars_read, get_token_string(), out_of_range::create(408,
+                                            exception_message(input_format, "integer value overflow", "size"), nullptr));
+                }
+                result = static_cast<std::size_t>(number);
+                return true;
+            }
+
+            case 'u':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint16_t number{};
+                if (JSON_HEDLEY_UNLIKELY(!get_number(input_format, number)))
+                {
+                    return false;
+                }
+                result = static_cast<std::size_t>(number);
+                return true;
+            }
+
+            case 'm':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint32_t number{};
+                if (JSON_HEDLEY_UNLIKELY(!get_number(input_format, number)))
+                {
+                    return false;
+                }
+                result = conditional_static_cast<std::size_t>(number);
+                return true;
+            }
+
+            case 'M':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint64_t number{};
+                if (JSON_HEDLEY_UNLIKELY(!get_number(input_format, number)))
+                {
+                    return false;
+                }
+                if (!value_in_range_of<std::size_t>(number))
+                {
+                    return sax->parse_error(chars_read, get_token_string(), out_of_range::create(408,
+                                            exception_message(input_format, "integer value overflow", "size"), nullptr));
+                }
+                result = detail::conditional_static_cast<std::size_t>(number);
+                return true;
+            }
+
+            case '[':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                if (is_ndarray) // ndarray dimensional vector can only contain integers, and can not embed another array
+                {
+                    return sax->parse_error(chars_read, get_token_string(), parse_error::create(113, chars_read, exception_message(input_format, "ndarray dimensional vector is not allowed", "size"), nullptr));
+                }
+                std::vector<size_t> dim;
+                if (JSON_HEDLEY_UNLIKELY(!get_ubjson_ndarray_size(dim)))
+                {
+                    return false;
+                }
+                if (dim.size() == 1 || (dim.size() == 2 && dim.at(0) == 1)) // return normal array size if 1D row vector
+                {
+                    result = dim.at(dim.size() - 1);
+                    return true;
+                }
+                if (!dim.empty())  // if ndarray, convert to an object in JData annotated array format
+                {
+                    for (auto i : dim) // test if any dimension in an ndarray is 0, if so, return a 1D empty container
+                    {
+                        if ( i == 0 )
+                        {
+                            result = 0;
+                            return true;
+                        }
+                    }
+
+                    string_t key = "_ArraySize_";
+                    if (JSON_HEDLEY_UNLIKELY(!sax->start_object(3) || !sax->key(key) || !sax->start_array(dim.size())))
+                    {
+                        return false;
+                    }
+                    result = 1;
+                    for (auto i : dim)
+                    {
+                        result *= i;
+                        if (result == 0 || result == npos) // because dim elements shall not have zeros, result = 0 means overflow happened; it also can't be npos as it is used to initialize size in get_ubjson_size_type()
+                        {
+                            return sax->parse_error(chars_read, get_token_string(), out_of_range::create(408, exception_message(input_format, "excessive ndarray size caused overflow", "size"), nullptr));
+                        }
+                        if (JSON_HEDLEY_UNLIKELY(!sax->number_unsigned(static_cast<number_unsigned_t>(i))))
+                        {
+                            return false;
+                        }
+                    }
+                    is_ndarray = true;
+                    return sax->end_array();
+                }
+                result = 0;
+                return true;
+            }
+
+            default:
+                break;
+        }
+        auto last_token = get_token_string();
+        std::string message;
+
+        if (input_format != input_format_t::bjdata)
+        {
+            message = "expected length type specification (U, i, I, l, L) after '#'; last byte: 0x" + last_token;
+        }
+        else
+        {
+            message = "expected length type specification (U, i, u, I, m, l, M, L) after '#'; last byte: 0x" + last_token;
+        }
+        return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read, exception_message(input_format, message, "size"), nullptr));
+    }
+
+    /*!
+    @brief determine the type and size for a container
+
+    In the optimized UBJSON format, a type and a size can be provided to allow
+    for a more compact representation.
+
+    @param[out] result  pair of the size and the type
+    @param[in] inside_ndarray  whether the parser is parsing an ND array dimensional vector
+
+    @return whether pair creation completed
+    */
+    bool get_ubjson_size_type(std::pair<std::size_t, char_int_type>& result, bool inside_ndarray = false)
+    {
+        result.first = npos; // size
+        result.second = 0; // type
+        bool is_ndarray = false;
+
+        get_ignore_noop();
+
+        if (current == '$')
+        {
+            result.second = get();  // must not ignore 'N', because 'N' maybe the type
+            if (input_format == input_format_t::bjdata
+                    && JSON_HEDLEY_UNLIKELY(std::binary_search(bjd_optimized_type_markers.begin(), bjd_optimized_type_markers.end(), result.second)))
+            {
+                auto last_token = get_token_string();
+                return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                        exception_message(input_format, concat("marker 0x", last_token, " is not a permitted optimized array type"), "type"), nullptr));
+            }
+
+            if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format, "type")))
+            {
+                return false;
+            }
+
+            get_ignore_noop();
+            if (JSON_HEDLEY_UNLIKELY(current != '#'))
+            {
+                if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format, "value")))
+                {
+                    return false;
+                }
+                auto last_token = get_token_string();
+                return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                        exception_message(input_format, concat("expected '#' after type information; last byte: 0x", last_token), "size"), nullptr));
+            }
+
+            const bool is_error = get_ubjson_size_value(result.first, is_ndarray);
+            if (input_format == input_format_t::bjdata && is_ndarray)
+            {
+                if (inside_ndarray)
+                {
+                    return sax->parse_error(chars_read, get_token_string(), parse_error::create(112, chars_read,
+                                            exception_message(input_format, "ndarray can not be recursive", "size"), nullptr));
+                }
+                result.second |= (1 << 8); // use bit 8 to indicate ndarray, all UBJSON and BJData markers should be ASCII letters
+            }
+            return is_error;
+        }
+
+        if (current == '#')
+        {
+            const bool is_error = get_ubjson_size_value(result.first, is_ndarray);
+            if (input_format == input_format_t::bjdata && is_ndarray)
+            {
+                return sax->parse_error(chars_read, get_token_string(), parse_error::create(112, chars_read,
+                                        exception_message(input_format, "ndarray requires both type and size", "size"), nullptr));
+            }
+            return is_error;
+        }
+
+        return true;
+    }
+
+    /*!
+    @param prefix  the previously read or set type prefix
+    @return whether value creation completed
+    */
+    bool get_ubjson_value(const char_int_type prefix)
+    {
+        switch (prefix)
+        {
+            case char_traits<char_type>::eof():  // EOF
+                return unexpect_eof(input_format, "value");
+
+            case 'T':  // true
+                return sax->boolean(true);
+            case 'F':  // false
+                return sax->boolean(false);
+
+            case 'Z':  // null
+                return sax->null();
+
+            case 'B':  // byte
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint8_t number{};
+                return get_number(input_format, number) && sax->number_unsigned(number);
+            }
+
+            case 'U':
+            {
+                std::uint8_t number{};
+                return get_number(input_format, number) && sax->number_unsigned(number);
+            }
+
+            case 'i':
+            {
+                std::int8_t number{};
+                return get_number(input_format, number) && sax->number_integer(number);
+            }
+
+            case 'I':
+            {
+                std::int16_t number{};
+                return get_number(input_format, number) && sax->number_integer(number);
+            }
+
+            case 'l':
+            {
+                std::int32_t number{};
+                return get_number(input_format, number) && sax->number_integer(number);
+            }
+
+            case 'L':
+            {
+                std::int64_t number{};
+                return get_number(input_format, number) && sax->number_integer(number);
+            }
+
+            case 'u':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint16_t number{};
+                return get_number(input_format, number) && sax->number_unsigned(number);
+            }
+
+            case 'm':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint32_t number{};
+                return get_number(input_format, number) && sax->number_unsigned(number);
+            }
+
+            case 'M':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                std::uint64_t number{};
+                return get_number(input_format, number) && sax->number_unsigned(number);
+            }
+
+            case 'h':
+            {
+                if (input_format != input_format_t::bjdata)
+                {
+                    break;
+                }
+                const auto byte1_raw = get();
+                if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format, "number")))
+                {
+                    return false;
+                }
+                const auto byte2_raw = get();
+                if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format, "number")))
+                {
+                    return false;
+                }
+
+                const auto byte1 = static_cast<unsigned char>(byte1_raw);
+                const auto byte2 = static_cast<unsigned char>(byte2_raw);
+
+                // code from RFC 7049, Appendix D, Figure 3:
+                // As half-precision floating-point numbers were only added
+                // to IEEE 754 in 2008, today's programming platforms often
+                // still only have limited support for them. It is very
+                // easy to include at least decoding support for them even
+                // without such support. An example of a small decoder for
+                // half-precision floating-point numbers in the C language
+                // is shown in Fig. 3.
+                const auto half = static_cast<unsigned int>((byte2 << 8u) + byte1);
+                const double val = [&half]
+                {
+                    const int exp = (half >> 10u) & 0x1Fu;
+                    const unsigned int mant = half & 0x3FFu;
+                    JSON_ASSERT(0 <= exp&& exp <= 32);
+                    JSON_ASSERT(mant <= 1024);
+                    switch (exp)
+                    {
+                        case 0:
+                            return std::ldexp(mant, -24);
+                        case 31:
+                            return (mant == 0)
+                            ? std::numeric_limits<double>::infinity()
+                            : std::numeric_limits<double>::quiet_NaN();
+                        default:
+                            return std::ldexp(mant + 1024, exp - 25);
+                    }
+                }();
+                return sax->number_float((half & 0x8000u) != 0
+                                         ? static_cast<number_float_t>(-val)
+                                         : static_cast<number_float_t>(val), "");
+            }
+
+            case 'd':
+            {
+                float number{};
+                return get_number(input_format, number) && sax->number_float(static_cast<number_float_t>(number), "");
+            }
+
+            case 'D':
+            {
+                double number{};
+                return get_number(input_format, number) && sax->number_float(static_cast<number_float_t>(number), "");
+            }
+
+            case 'H':
+            {
+                return get_ubjson_high_precision_number();
+            }
+
+            case 'C':  // char
+            {
+                get();
+                if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format, "char")))
+                {
+                    return false;
+                }
+                if (JSON_HEDLEY_UNLIKELY(current > 127))
+                {
+                    auto last_token = get_token_string();
+                    return sax->parse_error(chars_read, last_token, parse_error::create(113, chars_read,
+                                            exception_message(input_format, concat("byte after 'C' must be in range 0x00..0x7F; last byte: 0x", last_token), "char"), nullptr));
+                }
+                string_t s(1, static_cast<typename string_t::value_type>(current));
+                return sax->string(s);
+            }
+
+            case 'S':  // string
+            {
+                string_t s;
+                return get_ubjson_string(s) && sax->string(s);
+            }
+
+            case '[':  // array
+                return get_ubjson_array();
+
+            case '{':  // object
+                return get_ubjson_object();
+
+            default: // anything else
+                break;
+        }
+        auto last_token = get_token_string();
+        return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read, exception_message(input_format, "invalid byte: 0x" + last_token, "value"), nullptr));
+    }
+
+    /*!
+    @return whether array creation completed
+    */
+    bool get_ubjson_array()
+    {
+        std::pair<std::size_t, char_int_type> size_and_type;
+        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_type(size_and_type)))
+        {
+            return false;
+        }
+
+        // if bit-8 of size_and_type.second is set to 1, encode bjdata ndarray as an object in JData annotated array format (https://github.com/NeuroJSON/jdata):
+        // {"_ArrayType_" : "typeid", "_ArraySize_" : [n1, n2, ...], "_ArrayData_" : [v1, v2, ...]}
+
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
+        {
+            size_and_type.second &= ~(static_cast<char_int_type>(1) << 8);  // use bit 8 to indicate ndarray, here we remove the bit to restore the type marker
+            auto it = std::lower_bound(bjd_types_map.begin(), bjd_types_map.end(), size_and_type.second, [](const bjd_type & p, char_int_type t)
+            {
+                return p.first < t;
+            });
+            string_t key = "_ArrayType_";
+            if (JSON_HEDLEY_UNLIKELY(it == bjd_types_map.end() || it->first != size_and_type.second))
+            {
+                auto last_token = get_token_string();
+                return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                        exception_message(input_format, "invalid byte: 0x" + last_token, "type"), nullptr));
+            }
+
+            string_t type = it->second; // sax->string() takes a reference
+            if (JSON_HEDLEY_UNLIKELY(!sax->key(key) || !sax->string(type)))
+            {
+                return false;
+            }
+
+            if (size_and_type.second == 'C' || size_and_type.second == 'B')
+            {
+                size_and_type.second = 'U';
+            }
+
+            key = "_ArrayData_";
+            if (JSON_HEDLEY_UNLIKELY(!sax->key(key) || !sax->start_array(size_and_type.first) ))
+            {
+                return false;
+            }
+
+            for (std::size_t i = 0; i < size_and_type.first; ++i)
+            {
+                if (JSON_HEDLEY_UNLIKELY(!get_ubjson_value(size_and_type.second)))
+                {
+                    return false;
+                }
+            }
+
+            return (sax->end_array() && sax->end_object());
+        }
+
+        // If BJData type marker is 'B' decode as binary
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && size_and_type.second == 'B')
+        {
+            binary_t result;
+            return get_binary(input_format, size_and_type.first, result) && sax->binary(result);
+        }
+
+        if (size_and_type.first != npos)
+        {
+            if (JSON_HEDLEY_UNLIKELY(!sax->start_array(size_and_type.first)))
+            {
+                return false;
+            }
+
+            if (size_and_type.second != 0)
+            {
+                if (size_and_type.second != 'N')
+                {
+                    for (std::size_t i = 0; i < size_and_type.first; ++i)
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_value(size_and_type.second)))
+                        {
+                            return false;
+                        }
+                    }
+                }
+            }
+            else
+            {
+                for (std::size_t i = 0; i < size_and_type.first; ++i)
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!parse_ubjson_internal()))
+                    {
+                        return false;
+                    }
+                }
+            }
+        }
+        else
+        {
+            if (JSON_HEDLEY_UNLIKELY(!sax->start_array(detail::unknown_size())))
+            {
+                return false;
+            }
+
+            while (current != ']')
+            {
+                if (JSON_HEDLEY_UNLIKELY(!parse_ubjson_internal(false)))
+                {
+                    return false;
+                }
+                get_ignore_noop();
+            }
+        }
+
+        return sax->end_array();
+    }
+
+    /*!
+    @return whether object creation completed
+    */
+    bool get_ubjson_object()
+    {
+        std::pair<std::size_t, char_int_type> size_and_type;
+        if (JSON_HEDLEY_UNLIKELY(!get_ubjson_size_type(size_and_type)))
+        {
+            return false;
+        }
+
+        // do not accept ND-array size in objects in BJData
+        if (input_format == input_format_t::bjdata && size_and_type.first != npos && (size_and_type.second & (1 << 8)) != 0)
+        {
+            auto last_token = get_token_string();
+            return sax->parse_error(chars_read, last_token, parse_error::create(112, chars_read,
+                                    exception_message(input_format, "BJData object does not support ND-array size in optimized format", "object"), nullptr));
+        }
+
+        string_t key;
+        if (size_and_type.first != npos)
+        {
+            if (JSON_HEDLEY_UNLIKELY(!sax->start_object(size_and_type.first)))
+            {
+                return false;
+            }
+
+            if (size_and_type.second != 0)
+            {
+                for (std::size_t i = 0; i < size_and_type.first; ++i)
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!get_ubjson_string(key) || !sax->key(key)))
+                    {
+                        return false;
+                    }
+                    if (JSON_HEDLEY_UNLIKELY(!get_ubjson_value(size_and_type.second)))
+                    {
+                        return false;
+                    }
+                    key.clear();
+                }
+            }
+            else
+            {
+                for (std::size_t i = 0; i < size_and_type.first; ++i)
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!get_ubjson_string(key) || !sax->key(key)))
+                    {
+                        return false;
+                    }
+                    if (JSON_HEDLEY_UNLIKELY(!parse_ubjson_internal()))
+                    {
+                        return false;
+                    }
+                    key.clear();
+                }
+            }
+        }
+        else
+        {
+            if (JSON_HEDLEY_UNLIKELY(!sax->start_object(detail::unknown_size())))
+            {
+                return false;
+            }
+
+            while (current != '}')
+            {
+                if (JSON_HEDLEY_UNLIKELY(!get_ubjson_string(key, false) || !sax->key(key)))
+                {
+                    return false;
+                }
+                if (JSON_HEDLEY_UNLIKELY(!parse_ubjson_internal()))
+                {
+                    return false;
+                }
+                get_ignore_noop();
+                key.clear();
+            }
+        }
+
+        return sax->end_object();
+    }
+
+    // Note, no reader for UBJSON binary types is implemented because they do
+    // not exist
+
+    bool get_ubjson_high_precision_number()
+    {
+        // get size of following number string
+        std::size_t size{};
+        bool no_ndarray = true;
+        auto res = get_ubjson_size_value(size, no_ndarray);
+        if (JSON_HEDLEY_UNLIKELY(!res))
+        {
+            return res;
+        }
+
+        // get number string
+        std::vector<char> number_vector;
+        for (std::size_t i = 0; i < size; ++i)
+        {
+            get();
+            if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(input_format, "number")))
+            {
+                return false;
+            }
+            number_vector.push_back(static_cast<char>(current));
+        }
+
+        // parse number string
+        using ia_type = decltype(detail::input_adapter(number_vector));
+        auto number_lexer = detail::lexer<BasicJsonType, ia_type>(detail::input_adapter(number_vector), false);
+        const auto result_number = number_lexer.scan();
+        const auto number_string = number_lexer.get_token_string();
+        const auto result_remainder = number_lexer.scan();
+
+        using token_type = typename detail::lexer_base<BasicJsonType>::token_type;
+
+        if (JSON_HEDLEY_UNLIKELY(result_remainder != token_type::end_of_input))
+        {
+            return sax->parse_error(chars_read, number_string, parse_error::create(115, chars_read,
+                                    exception_message(input_format, concat("invalid number text: ", number_lexer.get_token_string()), "high-precision number"), nullptr));
+        }
+
+        switch (result_number)
+        {
+            case token_type::value_integer:
+                return sax->number_integer(number_lexer.get_number_integer());
+            case token_type::value_unsigned:
+                return sax->number_unsigned(number_lexer.get_number_unsigned());
+            case token_type::value_float:
+                return sax->number_float(number_lexer.get_number_float(), std::move(number_string));
+            case token_type::uninitialized:
+            case token_type::literal_true:
+            case token_type::literal_false:
+            case token_type::literal_null:
+            case token_type::value_string:
+            case token_type::begin_array:
+            case token_type::begin_object:
+            case token_type::end_array:
+            case token_type::end_object:
+            case token_type::name_separator:
+            case token_type::value_separator:
+            case token_type::parse_error:
+            case token_type::end_of_input:
+            case token_type::literal_or_value:
+            default:
+                return sax->parse_error(chars_read, number_string, parse_error::create(115, chars_read,
+                                        exception_message(input_format, concat("invalid number text: ", number_lexer.get_token_string()), "high-precision number"), nullptr));
+        }
+    }
+
+    ///////////////////////
+    // Utility functions //
+    ///////////////////////
+
+    /*!
+    @brief get next character from the input
+
+    This function provides the interface to the used input adapter. It does
+    not throw in case the input reached EOF, but returns a -'ve valued
+    `char_traits<char_type>::eof()` in that case.
+
+    @return character read from the input
+    */
+    char_int_type get()
+    {
+        ++chars_read;
+        return current = ia.get_character();
+    }
+
+    /*!
+    @brief get_to read into a primitive type
+
+    This function provides the interface to the used input adapter. It does
+    not throw in case the input reached EOF, but returns false instead
+
+    @return bool, whether the read was successful
+    */
+    template<class T>
+    bool get_to(T& dest, const input_format_t format, const char* context)
+    {
+        auto new_chars_read = ia.get_elements(&dest);
+        chars_read += new_chars_read;
+        if (JSON_HEDLEY_UNLIKELY(new_chars_read < sizeof(T)))
+        {
+            // in case of failure, advance position by 1 to report failing location
+            ++chars_read;
+            sax->parse_error(chars_read, "<end of file>", parse_error::create(110, chars_read, exception_message(format, "unexpected end of input", context), nullptr));
+            return false;
+        }
+        return true;
+    }
+
+    /*!
+    @return character read from the input after ignoring all 'N' entries
+    */
+    char_int_type get_ignore_noop()
+    {
+        do
+        {
+            get();
+        }
+        while (current == 'N');
+
+        return current;
+    }
+
+    template<class NumberType>
+    static void byte_swap(NumberType& number)
+    {
+        constexpr std::size_t sz = sizeof(number);
+#ifdef __cpp_lib_byteswap
+        if constexpr (sz == 1)
+        {
+            return;
+        }
+        if constexpr(std::is_integral_v<NumberType>)
+        {
+            number = std::byteswap(number);
+            return;
+        }
+#endif
+        auto* ptr = reinterpret_cast<std::uint8_t*>(&number);
+        for (std::size_t i = 0; i < sz / 2; ++i)
+        {
+            std::swap(ptr[i], ptr[sz - i - 1]);
+        }
+    }
+
+    /*
+    @brief read a number from the input
+
+    @tparam NumberType the type of the number
+    @param[in] format   the current format (for diagnostics)
+    @param[out] result  number of type @a NumberType
+
+    @return whether conversion completed
+
+    @note This function needs to respect the system's endianness, because
+          bytes in CBOR, MessagePack, and UBJSON are stored in network order
+          (big endian) and therefore need reordering on little endian systems.
+          On the other hand, BSON and BJData use little endian and should reorder
+          on big endian systems.
+    */
+    template<typename NumberType, bool InputIsLittleEndian = false>
+    bool get_number(const input_format_t format, NumberType& result)
+    {
+        // read in the original format
+
+        if (JSON_HEDLEY_UNLIKELY(!get_to(result, format, "number")))
+        {
+            return false;
+        }
+        if (is_little_endian != (InputIsLittleEndian || format == input_format_t::bjdata))
+        {
+            byte_swap(result);
+        }
+        return true;
+    }
+
+    /*!
+    @brief create a string by reading characters from the input
+
+    @tparam NumberType the type of the number
+    @param[in] format the current format (for diagnostics)
+    @param[in] len number of characters to read
+    @param[out] result string created by reading @a len bytes
+
+    @return whether string creation completed
+
+    @note We can not reserve @a len bytes for the result, because @a len
+          may be too large. Usually, @ref unexpect_eof() detects the end of
+          the input before we run out of string memory.
+    */
+    template<typename NumberType>
+    bool get_string(const input_format_t format,
+                    const NumberType len,
+                    string_t& result)
+    {
+        bool success = true;
+        for (NumberType i = 0; i < len; i++)
+        {
+            get();
+            if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(format, "string")))
+            {
+                success = false;
+                break;
+            }
+            result.push_back(static_cast<typename string_t::value_type>(current));
+        }
+        return success;
+    }
+
+    /*!
+    @brief create a byte array by reading bytes from the input
+
+    @tparam NumberType the type of the number
+    @param[in] format the current format (for diagnostics)
+    @param[in] len number of bytes to read
+    @param[out] result byte array created by reading @a len bytes
+
+    @return whether byte array creation completed
+
+    @note We can not reserve @a len bytes for the result, because @a len
+          may be too large. Usually, @ref unexpect_eof() detects the end of
+          the input before we run out of memory.
+    */
+    template<typename NumberType>
+    bool get_binary(const input_format_t format,
+                    const NumberType len,
+                    binary_t& result)
+    {
+        bool success = true;
+        for (NumberType i = 0; i < len; i++)
+        {
+            get();
+            if (JSON_HEDLEY_UNLIKELY(!unexpect_eof(format, "binary")))
+            {
+                success = false;
+                break;
+            }
+            result.push_back(static_cast<std::uint8_t>(current));
+        }
+        return success;
+    }
+
+    /*!
+    @param[in] format   the current format (for diagnostics)
+    @param[in] context  further context information (for diagnostics)
+    @return whether the last read character is not EOF
+    */
+    JSON_HEDLEY_NON_NULL(3)
+    bool unexpect_eof(const input_format_t format, const char* context) const
+    {
+        if (JSON_HEDLEY_UNLIKELY(current == char_traits<char_type>::eof()))
+        {
+            return sax->parse_error(chars_read, "<end of file>",
+                                    parse_error::create(110, chars_read, exception_message(format, "unexpected end of input", context), nullptr));
+        }
+        return true;
+    }
+
+    /*!
+    @return a string representation of the last read byte
+    */
+    std::string get_token_string() const
+    {
+        std::array<char, 3> cr{{}};
+        static_cast<void>((std::snprintf)(cr.data(), cr.size(), "%.2hhX", static_cast<unsigned char>(current))); // NOLINT(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+        return std::string{cr.data()};
+    }
+
+    /*!
+    @param[in] format   the current format
+    @param[in] detail   a detailed error message
+    @param[in] context  further context information
+    @return a message string to use in the parse_error exceptions
+    */
+    std::string exception_message(const input_format_t format,
+                                  const std::string& detail,
+                                  const std::string& context) const
+    {
+        std::string error_msg = "syntax error while parsing ";
+
+        switch (format)
+        {
+            case input_format_t::cbor:
+                error_msg += "CBOR";
+                break;
+
+            case input_format_t::msgpack:
+                error_msg += "MessagePack";
+                break;
+
+            case input_format_t::ubjson:
+                error_msg += "UBJSON";
+                break;
+
+            case input_format_t::bson:
+                error_msg += "BSON";
+                break;
+
+            case input_format_t::bjdata:
+                error_msg += "BJData";
+                break;
+
+            case input_format_t::json: // LCOV_EXCL_LINE
+            default:            // LCOV_EXCL_LINE
+                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+        }
+
+        return concat(error_msg, ' ', context, ": ", detail);
+    }
+
+  private:
+    static JSON_INLINE_VARIABLE constexpr std::size_t npos = detail::unknown_size();
+
+    /// input adapter
+    InputAdapterType ia;
+
+    /// the current character
+    char_int_type current = char_traits<char_type>::eof();
+
+    /// the number of characters read
+    std::size_t chars_read = 0;
+
+    /// whether we can assume little endianness
+    const bool is_little_endian = little_endianness();
+
+    /// input format
+    const input_format_t input_format = input_format_t::json;
+
+    /// the SAX parser
+    json_sax_t* sax = nullptr;
+
+    // excluded markers in bjdata optimized type
+#define JSON_BINARY_READER_MAKE_BJD_OPTIMIZED_TYPE_MARKERS_ \
+    make_array<char_int_type>('F', 'H', 'N', 'S', 'T', 'Z', '[', '{')
+
+#define JSON_BINARY_READER_MAKE_BJD_TYPES_MAP_ \
+    make_array<bjd_type>(                      \
+    bjd_type{'B', "byte"},                     \
+    bjd_type{'C', "char"},                     \
+    bjd_type{'D', "double"},                   \
+    bjd_type{'I', "int16"},                    \
+    bjd_type{'L', "int64"},                    \
+    bjd_type{'M', "uint64"},                   \
+    bjd_type{'U', "uint8"},                    \
+    bjd_type{'d', "single"},                   \
+    bjd_type{'i', "int8"},                     \
+    bjd_type{'l', "int32"},                    \
+    bjd_type{'m', "uint32"},                   \
+    bjd_type{'u', "uint16"})
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    // lookup tables
+    // NOLINTNEXTLINE(cppcoreguidelines-non-private-member-variables-in-classes)
+    const decltype(JSON_BINARY_READER_MAKE_BJD_OPTIMIZED_TYPE_MARKERS_) bjd_optimized_type_markers =
+        JSON_BINARY_READER_MAKE_BJD_OPTIMIZED_TYPE_MARKERS_;
+
+    using bjd_type = std::pair<char_int_type, string_t>;
+    // NOLINTNEXTLINE(cppcoreguidelines-non-private-member-variables-in-classes)
+    const decltype(JSON_BINARY_READER_MAKE_BJD_TYPES_MAP_) bjd_types_map =
+        JSON_BINARY_READER_MAKE_BJD_TYPES_MAP_;
+
+#undef JSON_BINARY_READER_MAKE_BJD_OPTIMIZED_TYPE_MARKERS_
+#undef JSON_BINARY_READER_MAKE_BJD_TYPES_MAP_
+};
+
+#ifndef JSON_HAS_CPP_17
+    template<typename BasicJsonType, typename InputAdapterType, typename SAX>
+    constexpr std::size_t binary_reader<BasicJsonType, InputAdapterType, SAX>::npos;
+#endif
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/input/input_adapters.hpp>
+
+// #include <nlohmann/detail/input/lexer.hpp>
+
+// #include <nlohmann/detail/input/parser.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cmath> // isfinite
+#include <cstdint> // uint8_t
+#include <functional> // function
+#include <string> // string
+#include <utility> // move
+#include <vector> // vector
+
+// #include <nlohmann/detail/exceptions.hpp>
+
+// #include <nlohmann/detail/input/input_adapters.hpp>
+
+// #include <nlohmann/detail/input/json_sax.hpp>
+
+// #include <nlohmann/detail/input/lexer.hpp>
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/is_sax.hpp>
+
+// #include <nlohmann/detail/string_concat.hpp>
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+////////////
+// parser //
+////////////
+
+enum class parse_event_t : std::uint8_t
+{
+    /// the parser read `{` and started to process a JSON object
+    object_start,
+    /// the parser read `}` and finished processing a JSON object
+    object_end,
+    /// the parser read `[` and started to process a JSON array
+    array_start,
+    /// the parser read `]` and finished processing a JSON array
+    array_end,
+    /// the parser read a key of a value in an object
+    key,
+    /// the parser finished reading a JSON value
+    value
+};
+
+template<typename BasicJsonType>
+using parser_callback_t =
+    std::function<bool(int /*depth*/, parse_event_t /*event*/, BasicJsonType& /*parsed*/)>;
+
+/*!
+@brief syntax analysis
+
+This class implements a recursive descent parser.
+*/
+template<typename BasicJsonType, typename InputAdapterType>
+class parser
+{
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using string_t = typename BasicJsonType::string_t;
+    using lexer_t = lexer<BasicJsonType, InputAdapterType>;
+    using token_type = typename lexer_t::token_type;
+
+  public:
+    /// a parser reading from an input adapter
+    explicit parser(InputAdapterType&& adapter,
+                    parser_callback_t<BasicJsonType> cb = nullptr,
+                    const bool allow_exceptions_ = true,
+                    const bool skip_comments = false)
+        : callback(std::move(cb))
+        , m_lexer(std::move(adapter), skip_comments)
+        , allow_exceptions(allow_exceptions_)
+    {
+        // read first token
+        get_token();
+    }
+
+    /*!
+    @brief public parser interface
+
+    @param[in] strict      whether to expect the last token to be EOF
+    @param[in,out] result  parsed JSON value
+
+    @throw parse_error.101 in case of an unexpected token
+    @throw parse_error.102 if to_unicode fails or surrogate error
+    @throw parse_error.103 if to_unicode fails
+    */
+    void parse(const bool strict, BasicJsonType& result)
+    {
+        if (callback)
+        {
+            json_sax_dom_callback_parser<BasicJsonType, InputAdapterType> sdp(result, callback, allow_exceptions, &m_lexer);
+            sax_parse_internal(&sdp);
+
+            // in strict mode, input must be completely read
+            if (strict && (get_token() != token_type::end_of_input))
+            {
+                sdp.parse_error(m_lexer.get_position(),
+                                m_lexer.get_token_string(),
+                                parse_error::create(101, m_lexer.get_position(),
+                                                    exception_message(token_type::end_of_input, "value"), nullptr));
+            }
+
+            // in case of an error, return discarded value
+            if (sdp.is_errored())
+            {
+                result = value_t::discarded;
+                return;
+            }
+
+            // set top-level value to null if it was discarded by the callback
+            // function
+            if (result.is_discarded())
+            {
+                result = nullptr;
+            }
+        }
+        else
+        {
+            json_sax_dom_parser<BasicJsonType, InputAdapterType> sdp(result, allow_exceptions, &m_lexer);
+            sax_parse_internal(&sdp);
+
+            // in strict mode, input must be completely read
+            if (strict && (get_token() != token_type::end_of_input))
+            {
+                sdp.parse_error(m_lexer.get_position(),
+                                m_lexer.get_token_string(),
+                                parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_of_input, "value"), nullptr));
+            }
+
+            // in case of an error, return discarded value
+            if (sdp.is_errored())
+            {
+                result = value_t::discarded;
+                return;
+            }
+        }
+
+        result.assert_invariant();
+    }
+
+    /*!
+    @brief public accept interface
+
+    @param[in] strict  whether to expect the last token to be EOF
+    @return whether the input is a proper JSON text
+    */
+    bool accept(const bool strict = true)
+    {
+        json_sax_acceptor<BasicJsonType> sax_acceptor;
+        return sax_parse(&sax_acceptor, strict);
+    }
+
+    template<typename SAX>
+    JSON_HEDLEY_NON_NULL(2)
+    bool sax_parse(SAX* sax, const bool strict = true)
+    {
+        (void)detail::is_sax_static_asserts<SAX, BasicJsonType> {};
+        const bool result = sax_parse_internal(sax);
+
+        // strict mode: next byte must be EOF
+        if (result && strict && (get_token() != token_type::end_of_input))
+        {
+            return sax->parse_error(m_lexer.get_position(),
+                                    m_lexer.get_token_string(),
+                                    parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_of_input, "value"), nullptr));
+        }
+
+        return result;
+    }
+
+  private:
+    template<typename SAX>
+    JSON_HEDLEY_NON_NULL(2)
+    bool sax_parse_internal(SAX* sax)
+    {
+        // stack to remember the hierarchy of structured values we are parsing
+        // true = array; false = object
+        std::vector<bool> states;
+        // value to avoid a goto (see comment where set to true)
+        bool skip_to_state_evaluation = false;
+
+        while (true)
+        {
+            if (!skip_to_state_evaluation)
+            {
+                // invariant: get_token() was called before each iteration
+                switch (last_token)
+                {
+                    case token_type::begin_object:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!sax->start_object(detail::unknown_size())))
+                        {
+                            return false;
+                        }
+
+                        // closing } -> we are done
+                        if (get_token() == token_type::end_object)
+                        {
+                            if (JSON_HEDLEY_UNLIKELY(!sax->end_object()))
+                            {
+                                return false;
+                            }
+                            break;
+                        }
+
+                        // parse key
+                        if (JSON_HEDLEY_UNLIKELY(last_token != token_type::value_string))
+                        {
+                            return sax->parse_error(m_lexer.get_position(),
+                                                    m_lexer.get_token_string(),
+                                                    parse_error::create(101, m_lexer.get_position(), exception_message(token_type::value_string, "object key"), nullptr));
+                        }
+                        if (JSON_HEDLEY_UNLIKELY(!sax->key(m_lexer.get_string())))
+                        {
+                            return false;
+                        }
+
+                        // parse separator (:)
+                        if (JSON_HEDLEY_UNLIKELY(get_token() != token_type::name_separator))
+                        {
+                            return sax->parse_error(m_lexer.get_position(),
+                                                    m_lexer.get_token_string(),
+                                                    parse_error::create(101, m_lexer.get_position(), exception_message(token_type::name_separator, "object separator"), nullptr));
+                        }
+
+                        // remember we are now inside an object
+                        states.push_back(false);
+
+                        // parse values
+                        get_token();
+                        continue;
+                    }
+
+                    case token_type::begin_array:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!sax->start_array(detail::unknown_size())))
+                        {
+                            return false;
+                        }
+
+                        // closing ] -> we are done
+                        if (get_token() == token_type::end_array)
+                        {
+                            if (JSON_HEDLEY_UNLIKELY(!sax->end_array()))
+                            {
+                                return false;
+                            }
+                            break;
+                        }
+
+                        // remember we are now inside an array
+                        states.push_back(true);
+
+                        // parse values (no need to call get_token)
+                        continue;
+                    }
+
+                    case token_type::value_float:
+                    {
+                        const auto res = m_lexer.get_number_float();
+
+                        if (JSON_HEDLEY_UNLIKELY(!std::isfinite(res)))
+                        {
+                            return sax->parse_error(m_lexer.get_position(),
+                                                    m_lexer.get_token_string(),
+                                                    out_of_range::create(406, concat("number overflow parsing '", m_lexer.get_token_string(), '\''), nullptr));
+                        }
+
+                        if (JSON_HEDLEY_UNLIKELY(!sax->number_float(res, m_lexer.get_string())))
+                        {
+                            return false;
+                        }
+
+                        break;
+                    }
+
+                    case token_type::literal_false:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!sax->boolean(false)))
+                        {
+                            return false;
+                        }
+                        break;
+                    }
+
+                    case token_type::literal_null:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!sax->null()))
+                        {
+                            return false;
+                        }
+                        break;
+                    }
+
+                    case token_type::literal_true:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!sax->boolean(true)))
+                        {
+                            return false;
+                        }
+                        break;
+                    }
+
+                    case token_type::value_integer:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!sax->number_integer(m_lexer.get_number_integer())))
+                        {
+                            return false;
+                        }
+                        break;
+                    }
+
+                    case token_type::value_string:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!sax->string(m_lexer.get_string())))
+                        {
+                            return false;
+                        }
+                        break;
+                    }
+
+                    case token_type::value_unsigned:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!sax->number_unsigned(m_lexer.get_number_unsigned())))
+                        {
+                            return false;
+                        }
+                        break;
+                    }
+
+                    case token_type::parse_error:
+                    {
+                        // using "uninitialized" to avoid "expected" message
+                        return sax->parse_error(m_lexer.get_position(),
+                                                m_lexer.get_token_string(),
+                                                parse_error::create(101, m_lexer.get_position(), exception_message(token_type::uninitialized, "value"), nullptr));
+                    }
+                    case token_type::end_of_input:
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(m_lexer.get_position().chars_read_total == 1))
+                        {
+                            return sax->parse_error(m_lexer.get_position(),
+                                                    m_lexer.get_token_string(),
+                                                    parse_error::create(101, m_lexer.get_position(),
+                                                            "attempting to parse an empty input; check that your input string or stream contains the expected JSON", nullptr));
+                        }
+
+                        return sax->parse_error(m_lexer.get_position(),
+                                                m_lexer.get_token_string(),
+                                                parse_error::create(101, m_lexer.get_position(), exception_message(token_type::literal_or_value, "value"), nullptr));
+                    }
+                    case token_type::uninitialized:
+                    case token_type::end_array:
+                    case token_type::end_object:
+                    case token_type::name_separator:
+                    case token_type::value_separator:
+                    case token_type::literal_or_value:
+                    default: // the last token was unexpected
+                    {
+                        return sax->parse_error(m_lexer.get_position(),
+                                                m_lexer.get_token_string(),
+                                                parse_error::create(101, m_lexer.get_position(), exception_message(token_type::literal_or_value, "value"), nullptr));
+                    }
+                }
+            }
+            else
+            {
+                skip_to_state_evaluation = false;
+            }
+
+            // we reached this line after we successfully parsed a value
+            if (states.empty())
+            {
+                // empty stack: we reached the end of the hierarchy: done
+                return true;
+            }
+
+            if (states.back())  // array
+            {
+                // comma -> next value
+                if (get_token() == token_type::value_separator)
+                {
+                    // parse a new value
+                    get_token();
+                    continue;
+                }
+
+                // closing ]
+                if (JSON_HEDLEY_LIKELY(last_token == token_type::end_array))
+                {
+                    if (JSON_HEDLEY_UNLIKELY(!sax->end_array()))
+                    {
+                        return false;
+                    }
+
+                    // We are done with this array. Before we can parse a
+                    // new value, we need to evaluate the new state first.
+                    // By setting skip_to_state_evaluation to false, we
+                    // are effectively jumping to the beginning of this if.
+                    JSON_ASSERT(!states.empty());
+                    states.pop_back();
+                    skip_to_state_evaluation = true;
+                    continue;
+                }
+
+                return sax->parse_error(m_lexer.get_position(),
+                                        m_lexer.get_token_string(),
+                                        parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_array, "array"), nullptr));
+            }
+
+            // states.back() is false -> object
+
+            // comma -> next value
+            if (get_token() == token_type::value_separator)
+            {
+                // parse key
+                if (JSON_HEDLEY_UNLIKELY(get_token() != token_type::value_string))
+                {
+                    return sax->parse_error(m_lexer.get_position(),
+                                            m_lexer.get_token_string(),
+                                            parse_error::create(101, m_lexer.get_position(), exception_message(token_type::value_string, "object key"), nullptr));
+                }
+
+                if (JSON_HEDLEY_UNLIKELY(!sax->key(m_lexer.get_string())))
+                {
+                    return false;
+                }
+
+                // parse separator (:)
+                if (JSON_HEDLEY_UNLIKELY(get_token() != token_type::name_separator))
+                {
+                    return sax->parse_error(m_lexer.get_position(),
+                                            m_lexer.get_token_string(),
+                                            parse_error::create(101, m_lexer.get_position(), exception_message(token_type::name_separator, "object separator"), nullptr));
+                }
+
+                // parse values
+                get_token();
+                continue;
+            }
+
+            // closing }
+            if (JSON_HEDLEY_LIKELY(last_token == token_type::end_object))
+            {
+                if (JSON_HEDLEY_UNLIKELY(!sax->end_object()))
+                {
+                    return false;
+                }
+
+                // We are done with this object. Before we can parse a
+                // new value, we need to evaluate the new state first.
+                // By setting skip_to_state_evaluation to false, we
+                // are effectively jumping to the beginning of this if.
+                JSON_ASSERT(!states.empty());
+                states.pop_back();
+                skip_to_state_evaluation = true;
+                continue;
+            }
+
+            return sax->parse_error(m_lexer.get_position(),
+                                    m_lexer.get_token_string(),
+                                    parse_error::create(101, m_lexer.get_position(), exception_message(token_type::end_object, "object"), nullptr));
+        }
+    }
+
+    /// get next token from lexer
+    token_type get_token()
+    {
+        return last_token = m_lexer.scan();
+    }
+
+    std::string exception_message(const token_type expected, const std::string& context)
+    {
+        std::string error_msg = "syntax error ";
+
+        if (!context.empty())
+        {
+            error_msg += concat("while parsing ", context, ' ');
+        }
+
+        error_msg += "- ";
+
+        if (last_token == token_type::parse_error)
+        {
+            error_msg += concat(m_lexer.get_error_message(), "; last read: '",
+                                m_lexer.get_token_string(), '\'');
+        }
+        else
+        {
+            error_msg += concat("unexpected ", lexer_t::token_type_name(last_token));
+        }
+
+        if (expected != token_type::uninitialized)
+        {
+            error_msg += concat("; expected ", lexer_t::token_type_name(expected));
+        }
+
+        return error_msg;
+    }
+
+  private:
+    /// callback function
+    const parser_callback_t<BasicJsonType> callback = nullptr;
+    /// the type of the last read token
+    token_type last_token = token_type::uninitialized;
+    /// the lexer
+    lexer_t m_lexer;
+    /// whether to throw exceptions in case of errors
+    const bool allow_exceptions = true;
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/iterators/internal_iterator.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+// #include <nlohmann/detail/iterators/primitive_iterator.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstddef> // ptrdiff_t
+#include <limits>  // numeric_limits
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/*
+@brief an iterator for primitive JSON types
+
+This class models an iterator for primitive JSON types (boolean, number,
+string). It's only purpose is to allow the iterator/const_iterator classes
+to "iterate" over primitive values. Internally, the iterator is modeled by
+a `difference_type` variable. Value begin_value (`0`) models the begin,
+end_value (`1`) models past the end.
+*/
+class primitive_iterator_t
+{
+  private:
+    using difference_type = std::ptrdiff_t;
+    static constexpr difference_type begin_value = 0;
+    static constexpr difference_type end_value = begin_value + 1;
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    /// iterator as signed integer type
+    difference_type m_it = (std::numeric_limits<std::ptrdiff_t>::min)();
+
+  public:
+    constexpr difference_type get_value() const noexcept
+    {
+        return m_it;
+    }
+
+    /// set iterator to a defined beginning
+    void set_begin() noexcept
+    {
+        m_it = begin_value;
+    }
+
+    /// set iterator to a defined past the end
+    void set_end() noexcept
+    {
+        m_it = end_value;
+    }
+
+    /// return whether the iterator can be dereferenced
+    constexpr bool is_begin() const noexcept
+    {
+        return m_it == begin_value;
+    }
+
+    /// return whether the iterator is at end
+    constexpr bool is_end() const noexcept
+    {
+        return m_it == end_value;
+    }
+
+    friend constexpr bool operator==(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept
+    {
+        return lhs.m_it == rhs.m_it;
+    }
+
+    friend constexpr bool operator<(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept
+    {
+        return lhs.m_it < rhs.m_it;
+    }
+
+    primitive_iterator_t operator+(difference_type n) noexcept
+    {
+        auto result = *this;
+        result += n;
+        return result;
+    }
+
+    friend constexpr difference_type operator-(primitive_iterator_t lhs, primitive_iterator_t rhs) noexcept
+    {
+        return lhs.m_it - rhs.m_it;
+    }
+
+    primitive_iterator_t& operator++() noexcept
+    {
+        ++m_it;
+        return *this;
+    }
+
+    primitive_iterator_t operator++(int)& noexcept // NOLINT(cert-dcl21-cpp)
+    {
+        auto result = *this;
+        ++m_it;
+        return result;
+    }
+
+    primitive_iterator_t& operator--() noexcept
+    {
+        --m_it;
+        return *this;
+    }
+
+    primitive_iterator_t operator--(int)& noexcept // NOLINT(cert-dcl21-cpp)
+    {
+        auto result = *this;
+        --m_it;
+        return result;
+    }
+
+    primitive_iterator_t& operator+=(difference_type n) noexcept
+    {
+        m_it += n;
+        return *this;
+    }
+
+    primitive_iterator_t& operator-=(difference_type n) noexcept
+    {
+        m_it -= n;
+        return *this;
+    }
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/*!
+@brief an iterator value
+
+@note This structure could easily be a union, but MSVC currently does not allow
+unions members with complex constructors, see https://github.com/nlohmann/json/pull/105.
+*/
+template<typename BasicJsonType> struct internal_iterator
+{
+    /// iterator for JSON objects
+    typename BasicJsonType::object_t::iterator object_iterator {};
+    /// iterator for JSON arrays
+    typename BasicJsonType::array_t::iterator array_iterator {};
+    /// generic iterator for all other types
+    primitive_iterator_t primitive_iterator {};
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/iterators/iter_impl.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <iterator> // iterator, random_access_iterator_tag, bidirectional_iterator_tag, advance, next
+#include <type_traits> // conditional, is_const, remove_const
+
+// #include <nlohmann/detail/exceptions.hpp>
+
+// #include <nlohmann/detail/iterators/internal_iterator.hpp>
+
+// #include <nlohmann/detail/iterators/primitive_iterator.hpp>
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+// forward declare, to be able to friend it later on
+template<typename IteratorType> class iteration_proxy;
+template<typename IteratorType> class iteration_proxy_value;
+
+/*!
+@brief a template for a bidirectional iterator for the @ref basic_json class
+This class implements a both iterators (iterator and const_iterator) for the
+@ref basic_json class.
+@note An iterator is called *initialized* when a pointer to a JSON value has
+      been set (e.g., by a constructor or a copy assignment). If the iterator is
+      default-constructed, it is *uninitialized* and most methods are undefined.
+      **The library uses assertions to detect calls on uninitialized iterators.**
+@requirement The class satisfies the following concept requirements:
+-
+[BidirectionalIterator](https://en.cppreference.com/w/cpp/named_req/BidirectionalIterator):
+  The iterator that can be moved can be moved in both directions (i.e.
+  incremented and decremented).
+@since version 1.0.0, simplified in version 2.0.9, change to bidirectional
+       iterators in version 3.0.0 (see https://github.com/nlohmann/json/issues/593)
+*/
+template<typename BasicJsonType>
+class iter_impl // NOLINT(cppcoreguidelines-special-member-functions,hicpp-special-member-functions)
+{
+    /// the iterator with BasicJsonType of different const-ness
+    using other_iter_impl = iter_impl<typename std::conditional<std::is_const<BasicJsonType>::value, typename std::remove_const<BasicJsonType>::type, const BasicJsonType>::type>;
+    /// allow basic_json to access private members
+    friend other_iter_impl;
+    friend BasicJsonType;
+    friend iteration_proxy<iter_impl>;
+    friend iteration_proxy_value<iter_impl>;
+
+    using object_t = typename BasicJsonType::object_t;
+    using array_t = typename BasicJsonType::array_t;
+    // make sure BasicJsonType is basic_json or const basic_json
+    static_assert(is_basic_json<typename std::remove_const<BasicJsonType>::type>::value,
+                  "iter_impl only accepts (const) basic_json");
+    // superficial check for the LegacyBidirectionalIterator named requirement
+    static_assert(std::is_base_of<std::bidirectional_iterator_tag, std::bidirectional_iterator_tag>::value
+                  &&  std::is_base_of<std::bidirectional_iterator_tag, typename std::iterator_traits<typename array_t::iterator>::iterator_category>::value,
+                  "basic_json iterator assumes array and object type iterators satisfy the LegacyBidirectionalIterator named requirement.");
+
+  public:
+    /// The std::iterator class template (used as a base class to provide typedefs) is deprecated in C++17.
+    /// The C++ Standard has never required user-defined iterators to derive from std::iterator.
+    /// A user-defined iterator should provide publicly accessible typedefs named
+    /// iterator_category, value_type, difference_type, pointer, and reference.
+    /// Note that value_type is required to be non-const, even for constant iterators.
+    using iterator_category = std::bidirectional_iterator_tag;
+
+    /// the type of the values when the iterator is dereferenced
+    using value_type = typename BasicJsonType::value_type;
+    /// a type to represent differences between iterators
+    using difference_type = typename BasicJsonType::difference_type;
+    /// defines a pointer to the type iterated over (value_type)
+    using pointer = typename std::conditional<std::is_const<BasicJsonType>::value,
+          typename BasicJsonType::const_pointer,
+          typename BasicJsonType::pointer>::type;
+    /// defines a reference to the type iterated over (value_type)
+    using reference =
+        typename std::conditional<std::is_const<BasicJsonType>::value,
+        typename BasicJsonType::const_reference,
+        typename BasicJsonType::reference>::type;
+
+    iter_impl() = default;
+    ~iter_impl() = default;
+    iter_impl(iter_impl&&) noexcept = default;
+    iter_impl& operator=(iter_impl&&) noexcept = default;
+
+    /*!
+    @brief constructor for a given JSON instance
+    @param[in] object  pointer to a JSON object for this iterator
+    @pre object != nullptr
+    @post The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    explicit iter_impl(pointer object) noexcept : m_object(object)
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+            {
+                m_it.object_iterator = typename object_t::iterator();
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_it.array_iterator = typename array_t::iterator();
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                m_it.primitive_iterator = primitive_iterator_t();
+                break;
+            }
+        }
+    }
+
+    /*!
+    @note The conventional copy constructor and copy assignment are implicitly
+          defined. Combined with the following converting constructor and
+          assignment, they support: (1) copy from iterator to iterator, (2)
+          copy from const iterator to const iterator, and (3) conversion from
+          iterator to const iterator. However conversion from const iterator
+          to iterator is not defined.
+    */
+
+    /*!
+    @brief const copy constructor
+    @param[in] other const iterator to copy from
+    @note This copy constructor had to be defined explicitly to circumvent a bug
+          occurring on msvc v19.0 compiler (VS 2015) debug build. For more
+          information refer to: https://github.com/nlohmann/json/issues/1608
+    */
+    iter_impl(const iter_impl<const BasicJsonType>& other) noexcept
+        : m_object(other.m_object), m_it(other.m_it)
+    {}
+
+    /*!
+    @brief converting assignment
+    @param[in] other const iterator to copy from
+    @return const/non-const iterator
+    @note It is not checked whether @a other is initialized.
+    */
+    iter_impl& operator=(const iter_impl<const BasicJsonType>& other) noexcept
+    {
+        if (&other != this)
+        {
+            m_object = other.m_object;
+            m_it = other.m_it;
+        }
+        return *this;
+    }
+
+    /*!
+    @brief converting constructor
+    @param[in] other  non-const iterator to copy from
+    @note It is not checked whether @a other is initialized.
+    */
+    iter_impl(const iter_impl<typename std::remove_const<BasicJsonType>::type>& other) noexcept
+        : m_object(other.m_object), m_it(other.m_it)
+    {}
+
+    /*!
+    @brief converting assignment
+    @param[in] other  non-const iterator to copy from
+    @return const/non-const iterator
+    @note It is not checked whether @a other is initialized.
+    */
+    iter_impl& operator=(const iter_impl<typename std::remove_const<BasicJsonType>::type>& other) noexcept // NOLINT(cert-oop54-cpp)
+    {
+        m_object = other.m_object;
+        m_it = other.m_it;
+        return *this;
+    }
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    /*!
+    @brief set the iterator to the first value
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    void set_begin() noexcept
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+            {
+                m_it.object_iterator = m_object->m_data.m_value.object->begin();
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_it.array_iterator = m_object->m_data.m_value.array->begin();
+                break;
+            }
+
+            case value_t::null:
+            {
+                // set to end so begin()==end() is true: null is empty
+                m_it.primitive_iterator.set_end();
+                break;
+            }
+
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                m_it.primitive_iterator.set_begin();
+                break;
+            }
+        }
+    }
+
+    /*!
+    @brief set the iterator past the last value
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    void set_end() noexcept
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+            {
+                m_it.object_iterator = m_object->m_data.m_value.object->end();
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_it.array_iterator = m_object->m_data.m_value.array->end();
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                m_it.primitive_iterator.set_end();
+                break;
+            }
+        }
+    }
+
+  public:
+    /*!
+    @brief return a reference to the value pointed to by the iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    reference operator*() const
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+            {
+                JSON_ASSERT(m_it.object_iterator != m_object->m_data.m_value.object->end());
+                return m_it.object_iterator->second;
+            }
+
+            case value_t::array:
+            {
+                JSON_ASSERT(m_it.array_iterator != m_object->m_data.m_value.array->end());
+                return *m_it.array_iterator;
+            }
+
+            case value_t::null:
+                JSON_THROW(invalid_iterator::create(214, "cannot get value", m_object));
+
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                if (JSON_HEDLEY_LIKELY(m_it.primitive_iterator.is_begin()))
+                {
+                    return *m_object;
+                }
+
+                JSON_THROW(invalid_iterator::create(214, "cannot get value", m_object));
+            }
+        }
+    }
+
+    /*!
+    @brief dereference the iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    pointer operator->() const
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+            {
+                JSON_ASSERT(m_it.object_iterator != m_object->m_data.m_value.object->end());
+                return &(m_it.object_iterator->second);
+            }
+
+            case value_t::array:
+            {
+                JSON_ASSERT(m_it.array_iterator != m_object->m_data.m_value.array->end());
+                return &*m_it.array_iterator;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                if (JSON_HEDLEY_LIKELY(m_it.primitive_iterator.is_begin()))
+                {
+                    return m_object;
+                }
+
+                JSON_THROW(invalid_iterator::create(214, "cannot get value", m_object));
+            }
+        }
+    }
+
+    /*!
+    @brief post-increment (it++)
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl operator++(int)& // NOLINT(cert-dcl21-cpp)
+    {
+        auto result = *this;
+        ++(*this);
+        return result;
+    }
+
+    /*!
+    @brief pre-increment (++it)
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl& operator++()
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+            {
+                std::advance(m_it.object_iterator, 1);
+                break;
+            }
+
+            case value_t::array:
+            {
+                std::advance(m_it.array_iterator, 1);
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                ++m_it.primitive_iterator;
+                break;
+            }
+        }
+
+        return *this;
+    }
+
+    /*!
+    @brief post-decrement (it--)
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl operator--(int)& // NOLINT(cert-dcl21-cpp)
+    {
+        auto result = *this;
+        --(*this);
+        return result;
+    }
+
+    /*!
+    @brief pre-decrement (--it)
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl& operator--()
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+            {
+                std::advance(m_it.object_iterator, -1);
+                break;
+            }
+
+            case value_t::array:
+            {
+                std::advance(m_it.array_iterator, -1);
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                --m_it.primitive_iterator;
+                break;
+            }
+        }
+
+        return *this;
+    }
+
+    /*!
+    @brief comparison: equal
+    @pre (1) Both iterators are initialized to point to the same object, or (2) both iterators are value-initialized.
+    */
+    template < typename IterImpl, detail::enable_if_t < (std::is_same<IterImpl, iter_impl>::value || std::is_same<IterImpl, other_iter_impl>::value), std::nullptr_t > = nullptr >
+    bool operator==(const IterImpl& other) const
+    {
+        // if objects are not the same, the comparison is undefined
+        if (JSON_HEDLEY_UNLIKELY(m_object != other.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(212, "cannot compare iterators of different containers", m_object));
+        }
+
+        // value-initialized forward iterators can be compared, and must compare equal to other value-initialized iterators of the same type #4493
+        if (m_object == nullptr)
+        {
+            return true;
+        }
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+                return (m_it.object_iterator == other.m_it.object_iterator);
+
+            case value_t::array:
+                return (m_it.array_iterator == other.m_it.array_iterator);
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+                return (m_it.primitive_iterator == other.m_it.primitive_iterator);
+        }
+    }
+
+    /*!
+    @brief comparison: not equal
+    @pre (1) Both iterators are initialized to point to the same object, or (2) both iterators are value-initialized.
+    */
+    template < typename IterImpl, detail::enable_if_t < (std::is_same<IterImpl, iter_impl>::value || std::is_same<IterImpl, other_iter_impl>::value), std::nullptr_t > = nullptr >
+    bool operator!=(const IterImpl& other) const
+    {
+        return !operator==(other);
+    }
+
+    /*!
+    @brief comparison: smaller
+    @pre (1) Both iterators are initialized to point to the same object, or (2) both iterators are value-initialized.
+    */
+    bool operator<(const iter_impl& other) const
+    {
+        // if objects are not the same, the comparison is undefined
+        if (JSON_HEDLEY_UNLIKELY(m_object != other.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(212, "cannot compare iterators of different containers", m_object));
+        }
+
+        // value-initialized forward iterators can be compared, and must compare equal to other value-initialized iterators of the same type #4493
+        if (m_object == nullptr)
+        {
+            // the iterators are both value-initialized and are to be considered equal, but this function checks for smaller, so we return false
+            return false;
+        }
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+                JSON_THROW(invalid_iterator::create(213, "cannot compare order of object iterators", m_object));
+
+            case value_t::array:
+                return (m_it.array_iterator < other.m_it.array_iterator);
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+                return (m_it.primitive_iterator < other.m_it.primitive_iterator);
+        }
+    }
+
+    /*!
+    @brief comparison: less than or equal
+    @pre (1) Both iterators are initialized to point to the same object, or (2) both iterators are value-initialized.
+    */
+    bool operator<=(const iter_impl& other) const
+    {
+        return !other.operator < (*this);
+    }
+
+    /*!
+    @brief comparison: greater than
+    @pre (1) Both iterators are initialized to point to the same object, or (2) both iterators are value-initialized.
+    */
+    bool operator>(const iter_impl& other) const
+    {
+        return !operator<=(other);
+    }
+
+    /*!
+    @brief comparison: greater than or equal
+    @pre (1) The iterator is initialized; i.e. `m_object != nullptr`, or (2) both iterators are value-initialized.
+    */
+    bool operator>=(const iter_impl& other) const
+    {
+        return !operator<(other);
+    }
+
+    /*!
+    @brief add to iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl& operator+=(difference_type i)
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+                JSON_THROW(invalid_iterator::create(209, "cannot use offsets with object iterators", m_object));
+
+            case value_t::array:
+            {
+                std::advance(m_it.array_iterator, i);
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                m_it.primitive_iterator += i;
+                break;
+            }
+        }
+
+        return *this;
+    }
+
+    /*!
+    @brief subtract from iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl& operator-=(difference_type i)
+    {
+        return operator+=(-i);
+    }
+
+    /*!
+    @brief add to iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl operator+(difference_type i) const
+    {
+        auto result = *this;
+        result += i;
+        return result;
+    }
+
+    /*!
+    @brief addition of distance and iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    friend iter_impl operator+(difference_type i, const iter_impl& it)
+    {
+        auto result = it;
+        result += i;
+        return result;
+    }
+
+    /*!
+    @brief subtract from iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    iter_impl operator-(difference_type i) const
+    {
+        auto result = *this;
+        result -= i;
+        return result;
+    }
+
+    /*!
+    @brief return difference
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    difference_type operator-(const iter_impl& other) const
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+                JSON_THROW(invalid_iterator::create(209, "cannot use offsets with object iterators", m_object));
+
+            case value_t::array:
+                return m_it.array_iterator - other.m_it.array_iterator;
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+                return m_it.primitive_iterator - other.m_it.primitive_iterator;
+        }
+    }
+
+    /*!
+    @brief access to successor
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    reference operator[](difference_type n) const
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        switch (m_object->m_data.m_type)
+        {
+            case value_t::object:
+                JSON_THROW(invalid_iterator::create(208, "cannot use operator[] for object iterators", m_object));
+
+            case value_t::array:
+                return *std::next(m_it.array_iterator, n);
+
+            case value_t::null:
+                JSON_THROW(invalid_iterator::create(214, "cannot get value", m_object));
+
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                if (JSON_HEDLEY_LIKELY(m_it.primitive_iterator.get_value() == -n))
+                {
+                    return *m_object;
+                }
+
+                JSON_THROW(invalid_iterator::create(214, "cannot get value", m_object));
+            }
+        }
+    }
+
+    /*!
+    @brief return the key of an object iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    const typename object_t::key_type& key() const
+    {
+        JSON_ASSERT(m_object != nullptr);
+
+        if (JSON_HEDLEY_LIKELY(m_object->is_object()))
+        {
+            return m_it.object_iterator->first;
+        }
+
+        JSON_THROW(invalid_iterator::create(207, "cannot use key() for non-object iterators", m_object));
+    }
+
+    /*!
+    @brief return the value of an iterator
+    @pre The iterator is initialized; i.e. `m_object != nullptr`.
+    */
+    reference value() const
+    {
+        return operator*();
+    }
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    /// associated JSON instance
+    pointer m_object = nullptr;
+    /// the actual iterator of the associated instance
+    internal_iterator<typename std::remove_const<BasicJsonType>::type> m_it {};
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/iterators/iteration_proxy.hpp>
+
+// #include <nlohmann/detail/iterators/json_reverse_iterator.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <cstddef> // ptrdiff_t
+#include <iterator> // reverse_iterator
+#include <utility> // declval
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+//////////////////////
+// reverse_iterator //
+//////////////////////
+
+/*!
+@brief a template for a reverse iterator class
+
+@tparam Base the base iterator type to reverse. Valid types are @ref
+iterator (to create @ref reverse_iterator) and @ref const_iterator (to
+create @ref const_reverse_iterator).
+
+@requirement The class satisfies the following concept requirements:
+-
+[BidirectionalIterator](https://en.cppreference.com/w/cpp/named_req/BidirectionalIterator):
+  The iterator that can be moved can be moved in both directions (i.e.
+  incremented and decremented).
+- [OutputIterator](https://en.cppreference.com/w/cpp/named_req/OutputIterator):
+  It is possible to write to the pointed-to element (only if @a Base is
+  @ref iterator).
+
+@since version 1.0.0
+*/
+template<typename Base>
+class json_reverse_iterator : public std::reverse_iterator<Base>
+{
+  public:
+    using difference_type = std::ptrdiff_t;
+    /// shortcut to the reverse iterator adapter
+    using base_iterator = std::reverse_iterator<Base>;
+    /// the reference type for the pointed-to element
+    using reference = typename Base::reference;
+
+    /// create reverse iterator from iterator
+    explicit json_reverse_iterator(const typename base_iterator::iterator_type& it) noexcept
+        : base_iterator(it) {}
+
+    /// create reverse iterator from base class
+    explicit json_reverse_iterator(const base_iterator& it) noexcept : base_iterator(it) {}
+
+    /// post-increment (it++)
+    json_reverse_iterator operator++(int)& // NOLINT(cert-dcl21-cpp)
+    {
+        return static_cast<json_reverse_iterator>(base_iterator::operator++(1));
+    }
+
+    /// pre-increment (++it)
+    json_reverse_iterator& operator++()
+    {
+        return static_cast<json_reverse_iterator&>(base_iterator::operator++());
+    }
+
+    /// post-decrement (it--)
+    json_reverse_iterator operator--(int)& // NOLINT(cert-dcl21-cpp)
+    {
+        return static_cast<json_reverse_iterator>(base_iterator::operator--(1));
+    }
+
+    /// pre-decrement (--it)
+    json_reverse_iterator& operator--()
+    {
+        return static_cast<json_reverse_iterator&>(base_iterator::operator--());
+    }
+
+    /// add to iterator
+    json_reverse_iterator& operator+=(difference_type i)
+    {
+        return static_cast<json_reverse_iterator&>(base_iterator::operator+=(i));
+    }
+
+    /// add to iterator
+    json_reverse_iterator operator+(difference_type i) const
+    {
+        return static_cast<json_reverse_iterator>(base_iterator::operator+(i));
+    }
+
+    /// subtract from iterator
+    json_reverse_iterator operator-(difference_type i) const
+    {
+        return static_cast<json_reverse_iterator>(base_iterator::operator-(i));
+    }
+
+    /// return difference
+    difference_type operator-(const json_reverse_iterator& other) const
+    {
+        return base_iterator(*this) - base_iterator(other);
+    }
+
+    /// access to successor
+    reference operator[](difference_type n) const
+    {
+        return *(this->operator+(n));
+    }
+
+    /// return the key of an object iterator
+    auto key() const -> decltype(std::declval<Base>().key())
+    {
+        auto it = --this->base();
+        return it.key();
+    }
+
+    /// return the value of an iterator
+    reference value() const
+    {
+        auto it = --this->base();
+        return it.operator * ();
+    }
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/iterators/primitive_iterator.hpp>
+
+// #include <nlohmann/detail/json_custom_base_class.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <type_traits> // conditional, is_same
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/*!
+@brief Default base class of the @ref basic_json class.
+
+So that the correct implementations of the copy / move ctors / assign operators
+of @ref basic_json do not require complex case distinctions
+(no base class / custom base class used as customization point),
+@ref basic_json always has a base class.
+By default, this class is used because it is empty and thus has no effect
+on the behavior of @ref basic_json.
+*/
+struct json_default_base {};
+
+template<class T>
+using json_base_class = typename std::conditional <
+                        std::is_same<T, void>::value,
+                        json_default_base,
+                        T
+                        >::type;
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/json_pointer.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <algorithm> // all_of
+#include <cctype> // isdigit
+#include <cerrno> // errno, ERANGE
+#include <cstdlib> // strtoull
+#ifndef JSON_NO_IO
+    #include <iosfwd> // ostream
+#endif  // JSON_NO_IO
+#include <limits> // max
+#include <numeric> // accumulate
+#include <string> // string
+#include <utility> // move
+#include <vector> // vector
+
+// #include <nlohmann/detail/exceptions.hpp>
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/string_concat.hpp>
+
+// #include <nlohmann/detail/string_escape.hpp>
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+/// @brief JSON Pointer defines a string syntax for identifying a specific value within a JSON document
+/// @sa https://json.nlohmann.me/api/json_pointer/
+template<typename RefStringType>
+class json_pointer
+{
+    // allow basic_json to access private members
+    NLOHMANN_BASIC_JSON_TPL_DECLARATION
+    friend class basic_json;
+
+    template<typename>
+    friend class json_pointer;
+
+    template<typename T>
+    struct string_t_helper
+    {
+        using type = T;
+    };
+
+    NLOHMANN_BASIC_JSON_TPL_DECLARATION
+    struct string_t_helper<NLOHMANN_BASIC_JSON_TPL>
+    {
+        using type = StringType;
+    };
+
+  public:
+    // for backwards compatibility accept BasicJsonType
+    using string_t = typename string_t_helper<RefStringType>::type;
+
+    /// @brief create JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/json_pointer/
+    explicit json_pointer(const string_t& s = "")
+        : reference_tokens(split(s))
+    {}
+
+    /// @brief return a string representation of the JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/to_string/
+    string_t to_string() const
+    {
+        return std::accumulate(reference_tokens.begin(), reference_tokens.end(),
+                               string_t{},
+                               [](const string_t& a, const string_t& b)
+        {
+            return detail::concat(a, '/', detail::escape(b));
+        });
+    }
+
+    /// @brief return a string representation of the JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_string/
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, to_string())
+    operator string_t() const
+    {
+        return to_string();
+    }
+
+#ifndef JSON_NO_IO
+    /// @brief write string representation of the JSON pointer to stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ltlt/
+    friend std::ostream& operator<<(std::ostream& o, const json_pointer& ptr)
+    {
+        o << ptr.to_string();
+        return o;
+    }
+#endif
+
+    /// @brief append another JSON pointer at the end of this JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_slasheq/
+    json_pointer& operator/=(const json_pointer& ptr)
+    {
+        reference_tokens.insert(reference_tokens.end(),
+                                ptr.reference_tokens.begin(),
+                                ptr.reference_tokens.end());
+        return *this;
+    }
+
+    /// @brief append an unescaped reference token at the end of this JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_slasheq/
+    json_pointer& operator/=(string_t token)
+    {
+        push_back(std::move(token));
+        return *this;
+    }
+
+    /// @brief append an array index at the end of this JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_slasheq/
+    json_pointer& operator/=(std::size_t array_idx)
+    {
+        return *this /= std::to_string(array_idx);
+    }
+
+    /// @brief create a new JSON pointer by appending the right JSON pointer at the end of the left JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_slash/
+    friend json_pointer operator/(const json_pointer& lhs,
+                                  const json_pointer& rhs)
+    {
+        return json_pointer(lhs) /= rhs;
+    }
+
+    /// @brief create a new JSON pointer by appending the unescaped token at the end of the JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_slash/
+    friend json_pointer operator/(const json_pointer& lhs, string_t token) // NOLINT(performance-unnecessary-value-param)
+    {
+        return json_pointer(lhs) /= std::move(token);
+    }
+
+    /// @brief create a new JSON pointer by appending the array-index-token at the end of the JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_slash/
+    friend json_pointer operator/(const json_pointer& lhs, std::size_t array_idx)
+    {
+        return json_pointer(lhs) /= array_idx;
+    }
+
+    /// @brief returns the parent of this JSON pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/parent_pointer/
+    json_pointer parent_pointer() const
+    {
+        if (empty())
+        {
+            return *this;
+        }
+
+        json_pointer res = *this;
+        res.pop_back();
+        return res;
+    }
+
+    /// @brief remove last reference token
+    /// @sa https://json.nlohmann.me/api/json_pointer/pop_back/
+    void pop_back()
+    {
+        if (JSON_HEDLEY_UNLIKELY(empty()))
+        {
+            JSON_THROW(detail::out_of_range::create(405, "JSON pointer has no parent", nullptr));
+        }
+
+        reference_tokens.pop_back();
+    }
+
+    /// @brief return last reference token
+    /// @sa https://json.nlohmann.me/api/json_pointer/back/
+    const string_t& back() const
+    {
+        if (JSON_HEDLEY_UNLIKELY(empty()))
+        {
+            JSON_THROW(detail::out_of_range::create(405, "JSON pointer has no parent", nullptr));
+        }
+
+        return reference_tokens.back();
+    }
+
+    /// @brief append an unescaped token at the end of the reference pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/push_back/
+    void push_back(const string_t& token)
+    {
+        reference_tokens.push_back(token);
+    }
+
+    /// @brief append an unescaped token at the end of the reference pointer
+    /// @sa https://json.nlohmann.me/api/json_pointer/push_back/
+    void push_back(string_t&& token)
+    {
+        reference_tokens.push_back(std::move(token));
+    }
+
+    /// @brief return whether pointer points to the root document
+    /// @sa https://json.nlohmann.me/api/json_pointer/empty/
+    bool empty() const noexcept
+    {
+        return reference_tokens.empty();
+    }
+
+  private:
+    /*!
+    @param[in] s  reference token to be converted into an array index
+
+    @return integer representation of @a s
+
+    @throw parse_error.106  if an array index begins with '0'
+    @throw parse_error.109  if an array index begins not with a digit
+    @throw out_of_range.404 if string @a s could not be converted to an integer
+    @throw out_of_range.410 if an array index exceeds size_type
+    */
+    template<typename BasicJsonType>
+    static typename BasicJsonType::size_type array_index(const string_t& s)
+    {
+        using size_type = typename BasicJsonType::size_type;
+
+        // error condition (cf. RFC 6901, Sect. 4)
+        if (JSON_HEDLEY_UNLIKELY(s.size() > 1 && s[0] == '0'))
+        {
+            JSON_THROW(detail::parse_error::create(106, 0, detail::concat("array index '", s, "' must not begin with '0'"), nullptr));
+        }
+
+        // error condition (cf. RFC 6901, Sect. 4)
+        if (JSON_HEDLEY_UNLIKELY(s.size() > 1 && !(s[0] >= '1' && s[0] <= '9')))
+        {
+            JSON_THROW(detail::parse_error::create(109, 0, detail::concat("array index '", s, "' is not a number"), nullptr));
+        }
+
+        const char* p = s.c_str();
+        char* p_end = nullptr;
+        errno = 0; // strtoull doesn't reset errno
+        const unsigned long long res = std::strtoull(p, &p_end, 10); // NOLINT(runtime/int)
+        if (p == p_end // invalid input or empty string
+                || errno == ERANGE // out of range
+                || JSON_HEDLEY_UNLIKELY(static_cast<std::size_t>(p_end - p) != s.size())) // incomplete read
+        {
+            JSON_THROW(detail::out_of_range::create(404, detail::concat("unresolved reference token '", s, "'"), nullptr));
+        }
+
+        // only triggered on special platforms (like 32bit), see also
+        // https://github.com/nlohmann/json/pull/2203
+        if (res >= static_cast<unsigned long long>((std::numeric_limits<size_type>::max)()))  // NOLINT(runtime/int)
+        {
+            JSON_THROW(detail::out_of_range::create(410, detail::concat("array index ", s, " exceeds size_type"), nullptr));   // LCOV_EXCL_LINE
+        }
+
+        return static_cast<size_type>(res);
+    }
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    json_pointer top() const
+    {
+        if (JSON_HEDLEY_UNLIKELY(empty()))
+        {
+            JSON_THROW(detail::out_of_range::create(405, "JSON pointer has no parent", nullptr));
+        }
+
+        json_pointer result = *this;
+        result.reference_tokens = {reference_tokens[0]};
+        return result;
+    }
+
+  private:
+    /*!
+    @brief create and return a reference to the pointed to value
+
+    @complexity Linear in the number of reference tokens.
+
+    @throw parse_error.109 if array index is not a number
+    @throw type_error.313 if value cannot be unflattened
+    */
+    template<typename BasicJsonType>
+    BasicJsonType& get_and_create(BasicJsonType& j) const
+    {
+        auto* result = &j;
+
+        // in case no reference tokens exist, return a reference to the JSON value
+        // j which will be overwritten by a primitive value
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (result->type())
+            {
+                case detail::value_t::null:
+                {
+                    if (reference_token == "0")
+                    {
+                        // start a new array if reference token is 0
+                        result = &result->operator[](0);
+                    }
+                    else
+                    {
+                        // start a new object otherwise
+                        result = &result->operator[](reference_token);
+                    }
+                    break;
+                }
+
+                case detail::value_t::object:
+                {
+                    // create an entry in the object
+                    result = &result->operator[](reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    // create an entry in the array
+                    result = &result->operator[](array_index<BasicJsonType>(reference_token));
+                    break;
+                }
+
+                /*
+                The following code is only reached if there exists a reference
+                token _and_ the current value is primitive. In this case, we have
+                an error situation, because primitive values may only occur as
+                single value; that is, with an empty list of reference tokens.
+                */
+                case detail::value_t::string:
+                case detail::value_t::boolean:
+                case detail::value_t::number_integer:
+                case detail::value_t::number_unsigned:
+                case detail::value_t::number_float:
+                case detail::value_t::binary:
+                case detail::value_t::discarded:
+                default:
+                    JSON_THROW(detail::type_error::create(313, "invalid value to unflatten", &j));
+            }
+        }
+
+        return *result;
+    }
+
+    /*!
+    @brief return a reference to the pointed to value
+
+    @note This version does not throw if a value is not present, but tries to
+          create nested values instead. For instance, calling this function
+          with pointer `"/this/that"` on a null value is equivalent to calling
+          `operator[]("this").operator[]("that")` on that value, effectively
+          changing the null value to an object.
+
+    @param[in] ptr  a JSON value
+
+    @return reference to the JSON value pointed to by the JSON pointer
+
+    @complexity Linear in the length of the JSON pointer.
+
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    @throw out_of_range.404  if the JSON pointer can not be resolved
+    */
+    template<typename BasicJsonType>
+    BasicJsonType& get_unchecked(BasicJsonType* ptr) const
+    {
+        for (const auto& reference_token : reference_tokens)
+        {
+            // convert null values to arrays or objects before continuing
+            if (ptr->is_null())
+            {
+                // check if reference token is a number
+                const bool nums =
+                    std::all_of(reference_token.begin(), reference_token.end(),
+                                [](const unsigned char x)
+                {
+                    return std::isdigit(x);
+                });
+
+                // change value to array for numbers or "-" or to object otherwise
+                *ptr = (nums || reference_token == "-")
+                       ? detail::value_t::array
+                       : detail::value_t::object;
+            }
+
+            switch (ptr->type())
+            {
+                case detail::value_t::object:
+                {
+                    // use unchecked object access
+                    ptr = &ptr->operator[](reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    if (reference_token == "-")
+                    {
+                        // explicitly treat "-" as index beyond the end
+                        ptr = &ptr->operator[](ptr->m_data.m_value.array->size());
+                    }
+                    else
+                    {
+                        // convert array index to number; unchecked access
+                        ptr = &ptr->operator[](array_index<BasicJsonType>(reference_token));
+                    }
+                    break;
+                }
+
+                case detail::value_t::null:
+                case detail::value_t::string:
+                case detail::value_t::boolean:
+                case detail::value_t::number_integer:
+                case detail::value_t::number_unsigned:
+                case detail::value_t::number_float:
+                case detail::value_t::binary:
+                case detail::value_t::discarded:
+                default:
+                    JSON_THROW(detail::out_of_range::create(404, detail::concat("unresolved reference token '", reference_token, "'"), ptr));
+            }
+        }
+
+        return *ptr;
+    }
+
+    /*!
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    @throw out_of_range.402  if the array index '-' is used
+    @throw out_of_range.404  if the JSON pointer can not be resolved
+    */
+    template<typename BasicJsonType>
+    BasicJsonType& get_checked(BasicJsonType* ptr) const
+    {
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (ptr->type())
+            {
+                case detail::value_t::object:
+                {
+                    // note: at performs range check
+                    ptr = &ptr->at(reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(reference_token == "-"))
+                    {
+                        // "-" always fails the range check
+                        JSON_THROW(detail::out_of_range::create(402, detail::concat(
+                                "array index '-' (", std::to_string(ptr->m_data.m_value.array->size()),
+                                ") is out of range"), ptr));
+                    }
+
+                    // note: at performs range check
+                    ptr = &ptr->at(array_index<BasicJsonType>(reference_token));
+                    break;
+                }
+
+                case detail::value_t::null:
+                case detail::value_t::string:
+                case detail::value_t::boolean:
+                case detail::value_t::number_integer:
+                case detail::value_t::number_unsigned:
+                case detail::value_t::number_float:
+                case detail::value_t::binary:
+                case detail::value_t::discarded:
+                default:
+                    JSON_THROW(detail::out_of_range::create(404, detail::concat("unresolved reference token '", reference_token, "'"), ptr));
+            }
+        }
+
+        return *ptr;
+    }
+
+    /*!
+    @brief return a const reference to the pointed to value
+
+    @param[in] ptr  a JSON value
+
+    @return const reference to the JSON value pointed to by the JSON
+    pointer
+
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    @throw out_of_range.402  if the array index '-' is used
+    @throw out_of_range.404  if the JSON pointer can not be resolved
+    */
+    template<typename BasicJsonType>
+    const BasicJsonType& get_unchecked(const BasicJsonType* ptr) const
+    {
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (ptr->type())
+            {
+                case detail::value_t::object:
+                {
+                    // use unchecked object access
+                    ptr = &ptr->operator[](reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(reference_token == "-"))
+                    {
+                        // "-" cannot be used for const access
+                        JSON_THROW(detail::out_of_range::create(402, detail::concat("array index '-' (", std::to_string(ptr->m_data.m_value.array->size()), ") is out of range"), ptr));
+                    }
+
+                    // use unchecked array access
+                    ptr = &ptr->operator[](array_index<BasicJsonType>(reference_token));
+                    break;
+                }
+
+                case detail::value_t::null:
+                case detail::value_t::string:
+                case detail::value_t::boolean:
+                case detail::value_t::number_integer:
+                case detail::value_t::number_unsigned:
+                case detail::value_t::number_float:
+                case detail::value_t::binary:
+                case detail::value_t::discarded:
+                default:
+                    JSON_THROW(detail::out_of_range::create(404, detail::concat("unresolved reference token '", reference_token, "'"), ptr));
+            }
+        }
+
+        return *ptr;
+    }
+
+    /*!
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    @throw out_of_range.402  if the array index '-' is used
+    @throw out_of_range.404  if the JSON pointer can not be resolved
+    */
+    template<typename BasicJsonType>
+    const BasicJsonType& get_checked(const BasicJsonType* ptr) const
+    {
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (ptr->type())
+            {
+                case detail::value_t::object:
+                {
+                    // note: at performs range check
+                    ptr = &ptr->at(reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(reference_token == "-"))
+                    {
+                        // "-" always fails the range check
+                        JSON_THROW(detail::out_of_range::create(402, detail::concat(
+                                "array index '-' (", std::to_string(ptr->m_data.m_value.array->size()),
+                                ") is out of range"), ptr));
+                    }
+
+                    // note: at performs range check
+                    ptr = &ptr->at(array_index<BasicJsonType>(reference_token));
+                    break;
+                }
+
+                case detail::value_t::null:
+                case detail::value_t::string:
+                case detail::value_t::boolean:
+                case detail::value_t::number_integer:
+                case detail::value_t::number_unsigned:
+                case detail::value_t::number_float:
+                case detail::value_t::binary:
+                case detail::value_t::discarded:
+                default:
+                    JSON_THROW(detail::out_of_range::create(404, detail::concat("unresolved reference token '", reference_token, "'"), ptr));
+            }
+        }
+
+        return *ptr;
+    }
+
+    /*!
+    @throw parse_error.106   if an array index begins with '0'
+    @throw parse_error.109   if an array index was not a number
+    */
+    template<typename BasicJsonType>
+    bool contains(const BasicJsonType* ptr) const
+    {
+        for (const auto& reference_token : reference_tokens)
+        {
+            switch (ptr->type())
+            {
+                case detail::value_t::object:
+                {
+                    if (!ptr->contains(reference_token))
+                    {
+                        // we did not find the key in the object
+                        return false;
+                    }
+
+                    ptr = &ptr->operator[](reference_token);
+                    break;
+                }
+
+                case detail::value_t::array:
+                {
+                    if (JSON_HEDLEY_UNLIKELY(reference_token == "-"))
+                    {
+                        // "-" always fails the range check
+                        return false;
+                    }
+                    if (JSON_HEDLEY_UNLIKELY(reference_token.size() == 1 && !("0" <= reference_token && reference_token <= "9")))
+                    {
+                        // invalid char
+                        return false;
+                    }
+                    if (JSON_HEDLEY_UNLIKELY(reference_token.size() > 1))
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!('1' <= reference_token[0] && reference_token[0] <= '9')))
+                        {
+                            // first char should be between '1' and '9'
+                            return false;
+                        }
+                        for (std::size_t i = 1; i < reference_token.size(); i++)
+                        {
+                            if (JSON_HEDLEY_UNLIKELY(!('0' <= reference_token[i] && reference_token[i] <= '9')))
+                            {
+                                // other char should be between '0' and '9'
+                                return false;
+                            }
+                        }
+                    }
+
+                    const auto idx = array_index<BasicJsonType>(reference_token);
+                    if (idx >= ptr->size())
+                    {
+                        // index out of range
+                        return false;
+                    }
+
+                    ptr = &ptr->operator[](idx);
+                    break;
+                }
+
+                case detail::value_t::null:
+                case detail::value_t::string:
+                case detail::value_t::boolean:
+                case detail::value_t::number_integer:
+                case detail::value_t::number_unsigned:
+                case detail::value_t::number_float:
+                case detail::value_t::binary:
+                case detail::value_t::discarded:
+                default:
+                {
+                    // we do not expect primitive values if there is still a
+                    // reference token to process
+                    return false;
+                }
+            }
+        }
+
+        // no reference token left means we found a primitive value
+        return true;
+    }
+
+    /*!
+    @brief split the string input to reference tokens
+
+    @note This function is only called by the json_pointer constructor.
+          All exceptions below are documented there.
+
+    @throw parse_error.107  if the pointer is not empty or begins with '/'
+    @throw parse_error.108  if character '~' is not followed by '0' or '1'
+    */
+    static std::vector<string_t> split(const string_t& reference_string)
+    {
+        std::vector<string_t> result;
+
+        // special case: empty reference string -> no reference tokens
+        if (reference_string.empty())
+        {
+            return result;
+        }
+
+        // check if nonempty reference string begins with slash
+        if (JSON_HEDLEY_UNLIKELY(reference_string[0] != '/'))
+        {
+            JSON_THROW(detail::parse_error::create(107, 1, detail::concat("JSON pointer must be empty or begin with '/' - was: '", reference_string, "'"), nullptr));
+        }
+
+        // extract the reference tokens:
+        // - slash: position of the last read slash (or end of string)
+        // - start: position after the previous slash
+        for (
+            // search for the first slash after the first character
+            std::size_t slash = reference_string.find_first_of('/', 1),
+            // set the beginning of the first reference token
+            start = 1;
+            // we can stop if start == 0 (if slash == string_t::npos)
+            start != 0;
+            // set the beginning of the next reference token
+            // (will eventually be 0 if slash == string_t::npos)
+            start = (slash == string_t::npos) ? 0 : slash + 1,
+            // find next slash
+            slash = reference_string.find_first_of('/', start))
+        {
+            // use the text between the beginning of the reference token
+            // (start) and the last slash (slash).
+            auto reference_token = reference_string.substr(start, slash - start);
+
+            // check reference tokens are properly escaped
+            for (std::size_t pos = reference_token.find_first_of('~');
+                    pos != string_t::npos;
+                    pos = reference_token.find_first_of('~', pos + 1))
+            {
+                JSON_ASSERT(reference_token[pos] == '~');
+
+                // ~ must be followed by 0 or 1
+                if (JSON_HEDLEY_UNLIKELY(pos == reference_token.size() - 1 ||
+                                         (reference_token[pos + 1] != '0' &&
+                                          reference_token[pos + 1] != '1')))
+                {
+                    JSON_THROW(detail::parse_error::create(108, 0, "escape character '~' must be followed with '0' or '1'", nullptr));
+                }
+            }
+
+            // finally, store the reference token
+            detail::unescape(reference_token);
+            result.push_back(reference_token);
+        }
+
+        return result;
+    }
+
+  private:
+    /*!
+    @param[in] reference_string  the reference string to the current value
+    @param[in] value             the value to consider
+    @param[in,out] result        the result object to insert values to
+
+    @note Empty objects or arrays are flattened to `null`.
+    */
+    template<typename BasicJsonType>
+    static void flatten(const string_t& reference_string,
+                        const BasicJsonType& value,
+                        BasicJsonType& result)
+    {
+        switch (value.type())
+        {
+            case detail::value_t::array:
+            {
+                if (value.m_data.m_value.array->empty())
+                {
+                    // flatten empty array as null
+                    result[reference_string] = nullptr;
+                }
+                else
+                {
+                    // iterate array and use index as reference string
+                    for (std::size_t i = 0; i < value.m_data.m_value.array->size(); ++i)
+                    {
+                        flatten(detail::concat<string_t>(reference_string, '/', std::to_string(i)),
+                                value.m_data.m_value.array->operator[](i), result);
+                    }
+                }
+                break;
+            }
+
+            case detail::value_t::object:
+            {
+                if (value.m_data.m_value.object->empty())
+                {
+                    // flatten empty object as null
+                    result[reference_string] = nullptr;
+                }
+                else
+                {
+                    // iterate object and use keys as reference string
+                    for (const auto& element : *value.m_data.m_value.object)
+                    {
+                        flatten(detail::concat<string_t>(reference_string, '/', detail::escape(element.first)), element.second, result);
+                    }
+                }
+                break;
+            }
+
+            case detail::value_t::null:
+            case detail::value_t::string:
+            case detail::value_t::boolean:
+            case detail::value_t::number_integer:
+            case detail::value_t::number_unsigned:
+            case detail::value_t::number_float:
+            case detail::value_t::binary:
+            case detail::value_t::discarded:
+            default:
+            {
+                // add primitive value with its reference string
+                result[reference_string] = value;
+                break;
+            }
+        }
+    }
+
+    /*!
+    @param[in] value  flattened JSON
+
+    @return unflattened JSON
+
+    @throw parse_error.109 if array index is not a number
+    @throw type_error.314  if value is not an object
+    @throw type_error.315  if object values are not primitive
+    @throw type_error.313  if value cannot be unflattened
+    */
+    template<typename BasicJsonType>
+    static BasicJsonType
+    unflatten(const BasicJsonType& value)
+    {
+        if (JSON_HEDLEY_UNLIKELY(!value.is_object()))
+        {
+            JSON_THROW(detail::type_error::create(314, "only objects can be unflattened", &value));
+        }
+
+        BasicJsonType result;
+
+        // iterate the JSON object values
+        for (const auto& element : *value.m_data.m_value.object)
+        {
+            if (JSON_HEDLEY_UNLIKELY(!element.second.is_primitive()))
+            {
+                JSON_THROW(detail::type_error::create(315, "values in object must be primitive", &element.second));
+            }
+
+            // assign value to reference pointed to by JSON pointer; Note that if
+            // the JSON pointer is "" (i.e., points to the whole value), function
+            // get_and_create returns a reference to result itself. An assignment
+            // will then create a primitive value.
+            json_pointer(element.first).get_and_create(result) = element.second;
+        }
+
+        return result;
+    }
+
+    // can't use conversion operator because of ambiguity
+    json_pointer<string_t> convert() const&
+    {
+        json_pointer<string_t> result;
+        result.reference_tokens = reference_tokens;
+        return result;
+    }
+
+    json_pointer<string_t> convert()&&
+    {
+        json_pointer<string_t> result;
+        result.reference_tokens = std::move(reference_tokens);
+        return result;
+    }
+
+  public:
+#if JSON_HAS_THREE_WAY_COMPARISON
+    /// @brief compares two JSON pointers for equality
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_eq/
+    template<typename RefStringTypeRhs>
+    bool operator==(const json_pointer<RefStringTypeRhs>& rhs) const noexcept
+    {
+        return reference_tokens == rhs.reference_tokens;
+    }
+
+    /// @brief compares JSON pointer and string for equality
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_eq/
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.2, operator==(json_pointer))
+    bool operator==(const string_t& rhs) const
+    {
+        return *this == json_pointer(rhs);
+    }
+
+    /// @brief 3-way compares two JSON pointers
+    template<typename RefStringTypeRhs>
+    std::strong_ordering operator<=>(const json_pointer<RefStringTypeRhs>& rhs) const noexcept // *NOPAD*
+    {
+        return  reference_tokens <=> rhs.reference_tokens; // *NOPAD*
+    }
+#else
+    /// @brief compares two JSON pointers for equality
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_eq/
+    template<typename RefStringTypeLhs, typename RefStringTypeRhs>
+    // NOLINTNEXTLINE(readability-redundant-declaration)
+    friend bool operator==(const json_pointer<RefStringTypeLhs>& lhs,
+                           const json_pointer<RefStringTypeRhs>& rhs) noexcept;
+
+    /// @brief compares JSON pointer and string for equality
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_eq/
+    template<typename RefStringTypeLhs, typename StringType>
+    // NOLINTNEXTLINE(readability-redundant-declaration)
+    friend bool operator==(const json_pointer<RefStringTypeLhs>& lhs,
+                           const StringType& rhs);
+
+    /// @brief compares string and JSON pointer for equality
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_eq/
+    template<typename RefStringTypeRhs, typename StringType>
+    // NOLINTNEXTLINE(readability-redundant-declaration)
+    friend bool operator==(const StringType& lhs,
+                           const json_pointer<RefStringTypeRhs>& rhs);
+
+    /// @brief compares two JSON pointers for inequality
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_ne/
+    template<typename RefStringTypeLhs, typename RefStringTypeRhs>
+    // NOLINTNEXTLINE(readability-redundant-declaration)
+    friend bool operator!=(const json_pointer<RefStringTypeLhs>& lhs,
+                           const json_pointer<RefStringTypeRhs>& rhs) noexcept;
+
+    /// @brief compares JSON pointer and string for inequality
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_ne/
+    template<typename RefStringTypeLhs, typename StringType>
+    // NOLINTNEXTLINE(readability-redundant-declaration)
+    friend bool operator!=(const json_pointer<RefStringTypeLhs>& lhs,
+                           const StringType& rhs);
+
+    /// @brief compares string and JSON pointer for inequality
+    /// @sa https://json.nlohmann.me/api/json_pointer/operator_ne/
+    template<typename RefStringTypeRhs, typename StringType>
+    // NOLINTNEXTLINE(readability-redundant-declaration)
+    friend bool operator!=(const StringType& lhs,
+                           const json_pointer<RefStringTypeRhs>& rhs);
+
+    /// @brief compares two JSON pointer for less-than
+    template<typename RefStringTypeLhs, typename RefStringTypeRhs>
+    // NOLINTNEXTLINE(readability-redundant-declaration)
+    friend bool operator<(const json_pointer<RefStringTypeLhs>& lhs,
+                          const json_pointer<RefStringTypeRhs>& rhs) noexcept;
+#endif
+
+  private:
+    /// the reference tokens
+    std::vector<string_t> reference_tokens;
+};
+
+#if !JSON_HAS_THREE_WAY_COMPARISON
+// functions cannot be defined inside class due to ODR violations
+template<typename RefStringTypeLhs, typename RefStringTypeRhs>
+inline bool operator==(const json_pointer<RefStringTypeLhs>& lhs,
+                       const json_pointer<RefStringTypeRhs>& rhs) noexcept
+{
+    return lhs.reference_tokens == rhs.reference_tokens;
+}
+
+template<typename RefStringTypeLhs,
+         typename StringType = typename json_pointer<RefStringTypeLhs>::string_t>
+JSON_HEDLEY_DEPRECATED_FOR(3.11.2, operator==(json_pointer, json_pointer))
+inline bool operator==(const json_pointer<RefStringTypeLhs>& lhs,
+                       const StringType& rhs)
+{
+    return lhs == json_pointer<RefStringTypeLhs>(rhs);
+}
+
+template<typename RefStringTypeRhs,
+         typename StringType = typename json_pointer<RefStringTypeRhs>::string_t>
+JSON_HEDLEY_DEPRECATED_FOR(3.11.2, operator==(json_pointer, json_pointer))
+inline bool operator==(const StringType& lhs,
+                       const json_pointer<RefStringTypeRhs>& rhs)
+{
+    return json_pointer<RefStringTypeRhs>(lhs) == rhs;
+}
+
+template<typename RefStringTypeLhs, typename RefStringTypeRhs>
+inline bool operator!=(const json_pointer<RefStringTypeLhs>& lhs,
+                       const json_pointer<RefStringTypeRhs>& rhs) noexcept
+{
+    return !(lhs == rhs);
+}
+
+template<typename RefStringTypeLhs,
+         typename StringType = typename json_pointer<RefStringTypeLhs>::string_t>
+JSON_HEDLEY_DEPRECATED_FOR(3.11.2, operator!=(json_pointer, json_pointer))
+inline bool operator!=(const json_pointer<RefStringTypeLhs>& lhs,
+                       const StringType& rhs)
+{
+    return !(lhs == rhs);
+}
+
+template<typename RefStringTypeRhs,
+         typename StringType = typename json_pointer<RefStringTypeRhs>::string_t>
+JSON_HEDLEY_DEPRECATED_FOR(3.11.2, operator!=(json_pointer, json_pointer))
+inline bool operator!=(const StringType& lhs,
+                       const json_pointer<RefStringTypeRhs>& rhs)
+{
+    return !(lhs == rhs);
+}
+
+template<typename RefStringTypeLhs, typename RefStringTypeRhs>
+inline bool operator<(const json_pointer<RefStringTypeLhs>& lhs,
+                      const json_pointer<RefStringTypeRhs>& rhs) noexcept
+{
+    return lhs.reference_tokens < rhs.reference_tokens;
+}
+#endif
+
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/json_ref.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <initializer_list>
+#include <utility>
+
+// #include <nlohmann/detail/abi_macros.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+template<typename BasicJsonType>
+class json_ref
+{
+  public:
+    using value_type = BasicJsonType;
+
+    json_ref(value_type&& value)
+        : owned_value(std::move(value))
+    {}
+
+    json_ref(const value_type& value)
+        : value_ref(&value)
+    {}
+
+    json_ref(std::initializer_list<json_ref> init)
+        : owned_value(init)
+    {}
+
+    template <
+        class... Args,
+        enable_if_t<std::is_constructible<value_type, Args...>::value, int> = 0 >
+    json_ref(Args && ... args)
+        : owned_value(std::forward<Args>(args)...)
+    {}
+
+    // class should be movable only
+    json_ref(json_ref&&) noexcept = default;
+    json_ref(const json_ref&) = delete;
+    json_ref& operator=(const json_ref&) = delete;
+    json_ref& operator=(json_ref&&) = delete;
+    ~json_ref() = default;
+
+    value_type moved_or_copied() const
+    {
+        if (value_ref == nullptr)
+        {
+            return std::move(owned_value);
+        }
+        return *value_ref;
+    }
+
+    value_type const& operator*() const
+    {
+        return value_ref ? *value_ref : owned_value;
+    }
+
+    value_type const* operator->() const
+    {
+        return &** this;
+    }
+
+  private:
+    mutable value_type owned_value = nullptr;
+    value_type const* value_ref = nullptr;
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/string_concat.hpp>
+
+// #include <nlohmann/detail/string_escape.hpp>
+
+// #include <nlohmann/detail/string_utils.hpp>
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+// #include <nlohmann/detail/output/binary_writer.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <algorithm> // reverse
+#include <array> // array
+#include <map> // map
+#include <cmath> // isnan, isinf
+#include <cstdint> // uint8_t, uint16_t, uint32_t, uint64_t
+#include <cstring> // memcpy
+#include <limits> // numeric_limits
+#include <string> // string
+#include <utility> // move
+#include <vector> // vector
+
+// #include <nlohmann/detail/input/binary_reader.hpp>
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/output/output_adapters.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <algorithm> // copy
+#include <cstddef> // size_t
+#include <iterator> // back_inserter
+#include <memory> // shared_ptr, make_shared
+#include <string> // basic_string
+#include <vector> // vector
+
+#ifndef JSON_NO_IO
+    #include <ios>      // streamsize
+    #include <ostream>  // basic_ostream
+#endif  // JSON_NO_IO
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/// abstract output adapter interface
+template<typename CharType> struct output_adapter_protocol
+{
+    virtual void write_character(CharType c) = 0;
+    virtual void write_characters(const CharType* s, std::size_t length) = 0;
+    virtual ~output_adapter_protocol() = default;
+
+    output_adapter_protocol() = default;
+    output_adapter_protocol(const output_adapter_protocol&) = default;
+    output_adapter_protocol(output_adapter_protocol&&) noexcept = default;
+    output_adapter_protocol& operator=(const output_adapter_protocol&) = default;
+    output_adapter_protocol& operator=(output_adapter_protocol&&) noexcept = default;
+};
+
+/// a type to simplify interfaces
+template<typename CharType>
+using output_adapter_t = std::shared_ptr<output_adapter_protocol<CharType>>;
+
+/// output adapter for byte vectors
+template<typename CharType, typename AllocatorType = std::allocator<CharType>>
+class output_vector_adapter : public output_adapter_protocol<CharType>
+{
+  public:
+    explicit output_vector_adapter(std::vector<CharType, AllocatorType>& vec) noexcept
+        : v(vec)
+    {}
+
+    void write_character(CharType c) override
+    {
+        v.push_back(c);
+    }
+
+    JSON_HEDLEY_NON_NULL(2)
+    void write_characters(const CharType* s, std::size_t length) override
+    {
+        v.insert(v.end(), s, s + length);
+    }
+
+  private:
+    std::vector<CharType, AllocatorType>& v;
+};
+
+#ifndef JSON_NO_IO
+/// output adapter for output streams
+template<typename CharType>
+class output_stream_adapter : public output_adapter_protocol<CharType>
+{
+  public:
+    explicit output_stream_adapter(std::basic_ostream<CharType>& s) noexcept
+        : stream(s)
+    {}
+
+    void write_character(CharType c) override
+    {
+        stream.put(c);
+    }
+
+    JSON_HEDLEY_NON_NULL(2)
+    void write_characters(const CharType* s, std::size_t length) override
+    {
+        stream.write(s, static_cast<std::streamsize>(length));
+    }
+
+  private:
+    std::basic_ostream<CharType>& stream;
+};
+#endif  // JSON_NO_IO
+
+/// output adapter for basic_string
+template<typename CharType, typename StringType = std::basic_string<CharType>>
+class output_string_adapter : public output_adapter_protocol<CharType>
+{
+  public:
+    explicit output_string_adapter(StringType& s) noexcept
+        : str(s)
+    {}
+
+    void write_character(CharType c) override
+    {
+        str.push_back(c);
+    }
+
+    JSON_HEDLEY_NON_NULL(2)
+    void write_characters(const CharType* s, std::size_t length) override
+    {
+        str.append(s, length);
+    }
+
+  private:
+    StringType& str;
+};
+
+template<typename CharType, typename StringType = std::basic_string<CharType>>
+class output_adapter
+{
+  public:
+    template<typename AllocatorType = std::allocator<CharType>>
+    output_adapter(std::vector<CharType, AllocatorType>& vec)
+        : oa(std::make_shared<output_vector_adapter<CharType, AllocatorType>>(vec)) {}
+
+#ifndef JSON_NO_IO
+    output_adapter(std::basic_ostream<CharType>& s)
+        : oa(std::make_shared<output_stream_adapter<CharType>>(s)) {}
+#endif  // JSON_NO_IO
+
+    output_adapter(StringType& s)
+        : oa(std::make_shared<output_string_adapter<CharType, StringType>>(s)) {}
+
+    operator output_adapter_t<CharType>()
+    {
+        return oa;
+    }
+
+  private:
+    output_adapter_t<CharType> oa = nullptr;
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/string_concat.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/// how to encode BJData
+enum class bjdata_version_t
+{
+    draft2,
+    draft3,
+};
+
+///////////////////
+// binary writer //
+///////////////////
+
+/*!
+@brief serialization to CBOR and MessagePack values
+*/
+template<typename BasicJsonType, typename CharType>
+class binary_writer
+{
+    using string_t = typename BasicJsonType::string_t;
+    using binary_t = typename BasicJsonType::binary_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+
+  public:
+    /*!
+    @brief create a binary writer
+
+    @param[in] adapter  output adapter to write to
+    */
+    explicit binary_writer(output_adapter_t<CharType> adapter) : oa(std::move(adapter))
+    {
+        JSON_ASSERT(oa);
+    }
+
+    /*!
+    @param[in] j  JSON value to serialize
+    @pre       j.type() == value_t::object
+    */
+    void write_bson(const BasicJsonType& j)
+    {
+        switch (j.type())
+        {
+            case value_t::object:
+            {
+                write_bson_object(*j.m_data.m_value.object);
+                break;
+            }
+
+            case value_t::null:
+            case value_t::array:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                JSON_THROW(type_error::create(317, concat("to serialize to BSON, top-level type must be object, but is ", j.type_name()), &j));
+            }
+        }
+    }
+
+    /*!
+    @param[in] j  JSON value to serialize
+    */
+    void write_cbor(const BasicJsonType& j)
+    {
+        switch (j.type())
+        {
+            case value_t::null:
+            {
+                oa->write_character(to_char_type(0xF6));
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                oa->write_character(j.m_data.m_value.boolean
+                                    ? to_char_type(0xF5)
+                                    : to_char_type(0xF4));
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                if (j.m_data.m_value.number_integer >= 0)
+                {
+                    // CBOR does not differentiate between positive signed
+                    // integers and unsigned integers. Therefore, we used the
+                    // code from the value_t::number_unsigned case here.
+                    if (j.m_data.m_value.number_integer <= 0x17)
+                    {
+                        write_number(static_cast<std::uint8_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_integer <= (std::numeric_limits<std::uint8_t>::max)())
+                    {
+                        oa->write_character(to_char_type(0x18));
+                        write_number(static_cast<std::uint8_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_integer <= (std::numeric_limits<std::uint16_t>::max)())
+                    {
+                        oa->write_character(to_char_type(0x19));
+                        write_number(static_cast<std::uint16_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_integer <= (std::numeric_limits<std::uint32_t>::max)())
+                    {
+                        oa->write_character(to_char_type(0x1A));
+                        write_number(static_cast<std::uint32_t>(j.m_data.m_value.number_integer));
+                    }
+                    else
+                    {
+                        oa->write_character(to_char_type(0x1B));
+                        write_number(static_cast<std::uint64_t>(j.m_data.m_value.number_integer));
+                    }
+                }
+                else
+                {
+                    // The conversions below encode the sign in the first
+                    // byte, and the value is converted to a positive number.
+                    const auto positive_number = -1 - j.m_data.m_value.number_integer;
+                    if (j.m_data.m_value.number_integer >= -24)
+                    {
+                        write_number(static_cast<std::uint8_t>(0x20 + positive_number));
+                    }
+                    else if (positive_number <= (std::numeric_limits<std::uint8_t>::max)())
+                    {
+                        oa->write_character(to_char_type(0x38));
+                        write_number(static_cast<std::uint8_t>(positive_number));
+                    }
+                    else if (positive_number <= (std::numeric_limits<std::uint16_t>::max)())
+                    {
+                        oa->write_character(to_char_type(0x39));
+                        write_number(static_cast<std::uint16_t>(positive_number));
+                    }
+                    else if (positive_number <= (std::numeric_limits<std::uint32_t>::max)())
+                    {
+                        oa->write_character(to_char_type(0x3A));
+                        write_number(static_cast<std::uint32_t>(positive_number));
+                    }
+                    else
+                    {
+                        oa->write_character(to_char_type(0x3B));
+                        write_number(static_cast<std::uint64_t>(positive_number));
+                    }
+                }
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                if (j.m_data.m_value.number_unsigned <= 0x17)
+                {
+                    write_number(static_cast<std::uint8_t>(j.m_data.m_value.number_unsigned));
+                }
+                else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x18));
+                    write_number(static_cast<std::uint8_t>(j.m_data.m_value.number_unsigned));
+                }
+                else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x19));
+                    write_number(static_cast<std::uint16_t>(j.m_data.m_value.number_unsigned));
+                }
+                else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x1A));
+                    write_number(static_cast<std::uint32_t>(j.m_data.m_value.number_unsigned));
+                }
+                else
+                {
+                    oa->write_character(to_char_type(0x1B));
+                    write_number(static_cast<std::uint64_t>(j.m_data.m_value.number_unsigned));
+                }
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                if (std::isnan(j.m_data.m_value.number_float))
+                {
+                    // NaN is 0xf97e00 in CBOR
+                    oa->write_character(to_char_type(0xF9));
+                    oa->write_character(to_char_type(0x7E));
+                    oa->write_character(to_char_type(0x00));
+                }
+                else if (std::isinf(j.m_data.m_value.number_float))
+                {
+                    // Infinity is 0xf97c00, -Infinity is 0xf9fc00
+                    oa->write_character(to_char_type(0xf9));
+                    oa->write_character(j.m_data.m_value.number_float > 0 ? to_char_type(0x7C) : to_char_type(0xFC));
+                    oa->write_character(to_char_type(0x00));
+                }
+                else
+                {
+                    write_compact_float(j.m_data.m_value.number_float, detail::input_format_t::cbor);
+                }
+                break;
+            }
+
+            case value_t::string:
+            {
+                // step 1: write control byte and the string length
+                const auto N = j.m_data.m_value.string->size();
+                if (N <= 0x17)
+                {
+                    write_number(static_cast<std::uint8_t>(0x60 + N));
+                }
+                else if (N <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x78));
+                    write_number(static_cast<std::uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x79));
+                    write_number(static_cast<std::uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x7A));
+                    write_number(static_cast<std::uint32_t>(N));
+                }
+                // LCOV_EXCL_START
+                else if (N <= (std::numeric_limits<std::uint64_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x7B));
+                    write_number(static_cast<std::uint64_t>(N));
+                }
+                // LCOV_EXCL_STOP
+
+                // step 2: write the string
+                oa->write_characters(
+                    reinterpret_cast<const CharType*>(j.m_data.m_value.string->c_str()),
+                    j.m_data.m_value.string->size());
+                break;
+            }
+
+            case value_t::array:
+            {
+                // step 1: write control byte and the array size
+                const auto N = j.m_data.m_value.array->size();
+                if (N <= 0x17)
+                {
+                    write_number(static_cast<std::uint8_t>(0x80 + N));
+                }
+                else if (N <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x98));
+                    write_number(static_cast<std::uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x99));
+                    write_number(static_cast<std::uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x9A));
+                    write_number(static_cast<std::uint32_t>(N));
+                }
+                // LCOV_EXCL_START
+                else if (N <= (std::numeric_limits<std::uint64_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x9B));
+                    write_number(static_cast<std::uint64_t>(N));
+                }
+                // LCOV_EXCL_STOP
+
+                // step 2: write each element
+                for (const auto& el : *j.m_data.m_value.array)
+                {
+                    write_cbor(el);
+                }
+                break;
+            }
+
+            case value_t::binary:
+            {
+                if (j.m_data.m_value.binary->has_subtype())
+                {
+                    if (j.m_data.m_value.binary->subtype() <= (std::numeric_limits<std::uint8_t>::max)())
+                    {
+                        write_number(static_cast<std::uint8_t>(0xd8));
+                        write_number(static_cast<std::uint8_t>(j.m_data.m_value.binary->subtype()));
+                    }
+                    else if (j.m_data.m_value.binary->subtype() <= (std::numeric_limits<std::uint16_t>::max)())
+                    {
+                        write_number(static_cast<std::uint8_t>(0xd9));
+                        write_number(static_cast<std::uint16_t>(j.m_data.m_value.binary->subtype()));
+                    }
+                    else if (j.m_data.m_value.binary->subtype() <= (std::numeric_limits<std::uint32_t>::max)())
+                    {
+                        write_number(static_cast<std::uint8_t>(0xda));
+                        write_number(static_cast<std::uint32_t>(j.m_data.m_value.binary->subtype()));
+                    }
+                    else if (j.m_data.m_value.binary->subtype() <= (std::numeric_limits<std::uint64_t>::max)())
+                    {
+                        write_number(static_cast<std::uint8_t>(0xdb));
+                        write_number(static_cast<std::uint64_t>(j.m_data.m_value.binary->subtype()));
+                    }
+                }
+
+                // step 1: write control byte and the binary array size
+                const auto N = j.m_data.m_value.binary->size();
+                if (N <= 0x17)
+                {
+                    write_number(static_cast<std::uint8_t>(0x40 + N));
+                }
+                else if (N <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x58));
+                    write_number(static_cast<std::uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x59));
+                    write_number(static_cast<std::uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x5A));
+                    write_number(static_cast<std::uint32_t>(N));
+                }
+                // LCOV_EXCL_START
+                else if (N <= (std::numeric_limits<std::uint64_t>::max)())
+                {
+                    oa->write_character(to_char_type(0x5B));
+                    write_number(static_cast<std::uint64_t>(N));
+                }
+                // LCOV_EXCL_STOP
+
+                // step 2: write each element
+                oa->write_characters(
+                    reinterpret_cast<const CharType*>(j.m_data.m_value.binary->data()),
+                    N);
+
+                break;
+            }
+
+            case value_t::object:
+            {
+                // step 1: write control byte and the object size
+                const auto N = j.m_data.m_value.object->size();
+                if (N <= 0x17)
+                {
+                    write_number(static_cast<std::uint8_t>(0xA0 + N));
+                }
+                else if (N <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    oa->write_character(to_char_type(0xB8));
+                    write_number(static_cast<std::uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    oa->write_character(to_char_type(0xB9));
+                    write_number(static_cast<std::uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    oa->write_character(to_char_type(0xBA));
+                    write_number(static_cast<std::uint32_t>(N));
+                }
+                // LCOV_EXCL_START
+                else if (N <= (std::numeric_limits<std::uint64_t>::max)())
+                {
+                    oa->write_character(to_char_type(0xBB));
+                    write_number(static_cast<std::uint64_t>(N));
+                }
+                // LCOV_EXCL_STOP
+
+                // step 2: write each element
+                for (const auto& el : *j.m_data.m_value.object)
+                {
+                    write_cbor(el.first);
+                    write_cbor(el.second);
+                }
+                break;
+            }
+
+            case value_t::discarded:
+            default:
+                break;
+        }
+    }
+
+    /*!
+    @param[in] j  JSON value to serialize
+    */
+    void write_msgpack(const BasicJsonType& j)
+    {
+        switch (j.type())
+        {
+            case value_t::null: // nil
+            {
+                oa->write_character(to_char_type(0xC0));
+                break;
+            }
+
+            case value_t::boolean: // true and false
+            {
+                oa->write_character(j.m_data.m_value.boolean
+                                    ? to_char_type(0xC3)
+                                    : to_char_type(0xC2));
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                if (j.m_data.m_value.number_integer >= 0)
+                {
+                    // MessagePack does not differentiate between positive
+                    // signed integers and unsigned integers. Therefore, we used
+                    // the code from the value_t::number_unsigned case here.
+                    if (j.m_data.m_value.number_unsigned < 128)
+                    {
+                        // positive fixnum
+                        write_number(static_cast<std::uint8_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint8_t>::max)())
+                    {
+                        // uint 8
+                        oa->write_character(to_char_type(0xCC));
+                        write_number(static_cast<std::uint8_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint16_t>::max)())
+                    {
+                        // uint 16
+                        oa->write_character(to_char_type(0xCD));
+                        write_number(static_cast<std::uint16_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint32_t>::max)())
+                    {
+                        // uint 32
+                        oa->write_character(to_char_type(0xCE));
+                        write_number(static_cast<std::uint32_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint64_t>::max)())
+                    {
+                        // uint 64
+                        oa->write_character(to_char_type(0xCF));
+                        write_number(static_cast<std::uint64_t>(j.m_data.m_value.number_integer));
+                    }
+                }
+                else
+                {
+                    if (j.m_data.m_value.number_integer >= -32)
+                    {
+                        // negative fixnum
+                        write_number(static_cast<std::int8_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int8_t>::min)() &&
+                             j.m_data.m_value.number_integer <= (std::numeric_limits<std::int8_t>::max)())
+                    {
+                        // int 8
+                        oa->write_character(to_char_type(0xD0));
+                        write_number(static_cast<std::int8_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int16_t>::min)() &&
+                             j.m_data.m_value.number_integer <= (std::numeric_limits<std::int16_t>::max)())
+                    {
+                        // int 16
+                        oa->write_character(to_char_type(0xD1));
+                        write_number(static_cast<std::int16_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int32_t>::min)() &&
+                             j.m_data.m_value.number_integer <= (std::numeric_limits<std::int32_t>::max)())
+                    {
+                        // int 32
+                        oa->write_character(to_char_type(0xD2));
+                        write_number(static_cast<std::int32_t>(j.m_data.m_value.number_integer));
+                    }
+                    else if (j.m_data.m_value.number_integer >= (std::numeric_limits<std::int64_t>::min)() &&
+                             j.m_data.m_value.number_integer <= (std::numeric_limits<std::int64_t>::max)())
+                    {
+                        // int 64
+                        oa->write_character(to_char_type(0xD3));
+                        write_number(static_cast<std::int64_t>(j.m_data.m_value.number_integer));
+                    }
+                }
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                if (j.m_data.m_value.number_unsigned < 128)
+                {
+                    // positive fixnum
+                    write_number(static_cast<std::uint8_t>(j.m_data.m_value.number_integer));
+                }
+                else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    // uint 8
+                    oa->write_character(to_char_type(0xCC));
+                    write_number(static_cast<std::uint8_t>(j.m_data.m_value.number_integer));
+                }
+                else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    // uint 16
+                    oa->write_character(to_char_type(0xCD));
+                    write_number(static_cast<std::uint16_t>(j.m_data.m_value.number_integer));
+                }
+                else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    // uint 32
+                    oa->write_character(to_char_type(0xCE));
+                    write_number(static_cast<std::uint32_t>(j.m_data.m_value.number_integer));
+                }
+                else if (j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint64_t>::max)())
+                {
+                    // uint 64
+                    oa->write_character(to_char_type(0xCF));
+                    write_number(static_cast<std::uint64_t>(j.m_data.m_value.number_integer));
+                }
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                write_compact_float(j.m_data.m_value.number_float, detail::input_format_t::msgpack);
+                break;
+            }
+
+            case value_t::string:
+            {
+                // step 1: write control byte and the string length
+                const auto N = j.m_data.m_value.string->size();
+                if (N <= 31)
+                {
+                    // fixstr
+                    write_number(static_cast<std::uint8_t>(0xA0 | N));
+                }
+                else if (N <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    // str 8
+                    oa->write_character(to_char_type(0xD9));
+                    write_number(static_cast<std::uint8_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    // str 16
+                    oa->write_character(to_char_type(0xDA));
+                    write_number(static_cast<std::uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    // str 32
+                    oa->write_character(to_char_type(0xDB));
+                    write_number(static_cast<std::uint32_t>(N));
+                }
+
+                // step 2: write the string
+                oa->write_characters(
+                    reinterpret_cast<const CharType*>(j.m_data.m_value.string->c_str()),
+                    j.m_data.m_value.string->size());
+                break;
+            }
+
+            case value_t::array:
+            {
+                // step 1: write control byte and the array size
+                const auto N = j.m_data.m_value.array->size();
+                if (N <= 15)
+                {
+                    // fixarray
+                    write_number(static_cast<std::uint8_t>(0x90 | N));
+                }
+                else if (N <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    // array 16
+                    oa->write_character(to_char_type(0xDC));
+                    write_number(static_cast<std::uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    // array 32
+                    oa->write_character(to_char_type(0xDD));
+                    write_number(static_cast<std::uint32_t>(N));
+                }
+
+                // step 2: write each element
+                for (const auto& el : *j.m_data.m_value.array)
+                {
+                    write_msgpack(el);
+                }
+                break;
+            }
+
+            case value_t::binary:
+            {
+                // step 0: determine if the binary type has a set subtype to
+                // determine whether to use the ext or fixext types
+                const bool use_ext = j.m_data.m_value.binary->has_subtype();
+
+                // step 1: write control byte and the byte string length
+                const auto N = j.m_data.m_value.binary->size();
+                if (N <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    std::uint8_t output_type{};
+                    bool fixed = true;
+                    if (use_ext)
+                    {
+                        switch (N)
+                        {
+                            case 1:
+                                output_type = 0xD4; // fixext 1
+                                break;
+                            case 2:
+                                output_type = 0xD5; // fixext 2
+                                break;
+                            case 4:
+                                output_type = 0xD6; // fixext 4
+                                break;
+                            case 8:
+                                output_type = 0xD7; // fixext 8
+                                break;
+                            case 16:
+                                output_type = 0xD8; // fixext 16
+                                break;
+                            default:
+                                output_type = 0xC7; // ext 8
+                                fixed = false;
+                                break;
+                        }
+
+                    }
+                    else
+                    {
+                        output_type = 0xC4; // bin 8
+                        fixed = false;
+                    }
+
+                    oa->write_character(to_char_type(output_type));
+                    if (!fixed)
+                    {
+                        write_number(static_cast<std::uint8_t>(N));
+                    }
+                }
+                else if (N <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    const std::uint8_t output_type = use_ext
+                                                     ? 0xC8 // ext 16
+                                                     : 0xC5; // bin 16
+
+                    oa->write_character(to_char_type(output_type));
+                    write_number(static_cast<std::uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    const std::uint8_t output_type = use_ext
+                                                     ? 0xC9 // ext 32
+                                                     : 0xC6; // bin 32
+
+                    oa->write_character(to_char_type(output_type));
+                    write_number(static_cast<std::uint32_t>(N));
+                }
+
+                // step 1.5: if this is an ext type, write the subtype
+                if (use_ext)
+                {
+                    write_number(static_cast<std::int8_t>(j.m_data.m_value.binary->subtype()));
+                }
+
+                // step 2: write the byte string
+                oa->write_characters(
+                    reinterpret_cast<const CharType*>(j.m_data.m_value.binary->data()),
+                    N);
+
+                break;
+            }
+
+            case value_t::object:
+            {
+                // step 1: write control byte and the object size
+                const auto N = j.m_data.m_value.object->size();
+                if (N <= 15)
+                {
+                    // fixmap
+                    write_number(static_cast<std::uint8_t>(0x80 | (N & 0xF)));
+                }
+                else if (N <= (std::numeric_limits<std::uint16_t>::max)())
+                {
+                    // map 16
+                    oa->write_character(to_char_type(0xDE));
+                    write_number(static_cast<std::uint16_t>(N));
+                }
+                else if (N <= (std::numeric_limits<std::uint32_t>::max)())
+                {
+                    // map 32
+                    oa->write_character(to_char_type(0xDF));
+                    write_number(static_cast<std::uint32_t>(N));
+                }
+
+                // step 2: write each element
+                for (const auto& el : *j.m_data.m_value.object)
+                {
+                    write_msgpack(el.first);
+                    write_msgpack(el.second);
+                }
+                break;
+            }
+
+            case value_t::discarded:
+            default:
+                break;
+        }
+    }
+
+    /*!
+    @param[in] j  JSON value to serialize
+    @param[in] use_count   whether to use '#' prefixes (optimized format)
+    @param[in] use_type    whether to use '$' prefixes (optimized format)
+    @param[in] add_prefix  whether prefixes need to be used for this value
+    @param[in] use_bjdata  whether write in BJData format, default is false
+    @param[in] bjdata_version  which BJData version to use, default is draft2
+    */
+    void write_ubjson(const BasicJsonType& j, const bool use_count,
+                      const bool use_type, const bool add_prefix = true,
+                      const bool use_bjdata = false, const bjdata_version_t bjdata_version = bjdata_version_t::draft2)
+    {
+        const bool bjdata_draft3 = use_bjdata && bjdata_version == bjdata_version_t::draft3;
+
+        switch (j.type())
+        {
+            case value_t::null:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(to_char_type('Z'));
+                }
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(j.m_data.m_value.boolean
+                                        ? to_char_type('T')
+                                        : to_char_type('F'));
+                }
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                write_number_with_ubjson_prefix(j.m_data.m_value.number_integer, add_prefix, use_bjdata);
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                write_number_with_ubjson_prefix(j.m_data.m_value.number_unsigned, add_prefix, use_bjdata);
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                write_number_with_ubjson_prefix(j.m_data.m_value.number_float, add_prefix, use_bjdata);
+                break;
+            }
+
+            case value_t::string:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(to_char_type('S'));
+                }
+                write_number_with_ubjson_prefix(j.m_data.m_value.string->size(), true, use_bjdata);
+                oa->write_characters(
+                    reinterpret_cast<const CharType*>(j.m_data.m_value.string->c_str()),
+                    j.m_data.m_value.string->size());
+                break;
+            }
+
+            case value_t::array:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(to_char_type('['));
+                }
+
+                bool prefix_required = true;
+                if (use_type && !j.m_data.m_value.array->empty())
+                {
+                    JSON_ASSERT(use_count);
+                    const CharType first_prefix = ubjson_prefix(j.front(), use_bjdata);
+                    const bool same_prefix = std::all_of(j.begin() + 1, j.end(),
+                                                         [this, first_prefix, use_bjdata](const BasicJsonType & v)
+                    {
+                        return ubjson_prefix(v, use_bjdata) == first_prefix;
+                    });
+
+                    std::vector<CharType> bjdx = {'[', '{', 'S', 'H', 'T', 'F', 'N', 'Z'}; // excluded markers in bjdata optimized type
+
+                    if (same_prefix && !(use_bjdata && std::find(bjdx.begin(), bjdx.end(), first_prefix) != bjdx.end()))
+                    {
+                        prefix_required = false;
+                        oa->write_character(to_char_type('$'));
+                        oa->write_character(first_prefix);
+                    }
+                }
+
+                if (use_count)
+                {
+                    oa->write_character(to_char_type('#'));
+                    write_number_with_ubjson_prefix(j.m_data.m_value.array->size(), true, use_bjdata);
+                }
+
+                for (const auto& el : *j.m_data.m_value.array)
+                {
+                    write_ubjson(el, use_count, use_type, prefix_required, use_bjdata, bjdata_version);
+                }
+
+                if (!use_count)
+                {
+                    oa->write_character(to_char_type(']'));
+                }
+
+                break;
+            }
+
+            case value_t::binary:
+            {
+                if (add_prefix)
+                {
+                    oa->write_character(to_char_type('['));
+                }
+
+                if (use_type && (bjdata_draft3 || !j.m_data.m_value.binary->empty()))
+                {
+                    JSON_ASSERT(use_count);
+                    oa->write_character(to_char_type('$'));
+                    oa->write_character(bjdata_draft3 ? 'B' : 'U');
+                }
+
+                if (use_count)
+                {
+                    oa->write_character(to_char_type('#'));
+                    write_number_with_ubjson_prefix(j.m_data.m_value.binary->size(), true, use_bjdata);
+                }
+
+                if (use_type)
+                {
+                    oa->write_characters(
+                        reinterpret_cast<const CharType*>(j.m_data.m_value.binary->data()),
+                        j.m_data.m_value.binary->size());
+                }
+                else
+                {
+                    for (size_t i = 0; i < j.m_data.m_value.binary->size(); ++i)
+                    {
+                        oa->write_character(to_char_type(bjdata_draft3 ? 'B' : 'U'));
+                        oa->write_character(j.m_data.m_value.binary->data()[i]);
+                    }
+                }
+
+                if (!use_count)
+                {
+                    oa->write_character(to_char_type(']'));
+                }
+
+                break;
+            }
+
+            case value_t::object:
+            {
+                if (use_bjdata && j.m_data.m_value.object->size() == 3 && j.m_data.m_value.object->find("_ArrayType_") != j.m_data.m_value.object->end() && j.m_data.m_value.object->find("_ArraySize_") != j.m_data.m_value.object->end() && j.m_data.m_value.object->find("_ArrayData_") != j.m_data.m_value.object->end())
+                {
+                    if (!write_bjdata_ndarray(*j.m_data.m_value.object, use_count, use_type, bjdata_version))  // decode bjdata ndarray in the JData format (https://github.com/NeuroJSON/jdata)
+                    {
+                        break;
+                    }
+                }
+
+                if (add_prefix)
+                {
+                    oa->write_character(to_char_type('{'));
+                }
+
+                bool prefix_required = true;
+                if (use_type && !j.m_data.m_value.object->empty())
+                {
+                    JSON_ASSERT(use_count);
+                    const CharType first_prefix = ubjson_prefix(j.front(), use_bjdata);
+                    const bool same_prefix = std::all_of(j.begin(), j.end(),
+                                                         [this, first_prefix, use_bjdata](const BasicJsonType & v)
+                    {
+                        return ubjson_prefix(v, use_bjdata) == first_prefix;
+                    });
+
+                    std::vector<CharType> bjdx = {'[', '{', 'S', 'H', 'T', 'F', 'N', 'Z'}; // excluded markers in bjdata optimized type
+
+                    if (same_prefix && !(use_bjdata && std::find(bjdx.begin(), bjdx.end(), first_prefix) != bjdx.end()))
+                    {
+                        prefix_required = false;
+                        oa->write_character(to_char_type('$'));
+                        oa->write_character(first_prefix);
+                    }
+                }
+
+                if (use_count)
+                {
+                    oa->write_character(to_char_type('#'));
+                    write_number_with_ubjson_prefix(j.m_data.m_value.object->size(), true, use_bjdata);
+                }
+
+                for (const auto& el : *j.m_data.m_value.object)
+                {
+                    write_number_with_ubjson_prefix(el.first.size(), true, use_bjdata);
+                    oa->write_characters(
+                        reinterpret_cast<const CharType*>(el.first.c_str()),
+                        el.first.size());
+                    write_ubjson(el.second, use_count, use_type, prefix_required, use_bjdata, bjdata_version);
+                }
+
+                if (!use_count)
+                {
+                    oa->write_character(to_char_type('}'));
+                }
+
+                break;
+            }
+
+            case value_t::discarded:
+            default:
+                break;
+        }
+    }
+
+  private:
+    //////////
+    // BSON //
+    //////////
+
+    /*!
+    @return The size of a BSON document entry header, including the id marker
+            and the entry name size (and its null-terminator).
+    */
+    static std::size_t calc_bson_entry_header_size(const string_t& name, const BasicJsonType& j)
+    {
+        const auto it = name.find(static_cast<typename string_t::value_type>(0));
+        if (JSON_HEDLEY_UNLIKELY(it != BasicJsonType::string_t::npos))
+        {
+            JSON_THROW(out_of_range::create(409, concat("BSON key cannot contain code point U+0000 (at byte ", std::to_string(it), ")"), &j));
+            static_cast<void>(j);
+        }
+
+        return /*id*/ 1ul + name.size() + /*zero-terminator*/1u;
+    }
+
+    /*!
+    @brief Writes the given @a element_type and @a name to the output adapter
+    */
+    void write_bson_entry_header(const string_t& name,
+                                 const std::uint8_t element_type)
+    {
+        oa->write_character(to_char_type(element_type)); // boolean
+        oa->write_characters(
+            reinterpret_cast<const CharType*>(name.c_str()),
+            name.size() + 1u);
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and boolean value @a value
+    */
+    void write_bson_boolean(const string_t& name,
+                            const bool value)
+    {
+        write_bson_entry_header(name, 0x08);
+        oa->write_character(value ? to_char_type(0x01) : to_char_type(0x00));
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and double value @a value
+    */
+    void write_bson_double(const string_t& name,
+                           const double value)
+    {
+        write_bson_entry_header(name, 0x01);
+        write_number<double>(value, true);
+    }
+
+    /*!
+    @return The size of the BSON-encoded string in @a value
+    */
+    static std::size_t calc_bson_string_size(const string_t& value)
+    {
+        return sizeof(std::int32_t) + value.size() + 1ul;
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and string value @a value
+    */
+    void write_bson_string(const string_t& name,
+                           const string_t& value)
+    {
+        write_bson_entry_header(name, 0x02);
+
+        write_number<std::int32_t>(static_cast<std::int32_t>(value.size() + 1ul), true);
+        oa->write_characters(
+            reinterpret_cast<const CharType*>(value.c_str()),
+            value.size() + 1);
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and null value
+    */
+    void write_bson_null(const string_t& name)
+    {
+        write_bson_entry_header(name, 0x0A);
+    }
+
+    /*!
+    @return The size of the BSON-encoded integer @a value
+    */
+    static std::size_t calc_bson_integer_size(const std::int64_t value)
+    {
+        return (std::numeric_limits<std::int32_t>::min)() <= value && value <= (std::numeric_limits<std::int32_t>::max)()
+               ? sizeof(std::int32_t)
+               : sizeof(std::int64_t);
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and integer @a value
+    */
+    void write_bson_integer(const string_t& name,
+                            const std::int64_t value)
+    {
+        if ((std::numeric_limits<std::int32_t>::min)() <= value && value <= (std::numeric_limits<std::int32_t>::max)())
+        {
+            write_bson_entry_header(name, 0x10); // int32
+            write_number<std::int32_t>(static_cast<std::int32_t>(value), true);
+        }
+        else
+        {
+            write_bson_entry_header(name, 0x12); // int64
+            write_number<std::int64_t>(static_cast<std::int64_t>(value), true);
+        }
+    }
+
+    /*!
+    @return The size of the BSON-encoded unsigned integer in @a j
+    */
+    static constexpr std::size_t calc_bson_unsigned_size(const std::uint64_t value) noexcept
+    {
+        return (value <= static_cast<std::uint64_t>((std::numeric_limits<std::int32_t>::max)()))
+               ? sizeof(std::int32_t)
+               : sizeof(std::int64_t);
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and unsigned @a value
+    */
+    void write_bson_unsigned(const string_t& name,
+                             const BasicJsonType& j)
+    {
+        if (j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::int32_t>::max)()))
+        {
+            write_bson_entry_header(name, 0x10 /* int32 */);
+            write_number<std::int32_t>(static_cast<std::int32_t>(j.m_data.m_value.number_unsigned), true);
+        }
+        else if (j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::int64_t>::max)()))
+        {
+            write_bson_entry_header(name, 0x12 /* int64 */);
+            write_number<std::int64_t>(static_cast<std::int64_t>(j.m_data.m_value.number_unsigned), true);
+        }
+        else
+        {
+            write_bson_entry_header(name, 0x11 /* uint64 */);
+            write_number<std::uint64_t>(static_cast<std::uint64_t>(j.m_data.m_value.number_unsigned), true);
+        }
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and object @a value
+    */
+    void write_bson_object_entry(const string_t& name,
+                                 const typename BasicJsonType::object_t& value)
+    {
+        write_bson_entry_header(name, 0x03); // object
+        write_bson_object(value);
+    }
+
+    /*!
+    @return The size of the BSON-encoded array @a value
+    */
+    static std::size_t calc_bson_array_size(const typename BasicJsonType::array_t& value)
+    {
+        std::size_t array_index = 0ul;
+
+        const std::size_t embedded_document_size = std::accumulate(std::begin(value), std::end(value), static_cast<std::size_t>(0), [&array_index](std::size_t result, const typename BasicJsonType::array_t::value_type & el)
+        {
+            return result + calc_bson_element_size(std::to_string(array_index++), el);
+        });
+
+        return sizeof(std::int32_t) + embedded_document_size + 1ul;
+    }
+
+    /*!
+    @return The size of the BSON-encoded binary array @a value
+    */
+    static std::size_t calc_bson_binary_size(const typename BasicJsonType::binary_t& value)
+    {
+        return sizeof(std::int32_t) + value.size() + 1ul;
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and array @a value
+    */
+    void write_bson_array(const string_t& name,
+                          const typename BasicJsonType::array_t& value)
+    {
+        write_bson_entry_header(name, 0x04); // array
+        write_number<std::int32_t>(static_cast<std::int32_t>(calc_bson_array_size(value)), true);
+
+        std::size_t array_index = 0ul;
+
+        for (const auto& el : value)
+        {
+            write_bson_element(std::to_string(array_index++), el);
+        }
+
+        oa->write_character(to_char_type(0x00));
+    }
+
+    /*!
+    @brief Writes a BSON element with key @a name and binary value @a value
+    */
+    void write_bson_binary(const string_t& name,
+                           const binary_t& value)
+    {
+        write_bson_entry_header(name, 0x05);
+
+        write_number<std::int32_t>(static_cast<std::int32_t>(value.size()), true);
+        write_number(value.has_subtype() ? static_cast<std::uint8_t>(value.subtype()) : static_cast<std::uint8_t>(0x00));
+
+        oa->write_characters(reinterpret_cast<const CharType*>(value.data()), value.size());
+    }
+
+    /*!
+    @brief Calculates the size necessary to serialize the JSON value @a j with its @a name
+    @return The calculated size for the BSON document entry for @a j with the given @a name.
+    */
+    static std::size_t calc_bson_element_size(const string_t& name,
+            const BasicJsonType& j)
+    {
+        const auto header_size = calc_bson_entry_header_size(name, j);
+        switch (j.type())
+        {
+            case value_t::object:
+                return header_size + calc_bson_object_size(*j.m_data.m_value.object);
+
+            case value_t::array:
+                return header_size + calc_bson_array_size(*j.m_data.m_value.array);
+
+            case value_t::binary:
+                return header_size + calc_bson_binary_size(*j.m_data.m_value.binary);
+
+            case value_t::boolean:
+                return header_size + 1ul;
+
+            case value_t::number_float:
+                return header_size + 8ul;
+
+            case value_t::number_integer:
+                return header_size + calc_bson_integer_size(j.m_data.m_value.number_integer);
+
+            case value_t::number_unsigned:
+                return header_size + calc_bson_unsigned_size(j.m_data.m_value.number_unsigned);
+
+            case value_t::string:
+                return header_size + calc_bson_string_size(*j.m_data.m_value.string);
+
+            case value_t::null:
+                return header_size + 0ul;
+
+            // LCOV_EXCL_START
+            case value_t::discarded:
+            default:
+                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert)
+                return 0ul;
+                // LCOV_EXCL_STOP
+        }
+    }
+
+    /*!
+    @brief Serializes the JSON value @a j to BSON and associates it with the
+           key @a name.
+    @param name The name to associate with the JSON entity @a j within the
+                current BSON document
+    */
+    void write_bson_element(const string_t& name,
+                            const BasicJsonType& j)
+    {
+        switch (j.type())
+        {
+            case value_t::object:
+                return write_bson_object_entry(name, *j.m_data.m_value.object);
+
+            case value_t::array:
+                return write_bson_array(name, *j.m_data.m_value.array);
+
+            case value_t::binary:
+                return write_bson_binary(name, *j.m_data.m_value.binary);
+
+            case value_t::boolean:
+                return write_bson_boolean(name, j.m_data.m_value.boolean);
+
+            case value_t::number_float:
+                return write_bson_double(name, j.m_data.m_value.number_float);
+
+            case value_t::number_integer:
+                return write_bson_integer(name, j.m_data.m_value.number_integer);
+
+            case value_t::number_unsigned:
+                return write_bson_unsigned(name, j);
+
+            case value_t::string:
+                return write_bson_string(name, *j.m_data.m_value.string);
+
+            case value_t::null:
+                return write_bson_null(name);
+
+            // LCOV_EXCL_START
+            case value_t::discarded:
+            default:
+                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert)
+                return;
+                // LCOV_EXCL_STOP
+        }
+    }
+
+    /*!
+    @brief Calculates the size of the BSON serialization of the given
+           JSON-object @a j.
+    @param[in] value  JSON value to serialize
+    @pre       value.type() == value_t::object
+    */
+    static std::size_t calc_bson_object_size(const typename BasicJsonType::object_t& value)
+    {
+        const std::size_t document_size = std::accumulate(value.begin(), value.end(), static_cast<std::size_t>(0),
+                                          [](size_t result, const typename BasicJsonType::object_t::value_type & el)
+        {
+            return result += calc_bson_element_size(el.first, el.second);
+        });
+
+        return sizeof(std::int32_t) + document_size + 1ul;
+    }
+
+    /*!
+    @param[in] value  JSON value to serialize
+    @pre       value.type() == value_t::object
+    */
+    void write_bson_object(const typename BasicJsonType::object_t& value)
+    {
+        write_number<std::int32_t>(static_cast<std::int32_t>(calc_bson_object_size(value)), true);
+
+        for (const auto& el : value)
+        {
+            write_bson_element(el.first, el.second);
+        }
+
+        oa->write_character(to_char_type(0x00));
+    }
+
+    //////////
+    // CBOR //
+    //////////
+
+    static constexpr CharType get_cbor_float_prefix(float /*unused*/)
+    {
+        return to_char_type(0xFA);  // Single-Precision Float
+    }
+
+    static constexpr CharType get_cbor_float_prefix(double /*unused*/)
+    {
+        return to_char_type(0xFB);  // Double-Precision Float
+    }
+
+    /////////////
+    // MsgPack //
+    /////////////
+
+    static constexpr CharType get_msgpack_float_prefix(float /*unused*/)
+    {
+        return to_char_type(0xCA);  // float 32
+    }
+
+    static constexpr CharType get_msgpack_float_prefix(double /*unused*/)
+    {
+        return to_char_type(0xCB);  // float 64
+    }
+
+    ////////////
+    // UBJSON //
+    ////////////
+
+    // UBJSON: write number (floating point)
+    template<typename NumberType, typename std::enable_if<
+                 std::is_floating_point<NumberType>::value, int>::type = 0>
+    void write_number_with_ubjson_prefix(const NumberType n,
+                                         const bool add_prefix,
+                                         const bool use_bjdata)
+    {
+        if (add_prefix)
+        {
+            oa->write_character(get_ubjson_float_prefix(n));
+        }
+        write_number(n, use_bjdata);
+    }
+
+    // UBJSON: write number (unsigned integer)
+    template<typename NumberType, typename std::enable_if<
+                 std::is_unsigned<NumberType>::value, int>::type = 0>
+    void write_number_with_ubjson_prefix(const NumberType n,
+                                         const bool add_prefix,
+                                         const bool use_bjdata)
+    {
+        if (n <= static_cast<std::uint64_t>((std::numeric_limits<std::int8_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('i'));  // int8
+            }
+            write_number(static_cast<std::uint8_t>(n), use_bjdata);
+        }
+        else if (n <= (std::numeric_limits<std::uint8_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('U'));  // uint8
+            }
+            write_number(static_cast<std::uint8_t>(n), use_bjdata);
+        }
+        else if (n <= static_cast<std::uint64_t>((std::numeric_limits<std::int16_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('I'));  // int16
+            }
+            write_number(static_cast<std::int16_t>(n), use_bjdata);
+        }
+        else if (use_bjdata && n <= static_cast<uint64_t>((std::numeric_limits<uint16_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('u'));  // uint16 - bjdata only
+            }
+            write_number(static_cast<std::uint16_t>(n), use_bjdata);
+        }
+        else if (n <= static_cast<std::uint64_t>((std::numeric_limits<std::int32_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('l'));  // int32
+            }
+            write_number(static_cast<std::int32_t>(n), use_bjdata);
+        }
+        else if (use_bjdata && n <= static_cast<uint64_t>((std::numeric_limits<uint32_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('m'));  // uint32 - bjdata only
+            }
+            write_number(static_cast<std::uint32_t>(n), use_bjdata);
+        }
+        else if (n <= static_cast<std::uint64_t>((std::numeric_limits<std::int64_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('L'));  // int64
+            }
+            write_number(static_cast<std::int64_t>(n), use_bjdata);
+        }
+        else if (use_bjdata && n <= (std::numeric_limits<uint64_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('M'));  // uint64 - bjdata only
+            }
+            write_number(static_cast<std::uint64_t>(n), use_bjdata);
+        }
+        else
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('H'));  // high-precision number
+            }
+
+            const auto number = BasicJsonType(n).dump();
+            write_number_with_ubjson_prefix(number.size(), true, use_bjdata);
+            for (std::size_t i = 0; i < number.size(); ++i)
+            {
+                oa->write_character(to_char_type(static_cast<std::uint8_t>(number[i])));
+            }
+        }
+    }
+
+    // UBJSON: write number (signed integer)
+    template < typename NumberType, typename std::enable_if <
+                   std::is_signed<NumberType>::value&&
+                   !std::is_floating_point<NumberType>::value, int >::type = 0 >
+    void write_number_with_ubjson_prefix(const NumberType n,
+                                         const bool add_prefix,
+                                         const bool use_bjdata)
+    {
+        if ((std::numeric_limits<std::int8_t>::min)() <= n && n <= (std::numeric_limits<std::int8_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('i'));  // int8
+            }
+            write_number(static_cast<std::int8_t>(n), use_bjdata);
+        }
+        else if (static_cast<std::int64_t>((std::numeric_limits<std::uint8_t>::min)()) <= n && n <= static_cast<std::int64_t>((std::numeric_limits<std::uint8_t>::max)()))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('U'));  // uint8
+            }
+            write_number(static_cast<std::uint8_t>(n), use_bjdata);
+        }
+        else if ((std::numeric_limits<std::int16_t>::min)() <= n && n <= (std::numeric_limits<std::int16_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('I'));  // int16
+            }
+            write_number(static_cast<std::int16_t>(n), use_bjdata);
+        }
+        else if (use_bjdata && (static_cast<std::int64_t>((std::numeric_limits<std::uint16_t>::min)()) <= n && n <= static_cast<std::int64_t>((std::numeric_limits<std::uint16_t>::max)())))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('u'));  // uint16 - bjdata only
+            }
+            write_number(static_cast<uint16_t>(n), use_bjdata);
+        }
+        else if ((std::numeric_limits<std::int32_t>::min)() <= n && n <= (std::numeric_limits<std::int32_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('l'));  // int32
+            }
+            write_number(static_cast<std::int32_t>(n), use_bjdata);
+        }
+        else if (use_bjdata && (static_cast<std::int64_t>((std::numeric_limits<std::uint32_t>::min)()) <= n && n <= static_cast<std::int64_t>((std::numeric_limits<std::uint32_t>::max)())))
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('m'));  // uint32 - bjdata only
+            }
+            write_number(static_cast<uint32_t>(n), use_bjdata);
+        }
+        else if ((std::numeric_limits<std::int64_t>::min)() <= n && n <= (std::numeric_limits<std::int64_t>::max)())
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('L'));  // int64
+            }
+            write_number(static_cast<std::int64_t>(n), use_bjdata);
+        }
+        // LCOV_EXCL_START
+        else
+        {
+            if (add_prefix)
+            {
+                oa->write_character(to_char_type('H'));  // high-precision number
+            }
+
+            const auto number = BasicJsonType(n).dump();
+            write_number_with_ubjson_prefix(number.size(), true, use_bjdata);
+            for (std::size_t i = 0; i < number.size(); ++i)
+            {
+                oa->write_character(to_char_type(static_cast<std::uint8_t>(number[i])));
+            }
+        }
+        // LCOV_EXCL_STOP
+    }
+
+    /*!
+    @brief determine the type prefix of container values
+    */
+    CharType ubjson_prefix(const BasicJsonType& j, const bool use_bjdata) const noexcept
+    {
+        switch (j.type())
+        {
+            case value_t::null:
+                return 'Z';
+
+            case value_t::boolean:
+                return j.m_data.m_value.boolean ? 'T' : 'F';
+
+            case value_t::number_integer:
+            {
+                if ((std::numeric_limits<std::int8_t>::min)() <= j.m_data.m_value.number_integer && j.m_data.m_value.number_integer <= (std::numeric_limits<std::int8_t>::max)())
+                {
+                    return 'i';
+                }
+                if ((std::numeric_limits<std::uint8_t>::min)() <= j.m_data.m_value.number_integer && j.m_data.m_value.number_integer <= (std::numeric_limits<std::uint8_t>::max)())
+                {
+                    return 'U';
+                }
+                if ((std::numeric_limits<std::int16_t>::min)() <= j.m_data.m_value.number_integer && j.m_data.m_value.number_integer <= (std::numeric_limits<std::int16_t>::max)())
+                {
+                    return 'I';
+                }
+                if (use_bjdata && ((std::numeric_limits<std::uint16_t>::min)() <= j.m_data.m_value.number_integer && j.m_data.m_value.number_integer <= (std::numeric_limits<std::uint16_t>::max)()))
+                {
+                    return 'u';
+                }
+                if ((std::numeric_limits<std::int32_t>::min)() <= j.m_data.m_value.number_integer && j.m_data.m_value.number_integer <= (std::numeric_limits<std::int32_t>::max)())
+                {
+                    return 'l';
+                }
+                if (use_bjdata && ((std::numeric_limits<std::uint32_t>::min)() <= j.m_data.m_value.number_integer && j.m_data.m_value.number_integer <= (std::numeric_limits<std::uint32_t>::max)()))
+                {
+                    return 'm';
+                }
+                if ((std::numeric_limits<std::int64_t>::min)() <= j.m_data.m_value.number_integer && j.m_data.m_value.number_integer <= (std::numeric_limits<std::int64_t>::max)())
+                {
+                    return 'L';
+                }
+                // anything else is treated as high-precision number
+                return 'H'; // LCOV_EXCL_LINE
+            }
+
+            case value_t::number_unsigned:
+            {
+                if (j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::int8_t>::max)()))
+                {
+                    return 'i';
+                }
+                if (j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::uint8_t>::max)()))
+                {
+                    return 'U';
+                }
+                if (j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::int16_t>::max)()))
+                {
+                    return 'I';
+                }
+                if (use_bjdata && j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::uint16_t>::max)()))
+                {
+                    return 'u';
+                }
+                if (j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::int32_t>::max)()))
+                {
+                    return 'l';
+                }
+                if (use_bjdata && j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::uint32_t>::max)()))
+                {
+                    return 'm';
+                }
+                if (j.m_data.m_value.number_unsigned <= static_cast<std::uint64_t>((std::numeric_limits<std::int64_t>::max)()))
+                {
+                    return 'L';
+                }
+                if (use_bjdata && j.m_data.m_value.number_unsigned <= (std::numeric_limits<std::uint64_t>::max)())
+                {
+                    return 'M';
+                }
+                // anything else is treated as high-precision number
+                return 'H'; // LCOV_EXCL_LINE
+            }
+
+            case value_t::number_float:
+                return get_ubjson_float_prefix(j.m_data.m_value.number_float);
+
+            case value_t::string:
+                return 'S';
+
+            case value_t::array: // fallthrough
+            case value_t::binary:
+                return '[';
+
+            case value_t::object:
+                return '{';
+
+            case value_t::discarded:
+            default:  // discarded values
+                return 'N';
+        }
+    }
+
+    static constexpr CharType get_ubjson_float_prefix(float /*unused*/)
+    {
+        return 'd';  // float 32
+    }
+
+    static constexpr CharType get_ubjson_float_prefix(double /*unused*/)
+    {
+        return 'D';  // float 64
+    }
+
+    /*!
+    @return false if the object is successfully converted to a bjdata ndarray, true if the type or size is invalid
+    */
+    bool write_bjdata_ndarray(const typename BasicJsonType::object_t& value, const bool use_count, const bool use_type, const bjdata_version_t bjdata_version)
+    {
+        std::map<string_t, CharType> bjdtype = {{"uint8", 'U'},  {"int8", 'i'},  {"uint16", 'u'}, {"int16", 'I'},
+            {"uint32", 'm'}, {"int32", 'l'}, {"uint64", 'M'}, {"int64", 'L'}, {"single", 'd'}, {"double", 'D'},
+            {"char", 'C'}, {"byte", 'B'}
+        };
+
+        string_t key = "_ArrayType_";
+        auto it = bjdtype.find(static_cast<string_t>(value.at(key)));
+        if (it == bjdtype.end())
+        {
+            return true;
+        }
+        CharType dtype = it->second;
+
+        key = "_ArraySize_";
+        std::size_t len = (value.at(key).empty() ? 0 : 1);
+        for (const auto& el : value.at(key))
+        {
+            len *= static_cast<std::size_t>(el.m_data.m_value.number_unsigned);
+        }
+
+        key = "_ArrayData_";
+        if (value.at(key).size() != len)
+        {
+            return true;
+        }
+
+        oa->write_character('[');
+        oa->write_character('$');
+        oa->write_character(dtype);
+        oa->write_character('#');
+
+        key = "_ArraySize_";
+        write_ubjson(value.at(key), use_count, use_type, true,  true, bjdata_version);
+
+        key = "_ArrayData_";
+        if (dtype == 'U' || dtype == 'C' || dtype == 'B')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<std::uint8_t>(el.m_data.m_value.number_unsigned), true);
+            }
+        }
+        else if (dtype == 'i')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<std::int8_t>(el.m_data.m_value.number_integer), true);
+            }
+        }
+        else if (dtype == 'u')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<std::uint16_t>(el.m_data.m_value.number_unsigned), true);
+            }
+        }
+        else if (dtype == 'I')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<std::int16_t>(el.m_data.m_value.number_integer), true);
+            }
+        }
+        else if (dtype == 'm')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<std::uint32_t>(el.m_data.m_value.number_unsigned), true);
+            }
+        }
+        else if (dtype == 'l')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<std::int32_t>(el.m_data.m_value.number_integer), true);
+            }
+        }
+        else if (dtype == 'M')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<std::uint64_t>(el.m_data.m_value.number_unsigned), true);
+            }
+        }
+        else if (dtype == 'L')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<std::int64_t>(el.m_data.m_value.number_integer), true);
+            }
+        }
+        else if (dtype == 'd')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<float>(el.m_data.m_value.number_float), true);
+            }
+        }
+        else if (dtype == 'D')
+        {
+            for (const auto& el : value.at(key))
+            {
+                write_number(static_cast<double>(el.m_data.m_value.number_float), true);
+            }
+        }
+        return false;
+    }
+
+    ///////////////////////
+    // Utility functions //
+    ///////////////////////
+
+    /*
+    @brief write a number to output input
+    @param[in] n number of type @a NumberType
+    @param[in] OutputIsLittleEndian Set to true if output data is
+                                 required to be little endian
+    @tparam NumberType the type of the number
+
+    @note This function needs to respect the system's endianness, because bytes
+          in CBOR, MessagePack, and UBJSON are stored in network order (big
+          endian) and therefore need reordering on little endian systems.
+          On the other hand, BSON and BJData use little endian and should reorder
+          on big endian systems.
+    */
+    template<typename NumberType>
+    void write_number(const NumberType n, const bool OutputIsLittleEndian = false)
+    {
+        // step 1: write number to array of length NumberType
+        std::array<CharType, sizeof(NumberType)> vec{};
+        std::memcpy(vec.data(), &n, sizeof(NumberType));
+
+        // step 2: write array to output (with possible reordering)
+        if (is_little_endian != OutputIsLittleEndian)
+        {
+            // reverse byte order prior to conversion if necessary
+            std::reverse(vec.begin(), vec.end());
+        }
+
+        oa->write_characters(vec.data(), sizeof(NumberType));
+    }
+
+    void write_compact_float(const number_float_t n, detail::input_format_t format)
+    {
+#ifdef __GNUC__
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wfloat-equal"
+#endif
+        if (static_cast<double>(n) >= static_cast<double>(std::numeric_limits<float>::lowest()) &&
+                static_cast<double>(n) <= static_cast<double>((std::numeric_limits<float>::max)()) &&
+                static_cast<double>(static_cast<float>(n)) == static_cast<double>(n))
+        {
+            oa->write_character(format == detail::input_format_t::cbor
+                                ? get_cbor_float_prefix(static_cast<float>(n))
+                                : get_msgpack_float_prefix(static_cast<float>(n)));
+            write_number(static_cast<float>(n));
+        }
+        else
+        {
+            oa->write_character(format == detail::input_format_t::cbor
+                                ? get_cbor_float_prefix(n)
+                                : get_msgpack_float_prefix(n));
+            write_number(n);
+        }
+#ifdef __GNUC__
+#pragma GCC diagnostic pop
+#endif
+    }
+
+  public:
+    // The following to_char_type functions are implement the conversion
+    // between uint8_t and CharType. In case CharType is not unsigned,
+    // such a conversion is required to allow values greater than 128.
+    // See <https://github.com/nlohmann/json/issues/1286> for a discussion.
+    template < typename C = CharType,
+               enable_if_t < std::is_signed<C>::value && std::is_signed<char>::value > * = nullptr >
+    static constexpr CharType to_char_type(std::uint8_t x) noexcept
+    {
+        return *reinterpret_cast<char*>(&x);
+    }
+
+    template < typename C = CharType,
+               enable_if_t < std::is_signed<C>::value && std::is_unsigned<char>::value > * = nullptr >
+    static CharType to_char_type(std::uint8_t x) noexcept
+    {
+        static_assert(sizeof(std::uint8_t) == sizeof(CharType), "size of CharType must be equal to std::uint8_t");
+        static_assert(std::is_trivial<CharType>::value, "CharType must be trivial");
+        CharType result;
+        std::memcpy(&result, &x, sizeof(x));
+        return result;
+    }
+
+    template<typename C = CharType,
+             enable_if_t<std::is_unsigned<C>::value>* = nullptr>
+    static constexpr CharType to_char_type(std::uint8_t x) noexcept
+    {
+        return x;
+    }
+
+    template < typename InputCharType, typename C = CharType,
+               enable_if_t <
+                   std::is_signed<C>::value &&
+                   std::is_signed<char>::value &&
+                   std::is_same<char, typename std::remove_cv<InputCharType>::type>::value
+                   > * = nullptr >
+    static constexpr CharType to_char_type(InputCharType x) noexcept
+    {
+        return x;
+    }
+
+  private:
+    /// whether we can assume little endianness
+    const bool is_little_endian = little_endianness();
+
+    /// the output
+    output_adapter_t<CharType> oa = nullptr;
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/output/output_adapters.hpp>
+
+// #include <nlohmann/detail/output/serializer.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2008 - 2009 Björn Hoehrmann <bjoern@hoehrmann.de>
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <algorithm> // reverse, remove, fill, find, none_of
+#include <array> // array
+#include <clocale> // localeconv, lconv
+#include <cmath> // labs, isfinite, isnan, signbit
+#include <cstddef> // size_t, ptrdiff_t
+#include <cstdint> // uint8_t
+#include <cstdio> // snprintf
+#include <limits> // numeric_limits
+#include <string> // string, char_traits
+#include <iomanip> // setfill, setw
+#include <type_traits> // is_same
+#include <utility> // move
+
+// #include <nlohmann/detail/conversions/to_chars.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2009 Florian Loitsch <https://florian.loitsch.com/>
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <array> // array
+#include <cmath>   // signbit, isfinite
+#include <cstdint> // intN_t, uintN_t
+#include <cstring> // memcpy, memmove
+#include <limits> // numeric_limits
+#include <type_traits> // conditional
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+/*!
+@brief implements the Grisu2 algorithm for binary to decimal floating-point
+conversion.
+
+This implementation is a slightly modified version of the reference
+implementation which may be obtained from
+http://florian.loitsch.com/publications (bench.tar.gz).
+
+The code is distributed under the MIT license, Copyright (c) 2009 Florian Loitsch.
+
+For a detailed description of the algorithm see:
+
+[1] Loitsch, "Printing Floating-Point Numbers Quickly and Accurately with
+    Integers", Proceedings of the ACM SIGPLAN 2010 Conference on Programming
+    Language Design and Implementation, PLDI 2010
+[2] Burger, Dybvig, "Printing Floating-Point Numbers Quickly and Accurately",
+    Proceedings of the ACM SIGPLAN 1996 Conference on Programming Language
+    Design and Implementation, PLDI 1996
+*/
+namespace dtoa_impl
+{
+
+template<typename Target, typename Source>
+Target reinterpret_bits(const Source source)
+{
+    static_assert(sizeof(Target) == sizeof(Source), "size mismatch");
+
+    Target target;
+    std::memcpy(&target, &source, sizeof(Source));
+    return target;
+}
+
+struct diyfp // f * 2^e
+{
+    static constexpr int kPrecision = 64; // = q
+
+    std::uint64_t f = 0;
+    int e = 0;
+
+    constexpr diyfp(std::uint64_t f_, int e_) noexcept : f(f_), e(e_) {}
+
+    /*!
+    @brief returns x - y
+    @pre x.e == y.e and x.f >= y.f
+    */
+    static diyfp sub(const diyfp& x, const diyfp& y) noexcept
+    {
+        JSON_ASSERT(x.e == y.e);
+        JSON_ASSERT(x.f >= y.f);
+
+        return {x.f - y.f, x.e};
+    }
+
+    /*!
+    @brief returns x * y
+    @note The result is rounded. (Only the upper q bits are returned.)
+    */
+    static diyfp mul(const diyfp& x, const diyfp& y) noexcept
+    {
+        static_assert(kPrecision == 64, "internal error");
+
+        // Computes:
+        //  f = round((x.f * y.f) / 2^q)
+        //  e = x.e + y.e + q
+
+        // Emulate the 64-bit * 64-bit multiplication:
+        //
+        // p = u * v
+        //   = (u_lo + 2^32 u_hi) (v_lo + 2^32 v_hi)
+        //   = (u_lo v_lo         ) + 2^32 ((u_lo v_hi         ) + (u_hi v_lo         )) + 2^64 (u_hi v_hi         )
+        //   = (p0                ) + 2^32 ((p1                ) + (p2                )) + 2^64 (p3                )
+        //   = (p0_lo + 2^32 p0_hi) + 2^32 ((p1_lo + 2^32 p1_hi) + (p2_lo + 2^32 p2_hi)) + 2^64 (p3                )
+        //   = (p0_lo             ) + 2^32 (p0_hi + p1_lo + p2_lo                      ) + 2^64 (p1_hi + p2_hi + p3)
+        //   = (p0_lo             ) + 2^32 (Q                                          ) + 2^64 (H                 )
+        //   = (p0_lo             ) + 2^32 (Q_lo + 2^32 Q_hi                           ) + 2^64 (H                 )
+        //
+        // (Since Q might be larger than 2^32 - 1)
+        //
+        //   = (p0_lo + 2^32 Q_lo) + 2^64 (Q_hi + H)
+        //
+        // (Q_hi + H does not overflow a 64-bit int)
+        //
+        //   = p_lo + 2^64 p_hi
+
+        const std::uint64_t u_lo = x.f & 0xFFFFFFFFu;
+        const std::uint64_t u_hi = x.f >> 32u;
+        const std::uint64_t v_lo = y.f & 0xFFFFFFFFu;
+        const std::uint64_t v_hi = y.f >> 32u;
+
+        const std::uint64_t p0 = u_lo * v_lo;
+        const std::uint64_t p1 = u_lo * v_hi;
+        const std::uint64_t p2 = u_hi * v_lo;
+        const std::uint64_t p3 = u_hi * v_hi;
+
+        const std::uint64_t p0_hi = p0 >> 32u;
+        const std::uint64_t p1_lo = p1 & 0xFFFFFFFFu;
+        const std::uint64_t p1_hi = p1 >> 32u;
+        const std::uint64_t p2_lo = p2 & 0xFFFFFFFFu;
+        const std::uint64_t p2_hi = p2 >> 32u;
+
+        std::uint64_t Q = p0_hi + p1_lo + p2_lo;
+
+        // The full product might now be computed as
+        //
+        // p_hi = p3 + p2_hi + p1_hi + (Q >> 32)
+        // p_lo = p0_lo + (Q << 32)
+        //
+        // But in this particular case here, the full p_lo is not required.
+        // Effectively we only need to add the highest bit in p_lo to p_hi (and
+        // Q_hi + 1 does not overflow).
+
+        Q += std::uint64_t{1} << (64u - 32u - 1u); // round, ties up
+
+        const std::uint64_t h = p3 + p2_hi + p1_hi + (Q >> 32u);
+
+        return {h, x.e + y.e + 64};
+    }
+
+    /*!
+    @brief normalize x such that the significand is >= 2^(q-1)
+    @pre x.f != 0
+    */
+    static diyfp normalize(diyfp x) noexcept
+    {
+        JSON_ASSERT(x.f != 0);
+
+        while ((x.f >> 63u) == 0)
+        {
+            x.f <<= 1u;
+            x.e--;
+        }
+
+        return x;
+    }
+
+    /*!
+    @brief normalize x such that the result has the exponent E
+    @pre e >= x.e and the upper e - x.e bits of x.f must be zero.
+    */
+    static diyfp normalize_to(const diyfp& x, const int target_exponent) noexcept
+    {
+        const int delta = x.e - target_exponent;
+
+        JSON_ASSERT(delta >= 0);
+        JSON_ASSERT(((x.f << delta) >> delta) == x.f);
+
+        return {x.f << delta, target_exponent};
+    }
+};
+
+struct boundaries
+{
+    diyfp w;
+    diyfp minus;
+    diyfp plus;
+};
+
+/*!
+Compute the (normalized) diyfp representing the input number 'value' and its
+boundaries.
+
+@pre value must be finite and positive
+*/
+template<typename FloatType>
+boundaries compute_boundaries(FloatType value)
+{
+    JSON_ASSERT(std::isfinite(value));
+    JSON_ASSERT(value > 0);
+
+    // Convert the IEEE representation into a diyfp.
+    //
+    // If v is denormal:
+    //      value = 0.F * 2^(1 - bias) = (          F) * 2^(1 - bias - (p-1))
+    // If v is normalized:
+    //      value = 1.F * 2^(E - bias) = (2^(p-1) + F) * 2^(E - bias - (p-1))
+
+    static_assert(std::numeric_limits<FloatType>::is_iec559,
+                  "internal error: dtoa_short requires an IEEE-754 floating-point implementation");
+
+    constexpr int      kPrecision = std::numeric_limits<FloatType>::digits; // = p (includes the hidden bit)
+    constexpr int      kBias      = std::numeric_limits<FloatType>::max_exponent - 1 + (kPrecision - 1);
+    constexpr int      kMinExp    = 1 - kBias;
+    constexpr std::uint64_t kHiddenBit = std::uint64_t{1} << (kPrecision - 1); // = 2^(p-1)
+
+    using bits_type = typename std::conditional<kPrecision == 24, std::uint32_t, std::uint64_t >::type;
+
+    const auto bits = static_cast<std::uint64_t>(reinterpret_bits<bits_type>(value));
+    const std::uint64_t E = bits >> (kPrecision - 1);
+    const std::uint64_t F = bits & (kHiddenBit - 1);
+
+    const bool is_denormal = E == 0;
+    const diyfp v = is_denormal
+                    ? diyfp(F, kMinExp)
+                    : diyfp(F + kHiddenBit, static_cast<int>(E) - kBias);
+
+    // Compute the boundaries m- and m+ of the floating-point value
+    // v = f * 2^e.
+    //
+    // Determine v- and v+, the floating-point predecessor and successor if v,
+    // respectively.
+    //
+    //      v- = v - 2^e        if f != 2^(p-1) or e == e_min                (A)
+    //         = v - 2^(e-1)    if f == 2^(p-1) and e > e_min                (B)
+    //
+    //      v+ = v + 2^e
+    //
+    // Let m- = (v- + v) / 2 and m+ = (v + v+) / 2. All real numbers _strictly_
+    // between m- and m+ round to v, regardless of how the input rounding
+    // algorithm breaks ties.
+    //
+    //      ---+-------------+-------------+-------------+-------------+---  (A)
+    //         v-            m-            v             m+            v+
+    //
+    //      -----------------+------+------+-------------+-------------+---  (B)
+    //                       v-     m-     v             m+            v+
+
+    const bool lower_boundary_is_closer = F == 0 && E > 1;
+    const diyfp m_plus = diyfp((2 * v.f) + 1, v.e - 1);
+    const diyfp m_minus = lower_boundary_is_closer
+                          ? diyfp((4 * v.f) - 1, v.e - 2)  // (B)
+                          : diyfp((2 * v.f) - 1, v.e - 1); // (A)
+
+    // Determine the normalized w+ = m+.
+    const diyfp w_plus = diyfp::normalize(m_plus);
+
+    // Determine w- = m- such that e_(w-) = e_(w+).
+    const diyfp w_minus = diyfp::normalize_to(m_minus, w_plus.e);
+
+    return {diyfp::normalize(v), w_minus, w_plus};
+}
+
+// Given normalized diyfp w, Grisu needs to find a (normalized) cached
+// power-of-ten c, such that the exponent of the product c * w = f * 2^e lies
+// within a certain range [alpha, gamma] (Definition 3.2 from [1])
+//
+//      alpha <= e = e_c + e_w + q <= gamma
+//
+// or
+//
+//      f_c * f_w * 2^alpha <= f_c 2^(e_c) * f_w 2^(e_w) * 2^q
+//                          <= f_c * f_w * 2^gamma
+//
+// Since c and w are normalized, i.e. 2^(q-1) <= f < 2^q, this implies
+//
+//      2^(q-1) * 2^(q-1) * 2^alpha <= c * w * 2^q < 2^q * 2^q * 2^gamma
+//
+// or
+//
+//      2^(q - 2 + alpha) <= c * w < 2^(q + gamma)
+//
+// The choice of (alpha,gamma) determines the size of the table and the form of
+// the digit generation procedure. Using (alpha,gamma)=(-60,-32) works out well
+// in practice:
+//
+// The idea is to cut the number c * w = f * 2^e into two parts, which can be
+// processed independently: An integral part p1, and a fractional part p2:
+//
+//      f * 2^e = ( (f div 2^-e) * 2^-e + (f mod 2^-e) ) * 2^e
+//              = (f div 2^-e) + (f mod 2^-e) * 2^e
+//              = p1 + p2 * 2^e
+//
+// The conversion of p1 into decimal form requires a series of divisions and
+// modulos by (a power of) 10. These operations are faster for 32-bit than for
+// 64-bit integers, so p1 should ideally fit into a 32-bit integer. This can be
+// achieved by choosing
+//
+//      -e >= 32   or   e <= -32 := gamma
+//
+// In order to convert the fractional part
+//
+//      p2 * 2^e = p2 / 2^-e = d[-1] / 10^1 + d[-2] / 10^2 + ...
+//
+// into decimal form, the fraction is repeatedly multiplied by 10 and the digits
+// d[-i] are extracted in order:
+//
+//      (10 * p2) div 2^-e = d[-1]
+//      (10 * p2) mod 2^-e = d[-2] / 10^1 + ...
+//
+// The multiplication by 10 must not overflow. It is sufficient to choose
+//
+//      10 * p2 < 16 * p2 = 2^4 * p2 <= 2^64.
+//
+// Since p2 = f mod 2^-e < 2^-e,
+//
+//      -e <= 60   or   e >= -60 := alpha
+
+constexpr int kAlpha = -60;
+constexpr int kGamma = -32;
+
+struct cached_power // c = f * 2^e ~= 10^k
+{
+    std::uint64_t f;
+    int e;
+    int k;
+};
+
+/*!
+For a normalized diyfp w = f * 2^e, this function returns a (normalized) cached
+power-of-ten c = f_c * 2^e_c, such that the exponent of the product w * c
+satisfies (Definition 3.2 from [1])
+
+     alpha <= e_c + e + q <= gamma.
+*/
+inline cached_power get_cached_power_for_binary_exponent(int e)
+{
+    // Now
+    //
+    //      alpha <= e_c + e + q <= gamma                                    (1)
+    //      ==> f_c * 2^alpha <= c * 2^e * 2^q
+    //
+    // and since the c's are normalized, 2^(q-1) <= f_c,
+    //
+    //      ==> 2^(q - 1 + alpha) <= c * 2^(e + q)
+    //      ==> 2^(alpha - e - 1) <= c
+    //
+    // If c were an exact power of ten, i.e. c = 10^k, one may determine k as
+    //
+    //      k = ceil( log_10( 2^(alpha - e - 1) ) )
+    //        = ceil( (alpha - e - 1) * log_10(2) )
+    //
+    // From the paper:
+    // "In theory the result of the procedure could be wrong since c is rounded,
+    //  and the computation itself is approximated [...]. In practice, however,
+    //  this simple function is sufficient."
+    //
+    // For IEEE double precision floating-point numbers converted into
+    // normalized diyfp's w = f * 2^e, with q = 64,
+    //
+    //      e >= -1022      (min IEEE exponent)
+    //           -52        (p - 1)
+    //           -52        (p - 1, possibly normalize denormal IEEE numbers)
+    //           -11        (normalize the diyfp)
+    //         = -1137
+    //
+    // and
+    //
+    //      e <= +1023      (max IEEE exponent)
+    //           -52        (p - 1)
+    //           -11        (normalize the diyfp)
+    //         = 960
+    //
+    // This binary exponent range [-1137,960] results in a decimal exponent
+    // range [-307,324]. One does not need to store a cached power for each
+    // k in this range. For each such k it suffices to find a cached power
+    // such that the exponent of the product lies in [alpha,gamma].
+    // This implies that the difference of the decimal exponents of adjacent
+    // table entries must be less than or equal to
+    //
+    //      floor( (gamma - alpha) * log_10(2) ) = 8.
+    //
+    // (A smaller distance gamma-alpha would require a larger table.)
+
+    // NB:
+    // Actually this function returns c, such that -60 <= e_c + e + 64 <= -34.
+
+    constexpr int kCachedPowersMinDecExp = -300;
+    constexpr int kCachedPowersDecStep = 8;
+
+    static constexpr std::array<cached_power, 79> kCachedPowers =
+    {
+        {
+            { 0xAB70FE17C79AC6CA, -1060, -300 },
+            { 0xFF77B1FCBEBCDC4F, -1034, -292 },
+            { 0xBE5691EF416BD60C, -1007, -284 },
+            { 0x8DD01FAD907FFC3C,  -980, -276 },
+            { 0xD3515C2831559A83,  -954, -268 },
+            { 0x9D71AC8FADA6C9B5,  -927, -260 },
+            { 0xEA9C227723EE8BCB,  -901, -252 },
+            { 0xAECC49914078536D,  -874, -244 },
+            { 0x823C12795DB6CE57,  -847, -236 },
+            { 0xC21094364DFB5637,  -821, -228 },
+            { 0x9096EA6F3848984F,  -794, -220 },
+            { 0xD77485CB25823AC7,  -768, -212 },
+            { 0xA086CFCD97BF97F4,  -741, -204 },
+            { 0xEF340A98172AACE5,  -715, -196 },
+            { 0xB23867FB2A35B28E,  -688, -188 },
+            { 0x84C8D4DFD2C63F3B,  -661, -180 },
+            { 0xC5DD44271AD3CDBA,  -635, -172 },
+            { 0x936B9FCEBB25C996,  -608, -164 },
+            { 0xDBAC6C247D62A584,  -582, -156 },
+            { 0xA3AB66580D5FDAF6,  -555, -148 },
+            { 0xF3E2F893DEC3F126,  -529, -140 },
+            { 0xB5B5ADA8AAFF80B8,  -502, -132 },
+            { 0x87625F056C7C4A8B,  -475, -124 },
+            { 0xC9BCFF6034C13053,  -449, -116 },
+            { 0x964E858C91BA2655,  -422, -108 },
+            { 0xDFF9772470297EBD,  -396, -100 },
+            { 0xA6DFBD9FB8E5B88F,  -369,  -92 },
+            { 0xF8A95FCF88747D94,  -343,  -84 },
+            { 0xB94470938FA89BCF,  -316,  -76 },
+            { 0x8A08F0F8BF0F156B,  -289,  -68 },
+            { 0xCDB02555653131B6,  -263,  -60 },
+            { 0x993FE2C6D07B7FAC,  -236,  -52 },
+            { 0xE45C10C42A2B3B06,  -210,  -44 },
+            { 0xAA242499697392D3,  -183,  -36 },
+            { 0xFD87B5F28300CA0E,  -157,  -28 },
+            { 0xBCE5086492111AEB,  -130,  -20 },
+            { 0x8CBCCC096F5088CC,  -103,  -12 },
+            { 0xD1B71758E219652C,   -77,   -4 },
+            { 0x9C40000000000000,   -50,    4 },
+            { 0xE8D4A51000000000,   -24,   12 },
+            { 0xAD78EBC5AC620000,     3,   20 },
+            { 0x813F3978F8940984,    30,   28 },
+            { 0xC097CE7BC90715B3,    56,   36 },
+            { 0x8F7E32CE7BEA5C70,    83,   44 },
+            { 0xD5D238A4ABE98068,   109,   52 },
+            { 0x9F4F2726179A2245,   136,   60 },
+            { 0xED63A231D4C4FB27,   162,   68 },
+            { 0xB0DE65388CC8ADA8,   189,   76 },
+            { 0x83C7088E1AAB65DB,   216,   84 },
+            { 0xC45D1DF942711D9A,   242,   92 },
+            { 0x924D692CA61BE758,   269,  100 },
+            { 0xDA01EE641A708DEA,   295,  108 },
+            { 0xA26DA3999AEF774A,   322,  116 },
+            { 0xF209787BB47D6B85,   348,  124 },
+            { 0xB454E4A179DD1877,   375,  132 },
+            { 0x865B86925B9BC5C2,   402,  140 },
+            { 0xC83553C5C8965D3D,   428,  148 },
+            { 0x952AB45CFA97A0B3,   455,  156 },
+            { 0xDE469FBD99A05FE3,   481,  164 },
+            { 0xA59BC234DB398C25,   508,  172 },
+            { 0xF6C69A72A3989F5C,   534,  180 },
+            { 0xB7DCBF5354E9BECE,   561,  188 },
+            { 0x88FCF317F22241E2,   588,  196 },
+            { 0xCC20CE9BD35C78A5,   614,  204 },
+            { 0x98165AF37B2153DF,   641,  212 },
+            { 0xE2A0B5DC971F303A,   667,  220 },
+            { 0xA8D9D1535CE3B396,   694,  228 },
+            { 0xFB9B7CD9A4A7443C,   720,  236 },
+            { 0xBB764C4CA7A44410,   747,  244 },
+            { 0x8BAB8EEFB6409C1A,   774,  252 },
+            { 0xD01FEF10A657842C,   800,  260 },
+            { 0x9B10A4E5E9913129,   827,  268 },
+            { 0xE7109BFBA19C0C9D,   853,  276 },
+            { 0xAC2820D9623BF429,   880,  284 },
+            { 0x80444B5E7AA7CF85,   907,  292 },
+            { 0xBF21E44003ACDD2D,   933,  300 },
+            { 0x8E679C2F5E44FF8F,   960,  308 },
+            { 0xD433179D9C8CB841,   986,  316 },
+            { 0x9E19DB92B4E31BA9,  1013,  324 },
+        }
+    };
+
+    // This computation gives exactly the same results for k as
+    //      k = ceil((kAlpha - e - 1) * 0.30102999566398114)
+    // for |e| <= 1500, but doesn't require floating-point operations.
+    // NB: log_10(2) ~= 78913 / 2^18
+    JSON_ASSERT(e >= -1500);
+    JSON_ASSERT(e <=  1500);
+    const int f = kAlpha - e - 1;
+    const int k = ((f * 78913) / (1 << 18)) + static_cast<int>(f > 0);
+
+    const int index = (-kCachedPowersMinDecExp + k + (kCachedPowersDecStep - 1)) / kCachedPowersDecStep;
+    JSON_ASSERT(index >= 0);
+    JSON_ASSERT(static_cast<std::size_t>(index) < kCachedPowers.size());
+
+    const cached_power cached = kCachedPowers[static_cast<std::size_t>(index)];
+    JSON_ASSERT(kAlpha <= cached.e + e + 64);
+    JSON_ASSERT(kGamma >= cached.e + e + 64);
+
+    return cached;
+}
+
+/*!
+For n != 0, returns k, such that pow10 := 10^(k-1) <= n < 10^k.
+For n == 0, returns 1 and sets pow10 := 1.
+*/
+inline int find_largest_pow10(const std::uint32_t n, std::uint32_t& pow10)
+{
+    // LCOV_EXCL_START
+    if (n >= 1000000000)
+    {
+        pow10 = 1000000000;
+        return 10;
+    }
+    // LCOV_EXCL_STOP
+    if (n >= 100000000)
+    {
+        pow10 = 100000000;
+        return  9;
+    }
+    if (n >= 10000000)
+    {
+        pow10 = 10000000;
+        return  8;
+    }
+    if (n >= 1000000)
+    {
+        pow10 = 1000000;
+        return  7;
+    }
+    if (n >= 100000)
+    {
+        pow10 = 100000;
+        return  6;
+    }
+    if (n >= 10000)
+    {
+        pow10 = 10000;
+        return  5;
+    }
+    if (n >= 1000)
+    {
+        pow10 = 1000;
+        return  4;
+    }
+    if (n >= 100)
+    {
+        pow10 = 100;
+        return  3;
+    }
+    if (n >= 10)
+    {
+        pow10 = 10;
+        return  2;
+    }
+
+    pow10 = 1;
+    return 1;
+}
+
+inline void grisu2_round(char* buf, int len, std::uint64_t dist, std::uint64_t delta,
+                         std::uint64_t rest, std::uint64_t ten_k)
+{
+    JSON_ASSERT(len >= 1);
+    JSON_ASSERT(dist <= delta);
+    JSON_ASSERT(rest <= delta);
+    JSON_ASSERT(ten_k > 0);
+
+    //               <--------------------------- delta ---->
+    //                                  <---- dist --------->
+    // --------------[------------------+-------------------]--------------
+    //               M-                 w                   M+
+    //
+    //                                  ten_k
+    //                                <------>
+    //                                       <---- rest ---->
+    // --------------[------------------+----+--------------]--------------
+    //                                  w    V
+    //                                       = buf * 10^k
+    //
+    // ten_k represents a unit-in-the-last-place in the decimal representation
+    // stored in buf.
+    // Decrement buf by ten_k while this takes buf closer to w.
+
+    // The tests are written in this order to avoid overflow in unsigned
+    // integer arithmetic.
+
+    while (rest < dist
+            && delta - rest >= ten_k
+            && (rest + ten_k < dist || dist - rest > rest + ten_k - dist))
+    {
+        JSON_ASSERT(buf[len - 1] != '0');
+        buf[len - 1]--;
+        rest += ten_k;
+    }
+}
+
+/*!
+Generates V = buffer * 10^decimal_exponent, such that M- <= V <= M+.
+M- and M+ must be normalized and share the same exponent -60 <= e <= -32.
+*/
+inline void grisu2_digit_gen(char* buffer, int& length, int& decimal_exponent,
+                             diyfp M_minus, diyfp w, diyfp M_plus)
+{
+    static_assert(kAlpha >= -60, "internal error");
+    static_assert(kGamma <= -32, "internal error");
+
+    // Generates the digits (and the exponent) of a decimal floating-point
+    // number V = buffer * 10^decimal_exponent in the range [M-, M+]. The diyfp's
+    // w, M- and M+ share the same exponent e, which satisfies alpha <= e <= gamma.
+    //
+    //               <--------------------------- delta ---->
+    //                                  <---- dist --------->
+    // --------------[------------------+-------------------]--------------
+    //               M-                 w                   M+
+    //
+    // Grisu2 generates the digits of M+ from left to right and stops as soon as
+    // V is in [M-,M+].
+
+    JSON_ASSERT(M_plus.e >= kAlpha);
+    JSON_ASSERT(M_plus.e <= kGamma);
+
+    std::uint64_t delta = diyfp::sub(M_plus, M_minus).f; // (significand of (M+ - M-), implicit exponent is e)
+    std::uint64_t dist  = diyfp::sub(M_plus, w      ).f; // (significand of (M+ - w ), implicit exponent is e)
+
+    // Split M+ = f * 2^e into two parts p1 and p2 (note: e < 0):
+    //
+    //      M+ = f * 2^e
+    //         = ((f div 2^-e) * 2^-e + (f mod 2^-e)) * 2^e
+    //         = ((p1        ) * 2^-e + (p2        )) * 2^e
+    //         = p1 + p2 * 2^e
+
+    const diyfp one(std::uint64_t{1} << -M_plus.e, M_plus.e);
+
+    auto p1 = static_cast<std::uint32_t>(M_plus.f >> -one.e); // p1 = f div 2^-e (Since -e >= 32, p1 fits into a 32-bit int.)
+    std::uint64_t p2 = M_plus.f & (one.f - 1);                    // p2 = f mod 2^-e
+
+    // 1)
+    //
+    // Generate the digits of the integral part p1 = d[n-1]...d[1]d[0]
+
+    JSON_ASSERT(p1 > 0);
+
+    std::uint32_t pow10{};
+    const int k = find_largest_pow10(p1, pow10);
+
+    //      10^(k-1) <= p1 < 10^k, pow10 = 10^(k-1)
+    //
+    //      p1 = (p1 div 10^(k-1)) * 10^(k-1) + (p1 mod 10^(k-1))
+    //         = (d[k-1]         ) * 10^(k-1) + (p1 mod 10^(k-1))
+    //
+    //      M+ = p1                                             + p2 * 2^e
+    //         = d[k-1] * 10^(k-1) + (p1 mod 10^(k-1))          + p2 * 2^e
+    //         = d[k-1] * 10^(k-1) + ((p1 mod 10^(k-1)) * 2^-e + p2) * 2^e
+    //         = d[k-1] * 10^(k-1) + (                         rest) * 2^e
+    //
+    // Now generate the digits d[n] of p1 from left to right (n = k-1,...,0)
+    //
+    //      p1 = d[k-1]...d[n] * 10^n + d[n-1]...d[0]
+    //
+    // but stop as soon as
+    //
+    //      rest * 2^e = (d[n-1]...d[0] * 2^-e + p2) * 2^e <= delta * 2^e
+
+    int n = k;
+    while (n > 0)
+    {
+        // Invariants:
+        //      M+ = buffer * 10^n + (p1 + p2 * 2^e)    (buffer = 0 for n = k)
+        //      pow10 = 10^(n-1) <= p1 < 10^n
+        //
+        const std::uint32_t d = p1 / pow10;  // d = p1 div 10^(n-1)
+        const std::uint32_t r = p1 % pow10;  // r = p1 mod 10^(n-1)
+        //
+        //      M+ = buffer * 10^n + (d * 10^(n-1) + r) + p2 * 2^e
+        //         = (buffer * 10 + d) * 10^(n-1) + (r + p2 * 2^e)
+        //
+        JSON_ASSERT(d <= 9);
+        buffer[length++] = static_cast<char>('0' + d); // buffer := buffer * 10 + d
+        //
+        //      M+ = buffer * 10^(n-1) + (r + p2 * 2^e)
+        //
+        p1 = r;
+        n--;
+        //
+        //      M+ = buffer * 10^n + (p1 + p2 * 2^e)
+        //      pow10 = 10^n
+        //
+
+        // Now check if enough digits have been generated.
+        // Compute
+        //
+        //      p1 + p2 * 2^e = (p1 * 2^-e + p2) * 2^e = rest * 2^e
+        //
+        // Note:
+        // Since rest and delta share the same exponent e, it suffices to
+        // compare the significands.
+        const std::uint64_t rest = (std::uint64_t{p1} << -one.e) + p2;
+        if (rest <= delta)
+        {
+            // V = buffer * 10^n, with M- <= V <= M+.
+
+            decimal_exponent += n;
+
+            // We may now just stop. But instead look if the buffer could be
+            // decremented to bring V closer to w.
+            //
+            // pow10 = 10^n is now 1 ulp in the decimal representation V.
+            // The rounding procedure works with diyfp's with an implicit
+            // exponent of e.
+            //
+            //      10^n = (10^n * 2^-e) * 2^e = ulp * 2^e
+            //
+            const std::uint64_t ten_n = std::uint64_t{pow10} << -one.e;
+            grisu2_round(buffer, length, dist, delta, rest, ten_n);
+
+            return;
+        }
+
+        pow10 /= 10;
+        //
+        //      pow10 = 10^(n-1) <= p1 < 10^n
+        // Invariants restored.
+    }
+
+    // 2)
+    //
+    // The digits of the integral part have been generated:
+    //
+    //      M+ = d[k-1]...d[1]d[0] + p2 * 2^e
+    //         = buffer            + p2 * 2^e
+    //
+    // Now generate the digits of the fractional part p2 * 2^e.
+    //
+    // Note:
+    // No decimal point is generated: the exponent is adjusted instead.
+    //
+    // p2 actually represents the fraction
+    //
+    //      p2 * 2^e
+    //          = p2 / 2^-e
+    //          = d[-1] / 10^1 + d[-2] / 10^2 + ...
+    //
+    // Now generate the digits d[-m] of p1 from left to right (m = 1,2,...)
+    //
+    //      p2 * 2^e = d[-1]d[-2]...d[-m] * 10^-m
+    //                      + 10^-m * (d[-m-1] / 10^1 + d[-m-2] / 10^2 + ...)
+    //
+    // using
+    //
+    //      10^m * p2 = ((10^m * p2) div 2^-e) * 2^-e + ((10^m * p2) mod 2^-e)
+    //                = (                   d) * 2^-e + (                   r)
+    //
+    // or
+    //      10^m * p2 * 2^e = d + r * 2^e
+    //
+    // i.e.
+    //
+    //      M+ = buffer + p2 * 2^e
+    //         = buffer + 10^-m * (d + r * 2^e)
+    //         = (buffer * 10^m + d) * 10^-m + 10^-m * r * 2^e
+    //
+    // and stop as soon as 10^-m * r * 2^e <= delta * 2^e
+
+    JSON_ASSERT(p2 > delta);
+
+    int m = 0;
+    for (;;)
+    {
+        // Invariant:
+        //      M+ = buffer * 10^-m + 10^-m * (d[-m-1] / 10 + d[-m-2] / 10^2 + ...) * 2^e
+        //         = buffer * 10^-m + 10^-m * (p2                                 ) * 2^e
+        //         = buffer * 10^-m + 10^-m * (1/10 * (10 * p2)                   ) * 2^e
+        //         = buffer * 10^-m + 10^-m * (1/10 * ((10*p2 div 2^-e) * 2^-e + (10*p2 mod 2^-e)) * 2^e
+        //
+        JSON_ASSERT(p2 <= (std::numeric_limits<std::uint64_t>::max)() / 10);
+        p2 *= 10;
+        const std::uint64_t d = p2 >> -one.e;     // d = (10 * p2) div 2^-e
+        const std::uint64_t r = p2 & (one.f - 1); // r = (10 * p2) mod 2^-e
+        //
+        //      M+ = buffer * 10^-m + 10^-m * (1/10 * (d * 2^-e + r) * 2^e
+        //         = buffer * 10^-m + 10^-m * (1/10 * (d + r * 2^e))
+        //         = (buffer * 10 + d) * 10^(-m-1) + 10^(-m-1) * r * 2^e
+        //
+        JSON_ASSERT(d <= 9);
+        buffer[length++] = static_cast<char>('0' + d); // buffer := buffer * 10 + d
+        //
+        //      M+ = buffer * 10^(-m-1) + 10^(-m-1) * r * 2^e
+        //
+        p2 = r;
+        m++;
+        //
+        //      M+ = buffer * 10^-m + 10^-m * p2 * 2^e
+        // Invariant restored.
+
+        // Check if enough digits have been generated.
+        //
+        //      10^-m * p2 * 2^e <= delta * 2^e
+        //              p2 * 2^e <= 10^m * delta * 2^e
+        //                    p2 <= 10^m * delta
+        delta *= 10;
+        dist  *= 10;
+        if (p2 <= delta)
+        {
+            break;
+        }
+    }
+
+    // V = buffer * 10^-m, with M- <= V <= M+.
+
+    decimal_exponent -= m;
+
+    // 1 ulp in the decimal representation is now 10^-m.
+    // Since delta and dist are now scaled by 10^m, we need to do the
+    // same with ulp in order to keep the units in sync.
+    //
+    //      10^m * 10^-m = 1 = 2^-e * 2^e = ten_m * 2^e
+    //
+    const std::uint64_t ten_m = one.f;
+    grisu2_round(buffer, length, dist, delta, p2, ten_m);
+
+    // By construction this algorithm generates the shortest possible decimal
+    // number (Loitsch, Theorem 6.2) which rounds back to w.
+    // For an input number of precision p, at least
+    //
+    //      N = 1 + ceil(p * log_10(2))
+    //
+    // decimal digits are sufficient to identify all binary floating-point
+    // numbers (Matula, "In-and-Out conversions").
+    // This implies that the algorithm does not produce more than N decimal
+    // digits.
+    //
+    //      N = 17 for p = 53 (IEEE double precision)
+    //      N = 9  for p = 24 (IEEE single precision)
+}
+
+/*!
+v = buf * 10^decimal_exponent
+len is the length of the buffer (number of decimal digits)
+The buffer must be large enough, i.e. >= max_digits10.
+*/
+JSON_HEDLEY_NON_NULL(1)
+inline void grisu2(char* buf, int& len, int& decimal_exponent,
+                   diyfp m_minus, diyfp v, diyfp m_plus)
+{
+    JSON_ASSERT(m_plus.e == m_minus.e);
+    JSON_ASSERT(m_plus.e == v.e);
+
+    //  --------(-----------------------+-----------------------)--------    (A)
+    //          m-                      v                       m+
+    //
+    //  --------------------(-----------+-----------------------)--------    (B)
+    //                      m-          v                       m+
+    //
+    // First scale v (and m- and m+) such that the exponent is in the range
+    // [alpha, gamma].
+
+    const cached_power cached = get_cached_power_for_binary_exponent(m_plus.e);
+
+    const diyfp c_minus_k(cached.f, cached.e); // = c ~= 10^-k
+
+    // The exponent of the products is = v.e + c_minus_k.e + q and is in the range [alpha,gamma]
+    const diyfp w       = diyfp::mul(v,       c_minus_k);
+    const diyfp w_minus = diyfp::mul(m_minus, c_minus_k);
+    const diyfp w_plus  = diyfp::mul(m_plus,  c_minus_k);
+
+    //  ----(---+---)---------------(---+---)---------------(---+---)----
+    //          w-                      w                       w+
+    //          = c*m-                  = c*v                   = c*m+
+    //
+    // diyfp::mul rounds its result and c_minus_k is approximated too. w, w- and
+    // w+ are now off by a small amount.
+    // In fact:
+    //
+    //      w - v * 10^k < 1 ulp
+    //
+    // To account for this inaccuracy, add resp. subtract 1 ulp.
+    //
+    //  --------+---[---------------(---+---)---------------]---+--------
+    //          w-  M-                  w                   M+  w+
+    //
+    // Now any number in [M-, M+] (bounds included) will round to w when input,
+    // regardless of how the input rounding algorithm breaks ties.
+    //
+    // And digit_gen generates the shortest possible such number in [M-, M+].
+    // Note that this does not mean that Grisu2 always generates the shortest
+    // possible number in the interval (m-, m+).
+    const diyfp M_minus(w_minus.f + 1, w_minus.e);
+    const diyfp M_plus (w_plus.f  - 1, w_plus.e );
+
+    decimal_exponent = -cached.k; // = -(-k) = k
+
+    grisu2_digit_gen(buf, len, decimal_exponent, M_minus, w, M_plus);
+}
+
+/*!
+v = buf * 10^decimal_exponent
+len is the length of the buffer (number of decimal digits)
+The buffer must be large enough, i.e. >= max_digits10.
+*/
+template<typename FloatType>
+JSON_HEDLEY_NON_NULL(1)
+void grisu2(char* buf, int& len, int& decimal_exponent, FloatType value)
+{
+    static_assert(diyfp::kPrecision >= std::numeric_limits<FloatType>::digits + 3,
+                  "internal error: not enough precision");
+
+    JSON_ASSERT(std::isfinite(value));
+    JSON_ASSERT(value > 0);
+
+    // If the neighbors (and boundaries) of 'value' are always computed for double-precision
+    // numbers, all float's can be recovered using strtod (and strtof). However, the resulting
+    // decimal representations are not exactly "short".
+    //
+    // The documentation for 'std::to_chars' (https://en.cppreference.com/w/cpp/utility/to_chars)
+    // says "value is converted to a string as if by std::sprintf in the default ("C") locale"
+    // and since sprintf promotes floats to doubles, I think this is exactly what 'std::to_chars'
+    // does.
+    // On the other hand, the documentation for 'std::to_chars' requires that "parsing the
+    // representation using the corresponding std::from_chars function recovers value exactly". That
+    // indicates that single precision floating-point numbers should be recovered using
+    // 'std::strtof'.
+    //
+    // NB: If the neighbors are computed for single-precision numbers, there is a single float
+    //     (7.0385307e-26f) which can't be recovered using strtod. The resulting double precision
+    //     value is off by 1 ulp.
+#if 0 // NOLINT(readability-avoid-unconditional-preprocessor-if)
+    const boundaries w = compute_boundaries(static_cast<double>(value));
+#else
+    const boundaries w = compute_boundaries(value);
+#endif
+
+    grisu2(buf, len, decimal_exponent, w.minus, w.w, w.plus);
+}
+
+/*!
+@brief appends a decimal representation of e to buf
+@return a pointer to the element following the exponent.
+@pre -1000 < e < 1000
+*/
+JSON_HEDLEY_NON_NULL(1)
+JSON_HEDLEY_RETURNS_NON_NULL
+inline char* append_exponent(char* buf, int e)
+{
+    JSON_ASSERT(e > -1000);
+    JSON_ASSERT(e <  1000);
+
+    if (e < 0)
+    {
+        e = -e;
+        *buf++ = '-';
+    }
+    else
+    {
+        *buf++ = '+';
+    }
+
+    auto k = static_cast<std::uint32_t>(e);
+    if (k < 10)
+    {
+        // Always print at least two digits in the exponent.
+        // This is for compatibility with printf("%g").
+        *buf++ = '0';
+        *buf++ = static_cast<char>('0' + k);
+    }
+    else if (k < 100)
+    {
+        *buf++ = static_cast<char>('0' + (k / 10));
+        k %= 10;
+        *buf++ = static_cast<char>('0' + k);
+    }
+    else
+    {
+        *buf++ = static_cast<char>('0' + (k / 100));
+        k %= 100;
+        *buf++ = static_cast<char>('0' + (k / 10));
+        k %= 10;
+        *buf++ = static_cast<char>('0' + k);
+    }
+
+    return buf;
+}
+
+/*!
+@brief prettify v = buf * 10^decimal_exponent
+
+If v is in the range [10^min_exp, 10^max_exp) it will be printed in fixed-point
+notation. Otherwise it will be printed in exponential notation.
+
+@pre min_exp < 0
+@pre max_exp > 0
+*/
+JSON_HEDLEY_NON_NULL(1)
+JSON_HEDLEY_RETURNS_NON_NULL
+inline char* format_buffer(char* buf, int len, int decimal_exponent,
+                           int min_exp, int max_exp)
+{
+    JSON_ASSERT(min_exp < 0);
+    JSON_ASSERT(max_exp > 0);
+
+    const int k = len;
+    const int n = len + decimal_exponent;
+
+    // v = buf * 10^(n-k)
+    // k is the length of the buffer (number of decimal digits)
+    // n is the position of the decimal point relative to the start of the buffer.
+
+    if (k <= n && n <= max_exp)
+    {
+        // digits[000]
+        // len <= max_exp + 2
+
+        std::memset(buf + k, '0', static_cast<size_t>(n) - static_cast<size_t>(k));
+        // Make it look like a floating-point number (#362, #378)
+        buf[n + 0] = '.';
+        buf[n + 1] = '0';
+        return buf + (static_cast<size_t>(n) + 2);
+    }
+
+    if (0 < n && n <= max_exp)
+    {
+        // dig.its
+        // len <= max_digits10 + 1
+
+        JSON_ASSERT(k > n);
+
+        std::memmove(buf + (static_cast<size_t>(n) + 1), buf + n, static_cast<size_t>(k) - static_cast<size_t>(n));
+        buf[n] = '.';
+        return buf + (static_cast<size_t>(k) + 1U);
+    }
+
+    if (min_exp < n && n <= 0)
+    {
+        // 0.[000]digits
+        // len <= 2 + (-min_exp - 1) + max_digits10
+
+        std::memmove(buf + (2 + static_cast<size_t>(-n)), buf, static_cast<size_t>(k));
+        buf[0] = '0';
+        buf[1] = '.';
+        std::memset(buf + 2, '0', static_cast<size_t>(-n));
+        return buf + (2U + static_cast<size_t>(-n) + static_cast<size_t>(k));
+    }
+
+    if (k == 1)
+    {
+        // dE+123
+        // len <= 1 + 5
+
+        buf += 1;
+    }
+    else
+    {
+        // d.igitsE+123
+        // len <= max_digits10 + 1 + 5
+
+        std::memmove(buf + 2, buf + 1, static_cast<size_t>(k) - 1);
+        buf[1] = '.';
+        buf += 1 + static_cast<size_t>(k);
+    }
+
+    *buf++ = 'e';
+    return append_exponent(buf, n - 1);
+}
+
+}  // namespace dtoa_impl
+
+/*!
+@brief generates a decimal representation of the floating-point number value in [first, last).
+
+The format of the resulting decimal representation is similar to printf's %g
+format. Returns an iterator pointing past-the-end of the decimal representation.
+
+@note The input number must be finite, i.e. NaN's and Inf's are not supported.
+@note The buffer must be large enough.
+@note The result is NOT null-terminated.
+*/
+template<typename FloatType>
+JSON_HEDLEY_NON_NULL(1, 2)
+JSON_HEDLEY_RETURNS_NON_NULL
+char* to_chars(char* first, const char* last, FloatType value)
+{
+    static_cast<void>(last); // maybe unused - fix warning
+    JSON_ASSERT(std::isfinite(value));
+
+    // Use signbit(value) instead of (value < 0) since signbit works for -0.
+    if (std::signbit(value))
+    {
+        value = -value;
+        *first++ = '-';
+    }
+
+#ifdef __GNUC__
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wfloat-equal"
+#endif
+    if (value == 0) // +-0
+    {
+        *first++ = '0';
+        // Make it look like a floating-point number (#362, #378)
+        *first++ = '.';
+        *first++ = '0';
+        return first;
+    }
+#ifdef __GNUC__
+#pragma GCC diagnostic pop
+#endif
+
+    JSON_ASSERT(last - first >= std::numeric_limits<FloatType>::max_digits10);
+
+    // Compute v = buffer * 10^decimal_exponent.
+    // The decimal digits are stored in the buffer, which needs to be interpreted
+    // as an unsigned decimal integer.
+    // len is the length of the buffer, i.e. the number of decimal digits.
+    int len = 0;
+    int decimal_exponent = 0;
+    dtoa_impl::grisu2(first, len, decimal_exponent, value);
+
+    JSON_ASSERT(len <= std::numeric_limits<FloatType>::max_digits10);
+
+    // Format the buffer like printf("%.*g", prec, value)
+    constexpr int kMinExp = -4;
+    // Use digits10 here to increase compatibility with version 2.
+    constexpr int kMaxExp = std::numeric_limits<FloatType>::digits10;
+
+    JSON_ASSERT(last - first >= kMaxExp + 2);
+    JSON_ASSERT(last - first >= 2 + (-kMinExp - 1) + std::numeric_limits<FloatType>::max_digits10);
+    JSON_ASSERT(last - first >= std::numeric_limits<FloatType>::max_digits10 + 6);
+
+    return dtoa_impl::format_buffer(first, len, decimal_exponent, kMinExp, kMaxExp);
+}
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/exceptions.hpp>
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/cpp_future.hpp>
+
+// #include <nlohmann/detail/output/binary_writer.hpp>
+
+// #include <nlohmann/detail/output/output_adapters.hpp>
+
+// #include <nlohmann/detail/string_concat.hpp>
+
+// #include <nlohmann/detail/value_t.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+namespace detail
+{
+
+///////////////////
+// serialization //
+///////////////////
+
+/// how to treat decoding errors
+enum class error_handler_t
+{
+    strict,  ///< throw a type_error exception in case of invalid UTF-8
+    replace, ///< replace invalid UTF-8 sequences with U+FFFD
+    ignore   ///< ignore invalid UTF-8 sequences
+};
+
+template<typename BasicJsonType>
+class serializer
+{
+    using string_t = typename BasicJsonType::string_t;
+    using number_float_t = typename BasicJsonType::number_float_t;
+    using number_integer_t = typename BasicJsonType::number_integer_t;
+    using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+    using binary_char_t = typename BasicJsonType::binary_t::value_type;
+    static constexpr std::uint8_t UTF8_ACCEPT = 0;
+    static constexpr std::uint8_t UTF8_REJECT = 1;
+
+  public:
+    /*!
+    @param[in] s  output stream to serialize to
+    @param[in] ichar  indentation character to use
+    @param[in] error_handler_  how to react on decoding errors
+    */
+    serializer(output_adapter_t<char> s, const char ichar,
+               error_handler_t error_handler_ = error_handler_t::strict)
+        : o(std::move(s))
+        , loc(std::localeconv())
+        , thousands_sep(loc->thousands_sep == nullptr ? '\0' : std::char_traits<char>::to_char_type(* (loc->thousands_sep)))
+        , decimal_point(loc->decimal_point == nullptr ? '\0' : std::char_traits<char>::to_char_type(* (loc->decimal_point)))
+        , indent_char(ichar)
+        , indent_string(512, indent_char)
+        , error_handler(error_handler_)
+    {}
+
+    // delete because of pointer members
+    serializer(const serializer&) = delete;
+    serializer& operator=(const serializer&) = delete;
+    serializer(serializer&&) = delete;
+    serializer& operator=(serializer&&) = delete;
+    ~serializer() = default;
+
+    /*!
+    @brief internal implementation of the serialization function
+
+    This function is called by the public member function dump and organizes
+    the serialization internally. The indentation level is propagated as
+    additional parameter. In case of arrays and objects, the function is
+    called recursively.
+
+    - strings and object keys are escaped using `escape_string()`
+    - integer numbers are converted implicitly via `operator<<`
+    - floating-point numbers are converted to a string using `"%g"` format
+    - binary values are serialized as objects containing the subtype and the
+      byte array
+
+    @param[in] val               value to serialize
+    @param[in] pretty_print      whether the output shall be pretty-printed
+    @param[in] ensure_ascii If @a ensure_ascii is true, all non-ASCII characters
+    in the output are escaped with `\uXXXX` sequences, and the result consists
+    of ASCII characters only.
+    @param[in] indent_step       the indent level
+    @param[in] current_indent    the current indent level (only used internally)
+    */
+    void dump(const BasicJsonType& val,
+              const bool pretty_print,
+              const bool ensure_ascii,
+              const unsigned int indent_step,
+              const unsigned int current_indent = 0)
+    {
+        switch (val.m_data.m_type)
+        {
+            case value_t::object:
+            {
+                if (val.m_data.m_value.object->empty())
+                {
+                    o->write_characters("{}", 2);
+                    return;
+                }
+
+                if (pretty_print)
+                {
+                    o->write_characters("{\n", 2);
+
+                    // variable to hold indentation for recursive calls
+                    const auto new_indent = current_indent + indent_step;
+                    if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
+                    {
+                        indent_string.resize(indent_string.size() * 2, ' ');
+                    }
+
+                    // first n-1 elements
+                    auto i = val.m_data.m_value.object->cbegin();
+                    for (std::size_t cnt = 0; cnt < val.m_data.m_value.object->size() - 1; ++cnt, ++i)
+                    {
+                        o->write_characters(indent_string.c_str(), new_indent);
+                        o->write_character('\"');
+                        dump_escaped(i->first, ensure_ascii);
+                        o->write_characters("\": ", 3);
+                        dump(i->second, true, ensure_ascii, indent_step, new_indent);
+                        o->write_characters(",\n", 2);
+                    }
+
+                    // last element
+                    JSON_ASSERT(i != val.m_data.m_value.object->cend());
+                    JSON_ASSERT(std::next(i) == val.m_data.m_value.object->cend());
+                    o->write_characters(indent_string.c_str(), new_indent);
+                    o->write_character('\"');
+                    dump_escaped(i->first, ensure_ascii);
+                    o->write_characters("\": ", 3);
+                    dump(i->second, true, ensure_ascii, indent_step, new_indent);
+
+                    o->write_character('\n');
+                    o->write_characters(indent_string.c_str(), current_indent);
+                    o->write_character('}');
+                }
+                else
+                {
+                    o->write_character('{');
+
+                    // first n-1 elements
+                    auto i = val.m_data.m_value.object->cbegin();
+                    for (std::size_t cnt = 0; cnt < val.m_data.m_value.object->size() - 1; ++cnt, ++i)
+                    {
+                        o->write_character('\"');
+                        dump_escaped(i->first, ensure_ascii);
+                        o->write_characters("\":", 2);
+                        dump(i->second, false, ensure_ascii, indent_step, current_indent);
+                        o->write_character(',');
+                    }
+
+                    // last element
+                    JSON_ASSERT(i != val.m_data.m_value.object->cend());
+                    JSON_ASSERT(std::next(i) == val.m_data.m_value.object->cend());
+                    o->write_character('\"');
+                    dump_escaped(i->first, ensure_ascii);
+                    o->write_characters("\":", 2);
+                    dump(i->second, false, ensure_ascii, indent_step, current_indent);
+
+                    o->write_character('}');
+                }
+
+                return;
+            }
+
+            case value_t::array:
+            {
+                if (val.m_data.m_value.array->empty())
+                {
+                    o->write_characters("[]", 2);
+                    return;
+                }
+
+                if (pretty_print)
+                {
+                    o->write_characters("[\n", 2);
+
+                    // variable to hold indentation for recursive calls
+                    const auto new_indent = current_indent + indent_step;
+                    if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
+                    {
+                        indent_string.resize(indent_string.size() * 2, ' ');
+                    }
+
+                    // first n-1 elements
+                    for (auto i = val.m_data.m_value.array->cbegin();
+                            i != val.m_data.m_value.array->cend() - 1; ++i)
+                    {
+                        o->write_characters(indent_string.c_str(), new_indent);
+                        dump(*i, true, ensure_ascii, indent_step, new_indent);
+                        o->write_characters(",\n", 2);
+                    }
+
+                    // last element
+                    JSON_ASSERT(!val.m_data.m_value.array->empty());
+                    o->write_characters(indent_string.c_str(), new_indent);
+                    dump(val.m_data.m_value.array->back(), true, ensure_ascii, indent_step, new_indent);
+
+                    o->write_character('\n');
+                    o->write_characters(indent_string.c_str(), current_indent);
+                    o->write_character(']');
+                }
+                else
+                {
+                    o->write_character('[');
+
+                    // first n-1 elements
+                    for (auto i = val.m_data.m_value.array->cbegin();
+                            i != val.m_data.m_value.array->cend() - 1; ++i)
+                    {
+                        dump(*i, false, ensure_ascii, indent_step, current_indent);
+                        o->write_character(',');
+                    }
+
+                    // last element
+                    JSON_ASSERT(!val.m_data.m_value.array->empty());
+                    dump(val.m_data.m_value.array->back(), false, ensure_ascii, indent_step, current_indent);
+
+                    o->write_character(']');
+                }
+
+                return;
+            }
+
+            case value_t::string:
+            {
+                o->write_character('\"');
+                dump_escaped(*val.m_data.m_value.string, ensure_ascii);
+                o->write_character('\"');
+                return;
+            }
+
+            case value_t::binary:
+            {
+                if (pretty_print)
+                {
+                    o->write_characters("{\n", 2);
+
+                    // variable to hold indentation for recursive calls
+                    const auto new_indent = current_indent + indent_step;
+                    if (JSON_HEDLEY_UNLIKELY(indent_string.size() < new_indent))
+                    {
+                        indent_string.resize(indent_string.size() * 2, ' ');
+                    }
+
+                    o->write_characters(indent_string.c_str(), new_indent);
+
+                    o->write_characters("\"bytes\": [", 10);
+
+                    if (!val.m_data.m_value.binary->empty())
+                    {
+                        for (auto i = val.m_data.m_value.binary->cbegin();
+                                i != val.m_data.m_value.binary->cend() - 1; ++i)
+                        {
+                            dump_integer(*i);
+                            o->write_characters(", ", 2);
+                        }
+                        dump_integer(val.m_data.m_value.binary->back());
+                    }
+
+                    o->write_characters("],\n", 3);
+                    o->write_characters(indent_string.c_str(), new_indent);
+
+                    o->write_characters("\"subtype\": ", 11);
+                    if (val.m_data.m_value.binary->has_subtype())
+                    {
+                        dump_integer(val.m_data.m_value.binary->subtype());
+                    }
+                    else
+                    {
+                        o->write_characters("null", 4);
+                    }
+                    o->write_character('\n');
+                    o->write_characters(indent_string.c_str(), current_indent);
+                    o->write_character('}');
+                }
+                else
+                {
+                    o->write_characters("{\"bytes\":[", 10);
+
+                    if (!val.m_data.m_value.binary->empty())
+                    {
+                        for (auto i = val.m_data.m_value.binary->cbegin();
+                                i != val.m_data.m_value.binary->cend() - 1; ++i)
+                        {
+                            dump_integer(*i);
+                            o->write_character(',');
+                        }
+                        dump_integer(val.m_data.m_value.binary->back());
+                    }
+
+                    o->write_characters("],\"subtype\":", 12);
+                    if (val.m_data.m_value.binary->has_subtype())
+                    {
+                        dump_integer(val.m_data.m_value.binary->subtype());
+                        o->write_character('}');
+                    }
+                    else
+                    {
+                        o->write_characters("null}", 5);
+                    }
+                }
+                return;
+            }
+
+            case value_t::boolean:
+            {
+                if (val.m_data.m_value.boolean)
+                {
+                    o->write_characters("true", 4);
+                }
+                else
+                {
+                    o->write_characters("false", 5);
+                }
+                return;
+            }
+
+            case value_t::number_integer:
+            {
+                dump_integer(val.m_data.m_value.number_integer);
+                return;
+            }
+
+            case value_t::number_unsigned:
+            {
+                dump_integer(val.m_data.m_value.number_unsigned);
+                return;
+            }
+
+            case value_t::number_float:
+            {
+                dump_float(val.m_data.m_value.number_float);
+                return;
+            }
+
+            case value_t::discarded:
+            {
+                o->write_characters("<discarded>", 11);
+                return;
+            }
+
+            case value_t::null:
+            {
+                o->write_characters("null", 4);
+                return;
+            }
+
+            default:            // LCOV_EXCL_LINE
+                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+        }
+    }
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    /*!
+    @brief dump escaped string
+
+    Escape a string by replacing certain special characters by a sequence of an
+    escape character (backslash) and another character and other control
+    characters by a sequence of "\u" followed by a four-digit hex
+    representation. The escaped string is written to output stream @a o.
+
+    @param[in] s  the string to escape
+    @param[in] ensure_ascii  whether to escape non-ASCII characters with
+                             \uXXXX sequences
+
+    @complexity Linear in the length of string @a s.
+    */
+    void dump_escaped(const string_t& s, const bool ensure_ascii)
+    {
+        std::uint32_t codepoint{};
+        std::uint8_t state = UTF8_ACCEPT;
+        std::size_t bytes = 0;  // number of bytes written to string_buffer
+
+        // number of bytes written at the point of the last valid byte
+        std::size_t bytes_after_last_accept = 0;
+        std::size_t undumped_chars = 0;
+
+        for (std::size_t i = 0; i < s.size(); ++i)
+        {
+            const auto byte = static_cast<std::uint8_t>(s[i]);
+
+            switch (decode(state, codepoint, byte))
+            {
+                case UTF8_ACCEPT:  // decode found a new code point
+                {
+                    switch (codepoint)
+                    {
+                        case 0x08: // backspace
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 'b';
+                            break;
+                        }
+
+                        case 0x09: // horizontal tab
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 't';
+                            break;
+                        }
+
+                        case 0x0A: // newline
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 'n';
+                            break;
+                        }
+
+                        case 0x0C: // formfeed
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 'f';
+                            break;
+                        }
+
+                        case 0x0D: // carriage return
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = 'r';
+                            break;
+                        }
+
+                        case 0x22: // quotation mark
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = '\"';
+                            break;
+                        }
+
+                        case 0x5C: // reverse solidus
+                        {
+                            string_buffer[bytes++] = '\\';
+                            string_buffer[bytes++] = '\\';
+                            break;
+                        }
+
+                        default:
+                        {
+                            // escape control characters (0x00..0x1F) or, if
+                            // ensure_ascii parameter is used, non-ASCII characters
+                            if ((codepoint <= 0x1F) || (ensure_ascii && (codepoint >= 0x7F)))
+                            {
+                                if (codepoint <= 0xFFFF)
+                                {
+                                    // NOLINTNEXTLINE(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+                                    static_cast<void>((std::snprintf)(string_buffer.data() + bytes, 7, "\\u%04x",
+                                                                      static_cast<std::uint16_t>(codepoint)));
+                                    bytes += 6;
+                                }
+                                else
+                                {
+                                    // NOLINTNEXTLINE(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+                                    static_cast<void>((std::snprintf)(string_buffer.data() + bytes, 13, "\\u%04x\\u%04x",
+                                                                      static_cast<std::uint16_t>(0xD7C0u + (codepoint >> 10u)),
+                                                                      static_cast<std::uint16_t>(0xDC00u + (codepoint & 0x3FFu))));
+                                    bytes += 12;
+                                }
+                            }
+                            else
+                            {
+                                // copy byte to buffer (all previous bytes
+                                // been copied have in default case above)
+                                string_buffer[bytes++] = s[i];
+                            }
+                            break;
+                        }
+                    }
+
+                    // write buffer and reset index; there must be 13 bytes
+                    // left, as this is the maximal number of bytes to be
+                    // written ("\uxxxx\uxxxx\0") for one code point
+                    if (string_buffer.size() - bytes < 13)
+                    {
+                        o->write_characters(string_buffer.data(), bytes);
+                        bytes = 0;
+                    }
+
+                    // remember the byte position of this accept
+                    bytes_after_last_accept = bytes;
+                    undumped_chars = 0;
+                    break;
+                }
+
+                case UTF8_REJECT:  // decode found invalid UTF-8 byte
+                {
+                    switch (error_handler)
+                    {
+                        case error_handler_t::strict:
+                        {
+                            JSON_THROW(type_error::create(316, concat("invalid UTF-8 byte at index ", std::to_string(i), ": 0x", hex_bytes(byte | 0)), nullptr));
+                        }
+
+                        case error_handler_t::ignore:
+                        case error_handler_t::replace:
+                        {
+                            // in case we saw this character the first time, we
+                            // would like to read it again, because the byte
+                            // may be OK for itself, but just not OK for the
+                            // previous sequence
+                            if (undumped_chars > 0)
+                            {
+                                --i;
+                            }
+
+                            // reset length buffer to the last accepted index;
+                            // thus removing/ignoring the invalid characters
+                            bytes = bytes_after_last_accept;
+
+                            if (error_handler == error_handler_t::replace)
+                            {
+                                // add a replacement character
+                                if (ensure_ascii)
+                                {
+                                    string_buffer[bytes++] = '\\';
+                                    string_buffer[bytes++] = 'u';
+                                    string_buffer[bytes++] = 'f';
+                                    string_buffer[bytes++] = 'f';
+                                    string_buffer[bytes++] = 'f';
+                                    string_buffer[bytes++] = 'd';
+                                }
+                                else
+                                {
+                                    string_buffer[bytes++] = detail::binary_writer<BasicJsonType, char>::to_char_type('\xEF');
+                                    string_buffer[bytes++] = detail::binary_writer<BasicJsonType, char>::to_char_type('\xBF');
+                                    string_buffer[bytes++] = detail::binary_writer<BasicJsonType, char>::to_char_type('\xBD');
+                                }
+
+                                // write buffer and reset index; there must be 13 bytes
+                                // left, as this is the maximal number of bytes to be
+                                // written ("\uxxxx\uxxxx\0") for one code point
+                                if (string_buffer.size() - bytes < 13)
+                                {
+                                    o->write_characters(string_buffer.data(), bytes);
+                                    bytes = 0;
+                                }
+
+                                bytes_after_last_accept = bytes;
+                            }
+
+                            undumped_chars = 0;
+
+                            // continue processing the string
+                            state = UTF8_ACCEPT;
+                            break;
+                        }
+
+                        default:            // LCOV_EXCL_LINE
+                            JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+                    }
+                    break;
+                }
+
+                default:  // decode found yet incomplete multi-byte code point
+                {
+                    if (!ensure_ascii)
+                    {
+                        // code point will not be escaped - copy byte to buffer
+                        string_buffer[bytes++] = s[i];
+                    }
+                    ++undumped_chars;
+                    break;
+                }
+            }
+        }
+
+        // we finished processing the string
+        if (JSON_HEDLEY_LIKELY(state == UTF8_ACCEPT))
+        {
+            // write buffer
+            if (bytes > 0)
+            {
+                o->write_characters(string_buffer.data(), bytes);
+            }
+        }
+        else
+        {
+            // we finish reading, but do not accept: string was incomplete
+            switch (error_handler)
+            {
+                case error_handler_t::strict:
+                {
+                    JSON_THROW(type_error::create(316, concat("incomplete UTF-8 string; last byte: 0x", hex_bytes(static_cast<std::uint8_t>(s.back() | 0))), nullptr));
+                }
+
+                case error_handler_t::ignore:
+                {
+                    // write all accepted bytes
+                    o->write_characters(string_buffer.data(), bytes_after_last_accept);
+                    break;
+                }
+
+                case error_handler_t::replace:
+                {
+                    // write all accepted bytes
+                    o->write_characters(string_buffer.data(), bytes_after_last_accept);
+                    // add a replacement character
+                    if (ensure_ascii)
+                    {
+                        o->write_characters("\\ufffd", 6);
+                    }
+                    else
+                    {
+                        o->write_characters("\xEF\xBF\xBD", 3);
+                    }
+                    break;
+                }
+
+                default:            // LCOV_EXCL_LINE
+                    JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+            }
+        }
+    }
+
+  private:
+    /*!
+    @brief count digits
+
+    Count the number of decimal (base 10) digits for an input unsigned integer.
+
+    @param[in] x  unsigned integer number to count its digits
+    @return    number of decimal digits
+    */
+    unsigned int count_digits(number_unsigned_t x) noexcept
+    {
+        unsigned int n_digits = 1;
+        for (;;)
+        {
+            if (x < 10)
+            {
+                return n_digits;
+            }
+            if (x < 100)
+            {
+                return n_digits + 1;
+            }
+            if (x < 1000)
+            {
+                return n_digits + 2;
+            }
+            if (x < 10000)
+            {
+                return n_digits + 3;
+            }
+            x = x / 10000u;
+            n_digits += 4;
+        }
+    }
+
+    /*!
+     * @brief convert a byte to a uppercase hex representation
+     * @param[in] byte byte to represent
+     * @return representation ("00".."FF")
+     */
+    static std::string hex_bytes(std::uint8_t byte)
+    {
+        std::string result = "FF";
+        constexpr const char* nibble_to_hex = "0123456789ABCDEF";
+        result[0] = nibble_to_hex[byte / 16];
+        result[1] = nibble_to_hex[byte % 16];
+        return result;
+    }
+
+    // templates to avoid warnings about useless casts
+    template <typename NumberType, enable_if_t<std::is_signed<NumberType>::value, int> = 0>
+    bool is_negative_number(NumberType x)
+    {
+        return x < 0;
+    }
+
+    template < typename NumberType, enable_if_t <std::is_unsigned<NumberType>::value, int > = 0 >
+    bool is_negative_number(NumberType /*unused*/)
+    {
+        return false;
+    }
+
+    /*!
+    @brief dump an integer
+
+    Dump a given integer to output stream @a o. Works internally with
+    @a number_buffer.
+
+    @param[in] x  integer number (signed or unsigned) to dump
+    @tparam NumberType either @a number_integer_t or @a number_unsigned_t
+    */
+    template < typename NumberType, detail::enable_if_t <
+                   std::is_integral<NumberType>::value ||
+                   std::is_same<NumberType, number_unsigned_t>::value ||
+                   std::is_same<NumberType, number_integer_t>::value ||
+                   std::is_same<NumberType, binary_char_t>::value,
+                   int > = 0 >
+    void dump_integer(NumberType x)
+    {
+        static constexpr std::array<std::array<char, 2>, 100> digits_to_99
+        {
+            {
+                {{'0', '0'}}, {{'0', '1'}}, {{'0', '2'}}, {{'0', '3'}}, {{'0', '4'}}, {{'0', '5'}}, {{'0', '6'}}, {{'0', '7'}}, {{'0', '8'}}, {{'0', '9'}},
+                {{'1', '0'}}, {{'1', '1'}}, {{'1', '2'}}, {{'1', '3'}}, {{'1', '4'}}, {{'1', '5'}}, {{'1', '6'}}, {{'1', '7'}}, {{'1', '8'}}, {{'1', '9'}},
+                {{'2', '0'}}, {{'2', '1'}}, {{'2', '2'}}, {{'2', '3'}}, {{'2', '4'}}, {{'2', '5'}}, {{'2', '6'}}, {{'2', '7'}}, {{'2', '8'}}, {{'2', '9'}},
+                {{'3', '0'}}, {{'3', '1'}}, {{'3', '2'}}, {{'3', '3'}}, {{'3', '4'}}, {{'3', '5'}}, {{'3', '6'}}, {{'3', '7'}}, {{'3', '8'}}, {{'3', '9'}},
+                {{'4', '0'}}, {{'4', '1'}}, {{'4', '2'}}, {{'4', '3'}}, {{'4', '4'}}, {{'4', '5'}}, {{'4', '6'}}, {{'4', '7'}}, {{'4', '8'}}, {{'4', '9'}},
+                {{'5', '0'}}, {{'5', '1'}}, {{'5', '2'}}, {{'5', '3'}}, {{'5', '4'}}, {{'5', '5'}}, {{'5', '6'}}, {{'5', '7'}}, {{'5', '8'}}, {{'5', '9'}},
+                {{'6', '0'}}, {{'6', '1'}}, {{'6', '2'}}, {{'6', '3'}}, {{'6', '4'}}, {{'6', '5'}}, {{'6', '6'}}, {{'6', '7'}}, {{'6', '8'}}, {{'6', '9'}},
+                {{'7', '0'}}, {{'7', '1'}}, {{'7', '2'}}, {{'7', '3'}}, {{'7', '4'}}, {{'7', '5'}}, {{'7', '6'}}, {{'7', '7'}}, {{'7', '8'}}, {{'7', '9'}},
+                {{'8', '0'}}, {{'8', '1'}}, {{'8', '2'}}, {{'8', '3'}}, {{'8', '4'}}, {{'8', '5'}}, {{'8', '6'}}, {{'8', '7'}}, {{'8', '8'}}, {{'8', '9'}},
+                {{'9', '0'}}, {{'9', '1'}}, {{'9', '2'}}, {{'9', '3'}}, {{'9', '4'}}, {{'9', '5'}}, {{'9', '6'}}, {{'9', '7'}}, {{'9', '8'}}, {{'9', '9'}},
+            }
+        };
+
+        // special case for "0"
+        if (x == 0)
+        {
+            o->write_character('0');
+            return;
+        }
+
+        // use a pointer to fill the buffer
+        auto buffer_ptr = number_buffer.begin(); // NOLINT(llvm-qualified-auto,readability-qualified-auto,cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+
+        number_unsigned_t abs_value;
+
+        unsigned int n_chars{};
+
+        if (is_negative_number(x))
+        {
+            *buffer_ptr = '-';
+            abs_value = remove_sign(static_cast<number_integer_t>(x));
+
+            // account one more byte for the minus sign
+            n_chars = 1 + count_digits(abs_value);
+        }
+        else
+        {
+            abs_value = static_cast<number_unsigned_t>(x);
+            n_chars = count_digits(abs_value);
+        }
+
+        // spare 1 byte for '\0'
+        JSON_ASSERT(n_chars < number_buffer.size() - 1);
+
+        // jump to the end to generate the string from backward,
+        // so we later avoid reversing the result
+        buffer_ptr += n_chars;
+
+        // Fast int2ascii implementation inspired by "Fastware" talk by Andrei Alexandrescu
+        // See: https://www.youtube.com/watch?v=o4-CwDo2zpg
+        while (abs_value >= 100)
+        {
+            const auto digits_index = static_cast<unsigned>((abs_value % 100));
+            abs_value /= 100;
+            *(--buffer_ptr) = digits_to_99[digits_index][1];
+            *(--buffer_ptr) = digits_to_99[digits_index][0];
+        }
+
+        if (abs_value >= 10)
+        {
+            const auto digits_index = static_cast<unsigned>(abs_value);
+            *(--buffer_ptr) = digits_to_99[digits_index][1];
+            *(--buffer_ptr) = digits_to_99[digits_index][0];
+        }
+        else
+        {
+            *(--buffer_ptr) = static_cast<char>('0' + abs_value);
+        }
+
+        o->write_characters(number_buffer.data(), n_chars);
+    }
+
+    /*!
+    @brief dump a floating-point number
+
+    Dump a given floating-point number to output stream @a o. Works internally
+    with @a number_buffer.
+
+    @param[in] x  floating-point number to dump
+    */
+    void dump_float(number_float_t x)
+    {
+        // NaN / inf
+        if (!std::isfinite(x))
+        {
+            o->write_characters("null", 4);
+            return;
+        }
+
+        // If number_float_t is an IEEE-754 single or double precision number,
+        // use the Grisu2 algorithm to produce short numbers which are
+        // guaranteed to round-trip, using strtof and strtod, resp.
+        //
+        // NB: The test below works if <long double> == <double>.
+        static constexpr bool is_ieee_single_or_double
+            = (std::numeric_limits<number_float_t>::is_iec559 && std::numeric_limits<number_float_t>::digits == 24 && std::numeric_limits<number_float_t>::max_exponent == 128) ||
+              (std::numeric_limits<number_float_t>::is_iec559 && std::numeric_limits<number_float_t>::digits == 53 && std::numeric_limits<number_float_t>::max_exponent == 1024);
+
+        dump_float(x, std::integral_constant<bool, is_ieee_single_or_double>());
+    }
+
+    void dump_float(number_float_t x, std::true_type /*is_ieee_single_or_double*/)
+    {
+        auto* begin = number_buffer.data();
+        auto* end = ::nlohmann::detail::to_chars(begin, begin + number_buffer.size(), x);
+
+        o->write_characters(begin, static_cast<size_t>(end - begin));
+    }
+
+    void dump_float(number_float_t x, std::false_type /*is_ieee_single_or_double*/)
+    {
+        // get number of digits for a float -> text -> float round-trip
+        static constexpr auto d = std::numeric_limits<number_float_t>::max_digits10;
+
+        // the actual conversion
+        // NOLINTNEXTLINE(cppcoreguidelines-pro-type-vararg,hicpp-vararg)
+        std::ptrdiff_t len = (std::snprintf)(number_buffer.data(), number_buffer.size(), "%.*g", d, x);
+
+        // negative value indicates an error
+        JSON_ASSERT(len > 0);
+        // check if buffer was large enough
+        JSON_ASSERT(static_cast<std::size_t>(len) < number_buffer.size());
+
+        // erase thousands separator
+        if (thousands_sep != '\0')
+        {
+            // NOLINTNEXTLINE(readability-qualified-auto,llvm-qualified-auto): std::remove returns an iterator, see https://github.com/nlohmann/json/issues/3081
+            const auto end = std::remove(number_buffer.begin(), number_buffer.begin() + len, thousands_sep);
+            std::fill(end, number_buffer.end(), '\0');
+            JSON_ASSERT((end - number_buffer.begin()) <= len);
+            len = (end - number_buffer.begin());
+        }
+
+        // convert decimal point to '.'
+        if (decimal_point != '\0' && decimal_point != '.')
+        {
+            // NOLINTNEXTLINE(readability-qualified-auto,llvm-qualified-auto): std::find returns an iterator, see https://github.com/nlohmann/json/issues/3081
+            const auto dec_pos = std::find(number_buffer.begin(), number_buffer.end(), decimal_point);
+            if (dec_pos != number_buffer.end())
+            {
+                *dec_pos = '.';
+            }
+        }
+
+        o->write_characters(number_buffer.data(), static_cast<std::size_t>(len));
+
+        // determine if we need to append ".0"
+        const bool value_is_int_like =
+            std::none_of(number_buffer.begin(), number_buffer.begin() + len + 1,
+                         [](char c)
+        {
+            return c == '.' || c == 'e';
+        });
+
+        if (value_is_int_like)
+        {
+            o->write_characters(".0", 2);
+        }
+    }
+
+    /*!
+    @brief check whether a string is UTF-8 encoded
+
+    The function checks each byte of a string whether it is UTF-8 encoded. The
+    result of the check is stored in the @a state parameter. The function must
+    be called initially with state 0 (accept). State 1 means the string must
+    be rejected, because the current byte is not allowed. If the string is
+    completely processed, but the state is non-zero, the string ended
+    prematurely; that is, the last byte indicated more bytes should have
+    followed.
+
+    @param[in,out] state  the state of the decoding
+    @param[in,out] codep  codepoint (valid only if resulting state is UTF8_ACCEPT)
+    @param[in] byte       next byte to decode
+    @return               new state
+
+    @note The function has been edited: a std::array is used.
+
+    @copyright Copyright (c) 2008-2009 Bjoern Hoehrmann <bjoern@hoehrmann.de>
+    @sa http://bjoern.hoehrmann.de/utf-8/decoder/dfa/
+    */
+    static std::uint8_t decode(std::uint8_t& state, std::uint32_t& codep, const std::uint8_t byte) noexcept
+    {
+        static const std::array<std::uint8_t, 400> utf8d =
+        {
+            {
+                0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 00..1F
+                0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 20..3F
+                0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 40..5F
+                0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 60..7F
+                1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, 9, // 80..9F
+                7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, // A0..BF
+                8, 8, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, // C0..DF
+                0xA, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x3, 0x4, 0x3, 0x3, // E0..EF
+                0xB, 0x6, 0x6, 0x6, 0x5, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, 0x8, // F0..FF
+                0x0, 0x1, 0x2, 0x3, 0x5, 0x8, 0x7, 0x1, 0x1, 0x1, 0x4, 0x6, 0x1, 0x1, 0x1, 0x1, // s0..s0
+                1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, // s1..s2
+                1, 2, 1, 1, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, // s3..s4
+                1, 2, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 1, 3, 1, 1, 1, 1, 1, 1, // s5..s6
+                1, 3, 1, 1, 1, 1, 1, 3, 1, 3, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 // s7..s8
+            }
+        };
+
+        JSON_ASSERT(byte < utf8d.size());
+        const std::uint8_t type = utf8d[byte];
+
+        codep = (state != UTF8_ACCEPT)
+                ? (byte & 0x3fu) | (codep << 6u)
+                : (0xFFu >> type) & (byte);
+
+        const std::size_t index = 256u + (static_cast<size_t>(state) * 16u) + static_cast<size_t>(type);
+        JSON_ASSERT(index < utf8d.size());
+        state = utf8d[index];
+        return state;
+    }
+
+    /*
+     * Overload to make the compiler happy while it is instantiating
+     * dump_integer for number_unsigned_t.
+     * Must never be called.
+     */
+    number_unsigned_t remove_sign(number_unsigned_t x)
+    {
+        JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+        return x; // LCOV_EXCL_LINE
+    }
+
+    /*
+     * Helper function for dump_integer
+     *
+     * This function takes a negative signed integer and returns its absolute
+     * value as unsigned integer. The plus/minus shuffling is necessary as we can
+     * not directly remove the sign of an arbitrary signed integer as the
+     * absolute values of INT_MIN and INT_MAX are usually not the same. See
+     * #1708 for details.
+     */
+    number_unsigned_t remove_sign(number_integer_t x) noexcept
+    {
+        JSON_ASSERT(x < 0 && x < (std::numeric_limits<number_integer_t>::max)()); // NOLINT(misc-redundant-expression)
+        return static_cast<number_unsigned_t>(-(x + 1)) + 1;
+    }
+
+  private:
+    /// the output of the serializer
+    output_adapter_t<char> o = nullptr;
+
+    /// a (hopefully) large enough character buffer
+    std::array<char, 64> number_buffer{{}};
+
+    /// the locale
+    const std::lconv* loc = nullptr;
+    /// the locale's thousand separator character
+    const char thousands_sep = '\0';
+    /// the locale's decimal point character
+    const char decimal_point = '\0';
+
+    /// string buffer
+    std::array<char, 512> string_buffer{{}};
+
+    /// the indentation character
+    const char indent_char;
+    /// the indentation string
+    string_t indent_string;
+
+    /// error_handler how to react on decoding errors
+    const error_handler_t error_handler;
+};
+
+}  // namespace detail
+NLOHMANN_JSON_NAMESPACE_END
+
+// #include <nlohmann/detail/value_t.hpp>
+
+// #include <nlohmann/json_fwd.hpp>
+
+// #include <nlohmann/ordered_map.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#include <functional> // equal_to, less
+#include <initializer_list> // initializer_list
+#include <iterator> // input_iterator_tag, iterator_traits
+#include <memory> // allocator
+#include <stdexcept> // for out_of_range
+#include <type_traits> // enable_if, is_convertible
+#include <utility> // pair
+#include <vector> // vector
+
+// #include <nlohmann/detail/macro_scope.hpp>
+
+// #include <nlohmann/detail/meta/type_traits.hpp>
+
+
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+/// ordered_map: a minimal map-like container that preserves insertion order
+/// for use within nlohmann::basic_json<ordered_map>
+template <class Key, class T, class IgnoredLess = std::less<Key>,
+          class Allocator = std::allocator<std::pair<const Key, T>>>
+              struct ordered_map : std::vector<std::pair<const Key, T>, Allocator>
+{
+    using key_type = Key;
+    using mapped_type = T;
+    using Container = std::vector<std::pair<const Key, T>, Allocator>;
+    using iterator = typename Container::iterator;
+    using const_iterator = typename Container::const_iterator;
+    using size_type = typename Container::size_type;
+    using value_type = typename Container::value_type;
+#ifdef JSON_HAS_CPP_14
+    using key_compare = std::equal_to<>;
+#else
+    using key_compare = std::equal_to<Key>;
+#endif
+
+    // Explicit constructors instead of `using Container::Container`
+    // otherwise older compilers choke on it (GCC <= 5.5, xcode <= 9.4)
+    ordered_map() noexcept(noexcept(Container())) : Container{} {}
+    explicit ordered_map(const Allocator& alloc) noexcept(noexcept(Container(alloc))) : Container{alloc} {}
+    template <class It>
+    ordered_map(It first, It last, const Allocator& alloc = Allocator())
+        : Container{first, last, alloc} {}
+    ordered_map(std::initializer_list<value_type> init, const Allocator& alloc = Allocator() )
+        : Container{init, alloc} {}
+
+    std::pair<iterator, bool> emplace(const key_type& key, T&& t)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return {it, false};
+            }
+        }
+        Container::emplace_back(key, std::forward<T>(t));
+        return {std::prev(this->end()), true};
+    }
+
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_key_type<key_compare, key_type, KeyType>::value, int> = 0>
+    std::pair<iterator, bool> emplace(KeyType && key, T && t)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return {it, false};
+            }
+        }
+        Container::emplace_back(std::forward<KeyType>(key), std::forward<T>(t));
+        return {std::prev(this->end()), true};
+    }
+
+    T& operator[](const key_type& key)
+    {
+        return emplace(key, T{}).first->second;
+    }
+
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_key_type<key_compare, key_type, KeyType>::value, int> = 0>
+    T & operator[](KeyType && key)
+    {
+        return emplace(std::forward<KeyType>(key), T{}).first->second;
+    }
+
+    const T& operator[](const key_type& key) const
+    {
+        return at(key);
+    }
+
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_key_type<key_compare, key_type, KeyType>::value, int> = 0>
+    const T & operator[](KeyType && key) const
+    {
+        return at(std::forward<KeyType>(key));
+    }
+
+    T& at(const key_type& key)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return it->second;
+            }
+        }
+
+        JSON_THROW(std::out_of_range("key not found"));
+    }
+
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_key_type<key_compare, key_type, KeyType>::value, int> = 0>
+    T & at(KeyType && key) // NOLINT(cppcoreguidelines-missing-std-forward)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return it->second;
+            }
+        }
+
+        JSON_THROW(std::out_of_range("key not found"));
+    }
+
+    const T& at(const key_type& key) const
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return it->second;
+            }
+        }
+
+        JSON_THROW(std::out_of_range("key not found"));
+    }
+
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_key_type<key_compare, key_type, KeyType>::value, int> = 0>
+    const T & at(KeyType && key) const // NOLINT(cppcoreguidelines-missing-std-forward)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return it->second;
+            }
+        }
+
+        JSON_THROW(std::out_of_range("key not found"));
+    }
+
+    size_type erase(const key_type& key)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                // Since we cannot move const Keys, re-construct them in place
+                for (auto next = it; ++next != this->end(); ++it)
+                {
+                    it->~value_type(); // Destroy but keep allocation
+                    new (&*it) value_type{std::move(*next)};
+                }
+                Container::pop_back();
+                return 1;
+            }
+        }
+        return 0;
+    }
+
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_key_type<key_compare, key_type, KeyType>::value, int> = 0>
+    size_type erase(KeyType && key) // NOLINT(cppcoreguidelines-missing-std-forward)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                // Since we cannot move const Keys, re-construct them in place
+                for (auto next = it; ++next != this->end(); ++it)
+                {
+                    it->~value_type(); // Destroy but keep allocation
+                    new (&*it) value_type{std::move(*next)};
+                }
+                Container::pop_back();
+                return 1;
+            }
+        }
+        return 0;
+    }
+
+    iterator erase(iterator pos)
+    {
+        return erase(pos, std::next(pos));
+    }
+
+    iterator erase(iterator first, iterator last)
+    {
+        if (first == last)
+        {
+            return first;
+        }
+
+        const auto elements_affected = std::distance(first, last);
+        const auto offset = std::distance(Container::begin(), first);
+
+        // This is the start situation. We need to delete elements_affected
+        // elements (3 in this example: e, f, g), and need to return an
+        // iterator past the last deleted element (h in this example).
+        // Note that offset is the distance from the start of the vector
+        // to first. We will need this later.
+
+        // [ a, b, c, d, e, f, g, h, i, j ]
+        //               ^        ^
+        //             first    last
+
+        // Since we cannot move const Keys, we re-construct them in place.
+        // We start at first and re-construct (viz. copy) the elements from
+        // the back of the vector. Example for first iteration:
+
+        //               ,--------.
+        //               v        |   destroy e and re-construct with h
+        // [ a, b, c, d, e, f, g, h, i, j ]
+        //               ^        ^
+        //               it       it + elements_affected
+
+        for (auto it = first; std::next(it, elements_affected) != Container::end(); ++it)
+        {
+            it->~value_type(); // destroy but keep allocation
+            new (&*it) value_type{std::move(*std::next(it, elements_affected))}; // "move" next element to it
+        }
+
+        // [ a, b, c, d, h, i, j, h, i, j ]
+        //               ^        ^
+        //             first    last
+
+        // remove the unneeded elements at the end of the vector
+        Container::resize(this->size() - static_cast<size_type>(elements_affected));
+
+        // [ a, b, c, d, h, i, j ]
+        //               ^        ^
+        //             first    last
+
+        // first is now pointing past the last deleted element, but we cannot
+        // use this iterator, because it may have been invalidated by the
+        // resize call. Instead, we can return begin() + offset.
+        return Container::begin() + offset;
+    }
+
+    size_type count(const key_type& key) const
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return 1;
+            }
+        }
+        return 0;
+    }
+
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_key_type<key_compare, key_type, KeyType>::value, int> = 0>
+    size_type count(KeyType && key) const // NOLINT(cppcoreguidelines-missing-std-forward)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return 1;
+            }
+        }
+        return 0;
+    }
+
+    iterator find(const key_type& key)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return it;
+            }
+        }
+        return Container::end();
+    }
+
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_key_type<key_compare, key_type, KeyType>::value, int> = 0>
+    iterator find(KeyType && key) // NOLINT(cppcoreguidelines-missing-std-forward)
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return it;
+            }
+        }
+        return Container::end();
+    }
+
+    const_iterator find(const key_type& key) const
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, key))
+            {
+                return it;
+            }
+        }
+        return Container::end();
+    }
+
+    std::pair<iterator, bool> insert( value_type&& value )
+    {
+        return emplace(value.first, std::move(value.second));
+    }
+
+    std::pair<iterator, bool> insert( const value_type& value )
+    {
+        for (auto it = this->begin(); it != this->end(); ++it)
+        {
+            if (m_compare(it->first, value.first))
+            {
+                return {it, false};
+            }
+        }
+        Container::push_back(value);
+        return {--this->end(), true};
+    }
+
+    template<typename InputIt>
+    using require_input_iter = typename std::enable_if<std::is_convertible<typename std::iterator_traits<InputIt>::iterator_category,
+        std::input_iterator_tag>::value>::type;
+
+    template<typename InputIt, typename = require_input_iter<InputIt>>
+    void insert(InputIt first, InputIt last)
+    {
+        for (auto it = first; it != last; ++it)
+        {
+            insert(*it);
+        }
+    }
+
+private:
+    JSON_NO_UNIQUE_ADDRESS key_compare m_compare = key_compare();
+};
+
+NLOHMANN_JSON_NAMESPACE_END
+
+
+#if defined(JSON_HAS_CPP_17)
+    #if JSON_HAS_STATIC_RTTI
+        #include <any>
+    #endif
+    #include <string_view>
+#endif
+
+/*!
+@brief namespace for Niels Lohmann
+@see https://github.com/nlohmann
+@since version 1.0.0
+*/
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+/*!
+@brief a class to store JSON values
+
+@internal
+@invariant The member variables @a m_value and @a m_type have the following
+relationship:
+- If `m_type == value_t::object`, then `m_value.object != nullptr`.
+- If `m_type == value_t::array`, then `m_value.array != nullptr`.
+- If `m_type == value_t::string`, then `m_value.string != nullptr`.
+The invariants are checked by member function assert_invariant().
+
+@note ObjectType trick from https://stackoverflow.com/a/9860911
+@endinternal
+
+@since version 1.0.0
+
+@nosubgrouping
+*/
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-special-member-functions)
+    : public ::nlohmann::detail::json_base_class<CustomBaseClass>
+{
+  private:
+    template<detail::value_t> friend struct detail::external_constructor;
+
+    template<typename>
+    friend class ::nlohmann::json_pointer;
+    // can be restored when json_pointer backwards compatibility is removed
+    // friend ::nlohmann::json_pointer<StringType>;
+
+    template<typename BasicJsonType, typename InputType>
+    friend class ::nlohmann::detail::parser;
+    friend ::nlohmann::detail::serializer<basic_json>;
+    template<typename BasicJsonType>
+    friend class ::nlohmann::detail::iter_impl;
+    template<typename BasicJsonType, typename CharType>
+    friend class ::nlohmann::detail::binary_writer;
+    template<typename BasicJsonType, typename InputType, typename SAX>
+    friend class ::nlohmann::detail::binary_reader;
+    template<typename BasicJsonType, typename InputAdapterType>
+    friend class ::nlohmann::detail::json_sax_dom_parser;
+    template<typename BasicJsonType, typename InputAdapterType>
+    friend class ::nlohmann::detail::json_sax_dom_callback_parser;
+    friend class ::nlohmann::detail::exception;
+
+    /// workaround type for MSVC
+    using basic_json_t = NLOHMANN_BASIC_JSON_TPL;
+    using json_base_class_t = ::nlohmann::detail::json_base_class<CustomBaseClass>;
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    // convenience aliases for types residing in namespace detail;
+    using lexer = ::nlohmann::detail::lexer_base<basic_json>;
+
+    template<typename InputAdapterType>
+    static ::nlohmann::detail::parser<basic_json, InputAdapterType> parser(
+        InputAdapterType adapter,
+        detail::parser_callback_t<basic_json>cb = nullptr,
+        const bool allow_exceptions = true,
+        const bool ignore_comments = false
+                                 )
+    {
+        return ::nlohmann::detail::parser<basic_json, InputAdapterType>(std::move(adapter),
+            std::move(cb), allow_exceptions, ignore_comments);
+    }
+
+  private:
+    using primitive_iterator_t = ::nlohmann::detail::primitive_iterator_t;
+    template<typename BasicJsonType>
+    using internal_iterator = ::nlohmann::detail::internal_iterator<BasicJsonType>;
+    template<typename BasicJsonType>
+    using iter_impl = ::nlohmann::detail::iter_impl<BasicJsonType>;
+    template<typename Iterator>
+    using iteration_proxy = ::nlohmann::detail::iteration_proxy<Iterator>;
+    template<typename Base> using json_reverse_iterator = ::nlohmann::detail::json_reverse_iterator<Base>;
+
+    template<typename CharType>
+    using output_adapter_t = ::nlohmann::detail::output_adapter_t<CharType>;
+
+    template<typename InputType>
+    using binary_reader = ::nlohmann::detail::binary_reader<basic_json, InputType>;
+    template<typename CharType> using binary_writer = ::nlohmann::detail::binary_writer<basic_json, CharType>;
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    using serializer = ::nlohmann::detail::serializer<basic_json>;
+
+  public:
+    using value_t = detail::value_t;
+    /// JSON Pointer, see @ref nlohmann::json_pointer
+    using json_pointer = ::nlohmann::json_pointer<StringType>;
+    template<typename T, typename SFINAE>
+    using json_serializer = JSONSerializer<T, SFINAE>;
+    /// how to treat decoding errors
+    using error_handler_t = detail::error_handler_t;
+    /// how to treat CBOR tags
+    using cbor_tag_handler_t = detail::cbor_tag_handler_t;
+    /// how to encode BJData
+    using bjdata_version_t = detail::bjdata_version_t;
+    /// helper type for initializer lists of basic_json values
+    using initializer_list_t = std::initializer_list<detail::json_ref<basic_json>>;
+
+    using input_format_t = detail::input_format_t;
+    /// SAX interface type, see @ref nlohmann::json_sax
+    using json_sax_t = json_sax<basic_json>;
+
+    ////////////////
+    // exceptions //
+    ////////////////
+
+    /// @name exceptions
+    /// Classes to implement user-defined exceptions.
+    /// @{
+
+    using exception = detail::exception;
+    using parse_error = detail::parse_error;
+    using invalid_iterator = detail::invalid_iterator;
+    using type_error = detail::type_error;
+    using out_of_range = detail::out_of_range;
+    using other_error = detail::other_error;
+
+    /// @}
+
+    /////////////////////
+    // container types //
+    /////////////////////
+
+    /// @name container types
+    /// The canonic container types to use @ref basic_json like any other STL
+    /// container.
+    /// @{
+
+    /// the type of elements in a basic_json container
+    using value_type = basic_json;
+
+    /// the type of an element reference
+    using reference = value_type&;
+    /// the type of an element const reference
+    using const_reference = const value_type&;
+
+    /// a type to represent differences between iterators
+    using difference_type = std::ptrdiff_t;
+    /// a type to represent container sizes
+    using size_type = std::size_t;
+
+    /// the allocator type
+    using allocator_type = AllocatorType<basic_json>;
+
+    /// the type of an element pointer
+    using pointer = typename std::allocator_traits<allocator_type>::pointer;
+    /// the type of an element const pointer
+    using const_pointer = typename std::allocator_traits<allocator_type>::const_pointer;
+
+    /// an iterator for a basic_json container
+    using iterator = iter_impl<basic_json>;
+    /// a const iterator for a basic_json container
+    using const_iterator = iter_impl<const basic_json>;
+    /// a reverse iterator for a basic_json container
+    using reverse_iterator = json_reverse_iterator<typename basic_json::iterator>;
+    /// a const reverse iterator for a basic_json container
+    using const_reverse_iterator = json_reverse_iterator<typename basic_json::const_iterator>;
+
+    /// @}
+
+    /// @brief returns the allocator associated with the container
+    /// @sa https://json.nlohmann.me/api/basic_json/get_allocator/
+    static allocator_type get_allocator()
+    {
+        return allocator_type();
+    }
+
+    /// @brief returns version information on the library
+    /// @sa https://json.nlohmann.me/api/basic_json/meta/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json meta()
+    {
+        basic_json result;
+
+        result["copyright"] = "(C) 2013-2023 Niels Lohmann";
+        result["name"] = "JSON for Modern C++";
+        result["url"] = "https://github.com/nlohmann/json";
+        result["version"]["string"] =
+            detail::concat(std::to_string(NLOHMANN_JSON_VERSION_MAJOR), '.',
+                           std::to_string(NLOHMANN_JSON_VERSION_MINOR), '.',
+                           std::to_string(NLOHMANN_JSON_VERSION_PATCH));
+        result["version"]["major"] = NLOHMANN_JSON_VERSION_MAJOR;
+        result["version"]["minor"] = NLOHMANN_JSON_VERSION_MINOR;
+        result["version"]["patch"] = NLOHMANN_JSON_VERSION_PATCH;
+
+#ifdef _WIN32
+        result["platform"] = "win32";
+#elif defined __linux__
+        result["platform"] = "linux";
+#elif defined __APPLE__
+        result["platform"] = "apple";
+#elif defined __unix__
+        result["platform"] = "unix";
+#else
+        result["platform"] = "unknown";
+#endif
+
+#if defined(__ICC) || defined(__INTEL_COMPILER)
+        result["compiler"] = {{"family", "icc"}, {"version", __INTEL_COMPILER}};
+#elif defined(__clang__)
+        result["compiler"] = {{"family", "clang"}, {"version", __clang_version__}};
+#elif defined(__GNUC__) || defined(__GNUG__)
+        result["compiler"] = {{"family", "gcc"}, {"version", detail::concat(
+                    std::to_string(__GNUC__), '.',
+                    std::to_string(__GNUC_MINOR__), '.',
+                    std::to_string(__GNUC_PATCHLEVEL__))
+            }
+        };
+#elif defined(__HP_cc) || defined(__HP_aCC)
+        result["compiler"] = "hp"
+#elif defined(__IBMCPP__)
+        result["compiler"] = {{"family", "ilecpp"}, {"version", __IBMCPP__}};
+#elif defined(_MSC_VER)
+        result["compiler"] = {{"family", "msvc"}, {"version", _MSC_VER}};
+#elif defined(__PGI)
+        result["compiler"] = {{"family", "pgcpp"}, {"version", __PGI}};
+#elif defined(__SUNPRO_CC)
+        result["compiler"] = {{"family", "sunpro"}, {"version", __SUNPRO_CC}};
+#else
+        result["compiler"] = {{"family", "unknown"}, {"version", "unknown"}};
+#endif
+
+#if defined(_MSVC_LANG)
+        result["compiler"]["c++"] = std::to_string(_MSVC_LANG);
+#elif defined(__cplusplus)
+        result["compiler"]["c++"] = std::to_string(__cplusplus);
+#else
+        result["compiler"]["c++"] = "unknown";
+#endif
+        return result;
+    }
+
+    ///////////////////////////
+    // JSON value data types //
+    ///////////////////////////
+
+    /// @name JSON value data types
+    /// The data types to store a JSON value. These types are derived from
+    /// the template arguments passed to class @ref basic_json.
+    /// @{
+
+    /// @brief default object key comparator type
+    /// The actual object key comparator type (@ref object_comparator_t) may be
+    /// different.
+    /// @sa https://json.nlohmann.me/api/basic_json/default_object_comparator_t/
+#if defined(JSON_HAS_CPP_14)
+    // use of transparent comparator avoids unnecessary repeated construction of temporaries
+    // in functions involving lookup by key with types other than object_t::key_type (aka. StringType)
+    using default_object_comparator_t = std::less<>;
+#else
+    using default_object_comparator_t = std::less<StringType>;
+#endif
+
+    /// @brief a type for an object
+    /// @sa https://json.nlohmann.me/api/basic_json/object_t/
+    using object_t = ObjectType<StringType,
+          basic_json,
+          default_object_comparator_t,
+          AllocatorType<std::pair<const StringType,
+          basic_json>>>;
+
+    /// @brief a type for an array
+    /// @sa https://json.nlohmann.me/api/basic_json/array_t/
+    using array_t = ArrayType<basic_json, AllocatorType<basic_json>>;
+
+    /// @brief a type for a string
+    /// @sa https://json.nlohmann.me/api/basic_json/string_t/
+    using string_t = StringType;
+
+    /// @brief a type for a boolean
+    /// @sa https://json.nlohmann.me/api/basic_json/boolean_t/
+    using boolean_t = BooleanType;
+
+    /// @brief a type for a number (integer)
+    /// @sa https://json.nlohmann.me/api/basic_json/number_integer_t/
+    using number_integer_t = NumberIntegerType;
+
+    /// @brief a type for a number (unsigned)
+    /// @sa https://json.nlohmann.me/api/basic_json/number_unsigned_t/
+    using number_unsigned_t = NumberUnsignedType;
+
+    /// @brief a type for a number (floating-point)
+    /// @sa https://json.nlohmann.me/api/basic_json/number_float_t/
+    using number_float_t = NumberFloatType;
+
+    /// @brief a type for a packed binary type
+    /// @sa https://json.nlohmann.me/api/basic_json/binary_t/
+    using binary_t = nlohmann::byte_container_with_subtype<BinaryType>;
+
+    /// @brief object key comparator type
+    /// @sa https://json.nlohmann.me/api/basic_json/object_comparator_t/
+    using object_comparator_t = detail::actual_object_comparator_t<basic_json>;
+
+    /// @}
+
+  private:
+
+    /// helper for exception-safe object creation
+    template<typename T, typename... Args>
+    JSON_HEDLEY_RETURNS_NON_NULL
+    static T* create(Args&& ... args)
+    {
+        AllocatorType<T> alloc;
+        using AllocatorTraits = std::allocator_traits<AllocatorType<T>>;
+
+        auto deleter = [&](T * obj)
+        {
+            AllocatorTraits::deallocate(alloc, obj, 1);
+        };
+        std::unique_ptr<T, decltype(deleter)> obj(AllocatorTraits::allocate(alloc, 1), deleter);
+        AllocatorTraits::construct(alloc, obj.get(), std::forward<Args>(args)...);
+        JSON_ASSERT(obj != nullptr);
+        return obj.release();
+    }
+
+    ////////////////////////
+    // JSON value storage //
+    ////////////////////////
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    /*!
+    @brief a JSON value
+
+    The actual storage for a JSON value of the @ref basic_json class. This
+    union combines the different storage types for the JSON value types
+    defined in @ref value_t.
+
+    JSON type | value_t type    | used type
+    --------- | --------------- | ------------------------
+    object    | object          | pointer to @ref object_t
+    array     | array           | pointer to @ref array_t
+    string    | string          | pointer to @ref string_t
+    boolean   | boolean         | @ref boolean_t
+    number    | number_integer  | @ref number_integer_t
+    number    | number_unsigned | @ref number_unsigned_t
+    number    | number_float    | @ref number_float_t
+    binary    | binary          | pointer to @ref binary_t
+    null      | null            | *no value is stored*
+
+    @note Variable-length types (objects, arrays, and strings) are stored as
+    pointers. The size of the union should not exceed 64 bits if the default
+    value types are used.
+
+    @since version 1.0.0
+    */
+    union json_value
+    {
+        /// object (stored with pointer to save storage)
+        object_t* object;
+        /// array (stored with pointer to save storage)
+        array_t* array;
+        /// string (stored with pointer to save storage)
+        string_t* string;
+        /// binary (stored with pointer to save storage)
+        binary_t* binary;
+        /// boolean
+        boolean_t boolean;
+        /// number (integer)
+        number_integer_t number_integer;
+        /// number (unsigned integer)
+        number_unsigned_t number_unsigned;
+        /// number (floating-point)
+        number_float_t number_float;
+
+        /// default constructor (for null values)
+        json_value() = default;
+        /// constructor for booleans
+        json_value(boolean_t v) noexcept : boolean(v) {}
+        /// constructor for numbers (integer)
+        json_value(number_integer_t v) noexcept : number_integer(v) {}
+        /// constructor for numbers (unsigned)
+        json_value(number_unsigned_t v) noexcept : number_unsigned(v) {}
+        /// constructor for numbers (floating-point)
+        json_value(number_float_t v) noexcept : number_float(v) {}
+        /// constructor for empty values of a given type
+        json_value(value_t t)
+        {
+            switch (t)
+            {
+                case value_t::object:
+                {
+                    object = create<object_t>();
+                    break;
+                }
+
+                case value_t::array:
+                {
+                    array = create<array_t>();
+                    break;
+                }
+
+                case value_t::string:
+                {
+                    string = create<string_t>("");
+                    break;
+                }
+
+                case value_t::binary:
+                {
+                    binary = create<binary_t>();
+                    break;
+                }
+
+                case value_t::boolean:
+                {
+                    boolean = static_cast<boolean_t>(false);
+                    break;
+                }
+
+                case value_t::number_integer:
+                {
+                    number_integer = static_cast<number_integer_t>(0);
+                    break;
+                }
+
+                case value_t::number_unsigned:
+                {
+                    number_unsigned = static_cast<number_unsigned_t>(0);
+                    break;
+                }
+
+                case value_t::number_float:
+                {
+                    number_float = static_cast<number_float_t>(0.0);
+                    break;
+                }
+
+                case value_t::null:
+                {
+                    object = nullptr;  // silence warning, see #821
+                    break;
+                }
+
+                case value_t::discarded:
+                default:
+                {
+                    object = nullptr;  // silence warning, see #821
+                    if (JSON_HEDLEY_UNLIKELY(t == value_t::null))
+                    {
+                        JSON_THROW(other_error::create(500, "961c151d2e87f2686a955a9be24d316f1362bf21 3.11.3", nullptr)); // LCOV_EXCL_LINE
+                    }
+                    break;
+                }
+            }
+        }
+
+        /// constructor for strings
+        json_value(const string_t& value) : string(create<string_t>(value)) {}
+
+        /// constructor for rvalue strings
+        json_value(string_t&& value) : string(create<string_t>(std::move(value))) {}
+
+        /// constructor for objects
+        json_value(const object_t& value) : object(create<object_t>(value)) {}
+
+        /// constructor for rvalue objects
+        json_value(object_t&& value) : object(create<object_t>(std::move(value))) {}
+
+        /// constructor for arrays
+        json_value(const array_t& value) : array(create<array_t>(value)) {}
+
+        /// constructor for rvalue arrays
+        json_value(array_t&& value) : array(create<array_t>(std::move(value))) {}
+
+        /// constructor for binary arrays
+        json_value(const typename binary_t::container_type& value) : binary(create<binary_t>(value)) {}
+
+        /// constructor for rvalue binary arrays
+        json_value(typename binary_t::container_type&& value) : binary(create<binary_t>(std::move(value))) {}
+
+        /// constructor for binary arrays (internal type)
+        json_value(const binary_t& value) : binary(create<binary_t>(value)) {}
+
+        /// constructor for rvalue binary arrays (internal type)
+        json_value(binary_t&& value) : binary(create<binary_t>(std::move(value))) {}
+
+        void destroy(value_t t)
+        {
+            if (
+                (t == value_t::object && object == nullptr) ||
+                (t == value_t::array && array == nullptr) ||
+                (t == value_t::string && string == nullptr) ||
+                (t == value_t::binary && binary == nullptr)
+            )
+            {
+                //not initialized (e.g. due to exception in the ctor)
+                return;
+            }
+            if (t == value_t::array || t == value_t::object)
+            {
+                // flatten the current json_value to a heap-allocated stack
+                std::vector<basic_json> stack;
+
+                // move the top-level items to stack
+                if (t == value_t::array)
+                {
+                    stack.reserve(array->size());
+                    std::move(array->begin(), array->end(), std::back_inserter(stack));
+                }
+                else
+                {
+                    stack.reserve(object->size());
+                    for (auto&& it : *object)
+                    {
+                        stack.push_back(std::move(it.second));
+                    }
+                }
+
+                while (!stack.empty())
+                {
+                    // move the last item to local variable to be processed
+                    basic_json current_item(std::move(stack.back()));
+                    stack.pop_back();
+
+                    // if current_item is array/object, move
+                    // its children to the stack to be processed later
+                    if (current_item.is_array())
+                    {
+                        std::move(current_item.m_data.m_value.array->begin(), current_item.m_data.m_value.array->end(), std::back_inserter(stack));
+
+                        current_item.m_data.m_value.array->clear();
+                    }
+                    else if (current_item.is_object())
+                    {
+                        for (auto&& it : *current_item.m_data.m_value.object)
+                        {
+                            stack.push_back(std::move(it.second));
+                        }
+
+                        current_item.m_data.m_value.object->clear();
+                    }
+
+                    // it's now safe that current_item get destructed
+                    // since it doesn't have any children
+                }
+            }
+
+            switch (t)
+            {
+                case value_t::object:
+                {
+                    AllocatorType<object_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, object);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, object, 1);
+                    break;
+                }
+
+                case value_t::array:
+                {
+                    AllocatorType<array_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, array);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, array, 1);
+                    break;
+                }
+
+                case value_t::string:
+                {
+                    AllocatorType<string_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, string);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, string, 1);
+                    break;
+                }
+
+                case value_t::binary:
+                {
+                    AllocatorType<binary_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, binary);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, binary, 1);
+                    break;
+                }
+
+                case value_t::null:
+                case value_t::boolean:
+                case value_t::number_integer:
+                case value_t::number_unsigned:
+                case value_t::number_float:
+                case value_t::discarded:
+                default:
+                {
+                    break;
+                }
+            }
+        }
+    };
+
+  private:
+    /*!
+    @brief checks the class invariants
+
+    This function asserts the class invariants. It needs to be called at the
+    end of every constructor to make sure that created objects respect the
+    invariant. Furthermore, it has to be called each time the type of a JSON
+    value is changed, because the invariant expresses a relationship between
+    @a m_type and @a m_value.
+
+    Furthermore, the parent relation is checked for arrays and objects: If
+    @a check_parents true and the value is an array or object, then the
+    container's elements must have the current value as parent.
+
+    @param[in] check_parents  whether the parent relation should be checked.
+               The value is true by default and should only be set to false
+               during destruction of objects when the invariant does not
+               need to hold.
+    */
+    void assert_invariant(bool check_parents = true) const noexcept
+    {
+        JSON_ASSERT(m_data.m_type != value_t::object || m_data.m_value.object != nullptr);
+        JSON_ASSERT(m_data.m_type != value_t::array || m_data.m_value.array != nullptr);
+        JSON_ASSERT(m_data.m_type != value_t::string || m_data.m_value.string != nullptr);
+        JSON_ASSERT(m_data.m_type != value_t::binary || m_data.m_value.binary != nullptr);
+
+#if JSON_DIAGNOSTICS
+        JSON_TRY
+        {
+            // cppcheck-suppress assertWithSideEffect
+            JSON_ASSERT(!check_parents || !is_structured() || std::all_of(begin(), end(), [this](const basic_json & j)
+            {
+                return j.m_parent == this;
+            }));
+        }
+        JSON_CATCH(...) {} // LCOV_EXCL_LINE
+#endif
+        static_cast<void>(check_parents);
+    }
+
+    void set_parents()
+    {
+#if JSON_DIAGNOSTICS
+        switch (m_data.m_type)
+        {
+            case value_t::array:
+            {
+                for (auto& element : *m_data.m_value.array)
+                {
+                    element.m_parent = this;
+                }
+                break;
+            }
+
+            case value_t::object:
+            {
+                for (auto& element : *m_data.m_value.object)
+                {
+                    element.second.m_parent = this;
+                }
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+                break;
+        }
+#endif
+    }
+
+    iterator set_parents(iterator it, typename iterator::difference_type count_set_parents)
+    {
+#if JSON_DIAGNOSTICS
+        for (typename iterator::difference_type i = 0; i < count_set_parents; ++i)
+        {
+            (it + i)->m_parent = this;
+        }
+#else
+        static_cast<void>(count_set_parents);
+#endif
+        return it;
+    }
+
+    reference set_parent(reference j, std::size_t old_capacity = detail::unknown_size())
+    {
+#if JSON_DIAGNOSTICS
+        if (old_capacity != detail::unknown_size())
+        {
+            // see https://github.com/nlohmann/json/issues/2838
+            JSON_ASSERT(type() == value_t::array);
+            if (JSON_HEDLEY_UNLIKELY(m_data.m_value.array->capacity() != old_capacity))
+            {
+                // capacity has changed: update all parents
+                set_parents();
+                return j;
+            }
+        }
+
+        // ordered_json uses a vector internally, so pointers could have
+        // been invalidated; see https://github.com/nlohmann/json/issues/2962
+#ifdef JSON_HEDLEY_MSVC_VERSION
+#pragma warning(push )
+#pragma warning(disable : 4127) // ignore warning to replace if with if constexpr
+#endif
+        if (detail::is_ordered_map<object_t>::value)
+        {
+            set_parents();
+            return j;
+        }
+#ifdef JSON_HEDLEY_MSVC_VERSION
+#pragma warning( pop )
+#endif
+
+        j.m_parent = this;
+#else
+        static_cast<void>(j);
+        static_cast<void>(old_capacity);
+#endif
+        return j;
+    }
+
+  public:
+    //////////////////////////
+    // JSON parser callback //
+    //////////////////////////
+
+    /// @brief parser event types
+    /// @sa https://json.nlohmann.me/api/basic_json/parse_event_t/
+    using parse_event_t = detail::parse_event_t;
+
+    /// @brief per-element parser callback type
+    /// @sa https://json.nlohmann.me/api/basic_json/parser_callback_t/
+    using parser_callback_t = detail::parser_callback_t<basic_json>;
+
+    //////////////////
+    // constructors //
+    //////////////////
+
+    /// @name constructors and destructors
+    /// Constructors of class @ref basic_json, copy/move constructor, copy
+    /// assignment, static functions creating objects, and the destructor.
+    /// @{
+
+    /// @brief create an empty value with a given type
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(const value_t v)
+        : m_data(v)
+    {
+        assert_invariant();
+    }
+
+    /// @brief create a null object
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(std::nullptr_t = nullptr) noexcept // NOLINT(bugprone-exception-escape)
+        : basic_json(value_t::null)
+    {
+        assert_invariant();
+    }
+
+    /// @brief create a JSON value from compatible types
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    template < typename CompatibleType,
+               typename U = detail::uncvref_t<CompatibleType>,
+               detail::enable_if_t <
+                   !detail::is_basic_json<U>::value && detail::is_compatible_type<basic_json_t, U>::value, int > = 0 >
+    basic_json(CompatibleType && val) noexcept(noexcept( // NOLINT(bugprone-forwarding-reference-overload,bugprone-exception-escape)
+            JSONSerializer<U>::to_json(std::declval<basic_json_t&>(),
+                                       std::forward<CompatibleType>(val))))
+    {
+        JSONSerializer<U>::to_json(*this, std::forward<CompatibleType>(val));
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief create a JSON value from an existing one
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    template < typename BasicJsonType,
+               detail::enable_if_t <
+                   detail::is_basic_json<BasicJsonType>::value&& !std::is_same<basic_json, BasicJsonType>::value, int > = 0 >
+    basic_json(const BasicJsonType& val)
+#if JSON_DIAGNOSTIC_POSITIONS
+        : start_position(val.start_pos()),
+          end_position(val.end_pos())
+#endif
+    {
+        using other_boolean_t = typename BasicJsonType::boolean_t;
+        using other_number_float_t = typename BasicJsonType::number_float_t;
+        using other_number_integer_t = typename BasicJsonType::number_integer_t;
+        using other_number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+        using other_string_t = typename BasicJsonType::string_t;
+        using other_object_t = typename BasicJsonType::object_t;
+        using other_array_t = typename BasicJsonType::array_t;
+        using other_binary_t = typename BasicJsonType::binary_t;
+
+        switch (val.type())
+        {
+            case value_t::boolean:
+                JSONSerializer<other_boolean_t>::to_json(*this, val.template get<other_boolean_t>());
+                break;
+            case value_t::number_float:
+                JSONSerializer<other_number_float_t>::to_json(*this, val.template get<other_number_float_t>());
+                break;
+            case value_t::number_integer:
+                JSONSerializer<other_number_integer_t>::to_json(*this, val.template get<other_number_integer_t>());
+                break;
+            case value_t::number_unsigned:
+                JSONSerializer<other_number_unsigned_t>::to_json(*this, val.template get<other_number_unsigned_t>());
+                break;
+            case value_t::string:
+                JSONSerializer<other_string_t>::to_json(*this, val.template get_ref<const other_string_t&>());
+                break;
+            case value_t::object:
+                JSONSerializer<other_object_t>::to_json(*this, val.template get_ref<const other_object_t&>());
+                break;
+            case value_t::array:
+                JSONSerializer<other_array_t>::to_json(*this, val.template get_ref<const other_array_t&>());
+                break;
+            case value_t::binary:
+                JSONSerializer<other_binary_t>::to_json(*this, val.template get_ref<const other_binary_t&>());
+                break;
+            case value_t::null:
+                *this = nullptr;
+                break;
+            case value_t::discarded:
+                m_data.m_type = value_t::discarded;
+                break;
+            default:            // LCOV_EXCL_LINE
+                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+        }
+        JSON_ASSERT(m_data.m_type == val.type());
+
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief create a container (array or object) from an initializer list
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(initializer_list_t init,
+               bool type_deduction = true,
+               value_t manual_type = value_t::array)
+    {
+        // check if each element is an array with two elements whose first
+        // element is a string
+        bool is_an_object = std::all_of(init.begin(), init.end(),
+                                        [](const detail::json_ref<basic_json>& element_ref)
+        {
+            // The cast is to ensure op[size_type] is called, bearing in mind size_type may not be int;
+            // (many string types can be constructed from 0 via its null-pointer guise, so we get a
+            // broken call to op[key_type], the wrong semantics and a 4804 warning on Windows)
+            return element_ref->is_array() && element_ref->size() == 2 && (*element_ref)[static_cast<size_type>(0)].is_string();
+        });
+
+        // adjust type if type deduction is not wanted
+        if (!type_deduction)
+        {
+            // if array is wanted, do not create an object though possible
+            if (manual_type == value_t::array)
+            {
+                is_an_object = false;
+            }
+
+            // if object is wanted but impossible, throw an exception
+            if (JSON_HEDLEY_UNLIKELY(manual_type == value_t::object && !is_an_object))
+            {
+                JSON_THROW(type_error::create(301, "cannot create object from initializer list", nullptr));
+            }
+        }
+
+        if (is_an_object)
+        {
+            // the initializer list is a list of pairs -> create object
+            m_data.m_type = value_t::object;
+            m_data.m_value = value_t::object;
+
+            for (auto& element_ref : init)
+            {
+                auto element = element_ref.moved_or_copied();
+                m_data.m_value.object->emplace(
+                    std::move(*((*element.m_data.m_value.array)[0].m_data.m_value.string)),
+                    std::move((*element.m_data.m_value.array)[1]));
+            }
+        }
+        else
+        {
+            // the initializer list describes an array -> create array
+            m_data.m_type = value_t::array;
+            m_data.m_value.array = create<array_t>(init.begin(), init.end());
+        }
+
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief explicitly create a binary array (without subtype)
+    /// @sa https://json.nlohmann.me/api/basic_json/binary/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json binary(const typename binary_t::container_type& init)
+    {
+        auto res = basic_json();
+        res.m_data.m_type = value_t::binary;
+        res.m_data.m_value = init;
+        return res;
+    }
+
+    /// @brief explicitly create a binary array (with subtype)
+    /// @sa https://json.nlohmann.me/api/basic_json/binary/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json binary(const typename binary_t::container_type& init, typename binary_t::subtype_type subtype)
+    {
+        auto res = basic_json();
+        res.m_data.m_type = value_t::binary;
+        res.m_data.m_value = binary_t(init, subtype);
+        return res;
+    }
+
+    /// @brief explicitly create a binary array
+    /// @sa https://json.nlohmann.me/api/basic_json/binary/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json binary(typename binary_t::container_type&& init)
+    {
+        auto res = basic_json();
+        res.m_data.m_type = value_t::binary;
+        res.m_data.m_value = std::move(init);
+        return res;
+    }
+
+    /// @brief explicitly create a binary array (with subtype)
+    /// @sa https://json.nlohmann.me/api/basic_json/binary/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json binary(typename binary_t::container_type&& init, typename binary_t::subtype_type subtype)
+    {
+        auto res = basic_json();
+        res.m_data.m_type = value_t::binary;
+        res.m_data.m_value = binary_t(std::move(init), subtype);
+        return res;
+    }
+
+    /// @brief explicitly create an array from an initializer list
+    /// @sa https://json.nlohmann.me/api/basic_json/array/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json array(initializer_list_t init = {})
+    {
+        return basic_json(init, false, value_t::array);
+    }
+
+    /// @brief explicitly create an object from an initializer list
+    /// @sa https://json.nlohmann.me/api/basic_json/object/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json object(initializer_list_t init = {})
+    {
+        return basic_json(init, false, value_t::object);
+    }
+
+    /// @brief construct an array with count copies of given value
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(size_type cnt, const basic_json& val):
+        m_data{cnt, val}
+    {
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief construct a JSON container given an iterator range
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    template < class InputIT, typename std::enable_if <
+                   std::is_same<InputIT, typename basic_json_t::iterator>::value ||
+                   std::is_same<InputIT, typename basic_json_t::const_iterator>::value, int >::type = 0 >
+    basic_json(InputIT first, InputIT last) // NOLINT(performance-unnecessary-value-param)
+    {
+        JSON_ASSERT(first.m_object != nullptr);
+        JSON_ASSERT(last.m_object != nullptr);
+
+        // make sure iterator fits the current value
+        if (JSON_HEDLEY_UNLIKELY(first.m_object != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(201, "iterators are not compatible", nullptr));
+        }
+
+        // copy type from first iterator
+        m_data.m_type = first.m_object->m_data.m_type;
+
+        // check if iterator range is complete for primitive values
+        switch (m_data.m_type)
+        {
+            case value_t::boolean:
+            case value_t::number_float:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::string:
+            {
+                if (JSON_HEDLEY_UNLIKELY(!first.m_it.primitive_iterator.is_begin()
+                                         || !last.m_it.primitive_iterator.is_end()))
+                {
+                    JSON_THROW(invalid_iterator::create(204, "iterators out of range", first.m_object));
+                }
+                break;
+            }
+
+            case value_t::null:
+            case value_t::object:
+            case value_t::array:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+                break;
+        }
+
+        switch (m_data.m_type)
+        {
+            case value_t::number_integer:
+            {
+                m_data.m_value.number_integer = first.m_object->m_data.m_value.number_integer;
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                m_data.m_value.number_unsigned = first.m_object->m_data.m_value.number_unsigned;
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                m_data.m_value.number_float = first.m_object->m_data.m_value.number_float;
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                m_data.m_value.boolean = first.m_object->m_data.m_value.boolean;
+                break;
+            }
+
+            case value_t::string:
+            {
+                m_data.m_value = *first.m_object->m_data.m_value.string;
+                break;
+            }
+
+            case value_t::object:
+            {
+                m_data.m_value.object = create<object_t>(first.m_it.object_iterator,
+                                        last.m_it.object_iterator);
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_data.m_value.array = create<array_t>(first.m_it.array_iterator,
+                                                       last.m_it.array_iterator);
+                break;
+            }
+
+            case value_t::binary:
+            {
+                m_data.m_value = *first.m_object->m_data.m_value.binary;
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                JSON_THROW(invalid_iterator::create(206, detail::concat("cannot construct with iterators from ", first.m_object->type_name()), first.m_object));
+        }
+
+        set_parents();
+        assert_invariant();
+    }
+
+    ///////////////////////////////////////
+    // other constructors and destructor //
+    ///////////////////////////////////////
+
+    template<typename JsonRef,
+             detail::enable_if_t<detail::conjunction<detail::is_json_ref<JsonRef>,
+                                 std::is_same<typename JsonRef::value_type, basic_json>>::value, int> = 0 >
+    basic_json(const JsonRef& ref) : basic_json(ref.moved_or_copied()) {}
+
+    /// @brief copy constructor
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(const basic_json& other)
+        : json_base_class_t(other)
+#if JSON_DIAGNOSTIC_POSITIONS
+        , start_position(other.start_position)
+        , end_position(other.end_position)
+#endif
+    {
+        m_data.m_type = other.m_data.m_type;
+        // check of passed value is valid
+        other.assert_invariant();
+
+        switch (m_data.m_type)
+        {
+            case value_t::object:
+            {
+                m_data.m_value = *other.m_data.m_value.object;
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_data.m_value = *other.m_data.m_value.array;
+                break;
+            }
+
+            case value_t::string:
+            {
+                m_data.m_value = *other.m_data.m_value.string;
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                m_data.m_value = other.m_data.m_value.boolean;
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                m_data.m_value = other.m_data.m_value.number_integer;
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                m_data.m_value = other.m_data.m_value.number_unsigned;
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                m_data.m_value = other.m_data.m_value.number_float;
+                break;
+            }
+
+            case value_t::binary:
+            {
+                m_data.m_value = *other.m_data.m_value.binary;
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                break;
+        }
+
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief move constructor
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(basic_json&& other) noexcept
+        : json_base_class_t(std::forward<json_base_class_t>(other)),
+          m_data(std::move(other.m_data)) // cppcheck-suppress[accessForwarded] TODO check
+#if JSON_DIAGNOSTIC_POSITIONS
+        , start_position(other.start_position) // cppcheck-suppress[accessForwarded] TODO check
+        , end_position(other.end_position) // cppcheck-suppress[accessForwarded] TODO check
+#endif
+    {
+        // check that passed value is valid
+        other.assert_invariant(false); // cppcheck-suppress[accessForwarded]
+
+        // invalidate payload
+        other.m_data.m_type = value_t::null;
+        other.m_data.m_value = {};
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        other.start_position = std::string::npos;
+        other.end_position = std::string::npos;
+#endif
+
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief copy assignment
+    /// @sa https://json.nlohmann.me/api/basic_json/operator=/
+    basic_json& operator=(basic_json other) noexcept (
+        std::is_nothrow_move_constructible<value_t>::value&&
+        std::is_nothrow_move_assignable<value_t>::value&&
+        std::is_nothrow_move_constructible<json_value>::value&&
+        std::is_nothrow_move_assignable<json_value>::value&&
+        std::is_nothrow_move_assignable<json_base_class_t>::value
+    )
+    {
+        // check that passed value is valid
+        other.assert_invariant();
+
+        using std::swap;
+        swap(m_data.m_type, other.m_data.m_type);
+        swap(m_data.m_value, other.m_data.m_value);
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        swap(start_position, other.start_position);
+        swap(end_position, other.end_position);
+#endif
+
+        json_base_class_t::operator=(std::move(other));
+
+        set_parents();
+        assert_invariant();
+        return *this;
+    }
+
+    /// @brief destructor
+    /// @sa https://json.nlohmann.me/api/basic_json/~basic_json/
+    ~basic_json() noexcept
+    {
+        assert_invariant(false);
+    }
+
+    /// @}
+
+  public:
+    ///////////////////////
+    // object inspection //
+    ///////////////////////
+
+    /// @name object inspection
+    /// Functions to inspect the type of a JSON value.
+    /// @{
+
+    /// @brief serialization
+    /// @sa https://json.nlohmann.me/api/basic_json/dump/
+    string_t dump(const int indent = -1,
+                  const char indent_char = ' ',
+                  const bool ensure_ascii = false,
+                  const error_handler_t error_handler = error_handler_t::strict) const
+    {
+        string_t result;
+        serializer s(detail::output_adapter<char, string_t>(result), indent_char, error_handler);
+
+        if (indent >= 0)
+        {
+            s.dump(*this, true, ensure_ascii, static_cast<unsigned int>(indent));
+        }
+        else
+        {
+            s.dump(*this, false, ensure_ascii, 0);
+        }
+
+        return result;
+    }
+
+    /// @brief return the type of the JSON value (explicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/type/
+    constexpr value_t type() const noexcept
+    {
+        return m_data.m_type;
+    }
+
+    /// @brief return whether type is primitive
+    /// @sa https://json.nlohmann.me/api/basic_json/is_primitive/
+    constexpr bool is_primitive() const noexcept
+    {
+        return is_null() || is_string() || is_boolean() || is_number() || is_binary();
+    }
+
+    /// @brief return whether type is structured
+    /// @sa https://json.nlohmann.me/api/basic_json/is_structured/
+    constexpr bool is_structured() const noexcept
+    {
+        return is_array() || is_object();
+    }
+
+    /// @brief return whether value is null
+    /// @sa https://json.nlohmann.me/api/basic_json/is_null/
+    constexpr bool is_null() const noexcept
+    {
+        return m_data.m_type == value_t::null;
+    }
+
+    /// @brief return whether value is a boolean
+    /// @sa https://json.nlohmann.me/api/basic_json/is_boolean/
+    constexpr bool is_boolean() const noexcept
+    {
+        return m_data.m_type == value_t::boolean;
+    }
+
+    /// @brief return whether value is a number
+    /// @sa https://json.nlohmann.me/api/basic_json/is_number/
+    constexpr bool is_number() const noexcept
+    {
+        return is_number_integer() || is_number_float();
+    }
+
+    /// @brief return whether value is an integer number
+    /// @sa https://json.nlohmann.me/api/basic_json/is_number_integer/
+    constexpr bool is_number_integer() const noexcept
+    {
+        return m_data.m_type == value_t::number_integer || m_data.m_type == value_t::number_unsigned;
+    }
+
+    /// @brief return whether value is an unsigned integer number
+    /// @sa https://json.nlohmann.me/api/basic_json/is_number_unsigned/
+    constexpr bool is_number_unsigned() const noexcept
+    {
+        return m_data.m_type == value_t::number_unsigned;
+    }
+
+    /// @brief return whether value is a floating-point number
+    /// @sa https://json.nlohmann.me/api/basic_json/is_number_float/
+    constexpr bool is_number_float() const noexcept
+    {
+        return m_data.m_type == value_t::number_float;
+    }
+
+    /// @brief return whether value is an object
+    /// @sa https://json.nlohmann.me/api/basic_json/is_object/
+    constexpr bool is_object() const noexcept
+    {
+        return m_data.m_type == value_t::object;
+    }
+
+    /// @brief return whether value is an array
+    /// @sa https://json.nlohmann.me/api/basic_json/is_array/
+    constexpr bool is_array() const noexcept
+    {
+        return m_data.m_type == value_t::array;
+    }
+
+    /// @brief return whether value is a string
+    /// @sa https://json.nlohmann.me/api/basic_json/is_string/
+    constexpr bool is_string() const noexcept
+    {
+        return m_data.m_type == value_t::string;
+    }
+
+    /// @brief return whether value is a binary array
+    /// @sa https://json.nlohmann.me/api/basic_json/is_binary/
+    constexpr bool is_binary() const noexcept
+    {
+        return m_data.m_type == value_t::binary;
+    }
+
+    /// @brief return whether value is discarded
+    /// @sa https://json.nlohmann.me/api/basic_json/is_discarded/
+    constexpr bool is_discarded() const noexcept
+    {
+        return m_data.m_type == value_t::discarded;
+    }
+
+    /// @brief return the type of the JSON value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_value_t/
+    constexpr operator value_t() const noexcept
+    {
+        return m_data.m_type;
+    }
+
+    /// @}
+
+  private:
+    //////////////////
+    // value access //
+    //////////////////
+
+    /// get a boolean (explicit)
+    boolean_t get_impl(boolean_t* /*unused*/) const
+    {
+        if (JSON_HEDLEY_LIKELY(is_boolean()))
+        {
+            return m_data.m_value.boolean;
+        }
+
+        JSON_THROW(type_error::create(302, detail::concat("type must be boolean, but is ", type_name()), this));
+    }
+
+    /// get a pointer to the value (object)
+    object_t* get_impl_ptr(object_t* /*unused*/) noexcept
+    {
+        return is_object() ? m_data.m_value.object : nullptr;
+    }
+
+    /// get a pointer to the value (object)
+    constexpr const object_t* get_impl_ptr(const object_t* /*unused*/) const noexcept
+    {
+        return is_object() ? m_data.m_value.object : nullptr;
+    }
+
+    /// get a pointer to the value (array)
+    array_t* get_impl_ptr(array_t* /*unused*/) noexcept
+    {
+        return is_array() ? m_data.m_value.array : nullptr;
+    }
+
+    /// get a pointer to the value (array)
+    constexpr const array_t* get_impl_ptr(const array_t* /*unused*/) const noexcept
+    {
+        return is_array() ? m_data.m_value.array : nullptr;
+    }
+
+    /// get a pointer to the value (string)
+    string_t* get_impl_ptr(string_t* /*unused*/) noexcept
+    {
+        return is_string() ? m_data.m_value.string : nullptr;
+    }
+
+    /// get a pointer to the value (string)
+    constexpr const string_t* get_impl_ptr(const string_t* /*unused*/) const noexcept
+    {
+        return is_string() ? m_data.m_value.string : nullptr;
+    }
+
+    /// get a pointer to the value (boolean)
+    boolean_t* get_impl_ptr(boolean_t* /*unused*/) noexcept
+    {
+        return is_boolean() ? &m_data.m_value.boolean : nullptr;
+    }
+
+    /// get a pointer to the value (boolean)
+    constexpr const boolean_t* get_impl_ptr(const boolean_t* /*unused*/) const noexcept
+    {
+        return is_boolean() ? &m_data.m_value.boolean : nullptr;
+    }
+
+    /// get a pointer to the value (integer number)
+    number_integer_t* get_impl_ptr(number_integer_t* /*unused*/) noexcept
+    {
+        return m_data.m_type == value_t::number_integer ? &m_data.m_value.number_integer : nullptr;
+    }
+
+    /// get a pointer to the value (integer number)
+    constexpr const number_integer_t* get_impl_ptr(const number_integer_t* /*unused*/) const noexcept
+    {
+        return m_data.m_type == value_t::number_integer ? &m_data.m_value.number_integer : nullptr;
+    }
+
+    /// get a pointer to the value (unsigned number)
+    number_unsigned_t* get_impl_ptr(number_unsigned_t* /*unused*/) noexcept
+    {
+        return is_number_unsigned() ? &m_data.m_value.number_unsigned : nullptr;
+    }
+
+    /// get a pointer to the value (unsigned number)
+    constexpr const number_unsigned_t* get_impl_ptr(const number_unsigned_t* /*unused*/) const noexcept
+    {
+        return is_number_unsigned() ? &m_data.m_value.number_unsigned : nullptr;
+    }
+
+    /// get a pointer to the value (floating-point number)
+    number_float_t* get_impl_ptr(number_float_t* /*unused*/) noexcept
+    {
+        return is_number_float() ? &m_data.m_value.number_float : nullptr;
+    }
+
+    /// get a pointer to the value (floating-point number)
+    constexpr const number_float_t* get_impl_ptr(const number_float_t* /*unused*/) const noexcept
+    {
+        return is_number_float() ? &m_data.m_value.number_float : nullptr;
+    }
+
+    /// get a pointer to the value (binary)
+    binary_t* get_impl_ptr(binary_t* /*unused*/) noexcept
+    {
+        return is_binary() ? m_data.m_value.binary : nullptr;
+    }
+
+    /// get a pointer to the value (binary)
+    constexpr const binary_t* get_impl_ptr(const binary_t* /*unused*/) const noexcept
+    {
+        return is_binary() ? m_data.m_value.binary : nullptr;
+    }
+
+    /*!
+    @brief helper function to implement get_ref()
+
+    This function helps to implement get_ref() without code duplication for
+    const and non-const overloads
+
+    @tparam ThisType will be deduced as `basic_json` or `const basic_json`
+
+    @throw type_error.303 if ReferenceType does not match underlying value
+    type of the current JSON
+    */
+    template<typename ReferenceType, typename ThisType>
+    static ReferenceType get_ref_impl(ThisType& obj)
+    {
+        // delegate the call to get_ptr<>()
+        auto* ptr = obj.template get_ptr<typename std::add_pointer<ReferenceType>::type>();
+
+        if (JSON_HEDLEY_LIKELY(ptr != nullptr))
+        {
+            return *ptr;
+        }
+
+        JSON_THROW(type_error::create(303, detail::concat("incompatible ReferenceType for get_ref, actual type is ", obj.type_name()), &obj));
+    }
+
+  public:
+    /// @name value access
+    /// Direct access to the stored value of a JSON value.
+    /// @{
+
+    /// @brief get a pointer value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_ptr/
+    template<typename PointerType, typename std::enable_if<
+                 std::is_pointer<PointerType>::value, int>::type = 0>
+    auto get_ptr() noexcept -> decltype(std::declval<basic_json_t&>().get_impl_ptr(std::declval<PointerType>()))
+    {
+        // delegate the call to get_impl_ptr<>()
+        return get_impl_ptr(static_cast<PointerType>(nullptr));
+    }
+
+    /// @brief get a pointer value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_ptr/
+    template < typename PointerType, typename std::enable_if <
+                   std::is_pointer<PointerType>::value&&
+                   std::is_const<typename std::remove_pointer<PointerType>::type>::value, int >::type = 0 >
+    constexpr auto get_ptr() const noexcept -> decltype(std::declval<const basic_json_t&>().get_impl_ptr(std::declval<PointerType>()))
+    {
+        // delegate the call to get_impl_ptr<>() const
+        return get_impl_ptr(static_cast<PointerType>(nullptr));
+    }
+
+  private:
+    /*!
+    @brief get a value (explicit)
+
+    Explicit type conversion between the JSON value and a compatible value
+    which is [CopyConstructible](https://en.cppreference.com/w/cpp/named_req/CopyConstructible)
+    and [DefaultConstructible](https://en.cppreference.com/w/cpp/named_req/DefaultConstructible).
+    The value is converted by calling the @ref json_serializer<ValueType>
+    `from_json()` method.
+
+    The function is equivalent to executing
+    @code {.cpp}
+    ValueType ret;
+    JSONSerializer<ValueType>::from_json(*this, ret);
+    return ret;
+    @endcode
+
+    This overloads is chosen if:
+    - @a ValueType is not @ref basic_json,
+    - @ref json_serializer<ValueType> has a `from_json()` method of the form
+      `void from_json(const basic_json&, ValueType&)`, and
+    - @ref json_serializer<ValueType> does not have a `from_json()` method of
+      the form `ValueType from_json(const basic_json&)`
+
+    @tparam ValueType the returned value type
+
+    @return copy of the JSON value, converted to @a ValueType
+
+    @throw what @ref json_serializer<ValueType> `from_json()` method throws
+
+    @liveexample{The example below shows several conversions from JSON values
+    to other types. There a few things to note: (1) Floating-point numbers can
+    be converted to integers\, (2) A JSON array can be converted to a standard
+    `std::vector<short>`\, (3) A JSON object can be converted to C++
+    associative containers such as `std::unordered_map<std::string\,
+    json>`.,get__ValueType_const}
+
+    @since version 2.1.0
+    */
+    template < typename ValueType,
+               detail::enable_if_t <
+                   detail::is_default_constructible<ValueType>::value&&
+                   detail::has_from_json<basic_json_t, ValueType>::value,
+                   int > = 0 >
+    ValueType get_impl(detail::priority_tag<0> /*unused*/) const noexcept(noexcept(
+            JSONSerializer<ValueType>::from_json(std::declval<const basic_json_t&>(), std::declval<ValueType&>())))
+    {
+        auto ret = ValueType();
+        JSONSerializer<ValueType>::from_json(*this, ret);
+        return ret;
+    }
+
+    /*!
+    @brief get a value (explicit); special case
+
+    Explicit type conversion between the JSON value and a compatible value
+    which is **not** [CopyConstructible](https://en.cppreference.com/w/cpp/named_req/CopyConstructible)
+    and **not** [DefaultConstructible](https://en.cppreference.com/w/cpp/named_req/DefaultConstructible).
+    The value is converted by calling the @ref json_serializer<ValueType>
+    `from_json()` method.
+
+    The function is equivalent to executing
+    @code {.cpp}
+    return JSONSerializer<ValueType>::from_json(*this);
+    @endcode
+
+    This overloads is chosen if:
+    - @a ValueType is not @ref basic_json and
+    - @ref json_serializer<ValueType> has a `from_json()` method of the form
+      `ValueType from_json(const basic_json&)`
+
+    @note If @ref json_serializer<ValueType> has both overloads of
+    `from_json()`, this one is chosen.
+
+    @tparam ValueType the returned value type
+
+    @return copy of the JSON value, converted to @a ValueType
+
+    @throw what @ref json_serializer<ValueType> `from_json()` method throws
+
+    @since version 2.1.0
+    */
+    template < typename ValueType,
+               detail::enable_if_t <
+                   detail::has_non_default_from_json<basic_json_t, ValueType>::value,
+                   int > = 0 >
+    ValueType get_impl(detail::priority_tag<1> /*unused*/) const noexcept(noexcept(
+            JSONSerializer<ValueType>::from_json(std::declval<const basic_json_t&>())))
+    {
+        return JSONSerializer<ValueType>::from_json(*this);
+    }
+
+    /*!
+    @brief get special-case overload
+
+    This overloads converts the current @ref basic_json in a different
+    @ref basic_json type
+
+    @tparam BasicJsonType == @ref basic_json
+
+    @return a copy of *this, converted into @a BasicJsonType
+
+    @complexity Depending on the implementation of the called `from_json()`
+                method.
+
+    @since version 3.2.0
+    */
+    template < typename BasicJsonType,
+               detail::enable_if_t <
+                   detail::is_basic_json<BasicJsonType>::value,
+                   int > = 0 >
+    BasicJsonType get_impl(detail::priority_tag<2> /*unused*/) const
+    {
+        return *this;
+    }
+
+    /*!
+    @brief get special-case overload
+
+    This overloads avoids a lot of template boilerplate, it can be seen as the
+    identity method
+
+    @tparam BasicJsonType == @ref basic_json
+
+    @return a copy of *this
+
+    @complexity Constant.
+
+    @since version 2.1.0
+    */
+    template<typename BasicJsonType,
+             detail::enable_if_t<
+                 std::is_same<BasicJsonType, basic_json_t>::value,
+                 int> = 0>
+    basic_json get_impl(detail::priority_tag<3> /*unused*/) const
+    {
+        return *this;
+    }
+
+    /*!
+    @brief get a pointer value (explicit)
+    @copydoc get()
+    */
+    template<typename PointerType,
+             detail::enable_if_t<
+                 std::is_pointer<PointerType>::value,
+                 int> = 0>
+    constexpr auto get_impl(detail::priority_tag<4> /*unused*/) const noexcept
+    -> decltype(std::declval<const basic_json_t&>().template get_ptr<PointerType>())
+    {
+        // delegate the call to get_ptr
+        return get_ptr<PointerType>();
+    }
+
+  public:
+    /*!
+    @brief get a (pointer) value (explicit)
+
+    Performs explicit type conversion between the JSON value and a compatible value if required.
+
+    - If the requested type is a pointer to the internally stored JSON value that pointer is returned.
+    No copies are made.
+
+    - If the requested type is the current @ref basic_json, or a different @ref basic_json convertible
+    from the current @ref basic_json.
+
+    - Otherwise the value is converted by calling the @ref json_serializer<ValueType> `from_json()`
+    method.
+
+    @tparam ValueTypeCV the provided value type
+    @tparam ValueType the returned value type
+
+    @return copy of the JSON value, converted to @tparam ValueType if necessary
+
+    @throw what @ref json_serializer<ValueType> `from_json()` method throws if conversion is required
+
+    @since version 2.1.0
+    */
+    template < typename ValueTypeCV, typename ValueType = detail::uncvref_t<ValueTypeCV>>
+#if defined(JSON_HAS_CPP_14)
+    constexpr
+#endif
+    auto get() const noexcept(
+    noexcept(std::declval<const basic_json_t&>().template get_impl<ValueType>(detail::priority_tag<4> {})))
+    -> decltype(std::declval<const basic_json_t&>().template get_impl<ValueType>(detail::priority_tag<4> {}))
+    {
+        // we cannot static_assert on ValueTypeCV being non-const, because
+        // there is support for get<const basic_json_t>(), which is why we
+        // still need the uncvref
+        static_assert(!std::is_reference<ValueTypeCV>::value,
+                      "get() cannot be used with reference types, you might want to use get_ref()");
+        return get_impl<ValueType>(detail::priority_tag<4> {});
+    }
+
+    /*!
+    @brief get a pointer value (explicit)
+
+    Explicit pointer access to the internally stored JSON value. No copies are
+    made.
+
+    @warning The pointer becomes invalid if the underlying JSON object
+    changes.
+
+    @tparam PointerType pointer type; must be a pointer to @ref array_t, @ref
+    object_t, @ref string_t, @ref boolean_t, @ref number_integer_t,
+    @ref number_unsigned_t, or @ref number_float_t.
+
+    @return pointer to the internally stored JSON value if the requested
+    pointer type @a PointerType fits to the JSON value; `nullptr` otherwise
+
+    @complexity Constant.
+
+    @liveexample{The example below shows how pointers to internal values of a
+    JSON value can be requested. Note that no type conversions are made and a
+    `nullptr` is returned if the value and the requested pointer type does not
+    match.,get__PointerType}
+
+    @sa see @ref get_ptr() for explicit pointer-member access
+
+    @since version 1.0.0
+    */
+    template<typename PointerType, typename std::enable_if<
+                 std::is_pointer<PointerType>::value, int>::type = 0>
+    auto get() noexcept -> decltype(std::declval<basic_json_t&>().template get_ptr<PointerType>())
+    {
+        // delegate the call to get_ptr
+        return get_ptr<PointerType>();
+    }
+
+    /// @brief get a value (explicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_to/
+    template < typename ValueType,
+               detail::enable_if_t <
+                   !detail::is_basic_json<ValueType>::value&&
+                   detail::has_from_json<basic_json_t, ValueType>::value,
+                   int > = 0 >
+    ValueType & get_to(ValueType& v) const noexcept(noexcept(
+            JSONSerializer<ValueType>::from_json(std::declval<const basic_json_t&>(), v)))
+    {
+        JSONSerializer<ValueType>::from_json(*this, v);
+        return v;
+    }
+
+    // specialization to allow calling get_to with a basic_json value
+    // see https://github.com/nlohmann/json/issues/2175
+    template<typename ValueType,
+             detail::enable_if_t <
+                 detail::is_basic_json<ValueType>::value,
+                 int> = 0>
+    ValueType & get_to(ValueType& v) const
+    {
+        v = *this;
+        return v;
+    }
+
+    template <
+        typename T, std::size_t N,
+        typename Array = T (&)[N], // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+        detail::enable_if_t <
+            detail::has_from_json<basic_json_t, Array>::value, int > = 0 >
+    Array get_to(T (&v)[N]) const // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+    noexcept(noexcept(JSONSerializer<Array>::from_json(
+                          std::declval<const basic_json_t&>(), v)))
+    {
+        JSONSerializer<Array>::from_json(*this, v);
+        return v;
+    }
+
+    /// @brief get a reference value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_ref/
+    template<typename ReferenceType, typename std::enable_if<
+                 std::is_reference<ReferenceType>::value, int>::type = 0>
+    ReferenceType get_ref()
+    {
+        // delegate call to get_ref_impl
+        return get_ref_impl<ReferenceType>(*this);
+    }
+
+    /// @brief get a reference value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_ref/
+    template < typename ReferenceType, typename std::enable_if <
+                   std::is_reference<ReferenceType>::value&&
+                   std::is_const<typename std::remove_reference<ReferenceType>::type>::value, int >::type = 0 >
+    ReferenceType get_ref() const
+    {
+        // delegate call to get_ref_impl
+        return get_ref_impl<ReferenceType>(*this);
+    }
+
+    /*!
+    @brief get a value (implicit)
+
+    Implicit type conversion between the JSON value and a compatible value.
+    The call is realized by calling @ref get() const.
+
+    @tparam ValueType non-pointer type compatible to the JSON value, for
+    instance `int` for JSON integer numbers, `bool` for JSON booleans, or
+    `std::vector` types for JSON arrays. The character type of @ref string_t
+    as well as an initializer list of this type is excluded to avoid
+    ambiguities as these types implicitly convert to `std::string`.
+
+    @return copy of the JSON value, converted to type @a ValueType
+
+    @throw type_error.302 in case passed type @a ValueType is incompatible
+    to the JSON value type (e.g., the JSON value is of type boolean, but a
+    string is requested); see example below
+
+    @complexity Linear in the size of the JSON value.
+
+    @liveexample{The example below shows several conversions from JSON values
+    to other types. There a few things to note: (1) Floating-point numbers can
+    be converted to integers\, (2) A JSON array can be converted to a standard
+    `std::vector<short>`\, (3) A JSON object can be converted to C++
+    associative containers such as `std::unordered_map<std::string\,
+    json>`.,operator__ValueType}
+
+    @since version 1.0.0
+    */
+    template < typename ValueType, typename std::enable_if <
+                   detail::conjunction <
+                       detail::negation<std::is_pointer<ValueType>>,
+                       detail::negation<std::is_same<ValueType, std::nullptr_t>>,
+                       detail::negation<std::is_same<ValueType, detail::json_ref<basic_json>>>,
+                                        detail::negation<std::is_same<ValueType, typename string_t::value_type>>,
+                                        detail::negation<detail::is_basic_json<ValueType>>,
+                                        detail::negation<std::is_same<ValueType, std::initializer_list<typename string_t::value_type>>>,
+#if defined(JSON_HAS_CPP_17) && (defined(__GNUC__) || (defined(_MSC_VER) && _MSC_VER >= 1910 && _MSC_VER <= 1914))
+                                                detail::negation<std::is_same<ValueType, std::string_view>>,
+#endif
+#if defined(JSON_HAS_CPP_17) && JSON_HAS_STATIC_RTTI
+                                                detail::negation<std::is_same<ValueType, std::any>>,
+#endif
+                                                detail::is_detected_lazy<detail::get_template_function, const basic_json_t&, ValueType>
+                                                >::value, int >::type = 0 >
+                                        JSON_EXPLICIT operator ValueType() const
+    {
+        // delegate the call to get<>() const
+        return get<ValueType>();
+    }
+
+    /// @brief get a binary value
+    /// @sa https://json.nlohmann.me/api/basic_json/get_binary/
+    binary_t& get_binary()
+    {
+        if (!is_binary())
+        {
+            JSON_THROW(type_error::create(302, detail::concat("type must be binary, but is ", type_name()), this));
+        }
+
+        return *get_ptr<binary_t*>();
+    }
+
+    /// @brief get a binary value
+    /// @sa https://json.nlohmann.me/api/basic_json/get_binary/
+    const binary_t& get_binary() const
+    {
+        if (!is_binary())
+        {
+            JSON_THROW(type_error::create(302, detail::concat("type must be binary, but is ", type_name()), this));
+        }
+
+        return *get_ptr<const binary_t*>();
+    }
+
+    /// @}
+
+    ////////////////////
+    // element access //
+    ////////////////////
+
+    /// @name element access
+    /// Access to the JSON value.
+    /// @{
+
+    /// @brief access specified array element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    reference at(size_type idx)
+    {
+        // at only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            JSON_TRY
+            {
+                return set_parent(m_data.m_value.array->at(idx));
+            }
+            JSON_CATCH (std::out_of_range&)
+            {
+                // create better exception explanation
+                JSON_THROW(out_of_range::create(401, detail::concat("array index ", std::to_string(idx), " is out of range"), this));
+            } // cppcheck-suppress[missingReturn]
+        }
+        else
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+    }
+
+    /// @brief access specified array element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    const_reference at(size_type idx) const
+    {
+        // at only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            JSON_TRY
+            {
+                return m_data.m_value.array->at(idx);
+            }
+            JSON_CATCH (std::out_of_range&)
+            {
+                // create better exception explanation
+                JSON_THROW(out_of_range::create(401, detail::concat("array index ", std::to_string(idx), " is out of range"), this));
+            } // cppcheck-suppress[missingReturn]
+        }
+        else
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+    }
+
+    /// @brief access specified object element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    reference at(const typename object_t::key_type& key)
+    {
+        // at only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+
+        auto it = m_data.m_value.object->find(key);
+        if (it == m_data.m_value.object->end())
+        {
+            JSON_THROW(out_of_range::create(403, detail::concat("key '", key, "' not found"), this));
+        }
+        return set_parent(it->second);
+    }
+
+    /// @brief access specified object element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    reference at(KeyType && key)
+    {
+        // at only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+
+        auto it = m_data.m_value.object->find(std::forward<KeyType>(key));
+        if (it == m_data.m_value.object->end())
+        {
+            JSON_THROW(out_of_range::create(403, detail::concat("key '", string_t(std::forward<KeyType>(key)), "' not found"), this));
+        }
+        return set_parent(it->second);
+    }
+
+    /// @brief access specified object element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    const_reference at(const typename object_t::key_type& key) const
+    {
+        // at only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+
+        auto it = m_data.m_value.object->find(key);
+        if (it == m_data.m_value.object->end())
+        {
+            JSON_THROW(out_of_range::create(403, detail::concat("key '", key, "' not found"), this));
+        }
+        return it->second;
+    }
+
+    /// @brief access specified object element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    const_reference at(KeyType && key) const
+    {
+        // at only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+
+        auto it = m_data.m_value.object->find(std::forward<KeyType>(key));
+        if (it == m_data.m_value.object->end())
+        {
+            JSON_THROW(out_of_range::create(403, detail::concat("key '", string_t(std::forward<KeyType>(key)), "' not found"), this));
+        }
+        return it->second;
+    }
+
+    /// @brief access specified array element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    reference operator[](size_type idx)
+    {
+        // implicitly convert null value to an empty array
+        if (is_null())
+        {
+            m_data.m_type = value_t::array;
+            m_data.m_value.array = create<array_t>();
+            assert_invariant();
+        }
+
+        // operator[] only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            // fill up array with null values if given idx is outside range
+            if (idx >= m_data.m_value.array->size())
+            {
+#if JSON_DIAGNOSTICS
+                // remember array size & capacity before resizing
+                const auto old_size = m_data.m_value.array->size();
+                const auto old_capacity = m_data.m_value.array->capacity();
+#endif
+                m_data.m_value.array->resize(idx + 1);
+
+#if JSON_DIAGNOSTICS
+                if (JSON_HEDLEY_UNLIKELY(m_data.m_value.array->capacity() != old_capacity))
+                {
+                    // capacity has changed: update all parents
+                    set_parents();
+                }
+                else
+                {
+                    // set parent for values added above
+                    set_parents(begin() + static_cast<typename iterator::difference_type>(old_size), static_cast<typename iterator::difference_type>(idx + 1 - old_size));
+                }
+#endif
+                assert_invariant();
+            }
+
+            return m_data.m_value.array->operator[](idx);
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a numeric argument with ", type_name()), this));
+    }
+
+    /// @brief access specified array element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    const_reference operator[](size_type idx) const
+    {
+        // const operator[] only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            return m_data.m_value.array->operator[](idx);
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a numeric argument with ", type_name()), this));
+    }
+
+    /// @brief access specified object element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    reference operator[](typename object_t::key_type key) // NOLINT(performance-unnecessary-value-param)
+    {
+        // implicitly convert null value to an empty object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value.object = create<object_t>();
+            assert_invariant();
+        }
+
+        // operator[] only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            auto result = m_data.m_value.object->emplace(std::move(key), nullptr);
+            return set_parent(result.first->second);
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a string argument with ", type_name()), this));
+    }
+
+    /// @brief access specified object element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    const_reference operator[](const typename object_t::key_type& key) const
+    {
+        // const operator[] only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            auto it = m_data.m_value.object->find(key);
+            JSON_ASSERT(it != m_data.m_value.object->end());
+            return it->second;
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a string argument with ", type_name()), this));
+    }
+
+    // these two functions resolve a (const) char * ambiguity affecting Clang and MSVC
+    // (they seemingly cannot be constrained to resolve the ambiguity)
+    template<typename T>
+    reference operator[](T* key)
+    {
+        return operator[](typename object_t::key_type(key));
+    }
+
+    template<typename T>
+    const_reference operator[](T* key) const
+    {
+        return operator[](typename object_t::key_type(key));
+    }
+
+    /// @brief access specified object element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int > = 0 >
+    reference operator[](KeyType && key)
+    {
+        // implicitly convert null value to an empty object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value.object = create<object_t>();
+            assert_invariant();
+        }
+
+        // operator[] only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            auto result = m_data.m_value.object->emplace(std::forward<KeyType>(key), nullptr);
+            return set_parent(result.first->second);
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a string argument with ", type_name()), this));
+    }
+
+    /// @brief access specified object element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int > = 0 >
+    const_reference operator[](KeyType && key) const
+    {
+        // const operator[] only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            auto it = m_data.m_value.object->find(std::forward<KeyType>(key));
+            JSON_ASSERT(it != m_data.m_value.object->end());
+            return it->second;
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a string argument with ", type_name()), this));
+    }
+
+  private:
+    template<typename KeyType>
+    using is_comparable_with_object_key = detail::is_comparable <
+        object_comparator_t, const typename object_t::key_type&, KeyType >;
+
+    template<typename ValueType>
+    using value_return_type = std::conditional <
+        detail::is_c_string_uncvref<ValueType>::value,
+        string_t, typename std::decay<ValueType>::type >;
+
+  public:
+    /// @brief access specified object element with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, detail::enable_if_t <
+                   !detail::is_transparent<object_comparator_t>::value
+                   && detail::is_getable<basic_json_t, ValueType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ValueType value(const typename object_t::key_type& key, const ValueType& default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if key is found, return value and given default value otherwise
+            const auto it = find(key);
+            if (it != end())
+            {
+                return it->template get<ValueType>();
+            }
+
+            return default_value;
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, class ReturnType = typename value_return_type<ValueType>::type,
+               detail::enable_if_t <
+                   !detail::is_transparent<object_comparator_t>::value
+                   && detail::is_getable<basic_json_t, ReturnType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ReturnType value(const typename object_t::key_type& key, ValueType && default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if key is found, return value and given default value otherwise
+            const auto it = find(key);
+            if (it != end())
+            {
+                return it->template get<ReturnType>();
+            }
+
+            return std::forward<ValueType>(default_value);
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, class KeyType, detail::enable_if_t <
+                   detail::is_transparent<object_comparator_t>::value
+                   && !detail::is_json_pointer<KeyType>::value
+                   && is_comparable_with_object_key<KeyType>::value
+                   && detail::is_getable<basic_json_t, ValueType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ValueType value(KeyType && key, const ValueType& default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if key is found, return value and given default value otherwise
+            const auto it = find(std::forward<KeyType>(key));
+            if (it != end())
+            {
+                return it->template get<ValueType>();
+            }
+
+            return default_value;
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element via JSON Pointer with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, class KeyType, class ReturnType = typename value_return_type<ValueType>::type,
+               detail::enable_if_t <
+                   detail::is_transparent<object_comparator_t>::value
+                   && !detail::is_json_pointer<KeyType>::value
+                   && is_comparable_with_object_key<KeyType>::value
+                   && detail::is_getable<basic_json_t, ReturnType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ReturnType value(KeyType && key, ValueType && default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if key is found, return value and given default value otherwise
+            const auto it = find(std::forward<KeyType>(key));
+            if (it != end())
+            {
+                return it->template get<ReturnType>();
+            }
+
+            return std::forward<ValueType>(default_value);
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element via JSON Pointer with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, detail::enable_if_t <
+                   detail::is_getable<basic_json_t, ValueType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ValueType value(const json_pointer& ptr, const ValueType& default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if pointer resolves a value, return it or use default value
+            JSON_TRY
+            {
+                return ptr.get_checked(this).template get<ValueType>();
+            }
+            JSON_INTERNAL_CATCH (out_of_range&)
+            {
+                return default_value;
+            }
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element via JSON Pointer with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, class ReturnType = typename value_return_type<ValueType>::type,
+               detail::enable_if_t <
+                   detail::is_getable<basic_json_t, ReturnType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ReturnType value(const json_pointer& ptr, ValueType && default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if pointer resolves a value, return it or use default value
+            JSON_TRY
+            {
+                return ptr.get_checked(this).template get<ReturnType>();
+            }
+            JSON_INTERNAL_CATCH (out_of_range&)
+            {
+                return std::forward<ValueType>(default_value);
+            }
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    template < class ValueType, class BasicJsonType, detail::enable_if_t <
+                   detail::is_basic_json<BasicJsonType>::value
+                   && detail::is_getable<basic_json_t, ValueType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    ValueType value(const ::nlohmann::json_pointer<BasicJsonType>& ptr, const ValueType& default_value) const
+    {
+        return value(ptr.convert(), default_value);
+    }
+
+    template < class ValueType, class BasicJsonType, class ReturnType = typename value_return_type<ValueType>::type,
+               detail::enable_if_t <
+                   detail::is_basic_json<BasicJsonType>::value
+                   && detail::is_getable<basic_json_t, ReturnType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    ReturnType value(const ::nlohmann::json_pointer<BasicJsonType>& ptr, ValueType && default_value) const
+    {
+        return value(ptr.convert(), std::forward<ValueType>(default_value));
+    }
+
+    /// @brief access the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/front/
+    reference front()
+    {
+        return *begin();
+    }
+
+    /// @brief access the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/front/
+    const_reference front() const
+    {
+        return *cbegin();
+    }
+
+    /// @brief access the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/back/
+    reference back()
+    {
+        auto tmp = end();
+        --tmp;
+        return *tmp;
+    }
+
+    /// @brief access the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/back/
+    const_reference back() const
+    {
+        auto tmp = cend();
+        --tmp;
+        return *tmp;
+    }
+
+    /// @brief remove element given an iterator
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    template < class IteratorType, detail::enable_if_t <
+                   std::is_same<IteratorType, typename basic_json_t::iterator>::value ||
+                   std::is_same<IteratorType, typename basic_json_t::const_iterator>::value, int > = 0 >
+    IteratorType erase(IteratorType pos) // NOLINT(performance-unnecessary-value-param)
+    {
+        // make sure iterator fits the current value
+        if (JSON_HEDLEY_UNLIKELY(this != pos.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+        }
+
+        IteratorType result = end();
+
+        switch (m_data.m_type)
+        {
+            case value_t::boolean:
+            case value_t::number_float:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::string:
+            case value_t::binary:
+            {
+                if (JSON_HEDLEY_UNLIKELY(!pos.m_it.primitive_iterator.is_begin()))
+                {
+                    JSON_THROW(invalid_iterator::create(205, "iterator out of range", this));
+                }
+
+                if (is_string())
+                {
+                    AllocatorType<string_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, m_data.m_value.string);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, m_data.m_value.string, 1);
+                    m_data.m_value.string = nullptr;
+                }
+                else if (is_binary())
+                {
+                    AllocatorType<binary_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, m_data.m_value.binary);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, m_data.m_value.binary, 1);
+                    m_data.m_value.binary = nullptr;
+                }
+
+                m_data.m_type = value_t::null;
+                assert_invariant();
+                break;
+            }
+
+            case value_t::object:
+            {
+                result.m_it.object_iterator = m_data.m_value.object->erase(pos.m_it.object_iterator);
+                break;
+            }
+
+            case value_t::array:
+            {
+                result.m_it.array_iterator = m_data.m_value.array->erase(pos.m_it.array_iterator);
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+
+        return result;
+    }
+
+    /// @brief remove elements given an iterator range
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    template < class IteratorType, detail::enable_if_t <
+                   std::is_same<IteratorType, typename basic_json_t::iterator>::value ||
+                   std::is_same<IteratorType, typename basic_json_t::const_iterator>::value, int > = 0 >
+    IteratorType erase(IteratorType first, IteratorType last) // NOLINT(performance-unnecessary-value-param)
+    {
+        // make sure iterator fits the current value
+        if (JSON_HEDLEY_UNLIKELY(this != first.m_object || this != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(203, "iterators do not fit current value", this));
+        }
+
+        IteratorType result = end();
+
+        switch (m_data.m_type)
+        {
+            case value_t::boolean:
+            case value_t::number_float:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::string:
+            case value_t::binary:
+            {
+                if (JSON_HEDLEY_LIKELY(!first.m_it.primitive_iterator.is_begin()
+                                       || !last.m_it.primitive_iterator.is_end()))
+                {
+                    JSON_THROW(invalid_iterator::create(204, "iterators out of range", this));
+                }
+
+                if (is_string())
+                {
+                    AllocatorType<string_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, m_data.m_value.string);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, m_data.m_value.string, 1);
+                    m_data.m_value.string = nullptr;
+                }
+                else if (is_binary())
+                {
+                    AllocatorType<binary_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, m_data.m_value.binary);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, m_data.m_value.binary, 1);
+                    m_data.m_value.binary = nullptr;
+                }
+
+                m_data.m_type = value_t::null;
+                assert_invariant();
+                break;
+            }
+
+            case value_t::object:
+            {
+                result.m_it.object_iterator = m_data.m_value.object->erase(first.m_it.object_iterator,
+                                              last.m_it.object_iterator);
+                break;
+            }
+
+            case value_t::array:
+            {
+                result.m_it.array_iterator = m_data.m_value.array->erase(first.m_it.array_iterator,
+                                             last.m_it.array_iterator);
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+
+        return result;
+    }
+
+  private:
+    template < typename KeyType, detail::enable_if_t <
+                   detail::has_erase_with_key_type<basic_json_t, KeyType>::value, int > = 0 >
+    size_type erase_internal(KeyType && key)
+    {
+        // this erase only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+
+        return m_data.m_value.object->erase(std::forward<KeyType>(key));
+    }
+
+    template < typename KeyType, detail::enable_if_t <
+                   !detail::has_erase_with_key_type<basic_json_t, KeyType>::value, int > = 0 >
+    size_type erase_internal(KeyType && key)
+    {
+        // this erase only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+
+        const auto it = m_data.m_value.object->find(std::forward<KeyType>(key));
+        if (it != m_data.m_value.object->end())
+        {
+            m_data.m_value.object->erase(it);
+            return 1;
+        }
+        return 0;
+    }
+
+  public:
+
+    /// @brief remove element from a JSON object given a key
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    size_type erase(const typename object_t::key_type& key)
+    {
+        // the indirection via erase_internal() is added to avoid making this
+        // function a template and thus de-rank it during overload resolution
+        return erase_internal(key);
+    }
+
+    /// @brief remove element from a JSON object given a key
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    size_type erase(KeyType && key)
+    {
+        return erase_internal(std::forward<KeyType>(key));
+    }
+
+    /// @brief remove element from a JSON array given an index
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    void erase(const size_type idx)
+    {
+        // this erase only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            if (JSON_HEDLEY_UNLIKELY(idx >= size()))
+            {
+                JSON_THROW(out_of_range::create(401, detail::concat("array index ", std::to_string(idx), " is out of range"), this));
+            }
+
+            m_data.m_value.array->erase(m_data.m_value.array->begin() + static_cast<difference_type>(idx));
+        }
+        else
+        {
+            JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+    }
+
+    /// @}
+
+    ////////////
+    // lookup //
+    ////////////
+
+    /// @name lookup
+    /// @{
+
+    /// @brief find an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/find/
+    iterator find(const typename object_t::key_type& key)
+    {
+        auto result = end();
+
+        if (is_object())
+        {
+            result.m_it.object_iterator = m_data.m_value.object->find(key);
+        }
+
+        return result;
+    }
+
+    /// @brief find an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/find/
+    const_iterator find(const typename object_t::key_type& key) const
+    {
+        auto result = cend();
+
+        if (is_object())
+        {
+            result.m_it.object_iterator = m_data.m_value.object->find(key);
+        }
+
+        return result;
+    }
+
+    /// @brief find an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/find/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    iterator find(KeyType && key)
+    {
+        auto result = end();
+
+        if (is_object())
+        {
+            result.m_it.object_iterator = m_data.m_value.object->find(std::forward<KeyType>(key));
+        }
+
+        return result;
+    }
+
+    /// @brief find an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/find/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    const_iterator find(KeyType && key) const
+    {
+        auto result = cend();
+
+        if (is_object())
+        {
+            result.m_it.object_iterator = m_data.m_value.object->find(std::forward<KeyType>(key));
+        }
+
+        return result;
+    }
+
+    /// @brief returns the number of occurrences of a key in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/count/
+    size_type count(const typename object_t::key_type& key) const
+    {
+        // return 0 for all nonobject types
+        return is_object() ? m_data.m_value.object->count(key) : 0;
+    }
+
+    /// @brief returns the number of occurrences of a key in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/count/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    size_type count(KeyType && key) const
+    {
+        // return 0 for all nonobject types
+        return is_object() ? m_data.m_value.object->count(std::forward<KeyType>(key)) : 0;
+    }
+
+    /// @brief check the existence of an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/contains/
+    bool contains(const typename object_t::key_type& key) const
+    {
+        return is_object() && m_data.m_value.object->find(key) != m_data.m_value.object->end();
+    }
+
+    /// @brief check the existence of an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/contains/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    bool contains(KeyType && key) const
+    {
+        return is_object() && m_data.m_value.object->find(std::forward<KeyType>(key)) != m_data.m_value.object->end();
+    }
+
+    /// @brief check the existence of an element in a JSON object given a JSON pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/contains/
+    bool contains(const json_pointer& ptr) const
+    {
+        return ptr.contains(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    bool contains(const typename ::nlohmann::json_pointer<BasicJsonType>& ptr) const
+    {
+        return ptr.contains(this);
+    }
+
+    /// @}
+
+    ///////////////
+    // iterators //
+    ///////////////
+
+    /// @name iterators
+    /// @{
+
+    /// @brief returns an iterator to the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/begin/
+    iterator begin() noexcept
+    {
+        iterator result(this);
+        result.set_begin();
+        return result;
+    }
+
+    /// @brief returns an iterator to the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/begin/
+    const_iterator begin() const noexcept
+    {
+        return cbegin();
+    }
+
+    /// @brief returns a const iterator to the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/cbegin/
+    const_iterator cbegin() const noexcept
+    {
+        const_iterator result(this);
+        result.set_begin();
+        return result;
+    }
+
+    /// @brief returns an iterator to one past the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/end/
+    iterator end() noexcept
+    {
+        iterator result(this);
+        result.set_end();
+        return result;
+    }
+
+    /// @brief returns an iterator to one past the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/end/
+    const_iterator end() const noexcept
+    {
+        return cend();
+    }
+
+    /// @brief returns an iterator to one past the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/cend/
+    const_iterator cend() const noexcept
+    {
+        const_iterator result(this);
+        result.set_end();
+        return result;
+    }
+
+    /// @brief returns an iterator to the reverse-beginning
+    /// @sa https://json.nlohmann.me/api/basic_json/rbegin/
+    reverse_iterator rbegin() noexcept
+    {
+        return reverse_iterator(end());
+    }
+
+    /// @brief returns an iterator to the reverse-beginning
+    /// @sa https://json.nlohmann.me/api/basic_json/rbegin/
+    const_reverse_iterator rbegin() const noexcept
+    {
+        return crbegin();
+    }
+
+    /// @brief returns an iterator to the reverse-end
+    /// @sa https://json.nlohmann.me/api/basic_json/rend/
+    reverse_iterator rend() noexcept
+    {
+        return reverse_iterator(begin());
+    }
+
+    /// @brief returns an iterator to the reverse-end
+    /// @sa https://json.nlohmann.me/api/basic_json/rend/
+    const_reverse_iterator rend() const noexcept
+    {
+        return crend();
+    }
+
+    /// @brief returns a const reverse iterator to the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/crbegin/
+    const_reverse_iterator crbegin() const noexcept
+    {
+        return const_reverse_iterator(cend());
+    }
+
+    /// @brief returns a const reverse iterator to one before the first
+    /// @sa https://json.nlohmann.me/api/basic_json/crend/
+    const_reverse_iterator crend() const noexcept
+    {
+        return const_reverse_iterator(cbegin());
+    }
+
+  public:
+    /// @brief wrapper to access iterator member functions in range-based for
+    /// @sa https://json.nlohmann.me/api/basic_json/items/
+    /// @deprecated This function is deprecated since 3.1.0 and will be removed in
+    ///             version 4.0.0 of the library. Please use @ref items() instead;
+    ///             that is, replace `json::iterator_wrapper(j)` with `j.items()`.
+    JSON_HEDLEY_DEPRECATED_FOR(3.1.0, items())
+    static iteration_proxy<iterator> iterator_wrapper(reference ref) noexcept
+    {
+        return ref.items();
+    }
+
+    /// @brief wrapper to access iterator member functions in range-based for
+    /// @sa https://json.nlohmann.me/api/basic_json/items/
+    /// @deprecated This function is deprecated since 3.1.0 and will be removed in
+    ///         version 4.0.0 of the library. Please use @ref items() instead;
+    ///         that is, replace `json::iterator_wrapper(j)` with `j.items()`.
+    JSON_HEDLEY_DEPRECATED_FOR(3.1.0, items())
+    static iteration_proxy<const_iterator> iterator_wrapper(const_reference ref) noexcept
+    {
+        return ref.items();
+    }
+
+    /// @brief helper to access iterator member functions in range-based for
+    /// @sa https://json.nlohmann.me/api/basic_json/items/
+    iteration_proxy<iterator> items() noexcept
+    {
+        return iteration_proxy<iterator>(*this);
+    }
+
+    /// @brief helper to access iterator member functions in range-based for
+    /// @sa https://json.nlohmann.me/api/basic_json/items/
+    iteration_proxy<const_iterator> items() const noexcept
+    {
+        return iteration_proxy<const_iterator>(*this);
+    }
+
+    /// @}
+
+    //////////////
+    // capacity //
+    //////////////
+
+    /// @name capacity
+    /// @{
+
+    /// @brief checks whether the container is empty.
+    /// @sa https://json.nlohmann.me/api/basic_json/empty/
+    bool empty() const noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::null:
+            {
+                // null values are empty
+                return true;
+            }
+
+            case value_t::array:
+            {
+                // delegate call to array_t::empty()
+                return m_data.m_value.array->empty();
+            }
+
+            case value_t::object:
+            {
+                // delegate call to object_t::empty()
+                return m_data.m_value.object->empty();
+            }
+
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                // all other types are nonempty
+                return false;
+            }
+        }
+    }
+
+    /// @brief returns the number of elements
+    /// @sa https://json.nlohmann.me/api/basic_json/size/
+    size_type size() const noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::null:
+            {
+                // null values are empty
+                return 0;
+            }
+
+            case value_t::array:
+            {
+                // delegate call to array_t::size()
+                return m_data.m_value.array->size();
+            }
+
+            case value_t::object:
+            {
+                // delegate call to object_t::size()
+                return m_data.m_value.object->size();
+            }
+
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                // all other types have size 1
+                return 1;
+            }
+        }
+    }
+
+    /// @brief returns the maximum possible number of elements
+    /// @sa https://json.nlohmann.me/api/basic_json/max_size/
+    size_type max_size() const noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::array:
+            {
+                // delegate call to array_t::max_size()
+                return m_data.m_value.array->max_size();
+            }
+
+            case value_t::object:
+            {
+                // delegate call to object_t::max_size()
+                return m_data.m_value.object->max_size();
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                // all other types have max_size() == size()
+                return size();
+            }
+        }
+    }
+
+    /// @}
+
+    ///////////////
+    // modifiers //
+    ///////////////
+
+    /// @name modifiers
+    /// @{
+
+    /// @brief clears the contents
+    /// @sa https://json.nlohmann.me/api/basic_json/clear/
+    void clear() noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::number_integer:
+            {
+                m_data.m_value.number_integer = 0;
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                m_data.m_value.number_unsigned = 0;
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                m_data.m_value.number_float = 0.0;
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                m_data.m_value.boolean = false;
+                break;
+            }
+
+            case value_t::string:
+            {
+                m_data.m_value.string->clear();
+                break;
+            }
+
+            case value_t::binary:
+            {
+                m_data.m_value.binary->clear();
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_data.m_value.array->clear();
+                break;
+            }
+
+            case value_t::object:
+            {
+                m_data.m_value.object->clear();
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                break;
+        }
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/push_back/
+    void push_back(basic_json&& val)
+    {
+        // push_back only works for null objects or arrays
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_array())))
+        {
+            JSON_THROW(type_error::create(308, detail::concat("cannot use push_back() with ", type_name()), this));
+        }
+
+        // transform null object into an array
+        if (is_null())
+        {
+            m_data.m_type = value_t::array;
+            m_data.m_value = value_t::array;
+            assert_invariant();
+        }
+
+        // add element to array (move semantics)
+        const auto old_capacity = m_data.m_value.array->capacity();
+        m_data.m_value.array->push_back(std::move(val));
+        set_parent(m_data.m_value.array->back(), old_capacity);
+        // if val is moved from, basic_json move constructor marks it null, so we do not call the destructor
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/operator+=/
+    reference operator+=(basic_json&& val)
+    {
+        push_back(std::move(val));
+        return *this;
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/push_back/
+    void push_back(const basic_json& val)
+    {
+        // push_back only works for null objects or arrays
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_array())))
+        {
+            JSON_THROW(type_error::create(308, detail::concat("cannot use push_back() with ", type_name()), this));
+        }
+
+        // transform null object into an array
+        if (is_null())
+        {
+            m_data.m_type = value_t::array;
+            m_data.m_value = value_t::array;
+            assert_invariant();
+        }
+
+        // add element to array
+        const auto old_capacity = m_data.m_value.array->capacity();
+        m_data.m_value.array->push_back(val);
+        set_parent(m_data.m_value.array->back(), old_capacity);
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/operator+=/
+    reference operator+=(const basic_json& val)
+    {
+        push_back(val);
+        return *this;
+    }
+
+    /// @brief add an object to an object
+    /// @sa https://json.nlohmann.me/api/basic_json/push_back/
+    void push_back(const typename object_t::value_type& val)
+    {
+        // push_back only works for null objects or objects
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_object())))
+        {
+            JSON_THROW(type_error::create(308, detail::concat("cannot use push_back() with ", type_name()), this));
+        }
+
+        // transform null object into an object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value = value_t::object;
+            assert_invariant();
+        }
+
+        // add element to object
+        auto res = m_data.m_value.object->insert(val);
+        set_parent(res.first->second);
+    }
+
+    /// @brief add an object to an object
+    /// @sa https://json.nlohmann.me/api/basic_json/operator+=/
+    reference operator+=(const typename object_t::value_type& val)
+    {
+        push_back(val);
+        return *this;
+    }
+
+    /// @brief add an object to an object
+    /// @sa https://json.nlohmann.me/api/basic_json/push_back/
+    void push_back(initializer_list_t init)
+    {
+        if (is_object() && init.size() == 2 && (*init.begin())->is_string())
+        {
+            basic_json&& key = init.begin()->moved_or_copied();
+            push_back(typename object_t::value_type(
+                          std::move(key.get_ref<string_t&>()), (init.begin() + 1)->moved_or_copied()));
+        }
+        else
+        {
+            push_back(basic_json(init));
+        }
+    }
+
+    /// @brief add an object to an object
+    /// @sa https://json.nlohmann.me/api/basic_json/operator+=/
+    reference operator+=(initializer_list_t init)
+    {
+        push_back(init);
+        return *this;
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/emplace_back/
+    template<class... Args>
+    reference emplace_back(Args&& ... args)
+    {
+        // emplace_back only works for null objects or arrays
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_array())))
+        {
+            JSON_THROW(type_error::create(311, detail::concat("cannot use emplace_back() with ", type_name()), this));
+        }
+
+        // transform null object into an array
+        if (is_null())
+        {
+            m_data.m_type = value_t::array;
+            m_data.m_value = value_t::array;
+            assert_invariant();
+        }
+
+        // add element to array (perfect forwarding)
+        const auto old_capacity = m_data.m_value.array->capacity();
+        m_data.m_value.array->emplace_back(std::forward<Args>(args)...);
+        return set_parent(m_data.m_value.array->back(), old_capacity);
+    }
+
+    /// @brief add an object to an object if key does not exist
+    /// @sa https://json.nlohmann.me/api/basic_json/emplace/
+    template<class... Args>
+    std::pair<iterator, bool> emplace(Args&& ... args)
+    {
+        // emplace only works for null objects or arrays
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_object())))
+        {
+            JSON_THROW(type_error::create(311, detail::concat("cannot use emplace() with ", type_name()), this));
+        }
+
+        // transform null object into an object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value = value_t::object;
+            assert_invariant();
+        }
+
+        // add element to array (perfect forwarding)
+        auto res = m_data.m_value.object->emplace(std::forward<Args>(args)...);
+        set_parent(res.first->second);
+
+        // create result iterator and set iterator to the result of emplace
+        auto it = begin();
+        it.m_it.object_iterator = res.first;
+
+        // return pair of iterator and boolean
+        return {it, res.second};
+    }
+
+    /// Helper for insertion of an iterator
+    /// @note: This uses std::distance to support GCC 4.8,
+    ///        see https://github.com/nlohmann/json/pull/1257
+    template<typename... Args>
+    iterator insert_iterator(const_iterator pos, Args&& ... args) // NOLINT(performance-unnecessary-value-param)
+    {
+        iterator result(this);
+        JSON_ASSERT(m_data.m_value.array != nullptr);
+
+        auto insert_pos = std::distance(m_data.m_value.array->begin(), pos.m_it.array_iterator);
+        m_data.m_value.array->insert(pos.m_it.array_iterator, std::forward<Args>(args)...);
+        result.m_it.array_iterator = m_data.m_value.array->begin() + insert_pos;
+
+        // This could have been written as:
+        // result.m_it.array_iterator = m_data.m_value.array->insert(pos.m_it.array_iterator, cnt, val);
+        // but the return value of insert is missing in GCC 4.8, so it is written this way instead.
+
+        set_parents();
+        return result;
+    }
+
+    /// @brief inserts element into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, const basic_json& val) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            // check if iterator pos fits to this JSON value
+            if (JSON_HEDLEY_UNLIKELY(pos.m_object != this))
+            {
+                JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+            }
+
+            // insert to array and return iterator
+            return insert_iterator(pos, val);
+        }
+
+        JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+    }
+
+    /// @brief inserts element into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, basic_json&& val) // NOLINT(performance-unnecessary-value-param)
+    {
+        return insert(pos, val);
+    }
+
+    /// @brief inserts copies of element into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, size_type cnt, const basic_json& val) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            // check if iterator pos fits to this JSON value
+            if (JSON_HEDLEY_UNLIKELY(pos.m_object != this))
+            {
+                JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+            }
+
+            // insert to array and return iterator
+            return insert_iterator(pos, cnt, val);
+        }
+
+        JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+    }
+
+    /// @brief inserts range of elements into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, const_iterator first, const_iterator last) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for arrays
+        if (JSON_HEDLEY_UNLIKELY(!is_array()))
+        {
+            JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+        }
+
+        // check if iterator pos fits to this JSON value
+        if (JSON_HEDLEY_UNLIKELY(pos.m_object != this))
+        {
+            JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+        }
+
+        // check if range iterators belong to the same JSON object
+        if (JSON_HEDLEY_UNLIKELY(first.m_object != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(210, "iterators do not fit", this));
+        }
+
+        if (JSON_HEDLEY_UNLIKELY(first.m_object == this))
+        {
+            JSON_THROW(invalid_iterator::create(211, "passed iterators may not belong to container", this));
+        }
+
+        // insert to array and return iterator
+        return insert_iterator(pos, first.m_it.array_iterator, last.m_it.array_iterator);
+    }
+
+    /// @brief inserts elements from initializer list into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, initializer_list_t ilist) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for arrays
+        if (JSON_HEDLEY_UNLIKELY(!is_array()))
+        {
+            JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+        }
+
+        // check if iterator pos fits to this JSON value
+        if (JSON_HEDLEY_UNLIKELY(pos.m_object != this))
+        {
+            JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+        }
+
+        // insert to array and return iterator
+        return insert_iterator(pos, ilist.begin(), ilist.end());
+    }
+
+    /// @brief inserts range of elements into object
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    void insert(const_iterator first, const_iterator last) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+        }
+
+        // check if range iterators belong to the same JSON object
+        if (JSON_HEDLEY_UNLIKELY(first.m_object != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(210, "iterators do not fit", this));
+        }
+
+        // passed iterators must belong to objects
+        if (JSON_HEDLEY_UNLIKELY(!first.m_object->is_object()))
+        {
+            JSON_THROW(invalid_iterator::create(202, "iterators first and last must point to objects", this));
+        }
+
+        m_data.m_value.object->insert(first.m_it.object_iterator, last.m_it.object_iterator);
+        set_parents();
+    }
+
+    /// @brief updates a JSON object from another object, overwriting existing keys
+    /// @sa https://json.nlohmann.me/api/basic_json/update/
+    void update(const_reference j, bool merge_objects = false)
+    {
+        update(j.begin(), j.end(), merge_objects);
+    }
+
+    /// @brief updates a JSON object from another object, overwriting existing keys
+    /// @sa https://json.nlohmann.me/api/basic_json/update/
+    void update(const_iterator first, const_iterator last, bool merge_objects = false) // NOLINT(performance-unnecessary-value-param)
+    {
+        // implicitly convert null value to an empty object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value.object = create<object_t>();
+            assert_invariant();
+        }
+
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(312, detail::concat("cannot use update() with ", type_name()), this));
+        }
+
+        // check if range iterators belong to the same JSON object
+        if (JSON_HEDLEY_UNLIKELY(first.m_object != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(210, "iterators do not fit", this));
+        }
+
+        // passed iterators must belong to objects
+        if (JSON_HEDLEY_UNLIKELY(!first.m_object->is_object()))
+        {
+            JSON_THROW(type_error::create(312, detail::concat("cannot use update() with ", first.m_object->type_name()), first.m_object));
+        }
+
+        for (auto it = first; it != last; ++it)
+        {
+            if (merge_objects && it.value().is_object())
+            {
+                auto it2 = m_data.m_value.object->find(it.key());
+                if (it2 != m_data.m_value.object->end())
+                {
+                    it2->second.update(it.value(), true);
+                    continue;
+                }
+            }
+            m_data.m_value.object->operator[](it.key()) = it.value();
+#if JSON_DIAGNOSTICS
+            m_data.m_value.object->operator[](it.key()).m_parent = this;
+#endif
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(reference other) noexcept (
+        std::is_nothrow_move_constructible<value_t>::value&&
+        std::is_nothrow_move_assignable<value_t>::value&&
+        std::is_nothrow_move_constructible<json_value>::value&& // NOLINT(cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+        std::is_nothrow_move_assignable<json_value>::value
+    )
+    {
+        std::swap(m_data.m_type, other.m_data.m_type);
+        std::swap(m_data.m_value, other.m_data.m_value);
+
+        set_parents();
+        other.set_parents();
+        assert_invariant();
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    friend void swap(reference left, reference right) noexcept (
+        std::is_nothrow_move_constructible<value_t>::value&&
+        std::is_nothrow_move_assignable<value_t>::value&&
+        std::is_nothrow_move_constructible<json_value>::value&& // NOLINT(cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+        std::is_nothrow_move_assignable<json_value>::value
+    )
+    {
+        left.swap(right);
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(array_t& other) // NOLINT(bugprone-exception-escape,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    {
+        // swap only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.array), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(array_t&) with ", type_name()), this));
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(object_t& other) // NOLINT(bugprone-exception-escape,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    {
+        // swap only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.object), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(object_t&) with ", type_name()), this));
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(string_t& other) // NOLINT(bugprone-exception-escape,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    {
+        // swap only works for strings
+        if (JSON_HEDLEY_LIKELY(is_string()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.string), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(string_t&) with ", type_name()), this));
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(binary_t& other) // NOLINT(bugprone-exception-escape,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    {
+        // swap only works for strings
+        if (JSON_HEDLEY_LIKELY(is_binary()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.binary), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(binary_t&) with ", type_name()), this));
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(typename binary_t::container_type& other) // NOLINT(bugprone-exception-escape)
+    {
+        // swap only works for strings
+        if (JSON_HEDLEY_LIKELY(is_binary()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.binary), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(binary_t::container_type&) with ", type_name()), this));
+        }
+    }
+
+    /// @}
+
+    //////////////////////////////////////////
+    // lexicographical comparison operators //
+    //////////////////////////////////////////
+
+    /// @name lexicographical comparison operators
+    /// @{
+
+    // note parentheses around operands are necessary; see
+    // https://github.com/nlohmann/json/issues/1530
+#define JSON_IMPLEMENT_OPERATOR(op, null_result, unordered_result, default_result)                       \
+    const auto lhs_type = lhs.type();                                                                    \
+    const auto rhs_type = rhs.type();                                                                    \
+    \
+    if (lhs_type == rhs_type) /* NOLINT(readability/braces) */                                           \
+    {                                                                                                    \
+        switch (lhs_type)                                                                                \
+        {                                                                                                \
+            case value_t::array:                                                                         \
+                return (*lhs.m_data.m_value.array) op (*rhs.m_data.m_value.array);                                     \
+                \
+            case value_t::object:                                                                        \
+                return (*lhs.m_data.m_value.object) op (*rhs.m_data.m_value.object);                                   \
+                \
+            case value_t::null:                                                                          \
+                return (null_result);                                                                    \
+                \
+            case value_t::string:                                                                        \
+                return (*lhs.m_data.m_value.string) op (*rhs.m_data.m_value.string);                                   \
+                \
+            case value_t::boolean:                                                                       \
+                return (lhs.m_data.m_value.boolean) op (rhs.m_data.m_value.boolean);                                   \
+                \
+            case value_t::number_integer:                                                                \
+                return (lhs.m_data.m_value.number_integer) op (rhs.m_data.m_value.number_integer);                     \
+                \
+            case value_t::number_unsigned:                                                               \
+                return (lhs.m_data.m_value.number_unsigned) op (rhs.m_data.m_value.number_unsigned);                   \
+                \
+            case value_t::number_float:                                                                  \
+                return (lhs.m_data.m_value.number_float) op (rhs.m_data.m_value.number_float);                         \
+                \
+            case value_t::binary:                                                                        \
+                return (*lhs.m_data.m_value.binary) op (*rhs.m_data.m_value.binary);                                   \
+                \
+            case value_t::discarded:                                                                     \
+            default:                                                                                     \
+                return (unordered_result);                                                               \
+        }                                                                                                \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_integer && rhs_type == value_t::number_float)                   \
+    {                                                                                                    \
+        return static_cast<number_float_t>(lhs.m_data.m_value.number_integer) op rhs.m_data.m_value.number_float;      \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_float && rhs_type == value_t::number_integer)                   \
+    {                                                                                                    \
+        return lhs.m_data.m_value.number_float op static_cast<number_float_t>(rhs.m_data.m_value.number_integer);      \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_unsigned && rhs_type == value_t::number_float)                  \
+    {                                                                                                    \
+        return static_cast<number_float_t>(lhs.m_data.m_value.number_unsigned) op rhs.m_data.m_value.number_float;     \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_float && rhs_type == value_t::number_unsigned)                  \
+    {                                                                                                    \
+        return lhs.m_data.m_value.number_float op static_cast<number_float_t>(rhs.m_data.m_value.number_unsigned);     \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_unsigned && rhs_type == value_t::number_integer)                \
+    {                                                                                                    \
+        return static_cast<number_integer_t>(lhs.m_data.m_value.number_unsigned) op rhs.m_data.m_value.number_integer; \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_integer && rhs_type == value_t::number_unsigned)                \
+    {                                                                                                    \
+        return lhs.m_data.m_value.number_integer op static_cast<number_integer_t>(rhs.m_data.m_value.number_unsigned); \
+    }                                                                                                    \
+    else if(compares_unordered(lhs, rhs))\
+    {\
+        return (unordered_result);\
+    }\
+    \
+    return (default_result);
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    // returns true if:
+    // - any operand is NaN and the other operand is of number type
+    // - any operand is discarded
+    // in legacy mode, discarded values are considered ordered if
+    // an operation is computed as an odd number of inverses of others
+    static bool compares_unordered(const_reference lhs, const_reference rhs, bool inverse = false) noexcept
+    {
+        if ((lhs.is_number_float() && std::isnan(lhs.m_data.m_value.number_float) && rhs.is_number())
+                || (rhs.is_number_float() && std::isnan(rhs.m_data.m_value.number_float) && lhs.is_number()))
+        {
+            return true;
+        }
+#if JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON
+        return (lhs.is_discarded() || rhs.is_discarded()) && !inverse;
+#else
+        static_cast<void>(inverse);
+        return lhs.is_discarded() || rhs.is_discarded();
+#endif
+    }
+
+  private:
+    bool compares_unordered(const_reference rhs, bool inverse = false) const noexcept
+    {
+        return compares_unordered(*this, rhs, inverse);
+    }
+
+  public:
+#if JSON_HAS_THREE_WAY_COMPARISON
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    bool operator==(const_reference rhs) const noexcept
+    {
+#ifdef __GNUC__
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wfloat-equal"
+#endif
+        const_reference lhs = *this;
+        JSON_IMPLEMENT_OPERATOR( ==, true, false, false)
+#ifdef __GNUC__
+#pragma GCC diagnostic pop
+#endif
+    }
+
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    template<typename ScalarType>
+    requires std::is_scalar_v<ScalarType>
+    bool operator==(ScalarType rhs) const noexcept
+    {
+        return *this == basic_json(rhs);
+    }
+
+    /// @brief comparison: not equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ne/
+    bool operator!=(const_reference rhs) const noexcept
+    {
+        if (compares_unordered(rhs, true))
+        {
+            return false;
+        }
+        return !operator==(rhs);
+    }
+
+    /// @brief comparison: 3-way
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_spaceship/
+    std::partial_ordering operator<=>(const_reference rhs) const noexcept // *NOPAD*
+    {
+        const_reference lhs = *this;
+        // default_result is used if we cannot compare values. In that case,
+        // we compare types.
+        JSON_IMPLEMENT_OPERATOR(<=>, // *NOPAD*
+                                std::partial_ordering::equivalent,
+                                std::partial_ordering::unordered,
+                                lhs_type <=> rhs_type) // *NOPAD*
+    }
+
+    /// @brief comparison: 3-way
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_spaceship/
+    template<typename ScalarType>
+    requires std::is_scalar_v<ScalarType>
+    std::partial_ordering operator<=>(ScalarType rhs) const noexcept // *NOPAD*
+    {
+        return *this <=> basic_json(rhs); // *NOPAD*
+    }
+
+#if JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON
+    // all operators that are computed as an odd number of inverses of others
+    // need to be overloaded to emulate the legacy comparison behavior
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, undef JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON)
+    bool operator<=(const_reference rhs) const noexcept
+    {
+        if (compares_unordered(rhs, true))
+        {
+            return false;
+        }
+        return !(rhs < *this);
+    }
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    template<typename ScalarType>
+    requires std::is_scalar_v<ScalarType>
+    bool operator<=(ScalarType rhs) const noexcept
+    {
+        return *this <= basic_json(rhs);
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, undef JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON)
+    bool operator>=(const_reference rhs) const noexcept
+    {
+        if (compares_unordered(rhs, true))
+        {
+            return false;
+        }
+        return !(*this < rhs);
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    template<typename ScalarType>
+    requires std::is_scalar_v<ScalarType>
+    bool operator>=(ScalarType rhs) const noexcept
+    {
+        return *this >= basic_json(rhs);
+    }
+#endif
+#else
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    friend bool operator==(const_reference lhs, const_reference rhs) noexcept
+    {
+#ifdef __GNUC__
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wfloat-equal"
+#endif
+        JSON_IMPLEMENT_OPERATOR( ==, true, false, false)
+#ifdef __GNUC__
+#pragma GCC diagnostic pop
+#endif
+    }
+
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator==(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs == basic_json(rhs);
+    }
+
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator==(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) == rhs;
+    }
+
+    /// @brief comparison: not equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ne/
+    friend bool operator!=(const_reference lhs, const_reference rhs) noexcept
+    {
+        if (compares_unordered(lhs, rhs, true))
+        {
+            return false;
+        }
+        return !(lhs == rhs);
+    }
+
+    /// @brief comparison: not equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ne/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator!=(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs != basic_json(rhs);
+    }
+
+    /// @brief comparison: not equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ne/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator!=(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) != rhs;
+    }
+
+    /// @brief comparison: less than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_lt/
+    friend bool operator<(const_reference lhs, const_reference rhs) noexcept
+    {
+        // default_result is used if we cannot compare values. In that case,
+        // we compare types. Note we have to call the operator explicitly,
+        // because MSVC has problems otherwise.
+        JSON_IMPLEMENT_OPERATOR( <, false, false, operator<(lhs_type, rhs_type))
+    }
+
+    /// @brief comparison: less than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_lt/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator<(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs < basic_json(rhs);
+    }
+
+    /// @brief comparison: less than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_lt/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator<(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) < rhs;
+    }
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    friend bool operator<=(const_reference lhs, const_reference rhs) noexcept
+    {
+        if (compares_unordered(lhs, rhs, true))
+        {
+            return false;
+        }
+        return !(rhs < lhs);
+    }
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator<=(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs <= basic_json(rhs);
+    }
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator<=(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) <= rhs;
+    }
+
+    /// @brief comparison: greater than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gt/
+    friend bool operator>(const_reference lhs, const_reference rhs) noexcept
+    {
+        // double inverse
+        if (compares_unordered(lhs, rhs))
+        {
+            return false;
+        }
+        return !(lhs <= rhs);
+    }
+
+    /// @brief comparison: greater than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gt/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator>(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs > basic_json(rhs);
+    }
+
+    /// @brief comparison: greater than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gt/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator>(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) > rhs;
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    friend bool operator>=(const_reference lhs, const_reference rhs) noexcept
+    {
+        if (compares_unordered(lhs, rhs, true))
+        {
+            return false;
+        }
+        return !(lhs < rhs);
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator>=(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs >= basic_json(rhs);
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator>=(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) >= rhs;
+    }
+#endif
+
+#undef JSON_IMPLEMENT_OPERATOR
+
+    /// @}
+
+    ///////////////////
+    // serialization //
+    ///////////////////
+
+    /// @name serialization
+    /// @{
+#ifndef JSON_NO_IO
+    /// @brief serialize to stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ltlt/
+    friend std::ostream& operator<<(std::ostream& o, const basic_json& j)
+    {
+        // read width member and use it as indentation parameter if nonzero
+        const bool pretty_print = o.width() > 0;
+        const auto indentation = pretty_print ? o.width() : 0;
+
+        // reset width to 0 for subsequent calls to this stream
+        o.width(0);
+
+        // do the actual serialization
+        serializer s(detail::output_adapter<char>(o), o.fill());
+        s.dump(j, pretty_print, false, static_cast<unsigned int>(indentation));
+        return o;
+    }
+
+    /// @brief serialize to stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ltlt/
+    /// @deprecated This function is deprecated since 3.0.0 and will be removed in
+    ///             version 4.0.0 of the library. Please use
+    ///             operator<<(std::ostream&, const basic_json&) instead; that is,
+    ///             replace calls like `j >> o;` with `o << j;`.
+    JSON_HEDLEY_DEPRECATED_FOR(3.0.0, operator<<(std::ostream&, const basic_json&))
+    friend std::ostream& operator>>(const basic_json& j, std::ostream& o)
+    {
+        return o << j;
+    }
+#endif  // JSON_NO_IO
+    /// @}
+
+    /////////////////////
+    // deserialization //
+    /////////////////////
+
+    /// @name deserialization
+    /// @{
+
+    /// @brief deserialize from a compatible input
+    /// @sa https://json.nlohmann.me/api/basic_json/parse/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json parse(InputType&& i,
+                            parser_callback_t cb = nullptr,
+                            const bool allow_exceptions = true,
+                            const bool ignore_comments = false)
+    {
+        basic_json result;
+        parser(detail::input_adapter(std::forward<InputType>(i)), std::move(cb), allow_exceptions, ignore_comments).parse(true, result); // cppcheck-suppress[accessMoved,accessForwarded]
+        return result;
+    }
+
+    /// @brief deserialize from a pair of character iterators
+    /// @sa https://json.nlohmann.me/api/basic_json/parse/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json parse(IteratorType first,
+                            IteratorType last,
+                            parser_callback_t cb = nullptr,
+                            const bool allow_exceptions = true,
+                            const bool ignore_comments = false)
+    {
+        basic_json result;
+        parser(detail::input_adapter(std::move(first), std::move(last)), std::move(cb), allow_exceptions, ignore_comments).parse(true, result); // cppcheck-suppress[accessMoved]
+        return result;
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, parse(ptr, ptr + len))
+    static basic_json parse(detail::span_input_adapter&& i,
+                            parser_callback_t cb = nullptr,
+                            const bool allow_exceptions = true,
+                            const bool ignore_comments = false)
+    {
+        basic_json result;
+        parser(i.get(), std::move(cb), allow_exceptions, ignore_comments).parse(true, result); // cppcheck-suppress[accessMoved]
+        return result;
+    }
+
+    /// @brief check if the input is valid JSON
+    /// @sa https://json.nlohmann.me/api/basic_json/accept/
+    template<typename InputType>
+    static bool accept(InputType&& i,
+                       const bool ignore_comments = false)
+    {
+        return parser(detail::input_adapter(std::forward<InputType>(i)), nullptr, false, ignore_comments).accept(true);
+    }
+
+    /// @brief check if the input is valid JSON
+    /// @sa https://json.nlohmann.me/api/basic_json/accept/
+    template<typename IteratorType>
+    static bool accept(IteratorType first, IteratorType last,
+                       const bool ignore_comments = false)
+    {
+        return parser(detail::input_adapter(std::move(first), std::move(last)), nullptr, false, ignore_comments).accept(true);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, accept(ptr, ptr + len))
+    static bool accept(detail::span_input_adapter&& i,
+                       const bool ignore_comments = false)
+    {
+        return parser(i.get(), nullptr, false, ignore_comments).accept(true);
+    }
+
+    /// @brief generate SAX events
+    /// @sa https://json.nlohmann.me/api/basic_json/sax_parse/
+    template <typename InputType, typename SAX>
+    JSON_HEDLEY_NON_NULL(2)
+    static bool sax_parse(InputType&& i, SAX* sax,
+                          input_format_t format = input_format_t::json,
+                          const bool strict = true,
+                          const bool ignore_comments = false)
+    {
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        return format == input_format_t::json
+               ? parser(std::move(ia), nullptr, true, ignore_comments).sax_parse(sax, strict)
+               : detail::binary_reader<basic_json, decltype(ia), SAX>(std::move(ia), format).sax_parse(format, sax, strict);
+    }
+
+    /// @brief generate SAX events
+    /// @sa https://json.nlohmann.me/api/basic_json/sax_parse/
+    template<class IteratorType, class SAX>
+    JSON_HEDLEY_NON_NULL(3)
+    static bool sax_parse(IteratorType first, IteratorType last, SAX* sax,
+                          input_format_t format = input_format_t::json,
+                          const bool strict = true,
+                          const bool ignore_comments = false)
+    {
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        return format == input_format_t::json
+               ? parser(std::move(ia), nullptr, true, ignore_comments).sax_parse(sax, strict)
+               : detail::binary_reader<basic_json, decltype(ia), SAX>(std::move(ia), format).sax_parse(format, sax, strict);
+    }
+
+    /// @brief generate SAX events
+    /// @sa https://json.nlohmann.me/api/basic_json/sax_parse/
+    /// @deprecated This function is deprecated since 3.8.0 and will be removed in
+    ///             version 4.0.0 of the library. Please use
+    ///             sax_parse(ptr, ptr + len) instead.
+    template <typename SAX>
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, sax_parse(ptr, ptr + len, ...))
+    JSON_HEDLEY_NON_NULL(2)
+    static bool sax_parse(detail::span_input_adapter&& i, SAX* sax,
+                          input_format_t format = input_format_t::json,
+                          const bool strict = true,
+                          const bool ignore_comments = false)
+    {
+        auto ia = i.get();
+        return format == input_format_t::json
+               // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+               ? parser(std::move(ia), nullptr, true, ignore_comments).sax_parse(sax, strict)
+               // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+               : detail::binary_reader<basic_json, decltype(ia), SAX>(std::move(ia), format).sax_parse(format, sax, strict);
+    }
+#ifndef JSON_NO_IO
+    /// @brief deserialize from stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gtgt/
+    /// @deprecated This stream operator is deprecated since 3.0.0 and will be removed in
+    ///             version 4.0.0 of the library. Please use
+    ///             operator>>(std::istream&, basic_json&) instead; that is,
+    ///             replace calls like `j << i;` with `i >> j;`.
+    JSON_HEDLEY_DEPRECATED_FOR(3.0.0, operator>>(std::istream&, basic_json&))
+    friend std::istream& operator<<(basic_json& j, std::istream& i)
+    {
+        return operator>>(i, j);
+    }
+
+    /// @brief deserialize from stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gtgt/
+    friend std::istream& operator>>(std::istream& i, basic_json& j)
+    {
+        parser(detail::input_adapter(i)).parse(false, j);
+        return i;
+    }
+#endif  // JSON_NO_IO
+    /// @}
+
+    ///////////////////////////
+    // convenience functions //
+    ///////////////////////////
+
+    /// @brief return the type as string
+    /// @sa https://json.nlohmann.me/api/basic_json/type_name/
+    JSON_HEDLEY_RETURNS_NON_NULL
+    const char* type_name() const noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::null:
+                return "null";
+            case value_t::object:
+                return "object";
+            case value_t::array:
+                return "array";
+            case value_t::string:
+                return "string";
+            case value_t::boolean:
+                return "boolean";
+            case value_t::binary:
+                return "binary";
+            case value_t::discarded:
+                return "discarded";
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            default:
+                return "number";
+        }
+    }
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    //////////////////////
+    // member variables //
+    //////////////////////
+
+    struct data
+    {
+        /// the type of the current element
+        value_t m_type = value_t::null;
+
+        /// the value of the current element
+        json_value m_value = {};
+
+        data(const value_t v)
+            : m_type(v), m_value(v)
+        {
+        }
+
+        data(size_type cnt, const basic_json& val)
+            : m_type(value_t::array)
+        {
+            m_value.array = create<array_t>(cnt, val);
+        }
+
+        data() noexcept = default;
+        data(data&&) noexcept = default;
+        data(const data&) noexcept = delete;
+        data& operator=(data&&) noexcept = delete;
+        data& operator=(const data&) noexcept = delete;
+
+        ~data() noexcept
+        {
+            m_value.destroy(m_type);
+        }
+    };
+
+    data m_data = {};
+
+#if JSON_DIAGNOSTICS
+    /// a pointer to a parent value (for debugging purposes)
+    basic_json* m_parent = nullptr;
+#endif
+
+#if JSON_DIAGNOSTIC_POSITIONS
+    /// the start position of the value
+    std::size_t start_position = std::string::npos;
+    /// the end position of the value
+    std::size_t end_position = std::string::npos;
+  public:
+    constexpr std::size_t start_pos() const noexcept
+    {
+        return start_position;
+    }
+
+    constexpr std::size_t end_pos() const noexcept
+    {
+        return end_position;
+    }
+#endif
+
+    //////////////////////////////////////////
+    // binary serialization/deserialization //
+    //////////////////////////////////////////
+
+    /// @name binary serialization/deserialization support
+    /// @{
+
+  public:
+    /// @brief create a CBOR serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_cbor/
+    static std::vector<std::uint8_t> to_cbor(const basic_json& j)
+    {
+        std::vector<std::uint8_t> result;
+        to_cbor(j, result);
+        return result;
+    }
+
+    /// @brief create a CBOR serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_cbor/
+    static void to_cbor(const basic_json& j, detail::output_adapter<std::uint8_t> o)
+    {
+        binary_writer<std::uint8_t>(o).write_cbor(j);
+    }
+
+    /// @brief create a CBOR serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_cbor/
+    static void to_cbor(const basic_json& j, detail::output_adapter<char> o)
+    {
+        binary_writer<char>(o).write_cbor(j);
+    }
+
+    /// @brief create a MessagePack serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_msgpack/
+    static std::vector<std::uint8_t> to_msgpack(const basic_json& j)
+    {
+        std::vector<std::uint8_t> result;
+        to_msgpack(j, result);
+        return result;
+    }
+
+    /// @brief create a MessagePack serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_msgpack/
+    static void to_msgpack(const basic_json& j, detail::output_adapter<std::uint8_t> o)
+    {
+        binary_writer<std::uint8_t>(o).write_msgpack(j);
+    }
+
+    /// @brief create a MessagePack serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_msgpack/
+    static void to_msgpack(const basic_json& j, detail::output_adapter<char> o)
+    {
+        binary_writer<char>(o).write_msgpack(j);
+    }
+
+    /// @brief create a UBJSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_ubjson/
+    static std::vector<std::uint8_t> to_ubjson(const basic_json& j,
+            const bool use_size = false,
+            const bool use_type = false)
+    {
+        std::vector<std::uint8_t> result;
+        to_ubjson(j, result, use_size, use_type);
+        return result;
+    }
+
+    /// @brief create a UBJSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_ubjson/
+    static void to_ubjson(const basic_json& j, detail::output_adapter<std::uint8_t> o,
+                          const bool use_size = false, const bool use_type = false)
+    {
+        binary_writer<std::uint8_t>(o).write_ubjson(j, use_size, use_type);
+    }
+
+    /// @brief create a UBJSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_ubjson/
+    static void to_ubjson(const basic_json& j, detail::output_adapter<char> o,
+                          const bool use_size = false, const bool use_type = false)
+    {
+        binary_writer<char>(o).write_ubjson(j, use_size, use_type);
+    }
+
+    /// @brief create a BJData serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bjdata/
+    static std::vector<std::uint8_t> to_bjdata(const basic_json& j,
+            const bool use_size = false,
+            const bool use_type = false,
+            const bjdata_version_t version = bjdata_version_t::draft2)
+    {
+        std::vector<std::uint8_t> result;
+        to_bjdata(j, result, use_size, use_type, version);
+        return result;
+    }
+
+    /// @brief create a BJData serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bjdata/
+    static void to_bjdata(const basic_json& j, detail::output_adapter<std::uint8_t> o,
+                          const bool use_size = false, const bool use_type = false,
+                          const bjdata_version_t version = bjdata_version_t::draft2)
+    {
+        binary_writer<std::uint8_t>(o).write_ubjson(j, use_size, use_type, true, true, version);
+    }
+
+    /// @brief create a BJData serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bjdata/
+    static void to_bjdata(const basic_json& j, detail::output_adapter<char> o,
+                          const bool use_size = false, const bool use_type = false,
+                          const bjdata_version_t version = bjdata_version_t::draft2)
+    {
+        binary_writer<char>(o).write_ubjson(j, use_size, use_type, true, true, version);
+    }
+
+    /// @brief create a BSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bson/
+    static std::vector<std::uint8_t> to_bson(const basic_json& j)
+    {
+        std::vector<std::uint8_t> result;
+        to_bson(j, result);
+        return result;
+    }
+
+    /// @brief create a BSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bson/
+    static void to_bson(const basic_json& j, detail::output_adapter<std::uint8_t> o)
+    {
+        binary_writer<std::uint8_t>(o).write_bson(j);
+    }
+
+    /// @brief create a BSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bson/
+    static void to_bson(const basic_json& j, detail::output_adapter<char> o)
+    {
+        binary_writer<char>(o).write_bson(j);
+    }
+
+    /// @brief create a JSON value from an input in CBOR format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_cbor/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_cbor(InputType&& i,
+                                const bool strict = true,
+                                const bool allow_exceptions = true,
+                                const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::cbor).sax_parse(input_format_t::cbor, &sdp, strict, tag_handler); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in CBOR format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_cbor/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_cbor(IteratorType first, IteratorType last,
+                                const bool strict = true,
+                                const bool allow_exceptions = true,
+                                const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::cbor).sax_parse(input_format_t::cbor, &sdp, strict, tag_handler); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    template<typename T>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_cbor(ptr, ptr + len))
+    static basic_json from_cbor(const T* ptr, std::size_t len,
+                                const bool strict = true,
+                                const bool allow_exceptions = true,
+                                const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        return from_cbor(ptr, ptr + len, strict, allow_exceptions, tag_handler);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_cbor(ptr, ptr + len))
+    static basic_json from_cbor(detail::span_input_adapter&& i,
+                                const bool strict = true,
+                                const bool allow_exceptions = true,
+                                const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        basic_json result;
+        auto ia = i.get();
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::cbor).sax_parse(input_format_t::cbor, &sdp, strict, tag_handler); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in MessagePack format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_msgpack/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_msgpack(InputType&& i,
+                                   const bool strict = true,
+                                   const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::msgpack).sax_parse(input_format_t::msgpack, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in MessagePack format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_msgpack/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_msgpack(IteratorType first, IteratorType last,
+                                   const bool strict = true,
+                                   const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::msgpack).sax_parse(input_format_t::msgpack, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    template<typename T>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_msgpack(ptr, ptr + len))
+    static basic_json from_msgpack(const T* ptr, std::size_t len,
+                                   const bool strict = true,
+                                   const bool allow_exceptions = true)
+    {
+        return from_msgpack(ptr, ptr + len, strict, allow_exceptions);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_msgpack(ptr, ptr + len))
+    static basic_json from_msgpack(detail::span_input_adapter&& i,
+                                   const bool strict = true,
+                                   const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = i.get();
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::msgpack).sax_parse(input_format_t::msgpack, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in UBJSON format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_ubjson/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_ubjson(InputType&& i,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::ubjson).sax_parse(input_format_t::ubjson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in UBJSON format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_ubjson/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_ubjson(IteratorType first, IteratorType last,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::ubjson).sax_parse(input_format_t::ubjson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    template<typename T>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_ubjson(ptr, ptr + len))
+    static basic_json from_ubjson(const T* ptr, std::size_t len,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        return from_ubjson(ptr, ptr + len, strict, allow_exceptions);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_ubjson(ptr, ptr + len))
+    static basic_json from_ubjson(detail::span_input_adapter&& i,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = i.get();
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::ubjson).sax_parse(input_format_t::ubjson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in BJData format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_bjdata/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_bjdata(InputType&& i,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bjdata).sax_parse(input_format_t::bjdata, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in BJData format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_bjdata/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_bjdata(IteratorType first, IteratorType last,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bjdata).sax_parse(input_format_t::bjdata, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in BSON format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_bson/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_bson(InputType&& i,
+                                const bool strict = true,
+                                const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bson).sax_parse(input_format_t::bson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in BSON format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_bson/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_bson(IteratorType first, IteratorType last,
+                                const bool strict = true,
+                                const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bson).sax_parse(input_format_t::bson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    template<typename T>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_bson(ptr, ptr + len))
+    static basic_json from_bson(const T* ptr, std::size_t len,
+                                const bool strict = true,
+                                const bool allow_exceptions = true)
+    {
+        return from_bson(ptr, ptr + len, strict, allow_exceptions);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_bson(ptr, ptr + len))
+    static basic_json from_bson(detail::span_input_adapter&& i,
+                                const bool strict = true,
+                                const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = i.get();
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bson).sax_parse(input_format_t::bson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+    /// @}
+
+    //////////////////////////
+    // JSON Pointer support //
+    //////////////////////////
+
+    /// @name JSON Pointer functions
+    /// @{
+
+    /// @brief access specified element via JSON Pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    reference operator[](const json_pointer& ptr)
+    {
+        return ptr.get_unchecked(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    reference operator[](const ::nlohmann::json_pointer<BasicJsonType>& ptr)
+    {
+        return ptr.get_unchecked(this);
+    }
+
+    /// @brief access specified element via JSON Pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    const_reference operator[](const json_pointer& ptr) const
+    {
+        return ptr.get_unchecked(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    const_reference operator[](const ::nlohmann::json_pointer<BasicJsonType>& ptr) const
+    {
+        return ptr.get_unchecked(this);
+    }
+
+    /// @brief access specified element via JSON Pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    reference at(const json_pointer& ptr)
+    {
+        return ptr.get_checked(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    reference at(const ::nlohmann::json_pointer<BasicJsonType>& ptr)
+    {
+        return ptr.get_checked(this);
+    }
+
+    /// @brief access specified element via JSON Pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    const_reference at(const json_pointer& ptr) const
+    {
+        return ptr.get_checked(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    const_reference at(const ::nlohmann::json_pointer<BasicJsonType>& ptr) const
+    {
+        return ptr.get_checked(this);
+    }
+
+    /// @brief return flattened JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/flatten/
+    basic_json flatten() const
+    {
+        basic_json result(value_t::object);
+        json_pointer::flatten("", *this, result);
+        return result;
+    }
+
+    /// @brief unflatten a previously flattened JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/unflatten/
+    basic_json unflatten() const
+    {
+        return json_pointer::unflatten(*this);
+    }
+
+    /// @}
+
+    //////////////////////////
+    // JSON Patch functions //
+    //////////////////////////
+
+    /// @name JSON Patch functions
+    /// @{
+
+    /// @brief applies a JSON patch in-place without copying the object
+    /// @sa https://json.nlohmann.me/api/basic_json/patch/
+    void patch_inplace(const basic_json& json_patch)
+    {
+        basic_json& result = *this;
+        // the valid JSON Patch operations
+        enum class patch_operations {add, remove, replace, move, copy, test, invalid};
+
+        const auto get_op = [](const string_t& op)
+        {
+            if (op == "add")
+            {
+                return patch_operations::add;
+            }
+            if (op == "remove")
+            {
+                return patch_operations::remove;
+            }
+            if (op == "replace")
+            {
+                return patch_operations::replace;
+            }
+            if (op == "move")
+            {
+                return patch_operations::move;
+            }
+            if (op == "copy")
+            {
+                return patch_operations::copy;
+            }
+            if (op == "test")
+            {
+                return patch_operations::test;
+            }
+
+            return patch_operations::invalid;
+        };
+
+        // wrapper for "add" operation; add value at ptr
+        const auto operation_add = [&result](json_pointer & ptr, const basic_json & val)
+        {
+            // adding to the root of the target document means replacing it
+            if (ptr.empty())
+            {
+                result = val;
+                return;
+            }
+
+            // make sure the top element of the pointer exists
+            json_pointer const top_pointer = ptr.top();
+            if (top_pointer != ptr)
+            {
+                result.at(top_pointer);
+            }
+
+            // get reference to parent of JSON pointer ptr
+            const auto last_path = ptr.back();
+            ptr.pop_back();
+            // parent must exist when performing patch add per RFC6902 specs
+            basic_json& parent = result.at(ptr);
+
+            switch (parent.m_data.m_type)
+            {
+                case value_t::null:
+                case value_t::object:
+                {
+                    // use operator[] to add value
+                    parent[last_path] = val;
+                    break;
+                }
+
+                case value_t::array:
+                {
+                    if (last_path == "-")
+                    {
+                        // special case: append to back
+                        parent.push_back(val);
+                    }
+                    else
+                    {
+                        const auto idx = json_pointer::template array_index<basic_json_t>(last_path);
+                        if (JSON_HEDLEY_UNLIKELY(idx > parent.size()))
+                        {
+                            // avoid undefined behavior
+                            JSON_THROW(out_of_range::create(401, detail::concat("array index ", std::to_string(idx), " is out of range"), &parent));
+                        }
+
+                        // default case: insert add offset
+                        parent.insert(parent.begin() + static_cast<difference_type>(idx), val);
+                    }
+                    break;
+                }
+
+                // if there exists a parent it cannot be primitive
+                case value_t::string: // LCOV_EXCL_LINE
+                case value_t::boolean: // LCOV_EXCL_LINE
+                case value_t::number_integer: // LCOV_EXCL_LINE
+                case value_t::number_unsigned: // LCOV_EXCL_LINE
+                case value_t::number_float: // LCOV_EXCL_LINE
+                case value_t::binary: // LCOV_EXCL_LINE
+                case value_t::discarded: // LCOV_EXCL_LINE
+                default:            // LCOV_EXCL_LINE
+                    JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+            }
+        };
+
+        // wrapper for "remove" operation; remove value at ptr
+        const auto operation_remove = [this, & result](json_pointer & ptr)
+        {
+            // get reference to parent of JSON pointer ptr
+            const auto last_path = ptr.back();
+            ptr.pop_back();
+            basic_json& parent = result.at(ptr);
+
+            // remove child
+            if (parent.is_object())
+            {
+                // perform range check
+                auto it = parent.find(last_path);
+                if (JSON_HEDLEY_LIKELY(it != parent.end()))
+                {
+                    parent.erase(it);
+                }
+                else
+                {
+                    JSON_THROW(out_of_range::create(403, detail::concat("key '", last_path, "' not found"), this));
+                }
+            }
+            else if (parent.is_array())
+            {
+                // note erase performs range check
+                parent.erase(json_pointer::template array_index<basic_json_t>(last_path));
+            }
+        };
+
+        // type check: top level value must be an array
+        if (JSON_HEDLEY_UNLIKELY(!json_patch.is_array()))
+        {
+            JSON_THROW(parse_error::create(104, 0, "JSON patch must be an array of objects", &json_patch));
+        }
+
+        // iterate and apply the operations
+        for (const auto& val : json_patch)
+        {
+            // wrapper to get a value for an operation
+            const auto get_value = [&val](const string_t& op,
+                                          const string_t& member,
+                                          bool string_type) -> basic_json &
+            {
+                // find value
+                auto it = val.m_data.m_value.object->find(member);
+
+                // context-sensitive error message
+                const auto error_msg = (op == "op") ? "operation" : detail::concat("operation '", op, '\''); // NOLINT(bugprone-unused-local-non-trivial-variable)
+
+                // check if desired value is present
+                if (JSON_HEDLEY_UNLIKELY(it == val.m_data.m_value.object->end()))
+                {
+                    // NOLINTNEXTLINE(performance-inefficient-string-concatenation)
+                    JSON_THROW(parse_error::create(105, 0, detail::concat(error_msg, " must have member '", member, "'"), &val));
+                }
+
+                // check if result is of type string
+                if (JSON_HEDLEY_UNLIKELY(string_type && !it->second.is_string()))
+                {
+                    // NOLINTNEXTLINE(performance-inefficient-string-concatenation)
+                    JSON_THROW(parse_error::create(105, 0, detail::concat(error_msg, " must have string member '", member, "'"), &val));
+                }
+
+                // no error: return value
+                return it->second;
+            };
+
+            // type check: every element of the array must be an object
+            if (JSON_HEDLEY_UNLIKELY(!val.is_object()))
+            {
+                JSON_THROW(parse_error::create(104, 0, "JSON patch must be an array of objects", &val));
+            }
+
+            // collect mandatory members
+            const auto op = get_value("op", "op", true).template get<string_t>();
+            const auto path = get_value(op, "path", true).template get<string_t>();
+            json_pointer ptr(path);
+
+            switch (get_op(op))
+            {
+                case patch_operations::add:
+                {
+                    operation_add(ptr, get_value("add", "value", false));
+                    break;
+                }
+
+                case patch_operations::remove:
+                {
+                    operation_remove(ptr);
+                    break;
+                }
+
+                case patch_operations::replace:
+                {
+                    // the "path" location must exist - use at()
+                    result.at(ptr) = get_value("replace", "value", false);
+                    break;
+                }
+
+                case patch_operations::move:
+                {
+                    const auto from_path = get_value("move", "from", true).template get<string_t>();
+                    json_pointer from_ptr(from_path);
+
+                    // the "from" location must exist - use at()
+                    basic_json const v = result.at(from_ptr);
+
+                    // The move operation is functionally identical to a
+                    // "remove" operation on the "from" location, followed
+                    // immediately by an "add" operation at the target
+                    // location with the value that was just removed.
+                    operation_remove(from_ptr);
+                    operation_add(ptr, v);
+                    break;
+                }
+
+                case patch_operations::copy:
+                {
+                    const auto from_path = get_value("copy", "from", true).template get<string_t>();
+                    const json_pointer from_ptr(from_path);
+
+                    // the "from" location must exist - use at()
+                    basic_json const v = result.at(from_ptr);
+
+                    // The copy is functionally identical to an "add"
+                    // operation at the target location using the value
+                    // specified in the "from" member.
+                    operation_add(ptr, v);
+                    break;
+                }
+
+                case patch_operations::test:
+                {
+                    bool success = false;
+                    JSON_TRY
+                    {
+                        // check if "value" matches the one at "path"
+                        // the "path" location must exist - use at()
+                        success = (result.at(ptr) == get_value("test", "value", false));
+                    }
+                    JSON_INTERNAL_CATCH (out_of_range&)
+                    {
+                        // ignore out of range errors: success remains false
+                    }
+
+                    // throw an exception if test fails
+                    if (JSON_HEDLEY_UNLIKELY(!success))
+                    {
+                        JSON_THROW(other_error::create(501, detail::concat("unsuccessful: ", val.dump()), &val));
+                    }
+
+                    break;
+                }
+
+                case patch_operations::invalid:
+                default:
+                {
+                    // op must be "add", "remove", "replace", "move", "copy", or
+                    // "test"
+                    JSON_THROW(parse_error::create(105, 0, detail::concat("operation value '", op, "' is invalid"), &val));
+                }
+            }
+        }
+    }
+
+    /// @brief applies a JSON patch to a copy of the current object
+    /// @sa https://json.nlohmann.me/api/basic_json/patch/
+    basic_json patch(const basic_json& json_patch) const
+    {
+        basic_json result = *this;
+        result.patch_inplace(json_patch);
+        return result;
+    }
+
+    /// @brief creates a diff as a JSON patch
+    /// @sa https://json.nlohmann.me/api/basic_json/diff/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json diff(const basic_json& source, const basic_json& target,
+                           const string_t& path = "")
+    {
+        // the patch
+        basic_json result(value_t::array);
+
+        // if the values are the same, return empty patch
+        if (source == target)
+        {
+            return result;
+        }
+
+        if (source.type() != target.type())
+        {
+            // different types: replace value
+            result.push_back(
+            {
+                {"op", "replace"}, {"path", path}, {"value", target}
+            });
+            return result;
+        }
+
+        switch (source.type())
+        {
+            case value_t::array:
+            {
+                // first pass: traverse common elements
+                std::size_t i = 0;
+                while (i < source.size() && i < target.size())
+                {
+                    // recursive call to compare array values at index i
+                    auto temp_diff = diff(source[i], target[i], detail::concat<string_t>(path, '/', detail::to_string<string_t>(i)));
+                    result.insert(result.end(), temp_diff.begin(), temp_diff.end());
+                    ++i;
+                }
+
+                // We now reached the end of at least one array
+                // in a second pass, traverse the remaining elements
+
+                // remove my remaining elements
+                const auto end_index = static_cast<difference_type>(result.size());
+                while (i < source.size())
+                {
+                    // add operations in reverse order to avoid invalid
+                    // indices
+                    result.insert(result.begin() + end_index, object(
+                    {
+                        {"op", "remove"},
+                        {"path", detail::concat<string_t>(path, '/', detail::to_string<string_t>(i))}
+                    }));
+                    ++i;
+                }
+
+                // add other remaining elements
+                while (i < target.size())
+                {
+                    result.push_back(
+                    {
+                        {"op", "add"},
+                        {"path", detail::concat<string_t>(path, "/-")},
+                        {"value", target[i]}
+                    });
+                    ++i;
+                }
+
+                break;
+            }
+
+            case value_t::object:
+            {
+                // first pass: traverse this object's elements
+                for (auto it = source.cbegin(); it != source.cend(); ++it)
+                {
+                    // escape the key name to be used in a JSON patch
+                    const auto path_key = detail::concat<string_t>(path, '/', detail::escape(it.key()));
+
+                    if (target.find(it.key()) != target.end())
+                    {
+                        // recursive call to compare object values at key it
+                        auto temp_diff = diff(it.value(), target[it.key()], path_key);
+                        result.insert(result.end(), temp_diff.begin(), temp_diff.end());
+                    }
+                    else
+                    {
+                        // found a key that is not in o -> remove it
+                        result.push_back(object(
+                        {
+                            {"op", "remove"}, {"path", path_key}
+                        }));
+                    }
+                }
+
+                // second pass: traverse other object's elements
+                for (auto it = target.cbegin(); it != target.cend(); ++it)
+                {
+                    if (source.find(it.key()) == source.end())
+                    {
+                        // found a key that is not in this -> add it
+                        const auto path_key = detail::concat<string_t>(path, '/', detail::escape(it.key()));
+                        result.push_back(
+                        {
+                            {"op", "add"}, {"path", path_key},
+                            {"value", it.value()}
+                        });
+                    }
+                }
+
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                // both primitive type: replace value
+                result.push_back(
+                {
+                    {"op", "replace"}, {"path", path}, {"value", target}
+                });
+                break;
+            }
+        }
+
+        return result;
+    }
+    /// @}
+
+    ////////////////////////////////
+    // JSON Merge Patch functions //
+    ////////////////////////////////
+
+    /// @name JSON Merge Patch functions
+    /// @{
+
+    /// @brief applies a JSON Merge Patch
+    /// @sa https://json.nlohmann.me/api/basic_json/merge_patch/
+    void merge_patch(const basic_json& apply_patch)
+    {
+        if (apply_patch.is_object())
+        {
+            if (!is_object())
+            {
+                *this = object();
+            }
+            for (auto it = apply_patch.begin(); it != apply_patch.end(); ++it)
+            {
+                if (it.value().is_null())
+                {
+                    erase(it.key());
+                }
+                else
+                {
+                    operator[](it.key()).merge_patch(it.value());
+                }
+            }
+        }
+        else
+        {
+            *this = apply_patch;
+        }
+    }
+
+    /// @}
+};
+
+/// @brief user-defined to_string function for JSON values
+/// @sa https://json.nlohmann.me/api/basic_json/to_string/
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+std::string to_string(const NLOHMANN_BASIC_JSON_TPL& j)
+{
+    return j.dump();
+}
+
+inline namespace literals
+{
+inline namespace json_literals
+{
+
+/// @brief user-defined string literal for JSON values
+/// @sa https://json.nlohmann.me/api/basic_json/operator_literal_json/
+JSON_HEDLEY_NON_NULL(1)
+#if !defined(JSON_HEDLEY_GCC_VERSION) || JSON_HEDLEY_GCC_VERSION_CHECK(4,9,0)
+    inline nlohmann::json operator ""_json(const char* s, std::size_t n)
+#else
+    inline nlohmann::json operator "" _json(const char* s, std::size_t n)
+#endif
+{
+    return nlohmann::json::parse(s, s + n);
+}
+
+/// @brief user-defined string literal for JSON pointer
+/// @sa https://json.nlohmann.me/api/basic_json/operator_literal_json_pointer/
+JSON_HEDLEY_NON_NULL(1)
+#if !defined(JSON_HEDLEY_GCC_VERSION) || JSON_HEDLEY_GCC_VERSION_CHECK(4,9,0)
+    inline nlohmann::json::json_pointer operator ""_json_pointer(const char* s, std::size_t n)
+#else
+    inline nlohmann::json::json_pointer operator "" _json_pointer(const char* s, std::size_t n)
+#endif
+{
+    return nlohmann::json::json_pointer(std::string(s, n));
+}
+
+}  // namespace json_literals
+}  // namespace literals
+NLOHMANN_JSON_NAMESPACE_END
+
+///////////////////////
+// nonmember support //
+///////////////////////
+
+namespace std // NOLINT(cert-dcl58-cpp)
+{
+
+/// @brief hash value for JSON objects
+/// @sa https://json.nlohmann.me/api/basic_json/std_hash/
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+struct hash<nlohmann::NLOHMANN_BASIC_JSON_TPL> // NOLINT(cert-dcl58-cpp)
+{
+    std::size_t operator()(const nlohmann::NLOHMANN_BASIC_JSON_TPL& j) const
+    {
+        return nlohmann::detail::hash(j);
+    }
+};
+
+// specialization for std::less<value_t>
+template<>
+struct less< ::nlohmann::detail::value_t> // do not remove the space after '<', see https://github.com/nlohmann/json/pull/679
+{
+    /*!
+    @brief compare two value_t enum values
+    @since version 3.0.0
+    */
+    bool operator()(::nlohmann::detail::value_t lhs,
+                    ::nlohmann::detail::value_t rhs) const noexcept
+    {
+#if JSON_HAS_THREE_WAY_COMPARISON
+        return std::is_lt(lhs <=> rhs); // *NOPAD*
+#else
+        return ::nlohmann::detail::operator<(lhs, rhs);
+#endif
+    }
+};
+
+// C++20 prohibit function specialization in the std namespace.
+#ifndef JSON_HAS_CPP_20
+
+/// @brief exchanges the values of two JSON objects
+/// @sa https://json.nlohmann.me/api/basic_json/std_swap/
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+inline void swap(nlohmann::NLOHMANN_BASIC_JSON_TPL& j1, nlohmann::NLOHMANN_BASIC_JSON_TPL& j2) noexcept(  // NOLINT(readability-inconsistent-declaration-parameter-name, cert-dcl58-cpp)
+    is_nothrow_move_constructible<nlohmann::NLOHMANN_BASIC_JSON_TPL>::value&&                          // NOLINT(misc-redundant-expression,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    is_nothrow_move_assignable<nlohmann::NLOHMANN_BASIC_JSON_TPL>::value)
+{
+    j1.swap(j2);
+}
+
+#endif
+
+}  // namespace std
+
+#if JSON_USE_GLOBAL_UDLS
+    #if !defined(JSON_HEDLEY_GCC_VERSION) || JSON_HEDLEY_GCC_VERSION_CHECK(4,9,0)
+        using nlohmann::literals::json_literals::operator ""_json; // NOLINT(misc-unused-using-decls,google-global-names-in-headers)
+        using nlohmann::literals::json_literals::operator ""_json_pointer; //NOLINT(misc-unused-using-decls,google-global-names-in-headers)
+    #else
+        using nlohmann::literals::json_literals::operator "" _json; // NOLINT(misc-unused-using-decls,google-global-names-in-headers)
+        using nlohmann::literals::json_literals::operator "" _json_pointer; //NOLINT(misc-unused-using-decls,google-global-names-in-headers)
+    #endif
+#endif
+
+// #include <nlohmann/detail/macro_unscope.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+// restore clang diagnostic settings
+#if defined(__clang__)
+    #pragma clang diagnostic pop
+#endif
+
+// clean up
+#undef JSON_ASSERT
+#undef JSON_INTERNAL_CATCH
+#undef JSON_THROW
+#undef JSON_PRIVATE_UNLESS_TESTED
+#undef NLOHMANN_BASIC_JSON_TPL_DECLARATION
+#undef NLOHMANN_BASIC_JSON_TPL
+#undef JSON_EXPLICIT
+#undef NLOHMANN_CAN_CALL_STD_FUNC_IMPL
+#undef JSON_INLINE_VARIABLE
+#undef JSON_NO_UNIQUE_ADDRESS
+#undef JSON_DISABLE_ENUM_SERIALIZATION
+#undef JSON_USE_GLOBAL_UDLS
+
+#ifndef JSON_TEST_KEEP_MACROS
+    #undef JSON_CATCH
+    #undef JSON_TRY
+    #undef JSON_HAS_CPP_11
+    #undef JSON_HAS_CPP_14
+    #undef JSON_HAS_CPP_17
+    #undef JSON_HAS_CPP_20
+    #undef JSON_HAS_FILESYSTEM
+    #undef JSON_HAS_EXPERIMENTAL_FILESYSTEM
+    #undef JSON_HAS_THREE_WAY_COMPARISON
+    #undef JSON_HAS_RANGES
+    #undef JSON_HAS_STATIC_RTTI
+    #undef JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON
+#endif
+
+// #include <nlohmann/thirdparty/hedley/hedley_undef.hpp>
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+
+
+#undef JSON_HEDLEY_ALWAYS_INLINE
+#undef JSON_HEDLEY_ARM_VERSION
+#undef JSON_HEDLEY_ARM_VERSION_CHECK
+#undef JSON_HEDLEY_ARRAY_PARAM
+#undef JSON_HEDLEY_ASSUME
+#undef JSON_HEDLEY_BEGIN_C_DECLS
+#undef JSON_HEDLEY_CLANG_HAS_ATTRIBUTE
+#undef JSON_HEDLEY_CLANG_HAS_BUILTIN
+#undef JSON_HEDLEY_CLANG_HAS_CPP_ATTRIBUTE
+#undef JSON_HEDLEY_CLANG_HAS_DECLSPEC_DECLSPEC_ATTRIBUTE
+#undef JSON_HEDLEY_CLANG_HAS_EXTENSION
+#undef JSON_HEDLEY_CLANG_HAS_FEATURE
+#undef JSON_HEDLEY_CLANG_HAS_WARNING
+#undef JSON_HEDLEY_COMPCERT_VERSION
+#undef JSON_HEDLEY_COMPCERT_VERSION_CHECK
+#undef JSON_HEDLEY_CONCAT
+#undef JSON_HEDLEY_CONCAT3
+#undef JSON_HEDLEY_CONCAT3_EX
+#undef JSON_HEDLEY_CONCAT_EX
+#undef JSON_HEDLEY_CONST
+#undef JSON_HEDLEY_CONSTEXPR
+#undef JSON_HEDLEY_CONST_CAST
+#undef JSON_HEDLEY_CPP_CAST
+#undef JSON_HEDLEY_CRAY_VERSION
+#undef JSON_HEDLEY_CRAY_VERSION_CHECK
+#undef JSON_HEDLEY_C_DECL
+#undef JSON_HEDLEY_DEPRECATED
+#undef JSON_HEDLEY_DEPRECATED_FOR
+#undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_CAST_QUAL
+#undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_CPP98_COMPAT_WRAP_
+#undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_DEPRECATED
+#undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_CPP_ATTRIBUTES
+#undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNKNOWN_PRAGMAS
+#undef JSON_HEDLEY_DIAGNOSTIC_DISABLE_UNUSED_FUNCTION
+#undef JSON_HEDLEY_DIAGNOSTIC_POP
+#undef JSON_HEDLEY_DIAGNOSTIC_PUSH
+#undef JSON_HEDLEY_DMC_VERSION
+#undef JSON_HEDLEY_DMC_VERSION_CHECK
+#undef JSON_HEDLEY_EMPTY_BASES
+#undef JSON_HEDLEY_EMSCRIPTEN_VERSION
+#undef JSON_HEDLEY_EMSCRIPTEN_VERSION_CHECK
+#undef JSON_HEDLEY_END_C_DECLS
+#undef JSON_HEDLEY_FLAGS
+#undef JSON_HEDLEY_FLAGS_CAST
+#undef JSON_HEDLEY_GCC_HAS_ATTRIBUTE
+#undef JSON_HEDLEY_GCC_HAS_BUILTIN
+#undef JSON_HEDLEY_GCC_HAS_CPP_ATTRIBUTE
+#undef JSON_HEDLEY_GCC_HAS_DECLSPEC_ATTRIBUTE
+#undef JSON_HEDLEY_GCC_HAS_EXTENSION
+#undef JSON_HEDLEY_GCC_HAS_FEATURE
+#undef JSON_HEDLEY_GCC_HAS_WARNING
+#undef JSON_HEDLEY_GCC_NOT_CLANG_VERSION_CHECK
+#undef JSON_HEDLEY_GCC_VERSION
+#undef JSON_HEDLEY_GCC_VERSION_CHECK
+#undef JSON_HEDLEY_GNUC_HAS_ATTRIBUTE
+#undef JSON_HEDLEY_GNUC_HAS_BUILTIN
+#undef JSON_HEDLEY_GNUC_HAS_CPP_ATTRIBUTE
+#undef JSON_HEDLEY_GNUC_HAS_DECLSPEC_ATTRIBUTE
+#undef JSON_HEDLEY_GNUC_HAS_EXTENSION
+#undef JSON_HEDLEY_GNUC_HAS_FEATURE
+#undef JSON_HEDLEY_GNUC_HAS_WARNING
+#undef JSON_HEDLEY_GNUC_VERSION
+#undef JSON_HEDLEY_GNUC_VERSION_CHECK
+#undef JSON_HEDLEY_HAS_ATTRIBUTE
+#undef JSON_HEDLEY_HAS_BUILTIN
+#undef JSON_HEDLEY_HAS_CPP_ATTRIBUTE
+#undef JSON_HEDLEY_HAS_CPP_ATTRIBUTE_NS
+#undef JSON_HEDLEY_HAS_DECLSPEC_ATTRIBUTE
+#undef JSON_HEDLEY_HAS_EXTENSION
+#undef JSON_HEDLEY_HAS_FEATURE
+#undef JSON_HEDLEY_HAS_WARNING
+#undef JSON_HEDLEY_IAR_VERSION
+#undef JSON_HEDLEY_IAR_VERSION_CHECK
+#undef JSON_HEDLEY_IBM_VERSION
+#undef JSON_HEDLEY_IBM_VERSION_CHECK
+#undef JSON_HEDLEY_IMPORT
+#undef JSON_HEDLEY_INLINE
+#undef JSON_HEDLEY_INTEL_CL_VERSION
+#undef JSON_HEDLEY_INTEL_CL_VERSION_CHECK
+#undef JSON_HEDLEY_INTEL_VERSION
+#undef JSON_HEDLEY_INTEL_VERSION_CHECK
+#undef JSON_HEDLEY_IS_CONSTANT
+#undef JSON_HEDLEY_IS_CONSTEXPR_
+#undef JSON_HEDLEY_LIKELY
+#undef JSON_HEDLEY_MALLOC
+#undef JSON_HEDLEY_MCST_LCC_VERSION
+#undef JSON_HEDLEY_MCST_LCC_VERSION_CHECK
+#undef JSON_HEDLEY_MESSAGE
+#undef JSON_HEDLEY_MSVC_VERSION
+#undef JSON_HEDLEY_MSVC_VERSION_CHECK
+#undef JSON_HEDLEY_NEVER_INLINE
+#undef JSON_HEDLEY_NON_NULL
+#undef JSON_HEDLEY_NO_ESCAPE
+#undef JSON_HEDLEY_NO_RETURN
+#undef JSON_HEDLEY_NO_THROW
+#undef JSON_HEDLEY_NULL
+#undef JSON_HEDLEY_PELLES_VERSION
+#undef JSON_HEDLEY_PELLES_VERSION_CHECK
+#undef JSON_HEDLEY_PGI_VERSION
+#undef JSON_HEDLEY_PGI_VERSION_CHECK
+#undef JSON_HEDLEY_PREDICT
+#undef JSON_HEDLEY_PRINTF_FORMAT
+#undef JSON_HEDLEY_PRIVATE
+#undef JSON_HEDLEY_PUBLIC
+#undef JSON_HEDLEY_PURE
+#undef JSON_HEDLEY_REINTERPRET_CAST
+#undef JSON_HEDLEY_REQUIRE
+#undef JSON_HEDLEY_REQUIRE_CONSTEXPR
+#undef JSON_HEDLEY_REQUIRE_MSG
+#undef JSON_HEDLEY_RESTRICT
+#undef JSON_HEDLEY_RETURNS_NON_NULL
+#undef JSON_HEDLEY_SENTINEL
+#undef JSON_HEDLEY_STATIC_ASSERT
+#undef JSON_HEDLEY_STATIC_CAST
+#undef JSON_HEDLEY_STRINGIFY
+#undef JSON_HEDLEY_STRINGIFY_EX
+#undef JSON_HEDLEY_SUNPRO_VERSION
+#undef JSON_HEDLEY_SUNPRO_VERSION_CHECK
+#undef JSON_HEDLEY_TINYC_VERSION
+#undef JSON_HEDLEY_TINYC_VERSION_CHECK
+#undef JSON_HEDLEY_TI_ARMCL_VERSION
+#undef JSON_HEDLEY_TI_ARMCL_VERSION_CHECK
+#undef JSON_HEDLEY_TI_CL2000_VERSION
+#undef JSON_HEDLEY_TI_CL2000_VERSION_CHECK
+#undef JSON_HEDLEY_TI_CL430_VERSION
+#undef JSON_HEDLEY_TI_CL430_VERSION_CHECK
+#undef JSON_HEDLEY_TI_CL6X_VERSION
+#undef JSON_HEDLEY_TI_CL6X_VERSION_CHECK
+#undef JSON_HEDLEY_TI_CL7X_VERSION
+#undef JSON_HEDLEY_TI_CL7X_VERSION_CHECK
+#undef JSON_HEDLEY_TI_CLPRU_VERSION
+#undef JSON_HEDLEY_TI_CLPRU_VERSION_CHECK
+#undef JSON_HEDLEY_TI_VERSION
+#undef JSON_HEDLEY_TI_VERSION_CHECK
+#undef JSON_HEDLEY_UNAVAILABLE
+#undef JSON_HEDLEY_UNLIKELY
+#undef JSON_HEDLEY_UNPREDICTABLE
+#undef JSON_HEDLEY_UNREACHABLE
+#undef JSON_HEDLEY_UNREACHABLE_RETURN
+#undef JSON_HEDLEY_VERSION
+#undef JSON_HEDLEY_VERSION_DECODE_MAJOR
+#undef JSON_HEDLEY_VERSION_DECODE_MINOR
+#undef JSON_HEDLEY_VERSION_DECODE_REVISION
+#undef JSON_HEDLEY_VERSION_ENCODE
+#undef JSON_HEDLEY_WARNING
+#undef JSON_HEDLEY_WARN_UNUSED_RESULT
+#undef JSON_HEDLEY_WARN_UNUSED_RESULT_MSG
+#undef JSON_HEDLEY_FALL_THROUGH
+
+
+
+#endif  // INCLUDE_NLOHMANN_JSON_HPP_
diff --git a/include/nlohmann/json_original.hpp b/include/nlohmann/json_original.hpp
new file mode 100644
index 0000000..7a857e6
--- /dev/null
+++ b/include/nlohmann/json_original.hpp
@@ -0,0 +1,5306 @@
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++
+// |  |  |__   |  |  | | | |  version 3.11.3
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013 - 2025 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+/****************************************************************************\
+ * Note on documentation: The source files contain links to the online      *
+ * documentation of the public API at https://json.nlohmann.me. This URL    *
+ * contains the most recent documentation and should also be applicable to  *
+ * previous versions; documentation for deprecated functions is not         *
+ * removed, but marked deprecated. See "Generate documentation" section in  *
+ * file docs/README.md.                                                     *
+\****************************************************************************/
+
+#ifndef INCLUDE_NLOHMANN_JSON_HPP_
+#define INCLUDE_NLOHMANN_JSON_HPP_
+
+#include <algorithm> // all_of, find, for_each
+#include <cstddef> // nullptr_t, ptrdiff_t, size_t
+#include <functional> // hash, less
+#include <initializer_list> // initializer_list
+#ifndef JSON_NO_IO
+    #include <iosfwd> // istream, ostream
+#endif  // JSON_NO_IO
+#include <iterator> // random_access_iterator_tag
+#include <memory> // unique_ptr
+#include <string> // string, stoi, to_string
+#include <utility> // declval, forward, move, pair, swap
+#include <vector> // vector
+
+#include <nlohmann/adl_serializer.hpp>
+#include <nlohmann/byte_container_with_subtype.hpp>
+#include <nlohmann/detail/conversions/from_json.hpp>
+#include <nlohmann/detail/conversions/to_json.hpp>
+#include <nlohmann/detail/exceptions.hpp>
+#include <nlohmann/detail/hash.hpp>
+#include <nlohmann/detail/input/binary_reader.hpp>
+#include <nlohmann/detail/input/input_adapters.hpp>
+#include <nlohmann/detail/input/lexer.hpp>
+#include <nlohmann/detail/input/parser.hpp>
+#include <nlohmann/detail/iterators/internal_iterator.hpp>
+#include <nlohmann/detail/iterators/iter_impl.hpp>
+#include <nlohmann/detail/iterators/iteration_proxy.hpp>
+#include <nlohmann/detail/iterators/json_reverse_iterator.hpp>
+#include <nlohmann/detail/iterators/primitive_iterator.hpp>
+#include <nlohmann/detail/json_custom_base_class.hpp>
+#include <nlohmann/detail/json_pointer.hpp>
+#include <nlohmann/detail/json_ref.hpp>
+#include <nlohmann/detail/macro_scope.hpp>
+#include <nlohmann/detail/string_concat.hpp>
+#include <nlohmann/detail/string_escape.hpp>
+#include <nlohmann/detail/string_utils.hpp>
+#include <nlohmann/detail/meta/cpp_future.hpp>
+#include <nlohmann/detail/meta/type_traits.hpp>
+#include <nlohmann/detail/output/binary_writer.hpp>
+#include <nlohmann/detail/output/output_adapters.hpp>
+#include <nlohmann/detail/output/serializer.hpp>
+#include <nlohmann/detail/value_t.hpp>
+#include <nlohmann/json_fwd.hpp>
+#include <nlohmann/ordered_map.hpp>
+
+#if defined(JSON_HAS_CPP_17)
+    #if JSON_HAS_STATIC_RTTI
+        #include <any>
+    #endif
+    #include <string_view>
+#endif
+
+/*!
+@brief namespace for Niels Lohmann
+@see https://github.com/nlohmann
+@since version 1.0.0
+*/
+NLOHMANN_JSON_NAMESPACE_BEGIN
+
+/*!
+@brief a class to store JSON values
+
+@internal
+@invariant The member variables @a m_value and @a m_type have the following
+relationship:
+- If `m_type == value_t::object`, then `m_value.object != nullptr`.
+- If `m_type == value_t::array`, then `m_value.array != nullptr`.
+- If `m_type == value_t::string`, then `m_value.string != nullptr`.
+The invariants are checked by member function assert_invariant().
+
+@note ObjectType trick from https://stackoverflow.com/a/9860911
+@endinternal
+
+@since version 1.0.0
+
+@nosubgrouping
+*/
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-special-member-functions)
+    : public ::nlohmann::detail::json_base_class<CustomBaseClass>
+{
+  private:
+    template<detail::value_t> friend struct detail::external_constructor;
+
+    template<typename>
+    friend class ::nlohmann::json_pointer;
+    // can be restored when json_pointer backwards compatibility is removed
+    // friend ::nlohmann::json_pointer<StringType>;
+
+    template<typename BasicJsonType, typename InputType>
+    friend class ::nlohmann::detail::parser;
+    friend ::nlohmann::detail::serializer<basic_json>;
+    template<typename BasicJsonType>
+    friend class ::nlohmann::detail::iter_impl;
+    template<typename BasicJsonType, typename CharType>
+    friend class ::nlohmann::detail::binary_writer;
+    template<typename BasicJsonType, typename InputType, typename SAX>
+    friend class ::nlohmann::detail::binary_reader;
+    template<typename BasicJsonType, typename InputAdapterType>
+    friend class ::nlohmann::detail::json_sax_dom_parser;
+    template<typename BasicJsonType, typename InputAdapterType>
+    friend class ::nlohmann::detail::json_sax_dom_callback_parser;
+    friend class ::nlohmann::detail::exception;
+
+    /// workaround type for MSVC
+    using basic_json_t = NLOHMANN_BASIC_JSON_TPL;
+    using json_base_class_t = ::nlohmann::detail::json_base_class<CustomBaseClass>;
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    // convenience aliases for types residing in namespace detail;
+    using lexer = ::nlohmann::detail::lexer_base<basic_json>;
+
+    template<typename InputAdapterType>
+    static ::nlohmann::detail::parser<basic_json, InputAdapterType> parser(
+        InputAdapterType adapter,
+        detail::parser_callback_t<basic_json>cb = nullptr,
+        const bool allow_exceptions = true,
+        const bool ignore_comments = false
+    )
+    {
+        return ::nlohmann::detail::parser<basic_json, InputAdapterType>(std::move(adapter),
+               std::move(cb), allow_exceptions, ignore_comments);
+    }
+
+  private:
+    using primitive_iterator_t = ::nlohmann::detail::primitive_iterator_t;
+    template<typename BasicJsonType>
+    using internal_iterator = ::nlohmann::detail::internal_iterator<BasicJsonType>;
+    template<typename BasicJsonType>
+    using iter_impl = ::nlohmann::detail::iter_impl<BasicJsonType>;
+    template<typename Iterator>
+    using iteration_proxy = ::nlohmann::detail::iteration_proxy<Iterator>;
+    template<typename Base> using json_reverse_iterator = ::nlohmann::detail::json_reverse_iterator<Base>;
+
+    template<typename CharType>
+    using output_adapter_t = ::nlohmann::detail::output_adapter_t<CharType>;
+
+    template<typename InputType>
+    using binary_reader = ::nlohmann::detail::binary_reader<basic_json, InputType>;
+    template<typename CharType> using binary_writer = ::nlohmann::detail::binary_writer<basic_json, CharType>;
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    using serializer = ::nlohmann::detail::serializer<basic_json>;
+
+  public:
+    using value_t = detail::value_t;
+    /// JSON Pointer, see @ref nlohmann::json_pointer
+    using json_pointer = ::nlohmann::json_pointer<StringType>;
+    template<typename T, typename SFINAE>
+    using json_serializer = JSONSerializer<T, SFINAE>;
+    /// how to treat decoding errors
+    using error_handler_t = detail::error_handler_t;
+    /// how to treat CBOR tags
+    using cbor_tag_handler_t = detail::cbor_tag_handler_t;
+    /// how to encode BJData
+    using bjdata_version_t = detail::bjdata_version_t;
+    /// helper type for initializer lists of basic_json values
+    using initializer_list_t = std::initializer_list<detail::json_ref<basic_json>>;
+
+    using input_format_t = detail::input_format_t;
+    /// SAX interface type, see @ref nlohmann::json_sax
+    using json_sax_t = json_sax<basic_json>;
+
+    ////////////////
+    // exceptions //
+    ////////////////
+
+    /// @name exceptions
+    /// Classes to implement user-defined exceptions.
+    /// @{
+
+    using exception = detail::exception;
+    using parse_error = detail::parse_error;
+    using invalid_iterator = detail::invalid_iterator;
+    using type_error = detail::type_error;
+    using out_of_range = detail::out_of_range;
+    using other_error = detail::other_error;
+
+    /// @}
+
+    /////////////////////
+    // container types //
+    /////////////////////
+
+    /// @name container types
+    /// The canonic container types to use @ref basic_json like any other STL
+    /// container.
+    /// @{
+
+    /// the type of elements in a basic_json container
+    using value_type = basic_json;
+
+    /// the type of an element reference
+    using reference = value_type&;
+    /// the type of an element const reference
+    using const_reference = const value_type&;
+
+    /// a type to represent differences between iterators
+    using difference_type = std::ptrdiff_t;
+    /// a type to represent container sizes
+    using size_type = std::size_t;
+
+    /// the allocator type
+    using allocator_type = AllocatorType<basic_json>;
+
+    /// the type of an element pointer
+    using pointer = typename std::allocator_traits<allocator_type>::pointer;
+    /// the type of an element const pointer
+    using const_pointer = typename std::allocator_traits<allocator_type>::const_pointer;
+
+    /// an iterator for a basic_json container
+    using iterator = iter_impl<basic_json>;
+    /// a const iterator for a basic_json container
+    using const_iterator = iter_impl<const basic_json>;
+    /// a reverse iterator for a basic_json container
+    using reverse_iterator = json_reverse_iterator<typename basic_json::iterator>;
+    /// a const reverse iterator for a basic_json container
+    using const_reverse_iterator = json_reverse_iterator<typename basic_json::const_iterator>;
+
+    /// @}
+
+    /// @brief returns the allocator associated with the container
+    /// @sa https://json.nlohmann.me/api/basic_json/get_allocator/
+    static allocator_type get_allocator()
+    {
+        return allocator_type();
+    }
+
+    /// @brief returns version information on the library
+    /// @sa https://json.nlohmann.me/api/basic_json/meta/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json meta()
+    {
+        basic_json result;
+
+        result["copyright"] = "(C) 2013-2023 Niels Lohmann";
+        result["name"] = "JSON for Modern C++";
+        result["url"] = "https://github.com/nlohmann/json";
+        result["version"]["string"] =
+            detail::concat(std::to_string(NLOHMANN_JSON_VERSION_MAJOR), '.',
+                           std::to_string(NLOHMANN_JSON_VERSION_MINOR), '.',
+                           std::to_string(NLOHMANN_JSON_VERSION_PATCH));
+        result["version"]["major"] = NLOHMANN_JSON_VERSION_MAJOR;
+        result["version"]["minor"] = NLOHMANN_JSON_VERSION_MINOR;
+        result["version"]["patch"] = NLOHMANN_JSON_VERSION_PATCH;
+
+#ifdef _WIN32
+        result["platform"] = "win32";
+#elif defined __linux__
+        result["platform"] = "linux";
+#elif defined __APPLE__
+        result["platform"] = "apple";
+#elif defined __unix__
+        result["platform"] = "unix";
+#else
+        result["platform"] = "unknown";
+#endif
+
+#if defined(__ICC) || defined(__INTEL_COMPILER)
+        result["compiler"] = {{"family", "icc"}, {"version", __INTEL_COMPILER}};
+#elif defined(__clang__)
+        result["compiler"] = {{"family", "clang"}, {"version", __clang_version__}};
+#elif defined(__GNUC__) || defined(__GNUG__)
+        result["compiler"] = {{"family", "gcc"}, {"version", detail::concat(
+                    std::to_string(__GNUC__), '.',
+                    std::to_string(__GNUC_MINOR__), '.',
+                    std::to_string(__GNUC_PATCHLEVEL__))
+            }
+        };
+#elif defined(__HP_cc) || defined(__HP_aCC)
+        result["compiler"] = "hp"
+#elif defined(__IBMCPP__)
+        result["compiler"] = {{"family", "ilecpp"}, {"version", __IBMCPP__}};
+#elif defined(_MSC_VER)
+        result["compiler"] = {{"family", "msvc"}, {"version", _MSC_VER}};
+#elif defined(__PGI)
+        result["compiler"] = {{"family", "pgcpp"}, {"version", __PGI}};
+#elif defined(__SUNPRO_CC)
+        result["compiler"] = {{"family", "sunpro"}, {"version", __SUNPRO_CC}};
+#else
+        result["compiler"] = {{"family", "unknown"}, {"version", "unknown"}};
+#endif
+
+#if defined(_MSVC_LANG)
+        result["compiler"]["c++"] = std::to_string(_MSVC_LANG);
+#elif defined(__cplusplus)
+        result["compiler"]["c++"] = std::to_string(__cplusplus);
+#else
+        result["compiler"]["c++"] = "unknown";
+#endif
+        return result;
+    }
+
+    ///////////////////////////
+    // JSON value data types //
+    ///////////////////////////
+
+    /// @name JSON value data types
+    /// The data types to store a JSON value. These types are derived from
+    /// the template arguments passed to class @ref basic_json.
+    /// @{
+
+    /// @brief default object key comparator type
+    /// The actual object key comparator type (@ref object_comparator_t) may be
+    /// different.
+    /// @sa https://json.nlohmann.me/api/basic_json/default_object_comparator_t/
+#if defined(JSON_HAS_CPP_14)
+    // use of transparent comparator avoids unnecessary repeated construction of temporaries
+    // in functions involving lookup by key with types other than object_t::key_type (aka. StringType)
+    using default_object_comparator_t = std::less<>;
+#else
+    using default_object_comparator_t = std::less<StringType>;
+#endif
+
+    /// @brief a type for an object
+    /// @sa https://json.nlohmann.me/api/basic_json/object_t/
+    using object_t = ObjectType<StringType,
+          basic_json,
+          default_object_comparator_t,
+          AllocatorType<std::pair<const StringType,
+          basic_json>>>;
+
+    /// @brief a type for an array
+    /// @sa https://json.nlohmann.me/api/basic_json/array_t/
+    using array_t = ArrayType<basic_json, AllocatorType<basic_json>>;
+
+    /// @brief a type for a string
+    /// @sa https://json.nlohmann.me/api/basic_json/string_t/
+    using string_t = StringType;
+
+    /// @brief a type for a boolean
+    /// @sa https://json.nlohmann.me/api/basic_json/boolean_t/
+    using boolean_t = BooleanType;
+
+    /// @brief a type for a number (integer)
+    /// @sa https://json.nlohmann.me/api/basic_json/number_integer_t/
+    using number_integer_t = NumberIntegerType;
+
+    /// @brief a type for a number (unsigned)
+    /// @sa https://json.nlohmann.me/api/basic_json/number_unsigned_t/
+    using number_unsigned_t = NumberUnsignedType;
+
+    /// @brief a type for a number (floating-point)
+    /// @sa https://json.nlohmann.me/api/basic_json/number_float_t/
+    using number_float_t = NumberFloatType;
+
+    /// @brief a type for a packed binary type
+    /// @sa https://json.nlohmann.me/api/basic_json/binary_t/
+    using binary_t = nlohmann::byte_container_with_subtype<BinaryType>;
+
+    /// @brief object key comparator type
+    /// @sa https://json.nlohmann.me/api/basic_json/object_comparator_t/
+    using object_comparator_t = detail::actual_object_comparator_t<basic_json>;
+
+    /// @}
+
+  private:
+
+    /// helper for exception-safe object creation
+    template<typename T, typename... Args>
+    JSON_HEDLEY_RETURNS_NON_NULL
+    static T* create(Args&& ... args)
+    {
+        AllocatorType<T> alloc;
+        using AllocatorTraits = std::allocator_traits<AllocatorType<T>>;
+
+        auto deleter = [&](T * obj)
+        {
+            AllocatorTraits::deallocate(alloc, obj, 1);
+        };
+        std::unique_ptr<T, decltype(deleter)> obj(AllocatorTraits::allocate(alloc, 1), deleter);
+        AllocatorTraits::construct(alloc, obj.get(), std::forward<Args>(args)...);
+        JSON_ASSERT(obj != nullptr);
+        return obj.release();
+    }
+
+    ////////////////////////
+    // JSON value storage //
+    ////////////////////////
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    /*!
+    @brief a JSON value
+
+    The actual storage for a JSON value of the @ref basic_json class. This
+    union combines the different storage types for the JSON value types
+    defined in @ref value_t.
+
+    JSON type | value_t type    | used type
+    --------- | --------------- | ------------------------
+    object    | object          | pointer to @ref object_t
+    array     | array           | pointer to @ref array_t
+    string    | string          | pointer to @ref string_t
+    boolean   | boolean         | @ref boolean_t
+    number    | number_integer  | @ref number_integer_t
+    number    | number_unsigned | @ref number_unsigned_t
+    number    | number_float    | @ref number_float_t
+    binary    | binary          | pointer to @ref binary_t
+    null      | null            | *no value is stored*
+
+    @note Variable-length types (objects, arrays, and strings) are stored as
+    pointers. The size of the union should not exceed 64 bits if the default
+    value types are used.
+
+    @since version 1.0.0
+    */
+    union json_value
+    {
+        /// object (stored with pointer to save storage)
+        object_t* object;
+        /// array (stored with pointer to save storage)
+        array_t* array;
+        /// string (stored with pointer to save storage)
+        string_t* string;
+        /// binary (stored with pointer to save storage)
+        binary_t* binary;
+        /// boolean
+        boolean_t boolean;
+        /// number (integer)
+        number_integer_t number_integer;
+        /// number (unsigned integer)
+        number_unsigned_t number_unsigned;
+        /// number (floating-point)
+        number_float_t number_float;
+
+        /// default constructor (for null values)
+        json_value() = default;
+        /// constructor for booleans
+        json_value(boolean_t v) noexcept : boolean(v) {}
+        /// constructor for numbers (integer)
+        json_value(number_integer_t v) noexcept : number_integer(v) {}
+        /// constructor for numbers (unsigned)
+        json_value(number_unsigned_t v) noexcept : number_unsigned(v) {}
+        /// constructor for numbers (floating-point)
+        json_value(number_float_t v) noexcept : number_float(v) {}
+        /// constructor for empty values of a given type
+        json_value(value_t t)
+        {
+            switch (t)
+            {
+                case value_t::object:
+                {
+                    object = create<object_t>();
+                    break;
+                }
+
+                case value_t::array:
+                {
+                    array = create<array_t>();
+                    break;
+                }
+
+                case value_t::string:
+                {
+                    string = create<string_t>("");
+                    break;
+                }
+
+                case value_t::binary:
+                {
+                    binary = create<binary_t>();
+                    break;
+                }
+
+                case value_t::boolean:
+                {
+                    boolean = static_cast<boolean_t>(false);
+                    break;
+                }
+
+                case value_t::number_integer:
+                {
+                    number_integer = static_cast<number_integer_t>(0);
+                    break;
+                }
+
+                case value_t::number_unsigned:
+                {
+                    number_unsigned = static_cast<number_unsigned_t>(0);
+                    break;
+                }
+
+                case value_t::number_float:
+                {
+                    number_float = static_cast<number_float_t>(0.0);
+                    break;
+                }
+
+                case value_t::null:
+                {
+                    object = nullptr;  // silence warning, see #821
+                    break;
+                }
+
+                case value_t::discarded:
+                default:
+                {
+                    object = nullptr;  // silence warning, see #821
+                    if (JSON_HEDLEY_UNLIKELY(t == value_t::null))
+                    {
+                        JSON_THROW(other_error::create(500, "961c151d2e87f2686a955a9be24d316f1362bf21 3.11.3", nullptr)); // LCOV_EXCL_LINE
+                    }
+                    break;
+                }
+            }
+        }
+
+        /// constructor for strings
+        json_value(const string_t& value) : string(create<string_t>(value)) {}
+
+        /// constructor for rvalue strings
+        json_value(string_t&& value) : string(create<string_t>(std::move(value))) {}
+
+        /// constructor for objects
+        json_value(const object_t& value) : object(create<object_t>(value)) {}
+
+        /// constructor for rvalue objects
+        json_value(object_t&& value) : object(create<object_t>(std::move(value))) {}
+
+        /// constructor for arrays
+        json_value(const array_t& value) : array(create<array_t>(value)) {}
+
+        /// constructor for rvalue arrays
+        json_value(array_t&& value) : array(create<array_t>(std::move(value))) {}
+
+        /// constructor for binary arrays
+        json_value(const typename binary_t::container_type& value) : binary(create<binary_t>(value)) {}
+
+        /// constructor for rvalue binary arrays
+        json_value(typename binary_t::container_type&& value) : binary(create<binary_t>(std::move(value))) {}
+
+        /// constructor for binary arrays (internal type)
+        json_value(const binary_t& value) : binary(create<binary_t>(value)) {}
+
+        /// constructor for rvalue binary arrays (internal type)
+        json_value(binary_t&& value) : binary(create<binary_t>(std::move(value))) {}
+
+        void destroy(value_t t)
+        {
+            if (
+                (t == value_t::object && object == nullptr) ||
+                (t == value_t::array && array == nullptr) ||
+                (t == value_t::string && string == nullptr) ||
+                (t == value_t::binary && binary == nullptr)
+            )
+            {
+                //not initialized (e.g. due to exception in the ctor)
+                return;
+            }
+            if (t == value_t::array || t == value_t::object)
+            {
+                // flatten the current json_value to a heap-allocated stack
+                std::vector<basic_json> stack;
+
+                // move the top-level items to stack
+                if (t == value_t::array)
+                {
+                    stack.reserve(array->size());
+                    std::move(array->begin(), array->end(), std::back_inserter(stack));
+                }
+                else
+                {
+                    stack.reserve(object->size());
+                    for (auto&& it : *object)
+                    {
+                        stack.push_back(std::move(it.second));
+                    }
+                }
+
+                while (!stack.empty())
+                {
+                    // move the last item to local variable to be processed
+                    basic_json current_item(std::move(stack.back()));
+                    stack.pop_back();
+
+                    // if current_item is array/object, move
+                    // its children to the stack to be processed later
+                    if (current_item.is_array())
+                    {
+                        std::move(current_item.m_data.m_value.array->begin(), current_item.m_data.m_value.array->end(), std::back_inserter(stack));
+
+                        current_item.m_data.m_value.array->clear();
+                    }
+                    else if (current_item.is_object())
+                    {
+                        for (auto&& it : *current_item.m_data.m_value.object)
+                        {
+                            stack.push_back(std::move(it.second));
+                        }
+
+                        current_item.m_data.m_value.object->clear();
+                    }
+
+                    // it's now safe that current_item get destructed
+                    // since it doesn't have any children
+                }
+            }
+
+            switch (t)
+            {
+                case value_t::object:
+                {
+                    AllocatorType<object_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, object);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, object, 1);
+                    break;
+                }
+
+                case value_t::array:
+                {
+                    AllocatorType<array_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, array);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, array, 1);
+                    break;
+                }
+
+                case value_t::string:
+                {
+                    AllocatorType<string_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, string);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, string, 1);
+                    break;
+                }
+
+                case value_t::binary:
+                {
+                    AllocatorType<binary_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, binary);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, binary, 1);
+                    break;
+                }
+
+                case value_t::null:
+                case value_t::boolean:
+                case value_t::number_integer:
+                case value_t::number_unsigned:
+                case value_t::number_float:
+                case value_t::discarded:
+                default:
+                {
+                    break;
+                }
+            }
+        }
+    };
+
+  private:
+    /*!
+    @brief checks the class invariants
+
+    This function asserts the class invariants. It needs to be called at the
+    end of every constructor to make sure that created objects respect the
+    invariant. Furthermore, it has to be called each time the type of a JSON
+    value is changed, because the invariant expresses a relationship between
+    @a m_type and @a m_value.
+
+    Furthermore, the parent relation is checked for arrays and objects: If
+    @a check_parents true and the value is an array or object, then the
+    container's elements must have the current value as parent.
+
+    @param[in] check_parents  whether the parent relation should be checked.
+               The value is true by default and should only be set to false
+               during destruction of objects when the invariant does not
+               need to hold.
+    */
+    void assert_invariant(bool check_parents = true) const noexcept
+    {
+        JSON_ASSERT(m_data.m_type != value_t::object || m_data.m_value.object != nullptr);
+        JSON_ASSERT(m_data.m_type != value_t::array || m_data.m_value.array != nullptr);
+        JSON_ASSERT(m_data.m_type != value_t::string || m_data.m_value.string != nullptr);
+        JSON_ASSERT(m_data.m_type != value_t::binary || m_data.m_value.binary != nullptr);
+
+#if JSON_DIAGNOSTICS
+        JSON_TRY
+        {
+            // cppcheck-suppress assertWithSideEffect
+            JSON_ASSERT(!check_parents || !is_structured() || std::all_of(begin(), end(), [this](const basic_json & j)
+            {
+                return j.m_parent == this;
+            }));
+        }
+        JSON_CATCH(...) {} // LCOV_EXCL_LINE
+#endif
+        static_cast<void>(check_parents);
+    }
+
+    void set_parents()
+    {
+#if JSON_DIAGNOSTICS
+        switch (m_data.m_type)
+        {
+            case value_t::array:
+            {
+                for (auto& element : *m_data.m_value.array)
+                {
+                    element.m_parent = this;
+                }
+                break;
+            }
+
+            case value_t::object:
+            {
+                for (auto& element : *m_data.m_value.object)
+                {
+                    element.second.m_parent = this;
+                }
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+                break;
+        }
+#endif
+    }
+
+    iterator set_parents(iterator it, typename iterator::difference_type count_set_parents)
+    {
+#if JSON_DIAGNOSTICS
+        for (typename iterator::difference_type i = 0; i < count_set_parents; ++i)
+        {
+            (it + i)->m_parent = this;
+        }
+#else
+        static_cast<void>(count_set_parents);
+#endif
+        return it;
+    }
+
+    reference set_parent(reference j, std::size_t old_capacity = detail::unknown_size())
+    {
+#if JSON_DIAGNOSTICS
+        if (old_capacity != detail::unknown_size())
+        {
+            // see https://github.com/nlohmann/json/issues/2838
+            JSON_ASSERT(type() == value_t::array);
+            if (JSON_HEDLEY_UNLIKELY(m_data.m_value.array->capacity() != old_capacity))
+            {
+                // capacity has changed: update all parents
+                set_parents();
+                return j;
+            }
+        }
+
+        // ordered_json uses a vector internally, so pointers could have
+        // been invalidated; see https://github.com/nlohmann/json/issues/2962
+#ifdef JSON_HEDLEY_MSVC_VERSION
+#pragma warning(push )
+#pragma warning(disable : 4127) // ignore warning to replace if with if constexpr
+#endif
+        if (detail::is_ordered_map<object_t>::value)
+        {
+            set_parents();
+            return j;
+        }
+#ifdef JSON_HEDLEY_MSVC_VERSION
+#pragma warning( pop )
+#endif
+
+        j.m_parent = this;
+#else
+        static_cast<void>(j);
+        static_cast<void>(old_capacity);
+#endif
+        return j;
+    }
+
+  public:
+    //////////////////////////
+    // JSON parser callback //
+    //////////////////////////
+
+    /// @brief parser event types
+    /// @sa https://json.nlohmann.me/api/basic_json/parse_event_t/
+    using parse_event_t = detail::parse_event_t;
+
+    /// @brief per-element parser callback type
+    /// @sa https://json.nlohmann.me/api/basic_json/parser_callback_t/
+    using parser_callback_t = detail::parser_callback_t<basic_json>;
+
+    //////////////////
+    // constructors //
+    //////////////////
+
+    /// @name constructors and destructors
+    /// Constructors of class @ref basic_json, copy/move constructor, copy
+    /// assignment, static functions creating objects, and the destructor.
+    /// @{
+
+    /// @brief create an empty value with a given type
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(const value_t v)
+        : m_data(v)
+    {
+        assert_invariant();
+    }
+
+    /// @brief create a null object
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(std::nullptr_t = nullptr) noexcept // NOLINT(bugprone-exception-escape)
+        : basic_json(value_t::null)
+    {
+        assert_invariant();
+    }
+
+    /// @brief create a JSON value from compatible types
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    template < typename CompatibleType,
+               typename U = detail::uncvref_t<CompatibleType>,
+               detail::enable_if_t <
+                   !detail::is_basic_json<U>::value && detail::is_compatible_type<basic_json_t, U>::value, int > = 0 >
+    basic_json(CompatibleType && val) noexcept(noexcept( // NOLINT(bugprone-forwarding-reference-overload,bugprone-exception-escape)
+                JSONSerializer<U>::to_json(std::declval<basic_json_t&>(),
+                                           std::forward<CompatibleType>(val))))
+    {
+        JSONSerializer<U>::to_json(*this, std::forward<CompatibleType>(val));
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief create a JSON value from an existing one
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    template < typename BasicJsonType,
+               detail::enable_if_t <
+                   detail::is_basic_json<BasicJsonType>::value&& !std::is_same<basic_json, BasicJsonType>::value, int > = 0 >
+    basic_json(const BasicJsonType& val)
+#if JSON_DIAGNOSTIC_POSITIONS
+        : start_position(val.start_pos()),
+          end_position(val.end_pos())
+#endif
+    {
+        using other_boolean_t = typename BasicJsonType::boolean_t;
+        using other_number_float_t = typename BasicJsonType::number_float_t;
+        using other_number_integer_t = typename BasicJsonType::number_integer_t;
+        using other_number_unsigned_t = typename BasicJsonType::number_unsigned_t;
+        using other_string_t = typename BasicJsonType::string_t;
+        using other_object_t = typename BasicJsonType::object_t;
+        using other_array_t = typename BasicJsonType::array_t;
+        using other_binary_t = typename BasicJsonType::binary_t;
+
+        switch (val.type())
+        {
+            case value_t::boolean:
+                JSONSerializer<other_boolean_t>::to_json(*this, val.template get<other_boolean_t>());
+                break;
+            case value_t::number_float:
+                JSONSerializer<other_number_float_t>::to_json(*this, val.template get<other_number_float_t>());
+                break;
+            case value_t::number_integer:
+                JSONSerializer<other_number_integer_t>::to_json(*this, val.template get<other_number_integer_t>());
+                break;
+            case value_t::number_unsigned:
+                JSONSerializer<other_number_unsigned_t>::to_json(*this, val.template get<other_number_unsigned_t>());
+                break;
+            case value_t::string:
+                JSONSerializer<other_string_t>::to_json(*this, val.template get_ref<const other_string_t&>());
+                break;
+            case value_t::object:
+                JSONSerializer<other_object_t>::to_json(*this, val.template get_ref<const other_object_t&>());
+                break;
+            case value_t::array:
+                JSONSerializer<other_array_t>::to_json(*this, val.template get_ref<const other_array_t&>());
+                break;
+            case value_t::binary:
+                JSONSerializer<other_binary_t>::to_json(*this, val.template get_ref<const other_binary_t&>());
+                break;
+            case value_t::null:
+                *this = nullptr;
+                break;
+            case value_t::discarded:
+                m_data.m_type = value_t::discarded;
+                break;
+            default:            // LCOV_EXCL_LINE
+                JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+        }
+        JSON_ASSERT(m_data.m_type == val.type());
+
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief create a container (array or object) from an initializer list
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(initializer_list_t init,
+               bool type_deduction = true,
+               value_t manual_type = value_t::array)
+    {
+        // check if each element is an array with two elements whose first
+        // element is a string
+        bool is_an_object = std::all_of(init.begin(), init.end(),
+                                        [](const detail::json_ref<basic_json>& element_ref)
+        {
+            // The cast is to ensure op[size_type] is called, bearing in mind size_type may not be int;
+            // (many string types can be constructed from 0 via its null-pointer guise, so we get a
+            // broken call to op[key_type], the wrong semantics and a 4804 warning on Windows)
+            return element_ref->is_array() && element_ref->size() == 2 && (*element_ref)[static_cast<size_type>(0)].is_string();
+        });
+
+        // adjust type if type deduction is not wanted
+        if (!type_deduction)
+        {
+            // if array is wanted, do not create an object though possible
+            if (manual_type == value_t::array)
+            {
+                is_an_object = false;
+            }
+
+            // if object is wanted but impossible, throw an exception
+            if (JSON_HEDLEY_UNLIKELY(manual_type == value_t::object && !is_an_object))
+            {
+                JSON_THROW(type_error::create(301, "cannot create object from initializer list", nullptr));
+            }
+        }
+
+        if (is_an_object)
+        {
+            // the initializer list is a list of pairs -> create object
+            m_data.m_type = value_t::object;
+            m_data.m_value = value_t::object;
+
+            for (auto& element_ref : init)
+            {
+                auto element = element_ref.moved_or_copied();
+                m_data.m_value.object->emplace(
+                    std::move(*((*element.m_data.m_value.array)[0].m_data.m_value.string)),
+                    std::move((*element.m_data.m_value.array)[1]));
+            }
+        }
+        else
+        {
+            // the initializer list describes an array -> create array
+            m_data.m_type = value_t::array;
+            m_data.m_value.array = create<array_t>(init.begin(), init.end());
+        }
+
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief explicitly create a binary array (without subtype)
+    /// @sa https://json.nlohmann.me/api/basic_json/binary/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json binary(const typename binary_t::container_type& init)
+    {
+        auto res = basic_json();
+        res.m_data.m_type = value_t::binary;
+        res.m_data.m_value = init;
+        return res;
+    }
+
+    /// @brief explicitly create a binary array (with subtype)
+    /// @sa https://json.nlohmann.me/api/basic_json/binary/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json binary(const typename binary_t::container_type& init, typename binary_t::subtype_type subtype)
+    {
+        auto res = basic_json();
+        res.m_data.m_type = value_t::binary;
+        res.m_data.m_value = binary_t(init, subtype);
+        return res;
+    }
+
+    /// @brief explicitly create a binary array
+    /// @sa https://json.nlohmann.me/api/basic_json/binary/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json binary(typename binary_t::container_type&& init)
+    {
+        auto res = basic_json();
+        res.m_data.m_type = value_t::binary;
+        res.m_data.m_value = std::move(init);
+        return res;
+    }
+
+    /// @brief explicitly create a binary array (with subtype)
+    /// @sa https://json.nlohmann.me/api/basic_json/binary/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json binary(typename binary_t::container_type&& init, typename binary_t::subtype_type subtype)
+    {
+        auto res = basic_json();
+        res.m_data.m_type = value_t::binary;
+        res.m_data.m_value = binary_t(std::move(init), subtype);
+        return res;
+    }
+
+    /// @brief explicitly create an array from an initializer list
+    /// @sa https://json.nlohmann.me/api/basic_json/array/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json array(initializer_list_t init = {})
+    {
+        return basic_json(init, false, value_t::array);
+    }
+
+    /// @brief explicitly create an object from an initializer list
+    /// @sa https://json.nlohmann.me/api/basic_json/object/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json object(initializer_list_t init = {})
+    {
+        return basic_json(init, false, value_t::object);
+    }
+
+    /// @brief construct an array with count copies of given value
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(size_type cnt, const basic_json& val):
+        m_data{cnt, val}
+    {
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief construct a JSON container given an iterator range
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    template < class InputIT, typename std::enable_if <
+                   std::is_same<InputIT, typename basic_json_t::iterator>::value ||
+                   std::is_same<InputIT, typename basic_json_t::const_iterator>::value, int >::type = 0 >
+    basic_json(InputIT first, InputIT last) // NOLINT(performance-unnecessary-value-param)
+    {
+        JSON_ASSERT(first.m_object != nullptr);
+        JSON_ASSERT(last.m_object != nullptr);
+
+        // make sure iterator fits the current value
+        if (JSON_HEDLEY_UNLIKELY(first.m_object != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(201, "iterators are not compatible", nullptr));
+        }
+
+        // copy type from first iterator
+        m_data.m_type = first.m_object->m_data.m_type;
+
+        // check if iterator range is complete for primitive values
+        switch (m_data.m_type)
+        {
+            case value_t::boolean:
+            case value_t::number_float:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::string:
+            {
+                if (JSON_HEDLEY_UNLIKELY(!first.m_it.primitive_iterator.is_begin()
+                                         || !last.m_it.primitive_iterator.is_end()))
+                {
+                    JSON_THROW(invalid_iterator::create(204, "iterators out of range", first.m_object));
+                }
+                break;
+            }
+
+            case value_t::null:
+            case value_t::object:
+            case value_t::array:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+                break;
+        }
+
+        switch (m_data.m_type)
+        {
+            case value_t::number_integer:
+            {
+                m_data.m_value.number_integer = first.m_object->m_data.m_value.number_integer;
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                m_data.m_value.number_unsigned = first.m_object->m_data.m_value.number_unsigned;
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                m_data.m_value.number_float = first.m_object->m_data.m_value.number_float;
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                m_data.m_value.boolean = first.m_object->m_data.m_value.boolean;
+                break;
+            }
+
+            case value_t::string:
+            {
+                m_data.m_value = *first.m_object->m_data.m_value.string;
+                break;
+            }
+
+            case value_t::object:
+            {
+                m_data.m_value.object = create<object_t>(first.m_it.object_iterator,
+                                        last.m_it.object_iterator);
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_data.m_value.array = create<array_t>(first.m_it.array_iterator,
+                                                       last.m_it.array_iterator);
+                break;
+            }
+
+            case value_t::binary:
+            {
+                m_data.m_value = *first.m_object->m_data.m_value.binary;
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                JSON_THROW(invalid_iterator::create(206, detail::concat("cannot construct with iterators from ", first.m_object->type_name()), first.m_object));
+        }
+
+        set_parents();
+        assert_invariant();
+    }
+
+    ///////////////////////////////////////
+    // other constructors and destructor //
+    ///////////////////////////////////////
+
+    template<typename JsonRef,
+             detail::enable_if_t<detail::conjunction<detail::is_json_ref<JsonRef>,
+                                 std::is_same<typename JsonRef::value_type, basic_json>>::value, int> = 0 >
+    basic_json(const JsonRef& ref) : basic_json(ref.moved_or_copied()) {}
+
+    /// @brief copy constructor
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(const basic_json& other)
+        : json_base_class_t(other)
+#if JSON_DIAGNOSTIC_POSITIONS
+        , start_position(other.start_position)
+        , end_position(other.end_position)
+#endif
+    {
+        m_data.m_type = other.m_data.m_type;
+        // check of passed value is valid
+        other.assert_invariant();
+
+        switch (m_data.m_type)
+        {
+            case value_t::object:
+            {
+                m_data.m_value = *other.m_data.m_value.object;
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_data.m_value = *other.m_data.m_value.array;
+                break;
+            }
+
+            case value_t::string:
+            {
+                m_data.m_value = *other.m_data.m_value.string;
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                m_data.m_value = other.m_data.m_value.boolean;
+                break;
+            }
+
+            case value_t::number_integer:
+            {
+                m_data.m_value = other.m_data.m_value.number_integer;
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                m_data.m_value = other.m_data.m_value.number_unsigned;
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                m_data.m_value = other.m_data.m_value.number_float;
+                break;
+            }
+
+            case value_t::binary:
+            {
+                m_data.m_value = *other.m_data.m_value.binary;
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                break;
+        }
+
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief move constructor
+    /// @sa https://json.nlohmann.me/api/basic_json/basic_json/
+    basic_json(basic_json&& other) noexcept
+        : json_base_class_t(std::forward<json_base_class_t>(other)),
+          m_data(std::move(other.m_data)) // cppcheck-suppress[accessForwarded] TODO check
+#if JSON_DIAGNOSTIC_POSITIONS
+        , start_position(other.start_position) // cppcheck-suppress[accessForwarded] TODO check
+        , end_position(other.end_position) // cppcheck-suppress[accessForwarded] TODO check
+#endif
+    {
+        // check that passed value is valid
+        other.assert_invariant(false); // cppcheck-suppress[accessForwarded]
+
+        // invalidate payload
+        other.m_data.m_type = value_t::null;
+        other.m_data.m_value = {};
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        other.start_position = std::string::npos;
+        other.end_position = std::string::npos;
+#endif
+
+        set_parents();
+        assert_invariant();
+    }
+
+    /// @brief copy assignment
+    /// @sa https://json.nlohmann.me/api/basic_json/operator=/
+    basic_json& operator=(basic_json other) noexcept (
+        std::is_nothrow_move_constructible<value_t>::value&&
+        std::is_nothrow_move_assignable<value_t>::value&&
+        std::is_nothrow_move_constructible<json_value>::value&&
+        std::is_nothrow_move_assignable<json_value>::value&&
+        std::is_nothrow_move_assignable<json_base_class_t>::value
+    )
+    {
+        // check that passed value is valid
+        other.assert_invariant();
+
+        using std::swap;
+        swap(m_data.m_type, other.m_data.m_type);
+        swap(m_data.m_value, other.m_data.m_value);
+
+#if JSON_DIAGNOSTIC_POSITIONS
+        swap(start_position, other.start_position);
+        swap(end_position, other.end_position);
+#endif
+
+        json_base_class_t::operator=(std::move(other));
+
+        set_parents();
+        assert_invariant();
+        return *this;
+    }
+
+    /// @brief destructor
+    /// @sa https://json.nlohmann.me/api/basic_json/~basic_json/
+    ~basic_json() noexcept
+    {
+        assert_invariant(false);
+    }
+
+    /// @}
+
+  public:
+    ///////////////////////
+    // object inspection //
+    ///////////////////////
+
+    /// @name object inspection
+    /// Functions to inspect the type of a JSON value.
+    /// @{
+
+    /// @brief serialization
+    /// @sa https://json.nlohmann.me/api/basic_json/dump/
+    string_t dump(const int indent = -1,
+                  const char indent_char = ' ',
+                  const bool ensure_ascii = false,
+                  const error_handler_t error_handler = error_handler_t::strict) const
+    {
+        string_t result;
+        serializer s(detail::output_adapter<char, string_t>(result), indent_char, error_handler);
+
+        if (indent >= 0)
+        {
+            s.dump(*this, true, ensure_ascii, static_cast<unsigned int>(indent));
+        }
+        else
+        {
+            s.dump(*this, false, ensure_ascii, 0);
+        }
+
+        return result;
+    }
+
+    /// @brief return the type of the JSON value (explicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/type/
+    constexpr value_t type() const noexcept
+    {
+        return m_data.m_type;
+    }
+
+    /// @brief return whether type is primitive
+    /// @sa https://json.nlohmann.me/api/basic_json/is_primitive/
+    constexpr bool is_primitive() const noexcept
+    {
+        return is_null() || is_string() || is_boolean() || is_number() || is_binary();
+    }
+
+    /// @brief return whether type is structured
+    /// @sa https://json.nlohmann.me/api/basic_json/is_structured/
+    constexpr bool is_structured() const noexcept
+    {
+        return is_array() || is_object();
+    }
+
+    /// @brief return whether value is null
+    /// @sa https://json.nlohmann.me/api/basic_json/is_null/
+    constexpr bool is_null() const noexcept
+    {
+        return m_data.m_type == value_t::null;
+    }
+
+    /// @brief return whether value is a boolean
+    /// @sa https://json.nlohmann.me/api/basic_json/is_boolean/
+    constexpr bool is_boolean() const noexcept
+    {
+        return m_data.m_type == value_t::boolean;
+    }
+
+    /// @brief return whether value is a number
+    /// @sa https://json.nlohmann.me/api/basic_json/is_number/
+    constexpr bool is_number() const noexcept
+    {
+        return is_number_integer() || is_number_float();
+    }
+
+    /// @brief return whether value is an integer number
+    /// @sa https://json.nlohmann.me/api/basic_json/is_number_integer/
+    constexpr bool is_number_integer() const noexcept
+    {
+        return m_data.m_type == value_t::number_integer || m_data.m_type == value_t::number_unsigned;
+    }
+
+    /// @brief return whether value is an unsigned integer number
+    /// @sa https://json.nlohmann.me/api/basic_json/is_number_unsigned/
+    constexpr bool is_number_unsigned() const noexcept
+    {
+        return m_data.m_type == value_t::number_unsigned;
+    }
+
+    /// @brief return whether value is a floating-point number
+    /// @sa https://json.nlohmann.me/api/basic_json/is_number_float/
+    constexpr bool is_number_float() const noexcept
+    {
+        return m_data.m_type == value_t::number_float;
+    }
+
+    /// @brief return whether value is an object
+    /// @sa https://json.nlohmann.me/api/basic_json/is_object/
+    constexpr bool is_object() const noexcept
+    {
+        return m_data.m_type == value_t::object;
+    }
+
+    /// @brief return whether value is an array
+    /// @sa https://json.nlohmann.me/api/basic_json/is_array/
+    constexpr bool is_array() const noexcept
+    {
+        return m_data.m_type == value_t::array;
+    }
+
+    /// @brief return whether value is a string
+    /// @sa https://json.nlohmann.me/api/basic_json/is_string/
+    constexpr bool is_string() const noexcept
+    {
+        return m_data.m_type == value_t::string;
+    }
+
+    /// @brief return whether value is a binary array
+    /// @sa https://json.nlohmann.me/api/basic_json/is_binary/
+    constexpr bool is_binary() const noexcept
+    {
+        return m_data.m_type == value_t::binary;
+    }
+
+    /// @brief return whether value is discarded
+    /// @sa https://json.nlohmann.me/api/basic_json/is_discarded/
+    constexpr bool is_discarded() const noexcept
+    {
+        return m_data.m_type == value_t::discarded;
+    }
+
+    /// @brief return the type of the JSON value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_value_t/
+    constexpr operator value_t() const noexcept
+    {
+        return m_data.m_type;
+    }
+
+    /// @}
+
+  private:
+    //////////////////
+    // value access //
+    //////////////////
+
+    /// get a boolean (explicit)
+    boolean_t get_impl(boolean_t* /*unused*/) const
+    {
+        if (JSON_HEDLEY_LIKELY(is_boolean()))
+        {
+            return m_data.m_value.boolean;
+        }
+
+        JSON_THROW(type_error::create(302, detail::concat("type must be boolean, but is ", type_name()), this));
+    }
+
+    /// get a pointer to the value (object)
+    object_t* get_impl_ptr(object_t* /*unused*/) noexcept
+    {
+        return is_object() ? m_data.m_value.object : nullptr;
+    }
+
+    /// get a pointer to the value (object)
+    constexpr const object_t* get_impl_ptr(const object_t* /*unused*/) const noexcept
+    {
+        return is_object() ? m_data.m_value.object : nullptr;
+    }
+
+    /// get a pointer to the value (array)
+    array_t* get_impl_ptr(array_t* /*unused*/) noexcept
+    {
+        return is_array() ? m_data.m_value.array : nullptr;
+    }
+
+    /// get a pointer to the value (array)
+    constexpr const array_t* get_impl_ptr(const array_t* /*unused*/) const noexcept
+    {
+        return is_array() ? m_data.m_value.array : nullptr;
+    }
+
+    /// get a pointer to the value (string)
+    string_t* get_impl_ptr(string_t* /*unused*/) noexcept
+    {
+        return is_string() ? m_data.m_value.string : nullptr;
+    }
+
+    /// get a pointer to the value (string)
+    constexpr const string_t* get_impl_ptr(const string_t* /*unused*/) const noexcept
+    {
+        return is_string() ? m_data.m_value.string : nullptr;
+    }
+
+    /// get a pointer to the value (boolean)
+    boolean_t* get_impl_ptr(boolean_t* /*unused*/) noexcept
+    {
+        return is_boolean() ? &m_data.m_value.boolean : nullptr;
+    }
+
+    /// get a pointer to the value (boolean)
+    constexpr const boolean_t* get_impl_ptr(const boolean_t* /*unused*/) const noexcept
+    {
+        return is_boolean() ? &m_data.m_value.boolean : nullptr;
+    }
+
+    /// get a pointer to the value (integer number)
+    number_integer_t* get_impl_ptr(number_integer_t* /*unused*/) noexcept
+    {
+        return m_data.m_type == value_t::number_integer ? &m_data.m_value.number_integer : nullptr;
+    }
+
+    /// get a pointer to the value (integer number)
+    constexpr const number_integer_t* get_impl_ptr(const number_integer_t* /*unused*/) const noexcept
+    {
+        return m_data.m_type == value_t::number_integer ? &m_data.m_value.number_integer : nullptr;
+    }
+
+    /// get a pointer to the value (unsigned number)
+    number_unsigned_t* get_impl_ptr(number_unsigned_t* /*unused*/) noexcept
+    {
+        return is_number_unsigned() ? &m_data.m_value.number_unsigned : nullptr;
+    }
+
+    /// get a pointer to the value (unsigned number)
+    constexpr const number_unsigned_t* get_impl_ptr(const number_unsigned_t* /*unused*/) const noexcept
+    {
+        return is_number_unsigned() ? &m_data.m_value.number_unsigned : nullptr;
+    }
+
+    /// get a pointer to the value (floating-point number)
+    number_float_t* get_impl_ptr(number_float_t* /*unused*/) noexcept
+    {
+        return is_number_float() ? &m_data.m_value.number_float : nullptr;
+    }
+
+    /// get a pointer to the value (floating-point number)
+    constexpr const number_float_t* get_impl_ptr(const number_float_t* /*unused*/) const noexcept
+    {
+        return is_number_float() ? &m_data.m_value.number_float : nullptr;
+    }
+
+    /// get a pointer to the value (binary)
+    binary_t* get_impl_ptr(binary_t* /*unused*/) noexcept
+    {
+        return is_binary() ? m_data.m_value.binary : nullptr;
+    }
+
+    /// get a pointer to the value (binary)
+    constexpr const binary_t* get_impl_ptr(const binary_t* /*unused*/) const noexcept
+    {
+        return is_binary() ? m_data.m_value.binary : nullptr;
+    }
+
+    /*!
+    @brief helper function to implement get_ref()
+
+    This function helps to implement get_ref() without code duplication for
+    const and non-const overloads
+
+    @tparam ThisType will be deduced as `basic_json` or `const basic_json`
+
+    @throw type_error.303 if ReferenceType does not match underlying value
+    type of the current JSON
+    */
+    template<typename ReferenceType, typename ThisType>
+    static ReferenceType get_ref_impl(ThisType& obj)
+    {
+        // delegate the call to get_ptr<>()
+        auto* ptr = obj.template get_ptr<typename std::add_pointer<ReferenceType>::type>();
+
+        if (JSON_HEDLEY_LIKELY(ptr != nullptr))
+        {
+            return *ptr;
+        }
+
+        JSON_THROW(type_error::create(303, detail::concat("incompatible ReferenceType for get_ref, actual type is ", obj.type_name()), &obj));
+    }
+
+  public:
+    /// @name value access
+    /// Direct access to the stored value of a JSON value.
+    /// @{
+
+    /// @brief get a pointer value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_ptr/
+    template<typename PointerType, typename std::enable_if<
+                 std::is_pointer<PointerType>::value, int>::type = 0>
+    auto get_ptr() noexcept -> decltype(std::declval<basic_json_t&>().get_impl_ptr(std::declval<PointerType>()))
+    {
+        // delegate the call to get_impl_ptr<>()
+        return get_impl_ptr(static_cast<PointerType>(nullptr));
+    }
+
+    /// @brief get a pointer value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_ptr/
+    template < typename PointerType, typename std::enable_if <
+                   std::is_pointer<PointerType>::value&&
+                   std::is_const<typename std::remove_pointer<PointerType>::type>::value, int >::type = 0 >
+    constexpr auto get_ptr() const noexcept -> decltype(std::declval<const basic_json_t&>().get_impl_ptr(std::declval<PointerType>()))
+    {
+        // delegate the call to get_impl_ptr<>() const
+        return get_impl_ptr(static_cast<PointerType>(nullptr));
+    }
+
+  private:
+    /*!
+    @brief get a value (explicit)
+
+    Explicit type conversion between the JSON value and a compatible value
+    which is [CopyConstructible](https://en.cppreference.com/w/cpp/named_req/CopyConstructible)
+    and [DefaultConstructible](https://en.cppreference.com/w/cpp/named_req/DefaultConstructible).
+    The value is converted by calling the @ref json_serializer<ValueType>
+    `from_json()` method.
+
+    The function is equivalent to executing
+    @code {.cpp}
+    ValueType ret;
+    JSONSerializer<ValueType>::from_json(*this, ret);
+    return ret;
+    @endcode
+
+    This overloads is chosen if:
+    - @a ValueType is not @ref basic_json,
+    - @ref json_serializer<ValueType> has a `from_json()` method of the form
+      `void from_json(const basic_json&, ValueType&)`, and
+    - @ref json_serializer<ValueType> does not have a `from_json()` method of
+      the form `ValueType from_json(const basic_json&)`
+
+    @tparam ValueType the returned value type
+
+    @return copy of the JSON value, converted to @a ValueType
+
+    @throw what @ref json_serializer<ValueType> `from_json()` method throws
+
+    @liveexample{The example below shows several conversions from JSON values
+    to other types. There a few things to note: (1) Floating-point numbers can
+    be converted to integers\, (2) A JSON array can be converted to a standard
+    `std::vector<short>`\, (3) A JSON object can be converted to C++
+    associative containers such as `std::unordered_map<std::string\,
+    json>`.,get__ValueType_const}
+
+    @since version 2.1.0
+    */
+    template < typename ValueType,
+               detail::enable_if_t <
+                   detail::is_default_constructible<ValueType>::value&&
+                   detail::has_from_json<basic_json_t, ValueType>::value,
+                   int > = 0 >
+    ValueType get_impl(detail::priority_tag<0> /*unused*/) const noexcept(noexcept(
+                JSONSerializer<ValueType>::from_json(std::declval<const basic_json_t&>(), std::declval<ValueType&>())))
+    {
+        auto ret = ValueType();
+        JSONSerializer<ValueType>::from_json(*this, ret);
+        return ret;
+    }
+
+    /*!
+    @brief get a value (explicit); special case
+
+    Explicit type conversion between the JSON value and a compatible value
+    which is **not** [CopyConstructible](https://en.cppreference.com/w/cpp/named_req/CopyConstructible)
+    and **not** [DefaultConstructible](https://en.cppreference.com/w/cpp/named_req/DefaultConstructible).
+    The value is converted by calling the @ref json_serializer<ValueType>
+    `from_json()` method.
+
+    The function is equivalent to executing
+    @code {.cpp}
+    return JSONSerializer<ValueType>::from_json(*this);
+    @endcode
+
+    This overloads is chosen if:
+    - @a ValueType is not @ref basic_json and
+    - @ref json_serializer<ValueType> has a `from_json()` method of the form
+      `ValueType from_json(const basic_json&)`
+
+    @note If @ref json_serializer<ValueType> has both overloads of
+    `from_json()`, this one is chosen.
+
+    @tparam ValueType the returned value type
+
+    @return copy of the JSON value, converted to @a ValueType
+
+    @throw what @ref json_serializer<ValueType> `from_json()` method throws
+
+    @since version 2.1.0
+    */
+    template < typename ValueType,
+               detail::enable_if_t <
+                   detail::has_non_default_from_json<basic_json_t, ValueType>::value,
+                   int > = 0 >
+    ValueType get_impl(detail::priority_tag<1> /*unused*/) const noexcept(noexcept(
+                JSONSerializer<ValueType>::from_json(std::declval<const basic_json_t&>())))
+    {
+        return JSONSerializer<ValueType>::from_json(*this);
+    }
+
+    /*!
+    @brief get special-case overload
+
+    This overloads converts the current @ref basic_json in a different
+    @ref basic_json type
+
+    @tparam BasicJsonType == @ref basic_json
+
+    @return a copy of *this, converted into @a BasicJsonType
+
+    @complexity Depending on the implementation of the called `from_json()`
+                method.
+
+    @since version 3.2.0
+    */
+    template < typename BasicJsonType,
+               detail::enable_if_t <
+                   detail::is_basic_json<BasicJsonType>::value,
+                   int > = 0 >
+    BasicJsonType get_impl(detail::priority_tag<2> /*unused*/) const
+    {
+        return *this;
+    }
+
+    /*!
+    @brief get special-case overload
+
+    This overloads avoids a lot of template boilerplate, it can be seen as the
+    identity method
+
+    @tparam BasicJsonType == @ref basic_json
+
+    @return a copy of *this
+
+    @complexity Constant.
+
+    @since version 2.1.0
+    */
+    template<typename BasicJsonType,
+             detail::enable_if_t<
+                 std::is_same<BasicJsonType, basic_json_t>::value,
+                 int> = 0>
+    basic_json get_impl(detail::priority_tag<3> /*unused*/) const
+    {
+        return *this;
+    }
+
+    /*!
+    @brief get a pointer value (explicit)
+    @copydoc get()
+    */
+    template<typename PointerType,
+             detail::enable_if_t<
+                 std::is_pointer<PointerType>::value,
+                 int> = 0>
+    constexpr auto get_impl(detail::priority_tag<4> /*unused*/) const noexcept
+    -> decltype(std::declval<const basic_json_t&>().template get_ptr<PointerType>())
+    {
+        // delegate the call to get_ptr
+        return get_ptr<PointerType>();
+    }
+
+  public:
+    /*!
+    @brief get a (pointer) value (explicit)
+
+    Performs explicit type conversion between the JSON value and a compatible value if required.
+
+    - If the requested type is a pointer to the internally stored JSON value that pointer is returned.
+    No copies are made.
+
+    - If the requested type is the current @ref basic_json, or a different @ref basic_json convertible
+    from the current @ref basic_json.
+
+    - Otherwise the value is converted by calling the @ref json_serializer<ValueType> `from_json()`
+    method.
+
+    @tparam ValueTypeCV the provided value type
+    @tparam ValueType the returned value type
+
+    @return copy of the JSON value, converted to @tparam ValueType if necessary
+
+    @throw what @ref json_serializer<ValueType> `from_json()` method throws if conversion is required
+
+    @since version 2.1.0
+    */
+    template < typename ValueTypeCV, typename ValueType = detail::uncvref_t<ValueTypeCV>>
+#if defined(JSON_HAS_CPP_14)
+    constexpr
+#endif
+    auto get() const noexcept(
+    noexcept(std::declval<const basic_json_t&>().template get_impl<ValueType>(detail::priority_tag<4> {})))
+    -> decltype(std::declval<const basic_json_t&>().template get_impl<ValueType>(detail::priority_tag<4> {}))
+    {
+        // we cannot static_assert on ValueTypeCV being non-const, because
+        // there is support for get<const basic_json_t>(), which is why we
+        // still need the uncvref
+        static_assert(!std::is_reference<ValueTypeCV>::value,
+                      "get() cannot be used with reference types, you might want to use get_ref()");
+        return get_impl<ValueType>(detail::priority_tag<4> {});
+    }
+
+    /*!
+    @brief get a pointer value (explicit)
+
+    Explicit pointer access to the internally stored JSON value. No copies are
+    made.
+
+    @warning The pointer becomes invalid if the underlying JSON object
+    changes.
+
+    @tparam PointerType pointer type; must be a pointer to @ref array_t, @ref
+    object_t, @ref string_t, @ref boolean_t, @ref number_integer_t,
+    @ref number_unsigned_t, or @ref number_float_t.
+
+    @return pointer to the internally stored JSON value if the requested
+    pointer type @a PointerType fits to the JSON value; `nullptr` otherwise
+
+    @complexity Constant.
+
+    @liveexample{The example below shows how pointers to internal values of a
+    JSON value can be requested. Note that no type conversions are made and a
+    `nullptr` is returned if the value and the requested pointer type does not
+    match.,get__PointerType}
+
+    @sa see @ref get_ptr() for explicit pointer-member access
+
+    @since version 1.0.0
+    */
+    template<typename PointerType, typename std::enable_if<
+                 std::is_pointer<PointerType>::value, int>::type = 0>
+    auto get() noexcept -> decltype(std::declval<basic_json_t&>().template get_ptr<PointerType>())
+    {
+        // delegate the call to get_ptr
+        return get_ptr<PointerType>();
+    }
+
+    /// @brief get a value (explicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_to/
+    template < typename ValueType,
+               detail::enable_if_t <
+                   !detail::is_basic_json<ValueType>::value&&
+                   detail::has_from_json<basic_json_t, ValueType>::value,
+                   int > = 0 >
+    ValueType & get_to(ValueType& v) const noexcept(noexcept(
+                JSONSerializer<ValueType>::from_json(std::declval<const basic_json_t&>(), v)))
+    {
+        JSONSerializer<ValueType>::from_json(*this, v);
+        return v;
+    }
+
+    // specialization to allow calling get_to with a basic_json value
+    // see https://github.com/nlohmann/json/issues/2175
+    template<typename ValueType,
+             detail::enable_if_t <
+                 detail::is_basic_json<ValueType>::value,
+                 int> = 0>
+    ValueType & get_to(ValueType& v) const
+    {
+        v = *this;
+        return v;
+    }
+
+    template <
+        typename T, std::size_t N,
+        typename Array = T (&)[N], // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+        detail::enable_if_t <
+            detail::has_from_json<basic_json_t, Array>::value, int > = 0 >
+    Array get_to(T (&v)[N]) const // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+    noexcept(noexcept(JSONSerializer<Array>::from_json(
+                          std::declval<const basic_json_t&>(), v)))
+    {
+        JSONSerializer<Array>::from_json(*this, v);
+        return v;
+    }
+
+    /// @brief get a reference value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_ref/
+    template<typename ReferenceType, typename std::enable_if<
+                 std::is_reference<ReferenceType>::value, int>::type = 0>
+    ReferenceType get_ref()
+    {
+        // delegate call to get_ref_impl
+        return get_ref_impl<ReferenceType>(*this);
+    }
+
+    /// @brief get a reference value (implicit)
+    /// @sa https://json.nlohmann.me/api/basic_json/get_ref/
+    template < typename ReferenceType, typename std::enable_if <
+                   std::is_reference<ReferenceType>::value&&
+                   std::is_const<typename std::remove_reference<ReferenceType>::type>::value, int >::type = 0 >
+    ReferenceType get_ref() const
+    {
+        // delegate call to get_ref_impl
+        return get_ref_impl<ReferenceType>(*this);
+    }
+
+    /*!
+    @brief get a value (implicit)
+
+    Implicit type conversion between the JSON value and a compatible value.
+    The call is realized by calling @ref get() const.
+
+    @tparam ValueType non-pointer type compatible to the JSON value, for
+    instance `int` for JSON integer numbers, `bool` for JSON booleans, or
+    `std::vector` types for JSON arrays. The character type of @ref string_t
+    as well as an initializer list of this type is excluded to avoid
+    ambiguities as these types implicitly convert to `std::string`.
+
+    @return copy of the JSON value, converted to type @a ValueType
+
+    @throw type_error.302 in case passed type @a ValueType is incompatible
+    to the JSON value type (e.g., the JSON value is of type boolean, but a
+    string is requested); see example below
+
+    @complexity Linear in the size of the JSON value.
+
+    @liveexample{The example below shows several conversions from JSON values
+    to other types. There a few things to note: (1) Floating-point numbers can
+    be converted to integers\, (2) A JSON array can be converted to a standard
+    `std::vector<short>`\, (3) A JSON object can be converted to C++
+    associative containers such as `std::unordered_map<std::string\,
+    json>`.,operator__ValueType}
+
+    @since version 1.0.0
+    */
+    template < typename ValueType, typename std::enable_if <
+                   detail::conjunction <
+                       detail::negation<std::is_pointer<ValueType>>,
+                       detail::negation<std::is_same<ValueType, std::nullptr_t>>,
+                       detail::negation<std::is_same<ValueType, detail::json_ref<basic_json>>>,
+                                        detail::negation<std::is_same<ValueType, typename string_t::value_type>>,
+                                        detail::negation<detail::is_basic_json<ValueType>>,
+                                        detail::negation<std::is_same<ValueType, std::initializer_list<typename string_t::value_type>>>,
+#if defined(JSON_HAS_CPP_17) && (defined(__GNUC__) || (defined(_MSC_VER) && _MSC_VER >= 1910 && _MSC_VER <= 1914))
+                                                detail::negation<std::is_same<ValueType, std::string_view>>,
+#endif
+#if defined(JSON_HAS_CPP_17) && JSON_HAS_STATIC_RTTI
+                                                detail::negation<std::is_same<ValueType, std::any>>,
+#endif
+                                                detail::is_detected_lazy<detail::get_template_function, const basic_json_t&, ValueType>
+                                                >::value, int >::type = 0 >
+                                        JSON_EXPLICIT operator ValueType() const
+    {
+        // delegate the call to get<>() const
+        return get<ValueType>();
+    }
+
+    /// @brief get a binary value
+    /// @sa https://json.nlohmann.me/api/basic_json/get_binary/
+    binary_t& get_binary()
+    {
+        if (!is_binary())
+        {
+            JSON_THROW(type_error::create(302, detail::concat("type must be binary, but is ", type_name()), this));
+        }
+
+        return *get_ptr<binary_t*>();
+    }
+
+    /// @brief get a binary value
+    /// @sa https://json.nlohmann.me/api/basic_json/get_binary/
+    const binary_t& get_binary() const
+    {
+        if (!is_binary())
+        {
+            JSON_THROW(type_error::create(302, detail::concat("type must be binary, but is ", type_name()), this));
+        }
+
+        return *get_ptr<const binary_t*>();
+    }
+
+    /// @}
+
+    ////////////////////
+    // element access //
+    ////////////////////
+
+    /// @name element access
+    /// Access to the JSON value.
+    /// @{
+
+    /// @brief access specified array element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    reference at(size_type idx)
+    {
+        // at only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            JSON_TRY
+            {
+                return set_parent(m_data.m_value.array->at(idx));
+            }
+            JSON_CATCH (std::out_of_range&)
+            {
+                // create better exception explanation
+                JSON_THROW(out_of_range::create(401, detail::concat("array index ", std::to_string(idx), " is out of range"), this));
+            } // cppcheck-suppress[missingReturn]
+        }
+        else
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+    }
+
+    /// @brief access specified array element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    const_reference at(size_type idx) const
+    {
+        // at only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            JSON_TRY
+            {
+                return m_data.m_value.array->at(idx);
+            }
+            JSON_CATCH (std::out_of_range&)
+            {
+                // create better exception explanation
+                JSON_THROW(out_of_range::create(401, detail::concat("array index ", std::to_string(idx), " is out of range"), this));
+            } // cppcheck-suppress[missingReturn]
+        }
+        else
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+    }
+
+    /// @brief access specified object element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    reference at(const typename object_t::key_type& key)
+    {
+        // at only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+
+        auto it = m_data.m_value.object->find(key);
+        if (it == m_data.m_value.object->end())
+        {
+            JSON_THROW(out_of_range::create(403, detail::concat("key '", key, "' not found"), this));
+        }
+        return set_parent(it->second);
+    }
+
+    /// @brief access specified object element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    reference at(KeyType && key)
+    {
+        // at only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+
+        auto it = m_data.m_value.object->find(std::forward<KeyType>(key));
+        if (it == m_data.m_value.object->end())
+        {
+            JSON_THROW(out_of_range::create(403, detail::concat("key '", string_t(std::forward<KeyType>(key)), "' not found"), this));
+        }
+        return set_parent(it->second);
+    }
+
+    /// @brief access specified object element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    const_reference at(const typename object_t::key_type& key) const
+    {
+        // at only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+
+        auto it = m_data.m_value.object->find(key);
+        if (it == m_data.m_value.object->end())
+        {
+            JSON_THROW(out_of_range::create(403, detail::concat("key '", key, "' not found"), this));
+        }
+        return it->second;
+    }
+
+    /// @brief access specified object element with bounds checking
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    const_reference at(KeyType && key) const
+    {
+        // at only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(304, detail::concat("cannot use at() with ", type_name()), this));
+        }
+
+        auto it = m_data.m_value.object->find(std::forward<KeyType>(key));
+        if (it == m_data.m_value.object->end())
+        {
+            JSON_THROW(out_of_range::create(403, detail::concat("key '", string_t(std::forward<KeyType>(key)), "' not found"), this));
+        }
+        return it->second;
+    }
+
+    /// @brief access specified array element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    reference operator[](size_type idx)
+    {
+        // implicitly convert null value to an empty array
+        if (is_null())
+        {
+            m_data.m_type = value_t::array;
+            m_data.m_value.array = create<array_t>();
+            assert_invariant();
+        }
+
+        // operator[] only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            // fill up array with null values if given idx is outside range
+            if (idx >= m_data.m_value.array->size())
+            {
+#if JSON_DIAGNOSTICS
+                // remember array size & capacity before resizing
+                const auto old_size = m_data.m_value.array->size();
+                const auto old_capacity = m_data.m_value.array->capacity();
+#endif
+                m_data.m_value.array->resize(idx + 1);
+
+#if JSON_DIAGNOSTICS
+                if (JSON_HEDLEY_UNLIKELY(m_data.m_value.array->capacity() != old_capacity))
+                {
+                    // capacity has changed: update all parents
+                    set_parents();
+                }
+                else
+                {
+                    // set parent for values added above
+                    set_parents(begin() + static_cast<typename iterator::difference_type>(old_size), static_cast<typename iterator::difference_type>(idx + 1 - old_size));
+                }
+#endif
+                assert_invariant();
+            }
+
+            return m_data.m_value.array->operator[](idx);
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a numeric argument with ", type_name()), this));
+    }
+
+    /// @brief access specified array element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    const_reference operator[](size_type idx) const
+    {
+        // const operator[] only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            return m_data.m_value.array->operator[](idx);
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a numeric argument with ", type_name()), this));
+    }
+
+    /// @brief access specified object element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    reference operator[](typename object_t::key_type key) // NOLINT(performance-unnecessary-value-param)
+    {
+        // implicitly convert null value to an empty object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value.object = create<object_t>();
+            assert_invariant();
+        }
+
+        // operator[] only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            auto result = m_data.m_value.object->emplace(std::move(key), nullptr);
+            return set_parent(result.first->second);
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a string argument with ", type_name()), this));
+    }
+
+    /// @brief access specified object element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    const_reference operator[](const typename object_t::key_type& key) const
+    {
+        // const operator[] only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            auto it = m_data.m_value.object->find(key);
+            JSON_ASSERT(it != m_data.m_value.object->end());
+            return it->second;
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a string argument with ", type_name()), this));
+    }
+
+    // these two functions resolve a (const) char * ambiguity affecting Clang and MSVC
+    // (they seemingly cannot be constrained to resolve the ambiguity)
+    template<typename T>
+    reference operator[](T* key)
+    {
+        return operator[](typename object_t::key_type(key));
+    }
+
+    template<typename T>
+    const_reference operator[](T* key) const
+    {
+        return operator[](typename object_t::key_type(key));
+    }
+
+    /// @brief access specified object element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int > = 0 >
+    reference operator[](KeyType && key)
+    {
+        // implicitly convert null value to an empty object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value.object = create<object_t>();
+            assert_invariant();
+        }
+
+        // operator[] only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            auto result = m_data.m_value.object->emplace(std::forward<KeyType>(key), nullptr);
+            return set_parent(result.first->second);
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a string argument with ", type_name()), this));
+    }
+
+    /// @brief access specified object element
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int > = 0 >
+    const_reference operator[](KeyType && key) const
+    {
+        // const operator[] only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            auto it = m_data.m_value.object->find(std::forward<KeyType>(key));
+            JSON_ASSERT(it != m_data.m_value.object->end());
+            return it->second;
+        }
+
+        JSON_THROW(type_error::create(305, detail::concat("cannot use operator[] with a string argument with ", type_name()), this));
+    }
+
+  private:
+    template<typename KeyType>
+    using is_comparable_with_object_key = detail::is_comparable <
+        object_comparator_t, const typename object_t::key_type&, KeyType >;
+
+    template<typename ValueType>
+    using value_return_type = std::conditional <
+        detail::is_c_string_uncvref<ValueType>::value,
+        string_t, typename std::decay<ValueType>::type >;
+
+  public:
+    /// @brief access specified object element with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, detail::enable_if_t <
+                   !detail::is_transparent<object_comparator_t>::value
+                   && detail::is_getable<basic_json_t, ValueType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ValueType value(const typename object_t::key_type& key, const ValueType& default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if key is found, return value and given default value otherwise
+            const auto it = find(key);
+            if (it != end())
+            {
+                return it->template get<ValueType>();
+            }
+
+            return default_value;
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, class ReturnType = typename value_return_type<ValueType>::type,
+               detail::enable_if_t <
+                   !detail::is_transparent<object_comparator_t>::value
+                   && detail::is_getable<basic_json_t, ReturnType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ReturnType value(const typename object_t::key_type& key, ValueType && default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if key is found, return value and given default value otherwise
+            const auto it = find(key);
+            if (it != end())
+            {
+                return it->template get<ReturnType>();
+            }
+
+            return std::forward<ValueType>(default_value);
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, class KeyType, detail::enable_if_t <
+                   detail::is_transparent<object_comparator_t>::value
+                   && !detail::is_json_pointer<KeyType>::value
+                   && is_comparable_with_object_key<KeyType>::value
+                   && detail::is_getable<basic_json_t, ValueType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ValueType value(KeyType && key, const ValueType& default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if key is found, return value and given default value otherwise
+            const auto it = find(std::forward<KeyType>(key));
+            if (it != end())
+            {
+                return it->template get<ValueType>();
+            }
+
+            return default_value;
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element via JSON Pointer with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, class KeyType, class ReturnType = typename value_return_type<ValueType>::type,
+               detail::enable_if_t <
+                   detail::is_transparent<object_comparator_t>::value
+                   && !detail::is_json_pointer<KeyType>::value
+                   && is_comparable_with_object_key<KeyType>::value
+                   && detail::is_getable<basic_json_t, ReturnType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ReturnType value(KeyType && key, ValueType && default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if key is found, return value and given default value otherwise
+            const auto it = find(std::forward<KeyType>(key));
+            if (it != end())
+            {
+                return it->template get<ReturnType>();
+            }
+
+            return std::forward<ValueType>(default_value);
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element via JSON Pointer with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, detail::enable_if_t <
+                   detail::is_getable<basic_json_t, ValueType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ValueType value(const json_pointer& ptr, const ValueType& default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if pointer resolves a value, return it or use default value
+            JSON_TRY
+            {
+                return ptr.get_checked(this).template get<ValueType>();
+            }
+            JSON_INTERNAL_CATCH (out_of_range&)
+            {
+                return default_value;
+            }
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    /// @brief access specified object element via JSON Pointer with default value
+    /// @sa https://json.nlohmann.me/api/basic_json/value/
+    template < class ValueType, class ReturnType = typename value_return_type<ValueType>::type,
+               detail::enable_if_t <
+                   detail::is_getable<basic_json_t, ReturnType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    ReturnType value(const json_pointer& ptr, ValueType && default_value) const
+    {
+        // value only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            // if pointer resolves a value, return it or use default value
+            JSON_TRY
+            {
+                return ptr.get_checked(this).template get<ReturnType>();
+            }
+            JSON_INTERNAL_CATCH (out_of_range&)
+            {
+                return std::forward<ValueType>(default_value);
+            }
+        }
+
+        JSON_THROW(type_error::create(306, detail::concat("cannot use value() with ", type_name()), this));
+    }
+
+    template < class ValueType, class BasicJsonType, detail::enable_if_t <
+                   detail::is_basic_json<BasicJsonType>::value
+                   && detail::is_getable<basic_json_t, ValueType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    ValueType value(const ::nlohmann::json_pointer<BasicJsonType>& ptr, const ValueType& default_value) const
+    {
+        return value(ptr.convert(), default_value);
+    }
+
+    template < class ValueType, class BasicJsonType, class ReturnType = typename value_return_type<ValueType>::type,
+               detail::enable_if_t <
+                   detail::is_basic_json<BasicJsonType>::value
+                   && detail::is_getable<basic_json_t, ReturnType>::value
+                   && !std::is_same<value_t, detail::uncvref_t<ValueType>>::value, int > = 0 >
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    ReturnType value(const ::nlohmann::json_pointer<BasicJsonType>& ptr, ValueType && default_value) const
+    {
+        return value(ptr.convert(), std::forward<ValueType>(default_value));
+    }
+
+    /// @brief access the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/front/
+    reference front()
+    {
+        return *begin();
+    }
+
+    /// @brief access the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/front/
+    const_reference front() const
+    {
+        return *cbegin();
+    }
+
+    /// @brief access the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/back/
+    reference back()
+    {
+        auto tmp = end();
+        --tmp;
+        return *tmp;
+    }
+
+    /// @brief access the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/back/
+    const_reference back() const
+    {
+        auto tmp = cend();
+        --tmp;
+        return *tmp;
+    }
+
+    /// @brief remove element given an iterator
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    template < class IteratorType, detail::enable_if_t <
+                   std::is_same<IteratorType, typename basic_json_t::iterator>::value ||
+                   std::is_same<IteratorType, typename basic_json_t::const_iterator>::value, int > = 0 >
+    IteratorType erase(IteratorType pos) // NOLINT(performance-unnecessary-value-param)
+    {
+        // make sure iterator fits the current value
+        if (JSON_HEDLEY_UNLIKELY(this != pos.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+        }
+
+        IteratorType result = end();
+
+        switch (m_data.m_type)
+        {
+            case value_t::boolean:
+            case value_t::number_float:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::string:
+            case value_t::binary:
+            {
+                if (JSON_HEDLEY_UNLIKELY(!pos.m_it.primitive_iterator.is_begin()))
+                {
+                    JSON_THROW(invalid_iterator::create(205, "iterator out of range", this));
+                }
+
+                if (is_string())
+                {
+                    AllocatorType<string_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, m_data.m_value.string);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, m_data.m_value.string, 1);
+                    m_data.m_value.string = nullptr;
+                }
+                else if (is_binary())
+                {
+                    AllocatorType<binary_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, m_data.m_value.binary);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, m_data.m_value.binary, 1);
+                    m_data.m_value.binary = nullptr;
+                }
+
+                m_data.m_type = value_t::null;
+                assert_invariant();
+                break;
+            }
+
+            case value_t::object:
+            {
+                result.m_it.object_iterator = m_data.m_value.object->erase(pos.m_it.object_iterator);
+                break;
+            }
+
+            case value_t::array:
+            {
+                result.m_it.array_iterator = m_data.m_value.array->erase(pos.m_it.array_iterator);
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+
+        return result;
+    }
+
+    /// @brief remove elements given an iterator range
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    template < class IteratorType, detail::enable_if_t <
+                   std::is_same<IteratorType, typename basic_json_t::iterator>::value ||
+                   std::is_same<IteratorType, typename basic_json_t::const_iterator>::value, int > = 0 >
+    IteratorType erase(IteratorType first, IteratorType last) // NOLINT(performance-unnecessary-value-param)
+    {
+        // make sure iterator fits the current value
+        if (JSON_HEDLEY_UNLIKELY(this != first.m_object || this != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(203, "iterators do not fit current value", this));
+        }
+
+        IteratorType result = end();
+
+        switch (m_data.m_type)
+        {
+            case value_t::boolean:
+            case value_t::number_float:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::string:
+            case value_t::binary:
+            {
+                if (JSON_HEDLEY_LIKELY(!first.m_it.primitive_iterator.is_begin()
+                                       || !last.m_it.primitive_iterator.is_end()))
+                {
+                    JSON_THROW(invalid_iterator::create(204, "iterators out of range", this));
+                }
+
+                if (is_string())
+                {
+                    AllocatorType<string_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, m_data.m_value.string);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, m_data.m_value.string, 1);
+                    m_data.m_value.string = nullptr;
+                }
+                else if (is_binary())
+                {
+                    AllocatorType<binary_t> alloc;
+                    std::allocator_traits<decltype(alloc)>::destroy(alloc, m_data.m_value.binary);
+                    std::allocator_traits<decltype(alloc)>::deallocate(alloc, m_data.m_value.binary, 1);
+                    m_data.m_value.binary = nullptr;
+                }
+
+                m_data.m_type = value_t::null;
+                assert_invariant();
+                break;
+            }
+
+            case value_t::object:
+            {
+                result.m_it.object_iterator = m_data.m_value.object->erase(first.m_it.object_iterator,
+                                              last.m_it.object_iterator);
+                break;
+            }
+
+            case value_t::array:
+            {
+                result.m_it.array_iterator = m_data.m_value.array->erase(first.m_it.array_iterator,
+                                             last.m_it.array_iterator);
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+
+        return result;
+    }
+
+  private:
+    template < typename KeyType, detail::enable_if_t <
+                   detail::has_erase_with_key_type<basic_json_t, KeyType>::value, int > = 0 >
+    size_type erase_internal(KeyType && key)
+    {
+        // this erase only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+
+        return m_data.m_value.object->erase(std::forward<KeyType>(key));
+    }
+
+    template < typename KeyType, detail::enable_if_t <
+                   !detail::has_erase_with_key_type<basic_json_t, KeyType>::value, int > = 0 >
+    size_type erase_internal(KeyType && key)
+    {
+        // this erase only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+
+        const auto it = m_data.m_value.object->find(std::forward<KeyType>(key));
+        if (it != m_data.m_value.object->end())
+        {
+            m_data.m_value.object->erase(it);
+            return 1;
+        }
+        return 0;
+    }
+
+  public:
+
+    /// @brief remove element from a JSON object given a key
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    size_type erase(const typename object_t::key_type& key)
+    {
+        // the indirection via erase_internal() is added to avoid making this
+        // function a template and thus de-rank it during overload resolution
+        return erase_internal(key);
+    }
+
+    /// @brief remove element from a JSON object given a key
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    size_type erase(KeyType && key)
+    {
+        return erase_internal(std::forward<KeyType>(key));
+    }
+
+    /// @brief remove element from a JSON array given an index
+    /// @sa https://json.nlohmann.me/api/basic_json/erase/
+    void erase(const size_type idx)
+    {
+        // this erase only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            if (JSON_HEDLEY_UNLIKELY(idx >= size()))
+            {
+                JSON_THROW(out_of_range::create(401, detail::concat("array index ", std::to_string(idx), " is out of range"), this));
+            }
+
+            m_data.m_value.array->erase(m_data.m_value.array->begin() + static_cast<difference_type>(idx));
+        }
+        else
+        {
+            JSON_THROW(type_error::create(307, detail::concat("cannot use erase() with ", type_name()), this));
+        }
+    }
+
+    /// @}
+
+    ////////////
+    // lookup //
+    ////////////
+
+    /// @name lookup
+    /// @{
+
+    /// @brief find an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/find/
+    iterator find(const typename object_t::key_type& key)
+    {
+        auto result = end();
+
+        if (is_object())
+        {
+            result.m_it.object_iterator = m_data.m_value.object->find(key);
+        }
+
+        return result;
+    }
+
+    /// @brief find an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/find/
+    const_iterator find(const typename object_t::key_type& key) const
+    {
+        auto result = cend();
+
+        if (is_object())
+        {
+            result.m_it.object_iterator = m_data.m_value.object->find(key);
+        }
+
+        return result;
+    }
+
+    /// @brief find an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/find/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    iterator find(KeyType && key)
+    {
+        auto result = end();
+
+        if (is_object())
+        {
+            result.m_it.object_iterator = m_data.m_value.object->find(std::forward<KeyType>(key));
+        }
+
+        return result;
+    }
+
+    /// @brief find an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/find/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    const_iterator find(KeyType && key) const
+    {
+        auto result = cend();
+
+        if (is_object())
+        {
+            result.m_it.object_iterator = m_data.m_value.object->find(std::forward<KeyType>(key));
+        }
+
+        return result;
+    }
+
+    /// @brief returns the number of occurrences of a key in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/count/
+    size_type count(const typename object_t::key_type& key) const
+    {
+        // return 0 for all nonobject types
+        return is_object() ? m_data.m_value.object->count(key) : 0;
+    }
+
+    /// @brief returns the number of occurrences of a key in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/count/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    size_type count(KeyType && key) const
+    {
+        // return 0 for all nonobject types
+        return is_object() ? m_data.m_value.object->count(std::forward<KeyType>(key)) : 0;
+    }
+
+    /// @brief check the existence of an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/contains/
+    bool contains(const typename object_t::key_type& key) const
+    {
+        return is_object() && m_data.m_value.object->find(key) != m_data.m_value.object->end();
+    }
+
+    /// @brief check the existence of an element in a JSON object
+    /// @sa https://json.nlohmann.me/api/basic_json/contains/
+    template<class KeyType, detail::enable_if_t<
+                 detail::is_usable_as_basic_json_key_type<basic_json_t, KeyType>::value, int> = 0>
+    bool contains(KeyType && key) const
+    {
+        return is_object() && m_data.m_value.object->find(std::forward<KeyType>(key)) != m_data.m_value.object->end();
+    }
+
+    /// @brief check the existence of an element in a JSON object given a JSON pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/contains/
+    bool contains(const json_pointer& ptr) const
+    {
+        return ptr.contains(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    bool contains(const typename ::nlohmann::json_pointer<BasicJsonType>& ptr) const
+    {
+        return ptr.contains(this);
+    }
+
+    /// @}
+
+    ///////////////
+    // iterators //
+    ///////////////
+
+    /// @name iterators
+    /// @{
+
+    /// @brief returns an iterator to the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/begin/
+    iterator begin() noexcept
+    {
+        iterator result(this);
+        result.set_begin();
+        return result;
+    }
+
+    /// @brief returns an iterator to the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/begin/
+    const_iterator begin() const noexcept
+    {
+        return cbegin();
+    }
+
+    /// @brief returns a const iterator to the first element
+    /// @sa https://json.nlohmann.me/api/basic_json/cbegin/
+    const_iterator cbegin() const noexcept
+    {
+        const_iterator result(this);
+        result.set_begin();
+        return result;
+    }
+
+    /// @brief returns an iterator to one past the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/end/
+    iterator end() noexcept
+    {
+        iterator result(this);
+        result.set_end();
+        return result;
+    }
+
+    /// @brief returns an iterator to one past the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/end/
+    const_iterator end() const noexcept
+    {
+        return cend();
+    }
+
+    /// @brief returns an iterator to one past the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/cend/
+    const_iterator cend() const noexcept
+    {
+        const_iterator result(this);
+        result.set_end();
+        return result;
+    }
+
+    /// @brief returns an iterator to the reverse-beginning
+    /// @sa https://json.nlohmann.me/api/basic_json/rbegin/
+    reverse_iterator rbegin() noexcept
+    {
+        return reverse_iterator(end());
+    }
+
+    /// @brief returns an iterator to the reverse-beginning
+    /// @sa https://json.nlohmann.me/api/basic_json/rbegin/
+    const_reverse_iterator rbegin() const noexcept
+    {
+        return crbegin();
+    }
+
+    /// @brief returns an iterator to the reverse-end
+    /// @sa https://json.nlohmann.me/api/basic_json/rend/
+    reverse_iterator rend() noexcept
+    {
+        return reverse_iterator(begin());
+    }
+
+    /// @brief returns an iterator to the reverse-end
+    /// @sa https://json.nlohmann.me/api/basic_json/rend/
+    const_reverse_iterator rend() const noexcept
+    {
+        return crend();
+    }
+
+    /// @brief returns a const reverse iterator to the last element
+    /// @sa https://json.nlohmann.me/api/basic_json/crbegin/
+    const_reverse_iterator crbegin() const noexcept
+    {
+        return const_reverse_iterator(cend());
+    }
+
+    /// @brief returns a const reverse iterator to one before the first
+    /// @sa https://json.nlohmann.me/api/basic_json/crend/
+    const_reverse_iterator crend() const noexcept
+    {
+        return const_reverse_iterator(cbegin());
+    }
+
+  public:
+    /// @brief wrapper to access iterator member functions in range-based for
+    /// @sa https://json.nlohmann.me/api/basic_json/items/
+    /// @deprecated This function is deprecated since 3.1.0 and will be removed in
+    ///             version 4.0.0 of the library. Please use @ref items() instead;
+    ///             that is, replace `json::iterator_wrapper(j)` with `j.items()`.
+    JSON_HEDLEY_DEPRECATED_FOR(3.1.0, items())
+    static iteration_proxy<iterator> iterator_wrapper(reference ref) noexcept
+    {
+        return ref.items();
+    }
+
+    /// @brief wrapper to access iterator member functions in range-based for
+    /// @sa https://json.nlohmann.me/api/basic_json/items/
+    /// @deprecated This function is deprecated since 3.1.0 and will be removed in
+    ///         version 4.0.0 of the library. Please use @ref items() instead;
+    ///         that is, replace `json::iterator_wrapper(j)` with `j.items()`.
+    JSON_HEDLEY_DEPRECATED_FOR(3.1.0, items())
+    static iteration_proxy<const_iterator> iterator_wrapper(const_reference ref) noexcept
+    {
+        return ref.items();
+    }
+
+    /// @brief helper to access iterator member functions in range-based for
+    /// @sa https://json.nlohmann.me/api/basic_json/items/
+    iteration_proxy<iterator> items() noexcept
+    {
+        return iteration_proxy<iterator>(*this);
+    }
+
+    /// @brief helper to access iterator member functions in range-based for
+    /// @sa https://json.nlohmann.me/api/basic_json/items/
+    iteration_proxy<const_iterator> items() const noexcept
+    {
+        return iteration_proxy<const_iterator>(*this);
+    }
+
+    /// @}
+
+    //////////////
+    // capacity //
+    //////////////
+
+    /// @name capacity
+    /// @{
+
+    /// @brief checks whether the container is empty.
+    /// @sa https://json.nlohmann.me/api/basic_json/empty/
+    bool empty() const noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::null:
+            {
+                // null values are empty
+                return true;
+            }
+
+            case value_t::array:
+            {
+                // delegate call to array_t::empty()
+                return m_data.m_value.array->empty();
+            }
+
+            case value_t::object:
+            {
+                // delegate call to object_t::empty()
+                return m_data.m_value.object->empty();
+            }
+
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                // all other types are nonempty
+                return false;
+            }
+        }
+    }
+
+    /// @brief returns the number of elements
+    /// @sa https://json.nlohmann.me/api/basic_json/size/
+    size_type size() const noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::null:
+            {
+                // null values are empty
+                return 0;
+            }
+
+            case value_t::array:
+            {
+                // delegate call to array_t::size()
+                return m_data.m_value.array->size();
+            }
+
+            case value_t::object:
+            {
+                // delegate call to object_t::size()
+                return m_data.m_value.object->size();
+            }
+
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                // all other types have size 1
+                return 1;
+            }
+        }
+    }
+
+    /// @brief returns the maximum possible number of elements
+    /// @sa https://json.nlohmann.me/api/basic_json/max_size/
+    size_type max_size() const noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::array:
+            {
+                // delegate call to array_t::max_size()
+                return m_data.m_value.array->max_size();
+            }
+
+            case value_t::object:
+            {
+                // delegate call to object_t::max_size()
+                return m_data.m_value.object->max_size();
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                // all other types have max_size() == size()
+                return size();
+            }
+        }
+    }
+
+    /// @}
+
+    ///////////////
+    // modifiers //
+    ///////////////
+
+    /// @name modifiers
+    /// @{
+
+    /// @brief clears the contents
+    /// @sa https://json.nlohmann.me/api/basic_json/clear/
+    void clear() noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::number_integer:
+            {
+                m_data.m_value.number_integer = 0;
+                break;
+            }
+
+            case value_t::number_unsigned:
+            {
+                m_data.m_value.number_unsigned = 0;
+                break;
+            }
+
+            case value_t::number_float:
+            {
+                m_data.m_value.number_float = 0.0;
+                break;
+            }
+
+            case value_t::boolean:
+            {
+                m_data.m_value.boolean = false;
+                break;
+            }
+
+            case value_t::string:
+            {
+                m_data.m_value.string->clear();
+                break;
+            }
+
+            case value_t::binary:
+            {
+                m_data.m_value.binary->clear();
+                break;
+            }
+
+            case value_t::array:
+            {
+                m_data.m_value.array->clear();
+                break;
+            }
+
+            case value_t::object:
+            {
+                m_data.m_value.object->clear();
+                break;
+            }
+
+            case value_t::null:
+            case value_t::discarded:
+            default:
+                break;
+        }
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/push_back/
+    void push_back(basic_json&& val)
+    {
+        // push_back only works for null objects or arrays
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_array())))
+        {
+            JSON_THROW(type_error::create(308, detail::concat("cannot use push_back() with ", type_name()), this));
+        }
+
+        // transform null object into an array
+        if (is_null())
+        {
+            m_data.m_type = value_t::array;
+            m_data.m_value = value_t::array;
+            assert_invariant();
+        }
+
+        // add element to array (move semantics)
+        const auto old_capacity = m_data.m_value.array->capacity();
+        m_data.m_value.array->push_back(std::move(val));
+        set_parent(m_data.m_value.array->back(), old_capacity);
+        // if val is moved from, basic_json move constructor marks it null, so we do not call the destructor
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/operator+=/
+    reference operator+=(basic_json&& val)
+    {
+        push_back(std::move(val));
+        return *this;
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/push_back/
+    void push_back(const basic_json& val)
+    {
+        // push_back only works for null objects or arrays
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_array())))
+        {
+            JSON_THROW(type_error::create(308, detail::concat("cannot use push_back() with ", type_name()), this));
+        }
+
+        // transform null object into an array
+        if (is_null())
+        {
+            m_data.m_type = value_t::array;
+            m_data.m_value = value_t::array;
+            assert_invariant();
+        }
+
+        // add element to array
+        const auto old_capacity = m_data.m_value.array->capacity();
+        m_data.m_value.array->push_back(val);
+        set_parent(m_data.m_value.array->back(), old_capacity);
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/operator+=/
+    reference operator+=(const basic_json& val)
+    {
+        push_back(val);
+        return *this;
+    }
+
+    /// @brief add an object to an object
+    /// @sa https://json.nlohmann.me/api/basic_json/push_back/
+    void push_back(const typename object_t::value_type& val)
+    {
+        // push_back only works for null objects or objects
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_object())))
+        {
+            JSON_THROW(type_error::create(308, detail::concat("cannot use push_back() with ", type_name()), this));
+        }
+
+        // transform null object into an object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value = value_t::object;
+            assert_invariant();
+        }
+
+        // add element to object
+        auto res = m_data.m_value.object->insert(val);
+        set_parent(res.first->second);
+    }
+
+    /// @brief add an object to an object
+    /// @sa https://json.nlohmann.me/api/basic_json/operator+=/
+    reference operator+=(const typename object_t::value_type& val)
+    {
+        push_back(val);
+        return *this;
+    }
+
+    /// @brief add an object to an object
+    /// @sa https://json.nlohmann.me/api/basic_json/push_back/
+    void push_back(initializer_list_t init)
+    {
+        if (is_object() && init.size() == 2 && (*init.begin())->is_string())
+        {
+            basic_json&& key = init.begin()->moved_or_copied();
+            push_back(typename object_t::value_type(
+                          std::move(key.get_ref<string_t&>()), (init.begin() + 1)->moved_or_copied()));
+        }
+        else
+        {
+            push_back(basic_json(init));
+        }
+    }
+
+    /// @brief add an object to an object
+    /// @sa https://json.nlohmann.me/api/basic_json/operator+=/
+    reference operator+=(initializer_list_t init)
+    {
+        push_back(init);
+        return *this;
+    }
+
+    /// @brief add an object to an array
+    /// @sa https://json.nlohmann.me/api/basic_json/emplace_back/
+    template<class... Args>
+    reference emplace_back(Args&& ... args)
+    {
+        // emplace_back only works for null objects or arrays
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_array())))
+        {
+            JSON_THROW(type_error::create(311, detail::concat("cannot use emplace_back() with ", type_name()), this));
+        }
+
+        // transform null object into an array
+        if (is_null())
+        {
+            m_data.m_type = value_t::array;
+            m_data.m_value = value_t::array;
+            assert_invariant();
+        }
+
+        // add element to array (perfect forwarding)
+        const auto old_capacity = m_data.m_value.array->capacity();
+        m_data.m_value.array->emplace_back(std::forward<Args>(args)...);
+        return set_parent(m_data.m_value.array->back(), old_capacity);
+    }
+
+    /// @brief add an object to an object if key does not exist
+    /// @sa https://json.nlohmann.me/api/basic_json/emplace/
+    template<class... Args>
+    std::pair<iterator, bool> emplace(Args&& ... args)
+    {
+        // emplace only works for null objects or arrays
+        if (JSON_HEDLEY_UNLIKELY(!(is_null() || is_object())))
+        {
+            JSON_THROW(type_error::create(311, detail::concat("cannot use emplace() with ", type_name()), this));
+        }
+
+        // transform null object into an object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value = value_t::object;
+            assert_invariant();
+        }
+
+        // add element to array (perfect forwarding)
+        auto res = m_data.m_value.object->emplace(std::forward<Args>(args)...);
+        set_parent(res.first->second);
+
+        // create result iterator and set iterator to the result of emplace
+        auto it = begin();
+        it.m_it.object_iterator = res.first;
+
+        // return pair of iterator and boolean
+        return {it, res.second};
+    }
+
+    /// Helper for insertion of an iterator
+    /// @note: This uses std::distance to support GCC 4.8,
+    ///        see https://github.com/nlohmann/json/pull/1257
+    template<typename... Args>
+    iterator insert_iterator(const_iterator pos, Args&& ... args) // NOLINT(performance-unnecessary-value-param)
+    {
+        iterator result(this);
+        JSON_ASSERT(m_data.m_value.array != nullptr);
+
+        auto insert_pos = std::distance(m_data.m_value.array->begin(), pos.m_it.array_iterator);
+        m_data.m_value.array->insert(pos.m_it.array_iterator, std::forward<Args>(args)...);
+        result.m_it.array_iterator = m_data.m_value.array->begin() + insert_pos;
+
+        // This could have been written as:
+        // result.m_it.array_iterator = m_data.m_value.array->insert(pos.m_it.array_iterator, cnt, val);
+        // but the return value of insert is missing in GCC 4.8, so it is written this way instead.
+
+        set_parents();
+        return result;
+    }
+
+    /// @brief inserts element into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, const basic_json& val) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            // check if iterator pos fits to this JSON value
+            if (JSON_HEDLEY_UNLIKELY(pos.m_object != this))
+            {
+                JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+            }
+
+            // insert to array and return iterator
+            return insert_iterator(pos, val);
+        }
+
+        JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+    }
+
+    /// @brief inserts element into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, basic_json&& val) // NOLINT(performance-unnecessary-value-param)
+    {
+        return insert(pos, val);
+    }
+
+    /// @brief inserts copies of element into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, size_type cnt, const basic_json& val) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            // check if iterator pos fits to this JSON value
+            if (JSON_HEDLEY_UNLIKELY(pos.m_object != this))
+            {
+                JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+            }
+
+            // insert to array and return iterator
+            return insert_iterator(pos, cnt, val);
+        }
+
+        JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+    }
+
+    /// @brief inserts range of elements into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, const_iterator first, const_iterator last) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for arrays
+        if (JSON_HEDLEY_UNLIKELY(!is_array()))
+        {
+            JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+        }
+
+        // check if iterator pos fits to this JSON value
+        if (JSON_HEDLEY_UNLIKELY(pos.m_object != this))
+        {
+            JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+        }
+
+        // check if range iterators belong to the same JSON object
+        if (JSON_HEDLEY_UNLIKELY(first.m_object != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(210, "iterators do not fit", this));
+        }
+
+        if (JSON_HEDLEY_UNLIKELY(first.m_object == this))
+        {
+            JSON_THROW(invalid_iterator::create(211, "passed iterators may not belong to container", this));
+        }
+
+        // insert to array and return iterator
+        return insert_iterator(pos, first.m_it.array_iterator, last.m_it.array_iterator);
+    }
+
+    /// @brief inserts elements from initializer list into array
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    iterator insert(const_iterator pos, initializer_list_t ilist) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for arrays
+        if (JSON_HEDLEY_UNLIKELY(!is_array()))
+        {
+            JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+        }
+
+        // check if iterator pos fits to this JSON value
+        if (JSON_HEDLEY_UNLIKELY(pos.m_object != this))
+        {
+            JSON_THROW(invalid_iterator::create(202, "iterator does not fit current value", this));
+        }
+
+        // insert to array and return iterator
+        return insert_iterator(pos, ilist.begin(), ilist.end());
+    }
+
+    /// @brief inserts range of elements into object
+    /// @sa https://json.nlohmann.me/api/basic_json/insert/
+    void insert(const_iterator first, const_iterator last) // NOLINT(performance-unnecessary-value-param)
+    {
+        // insert only works for objects
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(309, detail::concat("cannot use insert() with ", type_name()), this));
+        }
+
+        // check if range iterators belong to the same JSON object
+        if (JSON_HEDLEY_UNLIKELY(first.m_object != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(210, "iterators do not fit", this));
+        }
+
+        // passed iterators must belong to objects
+        if (JSON_HEDLEY_UNLIKELY(!first.m_object->is_object()))
+        {
+            JSON_THROW(invalid_iterator::create(202, "iterators first and last must point to objects", this));
+        }
+
+        m_data.m_value.object->insert(first.m_it.object_iterator, last.m_it.object_iterator);
+        set_parents();
+    }
+
+    /// @brief updates a JSON object from another object, overwriting existing keys
+    /// @sa https://json.nlohmann.me/api/basic_json/update/
+    void update(const_reference j, bool merge_objects = false)
+    {
+        update(j.begin(), j.end(), merge_objects);
+    }
+
+    /// @brief updates a JSON object from another object, overwriting existing keys
+    /// @sa https://json.nlohmann.me/api/basic_json/update/
+    void update(const_iterator first, const_iterator last, bool merge_objects = false) // NOLINT(performance-unnecessary-value-param)
+    {
+        // implicitly convert null value to an empty object
+        if (is_null())
+        {
+            m_data.m_type = value_t::object;
+            m_data.m_value.object = create<object_t>();
+            assert_invariant();
+        }
+
+        if (JSON_HEDLEY_UNLIKELY(!is_object()))
+        {
+            JSON_THROW(type_error::create(312, detail::concat("cannot use update() with ", type_name()), this));
+        }
+
+        // check if range iterators belong to the same JSON object
+        if (JSON_HEDLEY_UNLIKELY(first.m_object != last.m_object))
+        {
+            JSON_THROW(invalid_iterator::create(210, "iterators do not fit", this));
+        }
+
+        // passed iterators must belong to objects
+        if (JSON_HEDLEY_UNLIKELY(!first.m_object->is_object()))
+        {
+            JSON_THROW(type_error::create(312, detail::concat("cannot use update() with ", first.m_object->type_name()), first.m_object));
+        }
+
+        for (auto it = first; it != last; ++it)
+        {
+            if (merge_objects && it.value().is_object())
+            {
+                auto it2 = m_data.m_value.object->find(it.key());
+                if (it2 != m_data.m_value.object->end())
+                {
+                    it2->second.update(it.value(), true);
+                    continue;
+                }
+            }
+            m_data.m_value.object->operator[](it.key()) = it.value();
+#if JSON_DIAGNOSTICS
+            m_data.m_value.object->operator[](it.key()).m_parent = this;
+#endif
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(reference other) noexcept (
+        std::is_nothrow_move_constructible<value_t>::value&&
+        std::is_nothrow_move_assignable<value_t>::value&&
+        std::is_nothrow_move_constructible<json_value>::value&& // NOLINT(cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+        std::is_nothrow_move_assignable<json_value>::value
+    )
+    {
+        std::swap(m_data.m_type, other.m_data.m_type);
+        std::swap(m_data.m_value, other.m_data.m_value);
+
+        set_parents();
+        other.set_parents();
+        assert_invariant();
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    friend void swap(reference left, reference right) noexcept (
+        std::is_nothrow_move_constructible<value_t>::value&&
+        std::is_nothrow_move_assignable<value_t>::value&&
+        std::is_nothrow_move_constructible<json_value>::value&& // NOLINT(cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+        std::is_nothrow_move_assignable<json_value>::value
+    )
+    {
+        left.swap(right);
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(array_t& other) // NOLINT(bugprone-exception-escape,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    {
+        // swap only works for arrays
+        if (JSON_HEDLEY_LIKELY(is_array()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.array), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(array_t&) with ", type_name()), this));
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(object_t& other) // NOLINT(bugprone-exception-escape,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    {
+        // swap only works for objects
+        if (JSON_HEDLEY_LIKELY(is_object()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.object), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(object_t&) with ", type_name()), this));
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(string_t& other) // NOLINT(bugprone-exception-escape,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    {
+        // swap only works for strings
+        if (JSON_HEDLEY_LIKELY(is_string()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.string), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(string_t&) with ", type_name()), this));
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(binary_t& other) // NOLINT(bugprone-exception-escape,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    {
+        // swap only works for strings
+        if (JSON_HEDLEY_LIKELY(is_binary()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.binary), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(binary_t&) with ", type_name()), this));
+        }
+    }
+
+    /// @brief exchanges the values
+    /// @sa https://json.nlohmann.me/api/basic_json/swap/
+    void swap(typename binary_t::container_type& other) // NOLINT(bugprone-exception-escape)
+    {
+        // swap only works for strings
+        if (JSON_HEDLEY_LIKELY(is_binary()))
+        {
+            using std::swap;
+            swap(*(m_data.m_value.binary), other);
+        }
+        else
+        {
+            JSON_THROW(type_error::create(310, detail::concat("cannot use swap(binary_t::container_type&) with ", type_name()), this));
+        }
+    }
+
+    /// @}
+
+    //////////////////////////////////////////
+    // lexicographical comparison operators //
+    //////////////////////////////////////////
+
+    /// @name lexicographical comparison operators
+    /// @{
+
+    // note parentheses around operands are necessary; see
+    // https://github.com/nlohmann/json/issues/1530
+#define JSON_IMPLEMENT_OPERATOR(op, null_result, unordered_result, default_result)                       \
+    const auto lhs_type = lhs.type();                                                                    \
+    const auto rhs_type = rhs.type();                                                                    \
+    \
+    if (lhs_type == rhs_type) /* NOLINT(readability/braces) */                                           \
+    {                                                                                                    \
+        switch (lhs_type)                                                                                \
+        {                                                                                                \
+            case value_t::array:                                                                         \
+                return (*lhs.m_data.m_value.array) op (*rhs.m_data.m_value.array);                                     \
+                \
+            case value_t::object:                                                                        \
+                return (*lhs.m_data.m_value.object) op (*rhs.m_data.m_value.object);                                   \
+                \
+            case value_t::null:                                                                          \
+                return (null_result);                                                                    \
+                \
+            case value_t::string:                                                                        \
+                return (*lhs.m_data.m_value.string) op (*rhs.m_data.m_value.string);                                   \
+                \
+            case value_t::boolean:                                                                       \
+                return (lhs.m_data.m_value.boolean) op (rhs.m_data.m_value.boolean);                                   \
+                \
+            case value_t::number_integer:                                                                \
+                return (lhs.m_data.m_value.number_integer) op (rhs.m_data.m_value.number_integer);                     \
+                \
+            case value_t::number_unsigned:                                                               \
+                return (lhs.m_data.m_value.number_unsigned) op (rhs.m_data.m_value.number_unsigned);                   \
+                \
+            case value_t::number_float:                                                                  \
+                return (lhs.m_data.m_value.number_float) op (rhs.m_data.m_value.number_float);                         \
+                \
+            case value_t::binary:                                                                        \
+                return (*lhs.m_data.m_value.binary) op (*rhs.m_data.m_value.binary);                                   \
+                \
+            case value_t::discarded:                                                                     \
+            default:                                                                                     \
+                return (unordered_result);                                                               \
+        }                                                                                                \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_integer && rhs_type == value_t::number_float)                   \
+    {                                                                                                    \
+        return static_cast<number_float_t>(lhs.m_data.m_value.number_integer) op rhs.m_data.m_value.number_float;      \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_float && rhs_type == value_t::number_integer)                   \
+    {                                                                                                    \
+        return lhs.m_data.m_value.number_float op static_cast<number_float_t>(rhs.m_data.m_value.number_integer);      \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_unsigned && rhs_type == value_t::number_float)                  \
+    {                                                                                                    \
+        return static_cast<number_float_t>(lhs.m_data.m_value.number_unsigned) op rhs.m_data.m_value.number_float;     \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_float && rhs_type == value_t::number_unsigned)                  \
+    {                                                                                                    \
+        return lhs.m_data.m_value.number_float op static_cast<number_float_t>(rhs.m_data.m_value.number_unsigned);     \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_unsigned && rhs_type == value_t::number_integer)                \
+    {                                                                                                    \
+        return static_cast<number_integer_t>(lhs.m_data.m_value.number_unsigned) op rhs.m_data.m_value.number_integer; \
+    }                                                                                                    \
+    else if (lhs_type == value_t::number_integer && rhs_type == value_t::number_unsigned)                \
+    {                                                                                                    \
+        return lhs.m_data.m_value.number_integer op static_cast<number_integer_t>(rhs.m_data.m_value.number_unsigned); \
+    }                                                                                                    \
+    else if(compares_unordered(lhs, rhs))\
+    {\
+        return (unordered_result);\
+    }\
+    \
+    return (default_result);
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    // returns true if:
+    // - any operand is NaN and the other operand is of number type
+    // - any operand is discarded
+    // in legacy mode, discarded values are considered ordered if
+    // an operation is computed as an odd number of inverses of others
+    static bool compares_unordered(const_reference lhs, const_reference rhs, bool inverse = false) noexcept
+    {
+        if ((lhs.is_number_float() && std::isnan(lhs.m_data.m_value.number_float) && rhs.is_number())
+                || (rhs.is_number_float() && std::isnan(rhs.m_data.m_value.number_float) && lhs.is_number()))
+        {
+            return true;
+        }
+#if JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON
+        return (lhs.is_discarded() || rhs.is_discarded()) && !inverse;
+#else
+        static_cast<void>(inverse);
+        return lhs.is_discarded() || rhs.is_discarded();
+#endif
+    }
+
+  private:
+    bool compares_unordered(const_reference rhs, bool inverse = false) const noexcept
+    {
+        return compares_unordered(*this, rhs, inverse);
+    }
+
+  public:
+#if JSON_HAS_THREE_WAY_COMPARISON
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    bool operator==(const_reference rhs) const noexcept
+    {
+#ifdef __GNUC__
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wfloat-equal"
+#endif
+        const_reference lhs = *this;
+        JSON_IMPLEMENT_OPERATOR( ==, true, false, false)
+#ifdef __GNUC__
+#pragma GCC diagnostic pop
+#endif
+    }
+
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    template<typename ScalarType>
+    requires std::is_scalar_v<ScalarType>
+    bool operator==(ScalarType rhs) const noexcept
+    {
+        return *this == basic_json(rhs);
+    }
+
+    /// @brief comparison: not equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ne/
+    bool operator!=(const_reference rhs) const noexcept
+    {
+        if (compares_unordered(rhs, true))
+        {
+            return false;
+        }
+        return !operator==(rhs);
+    }
+
+    /// @brief comparison: 3-way
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_spaceship/
+    std::partial_ordering operator<=>(const_reference rhs) const noexcept // *NOPAD*
+    {
+        const_reference lhs = *this;
+        // default_result is used if we cannot compare values. In that case,
+        // we compare types.
+        JSON_IMPLEMENT_OPERATOR(<=>, // *NOPAD*
+                                std::partial_ordering::equivalent,
+                                std::partial_ordering::unordered,
+                                lhs_type <=> rhs_type) // *NOPAD*
+    }
+
+    /// @brief comparison: 3-way
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_spaceship/
+    template<typename ScalarType>
+    requires std::is_scalar_v<ScalarType>
+    std::partial_ordering operator<=>(ScalarType rhs) const noexcept // *NOPAD*
+    {
+        return *this <=> basic_json(rhs); // *NOPAD*
+    }
+
+#if JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON
+    // all operators that are computed as an odd number of inverses of others
+    // need to be overloaded to emulate the legacy comparison behavior
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, undef JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON)
+    bool operator<=(const_reference rhs) const noexcept
+    {
+        if (compares_unordered(rhs, true))
+        {
+            return false;
+        }
+        return !(rhs < *this);
+    }
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    template<typename ScalarType>
+    requires std::is_scalar_v<ScalarType>
+    bool operator<=(ScalarType rhs) const noexcept
+    {
+        return *this <= basic_json(rhs);
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, undef JSON_USE_LEGACY_DISCARDED_VALUE_COMPARISON)
+    bool operator>=(const_reference rhs) const noexcept
+    {
+        if (compares_unordered(rhs, true))
+        {
+            return false;
+        }
+        return !(*this < rhs);
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    template<typename ScalarType>
+    requires std::is_scalar_v<ScalarType>
+    bool operator>=(ScalarType rhs) const noexcept
+    {
+        return *this >= basic_json(rhs);
+    }
+#endif
+#else
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    friend bool operator==(const_reference lhs, const_reference rhs) noexcept
+    {
+#ifdef __GNUC__
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wfloat-equal"
+#endif
+        JSON_IMPLEMENT_OPERATOR( ==, true, false, false)
+#ifdef __GNUC__
+#pragma GCC diagnostic pop
+#endif
+    }
+
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator==(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs == basic_json(rhs);
+    }
+
+    /// @brief comparison: equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_eq/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator==(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) == rhs;
+    }
+
+    /// @brief comparison: not equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ne/
+    friend bool operator!=(const_reference lhs, const_reference rhs) noexcept
+    {
+        if (compares_unordered(lhs, rhs, true))
+        {
+            return false;
+        }
+        return !(lhs == rhs);
+    }
+
+    /// @brief comparison: not equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ne/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator!=(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs != basic_json(rhs);
+    }
+
+    /// @brief comparison: not equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ne/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator!=(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) != rhs;
+    }
+
+    /// @brief comparison: less than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_lt/
+    friend bool operator<(const_reference lhs, const_reference rhs) noexcept
+    {
+        // default_result is used if we cannot compare values. In that case,
+        // we compare types. Note we have to call the operator explicitly,
+        // because MSVC has problems otherwise.
+        JSON_IMPLEMENT_OPERATOR( <, false, false, operator<(lhs_type, rhs_type))
+    }
+
+    /// @brief comparison: less than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_lt/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator<(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs < basic_json(rhs);
+    }
+
+    /// @brief comparison: less than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_lt/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator<(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) < rhs;
+    }
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    friend bool operator<=(const_reference lhs, const_reference rhs) noexcept
+    {
+        if (compares_unordered(lhs, rhs, true))
+        {
+            return false;
+        }
+        return !(rhs < lhs);
+    }
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator<=(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs <= basic_json(rhs);
+    }
+
+    /// @brief comparison: less than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_le/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator<=(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) <= rhs;
+    }
+
+    /// @brief comparison: greater than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gt/
+    friend bool operator>(const_reference lhs, const_reference rhs) noexcept
+    {
+        // double inverse
+        if (compares_unordered(lhs, rhs))
+        {
+            return false;
+        }
+        return !(lhs <= rhs);
+    }
+
+    /// @brief comparison: greater than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gt/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator>(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs > basic_json(rhs);
+    }
+
+    /// @brief comparison: greater than
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gt/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator>(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) > rhs;
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    friend bool operator>=(const_reference lhs, const_reference rhs) noexcept
+    {
+        if (compares_unordered(lhs, rhs, true))
+        {
+            return false;
+        }
+        return !(lhs < rhs);
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator>=(const_reference lhs, ScalarType rhs) noexcept
+    {
+        return lhs >= basic_json(rhs);
+    }
+
+    /// @brief comparison: greater than or equal
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ge/
+    template<typename ScalarType, typename std::enable_if<
+                 std::is_scalar<ScalarType>::value, int>::type = 0>
+    friend bool operator>=(ScalarType lhs, const_reference rhs) noexcept
+    {
+        return basic_json(lhs) >= rhs;
+    }
+#endif
+
+#undef JSON_IMPLEMENT_OPERATOR
+
+    /// @}
+
+    ///////////////////
+    // serialization //
+    ///////////////////
+
+    /// @name serialization
+    /// @{
+#ifndef JSON_NO_IO
+    /// @brief serialize to stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ltlt/
+    friend std::ostream& operator<<(std::ostream& o, const basic_json& j)
+    {
+        // read width member and use it as indentation parameter if nonzero
+        const bool pretty_print = o.width() > 0;
+        const auto indentation = pretty_print ? o.width() : 0;
+
+        // reset width to 0 for subsequent calls to this stream
+        o.width(0);
+
+        // do the actual serialization
+        serializer s(detail::output_adapter<char>(o), o.fill());
+        s.dump(j, pretty_print, false, static_cast<unsigned int>(indentation));
+        return o;
+    }
+
+    /// @brief serialize to stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_ltlt/
+    /// @deprecated This function is deprecated since 3.0.0 and will be removed in
+    ///             version 4.0.0 of the library. Please use
+    ///             operator<<(std::ostream&, const basic_json&) instead; that is,
+    ///             replace calls like `j >> o;` with `o << j;`.
+    JSON_HEDLEY_DEPRECATED_FOR(3.0.0, operator<<(std::ostream&, const basic_json&))
+    friend std::ostream& operator>>(const basic_json& j, std::ostream& o)
+    {
+        return o << j;
+    }
+#endif  // JSON_NO_IO
+    /// @}
+
+    /////////////////////
+    // deserialization //
+    /////////////////////
+
+    /// @name deserialization
+    /// @{
+
+    /// @brief deserialize from a compatible input
+    /// @sa https://json.nlohmann.me/api/basic_json/parse/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json parse(InputType&& i,
+                            parser_callback_t cb = nullptr,
+                            const bool allow_exceptions = true,
+                            const bool ignore_comments = false)
+    {
+        basic_json result;
+        parser(detail::input_adapter(std::forward<InputType>(i)), std::move(cb), allow_exceptions, ignore_comments).parse(true, result); // cppcheck-suppress[accessMoved,accessForwarded]
+        return result;
+    }
+
+    /// @brief deserialize from a pair of character iterators
+    /// @sa https://json.nlohmann.me/api/basic_json/parse/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json parse(IteratorType first,
+                            IteratorType last,
+                            parser_callback_t cb = nullptr,
+                            const bool allow_exceptions = true,
+                            const bool ignore_comments = false)
+    {
+        basic_json result;
+        parser(detail::input_adapter(std::move(first), std::move(last)), std::move(cb), allow_exceptions, ignore_comments).parse(true, result); // cppcheck-suppress[accessMoved]
+        return result;
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, parse(ptr, ptr + len))
+    static basic_json parse(detail::span_input_adapter&& i,
+                            parser_callback_t cb = nullptr,
+                            const bool allow_exceptions = true,
+                            const bool ignore_comments = false)
+    {
+        basic_json result;
+        parser(i.get(), std::move(cb), allow_exceptions, ignore_comments).parse(true, result); // cppcheck-suppress[accessMoved]
+        return result;
+    }
+
+    /// @brief check if the input is valid JSON
+    /// @sa https://json.nlohmann.me/api/basic_json/accept/
+    template<typename InputType>
+    static bool accept(InputType&& i,
+                       const bool ignore_comments = false)
+    {
+        return parser(detail::input_adapter(std::forward<InputType>(i)), nullptr, false, ignore_comments).accept(true);
+    }
+
+    /// @brief check if the input is valid JSON
+    /// @sa https://json.nlohmann.me/api/basic_json/accept/
+    template<typename IteratorType>
+    static bool accept(IteratorType first, IteratorType last,
+                       const bool ignore_comments = false)
+    {
+        return parser(detail::input_adapter(std::move(first), std::move(last)), nullptr, false, ignore_comments).accept(true);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, accept(ptr, ptr + len))
+    static bool accept(detail::span_input_adapter&& i,
+                       const bool ignore_comments = false)
+    {
+        return parser(i.get(), nullptr, false, ignore_comments).accept(true);
+    }
+
+    /// @brief generate SAX events
+    /// @sa https://json.nlohmann.me/api/basic_json/sax_parse/
+    template <typename InputType, typename SAX>
+    JSON_HEDLEY_NON_NULL(2)
+    static bool sax_parse(InputType&& i, SAX* sax,
+                          input_format_t format = input_format_t::json,
+                          const bool strict = true,
+                          const bool ignore_comments = false)
+    {
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        return format == input_format_t::json
+               ? parser(std::move(ia), nullptr, true, ignore_comments).sax_parse(sax, strict)
+               : detail::binary_reader<basic_json, decltype(ia), SAX>(std::move(ia), format).sax_parse(format, sax, strict);
+    }
+
+    /// @brief generate SAX events
+    /// @sa https://json.nlohmann.me/api/basic_json/sax_parse/
+    template<class IteratorType, class SAX>
+    JSON_HEDLEY_NON_NULL(3)
+    static bool sax_parse(IteratorType first, IteratorType last, SAX* sax,
+                          input_format_t format = input_format_t::json,
+                          const bool strict = true,
+                          const bool ignore_comments = false)
+    {
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        return format == input_format_t::json
+               ? parser(std::move(ia), nullptr, true, ignore_comments).sax_parse(sax, strict)
+               : detail::binary_reader<basic_json, decltype(ia), SAX>(std::move(ia), format).sax_parse(format, sax, strict);
+    }
+
+    /// @brief generate SAX events
+    /// @sa https://json.nlohmann.me/api/basic_json/sax_parse/
+    /// @deprecated This function is deprecated since 3.8.0 and will be removed in
+    ///             version 4.0.0 of the library. Please use
+    ///             sax_parse(ptr, ptr + len) instead.
+    template <typename SAX>
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, sax_parse(ptr, ptr + len, ...))
+    JSON_HEDLEY_NON_NULL(2)
+    static bool sax_parse(detail::span_input_adapter&& i, SAX* sax,
+                          input_format_t format = input_format_t::json,
+                          const bool strict = true,
+                          const bool ignore_comments = false)
+    {
+        auto ia = i.get();
+        return format == input_format_t::json
+               // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+               ? parser(std::move(ia), nullptr, true, ignore_comments).sax_parse(sax, strict)
+               // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+               : detail::binary_reader<basic_json, decltype(ia), SAX>(std::move(ia), format).sax_parse(format, sax, strict);
+    }
+#ifndef JSON_NO_IO
+    /// @brief deserialize from stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gtgt/
+    /// @deprecated This stream operator is deprecated since 3.0.0 and will be removed in
+    ///             version 4.0.0 of the library. Please use
+    ///             operator>>(std::istream&, basic_json&) instead; that is,
+    ///             replace calls like `j << i;` with `i >> j;`.
+    JSON_HEDLEY_DEPRECATED_FOR(3.0.0, operator>>(std::istream&, basic_json&))
+    friend std::istream& operator<<(basic_json& j, std::istream& i)
+    {
+        return operator>>(i, j);
+    }
+
+    /// @brief deserialize from stream
+    /// @sa https://json.nlohmann.me/api/basic_json/operator_gtgt/
+    friend std::istream& operator>>(std::istream& i, basic_json& j)
+    {
+        parser(detail::input_adapter(i)).parse(false, j);
+        return i;
+    }
+#endif  // JSON_NO_IO
+    /// @}
+
+    ///////////////////////////
+    // convenience functions //
+    ///////////////////////////
+
+    /// @brief return the type as string
+    /// @sa https://json.nlohmann.me/api/basic_json/type_name/
+    JSON_HEDLEY_RETURNS_NON_NULL
+    const char* type_name() const noexcept
+    {
+        switch (m_data.m_type)
+        {
+            case value_t::null:
+                return "null";
+            case value_t::object:
+                return "object";
+            case value_t::array:
+                return "array";
+            case value_t::string:
+                return "string";
+            case value_t::boolean:
+                return "boolean";
+            case value_t::binary:
+                return "binary";
+            case value_t::discarded:
+                return "discarded";
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            default:
+                return "number";
+        }
+    }
+
+  JSON_PRIVATE_UNLESS_TESTED:
+    //////////////////////
+    // member variables //
+    //////////////////////
+
+    struct data
+    {
+        /// the type of the current element
+        value_t m_type = value_t::null;
+
+        /// the value of the current element
+        json_value m_value = {};
+
+        data(const value_t v)
+            : m_type(v), m_value(v)
+        {
+        }
+
+        data(size_type cnt, const basic_json& val)
+            : m_type(value_t::array)
+        {
+            m_value.array = create<array_t>(cnt, val);
+        }
+
+        data() noexcept = default;
+        data(data&&) noexcept = default;
+        data(const data&) noexcept = delete;
+        data& operator=(data&&) noexcept = delete;
+        data& operator=(const data&) noexcept = delete;
+
+        ~data() noexcept
+        {
+            m_value.destroy(m_type);
+        }
+    };
+
+    data m_data = {};
+
+#if JSON_DIAGNOSTICS
+    /// a pointer to a parent value (for debugging purposes)
+    basic_json* m_parent = nullptr;
+#endif
+
+#if JSON_DIAGNOSTIC_POSITIONS
+    /// the start position of the value
+    std::size_t start_position = std::string::npos;
+    /// the end position of the value
+    std::size_t end_position = std::string::npos;
+  public:
+    constexpr std::size_t start_pos() const noexcept
+    {
+        return start_position;
+    }
+
+    constexpr std::size_t end_pos() const noexcept
+    {
+        return end_position;
+    }
+#endif
+
+    //////////////////////////////////////////
+    // binary serialization/deserialization //
+    //////////////////////////////////////////
+
+    /// @name binary serialization/deserialization support
+    /// @{
+
+  public:
+    /// @brief create a CBOR serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_cbor/
+    static std::vector<std::uint8_t> to_cbor(const basic_json& j)
+    {
+        std::vector<std::uint8_t> result;
+        to_cbor(j, result);
+        return result;
+    }
+
+    /// @brief create a CBOR serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_cbor/
+    static void to_cbor(const basic_json& j, detail::output_adapter<std::uint8_t> o)
+    {
+        binary_writer<std::uint8_t>(o).write_cbor(j);
+    }
+
+    /// @brief create a CBOR serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_cbor/
+    static void to_cbor(const basic_json& j, detail::output_adapter<char> o)
+    {
+        binary_writer<char>(o).write_cbor(j);
+    }
+
+    /// @brief create a MessagePack serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_msgpack/
+    static std::vector<std::uint8_t> to_msgpack(const basic_json& j)
+    {
+        std::vector<std::uint8_t> result;
+        to_msgpack(j, result);
+        return result;
+    }
+
+    /// @brief create a MessagePack serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_msgpack/
+    static void to_msgpack(const basic_json& j, detail::output_adapter<std::uint8_t> o)
+    {
+        binary_writer<std::uint8_t>(o).write_msgpack(j);
+    }
+
+    /// @brief create a MessagePack serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_msgpack/
+    static void to_msgpack(const basic_json& j, detail::output_adapter<char> o)
+    {
+        binary_writer<char>(o).write_msgpack(j);
+    }
+
+    /// @brief create a UBJSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_ubjson/
+    static std::vector<std::uint8_t> to_ubjson(const basic_json& j,
+            const bool use_size = false,
+            const bool use_type = false)
+    {
+        std::vector<std::uint8_t> result;
+        to_ubjson(j, result, use_size, use_type);
+        return result;
+    }
+
+    /// @brief create a UBJSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_ubjson/
+    static void to_ubjson(const basic_json& j, detail::output_adapter<std::uint8_t> o,
+                          const bool use_size = false, const bool use_type = false)
+    {
+        binary_writer<std::uint8_t>(o).write_ubjson(j, use_size, use_type);
+    }
+
+    /// @brief create a UBJSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_ubjson/
+    static void to_ubjson(const basic_json& j, detail::output_adapter<char> o,
+                          const bool use_size = false, const bool use_type = false)
+    {
+        binary_writer<char>(o).write_ubjson(j, use_size, use_type);
+    }
+
+    /// @brief create a BJData serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bjdata/
+    static std::vector<std::uint8_t> to_bjdata(const basic_json& j,
+            const bool use_size = false,
+            const bool use_type = false,
+            const bjdata_version_t version = bjdata_version_t::draft2)
+    {
+        std::vector<std::uint8_t> result;
+        to_bjdata(j, result, use_size, use_type, version);
+        return result;
+    }
+
+    /// @brief create a BJData serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bjdata/
+    static void to_bjdata(const basic_json& j, detail::output_adapter<std::uint8_t> o,
+                          const bool use_size = false, const bool use_type = false,
+                          const bjdata_version_t version = bjdata_version_t::draft2)
+    {
+        binary_writer<std::uint8_t>(o).write_ubjson(j, use_size, use_type, true, true, version);
+    }
+
+    /// @brief create a BJData serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bjdata/
+    static void to_bjdata(const basic_json& j, detail::output_adapter<char> o,
+                          const bool use_size = false, const bool use_type = false,
+                          const bjdata_version_t version = bjdata_version_t::draft2)
+    {
+        binary_writer<char>(o).write_ubjson(j, use_size, use_type, true, true, version);
+    }
+
+    /// @brief create a BSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bson/
+    static std::vector<std::uint8_t> to_bson(const basic_json& j)
+    {
+        std::vector<std::uint8_t> result;
+        to_bson(j, result);
+        return result;
+    }
+
+    /// @brief create a BSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bson/
+    static void to_bson(const basic_json& j, detail::output_adapter<std::uint8_t> o)
+    {
+        binary_writer<std::uint8_t>(o).write_bson(j);
+    }
+
+    /// @brief create a BSON serialization of a given JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/to_bson/
+    static void to_bson(const basic_json& j, detail::output_adapter<char> o)
+    {
+        binary_writer<char>(o).write_bson(j);
+    }
+
+    /// @brief create a JSON value from an input in CBOR format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_cbor/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_cbor(InputType&& i,
+                                const bool strict = true,
+                                const bool allow_exceptions = true,
+                                const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::cbor).sax_parse(input_format_t::cbor, &sdp, strict, tag_handler); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in CBOR format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_cbor/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_cbor(IteratorType first, IteratorType last,
+                                const bool strict = true,
+                                const bool allow_exceptions = true,
+                                const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::cbor).sax_parse(input_format_t::cbor, &sdp, strict, tag_handler); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    template<typename T>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_cbor(ptr, ptr + len))
+    static basic_json from_cbor(const T* ptr, std::size_t len,
+                                const bool strict = true,
+                                const bool allow_exceptions = true,
+                                const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        return from_cbor(ptr, ptr + len, strict, allow_exceptions, tag_handler);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_cbor(ptr, ptr + len))
+    static basic_json from_cbor(detail::span_input_adapter&& i,
+                                const bool strict = true,
+                                const bool allow_exceptions = true,
+                                const cbor_tag_handler_t tag_handler = cbor_tag_handler_t::error)
+    {
+        basic_json result;
+        auto ia = i.get();
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::cbor).sax_parse(input_format_t::cbor, &sdp, strict, tag_handler); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in MessagePack format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_msgpack/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_msgpack(InputType&& i,
+                                   const bool strict = true,
+                                   const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::msgpack).sax_parse(input_format_t::msgpack, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in MessagePack format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_msgpack/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_msgpack(IteratorType first, IteratorType last,
+                                   const bool strict = true,
+                                   const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::msgpack).sax_parse(input_format_t::msgpack, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    template<typename T>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_msgpack(ptr, ptr + len))
+    static basic_json from_msgpack(const T* ptr, std::size_t len,
+                                   const bool strict = true,
+                                   const bool allow_exceptions = true)
+    {
+        return from_msgpack(ptr, ptr + len, strict, allow_exceptions);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_msgpack(ptr, ptr + len))
+    static basic_json from_msgpack(detail::span_input_adapter&& i,
+                                   const bool strict = true,
+                                   const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = i.get();
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::msgpack).sax_parse(input_format_t::msgpack, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in UBJSON format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_ubjson/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_ubjson(InputType&& i,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::ubjson).sax_parse(input_format_t::ubjson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in UBJSON format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_ubjson/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_ubjson(IteratorType first, IteratorType last,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::ubjson).sax_parse(input_format_t::ubjson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    template<typename T>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_ubjson(ptr, ptr + len))
+    static basic_json from_ubjson(const T* ptr, std::size_t len,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        return from_ubjson(ptr, ptr + len, strict, allow_exceptions);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_ubjson(ptr, ptr + len))
+    static basic_json from_ubjson(detail::span_input_adapter&& i,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = i.get();
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::ubjson).sax_parse(input_format_t::ubjson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in BJData format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_bjdata/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_bjdata(InputType&& i,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bjdata).sax_parse(input_format_t::bjdata, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in BJData format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_bjdata/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_bjdata(IteratorType first, IteratorType last,
+                                  const bool strict = true,
+                                  const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bjdata).sax_parse(input_format_t::bjdata, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in BSON format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_bson/
+    template<typename InputType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_bson(InputType&& i,
+                                const bool strict = true,
+                                const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::forward<InputType>(i));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bson).sax_parse(input_format_t::bson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    /// @brief create a JSON value from an input in BSON format
+    /// @sa https://json.nlohmann.me/api/basic_json/from_bson/
+    template<typename IteratorType>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json from_bson(IteratorType first, IteratorType last,
+                                const bool strict = true,
+                                const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = detail::input_adapter(std::move(first), std::move(last));
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bson).sax_parse(input_format_t::bson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+
+    template<typename T>
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_bson(ptr, ptr + len))
+    static basic_json from_bson(const T* ptr, std::size_t len,
+                                const bool strict = true,
+                                const bool allow_exceptions = true)
+    {
+        return from_bson(ptr, ptr + len, strict, allow_exceptions);
+    }
+
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    JSON_HEDLEY_DEPRECATED_FOR(3.8.0, from_bson(ptr, ptr + len))
+    static basic_json from_bson(detail::span_input_adapter&& i,
+                                const bool strict = true,
+                                const bool allow_exceptions = true)
+    {
+        basic_json result;
+        auto ia = i.get();
+        detail::json_sax_dom_parser<basic_json, decltype(ia)> sdp(result, allow_exceptions);
+        // NOLINTNEXTLINE(hicpp-move-const-arg,performance-move-const-arg)
+        const bool res = binary_reader<decltype(ia)>(std::move(ia), input_format_t::bson).sax_parse(input_format_t::bson, &sdp, strict); // cppcheck-suppress[accessMoved]
+        return res ? result : basic_json(value_t::discarded);
+    }
+    /// @}
+
+    //////////////////////////
+    // JSON Pointer support //
+    //////////////////////////
+
+    /// @name JSON Pointer functions
+    /// @{
+
+    /// @brief access specified element via JSON Pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    reference operator[](const json_pointer& ptr)
+    {
+        return ptr.get_unchecked(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    reference operator[](const ::nlohmann::json_pointer<BasicJsonType>& ptr)
+    {
+        return ptr.get_unchecked(this);
+    }
+
+    /// @brief access specified element via JSON Pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/operator%5B%5D/
+    const_reference operator[](const json_pointer& ptr) const
+    {
+        return ptr.get_unchecked(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    const_reference operator[](const ::nlohmann::json_pointer<BasicJsonType>& ptr) const
+    {
+        return ptr.get_unchecked(this);
+    }
+
+    /// @brief access specified element via JSON Pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    reference at(const json_pointer& ptr)
+    {
+        return ptr.get_checked(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    reference at(const ::nlohmann::json_pointer<BasicJsonType>& ptr)
+    {
+        return ptr.get_checked(this);
+    }
+
+    /// @brief access specified element via JSON Pointer
+    /// @sa https://json.nlohmann.me/api/basic_json/at/
+    const_reference at(const json_pointer& ptr) const
+    {
+        return ptr.get_checked(this);
+    }
+
+    template<typename BasicJsonType, detail::enable_if_t<detail::is_basic_json<BasicJsonType>::value, int> = 0>
+    JSON_HEDLEY_DEPRECATED_FOR(3.11.0, basic_json::json_pointer or nlohmann::json_pointer<basic_json::string_t>) // NOLINT(readability/alt_tokens)
+    const_reference at(const ::nlohmann::json_pointer<BasicJsonType>& ptr) const
+    {
+        return ptr.get_checked(this);
+    }
+
+    /// @brief return flattened JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/flatten/
+    basic_json flatten() const
+    {
+        basic_json result(value_t::object);
+        json_pointer::flatten("", *this, result);
+        return result;
+    }
+
+    /// @brief unflatten a previously flattened JSON value
+    /// @sa https://json.nlohmann.me/api/basic_json/unflatten/
+    basic_json unflatten() const
+    {
+        return json_pointer::unflatten(*this);
+    }
+
+    /// @}
+
+    //////////////////////////
+    // JSON Patch functions //
+    //////////////////////////
+
+    /// @name JSON Patch functions
+    /// @{
+
+    /// @brief applies a JSON patch in-place without copying the object
+    /// @sa https://json.nlohmann.me/api/basic_json/patch/
+    void patch_inplace(const basic_json& json_patch)
+    {
+        basic_json& result = *this;
+        // the valid JSON Patch operations
+        enum class patch_operations {add, remove, replace, move, copy, test, invalid};
+
+        const auto get_op = [](const string_t& op)
+        {
+            if (op == "add")
+            {
+                return patch_operations::add;
+            }
+            if (op == "remove")
+            {
+                return patch_operations::remove;
+            }
+            if (op == "replace")
+            {
+                return patch_operations::replace;
+            }
+            if (op == "move")
+            {
+                return patch_operations::move;
+            }
+            if (op == "copy")
+            {
+                return patch_operations::copy;
+            }
+            if (op == "test")
+            {
+                return patch_operations::test;
+            }
+
+            return patch_operations::invalid;
+        };
+
+        // wrapper for "add" operation; add value at ptr
+        const auto operation_add = [&result](json_pointer & ptr, const basic_json & val)
+        {
+            // adding to the root of the target document means replacing it
+            if (ptr.empty())
+            {
+                result = val;
+                return;
+            }
+
+            // make sure the top element of the pointer exists
+            json_pointer const top_pointer = ptr.top();
+            if (top_pointer != ptr)
+            {
+                result.at(top_pointer);
+            }
+
+            // get reference to parent of JSON pointer ptr
+            const auto last_path = ptr.back();
+            ptr.pop_back();
+            // parent must exist when performing patch add per RFC6902 specs
+            basic_json& parent = result.at(ptr);
+
+            switch (parent.m_data.m_type)
+            {
+                case value_t::null:
+                case value_t::object:
+                {
+                    // use operator[] to add value
+                    parent[last_path] = val;
+                    break;
+                }
+
+                case value_t::array:
+                {
+                    if (last_path == "-")
+                    {
+                        // special case: append to back
+                        parent.push_back(val);
+                    }
+                    else
+                    {
+                        const auto idx = json_pointer::template array_index<basic_json_t>(last_path);
+                        if (JSON_HEDLEY_UNLIKELY(idx > parent.size()))
+                        {
+                            // avoid undefined behavior
+                            JSON_THROW(out_of_range::create(401, detail::concat("array index ", std::to_string(idx), " is out of range"), &parent));
+                        }
+
+                        // default case: insert add offset
+                        parent.insert(parent.begin() + static_cast<difference_type>(idx), val);
+                    }
+                    break;
+                }
+
+                // if there exists a parent it cannot be primitive
+                case value_t::string: // LCOV_EXCL_LINE
+                case value_t::boolean: // LCOV_EXCL_LINE
+                case value_t::number_integer: // LCOV_EXCL_LINE
+                case value_t::number_unsigned: // LCOV_EXCL_LINE
+                case value_t::number_float: // LCOV_EXCL_LINE
+                case value_t::binary: // LCOV_EXCL_LINE
+                case value_t::discarded: // LCOV_EXCL_LINE
+                default:            // LCOV_EXCL_LINE
+                    JSON_ASSERT(false); // NOLINT(cert-dcl03-c,hicpp-static-assert,misc-static-assert) LCOV_EXCL_LINE
+            }
+        };
+
+        // wrapper for "remove" operation; remove value at ptr
+        const auto operation_remove = [this, & result](json_pointer & ptr)
+        {
+            // get reference to parent of JSON pointer ptr
+            const auto last_path = ptr.back();
+            ptr.pop_back();
+            basic_json& parent = result.at(ptr);
+
+            // remove child
+            if (parent.is_object())
+            {
+                // perform range check
+                auto it = parent.find(last_path);
+                if (JSON_HEDLEY_LIKELY(it != parent.end()))
+                {
+                    parent.erase(it);
+                }
+                else
+                {
+                    JSON_THROW(out_of_range::create(403, detail::concat("key '", last_path, "' not found"), this));
+                }
+            }
+            else if (parent.is_array())
+            {
+                // note erase performs range check
+                parent.erase(json_pointer::template array_index<basic_json_t>(last_path));
+            }
+        };
+
+        // type check: top level value must be an array
+        if (JSON_HEDLEY_UNLIKELY(!json_patch.is_array()))
+        {
+            JSON_THROW(parse_error::create(104, 0, "JSON patch must be an array of objects", &json_patch));
+        }
+
+        // iterate and apply the operations
+        for (const auto& val : json_patch)
+        {
+            // wrapper to get a value for an operation
+            const auto get_value = [&val](const string_t& op,
+                                          const string_t& member,
+                                          bool string_type) -> basic_json &
+            {
+                // find value
+                auto it = val.m_data.m_value.object->find(member);
+
+                // context-sensitive error message
+                const auto error_msg = (op == "op") ? "operation" : detail::concat("operation '", op, '\''); // NOLINT(bugprone-unused-local-non-trivial-variable)
+
+                // check if desired value is present
+                if (JSON_HEDLEY_UNLIKELY(it == val.m_data.m_value.object->end()))
+                {
+                    // NOLINTNEXTLINE(performance-inefficient-string-concatenation)
+                    JSON_THROW(parse_error::create(105, 0, detail::concat(error_msg, " must have member '", member, "'"), &val));
+                }
+
+                // check if result is of type string
+                if (JSON_HEDLEY_UNLIKELY(string_type && !it->second.is_string()))
+                {
+                    // NOLINTNEXTLINE(performance-inefficient-string-concatenation)
+                    JSON_THROW(parse_error::create(105, 0, detail::concat(error_msg, " must have string member '", member, "'"), &val));
+                }
+
+                // no error: return value
+                return it->second;
+            };
+
+            // type check: every element of the array must be an object
+            if (JSON_HEDLEY_UNLIKELY(!val.is_object()))
+            {
+                JSON_THROW(parse_error::create(104, 0, "JSON patch must be an array of objects", &val));
+            }
+
+            // collect mandatory members
+            const auto op = get_value("op", "op", true).template get<string_t>();
+            const auto path = get_value(op, "path", true).template get<string_t>();
+            json_pointer ptr(path);
+
+            switch (get_op(op))
+            {
+                case patch_operations::add:
+                {
+                    operation_add(ptr, get_value("add", "value", false));
+                    break;
+                }
+
+                case patch_operations::remove:
+                {
+                    operation_remove(ptr);
+                    break;
+                }
+
+                case patch_operations::replace:
+                {
+                    // the "path" location must exist - use at()
+                    result.at(ptr) = get_value("replace", "value", false);
+                    break;
+                }
+
+                case patch_operations::move:
+                {
+                    const auto from_path = get_value("move", "from", true).template get<string_t>();
+                    json_pointer from_ptr(from_path);
+
+                    // the "from" location must exist - use at()
+                    basic_json const v = result.at(from_ptr);
+
+                    // The move operation is functionally identical to a
+                    // "remove" operation on the "from" location, followed
+                    // immediately by an "add" operation at the target
+                    // location with the value that was just removed.
+                    operation_remove(from_ptr);
+                    operation_add(ptr, v);
+                    break;
+                }
+
+                case patch_operations::copy:
+                {
+                    const auto from_path = get_value("copy", "from", true).template get<string_t>();
+                    const json_pointer from_ptr(from_path);
+
+                    // the "from" location must exist - use at()
+                    basic_json const v = result.at(from_ptr);
+
+                    // The copy is functionally identical to an "add"
+                    // operation at the target location using the value
+                    // specified in the "from" member.
+                    operation_add(ptr, v);
+                    break;
+                }
+
+                case patch_operations::test:
+                {
+                    bool success = false;
+                    JSON_TRY
+                    {
+                        // check if "value" matches the one at "path"
+                        // the "path" location must exist - use at()
+                        success = (result.at(ptr) == get_value("test", "value", false));
+                    }
+                    JSON_INTERNAL_CATCH (out_of_range&)
+                    {
+                        // ignore out of range errors: success remains false
+                    }
+
+                    // throw an exception if test fails
+                    if (JSON_HEDLEY_UNLIKELY(!success))
+                    {
+                        JSON_THROW(other_error::create(501, detail::concat("unsuccessful: ", val.dump()), &val));
+                    }
+
+                    break;
+                }
+
+                case patch_operations::invalid:
+                default:
+                {
+                    // op must be "add", "remove", "replace", "move", "copy", or
+                    // "test"
+                    JSON_THROW(parse_error::create(105, 0, detail::concat("operation value '", op, "' is invalid"), &val));
+                }
+            }
+        }
+    }
+
+    /// @brief applies a JSON patch to a copy of the current object
+    /// @sa https://json.nlohmann.me/api/basic_json/patch/
+    basic_json patch(const basic_json& json_patch) const
+    {
+        basic_json result = *this;
+        result.patch_inplace(json_patch);
+        return result;
+    }
+
+    /// @brief creates a diff as a JSON patch
+    /// @sa https://json.nlohmann.me/api/basic_json/diff/
+    JSON_HEDLEY_WARN_UNUSED_RESULT
+    static basic_json diff(const basic_json& source, const basic_json& target,
+                           const string_t& path = "")
+    {
+        // the patch
+        basic_json result(value_t::array);
+
+        // if the values are the same, return empty patch
+        if (source == target)
+        {
+            return result;
+        }
+
+        if (source.type() != target.type())
+        {
+            // different types: replace value
+            result.push_back(
+            {
+                {"op", "replace"}, {"path", path}, {"value", target}
+            });
+            return result;
+        }
+
+        switch (source.type())
+        {
+            case value_t::array:
+            {
+                // first pass: traverse common elements
+                std::size_t i = 0;
+                while (i < source.size() && i < target.size())
+                {
+                    // recursive call to compare array values at index i
+                    auto temp_diff = diff(source[i], target[i], detail::concat<string_t>(path, '/', detail::to_string<string_t>(i)));
+                    result.insert(result.end(), temp_diff.begin(), temp_diff.end());
+                    ++i;
+                }
+
+                // We now reached the end of at least one array
+                // in a second pass, traverse the remaining elements
+
+                // remove my remaining elements
+                const auto end_index = static_cast<difference_type>(result.size());
+                while (i < source.size())
+                {
+                    // add operations in reverse order to avoid invalid
+                    // indices
+                    result.insert(result.begin() + end_index, object(
+                    {
+                        {"op", "remove"},
+                        {"path", detail::concat<string_t>(path, '/', detail::to_string<string_t>(i))}
+                    }));
+                    ++i;
+                }
+
+                // add other remaining elements
+                while (i < target.size())
+                {
+                    result.push_back(
+                    {
+                        {"op", "add"},
+                        {"path", detail::concat<string_t>(path, "/-")},
+                        {"value", target[i]}
+                    });
+                    ++i;
+                }
+
+                break;
+            }
+
+            case value_t::object:
+            {
+                // first pass: traverse this object's elements
+                for (auto it = source.cbegin(); it != source.cend(); ++it)
+                {
+                    // escape the key name to be used in a JSON patch
+                    const auto path_key = detail::concat<string_t>(path, '/', detail::escape(it.key()));
+
+                    if (target.find(it.key()) != target.end())
+                    {
+                        // recursive call to compare object values at key it
+                        auto temp_diff = diff(it.value(), target[it.key()], path_key);
+                        result.insert(result.end(), temp_diff.begin(), temp_diff.end());
+                    }
+                    else
+                    {
+                        // found a key that is not in o -> remove it
+                        result.push_back(object(
+                        {
+                            {"op", "remove"}, {"path", path_key}
+                        }));
+                    }
+                }
+
+                // second pass: traverse other object's elements
+                for (auto it = target.cbegin(); it != target.cend(); ++it)
+                {
+                    if (source.find(it.key()) == source.end())
+                    {
+                        // found a key that is not in this -> add it
+                        const auto path_key = detail::concat<string_t>(path, '/', detail::escape(it.key()));
+                        result.push_back(
+                        {
+                            {"op", "add"}, {"path", path_key},
+                            {"value", it.value()}
+                        });
+                    }
+                }
+
+                break;
+            }
+
+            case value_t::null:
+            case value_t::string:
+            case value_t::boolean:
+            case value_t::number_integer:
+            case value_t::number_unsigned:
+            case value_t::number_float:
+            case value_t::binary:
+            case value_t::discarded:
+            default:
+            {
+                // both primitive type: replace value
+                result.push_back(
+                {
+                    {"op", "replace"}, {"path", path}, {"value", target}
+                });
+                break;
+            }
+        }
+
+        return result;
+    }
+    /// @}
+
+    ////////////////////////////////
+    // JSON Merge Patch functions //
+    ////////////////////////////////
+
+    /// @name JSON Merge Patch functions
+    /// @{
+
+    /// @brief applies a JSON Merge Patch
+    /// @sa https://json.nlohmann.me/api/basic_json/merge_patch/
+    void merge_patch(const basic_json& apply_patch)
+    {
+        if (apply_patch.is_object())
+        {
+            if (!is_object())
+            {
+                *this = object();
+            }
+            for (auto it = apply_patch.begin(); it != apply_patch.end(); ++it)
+            {
+                if (it.value().is_null())
+                {
+                    erase(it.key());
+                }
+                else
+                {
+                    operator[](it.key()).merge_patch(it.value());
+                }
+            }
+        }
+        else
+        {
+            *this = apply_patch;
+        }
+    }
+
+    /// @}
+};
+
+/// @brief user-defined to_string function for JSON values
+/// @sa https://json.nlohmann.me/api/basic_json/to_string/
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+std::string to_string(const NLOHMANN_BASIC_JSON_TPL& j)
+{
+    return j.dump();
+}
+
+inline namespace literals
+{
+inline namespace json_literals
+{
+
+/// @brief user-defined string literal for JSON values
+/// @sa https://json.nlohmann.me/api/basic_json/operator_literal_json/
+JSON_HEDLEY_NON_NULL(1)
+#if !defined(JSON_HEDLEY_GCC_VERSION) || JSON_HEDLEY_GCC_VERSION_CHECK(4,9,0)
+    inline nlohmann::json operator ""_json(const char* s, std::size_t n)
+#else
+    inline nlohmann::json operator "" _json(const char* s, std::size_t n)
+#endif
+{
+    return nlohmann::json::parse(s, s + n);
+}
+
+/// @brief user-defined string literal for JSON pointer
+/// @sa https://json.nlohmann.me/api/basic_json/operator_literal_json_pointer/
+JSON_HEDLEY_NON_NULL(1)
+#if !defined(JSON_HEDLEY_GCC_VERSION) || JSON_HEDLEY_GCC_VERSION_CHECK(4,9,0)
+    inline nlohmann::json::json_pointer operator ""_json_pointer(const char* s, std::size_t n)
+#else
+    inline nlohmann::json::json_pointer operator "" _json_pointer(const char* s, std::size_t n)
+#endif
+{
+    return nlohmann::json::json_pointer(std::string(s, n));
+}
+
+}  // namespace json_literals
+}  // namespace literals
+NLOHMANN_JSON_NAMESPACE_END
+
+///////////////////////
+// nonmember support //
+///////////////////////
+
+namespace std // NOLINT(cert-dcl58-cpp)
+{
+
+/// @brief hash value for JSON objects
+/// @sa https://json.nlohmann.me/api/basic_json/std_hash/
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+struct hash<nlohmann::NLOHMANN_BASIC_JSON_TPL> // NOLINT(cert-dcl58-cpp)
+{
+    std::size_t operator()(const nlohmann::NLOHMANN_BASIC_JSON_TPL& j) const
+    {
+        return nlohmann::detail::hash(j);
+    }
+};
+
+// specialization for std::less<value_t>
+template<>
+struct less< ::nlohmann::detail::value_t> // do not remove the space after '<', see https://github.com/nlohmann/json/pull/679
+{
+    /*!
+    @brief compare two value_t enum values
+    @since version 3.0.0
+    */
+    bool operator()(::nlohmann::detail::value_t lhs,
+                    ::nlohmann::detail::value_t rhs) const noexcept
+    {
+#if JSON_HAS_THREE_WAY_COMPARISON
+        return std::is_lt(lhs <=> rhs); // *NOPAD*
+#else
+        return ::nlohmann::detail::operator<(lhs, rhs);
+#endif
+    }
+};
+
+// C++20 prohibit function specialization in the std namespace.
+#ifndef JSON_HAS_CPP_20
+
+/// @brief exchanges the values of two JSON objects
+/// @sa https://json.nlohmann.me/api/basic_json/std_swap/
+NLOHMANN_BASIC_JSON_TPL_DECLARATION
+inline void swap(nlohmann::NLOHMANN_BASIC_JSON_TPL& j1, nlohmann::NLOHMANN_BASIC_JSON_TPL& j2) noexcept(  // NOLINT(readability-inconsistent-declaration-parameter-name, cert-dcl58-cpp)
+    is_nothrow_move_constructible<nlohmann::NLOHMANN_BASIC_JSON_TPL>::value&&                          // NOLINT(misc-redundant-expression,cppcoreguidelines-noexcept-swap,performance-noexcept-swap)
+    is_nothrow_move_assignable<nlohmann::NLOHMANN_BASIC_JSON_TPL>::value)
+{
+    j1.swap(j2);
+}
+
+#endif
+
+}  // namespace std
+
+#if JSON_USE_GLOBAL_UDLS
+    #if !defined(JSON_HEDLEY_GCC_VERSION) || JSON_HEDLEY_GCC_VERSION_CHECK(4,9,0)
+        using nlohmann::literals::json_literals::operator ""_json; // NOLINT(misc-unused-using-decls,google-global-names-in-headers)
+        using nlohmann::literals::json_literals::operator ""_json_pointer; //NOLINT(misc-unused-using-decls,google-global-names-in-headers)
+    #else
+        using nlohmann::literals::json_literals::operator "" _json; // NOLINT(misc-unused-using-decls,google-global-names-in-headers)
+        using nlohmann::literals::json_literals::operator "" _json_pointer; //NOLINT(misc-unused-using-decls,google-global-names-in-headers)
+    #endif
+#endif
+
+#include <nlohmann/detail/macro_unscope.hpp>
+
+#endif  // INCLUDE_NLOHMANN_JSON_HPP_

From c76ad774c5807e8d10587fbb8c0853846987c5ea Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Thu, 6 Mar 2025 18:30:58 +0000
Subject: [PATCH 04/33] image resize

---
 include/stb_image_resize.h | 2585 ++++++++++++++++++++++++++++++++++++
 1 file changed, 2585 insertions(+)
 create mode 100644 include/stb_image_resize.h

diff --git a/include/stb_image_resize.h b/include/stb_image_resize.h
new file mode 100644
index 0000000..bcca92c
--- /dev/null
+++ b/include/stb_image_resize.h
@@ -0,0 +1,2585 @@
+/* stb_image_resize - v0.90 - public domain image resizing
+   by Jorge L Rodriguez (@VinoBS) - 2014
+   http://github.com/nothings/stb
+
+   Written with emphasis on usability, portability, and efficiency. (No
+   SIMD or threads, so it be easily outperformed by libs that use those.)
+   Only scaling and translation is supported, no rotations or shears.
+   Easy API downsamples w/Mitchell filter, upsamples w/cubic interpolation.
+
+   COMPILING & LINKING
+      In one C/C++ file that #includes this file, do this:
+         #define STB_IMAGE_RESIZE_IMPLEMENTATION
+      before the #include. That will create the implementation in that file.
+
+   QUICKSTART
+      stbir_resize_uint8(      input_pixels , in_w , in_h , 0,
+                               output_pixels, out_w, out_h, 0, num_channels)
+      stbir_resize_float(...)
+      stbir_resize_uint8_srgb( input_pixels , in_w , in_h , 0,
+                               output_pixels, out_w, out_h, 0,
+                               num_channels , alpha_chan  , 0)
+      stbir_resize_uint8_srgb_edgemode(
+                               input_pixels , in_w , in_h , 0, 
+                               output_pixels, out_w, out_h, 0, 
+                               num_channels , alpha_chan  , 0, STBIR_EDGE_CLAMP)
+                                                            // WRAP/REFLECT/ZERO
+
+   FULL API
+      See the "header file" section of the source for API documentation.
+
+   ADDITIONAL DOCUMENTATION
+
+      SRGB & FLOATING POINT REPRESENTATION
+         The sRGB functions presume IEEE floating point. If you do not have
+         IEEE floating point, define STBIR_NON_IEEE_FLOAT. This will use
+         a slower implementation.
+
+      MEMORY ALLOCATION
+         The resize functions here perform a single memory allocation using
+         malloc. To control the memory allocation, before the #include that
+         triggers the implementation, do:
+
+            #define STBIR_MALLOC(size,context) ...
+            #define STBIR_FREE(ptr,context)   ...
+
+         Each resize function makes exactly one call to malloc/free, so to use
+         temp memory, store the temp memory in the context and return that.
+
+      ASSERT
+         Define STBIR_ASSERT(boolval) to override assert() and not use assert.h
+
+      OPTIMIZATION
+         Define STBIR_SATURATE_INT to compute clamp values in-range using
+         integer operations instead of float operations. This may be faster
+         on some platforms.
+
+      DEFAULT FILTERS
+         For functions which don't provide explicit control over what filters
+         to use, you can change the compile-time defaults with
+
+            #define STBIR_DEFAULT_FILTER_UPSAMPLE     STBIR_FILTER_something
+            #define STBIR_DEFAULT_FILTER_DOWNSAMPLE   STBIR_FILTER_something
+
+         See stbir_filter in the header-file section for the list of filters.
+
+      NEW FILTERS
+         A number of 1D filter kernels are used. For a list of
+         supported filters see the stbir_filter enum. To add a new filter,
+         write a filter function and add it to stbir__filter_info_table.
+
+      PROGRESS
+         For interactive use with slow resize operations, you can install
+         a progress-report callback:
+
+            #define STBIR_PROGRESS_REPORT(val)   some_func(val)
+
+         The parameter val is a float which goes from 0 to 1 as progress is made.
+
+         For example:
+
+            static void my_progress_report(float progress);
+            #define STBIR_PROGRESS_REPORT(val) my_progress_report(val)
+
+            #define STB_IMAGE_RESIZE_IMPLEMENTATION
+            #include "stb_image_resize.h"
+
+            static void my_progress_report(float progress)
+            {
+               printf("Progress: %f%%\n", progress*100);
+            }
+
+      MAX CHANNELS
+         If your image has more than 64 channels, define STBIR_MAX_CHANNELS
+         to the max you'll have.
+
+      ALPHA CHANNEL
+         Most of the resizing functions provide the ability to control how
+         the alpha channel of an image is processed. The important things
+         to know about this:
+
+         1. The best mathematically-behaved version of alpha to use is
+         called "premultiplied alpha", in which the other color channels
+         have had the alpha value multiplied in. If you use premultiplied
+         alpha, linear filtering (such as image resampling done by this
+         library, or performed in texture units on GPUs) does the "right
+         thing". While premultiplied alpha is standard in the movie CGI
+         industry, it is still uncommon in the videogame/real-time world.
+
+         If you linearly filter non-premultiplied alpha, strange effects
+         occur. (For example, the average of 1% opaque bright green
+         and 99% opaque black produces 50% transparent dark green when
+         non-premultiplied, whereas premultiplied it produces 50%
+         transparent near-black. The former introduces green energy
+         that doesn't exist in the source image.)
+
+         2. Artists should not edit premultiplied-alpha images; artists
+         want non-premultiplied alpha images. Thus, art tools generally output
+         non-premultiplied alpha images.
+
+         3. You will get best results in most cases by converting images
+         to premultiplied alpha before processing them mathematically.
+
+         4. If you pass the flag STBIR_FLAG_ALPHA_PREMULTIPLIED, the
+         resizer does not do anything special for the alpha channel;
+         it is resampled identically to other channels. This produces
+         the correct results for premultiplied-alpha images, but produces
+         less-than-ideal results for non-premultiplied-alpha images.
+
+         5. If you do not pass the flag STBIR_FLAG_ALPHA_PREMULTIPLIED,
+         then the resizer weights the contribution of input pixels
+         based on their alpha values, or, equivalently, it multiplies
+         the alpha value into the color channels, resamples, then divides
+         by the resultant alpha value. Input pixels which have alpha=0 do
+         not contribute at all to output pixels unless _all_ of the input
+         pixels affecting that output pixel have alpha=0, in which case
+         the result for that pixel is the same as it would be without
+         STBIR_FLAG_ALPHA_PREMULTIPLIED. However, this is only true for
+         input images in integer formats. For input images in float format,
+         input pixels with alpha=0 have no effect, and output pixels
+         which have alpha=0 will be 0 in all channels. (For float images,
+         you can manually achieve the same result by adding a tiny epsilon
+         value to the alpha channel of every image, and then subtracting
+         or clamping it at the end.)
+
+         6. You can suppress the behavior described in #5 and make
+         all-0-alpha pixels have 0 in all channels by #defining
+         STBIR_NO_ALPHA_EPSILON.
+
+         7. You can separately control whether the alpha channel is
+         interpreted as linear or affected by the colorspace. By default
+         it is linear; you almost never want to apply the colorspace.
+         (For example, graphics hardware does not apply sRGB conversion
+         to the alpha channel.)
+
+   ADDITIONAL CONTRIBUTORS
+      Sean Barrett: API design, optimizations
+         
+   REVISIONS
+      0.90 (2014-09-17) first released version
+
+   LICENSE
+      This software is in the public domain. Where that dedication is not
+      recognized, you are granted a perpetual, irrevocable license to copy
+      and modify this file as you see fit.
+
+   TODO
+      Don't decode all of the image data when only processing a partial tile
+      Don't use full-width decode buffers when only processing a partial tile
+      When processing wide images, break processing into tiles so data fits in L1 cache
+      Installable filters?
+      Resize that respects alpha test coverage
+         (Reference code: FloatImage::alphaTestCoverage and FloatImage::scaleAlphaToCoverage:
+         https://code.google.com/p/nvidia-texture-tools/source/browse/trunk/src/nvimage/FloatImage.cpp )
+*/
+
+#ifndef STBIR_INCLUDE_STB_IMAGE_RESIZE_H
+#define STBIR_INCLUDE_STB_IMAGE_RESIZE_H
+
+#ifdef _MSC_VER
+typedef unsigned char  stbir_uint8;
+typedef unsigned short stbir_uint16;
+typedef unsigned int   stbir_uint32;
+#else
+#include <stdint.h>
+typedef uint8_t  stbir_uint8;
+typedef uint16_t stbir_uint16;
+typedef uint32_t stbir_uint32;
+#endif
+
+#ifdef STB_IMAGE_RESIZE_STATIC
+#define STBIRDEF static
+#else
+#ifdef __cplusplus
+#define STBIRDEF extern "C"
+#else
+#define STBIRDEF extern
+#endif
+#endif
+
+
+//////////////////////////////////////////////////////////////////////////////
+//
+// Easy-to-use API:
+//
+//     * "input pixels" points to an array of image data with 'num_channels' channels (e.g. RGB=3, RGBA=4)
+//     * input_w is input image width (x-axis), input_h is input image height (y-axis)
+//     * stride is the offset between successive rows of image data in memory, in bytes. you can
+//       specify 0 to mean packed continuously in memory
+//     * alpha channel is treated identically to other channels.
+//     * colorspace is linear or sRGB as specified by function name
+//     * returned result is 1 for success or 0 in case of an error.
+//       #define STBIR_ASSERT() to trigger an assert on parameter validation errors.
+//     * Memory required grows approximately linearly with input and output size, but with
+//       discontinuities at input_w == output_w and input_h == output_h.
+//     * These functions use a "default" resampling filter defined at compile time. To change the filter,
+//       you can change the compile-time defaults by #defining STBIR_DEFAULT_FILTER_UPSAMPLE
+//       and STBIR_DEFAULT_FILTER_DOWNSAMPLE, or you can use the medium-complexity API.
+
+STBIRDEF int stbir_resize_uint8(     const unsigned char *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                           unsigned char *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                     int num_channels);
+
+STBIRDEF int stbir_resize_float(     const float *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                           float *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                     int num_channels);
+
+
+// The following functions interpret image data as gamma-corrected sRGB. 
+// Specify STBIR_ALPHA_CHANNEL_NONE if you have no alpha channel,
+// or otherwise provide the index of the alpha channel. Flags value
+// of 0 will probably do the right thing if you're not sure what
+// the flags mean.
+
+#define STBIR_ALPHA_CHANNEL_NONE       -1
+
+// Set this flag if your texture has premultiplied alpha. Otherwise, stbir will
+// use alpha-weighted resampling (effectively premultiplying, resampling,
+// then unpremultiplying).
+#define STBIR_FLAG_ALPHA_PREMULTIPLIED    (1 << 0)
+// The specified alpha channel should be handled as gamma-corrected value even
+// when doing sRGB operations.
+#define STBIR_FLAG_ALPHA_USES_COLORSPACE  (1 << 1)
+
+STBIRDEF int stbir_resize_uint8_srgb(const unsigned char *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                           unsigned char *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                     int num_channels, int alpha_channel, int flags);
+
+
+typedef enum
+{
+    STBIR_EDGE_CLAMP   = 1,
+    STBIR_EDGE_REFLECT = 2,
+    STBIR_EDGE_WRAP    = 3,
+    STBIR_EDGE_ZERO    = 4,
+} stbir_edge;
+
+// This function adds the ability to specify how requests to sample off the edge of the image are handled.
+STBIRDEF int stbir_resize_uint8_srgb_edgemode(const unsigned char *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                                    unsigned char *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                              int num_channels, int alpha_channel, int flags,
+                                              stbir_edge edge_wrap_mode);
+
+//////////////////////////////////////////////////////////////////////////////
+//
+// Medium-complexity API
+//
+// This extends the easy-to-use API as follows:
+//
+//     * Alpha-channel can be processed separately
+//       * If alpha_channel is not STBIR_ALPHA_CHANNEL_NONE
+//         * Alpha channel will not be gamma corrected (unless flags&STBIR_FLAG_GAMMA_CORRECT)
+//         * Filters will be weighted by alpha channel (unless flags&STBIR_FLAG_ALPHA_PREMULTIPLIED)
+//     * Filter can be selected explicitly
+//     * uint16 image type
+//     * sRGB colorspace available for all types
+//     * context parameter for passing to STBIR_MALLOC
+
+typedef enum
+{
+    STBIR_FILTER_DEFAULT      = 0,  // use same filter type that easy-to-use API chooses
+    STBIR_FILTER_BOX          = 1,  // A trapezoid w/1-pixel wide ramps, same result as box for integer scale ratios
+    STBIR_FILTER_TRIANGLE     = 2,  // On upsampling, produces same results as bilinear texture filtering
+    STBIR_FILTER_CUBICBSPLINE = 3,  // The cubic b-spline (aka Mitchell-Netrevalli with B=1,C=0), gaussian-esque
+    STBIR_FILTER_CATMULLROM   = 4,  // An interpolating cubic spline
+    STBIR_FILTER_MITCHELL     = 5,  // Mitchell-Netrevalli filter with B=1/3, C=1/3
+} stbir_filter;
+
+typedef enum
+{
+    STBIR_COLORSPACE_LINEAR,
+    STBIR_COLORSPACE_SRGB,
+
+    STBIR_MAX_COLORSPACES,
+} stbir_colorspace;
+
+// The following functions are all identical except for the type of the image data
+
+STBIRDEF int stbir_resize_uint8_generic( const unsigned char *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                               unsigned char *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                         int num_channels, int alpha_channel, int flags,
+                                         stbir_edge edge_wrap_mode, stbir_filter filter, stbir_colorspace space, 
+                                         void *alloc_context);
+
+STBIRDEF int stbir_resize_uint16_generic(const stbir_uint16 *input_pixels  , int input_w , int input_h , int input_stride_in_bytes,
+                                               stbir_uint16 *output_pixels , int output_w, int output_h, int output_stride_in_bytes,
+                                         int num_channels, int alpha_channel, int flags,
+                                         stbir_edge edge_wrap_mode, stbir_filter filter, stbir_colorspace space, 
+                                         void *alloc_context);
+
+STBIRDEF int stbir_resize_float_generic( const float *input_pixels         , int input_w , int input_h , int input_stride_in_bytes,
+                                               float *output_pixels        , int output_w, int output_h, int output_stride_in_bytes,
+                                         int num_channels, int alpha_channel, int flags,
+                                         stbir_edge edge_wrap_mode, stbir_filter filter, stbir_colorspace space, 
+                                         void *alloc_context);
+
+
+
+//////////////////////////////////////////////////////////////////////////////
+//
+// Full-complexity API
+//
+// This extends the medium API as follows:
+//
+//       * uint32 image type
+//     * not typesafe
+//     * separate filter types for each axis
+//     * separate edge modes for each axis
+//     * can specify scale explicitly for subpixel correctness
+//     * can specify image source tile using texture coordinates
+
+typedef enum
+{
+    STBIR_TYPE_UINT8 ,
+    STBIR_TYPE_UINT16,
+    STBIR_TYPE_UINT32,
+    STBIR_TYPE_FLOAT ,
+
+    STBIR_MAX_TYPES
+} stbir_datatype;
+
+STBIRDEF int stbir_resize(         const void *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                         void *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                   stbir_datatype datatype,
+                                   int num_channels, int alpha_channel, int flags,
+                                   stbir_edge edge_mode_horizontal, stbir_edge edge_mode_vertical, 
+                                   stbir_filter filter_horizontal,  stbir_filter filter_vertical,
+                                   stbir_colorspace space, void *alloc_context);
+
+STBIRDEF int stbir_resize_subpixel(const void *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                         void *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                   stbir_datatype datatype,
+                                   int num_channels, int alpha_channel, int flags,
+                                   stbir_edge edge_mode_horizontal, stbir_edge edge_mode_vertical, 
+                                   stbir_filter filter_horizontal,  stbir_filter filter_vertical,
+                                   stbir_colorspace space, void *alloc_context,
+                                   float x_scale, float y_scale,
+                                   float x_offset, float y_offset);
+
+STBIRDEF int stbir_resize_region(  const void *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                         void *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                   stbir_datatype datatype,
+                                   int num_channels, int alpha_channel, int flags,
+                                   stbir_edge edge_mode_horizontal, stbir_edge edge_mode_vertical, 
+                                   stbir_filter filter_horizontal,  stbir_filter filter_vertical,
+                                   stbir_colorspace space, void *alloc_context,
+                                   float s0, float t0, float s1, float t1);
+// (s0, t0) & (s1, t1) are the top-left and bottom right corner (uv addressing style: [0, 1]x[0, 1]) of a region of the input image to use.
+
+//
+//
+////   end header file   /////////////////////////////////////////////////////
+#endif // STBIR_INCLUDE_STB_IMAGE_RESIZE_H
+
+
+
+
+
+#ifdef STB_IMAGE_RESIZE_IMPLEMENTATION
+
+#ifndef STBIR_ASSERT
+#include <assert.h>
+#define STBIR_ASSERT(x) assert(x)
+#endif
+
+#ifdef STBIR_DEBUG
+#define STBIR__DEBUG_ASSERT STBIR_ASSERT
+#else
+#define STBIR__DEBUG_ASSERT
+#endif
+
+// If you hit this it means I haven't done it yet.
+#define STBIR__UNIMPLEMENTED(x) STBIR_ASSERT(!(x))
+
+// For memset
+#include <string.h>
+
+#include <math.h>
+
+#ifndef STBIR_MALLOC
+#include <stdlib.h>
+#define STBIR_MALLOC(size,c) malloc(size)
+#define STBIR_FREE(ptr,c)    free(ptr)
+#endif
+
+#ifndef _MSC_VER
+#ifdef __cplusplus
+#define stbir__inline inline
+#else
+#define stbir__inline
+#endif
+#else
+#define stbir__inline __forceinline
+#endif
+
+
+// should produce compiler error if size is wrong
+typedef unsigned char stbir__validate_uint32[sizeof(stbir_uint32) == 4 ? 1 : -1];
+
+#ifdef _MSC_VER
+#define STBIR__NOTUSED(v)  (void)(v)
+#else
+#define STBIR__NOTUSED(v)  (void)sizeof(v)
+#endif
+
+#define STBIR__ARRAY_SIZE(a) (sizeof((a))/sizeof((a)[0]))
+
+#ifndef STBIR_DEFAULT_FILTER_UPSAMPLE
+#define STBIR_DEFAULT_FILTER_UPSAMPLE    STBIR_FILTER_CATMULLROM
+#endif
+
+#ifndef STBIR_DEFAULT_FILTER_DOWNSAMPLE
+#define STBIR_DEFAULT_FILTER_DOWNSAMPLE  STBIR_FILTER_MITCHELL
+#endif
+
+#ifndef STBIR_PROGRESS_REPORT
+#define STBIR_PROGRESS_REPORT(float_0_to_1)
+#endif
+
+#ifndef STBIR_MAX_CHANNELS
+#define STBIR_MAX_CHANNELS 64
+#endif
+
+#if STBIR_MAX_CHANNELS > 65536
+#error "Too many channels; STBIR_MAX_CHANNELS must be no more than 65536."
+// because we store the indices in 16-bit variables
+#endif
+
+// This value is added to alpha just before premultiplication to avoid
+// zeroing out color values. It is equivalent to 2^-80. If you don't want
+// that behavior (it may interfere if you have floating point images with
+// very small alpha values) then you can define STBIR_NO_ALPHA_EPSILON to
+// disable it.
+#ifndef STBIR_ALPHA_EPSILON
+#define STBIR_ALPHA_EPSILON ((float)1 / (1 << 20) / (1 << 20) / (1 << 20) / (1 << 20))
+#endif
+
+
+
+#ifdef _MSC_VER
+#define STBIR__UNUSED_PARAM(v)  (void)(v)
+#else
+#define STBIR__UNUSED_PARAM(v)  (void)sizeof(v)
+#endif
+
+// must match stbir_datatype
+static unsigned char stbir__type_size[] = {
+    1, // STBIR_TYPE_UINT8
+    2, // STBIR_TYPE_UINT16
+    4, // STBIR_TYPE_UINT32
+    4, // STBIR_TYPE_FLOAT
+};
+
+// Kernel function centered at 0
+typedef float (stbir__kernel_fn)(float x, float scale);
+typedef float (stbir__support_fn)(float scale);
+
+typedef struct
+{
+    stbir__kernel_fn* kernel;
+    stbir__support_fn* support;
+} stbir__filter_info;
+
+// When upsampling, the contributors are which source pixels contribute.
+// When downsampling, the contributors are which destination pixels are contributed to.
+typedef struct
+{
+    int n0; // First contributing pixel
+    int n1; // Last contributing pixel
+} stbir__contributors;
+
+typedef struct
+{
+    const void* input_data;
+    int input_w;
+    int input_h;
+    int input_stride_bytes;
+
+    void* output_data;
+    int output_w;
+    int output_h;
+    int output_stride_bytes;
+
+    float s0, t0, s1, t1;
+
+    float horizontal_shift; // Units: output pixels
+    float vertical_shift;   // Units: output pixels
+    float horizontal_scale;
+    float vertical_scale;
+
+    int channels;
+    int alpha_channel;
+    stbir_uint32 flags;
+    stbir_datatype type;
+    stbir_filter horizontal_filter;
+    stbir_filter vertical_filter;
+    stbir_edge edge_horizontal;
+    stbir_edge edge_vertical;
+    stbir_colorspace colorspace;
+
+    stbir__contributors* horizontal_contributors;
+    float* horizontal_coefficients;
+
+    stbir__contributors* vertical_contributors;
+    float* vertical_coefficients;
+
+    int decode_buffer_pixels;
+    float* decode_buffer;
+
+    float* horizontal_buffer;
+
+    // cache these because ceil/floor are inexplicably showing up in profile
+    int horizontal_coefficient_width;
+    int vertical_coefficient_width;
+    int horizontal_filter_pixel_width;
+    int vertical_filter_pixel_width;
+    int horizontal_filter_pixel_margin;
+    int vertical_filter_pixel_margin;
+    int horizontal_num_contributors;
+    int vertical_num_contributors;
+
+    int ring_buffer_length_bytes; // The length of an individual entry in the ring buffer. The total number of ring buffers is stbir__get_filter_pixel_width(filter)
+    int ring_buffer_first_scanline;
+    int ring_buffer_last_scanline;
+    int ring_buffer_begin_index;
+    float* ring_buffer;
+
+    float* encode_buffer; // A temporary buffer to store floats so we don't lose precision while we do multiply-adds.
+
+    int horizontal_contributors_size;
+    int horizontal_coefficients_size;
+    int vertical_contributors_size;
+    int vertical_coefficients_size;
+    int decode_buffer_size;
+    int horizontal_buffer_size;
+    int ring_buffer_size;
+    int encode_buffer_size;
+} stbir__info;
+
+static stbir__inline int stbir__min(int a, int b)
+{
+    return a < b ? a : b;
+}
+
+static stbir__inline int stbir__max(int a, int b)
+{
+    return a > b ? a : b;
+}
+
+static stbir__inline float stbir__saturate(float x)
+{
+    if (x < 0)
+        return 0;
+
+    if (x > 1)
+        return 1;
+
+    return x;
+}
+
+#ifdef STBIR_SATURATE_INT
+static stbir__inline stbir_uint8 stbir__saturate8(int x)
+{
+    if ((unsigned int) x <= 255)
+        return x;
+
+    if (x < 0)
+        return 0;
+
+    return 255;
+}
+
+static stbir__inline stbir_uint16 stbir__saturate16(int x)
+{
+    if ((unsigned int) x <= 65535)
+        return x;
+
+    if (x < 0)
+        return 0;
+
+    return 65535;
+}
+#endif
+
+static float stbir__srgb_uchar_to_linear_float[256] = {
+    0.000000f, 0.000304f, 0.000607f, 0.000911f, 0.001214f, 0.001518f, 0.001821f, 0.002125f, 0.002428f, 0.002732f, 0.003035f,
+    0.003347f, 0.003677f, 0.004025f, 0.004391f, 0.004777f, 0.005182f, 0.005605f, 0.006049f, 0.006512f, 0.006995f, 0.007499f,
+    0.008023f, 0.008568f, 0.009134f, 0.009721f, 0.010330f, 0.010960f, 0.011612f, 0.012286f, 0.012983f, 0.013702f, 0.014444f,
+    0.015209f, 0.015996f, 0.016807f, 0.017642f, 0.018500f, 0.019382f, 0.020289f, 0.021219f, 0.022174f, 0.023153f, 0.024158f,
+    0.025187f, 0.026241f, 0.027321f, 0.028426f, 0.029557f, 0.030713f, 0.031896f, 0.033105f, 0.034340f, 0.035601f, 0.036889f,
+    0.038204f, 0.039546f, 0.040915f, 0.042311f, 0.043735f, 0.045186f, 0.046665f, 0.048172f, 0.049707f, 0.051269f, 0.052861f,
+    0.054480f, 0.056128f, 0.057805f, 0.059511f, 0.061246f, 0.063010f, 0.064803f, 0.066626f, 0.068478f, 0.070360f, 0.072272f,
+    0.074214f, 0.076185f, 0.078187f, 0.080220f, 0.082283f, 0.084376f, 0.086500f, 0.088656f, 0.090842f, 0.093059f, 0.095307f,
+    0.097587f, 0.099899f, 0.102242f, 0.104616f, 0.107023f, 0.109462f, 0.111932f, 0.114435f, 0.116971f, 0.119538f, 0.122139f,
+    0.124772f, 0.127438f, 0.130136f, 0.132868f, 0.135633f, 0.138432f, 0.141263f, 0.144128f, 0.147027f, 0.149960f, 0.152926f,
+    0.155926f, 0.158961f, 0.162029f, 0.165132f, 0.168269f, 0.171441f, 0.174647f, 0.177888f, 0.181164f, 0.184475f, 0.187821f,
+    0.191202f, 0.194618f, 0.198069f, 0.201556f, 0.205079f, 0.208637f, 0.212231f, 0.215861f, 0.219526f, 0.223228f, 0.226966f,
+    0.230740f, 0.234551f, 0.238398f, 0.242281f, 0.246201f, 0.250158f, 0.254152f, 0.258183f, 0.262251f, 0.266356f, 0.270498f,
+    0.274677f, 0.278894f, 0.283149f, 0.287441f, 0.291771f, 0.296138f, 0.300544f, 0.304987f, 0.309469f, 0.313989f, 0.318547f,
+    0.323143f, 0.327778f, 0.332452f, 0.337164f, 0.341914f, 0.346704f, 0.351533f, 0.356400f, 0.361307f, 0.366253f, 0.371238f,
+    0.376262f, 0.381326f, 0.386430f, 0.391573f, 0.396755f, 0.401978f, 0.407240f, 0.412543f, 0.417885f, 0.423268f, 0.428691f,
+    0.434154f, 0.439657f, 0.445201f, 0.450786f, 0.456411f, 0.462077f, 0.467784f, 0.473532f, 0.479320f, 0.485150f, 0.491021f,
+    0.496933f, 0.502887f, 0.508881f, 0.514918f, 0.520996f, 0.527115f, 0.533276f, 0.539480f, 0.545725f, 0.552011f, 0.558340f,
+    0.564712f, 0.571125f, 0.577581f, 0.584078f, 0.590619f, 0.597202f, 0.603827f, 0.610496f, 0.617207f, 0.623960f, 0.630757f,
+    0.637597f, 0.644480f, 0.651406f, 0.658375f, 0.665387f, 0.672443f, 0.679543f, 0.686685f, 0.693872f, 0.701102f, 0.708376f,
+    0.715694f, 0.723055f, 0.730461f, 0.737911f, 0.745404f, 0.752942f, 0.760525f, 0.768151f, 0.775822f, 0.783538f, 0.791298f,
+    0.799103f, 0.806952f, 0.814847f, 0.822786f, 0.830770f, 0.838799f, 0.846873f, 0.854993f, 0.863157f, 0.871367f, 0.879622f,
+    0.887923f, 0.896269f, 0.904661f, 0.913099f, 0.921582f, 0.930111f, 0.938686f, 0.947307f, 0.955974f, 0.964686f, 0.973445f,
+    0.982251f, 0.991102f, 1.0f
+};
+
+static float stbir__srgb_to_linear(float f)
+{
+    if (f <= 0.04045f)
+        return f / 12.92f;
+    else
+        return (float)pow((f + 0.055f) / 1.055f, 2.4f);
+}
+
+static float stbir__linear_to_srgb(float f)
+{
+    if (f <= 0.0031308f)
+        return f * 12.92f;
+    else
+        return 1.055f * (float)pow(f, 1 / 2.4f) - 0.055f;
+}
+
+#ifndef STBIR_NON_IEEE_FLOAT
+// From https://gist.github.com/rygorous/2203834
+
+typedef union
+{
+    stbir_uint32 u;
+    float f;
+} stbir__FP32;
+
+static const stbir_uint32 fp32_to_srgb8_tab4[104] = {
+    0x0073000d, 0x007a000d, 0x0080000d, 0x0087000d, 0x008d000d, 0x0094000d, 0x009a000d, 0x00a1000d,
+    0x00a7001a, 0x00b4001a, 0x00c1001a, 0x00ce001a, 0x00da001a, 0x00e7001a, 0x00f4001a, 0x0101001a,
+    0x010e0033, 0x01280033, 0x01410033, 0x015b0033, 0x01750033, 0x018f0033, 0x01a80033, 0x01c20033,
+    0x01dc0067, 0x020f0067, 0x02430067, 0x02760067, 0x02aa0067, 0x02dd0067, 0x03110067, 0x03440067,
+    0x037800ce, 0x03df00ce, 0x044600ce, 0x04ad00ce, 0x051400ce, 0x057b00c5, 0x05dd00bc, 0x063b00b5,
+    0x06970158, 0x07420142, 0x07e30130, 0x087b0120, 0x090b0112, 0x09940106, 0x0a1700fc, 0x0a9500f2,
+    0x0b0f01cb, 0x0bf401ae, 0x0ccb0195, 0x0d950180, 0x0e56016e, 0x0f0d015e, 0x0fbc0150, 0x10630143,
+    0x11070264, 0x1238023e, 0x1357021d, 0x14660201, 0x156601e9, 0x165a01d3, 0x174401c0, 0x182401af,
+    0x18fe0331, 0x1a9602fe, 0x1c1502d2, 0x1d7e02ad, 0x1ed4028d, 0x201a0270, 0x21520256, 0x227d0240,
+    0x239f0443, 0x25c003fe, 0x27bf03c4, 0x29a10392, 0x2b6a0367, 0x2d1d0341, 0x2ebe031f, 0x304d0300,
+    0x31d105b0, 0x34a80555, 0x37520507, 0x39d504c5, 0x3c37048b, 0x3e7c0458, 0x40a8042a, 0x42bd0401,
+    0x44c20798, 0x488e071e, 0x4c1c06b6, 0x4f76065d, 0x52a50610, 0x55ac05cc, 0x5892058f, 0x5b590559,
+    0x5e0c0a23, 0x631c0980, 0x67db08f6, 0x6c55087f, 0x70940818, 0x74a007bd, 0x787d076c, 0x7c330723,
+};
+ 
+static stbir_uint8 stbir__linear_to_srgb_uchar(float in)
+{
+    static const stbir__FP32 almostone = { 0x3f7fffff }; // 1-eps
+    static const stbir__FP32 minval = { (127-13) << 23 };
+    stbir_uint32 tab,bias,scale,t;
+    stbir__FP32 f;
+ 
+    // Clamp to [2^(-13), 1-eps]; these two values map to 0 and 1, respectively.
+    // The tests are carefully written so that NaNs map to 0, same as in the reference
+    // implementation.
+    if (!(in > minval.f)) // written this way to catch NaNs
+        in = minval.f;
+    if (in > almostone.f)
+        in = almostone.f;
+ 
+    // Do the table lookup and unpack bias, scale
+    f.f = in;
+    tab = fp32_to_srgb8_tab4[(f.u - minval.u) >> 20];
+    bias = (tab >> 16) << 9;
+    scale = tab & 0xffff;
+ 
+    // Grab next-highest mantissa bits and perform linear interpolation
+    t = (f.u >> 12) & 0xff;
+    return (unsigned char) ((bias + scale*t) >> 16);
+}
+
+#else
+// sRGB transition values, scaled by 1<<28
+static int stbir__srgb_offset_to_linear_scaled[256] =
+{
+            0,     40738,    122216,    203693,    285170,    366648,    448125,    529603,
+       611080,    692557,    774035,    855852,    942009,   1033024,   1128971,   1229926,
+      1335959,   1447142,   1563542,   1685229,   1812268,   1944725,   2082664,   2226148,
+      2375238,   2529996,   2690481,   2856753,   3028870,   3206888,   3390865,   3580856,
+      3776916,   3979100,   4187460,   4402049,   4622919,   4850123,   5083710,   5323731,
+      5570236,   5823273,   6082892,   6349140,   6622065,   6901714,   7188133,   7481369,
+      7781466,   8088471,   8402427,   8723380,   9051372,   9386448,   9728650,  10078021,
+     10434603,  10798439,  11169569,  11548036,  11933879,  12327139,  12727857,  13136073,
+     13551826,  13975156,  14406100,  14844697,  15290987,  15745007,  16206795,  16676389,
+     17153826,  17639142,  18132374,  18633560,  19142734,  19659934,  20185196,  20718552,
+     21260042,  21809696,  22367554,  22933648,  23508010,  24090680,  24681686,  25281066,
+     25888850,  26505076,  27129772,  27762974,  28404716,  29055026,  29713942,  30381490,
+     31057708,  31742624,  32436272,  33138682,  33849884,  34569912,  35298800,  36036568,
+     36783260,  37538896,  38303512,  39077136,  39859796,  40651528,  41452360,  42262316,
+     43081432,  43909732,  44747252,  45594016,  46450052,  47315392,  48190064,  49074096,
+     49967516,  50870356,  51782636,  52704392,  53635648,  54576432,  55526772,  56486700,
+     57456236,  58435408,  59424248,  60422780,  61431036,  62449032,  63476804,  64514376,
+     65561776,  66619028,  67686160,  68763192,  69850160,  70947088,  72053992,  73170912,
+     74297864,  75434880,  76581976,  77739184,  78906536,  80084040,  81271736,  82469648,
+     83677792,  84896192,  86124888,  87363888,  88613232,  89872928,  91143016,  92423512,
+     93714432,  95015816,  96327688,  97650056,  98982952, 100326408, 101680440, 103045072,
+    104420320, 105806224, 107202800, 108610064, 110028048, 111456776, 112896264, 114346544,
+    115807632, 117279552, 118762328, 120255976, 121760536, 123276016, 124802440, 126339832,
+    127888216, 129447616, 131018048, 132599544, 134192112, 135795792, 137410592, 139036528,
+    140673648, 142321952, 143981456, 145652208, 147334208, 149027488, 150732064, 152447968,
+    154175200, 155913792, 157663776, 159425168, 161197984, 162982240, 164777968, 166585184,
+    168403904, 170234160, 172075968, 173929344, 175794320, 177670896, 179559120, 181458992,
+    183370528, 185293776, 187228736, 189175424, 191133888, 193104112, 195086128, 197079968,
+    199085648, 201103184, 203132592, 205173888, 207227120, 209292272, 211369392, 213458480,
+    215559568, 217672656, 219797792, 221934976, 224084240, 226245600, 228419056, 230604656,
+    232802400, 235012320, 237234432, 239468736, 241715280, 243974080, 246245120, 248528464,
+    250824112, 253132064, 255452368, 257785040, 260130080, 262487520, 264857376, 267239664,
+};
+
+static stbir_uint8 stbir__linear_to_srgb_uchar(float f)
+{
+    int x = (int) (f * (1 << 28)); // has headroom so you don't need to clamp
+    int v = 0;
+    int i;
+
+    // Refine the guess with a short binary search.
+    i = v + 128; if (x >= stbir__srgb_offset_to_linear_scaled[i]) v = i;
+    i = v +  64; if (x >= stbir__srgb_offset_to_linear_scaled[i]) v = i;
+    i = v +  32; if (x >= stbir__srgb_offset_to_linear_scaled[i]) v = i;
+    i = v +  16; if (x >= stbir__srgb_offset_to_linear_scaled[i]) v = i;
+    i = v +   8; if (x >= stbir__srgb_offset_to_linear_scaled[i]) v = i;
+    i = v +   4; if (x >= stbir__srgb_offset_to_linear_scaled[i]) v = i;
+    i = v +   2; if (x >= stbir__srgb_offset_to_linear_scaled[i]) v = i;
+    i = v +   1; if (x >= stbir__srgb_offset_to_linear_scaled[i]) v = i;
+
+    return (stbir_uint8) v;
+}
+#endif
+
+static float stbir__filter_trapezoid(float x, float scale)
+{
+    float halfscale = scale / 2;
+    float t = 0.5f + halfscale;
+    STBIR__DEBUG_ASSERT(scale <= 1);
+
+    x = (float)fabs(x);
+
+    if (x >= t)
+        return 0;
+    else
+    {
+        float r = 0.5f - halfscale;
+        if (x <= r)
+            return 1;
+        else
+            return (t - x) / scale;
+    }
+}
+
+static float stbir__support_trapezoid(float scale)
+{
+    STBIR__DEBUG_ASSERT(scale <= 1);
+    return 0.5f + scale / 2;
+}
+
+static float stbir__filter_triangle(float x, float s)
+{
+    STBIR__UNUSED_PARAM(s);
+
+    x = (float)fabs(x);
+
+    if (x <= 1.0f)
+        return 1 - x;
+    else
+        return 0;
+}
+
+static float stbir__filter_cubic(float x, float s)
+{
+    STBIR__UNUSED_PARAM(s);
+
+    x = (float)fabs(x);
+
+    if (x < 1.0f)
+        return (4 + x*x*(3*x - 6))/6;
+    else if (x < 2.0f)
+        return (8 + x*(-12 + x*(6 - x)))/6;
+
+    return (0.0f);
+}
+
+static float stbir__filter_catmullrom(float x, float s)
+{
+    STBIR__UNUSED_PARAM(s);
+
+    x = (float)fabs(x);
+
+    if (x < 1.0f)
+        return 1 - x*x*(2.5f - 1.5f*x);
+    else if (x < 2.0f)
+        return 2 - x*(4 + x*(0.5f*x - 2.5f));
+
+    return (0.0f);
+}
+
+static float stbir__filter_mitchell(float x, float s)
+{
+    STBIR__UNUSED_PARAM(s);
+
+    x = (float)fabs(x);
+
+    if (x < 1.0f)
+        return (16 + x*x*(21 * x - 36))/18;
+    else if (x < 2.0f)
+        return (32 + x*(-60 + x*(36 - 7*x)))/18;
+
+    return (0.0f);
+}
+
+static float stbir__support_zero(float s)
+{
+    STBIR__UNUSED_PARAM(s);
+    return 0;
+}
+
+static float stbir__support_one(float s)
+{
+    STBIR__UNUSED_PARAM(s);
+    return 1;
+}
+
+static float stbir__support_two(float s)
+{
+    STBIR__UNUSED_PARAM(s);
+    return 2;
+}
+
+static stbir__filter_info stbir__filter_info_table[] = {
+        { NULL,                     stbir__support_zero },
+        { stbir__filter_trapezoid,  stbir__support_trapezoid },
+        { stbir__filter_triangle,   stbir__support_one },
+        { stbir__filter_cubic,      stbir__support_two },
+        { stbir__filter_catmullrom, stbir__support_two },
+        { stbir__filter_mitchell,   stbir__support_two },
+};
+
+stbir__inline static int stbir__use_upsampling(float ratio)
+{
+    return ratio > 1;
+}
+
+stbir__inline static int stbir__use_width_upsampling(stbir__info* stbir_info)
+{
+    return stbir__use_upsampling(stbir_info->horizontal_scale);
+}
+
+stbir__inline static int stbir__use_height_upsampling(stbir__info* stbir_info)
+{
+    return stbir__use_upsampling(stbir_info->vertical_scale);
+}
+
+// This is the maximum number of input samples that can affect an output sample
+// with the given filter
+static int stbir__get_filter_pixel_width(stbir_filter filter, float scale)
+{
+    STBIR_ASSERT(filter != 0);
+    STBIR_ASSERT(filter < STBIR__ARRAY_SIZE(stbir__filter_info_table));
+
+    if (stbir__use_upsampling(scale))
+        return (int)ceil(stbir__filter_info_table[filter].support(1/scale) * 2);
+    else
+        return (int)ceil(stbir__filter_info_table[filter].support(scale) * 2 / scale);
+}
+
+// This is how much to expand buffers to account for filters seeking outside
+// the image boundaries.
+static int stbir__get_filter_pixel_margin(stbir_filter filter, float scale)
+{
+    return stbir__get_filter_pixel_width(filter, scale) / 2;
+}
+
+static int stbir__get_coefficient_width(stbir_filter filter, float scale)
+{
+    if (stbir__use_upsampling(scale))
+        return (int)ceil(stbir__filter_info_table[filter].support(1 / scale) * 2);
+    else
+        return (int)ceil(stbir__filter_info_table[filter].support(scale) * 2);
+}
+
+static int stbir__get_contributors(float scale, stbir_filter filter, int input_size, int output_size)
+{
+    if (stbir__use_upsampling(scale))
+        return output_size;
+    else
+        return (input_size + stbir__get_filter_pixel_margin(filter, scale) * 2);
+}
+
+static int stbir__get_total_horizontal_coefficients(stbir__info* info)
+{
+    return info->horizontal_num_contributors
+         * stbir__get_coefficient_width      (info->horizontal_filter, info->horizontal_scale);
+}
+
+static int stbir__get_total_vertical_coefficients(stbir__info* info)
+{
+    return info->vertical_num_contributors
+         * stbir__get_coefficient_width      (info->vertical_filter, info->vertical_scale);
+}
+
+static stbir__contributors* stbir__get_contributor(stbir__contributors* contributors, int n)
+{
+    return &contributors[n];
+}
+
+// For perf reasons this code is duplicated in stbir__resample_horizontal_upsample/downsample,
+// if you change it here change it there too.
+static float* stbir__get_coefficient(float* coefficients, stbir_filter filter, float scale, int n, int c)
+{
+    int width = stbir__get_coefficient_width(filter, scale);
+    return &coefficients[width*n + c];
+}
+
+static int stbir__edge_wrap_slow(stbir_edge edge, int n, int max)
+{
+    switch (edge)
+    {
+    case STBIR_EDGE_ZERO:
+        return 0; // we'll decode the wrong pixel here, and then overwrite with 0s later
+
+    case STBIR_EDGE_CLAMP:
+        if (n < 0)
+            return 0;
+
+        if (n >= max)
+            return max - 1;
+
+        return n; // NOTREACHED
+
+    case STBIR_EDGE_REFLECT:
+    {
+        if (n < 0)
+        {
+            if (n < max)
+                return -n;
+            else
+                return max - 1;
+        }
+
+        if (n >= max)
+        {
+            int max2 = max * 2;
+            if (n >= max2)
+                return 0;
+            else
+                return max2 - n - 1;
+        }
+
+        return n; // NOTREACHED
+    }
+
+    case STBIR_EDGE_WRAP:
+        if (n >= 0)
+            return (n % max);
+        else
+        {
+            int m = (-n) % max;
+
+            if (m != 0)
+                m = max - m;
+
+            return (m);
+        }
+        return n;  // NOTREACHED
+
+    default:
+        STBIR__UNIMPLEMENTED("Unimplemented edge type");
+        return 0;
+    }
+}
+
+stbir__inline static int stbir__edge_wrap(stbir_edge edge, int n, int max)
+{
+    // avoid per-pixel switch
+    if (n >= 0 && n < max)
+        return n;
+    return stbir__edge_wrap_slow(edge, n, max);
+}
+
+// What input pixels contribute to this output pixel?
+static void stbir__calculate_sample_range_upsample(int n, float out_filter_radius, float scale_ratio, float out_shift, int* in_first_pixel, int* in_last_pixel, float* in_center_of_out)
+{
+    float out_pixel_center = (float)n + 0.5f;
+    float out_pixel_influence_lowerbound = out_pixel_center - out_filter_radius;
+    float out_pixel_influence_upperbound = out_pixel_center + out_filter_radius;
+
+    float in_pixel_influence_lowerbound = (out_pixel_influence_lowerbound + out_shift) / scale_ratio;
+    float in_pixel_influence_upperbound = (out_pixel_influence_upperbound + out_shift) / scale_ratio;
+
+    *in_center_of_out = (out_pixel_center + out_shift) / scale_ratio;
+    *in_first_pixel = (int)(floor(in_pixel_influence_lowerbound + 0.5));
+    *in_last_pixel = (int)(floor(in_pixel_influence_upperbound - 0.5));
+}
+
+// What output pixels does this input pixel contribute to?
+static void stbir__calculate_sample_range_downsample(int n, float in_pixels_radius, float scale_ratio, float out_shift, int* out_first_pixel, int* out_last_pixel, float* out_center_of_in)
+{
+    float in_pixel_center = (float)n + 0.5f;
+    float in_pixel_influence_lowerbound = in_pixel_center - in_pixels_radius;
+    float in_pixel_influence_upperbound = in_pixel_center + in_pixels_radius;
+
+    float out_pixel_influence_lowerbound = in_pixel_influence_lowerbound * scale_ratio - out_shift;
+    float out_pixel_influence_upperbound = in_pixel_influence_upperbound * scale_ratio - out_shift;
+
+    *out_center_of_in = in_pixel_center * scale_ratio - out_shift;
+    *out_first_pixel = (int)(floor(out_pixel_influence_lowerbound + 0.5));
+    *out_last_pixel = (int)(floor(out_pixel_influence_upperbound - 0.5));
+}
+
+static void stbir__calculate_coefficients_upsample(stbir__info* stbir_info, stbir_filter filter, float scale, int in_first_pixel, int in_last_pixel, float in_center_of_out, stbir__contributors* contributor, float* coefficient_group)
+{
+    int i;
+    float total_filter = 0;
+    float filter_scale;
+
+    STBIR__DEBUG_ASSERT(in_last_pixel - in_first_pixel <= (int)ceil(stbir__filter_info_table[filter].support(1/scale) * 2)); // Taken directly from stbir__get_coefficient_width() which we can't call because we don't know if we're horizontal or vertical.
+
+    contributor->n0 = in_first_pixel;
+    contributor->n1 = in_last_pixel;
+
+    STBIR__DEBUG_ASSERT(contributor->n1 >= contributor->n0);
+
+    for (i = 0; i <= in_last_pixel - in_first_pixel; i++)
+    {
+        float in_pixel_center = (float)(i + in_first_pixel) + 0.5f;
+        coefficient_group[i] = stbir__filter_info_table[filter].kernel(in_center_of_out - in_pixel_center, 1 / scale);
+
+        // If the coefficient is zero, skip it. (Don't do the <0 check here, we want the influence of those outside pixels.)
+        if (i == 0 && !coefficient_group[i])
+        {
+            contributor->n0 = ++in_first_pixel;
+            i--;
+            continue;
+        }
+
+        total_filter += coefficient_group[i];
+    }
+
+    STBIR__DEBUG_ASSERT(stbir__filter_info_table[filter].kernel((float)(in_last_pixel + 1) + 0.5f - in_center_of_out, 1/scale) == 0);
+
+    STBIR__DEBUG_ASSERT(total_filter > 0.9);
+    STBIR__DEBUG_ASSERT(total_filter < 1.1f); // Make sure it's not way off.
+
+    // Make sure the sum of all coefficients is 1.
+    filter_scale = 1 / total_filter;
+
+    for (i = 0; i <= in_last_pixel - in_first_pixel; i++)
+        coefficient_group[i] *= filter_scale;
+
+    for (i = in_last_pixel - in_first_pixel; i >= 0; i--)
+    {
+        if (coefficient_group[i])
+            break;
+
+        // This line has no weight. We can skip it.
+        contributor->n1 = contributor->n0 + i - 1;
+    }
+}
+
+static void stbir__calculate_coefficients_downsample(stbir__info* stbir_info, stbir_filter filter, float scale_ratio, int out_first_pixel, int out_last_pixel, float out_center_of_in, stbir__contributors* contributor, float* coefficient_group)
+{
+    int i;
+
+     STBIR__DEBUG_ASSERT(out_last_pixel - out_first_pixel <= (int)ceil(stbir__filter_info_table[filter].support(scale_ratio) * 2)); // Taken directly from stbir__get_coefficient_width() which we can't call because we don't know if we're horizontal or vertical.
+
+    contributor->n0 = out_first_pixel;
+    contributor->n1 = out_last_pixel;
+
+    STBIR__DEBUG_ASSERT(contributor->n1 >= contributor->n0);
+
+    for (i = 0; i <= out_last_pixel - out_first_pixel; i++)
+    {
+        float out_pixel_center = (float)(i + out_first_pixel) + 0.5f;
+        float x = out_pixel_center - out_center_of_in;
+        coefficient_group[i] = stbir__filter_info_table[filter].kernel(x, scale_ratio) * scale_ratio;
+    }
+
+    STBIR__DEBUG_ASSERT(stbir__filter_info_table[filter].kernel((float)(out_last_pixel + 1) + 0.5f - out_center_of_in, scale_ratio) == 0);
+
+    for (i = out_last_pixel - out_first_pixel; i >= 0; i--)
+    {
+        if (coefficient_group[i])
+            break;
+
+        // This line has no weight. We can skip it.
+        contributor->n1 = contributor->n0 + i - 1;
+    }
+}
+
+static void stbir__normalize_downsample_coefficients(stbir__info* stbir_info, stbir__contributors* contributors, float* coefficients, stbir_filter filter, float scale_ratio, float shift, int input_size, int output_size)
+{
+    int num_contributors = stbir__get_contributors(scale_ratio, filter, input_size, output_size);
+    int num_coefficients = stbir__get_coefficient_width(filter, scale_ratio);
+    int i, j;
+    int skip;
+
+    for (i = 0; i < output_size; i++)
+    {
+        float scale;
+        float total = 0;
+
+        for (j = 0; j < num_contributors; j++)
+        {
+            if (i >= contributors[j].n0 && i <= contributors[j].n1)
+            {
+                float coefficient = *stbir__get_coefficient(coefficients, filter, scale_ratio, j, i - contributors[j].n0);
+                total += coefficient;
+            }
+            else if (i < contributors[j].n0)
+                break;
+        }
+
+        STBIR__DEBUG_ASSERT(total > 0.9f);
+        STBIR__DEBUG_ASSERT(total < 1.1f);
+
+        scale = 1 / total;
+
+        for (j = 0; j < num_contributors; j++)
+        {
+            if (i >= contributors[j].n0 && i <= contributors[j].n1)
+                *stbir__get_coefficient(coefficients, filter, scale_ratio, j, i - contributors[j].n0) *= scale;
+            else if (i < contributors[j].n0)
+                break;
+        }
+    }
+
+    // Optimize: Skip zero coefficients and contributions outside of image bounds.
+    // Do this after normalizing because normalization depends on the n0/n1 values.
+    for (j = 0; j < num_contributors; j++)
+    {
+        int range, max, width;
+
+        skip = 0;
+        while (*stbir__get_coefficient(coefficients, filter, scale_ratio, j, skip) == 0)
+            skip++;
+
+        contributors[j].n0 += skip;
+
+        while (contributors[j].n0 < 0)
+        {
+            contributors[j].n0++;
+            skip++;
+        }
+
+        range = contributors[j].n1 - contributors[j].n0 + 1;
+        max = stbir__min(num_coefficients, range);
+
+        width = stbir__get_coefficient_width(filter, scale_ratio);
+        for (i = 0; i < max; i++)
+        {
+            if (i + skip >= width)
+                break;
+
+            *stbir__get_coefficient(coefficients, filter, scale_ratio, j, i) = *stbir__get_coefficient(coefficients, filter, scale_ratio, j, i + skip);
+        }
+
+        continue;
+    }
+
+    // Using min to avoid writing into invalid pixels.
+    for (i = 0; i < num_contributors; i++)
+        contributors[i].n1 = stbir__min(contributors[i].n1, output_size - 1);
+}
+
+// Each scan line uses the same kernel values so we should calculate the kernel
+// values once and then we can use them for every scan line.
+static void stbir__calculate_filters(stbir__info* stbir_info, stbir__contributors* contributors, float* coefficients, stbir_filter filter, float scale_ratio, float shift, int input_size, int output_size)
+{
+    int n;
+    int total_contributors = stbir__get_contributors(scale_ratio, filter, input_size, output_size);
+
+    if (stbir__use_upsampling(scale_ratio))
+    {
+        float out_pixels_radius = stbir__filter_info_table[filter].support(1 / scale_ratio) * scale_ratio;
+
+        // Looping through out pixels
+        for (n = 0; n < total_contributors; n++)
+        {
+            float in_center_of_out; // Center of the current out pixel in the in pixel space
+            int in_first_pixel, in_last_pixel;
+
+            stbir__calculate_sample_range_upsample(n, out_pixels_radius, scale_ratio, shift, &in_first_pixel, &in_last_pixel, &in_center_of_out);
+
+            stbir__calculate_coefficients_upsample(stbir_info, filter, scale_ratio, in_first_pixel, in_last_pixel, in_center_of_out, stbir__get_contributor(contributors, n), stbir__get_coefficient(coefficients, filter, scale_ratio, n, 0));
+        }
+    }
+    else
+    {
+        float in_pixels_radius = stbir__filter_info_table[filter].support(scale_ratio) / scale_ratio;
+
+        // Looping through in pixels
+        for (n = 0; n < total_contributors; n++)
+        {
+            float out_center_of_in; // Center of the current out pixel in the in pixel space
+            int out_first_pixel, out_last_pixel;
+            int n_adjusted = n - stbir__get_filter_pixel_margin(filter, scale_ratio);
+
+            stbir__calculate_sample_range_downsample(n_adjusted, in_pixels_radius, scale_ratio, shift, &out_first_pixel, &out_last_pixel, &out_center_of_in);
+
+            stbir__calculate_coefficients_downsample(stbir_info, filter, scale_ratio, out_first_pixel, out_last_pixel, out_center_of_in, stbir__get_contributor(contributors, n), stbir__get_coefficient(coefficients, filter, scale_ratio, n, 0));
+        }
+
+        stbir__normalize_downsample_coefficients(stbir_info, contributors, coefficients, filter, scale_ratio, shift, input_size, output_size);
+    }
+}
+
+static float* stbir__get_decode_buffer(stbir__info* stbir_info)
+{
+    // The 0 index of the decode buffer starts after the margin. This makes
+    // it okay to use negative indexes on the decode buffer.
+    return &stbir_info->decode_buffer[stbir_info->horizontal_filter_pixel_margin * stbir_info->channels];
+}
+
+#define STBIR__DECODE(type, colorspace) ((type) * (STBIR_MAX_COLORSPACES) + (colorspace))
+
+static void stbir__decode_scanline(stbir__info* stbir_info, int n)
+{
+    int c;
+    int channels = stbir_info->channels;
+    int alpha_channel = stbir_info->alpha_channel;
+    int type = stbir_info->type;
+    int colorspace = stbir_info->colorspace;
+    int input_w = stbir_info->input_w;
+    int input_stride_bytes = stbir_info->input_stride_bytes;
+    float* decode_buffer = stbir__get_decode_buffer(stbir_info);
+    stbir_edge edge_horizontal = stbir_info->edge_horizontal;
+    stbir_edge edge_vertical = stbir_info->edge_vertical;
+    int in_buffer_row_offset = stbir__edge_wrap(edge_vertical, n, stbir_info->input_h) * input_stride_bytes;
+    const void* input_data = (char *) stbir_info->input_data + in_buffer_row_offset;
+    int max_x = input_w + stbir_info->horizontal_filter_pixel_margin;
+    int decode = STBIR__DECODE(type, colorspace);
+
+    int x = -stbir_info->horizontal_filter_pixel_margin;
+
+    // special handling for STBIR_EDGE_ZERO because it needs to return an item that doesn't appear in the input,
+    // and we want to avoid paying overhead on every pixel if not STBIR_EDGE_ZERO
+    if (edge_vertical == STBIR_EDGE_ZERO && (n < 0 || n >= stbir_info->input_h))
+    {
+        for (; x < max_x; x++)
+            for (c = 0; c < channels; c++)
+                decode_buffer[x*channels + c] = 0;
+        return;
+    }
+
+    switch (decode)
+    {
+    case STBIR__DECODE(STBIR_TYPE_UINT8, STBIR_COLORSPACE_LINEAR):
+        for (; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+            int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels;
+            for (c = 0; c < channels; c++)
+                decode_buffer[decode_pixel_index + c] = ((float)((const unsigned char*)input_data)[input_pixel_index + c]) / 255;
+        }
+        break;
+
+    case STBIR__DECODE(STBIR_TYPE_UINT8, STBIR_COLORSPACE_SRGB):
+        for (; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+            int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels;
+            for (c = 0; c < channels; c++)
+                decode_buffer[decode_pixel_index + c] = stbir__srgb_uchar_to_linear_float[((const unsigned char*)input_data)[input_pixel_index + c]];
+
+            if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE))
+                decode_buffer[decode_pixel_index + alpha_channel] = ((float)((const unsigned char*)input_data)[input_pixel_index + alpha_channel]) / 255;
+        }
+        break;
+
+    case STBIR__DECODE(STBIR_TYPE_UINT16, STBIR_COLORSPACE_LINEAR):
+        for (; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+            int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels;
+            for (c = 0; c < channels; c++)
+                decode_buffer[decode_pixel_index + c] = ((float)((const unsigned short*)input_data)[input_pixel_index + c]) / 65535;
+        }
+        break;
+
+    case STBIR__DECODE(STBIR_TYPE_UINT16, STBIR_COLORSPACE_SRGB):
+        for (; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+            int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels;
+            for (c = 0; c < channels; c++)
+                decode_buffer[decode_pixel_index + c] = stbir__srgb_to_linear(((float)((const unsigned short*)input_data)[input_pixel_index + c]) / 65535);
+
+            if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE))
+                decode_buffer[decode_pixel_index + alpha_channel] = ((float)((const unsigned short*)input_data)[input_pixel_index + alpha_channel]) / 65535;
+        }
+        break;
+
+    case STBIR__DECODE(STBIR_TYPE_UINT32, STBIR_COLORSPACE_LINEAR):
+        for (; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+            int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels;
+            for (c = 0; c < channels; c++)
+                decode_buffer[decode_pixel_index + c] = (float)(((double)((const unsigned int*)input_data)[input_pixel_index + c]) / 4294967295);
+        }
+        break;
+
+    case STBIR__DECODE(STBIR_TYPE_UINT32, STBIR_COLORSPACE_SRGB):
+        for (; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+            int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels;
+            for (c = 0; c < channels; c++)
+                decode_buffer[decode_pixel_index + c] = stbir__srgb_to_linear((float)(((double)((const unsigned int*)input_data)[input_pixel_index + c]) / 4294967295));
+
+            if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE))
+                decode_buffer[decode_pixel_index + alpha_channel] = (float)(((double)((const unsigned int*)input_data)[input_pixel_index + alpha_channel]) / 4294967295);
+        }
+        break;
+
+    case STBIR__DECODE(STBIR_TYPE_FLOAT, STBIR_COLORSPACE_LINEAR):
+        for (; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+            int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels;
+            for (c = 0; c < channels; c++)
+                decode_buffer[decode_pixel_index + c] = ((const float*)input_data)[input_pixel_index + c];
+        }
+        break;
+
+    case STBIR__DECODE(STBIR_TYPE_FLOAT, STBIR_COLORSPACE_SRGB):
+        for (; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+            int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels;
+            for (c = 0; c < channels; c++)
+                decode_buffer[decode_pixel_index + c] = stbir__srgb_to_linear(((const float*)input_data)[input_pixel_index + c]);
+
+            if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE))
+                decode_buffer[decode_pixel_index + alpha_channel] = ((const float*)input_data)[input_pixel_index + alpha_channel];
+        }
+
+        break;
+
+    default:
+        STBIR__UNIMPLEMENTED("Unknown type/colorspace/channels combination.");
+        break;
+    }
+
+    if (!(stbir_info->flags & STBIR_FLAG_ALPHA_PREMULTIPLIED))
+    {
+        for (x = -stbir_info->horizontal_filter_pixel_margin; x < max_x; x++)
+        {
+            int decode_pixel_index = x * channels;
+
+            // If the alpha value is 0 it will clobber the color values. Make sure it's not.
+            float alpha = decode_buffer[decode_pixel_index + alpha_channel];
+#ifndef STBIR_NO_ALPHA_EPSILON
+            if (stbir_info->type != STBIR_TYPE_FLOAT) {
+                alpha += STBIR_ALPHA_EPSILON;
+                decode_buffer[decode_pixel_index + alpha_channel] = alpha;
+            }
+#endif
+            for (c = 0; c < channels; c++)
+            {
+                if (c == alpha_channel)
+                    continue;
+
+                decode_buffer[decode_pixel_index + c] *= alpha;
+            }
+        }
+    }
+
+    if (edge_horizontal == STBIR_EDGE_ZERO)
+    {
+        for (x = -stbir_info->horizontal_filter_pixel_margin; x < 0; x++)
+        {
+            for (c = 0; c < channels; c++)
+                decode_buffer[x*channels + c] = 0;
+        }
+        for (x = input_w; x < max_x; x++)
+        {
+            for (c = 0; c < channels; c++)
+                decode_buffer[x*channels + c] = 0;
+        }
+    }
+}
+
+static float* stbir__get_ring_buffer_entry(float* ring_buffer, int index, int ring_buffer_length)
+{
+    return &ring_buffer[index * ring_buffer_length];
+}
+
+static float* stbir__add_empty_ring_buffer_entry(stbir__info* stbir_info, int n)
+{
+    int ring_buffer_index;
+    float* ring_buffer;
+
+    if (stbir_info->ring_buffer_begin_index < 0)
+    {
+        ring_buffer_index = stbir_info->ring_buffer_begin_index = 0;
+        stbir_info->ring_buffer_first_scanline = n;
+    }
+    else
+    {
+        ring_buffer_index = (stbir_info->ring_buffer_begin_index + (stbir_info->ring_buffer_last_scanline - stbir_info->ring_buffer_first_scanline) + 1) % stbir_info->vertical_filter_pixel_width;
+        STBIR__DEBUG_ASSERT(ring_buffer_index != stbir_info->ring_buffer_begin_index);
+    }
+
+    ring_buffer = stbir__get_ring_buffer_entry(stbir_info->ring_buffer, ring_buffer_index, stbir_info->ring_buffer_length_bytes / sizeof(float));
+    memset(ring_buffer, 0, stbir_info->ring_buffer_length_bytes);
+
+    stbir_info->ring_buffer_last_scanline = n;
+
+    return ring_buffer;
+}
+
+
+static void stbir__resample_horizontal_upsample(stbir__info* stbir_info, int n, float* output_buffer)
+{
+    int x, k;
+    int output_w = stbir_info->output_w;
+    int kernel_pixel_width = stbir_info->horizontal_filter_pixel_width;
+    int channels = stbir_info->channels;
+    float* decode_buffer = stbir__get_decode_buffer(stbir_info);
+    stbir__contributors* horizontal_contributors = stbir_info->horizontal_contributors;
+    float* horizontal_coefficients = stbir_info->horizontal_coefficients;
+    int coefficient_width = stbir_info->horizontal_coefficient_width;
+
+    for (x = 0; x < output_w; x++)
+    {
+        int n0 = horizontal_contributors[x].n0;
+        int n1 = horizontal_contributors[x].n1;
+
+        int out_pixel_index = x * channels;
+        int coefficient_group = coefficient_width * x;
+        int coefficient_counter = 0;
+
+        STBIR__DEBUG_ASSERT(n1 >= n0);
+        STBIR__DEBUG_ASSERT(n0 >= -stbir_info->horizontal_filter_pixel_margin);
+        STBIR__DEBUG_ASSERT(n1 >= -stbir_info->horizontal_filter_pixel_margin);
+        STBIR__DEBUG_ASSERT(n0 < stbir_info->input_w + stbir_info->horizontal_filter_pixel_margin);
+        STBIR__DEBUG_ASSERT(n1 < stbir_info->input_w + stbir_info->horizontal_filter_pixel_margin);
+
+        switch (channels) {
+            case 1:
+                for (k = n0; k <= n1; k++)
+                {
+                    int in_pixel_index = k * 1;
+                    float coefficient = horizontal_coefficients[coefficient_group + coefficient_counter++];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    output_buffer[out_pixel_index + 0] += decode_buffer[in_pixel_index + 0] * coefficient;
+                }
+                break;
+            case 2:
+                for (k = n0; k <= n1; k++)
+                {
+                    int in_pixel_index = k * 2;
+                    float coefficient = horizontal_coefficients[coefficient_group + coefficient_counter++];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    output_buffer[out_pixel_index + 0] += decode_buffer[in_pixel_index + 0] * coefficient;
+                    output_buffer[out_pixel_index + 1] += decode_buffer[in_pixel_index + 1] * coefficient;
+                }
+                break;
+            case 3:
+                for (k = n0; k <= n1; k++)
+                {
+                    int in_pixel_index = k * 3;
+                    float coefficient = horizontal_coefficients[coefficient_group + coefficient_counter++];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    output_buffer[out_pixel_index + 0] += decode_buffer[in_pixel_index + 0] * coefficient;
+                    output_buffer[out_pixel_index + 1] += decode_buffer[in_pixel_index + 1] * coefficient;
+                    output_buffer[out_pixel_index + 2] += decode_buffer[in_pixel_index + 2] * coefficient;
+                }
+                break;
+            case 4:
+                for (k = n0; k <= n1; k++)
+                {
+                    int in_pixel_index = k * 4;
+                    float coefficient = horizontal_coefficients[coefficient_group + coefficient_counter++];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    output_buffer[out_pixel_index + 0] += decode_buffer[in_pixel_index + 0] * coefficient;
+                    output_buffer[out_pixel_index + 1] += decode_buffer[in_pixel_index + 1] * coefficient;
+                    output_buffer[out_pixel_index + 2] += decode_buffer[in_pixel_index + 2] * coefficient;
+                    output_buffer[out_pixel_index + 3] += decode_buffer[in_pixel_index + 3] * coefficient;
+                }
+                break;
+            default:
+                for (k = n0; k <= n1; k++)
+                {
+                    int in_pixel_index = k * channels;
+                    float coefficient = horizontal_coefficients[coefficient_group + coefficient_counter++];
+                    int c;
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    for (c = 0; c < channels; c++)
+                        output_buffer[out_pixel_index + c] += decode_buffer[in_pixel_index + c] * coefficient;
+                }
+                break;
+        }
+    }
+}
+
+static void stbir__resample_horizontal_downsample(stbir__info* stbir_info, int n, float* output_buffer)
+{
+    int x, k;
+    int input_w = stbir_info->input_w;
+    int output_w = stbir_info->output_w;
+    int kernel_pixel_width = stbir_info->horizontal_filter_pixel_width;
+    int channels = stbir_info->channels;
+    float* decode_buffer = stbir__get_decode_buffer(stbir_info);
+    stbir__contributors* horizontal_contributors = stbir_info->horizontal_contributors;
+    float* horizontal_coefficients = stbir_info->horizontal_coefficients;
+    int coefficient_width = stbir_info->horizontal_coefficient_width;
+    int filter_pixel_margin = stbir_info->horizontal_filter_pixel_margin;
+    int max_x = input_w + filter_pixel_margin * 2;
+
+    STBIR__DEBUG_ASSERT(!stbir__use_width_upsampling(stbir_info));
+
+    switch (channels) {
+        case 1:
+            for (x = 0; x < max_x; x++)
+            {
+                int n0 = horizontal_contributors[x].n0;
+                int n1 = horizontal_contributors[x].n1;
+
+                int in_x = x - filter_pixel_margin;
+                int in_pixel_index = in_x * 1;
+                int max_n = n1;
+                int coefficient_group = coefficient_width * x;
+
+                for (k = n0; k <= max_n; k++)
+                {
+                    int out_pixel_index = k * 1;
+                    float coefficient = horizontal_coefficients[coefficient_group + k - n0];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    output_buffer[out_pixel_index + 0] += decode_buffer[in_pixel_index + 0] * coefficient;
+                }
+            }
+            break;
+
+        case 2:
+            for (x = 0; x < max_x; x++)
+            {
+                int n0 = horizontal_contributors[x].n0;
+                int n1 = horizontal_contributors[x].n1;
+
+                int in_x = x - filter_pixel_margin;
+                int in_pixel_index = in_x * 2;
+                int max_n = n1;
+                int coefficient_group = coefficient_width * x;
+
+                for (k = n0; k <= max_n; k++)
+                {
+                    int out_pixel_index = k * 2;
+                    float coefficient = horizontal_coefficients[coefficient_group + k - n0];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    output_buffer[out_pixel_index + 0] += decode_buffer[in_pixel_index + 0] * coefficient;
+                    output_buffer[out_pixel_index + 1] += decode_buffer[in_pixel_index + 1] * coefficient;
+                }
+            }
+            break;
+
+        case 3:
+            for (x = 0; x < max_x; x++)
+            {
+                int n0 = horizontal_contributors[x].n0;
+                int n1 = horizontal_contributors[x].n1;
+
+                int in_x = x - filter_pixel_margin;
+                int in_pixel_index = in_x * 3;
+                int max_n = n1;
+                int coefficient_group = coefficient_width * x;
+
+                for (k = n0; k <= max_n; k++)
+                {
+                    int out_pixel_index = k * 3;
+                    float coefficient = horizontal_coefficients[coefficient_group + k - n0];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    output_buffer[out_pixel_index + 0] += decode_buffer[in_pixel_index + 0] * coefficient;
+                    output_buffer[out_pixel_index + 1] += decode_buffer[in_pixel_index + 1] * coefficient;
+                    output_buffer[out_pixel_index + 2] += decode_buffer[in_pixel_index + 2] * coefficient;
+                }
+            }
+            break;
+
+        case 4:
+            for (x = 0; x < max_x; x++)
+            {
+                int n0 = horizontal_contributors[x].n0;
+                int n1 = horizontal_contributors[x].n1;
+
+                int in_x = x - filter_pixel_margin;
+                int in_pixel_index = in_x * 4;
+                int max_n = n1;
+                int coefficient_group = coefficient_width * x;
+
+                for (k = n0; k <= max_n; k++)
+                {
+                    int out_pixel_index = k * 4;
+                    float coefficient = horizontal_coefficients[coefficient_group + k - n0];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    output_buffer[out_pixel_index + 0] += decode_buffer[in_pixel_index + 0] * coefficient;
+                    output_buffer[out_pixel_index + 1] += decode_buffer[in_pixel_index + 1] * coefficient;
+                    output_buffer[out_pixel_index + 2] += decode_buffer[in_pixel_index + 2] * coefficient;
+                    output_buffer[out_pixel_index + 3] += decode_buffer[in_pixel_index + 3] * coefficient;
+                }
+            }
+            break;
+
+        default:
+            for (x = 0; x < max_x; x++)
+            {
+                int n0 = horizontal_contributors[x].n0;
+                int n1 = horizontal_contributors[x].n1;
+
+                int in_x = x - filter_pixel_margin;
+                int in_pixel_index = in_x * channels;
+                int max_n = n1;
+                int coefficient_group = coefficient_width * x;
+
+                for (k = n0; k <= max_n; k++)
+                {
+                    int c;
+                    int out_pixel_index = k * channels;
+                    float coefficient = horizontal_coefficients[coefficient_group + k - n0];
+                    STBIR__DEBUG_ASSERT(coefficient != 0);
+                    for (c = 0; c < channels; c++)
+                        output_buffer[out_pixel_index + c] += decode_buffer[in_pixel_index + c] * coefficient;
+                }
+            }
+            break;
+    }
+}
+
+static void stbir__decode_and_resample_upsample(stbir__info* stbir_info, int n)
+{
+    // Decode the nth scanline from the source image into the decode buffer.
+    stbir__decode_scanline(stbir_info, n);
+
+    // Now resample it into the ring buffer.
+    if (stbir__use_width_upsampling(stbir_info))
+        stbir__resample_horizontal_upsample(stbir_info, n, stbir__add_empty_ring_buffer_entry(stbir_info, n));
+    else
+        stbir__resample_horizontal_downsample(stbir_info, n, stbir__add_empty_ring_buffer_entry(stbir_info, n));
+
+    // Now it's sitting in the ring buffer ready to be used as source for the vertical sampling.
+}
+
+static void stbir__decode_and_resample_downsample(stbir__info* stbir_info, int n)
+{
+    // Decode the nth scanline from the source image into the decode buffer.
+    stbir__decode_scanline(stbir_info, n);
+
+    memset(stbir_info->horizontal_buffer, 0, stbir_info->output_w * stbir_info->channels * sizeof(float));
+
+    // Now resample it into the horizontal buffer.
+    if (stbir__use_width_upsampling(stbir_info))
+        stbir__resample_horizontal_upsample(stbir_info, n, stbir_info->horizontal_buffer);
+    else
+        stbir__resample_horizontal_downsample(stbir_info, n, stbir_info->horizontal_buffer);
+
+    // Now it's sitting in the horizontal buffer ready to be distributed into the ring buffers.
+}
+
+// Get the specified scan line from the ring buffer.
+static float* stbir__get_ring_buffer_scanline(int get_scanline, float* ring_buffer, int begin_index, int first_scanline, int ring_buffer_size, int ring_buffer_length)
+{
+    int ring_buffer_index = (begin_index + (get_scanline - first_scanline)) % ring_buffer_size;
+    return stbir__get_ring_buffer_entry(ring_buffer, ring_buffer_index, ring_buffer_length);
+}
+
+
+static void stbir__encode_scanline(stbir__info* stbir_info, int num_pixels, void *output_buffer, float *encode_buffer, int channels, int alpha_channel, int decode)
+{
+    int x;
+    int n;
+    int num_nonalpha;
+    stbir_uint16 nonalpha[STBIR_MAX_CHANNELS];
+
+    if (!(stbir_info->flags&STBIR_FLAG_ALPHA_PREMULTIPLIED))
+    {
+        for (x=0; x < num_pixels; ++x)
+        {
+            int pixel_index = x*channels;
+
+            float alpha = encode_buffer[pixel_index + alpha_channel];
+            float reciprocal_alpha = alpha ? 1.0f / alpha : 0;
+
+            // unrolling this produced a 1% slowdown upscaling a large RGBA linear-space image on my machine - stb
+            for (n = 0; n < channels; n++)
+                if (n != alpha_channel)
+                    encode_buffer[pixel_index + n] *= reciprocal_alpha;
+
+            // We added in a small epsilon to prevent the color channel from being deleted with zero alpha.
+            // Because we only add it for integer types, it will automatically be discarded on integer
+            // conversion, so we don't need to subtract it back out (which would be problematic for
+            // numeric precision reasons).
+        }
+    }
+
+    // build a table of all channels that need colorspace correction, so
+    // we don't perform colorspace correction on channels that don't need it.
+    for (x=0, num_nonalpha=0; x < channels; ++x)
+        if (x != alpha_channel || (stbir_info->flags & STBIR_FLAG_ALPHA_USES_COLORSPACE))
+            nonalpha[num_nonalpha++] = x;
+
+    #define STBIR__ROUND_INT(f)    ((int)          ((f)+0.5))
+    #define STBIR__ROUND_UINT(f)   ((stbir_uint32) ((f)+0.5))
+
+    #ifdef STBIR__SATURATE_INT
+    #define STBIR__ENCODE_LINEAR8(f)   stbir__saturate8 (STBIR__ROUND_INT((f) * 255  ))
+    #define STBIR__ENCODE_LINEAR16(f)  stbir__saturate16(STBIR__ROUND_INT((f) * 65535))
+    #else
+    #define STBIR__ENCODE_LINEAR8(f)   (unsigned char ) STBIR__ROUND_INT(stbir__saturate(f) * 255  )
+    #define STBIR__ENCODE_LINEAR16(f)  (unsigned short) STBIR__ROUND_INT(stbir__saturate(f) * 65535)
+    #endif
+
+    switch (decode)
+    {
+        case STBIR__DECODE(STBIR_TYPE_UINT8, STBIR_COLORSPACE_LINEAR):
+            for (x=0; x < num_pixels; ++x)
+            {
+                int pixel_index = x*channels;
+
+                for (n = 0; n < channels; n++)
+                {
+                    int index = pixel_index + n;
+                    ((unsigned char*)output_buffer)[index] = STBIR__ENCODE_LINEAR8(encode_buffer[index]);
+                }
+            }
+            break;
+
+        case STBIR__DECODE(STBIR_TYPE_UINT8, STBIR_COLORSPACE_SRGB):
+            for (x=0; x < num_pixels; ++x)
+            {
+                int pixel_index = x*channels;
+
+                for (n = 0; n < num_nonalpha; n++)
+                {
+                    int index = pixel_index + nonalpha[n];
+                    ((unsigned char*)output_buffer)[index] = stbir__linear_to_srgb_uchar(encode_buffer[index]);
+                }
+
+                if (!(stbir_info->flags & STBIR_FLAG_ALPHA_USES_COLORSPACE))
+                    ((unsigned char *)output_buffer)[pixel_index + alpha_channel] = STBIR__ENCODE_LINEAR8(encode_buffer[pixel_index+alpha_channel]);
+            }
+            break;
+
+        case STBIR__DECODE(STBIR_TYPE_UINT16, STBIR_COLORSPACE_LINEAR):
+            for (x=0; x < num_pixels; ++x)
+            {
+                int pixel_index = x*channels;
+
+                for (n = 0; n < channels; n++)
+                {
+                    int index = pixel_index + n;
+                    ((unsigned short*)output_buffer)[index] = STBIR__ENCODE_LINEAR16(encode_buffer[index]);
+                }
+            }
+            break;
+
+        case STBIR__DECODE(STBIR_TYPE_UINT16, STBIR_COLORSPACE_SRGB):
+            for (x=0; x < num_pixels; ++x)
+            {
+                int pixel_index = x*channels;
+
+                for (n = 0; n < num_nonalpha; n++)
+                {
+                    int index = pixel_index + nonalpha[n];
+                    ((unsigned short*)output_buffer)[index] = (unsigned short)STBIR__ROUND_INT(stbir__linear_to_srgb(stbir__saturate(encode_buffer[index])) * 65535);
+                }
+
+                if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE))
+                    ((unsigned short*)output_buffer)[pixel_index + alpha_channel] = STBIR__ENCODE_LINEAR16(encode_buffer[pixel_index + alpha_channel]);
+            }
+
+            break;
+
+        case STBIR__DECODE(STBIR_TYPE_UINT32, STBIR_COLORSPACE_LINEAR):
+            for (x=0; x < num_pixels; ++x)
+            {
+                int pixel_index = x*channels;
+
+                for (n = 0; n < channels; n++)
+                {
+                    int index = pixel_index + n;
+                    ((unsigned int*)output_buffer)[index] = (unsigned int)STBIR__ROUND_UINT(((double)stbir__saturate(encode_buffer[index])) * 4294967295);
+                }
+            }
+            break;
+
+        case STBIR__DECODE(STBIR_TYPE_UINT32, STBIR_COLORSPACE_SRGB):
+            for (x=0; x < num_pixels; ++x)
+            {
+                int pixel_index = x*channels;
+
+                for (n = 0; n < num_nonalpha; n++)
+                {
+                    int index = pixel_index + nonalpha[n];
+                    ((unsigned int*)output_buffer)[index] = (unsigned int)STBIR__ROUND_UINT(((double)stbir__linear_to_srgb(stbir__saturate(encode_buffer[index]))) * 4294967295);
+                }
+
+                if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE))
+                    ((unsigned int*)output_buffer)[pixel_index + alpha_channel] = (unsigned int)STBIR__ROUND_INT(((double)stbir__saturate(encode_buffer[pixel_index + alpha_channel])) * 4294967295);
+            }
+            break;
+
+        case STBIR__DECODE(STBIR_TYPE_FLOAT, STBIR_COLORSPACE_LINEAR):
+            for (x=0; x < num_pixels; ++x)
+            {
+                int pixel_index = x*channels;
+
+                for (n = 0; n < channels; n++)
+                {
+                    int index = pixel_index + n;
+                    ((float*)output_buffer)[index] = encode_buffer[index];
+                }
+            }
+            break;
+
+        case STBIR__DECODE(STBIR_TYPE_FLOAT, STBIR_COLORSPACE_SRGB):
+            for (x=0; x < num_pixels; ++x)
+            {
+                int pixel_index = x*channels;
+
+                for (n = 0; n < num_nonalpha; n++)
+                {
+                    int index = pixel_index + nonalpha[n];
+                    ((float*)output_buffer)[index] = stbir__linear_to_srgb(encode_buffer[index]);
+                }
+
+                if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE))
+                    ((float*)output_buffer)[pixel_index + alpha_channel] = encode_buffer[pixel_index + alpha_channel];
+            }
+            break;
+
+        default:
+            STBIR__UNIMPLEMENTED("Unknown type/colorspace/channels combination.");
+            break;
+    }
+}
+
+static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, int in_first_scanline, int in_last_scanline, float in_center_of_out)
+{
+    int x, k;
+    int output_w = stbir_info->output_w;
+    stbir__contributors* vertical_contributors = stbir_info->vertical_contributors;
+    float* vertical_coefficients = stbir_info->vertical_coefficients;
+    int channels = stbir_info->channels;
+    int alpha_channel = stbir_info->alpha_channel;
+    int type = stbir_info->type;
+    int colorspace = stbir_info->colorspace;
+    int kernel_pixel_width = stbir_info->vertical_filter_pixel_width;
+    void* output_data = stbir_info->output_data;
+    float* encode_buffer = stbir_info->encode_buffer;
+    int decode = STBIR__DECODE(type, colorspace);
+    int coefficient_width = stbir_info->vertical_coefficient_width;
+    int coefficient_counter;
+    int contributor = n;
+
+    float* ring_buffer = stbir_info->ring_buffer;
+    int ring_buffer_begin_index = stbir_info->ring_buffer_begin_index;
+    int ring_buffer_first_scanline = stbir_info->ring_buffer_first_scanline;
+    int ring_buffer_last_scanline = stbir_info->ring_buffer_last_scanline;
+    int ring_buffer_length = stbir_info->ring_buffer_length_bytes/sizeof(float);
+
+    int n0,n1, output_row_start;
+    int coefficient_group = coefficient_width * contributor;
+
+    n0 = vertical_contributors[contributor].n0;
+    n1 = vertical_contributors[contributor].n1;
+
+    output_row_start = n * stbir_info->output_stride_bytes;
+
+    STBIR__DEBUG_ASSERT(stbir__use_height_upsampling(stbir_info));
+
+    memset(encode_buffer, 0, output_w * sizeof(float) * channels);
+
+    // I tried reblocking this for better cache usage of encode_buffer
+    // (using x_outer, k, x_inner), but it lost speed. -- stb
+
+    coefficient_counter = 0;
+    switch (channels) {
+        case 1:
+            for (k = n0; k <= n1; k++)
+            {
+                int coefficient_index = coefficient_counter++;
+                float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length);
+                float coefficient = vertical_coefficients[coefficient_group + coefficient_index];
+                for (x = 0; x < output_w; ++x)
+                {
+                    int in_pixel_index = x * 1;
+                    encode_buffer[in_pixel_index + 0] += ring_buffer_entry[in_pixel_index + 0] * coefficient;
+                }
+            }
+            break;
+        case 2:
+            for (k = n0; k <= n1; k++)
+            {
+                int coefficient_index = coefficient_counter++;
+                float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length);
+                float coefficient = vertical_coefficients[coefficient_group + coefficient_index];
+                for (x = 0; x < output_w; ++x)
+                {
+                    int in_pixel_index = x * 2;
+                    encode_buffer[in_pixel_index + 0] += ring_buffer_entry[in_pixel_index + 0] * coefficient;
+                    encode_buffer[in_pixel_index + 1] += ring_buffer_entry[in_pixel_index + 1] * coefficient;
+                }
+            }
+            break;
+        case 3:
+            for (k = n0; k <= n1; k++)
+            {
+                int coefficient_index = coefficient_counter++;
+                float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length);
+                float coefficient = vertical_coefficients[coefficient_group + coefficient_index];
+                for (x = 0; x < output_w; ++x)
+                {
+                    int in_pixel_index = x * 3;
+                    encode_buffer[in_pixel_index + 0] += ring_buffer_entry[in_pixel_index + 0] * coefficient;
+                    encode_buffer[in_pixel_index + 1] += ring_buffer_entry[in_pixel_index + 1] * coefficient;
+                    encode_buffer[in_pixel_index + 2] += ring_buffer_entry[in_pixel_index + 2] * coefficient;
+                }
+            }
+            break;
+        case 4:
+            for (k = n0; k <= n1; k++)
+            {
+                int coefficient_index = coefficient_counter++;
+                float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length);
+                float coefficient = vertical_coefficients[coefficient_group + coefficient_index];
+                for (x = 0; x < output_w; ++x)
+                {
+                    int in_pixel_index = x * 4;
+                    encode_buffer[in_pixel_index + 0] += ring_buffer_entry[in_pixel_index + 0] * coefficient;
+                    encode_buffer[in_pixel_index + 1] += ring_buffer_entry[in_pixel_index + 1] * coefficient;
+                    encode_buffer[in_pixel_index + 2] += ring_buffer_entry[in_pixel_index + 2] * coefficient;
+                    encode_buffer[in_pixel_index + 3] += ring_buffer_entry[in_pixel_index + 3] * coefficient;
+                }
+            }
+            break;
+        default:
+            for (k = n0; k <= n1; k++)
+            {
+                int coefficient_index = coefficient_counter++;
+                float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length);
+                float coefficient = vertical_coefficients[coefficient_group + coefficient_index];
+                for (x = 0; x < output_w; ++x)
+                {
+                    int in_pixel_index = x * channels;
+                    int c;
+                    for (c = 0; c < channels; c++)
+                        encode_buffer[in_pixel_index + c] += ring_buffer_entry[in_pixel_index + c] * coefficient;
+                }
+            }
+            break;
+    }
+    stbir__encode_scanline(stbir_info, output_w, (char *) output_data + output_row_start, encode_buffer, channels, alpha_channel, decode);
+}
+
+static void stbir__resample_vertical_downsample(stbir__info* stbir_info, int n, int in_first_scanline, int in_last_scanline, float in_center_of_out)
+{
+    int x, k;
+    int output_w = stbir_info->output_w;
+    int output_h = stbir_info->output_h;
+    stbir__contributors* vertical_contributors = stbir_info->vertical_contributors;
+    float* vertical_coefficients = stbir_info->vertical_coefficients;
+    int channels = stbir_info->channels;
+    int kernel_pixel_width = stbir_info->vertical_filter_pixel_width;
+    void* output_data = stbir_info->output_data;
+    float* horizontal_buffer = stbir_info->horizontal_buffer;
+    int coefficient_width = stbir_info->vertical_coefficient_width;
+    int contributor = n + stbir_info->vertical_filter_pixel_margin;
+
+    float* ring_buffer = stbir_info->ring_buffer;
+    int ring_buffer_begin_index = stbir_info->ring_buffer_begin_index;
+    int ring_buffer_first_scanline = stbir_info->ring_buffer_first_scanline;
+    int ring_buffer_last_scanline = stbir_info->ring_buffer_last_scanline;
+    int ring_buffer_length = stbir_info->ring_buffer_length_bytes/sizeof(float);
+    int n0,n1;
+
+    n0 = vertical_contributors[contributor].n0;
+    n1 = vertical_contributors[contributor].n1;
+
+    STBIR__DEBUG_ASSERT(!stbir__use_height_upsampling(stbir_info));
+
+    for (k = n0; k <= n1; k++)
+    {
+        int coefficient_index = k - n0;
+        int coefficient_group = coefficient_width * contributor;
+        float coefficient = vertical_coefficients[coefficient_group + coefficient_index];
+
+        float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length);
+
+        switch (channels) {
+            case 1:
+                for (x = 0; x < output_w; x++)
+                {
+                    int in_pixel_index = x * 1;
+                    ring_buffer_entry[in_pixel_index + 0] += horizontal_buffer[in_pixel_index + 0] * coefficient;
+                }
+                break;
+            case 2:
+                for (x = 0; x < output_w; x++)
+                {
+                    int in_pixel_index = x * 2;
+                    ring_buffer_entry[in_pixel_index + 0] += horizontal_buffer[in_pixel_index + 0] * coefficient;
+                    ring_buffer_entry[in_pixel_index + 1] += horizontal_buffer[in_pixel_index + 1] * coefficient;
+                }
+                break;
+            case 3:
+                for (x = 0; x < output_w; x++)
+                {
+                    int in_pixel_index = x * 3;
+                    ring_buffer_entry[in_pixel_index + 0] += horizontal_buffer[in_pixel_index + 0] * coefficient;
+                    ring_buffer_entry[in_pixel_index + 1] += horizontal_buffer[in_pixel_index + 1] * coefficient;
+                    ring_buffer_entry[in_pixel_index + 2] += horizontal_buffer[in_pixel_index + 2] * coefficient;
+                }
+                break;
+            case 4:
+                for (x = 0; x < output_w; x++)
+                {
+                    int in_pixel_index = x * 4;
+                    ring_buffer_entry[in_pixel_index + 0] += horizontal_buffer[in_pixel_index + 0] * coefficient;
+                    ring_buffer_entry[in_pixel_index + 1] += horizontal_buffer[in_pixel_index + 1] * coefficient;
+                    ring_buffer_entry[in_pixel_index + 2] += horizontal_buffer[in_pixel_index + 2] * coefficient;
+                    ring_buffer_entry[in_pixel_index + 3] += horizontal_buffer[in_pixel_index + 3] * coefficient;
+                }
+                break;
+            default:
+                for (x = 0; x < output_w; x++)
+                {
+                    int in_pixel_index = x * channels;
+
+                    int c;
+                    for (c = 0; c < channels; c++)
+                        ring_buffer_entry[in_pixel_index + c] += horizontal_buffer[in_pixel_index + c] * coefficient;
+                }
+                break;
+        }
+    }
+}
+
+static void stbir__buffer_loop_upsample(stbir__info* stbir_info)
+{
+    int y;
+    float scale_ratio = stbir_info->vertical_scale;
+    float out_scanlines_radius = stbir__filter_info_table[stbir_info->vertical_filter].support(1/scale_ratio) * scale_ratio;
+
+    STBIR__DEBUG_ASSERT(stbir__use_height_upsampling(stbir_info));
+
+    for (y = 0; y < stbir_info->output_h; y++)
+    {
+        float in_center_of_out = 0; // Center of the current out scanline in the in scanline space
+        int in_first_scanline = 0, in_last_scanline = 0;
+
+        stbir__calculate_sample_range_upsample(y, out_scanlines_radius, scale_ratio, stbir_info->vertical_shift, &in_first_scanline, &in_last_scanline, &in_center_of_out);
+
+        STBIR__DEBUG_ASSERT(in_last_scanline - in_first_scanline <= stbir_info->vertical_filter_pixel_width);
+
+        if (stbir_info->ring_buffer_begin_index >= 0)
+        {
+            // Get rid of whatever we don't need anymore.
+            while (in_first_scanline > stbir_info->ring_buffer_first_scanline)
+            {
+                if (stbir_info->ring_buffer_first_scanline == stbir_info->ring_buffer_last_scanline)
+                {
+                    // We just popped the last scanline off the ring buffer.
+                    // Reset it to the empty state.
+                    stbir_info->ring_buffer_begin_index = -1;
+                    stbir_info->ring_buffer_first_scanline = 0;
+                    stbir_info->ring_buffer_last_scanline = 0;
+                    break;
+                }
+                else
+                {
+                    stbir_info->ring_buffer_first_scanline++;
+                    stbir_info->ring_buffer_begin_index = (stbir_info->ring_buffer_begin_index + 1) % stbir_info->vertical_filter_pixel_width;
+                }
+            }
+        }
+
+        // Load in new ones.
+        if (stbir_info->ring_buffer_begin_index < 0)
+            stbir__decode_and_resample_upsample(stbir_info, in_first_scanline);
+
+        while (in_last_scanline > stbir_info->ring_buffer_last_scanline)
+            stbir__decode_and_resample_upsample(stbir_info, stbir_info->ring_buffer_last_scanline + 1);
+
+        // Now all buffers should be ready to write a row of vertical sampling.
+        stbir__resample_vertical_upsample(stbir_info, y, in_first_scanline, in_last_scanline, in_center_of_out);
+
+        STBIR_PROGRESS_REPORT((float)y / stbir_info->output_h);
+    }
+}
+
+static void stbir__empty_ring_buffer(stbir__info* stbir_info, int first_necessary_scanline)
+{
+    int output_stride_bytes = stbir_info->output_stride_bytes;
+    int channels = stbir_info->channels;
+    int alpha_channel = stbir_info->alpha_channel;
+    int type = stbir_info->type;
+    int colorspace = stbir_info->colorspace;
+    int output_w = stbir_info->output_w;
+    void* output_data = stbir_info->output_data;
+    int decode = STBIR__DECODE(type, colorspace);
+
+    float* ring_buffer = stbir_info->ring_buffer;
+    int ring_buffer_length = stbir_info->ring_buffer_length_bytes/sizeof(float);
+
+    if (stbir_info->ring_buffer_begin_index >= 0)
+    {
+        // Get rid of whatever we don't need anymore.
+        while (first_necessary_scanline > stbir_info->ring_buffer_first_scanline)
+        {
+            if (stbir_info->ring_buffer_first_scanline >= 0 && stbir_info->ring_buffer_first_scanline < stbir_info->output_h)
+            {
+                int output_row_start = stbir_info->ring_buffer_first_scanline * output_stride_bytes;
+                float* ring_buffer_entry = stbir__get_ring_buffer_entry(ring_buffer, stbir_info->ring_buffer_begin_index, ring_buffer_length);
+                stbir__encode_scanline(stbir_info, output_w, (char *) output_data + output_row_start, ring_buffer_entry, channels, alpha_channel, decode);
+                STBIR_PROGRESS_REPORT((float)stbir_info->ring_buffer_first_scanline / stbir_info->output_h);
+            }
+
+            if (stbir_info->ring_buffer_first_scanline == stbir_info->ring_buffer_last_scanline)
+            {
+                // We just popped the last scanline off the ring buffer.
+                // Reset it to the empty state.
+                stbir_info->ring_buffer_begin_index = -1;
+                stbir_info->ring_buffer_first_scanline = 0;
+                stbir_info->ring_buffer_last_scanline = 0;
+                break;
+            }
+            else
+            {
+                stbir_info->ring_buffer_first_scanline++;
+                stbir_info->ring_buffer_begin_index = (stbir_info->ring_buffer_begin_index + 1) % stbir_info->vertical_filter_pixel_width;
+            }
+        }
+    }
+}
+
+static void stbir__buffer_loop_downsample(stbir__info* stbir_info)
+{
+    int y;
+    float scale_ratio = stbir_info->vertical_scale;
+    int output_h = stbir_info->output_h;
+    float in_pixels_radius = stbir__filter_info_table[stbir_info->vertical_filter].support(scale_ratio) / scale_ratio;
+    int pixel_margin = stbir_info->vertical_filter_pixel_margin;
+    int max_y = stbir_info->input_h + pixel_margin;
+
+    STBIR__DEBUG_ASSERT(!stbir__use_height_upsampling(stbir_info));
+
+    for (y = -pixel_margin; y < max_y; y++)
+    {
+        float out_center_of_in; // Center of the current out scanline in the in scanline space
+        int out_first_scanline, out_last_scanline;
+
+        stbir__calculate_sample_range_downsample(y, in_pixels_radius, scale_ratio, stbir_info->vertical_shift, &out_first_scanline, &out_last_scanline, &out_center_of_in);
+
+        STBIR__DEBUG_ASSERT(out_last_scanline - out_first_scanline <= stbir_info->vertical_filter_pixel_width);
+
+        if (out_last_scanline < 0 || out_first_scanline >= output_h)
+            continue;
+
+        stbir__empty_ring_buffer(stbir_info, out_first_scanline);
+
+        stbir__decode_and_resample_downsample(stbir_info, y);
+
+        // Load in new ones.
+        if (stbir_info->ring_buffer_begin_index < 0)
+            stbir__add_empty_ring_buffer_entry(stbir_info, out_first_scanline);
+
+        while (out_last_scanline > stbir_info->ring_buffer_last_scanline)
+            stbir__add_empty_ring_buffer_entry(stbir_info, stbir_info->ring_buffer_last_scanline + 1);
+
+        // Now the horizontal buffer is ready to write to all ring buffer rows.
+        stbir__resample_vertical_downsample(stbir_info, y, out_first_scanline, out_last_scanline, out_center_of_in);
+    }
+
+    stbir__empty_ring_buffer(stbir_info, stbir_info->output_h);
+}
+
+static void stbir__setup(stbir__info *info, int input_w, int input_h, int output_w, int output_h, int channels)
+{
+    info->input_w = input_w;
+    info->input_h = input_h;
+    info->output_w = output_w;
+    info->output_h = output_h;
+    info->channels = channels;
+}
+
+static void stbir__calculate_transform(stbir__info *info, float s0, float t0, float s1, float t1, float *transform)
+{
+    info->s0 = s0;
+    info->t0 = t0;
+    info->s1 = s1;
+    info->t1 = t1;
+
+    if (transform)
+    {
+        info->horizontal_scale = transform[0];
+        info->vertical_scale   = transform[1];
+        info->horizontal_shift = transform[2];
+        info->vertical_shift   = transform[3];
+    }
+    else
+    {
+        info->horizontal_scale = ((float)info->output_w / info->input_w) / (s1 - s0);
+        info->vertical_scale = ((float)info->output_h / info->input_h) / (t1 - t0);
+
+        info->horizontal_shift = s0 * info->input_w / (s1 - s0);
+        info->vertical_shift = t0 * info->input_h / (t1 - t0);
+    }
+}
+
+static void stbir__choose_filter(stbir__info *info, stbir_filter h_filter, stbir_filter v_filter)
+{
+    if (h_filter == 0)
+        h_filter = stbir__use_upsampling(info->horizontal_scale) ? STBIR_DEFAULT_FILTER_UPSAMPLE : STBIR_DEFAULT_FILTER_DOWNSAMPLE;
+    if (v_filter == 0)
+        v_filter = stbir__use_upsampling(info->vertical_scale)   ? STBIR_DEFAULT_FILTER_UPSAMPLE : STBIR_DEFAULT_FILTER_DOWNSAMPLE;
+    info->horizontal_filter = h_filter;
+    info->vertical_filter = v_filter;
+}
+
+static stbir_uint32 stbir__calculate_memory(stbir__info *info)
+{
+    int pixel_margin = stbir__get_filter_pixel_margin(info->horizontal_filter, info->horizontal_scale);
+    int filter_height = stbir__get_filter_pixel_width(info->vertical_filter, info->vertical_scale);
+
+    info->horizontal_num_contributors = stbir__get_contributors(info->horizontal_scale, info->horizontal_filter, info->input_w, info->output_w);
+    info->vertical_num_contributors   = stbir__get_contributors(info->vertical_scale  , info->vertical_filter  , info->input_h, info->output_h);
+
+    info->horizontal_contributors_size = info->horizontal_num_contributors * sizeof(stbir__contributors);
+    info->horizontal_coefficients_size = stbir__get_total_horizontal_coefficients(info) * sizeof(float);
+    info->vertical_contributors_size = info->vertical_num_contributors * sizeof(stbir__contributors);
+    info->vertical_coefficients_size = stbir__get_total_vertical_coefficients(info) * sizeof(float);
+    info->decode_buffer_size = (info->input_w + pixel_margin * 2) * info->channels * sizeof(float);
+    info->horizontal_buffer_size = info->output_w * info->channels * sizeof(float);
+    info->ring_buffer_size = info->output_w * info->channels * filter_height * sizeof(float);
+    info->encode_buffer_size = info->output_w * info->channels * sizeof(float);
+
+    STBIR_ASSERT(info->horizontal_filter != 0);
+    STBIR_ASSERT(info->horizontal_filter < STBIR__ARRAY_SIZE(stbir__filter_info_table)); // this now happens too late
+    STBIR_ASSERT(info->vertical_filter != 0);
+    STBIR_ASSERT(info->vertical_filter < STBIR__ARRAY_SIZE(stbir__filter_info_table)); // this now happens too late
+
+    if (stbir__use_height_upsampling(info))
+        // The horizontal buffer is for when we're downsampling the height and we
+        // can't output the result of sampling the decode buffer directly into the
+        // ring buffers.
+        info->horizontal_buffer_size = 0;
+    else
+        // The encode buffer is to retain precision in the height upsampling method
+        // and isn't used when height downsampling.
+        info->encode_buffer_size = 0;
+
+    return info->horizontal_contributors_size + info->horizontal_coefficients_size
+        + info->vertical_contributors_size + info->vertical_coefficients_size
+        + info->decode_buffer_size + info->horizontal_buffer_size
+        + info->ring_buffer_size + info->encode_buffer_size;
+}
+
+static int stbir__resize_allocated(stbir__info *info,
+    const void* input_data, int input_stride_in_bytes,
+    void* output_data, int output_stride_in_bytes,
+    int alpha_channel, stbir_uint32 flags, stbir_datatype type,
+    stbir_edge edge_horizontal, stbir_edge edge_vertical, stbir_colorspace colorspace,
+    void* tempmem, size_t tempmem_size_in_bytes)
+{
+    size_t memory_required = stbir__calculate_memory(info);
+
+    int width_stride_input = input_stride_in_bytes ? input_stride_in_bytes : info->channels * info->input_w * stbir__type_size[type];
+    int width_stride_output = output_stride_in_bytes ? output_stride_in_bytes : info->channels * info->output_w * stbir__type_size[type];
+
+#ifdef STBIR_DEBUG_OVERWRITE_TEST
+#define OVERWRITE_ARRAY_SIZE 8
+    unsigned char overwrite_output_before_pre[OVERWRITE_ARRAY_SIZE];
+    unsigned char overwrite_tempmem_before_pre[OVERWRITE_ARRAY_SIZE];
+    unsigned char overwrite_output_after_pre[OVERWRITE_ARRAY_SIZE];
+    unsigned char overwrite_tempmem_after_pre[OVERWRITE_ARRAY_SIZE];
+
+    size_t begin_forbidden = width_stride_output * (info->output_h - 1) + info->output_w * info->channels * stbir__type_size[type];
+    memcpy(overwrite_output_before_pre, &((unsigned char*)output_data)[-OVERWRITE_ARRAY_SIZE], OVERWRITE_ARRAY_SIZE);
+    memcpy(overwrite_output_after_pre, &((unsigned char*)output_data)[begin_forbidden], OVERWRITE_ARRAY_SIZE);
+    memcpy(overwrite_tempmem_before_pre, &((unsigned char*)tempmem)[-OVERWRITE_ARRAY_SIZE], OVERWRITE_ARRAY_SIZE);
+    memcpy(overwrite_tempmem_after_pre, &((unsigned char*)tempmem)[tempmem_size_in_bytes], OVERWRITE_ARRAY_SIZE);
+#endif
+
+    STBIR_ASSERT(info->channels >= 0);
+    STBIR_ASSERT(info->channels <= STBIR_MAX_CHANNELS);
+
+    if (info->channels < 0 || info->channels > STBIR_MAX_CHANNELS)
+        return 0;
+
+    STBIR_ASSERT(info->horizontal_filter < STBIR__ARRAY_SIZE(stbir__filter_info_table));
+    STBIR_ASSERT(info->vertical_filter < STBIR__ARRAY_SIZE(stbir__filter_info_table));
+
+    if (info->horizontal_filter >= STBIR__ARRAY_SIZE(stbir__filter_info_table))
+        return 0;
+    if (info->vertical_filter >= STBIR__ARRAY_SIZE(stbir__filter_info_table))
+        return 0;
+
+    if (alpha_channel < 0)
+        flags |= STBIR_FLAG_ALPHA_USES_COLORSPACE | STBIR_FLAG_ALPHA_PREMULTIPLIED;
+
+    if (!(flags&STBIR_FLAG_ALPHA_USES_COLORSPACE) || !(flags&STBIR_FLAG_ALPHA_PREMULTIPLIED))
+        STBIR_ASSERT(alpha_channel >= 0 && alpha_channel < info->channels);
+
+    if (alpha_channel >= info->channels)
+        return 0;
+
+    STBIR_ASSERT(tempmem);
+
+    if (!tempmem)
+        return 0;
+
+    STBIR_ASSERT(tempmem_size_in_bytes >= memory_required);
+
+    if (tempmem_size_in_bytes < memory_required)
+        return 0;
+
+    memset(tempmem, 0, tempmem_size_in_bytes);
+
+    info->input_data = input_data;
+    info->input_stride_bytes = width_stride_input;
+
+    info->output_data = output_data;
+    info->output_stride_bytes = width_stride_output;
+
+    info->alpha_channel = alpha_channel;
+    info->flags = flags;
+    info->type = type;
+    info->edge_horizontal = edge_horizontal;
+    info->edge_vertical = edge_vertical;
+    info->colorspace = colorspace;
+
+    info->horizontal_coefficient_width   = stbir__get_coefficient_width  (info->horizontal_filter, info->horizontal_scale);
+    info->vertical_coefficient_width     = stbir__get_coefficient_width  (info->vertical_filter  , info->vertical_scale  );
+    info->horizontal_filter_pixel_width  = stbir__get_filter_pixel_width (info->horizontal_filter, info->horizontal_scale);
+    info->vertical_filter_pixel_width    = stbir__get_filter_pixel_width (info->vertical_filter  , info->vertical_scale  );
+    info->horizontal_filter_pixel_margin = stbir__get_filter_pixel_margin(info->horizontal_filter, info->horizontal_scale);
+    info->vertical_filter_pixel_margin   = stbir__get_filter_pixel_margin(info->vertical_filter  , info->vertical_scale  );
+
+    info->ring_buffer_length_bytes = info->output_w * info->channels * sizeof(float);
+    info->decode_buffer_pixels = info->input_w + info->horizontal_filter_pixel_margin * 2;
+
+#define STBIR__NEXT_MEMPTR(current, newtype) (newtype*)(((unsigned char*)current) + current##_size)
+
+    info->horizontal_contributors = (stbir__contributors *) tempmem;
+    info->horizontal_coefficients = STBIR__NEXT_MEMPTR(info->horizontal_contributors, float);
+    info->vertical_contributors = STBIR__NEXT_MEMPTR(info->horizontal_coefficients, stbir__contributors);
+    info->vertical_coefficients = STBIR__NEXT_MEMPTR(info->vertical_contributors, float);
+    info->decode_buffer = STBIR__NEXT_MEMPTR(info->vertical_coefficients, float);
+
+    if (stbir__use_height_upsampling(info))
+    {
+        info->horizontal_buffer = NULL;
+        info->ring_buffer = STBIR__NEXT_MEMPTR(info->decode_buffer, float);
+        info->encode_buffer = STBIR__NEXT_MEMPTR(info->ring_buffer, float);
+
+        STBIR__DEBUG_ASSERT((size_t)STBIR__NEXT_MEMPTR(info->encode_buffer, unsigned char) == (size_t)tempmem + tempmem_size_in_bytes);
+    }
+    else
+    {
+        info->horizontal_buffer = STBIR__NEXT_MEMPTR(info->decode_buffer, float);
+        info->ring_buffer = STBIR__NEXT_MEMPTR(info->horizontal_buffer, float);
+        info->encode_buffer = NULL;
+
+        STBIR__DEBUG_ASSERT((size_t)STBIR__NEXT_MEMPTR(info->ring_buffer, unsigned char) == (size_t)tempmem + tempmem_size_in_bytes);
+    }
+
+#undef STBIR__NEXT_MEMPTR
+
+    // This signals that the ring buffer is empty
+    info->ring_buffer_begin_index = -1;
+
+    stbir__calculate_filters(info, info->horizontal_contributors, info->horizontal_coefficients, info->horizontal_filter, info->horizontal_scale, info->horizontal_shift, info->input_w, info->output_w);
+    stbir__calculate_filters(info, info->vertical_contributors, info->vertical_coefficients, info->vertical_filter, info->vertical_scale, info->vertical_shift, info->input_h, info->output_h);
+
+    STBIR_PROGRESS_REPORT(0);
+
+    if (stbir__use_height_upsampling(info))
+        stbir__buffer_loop_upsample(info);
+    else
+        stbir__buffer_loop_downsample(info);
+
+    STBIR_PROGRESS_REPORT(1);
+
+#ifdef STBIR_DEBUG_OVERWRITE_TEST
+    STBIR__DEBUG_ASSERT(memcmp(overwrite_output_before_pre, &((unsigned char*)output_data)[-OVERWRITE_ARRAY_SIZE], OVERWRITE_ARRAY_SIZE) == 0);
+    STBIR__DEBUG_ASSERT(memcmp(overwrite_output_after_pre, &((unsigned char*)output_data)[begin_forbidden], OVERWRITE_ARRAY_SIZE) == 0);
+    STBIR__DEBUG_ASSERT(memcmp(overwrite_tempmem_before_pre, &((unsigned char*)tempmem)[-OVERWRITE_ARRAY_SIZE], OVERWRITE_ARRAY_SIZE) == 0);
+    STBIR__DEBUG_ASSERT(memcmp(overwrite_tempmem_after_pre, &((unsigned char*)tempmem)[tempmem_size_in_bytes], OVERWRITE_ARRAY_SIZE) == 0);
+#endif
+
+    return 1;
+}
+
+
+static int stbir__resize_arbitrary(
+    void *alloc_context,
+    const void* input_data, int input_w, int input_h, int input_stride_in_bytes,
+    void* output_data, int output_w, int output_h, int output_stride_in_bytes,
+    float s0, float t0, float s1, float t1, float *transform,
+    int channels, int alpha_channel, stbir_uint32 flags, stbir_datatype type,
+    stbir_filter h_filter, stbir_filter v_filter,
+    stbir_edge edge_horizontal, stbir_edge edge_vertical, stbir_colorspace colorspace)
+{
+    stbir__info info;
+    int result;
+    size_t memory_required;
+    void* extra_memory;
+
+    stbir__setup(&info, input_w, input_h, output_w, output_h, channels);
+    stbir__calculate_transform(&info, s0,t0,s1,t1,transform);
+    stbir__choose_filter(&info, h_filter, v_filter);
+    memory_required = stbir__calculate_memory(&info);
+    extra_memory = STBIR_MALLOC(memory_required, alloc_context);
+
+    if (!extra_memory)
+        return 0;
+
+    result = stbir__resize_allocated(&info, input_data, input_stride_in_bytes,
+                                            output_data, output_stride_in_bytes, 
+                                            alpha_channel, flags, type,
+                                            edge_horizontal, edge_vertical,
+                                            colorspace, extra_memory, memory_required);
+
+    STBIR_FREE(extra_memory, alloc_context);
+
+    return result;
+}
+
+STBIRDEF int stbir_resize_uint8(     const unsigned char *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                           unsigned char *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                     int num_channels)
+{
+    return stbir__resize_arbitrary(NULL, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,NULL,num_channels,-1,0, STBIR_TYPE_UINT8, STBIR_FILTER_DEFAULT, STBIR_FILTER_DEFAULT,
+        STBIR_EDGE_CLAMP, STBIR_EDGE_CLAMP, STBIR_COLORSPACE_LINEAR);
+}
+
+STBIRDEF int stbir_resize_float(     const float *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                           float *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                     int num_channels)
+{
+    return stbir__resize_arbitrary(NULL, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,NULL,num_channels,-1,0, STBIR_TYPE_FLOAT, STBIR_FILTER_DEFAULT, STBIR_FILTER_DEFAULT,
+        STBIR_EDGE_CLAMP, STBIR_EDGE_CLAMP, STBIR_COLORSPACE_LINEAR);
+}
+
+STBIRDEF int stbir_resize_uint8_srgb(const unsigned char *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                           unsigned char *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                     int num_channels, int alpha_channel, int flags)
+{
+    return stbir__resize_arbitrary(NULL, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,NULL,num_channels,alpha_channel,flags, STBIR_TYPE_UINT8, STBIR_FILTER_DEFAULT, STBIR_FILTER_DEFAULT,
+        STBIR_EDGE_CLAMP, STBIR_EDGE_CLAMP, STBIR_COLORSPACE_SRGB);
+}
+
+STBIRDEF int stbir_resize_uint8_srgb_edgemode(const unsigned char *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                                    unsigned char *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                              int num_channels, int alpha_channel, int flags,
+                                              stbir_edge edge_wrap_mode)
+{
+    return stbir__resize_arbitrary(NULL, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,NULL,num_channels,alpha_channel,flags, STBIR_TYPE_UINT8, STBIR_FILTER_DEFAULT, STBIR_FILTER_DEFAULT,
+        edge_wrap_mode, edge_wrap_mode, STBIR_COLORSPACE_SRGB);
+}
+
+STBIRDEF int stbir_resize_uint8_generic( const unsigned char *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                               unsigned char *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                         int num_channels, int alpha_channel, int flags,
+                                         stbir_edge edge_wrap_mode, stbir_filter filter, stbir_colorspace space, 
+                                         void *alloc_context)
+{
+    return stbir__resize_arbitrary(alloc_context, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,NULL,num_channels,alpha_channel,flags, STBIR_TYPE_UINT8, filter, filter,
+        edge_wrap_mode, edge_wrap_mode, space);
+}
+
+STBIRDEF int stbir_resize_uint16_generic(const stbir_uint16 *input_pixels  , int input_w , int input_h , int input_stride_in_bytes,
+                                               stbir_uint16 *output_pixels , int output_w, int output_h, int output_stride_in_bytes,
+                                         int num_channels, int alpha_channel, int flags,
+                                         stbir_edge edge_wrap_mode, stbir_filter filter, stbir_colorspace space, 
+                                         void *alloc_context)
+{
+    return stbir__resize_arbitrary(alloc_context, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,NULL,num_channels,alpha_channel,flags, STBIR_TYPE_UINT16, filter, filter,
+        edge_wrap_mode, edge_wrap_mode, space);
+}
+
+
+STBIRDEF int stbir_resize_float_generic( const float *input_pixels         , int input_w , int input_h , int input_stride_in_bytes,
+                                               float *output_pixels        , int output_w, int output_h, int output_stride_in_bytes,
+                                         int num_channels, int alpha_channel, int flags,
+                                         stbir_edge edge_wrap_mode, stbir_filter filter, stbir_colorspace space, 
+                                         void *alloc_context)
+{
+    return stbir__resize_arbitrary(alloc_context, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,NULL,num_channels,alpha_channel,flags, STBIR_TYPE_FLOAT, filter, filter,
+        edge_wrap_mode, edge_wrap_mode, space);
+}
+
+
+STBIRDEF int stbir_resize(         const void *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                         void *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                   stbir_datatype datatype,
+                                   int num_channels, int alpha_channel, int flags,
+                                   stbir_edge edge_mode_horizontal, stbir_edge edge_mode_vertical, 
+                                   stbir_filter filter_horizontal,  stbir_filter filter_vertical,
+                                   stbir_colorspace space, void *alloc_context)
+{
+    return stbir__resize_arbitrary(alloc_context, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,NULL,num_channels,alpha_channel,flags, datatype, filter_horizontal, filter_vertical,
+        edge_mode_horizontal, edge_mode_vertical, space);
+}
+
+
+STBIRDEF int stbir_resize_subpixel(const void *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                         void *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                   stbir_datatype datatype,
+                                   int num_channels, int alpha_channel, int flags,
+                                   stbir_edge edge_mode_horizontal, stbir_edge edge_mode_vertical, 
+                                   stbir_filter filter_horizontal,  stbir_filter filter_vertical,
+                                   stbir_colorspace space, void *alloc_context,
+                                   float x_scale, float y_scale,
+                                   float x_offset, float y_offset)
+{
+    float transform[4];
+    transform[0] = x_scale;
+    transform[1] = y_scale;
+    transform[2] = x_offset;
+    transform[3] = y_offset;
+    return stbir__resize_arbitrary(alloc_context, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        0,0,1,1,transform,num_channels,alpha_channel,flags, datatype, filter_horizontal, filter_vertical,
+        edge_mode_horizontal, edge_mode_vertical, space);
+}
+
+STBIRDEF int stbir_resize_region(  const void *input_pixels , int input_w , int input_h , int input_stride_in_bytes,
+                                         void *output_pixels, int output_w, int output_h, int output_stride_in_bytes,
+                                   stbir_datatype datatype,
+                                   int num_channels, int alpha_channel, int flags,
+                                   stbir_edge edge_mode_horizontal, stbir_edge edge_mode_vertical, 
+                                   stbir_filter filter_horizontal,  stbir_filter filter_vertical,
+                                   stbir_colorspace space, void *alloc_context,
+                                   float s0, float t0, float s1, float t1)
+{
+    return stbir__resize_arbitrary(alloc_context, input_pixels, input_w, input_h, input_stride_in_bytes,
+        output_pixels, output_w, output_h, output_stride_in_bytes,
+        s0,t0,s1,t1,NULL,num_channels,alpha_channel,flags, datatype, filter_horizontal, filter_vertical,
+        edge_mode_horizontal, edge_mode_vertical, space);
+}
+
+#endif // STB_IMAGE_RESIZE_IMPLEMENTATION

From e10b57d27d80aa0c7ff37e45adc5649f4d107316 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Thu, 6 Mar 2025 18:31:38 +0000
Subject: [PATCH 05/33] auto cut

---
 .gitignore               |   3 +
 .gitmodules              |   4 +-
 CMakeLists.txt           |   1 -
 PretextView.cpp          | 132 +++++++++-
 src/ai_cut_json_parse.h  | 116 +++++++++
 src/copy_texture.cpp     |  44 ++--
 src/copy_texture.h       |  10 +-
 src/frag_sort.cpp        |  14 +-
 src/frag_sort.h          |   6 +-
 src/hic_figure.cpp       |   1 +
 src/hic_figure.h         | 516 +++++++++++++++++++++++++++++++++++++++
 src/utilsPretextView.cpp |  13 +-
 12 files changed, 810 insertions(+), 50 deletions(-)
 create mode 100644 src/ai_cut_json_parse.h
 create mode 100644 src/hic_figure.cpp
 create mode 100644 src/hic_figure.h

diff --git a/.gitignore b/.gitignore
index 608b4a4..9a429af 100644
--- a/.gitignore
+++ b/.gitignore
@@ -46,3 +46,6 @@ PretextViewAI.zip
 
 test_data/
 subprojects/libtorch(not needed in this version)/
+
+hic_figures/
+auto_cut_output/
\ No newline at end of file
diff --git a/.gitmodules b/.gitmodules
index a18765f..1411617 100644
--- a/.gitmodules
+++ b/.gitmodules
@@ -14,6 +14,4 @@
 [submodule "subprojects/pytorch_scatter"]
 	path = subprojects/pytorch_scatter
 	url = https://github.com/guanshaoheng/pytorch_scatter.git 
-[submodule "subprojects/opencv"]
-	path = subprojects/opencv
-	url = https://github.com/opencv/opencv.git
+
diff --git a/CMakeLists.txt b/CMakeLists.txt
index f898a7c..3c27166 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -81,7 +81,6 @@ include_directories(
 add_subdirectory(subprojects/glfw)
 add_subdirectory(subprojects/glm)
 
-
 find_library(
     lib_deflate_static 
     NAMES deflate deflatestatic
diff --git a/PretextView.cpp b/PretextView.cpp
index 440a1a5..4e151b6 100644
--- a/PretextView.cpp
+++ b/PretextView.cpp
@@ -103,7 +103,8 @@ SOFTWARE.
 
 #include "genomeData.h"
 #include "showWindowData.h"
-
+#include "hic_figure.h"
+#include "ai_cut_json_parse.h"
 
 #include "shaderSource.h"
 /*
@@ -659,7 +660,11 @@ global_variable
 u32
 Yahs_sorting = 0;
 
-global_variable s32 ai_sort_button = 0; 
+global_variable
+u32 
+AI_cutting = 0;
+
+global_variable s32 ai_cutting_button = 0; 
 global_variable s32 yahs_sort_button = 0;
 
 
@@ -4335,6 +4340,24 @@ Render() {
 
             ChangeSize((s32)width, (s32)height);
         }
+
+        if (AI_cutting)
+        {
+            u32 colour = glfonsRGBA(Theme_Colour.r, Theme_Colour.g, Theme_Colour.b, Theme_Colour.a);
+
+            fonsClearState(FontStash_Context);
+            fonsSetSize(FontStash_Context, 64.0f * Screen_Scale.x);
+            fonsSetAlign(FontStash_Context, FONS_ALIGN_CENTER | FONS_ALIGN_MIDDLE);
+            fonsSetFont(FontStash_Context, Font_Bold);
+            fonsSetColor(FontStash_Context, colour);
+
+            glUseProgram(UI_Shader->shaderProgram);
+            glUniformMatrix4fv(Flat_Shader->matLocation, 1, GL_FALSE, textNormalMat);
+            glViewport(0, 0, (s32)width, (s32)height);
+            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "AI cutting...", 0);
+
+            ChangeSize((s32)width, (s32)height);
+        }
     }
 }
 
@@ -6213,6 +6236,28 @@ void EraseAllEdits(map_editor* map_editor_, u32 max_edit_recorded=Edits_Stack_Si
 }
 
 
+void run_ai_detection()
+{
+    std::cout << "Running AI detection..." << std::endl;
+    int status = std::system("/Users/sg35/miniconda3/envs/auto_cut/bin/python python/autoCut/auto_cut.py");  
+
+    // 检查命令执行状态
+    if (status == 0) {
+        std::cout << "命令执行成功!" << std::endl;
+    } else {
+        std::cerr << "命令执行失败，错误码: " << status << std::endl;
+    }
+}
+
+void cut_frags(std::vector<std::vector<u32>> & problem_locs)
+{
+    for (auto & i : problem_locs)
+    {
+        InvertMap(i[0], i[1]);
+        AddMapEdit(0, {i[0], i[1]}, true);
+    }
+}
+
 
 void AutoCurationFromFragsOrder(
     const FragsOrder* frags_order_,
@@ -6232,7 +6277,7 @@ void AutoCurationFromFragsOrder(
 
     // check the difference between the contigs, order and the new frags order
     u32 start_loc = 0;
-    std::vector<std::pair<s32, s32>> current_order(num_frags);
+    std::vector<std::pair<s32, s32>> current_order(num_frags); // [contig_id, contig_length]
     std::vector<s32> predicted_order = frags_order_->get_order_without_chromosomeInfor(); // start from 1
     if (using_select_area)
     {
@@ -6295,15 +6340,19 @@ void AutoCurationFromFragsOrder(
             while (std::abs(predicted_order[i])!=std::abs(current_order[tmp_i].first))
             {
                 if (tmp_i >= num_frags)
-                {
-                    fprintf(stderr, "Error: contig %d not found in the current order.\n", current_order[i].first);
-                    MY_CHECK(0);
+                {   
+                    char buff[256];
+                    snprintf(buff, sizeof(buff), "Error: contig %d not found in the current order.\n", current_order[i].first);
+                    MY_CHECK(buff);
+                    assert(0);
                 }
                 pixelFrom += current_order[tmp_i].second; // length updated based on the current order
                 if (pixelFrom>=Number_of_Pixels_1D)
-                {
-                    fprintf(stderr, "Error: pixelFrom(%d) >= Number_of_Pixels_1D(%d).\n", pixelFrom, Number_of_Pixels_1D);
-                    MY_CHECK(0);
+                {   
+                    char buff[256];
+                    snprintf(buff, sizeof(buff), "Error: pixelFrom(%d) >= Number_of_Pixels_1D(%d).\n", pixelFrom, Number_of_Pixels_1D);
+                    MY_CHECK(buff);
+                    assert(0);
                 }
                 ++tmp_i;
             }
@@ -6338,6 +6387,50 @@ void AutoCurationFromFragsOrder(
     return ;
 }
 
+global_function
+void
+ai_cutting_func(
+    char * currFileName,
+    std::string file_save_path = std::string("./hic_figures")
+)
+{   
+    if (!currFileName || !AI_cutting) return;
+
+    fprintf(stdout, "========================\n");
+    // copy textures to cpu
+    fprintf(stdout, "[AI cutting] coping textures...\n");
+    auto texture_array_4_ai = TexturesArray4AI(Number_of_Textures_1D, Texture_Resolution, (char*)currFileName, Contigs);
+    // prepare before reading textures
+    f32 original_control_points[3];
+    prepare_before_copy(original_control_points);
+    texture_array_4_ai.copy_buffer_to_textures_dynamic(Contact_Matrix, false); 
+    restore_settings_after_copy(original_control_points);
+    
+    // generate figures
+    if (1)
+    {
+        fprintf(stdout, "[AI cutting] generating hic figures...\n");
+        Hic_Figure hic_figure(
+            Total_Genome_Length, 
+            file_save_path,
+            &texture_array_4_ai
+        );
+        hic_figure.generate_hic_figures();
+    }
+
+    // run ai model with python script
+    run_ai_detection();
+
+    // restore the error info via reading the error json
+    HIC_Problems hic_problems("auto_cut_output/infer_result/infer_result.json", (f32) Total_Genome_Length / (f32)Number_of_Pixels_1D ); 
+    std::vector<std::vector<u32>> problem_loc = hic_problems.get_problem_loci();
+
+    // cut the contigs 
+    cut_frags(problem_loc);
+
+    return ;
+}
+
 
 global_function
 void
@@ -6858,7 +6951,7 @@ global_function
 void
 KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
 {
-    if (!Loading && !Yahs_sorting && (action != GLFW_RELEASE || key == GLFW_KEY_SPACE || key == GLFW_KEY_A || key == GLFW_KEY_LEFT_SHIFT))
+    if (!Loading && !Yahs_sorting && !AI_cutting && (action != GLFW_RELEASE || key == GLFW_KEY_SPACE || key == GLFW_KEY_A || key == GLFW_KEY_LEFT_SHIFT))
     {
         if (UI_On)
         {
@@ -10464,6 +10557,14 @@ MainArgs {
             Redisplay = 1;
         }
 
+        if (AI_cutting)
+        {   
+            if (currFileName) SaveState(headerHash);
+            ai_cutting_func((char*)currFileName);
+            AI_cutting = 0;
+            Redisplay = 1;
+        }
+
         if (UI_On) 
         {
             glfwSetCursor(window, arrowCursor);
@@ -10572,9 +10673,10 @@ MainArgs {
                     struct nk_color sort_button_active = nk_rgb(128, 0, 128);  
                     push_nk_style(NK_Context, sort_button_normal, sort_button_hover, sort_button_active);
                     nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 3);
-                    // ai_sort_button = nk_button_label(NK_Context, "AI Sort");
+                    ai_cutting_button = nk_button_label(NK_Context, "AI cut");
                     bounds = nk_widget_bounds(NK_Context);
                     yahs_sort_button = nk_button_label(NK_Context, "YaHS Sort");
+
                     // AI sort button
                     /*
                     {
@@ -10617,6 +10719,14 @@ MainArgs {
                         }
                     }
                     */
+                    // ai cut
+                    {
+                        if (ai_cutting_button && currFileName)
+                        {
+                            AI_cutting = 1;
+                        }
+                    }
+
                     // YaHS sort button
                     {   
                         if (yahs_sort_button) 
diff --git a/src/ai_cut_json_parse.h b/src/ai_cut_json_parse.h
new file mode 100644
index 0000000..3f2e890
--- /dev/null
+++ b/src/ai_cut_json_parse.h
@@ -0,0 +1,116 @@
+
+
+#ifndef auto_cut_h
+#define auto_cut_h
+
+#include <iostream>
+#include <fstream>
+#include <string>
+#include <nlohmann/json.hpp>
+
+
+#include "utilsPretextView.h"
+
+
+using json = nlohmann::json;
+
+
+
+struct Problem
+{   
+    int id;
+    std::string image_id;
+    std::string category;
+    std::vector<f32> bbox;       // [x1, y1, x2, y2]
+    f32 score;
+    f32 resolution;
+    std::vector<u64> bp_loci;     // [start1, end1, start2, end2]
+    std::vector<u32> hic_loci_pixel; 
+    std::string infer_image;
+};
+
+
+void from_json(const json& j, Problem& t) 
+{
+    j.at("id").get_to(t.id);
+    j.at("image_id").get_to(t.image_id);
+    j.at("category").get_to(t.category);
+    j.at("bbox").get_to(t.bbox);
+    j.at("score").get_to(t.score);
+    j.at("resolution").get_to(t.resolution);
+    j.at("hic_loci").get_to(t.bp_loci);
+    j.at("infer_image").get_to(t.infer_image);
+    t.hic_loci_pixel = std::vector<u32>(4, 0);
+}
+
+
+
+class HIC_Problems
+{   
+public:
+    f32 bp_per_pixel;
+    std::vector<std::string> types={"translocation", "inversion", "debris"};
+    std::map<std::string, std::vector<Problem>> problems;
+
+    HIC_Problems(std::string file_path, f32 bp_per_pixel_)
+    : bp_per_pixel(bp_per_pixel_)
+    {
+        std::ifstream file(file_path, std::ios::in);
+        if (!file.is_open())
+        {   
+            throw std::runtime_error("Cannot open the file: " + file_path);
+        }
+        
+        // 解析 JSON
+        json data = json::parse(file);
+
+        for (const auto& i : this->types)
+        {
+            if (data.contains(i) && data[i].is_array()) 
+            {
+                for (const auto& item : data[i]) 
+                {   
+                    auto tmp = item.get<Problem>();
+                    if (tmp.bp_loci.size() != 4)
+                    {
+                        throw std::runtime_error("The hic_loci should contain 4 elements, tmp_id:" + tmp.image_id);
+                    }
+                    for (int j = 0; j < 4; j++)
+                    {
+                        tmp.hic_loci_pixel[j] = (u32) ((f32)tmp.bp_loci[j] / bp_per_pixel);
+                    }
+                    this->problems[i].push_back(tmp);
+                }
+            } else 
+            {   
+                #ifdef DEBUG
+                    std::cerr << "Warning: No valid " << i << " data found" << std::endl;
+                #endif // DEBUG
+            }
+        }
+    }
+
+    std::vector<std::vector<u32>> get_problem_loci()
+    {
+        std::vector<std::vector<u32>> output;
+
+        for (const auto& type : this->types)
+        {
+            for (auto item : this->problems[type])
+            {   
+                output.push_back({
+                    item.hic_loci_pixel[0],
+                    item.hic_loci_pixel[1]
+                });
+                std::cout << type << " loci: " << item.hic_loci_pixel[0] << " " << item.hic_loci_pixel[1] << std::endl;
+            }
+        }
+        return output;
+    } 
+
+
+};
+
+
+
+#endif // auto_cut_h
\ No newline at end of file
diff --git a/src/copy_texture.cpp b/src/copy_texture.cpp
index cb1f214..d67e755 100644
--- a/src/copy_texture.cpp
+++ b/src/copy_texture.cpp
@@ -126,19 +126,12 @@ TexturesArray4AI::TexturesArray4AI(
     is_copied_from_buffer(false), is_compressed(false), hic_shader_initilised(false)
 {   
     
-    MY_CHECK(0);
-
     this->frags = new Frag4compress(Contigs);
 
-    MY_CHECK(0);
     this->compressed_hic = new Matrix3D<f32>(1, 1, 5);  // initialize the compressed_hic with 1x1x5
 
-    MY_CHECK(0);
     this->compressed_extensions = new CompressedExtensions(1); // initialize the compressed_extensions with 1
 
-
-    MY_CHECK(0);
-
     char* tmp = fileName;
     while (*tmp != '\0' && *tmp != '.') 
     {
@@ -177,7 +170,7 @@ TexturesArray4AI::TexturesArray4AI(
         std::cerr << e.what() << " Allocate mem error\n";
         assert(0);
     }
-    MY_CHECK(0);
+    
     // Shader setup and texture binding
     this->shaderProgram = CreateShader(FragmentSource_copyTexture.c_str(), VertexSource_copyTexture.c_str());
 
@@ -265,7 +258,8 @@ TexturesArray4AI::~TexturesArray4AI()
     }
 }
 
-
+/* 
+*/
 u08 TexturesArray4AI::operator()(u32 row, u32 column) const
 {   
     if (row>=num_pixels_1d || column>=num_pixels_1d) 
@@ -273,6 +267,15 @@ u08 TexturesArray4AI::operator()(u32 row, u32 column) const
         printf("Index [%d, %d] is out of range [%d, %d]\n", row, column, num_pixels_1d - 1, num_pixels_1d - 1);
         assert(0);
     }
+    if (!this->textures) 
+    {   
+        char buff[512];
+        snprintf(buff, sizeof(buff), "TexturesArray4AI is not initialized, "
+            "please first copy the buffer textures to this->textures[], "
+            "row: %d, column: %d\n", row, column);
+        MY_CHECK(buff);
+        assert(0);
+    }
     if (row>column) std::swap(row, column);
     // cal texture_id 
     // index within the texture
@@ -316,7 +319,7 @@ void TexturesArray4AI::copy_buffer_to_textures(
 
     const GLuint &contact_matrix_textures = contact_matrix_->textures;
     if (is_copied_from_buffer) return;
-    MY_CHECK(0);
+    
     // Temporary texture setup
     GLuint temptexture;
     glGenTextures(1, &temptexture);
@@ -326,7 +329,7 @@ void TexturesArray4AI::copy_buffer_to_textures(
     glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
     glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);
     glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);
-    MY_CHECK(0);
+    
 
     GLuint framebuffer;
     glGenFramebuffers(1, &framebuffer);
@@ -338,7 +341,7 @@ void TexturesArray4AI::copy_buffer_to_textures(
         assert(0);
     }
 
-    MY_CHECK(0);
+    
 
     s32 orignal_viewport[4];
     glGetIntegerv(GL_VIEWPORT, orignal_viewport);
@@ -396,8 +399,6 @@ void TexturesArray4AI::copy_buffer_to_textures_dynamic(
     bool show_flag) 
 {
 
-    MY_CHECK(0);
-    // Temporary texture setup
     GLuint temptexture;
     glGenTextures(1, &temptexture);
     glBindTexture(GL_TEXTURE_2D, temptexture);
@@ -416,12 +417,9 @@ void TexturesArray4AI::copy_buffer_to_textures_dynamic(
         assert(0);
     }
 
-    MY_CHECK(0);
-
     s32 orignal_viewport[4];
     glGetIntegerv(GL_VIEWPORT, orignal_viewport);
 
-    // TODO (shaoheng): the shader is using all 4 channels but readpixel only the red channel
     auto model = glm::mat4(1.0f);
     glUseProgram(contact_matrix_->shaderProgram);
     glUniformMatrix4fv(contact_matrix_->matLocation, 1, GL_FALSE, glm::value_ptr(model));
@@ -849,16 +847,16 @@ void TexturesArray4AI::cal_compressed_hic(
 
     // cal the maximum number of shift
     u32 D = 0;
-    f32 global_shift0_diag_mean = cal_diagonal_mean_within_fragments(D, Contigs);
+    f32 global_shift0_diag_mean = this->cal_diagonal_mean_within_fragments(D, Contigs);
     std::vector<f32> norm_diag_mean = {global_shift0_diag_mean};
-    f32 tmp_diag_mean = cal_diagonal_mean_within_fragments(++D, Contigs);
+    f32 tmp_diag_mean = this->cal_diagonal_mean_within_fragments(++D, Contigs);
     while (
         tmp_diag_mean > min_hic_density && 
         tmp_diag_mean > D_hic_ratio * global_shift0_diag_mean && 
         D < maximum_D
     ) {
         norm_diag_mean.push_back(tmp_diag_mean);
-        tmp_diag_mean = cal_diagonal_mean_within_fragments(++D, Contigs); 
+        tmp_diag_mean = this->cal_diagonal_mean_within_fragments(++D, Contigs); 
     }
 
     Values_on_Channel buffer_values_on_channel;
@@ -867,7 +865,7 @@ void TexturesArray4AI::cal_compressed_hic(
     {
         for (u32 j = i+1; j < frags->num; j++)
         {   
-            get_interaction_score(
+            this->get_interaction_score(
                     frags->startCoord[i], 
                     frags->startCoord[j], 
                     frags->length[i], 
@@ -902,7 +900,7 @@ f32 TexturesArray4AI::cal_diagonal_mean_within_fragments(int shift, const contig
     for (int i = 0 ; i < Contigs->numberOfContigs; i ++ )
     {   
         maxlen = std::max(maxlen, Contigs->contigs_arr[i].length);
-        Sum_and_Number tmp = get_fragement_diag_mean_square( 
+        Sum_and_Number tmp = this->get_fragement_diag_mean_square( 
             start_coord, 
             Contigs->contigs_arr[i].length, 
             shift );
@@ -1157,7 +1155,7 @@ Sum_and_Number TexturesArray4AI::get_fragement_diag_mean_square(
         std::cerr << "The shift is less than 0" << std::endl;
         assert(0);
     }
-    check_copied_from_buffer();
+    this->check_copied_from_buffer();
     if (shift >= length)
     {
         return {0, 0};
diff --git a/src/copy_texture.h b/src/copy_texture.h
index 40df3a3..c72b4e5 100644
--- a/src/copy_texture.h
+++ b/src/copy_texture.h
@@ -409,7 +409,7 @@ struct CompressedExtensions
 };
 
 
-inline // c++17 needed
+inline
 std::unordered_map<u32, u32> 
 Switch_Channel_Symetric = 
 {   
@@ -453,7 +453,9 @@ struct Sum_and_Number
 };
     
 
-
+/*
+GraphData类用于存储图数据，将压缩后的hic数据转化为图数据，从而输入到图网络中预测两个片段之间的连接分数。
+*/
 class GraphData
 {
 private:
@@ -589,7 +591,7 @@ class TexturesArray4AI
     Matrix3D<f32>* compressed_hic = nullptr; // [0 - 3] channel represent the 4 directions, channel 4 represents the average of the interaction block
     CompressedExtensions* compressed_extensions = nullptr;
     MassCentre* mass_centres = nullptr;
-    unsigned char** textures = nullptr;
+    unsigned char** textures = nullptr; // used to save the copied textures
     bool is_copied_from_buffer;
     bool is_compressed;
 public:
@@ -639,6 +641,8 @@ class TexturesArray4AI
 
     u32 get_num_textures_1d(){return num_textures_1d;}
 
+    u32 get_num_pixels_1d() {return num_pixels_1d;}
+
     void check_copied_from_buffer() const;
     void copy_buffer_to_textures(const contact_matrix *contact_matrix_, bool show_flag=false);
     void copy_buffer_to_textures_dynamic(const contact_matrix *contact_matrix_, bool show_flag=false);
diff --git a/src/frag_sort.cpp b/src/frag_sort.cpp
index 5a6ca10..1328415 100644
--- a/src/frag_sort.cpp
+++ b/src/frag_sort.cpp
@@ -227,7 +227,8 @@ void FragSortTool::sort_according_likelihood_unionFind(
     const LikelihoodTable& likelihood_table, 
     FragsOrder& frags_order, 
     const f32 threshold, 
-    const Frag4compress* frags) const 
+    const Frag4compress* frags, 
+    bool sort_according_len_flag) const
 {
     u32 n = likelihood_table.get_num_frags();
     std::vector<Link> all_links;
@@ -323,7 +324,7 @@ void FragSortTool::sort_according_likelihood_unionFind(
     }
 
     // sort the chromosomes according to the length
-    if (frags)
+    if (frags && sort_according_len_flag)
     {
         std::sort(chromosomes.begin(), chromosomes.end(), 
             [&frags](const std::deque<s32>& a, const std::deque<s32>& b) {
@@ -566,7 +567,8 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
     const f32 threshold, 
     const Frag4compress* frags,
     const bool doStageOne,
-    const bool doStageTwo) const 
+    const bool doStageTwo, 
+    bool sort_according_len_flag) const 
 {   
     u32 n = likelihood_table.get_num_frags();
     std::vector<Link> all_links;
@@ -705,8 +707,8 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
             // and the fragments [beg...j) on chb
             // the sequence order in each chromosome keep unchanged
             // are fused to maximum the overall score (measured by link strength)
-            std::deque<s32> atail(0);
-            if (cha.size() > i+1)
+            std::deque<s32> atail(0); // TODO (shaoheng): 检查这里是不是应该填 a 的上一个片段
+            if (cha.size() > i+1) // TODO (shaoheng): 检查这里到底是 i + 1 还是 i
                 std::copy(cha.begin()+i+1, cha.end(), std::back_inserter(atail));
             std::deque<s32> bhead(chb.begin(), chb.begin()+j+1);
 
@@ -736,7 +738,7 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
     fprintf(stdout, "[Fusing]: consumed time %.2fs\n", elapsed_seconds.count());
 
     // sort the chromosomes according to the length
-    if (frags)
+    if (frags && sort_according_len_flag)
     {
         std::sort(chromosomes.begin(), chromosomes.end(), 
             [&frags](const std::deque<s32>& a, const std::deque<s32>& b) {
diff --git a/src/frag_sort.h b/src/frag_sort.h
index 269ef79..50bbc18 100644
--- a/src/frag_sort.h
+++ b/src/frag_sort.h
@@ -266,7 +266,8 @@ class FragSortTool
         const LikelihoodTable& likelihood_table, 
         FragsOrder& frags_order,
         const f32 threshold=-0.001, 
-        const Frag4compress* frags=nullptr) const;
+        const Frag4compress* frags=nullptr, 
+        bool sort_according_len_flag=true) const;
 
     void sort_according_likelihood_unionFind_doFuse(
         const LikelihoodTable& likelihood_table, 
@@ -274,7 +275,8 @@ class FragSortTool
         const f32 threshold=-0.001, 
         const Frag4compress* frags=nullptr,
         const bool doStageOne=true,
-        const bool doStageTwo=false) const;
+        const bool doStageTwo=false, 
+        bool sort_according_len_flag=true) const;
 };
 
 
diff --git a/src/hic_figure.cpp b/src/hic_figure.cpp
new file mode 100644
index 0000000..ce14611
--- /dev/null
+++ b/src/hic_figure.cpp
@@ -0,0 +1 @@
+#include "hic_figure.h"
\ No newline at end of file
diff --git a/src/hic_figure.h b/src/hic_figure.h
new file mode 100644
index 0000000..9658569
--- /dev/null
+++ b/src/hic_figure.h
@@ -0,0 +1,516 @@
+
+
+#ifndef HIC_FIGURE_H
+#define HIC_FIGURE_H
+
+
+#include <iostream>
+#include <random>
+#include <sstream>
+#include <vector>
+#include <queue>
+#include <string>
+#include <filesystem>
+#include <map>
+#include <fstream>
+#include <variant>
+#include <type_traits>
+#include "genomeData.h"
+#include "showWindowData.h" 
+#include "utilsPretextView.h"
+#include "copy_texture.h"
+
+#ifndef STB_IMAGE_RESIZE_IMPLEMENTATION
+    #define STB_IMAGE_RESIZE_IMPLEMENTATION
+    #include "stb_image_resize.h"
+#endif // STB_IMAGE_RESIZE_IMPLEMENTATION
+
+#ifndef STB_IMAGE_WRITE_IMPLEMENTATION
+    #define STB_IMAGE_WRITE_IMPLEMENTATION
+    #include "stb_image_write.h"
+#endif // STB_IMAGE_WRITE_IMPLEMENTATION
+
+
+
+using Dict_json = std::map<std::string, std::variant<u64, f32, std::string>>;
+
+
+/*计算数组的 95% 分位数*/
+template <typename T>
+T percentile_cal(T* data, u32 size, f32 percentile=0.95)
+{   
+    u32 heap_size = (u32)((f32)size * (1-percentile));
+    std::priority_queue<T, std::vector<T>, std::greater<T>> min_heap;
+    for (auto i = 0; i < size; i++)
+    {
+        if (min_heap.size() < heap_size)
+        {
+            min_heap.push(data[i]);
+        }
+        else if (data[i] > min_heap.top())
+        {
+            min_heap.pop();
+            min_heap.push(data[i]);
+        }
+    }
+    if (min_heap.size() > 0) 
+        return min_heap.top();
+    else return 0;
+}
+
+
+
+
+std::string generate_random_string(u32 length) 
+{
+    const std::string chars = 
+        "abcdefghijklmnopqrstuvwxyz"
+        "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
+        "0123456789"; 
+    std::random_device rd;
+    std::mt19937 generator(rd()); 
+    std::uniform_int_distribution<> distrib(0, chars.size() - 1); 
+    std::string result(length, 0);
+    for (u32 i = 0; i < length; ++i) 
+    {
+        result[i] = chars[distrib(generator)]; 
+    }
+    return result;
+}
+
+struct ValueVisitor 
+{
+    std::stringstream& ss;
+
+    void operator()(uint64_t val) 
+    {
+        ss << val;  // 
+    }
+
+    void operator()(float val) 
+    {
+        ss << val;  
+    }
+
+    void operator()(const std::string& val) 
+    {
+        ss << "\"" << val << "\"";  
+    }
+};
+
+
+/*
+hic图片信息
+*/
+struct Figure_Info
+{
+    u64 start_bp;
+    u64 end_bp;
+    u64 genome_bp;
+    f32 bp_per_pixel_figure;
+    f32 bp_per_pixel_hic;
+    u32 figure_pixel_1d;
+    u32 hic_pixel_1d;
+    u32 num_pixels_1d_from_hic;
+    f32 multiplier;
+    std::string file_save_path;    // dir_path/random_name without extension
+    Dict_json info_dic;
+    std::string img_id;
+
+    Figure_Info(
+        u32 figure_pixel_1d_,
+        u32 hic_pixel_1d_,
+        u64 start_bp_, 
+        u64 end_bp_,
+        u64 genome_bp_,
+        f32 bp_per_pixel_figure_, 
+        f32 bp_per_pixel_hic_,
+        f32 multiplier_,
+        u32 num_pixels_1d_from_hic_, 
+        std::string file_save_path_, 
+        std::string img_id_)
+        : 
+        figure_pixel_1d(figure_pixel_1d_),
+        hic_pixel_1d(hic_pixel_1d_),
+        start_bp(start_bp_), 
+        end_bp(end_bp_), 
+        genome_bp(genome_bp_),
+        bp_per_pixel_figure(bp_per_pixel_figure_), 
+        bp_per_pixel_hic(bp_per_pixel_hic_),
+        multiplier(multiplier_),
+        num_pixels_1d_from_hic(num_pixels_1d_from_hic_), 
+        file_save_path(file_save_path_), 
+        img_id(img_id_)
+    {   
+        /*
+        sample: 
+
+        '/lustre/scratch125/ids/team117-assembly/sg35/hic/hali_lig/results/AutoHiC/autohic_results/0/png/500000/00aea05fb7824b5dafb68fa6e280816b.jpg': {
+            'genome_id': 'png', 
+            'resolution': 500000, 
+            'chr_A': 'assembly', 
+            'chr_A_start': 0, 
+            'chr_A_end': 312298520, 
+            'chr_B': 'assembly', 
+            'chr_B_start': 0, 
+            'chr_B_end': 312298520
+        }
+        
+        */
+        this->info_dic["figure_pixel_1d"] = figure_pixel_1d;
+        this->info_dic["hic_pixel_1d"] = hic_pixel_1d;
+        this->info_dic["start_bp"] = start_bp;
+        this->info_dic["end_bp"] = end_bp;
+        this->info_dic["genome_bp"] = genome_bp;
+        this->info_dic["bp_per_pixel_figure"] = bp_per_pixel_figure;
+        this->info_dic["bp_per_pixel_hic"] = bp_per_pixel_hic;
+        this->info_dic["multiplier"] = multiplier;
+        this->info_dic["img_id"] = img_id;
+
+        this->info_dic["genome_id"] = "png";
+        this->info_dic["resolution"] = bp_per_pixel_figure;
+        this->info_dic["chr_A"] = "assembly";
+        this->info_dic["chr_A_start"] = start_bp;
+        this->info_dic["chr_A_end"] = end_bp;
+        this->info_dic["chr_B"] = "assembly";
+        this->info_dic["chr_B_start"] = start_bp;
+        this->info_dic["chr_B_end"] = end_bp;
+    }
+
+    std::string dump_info() {
+        std::stringstream ss;
+        ss << "{";
+        bool first = true;
+        
+        for (const auto& [key, value] : info_dic) {
+            if (!first) ss << ", ";
+            ss << "\"" << key << "\": ";
+            
+            ValueVisitor visitor{ss};
+            std::visit(visitor, value);
+            
+            first = false;
+        }
+        ss << "}";
+        return ss.str();
+    }
+};
+
+/*
+保存 Hi-C 数据，以给定的
+    位置、大小和分辨率 
+    保存 Hi-C 数据图表。
+*/
+class Hic_Figure
+{
+public:
+    u64 genome_bp;
+    u32 hic_pixel_1d;
+    u32 figure_pixel_1d=1116;
+    TexturesArray4AI* textures_array;
+    f32 bp_per_pixel_hic;
+    std::string file_save_dir;
+    std::string auto_cut_output_dir;
+    #ifdef DEBUG
+        std::vector<f32> resolution_mutiplier = {2, 4};
+    #else
+        std::vector<f32> resolution_mutiplier = {
+            0.5, 1, 2, 4, 8, 16, 32};
+    #endif // DEBUG
+    Figure_Info* figure_info=nullptr;
+    
+public:
+    Hic_Figure(
+        u64 genome_bp_,
+        std::string file_save_dir_, 
+        TexturesArray4AI* textures_array_,
+        std::string auto_cut_output_dir_ = "auto_cut_output"
+    ) 
+        : 
+        genome_bp(genome_bp_),
+        hic_pixel_1d(textures_array_->get_num_pixels_1d()),
+        file_save_dir(file_save_dir_),
+        auto_cut_output_dir(auto_cut_output_dir_)
+    {   
+        this->textures_array = textures_array_;
+        this->bp_per_pixel_hic = (f32)this->genome_bp / (f32)this->hic_pixel_1d;
+        if (std::filesystem::exists(this->auto_cut_output_dir))
+        {
+            std::cout << "Remove folder: " 
+                      << (std::filesystem::current_path().string() + this->auto_cut_output_dir) 
+                      << std::endl;
+            std::filesystem::remove_all(this->auto_cut_output_dir);
+            std::filesystem::create_directories(this->auto_cut_output_dir);
+        }
+        // if the folder does not exist, create it
+        if (std::filesystem::exists(file_save_dir))
+        {
+            // remove this one 
+            std::cout << "Remove folder: " 
+                      << (std::filesystem::current_path().string() + file_save_dir) 
+                      << std::endl;
+            std::filesystem::remove_all(file_save_dir);
+        }
+        std::cout << "Creat folder: " 
+                  << (std::filesystem::current_path().string() + file_save_dir) 
+                  << std::endl;
+        std::filesystem::create_directories(file_save_dir);
+        // clear the saved info
+        std::string info_save_file_full_path = this->file_save_dir+"/info.json";
+        std::ofstream out_file(info_save_file_full_path);
+        if (!out_file.is_open())
+        {   
+            std::stringstream ss;
+            ss << "Cannot open: " << info_save_file_full_path << std::endl;
+            MY_CHECK(ss.str().c_str());
+            assert(0);
+        }
+        out_file << "{" << std::endl << "}";
+        out_file.close();
+    }
+
+    ~Hic_Figure(){
+        if (this->figure_info)
+        {
+            delete this->figure_info;
+            this->figure_info = nullptr;
+        }
+    }
+
+    void single_hic_figure_config(u32 start_pixel_figure, f32 bp_per_pixel_figure, f32 multiplier)
+    {
+        u32 start_bp = (u32) (start_pixel_figure * bp_per_pixel_figure);
+        u32 end_bp = (u32) ((start_pixel_figure + this->figure_pixel_1d) * bp_per_pixel_figure);
+        
+        std::string img_id = generate_random_string(16);
+        std::string file_save_path = this->file_save_dir + "/" + img_id;
+
+        if (this->figure_info)
+        {
+            delete this->figure_info;
+            this->figure_info = nullptr;
+        }
+        this->figure_info = new Figure_Info(
+            this->figure_pixel_1d,
+            this->hic_pixel_1d,
+            start_bp, 
+            end_bp, 
+            this->genome_bp,
+            bp_per_pixel_figure, 
+            this->bp_per_pixel_hic,
+            multiplier,
+            this->hic_pixel_1d,
+            file_save_path,
+            img_id);
+        
+        this->dump_info_json(this->figure_info->info_dic);
+    }
+
+
+    void dump_info_json(std::map<std::string, std::variant<u64, f32, std::string>> info)
+    {
+        if (info.size() == 0) 
+        {
+            std::cerr << "The info is empty" << std::endl;
+            assert(0);
+        }
+        std::string info_save_file_full_path = this->file_save_dir+"/info.json";
+        std::fstream out_file = std::fstream(info_save_file_full_path, std::ios::in | std::ios::out);
+        if (!out_file.is_open())
+        {   
+            std::stringstream ss;
+            ss << "Cannot open: " << info_save_file_full_path << std::endl;
+            MY_CHECK(ss.str().c_str());
+            assert(0);
+        }
+        // move cursor to the second last line
+        out_file.seekg(0, std::ios_base::end);
+        u64 ptr = (u64) out_file.tellg() - 1 ;
+        char ch;
+        while (ptr > 0) 
+        {   
+            out_file.seekg(ptr);
+            out_file.get(ch);
+            if (ch == '}') 
+            {   
+                out_file.seekg(ptr-2);
+                out_file.get(ch);
+                if (ch == '}')
+                {   
+                    ptr -- ;
+                    out_file.seekp(ptr);
+                    out_file << "," << std::endl;
+                    ptr += 2 ;
+                }
+                out_file.seekp(ptr);
+                out_file << "\"" << this->file_save_dir << "/" ; 
+                std::visit(
+                    [&out_file](const auto& v) 
+                    {
+                        out_file << v;
+                    }, 
+                    info["img_id"]);
+                out_file << ".png" << "\"" << ": " 
+                         << this->figure_info->dump_info() << std::endl << "}";
+                break;
+            }
+            ptr--;
+        }
+        if (ptr == 0)
+        {
+            char buff[512];
+            snprintf(buff, sizeof(buff), "Cannot find ']' or '}' in the json file: %s\n", info_save_file_full_path.c_str());
+            MY_CHECK(buff);
+            assert(0);
+        }
+        out_file.close();
+    }
+
+    void copy_data_from_hic(
+        u08* data_figure, 
+        u32 start_pixel_hic, 
+        u32 num_pixels_to_copy_from_hic)
+    {   
+        if (num_pixels_to_copy_from_hic > this->hic_pixel_1d)
+        {
+            char buff[512];
+            snprintf(buff, sizeof(buff), "num_pixels_to_copy_from_hic(%d) > total_pixels_1d(%d)\n", num_pixels_to_copy_from_hic, this->hic_pixel_1d);
+            MY_CHECK(buff);
+            assert(0);
+        }
+        start_pixel_hic = my_Min(this->hic_pixel_1d - num_pixels_to_copy_from_hic, start_pixel_hic);
+        for (u32 i = 0; i < num_pixels_to_copy_from_hic; i++)
+        {
+            for (u32 j = 0; j < num_pixels_to_copy_from_hic; j++)
+            {
+                data_figure[i*num_pixels_to_copy_from_hic + j] = textures_array->operator()(i + start_pixel_hic, j + start_pixel_hic);
+            }
+        }
+    }
+
+    void resize_and_save_figure(u08* data_figure, u32 num_pixels_copy_from_hic, f32 gamma = 0.5)
+    {
+        int num_pixels = this->figure_pixel_1d * this->figure_pixel_1d;
+        u08* data_out = new u08[num_pixels ];
+        stbir_resize_uint8(
+            (const u08*)data_figure,      // input buffer
+            num_pixels_copy_from_hic,     // input width
+            num_pixels_copy_from_hic,     // input height
+            0,                            // input stride
+            (u08*)data_out,               // output buffer
+            this->figure_pixel_1d,        // output width
+            this->figure_pixel_1d,        // output height
+            0,                            // output stride
+            1                             // number of channels
+        );
+        char buff[512];
+        snprintf(
+            buff, sizeof(buff), 
+            "%s/%s.png", 
+            this->file_save_dir.c_str(),  
+            this->figure_info->img_id.c_str()
+            // "%s/multiplier%.2f_%.2f-%.2f_%s.png", 
+            // this->file_save_dir.c_str(), 
+            // this->figure_info->multiplier, 
+            // (f32)this->figure_info->start_bp / (f32)this->figure_info->genome_bp,
+            // (f32)this->figure_info->end_bp / (f32)this->figure_info->genome_bp, 
+            // this->figure_info->img_id.c_str()
+        );
+        // save the figure
+
+        // 95 percentile
+        u08 percentile_95 = percentile_cal(data_out, num_pixels, 0.95);
+        f32* data_out_f32 = new f32[num_pixels];
+        for (int i = 0; i < num_pixels; ++i) 
+        {
+            data_out_f32[i] = (f32)my_Min(data_out[i], percentile_95) / (f32)percentile_95;
+        }
+
+        // 分配三通道数据内存（R, G, B）
+        u32 num_channels = 3;
+        u08* rgb_data = new u08[num_pixels * num_channels];
+
+        // 转换逻辑：0 → 白色 (255,255,255)，非0 → 红色 (val,0,0)
+        for (int i = 0; i < num_pixels; ++i) 
+        {   
+            // rgb_data[i*3]   = (u08) (data_out_f32[i] < gamma? 255 : data_out_f32[i] * 255.f);         // R
+            rgb_data[i*num_channels]   = 255;         // R
+            rgb_data[i*num_channels+2] = rgb_data[i*num_channels+1] = (u08) ((1.f - data_out_f32[i]) * 255.f); // G & B
+            // rgb_data[i*num_channels+3] = 255;          // A
+        }
+        delete[] data_out_f32;
+
+        // 生成保存路径（添加_red后缀）
+        snprintf(
+            buff, sizeof(buff), 
+            "%s/%s.png",  
+            this->file_save_dir.c_str(),  
+            this->figure_info->img_id.c_str()
+        );
+
+        stbi_write_png(
+            buff, 
+            this->figure_pixel_1d, 
+            this->figure_pixel_1d, 
+            num_channels,  
+            rgb_data,           
+            this->figure_pixel_1d * num_channels         
+        );
+        delete[] rgb_data;
+        delete[] data_out;
+    }
+
+    void generate_hic_figures()
+    {   
+        u32 figure_num = 0;
+        for (f32 multiplier : this->resolution_mutiplier)
+        {   
+            u32 start_pixel_figure = 0;
+            f32 bp_per_pixel_figure = (f32)this->bp_per_pixel_hic * (f32)multiplier;
+            f32 total_pixel_figure = (f32)this->genome_bp / bp_per_pixel_figure;
+            u32 num_pixels_copy_from_hic = (u32)((f32)this->figure_pixel_1d * multiplier);
+            if (this->figure_pixel_1d > total_pixel_figure) continue;
+            u08* data_figure = new u08[num_pixels_copy_from_hic * num_pixels_copy_from_hic];
+            while (start_pixel_figure < (total_pixel_figure - (this->figure_pixel_1d>>1)))
+            {   
+                if (start_pixel_figure + this->figure_pixel_1d > total_pixel_figure)
+                {
+                    start_pixel_figure = (u32)total_pixel_figure - this->figure_pixel_1d;
+                }
+                this->single_hic_figure_config(
+                    start_pixel_figure, 
+                    bp_per_pixel_figure, 
+                    multiplier);
+                
+                std::cout   << "[" << ++figure_num << "] "
+                            << "Multiplier: " << multiplier 
+                            << ", start_pixel_figure: " << start_pixel_figure << "/" << total_pixel_figure 
+                            << ", start_hic: " << start_pixel_figure * multiplier << "/" << this->hic_pixel_1d
+                            << ", start_bp: " << this->figure_info->start_bp << "/" << this->genome_bp << std::endl;
+
+                this->copy_data_from_hic(
+                    data_figure,
+                    (u32)((f32)start_pixel_figure * multiplier), 
+                    num_pixels_copy_from_hic);
+                this->resize_and_save_figure(data_figure, num_pixels_copy_from_hic);
+                if (start_pixel_figure+this->figure_pixel_1d >= (u32)total_pixel_figure)
+                    break;
+                start_pixel_figure += (this->figure_pixel_1d>>1);
+
+                #ifdef DEBUG    
+                    if (figure_num >= 100) 
+                    {
+                        std::cout << "Debug: figure_num = " << figure_num <<  ", break" << std::endl;
+                        delete[] data_figure;
+                        return;
+                    }
+                #endif // DEBUG
+            }
+            delete[] data_figure;
+        }
+    }
+};
+
+
+
+#endif // HIC_FIGURE_H
\ No newline at end of file
diff --git a/src/utilsPretextView.cpp b/src/utilsPretextView.cpp
index d68c63e..201d24b 100644
--- a/src/utilsPretextView.cpp
+++ b/src/utilsPretextView.cpp
@@ -255,7 +255,7 @@ std::string getResourcesPath() {
 
 
 void my_code_position_handler(const char* file, int line, const char* message) {
-    if (0)
+    if (message)
     {
         std::cerr << "File: " << file << " Line: " << line ;
         if (message) std::cerr << " Message: " << message;
@@ -265,6 +265,17 @@ void my_code_position_handler(const char* file, int line, const char* message) {
 }
 
 
+void my_code_position_handler(const char* file, int line, std::string message) {
+    if (message.size()>0)
+    {
+        std::cerr << "File: " << file << " Line: " << line ;
+        if (message.size()>0) std::cerr << " Message: " << message;
+        std::cerr << std::endl;
+    }
+    return ;
+}
+
+
 
 f64 GetTime() {
     return std::chrono::duration<f64>(

From 2f9b4c35c3f853dd47b59ee39eb1ce4ad5e3664f Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Fri, 7 Mar 2025 13:01:37 +0000
Subject: [PATCH 06/33] auto cut python script

---
 .gitignore                                    |   32 +-
 python/autoCut/auto_cut.py                    |   42 +
 python/autoCut/error_pd.py                    |  879 ++++++++
 python/autoCut/global_settings.py             |   18 +
 python/autoCut/install.sh                     |   13 +
 python/autoCut/logger.py                      |   57 +
 .../cfgs/chr_cascade_mask_rcnn_swin_fpn.py    |  207 ++
 python/autoCut/models/cfgs/chr_model.py       |  140 ++
 python/autoCut/models/cfgs/coco_instance.py   |   48 +
 python/autoCut/models/cfgs/default_runtime.py |   16 +
 .../cfgs/error_cascade_mask_rcnn_swin_fpn.py  |  207 ++
 python/autoCut/models/cfgs/error_model.py     |  140 ++
 python/autoCut/models/cfgs/schedule_1x.py     |   11 +
 python/autoCut/models/swin/mmdet/__init__.py  |   28 +
 .../models/swin/mmdet/apis/__init__.py        |   10 +
 .../models/swin/mmdet/apis/inference.py       |  217 ++
 python/autoCut/models/swin/mmdet/apis/test.py |  190 ++
 .../autoCut/models/swin/mmdet/apis/train.py   |  185 ++
 .../models/swin/mmdet/core/__init__.py        |    7 +
 .../models/swin/mmdet/core/anchor/__init__.py |   11 +
 .../mmdet/core/anchor/anchor_generator.py     |  727 +++++++
 .../models/swin/mmdet/core/anchor/builder.py  |    7 +
 .../swin/mmdet/core/anchor/point_generator.py |   37 +
 .../models/swin/mmdet/core/anchor/utils.py    |   71 +
 .../models/swin/mmdet/core/bbox/__init__.py   |   27 +
 .../mmdet/core/bbox/assigners/__init__.py     |   19 +
 .../bbox/assigners/approx_max_iou_assigner.py |  145 ++
 .../core/bbox/assigners/assign_result.py      |  204 ++
 .../core/bbox/assigners/atss_assigner.py      |  178 ++
 .../core/bbox/assigners/atss_assigner_v2.py   |  205 ++
 .../core/bbox/assigners/base_assigner.py      |    9 +
 .../bbox/assigners/center_region_assigner.py  |  335 +++
 .../core/bbox/assigners/grid_assigner.py      |  155 ++
 .../core/bbox/assigners/hungarian_assigner.py |  145 ++
 .../core/bbox/assigners/max_iou_assigner.py   |  212 ++
 .../core/bbox/assigners/point_assigner.py     |  133 ++
 .../core/bbox/assigners/point_assigner_v2.py  |  127 ++
 .../core/bbox/assigners/point_hm_assigner.py  |  167 ++
 .../core/bbox/assigners/region_assigner.py    |  221 ++
 .../models/swin/mmdet/core/bbox/builder.py    |   20 +
 .../swin/mmdet/core/bbox/coder/__init__.py    |   13 +
 .../mmdet/core/bbox/coder/base_bbox_coder.py  |   17 +
 .../core/bbox/coder/bucketing_bbox_coder.py   |  350 ++++
 .../core/bbox/coder/delta_xywh_bbox_coder.py  |  237 +++
 .../coder/legacy_delta_xywh_bbox_coder.py     |  215 ++
 .../core/bbox/coder/pseudo_bbox_coder.py      |   18 +
 .../mmdet/core/bbox/coder/tblr_bbox_coder.py  |  198 ++
 .../mmdet/core/bbox/coder/yolo_bbox_coder.py  |   89 +
 .../models/swin/mmdet/core/bbox/demodata.py   |   41 +
 .../core/bbox/iou_calculators/__init__.py     |    4 +
 .../core/bbox/iou_calculators/builder.py      |    8 +
 .../bbox/iou_calculators/iou2d_calculator.py  |  159 ++
 .../mmdet/core/bbox/match_costs/__init__.py   |    7 +
 .../mmdet/core/bbox/match_costs/builder.py    |    8 +
 .../mmdet/core/bbox/match_costs/match_cost.py |  184 ++
 .../swin/mmdet/core/bbox/samplers/__init__.py |   15 +
 .../mmdet/core/bbox/samplers/base_sampler.py  |  101 +
 .../core/bbox/samplers/combined_sampler.py    |   20 +
 .../samplers/instance_balanced_pos_sampler.py |   55 +
 .../bbox/samplers/iou_balanced_neg_sampler.py |  157 ++
 .../mmdet/core/bbox/samplers/ohem_sampler.py  |  107 +
 .../core/bbox/samplers/pseudo_sampler.py      |   41 +
 .../core/bbox/samplers/random_sampler.py      |   81 +
 .../core/bbox/samplers/sampling_result.py     |  153 ++
 .../core/bbox/samplers/score_hlr_sampler.py   |  264 +++
 .../models/swin/mmdet/core/bbox/transforms.py |  240 +++
 .../swin/mmdet/core/evaluation/__init__.py    |   15 +
 .../mmdet/core/evaluation/bbox_overlaps.py    |   48 +
 .../swin/mmdet/core/evaluation/class_names.py |  116 ++
 .../swin/mmdet/core/evaluation/eval_hooks.py  |  303 +++
 .../swin/mmdet/core/evaluation/mean_ap.py     |  469 +++++
 .../swin/mmdet/core/evaluation/recall.py      |  189 ++
 .../models/swin/mmdet/core/export/__init__.py |    8 +
 .../swin/mmdet/core/export/pytorch2onnx.py    |  154 ++
 .../models/swin/mmdet/core/mask/__init__.py   |    8 +
 .../swin/mmdet/core/mask/mask_target.py       |  122 ++
 .../models/swin/mmdet/core/mask/structures.py | 1024 ++++++++++
 .../models/swin/mmdet/core/mask/utils.py      |   63 +
 .../mmdet/core/post_processing/__init__.py    |    8 +
 .../mmdet/core/post_processing/bbox_nms.py    |  259 +++
 .../mmdet/core/post_processing/merge_augs.py  |  150 ++
 .../models/swin/mmdet/core/utils/__init__.py  |    7 +
 .../swin/mmdet/core/utils/dist_utils.py       |   69 +
 .../models/swin/mmdet/core/utils/misc.py      |   61 +
 .../swin/mmdet/core/visualization/__init__.py |    4 +
 .../swin/mmdet/core/visualization/image.py    |  303 +++
 .../models/swin/mmdet/datasets/__init__.py    |   24 +
 .../models/swin/mmdet/datasets/builder.py     |  143 ++
 .../models/swin/mmdet/datasets/cityscapes.py  |  334 +++
 .../models/swin/mmdet/datasets/coco.py        |  546 +++++
 .../models/swin/mmdet/datasets/custom.py      |  323 +++
 .../swin/mmdet/datasets/dataset_wrappers.py   |  282 +++
 .../models/swin/mmdet/datasets/deepfashion.py |   10 +
 .../models/swin/mmdet/datasets/lvis.py        |  742 +++++++
 .../swin/mmdet/datasets/pipelines/__init__.py |   25 +
 .../mmdet/datasets/pipelines/auto_augment.py  |  890 ++++++++
 .../swin/mmdet/datasets/pipelines/compose.py  |   51 +
 .../mmdet/datasets/pipelines/formating.py     |  451 ++++
 .../mmdet/datasets/pipelines/instaboost.py    |   98 +
 .../swin/mmdet/datasets/pipelines/loading.py  |  521 +++++
 .../mmdet/datasets/pipelines/test_time_aug.py |  119 ++
 .../mmdet/datasets/pipelines/transforms.py    | 1811 +++++++++++++++++
 .../swin/mmdet/datasets/samplers/__init__.py  |    4 +
 .../datasets/samplers/distributed_sampler.py  |   39 +
 .../mmdet/datasets/samplers/group_sampler.py  |  148 ++
 .../models/swin/mmdet/datasets/utils.py       |  158 ++
 .../autoCut/models/swin/mmdet/datasets/voc.py |   93 +
 .../models/swin/mmdet/datasets/wider_face.py  |   51 +
 .../models/swin/mmdet/datasets/xml_style.py   |  170 ++
 .../models/swin/mmdet/models/__init__.py      |   16 +
 .../swin/mmdet/models/backbones/__init__.py   |   19 +
 .../swin/mmdet/models/backbones/darknet.py    |  199 ++
 .../models/backbones/detectors_resnet.py      |  305 +++
 .../models/backbones/detectors_resnext.py     |  122 ++
 .../swin/mmdet/models/backbones/hourglass.py  |  198 ++
 .../swin/mmdet/models/backbones/hrnet.py      |  537 +++++
 .../swin/mmdet/models/backbones/regnet.py     |  325 +++
 .../swin/mmdet/models/backbones/res2net.py    |  351 ++++
 .../swin/mmdet/models/backbones/resnest.py    |  317 +++
 .../swin/mmdet/models/backbones/resnet.py     |  663 ++++++
 .../swin/mmdet/models/backbones/resnext.py    |  153 ++
 .../swin/mmdet/models/backbones/ssd_vgg.py    |  169 ++
 .../models/backbones/swin_transformer.py      |  630 ++++++
 .../mmdet/models/backbones/trident_resnet.py  |  292 +++
 .../models/swin/mmdet/models/builder.py       |   77 +
 .../swin/mmdet/models/dense_heads/__init__.py |   42 +
 .../models/dense_heads/anchor_free_head.py    |  340 ++++
 .../mmdet/models/dense_heads/anchor_head.py   |  751 +++++++
 .../mmdet/models/dense_heads/atss_head.py     |  689 +++++++
 .../models/dense_heads/base_dense_head.py     |   59 +
 .../models/dense_heads/cascade_rpn_head.py    |  784 +++++++
 .../models/dense_heads/centripetal_head.py    |  421 ++++
 .../mmdet/models/dense_heads/corner_head.py   | 1074 ++++++++++
 .../models/dense_heads/dense_test_mixins.py   |  100 +
 .../models/dense_heads/embedding_rpn_head.py  |  100 +
 .../mmdet/models/dense_heads/fcos_head.py     |  629 ++++++
 .../mmdet/models/dense_heads/fovea_head.py    |  341 ++++
 .../dense_heads/free_anchor_retina_head.py    |  270 +++
 .../mmdet/models/dense_heads/fsaf_head.py     |  422 ++++
 .../models/dense_heads/ga_retina_head.py      |  109 +
 .../mmdet/models/dense_heads/ga_rpn_head.py   |  171 ++
 .../swin/mmdet/models/dense_heads/gfl_head.py |  647 ++++++
 .../models/dense_heads/guided_anchor_head.py  |  860 ++++++++
 .../swin/mmdet/models/dense_heads/ld_head.py  |  261 +++
 .../mmdet/models/dense_heads/nasfcos_head.py  |   75 +
 .../swin/mmdet/models/dense_heads/paa_head.py |  671 ++++++
 .../models/dense_heads/pisa_retinanet_head.py |  154 ++
 .../mmdet/models/dense_heads/pisa_ssd_head.py |  139 ++
 .../models/dense_heads/reppoints_head.py      |  763 +++++++
 .../models/dense_heads/reppoints_v2_head.py   | 1318 ++++++++++++
 .../mmdet/models/dense_heads/retina_head.py   |  114 ++
 .../models/dense_heads/retina_sepbn_head.py   |  113 +
 .../swin/mmdet/models/dense_heads/rpn_head.py |  236 +++
 .../models/dense_heads/rpn_test_mixin.py      |   59 +
 .../models/dense_heads/sabl_retina_head.py    |  621 ++++++
 .../swin/mmdet/models/dense_heads/ssd_head.py |  265 +++
 .../models/dense_heads/transformer_head.py    |  654 ++++++
 .../mmdet/models/dense_heads/vfnet_head.py    |  794 ++++++++
 .../mmdet/models/dense_heads/yolact_head.py   |  943 +++++++++
 .../mmdet/models/dense_heads/yolo_head.py     |  577 ++++++
 .../swin/mmdet/models/detectors/__init__.py   |   42 +
 .../swin/mmdet/models/detectors/atss.py       |   17 +
 .../swin/mmdet/models/detectors/base.py       |  355 ++++
 .../mmdet/models/detectors/cascade_rcnn.py    |   46 +
 .../swin/mmdet/models/detectors/cornernet.py  |   95 +
 .../swin/mmdet/models/detectors/detr.py       |   46 +
 .../swin/mmdet/models/detectors/fast_rcnn.py  |   52 +
 .../mmdet/models/detectors/faster_rcnn.py     |   24 +
 .../swin/mmdet/models/detectors/fcos.py       |   17 +
 .../swin/mmdet/models/detectors/fovea.py      |   17 +
 .../swin/mmdet/models/detectors/fsaf.py       |   17 +
 .../models/swin/mmdet/models/detectors/gfl.py |   16 +
 .../swin/mmdet/models/detectors/grid_rcnn.py  |   29 +
 .../models/swin/mmdet/models/detectors/htc.py |   15 +
 .../mmdet/models/detectors/kd_one_stage.py    |  100 +
 .../swin/mmdet/models/detectors/mask_rcnn.py  |   24 +
 .../detectors/mask_reppoints_v2_detector.py   |  202 ++
 .../models/detectors/mask_scoring_rcnn.py     |   27 +
 .../swin/mmdet/models/detectors/nasfcos.py    |   20 +
 .../models/swin/mmdet/models/detectors/paa.py |   17 +
 .../swin/mmdet/models/detectors/point_rend.py |   29 +
 .../models/detectors/reppoints_detector.py    |   22 +
 .../models/detectors/reppoints_v2_detector.py |  109 +
 .../swin/mmdet/models/detectors/retinanet.py  |   17 +
 .../models/swin/mmdet/models/detectors/rpn.py |  154 ++
 .../swin/mmdet/models/detectors/scnet.py      |   10 +
 .../mmdet/models/detectors/single_stage.py    |  154 ++
 .../mmdet/models/detectors/sparse_rcnn.py     |  110 +
 .../models/detectors/trident_faster_rcnn.py   |   66 +
 .../swin/mmdet/models/detectors/two_stage.py  |  215 ++
 .../swin/mmdet/models/detectors/vfnet.py      |   18 +
 .../swin/mmdet/models/detectors/yolact.py     |  146 ++
 .../swin/mmdet/models/detectors/yolo.py       |   18 +
 .../swin/mmdet/models/losses/__init__.py      |   29 +
 .../swin/mmdet/models/losses/accuracy.py      |   78 +
 .../swin/mmdet/models/losses/ae_loss.py       |  102 +
 .../mmdet/models/losses/balanced_l1_loss.py   |  120 ++
 .../mmdet/models/losses/cross_entropy_loss.py |  214 ++
 .../swin/mmdet/models/losses/focal_loss.py    |  244 +++
 .../models/losses/gaussian_focal_loss.py      |   91 +
 .../swin/mmdet/models/losses/gfocal_loss.py   |  274 +++
 .../swin/mmdet/models/losses/ghm_loss.py      |  172 ++
 .../swin/mmdet/models/losses/iou_loss.py      |  520 +++++
 .../swin/mmdet/models/losses/kd_loss.py       |   87 +
 .../swin/mmdet/models/losses/mse_loss.py      |   49 +
 .../swin/mmdet/models/losses/pisa_loss.py     |  183 ++
 .../mmdet/models/losses/smooth_l1_loss.py     |  139 ++
 .../models/swin/mmdet/models/losses/utils.py  |  100 +
 .../mmdet/models/losses/varifocal_loss.py     |  133 ++
 .../swin/mmdet/models/necks/__init__.py       |   17 +
 .../models/swin/mmdet/models/necks/bfp.py     |  104 +
 .../models/swin/mmdet/models/necks/bifpn.py   |  240 +++
 .../swin/mmdet/models/necks/channel_mapper.py |   74 +
 .../models/swin/mmdet/models/necks/fpg.py     |  398 ++++
 .../models/swin/mmdet/models/necks/fpn.py     |  221 ++
 .../swin/mmdet/models/necks/fpn_carafe.py     |  267 +++
 .../models/swin/mmdet/models/necks/hrfpn.py   |  102 +
 .../models/swin/mmdet/models/necks/nas_fpn.py |  160 ++
 .../swin/mmdet/models/necks/nasfcos_fpn.py    |  161 ++
 .../models/swin/mmdet/models/necks/pafpn.py   |  142 ++
 .../models/swin/mmdet/models/necks/rfp.py     |  128 ++
 .../swin/mmdet/models/necks/yolo_neck.py      |  136 ++
 .../swin/mmdet/models/roi_heads/__init__.py   |   34 +
 .../mmdet/models/roi_heads/base_roi_head.py   |  103 +
 .../models/roi_heads/bbox_heads/__init__.py   |   13 +
 .../models/roi_heads/bbox_heads/bbox_head.py  |  483 +++++
 .../roi_heads/bbox_heads/convfc_bbox_head.py  |  205 ++
 .../models/roi_heads/bbox_heads/dii_head.py   |  415 ++++
 .../roi_heads/bbox_heads/double_bbox_head.py  |  172 ++
 .../models/roi_heads/bbox_heads/sabl_head.py  |  572 ++++++
 .../roi_heads/bbox_heads/scnet_bbox_head.py   |   76 +
 .../models/roi_heads/cascade_roi_head.py      |  507 +++++
 .../mmdet/models/roi_heads/double_roi_head.py |   33 +
 .../models/roi_heads/dynamic_roi_head.py      |  154 ++
 .../mmdet/models/roi_heads/grid_roi_head.py   |  176 ++
 .../mmdet/models/roi_heads/htc_roi_head.py    |  589 ++++++
 .../models/roi_heads/mask_heads/__init__.py   |   18 +
 .../roi_heads/mask_heads/coarse_mask_head.py  |   91 +
 .../mask_heads/condconv_mask_head.py          |  517 +++++
 .../roi_heads/mask_heads/fcn_mask_head.py     |  377 ++++
 .../mask_heads/feature_relay_head.py          |   55 +
 .../mask_heads/fused_semantic_head.py         |  107 +
 .../mask_heads/global_context_head.py         |  102 +
 .../models/roi_heads/mask_heads/grid_head.py  |  359 ++++
 .../roi_heads/mask_heads/htc_mask_head.py     |   43 +
 .../roi_heads/mask_heads/mask_point_head.py   |  300 +++
 .../roi_heads/mask_heads/maskiou_head.py      |  186 ++
 .../roi_heads/mask_heads/scnet_mask_head.py   |   27 +
 .../mask_heads/scnet_semantic_head.py         |   27 +
 .../models/roi_heads/mask_scoring_roi_head.py |  122 ++
 .../mmdet/models/roi_heads/pisa_roi_head.py   |  159 ++
 .../models/roi_heads/point_rend_roi_head.py   |  218 ++
 .../roi_heads/roi_extractors/__init__.py      |    7 +
 .../roi_extractors/base_roi_extractor.py      |   83 +
 .../roi_extractors/generic_roi_extractor.py   |   83 +
 .../single_level_roi_extractor.py             |  108 +
 .../mmdet/models/roi_heads/scnet_roi_head.py  |  582 ++++++
 .../models/roi_heads/shared_heads/__init__.py |    3 +
 .../roi_heads/shared_heads/res_layer.py       |   77 +
 .../mmdet/models/roi_heads/sparse_roi_head.py |  311 +++
 .../models/roi_heads/standard_roi_head.py     |  295 +++
 .../mmdet/models/roi_heads/test_mixins.py     |  348 ++++
 .../models/roi_heads/trident_roi_head.py      |  119 ++
 .../swin/mmdet/models/utils/__init__.py       |   16 +
 .../models/swin/mmdet/models/utils/builder.py |   14 +
 .../mmdet/models/utils/gaussian_target.py     |  185 ++
 .../mmdet/models/utils/positional_encoding.py |  150 ++
 .../swin/mmdet/models/utils/res_layer.py      |  187 ++
 .../swin/mmdet/models/utils/transformer.py    |  860 ++++++++
 .../models/swin/mmdet/utils/__init__.py       |    5 +
 .../models/swin/mmdet/utils/collect_env.py    |   16 +
 .../autoCut/models/swin/mmdet/utils/common.py |   15 +
 .../swin/mmdet/utils/contextmanagers.py       |  121 ++
 .../models/swin/mmdet/utils/instances.py      |  185 ++
 .../autoCut/models/swin/mmdet/utils/logger.py |   19 +
 .../models/swin/mmdet/utils/optimizer.py      |   33 +
 .../models/swin/mmdet/utils/profiling.py      |   39 +
 .../models/swin/mmdet/utils/util_mixins.py    |  104 +
 .../models/swin/mmdet/utils/util_random.py    |   33 +
 python/autoCut/models/swin/mmdet/version.py   |   19 +
 python/autoCut/models/swin/requirements.txt   |    4 +
 .../models/swin/requirements/build.txt        |    3 +
 .../autoCut/models/swin/requirements/docs.txt |    4 +
 .../models/swin/requirements/optional.txt     |    6 +
 .../models/swin/requirements/readthedocs.txt  |    3 +
 .../models/swin/requirements/runtime.txt      |    6 +
 .../models/swin/requirements/tests.txt        |   13 +
 python/autoCut/models/swin/setup.cfg          |   13 +
 python/autoCut/models/swin/setup.py           |  161 ++
 python/autoCut/utils.py                       |   12 +
 290 files changed, 56146 insertions(+), 3 deletions(-)
 create mode 100644 python/autoCut/auto_cut.py
 create mode 100644 python/autoCut/error_pd.py
 create mode 100644 python/autoCut/global_settings.py
 create mode 100644 python/autoCut/install.sh
 create mode 100644 python/autoCut/logger.py
 create mode 100644 python/autoCut/models/cfgs/chr_cascade_mask_rcnn_swin_fpn.py
 create mode 100644 python/autoCut/models/cfgs/chr_model.py
 create mode 100644 python/autoCut/models/cfgs/coco_instance.py
 create mode 100644 python/autoCut/models/cfgs/default_runtime.py
 create mode 100644 python/autoCut/models/cfgs/error_cascade_mask_rcnn_swin_fpn.py
 create mode 100644 python/autoCut/models/cfgs/error_model.py
 create mode 100644 python/autoCut/models/cfgs/schedule_1x.py
 create mode 100644 python/autoCut/models/swin/mmdet/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/apis/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/apis/inference.py
 create mode 100644 python/autoCut/models/swin/mmdet/apis/test.py
 create mode 100644 python/autoCut/models/swin/mmdet/apis/train.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/anchor/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/anchor/anchor_generator.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/anchor/builder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/anchor/point_generator.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/anchor/utils.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/approx_max_iou_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/assign_result.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/atss_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/atss_assigner_v2.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/base_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/center_region_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/grid_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/hungarian_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/max_iou_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/point_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/point_assigner_v2.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/point_hm_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/assigners/region_assigner.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/builder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/coder/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/coder/base_bbox_coder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/coder/bucketing_bbox_coder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/coder/delta_xywh_bbox_coder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/coder/legacy_delta_xywh_bbox_coder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/coder/pseudo_bbox_coder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/coder/tblr_bbox_coder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/coder/yolo_bbox_coder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/demodata.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/builder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/iou2d_calculator.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/match_costs/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/match_costs/builder.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/match_costs/match_cost.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/base_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/combined_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/instance_balanced_pos_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/iou_balanced_neg_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/ohem_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/pseudo_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/random_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/sampling_result.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/samplers/score_hlr_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/bbox/transforms.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/evaluation/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/evaluation/bbox_overlaps.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/evaluation/class_names.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/evaluation/eval_hooks.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/evaluation/mean_ap.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/evaluation/recall.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/export/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/export/pytorch2onnx.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/mask/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/mask/mask_target.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/mask/structures.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/mask/utils.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/post_processing/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/post_processing/bbox_nms.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/post_processing/merge_augs.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/utils/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/utils/dist_utils.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/utils/misc.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/visualization/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/core/visualization/image.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/builder.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/cityscapes.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/coco.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/custom.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/dataset_wrappers.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/deepfashion.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/lvis.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/pipelines/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/pipelines/auto_augment.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/pipelines/compose.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/pipelines/formating.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/pipelines/instaboost.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/pipelines/loading.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/pipelines/test_time_aug.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/pipelines/transforms.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/samplers/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/samplers/distributed_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/samplers/group_sampler.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/utils.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/voc.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/wider_face.py
 create mode 100644 python/autoCut/models/swin/mmdet/datasets/xml_style.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/darknet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/detectors_resnet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/detectors_resnext.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/hourglass.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/hrnet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/regnet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/res2net.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/resnest.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/resnet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/resnext.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/ssd_vgg.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/swin_transformer.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/backbones/trident_resnet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/builder.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/anchor_free_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/anchor_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/atss_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/base_dense_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/cascade_rpn_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/centripetal_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/corner_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/dense_test_mixins.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/embedding_rpn_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/fcos_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/fovea_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/free_anchor_retina_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/fsaf_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/ga_retina_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/ga_rpn_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/gfl_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/guided_anchor_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/ld_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/nasfcos_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/paa_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/pisa_retinanet_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/pisa_ssd_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/reppoints_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/reppoints_v2_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/retina_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/retina_sepbn_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/rpn_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/rpn_test_mixin.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/sabl_retina_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/ssd_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/transformer_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/vfnet_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/yolact_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/dense_heads/yolo_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/atss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/base.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/cascade_rcnn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/cornernet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/detr.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/fast_rcnn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/faster_rcnn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/fcos.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/fovea.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/fsaf.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/gfl.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/grid_rcnn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/htc.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/kd_one_stage.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/mask_rcnn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/mask_reppoints_v2_detector.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/mask_scoring_rcnn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/nasfcos.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/paa.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/point_rend.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/reppoints_detector.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/reppoints_v2_detector.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/retinanet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/rpn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/scnet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/single_stage.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/sparse_rcnn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/trident_faster_rcnn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/two_stage.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/vfnet.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/yolact.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/detectors/yolo.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/accuracy.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/ae_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/balanced_l1_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/cross_entropy_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/focal_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/gaussian_focal_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/gfocal_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/ghm_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/iou_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/kd_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/mse_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/pisa_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/smooth_l1_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/utils.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/losses/varifocal_loss.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/bfp.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/bifpn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/channel_mapper.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/fpg.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/fpn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/fpn_carafe.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/hrfpn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/nas_fpn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/nasfcos_fpn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/pafpn.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/rfp.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/necks/yolo_neck.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/base_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/bbox_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/convfc_bbox_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/dii_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/double_bbox_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/sabl_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/scnet_bbox_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/cascade_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/double_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/dynamic_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/grid_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/htc_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/coarse_mask_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/condconv_mask_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/fcn_mask_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/feature_relay_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/fused_semantic_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/global_context_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/grid_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/htc_mask_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/mask_point_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/maskiou_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/scnet_mask_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/scnet_semantic_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/mask_scoring_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/pisa_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/point_rend_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/base_roi_extractor.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/generic_roi_extractor.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/single_level_roi_extractor.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/scnet_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/shared_heads/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/shared_heads/res_layer.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/sparse_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/standard_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/test_mixins.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/roi_heads/trident_roi_head.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/utils/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/utils/builder.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/utils/gaussian_target.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/utils/positional_encoding.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/utils/res_layer.py
 create mode 100644 python/autoCut/models/swin/mmdet/models/utils/transformer.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/__init__.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/collect_env.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/common.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/contextmanagers.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/instances.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/logger.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/optimizer.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/profiling.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/util_mixins.py
 create mode 100644 python/autoCut/models/swin/mmdet/utils/util_random.py
 create mode 100644 python/autoCut/models/swin/mmdet/version.py
 create mode 100644 python/autoCut/models/swin/requirements.txt
 create mode 100644 python/autoCut/models/swin/requirements/build.txt
 create mode 100644 python/autoCut/models/swin/requirements/docs.txt
 create mode 100644 python/autoCut/models/swin/requirements/optional.txt
 create mode 100644 python/autoCut/models/swin/requirements/readthedocs.txt
 create mode 100644 python/autoCut/models/swin/requirements/runtime.txt
 create mode 100644 python/autoCut/models/swin/requirements/tests.txt
 create mode 100644 python/autoCut/models/swin/setup.cfg
 create mode 100755 python/autoCut/models/swin/setup.py
 create mode 100644 python/autoCut/utils.py

diff --git a/.gitignore b/.gitignore
index 9a429af..4deba74 100644
--- a/.gitignore
+++ b/.gitignore
@@ -3,9 +3,34 @@ app/
 ascii.dSYM/
 build_cmake/
 builddir/
-.DS_Store
+*.DS_Store
+__pycache__/
+*.pyc
+.idea/          # JetBrains IDE
+.vscode/        # Visual Studio Code
+*.suo           # VS Code user settings
+*.swp           # Vim swap files
+*.log          
+*.tmp          
+*.bak           # backup files
+venv/           # 
+env/            # 
+*.egg-info/     # Python egg files
+*.eggs/
 
-python/
+python/comparison_fig/
+python/test_data_from_openglbuffer/
+python/graphData_constructor.py
+python/open_plot_saved_compressed_extension.py
+python/open_plot_saved_compressed_hic.py
+python/open_plot_saved_textures.py
+python/test_cal_mass_center.py
+python/test_compresed_data.py
+python/test.py
+python/compressed_data_test/
+python/autoCut/models/cfgs/error_model.pth
+python/autoCut/models/cfgs/chr_model.pth
+python/autoCut/models/swin/mmdet.egg-info/*
 
 subprojects/eigen
 subprojects/glfw
@@ -48,4 +73,5 @@ test_data/
 subprojects/libtorch(not needed in this version)/
 
 hic_figures/
-auto_cut_output/
\ No newline at end of file
+auto_cut_output/
+python/autoCut/models/swin/README.md
diff --git a/python/autoCut/auto_cut.py b/python/autoCut/auto_cut.py
new file mode 100644
index 0000000..a119e29
--- /dev/null
+++ b/python/autoCut/auto_cut.py
@@ -0,0 +1,42 @@
+import torch 
+import typer
+
+
+from error_pd import infer_error
+from logger import logger
+from global_settings import *
+from utils import find_device
+
+
+def main():
+
+    device = find_device()
+
+    # add hic figures into list
+    hic_figures_list = []
+    for root, dirs, files in os.walk(HIC_FIGURE_DIR):
+        for file in files:
+            if file.endswith(".png"):
+                hic_figures_list.append(os.path.join(root, file))
+    logger.info(f"Found {len(hic_figures_list)} hic figures.")
+
+    # 
+    infer_error_results = infer_error(
+        MODEL_CONFIG, 
+        PRETRAINED_MODEL,
+        HIC_FIGURE_DIR,
+        AUTO_CUT_OUTPUT_DIR,
+        device=device,
+        score = ERROR_SCORE_MIN,
+        error_min_len=15000,
+        error_max_len=int(2e7),
+        iou_score=ERROR_FILTER_IOU_SCORE,
+        chr_len=None
+    )
+    
+    return 
+
+
+
+if __name__ == "__main__":
+    typer.run(main)
diff --git a/python/autoCut/error_pd.py b/python/autoCut/error_pd.py
new file mode 100644
index 0000000..6036375
--- /dev/null
+++ b/python/autoCut/error_pd.py
@@ -0,0 +1,879 @@
+#!/usr/bin/env python3
+# encoding: utf-8 
+
+"""
+@author: jzj
+@contact: jzjlab@163.com
+@file: error_pd.py
+@time: 4/18/23 8:48 PM
+@function: construct error pd
+"""
+
+import json
+import os
+import random
+from collections import defaultdict
+
+import cv2
+import pandas as pd
+from PIL import Image
+from mmdet.apis.inference import init_detector, inference_detector
+
+from logger import logger
+
+from global_settings import *
+
+
+class ERRORS:
+    """
+        Infer error class
+    """
+    __slots__ = "filter_dict", "df", "info_file", "classes", "out_path", "img_size"
+
+    def __init__(self, classes, info_file, out_path, img_size):
+        self.info_file = info_file
+        self.classes = classes
+        self.out_path = out_path
+        self.img_size = img_size
+
+        # 创建一个空的DataFrame
+        self.df = pd.DataFrame()
+
+        # 创建一个记录每次过滤的字典
+        self.filter_dict = dict()
+
+    # generate error structure
+    def create_structure(self, img_info, detection_result):
+        """
+
+        Args:
+            img_info:
+            detection_result:
+
+        Returns:
+
+        """
+        for category, classes in zip(detection_result, self.classes): # 遍历种类
+            for index, error in enumerate(category): # 遍历预测的个数
+                error = error.tolist()
+                temp_dict = dict()
+                temp_dict["image_id"] = list(img_info.keys())[0]
+                temp_dict["category"] = classes
+                temp_dict["bbox_1"], temp_dict["bbox_2"], temp_dict["bbox_3"], temp_dict["bbox_4"] = error[0:4]
+                temp_dict["score"] = round(error[4], 2)
+                temp_dict["resolution"] = img_info[list(img_info.keys())[0]]["resolution"]
+                temp_dict["hic_loci_1"], temp_dict["hic_loci_2"], temp_dict["hic_loci_3"], temp_dict[
+                    "hic_loci_4"] = self.bbox2hic(error[0:4], img_info)
+                
+                logger.info(f"Error score: {error[4]}")
+                # 将字典数据添加到DataFrame中
+                if error[4] > ERROR_SCORE_MIN:
+                    df_new_row = pd.DataFrame(temp_dict, index=[0])
+                    self.df = pd.concat([self.df, df_new_row])
+
+        return self.df
+
+    def zoom_error2excel(self, zoom_error_json_file, output_excel_file="error_summary.xlsx"):
+
+        """
+        保存结果到excel
+        """
+        with open(zoom_error_json_file) as f:
+            json_data = json.load(f)
+
+        error_dict = {
+            "type": [],
+            "start": [],
+            "end": [],
+            "path": []
+        }
+
+        directory = os.path.dirname(zoom_error_json_file)
+        for error_type in json_data:
+            if len(json_data[error_type]) == 0:
+                continue
+            for error in json_data[error_type]:
+                img_basename = file_name = os.path.basename(error["image_id"])
+                error_dict["type"].append(error["category"])
+                error_dict["start"].append(error["hic_loci"][0])
+                error_dict["end"].append(error["hic_loci"][1])
+                error_dict["path"].append(os.path.join(directory, "infer_result", img_basename))
+
+        df = pd.DataFrame(error_dict)
+
+        df.to_excel(output_excel_file, index=False, engine='openpyxl')
+
+    # convert bbox coordinate to hic coordinate
+    def bbox2hic(self, bbox, img_info):
+        """
+            bbox coordinate to hic coordinate
+        Args:
+            bbox: bbox coordinate
+            img_info: hic info
+
+        Returns:
+            hic coordinate
+        """
+        img_size = self.img_size
+        key = list(img_info.keys())[0]
+        # Straw b chromosome
+        img_chr_a_s = img_info[key]["chr_A_start"]
+        img_chr_a_e = img_info[key]["chr_A_end"]
+
+        # Straw a chromosome
+        img_chr_b_s = img_info[key]["chr_B_start"]
+        img_chr_b_e = img_info[key]["chr_B_end"]
+
+        img_chr_w = img_chr_a_e - img_chr_a_s
+        img_chr_h = img_chr_b_e - img_chr_b_s
+
+        w_ration = img_chr_w / img_size[0]
+        h_ration = img_chr_h / img_size[1]
+
+        x, y, w, h = bbox
+
+        a_s = x * w_ration + img_chr_a_s
+        a_e = w * w_ration + img_chr_a_s
+        b_s = y * h_ration + img_chr_b_s
+        b_e = h * h_ration + img_chr_b_s
+
+        # to int
+        hic_loci = list(map(lambda temp: int(temp), [a_s, a_e, b_s, b_e]))
+
+        return hic_loci
+
+    @staticmethod
+    def cal_iou(box1, box2):
+        """
+            calculate iou
+        Args:
+            box1: bbox1
+            box2: bbox2
+
+        Returns:
+            iou
+        """
+        x1min, y1min, x1max, y1max = box1[0], box1[2], box1[1], box1[3]
+        x2min, y2min, x2max, y2max = box2[0], box2[2], box2[1], box2[3]
+
+        # calculate box area
+        s1 = (y1max - y1min + 1.) * (x1max - x1min + 1.)
+        s2 = (y2max - y2min + 1.) * (x2max - x2min + 1.)
+
+        # calculate overlap area
+        x_min = max(x1min, x2min)
+        y_min = max(y1min, y2min)
+        x_max = min(x1max, x2max)
+        y_max = min(y1max, y2max)
+
+        inter_h = max(y_max - y_min + 1, 0)
+        inter_w = max(x_max - x_min + 1, 0)
+
+        intersection = inter_h * inter_w
+        union = s1 + s2 - intersection
+
+        # calculate iou
+        iou = intersection / union
+        return iou
+
+    @staticmethod
+    def transform_bbox(detection_bbox):
+        """
+            transform bbox to [x1, y1, x2, y2]
+        Args:
+            detection_bbox: bbox
+
+        Returns:
+            [x1, y1, x2, y2]
+        """
+        x1 = detection_bbox[0]
+        y1 = detection_bbox[1]
+        x2 = detection_bbox[0] + detection_bbox[2]
+        y2 = detection_bbox[1] + detection_bbox[3]
+        return [x1, y1, x2, y2]
+
+    @staticmethod
+    def if_include(bbox1, bbox2):
+        """
+            if bbox2 include bbox1
+        Args:
+            bbox1: bbox1
+            bbox2: bbox2
+
+        Returns:
+            True or False
+        """
+        # default bbox1 is previous bbox
+        if bbox2[1] <= bbox1[1] and bbox2[2] >= bbox1[2] and bbox2[3] <= bbox1[3]:
+            return True
+        else:
+            return False
+
+    # filter error according to score
+    def filter_all_errors(self, score: float = 0.9, out_path="score_filtered_errors.xlsx", filter_cls=None):
+        """
+            filter error according to score
+        Args:
+            score: score threshold
+            out_path: output path
+            filter_cls: filter classes
+
+        Returns:
+            filtered errors
+        """
+        score_filtered_errors_counter = dict()  # record the number of filtered errors
+        all_errors_counter = dict()
+        if filter_cls is None:
+            filter_cls = self.classes
+        filtered_errors = self.df[self.df['score'] > score]
+
+        # save to excel
+        filtered_errors.to_excel(os.path.join(self.out_path, out_path), sheet_name='Sheet1', index=False)
+
+        for key in filter_cls:
+            score_filtered_errors_counter[key] = filtered_errors[filtered_errors['category'] == key].shape[0]
+            all_errors_counter[key] = self.df[self.df['category'] == key].shape[0]
+
+        self.filter_dict["Raw error number"] = all_errors_counter
+        self.filter_dict["Score filtered error number"] = score_filtered_errors_counter
+
+        return filtered_errors, score_filtered_errors_counter
+
+    def len_filter(self, errors_df, min_len: int = 50000, max_len: int = 10000000,
+                   out_path="len_filtered_errors.xlsx", remove_error_path="len_remove_error.xlsx", filter_cls=None):
+        """
+            filter error according to length
+        Args:
+            errors_df: errors dataframe
+            min_len: error min length
+            max_len: error max length
+            out_path: output path
+            remove_error_path: remove error path
+            filter_cls: filter classes
+
+        Returns:
+            length filtered errors
+        """
+        len_filtered_errors_counter = dict()
+        len_removed_errors_counter = dict()
+
+        if filter_cls is None:
+            filter_cls = self.classes
+
+        filtered_errors = errors_df[
+            (errors_df['hic_loci_2'] - errors_df['hic_loci_1'] > min_len) & (errors_df['hic_loci_2'] - errors_df[
+                'hic_loci_1'] < max_len)]
+        # save to excel
+        filtered_errors.to_excel(os.path.join(self.out_path, out_path), sheet_name='Sheet1', index=False)
+
+        len_removed_errors = errors_df[
+            (errors_df['hic_loci_2'] - errors_df['hic_loci_1'] < min_len) | (errors_df['hic_loci_2'] - errors_df[
+                'hic_loci_1'] > max_len)]
+
+        # save to excel
+        len_removed_errors.to_excel(os.path.join(self.out_path, remove_error_path), sheet_name='Sheet1', index=False)
+
+        for key in filter_cls:
+            try:
+                len_filtered_errors_counter[key] = filtered_errors[filtered_errors['category'] == key].shape[0]
+                len_removed_errors_counter[key] = len_removed_errors[len_removed_errors['category'] == key].shape[0]
+            except KeyError:
+                logger.info(f"KeyError: {key} not in errors_dict")
+                continue
+
+        self.filter_dict["Length filtered error number"] = len_filtered_errors_counter
+        self.filter_dict["Length removed error number"] = len_removed_errors_counter
+
+        return filtered_errors, len_filtered_errors_counter
+
+    def repeat_filter(self, errors_df, error_space, out_path="repeat_errors.xlsx"):
+        """
+            filter repeat error
+        Args:
+            errors_df: errors dataframe
+            error_space: error space
+            out_path: output path
+
+        Returns:
+            repeat filtered errors
+        """
+
+        # select min length translocation > dict default
+        # select translocation
+        tran_pd = errors_df[errors_df["category"] == "translocation"]
+        else__pd = errors_df[errors_df["category"] != "translocation"]
+
+        # sort by hic_loci_1
+        df_sorted = tran_pd.sort_values(by=['hic_loci_1'], na_position='first')
+
+        result_pd = None
+        repeat_pd = None
+        for index in range(len(df_sorted)):
+            if result_pd is None:
+                result_pd = df_sorted.iloc[0:1]
+                continue
+            else:
+                temp_value_1 = df_sorted.iloc[index:index + 1, 8].values[0]
+                temp_value_2 = result_pd.iloc[-1:, 9].values[0]
+                temp_value = abs(temp_value_1 - temp_value_2)
+                if temp_value < error_space:
+                    repeat_pd = pd.concat([repeat_pd, df_sorted[index:index + 1]], axis=0)
+                else:
+                    result_pd = pd.concat([result_pd, df_sorted[index:index + 1]], axis=0)
+        result_pd = pd.concat([result_pd, else__pd], axis=0)
+
+        # save to excel
+        repeat_pd.to_excel(os.path.join(self.out_path, out_path), sheet_name='Sheet1', index=False)
+
+        return result_pd
+
+    def pd2json(self, error_df, out_path="len_filtered_errors.json"):
+        """
+            convert dataframe to json
+        Args:
+            error_df: errors dataframe
+            out_path: output path
+
+        Returns:
+            json file
+        """
+        len_filtered_errors = dict()
+        df = error_df.sort_values("category", inplace=False)
+
+        error_class = df['category'].unique()
+        for class_ in error_class:
+            len_filtered_errors[class_] = []
+        id_counter = 0
+        for index, row in df.iterrows():
+            temp_dict = {
+                "id": id_counter,
+                "image_id": row["image_id"],
+                "category": row["category"],
+                "bbox": [row["bbox_1"], row["bbox_2"], row["bbox_3"], row["bbox_4"]],
+                "score": row["score"],
+                "resolution": row["resolution"],
+                "hic_loci": [row["hic_loci_1"], row["hic_loci_2"], row["hic_loci_3"], row["hic_loci_4"]]
+            }
+            len_filtered_errors[row["category"]].append(temp_dict)
+            id_counter += 1
+        with open(os.path.join(self.out_path, out_path), "w") as outfile:
+            json.dump(len_filtered_errors, outfile)
+
+        return len_filtered_errors
+
+    # filter error according to overlap and iou
+    def de_diff_overlap(self, errors_dict: dict, iou_score: float = 0.8, out_path="overlap_filtered_errors.json",
+                        remove_error_path="overlap_remove_error.txt"):
+        """
+            filter error according to overlap and iou
+        Args:
+            errors_dict: errors dict
+            iou_score: iou score
+            out_path: overlap filtered errors path
+            remove_error_path: overlap removed errors path
+
+        Returns:
+            overlap filtered errors
+        """
+        remove_list = list()  # save the key of the errors_dict which has been removed
+        ans = []  # store de_overlap errors
+        ans_dict = defaultdict()  # store de_overlap errors
+        all_errors = []
+
+        # record the number of overlap filtered errors
+        overlap_filtered_errors_counter = dict()
+
+        for class_ in errors_dict:  # loop classes
+            all_errors += errors_dict[class_]
+            overlap_filtered_errors_counter[class_] = 0  # error counter
+
+        # loop errors
+        sorted_errors_dict = sorted(all_errors, key=lambda itme: itme["hic_loci"][0], reverse=False)
+        temp_compare = None  # first error to compare
+        for error in sorted_errors_dict:
+            # whether there is overlap
+            if ans and error["hic_loci"][0] <= temp_compare["hic_loci"][1]:
+                remove_list.append((error, temp_compare))  # save the error which has overlap
+
+                # calculate overlap ratio
+                bbox1 = error["hic_loci"]
+                bbox2 = temp_compare["hic_loci"]
+                counted_score = self.cal_iou(bbox1, bbox2)
+                if float(counted_score) > float(iou_score): # if the overlap larger than iou_score, treat as same error. then save the error with higher score
+                    # judge which one's resolution is higher
+                    if int(error["resolution"]) == int(temp_compare["resolution"]):
+                        insert_result = max(error, temp_compare, key=lambda item: item["score"]) # select the error with the higher score
+                    else:
+                        # select the error with the highest resolution
+                        insert_result = min(error, temp_compare, key=lambda item: item["resolution"]) # select the error with the higher resolution (smaller in resolution value)
+
+                    if insert_result == temp_compare:
+                        continue
+                    else:
+                        del ans[-1]
+                        ans.append(insert_result)
+
+                else:  # not same error but have overlap, save the larger one, deprecate the smaller one
+                    error_len_1 = bbox1[1] - bbox1[0]
+                    temp_compare_len_2 = bbox2[1] - bbox2[0]
+                    if error_len_1 > temp_compare_len_2:
+                        del ans[-1]
+                        ans.append(error)
+            else:  # nought overlap
+                ans.append(error)
+            temp_compare = ans[-1]
+
+        # regenerate error structure
+        for _ in ans:
+            # ans_dict[_["category"]] = _
+            ans_dict.setdefault(_["category"], []).append(_)
+
+            # count the number of errors
+            overlap_filtered_errors_counter[_["category"]] += 1
+
+        # save remove error to file
+        with open(os.path.join(self.out_path, remove_error_path), "w") as f:
+            f.write(str(remove_list))
+
+        with open(os.path.join(self.out_path, out_path), "w") as outfile:
+            json.dump(ans_dict, outfile)
+
+        self.filter_dict["Overlap filtered error number"] = overlap_filtered_errors_counter
+
+        logger.info("Filter all error category Done")
+
+        self.zoom_error2excel(os.path.join(self.out_path, out_path), os.path.join(self.out_path, "overlap_filtered_errors.xlsx"))
+
+        return ans_dict, overlap_filtered_errors_counter
+
+    def chr_len_filter(self, errors_dict: dict, chr_len: int = None,
+                       out_path="chr_len_filtered_errors.json",
+                       remove_error_path="chr_len_remove_error.txt", filter_cls=None):
+        """
+            filter error according to chr_len
+        Args:
+            errors_dict: errors dict
+            chr_len: chromosome length
+            out_path: output path
+            remove_error_path: remove error path
+            filter_cls: filter class
+
+        Returns:
+            chromosome length filtered errors
+        """
+        if filter_cls is None:
+            filter_cls = self.classes
+        filtered_errors = dict()
+        chr_len_removed_errors = dict()
+        chr_len_filtered_errors_counter = dict()
+        if chr_len is None:
+            logger.info("chr_len is None, skip chr_len check")
+            filtered_errors = errors_dict
+            for key in filter_cls:
+                try:
+                    chr_len_filtered_errors_counter[key] = {
+                        "normal": len(filtered_errors[key]),
+                        "abnormal": 0
+                    }
+                except KeyError:
+                    logger.info(f"KeyError: {key} not in errors_dict")
+                    chr_len_filtered_errors_counter[key] = {
+                        "normal": 0,
+                        "abnormal": 0
+                    }
+            return filtered_errors, chr_len_filtered_errors_counter
+
+
+        for key in filter_cls:
+            try:
+                filtered_errors[key] = list(
+                    filter(lambda x: x["hic_loci"][1] < chr_len, errors_dict[key]))
+                chr_len_removed_errors[key] = list(
+                    filter(lambda x: x["hic_loci"][1] > chr_len, errors_dict[key]))
+
+                chr_len_filtered_errors_counter[key] = {
+                    "normal": len(filtered_errors[key]),
+                    "abnormal": len(chr_len_removed_errors[key])
+                }
+            except KeyError:
+                logger.info(f"KeyError: {key} not in errors_dict")
+                chr_len_filtered_errors_counter[key] = {
+                    "normal": 0,
+                    "abnormal": 0
+                }
+                continue
+
+        with open(os.path.join(self.out_path, out_path), "w") as outfile:
+            json.dump(filtered_errors, outfile)
+
+        with open(os.path.join(self.out_path, remove_error_path), "w") as outfile:
+            json.dump(chr_len_removed_errors, outfile)
+
+        self.filter_dict["Chromosome real length filtered error number"] = chr_len_filtered_errors_counter
+        with open(os.path.join(self.out_path, "error_summary.json"), "a") as outfile:
+            json.dump(self.filter_dict, outfile)
+            logger.info("Save \'error_summary.json\' within Function \'chr_len_filter\'")
+
+        self.zoom_error2excel(os.path.join(self.out_path, out_path),
+                              os.path.join(self.out_path, "chr_len_filtered_errors.xlsx"))
+
+        return filtered_errors, chr_len_filtered_errors_counter
+
+    def loci_zoom(self, errors_dict, threshold=3000, out_path="zoomed_errors.json", filter_cls=None):
+        """
+            zoom error loci
+        Args:
+            errors_dict: errors dict
+            threshold: zoom threshold
+            out_path: output path
+            filter_cls: filter class, if None, then filter self.classes
+
+        Returns:
+            zoomed errors
+        """
+        logger.info(f"zoom threshold: {threshold}")
+
+        if filter_cls is None:
+            filter_cls = self.classes
+        zoomed_errors = dict()
+
+        for key in filter_cls:
+            try:
+                zoomed_errors[key] = []
+                for error in errors_dict[key]:
+                    error["hic_loci"][0] = error["hic_loci"][0] + threshold
+                    error["hic_loci"][1] = error["hic_loci"][1] - threshold
+                    zoomed_errors[key].append(error)
+            except KeyError:
+                logger.info(f"KeyError: {key} not in errors_dict")
+                continue
+
+        with open(os.path.join(self.out_path, out_path), "w") as outfile:
+            json.dump(zoomed_errors, outfile)
+
+        self.zoom_error2excel(os.path.join(self.out_path, out_path), os.path.join(self.out_path, "error_summary.xlsx"))
+
+        return zoomed_errors
+
+    def divide_error(self, all_filtered_error: dict):
+        """
+            divide error according to class
+        Args:
+            all_filtered_error: all filtered error
+
+        Returns:
+            divided error
+        """
+        for _class in self.classes:
+            if _class in all_filtered_error.keys() and all_filtered_error[_class]:
+                divided_error = dict()
+                for tran_error in all_filtered_error[_class]:
+                    divided_error[tran_error["id"]] = {  # one key may have multiple values
+                        "start": tran_error["hic_loci"][0],
+                        "end": tran_error["hic_loci"][1],
+                    }
+                with open(os.path.join(self.out_path, _class + "_error.json"), "w") as outfile:
+                    json.dump(divided_error, outfile)
+            else:
+                continue
+        logger.info("Divide all error category Done")
+
+
+def draw_box_corner(draw_img, bbox, length, corner_color):
+    """
+        draw box corner
+    Args:
+        draw_img: image
+        bbox: bbox
+        length: length
+        corner_color: corner color
+
+    Returns:
+        None
+    """
+    # Top Left
+    cv2.line(draw_img, (bbox[0], bbox[1]), (bbox[0] + length, bbox[1]), corner_color, thickness=2)
+    cv2.line(draw_img, (bbox[0], bbox[1]), (bbox[0], bbox[1] + length), corner_color, thickness=2)
+    # Top Right
+    cv2.line(draw_img, (bbox[2], bbox[1]), (bbox[2] - length, bbox[1]), corner_color, thickness=2)
+    cv2.line(draw_img, (bbox[2], bbox[1]), (bbox[2], bbox[1] + length), corner_color, thickness=2)
+    # Bottom Left
+    cv2.line(draw_img, (bbox[0], bbox[3]), (bbox[0] + length, bbox[3]), corner_color, thickness=2)
+    cv2.line(draw_img, (bbox[0], bbox[3]), (bbox[0], bbox[3] - length), corner_color, thickness=2)
+    # Bottom Right
+    cv2.line(draw_img, (bbox[2], bbox[3]), (bbox[2] - length, bbox[3]), corner_color, thickness=2)
+    cv2.line(draw_img, (bbox[2], bbox[3]), (bbox[2], bbox[3] - length), corner_color, thickness=2)
+
+
+def draw_label_type(draw_img, bbox, label_color):
+    """
+        draw label type
+    Args:
+        draw_img: image
+        bbox: bbox
+        label_color: label color
+
+    Returns:
+        None
+    """
+    label = str(bbox[-1])
+    label_size = cv2.getTextSize(label + '0', cv2.FONT_HERSHEY_TRIPLEX, 0.7, 2)[0]
+    if bbox[1] - label_size[1] - 3 < 0:
+        cv2.rectangle(draw_img,
+                      (bbox[0], bbox[1] + 2),
+                      (bbox[0] + label_size[0], bbox[1] + label_size[1] + 3),
+                      color=label_color,
+                      thickness=-1
+                      )
+        cv2.putText(draw_img, label,
+                    (bbox[0], bbox[1] + label_size[1] + 3),
+                    cv2.FONT_HERSHEY_TRIPLEX,
+                    0.7,
+                    (0, 0, 0),
+                    thickness=1
+                    )
+    else:
+        cv2.rectangle(draw_img,
+                      (bbox[0], bbox[1] - label_size[1] - 3),
+                      (bbox[0] + label_size[0], bbox[1] - 3),
+                      color=label_color,
+                      thickness=-1
+                      )
+        cv2.putText(draw_img, label,
+                    (bbox[0], bbox[1] - 3),
+                    cv2.FONT_HERSHEY_TRIPLEX,
+                    0.7,
+                    (0, 0, 0),
+                    thickness=1
+                    )
+
+
+def test_corner_box(img, bbox, corner_l=20, is_transparent=False, draw_type=False, draw_corner=False,
+                    box_color=(255, 0, 255)):
+    """
+        test corner box
+    Args:
+        img: image
+        bbox: bbox
+        corner_l: corner length
+        is_transparent: transparent
+        draw_type: type
+        draw_corner: corner
+        box_color: box color
+
+    Returns:
+        None
+    """
+    draw_img = img.copy()
+    pt1 = (bbox[0], bbox[1])
+    pt2 = (bbox[2], bbox[3])
+
+    out_img = img
+    if is_transparent:
+        alpha = 0.8
+        # alpha = 0.5
+        cv2.rectangle(draw_img, pt1, pt2, color=box_color, thickness=-1)
+        out_img = cv2.addWeighted(img, alpha, draw_img, 1 - alpha, 0)
+
+    cv2.rectangle(out_img, pt1, pt2, color=box_color, thickness=2)
+
+    if draw_type:
+        draw_label_type(out_img, bbox, label_color=box_color)
+    if draw_corner:
+        draw_box_corner(out_img, bbox, length=corner_l, corner_color=(0, 255, 0))
+    return out_img
+
+
+def bbox2jpg(img_path, bbox, label, out_path):
+    """
+        bbox to jpg
+    Args:
+        img_path: image path
+        bbox: bbox
+        label: label
+        out_path: out path
+
+    Returns:
+        None
+    """
+    img = cv2.imread(img_path)
+
+    bbox = [int(x) for x in bbox] + [label]
+    box_color = (255, 144, 30)
+    out_img = test_corner_box(img, bbox, corner_l=30, is_transparent=True, draw_type=True, draw_corner=True,
+                              box_color=box_color)
+
+    cv2.imwrite(out_path, out_img)
+
+
+def vis_error(error_xlsx, out_dir):
+    """
+        vis error
+    Args:
+        error_xlsx: error xlsx
+        out_dir: out dir
+
+    Returns:
+        None
+    """
+    df = pd.read_excel(error_xlsx)
+    if not os.path.exists(out_dir):  # check if folder is existing
+        os.mkdir(out_dir)
+    for index, row in df.iterrows():
+        img_path = row['image_id']
+        bbox = [row['bbox_1'], row['bbox_2'], row['bbox_3'], row['bbox_4']]
+        label = row['category']
+        basename = str(index + 1) + "_" + os.path.basename(img_path)
+        out_path = os.path.join(out_dir, basename)
+        bbox2jpg(img_path, bbox, label, out_path)
+
+
+def json_vis(error_json, out_dir):
+    """
+        json visulization
+    Args:
+        error_json: error json
+        out_dir: out dir
+
+    Returns:
+        None
+    """
+    with open(error_json, 'r') as f:
+        error_dict = json.load(f)
+    logger.info("Done loading json file.")
+    for key in error_dict.keys():
+        for index, error in enumerate(error_dict[key]):
+            basename = str(index + 1) + "_" + os.path.basename(error["image_id"])
+            out_path = os.path.join(out_dir, basename)
+            error_dict[key][index]["infer_image"] = out_path
+            bbox2jpg(error["image_id"], error["bbox"], error["category"], out_path)
+            if DEBUG:
+                logger.info(f"Figure with error annotation generated: {basename}")
+
+    with open(os.path.join(out_dir, "infer_result.json"), "a") as outfile:
+        json.dump(error_dict, outfile)
+
+
+def infer_error(
+        model_cfg, 
+        pretrained_model, 
+        img_path, 
+        out_path, 
+        device='cuda:0', 
+        score=0.9, 
+        error_min_len=15000,
+        error_max_len=20000000, 
+        iou_score=0.8, 
+        chr_len=None):
+    """
+        infer error
+    Args:
+        model_cfg: model config path
+        pretrained_model: pretrained model path
+        img_path: image path
+        out_path: out path
+        device: GPU device or CPU
+        score: infer score
+        error_min_len: error min length
+        error_max_len: error max length
+        iou_score: iou score
+        chr_len: chromosome length
+
+    Returns:
+        None
+    """
+
+    # Initializing model
+    model = init_detector(model_cfg, pretrained_model, device=device)
+
+    info_file = os.path.join(img_path, "info.json")
+    infos = []
+    with open(info_file, "r") as f:
+        lines = ''.join(line.replace('\n', '') for line in f.readlines())
+    tmp = json.loads(lines)
+    infos = [{i: tmp[i]} for i in tmp]
+
+    classes = ("translocation", "inversion", "debris")
+
+    if not os.path.exists(out_path):  # check if folder is existing
+        os.mkdir(out_path)
+
+    # get img size
+    img_size = None
+
+    for item in os.listdir(img_path):
+        item_path = os.path.join(img_path, item)
+        if item_path.endswith(".png"):
+            img_size = Image.open(item_path).size
+            print("img size: ", img_size)
+            break
+
+    error_class = ERRORS(classes, info_file, out_path, img_size=img_size)
+
+    # for info in infos:
+    for i, info in enumerate(infos):
+        detection_result = inference_detector(
+            model, 
+            list(info.keys())[0])
+        logger.info(f"Processing: [{i}]: {list(info.keys())[0]}")
+
+        error_class.create_structure(info, detection_result[0])
+        # if DEBUG and error_class.df.size>0: break
+            
+
+    if len(error_class.df) == 0:  # no detect error
+        return True
+
+    # score filter
+    score_filtered_errors, score_filtered_errors_counter = error_class.filter_all_errors(
+        score=score,
+        filter_cls=classes)
+
+    # length filter
+    len_filtered_errors, len_filtered_errors_counter = error_class.len_filter(
+        score_filtered_errors,
+        min_len=error_min_len,
+        max_len=error_max_len,
+        out_path="len_filtered_errors.xlsx",
+        remove_error_path="len_remove_error.xlsx",
+        filter_cls=None)
+    # dataframe to json
+    len_filtered_errors_json = error_class.pd2json(len_filtered_errors)
+
+    # overlap filter
+    overlap_filtered_errors, overlap_filtered_errors_counter = error_class.de_diff_overlap(
+        len_filtered_errors_json,
+        iou_score=iou_score)
+
+    # chromosome length filter
+    chr_filtered_errors, chr_filtered_errors_counter = error_class.chr_len_filter(
+        overlap_filtered_errors,
+        chr_len=chr_len)
+
+    # zoom error
+    zoom_threshold = 0
+    zoomed_error = error_class.loci_zoom(
+        chr_filtered_errors, 
+        threshold=zoom_threshold, 
+        out_path="zoomed_errors.json",
+        filter_cls=None)
+
+    # divide error to each json
+    error_class.divide_error(zoomed_error)
+
+    # error visulization
+    # infer result out_dir
+    infer_out_dir = os.path.join(out_path, "infer_result")
+    if not os.path.exists(infer_out_dir):  # check if folder is existing
+        os.mkdir(infer_out_dir)
+
+    error_json = os.path.join(out_path, "chr_len_filtered_errors.json" if chr_len!=None else "overlap_filtered_errors.json")
+    json_vis(error_json, infer_out_dir)
+
+
+def main():
+    pass
+
+
+if __name__ == "__main__":
+    main()
diff --git a/python/autoCut/global_settings.py b/python/autoCut/global_settings.py
new file mode 100644
index 0000000..e887d8e
--- /dev/null
+++ b/python/autoCut/global_settings.py
@@ -0,0 +1,18 @@
+
+
+import os
+
+DEBUG = False
+SKIP_STEP_1=True         # Skip the step 1: generate the hic file (juicer) and assemble the genome (3D-DNA)
+SKIP_HIC_IMG_GEN=True    # Skip the hic figure generation (hicstraw)
+
+
+AUTO_CUT_DIR = "python/autoCut"
+MODEL_CONFIG = os.path.join(AUTO_CUT_DIR, "models/cfgs/error_model.py")
+PRETRAINED_MODEL = os.path.join(AUTO_CUT_DIR, "models/cfgs/error_model.pth")
+# path to hic figures
+HIC_FIGURE_DIR = "hic_figures"
+AUTO_CUT_OUTPUT_DIR = "auto_cut_output"
+ERROR_FILTER_IOU_SCORE=0.8 # used to filter out the errors with overlap area more than 0.8
+ERROR_SCORE_MIN = 0.2
+
diff --git a/python/autoCut/install.sh b/python/autoCut/install.sh
new file mode 100644
index 0000000..f5b5e64
--- /dev/null
+++ b/python/autoCut/install.sh
@@ -0,0 +1,13 @@
+#!/bin/bash
+
+
+# wget https://github.com/open-mmlab/mmdetection/archive/refs/tags/v2.12.0.tar.gz
+
+# tar -zxvf v2.12.0.tar.gz
+# cd mmdetection-2.12.0
+# cd python/autoCut/models
+conda create -n auto_cut python=3.9.16 cython openmim typer  -y
+conda activate auto_cut
+pip install torch==1.10.1 torchvision==0.11.2 matplotlib==3.5.3 pycocotools==2.0.6 numpy==1.24.3 six==1.16.0 terminaltables==3.1.10 timm==0.6.13 tomli==2.0.1 platformdirs==4.3.6 pandas==2.0.1 openpyxl==3.0.10
+mim install mmengine mmcv-full==1.3.7 
+pip install -e ./swin
diff --git a/python/autoCut/logger.py b/python/autoCut/logger.py
new file mode 100644
index 0000000..72639a4
--- /dev/null
+++ b/python/autoCut/logger.py
@@ -0,0 +1,57 @@
+#!/usr/scripts/env python
+# encoding: utf-8
+
+"""
+@author: jzj
+@contact: jzjlab@163.com
+@file: logger.py
+@time: 6/14/22 11:38 AM
+@function: configure logger
+"""
+
+import logging
+
+
+class LoggerHandler(logging.Logger):
+
+    def __init__(
+            self,
+            name="root",
+            level='DEBUG',
+            file=None):
+        super().__init__(name)
+
+        # set log level
+        self.setLevel(level)
+
+        # format setting
+        console_formatter = logging.Formatter(
+            fmt='[%(asctime)s] %(filename)s -> %(funcName)s line:%(lineno)d [%(levelname)s] : %(message)s',
+            datefmt='%Y-%m-%d  %H:%M:%S')
+
+        # initialing Handler
+        if file:
+            file_handle = logging.FileHandler(file)
+            file_handle.setLevel(level)
+
+            self.addHandler(file_handle)
+            file_handle.setFormatter(console_formatter)
+        stream_handler = logging.StreamHandler()
+
+        # set handler level
+        stream_handler.setLevel(level)
+
+        self.addHandler(stream_handler)
+        stream_handler.setFormatter(console_formatter)
+
+
+# initialing logger
+logger = LoggerHandler()
+
+if __name__ == '__main__':
+    logger = LoggerHandler()
+    logger.debug('debug ')
+    logger.info('info')
+    logger.warning('warning')
+    logger.error('error')
+    logger.critical('critical')
diff --git a/python/autoCut/models/cfgs/chr_cascade_mask_rcnn_swin_fpn.py b/python/autoCut/models/cfgs/chr_cascade_mask_rcnn_swin_fpn.py
new file mode 100644
index 0000000..7b8f016
--- /dev/null
+++ b/python/autoCut/models/cfgs/chr_cascade_mask_rcnn_swin_fpn.py
@@ -0,0 +1,207 @@
+# model settings
+model = dict(
+    type='CascadeRCNN',
+    pretrained=None,
+    backbone=dict(
+        type='SwinTransformer',
+        embed_dim=96,
+        depths=[2, 2, 6, 2],
+        num_heads=[3, 6, 12, 24],
+        window_size=7,
+        mlp_ratio=4.,
+        qkv_bias=True,
+        qk_scale=None,
+        drop_rate=0.,
+        attn_drop_rate=0.,
+        drop_path_rate=0.2,
+        ape=False,
+        patch_norm=True,
+        out_indices=(0, 1, 2, 3),
+        use_checkpoint=False),
+    neck=dict(
+        type='FPN',
+        in_channels=[96, 192, 384, 768],
+        out_channels=256,
+        num_outs=5),
+    rpn_head=dict(
+        type='RPNHead',
+        in_channels=256,
+        feat_channels=256,
+        anchor_generator=dict(
+            type='AnchorGenerator',
+            scales=[8],
+            ratios=[0.5, 1.0, 2.0],
+            strides=[4, 8, 16, 32, 64]),
+        bbox_coder=dict(
+            type='DeltaXYWHBBoxCoder',
+            target_means=[.0, .0, .0, .0],
+            target_stds=[1.0, 1.0, 1.0, 1.0]),
+        loss_cls=dict(
+            type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0)),
+    roi_head=dict(
+        type='CascadeRoIHead',
+        num_stages=3,
+        stage_loss_weights=[1, 0.5, 0.25],
+        bbox_roi_extractor=dict(
+            type='SingleRoIExtractor',
+            roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0),
+            out_channels=256,
+            featmap_strides=[4, 8, 16, 32]),
+        bbox_head=[
+            dict(
+                type='Shared2FCBBoxHead',
+                in_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=1,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.1, 0.1, 0.2, 0.2]),
+                reg_class_agnostic=True,
+                loss_cls=dict(
+                    type='CrossEntropyLoss',
+                    use_sigmoid=False,
+                    loss_weight=1.0),
+                loss_bbox=dict(type='SmoothL1Loss', beta=1.0,
+                               loss_weight=1.0)),
+            dict(
+                type='Shared2FCBBoxHead',
+                in_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=1,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.05, 0.05, 0.1, 0.1]),
+                reg_class_agnostic=True,
+                loss_cls=dict(
+                    type='CrossEntropyLoss',
+                    use_sigmoid=False,
+                    loss_weight=1.0),
+                loss_bbox=dict(type='SmoothL1Loss', beta=1.0,
+                               loss_weight=1.0)),
+            dict(
+                type='Shared2FCBBoxHead',
+                in_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=1,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.033, 0.033, 0.067, 0.067]),
+                reg_class_agnostic=True,
+                loss_cls=dict(
+                    type='CrossEntropyLoss',
+                    use_sigmoid=False,
+                    loss_weight=1.0),
+                loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0))
+        ],
+        mask_roi_extractor=dict(
+            type='SingleRoIExtractor',
+            roi_layer=dict(type='RoIAlign', output_size=14, sampling_ratio=0),
+            out_channels=256,
+            featmap_strides=[4, 8, 16, 32]),
+        mask_head=dict(
+            type='FCNMaskHead',
+            num_convs=4,
+            in_channels=256,
+            conv_out_channels=256,
+            num_classes=1,
+            loss_mask=dict(
+                type='CrossEntropyLoss', use_mask=True, loss_weight=1.0))),
+    # model training and testing settings
+    train_cfg = dict(
+        rpn=dict(
+            assigner=dict(
+                type='MaxIoUAssigner',
+                pos_iou_thr=0.7,
+                neg_iou_thr=0.3,
+                min_pos_iou=0.3,
+                match_low_quality=True,
+                ignore_iof_thr=-1),
+            sampler=dict(
+                type='RandomSampler',
+                num=256,
+                pos_fraction=0.5,
+                neg_pos_ub=-1,
+                add_gt_as_proposals=False),
+            allowed_border=0,
+            pos_weight=-1,
+            debug=False),
+        rpn_proposal=dict(
+            nms_across_levels=False,
+            nms_pre=2000,
+            nms_post=2000,
+            max_per_img=2000,
+            nms=dict(type='nms', iou_threshold=0.7),
+            min_bbox_size=0),
+        rcnn=[
+            dict(
+                assigner=dict(
+                    type='MaxIoUAssigner',
+                    pos_iou_thr=0.5,
+                    neg_iou_thr=0.5,
+                    min_pos_iou=0.5,
+                    match_low_quality=False,
+                    ignore_iof_thr=-1),
+                sampler=dict(
+                    type='RandomSampler',
+                    num=512,
+                    pos_fraction=0.25,
+                    neg_pos_ub=-1,
+                    add_gt_as_proposals=True),
+                mask_size=28,
+                pos_weight=-1,
+                debug=False),
+            dict(
+                assigner=dict(
+                    type='MaxIoUAssigner',
+                    pos_iou_thr=0.6,
+                    neg_iou_thr=0.6,
+                    min_pos_iou=0.6,
+                    match_low_quality=False,
+                    ignore_iof_thr=-1),
+                sampler=dict(
+                    type='RandomSampler',
+                    num=512,
+                    pos_fraction=0.25,
+                    neg_pos_ub=-1,
+                    add_gt_as_proposals=True),
+                mask_size=28,
+                pos_weight=-1,
+                debug=False),
+            dict(
+                assigner=dict(
+                    type='MaxIoUAssigner',
+                    pos_iou_thr=0.7,
+                    neg_iou_thr=0.7,
+                    min_pos_iou=0.7,
+                    match_low_quality=False,
+                    ignore_iof_thr=-1),
+                sampler=dict(
+                    type='RandomSampler',
+                    num=512,
+                    pos_fraction=0.25,
+                    neg_pos_ub=-1,
+                    add_gt_as_proposals=True),
+                mask_size=28,
+                pos_weight=-1,
+                debug=False)
+        ]),
+    test_cfg = dict(
+        rpn=dict(
+            nms_across_levels=False,
+            nms_pre=1000,
+            nms_post=1000,
+            max_per_img=1000,
+            nms=dict(type='nms', iou_threshold=0.7),
+            min_bbox_size=0),
+        rcnn=dict(
+            score_thr=0.05,
+            nms=dict(type='nms', iou_threshold=0.5),
+            max_per_img=100,
+            mask_thr_binary=0.5)))
diff --git a/python/autoCut/models/cfgs/chr_model.py b/python/autoCut/models/cfgs/chr_model.py
new file mode 100644
index 0000000..6a88a51
--- /dev/null
+++ b/python/autoCut/models/cfgs/chr_model.py
@@ -0,0 +1,140 @@
+_base_ = [
+    './chr_cascade_mask_rcnn_swin_fpn.py',
+    './coco_instance.py',
+    './schedule_1x.py', './default_runtime.py'
+]
+
+model = dict(
+    backbone=dict(
+        embed_dim=96,
+        depths=[2, 2, 6, 2],
+        num_heads=[3, 6, 12, 24],
+        window_size=7,
+        ape=False,
+        drop_path_rate=0.0,
+        patch_norm=True,
+        use_checkpoint=False
+    ),
+    neck=dict(in_channels=[96, 192, 384, 768]),
+    roi_head=dict(
+        bbox_head=[
+            dict(
+                type='ConvFCBBoxHead',
+                num_shared_convs=4,
+                num_shared_fcs=1,
+                in_channels=256,
+                conv_out_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=1,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.1, 0.1, 0.2, 0.2]),
+                reg_class_agnostic=False,
+                reg_decoded_bbox=True,
+                norm_cfg=dict(type='BN', requires_grad=True),
+                loss_cls=dict(
+                    type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0),
+                loss_bbox=dict(type='GIoULoss', loss_weight=10.0)),
+            dict(
+                type='ConvFCBBoxHead',
+                num_shared_convs=4,
+                num_shared_fcs=1,
+                in_channels=256,
+                conv_out_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=1,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.05, 0.05, 0.1, 0.1]),
+                reg_class_agnostic=False,
+                reg_decoded_bbox=True,
+                norm_cfg=dict(type='BN', requires_grad=True),
+                loss_cls=dict(
+                    type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0),
+                loss_bbox=dict(type='GIoULoss', loss_weight=10.0)),
+            dict(
+                type='ConvFCBBoxHead',
+                num_shared_convs=4,
+                num_shared_fcs=1,
+                in_channels=256,
+                conv_out_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=1,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.033, 0.033, 0.067, 0.067]),
+                reg_class_agnostic=False,
+                reg_decoded_bbox=True,
+                norm_cfg=dict(type='BN', requires_grad=True),
+                loss_cls=dict(
+                    type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0),
+                loss_bbox=dict(type='GIoULoss', loss_weight=10.0))
+        ]))
+
+img_norm_cfg = dict(
+    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
+
+# augmentation strategy originates from DETR / Sparse RCNN
+train_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='LoadAnnotations', with_bbox=True, with_mask=True),
+    dict(type='RandomFlip', flip_ratio=0.5),
+    dict(type='AutoAugment',
+         policies=[
+             [
+                 dict(type='Resize',
+                      img_scale=[(480, 1333), (512, 1333), (544, 1333), (576, 1333),
+                                 (608, 1333), (640, 1333), (672, 1333), (704, 1333),
+                                 (736, 1333), (768, 1333), (800, 1333)],
+                      multiscale_mode='value',
+                      keep_ratio=True)
+             ],
+             [
+                 dict(type='Resize',
+                      img_scale=[(400, 1333), (500, 1333), (600, 1333)],
+                      multiscale_mode='value',
+                      keep_ratio=True),
+                 dict(type='RandomCrop',
+                      crop_type='absolute_range',
+                      crop_size=(384, 600),
+                      allow_negative_crop=True),
+                 dict(type='Resize',
+                      img_scale=[(480, 1333), (512, 1333), (544, 1333),
+                                 (576, 1333), (608, 1333), (640, 1333),
+                                 (672, 1333), (704, 1333), (736, 1333),
+                                 (768, 1333), (800, 1333)],
+                      multiscale_mode='value',
+                      override=True,
+                      keep_ratio=True)
+             ]
+         ]),
+    dict(type='Normalize', **img_norm_cfg),
+    dict(type='Pad', size_divisor=32),
+    dict(type='DefaultFormatBundle'),
+    dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks']),
+]
+data = dict(train=dict(pipeline=train_pipeline))
+
+optimizer = dict(_delete_=True, type='AdamW', lr=0.0001, betas=(0.9, 0.999), weight_decay=0.05,
+                 paramwise_cfg=dict(custom_keys={'absolute_pos_embed': dict(decay_mult=0.),
+                                                 'relative_position_bias_table': dict(decay_mult=0.),
+                                                 'norm': dict(decay_mult=0.)}))
+lr_config = dict(step=[8, 11])
+runner = dict(type='EpochBasedRunnerAmp', max_epochs=200)
+
+# do not use mmdet version fp16
+fp16 = None
+optimizer_config = dict(
+    type="DistOptimizerHook",
+    update_interval=1,
+    grad_clip=None,
+    coalesce=True,
+    bucket_size_mb=-1,
+    use_fp16=True,
+)
diff --git a/python/autoCut/models/cfgs/coco_instance.py b/python/autoCut/models/cfgs/coco_instance.py
new file mode 100644
index 0000000..97d4c87
--- /dev/null
+++ b/python/autoCut/models/cfgs/coco_instance.py
@@ -0,0 +1,48 @@
+dataset_type = 'CocoDataset'
+data_root = 'data/hic_datasets/'
+img_norm_cfg = dict(
+    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
+train_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='LoadAnnotations', with_bbox=True, with_mask=True),
+    dict(type='Resize', img_scale=(1333, 800), keep_ratio=True),
+    dict(type='RandomFlip', flip_ratio=0.5),
+    dict(type='Normalize', **img_norm_cfg),
+    dict(type='Pad', size_divisor=32),
+    dict(type='DefaultFormatBundle'),
+    dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks']),
+]
+test_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(
+        type='MultiScaleFlipAug',
+        img_scale=(1333, 800),
+        flip=False,
+        transforms=[
+            dict(type='Resize', keep_ratio=True),
+            dict(type='RandomFlip'),
+            dict(type='Normalize', **img_norm_cfg),
+            dict(type='Pad', size_divisor=32),
+            dict(type='ImageToTensor', keys=['img']),
+            dict(type='Collect', keys=['img']),
+        ])
+]
+data = dict(
+    samples_per_gpu=2,
+    workers_per_gpu=2,
+    train=dict(
+        type=dataset_type,
+        ann_file=data_root + 'annotations/detection_train.json',
+        img_prefix=data_root + 'train/',
+        pipeline=train_pipeline),
+    val=dict(
+        type=dataset_type,
+        ann_file=data_root + 'annotations/detection_val.json',
+        img_prefix=data_root + 'val/',
+        pipeline=test_pipeline),
+    test=dict(
+        type=dataset_type,
+        ann_file=data_root + 'annotations/detection_val.json',
+        img_prefix=data_root + 'val/',
+        pipeline=test_pipeline))
+evaluation = dict(metric=['bbox', 'segm'])
diff --git a/python/autoCut/models/cfgs/default_runtime.py b/python/autoCut/models/cfgs/default_runtime.py
new file mode 100644
index 0000000..55097c5
--- /dev/null
+++ b/python/autoCut/models/cfgs/default_runtime.py
@@ -0,0 +1,16 @@
+checkpoint_config = dict(interval=1)
+# yapf:disable
+log_config = dict(
+    interval=50,
+    hooks=[
+        dict(type='TextLoggerHook'),
+        # dict(type='TensorboardLoggerHook')
+    ])
+# yapf:enable
+custom_hooks = [dict(type='NumClassCheckHook')]
+
+dist_params = dict(backend='nccl')
+log_level = 'INFO'
+load_from = None
+resume_from = None
+workflow = [('train', 1)]
diff --git a/python/autoCut/models/cfgs/error_cascade_mask_rcnn_swin_fpn.py b/python/autoCut/models/cfgs/error_cascade_mask_rcnn_swin_fpn.py
new file mode 100644
index 0000000..232fe5a
--- /dev/null
+++ b/python/autoCut/models/cfgs/error_cascade_mask_rcnn_swin_fpn.py
@@ -0,0 +1,207 @@
+# model settings
+model = dict(
+    type='CascadeRCNN',
+    pretrained=None,
+    backbone=dict(
+        type='SwinTransformer',
+        embed_dim=96,
+        depths=[2, 2, 6, 2],
+        num_heads=[3, 6, 12, 24],
+        window_size=7,
+        mlp_ratio=4.,
+        qkv_bias=True,
+        qk_scale=None,
+        drop_rate=0.,
+        attn_drop_rate=0.,
+        drop_path_rate=0.2,
+        ape=False,
+        patch_norm=True,
+        out_indices=(0, 1, 2, 3),
+        use_checkpoint=False),
+    neck=dict(
+        type='FPN',
+        in_channels=[96, 192, 384, 768],
+        out_channels=256,
+        num_outs=5),
+    rpn_head=dict(
+        type='RPNHead',
+        in_channels=256,
+        feat_channels=256,
+        anchor_generator=dict(
+            type='AnchorGenerator',
+            scales=[8],
+            ratios=[0.5, 1.0, 2.0],
+            strides=[4, 8, 16, 32, 64]),
+        bbox_coder=dict(
+            type='DeltaXYWHBBoxCoder',
+            target_means=[.0, .0, .0, .0],
+            target_stds=[1.0, 1.0, 1.0, 1.0]),
+        loss_cls=dict(
+            type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0)),
+    roi_head=dict(
+        type='CascadeRoIHead',
+        num_stages=3,
+        stage_loss_weights=[1, 0.5, 0.25],
+        bbox_roi_extractor=dict(
+            type='SingleRoIExtractor',
+            roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0),
+            out_channels=256,
+            featmap_strides=[4, 8, 16, 32]),
+        bbox_head=[
+            dict(
+                type='Shared2FCBBoxHead',
+                in_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=3,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.1, 0.1, 0.2, 0.2]),
+                reg_class_agnostic=True,
+                loss_cls=dict(
+                    type='CrossEntropyLoss',
+                    use_sigmoid=False,
+                    loss_weight=1.0),
+                loss_bbox=dict(type='SmoothL1Loss', beta=1.0,
+                               loss_weight=1.0)),
+            dict(
+                type='Shared2FCBBoxHead',
+                in_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=3,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.05, 0.05, 0.1, 0.1]),
+                reg_class_agnostic=True,
+                loss_cls=dict(
+                    type='CrossEntropyLoss',
+                    use_sigmoid=False,
+                    loss_weight=1.0),
+                loss_bbox=dict(type='SmoothL1Loss', beta=1.0,
+                               loss_weight=1.0)),
+            dict(
+                type='Shared2FCBBoxHead',
+                in_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=3,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.033, 0.033, 0.067, 0.067]),
+                reg_class_agnostic=True,
+                loss_cls=dict(
+                    type='CrossEntropyLoss',
+                    use_sigmoid=False,
+                    loss_weight=1.0),
+                loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0))
+        ],
+        mask_roi_extractor=dict(
+            type='SingleRoIExtractor',
+            roi_layer=dict(type='RoIAlign', output_size=14, sampling_ratio=0),
+            out_channels=256,
+            featmap_strides=[4, 8, 16, 32]),
+        mask_head=dict(
+            type='FCNMaskHead',
+            num_convs=4,
+            in_channels=256,
+            conv_out_channels=256,
+            num_classes=3,
+            loss_mask=dict(
+                type='CrossEntropyLoss', use_mask=True, loss_weight=1.0))),
+    # model training and testing settings
+    train_cfg = dict(
+        rpn=dict(
+            assigner=dict(
+                type='MaxIoUAssigner',
+                pos_iou_thr=0.7,
+                neg_iou_thr=0.3,
+                min_pos_iou=0.3,
+                match_low_quality=True,
+                ignore_iof_thr=-1),
+            sampler=dict(
+                type='RandomSampler',
+                num=256,
+                pos_fraction=0.5,
+                neg_pos_ub=-1,
+                add_gt_as_proposals=False),
+            allowed_border=0,
+            pos_weight=-1,
+            debug=False),
+        rpn_proposal=dict(
+            nms_across_levels=False,
+            nms_pre=2000,
+            nms_post=2000,
+            max_per_img=2000,
+            nms=dict(type='nms', iou_threshold=0.7),
+            min_bbox_size=0),
+        rcnn=[
+            dict(
+                assigner=dict(
+                    type='MaxIoUAssigner',
+                    pos_iou_thr=0.5,
+                    neg_iou_thr=0.5,
+                    min_pos_iou=0.5,
+                    match_low_quality=False,
+                    ignore_iof_thr=-1),
+                sampler=dict(
+                    type='RandomSampler',
+                    num=512,
+                    pos_fraction=0.25,
+                    neg_pos_ub=-1,
+                    add_gt_as_proposals=True),
+                mask_size=28,
+                pos_weight=-1,
+                debug=False),
+            dict(
+                assigner=dict(
+                    type='MaxIoUAssigner',
+                    pos_iou_thr=0.6,
+                    neg_iou_thr=0.6,
+                    min_pos_iou=0.6,
+                    match_low_quality=False,
+                    ignore_iof_thr=-1),
+                sampler=dict(
+                    type='RandomSampler',
+                    num=512,
+                    pos_fraction=0.25,
+                    neg_pos_ub=-1,
+                    add_gt_as_proposals=True),
+                mask_size=28,
+                pos_weight=-1,
+                debug=False),
+            dict(
+                assigner=dict(
+                    type='MaxIoUAssigner',
+                    pos_iou_thr=0.7,
+                    neg_iou_thr=0.7,
+                    min_pos_iou=0.7,
+                    match_low_quality=False,
+                    ignore_iof_thr=-1),
+                sampler=dict(
+                    type='RandomSampler',
+                    num=512,
+                    pos_fraction=0.25,
+                    neg_pos_ub=-1,
+                    add_gt_as_proposals=True),
+                mask_size=28,
+                pos_weight=-1,
+                debug=False)
+        ]),
+    test_cfg = dict(
+        rpn=dict(
+            nms_across_levels=False,
+            nms_pre=1000,
+            nms_post=1000,
+            max_per_img=1000,
+            nms=dict(type='nms', iou_threshold=0.7),
+            min_bbox_size=0),
+        rcnn=dict(
+            score_thr=0.05,
+            nms=dict(type='nms', iou_threshold=0.5),
+            max_per_img=100,
+            mask_thr_binary=0.5)))
diff --git a/python/autoCut/models/cfgs/error_model.py b/python/autoCut/models/cfgs/error_model.py
new file mode 100644
index 0000000..1f79a56
--- /dev/null
+++ b/python/autoCut/models/cfgs/error_model.py
@@ -0,0 +1,140 @@
+_base_ = [
+    './error_cascade_mask_rcnn_swin_fpn.py',
+    './coco_instance.py',
+    './schedule_1x.py', './default_runtime.py'
+]
+
+model = dict(
+    backbone=dict(
+        embed_dim=96,
+        depths=[2, 2, 6, 2],
+        num_heads=[3, 6, 12, 24],
+        window_size=7,
+        ape=False,
+        drop_path_rate=0.0,
+        patch_norm=True,
+        use_checkpoint=False
+    ),
+    neck=dict(in_channels=[96, 192, 384, 768]),
+    roi_head=dict(
+        bbox_head=[
+            dict(
+                type='ConvFCBBoxHead',
+                num_shared_convs=4,
+                num_shared_fcs=1,
+                in_channels=256,
+                conv_out_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=3,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.1, 0.1, 0.2, 0.2]),
+                reg_class_agnostic=False,
+                reg_decoded_bbox=True,
+                norm_cfg=dict(type='BN', requires_grad=True),
+                loss_cls=dict(
+                    type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0),
+                loss_bbox=dict(type='GIoULoss', loss_weight=10.0)),
+            dict(
+                type='ConvFCBBoxHead',
+                num_shared_convs=4,
+                num_shared_fcs=1,
+                in_channels=256,
+                conv_out_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=3,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.05, 0.05, 0.1, 0.1]),
+                reg_class_agnostic=False,
+                reg_decoded_bbox=True,
+                norm_cfg=dict(type='BN', requires_grad=True),
+                loss_cls=dict(
+                    type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0),
+                loss_bbox=dict(type='GIoULoss', loss_weight=10.0)),
+            dict(
+                type='ConvFCBBoxHead',
+                num_shared_convs=4,
+                num_shared_fcs=1,
+                in_channels=256,
+                conv_out_channels=256,
+                fc_out_channels=1024,
+                roi_feat_size=7,
+                num_classes=3,
+                bbox_coder=dict(
+                    type='DeltaXYWHBBoxCoder',
+                    target_means=[0., 0., 0., 0.],
+                    target_stds=[0.033, 0.033, 0.067, 0.067]),
+                reg_class_agnostic=False,
+                reg_decoded_bbox=True,
+                norm_cfg=dict(type='BN', requires_grad=True),
+                loss_cls=dict(
+                    type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0),
+                loss_bbox=dict(type='GIoULoss', loss_weight=10.0))
+        ]))
+
+img_norm_cfg = dict(
+    mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
+
+# augmentation strategy originates from DETR / Sparse RCNN
+train_pipeline = [
+    dict(type='LoadImageFromFile'),
+    dict(type='LoadAnnotations', with_bbox=True, with_mask=True),
+    dict(type='RandomFlip', flip_ratio=0.5),
+    dict(type='AutoAugment',
+         policies=[
+             [
+                 dict(type='Resize',
+                      img_scale=[(480, 1333), (512, 1333), (544, 1333), (576, 1333),
+                                 (608, 1333), (640, 1333), (672, 1333), (704, 1333),
+                                 (736, 1333), (768, 1333), (800, 1333)],
+                      multiscale_mode='value',
+                      keep_ratio=True)
+             ],
+             [
+                 dict(type='Resize',
+                      img_scale=[(400, 1333), (500, 1333), (600, 1333)],
+                      multiscale_mode='value',
+                      keep_ratio=True),
+                 dict(type='RandomCrop',
+                      crop_type='absolute_range',
+                      crop_size=(384, 600),
+                      allow_negative_crop=True),
+                 dict(type='Resize',
+                      img_scale=[(480, 1333), (512, 1333), (544, 1333),
+                                 (576, 1333), (608, 1333), (640, 1333),
+                                 (672, 1333), (704, 1333), (736, 1333),
+                                 (768, 1333), (800, 1333)],
+                      multiscale_mode='value',
+                      override=True,
+                      keep_ratio=True)
+             ]
+         ]),
+    dict(type='Normalize', **img_norm_cfg),
+    dict(type='Pad', size_divisor=32),
+    dict(type='DefaultFormatBundle'),
+    dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks']),
+]
+data = dict(train=dict(pipeline=train_pipeline))
+
+optimizer = dict(_delete_=True, type='AdamW', lr=0.0001, betas=(0.9, 0.999), weight_decay=0.05,
+                 paramwise_cfg=dict(custom_keys={'absolute_pos_embed': dict(decay_mult=0.),
+                                                 'relative_position_bias_table': dict(decay_mult=0.),
+                                                 'norm': dict(decay_mult=0.)}))
+lr_config = dict(step=[8, 11])
+runner = dict(type='EpochBasedRunnerAmp', max_epochs=200)
+
+# do not use mmdet version fp16
+fp16 = None
+optimizer_config = dict(
+    type="DistOptimizerHook",
+    update_interval=1,
+    grad_clip=None,
+    coalesce=True,
+    bucket_size_mb=-1,
+    use_fp16=True,
+)
diff --git a/python/autoCut/models/cfgs/schedule_1x.py b/python/autoCut/models/cfgs/schedule_1x.py
new file mode 100644
index 0000000..0da8c92
--- /dev/null
+++ b/python/autoCut/models/cfgs/schedule_1x.py
@@ -0,0 +1,11 @@
+# optimizer
+optimizer = dict(type='SGD', lr=0.02, momentum=0.9, weight_decay=0.0001)
+optimizer_config = dict(grad_clip=None)
+# learning policy
+lr_config = dict(
+    policy='step',
+    warmup='linear',
+    warmup_iters=500,
+    warmup_ratio=0.001,
+    step=[8, 11])
+runner = dict(type='EpochBasedRunner', max_epochs=100)
diff --git a/python/autoCut/models/swin/mmdet/__init__.py b/python/autoCut/models/swin/mmdet/__init__.py
new file mode 100644
index 0000000..ce2930f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/__init__.py
@@ -0,0 +1,28 @@
+import mmcv
+
+from .version import __version__, short_version
+
+
+def digit_version(version_str):
+    digit_version = []
+    for x in version_str.split('.'):
+        if x.isdigit():
+            digit_version.append(int(x))
+        elif x.find('rc') != -1:
+            patch_version = x.split('rc')
+            digit_version.append(int(patch_version[0]) - 1)
+            digit_version.append(int(patch_version[1]))
+    return digit_version
+
+
+mmcv_minimum_version = '1.2.4'
+mmcv_maximum_version = '1.4.0'
+mmcv_version = digit_version(mmcv.__version__)
+
+
+assert (mmcv_version >= digit_version(mmcv_minimum_version)
+        and mmcv_version <= digit_version(mmcv_maximum_version)), \
+    f'MMCV=={mmcv.__version__} is used but incompatible. ' \
+    f'Please install mmcv>={mmcv_minimum_version}, <={mmcv_maximum_version}.'
+
+__all__ = ['__version__', 'short_version']
diff --git a/python/autoCut/models/swin/mmdet/apis/__init__.py b/python/autoCut/models/swin/mmdet/apis/__init__.py
new file mode 100644
index 0000000..1d8035b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/apis/__init__.py
@@ -0,0 +1,10 @@
+from .inference import (async_inference_detector, inference_detector,
+                        init_detector, show_result_pyplot)
+from .test import multi_gpu_test, single_gpu_test
+from .train import get_root_logger, set_random_seed, train_detector
+
+__all__ = [
+    'get_root_logger', 'set_random_seed', 'train_detector', 'init_detector',
+    'async_inference_detector', 'inference_detector', 'show_result_pyplot',
+    'multi_gpu_test', 'single_gpu_test'
+]
diff --git a/python/autoCut/models/swin/mmdet/apis/inference.py b/python/autoCut/models/swin/mmdet/apis/inference.py
new file mode 100644
index 0000000..9acddb4
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/apis/inference.py
@@ -0,0 +1,217 @@
+import warnings
+
+import mmcv
+import numpy as np
+import torch
+from mmcv.ops import RoIPool
+from mmcv.parallel import collate, scatter
+from mmcv.runner import load_checkpoint
+
+from mmdet.core import get_classes
+from mmdet.datasets import replace_ImageToTensor
+from mmdet.datasets.pipelines import Compose
+from mmdet.models import build_detector
+
+
+def init_detector(config, checkpoint=None, device='cuda:0', cfg_options=None):
+    """Initialize a detector from config file.
+
+    Args:
+        config (str or :obj:`mmcv.Config`): Config file path or the config
+            object.
+        checkpoint (str, optional): Checkpoint path. If left as None, the model
+            will not load any weights.
+        cfg_options (dict): Options to override some settings in the used
+            config.
+
+    Returns:
+        nn.Module: The constructed detector.
+    """
+    if isinstance(config, str):
+        config = mmcv.Config.fromfile(config)
+    elif not isinstance(config, mmcv.Config):
+        raise TypeError('config must be a filename or Config object, '
+                        f'but got {type(config)}')
+    if cfg_options is not None:
+        config.merge_from_dict(cfg_options)
+    config.model.pretrained = None
+    config.model.train_cfg = None
+    model = build_detector(config.model, test_cfg=config.get('test_cfg'))
+    if checkpoint is not None:
+        map_loc = 'cpu' if (device == 'cpu' or device == torch.device("cpu")) else None
+        checkpoint = load_checkpoint(model, checkpoint, map_location=map_loc)
+        if 'CLASSES' in checkpoint.get('meta', {}):
+            model.CLASSES = checkpoint['meta']['CLASSES']
+        else:
+            warnings.simplefilter('once')
+            warnings.warn('Class names are not saved in the checkpoint\'s '
+                          'meta data, use COCO classes by default.')
+            model.CLASSES = get_classes('coco')
+    model.cfg = config  # save the config in the model for convenience
+    model.to(device)
+    model.eval()
+    return model
+
+
+class LoadImage(object):
+    """Deprecated.
+
+    A simple pipeline to load image.
+    """
+
+    def __call__(self, results):
+        """Call function to load images into results.
+
+        Args:
+            results (dict): A result dict contains the file name
+                of the image to be read.
+        Returns:
+            dict: ``results`` will be returned containing loaded image.
+        """
+        warnings.simplefilter('once')
+        warnings.warn('`LoadImage` is deprecated and will be removed in '
+                      'future releases. You may use `LoadImageFromWebcam` '
+                      'from `mmdet.datasets.pipelines.` instead.')
+        if isinstance(results['img'], str):
+            results['filename'] = results['img']
+            results['ori_filename'] = results['img']
+        else:
+            results['filename'] = None
+            results['ori_filename'] = None
+        img = mmcv.imread(results['img'])
+        results['img'] = img
+        results['img_fields'] = ['img']
+        results['img_shape'] = img.shape
+        results['ori_shape'] = img.shape
+        return results
+
+
+def inference_detector(model, imgs):
+    """Inference image(s) with the detector.
+
+    Args:
+        model (nn.Module): The loaded detector.
+        imgs (str/ndarray or list[str/ndarray] or tuple[str/ndarray]):
+           Either image files or loaded images.
+
+    Returns:
+        If imgs is a list or tuple, the same length list type results
+        will be returned, otherwise return the detection results directly.
+    """
+
+    if isinstance(imgs, (list, tuple)):
+        is_batch = True
+    else:
+        imgs = [imgs]
+        is_batch = False
+
+    cfg = model.cfg
+    device = next(model.parameters()).device  # model device
+
+    if isinstance(imgs[0], np.ndarray):
+        cfg = cfg.copy()
+        # set loading pipeline type
+        cfg.data.test.pipeline[0].type = 'LoadImageFromWebcam'
+
+    cfg.data.test.pipeline = replace_ImageToTensor(cfg.data.test.pipeline)
+    test_pipeline = Compose(cfg.data.test.pipeline)
+
+    datas = []
+    for img in imgs:
+        # prepare data
+        if isinstance(img, np.ndarray):
+            # directly add img
+            data = dict(img=img)
+        else:
+            # add information into dict
+            data = dict(img_info=dict(filename=img), img_prefix=None)
+        # build the data pipeline
+        data = test_pipeline(data)
+        datas.append(data)
+
+    data = collate(datas, samples_per_gpu=len(imgs))
+    # just get the actual data from DataContainer
+    data['img_metas'] = [img_metas.data[0] for img_metas in data['img_metas']]
+    data['img'] = [img.data[0] for img in data['img']]
+    if next(model.parameters()).is_cuda:
+        # scatter to specified GPU
+        data = scatter(data, [device])[0]
+    else:
+        for m in model.modules():
+            assert not isinstance(
+                m, RoIPool
+            ), 'CPU inference with RoIPool is not supported currently.'
+
+    # forward the model
+    with torch.no_grad():
+        results = model(return_loss=False, rescale=True, **data)
+
+    if not is_batch:
+        return results[0]
+    else:
+        return results
+
+
+async def async_inference_detector(model, img):
+    """Async inference image(s) with the detector.
+
+    Args:
+        model (nn.Module): The loaded detector.
+        img (str | ndarray): Either image files or loaded images.
+
+    Returns:
+        Awaitable detection results.
+    """
+    cfg = model.cfg
+    device = next(model.parameters()).device  # model device
+    # prepare data
+    if isinstance(img, np.ndarray):
+        # directly add img
+        data = dict(img=img)
+        cfg = cfg.copy()
+        # set loading pipeline type
+        cfg.data.test.pipeline[0].type = 'LoadImageFromWebcam'
+    else:
+        # add information into dict
+        data = dict(img_info=dict(filename=img), img_prefix=None)
+    # build the data pipeline
+    test_pipeline = Compose(cfg.data.test.pipeline)
+    data = test_pipeline(data)
+    data = scatter(collate([data], samples_per_gpu=1), [device])[0]
+
+    # We don't restore `torch.is_grad_enabled()` value during concurrent
+    # inference since execution can overlap
+    torch.set_grad_enabled(False)
+    result = await model.aforward_test(rescale=True, **data)
+    return result
+
+
+def show_result_pyplot(model,
+                       img,
+                       result,
+                       score_thr=0.3,
+                       title='result',
+                       wait_time=0):
+    """Visualize the detection results on the image.
+
+    Args:
+        model (nn.Module): The loaded detector.
+        img (str or np.ndarray): Image filename or loaded image.
+        result (tuple[list] or list): The detection result, can be either
+            (bbox, segm) or just bbox.
+        score_thr (float): The threshold to visualize the bboxes and masks.
+        title (str): Title of the pyplot figure.
+        wait_time (float): Value of waitKey param.
+                Default: 0.
+    """
+    if hasattr(model, 'module'):
+        model = model.module
+    model.show_result(
+        img,
+        result,
+        score_thr=score_thr,
+        show=True,
+        wait_time=wait_time,
+        win_name=title,
+        bbox_color=(72, 101, 241),
+        text_color=(72, 101, 241))
diff --git a/python/autoCut/models/swin/mmdet/apis/test.py b/python/autoCut/models/swin/mmdet/apis/test.py
new file mode 100644
index 0000000..e54b1b8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/apis/test.py
@@ -0,0 +1,190 @@
+import os.path as osp
+import pickle
+import shutil
+import tempfile
+import time
+
+import mmcv
+import torch
+import torch.distributed as dist
+from mmcv.image import tensor2imgs
+from mmcv.runner import get_dist_info
+
+from mmdet.core import encode_mask_results
+
+
+def single_gpu_test(model,
+                    data_loader,
+                    show=False,
+                    out_dir=None,
+                    show_score_thr=0.3):
+    model.eval()
+    results = []
+    dataset = data_loader.dataset
+    prog_bar = mmcv.ProgressBar(len(dataset))
+    for i, data in enumerate(data_loader):
+        with torch.no_grad():
+            result = model(return_loss=False, rescale=True, **data)
+
+        batch_size = len(result)
+        if show or out_dir:
+            if batch_size == 1 and isinstance(data['img'][0], torch.Tensor):
+                img_tensor = data['img'][0]
+            else:
+                img_tensor = data['img'][0].data[0]
+            img_metas = data['img_metas'][0].data[0]
+            imgs = tensor2imgs(img_tensor, **img_metas[0]['img_norm_cfg'])
+            assert len(imgs) == len(img_metas)
+
+            for i, (img, img_meta) in enumerate(zip(imgs, img_metas)):
+                h, w, _ = img_meta['img_shape']
+                img_show = img[:h, :w, :]
+
+                ori_h, ori_w = img_meta['ori_shape'][:-1]
+                img_show = mmcv.imresize(img_show, (ori_w, ori_h))
+
+                if out_dir:
+                    out_file = osp.join(out_dir, img_meta['ori_filename'])
+                else:
+                    out_file = None
+
+                model.module.show_result(
+                    img_show,
+                    result[i],
+                    show=show,
+                    out_file=out_file,
+                    score_thr=show_score_thr)
+
+        # encode mask results
+        if isinstance(result[0], tuple):
+            result = [(bbox_results, encode_mask_results(mask_results))
+                      for bbox_results, mask_results in result]
+        results.extend(result)
+
+        for _ in range(batch_size):
+            prog_bar.update()
+    return results
+
+
+def multi_gpu_test(model, data_loader, tmpdir=None, gpu_collect=False):
+    """Test model with multiple gpus.
+
+    This method tests model with multiple gpus and collects the results
+    under two different modes: gpu and cpu modes. By setting 'gpu_collect=True'
+    it encodes results to gpu tensors and use gpu communication for results
+    collection. On cpu mode it saves the results on different gpus to 'tmpdir'
+    and collects them by the rank 0 worker.
+
+    Args:
+        model (nn.Module): Model to be tested.
+        data_loader (nn.Dataloader): Pytorch data loader.
+        tmpdir (str): Path of directory to save the temporary results from
+            different gpus under cpu mode.
+        gpu_collect (bool): Option to use either gpu or cpu to collect results.
+
+    Returns:
+        list: The prediction results.
+    """
+    model.eval()
+    results = []
+    dataset = data_loader.dataset
+    rank, world_size = get_dist_info()
+    if rank == 0:
+        prog_bar = mmcv.ProgressBar(len(dataset))
+    time.sleep(2)  # This line can prevent deadlock problem in some cases.
+    for i, data in enumerate(data_loader):
+        with torch.no_grad():
+            result = model(return_loss=False, rescale=True, **data)
+            # encode mask results
+            if isinstance(result[0], tuple):
+                result = [(bbox_results, encode_mask_results(mask_results))
+                          for bbox_results, mask_results in result]
+        results.extend(result)
+
+        if rank == 0:
+            batch_size = len(result)
+            for _ in range(batch_size * world_size):
+                prog_bar.update()
+
+    # collect results from all ranks
+    if gpu_collect:
+        results = collect_results_gpu(results, len(dataset))
+    else:
+        results = collect_results_cpu(results, len(dataset), tmpdir)
+    return results
+
+
+def collect_results_cpu(result_part, size, tmpdir=None):
+    rank, world_size = get_dist_info()
+    # create a tmp dir if it is not specified
+    if tmpdir is None:
+        MAX_LEN = 512
+        # 32 is whitespace
+        dir_tensor = torch.full((MAX_LEN, ),
+                                32,
+                                dtype=torch.uint8,
+                                device='cuda')
+        if rank == 0:
+            mmcv.mkdir_or_exist('.dist_test')
+            tmpdir = tempfile.mkdtemp(dir='.dist_test')
+            tmpdir = torch.tensor(
+                bytearray(tmpdir.encode()), dtype=torch.uint8, device='cuda')
+            dir_tensor[:len(tmpdir)] = tmpdir
+        dist.broadcast(dir_tensor, 0)
+        tmpdir = dir_tensor.cpu().numpy().tobytes().decode().rstrip()
+    else:
+        mmcv.mkdir_or_exist(tmpdir)
+    # dump the part result to the dir
+    mmcv.dump(result_part, osp.join(tmpdir, f'part_{rank}.pkl'))
+    dist.barrier()
+    # collect all parts
+    if rank != 0:
+        return None
+    else:
+        # load results of all parts from tmp dir
+        part_list = []
+        for i in range(world_size):
+            part_file = osp.join(tmpdir, f'part_{i}.pkl')
+            part_list.append(mmcv.load(part_file))
+        # sort the results
+        ordered_results = []
+        for res in zip(*part_list):
+            ordered_results.extend(list(res))
+        # the dataloader may pad some samples
+        ordered_results = ordered_results[:size]
+        # remove tmp dir
+        shutil.rmtree(tmpdir)
+        return ordered_results
+
+
+def collect_results_gpu(result_part, size):
+    rank, world_size = get_dist_info()
+    # dump result part to tensor with pickle
+    part_tensor = torch.tensor(
+        bytearray(pickle.dumps(result_part)), dtype=torch.uint8, device='cuda')
+    # gather all result part tensor shape
+    shape_tensor = torch.tensor(part_tensor.shape, device='cuda')
+    shape_list = [shape_tensor.clone() for _ in range(world_size)]
+    dist.all_gather(shape_list, shape_tensor)
+    # padding result part tensor to max length
+    shape_max = torch.tensor(shape_list).max()
+    part_send = torch.zeros(shape_max, dtype=torch.uint8, device='cuda')
+    part_send[:shape_tensor[0]] = part_tensor
+    part_recv_list = [
+        part_tensor.new_zeros(shape_max) for _ in range(world_size)
+    ]
+    # gather all result part
+    dist.all_gather(part_recv_list, part_send)
+
+    if rank == 0:
+        part_list = []
+        for recv, shape in zip(part_recv_list, shape_list):
+            part_list.append(
+                pickle.loads(recv[:shape[0]].cpu().numpy().tobytes()))
+        # sort the results
+        ordered_results = []
+        for res in zip(*part_list):
+            ordered_results.extend(list(res))
+        # the dataloader may pad some samples
+        ordered_results = ordered_results[:size]
+        return ordered_results
diff --git a/python/autoCut/models/swin/mmdet/apis/train.py b/python/autoCut/models/swin/mmdet/apis/train.py
new file mode 100644
index 0000000..a41165c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/apis/train.py
@@ -0,0 +1,185 @@
+import random
+import warnings
+
+import numpy as np
+import torch
+from mmcv.parallel import MMDataParallel, MMDistributedDataParallel
+from mmcv.runner import (HOOKS, DistSamplerSeedHook, EpochBasedRunner,
+                         Fp16OptimizerHook, OptimizerHook, build_optimizer,
+                         build_runner)
+from mmcv.utils import build_from_cfg
+
+from mmdet.core import DistEvalHook, EvalHook
+from mmdet.datasets import (build_dataloader, build_dataset,
+                            replace_ImageToTensor)
+from mmdet.utils import get_root_logger
+from mmcv_custom.runner import EpochBasedRunnerAmp
+try:
+    import apex
+except:
+    pass
+
+
+def set_random_seed(seed, deterministic=False):
+    """Set random seed.
+
+    Args:
+        seed (int): Seed to be used.
+        deterministic (bool): Whether to set the deterministic option for
+            CUDNN backend, i.e., set `torch.backends.cudnn.deterministic`
+            to True and `torch.backends.cudnn.benchmark` to False.
+            Default: False.
+    """
+    random.seed(seed)
+    np.random.seed(seed)
+    torch.manual_seed(seed)
+    torch.cuda.manual_seed_all(seed)
+    if deterministic:
+        torch.backends.cudnn.deterministic = True
+        torch.backends.cudnn.benchmark = False
+
+
+def train_detector(model,
+                   dataset,
+                   cfg,
+                   distributed=False,
+                   validate=False,
+                   timestamp=None,
+                   meta=None):
+    logger = get_root_logger(cfg.log_level)
+
+    # prepare data loaders
+    dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset]
+    if 'imgs_per_gpu' in cfg.data:
+        logger.warning('"imgs_per_gpu" is deprecated in MMDet V2.0. '
+                       'Please use "samples_per_gpu" instead')
+        if 'samples_per_gpu' in cfg.data:
+            logger.warning(
+                f'Got "imgs_per_gpu"={cfg.data.imgs_per_gpu} and '
+                f'"samples_per_gpu"={cfg.data.samples_per_gpu}, "imgs_per_gpu"'
+                f'={cfg.data.imgs_per_gpu} is used in this experiments')
+        else:
+            logger.warning(
+                'Automatically set "samples_per_gpu"="imgs_per_gpu"='
+                f'{cfg.data.imgs_per_gpu} in this experiments')
+        cfg.data.samples_per_gpu = cfg.data.imgs_per_gpu
+
+    data_loaders = [
+        build_dataloader(
+            ds,
+            cfg.data.samples_per_gpu,
+            cfg.data.workers_per_gpu,
+            # cfg.gpus will be ignored if distributed
+            len(cfg.gpu_ids),
+            dist=distributed,
+            seed=cfg.seed) for ds in dataset
+    ]
+
+    # build optimizer
+    optimizer = build_optimizer(model, cfg.optimizer)
+
+    # use apex fp16 optimizer
+    if cfg.optimizer_config.get("type", None) and cfg.optimizer_config["type"] == "DistOptimizerHook":
+        if cfg.optimizer_config.get("use_fp16", False):
+            model, optimizer = apex.amp.initialize(
+                model.cuda(), optimizer, opt_level="O1")
+            for m in model.modules():
+                if hasattr(m, "fp16_enabled"):
+                    m.fp16_enabled = True
+
+    # put model on gpus
+    if distributed:
+        find_unused_parameters = cfg.get('find_unused_parameters', False)
+        # Sets the `find_unused_parameters` parameter in
+        # torch.nn.parallel.DistributedDataParallel
+        model = MMDistributedDataParallel(
+            model.cuda(),
+            device_ids=[torch.cuda.current_device()],
+            broadcast_buffers=False,
+            find_unused_parameters=find_unused_parameters)
+    else:
+        model = MMDataParallel(
+            model.cuda(cfg.gpu_ids[0]), device_ids=cfg.gpu_ids)
+
+    if 'runner' not in cfg:
+        cfg.runner = {
+            'type': 'EpochBasedRunner',
+            'max_epochs': cfg.total_epochs
+        }
+        warnings.warn(
+            'config is now expected to have a `runner` section, '
+            'please set `runner` in your config.', UserWarning)
+    else:
+        if 'total_epochs' in cfg:
+            assert cfg.total_epochs == cfg.runner.max_epochs
+
+    # build runner
+    runner = build_runner(
+        cfg.runner,
+        default_args=dict(
+            model=model,
+            optimizer=optimizer,
+            work_dir=cfg.work_dir,
+            logger=logger,
+            meta=meta))
+
+    # an ugly workaround to make .log and .log.json filenames the same
+    runner.timestamp = timestamp
+
+    # fp16 setting
+    fp16_cfg = cfg.get('fp16', None)
+    if fp16_cfg is not None:
+        optimizer_config = Fp16OptimizerHook(
+            **cfg.optimizer_config, **fp16_cfg, distributed=distributed)
+    elif distributed and 'type' not in cfg.optimizer_config:
+        optimizer_config = OptimizerHook(**cfg.optimizer_config)
+    else:
+        optimizer_config = cfg.optimizer_config
+
+    # register hooks
+    runner.register_training_hooks(cfg.lr_config, optimizer_config,
+                                   cfg.checkpoint_config, cfg.log_config,
+                                   cfg.get('momentum_config', None))
+    if distributed:
+        if isinstance(runner, EpochBasedRunner):
+            runner.register_hook(DistSamplerSeedHook())
+
+    # register eval hooks
+    if validate:
+        # Support batch_size > 1 in validation
+        val_samples_per_gpu = cfg.data.val.pop('samples_per_gpu', 1)
+        if val_samples_per_gpu > 1:
+            # Replace 'ImageToTensor' to 'DefaultFormatBundle'
+            cfg.data.val.pipeline = replace_ImageToTensor(
+                cfg.data.val.pipeline)
+        val_dataset = build_dataset(cfg.data.val, dict(test_mode=True))
+        val_dataloader = build_dataloader(
+            val_dataset,
+            samples_per_gpu=val_samples_per_gpu,
+            workers_per_gpu=cfg.data.workers_per_gpu,
+            dist=distributed,
+            shuffle=False)
+        eval_cfg = cfg.get('evaluation', {})
+        eval_cfg['by_epoch'] = cfg.runner['type'] != 'IterBasedRunner'
+        eval_hook = DistEvalHook if distributed else EvalHook
+        runner.register_hook(eval_hook(val_dataloader, **eval_cfg))
+
+    # user-defined hooks
+    if cfg.get('custom_hooks', None):
+        custom_hooks = cfg.custom_hooks
+        assert isinstance(custom_hooks, list), \
+            f'custom_hooks expect list type, but got {type(custom_hooks)}'
+        for hook_cfg in cfg.custom_hooks:
+            assert isinstance(hook_cfg, dict), \
+                'Each item in custom_hooks expects dict type, but got ' \
+                f'{type(hook_cfg)}'
+            hook_cfg = hook_cfg.copy()
+            priority = hook_cfg.pop('priority', 'NORMAL')
+            hook = build_from_cfg(hook_cfg, HOOKS)
+            runner.register_hook(hook, priority=priority)
+
+    if cfg.resume_from:
+        runner.resume(cfg.resume_from)
+    elif cfg.load_from:
+        runner.load_checkpoint(cfg.load_from)
+    runner.run(data_loaders, cfg.workflow)
diff --git a/python/autoCut/models/swin/mmdet/core/__init__.py b/python/autoCut/models/swin/mmdet/core/__init__.py
new file mode 100644
index 0000000..e812391
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/__init__.py
@@ -0,0 +1,7 @@
+from .anchor import *  # noqa: F401, F403
+from .bbox import *  # noqa: F401, F403
+from .evaluation import *  # noqa: F401, F403
+from .export import *  # noqa: F401, F403
+from .mask import *  # noqa: F401, F403
+from .post_processing import *  # noqa: F401, F403
+from .utils import *  # noqa: F401, F403
diff --git a/python/autoCut/models/swin/mmdet/core/anchor/__init__.py b/python/autoCut/models/swin/mmdet/core/anchor/__init__.py
new file mode 100644
index 0000000..5838ff3
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/anchor/__init__.py
@@ -0,0 +1,11 @@
+from .anchor_generator import (AnchorGenerator, LegacyAnchorGenerator,
+                               YOLOAnchorGenerator)
+from .builder import ANCHOR_GENERATORS, build_anchor_generator
+from .point_generator import PointGenerator
+from .utils import anchor_inside_flags, calc_region, images_to_levels
+
+__all__ = [
+    'AnchorGenerator', 'LegacyAnchorGenerator', 'anchor_inside_flags',
+    'PointGenerator', 'images_to_levels', 'calc_region',
+    'build_anchor_generator', 'ANCHOR_GENERATORS', 'YOLOAnchorGenerator'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/anchor/anchor_generator.py b/python/autoCut/models/swin/mmdet/core/anchor/anchor_generator.py
new file mode 100644
index 0000000..388d260
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/anchor/anchor_generator.py
@@ -0,0 +1,727 @@
+import mmcv
+import numpy as np
+import torch
+from torch.nn.modules.utils import _pair
+
+from .builder import ANCHOR_GENERATORS
+
+
+@ANCHOR_GENERATORS.register_module()
+class AnchorGenerator(object):
+    """Standard anchor generator for 2D anchor-based detectors.
+
+    Args:
+        strides (list[int] | list[tuple[int, int]]): Strides of anchors
+            in multiple feature levels in order (w, h).
+        ratios (list[float]): The list of ratios between the height and width
+            of anchors in a single level.
+        scales (list[int] | None): Anchor scales for anchors in a single level.
+            It cannot be set at the same time if `octave_base_scale` and
+            `scales_per_octave` are set.
+        base_sizes (list[int] | None): The basic sizes
+            of anchors in multiple levels.
+            If None is given, strides will be used as base_sizes.
+            (If strides are non square, the shortest stride is taken.)
+        scale_major (bool): Whether to multiply scales first when generating
+            base anchors. If true, the anchors in the same row will have the
+            same scales. By default it is True in V2.0
+        octave_base_scale (int): The base scale of octave.
+        scales_per_octave (int): Number of scales for each octave.
+            `octave_base_scale` and `scales_per_octave` are usually used in
+            retinanet and the `scales` should be None when they are set.
+        centers (list[tuple[float, float]] | None): The centers of the anchor
+            relative to the feature grid center in multiple feature levels.
+            By default it is set to be None and not used. If a list of tuple of
+            float is given, they will be used to shift the centers of anchors.
+        center_offset (float): The offset of center in proportion to anchors'
+            width and height. By default it is 0 in V2.0.
+
+    Examples:
+        >>> from mmdet.core import AnchorGenerator
+        >>> self = AnchorGenerator([16], [1.], [1.], [9])
+        >>> all_anchors = self.grid_anchors([(2, 2)], device='cpu')
+        >>> print(all_anchors)
+        [tensor([[-4.5000, -4.5000,  4.5000,  4.5000],
+                [11.5000, -4.5000, 20.5000,  4.5000],
+                [-4.5000, 11.5000,  4.5000, 20.5000],
+                [11.5000, 11.5000, 20.5000, 20.5000]])]
+        >>> self = AnchorGenerator([16, 32], [1.], [1.], [9, 18])
+        >>> all_anchors = self.grid_anchors([(2, 2), (1, 1)], device='cpu')
+        >>> print(all_anchors)
+        [tensor([[-4.5000, -4.5000,  4.5000,  4.5000],
+                [11.5000, -4.5000, 20.5000,  4.5000],
+                [-4.5000, 11.5000,  4.5000, 20.5000],
+                [11.5000, 11.5000, 20.5000, 20.5000]]), \
+        tensor([[-9., -9., 9., 9.]])]
+    """
+
+    def __init__(self,
+                 strides,
+                 ratios,
+                 scales=None,
+                 base_sizes=None,
+                 scale_major=True,
+                 octave_base_scale=None,
+                 scales_per_octave=None,
+                 centers=None,
+                 center_offset=0.):
+        # check center and center_offset
+        if center_offset != 0:
+            assert centers is None, 'center cannot be set when center_offset' \
+                f'!=0, {centers} is given.'
+        if not (0 <= center_offset <= 1):
+            raise ValueError('center_offset should be in range [0, 1], '
+                             f'{center_offset} is given.')
+        if centers is not None:
+            assert len(centers) == len(strides), \
+                'The number of strides should be the same as centers, got ' \
+                f'{strides} and {centers}'
+
+        # calculate base sizes of anchors
+        self.strides = [_pair(stride) for stride in strides]
+        self.base_sizes = [min(stride) for stride in self.strides
+                           ] if base_sizes is None else base_sizes
+        assert len(self.base_sizes) == len(self.strides), \
+            'The number of strides should be the same as base sizes, got ' \
+            f'{self.strides} and {self.base_sizes}'
+
+        # calculate scales of anchors
+        assert ((octave_base_scale is not None
+                and scales_per_octave is not None) ^ (scales is not None)), \
+            'scales and octave_base_scale with scales_per_octave cannot' \
+            ' be set at the same time'
+        if scales is not None:
+            self.scales = torch.Tensor(scales)
+        elif octave_base_scale is not None and scales_per_octave is not None:
+            octave_scales = np.array(
+                [2**(i / scales_per_octave) for i in range(scales_per_octave)])
+            scales = octave_scales * octave_base_scale
+            self.scales = torch.Tensor(scales)
+        else:
+            raise ValueError('Either scales or octave_base_scale with '
+                             'scales_per_octave should be set')
+
+        self.octave_base_scale = octave_base_scale
+        self.scales_per_octave = scales_per_octave
+        self.ratios = torch.Tensor(ratios)
+        self.scale_major = scale_major
+        self.centers = centers
+        self.center_offset = center_offset
+        self.base_anchors = self.gen_base_anchors()
+
+    @property
+    def num_base_anchors(self):
+        """list[int]: total number of base anchors in a feature grid"""
+        return [base_anchors.size(0) for base_anchors in self.base_anchors]
+
+    @property
+    def num_levels(self):
+        """int: number of feature levels that the generator will be applied"""
+        return len(self.strides)
+
+    def gen_base_anchors(self):
+        """Generate base anchors.
+
+        Returns:
+            list(torch.Tensor): Base anchors of a feature grid in multiple \
+                feature levels.
+        """
+        multi_level_base_anchors = []
+        for i, base_size in enumerate(self.base_sizes):
+            center = None
+            if self.centers is not None:
+                center = self.centers[i]
+            multi_level_base_anchors.append(
+                self.gen_single_level_base_anchors(
+                    base_size,
+                    scales=self.scales,
+                    ratios=self.ratios,
+                    center=center))
+        return multi_level_base_anchors
+
+    def gen_single_level_base_anchors(self,
+                                      base_size,
+                                      scales,
+                                      ratios,
+                                      center=None):
+        """Generate base anchors of a single level.
+
+        Args:
+            base_size (int | float): Basic size of an anchor.
+            scales (torch.Tensor): Scales of the anchor.
+            ratios (torch.Tensor): The ratio between between the height
+                and width of anchors in a single level.
+            center (tuple[float], optional): The center of the base anchor
+                related to a single feature grid. Defaults to None.
+
+        Returns:
+            torch.Tensor: Anchors in a single-level feature maps.
+        """
+        w = base_size
+        h = base_size
+        if center is None:
+            x_center = self.center_offset * w
+            y_center = self.center_offset * h
+        else:
+            x_center, y_center = center
+
+        h_ratios = torch.sqrt(ratios)
+        w_ratios = 1 / h_ratios
+        if self.scale_major:
+            ws = (w * w_ratios[:, None] * scales[None, :]).view(-1)
+            hs = (h * h_ratios[:, None] * scales[None, :]).view(-1)
+        else:
+            ws = (w * scales[:, None] * w_ratios[None, :]).view(-1)
+            hs = (h * scales[:, None] * h_ratios[None, :]).view(-1)
+
+        # use float anchor and the anchor's center is aligned with the
+        # pixel center
+        base_anchors = [
+            x_center - 0.5 * ws, y_center - 0.5 * hs, x_center + 0.5 * ws,
+            y_center + 0.5 * hs
+        ]
+        base_anchors = torch.stack(base_anchors, dim=-1)
+
+        return base_anchors
+
+    def _meshgrid(self, x, y, row_major=True):
+        """Generate mesh grid of x and y.
+
+        Args:
+            x (torch.Tensor): Grids of x dimension.
+            y (torch.Tensor): Grids of y dimension.
+            row_major (bool, optional): Whether to return y grids first.
+                Defaults to True.
+
+        Returns:
+            tuple[torch.Tensor]: The mesh grids of x and y.
+        """
+        # use shape instead of len to keep tracing while exporting to onnx
+        xx = x.repeat(y.shape[0])
+        yy = y.view(-1, 1).repeat(1, x.shape[0]).view(-1)
+        if row_major:
+            return xx, yy
+        else:
+            return yy, xx
+
+    def grid_anchors(self, featmap_sizes, device='cuda'):
+        """Generate grid anchors in multiple feature levels.
+
+        Args:
+            featmap_sizes (list[tuple]): List of feature map sizes in
+                multiple feature levels.
+            device (str): Device where the anchors will be put on.
+
+        Return:
+            list[torch.Tensor]: Anchors in multiple feature levels. \
+                The sizes of each tensor should be [N, 4], where \
+                N = width * height * num_base_anchors, width and height \
+                are the sizes of the corresponding feature level, \
+                num_base_anchors is the number of anchors for that level.
+        """
+        assert self.num_levels == len(featmap_sizes)
+        multi_level_anchors = []
+        for i in range(self.num_levels):
+            anchors = self.single_level_grid_anchors(
+                self.base_anchors[i].to(device),
+                featmap_sizes[i],
+                self.strides[i],
+                device=device)
+            multi_level_anchors.append(anchors)
+        return multi_level_anchors
+
+    def single_level_grid_anchors(self,
+                                  base_anchors,
+                                  featmap_size,
+                                  stride=(16, 16),
+                                  device='cuda'):
+        """Generate grid anchors of a single level.
+
+        Note:
+            This function is usually called by method ``self.grid_anchors``.
+
+        Args:
+            base_anchors (torch.Tensor): The base anchors of a feature grid.
+            featmap_size (tuple[int]): Size of the feature maps.
+            stride (tuple[int], optional): Stride of the feature map in order
+                (w, h). Defaults to (16, 16).
+            device (str, optional): Device the tensor will be put on.
+                Defaults to 'cuda'.
+
+        Returns:
+            torch.Tensor: Anchors in the overall feature maps.
+        """
+        # keep as Tensor, so that we can covert to ONNX correctly
+        feat_h, feat_w = featmap_size
+        shift_x = torch.arange(0, feat_w, device=device) * stride[0]
+        shift_y = torch.arange(0, feat_h, device=device) * stride[1]
+
+        shift_xx, shift_yy = self._meshgrid(shift_x, shift_y)
+        shifts = torch.stack([shift_xx, shift_yy, shift_xx, shift_yy], dim=-1)
+        shifts = shifts.type_as(base_anchors)
+        # first feat_w elements correspond to the first row of shifts
+        # add A anchors (1, A, 4) to K shifts (K, 1, 4) to get
+        # shifted anchors (K, A, 4), reshape to (K*A, 4)
+
+        all_anchors = base_anchors[None, :, :] + shifts[:, None, :]
+        all_anchors = all_anchors.view(-1, 4)
+        # first A rows correspond to A anchors of (0, 0) in feature map,
+        # then (0, 1), (0, 2), ...
+        return all_anchors
+
+    def valid_flags(self, featmap_sizes, pad_shape, device='cuda'):
+        """Generate valid flags of anchors in multiple feature levels.
+
+        Args:
+            featmap_sizes (list(tuple)): List of feature map sizes in
+                multiple feature levels.
+            pad_shape (tuple): The padded shape of the image.
+            device (str): Device where the anchors will be put on.
+
+        Return:
+            list(torch.Tensor): Valid flags of anchors in multiple levels.
+        """
+        assert self.num_levels == len(featmap_sizes)
+        multi_level_flags = []
+        for i in range(self.num_levels):
+            anchor_stride = self.strides[i]
+            feat_h, feat_w = featmap_sizes[i]
+            h, w = pad_shape[:2]
+            valid_feat_h = min(int(np.ceil(h / anchor_stride[1])), feat_h)
+            valid_feat_w = min(int(np.ceil(w / anchor_stride[0])), feat_w)
+            flags = self.single_level_valid_flags((feat_h, feat_w),
+                                                  (valid_feat_h, valid_feat_w),
+                                                  self.num_base_anchors[i],
+                                                  device=device)
+            multi_level_flags.append(flags)
+        return multi_level_flags
+
+    def single_level_valid_flags(self,
+                                 featmap_size,
+                                 valid_size,
+                                 num_base_anchors,
+                                 device='cuda'):
+        """Generate the valid flags of anchor in a single feature map.
+
+        Args:
+            featmap_size (tuple[int]): The size of feature maps.
+            valid_size (tuple[int]): The valid size of the feature maps.
+            num_base_anchors (int): The number of base anchors.
+            device (str, optional): Device where the flags will be put on.
+                Defaults to 'cuda'.
+
+        Returns:
+            torch.Tensor: The valid flags of each anchor in a single level \
+                feature map.
+        """
+        feat_h, feat_w = featmap_size
+        valid_h, valid_w = valid_size
+        assert valid_h <= feat_h and valid_w <= feat_w
+        valid_x = torch.zeros(feat_w, dtype=torch.bool, device=device)
+        valid_y = torch.zeros(feat_h, dtype=torch.bool, device=device)
+        valid_x[:valid_w] = 1
+        valid_y[:valid_h] = 1
+        valid_xx, valid_yy = self._meshgrid(valid_x, valid_y)
+        valid = valid_xx & valid_yy
+        valid = valid[:, None].expand(valid.size(0),
+                                      num_base_anchors).contiguous().view(-1)
+        return valid
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        indent_str = '    '
+        repr_str = self.__class__.__name__ + '(\n'
+        repr_str += f'{indent_str}strides={self.strides},\n'
+        repr_str += f'{indent_str}ratios={self.ratios},\n'
+        repr_str += f'{indent_str}scales={self.scales},\n'
+        repr_str += f'{indent_str}base_sizes={self.base_sizes},\n'
+        repr_str += f'{indent_str}scale_major={self.scale_major},\n'
+        repr_str += f'{indent_str}octave_base_scale='
+        repr_str += f'{self.octave_base_scale},\n'
+        repr_str += f'{indent_str}scales_per_octave='
+        repr_str += f'{self.scales_per_octave},\n'
+        repr_str += f'{indent_str}num_levels={self.num_levels}\n'
+        repr_str += f'{indent_str}centers={self.centers},\n'
+        repr_str += f'{indent_str}center_offset={self.center_offset})'
+        return repr_str
+
+
+@ANCHOR_GENERATORS.register_module()
+class SSDAnchorGenerator(AnchorGenerator):
+    """Anchor generator for SSD.
+
+    Args:
+        strides (list[int]  | list[tuple[int, int]]): Strides of anchors
+            in multiple feature levels.
+        ratios (list[float]): The list of ratios between the height and width
+            of anchors in a single level.
+        basesize_ratio_range (tuple(float)): Ratio range of anchors.
+        input_size (int): Size of feature map, 300 for SSD300,
+            512 for SSD512.
+        scale_major (bool): Whether to multiply scales first when generating
+            base anchors. If true, the anchors in the same row will have the
+            same scales. It is always set to be False in SSD.
+    """
+
+    def __init__(self,
+                 strides,
+                 ratios,
+                 basesize_ratio_range,
+                 input_size=300,
+                 scale_major=True):
+        assert len(strides) == len(ratios)
+        assert mmcv.is_tuple_of(basesize_ratio_range, float)
+
+        self.strides = [_pair(stride) for stride in strides]
+        self.input_size = input_size
+        self.centers = [(stride[0] / 2., stride[1] / 2.)
+                        for stride in self.strides]
+        self.basesize_ratio_range = basesize_ratio_range
+
+        # calculate anchor ratios and sizes
+        min_ratio, max_ratio = basesize_ratio_range
+        min_ratio = int(min_ratio * 100)
+        max_ratio = int(max_ratio * 100)
+        step = int(np.floor(max_ratio - min_ratio) / (self.num_levels - 2))
+        min_sizes = []
+        max_sizes = []
+        for ratio in range(int(min_ratio), int(max_ratio) + 1, step):
+            min_sizes.append(int(self.input_size * ratio / 100))
+            max_sizes.append(int(self.input_size * (ratio + step) / 100))
+        if self.input_size == 300:
+            if basesize_ratio_range[0] == 0.15:  # SSD300 COCO
+                min_sizes.insert(0, int(self.input_size * 7 / 100))
+                max_sizes.insert(0, int(self.input_size * 15 / 100))
+            elif basesize_ratio_range[0] == 0.2:  # SSD300 VOC
+                min_sizes.insert(0, int(self.input_size * 10 / 100))
+                max_sizes.insert(0, int(self.input_size * 20 / 100))
+            else:
+                raise ValueError(
+                    'basesize_ratio_range[0] should be either 0.15'
+                    'or 0.2 when input_size is 300, got '
+                    f'{basesize_ratio_range[0]}.')
+        elif self.input_size == 512:
+            if basesize_ratio_range[0] == 0.1:  # SSD512 COCO
+                min_sizes.insert(0, int(self.input_size * 4 / 100))
+                max_sizes.insert(0, int(self.input_size * 10 / 100))
+            elif basesize_ratio_range[0] == 0.15:  # SSD512 VOC
+                min_sizes.insert(0, int(self.input_size * 7 / 100))
+                max_sizes.insert(0, int(self.input_size * 15 / 100))
+            else:
+                raise ValueError('basesize_ratio_range[0] should be either 0.1'
+                                 'or 0.15 when input_size is 512, got'
+                                 f' {basesize_ratio_range[0]}.')
+        else:
+            raise ValueError('Only support 300 or 512 in SSDAnchorGenerator'
+                             f', got {self.input_size}.')
+
+        anchor_ratios = []
+        anchor_scales = []
+        for k in range(len(self.strides)):
+            scales = [1., np.sqrt(max_sizes[k] / min_sizes[k])]
+            anchor_ratio = [1.]
+            for r in ratios[k]:
+                anchor_ratio += [1 / r, r]  # 4 or 6 ratio
+            anchor_ratios.append(torch.Tensor(anchor_ratio))
+            anchor_scales.append(torch.Tensor(scales))
+
+        self.base_sizes = min_sizes
+        self.scales = anchor_scales
+        self.ratios = anchor_ratios
+        self.scale_major = scale_major
+        self.center_offset = 0
+        self.base_anchors = self.gen_base_anchors()
+
+    def gen_base_anchors(self):
+        """Generate base anchors.
+
+        Returns:
+            list(torch.Tensor): Base anchors of a feature grid in multiple \
+                feature levels.
+        """
+        multi_level_base_anchors = []
+        for i, base_size in enumerate(self.base_sizes):
+            base_anchors = self.gen_single_level_base_anchors(
+                base_size,
+                scales=self.scales[i],
+                ratios=self.ratios[i],
+                center=self.centers[i])
+            indices = list(range(len(self.ratios[i])))
+            indices.insert(1, len(indices))
+            base_anchors = torch.index_select(base_anchors, 0,
+                                              torch.LongTensor(indices))
+            multi_level_base_anchors.append(base_anchors)
+        return multi_level_base_anchors
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        indent_str = '    '
+        repr_str = self.__class__.__name__ + '(\n'
+        repr_str += f'{indent_str}strides={self.strides},\n'
+        repr_str += f'{indent_str}scales={self.scales},\n'
+        repr_str += f'{indent_str}scale_major={self.scale_major},\n'
+        repr_str += f'{indent_str}input_size={self.input_size},\n'
+        repr_str += f'{indent_str}scales={self.scales},\n'
+        repr_str += f'{indent_str}ratios={self.ratios},\n'
+        repr_str += f'{indent_str}num_levels={self.num_levels},\n'
+        repr_str += f'{indent_str}base_sizes={self.base_sizes},\n'
+        repr_str += f'{indent_str}basesize_ratio_range='
+        repr_str += f'{self.basesize_ratio_range})'
+        return repr_str
+
+
+@ANCHOR_GENERATORS.register_module()
+class LegacyAnchorGenerator(AnchorGenerator):
+    """Legacy anchor generator used in MMDetection V1.x.
+
+    Note:
+        Difference to the V2.0 anchor generator:
+
+        1. The center offset of V1.x anchors are set to be 0.5 rather than 0.
+        2. The width/height are minused by 1 when calculating the anchors' \
+            centers and corners to meet the V1.x coordinate system.
+        3. The anchors' corners are quantized.
+
+    Args:
+        strides (list[int] | list[tuple[int]]): Strides of anchors
+            in multiple feature levels.
+        ratios (list[float]): The list of ratios between the height and width
+            of anchors in a single level.
+        scales (list[int] | None): Anchor scales for anchors in a single level.
+            It cannot be set at the same time if `octave_base_scale` and
+            `scales_per_octave` are set.
+        base_sizes (list[int]): The basic sizes of anchors in multiple levels.
+            If None is given, strides will be used to generate base_sizes.
+        scale_major (bool): Whether to multiply scales first when generating
+            base anchors. If true, the anchors in the same row will have the
+            same scales. By default it is True in V2.0
+        octave_base_scale (int): The base scale of octave.
+        scales_per_octave (int): Number of scales for each octave.
+            `octave_base_scale` and `scales_per_octave` are usually used in
+            retinanet and the `scales` should be None when they are set.
+        centers (list[tuple[float, float]] | None): The centers of the anchor
+            relative to the feature grid center in multiple feature levels.
+            By default it is set to be None and not used. It a list of float
+            is given, this list will be used to shift the centers of anchors.
+        center_offset (float): The offset of center in propotion to anchors'
+            width and height. By default it is 0.5 in V2.0 but it should be 0.5
+            in v1.x models.
+
+    Examples:
+        >>> from mmdet.core import LegacyAnchorGenerator
+        >>> self = LegacyAnchorGenerator(
+        >>>     [16], [1.], [1.], [9], center_offset=0.5)
+        >>> all_anchors = self.grid_anchors(((2, 2),), device='cpu')
+        >>> print(all_anchors)
+        [tensor([[ 0.,  0.,  8.,  8.],
+                [16.,  0., 24.,  8.],
+                [ 0., 16.,  8., 24.],
+                [16., 16., 24., 24.]])]
+    """
+
+    def gen_single_level_base_anchors(self,
+                                      base_size,
+                                      scales,
+                                      ratios,
+                                      center=None):
+        """Generate base anchors of a single level.
+
+        Note:
+            The width/height of anchors are minused by 1 when calculating \
+                the centers and corners to meet the V1.x coordinate system.
+
+        Args:
+            base_size (int | float): Basic size of an anchor.
+            scales (torch.Tensor): Scales of the anchor.
+            ratios (torch.Tensor): The ratio between between the height.
+                and width of anchors in a single level.
+            center (tuple[float], optional): The center of the base anchor
+                related to a single feature grid. Defaults to None.
+
+        Returns:
+            torch.Tensor: Anchors in a single-level feature map.
+        """
+        w = base_size
+        h = base_size
+        if center is None:
+            x_center = self.center_offset * (w - 1)
+            y_center = self.center_offset * (h - 1)
+        else:
+            x_center, y_center = center
+
+        h_ratios = torch.sqrt(ratios)
+        w_ratios = 1 / h_ratios
+        if self.scale_major:
+            ws = (w * w_ratios[:, None] * scales[None, :]).view(-1)
+            hs = (h * h_ratios[:, None] * scales[None, :]).view(-1)
+        else:
+            ws = (w * scales[:, None] * w_ratios[None, :]).view(-1)
+            hs = (h * scales[:, None] * h_ratios[None, :]).view(-1)
+
+        # use float anchor and the anchor's center is aligned with the
+        # pixel center
+        base_anchors = [
+            x_center - 0.5 * (ws - 1), y_center - 0.5 * (hs - 1),
+            x_center + 0.5 * (ws - 1), y_center + 0.5 * (hs - 1)
+        ]
+        base_anchors = torch.stack(base_anchors, dim=-1).round()
+
+        return base_anchors
+
+
+@ANCHOR_GENERATORS.register_module()
+class LegacySSDAnchorGenerator(SSDAnchorGenerator, LegacyAnchorGenerator):
+    """Legacy anchor generator used in MMDetection V1.x.
+
+    The difference between `LegacySSDAnchorGenerator` and `SSDAnchorGenerator`
+    can be found in `LegacyAnchorGenerator`.
+    """
+
+    def __init__(self,
+                 strides,
+                 ratios,
+                 basesize_ratio_range,
+                 input_size=300,
+                 scale_major=True):
+        super(LegacySSDAnchorGenerator,
+              self).__init__(strides, ratios, basesize_ratio_range, input_size,
+                             scale_major)
+        self.centers = [((stride - 1) / 2., (stride - 1) / 2.)
+                        for stride in strides]
+        self.base_anchors = self.gen_base_anchors()
+
+
+@ANCHOR_GENERATORS.register_module()
+class YOLOAnchorGenerator(AnchorGenerator):
+    """Anchor generator for YOLO.
+
+    Args:
+        strides (list[int] | list[tuple[int, int]]): Strides of anchors
+            in multiple feature levels.
+        base_sizes (list[list[tuple[int, int]]]): The basic sizes
+            of anchors in multiple levels.
+    """
+
+    def __init__(self, strides, base_sizes):
+        self.strides = [_pair(stride) for stride in strides]
+        self.centers = [(stride[0] / 2., stride[1] / 2.)
+                        for stride in self.strides]
+        self.base_sizes = []
+        num_anchor_per_level = len(base_sizes[0])
+        for base_sizes_per_level in base_sizes:
+            assert num_anchor_per_level == len(base_sizes_per_level)
+            self.base_sizes.append(
+                [_pair(base_size) for base_size in base_sizes_per_level])
+        self.base_anchors = self.gen_base_anchors()
+
+    @property
+    def num_levels(self):
+        """int: number of feature levels that the generator will be applied"""
+        return len(self.base_sizes)
+
+    def gen_base_anchors(self):
+        """Generate base anchors.
+
+        Returns:
+            list(torch.Tensor): Base anchors of a feature grid in multiple \
+                feature levels.
+        """
+        multi_level_base_anchors = []
+        for i, base_sizes_per_level in enumerate(self.base_sizes):
+            center = None
+            if self.centers is not None:
+                center = self.centers[i]
+            multi_level_base_anchors.append(
+                self.gen_single_level_base_anchors(base_sizes_per_level,
+                                                   center))
+        return multi_level_base_anchors
+
+    def gen_single_level_base_anchors(self, base_sizes_per_level, center=None):
+        """Generate base anchors of a single level.
+
+        Args:
+            base_sizes_per_level (list[tuple[int, int]]): Basic sizes of
+                anchors.
+            center (tuple[float], optional): The center of the base anchor
+                related to a single feature grid. Defaults to None.
+
+        Returns:
+            torch.Tensor: Anchors in a single-level feature maps.
+        """
+        x_center, y_center = center
+        base_anchors = []
+        for base_size in base_sizes_per_level:
+            w, h = base_size
+
+            # use float anchor and the anchor's center is aligned with the
+            # pixel center
+            base_anchor = torch.Tensor([
+                x_center - 0.5 * w, y_center - 0.5 * h, x_center + 0.5 * w,
+                y_center + 0.5 * h
+            ])
+            base_anchors.append(base_anchor)
+        base_anchors = torch.stack(base_anchors, dim=0)
+
+        return base_anchors
+
+    def responsible_flags(self, featmap_sizes, gt_bboxes, device='cuda'):
+        """Generate responsible anchor flags of grid cells in multiple scales.
+
+        Args:
+            featmap_sizes (list(tuple)): List of feature map sizes in multiple
+                feature levels.
+            gt_bboxes (Tensor): Ground truth boxes, shape (n, 4).
+            device (str): Device where the anchors will be put on.
+
+        Return:
+            list(torch.Tensor): responsible flags of anchors in multiple level
+        """
+        assert self.num_levels == len(featmap_sizes)
+        multi_level_responsible_flags = []
+        for i in range(self.num_levels):
+            anchor_stride = self.strides[i]
+            flags = self.single_level_responsible_flags(
+                featmap_sizes[i],
+                gt_bboxes,
+                anchor_stride,
+                self.num_base_anchors[i],
+                device=device)
+            multi_level_responsible_flags.append(flags)
+        return multi_level_responsible_flags
+
+    def single_level_responsible_flags(self,
+                                       featmap_size,
+                                       gt_bboxes,
+                                       stride,
+                                       num_base_anchors,
+                                       device='cuda'):
+        """Generate the responsible flags of anchor in a single feature map.
+
+        Args:
+            featmap_size (tuple[int]): The size of feature maps.
+            gt_bboxes (Tensor): Ground truth boxes, shape (n, 4).
+            stride (tuple(int)): stride of current level
+            num_base_anchors (int): The number of base anchors.
+            device (str, optional): Device where the flags will be put on.
+                Defaults to 'cuda'.
+
+        Returns:
+            torch.Tensor: The valid flags of each anchor in a single level \
+                feature map.
+        """
+        feat_h, feat_w = featmap_size
+        gt_bboxes_cx = ((gt_bboxes[:, 0] + gt_bboxes[:, 2]) * 0.5).to(device)
+        gt_bboxes_cy = ((gt_bboxes[:, 1] + gt_bboxes[:, 3]) * 0.5).to(device)
+        gt_bboxes_grid_x = torch.floor(gt_bboxes_cx / stride[0]).long()
+        gt_bboxes_grid_y = torch.floor(gt_bboxes_cy / stride[1]).long()
+
+        # row major indexing
+        gt_bboxes_grid_idx = gt_bboxes_grid_y * feat_w + gt_bboxes_grid_x
+
+        responsible_grid = torch.zeros(
+            feat_h * feat_w, dtype=torch.uint8, device=device)
+        responsible_grid[gt_bboxes_grid_idx] = 1
+
+        responsible_grid = responsible_grid[:, None].expand(
+            responsible_grid.size(0), num_base_anchors).contiguous().view(-1)
+        return responsible_grid
diff --git a/python/autoCut/models/swin/mmdet/core/anchor/builder.py b/python/autoCut/models/swin/mmdet/core/anchor/builder.py
new file mode 100644
index 0000000..d79b448
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/anchor/builder.py
@@ -0,0 +1,7 @@
+from mmcv.utils import Registry, build_from_cfg
+
+ANCHOR_GENERATORS = Registry('Anchor generator')
+
+
+def build_anchor_generator(cfg, default_args=None):
+    return build_from_cfg(cfg, ANCHOR_GENERATORS, default_args)
diff --git a/python/autoCut/models/swin/mmdet/core/anchor/point_generator.py b/python/autoCut/models/swin/mmdet/core/anchor/point_generator.py
new file mode 100644
index 0000000..e6fbd98
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/anchor/point_generator.py
@@ -0,0 +1,37 @@
+import torch
+
+from .builder import ANCHOR_GENERATORS
+
+
+@ANCHOR_GENERATORS.register_module()
+class PointGenerator(object):
+
+    def _meshgrid(self, x, y, row_major=True):
+        xx = x.repeat(len(y))
+        yy = y.view(-1, 1).repeat(1, len(x)).view(-1)
+        if row_major:
+            return xx, yy
+        else:
+            return yy, xx
+
+    def grid_points(self, featmap_size, stride=16, device='cuda'):
+        feat_h, feat_w = featmap_size
+        shift_x = torch.arange(0., feat_w, device=device) * stride
+        shift_y = torch.arange(0., feat_h, device=device) * stride
+        shift_xx, shift_yy = self._meshgrid(shift_x, shift_y)
+        stride = shift_x.new_full((shift_xx.shape[0], ), stride)
+        shifts = torch.stack([shift_xx, shift_yy, stride], dim=-1)
+        all_points = shifts.to(device)
+        return all_points
+
+    def valid_flags(self, featmap_size, valid_size, device='cuda'):
+        feat_h, feat_w = featmap_size
+        valid_h, valid_w = valid_size
+        assert valid_h <= feat_h and valid_w <= feat_w
+        valid_x = torch.zeros(feat_w, dtype=torch.bool, device=device)
+        valid_y = torch.zeros(feat_h, dtype=torch.bool, device=device)
+        valid_x[:valid_w] = 1
+        valid_y[:valid_h] = 1
+        valid_xx, valid_yy = self._meshgrid(valid_x, valid_y)
+        valid = valid_xx & valid_yy
+        return valid
diff --git a/python/autoCut/models/swin/mmdet/core/anchor/utils.py b/python/autoCut/models/swin/mmdet/core/anchor/utils.py
new file mode 100644
index 0000000..ab9b53f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/anchor/utils.py
@@ -0,0 +1,71 @@
+import torch
+
+
+def images_to_levels(target, num_levels):
+    """Convert targets by image to targets by feature level.
+
+    [target_img0, target_img1] -> [target_level0, target_level1, ...]
+    """
+    target = torch.stack(target, 0)
+    level_targets = []
+    start = 0
+    for n in num_levels:
+        end = start + n
+        # level_targets.append(target[:, start:end].squeeze(0))
+        level_targets.append(target[:, start:end])
+        start = end
+    return level_targets
+
+
+def anchor_inside_flags(flat_anchors,
+                        valid_flags,
+                        img_shape,
+                        allowed_border=0):
+    """Check whether the anchors are inside the border.
+
+    Args:
+        flat_anchors (torch.Tensor): Flatten anchors, shape (n, 4).
+        valid_flags (torch.Tensor): An existing valid flags of anchors.
+        img_shape (tuple(int)): Shape of current image.
+        allowed_border (int, optional): The border to allow the valid anchor.
+            Defaults to 0.
+
+    Returns:
+        torch.Tensor: Flags indicating whether the anchors are inside a \
+            valid range.
+    """
+    img_h, img_w = img_shape[:2]
+    if allowed_border >= 0:
+        inside_flags = valid_flags & \
+            (flat_anchors[:, 0] >= -allowed_border) & \
+            (flat_anchors[:, 1] >= -allowed_border) & \
+            (flat_anchors[:, 2] < img_w + allowed_border) & \
+            (flat_anchors[:, 3] < img_h + allowed_border)
+    else:
+        inside_flags = valid_flags
+    return inside_flags
+
+
+def calc_region(bbox, ratio, featmap_size=None):
+    """Calculate a proportional bbox region.
+
+    The bbox center are fixed and the new h' and w' is h * ratio and w * ratio.
+
+    Args:
+        bbox (Tensor): Bboxes to calculate regions, shape (n, 4).
+        ratio (float): Ratio of the output region.
+        featmap_size (tuple): Feature map size used for clipping the boundary.
+
+    Returns:
+        tuple: x1, y1, x2, y2
+    """
+    x1 = torch.round((1 - ratio) * bbox[0] + ratio * bbox[2]).long()
+    y1 = torch.round((1 - ratio) * bbox[1] + ratio * bbox[3]).long()
+    x2 = torch.round(ratio * bbox[0] + (1 - ratio) * bbox[2]).long()
+    y2 = torch.round(ratio * bbox[1] + (1 - ratio) * bbox[3]).long()
+    if featmap_size is not None:
+        x1 = x1.clamp(min=0, max=featmap_size[1])
+        y1 = y1.clamp(min=0, max=featmap_size[0])
+        x2 = x2.clamp(min=0, max=featmap_size[1])
+        y2 = y2.clamp(min=0, max=featmap_size[0])
+    return (x1, y1, x2, y2)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/__init__.py b/python/autoCut/models/swin/mmdet/core/bbox/__init__.py
new file mode 100644
index 0000000..a353729
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/__init__.py
@@ -0,0 +1,27 @@
+from .assigners import (AssignResult, BaseAssigner, CenterRegionAssigner,
+                        MaxIoUAssigner, RegionAssigner)
+from .builder import build_assigner, build_bbox_coder, build_sampler
+from .coder import (BaseBBoxCoder, DeltaXYWHBBoxCoder, PseudoBBoxCoder,
+                    TBLRBBoxCoder)
+from .iou_calculators import BboxOverlaps2D, bbox_overlaps
+from .samplers import (BaseSampler, CombinedSampler,
+                       InstanceBalancedPosSampler, IoUBalancedNegSampler,
+                       OHEMSampler, PseudoSampler, RandomSampler,
+                       SamplingResult, ScoreHLRSampler)
+from .transforms import (bbox2distance, bbox2result, bbox2roi,
+                         bbox_cxcywh_to_xyxy, bbox_flip, bbox_mapping,
+                         bbox_mapping_back, bbox_rescale, bbox_xyxy_to_cxcywh,
+                         distance2bbox, roi2bbox)
+
+__all__ = [
+    'bbox_overlaps', 'BboxOverlaps2D', 'BaseAssigner', 'MaxIoUAssigner',
+    'AssignResult', 'BaseSampler', 'PseudoSampler', 'RandomSampler',
+    'InstanceBalancedPosSampler', 'IoUBalancedNegSampler', 'CombinedSampler',
+    'OHEMSampler', 'SamplingResult', 'ScoreHLRSampler', 'build_assigner',
+    'build_sampler', 'bbox_flip', 'bbox_mapping', 'bbox_mapping_back',
+    'bbox2roi', 'roi2bbox', 'bbox2result', 'distance2bbox', 'bbox2distance',
+    'build_bbox_coder', 'BaseBBoxCoder', 'PseudoBBoxCoder',
+    'DeltaXYWHBBoxCoder', 'TBLRBBoxCoder', 'CenterRegionAssigner',
+    'bbox_rescale', 'bbox_cxcywh_to_xyxy', 'bbox_xyxy_to_cxcywh',
+    'RegionAssigner'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/__init__.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/__init__.py
new file mode 100644
index 0000000..3143fbb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/__init__.py
@@ -0,0 +1,19 @@
+from .approx_max_iou_assigner import ApproxMaxIoUAssigner
+from .assign_result import AssignResult
+from .atss_assigner import ATSSAssigner
+from .base_assigner import BaseAssigner
+from .center_region_assigner import CenterRegionAssigner
+from .grid_assigner import GridAssigner
+from .hungarian_assigner import HungarianAssigner
+from .max_iou_assigner import MaxIoUAssigner
+from .point_assigner import PointAssigner
+from .point_assigner_v2 import PointAssignerV2
+from .point_hm_assigner import PointHMAssigner
+from .atss_assigner_v2 import ATSSAssignerV2
+from .region_assigner import RegionAssigner
+
+__all__ = [
+    'BaseAssigner', 'MaxIoUAssigner', 'ApproxMaxIoUAssigner', 'AssignResult',
+    'PointAssigner', 'ATSSAssigner', 'CenterRegionAssigner', 'GridAssigner',
+    'HungarianAssigner', 'RegionAssigner', 'PointAssignerV2', 'PointHMAssigner', 'ATSSAssignerV2'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/approx_max_iou_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/approx_max_iou_assigner.py
new file mode 100644
index 0000000..6d07656
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/approx_max_iou_assigner.py
@@ -0,0 +1,145 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from ..iou_calculators import build_iou_calculator
+from .max_iou_assigner import MaxIoUAssigner
+
+
+@BBOX_ASSIGNERS.register_module()
+class ApproxMaxIoUAssigner(MaxIoUAssigner):
+    """Assign a corresponding gt bbox or background to each bbox.
+
+    Each proposals will be assigned with an integer indicating the ground truth
+     index. (semi-positive index: gt label (0-based), -1: background)
+
+    - -1: negative sample, no assigned gt
+    - semi-positive integer: positive sample, index (0-based) of assigned gt
+
+    Args:
+        pos_iou_thr (float): IoU threshold for positive bboxes.
+        neg_iou_thr (float or tuple): IoU threshold for negative bboxes.
+        min_pos_iou (float): Minimum iou for a bbox to be considered as a
+            positive bbox. Positive samples can have smaller IoU than
+            pos_iou_thr due to the 4th step (assign max IoU sample to each gt).
+        gt_max_assign_all (bool): Whether to assign all bboxes with the same
+            highest overlap with some gt to that gt.
+        ignore_iof_thr (float): IoF threshold for ignoring bboxes (if
+            `gt_bboxes_ignore` is specified). Negative values mean not
+            ignoring any bboxes.
+        ignore_wrt_candidates (bool): Whether to compute the iof between
+            `bboxes` and `gt_bboxes_ignore`, or the contrary.
+        match_low_quality (bool): Whether to allow quality matches. This is
+            usually allowed for RPN and single stage detectors, but not allowed
+            in the second stage.
+        gpu_assign_thr (int): The upper bound of the number of GT for GPU
+            assign. When the number of gt is above this threshold, will assign
+            on CPU device. Negative values mean not assign on CPU.
+    """
+
+    def __init__(self,
+                 pos_iou_thr,
+                 neg_iou_thr,
+                 min_pos_iou=.0,
+                 gt_max_assign_all=True,
+                 ignore_iof_thr=-1,
+                 ignore_wrt_candidates=True,
+                 match_low_quality=True,
+                 gpu_assign_thr=-1,
+                 iou_calculator=dict(type='BboxOverlaps2D')):
+        self.pos_iou_thr = pos_iou_thr
+        self.neg_iou_thr = neg_iou_thr
+        self.min_pos_iou = min_pos_iou
+        self.gt_max_assign_all = gt_max_assign_all
+        self.ignore_iof_thr = ignore_iof_thr
+        self.ignore_wrt_candidates = ignore_wrt_candidates
+        self.gpu_assign_thr = gpu_assign_thr
+        self.match_low_quality = match_low_quality
+        self.iou_calculator = build_iou_calculator(iou_calculator)
+
+    def assign(self,
+               approxs,
+               squares,
+               approxs_per_octave,
+               gt_bboxes,
+               gt_bboxes_ignore=None,
+               gt_labels=None):
+        """Assign gt to approxs.
+
+        This method assign a gt bbox to each group of approxs (bboxes),
+        each group of approxs is represent by a base approx (bbox) and
+        will be assigned with -1, or a semi-positive number.
+        background_label (-1) means negative sample,
+        semi-positive number is the index (0-based) of assigned gt.
+        The assignment is done in following steps, the order matters.
+
+        1. assign every bbox to background_label (-1)
+        2. use the max IoU of each group of approxs to assign
+        2. assign proposals whose iou with all gts < neg_iou_thr to background
+        3. for each bbox, if the iou with its nearest gt >= pos_iou_thr,
+           assign it to that bbox
+        4. for each gt bbox, assign its nearest proposals (may be more than
+           one) to itself
+
+        Args:
+            approxs (Tensor): Bounding boxes to be assigned,
+                shape(approxs_per_octave*n, 4).
+            squares (Tensor): Base Bounding boxes to be assigned,
+                shape(n, 4).
+            approxs_per_octave (int): number of approxs per octave
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`, e.g., crowd boxes in COCO.
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        num_squares = squares.size(0)
+        num_gts = gt_bboxes.size(0)
+
+        if num_squares == 0 or num_gts == 0:
+            # No predictions and/or truth, return empty assignment
+            overlaps = approxs.new(num_gts, num_squares)
+            assign_result = self.assign_wrt_overlaps(overlaps, gt_labels)
+            return assign_result
+
+        # re-organize anchors by approxs_per_octave x num_squares
+        approxs = torch.transpose(
+            approxs.view(num_squares, approxs_per_octave, 4), 0,
+            1).contiguous().view(-1, 4)
+        assign_on_cpu = True if (self.gpu_assign_thr > 0) and (
+            num_gts > self.gpu_assign_thr) else False
+        # compute overlap and assign gt on CPU when number of GT is large
+        if assign_on_cpu:
+            device = approxs.device
+            approxs = approxs.cpu()
+            gt_bboxes = gt_bboxes.cpu()
+            if gt_bboxes_ignore is not None:
+                gt_bboxes_ignore = gt_bboxes_ignore.cpu()
+            if gt_labels is not None:
+                gt_labels = gt_labels.cpu()
+        all_overlaps = self.iou_calculator(approxs, gt_bboxes)
+
+        overlaps, _ = all_overlaps.view(approxs_per_octave, num_squares,
+                                        num_gts).max(dim=0)
+        overlaps = torch.transpose(overlaps, 0, 1)
+
+        if (self.ignore_iof_thr > 0 and gt_bboxes_ignore is not None
+                and gt_bboxes_ignore.numel() > 0 and squares.numel() > 0):
+            if self.ignore_wrt_candidates:
+                ignore_overlaps = self.iou_calculator(
+                    squares, gt_bboxes_ignore, mode='iof')
+                ignore_max_overlaps, _ = ignore_overlaps.max(dim=1)
+            else:
+                ignore_overlaps = self.iou_calculator(
+                    gt_bboxes_ignore, squares, mode='iof')
+                ignore_max_overlaps, _ = ignore_overlaps.max(dim=0)
+            overlaps[:, ignore_max_overlaps > self.ignore_iof_thr] = -1
+
+        assign_result = self.assign_wrt_overlaps(overlaps, gt_labels)
+        if assign_on_cpu:
+            assign_result.gt_inds = assign_result.gt_inds.to(device)
+            assign_result.max_overlaps = assign_result.max_overlaps.to(device)
+            if assign_result.labels is not None:
+                assign_result.labels = assign_result.labels.to(device)
+        return assign_result
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/assign_result.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/assign_result.py
new file mode 100644
index 0000000..4639fbd
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/assign_result.py
@@ -0,0 +1,204 @@
+import torch
+
+from mmdet.utils import util_mixins
+
+
+class AssignResult(util_mixins.NiceRepr):
+    """Stores assignments between predicted and truth boxes.
+
+    Attributes:
+        num_gts (int): the number of truth boxes considered when computing this
+            assignment
+
+        gt_inds (LongTensor): for each predicted box indicates the 1-based
+            index of the assigned truth box. 0 means unassigned and -1 means
+            ignore.
+
+        max_overlaps (FloatTensor): the iou between the predicted box and its
+            assigned truth box.
+
+        labels (None | LongTensor): If specified, for each predicted box
+            indicates the category label of the assigned truth box.
+
+    Example:
+        >>> # An assign result between 4 predicted boxes and 9 true boxes
+        >>> # where only two boxes were assigned.
+        >>> num_gts = 9
+        >>> max_overlaps = torch.LongTensor([0, .5, .9, 0])
+        >>> gt_inds = torch.LongTensor([-1, 1, 2, 0])
+        >>> labels = torch.LongTensor([0, 3, 4, 0])
+        >>> self = AssignResult(num_gts, gt_inds, max_overlaps, labels)
+        >>> print(str(self))  # xdoctest: +IGNORE_WANT
+        <AssignResult(num_gts=9, gt_inds.shape=(4,), max_overlaps.shape=(4,),
+                      labels.shape=(4,))>
+        >>> # Force addition of gt labels (when adding gt as proposals)
+        >>> new_labels = torch.LongTensor([3, 4, 5])
+        >>> self.add_gt_(new_labels)
+        >>> print(str(self))  # xdoctest: +IGNORE_WANT
+        <AssignResult(num_gts=9, gt_inds.shape=(7,), max_overlaps.shape=(7,),
+                      labels.shape=(7,))>
+    """
+
+    def __init__(self, num_gts, gt_inds, max_overlaps, labels=None):
+        self.num_gts = num_gts
+        self.gt_inds = gt_inds
+        self.max_overlaps = max_overlaps
+        self.labels = labels
+        # Interface for possible user-defined properties
+        self._extra_properties = {}
+
+    @property
+    def num_preds(self):
+        """int: the number of predictions in this assignment"""
+        return len(self.gt_inds)
+
+    def set_extra_property(self, key, value):
+        """Set user-defined new property."""
+        assert key not in self.info
+        self._extra_properties[key] = value
+
+    def get_extra_property(self, key):
+        """Get user-defined property."""
+        return self._extra_properties.get(key, None)
+
+    @property
+    def info(self):
+        """dict: a dictionary of info about the object"""
+        basic_info = {
+            'num_gts': self.num_gts,
+            'num_preds': self.num_preds,
+            'gt_inds': self.gt_inds,
+            'max_overlaps': self.max_overlaps,
+            'labels': self.labels,
+        }
+        basic_info.update(self._extra_properties)
+        return basic_info
+
+    def __nice__(self):
+        """str: a "nice" summary string describing this assign result"""
+        parts = []
+        parts.append(f'num_gts={self.num_gts!r}')
+        if self.gt_inds is None:
+            parts.append(f'gt_inds={self.gt_inds!r}')
+        else:
+            parts.append(f'gt_inds.shape={tuple(self.gt_inds.shape)!r}')
+        if self.max_overlaps is None:
+            parts.append(f'max_overlaps={self.max_overlaps!r}')
+        else:
+            parts.append('max_overlaps.shape='
+                         f'{tuple(self.max_overlaps.shape)!r}')
+        if self.labels is None:
+            parts.append(f'labels={self.labels!r}')
+        else:
+            parts.append(f'labels.shape={tuple(self.labels.shape)!r}')
+        return ', '.join(parts)
+
+    @classmethod
+    def random(cls, **kwargs):
+        """Create random AssignResult for tests or debugging.
+
+        Args:
+            num_preds: number of predicted boxes
+            num_gts: number of true boxes
+            p_ignore (float): probability of a predicted box assinged to an
+                ignored truth
+            p_assigned (float): probability of a predicted box not being
+                assigned
+            p_use_label (float | bool): with labels or not
+            rng (None | int | numpy.random.RandomState): seed or state
+
+        Returns:
+            :obj:`AssignResult`: Randomly generated assign results.
+
+        Example:
+            >>> from mmdet.core.bbox.assigners.assign_result import *  # NOQA
+            >>> self = AssignResult.random()
+            >>> print(self.info)
+        """
+        from mmdet.core.bbox import demodata
+        rng = demodata.ensure_rng(kwargs.get('rng', None))
+
+        num_gts = kwargs.get('num_gts', None)
+        num_preds = kwargs.get('num_preds', None)
+        p_ignore = kwargs.get('p_ignore', 0.3)
+        p_assigned = kwargs.get('p_assigned', 0.7)
+        p_use_label = kwargs.get('p_use_label', 0.5)
+        num_classes = kwargs.get('p_use_label', 3)
+
+        if num_gts is None:
+            num_gts = rng.randint(0, 8)
+        if num_preds is None:
+            num_preds = rng.randint(0, 16)
+
+        if num_gts == 0:
+            max_overlaps = torch.zeros(num_preds, dtype=torch.float32)
+            gt_inds = torch.zeros(num_preds, dtype=torch.int64)
+            if p_use_label is True or p_use_label < rng.rand():
+                labels = torch.zeros(num_preds, dtype=torch.int64)
+            else:
+                labels = None
+        else:
+            import numpy as np
+            # Create an overlap for each predicted box
+            max_overlaps = torch.from_numpy(rng.rand(num_preds))
+
+            # Construct gt_inds for each predicted box
+            is_assigned = torch.from_numpy(rng.rand(num_preds) < p_assigned)
+            # maximum number of assignments constraints
+            n_assigned = min(num_preds, min(num_gts, is_assigned.sum()))
+
+            assigned_idxs = np.where(is_assigned)[0]
+            rng.shuffle(assigned_idxs)
+            assigned_idxs = assigned_idxs[0:n_assigned]
+            assigned_idxs.sort()
+
+            is_assigned[:] = 0
+            is_assigned[assigned_idxs] = True
+
+            is_ignore = torch.from_numpy(
+                rng.rand(num_preds) < p_ignore) & is_assigned
+
+            gt_inds = torch.zeros(num_preds, dtype=torch.int64)
+
+            true_idxs = np.arange(num_gts)
+            rng.shuffle(true_idxs)
+            true_idxs = torch.from_numpy(true_idxs)
+            gt_inds[is_assigned] = true_idxs[:n_assigned]
+
+            gt_inds = torch.from_numpy(
+                rng.randint(1, num_gts + 1, size=num_preds))
+            gt_inds[is_ignore] = -1
+            gt_inds[~is_assigned] = 0
+            max_overlaps[~is_assigned] = 0
+
+            if p_use_label is True or p_use_label < rng.rand():
+                if num_classes == 0:
+                    labels = torch.zeros(num_preds, dtype=torch.int64)
+                else:
+                    labels = torch.from_numpy(
+                        # remind that we set FG labels to [0, num_class-1]
+                        # since mmdet v2.0
+                        # BG cat_id: num_class
+                        rng.randint(0, num_classes, size=num_preds))
+                    labels[~is_assigned] = 0
+            else:
+                labels = None
+
+        self = cls(num_gts, gt_inds, max_overlaps, labels)
+        return self
+
+    def add_gt_(self, gt_labels):
+        """Add ground truth as assigned results.
+
+        Args:
+            gt_labels (torch.Tensor): Labels of gt boxes
+        """
+        self_inds = torch.arange(
+            1, len(gt_labels) + 1, dtype=torch.long, device=gt_labels.device)
+        self.gt_inds = torch.cat([self_inds, self.gt_inds])
+
+        self.max_overlaps = torch.cat(
+            [self.max_overlaps.new_ones(len(gt_labels)), self.max_overlaps])
+
+        if self.labels is not None:
+            self.labels = torch.cat([gt_labels, self.labels])
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/atss_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/atss_assigner.py
new file mode 100644
index 0000000..d4fe9d0
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/atss_assigner.py
@@ -0,0 +1,178 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from ..iou_calculators import build_iou_calculator
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+
+@BBOX_ASSIGNERS.register_module()
+class ATSSAssigner(BaseAssigner):
+    """Assign a corresponding gt bbox or background to each bbox.
+
+    Each proposals will be assigned with `0` or a positive integer
+    indicating the ground truth index.
+
+    - 0: negative sample, no assigned gt
+    - positive integer: positive sample, index (1-based) of assigned gt
+
+    Args:
+        topk (float): number of bbox selected in each level
+    """
+
+    def __init__(self,
+                 topk,
+                 iou_calculator=dict(type='BboxOverlaps2D'),
+                 ignore_iof_thr=-1):
+        self.topk = topk
+        self.iou_calculator = build_iou_calculator(iou_calculator)
+        self.ignore_iof_thr = ignore_iof_thr
+
+    # https://github.com/sfzhang15/ATSS/blob/master/atss_core/modeling/rpn/atss/loss.py
+
+    def assign(self,
+               bboxes,
+               num_level_bboxes,
+               gt_bboxes,
+               gt_bboxes_ignore=None,
+               gt_labels=None):
+        """Assign gt to bboxes.
+
+        The assignment is done in following steps
+
+        1. compute iou between all bbox (bbox of all pyramid levels) and gt
+        2. compute center distance between all bbox and gt
+        3. on each pyramid level, for each gt, select k bbox whose center
+           are closest to the gt center, so we total select k*l bbox as
+           candidates for each gt
+        4. get corresponding iou for the these candidates, and compute the
+           mean and std, set mean + std as the iou threshold
+        5. select these candidates whose iou are greater than or equal to
+           the threshold as positive
+        6. limit the positive sample's center in gt
+
+
+        Args:
+            bboxes (Tensor): Bounding boxes to be assigned, shape(n, 4).
+            num_level_bboxes (List): num of bboxes in each level
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`, e.g., crowd boxes in COCO.
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        INF = 100000000
+        bboxes = bboxes[:, :4]
+        num_gt, num_bboxes = gt_bboxes.size(0), bboxes.size(0)
+
+        # compute iou between all bbox and gt
+        overlaps = self.iou_calculator(bboxes, gt_bboxes)
+
+        # assign 0 by default
+        assigned_gt_inds = overlaps.new_full((num_bboxes, ),
+                                             0,
+                                             dtype=torch.long)
+
+        if num_gt == 0 or num_bboxes == 0:
+            # No ground truth or boxes, return empty assignment
+            max_overlaps = overlaps.new_zeros((num_bboxes, ))
+            if num_gt == 0:
+                # No truth, assign everything to background
+                assigned_gt_inds[:] = 0
+            if gt_labels is None:
+                assigned_labels = None
+            else:
+                assigned_labels = overlaps.new_full((num_bboxes, ),
+                                                    -1,
+                                                    dtype=torch.long)
+            return AssignResult(
+                num_gt, assigned_gt_inds, max_overlaps, labels=assigned_labels)
+
+        # compute center distance between all bbox and gt
+        gt_cx = (gt_bboxes[:, 0] + gt_bboxes[:, 2]) / 2.0
+        gt_cy = (gt_bboxes[:, 1] + gt_bboxes[:, 3]) / 2.0
+        gt_points = torch.stack((gt_cx, gt_cy), dim=1)
+
+        bboxes_cx = (bboxes[:, 0] + bboxes[:, 2]) / 2.0
+        bboxes_cy = (bboxes[:, 1] + bboxes[:, 3]) / 2.0
+        bboxes_points = torch.stack((bboxes_cx, bboxes_cy), dim=1)
+
+        distances = (bboxes_points[:, None, :] -
+                     gt_points[None, :, :]).pow(2).sum(-1).sqrt()
+
+        if (self.ignore_iof_thr > 0 and gt_bboxes_ignore is not None
+                and gt_bboxes_ignore.numel() > 0 and bboxes.numel() > 0):
+            ignore_overlaps = self.iou_calculator(
+                bboxes, gt_bboxes_ignore, mode='iof')
+            ignore_max_overlaps, _ = ignore_overlaps.max(dim=1)
+            ignore_idxs = ignore_max_overlaps > self.ignore_iof_thr
+            distances[ignore_idxs, :] = INF
+            assigned_gt_inds[ignore_idxs] = -1
+
+        # Selecting candidates based on the center distance
+        candidate_idxs = []
+        start_idx = 0
+        for level, bboxes_per_level in enumerate(num_level_bboxes):
+            # on each pyramid level, for each gt,
+            # select k bbox whose center are closest to the gt center
+            end_idx = start_idx + bboxes_per_level
+            distances_per_level = distances[start_idx:end_idx, :]
+            selectable_k = min(self.topk, bboxes_per_level)
+            _, topk_idxs_per_level = distances_per_level.topk(
+                selectable_k, dim=0, largest=False)
+            candidate_idxs.append(topk_idxs_per_level + start_idx)
+            start_idx = end_idx
+        candidate_idxs = torch.cat(candidate_idxs, dim=0)
+
+        # get corresponding iou for the these candidates, and compute the
+        # mean and std, set mean + std as the iou threshold
+        candidate_overlaps = overlaps[candidate_idxs, torch.arange(num_gt)]
+        overlaps_mean_per_gt = candidate_overlaps.mean(0)
+        overlaps_std_per_gt = candidate_overlaps.std(0)
+        overlaps_thr_per_gt = overlaps_mean_per_gt + overlaps_std_per_gt
+
+        is_pos = candidate_overlaps >= overlaps_thr_per_gt[None, :]
+
+        # limit the positive sample's center in gt
+        for gt_idx in range(num_gt):
+            candidate_idxs[:, gt_idx] += gt_idx * num_bboxes
+        ep_bboxes_cx = bboxes_cx.view(1, -1).expand(
+            num_gt, num_bboxes).contiguous().view(-1)
+        ep_bboxes_cy = bboxes_cy.view(1, -1).expand(
+            num_gt, num_bboxes).contiguous().view(-1)
+        candidate_idxs = candidate_idxs.view(-1)
+
+        # calculate the left, top, right, bottom distance between positive
+        # bbox center and gt side
+        l_ = ep_bboxes_cx[candidate_idxs].view(-1, num_gt) - gt_bboxes[:, 0]
+        t_ = ep_bboxes_cy[candidate_idxs].view(-1, num_gt) - gt_bboxes[:, 1]
+        r_ = gt_bboxes[:, 2] - ep_bboxes_cx[candidate_idxs].view(-1, num_gt)
+        b_ = gt_bboxes[:, 3] - ep_bboxes_cy[candidate_idxs].view(-1, num_gt)
+        is_in_gts = torch.stack([l_, t_, r_, b_], dim=1).min(dim=1)[0] > 0.01
+        is_pos = is_pos & is_in_gts
+
+        # if an anchor box is assigned to multiple gts,
+        # the one with the highest IoU will be selected.
+        overlaps_inf = torch.full_like(overlaps,
+                                       -INF).t().contiguous().view(-1)
+        index = candidate_idxs.view(-1)[is_pos.view(-1)]
+        overlaps_inf[index] = overlaps.t().contiguous().view(-1)[index]
+        overlaps_inf = overlaps_inf.view(num_gt, -1).t()
+
+        max_overlaps, argmax_overlaps = overlaps_inf.max(dim=1)
+        assigned_gt_inds[
+            max_overlaps != -INF] = argmax_overlaps[max_overlaps != -INF] + 1
+
+        if gt_labels is not None:
+            assigned_labels = assigned_gt_inds.new_full((num_bboxes, ), -1)
+            pos_inds = torch.nonzero(
+                assigned_gt_inds > 0, as_tuple=False).squeeze()
+            if pos_inds.numel() > 0:
+                assigned_labels[pos_inds] = gt_labels[
+                    assigned_gt_inds[pos_inds] - 1]
+        else:
+            assigned_labels = None
+        return AssignResult(
+            num_gt, assigned_gt_inds, max_overlaps, labels=assigned_labels)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/atss_assigner_v2.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/atss_assigner_v2.py
new file mode 100644
index 0000000..c7b3f5d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/atss_assigner_v2.py
@@ -0,0 +1,205 @@
+import torch
+import random 
+
+from ..builder import BBOX_ASSIGNERS
+from ..iou_calculators import build_iou_calculator
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+
+@BBOX_ASSIGNERS.register_module()
+class ATSSAssignerV2(BaseAssigner):
+    """Assign a corresponding gt bbox or background to each bbox.
+
+    Each proposals will be assigned with `0` or a positive integer
+    indicating the ground truth index.
+
+    - 0: negative sample, no assigned gt
+    - positive integer: positive sample, index (1-based) of assigned gt
+
+    Args:
+        topk (float): number of bbox selected in each level
+        mask_center_sample: used in mask reppoints v2
+    ##: a little difference in dealing with the situation when the number of bbox per level is smaller than topk
+    """
+
+    def __init__(self,
+                 topk,
+                 iou_calculator=dict(type='BboxOverlaps2D'),
+                 ignore_iof_thr=-1,
+                 mask_center_sample=False):
+        self.topk = topk
+        self.iou_calculator = build_iou_calculator(iou_calculator)
+        self.ignore_iof_thr = ignore_iof_thr
+        self.mask_center_sample = mask_center_sample
+
+    # https://github.com/sfzhang15/ATSS/blob/master/atss_core/modeling/rpn/atss/loss.py
+
+    def assign(self,
+               bboxes,
+               num_level_bboxes,
+               gt_bboxes,
+               gt_bboxes_ignore=None,
+               gt_labels=None,
+               gt_bit_masks=None):
+        """Assign gt to bboxes.
+
+        The assignment is done in following steps
+
+        1. compute iou between all bbox (bbox of all pyramid levels) and gt
+        2. compute center distance between all bbox and gt
+        3. on each pyramid level, for each gt, select k bbox whose center
+           are closest to the gt center, so we total select k*l bbox as
+           candidates for each gt
+        4. get corresponding iou for the these candidates, and compute the
+           mean and std, set mean + std as the iou threshold
+        5. select these candidates whose iou are greater than or equal to
+           the threshold as positive
+        6. limit the positive sample's center in gt
+
+
+        Args:
+            bboxes (Tensor): Bounding boxes to be assigned, shape(n, 4).
+            num_level_bboxes (List): num of bboxes in each level
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`, e.g., crowd boxes in COCO.
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        INF = 100000000
+        bboxes = bboxes[:, :4]
+        num_gt, num_bboxes = gt_bboxes.size(0), bboxes.size(0)
+
+        # compute iou between all bbox and gt
+        overlaps = self.iou_calculator(bboxes, gt_bboxes)
+
+        # assign 0 by default
+        assigned_gt_inds = overlaps.new_full((num_bboxes, ),
+                                             0,
+                                             dtype=torch.long)
+
+        if num_gt == 0 or num_bboxes == 0:
+            # No ground truth or boxes, return empty assignment
+            max_overlaps = overlaps.new_zeros((num_bboxes, ))
+            if num_gt == 0:
+                # No truth, assign everything to background
+                assigned_gt_inds[:] = 0
+            if gt_labels is None:
+                assigned_labels = None
+            else:
+                assigned_labels = overlaps.new_full((num_bboxes, ),
+                                                    -1,
+                                                    dtype=torch.long)
+            return AssignResult(
+                num_gt, assigned_gt_inds, max_overlaps, labels=assigned_labels)
+
+        # compute center distance between all bbox and gt
+        if self.mask_center_sample:
+            _, _h, _w = gt_bit_masks.size()
+
+            _ys = torch.arange(0, _h, dtype=torch.float32, device=gt_bit_masks.device)
+            _xs = torch.arange(0, _w, dtype=torch.float32, device=gt_bit_masks.device)
+
+            m00 = gt_bit_masks.sum(dim=-1).sum(dim=-1).clamp(min=1e-6)
+            m10 = (gt_bit_masks * _xs).sum(dim=-1).sum(dim=-1)
+            m01 = (gt_bit_masks * _ys[:, None]).sum(dim=-1).sum(dim=-1)
+            gt_cx = m10 / m00
+            gt_cy = m01 / m00
+        else:
+            gt_cx = (gt_bboxes[:, 0] + gt_bboxes[:, 2]) / 2.0
+            gt_cy = (gt_bboxes[:, 1] + gt_bboxes[:, 3]) / 2.0
+        gt_points = torch.stack((gt_cx, gt_cy), dim=1)
+
+        bboxes_cx = (bboxes[:, 0] + bboxes[:, 2]) / 2.0
+        bboxes_cy = (bboxes[:, 1] + bboxes[:, 3]) / 2.0
+        bboxes_points = torch.stack((bboxes_cx, bboxes_cy), dim=1)
+
+        distances = (bboxes_points[:, None, :] -
+                     gt_points[None, :, :]).pow(2).sum(-1).sqrt()
+
+        if (self.ignore_iof_thr > 0 and gt_bboxes_ignore is not None
+                and gt_bboxes_ignore.numel() > 0 and bboxes.numel() > 0):
+            ignore_overlaps = self.iou_calculator(
+                bboxes, gt_bboxes_ignore, mode='iof')
+            ignore_max_overlaps, _ = ignore_overlaps.max(dim=1)
+            ignore_idxs = ignore_max_overlaps > self.ignore_iof_thr
+            distances[ignore_idxs, :] = INF
+            assigned_gt_inds[ignore_idxs] = -1
+
+        # Selecting candidates based on the center distance
+        candidate_idxs = []
+        start_idx = 0
+        for level, bboxes_per_level in enumerate(num_level_bboxes):
+            # on each pyramid level, for each gt,
+            # select k bbox whose center are closest to the gt center
+            end_idx = start_idx + bboxes_per_level
+            if bboxes_per_level<self.topk:
+                provided_idxs = list(range(start_idx,end_idx))
+                if len(provided_idxs) >= self.topk-bboxes_per_level:
+                    pesudo_idxs_per_level = provided_idxs + random.sample(provided_idxs,self.topk-bboxes_per_level)
+                else:
+                    pesudo_idxs_per_level = provided_idxs + [random.choice(provided_idxs) for _ in range(self.topk-bboxes_per_level)]
+                pesudo_idxs_per_level = distances.new_tensor(pesudo_idxs_per_level).long().unsqueeze(1).expand(self.topk,gt_points.size(0))
+                candidate_idxs.append(pesudo_idxs_per_level)
+            else:
+                distances_per_level = distances[start_idx:end_idx, :]
+                selectable_k = min(self.topk, bboxes_per_level)
+                _, topk_idxs_per_level = distances_per_level.topk(
+                    selectable_k, dim=0, largest=False)
+                candidate_idxs.append(topk_idxs_per_level + start_idx)
+            start_idx = end_idx
+        candidate_idxs = torch.cat(candidate_idxs, dim=0)
+
+        # get corresponding iou for the these candidates, and compute the
+        # mean and std, set mean + std as the iou threshold
+        candidate_overlaps = overlaps[candidate_idxs, torch.arange(num_gt)]
+        overlaps_mean_per_gt = candidate_overlaps.mean(0)
+        overlaps_std_per_gt = candidate_overlaps.std(0)
+        overlaps_thr_per_gt = overlaps_mean_per_gt + overlaps_std_per_gt
+
+        is_pos = candidate_overlaps >= overlaps_thr_per_gt[None, :]
+
+        # limit the positive sample's center in gt
+        for gt_idx in range(num_gt):
+            candidate_idxs[:, gt_idx] += gt_idx * num_bboxes
+        ep_bboxes_cx = bboxes_cx.view(1, -1).expand(
+            num_gt, num_bboxes).contiguous().view(-1)
+        ep_bboxes_cy = bboxes_cy.view(1, -1).expand(
+            num_gt, num_bboxes).contiguous().view(-1)
+        candidate_idxs = candidate_idxs.view(-1)
+
+        # calculate the left, top, right, bottom distance between positive
+        # bbox center and gt side
+        l_ = ep_bboxes_cx[candidate_idxs].view(-1, num_gt) - gt_bboxes[:, 0]
+        t_ = ep_bboxes_cy[candidate_idxs].view(-1, num_gt) - gt_bboxes[:, 1]
+        r_ = gt_bboxes[:, 2] - ep_bboxes_cx[candidate_idxs].view(-1, num_gt)
+        b_ = gt_bboxes[:, 3] - ep_bboxes_cy[candidate_idxs].view(-1, num_gt)
+        is_in_gts = torch.stack([l_, t_, r_, b_], dim=1).min(dim=1)[0] > 0.01
+        is_pos = is_pos & is_in_gts
+
+        # if an anchor box is assigned to multiple gts,
+        # the one with the highest IoU will be selected.
+        overlaps_inf = torch.full_like(overlaps,
+                                       -INF).t().contiguous().view(-1)
+        index = candidate_idxs.view(-1)[is_pos.view(-1)]
+        overlaps_inf[index] = overlaps.t().contiguous().view(-1)[index]
+        overlaps_inf = overlaps_inf.view(num_gt, -1).t()
+
+        max_overlaps, argmax_overlaps = overlaps_inf.max(dim=1)
+        assigned_gt_inds[
+            max_overlaps != -INF] = argmax_overlaps[max_overlaps != -INF] + 1
+
+        if gt_labels is not None:
+            assigned_labels = assigned_gt_inds.new_full((num_bboxes, ), -1)
+            pos_inds = torch.nonzero(
+                assigned_gt_inds > 0, as_tuple=False).squeeze()
+            if pos_inds.numel() > 0:
+                assigned_labels[pos_inds] = gt_labels[
+                    assigned_gt_inds[pos_inds] - 1]
+        else:
+            assigned_labels = None
+        return AssignResult(
+            num_gt, assigned_gt_inds, max_overlaps, labels=assigned_labels)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/base_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/base_assigner.py
new file mode 100644
index 0000000..1ff0160
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/base_assigner.py
@@ -0,0 +1,9 @@
+from abc import ABCMeta, abstractmethod
+
+
+class BaseAssigner(metaclass=ABCMeta):
+    """Base assigner that assigns boxes to ground truth boxes."""
+
+    @abstractmethod
+    def assign(self, bboxes, gt_bboxes, gt_bboxes_ignore=None, gt_labels=None):
+        """Assign boxes to either a ground truth boxes or a negative boxes."""
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/center_region_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/center_region_assigner.py
new file mode 100644
index 0000000..488e3b6
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/center_region_assigner.py
@@ -0,0 +1,335 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from ..iou_calculators import build_iou_calculator
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+
+def scale_boxes(bboxes, scale):
+    """Expand an array of boxes by a given scale.
+
+    Args:
+        bboxes (Tensor): Shape (m, 4)
+        scale (float): The scale factor of bboxes
+
+    Returns:
+        (Tensor): Shape (m, 4). Scaled bboxes
+    """
+    assert bboxes.size(1) == 4
+    w_half = (bboxes[:, 2] - bboxes[:, 0]) * .5
+    h_half = (bboxes[:, 3] - bboxes[:, 1]) * .5
+    x_c = (bboxes[:, 2] + bboxes[:, 0]) * .5
+    y_c = (bboxes[:, 3] + bboxes[:, 1]) * .5
+
+    w_half *= scale
+    h_half *= scale
+
+    boxes_scaled = torch.zeros_like(bboxes)
+    boxes_scaled[:, 0] = x_c - w_half
+    boxes_scaled[:, 2] = x_c + w_half
+    boxes_scaled[:, 1] = y_c - h_half
+    boxes_scaled[:, 3] = y_c + h_half
+    return boxes_scaled
+
+
+def is_located_in(points, bboxes):
+    """Are points located in bboxes.
+
+    Args:
+      points (Tensor): Points, shape: (m, 2).
+      bboxes (Tensor): Bounding boxes, shape: (n, 4).
+
+    Return:
+      Tensor: Flags indicating if points are located in bboxes, shape: (m, n).
+    """
+    assert points.size(1) == 2
+    assert bboxes.size(1) == 4
+    return (points[:, 0].unsqueeze(1) > bboxes[:, 0].unsqueeze(0)) & \
+           (points[:, 0].unsqueeze(1) < bboxes[:, 2].unsqueeze(0)) & \
+           (points[:, 1].unsqueeze(1) > bboxes[:, 1].unsqueeze(0)) & \
+           (points[:, 1].unsqueeze(1) < bboxes[:, 3].unsqueeze(0))
+
+
+def bboxes_area(bboxes):
+    """Compute the area of an array of bboxes.
+
+    Args:
+        bboxes (Tensor): The coordinates ox bboxes. Shape: (m, 4)
+
+    Returns:
+        Tensor: Area of the bboxes. Shape: (m, )
+    """
+    assert bboxes.size(1) == 4
+    w = (bboxes[:, 2] - bboxes[:, 0])
+    h = (bboxes[:, 3] - bboxes[:, 1])
+    areas = w * h
+    return areas
+
+
+@BBOX_ASSIGNERS.register_module()
+class CenterRegionAssigner(BaseAssigner):
+    """Assign pixels at the center region of a bbox as positive.
+
+    Each proposals will be assigned with `-1`, `0`, or a positive integer
+    indicating the ground truth index.
+    - -1: negative samples
+    - semi-positive numbers: positive sample, index (0-based) of assigned gt
+
+    Args:
+        pos_scale (float): Threshold within which pixels are
+          labelled as positive.
+        neg_scale (float): Threshold above which pixels are
+          labelled as positive.
+        min_pos_iof (float): Minimum iof of a pixel with a gt to be
+          labelled as positive. Default: 1e-2
+        ignore_gt_scale (float): Threshold within which the pixels
+          are ignored when the gt is labelled as shadowed. Default: 0.5
+        foreground_dominate (bool): If True, the bbox will be assigned as
+          positive when a gt's kernel region overlaps with another's shadowed
+          (ignored) region, otherwise it is set as ignored. Default to False.
+    """
+
+    def __init__(self,
+                 pos_scale,
+                 neg_scale,
+                 min_pos_iof=1e-2,
+                 ignore_gt_scale=0.5,
+                 foreground_dominate=False,
+                 iou_calculator=dict(type='BboxOverlaps2D')):
+        self.pos_scale = pos_scale
+        self.neg_scale = neg_scale
+        self.min_pos_iof = min_pos_iof
+        self.ignore_gt_scale = ignore_gt_scale
+        self.foreground_dominate = foreground_dominate
+        self.iou_calculator = build_iou_calculator(iou_calculator)
+
+    def get_gt_priorities(self, gt_bboxes):
+        """Get gt priorities according to their areas.
+
+        Smaller gt has higher priority.
+
+        Args:
+            gt_bboxes (Tensor): Ground truth boxes, shape (k, 4).
+
+        Returns:
+            Tensor: The priority of gts so that gts with larger priority is \
+              more likely to be assigned. Shape (k, )
+        """
+        gt_areas = bboxes_area(gt_bboxes)
+        # Rank all gt bbox areas. Smaller objects has larger priority
+        _, sort_idx = gt_areas.sort(descending=True)
+        sort_idx = sort_idx.argsort()
+        return sort_idx
+
+    def assign(self, bboxes, gt_bboxes, gt_bboxes_ignore=None, gt_labels=None):
+        """Assign gt to bboxes.
+
+        This method assigns gts to every bbox (proposal/anchor), each bbox \
+        will be assigned with -1, or a semi-positive number. -1 means \
+        negative sample, semi-positive number is the index (0-based) of \
+        assigned gt.
+
+        Args:
+            bboxes (Tensor): Bounding boxes to be assigned, shape(n, 4).
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (tensor, optional): Ground truth bboxes that are
+              labelled as `ignored`, e.g., crowd boxes in COCO.
+            gt_labels (tensor, optional): Label of gt_bboxes, shape (num_gts,).
+
+        Returns:
+            :obj:`AssignResult`: The assigned result. Note that \
+              shadowed_labels of shape (N, 2) is also added as an \
+              `assign_result` attribute. `shadowed_labels` is a tensor \
+              composed of N pairs of anchor_ind, class_label], where N \
+              is the number of anchors that lie in the outer region of a \
+              gt, anchor_ind is the shadowed anchor index and class_label \
+              is the shadowed class label.
+
+        Example:
+            >>> self = CenterRegionAssigner(0.2, 0.2)
+            >>> bboxes = torch.Tensor([[0, 0, 10, 10], [10, 10, 20, 20]])
+            >>> gt_bboxes = torch.Tensor([[0, 0, 10, 10]])
+            >>> assign_result = self.assign(bboxes, gt_bboxes)
+            >>> expected_gt_inds = torch.LongTensor([1, 0])
+            >>> assert torch.all(assign_result.gt_inds == expected_gt_inds)
+        """
+        # There are in total 5 steps in the pixel assignment
+        # 1. Find core (the center region, say inner 0.2)
+        #     and shadow (the relatively ourter part, say inner 0.2-0.5)
+        #     regions of every gt.
+        # 2. Find all prior bboxes that lie in gt_core and gt_shadow regions
+        # 3. Assign prior bboxes in gt_core with a one-hot id of the gt in
+        #      the image.
+        #    3.1. For overlapping objects, the prior bboxes in gt_core is
+        #           assigned with the object with smallest area
+        # 4. Assign prior bboxes with class label according to its gt id.
+        #    4.1. Assign -1 to prior bboxes lying in shadowed gts
+        #    4.2. Assign positive prior boxes with the corresponding label
+        # 5. Find pixels lying in the shadow of an object and assign them with
+        #      background label, but set the loss weight of its corresponding
+        #      gt to zero.
+        assert bboxes.size(1) == 4, 'bboxes must have size of 4'
+        # 1. Find core positive and shadow region of every gt
+        gt_core = scale_boxes(gt_bboxes, self.pos_scale)
+        gt_shadow = scale_boxes(gt_bboxes, self.neg_scale)
+
+        # 2. Find prior bboxes that lie in gt_core and gt_shadow regions
+        bbox_centers = (bboxes[:, 2:4] + bboxes[:, 0:2]) / 2
+        # The center points lie within the gt boxes
+        is_bbox_in_gt = is_located_in(bbox_centers, gt_bboxes)
+        # Only calculate bbox and gt_core IoF. This enables small prior bboxes
+        #   to match large gts
+        bbox_and_gt_core_overlaps = self.iou_calculator(
+            bboxes, gt_core, mode='iof')
+        # The center point of effective priors should be within the gt box
+        is_bbox_in_gt_core = is_bbox_in_gt & (
+            bbox_and_gt_core_overlaps > self.min_pos_iof)  # shape (n, k)
+
+        is_bbox_in_gt_shadow = (
+            self.iou_calculator(bboxes, gt_shadow, mode='iof') >
+            self.min_pos_iof)
+        # Rule out center effective positive pixels
+        is_bbox_in_gt_shadow &= (~is_bbox_in_gt_core)
+
+        num_gts, num_bboxes = gt_bboxes.size(0), bboxes.size(0)
+        if num_gts == 0 or num_bboxes == 0:
+            # If no gts exist, assign all pixels to negative
+            assigned_gt_ids = \
+                is_bbox_in_gt_core.new_zeros((num_bboxes,),
+                                             dtype=torch.long)
+            pixels_in_gt_shadow = assigned_gt_ids.new_empty((0, 2))
+        else:
+            # Step 3: assign a one-hot gt id to each pixel, and smaller objects
+            #    have high priority to assign the pixel.
+            sort_idx = self.get_gt_priorities(gt_bboxes)
+            assigned_gt_ids, pixels_in_gt_shadow = \
+                self.assign_one_hot_gt_indices(is_bbox_in_gt_core,
+                                               is_bbox_in_gt_shadow,
+                                               gt_priority=sort_idx)
+
+        if gt_bboxes_ignore is not None and gt_bboxes_ignore.numel() > 0:
+            # No ground truth or boxes, return empty assignment
+            gt_bboxes_ignore = scale_boxes(
+                gt_bboxes_ignore, scale=self.ignore_gt_scale)
+            is_bbox_in_ignored_gts = is_located_in(bbox_centers,
+                                                   gt_bboxes_ignore)
+            is_bbox_in_ignored_gts = is_bbox_in_ignored_gts.any(dim=1)
+            assigned_gt_ids[is_bbox_in_ignored_gts] = -1
+
+        # 4. Assign prior bboxes with class label according to its gt id.
+        assigned_labels = None
+        shadowed_pixel_labels = None
+        if gt_labels is not None:
+            # Default assigned label is the background (-1)
+            assigned_labels = assigned_gt_ids.new_full((num_bboxes, ), -1)
+            pos_inds = torch.nonzero(
+                assigned_gt_ids > 0, as_tuple=False).squeeze()
+            if pos_inds.numel() > 0:
+                assigned_labels[pos_inds] = gt_labels[assigned_gt_ids[pos_inds]
+                                                      - 1]
+            # 5. Find pixels lying in the shadow of an object
+            shadowed_pixel_labels = pixels_in_gt_shadow.clone()
+            if pixels_in_gt_shadow.numel() > 0:
+                pixel_idx, gt_idx =\
+                    pixels_in_gt_shadow[:, 0], pixels_in_gt_shadow[:, 1]
+                assert (assigned_gt_ids[pixel_idx] != gt_idx).all(), \
+                    'Some pixels are dually assigned to ignore and gt!'
+                shadowed_pixel_labels[:, 1] = gt_labels[gt_idx - 1]
+                override = (
+                    assigned_labels[pixel_idx] == shadowed_pixel_labels[:, 1])
+                if self.foreground_dominate:
+                    # When a pixel is both positive and shadowed, set it as pos
+                    shadowed_pixel_labels = shadowed_pixel_labels[~override]
+                else:
+                    # When a pixel is both pos and shadowed, set it as shadowed
+                    assigned_labels[pixel_idx[override]] = -1
+                    assigned_gt_ids[pixel_idx[override]] = 0
+
+        assign_result = AssignResult(
+            num_gts, assigned_gt_ids, None, labels=assigned_labels)
+        # Add shadowed_labels as assign_result property. Shape: (num_shadow, 2)
+        assign_result.set_extra_property('shadowed_labels',
+                                         shadowed_pixel_labels)
+        return assign_result
+
+    def assign_one_hot_gt_indices(self,
+                                  is_bbox_in_gt_core,
+                                  is_bbox_in_gt_shadow,
+                                  gt_priority=None):
+        """Assign only one gt index to each prior box.
+
+        Gts with large gt_priority are more likely to be assigned.
+
+        Args:
+            is_bbox_in_gt_core (Tensor): Bool tensor indicating the bbox center
+              is in the core area of a gt (e.g. 0-0.2).
+              Shape: (num_prior, num_gt).
+            is_bbox_in_gt_shadow (Tensor): Bool tensor indicating the bbox
+              center is in the shadowed area of a gt (e.g. 0.2-0.5).
+              Shape: (num_prior, num_gt).
+            gt_priority (Tensor): Priorities of gts. The gt with a higher
+              priority is more likely to be assigned to the bbox when the bbox
+              match with multiple gts. Shape: (num_gt, ).
+
+        Returns:
+            tuple: Returns (assigned_gt_inds, shadowed_gt_inds).
+
+                - assigned_gt_inds: The assigned gt index of each prior bbox \
+                    (i.e. index from 1 to num_gts). Shape: (num_prior, ).
+                - shadowed_gt_inds: shadowed gt indices. It is a tensor of \
+                    shape (num_ignore, 2) with first column being the \
+                    shadowed prior bbox indices and the second column the \
+                    shadowed gt indices (1-based).
+        """
+        num_bboxes, num_gts = is_bbox_in_gt_core.shape
+
+        if gt_priority is None:
+            gt_priority = torch.arange(
+                num_gts, device=is_bbox_in_gt_core.device)
+        assert gt_priority.size(0) == num_gts
+        # The bigger gt_priority, the more preferable to be assigned
+        # The assigned inds are by default 0 (background)
+        assigned_gt_inds = is_bbox_in_gt_core.new_zeros((num_bboxes, ),
+                                                        dtype=torch.long)
+        # Shadowed bboxes are assigned to be background. But the corresponding
+        #   label is ignored during loss calculation, which is done through
+        #   shadowed_gt_inds
+        shadowed_gt_inds = torch.nonzero(is_bbox_in_gt_shadow, as_tuple=False)
+        if is_bbox_in_gt_core.sum() == 0:  # No gt match
+            shadowed_gt_inds[:, 1] += 1  # 1-based. For consistency issue
+            return assigned_gt_inds, shadowed_gt_inds
+
+        # The priority of each prior box and gt pair. If one prior box is
+        #  matched bo multiple gts. Only the pair with the highest priority
+        #  is saved
+        pair_priority = is_bbox_in_gt_core.new_full((num_bboxes, num_gts),
+                                                    -1,
+                                                    dtype=torch.long)
+
+        # Each bbox could match with multiple gts.
+        # The following codes deal with this situation
+        # Matched  bboxes (to any gt). Shape: (num_pos_anchor, )
+        inds_of_match = torch.any(is_bbox_in_gt_core, dim=1)
+        # The matched gt index of each positive bbox. Length >= num_pos_anchor
+        #   , since one bbox could match multiple gts
+        matched_bbox_gt_inds = torch.nonzero(
+            is_bbox_in_gt_core, as_tuple=False)[:, 1]
+        # Assign priority to each bbox-gt pair.
+        pair_priority[is_bbox_in_gt_core] = gt_priority[matched_bbox_gt_inds]
+        _, argmax_priority = pair_priority[inds_of_match].max(dim=1)
+        assigned_gt_inds[inds_of_match] = argmax_priority + 1  # 1-based
+        # Zero-out the assigned anchor box to filter the shadowed gt indices
+        is_bbox_in_gt_core[inds_of_match, argmax_priority] = 0
+        # Concat the shadowed indices due to overlapping with that out side of
+        #   effective scale. shape: (total_num_ignore, 2)
+        shadowed_gt_inds = torch.cat(
+            (shadowed_gt_inds, torch.nonzero(
+                is_bbox_in_gt_core, as_tuple=False)),
+            dim=0)
+        # `is_bbox_in_gt_core` should be changed back to keep arguments intact.
+        is_bbox_in_gt_core[inds_of_match, argmax_priority] = 1
+        # 1-based shadowed gt indices, to be consistent with `assigned_gt_inds`
+        if shadowed_gt_inds.numel() > 0:
+            shadowed_gt_inds[:, 1] += 1
+        return assigned_gt_inds, shadowed_gt_inds
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/grid_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/grid_assigner.py
new file mode 100644
index 0000000..7390ea6
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/grid_assigner.py
@@ -0,0 +1,155 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from ..iou_calculators import build_iou_calculator
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+
+@BBOX_ASSIGNERS.register_module()
+class GridAssigner(BaseAssigner):
+    """Assign a corresponding gt bbox or background to each bbox.
+
+    Each proposals will be assigned with `-1`, `0`, or a positive integer
+    indicating the ground truth index.
+
+    - -1: don't care
+    - 0: negative sample, no assigned gt
+    - positive integer: positive sample, index (1-based) of assigned gt
+
+    Args:
+        pos_iou_thr (float): IoU threshold for positive bboxes.
+        neg_iou_thr (float or tuple): IoU threshold for negative bboxes.
+        min_pos_iou (float): Minimum iou for a bbox to be considered as a
+            positive bbox. Positive samples can have smaller IoU than
+            pos_iou_thr due to the 4th step (assign max IoU sample to each gt).
+        gt_max_assign_all (bool): Whether to assign all bboxes with the same
+            highest overlap with some gt to that gt.
+    """
+
+    def __init__(self,
+                 pos_iou_thr,
+                 neg_iou_thr,
+                 min_pos_iou=.0,
+                 gt_max_assign_all=True,
+                 iou_calculator=dict(type='BboxOverlaps2D')):
+        self.pos_iou_thr = pos_iou_thr
+        self.neg_iou_thr = neg_iou_thr
+        self.min_pos_iou = min_pos_iou
+        self.gt_max_assign_all = gt_max_assign_all
+        self.iou_calculator = build_iou_calculator(iou_calculator)
+
+    def assign(self, bboxes, box_responsible_flags, gt_bboxes, gt_labels=None):
+        """Assign gt to bboxes. The process is very much like the max iou
+        assigner, except that positive samples are constrained within the cell
+        that the gt boxes fell in.
+
+        This method assign a gt bbox to every bbox (proposal/anchor), each bbox
+        will be assigned with -1, 0, or a positive number. -1 means don't care,
+        0 means negative sample, positive number is the index (1-based) of
+        assigned gt.
+        The assignment is done in following steps, the order matters.
+
+        1. assign every bbox to -1
+        2. assign proposals whose iou with all gts <= neg_iou_thr to 0
+        3. for each bbox within a cell, if the iou with its nearest gt >
+            pos_iou_thr and the center of that gt falls inside the cell,
+            assign it to that bbox
+        4. for each gt bbox, assign its nearest proposals within the cell the
+            gt bbox falls in to itself.
+
+        Args:
+            bboxes (Tensor): Bounding boxes to be assigned, shape(n, 4).
+            box_responsible_flags (Tensor): flag to indicate whether box is
+                responsible for prediction, shape(n, )
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        num_gts, num_bboxes = gt_bboxes.size(0), bboxes.size(0)
+
+        # compute iou between all gt and bboxes
+        overlaps = self.iou_calculator(gt_bboxes, bboxes)
+
+        # 1. assign -1 by default
+        assigned_gt_inds = overlaps.new_full((num_bboxes, ),
+                                             -1,
+                                             dtype=torch.long)
+
+        if num_gts == 0 or num_bboxes == 0:
+            # No ground truth or boxes, return empty assignment
+            max_overlaps = overlaps.new_zeros((num_bboxes, ))
+            if num_gts == 0:
+                # No truth, assign everything to background
+                assigned_gt_inds[:] = 0
+            if gt_labels is None:
+                assigned_labels = None
+            else:
+                assigned_labels = overlaps.new_full((num_bboxes, ),
+                                                    -1,
+                                                    dtype=torch.long)
+            return AssignResult(
+                num_gts,
+                assigned_gt_inds,
+                max_overlaps,
+                labels=assigned_labels)
+
+        # 2. assign negative: below
+        # for each anchor, which gt best overlaps with it
+        # for each anchor, the max iou of all gts
+        # shape of max_overlaps == argmax_overlaps == num_bboxes
+        max_overlaps, argmax_overlaps = overlaps.max(dim=0)
+
+        if isinstance(self.neg_iou_thr, float):
+            assigned_gt_inds[(max_overlaps >= 0)
+                             & (max_overlaps <= self.neg_iou_thr)] = 0
+        elif isinstance(self.neg_iou_thr, (tuple, list)):
+            assert len(self.neg_iou_thr) == 2
+            assigned_gt_inds[(max_overlaps > self.neg_iou_thr[0])
+                             & (max_overlaps <= self.neg_iou_thr[1])] = 0
+
+        # 3. assign positive: falls into responsible cell and above
+        # positive IOU threshold, the order matters.
+        # the prior condition of comparision is to filter out all
+        # unrelated anchors, i.e. not box_responsible_flags
+        overlaps[:, ~box_responsible_flags.type(torch.bool)] = -1.
+
+        # calculate max_overlaps again, but this time we only consider IOUs
+        # for anchors responsible for prediction
+        max_overlaps, argmax_overlaps = overlaps.max(dim=0)
+
+        # for each gt, which anchor best overlaps with it
+        # for each gt, the max iou of all proposals
+        # shape of gt_max_overlaps == gt_argmax_overlaps == num_gts
+        gt_max_overlaps, gt_argmax_overlaps = overlaps.max(dim=1)
+
+        pos_inds = (max_overlaps >
+                    self.pos_iou_thr) & box_responsible_flags.type(torch.bool)
+        assigned_gt_inds[pos_inds] = argmax_overlaps[pos_inds] + 1
+
+        # 4. assign positive to max overlapped anchors within responsible cell
+        for i in range(num_gts):
+            if gt_max_overlaps[i] > self.min_pos_iou:
+                if self.gt_max_assign_all:
+                    max_iou_inds = (overlaps[i, :] == gt_max_overlaps[i]) & \
+                         box_responsible_flags.type(torch.bool)
+                    assigned_gt_inds[max_iou_inds] = i + 1
+                elif box_responsible_flags[gt_argmax_overlaps[i]]:
+                    assigned_gt_inds[gt_argmax_overlaps[i]] = i + 1
+
+        # assign labels of positive anchors
+        if gt_labels is not None:
+            assigned_labels = assigned_gt_inds.new_full((num_bboxes, ), -1)
+            pos_inds = torch.nonzero(
+                assigned_gt_inds > 0, as_tuple=False).squeeze()
+            if pos_inds.numel() > 0:
+                assigned_labels[pos_inds] = gt_labels[
+                    assigned_gt_inds[pos_inds] - 1]
+
+        else:
+            assigned_labels = None
+
+        return AssignResult(
+            num_gts, assigned_gt_inds, max_overlaps, labels=assigned_labels)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/hungarian_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/hungarian_assigner.py
new file mode 100644
index 0000000..e10cc14
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/hungarian_assigner.py
@@ -0,0 +1,145 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from ..match_costs import build_match_cost
+from ..transforms import bbox_cxcywh_to_xyxy
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+try:
+    from scipy.optimize import linear_sum_assignment
+except ImportError:
+    linear_sum_assignment = None
+
+
+@BBOX_ASSIGNERS.register_module()
+class HungarianAssigner(BaseAssigner):
+    """Computes one-to-one matching between predictions and ground truth.
+
+    This class computes an assignment between the targets and the predictions
+    based on the costs. The costs are weighted sum of three components:
+    classification cost, regression L1 cost and regression iou cost. The
+    targets don't include the no_object, so generally there are more
+    predictions than targets. After the one-to-one matching, the un-matched
+    are treated as backgrounds. Thus each query prediction will be assigned
+    with `0` or a positive integer indicating the ground truth index:
+
+    - 0: negative sample, no assigned gt
+    - positive integer: positive sample, index (1-based) of assigned gt
+
+    Args:
+        cls_weight (int | float, optional): The scale factor for classification
+            cost. Default 1.0.
+        bbox_weight (int | float, optional): The scale factor for regression
+            L1 cost. Default 1.0.
+        iou_weight (int | float, optional): The scale factor for regression
+            iou cost. Default 1.0.
+        iou_calculator (dict | optional): The config for the iou calculation.
+            Default type `BboxOverlaps2D`.
+        iou_mode (str | optional): "iou" (intersection over union), "iof"
+                (intersection over foreground), or "giou" (generalized
+                intersection over union). Default "giou".
+    """
+
+    def __init__(self,
+                 cls_cost=dict(type='ClassificationCost', weight=1.),
+                 reg_cost=dict(type='BBoxL1Cost', weight=1.0),
+                 iou_cost=dict(type='IoUCost', iou_mode='giou', weight=1.0)):
+        self.cls_cost = build_match_cost(cls_cost)
+        self.reg_cost = build_match_cost(reg_cost)
+        self.iou_cost = build_match_cost(iou_cost)
+
+    def assign(self,
+               bbox_pred,
+               cls_pred,
+               gt_bboxes,
+               gt_labels,
+               img_meta,
+               gt_bboxes_ignore=None,
+               eps=1e-7):
+        """Computes one-to-one matching based on the weighted costs.
+
+        This method assign each query prediction to a ground truth or
+        background. The `assigned_gt_inds` with -1 means don't care,
+        0 means negative sample, and positive number is the index (1-based)
+        of assigned gt.
+        The assignment is done in the following steps, the order matters.
+
+        1. assign every prediction to -1
+        2. compute the weighted costs
+        3. do Hungarian matching on CPU based on the costs
+        4. assign all to 0 (background) first, then for each matched pair
+           between predictions and gts, treat this prediction as foreground
+           and assign the corresponding gt index (plus 1) to it.
+
+        Args:
+            bbox_pred (Tensor): Predicted boxes with normalized coordinates
+                (cx, cy, w, h), which are all in range [0, 1]. Shape
+                [num_query, 4].
+            cls_pred (Tensor): Predicted classification logits, shape
+                [num_query, num_class].
+            gt_bboxes (Tensor): Ground truth boxes with unnormalized
+                coordinates (x1, y1, x2, y2). Shape [num_gt, 4].
+            gt_labels (Tensor): Label of `gt_bboxes`, shape (num_gt,).
+            img_meta (dict): Meta information for current image.
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`. Default None.
+            eps (int | float, optional): A value added to the denominator for
+                numerical stability. Default 1e-7.
+
+        Returns:
+            :obj:`AssignResult`: The assigned result.
+        """
+        assert gt_bboxes_ignore is None, \
+            'Only case when gt_bboxes_ignore is None is supported.'
+        num_gts, num_bboxes = gt_bboxes.size(0), bbox_pred.size(0)
+
+        # 1. assign -1 by default
+        assigned_gt_inds = bbox_pred.new_full((num_bboxes, ),
+                                              -1,
+                                              dtype=torch.long)
+        assigned_labels = bbox_pred.new_full((num_bboxes, ),
+                                             -1,
+                                             dtype=torch.long)
+        if num_gts == 0 or num_bboxes == 0:
+            # No ground truth or boxes, return empty assignment
+            if num_gts == 0:
+                # No ground truth, assign all to background
+                assigned_gt_inds[:] = 0
+            return AssignResult(
+                num_gts, assigned_gt_inds, None, labels=assigned_labels)
+        img_h, img_w, _ = img_meta['img_shape']
+        factor = gt_bboxes.new_tensor([img_w, img_h, img_w,
+                                       img_h]).unsqueeze(0)
+
+        # 2. compute the weighted costs
+        # classification and bboxcost.
+        cls_cost = self.cls_cost(cls_pred, gt_labels)
+        # regression L1 cost
+        normalize_gt_bboxes = gt_bboxes / factor
+        reg_cost = self.reg_cost(bbox_pred, normalize_gt_bboxes)
+        # regression iou cost, defaultly giou is used in official DETR.
+        bboxes = bbox_cxcywh_to_xyxy(bbox_pred) * factor
+        iou_cost = self.iou_cost(bboxes, gt_bboxes)
+        # weighted sum of above three costs
+        cost = cls_cost + reg_cost + iou_cost
+
+        # 3. do Hungarian matching on CPU using linear_sum_assignment
+        cost = cost.detach().cpu()
+        if linear_sum_assignment is None:
+            raise ImportError('Please run "pip install scipy" '
+                              'to install scipy first.')
+        matched_row_inds, matched_col_inds = linear_sum_assignment(cost)
+        matched_row_inds = torch.from_numpy(matched_row_inds).to(
+            bbox_pred.device)
+        matched_col_inds = torch.from_numpy(matched_col_inds).to(
+            bbox_pred.device)
+
+        # 4. assign backgrounds and foregrounds
+        # assign all indices to backgrounds first
+        assigned_gt_inds[:] = 0
+        # assign foregrounds based on matching results
+        assigned_gt_inds[matched_row_inds] = matched_col_inds + 1
+        assigned_labels[matched_row_inds] = gt_labels[matched_col_inds]
+        return AssignResult(
+            num_gts, assigned_gt_inds, None, labels=assigned_labels)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/max_iou_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/max_iou_assigner.py
new file mode 100644
index 0000000..5cf4c4b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/max_iou_assigner.py
@@ -0,0 +1,212 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from ..iou_calculators import build_iou_calculator
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+
+@BBOX_ASSIGNERS.register_module()
+class MaxIoUAssigner(BaseAssigner):
+    """Assign a corresponding gt bbox or background to each bbox.
+
+    Each proposals will be assigned with `-1`, or a semi-positive integer
+    indicating the ground truth index.
+
+    - -1: negative sample, no assigned gt
+    - semi-positive integer: positive sample, index (0-based) of assigned gt
+
+    Args:
+        pos_iou_thr (float): IoU threshold for positive bboxes.
+        neg_iou_thr (float or tuple): IoU threshold for negative bboxes.
+        min_pos_iou (float): Minimum iou for a bbox to be considered as a
+            positive bbox. Positive samples can have smaller IoU than
+            pos_iou_thr due to the 4th step (assign max IoU sample to each gt).
+        gt_max_assign_all (bool): Whether to assign all bboxes with the same
+            highest overlap with some gt to that gt.
+        ignore_iof_thr (float): IoF threshold for ignoring bboxes (if
+            `gt_bboxes_ignore` is specified). Negative values mean not
+            ignoring any bboxes.
+        ignore_wrt_candidates (bool): Whether to compute the iof between
+            `bboxes` and `gt_bboxes_ignore`, or the contrary.
+        match_low_quality (bool): Whether to allow low quality matches. This is
+            usually allowed for RPN and single stage detectors, but not allowed
+            in the second stage. Details are demonstrated in Step 4.
+        gpu_assign_thr (int): The upper bound of the number of GT for GPU
+            assign. When the number of gt is above this threshold, will assign
+            on CPU device. Negative values mean not assign on CPU.
+    """
+
+    def __init__(self,
+                 pos_iou_thr,
+                 neg_iou_thr,
+                 min_pos_iou=.0,
+                 gt_max_assign_all=True,
+                 ignore_iof_thr=-1,
+                 ignore_wrt_candidates=True,
+                 match_low_quality=True,
+                 gpu_assign_thr=-1,
+                 iou_calculator=dict(type='BboxOverlaps2D')):
+        self.pos_iou_thr = pos_iou_thr
+        self.neg_iou_thr = neg_iou_thr
+        self.min_pos_iou = min_pos_iou
+        self.gt_max_assign_all = gt_max_assign_all
+        self.ignore_iof_thr = ignore_iof_thr
+        self.ignore_wrt_candidates = ignore_wrt_candidates
+        self.gpu_assign_thr = gpu_assign_thr
+        self.match_low_quality = match_low_quality
+        self.iou_calculator = build_iou_calculator(iou_calculator)
+
+    def assign(self, bboxes, gt_bboxes, gt_bboxes_ignore=None, gt_labels=None):
+        """Assign gt to bboxes.
+
+        This method assign a gt bbox to every bbox (proposal/anchor), each bbox
+        will be assigned with -1, or a semi-positive number. -1 means negative
+        sample, semi-positive number is the index (0-based) of assigned gt.
+        The assignment is done in following steps, the order matters.
+
+        1. assign every bbox to the background
+        2. assign proposals whose iou with all gts < neg_iou_thr to 0
+        3. for each bbox, if the iou with its nearest gt >= pos_iou_thr,
+           assign it to that bbox
+        4. for each gt bbox, assign its nearest proposals (may be more than
+           one) to itself
+
+        Args:
+            bboxes (Tensor): Bounding boxes to be assigned, shape(n, 4).
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`, e.g., crowd boxes in COCO.
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+
+        Example:
+            >>> self = MaxIoUAssigner(0.5, 0.5)
+            >>> bboxes = torch.Tensor([[0, 0, 10, 10], [10, 10, 20, 20]])
+            >>> gt_bboxes = torch.Tensor([[0, 0, 10, 9]])
+            >>> assign_result = self.assign(bboxes, gt_bboxes)
+            >>> expected_gt_inds = torch.LongTensor([1, 0])
+            >>> assert torch.all(assign_result.gt_inds == expected_gt_inds)
+        """
+        assign_on_cpu = True if (self.gpu_assign_thr > 0) and (
+            gt_bboxes.shape[0] > self.gpu_assign_thr) else False
+        # compute overlap and assign gt on CPU when number of GT is large
+        if assign_on_cpu:
+            device = bboxes.device
+            bboxes = bboxes.cpu()
+            gt_bboxes = gt_bboxes.cpu()
+            if gt_bboxes_ignore is not None:
+                gt_bboxes_ignore = gt_bboxes_ignore.cpu()
+            if gt_labels is not None:
+                gt_labels = gt_labels.cpu()
+
+        overlaps = self.iou_calculator(gt_bboxes, bboxes)
+
+        if (self.ignore_iof_thr > 0 and gt_bboxes_ignore is not None
+                and gt_bboxes_ignore.numel() > 0 and bboxes.numel() > 0):
+            if self.ignore_wrt_candidates:
+                ignore_overlaps = self.iou_calculator(
+                    bboxes, gt_bboxes_ignore, mode='iof')
+                ignore_max_overlaps, _ = ignore_overlaps.max(dim=1)
+            else:
+                ignore_overlaps = self.iou_calculator(
+                    gt_bboxes_ignore, bboxes, mode='iof')
+                ignore_max_overlaps, _ = ignore_overlaps.max(dim=0)
+            overlaps[:, ignore_max_overlaps > self.ignore_iof_thr] = -1
+
+        assign_result = self.assign_wrt_overlaps(overlaps, gt_labels)
+        if assign_on_cpu:
+            assign_result.gt_inds = assign_result.gt_inds.to(device)
+            assign_result.max_overlaps = assign_result.max_overlaps.to(device)
+            if assign_result.labels is not None:
+                assign_result.labels = assign_result.labels.to(device)
+        return assign_result
+
+    def assign_wrt_overlaps(self, overlaps, gt_labels=None):
+        """Assign w.r.t. the overlaps of bboxes with gts.
+
+        Args:
+            overlaps (Tensor): Overlaps between k gt_bboxes and n bboxes,
+                shape(k, n).
+            gt_labels (Tensor, optional): Labels of k gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        num_gts, num_bboxes = overlaps.size(0), overlaps.size(1)
+
+        # 1. assign -1 by default
+        assigned_gt_inds = overlaps.new_full((num_bboxes, ),
+                                             -1,
+                                             dtype=torch.long)
+
+        if num_gts == 0 or num_bboxes == 0:
+            # No ground truth or boxes, return empty assignment
+            max_overlaps = overlaps.new_zeros((num_bboxes, ))
+            if num_gts == 0:
+                # No truth, assign everything to background
+                assigned_gt_inds[:] = 0
+            if gt_labels is None:
+                assigned_labels = None
+            else:
+                assigned_labels = overlaps.new_full((num_bboxes, ),
+                                                    -1,
+                                                    dtype=torch.long)
+            return AssignResult(
+                num_gts,
+                assigned_gt_inds,
+                max_overlaps,
+                labels=assigned_labels)
+
+        # for each anchor, which gt best overlaps with it
+        # for each anchor, the max iou of all gts
+        max_overlaps, argmax_overlaps = overlaps.max(dim=0)
+        # for each gt, which anchor best overlaps with it
+        # for each gt, the max iou of all proposals
+        gt_max_overlaps, gt_argmax_overlaps = overlaps.max(dim=1)
+
+        # 2. assign negative: below
+        # the negative inds are set to be 0
+        if isinstance(self.neg_iou_thr, float):
+            assigned_gt_inds[(max_overlaps >= 0)
+                             & (max_overlaps < self.neg_iou_thr)] = 0
+        elif isinstance(self.neg_iou_thr, tuple):
+            assert len(self.neg_iou_thr) == 2
+            assigned_gt_inds[(max_overlaps >= self.neg_iou_thr[0])
+                             & (max_overlaps < self.neg_iou_thr[1])] = 0
+
+        # 3. assign positive: above positive IoU threshold
+        pos_inds = max_overlaps >= self.pos_iou_thr
+        assigned_gt_inds[pos_inds] = argmax_overlaps[pos_inds] + 1
+
+        if self.match_low_quality:
+            # Low-quality matching will overwrite the assigned_gt_inds assigned
+            # in Step 3. Thus, the assigned gt might not be the best one for
+            # prediction.
+            # For example, if bbox A has 0.9 and 0.8 iou with GT bbox 1 & 2,
+            # bbox 1 will be assigned as the best target for bbox A in step 3.
+            # However, if GT bbox 2's gt_argmax_overlaps = A, bbox A's
+            # assigned_gt_inds will be overwritten to be bbox B.
+            # This might be the reason that it is not used in ROI Heads.
+            for i in range(num_gts):
+                if gt_max_overlaps[i] >= self.min_pos_iou:
+                    if self.gt_max_assign_all:
+                        max_iou_inds = overlaps[i, :] == gt_max_overlaps[i]
+                        assigned_gt_inds[max_iou_inds] = i + 1
+                    else:
+                        assigned_gt_inds[gt_argmax_overlaps[i]] = i + 1
+
+        if gt_labels is not None:
+            assigned_labels = assigned_gt_inds.new_full((num_bboxes, ), -1)
+            pos_inds = torch.nonzero(
+                assigned_gt_inds > 0, as_tuple=False).squeeze()
+            if pos_inds.numel() > 0:
+                assigned_labels[pos_inds] = gt_labels[
+                    assigned_gt_inds[pos_inds] - 1]
+        else:
+            assigned_labels = None
+
+        return AssignResult(
+            num_gts, assigned_gt_inds, max_overlaps, labels=assigned_labels)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_assigner.py
new file mode 100644
index 0000000..fb8f5e4
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_assigner.py
@@ -0,0 +1,133 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+
+@BBOX_ASSIGNERS.register_module()
+class PointAssigner(BaseAssigner):
+    """Assign a corresponding gt bbox or background to each point.
+
+    Each proposals will be assigned with `0`, or a positive integer
+    indicating the ground truth index.
+
+    - 0: negative sample, no assigned gt
+    - positive integer: positive sample, index (1-based) of assigned gt
+    """
+
+    def __init__(self, scale=4, pos_num=3):
+        self.scale = scale
+        self.pos_num = pos_num
+
+    def assign(self, points, gt_bboxes, gt_bboxes_ignore=None, gt_labels=None):
+        """Assign gt to points.
+
+        This method assign a gt bbox to every points set, each points set
+        will be assigned with  the background_label (-1), or a label number.
+        -1 is background, and semi-positive number is the index (0-based) of
+        assigned gt.
+        The assignment is done in following steps, the order matters.
+
+        1. assign every points to the background_label (-1)
+        2. A point is assigned to some gt bbox if
+            (i) the point is within the k closest points to the gt bbox
+            (ii) the distance between this point and the gt is smaller than
+                other gt bboxes
+
+        Args:
+            points (Tensor): points to be assigned, shape(n, 3) while last
+                dimension stands for (x, y, stride).
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`, e.g., crowd boxes in COCO.
+                NOTE: currently unused.
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        num_points = points.shape[0]
+        num_gts = gt_bboxes.shape[0]
+
+        if num_gts == 0 or num_points == 0:
+            # If no truth assign everything to the background
+            assigned_gt_inds = points.new_full((num_points, ),
+                                               0,
+                                               dtype=torch.long)
+            if gt_labels is None:
+                assigned_labels = None
+            else:
+                assigned_labels = points.new_full((num_points, ),
+                                                  -1,
+                                                  dtype=torch.long)
+            return AssignResult(
+                num_gts, assigned_gt_inds, None, labels=assigned_labels)
+
+        points_xy = points[:, :2]
+        points_stride = points[:, 2]
+        points_lvl = torch.log2(
+            points_stride).int()  # [3...,4...,5...,6...,7...]
+        lvl_min, lvl_max = points_lvl.min(), points_lvl.max()
+
+        # assign gt box
+        gt_bboxes_xy = (gt_bboxes[:, :2] + gt_bboxes[:, 2:]) / 2
+        gt_bboxes_wh = (gt_bboxes[:, 2:] - gt_bboxes[:, :2]).clamp(min=1e-6)
+        scale = self.scale
+        gt_bboxes_lvl = ((torch.log2(gt_bboxes_wh[:, 0] / scale) +
+                          torch.log2(gt_bboxes_wh[:, 1] / scale)) / 2).int()
+        gt_bboxes_lvl = torch.clamp(gt_bboxes_lvl, min=lvl_min, max=lvl_max)
+
+        # stores the assigned gt index of each point
+        assigned_gt_inds = points.new_zeros((num_points, ), dtype=torch.long)
+        # stores the assigned gt dist (to this point) of each point
+        assigned_gt_dist = points.new_full((num_points, ), float('inf'))
+        points_range = torch.arange(points.shape[0])
+
+        for idx in range(num_gts):
+            gt_lvl = gt_bboxes_lvl[idx]
+            # get the index of points in this level
+            lvl_idx = gt_lvl == points_lvl
+            points_index = points_range[lvl_idx]
+            # get the points in this level
+            lvl_points = points_xy[lvl_idx, :]
+            # get the center point of gt
+            gt_point = gt_bboxes_xy[[idx], :]
+            # get width and height of gt
+            gt_wh = gt_bboxes_wh[[idx], :]
+            # compute the distance between gt center and
+            #   all points in this level
+            points_gt_dist = ((lvl_points - gt_point) / gt_wh).norm(dim=1)
+            # find the nearest k points to gt center in this level
+            min_dist, min_dist_index = torch.topk(
+                points_gt_dist, self.pos_num, largest=False)
+            # the index of nearest k points to gt center in this level
+            min_dist_points_index = points_index[min_dist_index]
+            # The less_than_recorded_index stores the index
+            #   of min_dist that is less then the assigned_gt_dist. Where
+            #   assigned_gt_dist stores the dist from previous assigned gt
+            #   (if exist) to each point.
+            less_than_recorded_index = min_dist < assigned_gt_dist[
+                min_dist_points_index]
+            # The min_dist_points_index stores the index of points satisfy:
+            #   (1) it is k nearest to current gt center in this level.
+            #   (2) it is closer to current gt center than other gt center.
+            min_dist_points_index = min_dist_points_index[
+                less_than_recorded_index]
+            # assign the result
+            assigned_gt_inds[min_dist_points_index] = idx + 1
+            assigned_gt_dist[min_dist_points_index] = min_dist[
+                less_than_recorded_index]
+
+        if gt_labels is not None:
+            assigned_labels = assigned_gt_inds.new_full((num_points, ), -1)
+            pos_inds = torch.nonzero(
+                assigned_gt_inds > 0, as_tuple=False).squeeze()
+            if pos_inds.numel() > 0:
+                assigned_labels[pos_inds] = gt_labels[
+                    assigned_gt_inds[pos_inds] - 1]
+        else:
+            assigned_labels = None
+
+        return AssignResult(
+            num_gts, assigned_gt_inds, None, labels=assigned_labels)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_assigner_v2.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_assigner_v2.py
new file mode 100644
index 0000000..1d6d970
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_assigner_v2.py
@@ -0,0 +1,127 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+
+
+@BBOX_ASSIGNERS.register_module()
+class PointAssignerV2(BaseAssigner):
+    """Assign a corresponding gt bbox or background to each point.
+
+    Each proposals will be assigned with `0`, or a positive integer
+    indicating the ground truth index.
+
+    - 0: negative sample, no assigned gt
+    - positive integer: positive sample, index (1-based) of assigned gt
+    """
+
+    def __init__(self, scale=4, pos_num=3, mask_center_sample=False, use_center=False):
+        self.scale = scale
+        self.pos_num = pos_num
+        self.mask_center_sample = mask_center_sample
+        self.use_center = use_center
+
+    def assign(self, points, gt_bboxes, gt_bboxes_ignore=None, gt_labels=None, gt_bit_masks=None):
+        """Assign gt to bboxes.
+
+        This method assign a gt bbox to every point, each bbox
+        will be assigned with  0, or a positive number.
+        0 means negative sample, positive number is the index (1-based) of
+        assigned gt.
+        The assignment is done in following steps, the order matters.
+
+        1. assign every points to 0
+        2. for each gt box, we find the k most closest points to the
+            box center and assign the gt bbox to those points, we also record
+            the minimum distance from each point to the closest gt box. When we
+            assign the bbox to the points, we check whether its distance to the
+            points is closest.
+
+        Args:
+            points (Tensor): points to be assigned, shape(n, 3) while last
+                dimension stands for (x, y, stride).
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`, e.g., crowd boxes in COCO.
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        INF = 1e8
+        num_gts, num_points = gt_bboxes.shape[0], points.shape[0]
+
+        if num_gts == 0 or num_points == 0:
+            # If no truth assign everything to the background
+            assigned_gt_inds = points.new_full((num_points, ),
+                                               0,
+                                               dtype=torch.long)
+            if gt_labels is None:
+                assigned_labels = None
+            else:
+                assigned_labels = points.new_full((num_points, ),
+                                                  -1,
+                                                  dtype=torch.long)
+            return AssignResult(
+                num_gts, assigned_gt_inds, None, labels=assigned_labels)
+
+        points_xy = points[:, :2]
+        points_stride = points[:, 2]
+        points_lvl = torch.log2(points_stride).int()  # [3...,4...,5...,6...,7...]
+        lvl_min, lvl_max = points_lvl.min(), points_lvl.max()
+
+        # assign gt box
+        gt_bboxes_xy = (gt_bboxes[:, :2] + gt_bboxes[:, 2:]) / 2
+        # assign gt box
+        if self.mask_center_sample:
+            _, _h, _w = gt_bit_masks.size()
+
+            _ys = torch.arange(0, _h, dtype=torch.float32, device=gt_bit_masks.device)
+            _xs = torch.arange(0, _w, dtype=torch.float32, device=gt_bit_masks.device)
+
+            m00 = gt_bit_masks.sum(dim=-1).sum(dim=-1).clamp(min=1e-6)
+            m10 = (gt_bit_masks * _xs).sum(dim=-1).sum(dim=-1)
+            m01 = (gt_bit_masks * _ys[:, None]).sum(dim=-1).sum(dim=-1)
+            center_xs = m10 / m00
+            center_ys = m01 / m00
+            gt_center_xy = torch.stack((center_xs, center_ys), dim=-1)
+        else:
+            gt_center_xy = (gt_bboxes[:, :2] + gt_bboxes[:, 2:]) / 2
+
+        gt_bboxes_wh = (gt_bboxes[:, 2:] - gt_bboxes[:, :2]).clamp(min=1e-6)
+        scale = self.scale
+        gt_bboxes_lvl = ((torch.log2(gt_bboxes_wh[:, 0] / scale) +
+                          torch.log2(gt_bboxes_wh[:, 1] / scale)) / 2).int()
+        gt_bboxes_lvl = torch.clamp(gt_bboxes_lvl, min=lvl_min, max=lvl_max)
+
+        if self.use_center:
+            distances = ((points_xy[:, None, :] - gt_center_xy[None, :, :]) / gt_bboxes_wh[None, :, :]).norm(dim=2)
+        else:
+            distances = ((points_xy[:, None, :] - gt_bboxes_xy[None, :, :]) / gt_bboxes_wh[None, :, :]).norm(dim=2)
+        distances[points_lvl[:, None] != gt_bboxes_lvl[None, :]] = INF
+
+        # stores the assigned gt index of each point
+        assigned_gt_inds = points.new_zeros((num_points, ), dtype=torch.long)
+
+        min_dist, min_dist_index = torch.topk(distances, self.pos_num, dim=0, largest=False)
+
+        distances_inf = torch.full_like(distances, INF)
+        distances_inf[min_dist_index, torch.arange(num_gts)] = min_dist
+
+        min_dist, min_dist_index = distances_inf.min(dim=1)
+        assigned_gt_inds[min_dist != INF] = min_dist_index[min_dist != INF] + 1
+
+        if gt_labels is not None:
+            assigned_labels = assigned_gt_inds.new_full((num_points, ), -1)
+            pos_inds = torch.nonzero(
+                assigned_gt_inds > 0, as_tuple=False).squeeze()
+            if pos_inds.numel() > 0:
+                assigned_labels[pos_inds] = gt_labels[
+                    assigned_gt_inds[pos_inds] - 1]
+        else:
+            assigned_labels = None
+
+        return AssignResult(
+            num_gts, assigned_gt_inds, None, labels=assigned_labels)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_hm_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_hm_assigner.py
new file mode 100644
index 0000000..9e06b9d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/point_hm_assigner.py
@@ -0,0 +1,167 @@
+import torch
+
+from ..builder import BBOX_ASSIGNERS
+from .base_assigner import BaseAssigner
+from .assign_result import AssignResult
+
+
+@BBOX_ASSIGNERS.register_module()
+class PointHMAssigner(BaseAssigner):
+    """Assign a corresponding gt bbox or background to each point.
+
+    Each proposals will be assigned with `0`, or a positive integer
+    indicating the ground truth index.
+
+    - 0: negative sample, no assigned gt
+    - positive integer: positive sample, index (1-based) of assigned gt
+
+    """
+
+    def __init__(self, gaussian_bump=False, gaussian_iou=0.7):
+        self.gaussian_bump = gaussian_bump
+        self.gaussian_iou = gaussian_iou
+
+    def assign(self, points, gt_bboxes, gt_labels=None):
+        """Assign gt to bboxes.
+
+        This method assign a gt bbox to every point, each bbox
+        will be assigned with  0, or a positive number.
+        0 means negative sample, positive number is the index (1-based) of
+        assigned gt.
+        The assignment is done in following steps, the order matters.
+
+        1. assign every points to 0
+        2. for each gt box, we find the k most closest points to the
+            box center and assign the gt bbox to those points, we also record
+            the minimum distance from each point to the closest gt box. When we
+            assign the bbox to the points, we check whether its distance to the
+            points is closest.
+
+        Args:
+            points (Tensor): points to be assigned, shape(n, 3) while last
+                dimension stands for (x, y, stride).
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`, e.g., crowd boxes in COCO.
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        INF = 1e8
+        num_gts, num_points = gt_bboxes.shape[0], points.shape[0]
+
+        if self.gaussian_bump:
+            dtype = torch.float32
+        else:
+            dtype = torch.long
+        if num_points == 0 or num_gts == 0:
+            assigned_gt_hm_tl = points.new_zeros((num_points,), dtype=dtype)
+            assigned_gt_hm_br = points.new_zeros((num_points,), dtype=dtype)
+            # stores the assigned gt dist (to this point) of each point
+            assigned_gt_offset_tl = points.new_zeros((num_points, 2), dtype=torch.float32)
+            assigned_gt_offset_br = points.new_zeros((num_points, 2), dtype=torch.float32)
+
+            pos_inds_tl = torch.nonzero(assigned_gt_hm_tl == 1, as_tuple=False).squeeze(-1).unique()
+            pos_inds_br = torch.nonzero(assigned_gt_hm_br == 1, as_tuple=False).squeeze(-1).unique()
+            neg_inds_tl = torch.nonzero(assigned_gt_hm_tl < 1, as_tuple=False).squeeze(-1).unique()
+            neg_inds_br = torch.nonzero(assigned_gt_hm_br < 1, as_tuple=False).squeeze(-1).unique()
+
+            return assigned_gt_hm_tl, assigned_gt_offset_tl, pos_inds_tl, neg_inds_tl, \
+                   assigned_gt_hm_br, assigned_gt_offset_br, pos_inds_br, neg_inds_br
+        points_range = torch.arange(num_points)
+        points_xy = points[:, :2]
+        points_stride = points[:, 2]
+        points_lvl = torch.log2(points_stride).int()  # [3...,4...,5...,6...,7...]
+        lvl_min, lvl_max = points_lvl.min(), points_lvl.max()
+
+        # assign gt box
+        gt_bboxes_xtl, gt_bboxes_ytl, gt_bboxes_xbr, gt_bboxes_ybr = torch.chunk(gt_bboxes, 4, dim=1)
+        gt_bboxes_xytl = torch.cat([gt_bboxes_xtl, gt_bboxes_ytl], -1)
+        gt_bboxes_xybr = torch.cat([gt_bboxes_xbr, gt_bboxes_ybr], -1)
+        if self.gaussian_bump:
+            gt_bboxes_w = gt_bboxes[:, 2] - gt_bboxes[:, 0]
+            gt_bboxes_h = gt_bboxes[:, 3] - gt_bboxes[:, 1]
+            radius = gaussian_radius((gt_bboxes_h, gt_bboxes_w), self.gaussian_iou)
+            diameter = 2 * radius + 1
+            sigma = diameter / 6
+        else:
+            radius = None
+
+        distances_tl = (points_xy[:, None, :] - gt_bboxes_xytl[None, :, :]).norm(dim=2)
+        distances_br = (points_xy[:, None, :] - gt_bboxes_xybr[None, :, :]).norm(dim=2)
+
+        # stores the assigned gt heatmap of each point
+        assigned_gt_hm_tl = points.new_zeros((num_points,), dtype=dtype)
+        assigned_gt_hm_br = points.new_zeros((num_points,), dtype=dtype)
+        # stores the assigned gt dist (to this point) of each point
+        assigned_gt_offset_tl = points.new_zeros((num_points, 2), dtype=torch.float32)
+        assigned_gt_offset_br = points.new_zeros((num_points, 2), dtype=torch.float32)
+
+        lvls = torch.arange(lvl_min, lvl_max + 1, dtype=points_lvl.dtype, device=points_lvl.device)
+        for gt_lvl in lvls:
+            lvl_idx = gt_lvl == points_lvl
+            points_index = points_range[lvl_idx]
+            lvl_points = points_xy[lvl_idx, :]
+
+            downscale_factor = torch.pow(2, gt_lvl)
+            lvl_distances_tl = distances_tl[lvl_idx, :]
+            lvl_distances_br = distances_br[lvl_idx, :]
+
+            _, min_dist_index_tl = lvl_distances_tl.min(dim=0)
+            min_dist_points_index_tl = points_index[min_dist_index_tl]
+
+            assigned_gt_offset_tl[min_dist_points_index_tl, :] = \
+                (gt_bboxes_xytl - lvl_points[min_dist_index_tl, :]) / downscale_factor
+
+            _, min_dist_index_br = lvl_distances_br.min(dim=0)
+            min_dist_points_index_br = points_index[min_dist_index_br]
+            assigned_gt_offset_br[min_dist_points_index_br, :] = \
+                (gt_bboxes_xybr - lvl_points[min_dist_index_br, :]) / downscale_factor
+            if self.gaussian_bump:
+                out_index_tl = lvl_distances_tl >= radius[None, :]
+                lvl_gaussian_tl = torch.exp(-torch.pow(lvl_distances_tl, 2) / (2 * sigma * sigma)[None, :])
+                lvl_gaussian_tl[out_index_tl] = -INF
+                max_gaussian_tl, _ = lvl_gaussian_tl.max(dim=1)
+                assigned_gt_hm_tl[points_index[max_gaussian_tl != -INF]] = max_gaussian_tl[max_gaussian_tl != -INF]
+
+                out_index_br = lvl_distances_br >= radius[None, :]
+                lvl_gaussian_br = torch.exp(-torch.pow(lvl_distances_br, 2) / (2 * sigma * sigma)[None, :])
+                lvl_gaussian_br[out_index_br] = -INF
+                max_gaussian_br, _ = lvl_gaussian_br.max(dim=1)
+                assigned_gt_hm_br[points_index[max_gaussian_br != -INF]] = max_gaussian_br[max_gaussian_br != -INF]
+            assigned_gt_hm_tl[min_dist_points_index_tl] = 1
+            assigned_gt_hm_br[min_dist_points_index_br] = 1
+
+        pos_inds_tl = torch.nonzero(assigned_gt_hm_tl == 1, as_tuple=False).squeeze(-1).unique()
+        pos_inds_br = torch.nonzero(assigned_gt_hm_br == 1, as_tuple=False).squeeze(-1).unique()
+        neg_inds_tl = torch.nonzero(assigned_gt_hm_tl < 1, as_tuple=False).squeeze(-1).unique()
+        neg_inds_br = torch.nonzero(assigned_gt_hm_br < 1, as_tuple=False).squeeze(-1).unique()
+
+        return assigned_gt_hm_tl, assigned_gt_offset_tl, pos_inds_tl, neg_inds_tl, \
+               assigned_gt_hm_br, assigned_gt_offset_br, pos_inds_br, neg_inds_br
+
+
+def gaussian_radius(det_size, min_overlap):
+    height, width = det_size
+
+    a1  = 1
+    b1  = (height + width)
+    c1  = width * height * (1 - min_overlap) / (1 + min_overlap)
+    sq1 = torch.sqrt(b1 ** 2 - 4 * a1 * c1)
+    r1  = (b1 - sq1) / (2 * a1)
+
+    a2  = 4
+    b2  = 2 * (height + width)
+    c2  = (1 - min_overlap) * width * height
+    sq2 = torch.sqrt(b2 ** 2 - 4 * a2 * c2)
+    r2  = (b2 - sq2) / (2 * a2)
+
+    a3  = 4 * min_overlap
+    b3  = -2 * min_overlap * (height + width)
+    c3  = (min_overlap - 1) * width * height
+    sq3 = torch.sqrt(b3 ** 2 - 4 * a3 * c3)
+    r3  = (b3 + sq3) / (2 * a3)
+
+    r = torch.stack([r1, r2, r3], dim=1)
+    return torch.min(r, dim=1)[0]
\ No newline at end of file
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/assigners/region_assigner.py b/python/autoCut/models/swin/mmdet/core/bbox/assigners/region_assigner.py
new file mode 100644
index 0000000..2e8464b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/assigners/region_assigner.py
@@ -0,0 +1,221 @@
+import torch
+
+from mmdet.core import anchor_inside_flags
+from ..builder import BBOX_ASSIGNERS
+from .assign_result import AssignResult
+from .base_assigner import BaseAssigner
+
+
+def calc_region(bbox, ratio, stride, featmap_size=None):
+    """Calculate region of the box defined by the ratio, the ratio is from the
+    center of the box to every edge."""
+    # project bbox on the feature
+    f_bbox = bbox / stride
+    x1 = torch.round((1 - ratio) * f_bbox[0] + ratio * f_bbox[2])
+    y1 = torch.round((1 - ratio) * f_bbox[1] + ratio * f_bbox[3])
+    x2 = torch.round(ratio * f_bbox[0] + (1 - ratio) * f_bbox[2])
+    y2 = torch.round(ratio * f_bbox[1] + (1 - ratio) * f_bbox[3])
+    if featmap_size is not None:
+        x1 = x1.clamp(min=0, max=featmap_size[1])
+        y1 = y1.clamp(min=0, max=featmap_size[0])
+        x2 = x2.clamp(min=0, max=featmap_size[1])
+        y2 = y2.clamp(min=0, max=featmap_size[0])
+    return (x1, y1, x2, y2)
+
+
+def anchor_ctr_inside_region_flags(anchors, stride, region):
+    """Get the flag indicate whether anchor centers are inside regions."""
+    x1, y1, x2, y2 = region
+    f_anchors = anchors / stride
+    x = (f_anchors[:, 0] + f_anchors[:, 2]) * 0.5
+    y = (f_anchors[:, 1] + f_anchors[:, 3]) * 0.5
+    flags = (x >= x1) & (x <= x2) & (y >= y1) & (y <= y2)
+    return flags
+
+
+@BBOX_ASSIGNERS.register_module()
+class RegionAssigner(BaseAssigner):
+    """Assign a corresponding gt bbox or background to each bbox.
+
+    Each proposals will be assigned with `-1`, `0`, or a positive integer
+    indicating the ground truth index.
+
+    - -1: don't care
+    - 0: negative sample, no assigned gt
+    - positive integer: positive sample, index (1-based) of assigned gt
+
+    Args:
+        center_ratio: ratio of the region in the center of the bbox to
+            define positive sample.
+        ignore_ratio: ratio of the region to define ignore samples.
+    """
+
+    def __init__(self, center_ratio=0.2, ignore_ratio=0.5):
+        self.center_ratio = center_ratio
+        self.ignore_ratio = ignore_ratio
+
+    def assign(self,
+               mlvl_anchors,
+               mlvl_valid_flags,
+               gt_bboxes,
+               img_meta,
+               featmap_sizes,
+               anchor_scale,
+               anchor_strides,
+               gt_bboxes_ignore=None,
+               gt_labels=None,
+               allowed_border=0):
+        """Assign gt to anchors.
+
+        This method assign a gt bbox to every bbox (proposal/anchor), each bbox
+        will be assigned with -1, 0, or a positive number. -1 means don't care,
+        0 means negative sample, positive number is the index (1-based) of
+        assigned gt.
+        The assignment is done in following steps, the order matters.
+
+        1. Assign every anchor to 0 (negative)
+        For each gt_bboxes:
+            2. Compute ignore flags based on ignore_region then
+                assign -1 to anchors w.r.t. ignore flags
+            3. Compute pos flags based on center_region then
+               assign gt_bboxes to anchors w.r.t. pos flags
+            4. Compute ignore flags based on adjacent anchor lvl then
+               assign -1 to anchors w.r.t. ignore flags
+            5. Assign anchor outside of image to -1
+
+        Args:
+            mlvl_anchors (list[Tensor]): Multi level anchors.
+            mlvl_valid_flags (list[Tensor]): Multi level valid flags.
+            gt_bboxes (Tensor): Ground truth bboxes of image
+            img_meta (dict): Meta info of image.
+            featmap_sizes (list[Tensor]): Feature mapsize each level
+            anchor_scale (int): Scale of the anchor.
+            anchor_strides (list[int]): Stride of the anchor.
+            gt_bboxes (Tensor): Groundtruth boxes, shape (k, 4).
+            gt_bboxes_ignore (Tensor, optional): Ground truth bboxes that are
+                labelled as `ignored`, e.g., crowd boxes in COCO.
+            gt_labels (Tensor, optional): Label of gt_bboxes, shape (k, ).
+            allowed_border (int, optional): The border to allow the valid
+                anchor. Defaults to 0.
+
+        Returns:
+            :obj:`AssignResult`: The assign result.
+        """
+        if gt_bboxes_ignore is not None:
+            raise NotImplementedError
+
+        num_gts = gt_bboxes.shape[0]
+        num_bboxes = sum(x.shape[0] for x in mlvl_anchors)
+
+        if num_gts == 0 or num_bboxes == 0:
+            # No ground truth or boxes, return empty assignment
+            max_overlaps = gt_bboxes.new_zeros((num_bboxes, ))
+            assigned_gt_inds = gt_bboxes.new_zeros((num_bboxes, ),
+                                                   dtype=torch.long)
+            if gt_labels is None:
+                assigned_labels = None
+            else:
+                assigned_labels = gt_bboxes.new_full((num_bboxes, ),
+                                                     -1,
+                                                     dtype=torch.long)
+            return AssignResult(
+                num_gts,
+                assigned_gt_inds,
+                max_overlaps,
+                labels=assigned_labels)
+
+        num_lvls = len(mlvl_anchors)
+        r1 = (1 - self.center_ratio) / 2
+        r2 = (1 - self.ignore_ratio) / 2
+
+        scale = torch.sqrt((gt_bboxes[:, 2] - gt_bboxes[:, 0]) *
+                           (gt_bboxes[:, 3] - gt_bboxes[:, 1]))
+        min_anchor_size = scale.new_full(
+            (1, ), float(anchor_scale * anchor_strides[0]))
+        target_lvls = torch.floor(
+            torch.log2(scale) - torch.log2(min_anchor_size) + 0.5)
+        target_lvls = target_lvls.clamp(min=0, max=num_lvls - 1).long()
+
+        # 1. assign 0 (negative) by default
+        mlvl_assigned_gt_inds = []
+        mlvl_ignore_flags = []
+        for lvl in range(num_lvls):
+            h, w = featmap_sizes[lvl]
+            assert h * w == mlvl_anchors[lvl].shape[0]
+            assigned_gt_inds = gt_bboxes.new_full((h * w, ),
+                                                  0,
+                                                  dtype=torch.long)
+            ignore_flags = torch.zeros_like(assigned_gt_inds)
+            mlvl_assigned_gt_inds.append(assigned_gt_inds)
+            mlvl_ignore_flags.append(ignore_flags)
+
+        for gt_id in range(num_gts):
+            lvl = target_lvls[gt_id].item()
+            featmap_size = featmap_sizes[lvl]
+            stride = anchor_strides[lvl]
+            anchors = mlvl_anchors[lvl]
+            gt_bbox = gt_bboxes[gt_id, :4]
+
+            # Compute regions
+            ignore_region = calc_region(gt_bbox, r2, stride, featmap_size)
+            ctr_region = calc_region(gt_bbox, r1, stride, featmap_size)
+
+            # 2. Assign -1 to ignore flags
+            ignore_flags = anchor_ctr_inside_region_flags(
+                anchors, stride, ignore_region)
+            mlvl_assigned_gt_inds[lvl][ignore_flags] = -1
+
+            # 3. Assign gt_bboxes to pos flags
+            pos_flags = anchor_ctr_inside_region_flags(anchors, stride,
+                                                       ctr_region)
+            mlvl_assigned_gt_inds[lvl][pos_flags] = gt_id + 1
+
+            # 4. Assign -1 to ignore adjacent lvl
+            if lvl > 0:
+                d_lvl = lvl - 1
+                d_anchors = mlvl_anchors[d_lvl]
+                d_featmap_size = featmap_sizes[d_lvl]
+                d_stride = anchor_strides[d_lvl]
+                d_ignore_region = calc_region(gt_bbox, r2, d_stride,
+                                              d_featmap_size)
+                ignore_flags = anchor_ctr_inside_region_flags(
+                    d_anchors, d_stride, d_ignore_region)
+                mlvl_ignore_flags[d_lvl][ignore_flags] = 1
+            if lvl < num_lvls - 1:
+                u_lvl = lvl + 1
+                u_anchors = mlvl_anchors[u_lvl]
+                u_featmap_size = featmap_sizes[u_lvl]
+                u_stride = anchor_strides[u_lvl]
+                u_ignore_region = calc_region(gt_bbox, r2, u_stride,
+                                              u_featmap_size)
+                ignore_flags = anchor_ctr_inside_region_flags(
+                    u_anchors, u_stride, u_ignore_region)
+                mlvl_ignore_flags[u_lvl][ignore_flags] = 1
+
+        # 4. (cont.) Assign -1 to ignore adjacent lvl
+        for lvl in range(num_lvls):
+            ignore_flags = mlvl_ignore_flags[lvl]
+            mlvl_assigned_gt_inds[lvl][ignore_flags] = -1
+
+        # 5. Assign -1 to anchor outside of image
+        flat_assigned_gt_inds = torch.cat(mlvl_assigned_gt_inds)
+        flat_anchors = torch.cat(mlvl_anchors)
+        flat_valid_flags = torch.cat(mlvl_valid_flags)
+        assert (flat_assigned_gt_inds.shape[0] == flat_anchors.shape[0] ==
+                flat_valid_flags.shape[0])
+        inside_flags = anchor_inside_flags(flat_anchors, flat_valid_flags,
+                                           img_meta['img_shape'],
+                                           allowed_border)
+        outside_flags = ~inside_flags
+        flat_assigned_gt_inds[outside_flags] = -1
+
+        if gt_labels is not None:
+            assigned_labels = torch.zeros_like(flat_assigned_gt_inds)
+            pos_flags = assigned_gt_inds > 0
+            assigned_labels[pos_flags] = gt_labels[
+                flat_assigned_gt_inds[pos_flags] - 1]
+        else:
+            assigned_labels = None
+
+        return AssignResult(
+            num_gts, flat_assigned_gt_inds, None, labels=assigned_labels)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/builder.py b/python/autoCut/models/swin/mmdet/core/bbox/builder.py
new file mode 100644
index 0000000..682683b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/builder.py
@@ -0,0 +1,20 @@
+from mmcv.utils import Registry, build_from_cfg
+
+BBOX_ASSIGNERS = Registry('bbox_assigner')
+BBOX_SAMPLERS = Registry('bbox_sampler')
+BBOX_CODERS = Registry('bbox_coder')
+
+
+def build_assigner(cfg, **default_args):
+    """Builder of box assigner."""
+    return build_from_cfg(cfg, BBOX_ASSIGNERS, default_args)
+
+
+def build_sampler(cfg, **default_args):
+    """Builder of box sampler."""
+    return build_from_cfg(cfg, BBOX_SAMPLERS, default_args)
+
+
+def build_bbox_coder(cfg, **default_args):
+    """Builder of box coder."""
+    return build_from_cfg(cfg, BBOX_CODERS, default_args)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/coder/__init__.py b/python/autoCut/models/swin/mmdet/core/bbox/coder/__init__.py
new file mode 100644
index 0000000..ae455ba
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/coder/__init__.py
@@ -0,0 +1,13 @@
+from .base_bbox_coder import BaseBBoxCoder
+from .bucketing_bbox_coder import BucketingBBoxCoder
+from .delta_xywh_bbox_coder import DeltaXYWHBBoxCoder
+from .legacy_delta_xywh_bbox_coder import LegacyDeltaXYWHBBoxCoder
+from .pseudo_bbox_coder import PseudoBBoxCoder
+from .tblr_bbox_coder import TBLRBBoxCoder
+from .yolo_bbox_coder import YOLOBBoxCoder
+
+__all__ = [
+    'BaseBBoxCoder', 'PseudoBBoxCoder', 'DeltaXYWHBBoxCoder',
+    'LegacyDeltaXYWHBBoxCoder', 'TBLRBBoxCoder', 'YOLOBBoxCoder',
+    'BucketingBBoxCoder'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/coder/base_bbox_coder.py b/python/autoCut/models/swin/mmdet/core/bbox/coder/base_bbox_coder.py
new file mode 100644
index 0000000..cf0b34c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/coder/base_bbox_coder.py
@@ -0,0 +1,17 @@
+from abc import ABCMeta, abstractmethod
+
+
+class BaseBBoxCoder(metaclass=ABCMeta):
+    """Base bounding box coder."""
+
+    def __init__(self, **kwargs):
+        pass
+
+    @abstractmethod
+    def encode(self, bboxes, gt_bboxes):
+        """Encode deltas between bboxes and ground truth boxes."""
+
+    @abstractmethod
+    def decode(self, bboxes, bboxes_pred):
+        """Decode the predicted bboxes according to prediction and base
+        boxes."""
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/coder/bucketing_bbox_coder.py b/python/autoCut/models/swin/mmdet/core/bbox/coder/bucketing_bbox_coder.py
new file mode 100644
index 0000000..92d24b4
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/coder/bucketing_bbox_coder.py
@@ -0,0 +1,350 @@
+import mmcv
+import numpy as np
+import torch
+import torch.nn.functional as F
+
+from ..builder import BBOX_CODERS
+from ..transforms import bbox_rescale
+from .base_bbox_coder import BaseBBoxCoder
+
+
+@BBOX_CODERS.register_module()
+class BucketingBBoxCoder(BaseBBoxCoder):
+    """Bucketing BBox Coder for Side-Aware Boundary Localization (SABL).
+
+    Boundary Localization with Bucketing and Bucketing Guided Rescoring
+    are implemented here.
+
+    Please refer to https://arxiv.org/abs/1912.04260 for more details.
+
+    Args:
+        num_buckets (int): Number of buckets.
+        scale_factor (int): Scale factor of proposals to generate buckets.
+        offset_topk (int): Topk buckets are used to generate
+             bucket fine regression targets. Defaults to 2.
+        offset_upperbound (float): Offset upperbound to generate
+             bucket fine regression targets.
+             To avoid too large offset displacements. Defaults to 1.0.
+        cls_ignore_neighbor (bool): Ignore second nearest bucket or Not.
+             Defaults to True.
+        clip_border (bool, optional): Whether clip the objects outside the
+            border of the image. Defaults to True.
+    """
+
+    def __init__(self,
+                 num_buckets,
+                 scale_factor,
+                 offset_topk=2,
+                 offset_upperbound=1.0,
+                 cls_ignore_neighbor=True,
+                 clip_border=True):
+        super(BucketingBBoxCoder, self).__init__()
+        self.num_buckets = num_buckets
+        self.scale_factor = scale_factor
+        self.offset_topk = offset_topk
+        self.offset_upperbound = offset_upperbound
+        self.cls_ignore_neighbor = cls_ignore_neighbor
+        self.clip_border = clip_border
+
+    def encode(self, bboxes, gt_bboxes):
+        """Get bucketing estimation and fine regression targets during
+        training.
+
+        Args:
+            bboxes (torch.Tensor): source boxes, e.g., object proposals.
+            gt_bboxes (torch.Tensor): target of the transformation, e.g.,
+                ground truth boxes.
+
+        Returns:
+           encoded_bboxes(tuple[Tensor]): bucketing estimation
+            and fine regression targets and weights
+        """
+
+        assert bboxes.size(0) == gt_bboxes.size(0)
+        assert bboxes.size(-1) == gt_bboxes.size(-1) == 4
+        encoded_bboxes = bbox2bucket(bboxes, gt_bboxes, self.num_buckets,
+                                     self.scale_factor, self.offset_topk,
+                                     self.offset_upperbound,
+                                     self.cls_ignore_neighbor)
+        return encoded_bboxes
+
+    def decode(self, bboxes, pred_bboxes, max_shape=None):
+        """Apply transformation `pred_bboxes` to `boxes`.
+        Args:
+            boxes (torch.Tensor): Basic boxes.
+            pred_bboxes (torch.Tensor): Predictions for bucketing estimation
+                and fine regression
+            max_shape (tuple[int], optional): Maximum shape of boxes.
+                Defaults to None.
+
+        Returns:
+            torch.Tensor: Decoded boxes.
+        """
+        assert len(pred_bboxes) == 2
+        cls_preds, offset_preds = pred_bboxes
+        assert cls_preds.size(0) == bboxes.size(0) and offset_preds.size(
+            0) == bboxes.size(0)
+        decoded_bboxes = bucket2bbox(bboxes, cls_preds, offset_preds,
+                                     self.num_buckets, self.scale_factor,
+                                     max_shape, self.clip_border)
+
+        return decoded_bboxes
+
+
+@mmcv.jit(coderize=True)
+def generat_buckets(proposals, num_buckets, scale_factor=1.0):
+    """Generate buckets w.r.t bucket number and scale factor of proposals.
+
+    Args:
+        proposals (Tensor): Shape (n, 4)
+        num_buckets (int): Number of buckets.
+        scale_factor (float): Scale factor to rescale proposals.
+
+    Returns:
+        tuple[Tensor]: (bucket_w, bucket_h, l_buckets, r_buckets,
+         t_buckets, d_buckets)
+
+            - bucket_w: Width of buckets on x-axis. Shape (n, ).
+            - bucket_h: Height of buckets on y-axis. Shape (n, ).
+            - l_buckets: Left buckets. Shape (n, ceil(side_num/2)).
+            - r_buckets: Right buckets. Shape (n, ceil(side_num/2)).
+            - t_buckets: Top buckets. Shape (n, ceil(side_num/2)).
+            - d_buckets: Down buckets. Shape (n, ceil(side_num/2)).
+    """
+    proposals = bbox_rescale(proposals, scale_factor)
+
+    # number of buckets in each side
+    side_num = int(np.ceil(num_buckets / 2.0))
+    pw = proposals[..., 2] - proposals[..., 0]
+    ph = proposals[..., 3] - proposals[..., 1]
+    px1 = proposals[..., 0]
+    py1 = proposals[..., 1]
+    px2 = proposals[..., 2]
+    py2 = proposals[..., 3]
+
+    bucket_w = pw / num_buckets
+    bucket_h = ph / num_buckets
+
+    # left buckets
+    l_buckets = px1[:, None] + (0.5 + torch.arange(
+        0, side_num).to(proposals).float())[None, :] * bucket_w[:, None]
+    # right buckets
+    r_buckets = px2[:, None] - (0.5 + torch.arange(
+        0, side_num).to(proposals).float())[None, :] * bucket_w[:, None]
+    # top buckets
+    t_buckets = py1[:, None] + (0.5 + torch.arange(
+        0, side_num).to(proposals).float())[None, :] * bucket_h[:, None]
+    # down buckets
+    d_buckets = py2[:, None] - (0.5 + torch.arange(
+        0, side_num).to(proposals).float())[None, :] * bucket_h[:, None]
+    return bucket_w, bucket_h, l_buckets, r_buckets, t_buckets, d_buckets
+
+
+@mmcv.jit(coderize=True)
+def bbox2bucket(proposals,
+                gt,
+                num_buckets,
+                scale_factor,
+                offset_topk=2,
+                offset_upperbound=1.0,
+                cls_ignore_neighbor=True):
+    """Generate buckets estimation and fine regression targets.
+
+    Args:
+        proposals (Tensor): Shape (n, 4)
+        gt (Tensor): Shape (n, 4)
+        num_buckets (int): Number of buckets.
+        scale_factor (float): Scale factor to rescale proposals.
+        offset_topk (int): Topk buckets are used to generate
+             bucket fine regression targets. Defaults to 2.
+        offset_upperbound (float): Offset allowance to generate
+             bucket fine regression targets.
+             To avoid too large offset displacements. Defaults to 1.0.
+        cls_ignore_neighbor (bool): Ignore second nearest bucket or Not.
+             Defaults to True.
+
+    Returns:
+        tuple[Tensor]: (offsets, offsets_weights, bucket_labels, cls_weights).
+
+            - offsets: Fine regression targets. \
+                Shape (n, num_buckets*2).
+            - offsets_weights: Fine regression weights. \
+                Shape (n, num_buckets*2).
+            - bucket_labels: Bucketing estimation labels. \
+                Shape (n, num_buckets*2).
+            - cls_weights: Bucketing estimation weights. \
+                Shape (n, num_buckets*2).
+    """
+    assert proposals.size() == gt.size()
+
+    # generate buckets
+    proposals = proposals.float()
+    gt = gt.float()
+    (bucket_w, bucket_h, l_buckets, r_buckets, t_buckets,
+     d_buckets) = generat_buckets(proposals, num_buckets, scale_factor)
+
+    gx1 = gt[..., 0]
+    gy1 = gt[..., 1]
+    gx2 = gt[..., 2]
+    gy2 = gt[..., 3]
+
+    # generate offset targets and weights
+    # offsets from buckets to gts
+    l_offsets = (l_buckets - gx1[:, None]) / bucket_w[:, None]
+    r_offsets = (r_buckets - gx2[:, None]) / bucket_w[:, None]
+    t_offsets = (t_buckets - gy1[:, None]) / bucket_h[:, None]
+    d_offsets = (d_buckets - gy2[:, None]) / bucket_h[:, None]
+
+    # select top-k nearset buckets
+    l_topk, l_label = l_offsets.abs().topk(
+        offset_topk, dim=1, largest=False, sorted=True)
+    r_topk, r_label = r_offsets.abs().topk(
+        offset_topk, dim=1, largest=False, sorted=True)
+    t_topk, t_label = t_offsets.abs().topk(
+        offset_topk, dim=1, largest=False, sorted=True)
+    d_topk, d_label = d_offsets.abs().topk(
+        offset_topk, dim=1, largest=False, sorted=True)
+
+    offset_l_weights = l_offsets.new_zeros(l_offsets.size())
+    offset_r_weights = r_offsets.new_zeros(r_offsets.size())
+    offset_t_weights = t_offsets.new_zeros(t_offsets.size())
+    offset_d_weights = d_offsets.new_zeros(d_offsets.size())
+    inds = torch.arange(0, proposals.size(0)).to(proposals).long()
+
+    # generate offset weights of top-k nearset buckets
+    for k in range(offset_topk):
+        if k >= 1:
+            offset_l_weights[inds, l_label[:,
+                                           k]] = (l_topk[:, k] <
+                                                  offset_upperbound).float()
+            offset_r_weights[inds, r_label[:,
+                                           k]] = (r_topk[:, k] <
+                                                  offset_upperbound).float()
+            offset_t_weights[inds, t_label[:,
+                                           k]] = (t_topk[:, k] <
+                                                  offset_upperbound).float()
+            offset_d_weights[inds, d_label[:,
+                                           k]] = (d_topk[:, k] <
+                                                  offset_upperbound).float()
+        else:
+            offset_l_weights[inds, l_label[:, k]] = 1.0
+            offset_r_weights[inds, r_label[:, k]] = 1.0
+            offset_t_weights[inds, t_label[:, k]] = 1.0
+            offset_d_weights[inds, d_label[:, k]] = 1.0
+
+    offsets = torch.cat([l_offsets, r_offsets, t_offsets, d_offsets], dim=-1)
+    offsets_weights = torch.cat([
+        offset_l_weights, offset_r_weights, offset_t_weights, offset_d_weights
+    ],
+                                dim=-1)
+
+    # generate bucket labels and weight
+    side_num = int(np.ceil(num_buckets / 2.0))
+    labels = torch.stack(
+        [l_label[:, 0], r_label[:, 0], t_label[:, 0], d_label[:, 0]], dim=-1)
+
+    batch_size = labels.size(0)
+    bucket_labels = F.one_hot(labels.view(-1), side_num).view(batch_size,
+                                                              -1).float()
+    bucket_cls_l_weights = (l_offsets.abs() < 1).float()
+    bucket_cls_r_weights = (r_offsets.abs() < 1).float()
+    bucket_cls_t_weights = (t_offsets.abs() < 1).float()
+    bucket_cls_d_weights = (d_offsets.abs() < 1).float()
+    bucket_cls_weights = torch.cat([
+        bucket_cls_l_weights, bucket_cls_r_weights, bucket_cls_t_weights,
+        bucket_cls_d_weights
+    ],
+                                   dim=-1)
+    # ignore second nearest buckets for cls if necessary
+    if cls_ignore_neighbor:
+        bucket_cls_weights = (~((bucket_cls_weights == 1) &
+                                (bucket_labels == 0))).float()
+    else:
+        bucket_cls_weights[:] = 1.0
+    return offsets, offsets_weights, bucket_labels, bucket_cls_weights
+
+
+@mmcv.jit(coderize=True)
+def bucket2bbox(proposals,
+                cls_preds,
+                offset_preds,
+                num_buckets,
+                scale_factor=1.0,
+                max_shape=None,
+                clip_border=True):
+    """Apply bucketing estimation (cls preds) and fine regression (offset
+    preds) to generate det bboxes.
+
+    Args:
+        proposals (Tensor): Boxes to be transformed. Shape (n, 4)
+        cls_preds (Tensor): bucketing estimation. Shape (n, num_buckets*2).
+        offset_preds (Tensor): fine regression. Shape (n, num_buckets*2).
+        num_buckets (int): Number of buckets.
+        scale_factor (float): Scale factor to rescale proposals.
+        max_shape (tuple[int, int]): Maximum bounds for boxes. specifies (H, W)
+        clip_border (bool, optional): Whether clip the objects outside the
+            border of the image. Defaults to True.
+
+    Returns:
+        tuple[Tensor]: (bboxes, loc_confidence).
+
+            - bboxes: predicted bboxes. Shape (n, 4)
+            - loc_confidence: localization confidence of predicted bboxes.
+                Shape (n,).
+    """
+
+    side_num = int(np.ceil(num_buckets / 2.0))
+    cls_preds = cls_preds.view(-1, side_num)
+    offset_preds = offset_preds.view(-1, side_num)
+
+    scores = F.softmax(cls_preds, dim=1)
+    score_topk, score_label = scores.topk(2, dim=1, largest=True, sorted=True)
+
+    rescaled_proposals = bbox_rescale(proposals, scale_factor)
+
+    pw = rescaled_proposals[..., 2] - rescaled_proposals[..., 0]
+    ph = rescaled_proposals[..., 3] - rescaled_proposals[..., 1]
+    px1 = rescaled_proposals[..., 0]
+    py1 = rescaled_proposals[..., 1]
+    px2 = rescaled_proposals[..., 2]
+    py2 = rescaled_proposals[..., 3]
+
+    bucket_w = pw / num_buckets
+    bucket_h = ph / num_buckets
+
+    score_inds_l = score_label[0::4, 0]
+    score_inds_r = score_label[1::4, 0]
+    score_inds_t = score_label[2::4, 0]
+    score_inds_d = score_label[3::4, 0]
+    l_buckets = px1 + (0.5 + score_inds_l.float()) * bucket_w
+    r_buckets = px2 - (0.5 + score_inds_r.float()) * bucket_w
+    t_buckets = py1 + (0.5 + score_inds_t.float()) * bucket_h
+    d_buckets = py2 - (0.5 + score_inds_d.float()) * bucket_h
+
+    offsets = offset_preds.view(-1, 4, side_num)
+    inds = torch.arange(proposals.size(0)).to(proposals).long()
+    l_offsets = offsets[:, 0, :][inds, score_inds_l]
+    r_offsets = offsets[:, 1, :][inds, score_inds_r]
+    t_offsets = offsets[:, 2, :][inds, score_inds_t]
+    d_offsets = offsets[:, 3, :][inds, score_inds_d]
+
+    x1 = l_buckets - l_offsets * bucket_w
+    x2 = r_buckets - r_offsets * bucket_w
+    y1 = t_buckets - t_offsets * bucket_h
+    y2 = d_buckets - d_offsets * bucket_h
+
+    if clip_border and max_shape is not None:
+        x1 = x1.clamp(min=0, max=max_shape[1] - 1)
+        y1 = y1.clamp(min=0, max=max_shape[0] - 1)
+        x2 = x2.clamp(min=0, max=max_shape[1] - 1)
+        y2 = y2.clamp(min=0, max=max_shape[0] - 1)
+    bboxes = torch.cat([x1[:, None], y1[:, None], x2[:, None], y2[:, None]],
+                       dim=-1)
+
+    # bucketing guided rescoring
+    loc_confidence = score_topk[:, 0]
+    top2_neighbor_inds = (score_label[:, 0] - score_label[:, 1]).abs() == 1
+    loc_confidence += score_topk[:, 1] * top2_neighbor_inds.float()
+    loc_confidence = loc_confidence.view(-1, 4).mean(dim=1)
+
+    return bboxes, loc_confidence
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/coder/delta_xywh_bbox_coder.py b/python/autoCut/models/swin/mmdet/core/bbox/coder/delta_xywh_bbox_coder.py
new file mode 100644
index 0000000..da31718
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/coder/delta_xywh_bbox_coder.py
@@ -0,0 +1,237 @@
+import mmcv
+import numpy as np
+import torch
+
+from ..builder import BBOX_CODERS
+from .base_bbox_coder import BaseBBoxCoder
+
+
+@BBOX_CODERS.register_module()
+class DeltaXYWHBBoxCoder(BaseBBoxCoder):
+    """Delta XYWH BBox coder.
+
+    Following the practice in `R-CNN <https://arxiv.org/abs/1311.2524>`_,
+    this coder encodes bbox (x1, y1, x2, y2) into delta (dx, dy, dw, dh) and
+    decodes delta (dx, dy, dw, dh) back to original bbox (x1, y1, x2, y2).
+
+    Args:
+        target_means (Sequence[float]): Denormalizing means of target for
+            delta coordinates
+        target_stds (Sequence[float]): Denormalizing standard deviation of
+            target for delta coordinates
+        clip_border (bool, optional): Whether clip the objects outside the
+            border of the image. Defaults to True.
+    """
+
+    def __init__(self,
+                 target_means=(0., 0., 0., 0.),
+                 target_stds=(1., 1., 1., 1.),
+                 clip_border=True):
+        super(BaseBBoxCoder, self).__init__()
+        self.means = target_means
+        self.stds = target_stds
+        self.clip_border = clip_border
+
+    def encode(self, bboxes, gt_bboxes):
+        """Get box regression transformation deltas that can be used to
+        transform the ``bboxes`` into the ``gt_bboxes``.
+
+        Args:
+            bboxes (torch.Tensor): Source boxes, e.g., object proposals.
+            gt_bboxes (torch.Tensor): Target of the transformation, e.g.,
+                ground-truth boxes.
+
+        Returns:
+            torch.Tensor: Box transformation deltas
+        """
+
+        assert bboxes.size(0) == gt_bboxes.size(0)
+        assert bboxes.size(-1) == gt_bboxes.size(-1) == 4
+        encoded_bboxes = bbox2delta(bboxes, gt_bboxes, self.means, self.stds)
+        return encoded_bboxes
+
+    def decode(self,
+               bboxes,
+               pred_bboxes,
+               max_shape=None,
+               wh_ratio_clip=16 / 1000):
+        """Apply transformation `pred_bboxes` to `boxes`.
+
+        Args:
+            bboxes (torch.Tensor): Basic boxes. Shape (B, N, 4) or (N, 4)
+            pred_bboxes (Tensor): Encoded offsets with respect to each roi.
+               Has shape (B, N, num_classes * 4) or (B, N, 4) or
+               (N, num_classes * 4) or (N, 4). Note N = num_anchors * W * H
+               when rois is a grid of anchors.Offset encoding follows [1]_.
+            max_shape (Sequence[int] or torch.Tensor or Sequence[
+               Sequence[int]],optional): Maximum bounds for boxes, specifies
+               (H, W, C) or (H, W). If bboxes shape is (B, N, 4), then
+               the max_shape should be a Sequence[Sequence[int]]
+               and the length of max_shape should also be B.
+            wh_ratio_clip (float, optional): The allowed ratio between
+                width and height.
+
+        Returns:
+            torch.Tensor: Decoded boxes.
+        """
+
+        assert pred_bboxes.size(0) == bboxes.size(0)
+        if pred_bboxes.ndim == 3:
+            assert pred_bboxes.size(1) == bboxes.size(1)
+        decoded_bboxes = delta2bbox(bboxes, pred_bboxes, self.means, self.stds,
+                                    max_shape, wh_ratio_clip, self.clip_border)
+
+        return decoded_bboxes
+
+
+@mmcv.jit(coderize=True)
+def bbox2delta(proposals, gt, means=(0., 0., 0., 0.), stds=(1., 1., 1., 1.)):
+    """Compute deltas of proposals w.r.t. gt.
+
+    We usually compute the deltas of x, y, w, h of proposals w.r.t ground
+    truth bboxes to get regression target.
+    This is the inverse function of :func:`delta2bbox`.
+
+    Args:
+        proposals (Tensor): Boxes to be transformed, shape (N, ..., 4)
+        gt (Tensor): Gt bboxes to be used as base, shape (N, ..., 4)
+        means (Sequence[float]): Denormalizing means for delta coordinates
+        stds (Sequence[float]): Denormalizing standard deviation for delta
+            coordinates
+
+    Returns:
+        Tensor: deltas with shape (N, 4), where columns represent dx, dy,
+            dw, dh.
+    """
+    assert proposals.size() == gt.size()
+
+    proposals = proposals.float()
+    gt = gt.float()
+    px = (proposals[..., 0] + proposals[..., 2]) * 0.5
+    py = (proposals[..., 1] + proposals[..., 3]) * 0.5
+    pw = proposals[..., 2] - proposals[..., 0]
+    ph = proposals[..., 3] - proposals[..., 1]
+
+    gx = (gt[..., 0] + gt[..., 2]) * 0.5
+    gy = (gt[..., 1] + gt[..., 3]) * 0.5
+    gw = gt[..., 2] - gt[..., 0]
+    gh = gt[..., 3] - gt[..., 1]
+
+    dx = (gx - px) / pw
+    dy = (gy - py) / ph
+    dw = torch.log(gw / pw)
+    dh = torch.log(gh / ph)
+    deltas = torch.stack([dx, dy, dw, dh], dim=-1)
+
+    means = deltas.new_tensor(means).unsqueeze(0)
+    stds = deltas.new_tensor(stds).unsqueeze(0)
+    deltas = deltas.sub_(means).div_(stds)
+
+    return deltas
+
+
+@mmcv.jit(coderize=True)
+def delta2bbox(rois,
+               deltas,
+               means=(0., 0., 0., 0.),
+               stds=(1., 1., 1., 1.),
+               max_shape=None,
+               wh_ratio_clip=16 / 1000,
+               clip_border=True):
+    """Apply deltas to shift/scale base boxes.
+
+    Typically the rois are anchor or proposed bounding boxes and the deltas are
+    network outputs used to shift/scale those boxes.
+    This is the inverse function of :func:`bbox2delta`.
+
+    Args:
+        rois (Tensor): Boxes to be transformed. Has shape (N, 4) or (B, N, 4)
+        deltas (Tensor): Encoded offsets with respect to each roi.
+            Has shape (B, N, num_classes * 4) or (B, N, 4) or
+            (N, num_classes * 4) or (N, 4). Note N = num_anchors * W * H
+            when rois is a grid of anchors.Offset encoding follows [1]_.
+        means (Sequence[float]): Denormalizing means for delta coordinates
+        stds (Sequence[float]): Denormalizing standard deviation for delta
+            coordinates
+        max_shape (Sequence[int] or torch.Tensor or Sequence[
+            Sequence[int]],optional): Maximum bounds for boxes, specifies
+            (H, W, C) or (H, W). If rois shape is (B, N, 4), then
+            the max_shape should be a Sequence[Sequence[int]]
+            and the length of max_shape should also be B.
+        wh_ratio_clip (float): Maximum aspect ratio for boxes.
+        clip_border (bool, optional): Whether clip the objects outside the
+            border of the image. Defaults to True.
+
+    Returns:
+        Tensor: Boxes with shape (B, N, num_classes * 4) or (B, N, 4) or
+           (N, num_classes * 4) or (N, 4), where 4 represent
+           tl_x, tl_y, br_x, br_y.
+
+    References:
+        .. [1] https://arxiv.org/abs/1311.2524
+
+    Example:
+        >>> rois = torch.Tensor([[ 0.,  0.,  1.,  1.],
+        >>>                      [ 0.,  0.,  1.,  1.],
+        >>>                      [ 0.,  0.,  1.,  1.],
+        >>>                      [ 5.,  5.,  5.,  5.]])
+        >>> deltas = torch.Tensor([[  0.,   0.,   0.,   0.],
+        >>>                        [  1.,   1.,   1.,   1.],
+        >>>                        [  0.,   0.,   2.,  -1.],
+        >>>                        [ 0.7, -1.9, -0.5,  0.3]])
+        >>> delta2bbox(rois, deltas, max_shape=(32, 32, 3))
+        tensor([[0.0000, 0.0000, 1.0000, 1.0000],
+                [0.1409, 0.1409, 2.8591, 2.8591],
+                [0.0000, 0.3161, 4.1945, 0.6839],
+                [5.0000, 5.0000, 5.0000, 5.0000]])
+    """
+    means = deltas.new_tensor(means).view(1,
+                                          -1).repeat(1,
+                                                     deltas.size(-1) // 4)
+    stds = deltas.new_tensor(stds).view(1, -1).repeat(1, deltas.size(-1) // 4)
+    denorm_deltas = deltas * stds + means
+    dx = denorm_deltas[..., 0::4]
+    dy = denorm_deltas[..., 1::4]
+    dw = denorm_deltas[..., 2::4]
+    dh = denorm_deltas[..., 3::4]
+    max_ratio = np.abs(np.log(wh_ratio_clip))
+    dw = dw.clamp(min=-max_ratio, max=max_ratio)
+    dh = dh.clamp(min=-max_ratio, max=max_ratio)
+    x1, y1 = rois[..., 0], rois[..., 1]
+    x2, y2 = rois[..., 2], rois[..., 3]
+    # Compute center of each roi
+    px = ((x1 + x2) * 0.5).unsqueeze(-1).expand_as(dx)
+    py = ((y1 + y2) * 0.5).unsqueeze(-1).expand_as(dy)
+    # Compute width/height of each roi
+    pw = (x2 - x1).unsqueeze(-1).expand_as(dw)
+    ph = (y2 - y1).unsqueeze(-1).expand_as(dh)
+    # Use exp(network energy) to enlarge/shrink each roi
+    gw = pw * dw.exp()
+    gh = ph * dh.exp()
+    # Use network energy to shift the center of each roi
+    gx = px + pw * dx
+    gy = py + ph * dy
+    # Convert center-xy/width/height to top-left, bottom-right
+    x1 = gx - gw * 0.5
+    y1 = gy - gh * 0.5
+    x2 = gx + gw * 0.5
+    y2 = gy + gh * 0.5
+
+    bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view(deltas.size())
+
+    if clip_border and max_shape is not None:
+        if not isinstance(max_shape, torch.Tensor):
+            max_shape = x1.new_tensor(max_shape)
+        max_shape = max_shape[..., :2].type_as(x1)
+        if max_shape.ndim == 2:
+            assert bboxes.ndim == 3
+            assert max_shape.size(0) == bboxes.size(0)
+
+        min_xy = x1.new_tensor(0)
+        max_xy = torch.cat(
+            [max_shape] * (deltas.size(-1) // 2),
+            dim=-1).flip(-1).unsqueeze(-2)
+        bboxes = torch.where(bboxes < min_xy, min_xy, bboxes)
+        bboxes = torch.where(bboxes > max_xy, max_xy, bboxes)
+
+    return bboxes
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/coder/legacy_delta_xywh_bbox_coder.py b/python/autoCut/models/swin/mmdet/core/bbox/coder/legacy_delta_xywh_bbox_coder.py
new file mode 100644
index 0000000..190309f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/coder/legacy_delta_xywh_bbox_coder.py
@@ -0,0 +1,215 @@
+import mmcv
+import numpy as np
+import torch
+
+from ..builder import BBOX_CODERS
+from .base_bbox_coder import BaseBBoxCoder
+
+
+@BBOX_CODERS.register_module()
+class LegacyDeltaXYWHBBoxCoder(BaseBBoxCoder):
+    """Legacy Delta XYWH BBox coder used in MMDet V1.x.
+
+    Following the practice in R-CNN [1]_, this coder encodes bbox (x1, y1, x2,
+    y2) into delta (dx, dy, dw, dh) and decodes delta (dx, dy, dw, dh)
+    back to original bbox (x1, y1, x2, y2).
+
+    Note:
+        The main difference between :class`LegacyDeltaXYWHBBoxCoder` and
+        :class:`DeltaXYWHBBoxCoder` is whether ``+ 1`` is used during width and
+        height calculation. We suggest to only use this coder when testing with
+        MMDet V1.x models.
+
+    References:
+        .. [1] https://arxiv.org/abs/1311.2524
+
+    Args:
+        target_means (Sequence[float]): denormalizing means of target for
+            delta coordinates
+        target_stds (Sequence[float]): denormalizing standard deviation of
+            target for delta coordinates
+    """
+
+    def __init__(self,
+                 target_means=(0., 0., 0., 0.),
+                 target_stds=(1., 1., 1., 1.)):
+        super(BaseBBoxCoder, self).__init__()
+        self.means = target_means
+        self.stds = target_stds
+
+    def encode(self, bboxes, gt_bboxes):
+        """Get box regression transformation deltas that can be used to
+        transform the ``bboxes`` into the ``gt_bboxes``.
+
+        Args:
+            bboxes (torch.Tensor): source boxes, e.g., object proposals.
+            gt_bboxes (torch.Tensor): target of the transformation, e.g.,
+                ground-truth boxes.
+
+        Returns:
+            torch.Tensor: Box transformation deltas
+        """
+        assert bboxes.size(0) == gt_bboxes.size(0)
+        assert bboxes.size(-1) == gt_bboxes.size(-1) == 4
+        encoded_bboxes = legacy_bbox2delta(bboxes, gt_bboxes, self.means,
+                                           self.stds)
+        return encoded_bboxes
+
+    def decode(self,
+               bboxes,
+               pred_bboxes,
+               max_shape=None,
+               wh_ratio_clip=16 / 1000):
+        """Apply transformation `pred_bboxes` to `boxes`.
+
+        Args:
+            boxes (torch.Tensor): Basic boxes.
+            pred_bboxes (torch.Tensor): Encoded boxes with shape
+            max_shape (tuple[int], optional): Maximum shape of boxes.
+                Defaults to None.
+            wh_ratio_clip (float, optional): The allowed ratio between
+                width and height.
+
+        Returns:
+            torch.Tensor: Decoded boxes.
+        """
+        assert pred_bboxes.size(0) == bboxes.size(0)
+        decoded_bboxes = legacy_delta2bbox(bboxes, pred_bboxes, self.means,
+                                           self.stds, max_shape, wh_ratio_clip)
+
+        return decoded_bboxes
+
+
+@mmcv.jit(coderize=True)
+def legacy_bbox2delta(proposals,
+                      gt,
+                      means=(0., 0., 0., 0.),
+                      stds=(1., 1., 1., 1.)):
+    """Compute deltas of proposals w.r.t. gt in the MMDet V1.x manner.
+
+    We usually compute the deltas of x, y, w, h of proposals w.r.t ground
+    truth bboxes to get regression target.
+    This is the inverse function of `delta2bbox()`
+
+    Args:
+        proposals (Tensor): Boxes to be transformed, shape (N, ..., 4)
+        gt (Tensor): Gt bboxes to be used as base, shape (N, ..., 4)
+        means (Sequence[float]): Denormalizing means for delta coordinates
+        stds (Sequence[float]): Denormalizing standard deviation for delta
+            coordinates
+
+    Returns:
+        Tensor: deltas with shape (N, 4), where columns represent dx, dy,
+            dw, dh.
+    """
+    assert proposals.size() == gt.size()
+
+    proposals = proposals.float()
+    gt = gt.float()
+    px = (proposals[..., 0] + proposals[..., 2]) * 0.5
+    py = (proposals[..., 1] + proposals[..., 3]) * 0.5
+    pw = proposals[..., 2] - proposals[..., 0] + 1.0
+    ph = proposals[..., 3] - proposals[..., 1] + 1.0
+
+    gx = (gt[..., 0] + gt[..., 2]) * 0.5
+    gy = (gt[..., 1] + gt[..., 3]) * 0.5
+    gw = gt[..., 2] - gt[..., 0] + 1.0
+    gh = gt[..., 3] - gt[..., 1] + 1.0
+
+    dx = (gx - px) / pw
+    dy = (gy - py) / ph
+    dw = torch.log(gw / pw)
+    dh = torch.log(gh / ph)
+    deltas = torch.stack([dx, dy, dw, dh], dim=-1)
+
+    means = deltas.new_tensor(means).unsqueeze(0)
+    stds = deltas.new_tensor(stds).unsqueeze(0)
+    deltas = deltas.sub_(means).div_(stds)
+
+    return deltas
+
+
+@mmcv.jit(coderize=True)
+def legacy_delta2bbox(rois,
+                      deltas,
+                      means=(0., 0., 0., 0.),
+                      stds=(1., 1., 1., 1.),
+                      max_shape=None,
+                      wh_ratio_clip=16 / 1000):
+    """Apply deltas to shift/scale base boxes in the MMDet V1.x manner.
+
+    Typically the rois are anchor or proposed bounding boxes and the deltas are
+    network outputs used to shift/scale those boxes.
+    This is the inverse function of `bbox2delta()`
+
+    Args:
+        rois (Tensor): Boxes to be transformed. Has shape (N, 4)
+        deltas (Tensor): Encoded offsets with respect to each roi.
+            Has shape (N, 4 * num_classes). Note N = num_anchors * W * H when
+            rois is a grid of anchors. Offset encoding follows [1]_.
+        means (Sequence[float]): Denormalizing means for delta coordinates
+        stds (Sequence[float]): Denormalizing standard deviation for delta
+            coordinates
+        max_shape (tuple[int, int]): Maximum bounds for boxes. specifies (H, W)
+        wh_ratio_clip (float): Maximum aspect ratio for boxes.
+
+    Returns:
+        Tensor: Boxes with shape (N, 4), where columns represent
+            tl_x, tl_y, br_x, br_y.
+
+    References:
+        .. [1] https://arxiv.org/abs/1311.2524
+
+    Example:
+        >>> rois = torch.Tensor([[ 0.,  0.,  1.,  1.],
+        >>>                      [ 0.,  0.,  1.,  1.],
+        >>>                      [ 0.,  0.,  1.,  1.],
+        >>>                      [ 5.,  5.,  5.,  5.]])
+        >>> deltas = torch.Tensor([[  0.,   0.,   0.,   0.],
+        >>>                        [  1.,   1.,   1.,   1.],
+        >>>                        [  0.,   0.,   2.,  -1.],
+        >>>                        [ 0.7, -1.9, -0.5,  0.3]])
+        >>> legacy_delta2bbox(rois, deltas, max_shape=(32, 32))
+        tensor([[0.0000, 0.0000, 1.5000, 1.5000],
+                [0.0000, 0.0000, 5.2183, 5.2183],
+                [0.0000, 0.1321, 7.8891, 0.8679],
+                [5.3967, 2.4251, 6.0033, 3.7749]])
+    """
+    means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
+    stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
+    denorm_deltas = deltas * stds + means
+    dx = denorm_deltas[:, 0::4]
+    dy = denorm_deltas[:, 1::4]
+    dw = denorm_deltas[:, 2::4]
+    dh = denorm_deltas[:, 3::4]
+    max_ratio = np.abs(np.log(wh_ratio_clip))
+    dw = dw.clamp(min=-max_ratio, max=max_ratio)
+    dh = dh.clamp(min=-max_ratio, max=max_ratio)
+    # Compute center of each roi
+    px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
+    py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
+    # Compute width/height of each roi
+    pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
+    ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
+    # Use exp(network energy) to enlarge/shrink each roi
+    gw = pw * dw.exp()
+    gh = ph * dh.exp()
+    # Use network energy to shift the center of each roi
+    gx = px + pw * dx
+    gy = py + ph * dy
+    # Convert center-xy/width/height to top-left, bottom-right
+
+    # The true legacy box coder should +- 0.5 here.
+    # However, current implementation improves the performance when testing
+    # the models trained in MMDetection 1.X (~0.5 bbox AP, 0.2 mask AP)
+    x1 = gx - gw * 0.5
+    y1 = gy - gh * 0.5
+    x2 = gx + gw * 0.5
+    y2 = gy + gh * 0.5
+    if max_shape is not None:
+        x1 = x1.clamp(min=0, max=max_shape[1] - 1)
+        y1 = y1.clamp(min=0, max=max_shape[0] - 1)
+        x2 = x2.clamp(min=0, max=max_shape[1] - 1)
+        y2 = y2.clamp(min=0, max=max_shape[0] - 1)
+    bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
+    return bboxes
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/coder/pseudo_bbox_coder.py b/python/autoCut/models/swin/mmdet/core/bbox/coder/pseudo_bbox_coder.py
new file mode 100644
index 0000000..1c8346f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/coder/pseudo_bbox_coder.py
@@ -0,0 +1,18 @@
+from ..builder import BBOX_CODERS
+from .base_bbox_coder import BaseBBoxCoder
+
+
+@BBOX_CODERS.register_module()
+class PseudoBBoxCoder(BaseBBoxCoder):
+    """Pseudo bounding box coder."""
+
+    def __init__(self, **kwargs):
+        super(BaseBBoxCoder, self).__init__(**kwargs)
+
+    def encode(self, bboxes, gt_bboxes):
+        """torch.Tensor: return the given ``bboxes``"""
+        return gt_bboxes
+
+    def decode(self, bboxes, pred_bboxes):
+        """torch.Tensor: return the given ``pred_bboxes``"""
+        return pred_bboxes
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/coder/tblr_bbox_coder.py b/python/autoCut/models/swin/mmdet/core/bbox/coder/tblr_bbox_coder.py
new file mode 100644
index 0000000..edaffaf
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/coder/tblr_bbox_coder.py
@@ -0,0 +1,198 @@
+import mmcv
+import torch
+
+from ..builder import BBOX_CODERS
+from .base_bbox_coder import BaseBBoxCoder
+
+
+@BBOX_CODERS.register_module()
+class TBLRBBoxCoder(BaseBBoxCoder):
+    """TBLR BBox coder.
+
+    Following the practice in `FSAF <https://arxiv.org/abs/1903.00621>`_,
+    this coder encodes gt bboxes (x1, y1, x2, y2) into (top, bottom, left,
+    right) and decode it back to the original.
+
+    Args:
+        normalizer (list | float): Normalization factor to be
+          divided with when coding the coordinates. If it is a list, it should
+          have length of 4 indicating normalization factor in tblr dims.
+          Otherwise it is a unified float factor for all dims. Default: 4.0
+        clip_border (bool, optional): Whether clip the objects outside the
+            border of the image. Defaults to True.
+    """
+
+    def __init__(self, normalizer=4.0, clip_border=True):
+        super(BaseBBoxCoder, self).__init__()
+        self.normalizer = normalizer
+        self.clip_border = clip_border
+
+    def encode(self, bboxes, gt_bboxes):
+        """Get box regression transformation deltas that can be used to
+        transform the ``bboxes`` into the ``gt_bboxes`` in the (top, left,
+        bottom, right) order.
+
+        Args:
+            bboxes (torch.Tensor): source boxes, e.g., object proposals.
+            gt_bboxes (torch.Tensor): target of the transformation, e.g.,
+                ground truth boxes.
+
+        Returns:
+            torch.Tensor: Box transformation deltas
+        """
+        assert bboxes.size(0) == gt_bboxes.size(0)
+        assert bboxes.size(-1) == gt_bboxes.size(-1) == 4
+        encoded_bboxes = bboxes2tblr(
+            bboxes, gt_bboxes, normalizer=self.normalizer)
+        return encoded_bboxes
+
+    def decode(self, bboxes, pred_bboxes, max_shape=None):
+        """Apply transformation `pred_bboxes` to `boxes`.
+
+        Args:
+            bboxes (torch.Tensor): Basic boxes.Shape (B, N, 4) or (N, 4)
+            pred_bboxes (torch.Tensor): Encoded boxes with shape
+               (B, N, 4) or (N, 4)
+            max_shape (Sequence[int] or torch.Tensor or Sequence[
+               Sequence[int]],optional): Maximum bounds for boxes, specifies
+               (H, W, C) or (H, W). If bboxes shape is (B, N, 4), then
+               the max_shape should be a Sequence[Sequence[int]]
+               and the length of max_shape should also be B.
+
+        Returns:
+            torch.Tensor: Decoded boxes.
+        """
+        decoded_bboxes = tblr2bboxes(
+            bboxes,
+            pred_bboxes,
+            normalizer=self.normalizer,
+            max_shape=max_shape,
+            clip_border=self.clip_border)
+
+        return decoded_bboxes
+
+
+@mmcv.jit(coderize=True)
+def bboxes2tblr(priors, gts, normalizer=4.0, normalize_by_wh=True):
+    """Encode ground truth boxes to tblr coordinate.
+
+    It first convert the gt coordinate to tblr format,
+     (top, bottom, left, right), relative to prior box centers.
+     The tblr coordinate may be normalized by the side length of prior bboxes
+     if `normalize_by_wh` is specified as True, and it is then normalized by
+     the `normalizer` factor.
+
+    Args:
+        priors (Tensor): Prior boxes in point form
+            Shape: (num_proposals,4).
+        gts (Tensor): Coords of ground truth for each prior in point-form
+            Shape: (num_proposals, 4).
+        normalizer (Sequence[float] | float): normalization parameter of
+            encoded boxes. If it is a list, it has to have length = 4.
+            Default: 4.0
+        normalize_by_wh (bool): Whether to normalize tblr coordinate by the
+            side length (wh) of prior bboxes.
+
+    Return:
+        encoded boxes (Tensor), Shape: (num_proposals, 4)
+    """
+
+    # dist b/t match center and prior's center
+    if not isinstance(normalizer, float):
+        normalizer = torch.tensor(normalizer, device=priors.device)
+        assert len(normalizer) == 4, 'Normalizer must have length = 4'
+    assert priors.size(0) == gts.size(0)
+    prior_centers = (priors[:, 0:2] + priors[:, 2:4]) / 2
+    xmin, ymin, xmax, ymax = gts.split(1, dim=1)
+    top = prior_centers[:, 1].unsqueeze(1) - ymin
+    bottom = ymax - prior_centers[:, 1].unsqueeze(1)
+    left = prior_centers[:, 0].unsqueeze(1) - xmin
+    right = xmax - prior_centers[:, 0].unsqueeze(1)
+    loc = torch.cat((top, bottom, left, right), dim=1)
+    if normalize_by_wh:
+        # Normalize tblr by anchor width and height
+        wh = priors[:, 2:4] - priors[:, 0:2]
+        w, h = torch.split(wh, 1, dim=1)
+        loc[:, :2] /= h  # tb is normalized by h
+        loc[:, 2:] /= w  # lr is normalized by w
+    # Normalize tblr by the given normalization factor
+    return loc / normalizer
+
+
+@mmcv.jit(coderize=True)
+def tblr2bboxes(priors,
+                tblr,
+                normalizer=4.0,
+                normalize_by_wh=True,
+                max_shape=None,
+                clip_border=True):
+    """Decode tblr outputs to prediction boxes.
+
+    The process includes 3 steps: 1) De-normalize tblr coordinates by
+    multiplying it with `normalizer`; 2) De-normalize tblr coordinates by the
+    prior bbox width and height if `normalize_by_wh` is `True`; 3) Convert
+    tblr (top, bottom, left, right) pair relative to the center of priors back
+    to (xmin, ymin, xmax, ymax) coordinate.
+
+    Args:
+        priors (Tensor): Prior boxes in point form (x0, y0, x1, y1)
+          Shape: (N,4) or (B, N, 4).
+        tblr (Tensor): Coords of network output in tblr form
+          Shape: (N, 4) or (B, N, 4).
+        normalizer (Sequence[float] | float): Normalization parameter of
+          encoded boxes. By list, it represents the normalization factors at
+          tblr dims. By float, it is the unified normalization factor at all
+          dims. Default: 4.0
+        normalize_by_wh (bool): Whether the tblr coordinates have been
+          normalized by the side length (wh) of prior bboxes.
+        max_shape (Sequence[int] or torch.Tensor or Sequence[
+            Sequence[int]],optional): Maximum bounds for boxes, specifies
+            (H, W, C) or (H, W). If priors shape is (B, N, 4), then
+            the max_shape should be a Sequence[Sequence[int]]
+            and the length of max_shape should also be B.
+        clip_border (bool, optional): Whether clip the objects outside the
+            border of the image. Defaults to True.
+
+    Return:
+        encoded boxes (Tensor): Boxes with shape (N, 4) or (B, N, 4)
+    """
+    if not isinstance(normalizer, float):
+        normalizer = torch.tensor(normalizer, device=priors.device)
+        assert len(normalizer) == 4, 'Normalizer must have length = 4'
+    assert priors.size(0) == tblr.size(0)
+    if priors.ndim == 3:
+        assert priors.size(1) == tblr.size(1)
+
+    loc_decode = tblr * normalizer
+    prior_centers = (priors[..., 0:2] + priors[..., 2:4]) / 2
+    if normalize_by_wh:
+        wh = priors[..., 2:4] - priors[..., 0:2]
+        w, h = torch.split(wh, 1, dim=-1)
+        # Inplace operation with slice would failed for exporting to ONNX
+        th = h * loc_decode[..., :2]  # tb
+        tw = w * loc_decode[..., 2:]  # lr
+        loc_decode = torch.cat([th, tw], dim=-1)
+    # Cannot be exported using onnx when loc_decode.split(1, dim=-1)
+    top, bottom, left, right = loc_decode.split((1, 1, 1, 1), dim=-1)
+    xmin = prior_centers[..., 0].unsqueeze(-1) - left
+    xmax = prior_centers[..., 0].unsqueeze(-1) + right
+    ymin = prior_centers[..., 1].unsqueeze(-1) - top
+    ymax = prior_centers[..., 1].unsqueeze(-1) + bottom
+
+    bboxes = torch.cat((xmin, ymin, xmax, ymax), dim=-1)
+
+    if clip_border and max_shape is not None:
+        if not isinstance(max_shape, torch.Tensor):
+            max_shape = priors.new_tensor(max_shape)
+        max_shape = max_shape[..., :2].type_as(priors)
+        if max_shape.ndim == 2:
+            assert bboxes.ndim == 3
+            assert max_shape.size(0) == bboxes.size(0)
+
+        min_xy = priors.new_tensor(0)
+        max_xy = torch.cat([max_shape, max_shape],
+                           dim=-1).flip(-1).unsqueeze(-2)
+        bboxes = torch.where(bboxes < min_xy, min_xy, bboxes)
+        bboxes = torch.where(bboxes > max_xy, max_xy, bboxes)
+
+    return bboxes
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/coder/yolo_bbox_coder.py b/python/autoCut/models/swin/mmdet/core/bbox/coder/yolo_bbox_coder.py
new file mode 100644
index 0000000..d6d0e82
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/coder/yolo_bbox_coder.py
@@ -0,0 +1,89 @@
+import mmcv
+import torch
+
+from ..builder import BBOX_CODERS
+from .base_bbox_coder import BaseBBoxCoder
+
+
+@BBOX_CODERS.register_module()
+class YOLOBBoxCoder(BaseBBoxCoder):
+    """YOLO BBox coder.
+
+    Following `YOLO <https://arxiv.org/abs/1506.02640>`_, this coder divide
+    image into grids, and encode bbox (x1, y1, x2, y2) into (cx, cy, dw, dh).
+    cx, cy in [0., 1.], denotes relative center position w.r.t the center of
+    bboxes. dw, dh are the same as :obj:`DeltaXYWHBBoxCoder`.
+
+    Args:
+        eps (float): Min value of cx, cy when encoding.
+    """
+
+    def __init__(self, eps=1e-6):
+        super(BaseBBoxCoder, self).__init__()
+        self.eps = eps
+
+    @mmcv.jit(coderize=True)
+    def encode(self, bboxes, gt_bboxes, stride):
+        """Get box regression transformation deltas that can be used to
+        transform the ``bboxes`` into the ``gt_bboxes``.
+
+        Args:
+            bboxes (torch.Tensor): Source boxes, e.g., anchors.
+            gt_bboxes (torch.Tensor): Target of the transformation, e.g.,
+                ground-truth boxes.
+            stride (torch.Tensor | int): Stride of bboxes.
+
+        Returns:
+            torch.Tensor: Box transformation deltas
+        """
+
+        assert bboxes.size(0) == gt_bboxes.size(0)
+        assert bboxes.size(-1) == gt_bboxes.size(-1) == 4
+        x_center_gt = (gt_bboxes[..., 0] + gt_bboxes[..., 2]) * 0.5
+        y_center_gt = (gt_bboxes[..., 1] + gt_bboxes[..., 3]) * 0.5
+        w_gt = gt_bboxes[..., 2] - gt_bboxes[..., 0]
+        h_gt = gt_bboxes[..., 3] - gt_bboxes[..., 1]
+        x_center = (bboxes[..., 0] + bboxes[..., 2]) * 0.5
+        y_center = (bboxes[..., 1] + bboxes[..., 3]) * 0.5
+        w = bboxes[..., 2] - bboxes[..., 0]
+        h = bboxes[..., 3] - bboxes[..., 1]
+        w_target = torch.log((w_gt / w).clamp(min=self.eps))
+        h_target = torch.log((h_gt / h).clamp(min=self.eps))
+        x_center_target = ((x_center_gt - x_center) / stride + 0.5).clamp(
+            self.eps, 1 - self.eps)
+        y_center_target = ((y_center_gt - y_center) / stride + 0.5).clamp(
+            self.eps, 1 - self.eps)
+        encoded_bboxes = torch.stack(
+            [x_center_target, y_center_target, w_target, h_target], dim=-1)
+        return encoded_bboxes
+
+    @mmcv.jit(coderize=True)
+    def decode(self, bboxes, pred_bboxes, stride):
+        """Apply transformation `pred_bboxes` to `boxes`.
+
+        Args:
+            boxes (torch.Tensor): Basic boxes, e.g. anchors.
+            pred_bboxes (torch.Tensor): Encoded boxes with shape
+            stride (torch.Tensor | int): Strides of bboxes.
+
+        Returns:
+            torch.Tensor: Decoded boxes.
+        """
+        assert pred_bboxes.size(0) == bboxes.size(0)
+        assert pred_bboxes.size(-1) == bboxes.size(-1) == 4
+        x_center = (bboxes[..., 0] + bboxes[..., 2]) * 0.5
+        y_center = (bboxes[..., 1] + bboxes[..., 3]) * 0.5
+        w = bboxes[..., 2] - bboxes[..., 0]
+        h = bboxes[..., 3] - bboxes[..., 1]
+        # Get outputs x, y
+        x_center_pred = (pred_bboxes[..., 0] - 0.5) * stride + x_center
+        y_center_pred = (pred_bboxes[..., 1] - 0.5) * stride + y_center
+        w_pred = torch.exp(pred_bboxes[..., 2]) * w
+        h_pred = torch.exp(pred_bboxes[..., 3]) * h
+
+        decoded_bboxes = torch.stack(
+            (x_center_pred - w_pred / 2, y_center_pred - h_pred / 2,
+             x_center_pred + w_pred / 2, y_center_pred + h_pred / 2),
+            dim=-1)
+
+        return decoded_bboxes
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/demodata.py b/python/autoCut/models/swin/mmdet/core/bbox/demodata.py
new file mode 100644
index 0000000..feecb69
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/demodata.py
@@ -0,0 +1,41 @@
+import numpy as np
+import torch
+
+from mmdet.utils.util_random import ensure_rng
+
+
+def random_boxes(num=1, scale=1, rng=None):
+    """Simple version of ``kwimage.Boxes.random``
+
+    Returns:
+        Tensor: shape (n, 4) in x1, y1, x2, y2 format.
+
+    References:
+        https://gitlab.kitware.com/computer-vision/kwimage/blob/master/kwimage/structs/boxes.py#L1390
+
+    Example:
+        >>> num = 3
+        >>> scale = 512
+        >>> rng = 0
+        >>> boxes = random_boxes(num, scale, rng)
+        >>> print(boxes)
+        tensor([[280.9925, 278.9802, 308.6148, 366.1769],
+                [216.9113, 330.6978, 224.0446, 456.5878],
+                [405.3632, 196.3221, 493.3953, 270.7942]])
+    """
+    rng = ensure_rng(rng)
+
+    tlbr = rng.rand(num, 4).astype(np.float32)
+
+    tl_x = np.minimum(tlbr[:, 0], tlbr[:, 2])
+    tl_y = np.minimum(tlbr[:, 1], tlbr[:, 3])
+    br_x = np.maximum(tlbr[:, 0], tlbr[:, 2])
+    br_y = np.maximum(tlbr[:, 1], tlbr[:, 3])
+
+    tlbr[:, 0] = tl_x * scale
+    tlbr[:, 1] = tl_y * scale
+    tlbr[:, 2] = br_x * scale
+    tlbr[:, 3] = br_y * scale
+
+    boxes = torch.from_numpy(tlbr)
+    return boxes
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/__init__.py b/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/__init__.py
new file mode 100644
index 0000000..e71369a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/__init__.py
@@ -0,0 +1,4 @@
+from .builder import build_iou_calculator
+from .iou2d_calculator import BboxOverlaps2D, bbox_overlaps
+
+__all__ = ['build_iou_calculator', 'BboxOverlaps2D', 'bbox_overlaps']
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/builder.py b/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/builder.py
new file mode 100644
index 0000000..09094d7
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/builder.py
@@ -0,0 +1,8 @@
+from mmcv.utils import Registry, build_from_cfg
+
+IOU_CALCULATORS = Registry('IoU calculator')
+
+
+def build_iou_calculator(cfg, default_args=None):
+    """Builder of IoU calculator."""
+    return build_from_cfg(cfg, IOU_CALCULATORS, default_args)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/iou2d_calculator.py b/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/iou2d_calculator.py
new file mode 100644
index 0000000..158b702
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/iou_calculators/iou2d_calculator.py
@@ -0,0 +1,159 @@
+import torch
+
+from .builder import IOU_CALCULATORS
+
+
+@IOU_CALCULATORS.register_module()
+class BboxOverlaps2D(object):
+    """2D Overlaps (e.g. IoUs, GIoUs) Calculator."""
+
+    def __call__(self, bboxes1, bboxes2, mode='iou', is_aligned=False):
+        """Calculate IoU between 2D bboxes.
+
+        Args:
+            bboxes1 (Tensor): bboxes have shape (m, 4) in <x1, y1, x2, y2>
+                format, or shape (m, 5) in <x1, y1, x2, y2, score> format.
+            bboxes2 (Tensor): bboxes have shape (m, 4) in <x1, y1, x2, y2>
+                format, shape (m, 5) in <x1, y1, x2, y2, score> format, or be
+                empty. If ``is_aligned `` is ``True``, then m and n must be
+                equal.
+            mode (str): "iou" (intersection over union), "iof" (intersection
+                over foreground), or "giou" (generalized intersection over
+                union).
+            is_aligned (bool, optional): If True, then m and n must be equal.
+                Default False.
+
+        Returns:
+            Tensor: shape (m, n) if ``is_aligned `` is False else shape (m,)
+        """
+        assert bboxes1.size(-1) in [0, 4, 5]
+        assert bboxes2.size(-1) in [0, 4, 5]
+        if bboxes2.size(-1) == 5:
+            bboxes2 = bboxes2[..., :4]
+        if bboxes1.size(-1) == 5:
+            bboxes1 = bboxes1[..., :4]
+        return bbox_overlaps(bboxes1, bboxes2, mode, is_aligned)
+
+    def __repr__(self):
+        """str: a string describing the module"""
+        repr_str = self.__class__.__name__ + '()'
+        return repr_str
+
+
+def bbox_overlaps(bboxes1, bboxes2, mode='iou', is_aligned=False, eps=1e-6):
+    """Calculate overlap between two set of bboxes.
+
+    If ``is_aligned `` is ``False``, then calculate the overlaps between each
+    bbox of bboxes1 and bboxes2, otherwise the overlaps between each aligned
+    pair of bboxes1 and bboxes2.
+
+    Args:
+        bboxes1 (Tensor): shape (B, m, 4) in <x1, y1, x2, y2> format or empty.
+        bboxes2 (Tensor): shape (B, n, 4) in <x1, y1, x2, y2> format or empty.
+            B indicates the batch dim, in shape (B1, B2, ..., Bn).
+            If ``is_aligned `` is ``True``, then m and n must be equal.
+        mode (str): "iou" (intersection over union), "iof" (intersection over
+            foreground) or "giou" (generalized intersection over union).
+            Default "iou".
+        is_aligned (bool, optional): If True, then m and n must be equal.
+            Default False.
+        eps (float, optional): A value added to the denominator for numerical
+            stability. Default 1e-6.
+
+    Returns:
+        Tensor: shape (m, n) if ``is_aligned `` is False else shape (m,)
+
+    Example:
+        >>> bboxes1 = torch.FloatTensor([
+        >>>     [0, 0, 10, 10],
+        >>>     [10, 10, 20, 20],
+        >>>     [32, 32, 38, 42],
+        >>> ])
+        >>> bboxes2 = torch.FloatTensor([
+        >>>     [0, 0, 10, 20],
+        >>>     [0, 10, 10, 19],
+        >>>     [10, 10, 20, 20],
+        >>> ])
+        >>> overlaps = bbox_overlaps(bboxes1, bboxes2)
+        >>> assert overlaps.shape == (3, 3)
+        >>> overlaps = bbox_overlaps(bboxes1, bboxes2, is_aligned=True)
+        >>> assert overlaps.shape == (3, )
+
+    Example:
+        >>> empty = torch.empty(0, 4)
+        >>> nonempty = torch.FloatTensor([[0, 0, 10, 9]])
+        >>> assert tuple(bbox_overlaps(empty, nonempty).shape) == (0, 1)
+        >>> assert tuple(bbox_overlaps(nonempty, empty).shape) == (1, 0)
+        >>> assert tuple(bbox_overlaps(empty, empty).shape) == (0, 0)
+    """
+
+    assert mode in ['iou', 'iof', 'giou'], f'Unsupported mode {mode}'
+    # Either the boxes are empty or the length of boxes' last dimension is 4
+    assert (bboxes1.size(-1) == 4 or bboxes1.size(0) == 0)
+    assert (bboxes2.size(-1) == 4 or bboxes2.size(0) == 0)
+
+    # Batch dim must be the same
+    # Batch dim: (B1, B2, ... Bn)
+    assert bboxes1.shape[:-2] == bboxes2.shape[:-2]
+    batch_shape = bboxes1.shape[:-2]
+
+    rows = bboxes1.size(-2)
+    cols = bboxes2.size(-2)
+    if is_aligned:
+        assert rows == cols
+
+    if rows * cols == 0:
+        if is_aligned:
+            return bboxes1.new(batch_shape + (rows, ))
+        else:
+            return bboxes1.new(batch_shape + (rows, cols))
+
+    area1 = (bboxes1[..., 2] - bboxes1[..., 0]) * (
+        bboxes1[..., 3] - bboxes1[..., 1])
+    area2 = (bboxes2[..., 2] - bboxes2[..., 0]) * (
+        bboxes2[..., 3] - bboxes2[..., 1])
+
+    if is_aligned:
+        lt = torch.max(bboxes1[..., :2], bboxes2[..., :2])  # [B, rows, 2]
+        rb = torch.min(bboxes1[..., 2:], bboxes2[..., 2:])  # [B, rows, 2]
+
+        wh = (rb - lt).clamp(min=0)  # [B, rows, 2]
+        overlap = wh[..., 0] * wh[..., 1]
+
+        if mode in ['iou', 'giou']:
+            union = area1 + area2 - overlap
+        else:
+            union = area1
+        if mode == 'giou':
+            enclosed_lt = torch.min(bboxes1[..., :2], bboxes2[..., :2])
+            enclosed_rb = torch.max(bboxes1[..., 2:], bboxes2[..., 2:])
+    else:
+        lt = torch.max(bboxes1[..., :, None, :2],
+                       bboxes2[..., None, :, :2])  # [B, rows, cols, 2]
+        rb = torch.min(bboxes1[..., :, None, 2:],
+                       bboxes2[..., None, :, 2:])  # [B, rows, cols, 2]
+
+        wh = (rb - lt).clamp(min=0)  # [B, rows, cols, 2]
+        overlap = wh[..., 0] * wh[..., 1]
+
+        if mode in ['iou', 'giou']:
+            union = area1[..., None] + area2[..., None, :] - overlap
+        else:
+            union = area1[..., None]
+        if mode == 'giou':
+            enclosed_lt = torch.min(bboxes1[..., :, None, :2],
+                                    bboxes2[..., None, :, :2])
+            enclosed_rb = torch.max(bboxes1[..., :, None, 2:],
+                                    bboxes2[..., None, :, 2:])
+
+    eps = union.new_tensor([eps])
+    union = torch.max(union, eps)
+    ious = overlap / union
+    if mode in ['iou', 'iof']:
+        return ious
+    # calculate gious
+    enclose_wh = (enclosed_rb - enclosed_lt).clamp(min=0)
+    enclose_area = enclose_wh[..., 0] * enclose_wh[..., 1]
+    enclose_area = torch.max(enclose_area, eps)
+    gious = ious - (enclose_area - union) / enclose_area
+    return gious
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/match_costs/__init__.py b/python/autoCut/models/swin/mmdet/core/bbox/match_costs/__init__.py
new file mode 100644
index 0000000..add5e0d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/match_costs/__init__.py
@@ -0,0 +1,7 @@
+from .builder import build_match_cost
+from .match_cost import BBoxL1Cost, ClassificationCost, FocalLossCost, IoUCost
+
+__all__ = [
+    'build_match_cost', 'ClassificationCost', 'BBoxL1Cost', 'IoUCost',
+    'FocalLossCost'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/match_costs/builder.py b/python/autoCut/models/swin/mmdet/core/bbox/match_costs/builder.py
new file mode 100644
index 0000000..6894017
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/match_costs/builder.py
@@ -0,0 +1,8 @@
+from mmcv.utils import Registry, build_from_cfg
+
+MATCH_COST = Registry('Match Cost')
+
+
+def build_match_cost(cfg, default_args=None):
+    """Builder of IoU calculator."""
+    return build_from_cfg(cfg, MATCH_COST, default_args)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/match_costs/match_cost.py b/python/autoCut/models/swin/mmdet/core/bbox/match_costs/match_cost.py
new file mode 100644
index 0000000..3886973
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/match_costs/match_cost.py
@@ -0,0 +1,184 @@
+import torch
+
+from mmdet.core.bbox.iou_calculators import bbox_overlaps
+from mmdet.core.bbox.transforms import bbox_cxcywh_to_xyxy, bbox_xyxy_to_cxcywh
+from .builder import MATCH_COST
+
+
+@MATCH_COST.register_module()
+class BBoxL1Cost(object):
+    """BBoxL1Cost.
+
+     Args:
+         weight (int | float, optional): loss_weight
+         box_format (str, optional): 'xyxy' for DETR, 'xywh' for Sparse_RCNN
+
+     Examples:
+         >>> from mmdet.core.bbox.match_costs.match_cost import BBoxL1Cost
+         >>> import torch
+         >>> self = BBoxL1Cost()
+         >>> bbox_pred = torch.rand(1, 4)
+         >>> gt_bboxes= torch.FloatTensor([[0, 0, 2, 4], [1, 2, 3, 4]])
+         >>> factor = torch.tensor([10, 8, 10, 8])
+         >>> self(bbox_pred, gt_bboxes, factor)
+         tensor([[1.6172, 1.6422]])
+    """
+
+    def __init__(self, weight=1., box_format='xyxy'):
+        self.weight = weight
+        assert box_format in ['xyxy', 'xywh']
+        self.box_format = box_format
+
+    def __call__(self, bbox_pred, gt_bboxes):
+        """
+        Args:
+            bbox_pred (Tensor): Predicted boxes with normalized coordinates
+                (cx, cy, w, h), which are all in range [0, 1]. Shape
+                [num_query, 4].
+            gt_bboxes (Tensor): Ground truth boxes with normalized
+                coordinates (x1, y1, x2, y2). Shape [num_gt, 4].
+
+        Returns:
+            torch.Tensor: bbox_cost value with weight
+        """
+        if self.box_format == 'xywh':
+            gt_bboxes = bbox_xyxy_to_cxcywh(gt_bboxes)
+        elif self.box_format == 'xyxy':
+            bbox_pred = bbox_cxcywh_to_xyxy(bbox_pred)
+        bbox_cost = torch.cdist(bbox_pred, gt_bboxes, p=1)
+        return bbox_cost * self.weight
+
+
+@MATCH_COST.register_module()
+class FocalLossCost(object):
+    """FocalLossCost.
+
+     Args:
+         weight (int | float, optional): loss_weight
+         alpha (int | float, optional): focal_loss alpha
+         gamma (int | float, optional): focal_loss gamma
+         eps (float, optional): default 1e-12
+
+     Examples:
+         >>> from mmdet.core.bbox.match_costs.match_cost import FocalLossCost
+         >>> import torch
+         >>> self = FocalLossCost()
+         >>> cls_pred = torch.rand(4, 3)
+         >>> gt_labels = torch.tensor([0, 1, 2])
+         >>> factor = torch.tensor([10, 8, 10, 8])
+         >>> self(cls_pred, gt_labels)
+         tensor([[-0.3236, -0.3364, -0.2699],
+                [-0.3439, -0.3209, -0.4807],
+                [-0.4099, -0.3795, -0.2929],
+                [-0.1950, -0.1207, -0.2626]])
+    """
+
+    def __init__(self, weight=1., alpha=0.25, gamma=2, eps=1e-12):
+        self.weight = weight
+        self.alpha = alpha
+        self.gamma = gamma
+        self.eps = eps
+
+    def __call__(self, cls_pred, gt_labels):
+        """
+        Args:
+            cls_pred (Tensor): Predicted classification logits, shape
+                [num_query, num_class].
+            gt_labels (Tensor): Label of `gt_bboxes`, shape (num_gt,).
+
+        Returns:
+            torch.Tensor: cls_cost value with weight
+        """
+        cls_pred = cls_pred.sigmoid()
+        neg_cost = -(1 - cls_pred + self.eps).log() * (
+            1 - self.alpha) * cls_pred.pow(self.gamma)
+        pos_cost = -(cls_pred + self.eps).log() * self.alpha * (
+            1 - cls_pred).pow(self.gamma)
+        cls_cost = pos_cost[:, gt_labels] - neg_cost[:, gt_labels]
+        return cls_cost * self.weight
+
+
+@MATCH_COST.register_module()
+class ClassificationCost(object):
+    """ClsSoftmaxCost.
+
+     Args:
+         weight (int | float, optional): loss_weight
+
+     Examples:
+         >>> from mmdet.core.bbox.match_costs.match_cost import \
+         ... ClassificationCost
+         >>> import torch
+         >>> self = ClassificationCost()
+         >>> cls_pred = torch.rand(4, 3)
+         >>> gt_labels = torch.tensor([0, 1, 2])
+         >>> factor = torch.tensor([10, 8, 10, 8])
+         >>> self(cls_pred, gt_labels)
+         tensor([[-0.3430, -0.3525, -0.3045],
+                [-0.3077, -0.2931, -0.3992],
+                [-0.3664, -0.3455, -0.2881],
+                [-0.3343, -0.2701, -0.3956]])
+    """
+
+    def __init__(self, weight=1.):
+        self.weight = weight
+
+    def __call__(self, cls_pred, gt_labels):
+        """
+        Args:
+            cls_pred (Tensor): Predicted classification logits, shape
+                [num_query, num_class].
+            gt_labels (Tensor): Label of `gt_bboxes`, shape (num_gt,).
+
+        Returns:
+            torch.Tensor: cls_cost value with weight
+        """
+        # Following the official DETR repo, contrary to the loss that
+        # NLL is used, we approximate it in 1 - cls_score[gt_label].
+        # The 1 is a constant that doesn't change the matching,
+        # so it can be omitted.
+        cls_score = cls_pred.softmax(-1)
+        cls_cost = -cls_score[:, gt_labels]
+        return cls_cost * self.weight
+
+
+@MATCH_COST.register_module()
+class IoUCost(object):
+    """IoUCost.
+
+     Args:
+         iou_mode (str, optional): iou mode such as 'iou' | 'giou'
+         weight (int | float, optional): loss weight
+
+     Examples:
+         >>> from mmdet.core.bbox.match_costs.match_cost import IoUCost
+         >>> import torch
+         >>> self = IoUCost()
+         >>> bboxes = torch.FloatTensor([[1,1, 2, 2], [2, 2, 3, 4]])
+         >>> gt_bboxes = torch.FloatTensor([[0, 0, 2, 4], [1, 2, 3, 4]])
+         >>> self(bboxes, gt_bboxes)
+         tensor([[-0.1250,  0.1667],
+                [ 0.1667, -0.5000]])
+    """
+
+    def __init__(self, iou_mode='giou', weight=1.):
+        self.weight = weight
+        self.iou_mode = iou_mode
+
+    def __call__(self, bboxes, gt_bboxes):
+        """
+        Args:
+            bboxes (Tensor): Predicted boxes with unnormalized coordinates
+                (x1, y1, x2, y2). Shape [num_query, 4].
+            gt_bboxes (Tensor): Ground truth boxes with unnormalized
+                coordinates (x1, y1, x2, y2). Shape [num_gt, 4].
+
+        Returns:
+            torch.Tensor: iou_cost value with weight
+        """
+        # overlaps: [num_bboxes, num_gt]
+        overlaps = bbox_overlaps(
+            bboxes, gt_bboxes, mode=self.iou_mode, is_aligned=False)
+        # The 1 is a constant that doesn't change the matching, so omitted.
+        iou_cost = -overlaps
+        return iou_cost * self.weight
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/__init__.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/__init__.py
new file mode 100644
index 0000000..0b06303
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/__init__.py
@@ -0,0 +1,15 @@
+from .base_sampler import BaseSampler
+from .combined_sampler import CombinedSampler
+from .instance_balanced_pos_sampler import InstanceBalancedPosSampler
+from .iou_balanced_neg_sampler import IoUBalancedNegSampler
+from .ohem_sampler import OHEMSampler
+from .pseudo_sampler import PseudoSampler
+from .random_sampler import RandomSampler
+from .sampling_result import SamplingResult
+from .score_hlr_sampler import ScoreHLRSampler
+
+__all__ = [
+    'BaseSampler', 'PseudoSampler', 'RandomSampler',
+    'InstanceBalancedPosSampler', 'IoUBalancedNegSampler', 'CombinedSampler',
+    'OHEMSampler', 'SamplingResult', 'ScoreHLRSampler'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/base_sampler.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/base_sampler.py
new file mode 100644
index 0000000..9ea35de
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/base_sampler.py
@@ -0,0 +1,101 @@
+from abc import ABCMeta, abstractmethod
+
+import torch
+
+from .sampling_result import SamplingResult
+
+
+class BaseSampler(metaclass=ABCMeta):
+    """Base class of samplers."""
+
+    def __init__(self,
+                 num,
+                 pos_fraction,
+                 neg_pos_ub=-1,
+                 add_gt_as_proposals=True,
+                 **kwargs):
+        self.num = num
+        self.pos_fraction = pos_fraction
+        self.neg_pos_ub = neg_pos_ub
+        self.add_gt_as_proposals = add_gt_as_proposals
+        self.pos_sampler = self
+        self.neg_sampler = self
+
+    @abstractmethod
+    def _sample_pos(self, assign_result, num_expected, **kwargs):
+        """Sample positive samples."""
+        pass
+
+    @abstractmethod
+    def _sample_neg(self, assign_result, num_expected, **kwargs):
+        """Sample negative samples."""
+        pass
+
+    def sample(self,
+               assign_result,
+               bboxes,
+               gt_bboxes,
+               gt_labels=None,
+               **kwargs):
+        """Sample positive and negative bboxes.
+
+        This is a simple implementation of bbox sampling given candidates,
+        assigning results and ground truth bboxes.
+
+        Args:
+            assign_result (:obj:`AssignResult`): Bbox assigning results.
+            bboxes (Tensor): Boxes to be sampled from.
+            gt_bboxes (Tensor): Ground truth bboxes.
+            gt_labels (Tensor, optional): Class labels of ground truth bboxes.
+
+        Returns:
+            :obj:`SamplingResult`: Sampling result.
+
+        Example:
+            >>> from mmdet.core.bbox import RandomSampler
+            >>> from mmdet.core.bbox import AssignResult
+            >>> from mmdet.core.bbox.demodata import ensure_rng, random_boxes
+            >>> rng = ensure_rng(None)
+            >>> assign_result = AssignResult.random(rng=rng)
+            >>> bboxes = random_boxes(assign_result.num_preds, rng=rng)
+            >>> gt_bboxes = random_boxes(assign_result.num_gts, rng=rng)
+            >>> gt_labels = None
+            >>> self = RandomSampler(num=32, pos_fraction=0.5, neg_pos_ub=-1,
+            >>>                      add_gt_as_proposals=False)
+            >>> self = self.sample(assign_result, bboxes, gt_bboxes, gt_labels)
+        """
+        if len(bboxes.shape) < 2:
+            bboxes = bboxes[None, :]
+
+        bboxes = bboxes[:, :4]
+
+        gt_flags = bboxes.new_zeros((bboxes.shape[0], ), dtype=torch.uint8)
+        if self.add_gt_as_proposals and len(gt_bboxes) > 0:
+            if gt_labels is None:
+                raise ValueError(
+                    'gt_labels must be given when add_gt_as_proposals is True')
+            bboxes = torch.cat([gt_bboxes, bboxes], dim=0)
+            assign_result.add_gt_(gt_labels)
+            gt_ones = bboxes.new_ones(gt_bboxes.shape[0], dtype=torch.uint8)
+            gt_flags = torch.cat([gt_ones, gt_flags])
+
+        num_expected_pos = int(self.num * self.pos_fraction)
+        pos_inds = self.pos_sampler._sample_pos(
+            assign_result, num_expected_pos, bboxes=bboxes, **kwargs)
+        # We found that sampled indices have duplicated items occasionally.
+        # (may be a bug of PyTorch)
+        pos_inds = pos_inds.unique()
+        num_sampled_pos = pos_inds.numel()
+        num_expected_neg = self.num - num_sampled_pos
+        if self.neg_pos_ub >= 0:
+            _pos = max(1, num_sampled_pos)
+            neg_upper_bound = int(self.neg_pos_ub * _pos)
+            if num_expected_neg > neg_upper_bound:
+                num_expected_neg = neg_upper_bound
+        neg_inds = self.neg_sampler._sample_neg(
+            assign_result, num_expected_neg, bboxes=bboxes, **kwargs)
+        neg_inds = neg_inds.unique()
+
+        sampling_result = SamplingResult(pos_inds, neg_inds, bboxes, gt_bboxes,
+                                         assign_result, gt_flags)
+        return sampling_result
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/combined_sampler.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/combined_sampler.py
new file mode 100644
index 0000000..564729f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/combined_sampler.py
@@ -0,0 +1,20 @@
+from ..builder import BBOX_SAMPLERS, build_sampler
+from .base_sampler import BaseSampler
+
+
+@BBOX_SAMPLERS.register_module()
+class CombinedSampler(BaseSampler):
+    """A sampler that combines positive sampler and negative sampler."""
+
+    def __init__(self, pos_sampler, neg_sampler, **kwargs):
+        super(CombinedSampler, self).__init__(**kwargs)
+        self.pos_sampler = build_sampler(pos_sampler, **kwargs)
+        self.neg_sampler = build_sampler(neg_sampler, **kwargs)
+
+    def _sample_pos(self, **kwargs):
+        """Sample positive samples."""
+        raise NotImplementedError
+
+    def _sample_neg(self, **kwargs):
+        """Sample negative samples."""
+        raise NotImplementedError
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/instance_balanced_pos_sampler.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/instance_balanced_pos_sampler.py
new file mode 100644
index 0000000..c735298
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/instance_balanced_pos_sampler.py
@@ -0,0 +1,55 @@
+import numpy as np
+import torch
+
+from ..builder import BBOX_SAMPLERS
+from .random_sampler import RandomSampler
+
+
+@BBOX_SAMPLERS.register_module()
+class InstanceBalancedPosSampler(RandomSampler):
+    """Instance balanced sampler that samples equal number of positive samples
+    for each instance."""
+
+    def _sample_pos(self, assign_result, num_expected, **kwargs):
+        """Sample positive boxes.
+
+        Args:
+            assign_result (:obj:`AssignResult`): The assigned results of boxes.
+            num_expected (int): The number of expected positive samples
+
+        Returns:
+            Tensor or ndarray: sampled indices.
+        """
+        pos_inds = torch.nonzero(assign_result.gt_inds > 0, as_tuple=False)
+        if pos_inds.numel() != 0:
+            pos_inds = pos_inds.squeeze(1)
+        if pos_inds.numel() <= num_expected:
+            return pos_inds
+        else:
+            unique_gt_inds = assign_result.gt_inds[pos_inds].unique()
+            num_gts = len(unique_gt_inds)
+            num_per_gt = int(round(num_expected / float(num_gts)) + 1)
+            sampled_inds = []
+            for i in unique_gt_inds:
+                inds = torch.nonzero(
+                    assign_result.gt_inds == i.item(), as_tuple=False)
+                if inds.numel() != 0:
+                    inds = inds.squeeze(1)
+                else:
+                    continue
+                if len(inds) > num_per_gt:
+                    inds = self.random_choice(inds, num_per_gt)
+                sampled_inds.append(inds)
+            sampled_inds = torch.cat(sampled_inds)
+            if len(sampled_inds) < num_expected:
+                num_extra = num_expected - len(sampled_inds)
+                extra_inds = np.array(
+                    list(set(pos_inds.cpu()) - set(sampled_inds.cpu())))
+                if len(extra_inds) > num_extra:
+                    extra_inds = self.random_choice(extra_inds, num_extra)
+                extra_inds = torch.from_numpy(extra_inds).to(
+                    assign_result.gt_inds.device).long()
+                sampled_inds = torch.cat([sampled_inds, extra_inds])
+            elif len(sampled_inds) > num_expected:
+                sampled_inds = self.random_choice(sampled_inds, num_expected)
+            return sampled_inds
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/iou_balanced_neg_sampler.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/iou_balanced_neg_sampler.py
new file mode 100644
index 0000000..f275e43
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/iou_balanced_neg_sampler.py
@@ -0,0 +1,157 @@
+import numpy as np
+import torch
+
+from ..builder import BBOX_SAMPLERS
+from .random_sampler import RandomSampler
+
+
+@BBOX_SAMPLERS.register_module()
+class IoUBalancedNegSampler(RandomSampler):
+    """IoU Balanced Sampling.
+
+    arXiv: https://arxiv.org/pdf/1904.02701.pdf (CVPR 2019)
+
+    Sampling proposals according to their IoU. `floor_fraction` of needed RoIs
+    are sampled from proposals whose IoU are lower than `floor_thr` randomly.
+    The others are sampled from proposals whose IoU are higher than
+    `floor_thr`. These proposals are sampled from some bins evenly, which are
+    split by `num_bins` via IoU evenly.
+
+    Args:
+        num (int): number of proposals.
+        pos_fraction (float): fraction of positive proposals.
+        floor_thr (float): threshold (minimum) IoU for IoU balanced sampling,
+            set to -1 if all using IoU balanced sampling.
+        floor_fraction (float): sampling fraction of proposals under floor_thr.
+        num_bins (int): number of bins in IoU balanced sampling.
+    """
+
+    def __init__(self,
+                 num,
+                 pos_fraction,
+                 floor_thr=-1,
+                 floor_fraction=0,
+                 num_bins=3,
+                 **kwargs):
+        super(IoUBalancedNegSampler, self).__init__(num, pos_fraction,
+                                                    **kwargs)
+        assert floor_thr >= 0 or floor_thr == -1
+        assert 0 <= floor_fraction <= 1
+        assert num_bins >= 1
+
+        self.floor_thr = floor_thr
+        self.floor_fraction = floor_fraction
+        self.num_bins = num_bins
+
+    def sample_via_interval(self, max_overlaps, full_set, num_expected):
+        """Sample according to the iou interval.
+
+        Args:
+            max_overlaps (torch.Tensor): IoU between bounding boxes and ground
+                truth boxes.
+            full_set (set(int)): A full set of indices of boxes。
+            num_expected (int): Number of expected samples。
+
+        Returns:
+            np.ndarray: Indices  of samples
+        """
+        max_iou = max_overlaps.max()
+        iou_interval = (max_iou - self.floor_thr) / self.num_bins
+        per_num_expected = int(num_expected / self.num_bins)
+
+        sampled_inds = []
+        for i in range(self.num_bins):
+            start_iou = self.floor_thr + i * iou_interval
+            end_iou = self.floor_thr + (i + 1) * iou_interval
+            tmp_set = set(
+                np.where(
+                    np.logical_and(max_overlaps >= start_iou,
+                                   max_overlaps < end_iou))[0])
+            tmp_inds = list(tmp_set & full_set)
+            if len(tmp_inds) > per_num_expected:
+                tmp_sampled_set = self.random_choice(tmp_inds,
+                                                     per_num_expected)
+            else:
+                tmp_sampled_set = np.array(tmp_inds, dtype=np.int)
+            sampled_inds.append(tmp_sampled_set)
+
+        sampled_inds = np.concatenate(sampled_inds)
+        if len(sampled_inds) < num_expected:
+            num_extra = num_expected - len(sampled_inds)
+            extra_inds = np.array(list(full_set - set(sampled_inds)))
+            if len(extra_inds) > num_extra:
+                extra_inds = self.random_choice(extra_inds, num_extra)
+            sampled_inds = np.concatenate([sampled_inds, extra_inds])
+
+        return sampled_inds
+
+    def _sample_neg(self, assign_result, num_expected, **kwargs):
+        """Sample negative boxes.
+
+        Args:
+            assign_result (:obj:`AssignResult`): The assigned results of boxes.
+            num_expected (int): The number of expected negative samples
+
+        Returns:
+            Tensor or ndarray: sampled indices.
+        """
+        neg_inds = torch.nonzero(assign_result.gt_inds == 0, as_tuple=False)
+        if neg_inds.numel() != 0:
+            neg_inds = neg_inds.squeeze(1)
+        if len(neg_inds) <= num_expected:
+            return neg_inds
+        else:
+            max_overlaps = assign_result.max_overlaps.cpu().numpy()
+            # balance sampling for negative samples
+            neg_set = set(neg_inds.cpu().numpy())
+
+            if self.floor_thr > 0:
+                floor_set = set(
+                    np.where(
+                        np.logical_and(max_overlaps >= 0,
+                                       max_overlaps < self.floor_thr))[0])
+                iou_sampling_set = set(
+                    np.where(max_overlaps >= self.floor_thr)[0])
+            elif self.floor_thr == 0:
+                floor_set = set(np.where(max_overlaps == 0)[0])
+                iou_sampling_set = set(
+                    np.where(max_overlaps > self.floor_thr)[0])
+            else:
+                floor_set = set()
+                iou_sampling_set = set(
+                    np.where(max_overlaps > self.floor_thr)[0])
+                # for sampling interval calculation
+                self.floor_thr = 0
+
+            floor_neg_inds = list(floor_set & neg_set)
+            iou_sampling_neg_inds = list(iou_sampling_set & neg_set)
+            num_expected_iou_sampling = int(num_expected *
+                                            (1 - self.floor_fraction))
+            if len(iou_sampling_neg_inds) > num_expected_iou_sampling:
+                if self.num_bins >= 2:
+                    iou_sampled_inds = self.sample_via_interval(
+                        max_overlaps, set(iou_sampling_neg_inds),
+                        num_expected_iou_sampling)
+                else:
+                    iou_sampled_inds = self.random_choice(
+                        iou_sampling_neg_inds, num_expected_iou_sampling)
+            else:
+                iou_sampled_inds = np.array(
+                    iou_sampling_neg_inds, dtype=np.int)
+            num_expected_floor = num_expected - len(iou_sampled_inds)
+            if len(floor_neg_inds) > num_expected_floor:
+                sampled_floor_inds = self.random_choice(
+                    floor_neg_inds, num_expected_floor)
+            else:
+                sampled_floor_inds = np.array(floor_neg_inds, dtype=np.int)
+            sampled_inds = np.concatenate(
+                (sampled_floor_inds, iou_sampled_inds))
+            if len(sampled_inds) < num_expected:
+                num_extra = num_expected - len(sampled_inds)
+                extra_inds = np.array(list(neg_set - set(sampled_inds)))
+                if len(extra_inds) > num_extra:
+                    extra_inds = self.random_choice(extra_inds, num_extra)
+                sampled_inds = np.concatenate((sampled_inds, extra_inds))
+            sampled_inds = torch.from_numpy(sampled_inds).long().to(
+                assign_result.gt_inds.device)
+            return sampled_inds
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/ohem_sampler.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/ohem_sampler.py
new file mode 100644
index 0000000..8b99f60
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/ohem_sampler.py
@@ -0,0 +1,107 @@
+import torch
+
+from ..builder import BBOX_SAMPLERS
+from ..transforms import bbox2roi
+from .base_sampler import BaseSampler
+
+
+@BBOX_SAMPLERS.register_module()
+class OHEMSampler(BaseSampler):
+    r"""Online Hard Example Mining Sampler described in `Training Region-based
+    Object Detectors with Online Hard Example Mining
+    <https://arxiv.org/abs/1604.03540>`_.
+    """
+
+    def __init__(self,
+                 num,
+                 pos_fraction,
+                 context,
+                 neg_pos_ub=-1,
+                 add_gt_as_proposals=True,
+                 **kwargs):
+        super(OHEMSampler, self).__init__(num, pos_fraction, neg_pos_ub,
+                                          add_gt_as_proposals)
+        self.context = context
+        if not hasattr(self.context, 'num_stages'):
+            self.bbox_head = self.context.bbox_head
+        else:
+            self.bbox_head = self.context.bbox_head[self.context.current_stage]
+
+    def hard_mining(self, inds, num_expected, bboxes, labels, feats):
+        with torch.no_grad():
+            rois = bbox2roi([bboxes])
+            if not hasattr(self.context, 'num_stages'):
+                bbox_results = self.context._bbox_forward(feats, rois)
+            else:
+                bbox_results = self.context._bbox_forward(
+                    self.context.current_stage, feats, rois)
+            cls_score = bbox_results['cls_score']
+            loss = self.bbox_head.loss(
+                cls_score=cls_score,
+                bbox_pred=None,
+                rois=rois,
+                labels=labels,
+                label_weights=cls_score.new_ones(cls_score.size(0)),
+                bbox_targets=None,
+                bbox_weights=None,
+                reduction_override='none')['loss_cls']
+            _, topk_loss_inds = loss.topk(num_expected)
+        return inds[topk_loss_inds]
+
+    def _sample_pos(self,
+                    assign_result,
+                    num_expected,
+                    bboxes=None,
+                    feats=None,
+                    **kwargs):
+        """Sample positive boxes.
+
+        Args:
+            assign_result (:obj:`AssignResult`): Assigned results
+            num_expected (int): Number of expected positive samples
+            bboxes (torch.Tensor, optional): Boxes. Defaults to None.
+            feats (list[torch.Tensor], optional): Multi-level features.
+                Defaults to None.
+
+        Returns:
+            torch.Tensor: Indices  of positive samples
+        """
+        # Sample some hard positive samples
+        pos_inds = torch.nonzero(assign_result.gt_inds > 0, as_tuple=False)
+        if pos_inds.numel() != 0:
+            pos_inds = pos_inds.squeeze(1)
+        if pos_inds.numel() <= num_expected:
+            return pos_inds
+        else:
+            return self.hard_mining(pos_inds, num_expected, bboxes[pos_inds],
+                                    assign_result.labels[pos_inds], feats)
+
+    def _sample_neg(self,
+                    assign_result,
+                    num_expected,
+                    bboxes=None,
+                    feats=None,
+                    **kwargs):
+        """Sample negative boxes.
+
+        Args:
+            assign_result (:obj:`AssignResult`): Assigned results
+            num_expected (int): Number of expected negative samples
+            bboxes (torch.Tensor, optional): Boxes. Defaults to None.
+            feats (list[torch.Tensor], optional): Multi-level features.
+                Defaults to None.
+
+        Returns:
+            torch.Tensor: Indices  of negative samples
+        """
+        # Sample some hard negative samples
+        neg_inds = torch.nonzero(assign_result.gt_inds == 0, as_tuple=False)
+        if neg_inds.numel() != 0:
+            neg_inds = neg_inds.squeeze(1)
+        if len(neg_inds) <= num_expected:
+            return neg_inds
+        else:
+            neg_labels = assign_result.labels.new_empty(
+                neg_inds.size(0)).fill_(self.bbox_head.num_classes)
+            return self.hard_mining(neg_inds, num_expected, bboxes[neg_inds],
+                                    neg_labels, feats)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/pseudo_sampler.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/pseudo_sampler.py
new file mode 100644
index 0000000..2bd81ab
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/pseudo_sampler.py
@@ -0,0 +1,41 @@
+import torch
+
+from ..builder import BBOX_SAMPLERS
+from .base_sampler import BaseSampler
+from .sampling_result import SamplingResult
+
+
+@BBOX_SAMPLERS.register_module()
+class PseudoSampler(BaseSampler):
+    """A pseudo sampler that does not do sampling actually."""
+
+    def __init__(self, **kwargs):
+        pass
+
+    def _sample_pos(self, **kwargs):
+        """Sample positive samples."""
+        raise NotImplementedError
+
+    def _sample_neg(self, **kwargs):
+        """Sample negative samples."""
+        raise NotImplementedError
+
+    def sample(self, assign_result, bboxes, gt_bboxes, **kwargs):
+        """Directly returns the positive and negative indices  of samples.
+
+        Args:
+            assign_result (:obj:`AssignResult`): Assigned results
+            bboxes (torch.Tensor): Bounding boxes
+            gt_bboxes (torch.Tensor): Ground truth boxes
+
+        Returns:
+            :obj:`SamplingResult`: sampler results
+        """
+        pos_inds = torch.nonzero(
+            assign_result.gt_inds > 0, as_tuple=False).squeeze(-1).unique()
+        neg_inds = torch.nonzero(
+            assign_result.gt_inds == 0, as_tuple=False).squeeze(-1).unique()
+        gt_flags = bboxes.new_zeros(bboxes.shape[0], dtype=torch.uint8)
+        sampling_result = SamplingResult(pos_inds, neg_inds, bboxes, gt_bboxes,
+                                         assign_result, gt_flags)
+        return sampling_result
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/random_sampler.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/random_sampler.py
new file mode 100644
index 0000000..c23a7a1
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/random_sampler.py
@@ -0,0 +1,81 @@
+import torch
+
+from ..builder import BBOX_SAMPLERS
+from .base_sampler import BaseSampler
+
+
+@BBOX_SAMPLERS.register_module()
+class RandomSampler(BaseSampler):
+    """Random sampler.
+
+    Args:
+        num (int): Number of samples
+        pos_fraction (float): Fraction of positive samples
+        neg_pos_up (int, optional): Upper bound number of negative and
+            positive samples. Defaults to -1.
+        add_gt_as_proposals (bool, optional): Whether to add ground truth
+            boxes as proposals. Defaults to True.
+    """
+
+    def __init__(self,
+                 num,
+                 pos_fraction,
+                 neg_pos_ub=-1,
+                 add_gt_as_proposals=True,
+                 **kwargs):
+        from mmdet.core.bbox import demodata
+        super(RandomSampler, self).__init__(num, pos_fraction, neg_pos_ub,
+                                            add_gt_as_proposals)
+        self.rng = demodata.ensure_rng(kwargs.get('rng', None))
+
+    def random_choice(self, gallery, num):
+        """Random select some elements from the gallery.
+
+        If `gallery` is a Tensor, the returned indices will be a Tensor;
+        If `gallery` is a ndarray or list, the returned indices will be a
+        ndarray.
+
+        Args:
+            gallery (Tensor | ndarray | list): indices pool.
+            num (int): expected sample num.
+
+        Returns:
+            Tensor or ndarray: sampled indices.
+        """
+        assert len(gallery) >= num
+
+        is_tensor = isinstance(gallery, torch.Tensor)
+        if not is_tensor:
+            if torch.cuda.is_available():
+                device = torch.cuda.current_device()
+            else:
+                device = 'cpu'
+            gallery = torch.tensor(gallery, dtype=torch.long, device=device)
+        # This is a temporary fix. We can revert the following code
+        # when PyTorch fixes the abnormal return of torch.randperm.
+        # See: https://github.com/open-mmlab/mmdetection/pull/5014
+        perm = torch.randperm(gallery.numel())[:num].to(device=gallery.device)
+        rand_inds = gallery[perm]
+        if not is_tensor:
+            rand_inds = rand_inds.cpu().numpy()
+        return rand_inds
+
+    def _sample_pos(self, assign_result, num_expected, **kwargs):
+        """Randomly sample some positive samples."""
+        pos_inds = torch.nonzero(assign_result.gt_inds > 0, as_tuple=False)
+        if pos_inds.numel() != 0:
+            pos_inds = pos_inds.squeeze(1)
+        if pos_inds.numel() <= num_expected:
+            return pos_inds
+        else:
+            return self.random_choice(pos_inds, num_expected)
+
+    def _sample_neg(self, assign_result, num_expected, **kwargs):
+        """Randomly sample some negative samples."""
+        neg_inds = torch.nonzero(assign_result.gt_inds == 0, as_tuple=False)
+        if neg_inds.numel() != 0:
+            neg_inds = neg_inds.squeeze(1)
+        if len(neg_inds) <= num_expected:
+            return neg_inds
+        else:
+            return self.random_choice(neg_inds, num_expected)
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/sampling_result.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/sampling_result.py
new file mode 100644
index 0000000..058bae1
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/sampling_result.py
@@ -0,0 +1,153 @@
+import torch
+
+from mmdet.utils import util_mixins
+
+
+class SamplingResult(util_mixins.NiceRepr):
+    """Bbox sampling result.
+
+    Example:
+        >>> # xdoctest: +IGNORE_WANT
+        >>> from mmdet.core.bbox.samplers.sampling_result import *  # NOQA
+        >>> self = SamplingResult.random(rng=10)
+        >>> print(f'self = {self}')
+        self = <SamplingResult({
+            'neg_bboxes': torch.Size([12, 4]),
+            'neg_inds': tensor([ 0,  1,  2,  4,  5,  6,  7,  8,  9, 10, 11, 12]),
+            'num_gts': 4,
+            'pos_assigned_gt_inds': tensor([], dtype=torch.int64),
+            'pos_bboxes': torch.Size([0, 4]),
+            'pos_inds': tensor([], dtype=torch.int64),
+            'pos_is_gt': tensor([], dtype=torch.uint8)
+        })>
+    """
+
+    def __init__(self, pos_inds, neg_inds, bboxes, gt_bboxes, assign_result,
+                 gt_flags):
+        self.pos_inds = pos_inds
+        self.neg_inds = neg_inds
+        self.pos_bboxes = bboxes[pos_inds]
+        self.neg_bboxes = bboxes[neg_inds]
+        self.pos_is_gt = gt_flags[pos_inds]
+
+        self.num_gts = gt_bboxes.shape[0]
+        self.pos_assigned_gt_inds = assign_result.gt_inds[pos_inds] - 1
+        self.assign_result = assign_result
+
+        if gt_bboxes.numel() == 0:
+            # hack for index error case
+            assert self.pos_assigned_gt_inds.numel() == 0
+            self.pos_gt_bboxes = torch.empty_like(gt_bboxes).view(-1, 4)
+        else:
+            if len(gt_bboxes.shape) < 2:
+                gt_bboxes = gt_bboxes.view(-1, 4)
+
+            self.pos_gt_bboxes = gt_bboxes[self.pos_assigned_gt_inds, :]
+
+        if assign_result.labels is not None:
+            self.pos_gt_labels = assign_result.labels[pos_inds]
+        else:
+            self.pos_gt_labels = None
+
+    @property
+    def bboxes(self):
+        """torch.Tensor: concatenated positive and negative boxes"""
+        return torch.cat([self.pos_bboxes, self.neg_bboxes])
+
+    def to(self, device):
+        """Change the device of the data inplace.
+
+        Example:
+            >>> self = SamplingResult.random()
+            >>> print(f'self = {self.to(None)}')
+            >>> # xdoctest: +REQUIRES(--gpu)
+            >>> print(f'self = {self.to(0)}')
+        """
+        _dict = self.__dict__
+        for key, value in _dict.items():
+            if isinstance(value, torch.Tensor):
+                _dict[key] = value.to(device)
+        return self
+
+    def __nice__(self):
+        data = self.info.copy()
+        data['pos_bboxes'] = data.pop('pos_bboxes').shape
+        data['neg_bboxes'] = data.pop('neg_bboxes').shape
+        parts = [f"'{k}': {v!r}" for k, v in sorted(data.items())]
+        body = '    ' + ',\n    '.join(parts)
+        return '{\n' + body + '\n}'
+
+    @property
+    def info(self):
+        """Returns a dictionary of info about the object."""
+        return {
+            'pos_inds': self.pos_inds,
+            'neg_inds': self.neg_inds,
+            'pos_bboxes': self.pos_bboxes,
+            'neg_bboxes': self.neg_bboxes,
+            'pos_is_gt': self.pos_is_gt,
+            'num_gts': self.num_gts,
+            'pos_assigned_gt_inds': self.pos_assigned_gt_inds,
+        }
+
+    @classmethod
+    def random(cls, rng=None, **kwargs):
+        """
+        Args:
+            rng (None | int | numpy.random.RandomState): seed or state.
+            kwargs (keyword arguments):
+                - num_preds: number of predicted boxes
+                - num_gts: number of true boxes
+                - p_ignore (float): probability of a predicted box assinged to \
+                    an ignored truth.
+                - p_assigned (float): probability of a predicted box not being \
+                    assigned.
+                - p_use_label (float | bool): with labels or not.
+
+        Returns:
+            :obj:`SamplingResult`: Randomly generated sampling result.
+
+        Example:
+            >>> from mmdet.core.bbox.samplers.sampling_result import *  # NOQA
+            >>> self = SamplingResult.random()
+            >>> print(self.__dict__)
+        """
+        from mmdet.core.bbox.samplers.random_sampler import RandomSampler
+        from mmdet.core.bbox.assigners.assign_result import AssignResult
+        from mmdet.core.bbox import demodata
+        rng = demodata.ensure_rng(rng)
+
+        # make probabalistic?
+        num = 32
+        pos_fraction = 0.5
+        neg_pos_ub = -1
+
+        assign_result = AssignResult.random(rng=rng, **kwargs)
+
+        # Note we could just compute an assignment
+        bboxes = demodata.random_boxes(assign_result.num_preds, rng=rng)
+        gt_bboxes = demodata.random_boxes(assign_result.num_gts, rng=rng)
+
+        if rng.rand() > 0.2:
+            # sometimes algorithms squeeze their data, be robust to that
+            gt_bboxes = gt_bboxes.squeeze()
+            bboxes = bboxes.squeeze()
+
+        if assign_result.labels is None:
+            gt_labels = None
+        else:
+            gt_labels = None  # todo
+
+        if gt_labels is None:
+            add_gt_as_proposals = False
+        else:
+            add_gt_as_proposals = True  # make probabalistic?
+
+        sampler = RandomSampler(
+            num,
+            pos_fraction,
+            neg_pos_ub=neg_pos_ub,
+            add_gt_as_proposals=add_gt_as_proposals,
+            rng=rng)
+        self = sampler.sample(assign_result, bboxes, gt_bboxes, gt_labels)
+        return self
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/samplers/score_hlr_sampler.py b/python/autoCut/models/swin/mmdet/core/bbox/samplers/score_hlr_sampler.py
new file mode 100644
index 0000000..11d46b9
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/samplers/score_hlr_sampler.py
@@ -0,0 +1,264 @@
+import torch
+from mmcv.ops import nms_match
+
+from ..builder import BBOX_SAMPLERS
+from ..transforms import bbox2roi
+from .base_sampler import BaseSampler
+from .sampling_result import SamplingResult
+
+
+@BBOX_SAMPLERS.register_module()
+class ScoreHLRSampler(BaseSampler):
+    r"""Importance-based Sample Reweighting (ISR_N), described in `Prime Sample
+    Attention in Object Detection <https://arxiv.org/abs/1904.04821>`_.
+
+    Score hierarchical local rank (HLR) differentiates with RandomSampler in
+    negative part. It firstly computes Score-HLR in a two-step way,
+    then linearly maps score hlr to the loss weights.
+
+    Args:
+        num (int): Total number of sampled RoIs.
+        pos_fraction (float): Fraction of positive samples.
+        context (:class:`BaseRoIHead`): RoI head that the sampler belongs to.
+        neg_pos_ub (int): Upper bound of the ratio of num negative to num
+            positive, -1 means no upper bound.
+        add_gt_as_proposals (bool): Whether to add ground truth as proposals.
+        k (float): Power of the non-linear mapping.
+        bias (float): Shift of the non-linear mapping.
+        score_thr (float): Minimum score that a negative sample is to be
+            considered as valid bbox.
+    """
+
+    def __init__(self,
+                 num,
+                 pos_fraction,
+                 context,
+                 neg_pos_ub=-1,
+                 add_gt_as_proposals=True,
+                 k=0.5,
+                 bias=0,
+                 score_thr=0.05,
+                 iou_thr=0.5,
+                 **kwargs):
+        super().__init__(num, pos_fraction, neg_pos_ub, add_gt_as_proposals)
+        self.k = k
+        self.bias = bias
+        self.score_thr = score_thr
+        self.iou_thr = iou_thr
+        self.context = context
+        # context of cascade detectors is a list, so distinguish them here.
+        if not hasattr(context, 'num_stages'):
+            self.bbox_roi_extractor = context.bbox_roi_extractor
+            self.bbox_head = context.bbox_head
+            self.with_shared_head = context.with_shared_head
+            if self.with_shared_head:
+                self.shared_head = context.shared_head
+        else:
+            self.bbox_roi_extractor = context.bbox_roi_extractor[
+                context.current_stage]
+            self.bbox_head = context.bbox_head[context.current_stage]
+
+    @staticmethod
+    def random_choice(gallery, num):
+        """Randomly select some elements from the gallery.
+
+        If `gallery` is a Tensor, the returned indices will be a Tensor;
+        If `gallery` is a ndarray or list, the returned indices will be a
+        ndarray.
+
+        Args:
+            gallery (Tensor | ndarray | list): indices pool.
+            num (int): expected sample num.
+
+        Returns:
+            Tensor or ndarray: sampled indices.
+        """
+        assert len(gallery) >= num
+
+        is_tensor = isinstance(gallery, torch.Tensor)
+        if not is_tensor:
+            if torch.cuda.is_available():
+                device = torch.cuda.current_device()
+            else:
+                device = 'cpu'
+            gallery = torch.tensor(gallery, dtype=torch.long, device=device)
+        perm = torch.randperm(gallery.numel(), device=gallery.device)[:num]
+        rand_inds = gallery[perm]
+        if not is_tensor:
+            rand_inds = rand_inds.cpu().numpy()
+        return rand_inds
+
+    def _sample_pos(self, assign_result, num_expected, **kwargs):
+        """Randomly sample some positive samples."""
+        pos_inds = torch.nonzero(assign_result.gt_inds > 0).flatten()
+        if pos_inds.numel() <= num_expected:
+            return pos_inds
+        else:
+            return self.random_choice(pos_inds, num_expected)
+
+    def _sample_neg(self,
+                    assign_result,
+                    num_expected,
+                    bboxes,
+                    feats=None,
+                    img_meta=None,
+                    **kwargs):
+        """Sample negative samples.
+
+        Score-HLR sampler is done in the following steps:
+        1. Take the maximum positive score prediction of each negative samples
+            as s_i.
+        2. Filter out negative samples whose s_i <= score_thr, the left samples
+            are called valid samples.
+        3. Use NMS-Match to divide valid samples into different groups,
+            samples in the same group will greatly overlap with each other
+        4. Rank the matched samples in two-steps to get Score-HLR.
+            (1) In the same group, rank samples with their scores.
+            (2) In the same score rank across different groups,
+                rank samples with their scores again.
+        5. Linearly map Score-HLR to the final label weights.
+
+        Args:
+            assign_result (:obj:`AssignResult`): result of assigner.
+            num_expected (int): Expected number of samples.
+            bboxes (Tensor): bbox to be sampled.
+            feats (Tensor): Features come from FPN.
+            img_meta (dict): Meta information dictionary.
+        """
+        neg_inds = torch.nonzero(assign_result.gt_inds == 0).flatten()
+        num_neg = neg_inds.size(0)
+        if num_neg == 0:
+            return neg_inds, None
+        with torch.no_grad():
+            neg_bboxes = bboxes[neg_inds]
+            neg_rois = bbox2roi([neg_bboxes])
+            bbox_result = self.context._bbox_forward(feats, neg_rois)
+            cls_score, bbox_pred = bbox_result['cls_score'], bbox_result[
+                'bbox_pred']
+
+            ori_loss = self.bbox_head.loss(
+                cls_score=cls_score,
+                bbox_pred=None,
+                rois=None,
+                labels=neg_inds.new_full((num_neg, ),
+                                         self.bbox_head.num_classes),
+                label_weights=cls_score.new_ones(num_neg),
+                bbox_targets=None,
+                bbox_weights=None,
+                reduction_override='none')['loss_cls']
+
+            # filter out samples with the max score lower than score_thr
+            max_score, argmax_score = cls_score.softmax(-1)[:, :-1].max(-1)
+            valid_inds = (max_score > self.score_thr).nonzero().view(-1)
+            invalid_inds = (max_score <= self.score_thr).nonzero().view(-1)
+            num_valid = valid_inds.size(0)
+            num_invalid = invalid_inds.size(0)
+
+            num_expected = min(num_neg, num_expected)
+            num_hlr = min(num_valid, num_expected)
+            num_rand = num_expected - num_hlr
+            if num_valid > 0:
+                valid_rois = neg_rois[valid_inds]
+                valid_max_score = max_score[valid_inds]
+                valid_argmax_score = argmax_score[valid_inds]
+                valid_bbox_pred = bbox_pred[valid_inds]
+
+                # valid_bbox_pred shape: [num_valid, #num_classes, 4]
+                valid_bbox_pred = valid_bbox_pred.view(
+                    valid_bbox_pred.size(0), -1, 4)
+                selected_bbox_pred = valid_bbox_pred[range(num_valid),
+                                                     valid_argmax_score]
+                pred_bboxes = self.bbox_head.bbox_coder.decode(
+                    valid_rois[:, 1:], selected_bbox_pred)
+                pred_bboxes_with_score = torch.cat(
+                    [pred_bboxes, valid_max_score[:, None]], -1)
+                group = nms_match(pred_bboxes_with_score, self.iou_thr)
+
+                # imp: importance
+                imp = cls_score.new_zeros(num_valid)
+                for g in group:
+                    g_score = valid_max_score[g]
+                    # g_score has already sorted
+                    rank = g_score.new_tensor(range(g_score.size(0)))
+                    imp[g] = num_valid - rank + g_score
+                _, imp_rank_inds = imp.sort(descending=True)
+                _, imp_rank = imp_rank_inds.sort()
+                hlr_inds = imp_rank_inds[:num_expected]
+
+                if num_rand > 0:
+                    rand_inds = torch.randperm(num_invalid)[:num_rand]
+                    select_inds = torch.cat(
+                        [valid_inds[hlr_inds], invalid_inds[rand_inds]])
+                else:
+                    select_inds = valid_inds[hlr_inds]
+
+                neg_label_weights = cls_score.new_ones(num_expected)
+
+                up_bound = max(num_expected, num_valid)
+                imp_weights = (up_bound -
+                               imp_rank[hlr_inds].float()) / up_bound
+                neg_label_weights[:num_hlr] = imp_weights
+                neg_label_weights[num_hlr:] = imp_weights.min()
+                neg_label_weights = (self.bias +
+                                     (1 - self.bias) * neg_label_weights).pow(
+                                         self.k)
+                ori_selected_loss = ori_loss[select_inds]
+                new_loss = ori_selected_loss * neg_label_weights
+                norm_ratio = ori_selected_loss.sum() / new_loss.sum()
+                neg_label_weights *= norm_ratio
+            else:
+                neg_label_weights = cls_score.new_ones(num_expected)
+                select_inds = torch.randperm(num_neg)[:num_expected]
+
+            return neg_inds[select_inds], neg_label_weights
+
+    def sample(self,
+               assign_result,
+               bboxes,
+               gt_bboxes,
+               gt_labels=None,
+               img_meta=None,
+               **kwargs):
+        """Sample positive and negative bboxes.
+
+        This is a simple implementation of bbox sampling given candidates,
+        assigning results and ground truth bboxes.
+
+        Args:
+            assign_result (:obj:`AssignResult`): Bbox assigning results.
+            bboxes (Tensor): Boxes to be sampled from.
+            gt_bboxes (Tensor): Ground truth bboxes.
+            gt_labels (Tensor, optional): Class labels of ground truth bboxes.
+
+        Returns:
+            tuple[:obj:`SamplingResult`, Tensor]: Sampling result and negetive
+                label weights.
+        """
+        bboxes = bboxes[:, :4]
+
+        gt_flags = bboxes.new_zeros((bboxes.shape[0], ), dtype=torch.uint8)
+        if self.add_gt_as_proposals:
+            bboxes = torch.cat([gt_bboxes, bboxes], dim=0)
+            assign_result.add_gt_(gt_labels)
+            gt_ones = bboxes.new_ones(gt_bboxes.shape[0], dtype=torch.uint8)
+            gt_flags = torch.cat([gt_ones, gt_flags])
+
+        num_expected_pos = int(self.num * self.pos_fraction)
+        pos_inds = self.pos_sampler._sample_pos(
+            assign_result, num_expected_pos, bboxes=bboxes, **kwargs)
+        num_sampled_pos = pos_inds.numel()
+        num_expected_neg = self.num - num_sampled_pos
+        if self.neg_pos_ub >= 0:
+            _pos = max(1, num_sampled_pos)
+            neg_upper_bound = int(self.neg_pos_ub * _pos)
+            if num_expected_neg > neg_upper_bound:
+                num_expected_neg = neg_upper_bound
+        neg_inds, neg_label_weights = self.neg_sampler._sample_neg(
+            assign_result,
+            num_expected_neg,
+            bboxes,
+            img_meta=img_meta,
+            **kwargs)
+
+        return SamplingResult(pos_inds, neg_inds, bboxes, gt_bboxes,
+                              assign_result, gt_flags), neg_label_weights
diff --git a/python/autoCut/models/swin/mmdet/core/bbox/transforms.py b/python/autoCut/models/swin/mmdet/core/bbox/transforms.py
new file mode 100644
index 0000000..df55b0a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/bbox/transforms.py
@@ -0,0 +1,240 @@
+import numpy as np
+import torch
+
+
+def bbox_flip(bboxes, img_shape, direction='horizontal'):
+    """Flip bboxes horizontally or vertically.
+
+    Args:
+        bboxes (Tensor): Shape (..., 4*k)
+        img_shape (tuple): Image shape.
+        direction (str): Flip direction, options are "horizontal", "vertical",
+            "diagonal". Default: "horizontal"
+
+    Returns:
+        Tensor: Flipped bboxes.
+    """
+    assert bboxes.shape[-1] % 4 == 0
+    assert direction in ['horizontal', 'vertical', 'diagonal']
+    flipped = bboxes.clone()
+    if direction == 'horizontal':
+        flipped[..., 0::4] = img_shape[1] - bboxes[..., 2::4]
+        flipped[..., 2::4] = img_shape[1] - bboxes[..., 0::4]
+    elif direction == 'vertical':
+        flipped[..., 1::4] = img_shape[0] - bboxes[..., 3::4]
+        flipped[..., 3::4] = img_shape[0] - bboxes[..., 1::4]
+    else:
+        flipped[..., 0::4] = img_shape[1] - bboxes[..., 2::4]
+        flipped[..., 1::4] = img_shape[0] - bboxes[..., 3::4]
+        flipped[..., 2::4] = img_shape[1] - bboxes[..., 0::4]
+        flipped[..., 3::4] = img_shape[0] - bboxes[..., 1::4]
+    return flipped
+
+
+def bbox_mapping(bboxes,
+                 img_shape,
+                 scale_factor,
+                 flip,
+                 flip_direction='horizontal'):
+    """Map bboxes from the original image scale to testing scale."""
+    new_bboxes = bboxes * bboxes.new_tensor(scale_factor)
+    if flip:
+        new_bboxes = bbox_flip(new_bboxes, img_shape, flip_direction)
+    return new_bboxes
+
+
+def bbox_mapping_back(bboxes,
+                      img_shape,
+                      scale_factor,
+                      flip,
+                      flip_direction='horizontal'):
+    """Map bboxes from testing scale to original image scale."""
+    new_bboxes = bbox_flip(bboxes, img_shape,
+                           flip_direction) if flip else bboxes
+    new_bboxes = new_bboxes.view(-1, 4) / new_bboxes.new_tensor(scale_factor)
+    return new_bboxes.view(bboxes.shape)
+
+
+def bbox2roi(bbox_list):
+    """Convert a list of bboxes to roi format.
+
+    Args:
+        bbox_list (list[Tensor]): a list of bboxes corresponding to a batch
+            of images.
+
+    Returns:
+        Tensor: shape (n, 5), [batch_ind, x1, y1, x2, y2]
+    """
+    rois_list = []
+    for img_id, bboxes in enumerate(bbox_list):
+        if bboxes.size(0) > 0:
+            img_inds = bboxes.new_full((bboxes.size(0), 1), img_id)
+            rois = torch.cat([img_inds, bboxes[:, :4]], dim=-1)
+        else:
+            rois = bboxes.new_zeros((0, 5))
+        rois_list.append(rois)
+    rois = torch.cat(rois_list, 0)
+    return rois
+
+
+def roi2bbox(rois):
+    """Convert rois to bounding box format.
+
+    Args:
+        rois (torch.Tensor): RoIs with the shape (n, 5) where the first
+            column indicates batch id of each RoI.
+
+    Returns:
+        list[torch.Tensor]: Converted boxes of corresponding rois.
+    """
+    bbox_list = []
+    img_ids = torch.unique(rois[:, 0].cpu(), sorted=True)
+    for img_id in img_ids:
+        inds = (rois[:, 0] == img_id.item())
+        bbox = rois[inds, 1:]
+        bbox_list.append(bbox)
+    return bbox_list
+
+
+def bbox2result(bboxes, labels, num_classes):
+    """Convert detection results to a list of numpy arrays.
+
+    Args:
+        bboxes (torch.Tensor | np.ndarray): shape (n, 5)
+        labels (torch.Tensor | np.ndarray): shape (n, )
+        num_classes (int): class number, including background class
+
+    Returns:
+        list(ndarray): bbox results of each class
+    """
+    if bboxes.shape[0] == 0:
+        return [np.zeros((0, 5), dtype=np.float32) for i in range(num_classes)]
+    else:
+        if isinstance(bboxes, torch.Tensor):
+            bboxes = bboxes.detach().cpu().numpy()
+            labels = labels.detach().cpu().numpy()
+        return [bboxes[labels == i, :] for i in range(num_classes)]
+
+
+def distance2bbox(points, distance, max_shape=None):
+    """Decode distance prediction to bounding box.
+
+    Args:
+        points (Tensor): Shape (B, N, 2) or (N, 2).
+        distance (Tensor): Distance from the given point to 4
+            boundaries (left, top, right, bottom). Shape (B, N, 4) or (N, 4)
+        max_shape (Sequence[int] or torch.Tensor or Sequence[
+            Sequence[int]],optional): Maximum bounds for boxes, specifies
+            (H, W, C) or (H, W). If priors shape is (B, N, 4), then
+            the max_shape should be a Sequence[Sequence[int]]
+            and the length of max_shape should also be B.
+
+    Returns:
+        Tensor: Boxes with shape (N, 4) or (B, N, 4)
+    """
+    x1 = points[..., 0] - distance[..., 0]
+    y1 = points[..., 1] - distance[..., 1]
+    x2 = points[..., 0] + distance[..., 2]
+    y2 = points[..., 1] + distance[..., 3]
+
+    bboxes = torch.stack([x1, y1, x2, y2], -1)
+
+    if max_shape is not None:
+        if not isinstance(max_shape, torch.Tensor):
+            max_shape = x1.new_tensor(max_shape)
+        max_shape = max_shape[..., :2].type_as(x1)
+        if max_shape.ndim == 2:
+            assert bboxes.ndim == 3
+            assert max_shape.size(0) == bboxes.size(0)
+
+        min_xy = x1.new_tensor(0)
+        max_xy = torch.cat([max_shape, max_shape],
+                           dim=-1).flip(-1).unsqueeze(-2)
+        bboxes = torch.where(bboxes < min_xy, min_xy, bboxes)
+        bboxes = torch.where(bboxes > max_xy, max_xy, bboxes)
+
+    return bboxes
+
+
+def bbox2distance(points, bbox, max_dis=None, eps=0.1):
+    """Decode bounding box based on distances.
+
+    Args:
+        points (Tensor): Shape (n, 2), [x, y].
+        bbox (Tensor): Shape (n, 4), "xyxy" format
+        max_dis (float): Upper bound of the distance.
+        eps (float): a small value to ensure target < max_dis, instead <=
+
+    Returns:
+        Tensor: Decoded distances.
+    """
+    left = points[:, 0] - bbox[:, 0]
+    top = points[:, 1] - bbox[:, 1]
+    right = bbox[:, 2] - points[:, 0]
+    bottom = bbox[:, 3] - points[:, 1]
+    if max_dis is not None:
+        left = left.clamp(min=0, max=max_dis - eps)
+        top = top.clamp(min=0, max=max_dis - eps)
+        right = right.clamp(min=0, max=max_dis - eps)
+        bottom = bottom.clamp(min=0, max=max_dis - eps)
+    return torch.stack([left, top, right, bottom], -1)
+
+
+def bbox_rescale(bboxes, scale_factor=1.0):
+    """Rescale bounding box w.r.t. scale_factor.
+
+    Args:
+        bboxes (Tensor): Shape (n, 4) for bboxes or (n, 5) for rois
+        scale_factor (float): rescale factor
+
+    Returns:
+        Tensor: Rescaled bboxes.
+    """
+    if bboxes.size(1) == 5:
+        bboxes_ = bboxes[:, 1:]
+        inds_ = bboxes[:, 0]
+    else:
+        bboxes_ = bboxes
+    cx = (bboxes_[:, 0] + bboxes_[:, 2]) * 0.5
+    cy = (bboxes_[:, 1] + bboxes_[:, 3]) * 0.5
+    w = bboxes_[:, 2] - bboxes_[:, 0]
+    h = bboxes_[:, 3] - bboxes_[:, 1]
+    w = w * scale_factor
+    h = h * scale_factor
+    x1 = cx - 0.5 * w
+    x2 = cx + 0.5 * w
+    y1 = cy - 0.5 * h
+    y2 = cy + 0.5 * h
+    if bboxes.size(1) == 5:
+        rescaled_bboxes = torch.stack([inds_, x1, y1, x2, y2], dim=-1)
+    else:
+        rescaled_bboxes = torch.stack([x1, y1, x2, y2], dim=-1)
+    return rescaled_bboxes
+
+
+def bbox_cxcywh_to_xyxy(bbox):
+    """Convert bbox coordinates from (cx, cy, w, h) to (x1, y1, x2, y2).
+
+    Args:
+        bbox (Tensor): Shape (n, 4) for bboxes.
+
+    Returns:
+        Tensor: Converted bboxes.
+    """
+    cx, cy, w, h = bbox.split((1, 1, 1, 1), dim=-1)
+    bbox_new = [(cx - 0.5 * w), (cy - 0.5 * h), (cx + 0.5 * w), (cy + 0.5 * h)]
+    return torch.cat(bbox_new, dim=-1)
+
+
+def bbox_xyxy_to_cxcywh(bbox):
+    """Convert bbox coordinates from (x1, y1, x2, y2) to (cx, cy, w, h).
+
+    Args:
+        bbox (Tensor): Shape (n, 4) for bboxes.
+
+    Returns:
+        Tensor: Converted bboxes.
+    """
+    x1, y1, x2, y2 = bbox.split((1, 1, 1, 1), dim=-1)
+    bbox_new = [(x1 + x2) / 2, (y1 + y2) / 2, (x2 - x1), (y2 - y1)]
+    return torch.cat(bbox_new, dim=-1)
diff --git a/python/autoCut/models/swin/mmdet/core/evaluation/__init__.py b/python/autoCut/models/swin/mmdet/core/evaluation/__init__.py
new file mode 100644
index 0000000..d11ef15
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/evaluation/__init__.py
@@ -0,0 +1,15 @@
+from .class_names import (cityscapes_classes, coco_classes, dataset_aliases,
+                          get_classes, imagenet_det_classes,
+                          imagenet_vid_classes, voc_classes)
+from .eval_hooks import DistEvalHook, EvalHook
+from .mean_ap import average_precision, eval_map, print_map_summary
+from .recall import (eval_recalls, plot_iou_recall, plot_num_recall,
+                     print_recall_summary)
+
+__all__ = [
+    'voc_classes', 'imagenet_det_classes', 'imagenet_vid_classes',
+    'coco_classes', 'cityscapes_classes', 'dataset_aliases', 'get_classes',
+    'DistEvalHook', 'EvalHook', 'average_precision', 'eval_map',
+    'print_map_summary', 'eval_recalls', 'print_recall_summary',
+    'plot_num_recall', 'plot_iou_recall'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/evaluation/bbox_overlaps.py b/python/autoCut/models/swin/mmdet/core/evaluation/bbox_overlaps.py
new file mode 100644
index 0000000..93559ea
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/evaluation/bbox_overlaps.py
@@ -0,0 +1,48 @@
+import numpy as np
+
+
+def bbox_overlaps(bboxes1, bboxes2, mode='iou', eps=1e-6):
+    """Calculate the ious between each bbox of bboxes1 and bboxes2.
+
+    Args:
+        bboxes1(ndarray): shape (n, 4)
+        bboxes2(ndarray): shape (k, 4)
+        mode(str): iou (intersection over union) or iof (intersection
+            over foreground)
+
+    Returns:
+        ious(ndarray): shape (n, k)
+    """
+
+    assert mode in ['iou', 'iof']
+
+    bboxes1 = bboxes1.astype(np.float32)
+    bboxes2 = bboxes2.astype(np.float32)
+    rows = bboxes1.shape[0]
+    cols = bboxes2.shape[0]
+    ious = np.zeros((rows, cols), dtype=np.float32)
+    if rows * cols == 0:
+        return ious
+    exchange = False
+    if bboxes1.shape[0] > bboxes2.shape[0]:
+        bboxes1, bboxes2 = bboxes2, bboxes1
+        ious = np.zeros((cols, rows), dtype=np.float32)
+        exchange = True
+    area1 = (bboxes1[:, 2] - bboxes1[:, 0]) * (bboxes1[:, 3] - bboxes1[:, 1])
+    area2 = (bboxes2[:, 2] - bboxes2[:, 0]) * (bboxes2[:, 3] - bboxes2[:, 1])
+    for i in range(bboxes1.shape[0]):
+        x_start = np.maximum(bboxes1[i, 0], bboxes2[:, 0])
+        y_start = np.maximum(bboxes1[i, 1], bboxes2[:, 1])
+        x_end = np.minimum(bboxes1[i, 2], bboxes2[:, 2])
+        y_end = np.minimum(bboxes1[i, 3], bboxes2[:, 3])
+        overlap = np.maximum(x_end - x_start, 0) * np.maximum(
+            y_end - y_start, 0)
+        if mode == 'iou':
+            union = area1[i] + area2 - overlap
+        else:
+            union = area1[i] if not exchange else area2
+        union = np.maximum(union, eps)
+        ious[i, :] = overlap / union
+    if exchange:
+        ious = ious.T
+    return ious
diff --git a/python/autoCut/models/swin/mmdet/core/evaluation/class_names.py b/python/autoCut/models/swin/mmdet/core/evaluation/class_names.py
new file mode 100644
index 0000000..c2487c2
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/evaluation/class_names.py
@@ -0,0 +1,116 @@
+import mmcv
+
+
+def wider_face_classes():
+    return ['face']
+
+
+def voc_classes():
+    return [
+        'aeroplane', 'bicycle', 'bird', 'boat', 'bottle', 'bus', 'car', 'cat',
+        'chair', 'cow', 'diningtable', 'dog', 'horse', 'motorbike', 'person',
+        'pottedplant', 'sheep', 'sofa', 'train', 'tvmonitor'
+    ]
+
+
+def imagenet_det_classes():
+    return [
+        'accordion', 'airplane', 'ant', 'antelope', 'apple', 'armadillo',
+        'artichoke', 'axe', 'baby_bed', 'backpack', 'bagel', 'balance_beam',
+        'banana', 'band_aid', 'banjo', 'baseball', 'basketball', 'bathing_cap',
+        'beaker', 'bear', 'bee', 'bell_pepper', 'bench', 'bicycle', 'binder',
+        'bird', 'bookshelf', 'bow_tie', 'bow', 'bowl', 'brassiere', 'burrito',
+        'bus', 'butterfly', 'camel', 'can_opener', 'car', 'cart', 'cattle',
+        'cello', 'centipede', 'chain_saw', 'chair', 'chime', 'cocktail_shaker',
+        'coffee_maker', 'computer_keyboard', 'computer_mouse', 'corkscrew',
+        'cream', 'croquet_ball', 'crutch', 'cucumber', 'cup_or_mug', 'diaper',
+        'digital_clock', 'dishwasher', 'dog', 'domestic_cat', 'dragonfly',
+        'drum', 'dumbbell', 'electric_fan', 'elephant', 'face_powder', 'fig',
+        'filing_cabinet', 'flower_pot', 'flute', 'fox', 'french_horn', 'frog',
+        'frying_pan', 'giant_panda', 'goldfish', 'golf_ball', 'golfcart',
+        'guacamole', 'guitar', 'hair_dryer', 'hair_spray', 'hamburger',
+        'hammer', 'hamster', 'harmonica', 'harp', 'hat_with_a_wide_brim',
+        'head_cabbage', 'helmet', 'hippopotamus', 'horizontal_bar', 'horse',
+        'hotdog', 'iPod', 'isopod', 'jellyfish', 'koala_bear', 'ladle',
+        'ladybug', 'lamp', 'laptop', 'lemon', 'lion', 'lipstick', 'lizard',
+        'lobster', 'maillot', 'maraca', 'microphone', 'microwave', 'milk_can',
+        'miniskirt', 'monkey', 'motorcycle', 'mushroom', 'nail', 'neck_brace',
+        'oboe', 'orange', 'otter', 'pencil_box', 'pencil_sharpener', 'perfume',
+        'person', 'piano', 'pineapple', 'ping-pong_ball', 'pitcher', 'pizza',
+        'plastic_bag', 'plate_rack', 'pomegranate', 'popsicle', 'porcupine',
+        'power_drill', 'pretzel', 'printer', 'puck', 'punching_bag', 'purse',
+        'rabbit', 'racket', 'ray', 'red_panda', 'refrigerator',
+        'remote_control', 'rubber_eraser', 'rugby_ball', 'ruler',
+        'salt_or_pepper_shaker', 'saxophone', 'scorpion', 'screwdriver',
+        'seal', 'sheep', 'ski', 'skunk', 'snail', 'snake', 'snowmobile',
+        'snowplow', 'soap_dispenser', 'soccer_ball', 'sofa', 'spatula',
+        'squirrel', 'starfish', 'stethoscope', 'stove', 'strainer',
+        'strawberry', 'stretcher', 'sunglasses', 'swimming_trunks', 'swine',
+        'syringe', 'table', 'tape_player', 'tennis_ball', 'tick', 'tie',
+        'tiger', 'toaster', 'traffic_light', 'train', 'trombone', 'trumpet',
+        'turtle', 'tv_or_monitor', 'unicycle', 'vacuum', 'violin',
+        'volleyball', 'waffle_iron', 'washer', 'water_bottle', 'watercraft',
+        'whale', 'wine_bottle', 'zebra'
+    ]
+
+
+def imagenet_vid_classes():
+    return [
+        'airplane', 'antelope', 'bear', 'bicycle', 'bird', 'bus', 'car',
+        'cattle', 'dog', 'domestic_cat', 'elephant', 'fox', 'giant_panda',
+        'hamster', 'horse', 'lion', 'lizard', 'monkey', 'motorcycle', 'rabbit',
+        'red_panda', 'sheep', 'snake', 'squirrel', 'tiger', 'train', 'turtle',
+        'watercraft', 'whale', 'zebra'
+    ]
+
+
+def coco_classes():
+    return [
+        'person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train',
+        'truck', 'boat', 'traffic_light', 'fire_hydrant', 'stop_sign',
+        'parking_meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep',
+        'cow', 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella',
+        'handbag', 'tie', 'suitcase', 'frisbee', 'skis', 'snowboard',
+        'sports_ball', 'kite', 'baseball_bat', 'baseball_glove', 'skateboard',
+        'surfboard', 'tennis_racket', 'bottle', 'wine_glass', 'cup', 'fork',
+        'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange',
+        'broccoli', 'carrot', 'hot_dog', 'pizza', 'donut', 'cake', 'chair',
+        'couch', 'potted_plant', 'bed', 'dining_table', 'toilet', 'tv',
+        'laptop', 'mouse', 'remote', 'keyboard', 'cell_phone', 'microwave',
+        'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase',
+        'scissors', 'teddy_bear', 'hair_drier', 'toothbrush'
+    ]
+
+
+def cityscapes_classes():
+    return [
+        'person', 'rider', 'car', 'truck', 'bus', 'train', 'motorcycle',
+        'bicycle'
+    ]
+
+
+dataset_aliases = {
+    'voc': ['voc', 'pascal_voc', 'voc07', 'voc12'],
+    'imagenet_det': ['det', 'imagenet_det', 'ilsvrc_det'],
+    'imagenet_vid': ['vid', 'imagenet_vid', 'ilsvrc_vid'],
+    'coco': ['coco', 'mscoco', 'ms_coco'],
+    'wider_face': ['WIDERFaceDataset', 'wider_face', 'WIDERFace'],
+    'cityscapes': ['cityscapes']
+}
+
+
+def get_classes(dataset):
+    """Get class names of a dataset."""
+    alias2name = {}
+    for name, aliases in dataset_aliases.items():
+        for alias in aliases:
+            alias2name[alias] = name
+
+    if mmcv.is_str(dataset):
+        if dataset in alias2name:
+            labels = eval(alias2name[dataset] + '_classes()')
+        else:
+            raise ValueError(f'Unrecognized dataset: {dataset}')
+    else:
+        raise TypeError(f'dataset must a str, but got {type(dataset)}')
+    return labels
diff --git a/python/autoCut/models/swin/mmdet/core/evaluation/eval_hooks.py b/python/autoCut/models/swin/mmdet/core/evaluation/eval_hooks.py
new file mode 100644
index 0000000..6fb932e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/evaluation/eval_hooks.py
@@ -0,0 +1,303 @@
+import os.path as osp
+import warnings
+from math import inf
+
+import mmcv
+import torch.distributed as dist
+from mmcv.runner import Hook
+from torch.nn.modules.batchnorm import _BatchNorm
+from torch.utils.data import DataLoader
+
+from mmdet.utils import get_root_logger
+
+
+class EvalHook(Hook):
+    """Evaluation hook.
+
+    Notes:
+        If new arguments are added for EvalHook, tools/test.py,
+        tools/analysis_tools/eval_metric.py may be effected.
+
+    Attributes:
+        dataloader (DataLoader): A PyTorch dataloader.
+        start (int, optional): Evaluation starting epoch. It enables evaluation
+            before the training starts if ``start`` <= the resuming epoch.
+            If None, whether to evaluate is merely decided by ``interval``.
+            Default: None.
+        interval (int): Evaluation interval (by epochs). Default: 1.
+        save_best (str, optional): If a metric is specified, it would measure
+            the best checkpoint during evaluation. The information about best
+            checkpoint would be save in best.json.
+            Options are the evaluation metrics to the test dataset. e.g.,
+            ``bbox_mAP``, ``segm_mAP`` for bbox detection and instance
+            segmentation. ``AR@100`` for proposal recall. If ``save_best`` is
+            ``auto``, the first key will be used. The interval of
+            ``CheckpointHook`` should device EvalHook. Default: None.
+        rule (str, optional): Comparison rule for best score. If set to None,
+            it will infer a reasonable rule. Keys such as 'mAP' or 'AR' will
+            be inferred by 'greater' rule. Keys contain 'loss' will be inferred
+             by 'less' rule. Options are 'greater', 'less'. Default: None.
+        **eval_kwargs: Evaluation arguments fed into the evaluate function of
+            the dataset.
+    """
+
+    rule_map = {'greater': lambda x, y: x > y, 'less': lambda x, y: x < y}
+    init_value_map = {'greater': -inf, 'less': inf}
+    greater_keys = ['mAP', 'AR']
+    less_keys = ['loss']
+
+    def __init__(self,
+                 dataloader,
+                 start=None,
+                 interval=1,
+                 by_epoch=True,
+                 save_best=None,
+                 rule=None,
+                 **eval_kwargs):
+        if not isinstance(dataloader, DataLoader):
+            raise TypeError('dataloader must be a pytorch DataLoader, but got'
+                            f' {type(dataloader)}')
+        if not interval > 0:
+            raise ValueError(f'interval must be positive, but got {interval}')
+        if start is not None and start < 0:
+            warnings.warn(
+                f'The evaluation start epoch {start} is smaller than 0, '
+                f'use 0 instead', UserWarning)
+            start = 0
+        self.dataloader = dataloader
+        self.interval = interval
+        self.by_epoch = by_epoch
+        self.start = start
+        assert isinstance(save_best, str) or save_best is None
+        self.save_best = save_best
+        self.eval_kwargs = eval_kwargs
+        self.initial_epoch_flag = True
+
+        self.logger = get_root_logger()
+
+        if self.save_best is not None:
+            self._init_rule(rule, self.save_best)
+
+    def _init_rule(self, rule, key_indicator):
+        """Initialize rule, key_indicator, comparison_func, and best score.
+
+        Args:
+            rule (str | None): Comparison rule for best score.
+            key_indicator (str | None): Key indicator to determine the
+                comparison rule.
+        """
+        if rule not in self.rule_map and rule is not None:
+            raise KeyError(f'rule must be greater, less or None, '
+                           f'but got {rule}.')
+
+        if rule is None:
+            if key_indicator != 'auto':
+                if any(key in key_indicator for key in self.greater_keys):
+                    rule = 'greater'
+                elif any(key in key_indicator for key in self.less_keys):
+                    rule = 'less'
+                else:
+                    raise ValueError(f'Cannot infer the rule for key '
+                                     f'{key_indicator}, thus a specific rule '
+                                     f'must be specified.')
+        self.rule = rule
+        self.key_indicator = key_indicator
+        if self.rule is not None:
+            self.compare_func = self.rule_map[self.rule]
+
+    def before_run(self, runner):
+        if self.save_best is not None:
+            if runner.meta is None:
+                warnings.warn('runner.meta is None. Creating a empty one.')
+                runner.meta = dict()
+            runner.meta.setdefault('hook_msgs', dict())
+
+    def before_train_epoch(self, runner):
+        """Evaluate the model only at the start of training."""
+        if not self.initial_epoch_flag:
+            return
+        if self.start is not None and runner.epoch >= self.start:
+            self.after_train_epoch(runner)
+        self.initial_epoch_flag = False
+
+    def evaluation_flag(self, runner):
+        """Judge whether to perform_evaluation after this epoch.
+
+        Returns:
+            bool: The flag indicating whether to perform evaluation.
+        """
+        if self.start is None:
+            if not self.every_n_epochs(runner, self.interval):
+                # No evaluation during the interval epochs.
+                return False
+        elif (runner.epoch + 1) < self.start:
+            # No evaluation if start is larger than the current epoch.
+            return False
+        else:
+            # Evaluation only at epochs 3, 5, 7... if start==3 and interval==2
+            if (runner.epoch + 1 - self.start) % self.interval:
+                return False
+        return True
+
+    def after_train_epoch(self, runner):
+        if not self.by_epoch or not self.evaluation_flag(runner):
+            return
+        from mmdet.apis import single_gpu_test
+        results = single_gpu_test(runner.model, self.dataloader, show=False)
+        key_score = self.evaluate(runner, results)
+        if self.save_best:
+            self.save_best_checkpoint(runner, key_score)
+
+    def after_train_iter(self, runner):
+        if self.by_epoch or not self.every_n_iters(runner, self.interval):
+            return
+        from mmdet.apis import single_gpu_test
+        results = single_gpu_test(runner.model, self.dataloader, show=False)
+        key_score = self.evaluate(runner, results)
+        if self.save_best:
+            self.save_best_checkpoint(runner, key_score)
+
+    def save_best_checkpoint(self, runner, key_score):
+        best_score = runner.meta['hook_msgs'].get(
+            'best_score', self.init_value_map[self.rule])
+        if self.compare_func(key_score, best_score):
+            best_score = key_score
+            runner.meta['hook_msgs']['best_score'] = best_score
+            last_ckpt = runner.meta['hook_msgs']['last_ckpt']
+            runner.meta['hook_msgs']['best_ckpt'] = last_ckpt
+            mmcv.symlink(
+                last_ckpt,
+                osp.join(runner.work_dir, f'best_{self.key_indicator}.pth'))
+            time_stamp = runner.epoch + 1 if self.by_epoch else runner.iter + 1
+            self.logger.info(f'Now best checkpoint is epoch_{time_stamp}.pth.'
+                             f'Best {self.key_indicator} is {best_score:0.4f}')
+
+    def evaluate(self, runner, results):
+        eval_res = self.dataloader.dataset.evaluate(
+            results, logger=runner.logger, **self.eval_kwargs)
+        for name, val in eval_res.items():
+            runner.log_buffer.output[name] = val
+        runner.log_buffer.ready = True
+        if self.save_best is not None:
+            if self.key_indicator == 'auto':
+                # infer from eval_results
+                self._init_rule(self.rule, list(eval_res.keys())[0])
+            return eval_res[self.key_indicator]
+        else:
+            return None
+
+
+class DistEvalHook(EvalHook):
+    """Distributed evaluation hook.
+
+    Notes:
+        If new arguments are added, tools/test.py may be effected.
+
+    Attributes:
+        dataloader (DataLoader): A PyTorch dataloader.
+        start (int, optional): Evaluation starting epoch. It enables evaluation
+            before the training starts if ``start`` <= the resuming epoch.
+            If None, whether to evaluate is merely decided by ``interval``.
+            Default: None.
+        interval (int): Evaluation interval (by epochs). Default: 1.
+        tmpdir (str | None): Temporary directory to save the results of all
+            processes. Default: None.
+        gpu_collect (bool): Whether to use gpu or cpu to collect results.
+            Default: False.
+        save_best (str, optional): If a metric is specified, it would measure
+            the best checkpoint during evaluation. The information about best
+            checkpoint would be save in best.json.
+            Options are the evaluation metrics to the test dataset. e.g.,
+            ``bbox_mAP``, ``segm_mAP`` for bbox detection and instance
+            segmentation. ``AR@100`` for proposal recall. If ``save_best`` is
+            ``auto``, the first key will be used. The interval of
+            ``CheckpointHook`` should device EvalHook. Default: None.
+        rule (str | None): Comparison rule for best score. If set to None,
+            it will infer a reasonable rule. Default: 'None'.
+        broadcast_bn_buffer (bool): Whether to broadcast the
+            buffer(running_mean and running_var) of rank 0 to other rank
+            before evaluation. Default: True.
+        **eval_kwargs: Evaluation arguments fed into the evaluate function of
+            the dataset.
+    """
+
+    def __init__(self,
+                 dataloader,
+                 start=None,
+                 interval=1,
+                 by_epoch=True,
+                 tmpdir=None,
+                 gpu_collect=False,
+                 save_best=None,
+                 rule=None,
+                 broadcast_bn_buffer=True,
+                 **eval_kwargs):
+        super().__init__(
+            dataloader,
+            start=start,
+            interval=interval,
+            by_epoch=by_epoch,
+            save_best=save_best,
+            rule=rule,
+            **eval_kwargs)
+        self.broadcast_bn_buffer = broadcast_bn_buffer
+        self.tmpdir = tmpdir
+        self.gpu_collect = gpu_collect
+
+    def _broadcast_bn_buffer(self, runner):
+        # Synchronization of BatchNorm's buffer (running_mean
+        # and running_var) is not supported in the DDP of pytorch,
+        # which may cause the inconsistent performance of models in
+        # different ranks, so we broadcast BatchNorm's buffers
+        # of rank 0 to other ranks to avoid this.
+        if self.broadcast_bn_buffer:
+            model = runner.model
+            for name, module in model.named_modules():
+                if isinstance(module,
+                              _BatchNorm) and module.track_running_stats:
+                    dist.broadcast(module.running_var, 0)
+                    dist.broadcast(module.running_mean, 0)
+
+    def after_train_epoch(self, runner):
+        if not self.by_epoch or not self.evaluation_flag(runner):
+            return
+
+        if self.broadcast_bn_buffer:
+            self._broadcast_bn_buffer(runner)
+
+        from mmdet.apis import multi_gpu_test
+        tmpdir = self.tmpdir
+        if tmpdir is None:
+            tmpdir = osp.join(runner.work_dir, '.eval_hook')
+        results = multi_gpu_test(
+            runner.model,
+            self.dataloader,
+            tmpdir=tmpdir,
+            gpu_collect=self.gpu_collect)
+        if runner.rank == 0:
+            print('\n')
+            key_score = self.evaluate(runner, results)
+            if self.save_best:
+                self.save_best_checkpoint(runner, key_score)
+
+    def after_train_iter(self, runner):
+        if self.by_epoch or not self.every_n_iters(runner, self.interval):
+            return
+
+        if self.broadcast_bn_buffer:
+            self._broadcast_bn_buffer(runner)
+
+        from mmdet.apis import multi_gpu_test
+        tmpdir = self.tmpdir
+        if tmpdir is None:
+            tmpdir = osp.join(runner.work_dir, '.eval_hook')
+        results = multi_gpu_test(
+            runner.model,
+            self.dataloader,
+            tmpdir=tmpdir,
+            gpu_collect=self.gpu_collect)
+        if runner.rank == 0:
+            print('\n')
+            key_score = self.evaluate(runner, results)
+            if self.save_best:
+                self.save_best_checkpoint(runner, key_score)
diff --git a/python/autoCut/models/swin/mmdet/core/evaluation/mean_ap.py b/python/autoCut/models/swin/mmdet/core/evaluation/mean_ap.py
new file mode 100644
index 0000000..1d653a3
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/evaluation/mean_ap.py
@@ -0,0 +1,469 @@
+from multiprocessing import Pool
+
+import mmcv
+import numpy as np
+from mmcv.utils import print_log
+from terminaltables import AsciiTable
+
+from .bbox_overlaps import bbox_overlaps
+from .class_names import get_classes
+
+
+def average_precision(recalls, precisions, mode='area'):
+    """Calculate average precision (for single or multiple scales).
+
+    Args:
+        recalls (ndarray): shape (num_scales, num_dets) or (num_dets, )
+        precisions (ndarray): shape (num_scales, num_dets) or (num_dets, )
+        mode (str): 'area' or '11points', 'area' means calculating the area
+            under precision-recall curve, '11points' means calculating
+            the average precision of recalls at [0, 0.1, ..., 1]
+
+    Returns:
+        float or ndarray: calculated average precision
+    """
+    no_scale = False
+    if recalls.ndim == 1:
+        no_scale = True
+        recalls = recalls[np.newaxis, :]
+        precisions = precisions[np.newaxis, :]
+    assert recalls.shape == precisions.shape and recalls.ndim == 2
+    num_scales = recalls.shape[0]
+    ap = np.zeros(num_scales, dtype=np.float32)
+    if mode == 'area':
+        zeros = np.zeros((num_scales, 1), dtype=recalls.dtype)
+        ones = np.ones((num_scales, 1), dtype=recalls.dtype)
+        mrec = np.hstack((zeros, recalls, ones))
+        mpre = np.hstack((zeros, precisions, zeros))
+        for i in range(mpre.shape[1] - 1, 0, -1):
+            mpre[:, i - 1] = np.maximum(mpre[:, i - 1], mpre[:, i])
+        for i in range(num_scales):
+            ind = np.where(mrec[i, 1:] != mrec[i, :-1])[0]
+            ap[i] = np.sum(
+                (mrec[i, ind + 1] - mrec[i, ind]) * mpre[i, ind + 1])
+    elif mode == '11points':
+        for i in range(num_scales):
+            for thr in np.arange(0, 1 + 1e-3, 0.1):
+                precs = precisions[i, recalls[i, :] >= thr]
+                prec = precs.max() if precs.size > 0 else 0
+                ap[i] += prec
+        ap /= 11
+    else:
+        raise ValueError(
+            'Unrecognized mode, only "area" and "11points" are supported')
+    if no_scale:
+        ap = ap[0]
+    return ap
+
+
+def tpfp_imagenet(det_bboxes,
+                  gt_bboxes,
+                  gt_bboxes_ignore=None,
+                  default_iou_thr=0.5,
+                  area_ranges=None):
+    """Check if detected bboxes are true positive or false positive.
+
+    Args:
+        det_bbox (ndarray): Detected bboxes of this image, of shape (m, 5).
+        gt_bboxes (ndarray): GT bboxes of this image, of shape (n, 4).
+        gt_bboxes_ignore (ndarray): Ignored gt bboxes of this image,
+            of shape (k, 4). Default: None
+        default_iou_thr (float): IoU threshold to be considered as matched for
+            medium and large bboxes (small ones have special rules).
+            Default: 0.5.
+        area_ranges (list[tuple] | None): Range of bbox areas to be evaluated,
+            in the format [(min1, max1), (min2, max2), ...]. Default: None.
+
+    Returns:
+        tuple[np.ndarray]: (tp, fp) whose elements are 0 and 1. The shape of
+            each array is (num_scales, m).
+    """
+    # an indicator of ignored gts
+    gt_ignore_inds = np.concatenate(
+        (np.zeros(gt_bboxes.shape[0], dtype=np.bool),
+         np.ones(gt_bboxes_ignore.shape[0], dtype=np.bool)))
+    # stack gt_bboxes and gt_bboxes_ignore for convenience
+    gt_bboxes = np.vstack((gt_bboxes, gt_bboxes_ignore))
+
+    num_dets = det_bboxes.shape[0]
+    num_gts = gt_bboxes.shape[0]
+    if area_ranges is None:
+        area_ranges = [(None, None)]
+    num_scales = len(area_ranges)
+    # tp and fp are of shape (num_scales, num_gts), each row is tp or fp
+    # of a certain scale.
+    tp = np.zeros((num_scales, num_dets), dtype=np.float32)
+    fp = np.zeros((num_scales, num_dets), dtype=np.float32)
+    if gt_bboxes.shape[0] == 0:
+        if area_ranges == [(None, None)]:
+            fp[...] = 1
+        else:
+            det_areas = (det_bboxes[:, 2] - det_bboxes[:, 0]) * (
+                det_bboxes[:, 3] - det_bboxes[:, 1])
+            for i, (min_area, max_area) in enumerate(area_ranges):
+                fp[i, (det_areas >= min_area) & (det_areas < max_area)] = 1
+        return tp, fp
+    ious = bbox_overlaps(det_bboxes, gt_bboxes - 1)
+    gt_w = gt_bboxes[:, 2] - gt_bboxes[:, 0]
+    gt_h = gt_bboxes[:, 3] - gt_bboxes[:, 1]
+    iou_thrs = np.minimum((gt_w * gt_h) / ((gt_w + 10.0) * (gt_h + 10.0)),
+                          default_iou_thr)
+    # sort all detections by scores in descending order
+    sort_inds = np.argsort(-det_bboxes[:, -1])
+    for k, (min_area, max_area) in enumerate(area_ranges):
+        gt_covered = np.zeros(num_gts, dtype=bool)
+        # if no area range is specified, gt_area_ignore is all False
+        if min_area is None:
+            gt_area_ignore = np.zeros_like(gt_ignore_inds, dtype=bool)
+        else:
+            gt_areas = gt_w * gt_h
+            gt_area_ignore = (gt_areas < min_area) | (gt_areas >= max_area)
+        for i in sort_inds:
+            max_iou = -1
+            matched_gt = -1
+            # find best overlapped available gt
+            for j in range(num_gts):
+                # different from PASCAL VOC: allow finding other gts if the
+                # best overlapped ones are already matched by other det bboxes
+                if gt_covered[j]:
+                    continue
+                elif ious[i, j] >= iou_thrs[j] and ious[i, j] > max_iou:
+                    max_iou = ious[i, j]
+                    matched_gt = j
+            # there are 4 cases for a det bbox:
+            # 1. it matches a gt, tp = 1, fp = 0
+            # 2. it matches an ignored gt, tp = 0, fp = 0
+            # 3. it matches no gt and within area range, tp = 0, fp = 1
+            # 4. it matches no gt but is beyond area range, tp = 0, fp = 0
+            if matched_gt >= 0:
+                gt_covered[matched_gt] = 1
+                if not (gt_ignore_inds[matched_gt]
+                        or gt_area_ignore[matched_gt]):
+                    tp[k, i] = 1
+            elif min_area is None:
+                fp[k, i] = 1
+            else:
+                bbox = det_bboxes[i, :4]
+                area = (bbox[2] - bbox[0]) * (bbox[3] - bbox[1])
+                if area >= min_area and area < max_area:
+                    fp[k, i] = 1
+    return tp, fp
+
+
+def tpfp_default(det_bboxes,
+                 gt_bboxes,
+                 gt_bboxes_ignore=None,
+                 iou_thr=0.5,
+                 area_ranges=None):
+    """Check if detected bboxes are true positive or false positive.
+
+    Args:
+        det_bbox (ndarray): Detected bboxes of this image, of shape (m, 5).
+        gt_bboxes (ndarray): GT bboxes of this image, of shape (n, 4).
+        gt_bboxes_ignore (ndarray): Ignored gt bboxes of this image,
+            of shape (k, 4). Default: None
+        iou_thr (float): IoU threshold to be considered as matched.
+            Default: 0.5.
+        area_ranges (list[tuple] | None): Range of bbox areas to be evaluated,
+            in the format [(min1, max1), (min2, max2), ...]. Default: None.
+
+    Returns:
+        tuple[np.ndarray]: (tp, fp) whose elements are 0 and 1. The shape of
+            each array is (num_scales, m).
+    """
+    # an indicator of ignored gts
+    gt_ignore_inds = np.concatenate(
+        (np.zeros(gt_bboxes.shape[0], dtype=np.bool),
+         np.ones(gt_bboxes_ignore.shape[0], dtype=np.bool)))
+    # stack gt_bboxes and gt_bboxes_ignore for convenience
+    gt_bboxes = np.vstack((gt_bboxes, gt_bboxes_ignore))
+
+    num_dets = det_bboxes.shape[0]
+    num_gts = gt_bboxes.shape[0]
+    if area_ranges is None:
+        area_ranges = [(None, None)]
+    num_scales = len(area_ranges)
+    # tp and fp are of shape (num_scales, num_gts), each row is tp or fp of
+    # a certain scale
+    tp = np.zeros((num_scales, num_dets), dtype=np.float32)
+    fp = np.zeros((num_scales, num_dets), dtype=np.float32)
+
+    # if there is no gt bboxes in this image, then all det bboxes
+    # within area range are false positives
+    if gt_bboxes.shape[0] == 0:
+        if area_ranges == [(None, None)]:
+            fp[...] = 1
+        else:
+            det_areas = (det_bboxes[:, 2] - det_bboxes[:, 0]) * (
+                det_bboxes[:, 3] - det_bboxes[:, 1])
+            for i, (min_area, max_area) in enumerate(area_ranges):
+                fp[i, (det_areas >= min_area) & (det_areas < max_area)] = 1
+        return tp, fp
+
+    ious = bbox_overlaps(det_bboxes, gt_bboxes)
+    # for each det, the max iou with all gts
+    ious_max = ious.max(axis=1)
+    # for each det, which gt overlaps most with it
+    ious_argmax = ious.argmax(axis=1)
+    # sort all dets in descending order by scores
+    sort_inds = np.argsort(-det_bboxes[:, -1])
+    for k, (min_area, max_area) in enumerate(area_ranges):
+        gt_covered = np.zeros(num_gts, dtype=bool)
+        # if no area range is specified, gt_area_ignore is all False
+        if min_area is None:
+            gt_area_ignore = np.zeros_like(gt_ignore_inds, dtype=bool)
+        else:
+            gt_areas = (gt_bboxes[:, 2] - gt_bboxes[:, 0]) * (
+                gt_bboxes[:, 3] - gt_bboxes[:, 1])
+            gt_area_ignore = (gt_areas < min_area) | (gt_areas >= max_area)
+        for i in sort_inds:
+            if ious_max[i] >= iou_thr:
+                matched_gt = ious_argmax[i]
+                if not (gt_ignore_inds[matched_gt]
+                        or gt_area_ignore[matched_gt]):
+                    if not gt_covered[matched_gt]:
+                        gt_covered[matched_gt] = True
+                        tp[k, i] = 1
+                    else:
+                        fp[k, i] = 1
+                # otherwise ignore this detected bbox, tp = 0, fp = 0
+            elif min_area is None:
+                fp[k, i] = 1
+            else:
+                bbox = det_bboxes[i, :4]
+                area = (bbox[2] - bbox[0]) * (bbox[3] - bbox[1])
+                if area >= min_area and area < max_area:
+                    fp[k, i] = 1
+    return tp, fp
+
+
+def get_cls_results(det_results, annotations, class_id):
+    """Get det results and gt information of a certain class.
+
+    Args:
+        det_results (list[list]): Same as `eval_map()`.
+        annotations (list[dict]): Same as `eval_map()`.
+        class_id (int): ID of a specific class.
+
+    Returns:
+        tuple[list[np.ndarray]]: detected bboxes, gt bboxes, ignored gt bboxes
+    """
+    cls_dets = [img_res[class_id] for img_res in det_results]
+    cls_gts = []
+    cls_gts_ignore = []
+    for ann in annotations:
+        gt_inds = ann['labels'] == class_id
+        cls_gts.append(ann['bboxes'][gt_inds, :])
+
+        if ann.get('labels_ignore', None) is not None:
+            ignore_inds = ann['labels_ignore'] == class_id
+            cls_gts_ignore.append(ann['bboxes_ignore'][ignore_inds, :])
+        else:
+            cls_gts_ignore.append(np.empty((0, 4), dtype=np.float32))
+
+    return cls_dets, cls_gts, cls_gts_ignore
+
+
+def eval_map(det_results,
+             annotations,
+             scale_ranges=None,
+             iou_thr=0.5,
+             dataset=None,
+             logger=None,
+             tpfp_fn=None,
+             nproc=4):
+    """Evaluate mAP of a dataset.
+
+    Args:
+        det_results (list[list]): [[cls1_det, cls2_det, ...], ...].
+            The outer list indicates images, and the inner list indicates
+            per-class detected bboxes.
+        annotations (list[dict]): Ground truth annotations where each item of
+            the list indicates an image. Keys of annotations are:
+
+            - `bboxes`: numpy array of shape (n, 4)
+            - `labels`: numpy array of shape (n, )
+            - `bboxes_ignore` (optional): numpy array of shape (k, 4)
+            - `labels_ignore` (optional): numpy array of shape (k, )
+        scale_ranges (list[tuple] | None): Range of scales to be evaluated,
+            in the format [(min1, max1), (min2, max2), ...]. A range of
+            (32, 64) means the area range between (32**2, 64**2).
+            Default: None.
+        iou_thr (float): IoU threshold to be considered as matched.
+            Default: 0.5.
+        dataset (list[str] | str | None): Dataset name or dataset classes,
+            there are minor differences in metrics for different datsets, e.g.
+            "voc07", "imagenet_det", etc. Default: None.
+        logger (logging.Logger | str | None): The way to print the mAP
+            summary. See `mmcv.utils.print_log()` for details. Default: None.
+        tpfp_fn (callable | None): The function used to determine true/
+            false positives. If None, :func:`tpfp_default` is used as default
+            unless dataset is 'det' or 'vid' (:func:`tpfp_imagenet` in this
+            case). If it is given as a function, then this function is used
+            to evaluate tp & fp. Default None.
+        nproc (int): Processes used for computing TP and FP.
+            Default: 4.
+
+    Returns:
+        tuple: (mAP, [dict, dict, ...])
+    """
+    assert len(det_results) == len(annotations)
+
+    num_imgs = len(det_results)
+    num_scales = len(scale_ranges) if scale_ranges is not None else 1
+    num_classes = len(det_results[0])  # positive class num
+    area_ranges = ([(rg[0]**2, rg[1]**2) for rg in scale_ranges]
+                   if scale_ranges is not None else None)
+
+    pool = Pool(nproc)
+    eval_results = []
+    for i in range(num_classes):
+        # get gt and det bboxes of this class
+        cls_dets, cls_gts, cls_gts_ignore = get_cls_results(
+            det_results, annotations, i)
+        # choose proper function according to datasets to compute tp and fp
+        if tpfp_fn is None:
+            if dataset in ['det', 'vid']:
+                tpfp_fn = tpfp_imagenet
+            else:
+                tpfp_fn = tpfp_default
+        if not callable(tpfp_fn):
+            raise ValueError(
+                f'tpfp_fn has to be a function or None, but got {tpfp_fn}')
+
+        # compute tp and fp for each image with multiple processes
+        tpfp = pool.starmap(
+            tpfp_fn,
+            zip(cls_dets, cls_gts, cls_gts_ignore,
+                [iou_thr for _ in range(num_imgs)],
+                [area_ranges for _ in range(num_imgs)]))
+        tp, fp = tuple(zip(*tpfp))
+        # calculate gt number of each scale
+        # ignored gts or gts beyond the specific scale are not counted
+        num_gts = np.zeros(num_scales, dtype=int)
+        for j, bbox in enumerate(cls_gts):
+            if area_ranges is None:
+                num_gts[0] += bbox.shape[0]
+            else:
+                gt_areas = (bbox[:, 2] - bbox[:, 0]) * (
+                    bbox[:, 3] - bbox[:, 1])
+                for k, (min_area, max_area) in enumerate(area_ranges):
+                    num_gts[k] += np.sum((gt_areas >= min_area)
+                                         & (gt_areas < max_area))
+        # sort all det bboxes by score, also sort tp and fp
+        cls_dets = np.vstack(cls_dets)
+        num_dets = cls_dets.shape[0]
+        sort_inds = np.argsort(-cls_dets[:, -1])
+        tp = np.hstack(tp)[:, sort_inds]
+        fp = np.hstack(fp)[:, sort_inds]
+        # calculate recall and precision with tp and fp
+        tp = np.cumsum(tp, axis=1)
+        fp = np.cumsum(fp, axis=1)
+        eps = np.finfo(np.float32).eps
+        recalls = tp / np.maximum(num_gts[:, np.newaxis], eps)
+        precisions = tp / np.maximum((tp + fp), eps)
+        # calculate AP
+        if scale_ranges is None:
+            recalls = recalls[0, :]
+            precisions = precisions[0, :]
+            num_gts = num_gts.item()
+        mode = 'area' if dataset != 'voc07' else '11points'
+        ap = average_precision(recalls, precisions, mode)
+        eval_results.append({
+            'num_gts': num_gts,
+            'num_dets': num_dets,
+            'recall': recalls,
+            'precision': precisions,
+            'ap': ap
+        })
+    pool.close()
+    if scale_ranges is not None:
+        # shape (num_classes, num_scales)
+        all_ap = np.vstack([cls_result['ap'] for cls_result in eval_results])
+        all_num_gts = np.vstack(
+            [cls_result['num_gts'] for cls_result in eval_results])
+        mean_ap = []
+        for i in range(num_scales):
+            if np.any(all_num_gts[:, i] > 0):
+                mean_ap.append(all_ap[all_num_gts[:, i] > 0, i].mean())
+            else:
+                mean_ap.append(0.0)
+    else:
+        aps = []
+        for cls_result in eval_results:
+            if cls_result['num_gts'] > 0:
+                aps.append(cls_result['ap'])
+        mean_ap = np.array(aps).mean().item() if aps else 0.0
+
+    print_map_summary(
+        mean_ap, eval_results, dataset, area_ranges, logger=logger)
+
+    return mean_ap, eval_results
+
+
+def print_map_summary(mean_ap,
+                      results,
+                      dataset=None,
+                      scale_ranges=None,
+                      logger=None):
+    """Print mAP and results of each class.
+
+    A table will be printed to show the gts/dets/recall/AP of each class and
+    the mAP.
+
+    Args:
+        mean_ap (float): Calculated from `eval_map()`.
+        results (list[dict]): Calculated from `eval_map()`.
+        dataset (list[str] | str | None): Dataset name or dataset classes.
+        scale_ranges (list[tuple] | None): Range of scales to be evaluated.
+        logger (logging.Logger | str | None): The way to print the mAP
+            summary. See `mmcv.utils.print_log()` for details. Default: None.
+    """
+
+    if logger == 'silent':
+        return
+
+    if isinstance(results[0]['ap'], np.ndarray):
+        num_scales = len(results[0]['ap'])
+    else:
+        num_scales = 1
+
+    if scale_ranges is not None:
+        assert len(scale_ranges) == num_scales
+
+    num_classes = len(results)
+
+    recalls = np.zeros((num_scales, num_classes), dtype=np.float32)
+    aps = np.zeros((num_scales, num_classes), dtype=np.float32)
+    num_gts = np.zeros((num_scales, num_classes), dtype=int)
+    for i, cls_result in enumerate(results):
+        if cls_result['recall'].size > 0:
+            recalls[:, i] = np.array(cls_result['recall'], ndmin=2)[:, -1]
+        aps[:, i] = cls_result['ap']
+        num_gts[:, i] = cls_result['num_gts']
+
+    if dataset is None:
+        label_names = [str(i) for i in range(num_classes)]
+    elif mmcv.is_str(dataset):
+        label_names = get_classes(dataset)
+    else:
+        label_names = dataset
+
+    if not isinstance(mean_ap, list):
+        mean_ap = [mean_ap]
+
+    header = ['class', 'gts', 'dets', 'recall', 'ap']
+    for i in range(num_scales):
+        if scale_ranges is not None:
+            print_log(f'Scale range {scale_ranges[i]}', logger=logger)
+        table_data = [header]
+        for j in range(num_classes):
+            row_data = [
+                label_names[j], num_gts[i, j], results[j]['num_dets'],
+                f'{recalls[i, j]:.3f}', f'{aps[i, j]:.3f}'
+            ]
+            table_data.append(row_data)
+        table_data.append(['mAP', '', '', '', f'{mean_ap[i]:.3f}'])
+        table = AsciiTable(table_data)
+        table.inner_footing_row_border = True
+        print_log('\n' + table.table, logger=logger)
diff --git a/python/autoCut/models/swin/mmdet/core/evaluation/recall.py b/python/autoCut/models/swin/mmdet/core/evaluation/recall.py
new file mode 100644
index 0000000..23ec744
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/evaluation/recall.py
@@ -0,0 +1,189 @@
+from collections.abc import Sequence
+
+import numpy as np
+from mmcv.utils import print_log
+from terminaltables import AsciiTable
+
+from .bbox_overlaps import bbox_overlaps
+
+
+def _recalls(all_ious, proposal_nums, thrs):
+
+    img_num = all_ious.shape[0]
+    total_gt_num = sum([ious.shape[0] for ious in all_ious])
+
+    _ious = np.zeros((proposal_nums.size, total_gt_num), dtype=np.float32)
+    for k, proposal_num in enumerate(proposal_nums):
+        tmp_ious = np.zeros(0)
+        for i in range(img_num):
+            ious = all_ious[i][:, :proposal_num].copy()
+            gt_ious = np.zeros((ious.shape[0]))
+            if ious.size == 0:
+                tmp_ious = np.hstack((tmp_ious, gt_ious))
+                continue
+            for j in range(ious.shape[0]):
+                gt_max_overlaps = ious.argmax(axis=1)
+                max_ious = ious[np.arange(0, ious.shape[0]), gt_max_overlaps]
+                gt_idx = max_ious.argmax()
+                gt_ious[j] = max_ious[gt_idx]
+                box_idx = gt_max_overlaps[gt_idx]
+                ious[gt_idx, :] = -1
+                ious[:, box_idx] = -1
+            tmp_ious = np.hstack((tmp_ious, gt_ious))
+        _ious[k, :] = tmp_ious
+
+    _ious = np.fliplr(np.sort(_ious, axis=1))
+    recalls = np.zeros((proposal_nums.size, thrs.size))
+    for i, thr in enumerate(thrs):
+        recalls[:, i] = (_ious >= thr).sum(axis=1) / float(total_gt_num)
+
+    return recalls
+
+
+def set_recall_param(proposal_nums, iou_thrs):
+    """Check proposal_nums and iou_thrs and set correct format."""
+    if isinstance(proposal_nums, Sequence):
+        _proposal_nums = np.array(proposal_nums)
+    elif isinstance(proposal_nums, int):
+        _proposal_nums = np.array([proposal_nums])
+    else:
+        _proposal_nums = proposal_nums
+
+    if iou_thrs is None:
+        _iou_thrs = np.array([0.5])
+    elif isinstance(iou_thrs, Sequence):
+        _iou_thrs = np.array(iou_thrs)
+    elif isinstance(iou_thrs, float):
+        _iou_thrs = np.array([iou_thrs])
+    else:
+        _iou_thrs = iou_thrs
+
+    return _proposal_nums, _iou_thrs
+
+
+def eval_recalls(gts,
+                 proposals,
+                 proposal_nums=None,
+                 iou_thrs=0.5,
+                 logger=None):
+    """Calculate recalls.
+
+    Args:
+        gts (list[ndarray]): a list of arrays of shape (n, 4)
+        proposals (list[ndarray]): a list of arrays of shape (k, 4) or (k, 5)
+        proposal_nums (int | Sequence[int]): Top N proposals to be evaluated.
+        iou_thrs (float | Sequence[float]): IoU thresholds. Default: 0.5.
+        logger (logging.Logger | str | None): The way to print the recall
+            summary. See `mmcv.utils.print_log()` for details. Default: None.
+
+    Returns:
+        ndarray: recalls of different ious and proposal nums
+    """
+
+    img_num = len(gts)
+    assert img_num == len(proposals)
+
+    proposal_nums, iou_thrs = set_recall_param(proposal_nums, iou_thrs)
+
+    all_ious = []
+    for i in range(img_num):
+        if proposals[i].ndim == 2 and proposals[i].shape[1] == 5:
+            scores = proposals[i][:, 4]
+            sort_idx = np.argsort(scores)[::-1]
+            img_proposal = proposals[i][sort_idx, :]
+        else:
+            img_proposal = proposals[i]
+        prop_num = min(img_proposal.shape[0], proposal_nums[-1])
+        if gts[i] is None or gts[i].shape[0] == 0:
+            ious = np.zeros((0, img_proposal.shape[0]), dtype=np.float32)
+        else:
+            ious = bbox_overlaps(gts[i], img_proposal[:prop_num, :4])
+        all_ious.append(ious)
+    all_ious = np.array(all_ious)
+    recalls = _recalls(all_ious, proposal_nums, iou_thrs)
+
+    print_recall_summary(recalls, proposal_nums, iou_thrs, logger=logger)
+    return recalls
+
+
+def print_recall_summary(recalls,
+                         proposal_nums,
+                         iou_thrs,
+                         row_idxs=None,
+                         col_idxs=None,
+                         logger=None):
+    """Print recalls in a table.
+
+    Args:
+        recalls (ndarray): calculated from `bbox_recalls`
+        proposal_nums (ndarray or list): top N proposals
+        iou_thrs (ndarray or list): iou thresholds
+        row_idxs (ndarray): which rows(proposal nums) to print
+        col_idxs (ndarray): which cols(iou thresholds) to print
+        logger (logging.Logger | str | None): The way to print the recall
+            summary. See `mmcv.utils.print_log()` for details. Default: None.
+    """
+    proposal_nums = np.array(proposal_nums, dtype=np.int32)
+    iou_thrs = np.array(iou_thrs)
+    if row_idxs is None:
+        row_idxs = np.arange(proposal_nums.size)
+    if col_idxs is None:
+        col_idxs = np.arange(iou_thrs.size)
+    row_header = [''] + iou_thrs[col_idxs].tolist()
+    table_data = [row_header]
+    for i, num in enumerate(proposal_nums[row_idxs]):
+        row = [f'{val:.3f}' for val in recalls[row_idxs[i], col_idxs].tolist()]
+        row.insert(0, num)
+        table_data.append(row)
+    table = AsciiTable(table_data)
+    print_log('\n' + table.table, logger=logger)
+
+
+def plot_num_recall(recalls, proposal_nums):
+    """Plot Proposal_num-Recalls curve.
+
+    Args:
+        recalls(ndarray or list): shape (k,)
+        proposal_nums(ndarray or list): same shape as `recalls`
+    """
+    if isinstance(proposal_nums, np.ndarray):
+        _proposal_nums = proposal_nums.tolist()
+    else:
+        _proposal_nums = proposal_nums
+    if isinstance(recalls, np.ndarray):
+        _recalls = recalls.tolist()
+    else:
+        _recalls = recalls
+
+    import matplotlib.pyplot as plt
+    f = plt.figure()
+    plt.plot([0] + _proposal_nums, [0] + _recalls)
+    plt.xlabel('Proposal num')
+    plt.ylabel('Recall')
+    plt.axis([0, proposal_nums.max(), 0, 1])
+    f.show()
+
+
+def plot_iou_recall(recalls, iou_thrs):
+    """Plot IoU-Recalls curve.
+
+    Args:
+        recalls(ndarray or list): shape (k,)
+        iou_thrs(ndarray or list): same shape as `recalls`
+    """
+    if isinstance(iou_thrs, np.ndarray):
+        _iou_thrs = iou_thrs.tolist()
+    else:
+        _iou_thrs = iou_thrs
+    if isinstance(recalls, np.ndarray):
+        _recalls = recalls.tolist()
+    else:
+        _recalls = recalls
+
+    import matplotlib.pyplot as plt
+    f = plt.figure()
+    plt.plot(_iou_thrs + [1.0], _recalls + [0.])
+    plt.xlabel('IoU')
+    plt.ylabel('Recall')
+    plt.axis([iou_thrs.min(), 1, 0, 1])
+    f.show()
diff --git a/python/autoCut/models/swin/mmdet/core/export/__init__.py b/python/autoCut/models/swin/mmdet/core/export/__init__.py
new file mode 100644
index 0000000..76589b1
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/export/__init__.py
@@ -0,0 +1,8 @@
+from .pytorch2onnx import (build_model_from_cfg,
+                           generate_inputs_and_wrap_model,
+                           preprocess_example_input)
+
+__all__ = [
+    'build_model_from_cfg', 'generate_inputs_and_wrap_model',
+    'preprocess_example_input'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/export/pytorch2onnx.py b/python/autoCut/models/swin/mmdet/core/export/pytorch2onnx.py
new file mode 100644
index 0000000..809a817
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/export/pytorch2onnx.py
@@ -0,0 +1,154 @@
+from functools import partial
+
+import mmcv
+import numpy as np
+import torch
+from mmcv.runner import load_checkpoint
+
+
+def generate_inputs_and_wrap_model(config_path,
+                                   checkpoint_path,
+                                   input_config,
+                                   cfg_options=None):
+    """Prepare sample input and wrap model for ONNX export.
+
+    The ONNX export API only accept args, and all inputs should be
+    torch.Tensor or corresponding types (such as tuple of tensor).
+    So we should call this function before exporting. This function will:
+
+    1. generate corresponding inputs which are used to execute the model.
+    2. Wrap the model's forward function.
+
+    For example, the MMDet models' forward function has a parameter
+    ``return_loss:bool``. As we want to set it as False while export API
+    supports neither bool type or kwargs. So we have to replace the forward
+    like: ``model.forward = partial(model.forward, return_loss=False)``
+
+    Args:
+        config_path (str): the OpenMMLab config for the model we want to
+            export to ONNX
+        checkpoint_path (str): Path to the corresponding checkpoint
+        input_config (dict): the exactly data in this dict depends on the
+            framework. For MMSeg, we can just declare the input shape,
+            and generate the dummy data accordingly. However, for MMDet,
+            we may pass the real img path, or the NMS will return None
+            as there is no legal bbox.
+
+    Returns:
+        tuple: (model, tensor_data) wrapped model which can be called by \
+        model(*tensor_data) and a list of inputs which are used to execute \
+            the model while exporting.
+    """
+
+    model = build_model_from_cfg(
+        config_path, checkpoint_path, cfg_options=cfg_options)
+    one_img, one_meta = preprocess_example_input(input_config)
+    tensor_data = [one_img]
+    model.forward = partial(
+        model.forward, img_metas=[[one_meta]], return_loss=False)
+
+    # pytorch has some bug in pytorch1.3, we have to fix it
+    # by replacing these existing op
+    opset_version = 11
+    # put the import within the function thus it will not cause import error
+    # when not using this function
+    try:
+        from mmcv.onnx.symbolic import register_extra_symbolics
+    except ModuleNotFoundError:
+        raise NotImplementedError('please update mmcv to version>=v1.0.4')
+    register_extra_symbolics(opset_version)
+
+    return model, tensor_data
+
+
+def build_model_from_cfg(config_path, checkpoint_path, cfg_options=None):
+    """Build a model from config and load the given checkpoint.
+
+    Args:
+        config_path (str): the OpenMMLab config for the model we want to
+            export to ONNX
+        checkpoint_path (str): Path to the corresponding checkpoint
+
+    Returns:
+        torch.nn.Module: the built model
+    """
+    from mmdet.models import build_detector
+
+    cfg = mmcv.Config.fromfile(config_path)
+    if cfg_options is not None:
+        cfg.merge_from_dict(cfg_options)
+    # import modules from string list.
+    if cfg.get('custom_imports', None):
+        from mmcv.utils import import_modules_from_strings
+        import_modules_from_strings(**cfg['custom_imports'])
+    # set cudnn_benchmark
+    if cfg.get('cudnn_benchmark', False):
+        torch.backends.cudnn.benchmark = True
+    cfg.model.pretrained = None
+    cfg.data.test.test_mode = True
+
+    # build the model
+    cfg.model.train_cfg = None
+    model = build_detector(cfg.model, test_cfg=cfg.get('test_cfg'))
+    load_checkpoint(model, checkpoint_path, map_location='cpu')
+    model.cpu().eval()
+    return model
+
+
+def preprocess_example_input(input_config):
+    """Prepare an example input image for ``generate_inputs_and_wrap_model``.
+
+    Args:
+        input_config (dict): customized config describing the example input.
+
+    Returns:
+        tuple: (one_img, one_meta), tensor of the example input image and \
+            meta information for the example input image.
+
+    Examples:
+        >>> from mmdet.core.export import preprocess_example_input
+        >>> input_config = {
+        >>>         'input_shape': (1,3,224,224),
+        >>>         'input_path': 'demo/demo.jpg',
+        >>>         'normalize_cfg': {
+        >>>             'mean': (123.675, 116.28, 103.53),
+        >>>             'std': (58.395, 57.12, 57.375)
+        >>>             }
+        >>>         }
+        >>> one_img, one_meta = preprocess_example_input(input_config)
+        >>> print(one_img.shape)
+        torch.Size([1, 3, 224, 224])
+        >>> print(one_meta)
+        {'img_shape': (224, 224, 3),
+        'ori_shape': (224, 224, 3),
+        'pad_shape': (224, 224, 3),
+        'filename': '<demo>.png',
+        'scale_factor': 1.0,
+        'flip': False}
+    """
+    input_path = input_config['input_path']
+    input_shape = input_config['input_shape']
+    one_img = mmcv.imread(input_path)
+    one_img = mmcv.imresize(one_img, input_shape[2:][::-1])
+    show_img = one_img.copy()
+    if 'normalize_cfg' in input_config.keys():
+        normalize_cfg = input_config['normalize_cfg']
+        mean = np.array(normalize_cfg['mean'], dtype=np.float32)
+        std = np.array(normalize_cfg['std'], dtype=np.float32)
+        to_rgb = normalize_cfg.get('to_rgb', True)
+        one_img = mmcv.imnormalize(one_img, mean, std, to_rgb=to_rgb)
+    one_img = one_img.transpose(2, 0, 1)
+    one_img = torch.from_numpy(one_img).unsqueeze(0).float().requires_grad_(
+        True)
+    (_, C, H, W) = input_shape
+    one_meta = {
+        'img_shape': (H, W, C),
+        'ori_shape': (H, W, C),
+        'pad_shape': (H, W, C),
+        'filename': '<demo>.png',
+        'scale_factor': 1.0,
+        'flip': False,
+        'show_img': show_img,
+    }
+
+    return one_img, one_meta
diff --git a/python/autoCut/models/swin/mmdet/core/mask/__init__.py b/python/autoCut/models/swin/mmdet/core/mask/__init__.py
new file mode 100644
index 0000000..ab1e88b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/mask/__init__.py
@@ -0,0 +1,8 @@
+from .mask_target import mask_target
+from .structures import BaseInstanceMasks, BitmapMasks, PolygonMasks
+from .utils import encode_mask_results, split_combined_polys
+
+__all__ = [
+    'split_combined_polys', 'mask_target', 'BaseInstanceMasks', 'BitmapMasks',
+    'PolygonMasks', 'encode_mask_results'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/mask/mask_target.py b/python/autoCut/models/swin/mmdet/core/mask/mask_target.py
new file mode 100644
index 0000000..15d26a8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/mask/mask_target.py
@@ -0,0 +1,122 @@
+import numpy as np
+import torch
+from torch.nn.modules.utils import _pair
+
+
+def mask_target(pos_proposals_list, pos_assigned_gt_inds_list, gt_masks_list,
+                cfg):
+    """Compute mask target for positive proposals in multiple images.
+
+    Args:
+        pos_proposals_list (list[Tensor]): Positive proposals in multiple
+            images.
+        pos_assigned_gt_inds_list (list[Tensor]): Assigned GT indices for each
+            positive proposals.
+        gt_masks_list (list[:obj:`BaseInstanceMasks`]): Ground truth masks of
+            each image.
+        cfg (dict): Config dict that specifies the mask size.
+
+    Returns:
+        list[Tensor]: Mask target of each image.
+
+    Example:
+        >>> import mmcv
+        >>> import mmdet
+        >>> from mmdet.core.mask import BitmapMasks
+        >>> from mmdet.core.mask.mask_target import *
+        >>> H, W = 17, 18
+        >>> cfg = mmcv.Config({'mask_size': (13, 14)})
+        >>> rng = np.random.RandomState(0)
+        >>> # Positive proposals (tl_x, tl_y, br_x, br_y) for each image
+        >>> pos_proposals_list = [
+        >>>     torch.Tensor([
+        >>>         [ 7.2425,  5.5929, 13.9414, 14.9541],
+        >>>         [ 7.3241,  3.6170, 16.3850, 15.3102],
+        >>>     ]),
+        >>>     torch.Tensor([
+        >>>         [ 4.8448, 6.4010, 7.0314, 9.7681],
+        >>>         [ 5.9790, 2.6989, 7.4416, 4.8580],
+        >>>         [ 0.0000, 0.0000, 0.1398, 9.8232],
+        >>>     ]),
+        >>> ]
+        >>> # Corresponding class index for each proposal for each image
+        >>> pos_assigned_gt_inds_list = [
+        >>>     torch.LongTensor([7, 0]),
+        >>>     torch.LongTensor([5, 4, 1]),
+        >>> ]
+        >>> # Ground truth mask for each true object for each image
+        >>> gt_masks_list = [
+        >>>     BitmapMasks(rng.rand(8, H, W), height=H, width=W),
+        >>>     BitmapMasks(rng.rand(6, H, W), height=H, width=W),
+        >>> ]
+        >>> mask_targets = mask_target(
+        >>>     pos_proposals_list, pos_assigned_gt_inds_list,
+        >>>     gt_masks_list, cfg)
+        >>> assert mask_targets.shape == (5,) + cfg['mask_size']
+    """
+    cfg_list = [cfg for _ in range(len(pos_proposals_list))]
+    mask_targets = map(mask_target_single, pos_proposals_list,
+                       pos_assigned_gt_inds_list, gt_masks_list, cfg_list)
+    mask_targets = list(mask_targets)
+    if len(mask_targets) > 0:
+        mask_targets = torch.cat(mask_targets)
+    return mask_targets
+
+
+def mask_target_single(pos_proposals, pos_assigned_gt_inds, gt_masks, cfg):
+    """Compute mask target for each positive proposal in the image.
+
+    Args:
+        pos_proposals (Tensor): Positive proposals.
+        pos_assigned_gt_inds (Tensor): Assigned GT inds of positive proposals.
+        gt_masks (:obj:`BaseInstanceMasks`): GT masks in the format of Bitmap
+            or Polygon.
+        cfg (dict): Config dict that indicate the mask size.
+
+    Returns:
+        Tensor: Mask target of each positive proposals in the image.
+
+    Example:
+        >>> import mmcv
+        >>> import mmdet
+        >>> from mmdet.core.mask import BitmapMasks
+        >>> from mmdet.core.mask.mask_target import *  # NOQA
+        >>> H, W = 32, 32
+        >>> cfg = mmcv.Config({'mask_size': (7, 11)})
+        >>> rng = np.random.RandomState(0)
+        >>> # Masks for each ground truth box (relative to the image)
+        >>> gt_masks_data = rng.rand(3, H, W)
+        >>> gt_masks = BitmapMasks(gt_masks_data, height=H, width=W)
+        >>> # Predicted positive boxes in one image
+        >>> pos_proposals = torch.FloatTensor([
+        >>>     [ 16.2,   5.5, 19.9, 20.9],
+        >>>     [ 17.3,  13.6, 19.3, 19.3],
+        >>>     [ 14.8,  16.4, 17.0, 23.7],
+        >>>     [  0.0,   0.0, 16.0, 16.0],
+        >>>     [  4.0,   0.0, 20.0, 16.0],
+        >>> ])
+        >>> # For each predicted proposal, its assignment to a gt mask
+        >>> pos_assigned_gt_inds = torch.LongTensor([0, 1, 2, 1, 1])
+        >>> mask_targets = mask_target_single(
+        >>>     pos_proposals, pos_assigned_gt_inds, gt_masks, cfg)
+        >>> assert mask_targets.shape == (5,) + cfg['mask_size']
+    """
+    device = pos_proposals.device
+    mask_size = _pair(cfg.mask_size)
+    num_pos = pos_proposals.size(0)
+    if num_pos > 0:
+        proposals_np = pos_proposals.cpu().numpy()
+        maxh, maxw = gt_masks.height, gt_masks.width
+        proposals_np[:, [0, 2]] = np.clip(proposals_np[:, [0, 2]], 0, maxw)
+        proposals_np[:, [1, 3]] = np.clip(proposals_np[:, [1, 3]], 0, maxh)
+        pos_assigned_gt_inds = pos_assigned_gt_inds.cpu().numpy()
+
+        mask_targets = gt_masks.crop_and_resize(
+            proposals_np, mask_size, device=device,
+            inds=pos_assigned_gt_inds).to_ndarray()
+
+        mask_targets = torch.from_numpy(mask_targets).float().to(device)
+    else:
+        mask_targets = pos_proposals.new_zeros((0, ) + mask_size)
+
+    return mask_targets
diff --git a/python/autoCut/models/swin/mmdet/core/mask/structures.py b/python/autoCut/models/swin/mmdet/core/mask/structures.py
new file mode 100644
index 0000000..d9ec577
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/mask/structures.py
@@ -0,0 +1,1024 @@
+from abc import ABCMeta, abstractmethod
+
+import cv2
+import mmcv
+import numpy as np
+import pycocotools.mask as maskUtils
+import torch
+from mmcv.ops.roi_align import roi_align
+
+
+class BaseInstanceMasks(metaclass=ABCMeta):
+    """Base class for instance masks."""
+
+    @abstractmethod
+    def rescale(self, scale, interpolation='nearest'):
+        """Rescale masks as large as possible while keeping the aspect ratio.
+        For details can refer to `mmcv.imrescale`.
+
+        Args:
+            scale (tuple[int]): The maximum size (h, w) of rescaled mask.
+            interpolation (str): Same as :func:`mmcv.imrescale`.
+
+        Returns:
+            BaseInstanceMasks: The rescaled masks.
+        """
+
+    @abstractmethod
+    def resize(self, out_shape, interpolation='nearest'):
+        """Resize masks to the given out_shape.
+
+        Args:
+            out_shape: Target (h, w) of resized mask.
+            interpolation (str): See :func:`mmcv.imresize`.
+
+        Returns:
+            BaseInstanceMasks: The resized masks.
+        """
+
+    @abstractmethod
+    def flip(self, flip_direction='horizontal'):
+        """Flip masks alone the given direction.
+
+        Args:
+            flip_direction (str): Either 'horizontal' or 'vertical'.
+
+        Returns:
+            BaseInstanceMasks: The flipped masks.
+        """
+
+    @abstractmethod
+    def pad(self, out_shape, pad_val):
+        """Pad masks to the given size of (h, w).
+
+        Args:
+            out_shape (tuple[int]): Target (h, w) of padded mask.
+            pad_val (int): The padded value.
+
+        Returns:
+            BaseInstanceMasks: The padded masks.
+        """
+
+    @abstractmethod
+    def crop(self, bbox):
+        """Crop each mask by the given bbox.
+
+        Args:
+            bbox (ndarray): Bbox in format [x1, y1, x2, y2], shape (4, ).
+
+        Return:
+            BaseInstanceMasks: The cropped masks.
+        """
+
+    @abstractmethod
+    def crop_and_resize(self,
+                        bboxes,
+                        out_shape,
+                        inds,
+                        device,
+                        interpolation='bilinear'):
+        """Crop and resize masks by the given bboxes.
+
+        This function is mainly used in mask targets computation.
+        It firstly align mask to bboxes by assigned_inds, then crop mask by the
+        assigned bbox and resize to the size of (mask_h, mask_w)
+
+        Args:
+            bboxes (Tensor): Bboxes in format [x1, y1, x2, y2], shape (N, 4)
+            out_shape (tuple[int]): Target (h, w) of resized mask
+            inds (ndarray): Indexes to assign masks to each bbox,
+                shape (N,) and values should be between [0, num_masks - 1].
+            device (str): Device of bboxes
+            interpolation (str): See `mmcv.imresize`
+
+        Return:
+            BaseInstanceMasks: the cropped and resized masks.
+        """
+
+    @abstractmethod
+    def expand(self, expanded_h, expanded_w, top, left):
+        """see :class:`Expand`."""
+
+    @property
+    @abstractmethod
+    def areas(self):
+        """ndarray: areas of each instance."""
+
+    @abstractmethod
+    def to_ndarray(self):
+        """Convert masks to the format of ndarray.
+
+        Return:
+            ndarray: Converted masks in the format of ndarray.
+        """
+
+    @abstractmethod
+    def to_tensor(self, dtype, device):
+        """Convert masks to the format of Tensor.
+
+        Args:
+            dtype (str): Dtype of converted mask.
+            device (torch.device): Device of converted masks.
+
+        Returns:
+            Tensor: Converted masks in the format of Tensor.
+        """
+
+    @abstractmethod
+    def translate(self,
+                  out_shape,
+                  offset,
+                  direction='horizontal',
+                  fill_val=0,
+                  interpolation='bilinear'):
+        """Translate the masks.
+
+        Args:
+            out_shape (tuple[int]): Shape for output mask, format (h, w).
+            offset (int | float): The offset for translate.
+            direction (str): The translate direction, either "horizontal"
+                or "vertical".
+            fill_val (int | float): Border value. Default 0.
+            interpolation (str): Same as :func:`mmcv.imtranslate`.
+
+        Returns:
+            Translated masks.
+        """
+
+    def shear(self,
+              out_shape,
+              magnitude,
+              direction='horizontal',
+              border_value=0,
+              interpolation='bilinear'):
+        """Shear the masks.
+
+        Args:
+            out_shape (tuple[int]): Shape for output mask, format (h, w).
+            magnitude (int | float): The magnitude used for shear.
+            direction (str): The shear direction, either "horizontal"
+                or "vertical".
+            border_value (int | tuple[int]): Value used in case of a
+                constant border. Default 0.
+            interpolation (str): Same as in :func:`mmcv.imshear`.
+
+        Returns:
+            ndarray: Sheared masks.
+        """
+
+    @abstractmethod
+    def rotate(self, out_shape, angle, center=None, scale=1.0, fill_val=0):
+        """Rotate the masks.
+
+        Args:
+            out_shape (tuple[int]): Shape for output mask, format (h, w).
+            angle (int | float): Rotation angle in degrees. Positive values
+                mean counter-clockwise rotation.
+            center (tuple[float], optional): Center point (w, h) of the
+                rotation in source image. If not specified, the center of
+                the image will be used.
+            scale (int | float): Isotropic scale factor.
+            fill_val (int | float): Border value. Default 0 for masks.
+
+        Returns:
+            Rotated masks.
+        """
+
+
+class BitmapMasks(BaseInstanceMasks):
+    """This class represents masks in the form of bitmaps.
+
+    Args:
+        masks (ndarray): ndarray of masks in shape (N, H, W), where N is
+            the number of objects.
+        height (int): height of masks
+        width (int): width of masks
+
+    Example:
+        >>> from mmdet.core.mask.structures import *  # NOQA
+        >>> num_masks, H, W = 3, 32, 32
+        >>> rng = np.random.RandomState(0)
+        >>> masks = (rng.rand(num_masks, H, W) > 0.1).astype(np.int)
+        >>> self = BitmapMasks(masks, height=H, width=W)
+
+        >>> # demo crop_and_resize
+        >>> num_boxes = 5
+        >>> bboxes = np.array([[0, 0, 30, 10.0]] * num_boxes)
+        >>> out_shape = (14, 14)
+        >>> inds = torch.randint(0, len(self), size=(num_boxes,))
+        >>> device = 'cpu'
+        >>> interpolation = 'bilinear'
+        >>> new = self.crop_and_resize(
+        ...     bboxes, out_shape, inds, device, interpolation)
+        >>> assert len(new) == num_boxes
+        >>> assert new.height, new.width == out_shape
+    """
+
+    def __init__(self, masks, height, width):
+        self.height = height
+        self.width = width
+        if len(masks) == 0:
+            self.masks = np.empty((0, self.height, self.width), dtype=np.uint8)
+        else:
+            assert isinstance(masks, (list, np.ndarray))
+            if isinstance(masks, list):
+                assert isinstance(masks[0], np.ndarray)
+                assert masks[0].ndim == 2  # (H, W)
+            else:
+                assert masks.ndim == 3  # (N, H, W)
+
+            self.masks = np.stack(masks).reshape(-1, height, width)
+            assert self.masks.shape[1] == self.height
+            assert self.masks.shape[2] == self.width
+
+    def __getitem__(self, index):
+        """Index the BitmapMask.
+
+        Args:
+            index (int | ndarray): Indices in the format of integer or ndarray.
+
+        Returns:
+            :obj:`BitmapMasks`: Indexed bitmap masks.
+        """
+        masks = self.masks[index].reshape(-1, self.height, self.width)
+        return BitmapMasks(masks, self.height, self.width)
+
+    def __iter__(self):
+        return iter(self.masks)
+
+    def __repr__(self):
+        s = self.__class__.__name__ + '('
+        s += f'num_masks={len(self.masks)}, '
+        s += f'height={self.height}, '
+        s += f'width={self.width})'
+        return s
+
+    def __len__(self):
+        """Number of masks."""
+        return len(self.masks)
+
+    def rescale(self, scale, interpolation='nearest'):
+        """See :func:`BaseInstanceMasks.rescale`."""
+        if len(self.masks) == 0:
+            new_w, new_h = mmcv.rescale_size((self.width, self.height), scale)
+            rescaled_masks = np.empty((0, new_h, new_w), dtype=np.uint8)
+        else:
+            rescaled_masks = np.stack([
+                mmcv.imrescale(mask, scale, interpolation=interpolation)
+                for mask in self.masks
+            ])
+        height, width = rescaled_masks.shape[1:]
+        return BitmapMasks(rescaled_masks, height, width)
+
+    def resize(self, out_shape, interpolation='nearest'):
+        """See :func:`BaseInstanceMasks.resize`."""
+        if len(self.masks) == 0:
+            resized_masks = np.empty((0, *out_shape), dtype=np.uint8)
+        else:
+            resized_masks = np.stack([
+                mmcv.imresize(
+                    mask, out_shape[::-1], interpolation=interpolation)
+                for mask in self.masks
+            ])
+        return BitmapMasks(resized_masks, *out_shape)
+
+    def flip(self, flip_direction='horizontal'):
+        """See :func:`BaseInstanceMasks.flip`."""
+        assert flip_direction in ('horizontal', 'vertical', 'diagonal')
+
+        if len(self.masks) == 0:
+            flipped_masks = self.masks
+        else:
+            flipped_masks = np.stack([
+                mmcv.imflip(mask, direction=flip_direction)
+                for mask in self.masks
+            ])
+        return BitmapMasks(flipped_masks, self.height, self.width)
+
+    def pad(self, out_shape, pad_val=0):
+        """See :func:`BaseInstanceMasks.pad`."""
+        if len(self.masks) == 0:
+            padded_masks = np.empty((0, *out_shape), dtype=np.uint8)
+        else:
+            padded_masks = np.stack([
+                mmcv.impad(mask, shape=out_shape, pad_val=pad_val)
+                for mask in self.masks
+            ])
+        return BitmapMasks(padded_masks, *out_shape)
+
+    def crop(self, bbox):
+        """See :func:`BaseInstanceMasks.crop`."""
+        assert isinstance(bbox, np.ndarray)
+        assert bbox.ndim == 1
+
+        # clip the boundary
+        bbox = bbox.copy()
+        bbox[0::2] = np.clip(bbox[0::2], 0, self.width)
+        bbox[1::2] = np.clip(bbox[1::2], 0, self.height)
+        x1, y1, x2, y2 = bbox
+        w = np.maximum(x2 - x1, 1)
+        h = np.maximum(y2 - y1, 1)
+
+        if len(self.masks) == 0:
+            cropped_masks = np.empty((0, h, w), dtype=np.uint8)
+        else:
+            cropped_masks = self.masks[:, y1:y1 + h, x1:x1 + w]
+        return BitmapMasks(cropped_masks, h, w)
+
+    def crop_and_resize(self,
+                        bboxes,
+                        out_shape,
+                        inds,
+                        device='cpu',
+                        interpolation='bilinear'):
+        """See :func:`BaseInstanceMasks.crop_and_resize`."""
+        if len(self.masks) == 0:
+            empty_masks = np.empty((0, *out_shape), dtype=np.uint8)
+            return BitmapMasks(empty_masks, *out_shape)
+
+        # convert bboxes to tensor
+        if isinstance(bboxes, np.ndarray):
+            bboxes = torch.from_numpy(bboxes).to(device=device)
+        if isinstance(inds, np.ndarray):
+            inds = torch.from_numpy(inds).to(device=device)
+
+        num_bbox = bboxes.shape[0]
+        fake_inds = torch.arange(
+            num_bbox, device=device).to(dtype=bboxes.dtype)[:, None]
+        rois = torch.cat([fake_inds, bboxes], dim=1)  # Nx5
+        rois = rois.to(device=device)
+        if num_bbox > 0:
+            gt_masks_th = torch.from_numpy(self.masks).to(device).index_select(
+                0, inds).to(dtype=rois.dtype)
+            targets = roi_align(gt_masks_th[:, None, :, :], rois, out_shape,
+                                1.0, 0, 'avg', True).squeeze(1)
+            resized_masks = (targets >= 0.5).cpu().numpy()
+        else:
+            resized_masks = []
+        return BitmapMasks(resized_masks, *out_shape)
+
+    def expand(self, expanded_h, expanded_w, top, left):
+        """See :func:`BaseInstanceMasks.expand`."""
+        if len(self.masks) == 0:
+            expanded_mask = np.empty((0, expanded_h, expanded_w),
+                                     dtype=np.uint8)
+        else:
+            expanded_mask = np.zeros((len(self), expanded_h, expanded_w),
+                                     dtype=np.uint8)
+            expanded_mask[:, top:top + self.height,
+                          left:left + self.width] = self.masks
+        return BitmapMasks(expanded_mask, expanded_h, expanded_w)
+
+    def translate(self,
+                  out_shape,
+                  offset,
+                  direction='horizontal',
+                  fill_val=0,
+                  interpolation='bilinear'):
+        """Translate the BitmapMasks.
+
+        Args:
+            out_shape (tuple[int]): Shape for output mask, format (h, w).
+            offset (int | float): The offset for translate.
+            direction (str): The translate direction, either "horizontal"
+                or "vertical".
+            fill_val (int | float): Border value. Default 0 for masks.
+            interpolation (str): Same as :func:`mmcv.imtranslate`.
+
+        Returns:
+            BitmapMasks: Translated BitmapMasks.
+
+        Example:
+            >>> from mmdet.core.mask.structures import BitmapMasks
+            >>> self = BitmapMasks.random(dtype=np.uint8)
+            >>> out_shape = (32, 32)
+            >>> offset = 4
+            >>> direction = 'horizontal'
+            >>> fill_val = 0
+            >>> interpolation = 'bilinear'
+            >>> # Note, There seem to be issues when:
+            >>> # * out_shape is different than self's shape
+            >>> # * the mask dtype is not supported by cv2.AffineWarp
+            >>> new = self.translate(out_shape, offset, direction, fill_val,
+            >>>                      interpolation)
+            >>> assert len(new) == len(self)
+            >>> assert new.height, new.width == out_shape
+        """
+        if len(self.masks) == 0:
+            translated_masks = np.empty((0, *out_shape), dtype=np.uint8)
+        else:
+            translated_masks = mmcv.imtranslate(
+                self.masks.transpose((1, 2, 0)),
+                offset,
+                direction,
+                border_value=fill_val,
+                interpolation=interpolation)
+            if translated_masks.ndim == 2:
+                translated_masks = translated_masks[:, :, None]
+            translated_masks = translated_masks.transpose(
+                (2, 0, 1)).astype(self.masks.dtype)
+        return BitmapMasks(translated_masks, *out_shape)
+
+    def shear(self,
+              out_shape,
+              magnitude,
+              direction='horizontal',
+              border_value=0,
+              interpolation='bilinear'):
+        """Shear the BitmapMasks.
+
+        Args:
+            out_shape (tuple[int]): Shape for output mask, format (h, w).
+            magnitude (int | float): The magnitude used for shear.
+            direction (str): The shear direction, either "horizontal"
+                or "vertical".
+            border_value (int | tuple[int]): Value used in case of a
+                constant border.
+            interpolation (str): Same as in :func:`mmcv.imshear`.
+
+        Returns:
+            BitmapMasks: The sheared masks.
+        """
+        if len(self.masks) == 0:
+            sheared_masks = np.empty((0, *out_shape), dtype=np.uint8)
+        else:
+            sheared_masks = mmcv.imshear(
+                self.masks.transpose((1, 2, 0)),
+                magnitude,
+                direction,
+                border_value=border_value,
+                interpolation=interpolation)
+            if sheared_masks.ndim == 2:
+                sheared_masks = sheared_masks[:, :, None]
+            sheared_masks = sheared_masks.transpose(
+                (2, 0, 1)).astype(self.masks.dtype)
+        return BitmapMasks(sheared_masks, *out_shape)
+
+    def rotate(self, out_shape, angle, center=None, scale=1.0, fill_val=0):
+        """Rotate the BitmapMasks.
+
+        Args:
+            out_shape (tuple[int]): Shape for output mask, format (h, w).
+            angle (int | float): Rotation angle in degrees. Positive values
+                mean counter-clockwise rotation.
+            center (tuple[float], optional): Center point (w, h) of the
+                rotation in source image. If not specified, the center of
+                the image will be used.
+            scale (int | float): Isotropic scale factor.
+            fill_val (int | float): Border value. Default 0 for masks.
+
+        Returns:
+            BitmapMasks: Rotated BitmapMasks.
+        """
+        if len(self.masks) == 0:
+            rotated_masks = np.empty((0, *out_shape), dtype=self.masks.dtype)
+        else:
+            rotated_masks = mmcv.imrotate(
+                self.masks.transpose((1, 2, 0)),
+                angle,
+                center=center,
+                scale=scale,
+                border_value=fill_val)
+            if rotated_masks.ndim == 2:
+                # case when only one mask, (h, w)
+                rotated_masks = rotated_masks[:, :, None]  # (h, w, 1)
+            rotated_masks = rotated_masks.transpose(
+                (2, 0, 1)).astype(self.masks.dtype)
+        return BitmapMasks(rotated_masks, *out_shape)
+
+    @property
+    def areas(self):
+        """See :py:attr:`BaseInstanceMasks.areas`."""
+        return self.masks.sum((1, 2))
+
+    def to_ndarray(self):
+        """See :func:`BaseInstanceMasks.to_ndarray`."""
+        return self.masks
+
+    def to_tensor(self, dtype, device):
+        """See :func:`BaseInstanceMasks.to_tensor`."""
+        return torch.tensor(self.masks, dtype=dtype, device=device)
+
+    @classmethod
+    def random(cls,
+               num_masks=3,
+               height=32,
+               width=32,
+               dtype=np.uint8,
+               rng=None):
+        """Generate random bitmap masks for demo / testing purposes.
+
+        Example:
+            >>> from mmdet.core.mask.structures import BitmapMasks
+            >>> self = BitmapMasks.random()
+            >>> print('self = {}'.format(self))
+            self = BitmapMasks(num_masks=3, height=32, width=32)
+        """
+        from mmdet.utils.util_random import ensure_rng
+        rng = ensure_rng(rng)
+        masks = (rng.rand(num_masks, height, width) > 0.1).astype(dtype)
+        self = cls(masks, height=height, width=width)
+        return self
+
+
+class PolygonMasks(BaseInstanceMasks):
+    """This class represents masks in the form of polygons.
+
+    Polygons is a list of three levels. The first level of the list
+    corresponds to objects, the second level to the polys that compose the
+    object, the third level to the poly coordinates
+
+    Args:
+        masks (list[list[ndarray]]): The first level of the list
+            corresponds to objects, the second level to the polys that
+            compose the object, the third level to the poly coordinates
+        height (int): height of masks
+        width (int): width of masks
+
+    Example:
+        >>> from mmdet.core.mask.structures import *  # NOQA
+        >>> masks = [
+        >>>     [ np.array([0, 0, 10, 0, 10, 10., 0, 10, 0, 0]) ]
+        >>> ]
+        >>> height, width = 16, 16
+        >>> self = PolygonMasks(masks, height, width)
+
+        >>> # demo translate
+        >>> new = self.translate((16, 16), 4., direction='horizontal')
+        >>> assert np.all(new.masks[0][0][1::2] == masks[0][0][1::2])
+        >>> assert np.all(new.masks[0][0][0::2] == masks[0][0][0::2] + 4)
+
+        >>> # demo crop_and_resize
+        >>> num_boxes = 3
+        >>> bboxes = np.array([[0, 0, 30, 10.0]] * num_boxes)
+        >>> out_shape = (16, 16)
+        >>> inds = torch.randint(0, len(self), size=(num_boxes,))
+        >>> device = 'cpu'
+        >>> interpolation = 'bilinear'
+        >>> new = self.crop_and_resize(
+        ...     bboxes, out_shape, inds, device, interpolation)
+        >>> assert len(new) == num_boxes
+        >>> assert new.height, new.width == out_shape
+    """
+
+    def __init__(self, masks, height, width):
+        assert isinstance(masks, list)
+        if len(masks) > 0:
+            assert isinstance(masks[0], list)
+            assert isinstance(masks[0][0], np.ndarray)
+
+        self.height = height
+        self.width = width
+        self.masks = masks
+
+    def __getitem__(self, index):
+        """Index the polygon masks.
+
+        Args:
+            index (ndarray | List): The indices.
+
+        Returns:
+            :obj:`PolygonMasks`: The indexed polygon masks.
+        """
+        if isinstance(index, np.ndarray):
+            index = index.tolist()
+        if isinstance(index, list):
+            masks = [self.masks[i] for i in index]
+        else:
+            try:
+                masks = self.masks[index]
+            except Exception:
+                raise ValueError(
+                    f'Unsupported input of type {type(index)} for indexing!')
+        if len(masks) and isinstance(masks[0], np.ndarray):
+            masks = [masks]  # ensure a list of three levels
+        return PolygonMasks(masks, self.height, self.width)
+
+    def __iter__(self):
+        return iter(self.masks)
+
+    def __repr__(self):
+        s = self.__class__.__name__ + '('
+        s += f'num_masks={len(self.masks)}, '
+        s += f'height={self.height}, '
+        s += f'width={self.width})'
+        return s
+
+    def __len__(self):
+        """Number of masks."""
+        return len(self.masks)
+
+    def rescale(self, scale, interpolation=None):
+        """see :func:`BaseInstanceMasks.rescale`"""
+        new_w, new_h = mmcv.rescale_size((self.width, self.height), scale)
+        if len(self.masks) == 0:
+            rescaled_masks = PolygonMasks([], new_h, new_w)
+        else:
+            rescaled_masks = self.resize((new_h, new_w))
+        return rescaled_masks
+
+    def resize(self, out_shape, interpolation=None):
+        """see :func:`BaseInstanceMasks.resize`"""
+        if len(self.masks) == 0:
+            resized_masks = PolygonMasks([], *out_shape)
+        else:
+            h_scale = out_shape[0] / self.height
+            w_scale = out_shape[1] / self.width
+            resized_masks = []
+            for poly_per_obj in self.masks:
+                resized_poly = []
+                for p in poly_per_obj:
+                    p = p.copy()
+                    p[0::2] *= w_scale
+                    p[1::2] *= h_scale
+                    resized_poly.append(p)
+                resized_masks.append(resized_poly)
+            resized_masks = PolygonMasks(resized_masks, *out_shape)
+        return resized_masks
+
+    def flip(self, flip_direction='horizontal'):
+        """see :func:`BaseInstanceMasks.flip`"""
+        assert flip_direction in ('horizontal', 'vertical', 'diagonal')
+        if len(self.masks) == 0:
+            flipped_masks = PolygonMasks([], self.height, self.width)
+        else:
+            flipped_masks = []
+            for poly_per_obj in self.masks:
+                flipped_poly_per_obj = []
+                for p in poly_per_obj:
+                    p = p.copy()
+                    if flip_direction == 'horizontal':
+                        p[0::2] = self.width - p[0::2]
+                    elif flip_direction == 'vertical':
+                        p[1::2] = self.height - p[1::2]
+                    else:
+                        p[0::2] = self.width - p[0::2]
+                        p[1::2] = self.height - p[1::2]
+                    flipped_poly_per_obj.append(p)
+                flipped_masks.append(flipped_poly_per_obj)
+            flipped_masks = PolygonMasks(flipped_masks, self.height,
+                                         self.width)
+        return flipped_masks
+
+    def crop(self, bbox):
+        """see :func:`BaseInstanceMasks.crop`"""
+        assert isinstance(bbox, np.ndarray)
+        assert bbox.ndim == 1
+
+        # clip the boundary
+        bbox = bbox.copy()
+        bbox[0::2] = np.clip(bbox[0::2], 0, self.width)
+        bbox[1::2] = np.clip(bbox[1::2], 0, self.height)
+        x1, y1, x2, y2 = bbox
+        w = np.maximum(x2 - x1, 1)
+        h = np.maximum(y2 - y1, 1)
+
+        if len(self.masks) == 0:
+            cropped_masks = PolygonMasks([], h, w)
+        else:
+            cropped_masks = []
+            for poly_per_obj in self.masks:
+                cropped_poly_per_obj = []
+                for p in poly_per_obj:
+                    # pycocotools will clip the boundary
+                    p = p.copy()
+                    p[0::2] -= bbox[0]
+                    p[1::2] -= bbox[1]
+                    cropped_poly_per_obj.append(p)
+                cropped_masks.append(cropped_poly_per_obj)
+            cropped_masks = PolygonMasks(cropped_masks, h, w)
+        return cropped_masks
+
+    def pad(self, out_shape, pad_val=0):
+        """padding has no effect on polygons`"""
+        return PolygonMasks(self.masks, *out_shape)
+
+    def expand(self, *args, **kwargs):
+        """TODO: Add expand for polygon"""
+        raise NotImplementedError
+
+    def crop_and_resize(self,
+                        bboxes,
+                        out_shape,
+                        inds,
+                        device='cpu',
+                        interpolation='bilinear'):
+        """see :func:`BaseInstanceMasks.crop_and_resize`"""
+        out_h, out_w = out_shape
+        if len(self.masks) == 0:
+            return PolygonMasks([], out_h, out_w)
+
+        resized_masks = []
+        for i in range(len(bboxes)):
+            mask = self.masks[inds[i]]
+            bbox = bboxes[i, :]
+            x1, y1, x2, y2 = bbox
+            w = np.maximum(x2 - x1, 1)
+            h = np.maximum(y2 - y1, 1)
+            h_scale = out_h / max(h, 0.1)  # avoid too large scale
+            w_scale = out_w / max(w, 0.1)
+
+            resized_mask = []
+            for p in mask:
+                p = p.copy()
+                # crop
+                # pycocotools will clip the boundary
+                p[0::2] -= bbox[0]
+                p[1::2] -= bbox[1]
+
+                # resize
+                p[0::2] *= w_scale
+                p[1::2] *= h_scale
+                resized_mask.append(p)
+            resized_masks.append(resized_mask)
+        return PolygonMasks(resized_masks, *out_shape)
+
+    def translate(self,
+                  out_shape,
+                  offset,
+                  direction='horizontal',
+                  fill_val=None,
+                  interpolation=None):
+        """Translate the PolygonMasks.
+
+        Example:
+            >>> self = PolygonMasks.random(dtype=np.int)
+            >>> out_shape = (self.height, self.width)
+            >>> new = self.translate(out_shape, 4., direction='horizontal')
+            >>> assert np.all(new.masks[0][0][1::2] == self.masks[0][0][1::2])
+            >>> assert np.all(new.masks[0][0][0::2] == self.masks[0][0][0::2] + 4)  # noqa: E501
+        """
+        assert fill_val is None or fill_val == 0, 'Here fill_val is not '\
+            f'used, and defaultly should be None or 0. got {fill_val}.'
+        if len(self.masks) == 0:
+            translated_masks = PolygonMasks([], *out_shape)
+        else:
+            translated_masks = []
+            for poly_per_obj in self.masks:
+                translated_poly_per_obj = []
+                for p in poly_per_obj:
+                    p = p.copy()
+                    if direction == 'horizontal':
+                        p[0::2] = np.clip(p[0::2] + offset, 0, out_shape[1])
+                    elif direction == 'vertical':
+                        p[1::2] = np.clip(p[1::2] + offset, 0, out_shape[0])
+                    translated_poly_per_obj.append(p)
+                translated_masks.append(translated_poly_per_obj)
+            translated_masks = PolygonMasks(translated_masks, *out_shape)
+        return translated_masks
+
+    def shear(self,
+              out_shape,
+              magnitude,
+              direction='horizontal',
+              border_value=0,
+              interpolation='bilinear'):
+        """See :func:`BaseInstanceMasks.shear`."""
+        if len(self.masks) == 0:
+            sheared_masks = PolygonMasks([], *out_shape)
+        else:
+            sheared_masks = []
+            if direction == 'horizontal':
+                shear_matrix = np.stack([[1, magnitude],
+                                         [0, 1]]).astype(np.float32)
+            elif direction == 'vertical':
+                shear_matrix = np.stack([[1, 0], [magnitude,
+                                                  1]]).astype(np.float32)
+            for poly_per_obj in self.masks:
+                sheared_poly = []
+                for p in poly_per_obj:
+                    p = np.stack([p[0::2], p[1::2]], axis=0)  # [2, n]
+                    new_coords = np.matmul(shear_matrix, p)  # [2, n]
+                    new_coords[0, :] = np.clip(new_coords[0, :], 0,
+                                               out_shape[1])
+                    new_coords[1, :] = np.clip(new_coords[1, :], 0,
+                                               out_shape[0])
+                    sheared_poly.append(
+                        new_coords.transpose((1, 0)).reshape(-1))
+                sheared_masks.append(sheared_poly)
+            sheared_masks = PolygonMasks(sheared_masks, *out_shape)
+        return sheared_masks
+
+    def rotate(self, out_shape, angle, center=None, scale=1.0, fill_val=0):
+        """See :func:`BaseInstanceMasks.rotate`."""
+        if len(self.masks) == 0:
+            rotated_masks = PolygonMasks([], *out_shape)
+        else:
+            rotated_masks = []
+            rotate_matrix = cv2.getRotationMatrix2D(center, -angle, scale)
+            for poly_per_obj in self.masks:
+                rotated_poly = []
+                for p in poly_per_obj:
+                    p = p.copy()
+                    coords = np.stack([p[0::2], p[1::2]], axis=1)  # [n, 2]
+                    # pad 1 to convert from format [x, y] to homogeneous
+                    # coordinates format [x, y, 1]
+                    coords = np.concatenate(
+                        (coords, np.ones((coords.shape[0], 1), coords.dtype)),
+                        axis=1)  # [n, 3]
+                    rotated_coords = np.matmul(
+                        rotate_matrix[None, :, :],
+                        coords[:, :, None])[..., 0]  # [n, 2, 1] -> [n, 2]
+                    rotated_coords[:, 0] = np.clip(rotated_coords[:, 0], 0,
+                                                   out_shape[1])
+                    rotated_coords[:, 1] = np.clip(rotated_coords[:, 1], 0,
+                                                   out_shape[0])
+                    rotated_poly.append(rotated_coords.reshape(-1))
+                rotated_masks.append(rotated_poly)
+            rotated_masks = PolygonMasks(rotated_masks, *out_shape)
+        return rotated_masks
+
+    def to_bitmap(self):
+        """convert polygon masks to bitmap masks."""
+        bitmap_masks = self.to_ndarray()
+        return BitmapMasks(bitmap_masks, self.height, self.width)
+
+    @property
+    def areas(self):
+        """Compute areas of masks.
+
+        This func is modified from `detectron2
+        <https://github.com/facebookresearch/detectron2/blob/ffff8acc35ea88ad1cb1806ab0f00b4c1c5dbfd9/detectron2/structures/masks.py#L387>`_.
+        The function only works with Polygons using the shoelace formula.
+
+        Return:
+            ndarray: areas of each instance
+        """  # noqa: W501
+        area = []
+        for polygons_per_obj in self.masks:
+            area_per_obj = 0
+            for p in polygons_per_obj:
+                area_per_obj += self._polygon_area(p[0::2], p[1::2])
+            area.append(area_per_obj)
+        return np.asarray(area)
+
+    def _polygon_area(self, x, y):
+        """Compute the area of a component of a polygon.
+
+        Using the shoelace formula:
+        https://stackoverflow.com/questions/24467972/calculate-area-of-polygon-given-x-y-coordinates
+
+        Args:
+            x (ndarray): x coordinates of the component
+            y (ndarray): y coordinates of the component
+
+        Return:
+            float: the are of the component
+        """  # noqa: 501
+        return 0.5 * np.abs(
+            np.dot(x, np.roll(y, 1)) - np.dot(y, np.roll(x, 1)))
+
+    def to_ndarray(self):
+        """Convert masks to the format of ndarray."""
+        if len(self.masks) == 0:
+            return np.empty((0, self.height, self.width), dtype=np.uint8)
+        bitmap_masks = []
+        for poly_per_obj in self.masks:
+            bitmap_masks.append(
+                polygon_to_bitmap(poly_per_obj, self.height, self.width))
+        return np.stack(bitmap_masks)
+
+    def to_tensor(self, dtype, device):
+        """See :func:`BaseInstanceMasks.to_tensor`."""
+        if len(self.masks) == 0:
+            return torch.empty((0, self.height, self.width),
+                               dtype=dtype,
+                               device=device)
+        ndarray_masks = self.to_ndarray()
+        return torch.tensor(ndarray_masks, dtype=dtype, device=device)
+
+    @classmethod
+    def random(cls,
+               num_masks=3,
+               height=32,
+               width=32,
+               n_verts=5,
+               dtype=np.float32,
+               rng=None):
+        """Generate random polygon masks for demo / testing purposes.
+
+        Adapted from [1]_
+
+        References:
+            .. [1] https://gitlab.kitware.com/computer-vision/kwimage/-/blob/928cae35ca8/kwimage/structs/polygon.py#L379  # noqa: E501
+
+        Example:
+            >>> from mmdet.core.mask.structures import PolygonMasks
+            >>> self = PolygonMasks.random()
+            >>> print('self = {}'.format(self))
+        """
+        from mmdet.utils.util_random import ensure_rng
+        rng = ensure_rng(rng)
+
+        def _gen_polygon(n, irregularity, spikeyness):
+            """Creates the polygon by sampling points on a circle around the
+            centre.  Random noise is added by varying the angular spacing
+            between sequential points, and by varying the radial distance of
+            each point from the centre.
+
+            Based on original code by Mike Ounsworth
+
+            Args:
+                n (int): number of vertices
+                irregularity (float): [0,1] indicating how much variance there
+                    is in the angular spacing of vertices. [0,1] will map to
+                    [0, 2pi/numberOfVerts]
+                spikeyness (float): [0,1] indicating how much variance there is
+                    in each vertex from the circle of radius aveRadius. [0,1]
+                    will map to [0, aveRadius]
+
+            Returns:
+                a list of vertices, in CCW order.
+            """
+            from scipy.stats import truncnorm
+            # Generate around the unit circle
+            cx, cy = (0.0, 0.0)
+            radius = 1
+
+            tau = np.pi * 2
+
+            irregularity = np.clip(irregularity, 0, 1) * 2 * np.pi / n
+            spikeyness = np.clip(spikeyness, 1e-9, 1)
+
+            # generate n angle steps
+            lower = (tau / n) - irregularity
+            upper = (tau / n) + irregularity
+            angle_steps = rng.uniform(lower, upper, n)
+
+            # normalize the steps so that point 0 and point n+1 are the same
+            k = angle_steps.sum() / (2 * np.pi)
+            angles = (angle_steps / k).cumsum() + rng.uniform(0, tau)
+
+            # Convert high and low values to be wrt the standard normal range
+            # https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.truncnorm.html
+            low = 0
+            high = 2 * radius
+            mean = radius
+            std = spikeyness
+            a = (low - mean) / std
+            b = (high - mean) / std
+            tnorm = truncnorm(a=a, b=b, loc=mean, scale=std)
+
+            # now generate the points
+            radii = tnorm.rvs(n, random_state=rng)
+            x_pts = cx + radii * np.cos(angles)
+            y_pts = cy + radii * np.sin(angles)
+
+            points = np.hstack([x_pts[:, None], y_pts[:, None]])
+
+            # Scale to 0-1 space
+            points = points - points.min(axis=0)
+            points = points / points.max(axis=0)
+
+            # Randomly place within 0-1 space
+            points = points * (rng.rand() * .8 + .2)
+            min_pt = points.min(axis=0)
+            max_pt = points.max(axis=0)
+
+            high = (1 - max_pt)
+            low = (0 - min_pt)
+            offset = (rng.rand(2) * (high - low)) + low
+            points = points + offset
+            return points
+
+        def _order_vertices(verts):
+            """
+            References:
+                https://stackoverflow.com/questions/1709283/how-can-i-sort-a-coordinate-list-for-a-rectangle-counterclockwise
+            """
+            mlat = verts.T[0].sum() / len(verts)
+            mlng = verts.T[1].sum() / len(verts)
+
+            tau = np.pi * 2
+            angle = (np.arctan2(mlat - verts.T[0], verts.T[1] - mlng) +
+                     tau) % tau
+            sortx = angle.argsort()
+            verts = verts.take(sortx, axis=0)
+            return verts
+
+        # Generate a random exterior for each requested mask
+        masks = []
+        for _ in range(num_masks):
+            exterior = _order_vertices(_gen_polygon(n_verts, 0.9, 0.9))
+            exterior = (exterior * [(width, height)]).astype(dtype)
+            masks.append([exterior.ravel()])
+
+        self = cls(masks, height, width)
+        return self
+
+
+def polygon_to_bitmap(polygons, height, width):
+    """Convert masks from the form of polygons to bitmaps.
+
+    Args:
+        polygons (list[ndarray]): masks in polygon representation
+        height (int): mask height
+        width (int): mask width
+
+    Return:
+        ndarray: the converted masks in bitmap representation
+    """
+    rles = maskUtils.frPyObjects(polygons, height, width)
+    rle = maskUtils.merge(rles)
+    bitmap_mask = maskUtils.decode(rle).astype(np.bool)
+    return bitmap_mask
diff --git a/python/autoCut/models/swin/mmdet/core/mask/utils.py b/python/autoCut/models/swin/mmdet/core/mask/utils.py
new file mode 100644
index 0000000..c882082
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/mask/utils.py
@@ -0,0 +1,63 @@
+import mmcv
+import numpy as np
+import pycocotools.mask as mask_util
+
+
+def split_combined_polys(polys, poly_lens, polys_per_mask):
+    """Split the combined 1-D polys into masks.
+
+    A mask is represented as a list of polys, and a poly is represented as
+    a 1-D array. In dataset, all masks are concatenated into a single 1-D
+    tensor. Here we need to split the tensor into original representations.
+
+    Args:
+        polys (list): a list (length = image num) of 1-D tensors
+        poly_lens (list): a list (length = image num) of poly length
+        polys_per_mask (list): a list (length = image num) of poly number
+            of each mask
+
+    Returns:
+        list: a list (length = image num) of list (length = mask num) of \
+            list (length = poly num) of numpy array.
+    """
+    mask_polys_list = []
+    for img_id in range(len(polys)):
+        polys_single = polys[img_id]
+        polys_lens_single = poly_lens[img_id].tolist()
+        polys_per_mask_single = polys_per_mask[img_id].tolist()
+
+        split_polys = mmcv.slice_list(polys_single, polys_lens_single)
+        mask_polys = mmcv.slice_list(split_polys, polys_per_mask_single)
+        mask_polys_list.append(mask_polys)
+    return mask_polys_list
+
+
+# TODO: move this function to more proper place
+def encode_mask_results(mask_results):
+    """Encode bitmap mask to RLE code.
+
+    Args:
+        mask_results (list | tuple[list]): bitmap mask results.
+            In mask scoring rcnn, mask_results is a tuple of (segm_results,
+            segm_cls_score).
+
+    Returns:
+        list | tuple: RLE encoded mask.
+    """
+    if isinstance(mask_results, tuple):  # mask scoring
+        cls_segms, cls_mask_scores = mask_results
+    else:
+        cls_segms = mask_results
+    num_classes = len(cls_segms)
+    encoded_mask_results = [[] for _ in range(num_classes)]
+    for i in range(len(cls_segms)):
+        for cls_segm in cls_segms[i]:
+            encoded_mask_results[i].append(
+                mask_util.encode(
+                    np.array(
+                        cls_segm[:, :, np.newaxis], order='F',
+                        dtype='uint8'))[0])  # encoded with RLE
+    if isinstance(mask_results, tuple):
+        return encoded_mask_results, cls_mask_scores
+    else:
+        return encoded_mask_results
diff --git a/python/autoCut/models/swin/mmdet/core/post_processing/__init__.py b/python/autoCut/models/swin/mmdet/core/post_processing/__init__.py
new file mode 100644
index 0000000..aaf72a9
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/post_processing/__init__.py
@@ -0,0 +1,8 @@
+from .bbox_nms import fast_nms, multiclass_nms, multiclass_nms_rpd
+from .merge_augs import (merge_aug_bboxes, merge_aug_masks,
+                         merge_aug_proposals, merge_aug_scores)
+
+__all__ = [
+    'multiclass_nms', 'merge_aug_proposals', 'merge_aug_bboxes',
+    'merge_aug_scores', 'merge_aug_masks', 'fast_nms', 'multiclass_nms_rpd'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/post_processing/bbox_nms.py b/python/autoCut/models/swin/mmdet/core/post_processing/bbox_nms.py
new file mode 100644
index 0000000..1417254
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/post_processing/bbox_nms.py
@@ -0,0 +1,259 @@
+import torch
+from mmcv.ops.nms import batched_nms
+
+from mmdet.core.bbox.iou_calculators import bbox_overlaps
+
+
+def multiclass_nms(multi_bboxes,
+                   multi_scores,
+                   score_thr,
+                   nms_cfg,
+                   max_num=-1,
+                   score_factors=None,
+                   return_inds=False):
+    """NMS for multi-class bboxes.
+
+    Args:
+        multi_bboxes (Tensor): shape (n, #class*4) or (n, 4)
+        multi_scores (Tensor): shape (n, #class), where the last column
+            contains scores of the background class, but this will be ignored.
+        score_thr (float): bbox threshold, bboxes with scores lower than it
+            will not be considered.
+        nms_thr (float): NMS IoU threshold
+        max_num (int, optional): if there are more than max_num bboxes after
+            NMS, only top max_num will be kept. Default to -1.
+        score_factors (Tensor, optional): The factors multiplied to scores
+            before applying NMS. Default to None.
+        return_inds (bool, optional): Whether return the indices of kept
+            bboxes. Default to False.
+
+    Returns:
+        tuple: (bboxes, labels, indices (optional)), tensors of shape (k, 5),
+            (k), and (k). Labels are 0-based.
+    """
+    num_classes = multi_scores.size(1) - 1
+    # exclude background category
+    if multi_bboxes.shape[1] > 4:
+        bboxes = multi_bboxes.view(multi_scores.size(0), -1, 4)
+    else:
+        bboxes = multi_bboxes[:, None].expand(
+            multi_scores.size(0), num_classes, 4)
+
+    scores = multi_scores[:, :-1]
+
+    labels = torch.arange(num_classes, dtype=torch.long)
+    labels = labels.view(1, -1).expand_as(scores)
+
+    bboxes = bboxes.reshape(-1, 4)
+    scores = scores.reshape(-1)
+    labels = labels.reshape(-1)
+
+    if not torch.onnx.is_in_onnx_export():
+        # NonZero not supported  in TensorRT
+        # remove low scoring boxes
+        valid_mask = scores > score_thr
+    # multiply score_factor after threshold to preserve more bboxes, improve
+    # mAP by 1% for YOLOv3
+    if score_factors is not None:
+        # expand the shape to match original shape of score
+        score_factors = score_factors.view(-1, 1).expand(
+            multi_scores.size(0), num_classes)
+        score_factors = score_factors.reshape(-1)
+        scores = scores * score_factors
+
+    if not torch.onnx.is_in_onnx_export():
+        # NonZero not supported  in TensorRT
+        inds = valid_mask.nonzero(as_tuple=False).squeeze(1)
+        bboxes, scores, labels = bboxes[inds], scores[inds], labels[inds]
+    else:
+        # TensorRT NMS plugin has invalid output filled with -1
+        # add dummy data to make detection output correct.
+        bboxes = torch.cat([bboxes, bboxes.new_zeros(1, 4)], dim=0)
+        scores = torch.cat([scores, scores.new_zeros(1)], dim=0)
+        labels = torch.cat([labels, labels.new_zeros(1)], dim=0)
+
+    if bboxes.numel() == 0:
+        if torch.onnx.is_in_onnx_export():
+            raise RuntimeError('[ONNX Error] Can not record NMS '
+                               'as it has not been executed this time')
+        if return_inds:
+            return bboxes, labels, inds
+        else:
+            return bboxes, labels
+
+    dets, keep = batched_nms(bboxes, scores, labels, nms_cfg)
+
+    if max_num > 0:
+        dets = dets[:max_num]
+        keep = keep[:max_num]
+
+    if return_inds:
+        return dets, labels[keep], keep
+    else:
+        return dets, labels[keep]
+
+
+def multiclass_nms_rpd(multi_bboxes,
+                   multi_scores,
+                   score_thr,
+                   nms_cfg,
+                   max_num=-1,
+                   score_factors=None,
+                   inst_inds=None):
+    """NMS for multi-class bboxes.
+
+    Args:
+        multi_bboxes (Tensor): shape (n, #class*4) or (n, 4)
+        multi_scores (Tensor): shape (n, #class), where the last column
+            contains scores of the background class, but this will be ignored.
+        score_thr (float): bbox threshold, bboxes with scores lower than it
+            will not be considered.
+        nms_thr (float): NMS IoU threshold
+        max_num (int, optional): if there are more than max_num bboxes after
+            NMS, only top max_num will be kept. Default to -1.
+        score_factors (Tensor, optional): The factors multiplied to scores
+            before applying NMS. Default to None.
+
+    Returns:
+        tuple: (bboxes, labels, indices), tensors of shape (k, 5),
+            (k), and (k). Labels are 0-based.
+    """
+    num_classes = multi_scores.size(1) - 1
+    # exclude background category
+    if multi_bboxes.shape[1] > 4:
+        bboxes = multi_bboxes.view(multi_scores.size(0), -1, 4)
+        if inst_inds is not None:
+            inst_inds = inst_inds.view(inst_inds.size(0), -1)
+    else:
+        bboxes = multi_bboxes[:, None].expand(-1, num_classes, 4)
+        if inst_inds is not None:
+            inst_inds = inst_inds[:, None].expand(-1, num_classes)
+    scores = multi_scores[:, :-1]
+    if inst_inds is not None:
+        inst_inds = inst_inds[scores > score_thr]
+
+    labels = torch.arange(num_classes, dtype=torch.long)
+    labels = labels.view(1, -1).expand_as(scores)
+
+    bboxes = bboxes.reshape(-1, 4)
+    scores = scores.reshape(-1)
+    labels = labels.reshape(-1)
+
+    if not torch.onnx.is_in_onnx_export():
+        # NonZero not supported  in TensorRT
+        # remove low scoring boxes
+        valid_mask = scores > score_thr
+    # multiply score_factor after threshold to preserve more bboxes, improve
+    # mAP by 1% for YOLOv3
+    if score_factors is not None:
+        # expand the shape to match original shape of score
+        score_factors = score_factors.view(-1, 1).expand(
+            multi_scores.size(0), num_classes)
+        score_factors = score_factors.reshape(-1)
+        scores = scores * score_factors
+
+    if not torch.onnx.is_in_onnx_export():
+        # NonZero not supported  in TensorRT
+        inds = valid_mask.nonzero(as_tuple=False).squeeze(1)
+        bboxes, scores, labels = bboxes[inds], scores[inds], labels[inds]
+    else:
+        # TensorRT NMS plugin has invalid output filled with -1
+        # add dummy data to make detection output correct.
+        bboxes = torch.cat([bboxes, bboxes.new_zeros(1, 4)], dim=0)
+        scores = torch.cat([scores, scores.new_zeros(1)], dim=0)
+        labels = torch.cat([labels, labels.new_zeros(1)], dim=0)
+
+    if bboxes.numel() == 0:
+        if torch.onnx.is_in_onnx_export():
+            raise RuntimeError('[ONNX Error] Can not record NMS '
+                               'as it has not been executed this time')
+        if inst_inds is not None:
+            return bboxes, labels, inst_inds, None
+        else:
+            return bboxes, labels, None
+
+    dets, keep = batched_nms(bboxes, scores.to(bboxes.dtype), labels, nms_cfg)
+
+    if max_num > 0:
+        dets = dets[:max_num]
+        keep = keep[:max_num]
+
+    if inst_inds is not None:
+        return dets, labels[keep], inst_inds[keep], keep
+    else:
+        return dets, labels[keep], keep
+
+
+def fast_nms(multi_bboxes,
+             multi_scores,
+             multi_coeffs,
+             score_thr,
+             iou_thr,
+             top_k,
+             max_num=-1):
+    """Fast NMS in `YOLACT <https://arxiv.org/abs/1904.02689>`_.
+
+    Fast NMS allows already-removed detections to suppress other detections so
+    that every instance can be decided to be kept or discarded in parallel,
+    which is not possible in traditional NMS. This relaxation allows us to
+    implement Fast NMS entirely in standard GPU-accelerated matrix operations.
+
+    Args:
+        multi_bboxes (Tensor): shape (n, #class*4) or (n, 4)
+        multi_scores (Tensor): shape (n, #class+1), where the last column
+            contains scores of the background class, but this will be ignored.
+        multi_coeffs (Tensor): shape (n, #class*coeffs_dim).
+        score_thr (float): bbox threshold, bboxes with scores lower than it
+            will not be considered.
+        iou_thr (float): IoU threshold to be considered as conflicted.
+        top_k (int): if there are more than top_k bboxes before NMS,
+            only top top_k will be kept.
+        max_num (int): if there are more than max_num bboxes after NMS,
+            only top max_num will be kept. If -1, keep all the bboxes.
+            Default: -1.
+
+    Returns:
+        tuple: (bboxes, labels, coefficients), tensors of shape (k, 5), (k, 1),
+            and (k, coeffs_dim). Labels are 0-based.
+    """
+
+    scores = multi_scores[:, :-1].t()  # [#class, n]
+    scores, idx = scores.sort(1, descending=True)
+
+    idx = idx[:, :top_k].contiguous()
+    scores = scores[:, :top_k]  # [#class, topk]
+    num_classes, num_dets = idx.size()
+    boxes = multi_bboxes[idx.view(-1), :].view(num_classes, num_dets, 4)
+    coeffs = multi_coeffs[idx.view(-1), :].view(num_classes, num_dets, -1)
+
+    iou = bbox_overlaps(boxes, boxes)  # [#class, topk, topk]
+    iou.triu_(diagonal=1)
+    iou_max, _ = iou.max(dim=1)
+
+    # Now just filter out the ones higher than the threshold
+    keep = iou_max <= iou_thr
+
+    # Second thresholding introduces 0.2 mAP gain at negligible time cost
+    keep *= scores > score_thr
+
+    # Assign each kept detection to its corresponding class
+    classes = torch.arange(
+        num_classes, device=boxes.device)[:, None].expand_as(keep)
+    classes = classes[keep]
+
+    boxes = boxes[keep]
+    coeffs = coeffs[keep]
+    scores = scores[keep]
+
+    # Only keep the top max_num highest scores across all classes
+    scores, idx = scores.sort(0, descending=True)
+    if max_num > 0:
+        idx = idx[:max_num]
+        scores = scores[:max_num]
+
+    classes = classes[idx]
+    boxes = boxes[idx]
+    coeffs = coeffs[idx]
+
+    cls_dets = torch.cat([boxes, scores[:, None]], dim=1)
+    return cls_dets, classes, coeffs
diff --git a/python/autoCut/models/swin/mmdet/core/post_processing/merge_augs.py b/python/autoCut/models/swin/mmdet/core/post_processing/merge_augs.py
new file mode 100644
index 0000000..dbcf79d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/post_processing/merge_augs.py
@@ -0,0 +1,150 @@
+import copy
+import warnings
+
+import numpy as np
+import torch
+from mmcv import ConfigDict
+from mmcv.ops import nms
+
+from ..bbox import bbox_mapping_back
+
+
+def merge_aug_proposals(aug_proposals, img_metas, cfg):
+    """Merge augmented proposals (multiscale, flip, etc.)
+
+    Args:
+        aug_proposals (list[Tensor]): proposals from different testing
+            schemes, shape (n, 5). Note that they are not rescaled to the
+            original image size.
+
+        img_metas (list[dict]): list of image info dict where each dict has:
+            'img_shape', 'scale_factor', 'flip', and may also contain
+            'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+            For details on the values of these keys see
+            `mmdet/datasets/pipelines/formatting.py:Collect`.
+
+        cfg (dict): rpn test config.
+
+    Returns:
+        Tensor: shape (n, 4), proposals corresponding to original image scale.
+    """
+
+    cfg = copy.deepcopy(cfg)
+
+    # deprecate arguments warning
+    if 'nms' not in cfg or 'max_num' in cfg or 'nms_thr' in cfg:
+        warnings.warn(
+            'In rpn_proposal or test_cfg, '
+            'nms_thr has been moved to a dict named nms as '
+            'iou_threshold, max_num has been renamed as max_per_img, '
+            'name of original arguments and the way to specify '
+            'iou_threshold of NMS will be deprecated.')
+    if 'nms' not in cfg:
+        cfg.nms = ConfigDict(dict(type='nms', iou_threshold=cfg.nms_thr))
+    if 'max_num' in cfg:
+        if 'max_per_img' in cfg:
+            assert cfg.max_num == cfg.max_per_img, f'You set max_num and ' \
+                f'max_per_img at the same time, but get {cfg.max_num} ' \
+                f'and {cfg.max_per_img} respectively' \
+                f'Please delete max_num which will be deprecated.'
+        else:
+            cfg.max_per_img = cfg.max_num
+    if 'nms_thr' in cfg:
+        assert cfg.nms.iou_threshold == cfg.nms_thr, f'You set ' \
+            f'iou_threshold in nms and ' \
+            f'nms_thr at the same time, but get ' \
+            f'{cfg.nms.iou_threshold} and {cfg.nms_thr}' \
+            f' respectively. Please delete the nms_thr ' \
+            f'which will be deprecated.'
+
+    recovered_proposals = []
+    for proposals, img_info in zip(aug_proposals, img_metas):
+        img_shape = img_info['img_shape']
+        scale_factor = img_info['scale_factor']
+        flip = img_info['flip']
+        flip_direction = img_info['flip_direction']
+        _proposals = proposals.clone()
+        _proposals[:, :4] = bbox_mapping_back(_proposals[:, :4], img_shape,
+                                              scale_factor, flip,
+                                              flip_direction)
+        recovered_proposals.append(_proposals)
+    aug_proposals = torch.cat(recovered_proposals, dim=0)
+    merged_proposals, _ = nms(aug_proposals[:, :4].contiguous(),
+                              aug_proposals[:, -1].contiguous(),
+                              cfg.nms.iou_threshold)
+    scores = merged_proposals[:, 4]
+    _, order = scores.sort(0, descending=True)
+    num = min(cfg.max_per_img, merged_proposals.shape[0])
+    order = order[:num]
+    merged_proposals = merged_proposals[order, :]
+    return merged_proposals
+
+
+def merge_aug_bboxes(aug_bboxes, aug_scores, img_metas, rcnn_test_cfg):
+    """Merge augmented detection bboxes and scores.
+
+    Args:
+        aug_bboxes (list[Tensor]): shape (n, 4*#class)
+        aug_scores (list[Tensor] or None): shape (n, #class)
+        img_shapes (list[Tensor]): shape (3, ).
+        rcnn_test_cfg (dict): rcnn test config.
+
+    Returns:
+        tuple: (bboxes, scores)
+    """
+    recovered_bboxes = []
+    for bboxes, img_info in zip(aug_bboxes, img_metas):
+        img_shape = img_info[0]['img_shape']
+        scale_factor = img_info[0]['scale_factor']
+        flip = img_info[0]['flip']
+        flip_direction = img_info[0]['flip_direction']
+        bboxes = bbox_mapping_back(bboxes, img_shape, scale_factor, flip,
+                                   flip_direction)
+        recovered_bboxes.append(bboxes)
+    bboxes = torch.stack(recovered_bboxes).mean(dim=0)
+    if aug_scores is None:
+        return bboxes
+    else:
+        scores = torch.stack(aug_scores).mean(dim=0)
+        return bboxes, scores
+
+
+def merge_aug_scores(aug_scores):
+    """Merge augmented bbox scores."""
+    if isinstance(aug_scores[0], torch.Tensor):
+        return torch.mean(torch.stack(aug_scores), dim=0)
+    else:
+        return np.mean(aug_scores, axis=0)
+
+
+def merge_aug_masks(aug_masks, img_metas, rcnn_test_cfg, weights=None):
+    """Merge augmented mask prediction.
+
+    Args:
+        aug_masks (list[ndarray]): shape (n, #class, h, w)
+        img_shapes (list[ndarray]): shape (3, ).
+        rcnn_test_cfg (dict): rcnn test config.
+
+    Returns:
+        tuple: (bboxes, scores)
+    """
+    recovered_masks = []
+    for mask, img_info in zip(aug_masks, img_metas):
+        flip = img_info[0]['flip']
+        flip_direction = img_info[0]['flip_direction']
+        if flip:
+            if flip_direction == 'horizontal':
+                mask = mask[:, :, :, ::-1]
+            elif flip_direction == 'vertical':
+                mask = mask[:, :, ::-1, :]
+            else:
+                raise ValueError(
+                    f"Invalid flipping direction '{flip_direction}'")
+        recovered_masks.append(mask)
+
+    if weights is None:
+        merged_masks = np.mean(recovered_masks, axis=0)
+    else:
+        merged_masks = np.average(
+            np.array(recovered_masks), axis=0, weights=np.array(weights))
+    return merged_masks
diff --git a/python/autoCut/models/swin/mmdet/core/utils/__init__.py b/python/autoCut/models/swin/mmdet/core/utils/__init__.py
new file mode 100644
index 0000000..5c51dac
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/utils/__init__.py
@@ -0,0 +1,7 @@
+from .dist_utils import DistOptimizerHook, allreduce_grads, reduce_mean
+from .misc import mask2ndarray, multi_apply, unmap
+
+__all__ = [
+    'allreduce_grads', 'DistOptimizerHook', 'reduce_mean', 'multi_apply',
+    'unmap', 'mask2ndarray'
+]
diff --git a/python/autoCut/models/swin/mmdet/core/utils/dist_utils.py b/python/autoCut/models/swin/mmdet/core/utils/dist_utils.py
new file mode 100644
index 0000000..5fe7775
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/utils/dist_utils.py
@@ -0,0 +1,69 @@
+import warnings
+from collections import OrderedDict
+
+import torch.distributed as dist
+from mmcv.runner import OptimizerHook
+from torch._utils import (_flatten_dense_tensors, _take_tensors,
+                          _unflatten_dense_tensors)
+
+
+def _allreduce_coalesced(tensors, world_size, bucket_size_mb=-1):
+    if bucket_size_mb > 0:
+        bucket_size_bytes = bucket_size_mb * 1024 * 1024
+        buckets = _take_tensors(tensors, bucket_size_bytes)
+    else:
+        buckets = OrderedDict()
+        for tensor in tensors:
+            tp = tensor.type()
+            if tp not in buckets:
+                buckets[tp] = []
+            buckets[tp].append(tensor)
+        buckets = buckets.values()
+
+    for bucket in buckets:
+        flat_tensors = _flatten_dense_tensors(bucket)
+        dist.all_reduce(flat_tensors)
+        flat_tensors.div_(world_size)
+        for tensor, synced in zip(
+                bucket, _unflatten_dense_tensors(flat_tensors, bucket)):
+            tensor.copy_(synced)
+
+
+def allreduce_grads(params, coalesce=True, bucket_size_mb=-1):
+    """Allreduce gradients.
+
+    Args:
+        params (list[torch.Parameters]): List of parameters of a model
+        coalesce (bool, optional): Whether allreduce parameters as a whole.
+            Defaults to True.
+        bucket_size_mb (int, optional): Size of bucket, the unit is MB.
+            Defaults to -1.
+    """
+    grads = [
+        param.grad.data for param in params
+        if param.requires_grad and param.grad is not None
+    ]
+    world_size = dist.get_world_size()
+    if coalesce:
+        _allreduce_coalesced(grads, world_size, bucket_size_mb)
+    else:
+        for tensor in grads:
+            dist.all_reduce(tensor.div_(world_size))
+
+
+class DistOptimizerHook(OptimizerHook):
+    """Deprecated optimizer hook for distributed training."""
+
+    def __init__(self, *args, **kwargs):
+        warnings.warn('"DistOptimizerHook" is deprecated, please switch to'
+                      '"mmcv.runner.OptimizerHook".')
+        super().__init__(*args, **kwargs)
+
+
+def reduce_mean(tensor):
+    """"Obtain the mean of tensor on different GPUs."""
+    if not (dist.is_available() and dist.is_initialized()):
+        return tensor
+    tensor = tensor.clone()
+    dist.all_reduce(tensor.div_(dist.get_world_size()), op=dist.ReduceOp.SUM)
+    return tensor
diff --git a/python/autoCut/models/swin/mmdet/core/utils/misc.py b/python/autoCut/models/swin/mmdet/core/utils/misc.py
new file mode 100644
index 0000000..3e22c7b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/utils/misc.py
@@ -0,0 +1,61 @@
+from functools import partial
+
+import numpy as np
+import torch
+from six.moves import map, zip
+
+from ..mask.structures import BitmapMasks, PolygonMasks
+
+
+def multi_apply(func, *args, **kwargs):
+    """Apply function to a list of arguments.
+
+    Note:
+        This function applies the ``func`` to multiple inputs and
+        map the multiple outputs of the ``func`` into different
+        list. Each list contains the same type of outputs corresponding
+        to different inputs.
+
+    Args:
+        func (Function): A function that will be applied to a list of
+            arguments
+
+    Returns:
+        tuple(list): A tuple containing multiple list, each list contains \
+            a kind of returned results by the function
+    """
+    pfunc = partial(func, **kwargs) if kwargs else func
+    map_results = map(pfunc, *args)
+    return tuple(map(list, zip(*map_results)))
+
+
+def unmap(data, count, inds, fill=0):
+    """Unmap a subset of item (data) back to the original set of items (of size
+    count)"""
+    if data.dim() == 1:
+        ret = data.new_full((count, ), fill)
+        ret[inds.type(torch.bool)] = data
+    else:
+        new_size = (count, ) + data.size()[1:]
+        ret = data.new_full(new_size, fill)
+        ret[inds.type(torch.bool), :] = data
+    return ret
+
+
+def mask2ndarray(mask):
+    """Convert Mask to ndarray..
+
+    Args:
+        mask (:obj:`BitmapMasks` or :obj:`PolygonMasks` or
+        torch.Tensor or np.ndarray): The mask to be converted.
+
+    Returns:
+        np.ndarray: Ndarray mask of shape (n, h, w) that has been converted
+    """
+    if isinstance(mask, (BitmapMasks, PolygonMasks)):
+        mask = mask.to_ndarray()
+    elif isinstance(mask, torch.Tensor):
+        mask = mask.detach().cpu().numpy()
+    elif not isinstance(mask, np.ndarray):
+        raise TypeError(f'Unsupported {type(mask)} data type')
+    return mask
diff --git a/python/autoCut/models/swin/mmdet/core/visualization/__init__.py b/python/autoCut/models/swin/mmdet/core/visualization/__init__.py
new file mode 100644
index 0000000..4ff995c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/visualization/__init__.py
@@ -0,0 +1,4 @@
+from .image import (color_val_matplotlib, imshow_det_bboxes,
+                    imshow_gt_det_bboxes)
+
+__all__ = ['imshow_det_bboxes', 'imshow_gt_det_bboxes', 'color_val_matplotlib']
diff --git a/python/autoCut/models/swin/mmdet/core/visualization/image.py b/python/autoCut/models/swin/mmdet/core/visualization/image.py
new file mode 100644
index 0000000..5a14838
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/core/visualization/image.py
@@ -0,0 +1,303 @@
+import matplotlib.pyplot as plt
+import mmcv
+import numpy as np
+import pycocotools.mask as mask_util
+from matplotlib.collections import PatchCollection
+from matplotlib.patches import Polygon
+
+from ..utils import mask2ndarray
+
+EPS = 1e-2
+
+
+def color_val_matplotlib(color):
+    """Convert various input in BGR order to normalized RGB matplotlib color
+    tuples,
+
+    Args:
+        color (:obj:`Color`/str/tuple/int/ndarray): Color inputs
+
+    Returns:
+        tuple[float]: A tuple of 3 normalized floats indicating RGB channels.
+    """
+    color = mmcv.color_val(color)
+    color = [color / 255 for color in color[::-1]]
+    return tuple(color)
+
+
+def imshow_det_bboxes(img,
+                      bboxes,
+                      labels,
+                      segms=None,
+                      class_names=None,
+                      score_thr=0,
+                      bbox_color='green',
+                      text_color='green',
+                      mask_color=None,
+                      thickness=2,
+                      font_size=13,
+                      win_name='',
+                      show=True,
+                      wait_time=0,
+                      out_file=None):
+    """Draw bboxes and class labels (with scores) on an image.
+
+    Args:
+        img (str or ndarray): The image to be displayed.
+        bboxes (ndarray): Bounding boxes (with scores), shaped (n, 4) or
+            (n, 5).
+        labels (ndarray): Labels of bboxes.
+        segms (ndarray or None): Masks, shaped (n,h,w) or None
+        class_names (list[str]): Names of each classes.
+        score_thr (float): Minimum score of bboxes to be shown.  Default: 0
+        bbox_color (str or tuple(int) or :obj:`Color`):Color of bbox lines.
+           The tuple of color should be in BGR order. Default: 'green'
+        text_color (str or tuple(int) or :obj:`Color`):Color of texts.
+           The tuple of color should be in BGR order. Default: 'green'
+        mask_color (str or tuple(int) or :obj:`Color`, optional):
+           Color of masks. The tuple of color should be in BGR order.
+           Default: None
+        thickness (int): Thickness of lines. Default: 2
+        font_size (int): Font size of texts. Default: 13
+        show (bool): Whether to show the image. Default: True
+        win_name (str): The window name. Default: ''
+        wait_time (float): Value of waitKey param. Default: 0.
+        out_file (str, optional): The filename to write the image.
+            Default: None
+
+    Returns:
+        ndarray: The image with bboxes drawn on it.
+    """
+    assert bboxes.ndim == 2, \
+        f' bboxes ndim should be 2, but its ndim is {bboxes.ndim}.'
+    assert labels.ndim == 1, \
+        f' labels ndim should be 1, but its ndim is {labels.ndim}.'
+    assert bboxes.shape[0] == labels.shape[0], \
+        'bboxes.shape[0] and labels.shape[0] should have the same length.'
+    assert bboxes.shape[1] == 4 or bboxes.shape[1] == 5, \
+        f' bboxes.shape[1] should be 4 or 5, but its {bboxes.shape[1]}.'
+    img = mmcv.imread(img).astype(np.uint8)
+
+    if score_thr > 0:
+        assert bboxes.shape[1] == 5
+        scores = bboxes[:, -1]
+        inds = scores > score_thr
+        bboxes = bboxes[inds, :]
+        labels = labels[inds]
+        if segms is not None:
+            segms = segms[inds, ...]
+
+    mask_colors = []
+    if labels.shape[0] > 0:
+        if mask_color is None:
+            # random color
+            np.random.seed(42)
+            mask_colors = [
+                np.random.randint(0, 256, (1, 3), dtype=np.uint8)
+                for _ in range(max(labels) + 1)
+            ]
+        else:
+            # specify  color
+            mask_colors = [
+                np.array(mmcv.color_val(mask_color)[::-1], dtype=np.uint8)
+            ] * (
+                max(labels) + 1)
+
+    bbox_color = color_val_matplotlib(bbox_color)
+    text_color = color_val_matplotlib(text_color)
+
+    img = mmcv.bgr2rgb(img)
+    width, height = img.shape[1], img.shape[0]
+    img = np.ascontiguousarray(img)
+
+    fig = plt.figure(win_name, frameon=False)
+    plt.title(win_name)
+    canvas = fig.canvas
+    dpi = fig.get_dpi()
+    # add a small EPS to avoid precision lost due to matplotlib's truncation
+    # (https://github.com/matplotlib/matplotlib/issues/15363)
+    fig.set_size_inches((width + EPS) / dpi, (height + EPS) / dpi)
+
+    # remove white edges by set subplot margin
+    plt.subplots_adjust(left=0, right=1, bottom=0, top=1)
+    ax = plt.gca()
+    ax.axis('off')
+
+    polygons = []
+    color = []
+    for i, (bbox, label) in enumerate(zip(bboxes, labels)):
+        bbox_int = bbox.astype(np.int32)
+        poly = [[bbox_int[0], bbox_int[1]], [bbox_int[0], bbox_int[3]],
+                [bbox_int[2], bbox_int[3]], [bbox_int[2], bbox_int[1]]]
+        np_poly = np.array(poly).reshape((4, 2))
+        polygons.append(Polygon(np_poly))
+        color.append(bbox_color)
+        label_text = class_names[
+            label] if class_names is not None else f'class {label}'
+        if len(bbox) > 4:
+            label_text += f'|{bbox[-1]:.02f}'
+        ax.text(
+            bbox_int[0],
+            bbox_int[1],
+            f'{label_text}',
+            bbox={
+                'facecolor': 'black',
+                'alpha': 0.8,
+                'pad': 0.7,
+                'edgecolor': 'none'
+            },
+            color=text_color,
+            fontsize=font_size,
+            verticalalignment='top',
+            horizontalalignment='left')
+        if segms is not None:
+            color_mask = mask_colors[labels[i]]
+            mask = segms[i].astype(bool)
+            img[mask] = img[mask] * 0.5 + color_mask * 0.5
+
+    plt.imshow(img)
+
+    p = PatchCollection(
+        polygons, facecolor='none', edgecolors=color, linewidths=thickness)
+    ax.add_collection(p)
+
+    stream, _ = canvas.print_to_buffer()
+    buffer = np.frombuffer(stream, dtype='uint8')
+    img_rgba = buffer.reshape(height, width, 4)
+    rgb, alpha = np.split(img_rgba, [3], axis=2)
+    img = rgb.astype('uint8')
+    img = mmcv.rgb2bgr(img)
+
+    if show:
+        # We do not use cv2 for display because in some cases, opencv will
+        # conflict with Qt, it will output a warning: Current thread
+        # is not the object's thread. You can refer to
+        # https://github.com/opencv/opencv-python/issues/46 for details
+        if wait_time == 0:
+            plt.show()
+        else:
+            plt.show(block=False)
+            plt.pause(wait_time)
+    if out_file is not None:
+        mmcv.imwrite(img, out_file)
+
+    plt.close()
+
+    return img
+
+
+def imshow_gt_det_bboxes(img,
+                         annotation,
+                         result,
+                         class_names=None,
+                         score_thr=0,
+                         gt_bbox_color=(255, 102, 61),
+                         gt_text_color=(255, 102, 61),
+                         gt_mask_color=(255, 102, 61),
+                         det_bbox_color=(72, 101, 241),
+                         det_text_color=(72, 101, 241),
+                         det_mask_color=(72, 101, 241),
+                         thickness=2,
+                         font_size=13,
+                         win_name='',
+                         show=True,
+                         wait_time=0,
+                         out_file=None):
+    """General visualization GT and result function.
+
+    Args:
+      img (str or ndarray): The image to be displayed.)
+      annotation (dict): Ground truth annotations where contain keys of
+          'gt_bboxes' and 'gt_labels' or 'gt_masks'
+      result (tuple[list] or list): The detection result, can be either
+          (bbox, segm) or just bbox.
+      class_names (list[str]): Names of each classes.
+      score_thr (float): Minimum score of bboxes to be shown.  Default: 0
+      gt_bbox_color (str or tuple(int) or :obj:`Color`):Color of bbox lines.
+           The tuple of color should be in BGR order. Default: (255, 102, 61)
+      gt_text_color (str or tuple(int) or :obj:`Color`):Color of texts.
+           The tuple of color should be in BGR order. Default: (255, 102, 61)
+      gt_mask_color (str or tuple(int) or :obj:`Color`, optional):
+           Color of masks. The tuple of color should be in BGR order.
+           Default: (255, 102, 61)
+      det_bbox_color (str or tuple(int) or :obj:`Color`):Color of bbox lines.
+           The tuple of color should be in BGR order. Default: (72, 101, 241)
+      det_text_color (str or tuple(int) or :obj:`Color`):Color of texts.
+           The tuple of color should be in BGR order. Default: (72, 101, 241)
+      det_mask_color (str or tuple(int) or :obj:`Color`, optional):
+           Color of masks. The tuple of color should be in BGR order.
+           Default: (72, 101, 241)
+      thickness (int): Thickness of lines. Default: 2
+      font_size (int): Font size of texts. Default: 13
+      win_name (str): The window name. Default: ''
+      show (bool): Whether to show the image. Default: True
+      wait_time (float): Value of waitKey param. Default: 0.
+      out_file (str, optional): The filename to write the image.
+         Default: None
+
+    Returns:
+        ndarray: The image with bboxes or masks drawn on it.
+    """
+    assert 'gt_bboxes' in annotation
+    assert 'gt_labels' in annotation
+    assert isinstance(
+        result,
+        (tuple, list)), f'Expected tuple or list, but get {type(result)}'
+
+    gt_masks = annotation.get('gt_masks', None)
+    if gt_masks is not None:
+        gt_masks = mask2ndarray(gt_masks)
+
+    img = mmcv.imread(img)
+
+    img = imshow_det_bboxes(
+        img,
+        annotation['gt_bboxes'],
+        annotation['gt_labels'],
+        gt_masks,
+        class_names=class_names,
+        bbox_color=gt_bbox_color,
+        text_color=gt_text_color,
+        mask_color=gt_mask_color,
+        thickness=thickness,
+        font_size=font_size,
+        win_name=win_name,
+        show=False)
+
+    if isinstance(result, tuple):
+        bbox_result, segm_result = result
+        if isinstance(segm_result, tuple):
+            segm_result = segm_result[0]  # ms rcnn
+    else:
+        bbox_result, segm_result = result, None
+
+    bboxes = np.vstack(bbox_result)
+    labels = [
+        np.full(bbox.shape[0], i, dtype=np.int32)
+        for i, bbox in enumerate(bbox_result)
+    ]
+    labels = np.concatenate(labels)
+
+    segms = None
+    if segm_result is not None and len(labels) > 0:  # non empty
+        segms = mmcv.concat_list(segm_result)
+        segms = mask_util.decode(segms)
+        segms = segms.transpose(2, 0, 1)
+
+    img = imshow_det_bboxes(
+        img,
+        bboxes,
+        labels,
+        segms=segms,
+        class_names=class_names,
+        score_thr=score_thr,
+        bbox_color=det_bbox_color,
+        text_color=det_text_color,
+        mask_color=det_mask_color,
+        thickness=thickness,
+        font_size=font_size,
+        win_name=win_name,
+        show=show,
+        wait_time=wait_time,
+        out_file=out_file)
+    return img
diff --git a/python/autoCut/models/swin/mmdet/datasets/__init__.py b/python/autoCut/models/swin/mmdet/datasets/__init__.py
new file mode 100644
index 0000000..9b18b30
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/__init__.py
@@ -0,0 +1,24 @@
+from .builder import DATASETS, PIPELINES, build_dataloader, build_dataset
+from .cityscapes import CityscapesDataset
+from .coco import CocoDataset
+from .custom import CustomDataset
+from .dataset_wrappers import (ClassBalancedDataset, ConcatDataset,
+                               RepeatDataset)
+from .deepfashion import DeepFashionDataset
+from .lvis import LVISDataset, LVISV1Dataset, LVISV05Dataset
+from .samplers import DistributedGroupSampler, DistributedSampler, GroupSampler
+from .utils import (NumClassCheckHook, get_loading_pipeline,
+                    replace_ImageToTensor)
+from .voc import VOCDataset
+from .wider_face import WIDERFaceDataset
+from .xml_style import XMLDataset
+
+__all__ = [
+    'CustomDataset', 'XMLDataset', 'CocoDataset', 'DeepFashionDataset',
+    'VOCDataset', 'CityscapesDataset', 'LVISDataset', 'LVISV05Dataset',
+    'LVISV1Dataset', 'GroupSampler', 'DistributedGroupSampler',
+    'DistributedSampler', 'build_dataloader', 'ConcatDataset', 'RepeatDataset',
+    'ClassBalancedDataset', 'WIDERFaceDataset', 'DATASETS', 'PIPELINES',
+    'build_dataset', 'replace_ImageToTensor', 'get_loading_pipeline',
+    'NumClassCheckHook'
+]
diff --git a/python/autoCut/models/swin/mmdet/datasets/builder.py b/python/autoCut/models/swin/mmdet/datasets/builder.py
new file mode 100644
index 0000000..c9466a5
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/builder.py
@@ -0,0 +1,143 @@
+import copy
+import platform
+import random
+from functools import partial
+
+import numpy as np
+from mmcv.parallel import collate
+from mmcv.runner import get_dist_info
+from mmcv.utils import Registry, build_from_cfg
+from torch.utils.data import DataLoader
+
+from .samplers import DistributedGroupSampler, DistributedSampler, GroupSampler
+
+if platform.system() != 'Windows':
+    # https://github.com/pytorch/pytorch/issues/973
+    import resource
+    rlimit = resource.getrlimit(resource.RLIMIT_NOFILE)
+    hard_limit = rlimit[1]
+    soft_limit = min(4096, hard_limit)
+    resource.setrlimit(resource.RLIMIT_NOFILE, (soft_limit, hard_limit))
+
+DATASETS = Registry('dataset')
+PIPELINES = Registry('pipeline')
+
+
+def _concat_dataset(cfg, default_args=None):
+    from .dataset_wrappers import ConcatDataset
+    ann_files = cfg['ann_file']
+    img_prefixes = cfg.get('img_prefix', None)
+    seg_prefixes = cfg.get('seg_prefix', None)
+    proposal_files = cfg.get('proposal_file', None)
+    separate_eval = cfg.get('separate_eval', True)
+
+    datasets = []
+    num_dset = len(ann_files)
+    for i in range(num_dset):
+        data_cfg = copy.deepcopy(cfg)
+        # pop 'separate_eval' since it is not a valid key for common datasets.
+        if 'separate_eval' in data_cfg:
+            data_cfg.pop('separate_eval')
+        data_cfg['ann_file'] = ann_files[i]
+        if isinstance(img_prefixes, (list, tuple)):
+            data_cfg['img_prefix'] = img_prefixes[i]
+        if isinstance(seg_prefixes, (list, tuple)):
+            data_cfg['seg_prefix'] = seg_prefixes[i]
+        if isinstance(proposal_files, (list, tuple)):
+            data_cfg['proposal_file'] = proposal_files[i]
+        datasets.append(build_dataset(data_cfg, default_args))
+
+    return ConcatDataset(datasets, separate_eval)
+
+
+def build_dataset(cfg, default_args=None):
+    from .dataset_wrappers import (ConcatDataset, RepeatDataset,
+                                   ClassBalancedDataset)
+    if isinstance(cfg, (list, tuple)):
+        dataset = ConcatDataset([build_dataset(c, default_args) for c in cfg])
+    elif cfg['type'] == 'ConcatDataset':
+        dataset = ConcatDataset(
+            [build_dataset(c, default_args) for c in cfg['datasets']],
+            cfg.get('separate_eval', True))
+    elif cfg['type'] == 'RepeatDataset':
+        dataset = RepeatDataset(
+            build_dataset(cfg['dataset'], default_args), cfg['times'])
+    elif cfg['type'] == 'ClassBalancedDataset':
+        dataset = ClassBalancedDataset(
+            build_dataset(cfg['dataset'], default_args), cfg['oversample_thr'])
+    elif isinstance(cfg.get('ann_file'), (list, tuple)):
+        dataset = _concat_dataset(cfg, default_args)
+    else:
+        dataset = build_from_cfg(cfg, DATASETS, default_args)
+
+    return dataset
+
+
+def build_dataloader(dataset,
+                     samples_per_gpu,
+                     workers_per_gpu,
+                     num_gpus=1,
+                     dist=True,
+                     shuffle=True,
+                     seed=None,
+                     **kwargs):
+    """Build PyTorch DataLoader.
+
+    In distributed training, each GPU/process has a dataloader.
+    In non-distributed training, there is only one dataloader for all GPUs.
+
+    Args:
+        dataset (Dataset): A PyTorch dataset.
+        samples_per_gpu (int): Number of training samples on each GPU, i.e.,
+            batch size of each GPU.
+        workers_per_gpu (int): How many subprocesses to use for data loading
+            for each GPU.
+        num_gpus (int): Number of GPUs. Only used in non-distributed training.
+        dist (bool): Distributed training/test or not. Default: True.
+        shuffle (bool): Whether to shuffle the data at every epoch.
+            Default: True.
+        kwargs: any keyword argument to be used to initialize DataLoader
+
+    Returns:
+        DataLoader: A PyTorch dataloader.
+    """
+    rank, world_size = get_dist_info()
+    if dist:
+        # DistributedGroupSampler will definitely shuffle the data to satisfy
+        # that images on each GPU are in the same group
+        if shuffle:
+            sampler = DistributedGroupSampler(
+                dataset, samples_per_gpu, world_size, rank, seed=seed)
+        else:
+            sampler = DistributedSampler(
+                dataset, world_size, rank, shuffle=False, seed=seed)
+        batch_size = samples_per_gpu
+        num_workers = workers_per_gpu
+    else:
+        sampler = GroupSampler(dataset, samples_per_gpu) if shuffle else None
+        batch_size = num_gpus * samples_per_gpu
+        num_workers = num_gpus * workers_per_gpu
+
+    init_fn = partial(
+        worker_init_fn, num_workers=num_workers, rank=rank,
+        seed=seed) if seed is not None else None
+
+    data_loader = DataLoader(
+        dataset,
+        batch_size=batch_size,
+        sampler=sampler,
+        num_workers=num_workers,
+        collate_fn=partial(collate, samples_per_gpu=samples_per_gpu),
+        pin_memory=False,
+        worker_init_fn=init_fn,
+        **kwargs)
+
+    return data_loader
+
+
+def worker_init_fn(worker_id, num_workers, rank, seed):
+    # The seed of each worker equals to
+    # num_worker * rank + worker_id + user_seed
+    worker_seed = num_workers * rank + worker_id + seed
+    np.random.seed(worker_seed)
+    random.seed(worker_seed)
diff --git a/python/autoCut/models/swin/mmdet/datasets/cityscapes.py b/python/autoCut/models/swin/mmdet/datasets/cityscapes.py
new file mode 100644
index 0000000..71eead8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/cityscapes.py
@@ -0,0 +1,334 @@
+# Modified from https://github.com/facebookresearch/detectron2/blob/master/detectron2/data/datasets/cityscapes.py # noqa
+# and https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/evaluation/evalInstanceLevelSemanticLabeling.py # noqa
+
+import glob
+import os
+import os.path as osp
+import tempfile
+from collections import OrderedDict
+
+import mmcv
+import numpy as np
+import pycocotools.mask as maskUtils
+from mmcv.utils import print_log
+
+from .builder import DATASETS
+from .coco import CocoDataset
+
+
+@DATASETS.register_module()
+class CityscapesDataset(CocoDataset):
+
+    CLASSES = ('person', 'rider', 'car', 'truck', 'bus', 'train', 'motorcycle',
+               'bicycle')
+
+    def _filter_imgs(self, min_size=32):
+        """Filter images too small or without ground truths."""
+        valid_inds = []
+        # obtain images that contain annotation
+        ids_with_ann = set(_['image_id'] for _ in self.coco.anns.values())
+        # obtain images that contain annotations of the required categories
+        ids_in_cat = set()
+        for i, class_id in enumerate(self.cat_ids):
+            ids_in_cat |= set(self.coco.cat_img_map[class_id])
+        # merge the image id sets of the two conditions and use the merged set
+        # to filter out images if self.filter_empty_gt=True
+        ids_in_cat &= ids_with_ann
+
+        valid_img_ids = []
+        for i, img_info in enumerate(self.data_infos):
+            img_id = img_info['id']
+            ann_ids = self.coco.getAnnIds(imgIds=[img_id])
+            ann_info = self.coco.loadAnns(ann_ids)
+            all_iscrowd = all([_['iscrowd'] for _ in ann_info])
+            if self.filter_empty_gt and (self.img_ids[i] not in ids_in_cat
+                                         or all_iscrowd):
+                continue
+            if min(img_info['width'], img_info['height']) >= min_size:
+                valid_inds.append(i)
+                valid_img_ids.append(img_id)
+        self.img_ids = valid_img_ids
+        return valid_inds
+
+    def _parse_ann_info(self, img_info, ann_info):
+        """Parse bbox and mask annotation.
+
+        Args:
+            img_info (dict): Image info of an image.
+            ann_info (list[dict]): Annotation info of an image.
+
+        Returns:
+            dict: A dict containing the following keys: bboxes, \
+                bboxes_ignore, labels, masks, seg_map. \
+                "masks" are already decoded into binary masks.
+        """
+        gt_bboxes = []
+        gt_labels = []
+        gt_bboxes_ignore = []
+        gt_masks_ann = []
+
+        for i, ann in enumerate(ann_info):
+            if ann.get('ignore', False):
+                continue
+            x1, y1, w, h = ann['bbox']
+            if ann['area'] <= 0 or w < 1 or h < 1:
+                continue
+            if ann['category_id'] not in self.cat_ids:
+                continue
+            bbox = [x1, y1, x1 + w, y1 + h]
+            if ann.get('iscrowd', False):
+                gt_bboxes_ignore.append(bbox)
+            else:
+                gt_bboxes.append(bbox)
+                gt_labels.append(self.cat2label[ann['category_id']])
+                gt_masks_ann.append(ann['segmentation'])
+
+        if gt_bboxes:
+            gt_bboxes = np.array(gt_bboxes, dtype=np.float32)
+            gt_labels = np.array(gt_labels, dtype=np.int64)
+        else:
+            gt_bboxes = np.zeros((0, 4), dtype=np.float32)
+            gt_labels = np.array([], dtype=np.int64)
+
+        if gt_bboxes_ignore:
+            gt_bboxes_ignore = np.array(gt_bboxes_ignore, dtype=np.float32)
+        else:
+            gt_bboxes_ignore = np.zeros((0, 4), dtype=np.float32)
+
+        ann = dict(
+            bboxes=gt_bboxes,
+            labels=gt_labels,
+            bboxes_ignore=gt_bboxes_ignore,
+            masks=gt_masks_ann,
+            seg_map=img_info['segm_file'])
+
+        return ann
+
+    def results2txt(self, results, outfile_prefix):
+        """Dump the detection results to a txt file.
+
+        Args:
+            results (list[list | tuple]): Testing results of the
+                dataset.
+            outfile_prefix (str): The filename prefix of the json files.
+                If the prefix is "somepath/xxx",
+                the txt files will be named "somepath/xxx.txt".
+
+        Returns:
+            list[str]: Result txt files which contains corresponding \
+                instance segmentation images.
+        """
+        try:
+            import cityscapesscripts.helpers.labels as CSLabels
+        except ImportError:
+            raise ImportError('Please run "pip install citscapesscripts" to '
+                              'install cityscapesscripts first.')
+        result_files = []
+        os.makedirs(outfile_prefix, exist_ok=True)
+        prog_bar = mmcv.ProgressBar(len(self))
+        for idx in range(len(self)):
+            result = results[idx]
+            filename = self.data_infos[idx]['filename']
+            basename = osp.splitext(osp.basename(filename))[0]
+            pred_txt = osp.join(outfile_prefix, basename + '_pred.txt')
+
+            bbox_result, segm_result = result
+            bboxes = np.vstack(bbox_result)
+            # segm results
+            if isinstance(segm_result, tuple):
+                # Some detectors use different scores for bbox and mask,
+                # like Mask Scoring R-CNN. Score of segm will be used instead
+                # of bbox score.
+                segms = mmcv.concat_list(segm_result[0])
+                mask_score = segm_result[1]
+            else:
+                # use bbox score for mask score
+                segms = mmcv.concat_list(segm_result)
+                mask_score = [bbox[-1] for bbox in bboxes]
+            labels = [
+                np.full(bbox.shape[0], i, dtype=np.int32)
+                for i, bbox in enumerate(bbox_result)
+            ]
+            labels = np.concatenate(labels)
+
+            assert len(bboxes) == len(segms) == len(labels)
+            num_instances = len(bboxes)
+            prog_bar.update()
+            with open(pred_txt, 'w') as fout:
+                for i in range(num_instances):
+                    pred_class = labels[i]
+                    classes = self.CLASSES[pred_class]
+                    class_id = CSLabels.name2label[classes].id
+                    score = mask_score[i]
+                    mask = maskUtils.decode(segms[i]).astype(np.uint8)
+                    png_filename = osp.join(outfile_prefix,
+                                            basename + f'_{i}_{classes}.png')
+                    mmcv.imwrite(mask, png_filename)
+                    fout.write(f'{osp.basename(png_filename)} {class_id} '
+                               f'{score}\n')
+            result_files.append(pred_txt)
+
+        return result_files
+
+    def format_results(self, results, txtfile_prefix=None):
+        """Format the results to txt (standard format for Cityscapes
+        evaluation).
+
+        Args:
+            results (list): Testing results of the dataset.
+            txtfile_prefix (str | None): The prefix of txt files. It includes
+                the file path and the prefix of filename, e.g., "a/b/prefix".
+                If not specified, a temp file will be created. Default: None.
+
+        Returns:
+            tuple: (result_files, tmp_dir), result_files is a dict containing \
+                the json filepaths, tmp_dir is the temporal directory created \
+                for saving txt/png files when txtfile_prefix is not specified.
+        """
+        assert isinstance(results, list), 'results must be a list'
+        assert len(results) == len(self), (
+            'The length of results is not equal to the dataset len: {} != {}'.
+            format(len(results), len(self)))
+
+        assert isinstance(results, list), 'results must be a list'
+        assert len(results) == len(self), (
+            'The length of results is not equal to the dataset len: {} != {}'.
+            format(len(results), len(self)))
+
+        if txtfile_prefix is None:
+            tmp_dir = tempfile.TemporaryDirectory()
+            txtfile_prefix = osp.join(tmp_dir.name, 'results')
+        else:
+            tmp_dir = None
+        result_files = self.results2txt(results, txtfile_prefix)
+
+        return result_files, tmp_dir
+
+    def evaluate(self,
+                 results,
+                 metric='bbox',
+                 logger=None,
+                 outfile_prefix=None,
+                 classwise=False,
+                 proposal_nums=(100, 300, 1000),
+                 iou_thrs=np.arange(0.5, 0.96, 0.05)):
+        """Evaluation in Cityscapes/COCO protocol.
+
+        Args:
+            results (list[list | tuple]): Testing results of the dataset.
+            metric (str | list[str]): Metrics to be evaluated. Options are
+                'bbox', 'segm', 'proposal', 'proposal_fast'.
+            logger (logging.Logger | str | None): Logger used for printing
+                related information during evaluation. Default: None.
+            outfile_prefix (str | None): The prefix of output file. It includes
+                the file path and the prefix of filename, e.g., "a/b/prefix".
+                If results are evaluated with COCO protocol, it would be the
+                prefix of output json file. For example, the metric is 'bbox'
+                and 'segm', then json files would be "a/b/prefix.bbox.json" and
+                "a/b/prefix.segm.json".
+                If results are evaluated with cityscapes protocol, it would be
+                the prefix of output txt/png files. The output files would be
+                png images under folder "a/b/prefix/xxx/" and the file name of
+                images would be written into a txt file
+                "a/b/prefix/xxx_pred.txt", where "xxx" is the video name of
+                cityscapes. If not specified, a temp file will be created.
+                Default: None.
+            classwise (bool): Whether to evaluating the AP for each class.
+            proposal_nums (Sequence[int]): Proposal number used for evaluating
+                recalls, such as recall@100, recall@1000.
+                Default: (100, 300, 1000).
+            iou_thrs (Sequence[float]): IoU threshold used for evaluating
+                recalls. If set to a list, the average recall of all IoUs will
+                also be computed. Default: 0.5.
+
+        Returns:
+            dict[str, float]: COCO style evaluation metric or cityscapes mAP \
+                and AP@50.
+        """
+        eval_results = dict()
+
+        metrics = metric.copy() if isinstance(metric, list) else [metric]
+
+        if 'cityscapes' in metrics:
+            eval_results.update(
+                self._evaluate_cityscapes(results, outfile_prefix, logger))
+            metrics.remove('cityscapes')
+
+        # left metrics are all coco metric
+        if len(metrics) > 0:
+            # create CocoDataset with CityscapesDataset annotation
+            self_coco = CocoDataset(self.ann_file, self.pipeline.transforms,
+                                    None, self.data_root, self.img_prefix,
+                                    self.seg_prefix, self.proposal_file,
+                                    self.test_mode, self.filter_empty_gt)
+            # TODO: remove this in the future
+            # reload annotations of correct class
+            self_coco.CLASSES = self.CLASSES
+            self_coco.data_infos = self_coco.load_annotations(self.ann_file)
+            eval_results.update(
+                self_coco.evaluate(results, metrics, logger, outfile_prefix,
+                                   classwise, proposal_nums, iou_thrs))
+
+        return eval_results
+
+    def _evaluate_cityscapes(self, results, txtfile_prefix, logger):
+        """Evaluation in Cityscapes protocol.
+
+        Args:
+            results (list): Testing results of the dataset.
+            txtfile_prefix (str | None): The prefix of output txt file
+            logger (logging.Logger | str | None): Logger used for printing
+                related information during evaluation. Default: None.
+
+        Returns:
+            dict[str: float]: Cityscapes evaluation results, contains 'mAP' \
+                and 'AP@50'.
+        """
+
+        try:
+            import cityscapesscripts.evaluation.evalInstanceLevelSemanticLabeling as CSEval  # noqa
+        except ImportError:
+            raise ImportError('Please run "pip install citscapesscripts" to '
+                              'install cityscapesscripts first.')
+        msg = 'Evaluating in Cityscapes style'
+        if logger is None:
+            msg = '\n' + msg
+        print_log(msg, logger=logger)
+
+        result_files, tmp_dir = self.format_results(results, txtfile_prefix)
+
+        if tmp_dir is None:
+            result_dir = osp.join(txtfile_prefix, 'results')
+        else:
+            result_dir = osp.join(tmp_dir.name, 'results')
+
+        eval_results = OrderedDict()
+        print_log(f'Evaluating results under {result_dir} ...', logger=logger)
+
+        # set global states in cityscapes evaluation API
+        CSEval.args.cityscapesPath = os.path.join(self.img_prefix, '../..')
+        CSEval.args.predictionPath = os.path.abspath(result_dir)
+        CSEval.args.predictionWalk = None
+        CSEval.args.JSONOutput = False
+        CSEval.args.colorized = False
+        CSEval.args.gtInstancesFile = os.path.join(result_dir,
+                                                   'gtInstances.json')
+        CSEval.args.groundTruthSearch = os.path.join(
+            self.img_prefix.replace('leftImg8bit', 'gtFine'),
+            '*/*_gtFine_instanceIds.png')
+
+        groundTruthImgList = glob.glob(CSEval.args.groundTruthSearch)
+        assert len(groundTruthImgList), 'Cannot find ground truth images' \
+            f' in {CSEval.args.groundTruthSearch}.'
+        predictionImgList = []
+        for gt in groundTruthImgList:
+            predictionImgList.append(CSEval.getPrediction(gt, CSEval.args))
+        CSEval_results = CSEval.evaluateImgLists(predictionImgList,
+                                                 groundTruthImgList,
+                                                 CSEval.args)['averages']
+
+        eval_results['mAP'] = CSEval_results['allAp']
+        eval_results['AP@50'] = CSEval_results['allAp50%']
+        if tmp_dir is not None:
+            tmp_dir.cleanup()
+        return eval_results
diff --git a/python/autoCut/models/swin/mmdet/datasets/coco.py b/python/autoCut/models/swin/mmdet/datasets/coco.py
new file mode 100644
index 0000000..3a8e1bc
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/coco.py
@@ -0,0 +1,546 @@
+import itertools
+import logging
+import os.path as osp
+import tempfile
+from collections import OrderedDict
+
+import mmcv
+import numpy as np
+import pycocotools
+from mmcv.utils import print_log
+from pycocotools.coco import COCO
+from pycocotools.cocoeval import COCOeval
+from terminaltables import AsciiTable
+
+from mmdet.core import eval_recalls
+from .builder import DATASETS
+from .custom import CustomDataset
+
+
+@DATASETS.register_module()
+class CocoDataset(CustomDataset):
+
+    CLASSES = ('person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus',
+               'train', 'truck', 'boat', 'traffic light', 'fire hydrant',
+               'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog',
+               'horse', 'sheep', 'cow', 'elephant', 'bear', 'zebra', 'giraffe',
+               'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee',
+               'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat',
+               'baseball glove', 'skateboard', 'surfboard', 'tennis racket',
+               'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl',
+               'banana', 'apple', 'sandwich', 'orange', 'broccoli', 'carrot',
+               'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch',
+               'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop',
+               'mouse', 'remote', 'keyboard', 'cell phone', 'microwave',
+               'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock',
+               'vase', 'scissors', 'teddy bear', 'hair drier', 'toothbrush')
+
+    def load_annotations(self, ann_file):
+        """Load annotation from COCO style annotation file.
+
+        Args:
+            ann_file (str): Path of annotation file.
+
+        Returns:
+            list[dict]: Annotation info from COCO api.
+        """
+        if not getattr(pycocotools, '__version__', '0') >= '12.0.2':
+            raise AssertionError(
+                'Incompatible version of pycocotools is installed. '
+                'Run pip uninstall pycocotools first. Then run pip '
+                'install mmpycocotools to install open-mmlab forked '
+                'pycocotools.')
+
+        self.coco = COCO(ann_file)
+        self.cat_ids = self.coco.get_cat_ids(cat_names=self.CLASSES)
+        self.cat2label = {cat_id: i for i, cat_id in enumerate(self.cat_ids)}
+        self.img_ids = self.coco.get_img_ids()
+        data_infos = []
+        total_ann_ids = []
+        for i in self.img_ids:
+            info = self.coco.load_imgs([i])[0]
+            info['filename'] = info['file_name']
+            data_infos.append(info)
+            ann_ids = self.coco.get_ann_ids(img_ids=[i])
+            total_ann_ids.extend(ann_ids)
+        assert len(set(total_ann_ids)) == len(
+            total_ann_ids), f"Annotation ids in '{ann_file}' are not unique!"
+        return data_infos
+
+    def get_ann_info(self, idx):
+        """Get COCO annotation by index.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            dict: Annotation info of specified index.
+        """
+
+        img_id = self.data_infos[idx]['id']
+        ann_ids = self.coco.get_ann_ids(img_ids=[img_id])
+        ann_info = self.coco.load_anns(ann_ids)
+        return self._parse_ann_info(self.data_infos[idx], ann_info)
+
+    def get_cat_ids(self, idx):
+        """Get COCO category ids by index.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            list[int]: All categories in the image of specified index.
+        """
+
+        img_id = self.data_infos[idx]['id']
+        ann_ids = self.coco.get_ann_ids(img_ids=[img_id])
+        ann_info = self.coco.load_anns(ann_ids)
+        return [ann['category_id'] for ann in ann_info]
+
+    def _filter_imgs(self, min_size=32):
+        """Filter images too small or without ground truths."""
+        valid_inds = []
+        # obtain images that contain annotation
+        ids_with_ann = set(_['image_id'] for _ in self.coco.anns.values())
+        # obtain images that contain annotations of the required categories
+        ids_in_cat = set()
+        for i, class_id in enumerate(self.cat_ids):
+            ids_in_cat |= set(self.coco.cat_img_map[class_id])
+        # merge the image id sets of the two conditions and use the merged set
+        # to filter out images if self.filter_empty_gt=True
+        ids_in_cat &= ids_with_ann
+
+        valid_img_ids = []
+        for i, img_info in enumerate(self.data_infos):
+            img_id = self.img_ids[i]
+            if self.filter_empty_gt and img_id not in ids_in_cat:
+                continue
+            if min(img_info['width'], img_info['height']) >= min_size:
+                valid_inds.append(i)
+                valid_img_ids.append(img_id)
+        self.img_ids = valid_img_ids
+        return valid_inds
+
+    def _parse_ann_info(self, img_info, ann_info):
+        """Parse bbox and mask annotation.
+
+        Args:
+            ann_info (list[dict]): Annotation info of an image.
+            with_mask (bool): Whether to parse mask annotations.
+
+        Returns:
+            dict: A dict containing the following keys: bboxes, bboxes_ignore,\
+                labels, masks, seg_map. "masks" are raw annotations and not \
+                decoded into binary masks.
+        """
+        gt_bboxes = []
+        gt_labels = []
+        gt_bboxes_ignore = []
+        gt_masks_ann = []
+        for i, ann in enumerate(ann_info):
+            if ann.get('ignore', False):
+                continue
+            x1, y1, w, h = ann['bbox']
+            inter_w = max(0, min(x1 + w, img_info['width']) - max(x1, 0))
+            inter_h = max(0, min(y1 + h, img_info['height']) - max(y1, 0))
+            if inter_w * inter_h == 0:
+                continue
+            if ann['area'] <= 0 or w < 1 or h < 1:
+                continue
+            if ann['category_id'] not in self.cat_ids:
+                continue
+            bbox = [x1, y1, x1 + w, y1 + h]
+            if ann.get('iscrowd', False):
+                gt_bboxes_ignore.append(bbox)
+            else:
+                gt_bboxes.append(bbox)
+                gt_labels.append(self.cat2label[ann['category_id']])
+                gt_masks_ann.append(ann.get('segmentation', None))
+
+        if gt_bboxes:
+            gt_bboxes = np.array(gt_bboxes, dtype=np.float32)
+            gt_labels = np.array(gt_labels, dtype=np.int64)
+        else:
+            gt_bboxes = np.zeros((0, 4), dtype=np.float32)
+            gt_labels = np.array([], dtype=np.int64)
+
+        if gt_bboxes_ignore:
+            gt_bboxes_ignore = np.array(gt_bboxes_ignore, dtype=np.float32)
+        else:
+            gt_bboxes_ignore = np.zeros((0, 4), dtype=np.float32)
+
+        seg_map = img_info['filename'].replace('jpg', 'png')
+
+        ann = dict(
+            bboxes=gt_bboxes,
+            labels=gt_labels,
+            bboxes_ignore=gt_bboxes_ignore,
+            masks=gt_masks_ann,
+            seg_map=seg_map)
+
+        return ann
+
+    def xyxy2xywh(self, bbox):
+        """Convert ``xyxy`` style bounding boxes to ``xywh`` style for COCO
+        evaluation.
+
+        Args:
+            bbox (numpy.ndarray): The bounding boxes, shape (4, ), in
+                ``xyxy`` order.
+
+        Returns:
+            list[float]: The converted bounding boxes, in ``xywh`` order.
+        """
+
+        _bbox = bbox.tolist()
+        return [
+            _bbox[0],
+            _bbox[1],
+            _bbox[2] - _bbox[0],
+            _bbox[3] - _bbox[1],
+        ]
+
+    def _proposal2json(self, results):
+        """Convert proposal results to COCO json style."""
+        json_results = []
+        for idx in range(len(self)):
+            img_id = self.img_ids[idx]
+            bboxes = results[idx]
+            for i in range(bboxes.shape[0]):
+                data = dict()
+                data['image_id'] = img_id
+                data['bbox'] = self.xyxy2xywh(bboxes[i])
+                data['score'] = float(bboxes[i][4])
+                data['category_id'] = 1
+                json_results.append(data)
+        return json_results
+
+    def _det2json(self, results):
+        """Convert detection results to COCO json style."""
+        json_results = []
+        for idx in range(len(self)):
+            img_id = self.img_ids[idx]
+            result = results[idx]
+            for label in range(len(result)):
+                bboxes = result[label]
+                for i in range(bboxes.shape[0]):
+                    data = dict()
+                    data['image_id'] = img_id
+                    data['bbox'] = self.xyxy2xywh(bboxes[i])
+                    data['score'] = float(bboxes[i][4])
+                    data['category_id'] = self.cat_ids[label]
+                    json_results.append(data)
+        return json_results
+
+    def _segm2json(self, results):
+        """Convert instance segmentation results to COCO json style."""
+        bbox_json_results = []
+        segm_json_results = []
+        for idx in range(len(self)):
+            img_id = self.img_ids[idx]
+            det, seg = results[idx]
+            for label in range(len(det)):
+                # bbox results
+                bboxes = det[label]
+                for i in range(bboxes.shape[0]):
+                    data = dict()
+                    data['image_id'] = img_id
+                    data['bbox'] = self.xyxy2xywh(bboxes[i])
+                    data['score'] = float(bboxes[i][4])
+                    data['category_id'] = self.cat_ids[label]
+                    bbox_json_results.append(data)
+
+                # segm results
+                # some detectors use different scores for bbox and mask
+                if isinstance(seg, tuple):
+                    segms = seg[0][label]
+                    mask_score = seg[1][label]
+                else:
+                    segms = seg[label]
+                    mask_score = [bbox[4] for bbox in bboxes]
+                for i in range(bboxes.shape[0]):
+                    data = dict()
+                    data['image_id'] = img_id
+                    data['bbox'] = self.xyxy2xywh(bboxes[i])
+                    data['score'] = float(mask_score[i])
+                    data['category_id'] = self.cat_ids[label]
+                    if isinstance(segms[i]['counts'], bytes):
+                        segms[i]['counts'] = segms[i]['counts'].decode()
+                    data['segmentation'] = segms[i]
+                    segm_json_results.append(data)
+        return bbox_json_results, segm_json_results
+
+    def results2json(self, results, outfile_prefix):
+        """Dump the detection results to a COCO style json file.
+
+        There are 3 types of results: proposals, bbox predictions, mask
+        predictions, and they have different data types. This method will
+        automatically recognize the type, and dump them to json files.
+
+        Args:
+            results (list[list | tuple | ndarray]): Testing results of the
+                dataset.
+            outfile_prefix (str): The filename prefix of the json files. If the
+                prefix is "somepath/xxx", the json files will be named
+                "somepath/xxx.bbox.json", "somepath/xxx.segm.json",
+                "somepath/xxx.proposal.json".
+
+        Returns:
+            dict[str: str]: Possible keys are "bbox", "segm", "proposal", and \
+                values are corresponding filenames.
+        """
+        result_files = dict()
+        if isinstance(results[0], list):
+            json_results = self._det2json(results)
+            result_files['bbox'] = f'{outfile_prefix}.bbox.json'
+            result_files['proposal'] = f'{outfile_prefix}.bbox.json'
+            mmcv.dump(json_results, result_files['bbox'])
+        elif isinstance(results[0], tuple):
+            json_results = self._segm2json(results)
+            result_files['bbox'] = f'{outfile_prefix}.bbox.json'
+            result_files['proposal'] = f'{outfile_prefix}.bbox.json'
+            result_files['segm'] = f'{outfile_prefix}.segm.json'
+            mmcv.dump(json_results[0], result_files['bbox'])
+            mmcv.dump(json_results[1], result_files['segm'])
+        elif isinstance(results[0], np.ndarray):
+            json_results = self._proposal2json(results)
+            result_files['proposal'] = f'{outfile_prefix}.proposal.json'
+            mmcv.dump(json_results, result_files['proposal'])
+        else:
+            raise TypeError('invalid type of results')
+        return result_files
+
+    def fast_eval_recall(self, results, proposal_nums, iou_thrs, logger=None):
+        gt_bboxes = []
+        for i in range(len(self.img_ids)):
+            ann_ids = self.coco.get_ann_ids(img_ids=self.img_ids[i])
+            ann_info = self.coco.load_anns(ann_ids)
+            if len(ann_info) == 0:
+                gt_bboxes.append(np.zeros((0, 4)))
+                continue
+            bboxes = []
+            for ann in ann_info:
+                if ann.get('ignore', False) or ann['iscrowd']:
+                    continue
+                x1, y1, w, h = ann['bbox']
+                bboxes.append([x1, y1, x1 + w, y1 + h])
+            bboxes = np.array(bboxes, dtype=np.float32)
+            if bboxes.shape[0] == 0:
+                bboxes = np.zeros((0, 4))
+            gt_bboxes.append(bboxes)
+
+        recalls = eval_recalls(
+            gt_bboxes, results, proposal_nums, iou_thrs, logger=logger)
+        ar = recalls.mean(axis=1)
+        return ar
+
+    def format_results(self, results, jsonfile_prefix=None, **kwargs):
+        """Format the results to json (standard format for COCO evaluation).
+
+        Args:
+            results (list[tuple | numpy.ndarray]): Testing results of the
+                dataset.
+            jsonfile_prefix (str | None): The prefix of json files. It includes
+                the file path and the prefix of filename, e.g., "a/b/prefix".
+                If not specified, a temp file will be created. Default: None.
+
+        Returns:
+            tuple: (result_files, tmp_dir), result_files is a dict containing \
+                the json filepaths, tmp_dir is the temporal directory created \
+                for saving json files when jsonfile_prefix is not specified.
+        """
+        assert isinstance(results, list), 'results must be a list'
+        assert len(results) == len(self), (
+            'The length of results is not equal to the dataset len: {} != {}'.
+            format(len(results), len(self)))
+
+        if jsonfile_prefix is None:
+            tmp_dir = tempfile.TemporaryDirectory()
+            jsonfile_prefix = osp.join(tmp_dir.name, 'results')
+        else:
+            tmp_dir = None
+        result_files = self.results2json(results, jsonfile_prefix)
+        return result_files, tmp_dir
+
+    def evaluate(self,
+                 results,
+                 metric='bbox',
+                 logger=None,
+                 jsonfile_prefix=None,
+                 classwise=False,
+                 proposal_nums=(100, 300, 1000),
+                 iou_thrs=None,
+                 metric_items=None):
+        """Evaluation in COCO protocol.
+
+        Args:
+            results (list[list | tuple]): Testing results of the dataset.
+            metric (str | list[str]): Metrics to be evaluated. Options are
+                'bbox', 'segm', 'proposal', 'proposal_fast'.
+            logger (logging.Logger | str | None): Logger used for printing
+                related information during evaluation. Default: None.
+            jsonfile_prefix (str | None): The prefix of json files. It includes
+                the file path and the prefix of filename, e.g., "a/b/prefix".
+                If not specified, a temp file will be created. Default: None.
+            classwise (bool): Whether to evaluating the AP for each class.
+            proposal_nums (Sequence[int]): Proposal number used for evaluating
+                recalls, such as recall@100, recall@1000.
+                Default: (100, 300, 1000).
+            iou_thrs (Sequence[float], optional): IoU threshold used for
+                evaluating recalls/mAPs. If set to a list, the average of all
+                IoUs will also be computed. If not specified, [0.50, 0.55,
+                0.60, 0.65, 0.70, 0.75, 0.80, 0.85, 0.90, 0.95] will be used.
+                Default: None.
+            metric_items (list[str] | str, optional): Metric items that will
+                be returned. If not specified, ``['AR@100', 'AR@300',
+                'AR@1000', 'AR_s@1000', 'AR_m@1000', 'AR_l@1000' ]`` will be
+                used when ``metric=='proposal'``, ``['mAP', 'mAP_50', 'mAP_75',
+                'mAP_s', 'mAP_m', 'mAP_l']`` will be used when
+                ``metric=='bbox' or metric=='segm'``.
+
+        Returns:
+            dict[str, float]: COCO style evaluation metric.
+        """
+
+        metrics = metric if isinstance(metric, list) else [metric]
+        allowed_metrics = ['bbox', 'segm', 'proposal', 'proposal_fast']
+        for metric in metrics:
+            if metric not in allowed_metrics:
+                raise KeyError(f'metric {metric} is not supported')
+        if iou_thrs is None:
+            iou_thrs = np.linspace(
+                .5, 0.95, int(np.round((0.95 - .5) / .05)) + 1, endpoint=True)
+        if metric_items is not None:
+            if not isinstance(metric_items, list):
+                metric_items = [metric_items]
+
+        result_files, tmp_dir = self.format_results(results, jsonfile_prefix)
+
+        eval_results = OrderedDict()
+        cocoGt = self.coco
+        for metric in metrics:
+            msg = f'Evaluating {metric}...'
+            if logger is None:
+                msg = '\n' + msg
+            print_log(msg, logger=logger)
+
+            if metric == 'proposal_fast':
+                ar = self.fast_eval_recall(
+                    results, proposal_nums, iou_thrs, logger='silent')
+                log_msg = []
+                for i, num in enumerate(proposal_nums):
+                    eval_results[f'AR@{num}'] = ar[i]
+                    log_msg.append(f'\nAR@{num}\t{ar[i]:.4f}')
+                log_msg = ''.join(log_msg)
+                print_log(log_msg, logger=logger)
+                continue
+
+            if metric not in result_files:
+                raise KeyError(f'{metric} is not in results')
+            try:
+                cocoDt = cocoGt.loadRes(result_files[metric])
+            except IndexError:
+                print_log(
+                    'The testing results of the whole dataset is empty.',
+                    logger=logger,
+                    level=logging.ERROR)
+                break
+
+            iou_type = 'bbox' if metric == 'proposal' else metric
+            cocoEval = COCOeval(cocoGt, cocoDt, iou_type)
+            cocoEval.params.catIds = self.cat_ids
+            cocoEval.params.imgIds = self.img_ids
+            cocoEval.params.maxDets = list(proposal_nums)
+            cocoEval.params.iouThrs = iou_thrs
+            # mapping of cocoEval.stats
+            coco_metric_names = {
+                'mAP': 0,
+                'mAP_50': 1,
+                'mAP_75': 2,
+                'mAP_s': 3,
+                'mAP_m': 4,
+                'mAP_l': 5,
+                'AR@100': 6,
+                'AR@300': 7,
+                'AR@1000': 8,
+                'AR_s@1000': 9,
+                'AR_m@1000': 10,
+                'AR_l@1000': 11
+            }
+            if metric_items is not None:
+                for metric_item in metric_items:
+                    if metric_item not in coco_metric_names:
+                        raise KeyError(
+                            f'metric item {metric_item} is not supported')
+
+            if metric == 'proposal':
+                cocoEval.params.useCats = 0
+                cocoEval.evaluate()
+                cocoEval.accumulate()
+                cocoEval.summarize()
+                if metric_items is None:
+                    metric_items = [
+                        'AR@100', 'AR@300', 'AR@1000', 'AR_s@1000',
+                        'AR_m@1000', 'AR_l@1000'
+                    ]
+
+                for item in metric_items:
+                    val = float(
+                        f'{cocoEval.stats[coco_metric_names[item]]:.3f}')
+                    eval_results[item] = val
+            else:
+                cocoEval.evaluate()
+                cocoEval.accumulate()
+                cocoEval.summarize()
+                if classwise:  # Compute per-category AP
+                    # Compute per-category AP
+                    # from https://github.com/facebookresearch/detectron2/
+                    precisions = cocoEval.eval['precision']
+                    # precision: (iou, recall, cls, area range, max dets)
+                    assert len(self.cat_ids) == precisions.shape[2]
+
+                    results_per_category = []
+                    for idx, catId in enumerate(self.cat_ids):
+                        # area range index 0: all area ranges
+                        # max dets index -1: typically 100 per image
+                        nm = self.coco.loadCats(catId)[0]
+                        precision = precisions[:, :, idx, 0, -1]
+                        precision = precision[precision > -1]
+                        if precision.size:
+                            ap = np.mean(precision)
+                        else:
+                            ap = float('nan')
+                        results_per_category.append(
+                            (f'{nm["name"]}', f'{float(ap):0.3f}'))
+
+                    num_columns = min(6, len(results_per_category) * 2)
+                    results_flatten = list(
+                        itertools.chain(*results_per_category))
+                    headers = ['category', 'AP'] * (num_columns // 2)
+                    results_2d = itertools.zip_longest(*[
+                        results_flatten[i::num_columns]
+                        for i in range(num_columns)
+                    ])
+                    table_data = [headers]
+                    table_data += [result for result in results_2d]
+                    table = AsciiTable(table_data)
+                    print_log('\n' + table.table, logger=logger)
+
+                if metric_items is None:
+                    metric_items = [
+                        'mAP', 'mAP_50', 'mAP_75', 'mAP_s', 'mAP_m', 'mAP_l'
+                    ]
+
+                for metric_item in metric_items:
+                    key = f'{metric}_{metric_item}'
+                    val = float(
+                        f'{cocoEval.stats[coco_metric_names[metric_item]]:.3f}'
+                    )
+                    eval_results[key] = val
+                ap = cocoEval.stats[:6]
+                eval_results[f'{metric}_mAP_copypaste'] = (
+                    f'{ap[0]:.3f} {ap[1]:.3f} {ap[2]:.3f} {ap[3]:.3f} '
+                    f'{ap[4]:.3f} {ap[5]:.3f}')
+        if tmp_dir is not None:
+            tmp_dir.cleanup()
+        return eval_results
diff --git a/python/autoCut/models/swin/mmdet/datasets/custom.py b/python/autoCut/models/swin/mmdet/datasets/custom.py
new file mode 100644
index 0000000..1a2351c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/custom.py
@@ -0,0 +1,323 @@
+import os.path as osp
+import warnings
+from collections import OrderedDict
+
+import mmcv
+import numpy as np
+from mmcv.utils import print_log
+from torch.utils.data import Dataset
+
+from mmdet.core import eval_map, eval_recalls
+from .builder import DATASETS
+from .pipelines import Compose
+
+
+@DATASETS.register_module()
+class CustomDataset(Dataset):
+    """Custom dataset for detection.
+
+    The annotation format is shown as follows. The `ann` field is optional for
+    testing.
+
+    .. code-block:: none
+
+        [
+            {
+                'filename': 'a.jpg',
+                'width': 1280,
+                'height': 720,
+                'ann': {
+                    'bboxes': <np.ndarray> (n, 4) in (x1, y1, x2, y2) order.
+                    'labels': <np.ndarray> (n, ),
+                    'bboxes_ignore': <np.ndarray> (k, 4), (optional field)
+                    'labels_ignore': <np.ndarray> (k, 4) (optional field)
+                }
+            },
+            ...
+        ]
+
+    Args:
+        ann_file (str): Annotation file path.
+        pipeline (list[dict]): Processing pipeline.
+        classes (str | Sequence[str], optional): Specify classes to load.
+            If is None, ``cls.CLASSES`` will be used. Default: None.
+        data_root (str, optional): Data root for ``ann_file``,
+            ``img_prefix``, ``seg_prefix``, ``proposal_file`` if specified.
+        test_mode (bool, optional): If set True, annotation will not be loaded.
+        filter_empty_gt (bool, optional): If set true, images without bounding
+            boxes of the dataset's classes will be filtered out. This option
+            only works when `test_mode=False`, i.e., we never filter images
+            during tests.
+    """
+
+    CLASSES = None
+
+    def __init__(self,
+                 ann_file,
+                 pipeline,
+                 classes=None,
+                 data_root=None,
+                 img_prefix='',
+                 seg_prefix=None,
+                 proposal_file=None,
+                 test_mode=False,
+                 filter_empty_gt=True):
+        self.ann_file = ann_file
+        self.data_root = data_root
+        self.img_prefix = img_prefix
+        self.seg_prefix = seg_prefix
+        self.proposal_file = proposal_file
+        self.test_mode = test_mode
+        self.filter_empty_gt = filter_empty_gt
+        self.CLASSES = self.get_classes(classes)
+
+        # join paths if data_root is specified
+        if self.data_root is not None:
+            if not osp.isabs(self.ann_file):
+                self.ann_file = osp.join(self.data_root, self.ann_file)
+            if not (self.img_prefix is None or osp.isabs(self.img_prefix)):
+                self.img_prefix = osp.join(self.data_root, self.img_prefix)
+            if not (self.seg_prefix is None or osp.isabs(self.seg_prefix)):
+                self.seg_prefix = osp.join(self.data_root, self.seg_prefix)
+            if not (self.proposal_file is None
+                    or osp.isabs(self.proposal_file)):
+                self.proposal_file = osp.join(self.data_root,
+                                              self.proposal_file)
+        # load annotations (and proposals)
+        self.data_infos = self.load_annotations(self.ann_file)
+
+        if self.proposal_file is not None:
+            self.proposals = self.load_proposals(self.proposal_file)
+        else:
+            self.proposals = None
+
+        # filter images too small and containing no annotations
+        if not test_mode:
+            valid_inds = self._filter_imgs()
+            self.data_infos = [self.data_infos[i] for i in valid_inds]
+            if self.proposals is not None:
+                self.proposals = [self.proposals[i] for i in valid_inds]
+            # set group flag for the sampler
+            self._set_group_flag()
+
+        # processing pipeline
+        self.pipeline = Compose(pipeline)
+
+    def __len__(self):
+        """Total number of samples of data."""
+        return len(self.data_infos)
+
+    def load_annotations(self, ann_file):
+        """Load annotation from annotation file."""
+        return mmcv.load(ann_file)
+
+    def load_proposals(self, proposal_file):
+        """Load proposal from proposal file."""
+        return mmcv.load(proposal_file)
+
+    def get_ann_info(self, idx):
+        """Get annotation by index.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            dict: Annotation info of specified index.
+        """
+
+        return self.data_infos[idx]['ann']
+
+    def get_cat_ids(self, idx):
+        """Get category ids by index.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            list[int]: All categories in the image of specified index.
+        """
+
+        return self.data_infos[idx]['ann']['labels'].astype(np.int).tolist()
+
+    def pre_pipeline(self, results):
+        """Prepare results dict for pipeline."""
+        results['img_prefix'] = self.img_prefix
+        results['seg_prefix'] = self.seg_prefix
+        results['proposal_file'] = self.proposal_file
+        results['bbox_fields'] = []
+        results['mask_fields'] = []
+        results['seg_fields'] = []
+
+    def _filter_imgs(self, min_size=32):
+        """Filter images too small."""
+        if self.filter_empty_gt:
+            warnings.warn(
+                'CustomDataset does not support filtering empty gt images.')
+        valid_inds = []
+        for i, img_info in enumerate(self.data_infos):
+            if min(img_info['width'], img_info['height']) >= min_size:
+                valid_inds.append(i)
+        return valid_inds
+
+    def _set_group_flag(self):
+        """Set flag according to image aspect ratio.
+
+        Images with aspect ratio greater than 1 will be set as group 1,
+        otherwise group 0.
+        """
+        self.flag = np.zeros(len(self), dtype=np.uint8)
+        for i in range(len(self)):
+            img_info = self.data_infos[i]
+            if img_info['width'] / img_info['height'] > 1:
+                self.flag[i] = 1
+
+    def _rand_another(self, idx):
+        """Get another random index from the same group as the given index."""
+        pool = np.where(self.flag == self.flag[idx])[0]
+        return np.random.choice(pool)
+
+    def __getitem__(self, idx):
+        """Get training/test data after pipeline.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            dict: Training/test data (with annotation if `test_mode` is set \
+                True).
+        """
+
+        if self.test_mode:
+            return self.prepare_test_img(idx)
+        while True:
+            data = self.prepare_train_img(idx)
+            if data is None:
+                idx = self._rand_another(idx)
+                continue
+            return data
+
+    def prepare_train_img(self, idx):
+        """Get training data and annotations after pipeline.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            dict: Training data and annotation after pipeline with new keys \
+                introduced by pipeline.
+        """
+
+        img_info = self.data_infos[idx]
+        ann_info = self.get_ann_info(idx)
+        results = dict(img_info=img_info, ann_info=ann_info)
+        if self.proposals is not None:
+            results['proposals'] = self.proposals[idx]
+        self.pre_pipeline(results)
+        return self.pipeline(results)
+
+    def prepare_test_img(self, idx):
+        """Get testing data  after pipeline.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            dict: Testing data after pipeline with new keys introduced by \
+                pipeline.
+        """
+
+        img_info = self.data_infos[idx]
+        results = dict(img_info=img_info)
+        if self.proposals is not None:
+            results['proposals'] = self.proposals[idx]
+        self.pre_pipeline(results)
+        return self.pipeline(results)
+
+    @classmethod
+    def get_classes(cls, classes=None):
+        """Get class names of current dataset.
+
+        Args:
+            classes (Sequence[str] | str | None): If classes is None, use
+                default CLASSES defined by builtin dataset. If classes is a
+                string, take it as a file name. The file contains the name of
+                classes where each line contains one class name. If classes is
+                a tuple or list, override the CLASSES defined by the dataset.
+
+        Returns:
+            tuple[str] or list[str]: Names of categories of the dataset.
+        """
+        if classes is None:
+            return cls.CLASSES
+
+        if isinstance(classes, str):
+            # take it as a file path
+            class_names = mmcv.list_from_file(classes)
+        elif isinstance(classes, (tuple, list)):
+            class_names = classes
+        else:
+            raise ValueError(f'Unsupported type {type(classes)} of classes.')
+
+        return class_names
+
+    def format_results(self, results, **kwargs):
+        """Place holder to format result to dataset specific output."""
+
+    def evaluate(self,
+                 results,
+                 metric='mAP',
+                 logger=None,
+                 proposal_nums=(100, 300, 1000),
+                 iou_thr=0.5,
+                 scale_ranges=None):
+        """Evaluate the dataset.
+
+        Args:
+            results (list): Testing results of the dataset.
+            metric (str | list[str]): Metrics to be evaluated.
+            logger (logging.Logger | None | str): Logger used for printing
+                related information during evaluation. Default: None.
+            proposal_nums (Sequence[int]): Proposal number used for evaluating
+                recalls, such as recall@100, recall@1000.
+                Default: (100, 300, 1000).
+            iou_thr (float | list[float]): IoU threshold. Default: 0.5.
+            scale_ranges (list[tuple] | None): Scale ranges for evaluating mAP.
+                Default: None.
+        """
+
+        if not isinstance(metric, str):
+            assert len(metric) == 1
+            metric = metric[0]
+        allowed_metrics = ['mAP', 'recall']
+        if metric not in allowed_metrics:
+            raise KeyError(f'metric {metric} is not supported')
+        annotations = [self.get_ann_info(i) for i in range(len(self))]
+        eval_results = OrderedDict()
+        iou_thrs = [iou_thr] if isinstance(iou_thr, float) else iou_thr
+        if metric == 'mAP':
+            assert isinstance(iou_thrs, list)
+            mean_aps = []
+            for iou_thr in iou_thrs:
+                print_log(f'\n{"-" * 15}iou_thr: {iou_thr}{"-" * 15}')
+                mean_ap, _ = eval_map(
+                    results,
+                    annotations,
+                    scale_ranges=scale_ranges,
+                    iou_thr=iou_thr,
+                    dataset=self.CLASSES,
+                    logger=logger)
+                mean_aps.append(mean_ap)
+                eval_results[f'AP{int(iou_thr * 100):02d}'] = round(mean_ap, 3)
+            eval_results['mAP'] = sum(mean_aps) / len(mean_aps)
+        elif metric == 'recall':
+            gt_bboxes = [ann['bboxes'] for ann in annotations]
+            recalls = eval_recalls(
+                gt_bboxes, results, proposal_nums, iou_thr, logger=logger)
+            for i, num in enumerate(proposal_nums):
+                for j, iou in enumerate(iou_thrs):
+                    eval_results[f'recall@{num}@{iou}'] = recalls[i, j]
+            if recalls.shape[1] > 1:
+                ar = recalls.mean(axis=1)
+                for i, num in enumerate(proposal_nums):
+                    eval_results[f'AR@{num}'] = ar[i]
+        return eval_results
diff --git a/python/autoCut/models/swin/mmdet/datasets/dataset_wrappers.py b/python/autoCut/models/swin/mmdet/datasets/dataset_wrappers.py
new file mode 100644
index 0000000..55ad5cb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/dataset_wrappers.py
@@ -0,0 +1,282 @@
+import bisect
+import math
+from collections import defaultdict
+
+import numpy as np
+from mmcv.utils import print_log
+from torch.utils.data.dataset import ConcatDataset as _ConcatDataset
+
+from .builder import DATASETS
+from .coco import CocoDataset
+
+
+@DATASETS.register_module()
+class ConcatDataset(_ConcatDataset):
+    """A wrapper of concatenated dataset.
+
+    Same as :obj:`torch.utils.data.dataset.ConcatDataset`, but
+    concat the group flag for image aspect ratio.
+
+    Args:
+        datasets (list[:obj:`Dataset`]): A list of datasets.
+        separate_eval (bool): Whether to evaluate the results
+            separately if it is used as validation dataset.
+            Defaults to True.
+    """
+
+    def __init__(self, datasets, separate_eval=True):
+        super(ConcatDataset, self).__init__(datasets)
+        self.CLASSES = datasets[0].CLASSES
+        self.separate_eval = separate_eval
+        if not separate_eval:
+            if any([isinstance(ds, CocoDataset) for ds in datasets]):
+                raise NotImplementedError(
+                    'Evaluating concatenated CocoDataset as a whole is not'
+                    ' supported! Please set "separate_eval=True"')
+            elif len(set([type(ds) for ds in datasets])) != 1:
+                raise NotImplementedError(
+                    'All the datasets should have same types')
+
+        if hasattr(datasets[0], 'flag'):
+            flags = []
+            for i in range(0, len(datasets)):
+                flags.append(datasets[i].flag)
+            self.flag = np.concatenate(flags)
+
+    def get_cat_ids(self, idx):
+        """Get category ids of concatenated dataset by index.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            list[int]: All categories in the image of specified index.
+        """
+
+        if idx < 0:
+            if -idx > len(self):
+                raise ValueError(
+                    'absolute value of index should not exceed dataset length')
+            idx = len(self) + idx
+        dataset_idx = bisect.bisect_right(self.cumulative_sizes, idx)
+        if dataset_idx == 0:
+            sample_idx = idx
+        else:
+            sample_idx = idx - self.cumulative_sizes[dataset_idx - 1]
+        return self.datasets[dataset_idx].get_cat_ids(sample_idx)
+
+    def evaluate(self, results, logger=None, **kwargs):
+        """Evaluate the results.
+
+        Args:
+            results (list[list | tuple]): Testing results of the dataset.
+            logger (logging.Logger | str | None): Logger used for printing
+                related information during evaluation. Default: None.
+
+        Returns:
+            dict[str: float]: AP results of the total dataset or each separate
+            dataset if `self.separate_eval=True`.
+        """
+        assert len(results) == self.cumulative_sizes[-1], \
+            ('Dataset and results have different sizes: '
+             f'{self.cumulative_sizes[-1]} v.s. {len(results)}')
+
+        # Check whether all the datasets support evaluation
+        for dataset in self.datasets:
+            assert hasattr(dataset, 'evaluate'), \
+                    f'{type(dataset)} does not implement evaluate function'
+
+        if self.separate_eval:
+            dataset_idx = -1
+            total_eval_results = dict()
+            for size, dataset in zip(self.cumulative_sizes, self.datasets):
+                start_idx = 0 if dataset_idx == -1 else \
+                    self.cumulative_sizes[dataset_idx]
+                end_idx = self.cumulative_sizes[dataset_idx + 1]
+
+                results_per_dataset = results[start_idx:end_idx]
+                print_log(
+                    f'\nEvaluateing {dataset.ann_file} with '
+                    f'{len(results_per_dataset)} images now',
+                    logger=logger)
+
+                eval_results_per_dataset = dataset.evaluate(
+                    results_per_dataset, logger=logger, **kwargs)
+                dataset_idx += 1
+                for k, v in eval_results_per_dataset.items():
+                    total_eval_results.update({f'{dataset_idx}_{k}': v})
+
+            return total_eval_results
+        elif any([isinstance(ds, CocoDataset) for ds in self.datasets]):
+            raise NotImplementedError(
+                'Evaluating concatenated CocoDataset as a whole is not'
+                ' supported! Please set "separate_eval=True"')
+        elif len(set([type(ds) for ds in self.datasets])) != 1:
+            raise NotImplementedError(
+                'All the datasets should have same types')
+        else:
+            original_data_infos = self.datasets[0].data_infos
+            self.datasets[0].data_infos = sum(
+                [dataset.data_infos for dataset in self.datasets], [])
+            eval_results = self.datasets[0].evaluate(
+                results, logger=logger, **kwargs)
+            self.datasets[0].data_infos = original_data_infos
+            return eval_results
+
+
+@DATASETS.register_module()
+class RepeatDataset(object):
+    """A wrapper of repeated dataset.
+
+    The length of repeated dataset will be `times` larger than the original
+    dataset. This is useful when the data loading time is long but the dataset
+    is small. Using RepeatDataset can reduce the data loading time between
+    epochs.
+
+    Args:
+        dataset (:obj:`Dataset`): The dataset to be repeated.
+        times (int): Repeat times.
+    """
+
+    def __init__(self, dataset, times):
+        self.dataset = dataset
+        self.times = times
+        self.CLASSES = dataset.CLASSES
+        if hasattr(self.dataset, 'flag'):
+            self.flag = np.tile(self.dataset.flag, times)
+
+        self._ori_len = len(self.dataset)
+
+    def __getitem__(self, idx):
+        return self.dataset[idx % self._ori_len]
+
+    def get_cat_ids(self, idx):
+        """Get category ids of repeat dataset by index.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            list[int]: All categories in the image of specified index.
+        """
+
+        return self.dataset.get_cat_ids(idx % self._ori_len)
+
+    def __len__(self):
+        """Length after repetition."""
+        return self.times * self._ori_len
+
+
+# Modified from https://github.com/facebookresearch/detectron2/blob/41d475b75a230221e21d9cac5d69655e3415e3a4/detectron2/data/samplers/distributed_sampler.py#L57 # noqa
+@DATASETS.register_module()
+class ClassBalancedDataset(object):
+    """A wrapper of repeated dataset with repeat factor.
+
+    Suitable for training on class imbalanced datasets like LVIS. Following
+    the sampling strategy in the `paper <https://arxiv.org/abs/1908.03195>`_,
+    in each epoch, an image may appear multiple times based on its
+    "repeat factor".
+    The repeat factor for an image is a function of the frequency the rarest
+    category labeled in that image. The "frequency of category c" in [0, 1]
+    is defined by the fraction of images in the training set (without repeats)
+    in which category c appears.
+    The dataset needs to instantiate :func:`self.get_cat_ids` to support
+    ClassBalancedDataset.
+
+    The repeat factor is computed as followed.
+
+    1. For each category c, compute the fraction # of images
+       that contain it: :math:`f(c)`
+    2. For each category c, compute the category-level repeat factor:
+       :math:`r(c) = max(1, sqrt(t/f(c)))`
+    3. For each image I, compute the image-level repeat factor:
+       :math:`r(I) = max_{c in I} r(c)`
+
+    Args:
+        dataset (:obj:`CustomDataset`): The dataset to be repeated.
+        oversample_thr (float): frequency threshold below which data is
+            repeated. For categories with ``f_c >= oversample_thr``, there is
+            no oversampling. For categories with ``f_c < oversample_thr``, the
+            degree of oversampling following the square-root inverse frequency
+            heuristic above.
+        filter_empty_gt (bool, optional): If set true, images without bounding
+            boxes will not be oversampled. Otherwise, they will be categorized
+            as the pure background class and involved into the oversampling.
+            Default: True.
+    """
+
+    def __init__(self, dataset, oversample_thr, filter_empty_gt=True):
+        self.dataset = dataset
+        self.oversample_thr = oversample_thr
+        self.filter_empty_gt = filter_empty_gt
+        self.CLASSES = dataset.CLASSES
+
+        repeat_factors = self._get_repeat_factors(dataset, oversample_thr)
+        repeat_indices = []
+        for dataset_idx, repeat_factor in enumerate(repeat_factors):
+            repeat_indices.extend([dataset_idx] * math.ceil(repeat_factor))
+        self.repeat_indices = repeat_indices
+
+        flags = []
+        if hasattr(self.dataset, 'flag'):
+            for flag, repeat_factor in zip(self.dataset.flag, repeat_factors):
+                flags.extend([flag] * int(math.ceil(repeat_factor)))
+            assert len(flags) == len(repeat_indices)
+        self.flag = np.asarray(flags, dtype=np.uint8)
+
+    def _get_repeat_factors(self, dataset, repeat_thr):
+        """Get repeat factor for each images in the dataset.
+
+        Args:
+            dataset (:obj:`CustomDataset`): The dataset
+            repeat_thr (float): The threshold of frequency. If an image
+                contains the categories whose frequency below the threshold,
+                it would be repeated.
+
+        Returns:
+            list[float]: The repeat factors for each images in the dataset.
+        """
+
+        # 1. For each category c, compute the fraction # of images
+        #   that contain it: f(c)
+        category_freq = defaultdict(int)
+        num_images = len(dataset)
+        for idx in range(num_images):
+            cat_ids = set(self.dataset.get_cat_ids(idx))
+            if len(cat_ids) == 0 and not self.filter_empty_gt:
+                cat_ids = set([len(self.CLASSES)])
+            for cat_id in cat_ids:
+                category_freq[cat_id] += 1
+        for k, v in category_freq.items():
+            category_freq[k] = v / num_images
+
+        # 2. For each category c, compute the category-level repeat factor:
+        #    r(c) = max(1, sqrt(t/f(c)))
+        category_repeat = {
+            cat_id: max(1.0, math.sqrt(repeat_thr / cat_freq))
+            for cat_id, cat_freq in category_freq.items()
+        }
+
+        # 3. For each image I, compute the image-level repeat factor:
+        #    r(I) = max_{c in I} r(c)
+        repeat_factors = []
+        for idx in range(num_images):
+            cat_ids = set(self.dataset.get_cat_ids(idx))
+            if len(cat_ids) == 0 and not self.filter_empty_gt:
+                cat_ids = set([len(self.CLASSES)])
+            repeat_factor = 1
+            if len(cat_ids) > 0:
+                repeat_factor = max(
+                    {category_repeat[cat_id]
+                     for cat_id in cat_ids})
+            repeat_factors.append(repeat_factor)
+
+        return repeat_factors
+
+    def __getitem__(self, idx):
+        ori_index = self.repeat_indices[idx]
+        return self.dataset[ori_index]
+
+    def __len__(self):
+        """Length after repetition."""
+        return len(self.repeat_indices)
diff --git a/python/autoCut/models/swin/mmdet/datasets/deepfashion.py b/python/autoCut/models/swin/mmdet/datasets/deepfashion.py
new file mode 100644
index 0000000..1125376
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/deepfashion.py
@@ -0,0 +1,10 @@
+from .builder import DATASETS
+from .coco import CocoDataset
+
+
+@DATASETS.register_module()
+class DeepFashionDataset(CocoDataset):
+
+    CLASSES = ('top', 'skirt', 'leggings', 'dress', 'outer', 'pants', 'bag',
+               'neckwear', 'headwear', 'eyeglass', 'belt', 'footwear', 'hair',
+               'skin', 'face')
diff --git a/python/autoCut/models/swin/mmdet/datasets/lvis.py b/python/autoCut/models/swin/mmdet/datasets/lvis.py
new file mode 100644
index 0000000..122c64e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/lvis.py
@@ -0,0 +1,742 @@
+import itertools
+import logging
+import os.path as osp
+import tempfile
+from collections import OrderedDict
+
+import numpy as np
+from mmcv.utils import print_log
+from terminaltables import AsciiTable
+
+from .builder import DATASETS
+from .coco import CocoDataset
+
+
+@DATASETS.register_module()
+class LVISV05Dataset(CocoDataset):
+
+    CLASSES = (
+        'acorn', 'aerosol_can', 'air_conditioner', 'airplane', 'alarm_clock',
+        'alcohol', 'alligator', 'almond', 'ambulance', 'amplifier', 'anklet',
+        'antenna', 'apple', 'apple_juice', 'applesauce', 'apricot', 'apron',
+        'aquarium', 'armband', 'armchair', 'armoire', 'armor', 'artichoke',
+        'trash_can', 'ashtray', 'asparagus', 'atomizer', 'avocado', 'award',
+        'awning', 'ax', 'baby_buggy', 'basketball_backboard', 'backpack',
+        'handbag', 'suitcase', 'bagel', 'bagpipe', 'baguet', 'bait', 'ball',
+        'ballet_skirt', 'balloon', 'bamboo', 'banana', 'Band_Aid', 'bandage',
+        'bandanna', 'banjo', 'banner', 'barbell', 'barge', 'barrel',
+        'barrette', 'barrow', 'baseball_base', 'baseball', 'baseball_bat',
+        'baseball_cap', 'baseball_glove', 'basket', 'basketball_hoop',
+        'basketball', 'bass_horn', 'bat_(animal)', 'bath_mat', 'bath_towel',
+        'bathrobe', 'bathtub', 'batter_(food)', 'battery', 'beachball', 'bead',
+        'beaker', 'bean_curd', 'beanbag', 'beanie', 'bear', 'bed',
+        'bedspread', 'cow', 'beef_(food)', 'beeper', 'beer_bottle', 'beer_can',
+        'beetle', 'bell', 'bell_pepper', 'belt', 'belt_buckle', 'bench',
+        'beret', 'bib', 'Bible', 'bicycle', 'visor', 'binder', 'binoculars',
+        'bird', 'birdfeeder', 'birdbath', 'birdcage', 'birdhouse',
+        'birthday_cake', 'birthday_card', 'biscuit_(bread)', 'pirate_flag',
+        'black_sheep', 'blackboard', 'blanket', 'blazer', 'blender', 'blimp',
+        'blinker', 'blueberry', 'boar', 'gameboard', 'boat', 'bobbin',
+        'bobby_pin', 'boiled_egg', 'bolo_tie', 'deadbolt', 'bolt', 'bonnet',
+        'book', 'book_bag', 'bookcase', 'booklet', 'bookmark',
+        'boom_microphone', 'boot', 'bottle', 'bottle_opener', 'bouquet',
+        'bow_(weapon)', 'bow_(decorative_ribbons)', 'bow-tie', 'bowl',
+        'pipe_bowl', 'bowler_hat', 'bowling_ball', 'bowling_pin',
+        'boxing_glove', 'suspenders', 'bracelet', 'brass_plaque', 'brassiere',
+        'bread-bin', 'breechcloth', 'bridal_gown', 'briefcase',
+        'bristle_brush', 'broccoli', 'broach', 'broom', 'brownie',
+        'brussels_sprouts', 'bubble_gum', 'bucket', 'horse_buggy', 'bull',
+        'bulldog', 'bulldozer', 'bullet_train', 'bulletin_board',
+        'bulletproof_vest', 'bullhorn', 'corned_beef', 'bun', 'bunk_bed',
+        'buoy', 'burrito', 'bus_(vehicle)', 'business_card', 'butcher_knife',
+        'butter', 'butterfly', 'button', 'cab_(taxi)', 'cabana', 'cabin_car',
+        'cabinet', 'locker', 'cake', 'calculator', 'calendar', 'calf',
+        'camcorder', 'camel', 'camera', 'camera_lens', 'camper_(vehicle)',
+        'can', 'can_opener', 'candelabrum', 'candle', 'candle_holder',
+        'candy_bar', 'candy_cane', 'walking_cane', 'canister', 'cannon',
+        'canoe', 'cantaloup', 'canteen', 'cap_(headwear)', 'bottle_cap',
+        'cape', 'cappuccino', 'car_(automobile)', 'railcar_(part_of_a_train)',
+        'elevator_car', 'car_battery', 'identity_card', 'card', 'cardigan',
+        'cargo_ship', 'carnation', 'horse_carriage', 'carrot', 'tote_bag',
+        'cart', 'carton', 'cash_register', 'casserole', 'cassette', 'cast',
+        'cat', 'cauliflower', 'caviar', 'cayenne_(spice)', 'CD_player',
+        'celery', 'cellular_telephone', 'chain_mail', 'chair', 'chaise_longue',
+        'champagne', 'chandelier', 'chap', 'checkbook', 'checkerboard',
+        'cherry', 'chessboard', 'chest_of_drawers_(furniture)',
+        'chicken_(animal)', 'chicken_wire', 'chickpea', 'Chihuahua',
+        'chili_(vegetable)', 'chime', 'chinaware', 'crisp_(potato_chip)',
+        'poker_chip', 'chocolate_bar', 'chocolate_cake', 'chocolate_milk',
+        'chocolate_mousse', 'choker', 'chopping_board', 'chopstick',
+        'Christmas_tree', 'slide', 'cider', 'cigar_box', 'cigarette',
+        'cigarette_case', 'cistern', 'clarinet', 'clasp', 'cleansing_agent',
+        'clementine', 'clip', 'clipboard', 'clock', 'clock_tower',
+        'clothes_hamper', 'clothespin', 'clutch_bag', 'coaster', 'coat',
+        'coat_hanger', 'coatrack', 'cock', 'coconut', 'coffee_filter',
+        'coffee_maker', 'coffee_table', 'coffeepot', 'coil', 'coin',
+        'colander', 'coleslaw', 'coloring_material', 'combination_lock',
+        'pacifier', 'comic_book', 'computer_keyboard', 'concrete_mixer',
+        'cone', 'control', 'convertible_(automobile)', 'sofa_bed', 'cookie',
+        'cookie_jar', 'cooking_utensil', 'cooler_(for_food)',
+        'cork_(bottle_plug)', 'corkboard', 'corkscrew', 'edible_corn',
+        'cornbread', 'cornet', 'cornice', 'cornmeal', 'corset',
+        'romaine_lettuce', 'costume', 'cougar', 'coverall', 'cowbell',
+        'cowboy_hat', 'crab_(animal)', 'cracker', 'crape', 'crate', 'crayon',
+        'cream_pitcher', 'credit_card', 'crescent_roll', 'crib', 'crock_pot',
+        'crossbar', 'crouton', 'crow', 'crown', 'crucifix', 'cruise_ship',
+        'police_cruiser', 'crumb', 'crutch', 'cub_(animal)', 'cube',
+        'cucumber', 'cufflink', 'cup', 'trophy_cup', 'cupcake', 'hair_curler',
+        'curling_iron', 'curtain', 'cushion', 'custard', 'cutting_tool',
+        'cylinder', 'cymbal', 'dachshund', 'dagger', 'dartboard',
+        'date_(fruit)', 'deck_chair', 'deer', 'dental_floss', 'desk',
+        'detergent', 'diaper', 'diary', 'die', 'dinghy', 'dining_table', 'tux',
+        'dish', 'dish_antenna', 'dishrag', 'dishtowel', 'dishwasher',
+        'dishwasher_detergent', 'diskette', 'dispenser', 'Dixie_cup', 'dog',
+        'dog_collar', 'doll', 'dollar', 'dolphin', 'domestic_ass', 'eye_mask',
+        'doorbell', 'doorknob', 'doormat', 'doughnut', 'dove', 'dragonfly',
+        'drawer', 'underdrawers', 'dress', 'dress_hat', 'dress_suit',
+        'dresser', 'drill', 'drinking_fountain', 'drone', 'dropper',
+        'drum_(musical_instrument)', 'drumstick', 'duck', 'duckling',
+        'duct_tape', 'duffel_bag', 'dumbbell', 'dumpster', 'dustpan',
+        'Dutch_oven', 'eagle', 'earphone', 'earplug', 'earring', 'easel',
+        'eclair', 'eel', 'egg', 'egg_roll', 'egg_yolk', 'eggbeater',
+        'eggplant', 'electric_chair', 'refrigerator', 'elephant', 'elk',
+        'envelope', 'eraser', 'escargot', 'eyepatch', 'falcon', 'fan',
+        'faucet', 'fedora', 'ferret', 'Ferris_wheel', 'ferry', 'fig_(fruit)',
+        'fighter_jet', 'figurine', 'file_cabinet', 'file_(tool)', 'fire_alarm',
+        'fire_engine', 'fire_extinguisher', 'fire_hose', 'fireplace',
+        'fireplug', 'fish', 'fish_(food)', 'fishbowl', 'fishing_boat',
+        'fishing_rod', 'flag', 'flagpole', 'flamingo', 'flannel', 'flash',
+        'flashlight', 'fleece', 'flip-flop_(sandal)', 'flipper_(footwear)',
+        'flower_arrangement', 'flute_glass', 'foal', 'folding_chair',
+        'food_processor', 'football_(American)', 'football_helmet',
+        'footstool', 'fork', 'forklift', 'freight_car', 'French_toast',
+        'freshener', 'frisbee', 'frog', 'fruit_juice', 'fruit_salad',
+        'frying_pan', 'fudge', 'funnel', 'futon', 'gag', 'garbage',
+        'garbage_truck', 'garden_hose', 'gargle', 'gargoyle', 'garlic',
+        'gasmask', 'gazelle', 'gelatin', 'gemstone', 'giant_panda',
+        'gift_wrap', 'ginger', 'giraffe', 'cincture',
+        'glass_(drink_container)', 'globe', 'glove', 'goat', 'goggles',
+        'goldfish', 'golf_club', 'golfcart', 'gondola_(boat)', 'goose',
+        'gorilla', 'gourd', 'surgical_gown', 'grape', 'grasshopper', 'grater',
+        'gravestone', 'gravy_boat', 'green_bean', 'green_onion', 'griddle',
+        'grillroom', 'grinder_(tool)', 'grits', 'grizzly', 'grocery_bag',
+        'guacamole', 'guitar', 'gull', 'gun', 'hair_spray', 'hairbrush',
+        'hairnet', 'hairpin', 'ham', 'hamburger', 'hammer', 'hammock',
+        'hamper', 'hamster', 'hair_dryer', 'hand_glass', 'hand_towel',
+        'handcart', 'handcuff', 'handkerchief', 'handle', 'handsaw',
+        'hardback_book', 'harmonium', 'hat', 'hatbox', 'hatch', 'veil',
+        'headband', 'headboard', 'headlight', 'headscarf', 'headset',
+        'headstall_(for_horses)', 'hearing_aid', 'heart', 'heater',
+        'helicopter', 'helmet', 'heron', 'highchair', 'hinge', 'hippopotamus',
+        'hockey_stick', 'hog', 'home_plate_(baseball)', 'honey', 'fume_hood',
+        'hook', 'horse', 'hose', 'hot-air_balloon', 'hotplate', 'hot_sauce',
+        'hourglass', 'houseboat', 'hummingbird', 'hummus', 'polar_bear',
+        'icecream', 'popsicle', 'ice_maker', 'ice_pack', 'ice_skate',
+        'ice_tea', 'igniter', 'incense', 'inhaler', 'iPod',
+        'iron_(for_clothing)', 'ironing_board', 'jacket', 'jam', 'jean',
+        'jeep', 'jelly_bean', 'jersey', 'jet_plane', 'jewelry', 'joystick',
+        'jumpsuit', 'kayak', 'keg', 'kennel', 'kettle', 'key', 'keycard',
+        'kilt', 'kimono', 'kitchen_sink', 'kitchen_table', 'kite', 'kitten',
+        'kiwi_fruit', 'knee_pad', 'knife', 'knight_(chess_piece)',
+        'knitting_needle', 'knob', 'knocker_(on_a_door)', 'koala', 'lab_coat',
+        'ladder', 'ladle', 'ladybug', 'lamb_(animal)', 'lamb-chop', 'lamp',
+        'lamppost', 'lampshade', 'lantern', 'lanyard', 'laptop_computer',
+        'lasagna', 'latch', 'lawn_mower', 'leather', 'legging_(clothing)',
+        'Lego', 'lemon', 'lemonade', 'lettuce', 'license_plate', 'life_buoy',
+        'life_jacket', 'lightbulb', 'lightning_rod', 'lime', 'limousine',
+        'linen_paper', 'lion', 'lip_balm', 'lipstick', 'liquor', 'lizard',
+        'Loafer_(type_of_shoe)', 'log', 'lollipop', 'lotion',
+        'speaker_(stero_equipment)', 'loveseat', 'machine_gun', 'magazine',
+        'magnet', 'mail_slot', 'mailbox_(at_home)', 'mallet', 'mammoth',
+        'mandarin_orange', 'manger', 'manhole', 'map', 'marker', 'martini',
+        'mascot', 'mashed_potato', 'masher', 'mask', 'mast',
+        'mat_(gym_equipment)', 'matchbox', 'mattress', 'measuring_cup',
+        'measuring_stick', 'meatball', 'medicine', 'melon', 'microphone',
+        'microscope', 'microwave_oven', 'milestone', 'milk', 'minivan',
+        'mint_candy', 'mirror', 'mitten', 'mixer_(kitchen_tool)', 'money',
+        'monitor_(computer_equipment) computer_monitor', 'monkey', 'motor',
+        'motor_scooter', 'motor_vehicle', 'motorboat', 'motorcycle',
+        'mound_(baseball)', 'mouse_(animal_rodent)',
+        'mouse_(computer_equipment)', 'mousepad', 'muffin', 'mug', 'mushroom',
+        'music_stool', 'musical_instrument', 'nailfile', 'nameplate', 'napkin',
+        'neckerchief', 'necklace', 'necktie', 'needle', 'nest', 'newsstand',
+        'nightshirt', 'nosebag_(for_animals)', 'noseband_(for_animals)',
+        'notebook', 'notepad', 'nut', 'nutcracker', 'oar', 'octopus_(food)',
+        'octopus_(animal)', 'oil_lamp', 'olive_oil', 'omelet', 'onion',
+        'orange_(fruit)', 'orange_juice', 'oregano', 'ostrich', 'ottoman',
+        'overalls_(clothing)', 'owl', 'packet', 'inkpad', 'pad', 'paddle',
+        'padlock', 'paintbox', 'paintbrush', 'painting', 'pajamas', 'palette',
+        'pan_(for_cooking)', 'pan_(metal_container)', 'pancake', 'pantyhose',
+        'papaya', 'paperclip', 'paper_plate', 'paper_towel', 'paperback_book',
+        'paperweight', 'parachute', 'parakeet', 'parasail_(sports)',
+        'parchment', 'parka', 'parking_meter', 'parrot',
+        'passenger_car_(part_of_a_train)', 'passenger_ship', 'passport',
+        'pastry', 'patty_(food)', 'pea_(food)', 'peach', 'peanut_butter',
+        'pear', 'peeler_(tool_for_fruit_and_vegetables)', 'pegboard',
+        'pelican', 'pen', 'pencil', 'pencil_box', 'pencil_sharpener',
+        'pendulum', 'penguin', 'pennant', 'penny_(coin)', 'pepper',
+        'pepper_mill', 'perfume', 'persimmon', 'baby', 'pet', 'petfood',
+        'pew_(church_bench)', 'phonebook', 'phonograph_record', 'piano',
+        'pickle', 'pickup_truck', 'pie', 'pigeon', 'piggy_bank', 'pillow',
+        'pin_(non_jewelry)', 'pineapple', 'pinecone', 'ping-pong_ball',
+        'pinwheel', 'tobacco_pipe', 'pipe', 'pistol', 'pita_(bread)',
+        'pitcher_(vessel_for_liquid)', 'pitchfork', 'pizza', 'place_mat',
+        'plate', 'platter', 'playing_card', 'playpen', 'pliers',
+        'plow_(farm_equipment)', 'pocket_watch', 'pocketknife',
+        'poker_(fire_stirring_tool)', 'pole', 'police_van', 'polo_shirt',
+        'poncho', 'pony', 'pool_table', 'pop_(soda)', 'portrait',
+        'postbox_(public)', 'postcard', 'poster', 'pot', 'flowerpot', 'potato',
+        'potholder', 'pottery', 'pouch', 'power_shovel', 'prawn', 'printer',
+        'projectile_(weapon)', 'projector', 'propeller', 'prune', 'pudding',
+        'puffer_(fish)', 'puffin', 'pug-dog', 'pumpkin', 'puncher', 'puppet',
+        'puppy', 'quesadilla', 'quiche', 'quilt', 'rabbit', 'race_car',
+        'racket', 'radar', 'radiator', 'radio_receiver', 'radish', 'raft',
+        'rag_doll', 'raincoat', 'ram_(animal)', 'raspberry', 'rat',
+        'razorblade', 'reamer_(juicer)', 'rearview_mirror', 'receipt',
+        'recliner', 'record_player', 'red_cabbage', 'reflector',
+        'remote_control', 'rhinoceros', 'rib_(food)', 'rifle', 'ring',
+        'river_boat', 'road_map', 'robe', 'rocking_chair', 'roller_skate',
+        'Rollerblade', 'rolling_pin', 'root_beer',
+        'router_(computer_equipment)', 'rubber_band', 'runner_(carpet)',
+        'plastic_bag', 'saddle_(on_an_animal)', 'saddle_blanket', 'saddlebag',
+        'safety_pin', 'sail', 'salad', 'salad_plate', 'salami',
+        'salmon_(fish)', 'salmon_(food)', 'salsa', 'saltshaker',
+        'sandal_(type_of_shoe)', 'sandwich', 'satchel', 'saucepan', 'saucer',
+        'sausage', 'sawhorse', 'saxophone', 'scale_(measuring_instrument)',
+        'scarecrow', 'scarf', 'school_bus', 'scissors', 'scoreboard',
+        'scrambled_eggs', 'scraper', 'scratcher', 'screwdriver',
+        'scrubbing_brush', 'sculpture', 'seabird', 'seahorse', 'seaplane',
+        'seashell', 'seedling', 'serving_dish', 'sewing_machine', 'shaker',
+        'shampoo', 'shark', 'sharpener', 'Sharpie', 'shaver_(electric)',
+        'shaving_cream', 'shawl', 'shears', 'sheep', 'shepherd_dog',
+        'sherbert', 'shield', 'shirt', 'shoe', 'shopping_bag', 'shopping_cart',
+        'short_pants', 'shot_glass', 'shoulder_bag', 'shovel', 'shower_head',
+        'shower_curtain', 'shredder_(for_paper)', 'sieve', 'signboard', 'silo',
+        'sink', 'skateboard', 'skewer', 'ski', 'ski_boot', 'ski_parka',
+        'ski_pole', 'skirt', 'sled', 'sleeping_bag', 'sling_(bandage)',
+        'slipper_(footwear)', 'smoothie', 'snake', 'snowboard', 'snowman',
+        'snowmobile', 'soap', 'soccer_ball', 'sock', 'soda_fountain',
+        'carbonated_water', 'sofa', 'softball', 'solar_array', 'sombrero',
+        'soup', 'soup_bowl', 'soupspoon', 'sour_cream', 'soya_milk',
+        'space_shuttle', 'sparkler_(fireworks)', 'spatula', 'spear',
+        'spectacles', 'spice_rack', 'spider', 'sponge', 'spoon', 'sportswear',
+        'spotlight', 'squirrel', 'stapler_(stapling_machine)', 'starfish',
+        'statue_(sculpture)', 'steak_(food)', 'steak_knife',
+        'steamer_(kitchen_appliance)', 'steering_wheel', 'stencil',
+        'stepladder', 'step_stool', 'stereo_(sound_system)', 'stew', 'stirrer',
+        'stirrup', 'stockings_(leg_wear)', 'stool', 'stop_sign', 'brake_light',
+        'stove', 'strainer', 'strap', 'straw_(for_drinking)', 'strawberry',
+        'street_sign', 'streetlight', 'string_cheese', 'stylus', 'subwoofer',
+        'sugar_bowl', 'sugarcane_(plant)', 'suit_(clothing)', 'sunflower',
+        'sunglasses', 'sunhat', 'sunscreen', 'surfboard', 'sushi', 'mop',
+        'sweat_pants', 'sweatband', 'sweater', 'sweatshirt', 'sweet_potato',
+        'swimsuit', 'sword', 'syringe', 'Tabasco_sauce', 'table-tennis_table',
+        'table', 'table_lamp', 'tablecloth', 'tachometer', 'taco', 'tag',
+        'taillight', 'tambourine', 'army_tank', 'tank_(storage_vessel)',
+        'tank_top_(clothing)', 'tape_(sticky_cloth_or_paper)', 'tape_measure',
+        'tapestry', 'tarp', 'tartan', 'tassel', 'tea_bag', 'teacup',
+        'teakettle', 'teapot', 'teddy_bear', 'telephone', 'telephone_booth',
+        'telephone_pole', 'telephoto_lens', 'television_camera',
+        'television_set', 'tennis_ball', 'tennis_racket', 'tequila',
+        'thermometer', 'thermos_bottle', 'thermostat', 'thimble', 'thread',
+        'thumbtack', 'tiara', 'tiger', 'tights_(clothing)', 'timer', 'tinfoil',
+        'tinsel', 'tissue_paper', 'toast_(food)', 'toaster', 'toaster_oven',
+        'toilet', 'toilet_tissue', 'tomato', 'tongs', 'toolbox', 'toothbrush',
+        'toothpaste', 'toothpick', 'cover', 'tortilla', 'tow_truck', 'towel',
+        'towel_rack', 'toy', 'tractor_(farm_equipment)', 'traffic_light',
+        'dirt_bike', 'trailer_truck', 'train_(railroad_vehicle)', 'trampoline',
+        'tray', 'tree_house', 'trench_coat', 'triangle_(musical_instrument)',
+        'tricycle', 'tripod', 'trousers', 'truck', 'truffle_(chocolate)',
+        'trunk', 'vat', 'turban', 'turkey_(bird)', 'turkey_(food)', 'turnip',
+        'turtle', 'turtleneck_(clothing)', 'typewriter', 'umbrella',
+        'underwear', 'unicycle', 'urinal', 'urn', 'vacuum_cleaner', 'valve',
+        'vase', 'vending_machine', 'vent', 'videotape', 'vinegar', 'violin',
+        'vodka', 'volleyball', 'vulture', 'waffle', 'waffle_iron', 'wagon',
+        'wagon_wheel', 'walking_stick', 'wall_clock', 'wall_socket', 'wallet',
+        'walrus', 'wardrobe', 'wasabi', 'automatic_washer', 'watch',
+        'water_bottle', 'water_cooler', 'water_faucet', 'water_filter',
+        'water_heater', 'water_jug', 'water_gun', 'water_scooter', 'water_ski',
+        'water_tower', 'watering_can', 'watermelon', 'weathervane', 'webcam',
+        'wedding_cake', 'wedding_ring', 'wet_suit', 'wheel', 'wheelchair',
+        'whipped_cream', 'whiskey', 'whistle', 'wick', 'wig', 'wind_chime',
+        'windmill', 'window_box_(for_plants)', 'windshield_wiper', 'windsock',
+        'wine_bottle', 'wine_bucket', 'wineglass', 'wing_chair',
+        'blinder_(for_horses)', 'wok', 'wolf', 'wooden_spoon', 'wreath',
+        'wrench', 'wristband', 'wristlet', 'yacht', 'yak', 'yogurt',
+        'yoke_(animal_equipment)', 'zebra', 'zucchini')
+
+    def load_annotations(self, ann_file):
+        """Load annotation from lvis style annotation file.
+
+        Args:
+            ann_file (str): Path of annotation file.
+
+        Returns:
+            list[dict]: Annotation info from LVIS api.
+        """
+
+        try:
+            import lvis
+            assert lvis.__version__ >= '10.5.3'
+            from lvis import LVIS
+        except AssertionError:
+            raise AssertionError('Incompatible version of lvis is installed. '
+                                 'Run pip uninstall lvis first. Then run pip '
+                                 'install mmlvis to install open-mmlab forked '
+                                 'lvis. ')
+        except ImportError:
+            raise ImportError('Package lvis is not installed. Please run pip '
+                              'install mmlvis to install open-mmlab forked '
+                              'lvis.')
+        self.coco = LVIS(ann_file)
+        self.cat_ids = self.coco.get_cat_ids()
+        self.cat2label = {cat_id: i for i, cat_id in enumerate(self.cat_ids)}
+        self.img_ids = self.coco.get_img_ids()
+        data_infos = []
+        for i in self.img_ids:
+            info = self.coco.load_imgs([i])[0]
+            if info['file_name'].startswith('COCO'):
+                # Convert form the COCO 2014 file naming convention of
+                # COCO_[train/val/test]2014_000000000000.jpg to the 2017
+                # naming convention of 000000000000.jpg
+                # (LVIS v1 will fix this naming issue)
+                info['filename'] = info['file_name'][-16:]
+            else:
+                info['filename'] = info['file_name']
+            data_infos.append(info)
+        return data_infos
+
+    def evaluate(self,
+                 results,
+                 metric='bbox',
+                 logger=None,
+                 jsonfile_prefix=None,
+                 classwise=False,
+                 proposal_nums=(100, 300, 1000),
+                 iou_thrs=np.arange(0.5, 0.96, 0.05)):
+        """Evaluation in LVIS protocol.
+
+        Args:
+            results (list[list | tuple]): Testing results of the dataset.
+            metric (str | list[str]): Metrics to be evaluated. Options are
+                'bbox', 'segm', 'proposal', 'proposal_fast'.
+            logger (logging.Logger | str | None): Logger used for printing
+                related information during evaluation. Default: None.
+            jsonfile_prefix (str | None):
+            classwise (bool): Whether to evaluating the AP for each class.
+            proposal_nums (Sequence[int]): Proposal number used for evaluating
+                recalls, such as recall@100, recall@1000.
+                Default: (100, 300, 1000).
+            iou_thrs (Sequence[float]): IoU threshold used for evaluating
+                recalls. If set to a list, the average recall of all IoUs will
+                also be computed. Default: 0.5.
+
+        Returns:
+            dict[str, float]: LVIS style metrics.
+        """
+
+        try:
+            import lvis
+            assert lvis.__version__ >= '10.5.3'
+            from lvis import LVISResults, LVISEval
+        except AssertionError:
+            raise AssertionError('Incompatible version of lvis is installed. '
+                                 'Run pip uninstall lvis first. Then run pip '
+                                 'install mmlvis to install open-mmlab forked '
+                                 'lvis. ')
+        except ImportError:
+            raise ImportError('Package lvis is not installed. Please run pip '
+                              'install mmlvis to install open-mmlab forked '
+                              'lvis.')
+        assert isinstance(results, list), 'results must be a list'
+        assert len(results) == len(self), (
+            'The length of results is not equal to the dataset len: {} != {}'.
+            format(len(results), len(self)))
+
+        metrics = metric if isinstance(metric, list) else [metric]
+        allowed_metrics = ['bbox', 'segm', 'proposal', 'proposal_fast']
+        for metric in metrics:
+            if metric not in allowed_metrics:
+                raise KeyError('metric {} is not supported'.format(metric))
+
+        if jsonfile_prefix is None:
+            tmp_dir = tempfile.TemporaryDirectory()
+            jsonfile_prefix = osp.join(tmp_dir.name, 'results')
+        else:
+            tmp_dir = None
+        result_files = self.results2json(results, jsonfile_prefix)
+
+        eval_results = OrderedDict()
+        # get original api
+        lvis_gt = self.coco
+        for metric in metrics:
+            msg = 'Evaluating {}...'.format(metric)
+            if logger is None:
+                msg = '\n' + msg
+            print_log(msg, logger=logger)
+
+            if metric == 'proposal_fast':
+                ar = self.fast_eval_recall(
+                    results, proposal_nums, iou_thrs, logger='silent')
+                log_msg = []
+                for i, num in enumerate(proposal_nums):
+                    eval_results['AR@{}'.format(num)] = ar[i]
+                    log_msg.append('\nAR@{}\t{:.4f}'.format(num, ar[i]))
+                log_msg = ''.join(log_msg)
+                print_log(log_msg, logger=logger)
+                continue
+
+            if metric not in result_files:
+                raise KeyError('{} is not in results'.format(metric))
+            try:
+                lvis_dt = LVISResults(lvis_gt, result_files[metric])
+            except IndexError:
+                print_log(
+                    'The testing results of the whole dataset is empty.',
+                    logger=logger,
+                    level=logging.ERROR)
+                break
+
+            iou_type = 'bbox' if metric == 'proposal' else metric
+            lvis_eval = LVISEval(lvis_gt, lvis_dt, iou_type)
+            lvis_eval.params.imgIds = self.img_ids
+            if metric == 'proposal':
+                lvis_eval.params.useCats = 0
+                lvis_eval.params.maxDets = list(proposal_nums)
+                lvis_eval.evaluate()
+                lvis_eval.accumulate()
+                lvis_eval.summarize()
+                for k, v in lvis_eval.get_results().items():
+                    if k.startswith('AR'):
+                        val = float('{:.3f}'.format(float(v)))
+                        eval_results[k] = val
+            else:
+                lvis_eval.evaluate()
+                lvis_eval.accumulate()
+                lvis_eval.summarize()
+                lvis_results = lvis_eval.get_results()
+                if classwise:  # Compute per-category AP
+                    # Compute per-category AP
+                    # from https://github.com/facebookresearch/detectron2/
+                    precisions = lvis_eval.eval['precision']
+                    # precision: (iou, recall, cls, area range, max dets)
+                    assert len(self.cat_ids) == precisions.shape[2]
+
+                    results_per_category = []
+                    for idx, catId in enumerate(self.cat_ids):
+                        # area range index 0: all area ranges
+                        # max dets index -1: typically 100 per image
+                        nm = self.coco.load_cats(catId)[0]
+                        precision = precisions[:, :, idx, 0, -1]
+                        precision = precision[precision > -1]
+                        if precision.size:
+                            ap = np.mean(precision)
+                        else:
+                            ap = float('nan')
+                        results_per_category.append(
+                            (f'{nm["name"]}', f'{float(ap):0.3f}'))
+
+                    num_columns = min(6, len(results_per_category) * 2)
+                    results_flatten = list(
+                        itertools.chain(*results_per_category))
+                    headers = ['category', 'AP'] * (num_columns // 2)
+                    results_2d = itertools.zip_longest(*[
+                        results_flatten[i::num_columns]
+                        for i in range(num_columns)
+                    ])
+                    table_data = [headers]
+                    table_data += [result for result in results_2d]
+                    table = AsciiTable(table_data)
+                    print_log('\n' + table.table, logger=logger)
+
+                for k, v in lvis_results.items():
+                    if k.startswith('AP'):
+                        key = '{}_{}'.format(metric, k)
+                        val = float('{:.3f}'.format(float(v)))
+                        eval_results[key] = val
+                ap_summary = ' '.join([
+                    '{}:{:.3f}'.format(k, float(v))
+                    for k, v in lvis_results.items() if k.startswith('AP')
+                ])
+                eval_results['{}_mAP_copypaste'.format(metric)] = ap_summary
+            lvis_eval.print_results()
+        if tmp_dir is not None:
+            tmp_dir.cleanup()
+        return eval_results
+
+
+LVISDataset = LVISV05Dataset
+DATASETS.register_module(name='LVISDataset', module=LVISDataset)
+
+
+@DATASETS.register_module()
+class LVISV1Dataset(LVISDataset):
+
+    CLASSES = (
+        'aerosol_can', 'air_conditioner', 'airplane', 'alarm_clock', 'alcohol',
+        'alligator', 'almond', 'ambulance', 'amplifier', 'anklet', 'antenna',
+        'apple', 'applesauce', 'apricot', 'apron', 'aquarium',
+        'arctic_(type_of_shoe)', 'armband', 'armchair', 'armoire', 'armor',
+        'artichoke', 'trash_can', 'ashtray', 'asparagus', 'atomizer',
+        'avocado', 'award', 'awning', 'ax', 'baboon', 'baby_buggy',
+        'basketball_backboard', 'backpack', 'handbag', 'suitcase', 'bagel',
+        'bagpipe', 'baguet', 'bait', 'ball', 'ballet_skirt', 'balloon',
+        'bamboo', 'banana', 'Band_Aid', 'bandage', 'bandanna', 'banjo',
+        'banner', 'barbell', 'barge', 'barrel', 'barrette', 'barrow',
+        'baseball_base', 'baseball', 'baseball_bat', 'baseball_cap',
+        'baseball_glove', 'basket', 'basketball', 'bass_horn', 'bat_(animal)',
+        'bath_mat', 'bath_towel', 'bathrobe', 'bathtub', 'batter_(food)',
+        'battery', 'beachball', 'bead', 'bean_curd', 'beanbag', 'beanie',
+        'bear', 'bed', 'bedpan', 'bedspread', 'cow', 'beef_(food)', 'beeper',
+        'beer_bottle', 'beer_can', 'beetle', 'bell', 'bell_pepper', 'belt',
+        'belt_buckle', 'bench', 'beret', 'bib', 'Bible', 'bicycle', 'visor',
+        'billboard', 'binder', 'binoculars', 'bird', 'birdfeeder', 'birdbath',
+        'birdcage', 'birdhouse', 'birthday_cake', 'birthday_card',
+        'pirate_flag', 'black_sheep', 'blackberry', 'blackboard', 'blanket',
+        'blazer', 'blender', 'blimp', 'blinker', 'blouse', 'blueberry',
+        'gameboard', 'boat', 'bob', 'bobbin', 'bobby_pin', 'boiled_egg',
+        'bolo_tie', 'deadbolt', 'bolt', 'bonnet', 'book', 'bookcase',
+        'booklet', 'bookmark', 'boom_microphone', 'boot', 'bottle',
+        'bottle_opener', 'bouquet', 'bow_(weapon)', 'bow_(decorative_ribbons)',
+        'bow-tie', 'bowl', 'pipe_bowl', 'bowler_hat', 'bowling_ball', 'box',
+        'boxing_glove', 'suspenders', 'bracelet', 'brass_plaque', 'brassiere',
+        'bread-bin', 'bread', 'breechcloth', 'bridal_gown', 'briefcase',
+        'broccoli', 'broach', 'broom', 'brownie', 'brussels_sprouts',
+        'bubble_gum', 'bucket', 'horse_buggy', 'bull', 'bulldog', 'bulldozer',
+        'bullet_train', 'bulletin_board', 'bulletproof_vest', 'bullhorn',
+        'bun', 'bunk_bed', 'buoy', 'burrito', 'bus_(vehicle)', 'business_card',
+        'butter', 'butterfly', 'button', 'cab_(taxi)', 'cabana', 'cabin_car',
+        'cabinet', 'locker', 'cake', 'calculator', 'calendar', 'calf',
+        'camcorder', 'camel', 'camera', 'camera_lens', 'camper_(vehicle)',
+        'can', 'can_opener', 'candle', 'candle_holder', 'candy_bar',
+        'candy_cane', 'walking_cane', 'canister', 'canoe', 'cantaloup',
+        'canteen', 'cap_(headwear)', 'bottle_cap', 'cape', 'cappuccino',
+        'car_(automobile)', 'railcar_(part_of_a_train)', 'elevator_car',
+        'car_battery', 'identity_card', 'card', 'cardigan', 'cargo_ship',
+        'carnation', 'horse_carriage', 'carrot', 'tote_bag', 'cart', 'carton',
+        'cash_register', 'casserole', 'cassette', 'cast', 'cat', 'cauliflower',
+        'cayenne_(spice)', 'CD_player', 'celery', 'cellular_telephone',
+        'chain_mail', 'chair', 'chaise_longue', 'chalice', 'chandelier',
+        'chap', 'checkbook', 'checkerboard', 'cherry', 'chessboard',
+        'chicken_(animal)', 'chickpea', 'chili_(vegetable)', 'chime',
+        'chinaware', 'crisp_(potato_chip)', 'poker_chip', 'chocolate_bar',
+        'chocolate_cake', 'chocolate_milk', 'chocolate_mousse', 'choker',
+        'chopping_board', 'chopstick', 'Christmas_tree', 'slide', 'cider',
+        'cigar_box', 'cigarette', 'cigarette_case', 'cistern', 'clarinet',
+        'clasp', 'cleansing_agent', 'cleat_(for_securing_rope)', 'clementine',
+        'clip', 'clipboard', 'clippers_(for_plants)', 'cloak', 'clock',
+        'clock_tower', 'clothes_hamper', 'clothespin', 'clutch_bag', 'coaster',
+        'coat', 'coat_hanger', 'coatrack', 'cock', 'cockroach',
+        'cocoa_(beverage)', 'coconut', 'coffee_maker', 'coffee_table',
+        'coffeepot', 'coil', 'coin', 'colander', 'coleslaw',
+        'coloring_material', 'combination_lock', 'pacifier', 'comic_book',
+        'compass', 'computer_keyboard', 'condiment', 'cone', 'control',
+        'convertible_(automobile)', 'sofa_bed', 'cooker', 'cookie',
+        'cooking_utensil', 'cooler_(for_food)', 'cork_(bottle_plug)',
+        'corkboard', 'corkscrew', 'edible_corn', 'cornbread', 'cornet',
+        'cornice', 'cornmeal', 'corset', 'costume', 'cougar', 'coverall',
+        'cowbell', 'cowboy_hat', 'crab_(animal)', 'crabmeat', 'cracker',
+        'crape', 'crate', 'crayon', 'cream_pitcher', 'crescent_roll', 'crib',
+        'crock_pot', 'crossbar', 'crouton', 'crow', 'crowbar', 'crown',
+        'crucifix', 'cruise_ship', 'police_cruiser', 'crumb', 'crutch',
+        'cub_(animal)', 'cube', 'cucumber', 'cufflink', 'cup', 'trophy_cup',
+        'cupboard', 'cupcake', 'hair_curler', 'curling_iron', 'curtain',
+        'cushion', 'cylinder', 'cymbal', 'dagger', 'dalmatian', 'dartboard',
+        'date_(fruit)', 'deck_chair', 'deer', 'dental_floss', 'desk',
+        'detergent', 'diaper', 'diary', 'die', 'dinghy', 'dining_table', 'tux',
+        'dish', 'dish_antenna', 'dishrag', 'dishtowel', 'dishwasher',
+        'dishwasher_detergent', 'dispenser', 'diving_board', 'Dixie_cup',
+        'dog', 'dog_collar', 'doll', 'dollar', 'dollhouse', 'dolphin',
+        'domestic_ass', 'doorknob', 'doormat', 'doughnut', 'dove', 'dragonfly',
+        'drawer', 'underdrawers', 'dress', 'dress_hat', 'dress_suit',
+        'dresser', 'drill', 'drone', 'dropper', 'drum_(musical_instrument)',
+        'drumstick', 'duck', 'duckling', 'duct_tape', 'duffel_bag', 'dumbbell',
+        'dumpster', 'dustpan', 'eagle', 'earphone', 'earplug', 'earring',
+        'easel', 'eclair', 'eel', 'egg', 'egg_roll', 'egg_yolk', 'eggbeater',
+        'eggplant', 'electric_chair', 'refrigerator', 'elephant', 'elk',
+        'envelope', 'eraser', 'escargot', 'eyepatch', 'falcon', 'fan',
+        'faucet', 'fedora', 'ferret', 'Ferris_wheel', 'ferry', 'fig_(fruit)',
+        'fighter_jet', 'figurine', 'file_cabinet', 'file_(tool)', 'fire_alarm',
+        'fire_engine', 'fire_extinguisher', 'fire_hose', 'fireplace',
+        'fireplug', 'first-aid_kit', 'fish', 'fish_(food)', 'fishbowl',
+        'fishing_rod', 'flag', 'flagpole', 'flamingo', 'flannel', 'flap',
+        'flash', 'flashlight', 'fleece', 'flip-flop_(sandal)',
+        'flipper_(footwear)', 'flower_arrangement', 'flute_glass', 'foal',
+        'folding_chair', 'food_processor', 'football_(American)',
+        'football_helmet', 'footstool', 'fork', 'forklift', 'freight_car',
+        'French_toast', 'freshener', 'frisbee', 'frog', 'fruit_juice',
+        'frying_pan', 'fudge', 'funnel', 'futon', 'gag', 'garbage',
+        'garbage_truck', 'garden_hose', 'gargle', 'gargoyle', 'garlic',
+        'gasmask', 'gazelle', 'gelatin', 'gemstone', 'generator',
+        'giant_panda', 'gift_wrap', 'ginger', 'giraffe', 'cincture',
+        'glass_(drink_container)', 'globe', 'glove', 'goat', 'goggles',
+        'goldfish', 'golf_club', 'golfcart', 'gondola_(boat)', 'goose',
+        'gorilla', 'gourd', 'grape', 'grater', 'gravestone', 'gravy_boat',
+        'green_bean', 'green_onion', 'griddle', 'grill', 'grits', 'grizzly',
+        'grocery_bag', 'guitar', 'gull', 'gun', 'hairbrush', 'hairnet',
+        'hairpin', 'halter_top', 'ham', 'hamburger', 'hammer', 'hammock',
+        'hamper', 'hamster', 'hair_dryer', 'hand_glass', 'hand_towel',
+        'handcart', 'handcuff', 'handkerchief', 'handle', 'handsaw',
+        'hardback_book', 'harmonium', 'hat', 'hatbox', 'veil', 'headband',
+        'headboard', 'headlight', 'headscarf', 'headset',
+        'headstall_(for_horses)', 'heart', 'heater', 'helicopter', 'helmet',
+        'heron', 'highchair', 'hinge', 'hippopotamus', 'hockey_stick', 'hog',
+        'home_plate_(baseball)', 'honey', 'fume_hood', 'hook', 'hookah',
+        'hornet', 'horse', 'hose', 'hot-air_balloon', 'hotplate', 'hot_sauce',
+        'hourglass', 'houseboat', 'hummingbird', 'hummus', 'polar_bear',
+        'icecream', 'popsicle', 'ice_maker', 'ice_pack', 'ice_skate',
+        'igniter', 'inhaler', 'iPod', 'iron_(for_clothing)', 'ironing_board',
+        'jacket', 'jam', 'jar', 'jean', 'jeep', 'jelly_bean', 'jersey',
+        'jet_plane', 'jewel', 'jewelry', 'joystick', 'jumpsuit', 'kayak',
+        'keg', 'kennel', 'kettle', 'key', 'keycard', 'kilt', 'kimono',
+        'kitchen_sink', 'kitchen_table', 'kite', 'kitten', 'kiwi_fruit',
+        'knee_pad', 'knife', 'knitting_needle', 'knob', 'knocker_(on_a_door)',
+        'koala', 'lab_coat', 'ladder', 'ladle', 'ladybug', 'lamb_(animal)',
+        'lamb-chop', 'lamp', 'lamppost', 'lampshade', 'lantern', 'lanyard',
+        'laptop_computer', 'lasagna', 'latch', 'lawn_mower', 'leather',
+        'legging_(clothing)', 'Lego', 'legume', 'lemon', 'lemonade', 'lettuce',
+        'license_plate', 'life_buoy', 'life_jacket', 'lightbulb',
+        'lightning_rod', 'lime', 'limousine', 'lion', 'lip_balm', 'liquor',
+        'lizard', 'log', 'lollipop', 'speaker_(stero_equipment)', 'loveseat',
+        'machine_gun', 'magazine', 'magnet', 'mail_slot', 'mailbox_(at_home)',
+        'mallard', 'mallet', 'mammoth', 'manatee', 'mandarin_orange', 'manger',
+        'manhole', 'map', 'marker', 'martini', 'mascot', 'mashed_potato',
+        'masher', 'mask', 'mast', 'mat_(gym_equipment)', 'matchbox',
+        'mattress', 'measuring_cup', 'measuring_stick', 'meatball', 'medicine',
+        'melon', 'microphone', 'microscope', 'microwave_oven', 'milestone',
+        'milk', 'milk_can', 'milkshake', 'minivan', 'mint_candy', 'mirror',
+        'mitten', 'mixer_(kitchen_tool)', 'money',
+        'monitor_(computer_equipment) computer_monitor', 'monkey', 'motor',
+        'motor_scooter', 'motor_vehicle', 'motorcycle', 'mound_(baseball)',
+        'mouse_(computer_equipment)', 'mousepad', 'muffin', 'mug', 'mushroom',
+        'music_stool', 'musical_instrument', 'nailfile', 'napkin',
+        'neckerchief', 'necklace', 'necktie', 'needle', 'nest', 'newspaper',
+        'newsstand', 'nightshirt', 'nosebag_(for_animals)',
+        'noseband_(for_animals)', 'notebook', 'notepad', 'nut', 'nutcracker',
+        'oar', 'octopus_(food)', 'octopus_(animal)', 'oil_lamp', 'olive_oil',
+        'omelet', 'onion', 'orange_(fruit)', 'orange_juice', 'ostrich',
+        'ottoman', 'oven', 'overalls_(clothing)', 'owl', 'packet', 'inkpad',
+        'pad', 'paddle', 'padlock', 'paintbrush', 'painting', 'pajamas',
+        'palette', 'pan_(for_cooking)', 'pan_(metal_container)', 'pancake',
+        'pantyhose', 'papaya', 'paper_plate', 'paper_towel', 'paperback_book',
+        'paperweight', 'parachute', 'parakeet', 'parasail_(sports)', 'parasol',
+        'parchment', 'parka', 'parking_meter', 'parrot',
+        'passenger_car_(part_of_a_train)', 'passenger_ship', 'passport',
+        'pastry', 'patty_(food)', 'pea_(food)', 'peach', 'peanut_butter',
+        'pear', 'peeler_(tool_for_fruit_and_vegetables)', 'wooden_leg',
+        'pegboard', 'pelican', 'pen', 'pencil', 'pencil_box',
+        'pencil_sharpener', 'pendulum', 'penguin', 'pennant', 'penny_(coin)',
+        'pepper', 'pepper_mill', 'perfume', 'persimmon', 'person', 'pet',
+        'pew_(church_bench)', 'phonebook', 'phonograph_record', 'piano',
+        'pickle', 'pickup_truck', 'pie', 'pigeon', 'piggy_bank', 'pillow',
+        'pin_(non_jewelry)', 'pineapple', 'pinecone', 'ping-pong_ball',
+        'pinwheel', 'tobacco_pipe', 'pipe', 'pistol', 'pita_(bread)',
+        'pitcher_(vessel_for_liquid)', 'pitchfork', 'pizza', 'place_mat',
+        'plate', 'platter', 'playpen', 'pliers', 'plow_(farm_equipment)',
+        'plume', 'pocket_watch', 'pocketknife', 'poker_(fire_stirring_tool)',
+        'pole', 'polo_shirt', 'poncho', 'pony', 'pool_table', 'pop_(soda)',
+        'postbox_(public)', 'postcard', 'poster', 'pot', 'flowerpot', 'potato',
+        'potholder', 'pottery', 'pouch', 'power_shovel', 'prawn', 'pretzel',
+        'printer', 'projectile_(weapon)', 'projector', 'propeller', 'prune',
+        'pudding', 'puffer_(fish)', 'puffin', 'pug-dog', 'pumpkin', 'puncher',
+        'puppet', 'puppy', 'quesadilla', 'quiche', 'quilt', 'rabbit',
+        'race_car', 'racket', 'radar', 'radiator', 'radio_receiver', 'radish',
+        'raft', 'rag_doll', 'raincoat', 'ram_(animal)', 'raspberry', 'rat',
+        'razorblade', 'reamer_(juicer)', 'rearview_mirror', 'receipt',
+        'recliner', 'record_player', 'reflector', 'remote_control',
+        'rhinoceros', 'rib_(food)', 'rifle', 'ring', 'river_boat', 'road_map',
+        'robe', 'rocking_chair', 'rodent', 'roller_skate', 'Rollerblade',
+        'rolling_pin', 'root_beer', 'router_(computer_equipment)',
+        'rubber_band', 'runner_(carpet)', 'plastic_bag',
+        'saddle_(on_an_animal)', 'saddle_blanket', 'saddlebag', 'safety_pin',
+        'sail', 'salad', 'salad_plate', 'salami', 'salmon_(fish)',
+        'salmon_(food)', 'salsa', 'saltshaker', 'sandal_(type_of_shoe)',
+        'sandwich', 'satchel', 'saucepan', 'saucer', 'sausage', 'sawhorse',
+        'saxophone', 'scale_(measuring_instrument)', 'scarecrow', 'scarf',
+        'school_bus', 'scissors', 'scoreboard', 'scraper', 'screwdriver',
+        'scrubbing_brush', 'sculpture', 'seabird', 'seahorse', 'seaplane',
+        'seashell', 'sewing_machine', 'shaker', 'shampoo', 'shark',
+        'sharpener', 'Sharpie', 'shaver_(electric)', 'shaving_cream', 'shawl',
+        'shears', 'sheep', 'shepherd_dog', 'sherbert', 'shield', 'shirt',
+        'shoe', 'shopping_bag', 'shopping_cart', 'short_pants', 'shot_glass',
+        'shoulder_bag', 'shovel', 'shower_head', 'shower_cap',
+        'shower_curtain', 'shredder_(for_paper)', 'signboard', 'silo', 'sink',
+        'skateboard', 'skewer', 'ski', 'ski_boot', 'ski_parka', 'ski_pole',
+        'skirt', 'skullcap', 'sled', 'sleeping_bag', 'sling_(bandage)',
+        'slipper_(footwear)', 'smoothie', 'snake', 'snowboard', 'snowman',
+        'snowmobile', 'soap', 'soccer_ball', 'sock', 'sofa', 'softball',
+        'solar_array', 'sombrero', 'soup', 'soup_bowl', 'soupspoon',
+        'sour_cream', 'soya_milk', 'space_shuttle', 'sparkler_(fireworks)',
+        'spatula', 'spear', 'spectacles', 'spice_rack', 'spider', 'crawfish',
+        'sponge', 'spoon', 'sportswear', 'spotlight', 'squid_(food)',
+        'squirrel', 'stagecoach', 'stapler_(stapling_machine)', 'starfish',
+        'statue_(sculpture)', 'steak_(food)', 'steak_knife', 'steering_wheel',
+        'stepladder', 'step_stool', 'stereo_(sound_system)', 'stew', 'stirrer',
+        'stirrup', 'stool', 'stop_sign', 'brake_light', 'stove', 'strainer',
+        'strap', 'straw_(for_drinking)', 'strawberry', 'street_sign',
+        'streetlight', 'string_cheese', 'stylus', 'subwoofer', 'sugar_bowl',
+        'sugarcane_(plant)', 'suit_(clothing)', 'sunflower', 'sunglasses',
+        'sunhat', 'surfboard', 'sushi', 'mop', 'sweat_pants', 'sweatband',
+        'sweater', 'sweatshirt', 'sweet_potato', 'swimsuit', 'sword',
+        'syringe', 'Tabasco_sauce', 'table-tennis_table', 'table',
+        'table_lamp', 'tablecloth', 'tachometer', 'taco', 'tag', 'taillight',
+        'tambourine', 'army_tank', 'tank_(storage_vessel)',
+        'tank_top_(clothing)', 'tape_(sticky_cloth_or_paper)', 'tape_measure',
+        'tapestry', 'tarp', 'tartan', 'tassel', 'tea_bag', 'teacup',
+        'teakettle', 'teapot', 'teddy_bear', 'telephone', 'telephone_booth',
+        'telephone_pole', 'telephoto_lens', 'television_camera',
+        'television_set', 'tennis_ball', 'tennis_racket', 'tequila',
+        'thermometer', 'thermos_bottle', 'thermostat', 'thimble', 'thread',
+        'thumbtack', 'tiara', 'tiger', 'tights_(clothing)', 'timer', 'tinfoil',
+        'tinsel', 'tissue_paper', 'toast_(food)', 'toaster', 'toaster_oven',
+        'toilet', 'toilet_tissue', 'tomato', 'tongs', 'toolbox', 'toothbrush',
+        'toothpaste', 'toothpick', 'cover', 'tortilla', 'tow_truck', 'towel',
+        'towel_rack', 'toy', 'tractor_(farm_equipment)', 'traffic_light',
+        'dirt_bike', 'trailer_truck', 'train_(railroad_vehicle)', 'trampoline',
+        'tray', 'trench_coat', 'triangle_(musical_instrument)', 'tricycle',
+        'tripod', 'trousers', 'truck', 'truffle_(chocolate)', 'trunk', 'vat',
+        'turban', 'turkey_(food)', 'turnip', 'turtle', 'turtleneck_(clothing)',
+        'typewriter', 'umbrella', 'underwear', 'unicycle', 'urinal', 'urn',
+        'vacuum_cleaner', 'vase', 'vending_machine', 'vent', 'vest',
+        'videotape', 'vinegar', 'violin', 'vodka', 'volleyball', 'vulture',
+        'waffle', 'waffle_iron', 'wagon', 'wagon_wheel', 'walking_stick',
+        'wall_clock', 'wall_socket', 'wallet', 'walrus', 'wardrobe',
+        'washbasin', 'automatic_washer', 'watch', 'water_bottle',
+        'water_cooler', 'water_faucet', 'water_heater', 'water_jug',
+        'water_gun', 'water_scooter', 'water_ski', 'water_tower',
+        'watering_can', 'watermelon', 'weathervane', 'webcam', 'wedding_cake',
+        'wedding_ring', 'wet_suit', 'wheel', 'wheelchair', 'whipped_cream',
+        'whistle', 'wig', 'wind_chime', 'windmill', 'window_box_(for_plants)',
+        'windshield_wiper', 'windsock', 'wine_bottle', 'wine_bucket',
+        'wineglass', 'blinder_(for_horses)', 'wok', 'wolf', 'wooden_spoon',
+        'wreath', 'wrench', 'wristband', 'wristlet', 'yacht', 'yogurt',
+        'yoke_(animal_equipment)', 'zebra', 'zucchini')
+
+    def load_annotations(self, ann_file):
+        try:
+            import lvis
+            assert lvis.__version__ >= '10.5.3'
+            from lvis import LVIS
+        except AssertionError:
+            raise AssertionError('Incompatible version of lvis is installed. '
+                                 'Run pip uninstall lvis first. Then run pip '
+                                 'install mmlvis to install open-mmlab forked '
+                                 'lvis. ')
+        except ImportError:
+            raise ImportError('Package lvis is not installed. Please run pip '
+                              'install mmlvis to install open-mmlab forked '
+                              'lvis.')
+        self.coco = LVIS(ann_file)
+        self.cat_ids = self.coco.get_cat_ids()
+        self.cat2label = {cat_id: i for i, cat_id in enumerate(self.cat_ids)}
+        self.img_ids = self.coco.get_img_ids()
+        data_infos = []
+        for i in self.img_ids:
+            info = self.coco.load_imgs([i])[0]
+            # coco_url is used in LVISv1 instead of file_name
+            # e.g. http://images.cocodataset.org/train2017/000000391895.jpg
+            # train/val split in specified in url
+            info['filename'] = info['coco_url'].replace(
+                'http://images.cocodataset.org/', '')
+            data_infos.append(info)
+        return data_infos
diff --git a/python/autoCut/models/swin/mmdet/datasets/pipelines/__init__.py b/python/autoCut/models/swin/mmdet/datasets/pipelines/__init__.py
new file mode 100644
index 0000000..dd1e3d6
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/pipelines/__init__.py
@@ -0,0 +1,25 @@
+from .auto_augment import (AutoAugment, BrightnessTransform, ColorTransform,
+                           ContrastTransform, EqualizeTransform, Rotate, Shear,
+                           Translate)
+from .compose import Compose
+from .formating import (Collect, DefaultFormatBundle, ImageToTensor,
+                        ToDataContainer, ToTensor, Transpose, to_tensor, RPDV2FormatBundle)
+from .instaboost import InstaBoost
+from .loading import (LoadAnnotations, LoadImageFromFile, LoadImageFromWebcam,
+                      LoadMultiChannelImageFromFiles, LoadProposals, LoadRPDV2Annotations)
+from .test_time_aug import MultiScaleFlipAug
+from .transforms import (Albu, CutOut, Expand, MinIoURandomCrop, Normalize,
+                         Pad, PhotoMetricDistortion, RandomCenterCropPad,
+                         RandomCrop, RandomFlip, Resize, SegRescale)
+
+__all__ = [
+    'Compose', 'to_tensor', 'ToTensor', 'ImageToTensor', 'ToDataContainer',
+    'Transpose', 'Collect', 'DefaultFormatBundle', 'LoadAnnotations',
+    'LoadImageFromFile', 'LoadImageFromWebcam',
+    'LoadMultiChannelImageFromFiles', 'LoadProposals', 'MultiScaleFlipAug',
+    'Resize', 'RandomFlip', 'Pad', 'RandomCrop', 'Normalize', 'SegRescale',
+    'MinIoURandomCrop', 'Expand', 'PhotoMetricDistortion', 'Albu',
+    'InstaBoost', 'RandomCenterCropPad', 'AutoAugment', 'CutOut', 'Shear',
+    'Rotate', 'ColorTransform', 'EqualizeTransform', 'BrightnessTransform',
+    'ContrastTransform', 'Translate', 'LoadRPDV2Annotations', 'RPDV2FormatBundle'
+]
diff --git a/python/autoCut/models/swin/mmdet/datasets/pipelines/auto_augment.py b/python/autoCut/models/swin/mmdet/datasets/pipelines/auto_augment.py
new file mode 100644
index 0000000..e19adae
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/pipelines/auto_augment.py
@@ -0,0 +1,890 @@
+import copy
+
+import cv2
+import mmcv
+import numpy as np
+
+from ..builder import PIPELINES
+from .compose import Compose
+
+_MAX_LEVEL = 10
+
+
+def level_to_value(level, max_value):
+    """Map from level to values based on max_value."""
+    return (level / _MAX_LEVEL) * max_value
+
+
+def enhance_level_to_value(level, a=1.8, b=0.1):
+    """Map from level to values."""
+    return (level / _MAX_LEVEL) * a + b
+
+
+def random_negative(value, random_negative_prob):
+    """Randomly negate value based on random_negative_prob."""
+    return -value if np.random.rand() < random_negative_prob else value
+
+
+def bbox2fields():
+    """The key correspondence from bboxes to labels, masks and
+    segmentations."""
+    bbox2label = {
+        'gt_bboxes': 'gt_labels',
+        'gt_bboxes_ignore': 'gt_labels_ignore'
+    }
+    bbox2mask = {
+        'gt_bboxes': 'gt_masks',
+        'gt_bboxes_ignore': 'gt_masks_ignore'
+    }
+    bbox2seg = {
+        'gt_bboxes': 'gt_semantic_seg',
+    }
+    return bbox2label, bbox2mask, bbox2seg
+
+
+@PIPELINES.register_module()
+class AutoAugment(object):
+    """Auto augmentation.
+
+    This data augmentation is proposed in `Learning Data Augmentation
+    Strategies for Object Detection <https://arxiv.org/pdf/1906.11172>`_.
+
+    TODO: Implement 'Shear', 'Sharpness' and 'Rotate' transforms
+
+    Args:
+        policies (list[list[dict]]): The policies of auto augmentation. Each
+            policy in ``policies`` is a specific augmentation policy, and is
+            composed by several augmentations (dict). When AutoAugment is
+            called, a random policy in ``policies`` will be selected to
+            augment images.
+
+    Examples:
+        >>> replace = (104, 116, 124)
+        >>> policies = [
+        >>>     [
+        >>>         dict(type='Sharpness', prob=0.0, level=8),
+        >>>         dict(
+        >>>             type='Shear',
+        >>>             prob=0.4,
+        >>>             level=0,
+        >>>             replace=replace,
+        >>>             axis='x')
+        >>>     ],
+        >>>     [
+        >>>         dict(
+        >>>             type='Rotate',
+        >>>             prob=0.6,
+        >>>             level=10,
+        >>>             replace=replace),
+        >>>         dict(type='Color', prob=1.0, level=6)
+        >>>     ]
+        >>> ]
+        >>> augmentation = AutoAugment(policies)
+        >>> img = np.ones(100, 100, 3)
+        >>> gt_bboxes = np.ones(10, 4)
+        >>> results = dict(img=img, gt_bboxes=gt_bboxes)
+        >>> results = augmentation(results)
+    """
+
+    def __init__(self, policies):
+        assert isinstance(policies, list) and len(policies) > 0, \
+            'Policies must be a non-empty list.'
+        for policy in policies:
+            assert isinstance(policy, list) and len(policy) > 0, \
+                'Each policy in policies must be a non-empty list.'
+            for augment in policy:
+                assert isinstance(augment, dict) and 'type' in augment, \
+                    'Each specific augmentation must be a dict with key' \
+                    ' "type".'
+
+        self.policies = copy.deepcopy(policies)
+        self.transforms = [Compose(policy) for policy in self.policies]
+
+    def __call__(self, results):
+        transform = np.random.choice(self.transforms)
+        return transform(results)
+
+    def __repr__(self):
+        return f'{self.__class__.__name__}(policies={self.policies})'
+
+
+@PIPELINES.register_module()
+class Shear(object):
+    """Apply Shear Transformation to image (and its corresponding bbox, mask,
+    segmentation).
+
+    Args:
+        level (int | float): The level should be in range [0,_MAX_LEVEL].
+        img_fill_val (int | float | tuple): The filled values for image border.
+            If float, the same fill value will be used for all the three
+            channels of image. If tuple, the should be 3 elements.
+        seg_ignore_label (int): The fill value used for segmentation map.
+            Note this value must equals ``ignore_label`` in ``semantic_head``
+            of the corresponding config. Default 255.
+        prob (float): The probability for performing Shear and should be in
+            range [0, 1].
+        direction (str): The direction for shear, either "horizontal"
+            or "vertical".
+        max_shear_magnitude (float): The maximum magnitude for Shear
+            transformation.
+        random_negative_prob (float): The probability that turns the
+                offset negative. Should be in range [0,1]
+        interpolation (str): Same as in :func:`mmcv.imshear`.
+    """
+
+    def __init__(self,
+                 level,
+                 img_fill_val=128,
+                 seg_ignore_label=255,
+                 prob=0.5,
+                 direction='horizontal',
+                 max_shear_magnitude=0.3,
+                 random_negative_prob=0.5,
+                 interpolation='bilinear'):
+        assert isinstance(level, (int, float)), 'The level must be type ' \
+            f'int or float, got {type(level)}.'
+        assert 0 <= level <= _MAX_LEVEL, 'The level should be in range ' \
+            f'[0,{_MAX_LEVEL}], got {level}.'
+        if isinstance(img_fill_val, (float, int)):
+            img_fill_val = tuple([float(img_fill_val)] * 3)
+        elif isinstance(img_fill_val, tuple):
+            assert len(img_fill_val) == 3, 'img_fill_val as tuple must ' \
+                f'have 3 elements. got {len(img_fill_val)}.'
+            img_fill_val = tuple([float(val) for val in img_fill_val])
+        else:
+            raise ValueError(
+                'img_fill_val must be float or tuple with 3 elements.')
+        assert np.all([0 <= val <= 255 for val in img_fill_val]), 'all ' \
+            'elements of img_fill_val should between range [0,255].' \
+            f'got {img_fill_val}.'
+        assert 0 <= prob <= 1.0, 'The probability of shear should be in ' \
+            f'range [0,1]. got {prob}.'
+        assert direction in ('horizontal', 'vertical'), 'direction must ' \
+            f'in be either "horizontal" or "vertical". got {direction}.'
+        assert isinstance(max_shear_magnitude, float), 'max_shear_magnitude ' \
+            f'should be type float. got {type(max_shear_magnitude)}.'
+        assert 0. <= max_shear_magnitude <= 1., 'Defaultly ' \
+            'max_shear_magnitude should be in range [0,1]. ' \
+            f'got {max_shear_magnitude}.'
+        self.level = level
+        self.magnitude = level_to_value(level, max_shear_magnitude)
+        self.img_fill_val = img_fill_val
+        self.seg_ignore_label = seg_ignore_label
+        self.prob = prob
+        self.direction = direction
+        self.max_shear_magnitude = max_shear_magnitude
+        self.random_negative_prob = random_negative_prob
+        self.interpolation = interpolation
+
+    def _shear_img(self,
+                   results,
+                   magnitude,
+                   direction='horizontal',
+                   interpolation='bilinear'):
+        """Shear the image.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+            magnitude (int | float): The magnitude used for shear.
+            direction (str): The direction for shear, either "horizontal"
+                or "vertical".
+            interpolation (str): Same as in :func:`mmcv.imshear`.
+        """
+        for key in results.get('img_fields', ['img']):
+            img = results[key]
+            img_sheared = mmcv.imshear(
+                img,
+                magnitude,
+                direction,
+                border_value=self.img_fill_val,
+                interpolation=interpolation)
+            results[key] = img_sheared.astype(img.dtype)
+
+    def _shear_bboxes(self, results, magnitude):
+        """Shear the bboxes."""
+        h, w, c = results['img_shape']
+        if self.direction == 'horizontal':
+            shear_matrix = np.stack([[1, magnitude],
+                                     [0, 1]]).astype(np.float32)  # [2, 2]
+        else:
+            shear_matrix = np.stack([[1, 0], [magnitude,
+                                              1]]).astype(np.float32)
+        for key in results.get('bbox_fields', []):
+            min_x, min_y, max_x, max_y = np.split(
+                results[key], results[key].shape[-1], axis=-1)
+            coordinates = np.stack([[min_x, min_y], [max_x, min_y],
+                                    [min_x, max_y],
+                                    [max_x, max_y]])  # [4, 2, nb_box, 1]
+            coordinates = coordinates[..., 0].transpose(
+                (2, 1, 0)).astype(np.float32)  # [nb_box, 2, 4]
+            new_coords = np.matmul(shear_matrix[None, :, :],
+                                   coordinates)  # [nb_box, 2, 4]
+            min_x = np.min(new_coords[:, 0, :], axis=-1)
+            min_y = np.min(new_coords[:, 1, :], axis=-1)
+            max_x = np.max(new_coords[:, 0, :], axis=-1)
+            max_y = np.max(new_coords[:, 1, :], axis=-1)
+            min_x = np.clip(min_x, a_min=0, a_max=w)
+            min_y = np.clip(min_y, a_min=0, a_max=h)
+            max_x = np.clip(max_x, a_min=min_x, a_max=w)
+            max_y = np.clip(max_y, a_min=min_y, a_max=h)
+            results[key] = np.stack([min_x, min_y, max_x, max_y],
+                                    axis=-1).astype(results[key].dtype)
+
+    def _shear_masks(self,
+                     results,
+                     magnitude,
+                     direction='horizontal',
+                     fill_val=0,
+                     interpolation='bilinear'):
+        """Shear the masks."""
+        h, w, c = results['img_shape']
+        for key in results.get('mask_fields', []):
+            masks = results[key]
+            results[key] = masks.shear((h, w),
+                                       magnitude,
+                                       direction,
+                                       border_value=fill_val,
+                                       interpolation=interpolation)
+
+    def _shear_seg(self,
+                   results,
+                   magnitude,
+                   direction='horizontal',
+                   fill_val=255,
+                   interpolation='bilinear'):
+        """Shear the segmentation maps."""
+        for key in results.get('seg_fields', []):
+            seg = results[key]
+            results[key] = mmcv.imshear(
+                seg,
+                magnitude,
+                direction,
+                border_value=fill_val,
+                interpolation=interpolation).astype(seg.dtype)
+
+    def _filter_invalid(self, results, min_bbox_size=0):
+        """Filter bboxes and corresponding masks too small after shear
+        augmentation."""
+        bbox2label, bbox2mask, _ = bbox2fields()
+        for key in results.get('bbox_fields', []):
+            bbox_w = results[key][:, 2] - results[key][:, 0]
+            bbox_h = results[key][:, 3] - results[key][:, 1]
+            valid_inds = (bbox_w > min_bbox_size) & (bbox_h > min_bbox_size)
+            valid_inds = np.nonzero(valid_inds)[0]
+            results[key] = results[key][valid_inds]
+            # label fields. e.g. gt_labels and gt_labels_ignore
+            label_key = bbox2label.get(key)
+            if label_key in results:
+                results[label_key] = results[label_key][valid_inds]
+            # mask fields, e.g. gt_masks and gt_masks_ignore
+            mask_key = bbox2mask.get(key)
+            if mask_key in results:
+                results[mask_key] = results[mask_key][valid_inds]
+
+    def __call__(self, results):
+        """Call function to shear images, bounding boxes, masks and semantic
+        segmentation maps.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Sheared results.
+        """
+        if np.random.rand() > self.prob:
+            return results
+        magnitude = random_negative(self.magnitude, self.random_negative_prob)
+        self._shear_img(results, magnitude, self.direction, self.interpolation)
+        self._shear_bboxes(results, magnitude)
+        # fill_val set to 0 for background of mask.
+        self._shear_masks(
+            results,
+            magnitude,
+            self.direction,
+            fill_val=0,
+            interpolation=self.interpolation)
+        self._shear_seg(
+            results,
+            magnitude,
+            self.direction,
+            fill_val=self.seg_ignore_label,
+            interpolation=self.interpolation)
+        self._filter_invalid(results)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(level={self.level}, '
+        repr_str += f'img_fill_val={self.img_fill_val}, '
+        repr_str += f'seg_ignore_label={self.seg_ignore_label}, '
+        repr_str += f'prob={self.prob}, '
+        repr_str += f'direction={self.direction}, '
+        repr_str += f'max_shear_magnitude={self.max_shear_magnitude}, '
+        repr_str += f'random_negative_prob={self.random_negative_prob}, '
+        repr_str += f'interpolation={self.interpolation})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class Rotate(object):
+    """Apply Rotate Transformation to image (and its corresponding bbox, mask,
+    segmentation).
+
+    Args:
+        level (int | float): The level should be in range (0,_MAX_LEVEL].
+        scale (int | float): Isotropic scale factor. Same in
+            ``mmcv.imrotate``.
+        center (int | float | tuple[float]): Center point (w, h) of the
+            rotation in the source image. If None, the center of the
+            image will be used. Same in ``mmcv.imrotate``.
+        img_fill_val (int | float | tuple): The fill value for image border.
+            If float, the same value will be used for all the three
+            channels of image. If tuple, the should be 3 elements (e.g.
+            equals the number of channels for image).
+        seg_ignore_label (int): The fill value used for segmentation map.
+            Note this value must equals ``ignore_label`` in ``semantic_head``
+            of the corresponding config. Default 255.
+        prob (float): The probability for perform transformation and
+            should be in range 0 to 1.
+        max_rotate_angle (int | float): The maximum angles for rotate
+            transformation.
+        random_negative_prob (float): The probability that turns the
+             offset negative.
+    """
+
+    def __init__(self,
+                 level,
+                 scale=1,
+                 center=None,
+                 img_fill_val=128,
+                 seg_ignore_label=255,
+                 prob=0.5,
+                 max_rotate_angle=30,
+                 random_negative_prob=0.5):
+        assert isinstance(level, (int, float)), \
+            f'The level must be type int or float. got {type(level)}.'
+        assert 0 <= level <= _MAX_LEVEL, \
+            f'The level should be in range (0,{_MAX_LEVEL}]. got {level}.'
+        assert isinstance(scale, (int, float)), \
+            f'The scale must be type int or float. got type {type(scale)}.'
+        if isinstance(center, (int, float)):
+            center = (center, center)
+        elif isinstance(center, tuple):
+            assert len(center) == 2, 'center with type tuple must have '\
+                f'2 elements. got {len(center)} elements.'
+        else:
+            assert center is None, 'center must be None or type int, '\
+                f'float or tuple, got type {type(center)}.'
+        if isinstance(img_fill_val, (float, int)):
+            img_fill_val = tuple([float(img_fill_val)] * 3)
+        elif isinstance(img_fill_val, tuple):
+            assert len(img_fill_val) == 3, 'img_fill_val as tuple must '\
+                f'have 3 elements. got {len(img_fill_val)}.'
+            img_fill_val = tuple([float(val) for val in img_fill_val])
+        else:
+            raise ValueError(
+                'img_fill_val must be float or tuple with 3 elements.')
+        assert np.all([0 <= val <= 255 for val in img_fill_val]), \
+            'all elements of img_fill_val should between range [0,255]. '\
+            f'got {img_fill_val}.'
+        assert 0 <= prob <= 1.0, 'The probability should be in range [0,1]. '\
+            'got {prob}.'
+        assert isinstance(max_rotate_angle, (int, float)), 'max_rotate_angle '\
+            f'should be type int or float. got type {type(max_rotate_angle)}.'
+        self.level = level
+        self.scale = scale
+        # Rotation angle in degrees. Positive values mean
+        # clockwise rotation.
+        self.angle = level_to_value(level, max_rotate_angle)
+        self.center = center
+        self.img_fill_val = img_fill_val
+        self.seg_ignore_label = seg_ignore_label
+        self.prob = prob
+        self.max_rotate_angle = max_rotate_angle
+        self.random_negative_prob = random_negative_prob
+
+    def _rotate_img(self, results, angle, center=None, scale=1.0):
+        """Rotate the image.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+            angle (float): Rotation angle in degrees, positive values
+                mean clockwise rotation. Same in ``mmcv.imrotate``.
+            center (tuple[float], optional): Center point (w, h) of the
+                rotation. Same in ``mmcv.imrotate``.
+            scale (int | float): Isotropic scale factor. Same in
+                ``mmcv.imrotate``.
+        """
+        for key in results.get('img_fields', ['img']):
+            img = results[key].copy()
+            img_rotated = mmcv.imrotate(
+                img, angle, center, scale, border_value=self.img_fill_val)
+            results[key] = img_rotated.astype(img.dtype)
+
+    def _rotate_bboxes(self, results, rotate_matrix):
+        """Rotate the bboxes."""
+        h, w, c = results['img_shape']
+        for key in results.get('bbox_fields', []):
+            min_x, min_y, max_x, max_y = np.split(
+                results[key], results[key].shape[-1], axis=-1)
+            coordinates = np.stack([[min_x, min_y], [max_x, min_y],
+                                    [min_x, max_y],
+                                    [max_x, max_y]])  # [4, 2, nb_bbox, 1]
+            # pad 1 to convert from format [x, y] to homogeneous
+            # coordinates format [x, y, 1]
+            coordinates = np.concatenate(
+                (coordinates,
+                 np.ones((4, 1, coordinates.shape[2], 1), coordinates.dtype)),
+                axis=1)  # [4, 3, nb_bbox, 1]
+            coordinates = coordinates.transpose(
+                (2, 0, 1, 3))  # [nb_bbox, 4, 3, 1]
+            rotated_coords = np.matmul(rotate_matrix,
+                                       coordinates)  # [nb_bbox, 4, 2, 1]
+            rotated_coords = rotated_coords[..., 0]  # [nb_bbox, 4, 2]
+            min_x, min_y = np.min(
+                rotated_coords[:, :, 0], axis=1), np.min(
+                    rotated_coords[:, :, 1], axis=1)
+            max_x, max_y = np.max(
+                rotated_coords[:, :, 0], axis=1), np.max(
+                    rotated_coords[:, :, 1], axis=1)
+            min_x, min_y = np.clip(
+                min_x, a_min=0, a_max=w), np.clip(
+                    min_y, a_min=0, a_max=h)
+            max_x, max_y = np.clip(
+                max_x, a_min=min_x, a_max=w), np.clip(
+                    max_y, a_min=min_y, a_max=h)
+            results[key] = np.stack([min_x, min_y, max_x, max_y],
+                                    axis=-1).astype(results[key].dtype)
+
+    def _rotate_masks(self,
+                      results,
+                      angle,
+                      center=None,
+                      scale=1.0,
+                      fill_val=0):
+        """Rotate the masks."""
+        h, w, c = results['img_shape']
+        for key in results.get('mask_fields', []):
+            masks = results[key]
+            results[key] = masks.rotate((h, w), angle, center, scale, fill_val)
+
+    def _rotate_seg(self,
+                    results,
+                    angle,
+                    center=None,
+                    scale=1.0,
+                    fill_val=255):
+        """Rotate the segmentation map."""
+        for key in results.get('seg_fields', []):
+            seg = results[key].copy()
+            results[key] = mmcv.imrotate(
+                seg, angle, center, scale,
+                border_value=fill_val).astype(seg.dtype)
+
+    def _filter_invalid(self, results, min_bbox_size=0):
+        """Filter bboxes and corresponding masks too small after rotate
+        augmentation."""
+        bbox2label, bbox2mask, _ = bbox2fields()
+        for key in results.get('bbox_fields', []):
+            bbox_w = results[key][:, 2] - results[key][:, 0]
+            bbox_h = results[key][:, 3] - results[key][:, 1]
+            valid_inds = (bbox_w > min_bbox_size) & (bbox_h > min_bbox_size)
+            valid_inds = np.nonzero(valid_inds)[0]
+            results[key] = results[key][valid_inds]
+            # label fields. e.g. gt_labels and gt_labels_ignore
+            label_key = bbox2label.get(key)
+            if label_key in results:
+                results[label_key] = results[label_key][valid_inds]
+            # mask fields, e.g. gt_masks and gt_masks_ignore
+            mask_key = bbox2mask.get(key)
+            if mask_key in results:
+                results[mask_key] = results[mask_key][valid_inds]
+
+    def __call__(self, results):
+        """Call function to rotate images, bounding boxes, masks and semantic
+        segmentation maps.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Rotated results.
+        """
+        if np.random.rand() > self.prob:
+            return results
+        h, w = results['img'].shape[:2]
+        center = self.center
+        if center is None:
+            center = ((w - 1) * 0.5, (h - 1) * 0.5)
+        angle = random_negative(self.angle, self.random_negative_prob)
+        self._rotate_img(results, angle, center, self.scale)
+        rotate_matrix = cv2.getRotationMatrix2D(center, -angle, self.scale)
+        self._rotate_bboxes(results, rotate_matrix)
+        self._rotate_masks(results, angle, center, self.scale, fill_val=0)
+        self._rotate_seg(
+            results, angle, center, self.scale, fill_val=self.seg_ignore_label)
+        self._filter_invalid(results)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(level={self.level}, '
+        repr_str += f'scale={self.scale}, '
+        repr_str += f'center={self.center}, '
+        repr_str += f'img_fill_val={self.img_fill_val}, '
+        repr_str += f'seg_ignore_label={self.seg_ignore_label}, '
+        repr_str += f'prob={self.prob}, '
+        repr_str += f'max_rotate_angle={self.max_rotate_angle}, '
+        repr_str += f'random_negative_prob={self.random_negative_prob})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class Translate(object):
+    """Translate the images, bboxes, masks and segmentation maps horizontally
+    or vertically.
+
+    Args:
+        level (int | float): The level for Translate and should be in
+            range [0,_MAX_LEVEL].
+        prob (float): The probability for performing translation and
+            should be in range [0, 1].
+        img_fill_val (int | float | tuple): The filled value for image
+            border. If float, the same fill value will be used for all
+            the three channels of image. If tuple, the should be 3
+            elements (e.g. equals the number of channels for image).
+        seg_ignore_label (int): The fill value used for segmentation map.
+            Note this value must equals ``ignore_label`` in ``semantic_head``
+            of the corresponding config. Default 255.
+        direction (str): The translate direction, either "horizontal"
+            or "vertical".
+        max_translate_offset (int | float): The maximum pixel's offset for
+            Translate.
+        random_negative_prob (float): The probability that turns the
+            offset negative.
+        min_size (int | float): The minimum pixel for filtering
+            invalid bboxes after the translation.
+    """
+
+    def __init__(self,
+                 level,
+                 prob=0.5,
+                 img_fill_val=128,
+                 seg_ignore_label=255,
+                 direction='horizontal',
+                 max_translate_offset=250.,
+                 random_negative_prob=0.5,
+                 min_size=0):
+        assert isinstance(level, (int, float)), \
+            'The level must be type int or float.'
+        assert 0 <= level <= _MAX_LEVEL, \
+            'The level used for calculating Translate\'s offset should be ' \
+            'in range [0,_MAX_LEVEL]'
+        assert 0 <= prob <= 1.0, \
+            'The probability of translation should be in range [0, 1].'
+        if isinstance(img_fill_val, (float, int)):
+            img_fill_val = tuple([float(img_fill_val)] * 3)
+        elif isinstance(img_fill_val, tuple):
+            assert len(img_fill_val) == 3, \
+                'img_fill_val as tuple must have 3 elements.'
+            img_fill_val = tuple([float(val) for val in img_fill_val])
+        else:
+            raise ValueError('img_fill_val must be type float or tuple.')
+        assert np.all([0 <= val <= 255 for val in img_fill_val]), \
+            'all elements of img_fill_val should between range [0,255].'
+        assert direction in ('horizontal', 'vertical'), \
+            'direction should be "horizontal" or "vertical".'
+        assert isinstance(max_translate_offset, (int, float)), \
+            'The max_translate_offset must be type int or float.'
+        # the offset used for translation
+        self.offset = int(level_to_value(level, max_translate_offset))
+        self.level = level
+        self.prob = prob
+        self.img_fill_val = img_fill_val
+        self.seg_ignore_label = seg_ignore_label
+        self.direction = direction
+        self.max_translate_offset = max_translate_offset
+        self.random_negative_prob = random_negative_prob
+        self.min_size = min_size
+
+    def _translate_img(self, results, offset, direction='horizontal'):
+        """Translate the image.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+            offset (int | float): The offset for translate.
+            direction (str): The translate direction, either "horizontal"
+                or "vertical".
+        """
+        for key in results.get('img_fields', ['img']):
+            img = results[key].copy()
+            results[key] = mmcv.imtranslate(
+                img, offset, direction, self.img_fill_val).astype(img.dtype)
+
+    def _translate_bboxes(self, results, offset):
+        """Shift bboxes horizontally or vertically, according to offset."""
+        h, w, c = results['img_shape']
+        for key in results.get('bbox_fields', []):
+            min_x, min_y, max_x, max_y = np.split(
+                results[key], results[key].shape[-1], axis=-1)
+            if self.direction == 'horizontal':
+                min_x = np.maximum(0, min_x + offset)
+                max_x = np.minimum(w, max_x + offset)
+            elif self.direction == 'vertical':
+                min_y = np.maximum(0, min_y + offset)
+                max_y = np.minimum(h, max_y + offset)
+
+            # the boxes translated outside of image will be filtered along with
+            # the corresponding masks, by invoking ``_filter_invalid``.
+            results[key] = np.concatenate([min_x, min_y, max_x, max_y],
+                                          axis=-1)
+
+    def _translate_masks(self,
+                         results,
+                         offset,
+                         direction='horizontal',
+                         fill_val=0):
+        """Translate masks horizontally or vertically."""
+        h, w, c = results['img_shape']
+        for key in results.get('mask_fields', []):
+            masks = results[key]
+            results[key] = masks.translate((h, w), offset, direction, fill_val)
+
+    def _translate_seg(self,
+                       results,
+                       offset,
+                       direction='horizontal',
+                       fill_val=255):
+        """Translate segmentation maps horizontally or vertically."""
+        for key in results.get('seg_fields', []):
+            seg = results[key].copy()
+            results[key] = mmcv.imtranslate(seg, offset, direction,
+                                            fill_val).astype(seg.dtype)
+
+    def _filter_invalid(self, results, min_size=0):
+        """Filter bboxes and masks too small or translated out of image."""
+        bbox2label, bbox2mask, _ = bbox2fields()
+        for key in results.get('bbox_fields', []):
+            bbox_w = results[key][:, 2] - results[key][:, 0]
+            bbox_h = results[key][:, 3] - results[key][:, 1]
+            valid_inds = (bbox_w > min_size) & (bbox_h > min_size)
+            valid_inds = np.nonzero(valid_inds)[0]
+            results[key] = results[key][valid_inds]
+            # label fields. e.g. gt_labels and gt_labels_ignore
+            label_key = bbox2label.get(key)
+            if label_key in results:
+                results[label_key] = results[label_key][valid_inds]
+            # mask fields, e.g. gt_masks and gt_masks_ignore
+            mask_key = bbox2mask.get(key)
+            if mask_key in results:
+                results[mask_key] = results[mask_key][valid_inds]
+        return results
+
+    def __call__(self, results):
+        """Call function to translate images, bounding boxes, masks and
+        semantic segmentation maps.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Translated results.
+        """
+        if np.random.rand() > self.prob:
+            return results
+        offset = random_negative(self.offset, self.random_negative_prob)
+        self._translate_img(results, offset, self.direction)
+        self._translate_bboxes(results, offset)
+        # fill_val defaultly 0 for BitmapMasks and None for PolygonMasks.
+        self._translate_masks(results, offset, self.direction)
+        # fill_val set to ``seg_ignore_label`` for the ignored value
+        # of segmentation map.
+        self._translate_seg(
+            results, offset, self.direction, fill_val=self.seg_ignore_label)
+        self._filter_invalid(results, min_size=self.min_size)
+        return results
+
+
+@PIPELINES.register_module()
+class ColorTransform(object):
+    """Apply Color transformation to image. The bboxes, masks, and
+    segmentations are not modified.
+
+    Args:
+        level (int | float): Should be in range [0,_MAX_LEVEL].
+        prob (float): The probability for performing Color transformation.
+    """
+
+    def __init__(self, level, prob=0.5):
+        assert isinstance(level, (int, float)), \
+            'The level must be type int or float.'
+        assert 0 <= level <= _MAX_LEVEL, \
+            'The level should be in range [0,_MAX_LEVEL].'
+        assert 0 <= prob <= 1.0, \
+            'The probability should be in range [0,1].'
+        self.level = level
+        self.prob = prob
+        self.factor = enhance_level_to_value(level)
+
+    def _adjust_color_img(self, results, factor=1.0):
+        """Apply Color transformation to image."""
+        for key in results.get('img_fields', ['img']):
+            # NOTE defaultly the image should be BGR format
+            img = results[key]
+            results[key] = mmcv.adjust_color(img, factor).astype(img.dtype)
+
+    def __call__(self, results):
+        """Call function for Color transformation.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Colored results.
+        """
+        if np.random.rand() > self.prob:
+            return results
+        self._adjust_color_img(results, self.factor)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(level={self.level}, '
+        repr_str += f'prob={self.prob})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class EqualizeTransform(object):
+    """Apply Equalize transformation to image. The bboxes, masks and
+    segmentations are not modified.
+
+    Args:
+        prob (float): The probability for performing Equalize transformation.
+    """
+
+    def __init__(self, prob=0.5):
+        assert 0 <= prob <= 1.0, \
+            'The probability should be in range [0,1].'
+        self.prob = prob
+
+    def _imequalize(self, results):
+        """Equalizes the histogram of one image."""
+        for key in results.get('img_fields', ['img']):
+            img = results[key]
+            results[key] = mmcv.imequalize(img).astype(img.dtype)
+
+    def __call__(self, results):
+        """Call function for Equalize transformation.
+
+        Args:
+            results (dict): Results dict from loading pipeline.
+
+        Returns:
+            dict: Results after the transformation.
+        """
+        if np.random.rand() > self.prob:
+            return results
+        self._imequalize(results)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(prob={self.prob})'
+
+
+@PIPELINES.register_module()
+class BrightnessTransform(object):
+    """Apply Brightness transformation to image. The bboxes, masks and
+    segmentations are not modified.
+
+    Args:
+        level (int | float): Should be in range [0,_MAX_LEVEL].
+        prob (float): The probability for performing Brightness transformation.
+    """
+
+    def __init__(self, level, prob=0.5):
+        assert isinstance(level, (int, float)), \
+            'The level must be type int or float.'
+        assert 0 <= level <= _MAX_LEVEL, \
+            'The level should be in range [0,_MAX_LEVEL].'
+        assert 0 <= prob <= 1.0, \
+            'The probability should be in range [0,1].'
+        self.level = level
+        self.prob = prob
+        self.factor = enhance_level_to_value(level)
+
+    def _adjust_brightness_img(self, results, factor=1.0):
+        """Adjust the brightness of image."""
+        for key in results.get('img_fields', ['img']):
+            img = results[key]
+            results[key] = mmcv.adjust_brightness(img,
+                                                  factor).astype(img.dtype)
+
+    def __call__(self, results):
+        """Call function for Brightness transformation.
+
+        Args:
+            results (dict): Results dict from loading pipeline.
+
+        Returns:
+            dict: Results after the transformation.
+        """
+        if np.random.rand() > self.prob:
+            return results
+        self._adjust_brightness_img(results, self.factor)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(level={self.level}, '
+        repr_str += f'prob={self.prob})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class ContrastTransform(object):
+    """Apply Contrast transformation to image. The bboxes, masks and
+    segmentations are not modified.
+
+    Args:
+        level (int | float): Should be in range [0,_MAX_LEVEL].
+        prob (float): The probability for performing Contrast transformation.
+    """
+
+    def __init__(self, level, prob=0.5):
+        assert isinstance(level, (int, float)), \
+            'The level must be type int or float.'
+        assert 0 <= level <= _MAX_LEVEL, \
+            'The level should be in range [0,_MAX_LEVEL].'
+        assert 0 <= prob <= 1.0, \
+            'The probability should be in range [0,1].'
+        self.level = level
+        self.prob = prob
+        self.factor = enhance_level_to_value(level)
+
+    def _adjust_contrast_img(self, results, factor=1.0):
+        """Adjust the image contrast."""
+        for key in results.get('img_fields', ['img']):
+            img = results[key]
+            results[key] = mmcv.adjust_contrast(img, factor).astype(img.dtype)
+
+    def __call__(self, results):
+        """Call function for Contrast transformation.
+
+        Args:
+            results (dict): Results dict from loading pipeline.
+
+        Returns:
+            dict: Results after the transformation.
+        """
+        if np.random.rand() > self.prob:
+            return results
+        self._adjust_contrast_img(results, self.factor)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(level={self.level}, '
+        repr_str += f'prob={self.prob})'
+        return repr_str
diff --git a/python/autoCut/models/swin/mmdet/datasets/pipelines/compose.py b/python/autoCut/models/swin/mmdet/datasets/pipelines/compose.py
new file mode 100644
index 0000000..ca48f1c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/pipelines/compose.py
@@ -0,0 +1,51 @@
+import collections
+
+from mmcv.utils import build_from_cfg
+
+from ..builder import PIPELINES
+
+
+@PIPELINES.register_module()
+class Compose(object):
+    """Compose multiple transforms sequentially.
+
+    Args:
+        transforms (Sequence[dict | callable]): Sequence of transform object or
+            config dict to be composed.
+    """
+
+    def __init__(self, transforms):
+        assert isinstance(transforms, collections.abc.Sequence)
+        self.transforms = []
+        for transform in transforms:
+            if isinstance(transform, dict):
+                transform = build_from_cfg(transform, PIPELINES)
+                self.transforms.append(transform)
+            elif callable(transform):
+                self.transforms.append(transform)
+            else:
+                raise TypeError('transform must be callable or a dict')
+
+    def __call__(self, data):
+        """Call function to apply transforms sequentially.
+
+        Args:
+            data (dict): A result dict contains the data to transform.
+
+        Returns:
+           dict: Transformed data.
+        """
+
+        for t in self.transforms:
+            data = t(data)
+            if data is None:
+                return None
+        return data
+
+    def __repr__(self):
+        format_string = self.__class__.__name__ + '('
+        for t in self.transforms:
+            format_string += '\n'
+            format_string += f'    {t}'
+        format_string += '\n)'
+        return format_string
diff --git a/python/autoCut/models/swin/mmdet/datasets/pipelines/formating.py b/python/autoCut/models/swin/mmdet/datasets/pipelines/formating.py
new file mode 100644
index 0000000..5b6145c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/pipelines/formating.py
@@ -0,0 +1,451 @@
+from collections.abc import Sequence
+
+import mmcv
+import numpy as np
+import torch
+from mmcv.parallel import DataContainer as DC
+
+from ..builder import PIPELINES
+
+
+def to_tensor(data):
+    """Convert objects of various python types to :obj:`torch.Tensor`.
+
+    Supported types are: :class:`numpy.ndarray`, :class:`torch.Tensor`,
+    :class:`Sequence`, :class:`int` and :class:`float`.
+
+    Args:
+        data (torch.Tensor | numpy.ndarray | Sequence | int | float): Data to
+            be converted.
+    """
+
+    if isinstance(data, torch.Tensor):
+        return data
+    elif isinstance(data, np.ndarray):
+        return torch.from_numpy(data)
+    elif isinstance(data, Sequence) and not mmcv.is_str(data):
+        return torch.tensor(data)
+    elif isinstance(data, int):
+        return torch.LongTensor([data])
+    elif isinstance(data, float):
+        return torch.FloatTensor([data])
+    else:
+        raise TypeError(f'type {type(data)} cannot be converted to tensor.')
+
+
+@PIPELINES.register_module()
+class ToTensor(object):
+    """Convert some results to :obj:`torch.Tensor` by given keys.
+
+    Args:
+        keys (Sequence[str]): Keys that need to be converted to Tensor.
+    """
+
+    def __init__(self, keys):
+        self.keys = keys
+
+    def __call__(self, results):
+        """Call function to convert data in results to :obj:`torch.Tensor`.
+
+        Args:
+            results (dict): Result dict contains the data to convert.
+
+        Returns:
+            dict: The result dict contains the data converted
+                to :obj:`torch.Tensor`.
+        """
+        for key in self.keys:
+            results[key] = to_tensor(results[key])
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__ + f'(keys={self.keys})'
+
+
+@PIPELINES.register_module()
+class ImageToTensor(object):
+    """Convert image to :obj:`torch.Tensor` by given keys.
+
+    The dimension order of input image is (H, W, C). The pipeline will convert
+    it to (C, H, W). If only 2 dimension (H, W) is given, the output would be
+    (1, H, W).
+
+    Args:
+        keys (Sequence[str]): Key of images to be converted to Tensor.
+    """
+
+    def __init__(self, keys):
+        self.keys = keys
+
+    def __call__(self, results):
+        """Call function to convert image in results to :obj:`torch.Tensor` and
+        transpose the channel order.
+
+        Args:
+            results (dict): Result dict contains the image data to convert.
+
+        Returns:
+            dict: The result dict contains the image converted
+                to :obj:`torch.Tensor` and transposed to (C, H, W) order.
+        """
+        for key in self.keys:
+            img = results[key]
+            if len(img.shape) < 3:
+                img = np.expand_dims(img, -1)
+            results[key] = to_tensor(img.transpose(2, 0, 1))
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__ + f'(keys={self.keys})'
+
+
+@PIPELINES.register_module()
+class Transpose(object):
+    """Transpose some results by given keys.
+
+    Args:
+        keys (Sequence[str]): Keys of results to be transposed.
+        order (Sequence[int]): Order of transpose.
+    """
+
+    def __init__(self, keys, order):
+        self.keys = keys
+        self.order = order
+
+    def __call__(self, results):
+        """Call function to transpose the channel order of data in results.
+
+        Args:
+            results (dict): Result dict contains the data to transpose.
+
+        Returns:
+            dict: The result dict contains the data transposed to \
+                ``self.order``.
+        """
+        for key in self.keys:
+            results[key] = results[key].transpose(self.order)
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__ + \
+            f'(keys={self.keys}, order={self.order})'
+
+
+@PIPELINES.register_module()
+class ToDataContainer(object):
+    """Convert results to :obj:`mmcv.DataContainer` by given fields.
+
+    Args:
+        fields (Sequence[dict]): Each field is a dict like
+            ``dict(key='xxx', **kwargs)``. The ``key`` in result will
+            be converted to :obj:`mmcv.DataContainer` with ``**kwargs``.
+            Default: ``(dict(key='img', stack=True), dict(key='gt_bboxes'),
+            dict(key='gt_labels'))``.
+    """
+
+    def __init__(self,
+                 fields=(dict(key='img', stack=True), dict(key='gt_bboxes'),
+                         dict(key='gt_labels'))):
+        self.fields = fields
+
+    def __call__(self, results):
+        """Call function to convert data in results to
+        :obj:`mmcv.DataContainer`.
+
+        Args:
+            results (dict): Result dict contains the data to convert.
+
+        Returns:
+            dict: The result dict contains the data converted to \
+                :obj:`mmcv.DataContainer`.
+        """
+
+        for field in self.fields:
+            field = field.copy()
+            key = field.pop('key')
+            results[key] = DC(results[key], **field)
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__ + f'(fields={self.fields})'
+
+
+@PIPELINES.register_module()
+class DefaultFormatBundle(object):
+    """Default formatting bundle.
+
+    It simplifies the pipeline of formatting common fields, including "img",
+    "proposals", "gt_bboxes", "gt_labels", "gt_masks" and "gt_semantic_seg".
+    These fields are formatted as follows.
+
+    - img: (1)transpose, (2)to tensor, (3)to DataContainer (stack=True)
+    - proposals: (1)to tensor, (2)to DataContainer
+    - gt_bboxes: (1)to tensor, (2)to DataContainer
+    - gt_bboxes_ignore: (1)to tensor, (2)to DataContainer
+    - gt_labels: (1)to tensor, (2)to DataContainer
+    - gt_masks: (1)to tensor, (2)to DataContainer (cpu_only=True)
+    - gt_semantic_seg: (1)unsqueeze dim-0 (2)to tensor, \
+                       (3)to DataContainer (stack=True)
+    """
+
+    def __call__(self, results):
+        """Call function to transform and format common fields in results.
+
+        Args:
+            results (dict): Result dict contains the data to convert.
+
+        Returns:
+            dict: The result dict contains the data that is formatted with \
+                default bundle.
+        """
+
+        if 'img' in results:
+            img = results['img']
+            # add default meta keys
+            results = self._add_default_meta_keys(results)
+            if len(img.shape) < 3:
+                img = np.expand_dims(img, -1)
+            img = np.ascontiguousarray(img.transpose(2, 0, 1))
+            results['img'] = DC(to_tensor(img), stack=True)
+        for key in ['proposals', 'gt_bboxes', 'gt_bboxes_ignore', 'gt_labels']:
+            if key not in results:
+                continue
+            results[key] = DC(to_tensor(results[key]))
+        if 'gt_masks' in results:
+            results['gt_masks'] = DC(results['gt_masks'], cpu_only=True)
+        if 'gt_semantic_seg' in results:
+            results['gt_semantic_seg'] = DC(
+                to_tensor(results['gt_semantic_seg'][None, ...]), stack=True)
+        return results
+
+    def _add_default_meta_keys(self, results):
+        """Add default meta keys.
+
+        We set default meta keys including `pad_shape`, `scale_factor` and
+        `img_norm_cfg` to avoid the case where no `Resize`, `Normalize` and
+        `Pad` are implemented during the whole pipeline.
+
+        Args:
+            results (dict): Result dict contains the data to convert.
+
+        Returns:
+            results (dict): Updated result dict contains the data to convert.
+        """
+        img = results['img']
+        results.setdefault('pad_shape', img.shape)
+        results.setdefault('scale_factor', 1.0)
+        num_channels = 1 if len(img.shape) < 3 else img.shape[2]
+        results.setdefault(
+            'img_norm_cfg',
+            dict(
+                mean=np.zeros(num_channels, dtype=np.float32),
+                std=np.ones(num_channels, dtype=np.float32),
+                to_rgb=False))
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__
+
+
+@PIPELINES.register_module()
+class RPDV2FormatBundle(object):
+    """Default formatting bundle.
+
+    It simplifies the pipeline of formatting common fields, including "img",
+    "proposals", "gt_bboxes", "gt_labels", "gt_masks" and "gt_semantic_seg".
+    These fields are formatted as follows.
+
+    - img: (1)transpose, (2)to tensor, (3)to DataContainer (stack=True)
+    - proposals: (1)to tensor, (2)to DataContainer
+    - gt_bboxes: (1)to tensor, (2)to DataContainer
+    - gt_bboxes_ignore: (1)to tensor, (2)to DataContainer
+    - gt_labels: (1)to tensor, (2)to DataContainer
+    - gt_masks: (1)to tensor, (2)to DataContainer (cpu_only=True)
+    - gt_semantic_seg: (1)unsqueeze dim-0 (2)to tensor,
+                       (3)to DataContainer (stack=True)
+    """
+    
+    def __init__(self):
+        super(RPDV2FormatBundle, self).__init__()
+
+    def __call__(self, results):
+        """Call function to transform and format common fields in results.
+
+        Args:
+            results (dict): Result dict contains the data to convert.
+
+        Returns:
+            dict: The result dict contains the data that is formatted with
+                default bundle.
+        """
+
+        if 'img' in results:
+            img = results['img']
+            # add default meta keys
+            results = self._add_default_meta_keys(results)
+            if len(img.shape) < 3:
+                img = np.expand_dims(img, -1)
+            img = np.ascontiguousarray(img.transpose(2, 0, 1))
+            results['img'] = DC(to_tensor(img), stack=True)
+        for key in ['proposals', 'gt_bboxes', 'gt_bboxes_ignore', 'gt_labels']:
+            if key not in results:
+                continue
+            results[key] = DC(to_tensor(results[key]))
+        if 'gt_masks' in results:
+            results['gt_masks'] = DC(results['gt_masks'], cpu_only=True)
+        if 'gt_semantic_seg' in results:
+            results['gt_semantic_seg'] = DC(
+                to_tensor(results['gt_semantic_seg'][None, ...]), stack=True)
+        if 'gt_sem_map' in results:
+            results['gt_sem_map'] = DC(to_tensor(results['gt_sem_map']), stack=True)
+        if 'gt_sem_weights' in results:
+            results['gt_sem_weights'] = DC(to_tensor(results['gt_sem_weights']), stack=True)
+        if 'gt_contours' in results:
+            results['gt_contours'] = DC(to_tensor(results['gt_contours']))
+
+        return results
+
+    def _add_default_meta_keys(self, results):
+        """Add default meta keys.
+
+        We set default meta keys including `pad_shape`, `scale_factor` and
+        `img_norm_cfg` to avoid the case where no `Resize`, `Normalize` and
+        `Pad` are implemented during the whole pipeline.
+
+        Args:
+            results (dict): Result dict contains the data to convert.
+
+        Returns:
+            results (dict): Updated result dict contains the data to convert.
+        """
+        img = results['img']
+        results.setdefault('pad_shape', img.shape)
+        results.setdefault('scale_factor', 1.0)
+        num_channels = 1 if len(img.shape) < 3 else img.shape[2]
+        results.setdefault(
+            'img_norm_cfg',
+            dict(
+                mean=np.zeros(num_channels, dtype=np.float32),
+                std=np.ones(num_channels, dtype=np.float32),
+                to_rgb=False))
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__
+
+
+@PIPELINES.register_module()
+class Collect(object):
+    """Collect data from the loader relevant to the specific task.
+
+    This is usually the last stage of the data loader pipeline. Typically keys
+    is set to some subset of "img", "proposals", "gt_bboxes",
+    "gt_bboxes_ignore", "gt_labels", and/or "gt_masks".
+
+    The "img_meta" item is always populated.  The contents of the "img_meta"
+    dictionary depends on "meta_keys". By default this includes:
+
+        - "img_shape": shape of the image input to the network as a tuple \
+            (h, w, c).  Note that images may be zero padded on the \
+            bottom/right if the batch tensor is larger than this shape.
+
+        - "scale_factor": a float indicating the preprocessing scale
+
+        - "flip": a boolean indicating if image flip transform was used
+
+        - "filename": path to the image file
+
+        - "ori_shape": original shape of the image as a tuple (h, w, c)
+
+        - "pad_shape": image shape after padding
+
+        - "img_norm_cfg": a dict of normalization information:
+
+            - mean - per channel mean subtraction
+            - std - per channel std divisor
+            - to_rgb - bool indicating if bgr was converted to rgb
+
+    Args:
+        keys (Sequence[str]): Keys of results to be collected in ``data``.
+        meta_keys (Sequence[str], optional): Meta keys to be converted to
+            ``mmcv.DataContainer`` and collected in ``data[img_metas]``.
+            Default: ``('filename', 'ori_filename', 'ori_shape', 'img_shape',
+            'pad_shape', 'scale_factor', 'flip', 'flip_direction',
+            'img_norm_cfg')``
+    """
+
+    def __init__(self,
+                 keys,
+                 meta_keys=('filename', 'ori_filename', 'ori_shape',
+                            'img_shape', 'pad_shape', 'scale_factor', 'flip',
+                            'flip_direction', 'img_norm_cfg')):
+        self.keys = keys
+        self.meta_keys = meta_keys
+
+    def __call__(self, results):
+        """Call function to collect keys in results. The keys in ``meta_keys``
+        will be converted to :obj:mmcv.DataContainer.
+
+        Args:
+            results (dict): Result dict contains the data to collect.
+
+        Returns:
+            dict: The result dict contains the following keys
+
+                - keys in``self.keys``
+                - ``img_metas``
+        """
+
+        data = {}
+        img_meta = {}
+        for key in self.meta_keys:
+            img_meta[key] = results[key]
+        data['img_metas'] = DC(img_meta, cpu_only=True)
+        for key in self.keys:
+            data[key] = results[key]
+        return data
+
+    def __repr__(self):
+        return self.__class__.__name__ + \
+            f'(keys={self.keys}, meta_keys={self.meta_keys})'
+
+
+@PIPELINES.register_module()
+class WrapFieldsToLists(object):
+    """Wrap fields of the data dictionary into lists for evaluation.
+
+    This class can be used as a last step of a test or validation
+    pipeline for single image evaluation or inference.
+
+    Example:
+        >>> test_pipeline = [
+        >>>    dict(type='LoadImageFromFile'),
+        >>>    dict(type='Normalize',
+                    mean=[123.675, 116.28, 103.53],
+                    std=[58.395, 57.12, 57.375],
+                    to_rgb=True),
+        >>>    dict(type='Pad', size_divisor=32),
+        >>>    dict(type='ImageToTensor', keys=['img']),
+        >>>    dict(type='Collect', keys=['img']),
+        >>>    dict(type='WrapFieldsToLists')
+        >>> ]
+    """
+
+    def __call__(self, results):
+        """Call function to wrap fields into lists.
+
+        Args:
+            results (dict): Result dict contains the data to wrap.
+
+        Returns:
+            dict: The result dict where value of ``self.keys`` are wrapped \
+                into list.
+        """
+
+        # Wrap dict fields into lists
+        for key, val in results.items():
+            results[key] = [val]
+        return results
+
+    def __repr__(self):
+        return f'{self.__class__.__name__}()'
diff --git a/python/autoCut/models/swin/mmdet/datasets/pipelines/instaboost.py b/python/autoCut/models/swin/mmdet/datasets/pipelines/instaboost.py
new file mode 100644
index 0000000..38b6819
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/pipelines/instaboost.py
@@ -0,0 +1,98 @@
+import numpy as np
+
+from ..builder import PIPELINES
+
+
+@PIPELINES.register_module()
+class InstaBoost(object):
+    r"""Data augmentation method in `InstaBoost: Boosting Instance
+    Segmentation Via Probability Map Guided Copy-Pasting
+    <https://arxiv.org/abs/1908.07801>`_.
+
+    Refer to https://github.com/GothicAi/Instaboost for implementation details.
+    """
+
+    def __init__(self,
+                 action_candidate=('normal', 'horizontal', 'skip'),
+                 action_prob=(1, 0, 0),
+                 scale=(0.8, 1.2),
+                 dx=15,
+                 dy=15,
+                 theta=(-1, 1),
+                 color_prob=0.5,
+                 hflag=False,
+                 aug_ratio=0.5):
+        try:
+            import instaboostfast as instaboost
+        except ImportError:
+            raise ImportError(
+                'Please run "pip install instaboostfast" '
+                'to install instaboostfast first for instaboost augmentation.')
+        self.cfg = instaboost.InstaBoostConfig(action_candidate, action_prob,
+                                               scale, dx, dy, theta,
+                                               color_prob, hflag)
+        self.aug_ratio = aug_ratio
+
+    def _load_anns(self, results):
+        labels = results['ann_info']['labels']
+        masks = results['ann_info']['masks']
+        bboxes = results['ann_info']['bboxes']
+        n = len(labels)
+
+        anns = []
+        for i in range(n):
+            label = labels[i]
+            bbox = bboxes[i]
+            mask = masks[i]
+            x1, y1, x2, y2 = bbox
+            # assert (x2 - x1) >= 1 and (y2 - y1) >= 1
+            bbox = [x1, y1, x2 - x1, y2 - y1]
+            anns.append({
+                'category_id': label,
+                'segmentation': mask,
+                'bbox': bbox
+            })
+
+        return anns
+
+    def _parse_anns(self, results, anns, img):
+        gt_bboxes = []
+        gt_labels = []
+        gt_masks_ann = []
+        for ann in anns:
+            x1, y1, w, h = ann['bbox']
+            # TODO: more essential bug need to be fixed in instaboost
+            if w <= 0 or h <= 0:
+                continue
+            bbox = [x1, y1, x1 + w, y1 + h]
+            gt_bboxes.append(bbox)
+            gt_labels.append(ann['category_id'])
+            gt_masks_ann.append(ann['segmentation'])
+        gt_bboxes = np.array(gt_bboxes, dtype=np.float32)
+        gt_labels = np.array(gt_labels, dtype=np.int64)
+        results['ann_info']['labels'] = gt_labels
+        results['ann_info']['bboxes'] = gt_bboxes
+        results['ann_info']['masks'] = gt_masks_ann
+        results['img'] = img
+        return results
+
+    def __call__(self, results):
+        img = results['img']
+        orig_type = img.dtype
+        anns = self._load_anns(results)
+        if np.random.choice([0, 1], p=[1 - self.aug_ratio, self.aug_ratio]):
+            try:
+                import instaboostfast as instaboost
+            except ImportError:
+                raise ImportError('Please run "pip install instaboostfast" '
+                                  'to install instaboostfast first.')
+            anns, img = instaboost.get_new_data(
+                anns, img.astype(np.uint8), self.cfg, background=None)
+
+        results = self._parse_anns(results, anns, img.astype(orig_type))
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(cfg={self.cfg}, aug_ratio={self.aug_ratio})'
+        return repr_str
diff --git a/python/autoCut/models/swin/mmdet/datasets/pipelines/loading.py b/python/autoCut/models/swin/mmdet/datasets/pipelines/loading.py
new file mode 100644
index 0000000..7d70c32
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/pipelines/loading.py
@@ -0,0 +1,521 @@
+import os.path as osp
+
+import mmcv
+import numpy as np
+import pycocotools.mask as maskUtils
+
+from mmdet.core import BitmapMasks, PolygonMasks
+from ..builder import PIPELINES
+
+
+@PIPELINES.register_module()
+class LoadImageFromFile(object):
+    """Load an image from file.
+
+    Required keys are "img_prefix" and "img_info" (a dict that must contain the
+    key "filename"). Added or updated keys are "filename", "img", "img_shape",
+    "ori_shape" (same as `img_shape`), "pad_shape" (same as `img_shape`),
+    "scale_factor" (1.0) and "img_norm_cfg" (means=0 and stds=1).
+
+    Args:
+        to_float32 (bool): Whether to convert the loaded image to a float32
+            numpy array. If set to False, the loaded image is an uint8 array.
+            Defaults to False.
+        color_type (str): The flag argument for :func:`mmcv.imfrombytes`.
+            Defaults to 'color'.
+        file_client_args (dict): Arguments to instantiate a FileClient.
+            See :class:`mmcv.fileio.FileClient` for details.
+            Defaults to ``dict(backend='disk')``.
+    """
+
+    def __init__(self,
+                 to_float32=False,
+                 color_type='color',
+                 file_client_args=dict(backend='disk')):
+        self.to_float32 = to_float32
+        self.color_type = color_type
+        self.file_client_args = file_client_args.copy()
+        self.file_client = None
+
+    def __call__(self, results):
+        """Call functions to load image and get image meta information.
+
+        Args:
+            results (dict): Result dict from :obj:`mmdet.CustomDataset`.
+
+        Returns:
+            dict: The dict contains loaded image and meta information.
+        """
+
+        if self.file_client is None:
+            self.file_client = mmcv.FileClient(**self.file_client_args)
+
+        if results['img_prefix'] is not None:
+            filename = osp.join(results['img_prefix'],
+                                results['img_info']['filename'])
+        else:
+            filename = results['img_info']['filename']
+
+        img_bytes = self.file_client.get(filename)
+        img = mmcv.imfrombytes(img_bytes, flag=self.color_type)
+        if self.to_float32:
+            img = img.astype(np.float32)
+
+        results['filename'] = filename
+        results['ori_filename'] = results['img_info']['filename']
+        results['img'] = img
+        results['img_shape'] = img.shape
+        results['ori_shape'] = img.shape
+        results['img_fields'] = ['img']
+        return results
+
+    def __repr__(self):
+        repr_str = (f'{self.__class__.__name__}('
+                    f'to_float32={self.to_float32}, '
+                    f"color_type='{self.color_type}', "
+                    f'file_client_args={self.file_client_args})')
+        return repr_str
+
+
+@PIPELINES.register_module()
+class LoadImageFromWebcam(LoadImageFromFile):
+    """Load an image from webcam.
+
+    Similar with :obj:`LoadImageFromFile`, but the image read from webcam is in
+    ``results['img']``.
+    """
+
+    def __call__(self, results):
+        """Call functions to add image meta information.
+
+        Args:
+            results (dict): Result dict with Webcam read image in
+                ``results['img']``.
+
+        Returns:
+            dict: The dict contains loaded image and meta information.
+        """
+
+        img = results['img']
+        if self.to_float32:
+            img = img.astype(np.float32)
+
+        results['filename'] = None
+        results['ori_filename'] = None
+        results['img'] = img
+        results['img_shape'] = img.shape
+        results['ori_shape'] = img.shape
+        results['img_fields'] = ['img']
+        return results
+
+
+@PIPELINES.register_module()
+class LoadMultiChannelImageFromFiles(object):
+    """Load multi-channel images from a list of separate channel files.
+
+    Required keys are "img_prefix" and "img_info" (a dict that must contain the
+    key "filename", which is expected to be a list of filenames).
+    Added or updated keys are "filename", "img", "img_shape",
+    "ori_shape" (same as `img_shape`), "pad_shape" (same as `img_shape`),
+    "scale_factor" (1.0) and "img_norm_cfg" (means=0 and stds=1).
+
+    Args:
+        to_float32 (bool): Whether to convert the loaded image to a float32
+            numpy array. If set to False, the loaded image is an uint8 array.
+            Defaults to False.
+        color_type (str): The flag argument for :func:`mmcv.imfrombytes`.
+            Defaults to 'color'.
+        file_client_args (dict): Arguments to instantiate a FileClient.
+            See :class:`mmcv.fileio.FileClient` for details.
+            Defaults to ``dict(backend='disk')``.
+    """
+
+    def __init__(self,
+                 to_float32=False,
+                 color_type='unchanged',
+                 file_client_args=dict(backend='disk')):
+        self.to_float32 = to_float32
+        self.color_type = color_type
+        self.file_client_args = file_client_args.copy()
+        self.file_client = None
+
+    def __call__(self, results):
+        """Call functions to load multiple images and get images meta
+        information.
+
+        Args:
+            results (dict): Result dict from :obj:`mmdet.CustomDataset`.
+
+        Returns:
+            dict: The dict contains loaded images and meta information.
+        """
+
+        if self.file_client is None:
+            self.file_client = mmcv.FileClient(**self.file_client_args)
+
+        if results['img_prefix'] is not None:
+            filename = [
+                osp.join(results['img_prefix'], fname)
+                for fname in results['img_info']['filename']
+            ]
+        else:
+            filename = results['img_info']['filename']
+
+        img = []
+        for name in filename:
+            img_bytes = self.file_client.get(name)
+            img.append(mmcv.imfrombytes(img_bytes, flag=self.color_type))
+        img = np.stack(img, axis=-1)
+        if self.to_float32:
+            img = img.astype(np.float32)
+
+        results['filename'] = filename
+        results['ori_filename'] = results['img_info']['filename']
+        results['img'] = img
+        results['img_shape'] = img.shape
+        results['ori_shape'] = img.shape
+        # Set initial values for default meta_keys
+        results['pad_shape'] = img.shape
+        results['scale_factor'] = 1.0
+        num_channels = 1 if len(img.shape) < 3 else img.shape[2]
+        results['img_norm_cfg'] = dict(
+            mean=np.zeros(num_channels, dtype=np.float32),
+            std=np.ones(num_channels, dtype=np.float32),
+            to_rgb=False)
+        return results
+
+    def __repr__(self):
+        repr_str = (f'{self.__class__.__name__}('
+                    f'to_float32={self.to_float32}, '
+                    f"color_type='{self.color_type}', "
+                    f'file_client_args={self.file_client_args})')
+        return repr_str
+
+
+@PIPELINES.register_module()
+class LoadAnnotations(object):
+    """Load mutiple types of annotations.
+
+    Args:
+        with_bbox (bool): Whether to parse and load the bbox annotation.
+             Default: True.
+        with_label (bool): Whether to parse and load the label annotation.
+            Default: True.
+        with_mask (bool): Whether to parse and load the mask annotation.
+             Default: False.
+        with_seg (bool): Whether to parse and load the semantic segmentation
+            annotation. Default: False.
+        poly2mask (bool): Whether to convert the instance masks from polygons
+            to bitmaps. Default: True.
+        file_client_args (dict): Arguments to instantiate a FileClient.
+            See :class:`mmcv.fileio.FileClient` for details.
+            Defaults to ``dict(backend='disk')``.
+    """
+
+    def __init__(self,
+                 with_bbox=True,
+                 with_label=True,
+                 with_mask=False,
+                 with_seg=False,
+                 poly2mask=True,
+                 file_client_args=dict(backend='disk')):
+        self.with_bbox = with_bbox
+        self.with_label = with_label
+        self.with_mask = with_mask
+        self.with_seg = with_seg
+        self.poly2mask = poly2mask
+        self.file_client_args = file_client_args.copy()
+        self.file_client = None
+
+    def _load_bboxes(self, results):
+        """Private function to load bounding box annotations.
+
+        Args:
+            results (dict): Result dict from :obj:`mmdet.CustomDataset`.
+
+        Returns:
+            dict: The dict contains loaded bounding box annotations.
+        """
+
+        ann_info = results['ann_info']
+        results['gt_bboxes'] = ann_info['bboxes'].copy()
+
+        gt_bboxes_ignore = ann_info.get('bboxes_ignore', None)
+        if gt_bboxes_ignore is not None:
+            results['gt_bboxes_ignore'] = gt_bboxes_ignore.copy()
+            results['bbox_fields'].append('gt_bboxes_ignore')
+        results['bbox_fields'].append('gt_bboxes')
+        return results
+
+    def _load_labels(self, results):
+        """Private function to load label annotations.
+
+        Args:
+            results (dict): Result dict from :obj:`mmdet.CustomDataset`.
+
+        Returns:
+            dict: The dict contains loaded label annotations.
+        """
+
+        results['gt_labels'] = results['ann_info']['labels'].copy()
+        return results
+
+    def _poly2mask(self, mask_ann, img_h, img_w):
+        """Private function to convert masks represented with polygon to
+        bitmaps.
+
+        Args:
+            mask_ann (list | dict): Polygon mask annotation input.
+            img_h (int): The height of output mask.
+            img_w (int): The width of output mask.
+
+        Returns:
+            numpy.ndarray: The decode bitmap mask of shape (img_h, img_w).
+        """
+
+        if isinstance(mask_ann, list):
+            # polygon -- a single object might consist of multiple parts
+            # we merge all parts into one mask rle code
+            rles = maskUtils.frPyObjects(mask_ann, img_h, img_w)
+            rle = maskUtils.merge(rles)
+        elif isinstance(mask_ann['counts'], list):
+            # uncompressed RLE
+            rle = maskUtils.frPyObjects(mask_ann, img_h, img_w)
+        else:
+            # rle
+            rle = mask_ann
+        mask = maskUtils.decode(rle)
+        return mask
+
+    def process_polygons(self, polygons):
+        """Convert polygons to list of ndarray and filter invalid polygons.
+
+        Args:
+            polygons (list[list]): Polygons of one instance.
+
+        Returns:
+            list[numpy.ndarray]: Processed polygons.
+        """
+
+        polygons = [np.array(p) for p in polygons]
+        valid_polygons = []
+        for polygon in polygons:
+            if len(polygon) % 2 == 0 and len(polygon) >= 6:
+                valid_polygons.append(polygon)
+        return valid_polygons
+
+    def _load_masks(self, results):
+        """Private function to load mask annotations.
+
+        Args:
+            results (dict): Result dict from :obj:`mmdet.CustomDataset`.
+
+        Returns:
+            dict: The dict contains loaded mask annotations.
+                If ``self.poly2mask`` is set ``True``, `gt_mask` will contain
+                :obj:`PolygonMasks`. Otherwise, :obj:`BitmapMasks` is used.
+        """
+
+        h, w = results['img_info']['height'], results['img_info']['width']
+        gt_masks = results['ann_info']['masks']
+        if self.poly2mask:
+            gt_masks = BitmapMasks(
+                [self._poly2mask(mask, h, w) for mask in gt_masks], h, w)
+        else:
+            gt_masks = PolygonMasks(
+                [self.process_polygons(polygons) for polygons in gt_masks], h,
+                w)
+        results['gt_masks'] = gt_masks
+        results['mask_fields'].append('gt_masks')
+        return results
+
+    def _load_semantic_seg(self, results):
+        """Private function to load semantic segmentation annotations.
+
+        Args:
+            results (dict): Result dict from :obj:`dataset`.
+
+        Returns:
+            dict: The dict contains loaded semantic segmentation annotations.
+        """
+
+        if self.file_client is None:
+            self.file_client = mmcv.FileClient(**self.file_client_args)
+
+        filename = osp.join(results['seg_prefix'],
+                            results['ann_info']['seg_map'])
+        img_bytes = self.file_client.get(filename)
+        results['gt_semantic_seg'] = mmcv.imfrombytes(
+            img_bytes, flag='unchanged').squeeze()
+        results['seg_fields'].append('gt_semantic_seg')
+        return results
+
+    def __call__(self, results):
+        """Call function to load multiple types annotations.
+
+        Args:
+            results (dict): Result dict from :obj:`mmdet.CustomDataset`.
+
+        Returns:
+            dict: The dict contains loaded bounding box, label, mask and
+                semantic segmentation annotations.
+        """
+
+        if self.with_bbox:
+            results = self._load_bboxes(results)
+            if results is None:
+                return None
+        if self.with_label:
+            results = self._load_labels(results)
+        if self.with_mask:
+            results = self._load_masks(results)
+        if self.with_seg:
+            results = self._load_semantic_seg(results)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(with_bbox={self.with_bbox}, '
+        repr_str += f'with_label={self.with_label}, '
+        repr_str += f'with_mask={self.with_mask}, '
+        repr_str += f'with_seg={self.with_seg}, '
+        repr_str += f'poly2mask={self.poly2mask}, '
+        repr_str += f'poly2mask={self.file_client_args})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class LoadProposals(object):
+    """Load proposal pipeline.
+
+    Required key is "proposals". Updated keys are "proposals", "bbox_fields".
+
+    Args:
+        num_max_proposals (int, optional): Maximum number of proposals to load.
+            If not specified, all proposals will be loaded.
+    """
+
+    def __init__(self, num_max_proposals=None):
+        self.num_max_proposals = num_max_proposals
+
+    def __call__(self, results):
+        """Call function to load proposals from file.
+
+        Args:
+            results (dict): Result dict from :obj:`mmdet.CustomDataset`.
+
+        Returns:
+            dict: The dict contains loaded proposal annotations.
+        """
+
+        proposals = results['proposals']
+        if proposals.shape[1] not in (4, 5):
+            raise AssertionError(
+                'proposals should have shapes (n, 4) or (n, 5), '
+                f'but found {proposals.shape}')
+        proposals = proposals[:, :4]
+
+        if self.num_max_proposals is not None:
+            proposals = proposals[:self.num_max_proposals]
+
+        if len(proposals) == 0:
+            proposals = np.array([[0, 0, 0, 0]], dtype=np.float32)
+        results['proposals'] = proposals
+        results['bbox_fields'].append('proposals')
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__ + \
+            f'(num_max_proposals={self.num_max_proposals})'
+
+
+@PIPELINES.register_module()
+class FilterAnnotations(object):
+    """Filter invalid annotations.
+
+    Args:
+        min_gt_bbox_wh (tuple[int]): Minimum width and height of ground truth
+            boxes.
+    """
+
+    def __init__(self, min_gt_bbox_wh):
+        # TODO: add more filter options
+        self.min_gt_bbox_wh = min_gt_bbox_wh
+
+    def __call__(self, results):
+        assert 'gt_bboxes' in results
+        gt_bboxes = results['gt_bboxes']
+        w = gt_bboxes[:, 2] - gt_bboxes[:, 0]
+        h = gt_bboxes[:, 3] - gt_bboxes[:, 1]
+        keep = (w > self.min_gt_bbox_wh[0]) & (h > self.min_gt_bbox_wh[1])
+        if not keep.any():
+            return None
+        else:
+            keys = ('gt_bboxes', 'gt_labels', 'gt_masks', 'gt_semantic_seg')
+            for key in keys:
+                if key in results:
+                    results[key] = results[key][keep]
+            return results
+
+
+@PIPELINES.register_module()
+class LoadRPDV2Annotations(object):
+    """Load mutiple types of annotations.
+
+    Args:
+        with_bbox (bool): Whether to parse and load the bbox annotation.
+             Default: True.
+        with_label (bool): Whether to parse and load the label annotation.
+            Default: True.
+        with_mask (bool): Whether to parse and load the mask annotation.
+             Default: False.
+        with_seg (bool): Whether to parse and load the semantic segmentation
+            annotation. Default: False.
+        poly2mask (bool): Whether to convert the instance masks from polygons
+            to bitmaps. Default: True.
+        file_client_args (dict): Arguments to instantiate a FileClient.
+            See :class:`mmcv.fileio.FileClient` for details.
+            Defaults to ``dict(backend='disk')``.
+    """
+    def __init__(self):
+        super(LoadRPDV2Annotations, self).__init__()
+
+    def _load_semantic_map_from_box(self, results):
+        gt_bboxes = results['gt_bboxes']
+        gt_labels = results['gt_labels']
+        pad_shape = results['pad_shape']
+        gt_areas = (gt_bboxes[:, 2] - gt_bboxes[:, 0]) * (gt_bboxes[:, 3] - gt_bboxes[:, 1])
+        gt_sem_map = np.zeros((80, int(pad_shape[0] / 8), int(pad_shape[1] / 8)), dtype=np.float32)
+        gt_sem_weights = np.zeros((80, int(pad_shape[0] / 8), int(pad_shape[1] / 8)), dtype=np.float32)
+
+        indexs = np.argsort(gt_areas)
+        for ind in indexs[::-1]:
+            box = gt_bboxes[ind]
+            box_mask = np.zeros((int(pad_shape[0] / 8), int(pad_shape[1] / 8)), dtype=np.int64)
+            box_mask[int(box[1] / 8):int(box[3] / 8) + 1, int(box[0] / 8):int(box[2] / 8) + 1] = 1
+            gt_sem_map[gt_labels[ind]][box_mask > 0] = 1
+            gt_sem_weights[gt_labels[ind]][box_mask > 0] = 1 / gt_areas[ind]
+
+        results['gt_sem_map'] = gt_sem_map
+        results['gt_sem_weights'] = gt_sem_weights
+
+        return results
+
+    def __call__(self, results):
+        """Call function to load multiple types annotations
+
+        Args:
+            results (dict): Result dict from :obj:`mmdet.CustomDataset`.
+
+        Returns:
+            dict: The dict contains loaded bounding box, label, mask and
+                semantic segmentation annotations.
+        """
+
+        results = self._load_semantic_map_from_box(results)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(with_bbox_semantic_map={True}, '
+        return repr_str
\ No newline at end of file
diff --git a/python/autoCut/models/swin/mmdet/datasets/pipelines/test_time_aug.py b/python/autoCut/models/swin/mmdet/datasets/pipelines/test_time_aug.py
new file mode 100644
index 0000000..b6226e0
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/pipelines/test_time_aug.py
@@ -0,0 +1,119 @@
+import warnings
+
+import mmcv
+
+from ..builder import PIPELINES
+from .compose import Compose
+
+
+@PIPELINES.register_module()
+class MultiScaleFlipAug(object):
+    """Test-time augmentation with multiple scales and flipping.
+
+    An example configuration is as followed:
+
+    .. code-block::
+
+        img_scale=[(1333, 400), (1333, 800)],
+        flip=True,
+        transforms=[
+            dict(type='Resize', keep_ratio=True),
+            dict(type='RandomFlip'),
+            dict(type='Normalize', **img_norm_cfg),
+            dict(type='Pad', size_divisor=32),
+            dict(type='ImageToTensor', keys=['img']),
+            dict(type='Collect', keys=['img']),
+        ]
+
+    After MultiScaleFLipAug with above configuration, the results are wrapped
+    into lists of the same length as followed:
+
+    .. code-block::
+
+        dict(
+            img=[...],
+            img_shape=[...],
+            scale=[(1333, 400), (1333, 400), (1333, 800), (1333, 800)]
+            flip=[False, True, False, True]
+            ...
+        )
+
+    Args:
+        transforms (list[dict]): Transforms to apply in each augmentation.
+        img_scale (tuple | list[tuple] | None): Images scales for resizing.
+        scale_factor (float | list[float] | None): Scale factors for resizing.
+        flip (bool): Whether apply flip augmentation. Default: False.
+        flip_direction (str | list[str]): Flip augmentation directions,
+            options are "horizontal" and "vertical". If flip_direction is list,
+            multiple flip augmentations will be applied.
+            It has no effect when flip == False. Default: "horizontal".
+    """
+
+    def __init__(self,
+                 transforms,
+                 img_scale=None,
+                 scale_factor=None,
+                 flip=False,
+                 flip_direction='horizontal'):
+        self.transforms = Compose(transforms)
+        assert (img_scale is None) ^ (scale_factor is None), (
+            'Must have but only one variable can be setted')
+        if img_scale is not None:
+            self.img_scale = img_scale if isinstance(img_scale,
+                                                     list) else [img_scale]
+            self.scale_key = 'scale'
+            assert mmcv.is_list_of(self.img_scale, tuple)
+        else:
+            self.img_scale = scale_factor if isinstance(
+                scale_factor, list) else [scale_factor]
+            self.scale_key = 'scale_factor'
+
+        self.flip = flip
+        self.flip_direction = flip_direction if isinstance(
+            flip_direction, list) else [flip_direction]
+        assert mmcv.is_list_of(self.flip_direction, str)
+        if not self.flip and self.flip_direction != ['horizontal']:
+            warnings.warn(
+                'flip_direction has no effect when flip is set to False')
+        if (self.flip
+                and not any([t['type'] == 'RandomFlip' for t in transforms])):
+            warnings.warn(
+                'flip has no effect when RandomFlip is not in transforms')
+
+    def __call__(self, results):
+        """Call function to apply test time augment transforms on results.
+
+        Args:
+            results (dict): Result dict contains the data to transform.
+
+        Returns:
+           dict[str: list]: The augmented data, where each value is wrapped
+               into a list.
+        """
+
+        aug_data = []
+        flip_args = [(False, None)]
+        if self.flip:
+            flip_args += [(True, direction)
+                          for direction in self.flip_direction]
+        for scale in self.img_scale:
+            for flip, direction in flip_args:
+                _results = results.copy()
+                _results[self.scale_key] = scale
+                _results['flip'] = flip
+                _results['flip_direction'] = direction
+                data = self.transforms(_results)
+                aug_data.append(data)
+        # list of dict to dict of list
+        aug_data_dict = {key: [] for key in aug_data[0]}
+        for data in aug_data:
+            for key, val in data.items():
+                aug_data_dict[key].append(val)
+        return aug_data_dict
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(transforms={self.transforms}, '
+        repr_str += f'img_scale={self.img_scale}, flip={self.flip}, '
+        repr_str += f'flip_direction={self.flip_direction})'
+        return repr_str
diff --git a/python/autoCut/models/swin/mmdet/datasets/pipelines/transforms.py b/python/autoCut/models/swin/mmdet/datasets/pipelines/transforms.py
new file mode 100644
index 0000000..caed51d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/pipelines/transforms.py
@@ -0,0 +1,1811 @@
+import copy
+import inspect
+
+import mmcv
+import numpy as np
+from numpy import random
+
+from mmdet.core import PolygonMasks
+from mmdet.core.evaluation.bbox_overlaps import bbox_overlaps
+from ..builder import PIPELINES
+
+try:
+    from imagecorruptions import corrupt
+except ImportError:
+    corrupt = None
+
+try:
+    import albumentations
+    from albumentations import Compose
+except ImportError:
+    albumentations = None
+    Compose = None
+
+
+@PIPELINES.register_module()
+class Resize(object):
+    """Resize images & bbox & mask.
+
+    This transform resizes the input image to some scale. Bboxes and masks are
+    then resized with the same scale factor. If the input dict contains the key
+    "scale", then the scale in the input dict is used, otherwise the specified
+    scale in the init method is used. If the input dict contains the key
+    "scale_factor" (if MultiScaleFlipAug does not give img_scale but
+    scale_factor), the actual scale will be computed by image shape and
+    scale_factor.
+
+    `img_scale` can either be a tuple (single-scale) or a list of tuple
+    (multi-scale). There are 3 multiscale modes:
+
+    - ``ratio_range is not None``: randomly sample a ratio from the ratio \
+      range and multiply it with the image scale.
+    - ``ratio_range is None`` and ``multiscale_mode == "range"``: randomly \
+      sample a scale from the multiscale range.
+    - ``ratio_range is None`` and ``multiscale_mode == "value"``: randomly \
+      sample a scale from multiple scales.
+
+    Args:
+        img_scale (tuple or list[tuple]): Images scales for resizing.
+        multiscale_mode (str): Either "range" or "value".
+        ratio_range (tuple[float]): (min_ratio, max_ratio)
+        keep_ratio (bool): Whether to keep the aspect ratio when resizing the
+            image.
+        bbox_clip_border (bool, optional): Whether clip the objects outside
+            the border of the image. Defaults to True.
+        backend (str): Image resize backend, choices are 'cv2' and 'pillow'.
+            These two backends generates slightly different results. Defaults
+            to 'cv2'.
+        override (bool, optional): Whether to override `scale` and
+            `scale_factor` so as to call resize twice. Default False. If True,
+            after the first resizing, the existed `scale` and `scale_factor`
+            will be ignored so the second resizing can be allowed.
+            This option is a work-around for multiple times of resize in DETR.
+            Defaults to False.
+    """
+
+    def __init__(self,
+                 img_scale=None,
+                 multiscale_mode='range',
+                 ratio_range=None,
+                 keep_ratio=True,
+                 bbox_clip_border=True,
+                 backend='cv2',
+                 override=False):
+        if img_scale is None:
+            self.img_scale = None
+        else:
+            if isinstance(img_scale, list):
+                self.img_scale = img_scale
+            else:
+                self.img_scale = [img_scale]
+            assert mmcv.is_list_of(self.img_scale, tuple)
+
+        if ratio_range is not None:
+            # mode 1: given a scale and a range of image ratio
+            assert len(self.img_scale) == 1
+        else:
+            # mode 2: given multiple scales or a range of scales
+            assert multiscale_mode in ['value', 'range']
+
+        self.backend = backend
+        self.multiscale_mode = multiscale_mode
+        self.ratio_range = ratio_range
+        self.keep_ratio = keep_ratio
+        # TODO: refactor the override option in Resize
+        self.override = override
+        self.bbox_clip_border = bbox_clip_border
+
+    @staticmethod
+    def random_select(img_scales):
+        """Randomly select an img_scale from given candidates.
+
+        Args:
+            img_scales (list[tuple]): Images scales for selection.
+
+        Returns:
+            (tuple, int): Returns a tuple ``(img_scale, scale_dix)``, \
+                where ``img_scale`` is the selected image scale and \
+                ``scale_idx`` is the selected index in the given candidates.
+        """
+
+        assert mmcv.is_list_of(img_scales, tuple)
+        scale_idx = np.random.randint(len(img_scales))
+        img_scale = img_scales[scale_idx]
+        return img_scale, scale_idx
+
+    @staticmethod
+    def random_sample(img_scales):
+        """Randomly sample an img_scale when ``multiscale_mode=='range'``.
+
+        Args:
+            img_scales (list[tuple]): Images scale range for sampling.
+                There must be two tuples in img_scales, which specify the lower
+                and upper bound of image scales.
+
+        Returns:
+            (tuple, None): Returns a tuple ``(img_scale, None)``, where \
+                ``img_scale`` is sampled scale and None is just a placeholder \
+                to be consistent with :func:`random_select`.
+        """
+
+        assert mmcv.is_list_of(img_scales, tuple) and len(img_scales) == 2
+        img_scale_long = [max(s) for s in img_scales]
+        img_scale_short = [min(s) for s in img_scales]
+        long_edge = np.random.randint(
+            min(img_scale_long),
+            max(img_scale_long) + 1)
+        short_edge = np.random.randint(
+            min(img_scale_short),
+            max(img_scale_short) + 1)
+        img_scale = (long_edge, short_edge)
+        return img_scale, None
+
+    @staticmethod
+    def random_sample_ratio(img_scale, ratio_range):
+        """Randomly sample an img_scale when ``ratio_range`` is specified.
+
+        A ratio will be randomly sampled from the range specified by
+        ``ratio_range``. Then it would be multiplied with ``img_scale`` to
+        generate sampled scale.
+
+        Args:
+            img_scale (tuple): Images scale base to multiply with ratio.
+            ratio_range (tuple[float]): The minimum and maximum ratio to scale
+                the ``img_scale``.
+
+        Returns:
+            (tuple, None): Returns a tuple ``(scale, None)``, where \
+                ``scale`` is sampled ratio multiplied with ``img_scale`` and \
+                None is just a placeholder to be consistent with \
+                :func:`random_select`.
+        """
+
+        assert isinstance(img_scale, tuple) and len(img_scale) == 2
+        min_ratio, max_ratio = ratio_range
+        assert min_ratio <= max_ratio
+        ratio = np.random.random_sample() * (max_ratio - min_ratio) + min_ratio
+        scale = int(img_scale[0] * ratio), int(img_scale[1] * ratio)
+        return scale, None
+
+    def _random_scale(self, results):
+        """Randomly sample an img_scale according to ``ratio_range`` and
+        ``multiscale_mode``.
+
+        If ``ratio_range`` is specified, a ratio will be sampled and be
+        multiplied with ``img_scale``.
+        If multiple scales are specified by ``img_scale``, a scale will be
+        sampled according to ``multiscale_mode``.
+        Otherwise, single scale will be used.
+
+        Args:
+            results (dict): Result dict from :obj:`dataset`.
+
+        Returns:
+            dict: Two new keys 'scale` and 'scale_idx` are added into \
+                ``results``, which would be used by subsequent pipelines.
+        """
+
+        if self.ratio_range is not None:
+            scale, scale_idx = self.random_sample_ratio(
+                self.img_scale[0], self.ratio_range)
+        elif len(self.img_scale) == 1:
+            scale, scale_idx = self.img_scale[0], 0
+        elif self.multiscale_mode == 'range':
+            scale, scale_idx = self.random_sample(self.img_scale)
+        elif self.multiscale_mode == 'value':
+            scale, scale_idx = self.random_select(self.img_scale)
+        else:
+            raise NotImplementedError
+
+        results['scale'] = scale
+        results['scale_idx'] = scale_idx
+
+    def _resize_img(self, results):
+        """Resize images with ``results['scale']``."""
+        for key in results.get('img_fields', ['img']):
+            if self.keep_ratio:
+                img, scale_factor = mmcv.imrescale(
+                    results[key],
+                    results['scale'],
+                    return_scale=True,
+                    backend=self.backend)
+                # the w_scale and h_scale has minor difference
+                # a real fix should be done in the mmcv.imrescale in the future
+                new_h, new_w = img.shape[:2]
+                h, w = results[key].shape[:2]
+                w_scale = new_w / w
+                h_scale = new_h / h
+            else:
+                img, w_scale, h_scale = mmcv.imresize(
+                    results[key],
+                    results['scale'],
+                    return_scale=True,
+                    backend=self.backend)
+            results[key] = img
+
+            scale_factor = np.array([w_scale, h_scale, w_scale, h_scale],
+                                    dtype=np.float32)
+            results['img_shape'] = img.shape
+            # in case that there is no padding
+            results['pad_shape'] = img.shape
+            results['scale_factor'] = scale_factor
+            results['keep_ratio'] = self.keep_ratio
+
+    def _resize_bboxes(self, results):
+        """Resize bounding boxes with ``results['scale_factor']``."""
+        for key in results.get('bbox_fields', []):
+            bboxes = results[key] * results['scale_factor']
+            if self.bbox_clip_border:
+                img_shape = results['img_shape']
+                bboxes[:, 0::2] = np.clip(bboxes[:, 0::2], 0, img_shape[1])
+                bboxes[:, 1::2] = np.clip(bboxes[:, 1::2], 0, img_shape[0])
+            results[key] = bboxes
+
+    def _resize_masks(self, results):
+        """Resize masks with ``results['scale']``"""
+        for key in results.get('mask_fields', []):
+            if results[key] is None:
+                continue
+            if self.keep_ratio:
+                results[key] = results[key].rescale(results['scale'])
+            else:
+                results[key] = results[key].resize(results['img_shape'][:2])
+
+    def _resize_seg(self, results):
+        """Resize semantic segmentation map with ``results['scale']``."""
+        for key in results.get('seg_fields', []):
+            if self.keep_ratio:
+                gt_seg = mmcv.imrescale(
+                    results[key],
+                    results['scale'],
+                    interpolation='nearest',
+                    backend=self.backend)
+            else:
+                gt_seg = mmcv.imresize(
+                    results[key],
+                    results['scale'],
+                    interpolation='nearest',
+                    backend=self.backend)
+            results['gt_semantic_seg'] = gt_seg
+
+    def __call__(self, results):
+        """Call function to resize images, bounding boxes, masks, semantic
+        segmentation map.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Resized results, 'img_shape', 'pad_shape', 'scale_factor', \
+                'keep_ratio' keys are added into result dict.
+        """
+
+        if 'scale' not in results:
+            if 'scale_factor' in results:
+                img_shape = results['img'].shape[:2]
+                scale_factor = results['scale_factor']
+                assert isinstance(scale_factor, float)
+                results['scale'] = tuple(
+                    [int(x * scale_factor) for x in img_shape][::-1])
+            else:
+                self._random_scale(results)
+        else:
+            if not self.override:
+                assert 'scale_factor' not in results, (
+                    'scale and scale_factor cannot be both set.')
+            else:
+                results.pop('scale')
+                if 'scale_factor' in results:
+                    results.pop('scale_factor')
+                self._random_scale(results)
+
+        self._resize_img(results)
+        self._resize_bboxes(results)
+        self._resize_masks(results)
+        self._resize_seg(results)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(img_scale={self.img_scale}, '
+        repr_str += f'multiscale_mode={self.multiscale_mode}, '
+        repr_str += f'ratio_range={self.ratio_range}, '
+        repr_str += f'keep_ratio={self.keep_ratio}, '
+        repr_str += f'bbox_clip_border={self.bbox_clip_border})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class RandomFlip(object):
+    """Flip the image & bbox & mask.
+
+    If the input dict contains the key "flip", then the flag will be used,
+    otherwise it will be randomly decided by a ratio specified in the init
+    method.
+
+    When random flip is enabled, ``flip_ratio``/``direction`` can either be a
+    float/string or tuple of float/string. There are 3 flip modes:
+
+    - ``flip_ratio`` is float, ``direction`` is string: the image will be
+        ``direction``ly flipped with probability of ``flip_ratio`` .
+        E.g., ``flip_ratio=0.5``, ``direction='horizontal'``,
+        then image will be horizontally flipped with probability of 0.5.
+    - ``flip_ratio`` is float, ``direction`` is list of string: the image wil
+        be ``direction[i]``ly flipped with probability of
+        ``flip_ratio/len(direction)``.
+        E.g., ``flip_ratio=0.5``, ``direction=['horizontal', 'vertical']``,
+        then image will be horizontally flipped with probability of 0.25,
+        vertically with probability of 0.25.
+    - ``flip_ratio`` is list of float, ``direction`` is list of string:
+        given ``len(flip_ratio) == len(direction)``, the image wil
+        be ``direction[i]``ly flipped with probability of ``flip_ratio[i]``.
+        E.g., ``flip_ratio=[0.3, 0.5]``, ``direction=['horizontal',
+        'vertical']``, then image will be horizontally flipped with probability
+         of 0.3, vertically with probability of 0.5
+
+    Args:
+        flip_ratio (float | list[float], optional): The flipping probability.
+            Default: None.
+        direction(str | list[str], optional): The flipping direction. Options
+            are 'horizontal', 'vertical', 'diagonal'. Default: 'horizontal'.
+            If input is a list, the length must equal ``flip_ratio``. Each
+            element in ``flip_ratio`` indicates the flip probability of
+            corresponding direction.
+    """
+
+    def __init__(self, flip_ratio=None, direction='horizontal'):
+        if isinstance(flip_ratio, list):
+            assert mmcv.is_list_of(flip_ratio, float)
+            assert 0 <= sum(flip_ratio) <= 1
+        elif isinstance(flip_ratio, float):
+            assert 0 <= flip_ratio <= 1
+        elif flip_ratio is None:
+            pass
+        else:
+            raise ValueError('flip_ratios must be None, float, '
+                             'or list of float')
+        self.flip_ratio = flip_ratio
+
+        valid_directions = ['horizontal', 'vertical', 'diagonal']
+        if isinstance(direction, str):
+            assert direction in valid_directions
+        elif isinstance(direction, list):
+            assert mmcv.is_list_of(direction, str)
+            assert set(direction).issubset(set(valid_directions))
+        else:
+            raise ValueError('direction must be either str or list of str')
+        self.direction = direction
+
+        if isinstance(flip_ratio, list):
+            assert len(self.flip_ratio) == len(self.direction)
+
+    def bbox_flip(self, bboxes, img_shape, direction):
+        """Flip bboxes horizontally.
+
+        Args:
+            bboxes (numpy.ndarray): Bounding boxes, shape (..., 4*k)
+            img_shape (tuple[int]): Image shape (height, width)
+            direction (str): Flip direction. Options are 'horizontal',
+                'vertical'.
+
+        Returns:
+            numpy.ndarray: Flipped bounding boxes.
+        """
+
+        assert bboxes.shape[-1] % 4 == 0
+        flipped = bboxes.copy()
+        if direction == 'horizontal':
+            w = img_shape[1]
+            flipped[..., 0::4] = w - bboxes[..., 2::4]
+            flipped[..., 2::4] = w - bboxes[..., 0::4]
+        elif direction == 'vertical':
+            h = img_shape[0]
+            flipped[..., 1::4] = h - bboxes[..., 3::4]
+            flipped[..., 3::4] = h - bboxes[..., 1::4]
+        elif direction == 'diagonal':
+            w = img_shape[1]
+            h = img_shape[0]
+            flipped[..., 0::4] = w - bboxes[..., 2::4]
+            flipped[..., 1::4] = h - bboxes[..., 3::4]
+            flipped[..., 2::4] = w - bboxes[..., 0::4]
+            flipped[..., 3::4] = h - bboxes[..., 1::4]
+        else:
+            raise ValueError(f"Invalid flipping direction '{direction}'")
+        return flipped
+
+    def __call__(self, results):
+        """Call function to flip bounding boxes, masks, semantic segmentation
+        maps.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Flipped results, 'flip', 'flip_direction' keys are added \
+                into result dict.
+        """
+
+        if 'flip' not in results:
+            if isinstance(self.direction, list):
+                # None means non-flip
+                direction_list = self.direction + [None]
+            else:
+                # None means non-flip
+                direction_list = [self.direction, None]
+
+            if isinstance(self.flip_ratio, list):
+                non_flip_ratio = 1 - sum(self.flip_ratio)
+                flip_ratio_list = self.flip_ratio + [non_flip_ratio]
+            else:
+                non_flip_ratio = 1 - self.flip_ratio
+                # exclude non-flip
+                single_ratio = self.flip_ratio / (len(direction_list) - 1)
+                flip_ratio_list = [single_ratio] * (len(direction_list) -
+                                                    1) + [non_flip_ratio]
+
+            cur_dir = np.random.choice(direction_list, p=flip_ratio_list)
+
+            results['flip'] = cur_dir is not None
+        if 'flip_direction' not in results:
+            results['flip_direction'] = cur_dir
+        if results['flip']:
+            # flip image
+            for key in results.get('img_fields', ['img']):
+                results[key] = mmcv.imflip(
+                    results[key], direction=results['flip_direction'])
+            # flip bboxes
+            for key in results.get('bbox_fields', []):
+                results[key] = self.bbox_flip(results[key],
+                                              results['img_shape'],
+                                              results['flip_direction'])
+            # flip masks
+            for key in results.get('mask_fields', []):
+                results[key] = results[key].flip(results['flip_direction'])
+
+            # flip segs
+            for key in results.get('seg_fields', []):
+                results[key] = mmcv.imflip(
+                    results[key], direction=results['flip_direction'])
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__ + f'(flip_ratio={self.flip_ratio})'
+
+
+@PIPELINES.register_module()
+class Pad(object):
+    """Pad the image & mask.
+
+    There are two padding modes: (1) pad to a fixed size and (2) pad to the
+    minimum size that is divisible by some number.
+    Added keys are "pad_shape", "pad_fixed_size", "pad_size_divisor",
+
+    Args:
+        size (tuple, optional): Fixed padding size.
+        size_divisor (int, optional): The divisor of padded size.
+        pad_val (float, optional): Padding value, 0 by default.
+    """
+
+    def __init__(self, size=None, size_divisor=None, pad_val=0):
+        self.size = size
+        self.size_divisor = size_divisor
+        self.pad_val = pad_val
+        # only one of size and size_divisor should be valid
+        assert size is not None or size_divisor is not None
+        assert size is None or size_divisor is None
+
+    def _pad_img(self, results):
+        """Pad images according to ``self.size``."""
+        for key in results.get('img_fields', ['img']):
+            if self.size is not None:
+                padded_img = mmcv.impad(
+                    results[key], shape=self.size, pad_val=self.pad_val)
+            elif self.size_divisor is not None:
+                padded_img = mmcv.impad_to_multiple(
+                    results[key], self.size_divisor, pad_val=self.pad_val)
+            results[key] = padded_img
+        results['pad_shape'] = padded_img.shape
+        results['pad_fixed_size'] = self.size
+        results['pad_size_divisor'] = self.size_divisor
+
+    def _pad_masks(self, results):
+        """Pad masks according to ``results['pad_shape']``."""
+        pad_shape = results['pad_shape'][:2]
+        for key in results.get('mask_fields', []):
+            results[key] = results[key].pad(pad_shape, pad_val=self.pad_val)
+
+    def _pad_seg(self, results):
+        """Pad semantic segmentation map according to
+        ``results['pad_shape']``."""
+        for key in results.get('seg_fields', []):
+            results[key] = mmcv.impad(
+                results[key], shape=results['pad_shape'][:2])
+
+    def __call__(self, results):
+        """Call function to pad images, masks, semantic segmentation maps.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Updated result dict.
+        """
+        self._pad_img(results)
+        self._pad_masks(results)
+        self._pad_seg(results)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(size={self.size}, '
+        repr_str += f'size_divisor={self.size_divisor}, '
+        repr_str += f'pad_val={self.pad_val})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class Normalize(object):
+    """Normalize the image.
+
+    Added key is "img_norm_cfg".
+
+    Args:
+        mean (sequence): Mean values of 3 channels.
+        std (sequence): Std values of 3 channels.
+        to_rgb (bool): Whether to convert the image from BGR to RGB,
+            default is true.
+    """
+
+    def __init__(self, mean, std, to_rgb=True):
+        self.mean = np.array(mean, dtype=np.float32)
+        self.std = np.array(std, dtype=np.float32)
+        self.to_rgb = to_rgb
+
+    def __call__(self, results):
+        """Call function to normalize images.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Normalized results, 'img_norm_cfg' key is added into
+                result dict.
+        """
+        for key in results.get('img_fields', ['img']):
+            results[key] = mmcv.imnormalize(results[key], self.mean, self.std,
+                                            self.to_rgb)
+        results['img_norm_cfg'] = dict(
+            mean=self.mean, std=self.std, to_rgb=self.to_rgb)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(mean={self.mean}, std={self.std}, to_rgb={self.to_rgb})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class RandomCrop(object):
+    """Random crop the image & bboxes & masks.
+
+    The absolute `crop_size` is sampled based on `crop_type` and `image_size`,
+    then the cropped results are generated.
+
+    Args:
+        crop_size (tuple): The relative ratio or absolute pixels of
+            height and width.
+        crop_type (str, optional): one of "relative_range", "relative",
+            "absolute", "absolute_range". "relative" randomly crops
+            (h * crop_size[0], w * crop_size[1]) part from an input of size
+            (h, w). "relative_range" uniformly samples relative crop size from
+            range [crop_size[0], 1] and [crop_size[1], 1] for height and width
+            respectively. "absolute" crops from an input with absolute size
+            (crop_size[0], crop_size[1]). "absolute_range" uniformly samples
+            crop_h in range [crop_size[0], min(h, crop_size[1])] and crop_w
+            in range [crop_size[0], min(w, crop_size[1])]. Default "absolute".
+        allow_negative_crop (bool, optional): Whether to allow a crop that does
+            not contain any bbox area. Default False.
+        bbox_clip_border (bool, optional): Whether clip the objects outside
+            the border of the image. Defaults to True.
+
+    Note:
+        - If the image is smaller than the absolute crop size, return the
+            original image.
+        - The keys for bboxes, labels and masks must be aligned. That is,
+          `gt_bboxes` corresponds to `gt_labels` and `gt_masks`, and
+          `gt_bboxes_ignore` corresponds to `gt_labels_ignore` and
+          `gt_masks_ignore`.
+        - If the crop does not contain any gt-bbox region and
+          `allow_negative_crop` is set to False, skip this image.
+    """
+
+    def __init__(self,
+                 crop_size,
+                 crop_type='absolute',
+                 allow_negative_crop=False,
+                 bbox_clip_border=True):
+        if crop_type not in [
+                'relative_range', 'relative', 'absolute', 'absolute_range'
+        ]:
+            raise ValueError(f'Invalid crop_type {crop_type}.')
+        if crop_type in ['absolute', 'absolute_range']:
+            assert crop_size[0] > 0 and crop_size[1] > 0
+            assert isinstance(crop_size[0], int) and isinstance(
+                crop_size[1], int)
+        else:
+            assert 0 < crop_size[0] <= 1 and 0 < crop_size[1] <= 1
+        self.crop_size = crop_size
+        self.crop_type = crop_type
+        self.allow_negative_crop = allow_negative_crop
+        self.bbox_clip_border = bbox_clip_border
+        # The key correspondence from bboxes to labels and masks.
+        self.bbox2label = {
+            'gt_bboxes': 'gt_labels',
+            'gt_bboxes_ignore': 'gt_labels_ignore'
+        }
+        self.bbox2mask = {
+            'gt_bboxes': 'gt_masks',
+            'gt_bboxes_ignore': 'gt_masks_ignore'
+        }
+
+    def _crop_data(self, results, crop_size, allow_negative_crop):
+        """Function to randomly crop images, bounding boxes, masks, semantic
+        segmentation maps.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+            crop_size (tuple): Expected absolute size after cropping, (h, w).
+            allow_negative_crop (bool): Whether to allow a crop that does not
+                contain any bbox area. Default to False.
+
+        Returns:
+            dict: Randomly cropped results, 'img_shape' key in result dict is
+                updated according to crop size.
+        """
+        assert crop_size[0] > 0 and crop_size[1] > 0
+        for key in results.get('img_fields', ['img']):
+            img = results[key]
+            margin_h = max(img.shape[0] - crop_size[0], 0)
+            margin_w = max(img.shape[1] - crop_size[1], 0)
+            offset_h = np.random.randint(0, margin_h + 1)
+            offset_w = np.random.randint(0, margin_w + 1)
+            crop_y1, crop_y2 = offset_h, offset_h + crop_size[0]
+            crop_x1, crop_x2 = offset_w, offset_w + crop_size[1]
+
+            # crop the image
+            img = img[crop_y1:crop_y2, crop_x1:crop_x2, ...]
+            img_shape = img.shape
+            results[key] = img
+        results['img_shape'] = img_shape
+
+        # crop bboxes accordingly and clip to the image boundary
+        for key in results.get('bbox_fields', []):
+            # e.g. gt_bboxes and gt_bboxes_ignore
+            bbox_offset = np.array([offset_w, offset_h, offset_w, offset_h],
+                                   dtype=np.float32)
+            bboxes = results[key] - bbox_offset
+            if self.bbox_clip_border:
+                bboxes[:, 0::2] = np.clip(bboxes[:, 0::2], 0, img_shape[1])
+                bboxes[:, 1::2] = np.clip(bboxes[:, 1::2], 0, img_shape[0])
+            valid_inds = (bboxes[:, 2] > bboxes[:, 0]) & (
+                bboxes[:, 3] > bboxes[:, 1])
+            # If the crop does not contain any gt-bbox area and
+            # allow_negative_crop is False, skip this image.
+            if (key == 'gt_bboxes' and not valid_inds.any()
+                    and not allow_negative_crop):
+                return None
+            results[key] = bboxes[valid_inds, :]
+            # label fields. e.g. gt_labels and gt_labels_ignore
+            label_key = self.bbox2label.get(key)
+            if label_key in results:
+                results[label_key] = results[label_key][valid_inds]
+
+            # mask fields, e.g. gt_masks and gt_masks_ignore
+            mask_key = self.bbox2mask.get(key)
+            if mask_key in results:
+                results[mask_key] = results[mask_key][
+                    valid_inds.nonzero()[0]].crop(
+                        np.asarray([crop_x1, crop_y1, crop_x2, crop_y2]))
+
+        # crop semantic seg
+        for key in results.get('seg_fields', []):
+            results[key] = results[key][crop_y1:crop_y2, crop_x1:crop_x2]
+
+        return results
+
+    def _get_crop_size(self, image_size):
+        """Randomly generates the absolute crop size based on `crop_type` and
+        `image_size`.
+
+        Args:
+            image_size (tuple): (h, w).
+
+        Returns:
+            crop_size (tuple): (crop_h, crop_w) in absolute pixels.
+        """
+        h, w = image_size
+        if self.crop_type == 'absolute':
+            return (min(self.crop_size[0], h), min(self.crop_size[1], w))
+        elif self.crop_type == 'absolute_range':
+            assert self.crop_size[0] <= self.crop_size[1]
+            crop_h = np.random.randint(
+                min(h, self.crop_size[0]),
+                min(h, self.crop_size[1]) + 1)
+            crop_w = np.random.randint(
+                min(w, self.crop_size[0]),
+                min(w, self.crop_size[1]) + 1)
+            return crop_h, crop_w
+        elif self.crop_type == 'relative':
+            crop_h, crop_w = self.crop_size
+            return int(h * crop_h + 0.5), int(w * crop_w + 0.5)
+        elif self.crop_type == 'relative_range':
+            crop_size = np.asarray(self.crop_size, dtype=np.float32)
+            crop_h, crop_w = crop_size + np.random.rand(2) * (1 - crop_size)
+            return int(h * crop_h + 0.5), int(w * crop_w + 0.5)
+
+    def __call__(self, results):
+        """Call function to randomly crop images, bounding boxes, masks,
+        semantic segmentation maps.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Randomly cropped results, 'img_shape' key in result dict is
+                updated according to crop size.
+        """
+        image_size = results['img'].shape[:2]
+        crop_size = self._get_crop_size(image_size)
+        results = self._crop_data(results, crop_size, self.allow_negative_crop)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(crop_size={self.crop_size}, '
+        repr_str += f'crop_type={self.crop_type}, '
+        repr_str += f'allow_negative_crop={self.allow_negative_crop}, '
+        repr_str += f'bbox_clip_border={self.bbox_clip_border})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class SegRescale(object):
+    """Rescale semantic segmentation maps.
+
+    Args:
+        scale_factor (float): The scale factor of the final output.
+        backend (str): Image rescale backend, choices are 'cv2' and 'pillow'.
+            These two backends generates slightly different results. Defaults
+            to 'cv2'.
+    """
+
+    def __init__(self, scale_factor=1, backend='cv2'):
+        self.scale_factor = scale_factor
+        self.backend = backend
+
+    def __call__(self, results):
+        """Call function to scale the semantic segmentation map.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Result dict with semantic segmentation map scaled.
+        """
+
+        for key in results.get('seg_fields', []):
+            if self.scale_factor != 1:
+                results[key] = mmcv.imrescale(
+                    results[key],
+                    self.scale_factor,
+                    interpolation='nearest',
+                    backend=self.backend)
+        return results
+
+    def __repr__(self):
+        return self.__class__.__name__ + f'(scale_factor={self.scale_factor})'
+
+
+@PIPELINES.register_module()
+class PhotoMetricDistortion(object):
+    """Apply photometric distortion to image sequentially, every transformation
+    is applied with a probability of 0.5. The position of random contrast is in
+    second or second to last.
+
+    1. random brightness
+    2. random contrast (mode 0)
+    3. convert color from BGR to HSV
+    4. random saturation
+    5. random hue
+    6. convert color from HSV to BGR
+    7. random contrast (mode 1)
+    8. randomly swap channels
+
+    Args:
+        brightness_delta (int): delta of brightness.
+        contrast_range (tuple): range of contrast.
+        saturation_range (tuple): range of saturation.
+        hue_delta (int): delta of hue.
+    """
+
+    def __init__(self,
+                 brightness_delta=32,
+                 contrast_range=(0.5, 1.5),
+                 saturation_range=(0.5, 1.5),
+                 hue_delta=18):
+        self.brightness_delta = brightness_delta
+        self.contrast_lower, self.contrast_upper = contrast_range
+        self.saturation_lower, self.saturation_upper = saturation_range
+        self.hue_delta = hue_delta
+
+    def __call__(self, results):
+        """Call function to perform photometric distortion on images.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Result dict with images distorted.
+        """
+
+        if 'img_fields' in results:
+            assert results['img_fields'] == ['img'], \
+                'Only single img_fields is allowed'
+        img = results['img']
+        assert img.dtype == np.float32, \
+            'PhotoMetricDistortion needs the input image of dtype np.float32,'\
+            ' please set "to_float32=True" in "LoadImageFromFile" pipeline'
+        # random brightness
+        if random.randint(2):
+            delta = random.uniform(-self.brightness_delta,
+                                   self.brightness_delta)
+            img += delta
+
+        # mode == 0 --> do random contrast first
+        # mode == 1 --> do random contrast last
+        mode = random.randint(2)
+        if mode == 1:
+            if random.randint(2):
+                alpha = random.uniform(self.contrast_lower,
+                                       self.contrast_upper)
+                img *= alpha
+
+        # convert color from BGR to HSV
+        img = mmcv.bgr2hsv(img)
+
+        # random saturation
+        if random.randint(2):
+            img[..., 1] *= random.uniform(self.saturation_lower,
+                                          self.saturation_upper)
+
+        # random hue
+        if random.randint(2):
+            img[..., 0] += random.uniform(-self.hue_delta, self.hue_delta)
+            img[..., 0][img[..., 0] > 360] -= 360
+            img[..., 0][img[..., 0] < 0] += 360
+
+        # convert color from HSV to BGR
+        img = mmcv.hsv2bgr(img)
+
+        # random contrast
+        if mode == 0:
+            if random.randint(2):
+                alpha = random.uniform(self.contrast_lower,
+                                       self.contrast_upper)
+                img *= alpha
+
+        # randomly swap channels
+        if random.randint(2):
+            img = img[..., random.permutation(3)]
+
+        results['img'] = img
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(\nbrightness_delta={self.brightness_delta},\n'
+        repr_str += 'contrast_range='
+        repr_str += f'{(self.contrast_lower, self.contrast_upper)},\n'
+        repr_str += 'saturation_range='
+        repr_str += f'{(self.saturation_lower, self.saturation_upper)},\n'
+        repr_str += f'hue_delta={self.hue_delta})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class Expand(object):
+    """Random expand the image & bboxes.
+
+    Randomly place the original image on a canvas of 'ratio' x original image
+    size filled with mean values. The ratio is in the range of ratio_range.
+
+    Args:
+        mean (tuple): mean value of dataset.
+        to_rgb (bool): if need to convert the order of mean to align with RGB.
+        ratio_range (tuple): range of expand ratio.
+        prob (float): probability of applying this transformation
+    """
+
+    def __init__(self,
+                 mean=(0, 0, 0),
+                 to_rgb=True,
+                 ratio_range=(1, 4),
+                 seg_ignore_label=None,
+                 prob=0.5):
+        self.to_rgb = to_rgb
+        self.ratio_range = ratio_range
+        if to_rgb:
+            self.mean = mean[::-1]
+        else:
+            self.mean = mean
+        self.min_ratio, self.max_ratio = ratio_range
+        self.seg_ignore_label = seg_ignore_label
+        self.prob = prob
+
+    def __call__(self, results):
+        """Call function to expand images, bounding boxes.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Result dict with images, bounding boxes expanded
+        """
+
+        if random.uniform(0, 1) > self.prob:
+            return results
+
+        if 'img_fields' in results:
+            assert results['img_fields'] == ['img'], \
+                'Only single img_fields is allowed'
+        img = results['img']
+
+        h, w, c = img.shape
+        ratio = random.uniform(self.min_ratio, self.max_ratio)
+        # speedup expand when meets large image
+        if np.all(self.mean == self.mean[0]):
+            expand_img = np.empty((int(h * ratio), int(w * ratio), c),
+                                  img.dtype)
+            expand_img.fill(self.mean[0])
+        else:
+            expand_img = np.full((int(h * ratio), int(w * ratio), c),
+                                 self.mean,
+                                 dtype=img.dtype)
+        left = int(random.uniform(0, w * ratio - w))
+        top = int(random.uniform(0, h * ratio - h))
+        expand_img[top:top + h, left:left + w] = img
+
+        results['img'] = expand_img
+        # expand bboxes
+        for key in results.get('bbox_fields', []):
+            results[key] = results[key] + np.tile(
+                (left, top), 2).astype(results[key].dtype)
+
+        # expand masks
+        for key in results.get('mask_fields', []):
+            results[key] = results[key].expand(
+                int(h * ratio), int(w * ratio), top, left)
+
+        # expand segs
+        for key in results.get('seg_fields', []):
+            gt_seg = results[key]
+            expand_gt_seg = np.full((int(h * ratio), int(w * ratio)),
+                                    self.seg_ignore_label,
+                                    dtype=gt_seg.dtype)
+            expand_gt_seg[top:top + h, left:left + w] = gt_seg
+            results[key] = expand_gt_seg
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(mean={self.mean}, to_rgb={self.to_rgb}, '
+        repr_str += f'ratio_range={self.ratio_range}, '
+        repr_str += f'seg_ignore_label={self.seg_ignore_label})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class MinIoURandomCrop(object):
+    """Random crop the image & bboxes, the cropped patches have minimum IoU
+    requirement with original image & bboxes, the IoU threshold is randomly
+    selected from min_ious.
+
+    Args:
+        min_ious (tuple): minimum IoU threshold for all intersections with
+        bounding boxes
+        min_crop_size (float): minimum crop's size (i.e. h,w := a*h, a*w,
+        where a >= min_crop_size).
+        bbox_clip_border (bool, optional): Whether clip the objects outside
+            the border of the image. Defaults to True.
+
+    Note:
+        The keys for bboxes, labels and masks should be paired. That is, \
+        `gt_bboxes` corresponds to `gt_labels` and `gt_masks`, and \
+        `gt_bboxes_ignore` to `gt_labels_ignore` and `gt_masks_ignore`.
+    """
+
+    def __init__(self,
+                 min_ious=(0.1, 0.3, 0.5, 0.7, 0.9),
+                 min_crop_size=0.3,
+                 bbox_clip_border=True):
+        # 1: return ori img
+        self.min_ious = min_ious
+        self.sample_mode = (1, *min_ious, 0)
+        self.min_crop_size = min_crop_size
+        self.bbox_clip_border = bbox_clip_border
+        self.bbox2label = {
+            'gt_bboxes': 'gt_labels',
+            'gt_bboxes_ignore': 'gt_labels_ignore'
+        }
+        self.bbox2mask = {
+            'gt_bboxes': 'gt_masks',
+            'gt_bboxes_ignore': 'gt_masks_ignore'
+        }
+
+    def __call__(self, results):
+        """Call function to crop images and bounding boxes with minimum IoU
+        constraint.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Result dict with images and bounding boxes cropped, \
+                'img_shape' key is updated.
+        """
+
+        if 'img_fields' in results:
+            assert results['img_fields'] == ['img'], \
+                'Only single img_fields is allowed'
+        img = results['img']
+        assert 'bbox_fields' in results
+        boxes = [results[key] for key in results['bbox_fields']]
+        boxes = np.concatenate(boxes, 0)
+        h, w, c = img.shape
+        while True:
+            mode = random.choice(self.sample_mode)
+            self.mode = mode
+            if mode == 1:
+                return results
+
+            min_iou = mode
+            for i in range(50):
+                new_w = random.uniform(self.min_crop_size * w, w)
+                new_h = random.uniform(self.min_crop_size * h, h)
+
+                # h / w in [0.5, 2]
+                if new_h / new_w < 0.5 or new_h / new_w > 2:
+                    continue
+
+                left = random.uniform(w - new_w)
+                top = random.uniform(h - new_h)
+
+                patch = np.array(
+                    (int(left), int(top), int(left + new_w), int(top + new_h)))
+                # Line or point crop is not allowed
+                if patch[2] == patch[0] or patch[3] == patch[1]:
+                    continue
+                overlaps = bbox_overlaps(
+                    patch.reshape(-1, 4), boxes.reshape(-1, 4)).reshape(-1)
+                if len(overlaps) > 0 and overlaps.min() < min_iou:
+                    continue
+
+                # center of boxes should inside the crop img
+                # only adjust boxes and instance masks when the gt is not empty
+                if len(overlaps) > 0:
+                    # adjust boxes
+                    def is_center_of_bboxes_in_patch(boxes, patch):
+                        center = (boxes[:, :2] + boxes[:, 2:]) / 2
+                        mask = ((center[:, 0] > patch[0]) *
+                                (center[:, 1] > patch[1]) *
+                                (center[:, 0] < patch[2]) *
+                                (center[:, 1] < patch[3]))
+                        return mask
+
+                    mask = is_center_of_bboxes_in_patch(boxes, patch)
+                    if not mask.any():
+                        continue
+                    for key in results.get('bbox_fields', []):
+                        boxes = results[key].copy()
+                        mask = is_center_of_bboxes_in_patch(boxes, patch)
+                        boxes = boxes[mask]
+                        if self.bbox_clip_border:
+                            boxes[:, 2:] = boxes[:, 2:].clip(max=patch[2:])
+                            boxes[:, :2] = boxes[:, :2].clip(min=patch[:2])
+                        boxes -= np.tile(patch[:2], 2)
+
+                        results[key] = boxes
+                        # labels
+                        label_key = self.bbox2label.get(key)
+                        if label_key in results:
+                            results[label_key] = results[label_key][mask]
+
+                        # mask fields
+                        mask_key = self.bbox2mask.get(key)
+                        if mask_key in results:
+                            results[mask_key] = results[mask_key][
+                                mask.nonzero()[0]].crop(patch)
+                # adjust the img no matter whether the gt is empty before crop
+                img = img[patch[1]:patch[3], patch[0]:patch[2]]
+                results['img'] = img
+                results['img_shape'] = img.shape
+
+                # seg fields
+                for key in results.get('seg_fields', []):
+                    results[key] = results[key][patch[1]:patch[3],
+                                                patch[0]:patch[2]]
+                return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(min_ious={self.min_ious}, '
+        repr_str += f'min_crop_size={self.min_crop_size}, '
+        repr_str += f'bbox_clip_border={self.bbox_clip_border})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class Corrupt(object):
+    """Corruption augmentation.
+
+    Corruption transforms implemented based on
+    `imagecorruptions <https://github.com/bethgelab/imagecorruptions>`_.
+
+    Args:
+        corruption (str): Corruption name.
+        severity (int, optional): The severity of corruption. Default: 1.
+    """
+
+    def __init__(self, corruption, severity=1):
+        self.corruption = corruption
+        self.severity = severity
+
+    def __call__(self, results):
+        """Call function to corrupt image.
+
+        Args:
+            results (dict): Result dict from loading pipeline.
+
+        Returns:
+            dict: Result dict with images corrupted.
+        """
+
+        if corrupt is None:
+            raise RuntimeError('imagecorruptions is not installed')
+        if 'img_fields' in results:
+            assert results['img_fields'] == ['img'], \
+                'Only single img_fields is allowed'
+        results['img'] = corrupt(
+            results['img'].astype(np.uint8),
+            corruption_name=self.corruption,
+            severity=self.severity)
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(corruption={self.corruption}, '
+        repr_str += f'severity={self.severity})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class Albu(object):
+    """Albumentation augmentation.
+
+    Adds custom transformations from Albumentations library.
+    Please, visit `https://albumentations.readthedocs.io`
+    to get more information.
+
+    An example of ``transforms`` is as followed:
+
+    .. code-block::
+
+        [
+            dict(
+                type='ShiftScaleRotate',
+                shift_limit=0.0625,
+                scale_limit=0.0,
+                rotate_limit=0,
+                interpolation=1,
+                p=0.5),
+            dict(
+                type='RandomBrightnessContrast',
+                brightness_limit=[0.1, 0.3],
+                contrast_limit=[0.1, 0.3],
+                p=0.2),
+            dict(type='ChannelShuffle', p=0.1),
+            dict(
+                type='OneOf',
+                transforms=[
+                    dict(type='Blur', blur_limit=3, p=1.0),
+                    dict(type='MedianBlur', blur_limit=3, p=1.0)
+                ],
+                p=0.1),
+        ]
+
+    Args:
+        transforms (list[dict]): A list of albu transformations
+        bbox_params (dict): Bbox_params for albumentation `Compose`
+        keymap (dict): Contains {'input key':'albumentation-style key'}
+        skip_img_without_anno (bool): Whether to skip the image if no ann left
+            after aug
+    """
+
+    def __init__(self,
+                 transforms,
+                 bbox_params=None,
+                 keymap=None,
+                 update_pad_shape=False,
+                 skip_img_without_anno=False):
+        if Compose is None:
+            raise RuntimeError('albumentations is not installed')
+
+        # Args will be modified later, copying it will be safer
+        transforms = copy.deepcopy(transforms)
+        if bbox_params is not None:
+            bbox_params = copy.deepcopy(bbox_params)
+        if keymap is not None:
+            keymap = copy.deepcopy(keymap)
+        self.transforms = transforms
+        self.filter_lost_elements = False
+        self.update_pad_shape = update_pad_shape
+        self.skip_img_without_anno = skip_img_without_anno
+
+        # A simple workaround to remove masks without boxes
+        if (isinstance(bbox_params, dict) and 'label_fields' in bbox_params
+                and 'filter_lost_elements' in bbox_params):
+            self.filter_lost_elements = True
+            self.origin_label_fields = bbox_params['label_fields']
+            bbox_params['label_fields'] = ['idx_mapper']
+            del bbox_params['filter_lost_elements']
+
+        self.bbox_params = (
+            self.albu_builder(bbox_params) if bbox_params else None)
+        self.aug = Compose([self.albu_builder(t) for t in self.transforms],
+                           bbox_params=self.bbox_params)
+
+        if not keymap:
+            self.keymap_to_albu = {
+                'img': 'image',
+                'gt_masks': 'masks',
+                'gt_bboxes': 'bboxes'
+            }
+        else:
+            self.keymap_to_albu = keymap
+        self.keymap_back = {v: k for k, v in self.keymap_to_albu.items()}
+
+    def albu_builder(self, cfg):
+        """Import a module from albumentations.
+
+        It inherits some of :func:`build_from_cfg` logic.
+
+        Args:
+            cfg (dict): Config dict. It should at least contain the key "type".
+
+        Returns:
+            obj: The constructed object.
+        """
+
+        assert isinstance(cfg, dict) and 'type' in cfg
+        args = cfg.copy()
+
+        obj_type = args.pop('type')
+        if mmcv.is_str(obj_type):
+            if albumentations is None:
+                raise RuntimeError('albumentations is not installed')
+            obj_cls = getattr(albumentations, obj_type)
+        elif inspect.isclass(obj_type):
+            obj_cls = obj_type
+        else:
+            raise TypeError(
+                f'type must be a str or valid type, but got {type(obj_type)}')
+
+        if 'transforms' in args:
+            args['transforms'] = [
+                self.albu_builder(transform)
+                for transform in args['transforms']
+            ]
+
+        return obj_cls(**args)
+
+    @staticmethod
+    def mapper(d, keymap):
+        """Dictionary mapper. Renames keys according to keymap provided.
+
+        Args:
+            d (dict): old dict
+            keymap (dict): {'old_key':'new_key'}
+        Returns:
+            dict: new dict.
+        """
+
+        updated_dict = {}
+        for k, v in zip(d.keys(), d.values()):
+            new_k = keymap.get(k, k)
+            updated_dict[new_k] = d[k]
+        return updated_dict
+
+    def __call__(self, results):
+        # dict to albumentations format
+        results = self.mapper(results, self.keymap_to_albu)
+        # TODO: add bbox_fields
+        if 'bboxes' in results:
+            # to list of boxes
+            if isinstance(results['bboxes'], np.ndarray):
+                results['bboxes'] = [x for x in results['bboxes']]
+            # add pseudo-field for filtration
+            if self.filter_lost_elements:
+                results['idx_mapper'] = np.arange(len(results['bboxes']))
+
+        # TODO: Support mask structure in albu
+        if 'masks' in results:
+            if isinstance(results['masks'], PolygonMasks):
+                raise NotImplementedError(
+                    'Albu only supports BitMap masks now')
+            ori_masks = results['masks']
+            if albumentations.__version__ < '0.5':
+                results['masks'] = results['masks'].masks
+            else:
+                results['masks'] = [mask for mask in results['masks'].masks]
+
+        results = self.aug(**results)
+
+        if 'bboxes' in results:
+            if isinstance(results['bboxes'], list):
+                results['bboxes'] = np.array(
+                    results['bboxes'], dtype=np.float32)
+            results['bboxes'] = results['bboxes'].reshape(-1, 4)
+
+            # filter label_fields
+            if self.filter_lost_elements:
+
+                for label in self.origin_label_fields:
+                    results[label] = np.array(
+                        [results[label][i] for i in results['idx_mapper']])
+                if 'masks' in results:
+                    results['masks'] = np.array(
+                        [results['masks'][i] for i in results['idx_mapper']])
+                    results['masks'] = ori_masks.__class__(
+                        results['masks'], results['image'].shape[0],
+                        results['image'].shape[1])
+
+                if (not len(results['idx_mapper'])
+                        and self.skip_img_without_anno):
+                    return None
+
+        if 'gt_labels' in results:
+            if isinstance(results['gt_labels'], list):
+                results['gt_labels'] = np.array(results['gt_labels'])
+            results['gt_labels'] = results['gt_labels'].astype(np.int64)
+
+        # back to the original format
+        results = self.mapper(results, self.keymap_back)
+
+        # update final shape
+        if self.update_pad_shape:
+            results['pad_shape'] = results['img'].shape
+
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__ + f'(transforms={self.transforms})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class RandomCenterCropPad(object):
+    """Random center crop and random around padding for CornerNet.
+
+    This operation generates randomly cropped image from the original image and
+    pads it simultaneously. Different from :class:`RandomCrop`, the output
+    shape may not equal to ``crop_size`` strictly. We choose a random value
+    from ``ratios`` and the output shape could be larger or smaller than
+    ``crop_size``. The padding operation is also different from :class:`Pad`,
+    here we use around padding instead of right-bottom padding.
+
+    The relation between output image (padding image) and original image:
+
+    .. code:: text
+
+                        output image
+
+               +----------------------------+
+               |          padded area       |
+        +------|----------------------------|----------+
+        |      |         cropped area       |          |
+        |      |         +---------------+  |          |
+        |      |         |    .   center |  |          | original image
+        |      |         |        range  |  |          |
+        |      |         +---------------+  |          |
+        +------|----------------------------|----------+
+               |          padded area       |
+               +----------------------------+
+
+    There are 5 main areas in the figure:
+
+    - output image: output image of this operation, also called padding
+      image in following instruction.
+    - original image: input image of this operation.
+    - padded area: non-intersect area of output image and original image.
+    - cropped area: the overlap of output image and original image.
+    - center range: a smaller area where random center chosen from.
+      center range is computed by ``border`` and original image's shape
+      to avoid our random center is too close to original image's border.
+
+    Also this operation act differently in train and test mode, the summary
+    pipeline is listed below.
+
+    Train pipeline:
+
+    1. Choose a ``random_ratio`` from ``ratios``, the shape of padding image
+       will be ``random_ratio * crop_size``.
+    2. Choose a ``random_center`` in center range.
+    3. Generate padding image with center matches the ``random_center``.
+    4. Initialize the padding image with pixel value equals to ``mean``.
+    5. Copy the cropped area to padding image.
+    6. Refine annotations.
+
+    Test pipeline:
+
+    1. Compute output shape according to ``test_pad_mode``.
+    2. Generate padding image with center matches the original image
+       center.
+    3. Initialize the padding image with pixel value equals to ``mean``.
+    4. Copy the ``cropped area`` to padding image.
+
+    Args:
+        crop_size (tuple | None): expected size after crop, final size will
+            computed according to ratio. Requires (h, w) in train mode, and
+            None in test mode.
+        ratios (tuple): random select a ratio from tuple and crop image to
+            (crop_size[0] * ratio) * (crop_size[1] * ratio).
+            Only available in train mode.
+        border (int): max distance from center select area to image border.
+            Only available in train mode.
+        mean (sequence): Mean values of 3 channels.
+        std (sequence): Std values of 3 channels.
+        to_rgb (bool): Whether to convert the image from BGR to RGB.
+        test_mode (bool): whether involve random variables in transform.
+            In train mode, crop_size is fixed, center coords and ratio is
+            random selected from predefined lists. In test mode, crop_size
+            is image's original shape, center coords and ratio is fixed.
+        test_pad_mode (tuple): padding method and padding shape value, only
+            available in test mode. Default is using 'logical_or' with
+            127 as padding shape value.
+
+            - 'logical_or': final_shape = input_shape | padding_shape_value
+            - 'size_divisor': final_shape = int(
+              ceil(input_shape / padding_shape_value) * padding_shape_value)
+        bbox_clip_border (bool, optional): Whether clip the objects outside
+            the border of the image. Defaults to True.
+    """
+
+    def __init__(self,
+                 crop_size=None,
+                 ratios=(0.9, 1.0, 1.1),
+                 border=128,
+                 mean=None,
+                 std=None,
+                 to_rgb=None,
+                 test_mode=False,
+                 test_pad_mode=('logical_or', 127),
+                 bbox_clip_border=True):
+        if test_mode:
+            assert crop_size is None, 'crop_size must be None in test mode'
+            assert ratios is None, 'ratios must be None in test mode'
+            assert border is None, 'border must be None in test mode'
+            assert isinstance(test_pad_mode, (list, tuple))
+            assert test_pad_mode[0] in ['logical_or', 'size_divisor']
+        else:
+            assert isinstance(crop_size, (list, tuple))
+            assert crop_size[0] > 0 and crop_size[1] > 0, (
+                'crop_size must > 0 in train mode')
+            assert isinstance(ratios, (list, tuple))
+            assert test_pad_mode is None, (
+                'test_pad_mode must be None in train mode')
+
+        self.crop_size = crop_size
+        self.ratios = ratios
+        self.border = border
+        # We do not set default value to mean, std and to_rgb because these
+        # hyper-parameters are easy to forget but could affect the performance.
+        # Please use the same setting as Normalize for performance assurance.
+        assert mean is not None and std is not None and to_rgb is not None
+        self.to_rgb = to_rgb
+        self.input_mean = mean
+        self.input_std = std
+        if to_rgb:
+            self.mean = mean[::-1]
+            self.std = std[::-1]
+        else:
+            self.mean = mean
+            self.std = std
+        self.test_mode = test_mode
+        self.test_pad_mode = test_pad_mode
+        self.bbox_clip_border = bbox_clip_border
+
+    def _get_border(self, border, size):
+        """Get final border for the target size.
+
+        This function generates a ``final_border`` according to image's shape.
+        The area between ``final_border`` and ``size - final_border`` is the
+        ``center range``. We randomly choose center from the ``center range``
+        to avoid our random center is too close to original image's border.
+        Also ``center range`` should be larger than 0.
+
+        Args:
+            border (int): The initial border, default is 128.
+            size (int): The width or height of original image.
+        Returns:
+            int: The final border.
+        """
+        k = 2 * border / size
+        i = pow(2, np.ceil(np.log2(np.ceil(k))) + (k == int(k)))
+        return border // i
+
+    def _filter_boxes(self, patch, boxes):
+        """Check whether the center of each box is in the patch.
+
+        Args:
+            patch (list[int]): The cropped area, [left, top, right, bottom].
+            boxes (numpy array, (N x 4)): Ground truth boxes.
+
+        Returns:
+            mask (numpy array, (N,)): Each box is inside or outside the patch.
+        """
+        center = (boxes[:, :2] + boxes[:, 2:]) / 2
+        mask = (center[:, 0] > patch[0]) * (center[:, 1] > patch[1]) * (
+            center[:, 0] < patch[2]) * (
+                center[:, 1] < patch[3])
+        return mask
+
+    def _crop_image_and_paste(self, image, center, size):
+        """Crop image with a given center and size, then paste the cropped
+        image to a blank image with two centers align.
+
+        This function is equivalent to generating a blank image with ``size``
+        as its shape. Then cover it on the original image with two centers (
+        the center of blank image and the random center of original image)
+        aligned. The overlap area is paste from the original image and the
+        outside area is filled with ``mean pixel``.
+
+        Args:
+            image (np array, H x W x C): Original image.
+            center (list[int]): Target crop center coord.
+            size (list[int]): Target crop size. [target_h, target_w]
+
+        Returns:
+            cropped_img (np array, target_h x target_w x C): Cropped image.
+            border (np array, 4): The distance of four border of
+                ``cropped_img`` to the original image area, [top, bottom,
+                left, right]
+            patch (list[int]): The cropped area, [left, top, right, bottom].
+        """
+        center_y, center_x = center
+        target_h, target_w = size
+        img_h, img_w, img_c = image.shape
+
+        x0 = max(0, center_x - target_w // 2)
+        x1 = min(center_x + target_w // 2, img_w)
+        y0 = max(0, center_y - target_h // 2)
+        y1 = min(center_y + target_h // 2, img_h)
+        patch = np.array((int(x0), int(y0), int(x1), int(y1)))
+
+        left, right = center_x - x0, x1 - center_x
+        top, bottom = center_y - y0, y1 - center_y
+
+        cropped_center_y, cropped_center_x = target_h // 2, target_w // 2
+        cropped_img = np.zeros((target_h, target_w, img_c), dtype=image.dtype)
+        for i in range(img_c):
+            cropped_img[:, :, i] += self.mean[i]
+        y_slice = slice(cropped_center_y - top, cropped_center_y + bottom)
+        x_slice = slice(cropped_center_x - left, cropped_center_x + right)
+        cropped_img[y_slice, x_slice, :] = image[y0:y1, x0:x1, :]
+
+        border = np.array([
+            cropped_center_y - top, cropped_center_y + bottom,
+            cropped_center_x - left, cropped_center_x + right
+        ],
+                          dtype=np.float32)
+
+        return cropped_img, border, patch
+
+    def _train_aug(self, results):
+        """Random crop and around padding the original image.
+
+        Args:
+            results (dict): Image infomations in the augment pipeline.
+
+        Returns:
+            results (dict): The updated dict.
+        """
+        img = results['img']
+        h, w, c = img.shape
+        boxes = results['gt_bboxes']
+        while True:
+            scale = random.choice(self.ratios)
+            new_h = int(self.crop_size[0] * scale)
+            new_w = int(self.crop_size[1] * scale)
+            h_border = self._get_border(self.border, h)
+            w_border = self._get_border(self.border, w)
+
+            for i in range(50):
+                center_x = random.randint(low=w_border, high=w - w_border)
+                center_y = random.randint(low=h_border, high=h - h_border)
+
+                cropped_img, border, patch = self._crop_image_and_paste(
+                    img, [center_y, center_x], [new_h, new_w])
+
+                mask = self._filter_boxes(patch, boxes)
+                # if image do not have valid bbox, any crop patch is valid.
+                if not mask.any() and len(boxes) > 0:
+                    continue
+
+                results['img'] = cropped_img
+                results['img_shape'] = cropped_img.shape
+                results['pad_shape'] = cropped_img.shape
+
+                x0, y0, x1, y1 = patch
+
+                left_w, top_h = center_x - x0, center_y - y0
+                cropped_center_x, cropped_center_y = new_w // 2, new_h // 2
+
+                # crop bboxes accordingly and clip to the image boundary
+                for key in results.get('bbox_fields', []):
+                    mask = self._filter_boxes(patch, results[key])
+                    bboxes = results[key][mask]
+                    bboxes[:, 0:4:2] += cropped_center_x - left_w - x0
+                    bboxes[:, 1:4:2] += cropped_center_y - top_h - y0
+                    if self.bbox_clip_border:
+                        bboxes[:, 0:4:2] = np.clip(bboxes[:, 0:4:2], 0, new_w)
+                        bboxes[:, 1:4:2] = np.clip(bboxes[:, 1:4:2], 0, new_h)
+                    keep = (bboxes[:, 2] > bboxes[:, 0]) & (
+                        bboxes[:, 3] > bboxes[:, 1])
+                    bboxes = bboxes[keep]
+                    results[key] = bboxes
+                    if key in ['gt_bboxes']:
+                        if 'gt_labels' in results:
+                            labels = results['gt_labels'][mask]
+                            labels = labels[keep]
+                            results['gt_labels'] = labels
+                        if 'gt_masks' in results:
+                            raise NotImplementedError(
+                                'RandomCenterCropPad only supports bbox.')
+
+                # crop semantic seg
+                for key in results.get('seg_fields', []):
+                    raise NotImplementedError(
+                        'RandomCenterCropPad only supports bbox.')
+                return results
+
+    def _test_aug(self, results):
+        """Around padding the original image without cropping.
+
+        The padding mode and value are from ``test_pad_mode``.
+
+        Args:
+            results (dict): Image infomations in the augment pipeline.
+
+        Returns:
+            results (dict): The updated dict.
+        """
+        img = results['img']
+        h, w, c = img.shape
+        results['img_shape'] = img.shape
+        if self.test_pad_mode[0] in ['logical_or']:
+            target_h = h | self.test_pad_mode[1]
+            target_w = w | self.test_pad_mode[1]
+        elif self.test_pad_mode[0] in ['size_divisor']:
+            divisor = self.test_pad_mode[1]
+            target_h = int(np.ceil(h / divisor)) * divisor
+            target_w = int(np.ceil(w / divisor)) * divisor
+        else:
+            raise NotImplementedError(
+                'RandomCenterCropPad only support two testing pad mode:'
+                'logical-or and size_divisor.')
+
+        cropped_img, border, _ = self._crop_image_and_paste(
+            img, [h // 2, w // 2], [target_h, target_w])
+        results['img'] = cropped_img
+        results['pad_shape'] = cropped_img.shape
+        results['border'] = border
+        return results
+
+    def __call__(self, results):
+        img = results['img']
+        assert img.dtype == np.float32, (
+            'RandomCenterCropPad needs the input image of dtype np.float32,'
+            ' please set "to_float32=True" in "LoadImageFromFile" pipeline')
+        h, w, c = img.shape
+        assert c == len(self.mean)
+        if self.test_mode:
+            return self._test_aug(results)
+        else:
+            return self._train_aug(results)
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(crop_size={self.crop_size}, '
+        repr_str += f'ratios={self.ratios}, '
+        repr_str += f'border={self.border}, '
+        repr_str += f'mean={self.input_mean}, '
+        repr_str += f'std={self.input_std}, '
+        repr_str += f'to_rgb={self.to_rgb}, '
+        repr_str += f'test_mode={self.test_mode}, '
+        repr_str += f'test_pad_mode={self.test_pad_mode}, '
+        repr_str += f'bbox_clip_border={self.bbox_clip_border})'
+        return repr_str
+
+
+@PIPELINES.register_module()
+class CutOut(object):
+    """CutOut operation.
+
+    Randomly drop some regions of image used in
+    `Cutout <https://arxiv.org/abs/1708.04552>`_.
+
+    Args:
+        n_holes (int | tuple[int, int]): Number of regions to be dropped.
+            If it is given as a list, number of holes will be randomly
+            selected from the closed interval [`n_holes[0]`, `n_holes[1]`].
+        cutout_shape (tuple[int, int] | list[tuple[int, int]]): The candidate
+            shape of dropped regions. It can be `tuple[int, int]` to use a
+            fixed cutout shape, or `list[tuple[int, int]]` to randomly choose
+            shape from the list.
+        cutout_ratio (tuple[float, float] | list[tuple[float, float]]): The
+            candidate ratio of dropped regions. It can be `tuple[float, float]`
+            to use a fixed ratio or `list[tuple[float, float]]` to randomly
+            choose ratio from the list. Please note that `cutout_shape`
+            and `cutout_ratio` cannot be both given at the same time.
+        fill_in (tuple[float, float, float] | tuple[int, int, int]): The value
+            of pixel to fill in the dropped regions. Default: (0, 0, 0).
+    """
+
+    def __init__(self,
+                 n_holes,
+                 cutout_shape=None,
+                 cutout_ratio=None,
+                 fill_in=(0, 0, 0)):
+
+        assert (cutout_shape is None) ^ (cutout_ratio is None), \
+            'Either cutout_shape or cutout_ratio should be specified.'
+        assert (isinstance(cutout_shape, (list, tuple))
+                or isinstance(cutout_ratio, (list, tuple)))
+        if isinstance(n_holes, tuple):
+            assert len(n_holes) == 2 and 0 <= n_holes[0] < n_holes[1]
+        else:
+            n_holes = (n_holes, n_holes)
+        self.n_holes = n_holes
+        self.fill_in = fill_in
+        self.with_ratio = cutout_ratio is not None
+        self.candidates = cutout_ratio if self.with_ratio else cutout_shape
+        if not isinstance(self.candidates, list):
+            self.candidates = [self.candidates]
+
+    def __call__(self, results):
+        """Call function to drop some regions of image."""
+        h, w, c = results['img'].shape
+        n_holes = np.random.randint(self.n_holes[0], self.n_holes[1] + 1)
+        for _ in range(n_holes):
+            x1 = np.random.randint(0, w)
+            y1 = np.random.randint(0, h)
+            index = np.random.randint(0, len(self.candidates))
+            if not self.with_ratio:
+                cutout_w, cutout_h = self.candidates[index]
+            else:
+                cutout_w = int(self.candidates[index][0] * w)
+                cutout_h = int(self.candidates[index][1] * h)
+
+            x2 = np.clip(x1 + cutout_w, 0, w)
+            y2 = np.clip(y1 + cutout_h, 0, h)
+            results['img'][y1:y2, x1:x2, :] = self.fill_in
+
+        return results
+
+    def __repr__(self):
+        repr_str = self.__class__.__name__
+        repr_str += f'(n_holes={self.n_holes}, '
+        repr_str += (f'cutout_ratio={self.candidates}, ' if self.with_ratio
+                     else f'cutout_shape={self.candidates}, ')
+        repr_str += f'fill_in={self.fill_in})'
+        return repr_str
diff --git a/python/autoCut/models/swin/mmdet/datasets/samplers/__init__.py b/python/autoCut/models/swin/mmdet/datasets/samplers/__init__.py
new file mode 100644
index 0000000..2596aeb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/samplers/__init__.py
@@ -0,0 +1,4 @@
+from .distributed_sampler import DistributedSampler
+from .group_sampler import DistributedGroupSampler, GroupSampler
+
+__all__ = ['DistributedSampler', 'DistributedGroupSampler', 'GroupSampler']
diff --git a/python/autoCut/models/swin/mmdet/datasets/samplers/distributed_sampler.py b/python/autoCut/models/swin/mmdet/datasets/samplers/distributed_sampler.py
new file mode 100644
index 0000000..cc61019
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/samplers/distributed_sampler.py
@@ -0,0 +1,39 @@
+import math
+
+import torch
+from torch.utils.data import DistributedSampler as _DistributedSampler
+
+
+class DistributedSampler(_DistributedSampler):
+
+    def __init__(self,
+                 dataset,
+                 num_replicas=None,
+                 rank=None,
+                 shuffle=True,
+                 seed=0):
+        super().__init__(
+            dataset, num_replicas=num_replicas, rank=rank, shuffle=shuffle)
+        # for the compatibility from PyTorch 1.3+
+        self.seed = seed if seed is not None else 0
+
+    def __iter__(self):
+        # deterministically shuffle based on epoch
+        if self.shuffle:
+            g = torch.Generator()
+            g.manual_seed(self.epoch + self.seed)
+            indices = torch.randperm(len(self.dataset), generator=g).tolist()
+        else:
+            indices = torch.arange(len(self.dataset)).tolist()
+
+        # add extra samples to make it evenly divisible
+        # in case that indices is shorter than half of total_size
+        indices = (indices *
+                   math.ceil(self.total_size / len(indices)))[:self.total_size]
+        assert len(indices) == self.total_size
+
+        # subsample
+        indices = indices[self.rank:self.total_size:self.num_replicas]
+        assert len(indices) == self.num_samples
+
+        return iter(indices)
diff --git a/python/autoCut/models/swin/mmdet/datasets/samplers/group_sampler.py b/python/autoCut/models/swin/mmdet/datasets/samplers/group_sampler.py
new file mode 100644
index 0000000..f88cf34
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/samplers/group_sampler.py
@@ -0,0 +1,148 @@
+from __future__ import division
+import math
+
+import numpy as np
+import torch
+from mmcv.runner import get_dist_info
+from torch.utils.data import Sampler
+
+
+class GroupSampler(Sampler):
+
+    def __init__(self, dataset, samples_per_gpu=1):
+        assert hasattr(dataset, 'flag')
+        self.dataset = dataset
+        self.samples_per_gpu = samples_per_gpu
+        self.flag = dataset.flag.astype(np.int64)
+        self.group_sizes = np.bincount(self.flag)
+        self.num_samples = 0
+        for i, size in enumerate(self.group_sizes):
+            self.num_samples += int(np.ceil(
+                size / self.samples_per_gpu)) * self.samples_per_gpu
+
+    def __iter__(self):
+        indices = []
+        for i, size in enumerate(self.group_sizes):
+            if size == 0:
+                continue
+            indice = np.where(self.flag == i)[0]
+            assert len(indice) == size
+            np.random.shuffle(indice)
+            num_extra = int(np.ceil(size / self.samples_per_gpu)
+                            ) * self.samples_per_gpu - len(indice)
+            indice = np.concatenate(
+                [indice, np.random.choice(indice, num_extra)])
+            indices.append(indice)
+        indices = np.concatenate(indices)
+        indices = [
+            indices[i * self.samples_per_gpu:(i + 1) * self.samples_per_gpu]
+            for i in np.random.permutation(
+                range(len(indices) // self.samples_per_gpu))
+        ]
+        indices = np.concatenate(indices)
+        indices = indices.astype(np.int64).tolist()
+        assert len(indices) == self.num_samples
+        return iter(indices)
+
+    def __len__(self):
+        return self.num_samples
+
+
+class DistributedGroupSampler(Sampler):
+    """Sampler that restricts data loading to a subset of the dataset.
+
+    It is especially useful in conjunction with
+    :class:`torch.nn.parallel.DistributedDataParallel`. In such case, each
+    process can pass a DistributedSampler instance as a DataLoader sampler,
+    and load a subset of the original dataset that is exclusive to it.
+
+    .. note::
+        Dataset is assumed to be of constant size.
+
+    Arguments:
+        dataset: Dataset used for sampling.
+        num_replicas (optional): Number of processes participating in
+            distributed training.
+        rank (optional): Rank of the current process within num_replicas.
+        seed (int, optional): random seed used to shuffle the sampler if
+            ``shuffle=True``. This number should be identical across all
+            processes in the distributed group. Default: 0.
+    """
+
+    def __init__(self,
+                 dataset,
+                 samples_per_gpu=1,
+                 num_replicas=None,
+                 rank=None,
+                 seed=0):
+        _rank, _num_replicas = get_dist_info()
+        if num_replicas is None:
+            num_replicas = _num_replicas
+        if rank is None:
+            rank = _rank
+        self.dataset = dataset
+        self.samples_per_gpu = samples_per_gpu
+        self.num_replicas = num_replicas
+        self.rank = rank
+        self.epoch = 0
+        self.seed = seed if seed is not None else 0
+
+        assert hasattr(self.dataset, 'flag')
+        self.flag = self.dataset.flag
+        self.group_sizes = np.bincount(self.flag)
+
+        self.num_samples = 0
+        for i, j in enumerate(self.group_sizes):
+            self.num_samples += int(
+                math.ceil(self.group_sizes[i] * 1.0 / self.samples_per_gpu /
+                          self.num_replicas)) * self.samples_per_gpu
+        self.total_size = self.num_samples * self.num_replicas
+
+    def __iter__(self):
+        # deterministically shuffle based on epoch
+        g = torch.Generator()
+        g.manual_seed(self.epoch + self.seed)
+
+        indices = []
+        for i, size in enumerate(self.group_sizes):
+            if size > 0:
+                indice = np.where(self.flag == i)[0]
+                assert len(indice) == size
+                # add .numpy() to avoid bug when selecting indice in parrots.
+                # TODO: check whether torch.randperm() can be replaced by
+                # numpy.random.permutation().
+                indice = indice[list(
+                    torch.randperm(int(size), generator=g).numpy())].tolist()
+                extra = int(
+                    math.ceil(
+                        size * 1.0 / self.samples_per_gpu / self.num_replicas)
+                ) * self.samples_per_gpu * self.num_replicas - len(indice)
+                # pad indice
+                tmp = indice.copy()
+                for _ in range(extra // size):
+                    indice.extend(tmp)
+                indice.extend(tmp[:extra % size])
+                indices.extend(indice)
+
+        assert len(indices) == self.total_size
+
+        indices = [
+            indices[j] for i in list(
+                torch.randperm(
+                    len(indices) // self.samples_per_gpu, generator=g))
+            for j in range(i * self.samples_per_gpu, (i + 1) *
+                           self.samples_per_gpu)
+        ]
+
+        # subsample
+        offset = self.num_samples * self.rank
+        indices = indices[offset:offset + self.num_samples]
+        assert len(indices) == self.num_samples
+
+        return iter(indices)
+
+    def __len__(self):
+        return self.num_samples
+
+    def set_epoch(self, epoch):
+        self.epoch = epoch
diff --git a/python/autoCut/models/swin/mmdet/datasets/utils.py b/python/autoCut/models/swin/mmdet/datasets/utils.py
new file mode 100644
index 0000000..157c9a2
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/utils.py
@@ -0,0 +1,158 @@
+import copy
+import warnings
+
+from mmcv.cnn import VGG
+from mmcv.runner.hooks import HOOKS, Hook
+
+from mmdet.datasets.builder import PIPELINES
+from mmdet.datasets.pipelines import LoadAnnotations, LoadImageFromFile
+from mmdet.models.dense_heads import GARPNHead, RPNHead
+from mmdet.models.roi_heads.mask_heads import FusedSemanticHead
+
+
+def replace_ImageToTensor(pipelines):
+    """Replace the ImageToTensor transform in a data pipeline to
+    DefaultFormatBundle, which is normally useful in batch inference.
+
+    Args:
+        pipelines (list[dict]): Data pipeline configs.
+
+    Returns:
+        list: The new pipeline list with all ImageToTensor replaced by
+            DefaultFormatBundle.
+
+    Examples:
+        >>> pipelines = [
+        ...    dict(type='LoadImageFromFile'),
+        ...    dict(
+        ...        type='MultiScaleFlipAug',
+        ...        img_scale=(1333, 800),
+        ...        flip=False,
+        ...        transforms=[
+        ...            dict(type='Resize', keep_ratio=True),
+        ...            dict(type='RandomFlip'),
+        ...            dict(type='Normalize', mean=[0, 0, 0], std=[1, 1, 1]),
+        ...            dict(type='Pad', size_divisor=32),
+        ...            dict(type='ImageToTensor', keys=['img']),
+        ...            dict(type='Collect', keys=['img']),
+        ...        ])
+        ...    ]
+        >>> expected_pipelines = [
+        ...    dict(type='LoadImageFromFile'),
+        ...    dict(
+        ...        type='MultiScaleFlipAug',
+        ...        img_scale=(1333, 800),
+        ...        flip=False,
+        ...        transforms=[
+        ...            dict(type='Resize', keep_ratio=True),
+        ...            dict(type='RandomFlip'),
+        ...            dict(type='Normalize', mean=[0, 0, 0], std=[1, 1, 1]),
+        ...            dict(type='Pad', size_divisor=32),
+        ...            dict(type='DefaultFormatBundle'),
+        ...            dict(type='Collect', keys=['img']),
+        ...        ])
+        ...    ]
+        >>> assert expected_pipelines == replace_ImageToTensor(pipelines)
+    """
+    pipelines = copy.deepcopy(pipelines)
+    for i, pipeline in enumerate(pipelines):
+        if pipeline['type'] == 'MultiScaleFlipAug':
+            assert 'transforms' in pipeline
+            pipeline['transforms'] = replace_ImageToTensor(
+                pipeline['transforms'])
+        elif pipeline['type'] == 'ImageToTensor':
+            warnings.warn(
+                '"ImageToTensor" pipeline is replaced by '
+                '"DefaultFormatBundle" for batch inference. It is '
+                'recommended to manually replace it in the test '
+                'data pipeline in your config file.', UserWarning)
+            pipelines[i] = {'type': 'DefaultFormatBundle'}
+    return pipelines
+
+
+def get_loading_pipeline(pipeline):
+    """Only keep loading image and annotations related configuration.
+
+    Args:
+        pipeline (list[dict]): Data pipeline configs.
+
+    Returns:
+        list[dict]: The new pipeline list with only keep
+            loading image and annotations related configuration.
+
+    Examples:
+        >>> pipelines = [
+        ...    dict(type='LoadImageFromFile'),
+        ...    dict(type='LoadAnnotations', with_bbox=True),
+        ...    dict(type='Resize', img_scale=(1333, 800), keep_ratio=True),
+        ...    dict(type='RandomFlip', flip_ratio=0.5),
+        ...    dict(type='Normalize', **img_norm_cfg),
+        ...    dict(type='Pad', size_divisor=32),
+        ...    dict(type='DefaultFormatBundle'),
+        ...    dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels'])
+        ...    ]
+        >>> expected_pipelines = [
+        ...    dict(type='LoadImageFromFile'),
+        ...    dict(type='LoadAnnotations', with_bbox=True)
+        ...    ]
+        >>> assert expected_pipelines ==\
+        ...        get_loading_pipeline(pipelines)
+    """
+    loading_pipeline_cfg = []
+    for cfg in pipeline:
+        obj_cls = PIPELINES.get(cfg['type'])
+        # TODO：use more elegant way to distinguish loading modules
+        if obj_cls is not None and obj_cls in (LoadImageFromFile,
+                                               LoadAnnotations):
+            loading_pipeline_cfg.append(cfg)
+    assert len(loading_pipeline_cfg) == 2, \
+        'The data pipeline in your config file must include ' \
+        'loading image and annotations related pipeline.'
+    return loading_pipeline_cfg
+
+
+@HOOKS.register_module()
+class NumClassCheckHook(Hook):
+
+    def _check_head(self, runner):
+        """Check whether the `num_classes` in head matches the length of
+        `CLASSSES` in `dataset`.
+
+        Args:
+            runner (obj:`EpochBasedRunner`): Epoch based Runner.
+        """
+        model = runner.model
+        dataset = runner.data_loader.dataset
+        if dataset.CLASSES is None:
+            runner.logger.warning(
+                f'Please set `CLASSES` '
+                f'in the {dataset.__class__.__name__} and'
+                f'check if it is consistent with the `num_classes` '
+                f'of head')
+        else:
+            for name, module in model.named_modules():
+                if hasattr(module, 'num_classes') and not isinstance(
+                        module, (RPNHead, VGG, FusedSemanticHead, GARPNHead)):
+                    assert module.num_classes == len(dataset.CLASSES), \
+                        (f'The `num_classes` ({module.num_classes}) in '
+                         f'{module.__class__.__name__} of '
+                         f'{model.__class__.__name__} does not matches '
+                         f'the length of `CLASSES` '
+                         f'{len(dataset.CLASSES)}) in '
+                         f'{dataset.__class__.__name__}')
+
+    def before_train_epoch(self, runner):
+        """Check whether the training dataset is compatible with head.
+
+        Args:
+            runner (obj:`EpochBasedRunner`): Epoch based Runner.
+        """
+        self._check_head(runner)
+
+    def before_val_epoch(self, runner):
+        """Check whether the dataset in val epoch is compatible with head.
+
+        Args:
+            runner (obj:`EpochBasedRunner`): Epoch based Runner.
+        """
+        self._check_head(runner)
diff --git a/python/autoCut/models/swin/mmdet/datasets/voc.py b/python/autoCut/models/swin/mmdet/datasets/voc.py
new file mode 100644
index 0000000..abd4cb8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/voc.py
@@ -0,0 +1,93 @@
+from collections import OrderedDict
+
+from mmcv.utils import print_log
+
+from mmdet.core import eval_map, eval_recalls
+from .builder import DATASETS
+from .xml_style import XMLDataset
+
+
+@DATASETS.register_module()
+class VOCDataset(XMLDataset):
+
+    CLASSES = ('aeroplane', 'bicycle', 'bird', 'boat', 'bottle', 'bus', 'car',
+               'cat', 'chair', 'cow', 'diningtable', 'dog', 'horse',
+               'motorbike', 'person', 'pottedplant', 'sheep', 'sofa', 'train',
+               'tvmonitor')
+
+    def __init__(self, **kwargs):
+        super(VOCDataset, self).__init__(**kwargs)
+        if 'VOC2007' in self.img_prefix:
+            self.year = 2007
+        elif 'VOC2012' in self.img_prefix:
+            self.year = 2012
+        else:
+            raise ValueError('Cannot infer dataset year from img_prefix')
+
+    def evaluate(self,
+                 results,
+                 metric='mAP',
+                 logger=None,
+                 proposal_nums=(100, 300, 1000),
+                 iou_thr=0.5,
+                 scale_ranges=None):
+        """Evaluate in VOC protocol.
+
+        Args:
+            results (list[list | tuple]): Testing results of the dataset.
+            metric (str | list[str]): Metrics to be evaluated. Options are
+                'mAP', 'recall'.
+            logger (logging.Logger | str, optional): Logger used for printing
+                related information during evaluation. Default: None.
+            proposal_nums (Sequence[int]): Proposal number used for evaluating
+                recalls, such as recall@100, recall@1000.
+                Default: (100, 300, 1000).
+            iou_thr (float | list[float]): IoU threshold. Default: 0.5.
+            scale_ranges (list[tuple], optional): Scale ranges for evaluating
+                mAP. If not specified, all bounding boxes would be included in
+                evaluation. Default: None.
+
+        Returns:
+            dict[str, float]: AP/recall metrics.
+        """
+
+        if not isinstance(metric, str):
+            assert len(metric) == 1
+            metric = metric[0]
+        allowed_metrics = ['mAP', 'recall']
+        if metric not in allowed_metrics:
+            raise KeyError(f'metric {metric} is not supported')
+        annotations = [self.get_ann_info(i) for i in range(len(self))]
+        eval_results = OrderedDict()
+        iou_thrs = [iou_thr] if isinstance(iou_thr, float) else iou_thr
+        if metric == 'mAP':
+            assert isinstance(iou_thrs, list)
+            if self.year == 2007:
+                ds_name = 'voc07'
+            else:
+                ds_name = self.CLASSES
+            mean_aps = []
+            for iou_thr in iou_thrs:
+                print_log(f'\n{"-" * 15}iou_thr: {iou_thr}{"-" * 15}')
+                mean_ap, _ = eval_map(
+                    results,
+                    annotations,
+                    scale_ranges=None,
+                    iou_thr=iou_thr,
+                    dataset=ds_name,
+                    logger=logger)
+                mean_aps.append(mean_ap)
+                eval_results[f'AP{int(iou_thr * 100):02d}'] = round(mean_ap, 3)
+            eval_results['mAP'] = sum(mean_aps) / len(mean_aps)
+        elif metric == 'recall':
+            gt_bboxes = [ann['bboxes'] for ann in annotations]
+            recalls = eval_recalls(
+                gt_bboxes, results, proposal_nums, iou_thr, logger=logger)
+            for i, num in enumerate(proposal_nums):
+                for j, iou in enumerate(iou_thr):
+                    eval_results[f'recall@{num}@{iou}'] = recalls[i, j]
+            if recalls.shape[1] > 1:
+                ar = recalls.mean(axis=1)
+                for i, num in enumerate(proposal_nums):
+                    eval_results[f'AR@{num}'] = ar[i]
+        return eval_results
diff --git a/python/autoCut/models/swin/mmdet/datasets/wider_face.py b/python/autoCut/models/swin/mmdet/datasets/wider_face.py
new file mode 100644
index 0000000..3a13907
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/wider_face.py
@@ -0,0 +1,51 @@
+import os.path as osp
+import xml.etree.ElementTree as ET
+
+import mmcv
+
+from .builder import DATASETS
+from .xml_style import XMLDataset
+
+
+@DATASETS.register_module()
+class WIDERFaceDataset(XMLDataset):
+    """Reader for the WIDER Face dataset in PASCAL VOC format.
+
+    Conversion scripts can be found in
+    https://github.com/sovrasov/wider-face-pascal-voc-annotations
+    """
+    CLASSES = ('face', )
+
+    def __init__(self, **kwargs):
+        super(WIDERFaceDataset, self).__init__(**kwargs)
+
+    def load_annotations(self, ann_file):
+        """Load annotation from WIDERFace XML style annotation file.
+
+        Args:
+            ann_file (str): Path of XML file.
+
+        Returns:
+            list[dict]: Annotation info from XML file.
+        """
+
+        data_infos = []
+        img_ids = mmcv.list_from_file(ann_file)
+        for img_id in img_ids:
+            filename = f'{img_id}.jpg'
+            xml_path = osp.join(self.img_prefix, 'Annotations',
+                                f'{img_id}.xml')
+            tree = ET.parse(xml_path)
+            root = tree.getroot()
+            size = root.find('size')
+            width = int(size.find('width').text)
+            height = int(size.find('height').text)
+            folder = root.find('folder').text
+            data_infos.append(
+                dict(
+                    id=img_id,
+                    filename=osp.join(folder, filename),
+                    width=width,
+                    height=height))
+
+        return data_infos
diff --git a/python/autoCut/models/swin/mmdet/datasets/xml_style.py b/python/autoCut/models/swin/mmdet/datasets/xml_style.py
new file mode 100644
index 0000000..7106948
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/datasets/xml_style.py
@@ -0,0 +1,170 @@
+import os.path as osp
+import xml.etree.ElementTree as ET
+
+import mmcv
+import numpy as np
+from PIL import Image
+
+from .builder import DATASETS
+from .custom import CustomDataset
+
+
+@DATASETS.register_module()
+class XMLDataset(CustomDataset):
+    """XML dataset for detection.
+
+    Args:
+        min_size (int | float, optional): The minimum size of bounding
+            boxes in the images. If the size of a bounding box is less than
+            ``min_size``, it would be add to ignored field.
+    """
+
+    def __init__(self, min_size=None, **kwargs):
+        assert self.CLASSES or kwargs.get(
+            'classes', None), 'CLASSES in `XMLDataset` can not be None.'
+        super(XMLDataset, self).__init__(**kwargs)
+        self.cat2label = {cat: i for i, cat in enumerate(self.CLASSES)}
+        self.min_size = min_size
+
+    def load_annotations(self, ann_file):
+        """Load annotation from XML style ann_file.
+
+        Args:
+            ann_file (str): Path of XML file.
+
+        Returns:
+            list[dict]: Annotation info from XML file.
+        """
+
+        data_infos = []
+        img_ids = mmcv.list_from_file(ann_file)
+        for img_id in img_ids:
+            filename = f'JPEGImages/{img_id}.jpg'
+            xml_path = osp.join(self.img_prefix, 'Annotations',
+                                f'{img_id}.xml')
+            tree = ET.parse(xml_path)
+            root = tree.getroot()
+            size = root.find('size')
+            if size is not None:
+                width = int(size.find('width').text)
+                height = int(size.find('height').text)
+            else:
+                img_path = osp.join(self.img_prefix, 'JPEGImages',
+                                    '{}.jpg'.format(img_id))
+                img = Image.open(img_path)
+                width, height = img.size
+            data_infos.append(
+                dict(id=img_id, filename=filename, width=width, height=height))
+
+        return data_infos
+
+    def _filter_imgs(self, min_size=32):
+        """Filter images too small or without annotation."""
+        valid_inds = []
+        for i, img_info in enumerate(self.data_infos):
+            if min(img_info['width'], img_info['height']) < min_size:
+                continue
+            if self.filter_empty_gt:
+                img_id = img_info['id']
+                xml_path = osp.join(self.img_prefix, 'Annotations',
+                                    f'{img_id}.xml')
+                tree = ET.parse(xml_path)
+                root = tree.getroot()
+                for obj in root.findall('object'):
+                    name = obj.find('name').text
+                    if name in self.CLASSES:
+                        valid_inds.append(i)
+                        break
+            else:
+                valid_inds.append(i)
+        return valid_inds
+
+    def get_ann_info(self, idx):
+        """Get annotation from XML file by index.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            dict: Annotation info of specified index.
+        """
+
+        img_id = self.data_infos[idx]['id']
+        xml_path = osp.join(self.img_prefix, 'Annotations', f'{img_id}.xml')
+        tree = ET.parse(xml_path)
+        root = tree.getroot()
+        bboxes = []
+        labels = []
+        bboxes_ignore = []
+        labels_ignore = []
+        for obj in root.findall('object'):
+            name = obj.find('name').text
+            if name not in self.CLASSES:
+                continue
+            label = self.cat2label[name]
+            difficult = obj.find('difficult')
+            difficult = 0 if difficult is None else int(difficult.text)
+            bnd_box = obj.find('bndbox')
+            # TODO: check whether it is necessary to use int
+            # Coordinates may be float type
+            bbox = [
+                int(float(bnd_box.find('xmin').text)),
+                int(float(bnd_box.find('ymin').text)),
+                int(float(bnd_box.find('xmax').text)),
+                int(float(bnd_box.find('ymax').text))
+            ]
+            ignore = False
+            if self.min_size:
+                assert not self.test_mode
+                w = bbox[2] - bbox[0]
+                h = bbox[3] - bbox[1]
+                if w < self.min_size or h < self.min_size:
+                    ignore = True
+            if difficult or ignore:
+                bboxes_ignore.append(bbox)
+                labels_ignore.append(label)
+            else:
+                bboxes.append(bbox)
+                labels.append(label)
+        if not bboxes:
+            bboxes = np.zeros((0, 4))
+            labels = np.zeros((0, ))
+        else:
+            bboxes = np.array(bboxes, ndmin=2) - 1
+            labels = np.array(labels)
+        if not bboxes_ignore:
+            bboxes_ignore = np.zeros((0, 4))
+            labels_ignore = np.zeros((0, ))
+        else:
+            bboxes_ignore = np.array(bboxes_ignore, ndmin=2) - 1
+            labels_ignore = np.array(labels_ignore)
+        ann = dict(
+            bboxes=bboxes.astype(np.float32),
+            labels=labels.astype(np.int64),
+            bboxes_ignore=bboxes_ignore.astype(np.float32),
+            labels_ignore=labels_ignore.astype(np.int64))
+        return ann
+
+    def get_cat_ids(self, idx):
+        """Get category ids in XML file by index.
+
+        Args:
+            idx (int): Index of data.
+
+        Returns:
+            list[int]: All categories in the image of specified index.
+        """
+
+        cat_ids = []
+        img_id = self.data_infos[idx]['id']
+        xml_path = osp.join(self.img_prefix, 'Annotations', f'{img_id}.xml')
+        tree = ET.parse(xml_path)
+        root = tree.getroot()
+        for obj in root.findall('object'):
+            name = obj.find('name').text
+            if name not in self.CLASSES:
+                continue
+            label = self.cat2label[name]
+            cat_ids.append(label)
+
+        return cat_ids
diff --git a/python/autoCut/models/swin/mmdet/models/__init__.py b/python/autoCut/models/swin/mmdet/models/__init__.py
new file mode 100644
index 0000000..44ac998
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/__init__.py
@@ -0,0 +1,16 @@
+from .backbones import *  # noqa: F401,F403
+from .builder import (BACKBONES, DETECTORS, HEADS, LOSSES, NECKS,
+                      ROI_EXTRACTORS, SHARED_HEADS, build_backbone,
+                      build_detector, build_head, build_loss, build_neck,
+                      build_roi_extractor, build_shared_head)
+from .dense_heads import *  # noqa: F401,F403
+from .detectors import *  # noqa: F401,F403
+from .losses import *  # noqa: F401,F403
+from .necks import *  # noqa: F401,F403
+from .roi_heads import *  # noqa: F401,F403
+
+__all__ = [
+    'BACKBONES', 'NECKS', 'ROI_EXTRACTORS', 'SHARED_HEADS', 'HEADS', 'LOSSES',
+    'DETECTORS', 'build_backbone', 'build_neck', 'build_roi_extractor',
+    'build_shared_head', 'build_head', 'build_loss', 'build_detector'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/__init__.py b/python/autoCut/models/swin/mmdet/models/backbones/__init__.py
new file mode 100644
index 0000000..5d35e48
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/__init__.py
@@ -0,0 +1,19 @@
+from .darknet import Darknet
+from .detectors_resnet import DetectoRS_ResNet
+from .detectors_resnext import DetectoRS_ResNeXt
+from .hourglass import HourglassNet
+from .hrnet import HRNet
+from .regnet import RegNet
+from .res2net import Res2Net
+from .resnest import ResNeSt
+from .resnet import ResNet, ResNetV1d
+from .resnext import ResNeXt
+from .ssd_vgg import SSDVGG
+from .trident_resnet import TridentResNet
+from .swin_transformer import SwinTransformer
+
+__all__ = [
+    'RegNet', 'ResNet', 'ResNetV1d', 'ResNeXt', 'SSDVGG', 'HRNet', 'Res2Net',
+    'HourglassNet', 'DetectoRS_ResNet', 'DetectoRS_ResNeXt', 'Darknet',
+    'ResNeSt', 'TridentResNet', 'SwinTransformer'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/darknet.py b/python/autoCut/models/swin/mmdet/models/backbones/darknet.py
new file mode 100644
index 0000000..517fe26
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/darknet.py
@@ -0,0 +1,199 @@
+# Copyright (c) 2019 Western Digital Corporation or its affiliates.
+
+import logging
+
+import torch.nn as nn
+from mmcv.cnn import ConvModule, constant_init, kaiming_init
+from mmcv.runner import load_checkpoint
+from torch.nn.modules.batchnorm import _BatchNorm
+
+from ..builder import BACKBONES
+
+
+class ResBlock(nn.Module):
+    """The basic residual block used in Darknet. Each ResBlock consists of two
+    ConvModules and the input is added to the final output. Each ConvModule is
+    composed of Conv, BN, and LeakyReLU. In YoloV3 paper, the first convLayer
+    has half of the number of the filters as much as the second convLayer. The
+    first convLayer has filter size of 1x1 and the second one has the filter
+    size of 3x3.
+
+    Args:
+        in_channels (int): The input channels. Must be even.
+        conv_cfg (dict): Config dict for convolution layer. Default: None.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+            Default: dict(type='BN', requires_grad=True)
+        act_cfg (dict): Config dict for activation layer.
+            Default: dict(type='LeakyReLU', negative_slope=0.1).
+    """
+
+    def __init__(self,
+                 in_channels,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN', requires_grad=True),
+                 act_cfg=dict(type='LeakyReLU', negative_slope=0.1)):
+        super(ResBlock, self).__init__()
+        assert in_channels % 2 == 0  # ensure the in_channels is even
+        half_in_channels = in_channels // 2
+
+        # shortcut
+        cfg = dict(conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=act_cfg)
+
+        self.conv1 = ConvModule(in_channels, half_in_channels, 1, **cfg)
+        self.conv2 = ConvModule(
+            half_in_channels, in_channels, 3, padding=1, **cfg)
+
+    def forward(self, x):
+        residual = x
+        out = self.conv1(x)
+        out = self.conv2(out)
+        out = out + residual
+
+        return out
+
+
+@BACKBONES.register_module()
+class Darknet(nn.Module):
+    """Darknet backbone.
+
+    Args:
+        depth (int): Depth of Darknet. Currently only support 53.
+        out_indices (Sequence[int]): Output from which stages.
+        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
+            -1 means not freezing any parameters. Default: -1.
+        conv_cfg (dict): Config dict for convolution layer. Default: None.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+            Default: dict(type='BN', requires_grad=True)
+        act_cfg (dict): Config dict for activation layer.
+            Default: dict(type='LeakyReLU', negative_slope=0.1).
+        norm_eval (bool): Whether to set norm layers to eval mode, namely,
+            freeze running stats (mean and var). Note: Effect on Batch Norm
+            and its variants only.
+
+    Example:
+        >>> from mmdet.models import Darknet
+        >>> import torch
+        >>> self = Darknet(depth=53)
+        >>> self.eval()
+        >>> inputs = torch.rand(1, 3, 416, 416)
+        >>> level_outputs = self.forward(inputs)
+        >>> for level_out in level_outputs:
+        ...     print(tuple(level_out.shape))
+        ...
+        (1, 256, 52, 52)
+        (1, 512, 26, 26)
+        (1, 1024, 13, 13)
+    """
+
+    # Dict(depth: (layers, channels))
+    arch_settings = {
+        53: ((1, 2, 8, 8, 4), ((32, 64), (64, 128), (128, 256), (256, 512),
+                               (512, 1024)))
+    }
+
+    def __init__(self,
+                 depth=53,
+                 out_indices=(3, 4, 5),
+                 frozen_stages=-1,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN', requires_grad=True),
+                 act_cfg=dict(type='LeakyReLU', negative_slope=0.1),
+                 norm_eval=True):
+        super(Darknet, self).__init__()
+        if depth not in self.arch_settings:
+            raise KeyError(f'invalid depth {depth} for darknet')
+        self.depth = depth
+        self.out_indices = out_indices
+        self.frozen_stages = frozen_stages
+        self.layers, self.channels = self.arch_settings[depth]
+
+        cfg = dict(conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=act_cfg)
+
+        self.conv1 = ConvModule(3, 32, 3, padding=1, **cfg)
+
+        self.cr_blocks = ['conv1']
+        for i, n_layers in enumerate(self.layers):
+            layer_name = f'conv_res_block{i + 1}'
+            in_c, out_c = self.channels[i]
+            self.add_module(
+                layer_name,
+                self.make_conv_res_block(in_c, out_c, n_layers, **cfg))
+            self.cr_blocks.append(layer_name)
+
+        self.norm_eval = norm_eval
+
+    def forward(self, x):
+        outs = []
+        for i, layer_name in enumerate(self.cr_blocks):
+            cr_block = getattr(self, layer_name)
+            x = cr_block(x)
+            if i in self.out_indices:
+                outs.append(x)
+
+        return tuple(outs)
+
+    def init_weights(self, pretrained=None):
+        if isinstance(pretrained, str):
+            logger = logging.getLogger()
+            load_checkpoint(self, pretrained, strict=False, logger=logger)
+        elif pretrained is None:
+            for m in self.modules():
+                if isinstance(m, nn.Conv2d):
+                    kaiming_init(m)
+                elif isinstance(m, (_BatchNorm, nn.GroupNorm)):
+                    constant_init(m, 1)
+
+        else:
+            raise TypeError('pretrained must be a str or None')
+
+    def _freeze_stages(self):
+        if self.frozen_stages >= 0:
+            for i in range(self.frozen_stages):
+                m = getattr(self, self.cr_blocks[i])
+                m.eval()
+                for param in m.parameters():
+                    param.requires_grad = False
+
+    def train(self, mode=True):
+        super(Darknet, self).train(mode)
+        self._freeze_stages()
+        if mode and self.norm_eval:
+            for m in self.modules():
+                if isinstance(m, _BatchNorm):
+                    m.eval()
+
+    @staticmethod
+    def make_conv_res_block(in_channels,
+                            out_channels,
+                            res_repeat,
+                            conv_cfg=None,
+                            norm_cfg=dict(type='BN', requires_grad=True),
+                            act_cfg=dict(type='LeakyReLU',
+                                         negative_slope=0.1)):
+        """In Darknet backbone, ConvLayer is usually followed by ResBlock. This
+        function will make that. The Conv layers always have 3x3 filters with
+        stride=2. The number of the filters in Conv layer is the same as the
+        out channels of the ResBlock.
+
+        Args:
+            in_channels (int): The number of input channels.
+            out_channels (int): The number of output channels.
+            res_repeat (int): The number of ResBlocks.
+            conv_cfg (dict): Config dict for convolution layer. Default: None.
+            norm_cfg (dict): Dictionary to construct and config norm layer.
+                Default: dict(type='BN', requires_grad=True)
+            act_cfg (dict): Config dict for activation layer.
+                Default: dict(type='LeakyReLU', negative_slope=0.1).
+        """
+
+        cfg = dict(conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=act_cfg)
+
+        model = nn.Sequential()
+        model.add_module(
+            'conv',
+            ConvModule(
+                in_channels, out_channels, 3, stride=2, padding=1, **cfg))
+        for idx in range(res_repeat):
+            model.add_module('res{}'.format(idx),
+                             ResBlock(out_channels, **cfg))
+        return model
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/detectors_resnet.py b/python/autoCut/models/swin/mmdet/models/backbones/detectors_resnet.py
new file mode 100644
index 0000000..519db46
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/detectors_resnet.py
@@ -0,0 +1,305 @@
+import torch.nn as nn
+import torch.utils.checkpoint as cp
+from mmcv.cnn import build_conv_layer, build_norm_layer, constant_init
+
+from ..builder import BACKBONES
+from .resnet import Bottleneck as _Bottleneck
+from .resnet import ResNet
+
+
+class Bottleneck(_Bottleneck):
+    r"""Bottleneck for the ResNet backbone in `DetectoRS
+    <https://arxiv.org/pdf/2006.02334.pdf>`_.
+
+    This bottleneck allows the users to specify whether to use
+    SAC (Switchable Atrous Convolution) and RFP (Recursive Feature Pyramid).
+
+    Args:
+         inplanes (int): The number of input channels.
+         planes (int): The number of output channels before expansion.
+         rfp_inplanes (int, optional): The number of channels from RFP.
+             Default: None. If specified, an additional conv layer will be
+             added for ``rfp_feat``. Otherwise, the structure is the same as
+             base class.
+         sac (dict, optional): Dictionary to construct SAC. Default: None.
+    """
+    expansion = 4
+
+    def __init__(self,
+                 inplanes,
+                 planes,
+                 rfp_inplanes=None,
+                 sac=None,
+                 **kwargs):
+        super(Bottleneck, self).__init__(inplanes, planes, **kwargs)
+
+        assert sac is None or isinstance(sac, dict)
+        self.sac = sac
+        self.with_sac = sac is not None
+        if self.with_sac:
+            self.conv2 = build_conv_layer(
+                self.sac,
+                planes,
+                planes,
+                kernel_size=3,
+                stride=self.conv2_stride,
+                padding=self.dilation,
+                dilation=self.dilation,
+                bias=False)
+
+        self.rfp_inplanes = rfp_inplanes
+        if self.rfp_inplanes:
+            self.rfp_conv = build_conv_layer(
+                None,
+                self.rfp_inplanes,
+                planes * self.expansion,
+                1,
+                stride=1,
+                bias=True)
+        self.init_weights()
+
+    def init_weights(self):
+        """Initialize the weights."""
+        if self.rfp_inplanes:
+            constant_init(self.rfp_conv, 0)
+
+    def rfp_forward(self, x, rfp_feat):
+        """The forward function that also takes the RFP features as input."""
+
+        def _inner_forward(x):
+            identity = x
+
+            out = self.conv1(x)
+            out = self.norm1(out)
+            out = self.relu(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv1_plugin_names)
+
+            out = self.conv2(out)
+            out = self.norm2(out)
+            out = self.relu(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv2_plugin_names)
+
+            out = self.conv3(out)
+            out = self.norm3(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv3_plugin_names)
+
+            if self.downsample is not None:
+                identity = self.downsample(x)
+
+            out += identity
+
+            return out
+
+        if self.with_cp and x.requires_grad:
+            out = cp.checkpoint(_inner_forward, x)
+        else:
+            out = _inner_forward(x)
+
+        if self.rfp_inplanes:
+            rfp_feat = self.rfp_conv(rfp_feat)
+            out = out + rfp_feat
+
+        out = self.relu(out)
+
+        return out
+
+
+class ResLayer(nn.Sequential):
+    """ResLayer to build ResNet style backbone for RPF in detectoRS.
+
+    The difference between this module and base class is that we pass
+    ``rfp_inplanes`` to the first block.
+
+    Args:
+        block (nn.Module): block used to build ResLayer.
+        inplanes (int): inplanes of block.
+        planes (int): planes of block.
+        num_blocks (int): number of blocks.
+        stride (int): stride of the first block. Default: 1
+        avg_down (bool): Use AvgPool instead of stride conv when
+            downsampling in the bottleneck. Default: False
+        conv_cfg (dict): dictionary to construct and config conv layer.
+            Default: None
+        norm_cfg (dict): dictionary to construct and config norm layer.
+            Default: dict(type='BN')
+        downsample_first (bool): Downsample at the first block or last block.
+            False for Hourglass, True for ResNet. Default: True
+        rfp_inplanes (int, optional): The number of channels from RFP.
+            Default: None. If specified, an additional conv layer will be
+            added for ``rfp_feat``. Otherwise, the structure is the same as
+            base class.
+    """
+
+    def __init__(self,
+                 block,
+                 inplanes,
+                 planes,
+                 num_blocks,
+                 stride=1,
+                 avg_down=False,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 downsample_first=True,
+                 rfp_inplanes=None,
+                 **kwargs):
+        self.block = block
+        assert downsample_first, f'downsample_first={downsample_first} is ' \
+                                 'not supported in DetectoRS'
+
+        downsample = None
+        if stride != 1 or inplanes != planes * block.expansion:
+            downsample = []
+            conv_stride = stride
+            if avg_down and stride != 1:
+                conv_stride = 1
+                downsample.append(
+                    nn.AvgPool2d(
+                        kernel_size=stride,
+                        stride=stride,
+                        ceil_mode=True,
+                        count_include_pad=False))
+            downsample.extend([
+                build_conv_layer(
+                    conv_cfg,
+                    inplanes,
+                    planes * block.expansion,
+                    kernel_size=1,
+                    stride=conv_stride,
+                    bias=False),
+                build_norm_layer(norm_cfg, planes * block.expansion)[1]
+            ])
+            downsample = nn.Sequential(*downsample)
+
+        layers = []
+        layers.append(
+            block(
+                inplanes=inplanes,
+                planes=planes,
+                stride=stride,
+                downsample=downsample,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                rfp_inplanes=rfp_inplanes,
+                **kwargs))
+        inplanes = planes * block.expansion
+        for _ in range(1, num_blocks):
+            layers.append(
+                block(
+                    inplanes=inplanes,
+                    planes=planes,
+                    stride=1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg,
+                    **kwargs))
+
+        super(ResLayer, self).__init__(*layers)
+
+
+@BACKBONES.register_module()
+class DetectoRS_ResNet(ResNet):
+    """ResNet backbone for DetectoRS.
+
+    Args:
+        sac (dict, optional): Dictionary to construct SAC (Switchable Atrous
+            Convolution). Default: None.
+        stage_with_sac (list): Which stage to use sac. Default: (False, False,
+            False, False).
+        rfp_inplanes (int, optional): The number of channels from RFP.
+            Default: None. If specified, an additional conv layer will be
+            added for ``rfp_feat``. Otherwise, the structure is the same as
+            base class.
+        output_img (bool): If ``True``, the input image will be inserted into
+            the starting position of output. Default: False.
+        pretrained (str, optional): The pretrained model to load.
+    """
+
+    arch_settings = {
+        50: (Bottleneck, (3, 4, 6, 3)),
+        101: (Bottleneck, (3, 4, 23, 3)),
+        152: (Bottleneck, (3, 8, 36, 3))
+    }
+
+    def __init__(self,
+                 sac=None,
+                 stage_with_sac=(False, False, False, False),
+                 rfp_inplanes=None,
+                 output_img=False,
+                 pretrained=None,
+                 **kwargs):
+        self.sac = sac
+        self.stage_with_sac = stage_with_sac
+        self.rfp_inplanes = rfp_inplanes
+        self.output_img = output_img
+        self.pretrained = pretrained
+        super(DetectoRS_ResNet, self).__init__(**kwargs)
+
+        self.inplanes = self.stem_channels
+        self.res_layers = []
+        for i, num_blocks in enumerate(self.stage_blocks):
+            stride = self.strides[i]
+            dilation = self.dilations[i]
+            dcn = self.dcn if self.stage_with_dcn[i] else None
+            sac = self.sac if self.stage_with_sac[i] else None
+            if self.plugins is not None:
+                stage_plugins = self.make_stage_plugins(self.plugins, i)
+            else:
+                stage_plugins = None
+            planes = self.base_channels * 2**i
+            res_layer = self.make_res_layer(
+                block=self.block,
+                inplanes=self.inplanes,
+                planes=planes,
+                num_blocks=num_blocks,
+                stride=stride,
+                dilation=dilation,
+                style=self.style,
+                avg_down=self.avg_down,
+                with_cp=self.with_cp,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg,
+                dcn=dcn,
+                sac=sac,
+                rfp_inplanes=rfp_inplanes if i > 0 else None,
+                plugins=stage_plugins)
+            self.inplanes = planes * self.block.expansion
+            layer_name = f'layer{i + 1}'
+            self.add_module(layer_name, res_layer)
+            self.res_layers.append(layer_name)
+
+        self._freeze_stages()
+
+    def make_res_layer(self, **kwargs):
+        """Pack all blocks in a stage into a ``ResLayer`` for DetectoRS."""
+        return ResLayer(**kwargs)
+
+    def forward(self, x):
+        """Forward function."""
+        outs = list(super(DetectoRS_ResNet, self).forward(x))
+        if self.output_img:
+            outs.insert(0, x)
+        return tuple(outs)
+
+    def rfp_forward(self, x, rfp_feats):
+        """Forward function for RFP."""
+        if self.deep_stem:
+            x = self.stem(x)
+        else:
+            x = self.conv1(x)
+            x = self.norm1(x)
+            x = self.relu(x)
+        x = self.maxpool(x)
+        outs = []
+        for i, layer_name in enumerate(self.res_layers):
+            res_layer = getattr(self, layer_name)
+            rfp_feat = rfp_feats[i] if i > 0 else None
+            for layer in res_layer:
+                x = layer.rfp_forward(x, rfp_feat)
+            if i in self.out_indices:
+                outs.append(x)
+        return tuple(outs)
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/detectors_resnext.py b/python/autoCut/models/swin/mmdet/models/backbones/detectors_resnext.py
new file mode 100644
index 0000000..57d032f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/detectors_resnext.py
@@ -0,0 +1,122 @@
+import math
+
+from mmcv.cnn import build_conv_layer, build_norm_layer
+
+from ..builder import BACKBONES
+from .detectors_resnet import Bottleneck as _Bottleneck
+from .detectors_resnet import DetectoRS_ResNet
+
+
+class Bottleneck(_Bottleneck):
+    expansion = 4
+
+    def __init__(self,
+                 inplanes,
+                 planes,
+                 groups=1,
+                 base_width=4,
+                 base_channels=64,
+                 **kwargs):
+        """Bottleneck block for ResNeXt.
+
+        If style is "pytorch", the stride-two layer is the 3x3 conv layer, if
+        it is "caffe", the stride-two layer is the first 1x1 conv layer.
+        """
+        super(Bottleneck, self).__init__(inplanes, planes, **kwargs)
+
+        if groups == 1:
+            width = self.planes
+        else:
+            width = math.floor(self.planes *
+                               (base_width / base_channels)) * groups
+
+        self.norm1_name, norm1 = build_norm_layer(
+            self.norm_cfg, width, postfix=1)
+        self.norm2_name, norm2 = build_norm_layer(
+            self.norm_cfg, width, postfix=2)
+        self.norm3_name, norm3 = build_norm_layer(
+            self.norm_cfg, self.planes * self.expansion, postfix=3)
+
+        self.conv1 = build_conv_layer(
+            self.conv_cfg,
+            self.inplanes,
+            width,
+            kernel_size=1,
+            stride=self.conv1_stride,
+            bias=False)
+        self.add_module(self.norm1_name, norm1)
+        fallback_on_stride = False
+        self.with_modulated_dcn = False
+        if self.with_dcn:
+            fallback_on_stride = self.dcn.pop('fallback_on_stride', False)
+        if self.with_sac:
+            self.conv2 = build_conv_layer(
+                self.sac,
+                width,
+                width,
+                kernel_size=3,
+                stride=self.conv2_stride,
+                padding=self.dilation,
+                dilation=self.dilation,
+                groups=groups,
+                bias=False)
+        elif not self.with_dcn or fallback_on_stride:
+            self.conv2 = build_conv_layer(
+                self.conv_cfg,
+                width,
+                width,
+                kernel_size=3,
+                stride=self.conv2_stride,
+                padding=self.dilation,
+                dilation=self.dilation,
+                groups=groups,
+                bias=False)
+        else:
+            assert self.conv_cfg is None, 'conv_cfg must be None for DCN'
+            self.conv2 = build_conv_layer(
+                self.dcn,
+                width,
+                width,
+                kernel_size=3,
+                stride=self.conv2_stride,
+                padding=self.dilation,
+                dilation=self.dilation,
+                groups=groups,
+                bias=False)
+
+        self.add_module(self.norm2_name, norm2)
+        self.conv3 = build_conv_layer(
+            self.conv_cfg,
+            width,
+            self.planes * self.expansion,
+            kernel_size=1,
+            bias=False)
+        self.add_module(self.norm3_name, norm3)
+
+
+@BACKBONES.register_module()
+class DetectoRS_ResNeXt(DetectoRS_ResNet):
+    """ResNeXt backbone for DetectoRS.
+
+    Args:
+        groups (int): The number of groups in ResNeXt.
+        base_width (int): The base width of ResNeXt.
+    """
+
+    arch_settings = {
+        50: (Bottleneck, (3, 4, 6, 3)),
+        101: (Bottleneck, (3, 4, 23, 3)),
+        152: (Bottleneck, (3, 8, 36, 3))
+    }
+
+    def __init__(self, groups=1, base_width=4, **kwargs):
+        self.groups = groups
+        self.base_width = base_width
+        super(DetectoRS_ResNeXt, self).__init__(**kwargs)
+
+    def make_res_layer(self, **kwargs):
+        return super().make_res_layer(
+            groups=self.groups,
+            base_width=self.base_width,
+            base_channels=self.base_channels,
+            **kwargs)
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/hourglass.py b/python/autoCut/models/swin/mmdet/models/backbones/hourglass.py
new file mode 100644
index 0000000..3422ace
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/hourglass.py
@@ -0,0 +1,198 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule
+
+from ..builder import BACKBONES
+from ..utils import ResLayer
+from .resnet import BasicBlock
+
+
+class HourglassModule(nn.Module):
+    """Hourglass Module for HourglassNet backbone.
+
+    Generate module recursively and use BasicBlock as the base unit.
+
+    Args:
+        depth (int): Depth of current HourglassModule.
+        stage_channels (list[int]): Feature channels of sub-modules in current
+            and follow-up HourglassModule.
+        stage_blocks (list[int]): Number of sub-modules stacked in current and
+            follow-up HourglassModule.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+    """
+
+    def __init__(self,
+                 depth,
+                 stage_channels,
+                 stage_blocks,
+                 norm_cfg=dict(type='BN', requires_grad=True)):
+        super(HourglassModule, self).__init__()
+
+        self.depth = depth
+
+        cur_block = stage_blocks[0]
+        next_block = stage_blocks[1]
+
+        cur_channel = stage_channels[0]
+        next_channel = stage_channels[1]
+
+        self.up1 = ResLayer(
+            BasicBlock, cur_channel, cur_channel, cur_block, norm_cfg=norm_cfg)
+
+        self.low1 = ResLayer(
+            BasicBlock,
+            cur_channel,
+            next_channel,
+            cur_block,
+            stride=2,
+            norm_cfg=norm_cfg)
+
+        if self.depth > 1:
+            self.low2 = HourglassModule(depth - 1, stage_channels[1:],
+                                        stage_blocks[1:])
+        else:
+            self.low2 = ResLayer(
+                BasicBlock,
+                next_channel,
+                next_channel,
+                next_block,
+                norm_cfg=norm_cfg)
+
+        self.low3 = ResLayer(
+            BasicBlock,
+            next_channel,
+            cur_channel,
+            cur_block,
+            norm_cfg=norm_cfg,
+            downsample_first=False)
+
+        self.up2 = nn.Upsample(scale_factor=2)
+
+    def forward(self, x):
+        """Forward function."""
+        up1 = self.up1(x)
+        low1 = self.low1(x)
+        low2 = self.low2(low1)
+        low3 = self.low3(low2)
+        up2 = self.up2(low3)
+        return up1 + up2
+
+
+@BACKBONES.register_module()
+class HourglassNet(nn.Module):
+    """HourglassNet backbone.
+
+    Stacked Hourglass Networks for Human Pose Estimation.
+    More details can be found in the `paper
+    <https://arxiv.org/abs/1603.06937>`_ .
+
+    Args:
+        downsample_times (int): Downsample times in a HourglassModule.
+        num_stacks (int): Number of HourglassModule modules stacked,
+            1 for Hourglass-52, 2 for Hourglass-104.
+        stage_channels (list[int]): Feature channel of each sub-module in a
+            HourglassModule.
+        stage_blocks (list[int]): Number of sub-modules stacked in a
+            HourglassModule.
+        feat_channel (int): Feature channel of conv after a HourglassModule.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+
+    Example:
+        >>> from mmdet.models import HourglassNet
+        >>> import torch
+        >>> self = HourglassNet()
+        >>> self.eval()
+        >>> inputs = torch.rand(1, 3, 511, 511)
+        >>> level_outputs = self.forward(inputs)
+        >>> for level_output in level_outputs:
+        ...     print(tuple(level_output.shape))
+        (1, 256, 128, 128)
+        (1, 256, 128, 128)
+    """
+
+    def __init__(self,
+                 downsample_times=5,
+                 num_stacks=2,
+                 stage_channels=(256, 256, 384, 384, 384, 512),
+                 stage_blocks=(2, 2, 2, 2, 2, 4),
+                 feat_channel=256,
+                 norm_cfg=dict(type='BN', requires_grad=True)):
+        super(HourglassNet, self).__init__()
+
+        self.num_stacks = num_stacks
+        assert self.num_stacks >= 1
+        assert len(stage_channels) == len(stage_blocks)
+        assert len(stage_channels) > downsample_times
+
+        cur_channel = stage_channels[0]
+
+        self.stem = nn.Sequential(
+            ConvModule(3, 128, 7, padding=3, stride=2, norm_cfg=norm_cfg),
+            ResLayer(BasicBlock, 128, 256, 1, stride=2, norm_cfg=norm_cfg))
+
+        self.hourglass_modules = nn.ModuleList([
+            HourglassModule(downsample_times, stage_channels, stage_blocks)
+            for _ in range(num_stacks)
+        ])
+
+        self.inters = ResLayer(
+            BasicBlock,
+            cur_channel,
+            cur_channel,
+            num_stacks - 1,
+            norm_cfg=norm_cfg)
+
+        self.conv1x1s = nn.ModuleList([
+            ConvModule(
+                cur_channel, cur_channel, 1, norm_cfg=norm_cfg, act_cfg=None)
+            for _ in range(num_stacks - 1)
+        ])
+
+        self.out_convs = nn.ModuleList([
+            ConvModule(
+                cur_channel, feat_channel, 3, padding=1, norm_cfg=norm_cfg)
+            for _ in range(num_stacks)
+        ])
+
+        self.remap_convs = nn.ModuleList([
+            ConvModule(
+                feat_channel, cur_channel, 1, norm_cfg=norm_cfg, act_cfg=None)
+            for _ in range(num_stacks - 1)
+        ])
+
+        self.relu = nn.ReLU(inplace=True)
+
+    def init_weights(self, pretrained=None):
+        """Init module weights.
+
+        We do nothing in this function because all modules we used
+        (ConvModule, BasicBlock and etc.) have default initialization, and
+        currently we don't provide pretrained model of HourglassNet.
+
+        Detector's __init__() will call backbone's init_weights() with
+        pretrained as input, so we keep this function.
+        """
+        # Training Centripetal Model needs to reset parameters for Conv2d
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                m.reset_parameters()
+
+    def forward(self, x):
+        """Forward function."""
+        inter_feat = self.stem(x)
+        out_feats = []
+
+        for ind in range(self.num_stacks):
+            single_hourglass = self.hourglass_modules[ind]
+            out_conv = self.out_convs[ind]
+
+            hourglass_feat = single_hourglass(inter_feat)
+            out_feat = out_conv(hourglass_feat)
+            out_feats.append(out_feat)
+
+            if ind < self.num_stacks - 1:
+                inter_feat = self.conv1x1s[ind](
+                    inter_feat) + self.remap_convs[ind](
+                        out_feat)
+                inter_feat = self.inters[ind](self.relu(inter_feat))
+
+        return out_feats
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/hrnet.py b/python/autoCut/models/swin/mmdet/models/backbones/hrnet.py
new file mode 100644
index 0000000..c0fd0a9
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/hrnet.py
@@ -0,0 +1,537 @@
+import torch.nn as nn
+from mmcv.cnn import (build_conv_layer, build_norm_layer, constant_init,
+                      kaiming_init)
+from mmcv.runner import load_checkpoint
+from torch.nn.modules.batchnorm import _BatchNorm
+
+from mmdet.utils import get_root_logger
+from ..builder import BACKBONES
+from .resnet import BasicBlock, Bottleneck
+
+
+class HRModule(nn.Module):
+    """High-Resolution Module for HRNet.
+
+    In this module, every branch has 4 BasicBlocks/Bottlenecks. Fusion/Exchange
+    is in this module.
+    """
+
+    def __init__(self,
+                 num_branches,
+                 blocks,
+                 num_blocks,
+                 in_channels,
+                 num_channels,
+                 multiscale_output=True,
+                 with_cp=False,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN')):
+        super(HRModule, self).__init__()
+        self._check_branches(num_branches, num_blocks, in_channels,
+                             num_channels)
+
+        self.in_channels = in_channels
+        self.num_branches = num_branches
+
+        self.multiscale_output = multiscale_output
+        self.norm_cfg = norm_cfg
+        self.conv_cfg = conv_cfg
+        self.with_cp = with_cp
+        self.branches = self._make_branches(num_branches, blocks, num_blocks,
+                                            num_channels)
+        self.fuse_layers = self._make_fuse_layers()
+        self.relu = nn.ReLU(inplace=False)
+
+    def _check_branches(self, num_branches, num_blocks, in_channels,
+                        num_channels):
+        if num_branches != len(num_blocks):
+            error_msg = f'NUM_BRANCHES({num_branches}) ' \
+                f'!= NUM_BLOCKS({len(num_blocks)})'
+            raise ValueError(error_msg)
+
+        if num_branches != len(num_channels):
+            error_msg = f'NUM_BRANCHES({num_branches}) ' \
+                f'!= NUM_CHANNELS({len(num_channels)})'
+            raise ValueError(error_msg)
+
+        if num_branches != len(in_channels):
+            error_msg = f'NUM_BRANCHES({num_branches}) ' \
+                f'!= NUM_INCHANNELS({len(in_channels)})'
+            raise ValueError(error_msg)
+
+    def _make_one_branch(self,
+                         branch_index,
+                         block,
+                         num_blocks,
+                         num_channels,
+                         stride=1):
+        downsample = None
+        if stride != 1 or \
+                self.in_channels[branch_index] != \
+                num_channels[branch_index] * block.expansion:
+            downsample = nn.Sequential(
+                build_conv_layer(
+                    self.conv_cfg,
+                    self.in_channels[branch_index],
+                    num_channels[branch_index] * block.expansion,
+                    kernel_size=1,
+                    stride=stride,
+                    bias=False),
+                build_norm_layer(self.norm_cfg, num_channels[branch_index] *
+                                 block.expansion)[1])
+
+        layers = []
+        layers.append(
+            block(
+                self.in_channels[branch_index],
+                num_channels[branch_index],
+                stride,
+                downsample=downsample,
+                with_cp=self.with_cp,
+                norm_cfg=self.norm_cfg,
+                conv_cfg=self.conv_cfg))
+        self.in_channels[branch_index] = \
+            num_channels[branch_index] * block.expansion
+        for i in range(1, num_blocks[branch_index]):
+            layers.append(
+                block(
+                    self.in_channels[branch_index],
+                    num_channels[branch_index],
+                    with_cp=self.with_cp,
+                    norm_cfg=self.norm_cfg,
+                    conv_cfg=self.conv_cfg))
+
+        return nn.Sequential(*layers)
+
+    def _make_branches(self, num_branches, block, num_blocks, num_channels):
+        branches = []
+
+        for i in range(num_branches):
+            branches.append(
+                self._make_one_branch(i, block, num_blocks, num_channels))
+
+        return nn.ModuleList(branches)
+
+    def _make_fuse_layers(self):
+        if self.num_branches == 1:
+            return None
+
+        num_branches = self.num_branches
+        in_channels = self.in_channels
+        fuse_layers = []
+        num_out_branches = num_branches if self.multiscale_output else 1
+        for i in range(num_out_branches):
+            fuse_layer = []
+            for j in range(num_branches):
+                if j > i:
+                    fuse_layer.append(
+                        nn.Sequential(
+                            build_conv_layer(
+                                self.conv_cfg,
+                                in_channels[j],
+                                in_channels[i],
+                                kernel_size=1,
+                                stride=1,
+                                padding=0,
+                                bias=False),
+                            build_norm_layer(self.norm_cfg, in_channels[i])[1],
+                            nn.Upsample(
+                                scale_factor=2**(j - i), mode='nearest')))
+                elif j == i:
+                    fuse_layer.append(None)
+                else:
+                    conv_downsamples = []
+                    for k in range(i - j):
+                        if k == i - j - 1:
+                            conv_downsamples.append(
+                                nn.Sequential(
+                                    build_conv_layer(
+                                        self.conv_cfg,
+                                        in_channels[j],
+                                        in_channels[i],
+                                        kernel_size=3,
+                                        stride=2,
+                                        padding=1,
+                                        bias=False),
+                                    build_norm_layer(self.norm_cfg,
+                                                     in_channels[i])[1]))
+                        else:
+                            conv_downsamples.append(
+                                nn.Sequential(
+                                    build_conv_layer(
+                                        self.conv_cfg,
+                                        in_channels[j],
+                                        in_channels[j],
+                                        kernel_size=3,
+                                        stride=2,
+                                        padding=1,
+                                        bias=False),
+                                    build_norm_layer(self.norm_cfg,
+                                                     in_channels[j])[1],
+                                    nn.ReLU(inplace=False)))
+                    fuse_layer.append(nn.Sequential(*conv_downsamples))
+            fuse_layers.append(nn.ModuleList(fuse_layer))
+
+        return nn.ModuleList(fuse_layers)
+
+    def forward(self, x):
+        """Forward function."""
+        if self.num_branches == 1:
+            return [self.branches[0](x[0])]
+
+        for i in range(self.num_branches):
+            x[i] = self.branches[i](x[i])
+
+        x_fuse = []
+        for i in range(len(self.fuse_layers)):
+            y = 0
+            for j in range(self.num_branches):
+                if i == j:
+                    y += x[j]
+                else:
+                    y += self.fuse_layers[i][j](x[j])
+            x_fuse.append(self.relu(y))
+        return x_fuse
+
+
+@BACKBONES.register_module()
+class HRNet(nn.Module):
+    """HRNet backbone.
+
+    High-Resolution Representations for Labeling Pixels and Regions
+    arXiv: https://arxiv.org/abs/1904.04514
+
+    Args:
+        extra (dict): detailed configuration for each stage of HRNet.
+        in_channels (int): Number of input image channels. Default: 3.
+        conv_cfg (dict): dictionary to construct and config conv layer.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+        norm_eval (bool): Whether to set norm layers to eval mode, namely,
+            freeze running stats (mean and var). Note: Effect on Batch Norm
+            and its variants only.
+        with_cp (bool): Use checkpoint or not. Using checkpoint will save some
+            memory while slowing down the training speed.
+        zero_init_residual (bool): whether to use zero init for last norm layer
+            in resblocks to let them behave as identity.
+
+    Example:
+        >>> from mmdet.models import HRNet
+        >>> import torch
+        >>> extra = dict(
+        >>>     stage1=dict(
+        >>>         num_modules=1,
+        >>>         num_branches=1,
+        >>>         block='BOTTLENECK',
+        >>>         num_blocks=(4, ),
+        >>>         num_channels=(64, )),
+        >>>     stage2=dict(
+        >>>         num_modules=1,
+        >>>         num_branches=2,
+        >>>         block='BASIC',
+        >>>         num_blocks=(4, 4),
+        >>>         num_channels=(32, 64)),
+        >>>     stage3=dict(
+        >>>         num_modules=4,
+        >>>         num_branches=3,
+        >>>         block='BASIC',
+        >>>         num_blocks=(4, 4, 4),
+        >>>         num_channels=(32, 64, 128)),
+        >>>     stage4=dict(
+        >>>         num_modules=3,
+        >>>         num_branches=4,
+        >>>         block='BASIC',
+        >>>         num_blocks=(4, 4, 4, 4),
+        >>>         num_channels=(32, 64, 128, 256)))
+        >>> self = HRNet(extra, in_channels=1)
+        >>> self.eval()
+        >>> inputs = torch.rand(1, 1, 32, 32)
+        >>> level_outputs = self.forward(inputs)
+        >>> for level_out in level_outputs:
+        ...     print(tuple(level_out.shape))
+        (1, 32, 8, 8)
+        (1, 64, 4, 4)
+        (1, 128, 2, 2)
+        (1, 256, 1, 1)
+    """
+
+    blocks_dict = {'BASIC': BasicBlock, 'BOTTLENECK': Bottleneck}
+
+    def __init__(self,
+                 extra,
+                 in_channels=3,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 norm_eval=True,
+                 with_cp=False,
+                 zero_init_residual=False):
+        super(HRNet, self).__init__()
+        self.extra = extra
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.norm_eval = norm_eval
+        self.with_cp = with_cp
+        self.zero_init_residual = zero_init_residual
+
+        # stem net
+        self.norm1_name, norm1 = build_norm_layer(self.norm_cfg, 64, postfix=1)
+        self.norm2_name, norm2 = build_norm_layer(self.norm_cfg, 64, postfix=2)
+
+        self.conv1 = build_conv_layer(
+            self.conv_cfg,
+            in_channels,
+            64,
+            kernel_size=3,
+            stride=2,
+            padding=1,
+            bias=False)
+
+        self.add_module(self.norm1_name, norm1)
+        self.conv2 = build_conv_layer(
+            self.conv_cfg,
+            64,
+            64,
+            kernel_size=3,
+            stride=2,
+            padding=1,
+            bias=False)
+
+        self.add_module(self.norm2_name, norm2)
+        self.relu = nn.ReLU(inplace=True)
+
+        # stage 1
+        self.stage1_cfg = self.extra['stage1']
+        num_channels = self.stage1_cfg['num_channels'][0]
+        block_type = self.stage1_cfg['block']
+        num_blocks = self.stage1_cfg['num_blocks'][0]
+
+        block = self.blocks_dict[block_type]
+        stage1_out_channels = num_channels * block.expansion
+        self.layer1 = self._make_layer(block, 64, num_channels, num_blocks)
+
+        # stage 2
+        self.stage2_cfg = self.extra['stage2']
+        num_channels = self.stage2_cfg['num_channels']
+        block_type = self.stage2_cfg['block']
+
+        block = self.blocks_dict[block_type]
+        num_channels = [channel * block.expansion for channel in num_channels]
+        self.transition1 = self._make_transition_layer([stage1_out_channels],
+                                                       num_channels)
+        self.stage2, pre_stage_channels = self._make_stage(
+            self.stage2_cfg, num_channels)
+
+        # stage 3
+        self.stage3_cfg = self.extra['stage3']
+        num_channels = self.stage3_cfg['num_channels']
+        block_type = self.stage3_cfg['block']
+
+        block = self.blocks_dict[block_type]
+        num_channels = [channel * block.expansion for channel in num_channels]
+        self.transition2 = self._make_transition_layer(pre_stage_channels,
+                                                       num_channels)
+        self.stage3, pre_stage_channels = self._make_stage(
+            self.stage3_cfg, num_channels)
+
+        # stage 4
+        self.stage4_cfg = self.extra['stage4']
+        num_channels = self.stage4_cfg['num_channels']
+        block_type = self.stage4_cfg['block']
+
+        block = self.blocks_dict[block_type]
+        num_channels = [channel * block.expansion for channel in num_channels]
+        self.transition3 = self._make_transition_layer(pre_stage_channels,
+                                                       num_channels)
+        self.stage4, pre_stage_channels = self._make_stage(
+            self.stage4_cfg, num_channels)
+
+    @property
+    def norm1(self):
+        """nn.Module: the normalization layer named "norm1" """
+        return getattr(self, self.norm1_name)
+
+    @property
+    def norm2(self):
+        """nn.Module: the normalization layer named "norm2" """
+        return getattr(self, self.norm2_name)
+
+    def _make_transition_layer(self, num_channels_pre_layer,
+                               num_channels_cur_layer):
+        num_branches_cur = len(num_channels_cur_layer)
+        num_branches_pre = len(num_channels_pre_layer)
+
+        transition_layers = []
+        for i in range(num_branches_cur):
+            if i < num_branches_pre:
+                if num_channels_cur_layer[i] != num_channels_pre_layer[i]:
+                    transition_layers.append(
+                        nn.Sequential(
+                            build_conv_layer(
+                                self.conv_cfg,
+                                num_channels_pre_layer[i],
+                                num_channels_cur_layer[i],
+                                kernel_size=3,
+                                stride=1,
+                                padding=1,
+                                bias=False),
+                            build_norm_layer(self.norm_cfg,
+                                             num_channels_cur_layer[i])[1],
+                            nn.ReLU(inplace=True)))
+                else:
+                    transition_layers.append(None)
+            else:
+                conv_downsamples = []
+                for j in range(i + 1 - num_branches_pre):
+                    in_channels = num_channels_pre_layer[-1]
+                    out_channels = num_channels_cur_layer[i] \
+                        if j == i - num_branches_pre else in_channels
+                    conv_downsamples.append(
+                        nn.Sequential(
+                            build_conv_layer(
+                                self.conv_cfg,
+                                in_channels,
+                                out_channels,
+                                kernel_size=3,
+                                stride=2,
+                                padding=1,
+                                bias=False),
+                            build_norm_layer(self.norm_cfg, out_channels)[1],
+                            nn.ReLU(inplace=True)))
+                transition_layers.append(nn.Sequential(*conv_downsamples))
+
+        return nn.ModuleList(transition_layers)
+
+    def _make_layer(self, block, inplanes, planes, blocks, stride=1):
+        downsample = None
+        if stride != 1 or inplanes != planes * block.expansion:
+            downsample = nn.Sequential(
+                build_conv_layer(
+                    self.conv_cfg,
+                    inplanes,
+                    planes * block.expansion,
+                    kernel_size=1,
+                    stride=stride,
+                    bias=False),
+                build_norm_layer(self.norm_cfg, planes * block.expansion)[1])
+
+        layers = []
+        layers.append(
+            block(
+                inplanes,
+                planes,
+                stride,
+                downsample=downsample,
+                with_cp=self.with_cp,
+                norm_cfg=self.norm_cfg,
+                conv_cfg=self.conv_cfg))
+        inplanes = planes * block.expansion
+        for i in range(1, blocks):
+            layers.append(
+                block(
+                    inplanes,
+                    planes,
+                    with_cp=self.with_cp,
+                    norm_cfg=self.norm_cfg,
+                    conv_cfg=self.conv_cfg))
+
+        return nn.Sequential(*layers)
+
+    def _make_stage(self, layer_config, in_channels, multiscale_output=True):
+        num_modules = layer_config['num_modules']
+        num_branches = layer_config['num_branches']
+        num_blocks = layer_config['num_blocks']
+        num_channels = layer_config['num_channels']
+        block = self.blocks_dict[layer_config['block']]
+
+        hr_modules = []
+        for i in range(num_modules):
+            # multi_scale_output is only used for the last module
+            if not multiscale_output and i == num_modules - 1:
+                reset_multiscale_output = False
+            else:
+                reset_multiscale_output = True
+
+            hr_modules.append(
+                HRModule(
+                    num_branches,
+                    block,
+                    num_blocks,
+                    in_channels,
+                    num_channels,
+                    reset_multiscale_output,
+                    with_cp=self.with_cp,
+                    norm_cfg=self.norm_cfg,
+                    conv_cfg=self.conv_cfg))
+
+        return nn.Sequential(*hr_modules), in_channels
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in backbone.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        if isinstance(pretrained, str):
+            logger = get_root_logger()
+            load_checkpoint(self, pretrained, strict=False, logger=logger)
+        elif pretrained is None:
+            for m in self.modules():
+                if isinstance(m, nn.Conv2d):
+                    kaiming_init(m)
+                elif isinstance(m, (_BatchNorm, nn.GroupNorm)):
+                    constant_init(m, 1)
+
+            if self.zero_init_residual:
+                for m in self.modules():
+                    if isinstance(m, Bottleneck):
+                        constant_init(m.norm3, 0)
+                    elif isinstance(m, BasicBlock):
+                        constant_init(m.norm2, 0)
+        else:
+            raise TypeError('pretrained must be a str or None')
+
+    def forward(self, x):
+        """Forward function."""
+        x = self.conv1(x)
+        x = self.norm1(x)
+        x = self.relu(x)
+        x = self.conv2(x)
+        x = self.norm2(x)
+        x = self.relu(x)
+        x = self.layer1(x)
+
+        x_list = []
+        for i in range(self.stage2_cfg['num_branches']):
+            if self.transition1[i] is not None:
+                x_list.append(self.transition1[i](x))
+            else:
+                x_list.append(x)
+        y_list = self.stage2(x_list)
+
+        x_list = []
+        for i in range(self.stage3_cfg['num_branches']):
+            if self.transition2[i] is not None:
+                x_list.append(self.transition2[i](y_list[-1]))
+            else:
+                x_list.append(y_list[i])
+        y_list = self.stage3(x_list)
+
+        x_list = []
+        for i in range(self.stage4_cfg['num_branches']):
+            if self.transition3[i] is not None:
+                x_list.append(self.transition3[i](y_list[-1]))
+            else:
+                x_list.append(y_list[i])
+        y_list = self.stage4(x_list)
+
+        return y_list
+
+    def train(self, mode=True):
+        """Convert the model into training mode will keeping the normalization
+        layer freezed."""
+        super(HRNet, self).train(mode)
+        if mode and self.norm_eval:
+            for m in self.modules():
+                # trick: eval have effect on BatchNorm only
+                if isinstance(m, _BatchNorm):
+                    m.eval()
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/regnet.py b/python/autoCut/models/swin/mmdet/models/backbones/regnet.py
new file mode 100644
index 0000000..91a602a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/regnet.py
@@ -0,0 +1,325 @@
+import numpy as np
+import torch.nn as nn
+from mmcv.cnn import build_conv_layer, build_norm_layer
+
+from ..builder import BACKBONES
+from .resnet import ResNet
+from .resnext import Bottleneck
+
+
+@BACKBONES.register_module()
+class RegNet(ResNet):
+    """RegNet backbone.
+
+    More details can be found in `paper <https://arxiv.org/abs/2003.13678>`_ .
+
+    Args:
+        arch (dict): The parameter of RegNets.
+
+            - w0 (int): initial width
+            - wa (float): slope of width
+            - wm (float): quantization parameter to quantize the width
+            - depth (int): depth of the backbone
+            - group_w (int): width of group
+            - bot_mul (float): bottleneck ratio, i.e. expansion of bottleneck.
+        strides (Sequence[int]): Strides of the first block of each stage.
+        base_channels (int): Base channels after stem layer.
+        in_channels (int): Number of input image channels. Default: 3.
+        dilations (Sequence[int]): Dilation of each stage.
+        out_indices (Sequence[int]): Output from which stages.
+        style (str): `pytorch` or `caffe`. If set to "pytorch", the stride-two
+            layer is the 3x3 conv layer, otherwise the stride-two layer is
+            the first 1x1 conv layer.
+        frozen_stages (int): Stages to be frozen (all param fixed). -1 means
+            not freezing any parameters.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+        norm_eval (bool): Whether to set norm layers to eval mode, namely,
+            freeze running stats (mean and var). Note: Effect on Batch Norm
+            and its variants only.
+        with_cp (bool): Use checkpoint or not. Using checkpoint will save some
+            memory while slowing down the training speed.
+        zero_init_residual (bool): whether to use zero init for last norm layer
+            in resblocks to let them behave as identity.
+
+    Example:
+        >>> from mmdet.models import RegNet
+        >>> import torch
+        >>> self = RegNet(
+                arch=dict(
+                    w0=88,
+                    wa=26.31,
+                    wm=2.25,
+                    group_w=48,
+                    depth=25,
+                    bot_mul=1.0))
+        >>> self.eval()
+        >>> inputs = torch.rand(1, 3, 32, 32)
+        >>> level_outputs = self.forward(inputs)
+        >>> for level_out in level_outputs:
+        ...     print(tuple(level_out.shape))
+        (1, 96, 8, 8)
+        (1, 192, 4, 4)
+        (1, 432, 2, 2)
+        (1, 1008, 1, 1)
+    """
+    arch_settings = {
+        'regnetx_400mf':
+        dict(w0=24, wa=24.48, wm=2.54, group_w=16, depth=22, bot_mul=1.0),
+        'regnetx_800mf':
+        dict(w0=56, wa=35.73, wm=2.28, group_w=16, depth=16, bot_mul=1.0),
+        'regnetx_1.6gf':
+        dict(w0=80, wa=34.01, wm=2.25, group_w=24, depth=18, bot_mul=1.0),
+        'regnetx_3.2gf':
+        dict(w0=88, wa=26.31, wm=2.25, group_w=48, depth=25, bot_mul=1.0),
+        'regnetx_4.0gf':
+        dict(w0=96, wa=38.65, wm=2.43, group_w=40, depth=23, bot_mul=1.0),
+        'regnetx_6.4gf':
+        dict(w0=184, wa=60.83, wm=2.07, group_w=56, depth=17, bot_mul=1.0),
+        'regnetx_8.0gf':
+        dict(w0=80, wa=49.56, wm=2.88, group_w=120, depth=23, bot_mul=1.0),
+        'regnetx_12gf':
+        dict(w0=168, wa=73.36, wm=2.37, group_w=112, depth=19, bot_mul=1.0),
+    }
+
+    def __init__(self,
+                 arch,
+                 in_channels=3,
+                 stem_channels=32,
+                 base_channels=32,
+                 strides=(2, 2, 2, 2),
+                 dilations=(1, 1, 1, 1),
+                 out_indices=(0, 1, 2, 3),
+                 style='pytorch',
+                 deep_stem=False,
+                 avg_down=False,
+                 frozen_stages=-1,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN', requires_grad=True),
+                 norm_eval=True,
+                 dcn=None,
+                 stage_with_dcn=(False, False, False, False),
+                 plugins=None,
+                 with_cp=False,
+                 zero_init_residual=True):
+        super(ResNet, self).__init__()
+
+        # Generate RegNet parameters first
+        if isinstance(arch, str):
+            assert arch in self.arch_settings, \
+                f'"arch": "{arch}" is not one of the' \
+                ' arch_settings'
+            arch = self.arch_settings[arch]
+        elif not isinstance(arch, dict):
+            raise ValueError('Expect "arch" to be either a string '
+                             f'or a dict, got {type(arch)}')
+
+        widths, num_stages = self.generate_regnet(
+            arch['w0'],
+            arch['wa'],
+            arch['wm'],
+            arch['depth'],
+        )
+        # Convert to per stage format
+        stage_widths, stage_blocks = self.get_stages_from_blocks(widths)
+        # Generate group widths and bot muls
+        group_widths = [arch['group_w'] for _ in range(num_stages)]
+        self.bottleneck_ratio = [arch['bot_mul'] for _ in range(num_stages)]
+        # Adjust the compatibility of stage_widths and group_widths
+        stage_widths, group_widths = self.adjust_width_group(
+            stage_widths, self.bottleneck_ratio, group_widths)
+
+        # Group params by stage
+        self.stage_widths = stage_widths
+        self.group_widths = group_widths
+        self.depth = sum(stage_blocks)
+        self.stem_channels = stem_channels
+        self.base_channels = base_channels
+        self.num_stages = num_stages
+        assert num_stages >= 1 and num_stages <= 4
+        self.strides = strides
+        self.dilations = dilations
+        assert len(strides) == len(dilations) == num_stages
+        self.out_indices = out_indices
+        assert max(out_indices) < num_stages
+        self.style = style
+        self.deep_stem = deep_stem
+        self.avg_down = avg_down
+        self.frozen_stages = frozen_stages
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.with_cp = with_cp
+        self.norm_eval = norm_eval
+        self.dcn = dcn
+        self.stage_with_dcn = stage_with_dcn
+        if dcn is not None:
+            assert len(stage_with_dcn) == num_stages
+        self.plugins = plugins
+        self.zero_init_residual = zero_init_residual
+        self.block = Bottleneck
+        expansion_bak = self.block.expansion
+        self.block.expansion = 1
+        self.stage_blocks = stage_blocks[:num_stages]
+
+        self._make_stem_layer(in_channels, stem_channels)
+
+        self.inplanes = stem_channels
+        self.res_layers = []
+        for i, num_blocks in enumerate(self.stage_blocks):
+            stride = self.strides[i]
+            dilation = self.dilations[i]
+            group_width = self.group_widths[i]
+            width = int(round(self.stage_widths[i] * self.bottleneck_ratio[i]))
+            stage_groups = width // group_width
+
+            dcn = self.dcn if self.stage_with_dcn[i] else None
+            if self.plugins is not None:
+                stage_plugins = self.make_stage_plugins(self.plugins, i)
+            else:
+                stage_plugins = None
+
+            res_layer = self.make_res_layer(
+                block=self.block,
+                inplanes=self.inplanes,
+                planes=self.stage_widths[i],
+                num_blocks=num_blocks,
+                stride=stride,
+                dilation=dilation,
+                style=self.style,
+                avg_down=self.avg_down,
+                with_cp=self.with_cp,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg,
+                dcn=dcn,
+                plugins=stage_plugins,
+                groups=stage_groups,
+                base_width=group_width,
+                base_channels=self.stage_widths[i])
+            self.inplanes = self.stage_widths[i]
+            layer_name = f'layer{i + 1}'
+            self.add_module(layer_name, res_layer)
+            self.res_layers.append(layer_name)
+
+        self._freeze_stages()
+
+        self.feat_dim = stage_widths[-1]
+        self.block.expansion = expansion_bak
+
+    def _make_stem_layer(self, in_channels, base_channels):
+        self.conv1 = build_conv_layer(
+            self.conv_cfg,
+            in_channels,
+            base_channels,
+            kernel_size=3,
+            stride=2,
+            padding=1,
+            bias=False)
+        self.norm1_name, norm1 = build_norm_layer(
+            self.norm_cfg, base_channels, postfix=1)
+        self.add_module(self.norm1_name, norm1)
+        self.relu = nn.ReLU(inplace=True)
+
+    def generate_regnet(self,
+                        initial_width,
+                        width_slope,
+                        width_parameter,
+                        depth,
+                        divisor=8):
+        """Generates per block width from RegNet parameters.
+
+        Args:
+            initial_width ([int]): Initial width of the backbone
+            width_slope ([float]): Slope of the quantized linear function
+            width_parameter ([int]): Parameter used to quantize the width.
+            depth ([int]): Depth of the backbone.
+            divisor (int, optional): The divisor of channels. Defaults to 8.
+
+        Returns:
+            list, int: return a list of widths of each stage and the number \
+                of stages
+        """
+        assert width_slope >= 0
+        assert initial_width > 0
+        assert width_parameter > 1
+        assert initial_width % divisor == 0
+        widths_cont = np.arange(depth) * width_slope + initial_width
+        ks = np.round(
+            np.log(widths_cont / initial_width) / np.log(width_parameter))
+        widths = initial_width * np.power(width_parameter, ks)
+        widths = np.round(np.divide(widths, divisor)) * divisor
+        num_stages = len(np.unique(widths))
+        widths, widths_cont = widths.astype(int).tolist(), widths_cont.tolist()
+        return widths, num_stages
+
+    @staticmethod
+    def quantize_float(number, divisor):
+        """Converts a float to closest non-zero int divisible by divisor.
+
+        Args:
+            number (int): Original number to be quantized.
+            divisor (int): Divisor used to quantize the number.
+
+        Returns:
+            int: quantized number that is divisible by devisor.
+        """
+        return int(round(number / divisor) * divisor)
+
+    def adjust_width_group(self, widths, bottleneck_ratio, groups):
+        """Adjusts the compatibility of widths and groups.
+
+        Args:
+            widths (list[int]): Width of each stage.
+            bottleneck_ratio (float): Bottleneck ratio.
+            groups (int): number of groups in each stage
+
+        Returns:
+            tuple(list): The adjusted widths and groups of each stage.
+        """
+        bottleneck_width = [
+            int(w * b) for w, b in zip(widths, bottleneck_ratio)
+        ]
+        groups = [min(g, w_bot) for g, w_bot in zip(groups, bottleneck_width)]
+        bottleneck_width = [
+            self.quantize_float(w_bot, g)
+            for w_bot, g in zip(bottleneck_width, groups)
+        ]
+        widths = [
+            int(w_bot / b)
+            for w_bot, b in zip(bottleneck_width, bottleneck_ratio)
+        ]
+        return widths, groups
+
+    def get_stages_from_blocks(self, widths):
+        """Gets widths/stage_blocks of network at each stage.
+
+        Args:
+            widths (list[int]): Width in each stage.
+
+        Returns:
+            tuple(list): width and depth of each stage
+        """
+        width_diff = [
+            width != width_prev
+            for width, width_prev in zip(widths + [0], [0] + widths)
+        ]
+        stage_widths = [
+            width for width, diff in zip(widths, width_diff[:-1]) if diff
+        ]
+        stage_blocks = np.diff([
+            depth for depth, diff in zip(range(len(width_diff)), width_diff)
+            if diff
+        ]).tolist()
+        return stage_widths, stage_blocks
+
+    def forward(self, x):
+        """Forward function."""
+        x = self.conv1(x)
+        x = self.norm1(x)
+        x = self.relu(x)
+
+        outs = []
+        for i, layer_name in enumerate(self.res_layers):
+            res_layer = getattr(self, layer_name)
+            x = res_layer(x)
+            if i in self.out_indices:
+                outs.append(x)
+        return tuple(outs)
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/res2net.py b/python/autoCut/models/swin/mmdet/models/backbones/res2net.py
new file mode 100644
index 0000000..7901b7f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/res2net.py
@@ -0,0 +1,351 @@
+import math
+
+import torch
+import torch.nn as nn
+import torch.utils.checkpoint as cp
+from mmcv.cnn import (build_conv_layer, build_norm_layer, constant_init,
+                      kaiming_init)
+from mmcv.runner import load_checkpoint
+from torch.nn.modules.batchnorm import _BatchNorm
+
+from mmdet.utils import get_root_logger
+from ..builder import BACKBONES
+from .resnet import Bottleneck as _Bottleneck
+from .resnet import ResNet
+
+
+class Bottle2neck(_Bottleneck):
+    expansion = 4
+
+    def __init__(self,
+                 inplanes,
+                 planes,
+                 scales=4,
+                 base_width=26,
+                 base_channels=64,
+                 stage_type='normal',
+                 **kwargs):
+        """Bottle2neck block for Res2Net.
+
+        If style is "pytorch", the stride-two layer is the 3x3 conv layer, if
+        it is "caffe", the stride-two layer is the first 1x1 conv layer.
+        """
+        super(Bottle2neck, self).__init__(inplanes, planes, **kwargs)
+        assert scales > 1, 'Res2Net degenerates to ResNet when scales = 1.'
+        width = int(math.floor(self.planes * (base_width / base_channels)))
+
+        self.norm1_name, norm1 = build_norm_layer(
+            self.norm_cfg, width * scales, postfix=1)
+        self.norm3_name, norm3 = build_norm_layer(
+            self.norm_cfg, self.planes * self.expansion, postfix=3)
+
+        self.conv1 = build_conv_layer(
+            self.conv_cfg,
+            self.inplanes,
+            width * scales,
+            kernel_size=1,
+            stride=self.conv1_stride,
+            bias=False)
+        self.add_module(self.norm1_name, norm1)
+
+        if stage_type == 'stage' and self.conv2_stride != 1:
+            self.pool = nn.AvgPool2d(
+                kernel_size=3, stride=self.conv2_stride, padding=1)
+        convs = []
+        bns = []
+
+        fallback_on_stride = False
+        if self.with_dcn:
+            fallback_on_stride = self.dcn.pop('fallback_on_stride', False)
+        if not self.with_dcn or fallback_on_stride:
+            for i in range(scales - 1):
+                convs.append(
+                    build_conv_layer(
+                        self.conv_cfg,
+                        width,
+                        width,
+                        kernel_size=3,
+                        stride=self.conv2_stride,
+                        padding=self.dilation,
+                        dilation=self.dilation,
+                        bias=False))
+                bns.append(
+                    build_norm_layer(self.norm_cfg, width, postfix=i + 1)[1])
+            self.convs = nn.ModuleList(convs)
+            self.bns = nn.ModuleList(bns)
+        else:
+            assert self.conv_cfg is None, 'conv_cfg must be None for DCN'
+            for i in range(scales - 1):
+                convs.append(
+                    build_conv_layer(
+                        self.dcn,
+                        width,
+                        width,
+                        kernel_size=3,
+                        stride=self.conv2_stride,
+                        padding=self.dilation,
+                        dilation=self.dilation,
+                        bias=False))
+                bns.append(
+                    build_norm_layer(self.norm_cfg, width, postfix=i + 1)[1])
+            self.convs = nn.ModuleList(convs)
+            self.bns = nn.ModuleList(bns)
+
+        self.conv3 = build_conv_layer(
+            self.conv_cfg,
+            width * scales,
+            self.planes * self.expansion,
+            kernel_size=1,
+            bias=False)
+        self.add_module(self.norm3_name, norm3)
+
+        self.stage_type = stage_type
+        self.scales = scales
+        self.width = width
+        delattr(self, 'conv2')
+        delattr(self, self.norm2_name)
+
+    def forward(self, x):
+        """Forward function."""
+
+        def _inner_forward(x):
+            identity = x
+
+            out = self.conv1(x)
+            out = self.norm1(out)
+            out = self.relu(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv1_plugin_names)
+
+            spx = torch.split(out, self.width, 1)
+            sp = self.convs[0](spx[0].contiguous())
+            sp = self.relu(self.bns[0](sp))
+            out = sp
+            for i in range(1, self.scales - 1):
+                if self.stage_type == 'stage':
+                    sp = spx[i]
+                else:
+                    sp = sp + spx[i]
+                sp = self.convs[i](sp.contiguous())
+                sp = self.relu(self.bns[i](sp))
+                out = torch.cat((out, sp), 1)
+
+            if self.stage_type == 'normal' or self.conv2_stride == 1:
+                out = torch.cat((out, spx[self.scales - 1]), 1)
+            elif self.stage_type == 'stage':
+                out = torch.cat((out, self.pool(spx[self.scales - 1])), 1)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv2_plugin_names)
+
+            out = self.conv3(out)
+            out = self.norm3(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv3_plugin_names)
+
+            if self.downsample is not None:
+                identity = self.downsample(x)
+
+            out += identity
+
+            return out
+
+        if self.with_cp and x.requires_grad:
+            out = cp.checkpoint(_inner_forward, x)
+        else:
+            out = _inner_forward(x)
+
+        out = self.relu(out)
+
+        return out
+
+
+class Res2Layer(nn.Sequential):
+    """Res2Layer to build Res2Net style backbone.
+
+    Args:
+        block (nn.Module): block used to build ResLayer.
+        inplanes (int): inplanes of block.
+        planes (int): planes of block.
+        num_blocks (int): number of blocks.
+        stride (int): stride of the first block. Default: 1
+        avg_down (bool): Use AvgPool instead of stride conv when
+            downsampling in the bottle2neck. Default: False
+        conv_cfg (dict): dictionary to construct and config conv layer.
+            Default: None
+        norm_cfg (dict): dictionary to construct and config norm layer.
+            Default: dict(type='BN')
+        scales (int): Scales used in Res2Net. Default: 4
+        base_width (int): Basic width of each scale. Default: 26
+    """
+
+    def __init__(self,
+                 block,
+                 inplanes,
+                 planes,
+                 num_blocks,
+                 stride=1,
+                 avg_down=True,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 scales=4,
+                 base_width=26,
+                 **kwargs):
+        self.block = block
+
+        downsample = None
+        if stride != 1 or inplanes != planes * block.expansion:
+            downsample = nn.Sequential(
+                nn.AvgPool2d(
+                    kernel_size=stride,
+                    stride=stride,
+                    ceil_mode=True,
+                    count_include_pad=False),
+                build_conv_layer(
+                    conv_cfg,
+                    inplanes,
+                    planes * block.expansion,
+                    kernel_size=1,
+                    stride=1,
+                    bias=False),
+                build_norm_layer(norm_cfg, planes * block.expansion)[1],
+            )
+
+        layers = []
+        layers.append(
+            block(
+                inplanes=inplanes,
+                planes=planes,
+                stride=stride,
+                downsample=downsample,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                scales=scales,
+                base_width=base_width,
+                stage_type='stage',
+                **kwargs))
+        inplanes = planes * block.expansion
+        for i in range(1, num_blocks):
+            layers.append(
+                block(
+                    inplanes=inplanes,
+                    planes=planes,
+                    stride=1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg,
+                    scales=scales,
+                    base_width=base_width,
+                    **kwargs))
+        super(Res2Layer, self).__init__(*layers)
+
+
+@BACKBONES.register_module()
+class Res2Net(ResNet):
+    """Res2Net backbone.
+
+    Args:
+        scales (int): Scales used in Res2Net. Default: 4
+        base_width (int): Basic width of each scale. Default: 26
+        depth (int): Depth of res2net, from {50, 101, 152}.
+        in_channels (int): Number of input image channels. Default: 3.
+        num_stages (int): Res2net stages. Default: 4.
+        strides (Sequence[int]): Strides of the first block of each stage.
+        dilations (Sequence[int]): Dilation of each stage.
+        out_indices (Sequence[int]): Output from which stages.
+        style (str): `pytorch` or `caffe`. If set to "pytorch", the stride-two
+            layer is the 3x3 conv layer, otherwise the stride-two layer is
+            the first 1x1 conv layer.
+        deep_stem (bool): Replace 7x7 conv in input stem with 3 3x3 conv
+        avg_down (bool): Use AvgPool instead of stride conv when
+            downsampling in the bottle2neck.
+        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
+            -1 means not freezing any parameters.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+        norm_eval (bool): Whether to set norm layers to eval mode, namely,
+            freeze running stats (mean and var). Note: Effect on Batch Norm
+            and its variants only.
+        plugins (list[dict]): List of plugins for stages, each dict contains:
+
+            - cfg (dict, required): Cfg dict to build plugin.
+            - position (str, required): Position inside block to insert
+              plugin, options are 'after_conv1', 'after_conv2', 'after_conv3'.
+            - stages (tuple[bool], optional): Stages to apply plugin, length
+              should be same as 'num_stages'.
+        with_cp (bool): Use checkpoint or not. Using checkpoint will save some
+            memory while slowing down the training speed.
+        zero_init_residual (bool): Whether to use zero init for last norm layer
+            in resblocks to let them behave as identity.
+
+    Example:
+        >>> from mmdet.models import Res2Net
+        >>> import torch
+        >>> self = Res2Net(depth=50, scales=4, base_width=26)
+        >>> self.eval()
+        >>> inputs = torch.rand(1, 3, 32, 32)
+        >>> level_outputs = self.forward(inputs)
+        >>> for level_out in level_outputs:
+        ...     print(tuple(level_out.shape))
+        (1, 256, 8, 8)
+        (1, 512, 4, 4)
+        (1, 1024, 2, 2)
+        (1, 2048, 1, 1)
+    """
+
+    arch_settings = {
+        50: (Bottle2neck, (3, 4, 6, 3)),
+        101: (Bottle2neck, (3, 4, 23, 3)),
+        152: (Bottle2neck, (3, 8, 36, 3))
+    }
+
+    def __init__(self,
+                 scales=4,
+                 base_width=26,
+                 style='pytorch',
+                 deep_stem=True,
+                 avg_down=True,
+                 **kwargs):
+        self.scales = scales
+        self.base_width = base_width
+        super(Res2Net, self).__init__(
+            style='pytorch', deep_stem=True, avg_down=True, **kwargs)
+
+    def make_res_layer(self, **kwargs):
+        return Res2Layer(
+            scales=self.scales,
+            base_width=self.base_width,
+            base_channels=self.base_channels,
+            **kwargs)
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in backbone.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        if isinstance(pretrained, str):
+            logger = get_root_logger()
+            load_checkpoint(self, pretrained, strict=False, logger=logger)
+        elif pretrained is None:
+            for m in self.modules():
+                if isinstance(m, nn.Conv2d):
+                    kaiming_init(m)
+                elif isinstance(m, (_BatchNorm, nn.GroupNorm)):
+                    constant_init(m, 1)
+
+            if self.dcn is not None:
+                for m in self.modules():
+                    if isinstance(m, Bottle2neck):
+                        # dcn in Res2Net bottle2neck is in ModuleList
+                        for n in m.convs:
+                            if hasattr(n, 'conv_offset'):
+                                constant_init(n.conv_offset, 0)
+
+            if self.zero_init_residual:
+                for m in self.modules():
+                    if isinstance(m, Bottle2neck):
+                        constant_init(m.norm3, 0)
+        else:
+            raise TypeError('pretrained must be a str or None')
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/resnest.py b/python/autoCut/models/swin/mmdet/models/backbones/resnest.py
new file mode 100644
index 0000000..48e1d8b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/resnest.py
@@ -0,0 +1,317 @@
+import math
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.utils.checkpoint as cp
+from mmcv.cnn import build_conv_layer, build_norm_layer
+
+from ..builder import BACKBONES
+from ..utils import ResLayer
+from .resnet import Bottleneck as _Bottleneck
+from .resnet import ResNetV1d
+
+
+class RSoftmax(nn.Module):
+    """Radix Softmax module in ``SplitAttentionConv2d``.
+
+    Args:
+        radix (int): Radix of input.
+        groups (int): Groups of input.
+    """
+
+    def __init__(self, radix, groups):
+        super().__init__()
+        self.radix = radix
+        self.groups = groups
+
+    def forward(self, x):
+        batch = x.size(0)
+        if self.radix > 1:
+            x = x.view(batch, self.groups, self.radix, -1).transpose(1, 2)
+            x = F.softmax(x, dim=1)
+            x = x.reshape(batch, -1)
+        else:
+            x = torch.sigmoid(x)
+        return x
+
+
+class SplitAttentionConv2d(nn.Module):
+    """Split-Attention Conv2d in ResNeSt.
+
+    Args:
+        in_channels (int): Number of channels in the input feature map.
+        channels (int): Number of intermediate channels.
+        kernel_size (int | tuple[int]): Size of the convolution kernel.
+        stride (int | tuple[int]): Stride of the convolution.
+        padding (int | tuple[int]): Zero-padding added to both sides of
+        dilation (int | tuple[int]): Spacing between kernel elements.
+        groups (int): Number of blocked connections from input channels to
+            output channels.
+        groups (int): Same as nn.Conv2d.
+        radix (int): Radix of SpltAtConv2d. Default: 2
+        reduction_factor (int): Reduction factor of inter_channels. Default: 4.
+        conv_cfg (dict): Config dict for convolution layer. Default: None,
+            which means using conv2d.
+        norm_cfg (dict): Config dict for normalization layer. Default: None.
+        dcn (dict): Config dict for DCN. Default: None.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 channels,
+                 kernel_size,
+                 stride=1,
+                 padding=0,
+                 dilation=1,
+                 groups=1,
+                 radix=2,
+                 reduction_factor=4,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 dcn=None):
+        super(SplitAttentionConv2d, self).__init__()
+        inter_channels = max(in_channels * radix // reduction_factor, 32)
+        self.radix = radix
+        self.groups = groups
+        self.channels = channels
+        self.with_dcn = dcn is not None
+        self.dcn = dcn
+        fallback_on_stride = False
+        if self.with_dcn:
+            fallback_on_stride = self.dcn.pop('fallback_on_stride', False)
+        if self.with_dcn and not fallback_on_stride:
+            assert conv_cfg is None, 'conv_cfg must be None for DCN'
+            conv_cfg = dcn
+        self.conv = build_conv_layer(
+            conv_cfg,
+            in_channels,
+            channels * radix,
+            kernel_size,
+            stride=stride,
+            padding=padding,
+            dilation=dilation,
+            groups=groups * radix,
+            bias=False)
+        # To be consistent with original implementation, starting from 0
+        self.norm0_name, norm0 = build_norm_layer(
+            norm_cfg, channels * radix, postfix=0)
+        self.add_module(self.norm0_name, norm0)
+        self.relu = nn.ReLU(inplace=True)
+        self.fc1 = build_conv_layer(
+            None, channels, inter_channels, 1, groups=self.groups)
+        self.norm1_name, norm1 = build_norm_layer(
+            norm_cfg, inter_channels, postfix=1)
+        self.add_module(self.norm1_name, norm1)
+        self.fc2 = build_conv_layer(
+            None, inter_channels, channels * radix, 1, groups=self.groups)
+        self.rsoftmax = RSoftmax(radix, groups)
+
+    @property
+    def norm0(self):
+        """nn.Module: the normalization layer named "norm0" """
+        return getattr(self, self.norm0_name)
+
+    @property
+    def norm1(self):
+        """nn.Module: the normalization layer named "norm1" """
+        return getattr(self, self.norm1_name)
+
+    def forward(self, x):
+        x = self.conv(x)
+        x = self.norm0(x)
+        x = self.relu(x)
+
+        batch, rchannel = x.shape[:2]
+        batch = x.size(0)
+        if self.radix > 1:
+            splits = x.view(batch, self.radix, -1, *x.shape[2:])
+            gap = splits.sum(dim=1)
+        else:
+            gap = x
+        gap = F.adaptive_avg_pool2d(gap, 1)
+        gap = self.fc1(gap)
+
+        gap = self.norm1(gap)
+        gap = self.relu(gap)
+
+        atten = self.fc2(gap)
+        atten = self.rsoftmax(atten).view(batch, -1, 1, 1)
+
+        if self.radix > 1:
+            attens = atten.view(batch, self.radix, -1, *atten.shape[2:])
+            out = torch.sum(attens * splits, dim=1)
+        else:
+            out = atten * x
+        return out.contiguous()
+
+
+class Bottleneck(_Bottleneck):
+    """Bottleneck block for ResNeSt.
+
+    Args:
+        inplane (int): Input planes of this block.
+        planes (int): Middle planes of this block.
+        groups (int): Groups of conv2.
+        base_width (int): Base of width in terms of base channels. Default: 4.
+        base_channels (int): Base of channels for calculating width.
+            Default: 64.
+        radix (int): Radix of SpltAtConv2d. Default: 2
+        reduction_factor (int): Reduction factor of inter_channels in
+            SplitAttentionConv2d. Default: 4.
+        avg_down_stride (bool): Whether to use average pool for stride in
+            Bottleneck. Default: True.
+        kwargs (dict): Key word arguments for base class.
+    """
+    expansion = 4
+
+    def __init__(self,
+                 inplanes,
+                 planes,
+                 groups=1,
+                 base_width=4,
+                 base_channels=64,
+                 radix=2,
+                 reduction_factor=4,
+                 avg_down_stride=True,
+                 **kwargs):
+        """Bottleneck block for ResNeSt."""
+        super(Bottleneck, self).__init__(inplanes, planes, **kwargs)
+
+        if groups == 1:
+            width = self.planes
+        else:
+            width = math.floor(self.planes *
+                               (base_width / base_channels)) * groups
+
+        self.avg_down_stride = avg_down_stride and self.conv2_stride > 1
+
+        self.norm1_name, norm1 = build_norm_layer(
+            self.norm_cfg, width, postfix=1)
+        self.norm3_name, norm3 = build_norm_layer(
+            self.norm_cfg, self.planes * self.expansion, postfix=3)
+
+        self.conv1 = build_conv_layer(
+            self.conv_cfg,
+            self.inplanes,
+            width,
+            kernel_size=1,
+            stride=self.conv1_stride,
+            bias=False)
+        self.add_module(self.norm1_name, norm1)
+        self.with_modulated_dcn = False
+        self.conv2 = SplitAttentionConv2d(
+            width,
+            width,
+            kernel_size=3,
+            stride=1 if self.avg_down_stride else self.conv2_stride,
+            padding=self.dilation,
+            dilation=self.dilation,
+            groups=groups,
+            radix=radix,
+            reduction_factor=reduction_factor,
+            conv_cfg=self.conv_cfg,
+            norm_cfg=self.norm_cfg,
+            dcn=self.dcn)
+        delattr(self, self.norm2_name)
+
+        if self.avg_down_stride:
+            self.avd_layer = nn.AvgPool2d(3, self.conv2_stride, padding=1)
+
+        self.conv3 = build_conv_layer(
+            self.conv_cfg,
+            width,
+            self.planes * self.expansion,
+            kernel_size=1,
+            bias=False)
+        self.add_module(self.norm3_name, norm3)
+
+    def forward(self, x):
+
+        def _inner_forward(x):
+            identity = x
+
+            out = self.conv1(x)
+            out = self.norm1(out)
+            out = self.relu(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv1_plugin_names)
+
+            out = self.conv2(out)
+
+            if self.avg_down_stride:
+                out = self.avd_layer(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv2_plugin_names)
+
+            out = self.conv3(out)
+            out = self.norm3(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv3_plugin_names)
+
+            if self.downsample is not None:
+                identity = self.downsample(x)
+
+            out += identity
+
+            return out
+
+        if self.with_cp and x.requires_grad:
+            out = cp.checkpoint(_inner_forward, x)
+        else:
+            out = _inner_forward(x)
+
+        out = self.relu(out)
+
+        return out
+
+
+@BACKBONES.register_module()
+class ResNeSt(ResNetV1d):
+    """ResNeSt backbone.
+
+    Args:
+        groups (int): Number of groups of Bottleneck. Default: 1
+        base_width (int): Base width of Bottleneck. Default: 4
+        radix (int): Radix of SplitAttentionConv2d. Default: 2
+        reduction_factor (int): Reduction factor of inter_channels in
+            SplitAttentionConv2d. Default: 4.
+        avg_down_stride (bool): Whether to use average pool for stride in
+            Bottleneck. Default: True.
+        kwargs (dict): Keyword arguments for ResNet.
+    """
+
+    arch_settings = {
+        50: (Bottleneck, (3, 4, 6, 3)),
+        101: (Bottleneck, (3, 4, 23, 3)),
+        152: (Bottleneck, (3, 8, 36, 3)),
+        200: (Bottleneck, (3, 24, 36, 3))
+    }
+
+    def __init__(self,
+                 groups=1,
+                 base_width=4,
+                 radix=2,
+                 reduction_factor=4,
+                 avg_down_stride=True,
+                 **kwargs):
+        self.groups = groups
+        self.base_width = base_width
+        self.radix = radix
+        self.reduction_factor = reduction_factor
+        self.avg_down_stride = avg_down_stride
+        super(ResNeSt, self).__init__(**kwargs)
+
+    def make_res_layer(self, **kwargs):
+        """Pack all blocks in a stage into a ``ResLayer``."""
+        return ResLayer(
+            groups=self.groups,
+            base_width=self.base_width,
+            base_channels=self.base_channels,
+            radix=self.radix,
+            reduction_factor=self.reduction_factor,
+            avg_down_stride=self.avg_down_stride,
+            **kwargs)
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/resnet.py b/python/autoCut/models/swin/mmdet/models/backbones/resnet.py
new file mode 100644
index 0000000..3826815
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/resnet.py
@@ -0,0 +1,663 @@
+import torch.nn as nn
+import torch.utils.checkpoint as cp
+from mmcv.cnn import (build_conv_layer, build_norm_layer, build_plugin_layer,
+                      constant_init, kaiming_init)
+from mmcv.runner import load_checkpoint
+from torch.nn.modules.batchnorm import _BatchNorm
+
+from mmdet.utils import get_root_logger
+from ..builder import BACKBONES
+from ..utils import ResLayer
+
+
+class BasicBlock(nn.Module):
+    expansion = 1
+
+    def __init__(self,
+                 inplanes,
+                 planes,
+                 stride=1,
+                 dilation=1,
+                 downsample=None,
+                 style='pytorch',
+                 with_cp=False,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 dcn=None,
+                 plugins=None):
+        super(BasicBlock, self).__init__()
+        assert dcn is None, 'Not implemented yet.'
+        assert plugins is None, 'Not implemented yet.'
+
+        self.norm1_name, norm1 = build_norm_layer(norm_cfg, planes, postfix=1)
+        self.norm2_name, norm2 = build_norm_layer(norm_cfg, planes, postfix=2)
+
+        self.conv1 = build_conv_layer(
+            conv_cfg,
+            inplanes,
+            planes,
+            3,
+            stride=stride,
+            padding=dilation,
+            dilation=dilation,
+            bias=False)
+        self.add_module(self.norm1_name, norm1)
+        self.conv2 = build_conv_layer(
+            conv_cfg, planes, planes, 3, padding=1, bias=False)
+        self.add_module(self.norm2_name, norm2)
+
+        self.relu = nn.ReLU(inplace=True)
+        self.downsample = downsample
+        self.stride = stride
+        self.dilation = dilation
+        self.with_cp = with_cp
+
+    @property
+    def norm1(self):
+        """nn.Module: normalization layer after the first convolution layer"""
+        return getattr(self, self.norm1_name)
+
+    @property
+    def norm2(self):
+        """nn.Module: normalization layer after the second convolution layer"""
+        return getattr(self, self.norm2_name)
+
+    def forward(self, x):
+        """Forward function."""
+
+        def _inner_forward(x):
+            identity = x
+
+            out = self.conv1(x)
+            out = self.norm1(out)
+            out = self.relu(out)
+
+            out = self.conv2(out)
+            out = self.norm2(out)
+
+            if self.downsample is not None:
+                identity = self.downsample(x)
+
+            out += identity
+
+            return out
+
+        if self.with_cp and x.requires_grad:
+            out = cp.checkpoint(_inner_forward, x)
+        else:
+            out = _inner_forward(x)
+
+        out = self.relu(out)
+
+        return out
+
+
+class Bottleneck(nn.Module):
+    expansion = 4
+
+    def __init__(self,
+                 inplanes,
+                 planes,
+                 stride=1,
+                 dilation=1,
+                 downsample=None,
+                 style='pytorch',
+                 with_cp=False,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 dcn=None,
+                 plugins=None):
+        """Bottleneck block for ResNet.
+
+        If style is "pytorch", the stride-two layer is the 3x3 conv layer, if
+        it is "caffe", the stride-two layer is the first 1x1 conv layer.
+        """
+        super(Bottleneck, self).__init__()
+        assert style in ['pytorch', 'caffe']
+        assert dcn is None or isinstance(dcn, dict)
+        assert plugins is None or isinstance(plugins, list)
+        if plugins is not None:
+            allowed_position = ['after_conv1', 'after_conv2', 'after_conv3']
+            assert all(p['position'] in allowed_position for p in plugins)
+
+        self.inplanes = inplanes
+        self.planes = planes
+        self.stride = stride
+        self.dilation = dilation
+        self.style = style
+        self.with_cp = with_cp
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.dcn = dcn
+        self.with_dcn = dcn is not None
+        self.plugins = plugins
+        self.with_plugins = plugins is not None
+
+        if self.with_plugins:
+            # collect plugins for conv1/conv2/conv3
+            self.after_conv1_plugins = [
+                plugin['cfg'] for plugin in plugins
+                if plugin['position'] == 'after_conv1'
+            ]
+            self.after_conv2_plugins = [
+                plugin['cfg'] for plugin in plugins
+                if plugin['position'] == 'after_conv2'
+            ]
+            self.after_conv3_plugins = [
+                plugin['cfg'] for plugin in plugins
+                if plugin['position'] == 'after_conv3'
+            ]
+
+        if self.style == 'pytorch':
+            self.conv1_stride = 1
+            self.conv2_stride = stride
+        else:
+            self.conv1_stride = stride
+            self.conv2_stride = 1
+
+        self.norm1_name, norm1 = build_norm_layer(norm_cfg, planes, postfix=1)
+        self.norm2_name, norm2 = build_norm_layer(norm_cfg, planes, postfix=2)
+        self.norm3_name, norm3 = build_norm_layer(
+            norm_cfg, planes * self.expansion, postfix=3)
+
+        self.conv1 = build_conv_layer(
+            conv_cfg,
+            inplanes,
+            planes,
+            kernel_size=1,
+            stride=self.conv1_stride,
+            bias=False)
+        self.add_module(self.norm1_name, norm1)
+        fallback_on_stride = False
+        if self.with_dcn:
+            fallback_on_stride = dcn.pop('fallback_on_stride', False)
+        if not self.with_dcn or fallback_on_stride:
+            self.conv2 = build_conv_layer(
+                conv_cfg,
+                planes,
+                planes,
+                kernel_size=3,
+                stride=self.conv2_stride,
+                padding=dilation,
+                dilation=dilation,
+                bias=False)
+        else:
+            assert self.conv_cfg is None, 'conv_cfg must be None for DCN'
+            self.conv2 = build_conv_layer(
+                dcn,
+                planes,
+                planes,
+                kernel_size=3,
+                stride=self.conv2_stride,
+                padding=dilation,
+                dilation=dilation,
+                bias=False)
+
+        self.add_module(self.norm2_name, norm2)
+        self.conv3 = build_conv_layer(
+            conv_cfg,
+            planes,
+            planes * self.expansion,
+            kernel_size=1,
+            bias=False)
+        self.add_module(self.norm3_name, norm3)
+
+        self.relu = nn.ReLU(inplace=True)
+        self.downsample = downsample
+
+        if self.with_plugins:
+            self.after_conv1_plugin_names = self.make_block_plugins(
+                planes, self.after_conv1_plugins)
+            self.after_conv2_plugin_names = self.make_block_plugins(
+                planes, self.after_conv2_plugins)
+            self.after_conv3_plugin_names = self.make_block_plugins(
+                planes * self.expansion, self.after_conv3_plugins)
+
+    def make_block_plugins(self, in_channels, plugins):
+        """make plugins for block.
+
+        Args:
+            in_channels (int): Input channels of plugin.
+            plugins (list[dict]): List of plugins cfg to build.
+
+        Returns:
+            list[str]: List of the names of plugin.
+        """
+        assert isinstance(plugins, list)
+        plugin_names = []
+        for plugin in plugins:
+            plugin = plugin.copy()
+            name, layer = build_plugin_layer(
+                plugin,
+                in_channels=in_channels,
+                postfix=plugin.pop('postfix', ''))
+            assert not hasattr(self, name), f'duplicate plugin {name}'
+            self.add_module(name, layer)
+            plugin_names.append(name)
+        return plugin_names
+
+    def forward_plugin(self, x, plugin_names):
+        out = x
+        for name in plugin_names:
+            out = getattr(self, name)(x)
+        return out
+
+    @property
+    def norm1(self):
+        """nn.Module: normalization layer after the first convolution layer"""
+        return getattr(self, self.norm1_name)
+
+    @property
+    def norm2(self):
+        """nn.Module: normalization layer after the second convolution layer"""
+        return getattr(self, self.norm2_name)
+
+    @property
+    def norm3(self):
+        """nn.Module: normalization layer after the third convolution layer"""
+        return getattr(self, self.norm3_name)
+
+    def forward(self, x):
+        """Forward function."""
+
+        def _inner_forward(x):
+            identity = x
+            out = self.conv1(x)
+            out = self.norm1(out)
+            out = self.relu(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv1_plugin_names)
+
+            out = self.conv2(out)
+            out = self.norm2(out)
+            out = self.relu(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv2_plugin_names)
+
+            out = self.conv3(out)
+            out = self.norm3(out)
+
+            if self.with_plugins:
+                out = self.forward_plugin(out, self.after_conv3_plugin_names)
+
+            if self.downsample is not None:
+                identity = self.downsample(x)
+
+            out += identity
+
+            return out
+
+        if self.with_cp and x.requires_grad:
+            out = cp.checkpoint(_inner_forward, x)
+        else:
+            out = _inner_forward(x)
+
+        out = self.relu(out)
+
+        return out
+
+
+@BACKBONES.register_module()
+class ResNet(nn.Module):
+    """ResNet backbone.
+
+    Args:
+        depth (int): Depth of resnet, from {18, 34, 50, 101, 152}.
+        stem_channels (int | None): Number of stem channels. If not specified,
+            it will be the same as `base_channels`. Default: None.
+        base_channels (int): Number of base channels of res layer. Default: 64.
+        in_channels (int): Number of input image channels. Default: 3.
+        num_stages (int): Resnet stages. Default: 4.
+        strides (Sequence[int]): Strides of the first block of each stage.
+        dilations (Sequence[int]): Dilation of each stage.
+        out_indices (Sequence[int]): Output from which stages.
+        style (str): `pytorch` or `caffe`. If set to "pytorch", the stride-two
+            layer is the 3x3 conv layer, otherwise the stride-two layer is
+            the first 1x1 conv layer.
+        deep_stem (bool): Replace 7x7 conv in input stem with 3 3x3 conv
+        avg_down (bool): Use AvgPool instead of stride conv when
+            downsampling in the bottleneck.
+        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
+            -1 means not freezing any parameters.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+        norm_eval (bool): Whether to set norm layers to eval mode, namely,
+            freeze running stats (mean and var). Note: Effect on Batch Norm
+            and its variants only.
+        plugins (list[dict]): List of plugins for stages, each dict contains:
+
+            - cfg (dict, required): Cfg dict to build plugin.
+            - position (str, required): Position inside block to insert
+              plugin, options are 'after_conv1', 'after_conv2', 'after_conv3'.
+            - stages (tuple[bool], optional): Stages to apply plugin, length
+              should be same as 'num_stages'.
+        with_cp (bool): Use checkpoint or not. Using checkpoint will save some
+            memory while slowing down the training speed.
+        zero_init_residual (bool): Whether to use zero init for last norm layer
+            in resblocks to let them behave as identity.
+
+    Example:
+        >>> from mmdet.models import ResNet
+        >>> import torch
+        >>> self = ResNet(depth=18)
+        >>> self.eval()
+        >>> inputs = torch.rand(1, 3, 32, 32)
+        >>> level_outputs = self.forward(inputs)
+        >>> for level_out in level_outputs:
+        ...     print(tuple(level_out.shape))
+        (1, 64, 8, 8)
+        (1, 128, 4, 4)
+        (1, 256, 2, 2)
+        (1, 512, 1, 1)
+    """
+
+    arch_settings = {
+        18: (BasicBlock, (2, 2, 2, 2)),
+        34: (BasicBlock, (3, 4, 6, 3)),
+        50: (Bottleneck, (3, 4, 6, 3)),
+        101: (Bottleneck, (3, 4, 23, 3)),
+        152: (Bottleneck, (3, 8, 36, 3))
+    }
+
+    def __init__(self,
+                 depth,
+                 in_channels=3,
+                 stem_channels=None,
+                 base_channels=64,
+                 num_stages=4,
+                 strides=(1, 2, 2, 2),
+                 dilations=(1, 1, 1, 1),
+                 out_indices=(0, 1, 2, 3),
+                 style='pytorch',
+                 deep_stem=False,
+                 avg_down=False,
+                 frozen_stages=-1,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN', requires_grad=True),
+                 norm_eval=True,
+                 dcn=None,
+                 stage_with_dcn=(False, False, False, False),
+                 plugins=None,
+                 with_cp=False,
+                 zero_init_residual=True):
+        super(ResNet, self).__init__()
+        if depth not in self.arch_settings:
+            raise KeyError(f'invalid depth {depth} for resnet')
+        self.depth = depth
+        if stem_channels is None:
+            stem_channels = base_channels
+        self.stem_channels = stem_channels
+        self.base_channels = base_channels
+        self.num_stages = num_stages
+        assert num_stages >= 1 and num_stages <= 4
+        self.strides = strides
+        self.dilations = dilations
+        assert len(strides) == len(dilations) == num_stages
+        self.out_indices = out_indices
+        assert max(out_indices) < num_stages
+        self.style = style
+        self.deep_stem = deep_stem
+        self.avg_down = avg_down
+        self.frozen_stages = frozen_stages
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.with_cp = with_cp
+        self.norm_eval = norm_eval
+        self.dcn = dcn
+        self.stage_with_dcn = stage_with_dcn
+        if dcn is not None:
+            assert len(stage_with_dcn) == num_stages
+        self.plugins = plugins
+        self.zero_init_residual = zero_init_residual
+        self.block, stage_blocks = self.arch_settings[depth]
+        self.stage_blocks = stage_blocks[:num_stages]
+        self.inplanes = stem_channels
+
+        self._make_stem_layer(in_channels, stem_channels)
+
+        self.res_layers = []
+        for i, num_blocks in enumerate(self.stage_blocks):
+            stride = strides[i]
+            dilation = dilations[i]
+            dcn = self.dcn if self.stage_with_dcn[i] else None
+            if plugins is not None:
+                stage_plugins = self.make_stage_plugins(plugins, i)
+            else:
+                stage_plugins = None
+            planes = base_channels * 2**i
+            res_layer = self.make_res_layer(
+                block=self.block,
+                inplanes=self.inplanes,
+                planes=planes,
+                num_blocks=num_blocks,
+                stride=stride,
+                dilation=dilation,
+                style=self.style,
+                avg_down=self.avg_down,
+                with_cp=with_cp,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                dcn=dcn,
+                plugins=stage_plugins)
+            self.inplanes = planes * self.block.expansion
+            layer_name = f'layer{i + 1}'
+            self.add_module(layer_name, res_layer)
+            self.res_layers.append(layer_name)
+
+        self._freeze_stages()
+
+        self.feat_dim = self.block.expansion * base_channels * 2**(
+            len(self.stage_blocks) - 1)
+
+    def make_stage_plugins(self, plugins, stage_idx):
+        """Make plugins for ResNet ``stage_idx`` th stage.
+
+        Currently we support to insert ``context_block``,
+        ``empirical_attention_block``, ``nonlocal_block`` into the backbone
+        like ResNet/ResNeXt. They could be inserted after conv1/conv2/conv3 of
+        Bottleneck.
+
+        An example of plugins format could be:
+
+        Examples:
+            >>> plugins=[
+            ...     dict(cfg=dict(type='xxx', arg1='xxx'),
+            ...          stages=(False, True, True, True),
+            ...          position='after_conv2'),
+            ...     dict(cfg=dict(type='yyy'),
+            ...          stages=(True, True, True, True),
+            ...          position='after_conv3'),
+            ...     dict(cfg=dict(type='zzz', postfix='1'),
+            ...          stages=(True, True, True, True),
+            ...          position='after_conv3'),
+            ...     dict(cfg=dict(type='zzz', postfix='2'),
+            ...          stages=(True, True, True, True),
+            ...          position='after_conv3')
+            ... ]
+            >>> self = ResNet(depth=18)
+            >>> stage_plugins = self.make_stage_plugins(plugins, 0)
+            >>> assert len(stage_plugins) == 3
+
+        Suppose ``stage_idx=0``, the structure of blocks in the stage would be:
+
+        .. code-block:: none
+
+            conv1-> conv2->conv3->yyy->zzz1->zzz2
+
+        Suppose 'stage_idx=1', the structure of blocks in the stage would be:
+
+        .. code-block:: none
+
+            conv1-> conv2->xxx->conv3->yyy->zzz1->zzz2
+
+        If stages is missing, the plugin would be applied to all stages.
+
+        Args:
+            plugins (list[dict]): List of plugins cfg to build. The postfix is
+                required if multiple same type plugins are inserted.
+            stage_idx (int): Index of stage to build
+
+        Returns:
+            list[dict]: Plugins for current stage
+        """
+        stage_plugins = []
+        for plugin in plugins:
+            plugin = plugin.copy()
+            stages = plugin.pop('stages', None)
+            assert stages is None or len(stages) == self.num_stages
+            # whether to insert plugin into current stage
+            if stages is None or stages[stage_idx]:
+                stage_plugins.append(plugin)
+
+        return stage_plugins
+
+    def make_res_layer(self, **kwargs):
+        """Pack all blocks in a stage into a ``ResLayer``."""
+        return ResLayer(**kwargs)
+
+    @property
+    def norm1(self):
+        """nn.Module: the normalization layer named "norm1" """
+        return getattr(self, self.norm1_name)
+
+    def _make_stem_layer(self, in_channels, stem_channels):
+        if self.deep_stem:
+            self.stem = nn.Sequential(
+                build_conv_layer(
+                    self.conv_cfg,
+                    in_channels,
+                    stem_channels // 2,
+                    kernel_size=3,
+                    stride=2,
+                    padding=1,
+                    bias=False),
+                build_norm_layer(self.norm_cfg, stem_channels // 2)[1],
+                nn.ReLU(inplace=True),
+                build_conv_layer(
+                    self.conv_cfg,
+                    stem_channels // 2,
+                    stem_channels // 2,
+                    kernel_size=3,
+                    stride=1,
+                    padding=1,
+                    bias=False),
+                build_norm_layer(self.norm_cfg, stem_channels // 2)[1],
+                nn.ReLU(inplace=True),
+                build_conv_layer(
+                    self.conv_cfg,
+                    stem_channels // 2,
+                    stem_channels,
+                    kernel_size=3,
+                    stride=1,
+                    padding=1,
+                    bias=False),
+                build_norm_layer(self.norm_cfg, stem_channels)[1],
+                nn.ReLU(inplace=True))
+        else:
+            self.conv1 = build_conv_layer(
+                self.conv_cfg,
+                in_channels,
+                stem_channels,
+                kernel_size=7,
+                stride=2,
+                padding=3,
+                bias=False)
+            self.norm1_name, norm1 = build_norm_layer(
+                self.norm_cfg, stem_channels, postfix=1)
+            self.add_module(self.norm1_name, norm1)
+            self.relu = nn.ReLU(inplace=True)
+        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
+
+    def _freeze_stages(self):
+        if self.frozen_stages >= 0:
+            if self.deep_stem:
+                self.stem.eval()
+                for param in self.stem.parameters():
+                    param.requires_grad = False
+            else:
+                self.norm1.eval()
+                for m in [self.conv1, self.norm1]:
+                    for param in m.parameters():
+                        param.requires_grad = False
+
+        for i in range(1, self.frozen_stages + 1):
+            m = getattr(self, f'layer{i}')
+            m.eval()
+            for param in m.parameters():
+                param.requires_grad = False
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in backbone.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        if isinstance(pretrained, str):
+            logger = get_root_logger()
+            load_checkpoint(self, pretrained, strict=False, logger=logger)
+        elif pretrained is None:
+            for m in self.modules():
+                if isinstance(m, nn.Conv2d):
+                    kaiming_init(m)
+                elif isinstance(m, (_BatchNorm, nn.GroupNorm)):
+                    constant_init(m, 1)
+
+            if self.dcn is not None:
+                for m in self.modules():
+                    if isinstance(m, Bottleneck) and hasattr(
+                            m.conv2, 'conv_offset'):
+                        constant_init(m.conv2.conv_offset, 0)
+
+            if self.zero_init_residual:
+                for m in self.modules():
+                    if isinstance(m, Bottleneck):
+                        constant_init(m.norm3, 0)
+                    elif isinstance(m, BasicBlock):
+                        constant_init(m.norm2, 0)
+        else:
+            raise TypeError('pretrained must be a str or None')
+
+    def forward(self, x):
+        """Forward function."""
+        if self.deep_stem:
+            x = self.stem(x)
+        else:
+            x = self.conv1(x)
+            x = self.norm1(x)
+            x = self.relu(x)
+        x = self.maxpool(x)
+        outs = []
+        for i, layer_name in enumerate(self.res_layers):
+            res_layer = getattr(self, layer_name)
+            x = res_layer(x)
+            if i in self.out_indices:
+                outs.append(x)
+        return tuple(outs)
+
+    def train(self, mode=True):
+        """Convert the model into training mode while keep normalization layer
+        freezed."""
+        super(ResNet, self).train(mode)
+        self._freeze_stages()
+        if mode and self.norm_eval:
+            for m in self.modules():
+                # trick: eval have effect on BatchNorm only
+                if isinstance(m, _BatchNorm):
+                    m.eval()
+
+
+@BACKBONES.register_module()
+class ResNetV1d(ResNet):
+    r"""ResNetV1d variant described in `Bag of Tricks
+    <https://arxiv.org/pdf/1812.01187.pdf>`_.
+
+    Compared with default ResNet(ResNetV1b), ResNetV1d replaces the 7x7 conv in
+    the input stem with three 3x3 convs. And in the downsampling block, a 2x2
+    avg_pool with stride 2 is added before conv, whose stride is changed to 1.
+    """
+
+    def __init__(self, **kwargs):
+        super(ResNetV1d, self).__init__(
+            deep_stem=True, avg_down=True, **kwargs)
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/resnext.py b/python/autoCut/models/swin/mmdet/models/backbones/resnext.py
new file mode 100644
index 0000000..6dbcbd5
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/resnext.py
@@ -0,0 +1,153 @@
+import math
+
+from mmcv.cnn import build_conv_layer, build_norm_layer
+
+from ..builder import BACKBONES
+from ..utils import ResLayer
+from .resnet import Bottleneck as _Bottleneck
+from .resnet import ResNet
+
+
+class Bottleneck(_Bottleneck):
+    expansion = 4
+
+    def __init__(self,
+                 inplanes,
+                 planes,
+                 groups=1,
+                 base_width=4,
+                 base_channels=64,
+                 **kwargs):
+        """Bottleneck block for ResNeXt.
+
+        If style is "pytorch", the stride-two layer is the 3x3 conv layer, if
+        it is "caffe", the stride-two layer is the first 1x1 conv layer.
+        """
+        super(Bottleneck, self).__init__(inplanes, planes, **kwargs)
+
+        if groups == 1:
+            width = self.planes
+        else:
+            width = math.floor(self.planes *
+                               (base_width / base_channels)) * groups
+
+        self.norm1_name, norm1 = build_norm_layer(
+            self.norm_cfg, width, postfix=1)
+        self.norm2_name, norm2 = build_norm_layer(
+            self.norm_cfg, width, postfix=2)
+        self.norm3_name, norm3 = build_norm_layer(
+            self.norm_cfg, self.planes * self.expansion, postfix=3)
+
+        self.conv1 = build_conv_layer(
+            self.conv_cfg,
+            self.inplanes,
+            width,
+            kernel_size=1,
+            stride=self.conv1_stride,
+            bias=False)
+        self.add_module(self.norm1_name, norm1)
+        fallback_on_stride = False
+        self.with_modulated_dcn = False
+        if self.with_dcn:
+            fallback_on_stride = self.dcn.pop('fallback_on_stride', False)
+        if not self.with_dcn or fallback_on_stride:
+            self.conv2 = build_conv_layer(
+                self.conv_cfg,
+                width,
+                width,
+                kernel_size=3,
+                stride=self.conv2_stride,
+                padding=self.dilation,
+                dilation=self.dilation,
+                groups=groups,
+                bias=False)
+        else:
+            assert self.conv_cfg is None, 'conv_cfg must be None for DCN'
+            self.conv2 = build_conv_layer(
+                self.dcn,
+                width,
+                width,
+                kernel_size=3,
+                stride=self.conv2_stride,
+                padding=self.dilation,
+                dilation=self.dilation,
+                groups=groups,
+                bias=False)
+
+        self.add_module(self.norm2_name, norm2)
+        self.conv3 = build_conv_layer(
+            self.conv_cfg,
+            width,
+            self.planes * self.expansion,
+            kernel_size=1,
+            bias=False)
+        self.add_module(self.norm3_name, norm3)
+
+        if self.with_plugins:
+            self._del_block_plugins(self.after_conv1_plugin_names +
+                                    self.after_conv2_plugin_names +
+                                    self.after_conv3_plugin_names)
+            self.after_conv1_plugin_names = self.make_block_plugins(
+                width, self.after_conv1_plugins)
+            self.after_conv2_plugin_names = self.make_block_plugins(
+                width, self.after_conv2_plugins)
+            self.after_conv3_plugin_names = self.make_block_plugins(
+                self.planes * self.expansion, self.after_conv3_plugins)
+
+    def _del_block_plugins(self, plugin_names):
+        """delete plugins for block if exist.
+
+        Args:
+            plugin_names (list[str]): List of plugins name to delete.
+        """
+        assert isinstance(plugin_names, list)
+        for plugin_name in plugin_names:
+            del self._modules[plugin_name]
+
+
+@BACKBONES.register_module()
+class ResNeXt(ResNet):
+    """ResNeXt backbone.
+
+    Args:
+        depth (int): Depth of resnet, from {18, 34, 50, 101, 152}.
+        in_channels (int): Number of input image channels. Default: 3.
+        num_stages (int): Resnet stages. Default: 4.
+        groups (int): Group of resnext.
+        base_width (int): Base width of resnext.
+        strides (Sequence[int]): Strides of the first block of each stage.
+        dilations (Sequence[int]): Dilation of each stage.
+        out_indices (Sequence[int]): Output from which stages.
+        style (str): `pytorch` or `caffe`. If set to "pytorch", the stride-two
+            layer is the 3x3 conv layer, otherwise the stride-two layer is
+            the first 1x1 conv layer.
+        frozen_stages (int): Stages to be frozen (all param fixed). -1 means
+            not freezing any parameters.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+        norm_eval (bool): Whether to set norm layers to eval mode, namely,
+            freeze running stats (mean and var). Note: Effect on Batch Norm
+            and its variants only.
+        with_cp (bool): Use checkpoint or not. Using checkpoint will save some
+            memory while slowing down the training speed.
+        zero_init_residual (bool): whether to use zero init for last norm layer
+            in resblocks to let them behave as identity.
+    """
+
+    arch_settings = {
+        50: (Bottleneck, (3, 4, 6, 3)),
+        101: (Bottleneck, (3, 4, 23, 3)),
+        152: (Bottleneck, (3, 8, 36, 3))
+    }
+
+    def __init__(self, groups=1, base_width=4, **kwargs):
+        self.groups = groups
+        self.base_width = base_width
+        super(ResNeXt, self).__init__(**kwargs)
+
+    def make_res_layer(self, **kwargs):
+        """Pack all blocks in a stage into a ``ResLayer``"""
+        return ResLayer(
+            groups=self.groups,
+            base_width=self.base_width,
+            base_channels=self.base_channels,
+            **kwargs)
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/ssd_vgg.py b/python/autoCut/models/swin/mmdet/models/backbones/ssd_vgg.py
new file mode 100644
index 0000000..cbc4fbb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/ssd_vgg.py
@@ -0,0 +1,169 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import VGG, constant_init, kaiming_init, normal_init, xavier_init
+from mmcv.runner import load_checkpoint
+
+from mmdet.utils import get_root_logger
+from ..builder import BACKBONES
+
+
+@BACKBONES.register_module()
+class SSDVGG(VGG):
+    """VGG Backbone network for single-shot-detection.
+
+    Args:
+        input_size (int): width and height of input, from {300, 512}.
+        depth (int): Depth of vgg, from {11, 13, 16, 19}.
+        out_indices (Sequence[int]): Output from which stages.
+
+    Example:
+        >>> self = SSDVGG(input_size=300, depth=11)
+        >>> self.eval()
+        >>> inputs = torch.rand(1, 3, 300, 300)
+        >>> level_outputs = self.forward(inputs)
+        >>> for level_out in level_outputs:
+        ...     print(tuple(level_out.shape))
+        (1, 1024, 19, 19)
+        (1, 512, 10, 10)
+        (1, 256, 5, 5)
+        (1, 256, 3, 3)
+        (1, 256, 1, 1)
+    """
+    extra_setting = {
+        300: (256, 'S', 512, 128, 'S', 256, 128, 256, 128, 256),
+        512: (256, 'S', 512, 128, 'S', 256, 128, 'S', 256, 128, 'S', 256, 128),
+    }
+
+    def __init__(self,
+                 input_size,
+                 depth,
+                 with_last_pool=False,
+                 ceil_mode=True,
+                 out_indices=(3, 4),
+                 out_feature_indices=(22, 34),
+                 l2_norm_scale=20.):
+        # TODO: in_channels for mmcv.VGG
+        super(SSDVGG, self).__init__(
+            depth,
+            with_last_pool=with_last_pool,
+            ceil_mode=ceil_mode,
+            out_indices=out_indices)
+        assert input_size in (300, 512)
+        self.input_size = input_size
+
+        self.features.add_module(
+            str(len(self.features)),
+            nn.MaxPool2d(kernel_size=3, stride=1, padding=1))
+        self.features.add_module(
+            str(len(self.features)),
+            nn.Conv2d(512, 1024, kernel_size=3, padding=6, dilation=6))
+        self.features.add_module(
+            str(len(self.features)), nn.ReLU(inplace=True))
+        self.features.add_module(
+            str(len(self.features)), nn.Conv2d(1024, 1024, kernel_size=1))
+        self.features.add_module(
+            str(len(self.features)), nn.ReLU(inplace=True))
+        self.out_feature_indices = out_feature_indices
+
+        self.inplanes = 1024
+        self.extra = self._make_extra_layers(self.extra_setting[input_size])
+        self.l2_norm = L2Norm(
+            self.features[out_feature_indices[0] - 1].out_channels,
+            l2_norm_scale)
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in backbone.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        if isinstance(pretrained, str):
+            logger = get_root_logger()
+            load_checkpoint(self, pretrained, strict=False, logger=logger)
+        elif pretrained is None:
+            for m in self.features.modules():
+                if isinstance(m, nn.Conv2d):
+                    kaiming_init(m)
+                elif isinstance(m, nn.BatchNorm2d):
+                    constant_init(m, 1)
+                elif isinstance(m, nn.Linear):
+                    normal_init(m, std=0.01)
+        else:
+            raise TypeError('pretrained must be a str or None')
+
+        for m in self.extra.modules():
+            if isinstance(m, nn.Conv2d):
+                xavier_init(m, distribution='uniform')
+
+        constant_init(self.l2_norm, self.l2_norm.scale)
+
+    def forward(self, x):
+        """Forward function."""
+        outs = []
+        for i, layer in enumerate(self.features):
+            x = layer(x)
+            if i in self.out_feature_indices:
+                outs.append(x)
+        for i, layer in enumerate(self.extra):
+            x = F.relu(layer(x), inplace=True)
+            if i % 2 == 1:
+                outs.append(x)
+        outs[0] = self.l2_norm(outs[0])
+        if len(outs) == 1:
+            return outs[0]
+        else:
+            return tuple(outs)
+
+    def _make_extra_layers(self, outplanes):
+        layers = []
+        kernel_sizes = (1, 3)
+        num_layers = 0
+        outplane = None
+        for i in range(len(outplanes)):
+            if self.inplanes == 'S':
+                self.inplanes = outplane
+                continue
+            k = kernel_sizes[num_layers % 2]
+            if outplanes[i] == 'S':
+                outplane = outplanes[i + 1]
+                conv = nn.Conv2d(
+                    self.inplanes, outplane, k, stride=2, padding=1)
+            else:
+                outplane = outplanes[i]
+                conv = nn.Conv2d(
+                    self.inplanes, outplane, k, stride=1, padding=0)
+            layers.append(conv)
+            self.inplanes = outplanes[i]
+            num_layers += 1
+        if self.input_size == 512:
+            layers.append(nn.Conv2d(self.inplanes, 256, 4, padding=1))
+
+        return nn.Sequential(*layers)
+
+
+class L2Norm(nn.Module):
+
+    def __init__(self, n_dims, scale=20., eps=1e-10):
+        """L2 normalization layer.
+
+        Args:
+            n_dims (int): Number of dimensions to be normalized
+            scale (float, optional): Defaults to 20..
+            eps (float, optional): Used to avoid division by zero.
+                Defaults to 1e-10.
+        """
+        super(L2Norm, self).__init__()
+        self.n_dims = n_dims
+        self.weight = nn.Parameter(torch.Tensor(self.n_dims))
+        self.eps = eps
+        self.scale = scale
+
+    def forward(self, x):
+        """Forward function."""
+        # normalization layer convert to FP32 in FP16 training
+        x_float = x.float()
+        norm = x_float.pow(2).sum(1, keepdim=True).sqrt() + self.eps
+        return (self.weight[None, :, None, None].float().expand_as(x_float) *
+                x_float / norm).type_as(x)
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/swin_transformer.py b/python/autoCut/models/swin/mmdet/models/backbones/swin_transformer.py
new file mode 100644
index 0000000..bb41850
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/swin_transformer.py
@@ -0,0 +1,630 @@
+# --------------------------------------------------------
+# Swin Transformer
+# Copyright (c) 2021 Microsoft
+# Licensed under The MIT License [see LICENSE for details]
+# Written by Ze Liu, Yutong Lin, Yixuan Wei
+# --------------------------------------------------------
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.utils.checkpoint as checkpoint
+import numpy as np
+from timm.models.layers import DropPath, to_2tuple, trunc_normal_
+
+from mmcv_custom import load_checkpoint
+from mmdet.utils import get_root_logger
+from ..builder import BACKBONES
+
+
+class Mlp(nn.Module):
+    """ Multilayer perceptron."""
+
+    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.fc1 = nn.Linear(in_features, hidden_features)
+        self.act = act_layer()
+        self.fc2 = nn.Linear(hidden_features, out_features)
+        self.drop = nn.Dropout(drop)
+
+    def forward(self, x):
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.drop(x)
+        x = self.fc2(x)
+        x = self.drop(x)
+        return x
+
+
+def window_partition(x, window_size):
+    """
+    Args:
+        x: (B, H, W, C)
+        window_size (int): window size
+
+    Returns:
+        windows: (num_windows*B, window_size, window_size, C)
+    """
+    B, H, W, C = x.shape
+    x = x.view(B, H // window_size, window_size, W // window_size, window_size, C)
+    windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C)
+    return windows
+
+
+def window_reverse(windows, window_size, H, W):
+    """
+    Args:
+        windows: (num_windows*B, window_size, window_size, C)
+        window_size (int): Window size
+        H (int): Height of image
+        W (int): Width of image
+
+    Returns:
+        x: (B, H, W, C)
+    """
+    B = int(windows.shape[0] / (H * W / window_size / window_size))
+    x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1)
+    x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1)
+    return x
+
+
+class WindowAttention(nn.Module):
+    """ Window based multi-head self attention (W-MSA) module with relative position bias.
+    It supports both of shifted and non-shifted window.
+
+    Args:
+        dim (int): Number of input channels.
+        window_size (tuple[int]): The height and width of the window.
+        num_heads (int): Number of attention heads.
+        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set
+        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0
+        proj_drop (float, optional): Dropout ratio of output. Default: 0.0
+    """
+
+    def __init__(self, dim, window_size, num_heads, qkv_bias=True, qk_scale=None, attn_drop=0., proj_drop=0.):
+
+        super().__init__()
+        self.dim = dim
+        self.window_size = window_size  # Wh, Ww
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+        self.scale = qk_scale or head_dim ** -0.5
+
+        # define a parameter table of relative position bias
+        self.relative_position_bias_table = nn.Parameter(
+            torch.zeros((2 * window_size[0] - 1) * (2 * window_size[1] - 1), num_heads))  # 2*Wh-1 * 2*Ww-1, nH
+
+        # get pair-wise relative position index for each token inside the window
+        coords_h = torch.arange(self.window_size[0])
+        coords_w = torch.arange(self.window_size[1])
+        coords = torch.stack(torch.meshgrid([coords_h, coords_w]))  # 2, Wh, Ww
+        coords_flatten = torch.flatten(coords, 1)  # 2, Wh*Ww
+        relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :]  # 2, Wh*Ww, Wh*Ww
+        relative_coords = relative_coords.permute(1, 2, 0).contiguous()  # Wh*Ww, Wh*Ww, 2
+        relative_coords[:, :, 0] += self.window_size[0] - 1  # shift to start from 0
+        relative_coords[:, :, 1] += self.window_size[1] - 1
+        relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1
+        relative_position_index = relative_coords.sum(-1)  # Wh*Ww, Wh*Ww
+        self.register_buffer("relative_position_index", relative_position_index)
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.attn_drop = nn.Dropout(attn_drop)
+        self.proj = nn.Linear(dim, dim)
+        self.proj_drop = nn.Dropout(proj_drop)
+
+        trunc_normal_(self.relative_position_bias_table, std=.02)
+        self.softmax = nn.Softmax(dim=-1)
+
+    def forward(self, x, mask=None):
+        """ Forward function.
+
+        Args:
+            x: input features with shape of (num_windows*B, N, C)
+            mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None
+        """
+        B_, N, C = x.shape
+        qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
+        q, k, v = qkv[0], qkv[1], qkv[2]  # make torchscript happy (cannot use tensor as tuple)
+
+        q = q * self.scale
+        attn = (q @ k.transpose(-2, -1))
+
+        relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
+            self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1)  # Wh*Ww,Wh*Ww,nH
+        relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww
+        attn = attn + relative_position_bias.unsqueeze(0)
+
+        if mask is not None:
+            nW = mask.shape[0]
+            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0)
+            attn = attn.view(-1, self.num_heads, N, N)
+            attn = self.softmax(attn)
+        else:
+            attn = self.softmax(attn)
+
+        attn = self.attn_drop(attn)
+
+        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
+        x = self.proj(x)
+        x = self.proj_drop(x)
+        return x
+
+
+class SwinTransformerBlock(nn.Module):
+    """ Swin Transformer Block.
+
+    Args:
+        dim (int): Number of input channels.
+        num_heads (int): Number of attention heads.
+        window_size (int): Window size.
+        shift_size (int): Shift size for SW-MSA.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
+        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
+        drop (float, optional): Dropout rate. Default: 0.0
+        attn_drop (float, optional): Attention dropout rate. Default: 0.0
+        drop_path (float, optional): Stochastic depth rate. Default: 0.0
+        act_layer (nn.Module, optional): Activation layer. Default: nn.GELU
+        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
+    """
+
+    def __init__(self, dim, num_heads, window_size=7, shift_size=0,
+                 mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0., drop_path=0.,
+                 act_layer=nn.GELU, norm_layer=nn.LayerNorm):
+        super().__init__()
+        self.dim = dim
+        self.num_heads = num_heads
+        self.window_size = window_size
+        self.shift_size = shift_size
+        self.mlp_ratio = mlp_ratio
+        assert 0 <= self.shift_size < self.window_size, "shift_size must in 0-window_size"
+
+        self.norm1 = norm_layer(dim)
+        self.attn = WindowAttention(
+            dim, window_size=to_2tuple(self.window_size), num_heads=num_heads,
+            qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
+
+        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
+        self.norm2 = norm_layer(dim)
+        mlp_hidden_dim = int(dim * mlp_ratio)
+        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)
+
+        self.H = None
+        self.W = None
+
+    def forward(self, x, mask_matrix):
+        """ Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, H*W, C).
+            H, W: Spatial resolution of the input feature.
+            mask_matrix: Attention mask for cyclic shift.
+        """
+        B, L, C = x.shape
+        H, W = self.H, self.W
+        assert L == H * W, "input feature has wrong size"
+
+        shortcut = x
+        x = self.norm1(x)
+        x = x.view(B, H, W, C)
+
+        # pad feature maps to multiples of window size
+        pad_l = pad_t = 0
+        pad_r = (self.window_size - W % self.window_size) % self.window_size
+        pad_b = (self.window_size - H % self.window_size) % self.window_size
+        x = F.pad(x, (0, 0, pad_l, pad_r, pad_t, pad_b))
+        _, Hp, Wp, _ = x.shape
+
+        # cyclic shift
+        if self.shift_size > 0:
+            shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
+            attn_mask = mask_matrix
+        else:
+            shifted_x = x
+            attn_mask = None
+
+        # partition windows
+        x_windows = window_partition(shifted_x, self.window_size)  # nW*B, window_size, window_size, C
+        x_windows = x_windows.view(-1, self.window_size * self.window_size, C)  # nW*B, window_size*window_size, C
+
+        # W-MSA/SW-MSA
+        attn_windows = self.attn(x_windows, mask=attn_mask)  # nW*B, window_size*window_size, C
+
+        # merge windows
+        attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C)
+        shifted_x = window_reverse(attn_windows, self.window_size, Hp, Wp)  # B H' W' C
+
+        # reverse cyclic shift
+        if self.shift_size > 0:
+            x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
+        else:
+            x = shifted_x
+
+        if pad_r > 0 or pad_b > 0:
+            x = x[:, :H, :W, :].contiguous()
+
+        x = x.view(B, H * W, C)
+
+        # FFN
+        x = shortcut + self.drop_path(x)
+        x = x + self.drop_path(self.mlp(self.norm2(x)))
+
+        return x
+
+
+class PatchMerging(nn.Module):
+    """ Patch Merging Layer
+
+    Args:
+        dim (int): Number of input channels.
+        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
+    """
+    def __init__(self, dim, norm_layer=nn.LayerNorm):
+        super().__init__()
+        self.dim = dim
+        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
+        self.norm = norm_layer(4 * dim)
+
+    def forward(self, x, H, W):
+        """ Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, H*W, C).
+            H, W: Spatial resolution of the input feature.
+        """
+        B, L, C = x.shape
+        assert L == H * W, "input feature has wrong size"
+
+        x = x.view(B, H, W, C)
+
+        # padding
+        pad_input = (H % 2 == 1) or (W % 2 == 1)
+        if pad_input:
+            x = F.pad(x, (0, 0, 0, W % 2, 0, H % 2))
+
+        x0 = x[:, 0::2, 0::2, :]  # B H/2 W/2 C
+        x1 = x[:, 1::2, 0::2, :]  # B H/2 W/2 C
+        x2 = x[:, 0::2, 1::2, :]  # B H/2 W/2 C
+        x3 = x[:, 1::2, 1::2, :]  # B H/2 W/2 C
+        x = torch.cat([x0, x1, x2, x3], -1)  # B H/2 W/2 4*C
+        x = x.view(B, -1, 4 * C)  # B H/2*W/2 4*C
+
+        x = self.norm(x)
+        x = self.reduction(x)
+
+        return x
+
+
+class BasicLayer(nn.Module):
+    """ A basic Swin Transformer layer for one stage.
+
+    Args:
+        dim (int): Number of feature channels
+        depth (int): Depths of this stage.
+        num_heads (int): Number of attention head.
+        window_size (int): Local window size. Default: 7.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
+        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
+        drop (float, optional): Dropout rate. Default: 0.0
+        attn_drop (float, optional): Attention dropout rate. Default: 0.0
+        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
+        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
+        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
+        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.
+    """
+
+    def __init__(self,
+                 dim,
+                 depth,
+                 num_heads,
+                 window_size=7,
+                 mlp_ratio=4.,
+                 qkv_bias=True,
+                 qk_scale=None,
+                 drop=0.,
+                 attn_drop=0.,
+                 drop_path=0.,
+                 norm_layer=nn.LayerNorm,
+                 downsample=None,
+                 use_checkpoint=False):
+        super().__init__()
+        self.window_size = window_size
+        self.shift_size = window_size // 2
+        self.depth = depth
+        self.use_checkpoint = use_checkpoint
+
+        # build blocks
+        self.blocks = nn.ModuleList([
+            SwinTransformerBlock(
+                dim=dim,
+                num_heads=num_heads,
+                window_size=window_size,
+                shift_size=0 if (i % 2 == 0) else window_size // 2,
+                mlp_ratio=mlp_ratio,
+                qkv_bias=qkv_bias,
+                qk_scale=qk_scale,
+                drop=drop,
+                attn_drop=attn_drop,
+                drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,
+                norm_layer=norm_layer)
+            for i in range(depth)])
+
+        # patch merging layer
+        if downsample is not None:
+            self.downsample = downsample(dim=dim, norm_layer=norm_layer)
+        else:
+            self.downsample = None
+
+    def forward(self, x, H, W):
+        """ Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, H*W, C).
+            H, W: Spatial resolution of the input feature.
+        """
+
+        # calculate attention mask for SW-MSA
+        Hp = int(np.ceil(H / self.window_size)) * self.window_size
+        Wp = int(np.ceil(W / self.window_size)) * self.window_size
+        img_mask = torch.zeros((1, Hp, Wp, 1), device=x.device)  # 1 Hp Wp 1
+        h_slices = (slice(0, -self.window_size),
+                    slice(-self.window_size, -self.shift_size),
+                    slice(-self.shift_size, None))
+        w_slices = (slice(0, -self.window_size),
+                    slice(-self.window_size, -self.shift_size),
+                    slice(-self.shift_size, None))
+        cnt = 0
+        for h in h_slices:
+            for w in w_slices:
+                img_mask[:, h, w, :] = cnt
+                cnt += 1
+
+        mask_windows = window_partition(img_mask, self.window_size)  # nW, window_size, window_size, 1
+        mask_windows = mask_windows.view(-1, self.window_size * self.window_size)
+        attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
+        attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))
+
+        for blk in self.blocks:
+            blk.H, blk.W = H, W
+            if self.use_checkpoint:
+                x = checkpoint.checkpoint(blk, x, attn_mask)
+            else:
+                x = blk(x, attn_mask)
+        if self.downsample is not None:
+            x_down = self.downsample(x, H, W)
+            Wh, Ww = (H + 1) // 2, (W + 1) // 2
+            return x, H, W, x_down, Wh, Ww
+        else:
+            return x, H, W, x, H, W
+
+
+class PatchEmbed(nn.Module):
+    """ Image to Patch Embedding
+
+    Args:
+        patch_size (int): Patch token size. Default: 4.
+        in_chans (int): Number of input image channels. Default: 3.
+        embed_dim (int): Number of linear projection output channels. Default: 96.
+        norm_layer (nn.Module, optional): Normalization layer. Default: None
+    """
+
+    def __init__(self, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None):
+        super().__init__()
+        patch_size = to_2tuple(patch_size)
+        self.patch_size = patch_size
+
+        self.in_chans = in_chans
+        self.embed_dim = embed_dim
+
+        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)
+        if norm_layer is not None:
+            self.norm = norm_layer(embed_dim)
+        else:
+            self.norm = None
+
+    def forward(self, x):
+        """Forward function."""
+        # padding
+        _, _, H, W = x.size()
+        if W % self.patch_size[1] != 0:
+            x = F.pad(x, (0, self.patch_size[1] - W % self.patch_size[1]))
+        if H % self.patch_size[0] != 0:
+            x = F.pad(x, (0, 0, 0, self.patch_size[0] - H % self.patch_size[0]))
+
+        x = self.proj(x)  # B C Wh Ww
+        if self.norm is not None:
+            Wh, Ww = x.size(2), x.size(3)
+            x = x.flatten(2).transpose(1, 2)
+            x = self.norm(x)
+            x = x.transpose(1, 2).view(-1, self.embed_dim, Wh, Ww)
+
+        return x
+
+
+@BACKBONES.register_module()
+class SwinTransformer(nn.Module):
+    """ Swin Transformer backbone.
+        A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows`  -
+          https://arxiv.org/pdf/2103.14030
+
+    Args:
+        pretrain_img_size (int): Input image size for training the pretrained model,
+            used in absolute postion embedding. Default 224.
+        patch_size (int | tuple(int)): Patch size. Default: 4.
+        in_chans (int): Number of input image channels. Default: 3.
+        embed_dim (int): Number of linear projection output channels. Default: 96.
+        depths (tuple[int]): Depths of each Swin Transformer stage.
+        num_heads (tuple[int]): Number of attention head of each stage.
+        window_size (int): Window size. Default: 7.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
+        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float): Override default qk scale of head_dim ** -0.5 if set.
+        drop_rate (float): Dropout rate.
+        attn_drop_rate (float): Attention dropout rate. Default: 0.
+        drop_path_rate (float): Stochastic depth rate. Default: 0.2.
+        norm_layer (nn.Module): Normalization layer. Default: nn.LayerNorm.
+        ape (bool): If True, add absolute position embedding to the patch embedding. Default: False.
+        patch_norm (bool): If True, add normalization after patch embedding. Default: True.
+        out_indices (Sequence[int]): Output from which stages.
+        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
+            -1 means not freezing any parameters.
+        use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False.
+    """
+
+    def __init__(self,
+                 pretrain_img_size=224,
+                 patch_size=4,
+                 in_chans=3,
+                 embed_dim=96,
+                 depths=[2, 2, 6, 2],
+                 num_heads=[3, 6, 12, 24],
+                 window_size=7,
+                 mlp_ratio=4.,
+                 qkv_bias=True,
+                 qk_scale=None,
+                 drop_rate=0.,
+                 attn_drop_rate=0.,
+                 drop_path_rate=0.2,
+                 norm_layer=nn.LayerNorm,
+                 ape=False,
+                 patch_norm=True,
+                 out_indices=(0, 1, 2, 3),
+                 frozen_stages=-1,
+                 use_checkpoint=False):
+        super().__init__()
+
+        self.pretrain_img_size = pretrain_img_size
+        self.num_layers = len(depths)
+        self.embed_dim = embed_dim
+        self.ape = ape
+        self.patch_norm = patch_norm
+        self.out_indices = out_indices
+        self.frozen_stages = frozen_stages
+
+        # split image into non-overlapping patches
+        self.patch_embed = PatchEmbed(
+            patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim,
+            norm_layer=norm_layer if self.patch_norm else None)
+
+        # absolute position embedding
+        if self.ape:
+            pretrain_img_size = to_2tuple(pretrain_img_size)
+            patch_size = to_2tuple(patch_size)
+            patches_resolution = [pretrain_img_size[0] // patch_size[0], pretrain_img_size[1] // patch_size[1]]
+
+            self.absolute_pos_embed = nn.Parameter(torch.zeros(1, embed_dim, patches_resolution[0], patches_resolution[1]))
+            trunc_normal_(self.absolute_pos_embed, std=.02)
+
+        self.pos_drop = nn.Dropout(p=drop_rate)
+
+        # stochastic depth
+        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))]  # stochastic depth decay rule
+
+        # build layers
+        self.layers = nn.ModuleList()
+        for i_layer in range(self.num_layers):
+            layer = BasicLayer(
+                dim=int(embed_dim * 2 ** i_layer),
+                depth=depths[i_layer],
+                num_heads=num_heads[i_layer],
+                window_size=window_size,
+                mlp_ratio=mlp_ratio,
+                qkv_bias=qkv_bias,
+                qk_scale=qk_scale,
+                drop=drop_rate,
+                attn_drop=attn_drop_rate,
+                drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])],
+                norm_layer=norm_layer,
+                downsample=PatchMerging if (i_layer < self.num_layers - 1) else None,
+                use_checkpoint=use_checkpoint)
+            self.layers.append(layer)
+
+        num_features = [int(embed_dim * 2 ** i) for i in range(self.num_layers)]
+        self.num_features = num_features
+
+        # add a norm layer for each output
+        for i_layer in out_indices:
+            layer = norm_layer(num_features[i_layer])
+            layer_name = f'norm{i_layer}'
+            self.add_module(layer_name, layer)
+
+        self._freeze_stages()
+
+    def _freeze_stages(self):
+        if self.frozen_stages >= 0:
+            self.patch_embed.eval()
+            for param in self.patch_embed.parameters():
+                param.requires_grad = False
+
+        if self.frozen_stages >= 1 and self.ape:
+            self.absolute_pos_embed.requires_grad = False
+
+        if self.frozen_stages >= 2:
+            self.pos_drop.eval()
+            for i in range(0, self.frozen_stages - 1):
+                m = self.layers[i]
+                m.eval()
+                for param in m.parameters():
+                    param.requires_grad = False
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in backbone.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+
+        def _init_weights(m):
+            if isinstance(m, nn.Linear):
+                trunc_normal_(m.weight, std=.02)
+                if isinstance(m, nn.Linear) and m.bias is not None:
+                    nn.init.constant_(m.bias, 0)
+            elif isinstance(m, nn.LayerNorm):
+                nn.init.constant_(m.bias, 0)
+                nn.init.constant_(m.weight, 1.0)
+
+        if isinstance(pretrained, str):
+            self.apply(_init_weights)
+            logger = get_root_logger()
+            load_checkpoint(self, pretrained, strict=False, logger=logger)
+        elif pretrained is None:
+            self.apply(_init_weights)
+        else:
+            raise TypeError('pretrained must be a str or None')
+
+    def forward(self, x):
+        """Forward function."""
+        x = self.patch_embed(x)
+
+        Wh, Ww = x.size(2), x.size(3)
+        if self.ape:
+            # interpolate the position embedding to the corresponding size
+            absolute_pos_embed = F.interpolate(self.absolute_pos_embed, size=(Wh, Ww), mode='bicubic')
+            x = (x + absolute_pos_embed).flatten(2).transpose(1, 2)  # B Wh*Ww C
+        else:
+            x = x.flatten(2).transpose(1, 2)
+        x = self.pos_drop(x)
+
+        outs = []
+        for i in range(self.num_layers):
+            layer = self.layers[i]
+            x_out, H, W, x, Wh, Ww = layer(x, Wh, Ww)
+
+            if i in self.out_indices:
+                norm_layer = getattr(self, f'norm{i}')
+                x_out = norm_layer(x_out)
+
+                out = x_out.view(-1, H, W, self.num_features[i]).permute(0, 3, 1, 2).contiguous()
+                outs.append(out)
+
+        return tuple(outs)
+
+    def train(self, mode=True):
+        """Convert the model into training mode while keep layers freezed."""
+        super(SwinTransformer, self).train(mode)
+        self._freeze_stages()
diff --git a/python/autoCut/models/swin/mmdet/models/backbones/trident_resnet.py b/python/autoCut/models/swin/mmdet/models/backbones/trident_resnet.py
new file mode 100644
index 0000000..e610013
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/backbones/trident_resnet.py
@@ -0,0 +1,292 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.utils.checkpoint as cp
+from mmcv.cnn import build_conv_layer, build_norm_layer, kaiming_init
+from torch.nn.modules.utils import _pair
+
+from mmdet.models.backbones.resnet import Bottleneck, ResNet
+from mmdet.models.builder import BACKBONES
+
+
+class TridentConv(nn.Module):
+    """Trident Convolution Module.
+
+    Args:
+        in_channels (int): Number of channels in input.
+        out_channels (int): Number of channels in output.
+        kernel_size (int): Size of convolution kernel.
+        stride (int, optional): Convolution stride. Default: 1.
+        trident_dilations (tuple[int, int, int], optional): Dilations of
+            different trident branch. Default: (1, 2, 3).
+        test_branch_idx (int, optional): In inference, all 3 branches will
+            be used if `test_branch_idx==-1`, otherwise only branch with
+            index `test_branch_idx` will be used. Default: 1.
+        bias (bool, optional): Whether to use bias in convolution or not.
+            Default: False.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 kernel_size,
+                 stride=1,
+                 trident_dilations=(1, 2, 3),
+                 test_branch_idx=1,
+                 bias=False):
+        super(TridentConv, self).__init__()
+        self.num_branch = len(trident_dilations)
+        self.with_bias = bias
+        self.test_branch_idx = test_branch_idx
+        self.stride = _pair(stride)
+        self.kernel_size = _pair(kernel_size)
+        self.paddings = _pair(trident_dilations)
+        self.dilations = trident_dilations
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.bias = bias
+
+        self.weight = nn.Parameter(
+            torch.Tensor(out_channels, in_channels, *self.kernel_size))
+        if bias:
+            self.bias = nn.Parameter(torch.Tensor(out_channels))
+        else:
+            self.bias = None
+        self.init_weights()
+
+    def init_weights(self):
+        kaiming_init(self, distribution='uniform', mode='fan_in')
+
+    def extra_repr(self):
+        tmpstr = f'in_channels={self.in_channels}'
+        tmpstr += f', out_channels={self.out_channels}'
+        tmpstr += f', kernel_size={self.kernel_size}'
+        tmpstr += f', num_branch={self.num_branch}'
+        tmpstr += f', test_branch_idx={self.test_branch_idx}'
+        tmpstr += f', stride={self.stride}'
+        tmpstr += f', paddings={self.paddings}'
+        tmpstr += f', dilations={self.dilations}'
+        tmpstr += f', bias={self.bias}'
+        return tmpstr
+
+    def forward(self, inputs):
+        if self.training or self.test_branch_idx == -1:
+            outputs = [
+                F.conv2d(input, self.weight, self.bias, self.stride, padding,
+                         dilation) for input, dilation, padding in zip(
+                             inputs, self.dilations, self.paddings)
+            ]
+        else:
+            assert len(inputs) == 1
+            outputs = [
+                F.conv2d(inputs[0], self.weight, self.bias, self.stride,
+                         self.paddings[self.test_branch_idx],
+                         self.dilations[self.test_branch_idx])
+            ]
+
+        return outputs
+
+
+# Since TridentNet is defined over ResNet50 and ResNet101, here we
+# only support TridentBottleneckBlock.
+class TridentBottleneck(Bottleneck):
+    """BottleBlock for TridentResNet.
+
+    Args:
+        trident_dilations (tuple[int, int, int]): Dilations of different
+            trident branch.
+        test_branch_idx (int): In inference, all 3 branches will be used
+            if `test_branch_idx==-1`, otherwise only branch with index
+            `test_branch_idx` will be used.
+        concat_output (bool): Whether to concat the output list to a Tensor.
+            `True` only in the last Block.
+    """
+
+    def __init__(self, trident_dilations, test_branch_idx, concat_output,
+                 **kwargs):
+
+        super(TridentBottleneck, self).__init__(**kwargs)
+        self.trident_dilations = trident_dilations
+        self.num_branch = len(trident_dilations)
+        self.concat_output = concat_output
+        self.test_branch_idx = test_branch_idx
+        self.conv2 = TridentConv(
+            self.planes,
+            self.planes,
+            kernel_size=3,
+            stride=self.conv2_stride,
+            bias=False,
+            trident_dilations=self.trident_dilations,
+            test_branch_idx=test_branch_idx)
+
+    def forward(self, x):
+
+        def _inner_forward(x):
+            num_branch = (
+                self.num_branch
+                if self.training or self.test_branch_idx == -1 else 1)
+            identity = x
+            if not isinstance(x, list):
+                x = (x, ) * num_branch
+                identity = x
+                if self.downsample is not None:
+                    identity = [self.downsample(b) for b in x]
+
+            out = [self.conv1(b) for b in x]
+            out = [self.norm1(b) for b in out]
+            out = [self.relu(b) for b in out]
+
+            if self.with_plugins:
+                for k in range(len(out)):
+                    out[k] = self.forward_plugin(out[k],
+                                                 self.after_conv1_plugin_names)
+
+            out = self.conv2(out)
+            out = [self.norm2(b) for b in out]
+            out = [self.relu(b) for b in out]
+            if self.with_plugins:
+                for k in range(len(out)):
+                    out[k] = self.forward_plugin(out[k],
+                                                 self.after_conv2_plugin_names)
+
+            out = [self.conv3(b) for b in out]
+            out = [self.norm3(b) for b in out]
+
+            if self.with_plugins:
+                for k in range(len(out)):
+                    out[k] = self.forward_plugin(out[k],
+                                                 self.after_conv3_plugin_names)
+
+            out = [
+                out_b + identity_b for out_b, identity_b in zip(out, identity)
+            ]
+            return out
+
+        if self.with_cp and x.requires_grad:
+            out = cp.checkpoint(_inner_forward, x)
+        else:
+            out = _inner_forward(x)
+
+        out = [self.relu(b) for b in out]
+        if self.concat_output:
+            out = torch.cat(out, dim=0)
+        return out
+
+
+def make_trident_res_layer(block,
+                           inplanes,
+                           planes,
+                           num_blocks,
+                           stride=1,
+                           trident_dilations=(1, 2, 3),
+                           style='pytorch',
+                           with_cp=False,
+                           conv_cfg=None,
+                           norm_cfg=dict(type='BN'),
+                           dcn=None,
+                           plugins=None,
+                           test_branch_idx=-1):
+    """Build Trident Res Layers."""
+
+    downsample = None
+    if stride != 1 or inplanes != planes * block.expansion:
+        downsample = []
+        conv_stride = stride
+        downsample.extend([
+            build_conv_layer(
+                conv_cfg,
+                inplanes,
+                planes * block.expansion,
+                kernel_size=1,
+                stride=conv_stride,
+                bias=False),
+            build_norm_layer(norm_cfg, planes * block.expansion)[1]
+        ])
+        downsample = nn.Sequential(*downsample)
+
+    layers = []
+    for i in range(num_blocks):
+        layers.append(
+            block(
+                inplanes=inplanes,
+                planes=planes,
+                stride=stride if i == 0 else 1,
+                trident_dilations=trident_dilations,
+                downsample=downsample if i == 0 else None,
+                style=style,
+                with_cp=with_cp,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                dcn=dcn,
+                plugins=plugins,
+                test_branch_idx=test_branch_idx,
+                concat_output=True if i == num_blocks - 1 else False))
+        inplanes = planes * block.expansion
+    return nn.Sequential(*layers)
+
+
+@BACKBONES.register_module()
+class TridentResNet(ResNet):
+    """The stem layer, stage 1 and stage 2 in Trident ResNet are identical to
+    ResNet, while in stage 3, Trident BottleBlock is utilized to replace the
+    normal BottleBlock to yield trident output. Different branch shares the
+    convolution weight but uses different dilations to achieve multi-scale
+    output.
+
+                               / stage3(b0) \
+    x - stem - stage1 - stage2 - stage3(b1) - output
+                               \ stage3(b2) /
+
+    Args:
+        depth (int): Depth of resnet, from {50, 101, 152}.
+        num_branch (int): Number of branches in TridentNet.
+        test_branch_idx (int): In inference, all 3 branches will be used
+            if `test_branch_idx==-1`, otherwise only branch with index
+            `test_branch_idx` will be used.
+        trident_dilations (tuple[int]): Dilations of different trident branch.
+            len(trident_dilations) should be equal to num_branch.
+    """  # noqa
+
+    def __init__(self, depth, num_branch, test_branch_idx, trident_dilations,
+                 **kwargs):
+
+        assert num_branch == len(trident_dilations)
+        assert depth in (50, 101, 152)
+        super(TridentResNet, self).__init__(depth, **kwargs)
+        assert self.num_stages == 3
+        self.test_branch_idx = test_branch_idx
+        self.num_branch = num_branch
+
+        last_stage_idx = self.num_stages - 1
+        stride = self.strides[last_stage_idx]
+        dilation = trident_dilations
+        dcn = self.dcn if self.stage_with_dcn[last_stage_idx] else None
+        if self.plugins is not None:
+            stage_plugins = self.make_stage_plugins(self.plugins,
+                                                    last_stage_idx)
+        else:
+            stage_plugins = None
+        planes = self.base_channels * 2**last_stage_idx
+        res_layer = make_trident_res_layer(
+            TridentBottleneck,
+            inplanes=(self.block.expansion * self.base_channels *
+                      2**(last_stage_idx - 1)),
+            planes=planes,
+            num_blocks=self.stage_blocks[last_stage_idx],
+            stride=stride,
+            trident_dilations=dilation,
+            style=self.style,
+            with_cp=self.with_cp,
+            conv_cfg=self.conv_cfg,
+            norm_cfg=self.norm_cfg,
+            dcn=dcn,
+            plugins=stage_plugins,
+            test_branch_idx=self.test_branch_idx)
+
+        layer_name = f'layer{last_stage_idx + 1}'
+
+        self.__setattr__(layer_name, res_layer)
+        self.res_layers.pop(last_stage_idx)
+        self.res_layers.insert(last_stage_idx, layer_name)
+
+        self._freeze_stages()
diff --git a/python/autoCut/models/swin/mmdet/models/builder.py b/python/autoCut/models/swin/mmdet/models/builder.py
new file mode 100644
index 0000000..81c927e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/builder.py
@@ -0,0 +1,77 @@
+import warnings
+
+from mmcv.utils import Registry, build_from_cfg
+from torch import nn
+
+BACKBONES = Registry('backbone')
+NECKS = Registry('neck')
+ROI_EXTRACTORS = Registry('roi_extractor')
+SHARED_HEADS = Registry('shared_head')
+HEADS = Registry('head')
+LOSSES = Registry('loss')
+DETECTORS = Registry('detector')
+
+
+def build(cfg, registry, default_args=None):
+    """Build a module.
+
+    Args:
+        cfg (dict, list[dict]): The config of modules, is is either a dict
+            or a list of configs.
+        registry (:obj:`Registry`): A registry the module belongs to.
+        default_args (dict, optional): Default arguments to build the module.
+            Defaults to None.
+
+    Returns:
+        nn.Module: A built nn module.
+    """
+    if isinstance(cfg, list):
+        modules = [
+            build_from_cfg(cfg_, registry, default_args) for cfg_ in cfg
+        ]
+        return nn.Sequential(*modules)
+    else:
+        return build_from_cfg(cfg, registry, default_args)
+
+
+def build_backbone(cfg):
+    """Build backbone."""
+    return build(cfg, BACKBONES)
+
+
+def build_neck(cfg):
+    """Build neck."""
+    return build(cfg, NECKS)
+
+
+def build_roi_extractor(cfg):
+    """Build roi extractor."""
+    return build(cfg, ROI_EXTRACTORS)
+
+
+def build_shared_head(cfg):
+    """Build shared head."""
+    return build(cfg, SHARED_HEADS)
+
+
+def build_head(cfg):
+    """Build head."""
+    return build(cfg, HEADS)
+
+
+def build_loss(cfg):
+    """Build loss."""
+    return build(cfg, LOSSES)
+
+
+def build_detector(cfg, train_cfg=None, test_cfg=None):
+    """Build detector."""
+    if train_cfg is not None or test_cfg is not None:
+        warnings.warn(
+            'train_cfg and test_cfg is deprecated, '
+            'please specify them in model', UserWarning)
+    assert cfg.get('train_cfg') is None or train_cfg is None, \
+        'train_cfg specified in both outer field and model field '
+    assert cfg.get('test_cfg') is None or test_cfg is None, \
+        'test_cfg specified in both outer field and model field '
+    return build(cfg, DETECTORS, dict(train_cfg=train_cfg, test_cfg=test_cfg))
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/__init__.py b/python/autoCut/models/swin/mmdet/models/dense_heads/__init__.py
new file mode 100644
index 0000000..9021dc8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/__init__.py
@@ -0,0 +1,42 @@
+from .anchor_free_head import AnchorFreeHead
+from .anchor_head import AnchorHead
+from .atss_head import ATSSHead
+from .cascade_rpn_head import CascadeRPNHead, StageCascadeRPNHead
+from .centripetal_head import CentripetalHead
+from .corner_head import CornerHead
+from .embedding_rpn_head import EmbeddingRPNHead
+from .fcos_head import FCOSHead
+from .fovea_head import FoveaHead
+from .free_anchor_retina_head import FreeAnchorRetinaHead
+from .fsaf_head import FSAFHead
+from .ga_retina_head import GARetinaHead
+from .ga_rpn_head import GARPNHead
+from .gfl_head import GFLHead
+from .guided_anchor_head import FeatureAdaption, GuidedAnchorHead
+from .ld_head import LDHead
+from .nasfcos_head import NASFCOSHead
+from .paa_head import PAAHead
+from .pisa_retinanet_head import PISARetinaHead
+from .pisa_ssd_head import PISASSDHead
+from .reppoints_head import RepPointsHead
+from .reppoints_v2_head import RepPointsV2Head
+from .retina_head import RetinaHead
+from .retina_sepbn_head import RetinaSepBNHead
+from .rpn_head import RPNHead
+from .sabl_retina_head import SABLRetinaHead
+from .ssd_head import SSDHead
+from .transformer_head import TransformerHead
+from .vfnet_head import VFNetHead
+from .yolact_head import YOLACTHead, YOLACTProtonet, YOLACTSegmHead
+from .yolo_head import YOLOV3Head
+
+__all__ = [
+    'AnchorFreeHead', 'AnchorHead', 'GuidedAnchorHead', 'FeatureAdaption',
+    'RPNHead', 'GARPNHead', 'RetinaHead', 'RetinaSepBNHead', 'GARetinaHead',
+    'SSDHead', 'FCOSHead', 'RepPointsHead', 'RepPointsV2Head', 'FoveaHead',
+    'FreeAnchorRetinaHead', 'ATSSHead', 'FSAFHead', 'NASFCOSHead',
+    'PISARetinaHead', 'PISASSDHead', 'GFLHead', 'CornerHead', 'YOLACTHead',
+    'YOLACTSegmHead', 'YOLACTProtonet', 'YOLOV3Head', 'PAAHead',
+    'SABLRetinaHead', 'CentripetalHead', 'VFNetHead', 'TransformerHead',
+    'StageCascadeRPNHead', 'CascadeRPNHead', 'EmbeddingRPNHead', 'LDHead'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/anchor_free_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/anchor_free_head.py
new file mode 100644
index 0000000..1814a0c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/anchor_free_head.py
@@ -0,0 +1,340 @@
+from abc import abstractmethod
+
+import torch
+import torch.nn as nn
+from mmcv.cnn import ConvModule, bias_init_with_prob, normal_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import multi_apply
+from ..builder import HEADS, build_loss
+from .base_dense_head import BaseDenseHead
+from .dense_test_mixins import BBoxTestMixin
+
+
+@HEADS.register_module()
+class AnchorFreeHead(BaseDenseHead, BBoxTestMixin):
+    """Anchor-free head (FCOS, Fovea, RepPoints, etc.).
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        feat_channels (int): Number of hidden channels. Used in child classes.
+        stacked_convs (int): Number of stacking convs of the head.
+        strides (tuple): Downsample factor of each feature map.
+        dcn_on_last_conv (bool): If true, use dcn in the last layer of
+            towers. Default: False.
+        conv_bias (bool | str): If specified as `auto`, it will be decided by
+            the norm_cfg. Bias of conv will be set as True if `norm_cfg` is
+            None, otherwise False. Default: "auto".
+        loss_cls (dict): Config of classification loss.
+        loss_bbox (dict): Config of localization loss.
+        conv_cfg (dict): Config dict for convolution layer. Default: None.
+        norm_cfg (dict): Config dict for normalization layer. Default: None.
+        train_cfg (dict): Training config of anchor head.
+        test_cfg (dict): Testing config of anchor head.
+    """  # noqa: W605
+
+    _version = 1
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 feat_channels=256,
+                 stacked_convs=4,
+                 strides=(4, 8, 16, 32, 64),
+                 dcn_on_last_conv=False,
+                 conv_bias='auto',
+                 loss_cls=dict(
+                     type='FocalLoss',
+                     use_sigmoid=True,
+                     gamma=2.0,
+                     alpha=0.25,
+                     loss_weight=1.0),
+                 loss_bbox=dict(type='IoULoss', loss_weight=1.0),
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 train_cfg=None,
+                 test_cfg=None):
+        super(AnchorFreeHead, self).__init__()
+        self.num_classes = num_classes
+        self.cls_out_channels = num_classes
+        self.in_channels = in_channels
+        self.feat_channels = feat_channels
+        self.stacked_convs = stacked_convs
+        self.strides = strides
+        self.dcn_on_last_conv = dcn_on_last_conv
+        assert conv_bias == 'auto' or isinstance(conv_bias, bool)
+        self.conv_bias = conv_bias
+        self.loss_cls = build_loss(loss_cls)
+        self.loss_bbox = build_loss(loss_bbox)
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.fp16_enabled = False
+
+        self._init_layers()
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self._init_cls_convs()
+        self._init_reg_convs()
+        self._init_predictor()
+
+    def _init_cls_convs(self):
+        """Initialize classification conv layers of the head."""
+        self.cls_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            if self.dcn_on_last_conv and i == self.stacked_convs - 1:
+                conv_cfg = dict(type='DCNv2')
+            else:
+                conv_cfg = self.conv_cfg
+            self.cls_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=self.norm_cfg,
+                    bias=self.conv_bias))
+
+    def _init_reg_convs(self):
+        """Initialize bbox regression conv layers of the head."""
+        self.reg_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            if self.dcn_on_last_conv and i == self.stacked_convs - 1:
+                conv_cfg = dict(type='DCNv2')
+            else:
+                conv_cfg = self.conv_cfg
+            self.reg_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=self.norm_cfg,
+                    bias=self.conv_bias))
+
+    def _init_predictor(self):
+        """Initialize predictor layers of the head."""
+        self.conv_cls = nn.Conv2d(
+            self.feat_channels, self.cls_out_channels, 3, padding=1)
+        self.conv_reg = nn.Conv2d(self.feat_channels, 4, 3, padding=1)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.cls_convs:
+            if isinstance(m.conv, nn.Conv2d):
+                normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            if isinstance(m.conv, nn.Conv2d):
+                normal_init(m.conv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.conv_cls, std=0.01, bias=bias_cls)
+        normal_init(self.conv_reg, std=0.01)
+
+    def _load_from_state_dict(self, state_dict, prefix, local_metadata, strict,
+                              missing_keys, unexpected_keys, error_msgs):
+        """Hack some keys of the model state dict so that can load checkpoints
+        of previous version."""
+        version = local_metadata.get('version', None)
+        if version is None:
+            # the key is different in early versions
+            # for example, 'fcos_cls' become 'conv_cls' now
+            bbox_head_keys = [
+                k for k in state_dict.keys() if k.startswith(prefix)
+            ]
+            ori_predictor_keys = []
+            new_predictor_keys = []
+            # e.g. 'fcos_cls' or 'fcos_reg'
+            for key in bbox_head_keys:
+                ori_predictor_keys.append(key)
+                key = key.split('.')
+                conv_name = None
+                if key[1].endswith('cls'):
+                    conv_name = 'conv_cls'
+                elif key[1].endswith('reg'):
+                    conv_name = 'conv_reg'
+                elif key[1].endswith('centerness'):
+                    conv_name = 'conv_centerness'
+                else:
+                    assert NotImplementedError
+                if conv_name is not None:
+                    key[1] = conv_name
+                    new_predictor_keys.append('.'.join(key))
+                else:
+                    ori_predictor_keys.pop(-1)
+            for i in range(len(new_predictor_keys)):
+                state_dict[new_predictor_keys[i]] = state_dict.pop(
+                    ori_predictor_keys[i])
+        super()._load_from_state_dict(state_dict, prefix, local_metadata,
+                                      strict, missing_keys, unexpected_keys,
+                                      error_msgs)
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple: Usually contain classification scores and bbox predictions.
+                cls_scores (list[Tensor]): Box scores for each scale level,
+                    each is a 4D-tensor, the channel number is
+                    num_points * num_classes.
+                bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                    level, each is a 4D-tensor, the channel number is
+                    num_points * 4.
+        """
+        return multi_apply(self.forward_single, feats)[:2]
+
+    def forward_single(self, x):
+        """Forward features of a single scale level.
+
+        Args:
+            x (Tensor): FPN feature maps of the specified stride.
+
+        Returns:
+            tuple: Scores for each class, bbox predictions, features
+                after classification and regression conv layers, some
+                models needs these features like FCOS.
+        """
+        cls_feat = x
+        reg_feat = x
+
+        for cls_layer in self.cls_convs:
+            cls_feat = cls_layer(cls_feat)
+        cls_score = self.conv_cls(cls_feat)
+
+        for reg_layer in self.reg_convs:
+            reg_feat = reg_layer(reg_feat)
+        bbox_pred = self.conv_reg(reg_feat)
+        return cls_score, bbox_pred, cls_feat, reg_feat
+
+    @abstractmethod
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute loss of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level,
+                each is a 4D-tensor, the channel number is
+                num_points * num_classes.
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level, each is a 4D-tensor, the channel number is
+                num_points * 4.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+        """
+
+        raise NotImplementedError
+
+    @abstractmethod
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   img_metas,
+                   cfg=None,
+                   rescale=None):
+        """Transform network output for a batch into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_points * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_points * 4, H, W)
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used
+            rescale (bool): If True, return boxes in original image space
+        """
+
+        raise NotImplementedError
+
+    @abstractmethod
+    def get_targets(self, points, gt_bboxes_list, gt_labels_list):
+        """Compute regression, classification and centerness targets for points
+        in multiple images.
+
+        Args:
+            points (list[Tensor]): Points of each fpn level, each has shape
+                (num_points, 2).
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image,
+                each has shape (num_gt, 4).
+            gt_labels_list (list[Tensor]): Ground truth labels of each box,
+                each has shape (num_gt,).
+        """
+        raise NotImplementedError
+
+    def _get_points_single(self,
+                           featmap_size,
+                           stride,
+                           dtype,
+                           device,
+                           flatten=False):
+        """Get points of a single scale level."""
+        h, w = featmap_size
+        x_range = torch.arange(w, dtype=dtype, device=device)
+        y_range = torch.arange(h, dtype=dtype, device=device)
+        y, x = torch.meshgrid(y_range, x_range)
+        if flatten:
+            y = y.flatten()
+            x = x.flatten()
+        return y, x
+
+    def get_points(self, featmap_sizes, dtype, device, flatten=False):
+        """Get points according to feature map sizes.
+
+        Args:
+            featmap_sizes (list[tuple]): Multi-level feature map sizes.
+            dtype (torch.dtype): Type of points.
+            device (torch.device): Device of points.
+
+        Returns:
+            tuple: points of each image.
+        """
+        mlvl_points = []
+        for i in range(len(featmap_sizes)):
+            mlvl_points.append(
+                self._get_points_single(featmap_sizes[i], self.strides[i],
+                                        dtype, device, flatten))
+        return mlvl_points
+
+    def aug_test(self, feats, img_metas, rescale=False):
+        """Test function with test time augmentation.
+
+        Args:
+            feats (list[Tensor]): the outer list indicates test-time
+                augmentations and inner Tensor should have a shape NxCxHxW,
+                which contains features for all images in the batch.
+            img_metas (list[list[dict]]): the outer list indicates test-time
+                augs (multiscale, flip, etc.) and the inner list indicates
+                images in a batch. each dict has image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[ndarray]: bbox results of each class
+        """
+        return self.aug_test_bboxes(feats, img_metas, rescale=rescale)
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/anchor_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/anchor_head.py
new file mode 100644
index 0000000..eea7352
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/anchor_head.py
@@ -0,0 +1,751 @@
+import torch
+import torch.nn as nn
+from mmcv.cnn import normal_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import (anchor_inside_flags, build_anchor_generator,
+                        build_assigner, build_bbox_coder, build_sampler,
+                        images_to_levels, multi_apply, multiclass_nms, unmap)
+from ..builder import HEADS, build_loss
+from .base_dense_head import BaseDenseHead
+from .dense_test_mixins import BBoxTestMixin
+
+
+@HEADS.register_module()
+class AnchorHead(BaseDenseHead, BBoxTestMixin):
+    """Anchor-based head (RPN, RetinaNet, SSD, etc.).
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        feat_channels (int): Number of hidden channels. Used in child classes.
+        anchor_generator (dict): Config dict for anchor generator
+        bbox_coder (dict): Config of bounding box coder.
+        reg_decoded_bbox (bool): If true, the regression loss would be
+            applied directly on decoded bounding boxes, converting both
+            the predicted boxes and regression targets to absolute
+            coordinates format. Default False. It should be `True` when
+            using `IoULoss`, `GIoULoss`, or `DIoULoss` in the bbox head.
+        loss_cls (dict): Config of classification loss.
+        loss_bbox (dict): Config of localization loss.
+        train_cfg (dict): Training config of anchor head.
+        test_cfg (dict): Testing config of anchor head.
+    """  # noqa: W605
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 feat_channels=256,
+                 anchor_generator=dict(
+                     type='AnchorGenerator',
+                     scales=[8, 16, 32],
+                     ratios=[0.5, 1.0, 2.0],
+                     strides=[4, 8, 16, 32, 64]),
+                 bbox_coder=dict(
+                     type='DeltaXYWHBBoxCoder',
+                     clip_border=True,
+                     target_means=(.0, .0, .0, .0),
+                     target_stds=(1.0, 1.0, 1.0, 1.0)),
+                 reg_decoded_bbox=False,
+                 loss_cls=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.0),
+                 loss_bbox=dict(
+                     type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0),
+                 train_cfg=None,
+                 test_cfg=None):
+        super(AnchorHead, self).__init__()
+        self.in_channels = in_channels
+        self.num_classes = num_classes
+        self.feat_channels = feat_channels
+        self.use_sigmoid_cls = loss_cls.get('use_sigmoid', False)
+        # TODO better way to determine whether sample or not
+        self.sampling = loss_cls['type'] not in [
+            'FocalLoss', 'GHMC', 'QualityFocalLoss'
+        ]
+        if self.use_sigmoid_cls:
+            self.cls_out_channels = num_classes
+        else:
+            self.cls_out_channels = num_classes + 1
+
+        if self.cls_out_channels <= 0:
+            raise ValueError(f'num_classes={num_classes} is too small')
+        self.reg_decoded_bbox = reg_decoded_bbox
+
+        self.bbox_coder = build_bbox_coder(bbox_coder)
+        self.loss_cls = build_loss(loss_cls)
+        self.loss_bbox = build_loss(loss_bbox)
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            # use PseudoSampler when sampling is False
+            if self.sampling and hasattr(self.train_cfg, 'sampler'):
+                sampler_cfg = self.train_cfg.sampler
+            else:
+                sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+        self.fp16_enabled = False
+
+        self.anchor_generator = build_anchor_generator(anchor_generator)
+        # usually the numbers of anchors for each level are the same
+        # except SSD detectors
+        self.num_anchors = self.anchor_generator.num_base_anchors[0]
+        self._init_layers()
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.conv_cls = nn.Conv2d(self.in_channels,
+                                  self.num_anchors * self.cls_out_channels, 1)
+        self.conv_reg = nn.Conv2d(self.in_channels, self.num_anchors * 4, 1)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        normal_init(self.conv_cls, std=0.01)
+        normal_init(self.conv_reg, std=0.01)
+
+    def forward_single(self, x):
+        """Forward feature of a single scale level.
+
+        Args:
+            x (Tensor): Features of a single scale level.
+
+        Returns:
+            tuple:
+                cls_score (Tensor): Cls scores for a single scale level \
+                    the channels number is num_anchors * num_classes.
+                bbox_pred (Tensor): Box energies / deltas for a single scale \
+                    level, the channels number is num_anchors * 4.
+        """
+        cls_score = self.conv_cls(x)
+        bbox_pred = self.conv_reg(x)
+        return cls_score, bbox_pred
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple: A tuple of classification scores and bbox prediction.
+
+                - cls_scores (list[Tensor]): Classification scores for all \
+                    scale levels, each is a 4D-tensor, the channels number \
+                    is num_anchors * num_classes.
+                - bbox_preds (list[Tensor]): Box energies / deltas for all \
+                    scale levels, each is a 4D-tensor, the channels number \
+                    is num_anchors * 4.
+        """
+        return multi_apply(self.forward_single, feats)
+
+    def get_anchors(self, featmap_sizes, img_metas, device='cuda'):
+        """Get anchors according to feature map sizes.
+
+        Args:
+            featmap_sizes (list[tuple]): Multi-level feature map sizes.
+            img_metas (list[dict]): Image meta info.
+            device (torch.device | str): Device for returned tensors
+
+        Returns:
+            tuple:
+                anchor_list (list[Tensor]): Anchors of each image.
+                valid_flag_list (list[Tensor]): Valid flags of each image.
+        """
+        num_imgs = len(img_metas)
+
+        # since feature map sizes of all images are the same, we only compute
+        # anchors for one time
+        multi_level_anchors = self.anchor_generator.grid_anchors(
+            featmap_sizes, device)
+        anchor_list = [multi_level_anchors for _ in range(num_imgs)]
+
+        # for each image, we compute valid flags of multi level anchors
+        valid_flag_list = []
+        for img_id, img_meta in enumerate(img_metas):
+            multi_level_flags = self.anchor_generator.valid_flags(
+                featmap_sizes, img_meta['pad_shape'], device)
+            valid_flag_list.append(multi_level_flags)
+
+        return anchor_list, valid_flag_list
+
+    def _get_targets_single(self,
+                            flat_anchors,
+                            valid_flags,
+                            gt_bboxes,
+                            gt_bboxes_ignore,
+                            gt_labels,
+                            img_meta,
+                            label_channels=1,
+                            unmap_outputs=True):
+        """Compute regression and classification targets for anchors in a
+        single image.
+
+        Args:
+            flat_anchors (Tensor): Multi-level anchors of the image, which are
+                concatenated into a single tensor of shape (num_anchors ,4)
+            valid_flags (Tensor): Multi level valid flags of the image,
+                which are concatenated into a single tensor of
+                    shape (num_anchors,).
+            gt_bboxes (Tensor): Ground truth bboxes of the image,
+                shape (num_gts, 4).
+            gt_bboxes_ignore (Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+            img_meta (dict): Meta info of the image.
+            gt_labels (Tensor): Ground truth labels of each box,
+                shape (num_gts,).
+            label_channels (int): Channel of label.
+            unmap_outputs (bool): Whether to map outputs back to the original
+                set of anchors.
+
+        Returns:
+            tuple:
+                labels_list (list[Tensor]): Labels of each level
+                label_weights_list (list[Tensor]): Label weights of each level
+                bbox_targets_list (list[Tensor]): BBox targets of each level
+                bbox_weights_list (list[Tensor]): BBox weights of each level
+                num_total_pos (int): Number of positive samples in all images
+                num_total_neg (int): Number of negative samples in all images
+        """
+        inside_flags = anchor_inside_flags(flat_anchors, valid_flags,
+                                           img_meta['img_shape'][:2],
+                                           self.train_cfg.allowed_border)
+        if not inside_flags.any():
+            return (None, ) * 7
+        # assign gt and sample anchors
+        anchors = flat_anchors[inside_flags, :]
+
+        assign_result = self.assigner.assign(
+            anchors, gt_bboxes, gt_bboxes_ignore,
+            None if self.sampling else gt_labels)
+        sampling_result = self.sampler.sample(assign_result, anchors,
+                                              gt_bboxes)
+
+        num_valid_anchors = anchors.shape[0]
+        bbox_targets = torch.zeros_like(anchors)
+        bbox_weights = torch.zeros_like(anchors)
+        labels = anchors.new_full((num_valid_anchors, ),
+                                  self.num_classes,
+                                  dtype=torch.long)
+        label_weights = anchors.new_zeros(num_valid_anchors, dtype=torch.float)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+        if len(pos_inds) > 0:
+            if not self.reg_decoded_bbox:
+                pos_bbox_targets = self.bbox_coder.encode(
+                    sampling_result.pos_bboxes, sampling_result.pos_gt_bboxes)
+            else:
+                pos_bbox_targets = sampling_result.pos_gt_bboxes
+            bbox_targets[pos_inds, :] = pos_bbox_targets
+            bbox_weights[pos_inds, :] = 1.0
+            if gt_labels is None:
+                # Only rpn gives gt_labels as None
+                # Foreground is the first class since v2.5.0
+                labels[pos_inds] = 0
+            else:
+                labels[pos_inds] = gt_labels[
+                    sampling_result.pos_assigned_gt_inds]
+            if self.train_cfg.pos_weight <= 0:
+                label_weights[pos_inds] = 1.0
+            else:
+                label_weights[pos_inds] = self.train_cfg.pos_weight
+        if len(neg_inds) > 0:
+            label_weights[neg_inds] = 1.0
+
+        # map up to original set of anchors
+        if unmap_outputs:
+            num_total_anchors = flat_anchors.size(0)
+            labels = unmap(
+                labels, num_total_anchors, inside_flags,
+                fill=self.num_classes)  # fill bg label
+            label_weights = unmap(label_weights, num_total_anchors,
+                                  inside_flags)
+            bbox_targets = unmap(bbox_targets, num_total_anchors, inside_flags)
+            bbox_weights = unmap(bbox_weights, num_total_anchors, inside_flags)
+
+        return (labels, label_weights, bbox_targets, bbox_weights, pos_inds,
+                neg_inds, sampling_result)
+
+    def get_targets(self,
+                    anchor_list,
+                    valid_flag_list,
+                    gt_bboxes_list,
+                    img_metas,
+                    gt_bboxes_ignore_list=None,
+                    gt_labels_list=None,
+                    label_channels=1,
+                    unmap_outputs=True,
+                    return_sampling_results=False):
+        """Compute regression and classification targets for anchors in
+        multiple images.
+
+        Args:
+            anchor_list (list[list[Tensor]]): Multi level anchors of each
+                image. The outer list indicates images, and the inner list
+                corresponds to feature levels of the image. Each element of
+                the inner list is a tensor of shape (num_anchors, 4).
+            valid_flag_list (list[list[Tensor]]): Multi level valid flags of
+                each image. The outer list indicates images, and the inner list
+                corresponds to feature levels of the image. Each element of
+                the inner list is a tensor of shape (num_anchors, )
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image.
+            img_metas (list[dict]): Meta info of each image.
+            gt_bboxes_ignore_list (list[Tensor]): Ground truth bboxes to be
+                ignored.
+            gt_labels_list (list[Tensor]): Ground truth labels of each box.
+            label_channels (int): Channel of label.
+            unmap_outputs (bool): Whether to map outputs back to the original
+                set of anchors.
+
+        Returns:
+            tuple: Usually returns a tuple containing learning targets.
+
+                - labels_list (list[Tensor]): Labels of each level.
+                - label_weights_list (list[Tensor]): Label weights of each \
+                    level.
+                - bbox_targets_list (list[Tensor]): BBox targets of each level.
+                - bbox_weights_list (list[Tensor]): BBox weights of each level.
+                - num_total_pos (int): Number of positive samples in all \
+                    images.
+                - num_total_neg (int): Number of negative samples in all \
+                    images.
+            additional_returns: This function enables user-defined returns from
+                `self._get_targets_single`. These returns are currently refined
+                to properties at each feature map (i.e. having HxW dimension).
+                The results will be concatenated after the end
+        """
+        num_imgs = len(img_metas)
+        assert len(anchor_list) == len(valid_flag_list) == num_imgs
+
+        # anchor number of multi levels
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+        # concat all level anchors to a single tensor
+        concat_anchor_list = []
+        concat_valid_flag_list = []
+        for i in range(num_imgs):
+            assert len(anchor_list[i]) == len(valid_flag_list[i])
+            concat_anchor_list.append(torch.cat(anchor_list[i]))
+            concat_valid_flag_list.append(torch.cat(valid_flag_list[i]))
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        results = multi_apply(
+            self._get_targets_single,
+            concat_anchor_list,
+            concat_valid_flag_list,
+            gt_bboxes_list,
+            gt_bboxes_ignore_list,
+            gt_labels_list,
+            img_metas,
+            label_channels=label_channels,
+            unmap_outputs=unmap_outputs)
+        (all_labels, all_label_weights, all_bbox_targets, all_bbox_weights,
+         pos_inds_list, neg_inds_list, sampling_results_list) = results[:7]
+        rest_results = list(results[7:])  # user-added return values
+        # no valid anchors
+        if any([labels is None for labels in all_labels]):
+            return None
+        # sampled anchors of all images
+        num_total_pos = sum([max(inds.numel(), 1) for inds in pos_inds_list])
+        num_total_neg = sum([max(inds.numel(), 1) for inds in neg_inds_list])
+        # split targets to a list w.r.t. multiple levels
+        labels_list = images_to_levels(all_labels, num_level_anchors)
+        label_weights_list = images_to_levels(all_label_weights,
+                                              num_level_anchors)
+        bbox_targets_list = images_to_levels(all_bbox_targets,
+                                             num_level_anchors)
+        bbox_weights_list = images_to_levels(all_bbox_weights,
+                                             num_level_anchors)
+        res = (labels_list, label_weights_list, bbox_targets_list,
+               bbox_weights_list, num_total_pos, num_total_neg)
+        if return_sampling_results:
+            res = res + (sampling_results_list, )
+        for i, r in enumerate(rest_results):  # user-added return values
+            rest_results[i] = images_to_levels(r, num_level_anchors)
+
+        return res + tuple(rest_results)
+
+    def loss_single(self, cls_score, bbox_pred, anchors, labels, label_weights,
+                    bbox_targets, bbox_weights, num_total_samples):
+        """Compute loss of a single scale level.
+
+        Args:
+            cls_score (Tensor): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W).
+            bbox_pred (Tensor): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W).
+            anchors (Tensor): Box reference for each scale level with shape
+                (N, num_total_anchors, 4).
+            labels (Tensor): Labels of each anchors with shape
+                (N, num_total_anchors).
+            label_weights (Tensor): Label weights of each anchor with shape
+                (N, num_total_anchors)
+            bbox_targets (Tensor): BBox regression targets of each anchor wight
+                shape (N, num_total_anchors, 4).
+            bbox_weights (Tensor): BBox regression loss weights of each anchor
+                with shape (N, num_total_anchors, 4).
+            num_total_samples (int): If sampling, num total samples equal to
+                the number of total anchors; Otherwise, it is the number of
+                positive anchors.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        # classification loss
+        labels = labels.reshape(-1)
+        label_weights = label_weights.reshape(-1)
+        cls_score = cls_score.permute(0, 2, 3,
+                                      1).reshape(-1, self.cls_out_channels)
+        loss_cls = self.loss_cls(
+            cls_score, labels, label_weights, avg_factor=num_total_samples)
+        # regression loss
+        bbox_targets = bbox_targets.reshape(-1, 4)
+        bbox_weights = bbox_weights.reshape(-1, 4)
+        bbox_pred = bbox_pred.permute(0, 2, 3, 1).reshape(-1, 4)
+        if self.reg_decoded_bbox:
+            # When the regression loss (e.g. `IouLoss`, `GIouLoss`)
+            # is applied directly on the decoded bounding boxes, it
+            # decodes the already encoded coordinates to absolute format.
+            anchors = anchors.reshape(-1, 4)
+            bbox_pred = self.bbox_coder.decode(anchors, bbox_pred)
+        loss_bbox = self.loss_bbox(
+            bbox_pred,
+            bbox_targets,
+            bbox_weights,
+            avg_factor=num_total_samples)
+        return loss_cls, loss_bbox
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss. Default: None
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg) = cls_reg_targets
+        num_total_samples = (
+            num_total_pos + num_total_neg if self.sampling else num_total_pos)
+
+        # anchor number of multi levels
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+        # concat all level anchors and flags to a single tensor
+        concat_anchor_list = []
+        for i in range(len(anchor_list)):
+            concat_anchor_list.append(torch.cat(anchor_list[i]))
+        all_anchor_list = images_to_levels(concat_anchor_list,
+                                           num_level_anchors)
+
+        losses_cls, losses_bbox = multi_apply(
+            self.loss_single,
+            cls_scores,
+            bbox_preds,
+            all_anchor_list,
+            labels_list,
+            label_weights_list,
+            bbox_targets_list,
+            bbox_weights_list,
+            num_total_samples=num_total_samples)
+        return dict(loss_cls=losses_cls, loss_bbox=losses_bbox)
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   img_metas,
+                   cfg=None,
+                   rescale=False,
+                   with_nms=True):
+        """Transform network output for a batch into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each level in the
+                feature pyramid, has shape
+                (N, num_anchors * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box energies / deltas for each
+                level in the feature pyramid, has shape
+                (N, num_anchors * 4, H, W).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where 5 represent
+                (tl_x, tl_y, br_x, br_y, score) and the score between 0 and 1.
+                The shape of the second tensor in the tuple is (n,), and
+                each element represents the class label of the corresponding
+                box.
+
+        Example:
+            >>> import mmcv
+            >>> self = AnchorHead(
+            >>>     num_classes=9,
+            >>>     in_channels=1,
+            >>>     anchor_generator=dict(
+            >>>         type='AnchorGenerator',
+            >>>         scales=[8],
+            >>>         ratios=[0.5, 1.0, 2.0],
+            >>>         strides=[4,]))
+            >>> img_metas = [{'img_shape': (32, 32, 3), 'scale_factor': 1}]
+            >>> cfg = mmcv.Config(dict(
+            >>>     score_thr=0.00,
+            >>>     nms=dict(type='nms', iou_thr=1.0),
+            >>>     max_per_img=10))
+            >>> feat = torch.rand(1, 1, 3, 3)
+            >>> cls_score, bbox_pred = self.forward_single(feat)
+            >>> # note the input lists are over different levels, not images
+            >>> cls_scores, bbox_preds = [cls_score], [bbox_pred]
+            >>> result_list = self.get_bboxes(cls_scores, bbox_preds,
+            >>>                               img_metas, cfg)
+            >>> det_bboxes, det_labels = result_list[0]
+            >>> assert len(result_list) == 1
+            >>> assert det_bboxes.shape[1] == 5
+            >>> assert len(det_bboxes) == len(det_labels) == cfg.max_per_img
+        """
+        assert len(cls_scores) == len(bbox_preds)
+        num_levels = len(cls_scores)
+
+        device = cls_scores[0].device
+        featmap_sizes = [cls_scores[i].shape[-2:] for i in range(num_levels)]
+        mlvl_anchors = self.anchor_generator.grid_anchors(
+            featmap_sizes, device=device)
+
+        mlvl_cls_scores = [cls_scores[i].detach() for i in range(num_levels)]
+        mlvl_bbox_preds = [bbox_preds[i].detach() for i in range(num_levels)]
+
+        if torch.onnx.is_in_onnx_export():
+            assert len(
+                img_metas
+            ) == 1, 'Only support one input image while in exporting to ONNX'
+            img_shapes = img_metas[0]['img_shape_for_onnx']
+        else:
+            img_shapes = [
+                img_metas[i]['img_shape']
+                for i in range(cls_scores[0].shape[0])
+            ]
+        scale_factors = [
+            img_metas[i]['scale_factor'] for i in range(cls_scores[0].shape[0])
+        ]
+
+        if with_nms:
+            # some heads don't support with_nms argument
+            result_list = self._get_bboxes(mlvl_cls_scores, mlvl_bbox_preds,
+                                           mlvl_anchors, img_shapes,
+                                           scale_factors, cfg, rescale)
+        else:
+            result_list = self._get_bboxes(mlvl_cls_scores, mlvl_bbox_preds,
+                                           mlvl_anchors, img_shapes,
+                                           scale_factors, cfg, rescale,
+                                           with_nms)
+        return result_list
+
+    def _get_bboxes(self,
+                    mlvl_cls_scores,
+                    mlvl_bbox_preds,
+                    mlvl_anchors,
+                    img_shapes,
+                    scale_factors,
+                    cfg,
+                    rescale=False,
+                    with_nms=True):
+        """Transform outputs for a batch item into bbox predictions.
+
+        Args:
+            mlvl_cls_scores (list[Tensor]): Each element in the list is
+                the scores of bboxes of single level in the feature pyramid,
+                has shape (N, num_anchors * num_classes, H, W).
+            mlvl_bbox_preds (list[Tensor]):  Each element in the list is the
+                bboxes predictions of single level in the feature pyramid,
+                has shape (N, num_anchors * 4, H, W).
+            mlvl_anchors (list[Tensor]): Each element in the list is
+                the anchors of single level in feature pyramid, has shape
+                (num_anchors, 4).
+            img_shapes (list[tuple[int]]): Each tuple in the list represent
+                the shape(height, width, 3) of single image in the batch.
+            scale_factors (list[ndarray]): Scale factor of the batch
+                image arange as list[(w_scale, h_scale, w_scale, h_scale)].
+            cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where 5 represent
+                (tl_x, tl_y, br_x, br_y, score) and the score between 0 and 1.
+                The shape of the second tensor in the tuple is (n,), and
+                each element represents the class label of the corresponding
+                box.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(mlvl_cls_scores) == len(mlvl_bbox_preds) == len(
+            mlvl_anchors)
+        batch_size = mlvl_cls_scores[0].shape[0]
+        # convert to tensor to keep tracing
+        nms_pre_tensor = torch.tensor(
+            cfg.get('nms_pre', -1),
+            device=mlvl_cls_scores[0].device,
+            dtype=torch.long)
+
+        mlvl_bboxes = []
+        mlvl_scores = []
+        for cls_score, bbox_pred, anchors in zip(mlvl_cls_scores,
+                                                 mlvl_bbox_preds,
+                                                 mlvl_anchors):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            cls_score = cls_score.permute(0, 2, 3,
+                                          1).reshape(batch_size, -1,
+                                                     self.cls_out_channels)
+            if self.use_sigmoid_cls:
+                scores = cls_score.sigmoid()
+            else:
+                scores = cls_score.softmax(-1)
+            bbox_pred = bbox_pred.permute(0, 2, 3,
+                                          1).reshape(batch_size, -1, 4)
+            anchors = anchors.expand_as(bbox_pred)
+            # Always keep topk op for dynamic input in onnx
+            if nms_pre_tensor > 0 and (torch.onnx.is_in_onnx_export()
+                                       or scores.shape[-2] > nms_pre_tensor):
+                from torch import _shape_as_tensor
+                # keep shape as tensor and get k
+                num_anchor = _shape_as_tensor(scores)[-2].to(
+                    nms_pre_tensor.device)
+                nms_pre = torch.where(nms_pre_tensor < num_anchor,
+                                      nms_pre_tensor, num_anchor)
+
+                # Get maximum scores for foreground classes.
+                if self.use_sigmoid_cls:
+                    max_scores, _ = scores.max(-1)
+                else:
+                    # remind that we set FG labels to [0, num_class-1]
+                    # since mmdet v2.0
+                    # BG cat_id: num_class
+                    max_scores, _ = scores[..., :-1].max(-1)
+
+                _, topk_inds = max_scores.topk(nms_pre)
+                batch_inds = torch.arange(batch_size).view(
+                    -1, 1).expand_as(topk_inds)
+                anchors = anchors[batch_inds, topk_inds, :]
+                bbox_pred = bbox_pred[batch_inds, topk_inds, :]
+                scores = scores[batch_inds, topk_inds, :]
+
+            bboxes = self.bbox_coder.decode(
+                anchors, bbox_pred, max_shape=img_shapes)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+
+        batch_mlvl_bboxes = torch.cat(mlvl_bboxes, dim=1)
+        if rescale:
+            batch_mlvl_bboxes /= batch_mlvl_bboxes.new_tensor(
+                scale_factors).unsqueeze(1)
+        batch_mlvl_scores = torch.cat(mlvl_scores, dim=1)
+
+        # Set max number of box to be feed into nms in deployment
+        deploy_nms_pre = cfg.get('deploy_nms_pre', -1)
+        if deploy_nms_pre > 0 and torch.onnx.is_in_onnx_export():
+            # Get maximum scores for foreground classes.
+            if self.use_sigmoid_cls:
+                max_scores, _ = batch_mlvl_scores.max(-1)
+            else:
+                # remind that we set FG labels to [0, num_class-1]
+                # since mmdet v2.0
+                # BG cat_id: num_class
+                max_scores, _ = batch_mlvl_scores[..., :-1].max(-1)
+            _, topk_inds = max_scores.topk(deploy_nms_pre)
+            batch_inds = torch.arange(batch_size).view(-1,
+                                                       1).expand_as(topk_inds)
+            batch_mlvl_scores = batch_mlvl_scores[batch_inds, topk_inds]
+            batch_mlvl_bboxes = batch_mlvl_bboxes[batch_inds, topk_inds]
+        if self.use_sigmoid_cls:
+            # Add a dummy background class to the backend when using sigmoid
+            # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+            # BG cat_id: num_class
+            padding = batch_mlvl_scores.new_zeros(batch_size,
+                                                  batch_mlvl_scores.shape[1],
+                                                  1)
+            batch_mlvl_scores = torch.cat([batch_mlvl_scores, padding], dim=-1)
+
+        if with_nms:
+            det_results = []
+            for (mlvl_bboxes, mlvl_scores) in zip(batch_mlvl_bboxes,
+                                                  batch_mlvl_scores):
+                det_bbox, det_label = multiclass_nms(mlvl_bboxes, mlvl_scores,
+                                                     cfg.score_thr, cfg.nms,
+                                                     cfg.max_per_img)
+                det_results.append(tuple([det_bbox, det_label]))
+        else:
+            det_results = [
+                tuple(mlvl_bs)
+                for mlvl_bs in zip(batch_mlvl_bboxes, batch_mlvl_scores)
+            ]
+        return det_results
+
+    def aug_test(self, feats, img_metas, rescale=False):
+        """Test function with test time augmentation.
+
+        Args:
+            feats (list[Tensor]): the outer list indicates test-time
+                augmentations and inner Tensor should have a shape NxCxHxW,
+                which contains features for all images in the batch.
+            img_metas (list[list[dict]]): the outer list indicates test-time
+                augs (multiscale, flip, etc.) and the inner list indicates
+                images in a batch. each dict has image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[ndarray]: bbox results of each class
+        """
+        return self.aug_test_bboxes(feats, img_metas, rescale=rescale)
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/atss_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/atss_head.py
new file mode 100644
index 0000000..ff55dfa
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/atss_head.py
@@ -0,0 +1,689 @@
+import torch
+import torch.nn as nn
+from mmcv.cnn import ConvModule, Scale, bias_init_with_prob, normal_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import (anchor_inside_flags, build_assigner, build_sampler,
+                        images_to_levels, multi_apply, multiclass_nms,
+                        reduce_mean, unmap)
+from ..builder import HEADS, build_loss
+from .anchor_head import AnchorHead
+
+EPS = 1e-12
+
+
+@HEADS.register_module()
+class ATSSHead(AnchorHead):
+    """Bridging the Gap Between Anchor-based and Anchor-free Detection via
+    Adaptive Training Sample Selection.
+
+    ATSS head structure is similar with FCOS, however ATSS use anchor boxes
+    and assign label by Adaptive Training Sample Selection instead max-iou.
+
+    https://arxiv.org/abs/1912.02424
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 stacked_convs=4,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='GN', num_groups=32, requires_grad=True),
+                 loss_centerness=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.0),
+                 **kwargs):
+        self.stacked_convs = stacked_convs
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        super(ATSSHead, self).__init__(num_classes, in_channels, **kwargs)
+
+        self.sampling = False
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            # SSD sampling=False so use PseudoSampler
+            sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+        self.loss_centerness = build_loss(loss_centerness)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.relu = nn.ReLU(inplace=True)
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            self.cls_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+            self.reg_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+        self.atss_cls = nn.Conv2d(
+            self.feat_channels,
+            self.num_anchors * self.cls_out_channels,
+            3,
+            padding=1)
+        self.atss_reg = nn.Conv2d(
+            self.feat_channels, self.num_anchors * 4, 3, padding=1)
+        self.atss_centerness = nn.Conv2d(
+            self.feat_channels, self.num_anchors * 1, 3, padding=1)
+        self.scales = nn.ModuleList(
+            [Scale(1.0) for _ in self.anchor_generator.strides])
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.cls_convs:
+            normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            normal_init(m.conv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.atss_cls, std=0.01, bias=bias_cls)
+        normal_init(self.atss_reg, std=0.01)
+        normal_init(self.atss_centerness, std=0.01)
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple: Usually a tuple of classification scores and bbox prediction
+                cls_scores (list[Tensor]): Classification scores for all scale
+                    levels, each is a 4D-tensor, the channels number is
+                    num_anchors * num_classes.
+                bbox_preds (list[Tensor]): Box energies / deltas for all scale
+                    levels, each is a 4D-tensor, the channels number is
+                    num_anchors * 4.
+        """
+        return multi_apply(self.forward_single, feats, self.scales)
+
+    def forward_single(self, x, scale):
+        """Forward feature of a single scale level.
+
+        Args:
+            x (Tensor): Features of a single scale level.
+            scale (:obj: `mmcv.cnn.Scale`): Learnable scale module to resize
+                the bbox prediction.
+
+        Returns:
+            tuple:
+                cls_score (Tensor): Cls scores for a single scale level
+                    the channels number is num_anchors * num_classes.
+                bbox_pred (Tensor): Box energies / deltas for a single scale
+                    level, the channels number is num_anchors * 4.
+                centerness (Tensor): Centerness for a single scale level, the
+                    channel number is (N, num_anchors * 1, H, W).
+        """
+        cls_feat = x
+        reg_feat = x
+        for cls_conv in self.cls_convs:
+            cls_feat = cls_conv(cls_feat)
+        for reg_conv in self.reg_convs:
+            reg_feat = reg_conv(reg_feat)
+        cls_score = self.atss_cls(cls_feat)
+        # we just follow atss, not apply exp in bbox_pred
+        bbox_pred = scale(self.atss_reg(reg_feat)).float()
+        centerness = self.atss_centerness(reg_feat)
+        return cls_score, bbox_pred, centerness
+
+    def loss_single(self, anchors, cls_score, bbox_pred, centerness, labels,
+                    label_weights, bbox_targets, num_total_samples):
+        """Compute loss of a single scale level.
+
+        Args:
+            cls_score (Tensor): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W).
+            bbox_pred (Tensor): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W).
+            anchors (Tensor): Box reference for each scale level with shape
+                (N, num_total_anchors, 4).
+            labels (Tensor): Labels of each anchors with shape
+                (N, num_total_anchors).
+            label_weights (Tensor): Label weights of each anchor with shape
+                (N, num_total_anchors)
+            bbox_targets (Tensor): BBox regression targets of each anchor wight
+                shape (N, num_total_anchors, 4).
+            num_total_samples (int): Number os positive samples that is
+                reduced over all GPUs.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+
+        anchors = anchors.reshape(-1, 4)
+        cls_score = cls_score.permute(0, 2, 3, 1).reshape(
+            -1, self.cls_out_channels).contiguous()
+        bbox_pred = bbox_pred.permute(0, 2, 3, 1).reshape(-1, 4)
+        centerness = centerness.permute(0, 2, 3, 1).reshape(-1)
+        bbox_targets = bbox_targets.reshape(-1, 4)
+        labels = labels.reshape(-1)
+        label_weights = label_weights.reshape(-1)
+
+        # classification loss
+        loss_cls = self.loss_cls(
+            cls_score, labels, label_weights, avg_factor=num_total_samples)
+
+        # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+        bg_class_ind = self.num_classes
+        pos_inds = ((labels >= 0)
+                    & (labels < bg_class_ind)).nonzero().squeeze(1)
+
+        if len(pos_inds) > 0:
+            pos_bbox_targets = bbox_targets[pos_inds]
+            pos_bbox_pred = bbox_pred[pos_inds]
+            pos_anchors = anchors[pos_inds]
+            pos_centerness = centerness[pos_inds]
+
+            centerness_targets = self.centerness_target(
+                pos_anchors, pos_bbox_targets)
+            pos_decode_bbox_pred = self.bbox_coder.decode(
+                pos_anchors, pos_bbox_pred)
+            pos_decode_bbox_targets = self.bbox_coder.decode(
+                pos_anchors, pos_bbox_targets)
+
+            # regression loss
+            loss_bbox = self.loss_bbox(
+                pos_decode_bbox_pred,
+                pos_decode_bbox_targets,
+                weight=centerness_targets,
+                avg_factor=1.0)
+
+            # centerness loss
+            loss_centerness = self.loss_centerness(
+                pos_centerness,
+                centerness_targets,
+                avg_factor=num_total_samples)
+
+        else:
+            loss_bbox = bbox_pred.sum() * 0
+            loss_centerness = centerness.sum() * 0
+            centerness_targets = bbox_targets.new_tensor(0.)
+
+        return loss_cls, loss_bbox, loss_centerness, centerness_targets.sum()
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds', 'centernesses'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             centernesses,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            centernesses (list[Tensor]): Centerness for each scale
+                level with shape (N, num_anchors * 1, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (list[Tensor] | None): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels)
+        if cls_reg_targets is None:
+            return None
+
+        (anchor_list, labels_list, label_weights_list, bbox_targets_list,
+         bbox_weights_list, num_total_pos, num_total_neg) = cls_reg_targets
+
+        num_total_samples = reduce_mean(
+            torch.tensor(num_total_pos, dtype=torch.float,
+                         device=device)).item()
+        num_total_samples = max(num_total_samples, 1.0)
+
+        losses_cls, losses_bbox, loss_centerness,\
+            bbox_avg_factor = multi_apply(
+                self.loss_single,
+                anchor_list,
+                cls_scores,
+                bbox_preds,
+                centernesses,
+                labels_list,
+                label_weights_list,
+                bbox_targets_list,
+                num_total_samples=num_total_samples)
+
+        bbox_avg_factor = sum(bbox_avg_factor)
+        bbox_avg_factor = reduce_mean(bbox_avg_factor).item()
+        if bbox_avg_factor < EPS:
+            bbox_avg_factor = 1
+        losses_bbox = list(map(lambda x: x / bbox_avg_factor, losses_bbox))
+        return dict(
+            loss_cls=losses_cls,
+            loss_bbox=losses_bbox,
+            loss_centerness=loss_centerness)
+
+    def centerness_target(self, anchors, bbox_targets):
+        # only calculate pos centerness targets, otherwise there may be nan
+        gts = self.bbox_coder.decode(anchors, bbox_targets)
+        anchors_cx = (anchors[:, 2] + anchors[:, 0]) / 2
+        anchors_cy = (anchors[:, 3] + anchors[:, 1]) / 2
+        l_ = anchors_cx - gts[:, 0]
+        t_ = anchors_cy - gts[:, 1]
+        r_ = gts[:, 2] - anchors_cx
+        b_ = gts[:, 3] - anchors_cy
+
+        left_right = torch.stack([l_, r_], dim=1)
+        top_bottom = torch.stack([t_, b_], dim=1)
+        centerness = torch.sqrt(
+            (left_right.min(dim=-1)[0] / left_right.max(dim=-1)[0]) *
+            (top_bottom.min(dim=-1)[0] / top_bottom.max(dim=-1)[0]))
+        assert not torch.isnan(centerness).any()
+        return centerness
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds', 'centernesses'))
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   centernesses,
+                   img_metas,
+                   cfg=None,
+                   rescale=False,
+                   with_nms=True):
+        """Transform network output for a batch into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                with shape (N, num_anchors * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W).
+            centernesses (list[Tensor]): Centerness for each scale level with
+                shape (N, num_anchors * 1, H, W).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used. Default: None.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where 5 represent
+                (tl_x, tl_y, br_x, br_y, score) and the score between 0 and 1.
+                The shape of the second tensor in the tuple is (n,), and
+                each element represents the class label of the corresponding
+                box.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_scores) == len(bbox_preds)
+        num_levels = len(cls_scores)
+        device = cls_scores[0].device
+        featmap_sizes = [cls_scores[i].shape[-2:] for i in range(num_levels)]
+        mlvl_anchors = self.anchor_generator.grid_anchors(
+            featmap_sizes, device=device)
+
+        cls_score_list = [cls_scores[i].detach() for i in range(num_levels)]
+        bbox_pred_list = [bbox_preds[i].detach() for i in range(num_levels)]
+        centerness_pred_list = [
+            centernesses[i].detach() for i in range(num_levels)
+        ]
+        img_shapes = [
+            img_metas[i]['img_shape'] for i in range(cls_scores[0].shape[0])
+        ]
+        scale_factors = [
+            img_metas[i]['scale_factor'] for i in range(cls_scores[0].shape[0])
+        ]
+        result_list = self._get_bboxes(cls_score_list, bbox_pred_list,
+                                       centerness_pred_list, mlvl_anchors,
+                                       img_shapes, scale_factors, cfg, rescale,
+                                       with_nms)
+        return result_list
+
+    def _get_bboxes(self,
+                    cls_scores,
+                    bbox_preds,
+                    centernesses,
+                    mlvl_anchors,
+                    img_shapes,
+                    scale_factors,
+                    cfg,
+                    rescale=False,
+                    with_nms=True):
+        """Transform outputs for a single batch item into labeled boxes.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for a single scale level
+                with shape (N, num_anchors * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box energies / deltas for a single
+                scale level with shape (N, num_anchors * 4, H, W).
+            centernesses (list[Tensor]): Centerness for a single scale level
+                with shape (N, num_anchors * 1, H, W).
+            mlvl_anchors (list[Tensor]): Box reference for a single scale level
+                with shape (num_total_anchors, 4).
+            img_shapes (list[tuple[int]]): Shape of the input image,
+                list[(height, width, 3)].
+            scale_factors (list[ndarray]): Scale factor of the image arrange as
+                (w_scale, h_scale, w_scale, h_scale).
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where 5 represent
+                (tl_x, tl_y, br_x, br_y, score) and the score between 0 and 1.
+                The shape of the second tensor in the tuple is (n,), and
+                each element represents the class label of the corresponding
+                box.
+        """
+        assert len(cls_scores) == len(bbox_preds) == len(mlvl_anchors)
+        device = cls_scores[0].device
+        batch_size = cls_scores[0].shape[0]
+        # convert to tensor to keep tracing
+        nms_pre_tensor = torch.tensor(
+            cfg.get('nms_pre', -1), device=device, dtype=torch.long)
+        mlvl_bboxes = []
+        mlvl_scores = []
+        mlvl_centerness = []
+        for cls_score, bbox_pred, centerness, anchors in zip(
+                cls_scores, bbox_preds, centernesses, mlvl_anchors):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            scores = cls_score.permute(0, 2, 3, 1).reshape(
+                batch_size, -1, self.cls_out_channels).sigmoid()
+            centerness = centerness.permute(0, 2, 3,
+                                            1).reshape(batch_size,
+                                                       -1).sigmoid()
+            bbox_pred = bbox_pred.permute(0, 2, 3,
+                                          1).reshape(batch_size, -1, 4)
+
+            # Always keep topk op for dynamic input in onnx
+            if nms_pre_tensor > 0 and (torch.onnx.is_in_onnx_export()
+                                       or scores.shape[-2] > nms_pre_tensor):
+                from torch import _shape_as_tensor
+                # keep shape as tensor and get k
+                num_anchor = _shape_as_tensor(scores)[-2].to(device)
+                nms_pre = torch.where(nms_pre_tensor < num_anchor,
+                                      nms_pre_tensor, num_anchor)
+
+                max_scores, _ = (scores * centerness[..., None]).max(-1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                anchors = anchors[topk_inds, :]
+                batch_inds = torch.arange(batch_size).view(
+                    -1, 1).expand_as(topk_inds).long()
+                bbox_pred = bbox_pred[batch_inds, topk_inds, :]
+                scores = scores[batch_inds, topk_inds, :]
+                centerness = centerness[batch_inds, topk_inds]
+            else:
+                anchors = anchors.expand_as(bbox_pred)
+
+            bboxes = self.bbox_coder.decode(
+                anchors, bbox_pred, max_shape=img_shapes)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+            mlvl_centerness.append(centerness)
+
+        batch_mlvl_bboxes = torch.cat(mlvl_bboxes, dim=1)
+        if rescale:
+            batch_mlvl_bboxes /= batch_mlvl_bboxes.new_tensor(
+                scale_factors).unsqueeze(1)
+        batch_mlvl_scores = torch.cat(mlvl_scores, dim=1)
+        batch_mlvl_centerness = torch.cat(mlvl_centerness, dim=1)
+
+        # Set max number of box to be feed into nms in deployment
+        deploy_nms_pre = cfg.get('deploy_nms_pre', -1)
+        if deploy_nms_pre > 0 and torch.onnx.is_in_onnx_export():
+            batch_mlvl_scores, _ = (
+                batch_mlvl_scores *
+                batch_mlvl_centerness.unsqueeze(2).expand_as(batch_mlvl_scores)
+            ).max(-1)
+            _, topk_inds = batch_mlvl_scores.topk(deploy_nms_pre)
+            batch_inds = torch.arange(batch_size).view(-1,
+                                                       1).expand_as(topk_inds)
+            batch_mlvl_scores = batch_mlvl_scores[batch_inds, topk_inds, :]
+            batch_mlvl_bboxes = batch_mlvl_bboxes[batch_inds, topk_inds, :]
+            batch_mlvl_centerness = batch_mlvl_centerness[batch_inds,
+                                                          topk_inds]
+        # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+        # BG cat_id: num_class
+        padding = batch_mlvl_scores.new_zeros(batch_size,
+                                              batch_mlvl_scores.shape[1], 1)
+        batch_mlvl_scores = torch.cat([batch_mlvl_scores, padding], dim=-1)
+
+        if with_nms:
+            det_results = []
+            for (mlvl_bboxes, mlvl_scores,
+                 mlvl_centerness) in zip(batch_mlvl_bboxes, batch_mlvl_scores,
+                                         batch_mlvl_centerness):
+                det_bbox, det_label = multiclass_nms(
+                    mlvl_bboxes,
+                    mlvl_scores,
+                    cfg.score_thr,
+                    cfg.nms,
+                    cfg.max_per_img,
+                    score_factors=mlvl_centerness)
+                det_results.append(tuple([det_bbox, det_label]))
+        else:
+            det_results = [
+                tuple(mlvl_bs)
+                for mlvl_bs in zip(batch_mlvl_bboxes, batch_mlvl_scores,
+                                   batch_mlvl_centerness)
+            ]
+        return det_results
+
+    def get_targets(self,
+                    anchor_list,
+                    valid_flag_list,
+                    gt_bboxes_list,
+                    img_metas,
+                    gt_bboxes_ignore_list=None,
+                    gt_labels_list=None,
+                    label_channels=1,
+                    unmap_outputs=True):
+        """Get targets for ATSS head.
+
+        This method is almost the same as `AnchorHead.get_targets()`. Besides
+        returning the targets as the parent method does, it also returns the
+        anchors as the first element of the returned tuple.
+        """
+        num_imgs = len(img_metas)
+        assert len(anchor_list) == len(valid_flag_list) == num_imgs
+
+        # anchor number of multi levels
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+        num_level_anchors_list = [num_level_anchors] * num_imgs
+
+        # concat all level anchors and flags to a single tensor
+        for i in range(num_imgs):
+            assert len(anchor_list[i]) == len(valid_flag_list[i])
+            anchor_list[i] = torch.cat(anchor_list[i])
+            valid_flag_list[i] = torch.cat(valid_flag_list[i])
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        (all_anchors, all_labels, all_label_weights, all_bbox_targets,
+         all_bbox_weights, pos_inds_list, neg_inds_list) = multi_apply(
+             self._get_target_single,
+             anchor_list,
+             valid_flag_list,
+             num_level_anchors_list,
+             gt_bboxes_list,
+             gt_bboxes_ignore_list,
+             gt_labels_list,
+             img_metas,
+             label_channels=label_channels,
+             unmap_outputs=unmap_outputs)
+        # no valid anchors
+        if any([labels is None for labels in all_labels]):
+            return None
+        # sampled anchors of all images
+        num_total_pos = sum([max(inds.numel(), 1) for inds in pos_inds_list])
+        num_total_neg = sum([max(inds.numel(), 1) for inds in neg_inds_list])
+        # split targets to a list w.r.t. multiple levels
+        anchors_list = images_to_levels(all_anchors, num_level_anchors)
+        labels_list = images_to_levels(all_labels, num_level_anchors)
+        label_weights_list = images_to_levels(all_label_weights,
+                                              num_level_anchors)
+        bbox_targets_list = images_to_levels(all_bbox_targets,
+                                             num_level_anchors)
+        bbox_weights_list = images_to_levels(all_bbox_weights,
+                                             num_level_anchors)
+        return (anchors_list, labels_list, label_weights_list,
+                bbox_targets_list, bbox_weights_list, num_total_pos,
+                num_total_neg)
+
+    def _get_target_single(self,
+                           flat_anchors,
+                           valid_flags,
+                           num_level_anchors,
+                           gt_bboxes,
+                           gt_bboxes_ignore,
+                           gt_labels,
+                           img_meta,
+                           label_channels=1,
+                           unmap_outputs=True):
+        """Compute regression, classification targets for anchors in a single
+        image.
+
+        Args:
+            flat_anchors (Tensor): Multi-level anchors of the image, which are
+                concatenated into a single tensor of shape (num_anchors ,4)
+            valid_flags (Tensor): Multi level valid flags of the image,
+                which are concatenated into a single tensor of
+                    shape (num_anchors,).
+            num_level_anchors Tensor): Number of anchors of each scale level.
+            gt_bboxes (Tensor): Ground truth bboxes of the image,
+                shape (num_gts, 4).
+            gt_bboxes_ignore (Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+            gt_labels (Tensor): Ground truth labels of each box,
+                shape (num_gts,).
+            img_meta (dict): Meta info of the image.
+            label_channels (int): Channel of label.
+            unmap_outputs (bool): Whether to map outputs back to the original
+                set of anchors.
+
+        Returns:
+            tuple: N is the number of total anchors in the image.
+                labels (Tensor): Labels of all anchors in the image with shape
+                    (N,).
+                label_weights (Tensor): Label weights of all anchor in the
+                    image with shape (N,).
+                bbox_targets (Tensor): BBox targets of all anchors in the
+                    image with shape (N, 4).
+                bbox_weights (Tensor): BBox weights of all anchors in the
+                    image with shape (N, 4)
+                pos_inds (Tensor): Indices of positive anchor with shape
+                    (num_pos,).
+                neg_inds (Tensor): Indices of negative anchor with shape
+                    (num_neg,).
+        """
+        inside_flags = anchor_inside_flags(flat_anchors, valid_flags,
+                                           img_meta['img_shape'][:2],
+                                           self.train_cfg.allowed_border)
+        if not inside_flags.any():
+            return (None, ) * 7
+        # assign gt and sample anchors
+        anchors = flat_anchors[inside_flags, :]
+
+        num_level_anchors_inside = self.get_num_level_anchors_inside(
+            num_level_anchors, inside_flags)
+        assign_result = self.assigner.assign(anchors, num_level_anchors_inside,
+                                             gt_bboxes, gt_bboxes_ignore,
+                                             gt_labels)
+
+        sampling_result = self.sampler.sample(assign_result, anchors,
+                                              gt_bboxes)
+
+        num_valid_anchors = anchors.shape[0]
+        bbox_targets = torch.zeros_like(anchors)
+        bbox_weights = torch.zeros_like(anchors)
+        labels = anchors.new_full((num_valid_anchors, ),
+                                  self.num_classes,
+                                  dtype=torch.long)
+        label_weights = anchors.new_zeros(num_valid_anchors, dtype=torch.float)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+        if len(pos_inds) > 0:
+            if hasattr(self, 'bbox_coder'):
+                pos_bbox_targets = self.bbox_coder.encode(
+                    sampling_result.pos_bboxes, sampling_result.pos_gt_bboxes)
+            else:
+                # used in VFNetHead
+                pos_bbox_targets = sampling_result.pos_gt_bboxes
+            bbox_targets[pos_inds, :] = pos_bbox_targets
+            bbox_weights[pos_inds, :] = 1.0
+            if gt_labels is None:
+                # Only rpn gives gt_labels as None
+                # Foreground is the first class since v2.5.0
+                labels[pos_inds] = 0
+            else:
+                labels[pos_inds] = gt_labels[
+                    sampling_result.pos_assigned_gt_inds]
+            if self.train_cfg.pos_weight <= 0:
+                label_weights[pos_inds] = 1.0
+            else:
+                label_weights[pos_inds] = self.train_cfg.pos_weight
+        if len(neg_inds) > 0:
+            label_weights[neg_inds] = 1.0
+
+        # map up to original set of anchors
+        if unmap_outputs:
+            num_total_anchors = flat_anchors.size(0)
+            anchors = unmap(anchors, num_total_anchors, inside_flags)
+            labels = unmap(
+                labels, num_total_anchors, inside_flags, fill=self.num_classes)
+            label_weights = unmap(label_weights, num_total_anchors,
+                                  inside_flags)
+            bbox_targets = unmap(bbox_targets, num_total_anchors, inside_flags)
+            bbox_weights = unmap(bbox_weights, num_total_anchors, inside_flags)
+
+        return (anchors, labels, label_weights, bbox_targets, bbox_weights,
+                pos_inds, neg_inds)
+
+    def get_num_level_anchors_inside(self, num_level_anchors, inside_flags):
+        split_inside_flags = torch.split(inside_flags, num_level_anchors)
+        num_level_anchors_inside = [
+            int(flags.sum()) for flags in split_inside_flags
+        ]
+        return num_level_anchors_inside
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/base_dense_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/base_dense_head.py
new file mode 100644
index 0000000..de11e4a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/base_dense_head.py
@@ -0,0 +1,59 @@
+from abc import ABCMeta, abstractmethod
+
+import torch.nn as nn
+
+
+class BaseDenseHead(nn.Module, metaclass=ABCMeta):
+    """Base class for DenseHeads."""
+
+    def __init__(self):
+        super(BaseDenseHead, self).__init__()
+
+    @abstractmethod
+    def loss(self, **kwargs):
+        """Compute losses of the head."""
+        pass
+
+    @abstractmethod
+    def get_bboxes(self, **kwargs):
+        """Transform network output for a batch into bbox predictions."""
+        pass
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels=None,
+                      gt_bboxes_ignore=None,
+                      proposal_cfg=None,
+                      **kwargs):
+        """
+        Args:
+            x (list[Tensor]): Features from FPN.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes (Tensor): Ground truth bboxes of the image,
+                shape (num_gts, 4).
+            gt_labels (Tensor): Ground truth labels of each box,
+                shape (num_gts,).
+            gt_bboxes_ignore (Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+            proposal_cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used
+
+        Returns:
+            tuple:
+                losses: (dict[str, Tensor]): A dictionary of loss components.
+                proposal_list (list[Tensor]): Proposals of each image.
+        """
+        outs = self(x)
+        if gt_labels is None:
+            loss_inputs = outs + (gt_bboxes, img_metas)
+        else:
+            loss_inputs = outs + (gt_bboxes, gt_labels, img_metas)
+        losses = self.loss(*loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore)
+        if proposal_cfg is None:
+            return losses
+        else:
+            proposal_list = self.get_bboxes(*outs, img_metas, cfg=proposal_cfg)
+            return losses, proposal_list
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/cascade_rpn_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/cascade_rpn_head.py
new file mode 100644
index 0000000..e32ee46
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/cascade_rpn_head.py
@@ -0,0 +1,784 @@
+from __future__ import division
+import copy
+import warnings
+
+import torch
+import torch.nn as nn
+from mmcv import ConfigDict
+from mmcv.cnn import normal_init
+from mmcv.ops import DeformConv2d, batched_nms
+
+from mmdet.core import (RegionAssigner, build_assigner, build_sampler,
+                        images_to_levels, multi_apply)
+from ..builder import HEADS, build_head
+from .base_dense_head import BaseDenseHead
+from .rpn_head import RPNHead
+
+
+class AdaptiveConv(nn.Module):
+    """AdaptiveConv used to adapt the sampling location with the anchors.
+
+    Args:
+        in_channels (int): Number of channels in the input image
+        out_channels (int): Number of channels produced by the convolution
+        kernel_size (int or tuple): Size of the conv kernel. Default: 3
+        stride (int or tuple, optional): Stride of the convolution. Default: 1
+        padding (int or tuple, optional): Zero-padding added to both sides of
+            the input. Default: 1
+        dilation (int or tuple, optional): Spacing between kernel elements.
+            Default: 3
+        groups (int, optional): Number of blocked connections from input
+            channels to output channels. Default: 1
+        bias (bool, optional): If set True, adds a learnable bias to the
+            output. Default: False.
+        type (str, optional): Type of adaptive conv, can be either 'offset'
+            (arbitrary anchors) or 'dilation' (uniform anchor).
+            Default: 'dilation'.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 kernel_size=3,
+                 stride=1,
+                 padding=1,
+                 dilation=3,
+                 groups=1,
+                 bias=False,
+                 type='dilation'):
+        super(AdaptiveConv, self).__init__()
+        assert type in ['offset', 'dilation']
+        self.adapt_type = type
+
+        assert kernel_size == 3, 'Adaptive conv only supports kernels 3'
+        if self.adapt_type == 'offset':
+            assert stride == 1 and padding == 1 and groups == 1, \
+                'Adaptive conv offset mode only supports padding: {1}, ' \
+                f'stride: {1}, groups: {1}'
+            self.conv = DeformConv2d(
+                in_channels,
+                out_channels,
+                kernel_size,
+                padding=padding,
+                stride=stride,
+                groups=groups,
+                bias=bias)
+        else:
+            self.conv = nn.Conv2d(
+                in_channels,
+                out_channels,
+                kernel_size,
+                padding=dilation,
+                dilation=dilation)
+
+    def init_weights(self):
+        """Init weights."""
+        normal_init(self.conv, std=0.01)
+
+    def forward(self, x, offset):
+        """Forward function."""
+        if self.adapt_type == 'offset':
+            N, _, H, W = x.shape
+            assert offset is not None
+            assert H * W == offset.shape[1]
+            # reshape [N, NA, 18] to (N, 18, H, W)
+            offset = offset.permute(0, 2, 1).reshape(N, -1, H, W)
+            offset = offset.contiguous()
+            x = self.conv(x, offset)
+        else:
+            assert offset is None
+            x = self.conv(x)
+        return x
+
+
+@HEADS.register_module()
+class StageCascadeRPNHead(RPNHead):
+    """Stage of CascadeRPNHead.
+
+    Args:
+        in_channels (int): Number of channels in the input feature map.
+        anchor_generator (dict): anchor generator config.
+        adapt_cfg (dict): adaptation config.
+        bridged_feature (bool, optional): whether update rpn feature.
+            Default: False.
+        with_cls (bool, optional): wheather use classification branch.
+            Default: True.
+        sampling (bool, optional): wheather use sampling. Default: True.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 anchor_generator=dict(
+                     type='AnchorGenerator',
+                     scales=[8],
+                     ratios=[1.0],
+                     strides=[4, 8, 16, 32, 64]),
+                 adapt_cfg=dict(type='dilation', dilation=3),
+                 bridged_feature=False,
+                 with_cls=True,
+                 sampling=True,
+                 **kwargs):
+        self.with_cls = with_cls
+        self.anchor_strides = anchor_generator['strides']
+        self.anchor_scales = anchor_generator['scales']
+        self.bridged_feature = bridged_feature
+        self.adapt_cfg = adapt_cfg
+        super(StageCascadeRPNHead, self).__init__(
+            in_channels, anchor_generator=anchor_generator, **kwargs)
+
+        # override sampling and sampler
+        self.sampling = sampling
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            # use PseudoSampler when sampling is False
+            if self.sampling and hasattr(self.train_cfg, 'sampler'):
+                sampler_cfg = self.train_cfg.sampler
+            else:
+                sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+
+    def _init_layers(self):
+        """Init layers of a CascadeRPN stage."""
+        self.rpn_conv = AdaptiveConv(self.in_channels, self.feat_channels,
+                                     **self.adapt_cfg)
+        if self.with_cls:
+            self.rpn_cls = nn.Conv2d(self.feat_channels,
+                                     self.num_anchors * self.cls_out_channels,
+                                     1)
+        self.rpn_reg = nn.Conv2d(self.feat_channels, self.num_anchors * 4, 1)
+        self.relu = nn.ReLU(inplace=True)
+
+    def init_weights(self):
+        """Init weights of a CascadeRPN stage."""
+        self.rpn_conv.init_weights()
+        normal_init(self.rpn_reg, std=0.01)
+        if self.with_cls:
+            normal_init(self.rpn_cls, std=0.01)
+
+    def forward_single(self, x, offset):
+        """Forward function of single scale."""
+        bridged_x = x
+        x = self.relu(self.rpn_conv(x, offset))
+        if self.bridged_feature:
+            bridged_x = x  # update feature
+        cls_score = self.rpn_cls(x) if self.with_cls else None
+        bbox_pred = self.rpn_reg(x)
+        return bridged_x, cls_score, bbox_pred
+
+    def forward(self, feats, offset_list=None):
+        """Forward function."""
+        if offset_list is None:
+            offset_list = [None for _ in range(len(feats))]
+        return multi_apply(self.forward_single, feats, offset_list)
+
+    def _region_targets_single(self,
+                               anchors,
+                               valid_flags,
+                               gt_bboxes,
+                               gt_bboxes_ignore,
+                               gt_labels,
+                               img_meta,
+                               featmap_sizes,
+                               label_channels=1):
+        """Get anchor targets based on region for single level."""
+        assign_result = self.assigner.assign(
+            anchors,
+            valid_flags,
+            gt_bboxes,
+            img_meta,
+            featmap_sizes,
+            self.anchor_scales[0],
+            self.anchor_strides,
+            gt_bboxes_ignore=gt_bboxes_ignore,
+            gt_labels=None,
+            allowed_border=self.train_cfg.allowed_border)
+        flat_anchors = torch.cat(anchors)
+        sampling_result = self.sampler.sample(assign_result, flat_anchors,
+                                              gt_bboxes)
+
+        num_anchors = flat_anchors.shape[0]
+        bbox_targets = torch.zeros_like(flat_anchors)
+        bbox_weights = torch.zeros_like(flat_anchors)
+        labels = flat_anchors.new_zeros(num_anchors, dtype=torch.long)
+        label_weights = flat_anchors.new_zeros(num_anchors, dtype=torch.float)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+        if len(pos_inds) > 0:
+            if not self.reg_decoded_bbox:
+                pos_bbox_targets = self.bbox_coder.encode(
+                    sampling_result.pos_bboxes, sampling_result.pos_gt_bboxes)
+            else:
+                pos_bbox_targets = sampling_result.pos_gt_bboxes
+            bbox_targets[pos_inds, :] = pos_bbox_targets
+            bbox_weights[pos_inds, :] = 1.0
+            if gt_labels is None:
+                labels[pos_inds] = 1
+            else:
+                labels[pos_inds] = gt_labels[
+                    sampling_result.pos_assigned_gt_inds]
+            if self.train_cfg.pos_weight <= 0:
+                label_weights[pos_inds] = 1.0
+            else:
+                label_weights[pos_inds] = self.train_cfg.pos_weight
+        if len(neg_inds) > 0:
+            label_weights[neg_inds] = 1.0
+
+        return (labels, label_weights, bbox_targets, bbox_weights, pos_inds,
+                neg_inds)
+
+    def region_targets(self,
+                       anchor_list,
+                       valid_flag_list,
+                       gt_bboxes_list,
+                       img_metas,
+                       featmap_sizes,
+                       gt_bboxes_ignore_list=None,
+                       gt_labels_list=None,
+                       label_channels=1,
+                       unmap_outputs=True):
+        """See :func:`StageCascadeRPNHead.get_targets`."""
+        num_imgs = len(img_metas)
+        assert len(anchor_list) == len(valid_flag_list) == num_imgs
+
+        # anchor number of multi levels
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        (all_labels, all_label_weights, all_bbox_targets, all_bbox_weights,
+         pos_inds_list, neg_inds_list) = multi_apply(
+             self._region_targets_single,
+             anchor_list,
+             valid_flag_list,
+             gt_bboxes_list,
+             gt_bboxes_ignore_list,
+             gt_labels_list,
+             img_metas,
+             featmap_sizes=featmap_sizes,
+             label_channels=label_channels)
+        # no valid anchors
+        if any([labels is None for labels in all_labels]):
+            return None
+        # sampled anchors of all images
+        num_total_pos = sum([max(inds.numel(), 1) for inds in pos_inds_list])
+        num_total_neg = sum([max(inds.numel(), 1) for inds in neg_inds_list])
+        # split targets to a list w.r.t. multiple levels
+        labels_list = images_to_levels(all_labels, num_level_anchors)
+        label_weights_list = images_to_levels(all_label_weights,
+                                              num_level_anchors)
+        bbox_targets_list = images_to_levels(all_bbox_targets,
+                                             num_level_anchors)
+        bbox_weights_list = images_to_levels(all_bbox_weights,
+                                             num_level_anchors)
+        return (labels_list, label_weights_list, bbox_targets_list,
+                bbox_weights_list, num_total_pos, num_total_neg)
+
+    def get_targets(self,
+                    anchor_list,
+                    valid_flag_list,
+                    gt_bboxes,
+                    img_metas,
+                    featmap_sizes,
+                    gt_bboxes_ignore=None,
+                    label_channels=1):
+        """Compute regression and classification targets for anchors.
+
+        Args:
+            anchor_list (list[list]): Multi level anchors of each image.
+            valid_flag_list (list[list]): Multi level valid flags of each
+                image.
+            gt_bboxes (list[Tensor]): Ground truth bboxes of each image.
+            img_metas (list[dict]): Meta info of each image.
+            featmap_sizes (list[Tensor]): Feature mapsize each level
+            gt_bboxes_ignore (list[Tensor]): Ignore bboxes of each images
+            label_channels (int): Channel of label.
+
+        Returns:
+            cls_reg_targets (tuple)
+        """
+        if isinstance(self.assigner, RegionAssigner):
+            cls_reg_targets = self.region_targets(
+                anchor_list,
+                valid_flag_list,
+                gt_bboxes,
+                img_metas,
+                featmap_sizes,
+                gt_bboxes_ignore_list=gt_bboxes_ignore,
+                label_channels=label_channels)
+        else:
+            cls_reg_targets = super(StageCascadeRPNHead, self).get_targets(
+                anchor_list,
+                valid_flag_list,
+                gt_bboxes,
+                img_metas,
+                gt_bboxes_ignore_list=gt_bboxes_ignore,
+                label_channels=label_channels)
+        return cls_reg_targets
+
+    def anchor_offset(self, anchor_list, anchor_strides, featmap_sizes):
+        """ Get offest for deformable conv based on anchor shape
+        NOTE: currently support deformable kernel_size=3 and dilation=1
+
+        Args:
+            anchor_list (list[list[tensor])): [NI, NLVL, NA, 4] list of
+                multi-level anchors
+            anchor_strides (list[int]): anchor stride of each level
+
+        Returns:
+            offset_list (list[tensor]): [NLVL, NA, 2, 18]: offset of DeformConv
+                kernel.
+        """
+
+        def _shape_offset(anchors, stride, ks=3, dilation=1):
+            # currently support kernel_size=3 and dilation=1
+            assert ks == 3 and dilation == 1
+            pad = (ks - 1) // 2
+            idx = torch.arange(-pad, pad + 1, dtype=dtype, device=device)
+            yy, xx = torch.meshgrid(idx, idx)  # return order matters
+            xx = xx.reshape(-1)
+            yy = yy.reshape(-1)
+            w = (anchors[:, 2] - anchors[:, 0]) / stride
+            h = (anchors[:, 3] - anchors[:, 1]) / stride
+            w = w / (ks - 1) - dilation
+            h = h / (ks - 1) - dilation
+            offset_x = w[:, None] * xx  # (NA, ks**2)
+            offset_y = h[:, None] * yy  # (NA, ks**2)
+            return offset_x, offset_y
+
+        def _ctr_offset(anchors, stride, featmap_size):
+            feat_h, feat_w = featmap_size
+            assert len(anchors) == feat_h * feat_w
+
+            x = (anchors[:, 0] + anchors[:, 2]) * 0.5
+            y = (anchors[:, 1] + anchors[:, 3]) * 0.5
+            # compute centers on feature map
+            x = x / stride
+            y = y / stride
+            # compute predefine centers
+            xx = torch.arange(0, feat_w, device=anchors.device)
+            yy = torch.arange(0, feat_h, device=anchors.device)
+            yy, xx = torch.meshgrid(yy, xx)
+            xx = xx.reshape(-1).type_as(x)
+            yy = yy.reshape(-1).type_as(y)
+
+            offset_x = x - xx  # (NA, )
+            offset_y = y - yy  # (NA, )
+            return offset_x, offset_y
+
+        num_imgs = len(anchor_list)
+        num_lvls = len(anchor_list[0])
+        dtype = anchor_list[0][0].dtype
+        device = anchor_list[0][0].device
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+
+        offset_list = []
+        for i in range(num_imgs):
+            mlvl_offset = []
+            for lvl in range(num_lvls):
+                c_offset_x, c_offset_y = _ctr_offset(anchor_list[i][lvl],
+                                                     anchor_strides[lvl],
+                                                     featmap_sizes[lvl])
+                s_offset_x, s_offset_y = _shape_offset(anchor_list[i][lvl],
+                                                       anchor_strides[lvl])
+
+                # offset = ctr_offset + shape_offset
+                offset_x = s_offset_x + c_offset_x[:, None]
+                offset_y = s_offset_y + c_offset_y[:, None]
+
+                # offset order (y0, x0, y1, x2, .., y8, x8, y9, x9)
+                offset = torch.stack([offset_y, offset_x], dim=-1)
+                offset = offset.reshape(offset.size(0), -1)  # [NA, 2*ks**2]
+                mlvl_offset.append(offset)
+            offset_list.append(torch.cat(mlvl_offset))  # [totalNA, 2*ks**2]
+        offset_list = images_to_levels(offset_list, num_level_anchors)
+        return offset_list
+
+    def loss_single(self, cls_score, bbox_pred, anchors, labels, label_weights,
+                    bbox_targets, bbox_weights, num_total_samples):
+        """Loss function on single scale."""
+        # classification loss
+        if self.with_cls:
+            labels = labels.reshape(-1)
+            label_weights = label_weights.reshape(-1)
+            cls_score = cls_score.permute(0, 2, 3,
+                                          1).reshape(-1, self.cls_out_channels)
+            loss_cls = self.loss_cls(
+                cls_score, labels, label_weights, avg_factor=num_total_samples)
+        # regression loss
+        bbox_targets = bbox_targets.reshape(-1, 4)
+        bbox_weights = bbox_weights.reshape(-1, 4)
+        bbox_pred = bbox_pred.permute(0, 2, 3, 1).reshape(-1, 4)
+        if self.reg_decoded_bbox:
+            # When the regression loss (e.g. `IouLoss`, `GIouLoss`)
+            # is applied directly on the decoded bounding boxes, it
+            # decodes the already encoded coordinates to absolute format.
+            anchors = anchors.reshape(-1, 4)
+            bbox_pred = self.bbox_coder.decode(anchors, bbox_pred)
+        loss_reg = self.loss_bbox(
+            bbox_pred,
+            bbox_targets,
+            bbox_weights,
+            avg_factor=num_total_samples)
+        if self.with_cls:
+            return loss_cls, loss_reg
+        return None, loss_reg
+
+    def loss(self,
+             anchor_list,
+             valid_flag_list,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            anchor_list (list[list]): Multi level anchors of each image.
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss. Default: None
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in bbox_preds]
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            featmap_sizes,
+            gt_bboxes_ignore=gt_bboxes_ignore,
+            label_channels=label_channels)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg) = cls_reg_targets
+        if self.sampling:
+            num_total_samples = num_total_pos + num_total_neg
+        else:
+            # 200 is hard-coded average factor,
+            # which follows guided anchoring.
+            num_total_samples = sum([label.numel()
+                                     for label in labels_list]) / 200.0
+
+        # change per image, per level anchor_list to per_level, per_image
+        mlvl_anchor_list = list(zip(*anchor_list))
+        # concat mlvl_anchor_list
+        mlvl_anchor_list = [
+            torch.cat(anchors, dim=0) for anchors in mlvl_anchor_list
+        ]
+
+        losses = multi_apply(
+            self.loss_single,
+            cls_scores,
+            bbox_preds,
+            mlvl_anchor_list,
+            labels_list,
+            label_weights_list,
+            bbox_targets_list,
+            bbox_weights_list,
+            num_total_samples=num_total_samples)
+        if self.with_cls:
+            return dict(loss_rpn_cls=losses[0], loss_rpn_reg=losses[1])
+        return dict(loss_rpn_reg=losses[1])
+
+    def get_bboxes(self,
+                   anchor_list,
+                   cls_scores,
+                   bbox_preds,
+                   img_metas,
+                   cfg,
+                   rescale=False):
+        """Get proposal predict."""
+        assert len(cls_scores) == len(bbox_preds)
+        num_levels = len(cls_scores)
+
+        result_list = []
+        for img_id in range(len(img_metas)):
+            cls_score_list = [
+                cls_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_pred_list = [
+                bbox_preds[i][img_id].detach() for i in range(num_levels)
+            ]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            proposals = self._get_bboxes_single(cls_score_list, bbox_pred_list,
+                                                anchor_list[img_id], img_shape,
+                                                scale_factor, cfg, rescale)
+            result_list.append(proposals)
+        return result_list
+
+    def refine_bboxes(self, anchor_list, bbox_preds, img_metas):
+        """Refine bboxes through stages."""
+        num_levels = len(bbox_preds)
+        new_anchor_list = []
+        for img_id in range(len(img_metas)):
+            mlvl_anchors = []
+            for i in range(num_levels):
+                bbox_pred = bbox_preds[i][img_id].detach()
+                bbox_pred = bbox_pred.permute(1, 2, 0).reshape(-1, 4)
+                img_shape = img_metas[img_id]['img_shape']
+                bboxes = self.bbox_coder.decode(anchor_list[img_id][i],
+                                                bbox_pred, img_shape)
+                mlvl_anchors.append(bboxes)
+            new_anchor_list.append(mlvl_anchors)
+        return new_anchor_list
+
+    # TODO: temporary plan
+    def _get_bboxes_single(self,
+                           cls_scores,
+                           bbox_preds,
+                           mlvl_anchors,
+                           img_shape,
+                           scale_factor,
+                           cfg,
+                           rescale=False):
+        """Transform outputs for a single batch item into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (num_anchors * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (num_anchors * 4, H, W).
+            mlvl_anchors (list[Tensor]): Box reference for each scale level
+                with shape (num_total_anchors, 4).
+            img_shape (tuple[int]): Shape of the input image,
+                (height, width, 3).
+            scale_factor (ndarray): Scale factor of the image arange as
+                (w_scale, h_scale, w_scale, h_scale).
+            cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+
+        Returns:
+            Tensor: Labeled boxes have the shape of (n,5), where the
+                first 4 columns are bounding box positions
+                (tl_x, tl_y, br_x, br_y) and the 5-th column is a score
+                between 0 and 1.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        cfg = copy.deepcopy(cfg)
+        # bboxes from different level should be independent during NMS,
+        # level_ids are used as labels for batched NMS to separate them
+        level_ids = []
+        mlvl_scores = []
+        mlvl_bbox_preds = []
+        mlvl_valid_anchors = []
+        for idx in range(len(cls_scores)):
+            rpn_cls_score = cls_scores[idx]
+            rpn_bbox_pred = bbox_preds[idx]
+            assert rpn_cls_score.size()[-2:] == rpn_bbox_pred.size()[-2:]
+            rpn_cls_score = rpn_cls_score.permute(1, 2, 0)
+            if self.use_sigmoid_cls:
+                rpn_cls_score = rpn_cls_score.reshape(-1)
+                scores = rpn_cls_score.sigmoid()
+            else:
+                rpn_cls_score = rpn_cls_score.reshape(-1, 2)
+                # We set FG labels to [0, num_class-1] and BG label to
+                # num_class in RPN head since mmdet v2.5, which is unified to
+                # be consistent with other head since mmdet v2.0. In mmdet v2.0
+                # to v2.4 we keep BG label as 0 and FG label as 1 in rpn head.
+                scores = rpn_cls_score.softmax(dim=1)[:, 0]
+            rpn_bbox_pred = rpn_bbox_pred.permute(1, 2, 0).reshape(-1, 4)
+            anchors = mlvl_anchors[idx]
+            if cfg.nms_pre > 0 and scores.shape[0] > cfg.nms_pre:
+                # sort is faster than topk
+                # _, topk_inds = scores.topk(cfg.nms_pre)
+                if torch.onnx.is_in_onnx_export():
+                    # sort op will be converted to TopK in onnx
+                    # and k<=3480 in TensorRT
+                    _, topk_inds = scores.topk(cfg.nms_pre)
+                    scores = scores[topk_inds]
+                else:
+                    ranked_scores, rank_inds = scores.sort(descending=True)
+                    topk_inds = rank_inds[:cfg.nms_pre]
+                    scores = ranked_scores[:cfg.nms_pre]
+                rpn_bbox_pred = rpn_bbox_pred[topk_inds, :]
+                anchors = anchors[topk_inds, :]
+            mlvl_scores.append(scores)
+            mlvl_bbox_preds.append(rpn_bbox_pred)
+            mlvl_valid_anchors.append(anchors)
+            level_ids.append(
+                scores.new_full((scores.size(0), ), idx, dtype=torch.long))
+
+        scores = torch.cat(mlvl_scores)
+        anchors = torch.cat(mlvl_valid_anchors)
+        rpn_bbox_pred = torch.cat(mlvl_bbox_preds)
+        proposals = self.bbox_coder.decode(
+            anchors, rpn_bbox_pred, max_shape=img_shape)
+        ids = torch.cat(level_ids)
+
+        # Skip nonzero op while exporting to ONNX
+        if cfg.min_bbox_size > 0 and (not torch.onnx.is_in_onnx_export()):
+            w = proposals[:, 2] - proposals[:, 0]
+            h = proposals[:, 3] - proposals[:, 1]
+            valid_inds = torch.nonzero(
+                (w >= cfg.min_bbox_size)
+                & (h >= cfg.min_bbox_size),
+                as_tuple=False).squeeze()
+            if valid_inds.sum().item() != len(proposals):
+                proposals = proposals[valid_inds, :]
+                scores = scores[valid_inds]
+                ids = ids[valid_inds]
+
+        # deprecate arguments warning
+        if 'nms' not in cfg or 'max_num' in cfg or 'nms_thr' in cfg:
+            warnings.warn(
+                'In rpn_proposal or test_cfg, '
+                'nms_thr has been moved to a dict named nms as '
+                'iou_threshold, max_num has been renamed as max_per_img, '
+                'name of original arguments and the way to specify '
+                'iou_threshold of NMS will be deprecated.')
+        if 'nms' not in cfg:
+            cfg.nms = ConfigDict(dict(type='nms', iou_threshold=cfg.nms_thr))
+        if 'max_num' in cfg:
+            if 'max_per_img' in cfg:
+                assert cfg.max_num == cfg.max_per_img, f'You ' \
+                    f'set max_num and ' \
+                    f'max_per_img at the same time, but get {cfg.max_num} ' \
+                    f'and {cfg.max_per_img} respectively' \
+                    'Please delete max_num which will be deprecated.'
+            else:
+                cfg.max_per_img = cfg.max_num
+        if 'nms_thr' in cfg:
+            assert cfg.nms.iou_threshold == cfg.nms_thr, f'You set' \
+                f' iou_threshold in nms and ' \
+                f'nms_thr at the same time, but get' \
+                f' {cfg.nms.iou_threshold} and {cfg.nms_thr}' \
+                f' respectively. Please delete the nms_thr ' \
+                f'which will be deprecated.'
+
+        dets, keep = batched_nms(proposals, scores, ids, cfg.nms)
+        return dets[:cfg.max_per_img]
+
+
+@HEADS.register_module()
+class CascadeRPNHead(BaseDenseHead):
+    """The CascadeRPNHead will predict more accurate region proposals, which is
+    required for two-stage detectors (such as Fast/Faster R-CNN). CascadeRPN
+    consists of a sequence of RPNStage to progressively improve the accuracy of
+    the detected proposals.
+
+    More details can be found in ``https://arxiv.org/abs/1909.06720``.
+
+    Args:
+        num_stages (int): number of CascadeRPN stages.
+        stages (list[dict]): list of configs to build the stages.
+        train_cfg (list[dict]): list of configs at training time each stage.
+        test_cfg (dict): config at testing time.
+    """
+
+    def __init__(self, num_stages, stages, train_cfg, test_cfg):
+        super(CascadeRPNHead, self).__init__()
+        assert num_stages == len(stages)
+        self.num_stages = num_stages
+        self.stages = nn.ModuleList()
+        for i in range(len(stages)):
+            train_cfg_i = train_cfg[i] if train_cfg is not None else None
+            stages[i].update(train_cfg=train_cfg_i)
+            stages[i].update(test_cfg=test_cfg)
+            self.stages.append(build_head(stages[i]))
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+
+    def init_weights(self):
+        """Init weight of CascadeRPN."""
+        for i in range(self.num_stages):
+            self.stages[i].init_weights()
+
+    def loss(self):
+        """loss() is implemented in StageCascadeRPNHead."""
+        pass
+
+    def get_bboxes(self):
+        """get_bboxes() is implemented in StageCascadeRPNHead."""
+        pass
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels=None,
+                      gt_bboxes_ignore=None,
+                      proposal_cfg=None):
+        """Forward train function."""
+        assert gt_labels is None, 'RPN does not require gt_labels'
+
+        featmap_sizes = [featmap.size()[-2:] for featmap in x]
+        device = x[0].device
+        anchor_list, valid_flag_list = self.stages[0].get_anchors(
+            featmap_sizes, img_metas, device=device)
+
+        losses = dict()
+
+        for i in range(self.num_stages):
+            stage = self.stages[i]
+
+            if stage.adapt_cfg['type'] == 'offset':
+                offset_list = stage.anchor_offset(anchor_list,
+                                                  stage.anchor_strides,
+                                                  featmap_sizes)
+            else:
+                offset_list = None
+            x, cls_score, bbox_pred = stage(x, offset_list)
+            rpn_loss_inputs = (anchor_list, valid_flag_list, cls_score,
+                               bbox_pred, gt_bboxes, img_metas)
+            stage_loss = stage.loss(*rpn_loss_inputs)
+            for name, value in stage_loss.items():
+                losses['s{}.{}'.format(i, name)] = value
+
+            # refine boxes
+            if i < self.num_stages - 1:
+                anchor_list = stage.refine_bboxes(anchor_list, bbox_pred,
+                                                  img_metas)
+        if proposal_cfg is None:
+            return losses
+        else:
+            proposal_list = self.stages[-1].get_bboxes(anchor_list, cls_score,
+                                                       bbox_pred, img_metas,
+                                                       self.test_cfg)
+            return losses, proposal_list
+
+    def simple_test_rpn(self, x, img_metas):
+        """Simple forward test function."""
+        featmap_sizes = [featmap.size()[-2:] for featmap in x]
+        device = x[0].device
+        anchor_list, _ = self.stages[0].get_anchors(
+            featmap_sizes, img_metas, device=device)
+
+        for i in range(self.num_stages):
+            stage = self.stages[i]
+            if stage.adapt_cfg['type'] == 'offset':
+                offset_list = stage.anchor_offset(anchor_list,
+                                                  stage.anchor_strides,
+                                                  featmap_sizes)
+            else:
+                offset_list = None
+            x, cls_score, bbox_pred = stage(x, offset_list)
+            if i < self.num_stages - 1:
+                anchor_list = stage.refine_bboxes(anchor_list, bbox_pred,
+                                                  img_metas)
+
+        proposal_list = self.stages[-1].get_bboxes(anchor_list, cls_score,
+                                                   bbox_pred, img_metas,
+                                                   self.test_cfg)
+        return proposal_list
+
+    def aug_test_rpn(self, x, img_metas):
+        """Augmented forward test function."""
+        raise NotImplementedError
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/centripetal_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/centripetal_head.py
new file mode 100644
index 0000000..6728218
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/centripetal_head.py
@@ -0,0 +1,421 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, normal_init
+from mmcv.ops import DeformConv2d
+
+from mmdet.core import multi_apply
+from ..builder import HEADS, build_loss
+from .corner_head import CornerHead
+
+
+@HEADS.register_module()
+class CentripetalHead(CornerHead):
+    """Head of CentripetalNet: Pursuing High-quality Keypoint Pairs for Object
+    Detection.
+
+    CentripetalHead inherits from :class:`CornerHead`. It removes the
+    embedding branch and adds guiding shift and centripetal shift branches.
+    More details can be found in the `paper
+    <https://arxiv.org/abs/2003.09119>`_ .
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        num_feat_levels (int): Levels of feature from the previous module. 2
+            for HourglassNet-104 and 1 for HourglassNet-52. HourglassNet-104
+            outputs the final feature and intermediate supervision feature and
+            HourglassNet-52 only outputs the final feature. Default: 2.
+        corner_emb_channels (int): Channel of embedding vector. Default: 1.
+        train_cfg (dict | None): Training config. Useless in CornerHead,
+            but we keep this variable for SingleStageDetector. Default: None.
+        test_cfg (dict | None): Testing config of CornerHead. Default: None.
+        loss_heatmap (dict | None): Config of corner heatmap loss. Default:
+            GaussianFocalLoss.
+        loss_embedding (dict | None): Config of corner embedding loss. Default:
+            AssociativeEmbeddingLoss.
+        loss_offset (dict | None): Config of corner offset loss. Default:
+            SmoothL1Loss.
+        loss_guiding_shift (dict): Config of guiding shift loss. Default:
+            SmoothL1Loss.
+        loss_centripetal_shift (dict): Config of centripetal shift loss.
+            Default: SmoothL1Loss.
+    """
+
+    def __init__(self,
+                 *args,
+                 centripetal_shift_channels=2,
+                 guiding_shift_channels=2,
+                 feat_adaption_conv_kernel=3,
+                 loss_guiding_shift=dict(
+                     type='SmoothL1Loss', beta=1.0, loss_weight=0.05),
+                 loss_centripetal_shift=dict(
+                     type='SmoothL1Loss', beta=1.0, loss_weight=1),
+                 **kwargs):
+        assert centripetal_shift_channels == 2, (
+            'CentripetalHead only support centripetal_shift_channels == 2')
+        self.centripetal_shift_channels = centripetal_shift_channels
+        assert guiding_shift_channels == 2, (
+            'CentripetalHead only support guiding_shift_channels == 2')
+        self.guiding_shift_channels = guiding_shift_channels
+        self.feat_adaption_conv_kernel = feat_adaption_conv_kernel
+        super(CentripetalHead, self).__init__(*args, **kwargs)
+        self.loss_guiding_shift = build_loss(loss_guiding_shift)
+        self.loss_centripetal_shift = build_loss(loss_centripetal_shift)
+
+    def _init_centripetal_layers(self):
+        """Initialize centripetal layers.
+
+        Including feature adaption deform convs (feat_adaption), deform offset
+        prediction convs (dcn_off), guiding shift (guiding_shift) and
+        centripetal shift ( centripetal_shift). Each branch has two parts:
+        prefix `tl_` for top-left and `br_` for bottom-right.
+        """
+        self.tl_feat_adaption = nn.ModuleList()
+        self.br_feat_adaption = nn.ModuleList()
+        self.tl_dcn_offset = nn.ModuleList()
+        self.br_dcn_offset = nn.ModuleList()
+        self.tl_guiding_shift = nn.ModuleList()
+        self.br_guiding_shift = nn.ModuleList()
+        self.tl_centripetal_shift = nn.ModuleList()
+        self.br_centripetal_shift = nn.ModuleList()
+
+        for _ in range(self.num_feat_levels):
+            self.tl_feat_adaption.append(
+                DeformConv2d(self.in_channels, self.in_channels,
+                             self.feat_adaption_conv_kernel, 1, 1))
+            self.br_feat_adaption.append(
+                DeformConv2d(self.in_channels, self.in_channels,
+                             self.feat_adaption_conv_kernel, 1, 1))
+
+            self.tl_guiding_shift.append(
+                self._make_layers(
+                    out_channels=self.guiding_shift_channels,
+                    in_channels=self.in_channels))
+            self.br_guiding_shift.append(
+                self._make_layers(
+                    out_channels=self.guiding_shift_channels,
+                    in_channels=self.in_channels))
+
+            self.tl_dcn_offset.append(
+                ConvModule(
+                    self.guiding_shift_channels,
+                    self.feat_adaption_conv_kernel**2 *
+                    self.guiding_shift_channels,
+                    1,
+                    bias=False,
+                    act_cfg=None))
+            self.br_dcn_offset.append(
+                ConvModule(
+                    self.guiding_shift_channels,
+                    self.feat_adaption_conv_kernel**2 *
+                    self.guiding_shift_channels,
+                    1,
+                    bias=False,
+                    act_cfg=None))
+
+            self.tl_centripetal_shift.append(
+                self._make_layers(
+                    out_channels=self.centripetal_shift_channels,
+                    in_channels=self.in_channels))
+            self.br_centripetal_shift.append(
+                self._make_layers(
+                    out_channels=self.centripetal_shift_channels,
+                    in_channels=self.in_channels))
+
+    def _init_layers(self):
+        """Initialize layers for CentripetalHead.
+
+        Including two parts: CornerHead layers and CentripetalHead layers
+        """
+        super()._init_layers()  # using _init_layers in CornerHead
+        self._init_centripetal_layers()
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        super().init_weights()
+        for i in range(self.num_feat_levels):
+            normal_init(self.tl_feat_adaption[i], std=0.01)
+            normal_init(self.br_feat_adaption[i], std=0.01)
+            normal_init(self.tl_dcn_offset[i].conv, std=0.1)
+            normal_init(self.br_dcn_offset[i].conv, std=0.1)
+            _ = [x.conv.reset_parameters() for x in self.tl_guiding_shift[i]]
+            _ = [x.conv.reset_parameters() for x in self.br_guiding_shift[i]]
+            _ = [
+                x.conv.reset_parameters() for x in self.tl_centripetal_shift[i]
+            ]
+            _ = [
+                x.conv.reset_parameters() for x in self.br_centripetal_shift[i]
+            ]
+
+    def forward_single(self, x, lvl_ind):
+        """Forward feature of a single level.
+
+        Args:
+            x (Tensor): Feature of a single level.
+            lvl_ind (int): Level index of current feature.
+
+        Returns:
+            tuple[Tensor]: A tuple of CentripetalHead's output for current
+            feature level. Containing the following Tensors:
+
+                - tl_heat (Tensor): Predicted top-left corner heatmap.
+                - br_heat (Tensor): Predicted bottom-right corner heatmap.
+                - tl_off (Tensor): Predicted top-left offset heatmap.
+                - br_off (Tensor): Predicted bottom-right offset heatmap.
+                - tl_guiding_shift (Tensor): Predicted top-left guiding shift
+                  heatmap.
+                - br_guiding_shift (Tensor): Predicted bottom-right guiding
+                  shift heatmap.
+                - tl_centripetal_shift (Tensor): Predicted top-left centripetal
+                  shift heatmap.
+                - br_centripetal_shift (Tensor): Predicted bottom-right
+                  centripetal shift heatmap.
+        """
+        tl_heat, br_heat, _, _, tl_off, br_off, tl_pool, br_pool = super(
+        ).forward_single(
+            x, lvl_ind, return_pool=True)
+
+        tl_guiding_shift = self.tl_guiding_shift[lvl_ind](tl_pool)
+        br_guiding_shift = self.br_guiding_shift[lvl_ind](br_pool)
+
+        tl_dcn_offset = self.tl_dcn_offset[lvl_ind](tl_guiding_shift.detach())
+        br_dcn_offset = self.br_dcn_offset[lvl_ind](br_guiding_shift.detach())
+
+        tl_feat_adaption = self.tl_feat_adaption[lvl_ind](tl_pool,
+                                                          tl_dcn_offset)
+        br_feat_adaption = self.br_feat_adaption[lvl_ind](br_pool,
+                                                          br_dcn_offset)
+
+        tl_centripetal_shift = self.tl_centripetal_shift[lvl_ind](
+            tl_feat_adaption)
+        br_centripetal_shift = self.br_centripetal_shift[lvl_ind](
+            br_feat_adaption)
+
+        result_list = [
+            tl_heat, br_heat, tl_off, br_off, tl_guiding_shift,
+            br_guiding_shift, tl_centripetal_shift, br_centripetal_shift
+        ]
+        return result_list
+
+    def loss(self,
+             tl_heats,
+             br_heats,
+             tl_offs,
+             br_offs,
+             tl_guiding_shifts,
+             br_guiding_shifts,
+             tl_centripetal_shifts,
+             br_centripetal_shifts,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            tl_heats (list[Tensor]): Top-left corner heatmaps for each level
+                with shape (N, num_classes, H, W).
+            br_heats (list[Tensor]): Bottom-right corner heatmaps for each
+                level with shape (N, num_classes, H, W).
+            tl_offs (list[Tensor]): Top-left corner offsets for each level
+                with shape (N, corner_offset_channels, H, W).
+            br_offs (list[Tensor]): Bottom-right corner offsets for each level
+                with shape (N, corner_offset_channels, H, W).
+            tl_guiding_shifts (list[Tensor]): Top-left guiding shifts for each
+                level with shape (N, guiding_shift_channels, H, W).
+            br_guiding_shifts (list[Tensor]): Bottom-right guiding shifts for
+                each level with shape (N, guiding_shift_channels, H, W).
+            tl_centripetal_shifts (list[Tensor]): Top-left centripetal shifts
+                for each level with shape (N, centripetal_shift_channels, H,
+                W).
+            br_centripetal_shifts (list[Tensor]): Bottom-right centripetal
+                shifts for each level with shape (N,
+                centripetal_shift_channels, H, W).
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [left, top, right, bottom] format.
+            gt_labels (list[Tensor]): Class indices corresponding to each box.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (list[Tensor] | None): Specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components. Containing the
+            following losses:
+
+                - det_loss (list[Tensor]): Corner keypoint losses of all
+                  feature levels.
+                - off_loss (list[Tensor]): Corner offset losses of all feature
+                  levels.
+                - guiding_loss (list[Tensor]): Guiding shift losses of all
+                  feature levels.
+                - centripetal_loss (list[Tensor]): Centripetal shift losses of
+                  all feature levels.
+        """
+        targets = self.get_targets(
+            gt_bboxes,
+            gt_labels,
+            tl_heats[-1].shape,
+            img_metas[0]['pad_shape'],
+            with_corner_emb=self.with_corner_emb,
+            with_guiding_shift=True,
+            with_centripetal_shift=True)
+        mlvl_targets = [targets for _ in range(self.num_feat_levels)]
+        [det_losses, off_losses, guiding_losses, centripetal_losses
+         ] = multi_apply(self.loss_single, tl_heats, br_heats, tl_offs,
+                         br_offs, tl_guiding_shifts, br_guiding_shifts,
+                         tl_centripetal_shifts, br_centripetal_shifts,
+                         mlvl_targets)
+        loss_dict = dict(
+            det_loss=det_losses,
+            off_loss=off_losses,
+            guiding_loss=guiding_losses,
+            centripetal_loss=centripetal_losses)
+        return loss_dict
+
+    def loss_single(self, tl_hmp, br_hmp, tl_off, br_off, tl_guiding_shift,
+                    br_guiding_shift, tl_centripetal_shift,
+                    br_centripetal_shift, targets):
+        """Compute losses for single level.
+
+        Args:
+            tl_hmp (Tensor): Top-left corner heatmap for current level with
+                shape (N, num_classes, H, W).
+            br_hmp (Tensor): Bottom-right corner heatmap for current level with
+                shape (N, num_classes, H, W).
+            tl_off (Tensor): Top-left corner offset for current level with
+                shape (N, corner_offset_channels, H, W).
+            br_off (Tensor): Bottom-right corner offset for current level with
+                shape (N, corner_offset_channels, H, W).
+            tl_guiding_shift (Tensor): Top-left guiding shift for current level
+                with shape (N, guiding_shift_channels, H, W).
+            br_guiding_shift (Tensor): Bottom-right guiding shift for current
+                level with shape (N, guiding_shift_channels, H, W).
+            tl_centripetal_shift (Tensor): Top-left centripetal shift for
+                current level with shape (N, centripetal_shift_channels, H, W).
+            br_centripetal_shift (Tensor): Bottom-right centripetal shift for
+                current level with shape (N, centripetal_shift_channels, H, W).
+            targets (dict): Corner target generated by `get_targets`.
+
+        Returns:
+            tuple[torch.Tensor]: Losses of the head's differnet branches
+            containing the following losses:
+
+                - det_loss (Tensor): Corner keypoint loss.
+                - off_loss (Tensor): Corner offset loss.
+                - guiding_loss (Tensor): Guiding shift loss.
+                - centripetal_loss (Tensor): Centripetal shift loss.
+        """
+        targets['corner_embedding'] = None
+
+        det_loss, _, _, off_loss = super().loss_single(tl_hmp, br_hmp, None,
+                                                       None, tl_off, br_off,
+                                                       targets)
+
+        gt_tl_guiding_shift = targets['topleft_guiding_shift']
+        gt_br_guiding_shift = targets['bottomright_guiding_shift']
+        gt_tl_centripetal_shift = targets['topleft_centripetal_shift']
+        gt_br_centripetal_shift = targets['bottomright_centripetal_shift']
+
+        gt_tl_heatmap = targets['topleft_heatmap']
+        gt_br_heatmap = targets['bottomright_heatmap']
+        # We only compute the offset loss at the real corner position.
+        # The value of real corner would be 1 in heatmap ground truth.
+        # The mask is computed in class agnostic mode and its shape is
+        # batch * 1 * width * height.
+        tl_mask = gt_tl_heatmap.eq(1).sum(1).gt(0).unsqueeze(1).type_as(
+            gt_tl_heatmap)
+        br_mask = gt_br_heatmap.eq(1).sum(1).gt(0).unsqueeze(1).type_as(
+            gt_br_heatmap)
+
+        # Guiding shift loss
+        tl_guiding_loss = self.loss_guiding_shift(
+            tl_guiding_shift,
+            gt_tl_guiding_shift,
+            tl_mask,
+            avg_factor=tl_mask.sum())
+        br_guiding_loss = self.loss_guiding_shift(
+            br_guiding_shift,
+            gt_br_guiding_shift,
+            br_mask,
+            avg_factor=br_mask.sum())
+        guiding_loss = (tl_guiding_loss + br_guiding_loss) / 2.0
+        # Centripetal shift loss
+        tl_centripetal_loss = self.loss_centripetal_shift(
+            tl_centripetal_shift,
+            gt_tl_centripetal_shift,
+            tl_mask,
+            avg_factor=tl_mask.sum())
+        br_centripetal_loss = self.loss_centripetal_shift(
+            br_centripetal_shift,
+            gt_br_centripetal_shift,
+            br_mask,
+            avg_factor=br_mask.sum())
+        centripetal_loss = (tl_centripetal_loss + br_centripetal_loss) / 2.0
+
+        return det_loss, off_loss, guiding_loss, centripetal_loss
+
+    def get_bboxes(self,
+                   tl_heats,
+                   br_heats,
+                   tl_offs,
+                   br_offs,
+                   tl_guiding_shifts,
+                   br_guiding_shifts,
+                   tl_centripetal_shifts,
+                   br_centripetal_shifts,
+                   img_metas,
+                   rescale=False,
+                   with_nms=True):
+        """Transform network output for a batch into bbox predictions.
+
+        Args:
+            tl_heats (list[Tensor]): Top-left corner heatmaps for each level
+                with shape (N, num_classes, H, W).
+            br_heats (list[Tensor]): Bottom-right corner heatmaps for each
+                level with shape (N, num_classes, H, W).
+            tl_offs (list[Tensor]): Top-left corner offsets for each level
+                with shape (N, corner_offset_channels, H, W).
+            br_offs (list[Tensor]): Bottom-right corner offsets for each level
+                with shape (N, corner_offset_channels, H, W).
+            tl_guiding_shifts (list[Tensor]): Top-left guiding shifts for each
+                level with shape (N, guiding_shift_channels, H, W). Useless in
+                this function, we keep this arg because it's the raw output
+                from CentripetalHead.
+            br_guiding_shifts (list[Tensor]): Bottom-right guiding shifts for
+                each level with shape (N, guiding_shift_channels, H, W).
+                Useless in this function, we keep this arg because it's the
+                raw output from CentripetalHead.
+            tl_centripetal_shifts (list[Tensor]): Top-left centripetal shifts
+                for each level with shape (N, centripetal_shift_channels, H,
+                W).
+            br_centripetal_shifts (list[Tensor]): Bottom-right centripetal
+                shifts for each level with shape (N,
+                centripetal_shift_channels, H, W).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+        """
+        assert tl_heats[-1].shape[0] == br_heats[-1].shape[0] == len(img_metas)
+        result_list = []
+        for img_id in range(len(img_metas)):
+            result_list.append(
+                self._get_bboxes_single(
+                    tl_heats[-1][img_id:img_id + 1, :],
+                    br_heats[-1][img_id:img_id + 1, :],
+                    tl_offs[-1][img_id:img_id + 1, :],
+                    br_offs[-1][img_id:img_id + 1, :],
+                    img_metas[img_id],
+                    tl_emb=None,
+                    br_emb=None,
+                    tl_centripetal_shift=tl_centripetal_shifts[-1][
+                        img_id:img_id + 1, :],
+                    br_centripetal_shift=br_centripetal_shifts[-1][
+                        img_id:img_id + 1, :],
+                    rescale=rescale,
+                    with_nms=with_nms))
+
+        return result_list
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/corner_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/corner_head.py
new file mode 100644
index 0000000..50cdb49
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/corner_head.py
@@ -0,0 +1,1074 @@
+from logging import warning
+from math import ceil, log
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, bias_init_with_prob
+from mmcv.ops import CornerPool, batched_nms
+
+from mmdet.core import multi_apply
+from ..builder import HEADS, build_loss
+from ..utils import gaussian_radius, gen_gaussian_target
+from .base_dense_head import BaseDenseHead
+
+
+class BiCornerPool(nn.Module):
+    """Bidirectional Corner Pooling Module (TopLeft, BottomRight, etc.)
+
+    Args:
+        in_channels (int): Input channels of module.
+        out_channels (int): Output channels of module.
+        feat_channels (int): Feature channels of module.
+        directions (list[str]): Directions of two CornerPools.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 directions,
+                 feat_channels=128,
+                 out_channels=128,
+                 norm_cfg=dict(type='BN', requires_grad=True)):
+        super(BiCornerPool, self).__init__()
+        self.direction1_conv = ConvModule(
+            in_channels, feat_channels, 3, padding=1, norm_cfg=norm_cfg)
+        self.direction2_conv = ConvModule(
+            in_channels, feat_channels, 3, padding=1, norm_cfg=norm_cfg)
+
+        self.aftpool_conv = ConvModule(
+            feat_channels,
+            out_channels,
+            3,
+            padding=1,
+            norm_cfg=norm_cfg,
+            act_cfg=None)
+
+        self.conv1 = ConvModule(
+            in_channels, out_channels, 1, norm_cfg=norm_cfg, act_cfg=None)
+        self.conv2 = ConvModule(
+            in_channels, out_channels, 3, padding=1, norm_cfg=norm_cfg)
+
+        self.direction1_pool = CornerPool(directions[0])
+        self.direction2_pool = CornerPool(directions[1])
+        self.relu = nn.ReLU(inplace=True)
+
+    def forward(self, x):
+        """Forward features from the upstream network.
+
+        Args:
+            x (tensor): Input feature of BiCornerPool.
+
+        Returns:
+            conv2 (tensor): Output feature of BiCornerPool.
+        """
+        direction1_conv = self.direction1_conv(x)
+        direction2_conv = self.direction2_conv(x)
+        direction1_feat = self.direction1_pool(direction1_conv)
+        direction2_feat = self.direction2_pool(direction2_conv)
+        aftpool_conv = self.aftpool_conv(direction1_feat + direction2_feat)
+        conv1 = self.conv1(x)
+        relu = self.relu(aftpool_conv + conv1)
+        conv2 = self.conv2(relu)
+        return conv2
+
+
+@HEADS.register_module()
+class CornerHead(BaseDenseHead):
+    """Head of CornerNet: Detecting Objects as Paired Keypoints.
+
+    Code is modified from the `official github repo
+    <https://github.com/princeton-vl/CornerNet/blob/master/models/py_utils/
+    kp.py#L73>`_ .
+
+    More details can be found in the `paper
+    <https://arxiv.org/abs/1808.01244>`_ .
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        num_feat_levels (int): Levels of feature from the previous module. 2
+            for HourglassNet-104 and 1 for HourglassNet-52. Because
+            HourglassNet-104 outputs the final feature and intermediate
+            supervision feature and HourglassNet-52 only outputs the final
+            feature. Default: 2.
+        corner_emb_channels (int): Channel of embedding vector. Default: 1.
+        train_cfg (dict | None): Training config. Useless in CornerHead,
+            but we keep this variable for SingleStageDetector. Default: None.
+        test_cfg (dict | None): Testing config of CornerHead. Default: None.
+        loss_heatmap (dict | None): Config of corner heatmap loss. Default:
+            GaussianFocalLoss.
+        loss_embedding (dict | None): Config of corner embedding loss. Default:
+            AssociativeEmbeddingLoss.
+        loss_offset (dict | None): Config of corner offset loss. Default:
+            SmoothL1Loss.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 num_feat_levels=2,
+                 corner_emb_channels=1,
+                 train_cfg=None,
+                 test_cfg=None,
+                 loss_heatmap=dict(
+                     type='GaussianFocalLoss',
+                     alpha=2.0,
+                     gamma=4.0,
+                     loss_weight=1),
+                 loss_embedding=dict(
+                     type='AssociativeEmbeddingLoss',
+                     pull_weight=0.25,
+                     push_weight=0.25),
+                 loss_offset=dict(
+                     type='SmoothL1Loss', beta=1.0, loss_weight=1)):
+        super(CornerHead, self).__init__()
+        self.num_classes = num_classes
+        self.in_channels = in_channels
+        self.corner_emb_channels = corner_emb_channels
+        self.with_corner_emb = self.corner_emb_channels > 0
+        self.corner_offset_channels = 2
+        self.num_feat_levels = num_feat_levels
+        self.loss_heatmap = build_loss(
+            loss_heatmap) if loss_heatmap is not None else None
+        self.loss_embedding = build_loss(
+            loss_embedding) if loss_embedding is not None else None
+        self.loss_offset = build_loss(
+            loss_offset) if loss_offset is not None else None
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+
+        self._init_layers()
+
+    def _make_layers(self, out_channels, in_channels=256, feat_channels=256):
+        """Initialize conv sequential for CornerHead."""
+        return nn.Sequential(
+            ConvModule(in_channels, feat_channels, 3, padding=1),
+            ConvModule(
+                feat_channels, out_channels, 1, norm_cfg=None, act_cfg=None))
+
+    def _init_corner_kpt_layers(self):
+        """Initialize corner keypoint layers.
+
+        Including corner heatmap branch and corner offset branch. Each branch
+        has two parts: prefix `tl_` for top-left and `br_` for bottom-right.
+        """
+        self.tl_pool, self.br_pool = nn.ModuleList(), nn.ModuleList()
+        self.tl_heat, self.br_heat = nn.ModuleList(), nn.ModuleList()
+        self.tl_off, self.br_off = nn.ModuleList(), nn.ModuleList()
+
+        for _ in range(self.num_feat_levels):
+            self.tl_pool.append(
+                BiCornerPool(
+                    self.in_channels, ['top', 'left'],
+                    out_channels=self.in_channels))
+            self.br_pool.append(
+                BiCornerPool(
+                    self.in_channels, ['bottom', 'right'],
+                    out_channels=self.in_channels))
+
+            self.tl_heat.append(
+                self._make_layers(
+                    out_channels=self.num_classes,
+                    in_channels=self.in_channels))
+            self.br_heat.append(
+                self._make_layers(
+                    out_channels=self.num_classes,
+                    in_channels=self.in_channels))
+
+            self.tl_off.append(
+                self._make_layers(
+                    out_channels=self.corner_offset_channels,
+                    in_channels=self.in_channels))
+            self.br_off.append(
+                self._make_layers(
+                    out_channels=self.corner_offset_channels,
+                    in_channels=self.in_channels))
+
+    def _init_corner_emb_layers(self):
+        """Initialize corner embedding layers.
+
+        Only include corner embedding branch with two parts: prefix `tl_` for
+        top-left and `br_` for bottom-right.
+        """
+        self.tl_emb, self.br_emb = nn.ModuleList(), nn.ModuleList()
+
+        for _ in range(self.num_feat_levels):
+            self.tl_emb.append(
+                self._make_layers(
+                    out_channels=self.corner_emb_channels,
+                    in_channels=self.in_channels))
+            self.br_emb.append(
+                self._make_layers(
+                    out_channels=self.corner_emb_channels,
+                    in_channels=self.in_channels))
+
+    def _init_layers(self):
+        """Initialize layers for CornerHead.
+
+        Including two parts: corner keypoint layers and corner embedding layers
+        """
+        self._init_corner_kpt_layers()
+        if self.with_corner_emb:
+            self._init_corner_emb_layers()
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        bias_init = bias_init_with_prob(0.1)
+        for i in range(self.num_feat_levels):
+            # The initialization of parameters are different between nn.Conv2d
+            # and ConvModule. Our experiments show that using the original
+            # initialization of nn.Conv2d increases the final mAP by about 0.2%
+            self.tl_heat[i][-1].conv.reset_parameters()
+            self.tl_heat[i][-1].conv.bias.data.fill_(bias_init)
+            self.br_heat[i][-1].conv.reset_parameters()
+            self.br_heat[i][-1].conv.bias.data.fill_(bias_init)
+            self.tl_off[i][-1].conv.reset_parameters()
+            self.br_off[i][-1].conv.reset_parameters()
+            if self.with_corner_emb:
+                self.tl_emb[i][-1].conv.reset_parameters()
+                self.br_emb[i][-1].conv.reset_parameters()
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple: Usually a tuple of corner heatmaps, offset heatmaps and
+            embedding heatmaps.
+                - tl_heats (list[Tensor]): Top-left corner heatmaps for all
+                  levels, each is a 4D-tensor, the channels number is
+                  num_classes.
+                - br_heats (list[Tensor]): Bottom-right corner heatmaps for all
+                  levels, each is a 4D-tensor, the channels number is
+                  num_classes.
+                - tl_embs (list[Tensor] | list[None]): Top-left embedding
+                  heatmaps for all levels, each is a 4D-tensor or None.
+                  If not None, the channels number is corner_emb_channels.
+                - br_embs (list[Tensor] | list[None]): Bottom-right embedding
+                  heatmaps for all levels, each is a 4D-tensor or None.
+                  If not None, the channels number is corner_emb_channels.
+                - tl_offs (list[Tensor]): Top-left offset heatmaps for all
+                  levels, each is a 4D-tensor. The channels number is
+                  corner_offset_channels.
+                - br_offs (list[Tensor]): Bottom-right offset heatmaps for all
+                  levels, each is a 4D-tensor. The channels number is
+                  corner_offset_channels.
+        """
+        lvl_ind = list(range(self.num_feat_levels))
+        return multi_apply(self.forward_single, feats, lvl_ind)
+
+    def forward_single(self, x, lvl_ind, return_pool=False):
+        """Forward feature of a single level.
+
+        Args:
+            x (Tensor): Feature of a single level.
+            lvl_ind (int): Level index of current feature.
+            return_pool (bool): Return corner pool feature or not.
+
+        Returns:
+            tuple[Tensor]: A tuple of CornerHead's output for current feature
+            level. Containing the following Tensors:
+
+                - tl_heat (Tensor): Predicted top-left corner heatmap.
+                - br_heat (Tensor): Predicted bottom-right corner heatmap.
+                - tl_emb (Tensor | None): Predicted top-left embedding heatmap.
+                  None for `self.with_corner_emb == False`.
+                - br_emb (Tensor | None): Predicted bottom-right embedding
+                  heatmap. None for `self.with_corner_emb == False`.
+                - tl_off (Tensor): Predicted top-left offset heatmap.
+                - br_off (Tensor): Predicted bottom-right offset heatmap.
+                - tl_pool (Tensor): Top-left corner pool feature. Not must
+                  have.
+                - br_pool (Tensor): Bottom-right corner pool feature. Not must
+                  have.
+        """
+        tl_pool = self.tl_pool[lvl_ind](x)
+        tl_heat = self.tl_heat[lvl_ind](tl_pool)
+        br_pool = self.br_pool[lvl_ind](x)
+        br_heat = self.br_heat[lvl_ind](br_pool)
+
+        tl_emb, br_emb = None, None
+        if self.with_corner_emb:
+            tl_emb = self.tl_emb[lvl_ind](tl_pool)
+            br_emb = self.br_emb[lvl_ind](br_pool)
+
+        tl_off = self.tl_off[lvl_ind](tl_pool)
+        br_off = self.br_off[lvl_ind](br_pool)
+
+        result_list = [tl_heat, br_heat, tl_emb, br_emb, tl_off, br_off]
+        if return_pool:
+            result_list.append(tl_pool)
+            result_list.append(br_pool)
+
+        return result_list
+
+    def get_targets(self,
+                    gt_bboxes,
+                    gt_labels,
+                    feat_shape,
+                    img_shape,
+                    with_corner_emb=False,
+                    with_guiding_shift=False,
+                    with_centripetal_shift=False):
+        """Generate corner targets.
+
+        Including corner heatmap, corner offset.
+
+        Optional: corner embedding, corner guiding shift, centripetal shift.
+
+        For CornerNet, we generate corner heatmap, corner offset and corner
+        embedding from this function.
+
+        For CentripetalNet, we generate corner heatmap, corner offset, guiding
+        shift and centripetal shift from this function.
+
+        Args:
+            gt_bboxes (list[Tensor]): Ground truth bboxes of each image, each
+                has shape (num_gt, 4).
+            gt_labels (list[Tensor]): Ground truth labels of each box, each has
+                shape (num_gt,).
+            feat_shape (list[int]): Shape of output feature,
+                [batch, channel, height, width].
+            img_shape (list[int]): Shape of input image,
+                [height, width, channel].
+            with_corner_emb (bool): Generate corner embedding target or not.
+                Default: False.
+            with_guiding_shift (bool): Generate guiding shift target or not.
+                Default: False.
+            with_centripetal_shift (bool): Generate centripetal shift target or
+                not. Default: False.
+
+        Returns:
+            dict: Ground truth of corner heatmap, corner offset, corner
+            embedding, guiding shift and centripetal shift. Containing the
+            following keys:
+
+                - topleft_heatmap (Tensor): Ground truth top-left corner
+                  heatmap.
+                - bottomright_heatmap (Tensor): Ground truth bottom-right
+                  corner heatmap.
+                - topleft_offset (Tensor): Ground truth top-left corner offset.
+                - bottomright_offset (Tensor): Ground truth bottom-right corner
+                  offset.
+                - corner_embedding (list[list[list[int]]]): Ground truth corner
+                  embedding. Not must have.
+                - topleft_guiding_shift (Tensor): Ground truth top-left corner
+                  guiding shift. Not must have.
+                - bottomright_guiding_shift (Tensor): Ground truth bottom-right
+                  corner guiding shift. Not must have.
+                - topleft_centripetal_shift (Tensor): Ground truth top-left
+                  corner centripetal shift. Not must have.
+                - bottomright_centripetal_shift (Tensor): Ground truth
+                  bottom-right corner centripetal shift. Not must have.
+        """
+        batch_size, _, height, width = feat_shape
+        img_h, img_w = img_shape[:2]
+
+        width_ratio = float(width / img_w)
+        height_ratio = float(height / img_h)
+
+        gt_tl_heatmap = gt_bboxes[-1].new_zeros(
+            [batch_size, self.num_classes, height, width])
+        gt_br_heatmap = gt_bboxes[-1].new_zeros(
+            [batch_size, self.num_classes, height, width])
+        gt_tl_offset = gt_bboxes[-1].new_zeros([batch_size, 2, height, width])
+        gt_br_offset = gt_bboxes[-1].new_zeros([batch_size, 2, height, width])
+
+        if with_corner_emb:
+            match = []
+
+        # Guiding shift is a kind of offset, from center to corner
+        if with_guiding_shift:
+            gt_tl_guiding_shift = gt_bboxes[-1].new_zeros(
+                [batch_size, 2, height, width])
+            gt_br_guiding_shift = gt_bboxes[-1].new_zeros(
+                [batch_size, 2, height, width])
+        # Centripetal shift is also a kind of offset, from center to corner
+        # and normalized by log.
+        if with_centripetal_shift:
+            gt_tl_centripetal_shift = gt_bboxes[-1].new_zeros(
+                [batch_size, 2, height, width])
+            gt_br_centripetal_shift = gt_bboxes[-1].new_zeros(
+                [batch_size, 2, height, width])
+
+        for batch_id in range(batch_size):
+            # Ground truth of corner embedding per image is a list of coord set
+            corner_match = []
+            for box_id in range(len(gt_labels[batch_id])):
+                left, top, right, bottom = gt_bboxes[batch_id][box_id]
+                center_x = (left + right) / 2.0
+                center_y = (top + bottom) / 2.0
+                label = gt_labels[batch_id][box_id]
+
+                # Use coords in the feature level to generate ground truth
+                scale_left = left * width_ratio
+                scale_right = right * width_ratio
+                scale_top = top * height_ratio
+                scale_bottom = bottom * height_ratio
+                scale_center_x = center_x * width_ratio
+                scale_center_y = center_y * height_ratio
+
+                # Int coords on feature map/ground truth tensor
+                left_idx = int(min(scale_left, width - 1))
+                right_idx = int(min(scale_right, width - 1))
+                top_idx = int(min(scale_top, height - 1))
+                bottom_idx = int(min(scale_bottom, height - 1))
+
+                # Generate gaussian heatmap
+                scale_box_width = ceil(scale_right - scale_left)
+                scale_box_height = ceil(scale_bottom - scale_top)
+                radius = gaussian_radius((scale_box_height, scale_box_width),
+                                         min_overlap=0.3)
+                radius = max(0, int(radius))
+                gt_tl_heatmap[batch_id, label] = gen_gaussian_target(
+                    gt_tl_heatmap[batch_id, label], [left_idx, top_idx],
+                    radius)
+                gt_br_heatmap[batch_id, label] = gen_gaussian_target(
+                    gt_br_heatmap[batch_id, label], [right_idx, bottom_idx],
+                    radius)
+
+                # Generate corner offset
+                left_offset = scale_left - left_idx
+                top_offset = scale_top - top_idx
+                right_offset = scale_right - right_idx
+                bottom_offset = scale_bottom - bottom_idx
+                gt_tl_offset[batch_id, 0, top_idx, left_idx] = left_offset
+                gt_tl_offset[batch_id, 1, top_idx, left_idx] = top_offset
+                gt_br_offset[batch_id, 0, bottom_idx, right_idx] = right_offset
+                gt_br_offset[batch_id, 1, bottom_idx,
+                             right_idx] = bottom_offset
+
+                # Generate corner embedding
+                if with_corner_emb:
+                    corner_match.append([[top_idx, left_idx],
+                                         [bottom_idx, right_idx]])
+                # Generate guiding shift
+                if with_guiding_shift:
+                    gt_tl_guiding_shift[batch_id, 0, top_idx,
+                                        left_idx] = scale_center_x - left_idx
+                    gt_tl_guiding_shift[batch_id, 1, top_idx,
+                                        left_idx] = scale_center_y - top_idx
+                    gt_br_guiding_shift[batch_id, 0, bottom_idx,
+                                        right_idx] = right_idx - scale_center_x
+                    gt_br_guiding_shift[
+                        batch_id, 1, bottom_idx,
+                        right_idx] = bottom_idx - scale_center_y
+                # Generate centripetal shift
+                if with_centripetal_shift:
+                    gt_tl_centripetal_shift[batch_id, 0, top_idx,
+                                            left_idx] = log(scale_center_x -
+                                                            scale_left)
+                    gt_tl_centripetal_shift[batch_id, 1, top_idx,
+                                            left_idx] = log(scale_center_y -
+                                                            scale_top)
+                    gt_br_centripetal_shift[batch_id, 0, bottom_idx,
+                                            right_idx] = log(scale_right -
+                                                             scale_center_x)
+                    gt_br_centripetal_shift[batch_id, 1, bottom_idx,
+                                            right_idx] = log(scale_bottom -
+                                                             scale_center_y)
+
+            if with_corner_emb:
+                match.append(corner_match)
+
+        target_result = dict(
+            topleft_heatmap=gt_tl_heatmap,
+            topleft_offset=gt_tl_offset,
+            bottomright_heatmap=gt_br_heatmap,
+            bottomright_offset=gt_br_offset)
+
+        if with_corner_emb:
+            target_result.update(corner_embedding=match)
+        if with_guiding_shift:
+            target_result.update(
+                topleft_guiding_shift=gt_tl_guiding_shift,
+                bottomright_guiding_shift=gt_br_guiding_shift)
+        if with_centripetal_shift:
+            target_result.update(
+                topleft_centripetal_shift=gt_tl_centripetal_shift,
+                bottomright_centripetal_shift=gt_br_centripetal_shift)
+
+        return target_result
+
+    def loss(self,
+             tl_heats,
+             br_heats,
+             tl_embs,
+             br_embs,
+             tl_offs,
+             br_offs,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            tl_heats (list[Tensor]): Top-left corner heatmaps for each level
+                with shape (N, num_classes, H, W).
+            br_heats (list[Tensor]): Bottom-right corner heatmaps for each
+                level with shape (N, num_classes, H, W).
+            tl_embs (list[Tensor]): Top-left corner embeddings for each level
+                with shape (N, corner_emb_channels, H, W).
+            br_embs (list[Tensor]): Bottom-right corner embeddings for each
+                level with shape (N, corner_emb_channels, H, W).
+            tl_offs (list[Tensor]): Top-left corner offsets for each level
+                with shape (N, corner_offset_channels, H, W).
+            br_offs (list[Tensor]): Bottom-right corner offsets for each level
+                with shape (N, corner_offset_channels, H, W).
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [left, top, right, bottom] format.
+            gt_labels (list[Tensor]): Class indices corresponding to each box.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (list[Tensor] | None): Specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components. Containing the
+            following losses:
+
+                - det_loss (list[Tensor]): Corner keypoint losses of all
+                  feature levels.
+                - pull_loss (list[Tensor]): Part one of AssociativeEmbedding
+                  losses of all feature levels.
+                - push_loss (list[Tensor]): Part two of AssociativeEmbedding
+                  losses of all feature levels.
+                - off_loss (list[Tensor]): Corner offset losses of all feature
+                  levels.
+        """
+        targets = self.get_targets(
+            gt_bboxes,
+            gt_labels,
+            tl_heats[-1].shape,
+            img_metas[0]['pad_shape'],
+            with_corner_emb=self.with_corner_emb)
+        mlvl_targets = [targets for _ in range(self.num_feat_levels)]
+        det_losses, pull_losses, push_losses, off_losses = multi_apply(
+            self.loss_single, tl_heats, br_heats, tl_embs, br_embs, tl_offs,
+            br_offs, mlvl_targets)
+        loss_dict = dict(det_loss=det_losses, off_loss=off_losses)
+        if self.with_corner_emb:
+            loss_dict.update(pull_loss=pull_losses, push_loss=push_losses)
+        return loss_dict
+
+    def loss_single(self, tl_hmp, br_hmp, tl_emb, br_emb, tl_off, br_off,
+                    targets):
+        """Compute losses for single level.
+
+        Args:
+            tl_hmp (Tensor): Top-left corner heatmap for current level with
+                shape (N, num_classes, H, W).
+            br_hmp (Tensor): Bottom-right corner heatmap for current level with
+                shape (N, num_classes, H, W).
+            tl_emb (Tensor): Top-left corner embedding for current level with
+                shape (N, corner_emb_channels, H, W).
+            br_emb (Tensor): Bottom-right corner embedding for current level
+                with shape (N, corner_emb_channels, H, W).
+            tl_off (Tensor): Top-left corner offset for current level with
+                shape (N, corner_offset_channels, H, W).
+            br_off (Tensor): Bottom-right corner offset for current level with
+                shape (N, corner_offset_channels, H, W).
+            targets (dict): Corner target generated by `get_targets`.
+
+        Returns:
+            tuple[torch.Tensor]: Losses of the head's differnet branches
+            containing the following losses:
+
+                - det_loss (Tensor): Corner keypoint loss.
+                - pull_loss (Tensor): Part one of AssociativeEmbedding loss.
+                - push_loss (Tensor): Part two of AssociativeEmbedding loss.
+                - off_loss (Tensor): Corner offset loss.
+        """
+        gt_tl_hmp = targets['topleft_heatmap']
+        gt_br_hmp = targets['bottomright_heatmap']
+        gt_tl_off = targets['topleft_offset']
+        gt_br_off = targets['bottomright_offset']
+        gt_embedding = targets['corner_embedding']
+
+        # Detection loss
+        tl_det_loss = self.loss_heatmap(
+            tl_hmp.sigmoid(),
+            gt_tl_hmp,
+            avg_factor=max(1,
+                           gt_tl_hmp.eq(1).sum()))
+        br_det_loss = self.loss_heatmap(
+            br_hmp.sigmoid(),
+            gt_br_hmp,
+            avg_factor=max(1,
+                           gt_br_hmp.eq(1).sum()))
+        det_loss = (tl_det_loss + br_det_loss) / 2.0
+
+        # AssociativeEmbedding loss
+        if self.with_corner_emb and self.loss_embedding is not None:
+            pull_loss, push_loss = self.loss_embedding(tl_emb, br_emb,
+                                                       gt_embedding)
+        else:
+            pull_loss, push_loss = None, None
+
+        # Offset loss
+        # We only compute the offset loss at the real corner position.
+        # The value of real corner would be 1 in heatmap ground truth.
+        # The mask is computed in class agnostic mode and its shape is
+        # batch * 1 * width * height.
+        tl_off_mask = gt_tl_hmp.eq(1).sum(1).gt(0).unsqueeze(1).type_as(
+            gt_tl_hmp)
+        br_off_mask = gt_br_hmp.eq(1).sum(1).gt(0).unsqueeze(1).type_as(
+            gt_br_hmp)
+        tl_off_loss = self.loss_offset(
+            tl_off,
+            gt_tl_off,
+            tl_off_mask,
+            avg_factor=max(1, tl_off_mask.sum()))
+        br_off_loss = self.loss_offset(
+            br_off,
+            gt_br_off,
+            br_off_mask,
+            avg_factor=max(1, br_off_mask.sum()))
+
+        off_loss = (tl_off_loss + br_off_loss) / 2.0
+
+        return det_loss, pull_loss, push_loss, off_loss
+
+    def get_bboxes(self,
+                   tl_heats,
+                   br_heats,
+                   tl_embs,
+                   br_embs,
+                   tl_offs,
+                   br_offs,
+                   img_metas,
+                   rescale=False,
+                   with_nms=True):
+        """Transform network output for a batch into bbox predictions.
+
+        Args:
+            tl_heats (list[Tensor]): Top-left corner heatmaps for each level
+                with shape (N, num_classes, H, W).
+            br_heats (list[Tensor]): Bottom-right corner heatmaps for each
+                level with shape (N, num_classes, H, W).
+            tl_embs (list[Tensor]): Top-left corner embeddings for each level
+                with shape (N, corner_emb_channels, H, W).
+            br_embs (list[Tensor]): Bottom-right corner embeddings for each
+                level with shape (N, corner_emb_channels, H, W).
+            tl_offs (list[Tensor]): Top-left corner offsets for each level
+                with shape (N, corner_offset_channels, H, W).
+            br_offs (list[Tensor]): Bottom-right corner offsets for each level
+                with shape (N, corner_offset_channels, H, W).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+        """
+        assert tl_heats[-1].shape[0] == br_heats[-1].shape[0] == len(img_metas)
+        result_list = []
+        for img_id in range(len(img_metas)):
+            result_list.append(
+                self._get_bboxes_single(
+                    tl_heats[-1][img_id:img_id + 1, :],
+                    br_heats[-1][img_id:img_id + 1, :],
+                    tl_offs[-1][img_id:img_id + 1, :],
+                    br_offs[-1][img_id:img_id + 1, :],
+                    img_metas[img_id],
+                    tl_emb=tl_embs[-1][img_id:img_id + 1, :],
+                    br_emb=br_embs[-1][img_id:img_id + 1, :],
+                    rescale=rescale,
+                    with_nms=with_nms))
+
+        return result_list
+
+    def _get_bboxes_single(self,
+                           tl_heat,
+                           br_heat,
+                           tl_off,
+                           br_off,
+                           img_meta,
+                           tl_emb=None,
+                           br_emb=None,
+                           tl_centripetal_shift=None,
+                           br_centripetal_shift=None,
+                           rescale=False,
+                           with_nms=True):
+        """Transform outputs for a single batch item into bbox predictions.
+
+        Args:
+            tl_heat (Tensor): Top-left corner heatmap for current level with
+                shape (N, num_classes, H, W).
+            br_heat (Tensor): Bottom-right corner heatmap for current level
+                with shape (N, num_classes, H, W).
+            tl_off (Tensor): Top-left corner offset for current level with
+                shape (N, corner_offset_channels, H, W).
+            br_off (Tensor): Bottom-right corner offset for current level with
+                shape (N, corner_offset_channels, H, W).
+            img_meta (dict): Meta information of current image, e.g.,
+                image size, scaling factor, etc.
+            tl_emb (Tensor): Top-left corner embedding for current level with
+                shape (N, corner_emb_channels, H, W).
+            br_emb (Tensor): Bottom-right corner embedding for current level
+                with shape (N, corner_emb_channels, H, W).
+            tl_centripetal_shift: Top-left corner's centripetal shift for
+                current level with shape (N, 2, H, W).
+            br_centripetal_shift: Bottom-right corner's centripetal shift for
+                current level with shape (N, 2, H, W).
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+        """
+        if isinstance(img_meta, (list, tuple)):
+            img_meta = img_meta[0]
+
+        batch_bboxes, batch_scores, batch_clses = self.decode_heatmap(
+            tl_heat=tl_heat.sigmoid(),
+            br_heat=br_heat.sigmoid(),
+            tl_off=tl_off,
+            br_off=br_off,
+            tl_emb=tl_emb,
+            br_emb=br_emb,
+            tl_centripetal_shift=tl_centripetal_shift,
+            br_centripetal_shift=br_centripetal_shift,
+            img_meta=img_meta,
+            k=self.test_cfg.corner_topk,
+            kernel=self.test_cfg.local_maximum_kernel,
+            distance_threshold=self.test_cfg.distance_threshold)
+
+        if rescale:
+            batch_bboxes /= batch_bboxes.new_tensor(img_meta['scale_factor'])
+
+        bboxes = batch_bboxes.view([-1, 4])
+        scores = batch_scores.view([-1, 1])
+        clses = batch_clses.view([-1, 1])
+
+        idx = scores.argsort(dim=0, descending=True)
+        bboxes = bboxes[idx].view([-1, 4])
+        scores = scores[idx].view(-1)
+        clses = clses[idx].view(-1)
+
+        detections = torch.cat([bboxes, scores.unsqueeze(-1)], -1)
+        keepinds = (detections[:, -1] > -0.1)
+        detections = detections[keepinds]
+        labels = clses[keepinds]
+
+        if with_nms:
+            detections, labels = self._bboxes_nms(detections, labels,
+                                                  self.test_cfg)
+
+        return detections, labels
+
+    def _bboxes_nms(self, bboxes, labels, cfg):
+        if labels.numel() == 0:
+            return bboxes, labels
+
+        if 'nms_cfg' in cfg:
+            warning.warn('nms_cfg in test_cfg will be deprecated. '
+                         'Please rename it as nms')
+        if 'nms' not in cfg:
+            cfg.nms = cfg.nms_cfg
+
+        out_bboxes, keep = batched_nms(bboxes[:, :4], bboxes[:, -1], labels,
+                                       cfg.nms)
+        out_labels = labels[keep]
+
+        if len(out_bboxes) > 0:
+            idx = torch.argsort(out_bboxes[:, -1], descending=True)
+            idx = idx[:cfg.max_per_img]
+            out_bboxes = out_bboxes[idx]
+            out_labels = out_labels[idx]
+
+        return out_bboxes, out_labels
+
+    def _gather_feat(self, feat, ind, mask=None):
+        """Gather feature according to index.
+
+        Args:
+            feat (Tensor): Target feature map.
+            ind (Tensor): Target coord index.
+            mask (Tensor | None): Mask of featuremap. Default: None.
+
+        Returns:
+            feat (Tensor): Gathered feature.
+        """
+        dim = feat.size(2)
+        ind = ind.unsqueeze(2).repeat(1, 1, dim)
+        feat = feat.gather(1, ind)
+        if mask is not None:
+            mask = mask.unsqueeze(2).expand_as(feat)
+            feat = feat[mask]
+            feat = feat.view(-1, dim)
+        return feat
+
+    def _local_maximum(self, heat, kernel=3):
+        """Extract local maximum pixel with given kernel.
+
+        Args:
+            heat (Tensor): Target heatmap.
+            kernel (int): Kernel size of max pooling. Default: 3.
+
+        Returns:
+            heat (Tensor): A heatmap where local maximum pixels maintain its
+                own value and other positions are 0.
+        """
+        pad = (kernel - 1) // 2
+        hmax = F.max_pool2d(heat, kernel, stride=1, padding=pad)
+        keep = (hmax == heat).float()
+        return heat * keep
+
+    def _transpose_and_gather_feat(self, feat, ind):
+        """Transpose and gather feature according to index.
+
+        Args:
+            feat (Tensor): Target feature map.
+            ind (Tensor): Target coord index.
+
+        Returns:
+            feat (Tensor): Transposed and gathered feature.
+        """
+        feat = feat.permute(0, 2, 3, 1).contiguous()
+        feat = feat.view(feat.size(0), -1, feat.size(3))
+        feat = self._gather_feat(feat, ind)
+        return feat
+
+    def _topk(self, scores, k=20):
+        """Get top k positions from heatmap.
+
+        Args:
+            scores (Tensor): Target heatmap with shape
+                [batch, num_classes, height, width].
+            k (int): Target number. Default: 20.
+
+        Returns:
+            tuple[torch.Tensor]: Scores, indexes, categories and coords of
+                topk keypoint. Containing following Tensors:
+
+            - topk_scores (Tensor): Max scores of each topk keypoint.
+            - topk_inds (Tensor): Indexes of each topk keypoint.
+            - topk_clses (Tensor): Categories of each topk keypoint.
+            - topk_ys (Tensor): Y-coord of each topk keypoint.
+            - topk_xs (Tensor): X-coord of each topk keypoint.
+        """
+        batch, _, height, width = scores.size()
+        topk_scores, topk_inds = torch.topk(scores.view(batch, -1), k)
+        topk_clses = topk_inds // (height * width)
+        topk_inds = topk_inds % (height * width)
+        topk_ys = topk_inds // width
+        topk_xs = (topk_inds % width).int().float()
+        return topk_scores, topk_inds, topk_clses, topk_ys, topk_xs
+
+    def decode_heatmap(self,
+                       tl_heat,
+                       br_heat,
+                       tl_off,
+                       br_off,
+                       tl_emb=None,
+                       br_emb=None,
+                       tl_centripetal_shift=None,
+                       br_centripetal_shift=None,
+                       img_meta=None,
+                       k=100,
+                       kernel=3,
+                       distance_threshold=0.5,
+                       num_dets=1000):
+        """Transform outputs for a single batch item into raw bbox predictions.
+
+        Args:
+            tl_heat (Tensor): Top-left corner heatmap for current level with
+                shape (N, num_classes, H, W).
+            br_heat (Tensor): Bottom-right corner heatmap for current level
+                with shape (N, num_classes, H, W).
+            tl_off (Tensor): Top-left corner offset for current level with
+                shape (N, corner_offset_channels, H, W).
+            br_off (Tensor): Bottom-right corner offset for current level with
+                shape (N, corner_offset_channels, H, W).
+            tl_emb (Tensor | None): Top-left corner embedding for current
+                level with shape (N, corner_emb_channels, H, W).
+            br_emb (Tensor | None): Bottom-right corner embedding for current
+                level with shape (N, corner_emb_channels, H, W).
+            tl_centripetal_shift (Tensor | None): Top-left centripetal shift
+                for current level with shape (N, 2, H, W).
+            br_centripetal_shift (Tensor | None): Bottom-right centripetal
+                shift for current level with shape (N, 2, H, W).
+            img_meta (dict): Meta information of current image, e.g.,
+                image size, scaling factor, etc.
+            k (int): Get top k corner keypoints from heatmap.
+            kernel (int): Max pooling kernel for extract local maximum pixels.
+            distance_threshold (float): Distance threshold. Top-left and
+                bottom-right corner keypoints with feature distance less than
+                the threshold will be regarded as keypoints from same object.
+            num_dets (int): Num of raw boxes before doing nms.
+
+        Returns:
+            tuple[torch.Tensor]: Decoded output of CornerHead, containing the
+            following Tensors:
+
+            - bboxes (Tensor): Coords of each box.
+            - scores (Tensor): Scores of each box.
+            - clses (Tensor): Categories of each box.
+        """
+        with_embedding = tl_emb is not None and br_emb is not None
+        with_centripetal_shift = (
+            tl_centripetal_shift is not None
+            and br_centripetal_shift is not None)
+        assert with_embedding + with_centripetal_shift == 1
+        batch, _, height, width = tl_heat.size()
+        inp_h, inp_w, _ = img_meta['pad_shape']
+
+        # perform nms on heatmaps
+        tl_heat = self._local_maximum(tl_heat, kernel=kernel)
+        br_heat = self._local_maximum(br_heat, kernel=kernel)
+
+        tl_scores, tl_inds, tl_clses, tl_ys, tl_xs = self._topk(tl_heat, k=k)
+        br_scores, br_inds, br_clses, br_ys, br_xs = self._topk(br_heat, k=k)
+
+        # We use repeat instead of expand here because expand is a
+        # shallow-copy function. Thus it could cause unexpected testing result
+        # sometimes. Using expand will decrease about 10% mAP during testing
+        # compared to repeat.
+        tl_ys = tl_ys.view(batch, k, 1).repeat(1, 1, k)
+        tl_xs = tl_xs.view(batch, k, 1).repeat(1, 1, k)
+        br_ys = br_ys.view(batch, 1, k).repeat(1, k, 1)
+        br_xs = br_xs.view(batch, 1, k).repeat(1, k, 1)
+
+        tl_off = self._transpose_and_gather_feat(tl_off, tl_inds)
+        tl_off = tl_off.view(batch, k, 1, 2)
+        br_off = self._transpose_and_gather_feat(br_off, br_inds)
+        br_off = br_off.view(batch, 1, k, 2)
+
+        tl_xs = tl_xs + tl_off[..., 0]
+        tl_ys = tl_ys + tl_off[..., 1]
+        br_xs = br_xs + br_off[..., 0]
+        br_ys = br_ys + br_off[..., 1]
+
+        if with_centripetal_shift:
+            tl_centripetal_shift = self._transpose_and_gather_feat(
+                tl_centripetal_shift, tl_inds).view(batch, k, 1, 2).exp()
+            br_centripetal_shift = self._transpose_and_gather_feat(
+                br_centripetal_shift, br_inds).view(batch, 1, k, 2).exp()
+
+            tl_ctxs = tl_xs + tl_centripetal_shift[..., 0]
+            tl_ctys = tl_ys + tl_centripetal_shift[..., 1]
+            br_ctxs = br_xs - br_centripetal_shift[..., 0]
+            br_ctys = br_ys - br_centripetal_shift[..., 1]
+
+        # all possible boxes based on top k corners (ignoring class)
+        tl_xs *= (inp_w / width)
+        tl_ys *= (inp_h / height)
+        br_xs *= (inp_w / width)
+        br_ys *= (inp_h / height)
+
+        if with_centripetal_shift:
+            tl_ctxs *= (inp_w / width)
+            tl_ctys *= (inp_h / height)
+            br_ctxs *= (inp_w / width)
+            br_ctys *= (inp_h / height)
+
+        x_off = img_meta['border'][2]
+        y_off = img_meta['border'][0]
+
+        tl_xs -= x_off
+        tl_ys -= y_off
+        br_xs -= x_off
+        br_ys -= y_off
+
+        tl_xs *= tl_xs.gt(0.0).type_as(tl_xs)
+        tl_ys *= tl_ys.gt(0.0).type_as(tl_ys)
+        br_xs *= br_xs.gt(0.0).type_as(br_xs)
+        br_ys *= br_ys.gt(0.0).type_as(br_ys)
+
+        bboxes = torch.stack((tl_xs, tl_ys, br_xs, br_ys), dim=3)
+        area_bboxes = ((br_xs - tl_xs) * (br_ys - tl_ys)).abs()
+
+        if with_centripetal_shift:
+            tl_ctxs -= x_off
+            tl_ctys -= y_off
+            br_ctxs -= x_off
+            br_ctys -= y_off
+
+            tl_ctxs *= tl_ctxs.gt(0.0).type_as(tl_ctxs)
+            tl_ctys *= tl_ctys.gt(0.0).type_as(tl_ctys)
+            br_ctxs *= br_ctxs.gt(0.0).type_as(br_ctxs)
+            br_ctys *= br_ctys.gt(0.0).type_as(br_ctys)
+
+            ct_bboxes = torch.stack((tl_ctxs, tl_ctys, br_ctxs, br_ctys),
+                                    dim=3)
+            area_ct_bboxes = ((br_ctxs - tl_ctxs) * (br_ctys - tl_ctys)).abs()
+
+            rcentral = torch.zeros_like(ct_bboxes)
+            # magic nums from paper section 4.1
+            mu = torch.ones_like(area_bboxes) / 2.4
+            mu[area_bboxes > 3500] = 1 / 2.1  # large bbox have smaller mu
+
+            bboxes_center_x = (bboxes[..., 0] + bboxes[..., 2]) / 2
+            bboxes_center_y = (bboxes[..., 1] + bboxes[..., 3]) / 2
+            rcentral[..., 0] = bboxes_center_x - mu * (bboxes[..., 2] -
+                                                       bboxes[..., 0]) / 2
+            rcentral[..., 1] = bboxes_center_y - mu * (bboxes[..., 3] -
+                                                       bboxes[..., 1]) / 2
+            rcentral[..., 2] = bboxes_center_x + mu * (bboxes[..., 2] -
+                                                       bboxes[..., 0]) / 2
+            rcentral[..., 3] = bboxes_center_y + mu * (bboxes[..., 3] -
+                                                       bboxes[..., 1]) / 2
+            area_rcentral = ((rcentral[..., 2] - rcentral[..., 0]) *
+                             (rcentral[..., 3] - rcentral[..., 1])).abs()
+            dists = area_ct_bboxes / area_rcentral
+
+            tl_ctx_inds = (ct_bboxes[..., 0] <= rcentral[..., 0]) | (
+                ct_bboxes[..., 0] >= rcentral[..., 2])
+            tl_cty_inds = (ct_bboxes[..., 1] <= rcentral[..., 1]) | (
+                ct_bboxes[..., 1] >= rcentral[..., 3])
+            br_ctx_inds = (ct_bboxes[..., 2] <= rcentral[..., 0]) | (
+                ct_bboxes[..., 2] >= rcentral[..., 2])
+            br_cty_inds = (ct_bboxes[..., 3] <= rcentral[..., 1]) | (
+                ct_bboxes[..., 3] >= rcentral[..., 3])
+
+        if with_embedding:
+            tl_emb = self._transpose_and_gather_feat(tl_emb, tl_inds)
+            tl_emb = tl_emb.view(batch, k, 1)
+            br_emb = self._transpose_and_gather_feat(br_emb, br_inds)
+            br_emb = br_emb.view(batch, 1, k)
+            dists = torch.abs(tl_emb - br_emb)
+
+        tl_scores = tl_scores.view(batch, k, 1).repeat(1, 1, k)
+        br_scores = br_scores.view(batch, 1, k).repeat(1, k, 1)
+
+        scores = (tl_scores + br_scores) / 2  # scores for all possible boxes
+
+        # tl and br should have same class
+        tl_clses = tl_clses.view(batch, k, 1).repeat(1, 1, k)
+        br_clses = br_clses.view(batch, 1, k).repeat(1, k, 1)
+        cls_inds = (tl_clses != br_clses)
+
+        # reject boxes based on distances
+        dist_inds = dists > distance_threshold
+
+        # reject boxes based on widths and heights
+        width_inds = (br_xs <= tl_xs)
+        height_inds = (br_ys <= tl_ys)
+
+        scores[cls_inds] = -1
+        scores[width_inds] = -1
+        scores[height_inds] = -1
+        scores[dist_inds] = -1
+        if with_centripetal_shift:
+            scores[tl_ctx_inds] = -1
+            scores[tl_cty_inds] = -1
+            scores[br_ctx_inds] = -1
+            scores[br_cty_inds] = -1
+
+        scores = scores.view(batch, -1)
+        scores, inds = torch.topk(scores, num_dets)
+        scores = scores.unsqueeze(2)
+
+        bboxes = bboxes.view(batch, -1, 4)
+        bboxes = self._gather_feat(bboxes, inds)
+
+        clses = tl_clses.contiguous().view(batch, -1, 1)
+        clses = self._gather_feat(clses, inds).float()
+
+        return bboxes, scores, clses
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/dense_test_mixins.py b/python/autoCut/models/swin/mmdet/models/dense_heads/dense_test_mixins.py
new file mode 100644
index 0000000..dd81364
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/dense_test_mixins.py
@@ -0,0 +1,100 @@
+from inspect import signature
+
+import torch
+
+from mmdet.core import bbox2result, bbox_mapping_back, multiclass_nms
+
+
+class BBoxTestMixin(object):
+    """Mixin class for test time augmentation of bboxes."""
+
+    def merge_aug_bboxes(self, aug_bboxes, aug_scores, img_metas):
+        """Merge augmented detection bboxes and scores.
+
+        Args:
+            aug_bboxes (list[Tensor]): shape (n, 4*#class)
+            aug_scores (list[Tensor] or None): shape (n, #class)
+            img_shapes (list[Tensor]): shape (3, ).
+
+        Returns:
+            tuple: (bboxes, scores)
+        """
+        recovered_bboxes = []
+        for bboxes, img_info in zip(aug_bboxes, img_metas):
+            img_shape = img_info[0]['img_shape']
+            scale_factor = img_info[0]['scale_factor']
+            flip = img_info[0]['flip']
+            flip_direction = img_info[0]['flip_direction']
+            bboxes = bbox_mapping_back(bboxes, img_shape, scale_factor, flip,
+                                       flip_direction)
+            recovered_bboxes.append(bboxes)
+        bboxes = torch.cat(recovered_bboxes, dim=0)
+        if aug_scores is None:
+            return bboxes
+        else:
+            scores = torch.cat(aug_scores, dim=0)
+            return bboxes, scores
+
+    def aug_test_bboxes(self, feats, img_metas, rescale=False):
+        """Test det bboxes with test time augmentation.
+
+        Args:
+            feats (list[Tensor]): the outer list indicates test-time
+                augmentations and inner Tensor should have a shape NxCxHxW,
+                which contains features for all images in the batch.
+            img_metas (list[list[dict]]): the outer list indicates test-time
+                augs (multiscale, flip, etc.) and the inner list indicates
+                images in a batch. each dict has image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[ndarray]: bbox results of each class
+        """
+        # check with_nms argument
+        gb_sig = signature(self.get_bboxes)
+        gb_args = [p.name for p in gb_sig.parameters.values()]
+        if hasattr(self, '_get_bboxes'):
+            gbs_sig = signature(self._get_bboxes)
+        else:
+            gbs_sig = signature(self._get_bboxes_single)
+        gbs_args = [p.name for p in gbs_sig.parameters.values()]
+        assert ('with_nms' in gb_args) and ('with_nms' in gbs_args), \
+            f'{self.__class__.__name__}' \
+            ' does not support test-time augmentation'
+
+        aug_bboxes = []
+        aug_scores = []
+        aug_factors = []  # score_factors for NMS
+        for x, img_meta in zip(feats, img_metas):
+            # only one image in the batch
+            outs = self.forward(x)
+            bbox_inputs = outs + (img_meta, self.test_cfg, False, False)
+            bbox_outputs = self.get_bboxes(*bbox_inputs)[0]
+            aug_bboxes.append(bbox_outputs[0])
+            aug_scores.append(bbox_outputs[1])
+            # bbox_outputs of some detectors (e.g., ATSS, FCOS, YOLOv3)
+            # contains additional element to adjust scores before NMS
+            if len(bbox_outputs) >= 3:
+                aug_factors.append(bbox_outputs[2])
+
+        # after merging, bboxes will be rescaled to the original image size
+        merged_bboxes, merged_scores = self.merge_aug_bboxes(
+            aug_bboxes, aug_scores, img_metas)
+        merged_factors = torch.cat(aug_factors, dim=0) if aug_factors else None
+        det_bboxes, det_labels = multiclass_nms(
+            merged_bboxes,
+            merged_scores,
+            self.test_cfg.score_thr,
+            self.test_cfg.nms,
+            self.test_cfg.max_per_img,
+            score_factors=merged_factors)
+
+        if rescale:
+            _det_bboxes = det_bboxes
+        else:
+            _det_bboxes = det_bboxes.clone()
+            _det_bboxes[:, :4] *= det_bboxes.new_tensor(
+                img_metas[0][0]['scale_factor'])
+        bbox_results = bbox2result(_det_bboxes, det_labels, self.num_classes)
+        return bbox_results
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/embedding_rpn_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/embedding_rpn_head.py
new file mode 100644
index 0000000..200ce8d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/embedding_rpn_head.py
@@ -0,0 +1,100 @@
+import torch
+import torch.nn as nn
+
+from mmdet.models.builder import HEADS
+from ...core import bbox_cxcywh_to_xyxy
+
+
+@HEADS.register_module()
+class EmbeddingRPNHead(nn.Module):
+    """RPNHead in the `Sparse R-CNN <https://arxiv.org/abs/2011.12450>`_ .
+
+    Unlike traditional RPNHead, this module does not need FPN input, but just
+    decode `init_proposal_bboxes` and expand the first dimension of
+    `init_proposal_bboxes` and `init_proposal_features` to the batch_size.
+
+    Args:
+        num_proposals (int): Number of init_proposals. Default 100.
+        proposal_feature_channel (int): Channel number of
+            init_proposal_feature. Defaults to 256.
+    """
+
+    def __init__(self,
+                 num_proposals=100,
+                 proposal_feature_channel=256,
+                 **kwargs):
+        super(EmbeddingRPNHead, self).__init__()
+        self.num_proposals = num_proposals
+        self.proposal_feature_channel = proposal_feature_channel
+        self._init_layers()
+
+    def _init_layers(self):
+        """Initialize a sparse set of proposal boxes and proposal features."""
+        self.init_proposal_bboxes = nn.Embedding(self.num_proposals, 4)
+        self.init_proposal_features = nn.Embedding(
+            self.num_proposals, self.proposal_feature_channel)
+
+    def init_weights(self):
+        """Initialize the init_proposal_bboxes as normalized.
+
+        [c_x, c_y, w, h], and we initialize it to the size of  the entire
+        image.
+        """
+        nn.init.constant_(self.init_proposal_bboxes.weight[:, :2], 0.5)
+        nn.init.constant_(self.init_proposal_bboxes.weight[:, 2:], 1)
+
+    def _decode_init_proposals(self, imgs, img_metas):
+        """Decode init_proposal_bboxes according to the size of images and
+        expand dimension of init_proposal_features to batch_size.
+
+        Args:
+            imgs (list[Tensor]): List of FPN features.
+            img_metas (list[dict]): List of meta-information of
+                images. Need the img_shape to decode the init_proposals.
+
+        Returns:
+            Tuple(Tensor):
+
+                - proposals (Tensor): Decoded proposal bboxes,
+                  has shape (batch_size, num_proposals, 4).
+                - init_proposal_features (Tensor): Expanded proposal
+                  features, has shape
+                  (batch_size, num_proposals, proposal_feature_channel).
+                - imgs_whwh (Tensor): Tensor with shape
+                  (batch_size, 4), the dimension means
+                  [img_width, img_height, img_width, img_height].
+        """
+        proposals = self.init_proposal_bboxes.weight.clone()
+        proposals = bbox_cxcywh_to_xyxy(proposals)
+        num_imgs = len(imgs[0])
+        imgs_whwh = []
+        for meta in img_metas:
+            h, w, _ = meta['img_shape']
+            imgs_whwh.append(imgs[0].new_tensor([[w, h, w, h]]))
+        imgs_whwh = torch.cat(imgs_whwh, dim=0)
+        imgs_whwh = imgs_whwh[:, None, :]
+
+        # imgs_whwh has shape (batch_size, 1, 4)
+        # The shape of proposals change from (num_proposals, 4)
+        # to (batch_size ,num_proposals, 4)
+        proposals = proposals * imgs_whwh
+
+        init_proposal_features = self.init_proposal_features.weight.clone()
+        init_proposal_features = init_proposal_features[None].expand(
+            num_imgs, *init_proposal_features.size())
+        return proposals, init_proposal_features, imgs_whwh
+
+    def forward_dummy(self, img, img_metas):
+        """Dummy forward function.
+
+        Used in flops calculation.
+        """
+        return self._decode_init_proposals(img, img_metas)
+
+    def forward_train(self, img, img_metas):
+        """Forward function in training stage."""
+        return self._decode_init_proposals(img, img_metas)
+
+    def simple_test_rpn(self, img, img_metas):
+        """Forward function in testing stage."""
+        return self._decode_init_proposals(img, img_metas)
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/fcos_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/fcos_head.py
new file mode 100644
index 0000000..905a703
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/fcos_head.py
@@ -0,0 +1,629 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import Scale, normal_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import distance2bbox, multi_apply, multiclass_nms, reduce_mean
+from ..builder import HEADS, build_loss
+from .anchor_free_head import AnchorFreeHead
+
+INF = 1e8
+
+
+@HEADS.register_module()
+class FCOSHead(AnchorFreeHead):
+    """Anchor-free head used in `FCOS <https://arxiv.org/abs/1904.01355>`_.
+
+    The FCOS head does not use anchor boxes. Instead bounding boxes are
+    predicted at each pixel and a centerness measure is used to suppress
+    low-quality predictions.
+    Here norm_on_bbox, centerness_on_reg, dcn_on_last_conv are training
+    tricks used in official repo, which will bring remarkable mAP gains
+    of up to 4.9. Please see https://github.com/tianzhi0549/FCOS for
+    more detail.
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        strides (list[int] | list[tuple[int, int]]): Strides of points
+            in multiple feature levels. Default: (4, 8, 16, 32, 64).
+        regress_ranges (tuple[tuple[int, int]]): Regress range of multiple
+            level points.
+        center_sampling (bool): If true, use center sampling. Default: False.
+        center_sample_radius (float): Radius of center sampling. Default: 1.5.
+        norm_on_bbox (bool): If true, normalize the regression targets
+            with FPN strides. Default: False.
+        centerness_on_reg (bool): If true, position centerness on the
+            regress branch. Please refer to https://github.com/tianzhi0549/FCOS/issues/89#issuecomment-516877042.
+            Default: False.
+        conv_bias (bool | str): If specified as `auto`, it will be decided by the
+            norm_cfg. Bias of conv will be set as True if `norm_cfg` is None, otherwise
+            False. Default: "auto".
+        loss_cls (dict): Config of classification loss.
+        loss_bbox (dict): Config of localization loss.
+        loss_centerness (dict): Config of centerness loss.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+            Default: norm_cfg=dict(type='GN', num_groups=32, requires_grad=True).
+
+    Example:
+        >>> self = FCOSHead(11, 7)
+        >>> feats = [torch.rand(1, 7, s, s) for s in [4, 8, 16, 32, 64]]
+        >>> cls_score, bbox_pred, centerness = self.forward(feats)
+        >>> assert len(cls_score) == len(self.scales)
+    """  # noqa: E501
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 regress_ranges=((-1, 64), (64, 128), (128, 256), (256, 512),
+                                 (512, INF)),
+                 center_sampling=False,
+                 center_sample_radius=1.5,
+                 norm_on_bbox=False,
+                 centerness_on_reg=False,
+                 loss_cls=dict(
+                     type='FocalLoss',
+                     use_sigmoid=True,
+                     gamma=2.0,
+                     alpha=0.25,
+                     loss_weight=1.0),
+                 loss_bbox=dict(type='IoULoss', loss_weight=1.0),
+                 loss_centerness=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.0),
+                 norm_cfg=dict(type='GN', num_groups=32, requires_grad=True),
+                 **kwargs):
+        self.regress_ranges = regress_ranges
+        self.center_sampling = center_sampling
+        self.center_sample_radius = center_sample_radius
+        self.norm_on_bbox = norm_on_bbox
+        self.centerness_on_reg = centerness_on_reg
+        super().__init__(
+            num_classes,
+            in_channels,
+            loss_cls=loss_cls,
+            loss_bbox=loss_bbox,
+            norm_cfg=norm_cfg,
+            **kwargs)
+        self.loss_centerness = build_loss(loss_centerness)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        super()._init_layers()
+        self.conv_centerness = nn.Conv2d(self.feat_channels, 1, 3, padding=1)
+        self.scales = nn.ModuleList([Scale(1.0) for _ in self.strides])
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        super().init_weights()
+        normal_init(self.conv_centerness, std=0.01)
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple:
+                cls_scores (list[Tensor]): Box scores for each scale level, \
+                    each is a 4D-tensor, the channel number is \
+                    num_points * num_classes.
+                bbox_preds (list[Tensor]): Box energies / deltas for each \
+                    scale level, each is a 4D-tensor, the channel number is \
+                    num_points * 4.
+                centernesses (list[Tensor]): centerness for each scale level, \
+                    each is a 4D-tensor, the channel number is num_points * 1.
+        """
+        return multi_apply(self.forward_single, feats, self.scales,
+                           self.strides)
+
+    def forward_single(self, x, scale, stride):
+        """Forward features of a single scale level.
+
+        Args:
+            x (Tensor): FPN feature maps of the specified stride.
+            scale (:obj: `mmcv.cnn.Scale`): Learnable scale module to resize
+                the bbox prediction.
+            stride (int): The corresponding stride for feature maps, only
+                used to normalize the bbox prediction when self.norm_on_bbox
+                is True.
+
+        Returns:
+            tuple: scores for each class, bbox predictions and centerness \
+                predictions of input feature maps.
+        """
+        cls_score, bbox_pred, cls_feat, reg_feat = super().forward_single(x)
+        if self.centerness_on_reg:
+            centerness = self.conv_centerness(reg_feat)
+        else:
+            centerness = self.conv_centerness(cls_feat)
+        # scale the bbox_pred of different level
+        # float to avoid overflow when enabling FP16
+        bbox_pred = scale(bbox_pred).float()
+        if self.norm_on_bbox:
+            bbox_pred = F.relu(bbox_pred)
+            if not self.training:
+                bbox_pred *= stride
+        else:
+            bbox_pred = bbox_pred.exp()
+        return cls_score, bbox_pred, centerness
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds', 'centernesses'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             centernesses,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute loss of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level,
+                each is a 4D-tensor, the channel number is
+                num_points * num_classes.
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level, each is a 4D-tensor, the channel number is
+                num_points * 4.
+            centernesses (list[Tensor]): centerness for each scale level, each
+                is a 4D-tensor, the channel number is num_points * 1.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        assert len(cls_scores) == len(bbox_preds) == len(centernesses)
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        all_level_points = self.get_points(featmap_sizes, bbox_preds[0].dtype,
+                                           bbox_preds[0].device)
+        labels, bbox_targets = self.get_targets(all_level_points, gt_bboxes,
+                                                gt_labels)
+
+        num_imgs = cls_scores[0].size(0)
+        # flatten cls_scores, bbox_preds and centerness
+        flatten_cls_scores = [
+            cls_score.permute(0, 2, 3, 1).reshape(-1, self.cls_out_channels)
+            for cls_score in cls_scores
+        ]
+        flatten_bbox_preds = [
+            bbox_pred.permute(0, 2, 3, 1).reshape(-1, 4)
+            for bbox_pred in bbox_preds
+        ]
+        flatten_centerness = [
+            centerness.permute(0, 2, 3, 1).reshape(-1)
+            for centerness in centernesses
+        ]
+        flatten_cls_scores = torch.cat(flatten_cls_scores)
+        flatten_bbox_preds = torch.cat(flatten_bbox_preds)
+        flatten_centerness = torch.cat(flatten_centerness)
+        flatten_labels = torch.cat(labels)
+        flatten_bbox_targets = torch.cat(bbox_targets)
+        # repeat points to align with bbox_preds
+        flatten_points = torch.cat(
+            [points.repeat(num_imgs, 1) for points in all_level_points])
+
+        # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+        bg_class_ind = self.num_classes
+        pos_inds = ((flatten_labels >= 0)
+                    & (flatten_labels < bg_class_ind)).nonzero().reshape(-1)
+        num_pos = torch.tensor(
+            len(pos_inds), dtype=torch.float, device=bbox_preds[0].device)
+        num_pos = max(reduce_mean(num_pos), 1.0)
+        loss_cls = self.loss_cls(
+            flatten_cls_scores, flatten_labels, avg_factor=num_pos)
+
+        pos_bbox_preds = flatten_bbox_preds[pos_inds]
+        pos_centerness = flatten_centerness[pos_inds]
+
+        if len(pos_inds) > 0:
+            pos_bbox_targets = flatten_bbox_targets[pos_inds]
+            pos_centerness_targets = self.centerness_target(pos_bbox_targets)
+            pos_points = flatten_points[pos_inds]
+            pos_decoded_bbox_preds = distance2bbox(pos_points, pos_bbox_preds)
+            pos_decoded_target_preds = distance2bbox(pos_points,
+                                                     pos_bbox_targets)
+            # centerness weighted iou loss
+            centerness_denorm = max(
+                reduce_mean(pos_centerness_targets.sum().detach()), 1e-6)
+            loss_bbox = self.loss_bbox(
+                pos_decoded_bbox_preds,
+                pos_decoded_target_preds,
+                weight=pos_centerness_targets,
+                avg_factor=centerness_denorm)
+            loss_centerness = self.loss_centerness(
+                pos_centerness, pos_centerness_targets, avg_factor=num_pos)
+        else:
+            loss_bbox = pos_bbox_preds.sum()
+            loss_centerness = pos_centerness.sum()
+
+        return dict(
+            loss_cls=loss_cls,
+            loss_bbox=loss_bbox,
+            loss_centerness=loss_centerness)
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds', 'centernesses'))
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   centernesses,
+                   img_metas,
+                   cfg=None,
+                   rescale=False,
+                   with_nms=True):
+        """Transform network output for a batch into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                with shape (N, num_points * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_points * 4, H, W).
+            centernesses (list[Tensor]): Centerness for each scale level with
+                shape (N, num_points * 1, H, W).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used. Default: None.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where 5 represent
+                (tl_x, tl_y, br_x, br_y, score) and the score between 0 and 1.
+                The shape of the second tensor in the tuple is (n,), and
+                each element represents the class label of the corresponding
+                box.
+        """
+        assert len(cls_scores) == len(bbox_preds)
+        num_levels = len(cls_scores)
+
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        mlvl_points = self.get_points(featmap_sizes, bbox_preds[0].dtype,
+                                      bbox_preds[0].device)
+
+        cls_score_list = [cls_scores[i].detach() for i in range(num_levels)]
+        bbox_pred_list = [bbox_preds[i].detach() for i in range(num_levels)]
+        centerness_pred_list = [
+            centernesses[i].detach() for i in range(num_levels)
+        ]
+        if torch.onnx.is_in_onnx_export():
+            assert len(
+                img_metas
+            ) == 1, 'Only support one input image while in exporting to ONNX'
+            img_shapes = img_metas[0]['img_shape_for_onnx']
+        else:
+            img_shapes = [
+                img_metas[i]['img_shape']
+                for i in range(cls_scores[0].shape[0])
+            ]
+        scale_factors = [
+            img_metas[i]['scale_factor'] for i in range(cls_scores[0].shape[0])
+        ]
+        result_list = self._get_bboxes(cls_score_list, bbox_pred_list,
+                                       centerness_pred_list, mlvl_points,
+                                       img_shapes, scale_factors, cfg, rescale,
+                                       with_nms)
+        return result_list
+
+    def _get_bboxes(self,
+                    cls_scores,
+                    bbox_preds,
+                    centernesses,
+                    mlvl_points,
+                    img_shapes,
+                    scale_factors,
+                    cfg,
+                    rescale=False,
+                    with_nms=True):
+        """Transform outputs for a single batch item into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for a single scale level
+                with shape (N, num_points * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box energies / deltas for a single scale
+                level with shape (N, num_points * 4, H, W).
+            centernesses (list[Tensor]): Centerness for a single scale level
+                with shape (N, num_points * 4, H, W).
+            mlvl_points (list[Tensor]): Box reference for a single scale level
+                with shape (num_total_points, 4).
+            img_shapes (list[tuple[int]]): Shape of the input image,
+                list[(height, width, 3)].
+            scale_factors (list[ndarray]): Scale factor of the image arrange as
+                (w_scale, h_scale, w_scale, h_scale).
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            tuple(Tensor):
+                det_bboxes (Tensor): BBox predictions in shape (n, 5), where
+                    the first 4 columns are bounding box positions
+                    (tl_x, tl_y, br_x, br_y) and the 5-th column is a score
+                    between 0 and 1.
+                det_labels (Tensor): A (n,) tensor where each item is the
+                    predicted class label of the corresponding box.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_scores) == len(bbox_preds) == len(mlvl_points)
+        device = cls_scores[0].device
+        batch_size = cls_scores[0].shape[0]
+        # convert to tensor to keep tracing
+        nms_pre_tensor = torch.tensor(
+            cfg.get('nms_pre', -1), device=device, dtype=torch.long)
+        mlvl_bboxes = []
+        mlvl_scores = []
+        mlvl_centerness = []
+        for cls_score, bbox_pred, centerness, points in zip(
+                cls_scores, bbox_preds, centernesses, mlvl_points):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            scores = cls_score.permute(0, 2, 3, 1).reshape(
+                batch_size, -1, self.cls_out_channels).sigmoid()
+            centerness = centerness.permute(0, 2, 3,
+                                            1).reshape(batch_size,
+                                                       -1).sigmoid()
+
+            bbox_pred = bbox_pred.permute(0, 2, 3,
+                                          1).reshape(batch_size, -1, 4)
+            # Always keep topk op for dynamic input in onnx
+            if nms_pre_tensor > 0 and (torch.onnx.is_in_onnx_export()
+                                       or scores.shape[-2] > nms_pre_tensor):
+                from torch import _shape_as_tensor
+                # keep shape as tensor and get k
+                num_anchor = _shape_as_tensor(scores)[-2].to(device)
+                nms_pre = torch.where(nms_pre_tensor < num_anchor,
+                                      nms_pre_tensor, num_anchor)
+
+                max_scores, _ = (scores * centerness[..., None]).max(-1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                points = points[topk_inds, :]
+                batch_inds = torch.arange(batch_size).view(
+                    -1, 1).expand_as(topk_inds).long()
+                bbox_pred = bbox_pred[batch_inds, topk_inds, :]
+                scores = scores[batch_inds, topk_inds, :]
+                centerness = centerness[batch_inds, topk_inds]
+
+            bboxes = distance2bbox(points, bbox_pred, max_shape=img_shapes)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+            mlvl_centerness.append(centerness)
+
+        batch_mlvl_bboxes = torch.cat(mlvl_bboxes, dim=1)
+        if rescale:
+            batch_mlvl_bboxes /= batch_mlvl_bboxes.new_tensor(
+                scale_factors).unsqueeze(1)
+        batch_mlvl_scores = torch.cat(mlvl_scores, dim=1)
+        batch_mlvl_centerness = torch.cat(mlvl_centerness, dim=1)
+
+        # Set max number of box to be feed into nms in deployment
+        deploy_nms_pre = cfg.get('deploy_nms_pre', -1)
+        if deploy_nms_pre > 0 and torch.onnx.is_in_onnx_export():
+            batch_mlvl_scores, _ = (
+                batch_mlvl_scores *
+                batch_mlvl_centerness.unsqueeze(2).expand_as(batch_mlvl_scores)
+            ).max(-1)
+            _, topk_inds = batch_mlvl_scores.topk(deploy_nms_pre)
+            batch_inds = torch.arange(batch_mlvl_scores.shape[0]).view(
+                -1, 1).expand_as(topk_inds)
+            batch_mlvl_scores = batch_mlvl_scores[batch_inds, topk_inds, :]
+            batch_mlvl_bboxes = batch_mlvl_bboxes[batch_inds, topk_inds, :]
+            batch_mlvl_centerness = batch_mlvl_centerness[batch_inds,
+                                                          topk_inds]
+
+        # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+        # BG cat_id: num_class
+        padding = batch_mlvl_scores.new_zeros(batch_size,
+                                              batch_mlvl_scores.shape[1], 1)
+        batch_mlvl_scores = torch.cat([batch_mlvl_scores, padding], dim=-1)
+
+        if with_nms:
+            det_results = []
+            for (mlvl_bboxes, mlvl_scores,
+                 mlvl_centerness) in zip(batch_mlvl_bboxes, batch_mlvl_scores,
+                                         batch_mlvl_centerness):
+                det_bbox, det_label = multiclass_nms(
+                    mlvl_bboxes,
+                    mlvl_scores,
+                    cfg.score_thr,
+                    cfg.nms,
+                    cfg.max_per_img,
+                    score_factors=mlvl_centerness)
+                det_results.append(tuple([det_bbox, det_label]))
+        else:
+            det_results = [
+                tuple(mlvl_bs)
+                for mlvl_bs in zip(batch_mlvl_bboxes, batch_mlvl_scores,
+                                   batch_mlvl_centerness)
+            ]
+        return det_results
+
+    def _get_points_single(self,
+                           featmap_size,
+                           stride,
+                           dtype,
+                           device,
+                           flatten=False):
+        """Get points according to feature map sizes."""
+        y, x = super()._get_points_single(featmap_size, stride, dtype, device)
+        points = torch.stack((x.reshape(-1) * stride, y.reshape(-1) * stride),
+                             dim=-1) + stride // 2
+        return points
+
+    def get_targets(self, points, gt_bboxes_list, gt_labels_list):
+        """Compute regression, classification and centerness targets for points
+        in multiple images.
+
+        Args:
+            points (list[Tensor]): Points of each fpn level, each has shape
+                (num_points, 2).
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image,
+                each has shape (num_gt, 4).
+            gt_labels_list (list[Tensor]): Ground truth labels of each box,
+                each has shape (num_gt,).
+
+        Returns:
+            tuple:
+                concat_lvl_labels (list[Tensor]): Labels of each level. \
+                concat_lvl_bbox_targets (list[Tensor]): BBox targets of each \
+                    level.
+        """
+        assert len(points) == len(self.regress_ranges)
+        num_levels = len(points)
+        # expand regress ranges to align with points
+        expanded_regress_ranges = [
+            points[i].new_tensor(self.regress_ranges[i])[None].expand_as(
+                points[i]) for i in range(num_levels)
+        ]
+        # concat all levels points and regress ranges
+        concat_regress_ranges = torch.cat(expanded_regress_ranges, dim=0)
+        concat_points = torch.cat(points, dim=0)
+
+        # the number of points per img, per lvl
+        num_points = [center.size(0) for center in points]
+
+        # get labels and bbox_targets of each image
+        labels_list, bbox_targets_list = multi_apply(
+            self._get_target_single,
+            gt_bboxes_list,
+            gt_labels_list,
+            points=concat_points,
+            regress_ranges=concat_regress_ranges,
+            num_points_per_lvl=num_points)
+
+        # split to per img, per level
+        labels_list = [labels.split(num_points, 0) for labels in labels_list]
+        bbox_targets_list = [
+            bbox_targets.split(num_points, 0)
+            for bbox_targets in bbox_targets_list
+        ]
+
+        # concat per level image
+        concat_lvl_labels = []
+        concat_lvl_bbox_targets = []
+        for i in range(num_levels):
+            concat_lvl_labels.append(
+                torch.cat([labels[i] for labels in labels_list]))
+            bbox_targets = torch.cat(
+                [bbox_targets[i] for bbox_targets in bbox_targets_list])
+            if self.norm_on_bbox:
+                bbox_targets = bbox_targets / self.strides[i]
+            concat_lvl_bbox_targets.append(bbox_targets)
+        return concat_lvl_labels, concat_lvl_bbox_targets
+
+    def _get_target_single(self, gt_bboxes, gt_labels, points, regress_ranges,
+                           num_points_per_lvl):
+        """Compute regression and classification targets for a single image."""
+        num_points = points.size(0)
+        num_gts = gt_labels.size(0)
+        if num_gts == 0:
+            return gt_labels.new_full((num_points,), self.num_classes), \
+                   gt_bboxes.new_zeros((num_points, 4))
+
+        areas = (gt_bboxes[:, 2] - gt_bboxes[:, 0]) * (
+            gt_bboxes[:, 3] - gt_bboxes[:, 1])
+        # TODO: figure out why these two are different
+        # areas = areas[None].expand(num_points, num_gts)
+        areas = areas[None].repeat(num_points, 1)
+        regress_ranges = regress_ranges[:, None, :].expand(
+            num_points, num_gts, 2)
+        gt_bboxes = gt_bboxes[None].expand(num_points, num_gts, 4)
+        xs, ys = points[:, 0], points[:, 1]
+        xs = xs[:, None].expand(num_points, num_gts)
+        ys = ys[:, None].expand(num_points, num_gts)
+
+        left = xs - gt_bboxes[..., 0]
+        right = gt_bboxes[..., 2] - xs
+        top = ys - gt_bboxes[..., 1]
+        bottom = gt_bboxes[..., 3] - ys
+        bbox_targets = torch.stack((left, top, right, bottom), -1)
+
+        if self.center_sampling:
+            # condition1: inside a `center bbox`
+            radius = self.center_sample_radius
+            center_xs = (gt_bboxes[..., 0] + gt_bboxes[..., 2]) / 2
+            center_ys = (gt_bboxes[..., 1] + gt_bboxes[..., 3]) / 2
+            center_gts = torch.zeros_like(gt_bboxes)
+            stride = center_xs.new_zeros(center_xs.shape)
+
+            # project the points on current lvl back to the `original` sizes
+            lvl_begin = 0
+            for lvl_idx, num_points_lvl in enumerate(num_points_per_lvl):
+                lvl_end = lvl_begin + num_points_lvl
+                stride[lvl_begin:lvl_end] = self.strides[lvl_idx] * radius
+                lvl_begin = lvl_end
+
+            x_mins = center_xs - stride
+            y_mins = center_ys - stride
+            x_maxs = center_xs + stride
+            y_maxs = center_ys + stride
+            center_gts[..., 0] = torch.where(x_mins > gt_bboxes[..., 0],
+                                             x_mins, gt_bboxes[..., 0])
+            center_gts[..., 1] = torch.where(y_mins > gt_bboxes[..., 1],
+                                             y_mins, gt_bboxes[..., 1])
+            center_gts[..., 2] = torch.where(x_maxs > gt_bboxes[..., 2],
+                                             gt_bboxes[..., 2], x_maxs)
+            center_gts[..., 3] = torch.where(y_maxs > gt_bboxes[..., 3],
+                                             gt_bboxes[..., 3], y_maxs)
+
+            cb_dist_left = xs - center_gts[..., 0]
+            cb_dist_right = center_gts[..., 2] - xs
+            cb_dist_top = ys - center_gts[..., 1]
+            cb_dist_bottom = center_gts[..., 3] - ys
+            center_bbox = torch.stack(
+                (cb_dist_left, cb_dist_top, cb_dist_right, cb_dist_bottom), -1)
+            inside_gt_bbox_mask = center_bbox.min(-1)[0] > 0
+        else:
+            # condition1: inside a gt bbox
+            inside_gt_bbox_mask = bbox_targets.min(-1)[0] > 0
+
+        # condition2: limit the regression range for each location
+        max_regress_distance = bbox_targets.max(-1)[0]
+        inside_regress_range = (
+            (max_regress_distance >= regress_ranges[..., 0])
+            & (max_regress_distance <= regress_ranges[..., 1]))
+
+        # if there are still more than one objects for a location,
+        # we choose the one with minimal area
+        areas[inside_gt_bbox_mask == 0] = INF
+        areas[inside_regress_range == 0] = INF
+        min_area, min_area_inds = areas.min(dim=1)
+
+        labels = gt_labels[min_area_inds]
+        labels[min_area == INF] = self.num_classes  # set as BG
+        bbox_targets = bbox_targets[range(num_points), min_area_inds]
+
+        return labels, bbox_targets
+
+    def centerness_target(self, pos_bbox_targets):
+        """Compute centerness targets.
+
+        Args:
+            pos_bbox_targets (Tensor): BBox targets of positive bboxes in shape
+                (num_pos, 4)
+
+        Returns:
+            Tensor: Centerness target.
+        """
+        # only calculate pos centerness targets, otherwise there may be nan
+        left_right = pos_bbox_targets[:, [0, 2]]
+        top_bottom = pos_bbox_targets[:, [1, 3]]
+        centerness_targets = (
+            left_right.min(dim=-1)[0] / left_right.max(dim=-1)[0]) * (
+                top_bottom.min(dim=-1)[0] / top_bottom.max(dim=-1)[0])
+        return torch.sqrt(centerness_targets)
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/fovea_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/fovea_head.py
new file mode 100644
index 0000000..c8ccea7
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/fovea_head.py
@@ -0,0 +1,341 @@
+import torch
+import torch.nn as nn
+from mmcv.cnn import ConvModule, normal_init
+from mmcv.ops import DeformConv2d
+
+from mmdet.core import multi_apply, multiclass_nms
+from ..builder import HEADS
+from .anchor_free_head import AnchorFreeHead
+
+INF = 1e8
+
+
+class FeatureAlign(nn.Module):
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 kernel_size=3,
+                 deform_groups=4):
+        super(FeatureAlign, self).__init__()
+        offset_channels = kernel_size * kernel_size * 2
+        self.conv_offset = nn.Conv2d(
+            4, deform_groups * offset_channels, 1, bias=False)
+        self.conv_adaption = DeformConv2d(
+            in_channels,
+            out_channels,
+            kernel_size=kernel_size,
+            padding=(kernel_size - 1) // 2,
+            deform_groups=deform_groups)
+        self.relu = nn.ReLU(inplace=True)
+
+    def init_weights(self):
+        normal_init(self.conv_offset, std=0.1)
+        normal_init(self.conv_adaption, std=0.01)
+
+    def forward(self, x, shape):
+        offset = self.conv_offset(shape)
+        x = self.relu(self.conv_adaption(x, offset))
+        return x
+
+
+@HEADS.register_module()
+class FoveaHead(AnchorFreeHead):
+    """FoveaBox: Beyond Anchor-based Object Detector
+    https://arxiv.org/abs/1904.03797
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 base_edge_list=(16, 32, 64, 128, 256),
+                 scale_ranges=((8, 32), (16, 64), (32, 128), (64, 256), (128,
+                                                                         512)),
+                 sigma=0.4,
+                 with_deform=False,
+                 deform_groups=4,
+                 **kwargs):
+        self.base_edge_list = base_edge_list
+        self.scale_ranges = scale_ranges
+        self.sigma = sigma
+        self.with_deform = with_deform
+        self.deform_groups = deform_groups
+        super().__init__(num_classes, in_channels, **kwargs)
+
+    def _init_layers(self):
+        # box branch
+        super()._init_reg_convs()
+        self.conv_reg = nn.Conv2d(self.feat_channels, 4, 3, padding=1)
+
+        # cls branch
+        if not self.with_deform:
+            super()._init_cls_convs()
+            self.conv_cls = nn.Conv2d(
+                self.feat_channels, self.cls_out_channels, 3, padding=1)
+        else:
+            self.cls_convs = nn.ModuleList()
+            self.cls_convs.append(
+                ConvModule(
+                    self.feat_channels, (self.feat_channels * 4),
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg,
+                    bias=self.norm_cfg is None))
+            self.cls_convs.append(
+                ConvModule((self.feat_channels * 4), (self.feat_channels * 4),
+                           1,
+                           stride=1,
+                           padding=0,
+                           conv_cfg=self.conv_cfg,
+                           norm_cfg=self.norm_cfg,
+                           bias=self.norm_cfg is None))
+            self.feature_adaption = FeatureAlign(
+                self.feat_channels,
+                self.feat_channels,
+                kernel_size=3,
+                deform_groups=self.deform_groups)
+            self.conv_cls = nn.Conv2d(
+                int(self.feat_channels * 4),
+                self.cls_out_channels,
+                3,
+                padding=1)
+
+    def init_weights(self):
+        super().init_weights()
+        if self.with_deform:
+            self.feature_adaption.init_weights()
+
+    def forward_single(self, x):
+        cls_feat = x
+        reg_feat = x
+        for reg_layer in self.reg_convs:
+            reg_feat = reg_layer(reg_feat)
+        bbox_pred = self.conv_reg(reg_feat)
+        if self.with_deform:
+            cls_feat = self.feature_adaption(cls_feat, bbox_pred.exp())
+        for cls_layer in self.cls_convs:
+            cls_feat = cls_layer(cls_feat)
+        cls_score = self.conv_cls(cls_feat)
+        return cls_score, bbox_pred
+
+    def _get_points_single(self, *args, **kwargs):
+        y, x = super()._get_points_single(*args, **kwargs)
+        return y + 0.5, x + 0.5
+
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bbox_list,
+             gt_label_list,
+             img_metas,
+             gt_bboxes_ignore=None):
+        assert len(cls_scores) == len(bbox_preds)
+
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        points = self.get_points(featmap_sizes, bbox_preds[0].dtype,
+                                 bbox_preds[0].device)
+        num_imgs = cls_scores[0].size(0)
+        flatten_cls_scores = [
+            cls_score.permute(0, 2, 3, 1).reshape(-1, self.cls_out_channels)
+            for cls_score in cls_scores
+        ]
+        flatten_bbox_preds = [
+            bbox_pred.permute(0, 2, 3, 1).reshape(-1, 4)
+            for bbox_pred in bbox_preds
+        ]
+        flatten_cls_scores = torch.cat(flatten_cls_scores)
+        flatten_bbox_preds = torch.cat(flatten_bbox_preds)
+        flatten_labels, flatten_bbox_targets = self.get_targets(
+            gt_bbox_list, gt_label_list, featmap_sizes, points)
+
+        # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+        pos_inds = ((flatten_labels >= 0)
+                    & (flatten_labels < self.num_classes)).nonzero().view(-1)
+        num_pos = len(pos_inds)
+
+        loss_cls = self.loss_cls(
+            flatten_cls_scores, flatten_labels, avg_factor=num_pos + num_imgs)
+        if num_pos > 0:
+            pos_bbox_preds = flatten_bbox_preds[pos_inds]
+            pos_bbox_targets = flatten_bbox_targets[pos_inds]
+            pos_weights = pos_bbox_targets.new_zeros(
+                pos_bbox_targets.size()) + 1.0
+            loss_bbox = self.loss_bbox(
+                pos_bbox_preds,
+                pos_bbox_targets,
+                pos_weights,
+                avg_factor=num_pos)
+        else:
+            loss_bbox = torch.tensor(
+                0,
+                dtype=flatten_bbox_preds.dtype,
+                device=flatten_bbox_preds.device)
+        return dict(loss_cls=loss_cls, loss_bbox=loss_bbox)
+
+    def get_targets(self, gt_bbox_list, gt_label_list, featmap_sizes, points):
+        label_list, bbox_target_list = multi_apply(
+            self._get_target_single,
+            gt_bbox_list,
+            gt_label_list,
+            featmap_size_list=featmap_sizes,
+            point_list=points)
+        flatten_labels = [
+            torch.cat([
+                labels_level_img.flatten() for labels_level_img in labels_level
+            ]) for labels_level in zip(*label_list)
+        ]
+        flatten_bbox_targets = [
+            torch.cat([
+                bbox_targets_level_img.reshape(-1, 4)
+                for bbox_targets_level_img in bbox_targets_level
+            ]) for bbox_targets_level in zip(*bbox_target_list)
+        ]
+        flatten_labels = torch.cat(flatten_labels)
+        flatten_bbox_targets = torch.cat(flatten_bbox_targets)
+        return flatten_labels, flatten_bbox_targets
+
+    def _get_target_single(self,
+                           gt_bboxes_raw,
+                           gt_labels_raw,
+                           featmap_size_list=None,
+                           point_list=None):
+
+        gt_areas = torch.sqrt((gt_bboxes_raw[:, 2] - gt_bboxes_raw[:, 0]) *
+                              (gt_bboxes_raw[:, 3] - gt_bboxes_raw[:, 1]))
+        label_list = []
+        bbox_target_list = []
+        # for each pyramid, find the cls and box target
+        for base_len, (lower_bound, upper_bound), stride, featmap_size, \
+            (y, x) in zip(self.base_edge_list, self.scale_ranges,
+                          self.strides, featmap_size_list, point_list):
+            # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+            labels = gt_labels_raw.new_zeros(featmap_size) + self.num_classes
+            bbox_targets = gt_bboxes_raw.new(featmap_size[0], featmap_size[1],
+                                             4) + 1
+            # scale assignment
+            hit_indices = ((gt_areas >= lower_bound) &
+                           (gt_areas <= upper_bound)).nonzero().flatten()
+            if len(hit_indices) == 0:
+                label_list.append(labels)
+                bbox_target_list.append(torch.log(bbox_targets))
+                continue
+            _, hit_index_order = torch.sort(-gt_areas[hit_indices])
+            hit_indices = hit_indices[hit_index_order]
+            gt_bboxes = gt_bboxes_raw[hit_indices, :] / stride
+            gt_labels = gt_labels_raw[hit_indices]
+            half_w = 0.5 * (gt_bboxes[:, 2] - gt_bboxes[:, 0])
+            half_h = 0.5 * (gt_bboxes[:, 3] - gt_bboxes[:, 1])
+            # valid fovea area: left, right, top, down
+            pos_left = torch.ceil(
+                gt_bboxes[:, 0] + (1 - self.sigma) * half_w - 0.5).long().\
+                clamp(0, featmap_size[1] - 1)
+            pos_right = torch.floor(
+                gt_bboxes[:, 0] + (1 + self.sigma) * half_w - 0.5).long().\
+                clamp(0, featmap_size[1] - 1)
+            pos_top = torch.ceil(
+                gt_bboxes[:, 1] + (1 - self.sigma) * half_h - 0.5).long().\
+                clamp(0, featmap_size[0] - 1)
+            pos_down = torch.floor(
+                gt_bboxes[:, 1] + (1 + self.sigma) * half_h - 0.5).long().\
+                clamp(0, featmap_size[0] - 1)
+            for px1, py1, px2, py2, label, (gt_x1, gt_y1, gt_x2, gt_y2) in \
+                    zip(pos_left, pos_top, pos_right, pos_down, gt_labels,
+                        gt_bboxes_raw[hit_indices, :]):
+                labels[py1:py2 + 1, px1:px2 + 1] = label
+                bbox_targets[py1:py2 + 1, px1:px2 + 1, 0] = \
+                    (stride * x[py1:py2 + 1, px1:px2 + 1] - gt_x1) / base_len
+                bbox_targets[py1:py2 + 1, px1:px2 + 1, 1] = \
+                    (stride * y[py1:py2 + 1, px1:px2 + 1] - gt_y1) / base_len
+                bbox_targets[py1:py2 + 1, px1:px2 + 1, 2] = \
+                    (gt_x2 - stride * x[py1:py2 + 1, px1:px2 + 1]) / base_len
+                bbox_targets[py1:py2 + 1, px1:px2 + 1, 3] = \
+                    (gt_y2 - stride * y[py1:py2 + 1, px1:px2 + 1]) / base_len
+            bbox_targets = bbox_targets.clamp(min=1. / 16, max=16.)
+            label_list.append(labels)
+            bbox_target_list.append(torch.log(bbox_targets))
+        return label_list, bbox_target_list
+
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   img_metas,
+                   cfg=None,
+                   rescale=None):
+        assert len(cls_scores) == len(bbox_preds)
+        num_levels = len(cls_scores)
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        points = self.get_points(
+            featmap_sizes,
+            bbox_preds[0].dtype,
+            bbox_preds[0].device,
+            flatten=True)
+        result_list = []
+        for img_id in range(len(img_metas)):
+            cls_score_list = [
+                cls_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_pred_list = [
+                bbox_preds[i][img_id].detach() for i in range(num_levels)
+            ]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            det_bboxes = self._get_bboxes_single(cls_score_list,
+                                                 bbox_pred_list, featmap_sizes,
+                                                 points, img_shape,
+                                                 scale_factor, cfg, rescale)
+            result_list.append(det_bboxes)
+        return result_list
+
+    def _get_bboxes_single(self,
+                           cls_scores,
+                           bbox_preds,
+                           featmap_sizes,
+                           point_list,
+                           img_shape,
+                           scale_factor,
+                           cfg,
+                           rescale=False):
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_scores) == len(bbox_preds) == len(point_list)
+        det_bboxes = []
+        det_scores = []
+        for cls_score, bbox_pred, featmap_size, stride, base_len, (y, x) \
+                in zip(cls_scores, bbox_preds, featmap_sizes, self.strides,
+                       self.base_edge_list, point_list):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            scores = cls_score.permute(1, 2, 0).reshape(
+                -1, self.cls_out_channels).sigmoid()
+            bbox_pred = bbox_pred.permute(1, 2, 0).reshape(-1, 4).exp()
+            nms_pre = cfg.get('nms_pre', -1)
+            if (nms_pre > 0) and (scores.shape[0] > nms_pre):
+                max_scores, _ = scores.max(dim=1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                bbox_pred = bbox_pred[topk_inds, :]
+                scores = scores[topk_inds, :]
+                y = y[topk_inds]
+                x = x[topk_inds]
+            x1 = (stride * x - base_len * bbox_pred[:, 0]).\
+                clamp(min=0, max=img_shape[1] - 1)
+            y1 = (stride * y - base_len * bbox_pred[:, 1]).\
+                clamp(min=0, max=img_shape[0] - 1)
+            x2 = (stride * x + base_len * bbox_pred[:, 2]).\
+                clamp(min=0, max=img_shape[1] - 1)
+            y2 = (stride * y + base_len * bbox_pred[:, 3]).\
+                clamp(min=0, max=img_shape[0] - 1)
+            bboxes = torch.stack([x1, y1, x2, y2], -1)
+            det_bboxes.append(bboxes)
+            det_scores.append(scores)
+        det_bboxes = torch.cat(det_bboxes)
+        if rescale:
+            det_bboxes /= det_bboxes.new_tensor(scale_factor)
+        det_scores = torch.cat(det_scores)
+        padding = det_scores.new_zeros(det_scores.shape[0], 1)
+        # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+        # BG cat_id: num_class
+        det_scores = torch.cat([det_scores, padding], dim=1)
+        det_bboxes, det_labels = multiclass_nms(det_bboxes, det_scores,
+                                                cfg.score_thr, cfg.nms,
+                                                cfg.max_per_img)
+        return det_bboxes, det_labels
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/free_anchor_retina_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/free_anchor_retina_head.py
new file mode 100644
index 0000000..79879fd
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/free_anchor_retina_head.py
@@ -0,0 +1,270 @@
+import torch
+import torch.nn.functional as F
+
+from mmdet.core import bbox_overlaps
+from ..builder import HEADS
+from .retina_head import RetinaHead
+
+EPS = 1e-12
+
+
+@HEADS.register_module()
+class FreeAnchorRetinaHead(RetinaHead):
+    """FreeAnchor RetinaHead used in https://arxiv.org/abs/1909.02466.
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        stacked_convs (int): Number of conv layers in cls and reg tower.
+            Default: 4.
+        conv_cfg (dict): dictionary to construct and config conv layer.
+            Default: None.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+            Default: norm_cfg=dict(type='GN', num_groups=32,
+            requires_grad=True).
+        pre_anchor_topk (int): Number of boxes that be token in each bag.
+        bbox_thr (float): The threshold of the saturated linear function. It is
+            usually the same with the IoU threshold used in NMS.
+        gamma (float): Gamma parameter in focal loss.
+        alpha (float): Alpha parameter in focal loss.
+    """  # noqa: W605
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 stacked_convs=4,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 pre_anchor_topk=50,
+                 bbox_thr=0.6,
+                 gamma=2.0,
+                 alpha=0.5,
+                 **kwargs):
+        super(FreeAnchorRetinaHead,
+              self).__init__(num_classes, in_channels, stacked_convs, conv_cfg,
+                             norm_cfg, **kwargs)
+
+        self.pre_anchor_topk = pre_anchor_topk
+        self.bbox_thr = bbox_thr
+        self.gamma = gamma
+        self.alpha = alpha
+
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            gt_bboxes (list[Tensor]): each item are the truth boxes for each
+                image in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == len(self.anchor_generator.base_anchors)
+
+        anchor_list, _ = self.get_anchors(featmap_sizes, img_metas)
+        anchors = [torch.cat(anchor) for anchor in anchor_list]
+
+        # concatenate each level
+        cls_scores = [
+            cls.permute(0, 2, 3,
+                        1).reshape(cls.size(0), -1, self.cls_out_channels)
+            for cls in cls_scores
+        ]
+        bbox_preds = [
+            bbox_pred.permute(0, 2, 3, 1).reshape(bbox_pred.size(0), -1, 4)
+            for bbox_pred in bbox_preds
+        ]
+        cls_scores = torch.cat(cls_scores, dim=1)
+        bbox_preds = torch.cat(bbox_preds, dim=1)
+
+        cls_prob = torch.sigmoid(cls_scores)
+        box_prob = []
+        num_pos = 0
+        positive_losses = []
+        for _, (anchors_, gt_labels_, gt_bboxes_, cls_prob_,
+                bbox_preds_) in enumerate(
+                    zip(anchors, gt_labels, gt_bboxes, cls_prob, bbox_preds)):
+
+            with torch.no_grad():
+                if len(gt_bboxes_) == 0:
+                    image_box_prob = torch.zeros(
+                        anchors_.size(0),
+                        self.cls_out_channels).type_as(bbox_preds_)
+                else:
+                    # box_localization: a_{j}^{loc}, shape: [j, 4]
+                    pred_boxes = self.bbox_coder.decode(anchors_, bbox_preds_)
+
+                    # object_box_iou: IoU_{ij}^{loc}, shape: [i, j]
+                    object_box_iou = bbox_overlaps(gt_bboxes_, pred_boxes)
+
+                    # object_box_prob: P{a_{j} -> b_{i}}, shape: [i, j]
+                    t1 = self.bbox_thr
+                    t2 = object_box_iou.max(
+                        dim=1, keepdim=True).values.clamp(min=t1 + 1e-12)
+                    object_box_prob = ((object_box_iou - t1) /
+                                       (t2 - t1)).clamp(
+                                           min=0, max=1)
+
+                    # object_cls_box_prob: P{a_{j} -> b_{i}}, shape: [i, c, j]
+                    num_obj = gt_labels_.size(0)
+                    indices = torch.stack([
+                        torch.arange(num_obj).type_as(gt_labels_), gt_labels_
+                    ],
+                                          dim=0)
+                    object_cls_box_prob = torch.sparse_coo_tensor(
+                        indices, object_box_prob)
+
+                    # image_box_iou: P{a_{j} \in A_{+}}, shape: [c, j]
+                    """
+                    from "start" to "end" implement:
+                    image_box_iou = torch.sparse.max(object_cls_box_prob,
+                                                     dim=0).t()
+
+                    """
+                    # start
+                    box_cls_prob = torch.sparse.sum(
+                        object_cls_box_prob, dim=0).to_dense()
+
+                    indices = torch.nonzero(box_cls_prob, as_tuple=False).t_()
+                    if indices.numel() == 0:
+                        image_box_prob = torch.zeros(
+                            anchors_.size(0),
+                            self.cls_out_channels).type_as(object_box_prob)
+                    else:
+                        nonzero_box_prob = torch.where(
+                            (gt_labels_.unsqueeze(dim=-1) == indices[0]),
+                            object_box_prob[:, indices[1]],
+                            torch.tensor([
+                                0
+                            ]).type_as(object_box_prob)).max(dim=0).values
+
+                        # upmap to shape [j, c]
+                        image_box_prob = torch.sparse_coo_tensor(
+                            indices.flip([0]),
+                            nonzero_box_prob,
+                            size=(anchors_.size(0),
+                                  self.cls_out_channels)).to_dense()
+                    # end
+
+                box_prob.append(image_box_prob)
+
+            # construct bags for objects
+            match_quality_matrix = bbox_overlaps(gt_bboxes_, anchors_)
+            _, matched = torch.topk(
+                match_quality_matrix,
+                self.pre_anchor_topk,
+                dim=1,
+                sorted=False)
+            del match_quality_matrix
+
+            # matched_cls_prob: P_{ij}^{cls}
+            matched_cls_prob = torch.gather(
+                cls_prob_[matched], 2,
+                gt_labels_.view(-1, 1, 1).repeat(1, self.pre_anchor_topk,
+                                                 1)).squeeze(2)
+
+            # matched_box_prob: P_{ij}^{loc}
+            matched_anchors = anchors_[matched]
+            matched_object_targets = self.bbox_coder.encode(
+                matched_anchors,
+                gt_bboxes_.unsqueeze(dim=1).expand_as(matched_anchors))
+            loss_bbox = self.loss_bbox(
+                bbox_preds_[matched],
+                matched_object_targets,
+                reduction_override='none').sum(-1)
+            matched_box_prob = torch.exp(-loss_bbox)
+
+            # positive_losses: {-log( Mean-max(P_{ij}^{cls} * P_{ij}^{loc}) )}
+            num_pos += len(gt_bboxes_)
+            positive_losses.append(
+                self.positive_bag_loss(matched_cls_prob, matched_box_prob))
+        positive_loss = torch.cat(positive_losses).sum() / max(1, num_pos)
+
+        # box_prob: P{a_{j} \in A_{+}}
+        box_prob = torch.stack(box_prob, dim=0)
+
+        # negative_loss:
+        # \sum_{j}{ FL((1 - P{a_{j} \in A_{+}}) * (1 - P_{j}^{bg})) } / n||B||
+        negative_loss = self.negative_bag_loss(cls_prob, box_prob).sum() / max(
+            1, num_pos * self.pre_anchor_topk)
+
+        # avoid the absence of gradients in regression subnet
+        # when no ground-truth in a batch
+        if num_pos == 0:
+            positive_loss = bbox_preds.sum() * 0
+
+        losses = {
+            'positive_bag_loss': positive_loss,
+            'negative_bag_loss': negative_loss
+        }
+        return losses
+
+    def positive_bag_loss(self, matched_cls_prob, matched_box_prob):
+        """Compute positive bag loss.
+
+        :math:`-log( Mean-max(P_{ij}^{cls} * P_{ij}^{loc}) )`.
+
+        :math:`P_{ij}^{cls}`: matched_cls_prob, classification probability of matched samples.
+
+        :math:`P_{ij}^{loc}`: matched_box_prob, box probability of matched samples.
+
+        Args:
+            matched_cls_prob (Tensor): Classification probabilty of matched
+                samples in shape (num_gt, pre_anchor_topk).
+            matched_box_prob (Tensor): BBox probability of matched samples,
+                in shape (num_gt, pre_anchor_topk).
+
+        Returns:
+            Tensor: Positive bag loss in shape (num_gt,).
+        """  # noqa: E501, W605
+        # bag_prob = Mean-max(matched_prob)
+        matched_prob = matched_cls_prob * matched_box_prob
+        weight = 1 / torch.clamp(1 - matched_prob, 1e-12, None)
+        weight /= weight.sum(dim=1).unsqueeze(dim=-1)
+        bag_prob = (weight * matched_prob).sum(dim=1)
+        # positive_bag_loss = -self.alpha * log(bag_prob)
+        return self.alpha * F.binary_cross_entropy(
+            bag_prob, torch.ones_like(bag_prob), reduction='none')
+
+    def negative_bag_loss(self, cls_prob, box_prob):
+        """Compute negative bag loss.
+
+        :math:`FL((1 - P_{a_{j} \in A_{+}}) * (1 - P_{j}^{bg}))`.
+
+        :math:`P_{a_{j} \in A_{+}}`: Box_probability of matched samples.
+
+        :math:`P_{j}^{bg}`: Classification probability of negative samples.
+
+        Args:
+            cls_prob (Tensor): Classification probability, in shape
+                (num_img, num_anchors, num_classes).
+            box_prob (Tensor): Box probability, in shape
+                (num_img, num_anchors, num_classes).
+
+        Returns:
+            Tensor: Negative bag loss in shape (num_img, num_anchors, num_classes).
+        """  # noqa: E501, W605
+        prob = cls_prob * (1 - box_prob)
+        # There are some cases when neg_prob = 0.
+        # This will cause the neg_prob.log() to be inf without clamp.
+        prob = prob.clamp(min=EPS, max=1 - EPS)
+        negative_bag_loss = prob**self.gamma * F.binary_cross_entropy(
+            prob, torch.zeros_like(prob), reduction='none')
+        return (1 - self.alpha) * negative_bag_loss
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/fsaf_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/fsaf_head.py
new file mode 100644
index 0000000..7183efc
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/fsaf_head.py
@@ -0,0 +1,422 @@
+import numpy as np
+import torch
+from mmcv.cnn import normal_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import (anchor_inside_flags, images_to_levels, multi_apply,
+                        unmap)
+from ..builder import HEADS
+from ..losses.accuracy import accuracy
+from ..losses.utils import weight_reduce_loss
+from .retina_head import RetinaHead
+
+
+@HEADS.register_module()
+class FSAFHead(RetinaHead):
+    """Anchor-free head used in `FSAF <https://arxiv.org/abs/1903.00621>`_.
+
+    The head contains two subnetworks. The first classifies anchor boxes and
+    the second regresses deltas for the anchors (num_anchors is 1 for anchor-
+    free methods)
+
+    Args:
+        *args: Same as its base class in :class:`RetinaHead`
+        score_threshold (float, optional): The score_threshold to calculate
+            positive recall. If given, prediction scores lower than this value
+            is counted as incorrect prediction. Default to None.
+        **kwargs: Same as its base class in :class:`RetinaHead`
+
+    Example:
+        >>> import torch
+        >>> self = FSAFHead(11, 7)
+        >>> x = torch.rand(1, 7, 32, 32)
+        >>> cls_score, bbox_pred = self.forward_single(x)
+        >>> # Each anchor predicts a score for each class except background
+        >>> cls_per_anchor = cls_score.shape[1] / self.num_anchors
+        >>> box_per_anchor = bbox_pred.shape[1] / self.num_anchors
+        >>> assert cls_per_anchor == self.num_classes
+        >>> assert box_per_anchor == 4
+    """
+
+    def __init__(self, *args, score_threshold=None, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.score_threshold = score_threshold
+
+    def forward_single(self, x):
+        """Forward feature map of a single scale level.
+
+        Args:
+            x (Tensor): Feature map of a single scale level.
+
+        Returns:
+            tuple (Tensor):
+                cls_score (Tensor): Box scores for each scale level
+                    Has shape (N, num_points * num_classes, H, W).
+                bbox_pred (Tensor): Box energies / deltas for each scale
+                    level with shape (N, num_points * 4, H, W).
+        """
+        cls_score, bbox_pred = super().forward_single(x)
+        # relu: TBLR encoder only accepts positive bbox_pred
+        return cls_score, self.relu(bbox_pred)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        super(FSAFHead, self).init_weights()
+        # The positive bias in self.retina_reg conv is to prevent predicted \
+        #  bbox with 0 area
+        normal_init(self.retina_reg, std=0.01, bias=0.25)
+
+    def _get_targets_single(self,
+                            flat_anchors,
+                            valid_flags,
+                            gt_bboxes,
+                            gt_bboxes_ignore,
+                            gt_labels,
+                            img_meta,
+                            label_channels=1,
+                            unmap_outputs=True):
+        """Compute regression and classification targets for anchors in a
+        single image.
+
+        Most of the codes are the same with the base class
+          :obj: `AnchorHead`, except that it also collects and returns
+          the matched gt index in the image (from 0 to num_gt-1). If the
+          anchor bbox is not matched to any gt, the corresponding value in
+          pos_gt_inds is -1.
+        """
+        inside_flags = anchor_inside_flags(flat_anchors, valid_flags,
+                                           img_meta['img_shape'][:2],
+                                           self.train_cfg.allowed_border)
+        if not inside_flags.any():
+            return (None, ) * 7
+        # Assign gt and sample anchors
+        anchors = flat_anchors[inside_flags.type(torch.bool), :]
+        assign_result = self.assigner.assign(
+            anchors, gt_bboxes, gt_bboxes_ignore,
+            None if self.sampling else gt_labels)
+
+        sampling_result = self.sampler.sample(assign_result, anchors,
+                                              gt_bboxes)
+
+        num_valid_anchors = anchors.shape[0]
+        bbox_targets = torch.zeros_like(anchors)
+        bbox_weights = torch.zeros_like(anchors)
+        labels = anchors.new_full((num_valid_anchors, ),
+                                  self.num_classes,
+                                  dtype=torch.long)
+        label_weights = anchors.new_zeros((num_valid_anchors, label_channels),
+                                          dtype=torch.float)
+        pos_gt_inds = anchors.new_full((num_valid_anchors, ),
+                                       -1,
+                                       dtype=torch.long)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+
+        if len(pos_inds) > 0:
+            if not self.reg_decoded_bbox:
+                pos_bbox_targets = self.bbox_coder.encode(
+                    sampling_result.pos_bboxes, sampling_result.pos_gt_bboxes)
+            else:
+                # When the regression loss (e.g. `IouLoss`, `GIouLoss`)
+                # is applied directly on the decoded bounding boxes, both
+                # the predicted boxes and regression targets should be with
+                # absolute coordinate format.
+                pos_bbox_targets = sampling_result.pos_gt_bboxes
+            bbox_targets[pos_inds, :] = pos_bbox_targets
+            bbox_weights[pos_inds, :] = 1.0
+            # The assigned gt_index for each anchor. (0-based)
+            pos_gt_inds[pos_inds] = sampling_result.pos_assigned_gt_inds
+            if gt_labels is None:
+                # Only rpn gives gt_labels as None
+                # Foreground is the first class
+                labels[pos_inds] = 0
+            else:
+                labels[pos_inds] = gt_labels[
+                    sampling_result.pos_assigned_gt_inds]
+            if self.train_cfg.pos_weight <= 0:
+                label_weights[pos_inds] = 1.0
+            else:
+                label_weights[pos_inds] = self.train_cfg.pos_weight
+
+        if len(neg_inds) > 0:
+            label_weights[neg_inds] = 1.0
+
+        # shadowed_labels is a tensor composed of tuples
+        #  (anchor_inds, class_label) that indicate those anchors lying in the
+        #  outer region of a gt or overlapped by another gt with a smaller
+        #  area.
+        #
+        # Therefore, only the shadowed labels are ignored for loss calculation.
+        # the key `shadowed_labels` is defined in :obj:`CenterRegionAssigner`
+        shadowed_labels = assign_result.get_extra_property('shadowed_labels')
+        if shadowed_labels is not None and shadowed_labels.numel():
+            if len(shadowed_labels.shape) == 2:
+                idx_, label_ = shadowed_labels[:, 0], shadowed_labels[:, 1]
+                assert (labels[idx_] != label_).all(), \
+                    'One label cannot be both positive and ignored'
+                label_weights[idx_, label_] = 0
+            else:
+                label_weights[shadowed_labels] = 0
+
+        # map up to original set of anchors
+        if unmap_outputs:
+            num_total_anchors = flat_anchors.size(0)
+            labels = unmap(labels, num_total_anchors, inside_flags)
+            label_weights = unmap(label_weights, num_total_anchors,
+                                  inside_flags)
+            bbox_targets = unmap(bbox_targets, num_total_anchors, inside_flags)
+            bbox_weights = unmap(bbox_weights, num_total_anchors, inside_flags)
+            pos_gt_inds = unmap(
+                pos_gt_inds, num_total_anchors, inside_flags, fill=-1)
+
+        return (labels, label_weights, bbox_targets, bbox_weights, pos_inds,
+                neg_inds, sampling_result, pos_gt_inds)
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute loss of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_points * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_points * 4, H, W).
+            gt_bboxes (list[Tensor]): each item are the truth boxes for each
+                image in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        for i in range(len(bbox_preds)):  # loop over fpn level
+            # avoid 0 area of the predicted bbox
+            bbox_preds[i] = bbox_preds[i].clamp(min=1e-4)
+        # TODO: It may directly use the base-class loss function.
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+        batch_size = len(gt_bboxes)
+        device = cls_scores[0].device
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg,
+         pos_assigned_gt_inds_list) = cls_reg_targets
+
+        num_gts = np.array(list(map(len, gt_labels)))
+        num_total_samples = (
+            num_total_pos + num_total_neg if self.sampling else num_total_pos)
+        # anchor number of multi levels
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+        # concat all level anchors and flags to a single tensor
+        concat_anchor_list = []
+        for i in range(len(anchor_list)):
+            concat_anchor_list.append(torch.cat(anchor_list[i]))
+        all_anchor_list = images_to_levels(concat_anchor_list,
+                                           num_level_anchors)
+        losses_cls, losses_bbox = multi_apply(
+            self.loss_single,
+            cls_scores,
+            bbox_preds,
+            all_anchor_list,
+            labels_list,
+            label_weights_list,
+            bbox_targets_list,
+            bbox_weights_list,
+            num_total_samples=num_total_samples)
+
+        # `pos_assigned_gt_inds_list` (length: fpn_levels) stores the assigned
+        # gt index of each anchor bbox in each fpn level.
+        cum_num_gts = list(np.cumsum(num_gts))  # length of batch_size
+        for i, assign in enumerate(pos_assigned_gt_inds_list):
+            # loop over fpn levels
+            for j in range(1, batch_size):
+                # loop over batch size
+                # Convert gt indices in each img to those in the batch
+                assign[j][assign[j] >= 0] += int(cum_num_gts[j - 1])
+            pos_assigned_gt_inds_list[i] = assign.flatten()
+            labels_list[i] = labels_list[i].flatten()
+        num_gts = sum(map(len, gt_labels))  # total number of gt in the batch
+        # The unique label index of each gt in the batch
+        label_sequence = torch.arange(num_gts, device=device)
+        # Collect the average loss of each gt in each level
+        with torch.no_grad():
+            loss_levels, = multi_apply(
+                self.collect_loss_level_single,
+                losses_cls,
+                losses_bbox,
+                pos_assigned_gt_inds_list,
+                labels_seq=label_sequence)
+            # Shape: (fpn_levels, num_gts). Loss of each gt at each fpn level
+            loss_levels = torch.stack(loss_levels, dim=0)
+            # Locate the best fpn level for loss back-propagation
+            if loss_levels.numel() == 0:  # zero gt
+                argmin = loss_levels.new_empty((num_gts, ), dtype=torch.long)
+            else:
+                _, argmin = loss_levels.min(dim=0)
+
+        # Reweight the loss of each (anchor, label) pair, so that only those
+        #  at the best gt level are back-propagated.
+        losses_cls, losses_bbox, pos_inds = multi_apply(
+            self.reweight_loss_single,
+            losses_cls,
+            losses_bbox,
+            pos_assigned_gt_inds_list,
+            labels_list,
+            list(range(len(losses_cls))),
+            min_levels=argmin)
+        num_pos = torch.cat(pos_inds, 0).sum().float()
+        pos_recall = self.calculate_pos_recall(cls_scores, labels_list,
+                                               pos_inds)
+
+        if num_pos == 0:  # No gt
+            avg_factor = num_pos + float(num_total_neg)
+        else:
+            avg_factor = num_pos
+        for i in range(len(losses_cls)):
+            losses_cls[i] /= avg_factor
+            losses_bbox[i] /= avg_factor
+        return dict(
+            loss_cls=losses_cls,
+            loss_bbox=losses_bbox,
+            num_pos=num_pos / batch_size,
+            pos_recall=pos_recall)
+
+    def calculate_pos_recall(self, cls_scores, labels_list, pos_inds):
+        """Calculate positive recall with score threshold.
+
+        Args:
+            cls_scores (list[Tensor]): Classification scores at all fpn levels.
+                Each tensor is in shape (N, num_classes * num_anchors, H, W)
+            labels_list (list[Tensor]): The label that each anchor is assigned
+                to. Shape (N * H * W * num_anchors, )
+            pos_inds (list[Tensor]): List of bool tensors indicating whether
+                the anchor is assigned to a positive label.
+                Shape (N * H * W * num_anchors, )
+
+        Returns:
+            Tensor: A single float number indicating the positive recall.
+        """
+        with torch.no_grad():
+            num_class = self.num_classes
+            scores = [
+                cls.permute(0, 2, 3, 1).reshape(-1, num_class)[pos]
+                for cls, pos in zip(cls_scores, pos_inds)
+            ]
+            labels = [
+                label.reshape(-1)[pos]
+                for label, pos in zip(labels_list, pos_inds)
+            ]
+            scores = torch.cat(scores, dim=0)
+            labels = torch.cat(labels, dim=0)
+            if self.use_sigmoid_cls:
+                scores = scores.sigmoid()
+            else:
+                scores = scores.softmax(dim=1)
+
+            return accuracy(scores, labels, thresh=self.score_threshold)
+
+    def collect_loss_level_single(self, cls_loss, reg_loss, assigned_gt_inds,
+                                  labels_seq):
+        """Get the average loss in each FPN level w.r.t. each gt label.
+
+        Args:
+            cls_loss (Tensor): Classification loss of each feature map pixel,
+              shape (num_anchor, num_class)
+            reg_loss (Tensor): Regression loss of each feature map pixel,
+              shape (num_anchor, 4)
+            assigned_gt_inds (Tensor): It indicates which gt the prior is
+              assigned to (0-based, -1: no assignment). shape (num_anchor),
+            labels_seq: The rank of labels. shape (num_gt)
+
+        Returns:
+            shape: (num_gt), average loss of each gt in this level
+        """
+        if len(reg_loss.shape) == 2:  # iou loss has shape (num_prior, 4)
+            reg_loss = reg_loss.sum(dim=-1)  # sum loss in tblr dims
+        if len(cls_loss.shape) == 2:
+            cls_loss = cls_loss.sum(dim=-1)  # sum loss in class dims
+        loss = cls_loss + reg_loss
+        assert loss.size(0) == assigned_gt_inds.size(0)
+        # Default loss value is 1e6 for a layer where no anchor is positive
+        #  to ensure it will not be chosen to back-propagate gradient
+        losses_ = loss.new_full(labels_seq.shape, 1e6)
+        for i, l in enumerate(labels_seq):
+            match = assigned_gt_inds == l
+            if match.any():
+                losses_[i] = loss[match].mean()
+        return losses_,
+
+    def reweight_loss_single(self, cls_loss, reg_loss, assigned_gt_inds,
+                             labels, level, min_levels):
+        """Reweight loss values at each level.
+
+        Reassign loss values at each level by masking those where the
+        pre-calculated loss is too large. Then return the reduced losses.
+
+        Args:
+            cls_loss (Tensor): Element-wise classification loss.
+              Shape: (num_anchors, num_classes)
+            reg_loss (Tensor): Element-wise regression loss.
+              Shape: (num_anchors, 4)
+            assigned_gt_inds (Tensor): The gt indices that each anchor bbox
+              is assigned to. -1 denotes a negative anchor, otherwise it is the
+              gt index (0-based). Shape: (num_anchors, ),
+            labels (Tensor): Label assigned to anchors. Shape: (num_anchors, ).
+            level (int): The current level index in the pyramid
+              (0-4 for RetinaNet)
+            min_levels (Tensor): The best-matching level for each gt.
+              Shape: (num_gts, ),
+
+        Returns:
+            tuple:
+                - cls_loss: Reduced corrected classification loss. Scalar.
+                - reg_loss: Reduced corrected regression loss. Scalar.
+                - pos_flags (Tensor): Corrected bool tensor indicating the
+                  final positive anchors. Shape: (num_anchors, ).
+        """
+        loc_weight = torch.ones_like(reg_loss)
+        cls_weight = torch.ones_like(cls_loss)
+        pos_flags = assigned_gt_inds >= 0  # positive pixel flag
+        pos_indices = torch.nonzero(pos_flags, as_tuple=False).flatten()
+
+        if pos_flags.any():  # pos pixels exist
+            pos_assigned_gt_inds = assigned_gt_inds[pos_flags]
+            zeroing_indices = (min_levels[pos_assigned_gt_inds] != level)
+            neg_indices = pos_indices[zeroing_indices]
+
+            if neg_indices.numel():
+                pos_flags[neg_indices] = 0
+                loc_weight[neg_indices] = 0
+                # Only the weight corresponding to the label is
+                #  zeroed out if not selected
+                zeroing_labels = labels[neg_indices]
+                assert (zeroing_labels >= 0).all()
+                cls_weight[neg_indices, zeroing_labels] = 0
+
+        # Weighted loss for both cls and reg loss
+        cls_loss = weight_reduce_loss(cls_loss, cls_weight, reduction='sum')
+        reg_loss = weight_reduce_loss(reg_loss, loc_weight, reduction='sum')
+
+        return cls_loss, reg_loss, pos_flags
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/ga_retina_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/ga_retina_head.py
new file mode 100644
index 0000000..8822d1c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/ga_retina_head.py
@@ -0,0 +1,109 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, bias_init_with_prob, normal_init
+from mmcv.ops import MaskedConv2d
+
+from ..builder import HEADS
+from .guided_anchor_head import FeatureAdaption, GuidedAnchorHead
+
+
+@HEADS.register_module()
+class GARetinaHead(GuidedAnchorHead):
+    """Guided-Anchor-based RetinaNet head."""
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 stacked_convs=4,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 **kwargs):
+        self.stacked_convs = stacked_convs
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        super(GARetinaHead, self).__init__(num_classes, in_channels, **kwargs)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.relu = nn.ReLU(inplace=True)
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            self.cls_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+            self.reg_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+
+        self.conv_loc = nn.Conv2d(self.feat_channels, 1, 1)
+        self.conv_shape = nn.Conv2d(self.feat_channels, self.num_anchors * 2,
+                                    1)
+        self.feature_adaption_cls = FeatureAdaption(
+            self.feat_channels,
+            self.feat_channels,
+            kernel_size=3,
+            deform_groups=self.deform_groups)
+        self.feature_adaption_reg = FeatureAdaption(
+            self.feat_channels,
+            self.feat_channels,
+            kernel_size=3,
+            deform_groups=self.deform_groups)
+        self.retina_cls = MaskedConv2d(
+            self.feat_channels,
+            self.num_anchors * self.cls_out_channels,
+            3,
+            padding=1)
+        self.retina_reg = MaskedConv2d(
+            self.feat_channels, self.num_anchors * 4, 3, padding=1)
+
+    def init_weights(self):
+        """Initialize weights of the layer."""
+        for m in self.cls_convs:
+            normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            normal_init(m.conv, std=0.01)
+
+        self.feature_adaption_cls.init_weights()
+        self.feature_adaption_reg.init_weights()
+
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.conv_loc, std=0.01, bias=bias_cls)
+        normal_init(self.conv_shape, std=0.01)
+        normal_init(self.retina_cls, std=0.01, bias=bias_cls)
+        normal_init(self.retina_reg, std=0.01)
+
+    def forward_single(self, x):
+        """Forward feature map of a single scale level."""
+        cls_feat = x
+        reg_feat = x
+        for cls_conv in self.cls_convs:
+            cls_feat = cls_conv(cls_feat)
+        for reg_conv in self.reg_convs:
+            reg_feat = reg_conv(reg_feat)
+
+        loc_pred = self.conv_loc(cls_feat)
+        shape_pred = self.conv_shape(reg_feat)
+
+        cls_feat = self.feature_adaption_cls(cls_feat, shape_pred)
+        reg_feat = self.feature_adaption_reg(reg_feat, shape_pred)
+
+        if not self.training:
+            mask = loc_pred.sigmoid()[0] >= self.loc_filter_thr
+        else:
+            mask = None
+        cls_score = self.retina_cls(cls_feat, mask)
+        bbox_pred = self.retina_reg(reg_feat, mask)
+        return cls_score, bbox_pred, shape_pred, loc_pred
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/ga_rpn_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/ga_rpn_head.py
new file mode 100644
index 0000000..2ec0d4f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/ga_rpn_head.py
@@ -0,0 +1,171 @@
+import copy
+import warnings
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv import ConfigDict
+from mmcv.cnn import normal_init
+from mmcv.ops import nms
+
+from ..builder import HEADS
+from .guided_anchor_head import GuidedAnchorHead
+from .rpn_test_mixin import RPNTestMixin
+
+
+@HEADS.register_module()
+class GARPNHead(RPNTestMixin, GuidedAnchorHead):
+    """Guided-Anchor-based RPN head."""
+
+    def __init__(self, in_channels, **kwargs):
+        super(GARPNHead, self).__init__(1, in_channels, **kwargs)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.rpn_conv = nn.Conv2d(
+            self.in_channels, self.feat_channels, 3, padding=1)
+        super(GARPNHead, self)._init_layers()
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        normal_init(self.rpn_conv, std=0.01)
+        super(GARPNHead, self).init_weights()
+
+    def forward_single(self, x):
+        """Forward feature of a single scale level."""
+
+        x = self.rpn_conv(x)
+        x = F.relu(x, inplace=True)
+        (cls_score, bbox_pred, shape_pred,
+         loc_pred) = super(GARPNHead, self).forward_single(x)
+        return cls_score, bbox_pred, shape_pred, loc_pred
+
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             shape_preds,
+             loc_preds,
+             gt_bboxes,
+             img_metas,
+             gt_bboxes_ignore=None):
+        losses = super(GARPNHead, self).loss(
+            cls_scores,
+            bbox_preds,
+            shape_preds,
+            loc_preds,
+            gt_bboxes,
+            None,
+            img_metas,
+            gt_bboxes_ignore=gt_bboxes_ignore)
+        return dict(
+            loss_rpn_cls=losses['loss_cls'],
+            loss_rpn_bbox=losses['loss_bbox'],
+            loss_anchor_shape=losses['loss_shape'],
+            loss_anchor_loc=losses['loss_loc'])
+
+    def _get_bboxes_single(self,
+                           cls_scores,
+                           bbox_preds,
+                           mlvl_anchors,
+                           mlvl_masks,
+                           img_shape,
+                           scale_factor,
+                           cfg,
+                           rescale=False):
+        cfg = self.test_cfg if cfg is None else cfg
+
+        cfg = copy.deepcopy(cfg)
+
+        # deprecate arguments warning
+        if 'nms' not in cfg or 'max_num' in cfg or 'nms_thr' in cfg:
+            warnings.warn(
+                'In rpn_proposal or test_cfg, '
+                'nms_thr has been moved to a dict named nms as '
+                'iou_threshold, max_num has been renamed as max_per_img, '
+                'name of original arguments and the way to specify '
+                'iou_threshold of NMS will be deprecated.')
+        if 'nms' not in cfg:
+            cfg.nms = ConfigDict(dict(type='nms', iou_threshold=cfg.nms_thr))
+        if 'max_num' in cfg:
+            if 'max_per_img' in cfg:
+                assert cfg.max_num == cfg.max_per_img, f'You ' \
+                    f'set max_num and max_per_img at the same time, ' \
+                    f'but get {cfg.max_num} ' \
+                    f'and {cfg.max_per_img} respectively' \
+                    'Please delete max_num which will be deprecated.'
+            else:
+                cfg.max_per_img = cfg.max_num
+        if 'nms_thr' in cfg:
+            assert cfg.nms.iou_threshold == cfg.nms_thr, f'You set ' \
+                f'iou_threshold in nms and ' \
+                f'nms_thr at the same time, but get ' \
+                f'{cfg.nms.iou_threshold} and {cfg.nms_thr}' \
+                f' respectively. Please delete the ' \
+                f'nms_thr which will be deprecated.'
+
+        assert cfg.nms.get('type', 'nms') == 'nms', 'GARPNHead only support ' \
+            'naive nms.'
+
+        mlvl_proposals = []
+        for idx in range(len(cls_scores)):
+            rpn_cls_score = cls_scores[idx]
+            rpn_bbox_pred = bbox_preds[idx]
+            anchors = mlvl_anchors[idx]
+            mask = mlvl_masks[idx]
+            assert rpn_cls_score.size()[-2:] == rpn_bbox_pred.size()[-2:]
+            # if no location is kept, end.
+            if mask.sum() == 0:
+                continue
+            rpn_cls_score = rpn_cls_score.permute(1, 2, 0)
+            if self.use_sigmoid_cls:
+                rpn_cls_score = rpn_cls_score.reshape(-1)
+                scores = rpn_cls_score.sigmoid()
+            else:
+                rpn_cls_score = rpn_cls_score.reshape(-1, 2)
+                # remind that we set FG labels to [0, num_class-1]
+                # since mmdet v2.0
+                # BG cat_id: num_class
+                scores = rpn_cls_score.softmax(dim=1)[:, :-1]
+            # filter scores, bbox_pred w.r.t. mask.
+            # anchors are filtered in get_anchors() beforehand.
+            scores = scores[mask]
+            rpn_bbox_pred = rpn_bbox_pred.permute(1, 2, 0).reshape(-1,
+                                                                   4)[mask, :]
+            if scores.dim() == 0:
+                rpn_bbox_pred = rpn_bbox_pred.unsqueeze(0)
+                anchors = anchors.unsqueeze(0)
+                scores = scores.unsqueeze(0)
+            # filter anchors, bbox_pred, scores w.r.t. scores
+            if cfg.nms_pre > 0 and scores.shape[0] > cfg.nms_pre:
+                _, topk_inds = scores.topk(cfg.nms_pre)
+                rpn_bbox_pred = rpn_bbox_pred[topk_inds, :]
+                anchors = anchors[topk_inds, :]
+                scores = scores[topk_inds]
+            # get proposals w.r.t. anchors and rpn_bbox_pred
+            proposals = self.bbox_coder.decode(
+                anchors, rpn_bbox_pred, max_shape=img_shape)
+            # filter out too small bboxes
+            if cfg.min_bbox_size > 0:
+                w = proposals[:, 2] - proposals[:, 0]
+                h = proposals[:, 3] - proposals[:, 1]
+                valid_inds = torch.nonzero(
+                    (w >= cfg.min_bbox_size) & (h >= cfg.min_bbox_size),
+                    as_tuple=False).squeeze()
+                proposals = proposals[valid_inds, :]
+                scores = scores[valid_inds]
+            # NMS in current level
+            proposals, _ = nms(proposals, scores, cfg.nms.iou_threshold)
+            proposals = proposals[:cfg.nms_post, :]
+            mlvl_proposals.append(proposals)
+        proposals = torch.cat(mlvl_proposals, 0)
+        if cfg.get('nms_across_levels', False):
+            # NMS across multi levels
+            proposals, _ = nms(proposals[:, :4], proposals[:, -1],
+                               cfg.nms.iou_threshold)
+            proposals = proposals[:cfg.max_per_img, :]
+        else:
+            scores = proposals[:, 4]
+            num = min(cfg.max_per_img, proposals.shape[0])
+            _, topk_inds = scores.topk(num)
+            proposals = proposals[topk_inds, :]
+        return proposals
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/gfl_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/gfl_head.py
new file mode 100644
index 0000000..961bc92
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/gfl_head.py
@@ -0,0 +1,647 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, Scale, bias_init_with_prob, normal_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import (anchor_inside_flags, bbox2distance, bbox_overlaps,
+                        build_assigner, build_sampler, distance2bbox,
+                        images_to_levels, multi_apply, multiclass_nms,
+                        reduce_mean, unmap)
+from ..builder import HEADS, build_loss
+from .anchor_head import AnchorHead
+
+
+class Integral(nn.Module):
+    """A fixed layer for calculating integral result from distribution.
+
+    This layer calculates the target location by :math: `sum{P(y_i) * y_i}`,
+    P(y_i) denotes the softmax vector that represents the discrete distribution
+    y_i denotes the discrete set, usually {0, 1, 2, ..., reg_max}
+
+    Args:
+        reg_max (int): The maximal value of the discrete set. Default: 16. You
+            may want to reset it according to your new dataset or related
+            settings.
+    """
+
+    def __init__(self, reg_max=16):
+        super(Integral, self).__init__()
+        self.reg_max = reg_max
+        self.register_buffer('project',
+                             torch.linspace(0, self.reg_max, self.reg_max + 1))
+
+    def forward(self, x):
+        """Forward feature from the regression head to get integral result of
+        bounding box location.
+
+        Args:
+            x (Tensor): Features of the regression head, shape (N, 4*(n+1)),
+                n is self.reg_max.
+
+        Returns:
+            x (Tensor): Integral result of box locations, i.e., distance
+                offsets from the box center in four directions, shape (N, 4).
+        """
+        x = F.softmax(x.reshape(-1, self.reg_max + 1), dim=1)
+        x = F.linear(x, self.project.type_as(x)).reshape(-1, 4)
+        return x
+
+
+@HEADS.register_module()
+class GFLHead(AnchorHead):
+    """Generalized Focal Loss: Learning Qualified and Distributed Bounding
+    Boxes for Dense Object Detection.
+
+    GFL head structure is similar with ATSS, however GFL uses
+    1) joint representation for classification and localization quality, and
+    2) flexible General distribution for bounding box locations,
+    which are supervised by
+    Quality Focal Loss (QFL) and Distribution Focal Loss (DFL), respectively
+
+    https://arxiv.org/abs/2006.04388
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        stacked_convs (int): Number of conv layers in cls and reg tower.
+            Default: 4.
+        conv_cfg (dict): dictionary to construct and config conv layer.
+            Default: None.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+            Default: dict(type='GN', num_groups=32, requires_grad=True).
+        loss_qfl (dict): Config of Quality Focal Loss (QFL).
+        reg_max (int): Max value of integral set :math: `{0, ..., reg_max}`
+            in QFL setting. Default: 16.
+    Example:
+        >>> self = GFLHead(11, 7)
+        >>> feats = [torch.rand(1, 7, s, s) for s in [4, 8, 16, 32, 64]]
+        >>> cls_quality_score, bbox_pred = self.forward(feats)
+        >>> assert len(cls_quality_score) == len(self.scales)
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 stacked_convs=4,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='GN', num_groups=32, requires_grad=True),
+                 loss_dfl=dict(type='DistributionFocalLoss', loss_weight=0.25),
+                 reg_max=16,
+                 **kwargs):
+        self.stacked_convs = stacked_convs
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.reg_max = reg_max
+        super(GFLHead, self).__init__(num_classes, in_channels, **kwargs)
+
+        self.sampling = False
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            # SSD sampling=False so use PseudoSampler
+            sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+
+        self.integral = Integral(self.reg_max)
+        self.loss_dfl = build_loss(loss_dfl)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.relu = nn.ReLU(inplace=True)
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            self.cls_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+            self.reg_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+        assert self.num_anchors == 1, 'anchor free version'
+        self.gfl_cls = nn.Conv2d(
+            self.feat_channels, self.cls_out_channels, 3, padding=1)
+        self.gfl_reg = nn.Conv2d(
+            self.feat_channels, 4 * (self.reg_max + 1), 3, padding=1)
+        self.scales = nn.ModuleList(
+            [Scale(1.0) for _ in self.anchor_generator.strides])
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.cls_convs:
+            normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            normal_init(m.conv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.gfl_cls, std=0.01, bias=bias_cls)
+        normal_init(self.gfl_reg, std=0.01)
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple: Usually a tuple of classification scores and bbox prediction
+                cls_scores (list[Tensor]): Classification and quality (IoU)
+                    joint scores for all scale levels, each is a 4D-tensor,
+                    the channel number is num_classes.
+                bbox_preds (list[Tensor]): Box distribution logits for all
+                    scale levels, each is a 4D-tensor, the channel number is
+                    4*(n+1), n is max value of integral set.
+        """
+        return multi_apply(self.forward_single, feats, self.scales)
+
+    def forward_single(self, x, scale):
+        """Forward feature of a single scale level.
+
+        Args:
+            x (Tensor): Features of a single scale level.
+            scale (:obj: `mmcv.cnn.Scale`): Learnable scale module to resize
+                the bbox prediction.
+
+        Returns:
+            tuple:
+                cls_score (Tensor): Cls and quality joint scores for a single
+                    scale level the channel number is num_classes.
+                bbox_pred (Tensor): Box distribution logits for a single scale
+                    level, the channel number is 4*(n+1), n is max value of
+                    integral set.
+        """
+        cls_feat = x
+        reg_feat = x
+        for cls_conv in self.cls_convs:
+            cls_feat = cls_conv(cls_feat)
+        for reg_conv in self.reg_convs:
+            reg_feat = reg_conv(reg_feat)
+        cls_score = self.gfl_cls(cls_feat)
+        bbox_pred = scale(self.gfl_reg(reg_feat)).float()
+        return cls_score, bbox_pred
+
+    def anchor_center(self, anchors):
+        """Get anchor centers from anchors.
+
+        Args:
+            anchors (Tensor): Anchor list with shape (N, 4), "xyxy" format.
+
+        Returns:
+            Tensor: Anchor centers with shape (N, 2), "xy" format.
+        """
+        anchors_cx = (anchors[..., 2] + anchors[..., 0]) / 2
+        anchors_cy = (anchors[..., 3] + anchors[..., 1]) / 2
+        return torch.stack([anchors_cx, anchors_cy], dim=-1)
+
+    def loss_single(self, anchors, cls_score, bbox_pred, labels, label_weights,
+                    bbox_targets, stride, num_total_samples):
+        """Compute loss of a single scale level.
+
+        Args:
+            anchors (Tensor): Box reference for each scale level with shape
+                (N, num_total_anchors, 4).
+            cls_score (Tensor): Cls and quality joint scores for each scale
+                level has shape (N, num_classes, H, W).
+            bbox_pred (Tensor): Box distribution logits for each scale
+                level with shape (N, 4*(n+1), H, W), n is max value of integral
+                set.
+            labels (Tensor): Labels of each anchors with shape
+                (N, num_total_anchors).
+            label_weights (Tensor): Label weights of each anchor with shape
+                (N, num_total_anchors)
+            bbox_targets (Tensor): BBox regression targets of each anchor wight
+                shape (N, num_total_anchors, 4).
+            stride (tuple): Stride in this scale level.
+            num_total_samples (int): Number of positive samples that is
+                reduced over all GPUs.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        assert stride[0] == stride[1], 'h stride is not equal to w stride!'
+        anchors = anchors.reshape(-1, 4)
+        cls_score = cls_score.permute(0, 2, 3,
+                                      1).reshape(-1, self.cls_out_channels)
+        bbox_pred = bbox_pred.permute(0, 2, 3,
+                                      1).reshape(-1, 4 * (self.reg_max + 1))
+        bbox_targets = bbox_targets.reshape(-1, 4)
+        labels = labels.reshape(-1)
+        label_weights = label_weights.reshape(-1)
+
+        # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+        bg_class_ind = self.num_classes
+        pos_inds = ((labels >= 0)
+                    & (labels < bg_class_ind)).nonzero().squeeze(1)
+        score = label_weights.new_zeros(labels.shape)
+
+        if len(pos_inds) > 0:
+            pos_bbox_targets = bbox_targets[pos_inds]
+            pos_bbox_pred = bbox_pred[pos_inds]
+            pos_anchors = anchors[pos_inds]
+            pos_anchor_centers = self.anchor_center(pos_anchors) / stride[0]
+
+            weight_targets = cls_score.detach().sigmoid()
+            weight_targets = weight_targets.max(dim=1)[0][pos_inds]
+            pos_bbox_pred_corners = self.integral(pos_bbox_pred)
+            pos_decode_bbox_pred = distance2bbox(pos_anchor_centers,
+                                                 pos_bbox_pred_corners)
+            pos_decode_bbox_targets = pos_bbox_targets / stride[0]
+            score[pos_inds] = bbox_overlaps(
+                pos_decode_bbox_pred.detach(),
+                pos_decode_bbox_targets,
+                is_aligned=True)
+            pred_corners = pos_bbox_pred.reshape(-1, self.reg_max + 1)
+            target_corners = bbox2distance(pos_anchor_centers,
+                                           pos_decode_bbox_targets,
+                                           self.reg_max).reshape(-1)
+
+            # regression loss
+            loss_bbox = self.loss_bbox(
+                pos_decode_bbox_pred,
+                pos_decode_bbox_targets,
+                weight=weight_targets,
+                avg_factor=1.0)
+
+            # dfl loss
+            loss_dfl = self.loss_dfl(
+                pred_corners,
+                target_corners,
+                weight=weight_targets[:, None].expand(-1, 4).reshape(-1),
+                avg_factor=4.0)
+        else:
+            loss_bbox = bbox_pred.sum() * 0
+            loss_dfl = bbox_pred.sum() * 0
+            weight_targets = bbox_pred.new_tensor(0)
+
+        # cls (qfl) loss
+        loss_cls = self.loss_cls(
+            cls_score, (labels, score),
+            weight=label_weights,
+            avg_factor=num_total_samples)
+
+        return loss_cls, loss_bbox, loss_dfl, weight_targets.sum()
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Cls and quality scores for each scale
+                level has shape (N, num_classes, H, W).
+            bbox_preds (list[Tensor]): Box distribution logits for each scale
+                level with shape (N, 4*(n+1), H, W), n is max value of integral
+                set.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (list[Tensor] | None): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels)
+        if cls_reg_targets is None:
+            return None
+
+        (anchor_list, labels_list, label_weights_list, bbox_targets_list,
+         bbox_weights_list, num_total_pos, num_total_neg) = cls_reg_targets
+
+        num_total_samples = reduce_mean(
+            torch.tensor(num_total_pos, dtype=torch.float,
+                         device=device)).item()
+        num_total_samples = max(num_total_samples, 1.0)
+
+        losses_cls, losses_bbox, losses_dfl,\
+            avg_factor = multi_apply(
+                self.loss_single,
+                anchor_list,
+                cls_scores,
+                bbox_preds,
+                labels_list,
+                label_weights_list,
+                bbox_targets_list,
+                self.anchor_generator.strides,
+                num_total_samples=num_total_samples)
+
+        avg_factor = sum(avg_factor)
+        avg_factor = reduce_mean(avg_factor).item()
+        losses_bbox = list(map(lambda x: x / avg_factor, losses_bbox))
+        losses_dfl = list(map(lambda x: x / avg_factor, losses_dfl))
+        return dict(
+            loss_cls=losses_cls, loss_bbox=losses_bbox, loss_dfl=losses_dfl)
+
+    def _get_bboxes(self,
+                    cls_scores,
+                    bbox_preds,
+                    mlvl_anchors,
+                    img_shapes,
+                    scale_factors,
+                    cfg,
+                    rescale=False,
+                    with_nms=True):
+        """Transform outputs for a single batch item into labeled boxes.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for a single scale level
+                has shape (N, num_classes, H, W).
+            bbox_preds (list[Tensor]): Box distribution logits for a single
+                scale level with shape (N, 4*(n+1), H, W), n is max value of
+                integral set.
+            mlvl_anchors (list[Tensor]): Box reference for a single scale level
+                with shape (num_total_anchors, 4).
+            img_shapes (list[tuple[int]]): Shape of the input image,
+                list[(height, width, 3)].
+            scale_factors (list[ndarray]): Scale factor of the image arange as
+                (w_scale, h_scale, w_scale, h_scale).
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where 5 represent
+                (tl_x, tl_y, br_x, br_y, score) and the score between 0 and 1.
+                The shape of the second tensor in the tuple is (n,), and
+                each element represents the class label of the corresponding
+                box.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_scores) == len(bbox_preds) == len(mlvl_anchors)
+        batch_size = cls_scores[0].shape[0]
+
+        mlvl_bboxes = []
+        mlvl_scores = []
+        for cls_score, bbox_pred, stride, anchors in zip(
+                cls_scores, bbox_preds, self.anchor_generator.strides,
+                mlvl_anchors):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            assert stride[0] == stride[1]
+            scores = cls_score.permute(0, 2, 3, 1).reshape(
+                batch_size, -1, self.cls_out_channels).sigmoid()
+            bbox_pred = bbox_pred.permute(0, 2, 3, 1)
+
+            bbox_pred = self.integral(bbox_pred) * stride[0]
+            bbox_pred = bbox_pred.reshape(batch_size, -1, 4)
+
+            nms_pre = cfg.get('nms_pre', -1)
+            if nms_pre > 0 and scores.shape[1] > nms_pre:
+                max_scores, _ = scores.max(-1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                batch_inds = torch.arange(batch_size).view(
+                    -1, 1).expand_as(topk_inds).long()
+                anchors = anchors[topk_inds, :]
+                bbox_pred = bbox_pred[batch_inds, topk_inds, :]
+                scores = scores[batch_inds, topk_inds, :]
+            else:
+                anchors = anchors.expand_as(bbox_pred)
+
+            bboxes = distance2bbox(
+                self.anchor_center(anchors), bbox_pred, max_shape=img_shapes)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+
+        batch_mlvl_bboxes = torch.cat(mlvl_bboxes, dim=1)
+        if rescale:
+            batch_mlvl_bboxes /= batch_mlvl_bboxes.new_tensor(
+                scale_factors).unsqueeze(1)
+
+        batch_mlvl_scores = torch.cat(mlvl_scores, dim=1)
+        # Add a dummy background class to the backend when using sigmoid
+        # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+        # BG cat_id: num_class
+        padding = batch_mlvl_scores.new_zeros(batch_size,
+                                              batch_mlvl_scores.shape[1], 1)
+        batch_mlvl_scores = torch.cat([batch_mlvl_scores, padding], dim=-1)
+
+        if with_nms:
+            det_results = []
+            for (mlvl_bboxes, mlvl_scores) in zip(batch_mlvl_bboxes,
+                                                  batch_mlvl_scores):
+                det_bbox, det_label = multiclass_nms(mlvl_bboxes, mlvl_scores,
+                                                     cfg.score_thr, cfg.nms,
+                                                     cfg.max_per_img)
+                det_results.append(tuple([det_bbox, det_label]))
+        else:
+            det_results = [
+                tuple(mlvl_bs)
+                for mlvl_bs in zip(batch_mlvl_bboxes, batch_mlvl_scores)
+            ]
+        return det_results
+
+    def get_targets(self,
+                    anchor_list,
+                    valid_flag_list,
+                    gt_bboxes_list,
+                    img_metas,
+                    gt_bboxes_ignore_list=None,
+                    gt_labels_list=None,
+                    label_channels=1,
+                    unmap_outputs=True):
+        """Get targets for GFL head.
+
+        This method is almost the same as `AnchorHead.get_targets()`. Besides
+        returning the targets as the parent method does, it also returns the
+        anchors as the first element of the returned tuple.
+        """
+        num_imgs = len(img_metas)
+        assert len(anchor_list) == len(valid_flag_list) == num_imgs
+
+        # anchor number of multi levels
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+        num_level_anchors_list = [num_level_anchors] * num_imgs
+
+        # concat all level anchors and flags to a single tensor
+        for i in range(num_imgs):
+            assert len(anchor_list[i]) == len(valid_flag_list[i])
+            anchor_list[i] = torch.cat(anchor_list[i])
+            valid_flag_list[i] = torch.cat(valid_flag_list[i])
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        (all_anchors, all_labels, all_label_weights, all_bbox_targets,
+         all_bbox_weights, pos_inds_list, neg_inds_list) = multi_apply(
+             self._get_target_single,
+             anchor_list,
+             valid_flag_list,
+             num_level_anchors_list,
+             gt_bboxes_list,
+             gt_bboxes_ignore_list,
+             gt_labels_list,
+             img_metas,
+             label_channels=label_channels,
+             unmap_outputs=unmap_outputs)
+        # no valid anchors
+        if any([labels is None for labels in all_labels]):
+            return None
+        # sampled anchors of all images
+        num_total_pos = sum([max(inds.numel(), 1) for inds in pos_inds_list])
+        num_total_neg = sum([max(inds.numel(), 1) for inds in neg_inds_list])
+        # split targets to a list w.r.t. multiple levels
+        anchors_list = images_to_levels(all_anchors, num_level_anchors)
+        labels_list = images_to_levels(all_labels, num_level_anchors)
+        label_weights_list = images_to_levels(all_label_weights,
+                                              num_level_anchors)
+        bbox_targets_list = images_to_levels(all_bbox_targets,
+                                             num_level_anchors)
+        bbox_weights_list = images_to_levels(all_bbox_weights,
+                                             num_level_anchors)
+        return (anchors_list, labels_list, label_weights_list,
+                bbox_targets_list, bbox_weights_list, num_total_pos,
+                num_total_neg)
+
+    def _get_target_single(self,
+                           flat_anchors,
+                           valid_flags,
+                           num_level_anchors,
+                           gt_bboxes,
+                           gt_bboxes_ignore,
+                           gt_labels,
+                           img_meta,
+                           label_channels=1,
+                           unmap_outputs=True):
+        """Compute regression, classification targets for anchors in a single
+        image.
+
+        Args:
+            flat_anchors (Tensor): Multi-level anchors of the image, which are
+                concatenated into a single tensor of shape (num_anchors, 4)
+            valid_flags (Tensor): Multi level valid flags of the image,
+                which are concatenated into a single tensor of
+                    shape (num_anchors,).
+            num_level_anchors Tensor): Number of anchors of each scale level.
+            gt_bboxes (Tensor): Ground truth bboxes of the image,
+                shape (num_gts, 4).
+            gt_bboxes_ignore (Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+            gt_labels (Tensor): Ground truth labels of each box,
+                shape (num_gts,).
+            img_meta (dict): Meta info of the image.
+            label_channels (int): Channel of label.
+            unmap_outputs (bool): Whether to map outputs back to the original
+                set of anchors.
+
+        Returns:
+            tuple: N is the number of total anchors in the image.
+                anchors (Tensor): All anchors in the image with shape (N, 4).
+                labels (Tensor): Labels of all anchors in the image with shape
+                    (N,).
+                label_weights (Tensor): Label weights of all anchor in the
+                    image with shape (N,).
+                bbox_targets (Tensor): BBox targets of all anchors in the
+                    image with shape (N, 4).
+                bbox_weights (Tensor): BBox weights of all anchors in the
+                    image with shape (N, 4).
+                pos_inds (Tensor): Indices of positive anchor with shape
+                    (num_pos,).
+                neg_inds (Tensor): Indices of negative anchor with shape
+                    (num_neg,).
+        """
+        inside_flags = anchor_inside_flags(flat_anchors, valid_flags,
+                                           img_meta['img_shape'][:2],
+                                           self.train_cfg.allowed_border)
+        if not inside_flags.any():
+            return (None, ) * 7
+        # assign gt and sample anchors
+        anchors = flat_anchors[inside_flags, :]
+
+        num_level_anchors_inside = self.get_num_level_anchors_inside(
+            num_level_anchors, inside_flags)
+        assign_result = self.assigner.assign(anchors, num_level_anchors_inside,
+                                             gt_bboxes, gt_bboxes_ignore,
+                                             gt_labels)
+
+        sampling_result = self.sampler.sample(assign_result, anchors,
+                                              gt_bboxes)
+
+        num_valid_anchors = anchors.shape[0]
+        bbox_targets = torch.zeros_like(anchors)
+        bbox_weights = torch.zeros_like(anchors)
+        labels = anchors.new_full((num_valid_anchors, ),
+                                  self.num_classes,
+                                  dtype=torch.long)
+        label_weights = anchors.new_zeros(num_valid_anchors, dtype=torch.float)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+        if len(pos_inds) > 0:
+            pos_bbox_targets = sampling_result.pos_gt_bboxes
+            bbox_targets[pos_inds, :] = pos_bbox_targets
+            bbox_weights[pos_inds, :] = 1.0
+            if gt_labels is None:
+                # Only rpn gives gt_labels as None
+                # Foreground is the first class
+                labels[pos_inds] = 0
+            else:
+                labels[pos_inds] = gt_labels[
+                    sampling_result.pos_assigned_gt_inds]
+            if self.train_cfg.pos_weight <= 0:
+                label_weights[pos_inds] = 1.0
+            else:
+                label_weights[pos_inds] = self.train_cfg.pos_weight
+        if len(neg_inds) > 0:
+            label_weights[neg_inds] = 1.0
+
+        # map up to original set of anchors
+        if unmap_outputs:
+            num_total_anchors = flat_anchors.size(0)
+            anchors = unmap(anchors, num_total_anchors, inside_flags)
+            labels = unmap(
+                labels, num_total_anchors, inside_flags, fill=self.num_classes)
+            label_weights = unmap(label_weights, num_total_anchors,
+                                  inside_flags)
+            bbox_targets = unmap(bbox_targets, num_total_anchors, inside_flags)
+            bbox_weights = unmap(bbox_weights, num_total_anchors, inside_flags)
+
+        return (anchors, labels, label_weights, bbox_targets, bbox_weights,
+                pos_inds, neg_inds)
+
+    def get_num_level_anchors_inside(self, num_level_anchors, inside_flags):
+        split_inside_flags = torch.split(inside_flags, num_level_anchors)
+        num_level_anchors_inside = [
+            int(flags.sum()) for flags in split_inside_flags
+        ]
+        return num_level_anchors_inside
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/guided_anchor_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/guided_anchor_head.py
new file mode 100644
index 0000000..997ebb7
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/guided_anchor_head.py
@@ -0,0 +1,860 @@
+import torch
+import torch.nn as nn
+from mmcv.cnn import bias_init_with_prob, normal_init
+from mmcv.ops import DeformConv2d, MaskedConv2d
+from mmcv.runner import force_fp32
+
+from mmdet.core import (anchor_inside_flags, build_anchor_generator,
+                        build_assigner, build_bbox_coder, build_sampler,
+                        calc_region, images_to_levels, multi_apply,
+                        multiclass_nms, unmap)
+from ..builder import HEADS, build_loss
+from .anchor_head import AnchorHead
+
+
+class FeatureAdaption(nn.Module):
+    """Feature Adaption Module.
+
+    Feature Adaption Module is implemented based on DCN v1.
+    It uses anchor shape prediction rather than feature map to
+    predict offsets of deform conv layer.
+
+    Args:
+        in_channels (int): Number of channels in the input feature map.
+        out_channels (int): Number of channels in the output feature map.
+        kernel_size (int): Deformable conv kernel size.
+        deform_groups (int): Deformable conv group size.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 kernel_size=3,
+                 deform_groups=4):
+        super(FeatureAdaption, self).__init__()
+        offset_channels = kernel_size * kernel_size * 2
+        self.conv_offset = nn.Conv2d(
+            2, deform_groups * offset_channels, 1, bias=False)
+        self.conv_adaption = DeformConv2d(
+            in_channels,
+            out_channels,
+            kernel_size=kernel_size,
+            padding=(kernel_size - 1) // 2,
+            deform_groups=deform_groups)
+        self.relu = nn.ReLU(inplace=True)
+
+    def init_weights(self):
+        normal_init(self.conv_offset, std=0.1)
+        normal_init(self.conv_adaption, std=0.01)
+
+    def forward(self, x, shape):
+        offset = self.conv_offset(shape.detach())
+        x = self.relu(self.conv_adaption(x, offset))
+        return x
+
+
+@HEADS.register_module()
+class GuidedAnchorHead(AnchorHead):
+    """Guided-Anchor-based head (GA-RPN, GA-RetinaNet, etc.).
+
+    This GuidedAnchorHead will predict high-quality feature guided
+    anchors and locations where anchors will be kept in inference.
+    There are mainly 3 categories of bounding-boxes.
+
+    - Sampled 9 pairs for target assignment. (approxes)
+    - The square boxes where the predicted anchors are based on. (squares)
+    - Guided anchors.
+
+    Please refer to https://arxiv.org/abs/1901.03278 for more details.
+
+    Args:
+        num_classes (int): Number of classes.
+        in_channels (int): Number of channels in the input feature map.
+        feat_channels (int): Number of hidden channels.
+        approx_anchor_generator (dict): Config dict for approx generator
+        square_anchor_generator (dict): Config dict for square generator
+        anchor_coder (dict): Config dict for anchor coder
+        bbox_coder (dict): Config dict for bbox coder
+        reg_decoded_bbox (bool): If true, the regression loss would be
+            applied directly on decoded bounding boxes, converting both
+            the predicted boxes and regression targets to absolute
+            coordinates format. Default False. It should be `True` when
+            using `IoULoss`, `GIoULoss`, or `DIoULoss` in the bbox head.
+        deform_groups: (int): Group number of DCN in
+            FeatureAdaption module.
+        loc_filter_thr (float): Threshold to filter out unconcerned regions.
+        loss_loc (dict): Config of location loss.
+        loss_shape (dict): Config of anchor shape loss.
+        loss_cls (dict): Config of classification loss.
+        loss_bbox (dict): Config of bbox regression loss.
+    """
+
+    def __init__(
+        self,
+        num_classes,
+        in_channels,
+        feat_channels=256,
+        approx_anchor_generator=dict(
+            type='AnchorGenerator',
+            octave_base_scale=8,
+            scales_per_octave=3,
+            ratios=[0.5, 1.0, 2.0],
+            strides=[4, 8, 16, 32, 64]),
+        square_anchor_generator=dict(
+            type='AnchorGenerator',
+            ratios=[1.0],
+            scales=[8],
+            strides=[4, 8, 16, 32, 64]),
+        anchor_coder=dict(
+            type='DeltaXYWHBBoxCoder',
+            target_means=[.0, .0, .0, .0],
+            target_stds=[1.0, 1.0, 1.0, 1.0]
+        ),
+        bbox_coder=dict(
+            type='DeltaXYWHBBoxCoder',
+            target_means=[.0, .0, .0, .0],
+            target_stds=[1.0, 1.0, 1.0, 1.0]
+        ),
+        reg_decoded_bbox=False,
+        deform_groups=4,
+        loc_filter_thr=0.01,
+        train_cfg=None,
+        test_cfg=None,
+        loss_loc=dict(
+            type='FocalLoss',
+            use_sigmoid=True,
+            gamma=2.0,
+            alpha=0.25,
+            loss_weight=1.0),
+        loss_shape=dict(type='BoundedIoULoss', beta=0.2, loss_weight=1.0),
+        loss_cls=dict(
+            type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0),
+        loss_bbox=dict(type='SmoothL1Loss', beta=1.0,
+                       loss_weight=1.0)):  # yapf: disable
+        super(AnchorHead, self).__init__()
+        self.in_channels = in_channels
+        self.num_classes = num_classes
+        self.feat_channels = feat_channels
+        self.deform_groups = deform_groups
+        self.loc_filter_thr = loc_filter_thr
+
+        # build approx_anchor_generator and square_anchor_generator
+        assert (approx_anchor_generator['octave_base_scale'] ==
+                square_anchor_generator['scales'][0])
+        assert (approx_anchor_generator['strides'] ==
+                square_anchor_generator['strides'])
+        self.approx_anchor_generator = build_anchor_generator(
+            approx_anchor_generator)
+        self.square_anchor_generator = build_anchor_generator(
+            square_anchor_generator)
+        self.approxs_per_octave = self.approx_anchor_generator \
+            .num_base_anchors[0]
+
+        self.reg_decoded_bbox = reg_decoded_bbox
+
+        # one anchor per location
+        self.num_anchors = 1
+        self.use_sigmoid_cls = loss_cls.get('use_sigmoid', False)
+        self.loc_focal_loss = loss_loc['type'] in ['FocalLoss']
+        self.sampling = loss_cls['type'] not in ['FocalLoss']
+        self.ga_sampling = train_cfg is not None and hasattr(
+            train_cfg, 'ga_sampler')
+        if self.use_sigmoid_cls:
+            self.cls_out_channels = self.num_classes
+        else:
+            self.cls_out_channels = self.num_classes + 1
+
+        # build bbox_coder
+        self.anchor_coder = build_bbox_coder(anchor_coder)
+        self.bbox_coder = build_bbox_coder(bbox_coder)
+
+        # build losses
+        self.loss_loc = build_loss(loss_loc)
+        self.loss_shape = build_loss(loss_shape)
+        self.loss_cls = build_loss(loss_cls)
+        self.loss_bbox = build_loss(loss_bbox)
+
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            # use PseudoSampler when sampling is False
+            if self.sampling and hasattr(self.train_cfg, 'sampler'):
+                sampler_cfg = self.train_cfg.sampler
+            else:
+                sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+
+            self.ga_assigner = build_assigner(self.train_cfg.ga_assigner)
+            if self.ga_sampling:
+                ga_sampler_cfg = self.train_cfg.ga_sampler
+            else:
+                ga_sampler_cfg = dict(type='PseudoSampler')
+            self.ga_sampler = build_sampler(ga_sampler_cfg, context=self)
+
+        self.fp16_enabled = False
+
+        self._init_layers()
+
+    def _init_layers(self):
+        self.relu = nn.ReLU(inplace=True)
+        self.conv_loc = nn.Conv2d(self.in_channels, 1, 1)
+        self.conv_shape = nn.Conv2d(self.in_channels, self.num_anchors * 2, 1)
+        self.feature_adaption = FeatureAdaption(
+            self.in_channels,
+            self.feat_channels,
+            kernel_size=3,
+            deform_groups=self.deform_groups)
+        self.conv_cls = MaskedConv2d(self.feat_channels,
+                                     self.num_anchors * self.cls_out_channels,
+                                     1)
+        self.conv_reg = MaskedConv2d(self.feat_channels, self.num_anchors * 4,
+                                     1)
+
+    def init_weights(self):
+        normal_init(self.conv_cls, std=0.01)
+        normal_init(self.conv_reg, std=0.01)
+
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.conv_loc, std=0.01, bias=bias_cls)
+        normal_init(self.conv_shape, std=0.01)
+
+        self.feature_adaption.init_weights()
+
+    def forward_single(self, x):
+        loc_pred = self.conv_loc(x)
+        shape_pred = self.conv_shape(x)
+        x = self.feature_adaption(x, shape_pred)
+        # masked conv is only used during inference for speed-up
+        if not self.training:
+            mask = loc_pred.sigmoid()[0] >= self.loc_filter_thr
+        else:
+            mask = None
+        cls_score = self.conv_cls(x, mask)
+        bbox_pred = self.conv_reg(x, mask)
+        return cls_score, bbox_pred, shape_pred, loc_pred
+
+    def forward(self, feats):
+        return multi_apply(self.forward_single, feats)
+
+    def get_sampled_approxs(self, featmap_sizes, img_metas, device='cuda'):
+        """Get sampled approxs and inside flags according to feature map sizes.
+
+        Args:
+            featmap_sizes (list[tuple]): Multi-level feature map sizes.
+            img_metas (list[dict]): Image meta info.
+            device (torch.device | str): device for returned tensors
+
+        Returns:
+            tuple: approxes of each image, inside flags of each image
+        """
+        num_imgs = len(img_metas)
+
+        # since feature map sizes of all images are the same, we only compute
+        # approxes for one time
+        multi_level_approxs = self.approx_anchor_generator.grid_anchors(
+            featmap_sizes, device=device)
+        approxs_list = [multi_level_approxs for _ in range(num_imgs)]
+
+        # for each image, we compute inside flags of multi level approxes
+        inside_flag_list = []
+        for img_id, img_meta in enumerate(img_metas):
+            multi_level_flags = []
+            multi_level_approxs = approxs_list[img_id]
+
+            # obtain valid flags for each approx first
+            multi_level_approx_flags = self.approx_anchor_generator \
+                .valid_flags(featmap_sizes,
+                             img_meta['pad_shape'],
+                             device=device)
+
+            for i, flags in enumerate(multi_level_approx_flags):
+                approxs = multi_level_approxs[i]
+                inside_flags_list = []
+                for i in range(self.approxs_per_octave):
+                    split_valid_flags = flags[i::self.approxs_per_octave]
+                    split_approxs = approxs[i::self.approxs_per_octave, :]
+                    inside_flags = anchor_inside_flags(
+                        split_approxs, split_valid_flags,
+                        img_meta['img_shape'][:2],
+                        self.train_cfg.allowed_border)
+                    inside_flags_list.append(inside_flags)
+                # inside_flag for a position is true if any anchor in this
+                # position is true
+                inside_flags = (
+                    torch.stack(inside_flags_list, 0).sum(dim=0) > 0)
+                multi_level_flags.append(inside_flags)
+            inside_flag_list.append(multi_level_flags)
+        return approxs_list, inside_flag_list
+
+    def get_anchors(self,
+                    featmap_sizes,
+                    shape_preds,
+                    loc_preds,
+                    img_metas,
+                    use_loc_filter=False,
+                    device='cuda'):
+        """Get squares according to feature map sizes and guided anchors.
+
+        Args:
+            featmap_sizes (list[tuple]): Multi-level feature map sizes.
+            shape_preds (list[tensor]): Multi-level shape predictions.
+            loc_preds (list[tensor]): Multi-level location predictions.
+            img_metas (list[dict]): Image meta info.
+            use_loc_filter (bool): Use loc filter or not.
+            device (torch.device | str): device for returned tensors
+
+        Returns:
+            tuple: square approxs of each image, guided anchors of each image,
+                loc masks of each image
+        """
+        num_imgs = len(img_metas)
+        num_levels = len(featmap_sizes)
+
+        # since feature map sizes of all images are the same, we only compute
+        # squares for one time
+        multi_level_squares = self.square_anchor_generator.grid_anchors(
+            featmap_sizes, device=device)
+        squares_list = [multi_level_squares for _ in range(num_imgs)]
+
+        # for each image, we compute multi level guided anchors
+        guided_anchors_list = []
+        loc_mask_list = []
+        for img_id, img_meta in enumerate(img_metas):
+            multi_level_guided_anchors = []
+            multi_level_loc_mask = []
+            for i in range(num_levels):
+                squares = squares_list[img_id][i]
+                shape_pred = shape_preds[i][img_id]
+                loc_pred = loc_preds[i][img_id]
+                guided_anchors, loc_mask = self._get_guided_anchors_single(
+                    squares,
+                    shape_pred,
+                    loc_pred,
+                    use_loc_filter=use_loc_filter)
+                multi_level_guided_anchors.append(guided_anchors)
+                multi_level_loc_mask.append(loc_mask)
+            guided_anchors_list.append(multi_level_guided_anchors)
+            loc_mask_list.append(multi_level_loc_mask)
+        return squares_list, guided_anchors_list, loc_mask_list
+
+    def _get_guided_anchors_single(self,
+                                   squares,
+                                   shape_pred,
+                                   loc_pred,
+                                   use_loc_filter=False):
+        """Get guided anchors and loc masks for a single level.
+
+        Args:
+            square (tensor): Squares of a single level.
+            shape_pred (tensor): Shape predections of a single level.
+            loc_pred (tensor): Loc predections of a single level.
+            use_loc_filter (list[tensor]): Use loc filter or not.
+
+        Returns:
+            tuple: guided anchors, location masks
+        """
+        # calculate location filtering mask
+        loc_pred = loc_pred.sigmoid().detach()
+        if use_loc_filter:
+            loc_mask = loc_pred >= self.loc_filter_thr
+        else:
+            loc_mask = loc_pred >= 0.0
+        mask = loc_mask.permute(1, 2, 0).expand(-1, -1, self.num_anchors)
+        mask = mask.contiguous().view(-1)
+        # calculate guided anchors
+        squares = squares[mask]
+        anchor_deltas = shape_pred.permute(1, 2, 0).contiguous().view(
+            -1, 2).detach()[mask]
+        bbox_deltas = anchor_deltas.new_full(squares.size(), 0)
+        bbox_deltas[:, 2:] = anchor_deltas
+        guided_anchors = self.anchor_coder.decode(
+            squares, bbox_deltas, wh_ratio_clip=1e-6)
+        return guided_anchors, mask
+
+    def ga_loc_targets(self, gt_bboxes_list, featmap_sizes):
+        """Compute location targets for guided anchoring.
+
+        Each feature map is divided into positive, negative and ignore regions.
+        - positive regions: target 1, weight 1
+        - ignore regions: target 0, weight 0
+        - negative regions: target 0, weight 0.1
+
+        Args:
+            gt_bboxes_list (list[Tensor]): Gt bboxes of each image.
+            featmap_sizes (list[tuple]): Multi level sizes of each feature
+                maps.
+
+        Returns:
+            tuple
+        """
+        anchor_scale = self.approx_anchor_generator.octave_base_scale
+        anchor_strides = self.approx_anchor_generator.strides
+        # Currently only supports same stride in x and y direction.
+        for stride in anchor_strides:
+            assert (stride[0] == stride[1])
+        anchor_strides = [stride[0] for stride in anchor_strides]
+
+        center_ratio = self.train_cfg.center_ratio
+        ignore_ratio = self.train_cfg.ignore_ratio
+        img_per_gpu = len(gt_bboxes_list)
+        num_lvls = len(featmap_sizes)
+        r1 = (1 - center_ratio) / 2
+        r2 = (1 - ignore_ratio) / 2
+        all_loc_targets = []
+        all_loc_weights = []
+        all_ignore_map = []
+        for lvl_id in range(num_lvls):
+            h, w = featmap_sizes[lvl_id]
+            loc_targets = torch.zeros(
+                img_per_gpu,
+                1,
+                h,
+                w,
+                device=gt_bboxes_list[0].device,
+                dtype=torch.float32)
+            loc_weights = torch.full_like(loc_targets, -1)
+            ignore_map = torch.zeros_like(loc_targets)
+            all_loc_targets.append(loc_targets)
+            all_loc_weights.append(loc_weights)
+            all_ignore_map.append(ignore_map)
+        for img_id in range(img_per_gpu):
+            gt_bboxes = gt_bboxes_list[img_id]
+            scale = torch.sqrt((gt_bboxes[:, 2] - gt_bboxes[:, 0]) *
+                               (gt_bboxes[:, 3] - gt_bboxes[:, 1]))
+            min_anchor_size = scale.new_full(
+                (1, ), float(anchor_scale * anchor_strides[0]))
+            # assign gt bboxes to different feature levels w.r.t. their scales
+            target_lvls = torch.floor(
+                torch.log2(scale) - torch.log2(min_anchor_size) + 0.5)
+            target_lvls = target_lvls.clamp(min=0, max=num_lvls - 1).long()
+            for gt_id in range(gt_bboxes.size(0)):
+                lvl = target_lvls[gt_id].item()
+                # rescaled to corresponding feature map
+                gt_ = gt_bboxes[gt_id, :4] / anchor_strides[lvl]
+                # calculate ignore regions
+                ignore_x1, ignore_y1, ignore_x2, ignore_y2 = calc_region(
+                    gt_, r2, featmap_sizes[lvl])
+                # calculate positive (center) regions
+                ctr_x1, ctr_y1, ctr_x2, ctr_y2 = calc_region(
+                    gt_, r1, featmap_sizes[lvl])
+                all_loc_targets[lvl][img_id, 0, ctr_y1:ctr_y2 + 1,
+                                     ctr_x1:ctr_x2 + 1] = 1
+                all_loc_weights[lvl][img_id, 0, ignore_y1:ignore_y2 + 1,
+                                     ignore_x1:ignore_x2 + 1] = 0
+                all_loc_weights[lvl][img_id, 0, ctr_y1:ctr_y2 + 1,
+                                     ctr_x1:ctr_x2 + 1] = 1
+                # calculate ignore map on nearby low level feature
+                if lvl > 0:
+                    d_lvl = lvl - 1
+                    # rescaled to corresponding feature map
+                    gt_ = gt_bboxes[gt_id, :4] / anchor_strides[d_lvl]
+                    ignore_x1, ignore_y1, ignore_x2, ignore_y2 = calc_region(
+                        gt_, r2, featmap_sizes[d_lvl])
+                    all_ignore_map[d_lvl][img_id, 0, ignore_y1:ignore_y2 + 1,
+                                          ignore_x1:ignore_x2 + 1] = 1
+                # calculate ignore map on nearby high level feature
+                if lvl < num_lvls - 1:
+                    u_lvl = lvl + 1
+                    # rescaled to corresponding feature map
+                    gt_ = gt_bboxes[gt_id, :4] / anchor_strides[u_lvl]
+                    ignore_x1, ignore_y1, ignore_x2, ignore_y2 = calc_region(
+                        gt_, r2, featmap_sizes[u_lvl])
+                    all_ignore_map[u_lvl][img_id, 0, ignore_y1:ignore_y2 + 1,
+                                          ignore_x1:ignore_x2 + 1] = 1
+        for lvl_id in range(num_lvls):
+            # ignore negative regions w.r.t. ignore map
+            all_loc_weights[lvl_id][(all_loc_weights[lvl_id] < 0)
+                                    & (all_ignore_map[lvl_id] > 0)] = 0
+            # set negative regions with weight 0.1
+            all_loc_weights[lvl_id][all_loc_weights[lvl_id] < 0] = 0.1
+        # loc average factor to balance loss
+        loc_avg_factor = sum(
+            [t.size(0) * t.size(-1) * t.size(-2)
+             for t in all_loc_targets]) / 200
+        return all_loc_targets, all_loc_weights, loc_avg_factor
+
+    def _ga_shape_target_single(self,
+                                flat_approxs,
+                                inside_flags,
+                                flat_squares,
+                                gt_bboxes,
+                                gt_bboxes_ignore,
+                                img_meta,
+                                unmap_outputs=True):
+        """Compute guided anchoring targets.
+
+        This function returns sampled anchors and gt bboxes directly
+        rather than calculates regression targets.
+
+        Args:
+            flat_approxs (Tensor): flat approxs of a single image,
+                shape (n, 4)
+            inside_flags (Tensor): inside flags of a single image,
+                shape (n, ).
+            flat_squares (Tensor): flat squares of a single image,
+                shape (approxs_per_octave * n, 4)
+            gt_bboxes (Tensor): Ground truth bboxes of a single image.
+            img_meta (dict): Meta info of a single image.
+            approxs_per_octave (int): number of approxs per octave
+            cfg (dict): RPN train configs.
+            unmap_outputs (bool): unmap outputs or not.
+
+        Returns:
+            tuple
+        """
+        if not inside_flags.any():
+            return (None, ) * 5
+        # assign gt and sample anchors
+        expand_inside_flags = inside_flags[:, None].expand(
+            -1, self.approxs_per_octave).reshape(-1)
+        approxs = flat_approxs[expand_inside_flags, :]
+        squares = flat_squares[inside_flags, :]
+
+        assign_result = self.ga_assigner.assign(approxs, squares,
+                                                self.approxs_per_octave,
+                                                gt_bboxes, gt_bboxes_ignore)
+        sampling_result = self.ga_sampler.sample(assign_result, squares,
+                                                 gt_bboxes)
+
+        bbox_anchors = torch.zeros_like(squares)
+        bbox_gts = torch.zeros_like(squares)
+        bbox_weights = torch.zeros_like(squares)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+        if len(pos_inds) > 0:
+            bbox_anchors[pos_inds, :] = sampling_result.pos_bboxes
+            bbox_gts[pos_inds, :] = sampling_result.pos_gt_bboxes
+            bbox_weights[pos_inds, :] = 1.0
+
+        # map up to original set of anchors
+        if unmap_outputs:
+            num_total_anchors = flat_squares.size(0)
+            bbox_anchors = unmap(bbox_anchors, num_total_anchors, inside_flags)
+            bbox_gts = unmap(bbox_gts, num_total_anchors, inside_flags)
+            bbox_weights = unmap(bbox_weights, num_total_anchors, inside_flags)
+
+        return (bbox_anchors, bbox_gts, bbox_weights, pos_inds, neg_inds)
+
+    def ga_shape_targets(self,
+                         approx_list,
+                         inside_flag_list,
+                         square_list,
+                         gt_bboxes_list,
+                         img_metas,
+                         gt_bboxes_ignore_list=None,
+                         unmap_outputs=True):
+        """Compute guided anchoring targets.
+
+        Args:
+            approx_list (list[list]): Multi level approxs of each image.
+            inside_flag_list (list[list]): Multi level inside flags of each
+                image.
+            square_list (list[list]): Multi level squares of each image.
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image.
+            img_metas (list[dict]): Meta info of each image.
+            gt_bboxes_ignore_list (list[Tensor]): ignore list of gt bboxes.
+            unmap_outputs (bool): unmap outputs or not.
+
+        Returns:
+            tuple
+        """
+        num_imgs = len(img_metas)
+        assert len(approx_list) == len(inside_flag_list) == len(
+            square_list) == num_imgs
+        # anchor number of multi levels
+        num_level_squares = [squares.size(0) for squares in square_list[0]]
+        # concat all level anchors and flags to a single tensor
+        inside_flag_flat_list = []
+        approx_flat_list = []
+        square_flat_list = []
+        for i in range(num_imgs):
+            assert len(square_list[i]) == len(inside_flag_list[i])
+            inside_flag_flat_list.append(torch.cat(inside_flag_list[i]))
+            approx_flat_list.append(torch.cat(approx_list[i]))
+            square_flat_list.append(torch.cat(square_list[i]))
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        (all_bbox_anchors, all_bbox_gts, all_bbox_weights, pos_inds_list,
+         neg_inds_list) = multi_apply(
+             self._ga_shape_target_single,
+             approx_flat_list,
+             inside_flag_flat_list,
+             square_flat_list,
+             gt_bboxes_list,
+             gt_bboxes_ignore_list,
+             img_metas,
+             unmap_outputs=unmap_outputs)
+        # no valid anchors
+        if any([bbox_anchors is None for bbox_anchors in all_bbox_anchors]):
+            return None
+        # sampled anchors of all images
+        num_total_pos = sum([max(inds.numel(), 1) for inds in pos_inds_list])
+        num_total_neg = sum([max(inds.numel(), 1) for inds in neg_inds_list])
+        # split targets to a list w.r.t. multiple levels
+        bbox_anchors_list = images_to_levels(all_bbox_anchors,
+                                             num_level_squares)
+        bbox_gts_list = images_to_levels(all_bbox_gts, num_level_squares)
+        bbox_weights_list = images_to_levels(all_bbox_weights,
+                                             num_level_squares)
+        return (bbox_anchors_list, bbox_gts_list, bbox_weights_list,
+                num_total_pos, num_total_neg)
+
+    def loss_shape_single(self, shape_pred, bbox_anchors, bbox_gts,
+                          anchor_weights, anchor_total_num):
+        shape_pred = shape_pred.permute(0, 2, 3, 1).contiguous().view(-1, 2)
+        bbox_anchors = bbox_anchors.contiguous().view(-1, 4)
+        bbox_gts = bbox_gts.contiguous().view(-1, 4)
+        anchor_weights = anchor_weights.contiguous().view(-1, 4)
+        bbox_deltas = bbox_anchors.new_full(bbox_anchors.size(), 0)
+        bbox_deltas[:, 2:] += shape_pred
+        # filter out negative samples to speed-up weighted_bounded_iou_loss
+        inds = torch.nonzero(
+            anchor_weights[:, 0] > 0, as_tuple=False).squeeze(1)
+        bbox_deltas_ = bbox_deltas[inds]
+        bbox_anchors_ = bbox_anchors[inds]
+        bbox_gts_ = bbox_gts[inds]
+        anchor_weights_ = anchor_weights[inds]
+        pred_anchors_ = self.anchor_coder.decode(
+            bbox_anchors_, bbox_deltas_, wh_ratio_clip=1e-6)
+        loss_shape = self.loss_shape(
+            pred_anchors_,
+            bbox_gts_,
+            anchor_weights_,
+            avg_factor=anchor_total_num)
+        return loss_shape
+
+    def loss_loc_single(self, loc_pred, loc_target, loc_weight,
+                        loc_avg_factor):
+        loss_loc = self.loss_loc(
+            loc_pred.reshape(-1, 1),
+            loc_target.reshape(-1).long(),
+            loc_weight.reshape(-1),
+            avg_factor=loc_avg_factor)
+        return loss_loc
+
+    @force_fp32(
+        apply_to=('cls_scores', 'bbox_preds', 'shape_preds', 'loc_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             shape_preds,
+             loc_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.approx_anchor_generator.num_levels
+
+        device = cls_scores[0].device
+
+        # get loc targets
+        loc_targets, loc_weights, loc_avg_factor = self.ga_loc_targets(
+            gt_bboxes, featmap_sizes)
+
+        # get sampled approxes
+        approxs_list, inside_flag_list = self.get_sampled_approxs(
+            featmap_sizes, img_metas, device=device)
+        # get squares and guided anchors
+        squares_list, guided_anchors_list, _ = self.get_anchors(
+            featmap_sizes, shape_preds, loc_preds, img_metas, device=device)
+
+        # get shape targets
+        shape_targets = self.ga_shape_targets(approxs_list, inside_flag_list,
+                                              squares_list, gt_bboxes,
+                                              img_metas)
+        if shape_targets is None:
+            return None
+        (bbox_anchors_list, bbox_gts_list, anchor_weights_list, anchor_fg_num,
+         anchor_bg_num) = shape_targets
+        anchor_total_num = (
+            anchor_fg_num if not self.ga_sampling else anchor_fg_num +
+            anchor_bg_num)
+
+        # get anchor targets
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+        cls_reg_targets = self.get_targets(
+            guided_anchors_list,
+            inside_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg) = cls_reg_targets
+        num_total_samples = (
+            num_total_pos + num_total_neg if self.sampling else num_total_pos)
+
+        # anchor number of multi levels
+        num_level_anchors = [
+            anchors.size(0) for anchors in guided_anchors_list[0]
+        ]
+        # concat all level anchors to a single tensor
+        concat_anchor_list = []
+        for i in range(len(guided_anchors_list)):
+            concat_anchor_list.append(torch.cat(guided_anchors_list[i]))
+        all_anchor_list = images_to_levels(concat_anchor_list,
+                                           num_level_anchors)
+
+        # get classification and bbox regression losses
+        losses_cls, losses_bbox = multi_apply(
+            self.loss_single,
+            cls_scores,
+            bbox_preds,
+            all_anchor_list,
+            labels_list,
+            label_weights_list,
+            bbox_targets_list,
+            bbox_weights_list,
+            num_total_samples=num_total_samples)
+
+        # get anchor location loss
+        losses_loc = []
+        for i in range(len(loc_preds)):
+            loss_loc = self.loss_loc_single(
+                loc_preds[i],
+                loc_targets[i],
+                loc_weights[i],
+                loc_avg_factor=loc_avg_factor)
+            losses_loc.append(loss_loc)
+
+        # get anchor shape loss
+        losses_shape = []
+        for i in range(len(shape_preds)):
+            loss_shape = self.loss_shape_single(
+                shape_preds[i],
+                bbox_anchors_list[i],
+                bbox_gts_list[i],
+                anchor_weights_list[i],
+                anchor_total_num=anchor_total_num)
+            losses_shape.append(loss_shape)
+
+        return dict(
+            loss_cls=losses_cls,
+            loss_bbox=losses_bbox,
+            loss_shape=losses_shape,
+            loss_loc=losses_loc)
+
+    @force_fp32(
+        apply_to=('cls_scores', 'bbox_preds', 'shape_preds', 'loc_preds'))
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   shape_preds,
+                   loc_preds,
+                   img_metas,
+                   cfg=None,
+                   rescale=False):
+        assert len(cls_scores) == len(bbox_preds) == len(shape_preds) == len(
+            loc_preds)
+        num_levels = len(cls_scores)
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        device = cls_scores[0].device
+        # get guided anchors
+        _, guided_anchors, loc_masks = self.get_anchors(
+            featmap_sizes,
+            shape_preds,
+            loc_preds,
+            img_metas,
+            use_loc_filter=not self.training,
+            device=device)
+        result_list = []
+        for img_id in range(len(img_metas)):
+            cls_score_list = [
+                cls_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_pred_list = [
+                bbox_preds[i][img_id].detach() for i in range(num_levels)
+            ]
+            guided_anchor_list = [
+                guided_anchors[img_id][i].detach() for i in range(num_levels)
+            ]
+            loc_mask_list = [
+                loc_masks[img_id][i].detach() for i in range(num_levels)
+            ]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            proposals = self._get_bboxes_single(cls_score_list, bbox_pred_list,
+                                                guided_anchor_list,
+                                                loc_mask_list, img_shape,
+                                                scale_factor, cfg, rescale)
+            result_list.append(proposals)
+        return result_list
+
+    def _get_bboxes_single(self,
+                           cls_scores,
+                           bbox_preds,
+                           mlvl_anchors,
+                           mlvl_masks,
+                           img_shape,
+                           scale_factor,
+                           cfg,
+                           rescale=False):
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_scores) == len(bbox_preds) == len(mlvl_anchors)
+        mlvl_bboxes = []
+        mlvl_scores = []
+        for cls_score, bbox_pred, anchors, mask in zip(cls_scores, bbox_preds,
+                                                       mlvl_anchors,
+                                                       mlvl_masks):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            # if no location is kept, end.
+            if mask.sum() == 0:
+                continue
+            # reshape scores and bbox_pred
+            cls_score = cls_score.permute(1, 2,
+                                          0).reshape(-1, self.cls_out_channels)
+            if self.use_sigmoid_cls:
+                scores = cls_score.sigmoid()
+            else:
+                scores = cls_score.softmax(-1)
+            bbox_pred = bbox_pred.permute(1, 2, 0).reshape(-1, 4)
+            # filter scores, bbox_pred w.r.t. mask.
+            # anchors are filtered in get_anchors() beforehand.
+            scores = scores[mask, :]
+            bbox_pred = bbox_pred[mask, :]
+            if scores.dim() == 0:
+                anchors = anchors.unsqueeze(0)
+                scores = scores.unsqueeze(0)
+                bbox_pred = bbox_pred.unsqueeze(0)
+            # filter anchors, bbox_pred, scores w.r.t. scores
+            nms_pre = cfg.get('nms_pre', -1)
+            if nms_pre > 0 and scores.shape[0] > nms_pre:
+                if self.use_sigmoid_cls:
+                    max_scores, _ = scores.max(dim=1)
+                else:
+                    # remind that we set FG labels to [0, num_class-1]
+                    # since mmdet v2.0
+                    # BG cat_id: num_class
+                    max_scores, _ = scores[:, :-1].max(dim=1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                anchors = anchors[topk_inds, :]
+                bbox_pred = bbox_pred[topk_inds, :]
+                scores = scores[topk_inds, :]
+            bboxes = self.bbox_coder.decode(
+                anchors, bbox_pred, max_shape=img_shape)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+        mlvl_bboxes = torch.cat(mlvl_bboxes)
+        if rescale:
+            mlvl_bboxes /= mlvl_bboxes.new_tensor(scale_factor)
+        mlvl_scores = torch.cat(mlvl_scores)
+        if self.use_sigmoid_cls:
+            # Add a dummy background class to the backend when using sigmoid
+            # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+            # BG cat_id: num_class
+            padding = mlvl_scores.new_zeros(mlvl_scores.shape[0], 1)
+            mlvl_scores = torch.cat([mlvl_scores, padding], dim=1)
+        # multi class NMS
+        det_bboxes, det_labels = multiclass_nms(mlvl_bboxes, mlvl_scores,
+                                                cfg.score_thr, cfg.nms,
+                                                cfg.max_per_img)
+        return det_bboxes, det_labels
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/ld_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/ld_head.py
new file mode 100644
index 0000000..501e1f7
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/ld_head.py
@@ -0,0 +1,261 @@
+import torch
+from mmcv.runner import force_fp32
+
+from mmdet.core import (bbox2distance, bbox_overlaps, distance2bbox,
+                        multi_apply, reduce_mean)
+from ..builder import HEADS, build_loss
+from .gfl_head import GFLHead
+
+
+@HEADS.register_module()
+class LDHead(GFLHead):
+    """Localization distillation Head. (Short description)
+
+    It utilizes the learned bbox distributions to transfer the localization
+    dark knowledge from teacher to student. Original paper: `Localization
+    Distillation for Object Detection. <https://arxiv.org/abs/2102.12252>`_
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        loss_ld (dict): Config of Localization Distillation Loss (LD),
+            T is the temperature for distillation.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 loss_ld=dict(
+                     type='LocalizationDistillationLoss',
+                     loss_weight=0.25,
+                     T=10),
+                 **kwargs):
+
+        super(LDHead, self).__init__(num_classes, in_channels, **kwargs)
+        self.loss_ld = build_loss(loss_ld)
+
+    def loss_single(self, anchors, cls_score, bbox_pred, labels, label_weights,
+                    bbox_targets, stride, soft_targets, num_total_samples):
+        """Compute loss of a single scale level.
+
+        Args:
+            anchors (Tensor): Box reference for each scale level with shape
+                (N, num_total_anchors, 4).
+            cls_score (Tensor): Cls and quality joint scores for each scale
+                level has shape (N, num_classes, H, W).
+            bbox_pred (Tensor): Box distribution logits for each scale
+                level with shape (N, 4*(n+1), H, W), n is max value of integral
+                set.
+            labels (Tensor): Labels of each anchors with shape
+                (N, num_total_anchors).
+            label_weights (Tensor): Label weights of each anchor with shape
+                (N, num_total_anchors)
+            bbox_targets (Tensor): BBox regression targets of each anchor wight
+                shape (N, num_total_anchors, 4).
+            stride (tuple): Stride in this scale level.
+            num_total_samples (int): Number of positive samples that is
+                reduced over all GPUs.
+
+        Returns:
+            dict[tuple, Tensor]: Loss components and weight targets.
+        """
+        assert stride[0] == stride[1], 'h stride is not equal to w stride!'
+        anchors = anchors.reshape(-1, 4)
+        cls_score = cls_score.permute(0, 2, 3,
+                                      1).reshape(-1, self.cls_out_channels)
+        bbox_pred = bbox_pred.permute(0, 2, 3,
+                                      1).reshape(-1, 4 * (self.reg_max + 1))
+        soft_targets = soft_targets.permute(0, 2, 3,
+                                            1).reshape(-1,
+                                                       4 * (self.reg_max + 1))
+
+        bbox_targets = bbox_targets.reshape(-1, 4)
+        labels = labels.reshape(-1)
+        label_weights = label_weights.reshape(-1)
+
+        # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+        bg_class_ind = self.num_classes
+        pos_inds = ((labels >= 0)
+                    & (labels < bg_class_ind)).nonzero().squeeze(1)
+        score = label_weights.new_zeros(labels.shape)
+
+        if len(pos_inds) > 0:
+            pos_bbox_targets = bbox_targets[pos_inds]
+            pos_bbox_pred = bbox_pred[pos_inds]
+            pos_anchors = anchors[pos_inds]
+            pos_anchor_centers = self.anchor_center(pos_anchors) / stride[0]
+
+            weight_targets = cls_score.detach().sigmoid()
+            weight_targets = weight_targets.max(dim=1)[0][pos_inds]
+            pos_bbox_pred_corners = self.integral(pos_bbox_pred)
+            pos_decode_bbox_pred = distance2bbox(pos_anchor_centers,
+                                                 pos_bbox_pred_corners)
+            pos_decode_bbox_targets = pos_bbox_targets / stride[0]
+            score[pos_inds] = bbox_overlaps(
+                pos_decode_bbox_pred.detach(),
+                pos_decode_bbox_targets,
+                is_aligned=True)
+            pred_corners = pos_bbox_pred.reshape(-1, self.reg_max + 1)
+            pos_soft_targets = soft_targets[pos_inds]
+            soft_corners = pos_soft_targets.reshape(-1, self.reg_max + 1)
+
+            target_corners = bbox2distance(pos_anchor_centers,
+                                           pos_decode_bbox_targets,
+                                           self.reg_max).reshape(-1)
+
+            # regression loss
+            loss_bbox = self.loss_bbox(
+                pos_decode_bbox_pred,
+                pos_decode_bbox_targets,
+                weight=weight_targets,
+                avg_factor=1.0)
+
+            # dfl loss
+            loss_dfl = self.loss_dfl(
+                pred_corners,
+                target_corners,
+                weight=weight_targets[:, None].expand(-1, 4).reshape(-1),
+                avg_factor=4.0)
+
+            # ld loss
+            loss_ld = self.loss_ld(
+                pred_corners,
+                soft_corners,
+                weight=weight_targets[:, None].expand(-1, 4).reshape(-1),
+                avg_factor=4.0)
+
+        else:
+            loss_ld = bbox_pred.sum() * 0
+            loss_bbox = bbox_pred.sum() * 0
+            loss_dfl = bbox_pred.sum() * 0
+            weight_targets = bbox_pred.new_tensor(0)
+
+        # cls (qfl) loss
+        loss_cls = self.loss_cls(
+            cls_score, (labels, score),
+            weight=label_weights,
+            avg_factor=num_total_samples)
+
+        return loss_cls, loss_bbox, loss_dfl, loss_ld, weight_targets.sum()
+
+    def forward_train(self,
+                      x,
+                      out_teacher,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels=None,
+                      gt_bboxes_ignore=None,
+                      proposal_cfg=None,
+                      **kwargs):
+        """
+        Args:
+            x (list[Tensor]): Features from FPN.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes (Tensor): Ground truth bboxes of the image,
+                shape (num_gts, 4).
+            gt_labels (Tensor): Ground truth labels of each box,
+                shape (num_gts,).
+            gt_bboxes_ignore (Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+            proposal_cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used
+
+        Returns:
+            tuple[dict, list]: The loss components and proposals of each image.
+
+            - losses (dict[str, Tensor]): A dictionary of loss components.
+            - proposal_list (list[Tensor]): Proposals of each image.
+        """
+        outs = self(x)
+        soft_target = out_teacher[1]
+        if gt_labels is None:
+            loss_inputs = outs + (gt_bboxes, soft_target, img_metas)
+        else:
+            loss_inputs = outs + (gt_bboxes, gt_labels, soft_target, img_metas)
+        losses = self.loss(*loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore)
+        if proposal_cfg is None:
+            return losses
+        else:
+            proposal_list = self.get_bboxes(*outs, img_metas, cfg=proposal_cfg)
+            return losses, proposal_list
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             soft_target,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Cls and quality scores for each scale
+                level has shape (N, num_classes, H, W).
+            bbox_preds (list[Tensor]): Box distribution logits for each scale
+                level with shape (N, 4*(n+1), H, W), n is max value of integral
+                set.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (list[Tensor] | None): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels)
+        if cls_reg_targets is None:
+            return None
+
+        (anchor_list, labels_list, label_weights_list, bbox_targets_list,
+         bbox_weights_list, num_total_pos, num_total_neg) = cls_reg_targets
+
+        num_total_samples = reduce_mean(
+            torch.tensor(num_total_pos, dtype=torch.float,
+                         device=device)).item()
+        num_total_samples = max(num_total_samples, 1.0)
+
+        losses_cls, losses_bbox, losses_dfl, losses_ld, \
+            avg_factor = multi_apply(
+                self.loss_single,
+                anchor_list,
+                cls_scores,
+                bbox_preds,
+                labels_list,
+                label_weights_list,
+                bbox_targets_list,
+                self.anchor_generator.strides,
+                soft_target,
+                num_total_samples=num_total_samples)
+
+        avg_factor = sum(avg_factor) + 1e-6
+        avg_factor = reduce_mean(avg_factor).item()
+        losses_bbox = [x / avg_factor for x in losses_bbox]
+        losses_dfl = [x / avg_factor for x in losses_dfl]
+        return dict(
+            loss_cls=losses_cls,
+            loss_bbox=losses_bbox,
+            loss_dfl=losses_dfl,
+            loss_ld=losses_ld)
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/nasfcos_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/nasfcos_head.py
new file mode 100644
index 0000000..994ce04
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/nasfcos_head.py
@@ -0,0 +1,75 @@
+import copy
+
+import torch.nn as nn
+from mmcv.cnn import (ConvModule, Scale, bias_init_with_prob,
+                      caffe2_xavier_init, normal_init)
+
+from mmdet.models.dense_heads.fcos_head import FCOSHead
+from ..builder import HEADS
+
+
+@HEADS.register_module()
+class NASFCOSHead(FCOSHead):
+    """Anchor-free head used in `NASFCOS <https://arxiv.org/abs/1906.04423>`_.
+
+    It is quite similar with FCOS head, except for the searched structure of
+    classification branch and bbox regression branch, where a structure of
+    "dconv3x3, conv3x3, dconv3x3, conv1x1" is utilized instead.
+    """
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        dconv3x3_config = dict(
+            type='DCNv2',
+            kernel_size=3,
+            use_bias=True,
+            deform_groups=2,
+            padding=1)
+        conv3x3_config = dict(type='Conv', kernel_size=3, padding=1)
+        conv1x1_config = dict(type='Conv', kernel_size=1)
+
+        self.arch_config = [
+            dconv3x3_config, conv3x3_config, dconv3x3_config, conv1x1_config
+        ]
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        for i, op_ in enumerate(self.arch_config):
+            op = copy.deepcopy(op_)
+            chn = self.in_channels if i == 0 else self.feat_channels
+            assert isinstance(op, dict)
+            use_bias = op.pop('use_bias', False)
+            padding = op.pop('padding', 0)
+            kernel_size = op.pop('kernel_size')
+            module = ConvModule(
+                chn,
+                self.feat_channels,
+                kernel_size,
+                stride=1,
+                padding=padding,
+                norm_cfg=self.norm_cfg,
+                bias=use_bias,
+                conv_cfg=op)
+
+            self.cls_convs.append(copy.deepcopy(module))
+            self.reg_convs.append(copy.deepcopy(module))
+
+        self.conv_cls = nn.Conv2d(
+            self.feat_channels, self.cls_out_channels, 3, padding=1)
+        self.conv_reg = nn.Conv2d(self.feat_channels, 4, 3, padding=1)
+        self.conv_centerness = nn.Conv2d(self.feat_channels, 1, 3, padding=1)
+
+        self.scales = nn.ModuleList([Scale(1.0) for _ in self.strides])
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        # retinanet_bias_init
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.conv_reg, std=0.01)
+        normal_init(self.conv_centerness, std=0.01)
+        normal_init(self.conv_cls, std=0.01, bias=bias_cls)
+
+        for branch in [self.cls_convs, self.reg_convs]:
+            for module in branch.modules():
+                if isinstance(module, ConvModule) \
+                        and isinstance(module.conv, nn.Conv2d):
+                    caffe2_xavier_init(module.conv)
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/paa_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/paa_head.py
new file mode 100644
index 0000000..e067b01
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/paa_head.py
@@ -0,0 +1,671 @@
+import numpy as np
+import torch
+from mmcv.runner import force_fp32
+
+from mmdet.core import multi_apply, multiclass_nms
+from mmdet.core.bbox.iou_calculators import bbox_overlaps
+from mmdet.models import HEADS
+from mmdet.models.dense_heads import ATSSHead
+
+EPS = 1e-12
+try:
+    import sklearn.mixture as skm
+except ImportError:
+    skm = None
+
+
+def levels_to_images(mlvl_tensor):
+    """Concat multi-level feature maps by image.
+
+    [feature_level0, feature_level1...] -> [feature_image0, feature_image1...]
+    Convert the shape of each element in mlvl_tensor from (N, C, H, W) to
+    (N, H*W , C), then split the element to N elements with shape (H*W, C), and
+    concat elements in same image of all level along first dimension.
+
+    Args:
+        mlvl_tensor (list[torch.Tensor]): list of Tensor which collect from
+            corresponding level. Each element is of shape (N, C, H, W)
+
+    Returns:
+        list[torch.Tensor]: A list that contains N tensors and each tensor is
+            of shape (num_elements, C)
+    """
+    batch_size = mlvl_tensor[0].size(0)
+    batch_list = [[] for _ in range(batch_size)]
+    channels = mlvl_tensor[0].size(1)
+    for t in mlvl_tensor:
+        t = t.permute(0, 2, 3, 1)
+        t = t.view(batch_size, -1, channels).contiguous()
+        for img in range(batch_size):
+            batch_list[img].append(t[img])
+    return [torch.cat(item, 0) for item in batch_list]
+
+
+@HEADS.register_module()
+class PAAHead(ATSSHead):
+    """Head of PAAAssignment: Probabilistic Anchor Assignment with IoU
+    Prediction for Object Detection.
+
+    Code is modified from the `official github repo
+    <https://github.com/kkhoot/PAA/blob/master/paa_core
+    /modeling/rpn/paa/loss.py>`_.
+
+    More details can be found in the `paper
+    <https://arxiv.org/abs/2007.08103>`_ .
+
+    Args:
+        topk (int): Select topk samples with smallest loss in
+            each level.
+        score_voting (bool): Whether to use score voting in post-process.
+        covariance_type : String describing the type of covariance parameters
+            to be used in :class:`sklearn.mixture.GaussianMixture`.
+            It must be one of:
+
+            - 'full': each component has its own general covariance matrix
+            - 'tied': all components share the same general covariance matrix
+            - 'diag': each component has its own diagonal covariance matrix
+            - 'spherical': each component has its own single variance
+            Default: 'diag'. From 'full' to 'spherical', the gmm fitting
+            process is faster yet the performance could be influenced. For most
+            cases, 'diag' should be a good choice.
+    """
+
+    def __init__(self,
+                 *args,
+                 topk=9,
+                 score_voting=True,
+                 covariance_type='diag',
+                 **kwargs):
+        # topk used in paa reassign process
+        self.topk = topk
+        self.with_score_voting = score_voting
+        self.covariance_type = covariance_type
+        super(PAAHead, self).__init__(*args, **kwargs)
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds', 'iou_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             iou_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            iou_preds (list[Tensor]): iou_preds for each scale
+                level with shape (N, num_anchors * 1, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (list[Tensor] | None): Specify which bounding
+                boxes can be ignored when are computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss gmm_assignment.
+        """
+
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels,
+        )
+        (labels, labels_weight, bboxes_target, bboxes_weight, pos_inds,
+         pos_gt_index) = cls_reg_targets
+        cls_scores = levels_to_images(cls_scores)
+        cls_scores = [
+            item.reshape(-1, self.cls_out_channels) for item in cls_scores
+        ]
+        bbox_preds = levels_to_images(bbox_preds)
+        bbox_preds = [item.reshape(-1, 4) for item in bbox_preds]
+        iou_preds = levels_to_images(iou_preds)
+        iou_preds = [item.reshape(-1, 1) for item in iou_preds]
+        pos_losses_list, = multi_apply(self.get_pos_loss, anchor_list,
+                                       cls_scores, bbox_preds, labels,
+                                       labels_weight, bboxes_target,
+                                       bboxes_weight, pos_inds)
+
+        with torch.no_grad():
+            reassign_labels, reassign_label_weight, \
+                reassign_bbox_weights, num_pos = multi_apply(
+                    self.paa_reassign,
+                    pos_losses_list,
+                    labels,
+                    labels_weight,
+                    bboxes_weight,
+                    pos_inds,
+                    pos_gt_index,
+                    anchor_list)
+            num_pos = sum(num_pos)
+        # convert all tensor list to a flatten tensor
+        cls_scores = torch.cat(cls_scores, 0).view(-1, cls_scores[0].size(-1))
+        bbox_preds = torch.cat(bbox_preds, 0).view(-1, bbox_preds[0].size(-1))
+        iou_preds = torch.cat(iou_preds, 0).view(-1, iou_preds[0].size(-1))
+        labels = torch.cat(reassign_labels, 0).view(-1)
+        flatten_anchors = torch.cat(
+            [torch.cat(item, 0) for item in anchor_list])
+        labels_weight = torch.cat(reassign_label_weight, 0).view(-1)
+        bboxes_target = torch.cat(bboxes_target,
+                                  0).view(-1, bboxes_target[0].size(-1))
+
+        pos_inds_flatten = ((labels >= 0)
+                            &
+                            (labels < self.num_classes)).nonzero().reshape(-1)
+
+        losses_cls = self.loss_cls(
+            cls_scores,
+            labels,
+            labels_weight,
+            avg_factor=max(num_pos, len(img_metas)))  # avoid num_pos=0
+        if num_pos:
+            pos_bbox_pred = self.bbox_coder.decode(
+                flatten_anchors[pos_inds_flatten],
+                bbox_preds[pos_inds_flatten])
+            pos_bbox_target = bboxes_target[pos_inds_flatten]
+            iou_target = bbox_overlaps(
+                pos_bbox_pred.detach(), pos_bbox_target, is_aligned=True)
+            losses_iou = self.loss_centerness(
+                iou_preds[pos_inds_flatten],
+                iou_target.unsqueeze(-1),
+                avg_factor=num_pos)
+            losses_bbox = self.loss_bbox(
+                pos_bbox_pred,
+                pos_bbox_target,
+                iou_target.clamp(min=EPS),
+                avg_factor=iou_target.sum())
+        else:
+            losses_iou = iou_preds.sum() * 0
+            losses_bbox = bbox_preds.sum() * 0
+
+        return dict(
+            loss_cls=losses_cls, loss_bbox=losses_bbox, loss_iou=losses_iou)
+
+    def get_pos_loss(self, anchors, cls_score, bbox_pred, label, label_weight,
+                     bbox_target, bbox_weight, pos_inds):
+        """Calculate loss of all potential positive samples obtained from first
+        match process.
+
+        Args:
+            anchors (list[Tensor]): Anchors of each scale.
+            cls_score (Tensor): Box scores of single image with shape
+                (num_anchors, num_classes)
+            bbox_pred (Tensor): Box energies / deltas of single image
+                with shape (num_anchors, 4)
+            label (Tensor): classification target of each anchor with
+                shape (num_anchors,)
+            label_weight (Tensor): Classification loss weight of each
+                anchor with shape (num_anchors).
+            bbox_target (dict): Regression target of each anchor with
+                shape (num_anchors, 4).
+            bbox_weight (Tensor): Bbox weight of each anchor with shape
+                (num_anchors, 4).
+            pos_inds (Tensor): Index of all positive samples got from
+                first assign process.
+
+        Returns:
+            Tensor: Losses of all positive samples in single image.
+        """
+        if not len(pos_inds):
+            return cls_score.new([]),
+        anchors_all_level = torch.cat(anchors, 0)
+        pos_scores = cls_score[pos_inds]
+        pos_bbox_pred = bbox_pred[pos_inds]
+        pos_label = label[pos_inds]
+        pos_label_weight = label_weight[pos_inds]
+        pos_bbox_target = bbox_target[pos_inds]
+        pos_bbox_weight = bbox_weight[pos_inds]
+        pos_anchors = anchors_all_level[pos_inds]
+        pos_bbox_pred = self.bbox_coder.decode(pos_anchors, pos_bbox_pred)
+
+        # to keep loss dimension
+        loss_cls = self.loss_cls(
+            pos_scores,
+            pos_label,
+            pos_label_weight,
+            avg_factor=self.loss_cls.loss_weight,
+            reduction_override='none')
+
+        loss_bbox = self.loss_bbox(
+            pos_bbox_pred,
+            pos_bbox_target,
+            pos_bbox_weight,
+            avg_factor=self.loss_cls.loss_weight,
+            reduction_override='none')
+
+        loss_cls = loss_cls.sum(-1)
+        pos_loss = loss_bbox + loss_cls
+        return pos_loss,
+
+    def paa_reassign(self, pos_losses, label, label_weight, bbox_weight,
+                     pos_inds, pos_gt_inds, anchors):
+        """Fit loss to GMM distribution and separate positive, ignore, negative
+        samples again with GMM model.
+
+        Args:
+            pos_losses (Tensor): Losses of all positive samples in
+                single image.
+            label (Tensor): classification target of each anchor with
+                shape (num_anchors,)
+            label_weight (Tensor): Classification loss weight of each
+                anchor with shape (num_anchors).
+            bbox_weight (Tensor): Bbox weight of each anchor with shape
+                (num_anchors, 4).
+            pos_inds (Tensor): Index of all positive samples got from
+                first assign process.
+            pos_gt_inds (Tensor): Gt_index of all positive samples got
+                from first assign process.
+            anchors (list[Tensor]): Anchors of each scale.
+
+        Returns:
+            tuple: Usually returns a tuple containing learning targets.
+
+                - label (Tensor): classification target of each anchor after
+                  paa assign, with shape (num_anchors,)
+                - label_weight (Tensor): Classification loss weight of each
+                  anchor after paa assign, with shape (num_anchors).
+                - bbox_weight (Tensor): Bbox weight of each anchor with shape
+                  (num_anchors, 4).
+                - num_pos (int): The number of positive samples after paa
+                  assign.
+        """
+        if not len(pos_inds):
+            return label, label_weight, bbox_weight, 0
+        label = label.clone()
+        label_weight = label_weight.clone()
+        bbox_weight = bbox_weight.clone()
+        num_gt = pos_gt_inds.max() + 1
+        num_level = len(anchors)
+        num_anchors_each_level = [item.size(0) for item in anchors]
+        num_anchors_each_level.insert(0, 0)
+        inds_level_interval = np.cumsum(num_anchors_each_level)
+        pos_level_mask = []
+        for i in range(num_level):
+            mask = (pos_inds >= inds_level_interval[i]) & (
+                pos_inds < inds_level_interval[i + 1])
+            pos_level_mask.append(mask)
+        pos_inds_after_paa = [label.new_tensor([])]
+        ignore_inds_after_paa = [label.new_tensor([])]
+        for gt_ind in range(num_gt):
+            pos_inds_gmm = []
+            pos_loss_gmm = []
+            gt_mask = pos_gt_inds == gt_ind
+            for level in range(num_level):
+                level_mask = pos_level_mask[level]
+                level_gt_mask = level_mask & gt_mask
+                value, topk_inds = pos_losses[level_gt_mask].topk(
+                    min(level_gt_mask.sum(), self.topk), largest=False)
+                pos_inds_gmm.append(pos_inds[level_gt_mask][topk_inds])
+                pos_loss_gmm.append(value)
+            pos_inds_gmm = torch.cat(pos_inds_gmm)
+            pos_loss_gmm = torch.cat(pos_loss_gmm)
+            # fix gmm need at least two sample
+            if len(pos_inds_gmm) < 2:
+                continue
+            device = pos_inds_gmm.device
+            pos_loss_gmm, sort_inds = pos_loss_gmm.sort()
+            pos_inds_gmm = pos_inds_gmm[sort_inds]
+            pos_loss_gmm = pos_loss_gmm.view(-1, 1).cpu().numpy()
+            min_loss, max_loss = pos_loss_gmm.min(), pos_loss_gmm.max()
+            means_init = np.array([min_loss, max_loss]).reshape(2, 1)
+            weights_init = np.array([0.5, 0.5])
+            precisions_init = np.array([1.0, 1.0]).reshape(2, 1, 1)  # full
+            if self.covariance_type == 'spherical':
+                precisions_init = precisions_init.reshape(2)
+            elif self.covariance_type == 'diag':
+                precisions_init = precisions_init.reshape(2, 1)
+            elif self.covariance_type == 'tied':
+                precisions_init = np.array([[1.0]])
+            if skm is None:
+                raise ImportError('Please run "pip install sklearn" '
+                                  'to install sklearn first.')
+            gmm = skm.GaussianMixture(
+                2,
+                weights_init=weights_init,
+                means_init=means_init,
+                precisions_init=precisions_init,
+                covariance_type=self.covariance_type)
+            gmm.fit(pos_loss_gmm)
+            gmm_assignment = gmm.predict(pos_loss_gmm)
+            scores = gmm.score_samples(pos_loss_gmm)
+            gmm_assignment = torch.from_numpy(gmm_assignment).to(device)
+            scores = torch.from_numpy(scores).to(device)
+
+            pos_inds_temp, ignore_inds_temp = self.gmm_separation_scheme(
+                gmm_assignment, scores, pos_inds_gmm)
+            pos_inds_after_paa.append(pos_inds_temp)
+            ignore_inds_after_paa.append(ignore_inds_temp)
+
+        pos_inds_after_paa = torch.cat(pos_inds_after_paa)
+        ignore_inds_after_paa = torch.cat(ignore_inds_after_paa)
+        reassign_mask = (pos_inds.unsqueeze(1) != pos_inds_after_paa).all(1)
+        reassign_ids = pos_inds[reassign_mask]
+        label[reassign_ids] = self.num_classes
+        label_weight[ignore_inds_after_paa] = 0
+        bbox_weight[reassign_ids] = 0
+        num_pos = len(pos_inds_after_paa)
+        return label, label_weight, bbox_weight, num_pos
+
+    def gmm_separation_scheme(self, gmm_assignment, scores, pos_inds_gmm):
+        """A general separation scheme for gmm model.
+
+        It separates a GMM distribution of candidate samples into three
+        parts, 0 1 and uncertain areas, and you can implement other
+        separation schemes by rewriting this function.
+
+        Args:
+            gmm_assignment (Tensor): The prediction of GMM which is of shape
+                (num_samples,). The 0/1 value indicates the distribution
+                that each sample comes from.
+            scores (Tensor): The probability of sample coming from the
+                fit GMM distribution. The tensor is of shape (num_samples,).
+            pos_inds_gmm (Tensor): All the indexes of samples which are used
+                to fit GMM model. The tensor is of shape (num_samples,)
+
+        Returns:
+            tuple[Tensor]: The indices of positive and ignored samples.
+
+                - pos_inds_temp (Tensor): Indices of positive samples.
+                - ignore_inds_temp (Tensor): Indices of ignore samples.
+        """
+        # The implementation is (c) in Fig.3 in origin paper instead of (b).
+        # You can refer to issues such as
+        # https://github.com/kkhoot/PAA/issues/8 and
+        # https://github.com/kkhoot/PAA/issues/9.
+        fgs = gmm_assignment == 0
+        pos_inds_temp = fgs.new_tensor([], dtype=torch.long)
+        ignore_inds_temp = fgs.new_tensor([], dtype=torch.long)
+        if fgs.nonzero().numel():
+            _, pos_thr_ind = scores[fgs].topk(1)
+            pos_inds_temp = pos_inds_gmm[fgs][:pos_thr_ind + 1]
+            ignore_inds_temp = pos_inds_gmm.new_tensor([])
+        return pos_inds_temp, ignore_inds_temp
+
+    def get_targets(
+        self,
+        anchor_list,
+        valid_flag_list,
+        gt_bboxes_list,
+        img_metas,
+        gt_bboxes_ignore_list=None,
+        gt_labels_list=None,
+        label_channels=1,
+        unmap_outputs=True,
+    ):
+        """Get targets for PAA head.
+
+        This method is almost the same as `AnchorHead.get_targets()`. We direct
+        return the results from _get_targets_single instead map it to levels
+        by images_to_levels function.
+
+        Args:
+            anchor_list (list[list[Tensor]]): Multi level anchors of each
+                image. The outer list indicates images, and the inner list
+                corresponds to feature levels of the image. Each element of
+                the inner list is a tensor of shape (num_anchors, 4).
+            valid_flag_list (list[list[Tensor]]): Multi level valid flags of
+                each image. The outer list indicates images, and the inner list
+                corresponds to feature levels of the image. Each element of
+                the inner list is a tensor of shape (num_anchors, )
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image.
+            img_metas (list[dict]): Meta info of each image.
+            gt_bboxes_ignore_list (list[Tensor]): Ground truth bboxes to be
+                ignored.
+            gt_labels_list (list[Tensor]): Ground truth labels of each box.
+            label_channels (int): Channel of label.
+            unmap_outputs (bool): Whether to map outputs back to the original
+                set of anchors.
+
+        Returns:
+            tuple: Usually returns a tuple containing learning targets.
+
+                - labels (list[Tensor]): Labels of all anchors, each with
+                    shape (num_anchors,).
+                - label_weights (list[Tensor]): Label weights of all anchor.
+                    each with shape (num_anchors,).
+                - bbox_targets (list[Tensor]): BBox targets of all anchors.
+                    each with shape (num_anchors, 4).
+                - bbox_weights (list[Tensor]): BBox weights of all anchors.
+                    each with shape (num_anchors, 4).
+                - pos_inds (list[Tensor]): Contains all index of positive
+                    sample in all anchor.
+                - gt_inds (list[Tensor]): Contains all gt_index of positive
+                    sample in all anchor.
+        """
+
+        num_imgs = len(img_metas)
+        assert len(anchor_list) == len(valid_flag_list) == num_imgs
+        concat_anchor_list = []
+        concat_valid_flag_list = []
+        for i in range(num_imgs):
+            assert len(anchor_list[i]) == len(valid_flag_list[i])
+            concat_anchor_list.append(torch.cat(anchor_list[i]))
+            concat_valid_flag_list.append(torch.cat(valid_flag_list[i]))
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        results = multi_apply(
+            self._get_targets_single,
+            concat_anchor_list,
+            concat_valid_flag_list,
+            gt_bboxes_list,
+            gt_bboxes_ignore_list,
+            gt_labels_list,
+            img_metas,
+            label_channels=label_channels,
+            unmap_outputs=unmap_outputs)
+
+        (labels, label_weights, bbox_targets, bbox_weights, valid_pos_inds,
+         valid_neg_inds, sampling_result) = results
+
+        # Due to valid flag of anchors, we have to calculate the real pos_inds
+        # in origin anchor set.
+        pos_inds = []
+        for i, single_labels in enumerate(labels):
+            pos_mask = (0 <= single_labels) & (
+                single_labels < self.num_classes)
+            pos_inds.append(pos_mask.nonzero().view(-1))
+
+        gt_inds = [item.pos_assigned_gt_inds for item in sampling_result]
+        return (labels, label_weights, bbox_targets, bbox_weights, pos_inds,
+                gt_inds)
+
+    def _get_targets_single(self,
+                            flat_anchors,
+                            valid_flags,
+                            gt_bboxes,
+                            gt_bboxes_ignore,
+                            gt_labels,
+                            img_meta,
+                            label_channels=1,
+                            unmap_outputs=True):
+        """Compute regression and classification targets for anchors in a
+        single image.
+
+        This method is same as `AnchorHead._get_targets_single()`.
+        """
+        assert unmap_outputs, 'We must map outputs back to the original' \
+            'set of anchors in PAAhead'
+        return super(ATSSHead, self)._get_targets_single(
+            flat_anchors,
+            valid_flags,
+            gt_bboxes,
+            gt_bboxes_ignore,
+            gt_labels,
+            img_meta,
+            label_channels=1,
+            unmap_outputs=True)
+
+    def _get_bboxes(self,
+                    cls_scores,
+                    bbox_preds,
+                    iou_preds,
+                    mlvl_anchors,
+                    img_shapes,
+                    scale_factors,
+                    cfg,
+                    rescale=False,
+                    with_nms=True):
+        """Transform outputs for a single batch item into labeled boxes.
+
+        This method is almost same as `ATSSHead._get_bboxes()`.
+        We use sqrt(iou_preds * cls_scores) in NMS process instead of just
+        cls_scores. Besides, score voting is used when `` score_voting``
+        is set to True.
+        """
+        assert with_nms, 'PAA only supports "with_nms=True" now'
+        assert len(cls_scores) == len(bbox_preds) == len(mlvl_anchors)
+        batch_size = cls_scores[0].shape[0]
+
+        mlvl_bboxes = []
+        mlvl_scores = []
+        mlvl_iou_preds = []
+        for cls_score, bbox_pred, iou_preds, anchors in zip(
+                cls_scores, bbox_preds, iou_preds, mlvl_anchors):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+
+            scores = cls_score.permute(0, 2, 3, 1).reshape(
+                batch_size, -1, self.cls_out_channels).sigmoid()
+            bbox_pred = bbox_pred.permute(0, 2, 3,
+                                          1).reshape(batch_size, -1, 4)
+            iou_preds = iou_preds.permute(0, 2, 3, 1).reshape(batch_size,
+                                                              -1).sigmoid()
+
+            nms_pre = cfg.get('nms_pre', -1)
+            if nms_pre > 0 and scores.shape[1] > nms_pre:
+                max_scores, _ = (scores * iou_preds[..., None]).sqrt().max(-1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                batch_inds = torch.arange(batch_size).view(
+                    -1, 1).expand_as(topk_inds).long()
+                anchors = anchors[topk_inds, :]
+                bbox_pred = bbox_pred[batch_inds, topk_inds, :]
+                scores = scores[batch_inds, topk_inds, :]
+                iou_preds = iou_preds[batch_inds, topk_inds]
+            else:
+                anchors = anchors.expand_as(bbox_pred)
+
+            bboxes = self.bbox_coder.decode(
+                anchors, bbox_pred, max_shape=img_shapes)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+            mlvl_iou_preds.append(iou_preds)
+
+        batch_mlvl_bboxes = torch.cat(mlvl_bboxes, dim=1)
+        if rescale:
+            batch_mlvl_bboxes /= batch_mlvl_bboxes.new_tensor(
+                scale_factors).unsqueeze(1)
+        batch_mlvl_scores = torch.cat(mlvl_scores, dim=1)
+        # Add a dummy background class to the backend when using sigmoid
+        # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+        # BG cat_id: num_class
+        padding = batch_mlvl_scores.new_zeros(batch_size,
+                                              batch_mlvl_scores.shape[1], 1)
+        batch_mlvl_scores = torch.cat([batch_mlvl_scores, padding], dim=-1)
+        batch_mlvl_iou_preds = torch.cat(mlvl_iou_preds, dim=1)
+        batch_mlvl_nms_scores = (batch_mlvl_scores *
+                                 batch_mlvl_iou_preds[..., None]).sqrt()
+
+        det_results = []
+        for (mlvl_bboxes, mlvl_scores) in zip(batch_mlvl_bboxes,
+                                              batch_mlvl_nms_scores):
+            det_bbox, det_label = multiclass_nms(
+                mlvl_bboxes,
+                mlvl_scores,
+                cfg.score_thr,
+                cfg.nms,
+                cfg.max_per_img,
+                score_factors=None)
+            if self.with_score_voting and len(det_bbox) > 0:
+                det_bbox, det_label = self.score_voting(
+                    det_bbox, det_label, mlvl_bboxes, mlvl_scores,
+                    cfg.score_thr)
+            det_results.append(tuple([det_bbox, det_label]))
+
+        return det_results
+
+    def score_voting(self, det_bboxes, det_labels, mlvl_bboxes,
+                     mlvl_nms_scores, score_thr):
+        """Implementation of score voting method works on each remaining boxes
+        after NMS procedure.
+
+        Args:
+            det_bboxes (Tensor): Remaining boxes after NMS procedure,
+                with shape (k, 5), each dimension means
+                (x1, y1, x2, y2, score).
+            det_labels (Tensor): The label of remaining boxes, with shape
+                (k, 1),Labels are 0-based.
+            mlvl_bboxes (Tensor): All boxes before the NMS procedure,
+                with shape (num_anchors,4).
+            mlvl_nms_scores (Tensor): The scores of all boxes which is used
+                in the NMS procedure, with shape (num_anchors, num_class)
+            mlvl_iou_preds (Tensor): The predictions of IOU of all boxes
+                before the NMS procedure, with shape (num_anchors, 1)
+            score_thr (float): The score threshold of bboxes.
+
+        Returns:
+            tuple: Usually returns a tuple containing voting results.
+
+                - det_bboxes_voted (Tensor): Remaining boxes after
+                    score voting procedure, with shape (k, 5), each
+                    dimension means (x1, y1, x2, y2, score).
+                - det_labels_voted (Tensor): Label of remaining bboxes
+                    after voting, with shape (num_anchors,).
+        """
+        candidate_mask = mlvl_nms_scores > score_thr
+        candidate_mask_nonzeros = candidate_mask.nonzero()
+        candidate_inds = candidate_mask_nonzeros[:, 0]
+        candidate_labels = candidate_mask_nonzeros[:, 1]
+        candidate_bboxes = mlvl_bboxes[candidate_inds]
+        candidate_scores = mlvl_nms_scores[candidate_mask]
+        det_bboxes_voted = []
+        det_labels_voted = []
+        for cls in range(self.cls_out_channels):
+            candidate_cls_mask = candidate_labels == cls
+            if not candidate_cls_mask.any():
+                continue
+            candidate_cls_scores = candidate_scores[candidate_cls_mask]
+            candidate_cls_bboxes = candidate_bboxes[candidate_cls_mask]
+            det_cls_mask = det_labels == cls
+            det_cls_bboxes = det_bboxes[det_cls_mask].view(
+                -1, det_bboxes.size(-1))
+            det_candidate_ious = bbox_overlaps(det_cls_bboxes[:, :4],
+                                               candidate_cls_bboxes)
+            for det_ind in range(len(det_cls_bboxes)):
+                single_det_ious = det_candidate_ious[det_ind]
+                pos_ious_mask = single_det_ious > 0.01
+                pos_ious = single_det_ious[pos_ious_mask]
+                pos_bboxes = candidate_cls_bboxes[pos_ious_mask]
+                pos_scores = candidate_cls_scores[pos_ious_mask]
+                pis = (torch.exp(-(1 - pos_ious)**2 / 0.025) *
+                       pos_scores)[:, None]
+                voted_box = torch.sum(
+                    pis * pos_bboxes, dim=0) / torch.sum(
+                        pis, dim=0)
+                voted_score = det_cls_bboxes[det_ind][-1:][None, :]
+                det_bboxes_voted.append(
+                    torch.cat((voted_box[None, :], voted_score), dim=1))
+                det_labels_voted.append(cls)
+
+        det_bboxes_voted = torch.cat(det_bboxes_voted, dim=0)
+        det_labels_voted = det_labels.new_tensor(det_labels_voted)
+        return det_bboxes_voted, det_labels_voted
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/pisa_retinanet_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/pisa_retinanet_head.py
new file mode 100644
index 0000000..bd87b9a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/pisa_retinanet_head.py
@@ -0,0 +1,154 @@
+import torch
+from mmcv.runner import force_fp32
+
+from mmdet.core import images_to_levels
+from ..builder import HEADS
+from ..losses import carl_loss, isr_p
+from .retina_head import RetinaHead
+
+
+@HEADS.register_module()
+class PISARetinaHead(RetinaHead):
+    """PISA Retinanet Head.
+
+    The head owns the same structure with Retinanet Head, but differs in two
+        aspects:
+        1. Importance-based Sample Reweighting Positive (ISR-P) is applied to
+            change the positive loss weights.
+        2. Classification-aware regression loss is adopted as a third loss.
+    """
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes of each image
+                with shape (num_obj, 4).
+            gt_labels (list[Tensor]): Ground truth labels of each image
+                with shape (num_obj, 4).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (list[Tensor]): Ignored gt bboxes of each image.
+                Default: None.
+
+        Returns:
+            dict: Loss dict, comprise classification loss, regression loss and
+                carl loss.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels,
+            return_sampling_results=True)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg, sampling_results_list) = cls_reg_targets
+        num_total_samples = (
+            num_total_pos + num_total_neg if self.sampling else num_total_pos)
+
+        # anchor number of multi levels
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+        # concat all level anchors and flags to a single tensor
+        concat_anchor_list = []
+        for i in range(len(anchor_list)):
+            concat_anchor_list.append(torch.cat(anchor_list[i]))
+        all_anchor_list = images_to_levels(concat_anchor_list,
+                                           num_level_anchors)
+
+        num_imgs = len(img_metas)
+        flatten_cls_scores = [
+            cls_score.permute(0, 2, 3, 1).reshape(num_imgs, -1, label_channels)
+            for cls_score in cls_scores
+        ]
+        flatten_cls_scores = torch.cat(
+            flatten_cls_scores, dim=1).reshape(-1,
+                                               flatten_cls_scores[0].size(-1))
+        flatten_bbox_preds = [
+            bbox_pred.permute(0, 2, 3, 1).reshape(num_imgs, -1, 4)
+            for bbox_pred in bbox_preds
+        ]
+        flatten_bbox_preds = torch.cat(
+            flatten_bbox_preds, dim=1).view(-1, flatten_bbox_preds[0].size(-1))
+        flatten_labels = torch.cat(labels_list, dim=1).reshape(-1)
+        flatten_label_weights = torch.cat(
+            label_weights_list, dim=1).reshape(-1)
+        flatten_anchors = torch.cat(all_anchor_list, dim=1).reshape(-1, 4)
+        flatten_bbox_targets = torch.cat(
+            bbox_targets_list, dim=1).reshape(-1, 4)
+        flatten_bbox_weights = torch.cat(
+            bbox_weights_list, dim=1).reshape(-1, 4)
+
+        # Apply ISR-P
+        isr_cfg = self.train_cfg.get('isr', None)
+        if isr_cfg is not None:
+            all_targets = (flatten_labels, flatten_label_weights,
+                           flatten_bbox_targets, flatten_bbox_weights)
+            with torch.no_grad():
+                all_targets = isr_p(
+                    flatten_cls_scores,
+                    flatten_bbox_preds,
+                    all_targets,
+                    flatten_anchors,
+                    sampling_results_list,
+                    bbox_coder=self.bbox_coder,
+                    loss_cls=self.loss_cls,
+                    num_class=self.num_classes,
+                    **self.train_cfg.isr)
+            (flatten_labels, flatten_label_weights, flatten_bbox_targets,
+             flatten_bbox_weights) = all_targets
+
+        # For convenience we compute loss once instead separating by fpn level,
+        # so that we don't need to separate the weights by level again.
+        # The result should be the same
+        losses_cls = self.loss_cls(
+            flatten_cls_scores,
+            flatten_labels,
+            flatten_label_weights,
+            avg_factor=num_total_samples)
+        losses_bbox = self.loss_bbox(
+            flatten_bbox_preds,
+            flatten_bbox_targets,
+            flatten_bbox_weights,
+            avg_factor=num_total_samples)
+        loss_dict = dict(loss_cls=losses_cls, loss_bbox=losses_bbox)
+
+        # CARL Loss
+        carl_cfg = self.train_cfg.get('carl', None)
+        if carl_cfg is not None:
+            loss_carl = carl_loss(
+                flatten_cls_scores,
+                flatten_labels,
+                flatten_bbox_preds,
+                flatten_bbox_targets,
+                self.loss_bbox,
+                **self.train_cfg.carl,
+                avg_factor=num_total_pos,
+                sigmoid=True,
+                num_class=self.num_classes)
+            loss_dict.update(loss_carl)
+
+        return loss_dict
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/pisa_ssd_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/pisa_ssd_head.py
new file mode 100644
index 0000000..90ef3c8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/pisa_ssd_head.py
@@ -0,0 +1,139 @@
+import torch
+
+from mmdet.core import multi_apply
+from ..builder import HEADS
+from ..losses import CrossEntropyLoss, SmoothL1Loss, carl_loss, isr_p
+from .ssd_head import SSDHead
+
+
+# TODO: add loss evaluator for SSD
+@HEADS.register_module()
+class PISASSDHead(SSDHead):
+
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes of each image
+                with shape (num_obj, 4).
+            gt_labels (list[Tensor]): Ground truth labels of each image
+                with shape (num_obj, 4).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (list[Tensor]): Ignored gt bboxes of each image.
+                Default: None.
+
+        Returns:
+            dict: Loss dict, comprise classification loss regression loss and
+                carl loss.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=1,
+            unmap_outputs=False,
+            return_sampling_results=True)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg, sampling_results_list) = cls_reg_targets
+
+        num_images = len(img_metas)
+        all_cls_scores = torch.cat([
+            s.permute(0, 2, 3, 1).reshape(
+                num_images, -1, self.cls_out_channels) for s in cls_scores
+        ], 1)
+        all_labels = torch.cat(labels_list, -1).view(num_images, -1)
+        all_label_weights = torch.cat(label_weights_list,
+                                      -1).view(num_images, -1)
+        all_bbox_preds = torch.cat([
+            b.permute(0, 2, 3, 1).reshape(num_images, -1, 4)
+            for b in bbox_preds
+        ], -2)
+        all_bbox_targets = torch.cat(bbox_targets_list,
+                                     -2).view(num_images, -1, 4)
+        all_bbox_weights = torch.cat(bbox_weights_list,
+                                     -2).view(num_images, -1, 4)
+
+        # concat all level anchors to a single tensor
+        all_anchors = []
+        for i in range(num_images):
+            all_anchors.append(torch.cat(anchor_list[i]))
+
+        isr_cfg = self.train_cfg.get('isr', None)
+        all_targets = (all_labels.view(-1), all_label_weights.view(-1),
+                       all_bbox_targets.view(-1,
+                                             4), all_bbox_weights.view(-1, 4))
+        # apply ISR-P
+        if isr_cfg is not None:
+            all_targets = isr_p(
+                all_cls_scores.view(-1, all_cls_scores.size(-1)),
+                all_bbox_preds.view(-1, 4),
+                all_targets,
+                torch.cat(all_anchors),
+                sampling_results_list,
+                loss_cls=CrossEntropyLoss(),
+                bbox_coder=self.bbox_coder,
+                **self.train_cfg.isr,
+                num_class=self.num_classes)
+            (new_labels, new_label_weights, new_bbox_targets,
+             new_bbox_weights) = all_targets
+            all_labels = new_labels.view(all_labels.shape)
+            all_label_weights = new_label_weights.view(all_label_weights.shape)
+            all_bbox_targets = new_bbox_targets.view(all_bbox_targets.shape)
+            all_bbox_weights = new_bbox_weights.view(all_bbox_weights.shape)
+
+        # add CARL loss
+        carl_loss_cfg = self.train_cfg.get('carl', None)
+        if carl_loss_cfg is not None:
+            loss_carl = carl_loss(
+                all_cls_scores.view(-1, all_cls_scores.size(-1)),
+                all_targets[0],
+                all_bbox_preds.view(-1, 4),
+                all_targets[2],
+                SmoothL1Loss(beta=1.),
+                **self.train_cfg.carl,
+                avg_factor=num_total_pos,
+                num_class=self.num_classes)
+
+        # check NaN and Inf
+        assert torch.isfinite(all_cls_scores).all().item(), \
+            'classification scores become infinite or NaN!'
+        assert torch.isfinite(all_bbox_preds).all().item(), \
+            'bbox predications become infinite or NaN!'
+
+        losses_cls, losses_bbox = multi_apply(
+            self.loss_single,
+            all_cls_scores,
+            all_bbox_preds,
+            all_anchors,
+            all_labels,
+            all_label_weights,
+            all_bbox_targets,
+            all_bbox_weights,
+            num_total_samples=num_total_pos)
+        loss_dict = dict(loss_cls=losses_cls, loss_bbox=losses_bbox)
+        if carl_loss_cfg is not None:
+            loss_dict.update(loss_carl)
+        return loss_dict
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/reppoints_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/reppoints_head.py
new file mode 100644
index 0000000..499cc4f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/reppoints_head.py
@@ -0,0 +1,763 @@
+import numpy as np
+import torch
+import torch.nn as nn
+from mmcv.cnn import ConvModule, bias_init_with_prob, normal_init
+from mmcv.ops import DeformConv2d
+
+from mmdet.core import (PointGenerator, build_assigner, build_sampler,
+                        images_to_levels, multi_apply, multiclass_nms, unmap)
+from ..builder import HEADS, build_loss
+from .anchor_free_head import AnchorFreeHead
+
+
+@HEADS.register_module()
+class RepPointsHead(AnchorFreeHead):
+    """RepPoint head.
+
+    Args:
+        point_feat_channels (int): Number of channels of points features.
+        gradient_mul (float): The multiplier to gradients from
+            points refinement and recognition.
+        point_strides (Iterable): points strides.
+        point_base_scale (int): bbox scale for assigning labels.
+        loss_cls (dict): Config of classification loss.
+        loss_bbox_init (dict): Config of initial points loss.
+        loss_bbox_refine (dict): Config of points loss in refinement.
+        use_grid_points (bool): If we use bounding box representation, the
+        reppoints is represented as grid points on the bounding box.
+        center_init (bool): Whether to use center point assignment.
+        transform_method (str): The methods to transform RepPoints to bbox.
+    """  # noqa: W605
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 point_feat_channels=256,
+                 num_points=9,
+                 gradient_mul=0.1,
+                 point_strides=[8, 16, 32, 64, 128],
+                 point_base_scale=4,
+                 loss_cls=dict(
+                     type='FocalLoss',
+                     use_sigmoid=True,
+                     gamma=2.0,
+                     alpha=0.25,
+                     loss_weight=1.0),
+                 loss_bbox_init=dict(
+                     type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=0.5),
+                 loss_bbox_refine=dict(
+                     type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0),
+                 use_grid_points=False,
+                 center_init=True,
+                 transform_method='moment',
+                 moment_mul=0.01,
+                 **kwargs):
+        self.num_points = num_points
+        self.point_feat_channels = point_feat_channels
+        self.use_grid_points = use_grid_points
+        self.center_init = center_init
+
+        # we use deform conv to extract points features
+        self.dcn_kernel = int(np.sqrt(num_points))
+        self.dcn_pad = int((self.dcn_kernel - 1) / 2)
+        assert self.dcn_kernel * self.dcn_kernel == num_points, \
+            'The points number should be a square number.'
+        assert self.dcn_kernel % 2 == 1, \
+            'The points number should be an odd square number.'
+        dcn_base = np.arange(-self.dcn_pad,
+                             self.dcn_pad + 1).astype(np.float64)
+        dcn_base_y = np.repeat(dcn_base, self.dcn_kernel)
+        dcn_base_x = np.tile(dcn_base, self.dcn_kernel)
+        dcn_base_offset = np.stack([dcn_base_y, dcn_base_x], axis=1).reshape(
+            (-1))
+        self.dcn_base_offset = torch.tensor(dcn_base_offset).view(1, -1, 1, 1)
+
+        super().__init__(num_classes, in_channels, loss_cls=loss_cls, **kwargs)
+
+        self.gradient_mul = gradient_mul
+        self.point_base_scale = point_base_scale
+        self.point_strides = point_strides
+        self.point_generators = [PointGenerator() for _ in self.point_strides]
+
+        self.sampling = loss_cls['type'] not in ['FocalLoss']
+        if self.train_cfg:
+            self.init_assigner = build_assigner(self.train_cfg.init.assigner)
+            self.refine_assigner = build_assigner(
+                self.train_cfg.refine.assigner)
+            # use PseudoSampler when sampling is False
+            if self.sampling and hasattr(self.train_cfg, 'sampler'):
+                sampler_cfg = self.train_cfg.sampler
+            else:
+                sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+        self.transform_method = transform_method
+        if self.transform_method == 'moment':
+            self.moment_transfer = nn.Parameter(
+                data=torch.zeros(2), requires_grad=True)
+            self.moment_mul = moment_mul
+
+        self.use_sigmoid_cls = loss_cls.get('use_sigmoid', False)
+        if self.use_sigmoid_cls:
+            self.cls_out_channels = self.num_classes
+        else:
+            self.cls_out_channels = self.num_classes + 1
+        self.loss_bbox_init = build_loss(loss_bbox_init)
+        self.loss_bbox_refine = build_loss(loss_bbox_refine)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.relu = nn.ReLU(inplace=True)
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            self.cls_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+            self.reg_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+        pts_out_dim = 4 if self.use_grid_points else 2 * self.num_points
+        self.reppoints_cls_conv = DeformConv2d(self.feat_channels,
+                                               self.point_feat_channels,
+                                               self.dcn_kernel, 1,
+                                               self.dcn_pad)
+        self.reppoints_cls_out = nn.Conv2d(self.point_feat_channels,
+                                           self.cls_out_channels, 1, 1, 0)
+        self.reppoints_pts_init_conv = nn.Conv2d(self.feat_channels,
+                                                 self.point_feat_channels, 3,
+                                                 1, 1)
+        self.reppoints_pts_init_out = nn.Conv2d(self.point_feat_channels,
+                                                pts_out_dim, 1, 1, 0)
+        self.reppoints_pts_refine_conv = DeformConv2d(self.feat_channels,
+                                                      self.point_feat_channels,
+                                                      self.dcn_kernel, 1,
+                                                      self.dcn_pad)
+        self.reppoints_pts_refine_out = nn.Conv2d(self.point_feat_channels,
+                                                  pts_out_dim, 1, 1, 0)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.cls_convs:
+            normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            normal_init(m.conv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.reppoints_cls_conv, std=0.01)
+        normal_init(self.reppoints_cls_out, std=0.01, bias=bias_cls)
+        normal_init(self.reppoints_pts_init_conv, std=0.01)
+        normal_init(self.reppoints_pts_init_out, std=0.01)
+        normal_init(self.reppoints_pts_refine_conv, std=0.01)
+        normal_init(self.reppoints_pts_refine_out, std=0.01)
+
+    def points2bbox(self, pts, y_first=True):
+        """Converting the points set into bounding box.
+
+        :param pts: the input points sets (fields), each points
+            set (fields) is represented as 2n scalar.
+        :param y_first: if y_first=True, the point set is represented as
+            [y1, x1, y2, x2 ... yn, xn], otherwise the point set is
+            represented as [x1, y1, x2, y2 ... xn, yn].
+        :return: each points set is converting to a bbox [x1, y1, x2, y2].
+        """
+        pts_reshape = pts.view(pts.shape[0], -1, 2, *pts.shape[2:])
+        pts_y = pts_reshape[:, :, 0, ...] if y_first else pts_reshape[:, :, 1,
+                                                                      ...]
+        pts_x = pts_reshape[:, :, 1, ...] if y_first else pts_reshape[:, :, 0,
+                                                                      ...]
+        if self.transform_method == 'minmax':
+            bbox_left = pts_x.min(dim=1, keepdim=True)[0]
+            bbox_right = pts_x.max(dim=1, keepdim=True)[0]
+            bbox_up = pts_y.min(dim=1, keepdim=True)[0]
+            bbox_bottom = pts_y.max(dim=1, keepdim=True)[0]
+            bbox = torch.cat([bbox_left, bbox_up, bbox_right, bbox_bottom],
+                             dim=1)
+        elif self.transform_method == 'partial_minmax':
+            pts_y = pts_y[:, :4, ...]
+            pts_x = pts_x[:, :4, ...]
+            bbox_left = pts_x.min(dim=1, keepdim=True)[0]
+            bbox_right = pts_x.max(dim=1, keepdim=True)[0]
+            bbox_up = pts_y.min(dim=1, keepdim=True)[0]
+            bbox_bottom = pts_y.max(dim=1, keepdim=True)[0]
+            bbox = torch.cat([bbox_left, bbox_up, bbox_right, bbox_bottom],
+                             dim=1)
+        elif self.transform_method == 'moment':
+            pts_y_mean = pts_y.mean(dim=1, keepdim=True)
+            pts_x_mean = pts_x.mean(dim=1, keepdim=True)
+            pts_y_std = torch.std(pts_y - pts_y_mean, dim=1, keepdim=True)
+            pts_x_std = torch.std(pts_x - pts_x_mean, dim=1, keepdim=True)
+            moment_transfer = (self.moment_transfer * self.moment_mul) + (
+                self.moment_transfer.detach() * (1 - self.moment_mul))
+            moment_width_transfer = moment_transfer[0]
+            moment_height_transfer = moment_transfer[1]
+            half_width = pts_x_std * torch.exp(moment_width_transfer)
+            half_height = pts_y_std * torch.exp(moment_height_transfer)
+            bbox = torch.cat([
+                pts_x_mean - half_width, pts_y_mean - half_height,
+                pts_x_mean + half_width, pts_y_mean + half_height
+            ],
+                             dim=1)
+        else:
+            raise NotImplementedError
+        return bbox
+
+    def gen_grid_from_reg(self, reg, previous_boxes):
+        """Base on the previous bboxes and regression values, we compute the
+        regressed bboxes and generate the grids on the bboxes.
+
+        :param reg: the regression value to previous bboxes.
+        :param previous_boxes: previous bboxes.
+        :return: generate grids on the regressed bboxes.
+        """
+        b, _, h, w = reg.shape
+        bxy = (previous_boxes[:, :2, ...] + previous_boxes[:, 2:, ...]) / 2.
+        bwh = (previous_boxes[:, 2:, ...] -
+               previous_boxes[:, :2, ...]).clamp(min=1e-6)
+        grid_topleft = bxy + bwh * reg[:, :2, ...] - 0.5 * bwh * torch.exp(
+            reg[:, 2:, ...])
+        grid_wh = bwh * torch.exp(reg[:, 2:, ...])
+        grid_left = grid_topleft[:, [0], ...]
+        grid_top = grid_topleft[:, [1], ...]
+        grid_width = grid_wh[:, [0], ...]
+        grid_height = grid_wh[:, [1], ...]
+        intervel = torch.linspace(0., 1., self.dcn_kernel).view(
+            1, self.dcn_kernel, 1, 1).type_as(reg)
+        grid_x = grid_left + grid_width * intervel
+        grid_x = grid_x.unsqueeze(1).repeat(1, self.dcn_kernel, 1, 1, 1)
+        grid_x = grid_x.view(b, -1, h, w)
+        grid_y = grid_top + grid_height * intervel
+        grid_y = grid_y.unsqueeze(2).repeat(1, 1, self.dcn_kernel, 1, 1)
+        grid_y = grid_y.view(b, -1, h, w)
+        grid_yx = torch.stack([grid_y, grid_x], dim=2)
+        grid_yx = grid_yx.view(b, -1, h, w)
+        regressed_bbox = torch.cat([
+            grid_left, grid_top, grid_left + grid_width, grid_top + grid_height
+        ], 1)
+        return grid_yx, regressed_bbox
+
+    def forward(self, feats):
+        return multi_apply(self.forward_single, feats)
+
+    def forward_single(self, x):
+        """Forward feature map of a single FPN level."""
+        dcn_base_offset = self.dcn_base_offset.type_as(x)
+        # If we use center_init, the initial reppoints is from center points.
+        # If we use bounding bbox representation, the initial reppoints is
+        #   from regular grid placed on a pre-defined bbox.
+        if self.use_grid_points or not self.center_init:
+            scale = self.point_base_scale / 2
+            points_init = dcn_base_offset / dcn_base_offset.max() * scale
+            bbox_init = x.new_tensor([-scale, -scale, scale,
+                                      scale]).view(1, 4, 1, 1)
+        else:
+            points_init = 0
+        cls_feat = x
+        pts_feat = x
+        for cls_conv in self.cls_convs:
+            cls_feat = cls_conv(cls_feat)
+        for reg_conv in self.reg_convs:
+            pts_feat = reg_conv(pts_feat)
+        # initialize reppoints
+        pts_out_init = self.reppoints_pts_init_out(
+            self.relu(self.reppoints_pts_init_conv(pts_feat)))
+        if self.use_grid_points:
+            pts_out_init, bbox_out_init = self.gen_grid_from_reg(
+                pts_out_init, bbox_init.detach())
+        else:
+            pts_out_init = pts_out_init + points_init
+        # refine and classify reppoints
+        pts_out_init_grad_mul = (1 - self.gradient_mul) * pts_out_init.detach(
+        ) + self.gradient_mul * pts_out_init
+        dcn_offset = pts_out_init_grad_mul - dcn_base_offset
+        cls_out = self.reppoints_cls_out(
+            self.relu(self.reppoints_cls_conv(cls_feat, dcn_offset)))
+        pts_out_refine = self.reppoints_pts_refine_out(
+            self.relu(self.reppoints_pts_refine_conv(pts_feat, dcn_offset)))
+        if self.use_grid_points:
+            pts_out_refine, bbox_out_refine = self.gen_grid_from_reg(
+                pts_out_refine, bbox_out_init.detach())
+        else:
+            pts_out_refine = pts_out_refine + pts_out_init.detach()
+        return cls_out, pts_out_init, pts_out_refine
+
+    def get_points(self, featmap_sizes, img_metas, device):
+        """Get points according to feature map sizes.
+
+        Args:
+            featmap_sizes (list[tuple]): Multi-level feature map sizes.
+            img_metas (list[dict]): Image meta info.
+
+        Returns:
+            tuple: points of each image, valid flags of each image
+        """
+        num_imgs = len(img_metas)
+        num_levels = len(featmap_sizes)
+
+        # since feature map sizes of all images are the same, we only compute
+        # points center for one time
+        multi_level_points = []
+        for i in range(num_levels):
+            points = self.point_generators[i].grid_points(
+                featmap_sizes[i], self.point_strides[i], device)
+            multi_level_points.append(points)
+        points_list = [[point.clone() for point in multi_level_points]
+                       for _ in range(num_imgs)]
+
+        # for each image, we compute valid flags of multi level grids
+        valid_flag_list = []
+        for img_id, img_meta in enumerate(img_metas):
+            multi_level_flags = []
+            for i in range(num_levels):
+                point_stride = self.point_strides[i]
+                feat_h, feat_w = featmap_sizes[i]
+                h, w = img_meta['pad_shape'][:2]
+                valid_feat_h = min(int(np.ceil(h / point_stride)), feat_h)
+                valid_feat_w = min(int(np.ceil(w / point_stride)), feat_w)
+                flags = self.point_generators[i].valid_flags(
+                    (feat_h, feat_w), (valid_feat_h, valid_feat_w), device)
+                multi_level_flags.append(flags)
+            valid_flag_list.append(multi_level_flags)
+
+        return points_list, valid_flag_list
+
+    def centers_to_bboxes(self, point_list):
+        """Get bboxes according to center points.
+
+        Only used in :class:`MaxIoUAssigner`.
+        """
+        bbox_list = []
+        for i_img, point in enumerate(point_list):
+            bbox = []
+            for i_lvl in range(len(self.point_strides)):
+                scale = self.point_base_scale * self.point_strides[i_lvl] * 0.5
+                bbox_shift = torch.Tensor([-scale, -scale, scale,
+                                           scale]).view(1, 4).type_as(point[0])
+                bbox_center = torch.cat(
+                    [point[i_lvl][:, :2], point[i_lvl][:, :2]], dim=1)
+                bbox.append(bbox_center + bbox_shift)
+            bbox_list.append(bbox)
+        return bbox_list
+
+    def offset_to_pts(self, center_list, pred_list):
+        """Change from point offset to point coordinate."""
+        pts_list = []
+        for i_lvl in range(len(self.point_strides)):
+            pts_lvl = []
+            for i_img in range(len(center_list)):
+                pts_center = center_list[i_img][i_lvl][:, :2].repeat(
+                    1, self.num_points)
+                pts_shift = pred_list[i_lvl][i_img]
+                yx_pts_shift = pts_shift.permute(1, 2, 0).view(
+                    -1, 2 * self.num_points)
+                y_pts_shift = yx_pts_shift[..., 0::2]
+                x_pts_shift = yx_pts_shift[..., 1::2]
+                xy_pts_shift = torch.stack([x_pts_shift, y_pts_shift], -1)
+                xy_pts_shift = xy_pts_shift.view(*yx_pts_shift.shape[:-1], -1)
+                pts = xy_pts_shift * self.point_strides[i_lvl] + pts_center
+                pts_lvl.append(pts)
+            pts_lvl = torch.stack(pts_lvl, 0)
+            pts_list.append(pts_lvl)
+        return pts_list
+
+    def _point_target_single(self,
+                             flat_proposals,
+                             valid_flags,
+                             gt_bboxes,
+                             gt_bboxes_ignore,
+                             gt_labels,
+                             label_channels=1,
+                             stage='init',
+                             unmap_outputs=True):
+        inside_flags = valid_flags
+        if not inside_flags.any():
+            return (None, ) * 7
+        # assign gt and sample proposals
+        proposals = flat_proposals[inside_flags, :]
+
+        if stage == 'init':
+            assigner = self.init_assigner
+            pos_weight = self.train_cfg.init.pos_weight
+        else:
+            assigner = self.refine_assigner
+            pos_weight = self.train_cfg.refine.pos_weight
+        assign_result = assigner.assign(proposals, gt_bboxes, gt_bboxes_ignore,
+                                        None if self.sampling else gt_labels)
+        sampling_result = self.sampler.sample(assign_result, proposals,
+                                              gt_bboxes)
+
+        num_valid_proposals = proposals.shape[0]
+        bbox_gt = proposals.new_zeros([num_valid_proposals, 4])
+        pos_proposals = torch.zeros_like(proposals)
+        proposals_weights = proposals.new_zeros([num_valid_proposals, 4])
+        labels = proposals.new_full((num_valid_proposals, ),
+                                    self.num_classes,
+                                    dtype=torch.long)
+        label_weights = proposals.new_zeros(
+            num_valid_proposals, dtype=torch.float)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+        if len(pos_inds) > 0:
+            pos_gt_bboxes = sampling_result.pos_gt_bboxes
+            bbox_gt[pos_inds, :] = pos_gt_bboxes
+            pos_proposals[pos_inds, :] = proposals[pos_inds, :]
+            proposals_weights[pos_inds, :] = 1.0
+            if gt_labels is None:
+                # Only rpn gives gt_labels as None
+                # Foreground is the first class
+                labels[pos_inds] = 0
+            else:
+                labels[pos_inds] = gt_labels[
+                    sampling_result.pos_assigned_gt_inds]
+            if pos_weight <= 0:
+                label_weights[pos_inds] = 1.0
+            else:
+                label_weights[pos_inds] = pos_weight
+        if len(neg_inds) > 0:
+            label_weights[neg_inds] = 1.0
+
+        # map up to original set of proposals
+        if unmap_outputs:
+            num_total_proposals = flat_proposals.size(0)
+            labels = unmap(labels, num_total_proposals, inside_flags)
+            label_weights = unmap(label_weights, num_total_proposals,
+                                  inside_flags)
+            bbox_gt = unmap(bbox_gt, num_total_proposals, inside_flags)
+            pos_proposals = unmap(pos_proposals, num_total_proposals,
+                                  inside_flags)
+            proposals_weights = unmap(proposals_weights, num_total_proposals,
+                                      inside_flags)
+
+        return (labels, label_weights, bbox_gt, pos_proposals,
+                proposals_weights, pos_inds, neg_inds)
+
+    def get_targets(self,
+                    proposals_list,
+                    valid_flag_list,
+                    gt_bboxes_list,
+                    img_metas,
+                    gt_bboxes_ignore_list=None,
+                    gt_labels_list=None,
+                    stage='init',
+                    label_channels=1,
+                    unmap_outputs=True):
+        """Compute corresponding GT box and classification targets for
+        proposals.
+
+        Args:
+            proposals_list (list[list]): Multi level points/bboxes of each
+                image.
+            valid_flag_list (list[list]): Multi level valid flags of each
+                image.
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image.
+            img_metas (list[dict]): Meta info of each image.
+            gt_bboxes_ignore_list (list[Tensor]): Ground truth bboxes to be
+                ignored.
+            gt_bboxes_list (list[Tensor]): Ground truth labels of each box.
+            stage (str): `init` or `refine`. Generate target for init stage or
+                refine stage
+            label_channels (int): Channel of label.
+            unmap_outputs (bool): Whether to map outputs back to the original
+                set of anchors.
+
+        Returns:
+            tuple:
+                - labels_list (list[Tensor]): Labels of each level.
+                - label_weights_list (list[Tensor]): Label weights of each level.  # noqa: E501
+                - bbox_gt_list (list[Tensor]): Ground truth bbox of each level.
+                - proposal_list (list[Tensor]): Proposals(points/bboxes) of each level.  # noqa: E501
+                - proposal_weights_list (list[Tensor]): Proposal weights of each level.  # noqa: E501
+                - num_total_pos (int): Number of positive samples in all images.  # noqa: E501
+                - num_total_neg (int): Number of negative samples in all images.  # noqa: E501
+        """
+        assert stage in ['init', 'refine']
+        num_imgs = len(img_metas)
+        assert len(proposals_list) == len(valid_flag_list) == num_imgs
+
+        # points number of multi levels
+        num_level_proposals = [points.size(0) for points in proposals_list[0]]
+
+        # concat all level points and flags to a single tensor
+        for i in range(num_imgs):
+            assert len(proposals_list[i]) == len(valid_flag_list[i])
+            proposals_list[i] = torch.cat(proposals_list[i])
+            valid_flag_list[i] = torch.cat(valid_flag_list[i])
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        (all_labels, all_label_weights, all_bbox_gt, all_proposals,
+         all_proposal_weights, pos_inds_list, neg_inds_list) = multi_apply(
+             self._point_target_single,
+             proposals_list,
+             valid_flag_list,
+             gt_bboxes_list,
+             gt_bboxes_ignore_list,
+             gt_labels_list,
+             stage=stage,
+             label_channels=label_channels,
+             unmap_outputs=unmap_outputs)
+        # no valid points
+        if any([labels is None for labels in all_labels]):
+            return None
+        # sampled points of all images
+        num_total_pos = sum([max(inds.numel(), 1) for inds in pos_inds_list])
+        num_total_neg = sum([max(inds.numel(), 1) for inds in neg_inds_list])
+        labels_list = images_to_levels(all_labels, num_level_proposals)
+        label_weights_list = images_to_levels(all_label_weights,
+                                              num_level_proposals)
+        bbox_gt_list = images_to_levels(all_bbox_gt, num_level_proposals)
+        proposals_list = images_to_levels(all_proposals, num_level_proposals)
+        proposal_weights_list = images_to_levels(all_proposal_weights,
+                                                 num_level_proposals)
+        return (labels_list, label_weights_list, bbox_gt_list, proposals_list,
+                proposal_weights_list, num_total_pos, num_total_neg)
+
+    def loss_single(self, cls_score, pts_pred_init, pts_pred_refine, labels,
+                    label_weights, bbox_gt_init, bbox_weights_init,
+                    bbox_gt_refine, bbox_weights_refine, stride,
+                    num_total_samples_init, num_total_samples_refine):
+        # classification loss
+        labels = labels.reshape(-1)
+        label_weights = label_weights.reshape(-1)
+        cls_score = cls_score.permute(0, 2, 3,
+                                      1).reshape(-1, self.cls_out_channels)
+        cls_score = cls_score.contiguous()
+        loss_cls = self.loss_cls(
+            cls_score,
+            labels,
+            label_weights,
+            avg_factor=num_total_samples_refine)
+
+        # points loss
+        bbox_gt_init = bbox_gt_init.reshape(-1, 4)
+        bbox_weights_init = bbox_weights_init.reshape(-1, 4)
+        bbox_pred_init = self.points2bbox(
+            pts_pred_init.reshape(-1, 2 * self.num_points), y_first=False)
+        bbox_gt_refine = bbox_gt_refine.reshape(-1, 4)
+        bbox_weights_refine = bbox_weights_refine.reshape(-1, 4)
+        bbox_pred_refine = self.points2bbox(
+            pts_pred_refine.reshape(-1, 2 * self.num_points), y_first=False)
+        normalize_term = self.point_base_scale * stride
+        loss_pts_init = self.loss_bbox_init(
+            bbox_pred_init / normalize_term,
+            bbox_gt_init / normalize_term,
+            bbox_weights_init,
+            avg_factor=num_total_samples_init)
+        loss_pts_refine = self.loss_bbox_refine(
+            bbox_pred_refine / normalize_term,
+            bbox_gt_refine / normalize_term,
+            bbox_weights_refine,
+            avg_factor=num_total_samples_refine)
+        return loss_cls, loss_pts_init, loss_pts_refine
+
+    def loss(self,
+             cls_scores,
+             pts_preds_init,
+             pts_preds_refine,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == len(self.point_generators)
+        device = cls_scores[0].device
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+
+        # target for initial stage
+        center_list, valid_flag_list = self.get_points(featmap_sizes,
+                                                       img_metas, device)
+        pts_coordinate_preds_init = self.offset_to_pts(center_list,
+                                                       pts_preds_init)
+        if self.train_cfg.init.assigner['type'] == 'PointAssigner':
+            # Assign target for center list
+            candidate_list = center_list
+        else:
+            # transform center list to bbox list and
+            #   assign target for bbox list
+            bbox_list = self.centers_to_bboxes(center_list)
+            candidate_list = bbox_list
+        cls_reg_targets_init = self.get_targets(
+            candidate_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            stage='init',
+            label_channels=label_channels)
+        (*_, bbox_gt_list_init, candidate_list_init, bbox_weights_list_init,
+         num_total_pos_init, num_total_neg_init) = cls_reg_targets_init
+        num_total_samples_init = (
+            num_total_pos_init +
+            num_total_neg_init if self.sampling else num_total_pos_init)
+
+        # target for refinement stage
+        center_list, valid_flag_list = self.get_points(featmap_sizes,
+                                                       img_metas, device)
+        pts_coordinate_preds_refine = self.offset_to_pts(
+            center_list, pts_preds_refine)
+        bbox_list = []
+        for i_img, center in enumerate(center_list):
+            bbox = []
+            for i_lvl in range(len(pts_preds_refine)):
+                bbox_preds_init = self.points2bbox(
+                    pts_preds_init[i_lvl].detach())
+                bbox_shift = bbox_preds_init * self.point_strides[i_lvl]
+                bbox_center = torch.cat(
+                    [center[i_lvl][:, :2], center[i_lvl][:, :2]], dim=1)
+                bbox.append(bbox_center +
+                            bbox_shift[i_img].permute(1, 2, 0).reshape(-1, 4))
+            bbox_list.append(bbox)
+        cls_reg_targets_refine = self.get_targets(
+            bbox_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            stage='refine',
+            label_channels=label_channels)
+        (labels_list, label_weights_list, bbox_gt_list_refine,
+         candidate_list_refine, bbox_weights_list_refine, num_total_pos_refine,
+         num_total_neg_refine) = cls_reg_targets_refine
+        num_total_samples_refine = (
+            num_total_pos_refine +
+            num_total_neg_refine if self.sampling else num_total_pos_refine)
+
+        # compute loss
+        losses_cls, losses_pts_init, losses_pts_refine = multi_apply(
+            self.loss_single,
+            cls_scores,
+            pts_coordinate_preds_init,
+            pts_coordinate_preds_refine,
+            labels_list,
+            label_weights_list,
+            bbox_gt_list_init,
+            bbox_weights_list_init,
+            bbox_gt_list_refine,
+            bbox_weights_list_refine,
+            self.point_strides,
+            num_total_samples_init=num_total_samples_init,
+            num_total_samples_refine=num_total_samples_refine)
+        loss_dict_all = {
+            'loss_cls': losses_cls,
+            'loss_pts_init': losses_pts_init,
+            'loss_pts_refine': losses_pts_refine
+        }
+        return loss_dict_all
+
+    def get_bboxes(self,
+                   cls_scores,
+                   pts_preds_init,
+                   pts_preds_refine,
+                   img_metas,
+                   cfg=None,
+                   rescale=False,
+                   with_nms=True):
+        assert len(cls_scores) == len(pts_preds_refine)
+        device = cls_scores[0].device
+        bbox_preds_refine = [
+            self.points2bbox(pts_pred_refine)
+            for pts_pred_refine in pts_preds_refine
+        ]
+        num_levels = len(cls_scores)
+        mlvl_points = [
+            self.point_generators[i].grid_points(cls_scores[i].size()[-2:],
+                                                 self.point_strides[i], device)
+            for i in range(num_levels)
+        ]
+        result_list = []
+        for img_id in range(len(img_metas)):
+            cls_score_list = [
+                cls_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_pred_list = [
+                bbox_preds_refine[i][img_id].detach()
+                for i in range(num_levels)
+            ]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            proposals = self._get_bboxes_single(cls_score_list, bbox_pred_list,
+                                                mlvl_points, img_shape,
+                                                scale_factor, cfg, rescale,
+                                                with_nms)
+            result_list.append(proposals)
+        return result_list
+
+    def _get_bboxes_single(self,
+                           cls_scores,
+                           bbox_preds,
+                           mlvl_points,
+                           img_shape,
+                           scale_factor,
+                           cfg,
+                           rescale=False,
+                           with_nms=True):
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_scores) == len(bbox_preds) == len(mlvl_points)
+        mlvl_bboxes = []
+        mlvl_scores = []
+        for i_lvl, (cls_score, bbox_pred, points) in enumerate(
+                zip(cls_scores, bbox_preds, mlvl_points)):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            cls_score = cls_score.permute(1, 2,
+                                          0).reshape(-1, self.cls_out_channels)
+            if self.use_sigmoid_cls:
+                scores = cls_score.sigmoid()
+            else:
+                scores = cls_score.softmax(-1)
+            bbox_pred = bbox_pred.permute(1, 2, 0).reshape(-1, 4)
+            nms_pre = cfg.get('nms_pre', -1)
+            if nms_pre > 0 and scores.shape[0] > nms_pre:
+                if self.use_sigmoid_cls:
+                    max_scores, _ = scores.max(dim=1)
+                else:
+                    # remind that we set FG labels to [0, num_class-1]
+                    # since mmdet v2.0
+                    # BG cat_id: num_class
+                    max_scores, _ = scores[:, :-1].max(dim=1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                points = points[topk_inds, :]
+                bbox_pred = bbox_pred[topk_inds, :]
+                scores = scores[topk_inds, :]
+            bbox_pos_center = torch.cat([points[:, :2], points[:, :2]], dim=1)
+            bboxes = bbox_pred * self.point_strides[i_lvl] + bbox_pos_center
+            x1 = bboxes[:, 0].clamp(min=0, max=img_shape[1])
+            y1 = bboxes[:, 1].clamp(min=0, max=img_shape[0])
+            x2 = bboxes[:, 2].clamp(min=0, max=img_shape[1])
+            y2 = bboxes[:, 3].clamp(min=0, max=img_shape[0])
+            bboxes = torch.stack([x1, y1, x2, y2], dim=-1)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+        mlvl_bboxes = torch.cat(mlvl_bboxes)
+        if rescale:
+            mlvl_bboxes /= mlvl_bboxes.new_tensor(scale_factor)
+        mlvl_scores = torch.cat(mlvl_scores)
+        if self.use_sigmoid_cls:
+            # Add a dummy background class to the backend when using sigmoid
+            # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+            # BG cat_id: num_class
+            padding = mlvl_scores.new_zeros(mlvl_scores.shape[0], 1)
+            mlvl_scores = torch.cat([mlvl_scores, padding], dim=1)
+        if with_nms:
+            det_bboxes, det_labels = multiclass_nms(mlvl_bboxes, mlvl_scores,
+                                                    cfg.score_thr, cfg.nms,
+                                                    cfg.max_per_img)
+            return det_bboxes, det_labels
+        else:
+            return mlvl_bboxes, mlvl_scores
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/reppoints_v2_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/reppoints_v2_head.py
new file mode 100644
index 0000000..270e612
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/reppoints_v2_head.py
@@ -0,0 +1,1318 @@
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, bias_init_with_prob, normal_init
+from mmcv.ops import DeformConv2d, CornerPool
+
+from mmdet.core import (PointGenerator, build_assigner, build_sampler,
+                        images_to_levels, multi_apply, multiclass_nms_rpd, unmap)
+from ..builder import HEADS, build_loss, build_head
+from .anchor_free_head import AnchorFreeHead
+
+from mmdet.utils.instances import Instances
+from mmdet.utils.common import compute_locations
+from itertools import accumulate
+
+class CornerPoolPack(nn.Module):
+    def __init__(self, dim, pool1, pool2, conv_cfg=None, norm_cfg=None, first_kernel_size=3, kernel_size=3, corner_dim=128):
+        super(CornerPoolPack, self).__init__()
+        self.p1_conv1 = ConvModule(
+            dim,
+            corner_dim,
+            first_kernel_size,
+            stride=1,
+            padding=(first_kernel_size - 1) // 2,
+            conv_cfg=conv_cfg,
+            norm_cfg=norm_cfg)
+        self.p2_conv1 = ConvModule(
+            dim,
+            corner_dim,
+            first_kernel_size,
+            stride=1,
+            padding=(first_kernel_size - 1) // 2,
+            conv_cfg=conv_cfg,
+            norm_cfg=norm_cfg)
+
+        self.p_conv1 = nn.Conv2d(corner_dim, dim, 3, padding=1, bias=False)
+        self.p_gn1   = nn.GroupNorm(num_groups=32, num_channels=dim)
+
+        self.conv1 = nn.Conv2d(dim, dim, 1, bias=False)
+        self.gn1   = nn.GroupNorm(num_groups=32, num_channels=dim)
+        self.relu1 = nn.ReLU(inplace=True)
+
+        self.conv2 = ConvModule(
+            dim,
+            dim,
+            kernel_size,
+            stride=1,
+            padding=(kernel_size - 1) // 2,
+            conv_cfg=conv_cfg,
+            norm_cfg=norm_cfg)
+
+        self.pool1 = pool1
+        self.pool2 = pool2
+
+    def forward(self, x):
+        # pool 1
+        p1_conv1 = self.p1_conv1(x)
+        pool1    = self.pool1(p1_conv1)
+
+        # pool 2
+        p2_conv1 = self.p2_conv1(x)
+        pool2    = self.pool2(p2_conv1)
+
+        # pool 1 + pool 2
+        p_conv1 = self.p_conv1(pool1 + pool2)
+        p_gn1   = self.p_gn1(p_conv1)
+
+        conv1 = self.conv1(x)
+        gn1   = self.gn1(conv1)
+        relu1 = self.relu1(p_gn1 + gn1)
+
+        conv2 = self.conv2(relu1)
+        return conv2
+
+
+class TLPool(CornerPoolPack):
+    def __init__(self, dim, conv_cfg=None, norm_cfg=None, first_kernel_size=3, kernel_size=3, corner_dim=128):
+        super(TLPool, self).__init__(dim, CornerPool('top'), CornerPool('left'), conv_cfg, norm_cfg, first_kernel_size, kernel_size, corner_dim)
+
+
+class BRPool(CornerPoolPack):
+    def __init__(self, dim, conv_cfg=None, norm_cfg=None, first_kernel_size=3, kernel_size=3, corner_dim=128):
+        super(BRPool, self).__init__(dim, CornerPool('bottom'), CornerPool('right'), conv_cfg, norm_cfg, first_kernel_size, kernel_size, corner_dim)
+
+
+@HEADS.register_module()
+class RepPointsV2Head(AnchorFreeHead):
+    """RepPointV2 head. With mask branch support
+
+    Args:
+        point_feat_channels (int): Number of channels of points features.
+        gradient_mul (float): The multiplier to gradients from
+            points refinement and recognition.
+        point_strides (Iterable): points strides.
+        point_base_scale (int): bbox scale for assigning labels.
+        loss_cls (dict): Config of classification loss.
+        loss_bbox_init (dict): Config of initial points loss.
+        loss_bbox_refine (dict): Config of points loss in refinement.
+        use_grid_points (bool): If we use bounding box representation, the
+        reppoints is represented as grid points on the bounding box.
+        center_init (bool): Whether to use center point assignment.
+        transform_method (str): The methods to transform RepPoints to bbox.
+    """  # noqa: W605
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 point_feat_channels=256,
+                 shared_stacked_convs=1,
+                 first_kernel_size=3,
+                 kernel_size=1,
+                 corner_dim=64,
+                 num_points=9,
+                 gradient_mul=0.1,
+                 point_strides=[8, 16, 32, 64, 128],
+                 point_base_scale=4,
+                 corner_refine=True,
+                 loss_cls=dict(
+                     type='FocalLoss',
+                     use_sigmoid=True,
+                     gamma=2.0,
+                     alpha=0.25,
+                     loss_weight=1.0),
+                 loss_bbox_init=dict(
+                     type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=0.5),
+                 loss_bbox_refine=dict(
+                     type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0),
+                 loss_heatmap=dict(
+                     type='GaussianFocalLoss',
+                     alpha=2.0,
+                     gamma=4.0,
+                     loss_weight=0.25),
+                 loss_offset=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0),
+                 loss_sem=dict(type='SEPFocalLoss', use_sigmoid=True, gamma=2.0, alpha=0.25, loss_weight=0.1),
+                 use_grid_points=False,
+                 center_init=True,
+                 transform_method='moment',
+                 moment_mul=0.01,
+                 mask_head=None,
+                 background_label=None,
+                 **kwargs):
+        self.background_label = (
+            num_classes if background_label is None else background_label)
+        # background_label should be either 0 or num_classes
+        assert (self.background_label == 0
+                or self.background_label == num_classes)
+        self.num_points = num_points
+        self.point_feat_channels = point_feat_channels
+        self.shared_stacked_convs = shared_stacked_convs
+        self.use_grid_points = use_grid_points
+        self.center_init = center_init
+
+        self.first_kernel_size = first_kernel_size
+        self.kernel_size = kernel_size
+        self.corner_dim = corner_dim
+        self.corner_refine = corner_refine
+
+        # we use deformable conv to extract points features
+        self.dcn_kernel = int(np.sqrt(num_points))
+        self.dcn_pad = int((self.dcn_kernel - 1) / 2)
+        assert self.dcn_kernel * self.dcn_kernel == num_points, \
+            'The points number should be a square number.'
+        assert self.dcn_kernel % 2 == 1, \
+            'The points number should be an odd square number.'
+        dcn_base = np.arange(-self.dcn_pad,
+                             self.dcn_pad + 1).astype(np.float64)
+        dcn_base_y = np.repeat(dcn_base, self.dcn_kernel)
+        dcn_base_x = np.tile(dcn_base, self.dcn_kernel)
+        dcn_base_offset = np.stack([dcn_base_y, dcn_base_x], axis=1).reshape(
+            (-1))
+        self.dcn_base_offset = torch.tensor(dcn_base_offset).view(1, -1, 1, 1)
+
+        self.controller_on = kwargs.pop('controller_on', 'cls')
+        self.coord_pos = kwargs.pop('coord_pos', 'grid')
+        if mask_head is not None and self.coord_pos in ['grid', 'center']:
+            mask_head.head_cfg.rel_num = 1
+        elif mask_head is not None and self.coord_pos in ['center-lt-rb']:
+            mask_head.head_cfg.rel_num = 3
+        super().__init__(num_classes, in_channels, loss_cls=loss_cls, **kwargs)
+
+        self.gradient_mul = gradient_mul
+        self.point_base_scale = point_base_scale
+        self.point_strides = point_strides
+        self.point_generators = [PointGenerator() for _ in self.point_strides]
+
+        if self.train_cfg:
+            self.init_assigner = build_assigner(self.train_cfg.init.assigner)
+            self.refine_assigner = build_assigner(self.train_cfg.refine.assigner)
+            self.hm_assigner = build_assigner(self.train_cfg.heatmap.assigner)
+            # use PseudoSampler when sampling is False
+            sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+        self.transform_method = transform_method
+        if self.transform_method == 'moment':
+            self.moment_transfer = nn.Parameter(data=torch.zeros(2), requires_grad=True)
+            self.moment_mul = moment_mul
+
+        self.cls_out_channels = self.num_classes
+        self.loss_bbox_init = build_loss(loss_bbox_init)
+        self.loss_bbox_refine = build_loss(loss_bbox_refine)
+        self.loss_heatmap = build_loss(loss_heatmap)
+        self.loss_offset = build_loss(loss_offset)
+        self.loss_sem = build_loss(loss_sem)
+        
+        # mask
+        self.mask_head = mask_head
+        if mask_head is not None:
+            self.mask_head = build_head(mask_head)
+
+            self.controller = nn.Conv2d(
+                point_feat_channels, self.mask_head.head.num_gen_params,
+                kernel_size=3, stride=1, padding=1
+            )
+            torch.nn.init.normal_(self.controller.weight, std=0.01)
+            torch.nn.init.constant_(self.controller.bias, 0)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.relu = nn.ReLU(inplace=True)
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        self.shared_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            self.cls_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+            self.reg_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+
+        for i in range(self.shared_stacked_convs):
+            self.shared_convs.append(
+                ConvModule(
+                    self.feat_channels,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+
+        self.hem_tl = TLPool(self.feat_channels, self.conv_cfg, self.norm_cfg, first_kernel_size=self.first_kernel_size, kernel_size=self.kernel_size, corner_dim=self.corner_dim)
+        self.hem_br = BRPool(self.feat_channels, self.conv_cfg, self.norm_cfg, first_kernel_size=self.first_kernel_size, kernel_size=self.kernel_size, corner_dim=self.corner_dim)
+
+        pts_out_dim = 4 if self.use_grid_points else 2 * self.num_points
+
+        cls_in_channels = self.feat_channels + 6
+        self.reppoints_cls_conv = DeformConv2d(cls_in_channels,
+                                             self.point_feat_channels,
+                                             self.dcn_kernel, 1, self.dcn_pad)
+        self.reppoints_cls_out = nn.Conv2d(self.point_feat_channels,
+                                           self.cls_out_channels, 1, 1, 0)
+
+        self.reppoints_pts_init_conv = nn.Conv2d(self.feat_channels,
+                                                 self.point_feat_channels, 3, 1, 1)
+        self.reppoints_pts_init_out = nn.Conv2d(self.point_feat_channels,
+                                                pts_out_dim, 1, 1, 0)
+        pts_in_channels = self.feat_channels + 6
+        self.reppoints_pts_refine_conv = DeformConv2d(pts_in_channels,
+                                                    self.point_feat_channels,
+                                                    self.dcn_kernel, 1,
+                                                    self.dcn_pad)
+        self.reppoints_pts_refine_out = nn.Conv2d(self.point_feat_channels,
+                                                  pts_out_dim, 1, 1, 0)
+
+        self.reppoints_hem_tl_score_out = nn.Conv2d(self.feat_channels, 1, 3, 1, 1)
+        self.reppoints_hem_br_score_out = nn.Conv2d(self.feat_channels, 1, 3, 1, 1)
+        self.reppoints_hem_tl_offset_out = nn.Conv2d(self.feat_channels, 2, 3, 1, 1)
+        self.reppoints_hem_br_offset_out = nn.Conv2d(self.feat_channels, 2, 3, 1, 1)
+
+        self.reppoints_sem_out = nn.Conv2d(self.feat_channels, self.cls_out_channels, 1, 1, 0)
+        self.reppoints_sem_embedding = ConvModule(
+            self.feat_channels,
+            self.feat_channels,
+            1,
+            conv_cfg=self.conv_cfg,
+            norm_cfg=self.norm_cfg)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.cls_convs:
+            normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            normal_init(m.conv, std=0.01)
+        for m in self.shared_convs:
+            normal_init(m.conv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.reppoints_cls_conv, std=0.01)
+        normal_init(self.reppoints_cls_out, std=0.01, bias=bias_cls)
+        normal_init(self.reppoints_pts_init_conv, std=0.01)
+        normal_init(self.reppoints_pts_init_out, std=0.01)
+        normal_init(self.reppoints_pts_refine_conv, std=0.01)
+        normal_init(self.reppoints_pts_refine_out, std=0.01)
+        normal_init(self.reppoints_hem_tl_score_out, std=0.01, bias=bias_cls)
+        normal_init(self.reppoints_hem_tl_offset_out, std=0.01)
+        normal_init(self.reppoints_hem_br_score_out, std=0.01, bias=bias_cls)
+        normal_init(self.reppoints_hem_br_offset_out, std=0.01)
+        normal_init(self.reppoints_sem_out, std=0.01, bias=bias_cls)
+
+    def points2bbox(self, pts, y_first=True):
+        """Converting the points set into bounding box.
+
+        :param pts: the input points sets (fields), each points
+            set (fields) is represented as 2n scalar.
+        :param y_first: if y_fisrt=True, the point set is represented as
+            [y1, x1, y2, x2 ... yn, xn], otherwise the point set is
+            represented as [x1, y1, x2, y2 ... xn, yn].
+        :return: each points set is converting to a bbox [x1, y1, x2, y2].
+        """
+        pts_reshape = pts.view(pts.shape[0], -1, 2, *pts.shape[2:])
+        pts_y = pts_reshape[:, :, 0, ...] if y_first else pts_reshape[:, :, 1, ...]
+        pts_x = pts_reshape[:, :, 1, ...] if y_first else pts_reshape[:, :, 0, ...]
+        if self.transform_method == 'minmax':
+            bbox_left = pts_x.min(dim=1, keepdim=True)[0]
+            bbox_right = pts_x.max(dim=1, keepdim=True)[0]
+            bbox_up = pts_y.min(dim=1, keepdim=True)[0]
+            bbox_bottom = pts_y.max(dim=1, keepdim=True)[0]
+            bbox = torch.cat([bbox_left, bbox_up, bbox_right, bbox_bottom],
+                             dim=1)
+        elif self.transform_method == 'partial_minmax':
+            pts_y = pts_y[:, :4, ...]
+            pts_x = pts_x[:, :4, ...]
+            bbox_left = pts_x.min(dim=1, keepdim=True)[0]
+            bbox_right = pts_x.max(dim=1, keepdim=True)[0]
+            bbox_up = pts_y.min(dim=1, keepdim=True)[0]
+            bbox_bottom = pts_y.max(dim=1, keepdim=True)[0]
+            bbox = torch.cat([bbox_left, bbox_up, bbox_right, bbox_bottom],
+                             dim=1)
+        elif self.transform_method == 'moment':
+            pts_y_mean = pts_y.mean(dim=1, keepdim=True)
+            pts_x_mean = pts_x.mean(dim=1, keepdim=True)
+            pts_y_std = torch.std(pts_y - pts_y_mean, dim=1, keepdim=True)
+            pts_x_std = torch.std(pts_x - pts_x_mean, dim=1, keepdim=True)
+            moment_transfer = (self.moment_transfer * self.moment_mul) + (
+                self.moment_transfer.detach() * (1 - self.moment_mul))
+            moment_width_transfer = moment_transfer[0]
+            moment_height_transfer = moment_transfer[1]
+            half_width = pts_x_std * torch.exp(moment_width_transfer)
+            half_height = pts_y_std * torch.exp(moment_height_transfer)
+            bbox = torch.cat([
+                pts_x_mean - half_width, pts_y_mean - half_height,
+                pts_x_mean + half_width, pts_y_mean + half_height
+            ],
+                             dim=1)
+        elif self.transform_method == "exact_minmax":
+            pts_reshape = pts.view(pts.shape[0], -1, 2, *pts.shape[2:])
+            pts_reshape = pts_reshape[:, :2, ...]
+            pts_y = pts_reshape[:, :, 0, ...] if y_first else pts_reshape[:, :, 1, ...]
+            pts_x = pts_reshape[:, :, 1, ...] if y_first else pts_reshape[:, :, 0, ...]
+            bbox_left = pts_x[:, 0:1, ...]
+            bbox_right = pts_x[:, 1:2, ...]
+            bbox_up = pts_y[:, 0:1, ...]
+            bbox_bottom = pts_y[:, 1:2, ...]
+            bbox = torch.cat([bbox_left, bbox_up, bbox_right, bbox_bottom], dim=1)
+        else:
+            raise NotImplementedError
+        return bbox
+
+    def gen_grid_from_reg(self, reg, previous_boxes):
+        """Base on the previous bboxes and regression values, we compute the
+        regressed bboxes and generate the grids on the bboxes.
+
+        :param reg: the regression value to previous bboxes.
+        :param previous_boxes: previous bboxes.
+        :return: generate grids on the regressed bboxes.
+        """
+        b, _, h, w = reg.shape
+        bxy = (previous_boxes[:, :2, ...] + previous_boxes[:, 2:, ...]) / 2.
+        bwh = (previous_boxes[:, 2:, ...] -
+               previous_boxes[:, :2, ...]).clamp(min=1e-6)
+        grid_topleft = bxy + bwh * reg[:, :2, ...] - 0.5 * bwh * torch.exp(
+            reg[:, 2:, ...])
+        grid_wh = bwh * torch.exp(reg[:, 2:, ...])
+        grid_left = grid_topleft[:, [0], ...]
+        grid_top = grid_topleft[:, [1], ...]
+        grid_width = grid_wh[:, [0], ...]
+        grid_height = grid_wh[:, [1], ...]
+        intervel = torch.linspace(0., 1., self.dcn_kernel).view(
+            1, self.dcn_kernel, 1, 1).type_as(reg)
+        grid_x = grid_left + grid_width * intervel
+        grid_x = grid_x.unsqueeze(1).repeat(1, self.dcn_kernel, 1, 1, 1)
+        grid_x = grid_x.view(b, -1, h, w)
+        grid_y = grid_top + grid_height * intervel
+        grid_y = grid_y.unsqueeze(2).repeat(1, 1, self.dcn_kernel, 1, 1)
+        grid_y = grid_y.view(b, -1, h, w)
+        grid_yx = torch.stack([grid_y, grid_x], dim=2)
+        grid_yx = grid_yx.view(b, -1, h, w)
+        regressed_bbox = torch.cat([
+            grid_left, grid_top, grid_left + grid_width, grid_top + grid_height
+        ], 1)
+        return grid_yx, regressed_bbox
+
+    def compute_coord_pos(self, points):
+        # points: N, HW, C
+        x, y = points[..., 0::2], points[..., 1::2]
+        out_x, out_y = None, None
+        if self.coord_pos == 'center':
+            out_x, out_y = x.mean(-1), y.mean(-1)
+        else:
+            assert 0
+        return torch.stack([out_x.flatten(), out_y.flatten()], dim=-1)
+
+    def forward(self, feats):
+        if not self.mask_head:
+            return multi_apply(self.forward_single, feats)
+        else:
+            self.pred_instances = Instances((0,0)) # just a dumy image shape
+            bbox_res = multi_apply(self.forward_single, feats, controller=self.controller) # cls_out, pts_out_init, pts_out_refine, controller_out
+            device = bbox_res[1][0].device
+            self.level_sum = tuple(accumulate([0]+[r[:,0].numel() for r in bbox_res[1]][:-1]))
+            self.image_sum = torch.tensor([r[0,0].numel() for r in bbox_res[1]])
+            self.pred_instances.fpn_levels = torch.cat([torch.tensor([i]*r[:,0].numel()).to(device) for i,r in enumerate(bbox_res[1])])
+            self.pred_instances.im_inds = torch.cat([torch.cat([torch.tensor([i]*img.numel()).to(device) for i,img in enumerate(r[:,0])]) for r in bbox_res[1]])
+            
+            pts_grid = torch.cat([compute_locations(*r.shape[-2:], self.point_strides[level], device).reshape(-1,2).repeat(len(r),1) for level, r in enumerate(bbox_res[1])], dim=0)
+            pts_center_list = []
+            for i_lvl in range(len(self.point_strides)):
+                pts_shift = bbox_res[1][i_lvl]
+                pts_shift = pts_shift.permute(0, 2, 3, 1)
+                yx_pts_shift = pts_shift.reshape(*pts_shift.shape[:-1], 2 * self.num_points)
+                y_pts_shift = yx_pts_shift[..., 0::2]
+                x_pts_shift = yx_pts_shift[..., 1::2]
+                xy_pts_shift = torch.stack([x_pts_shift, y_pts_shift], -1)
+                pts = xy_pts_shift * self.point_strides[i_lvl] # N, H, W, np, 2
+                pts_center = pts.mean(-2)
+                pts_center_list.append(pts_center)
+            self.pred_instances.locations = torch.cat([l.flatten(0, -2) for l in pts_center_list]) + pts_grid
+
+            self.pred_instances.mask_head_params = torch.cat([r.permute(0,2,3,1).flatten(0,-2) for r in bbox_res[-1]])
+            return bbox_res[:-1]
+
+    def forward_single(self, x, controller=None):
+        """ Forward feature map of a single FPN level."""
+        dcn_base_offset = self.dcn_base_offset.type_as(x)
+        # If we use center_init, the initial reppoints is from center points.
+        # If we use bounding bbox representation, the initial reppoints is
+        #   from regular grid placed on a pre-defined bbox.
+        if self.use_grid_points or not self.center_init:
+            scale = self.point_base_scale / 2
+            points_init = dcn_base_offset / dcn_base_offset.max() * scale
+            bbox_init = x.new_tensor([-scale, -scale, scale,
+                                      scale]).view(1, 4, 1, 1)
+        else:
+            points_init = 0
+        cls_feat = x
+        pts_feat = x
+        for cls_conv in self.cls_convs:
+            cls_feat = cls_conv(cls_feat)
+        for reg_conv in self.reg_convs:
+            pts_feat = reg_conv(pts_feat)
+
+        shared_feat = pts_feat
+        for shared_conv in self.shared_convs:
+            shared_feat = shared_conv(shared_feat)
+
+        sem_feat = shared_feat
+        hem_feat = shared_feat
+
+        sem_scores_out = self.reppoints_sem_out(sem_feat)
+        sem_feat = self.reppoints_sem_embedding(sem_feat)
+
+        cls_feat = cls_feat + sem_feat
+        pts_feat = pts_feat + sem_feat
+        hem_feat = hem_feat + sem_feat
+
+        # generate heatmap and offset
+        hem_tl_feat = self.hem_tl(hem_feat)
+        hem_br_feat = self.hem_br(hem_feat)
+
+        hem_tl_score_out = self.reppoints_hem_tl_score_out(hem_tl_feat)
+        hem_tl_offset_out = self.reppoints_hem_tl_offset_out(hem_tl_feat)
+        hem_br_score_out = self.reppoints_hem_br_score_out(hem_br_feat)
+        hem_br_offset_out = self.reppoints_hem_br_offset_out(hem_br_feat)
+
+        hem_score_out = torch.cat([hem_tl_score_out, hem_br_score_out], dim=1)
+        hem_offset_out = torch.cat([hem_tl_offset_out, hem_br_offset_out], dim=1)
+
+        # initialize reppoints
+        # pts_out_init = self.reppoints_pts_init_out(self.relu(self.reppoints_pts_init_conv(pts_feat)))
+        init_reg_tower = self.relu(self.reppoints_pts_init_conv(pts_feat))
+        pts_out_init = self.reppoints_pts_init_out(init_reg_tower)
+        if self.controller_on == 'init':
+            controller_out = controller(init_reg_tower)
+        if self.use_grid_points:
+            pts_out_init, bbox_out_init = self.gen_grid_from_reg(pts_out_init, bbox_init.detach())
+        else:
+            pts_out_init = pts_out_init + points_init
+        # refine and classify reppoints
+        pts_out_init_grad_mul = (1 - self.gradient_mul) * pts_out_init.detach() + self.gradient_mul * pts_out_init
+        dcn_offset = pts_out_init_grad_mul - dcn_base_offset
+
+        hem_feat = torch.cat([hem_score_out, hem_offset_out], dim=1)
+        cls_feat = torch.cat([cls_feat, hem_feat], dim=1)
+        pts_feat = torch.cat([pts_feat, hem_feat], dim=1)
+
+        cls_tower = self.relu(self.reppoints_cls_conv(cls_feat, dcn_offset))
+        cls_out = self.reppoints_cls_out(cls_tower)
+        if self.mask_head and controller is not None and self.controller_on == 'cls':
+            controller_out = controller(cls_tower)
+        refine_reg_tower = self.relu(self.reppoints_pts_refine_conv(pts_feat, dcn_offset))
+        pts_out_refine = self.reppoints_pts_refine_out(refine_reg_tower)
+        if self.controller_on == 'refine':
+            controller_out = controller(refine_reg_tower)
+        if self.use_grid_points:
+            pts_out_refine, bbox_out_refine = self.gen_grid_from_reg(pts_out_refine, bbox_out_init.detach())
+        else:
+            pts_out_refine = pts_out_refine + pts_out_init.detach()
+        
+        if self.mask_head:
+            return cls_out, pts_out_init, pts_out_refine, hem_score_out, hem_offset_out, sem_scores_out, controller_out
+        else:
+            return cls_out, pts_out_init, pts_out_refine, hem_score_out, hem_offset_out, sem_scores_out
+
+    def get_points(self, featmap_sizes, img_metas):
+        """Get points according to feature map sizes.
+
+        Args:
+            featmap_sizes (list[tuple]): Multi-level feature map sizes.
+            img_metas (list[dict]): Image meta info.
+
+        Returns:
+            tuple: points of each image, valid flags of each image
+        """
+        num_imgs = len(img_metas)
+        num_levels = len(featmap_sizes)
+
+        # since feature map sizes of all images are the same, we only compute
+        # points center for one time
+        multi_level_points = []
+        for i in range(num_levels):
+            points = self.point_generators[i].grid_points(
+                featmap_sizes[i], self.point_strides[i])
+            multi_level_points.append(points)
+        points_list = [[point.clone() for point in multi_level_points] for _ in range(num_imgs)]
+
+        # for each image, we compute valid flags of multi level grids
+        valid_flag_list = []
+        for img_id, img_meta in enumerate(img_metas):
+            multi_level_flags = []
+            for i in range(num_levels):
+                point_stride = self.point_strides[i]
+                feat_h, feat_w = featmap_sizes[i]
+                h, w = img_meta['pad_shape'][:2]
+                valid_feat_h = min(int(np.ceil(h / point_stride)), feat_h)
+                valid_feat_w = min(int(np.ceil(w / point_stride)), feat_w)
+                flags = self.point_generators[i].valid_flags(
+                    (feat_h, feat_w), (valid_feat_h, valid_feat_w))
+                multi_level_flags.append(flags)
+            valid_flag_list.append(multi_level_flags)
+
+        return points_list, valid_flag_list
+
+    def centers_to_bboxes(self, point_list):
+        """Get bboxes according to center points. Only used in MaxIOUAssigner.
+        """
+        bbox_list = []
+        for i_img, point in enumerate(point_list):
+            bbox = []
+            for i_lvl in range(len(self.point_strides)):
+                scale = self.point_base_scale * self.point_strides[i_lvl] * 0.5
+                bbox_shift = torch.Tensor([-scale, -scale, scale, scale]).view(1, 4).type_as(point[0])
+                bbox_center = torch.cat([point[i_lvl][:, :2], point[i_lvl][:, :2]], dim=1)
+                bbox.append(bbox_center + bbox_shift)
+            bbox_list.append(bbox)
+        return bbox_list
+
+    def offset_to_pts(self, center_list, pred_list):
+        """Change from point offset to point coordinate."""
+        pts_list = []
+        for i_lvl in range(len(self.point_strides)):
+            pts_lvl = []
+            for i_img in range(len(center_list)):
+                pts_center = center_list[i_img][i_lvl][:, :2].repeat(
+                    1, self.num_points)
+                pts_shift = pred_list[i_lvl][i_img]
+                yx_pts_shift = pts_shift.permute(1, 2, 0).view(
+                    -1, 2 * self.num_points)
+                y_pts_shift = yx_pts_shift[..., 0::2]
+                x_pts_shift = yx_pts_shift[..., 1::2]
+                xy_pts_shift = torch.stack([x_pts_shift, y_pts_shift], -1)
+                xy_pts_shift = xy_pts_shift.view(*yx_pts_shift.shape[:-1], -1)
+                pts = xy_pts_shift * self.point_strides[i_lvl] + pts_center
+                pts_lvl.append(pts)
+            pts_lvl = torch.stack(pts_lvl, 0)
+            pts_list.append(pts_lvl)
+        return pts_list
+
+    def _point_target_single(self,
+                             flat_proposals,
+                             valid_flags,
+                             num_level_proposals,
+                             gt_bboxes,
+                             gt_bboxes_ignore,
+                             gt_labels,
+                             gt_masks=None,
+                             label_channels=1,
+                             stage='init',
+                             unmap_outputs=True):
+        inside_flags = valid_flags
+        if not inside_flags.any():
+            return (None, ) * 6
+        # assign gt and sample proposals
+        proposals = flat_proposals[inside_flags, :]
+
+        num_level_proposals_inside = self.get_num_level_proposals_inside(num_level_proposals, inside_flags)
+        if stage == 'init':
+            assigner = self.init_assigner
+            assigner_type = self.train_cfg.init.assigner.type
+            pos_weight = self.train_cfg.init.pos_weight
+        else:
+            assigner = self.refine_assigner
+            assigner_type = self.train_cfg.refine.assigner.type
+            pos_weight = self.train_cfg.refine.pos_weight
+        if assigner_type != "ATSSAssignerV2":
+            assign_result = assigner.assign(proposals, gt_bboxes, gt_bboxes_ignore, gt_labels, gt_masks)
+        else:
+            assign_result = assigner.assign(proposals, num_level_proposals_inside, gt_bboxes, gt_bboxes_ignore, gt_labels, gt_masks)
+        sampling_result = self.sampler.sample(assign_result, proposals, gt_bboxes)
+
+        num_valid_proposals = proposals.shape[0]
+        bbox_gt = proposals.new_zeros([num_valid_proposals, 4])
+        bbox_weights = proposals.new_zeros([num_valid_proposals, 4])
+        labels = proposals.new_full((num_valid_proposals, ), self.background_label, dtype=torch.long)
+        label_weights = proposals.new_zeros(num_valid_proposals, dtype=torch.float)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+        if len(pos_inds) > 0:
+            pos_gt_bboxes = sampling_result.pos_gt_bboxes
+            bbox_gt[pos_inds, :] = pos_gt_bboxes
+            bbox_weights[pos_inds, :] = 1.0
+            if gt_labels is None:
+                labels[pos_inds] = 1
+            else:
+                labels[pos_inds] = gt_labels[sampling_result.pos_assigned_gt_inds]
+            if pos_weight <= 0:
+                label_weights[pos_inds] = 1.0
+            else:
+                label_weights[pos_inds] = pos_weight
+        if len(neg_inds) > 0:
+            label_weights[neg_inds] = 1.0
+
+        # map up to original set of proposals
+        if unmap_outputs:
+            num_total_proposals = flat_proposals.size(0)
+            labels = unmap(labels, num_total_proposals, inside_flags)
+            label_weights = unmap(label_weights, num_total_proposals, inside_flags)
+            bbox_gt = unmap(bbox_gt, num_total_proposals, inside_flags)
+            bbox_weights = unmap(bbox_weights, num_total_proposals, inside_flags)
+            gt_inds = unmap(
+                sampling_result.assign_result.gt_inds, num_total_proposals, inside_flags)
+
+        return labels, label_weights, bbox_gt, bbox_weights, pos_inds, neg_inds, gt_inds
+
+    def get_targets(self,
+                    proposals_list,
+                    valid_flag_list,
+                    gt_bboxes_list,
+                    img_metas,
+                    gt_bboxes_ignore_list=None,
+                    gt_labels_list=None,
+                    stage='init',
+                    label_channels=1,
+                    unmap_outputs=True,
+                    gt_masks=None):
+        """Compute corresponding GT box and classification targets for
+        proposals.
+
+        Args:
+            proposals_list (list[list]): Multi level points/bboxes of each
+                image.
+            valid_flag_list (list[list]): Multi level valid flags of each
+                image.
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image.
+            img_metas (list[dict]): Meta info of each image.
+            gt_bboxes_ignore_list (list[Tensor]): Ground truth bboxes to be
+                ignored.
+            gt_bboxes_list (list[Tensor]): Ground truth labels of each box.
+            stage (str): `init` or `refine`. Generate target for init stage or
+                refine stage
+            label_channels (int): Channel of label.
+            unmap_outputs (bool): Whether to map outputs back to the original
+                set of anchors.
+
+        Returns:
+            tuple:
+                - labels_list (list[Tensor]): Labels of each level.
+                - label_weights_list (list[Tensor]): Label weights of each level.  # noqa: E501
+                - bbox_gt_list (list[Tensor]): Ground truth bbox of each level.
+                - proposal_list (list[Tensor]): Proposals(points/bboxes) of each level.  # noqa: E501
+                - proposal_weights_list (list[Tensor]): Proposal weights of each level.  # noqa: E501
+                - num_total_pos (int): Number of positive samples in all images.  # noqa: E501
+                - num_total_neg (int): Number of negative samples in all images.  # noqa: E501
+        """
+        assert stage in ['init', 'refine']
+        num_imgs = len(img_metas)
+        assert len(proposals_list) == len(valid_flag_list) == num_imgs
+
+        # points number of multi levels
+        num_level_proposals = [points.size(0) for points in proposals_list[0]]
+        num_level_proposals_list = [num_level_proposals] * num_imgs
+
+        # concat all level points and flags to a single tensor
+        for i in range(num_imgs):
+            assert len(proposals_list[i]) == len(valid_flag_list[i])
+            proposals_list[i] = torch.cat(proposals_list[i])
+            valid_flag_list[i] = torch.cat(valid_flag_list[i])
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        (all_labels, all_label_weights, all_bbox_gt, all_bbox_weights,
+         pos_inds_list, neg_inds_list, gt_inds) = multi_apply(
+             self._point_target_single,
+             proposals_list,
+             valid_flag_list,
+             num_level_proposals_list,
+             gt_bboxes_list,
+             gt_bboxes_ignore_list,
+             gt_labels_list,
+             gt_masks if self.mask_head else [None] * len(proposals_list),
+             stage=stage,
+             label_channels=label_channels,
+             unmap_outputs=unmap_outputs)
+        # no valid points
+        if any([labels is None for labels in all_labels]):
+            return None
+        # sampled points of all images
+        num_total_pos = sum([max(inds.numel(), 1) for inds in pos_inds_list])
+        num_total_neg = sum([max(inds.numel(), 1) for inds in neg_inds_list])
+        labels_list = images_to_levels(all_labels, num_level_proposals)
+        label_weights_list = images_to_levels(all_label_weights, num_level_proposals)
+        bbox_gt_list = images_to_levels(all_bbox_gt, num_level_proposals)
+        bbox_weights_list = images_to_levels(all_bbox_weights,
+                                                 num_level_proposals)
+        shift_gt_inds = [] # shift gt_inds by previous mask number
+        if self.mask_head:
+            for gt_ind,nsum in zip(gt_inds,self.masknum_sum):
+                gt_ind[gt_ind>0] += nsum
+                shift_gt_inds.append(gt_ind)
+            shift_gt_inds = images_to_levels(shift_gt_inds, num_level_proposals)
+        return (labels_list, label_weights_list, bbox_gt_list, bbox_weights_list,
+                num_total_pos, num_total_neg, shift_gt_inds)
+
+    def _hm_target_single(self,
+                          flat_points,
+                          inside_flags,
+                          gt_bboxes,
+                          gt_labels,
+                          unmap_outputs=True):
+        # assign gt and sample points
+        if not inside_flags.any():
+            return (None, ) * 12
+        points = flat_points[inside_flags, :]
+
+        assigner = self.hm_assigner
+        gt_hm_tl, gt_offset_tl, pos_inds_tl, neg_inds_tl, \
+        gt_hm_br, gt_offset_br, pos_inds_br, neg_inds_br = \
+            assigner.assign(points, gt_bboxes, gt_labels)
+
+        num_valid_points = points.shape[0]
+        hm_tl_weights = points.new_zeros(num_valid_points, dtype=torch.float)
+        hm_br_weights = points.new_zeros(num_valid_points, dtype=torch.float)
+        offset_tl_weights = points.new_zeros([num_valid_points, 2], dtype=torch.float)
+        offset_br_weights = points.new_zeros([num_valid_points, 2], dtype=torch.float)
+
+        hm_tl_weights[pos_inds_tl] = 1.0
+        hm_tl_weights[neg_inds_tl] = 1.0
+        offset_tl_weights[pos_inds_tl, :] = 1.0
+
+        hm_br_weights[pos_inds_br] = 1.0
+        hm_br_weights[neg_inds_br] = 1.0
+        offset_br_weights[pos_inds_br, :] = 1.0
+
+        # map up to original set of grids
+        if unmap_outputs:
+            num_total_points = flat_points.shape[0]
+            gt_hm_tl = unmap(gt_hm_tl, num_total_points, inside_flags)
+            gt_offset_tl = unmap(gt_offset_tl, num_total_points, inside_flags)
+            hm_tl_weights = unmap(hm_tl_weights, num_total_points, inside_flags)
+            offset_tl_weights = unmap(offset_tl_weights, num_total_points, inside_flags)
+
+            gt_hm_br = unmap(gt_hm_br, num_total_points, inside_flags)
+            gt_offset_br = unmap(gt_offset_br, num_total_points, inside_flags)
+            hm_br_weights = unmap(hm_br_weights, num_total_points, inside_flags)
+            offset_br_weights = unmap(offset_br_weights, num_total_points, inside_flags)
+
+        return (gt_hm_tl, gt_offset_tl, hm_tl_weights, offset_tl_weights, pos_inds_tl, neg_inds_tl,
+                gt_hm_br, gt_offset_br, hm_br_weights, offset_br_weights, pos_inds_br, neg_inds_br)
+
+    def get_hm_targets(self,
+                       proposals_list,
+                       valid_flag_list,
+                       gt_bboxes_list,
+                       img_metas,
+                       gt_labels_list=None,
+                       unmap_outputs=True):
+        """Compute refinement and classification targets for points.
+
+        Args:
+            points_list (list[list]): Multi level points of each image.
+            valid_flag_list (list[list]): Multi level valid flags of each image.
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image.
+            img_metas (list[dict]): Meta info of each image.
+            cfg (dict): train sample configs.
+
+        Returns:
+            tuple
+        """
+        num_imgs = len(img_metas)
+        assert len(proposals_list) == len(valid_flag_list) == num_imgs
+
+        # points number of multi levels
+        num_level_proposals = [points.size(0) for points in proposals_list[0]]
+
+        # concat all level points and flags to a single tensor
+        for i in range(len(proposals_list)):
+            assert len(proposals_list[i]) == len(valid_flag_list[i])
+            proposals_list[i] = torch.cat(proposals_list[i])
+            valid_flag_list[i] = torch.cat(valid_flag_list[i])
+
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        (all_gt_hm_tl, all_gt_offset_tl, all_hm_tl_weights, all_offset_tl_weights, pos_inds_tl_list, neg_inds_tl_list,
+         all_gt_hm_br, all_gt_offset_br, all_hm_br_weights, all_offset_br_weights, pos_inds_br_list, neg_inds_br_list) = \
+            multi_apply(
+                self._hm_target_single,
+                proposals_list,
+                valid_flag_list,
+                gt_bboxes_list,
+                gt_labels_list,
+                unmap_outputs=unmap_outputs)
+        # no valid points
+        if any([gt_hm_tl is None for gt_hm_tl in all_gt_hm_tl]):
+            return None
+        # sampled points of all images
+        num_total_pos_tl = sum([max(inds.numel(), 1) for inds in pos_inds_tl_list])
+        num_total_neg_tl = sum([max(inds.numel(), 1) for inds in neg_inds_tl_list])
+        num_total_pos_br = sum([max(inds.numel(), 1) for inds in pos_inds_br_list])
+        num_total_neg_br = sum([max(inds.numel(), 1) for inds in neg_inds_br_list])
+
+        gt_hm_tl_list = images_to_levels(all_gt_hm_tl, num_level_proposals)
+        gt_offset_tl_list = images_to_levels(all_gt_offset_tl, num_level_proposals)
+        hm_tl_weight_list = images_to_levels(all_hm_tl_weights, num_level_proposals)
+        offset_tl_weight_list = images_to_levels(all_offset_tl_weights, num_level_proposals)
+
+        gt_hm_br_list = images_to_levels(all_gt_hm_br, num_level_proposals)
+        gt_offset_br_list = images_to_levels(all_gt_offset_br, num_level_proposals)
+        hm_br_weight_list = images_to_levels(all_hm_br_weights, num_level_proposals)
+        offset_br_weight_list = images_to_levels(all_offset_br_weights, num_level_proposals)
+
+        return (gt_hm_tl_list, gt_offset_tl_list, hm_tl_weight_list, offset_tl_weight_list,
+                gt_hm_br_list, gt_offset_br_list, hm_br_weight_list, offset_br_weight_list,
+                num_total_pos_tl, num_total_neg_tl, num_total_pos_br, num_total_neg_br)
+
+    def loss_single(self, cls_score, pts_pred_init, pts_pred_refine, hm_score, hm_offset,
+                    labels, label_weights,
+                    bbox_gt_init, bbox_weights_init,
+                    bbox_gt_refine, bbox_weights_refine,
+                    gt_hm_tl, gt_offset_tl, gt_hm_tl_weight, gt_offset_tl_weight,
+                    gt_hm_br, gt_offset_br, gt_hm_br_weight, gt_offset_br_weight,
+                    stride,
+                    num_total_samples_init, num_total_samples_refine,
+                    num_total_samples_tl, num_total_samples_br):
+        # classification loss
+        labels = labels.reshape(-1)
+        label_weights = label_weights.reshape(-1)
+        cls_score = cls_score.permute(0, 2, 3, 1).reshape(-1, self.cls_out_channels)
+
+
+        # points loss
+        bbox_gt_init = bbox_gt_init.reshape(-1, 4)
+        bbox_weights_init = bbox_weights_init.reshape(-1, 4)
+        bbox_pred_init = self.points2bbox(pts_pred_init.reshape(-1, 2 * self.num_points), y_first=False)
+        
+        bbox_gt_refine = bbox_gt_refine.reshape(-1, 4)
+        bbox_weights_refine = bbox_weights_refine.reshape(-1, 4)
+        bbox_pred_refine = self.points2bbox(pts_pred_refine.reshape(-1, 2 * self.num_points), y_first=False)
+        normalize_term = self.point_base_scale * stride
+        loss_pts_init = self.loss_bbox_init(
+            bbox_pred_init / normalize_term,
+            bbox_gt_init / normalize_term,
+            bbox_weights_init,
+            avg_factor=num_total_samples_init)
+        loss_pts_refine = self.loss_bbox_refine(
+            bbox_pred_refine / normalize_term,
+            bbox_gt_refine / normalize_term,
+            bbox_weights_refine,
+            avg_factor=num_total_samples_refine)
+        #cls loss
+        if hasattr(self.loss_cls,'requires_box') :
+            loss_cls = self.loss_cls(cls_score,labels, label_weights, bbox_pred_refine,bbox_gt_refine,bbox_weights_refine,background_label=self.background_label,avg_factor=num_total_samples_refine)
+        else:
+            loss_cls = self.loss_cls(
+                cls_score, labels, label_weights, avg_factor=num_total_samples_refine)
+        # heatmap cls loss
+        hm_score = hm_score.permute(0, 2, 3, 1).reshape(-1, 2)
+        hm_score_tl, hm_score_br = torch.chunk(hm_score, 2, dim=-1)
+        hm_score_tl = hm_score_tl.squeeze(1).sigmoid()
+        hm_score_br = hm_score_br.squeeze(1).sigmoid()
+
+        gt_hm_tl = gt_hm_tl.reshape(-1)
+        gt_hm_tl_weight = gt_hm_tl_weight.reshape(-1)
+        gt_hm_br = gt_hm_br.reshape(-1)
+        gt_hm_br_weight = gt_hm_br_weight.reshape(-1)
+
+        loss_heatmap = 0
+        loss_heatmap += self.loss_heatmap(
+            hm_score_tl, gt_hm_tl, gt_hm_tl_weight, avg_factor=num_total_samples_tl
+        )
+        loss_heatmap += self.loss_heatmap(
+            hm_score_br, gt_hm_br, gt_hm_br_weight, avg_factor=num_total_samples_br
+        )
+        loss_heatmap /= 2.0
+
+        # heatmap offset loss
+        hm_offset = hm_offset.permute(0, 2, 3, 1).reshape(-1, 4)
+        hm_offset_tl, hm_offset_br = torch.chunk(hm_offset, 2, dim=-1)
+
+        gt_offset_tl = gt_offset_tl.reshape(-1, 2)
+        gt_offset_tl_weight = gt_offset_tl_weight.reshape(-1, 2)
+        gt_offset_br = gt_offset_br.reshape(-1, 2)
+        gt_offset_br_weight = gt_offset_br_weight.reshape(-1, 2)
+
+        loss_offset = 0
+        loss_offset += self.loss_offset(
+            hm_offset_tl, gt_offset_tl, gt_offset_tl_weight,
+            avg_factor=num_total_samples_tl
+        )
+        loss_offset += self.loss_offset(
+            hm_offset_br, gt_offset_br, gt_offset_br_weight,
+            avg_factor=num_total_samples_br
+        )
+        loss_offset /= 2.0
+
+        return loss_cls, loss_pts_init, loss_pts_refine, loss_heatmap, loss_offset
+
+    def loss(self,
+             cls_scores,
+             pts_preds_init,
+             pts_preds_refine,
+             hm_scores,
+             hm_offsets,
+             sem_scores,
+             gt_bboxes,
+             gt_sem_map,
+             gt_sem_weights,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None,
+             gt_masks=None):
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == len(self.point_generators)
+        label_channels = self.cls_out_channels
+
+        # target for initial stage
+        # stride affect
+        center_list, valid_flag_list = self.get_points(featmap_sizes, img_metas)
+        pts_coordinate_preds_init = self.offset_to_pts(center_list, pts_preds_init)
+        if self.train_cfg.init.assigner['type'] != 'MaxIoUAssigner':
+            # Assign target for center list
+            candidate_list = center_list
+        else:
+            # transform center list to bbox list and
+            #   assign target for bbox list
+            bbox_list = self.centers_to_bboxes(center_list)
+            candidate_list = bbox_list
+        cls_reg_targets_init = self.get_targets(
+            candidate_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            stage='init',
+            label_channels=label_channels,
+            gt_masks=gt_masks)
+        (*_, bbox_gt_list_init, bbox_weights_list_init,
+         num_total_pos_init, num_total_neg_init, gt_inds) = cls_reg_targets_init
+
+        # target for heatmap in initial stage
+        # stride affect
+        proposal_list, valid_flag_list = self.get_points(featmap_sizes, img_metas)
+        heatmap_targets = self.get_hm_targets(
+            proposal_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_labels)
+        (gt_hm_tl_list, gt_offset_tl_list, gt_hm_tl_weight_list, gt_offset_tl_weight_list,
+         gt_hm_br_list, gt_offset_br_list, gt_hm_br_weight_list, gt_offset_br_weight_list,
+         num_total_pos_tl, num_total_neg_tl, num_total_pos_br, num_total_neg_br) = heatmap_targets
+
+        # target for refinement stage
+        center_list, valid_flag_list = self.get_points(featmap_sizes, img_metas) #[x,y,stride]
+        pts_coordinate_preds_refine = self.offset_to_pts(center_list, pts_preds_refine)
+        bbox_list = []
+        
+        for i_img, center in enumerate(center_list):
+            bbox = []
+            for i_lvl in range(len(pts_preds_refine)):
+                bbox_preds_init = self.points2bbox(
+                    pts_preds_init[i_lvl].detach())
+                bbox_shift = bbox_preds_init * self.point_strides[i_lvl]
+                bbox_center = torch.cat([center[i_lvl][:, :2], center[i_lvl][:, :2]], dim=1) # no stride
+                bbox.append(bbox_center + bbox_shift[i_img].permute(1, 2, 0).reshape(-1, 4))
+            bbox_list.append(bbox)
+        cls_reg_targets_refine = self.get_targets(
+            bbox_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            stage='refine',
+            label_channels=label_channels,
+            gt_masks=gt_masks)
+        (labels_list, label_weights_list,
+         bbox_gt_list_refine, bbox_weights_list_refine,
+         num_total_pos_refine, num_total_neg_refine, gt_inds) = cls_reg_targets_refine
+
+        if self.mask_head:
+            def get_shape(boxes):
+                return torch.stack((boxes[:,2]-boxes[:,0], boxes[:,3]-boxes[:,1]), axis=-1)
+            self.pred_instances.gt_inds = torch.cat([torch.flatten(gt_ind) for gt_ind in gt_inds])
+            # norm by gt box, size=(W,H)
+            self.pred_instances.boxsz = torch.cat([get_shape(gt_bbox.reshape(-1,4)) for gt_bbox in bbox_gt_list_refine])
+            # norm by pred box, size=(W,H)
+            # bbox_pred_refine = [self.points2bbox(pts.reshape(-1, 2 * self.num_points), y_first=False) for pts in pts_coordinate_preds_refine]
+            # self.pred_instances.boxsz = torch.cat([get_shape(bbox.reshape(-1,4)) for bbox in bbox_pred_refine])
+            
+        # compute loss
+        losses_cls, losses_pts_init, losses_pts_refine, losses_heatmap, losses_offset = multi_apply(
+            self.loss_single,
+            cls_scores,
+            pts_coordinate_preds_init,
+            pts_coordinate_preds_refine,
+            hm_scores,
+            hm_offsets,
+            labels_list,
+            label_weights_list,
+            bbox_gt_list_init,
+            bbox_weights_list_init,
+            bbox_gt_list_refine,
+            bbox_weights_list_refine,
+            gt_hm_tl_list,
+            gt_offset_tl_list,
+            gt_hm_tl_weight_list,
+            gt_offset_tl_weight_list,
+            gt_hm_br_list,
+            gt_offset_br_list,
+            gt_hm_br_weight_list,
+            gt_offset_br_weight_list,
+            self.point_strides,
+            num_total_samples_init=num_total_pos_init,
+            num_total_samples_refine=num_total_pos_refine,
+            num_total_samples_tl=num_total_pos_tl,
+            num_total_samples_br=num_total_pos_br)
+
+        # sem loss
+        concat_sem_scores = []
+        concat_gt_sem_map = []
+        concat_gt_sem_weights = []
+
+        for i in range(5):
+            sem_score = sem_scores[i]
+            gt_lvl_sem_map = F.interpolate(gt_sem_map, sem_score.shape[-2:]).reshape(-1)
+            gt_lvl_sem_weight = F.interpolate(gt_sem_weights, sem_score.shape[-2:]).reshape(-1)
+            sem_score = sem_score.reshape(-1)
+
+            try:
+                concat_sem_scores = torch.cat([concat_sem_scores, sem_score])
+                concat_gt_sem_map = torch.cat([concat_gt_sem_map, gt_lvl_sem_map])
+                concat_gt_sem_weights = torch.cat([concat_gt_sem_weights, gt_lvl_sem_weight])
+            except:
+                concat_sem_scores = sem_score
+                concat_gt_sem_map = gt_lvl_sem_map
+                concat_gt_sem_weights = gt_lvl_sem_weight
+
+        loss_sem = self.loss_sem(concat_sem_scores, concat_gt_sem_map, concat_gt_sem_weights, avg_factor=(concat_gt_sem_map > 0).sum())
+
+        loss_dict_all = {'loss_cls': losses_cls,
+                         'loss_pts_init': losses_pts_init,
+                         'loss_pts_refine': losses_pts_refine,
+                         'loss_heatmap': losses_heatmap,
+                         'loss_offset': losses_offset,
+                         'loss_sem': loss_sem,
+                         }
+        return loss_dict_all
+
+    def get_bboxes(self,
+                   cls_scores,
+                   pts_preds_init,
+                   pts_preds_refine,
+                   hm_scores,
+                   hm_offsets,
+                   sem_scores,
+                   img_metas,
+                   cfg=None,
+                   rescale=False,
+                   nms=True):
+        assert len(cls_scores) == len(pts_preds_refine)
+        bbox_preds_refine = [self.points2bbox(pts_pred_refine) for pts_pred_refine in pts_preds_refine]
+        num_levels = len(cls_scores)
+        mlvl_points = [
+            self.point_generators[i].grid_points(cls_scores[i].size()[-2:],
+                                                 self.point_strides[i])
+            for i in range(num_levels)
+        ]
+        result_list = []
+        for img_id in range(len(img_metas)):
+            cls_score_list = [
+                cls_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_pred_list = [
+                bbox_preds_refine[i][img_id].detach() for i in range(num_levels)
+            ]
+            hm_scores_list = [
+                hm_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            hm_offsets_list = [
+                hm_offsets[i][img_id].detach() for i in range(num_levels)
+            ]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            proposals = self._get_bboxes_single(cls_score_list, bbox_pred_list, hm_scores_list, hm_offsets_list,
+                                                mlvl_points, img_shape,
+                                                scale_factor, cfg, rescale,
+                                                nms, imid_shift=self.image_sum*img_id if self.mask_head else 0)
+            result_list.append(proposals)
+        return result_list
+
+    def _get_bboxes_single(self,
+                           cls_scores,
+                           bbox_preds,
+                           hm_scores,
+                           hm_offsets,
+                           mlvl_points,
+                           img_shape,
+                           scale_factor,
+                           cfg,
+                           rescale=False,
+                           nms=True,
+                           imid_shift=0):
+        def select(score_map, x, y, ks=2, i=0):
+            H, W = score_map.shape[-2], score_map.shape[-1]
+            score_map = score_map.sigmoid()
+            score_map_original = score_map.clone()
+
+            score_map, indices = F.max_pool2d_with_indices(score_map.unsqueeze(0), kernel_size=ks, stride=1, padding=(ks - 1) // 2)
+
+            indices = indices.squeeze(0).squeeze(0)
+
+            if ks % 2 == 0:
+                round_func = torch.floor
+            else:
+                round_func = torch.round
+
+            x_round = round_func((x / self.point_strides[i]).clamp(min=0, max=score_map.shape[-1] - 1))
+            y_round = round_func((y / self.point_strides[i]).clamp(min=0, max=score_map.shape[-2] - 1))
+
+            select_indices = indices[y_round.to(torch.long), x_round.to(torch.long)]
+            new_x = select_indices % W
+            new_y = select_indices // W
+
+            score_map_squeeze = score_map_original.squeeze(0)
+            score = score_map_squeeze[new_y, new_x]
+
+            new_x, new_y = new_x.to(torch.float), new_y.to(torch.float)
+
+            return new_x, new_y, score
+
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_scores) == len(bbox_preds) == len(mlvl_points)
+        mlvl_bboxes = []
+        mlvl_scores = []
+        inst_inds = [] if self.mask_head else None
+        for i_lvl, (cls_score, bbox_pred, points) in enumerate(zip(cls_scores, bbox_preds, mlvl_points)):
+            if self.mask_head:
+                inst_ind = torch.arange(cls_score.shape[-2]*cls_score.shape[-1], device=cls_score.device) + self.level_sum[i_lvl] + imid_shift[i_lvl]
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            scores = cls_score.permute(1, 2, 0).reshape(-1, self.cls_out_channels).sigmoid()
+            bbox_pred = bbox_pred.permute(1, 2, 0).reshape(-1, 4)
+            nms_pre = cfg.get('nms_pre', -1)
+            if nms_pre > 0 and scores.shape[0] > nms_pre:
+                max_scores, _ = scores.max(dim=1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                points = points[topk_inds, :]
+                bbox_pred = bbox_pred[topk_inds, :]
+                scores = scores[topk_inds, :]
+                if self.mask_head:
+                    inst_ind = inst_ind[topk_inds]
+            bbox_pos_center = torch.cat([points[:, :2], points[:, :2]], dim=1)
+            bboxes = bbox_pred * self.point_strides[i_lvl] + bbox_pos_center
+            x1 = bboxes[:, 0].clamp(min=0, max=img_shape[1])
+            y1 = bboxes[:, 1].clamp(min=0, max=img_shape[0])
+            x2 = bboxes[:, 2].clamp(min=0, max=img_shape[1])
+            y2 = bboxes[:, 3].clamp(min=0, max=img_shape[0])
+            if self.corner_refine:
+                if i_lvl > 0:
+                    i = 0 if i_lvl in (1, 2) else 1
+
+                    x1_new, y1_new, score1_new = select(hm_scores[i][0, ...], x1, y1, 2, i)
+                    x2_new, y2_new, score2_new = select(hm_scores[i][1, ...], x2, y2, 2, i)
+
+                    hm_offset = hm_offsets[i].permute(1, 2, 0)
+                    point_stride = self.point_strides[i]
+
+                    x1 = ((x1_new + hm_offset[y1_new.to(torch.long), x1_new.to(torch.long), 0]) * point_stride).clamp(min=0, max=img_shape[1])
+                    y1 = ((y1_new + hm_offset[y1_new.to(torch.long), x1_new.to(torch.long), 1]) * point_stride).clamp(min=0, max=img_shape[0])
+                    x2 = ((x2_new + hm_offset[y2_new.to(torch.long), x2_new.to(torch.long), 2]) * point_stride).clamp(min=0, max=img_shape[1])
+                    y2 = ((y2_new + hm_offset[y2_new.to(torch.long), x2_new.to(torch.long), 3]) * point_stride).clamp(min=0, max=img_shape[0])
+            bboxes = torch.stack([x1, y1, x2, y2], dim=-1)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+            if self.mask_head:
+                inst_inds.append(inst_ind)
+        if self.mask_head:
+            inst_inds = torch.cat(inst_inds)
+        mlvl_bboxes = torch.cat(mlvl_bboxes)
+        if rescale:
+            mlvl_bboxes /= mlvl_bboxes.new_tensor(scale_factor)
+        mlvl_scores = torch.cat(mlvl_scores)
+        padding = mlvl_scores.new_zeros(mlvl_scores.shape[0], 1)
+        mlvl_scores = torch.cat([mlvl_scores, padding], dim=1)
+        if nms:
+            if self.mask_head:
+                det_bboxes, det_labels, inst_inds, _ = multiclass_nms_rpd(mlvl_bboxes, mlvl_scores,
+                                                    cfg.score_thr, cfg.nms,
+                                                    cfg.max_per_img, inst_inds=inst_inds)
+                return det_bboxes, det_labels, inst_inds
+            else:
+                det_bboxes, det_labels, _ = multiclass_nms_rpd(mlvl_bboxes, mlvl_scores,
+                                                    cfg.score_thr, cfg.nms,
+                                                    cfg.max_per_img)
+                return det_bboxes, det_labels
+        else:
+            if self.mask_head:
+                return mlvl_bboxes, mlvl_scores, inst_inds
+            else:
+                return mlvl_bboxes, mlvl_scores
+
+    def get_num_level_proposals_inside(self, num_level_proposals, inside_flags):
+        split_inside_flags = torch.split(inside_flags, num_level_proposals)
+        num_level_proposals_inside = [
+            int(flags.sum()) for flags in split_inside_flags
+        ]
+        return num_level_proposals_inside
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels=None,
+                      gt_bboxes_ignore=None,
+                      gt_sem_map=None,
+                      gt_sem_weights=None,
+                      gt_masks=None,
+                      proposal_cfg=None,
+                      **kwargs):
+        """
+        # NOTE! for condinst only
+        Args:
+            x (list[Tensor]): Features from FPN.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes (Tensor): Ground truth bboxes of the image,
+                shape (num_gts, 4).
+            gt_labels (Tensor): Ground truth labels of each box,
+                shape (num_gts,).
+            gt_bboxes_ignore (Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+            gt_masks (None | list[Tensor]) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+            proposal_cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used
+
+        Returns:
+            tuple:
+                losses: (dict[str, Tensor]): A dictionary of loss components.
+                proposal_list (list[Tensor]): Proposals of each image.
+        """
+        if gt_masks:
+            self.masknum_sum = tuple(accumulate([0]+[len(mask) for mask in gt_masks][:-1]))
+        outs = self(x)
+        if gt_labels is None:
+            loss_inputs = outs + (gt_bboxes, gt_sem_map, gt_sem_weights, img_metas)
+        else:
+            loss_inputs = outs + (gt_bboxes, gt_sem_map, gt_sem_weights, gt_labels, img_metas)
+        
+        losses = self.loss(*loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore, gt_masks=gt_masks)
+        if self.mask_head:
+            self.pred_instances = self.pred_instances[self.pred_instances.gt_inds > 0] # select only foreground labels
+            self.pred_instances.gt_inds = self.pred_instances.gt_inds - 1
+            mask_loss = self.mask_head(x, self.pred_instances, gt_masks, gt_labels)
+            losses.update(mask_loss)
+        
+        if proposal_cfg is None:
+            return losses
+        else:
+            proposal_list = self.get_bboxes(*outs, img_metas, cfg=proposal_cfg)
+            return losses, proposal_list
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/retina_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/retina_head.py
new file mode 100644
index 0000000..b12416f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/retina_head.py
@@ -0,0 +1,114 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, bias_init_with_prob, normal_init
+
+from ..builder import HEADS
+from .anchor_head import AnchorHead
+
+
+@HEADS.register_module()
+class RetinaHead(AnchorHead):
+    r"""An anchor-based head used in `RetinaNet
+    <https://arxiv.org/pdf/1708.02002.pdf>`_.
+
+    The head contains two subnetworks. The first classifies anchor boxes and
+    the second regresses deltas for the anchors.
+
+    Example:
+        >>> import torch
+        >>> self = RetinaHead(11, 7)
+        >>> x = torch.rand(1, 7, 32, 32)
+        >>> cls_score, bbox_pred = self.forward_single(x)
+        >>> # Each anchor predicts a score for each class except background
+        >>> cls_per_anchor = cls_score.shape[1] / self.num_anchors
+        >>> box_per_anchor = bbox_pred.shape[1] / self.num_anchors
+        >>> assert cls_per_anchor == (self.num_classes)
+        >>> assert box_per_anchor == 4
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 stacked_convs=4,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 anchor_generator=dict(
+                     type='AnchorGenerator',
+                     octave_base_scale=4,
+                     scales_per_octave=3,
+                     ratios=[0.5, 1.0, 2.0],
+                     strides=[8, 16, 32, 64, 128]),
+                 **kwargs):
+        self.stacked_convs = stacked_convs
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        super(RetinaHead, self).__init__(
+            num_classes,
+            in_channels,
+            anchor_generator=anchor_generator,
+            **kwargs)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.relu = nn.ReLU(inplace=True)
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            self.cls_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+            self.reg_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+        self.retina_cls = nn.Conv2d(
+            self.feat_channels,
+            self.num_anchors * self.cls_out_channels,
+            3,
+            padding=1)
+        self.retina_reg = nn.Conv2d(
+            self.feat_channels, self.num_anchors * 4, 3, padding=1)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.cls_convs:
+            normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            normal_init(m.conv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.retina_cls, std=0.01, bias=bias_cls)
+        normal_init(self.retina_reg, std=0.01)
+
+    def forward_single(self, x):
+        """Forward feature of a single scale level.
+
+        Args:
+            x (Tensor): Features of a single scale level.
+
+        Returns:
+            tuple:
+                cls_score (Tensor): Cls scores for a single scale level
+                    the channels number is num_anchors * num_classes.
+                bbox_pred (Tensor): Box energies / deltas for a single scale
+                    level, the channels number is num_anchors * 4.
+        """
+        cls_feat = x
+        reg_feat = x
+        for cls_conv in self.cls_convs:
+            cls_feat = cls_conv(cls_feat)
+        for reg_conv in self.reg_convs:
+            reg_feat = reg_conv(reg_feat)
+        cls_score = self.retina_cls(cls_feat)
+        bbox_pred = self.retina_reg(reg_feat)
+        return cls_score, bbox_pred
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/retina_sepbn_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/retina_sepbn_head.py
new file mode 100644
index 0000000..6b8ce7f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/retina_sepbn_head.py
@@ -0,0 +1,113 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, bias_init_with_prob, normal_init
+
+from ..builder import HEADS
+from .anchor_head import AnchorHead
+
+
+@HEADS.register_module()
+class RetinaSepBNHead(AnchorHead):
+    """"RetinaHead with separate BN.
+
+    In RetinaHead, conv/norm layers are shared across different FPN levels,
+    while in RetinaSepBNHead, conv layers are shared across different FPN
+    levels, but BN layers are separated.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 num_ins,
+                 in_channels,
+                 stacked_convs=4,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 **kwargs):
+        self.stacked_convs = stacked_convs
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.num_ins = num_ins
+        super(RetinaSepBNHead, self).__init__(num_classes, in_channels,
+                                              **kwargs)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.relu = nn.ReLU(inplace=True)
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        for i in range(self.num_ins):
+            cls_convs = nn.ModuleList()
+            reg_convs = nn.ModuleList()
+            for i in range(self.stacked_convs):
+                chn = self.in_channels if i == 0 else self.feat_channels
+                cls_convs.append(
+                    ConvModule(
+                        chn,
+                        self.feat_channels,
+                        3,
+                        stride=1,
+                        padding=1,
+                        conv_cfg=self.conv_cfg,
+                        norm_cfg=self.norm_cfg))
+                reg_convs.append(
+                    ConvModule(
+                        chn,
+                        self.feat_channels,
+                        3,
+                        stride=1,
+                        padding=1,
+                        conv_cfg=self.conv_cfg,
+                        norm_cfg=self.norm_cfg))
+            self.cls_convs.append(cls_convs)
+            self.reg_convs.append(reg_convs)
+        for i in range(self.stacked_convs):
+            for j in range(1, self.num_ins):
+                self.cls_convs[j][i].conv = self.cls_convs[0][i].conv
+                self.reg_convs[j][i].conv = self.reg_convs[0][i].conv
+        self.retina_cls = nn.Conv2d(
+            self.feat_channels,
+            self.num_anchors * self.cls_out_channels,
+            3,
+            padding=1)
+        self.retina_reg = nn.Conv2d(
+            self.feat_channels, self.num_anchors * 4, 3, padding=1)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.cls_convs[0]:
+            normal_init(m.conv, std=0.01)
+        for m in self.reg_convs[0]:
+            normal_init(m.conv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.retina_cls, std=0.01, bias=bias_cls)
+        normal_init(self.retina_reg, std=0.01)
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple: Usually a tuple of classification scores and bbox prediction
+                cls_scores (list[Tensor]): Classification scores for all scale
+                    levels, each is a 4D-tensor, the channels number is
+                    num_anchors * num_classes.
+                bbox_preds (list[Tensor]): Box energies / deltas for all scale
+                    levels, each is a 4D-tensor, the channels number is
+                    num_anchors * 4.
+        """
+        cls_scores = []
+        bbox_preds = []
+        for i, x in enumerate(feats):
+            cls_feat = feats[i]
+            reg_feat = feats[i]
+            for cls_conv in self.cls_convs[i]:
+                cls_feat = cls_conv(cls_feat)
+            for reg_conv in self.reg_convs[i]:
+                reg_feat = reg_conv(reg_feat)
+            cls_score = self.retina_cls(cls_feat)
+            bbox_pred = self.retina_reg(reg_feat)
+            cls_scores.append(cls_score)
+            bbox_preds.append(bbox_pred)
+        return cls_scores, bbox_preds
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/rpn_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/rpn_head.py
new file mode 100644
index 0000000..a888cb8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/rpn_head.py
@@ -0,0 +1,236 @@
+import copy
+import warnings
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv import ConfigDict
+from mmcv.cnn import normal_init
+from mmcv.ops import batched_nms
+
+from ..builder import HEADS
+from .anchor_head import AnchorHead
+from .rpn_test_mixin import RPNTestMixin
+
+
+@HEADS.register_module()
+class RPNHead(RPNTestMixin, AnchorHead):
+    """RPN head.
+
+    Args:
+        in_channels (int): Number of channels in the input feature map.
+    """  # noqa: W605
+
+    def __init__(self, in_channels, **kwargs):
+        super(RPNHead, self).__init__(1, in_channels, **kwargs)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.rpn_conv = nn.Conv2d(
+            self.in_channels, self.feat_channels, 3, padding=1)
+        self.rpn_cls = nn.Conv2d(self.feat_channels,
+                                 self.num_anchors * self.cls_out_channels, 1)
+        self.rpn_reg = nn.Conv2d(self.feat_channels, self.num_anchors * 4, 1)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        normal_init(self.rpn_conv, std=0.01)
+        normal_init(self.rpn_cls, std=0.01)
+        normal_init(self.rpn_reg, std=0.01)
+
+    def forward_single(self, x):
+        """Forward feature map of a single scale level."""
+        x = self.rpn_conv(x)
+        x = F.relu(x, inplace=True)
+        rpn_cls_score = self.rpn_cls(x)
+        rpn_bbox_pred = self.rpn_reg(x)
+        return rpn_cls_score, rpn_bbox_pred
+
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        losses = super(RPNHead, self).loss(
+            cls_scores,
+            bbox_preds,
+            gt_bboxes,
+            None,
+            img_metas,
+            gt_bboxes_ignore=gt_bboxes_ignore)
+        return dict(
+            loss_rpn_cls=losses['loss_cls'], loss_rpn_bbox=losses['loss_bbox'])
+
+    def _get_bboxes(self,
+                    cls_scores,
+                    bbox_preds,
+                    mlvl_anchors,
+                    img_shapes,
+                    scale_factors,
+                    cfg,
+                    rescale=False):
+        """Transform outputs for a single batch item into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W).
+            mlvl_anchors (list[Tensor]): Box reference for each scale level
+                with shape (num_total_anchors, 4).
+            img_shapes (list[tuple[int]]): Shape of the input image,
+                (height, width, 3).
+            scale_factors (list[ndarray]): Scale factor of the image arange as
+                (w_scale, h_scale, w_scale, h_scale).
+            cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where the first 4 columns
+                are bounding box positions (tl_x, tl_y, br_x, br_y) and the
+                5-th column is a score between 0 and 1. The second item is a
+                (n,) tensor where each item is the predicted class labelof the
+                corresponding box.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        cfg = copy.deepcopy(cfg)
+        # bboxes from different level should be independent during NMS,
+        # level_ids are used as labels for batched NMS to separate them
+        level_ids = []
+        mlvl_scores = []
+        mlvl_bbox_preds = []
+        mlvl_valid_anchors = []
+        batch_size = cls_scores[0].shape[0]
+        nms_pre_tensor = torch.tensor(
+            cfg.nms_pre, device=cls_scores[0].device, dtype=torch.long)
+        for idx in range(len(cls_scores)):
+            rpn_cls_score = cls_scores[idx]
+            rpn_bbox_pred = bbox_preds[idx]
+            assert rpn_cls_score.size()[-2:] == rpn_bbox_pred.size()[-2:]
+            rpn_cls_score = rpn_cls_score.permute(0, 2, 3, 1)
+            if self.use_sigmoid_cls:
+                rpn_cls_score = rpn_cls_score.reshape(batch_size, -1)
+                scores = rpn_cls_score.sigmoid()
+            else:
+                rpn_cls_score = rpn_cls_score.reshape(batch_size, -1, 2)
+                # We set FG labels to [0, num_class-1] and BG label to
+                # num_class in RPN head since mmdet v2.5, which is unified to
+                # be consistent with other head since mmdet v2.0. In mmdet v2.0
+                # to v2.4 we keep BG label as 0 and FG label as 1 in rpn head.
+                scores = rpn_cls_score.softmax(-1)[..., 0]
+            rpn_bbox_pred = rpn_bbox_pred.permute(0, 2, 3, 1).reshape(
+                batch_size, -1, 4)
+            anchors = mlvl_anchors[idx]
+            anchors = anchors.expand_as(rpn_bbox_pred)
+            if nms_pre_tensor > 0:
+                # sort is faster than topk
+                # _, topk_inds = scores.topk(cfg.nms_pre)
+                # keep topk op for dynamic k in onnx model
+                if torch.onnx.is_in_onnx_export():
+                    # sort op will be converted to TopK in onnx
+                    # and k<=3480 in TensorRT
+                    scores_shape = torch._shape_as_tensor(scores)
+                    nms_pre = torch.where(scores_shape[1] < nms_pre_tensor,
+                                          scores_shape[1], nms_pre_tensor)
+                    _, topk_inds = scores.topk(nms_pre)
+                    batch_inds = torch.arange(batch_size).view(
+                        -1, 1).expand_as(topk_inds)
+                    scores = scores[batch_inds, topk_inds]
+                    rpn_bbox_pred = rpn_bbox_pred[batch_inds, topk_inds, :]
+                    anchors = anchors[batch_inds, topk_inds, :]
+
+                elif scores.shape[-1] > cfg.nms_pre:
+                    ranked_scores, rank_inds = scores.sort(descending=True)
+                    topk_inds = rank_inds[:, :cfg.nms_pre]
+                    scores = ranked_scores[:, :cfg.nms_pre]
+                    batch_inds = torch.arange(batch_size).view(
+                        -1, 1).expand_as(topk_inds)
+                    rpn_bbox_pred = rpn_bbox_pred[batch_inds, topk_inds, :]
+                    anchors = anchors[batch_inds, topk_inds, :]
+
+            mlvl_scores.append(scores)
+            mlvl_bbox_preds.append(rpn_bbox_pred)
+            mlvl_valid_anchors.append(anchors)
+            level_ids.append(
+                scores.new_full((
+                    batch_size,
+                    scores.size(1),
+                ),
+                                idx,
+                                dtype=torch.long))
+
+        batch_mlvl_scores = torch.cat(mlvl_scores, dim=1)
+        batch_mlvl_anchors = torch.cat(mlvl_valid_anchors, dim=1)
+        batch_mlvl_rpn_bbox_pred = torch.cat(mlvl_bbox_preds, dim=1)
+        batch_mlvl_proposals = self.bbox_coder.decode(
+            batch_mlvl_anchors, batch_mlvl_rpn_bbox_pred, max_shape=img_shapes)
+        batch_mlvl_ids = torch.cat(level_ids, dim=1)
+
+        # deprecate arguments warning
+        if 'nms' not in cfg or 'max_num' in cfg or 'nms_thr' in cfg:
+            warnings.warn(
+                'In rpn_proposal or test_cfg, '
+                'nms_thr has been moved to a dict named nms as '
+                'iou_threshold, max_num has been renamed as max_per_img, '
+                'name of original arguments and the way to specify '
+                'iou_threshold of NMS will be deprecated.')
+        if 'nms' not in cfg:
+            cfg.nms = ConfigDict(dict(type='nms', iou_threshold=cfg.nms_thr))
+        if 'max_num' in cfg:
+            if 'max_per_img' in cfg:
+                assert cfg.max_num == cfg.max_per_img, f'You ' \
+                    f'set max_num and ' \
+                    f'max_per_img at the same time, but get {cfg.max_num} ' \
+                    f'and {cfg.max_per_img} respectively' \
+                    'Please delete max_num which will be deprecated.'
+            else:
+                cfg.max_per_img = cfg.max_num
+        if 'nms_thr' in cfg:
+            assert cfg.nms.iou_threshold == cfg.nms_thr, f'You set' \
+                f' iou_threshold in nms and ' \
+                f'nms_thr at the same time, but get' \
+                f' {cfg.nms.iou_threshold} and {cfg.nms_thr}' \
+                f' respectively. Please delete the nms_thr ' \
+                f'which will be deprecated.'
+
+        result_list = []
+        for (mlvl_proposals, mlvl_scores,
+             mlvl_ids) in zip(batch_mlvl_proposals, batch_mlvl_scores,
+                              batch_mlvl_ids):
+            # Skip nonzero op while exporting to ONNX
+            if cfg.min_bbox_size > 0 and (not torch.onnx.is_in_onnx_export()):
+                w = mlvl_proposals[:, 2] - mlvl_proposals[:, 0]
+                h = mlvl_proposals[:, 3] - mlvl_proposals[:, 1]
+                valid_ind = torch.nonzero(
+                    (w >= cfg.min_bbox_size)
+                    & (h >= cfg.min_bbox_size),
+                    as_tuple=False).squeeze()
+                if valid_ind.sum().item() != len(mlvl_proposals):
+                    mlvl_proposals = mlvl_proposals[valid_ind, :]
+                    mlvl_scores = mlvl_scores[valid_ind]
+                    mlvl_ids = mlvl_ids[valid_ind]
+
+            dets, keep = batched_nms(mlvl_proposals, mlvl_scores, mlvl_ids,
+                                     cfg.nms)
+            result_list.append(dets[:cfg.max_per_img])
+        return result_list
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/rpn_test_mixin.py b/python/autoCut/models/swin/mmdet/models/dense_heads/rpn_test_mixin.py
new file mode 100644
index 0000000..4ce5c66
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/rpn_test_mixin.py
@@ -0,0 +1,59 @@
+import sys
+
+from mmdet.core import merge_aug_proposals
+
+if sys.version_info >= (3, 7):
+    from mmdet.utils.contextmanagers import completed
+
+
+class RPNTestMixin(object):
+    """Test methods of RPN."""
+
+    if sys.version_info >= (3, 7):
+
+        async def async_simple_test_rpn(self, x, img_metas):
+            sleep_interval = self.test_cfg.pop('async_sleep_interval', 0.025)
+            async with completed(
+                    __name__, 'rpn_head_forward',
+                    sleep_interval=sleep_interval):
+                rpn_outs = self(x)
+
+            proposal_list = self.get_bboxes(*rpn_outs, img_metas)
+            return proposal_list
+
+    def simple_test_rpn(self, x, img_metas):
+        """Test without augmentation.
+
+        Args:
+            x (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+            img_metas (list[dict]): Meta info of each image.
+
+        Returns:
+            list[Tensor]: Proposals of each image.
+        """
+        rpn_outs = self(x)
+        proposal_list = self.get_bboxes(*rpn_outs, img_metas)
+        return proposal_list
+
+    def aug_test_rpn(self, feats, img_metas):
+        samples_per_gpu = len(img_metas[0])
+        aug_proposals = [[] for _ in range(samples_per_gpu)]
+        for x, img_meta in zip(feats, img_metas):
+            proposal_list = self.simple_test_rpn(x, img_meta)
+            for i, proposals in enumerate(proposal_list):
+                aug_proposals[i].append(proposals)
+        # reorganize the order of 'img_metas' to match the dimensions
+        # of 'aug_proposals'
+        aug_img_metas = []
+        for i in range(samples_per_gpu):
+            aug_img_meta = []
+            for j in range(len(img_metas)):
+                aug_img_meta.append(img_metas[j][i])
+            aug_img_metas.append(aug_img_meta)
+        # after merging, proposals will be rescaled to the original image size
+        merged_proposals = [
+            merge_aug_proposals(proposals, aug_img_meta, self.test_cfg)
+            for proposals, aug_img_meta in zip(aug_proposals, aug_img_metas)
+        ]
+        return merged_proposals
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/sabl_retina_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/sabl_retina_head.py
new file mode 100644
index 0000000..4211622
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/sabl_retina_head.py
@@ -0,0 +1,621 @@
+import numpy as np
+import torch
+import torch.nn as nn
+from mmcv.cnn import ConvModule, bias_init_with_prob, normal_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import (build_anchor_generator, build_assigner,
+                        build_bbox_coder, build_sampler, images_to_levels,
+                        multi_apply, multiclass_nms, unmap)
+from ..builder import HEADS, build_loss
+from .base_dense_head import BaseDenseHead
+from .guided_anchor_head import GuidedAnchorHead
+
+
+@HEADS.register_module()
+class SABLRetinaHead(BaseDenseHead):
+    """Side-Aware Boundary Localization (SABL) for RetinaNet.
+
+    The anchor generation, assigning and sampling in SABLRetinaHead
+    are the same as GuidedAnchorHead for guided anchoring.
+
+    Please refer to https://arxiv.org/abs/1912.04260 for more details.
+
+    Args:
+        num_classes (int): Number of classes.
+        in_channels (int): Number of channels in the input feature map.
+        stacked_convs (int): Number of Convs for classification \
+            and regression branches. Defaults to 4.
+        feat_channels (int): Number of hidden channels. \
+            Defaults to 256.
+        approx_anchor_generator (dict): Config dict for approx generator.
+        square_anchor_generator (dict): Config dict for square generator.
+        conv_cfg (dict): Config dict for ConvModule. Defaults to None.
+        norm_cfg (dict): Config dict for Norm Layer. Defaults to None.
+        bbox_coder (dict): Config dict for bbox coder.
+        reg_decoded_bbox (bool): If true, the regression loss would be
+            applied directly on decoded bounding boxes, converting both
+            the predicted boxes and regression targets to absolute
+            coordinates format. Default False. It should be `True` when
+            using `IoULoss`, `GIoULoss`, or `DIoULoss` in the bbox head.
+        train_cfg (dict): Training config of SABLRetinaHead.
+        test_cfg (dict): Testing config of SABLRetinaHead.
+        loss_cls (dict): Config of classification loss.
+        loss_bbox_cls (dict): Config of classification loss for bbox branch.
+        loss_bbox_reg (dict): Config of regression loss for bbox branch.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 stacked_convs=4,
+                 feat_channels=256,
+                 approx_anchor_generator=dict(
+                     type='AnchorGenerator',
+                     octave_base_scale=4,
+                     scales_per_octave=3,
+                     ratios=[0.5, 1.0, 2.0],
+                     strides=[8, 16, 32, 64, 128]),
+                 square_anchor_generator=dict(
+                     type='AnchorGenerator',
+                     ratios=[1.0],
+                     scales=[4],
+                     strides=[8, 16, 32, 64, 128]),
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 bbox_coder=dict(
+                     type='BucketingBBoxCoder',
+                     num_buckets=14,
+                     scale_factor=3.0),
+                 reg_decoded_bbox=False,
+                 train_cfg=None,
+                 test_cfg=None,
+                 loss_cls=dict(
+                     type='FocalLoss',
+                     use_sigmoid=True,
+                     gamma=2.0,
+                     alpha=0.25,
+                     loss_weight=1.0),
+                 loss_bbox_cls=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.5),
+                 loss_bbox_reg=dict(
+                     type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.5)):
+        super(SABLRetinaHead, self).__init__()
+        self.in_channels = in_channels
+        self.num_classes = num_classes
+        self.feat_channels = feat_channels
+        self.num_buckets = bbox_coder['num_buckets']
+        self.side_num = int(np.ceil(self.num_buckets / 2))
+
+        assert (approx_anchor_generator['octave_base_scale'] ==
+                square_anchor_generator['scales'][0])
+        assert (approx_anchor_generator['strides'] ==
+                square_anchor_generator['strides'])
+
+        self.approx_anchor_generator = build_anchor_generator(
+            approx_anchor_generator)
+        self.square_anchor_generator = build_anchor_generator(
+            square_anchor_generator)
+        self.approxs_per_octave = (
+            self.approx_anchor_generator.num_base_anchors[0])
+
+        # one anchor per location
+        self.num_anchors = 1
+        self.stacked_convs = stacked_convs
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+
+        self.reg_decoded_bbox = reg_decoded_bbox
+
+        self.use_sigmoid_cls = loss_cls.get('use_sigmoid', False)
+        self.sampling = loss_cls['type'] not in [
+            'FocalLoss', 'GHMC', 'QualityFocalLoss'
+        ]
+        if self.use_sigmoid_cls:
+            self.cls_out_channels = num_classes
+        else:
+            self.cls_out_channels = num_classes + 1
+
+        self.bbox_coder = build_bbox_coder(bbox_coder)
+        self.loss_cls = build_loss(loss_cls)
+        self.loss_bbox_cls = build_loss(loss_bbox_cls)
+        self.loss_bbox_reg = build_loss(loss_bbox_reg)
+
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            # use PseudoSampler when sampling is False
+            if self.sampling and hasattr(self.train_cfg, 'sampler'):
+                sampler_cfg = self.train_cfg.sampler
+            else:
+                sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+
+        self.fp16_enabled = False
+        self._init_layers()
+
+    def _init_layers(self):
+        self.relu = nn.ReLU(inplace=True)
+        self.cls_convs = nn.ModuleList()
+        self.reg_convs = nn.ModuleList()
+        for i in range(self.stacked_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            self.cls_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+            self.reg_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+        self.retina_cls = nn.Conv2d(
+            self.feat_channels, self.cls_out_channels, 3, padding=1)
+        self.retina_bbox_reg = nn.Conv2d(
+            self.feat_channels, self.side_num * 4, 3, padding=1)
+        self.retina_bbox_cls = nn.Conv2d(
+            self.feat_channels, self.side_num * 4, 3, padding=1)
+
+    def init_weights(self):
+        for m in self.cls_convs:
+            normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            normal_init(m.conv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.retina_cls, std=0.01, bias=bias_cls)
+        normal_init(self.retina_bbox_reg, std=0.01)
+        normal_init(self.retina_bbox_cls, std=0.01)
+
+    def forward_single(self, x):
+        cls_feat = x
+        reg_feat = x
+        for cls_conv in self.cls_convs:
+            cls_feat = cls_conv(cls_feat)
+        for reg_conv in self.reg_convs:
+            reg_feat = reg_conv(reg_feat)
+        cls_score = self.retina_cls(cls_feat)
+        bbox_cls_pred = self.retina_bbox_cls(reg_feat)
+        bbox_reg_pred = self.retina_bbox_reg(reg_feat)
+        bbox_pred = (bbox_cls_pred, bbox_reg_pred)
+        return cls_score, bbox_pred
+
+    def forward(self, feats):
+        return multi_apply(self.forward_single, feats)
+
+    def get_anchors(self, featmap_sizes, img_metas, device='cuda'):
+        """Get squares according to feature map sizes and guided anchors.
+
+        Args:
+            featmap_sizes (list[tuple]): Multi-level feature map sizes.
+            img_metas (list[dict]): Image meta info.
+            device (torch.device | str): device for returned tensors
+
+        Returns:
+            tuple: square approxs of each image
+        """
+        num_imgs = len(img_metas)
+
+        # since feature map sizes of all images are the same, we only compute
+        # squares for one time
+        multi_level_squares = self.square_anchor_generator.grid_anchors(
+            featmap_sizes, device=device)
+        squares_list = [multi_level_squares for _ in range(num_imgs)]
+
+        return squares_list
+
+    def get_target(self,
+                   approx_list,
+                   inside_flag_list,
+                   square_list,
+                   gt_bboxes_list,
+                   img_metas,
+                   gt_bboxes_ignore_list=None,
+                   gt_labels_list=None,
+                   label_channels=None,
+                   sampling=True,
+                   unmap_outputs=True):
+        """Compute bucketing targets.
+        Args:
+            approx_list (list[list]): Multi level approxs of each image.
+            inside_flag_list (list[list]): Multi level inside flags of each
+                image.
+            square_list (list[list]): Multi level squares of each image.
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image.
+            img_metas (list[dict]): Meta info of each image.
+            gt_bboxes_ignore_list (list[Tensor]): ignore list of gt bboxes.
+            gt_bboxes_list (list[Tensor]): Gt bboxes of each image.
+            label_channels (int): Channel of label.
+            sampling (bool): Sample Anchors or not.
+            unmap_outputs (bool): unmap outputs or not.
+
+        Returns:
+            tuple: Returns a tuple containing learning targets.
+
+                - labels_list (list[Tensor]): Labels of each level.
+                - label_weights_list (list[Tensor]): Label weights of each \
+                    level.
+                - bbox_cls_targets_list (list[Tensor]): BBox cls targets of \
+                    each level.
+                - bbox_cls_weights_list (list[Tensor]): BBox cls weights of \
+                    each level.
+                - bbox_reg_targets_list (list[Tensor]): BBox reg targets of \
+                    each level.
+                - bbox_reg_weights_list (list[Tensor]): BBox reg weights of \
+                    each level.
+                - num_total_pos (int): Number of positive samples in all \
+                    images.
+                - num_total_neg (int): Number of negative samples in all \
+                    images.
+        """
+        num_imgs = len(img_metas)
+        assert len(approx_list) == len(inside_flag_list) == len(
+            square_list) == num_imgs
+        # anchor number of multi levels
+        num_level_squares = [squares.size(0) for squares in square_list[0]]
+        # concat all level anchors and flags to a single tensor
+        inside_flag_flat_list = []
+        approx_flat_list = []
+        square_flat_list = []
+        for i in range(num_imgs):
+            assert len(square_list[i]) == len(inside_flag_list[i])
+            inside_flag_flat_list.append(torch.cat(inside_flag_list[i]))
+            approx_flat_list.append(torch.cat(approx_list[i]))
+            square_flat_list.append(torch.cat(square_list[i]))
+
+        # compute targets for each image
+        if gt_bboxes_ignore_list is None:
+            gt_bboxes_ignore_list = [None for _ in range(num_imgs)]
+        if gt_labels_list is None:
+            gt_labels_list = [None for _ in range(num_imgs)]
+        (all_labels, all_label_weights, all_bbox_cls_targets,
+         all_bbox_cls_weights, all_bbox_reg_targets, all_bbox_reg_weights,
+         pos_inds_list, neg_inds_list) = multi_apply(
+             self._get_target_single,
+             approx_flat_list,
+             inside_flag_flat_list,
+             square_flat_list,
+             gt_bboxes_list,
+             gt_bboxes_ignore_list,
+             gt_labels_list,
+             img_metas,
+             label_channels=label_channels,
+             sampling=sampling,
+             unmap_outputs=unmap_outputs)
+        # no valid anchors
+        if any([labels is None for labels in all_labels]):
+            return None
+        # sampled anchors of all images
+        num_total_pos = sum([max(inds.numel(), 1) for inds in pos_inds_list])
+        num_total_neg = sum([max(inds.numel(), 1) for inds in neg_inds_list])
+        # split targets to a list w.r.t. multiple levels
+        labels_list = images_to_levels(all_labels, num_level_squares)
+        label_weights_list = images_to_levels(all_label_weights,
+                                              num_level_squares)
+        bbox_cls_targets_list = images_to_levels(all_bbox_cls_targets,
+                                                 num_level_squares)
+        bbox_cls_weights_list = images_to_levels(all_bbox_cls_weights,
+                                                 num_level_squares)
+        bbox_reg_targets_list = images_to_levels(all_bbox_reg_targets,
+                                                 num_level_squares)
+        bbox_reg_weights_list = images_to_levels(all_bbox_reg_weights,
+                                                 num_level_squares)
+        return (labels_list, label_weights_list, bbox_cls_targets_list,
+                bbox_cls_weights_list, bbox_reg_targets_list,
+                bbox_reg_weights_list, num_total_pos, num_total_neg)
+
+    def _get_target_single(self,
+                           flat_approxs,
+                           inside_flags,
+                           flat_squares,
+                           gt_bboxes,
+                           gt_bboxes_ignore,
+                           gt_labels,
+                           img_meta,
+                           label_channels=None,
+                           sampling=True,
+                           unmap_outputs=True):
+        """Compute regression and classification targets for anchors in a
+        single image.
+
+        Args:
+            flat_approxs (Tensor): flat approxs of a single image,
+                shape (n, 4)
+            inside_flags (Tensor): inside flags of a single image,
+                shape (n, ).
+            flat_squares (Tensor): flat squares of a single image,
+                shape (approxs_per_octave * n, 4)
+            gt_bboxes (Tensor): Ground truth bboxes of a single image, \
+                shape (num_gts, 4).
+            gt_bboxes_ignore (Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+            gt_labels (Tensor): Ground truth labels of each box,
+                shape (num_gts,).
+            img_meta (dict): Meta info of the image.
+            label_channels (int): Channel of label.
+            sampling (bool): Sample Anchors or not.
+            unmap_outputs (bool): unmap outputs or not.
+
+        Returns:
+            tuple:
+
+                - labels_list (Tensor): Labels in a single image
+                - label_weights (Tensor): Label weights in a single image
+                - bbox_cls_targets (Tensor): BBox cls targets in a single image
+                - bbox_cls_weights (Tensor): BBox cls weights in a single image
+                - bbox_reg_targets (Tensor): BBox reg targets in a single image
+                - bbox_reg_weights (Tensor): BBox reg weights in a single image
+                - num_total_pos (int): Number of positive samples \
+                    in a single image
+                - num_total_neg (int): Number of negative samples \
+                    in a single image
+        """
+        if not inside_flags.any():
+            return (None, ) * 8
+        # assign gt and sample anchors
+        expand_inside_flags = inside_flags[:, None].expand(
+            -1, self.approxs_per_octave).reshape(-1)
+        approxs = flat_approxs[expand_inside_flags, :]
+        squares = flat_squares[inside_flags, :]
+
+        assign_result = self.assigner.assign(approxs, squares,
+                                             self.approxs_per_octave,
+                                             gt_bboxes, gt_bboxes_ignore)
+        sampling_result = self.sampler.sample(assign_result, squares,
+                                              gt_bboxes)
+
+        num_valid_squares = squares.shape[0]
+        bbox_cls_targets = squares.new_zeros(
+            (num_valid_squares, self.side_num * 4))
+        bbox_cls_weights = squares.new_zeros(
+            (num_valid_squares, self.side_num * 4))
+        bbox_reg_targets = squares.new_zeros(
+            (num_valid_squares, self.side_num * 4))
+        bbox_reg_weights = squares.new_zeros(
+            (num_valid_squares, self.side_num * 4))
+        labels = squares.new_full((num_valid_squares, ),
+                                  self.num_classes,
+                                  dtype=torch.long)
+        label_weights = squares.new_zeros(num_valid_squares, dtype=torch.float)
+
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+        if len(pos_inds) > 0:
+            (pos_bbox_reg_targets, pos_bbox_reg_weights, pos_bbox_cls_targets,
+             pos_bbox_cls_weights) = self.bbox_coder.encode(
+                 sampling_result.pos_bboxes, sampling_result.pos_gt_bboxes)
+
+            bbox_cls_targets[pos_inds, :] = pos_bbox_cls_targets
+            bbox_reg_targets[pos_inds, :] = pos_bbox_reg_targets
+            bbox_cls_weights[pos_inds, :] = pos_bbox_cls_weights
+            bbox_reg_weights[pos_inds, :] = pos_bbox_reg_weights
+            if gt_labels is None:
+                # Only rpn gives gt_labels as None
+                # Foreground is the first class
+                labels[pos_inds] = 0
+            else:
+                labels[pos_inds] = gt_labels[
+                    sampling_result.pos_assigned_gt_inds]
+            if self.train_cfg.pos_weight <= 0:
+                label_weights[pos_inds] = 1.0
+            else:
+                label_weights[pos_inds] = self.train_cfg.pos_weight
+        if len(neg_inds) > 0:
+            label_weights[neg_inds] = 1.0
+
+        # map up to original set of anchors
+        if unmap_outputs:
+            num_total_anchors = flat_squares.size(0)
+            labels = unmap(
+                labels, num_total_anchors, inside_flags, fill=self.num_classes)
+            label_weights = unmap(label_weights, num_total_anchors,
+                                  inside_flags)
+            bbox_cls_targets = unmap(bbox_cls_targets, num_total_anchors,
+                                     inside_flags)
+            bbox_cls_weights = unmap(bbox_cls_weights, num_total_anchors,
+                                     inside_flags)
+            bbox_reg_targets = unmap(bbox_reg_targets, num_total_anchors,
+                                     inside_flags)
+            bbox_reg_weights = unmap(bbox_reg_weights, num_total_anchors,
+                                     inside_flags)
+        return (labels, label_weights, bbox_cls_targets, bbox_cls_weights,
+                bbox_reg_targets, bbox_reg_weights, pos_inds, neg_inds)
+
+    def loss_single(self, cls_score, bbox_pred, labels, label_weights,
+                    bbox_cls_targets, bbox_cls_weights, bbox_reg_targets,
+                    bbox_reg_weights, num_total_samples):
+        # classification loss
+        labels = labels.reshape(-1)
+        label_weights = label_weights.reshape(-1)
+        cls_score = cls_score.permute(0, 2, 3,
+                                      1).reshape(-1, self.cls_out_channels)
+        loss_cls = self.loss_cls(
+            cls_score, labels, label_weights, avg_factor=num_total_samples)
+        # regression loss
+        bbox_cls_targets = bbox_cls_targets.reshape(-1, self.side_num * 4)
+        bbox_cls_weights = bbox_cls_weights.reshape(-1, self.side_num * 4)
+        bbox_reg_targets = bbox_reg_targets.reshape(-1, self.side_num * 4)
+        bbox_reg_weights = bbox_reg_weights.reshape(-1, self.side_num * 4)
+        (bbox_cls_pred, bbox_reg_pred) = bbox_pred
+        bbox_cls_pred = bbox_cls_pred.permute(0, 2, 3, 1).reshape(
+            -1, self.side_num * 4)
+        bbox_reg_pred = bbox_reg_pred.permute(0, 2, 3, 1).reshape(
+            -1, self.side_num * 4)
+        loss_bbox_cls = self.loss_bbox_cls(
+            bbox_cls_pred,
+            bbox_cls_targets.long(),
+            bbox_cls_weights,
+            avg_factor=num_total_samples * 4 * self.side_num)
+        loss_bbox_reg = self.loss_bbox_reg(
+            bbox_reg_pred,
+            bbox_reg_targets,
+            bbox_reg_weights,
+            avg_factor=num_total_samples * 4 * self.bbox_coder.offset_topk)
+        return loss_cls, loss_bbox_cls, loss_bbox_reg
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.approx_anchor_generator.num_levels
+
+        device = cls_scores[0].device
+
+        # get sampled approxes
+        approxs_list, inside_flag_list = GuidedAnchorHead.get_sampled_approxs(
+            self, featmap_sizes, img_metas, device=device)
+
+        square_list = self.get_anchors(featmap_sizes, img_metas, device=device)
+
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+
+        cls_reg_targets = self.get_target(
+            approxs_list,
+            inside_flag_list,
+            square_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels,
+            sampling=self.sampling)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_cls_targets_list,
+         bbox_cls_weights_list, bbox_reg_targets_list, bbox_reg_weights_list,
+         num_total_pos, num_total_neg) = cls_reg_targets
+        num_total_samples = (
+            num_total_pos + num_total_neg if self.sampling else num_total_pos)
+        losses_cls, losses_bbox_cls, losses_bbox_reg = multi_apply(
+            self.loss_single,
+            cls_scores,
+            bbox_preds,
+            labels_list,
+            label_weights_list,
+            bbox_cls_targets_list,
+            bbox_cls_weights_list,
+            bbox_reg_targets_list,
+            bbox_reg_weights_list,
+            num_total_samples=num_total_samples)
+        return dict(
+            loss_cls=losses_cls,
+            loss_bbox_cls=losses_bbox_cls,
+            loss_bbox_reg=losses_bbox_reg)
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   img_metas,
+                   cfg=None,
+                   rescale=False):
+        assert len(cls_scores) == len(bbox_preds)
+        num_levels = len(cls_scores)
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+
+        device = cls_scores[0].device
+        mlvl_anchors = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        result_list = []
+        for img_id in range(len(img_metas)):
+            cls_score_list = [
+                cls_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_cls_pred_list = [
+                bbox_preds[i][0][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_reg_pred_list = [
+                bbox_preds[i][1][img_id].detach() for i in range(num_levels)
+            ]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            proposals = self.get_bboxes_single(cls_score_list,
+                                               bbox_cls_pred_list,
+                                               bbox_reg_pred_list,
+                                               mlvl_anchors[img_id], img_shape,
+                                               scale_factor, cfg, rescale)
+            result_list.append(proposals)
+        return result_list
+
+    def get_bboxes_single(self,
+                          cls_scores,
+                          bbox_cls_preds,
+                          bbox_reg_preds,
+                          mlvl_anchors,
+                          img_shape,
+                          scale_factor,
+                          cfg,
+                          rescale=False):
+        cfg = self.test_cfg if cfg is None else cfg
+        mlvl_bboxes = []
+        mlvl_scores = []
+        mlvl_confids = []
+        assert len(cls_scores) == len(bbox_cls_preds) == len(
+            bbox_reg_preds) == len(mlvl_anchors)
+        for cls_score, bbox_cls_pred, bbox_reg_pred, anchors in zip(
+                cls_scores, bbox_cls_preds, bbox_reg_preds, mlvl_anchors):
+            assert cls_score.size()[-2:] == bbox_cls_pred.size(
+            )[-2:] == bbox_reg_pred.size()[-2::]
+            cls_score = cls_score.permute(1, 2,
+                                          0).reshape(-1, self.cls_out_channels)
+            if self.use_sigmoid_cls:
+                scores = cls_score.sigmoid()
+            else:
+                scores = cls_score.softmax(-1)
+            bbox_cls_pred = bbox_cls_pred.permute(1, 2, 0).reshape(
+                -1, self.side_num * 4)
+            bbox_reg_pred = bbox_reg_pred.permute(1, 2, 0).reshape(
+                -1, self.side_num * 4)
+            nms_pre = cfg.get('nms_pre', -1)
+            if nms_pre > 0 and scores.shape[0] > nms_pre:
+                if self.use_sigmoid_cls:
+                    max_scores, _ = scores.max(dim=1)
+                else:
+                    max_scores, _ = scores[:, :-1].max(dim=1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                anchors = anchors[topk_inds, :]
+                bbox_cls_pred = bbox_cls_pred[topk_inds, :]
+                bbox_reg_pred = bbox_reg_pred[topk_inds, :]
+                scores = scores[topk_inds, :]
+            bbox_preds = [
+                bbox_cls_pred.contiguous(),
+                bbox_reg_pred.contiguous()
+            ]
+            bboxes, confids = self.bbox_coder.decode(
+                anchors.contiguous(), bbox_preds, max_shape=img_shape)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+            mlvl_confids.append(confids)
+        mlvl_bboxes = torch.cat(mlvl_bboxes)
+        if rescale:
+            mlvl_bboxes /= mlvl_bboxes.new_tensor(scale_factor)
+        mlvl_scores = torch.cat(mlvl_scores)
+        mlvl_confids = torch.cat(mlvl_confids)
+        if self.use_sigmoid_cls:
+            padding = mlvl_scores.new_zeros(mlvl_scores.shape[0], 1)
+            mlvl_scores = torch.cat([mlvl_scores, padding], dim=1)
+        det_bboxes, det_labels = multiclass_nms(
+            mlvl_bboxes,
+            mlvl_scores,
+            cfg.score_thr,
+            cfg.nms,
+            cfg.max_per_img,
+            score_factors=mlvl_confids)
+        return det_bboxes, det_labels
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/ssd_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/ssd_head.py
new file mode 100644
index 0000000..145622b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/ssd_head.py
@@ -0,0 +1,265 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import xavier_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import (build_anchor_generator, build_assigner,
+                        build_bbox_coder, build_sampler, multi_apply)
+from ..builder import HEADS
+from ..losses import smooth_l1_loss
+from .anchor_head import AnchorHead
+
+
+# TODO: add loss evaluator for SSD
+@HEADS.register_module()
+class SSDHead(AnchorHead):
+    """SSD head used in https://arxiv.org/abs/1512.02325.
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        anchor_generator (dict): Config dict for anchor generator
+        bbox_coder (dict): Config of bounding box coder.
+        reg_decoded_bbox (bool): If true, the regression loss would be
+            applied directly on decoded bounding boxes, converting both
+            the predicted boxes and regression targets to absolute
+            coordinates format. Default False. It should be `True` when
+            using `IoULoss`, `GIoULoss`, or `DIoULoss` in the bbox head.
+        train_cfg (dict): Training config of anchor head.
+        test_cfg (dict): Testing config of anchor head.
+    """  # noqa: W605
+
+    def __init__(self,
+                 num_classes=80,
+                 in_channels=(512, 1024, 512, 256, 256, 256),
+                 anchor_generator=dict(
+                     type='SSDAnchorGenerator',
+                     scale_major=False,
+                     input_size=300,
+                     strides=[8, 16, 32, 64, 100, 300],
+                     ratios=([2], [2, 3], [2, 3], [2, 3], [2], [2]),
+                     basesize_ratio_range=(0.1, 0.9)),
+                 bbox_coder=dict(
+                     type='DeltaXYWHBBoxCoder',
+                     clip_border=True,
+                     target_means=[.0, .0, .0, .0],
+                     target_stds=[1.0, 1.0, 1.0, 1.0],
+                 ),
+                 reg_decoded_bbox=False,
+                 train_cfg=None,
+                 test_cfg=None):
+        super(AnchorHead, self).__init__()
+        self.num_classes = num_classes
+        self.in_channels = in_channels
+        self.cls_out_channels = num_classes + 1  # add background class
+        self.anchor_generator = build_anchor_generator(anchor_generator)
+        num_anchors = self.anchor_generator.num_base_anchors
+
+        reg_convs = []
+        cls_convs = []
+        for i in range(len(in_channels)):
+            reg_convs.append(
+                nn.Conv2d(
+                    in_channels[i],
+                    num_anchors[i] * 4,
+                    kernel_size=3,
+                    padding=1))
+            cls_convs.append(
+                nn.Conv2d(
+                    in_channels[i],
+                    num_anchors[i] * (num_classes + 1),
+                    kernel_size=3,
+                    padding=1))
+        self.reg_convs = nn.ModuleList(reg_convs)
+        self.cls_convs = nn.ModuleList(cls_convs)
+
+        self.bbox_coder = build_bbox_coder(bbox_coder)
+        self.reg_decoded_bbox = reg_decoded_bbox
+        self.use_sigmoid_cls = False
+        self.cls_focal_loss = False
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+        # set sampling=False for archor_target
+        self.sampling = False
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            # SSD sampling=False so use PseudoSampler
+            sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+        self.fp16_enabled = False
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                xavier_init(m, distribution='uniform', bias=0)
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple:
+                cls_scores (list[Tensor]): Classification scores for all scale
+                    levels, each is a 4D-tensor, the channels number is
+                    num_anchors * num_classes.
+                bbox_preds (list[Tensor]): Box energies / deltas for all scale
+                    levels, each is a 4D-tensor, the channels number is
+                    num_anchors * 4.
+        """
+        cls_scores = []
+        bbox_preds = []
+        for feat, reg_conv, cls_conv in zip(feats, self.reg_convs,
+                                            self.cls_convs):
+            cls_scores.append(cls_conv(feat))
+            bbox_preds.append(reg_conv(feat))
+        return cls_scores, bbox_preds
+
+    def loss_single(self, cls_score, bbox_pred, anchor, labels, label_weights,
+                    bbox_targets, bbox_weights, num_total_samples):
+        """Compute loss of a single image.
+
+        Args:
+            cls_score (Tensor): Box scores for eachimage
+                Has shape (num_total_anchors, num_classes).
+            bbox_pred (Tensor): Box energies / deltas for each image
+                level with shape (num_total_anchors, 4).
+            anchors (Tensor): Box reference for each scale level with shape
+                (num_total_anchors, 4).
+            labels (Tensor): Labels of each anchors with shape
+                (num_total_anchors,).
+            label_weights (Tensor): Label weights of each anchor with shape
+                (num_total_anchors,)
+            bbox_targets (Tensor): BBox regression targets of each anchor wight
+                shape (num_total_anchors, 4).
+            bbox_weights (Tensor): BBox regression loss weights of each anchor
+                with shape (num_total_anchors, 4).
+            num_total_samples (int): If sampling, num total samples equal to
+                the number of total anchors; Otherwise, it is the number of
+                positive anchors.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+
+        loss_cls_all = F.cross_entropy(
+            cls_score, labels, reduction='none') * label_weights
+        # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+        pos_inds = ((labels >= 0) &
+                    (labels < self.num_classes)).nonzero().reshape(-1)
+        neg_inds = (labels == self.num_classes).nonzero().view(-1)
+
+        num_pos_samples = pos_inds.size(0)
+        num_neg_samples = self.train_cfg.neg_pos_ratio * num_pos_samples
+        if num_neg_samples > neg_inds.size(0):
+            num_neg_samples = neg_inds.size(0)
+        topk_loss_cls_neg, _ = loss_cls_all[neg_inds].topk(num_neg_samples)
+        loss_cls_pos = loss_cls_all[pos_inds].sum()
+        loss_cls_neg = topk_loss_cls_neg.sum()
+        loss_cls = (loss_cls_pos + loss_cls_neg) / num_total_samples
+
+        if self.reg_decoded_bbox:
+            # When the regression loss (e.g. `IouLoss`, `GIouLoss`)
+            # is applied directly on the decoded bounding boxes, it
+            # decodes the already encoded coordinates to absolute format.
+            bbox_pred = self.bbox_coder.decode(anchor, bbox_pred)
+
+        loss_bbox = smooth_l1_loss(
+            bbox_pred,
+            bbox_targets,
+            bbox_weights,
+            beta=self.train_cfg.smoothl1_beta,
+            avg_factor=num_total_samples)
+        return loss_cls[None], loss_bbox
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute losses of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            gt_bboxes (list[Tensor]): each item are the truth boxes for each
+                image in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=1,
+            unmap_outputs=False)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg) = cls_reg_targets
+
+        num_images = len(img_metas)
+        all_cls_scores = torch.cat([
+            s.permute(0, 2, 3, 1).reshape(
+                num_images, -1, self.cls_out_channels) for s in cls_scores
+        ], 1)
+        all_labels = torch.cat(labels_list, -1).view(num_images, -1)
+        all_label_weights = torch.cat(label_weights_list,
+                                      -1).view(num_images, -1)
+        all_bbox_preds = torch.cat([
+            b.permute(0, 2, 3, 1).reshape(num_images, -1, 4)
+            for b in bbox_preds
+        ], -2)
+        all_bbox_targets = torch.cat(bbox_targets_list,
+                                     -2).view(num_images, -1, 4)
+        all_bbox_weights = torch.cat(bbox_weights_list,
+                                     -2).view(num_images, -1, 4)
+
+        # concat all level anchors to a single tensor
+        all_anchors = []
+        for i in range(num_images):
+            all_anchors.append(torch.cat(anchor_list[i]))
+
+        # check NaN and Inf
+        assert torch.isfinite(all_cls_scores).all().item(), \
+            'classification scores become infinite or NaN!'
+        assert torch.isfinite(all_bbox_preds).all().item(), \
+            'bbox predications become infinite or NaN!'
+
+        losses_cls, losses_bbox = multi_apply(
+            self.loss_single,
+            all_cls_scores,
+            all_bbox_preds,
+            all_anchors,
+            all_labels,
+            all_label_weights,
+            all_bbox_targets,
+            all_bbox_weights,
+            num_total_samples=num_total_pos)
+        return dict(loss_cls=losses_cls, loss_bbox=losses_bbox)
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/transformer_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/transformer_head.py
new file mode 100644
index 0000000..820fd06
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/transformer_head.py
@@ -0,0 +1,654 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import Conv2d, Linear, build_activation_layer
+from mmcv.runner import force_fp32
+
+from mmdet.core import (bbox_cxcywh_to_xyxy, bbox_xyxy_to_cxcywh,
+                        build_assigner, build_sampler, multi_apply,
+                        reduce_mean)
+from mmdet.models.utils import (FFN, build_positional_encoding,
+                                build_transformer)
+from ..builder import HEADS, build_loss
+from .anchor_free_head import AnchorFreeHead
+
+
+@HEADS.register_module()
+class TransformerHead(AnchorFreeHead):
+    """Implements the DETR transformer head.
+
+    See `paper: End-to-End Object Detection with Transformers
+    <https://arxiv.org/pdf/2005.12872>`_ for details.
+
+    Args:
+        num_classes (int): Number of categories excluding the background.
+        in_channels (int): Number of channels in the input feature map.
+        num_fcs (int, optional): Number of fully-connected layers used in
+            `FFN`, which is then used for the regression head. Default 2.
+        transformer (dict, optional): Config for transformer.
+        positional_encoding (dict, optional): Config for position encoding.
+        loss_cls (dict, optional): Config of the classification loss.
+            Default `CrossEntropyLoss`.
+        loss_bbox (dict, optional): Config of the regression loss.
+            Default `L1Loss`.
+        loss_iou (dict, optional): Config of the regression iou loss.
+            Default `GIoULoss`.
+        tran_cfg (dict, optional): Training config of transformer head.
+        test_cfg (dict, optional): Testing config of transformer head.
+
+    Example:
+        >>> import torch
+        >>> self = TransformerHead(80, 2048)
+        >>> x = torch.rand(1, 2048, 32, 32)
+        >>> mask = torch.ones(1, 32, 32).to(x.dtype)
+        >>> mask[:, :16, :15] = 0
+        >>> all_cls_scores, all_bbox_preds = self(x, mask)
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 num_fcs=2,
+                 transformer=dict(
+                     type='Transformer',
+                     embed_dims=256,
+                     num_heads=8,
+                     num_encoder_layers=6,
+                     num_decoder_layers=6,
+                     feedforward_channels=2048,
+                     dropout=0.1,
+                     act_cfg=dict(type='ReLU', inplace=True),
+                     norm_cfg=dict(type='LN'),
+                     num_fcs=2,
+                     pre_norm=False,
+                     return_intermediate_dec=True),
+                 positional_encoding=dict(
+                     type='SinePositionalEncoding',
+                     num_feats=128,
+                     normalize=True),
+                 loss_cls=dict(
+                     type='CrossEntropyLoss',
+                     bg_cls_weight=0.1,
+                     use_sigmoid=False,
+                     loss_weight=1.0,
+                     class_weight=1.0),
+                 loss_bbox=dict(type='L1Loss', loss_weight=5.0),
+                 loss_iou=dict(type='GIoULoss', loss_weight=2.0),
+                 train_cfg=dict(
+                     assigner=dict(
+                         type='HungarianAssigner',
+                         cls_cost=dict(type='ClassificationCost', weight=1.),
+                         reg_cost=dict(type='BBoxL1Cost', weight=5.0),
+                         iou_cost=dict(
+                             type='IoUCost', iou_mode='giou', weight=2.0))),
+                 test_cfg=dict(max_per_img=100),
+                 **kwargs):
+        # NOTE here use `AnchorFreeHead` instead of `TransformerHead`,
+        # since it brings inconvenience when the initialization of
+        # `AnchorFreeHead` is called.
+        super(AnchorFreeHead, self).__init__()
+        use_sigmoid_cls = loss_cls.get('use_sigmoid', False)
+        assert not use_sigmoid_cls, 'setting use_sigmoid_cls as True is ' \
+            'not supported in DETR, since background is needed for the ' \
+            'matching process.'
+        assert 'embed_dims' in transformer \
+            and 'num_feats' in positional_encoding
+        num_feats = positional_encoding['num_feats']
+        embed_dims = transformer['embed_dims']
+        assert num_feats * 2 == embed_dims, 'embed_dims should' \
+            f' be exactly 2 times of num_feats. Found {embed_dims}' \
+            f' and {num_feats}.'
+        assert test_cfg is not None and 'max_per_img' in test_cfg
+
+        class_weight = loss_cls.get('class_weight', None)
+        if class_weight is not None:
+            assert isinstance(class_weight, float), 'Expected ' \
+                'class_weight to have type float. Found ' \
+                f'{type(class_weight)}.'
+            # NOTE following the official DETR rep0, bg_cls_weight means
+            # relative classification weight of the no-object class.
+            bg_cls_weight = loss_cls.get('bg_cls_weight', class_weight)
+            assert isinstance(bg_cls_weight, float), 'Expected ' \
+                'bg_cls_weight to have type float. Found ' \
+                f'{type(bg_cls_weight)}.'
+            class_weight = torch.ones(num_classes + 1) * class_weight
+            # set background class as the last indice
+            class_weight[num_classes] = bg_cls_weight
+            loss_cls.update({'class_weight': class_weight})
+            if 'bg_cls_weight' in loss_cls:
+                loss_cls.pop('bg_cls_weight')
+            self.bg_cls_weight = bg_cls_weight
+
+        if train_cfg:
+            assert 'assigner' in train_cfg, 'assigner should be provided '\
+                'when train_cfg is set.'
+            assigner = train_cfg['assigner']
+            assert loss_cls['loss_weight'] == assigner['cls_cost']['weight'], \
+                'The classification weight for loss and matcher should be' \
+                'exactly the same.'
+            assert loss_bbox['loss_weight'] == assigner['reg_cost'][
+                'weight'], 'The regression L1 weight for loss and matcher ' \
+                'should be exactly the same.'
+            assert loss_iou['loss_weight'] == assigner['iou_cost']['weight'], \
+                'The regression iou weight for loss and matcher should be' \
+                'exactly the same.'
+            self.assigner = build_assigner(assigner)
+            # DETR sampling=False, so use PseudoSampler
+            sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+        self.num_classes = num_classes
+        self.cls_out_channels = num_classes + 1
+        self.in_channels = in_channels
+        self.num_fcs = num_fcs
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+        self.use_sigmoid_cls = use_sigmoid_cls
+        self.embed_dims = embed_dims
+        self.num_query = test_cfg['max_per_img']
+        self.fp16_enabled = False
+        self.loss_cls = build_loss(loss_cls)
+        self.loss_bbox = build_loss(loss_bbox)
+        self.loss_iou = build_loss(loss_iou)
+        self.act_cfg = transformer.get('act_cfg',
+                                       dict(type='ReLU', inplace=True))
+        self.activate = build_activation_layer(self.act_cfg)
+        self.positional_encoding = build_positional_encoding(
+            positional_encoding)
+        self.transformer = build_transformer(transformer)
+        self._init_layers()
+
+    def _init_layers(self):
+        """Initialize layers of the transformer head."""
+        self.input_proj = Conv2d(
+            self.in_channels, self.embed_dims, kernel_size=1)
+        self.fc_cls = Linear(self.embed_dims, self.cls_out_channels)
+        self.reg_ffn = FFN(
+            self.embed_dims,
+            self.embed_dims,
+            self.num_fcs,
+            self.act_cfg,
+            dropout=0.0,
+            add_residual=False)
+        self.fc_reg = Linear(self.embed_dims, 4)
+        self.query_embedding = nn.Embedding(self.num_query, self.embed_dims)
+
+    def init_weights(self, distribution='uniform'):
+        """Initialize weights of the transformer head."""
+        # The initialization for transformer is important
+        self.transformer.init_weights()
+
+    def _load_from_state_dict(self, state_dict, prefix, local_metadata, strict,
+                              missing_keys, unexpected_keys, error_msgs):
+        """load checkpoints."""
+        # NOTE here use `AnchorFreeHead` instead of `TransformerHead`,
+        # since `AnchorFreeHead._load_from_state_dict` should not be
+        # called here. Invoking the default `Module._load_from_state_dict`
+        # is enough.
+        super(AnchorFreeHead,
+              self)._load_from_state_dict(state_dict, prefix, local_metadata,
+                                          strict, missing_keys,
+                                          unexpected_keys, error_msgs)
+
+    def forward(self, feats, img_metas):
+        """Forward function.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+            img_metas (list[dict]): List of image information.
+
+        Returns:
+            tuple[list[Tensor], list[Tensor]]: Outputs for all scale levels.
+
+                - all_cls_scores_list (list[Tensor]): Classification scores \
+                    for each scale level. Each is a 4D-tensor with shape \
+                    [nb_dec, bs, num_query, cls_out_channels]. Note \
+                    `cls_out_channels` should includes background.
+                - all_bbox_preds_list (list[Tensor]): Sigmoid regression \
+                    outputs for each scale level. Each is a 4D-tensor with \
+                    normalized coordinate format (cx, cy, w, h) and shape \
+                    [nb_dec, bs, num_query, 4].
+        """
+        num_levels = len(feats)
+        img_metas_list = [img_metas for _ in range(num_levels)]
+        return multi_apply(self.forward_single, feats, img_metas_list)
+
+    def forward_single(self, x, img_metas):
+        """"Forward function for a single feature level.
+
+        Args:
+            x (Tensor): Input feature from backbone's single stage, shape
+                [bs, c, h, w].
+            img_metas (list[dict]): List of image information.
+
+        Returns:
+            all_cls_scores (Tensor): Outputs from the classification head,
+                shape [nb_dec, bs, num_query, cls_out_channels]. Note
+                cls_out_channels should includes background.
+            all_bbox_preds (Tensor): Sigmoid outputs from the regression
+                head with normalized coordinate format (cx, cy, w, h).
+                Shape [nb_dec, bs, num_query, 4].
+        """
+        # construct binary masks which used for the transformer.
+        # NOTE following the official DETR repo, non-zero values representing
+        # ignored positions, while zero values means valid positions.
+        batch_size = x.size(0)
+        input_img_h, input_img_w = img_metas[0]['batch_input_shape']
+        masks = x.new_ones((batch_size, input_img_h, input_img_w))
+        for img_id in range(batch_size):
+            img_h, img_w, _ = img_metas[img_id]['img_shape']
+            masks[img_id, :img_h, :img_w] = 0
+
+        x = self.input_proj(x)
+        # interpolate masks to have the same spatial shape with x
+        masks = F.interpolate(
+            masks.unsqueeze(1), size=x.shape[-2:]).to(torch.bool).squeeze(1)
+        # position encoding
+        pos_embed = self.positional_encoding(masks)  # [bs, embed_dim, h, w]
+        # outs_dec: [nb_dec, bs, num_query, embed_dim]
+        outs_dec, _ = self.transformer(x, masks, self.query_embedding.weight,
+                                       pos_embed)
+
+        all_cls_scores = self.fc_cls(outs_dec)
+        all_bbox_preds = self.fc_reg(self.activate(
+            self.reg_ffn(outs_dec))).sigmoid()
+        return all_cls_scores, all_bbox_preds
+
+    @force_fp32(apply_to=('all_cls_scores_list', 'all_bbox_preds_list'))
+    def loss(self,
+             all_cls_scores_list,
+             all_bbox_preds_list,
+             gt_bboxes_list,
+             gt_labels_list,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """"Loss function.
+
+        Only outputs from the last feature level are used for computing
+        losses by default.
+
+        Args:
+            all_cls_scores_list (list[Tensor]): Classification outputs
+                for each feature level. Each is a 4D-tensor with shape
+                [nb_dec, bs, num_query, cls_out_channels].
+            all_bbox_preds_list (list[Tensor]): Sigmoid regression
+                outputs for each feature level. Each is a 4D-tensor with
+                normalized coordinate format (cx, cy, w, h) and shape
+                [nb_dec, bs, num_query, 4].
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes for each image
+                with shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels_list (list[Tensor]): Ground truth class indices for each
+                image with shape (num_gts, ).
+            img_metas (list[dict]): List of image meta information.
+            gt_bboxes_ignore (list[Tensor], optional): Bounding boxes
+                which can be ignored for each image. Default None.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        # NOTE defaultly only the outputs from the last feature scale is used.
+        all_cls_scores = all_cls_scores_list[-1]
+        all_bbox_preds = all_bbox_preds_list[-1]
+        assert gt_bboxes_ignore is None, \
+            'Only supports for gt_bboxes_ignore setting to None.'
+
+        num_dec_layers = len(all_cls_scores)
+        all_gt_bboxes_list = [gt_bboxes_list for _ in range(num_dec_layers)]
+        all_gt_labels_list = [gt_labels_list for _ in range(num_dec_layers)]
+        all_gt_bboxes_ignore_list = [
+            gt_bboxes_ignore for _ in range(num_dec_layers)
+        ]
+        img_metas_list = [img_metas for _ in range(num_dec_layers)]
+
+        losses_cls, losses_bbox, losses_iou = multi_apply(
+            self.loss_single, all_cls_scores, all_bbox_preds,
+            all_gt_bboxes_list, all_gt_labels_list, img_metas_list,
+            all_gt_bboxes_ignore_list)
+
+        loss_dict = dict()
+        # loss from the last decoder layer
+        loss_dict['loss_cls'] = losses_cls[-1]
+        loss_dict['loss_bbox'] = losses_bbox[-1]
+        loss_dict['loss_iou'] = losses_iou[-1]
+        # loss from other decoder layers
+        num_dec_layer = 0
+        for loss_cls_i, loss_bbox_i, loss_iou_i in zip(losses_cls[:-1],
+                                                       losses_bbox[:-1],
+                                                       losses_iou[:-1]):
+            loss_dict[f'd{num_dec_layer}.loss_cls'] = loss_cls_i
+            loss_dict[f'd{num_dec_layer}.loss_bbox'] = loss_bbox_i
+            loss_dict[f'd{num_dec_layer}.loss_iou'] = loss_iou_i
+            num_dec_layer += 1
+        return loss_dict
+
+    def loss_single(self,
+                    cls_scores,
+                    bbox_preds,
+                    gt_bboxes_list,
+                    gt_labels_list,
+                    img_metas,
+                    gt_bboxes_ignore_list=None):
+        """"Loss function for outputs from a single decoder layer of a single
+        feature level.
+
+        Args:
+            cls_scores (Tensor): Box score logits from a single decoder layer
+                for all images. Shape [bs, num_query, cls_out_channels].
+            bbox_preds (Tensor): Sigmoid outputs from a single decoder layer
+                for all images, with normalized coordinate (cx, cy, w, h) and
+                shape [bs, num_query, 4].
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes for each image
+                with shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels_list (list[Tensor]): Ground truth class indices for each
+                image with shape (num_gts, ).
+            img_metas (list[dict]): List of image meta information.
+            gt_bboxes_ignore_list (list[Tensor], optional): Bounding
+                boxes which can be ignored for each image. Default None.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components for outputs from
+                a single decoder layer.
+        """
+        num_imgs = cls_scores.size(0)
+        cls_scores_list = [cls_scores[i] for i in range(num_imgs)]
+        bbox_preds_list = [bbox_preds[i] for i in range(num_imgs)]
+        cls_reg_targets = self.get_targets(cls_scores_list, bbox_preds_list,
+                                           gt_bboxes_list, gt_labels_list,
+                                           img_metas, gt_bboxes_ignore_list)
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg) = cls_reg_targets
+        labels = torch.cat(labels_list, 0)
+        label_weights = torch.cat(label_weights_list, 0)
+        bbox_targets = torch.cat(bbox_targets_list, 0)
+        bbox_weights = torch.cat(bbox_weights_list, 0)
+
+        # classification loss
+        cls_scores = cls_scores.reshape(-1, self.cls_out_channels)
+        # construct weighted avg_factor to match with the official DETR repo
+        cls_avg_factor = num_total_pos * 1.0 + \
+            num_total_neg * self.bg_cls_weight
+        loss_cls = self.loss_cls(
+            cls_scores, labels, label_weights, avg_factor=cls_avg_factor)
+
+        # Compute the average number of gt boxes accross all gpus, for
+        # normalization purposes
+        num_total_pos = loss_cls.new_tensor([num_total_pos])
+        num_total_pos = torch.clamp(reduce_mean(num_total_pos), min=1).item()
+
+        # construct factors used for rescale bboxes
+        factors = []
+        for img_meta, bbox_pred in zip(img_metas, bbox_preds):
+            img_h, img_w, _ = img_meta['img_shape']
+            factor = bbox_pred.new_tensor([img_w, img_h, img_w,
+                                           img_h]).unsqueeze(0).repeat(
+                                               bbox_pred.size(0), 1)
+            factors.append(factor)
+        factors = torch.cat(factors, 0)
+
+        # DETR regress the relative position of boxes (cxcywh) in the image,
+        # thus the learning target is normalized by the image size. So here
+        # we need to re-scale them for calculating IoU loss
+        bbox_preds = bbox_preds.reshape(-1, 4)
+        bboxes = bbox_cxcywh_to_xyxy(bbox_preds) * factors
+        bboxes_gt = bbox_cxcywh_to_xyxy(bbox_targets) * factors
+
+        # regression IoU loss, defaultly GIoU loss
+        loss_iou = self.loss_iou(
+            bboxes, bboxes_gt, bbox_weights, avg_factor=num_total_pos)
+
+        # regression L1 loss
+        loss_bbox = self.loss_bbox(
+            bbox_preds, bbox_targets, bbox_weights, avg_factor=num_total_pos)
+        return loss_cls, loss_bbox, loss_iou
+
+    def get_targets(self,
+                    cls_scores_list,
+                    bbox_preds_list,
+                    gt_bboxes_list,
+                    gt_labels_list,
+                    img_metas,
+                    gt_bboxes_ignore_list=None):
+        """"Compute regression and classification targets for a batch image.
+
+        Outputs from a single decoder layer of a single feature level are used.
+
+        Args:
+            cls_scores_list (list[Tensor]): Box score logits from a single
+                decoder layer for each image with shape [num_query,
+                cls_out_channels].
+            bbox_preds_list (list[Tensor]): Sigmoid outputs from a single
+                decoder layer for each image, with normalized coordinate
+                (cx, cy, w, h) and shape [num_query, 4].
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes for each image
+                with shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels_list (list[Tensor]): Ground truth class indices for each
+                image with shape (num_gts, ).
+            img_metas (list[dict]): List of image meta information.
+            gt_bboxes_ignore_list (list[Tensor], optional): Bounding
+                boxes which can be ignored for each image. Default None.
+
+        Returns:
+            tuple: a tuple containing the following targets.
+
+                - labels_list (list[Tensor]): Labels for all images.
+                - label_weights_list (list[Tensor]): Label weights for all \
+                    images.
+                - bbox_targets_list (list[Tensor]): BBox targets for all \
+                    images.
+                - bbox_weights_list (list[Tensor]): BBox weights for all \
+                    images.
+                - num_total_pos (int): Number of positive samples in all \
+                    images.
+                - num_total_neg (int): Number of negative samples in all \
+                    images.
+        """
+        assert gt_bboxes_ignore_list is None, \
+            'Only supports for gt_bboxes_ignore setting to None.'
+        num_imgs = len(cls_scores_list)
+        gt_bboxes_ignore_list = [
+            gt_bboxes_ignore_list for _ in range(num_imgs)
+        ]
+
+        (labels_list, label_weights_list, bbox_targets_list,
+         bbox_weights_list, pos_inds_list, neg_inds_list) = multi_apply(
+             self._get_target_single, cls_scores_list, bbox_preds_list,
+             gt_bboxes_list, gt_labels_list, img_metas, gt_bboxes_ignore_list)
+        num_total_pos = sum((inds.numel() for inds in pos_inds_list))
+        num_total_neg = sum((inds.numel() for inds in neg_inds_list))
+        return (labels_list, label_weights_list, bbox_targets_list,
+                bbox_weights_list, num_total_pos, num_total_neg)
+
+    def _get_target_single(self,
+                           cls_score,
+                           bbox_pred,
+                           gt_bboxes,
+                           gt_labels,
+                           img_meta,
+                           gt_bboxes_ignore=None):
+        """"Compute regression and classification targets for one image.
+
+        Outputs from a single decoder layer of a single feature level are used.
+
+        Args:
+            cls_score (Tensor): Box score logits from a single decoder layer
+                for one image. Shape [num_query, cls_out_channels].
+            bbox_pred (Tensor): Sigmoid outputs from a single decoder layer
+                for one image, with normalized coordinate (cx, cy, w, h) and
+                shape [num_query, 4].
+            gt_bboxes (Tensor): Ground truth bboxes for one image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (Tensor): Ground truth class indices for one image
+                with shape (num_gts, ).
+            img_meta (dict): Meta information for one image.
+            gt_bboxes_ignore (Tensor, optional): Bounding boxes
+                which can be ignored. Default None.
+
+        Returns:
+            tuple[Tensor]: a tuple containing the following for one image.
+
+                - labels (Tensor): Labels of each image.
+                - label_weights (Tensor]): Label weights of each image.
+                - bbox_targets (Tensor): BBox targets of each image.
+                - bbox_weights (Tensor): BBox weights of each image.
+                - pos_inds (Tensor): Sampled positive indices for each image.
+                - neg_inds (Tensor): Sampled negative indices for each image.
+        """
+
+        num_bboxes = bbox_pred.size(0)
+        # assigner and sampler
+        assign_result = self.assigner.assign(bbox_pred, cls_score, gt_bboxes,
+                                             gt_labels, img_meta,
+                                             gt_bboxes_ignore)
+        sampling_result = self.sampler.sample(assign_result, bbox_pred,
+                                              gt_bboxes)
+        pos_inds = sampling_result.pos_inds
+        neg_inds = sampling_result.neg_inds
+
+        # label targets
+        labels = gt_bboxes.new_full((num_bboxes, ),
+                                    self.num_classes,
+                                    dtype=torch.long)
+        labels[pos_inds] = gt_labels[sampling_result.pos_assigned_gt_inds]
+        label_weights = gt_bboxes.new_ones(num_bboxes)
+
+        # bbox targets
+        bbox_targets = torch.zeros_like(bbox_pred)
+        bbox_weights = torch.zeros_like(bbox_pred)
+        bbox_weights[pos_inds] = 1.0
+        img_h, img_w, _ = img_meta['img_shape']
+
+        # DETR regress the relative position of boxes (cxcywh) in the image.
+        # Thus the learning target should be normalized by the image size, also
+        # the box format should be converted from defaultly x1y1x2y2 to cxcywh.
+        factor = bbox_pred.new_tensor([img_w, img_h, img_w,
+                                       img_h]).unsqueeze(0)
+        pos_gt_bboxes_normalized = sampling_result.pos_gt_bboxes / factor
+        pos_gt_bboxes_targets = bbox_xyxy_to_cxcywh(pos_gt_bboxes_normalized)
+        bbox_targets[pos_inds] = pos_gt_bboxes_targets
+        return (labels, label_weights, bbox_targets, bbox_weights, pos_inds,
+                neg_inds)
+
+    # over-write because img_metas are needed as inputs for bbox_head.
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels=None,
+                      gt_bboxes_ignore=None,
+                      proposal_cfg=None,
+                      **kwargs):
+        """Forward function for training mode.
+
+        Args:
+            x (list[Tensor]): Features from backbone.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes (Tensor): Ground truth bboxes of the image,
+                shape (num_gts, 4).
+            gt_labels (Tensor): Ground truth labels of each box,
+                shape (num_gts,).
+            gt_bboxes_ignore (Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+            proposal_cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        assert proposal_cfg is None, '"proposal_cfg" must be None'
+        outs = self(x, img_metas)
+        if gt_labels is None:
+            loss_inputs = outs + (gt_bboxes, img_metas)
+        else:
+            loss_inputs = outs + (gt_bboxes, gt_labels, img_metas)
+        losses = self.loss(*loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore)
+        return losses
+
+    @force_fp32(apply_to=('all_cls_scores_list', 'all_bbox_preds_list'))
+    def get_bboxes(self,
+                   all_cls_scores_list,
+                   all_bbox_preds_list,
+                   img_metas,
+                   rescale=False):
+        """Transform network outputs for a batch into bbox predictions.
+
+        Args:
+            all_cls_scores_list (list[Tensor]): Classification outputs
+                for each feature level. Each is a 4D-tensor with shape
+                [nb_dec, bs, num_query, cls_out_channels].
+            all_bbox_preds_list (list[Tensor]): Sigmoid regression
+                outputs for each feature level. Each is a 4D-tensor with
+                normalized coordinate format (cx, cy, w, h) and shape
+                [nb_dec, bs, num_query, 4].
+            img_metas (list[dict]): Meta information of each image.
+            rescale (bool, optional): If True, return boxes in original
+                image space. Default False.
+
+        Returns:
+            list[list[Tensor, Tensor]]: Each item in result_list is 2-tuple. \
+                The first item is an (n, 5) tensor, where the first 4 columns \
+                are bounding box positions (tl_x, tl_y, br_x, br_y) and the \
+                5-th column is a score between 0 and 1. The second item is a \
+                (n,) tensor where each item is the predicted class label of \
+                the corresponding box.
+        """
+        # NOTE defaultly only using outputs from the last feature level,
+        # and only the outputs from the last decoder layer is used.
+        cls_scores = all_cls_scores_list[-1][-1]
+        bbox_preds = all_bbox_preds_list[-1][-1]
+
+        result_list = []
+        for img_id in range(len(img_metas)):
+            cls_score = cls_scores[img_id]
+            bbox_pred = bbox_preds[img_id]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            proposals = self._get_bboxes_single(cls_score, bbox_pred,
+                                                img_shape, scale_factor,
+                                                rescale)
+            result_list.append(proposals)
+        return result_list
+
+    def _get_bboxes_single(self,
+                           cls_score,
+                           bbox_pred,
+                           img_shape,
+                           scale_factor,
+                           rescale=False):
+        """Transform outputs from the last decoder layer into bbox predictions
+        for each image.
+
+        Args:
+            cls_score (Tensor): Box score logits from the last decoder layer
+                for each image. Shape [num_query, cls_out_channels].
+            bbox_pred (Tensor): Sigmoid outputs from the last decoder layer
+                for each image, with coordinate format (cx, cy, w, h) and
+                shape [num_query, 4].
+            img_shape (tuple[int]): Shape of input image, (height, width, 3).
+            scale_factor (ndarray, optional): Scale factor of the image arange
+                as (w_scale, h_scale, w_scale, h_scale).
+            rescale (bool, optional): If True, return boxes in original image
+                space. Default False.
+
+        Returns:
+            tuple[Tensor]: Results of detected bboxes and labels.
+
+                - det_bboxes: Predicted bboxes with shape [num_query, 5], \
+                    where the first 4 columns are bounding box positions \
+                    (tl_x, tl_y, br_x, br_y) and the 5-th column are scores \
+                    between 0 and 1.
+                - det_labels: Predicted labels of the corresponding box with \
+                    shape [num_query].
+        """
+        assert len(cls_score) == len(bbox_pred)
+        # exclude background
+        scores, det_labels = F.softmax(cls_score, dim=-1)[..., :-1].max(-1)
+        det_bboxes = bbox_cxcywh_to_xyxy(bbox_pred)
+        det_bboxes[:, 0::2] = det_bboxes[:, 0::2] * img_shape[1]
+        det_bboxes[:, 1::2] = det_bboxes[:, 1::2] * img_shape[0]
+        det_bboxes[:, 0::2].clamp_(min=0, max=img_shape[1])
+        det_bboxes[:, 1::2].clamp_(min=0, max=img_shape[0])
+        if rescale:
+            det_bboxes /= det_bboxes.new_tensor(scale_factor)
+        det_bboxes = torch.cat((det_bboxes, scores.unsqueeze(1)), -1)
+        return det_bboxes, det_labels
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/vfnet_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/vfnet_head.py
new file mode 100644
index 0000000..7243bb6
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/vfnet_head.py
@@ -0,0 +1,794 @@
+import numpy as np
+import torch
+import torch.nn as nn
+from mmcv.cnn import ConvModule, Scale, bias_init_with_prob, normal_init
+from mmcv.ops import DeformConv2d
+from mmcv.runner import force_fp32
+
+from mmdet.core import (bbox2distance, bbox_overlaps, build_anchor_generator,
+                        build_assigner, build_sampler, distance2bbox,
+                        multi_apply, multiclass_nms, reduce_mean)
+from ..builder import HEADS, build_loss
+from .atss_head import ATSSHead
+from .fcos_head import FCOSHead
+
+INF = 1e8
+
+
+@HEADS.register_module()
+class VFNetHead(ATSSHead, FCOSHead):
+    """Head of `VarifocalNet (VFNet): An IoU-aware Dense Object
+    Detector.<https://arxiv.org/abs/2008.13367>`_.
+
+    The VFNet predicts IoU-aware classification scores which mix the
+    object presence confidence and object localization accuracy as the
+    detection score. It is built on the FCOS architecture and uses ATSS
+    for defining positive/negative training examples. The VFNet is trained
+    with Varifocal Loss and empolys star-shaped deformable convolution to
+    extract features for a bbox.
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        regress_ranges (tuple[tuple[int, int]]): Regress range of multiple
+            level points.
+        center_sampling (bool): If true, use center sampling. Default: False.
+        center_sample_radius (float): Radius of center sampling. Default: 1.5.
+        sync_num_pos (bool): If true, synchronize the number of positive
+            examples across GPUs. Default: True
+        gradient_mul (float): The multiplier to gradients from bbox refinement
+            and recognition. Default: 0.1.
+        bbox_norm_type (str): The bbox normalization type, 'reg_denom' or
+            'stride'. Default: reg_denom
+        loss_cls_fl (dict): Config of focal loss.
+        use_vfl (bool): If true, use varifocal loss for training.
+            Default: True.
+        loss_cls (dict): Config of varifocal loss.
+        loss_bbox (dict): Config of localization loss, GIoU Loss.
+        loss_bbox (dict): Config of localization refinement loss, GIoU Loss.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+            Default: norm_cfg=dict(type='GN', num_groups=32,
+            requires_grad=True).
+        use_atss (bool): If true, use ATSS to define positive/negative
+            examples. Default: True.
+        anchor_generator (dict): Config of anchor generator for ATSS.
+
+    Example:
+        >>> self = VFNetHead(11, 7)
+        >>> feats = [torch.rand(1, 7, s, s) for s in [4, 8, 16, 32, 64]]
+        >>> cls_score, bbox_pred, bbox_pred_refine= self.forward(feats)
+        >>> assert len(cls_score) == len(self.scales)
+    """  # noqa: E501
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 regress_ranges=((-1, 64), (64, 128), (128, 256), (256, 512),
+                                 (512, INF)),
+                 center_sampling=False,
+                 center_sample_radius=1.5,
+                 sync_num_pos=True,
+                 gradient_mul=0.1,
+                 bbox_norm_type='reg_denom',
+                 loss_cls_fl=dict(
+                     type='FocalLoss',
+                     use_sigmoid=True,
+                     gamma=2.0,
+                     alpha=0.25,
+                     loss_weight=1.0),
+                 use_vfl=True,
+                 loss_cls=dict(
+                     type='VarifocalLoss',
+                     use_sigmoid=True,
+                     alpha=0.75,
+                     gamma=2.0,
+                     iou_weighted=True,
+                     loss_weight=1.0),
+                 loss_bbox=dict(type='GIoULoss', loss_weight=1.5),
+                 loss_bbox_refine=dict(type='GIoULoss', loss_weight=2.0),
+                 norm_cfg=dict(type='GN', num_groups=32, requires_grad=True),
+                 use_atss=True,
+                 anchor_generator=dict(
+                     type='AnchorGenerator',
+                     ratios=[1.0],
+                     octave_base_scale=8,
+                     scales_per_octave=1,
+                     center_offset=0.0,
+                     strides=[8, 16, 32, 64, 128]),
+                 **kwargs):
+        # dcn base offsets, adapted from reppoints_head.py
+        self.num_dconv_points = 9
+        self.dcn_kernel = int(np.sqrt(self.num_dconv_points))
+        self.dcn_pad = int((self.dcn_kernel - 1) / 2)
+        dcn_base = np.arange(-self.dcn_pad,
+                             self.dcn_pad + 1).astype(np.float64)
+        dcn_base_y = np.repeat(dcn_base, self.dcn_kernel)
+        dcn_base_x = np.tile(dcn_base, self.dcn_kernel)
+        dcn_base_offset = np.stack([dcn_base_y, dcn_base_x], axis=1).reshape(
+            (-1))
+        self.dcn_base_offset = torch.tensor(dcn_base_offset).view(1, -1, 1, 1)
+
+        super(FCOSHead, self).__init__(
+            num_classes, in_channels, norm_cfg=norm_cfg, **kwargs)
+        self.regress_ranges = regress_ranges
+        self.reg_denoms = [
+            regress_range[-1] for regress_range in regress_ranges
+        ]
+        self.reg_denoms[-1] = self.reg_denoms[-2] * 2
+        self.center_sampling = center_sampling
+        self.center_sample_radius = center_sample_radius
+        self.sync_num_pos = sync_num_pos
+        self.bbox_norm_type = bbox_norm_type
+        self.gradient_mul = gradient_mul
+        self.use_vfl = use_vfl
+        if self.use_vfl:
+            self.loss_cls = build_loss(loss_cls)
+        else:
+            self.loss_cls = build_loss(loss_cls_fl)
+        self.loss_bbox = build_loss(loss_bbox)
+        self.loss_bbox_refine = build_loss(loss_bbox_refine)
+
+        # for getting ATSS targets
+        self.use_atss = use_atss
+        self.use_sigmoid_cls = loss_cls.get('use_sigmoid', False)
+        self.anchor_generator = build_anchor_generator(anchor_generator)
+        self.anchor_center_offset = anchor_generator['center_offset']
+        self.num_anchors = self.anchor_generator.num_base_anchors[0]
+        self.sampling = False
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        super(FCOSHead, self)._init_cls_convs()
+        super(FCOSHead, self)._init_reg_convs()
+        self.relu = nn.ReLU(inplace=True)
+        self.vfnet_reg_conv = ConvModule(
+            self.feat_channels,
+            self.feat_channels,
+            3,
+            stride=1,
+            padding=1,
+            conv_cfg=self.conv_cfg,
+            norm_cfg=self.norm_cfg,
+            bias=self.conv_bias)
+        self.vfnet_reg = nn.Conv2d(self.feat_channels, 4, 3, padding=1)
+        self.scales = nn.ModuleList([Scale(1.0) for _ in self.strides])
+
+        self.vfnet_reg_refine_dconv = DeformConv2d(
+            self.feat_channels,
+            self.feat_channels,
+            self.dcn_kernel,
+            1,
+            padding=self.dcn_pad)
+        self.vfnet_reg_refine = nn.Conv2d(self.feat_channels, 4, 3, padding=1)
+        self.scales_refine = nn.ModuleList([Scale(1.0) for _ in self.strides])
+
+        self.vfnet_cls_dconv = DeformConv2d(
+            self.feat_channels,
+            self.feat_channels,
+            self.dcn_kernel,
+            1,
+            padding=self.dcn_pad)
+        self.vfnet_cls = nn.Conv2d(
+            self.feat_channels, self.cls_out_channels, 3, padding=1)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.cls_convs:
+            if isinstance(m.conv, nn.Conv2d):
+                normal_init(m.conv, std=0.01)
+        for m in self.reg_convs:
+            if isinstance(m.conv, nn.Conv2d):
+                normal_init(m.conv, std=0.01)
+        normal_init(self.vfnet_reg_conv.conv, std=0.01)
+        normal_init(self.vfnet_reg, std=0.01)
+        normal_init(self.vfnet_reg_refine_dconv, std=0.01)
+        normal_init(self.vfnet_reg_refine, std=0.01)
+        normal_init(self.vfnet_cls_dconv, std=0.01)
+        bias_cls = bias_init_with_prob(0.01)
+        normal_init(self.vfnet_cls, std=0.01, bias=bias_cls)
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple:
+                cls_scores (list[Tensor]): Box iou-aware scores for each scale
+                    level, each is a 4D-tensor, the channel number is
+                    num_points * num_classes.
+                bbox_preds (list[Tensor]): Box offsets for each
+                    scale level, each is a 4D-tensor, the channel number is
+                    num_points * 4.
+                bbox_preds_refine (list[Tensor]): Refined Box offsets for
+                    each scale level, each is a 4D-tensor, the channel
+                    number is num_points * 4.
+        """
+        return multi_apply(self.forward_single, feats, self.scales,
+                           self.scales_refine, self.strides, self.reg_denoms)
+
+    def forward_single(self, x, scale, scale_refine, stride, reg_denom):
+        """Forward features of a single scale level.
+
+        Args:
+            x (Tensor): FPN feature maps of the specified stride.
+            scale (:obj: `mmcv.cnn.Scale`): Learnable scale module to resize
+                the bbox prediction.
+            scale_refine (:obj: `mmcv.cnn.Scale`): Learnable scale module to
+                resize the refined bbox prediction.
+            stride (int): The corresponding stride for feature maps,
+                used to normalize the bbox prediction when
+                bbox_norm_type = 'stride'.
+            reg_denom (int): The corresponding regression range for feature
+                maps, only used to normalize the bbox prediction when
+                bbox_norm_type = 'reg_denom'.
+
+        Returns:
+            tuple: iou-aware cls scores for each box, bbox predictions and
+                refined bbox predictions of input feature maps.
+        """
+        cls_feat = x
+        reg_feat = x
+
+        for cls_layer in self.cls_convs:
+            cls_feat = cls_layer(cls_feat)
+
+        for reg_layer in self.reg_convs:
+            reg_feat = reg_layer(reg_feat)
+
+        # predict the bbox_pred of different level
+        reg_feat_init = self.vfnet_reg_conv(reg_feat)
+        if self.bbox_norm_type == 'reg_denom':
+            bbox_pred = scale(
+                self.vfnet_reg(reg_feat_init)).float().exp() * reg_denom
+        elif self.bbox_norm_type == 'stride':
+            bbox_pred = scale(
+                self.vfnet_reg(reg_feat_init)).float().exp() * stride
+        else:
+            raise NotImplementedError
+
+        # compute star deformable convolution offsets
+        # converting dcn_offset to reg_feat.dtype thus VFNet can be
+        # trained with FP16
+        dcn_offset = self.star_dcn_offset(bbox_pred, self.gradient_mul,
+                                          stride).to(reg_feat.dtype)
+
+        # refine the bbox_pred
+        reg_feat = self.relu(self.vfnet_reg_refine_dconv(reg_feat, dcn_offset))
+        bbox_pred_refine = scale_refine(
+            self.vfnet_reg_refine(reg_feat)).float().exp()
+        bbox_pred_refine = bbox_pred_refine * bbox_pred.detach()
+
+        # predict the iou-aware cls score
+        cls_feat = self.relu(self.vfnet_cls_dconv(cls_feat, dcn_offset))
+        cls_score = self.vfnet_cls(cls_feat)
+
+        return cls_score, bbox_pred, bbox_pred_refine
+
+    def star_dcn_offset(self, bbox_pred, gradient_mul, stride):
+        """Compute the star deformable conv offsets.
+
+        Args:
+            bbox_pred (Tensor): Predicted bbox distance offsets (l, r, t, b).
+            gradient_mul (float): Gradient multiplier.
+            stride (int): The corresponding stride for feature maps,
+                used to project the bbox onto the feature map.
+
+        Returns:
+            dcn_offsets (Tensor): The offsets for deformable convolution.
+        """
+        dcn_base_offset = self.dcn_base_offset.type_as(bbox_pred)
+        bbox_pred_grad_mul = (1 - gradient_mul) * bbox_pred.detach() + \
+            gradient_mul * bbox_pred
+        # map to the feature map scale
+        bbox_pred_grad_mul = bbox_pred_grad_mul / stride
+        N, C, H, W = bbox_pred.size()
+
+        x1 = bbox_pred_grad_mul[:, 0, :, :]
+        y1 = bbox_pred_grad_mul[:, 1, :, :]
+        x2 = bbox_pred_grad_mul[:, 2, :, :]
+        y2 = bbox_pred_grad_mul[:, 3, :, :]
+        bbox_pred_grad_mul_offset = bbox_pred.new_zeros(
+            N, 2 * self.num_dconv_points, H, W)
+        bbox_pred_grad_mul_offset[:, 0, :, :] = -1.0 * y1  # -y1
+        bbox_pred_grad_mul_offset[:, 1, :, :] = -1.0 * x1  # -x1
+        bbox_pred_grad_mul_offset[:, 2, :, :] = -1.0 * y1  # -y1
+        bbox_pred_grad_mul_offset[:, 4, :, :] = -1.0 * y1  # -y1
+        bbox_pred_grad_mul_offset[:, 5, :, :] = x2  # x2
+        bbox_pred_grad_mul_offset[:, 7, :, :] = -1.0 * x1  # -x1
+        bbox_pred_grad_mul_offset[:, 11, :, :] = x2  # x2
+        bbox_pred_grad_mul_offset[:, 12, :, :] = y2  # y2
+        bbox_pred_grad_mul_offset[:, 13, :, :] = -1.0 * x1  # -x1
+        bbox_pred_grad_mul_offset[:, 14, :, :] = y2  # y2
+        bbox_pred_grad_mul_offset[:, 16, :, :] = y2  # y2
+        bbox_pred_grad_mul_offset[:, 17, :, :] = x2  # x2
+        dcn_offset = bbox_pred_grad_mul_offset - dcn_base_offset
+
+        return dcn_offset
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds', 'bbox_preds_refine'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             bbox_preds_refine,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute loss of the head.
+
+        Args:
+            cls_scores (list[Tensor]): Box iou-aware scores for each scale
+                level, each is a 4D-tensor, the channel number is
+                num_points * num_classes.
+            bbox_preds (list[Tensor]): Box offsets for each
+                scale level, each is a 4D-tensor, the channel number is
+                num_points * 4.
+            bbox_preds_refine (list[Tensor]): Refined Box offsets for
+                each scale level, each is a 4D-tensor, the channel
+                number is num_points * 4.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+                Default: None.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        assert len(cls_scores) == len(bbox_preds) == len(bbox_preds_refine)
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        all_level_points = self.get_points(featmap_sizes, bbox_preds[0].dtype,
+                                           bbox_preds[0].device)
+        labels, label_weights, bbox_targets, bbox_weights = self.get_targets(
+            cls_scores, all_level_points, gt_bboxes, gt_labels, img_metas,
+            gt_bboxes_ignore)
+
+        num_imgs = cls_scores[0].size(0)
+        # flatten cls_scores, bbox_preds and bbox_preds_refine
+        flatten_cls_scores = [
+            cls_score.permute(0, 2, 3,
+                              1).reshape(-1,
+                                         self.cls_out_channels).contiguous()
+            for cls_score in cls_scores
+        ]
+        flatten_bbox_preds = [
+            bbox_pred.permute(0, 2, 3, 1).reshape(-1, 4).contiguous()
+            for bbox_pred in bbox_preds
+        ]
+        flatten_bbox_preds_refine = [
+            bbox_pred_refine.permute(0, 2, 3, 1).reshape(-1, 4).contiguous()
+            for bbox_pred_refine in bbox_preds_refine
+        ]
+        flatten_cls_scores = torch.cat(flatten_cls_scores)
+        flatten_bbox_preds = torch.cat(flatten_bbox_preds)
+        flatten_bbox_preds_refine = torch.cat(flatten_bbox_preds_refine)
+        flatten_labels = torch.cat(labels)
+        flatten_bbox_targets = torch.cat(bbox_targets)
+        # repeat points to align with bbox_preds
+        flatten_points = torch.cat(
+            [points.repeat(num_imgs, 1) for points in all_level_points])
+
+        # FG cat_id: [0, num_classes - 1], BG cat_id: num_classes
+        bg_class_ind = self.num_classes
+        pos_inds = torch.where(
+            ((flatten_labels >= 0) & (flatten_labels < bg_class_ind)) > 0)[0]
+        num_pos = len(pos_inds)
+
+        pos_bbox_preds = flatten_bbox_preds[pos_inds]
+        pos_bbox_preds_refine = flatten_bbox_preds_refine[pos_inds]
+        pos_labels = flatten_labels[pos_inds]
+
+        # sync num_pos across all gpus
+        if self.sync_num_pos:
+            num_pos_avg_per_gpu = reduce_mean(
+                pos_inds.new_tensor(num_pos).float()).item()
+            num_pos_avg_per_gpu = max(num_pos_avg_per_gpu, 1.0)
+        else:
+            num_pos_avg_per_gpu = num_pos
+
+        if num_pos > 0:
+            pos_bbox_targets = flatten_bbox_targets[pos_inds]
+            pos_points = flatten_points[pos_inds]
+
+            pos_decoded_bbox_preds = distance2bbox(pos_points, pos_bbox_preds)
+            pos_decoded_target_preds = distance2bbox(pos_points,
+                                                     pos_bbox_targets)
+            iou_targets_ini = bbox_overlaps(
+                pos_decoded_bbox_preds,
+                pos_decoded_target_preds.detach(),
+                is_aligned=True).clamp(min=1e-6)
+            bbox_weights_ini = iou_targets_ini.clone().detach()
+            iou_targets_ini_avg_per_gpu = reduce_mean(
+                bbox_weights_ini.sum()).item()
+            bbox_avg_factor_ini = max(iou_targets_ini_avg_per_gpu, 1.0)
+            loss_bbox = self.loss_bbox(
+                pos_decoded_bbox_preds,
+                pos_decoded_target_preds.detach(),
+                weight=bbox_weights_ini,
+                avg_factor=bbox_avg_factor_ini)
+
+            pos_decoded_bbox_preds_refine = \
+                distance2bbox(pos_points, pos_bbox_preds_refine)
+            iou_targets_rf = bbox_overlaps(
+                pos_decoded_bbox_preds_refine,
+                pos_decoded_target_preds.detach(),
+                is_aligned=True).clamp(min=1e-6)
+            bbox_weights_rf = iou_targets_rf.clone().detach()
+            iou_targets_rf_avg_per_gpu = reduce_mean(
+                bbox_weights_rf.sum()).item()
+            bbox_avg_factor_rf = max(iou_targets_rf_avg_per_gpu, 1.0)
+            loss_bbox_refine = self.loss_bbox_refine(
+                pos_decoded_bbox_preds_refine,
+                pos_decoded_target_preds.detach(),
+                weight=bbox_weights_rf,
+                avg_factor=bbox_avg_factor_rf)
+
+            # build IoU-aware cls_score targets
+            if self.use_vfl:
+                pos_ious = iou_targets_rf.clone().detach()
+                cls_iou_targets = torch.zeros_like(flatten_cls_scores)
+                cls_iou_targets[pos_inds, pos_labels] = pos_ious
+        else:
+            loss_bbox = pos_bbox_preds.sum() * 0
+            loss_bbox_refine = pos_bbox_preds_refine.sum() * 0
+            if self.use_vfl:
+                cls_iou_targets = torch.zeros_like(flatten_cls_scores)
+
+        if self.use_vfl:
+            loss_cls = self.loss_cls(
+                flatten_cls_scores,
+                cls_iou_targets,
+                avg_factor=num_pos_avg_per_gpu)
+        else:
+            loss_cls = self.loss_cls(
+                flatten_cls_scores,
+                flatten_labels,
+                weight=label_weights,
+                avg_factor=num_pos_avg_per_gpu)
+
+        return dict(
+            loss_cls=loss_cls,
+            loss_bbox=loss_bbox,
+            loss_bbox_rf=loss_bbox_refine)
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds', 'bbox_preds_refine'))
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   bbox_preds_refine,
+                   img_metas,
+                   cfg=None,
+                   rescale=None,
+                   with_nms=True):
+        """Transform network outputs for a batch into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box iou-aware scores for each scale
+                level with shape (N, num_points * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box offsets for each scale
+                level with shape (N, num_points * 4, H, W).
+            bbox_preds_refine (list[Tensor]): Refined Box offsets for
+                each scale level with shape (N, num_points * 4, H, W).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used. Default: None.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before returning boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where the first 4 columns
+                are bounding box positions (tl_x, tl_y, br_x, br_y) and the
+                5-th column is a score between 0 and 1. The second item is a
+                (n,) tensor where each item is the predicted class label of
+                the corresponding box.
+        """
+        assert len(cls_scores) == len(bbox_preds) == len(bbox_preds_refine)
+        num_levels = len(cls_scores)
+
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        mlvl_points = self.get_points(featmap_sizes, bbox_preds[0].dtype,
+                                      bbox_preds[0].device)
+        result_list = []
+        for img_id in range(len(img_metas)):
+            cls_score_list = [
+                cls_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_pred_list = [
+                bbox_preds_refine[i][img_id].detach()
+                for i in range(num_levels)
+            ]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            det_bboxes = self._get_bboxes_single(cls_score_list,
+                                                 bbox_pred_list, mlvl_points,
+                                                 img_shape, scale_factor, cfg,
+                                                 rescale, with_nms)
+            result_list.append(det_bboxes)
+        return result_list
+
+    def _get_bboxes_single(self,
+                           cls_scores,
+                           bbox_preds,
+                           mlvl_points,
+                           img_shape,
+                           scale_factor,
+                           cfg,
+                           rescale=False,
+                           with_nms=True):
+        """Transform outputs for a single batch item into bbox predictions.
+
+        Args:
+            cls_scores (list[Tensor]): Box iou-aware scores for a single scale
+                level with shape (num_points * num_classes, H, W).
+            bbox_preds (list[Tensor]): Box offsets for a single scale
+                level with shape (num_points * 4, H, W).
+            mlvl_points (list[Tensor]): Box reference for a single scale level
+                with shape (num_total_points, 4).
+            img_shape (tuple[int]): Shape of the input image,
+                (height, width, 3).
+            scale_factor (ndarray): Scale factor of the image arrange as
+                (w_scale, h_scale, w_scale, h_scale).
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before returning boxes.
+                Default: True.
+
+        Returns:
+            tuple(Tensor):
+                det_bboxes (Tensor): BBox predictions in shape (n, 5), where
+                    the first 4 columns are bounding box positions
+                    (tl_x, tl_y, br_x, br_y) and the 5-th column is a score
+                    between 0 and 1.
+                det_labels (Tensor): A (n,) tensor where each item is the
+                    predicted class label of the corresponding box.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_scores) == len(bbox_preds) == len(mlvl_points)
+        mlvl_bboxes = []
+        mlvl_scores = []
+        for cls_score, bbox_pred, points in zip(cls_scores, bbox_preds,
+                                                mlvl_points):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            scores = cls_score.permute(1, 2, 0).reshape(
+                -1, self.cls_out_channels).contiguous().sigmoid()
+            bbox_pred = bbox_pred.permute(1, 2, 0).reshape(-1, 4).contiguous()
+
+            nms_pre = cfg.get('nms_pre', -1)
+            if 0 < nms_pre < scores.shape[0]:
+                max_scores, _ = scores.max(dim=1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                points = points[topk_inds, :]
+                bbox_pred = bbox_pred[topk_inds, :]
+                scores = scores[topk_inds, :]
+            bboxes = distance2bbox(points, bbox_pred, max_shape=img_shape)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+        mlvl_bboxes = torch.cat(mlvl_bboxes)
+        if rescale:
+            mlvl_bboxes /= mlvl_bboxes.new_tensor(scale_factor)
+        mlvl_scores = torch.cat(mlvl_scores)
+        padding = mlvl_scores.new_zeros(mlvl_scores.shape[0], 1)
+        # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+        # BG cat_id: num_class
+        mlvl_scores = torch.cat([mlvl_scores, padding], dim=1)
+        if with_nms:
+            det_bboxes, det_labels = multiclass_nms(mlvl_bboxes, mlvl_scores,
+                                                    cfg.score_thr, cfg.nms,
+                                                    cfg.max_per_img)
+            return det_bboxes, det_labels
+        else:
+            return mlvl_bboxes, mlvl_scores
+
+    def _get_points_single(self,
+                           featmap_size,
+                           stride,
+                           dtype,
+                           device,
+                           flatten=False):
+        """Get points according to feature map sizes."""
+        h, w = featmap_size
+        x_range = torch.arange(
+            0, w * stride, stride, dtype=dtype, device=device)
+        y_range = torch.arange(
+            0, h * stride, stride, dtype=dtype, device=device)
+        y, x = torch.meshgrid(y_range, x_range)
+        # to be compatible with anchor points in ATSS
+        if self.use_atss:
+            points = torch.stack(
+                (x.reshape(-1), y.reshape(-1)), dim=-1) + \
+                     stride * self.anchor_center_offset
+        else:
+            points = torch.stack(
+                (x.reshape(-1), y.reshape(-1)), dim=-1) + stride // 2
+        return points
+
+    def get_targets(self, cls_scores, mlvl_points, gt_bboxes, gt_labels,
+                    img_metas, gt_bboxes_ignore):
+        """A wrapper for computing ATSS and FCOS targets for points in multiple
+        images.
+
+        Args:
+            cls_scores (list[Tensor]): Box iou-aware scores for each scale
+                level with shape (N, num_points * num_classes, H, W).
+            mlvl_points (list[Tensor]): Points of each fpn level, each has
+                shape (num_points, 2).
+            gt_bboxes (list[Tensor]): Ground truth bboxes of each image,
+                each has shape (num_gt, 4).
+            gt_labels (list[Tensor]): Ground truth labels of each box,
+                each has shape (num_gt,).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4).
+
+        Returns:
+            tuple:
+                labels_list (list[Tensor]): Labels of each level.
+                label_weights (Tensor/None): Label weights of all levels.
+                bbox_targets_list (list[Tensor]): Regression targets of each
+                    level, (l, t, r, b).
+                bbox_weights (Tensor/None): Bbox weights of all levels.
+        """
+        if self.use_atss:
+            return self.get_atss_targets(cls_scores, mlvl_points, gt_bboxes,
+                                         gt_labels, img_metas,
+                                         gt_bboxes_ignore)
+        else:
+            self.norm_on_bbox = False
+            return self.get_fcos_targets(mlvl_points, gt_bboxes, gt_labels)
+
+    def _get_target_single(self, *args, **kwargs):
+        """Avoid ambiguity in multiple inheritance."""
+        if self.use_atss:
+            return ATSSHead._get_target_single(self, *args, **kwargs)
+        else:
+            return FCOSHead._get_target_single(self, *args, **kwargs)
+
+    def get_fcos_targets(self, points, gt_bboxes_list, gt_labels_list):
+        """Compute FCOS regression and classification targets for points in
+        multiple images.
+
+        Args:
+            points (list[Tensor]): Points of each fpn level, each has shape
+                (num_points, 2).
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image,
+                each has shape (num_gt, 4).
+            gt_labels_list (list[Tensor]): Ground truth labels of each box,
+                each has shape (num_gt,).
+
+        Returns:
+            tuple:
+                labels (list[Tensor]): Labels of each level.
+                label_weights: None, to be compatible with ATSS targets.
+                bbox_targets (list[Tensor]): BBox targets of each level.
+                bbox_weights: None, to be compatible with ATSS targets.
+        """
+        labels, bbox_targets = FCOSHead.get_targets(self, points,
+                                                    gt_bboxes_list,
+                                                    gt_labels_list)
+        label_weights = None
+        bbox_weights = None
+        return labels, label_weights, bbox_targets, bbox_weights
+
+    def get_atss_targets(self,
+                         cls_scores,
+                         mlvl_points,
+                         gt_bboxes,
+                         gt_labels,
+                         img_metas,
+                         gt_bboxes_ignore=None):
+        """A wrapper for computing ATSS targets for points in multiple images.
+
+        Args:
+            cls_scores (list[Tensor]): Box iou-aware scores for each scale
+                level with shape (N, num_points * num_classes, H, W).
+            mlvl_points (list[Tensor]): Points of each fpn level, each has
+                shape (num_points, 2).
+            gt_bboxes (list[Tensor]): Ground truth bboxes of each image,
+                each has shape (num_gt, 4).
+            gt_labels (list[Tensor]): Ground truth labels of each box,
+                each has shape (num_gt,).
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | Tensor): Ground truth bboxes to be
+                ignored, shape (num_ignored_gts, 4). Default: None.
+
+        Returns:
+            tuple:
+                labels_list (list[Tensor]): Labels of each level.
+                label_weights (Tensor): Label weights of all levels.
+                bbox_targets_list (list[Tensor]): Regression targets of each
+                    level, (l, t, r, b).
+                bbox_weights (Tensor): Bbox weights of all levels.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+
+        cls_reg_targets = ATSSHead.get_targets(
+            self,
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels,
+            unmap_outputs=True)
+        if cls_reg_targets is None:
+            return None
+
+        (anchor_list, labels_list, label_weights_list, bbox_targets_list,
+         bbox_weights_list, num_total_pos, num_total_neg) = cls_reg_targets
+
+        bbox_targets_list = [
+            bbox_targets.reshape(-1, 4) for bbox_targets in bbox_targets_list
+        ]
+
+        num_imgs = len(img_metas)
+        # transform bbox_targets (x1, y1, x2, y2) into (l, t, r, b) format
+        bbox_targets_list = self.transform_bbox_targets(
+            bbox_targets_list, mlvl_points, num_imgs)
+
+        labels_list = [labels.reshape(-1) for labels in labels_list]
+        label_weights_list = [
+            label_weights.reshape(-1) for label_weights in label_weights_list
+        ]
+        bbox_weights_list = [
+            bbox_weights.reshape(-1) for bbox_weights in bbox_weights_list
+        ]
+        label_weights = torch.cat(label_weights_list)
+        bbox_weights = torch.cat(bbox_weights_list)
+        return labels_list, label_weights, bbox_targets_list, bbox_weights
+
+    def transform_bbox_targets(self, decoded_bboxes, mlvl_points, num_imgs):
+        """Transform bbox_targets (x1, y1, x2, y2) into (l, t, r, b) format.
+
+        Args:
+            decoded_bboxes (list[Tensor]): Regression targets of each level,
+                in the form of (x1, y1, x2, y2).
+            mlvl_points (list[Tensor]): Points of each fpn level, each has
+                shape (num_points, 2).
+            num_imgs (int): the number of images in a batch.
+
+        Returns:
+            bbox_targets (list[Tensor]): Regression targets of each level in
+                the form of (l, t, r, b).
+        """
+        # TODO: Re-implemented in Class PointCoder
+        assert len(decoded_bboxes) == len(mlvl_points)
+        num_levels = len(decoded_bboxes)
+        mlvl_points = [points.repeat(num_imgs, 1) for points in mlvl_points]
+        bbox_targets = []
+        for i in range(num_levels):
+            bbox_target = bbox2distance(mlvl_points[i], decoded_bboxes[i])
+            bbox_targets.append(bbox_target)
+
+        return bbox_targets
+
+    def _load_from_state_dict(self, state_dict, prefix, local_metadata, strict,
+                              missing_keys, unexpected_keys, error_msgs):
+        """Override the method in the parent class to avoid changing para's
+        name."""
+        pass
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/yolact_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/yolact_head.py
new file mode 100644
index 0000000..10d311f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/yolact_head.py
@@ -0,0 +1,943 @@
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, xavier_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import build_sampler, fast_nms, images_to_levels, multi_apply
+from ..builder import HEADS, build_loss
+from .anchor_head import AnchorHead
+
+
+@HEADS.register_module()
+class YOLACTHead(AnchorHead):
+    """YOLACT box head used in https://arxiv.org/abs/1904.02689.
+
+    Note that YOLACT head is a light version of RetinaNet head.
+    Four differences are described as follows:
+
+    1. YOLACT box head has three-times fewer anchors.
+    2. YOLACT box head shares the convs for box and cls branches.
+    3. YOLACT box head uses OHEM instead of Focal loss.
+    4. YOLACT box head predicts a set of mask coefficients for each box.
+
+    Args:
+        num_classes (int): Number of categories excluding the background
+            category.
+        in_channels (int): Number of channels in the input feature map.
+        anchor_generator (dict): Config dict for anchor generator
+        loss_cls (dict): Config of classification loss.
+        loss_bbox (dict): Config of localization loss.
+        num_head_convs (int): Number of the conv layers shared by
+            box and cls branches.
+        num_protos (int): Number of the mask coefficients.
+        use_ohem (bool): If true, ``loss_single_OHEM`` will be used for
+            cls loss calculation. If false, ``loss_single`` will be used.
+        conv_cfg (dict): Dictionary to construct and config conv layer.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 anchor_generator=dict(
+                     type='AnchorGenerator',
+                     octave_base_scale=3,
+                     scales_per_octave=1,
+                     ratios=[0.5, 1.0, 2.0],
+                     strides=[8, 16, 32, 64, 128]),
+                 loss_cls=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=False,
+                     reduction='none',
+                     loss_weight=1.0),
+                 loss_bbox=dict(
+                     type='SmoothL1Loss', beta=1.0, loss_weight=1.5),
+                 num_head_convs=1,
+                 num_protos=32,
+                 use_ohem=True,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 **kwargs):
+        self.num_head_convs = num_head_convs
+        self.num_protos = num_protos
+        self.use_ohem = use_ohem
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        super(YOLACTHead, self).__init__(
+            num_classes,
+            in_channels,
+            loss_cls=loss_cls,
+            loss_bbox=loss_bbox,
+            anchor_generator=anchor_generator,
+            **kwargs)
+        if self.use_ohem:
+            sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+            self.sampling = False
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.relu = nn.ReLU(inplace=True)
+        self.head_convs = nn.ModuleList()
+        for i in range(self.num_head_convs):
+            chn = self.in_channels if i == 0 else self.feat_channels
+            self.head_convs.append(
+                ConvModule(
+                    chn,
+                    self.feat_channels,
+                    3,
+                    stride=1,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+        self.conv_cls = nn.Conv2d(
+            self.feat_channels,
+            self.num_anchors * self.cls_out_channels,
+            3,
+            padding=1)
+        self.conv_reg = nn.Conv2d(
+            self.feat_channels, self.num_anchors * 4, 3, padding=1)
+        self.conv_coeff = nn.Conv2d(
+            self.feat_channels,
+            self.num_anchors * self.num_protos,
+            3,
+            padding=1)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.head_convs:
+            xavier_init(m.conv, distribution='uniform', bias=0)
+        xavier_init(self.conv_cls, distribution='uniform', bias=0)
+        xavier_init(self.conv_reg, distribution='uniform', bias=0)
+        xavier_init(self.conv_coeff, distribution='uniform', bias=0)
+
+    def forward_single(self, x):
+        """Forward feature of a single scale level.
+
+        Args:
+            x (Tensor): Features of a single scale level.
+
+        Returns:
+            tuple:
+                cls_score (Tensor): Cls scores for a single scale level \
+                    the channels number is num_anchors * num_classes.
+                bbox_pred (Tensor): Box energies / deltas for a single scale \
+                    level, the channels number is num_anchors * 4.
+                coeff_pred (Tensor): Mask coefficients for a single scale \
+                    level, the channels number is num_anchors * num_protos.
+        """
+        for head_conv in self.head_convs:
+            x = head_conv(x)
+        cls_score = self.conv_cls(x)
+        bbox_pred = self.conv_reg(x)
+        coeff_pred = self.conv_coeff(x).tanh()
+        return cls_score, bbox_pred, coeff_pred
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds'))
+    def loss(self,
+             cls_scores,
+             bbox_preds,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """A combination of the func:``AnchorHead.loss`` and
+        func:``SSDHead.loss``.
+
+        When ``self.use_ohem == True``, it functions like ``SSDHead.loss``,
+        otherwise, it follows ``AnchorHead.loss``. Besides, it additionally
+        returns ``sampling_results``.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                Has shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): Class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): Specify which bounding
+                boxes can be ignored when computing the loss. Default: None
+
+        Returns:
+            tuple:
+                dict[str, Tensor]: A dictionary of loss components.
+                List[:obj:``SamplingResult``]: Sampler results for each image.
+        """
+        featmap_sizes = [featmap.size()[-2:] for featmap in cls_scores]
+        assert len(featmap_sizes) == self.anchor_generator.num_levels
+
+        device = cls_scores[0].device
+
+        anchor_list, valid_flag_list = self.get_anchors(
+            featmap_sizes, img_metas, device=device)
+        label_channels = self.cls_out_channels if self.use_sigmoid_cls else 1
+        cls_reg_targets = self.get_targets(
+            anchor_list,
+            valid_flag_list,
+            gt_bboxes,
+            img_metas,
+            gt_bboxes_ignore_list=gt_bboxes_ignore,
+            gt_labels_list=gt_labels,
+            label_channels=label_channels,
+            unmap_outputs=not self.use_ohem,
+            return_sampling_results=True)
+        if cls_reg_targets is None:
+            return None
+        (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list,
+         num_total_pos, num_total_neg, sampling_results) = cls_reg_targets
+
+        if self.use_ohem:
+            num_images = len(img_metas)
+            all_cls_scores = torch.cat([
+                s.permute(0, 2, 3, 1).reshape(
+                    num_images, -1, self.cls_out_channels) for s in cls_scores
+            ], 1)
+            all_labels = torch.cat(labels_list, -1).view(num_images, -1)
+            all_label_weights = torch.cat(label_weights_list,
+                                          -1).view(num_images, -1)
+            all_bbox_preds = torch.cat([
+                b.permute(0, 2, 3, 1).reshape(num_images, -1, 4)
+                for b in bbox_preds
+            ], -2)
+            all_bbox_targets = torch.cat(bbox_targets_list,
+                                         -2).view(num_images, -1, 4)
+            all_bbox_weights = torch.cat(bbox_weights_list,
+                                         -2).view(num_images, -1, 4)
+
+            # concat all level anchors to a single tensor
+            all_anchors = []
+            for i in range(num_images):
+                all_anchors.append(torch.cat(anchor_list[i]))
+
+            # check NaN and Inf
+            assert torch.isfinite(all_cls_scores).all().item(), \
+                'classification scores become infinite or NaN!'
+            assert torch.isfinite(all_bbox_preds).all().item(), \
+                'bbox predications become infinite or NaN!'
+
+            losses_cls, losses_bbox = multi_apply(
+                self.loss_single_OHEM,
+                all_cls_scores,
+                all_bbox_preds,
+                all_anchors,
+                all_labels,
+                all_label_weights,
+                all_bbox_targets,
+                all_bbox_weights,
+                num_total_samples=num_total_pos)
+        else:
+            num_total_samples = (
+                num_total_pos +
+                num_total_neg if self.sampling else num_total_pos)
+
+            # anchor number of multi levels
+            num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+            # concat all level anchors and flags to a single tensor
+            concat_anchor_list = []
+            for i in range(len(anchor_list)):
+                concat_anchor_list.append(torch.cat(anchor_list[i]))
+            all_anchor_list = images_to_levels(concat_anchor_list,
+                                               num_level_anchors)
+            losses_cls, losses_bbox = multi_apply(
+                self.loss_single,
+                cls_scores,
+                bbox_preds,
+                all_anchor_list,
+                labels_list,
+                label_weights_list,
+                bbox_targets_list,
+                bbox_weights_list,
+                num_total_samples=num_total_samples)
+
+        return dict(
+            loss_cls=losses_cls, loss_bbox=losses_bbox), sampling_results
+
+    def loss_single_OHEM(self, cls_score, bbox_pred, anchors, labels,
+                         label_weights, bbox_targets, bbox_weights,
+                         num_total_samples):
+        """"See func:``SSDHead.loss``."""
+        loss_cls_all = self.loss_cls(cls_score, labels, label_weights)
+
+        # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+        pos_inds = ((labels >= 0) & (labels < self.num_classes)).nonzero(
+            as_tuple=False).reshape(-1)
+        neg_inds = (labels == self.num_classes).nonzero(
+            as_tuple=False).view(-1)
+
+        num_pos_samples = pos_inds.size(0)
+        if num_pos_samples == 0:
+            num_neg_samples = neg_inds.size(0)
+        else:
+            num_neg_samples = self.train_cfg.neg_pos_ratio * num_pos_samples
+            if num_neg_samples > neg_inds.size(0):
+                num_neg_samples = neg_inds.size(0)
+        topk_loss_cls_neg, _ = loss_cls_all[neg_inds].topk(num_neg_samples)
+        loss_cls_pos = loss_cls_all[pos_inds].sum()
+        loss_cls_neg = topk_loss_cls_neg.sum()
+        loss_cls = (loss_cls_pos + loss_cls_neg) / num_total_samples
+        if self.reg_decoded_bbox:
+            # When the regression loss (e.g. `IouLoss`, `GIouLoss`)
+            # is applied directly on the decoded bounding boxes, it
+            # decodes the already encoded coordinates to absolute format.
+            bbox_pred = self.bbox_coder.decode(anchors, bbox_pred)
+        loss_bbox = self.loss_bbox(
+            bbox_pred,
+            bbox_targets,
+            bbox_weights,
+            avg_factor=num_total_samples)
+        return loss_cls[None], loss_bbox
+
+    @force_fp32(apply_to=('cls_scores', 'bbox_preds', 'coeff_preds'))
+    def get_bboxes(self,
+                   cls_scores,
+                   bbox_preds,
+                   coeff_preds,
+                   img_metas,
+                   cfg=None,
+                   rescale=False):
+        """"Similiar to func:``AnchorHead.get_bboxes``, but additionally
+        processes coeff_preds.
+
+        Args:
+            cls_scores (list[Tensor]): Box scores for each scale level
+                with shape (N, num_anchors * num_classes, H, W)
+            bbox_preds (list[Tensor]): Box energies / deltas for each scale
+                level with shape (N, num_anchors * 4, H, W)
+            coeff_preds (list[Tensor]): Mask coefficients for each scale
+                level with shape (N, num_anchors * num_protos, H, W)
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+
+        Returns:
+            list[tuple[Tensor, Tensor, Tensor]]: Each item in result_list is
+                a 3-tuple. The first item is an (n, 5) tensor, where the
+                first 4 columns are bounding box positions
+                (tl_x, tl_y, br_x, br_y) and the 5-th column is a score
+                between 0 and 1. The second item is an (n,) tensor where each
+                item is the predicted class label of the corresponding box.
+                The third item is an (n, num_protos) tensor where each item
+                is the predicted mask coefficients of instance inside the
+                corresponding box.
+        """
+        assert len(cls_scores) == len(bbox_preds)
+        num_levels = len(cls_scores)
+
+        device = cls_scores[0].device
+        featmap_sizes = [cls_scores[i].shape[-2:] for i in range(num_levels)]
+        mlvl_anchors = self.anchor_generator.grid_anchors(
+            featmap_sizes, device=device)
+
+        det_bboxes = []
+        det_labels = []
+        det_coeffs = []
+        for img_id in range(len(img_metas)):
+            cls_score_list = [
+                cls_scores[i][img_id].detach() for i in range(num_levels)
+            ]
+            bbox_pred_list = [
+                bbox_preds[i][img_id].detach() for i in range(num_levels)
+            ]
+            coeff_pred_list = [
+                coeff_preds[i][img_id].detach() for i in range(num_levels)
+            ]
+            img_shape = img_metas[img_id]['img_shape']
+            scale_factor = img_metas[img_id]['scale_factor']
+            bbox_res = self._get_bboxes_single(cls_score_list, bbox_pred_list,
+                                               coeff_pred_list, mlvl_anchors,
+                                               img_shape, scale_factor, cfg,
+                                               rescale)
+            det_bboxes.append(bbox_res[0])
+            det_labels.append(bbox_res[1])
+            det_coeffs.append(bbox_res[2])
+        return det_bboxes, det_labels, det_coeffs
+
+    def _get_bboxes_single(self,
+                           cls_score_list,
+                           bbox_pred_list,
+                           coeff_preds_list,
+                           mlvl_anchors,
+                           img_shape,
+                           scale_factor,
+                           cfg,
+                           rescale=False):
+        """"Similiar to func:``AnchorHead._get_bboxes_single``, but
+        additionally processes coeff_preds_list and uses fast NMS instead of
+        traditional NMS.
+
+        Args:
+            cls_score_list (list[Tensor]): Box scores for a single scale level
+                Has shape (num_anchors * num_classes, H, W).
+            bbox_pred_list (list[Tensor]): Box energies / deltas for a single
+                scale level with shape (num_anchors * 4, H, W).
+            coeff_preds_list (list[Tensor]): Mask coefficients for a single
+                scale level with shape (num_anchors * num_protos, H, W).
+            mlvl_anchors (list[Tensor]): Box reference for a single scale level
+                with shape (num_total_anchors, 4).
+            img_shape (tuple[int]): Shape of the input image,
+                (height, width, 3).
+            scale_factor (ndarray): Scale factor of the image arange as
+                (w_scale, h_scale, w_scale, h_scale).
+            cfg (mmcv.Config): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+
+        Returns:
+            tuple[Tensor, Tensor, Tensor]: The first item is an (n, 5) tensor,
+                where the first 4 columns are bounding box positions
+                (tl_x, tl_y, br_x, br_y) and the 5-th column is a score between
+                0 and 1. The second item is an (n,) tensor where each item is
+                the predicted class label of the corresponding box. The third
+                item is an (n, num_protos) tensor where each item is the
+                predicted mask coefficients of instance inside the
+                corresponding box.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(cls_score_list) == len(bbox_pred_list) == len(mlvl_anchors)
+        mlvl_bboxes = []
+        mlvl_scores = []
+        mlvl_coeffs = []
+        for cls_score, bbox_pred, coeff_pred, anchors in \
+                zip(cls_score_list, bbox_pred_list,
+                    coeff_preds_list, mlvl_anchors):
+            assert cls_score.size()[-2:] == bbox_pred.size()[-2:]
+            cls_score = cls_score.permute(1, 2,
+                                          0).reshape(-1, self.cls_out_channels)
+            if self.use_sigmoid_cls:
+                scores = cls_score.sigmoid()
+            else:
+                scores = cls_score.softmax(-1)
+            bbox_pred = bbox_pred.permute(1, 2, 0).reshape(-1, 4)
+            coeff_pred = coeff_pred.permute(1, 2,
+                                            0).reshape(-1, self.num_protos)
+            nms_pre = cfg.get('nms_pre', -1)
+            if nms_pre > 0 and scores.shape[0] > nms_pre:
+                # Get maximum scores for foreground classes.
+                if self.use_sigmoid_cls:
+                    max_scores, _ = scores.max(dim=1)
+                else:
+                    # remind that we set FG labels to [0, num_class-1]
+                    # since mmdet v2.0
+                    # BG cat_id: num_class
+                    max_scores, _ = scores[:, :-1].max(dim=1)
+                _, topk_inds = max_scores.topk(nms_pre)
+                anchors = anchors[topk_inds, :]
+                bbox_pred = bbox_pred[topk_inds, :]
+                scores = scores[topk_inds, :]
+                coeff_pred = coeff_pred[topk_inds, :]
+            bboxes = self.bbox_coder.decode(
+                anchors, bbox_pred, max_shape=img_shape)
+            mlvl_bboxes.append(bboxes)
+            mlvl_scores.append(scores)
+            mlvl_coeffs.append(coeff_pred)
+        mlvl_bboxes = torch.cat(mlvl_bboxes)
+        if rescale:
+            mlvl_bboxes /= mlvl_bboxes.new_tensor(scale_factor)
+        mlvl_scores = torch.cat(mlvl_scores)
+        mlvl_coeffs = torch.cat(mlvl_coeffs)
+        if self.use_sigmoid_cls:
+            # Add a dummy background class to the backend when using sigmoid
+            # remind that we set FG labels to [0, num_class-1] since mmdet v2.0
+            # BG cat_id: num_class
+            padding = mlvl_scores.new_zeros(mlvl_scores.shape[0], 1)
+            mlvl_scores = torch.cat([mlvl_scores, padding], dim=1)
+        det_bboxes, det_labels, det_coeffs = fast_nms(mlvl_bboxes, mlvl_scores,
+                                                      mlvl_coeffs,
+                                                      cfg.score_thr,
+                                                      cfg.iou_thr, cfg.top_k,
+                                                      cfg.max_per_img)
+        return det_bboxes, det_labels, det_coeffs
+
+
+@HEADS.register_module()
+class YOLACTSegmHead(nn.Module):
+    """YOLACT segmentation head used in https://arxiv.org/abs/1904.02689.
+
+    Apply a semantic segmentation loss on feature space using layers that are
+    only evaluated during training to increase performance with no speed
+    penalty.
+
+    Args:
+        in_channels (int): Number of channels in the input feature map.
+        num_classes (int): Number of categories excluding the background
+            category.
+        loss_segm (dict): Config of semantic segmentation loss.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels=256,
+                 loss_segm=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.0)):
+        super(YOLACTSegmHead, self).__init__()
+        self.in_channels = in_channels
+        self.num_classes = num_classes
+        self.loss_segm = build_loss(loss_segm)
+        self._init_layers()
+        self.fp16_enabled = False
+
+    def _init_layers(self):
+        """Initialize layers of the head."""
+        self.segm_conv = nn.Conv2d(
+            self.in_channels, self.num_classes, kernel_size=1)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        xavier_init(self.segm_conv, distribution='uniform')
+
+    def forward(self, x):
+        """Forward feature from the upstream network.
+
+        Args:
+            x (Tensor): Feature from the upstream network, which is
+                a 4D-tensor.
+
+        Returns:
+            Tensor: Predicted semantic segmentation map with shape
+                (N, num_classes, H, W).
+        """
+        return self.segm_conv(x)
+
+    @force_fp32(apply_to=('segm_pred', ))
+    def loss(self, segm_pred, gt_masks, gt_labels):
+        """Compute loss of the head.
+
+        Args:
+            segm_pred (list[Tensor]): Predicted semantic segmentation map
+                with shape (N, num_classes, H, W).
+            gt_masks (list[Tensor]): Ground truth masks for each image with
+                the same shape of the input image.
+            gt_labels (list[Tensor]): Class indices corresponding to each box.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        loss_segm = []
+        num_imgs, num_classes, mask_h, mask_w = segm_pred.size()
+        for idx in range(num_imgs):
+            cur_segm_pred = segm_pred[idx]
+            cur_gt_masks = gt_masks[idx].float()
+            cur_gt_labels = gt_labels[idx]
+            segm_targets = self.get_targets(cur_segm_pred, cur_gt_masks,
+                                            cur_gt_labels)
+            if segm_targets is None:
+                loss = self.loss_segm(cur_segm_pred,
+                                      torch.zeros_like(cur_segm_pred),
+                                      torch.zeros_like(cur_segm_pred))
+            else:
+                loss = self.loss_segm(
+                    cur_segm_pred,
+                    segm_targets,
+                    avg_factor=num_imgs * mask_h * mask_w)
+            loss_segm.append(loss)
+        return dict(loss_segm=loss_segm)
+
+    def get_targets(self, segm_pred, gt_masks, gt_labels):
+        """Compute semantic segmentation targets for each image.
+
+        Args:
+            segm_pred (Tensor): Predicted semantic segmentation map
+                with shape (num_classes, H, W).
+            gt_masks (Tensor): Ground truth masks for each image with
+                the same shape of the input image.
+            gt_labels (Tensor): Class indices corresponding to each box.
+
+        Returns:
+            Tensor: Semantic segmentation targets with shape
+                (num_classes, H, W).
+        """
+        if gt_masks.size(0) == 0:
+            return None
+        num_classes, mask_h, mask_w = segm_pred.size()
+        with torch.no_grad():
+            downsampled_masks = F.interpolate(
+                gt_masks.unsqueeze(0), (mask_h, mask_w),
+                mode='bilinear',
+                align_corners=False).squeeze(0)
+            downsampled_masks = downsampled_masks.gt(0.5).float()
+            segm_targets = torch.zeros_like(segm_pred, requires_grad=False)
+            for obj_idx in range(downsampled_masks.size(0)):
+                segm_targets[gt_labels[obj_idx] - 1] = torch.max(
+                    segm_targets[gt_labels[obj_idx] - 1],
+                    downsampled_masks[obj_idx])
+            return segm_targets
+
+
+@HEADS.register_module()
+class YOLACTProtonet(nn.Module):
+    """YOLACT mask head used in https://arxiv.org/abs/1904.02689.
+
+    This head outputs the mask prototypes for YOLACT.
+
+    Args:
+        in_channels (int): Number of channels in the input feature map.
+        proto_channels (tuple[int]): Output channels of protonet convs.
+        proto_kernel_sizes (tuple[int]): Kernel sizes of protonet convs.
+        include_last_relu (Bool): If keep the last relu of protonet.
+        num_protos (int): Number of prototypes.
+        num_classes (int): Number of categories excluding the background
+            category.
+        loss_mask_weight (float): Reweight the mask loss by this factor.
+        max_masks_to_train (int): Maximum number of masks to train for
+            each image.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels=256,
+                 proto_channels=(256, 256, 256, None, 256, 32),
+                 proto_kernel_sizes=(3, 3, 3, -2, 3, 1),
+                 include_last_relu=True,
+                 num_protos=32,
+                 loss_mask_weight=1.0,
+                 max_masks_to_train=100):
+        super(YOLACTProtonet, self).__init__()
+        self.in_channels = in_channels
+        self.proto_channels = proto_channels
+        self.proto_kernel_sizes = proto_kernel_sizes
+        self.include_last_relu = include_last_relu
+        self.protonet = self._init_layers()
+
+        self.loss_mask_weight = loss_mask_weight
+        self.num_protos = num_protos
+        self.num_classes = num_classes
+        self.max_masks_to_train = max_masks_to_train
+        self.fp16_enabled = False
+
+    def _init_layers(self):
+        """A helper function to take a config setting and turn it into a
+        network."""
+        # Possible patterns:
+        # ( 256, 3) -> conv
+        # ( 256,-2) -> deconv
+        # (None,-2) -> bilinear interpolate
+        in_channels = self.in_channels
+        protonets = nn.ModuleList()
+        for num_channels, kernel_size in zip(self.proto_channels,
+                                             self.proto_kernel_sizes):
+            if kernel_size > 0:
+                layer = nn.Conv2d(
+                    in_channels,
+                    num_channels,
+                    kernel_size,
+                    padding=kernel_size // 2)
+            else:
+                if num_channels is None:
+                    layer = InterpolateModule(
+                        scale_factor=-kernel_size,
+                        mode='bilinear',
+                        align_corners=False)
+                else:
+                    layer = nn.ConvTranspose2d(
+                        in_channels,
+                        num_channels,
+                        -kernel_size,
+                        padding=kernel_size // 2)
+            protonets.append(layer)
+            protonets.append(nn.ReLU(inplace=True))
+            in_channels = num_channels if num_channels is not None \
+                else in_channels
+        if not self.include_last_relu:
+            protonets = protonets[:-1]
+        return nn.Sequential(*protonets)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.protonet:
+            if isinstance(m, nn.Conv2d):
+                xavier_init(m, distribution='uniform')
+
+    def forward(self, x, coeff_pred, bboxes, img_meta, sampling_results=None):
+        """Forward feature from the upstream network to get prototypes and
+        linearly combine the prototypes, using masks coefficients, into
+        instance masks. Finally, crop the instance masks with given bboxes.
+
+        Args:
+            x (Tensor): Feature from the upstream network, which is
+                a 4D-tensor.
+            coeff_pred (list[Tensor]): Mask coefficients for each scale
+                level with shape (N, num_anchors * num_protos, H, W).
+            bboxes (list[Tensor]): Box used for cropping with shape
+                (N, num_anchors * 4, H, W). During training, they are
+                ground truth boxes. During testing, they are predicted
+                boxes.
+            img_meta (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            sampling_results (List[:obj:``SamplingResult``]): Sampler results
+                for each image.
+
+        Returns:
+            list[Tensor]: Predicted instance segmentation masks.
+        """
+        prototypes = self.protonet(x)
+        prototypes = prototypes.permute(0, 2, 3, 1).contiguous()
+
+        num_imgs = x.size(0)
+        # Training state
+        if self.training:
+            coeff_pred_list = []
+            for coeff_pred_per_level in coeff_pred:
+                coeff_pred_per_level = \
+                    coeff_pred_per_level.permute(0, 2, 3, 1)\
+                    .reshape(num_imgs, -1, self.num_protos)
+                coeff_pred_list.append(coeff_pred_per_level)
+            coeff_pred = torch.cat(coeff_pred_list, dim=1)
+
+        mask_pred_list = []
+        for idx in range(num_imgs):
+            cur_prototypes = prototypes[idx]
+            cur_coeff_pred = coeff_pred[idx]
+            cur_bboxes = bboxes[idx]
+            cur_img_meta = img_meta[idx]
+
+            # Testing state
+            if not self.training:
+                bboxes_for_cropping = cur_bboxes
+            else:
+                cur_sampling_results = sampling_results[idx]
+                pos_assigned_gt_inds = \
+                    cur_sampling_results.pos_assigned_gt_inds
+                bboxes_for_cropping = cur_bboxes[pos_assigned_gt_inds].clone()
+                pos_inds = cur_sampling_results.pos_inds
+                cur_coeff_pred = cur_coeff_pred[pos_inds]
+
+            # Linearly combine the prototypes with the mask coefficients
+            mask_pred = cur_prototypes @ cur_coeff_pred.t()
+            mask_pred = torch.sigmoid(mask_pred)
+
+            h, w = cur_img_meta['img_shape'][:2]
+            bboxes_for_cropping[:, 0] /= w
+            bboxes_for_cropping[:, 1] /= h
+            bboxes_for_cropping[:, 2] /= w
+            bboxes_for_cropping[:, 3] /= h
+
+            mask_pred = self.crop(mask_pred, bboxes_for_cropping)
+            mask_pred = mask_pred.permute(2, 0, 1).contiguous()
+            mask_pred_list.append(mask_pred)
+        return mask_pred_list
+
+    @force_fp32(apply_to=('mask_pred', ))
+    def loss(self, mask_pred, gt_masks, gt_bboxes, img_meta, sampling_results):
+        """Compute loss of the head.
+
+        Args:
+            mask_pred (list[Tensor]): Predicted prototypes with shape
+                (num_classes, H, W).
+            gt_masks (list[Tensor]): Ground truth masks for each image with
+                the same shape of the input image.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            img_meta (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            sampling_results (List[:obj:``SamplingResult``]): Sampler results
+                for each image.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        loss_mask = []
+        num_imgs = len(mask_pred)
+        total_pos = 0
+        for idx in range(num_imgs):
+            cur_mask_pred = mask_pred[idx]
+            cur_gt_masks = gt_masks[idx].float()
+            cur_gt_bboxes = gt_bboxes[idx]
+            cur_img_meta = img_meta[idx]
+            cur_sampling_results = sampling_results[idx]
+
+            pos_assigned_gt_inds = cur_sampling_results.pos_assigned_gt_inds
+            num_pos = pos_assigned_gt_inds.size(0)
+            # Since we're producing (near) full image masks,
+            # it'd take too much vram to backprop on every single mask.
+            # Thus we select only a subset.
+            if num_pos > self.max_masks_to_train:
+                perm = torch.randperm(num_pos)
+                select = perm[:self.max_masks_to_train]
+                cur_mask_pred = cur_mask_pred[select]
+                pos_assigned_gt_inds = pos_assigned_gt_inds[select]
+                num_pos = self.max_masks_to_train
+            total_pos += num_pos
+
+            gt_bboxes_for_reweight = cur_gt_bboxes[pos_assigned_gt_inds]
+
+            mask_targets = self.get_targets(cur_mask_pred, cur_gt_masks,
+                                            pos_assigned_gt_inds)
+            if num_pos == 0:
+                loss = cur_mask_pred.sum() * 0.
+            elif mask_targets is None:
+                loss = F.binary_cross_entropy(cur_mask_pred,
+                                              torch.zeros_like(cur_mask_pred),
+                                              torch.zeros_like(cur_mask_pred))
+            else:
+                cur_mask_pred = torch.clamp(cur_mask_pred, 0, 1)
+                loss = F.binary_cross_entropy(
+                    cur_mask_pred, mask_targets,
+                    reduction='none') * self.loss_mask_weight
+
+                h, w = cur_img_meta['img_shape'][:2]
+                gt_bboxes_width = (gt_bboxes_for_reweight[:, 2] -
+                                   gt_bboxes_for_reweight[:, 0]) / w
+                gt_bboxes_height = (gt_bboxes_for_reweight[:, 3] -
+                                    gt_bboxes_for_reweight[:, 1]) / h
+                loss = loss.mean(dim=(1,
+                                      2)) / gt_bboxes_width / gt_bboxes_height
+                loss = torch.sum(loss)
+            loss_mask.append(loss)
+
+        if total_pos == 0:
+            total_pos += 1  # avoid nan
+        loss_mask = [x / total_pos for x in loss_mask]
+
+        return dict(loss_mask=loss_mask)
+
+    def get_targets(self, mask_pred, gt_masks, pos_assigned_gt_inds):
+        """Compute instance segmentation targets for each image.
+
+        Args:
+            mask_pred (Tensor): Predicted prototypes with shape
+                (num_classes, H, W).
+            gt_masks (Tensor): Ground truth masks for each image with
+                the same shape of the input image.
+            pos_assigned_gt_inds (Tensor): GT indices of the corresponding
+                positive samples.
+        Returns:
+            Tensor: Instance segmentation targets with shape
+                (num_instances, H, W).
+        """
+        if gt_masks.size(0) == 0:
+            return None
+        mask_h, mask_w = mask_pred.shape[-2:]
+        gt_masks = F.interpolate(
+            gt_masks.unsqueeze(0), (mask_h, mask_w),
+            mode='bilinear',
+            align_corners=False).squeeze(0)
+        gt_masks = gt_masks.gt(0.5).float()
+        mask_targets = gt_masks[pos_assigned_gt_inds]
+        return mask_targets
+
+    def get_seg_masks(self, mask_pred, label_pred, img_meta, rescale):
+        """Resize, binarize, and format the instance mask predictions.
+
+        Args:
+            mask_pred (Tensor): shape (N, H, W).
+            label_pred (Tensor): shape (N, ).
+            img_meta (dict): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            rescale (bool): If rescale is False, then returned masks will
+                fit the scale of imgs[0].
+        Returns:
+            list[ndarray]: Mask predictions grouped by their predicted classes.
+        """
+        ori_shape = img_meta['ori_shape']
+        scale_factor = img_meta['scale_factor']
+        if rescale:
+            img_h, img_w = ori_shape[:2]
+        else:
+            img_h = np.round(ori_shape[0] * scale_factor[1]).astype(np.int32)
+            img_w = np.round(ori_shape[1] * scale_factor[0]).astype(np.int32)
+
+        cls_segms = [[] for _ in range(self.num_classes)]
+        if mask_pred.size(0) == 0:
+            return cls_segms
+
+        mask_pred = F.interpolate(
+            mask_pred.unsqueeze(0), (img_h, img_w),
+            mode='bilinear',
+            align_corners=False).squeeze(0) > 0.5
+        mask_pred = mask_pred.cpu().numpy().astype(np.uint8)
+
+        for m, l in zip(mask_pred, label_pred):
+            cls_segms[l].append(m)
+        return cls_segms
+
+    def crop(self, masks, boxes, padding=1):
+        """Crop predicted masks by zeroing out everything not in the predicted
+        bbox.
+
+        Args:
+            masks (Tensor): shape [H, W, N].
+            boxes (Tensor): bbox coords in relative point form with
+                shape [N, 4].
+
+        Return:
+            Tensor: The cropped masks.
+        """
+        h, w, n = masks.size()
+        x1, x2 = self.sanitize_coordinates(
+            boxes[:, 0], boxes[:, 2], w, padding, cast=False)
+        y1, y2 = self.sanitize_coordinates(
+            boxes[:, 1], boxes[:, 3], h, padding, cast=False)
+
+        rows = torch.arange(
+            w, device=masks.device, dtype=x1.dtype).view(1, -1,
+                                                         1).expand(h, w, n)
+        cols = torch.arange(
+            h, device=masks.device, dtype=x1.dtype).view(-1, 1,
+                                                         1).expand(h, w, n)
+
+        masks_left = rows >= x1.view(1, 1, -1)
+        masks_right = rows < x2.view(1, 1, -1)
+        masks_up = cols >= y1.view(1, 1, -1)
+        masks_down = cols < y2.view(1, 1, -1)
+
+        crop_mask = masks_left * masks_right * masks_up * masks_down
+
+        return masks * crop_mask.float()
+
+    def sanitize_coordinates(self, x1, x2, img_size, padding=0, cast=True):
+        """Sanitizes the input coordinates so that x1 < x2, x1 != x2, x1 >= 0,
+        and x2 <= image_size. Also converts from relative to absolute
+        coordinates and casts the results to long tensors.
+
+        Warning: this does things in-place behind the scenes so
+        copy if necessary.
+
+        Args:
+            _x1 (Tensor): shape (N, ).
+            _x2 (Tensor): shape (N, ).
+            img_size (int): Size of the input image.
+            padding (int): x1 >= padding, x2 <= image_size-padding.
+            cast (bool): If cast is false, the result won't be cast to longs.
+
+        Returns:
+            tuple:
+                x1 (Tensor): Sanitized _x1.
+                x2 (Tensor): Sanitized _x2.
+        """
+        x1 = x1 * img_size
+        x2 = x2 * img_size
+        if cast:
+            x1 = x1.long()
+            x2 = x2.long()
+        x1 = torch.min(x1, x2)
+        x2 = torch.max(x1, x2)
+        x1 = torch.clamp(x1 - padding, min=0)
+        x2 = torch.clamp(x2 + padding, max=img_size)
+        return x1, x2
+
+
+class InterpolateModule(nn.Module):
+    """This is a module version of F.interpolate.
+
+    Any arguments you give it just get passed along for the ride.
+    """
+
+    def __init__(self, *args, **kwargs):
+        super().__init__()
+
+        self.args = args
+        self.kwargs = kwargs
+
+    def forward(self, x):
+        """Forward features from the upstream network."""
+        return F.interpolate(x, *self.args, **self.kwargs)
diff --git a/python/autoCut/models/swin/mmdet/models/dense_heads/yolo_head.py b/python/autoCut/models/swin/mmdet/models/dense_heads/yolo_head.py
new file mode 100644
index 0000000..25a005d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/dense_heads/yolo_head.py
@@ -0,0 +1,577 @@
+# Copyright (c) 2019 Western Digital Corporation or its affiliates.
+
+import warnings
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, normal_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import (build_anchor_generator, build_assigner,
+                        build_bbox_coder, build_sampler, images_to_levels,
+                        multi_apply, multiclass_nms)
+from ..builder import HEADS, build_loss
+from .base_dense_head import BaseDenseHead
+from .dense_test_mixins import BBoxTestMixin
+
+
+@HEADS.register_module()
+class YOLOV3Head(BaseDenseHead, BBoxTestMixin):
+    """YOLOV3Head Paper link: https://arxiv.org/abs/1804.02767.
+
+    Args:
+        num_classes (int): The number of object classes (w/o background)
+        in_channels (List[int]): Number of input channels per scale.
+        out_channels (List[int]): The number of output channels per scale
+            before the final 1x1 layer. Default: (1024, 512, 256).
+        anchor_generator (dict): Config dict for anchor generator
+        bbox_coder (dict): Config of bounding box coder.
+        featmap_strides (List[int]): The stride of each scale.
+            Should be in descending order. Default: (32, 16, 8).
+        one_hot_smoother (float): Set a non-zero value to enable label-smooth
+            Default: 0.
+        conv_cfg (dict): Config dict for convolution layer. Default: None.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+            Default: dict(type='BN', requires_grad=True)
+        act_cfg (dict): Config dict for activation layer.
+            Default: dict(type='LeakyReLU', negative_slope=0.1).
+        loss_cls (dict): Config of classification loss.
+        loss_conf (dict): Config of confidence loss.
+        loss_xy (dict): Config of xy coordinate loss.
+        loss_wh (dict): Config of wh coordinate loss.
+        train_cfg (dict): Training config of YOLOV3 head. Default: None.
+        test_cfg (dict): Testing config of YOLOV3 head. Default: None.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 in_channels,
+                 out_channels=(1024, 512, 256),
+                 anchor_generator=dict(
+                     type='YOLOAnchorGenerator',
+                     base_sizes=[[(116, 90), (156, 198), (373, 326)],
+                                 [(30, 61), (62, 45), (59, 119)],
+                                 [(10, 13), (16, 30), (33, 23)]],
+                     strides=[32, 16, 8]),
+                 bbox_coder=dict(type='YOLOBBoxCoder'),
+                 featmap_strides=[32, 16, 8],
+                 one_hot_smoother=0.,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN', requires_grad=True),
+                 act_cfg=dict(type='LeakyReLU', negative_slope=0.1),
+                 loss_cls=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.0),
+                 loss_conf=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.0),
+                 loss_xy=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.0),
+                 loss_wh=dict(type='MSELoss', loss_weight=1.0),
+                 train_cfg=None,
+                 test_cfg=None):
+        super(YOLOV3Head, self).__init__()
+        # Check params
+        assert (len(in_channels) == len(out_channels) == len(featmap_strides))
+
+        self.num_classes = num_classes
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.featmap_strides = featmap_strides
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+        if self.train_cfg:
+            self.assigner = build_assigner(self.train_cfg.assigner)
+            if hasattr(self.train_cfg, 'sampler'):
+                sampler_cfg = self.train_cfg.sampler
+            else:
+                sampler_cfg = dict(type='PseudoSampler')
+            self.sampler = build_sampler(sampler_cfg, context=self)
+
+        self.one_hot_smoother = one_hot_smoother
+
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.act_cfg = act_cfg
+
+        self.bbox_coder = build_bbox_coder(bbox_coder)
+        self.anchor_generator = build_anchor_generator(anchor_generator)
+
+        self.loss_cls = build_loss(loss_cls)
+        self.loss_conf = build_loss(loss_conf)
+        self.loss_xy = build_loss(loss_xy)
+        self.loss_wh = build_loss(loss_wh)
+        # usually the numbers of anchors for each level are the same
+        # except SSD detectors
+        self.num_anchors = self.anchor_generator.num_base_anchors[0]
+        assert len(
+            self.anchor_generator.num_base_anchors) == len(featmap_strides)
+        self._init_layers()
+
+    @property
+    def num_levels(self):
+        return len(self.featmap_strides)
+
+    @property
+    def num_attrib(self):
+        """int: number of attributes in pred_map, bboxes (4) +
+        objectness (1) + num_classes"""
+
+        return 5 + self.num_classes
+
+    def _init_layers(self):
+        self.convs_bridge = nn.ModuleList()
+        self.convs_pred = nn.ModuleList()
+        for i in range(self.num_levels):
+            conv_bridge = ConvModule(
+                self.in_channels[i],
+                self.out_channels[i],
+                3,
+                padding=1,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg,
+                act_cfg=self.act_cfg)
+            conv_pred = nn.Conv2d(self.out_channels[i],
+                                  self.num_anchors * self.num_attrib, 1)
+
+            self.convs_bridge.append(conv_bridge)
+            self.convs_pred.append(conv_pred)
+
+    def init_weights(self):
+        """Initialize weights of the head."""
+        for m in self.convs_pred:
+            normal_init(m, std=0.01)
+
+    def forward(self, feats):
+        """Forward features from the upstream network.
+
+        Args:
+            feats (tuple[Tensor]): Features from the upstream network, each is
+                a 4D-tensor.
+
+        Returns:
+            tuple[Tensor]: A tuple of multi-level predication map, each is a
+                4D-tensor of shape (batch_size, 5+num_classes, height, width).
+        """
+
+        assert len(feats) == self.num_levels
+        pred_maps = []
+        for i in range(self.num_levels):
+            x = feats[i]
+            x = self.convs_bridge[i](x)
+            pred_map = self.convs_pred[i](x)
+            pred_maps.append(pred_map)
+
+        return tuple(pred_maps),
+
+    @force_fp32(apply_to=('pred_maps', ))
+    def get_bboxes(self,
+                   pred_maps,
+                   img_metas,
+                   cfg=None,
+                   rescale=False,
+                   with_nms=True):
+        """Transform network output for a batch into bbox predictions.
+
+        Args:
+            pred_maps (list[Tensor]): Raw predictions for a batch of images.
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used. Default: None.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where 5 represent
+                (tl_x, tl_y, br_x, br_y, score) and the score between 0 and 1.
+                The shape of the second tensor in the tuple is (n,), and
+                each element represents the class label of the corresponding
+                box.
+        """
+        num_levels = len(pred_maps)
+        pred_maps_list = [pred_maps[i].detach() for i in range(num_levels)]
+        scale_factors = [
+            img_metas[i]['scale_factor']
+            for i in range(pred_maps_list[0].shape[0])
+        ]
+        result_list = self._get_bboxes(pred_maps_list, scale_factors, cfg,
+                                       rescale, with_nms)
+        return result_list
+
+    def _get_bboxes(self,
+                    pred_maps_list,
+                    scale_factors,
+                    cfg,
+                    rescale=False,
+                    with_nms=True):
+        """Transform outputs for a single batch item into bbox predictions.
+
+        Args:
+            pred_maps_list (list[Tensor]): Prediction maps for different scales
+                of each single image in the batch.
+            scale_factors (list(ndarray)): Scale factor of the image arrange as
+                (w_scale, h_scale, w_scale, h_scale).
+            cfg (mmcv.Config | None): Test / postprocessing configuration,
+                if None, test_cfg would be used.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            with_nms (bool): If True, do nms before return boxes.
+                Default: True.
+
+        Returns:
+            list[tuple[Tensor, Tensor]]: Each item in result_list is 2-tuple.
+                The first item is an (n, 5) tensor, where 5 represent
+                (tl_x, tl_y, br_x, br_y, score) and the score between 0 and 1.
+                The shape of the second tensor in the tuple is (n,), and
+                each element represents the class label of the corresponding
+                box.
+        """
+        cfg = self.test_cfg if cfg is None else cfg
+        assert len(pred_maps_list) == self.num_levels
+
+        device = pred_maps_list[0].device
+        batch_size = pred_maps_list[0].shape[0]
+
+        featmap_sizes = [
+            pred_maps_list[i].shape[-2:] for i in range(self.num_levels)
+        ]
+        multi_lvl_anchors = self.anchor_generator.grid_anchors(
+            featmap_sizes, device)
+        # convert to tensor to keep tracing
+        nms_pre_tensor = torch.tensor(
+            cfg.get('nms_pre', -1), device=device, dtype=torch.long)
+
+        multi_lvl_bboxes = []
+        multi_lvl_cls_scores = []
+        multi_lvl_conf_scores = []
+        for i in range(self.num_levels):
+            # get some key info for current scale
+            pred_map = pred_maps_list[i]
+            stride = self.featmap_strides[i]
+            # (b,h, w, num_anchors*num_attrib) ->
+            # (b,h*w*num_anchors, num_attrib)
+            pred_map = pred_map.permute(0, 2, 3,
+                                        1).reshape(batch_size, -1,
+                                                   self.num_attrib)
+            # Inplace operation like
+            # ```pred_map[..., :2] = \torch.sigmoid(pred_map[..., :2])```
+            # would create constant tensor when exporting to onnx
+            pred_map_conf = torch.sigmoid(pred_map[..., :2])
+            pred_map_rest = pred_map[..., 2:]
+            pred_map = torch.cat([pred_map_conf, pred_map_rest], dim=-1)
+            pred_map_boxes = pred_map[..., :4]
+            multi_lvl_anchor = multi_lvl_anchors[i]
+            multi_lvl_anchor = multi_lvl_anchor.expand_as(pred_map_boxes)
+            bbox_pred = self.bbox_coder.decode(multi_lvl_anchor,
+                                               pred_map_boxes, stride)
+            # conf and cls
+            conf_pred = torch.sigmoid(pred_map[..., 4])
+            cls_pred = torch.sigmoid(pred_map[..., 5:]).view(
+                batch_size, -1, self.num_classes)  # Cls pred one-hot.
+
+            # Get top-k prediction
+            # Always keep topk op for dynamic input in onnx
+            if nms_pre_tensor > 0 and (torch.onnx.is_in_onnx_export()
+                                       or conf_pred.shape[1] > nms_pre_tensor):
+                from torch import _shape_as_tensor
+                # keep shape as tensor and get k
+                num_anchor = _shape_as_tensor(conf_pred)[1].to(device)
+                nms_pre = torch.where(nms_pre_tensor < num_anchor,
+                                      nms_pre_tensor, num_anchor)
+                _, topk_inds = conf_pred.topk(nms_pre)
+                batch_inds = torch.arange(batch_size).view(
+                    -1, 1).expand_as(topk_inds).long()
+                bbox_pred = bbox_pred[batch_inds, topk_inds, :]
+                cls_pred = cls_pred[batch_inds, topk_inds, :]
+                conf_pred = conf_pred[batch_inds, topk_inds]
+
+            # Save the result of current scale
+            multi_lvl_bboxes.append(bbox_pred)
+            multi_lvl_cls_scores.append(cls_pred)
+            multi_lvl_conf_scores.append(conf_pred)
+
+        # Merge the results of different scales together
+        batch_mlvl_bboxes = torch.cat(multi_lvl_bboxes, dim=1)
+        batch_mlvl_scores = torch.cat(multi_lvl_cls_scores, dim=1)
+        batch_mlvl_conf_scores = torch.cat(multi_lvl_conf_scores, dim=1)
+
+        # Set max number of box to be feed into nms in deployment
+        deploy_nms_pre = cfg.get('deploy_nms_pre', -1)
+        if deploy_nms_pre > 0 and torch.onnx.is_in_onnx_export():
+            _, topk_inds = batch_mlvl_conf_scores.topk(deploy_nms_pre)
+            batch_inds = torch.arange(batch_size).view(
+                -1, 1).expand_as(topk_inds).long()
+            batch_mlvl_bboxes = batch_mlvl_bboxes[batch_inds, topk_inds, :]
+            batch_mlvl_scores = batch_mlvl_scores[batch_inds, topk_inds, :]
+            batch_mlvl_conf_scores = batch_mlvl_conf_scores[batch_inds,
+                                                            topk_inds]
+
+        if with_nms and (batch_mlvl_conf_scores.size(0) == 0):
+            return torch.zeros((0, 5)), torch.zeros((0, ))
+
+        if rescale:
+            batch_mlvl_bboxes /= batch_mlvl_bboxes.new_tensor(
+                scale_factors).unsqueeze(1)
+
+        # In mmdet 2.x, the class_id for background is num_classes.
+        # i.e., the last column.
+        padding = batch_mlvl_scores.new_zeros(batch_size,
+                                              batch_mlvl_scores.shape[1], 1)
+        batch_mlvl_scores = torch.cat([batch_mlvl_scores, padding], dim=-1)
+
+        # Support exporting to onnx without nms
+        if with_nms and cfg.get('nms', None) is not None:
+            det_results = []
+            for (mlvl_bboxes, mlvl_scores,
+                 mlvl_conf_scores) in zip(batch_mlvl_bboxes, batch_mlvl_scores,
+                                          batch_mlvl_conf_scores):
+                # Filtering out all predictions with conf < conf_thr
+                conf_thr = cfg.get('conf_thr', -1)
+                if conf_thr > 0 and (not torch.onnx.is_in_onnx_export()):
+                    # TensorRT not support NonZero
+                    # add as_tuple=False for compatibility in Pytorch 1.6
+                    # flatten would create a Reshape op with constant values,
+                    # and raise RuntimeError when doing inference in ONNX
+                    # Runtime with a different input image (#4221).
+                    conf_inds = mlvl_conf_scores.ge(conf_thr).nonzero(
+                        as_tuple=False).squeeze(1)
+                    mlvl_bboxes = mlvl_bboxes[conf_inds, :]
+                    mlvl_scores = mlvl_scores[conf_inds, :]
+                    mlvl_conf_scores = mlvl_conf_scores[conf_inds]
+
+                det_bboxes, det_labels = multiclass_nms(
+                    mlvl_bboxes,
+                    mlvl_scores,
+                    cfg.score_thr,
+                    cfg.nms,
+                    cfg.max_per_img,
+                    score_factors=mlvl_conf_scores)
+                det_results.append(tuple([det_bboxes, det_labels]))
+
+        else:
+            det_results = [
+                tuple(mlvl_bs)
+                for mlvl_bs in zip(batch_mlvl_bboxes, batch_mlvl_scores,
+                                   batch_mlvl_conf_scores)
+            ]
+        return det_results
+
+    @force_fp32(apply_to=('pred_maps', ))
+    def loss(self,
+             pred_maps,
+             gt_bboxes,
+             gt_labels,
+             img_metas,
+             gt_bboxes_ignore=None):
+        """Compute loss of the head.
+
+        Args:
+            pred_maps (list[Tensor]): Prediction map for each scale level,
+                shape (N, num_anchors * num_attrib, H, W)
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            img_metas (list[dict]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        num_imgs = len(img_metas)
+        device = pred_maps[0][0].device
+
+        featmap_sizes = [
+            pred_maps[i].shape[-2:] for i in range(self.num_levels)
+        ]
+        multi_level_anchors = self.anchor_generator.grid_anchors(
+            featmap_sizes, device)
+        anchor_list = [multi_level_anchors for _ in range(num_imgs)]
+
+        responsible_flag_list = []
+        for img_id in range(len(img_metas)):
+            responsible_flag_list.append(
+                self.anchor_generator.responsible_flags(
+                    featmap_sizes, gt_bboxes[img_id], device))
+
+        target_maps_list, neg_maps_list = self.get_targets(
+            anchor_list, responsible_flag_list, gt_bboxes, gt_labels)
+
+        losses_cls, losses_conf, losses_xy, losses_wh = multi_apply(
+            self.loss_single, pred_maps, target_maps_list, neg_maps_list)
+
+        return dict(
+            loss_cls=losses_cls,
+            loss_conf=losses_conf,
+            loss_xy=losses_xy,
+            loss_wh=losses_wh)
+
+    def loss_single(self, pred_map, target_map, neg_map):
+        """Compute loss of a single image from a batch.
+
+        Args:
+            pred_map (Tensor): Raw predictions for a single level.
+            target_map (Tensor): The Ground-Truth target for a single level.
+            neg_map (Tensor): The negative masks for a single level.
+
+        Returns:
+            tuple:
+                loss_cls (Tensor): Classification loss.
+                loss_conf (Tensor): Confidence loss.
+                loss_xy (Tensor): Regression loss of x, y coordinate.
+                loss_wh (Tensor): Regression loss of w, h coordinate.
+        """
+
+        num_imgs = len(pred_map)
+        pred_map = pred_map.permute(0, 2, 3,
+                                    1).reshape(num_imgs, -1, self.num_attrib)
+        neg_mask = neg_map.float()
+        pos_mask = target_map[..., 4]
+        pos_and_neg_mask = neg_mask + pos_mask
+        pos_mask = pos_mask.unsqueeze(dim=-1)
+        if torch.max(pos_and_neg_mask) > 1.:
+            warnings.warn('There is overlap between pos and neg sample.')
+            pos_and_neg_mask = pos_and_neg_mask.clamp(min=0., max=1.)
+
+        pred_xy = pred_map[..., :2]
+        pred_wh = pred_map[..., 2:4]
+        pred_conf = pred_map[..., 4]
+        pred_label = pred_map[..., 5:]
+
+        target_xy = target_map[..., :2]
+        target_wh = target_map[..., 2:4]
+        target_conf = target_map[..., 4]
+        target_label = target_map[..., 5:]
+
+        loss_cls = self.loss_cls(pred_label, target_label, weight=pos_mask)
+        loss_conf = self.loss_conf(
+            pred_conf, target_conf, weight=pos_and_neg_mask)
+        loss_xy = self.loss_xy(pred_xy, target_xy, weight=pos_mask)
+        loss_wh = self.loss_wh(pred_wh, target_wh, weight=pos_mask)
+
+        return loss_cls, loss_conf, loss_xy, loss_wh
+
+    def get_targets(self, anchor_list, responsible_flag_list, gt_bboxes_list,
+                    gt_labels_list):
+        """Compute target maps for anchors in multiple images.
+
+        Args:
+            anchor_list (list[list[Tensor]]): Multi level anchors of each
+                image. The outer list indicates images, and the inner list
+                corresponds to feature levels of the image. Each element of
+                the inner list is a tensor of shape (num_total_anchors, 4).
+            responsible_flag_list (list[list[Tensor]]): Multi level responsible
+                flags of each image. Each element is a tensor of shape
+                (num_total_anchors, )
+            gt_bboxes_list (list[Tensor]): Ground truth bboxes of each image.
+            gt_labels_list (list[Tensor]): Ground truth labels of each box.
+
+        Returns:
+            tuple: Usually returns a tuple containing learning targets.
+                - target_map_list (list[Tensor]): Target map of each level.
+                - neg_map_list (list[Tensor]): Negative map of each level.
+        """
+        num_imgs = len(anchor_list)
+
+        # anchor number of multi levels
+        num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]]
+
+        results = multi_apply(self._get_targets_single, anchor_list,
+                              responsible_flag_list, gt_bboxes_list,
+                              gt_labels_list)
+
+        all_target_maps, all_neg_maps = results
+        assert num_imgs == len(all_target_maps) == len(all_neg_maps)
+        target_maps_list = images_to_levels(all_target_maps, num_level_anchors)
+        neg_maps_list = images_to_levels(all_neg_maps, num_level_anchors)
+
+        return target_maps_list, neg_maps_list
+
+    def _get_targets_single(self, anchors, responsible_flags, gt_bboxes,
+                            gt_labels):
+        """Generate matching bounding box prior and converted GT.
+
+        Args:
+            anchors (list[Tensor]): Multi-level anchors of the image.
+            responsible_flags (list[Tensor]): Multi-level responsible flags of
+                anchors
+            gt_bboxes (Tensor): Ground truth bboxes of single image.
+            gt_labels (Tensor): Ground truth labels of single image.
+
+        Returns:
+            tuple:
+                target_map (Tensor): Predication target map of each
+                    scale level, shape (num_total_anchors,
+                    5+num_classes)
+                neg_map (Tensor): Negative map of each scale level,
+                    shape (num_total_anchors,)
+        """
+
+        anchor_strides = []
+        for i in range(len(anchors)):
+            anchor_strides.append(
+                torch.tensor(self.featmap_strides[i],
+                             device=gt_bboxes.device).repeat(len(anchors[i])))
+        concat_anchors = torch.cat(anchors)
+        concat_responsible_flags = torch.cat(responsible_flags)
+
+        anchor_strides = torch.cat(anchor_strides)
+        assert len(anchor_strides) == len(concat_anchors) == \
+               len(concat_responsible_flags)
+        assign_result = self.assigner.assign(concat_anchors,
+                                             concat_responsible_flags,
+                                             gt_bboxes)
+        sampling_result = self.sampler.sample(assign_result, concat_anchors,
+                                              gt_bboxes)
+
+        target_map = concat_anchors.new_zeros(
+            concat_anchors.size(0), self.num_attrib)
+
+        target_map[sampling_result.pos_inds, :4] = self.bbox_coder.encode(
+            sampling_result.pos_bboxes, sampling_result.pos_gt_bboxes,
+            anchor_strides[sampling_result.pos_inds])
+
+        target_map[sampling_result.pos_inds, 4] = 1
+
+        gt_labels_one_hot = F.one_hot(
+            gt_labels, num_classes=self.num_classes).float()
+        if self.one_hot_smoother != 0:  # label smooth
+            gt_labels_one_hot = gt_labels_one_hot * (
+                1 - self.one_hot_smoother
+            ) + self.one_hot_smoother / self.num_classes
+        target_map[sampling_result.pos_inds, 5:] = gt_labels_one_hot[
+            sampling_result.pos_assigned_gt_inds]
+
+        neg_map = concat_anchors.new_zeros(
+            concat_anchors.size(0), dtype=torch.uint8)
+        neg_map[sampling_result.neg_inds] = 1
+
+        return target_map, neg_map
+
+    def aug_test(self, feats, img_metas, rescale=False):
+        """Test function with test time augmentation.
+
+        Args:
+            feats (list[Tensor]): the outer list indicates test-time
+                augmentations and inner Tensor should have a shape NxCxHxW,
+                which contains features for all images in the batch.
+            img_metas (list[list[dict]]): the outer list indicates test-time
+                augs (multiscale, flip, etc.) and the inner list indicates
+                images in a batch. each dict has image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[ndarray]: bbox results of each class
+        """
+        return self.aug_test_bboxes(feats, img_metas, rescale=rescale)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/__init__.py b/python/autoCut/models/swin/mmdet/models/detectors/__init__.py
new file mode 100644
index 0000000..0d7fe18
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/__init__.py
@@ -0,0 +1,42 @@
+from .atss import ATSS
+from .base import BaseDetector
+from .cascade_rcnn import CascadeRCNN
+from .cornernet import CornerNet
+from .detr import DETR
+from .fast_rcnn import FastRCNN
+from .faster_rcnn import FasterRCNN
+from .fcos import FCOS
+from .fovea import FOVEA
+from .fsaf import FSAF
+from .gfl import GFL
+from .grid_rcnn import GridRCNN
+from .htc import HybridTaskCascade
+from .kd_one_stage import KnowledgeDistillationSingleStageDetector
+from .mask_rcnn import MaskRCNN
+from .mask_scoring_rcnn import MaskScoringRCNN
+from .nasfcos import NASFCOS
+from .paa import PAA
+from .point_rend import PointRend
+from .reppoints_detector import RepPointsDetector
+from .reppoints_v2_detector import RepPointsV2Detector
+from .mask_reppoints_v2_detector import RepPointsV2MaskDetector
+from .retinanet import RetinaNet
+from .rpn import RPN
+from .scnet import SCNet
+from .single_stage import SingleStageDetector
+from .sparse_rcnn import SparseRCNN
+from .trident_faster_rcnn import TridentFasterRCNN
+from .two_stage import TwoStageDetector
+from .vfnet import VFNet
+from .yolact import YOLACT
+from .yolo import YOLOV3
+
+__all__ = [
+    'ATSS', 'BaseDetector', 'SingleStageDetector',
+    'KnowledgeDistillationSingleStageDetector', 'TwoStageDetector', 'RPN',
+    'FastRCNN', 'FasterRCNN', 'MaskRCNN', 'CascadeRCNN', 'HybridTaskCascade',
+    'RetinaNet', 'FCOS', 'GridRCNN', 'MaskScoringRCNN', 'RepPointsDetector',
+    'FOVEA', 'FSAF', 'NASFCOS', 'PointRend', 'GFL', 'CornerNet', 'PAA',
+    'YOLOV3', 'YOLACT', 'VFNet', 'DETR', 'TridentFasterRCNN', 'SparseRCNN',
+    'SCNet', 'RepPointsV2Detector', 'RepPointsV2MaskDetector'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/atss.py b/python/autoCut/models/swin/mmdet/models/detectors/atss.py
new file mode 100644
index 0000000..db7139c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/atss.py
@@ -0,0 +1,17 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class ATSS(SingleStageDetector):
+    """Implementation of `ATSS <https://arxiv.org/abs/1912.02424>`_."""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(ATSS, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                   test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/base.py b/python/autoCut/models/swin/mmdet/models/detectors/base.py
new file mode 100644
index 0000000..89134f3
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/base.py
@@ -0,0 +1,355 @@
+from abc import ABCMeta, abstractmethod
+from collections import OrderedDict
+
+import mmcv
+import numpy as np
+import torch
+import torch.distributed as dist
+import torch.nn as nn
+from mmcv.runner import auto_fp16
+from mmcv.utils import print_log
+
+from mmdet.core.visualization import imshow_det_bboxes
+from mmdet.utils import get_root_logger
+
+
+class BaseDetector(nn.Module, metaclass=ABCMeta):
+    """Base class for detectors."""
+
+    def __init__(self):
+        super(BaseDetector, self).__init__()
+        self.fp16_enabled = False
+
+    @property
+    def with_neck(self):
+        """bool: whether the detector has a neck"""
+        return hasattr(self, 'neck') and self.neck is not None
+
+    # TODO: these properties need to be carefully handled
+    # for both single stage & two stage detectors
+    @property
+    def with_shared_head(self):
+        """bool: whether the detector has a shared head in the RoI Head"""
+        return hasattr(self, 'roi_head') and self.roi_head.with_shared_head
+
+    @property
+    def with_bbox(self):
+        """bool: whether the detector has a bbox head"""
+        return ((hasattr(self, 'roi_head') and self.roi_head.with_bbox)
+                or (hasattr(self, 'bbox_head') and self.bbox_head is not None))
+
+    @property
+    def with_mask(self):
+        """bool: whether the detector has a mask head"""
+        return ((hasattr(self, 'roi_head') and self.roi_head.with_mask)
+                or (hasattr(self, 'mask_head') and self.mask_head is not None))
+
+    @abstractmethod
+    def extract_feat(self, imgs):
+        """Extract features from images."""
+        pass
+
+    def extract_feats(self, imgs):
+        """Extract features from multiple images.
+
+        Args:
+            imgs (list[torch.Tensor]): A list of images. The images are
+                augmented from the same image but in different ways.
+
+        Returns:
+            list[torch.Tensor]: Features of different images
+        """
+        assert isinstance(imgs, list)
+        return [self.extract_feat(img) for img in imgs]
+
+    def forward_train(self, imgs, img_metas, **kwargs):
+        """
+        Args:
+            img (list[Tensor]): List of tensors of shape (1, C, H, W).
+                Typically these should be mean centered and std scaled.
+            img_metas (list[dict]): List of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys, see
+                :class:`mmdet.datasets.pipelines.Collect`.
+            kwargs (keyword arguments): Specific to concrete implementation.
+        """
+        # NOTE the batched image size information may be useful, e.g.
+        # in DETR, this is needed for the construction of masks, which is
+        # then used for the transformer_head.
+        batch_input_shape = tuple(imgs[0].size()[-2:])
+        for img_meta in img_metas:
+            img_meta['batch_input_shape'] = batch_input_shape
+
+    async def async_simple_test(self, img, img_metas, **kwargs):
+        raise NotImplementedError
+
+    @abstractmethod
+    def simple_test(self, img, img_metas, **kwargs):
+        pass
+
+    @abstractmethod
+    def aug_test(self, imgs, img_metas, **kwargs):
+        """Test function with test time augmentation."""
+        pass
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in detector.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        if pretrained is not None:
+            logger = get_root_logger()
+            print_log(f'load model from: {pretrained}', logger=logger)
+
+    async def aforward_test(self, *, img, img_metas, **kwargs):
+        for var, name in [(img, 'img'), (img_metas, 'img_metas')]:
+            if not isinstance(var, list):
+                raise TypeError(f'{name} must be a list, but got {type(var)}')
+
+        num_augs = len(img)
+        if num_augs != len(img_metas):
+            raise ValueError(f'num of augmentations ({len(img)}) '
+                             f'!= num of image metas ({len(img_metas)})')
+        # TODO: remove the restriction of samples_per_gpu == 1 when prepared
+        samples_per_gpu = img[0].size(0)
+        assert samples_per_gpu == 1
+
+        if num_augs == 1:
+            return await self.async_simple_test(img[0], img_metas[0], **kwargs)
+        else:
+            raise NotImplementedError
+
+    def forward_test(self, imgs, img_metas, **kwargs):
+        """
+        Args:
+            imgs (List[Tensor]): the outer list indicates test-time
+                augmentations and inner Tensor should have a shape NxCxHxW,
+                which contains all images in the batch.
+            img_metas (List[List[dict]]): the outer list indicates test-time
+                augs (multiscale, flip, etc.) and the inner list indicates
+                images in a batch.
+        """
+        for var, name in [(imgs, 'imgs'), (img_metas, 'img_metas')]:
+            if not isinstance(var, list):
+                raise TypeError(f'{name} must be a list, but got {type(var)}')
+
+        num_augs = len(imgs)
+        if num_augs != len(img_metas):
+            raise ValueError(f'num of augmentations ({len(imgs)}) '
+                             f'!= num of image meta ({len(img_metas)})')
+
+        # NOTE the batched image size information may be useful, e.g.
+        # in DETR, this is needed for the construction of masks, which is
+        # then used for the transformer_head.
+        for img, img_meta in zip(imgs, img_metas):
+            batch_size = len(img_meta)
+            for img_id in range(batch_size):
+                img_meta[img_id]['batch_input_shape'] = tuple(img.size()[-2:])
+
+        if num_augs == 1:
+            # proposals (List[List[Tensor]]): the outer list indicates
+            # test-time augs (multiscale, flip, etc.) and the inner list
+            # indicates images in a batch.
+            # The Tensor should have a shape Px4, where P is the number of
+            # proposals.
+            if 'proposals' in kwargs:
+                kwargs['proposals'] = kwargs['proposals'][0]
+            return self.simple_test(imgs[0], img_metas[0], **kwargs)
+        else:
+            assert imgs[0].size(0) == 1, 'aug test does not support ' \
+                                         'inference with batch size ' \
+                                         f'{imgs[0].size(0)}'
+            # TODO: support test augmentation for predefined proposals
+            assert 'proposals' not in kwargs
+            return self.aug_test(imgs, img_metas, **kwargs)
+
+    @auto_fp16(apply_to=('img', ))
+    def forward(self, img, img_metas, return_loss=True, **kwargs):
+        """Calls either :func:`forward_train` or :func:`forward_test` depending
+        on whether ``return_loss`` is ``True``.
+
+        Note this setting will change the expected inputs. When
+        ``return_loss=True``, img and img_meta are single-nested (i.e. Tensor
+        and List[dict]), and when ``resturn_loss=False``, img and img_meta
+        should be double nested (i.e.  List[Tensor], List[List[dict]]), with
+        the outer list indicating test time augmentations.
+        """
+        if return_loss:
+            return self.forward_train(img, img_metas, **kwargs)
+        else:
+            return self.forward_test(img, img_metas, **kwargs)
+
+    def _parse_losses(self, losses):
+        """Parse the raw outputs (losses) of the network.
+
+        Args:
+            losses (dict): Raw output of the network, which usually contain
+                losses and other necessary infomation.
+
+        Returns:
+            tuple[Tensor, dict]: (loss, log_vars), loss is the loss tensor \
+                which may be a weighted sum of all losses, log_vars contains \
+                all the variables to be sent to the logger.
+        """
+        log_vars = OrderedDict()
+        for loss_name, loss_value in losses.items():
+            if isinstance(loss_value, torch.Tensor):
+                log_vars[loss_name] = loss_value.mean()
+            elif isinstance(loss_value, list):
+                log_vars[loss_name] = sum(_loss.mean() for _loss in loss_value)
+            else:
+                raise TypeError(
+                    f'{loss_name} is not a tensor or list of tensors')
+
+        loss = sum(_value for _key, _value in log_vars.items()
+                   if 'loss' in _key)
+
+        log_vars['loss'] = loss
+        for loss_name, loss_value in log_vars.items():
+            # reduce loss when distributed training
+            if dist.is_available() and dist.is_initialized():
+                loss_value = loss_value.data.clone()
+                dist.all_reduce(loss_value.div_(dist.get_world_size()))
+            log_vars[loss_name] = loss_value.item()
+
+        return loss, log_vars
+
+    def train_step(self, data, optimizer):
+        """The iteration step during training.
+
+        This method defines an iteration step during training, except for the
+        back propagation and optimizer updating, which are done in an optimizer
+        hook. Note that in some complicated cases or models, the whole process
+        including back propagation and optimizer updating is also defined in
+        this method, such as GAN.
+
+        Args:
+            data (dict): The output of dataloader.
+            optimizer (:obj:`torch.optim.Optimizer` | dict): The optimizer of
+                runner is passed to ``train_step()``. This argument is unused
+                and reserved.
+
+        Returns:
+            dict: It should contain at least 3 keys: ``loss``, ``log_vars``, \
+                ``num_samples``.
+
+                - ``loss`` is a tensor for back propagation, which can be a \
+                weighted sum of multiple losses.
+                - ``log_vars`` contains all the variables to be sent to the
+                logger.
+                - ``num_samples`` indicates the batch size (when the model is \
+                DDP, it means the batch size on each GPU), which is used for \
+                averaging the logs.
+        """
+        losses = self(**data)
+        loss, log_vars = self._parse_losses(losses)
+
+        outputs = dict(
+            loss=loss, log_vars=log_vars, num_samples=len(data['img_metas']))
+
+        return outputs
+
+    def val_step(self, data, optimizer):
+        """The iteration step during validation.
+
+        This method shares the same signature as :func:`train_step`, but used
+        during val epochs. Note that the evaluation after training epochs is
+        not implemented with this method, but an evaluation hook.
+        """
+        losses = self(**data)
+        loss, log_vars = self._parse_losses(losses)
+
+        outputs = dict(
+            loss=loss, log_vars=log_vars, num_samples=len(data['img_metas']))
+
+        return outputs
+
+    def show_result(self,
+                    img,
+                    result,
+                    score_thr=0.3,
+                    bbox_color=(72, 101, 241),
+                    text_color=(72, 101, 241),
+                    mask_color=None,
+                    thickness=2,
+                    font_size=13,
+                    win_name='',
+                    show=False,
+                    wait_time=0,
+                    out_file=None):
+        """Draw `result` over `img`.
+
+        Args:
+            img (str or Tensor): The image to be displayed.
+            result (Tensor or tuple): The results to draw over `img`
+                bbox_result or (bbox_result, segm_result).
+            score_thr (float, optional): Minimum score of bboxes to be shown.
+                Default: 0.3.
+            bbox_color (str or tuple(int) or :obj:`Color`):Color of bbox lines.
+               The tuple of color should be in BGR order. Default: 'green'
+            text_color (str or tuple(int) or :obj:`Color`):Color of texts.
+               The tuple of color should be in BGR order. Default: 'green'
+            mask_color (None or str or tuple(int) or :obj:`Color`):
+               Color of masks. The tuple of color should be in BGR order.
+               Default: None
+            thickness (int): Thickness of lines. Default: 2
+            font_size (int): Font size of texts. Default: 13
+            win_name (str): The window name. Default: ''
+            wait_time (float): Value of waitKey param.
+                Default: 0.
+            show (bool): Whether to show the image.
+                Default: False.
+            out_file (str or None): The filename to write the image.
+                Default: None.
+
+        Returns:
+            img (Tensor): Only if not `show` or `out_file`
+        """
+        img = mmcv.imread(img)
+        img = img.copy()
+        if isinstance(result, tuple):
+            bbox_result, segm_result = result
+            if isinstance(segm_result, tuple):
+                segm_result = segm_result[0]  # ms rcnn
+        else:
+            bbox_result, segm_result = result, None
+        bboxes = np.vstack(bbox_result)
+        labels = [
+            np.full(bbox.shape[0], i, dtype=np.int32)
+            for i, bbox in enumerate(bbox_result)
+        ]
+        labels = np.concatenate(labels)
+        # draw segmentation masks
+        segms = None
+        if segm_result is not None and len(labels) > 0:  # non empty
+            segms = mmcv.concat_list(segm_result)
+            if isinstance(segms[0], torch.Tensor):
+                segms = torch.stack(segms, dim=0).detach().cpu().numpy()
+            else:
+                segms = np.stack(segms, axis=0)
+        # if out_file specified, do not show image in window
+        if out_file is not None:
+            show = False
+        # draw bounding boxes
+        img = imshow_det_bboxes(
+            img,
+            bboxes,
+            labels,
+            segms,
+            class_names=self.CLASSES,
+            score_thr=score_thr,
+            bbox_color=bbox_color,
+            text_color=text_color,
+            mask_color=mask_color,
+            thickness=thickness,
+            font_size=font_size,
+            win_name=win_name,
+            show=show,
+            wait_time=wait_time,
+            out_file=out_file)
+
+        if not (show or out_file):
+            return img
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/cascade_rcnn.py b/python/autoCut/models/swin/mmdet/models/detectors/cascade_rcnn.py
new file mode 100644
index 0000000..d873dce
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/cascade_rcnn.py
@@ -0,0 +1,46 @@
+from ..builder import DETECTORS
+from .two_stage import TwoStageDetector
+
+
+@DETECTORS.register_module()
+class CascadeRCNN(TwoStageDetector):
+    r"""Implementation of `Cascade R-CNN: Delving into High Quality Object
+    Detection <https://arxiv.org/abs/1906.09756>`_"""
+
+    def __init__(self,
+                 backbone,
+                 neck=None,
+                 rpn_head=None,
+                 roi_head=None,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(CascadeRCNN, self).__init__(
+            backbone=backbone,
+            neck=neck,
+            rpn_head=rpn_head,
+            roi_head=roi_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg,
+            pretrained=pretrained)
+
+    def show_result(self, data, result, **kwargs):
+        """Show prediction results of the detector.
+
+        Args:
+            data (str or np.ndarray): Image filename or loaded image.
+            result (Tensor or tuple): The results to draw over `img`
+                bbox_result or (bbox_result, segm_result).
+
+        Returns:
+            np.ndarray: The image with bboxes drawn on it.
+        """
+        if self.with_mask:
+            ms_bbox_result, ms_segm_result = result
+            if isinstance(ms_bbox_result, dict):
+                result = (ms_bbox_result['ensemble'],
+                          ms_segm_result['ensemble'])
+        else:
+            if isinstance(result, dict):
+                result = result['ensemble']
+        return super(CascadeRCNN, self).show_result(data, result, **kwargs)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/cornernet.py b/python/autoCut/models/swin/mmdet/models/detectors/cornernet.py
new file mode 100644
index 0000000..bb8ccc1
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/cornernet.py
@@ -0,0 +1,95 @@
+import torch
+
+from mmdet.core import bbox2result, bbox_mapping_back
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class CornerNet(SingleStageDetector):
+    """CornerNet.
+
+    This detector is the implementation of the paper `CornerNet: Detecting
+    Objects as Paired Keypoints <https://arxiv.org/abs/1808.01244>`_ .
+    """
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(CornerNet, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                        test_cfg, pretrained)
+
+    def merge_aug_results(self, aug_results, img_metas):
+        """Merge augmented detection bboxes and score.
+
+        Args:
+            aug_results (list[list[Tensor]]): Det_bboxes and det_labels of each
+                image.
+            img_metas (list[list[dict]]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+
+        Returns:
+            tuple: (bboxes, labels)
+        """
+        recovered_bboxes, aug_labels = [], []
+        for bboxes_labels, img_info in zip(aug_results, img_metas):
+            img_shape = img_info[0]['img_shape']  # using shape before padding
+            scale_factor = img_info[0]['scale_factor']
+            flip = img_info[0]['flip']
+            bboxes, labels = bboxes_labels
+            bboxes, scores = bboxes[:, :4], bboxes[:, -1:]
+            bboxes = bbox_mapping_back(bboxes, img_shape, scale_factor, flip)
+            recovered_bboxes.append(torch.cat([bboxes, scores], dim=-1))
+            aug_labels.append(labels)
+
+        bboxes = torch.cat(recovered_bboxes, dim=0)
+        labels = torch.cat(aug_labels)
+
+        if bboxes.shape[0] > 0:
+            out_bboxes, out_labels = self.bbox_head._bboxes_nms(
+                bboxes, labels, self.bbox_head.test_cfg)
+        else:
+            out_bboxes, out_labels = bboxes, labels
+
+        return out_bboxes, out_labels
+
+    def aug_test(self, imgs, img_metas, rescale=False):
+        """Augment testing of CornerNet.
+
+        Args:
+            imgs (list[Tensor]): Augmented images.
+            img_metas (list[list[dict]]): Meta information of each image, e.g.,
+                image size, scaling factor, etc.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+
+        Note:
+            ``imgs`` must including flipped image pairs.
+
+        Returns:
+            list[list[np.ndarray]]: BBox results of each image and classes.
+                The outer list corresponds to each image. The inner list
+                corresponds to each class.
+        """
+        img_inds = list(range(len(imgs)))
+
+        assert img_metas[0][0]['flip'] + img_metas[1][0]['flip'], (
+            'aug test must have flipped image pair')
+        aug_results = []
+        for ind, flip_ind in zip(img_inds[0::2], img_inds[1::2]):
+            img_pair = torch.cat([imgs[ind], imgs[flip_ind]])
+            x = self.extract_feat(img_pair)
+            outs = self.bbox_head(x)
+            bbox_list = self.bbox_head.get_bboxes(
+                *outs, [img_metas[ind], img_metas[flip_ind]], False, False)
+            aug_results.append(bbox_list[0])
+            aug_results.append(bbox_list[1])
+
+        bboxes, labels = self.merge_aug_results(aug_results, img_metas)
+        bbox_results = bbox2result(bboxes, labels, self.bbox_head.num_classes)
+
+        return [bbox_results]
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/detr.py b/python/autoCut/models/swin/mmdet/models/detectors/detr.py
new file mode 100644
index 0000000..5ff82a2
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/detr.py
@@ -0,0 +1,46 @@
+from mmdet.core import bbox2result
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class DETR(SingleStageDetector):
+    r"""Implementation of `DETR: End-to-End Object Detection with
+    Transformers <https://arxiv.org/pdf/2005.12872>`_"""
+
+    def __init__(self,
+                 backbone,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(DETR, self).__init__(backbone, None, bbox_head, train_cfg,
+                                   test_cfg, pretrained)
+
+    def simple_test(self, img, img_metas, rescale=False):
+        """Test function without test time augmentation.
+
+        Args:
+            imgs (list[torch.Tensor]): List of multiple images
+            img_metas (list[dict]): List of image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[list[np.ndarray]]: BBox results of each image and classes.
+                The outer list corresponds to each image. The inner list
+                corresponds to each class.
+        """
+        batch_size = len(img_metas)
+        assert batch_size == 1, 'Currently only batch_size 1 for inference ' \
+            f'mode is supported. Found batch_size {batch_size}.'
+        x = self.extract_feat(img)
+        outs = self.bbox_head(x, img_metas)
+        bbox_list = self.bbox_head.get_bboxes(
+            *outs, img_metas, rescale=rescale)
+
+        bbox_results = [
+            bbox2result(det_bboxes, det_labels, self.bbox_head.num_classes)
+            for det_bboxes, det_labels in bbox_list
+        ]
+        return bbox_results
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/fast_rcnn.py b/python/autoCut/models/swin/mmdet/models/detectors/fast_rcnn.py
new file mode 100644
index 0000000..3d6e242
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/fast_rcnn.py
@@ -0,0 +1,52 @@
+from ..builder import DETECTORS
+from .two_stage import TwoStageDetector
+
+
+@DETECTORS.register_module()
+class FastRCNN(TwoStageDetector):
+    """Implementation of `Fast R-CNN <https://arxiv.org/abs/1504.08083>`_"""
+
+    def __init__(self,
+                 backbone,
+                 roi_head,
+                 train_cfg,
+                 test_cfg,
+                 neck=None,
+                 pretrained=None):
+        super(FastRCNN, self).__init__(
+            backbone=backbone,
+            neck=neck,
+            roi_head=roi_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg,
+            pretrained=pretrained)
+
+    def forward_test(self, imgs, img_metas, proposals, **kwargs):
+        """
+        Args:
+            imgs (List[Tensor]): the outer list indicates test-time
+                augmentations and inner Tensor should have a shape NxCxHxW,
+                which contains all images in the batch.
+            img_metas (List[List[dict]]): the outer list indicates test-time
+                augs (multiscale, flip, etc.) and the inner list indicates
+                images in a batch.
+            proposals (List[List[Tensor]]): the outer list indicates test-time
+                augs (multiscale, flip, etc.) and the inner list indicates
+                images in a batch. The Tensor should have a shape Px4, where
+                P is the number of proposals.
+        """
+        for var, name in [(imgs, 'imgs'), (img_metas, 'img_metas')]:
+            if not isinstance(var, list):
+                raise TypeError(f'{name} must be a list, but got {type(var)}')
+
+        num_augs = len(imgs)
+        if num_augs != len(img_metas):
+            raise ValueError(f'num of augmentations ({len(imgs)}) '
+                             f'!= num of image meta ({len(img_metas)})')
+
+        if num_augs == 1:
+            return self.simple_test(imgs[0], img_metas[0], proposals[0],
+                                    **kwargs)
+        else:
+            # TODO: support test-time augmentation
+            assert NotImplementedError
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/faster_rcnn.py b/python/autoCut/models/swin/mmdet/models/detectors/faster_rcnn.py
new file mode 100644
index 0000000..81bad0f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/faster_rcnn.py
@@ -0,0 +1,24 @@
+from ..builder import DETECTORS
+from .two_stage import TwoStageDetector
+
+
+@DETECTORS.register_module()
+class FasterRCNN(TwoStageDetector):
+    """Implementation of `Faster R-CNN <https://arxiv.org/abs/1506.01497>`_"""
+
+    def __init__(self,
+                 backbone,
+                 rpn_head,
+                 roi_head,
+                 train_cfg,
+                 test_cfg,
+                 neck=None,
+                 pretrained=None):
+        super(FasterRCNN, self).__init__(
+            backbone=backbone,
+            neck=neck,
+            rpn_head=rpn_head,
+            roi_head=roi_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg,
+            pretrained=pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/fcos.py b/python/autoCut/models/swin/mmdet/models/detectors/fcos.py
new file mode 100644
index 0000000..58485c1
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/fcos.py
@@ -0,0 +1,17 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class FCOS(SingleStageDetector):
+    """Implementation of `FCOS <https://arxiv.org/abs/1904.01355>`_"""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(FCOS, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                   test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/fovea.py b/python/autoCut/models/swin/mmdet/models/detectors/fovea.py
new file mode 100644
index 0000000..22a578e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/fovea.py
@@ -0,0 +1,17 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class FOVEA(SingleStageDetector):
+    """Implementation of `FoveaBox <https://arxiv.org/abs/1904.03797>`_"""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(FOVEA, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                    test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/fsaf.py b/python/autoCut/models/swin/mmdet/models/detectors/fsaf.py
new file mode 100644
index 0000000..9f10fa1
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/fsaf.py
@@ -0,0 +1,17 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class FSAF(SingleStageDetector):
+    """Implementation of `FSAF <https://arxiv.org/abs/1903.00621>`_"""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(FSAF, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                   test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/gfl.py b/python/autoCut/models/swin/mmdet/models/detectors/gfl.py
new file mode 100644
index 0000000..64d65cb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/gfl.py
@@ -0,0 +1,16 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class GFL(SingleStageDetector):
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(GFL, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                  test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/grid_rcnn.py b/python/autoCut/models/swin/mmdet/models/detectors/grid_rcnn.py
new file mode 100644
index 0000000..b6145a1
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/grid_rcnn.py
@@ -0,0 +1,29 @@
+from ..builder import DETECTORS
+from .two_stage import TwoStageDetector
+
+
+@DETECTORS.register_module()
+class GridRCNN(TwoStageDetector):
+    """Grid R-CNN.
+
+    This detector is the implementation of:
+    - Grid R-CNN (https://arxiv.org/abs/1811.12030)
+    - Grid R-CNN Plus: Faster and Better (https://arxiv.org/abs/1906.05688)
+    """
+
+    def __init__(self,
+                 backbone,
+                 rpn_head,
+                 roi_head,
+                 train_cfg,
+                 test_cfg,
+                 neck=None,
+                 pretrained=None):
+        super(GridRCNN, self).__init__(
+            backbone=backbone,
+            neck=neck,
+            rpn_head=rpn_head,
+            roi_head=roi_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg,
+            pretrained=pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/htc.py b/python/autoCut/models/swin/mmdet/models/detectors/htc.py
new file mode 100644
index 0000000..d9efdf4
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/htc.py
@@ -0,0 +1,15 @@
+from ..builder import DETECTORS
+from .cascade_rcnn import CascadeRCNN
+
+
+@DETECTORS.register_module()
+class HybridTaskCascade(CascadeRCNN):
+    """Implementation of `HTC <https://arxiv.org/abs/1901.07518>`_"""
+
+    def __init__(self, **kwargs):
+        super(HybridTaskCascade, self).__init__(**kwargs)
+
+    @property
+    def with_semantic(self):
+        """bool: whether the detector has a semantic head"""
+        return self.roi_head.with_semantic
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/kd_one_stage.py b/python/autoCut/models/swin/mmdet/models/detectors/kd_one_stage.py
new file mode 100644
index 0000000..671ec19
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/kd_one_stage.py
@@ -0,0 +1,100 @@
+import mmcv
+import torch
+from mmcv.runner import load_checkpoint
+
+from .. import build_detector
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class KnowledgeDistillationSingleStageDetector(SingleStageDetector):
+    r"""Implementation of `Distilling the Knowledge in a Neural Network.
+    <https://arxiv.org/abs/1503.02531>`_.
+
+    Args:
+        teacher_config (str | dict): Config file path
+            or the config object of teacher model.
+        teacher_ckpt (str, optional): Checkpoint path of teacher model.
+            If left as None, the model will not load any weights.
+    """
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 teacher_config,
+                 teacher_ckpt=None,
+                 eval_teacher=True,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super().__init__(backbone, neck, bbox_head, train_cfg, test_cfg,
+                         pretrained)
+        self.eval_teacher = eval_teacher
+        # Build teacher model
+        if isinstance(teacher_config, str):
+            teacher_config = mmcv.Config.fromfile(teacher_config)
+        self.teacher_model = build_detector(teacher_config['model'])
+        if teacher_ckpt is not None:
+            load_checkpoint(
+                self.teacher_model, teacher_ckpt, map_location='cpu')
+
+    def forward_train(self,
+                      img,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None):
+        """
+        Args:
+            img (Tensor): Input images of shape (N, C, H, W).
+                Typically these should be mean centered and std scaled.
+            img_metas (list[dict]): A List of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                :class:`mmdet.datasets.pipelines.Collect`.
+            gt_bboxes (list[Tensor]): Each item are the truth boxes for each
+                image in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): Class indices corresponding to each box
+            gt_bboxes_ignore (None | list[Tensor]): Specify which bounding
+                boxes can be ignored when computing the loss.
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        x = self.extract_feat(img)
+        with torch.no_grad():
+            teacher_x = self.teacher_model.extract_feat(img)
+            out_teacher = self.teacher_model.bbox_head(teacher_x)
+        losses = self.bbox_head.forward_train(x, out_teacher, img_metas,
+                                              gt_bboxes, gt_labels,
+                                              gt_bboxes_ignore)
+        return losses
+
+    def cuda(self, device=None):
+        """Since teacher_model is registered as a plain object, it is necessary
+        to put the teacher model to cuda when calling cuda function."""
+        self.teacher_model.cuda(device=device)
+        return super().cuda(device=device)
+
+    def train(self, mode=True):
+        """Set the same train mode for teacher and student model."""
+        if self.eval_teacher:
+            self.teacher_model.train(False)
+        else:
+            self.teacher_model.train(mode)
+        super().train(mode)
+
+    def __setattr__(self, name, value):
+        """Set attribute, i.e. self.name = value
+
+        This reloading prevent the teacher model from being registered as a
+        nn.Module. The teacher module is registered as a plain object, so that
+        the teacher parameters will not show up when calling
+        ``self.parameters``, ``self.modules``, ``self.children`` methods.
+        """
+        if name == 'teacher_model':
+            object.__setattr__(self, name, value)
+        else:
+            super().__setattr__(name, value)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/mask_rcnn.py b/python/autoCut/models/swin/mmdet/models/detectors/mask_rcnn.py
new file mode 100644
index 0000000..c15a773
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/mask_rcnn.py
@@ -0,0 +1,24 @@
+from ..builder import DETECTORS
+from .two_stage import TwoStageDetector
+
+
+@DETECTORS.register_module()
+class MaskRCNN(TwoStageDetector):
+    """Implementation of `Mask R-CNN <https://arxiv.org/abs/1703.06870>`_"""
+
+    def __init__(self,
+                 backbone,
+                 rpn_head,
+                 roi_head,
+                 train_cfg,
+                 test_cfg,
+                 neck=None,
+                 pretrained=None):
+        super(MaskRCNN, self).__init__(
+            backbone=backbone,
+            neck=neck,
+            rpn_head=rpn_head,
+            roi_head=roi_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg,
+            pretrained=pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/mask_reppoints_v2_detector.py b/python/autoCut/models/swin/mmdet/models/detectors/mask_reppoints_v2_detector.py
new file mode 100644
index 0000000..85c2fd6
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/mask_reppoints_v2_detector.py
@@ -0,0 +1,202 @@
+import numpy as np
+
+import torch
+
+from mmdet.core import bbox2result, bbox_mapping_back, multiclass_nms, merge_aug_masks, bbox_flip
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+import torch.nn.functional as F
+from mmdet.models.roi_heads.mask_heads.condconv_mask_head import aligned_bilinear
+import copy
+from PIL import Image 
+@DETECTORS.register_module()
+class RepPointsV2MaskDetector(SingleStageDetector):
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None,
+                 mask_inbox=False):
+        self.mask_inbox = mask_inbox
+        super(RepPointsV2MaskDetector, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                                test_cfg, pretrained)
+
+    def forward_train(self,
+                      img,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_sem_map=None,
+                      gt_sem_weights=None,
+                      gt_masks=None):
+        """
+        Args:
+            img (Tensor): of shape (N, C, H, W) encoding input images.
+                Typically these should be mean centered and std scaled.
+
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+
+            gt_labels (list[Tensor]): class indices corresponding to each box
+
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+            gt_masks (None | list[BitmapMasks]) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        masks = []
+        for mask in gt_masks:
+            mask_tensor = img.new_tensor(mask.masks)
+            mask_tensor = F.pad(mask_tensor, pad=(0, img.size(-1)-mask_tensor.size(-1), 0, img.size(-2)-mask_tensor.size(-2)))
+            masks.append(mask_tensor)
+        x = self.extract_feat(img)
+        losses = self.bbox_head.forward_train(x, img_metas, gt_bboxes,
+                                              gt_labels, gt_bboxes_ignore, gt_sem_map, gt_sem_weights, masks)
+        return losses
+        
+    def merge_aug_results(self, aug_bboxes, aug_scores, img_metas):
+        """Merge augmented detection bboxes and scores.
+
+        Args:
+            aug_bboxes (list[Tensor]): shape (n, 4*#class)
+            aug_scores (list[Tensor] or None): shape (n, #class)
+            img_shapes (list[Tensor]): shape (3, ).
+
+        Returns:
+            tuple: (bboxes, scores)
+        """
+        recovered_bboxes = []
+        for bboxes, img_info in zip(aug_bboxes, img_metas):
+            img_shape = img_info[0]['img_shape']
+            scale_factor = img_info[0]['scale_factor']
+            flip = img_info[0]['flip']
+            flip_direction = img_info[0]['flip_direction']
+            bboxes = bbox_mapping_back(bboxes, img_shape, scale_factor, flip,
+                                       flip_direction)
+            recovered_bboxes.append(bboxes)
+        bboxes = torch.cat(recovered_bboxes, dim=0)
+            
+        if aug_scores is None:
+            return bboxes
+        else:
+            scores = torch.cat(aug_scores, dim=0)
+            return bboxes, scores
+
+    def mask2result(self, x, det_labels, inst_inds, img_meta, det_bboxes, pred_instances=None, rescale=True, return_score=False):
+        resized_im_h, resized_im_w = img_meta['img_shape'][:2]
+        ori_h, ori_w = img_meta['ori_shape'][:2]
+        if pred_instances is not None:
+            pred_instances = pred_instances[inst_inds]
+        else:
+            pred_instances = self.bbox_head.pred_instances[inst_inds]
+
+        scale_factor = img_meta['scale_factor'] if rescale else [1, 1, 1, 1]
+        pred_instances.boxsz = torch.stack((det_bboxes[:, 2] * scale_factor[2] - det_bboxes[:, 0] * scale_factor[0],
+            det_bboxes[:, 3] * scale_factor[3] - det_bboxes[:, 1] * scale_factor[1]), axis=-1)
+        mask_logits = self.bbox_head.mask_head(x, pred_instances)
+        if len(pred_instances) > 0:
+            mask_logits = aligned_bilinear(mask_logits, self.bbox_head.mask_head.head.mask_out_stride)
+            mask_logits = mask_logits[:, :, :resized_im_h, :resized_im_w]
+            mask_logits = F.interpolate(
+                mask_logits,
+                size=(ori_h, ori_w),
+                mode="bilinear", align_corners=False
+            ).squeeze(1)
+            mask_pred = (mask_logits > 0.5)
+            flip = img_meta['flip']
+            flip_direction = img_meta['flip_direction']
+            if flip:
+                if flip_direction == 'horizontal':
+                    mask_pred = mask_pred.cpu().numpy()[:, :, ::-1]
+                elif flip_direction == 'vertical':
+                    mask_pred = mask_pred.cpu().numpy()[:, ::-1, :]
+                else:
+                    raise ValueError
+            else:
+               mask_pred = mask_pred.cpu().numpy()
+        else:
+            mask_pred = torch.zeros((self.bbox_head.num_classes, *img_meta['ori_shape'][:2]), dtype=torch.int)
+        cls_segms = [[] for _ in range(self.bbox_head.num_classes)]  # BG is not included in num_classes
+        cls_scores =[[] for _ in range(self.bbox_head.num_classes)]
+
+        for i, label in enumerate(det_labels):
+            score = det_bboxes[i][-1]
+            if self.mask_inbox:
+                mask_pred_ = torch.zeros_like(mask_pred[i])
+                det_bbox_ = det_bboxes[i, :-1].clone()
+                det_bbox_[[0, 1]], det_bbox_[[2, 3]] = det_bbox_[[0, 1]].floor(), det_bbox_[[2, 3]].ceil()
+                det_bbox_ = det_bbox_.int()
+                mask_pred_[det_bbox_[1]:det_bbox_[3], det_bbox_[0]:det_bbox_[2]] = mask_pred[i][det_bbox_[1]:det_bbox_[3], det_bbox_[0]:det_bbox_[2]]
+                cls_segms[label].append(mask_pred_.cpu().numpy())
+            else:
+                cls_segms[label].append(mask_pred[i].cpu().numpy() if isinstance(mask_pred[i], torch.Tensor) else mask_pred[i])
+            cls_scores[label].append(score.cpu().numpy())
+        
+        if return_score:
+            return cls_segms,cls_scores
+        return cls_segms
+
+    def simple_test(self, img, img_metas, rescale=False):
+        """Test function without test time augmentation
+
+        Args:
+            imgs (list[torch.Tensor]): List of multiple images
+            img_metas (list[dict]): List of image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+            [We all use True]
+        Returns:
+            np.ndarray: proposals
+        """
+        x = self.extract_feat(img)
+        outs = self.bbox_head(x)
+        bbox_list = self.bbox_head.get_bboxes(
+            *outs, img_metas, rescale=rescale)
+        if not self.bbox_head.mask_head: # detection only
+            bbox_results = [
+                bbox2result(det_bboxes, det_labels, self.bbox_head.num_classes)
+                for det_bboxes, det_labels in bbox_list
+            ]
+            return bbox_results
+        else:
+            bbox_results = [
+                bbox2result(det_bboxes, det_labels, self.bbox_head.num_classes)
+                for det_bboxes, det_labels, _ in bbox_list
+            ]
+            cls_segms = [
+                self.mask2result([xl[[i]] for xl in x], det_labels, inst_inds, img_metas[i], det_bboxes) 
+                for i, (det_bboxes, det_labels, inst_inds) in enumerate(bbox_list)
+            ]
+            
+            return list(zip(bbox_results, cls_segms))
+
+    def aug_test_simple(self, imgs, img_metas, rescale=False):
+        """Test function with test time augmentation
+
+        Args:
+            imgs (list[torch.Tensor]): List of multiple images
+            img_metas (list[dict]): List of image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[ndarray]: bbox results of each class
+        """
+        raise NotImplementedError
+
+    def aug_test(self, imgs, img_metas, rescale=False):
+        return self.aug_test_simple(imgs, img_metas, rescale)
\ No newline at end of file
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/mask_scoring_rcnn.py b/python/autoCut/models/swin/mmdet/models/detectors/mask_scoring_rcnn.py
new file mode 100644
index 0000000..b6252b6
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/mask_scoring_rcnn.py
@@ -0,0 +1,27 @@
+from ..builder import DETECTORS
+from .two_stage import TwoStageDetector
+
+
+@DETECTORS.register_module()
+class MaskScoringRCNN(TwoStageDetector):
+    """Mask Scoring RCNN.
+
+    https://arxiv.org/abs/1903.00241
+    """
+
+    def __init__(self,
+                 backbone,
+                 rpn_head,
+                 roi_head,
+                 train_cfg,
+                 test_cfg,
+                 neck=None,
+                 pretrained=None):
+        super(MaskScoringRCNN, self).__init__(
+            backbone=backbone,
+            neck=neck,
+            rpn_head=rpn_head,
+            roi_head=roi_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg,
+            pretrained=pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/nasfcos.py b/python/autoCut/models/swin/mmdet/models/detectors/nasfcos.py
new file mode 100644
index 0000000..fb01483
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/nasfcos.py
@@ -0,0 +1,20 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class NASFCOS(SingleStageDetector):
+    """NAS-FCOS: Fast Neural Architecture Search for Object Detection.
+
+    https://arxiv.org/abs/1906.0442
+    """
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(NASFCOS, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                      test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/paa.py b/python/autoCut/models/swin/mmdet/models/detectors/paa.py
new file mode 100644
index 0000000..9b4bb5e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/paa.py
@@ -0,0 +1,17 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class PAA(SingleStageDetector):
+    """Implementation of `PAA <https://arxiv.org/pdf/2007.08103.pdf>`_."""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(PAA, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                  test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/point_rend.py b/python/autoCut/models/swin/mmdet/models/detectors/point_rend.py
new file mode 100644
index 0000000..808ef22
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/point_rend.py
@@ -0,0 +1,29 @@
+from ..builder import DETECTORS
+from .two_stage import TwoStageDetector
+
+
+@DETECTORS.register_module()
+class PointRend(TwoStageDetector):
+    """PointRend: Image Segmentation as Rendering
+
+    This detector is the implementation of
+    `PointRend <https://arxiv.org/abs/1912.08193>`_.
+
+    """
+
+    def __init__(self,
+                 backbone,
+                 rpn_head,
+                 roi_head,
+                 train_cfg,
+                 test_cfg,
+                 neck=None,
+                 pretrained=None):
+        super(PointRend, self).__init__(
+            backbone=backbone,
+            neck=neck,
+            rpn_head=rpn_head,
+            roi_head=roi_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg,
+            pretrained=pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/reppoints_detector.py b/python/autoCut/models/swin/mmdet/models/detectors/reppoints_detector.py
new file mode 100644
index 0000000..a5f6be3
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/reppoints_detector.py
@@ -0,0 +1,22 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class RepPointsDetector(SingleStageDetector):
+    """RepPoints: Point Set Representation for Object Detection.
+
+        This detector is the implementation of:
+        - RepPoints detector (https://arxiv.org/pdf/1904.11490)
+    """
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(RepPointsDetector,
+              self).__init__(backbone, neck, bbox_head, train_cfg, test_cfg,
+                             pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/reppoints_v2_detector.py b/python/autoCut/models/swin/mmdet/models/detectors/reppoints_v2_detector.py
new file mode 100644
index 0000000..605996d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/reppoints_v2_detector.py
@@ -0,0 +1,109 @@
+import numpy as np
+
+import torch
+
+from mmdet.core import bbox2result, bbox_mapping_back, multiclass_nms
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class RepPointsV2Detector(SingleStageDetector):
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(RepPointsV2Detector, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                                test_cfg, pretrained)
+
+    def forward_train(self,
+                      img,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_sem_map=None,
+                      gt_sem_weights=None):
+        x = self.extract_feat(img)
+        outs = self.bbox_head(x)
+        loss_inputs = outs + (gt_bboxes, gt_sem_map, gt_sem_weights, gt_labels, img_metas)
+        losses = self.bbox_head.loss(
+            *loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore)
+        return losses
+
+    def merge_aug_results(self, aug_bboxes, aug_scores, img_metas):
+        """Merge augmented detection bboxes and scores.
+
+        Args:
+            aug_bboxes (list[Tensor]): shape (n, 4*#class)
+            aug_scores (list[Tensor] or None): shape (n, #class)
+            img_shapes (list[Tensor]): shape (3, ).
+
+        Returns:
+            tuple: (bboxes, scores)
+        """
+        recovered_bboxes = []
+        for bboxes, img_info in zip(aug_bboxes, img_metas):
+            img_shape = img_info[0]['img_shape']
+            scale_factor = img_info[0]['scale_factor']
+            flip = img_info[0]['flip']
+            flip_direction = img_info[0]['flip_direction']
+            bboxes = bbox_mapping_back(bboxes, img_shape, scale_factor, flip,
+                                       flip_direction)
+            recovered_bboxes.append(bboxes)
+        bboxes = torch.cat(recovered_bboxes, dim=0)
+        if aug_scores is None:
+            return bboxes
+        else:
+            scores = torch.cat(aug_scores, dim=0)
+            return bboxes, scores
+
+    def aug_test_simple(self, imgs, img_metas, rescale=False):
+        """Test function with test time augmentation
+
+        Args:
+            imgs (list[torch.Tensor]): List of multiple images
+            img_metas (list[dict]): List of image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[ndarray]: bbox results of each class
+        """
+        # recompute feats to save memory
+        feats = self.extract_feats(imgs)
+
+        aug_bboxes = []
+        aug_scores = []
+        for x, img_meta in zip(feats, img_metas):
+            # only one image in the batch
+            outs = self.bbox_head(x)
+            bbox_inputs = outs + (img_metas, self.test_cfg, False, False)
+            det_bboxes, det_scores = self.bbox_head.get_bboxes(*bbox_inputs)[0]
+            aug_bboxes.append(det_bboxes)
+            aug_scores.append(det_scores)
+
+        # after merging, bboxes will be rescaled to the original image size
+        merged_bboxes, merged_scores = self.merge_aug_results(
+            aug_bboxes, aug_scores, img_metas)
+        det_bboxes, det_labels = multiclass_nms(merged_bboxes, merged_scores,
+                                                self.test_cfg.score_thr,
+                                                self.test_cfg.nms,
+                                                self.test_cfg.max_per_img)
+
+        if rescale:
+            _det_bboxes = det_bboxes
+        else:
+            _det_bboxes = det_bboxes.clone()
+            _det_bboxes[:, :4] *= det_bboxes.new_tensor(
+                img_metas[0][0]['scale_factor'])
+        bbox_results = bbox2result(_det_bboxes, det_labels,
+                                   self.bbox_head.num_classes)
+        return bbox_results
+
+    def aug_test(self, imgs, img_metas, rescale=False):
+        return self.aug_test_simple(imgs, img_metas, rescale)
\ No newline at end of file
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/retinanet.py b/python/autoCut/models/swin/mmdet/models/detectors/retinanet.py
new file mode 100644
index 0000000..41378e8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/retinanet.py
@@ -0,0 +1,17 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class RetinaNet(SingleStageDetector):
+    """Implementation of `RetinaNet <https://arxiv.org/abs/1708.02002>`_"""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(RetinaNet, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                        test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/rpn.py b/python/autoCut/models/swin/mmdet/models/detectors/rpn.py
new file mode 100644
index 0000000..1a77294
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/rpn.py
@@ -0,0 +1,154 @@
+import mmcv
+from mmcv.image import tensor2imgs
+
+from mmdet.core import bbox_mapping
+from ..builder import DETECTORS, build_backbone, build_head, build_neck
+from .base import BaseDetector
+
+
+@DETECTORS.register_module()
+class RPN(BaseDetector):
+    """Implementation of Region Proposal Network."""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 rpn_head,
+                 train_cfg,
+                 test_cfg,
+                 pretrained=None):
+        super(RPN, self).__init__()
+        self.backbone = build_backbone(backbone)
+        self.neck = build_neck(neck) if neck is not None else None
+        rpn_train_cfg = train_cfg.rpn if train_cfg is not None else None
+        rpn_head.update(train_cfg=rpn_train_cfg)
+        rpn_head.update(test_cfg=test_cfg.rpn)
+        self.rpn_head = build_head(rpn_head)
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+        self.init_weights(pretrained=pretrained)
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in detector.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        super(RPN, self).init_weights(pretrained)
+        self.backbone.init_weights(pretrained=pretrained)
+        if self.with_neck:
+            self.neck.init_weights()
+        self.rpn_head.init_weights()
+
+    def extract_feat(self, img):
+        """Extract features.
+
+        Args:
+            img (torch.Tensor): Image tensor with shape (n, c, h ,w).
+
+        Returns:
+            list[torch.Tensor]: Multi-level features that may have
+                different resolutions.
+        """
+        x = self.backbone(img)
+        if self.with_neck:
+            x = self.neck(x)
+        return x
+
+    def forward_dummy(self, img):
+        """Dummy forward function."""
+        x = self.extract_feat(img)
+        rpn_outs = self.rpn_head(x)
+        return rpn_outs
+
+    def forward_train(self,
+                      img,
+                      img_metas,
+                      gt_bboxes=None,
+                      gt_bboxes_ignore=None):
+        """
+        Args:
+            img (Tensor): Input images of shape (N, C, H, W).
+                Typically these should be mean centered and std scaled.
+            img_metas (list[dict]): A List of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                :class:`mmdet.datasets.pipelines.Collect`.
+            gt_bboxes (list[Tensor]): Each item are the truth boxes for each
+                image in [tl_x, tl_y, br_x, br_y] format.
+            gt_bboxes_ignore (None | list[Tensor]): Specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        if (isinstance(self.train_cfg.rpn, dict)
+                and self.train_cfg.rpn.get('debug', False)):
+            self.rpn_head.debug_imgs = tensor2imgs(img)
+
+        x = self.extract_feat(img)
+        losses = self.rpn_head.forward_train(x, img_metas, gt_bboxes, None,
+                                             gt_bboxes_ignore)
+        return losses
+
+    def simple_test(self, img, img_metas, rescale=False):
+        """Test function without test time augmentation.
+
+        Args:
+            imgs (list[torch.Tensor]): List of multiple images
+            img_metas (list[dict]): List of image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[np.ndarray]: proposals
+        """
+        x = self.extract_feat(img)
+        proposal_list = self.rpn_head.simple_test_rpn(x, img_metas)
+        if rescale:
+            for proposals, meta in zip(proposal_list, img_metas):
+                proposals[:, :4] /= proposals.new_tensor(meta['scale_factor'])
+
+        return [proposal.cpu().numpy() for proposal in proposal_list]
+
+    def aug_test(self, imgs, img_metas, rescale=False):
+        """Test function with test time augmentation.
+
+        Args:
+            imgs (list[torch.Tensor]): List of multiple images
+            img_metas (list[dict]): List of image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[np.ndarray]: proposals
+        """
+        proposal_list = self.rpn_head.aug_test_rpn(
+            self.extract_feats(imgs), img_metas)
+        if not rescale:
+            for proposals, img_meta in zip(proposal_list, img_metas[0]):
+                img_shape = img_meta['img_shape']
+                scale_factor = img_meta['scale_factor']
+                flip = img_meta['flip']
+                flip_direction = img_meta['flip_direction']
+                proposals[:, :4] = bbox_mapping(proposals[:, :4], img_shape,
+                                                scale_factor, flip,
+                                                flip_direction)
+        return [proposal.cpu().numpy() for proposal in proposal_list]
+
+    def show_result(self, data, result, top_k=20, **kwargs):
+        """Show RPN proposals on the image.
+
+        Args:
+            data (str or np.ndarray): Image filename or loaded image.
+            result (Tensor or tuple): The results to draw over `img`
+                bbox_result or (bbox_result, segm_result).
+            top_k (int): Plot the first k bboxes only
+               if set positive. Default: 20
+
+        Returns:
+            np.ndarray: The image with bboxes drawn on it.
+        """
+        mmcv.imshow_bboxes(data, result, top_k=top_k)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/scnet.py b/python/autoCut/models/swin/mmdet/models/detectors/scnet.py
new file mode 100644
index 0000000..04a2347
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/scnet.py
@@ -0,0 +1,10 @@
+from ..builder import DETECTORS
+from .cascade_rcnn import CascadeRCNN
+
+
+@DETECTORS.register_module()
+class SCNet(CascadeRCNN):
+    """Implementation of `SCNet <https://arxiv.org/abs/2012.10150>`_"""
+
+    def __init__(self, **kwargs):
+        super(SCNet, self).__init__(**kwargs)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/single_stage.py b/python/autoCut/models/swin/mmdet/models/detectors/single_stage.py
new file mode 100644
index 0000000..5172bdb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/single_stage.py
@@ -0,0 +1,154 @@
+import torch
+import torch.nn as nn
+
+from mmdet.core import bbox2result
+from ..builder import DETECTORS, build_backbone, build_head, build_neck
+from .base import BaseDetector
+
+
+@DETECTORS.register_module()
+class SingleStageDetector(BaseDetector):
+    """Base class for single-stage detectors.
+
+    Single-stage detectors directly and densely predict bounding boxes on the
+    output features of the backbone+neck.
+    """
+
+    def __init__(self,
+                 backbone,
+                 neck=None,
+                 bbox_head=None,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(SingleStageDetector, self).__init__()
+        self.backbone = build_backbone(backbone)
+        if neck is not None:
+            self.neck = build_neck(neck)
+        bbox_head.update(train_cfg=train_cfg)
+        bbox_head.update(test_cfg=test_cfg)
+        self.bbox_head = build_head(bbox_head)
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+        self.init_weights(pretrained=pretrained)
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in detector.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        super(SingleStageDetector, self).init_weights(pretrained)
+        self.backbone.init_weights(pretrained=pretrained)
+        if self.with_neck:
+            if isinstance(self.neck, nn.Sequential):
+                for m in self.neck:
+                    m.init_weights()
+            else:
+                self.neck.init_weights()
+        self.bbox_head.init_weights()
+
+    def extract_feat(self, img):
+        """Directly extract features from the backbone+neck."""
+        x = self.backbone(img)
+        if self.with_neck:
+            x = self.neck(x)
+        return x
+
+    def forward_dummy(self, img):
+        """Used for computing network flops.
+
+        See `mmdetection/tools/analysis_tools/get_flops.py`
+        """
+        x = self.extract_feat(img)
+        outs = self.bbox_head(x)
+        return outs
+
+    def forward_train(self,
+                      img,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None):
+        """
+        Args:
+            img (Tensor): Input images of shape (N, C, H, W).
+                Typically these should be mean centered and std scaled.
+            img_metas (list[dict]): A List of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                :class:`mmdet.datasets.pipelines.Collect`.
+            gt_bboxes (list[Tensor]): Each item are the truth boxes for each
+                image in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): Class indices corresponding to each box
+            gt_bboxes_ignore (None | list[Tensor]): Specify which bounding
+                boxes can be ignored when computing the loss.
+
+        Returns:
+            dict[str, Tensor]: A dictionary of loss components.
+        """
+        super(SingleStageDetector, self).forward_train(img, img_metas)
+        x = self.extract_feat(img)
+        losses = self.bbox_head.forward_train(x, img_metas, gt_bboxes,
+                                              gt_labels, gt_bboxes_ignore)
+        return losses
+
+    def simple_test(self, img, img_metas, rescale=False):
+        """Test function without test time augmentation.
+
+        Args:
+            imgs (list[torch.Tensor]): List of multiple images
+            img_metas (list[dict]): List of image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[list[np.ndarray]]: BBox results of each image and classes.
+                The outer list corresponds to each image. The inner list
+                corresponds to each class.
+        """
+        x = self.extract_feat(img)
+        outs = self.bbox_head(x)
+        # get origin input shape to support onnx dynamic shape
+        if torch.onnx.is_in_onnx_export():
+            # get shape as tensor
+            img_shape = torch._shape_as_tensor(img)[2:]
+            img_metas[0]['img_shape_for_onnx'] = img_shape
+        bbox_list = self.bbox_head.get_bboxes(
+            *outs, img_metas, rescale=rescale)
+        # skip post-processing when exporting to ONNX
+        if torch.onnx.is_in_onnx_export():
+            return bbox_list
+
+        bbox_results = [
+            bbox2result(det_bboxes, det_labels, self.bbox_head.num_classes)
+            for det_bboxes, det_labels in bbox_list
+        ]
+        return bbox_results
+
+    def aug_test(self, imgs, img_metas, rescale=False):
+        """Test function with test time augmentation.
+
+        Args:
+            imgs (list[Tensor]): the outer list indicates test-time
+                augmentations and inner Tensor should have a shape NxCxHxW,
+                which contains all images in the batch.
+            img_metas (list[list[dict]]): the outer list indicates test-time
+                augs (multiscale, flip, etc.) and the inner list indicates
+                images in a batch. each dict has image information.
+            rescale (bool, optional): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[list[np.ndarray]]: BBox results of each image and classes.
+                The outer list corresponds to each image. The inner list
+                corresponds to each class.
+        """
+        assert hasattr(self.bbox_head, 'aug_test'), \
+            f'{self.bbox_head.__class__.__name__}' \
+            ' does not support test-time augmentation'
+
+        feats = self.extract_feats(imgs)
+        return [self.bbox_head.aug_test(feats, img_metas, rescale=rescale)]
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/sparse_rcnn.py b/python/autoCut/models/swin/mmdet/models/detectors/sparse_rcnn.py
new file mode 100644
index 0000000..0dbd025
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/sparse_rcnn.py
@@ -0,0 +1,110 @@
+from ..builder import DETECTORS
+from .two_stage import TwoStageDetector
+
+
+@DETECTORS.register_module()
+class SparseRCNN(TwoStageDetector):
+    r"""Implementation of `Sparse R-CNN: End-to-End Object Detection with
+    Learnable Proposals <https://arxiv.org/abs/2011.12450>`_"""
+
+    def __init__(self, *args, **kwargs):
+        super(SparseRCNN, self).__init__(*args, **kwargs)
+        assert self.with_rpn, 'Sparse R-CNN do not support external proposals'
+
+    def forward_train(self,
+                      img,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None,
+                      proposals=None,
+                      **kwargs):
+        """Forward function of SparseR-CNN in train stage.
+
+        Args:
+            img (Tensor): of shape (N, C, H, W) encoding input images.
+                Typically these should be mean centered and std scaled.
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                :class:`mmdet.datasets.pipelines.Collect`.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            gt_bboxes_ignore (None | list[Tensor): specify which bounding
+                boxes can be ignored when computing the loss.
+            gt_masks (List[Tensor], optional) : Segmentation masks for
+                each box. But we don't support it in this architecture.
+            proposals (List[Tensor], optional): override rpn proposals with
+                custom proposals. Use when `with_rpn` is False.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+
+        assert proposals is None, 'Sparse R-CNN does not support' \
+                                  ' external proposals'
+        assert gt_masks is None, 'Sparse R-CNN does not instance segmentation'
+
+        x = self.extract_feat(img)
+        proposal_boxes, proposal_features, imgs_whwh = \
+            self.rpn_head.forward_train(x, img_metas)
+        roi_losses = self.roi_head.forward_train(
+            x,
+            proposal_boxes,
+            proposal_features,
+            img_metas,
+            gt_bboxes,
+            gt_labels,
+            gt_bboxes_ignore=gt_bboxes_ignore,
+            gt_masks=gt_masks,
+            imgs_whwh=imgs_whwh)
+        return roi_losses
+
+    def simple_test(self, img, img_metas, rescale=False):
+        """Test function without test time augmentation.
+
+        Args:
+            imgs (list[torch.Tensor]): List of multiple images
+            img_metas (list[dict]): List of image information.
+            rescale (bool): Whether to rescale the results.
+                Defaults to False.
+
+        Returns:
+            list[list[np.ndarray]]: BBox results of each image and classes.
+                The outer list corresponds to each image. The inner list
+                corresponds to each class.
+        """
+        x = self.extract_feat(img)
+        proposal_boxes, proposal_features, imgs_whwh = \
+            self.rpn_head.simple_test_rpn(x, img_metas)
+        bbox_results = self.roi_head.simple_test(
+            x,
+            proposal_boxes,
+            proposal_features,
+            img_metas,
+            imgs_whwh=imgs_whwh,
+            rescale=rescale)
+        return bbox_results
+
+    def forward_dummy(self, img):
+        """Used for computing network flops.
+
+        See `mmdetection/tools/analysis_tools/get_flops.py`
+        """
+        # backbone
+        x = self.extract_feat(img)
+        # rpn
+        num_imgs = len(img)
+        dummy_img_metas = [
+            dict(img_shape=(800, 1333, 3)) for _ in range(num_imgs)
+        ]
+        proposal_boxes, proposal_features, imgs_whwh = \
+            self.rpn_head.simple_test_rpn(x, dummy_img_metas)
+        # roi_head
+        roi_outs = self.roi_head.forward_dummy(x, proposal_boxes,
+                                               proposal_features,
+                                               dummy_img_metas)
+        return roi_outs
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/trident_faster_rcnn.py b/python/autoCut/models/swin/mmdet/models/detectors/trident_faster_rcnn.py
new file mode 100644
index 0000000..f0fd80d
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/trident_faster_rcnn.py
@@ -0,0 +1,66 @@
+from ..builder import DETECTORS
+from .faster_rcnn import FasterRCNN
+
+
+@DETECTORS.register_module()
+class TridentFasterRCNN(FasterRCNN):
+    """Implementation of `TridentNet <https://arxiv.org/abs/1901.01892>`_"""
+
+    def __init__(self,
+                 backbone,
+                 rpn_head,
+                 roi_head,
+                 train_cfg,
+                 test_cfg,
+                 neck=None,
+                 pretrained=None):
+
+        super(TridentFasterRCNN, self).__init__(
+            backbone=backbone,
+            neck=neck,
+            rpn_head=rpn_head,
+            roi_head=roi_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg,
+            pretrained=pretrained)
+        assert self.backbone.num_branch == self.roi_head.num_branch
+        assert self.backbone.test_branch_idx == self.roi_head.test_branch_idx
+        self.num_branch = self.backbone.num_branch
+        self.test_branch_idx = self.backbone.test_branch_idx
+
+    def simple_test(self, img, img_metas, proposals=None, rescale=False):
+        """Test without augmentation."""
+        assert self.with_bbox, 'Bbox head must be implemented.'
+        x = self.extract_feat(img)
+        if proposals is None:
+            num_branch = (self.num_branch if self.test_branch_idx == -1 else 1)
+            trident_img_metas = img_metas * num_branch
+            proposal_list = self.rpn_head.simple_test_rpn(x, trident_img_metas)
+        else:
+            proposal_list = proposals
+
+        return self.roi_head.simple_test(
+            x, proposal_list, trident_img_metas, rescale=rescale)
+
+    def aug_test(self, imgs, img_metas, rescale=False):
+        """Test with augmentations.
+
+        If rescale is False, then returned bboxes and masks will fit the scale
+        of imgs[0].
+        """
+        x = self.extract_feats(imgs)
+        num_branch = (self.num_branch if self.test_branch_idx == -1 else 1)
+        trident_img_metas = [img_metas * num_branch for img_metas in img_metas]
+        proposal_list = self.rpn_head.aug_test_rpn(x, trident_img_metas)
+        return self.roi_head.aug_test(
+            x, proposal_list, img_metas, rescale=rescale)
+
+    def forward_train(self, img, img_metas, gt_bboxes, gt_labels, **kwargs):
+        """make copies of img and gts to fit multi-branch."""
+        trident_gt_bboxes = tuple(gt_bboxes * self.num_branch)
+        trident_gt_labels = tuple(gt_labels * self.num_branch)
+        trident_img_metas = tuple(img_metas * self.num_branch)
+
+        return super(TridentFasterRCNN,
+                     self).forward_train(img, trident_img_metas,
+                                         trident_gt_bboxes, trident_gt_labels)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/two_stage.py b/python/autoCut/models/swin/mmdet/models/detectors/two_stage.py
new file mode 100644
index 0000000..ba5bdde
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/two_stage.py
@@ -0,0 +1,215 @@
+import torch
+import torch.nn as nn
+
+# from mmdet.core import bbox2result, bbox2roi, build_assigner, build_sampler
+from ..builder import DETECTORS, build_backbone, build_head, build_neck
+from .base import BaseDetector
+
+
+@DETECTORS.register_module()
+class TwoStageDetector(BaseDetector):
+    """Base class for two-stage detectors.
+
+    Two-stage detectors typically consisting of a region proposal network and a
+    task-specific regression head.
+    """
+
+    def __init__(self,
+                 backbone,
+                 neck=None,
+                 rpn_head=None,
+                 roi_head=None,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(TwoStageDetector, self).__init__()
+        self.backbone = build_backbone(backbone)
+
+        if neck is not None:
+            self.neck = build_neck(neck)
+
+        if rpn_head is not None:
+            rpn_train_cfg = train_cfg.rpn if train_cfg is not None else None
+            rpn_head_ = rpn_head.copy()
+            rpn_head_.update(train_cfg=rpn_train_cfg, test_cfg=test_cfg.rpn)
+            self.rpn_head = build_head(rpn_head_)
+
+        if roi_head is not None:
+            # update train and test cfg here for now
+            # TODO: refactor assigner & sampler
+            rcnn_train_cfg = train_cfg.rcnn if train_cfg is not None else None
+            roi_head.update(train_cfg=rcnn_train_cfg)
+            roi_head.update(test_cfg=test_cfg.rcnn)
+            self.roi_head = build_head(roi_head)
+
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+
+        self.init_weights(pretrained=pretrained)
+
+    @property
+    def with_rpn(self):
+        """bool: whether the detector has RPN"""
+        return hasattr(self, 'rpn_head') and self.rpn_head is not None
+
+    @property
+    def with_roi_head(self):
+        """bool: whether the detector has a RoI head"""
+        return hasattr(self, 'roi_head') and self.roi_head is not None
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in detector.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        super(TwoStageDetector, self).init_weights(pretrained)
+        self.backbone.init_weights(pretrained=pretrained)
+        if self.with_neck:
+            if isinstance(self.neck, nn.Sequential):
+                for m in self.neck:
+                    m.init_weights()
+            else:
+                self.neck.init_weights()
+        if self.with_rpn:
+            self.rpn_head.init_weights()
+        if self.with_roi_head:
+            self.roi_head.init_weights(pretrained)
+
+    def extract_feat(self, img):
+        """Directly extract features from the backbone+neck."""
+        x = self.backbone(img)
+        if self.with_neck:
+            x = self.neck(x)
+        return x
+
+    def forward_dummy(self, img):
+        """Used for computing network flops.
+
+        See `mmdetection/tools/analysis_tools/get_flops.py`
+        """
+        outs = ()
+        # backbone
+        x = self.extract_feat(img)
+        # rpn
+        if self.with_rpn:
+            rpn_outs = self.rpn_head(x)
+            outs = outs + (rpn_outs, )
+        proposals = torch.randn(1000, 4).to(img.device)
+        # roi_head
+        roi_outs = self.roi_head.forward_dummy(x, proposals)
+        outs = outs + (roi_outs, )
+        return outs
+
+    def forward_train(self,
+                      img,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None,
+                      proposals=None,
+                      **kwargs):
+        """
+        Args:
+            img (Tensor): of shape (N, C, H, W) encoding input images.
+                Typically these should be mean centered and std scaled.
+
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+
+            gt_labels (list[Tensor]): class indices corresponding to each box
+
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+            gt_masks (None | Tensor) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+            proposals : override rpn proposals with custom proposals. Use when
+                `with_rpn` is False.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        x = self.extract_feat(img)
+
+        losses = dict()
+
+        # RPN forward and loss
+        if self.with_rpn:
+            proposal_cfg = self.train_cfg.get('rpn_proposal',
+                                              self.test_cfg.rpn)
+            rpn_losses, proposal_list = self.rpn_head.forward_train(
+                x,
+                img_metas,
+                gt_bboxes,
+                gt_labels=None,
+                gt_bboxes_ignore=gt_bboxes_ignore,
+                proposal_cfg=proposal_cfg)
+            losses.update(rpn_losses)
+        else:
+            proposal_list = proposals
+
+        roi_losses = self.roi_head.forward_train(x, img_metas, proposal_list,
+                                                 gt_bboxes, gt_labels,
+                                                 gt_bboxes_ignore, gt_masks,
+                                                 **kwargs)
+        losses.update(roi_losses)
+
+        return losses
+
+    async def async_simple_test(self,
+                                img,
+                                img_meta,
+                                proposals=None,
+                                rescale=False):
+        """Async test without augmentation."""
+        assert self.with_bbox, 'Bbox head must be implemented.'
+        x = self.extract_feat(img)
+
+        if proposals is None:
+            proposal_list = await self.rpn_head.async_simple_test_rpn(
+                x, img_meta)
+        else:
+            proposal_list = proposals
+
+        return await self.roi_head.async_simple_test(
+            x, proposal_list, img_meta, rescale=rescale)
+
+    def simple_test(self, img, img_metas, proposals=None, rescale=False):
+        """Test without augmentation."""
+        assert self.with_bbox, 'Bbox head must be implemented.'
+
+        x = self.extract_feat(img)
+
+        # get origin input shape to onnx dynamic input shape
+        if torch.onnx.is_in_onnx_export():
+            img_shape = torch._shape_as_tensor(img)[2:]
+            img_metas[0]['img_shape_for_onnx'] = img_shape
+
+        if proposals is None:
+            proposal_list = self.rpn_head.simple_test_rpn(x, img_metas)
+        else:
+            proposal_list = proposals
+
+        return self.roi_head.simple_test(
+            x, proposal_list, img_metas, rescale=rescale)
+
+    def aug_test(self, imgs, img_metas, rescale=False):
+        """Test with augmentations.
+
+        If rescale is False, then returned bboxes and masks will fit the scale
+        of imgs[0].
+        """
+        x = self.extract_feats(imgs)
+        proposal_list = self.rpn_head.aug_test_rpn(x, img_metas)
+        return self.roi_head.aug_test(
+            x, proposal_list, img_metas, rescale=rescale)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/vfnet.py b/python/autoCut/models/swin/mmdet/models/detectors/vfnet.py
new file mode 100644
index 0000000..e23f896
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/vfnet.py
@@ -0,0 +1,18 @@
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class VFNet(SingleStageDetector):
+    """Implementation of `VarifocalNet
+    (VFNet).<https://arxiv.org/abs/2008.13367>`_"""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(VFNet, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                    test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/yolact.py b/python/autoCut/models/swin/mmdet/models/detectors/yolact.py
new file mode 100644
index 0000000..f32fde0
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/yolact.py
@@ -0,0 +1,146 @@
+import torch
+
+from mmdet.core import bbox2result
+from ..builder import DETECTORS, build_head
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class YOLACT(SingleStageDetector):
+    """Implementation of `YOLACT <https://arxiv.org/abs/1904.02689>`_"""
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 segm_head,
+                 mask_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(YOLACT, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                     test_cfg, pretrained)
+        self.segm_head = build_head(segm_head)
+        self.mask_head = build_head(mask_head)
+        self.init_segm_mask_weights()
+
+    def init_segm_mask_weights(self):
+        """Initialize weights of the YOLACT segm head and YOLACT mask head."""
+        self.segm_head.init_weights()
+        self.mask_head.init_weights()
+
+    def forward_dummy(self, img):
+        """Used for computing network flops.
+
+        See `mmdetection/tools/analysis_tools/get_flops.py`
+        """
+        raise NotImplementedError
+
+    def forward_train(self,
+                      img,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None):
+        """
+        Args:
+            img (Tensor): of shape (N, C, H, W) encoding input images.
+                Typically these should be mean centered and std scaled.
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+            gt_masks (None | Tensor) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        # convert Bitmap mask or Polygon Mask to Tensor here
+        gt_masks = [
+            gt_mask.to_tensor(dtype=torch.uint8, device=img.device)
+            for gt_mask in gt_masks
+        ]
+
+        x = self.extract_feat(img)
+
+        cls_score, bbox_pred, coeff_pred = self.bbox_head(x)
+        bbox_head_loss_inputs = (cls_score, bbox_pred) + (gt_bboxes, gt_labels,
+                                                          img_metas)
+        losses, sampling_results = self.bbox_head.loss(
+            *bbox_head_loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore)
+
+        segm_head_outs = self.segm_head(x[0])
+        loss_segm = self.segm_head.loss(segm_head_outs, gt_masks, gt_labels)
+        losses.update(loss_segm)
+
+        mask_pred = self.mask_head(x[0], coeff_pred, gt_bboxes, img_metas,
+                                   sampling_results)
+        loss_mask = self.mask_head.loss(mask_pred, gt_masks, gt_bboxes,
+                                        img_metas, sampling_results)
+        losses.update(loss_mask)
+
+        # check NaN and Inf
+        for loss_name in losses.keys():
+            assert torch.isfinite(torch.stack(losses[loss_name]))\
+                .all().item(), '{} becomes infinite or NaN!'\
+                .format(loss_name)
+
+        return losses
+
+    def simple_test(self, img, img_metas, rescale=False):
+        """Test function without test time augmentation."""
+        x = self.extract_feat(img)
+
+        cls_score, bbox_pred, coeff_pred = self.bbox_head(x)
+
+        bbox_inputs = (cls_score, bbox_pred,
+                       coeff_pred) + (img_metas, self.test_cfg, rescale)
+        det_bboxes, det_labels, det_coeffs = self.bbox_head.get_bboxes(
+            *bbox_inputs)
+        bbox_results = [
+            bbox2result(det_bbox, det_label, self.bbox_head.num_classes)
+            for det_bbox, det_label in zip(det_bboxes, det_labels)
+        ]
+
+        num_imgs = len(img_metas)
+        scale_factors = tuple(meta['scale_factor'] for meta in img_metas)
+        if all(det_bbox.shape[0] == 0 for det_bbox in det_bboxes):
+            segm_results = [[[] for _ in range(self.mask_head.num_classes)]
+                            for _ in range(num_imgs)]
+        else:
+            # if det_bboxes is rescaled to the original image size, we need to
+            # rescale it back to the testing scale to obtain RoIs.
+            if rescale and not isinstance(scale_factors[0], float):
+                scale_factors = [
+                    torch.from_numpy(scale_factor).to(det_bboxes[0].device)
+                    for scale_factor in scale_factors
+                ]
+            _bboxes = [
+                det_bboxes[i][:, :4] *
+                scale_factors[i] if rescale else det_bboxes[i][:, :4]
+                for i in range(len(det_bboxes))
+            ]
+            mask_preds = self.mask_head(x[0], det_coeffs, _bboxes, img_metas)
+            # apply mask post-processing to each image individually
+            segm_results = []
+            for i in range(num_imgs):
+                if det_bboxes[i].shape[0] == 0:
+                    segm_results.append(
+                        [[] for _ in range(self.mask_head.num_classes)])
+                else:
+                    segm_result = self.mask_head.get_seg_masks(
+                        mask_preds[i], det_labels[i], img_metas[i], rescale)
+                    segm_results.append(segm_result)
+        return list(zip(bbox_results, segm_results))
+
+    def aug_test(self, imgs, img_metas, rescale=False):
+        """Test with augmentations."""
+        raise NotImplementedError
diff --git a/python/autoCut/models/swin/mmdet/models/detectors/yolo.py b/python/autoCut/models/swin/mmdet/models/detectors/yolo.py
new file mode 100644
index 0000000..240aab2
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/detectors/yolo.py
@@ -0,0 +1,18 @@
+# Copyright (c) 2019 Western Digital Corporation or its affiliates.
+
+from ..builder import DETECTORS
+from .single_stage import SingleStageDetector
+
+
+@DETECTORS.register_module()
+class YOLOV3(SingleStageDetector):
+
+    def __init__(self,
+                 backbone,
+                 neck,
+                 bbox_head,
+                 train_cfg=None,
+                 test_cfg=None,
+                 pretrained=None):
+        super(YOLOV3, self).__init__(backbone, neck, bbox_head, train_cfg,
+                                     test_cfg, pretrained)
diff --git a/python/autoCut/models/swin/mmdet/models/losses/__init__.py b/python/autoCut/models/swin/mmdet/models/losses/__init__.py
new file mode 100644
index 0000000..6c80c69
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/__init__.py
@@ -0,0 +1,29 @@
+from .accuracy import Accuracy, accuracy
+from .ae_loss import AssociativeEmbeddingLoss
+from .balanced_l1_loss import BalancedL1Loss, balanced_l1_loss
+from .cross_entropy_loss import (CrossEntropyLoss, binary_cross_entropy,
+                                 cross_entropy, mask_cross_entropy)
+from .focal_loss import FocalLoss, sigmoid_focal_loss, SEPFocalLoss
+from .gaussian_focal_loss import GaussianFocalLoss
+from .gfocal_loss import DistributionFocalLoss, QualityFocalLoss, RPDQualityFocalLoss
+from .ghm_loss import GHMC, GHMR
+from .iou_loss import (BoundedIoULoss, CIoULoss, DIoULoss, GIoULoss, IoULoss,
+                       bounded_iou_loss, iou_loss)
+from .kd_loss import KnowledgeDistillationKLDivLoss
+from .mse_loss import MSELoss, mse_loss
+from .pisa_loss import carl_loss, isr_p
+from .smooth_l1_loss import L1Loss, SmoothL1Loss, l1_loss, smooth_l1_loss
+from .utils import reduce_loss, weight_reduce_loss, weighted_loss
+from .varifocal_loss import VarifocalLoss
+
+__all__ = [
+    'accuracy', 'Accuracy', 'cross_entropy', 'binary_cross_entropy',
+    'mask_cross_entropy', 'CrossEntropyLoss', 'sigmoid_focal_loss', 'SEPFocalLoss',
+    'FocalLoss', 'smooth_l1_loss', 'SmoothL1Loss', 'balanced_l1_loss',
+    'BalancedL1Loss', 'mse_loss', 'MSELoss', 'iou_loss', 'bounded_iou_loss',
+    'IoULoss', 'BoundedIoULoss', 'GIoULoss', 'DIoULoss', 'CIoULoss', 'GHMC',
+    'GHMR', 'reduce_loss', 'weight_reduce_loss', 'weighted_loss', 'L1Loss',
+    'l1_loss', 'isr_p', 'carl_loss', 'AssociativeEmbeddingLoss',
+    'GaussianFocalLoss', 'QualityFocalLoss', 'DistributionFocalLoss', 'RPDQualityFocalLoss',
+    'VarifocalLoss', 'KnowledgeDistillationKLDivLoss'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/losses/accuracy.py b/python/autoCut/models/swin/mmdet/models/losses/accuracy.py
new file mode 100644
index 0000000..789a224
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/accuracy.py
@@ -0,0 +1,78 @@
+import mmcv
+import torch.nn as nn
+
+
+@mmcv.jit(coderize=True)
+def accuracy(pred, target, topk=1, thresh=None):
+    """Calculate accuracy according to the prediction and target.
+
+    Args:
+        pred (torch.Tensor): The model prediction, shape (N, num_class)
+        target (torch.Tensor): The target of each prediction, shape (N, )
+        topk (int | tuple[int], optional): If the predictions in ``topk``
+            matches the target, the predictions will be regarded as
+            correct ones. Defaults to 1.
+        thresh (float, optional): If not None, predictions with scores under
+            this threshold are considered incorrect. Default to None.
+
+    Returns:
+        float | tuple[float]: If the input ``topk`` is a single integer,
+            the function will return a single float as accuracy. If
+            ``topk`` is a tuple containing multiple integers, the
+            function will return a tuple containing accuracies of
+            each ``topk`` number.
+    """
+    assert isinstance(topk, (int, tuple))
+    if isinstance(topk, int):
+        topk = (topk, )
+        return_single = True
+    else:
+        return_single = False
+
+    maxk = max(topk)
+    if pred.size(0) == 0:
+        accu = [pred.new_tensor(0.) for i in range(len(topk))]
+        return accu[0] if return_single else accu
+    assert pred.ndim == 2 and target.ndim == 1
+    assert pred.size(0) == target.size(0)
+    assert maxk <= pred.size(1), \
+        f'maxk {maxk} exceeds pred dimension {pred.size(1)}'
+    pred_value, pred_label = pred.topk(maxk, dim=1)
+    pred_label = pred_label.t()  # transpose to shape (maxk, N)
+    correct = pred_label.eq(target.view(1, -1).expand_as(pred_label))
+    if thresh is not None:
+        # Only prediction values larger than thresh are counted as correct
+        correct = correct & (pred_value > thresh).t()
+    res = []
+    for k in topk:
+        correct_k = correct[:k].reshape(-1).float().sum(0, keepdim=True)
+        res.append(correct_k.mul_(100.0 / pred.size(0)))
+    return res[0] if return_single else res
+
+
+class Accuracy(nn.Module):
+
+    def __init__(self, topk=(1, ), thresh=None):
+        """Module to calculate the accuracy.
+
+        Args:
+            topk (tuple, optional): The criterion used to calculate the
+                accuracy. Defaults to (1,).
+            thresh (float, optional): If not None, predictions with scores
+                under this threshold are considered incorrect. Default to None.
+        """
+        super().__init__()
+        self.topk = topk
+        self.thresh = thresh
+
+    def forward(self, pred, target):
+        """Forward function to calculate accuracy.
+
+        Args:
+            pred (torch.Tensor): Prediction of models.
+            target (torch.Tensor): Target for each prediction.
+
+        Returns:
+            tuple[float]: The accuracies under different topk criterions.
+        """
+        return accuracy(pred, target, self.topk, self.thresh)
diff --git a/python/autoCut/models/swin/mmdet/models/losses/ae_loss.py b/python/autoCut/models/swin/mmdet/models/losses/ae_loss.py
new file mode 100644
index 0000000..cff472a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/ae_loss.py
@@ -0,0 +1,102 @@
+import mmcv
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from ..builder import LOSSES
+
+
+@mmcv.jit(derivate=True, coderize=True)
+def ae_loss_per_image(tl_preds, br_preds, match):
+    """Associative Embedding Loss in one image.
+
+    Associative Embedding Loss including two parts: pull loss and push loss.
+    Pull loss makes embedding vectors from same object closer to each other.
+    Push loss distinguish embedding vector from different objects, and makes
+        the gap between them is large enough.
+
+    During computing, usually there are 3 cases:
+        - no object in image: both pull loss and push loss will be 0.
+        - one object in image: push loss will be 0 and pull loss is computed
+            by the two corner of the only object.
+        - more than one objects in image: pull loss is computed by corner pairs
+            from each object, push loss is computed by each object with all
+            other objects. We use confusion matrix with 0 in diagonal to
+            compute the push loss.
+
+    Args:
+        tl_preds (tensor): Embedding feature map of left-top corner.
+        br_preds (tensor): Embedding feature map of bottim-right corner.
+        match (list): Downsampled coordinates pair of each ground truth box.
+    """
+
+    tl_list, br_list, me_list = [], [], []
+    if len(match) == 0:  # no object in image
+        pull_loss = tl_preds.sum() * 0.
+        push_loss = tl_preds.sum() * 0.
+    else:
+        for m in match:
+            [tl_y, tl_x], [br_y, br_x] = m
+            tl_e = tl_preds[:, tl_y, tl_x].view(-1, 1)
+            br_e = br_preds[:, br_y, br_x].view(-1, 1)
+            tl_list.append(tl_e)
+            br_list.append(br_e)
+            me_list.append((tl_e + br_e) / 2.0)
+
+        tl_list = torch.cat(tl_list)
+        br_list = torch.cat(br_list)
+        me_list = torch.cat(me_list)
+
+        assert tl_list.size() == br_list.size()
+
+        # N is object number in image, M is dimension of embedding vector
+        N, M = tl_list.size()
+
+        pull_loss = (tl_list - me_list).pow(2) + (br_list - me_list).pow(2)
+        pull_loss = pull_loss.sum() / N
+
+        margin = 1  # exp setting of CornerNet, details in section 3.3 of paper
+
+        # confusion matrix of push loss
+        conf_mat = me_list.expand((N, N, M)).permute(1, 0, 2) - me_list
+        conf_weight = 1 - torch.eye(N).type_as(me_list)
+        conf_mat = conf_weight * (margin - conf_mat.sum(-1).abs())
+
+        if N > 1:  # more than one object in current image
+            push_loss = F.relu(conf_mat).sum() / (N * (N - 1))
+        else:
+            push_loss = tl_preds.sum() * 0.
+
+    return pull_loss, push_loss
+
+
+@LOSSES.register_module()
+class AssociativeEmbeddingLoss(nn.Module):
+    """Associative Embedding Loss.
+
+    More details can be found in
+    `Associative Embedding <https://arxiv.org/abs/1611.05424>`_ and
+    `CornerNet <https://arxiv.org/abs/1808.01244>`_ .
+    Code is modified from `kp_utils.py <https://github.com/princeton-vl/CornerNet/blob/master/models/py_utils/kp_utils.py#L180>`_  # noqa: E501
+
+    Args:
+        pull_weight (float): Loss weight for corners from same object.
+        push_weight (float): Loss weight for corners from different object.
+    """
+
+    def __init__(self, pull_weight=0.25, push_weight=0.25):
+        super(AssociativeEmbeddingLoss, self).__init__()
+        self.pull_weight = pull_weight
+        self.push_weight = push_weight
+
+    def forward(self, pred, target, match):
+        """Forward function."""
+        batch = pred.size(0)
+        pull_all, push_all = 0.0, 0.0
+        for i in range(batch):
+            pull, push = ae_loss_per_image(pred[i], target[i], match[i])
+
+            pull_all += self.pull_weight * pull
+            push_all += self.push_weight * push
+
+        return pull_all, push_all
diff --git a/python/autoCut/models/swin/mmdet/models/losses/balanced_l1_loss.py b/python/autoCut/models/swin/mmdet/models/losses/balanced_l1_loss.py
new file mode 100644
index 0000000..7bcd13f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/balanced_l1_loss.py
@@ -0,0 +1,120 @@
+import mmcv
+import numpy as np
+import torch
+import torch.nn as nn
+
+from ..builder import LOSSES
+from .utils import weighted_loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def balanced_l1_loss(pred,
+                     target,
+                     beta=1.0,
+                     alpha=0.5,
+                     gamma=1.5,
+                     reduction='mean'):
+    """Calculate balanced L1 loss.
+
+    Please see the `Libra R-CNN <https://arxiv.org/pdf/1904.02701.pdf>`_
+
+    Args:
+        pred (torch.Tensor): The prediction with shape (N, 4).
+        target (torch.Tensor): The learning target of the prediction with
+            shape (N, 4).
+        beta (float): The loss is a piecewise function of prediction and target
+            and ``beta`` serves as a threshold for the difference between the
+            prediction and target. Defaults to 1.0.
+        alpha (float): The denominator ``alpha`` in the balanced L1 loss.
+            Defaults to 0.5.
+        gamma (float): The ``gamma`` in the balanced L1 loss.
+            Defaults to 1.5.
+        reduction (str, optional): The method that reduces the loss to a
+            scalar. Options are "none", "mean" and "sum".
+
+    Returns:
+        torch.Tensor: The calculated loss
+    """
+    assert beta > 0
+    assert pred.size() == target.size() and target.numel() > 0
+
+    diff = torch.abs(pred - target)
+    b = np.e**(gamma / alpha) - 1
+    loss = torch.where(
+        diff < beta, alpha / b *
+        (b * diff + 1) * torch.log(b * diff / beta + 1) - alpha * diff,
+        gamma * diff + gamma / b - alpha * beta)
+
+    return loss
+
+
+@LOSSES.register_module()
+class BalancedL1Loss(nn.Module):
+    """Balanced L1 Loss.
+
+    arXiv: https://arxiv.org/pdf/1904.02701.pdf (CVPR 2019)
+
+    Args:
+        alpha (float): The denominator ``alpha`` in the balanced L1 loss.
+            Defaults to 0.5.
+        gamma (float): The ``gamma`` in the balanced L1 loss. Defaults to 1.5.
+        beta (float, optional): The loss is a piecewise function of prediction
+            and target. ``beta`` serves as a threshold for the difference
+            between the prediction and target. Defaults to 1.0.
+        reduction (str, optional): The method that reduces the loss to a
+            scalar. Options are "none", "mean" and "sum".
+        loss_weight (float, optional): The weight of the loss. Defaults to 1.0
+    """
+
+    def __init__(self,
+                 alpha=0.5,
+                 gamma=1.5,
+                 beta=1.0,
+                 reduction='mean',
+                 loss_weight=1.0):
+        super(BalancedL1Loss, self).__init__()
+        self.alpha = alpha
+        self.gamma = gamma
+        self.beta = beta
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        """Forward function of loss.
+
+        Args:
+            pred (torch.Tensor): The prediction with shape (N, 4).
+            target (torch.Tensor): The learning target of the prediction with
+                shape (N, 4).
+            weight (torch.Tensor, optional): Sample-wise loss weight with
+                shape (N, ).
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Options are "none", "mean" and "sum".
+
+        Returns:
+            torch.Tensor: The calculated loss
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        loss_bbox = self.loss_weight * balanced_l1_loss(
+            pred,
+            target,
+            weight,
+            alpha=self.alpha,
+            gamma=self.gamma,
+            beta=self.beta,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss_bbox
diff --git a/python/autoCut/models/swin/mmdet/models/losses/cross_entropy_loss.py b/python/autoCut/models/swin/mmdet/models/losses/cross_entropy_loss.py
new file mode 100644
index 0000000..5799415
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/cross_entropy_loss.py
@@ -0,0 +1,214 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from ..builder import LOSSES
+from .utils import weight_reduce_loss
+
+
+def cross_entropy(pred,
+                  label,
+                  weight=None,
+                  reduction='mean',
+                  avg_factor=None,
+                  class_weight=None):
+    """Calculate the CrossEntropy loss.
+
+    Args:
+        pred (torch.Tensor): The prediction with shape (N, C), C is the number
+            of classes.
+        label (torch.Tensor): The learning label of the prediction.
+        weight (torch.Tensor, optional): Sample-wise loss weight.
+        reduction (str, optional): The method used to reduce the loss.
+        avg_factor (int, optional): Average factor that is used to average
+            the loss. Defaults to None.
+        class_weight (list[float], optional): The weight for each class.
+
+    Returns:
+        torch.Tensor: The calculated loss
+    """
+    # element-wise losses
+    loss = F.cross_entropy(pred, label, weight=class_weight, reduction='none')
+
+    # apply weights and do the reduction
+    if weight is not None:
+        weight = weight.float()
+    loss = weight_reduce_loss(
+        loss, weight=weight, reduction=reduction, avg_factor=avg_factor)
+
+    return loss
+
+
+def _expand_onehot_labels(labels, label_weights, label_channels):
+    bin_labels = labels.new_full((labels.size(0), label_channels), 0)
+    inds = torch.nonzero(
+        (labels >= 0) & (labels < label_channels), as_tuple=False).squeeze()
+    if inds.numel() > 0:
+        bin_labels[inds, labels[inds]] = 1
+
+    if label_weights is None:
+        bin_label_weights = None
+    else:
+        bin_label_weights = label_weights.view(-1, 1).expand(
+            label_weights.size(0), label_channels)
+
+    return bin_labels, bin_label_weights
+
+
+def binary_cross_entropy(pred,
+                         label,
+                         weight=None,
+                         reduction='mean',
+                         avg_factor=None,
+                         class_weight=None):
+    """Calculate the binary CrossEntropy loss.
+
+    Args:
+        pred (torch.Tensor): The prediction with shape (N, 1).
+        label (torch.Tensor): The learning label of the prediction.
+        weight (torch.Tensor, optional): Sample-wise loss weight.
+        reduction (str, optional): The method used to reduce the loss.
+            Options are "none", "mean" and "sum".
+        avg_factor (int, optional): Average factor that is used to average
+            the loss. Defaults to None.
+        class_weight (list[float], optional): The weight for each class.
+
+    Returns:
+        torch.Tensor: The calculated loss
+    """
+    if pred.dim() != label.dim():
+        label, weight = _expand_onehot_labels(label, weight, pred.size(-1))
+
+    # weighted element-wise losses
+    if weight is not None:
+        weight = weight.float()
+    loss = F.binary_cross_entropy_with_logits(
+        pred, label.float(), pos_weight=class_weight, reduction='none')
+    # do the reduction for the weighted loss
+    loss = weight_reduce_loss(
+        loss, weight, reduction=reduction, avg_factor=avg_factor)
+
+    return loss
+
+
+def mask_cross_entropy(pred,
+                       target,
+                       label,
+                       reduction='mean',
+                       avg_factor=None,
+                       class_weight=None):
+    """Calculate the CrossEntropy loss for masks.
+
+    Args:
+        pred (torch.Tensor): The prediction with shape (N, C, *), C is the
+            number of classes. The trailing * indicates arbitrary shape.
+        target (torch.Tensor): The learning label of the prediction.
+        label (torch.Tensor): ``label`` indicates the class label of the mask
+            corresponding object. This will be used to select the mask in the
+            of the class which the object belongs to when the mask prediction
+            if not class-agnostic.
+        reduction (str, optional): The method used to reduce the loss.
+            Options are "none", "mean" and "sum".
+        avg_factor (int, optional): Average factor that is used to average
+            the loss. Defaults to None.
+        class_weight (list[float], optional): The weight for each class.
+
+    Returns:
+        torch.Tensor: The calculated loss
+
+    Example:
+        >>> N, C = 3, 11
+        >>> H, W = 2, 2
+        >>> pred = torch.randn(N, C, H, W) * 1000
+        >>> target = torch.rand(N, H, W)
+        >>> label = torch.randint(0, C, size=(N,))
+        >>> reduction = 'mean'
+        >>> avg_factor = None
+        >>> class_weights = None
+        >>> loss = mask_cross_entropy(pred, target, label, reduction,
+        >>>                           avg_factor, class_weights)
+        >>> assert loss.shape == (1,)
+    """
+    # TODO: handle these two reserved arguments
+    assert reduction == 'mean' and avg_factor is None
+    num_rois = pred.size()[0]
+    inds = torch.arange(0, num_rois, dtype=torch.long, device=pred.device)
+    pred_slice = pred[inds, label].squeeze(1)
+    return F.binary_cross_entropy_with_logits(
+        pred_slice, target, weight=class_weight, reduction='mean')[None]
+
+
+@LOSSES.register_module()
+class CrossEntropyLoss(nn.Module):
+
+    def __init__(self,
+                 use_sigmoid=False,
+                 use_mask=False,
+                 reduction='mean',
+                 class_weight=None,
+                 loss_weight=1.0):
+        """CrossEntropyLoss.
+
+        Args:
+            use_sigmoid (bool, optional): Whether the prediction uses sigmoid
+                of softmax. Defaults to False.
+            use_mask (bool, optional): Whether to use mask cross entropy loss.
+                Defaults to False.
+            reduction (str, optional): . Defaults to 'mean'.
+                Options are "none", "mean" and "sum".
+            class_weight (list[float], optional): Weight of each class.
+                Defaults to None.
+            loss_weight (float, optional): Weight of the loss. Defaults to 1.0.
+        """
+        super(CrossEntropyLoss, self).__init__()
+        assert (use_sigmoid is False) or (use_mask is False)
+        self.use_sigmoid = use_sigmoid
+        self.use_mask = use_mask
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+        self.class_weight = class_weight
+
+        if self.use_sigmoid:
+            self.cls_criterion = binary_cross_entropy
+        elif self.use_mask:
+            self.cls_criterion = mask_cross_entropy
+        else:
+            self.cls_criterion = cross_entropy
+
+    def forward(self,
+                cls_score,
+                label,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        """Forward function.
+
+        Args:
+            cls_score (torch.Tensor): The prediction.
+            label (torch.Tensor): The learning label of the prediction.
+            weight (torch.Tensor, optional): Sample-wise loss weight.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction (str, optional): The method used to reduce the loss.
+                Options are "none", "mean" and "sum".
+        Returns:
+            torch.Tensor: The calculated loss
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if self.class_weight is not None:
+            class_weight = cls_score.new_tensor(
+                self.class_weight, device=cls_score.device)
+        else:
+            class_weight = None
+        loss_cls = self.loss_weight * self.cls_criterion(
+            cls_score,
+            label,
+            weight,
+            class_weight=class_weight,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss_cls
diff --git a/python/autoCut/models/swin/mmdet/models/losses/focal_loss.py b/python/autoCut/models/swin/mmdet/models/losses/focal_loss.py
new file mode 100644
index 0000000..4446174
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/focal_loss.py
@@ -0,0 +1,244 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.ops import sigmoid_focal_loss as _sigmoid_focal_loss
+
+from ..builder import LOSSES
+from .utils import weight_reduce_loss
+
+
+# This method is only for debugging
+def py_sigmoid_focal_loss(pred,
+                          target,
+                          weight=None,
+                          gamma=2.0,
+                          alpha=0.25,
+                          reduction='mean',
+                          avg_factor=None):
+    """PyTorch version of `Focal Loss <https://arxiv.org/abs/1708.02002>`_.
+
+    Args:
+        pred (torch.Tensor): The prediction with shape (N, C), C is the
+            number of classes
+        target (torch.Tensor): The learning label of the prediction.
+        weight (torch.Tensor, optional): Sample-wise loss weight.
+        gamma (float, optional): The gamma for calculating the modulating
+            factor. Defaults to 2.0.
+        alpha (float, optional): A balanced form for Focal Loss.
+            Defaults to 0.25.
+        reduction (str, optional): The method used to reduce the loss into
+            a scalar. Defaults to 'mean'.
+        avg_factor (int, optional): Average factor that is used to average
+            the loss. Defaults to None.
+    """
+    pred_sigmoid = pred.sigmoid()
+    target = target.type_as(pred)
+    pt = (1 - pred_sigmoid) * target + pred_sigmoid * (1 - target)
+    focal_weight = (alpha * target + (1 - alpha) *
+                    (1 - target)) * pt.pow(gamma)
+    loss = F.binary_cross_entropy_with_logits(
+        pred, target, reduction='none') * focal_weight
+    if weight is not None:
+        if weight.shape != loss.shape:
+            if weight.size(0) == loss.size(0):
+                # For most cases, weight is of shape (num_priors, ),
+                #  which means it does not have the second axis num_class
+                weight = weight.view(-1, 1)
+            else:
+                # Sometimes, weight per anchor per class is also needed. e.g.
+                #  in FSAF. But it may be flattened of shape
+                #  (num_priors x num_class, ), while loss is still of shape
+                #  (num_priors, num_class).
+                assert weight.numel() == loss.numel()
+                weight = weight.view(loss.size(0), -1)
+        assert weight.ndim == loss.ndim
+    loss = weight_reduce_loss(loss, weight, reduction, avg_factor)
+    return loss
+
+def separate_sigmoid_focal_loss(pred,
+                                target,
+                                weight=None,
+                                gamma=2.0,
+                                alpha=0.25,
+                                reduction='mean',
+                                avg_factor=None):
+    pred_sigmoid = pred.sigmoid()
+    target = target.type_as(pred)
+
+    pos_inds = target.eq(1)
+    neg_inds = target.lt(1)
+
+    pos_weights = weight[pos_inds]
+
+    pos_pred = pred_sigmoid[pos_inds]
+    neg_pred = pred_sigmoid[neg_inds]
+
+    pos_loss = -torch.log(pos_pred) * torch.pow(1 -
+                                                pos_pred, gamma) * pos_weights * alpha
+    neg_loss = -torch.log(1 - neg_pred) * \
+        torch.pow(neg_pred, gamma) * (1 - alpha)
+
+    if pos_pred.nelement() == 0:
+        loss = neg_loss.sum() / avg_factor
+    else:
+        loss = pos_loss.sum() / pos_weights.sum() + neg_loss.sum() / avg_factor
+
+    return loss
+    
+def sigmoid_focal_loss(pred,
+                       target,
+                       weight=None,
+                       gamma=2.0,
+                       alpha=0.25,
+                       reduction='mean',
+                       avg_factor=None):
+    r"""A warpper of cuda version `Focal Loss
+    <https://arxiv.org/abs/1708.02002>`_.
+
+    Args:
+        pred (torch.Tensor): The prediction with shape (N, C), C is the number
+            of classes.
+        target (torch.Tensor): The learning label of the prediction.
+        weight (torch.Tensor, optional): Sample-wise loss weight.
+        gamma (float, optional): The gamma for calculating the modulating
+            factor. Defaults to 2.0.
+        alpha (float, optional): A balanced form for Focal Loss.
+            Defaults to 0.25.
+        reduction (str, optional): The method used to reduce the loss into
+            a scalar. Defaults to 'mean'. Options are "none", "mean" and "sum".
+        avg_factor (int, optional): Average factor that is used to average
+            the loss. Defaults to None.
+    """
+    # Function.apply does not accept keyword arguments, so the decorator
+    # "weighted_loss" is not applicable
+    loss = _sigmoid_focal_loss(pred.contiguous(), target, gamma, alpha, None,
+                               'none')
+    if weight is not None:
+        if weight.shape != loss.shape:
+            if weight.size(0) == loss.size(0):
+                # For most cases, weight is of shape (num_priors, ),
+                #  which means it does not have the second axis num_class
+                weight = weight.view(-1, 1)
+            else:
+                # Sometimes, weight per anchor per class is also needed. e.g.
+                #  in FSAF. But it may be flattened of shape
+                #  (num_priors x num_class, ), while loss is still of shape
+                #  (num_priors, num_class).
+                assert weight.numel() == loss.numel()
+                weight = weight.view(loss.size(0), -1)
+        assert weight.ndim == loss.ndim
+    loss = weight_reduce_loss(loss, weight, reduction, avg_factor)
+    return loss
+
+
+@LOSSES.register_module()
+class FocalLoss(nn.Module):
+
+    def __init__(self,
+                 use_sigmoid=True,
+                 gamma=2.0,
+                 alpha=0.25,
+                 reduction='mean',
+                 loss_weight=1.0):
+        """`Focal Loss <https://arxiv.org/abs/1708.02002>`_
+
+        Args:
+            use_sigmoid (bool, optional): Whether to the prediction is
+                used for sigmoid or softmax. Defaults to True.
+            gamma (float, optional): The gamma for calculating the modulating
+                factor. Defaults to 2.0.
+            alpha (float, optional): A balanced form for Focal Loss.
+                Defaults to 0.25.
+            reduction (str, optional): The method used to reduce the loss into
+                a scalar. Defaults to 'mean'. Options are "none", "mean" and
+                "sum".
+            loss_weight (float, optional): Weight of loss. Defaults to 1.0.
+        """
+        super(FocalLoss, self).__init__()
+        assert use_sigmoid is True, 'Only sigmoid focal loss supported now.'
+        self.use_sigmoid = use_sigmoid
+        self.gamma = gamma
+        self.alpha = alpha
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): The prediction.
+            target (torch.Tensor): The learning label of the prediction.
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Options are "none", "mean" and "sum".
+
+        Returns:
+            torch.Tensor: The calculated loss
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if self.use_sigmoid:
+            if torch.cuda.is_available() and pred.is_cuda:
+                calculate_loss_func = sigmoid_focal_loss
+            else:
+                num_classes = pred.size(1)
+                target = F.one_hot(target, num_classes=num_classes + 1)
+                target = target[:, :num_classes]
+                calculate_loss_func = py_sigmoid_focal_loss
+
+            loss_cls = self.loss_weight * calculate_loss_func(
+                pred,
+                target,
+                weight,
+                gamma=self.gamma,
+                alpha=self.alpha,
+                reduction=reduction,
+                avg_factor=avg_factor)
+
+        else:
+            raise NotImplementedError
+        return loss_cls
+
+@LOSSES.register_module()
+class SEPFocalLoss(nn.Module):
+
+    def __init__(self,
+                 gamma=2.0,
+                 alpha=0.25,
+                 reduction='mean',
+                 loss_weight=1.0):
+        super(SEPFocalLoss, self).__init__()
+        self.gamma = gamma
+        self.alpha = alpha
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None):
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        loss_cls = self.loss_weight * separate_sigmoid_focal_loss(
+            pred,
+            target,
+            weight,
+            gamma=self.gamma,
+            alpha=self.alpha,
+            reduction=reduction,
+            avg_factor=avg_factor)
+
+        return loss_cls
diff --git a/python/autoCut/models/swin/mmdet/models/losses/gaussian_focal_loss.py b/python/autoCut/models/swin/mmdet/models/losses/gaussian_focal_loss.py
new file mode 100644
index 0000000..e45506a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/gaussian_focal_loss.py
@@ -0,0 +1,91 @@
+import mmcv
+import torch.nn as nn
+
+from ..builder import LOSSES
+from .utils import weighted_loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def gaussian_focal_loss(pred, gaussian_target, alpha=2.0, gamma=4.0):
+    """`Focal Loss <https://arxiv.org/abs/1708.02002>`_ for targets in gaussian
+    distribution.
+
+    Args:
+        pred (torch.Tensor): The prediction.
+        gaussian_target (torch.Tensor): The learning target of the prediction
+            in gaussian distribution.
+        alpha (float, optional): A balanced form for Focal Loss.
+            Defaults to 2.0.
+        gamma (float, optional): The gamma for calculating the modulating
+            factor. Defaults to 4.0.
+    """
+    eps = 1e-12
+    pos_weights = gaussian_target.eq(1)
+    neg_weights = (1 - gaussian_target).pow(gamma)
+    pos_loss = -(pred + eps).log() * (1 - pred).pow(alpha) * pos_weights
+    neg_loss = -(1 - pred + eps).log() * pred.pow(alpha) * neg_weights
+    return pos_loss + neg_loss
+
+
+@LOSSES.register_module()
+class GaussianFocalLoss(nn.Module):
+    """GaussianFocalLoss is a variant of focal loss.
+
+    More details can be found in the `paper
+    <https://arxiv.org/abs/1808.01244>`_
+    Code is modified from `kp_utils.py
+    <https://github.com/princeton-vl/CornerNet/blob/master/models/py_utils/kp_utils.py#L152>`_  # noqa: E501
+    Please notice that the target in GaussianFocalLoss is a gaussian heatmap,
+    not 0/1 binary target.
+
+    Args:
+        alpha (float): Power of prediction.
+        gamma (float): Power of target for negative samples.
+        reduction (str): Options are "none", "mean" and "sum".
+        loss_weight (float): Loss weight of current loss.
+    """
+
+    def __init__(self,
+                 alpha=2.0,
+                 gamma=4.0,
+                 reduction='mean',
+                 loss_weight=1.0):
+        super(GaussianFocalLoss, self).__init__()
+        self.alpha = alpha
+        self.gamma = gamma
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): The prediction.
+            target (torch.Tensor): The learning target of the prediction
+                in gaussian distribution.
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Defaults to None.
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        loss_reg = self.loss_weight * gaussian_focal_loss(
+            pred,
+            target,
+            weight,
+            alpha=self.alpha,
+            gamma=self.gamma,
+            reduction=reduction,
+            avg_factor=avg_factor)
+        return loss_reg
diff --git a/python/autoCut/models/swin/mmdet/models/losses/gfocal_loss.py b/python/autoCut/models/swin/mmdet/models/losses/gfocal_loss.py
new file mode 100644
index 0000000..c590539
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/gfocal_loss.py
@@ -0,0 +1,274 @@
+import mmcv
+import torch.nn as nn
+import torch.nn.functional as F
+
+from ..builder import LOSSES
+from .utils import weighted_loss
+
+from mmdet.core import bbox_overlaps
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def quality_focal_loss(pred, target, beta=2.0, background_label=None):
+    r"""Quality Focal Loss (QFL) is from `Generalized Focal Loss: Learning
+    Qualified and Distributed Bounding Boxes for Dense Object Detection
+    <https://arxiv.org/abs/2006.04388>`_.
+
+    Args:
+        pred (torch.Tensor): Predicted joint representation of classification
+            and quality (IoU) estimation with shape (N, C), C is the number of
+            classes.
+        target (tuple([torch.Tensor])): Target category label with shape (N,)
+            and target quality label with shape (N,).
+        beta (float): The beta parameter for calculating the modulating factor.
+            Defaults to 2.0.
+
+    Returns:
+        torch.Tensor: Loss tensor with shape (N,).
+    """
+    assert len(target) == 2, """target for QFL must be a tuple of two elements,
+        including category label and quality label, respectively"""
+    # label denotes the category id, score denotes the quality score
+    label, score = target
+
+    # negatives are supervised by 0 quality score
+    pred_sigmoid = pred.sigmoid()
+    scale_factor = pred_sigmoid
+    zerolabel = scale_factor.new_zeros(pred.shape)
+    loss = F.binary_cross_entropy_with_logits(
+        pred, zerolabel, reduction='none') * scale_factor.pow(beta)
+
+    # FG cat_id: [0, num_classes -1], BG cat_id: num_classes
+    if background_label is None:
+        bg_class_ind = pred.size(1)
+        pos = ((label >= 0) & (label < bg_class_ind)).nonzero().squeeze(1)
+    else:
+        pos = (label!=background_label).nonzero().squeeze(1)
+    pos_label = label[pos].long()
+    # positives are supervised by bbox quality (IoU) score
+    if len(pos) == 0:
+        pass
+    else:
+        scale_factor = score[pos] - pred_sigmoid[pos, pos_label]
+        loss[pos, pos_label] = F.binary_cross_entropy_with_logits(
+            pred[pos, pos_label], score[pos],
+            reduction='none') * scale_factor.abs().pow(beta)
+
+    loss = loss.sum(dim=1, keepdim=False)
+    return loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def distribution_focal_loss(pred, label):
+    r"""Distribution Focal Loss (DFL) is from `Generalized Focal Loss: Learning
+    Qualified and Distributed Bounding Boxes for Dense Object Detection
+    <https://arxiv.org/abs/2006.04388>`_.
+
+    Args:
+        pred (torch.Tensor): Predicted general distribution of bounding boxes
+            (before softmax) with shape (N, n+1), n is the max value of the
+            integral set `{0, ..., n}` in paper.
+        label (torch.Tensor): Target distance label for bounding boxes with
+            shape (N,).
+
+    Returns:
+        torch.Tensor: Loss tensor with shape (N,).
+    """
+    dis_left = label.long()
+    dis_right = dis_left + 1
+    weight_left = dis_right.float() - label
+    weight_right = label - dis_left.float()
+    loss = F.cross_entropy(pred, dis_left, reduction='none') * weight_left \
+        + F.cross_entropy(pred, dis_right, reduction='none') * weight_right
+    return loss
+
+
+@LOSSES.register_module()
+class QualityFocalLoss(nn.Module):
+    r"""Quality Focal Loss (QFL) is a variant of `Generalized Focal Loss:
+    Learning Qualified and Distributed Bounding Boxes for Dense Object
+    Detection <https://arxiv.org/abs/2006.04388>`_.
+
+    Args:
+        use_sigmoid (bool): Whether sigmoid operation is conducted in QFL.
+            Defaults to True.
+        beta (float): The beta parameter for calculating the modulating factor.
+            Defaults to 2.0.
+        reduction (str): Options are "none", "mean" and "sum".
+        loss_weight (float): Loss weight of current loss.
+    """
+
+    def __init__(self,
+                 use_sigmoid=True,
+                 beta=2.0,
+                 reduction='mean',
+                 loss_weight=1.0):
+        super(QualityFocalLoss, self).__init__()
+        assert use_sigmoid is True, 'Only sigmoid in QFL supported now.'
+        self.use_sigmoid = use_sigmoid
+        self.beta = beta
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): Predicted joint representation of
+                classification and quality (IoU) estimation with shape (N, C),
+                C is the number of classes.
+            target (tuple([torch.Tensor])): Target category label with shape
+                (N,) and target quality label with shape (N,).
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Defaults to None.
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if self.use_sigmoid:
+            loss_cls = self.loss_weight * quality_focal_loss(
+                pred,
+                target,
+                weight,
+                beta=self.beta,
+                reduction=reduction,
+                avg_factor=avg_factor)
+        else:
+            raise NotImplementedError
+        return loss_cls
+
+@LOSSES.register_module()
+class RPDQualityFocalLoss(nn.Module):
+    r"""Quality Focal Loss (QFL) is a variant of `Generalized Focal Loss:
+    Learning Qualified and Distributed Bounding Boxes for Dense Object
+    Detection <https://arxiv.org/abs/2006.04388>`_.
+
+    Args:
+        use_sigmoid (bool): Whether sigmoid operation is conducted in QFL.
+            Defaults to True.
+        beta (float): The beta parameter for calculating the modulating factor.
+            Defaults to 2.0.
+        reduction (str): Options are "none", "mean" and "sum".
+        loss_weight (float): Loss weight of current loss.
+    """
+
+    def __init__(self,
+                 use_sigmoid=True,
+                 beta=2.0,
+                 reduction='mean',
+                 loss_weight=1.0):
+        super(RPDQualityFocalLoss, self).__init__()
+        assert use_sigmoid is True, 'Only sigmoid in QFL supported now.'
+        self.use_sigmoid = use_sigmoid
+        self.beta = beta
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+        self.requires_box=True
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                bbox_pred=None,
+                bbox_target=None,
+                bbox_weight=None,
+                background_label=0,
+                avg_factor=None,
+                reduction_override=None):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): Predicted joint representation of
+                classification and quality (IoU) estimation with shape (N, C),
+                C is the number of classes.
+            target (tuple([torch.Tensor])): Target category label with shape
+                (N,) and target quality label with shape (N,).
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Defaults to None.
+        """
+        #compute iou score
+        iou_score = bbox_target.new_zeros(target.size()).squeeze()
+        pos_inds = (target!=background_label).nonzero().squeeze(1)
+        if len(pos_inds)>0:
+            score=bbox_overlaps(bbox_pred[pos_inds],bbox_target[pos_inds],is_aligned=True)
+            #print(score)
+            iou_score[pos_inds]=score.detach()
+        iou_score = iou_score.detach()
+
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if self.use_sigmoid:
+            loss_cls = self.loss_weight * quality_focal_loss(
+                pred,
+                (target,iou_score),
+                weight,
+                beta=self.beta,
+                background_label=background_label,
+                reduction=reduction,
+                avg_factor=avg_factor)
+        else:
+            raise NotImplementedError
+        return loss_cls
+
+
+@LOSSES.register_module()
+class DistributionFocalLoss(nn.Module):
+    r"""Distribution Focal Loss (DFL) is a variant of `Generalized Focal Loss:
+    Learning Qualified and Distributed Bounding Boxes for Dense Object
+    Detection <https://arxiv.org/abs/2006.04388>`_.
+
+    Args:
+        reduction (str): Options are `'none'`, `'mean'` and `'sum'`.
+        loss_weight (float): Loss weight of current loss.
+    """
+
+    def __init__(self, reduction='mean', loss_weight=1.0):
+        super(DistributionFocalLoss, self).__init__()
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): Predicted general distribution of bounding
+                boxes (before softmax) with shape (N, n+1), n is the max value
+                of the integral set `{0, ..., n}` in paper.
+            target (torch.Tensor): Target distance label for bounding boxes
+                with shape (N,).
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Defaults to None.
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        loss_cls = self.loss_weight * distribution_focal_loss(
+            pred, target, weight, reduction=reduction, avg_factor=avg_factor)
+        return loss_cls
diff --git a/python/autoCut/models/swin/mmdet/models/losses/ghm_loss.py b/python/autoCut/models/swin/mmdet/models/losses/ghm_loss.py
new file mode 100644
index 0000000..8969a23
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/ghm_loss.py
@@ -0,0 +1,172 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from ..builder import LOSSES
+
+
+def _expand_onehot_labels(labels, label_weights, label_channels):
+    bin_labels = labels.new_full((labels.size(0), label_channels), 0)
+    inds = torch.nonzero(
+        (labels >= 0) & (labels < label_channels), as_tuple=False).squeeze()
+    if inds.numel() > 0:
+        bin_labels[inds, labels[inds]] = 1
+    bin_label_weights = label_weights.view(-1, 1).expand(
+        label_weights.size(0), label_channels)
+    return bin_labels, bin_label_weights
+
+
+# TODO: code refactoring to make it consistent with other losses
+@LOSSES.register_module()
+class GHMC(nn.Module):
+    """GHM Classification Loss.
+
+    Details of the theorem can be viewed in the paper
+    `Gradient Harmonized Single-stage Detector
+    <https://arxiv.org/abs/1811.05181>`_.
+
+    Args:
+        bins (int): Number of the unit regions for distribution calculation.
+        momentum (float): The parameter for moving average.
+        use_sigmoid (bool): Can only be true for BCE based loss now.
+        loss_weight (float): The weight of the total GHM-C loss.
+    """
+
+    def __init__(self, bins=10, momentum=0, use_sigmoid=True, loss_weight=1.0):
+        super(GHMC, self).__init__()
+        self.bins = bins
+        self.momentum = momentum
+        edges = torch.arange(bins + 1).float() / bins
+        self.register_buffer('edges', edges)
+        self.edges[-1] += 1e-6
+        if momentum > 0:
+            acc_sum = torch.zeros(bins)
+            self.register_buffer('acc_sum', acc_sum)
+        self.use_sigmoid = use_sigmoid
+        if not self.use_sigmoid:
+            raise NotImplementedError
+        self.loss_weight = loss_weight
+
+    def forward(self, pred, target, label_weight, *args, **kwargs):
+        """Calculate the GHM-C loss.
+
+        Args:
+            pred (float tensor of size [batch_num, class_num]):
+                The direct prediction of classification fc layer.
+            target (float tensor of size [batch_num, class_num]):
+                Binary class target for each sample.
+            label_weight (float tensor of size [batch_num, class_num]):
+                the value is 1 if the sample is valid and 0 if ignored.
+        Returns:
+            The gradient harmonized loss.
+        """
+        # the target should be binary class label
+        if pred.dim() != target.dim():
+            target, label_weight = _expand_onehot_labels(
+                target, label_weight, pred.size(-1))
+        target, label_weight = target.float(), label_weight.float()
+        edges = self.edges
+        mmt = self.momentum
+        weights = torch.zeros_like(pred)
+
+        # gradient length
+        g = torch.abs(pred.sigmoid().detach() - target)
+
+        valid = label_weight > 0
+        tot = max(valid.float().sum().item(), 1.0)
+        n = 0  # n valid bins
+        for i in range(self.bins):
+            inds = (g >= edges[i]) & (g < edges[i + 1]) & valid
+            num_in_bin = inds.sum().item()
+            if num_in_bin > 0:
+                if mmt > 0:
+                    self.acc_sum[i] = mmt * self.acc_sum[i] \
+                        + (1 - mmt) * num_in_bin
+                    weights[inds] = tot / self.acc_sum[i]
+                else:
+                    weights[inds] = tot / num_in_bin
+                n += 1
+        if n > 0:
+            weights = weights / n
+
+        loss = F.binary_cross_entropy_with_logits(
+            pred, target, weights, reduction='sum') / tot
+        return loss * self.loss_weight
+
+
+# TODO: code refactoring to make it consistent with other losses
+@LOSSES.register_module()
+class GHMR(nn.Module):
+    """GHM Regression Loss.
+
+    Details of the theorem can be viewed in the paper
+    `Gradient Harmonized Single-stage Detector
+    <https://arxiv.org/abs/1811.05181>`_.
+
+    Args:
+        mu (float): The parameter for the Authentic Smooth L1 loss.
+        bins (int): Number of the unit regions for distribution calculation.
+        momentum (float): The parameter for moving average.
+        loss_weight (float): The weight of the total GHM-R loss.
+    """
+
+    def __init__(self, mu=0.02, bins=10, momentum=0, loss_weight=1.0):
+        super(GHMR, self).__init__()
+        self.mu = mu
+        self.bins = bins
+        edges = torch.arange(bins + 1).float() / bins
+        self.register_buffer('edges', edges)
+        self.edges[-1] = 1e3
+        self.momentum = momentum
+        if momentum > 0:
+            acc_sum = torch.zeros(bins)
+            self.register_buffer('acc_sum', acc_sum)
+        self.loss_weight = loss_weight
+
+    # TODO: support reduction parameter
+    def forward(self, pred, target, label_weight, avg_factor=None):
+        """Calculate the GHM-R loss.
+
+        Args:
+            pred (float tensor of size [batch_num, 4 (* class_num)]):
+                The prediction of box regression layer. Channel number can be 4
+                or 4 * class_num depending on whether it is class-agnostic.
+            target (float tensor of size [batch_num, 4 (* class_num)]):
+                The target regression values with the same size of pred.
+            label_weight (float tensor of size [batch_num, 4 (* class_num)]):
+                The weight of each sample, 0 if ignored.
+        Returns:
+            The gradient harmonized loss.
+        """
+        mu = self.mu
+        edges = self.edges
+        mmt = self.momentum
+
+        # ASL1 loss
+        diff = pred - target
+        loss = torch.sqrt(diff * diff + mu * mu) - mu
+
+        # gradient length
+        g = torch.abs(diff / torch.sqrt(mu * mu + diff * diff)).detach()
+        weights = torch.zeros_like(g)
+
+        valid = label_weight > 0
+        tot = max(label_weight.float().sum().item(), 1.0)
+        n = 0  # n: valid bins
+        for i in range(self.bins):
+            inds = (g >= edges[i]) & (g < edges[i + 1]) & valid
+            num_in_bin = inds.sum().item()
+            if num_in_bin > 0:
+                n += 1
+                if mmt > 0:
+                    self.acc_sum[i] = mmt * self.acc_sum[i] \
+                        + (1 - mmt) * num_in_bin
+                    weights[inds] = tot / self.acc_sum[i]
+                else:
+                    weights[inds] = tot / num_in_bin
+        if n > 0:
+            weights /= n
+
+        loss = loss * weights
+        loss = loss.sum() / tot
+        return loss * self.loss_weight
diff --git a/python/autoCut/models/swin/mmdet/models/losses/iou_loss.py b/python/autoCut/models/swin/mmdet/models/losses/iou_loss.py
new file mode 100644
index 0000000..a5ebe70
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/iou_loss.py
@@ -0,0 +1,520 @@
+import math
+
+import mmcv
+import torch
+import torch.nn as nn
+
+from mmdet.core import bbox_overlaps
+from ..builder import LOSSES
+from .utils import weighted_loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def iou_loss(pred, target, linear=False, eps=1e-6):
+    """IoU loss.
+
+    Computing the IoU loss between a set of predicted bboxes and target bboxes.
+    The loss is calculated as negative log of IoU.
+
+    Args:
+        pred (torch.Tensor): Predicted bboxes of format (x1, y1, x2, y2),
+            shape (n, 4).
+        target (torch.Tensor): Corresponding gt bboxes, shape (n, 4).
+        linear (bool, optional): If True, use linear scale of loss instead of
+            log scale. Default: False.
+        eps (float): Eps to avoid log(0).
+
+    Return:
+        torch.Tensor: Loss tensor.
+    """
+    ious = bbox_overlaps(pred, target, is_aligned=True).clamp(min=eps)
+    if linear:
+        loss = 1 - ious
+    else:
+        loss = -ious.log()
+    return loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def bounded_iou_loss(pred, target, beta=0.2, eps=1e-3):
+    """BIoULoss.
+
+    This is an implementation of paper
+    `Improving Object Localization with Fitness NMS and Bounded IoU Loss.
+    <https://arxiv.org/abs/1711.00164>`_.
+
+    Args:
+        pred (torch.Tensor): Predicted bboxes.
+        target (torch.Tensor): Target bboxes.
+        beta (float): beta parameter in smoothl1.
+        eps (float): eps to avoid NaN.
+    """
+    pred_ctrx = (pred[:, 0] + pred[:, 2]) * 0.5
+    pred_ctry = (pred[:, 1] + pred[:, 3]) * 0.5
+    pred_w = pred[:, 2] - pred[:, 0]
+    pred_h = pred[:, 3] - pred[:, 1]
+    with torch.no_grad():
+        target_ctrx = (target[:, 0] + target[:, 2]) * 0.5
+        target_ctry = (target[:, 1] + target[:, 3]) * 0.5
+        target_w = target[:, 2] - target[:, 0]
+        target_h = target[:, 3] - target[:, 1]
+
+    dx = target_ctrx - pred_ctrx
+    dy = target_ctry - pred_ctry
+
+    loss_dx = 1 - torch.max(
+        (target_w - 2 * dx.abs()) /
+        (target_w + 2 * dx.abs() + eps), torch.zeros_like(dx))
+    loss_dy = 1 - torch.max(
+        (target_h - 2 * dy.abs()) /
+        (target_h + 2 * dy.abs() + eps), torch.zeros_like(dy))
+    loss_dw = 1 - torch.min(target_w / (pred_w + eps), pred_w /
+                            (target_w + eps))
+    loss_dh = 1 - torch.min(target_h / (pred_h + eps), pred_h /
+                            (target_h + eps))
+    loss_comb = torch.stack([loss_dx, loss_dy, loss_dw, loss_dh],
+                            dim=-1).view(loss_dx.size(0), -1)
+
+    loss = torch.where(loss_comb < beta, 0.5 * loss_comb * loss_comb / beta,
+                       loss_comb - 0.5 * beta)
+    return loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def giou_loss(pred, target, eps=1e-7):
+    r"""`Generalized Intersection over Union: A Metric and A Loss for Bounding
+    Box Regression <https://arxiv.org/abs/1902.09630>`_.
+
+    Args:
+        pred (torch.Tensor): Predicted bboxes of format (x1, y1, x2, y2),
+            shape (n, 4).
+        target (torch.Tensor): Corresponding gt bboxes, shape (n, 4).
+        eps (float): Eps to avoid log(0).
+
+    Return:
+        Tensor: Loss tensor.
+    """
+    gious = bbox_overlaps(pred, target, mode='giou', is_aligned=True, eps=eps)
+    loss = 1 - gious
+    return loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def giou_loss_oldver(pred, target, eps=1e-7):
+    """
+    Generalized Intersection over Union: A Metric and A Loss for
+    Bounding Box Regression
+    https://arxiv.org/abs/1902.09630
+
+    code refer to:
+    https://github.com/sfzhang15/ATSS/blob/master/atss_core/modeling/rpn/atss/loss.py#L36
+
+    Args:
+        pred (torch.Tensor): Predicted bboxes of format (x1, y1, x2, y2),
+            shape (n, 4).
+        target (torch.Tensor): Corresponding gt bboxes, shape (n, 4).
+        eps (float): Eps to avoid log(0).
+
+    Return:
+        Tensor: Loss tensor.
+    """
+    # overlap
+    lt = torch.max(pred[:, :2], target[:, :2])
+    rb = torch.min(pred[:, 2:], target[:, 2:])
+    wh = (rb - lt + 1).clamp(min=0)
+    overlap = wh[:, 0] * wh[:, 1]
+
+    # union
+    ap = (pred[:, 2] - pred[:, 0] + 1) * (pred[:, 3] - pred[:, 1] + 1)
+    ag = (target[:, 2] - target[:, 0] + 1) * (target[:, 3] - target[:, 1] + 1)
+    union = ap + ag - overlap + eps
+
+    # IoU
+    ious = overlap / union
+
+    # enclose area
+    enclose_x1y1 = torch.min(pred[:, :2], target[:, :2])
+    enclose_x2y2 = torch.max(pred[:, 2:], target[:, 2:])
+    enclose_wh = (enclose_x2y2 - enclose_x1y1 + 1).clamp(min=0)
+    enclose_area = enclose_wh[:, 0] * enclose_wh[:, 1] + eps
+
+    # GIoU
+    gious = ious - (enclose_area - union) / enclose_area
+    loss = 1 - gious
+    return loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def diou_loss(pred, target, eps=1e-7):
+    r"""`Implementation of Distance-IoU Loss: Faster and Better
+    Learning for Bounding Box Regression, https://arxiv.org/abs/1911.08287`_.
+
+    Code is modified from https://github.com/Zzh-tju/DIoU.
+
+    Args:
+        pred (Tensor): Predicted bboxes of format (x1, y1, x2, y2),
+            shape (n, 4).
+        target (Tensor): Corresponding gt bboxes, shape (n, 4).
+        eps (float): Eps to avoid log(0).
+    Return:
+        Tensor: Loss tensor.
+    """
+    # overlap
+    lt = torch.max(pred[:, :2], target[:, :2])
+    rb = torch.min(pred[:, 2:], target[:, 2:])
+    wh = (rb - lt).clamp(min=0)
+    overlap = wh[:, 0] * wh[:, 1]
+
+    # union
+    ap = (pred[:, 2] - pred[:, 0]) * (pred[:, 3] - pred[:, 1])
+    ag = (target[:, 2] - target[:, 0]) * (target[:, 3] - target[:, 1])
+    union = ap + ag - overlap + eps
+
+    # IoU
+    ious = overlap / union
+
+    # enclose area
+    enclose_x1y1 = torch.min(pred[:, :2], target[:, :2])
+    enclose_x2y2 = torch.max(pred[:, 2:], target[:, 2:])
+    enclose_wh = (enclose_x2y2 - enclose_x1y1).clamp(min=0)
+
+    cw = enclose_wh[:, 0]
+    ch = enclose_wh[:, 1]
+
+    c2 = cw**2 + ch**2 + eps
+
+    b1_x1, b1_y1 = pred[:, 0], pred[:, 1]
+    b1_x2, b1_y2 = pred[:, 2], pred[:, 3]
+    b2_x1, b2_y1 = target[:, 0], target[:, 1]
+    b2_x2, b2_y2 = target[:, 2], target[:, 3]
+
+    left = ((b2_x1 + b2_x2) - (b1_x1 + b1_x2))**2 / 4
+    right = ((b2_y1 + b2_y2) - (b1_y1 + b1_y2))**2 / 4
+    rho2 = left + right
+
+    # DIoU
+    dious = ious - rho2 / c2
+    loss = 1 - dious
+    return loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def ciou_loss(pred, target, eps=1e-7):
+    r"""`Implementation of paper `Enhancing Geometric Factors into
+    Model Learning and Inference for Object Detection and Instance
+    Segmentation <https://arxiv.org/abs/2005.03572>`_.
+
+    Code is modified from https://github.com/Zzh-tju/CIoU.
+
+    Args:
+        pred (Tensor): Predicted bboxes of format (x1, y1, x2, y2),
+            shape (n, 4).
+        target (Tensor): Corresponding gt bboxes, shape (n, 4).
+        eps (float): Eps to avoid log(0).
+    Return:
+        Tensor: Loss tensor.
+    """
+    # overlap
+    lt = torch.max(pred[:, :2], target[:, :2])
+    rb = torch.min(pred[:, 2:], target[:, 2:])
+    wh = (rb - lt).clamp(min=0)
+    overlap = wh[:, 0] * wh[:, 1]
+
+    # union
+    ap = (pred[:, 2] - pred[:, 0]) * (pred[:, 3] - pred[:, 1])
+    ag = (target[:, 2] - target[:, 0]) * (target[:, 3] - target[:, 1])
+    union = ap + ag - overlap + eps
+
+    # IoU
+    ious = overlap / union
+
+    # enclose area
+    enclose_x1y1 = torch.min(pred[:, :2], target[:, :2])
+    enclose_x2y2 = torch.max(pred[:, 2:], target[:, 2:])
+    enclose_wh = (enclose_x2y2 - enclose_x1y1).clamp(min=0)
+
+    cw = enclose_wh[:, 0]
+    ch = enclose_wh[:, 1]
+
+    c2 = cw**2 + ch**2 + eps
+
+    b1_x1, b1_y1 = pred[:, 0], pred[:, 1]
+    b1_x2, b1_y2 = pred[:, 2], pred[:, 3]
+    b2_x1, b2_y1 = target[:, 0], target[:, 1]
+    b2_x2, b2_y2 = target[:, 2], target[:, 3]
+
+    w1, h1 = b1_x2 - b1_x1, b1_y2 - b1_y1 + eps
+    w2, h2 = b2_x2 - b2_x1, b2_y2 - b2_y1 + eps
+
+    left = ((b2_x1 + b2_x2) - (b1_x1 + b1_x2))**2 / 4
+    right = ((b2_y1 + b2_y2) - (b1_y1 + b1_y2))**2 / 4
+    rho2 = left + right
+
+    factor = 4 / math.pi**2
+    v = factor * torch.pow(torch.atan(w2 / h2) - torch.atan(w1 / h1), 2)
+
+    # CIoU
+    cious = ious - (rho2 / c2 + v**2 / (1 - ious + v))
+    loss = 1 - cious
+    return loss
+
+
+@LOSSES.register_module()
+class IoULoss(nn.Module):
+    """IoULoss.
+
+    Computing the IoU loss between a set of predicted bboxes and target bboxes.
+
+    Args:
+        linear (bool): If True, use linear scale of loss instead of log scale.
+            Default: False.
+        eps (float): Eps to avoid log(0).
+        reduction (str): Options are "none", "mean" and "sum".
+        loss_weight (float): Weight of loss.
+    """
+
+    def __init__(self,
+                 linear=False,
+                 eps=1e-6,
+                 reduction='mean',
+                 loss_weight=1.0):
+        super(IoULoss, self).__init__()
+        self.linear = linear
+        self.eps = eps
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): The prediction.
+            target (torch.Tensor): The learning target of the prediction.
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Defaults to None. Options are "none", "mean" and "sum".
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if (weight is not None) and (not torch.any(weight > 0)) and (
+                reduction != 'none'):
+            return (pred * weight).sum()  # 0
+        if weight is not None and weight.dim() > 1:
+            # TODO: remove this in the future
+            # reduce the weight of shape (n, 4) to (n,) to match the
+            # iou_loss of shape (n,)
+            assert weight.shape == pred.shape
+            weight = weight.mean(-1)
+        loss = self.loss_weight * iou_loss(
+            pred,
+            target,
+            weight,
+            linear=self.linear,
+            eps=self.eps,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss
+
+
+@LOSSES.register_module()
+class BoundedIoULoss(nn.Module):
+
+    def __init__(self, beta=0.2, eps=1e-3, reduction='mean', loss_weight=1.0):
+        super(BoundedIoULoss, self).__init__()
+        self.beta = beta
+        self.eps = eps
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        if weight is not None and not torch.any(weight > 0):
+            return (pred * weight).sum()  # 0
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        loss = self.loss_weight * bounded_iou_loss(
+            pred,
+            target,
+            weight,
+            beta=self.beta,
+            eps=self.eps,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss
+
+
+@LOSSES.register_module()
+class GIoULoss(nn.Module):
+
+    def __init__(self, eps=1e-6, reduction='mean', loss_weight=1.0):
+        super(GIoULoss, self).__init__()
+        self.eps = eps
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        if weight is not None and not torch.any(weight > 0):
+            return (pred * weight).sum()  # 0
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if weight is not None and weight.dim() > 1:
+            # TODO: remove this in the future
+            # reduce the weight of shape (n, 4) to (n,) to match the
+            # giou_loss of shape (n,)
+            assert weight.shape == pred.shape
+            weight = weight.mean(-1)
+        loss = self.loss_weight * giou_loss(
+            pred,
+            target,
+            weight,
+            eps=self.eps,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss
+
+
+@LOSSES.register_module()
+class RPDGIoULoss(nn.Module):
+
+    def __init__(self, eps=1e-6, reduction='mean', loss_weight=1.0):
+        super(RPDGIoULoss, self).__init__()
+        self.eps = eps
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        if weight is not None and not torch.any(weight > 0):
+            return (pred * weight).sum()  # 0
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if weight is not None and weight.dim() > 1:
+            # TODO: remove this in the future
+            # reduce the weight of shape (n, 4) to (n,) to match the
+            # giou_loss of shape (n,)
+            assert weight.shape == pred.shape
+            weight = weight.mean(-1)
+        loss = self.loss_weight * giou_loss_oldver(
+            pred,
+            target,
+            weight,
+            eps=self.eps,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss
+
+
+@LOSSES.register_module()
+class DIoULoss(nn.Module):
+
+    def __init__(self, eps=1e-6, reduction='mean', loss_weight=1.0):
+        super(DIoULoss, self).__init__()
+        self.eps = eps
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        if weight is not None and not torch.any(weight > 0):
+            return (pred * weight).sum()  # 0
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if weight is not None and weight.dim() > 1:
+            # TODO: remove this in the future
+            # reduce the weight of shape (n, 4) to (n,) to match the
+            # giou_loss of shape (n,)
+            assert weight.shape == pred.shape
+            weight = weight.mean(-1)
+        loss = self.loss_weight * diou_loss(
+            pred,
+            target,
+            weight,
+            eps=self.eps,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss
+
+
+@LOSSES.register_module()
+class CIoULoss(nn.Module):
+
+    def __init__(self, eps=1e-6, reduction='mean', loss_weight=1.0):
+        super(CIoULoss, self).__init__()
+        self.eps = eps
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        if weight is not None and not torch.any(weight > 0):
+            return (pred * weight).sum()  # 0
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if weight is not None and weight.dim() > 1:
+            # TODO: remove this in the future
+            # reduce the weight of shape (n, 4) to (n,) to match the
+            # giou_loss of shape (n,)
+            assert weight.shape == pred.shape
+            weight = weight.mean(-1)
+        loss = self.loss_weight * ciou_loss(
+            pred,
+            target,
+            weight,
+            eps=self.eps,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss
diff --git a/python/autoCut/models/swin/mmdet/models/losses/kd_loss.py b/python/autoCut/models/swin/mmdet/models/losses/kd_loss.py
new file mode 100644
index 0000000..f3abb68
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/kd_loss.py
@@ -0,0 +1,87 @@
+import mmcv
+import torch.nn as nn
+import torch.nn.functional as F
+
+from ..builder import LOSSES
+from .utils import weighted_loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def knowledge_distillation_kl_div_loss(pred,
+                                       soft_label,
+                                       T,
+                                       detach_target=True):
+    r"""Loss function for knowledge distilling using KL divergence.
+
+    Args:
+        pred (Tensor): Predicted logits with shape (N, n + 1).
+        soft_label (Tensor): Target logits with shape (N, N + 1).
+        T (int): Temperature for distillation.
+        detach_target (bool): Remove soft_label from automatic differentiation
+
+    Returns:
+        torch.Tensor: Loss tensor with shape (N,).
+    """
+    assert pred.size() == soft_label.size()
+    target = F.softmax(soft_label / T, dim=1)
+    if detach_target:
+        target = target.detach()
+
+    kd_loss = F.kl_div(
+        F.log_softmax(pred / T, dim=1), target, reduction='none').mean(1) * (
+            T * T)
+
+    return kd_loss
+
+
+@LOSSES.register_module()
+class KnowledgeDistillationKLDivLoss(nn.Module):
+    """Loss function for knowledge distilling using KL divergence.
+
+    Args:
+        reduction (str): Options are `'none'`, `'mean'` and `'sum'`.
+        loss_weight (float): Loss weight of current loss.
+        T (int): Temperature for distillation.
+    """
+
+    def __init__(self, reduction='mean', loss_weight=1.0, T=10):
+        super(KnowledgeDistillationKLDivLoss, self).__init__()
+        assert T >= 1
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+        self.T = T
+
+    def forward(self,
+                pred,
+                soft_label,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None):
+        """Forward function.
+
+        Args:
+            pred (Tensor): Predicted logits with shape (N, n + 1).
+            soft_label (Tensor): Target logits with shape (N, N + 1).
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Defaults to None.
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+
+        loss_kd = self.loss_weight * knowledge_distillation_kl_div_loss(
+            pred,
+            soft_label,
+            weight,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            T=self.T)
+
+        return loss_kd
diff --git a/python/autoCut/models/swin/mmdet/models/losses/mse_loss.py b/python/autoCut/models/swin/mmdet/models/losses/mse_loss.py
new file mode 100644
index 0000000..68d0575
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/mse_loss.py
@@ -0,0 +1,49 @@
+import torch.nn as nn
+import torch.nn.functional as F
+
+from ..builder import LOSSES
+from .utils import weighted_loss
+
+
+@weighted_loss
+def mse_loss(pred, target):
+    """Warpper of mse loss."""
+    return F.mse_loss(pred, target, reduction='none')
+
+
+@LOSSES.register_module()
+class MSELoss(nn.Module):
+    """MSELoss.
+
+    Args:
+        reduction (str, optional): The method that reduces the loss to a
+            scalar. Options are "none", "mean" and "sum".
+        loss_weight (float, optional): The weight of the loss. Defaults to 1.0
+    """
+
+    def __init__(self, reduction='mean', loss_weight=1.0):
+        super().__init__()
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self, pred, target, weight=None, avg_factor=None):
+        """Forward function of loss.
+
+        Args:
+            pred (torch.Tensor): The prediction.
+            target (torch.Tensor): The learning target of the prediction.
+            weight (torch.Tensor, optional): Weight of the loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+
+        Returns:
+            torch.Tensor: The calculated loss
+        """
+        loss = self.loss_weight * mse_loss(
+            pred,
+            target,
+            weight,
+            reduction=self.reduction,
+            avg_factor=avg_factor)
+        return loss
diff --git a/python/autoCut/models/swin/mmdet/models/losses/pisa_loss.py b/python/autoCut/models/swin/mmdet/models/losses/pisa_loss.py
new file mode 100644
index 0000000..4a48adf
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/pisa_loss.py
@@ -0,0 +1,183 @@
+import mmcv
+import torch
+
+from mmdet.core import bbox_overlaps
+
+
+@mmcv.jit(derivate=True, coderize=True)
+def isr_p(cls_score,
+          bbox_pred,
+          bbox_targets,
+          rois,
+          sampling_results,
+          loss_cls,
+          bbox_coder,
+          k=2,
+          bias=0,
+          num_class=80):
+    """Importance-based Sample Reweighting (ISR_P), positive part.
+
+    Args:
+        cls_score (Tensor): Predicted classification scores.
+        bbox_pred (Tensor): Predicted bbox deltas.
+        bbox_targets (tuple[Tensor]): A tuple of bbox targets, the are
+            labels, label_weights, bbox_targets, bbox_weights, respectively.
+        rois (Tensor): Anchors (single_stage) in shape (n, 4) or RoIs
+            (two_stage) in shape (n, 5).
+        sampling_results (obj): Sampling results.
+        loss_cls (func): Classification loss func of the head.
+        bbox_coder (obj): BBox coder of the head.
+        k (float): Power of the non-linear mapping.
+        bias (float): Shift of the non-linear mapping.
+        num_class (int): Number of classes, default: 80.
+
+    Return:
+        tuple([Tensor]): labels, imp_based_label_weights, bbox_targets,
+            bbox_target_weights
+    """
+
+    labels, label_weights, bbox_targets, bbox_weights = bbox_targets
+    pos_label_inds = ((labels >= 0) &
+                      (labels < num_class)).nonzero().reshape(-1)
+    pos_labels = labels[pos_label_inds]
+
+    # if no positive samples, return the original targets
+    num_pos = float(pos_label_inds.size(0))
+    if num_pos == 0:
+        return labels, label_weights, bbox_targets, bbox_weights
+
+    # merge pos_assigned_gt_inds of per image to a single tensor
+    gts = list()
+    last_max_gt = 0
+    for i in range(len(sampling_results)):
+        gt_i = sampling_results[i].pos_assigned_gt_inds
+        gts.append(gt_i + last_max_gt)
+        if len(gt_i) != 0:
+            last_max_gt = gt_i.max() + 1
+    gts = torch.cat(gts)
+    assert len(gts) == num_pos
+
+    cls_score = cls_score.detach()
+    bbox_pred = bbox_pred.detach()
+
+    # For single stage detectors, rois here indicate anchors, in shape (N, 4)
+    # For two stage detectors, rois are in shape (N, 5)
+    if rois.size(-1) == 5:
+        pos_rois = rois[pos_label_inds][:, 1:]
+    else:
+        pos_rois = rois[pos_label_inds]
+
+    if bbox_pred.size(-1) > 4:
+        bbox_pred = bbox_pred.view(bbox_pred.size(0), -1, 4)
+        pos_delta_pred = bbox_pred[pos_label_inds, pos_labels].view(-1, 4)
+    else:
+        pos_delta_pred = bbox_pred[pos_label_inds].view(-1, 4)
+
+    # compute iou of the predicted bbox and the corresponding GT
+    pos_delta_target = bbox_targets[pos_label_inds].view(-1, 4)
+    pos_bbox_pred = bbox_coder.decode(pos_rois, pos_delta_pred)
+    target_bbox_pred = bbox_coder.decode(pos_rois, pos_delta_target)
+    ious = bbox_overlaps(pos_bbox_pred, target_bbox_pred, is_aligned=True)
+
+    pos_imp_weights = label_weights[pos_label_inds]
+    # Two steps to compute IoU-HLR. Samples are first sorted by IoU locally,
+    # then sorted again within the same-rank group
+    max_l_num = pos_labels.bincount().max()
+    for label in pos_labels.unique():
+        l_inds = (pos_labels == label).nonzero().view(-1)
+        l_gts = gts[l_inds]
+        for t in l_gts.unique():
+            t_inds = l_inds[l_gts == t]
+            t_ious = ious[t_inds]
+            _, t_iou_rank_idx = t_ious.sort(descending=True)
+            _, t_iou_rank = t_iou_rank_idx.sort()
+            ious[t_inds] += max_l_num - t_iou_rank.float()
+        l_ious = ious[l_inds]
+        _, l_iou_rank_idx = l_ious.sort(descending=True)
+        _, l_iou_rank = l_iou_rank_idx.sort()  # IoU-HLR
+        # linearly map HLR to label weights
+        pos_imp_weights[l_inds] *= (max_l_num - l_iou_rank.float()) / max_l_num
+
+    pos_imp_weights = (bias + pos_imp_weights * (1 - bias)).pow(k)
+
+    # normalize to make the new weighted loss value equal to the original loss
+    pos_loss_cls = loss_cls(
+        cls_score[pos_label_inds], pos_labels, reduction_override='none')
+    if pos_loss_cls.dim() > 1:
+        ori_pos_loss_cls = pos_loss_cls * label_weights[pos_label_inds][:,
+                                                                        None]
+        new_pos_loss_cls = pos_loss_cls * pos_imp_weights[:, None]
+    else:
+        ori_pos_loss_cls = pos_loss_cls * label_weights[pos_label_inds]
+        new_pos_loss_cls = pos_loss_cls * pos_imp_weights
+    pos_loss_cls_ratio = ori_pos_loss_cls.sum() / new_pos_loss_cls.sum()
+    pos_imp_weights = pos_imp_weights * pos_loss_cls_ratio
+    label_weights[pos_label_inds] = pos_imp_weights
+
+    bbox_targets = labels, label_weights, bbox_targets, bbox_weights
+    return bbox_targets
+
+
+@mmcv.jit(derivate=True, coderize=True)
+def carl_loss(cls_score,
+              labels,
+              bbox_pred,
+              bbox_targets,
+              loss_bbox,
+              k=1,
+              bias=0.2,
+              avg_factor=None,
+              sigmoid=False,
+              num_class=80):
+    """Classification-Aware Regression Loss (CARL).
+
+    Args:
+        cls_score (Tensor): Predicted classification scores.
+        labels (Tensor): Targets of classification.
+        bbox_pred (Tensor): Predicted bbox deltas.
+        bbox_targets (Tensor): Target of bbox regression.
+        loss_bbox (func): Regression loss func of the head.
+        bbox_coder (obj): BBox coder of the head.
+        k (float): Power of the non-linear mapping.
+        bias (float): Shift of the non-linear mapping.
+        avg_factor (int): Average factor used in regression loss.
+        sigmoid (bool): Activation of the classification score.
+        num_class (int): Number of classes, default: 80.
+
+    Return:
+        dict: CARL loss dict.
+    """
+    pos_label_inds = ((labels >= 0) &
+                      (labels < num_class)).nonzero().reshape(-1)
+    if pos_label_inds.numel() == 0:
+        return dict(loss_carl=cls_score.sum()[None] * 0.)
+    pos_labels = labels[pos_label_inds]
+
+    # multiply pos_cls_score with the corresponding bbox weight
+    # and remain gradient
+    if sigmoid:
+        pos_cls_score = cls_score.sigmoid()[pos_label_inds, pos_labels]
+    else:
+        pos_cls_score = cls_score.softmax(-1)[pos_label_inds, pos_labels]
+    carl_loss_weights = (bias + (1 - bias) * pos_cls_score).pow(k)
+
+    # normalize carl_loss_weight to make its sum equal to num positive
+    num_pos = float(pos_cls_score.size(0))
+    weight_ratio = num_pos / carl_loss_weights.sum()
+    carl_loss_weights *= weight_ratio
+
+    if avg_factor is None:
+        avg_factor = bbox_targets.size(0)
+    # if is class agnostic, bbox pred is in shape (N, 4)
+    # otherwise, bbox pred is in shape (N, #classes, 4)
+    if bbox_pred.size(-1) > 4:
+        bbox_pred = bbox_pred.view(bbox_pred.size(0), -1, 4)
+        pos_bbox_preds = bbox_pred[pos_label_inds, pos_labels]
+    else:
+        pos_bbox_preds = bbox_pred[pos_label_inds]
+    ori_loss_reg = loss_bbox(
+        pos_bbox_preds,
+        bbox_targets[pos_label_inds],
+        reduction_override='none') / avg_factor
+    loss_carl = (ori_loss_reg * carl_loss_weights[:, None]).sum()
+    return dict(loss_carl=loss_carl[None])
diff --git a/python/autoCut/models/swin/mmdet/models/losses/smooth_l1_loss.py b/python/autoCut/models/swin/mmdet/models/losses/smooth_l1_loss.py
new file mode 100644
index 0000000..ec9c98a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/smooth_l1_loss.py
@@ -0,0 +1,139 @@
+import mmcv
+import torch
+import torch.nn as nn
+
+from ..builder import LOSSES
+from .utils import weighted_loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def smooth_l1_loss(pred, target, beta=1.0):
+    """Smooth L1 loss.
+
+    Args:
+        pred (torch.Tensor): The prediction.
+        target (torch.Tensor): The learning target of the prediction.
+        beta (float, optional): The threshold in the piecewise function.
+            Defaults to 1.0.
+
+    Returns:
+        torch.Tensor: Calculated loss
+    """
+    assert beta > 0
+    assert pred.size() == target.size() and target.numel() > 0
+    diff = torch.abs(pred - target)
+    loss = torch.where(diff < beta, 0.5 * diff * diff / beta,
+                       diff - 0.5 * beta)
+    return loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+@weighted_loss
+def l1_loss(pred, target):
+    """L1 loss.
+
+    Args:
+        pred (torch.Tensor): The prediction.
+        target (torch.Tensor): The learning target of the prediction.
+
+    Returns:
+        torch.Tensor: Calculated loss
+    """
+    assert pred.size() == target.size() and target.numel() > 0
+    loss = torch.abs(pred - target)
+    return loss
+
+
+@LOSSES.register_module()
+class SmoothL1Loss(nn.Module):
+    """Smooth L1 loss.
+
+    Args:
+        beta (float, optional): The threshold in the piecewise function.
+            Defaults to 1.0.
+        reduction (str, optional): The method to reduce the loss.
+            Options are "none", "mean" and "sum". Defaults to "mean".
+        loss_weight (float, optional): The weight of loss.
+    """
+
+    def __init__(self, beta=1.0, reduction='mean', loss_weight=1.0):
+        super(SmoothL1Loss, self).__init__()
+        self.beta = beta
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None,
+                **kwargs):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): The prediction.
+            target (torch.Tensor): The learning target of the prediction.
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Defaults to None.
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        loss_bbox = self.loss_weight * smooth_l1_loss(
+            pred,
+            target,
+            weight,
+            beta=self.beta,
+            reduction=reduction,
+            avg_factor=avg_factor,
+            **kwargs)
+        return loss_bbox
+
+
+@LOSSES.register_module()
+class L1Loss(nn.Module):
+    """L1 loss.
+
+    Args:
+        reduction (str, optional): The method to reduce the loss.
+            Options are "none", "mean" and "sum".
+        loss_weight (float, optional): The weight of loss.
+    """
+
+    def __init__(self, reduction='mean', loss_weight=1.0):
+        super(L1Loss, self).__init__()
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): The prediction.
+            target (torch.Tensor): The learning target of the prediction.
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Defaults to None.
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        loss_bbox = self.loss_weight * l1_loss(
+            pred, target, weight, reduction=reduction, avg_factor=avg_factor)
+        return loss_bbox
diff --git a/python/autoCut/models/swin/mmdet/models/losses/utils.py b/python/autoCut/models/swin/mmdet/models/losses/utils.py
new file mode 100644
index 0000000..4756d7f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/utils.py
@@ -0,0 +1,100 @@
+import functools
+
+import mmcv
+import torch.nn.functional as F
+
+
+def reduce_loss(loss, reduction):
+    """Reduce loss as specified.
+
+    Args:
+        loss (Tensor): Elementwise loss tensor.
+        reduction (str): Options are "none", "mean" and "sum".
+
+    Return:
+        Tensor: Reduced loss tensor.
+    """
+    reduction_enum = F._Reduction.get_enum(reduction)
+    # none: 0, elementwise_mean:1, sum: 2
+    if reduction_enum == 0:
+        return loss
+    elif reduction_enum == 1:
+        return loss.mean()
+    elif reduction_enum == 2:
+        return loss.sum()
+
+
+@mmcv.jit(derivate=True, coderize=True)
+def weight_reduce_loss(loss, weight=None, reduction='mean', avg_factor=None):
+    """Apply element-wise weight and reduce loss.
+
+    Args:
+        loss (Tensor): Element-wise loss.
+        weight (Tensor): Element-wise weights.
+        reduction (str): Same as built-in losses of PyTorch.
+        avg_factor (float): Avarage factor when computing the mean of losses.
+
+    Returns:
+        Tensor: Processed loss values.
+    """
+    # if weight is specified, apply element-wise weight
+    if weight is not None:
+        loss = loss * weight
+
+    # if avg_factor is not specified, just reduce the loss
+    if avg_factor is None:
+        loss = reduce_loss(loss, reduction)
+    else:
+        # if reduction is mean, then average the loss by avg_factor
+        if reduction == 'mean':
+            loss = loss.sum() / avg_factor
+        # if reduction is 'none', then do nothing, otherwise raise an error
+        elif reduction != 'none':
+            raise ValueError('avg_factor can not be used with reduction="sum"')
+    return loss
+
+
+def weighted_loss(loss_func):
+    """Create a weighted version of a given loss function.
+
+    To use this decorator, the loss function must have the signature like
+    `loss_func(pred, target, **kwargs)`. The function only needs to compute
+    element-wise loss without any reduction. This decorator will add weight
+    and reduction arguments to the function. The decorated function will have
+    the signature like `loss_func(pred, target, weight=None, reduction='mean',
+    avg_factor=None, **kwargs)`.
+
+    :Example:
+
+    >>> import torch
+    >>> @weighted_loss
+    >>> def l1_loss(pred, target):
+    >>>     return (pred - target).abs()
+
+    >>> pred = torch.Tensor([0, 2, 3])
+    >>> target = torch.Tensor([1, 1, 1])
+    >>> weight = torch.Tensor([1, 0, 1])
+
+    >>> l1_loss(pred, target)
+    tensor(1.3333)
+    >>> l1_loss(pred, target, weight)
+    tensor(1.)
+    >>> l1_loss(pred, target, reduction='none')
+    tensor([1., 1., 2.])
+    >>> l1_loss(pred, target, weight, avg_factor=2)
+    tensor(1.5000)
+    """
+
+    @functools.wraps(loss_func)
+    def wrapper(pred,
+                target,
+                weight=None,
+                reduction='mean',
+                avg_factor=None,
+                **kwargs):
+        # get element-wise loss
+        loss = loss_func(pred, target, **kwargs)
+        loss = weight_reduce_loss(loss, weight, reduction, avg_factor)
+        return loss
+
+    return wrapper
diff --git a/python/autoCut/models/swin/mmdet/models/losses/varifocal_loss.py b/python/autoCut/models/swin/mmdet/models/losses/varifocal_loss.py
new file mode 100644
index 0000000..7f00bd6
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/losses/varifocal_loss.py
@@ -0,0 +1,133 @@
+import mmcv
+import torch.nn as nn
+import torch.nn.functional as F
+
+from ..builder import LOSSES
+from .utils import weight_reduce_loss
+
+
+@mmcv.jit(derivate=True, coderize=True)
+def varifocal_loss(pred,
+                   target,
+                   weight=None,
+                   alpha=0.75,
+                   gamma=2.0,
+                   iou_weighted=True,
+                   reduction='mean',
+                   avg_factor=None):
+    """`Varifocal Loss <https://arxiv.org/abs/2008.13367>`_
+
+    Args:
+        pred (torch.Tensor): The prediction with shape (N, C), C is the
+            number of classes
+        target (torch.Tensor): The learning target of the iou-aware
+            classification score with shape (N, C), C is the number of classes.
+        weight (torch.Tensor, optional): The weight of loss for each
+            prediction. Defaults to None.
+        alpha (float, optional): A balance factor for the negative part of
+            Varifocal Loss, which is different from the alpha of Focal Loss.
+            Defaults to 0.75.
+        gamma (float, optional): The gamma for calculating the modulating
+            factor. Defaults to 2.0.
+        iou_weighted (bool, optional): Whether to weight the loss of the
+            positive example with the iou target. Defaults to True.
+        reduction (str, optional): The method used to reduce the loss into
+            a scalar. Defaults to 'mean'. Options are "none", "mean" and
+            "sum".
+        avg_factor (int, optional): Average factor that is used to average
+            the loss. Defaults to None.
+    """
+    # pred and target should be of the same size
+    assert pred.size() == target.size()
+    pred_sigmoid = pred.sigmoid()
+    target = target.type_as(pred)
+    if iou_weighted:
+        focal_weight = target * (target > 0.0).float() + \
+            alpha * (pred_sigmoid - target).abs().pow(gamma) * \
+            (target <= 0.0).float()
+    else:
+        focal_weight = (target > 0.0).float() + \
+            alpha * (pred_sigmoid - target).abs().pow(gamma) * \
+            (target <= 0.0).float()
+    loss = F.binary_cross_entropy_with_logits(
+        pred, target, reduction='none') * focal_weight
+    loss = weight_reduce_loss(loss, weight, reduction, avg_factor)
+    return loss
+
+
+@LOSSES.register_module()
+class VarifocalLoss(nn.Module):
+
+    def __init__(self,
+                 use_sigmoid=True,
+                 alpha=0.75,
+                 gamma=2.0,
+                 iou_weighted=True,
+                 reduction='mean',
+                 loss_weight=1.0):
+        """`Varifocal Loss <https://arxiv.org/abs/2008.13367>`_
+
+        Args:
+            use_sigmoid (bool, optional): Whether the prediction is
+                used for sigmoid or softmax. Defaults to True.
+            alpha (float, optional): A balance factor for the negative part of
+                Varifocal Loss, which is different from the alpha of Focal
+                Loss. Defaults to 0.75.
+            gamma (float, optional): The gamma for calculating the modulating
+                factor. Defaults to 2.0.
+            iou_weighted (bool, optional): Whether to weight the loss of the
+                positive examples with the iou target. Defaults to True.
+            reduction (str, optional): The method used to reduce the loss into
+                a scalar. Defaults to 'mean'. Options are "none", "mean" and
+                "sum".
+            loss_weight (float, optional): Weight of loss. Defaults to 1.0.
+        """
+        super(VarifocalLoss, self).__init__()
+        assert use_sigmoid is True, \
+            'Only sigmoid varifocal loss supported now.'
+        assert alpha >= 0.0
+        self.use_sigmoid = use_sigmoid
+        self.alpha = alpha
+        self.gamma = gamma
+        self.iou_weighted = iou_weighted
+        self.reduction = reduction
+        self.loss_weight = loss_weight
+
+    def forward(self,
+                pred,
+                target,
+                weight=None,
+                avg_factor=None,
+                reduction_override=None):
+        """Forward function.
+
+        Args:
+            pred (torch.Tensor): The prediction.
+            target (torch.Tensor): The learning target of the prediction.
+            weight (torch.Tensor, optional): The weight of loss for each
+                prediction. Defaults to None.
+            avg_factor (int, optional): Average factor that is used to average
+                the loss. Defaults to None.
+            reduction_override (str, optional): The reduction method used to
+                override the original reduction method of the loss.
+                Options are "none", "mean" and "sum".
+
+        Returns:
+            torch.Tensor: The calculated loss
+        """
+        assert reduction_override in (None, 'none', 'mean', 'sum')
+        reduction = (
+            reduction_override if reduction_override else self.reduction)
+        if self.use_sigmoid:
+            loss_cls = self.loss_weight * varifocal_loss(
+                pred,
+                target,
+                weight,
+                alpha=self.alpha,
+                gamma=self.gamma,
+                iou_weighted=self.iou_weighted,
+                reduction=reduction,
+                avg_factor=avg_factor)
+        else:
+            raise NotImplementedError
+        return loss_cls
diff --git a/python/autoCut/models/swin/mmdet/models/necks/__init__.py b/python/autoCut/models/swin/mmdet/models/necks/__init__.py
new file mode 100644
index 0000000..291262b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/__init__.py
@@ -0,0 +1,17 @@
+from .bfp import BFP
+from .channel_mapper import ChannelMapper
+from .fpg import FPG
+from .fpn import FPN
+from .fpn_carafe import FPN_CARAFE
+from .hrfpn import HRFPN
+from .nas_fpn import NASFPN
+from .nasfcos_fpn import NASFCOS_FPN
+from .pafpn import PAFPN
+from .rfp import RFP
+from .bifpn import BiFPN
+from .yolo_neck import YOLOV3Neck
+
+__all__ = [
+    'FPN', 'BFP', 'ChannelMapper', 'HRFPN', 'NASFPN', 'FPN_CARAFE', 'PAFPN',
+    'NASFCOS_FPN', 'RFP', 'YOLOV3Neck', 'FPG', 'BiFPN'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/necks/bfp.py b/python/autoCut/models/swin/mmdet/models/necks/bfp.py
new file mode 100644
index 0000000..123f551
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/bfp.py
@@ -0,0 +1,104 @@
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, xavier_init
+from mmcv.cnn.bricks import NonLocal2d
+
+from ..builder import NECKS
+
+
+@NECKS.register_module()
+class BFP(nn.Module):
+    """BFP (Balanced Feature Pyramids)
+
+    BFP takes multi-level features as inputs and gather them into a single one,
+    then refine the gathered feature and scatter the refined results to
+    multi-level features. This module is used in Libra R-CNN (CVPR 2019), see
+    the paper `Libra R-CNN: Towards Balanced Learning for Object Detection
+    <https://arxiv.org/abs/1904.02701>`_ for details.
+
+    Args:
+        in_channels (int): Number of input channels (feature maps of all levels
+            should have the same channels).
+        num_levels (int): Number of input feature levels.
+        conv_cfg (dict): The config dict for convolution layers.
+        norm_cfg (dict): The config dict for normalization layers.
+        refine_level (int): Index of integration and refine level of BSF in
+            multi-level features from bottom to top.
+        refine_type (str): Type of the refine op, currently support
+            [None, 'conv', 'non_local'].
+    """
+
+    def __init__(self,
+                 in_channels,
+                 num_levels,
+                 refine_level=2,
+                 refine_type=None,
+                 conv_cfg=None,
+                 norm_cfg=None):
+        super(BFP, self).__init__()
+        assert refine_type in [None, 'conv', 'non_local']
+
+        self.in_channels = in_channels
+        self.num_levels = num_levels
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+
+        self.refine_level = refine_level
+        self.refine_type = refine_type
+        assert 0 <= self.refine_level < self.num_levels
+
+        if self.refine_type == 'conv':
+            self.refine = ConvModule(
+                self.in_channels,
+                self.in_channels,
+                3,
+                padding=1,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg)
+        elif self.refine_type == 'non_local':
+            self.refine = NonLocal2d(
+                self.in_channels,
+                reduction=1,
+                use_scale=False,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg)
+
+    def init_weights(self):
+        """Initialize the weights of FPN module."""
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                xavier_init(m, distribution='uniform')
+
+    def forward(self, inputs):
+        """Forward function."""
+        assert len(inputs) == self.num_levels
+
+        # step 1: gather multi-level features by resize and average
+        feats = []
+        gather_size = inputs[self.refine_level].size()[2:]
+        for i in range(self.num_levels):
+            if i < self.refine_level:
+                gathered = F.adaptive_max_pool2d(
+                    inputs[i], output_size=gather_size)
+            else:
+                gathered = F.interpolate(
+                    inputs[i], size=gather_size, mode='nearest')
+            feats.append(gathered)
+
+        bsf = sum(feats) / len(feats)
+
+        # step 2: refine gathered features
+        if self.refine_type is not None:
+            bsf = self.refine(bsf)
+
+        # step 3: scatter refined features to multi-levels by a residual path
+        outs = []
+        for i in range(self.num_levels):
+            out_size = inputs[i].size()[2:]
+            if i < self.refine_level:
+                residual = F.interpolate(bsf, size=out_size, mode='nearest')
+            else:
+                residual = F.adaptive_max_pool2d(bsf, output_size=out_size)
+            outs.append(residual + inputs[i])
+
+        return tuple(outs)
diff --git a/python/autoCut/models/swin/mmdet/models/necks/bifpn.py b/python/autoCut/models/swin/mmdet/models/necks/bifpn.py
new file mode 100644
index 0000000..34f3ea1
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/bifpn.py
@@ -0,0 +1,240 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, xavier_init
+from mmcv.runner import auto_fp16
+
+from ..builder import NECKS
+
+
+def swish(x):
+    return x * x.sigmoid()
+
+
+class LayerCombineModule(nn.Module):
+    def __init__(self, num_input=2):
+        super().__init__()
+        self.weights = nn.Parameter(
+            torch.ones(num_input, dtype=torch.float32).view(1, 1, 1, 1, -1),
+            requires_grad=True
+        )
+
+    def forward(self, inputs):
+
+        weights = self.weights.relu()
+        norm_weights = weights / (weights.sum() + 0.0001)
+
+        out = (norm_weights*torch.stack(inputs, dim=-1)).sum(dim=-1)
+        return swish(out)
+
+class Identity(nn.Module):
+    def __init__(self):
+        super().__init__()
+    
+    def forward(self,x):
+        return x 
+
+class SingleBiFPN(nn.Module):
+    def __init__(self, in_channels, out_channels, no_norm_on_lateral=True, conv_cfg=None,
+                 norm_cfg=None,
+                 act_cfg=None,
+                 upsample_cfg=dict(mode='nearest')):
+        super().__init__()
+
+        self.no_norm_on_lateral = no_norm_on_lateral
+        self.upsample_cfg = upsample_cfg
+
+        self.lateral_convs = nn.ModuleList()
+        self.lateral_combine = nn.ModuleList()
+        self.lateral_combine_conv = nn.ModuleList()
+        self.out_combine = nn.ModuleList()
+        self.out_combine_conv = nn.ModuleList()
+
+        for i, in_channel in enumerate(in_channels):
+            if in_channel != out_channels:
+                self.lateral_convs.append(ConvModule(
+                    in_channel,
+                    out_channels,
+                    1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg if not self.no_norm_on_lateral else None,
+                    act_cfg=act_cfg,
+                    inplace=False))
+            else:
+                self.lateral_convs.append(Identity())
+            if i != len(in_channels)-1:
+                self.lateral_combine.append(LayerCombineModule(2))
+                self.lateral_combine_conv.append(ConvModule(
+                    out_channels,
+                    out_channels,
+                    3,
+                    padding=1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg if not self.no_norm_on_lateral else None,
+                    act_cfg=None,
+                    inplace=False)
+                )
+            if i != 0:
+                self.out_combine.append(LayerCombineModule(
+                    3 if i != len(in_channels)-1 else 2))
+                self.out_combine_conv.append(ConvModule(
+                    out_channels,
+                    out_channels,
+                    3,
+                    padding=1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg if not self.no_norm_on_lateral else None,
+                    act_cfg=None,
+                    inplace=False))
+
+    def forward(self, inputs):
+
+        laterals = [
+            lateral_conv(inputs[i])
+            for i, lateral_conv in enumerate(self.lateral_convs)
+        ]
+        laterals = laterals + \
+            inputs[len(self.lateral_convs):]  # p3,p4,p5,p6,p7
+
+        # top to down
+        outs = [laterals[i] for i in range(len(laterals))]
+
+        for i in range(len(laterals)-1, 0, -1):
+            # In some cases, fixing `scale factor` (e.g. 2) is preferred, but
+            #  it cannot co-exist with `size` in `F.interpolate`.
+
+            if 'scale_factor' in self.upsample_cfg:
+                up_feat = F.interpolate(outs[i],
+                                        **self.upsample_cfg)
+            else:
+                prev_shape = outs[i-1].shape[2:]
+                up_feat = F.interpolate(
+                    outs[i], size=prev_shape, **self.upsample_cfg)
+            # weight combine
+            outs[i-1] = self.lateral_combine_conv[i -
+                                                  1](self.lateral_combine[i-1]([outs[i-1], up_feat]))
+
+        # down to top
+        for i in range(len(outs)-1):
+            # print(laterals[i].size())
+            down_feat = F.max_pool2d(outs[i], 3, stride=2, padding=1)
+            # print(down_feat.size())
+            cur_outs = outs[i+1]
+            if i != len(laterals)-2:
+                cur_inputs = laterals[i+1]
+                outs[i +
+                     1] = self.out_combine[i]([down_feat, cur_outs, cur_inputs])
+            else:
+                outs[i+1] = self.out_combine[i]([down_feat, cur_outs])
+            outs[i+1] = self.out_combine_conv[i](outs[i+1])
+
+        return outs
+
+
+@NECKS.register_module()
+class BiFPN(nn.Module):
+    def __init__(self,
+                 in_channels,
+                 out_channels=160,
+                 num_outs=5,
+                 start_level=0,
+                 end_level=-1,
+                 num_repeat=6,
+                 add_extra_convs=False,
+                 relu_before_extra_convs=False,
+                 no_norm_on_lateral=True,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 act_cfg=None,
+                 upsample_cfg=dict(mode='nearest')):
+        super(BiFPN, self).__init__()
+        assert isinstance(in_channels, list)
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.num_ins = len(in_channels)
+        self.num_outs = num_outs
+        self.num_repeat = num_repeat
+        self.relu_before_extra_convs = relu_before_extra_convs
+        self.no_norm_on_lateral = no_norm_on_lateral
+        self.fp16_enabled = False
+        self.upsample_cfg = upsample_cfg.copy()
+
+        if end_level == -1:
+            self.backbone_end_level = self.num_ins
+            assert num_outs >= self.num_ins - start_level
+        else:
+            # if end_level < inputs, no extra level is allowed
+            self.backbone_end_level = end_level
+            assert end_level <= len(in_channels)
+            assert num_outs == end_level - start_level
+        self.start_level = start_level
+        self.end_level = end_level
+        self.add_extra_convs = add_extra_convs
+
+        self.downsample_convs = nn.ModuleList()
+        # add extra conv layers (e.g., RetinaNet)
+        extra_levels = num_outs - self.backbone_end_level + self.start_level
+        if self.add_extra_convs and extra_levels >= 1:
+            for i in range(extra_levels):
+                if i == 0:
+                    in_channels = self.in_channels[self.backbone_end_level - 1]
+                else:
+                    in_channels = out_channels
+                extra_conv = nn.Sequential(
+                    ConvModule(
+                    in_channels,
+                    out_channels,
+                    1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg,
+                    act_cfg=act_cfg,
+                    inplace=False),
+                    nn.MaxPool2d(3,2,1)
+                    )
+                self.downsample_convs.append(extra_conv)
+
+        out_channels = out_channels if self.add_extra_convs else self.in_channels[
+            self.backbone_end_level-1]
+        self.bi_fpn = nn.ModuleList()
+        for i in range(self.num_repeat):
+            if i == 0:
+                in_channels = self.in_channels[self.start_level:self.backbone_end_level]+[
+                    out_channels]*extra_levels
+            else:
+                in_channels = [self.out_channels]*num_outs
+            self.bi_fpn.append(SingleBiFPN(in_channels, self.out_channels, no_norm_on_lateral=no_norm_on_lateral,
+                                           conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=act_cfg, upsample_cfg=upsample_cfg))
+
+    # default init_weights for conv(msra) and norm in ConvModule
+    def init_weights(self):
+        """Initialize the weights of FPN module"""
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                xavier_init(m, distribution='uniform')
+
+    def forward(self, inputs):
+        """Forward function"""
+        assert len(inputs) == len(self.in_channels)
+        # build laterals
+        outs = list(inputs[self.start_level:self.backbone_end_level])
+        used_backbone_levels = len(outs)
+        # part 2: add extra levels
+        if self.num_outs > len(outs):
+            # use max pool to get more levels on top of outputs
+            # (e.g., Faster R-CNN, Mask R-CNN)
+            if not self.add_extra_convs:
+                for i in range(self.num_outs - used_backbone_levels):
+                    outs.append(F.max_pool2d(outs[-1], 3, stride=2, padding=1))
+            # add conv layers on top of original feature maps (RetinaNet)
+            else:
+                for i in range(self.num_outs-used_backbone_levels):
+                    if self.relu_before_extra_convs:
+                        outs.append(self.downsample_convs[i](F.relu(outs[-1])))
+                    else:
+                        outs.append(self.downsample_convs[i](outs[-1]))
+
+        # p2,p3,p4,p5,p6,p7
+        # forward to bifpn
+        for i in range(self.num_repeat):
+            outs = self.bi_fpn[i](outs)
+        return tuple(outs)
diff --git a/python/autoCut/models/swin/mmdet/models/necks/channel_mapper.py b/python/autoCut/models/swin/mmdet/models/necks/channel_mapper.py
new file mode 100644
index 0000000..a4f5ed4
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/channel_mapper.py
@@ -0,0 +1,74 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, xavier_init
+
+from ..builder import NECKS
+
+
+@NECKS.register_module()
+class ChannelMapper(nn.Module):
+    r"""Channel Mapper to reduce/increase channels of backbone features.
+
+    This is used to reduce/increase channels of backbone features.
+
+    Args:
+        in_channels (List[int]): Number of input channels per scale.
+        out_channels (int): Number of output channels (used at each scale).
+        kernel_size (int, optional): kernel_size for reducing channels (used
+            at each scale). Default: 3.
+        conv_cfg (dict, optional): Config dict for convolution layer.
+            Default: None.
+        norm_cfg (dict, optional): Config dict for normalization layer.
+            Default: None.
+        act_cfg (dict, optional): Config dict for activation layer in
+            ConvModule. Default: dict(type='ReLU').
+
+    Example:
+        >>> import torch
+        >>> in_channels = [2, 3, 5, 7]
+        >>> scales = [340, 170, 84, 43]
+        >>> inputs = [torch.rand(1, c, s, s)
+        ...           for c, s in zip(in_channels, scales)]
+        >>> self = ChannelMapper(in_channels, 11, 3).eval()
+        >>> outputs = self.forward(inputs)
+        >>> for i in range(len(outputs)):
+        ...     print(f'outputs[{i}].shape = {outputs[i].shape}')
+        outputs[0].shape = torch.Size([1, 11, 340, 340])
+        outputs[1].shape = torch.Size([1, 11, 170, 170])
+        outputs[2].shape = torch.Size([1, 11, 84, 84])
+        outputs[3].shape = torch.Size([1, 11, 43, 43])
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 kernel_size=3,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 act_cfg=dict(type='ReLU')):
+        super(ChannelMapper, self).__init__()
+        assert isinstance(in_channels, list)
+
+        self.convs = nn.ModuleList()
+        for in_channel in in_channels:
+            self.convs.append(
+                ConvModule(
+                    in_channel,
+                    out_channels,
+                    kernel_size,
+                    padding=(kernel_size - 1) // 2,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg,
+                    act_cfg=act_cfg))
+
+    # default init_weights for conv(msra) and norm in ConvModule
+    def init_weights(self):
+        """Initialize the weights of ChannelMapper module."""
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                xavier_init(m, distribution='uniform')
+
+    def forward(self, inputs):
+        """Forward function."""
+        assert len(inputs) == len(self.convs)
+        outs = [self.convs[i](inputs[i]) for i in range(len(inputs))]
+        return tuple(outs)
diff --git a/python/autoCut/models/swin/mmdet/models/necks/fpg.py b/python/autoCut/models/swin/mmdet/models/necks/fpg.py
new file mode 100644
index 0000000..c8e0d16
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/fpg.py
@@ -0,0 +1,398 @@
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, caffe2_xavier_init, constant_init, is_norm
+
+from ..builder import NECKS
+
+
+class Transition(nn.Module):
+    """Base class for transition.
+
+    Args:
+        in_channels (int): Number of input channels.
+        out_channels (int): Number of output channels.
+    """
+
+    def __init__(self, in_channels, out_channels):
+        super().__init__()
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+
+    def forward(x):
+        pass
+
+
+class UpInterpolationConv(Transition):
+    """A transition used for up-sampling.
+
+    Up-sample the input by interpolation then refines the feature by
+    a convolution layer.
+
+    Args:
+        in_channels (int): Number of input channels.
+        out_channels (int): Number of output channels.
+        scale_factor (int): Up-sampling factor. Default: 2.
+        mode (int): Interpolation mode. Default: nearest.
+        align_corners (bool): Whether align corners when interpolation.
+            Default: None.
+        kernel_size (int): Kernel size for the conv. Default: 3.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 scale_factor=2,
+                 mode='nearest',
+                 align_corners=None,
+                 kernel_size=3,
+                 **kwargs):
+        super().__init__(in_channels, out_channels)
+        self.mode = mode
+        self.scale_factor = scale_factor
+        self.align_corners = align_corners
+        self.conv = ConvModule(
+            in_channels,
+            out_channels,
+            kernel_size,
+            padding=(kernel_size - 1) // 2,
+            **kwargs)
+
+    def forward(self, x):
+        x = F.interpolate(
+            x,
+            scale_factor=self.scale_factor,
+            mode=self.mode,
+            align_corners=self.align_corners)
+        x = self.conv(x)
+        return x
+
+
+class LastConv(Transition):
+    """A transition used for refining the output of the last stage.
+
+    Args:
+        in_channels (int): Number of input channels.
+        out_channels (int): Number of output channels.
+        num_inputs (int): Number of inputs of the FPN features.
+        kernel_size (int): Kernel size for the conv. Default: 3.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 num_inputs,
+                 kernel_size=3,
+                 **kwargs):
+        super().__init__(in_channels, out_channels)
+        self.num_inputs = num_inputs
+        self.conv_out = ConvModule(
+            in_channels,
+            out_channels,
+            kernel_size,
+            padding=(kernel_size - 1) // 2,
+            **kwargs)
+
+    def forward(self, inputs):
+        assert len(inputs) == self.num_inputs
+        return self.conv_out(inputs[-1])
+
+
+@NECKS.register_module()
+class FPG(nn.Module):
+    """FPG.
+
+    Implementation of `Feature Pyramid Grids (FPG)
+    <https://arxiv.org/abs/2004.03580>`_.
+    This implementation only gives the basic structure stated in the paper.
+    But users can implement different type of transitions to fully explore the
+    the potential power of the structure of FPG.
+
+    Args:
+        in_channels (int): Number of input channels (feature maps of all levels
+            should have the same channels).
+        out_channels (int): Number of output channels (used at each scale)
+        num_outs (int): Number of output scales.
+        stack_times (int): The number of times the pyramid architecture will
+            be stacked.
+        paths (list[str]): Specify the path order of each stack level.
+            Each element in the list should be either 'bu' (bottom-up) or
+            'td' (top-down).
+        inter_channels (int): Number of inter channels.
+        same_up_trans (dict): Transition that goes down at the same stage.
+        same_down_trans (dict): Transition that goes up at the same stage.
+        across_lateral_trans (dict): Across-pathway same-stage
+        across_down_trans (dict): Across-pathway bottom-up connection.
+        across_up_trans (dict): Across-pathway top-down connection.
+        across_skip_trans (dict): Across-pathway skip connection.
+        output_trans (dict): Transition that trans the output of the
+            last stage.
+        start_level (int): Index of the start input backbone level used to
+            build the feature pyramid. Default: 0.
+        end_level (int): Index of the end input backbone level (exclusive) to
+            build the feature pyramid. Default: -1, which means the last level.
+        add_extra_convs (bool): It decides whether to add conv
+            layers on top of the original feature maps. Default to False.
+            If True, its actual mode is specified by `extra_convs_on_inputs`.
+        norm_cfg (dict): Config dict for normalization layer. Default: None.
+    """
+
+    transition_types = {
+        'conv': ConvModule,
+        'interpolation_conv': UpInterpolationConv,
+        'last_conv': LastConv,
+    }
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 num_outs,
+                 stack_times,
+                 paths,
+                 inter_channels=None,
+                 same_down_trans=None,
+                 same_up_trans=dict(
+                     type='conv', kernel_size=3, stride=2, padding=1),
+                 across_lateral_trans=dict(type='conv', kernel_size=1),
+                 across_down_trans=dict(type='conv', kernel_size=3),
+                 across_up_trans=None,
+                 across_skip_trans=dict(type='identity'),
+                 output_trans=dict(type='last_conv', kernel_size=3),
+                 start_level=0,
+                 end_level=-1,
+                 add_extra_convs=False,
+                 norm_cfg=None,
+                 skip_inds=None):
+        super(FPG, self).__init__()
+        assert isinstance(in_channels, list)
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.num_ins = len(in_channels)
+        self.num_outs = num_outs
+        if inter_channels is None:
+            self.inter_channels = [out_channels for _ in range(num_outs)]
+        elif isinstance(inter_channels, int):
+            self.inter_channels = [inter_channels for _ in range(num_outs)]
+        else:
+            assert isinstance(inter_channels, list)
+            assert len(inter_channels) == num_outs
+            self.inter_channels = inter_channels
+        self.stack_times = stack_times
+        self.paths = paths
+        assert isinstance(paths, list) and len(paths) == stack_times
+        for d in paths:
+            assert d in ('bu', 'td')
+
+        self.same_down_trans = same_down_trans
+        self.same_up_trans = same_up_trans
+        self.across_lateral_trans = across_lateral_trans
+        self.across_down_trans = across_down_trans
+        self.across_up_trans = across_up_trans
+        self.output_trans = output_trans
+        self.across_skip_trans = across_skip_trans
+
+        self.with_bias = norm_cfg is None
+        # skip inds must be specified if across skip trans is not None
+        if self.across_skip_trans is not None:
+            skip_inds is not None
+        self.skip_inds = skip_inds
+        assert len(self.skip_inds[0]) <= self.stack_times
+
+        if end_level == -1:
+            self.backbone_end_level = self.num_ins
+            assert num_outs >= self.num_ins - start_level
+        else:
+            # if end_level < inputs, no extra level is allowed
+            self.backbone_end_level = end_level
+            assert end_level <= len(in_channels)
+            assert num_outs == end_level - start_level
+        self.start_level = start_level
+        self.end_level = end_level
+        self.add_extra_convs = add_extra_convs
+
+        # build lateral 1x1 convs to reduce channels
+        self.lateral_convs = nn.ModuleList()
+        for i in range(self.start_level, self.backbone_end_level):
+            l_conv = nn.Conv2d(self.in_channels[i],
+                               self.inter_channels[i - self.start_level], 1)
+            self.lateral_convs.append(l_conv)
+
+        extra_levels = num_outs - self.backbone_end_level + self.start_level
+        self.extra_downsamples = nn.ModuleList()
+        for i in range(extra_levels):
+            if self.add_extra_convs:
+                fpn_idx = self.backbone_end_level - self.start_level + i
+                extra_conv = nn.Conv2d(
+                    self.inter_channels[fpn_idx - 1],
+                    self.inter_channels[fpn_idx],
+                    3,
+                    stride=2,
+                    padding=1)
+                self.extra_downsamples.append(extra_conv)
+            else:
+                self.extra_downsamples.append(nn.MaxPool2d(1, stride=2))
+
+        self.fpn_transitions = nn.ModuleList()  # stack times
+        for s in range(self.stack_times):
+            stage_trans = nn.ModuleList()  # num of feature levels
+            for i in range(self.num_outs):
+                # same, across_lateral, across_down, across_up
+                trans = nn.ModuleDict()
+                if s in self.skip_inds[i]:
+                    stage_trans.append(trans)
+                    continue
+                # build same-stage down trans (used in bottom-up paths)
+                if i == 0 or self.same_up_trans is None:
+                    same_up_trans = None
+                else:
+                    same_up_trans = self.build_trans(
+                        self.same_up_trans, self.inter_channels[i - 1],
+                        self.inter_channels[i])
+                trans['same_up'] = same_up_trans
+                # build same-stage up trans (used in top-down paths)
+                if i == self.num_outs - 1 or self.same_down_trans is None:
+                    same_down_trans = None
+                else:
+                    same_down_trans = self.build_trans(
+                        self.same_down_trans, self.inter_channels[i + 1],
+                        self.inter_channels[i])
+                trans['same_down'] = same_down_trans
+                # build across lateral trans
+                across_lateral_trans = self.build_trans(
+                    self.across_lateral_trans, self.inter_channels[i],
+                    self.inter_channels[i])
+                trans['across_lateral'] = across_lateral_trans
+                # build across down trans
+                if i == self.num_outs - 1 or self.across_down_trans is None:
+                    across_down_trans = None
+                else:
+                    across_down_trans = self.build_trans(
+                        self.across_down_trans, self.inter_channels[i + 1],
+                        self.inter_channels[i])
+                trans['across_down'] = across_down_trans
+                # build across up trans
+                if i == 0 or self.across_up_trans is None:
+                    across_up_trans = None
+                else:
+                    across_up_trans = self.build_trans(
+                        self.across_up_trans, self.inter_channels[i - 1],
+                        self.inter_channels[i])
+                trans['across_up'] = across_up_trans
+                if self.across_skip_trans is None:
+                    across_skip_trans = None
+                else:
+                    across_skip_trans = self.build_trans(
+                        self.across_skip_trans, self.inter_channels[i - 1],
+                        self.inter_channels[i])
+                trans['across_skip'] = across_skip_trans
+                # build across_skip trans
+                stage_trans.append(trans)
+            self.fpn_transitions.append(stage_trans)
+
+        self.output_transition = nn.ModuleList()  # output levels
+        for i in range(self.num_outs):
+            trans = self.build_trans(
+                self.output_trans,
+                self.inter_channels[i],
+                self.out_channels,
+                num_inputs=self.stack_times + 1)
+            self.output_transition.append(trans)
+
+        self.relu = nn.ReLU(inplace=True)
+
+    def build_trans(self, cfg, in_channels, out_channels, **extra_args):
+        cfg_ = cfg.copy()
+        trans_type = cfg_.pop('type')
+        trans_cls = self.transition_types[trans_type]
+        return trans_cls(in_channels, out_channels, **cfg_, **extra_args)
+
+    def init_weights(self):
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                caffe2_xavier_init(m)
+            elif is_norm(m):
+                constant_init(m, 1.0)
+
+    def fuse(self, fuse_dict):
+        out = None
+        for item in fuse_dict.values():
+            if item is not None:
+                if out is None:
+                    out = item
+                else:
+                    out = out + item
+        return out
+
+    def forward(self, inputs):
+        assert len(inputs) == len(self.in_channels)
+
+        # build all levels from original feature maps
+        feats = [
+            lateral_conv(inputs[i + self.start_level])
+            for i, lateral_conv in enumerate(self.lateral_convs)
+        ]
+        for downsample in self.extra_downsamples:
+            feats.append(downsample(feats[-1]))
+
+        outs = [feats]
+
+        for i in range(self.stack_times):
+            current_outs = outs[-1]
+            next_outs = []
+            direction = self.paths[i]
+            for j in range(self.num_outs):
+                if i in self.skip_inds[j]:
+                    next_outs.append(outs[-1][j])
+                    continue
+                # feature level
+                if direction == 'td':
+                    lvl = self.num_outs - j - 1
+                else:
+                    lvl = j
+                # get transitions
+                if direction == 'td':
+                    same_trans = self.fpn_transitions[i][lvl]['same_down']
+                else:
+                    same_trans = self.fpn_transitions[i][lvl]['same_up']
+                across_lateral_trans = self.fpn_transitions[i][lvl][
+                    'across_lateral']
+                across_down_trans = self.fpn_transitions[i][lvl]['across_down']
+                across_up_trans = self.fpn_transitions[i][lvl]['across_up']
+                across_skip_trans = self.fpn_transitions[i][lvl]['across_skip']
+                # init output
+                to_fuse = dict(
+                    same=None, lateral=None, across_up=None, across_down=None)
+                # same downsample/upsample
+                if same_trans is not None:
+                    to_fuse['same'] = same_trans(next_outs[-1])
+                # across lateral
+                if across_lateral_trans is not None:
+                    to_fuse['lateral'] = across_lateral_trans(
+                        current_outs[lvl])
+                # across downsample
+                if lvl > 0 and across_up_trans is not None:
+                    to_fuse['across_up'] = across_up_trans(current_outs[lvl -
+                                                                        1])
+                # across upsample
+                if (lvl < self.num_outs - 1 and across_down_trans is not None):
+                    to_fuse['across_down'] = across_down_trans(
+                        current_outs[lvl + 1])
+                if across_skip_trans is not None:
+                    to_fuse['across_skip'] = across_skip_trans(outs[0][lvl])
+                x = self.fuse(to_fuse)
+                next_outs.append(x)
+
+            if direction == 'td':
+                outs.append(next_outs[::-1])
+            else:
+                outs.append(next_outs)
+
+        # output trans
+        final_outs = []
+        for i in range(self.num_outs):
+            lvl_out_list = []
+            for s in range(len(outs)):
+                lvl_out_list.append(outs[s][i])
+            lvl_out = self.output_transition[i](lvl_out_list)
+            final_outs.append(lvl_out)
+
+        return final_outs
diff --git a/python/autoCut/models/swin/mmdet/models/necks/fpn.py b/python/autoCut/models/swin/mmdet/models/necks/fpn.py
new file mode 100644
index 0000000..5e5dfe6
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/fpn.py
@@ -0,0 +1,221 @@
+import warnings
+
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, xavier_init
+from mmcv.runner import auto_fp16
+
+from ..builder import NECKS
+
+
+@NECKS.register_module()
+class FPN(nn.Module):
+    r"""Feature Pyramid Network.
+
+    This is an implementation of paper `Feature Pyramid Networks for Object
+    Detection <https://arxiv.org/abs/1612.03144>`_.
+
+    Args:
+        in_channels (List[int]): Number of input channels per scale.
+        out_channels (int): Number of output channels (used at each scale)
+        num_outs (int): Number of output scales.
+        start_level (int): Index of the start input backbone level used to
+            build the feature pyramid. Default: 0.
+        end_level (int): Index of the end input backbone level (exclusive) to
+            build the feature pyramid. Default: -1, which means the last level.
+        add_extra_convs (bool | str): If bool, it decides whether to add conv
+            layers on top of the original feature maps. Default to False.
+            If True, its actual mode is specified by `extra_convs_on_inputs`.
+            If str, it specifies the source feature map of the extra convs.
+            Only the following options are allowed
+
+            - 'on_input': Last feat map of neck inputs (i.e. backbone feature).
+            - 'on_lateral':  Last feature map after lateral convs.
+            - 'on_output': The last output feature map after fpn convs.
+        extra_convs_on_inputs (bool, deprecated): Whether to apply extra convs
+            on the original feature from the backbone. If True,
+            it is equivalent to `add_extra_convs='on_input'`. If False, it is
+            equivalent to set `add_extra_convs='on_output'`. Default to True.
+        relu_before_extra_convs (bool): Whether to apply relu before the extra
+            conv. Default: False.
+        no_norm_on_lateral (bool): Whether to apply norm on lateral.
+            Default: False.
+        conv_cfg (dict): Config dict for convolution layer. Default: None.
+        norm_cfg (dict): Config dict for normalization layer. Default: None.
+        act_cfg (str): Config dict for activation layer in ConvModule.
+            Default: None.
+        upsample_cfg (dict): Config dict for interpolate layer.
+            Default: `dict(mode='nearest')`
+
+    Example:
+        >>> import torch
+        >>> in_channels = [2, 3, 5, 7]
+        >>> scales = [340, 170, 84, 43]
+        >>> inputs = [torch.rand(1, c, s, s)
+        ...           for c, s in zip(in_channels, scales)]
+        >>> self = FPN(in_channels, 11, len(in_channels)).eval()
+        >>> outputs = self.forward(inputs)
+        >>> for i in range(len(outputs)):
+        ...     print(f'outputs[{i}].shape = {outputs[i].shape}')
+        outputs[0].shape = torch.Size([1, 11, 340, 340])
+        outputs[1].shape = torch.Size([1, 11, 170, 170])
+        outputs[2].shape = torch.Size([1, 11, 84, 84])
+        outputs[3].shape = torch.Size([1, 11, 43, 43])
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 num_outs,
+                 start_level=0,
+                 end_level=-1,
+                 add_extra_convs=False,
+                 extra_convs_on_inputs=True,
+                 relu_before_extra_convs=False,
+                 no_norm_on_lateral=False,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 act_cfg=None,
+                 upsample_cfg=dict(mode='nearest')):
+        super(FPN, self).__init__()
+        assert isinstance(in_channels, list)
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.num_ins = len(in_channels)
+        self.num_outs = num_outs
+        self.relu_before_extra_convs = relu_before_extra_convs
+        self.no_norm_on_lateral = no_norm_on_lateral
+        self.fp16_enabled = False
+        self.upsample_cfg = upsample_cfg.copy()
+
+        if end_level == -1:
+            self.backbone_end_level = self.num_ins
+            assert num_outs >= self.num_ins - start_level
+        else:
+            # if end_level < inputs, no extra level is allowed
+            self.backbone_end_level = end_level
+            assert end_level <= len(in_channels)
+            assert num_outs == end_level - start_level
+        self.start_level = start_level
+        self.end_level = end_level
+        self.add_extra_convs = add_extra_convs
+        assert isinstance(add_extra_convs, (str, bool))
+        if isinstance(add_extra_convs, str):
+            # Extra_convs_source choices: 'on_input', 'on_lateral', 'on_output'
+            assert add_extra_convs in ('on_input', 'on_lateral', 'on_output')
+        elif add_extra_convs:  # True
+            if extra_convs_on_inputs:
+                # TODO: deprecate `extra_convs_on_inputs`
+                warnings.simplefilter('once')
+                warnings.warn(
+                    '"extra_convs_on_inputs" will be deprecated in v2.9.0,'
+                    'Please use "add_extra_convs"', DeprecationWarning)
+                self.add_extra_convs = 'on_input'
+            else:
+                self.add_extra_convs = 'on_output'
+
+        self.lateral_convs = nn.ModuleList()
+        self.fpn_convs = nn.ModuleList()
+
+        for i in range(self.start_level, self.backbone_end_level):
+            l_conv = ConvModule(
+                in_channels[i],
+                out_channels,
+                1,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg if not self.no_norm_on_lateral else None,
+                act_cfg=act_cfg,
+                inplace=False)
+            fpn_conv = ConvModule(
+                out_channels,
+                out_channels,
+                3,
+                padding=1,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                act_cfg=act_cfg,
+                inplace=False)
+
+            self.lateral_convs.append(l_conv)
+            self.fpn_convs.append(fpn_conv)
+
+        # add extra conv layers (e.g., RetinaNet)
+        extra_levels = num_outs - self.backbone_end_level + self.start_level
+        if self.add_extra_convs and extra_levels >= 1:
+            for i in range(extra_levels):
+                if i == 0 and self.add_extra_convs == 'on_input':
+                    in_channels = self.in_channels[self.backbone_end_level - 1]
+                else:
+                    in_channels = out_channels
+                extra_fpn_conv = ConvModule(
+                    in_channels,
+                    out_channels,
+                    3,
+                    stride=2,
+                    padding=1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg,
+                    act_cfg=act_cfg,
+                    inplace=False)
+                self.fpn_convs.append(extra_fpn_conv)
+
+    # default init_weights for conv(msra) and norm in ConvModule
+    def init_weights(self):
+        """Initialize the weights of FPN module."""
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                xavier_init(m, distribution='uniform')
+
+    @auto_fp16()
+    def forward(self, inputs):
+        """Forward function."""
+        assert len(inputs) == len(self.in_channels)
+
+        # build laterals
+        laterals = [
+            lateral_conv(inputs[i + self.start_level])
+            for i, lateral_conv in enumerate(self.lateral_convs)
+        ]
+
+        # build top-down path
+        used_backbone_levels = len(laterals)
+        for i in range(used_backbone_levels - 1, 0, -1):
+            # In some cases, fixing `scale factor` (e.g. 2) is preferred, but
+            #  it cannot co-exist with `size` in `F.interpolate`.
+            if 'scale_factor' in self.upsample_cfg:
+                laterals[i - 1] += F.interpolate(laterals[i],
+                                                 **self.upsample_cfg)
+            else:
+                prev_shape = laterals[i - 1].shape[2:]
+                laterals[i - 1] += F.interpolate(
+                    laterals[i], size=prev_shape, **self.upsample_cfg)
+
+        # build outputs
+        # part 1: from original levels
+        outs = [
+            self.fpn_convs[i](laterals[i]) for i in range(used_backbone_levels)
+        ]
+        # part 2: add extra levels
+        if self.num_outs > len(outs):
+            # use max pool to get more levels on top of outputs
+            # (e.g., Faster R-CNN, Mask R-CNN)
+            if not self.add_extra_convs:
+                for i in range(self.num_outs - used_backbone_levels):
+                    outs.append(F.max_pool2d(outs[-1], 1, stride=2))
+            # add conv layers on top of original feature maps (RetinaNet)
+            else:
+                if self.add_extra_convs == 'on_input':
+                    extra_source = inputs[self.backbone_end_level - 1]
+                elif self.add_extra_convs == 'on_lateral':
+                    extra_source = laterals[-1]
+                elif self.add_extra_convs == 'on_output':
+                    extra_source = outs[-1]
+                else:
+                    raise NotImplementedError
+                outs.append(self.fpn_convs[used_backbone_levels](extra_source))
+                for i in range(used_backbone_levels + 1, self.num_outs):
+                    if self.relu_before_extra_convs:
+                        outs.append(self.fpn_convs[i](F.relu(outs[-1])))
+                    else:
+                        outs.append(self.fpn_convs[i](outs[-1]))
+        return tuple(outs)
diff --git a/python/autoCut/models/swin/mmdet/models/necks/fpn_carafe.py b/python/autoCut/models/swin/mmdet/models/necks/fpn_carafe.py
new file mode 100644
index 0000000..302e657
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/fpn_carafe.py
@@ -0,0 +1,267 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, build_upsample_layer, xavier_init
+from mmcv.ops.carafe import CARAFEPack
+
+from ..builder import NECKS
+
+
+@NECKS.register_module()
+class FPN_CARAFE(nn.Module):
+    """FPN_CARAFE is a more flexible implementation of FPN. It allows more
+    choice for upsample methods during the top-down pathway.
+
+    It can reproduce the performance of ICCV 2019 paper
+    CARAFE: Content-Aware ReAssembly of FEatures
+    Please refer to https://arxiv.org/abs/1905.02188 for more details.
+
+    Args:
+        in_channels (list[int]): Number of channels for each input feature map.
+        out_channels (int): Output channels of feature pyramids.
+        num_outs (int): Number of output stages.
+        start_level (int): Start level of feature pyramids.
+            (Default: 0)
+        end_level (int): End level of feature pyramids.
+            (Default: -1 indicates the last level).
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+        activate (str): Type of activation function in ConvModule
+            (Default: None indicates w/o activation).
+        order (dict): Order of components in ConvModule.
+        upsample (str): Type of upsample layer.
+        upsample_cfg (dict): Dictionary to construct and config upsample layer.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 num_outs,
+                 start_level=0,
+                 end_level=-1,
+                 norm_cfg=None,
+                 act_cfg=None,
+                 order=('conv', 'norm', 'act'),
+                 upsample_cfg=dict(
+                     type='carafe',
+                     up_kernel=5,
+                     up_group=1,
+                     encoder_kernel=3,
+                     encoder_dilation=1)):
+        super(FPN_CARAFE, self).__init__()
+        assert isinstance(in_channels, list)
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.num_ins = len(in_channels)
+        self.num_outs = num_outs
+        self.norm_cfg = norm_cfg
+        self.act_cfg = act_cfg
+        self.with_bias = norm_cfg is None
+        self.upsample_cfg = upsample_cfg.copy()
+        self.upsample = self.upsample_cfg.get('type')
+        self.relu = nn.ReLU(inplace=False)
+
+        self.order = order
+        assert order in [('conv', 'norm', 'act'), ('act', 'conv', 'norm')]
+
+        assert self.upsample in [
+            'nearest', 'bilinear', 'deconv', 'pixel_shuffle', 'carafe', None
+        ]
+        if self.upsample in ['deconv', 'pixel_shuffle']:
+            assert hasattr(
+                self.upsample_cfg,
+                'upsample_kernel') and self.upsample_cfg.upsample_kernel > 0
+            self.upsample_kernel = self.upsample_cfg.pop('upsample_kernel')
+
+        if end_level == -1:
+            self.backbone_end_level = self.num_ins
+            assert num_outs >= self.num_ins - start_level
+        else:
+            # if end_level < inputs, no extra level is allowed
+            self.backbone_end_level = end_level
+            assert end_level <= len(in_channels)
+            assert num_outs == end_level - start_level
+        self.start_level = start_level
+        self.end_level = end_level
+
+        self.lateral_convs = nn.ModuleList()
+        self.fpn_convs = nn.ModuleList()
+        self.upsample_modules = nn.ModuleList()
+
+        for i in range(self.start_level, self.backbone_end_level):
+            l_conv = ConvModule(
+                in_channels[i],
+                out_channels,
+                1,
+                norm_cfg=norm_cfg,
+                bias=self.with_bias,
+                act_cfg=act_cfg,
+                inplace=False,
+                order=self.order)
+            fpn_conv = ConvModule(
+                out_channels,
+                out_channels,
+                3,
+                padding=1,
+                norm_cfg=self.norm_cfg,
+                bias=self.with_bias,
+                act_cfg=act_cfg,
+                inplace=False,
+                order=self.order)
+            if i != self.backbone_end_level - 1:
+                upsample_cfg_ = self.upsample_cfg.copy()
+                if self.upsample == 'deconv':
+                    upsample_cfg_.update(
+                        in_channels=out_channels,
+                        out_channels=out_channels,
+                        kernel_size=self.upsample_kernel,
+                        stride=2,
+                        padding=(self.upsample_kernel - 1) // 2,
+                        output_padding=(self.upsample_kernel - 1) // 2)
+                elif self.upsample == 'pixel_shuffle':
+                    upsample_cfg_.update(
+                        in_channels=out_channels,
+                        out_channels=out_channels,
+                        scale_factor=2,
+                        upsample_kernel=self.upsample_kernel)
+                elif self.upsample == 'carafe':
+                    upsample_cfg_.update(channels=out_channels, scale_factor=2)
+                else:
+                    # suppress warnings
+                    align_corners = (None
+                                     if self.upsample == 'nearest' else False)
+                    upsample_cfg_.update(
+                        scale_factor=2,
+                        mode=self.upsample,
+                        align_corners=align_corners)
+                upsample_module = build_upsample_layer(upsample_cfg_)
+                self.upsample_modules.append(upsample_module)
+            self.lateral_convs.append(l_conv)
+            self.fpn_convs.append(fpn_conv)
+
+        # add extra conv layers (e.g., RetinaNet)
+        extra_out_levels = (
+            num_outs - self.backbone_end_level + self.start_level)
+        if extra_out_levels >= 1:
+            for i in range(extra_out_levels):
+                in_channels = (
+                    self.in_channels[self.backbone_end_level -
+                                     1] if i == 0 else out_channels)
+                extra_l_conv = ConvModule(
+                    in_channels,
+                    out_channels,
+                    3,
+                    stride=2,
+                    padding=1,
+                    norm_cfg=norm_cfg,
+                    bias=self.with_bias,
+                    act_cfg=act_cfg,
+                    inplace=False,
+                    order=self.order)
+                if self.upsample == 'deconv':
+                    upsampler_cfg_ = dict(
+                        in_channels=out_channels,
+                        out_channels=out_channels,
+                        kernel_size=self.upsample_kernel,
+                        stride=2,
+                        padding=(self.upsample_kernel - 1) // 2,
+                        output_padding=(self.upsample_kernel - 1) // 2)
+                elif self.upsample == 'pixel_shuffle':
+                    upsampler_cfg_ = dict(
+                        in_channels=out_channels,
+                        out_channels=out_channels,
+                        scale_factor=2,
+                        upsample_kernel=self.upsample_kernel)
+                elif self.upsample == 'carafe':
+                    upsampler_cfg_ = dict(
+                        channels=out_channels,
+                        scale_factor=2,
+                        **self.upsample_cfg)
+                else:
+                    # suppress warnings
+                    align_corners = (None
+                                     if self.upsample == 'nearest' else False)
+                    upsampler_cfg_ = dict(
+                        scale_factor=2,
+                        mode=self.upsample,
+                        align_corners=align_corners)
+                upsampler_cfg_['type'] = self.upsample
+                upsample_module = build_upsample_layer(upsampler_cfg_)
+                extra_fpn_conv = ConvModule(
+                    out_channels,
+                    out_channels,
+                    3,
+                    padding=1,
+                    norm_cfg=self.norm_cfg,
+                    bias=self.with_bias,
+                    act_cfg=act_cfg,
+                    inplace=False,
+                    order=self.order)
+                self.upsample_modules.append(upsample_module)
+                self.fpn_convs.append(extra_fpn_conv)
+                self.lateral_convs.append(extra_l_conv)
+
+    # default init_weights for conv(msra) and norm in ConvModule
+    def init_weights(self):
+        """Initialize the weights of module."""
+        for m in self.modules():
+            if isinstance(m, (nn.Conv2d, nn.ConvTranspose2d)):
+                xavier_init(m, distribution='uniform')
+        for m in self.modules():
+            if isinstance(m, CARAFEPack):
+                m.init_weights()
+
+    def slice_as(self, src, dst):
+        """Slice ``src`` as ``dst``
+
+        Note:
+            ``src`` should have the same or larger size than ``dst``.
+
+        Args:
+            src (torch.Tensor): Tensors to be sliced.
+            dst (torch.Tensor): ``src`` will be sliced to have the same
+                size as ``dst``.
+
+        Returns:
+            torch.Tensor: Sliced tensor.
+        """
+        assert (src.size(2) >= dst.size(2)) and (src.size(3) >= dst.size(3))
+        if src.size(2) == dst.size(2) and src.size(3) == dst.size(3):
+            return src
+        else:
+            return src[:, :, :dst.size(2), :dst.size(3)]
+
+    def tensor_add(self, a, b):
+        """Add tensors ``a`` and ``b`` that might have different sizes."""
+        if a.size() == b.size():
+            c = a + b
+        else:
+            c = a + self.slice_as(b, a)
+        return c
+
+    def forward(self, inputs):
+        """Forward function."""
+        assert len(inputs) == len(self.in_channels)
+
+        # build laterals
+        laterals = []
+        for i, lateral_conv in enumerate(self.lateral_convs):
+            if i <= self.backbone_end_level - self.start_level:
+                input = inputs[min(i + self.start_level, len(inputs) - 1)]
+            else:
+                input = laterals[-1]
+            lateral = lateral_conv(input)
+            laterals.append(lateral)
+
+        # build top-down path
+        for i in range(len(laterals) - 1, 0, -1):
+            if self.upsample is not None:
+                upsample_feat = self.upsample_modules[i - 1](laterals[i])
+            else:
+                upsample_feat = laterals[i]
+            laterals[i - 1] = self.tensor_add(laterals[i - 1], upsample_feat)
+
+        # build outputs
+        num_conv_outs = len(self.fpn_convs)
+        outs = []
+        for i in range(num_conv_outs):
+            out = self.fpn_convs[i](laterals[i])
+            outs.append(out)
+        return tuple(outs)
diff --git a/python/autoCut/models/swin/mmdet/models/necks/hrfpn.py b/python/autoCut/models/swin/mmdet/models/necks/hrfpn.py
new file mode 100644
index 0000000..ed4f194
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/hrfpn.py
@@ -0,0 +1,102 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, caffe2_xavier_init
+from torch.utils.checkpoint import checkpoint
+
+from ..builder import NECKS
+
+
+@NECKS.register_module()
+class HRFPN(nn.Module):
+    """HRFPN (High Resolution Feature Pyramids)
+
+    paper: `High-Resolution Representations for Labeling Pixels and Regions
+    <https://arxiv.org/abs/1904.04514>`_.
+
+    Args:
+        in_channels (list): number of channels for each branch.
+        out_channels (int): output channels of feature pyramids.
+        num_outs (int): number of output stages.
+        pooling_type (str): pooling for generating feature pyramids
+            from {MAX, AVG}.
+        conv_cfg (dict): dictionary to construct and config conv layer.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+        with_cp  (bool): Use checkpoint or not. Using checkpoint will save some
+            memory while slowing down the training speed.
+        stride (int): stride of 3x3 convolutional layers
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 num_outs=5,
+                 pooling_type='AVG',
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 with_cp=False,
+                 stride=1):
+        super(HRFPN, self).__init__()
+        assert isinstance(in_channels, list)
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.num_ins = len(in_channels)
+        self.num_outs = num_outs
+        self.with_cp = with_cp
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+
+        self.reduction_conv = ConvModule(
+            sum(in_channels),
+            out_channels,
+            kernel_size=1,
+            conv_cfg=self.conv_cfg,
+            act_cfg=None)
+
+        self.fpn_convs = nn.ModuleList()
+        for i in range(self.num_outs):
+            self.fpn_convs.append(
+                ConvModule(
+                    out_channels,
+                    out_channels,
+                    kernel_size=3,
+                    padding=1,
+                    stride=stride,
+                    conv_cfg=self.conv_cfg,
+                    act_cfg=None))
+
+        if pooling_type == 'MAX':
+            self.pooling = F.max_pool2d
+        else:
+            self.pooling = F.avg_pool2d
+
+    def init_weights(self):
+        """Initialize the weights of module."""
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                caffe2_xavier_init(m)
+
+    def forward(self, inputs):
+        """Forward function."""
+        assert len(inputs) == self.num_ins
+        outs = [inputs[0]]
+        for i in range(1, self.num_ins):
+            outs.append(
+                F.interpolate(inputs[i], scale_factor=2**i, mode='bilinear'))
+        out = torch.cat(outs, dim=1)
+        if out.requires_grad and self.with_cp:
+            out = checkpoint(self.reduction_conv, out)
+        else:
+            out = self.reduction_conv(out)
+        outs = [out]
+        for i in range(1, self.num_outs):
+            outs.append(self.pooling(out, kernel_size=2**i, stride=2**i))
+        outputs = []
+
+        for i in range(self.num_outs):
+            if outs[i].requires_grad and self.with_cp:
+                tmp_out = checkpoint(self.fpn_convs[i], outs[i])
+            else:
+                tmp_out = self.fpn_convs[i](outs[i])
+            outputs.append(tmp_out)
+        return tuple(outputs)
diff --git a/python/autoCut/models/swin/mmdet/models/necks/nas_fpn.py b/python/autoCut/models/swin/mmdet/models/necks/nas_fpn.py
new file mode 100644
index 0000000..8e333ce
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/nas_fpn.py
@@ -0,0 +1,160 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, caffe2_xavier_init
+from mmcv.ops.merge_cells import GlobalPoolingCell, SumCell
+
+from ..builder import NECKS
+
+
+@NECKS.register_module()
+class NASFPN(nn.Module):
+    """NAS-FPN.
+
+    Implementation of `NAS-FPN: Learning Scalable Feature Pyramid Architecture
+    for Object Detection <https://arxiv.org/abs/1904.07392>`_
+
+    Args:
+        in_channels (List[int]): Number of input channels per scale.
+        out_channels (int): Number of output channels (used at each scale)
+        num_outs (int): Number of output scales.
+        stack_times (int): The number of times the pyramid architecture will
+            be stacked.
+        start_level (int): Index of the start input backbone level used to
+            build the feature pyramid. Default: 0.
+        end_level (int): Index of the end input backbone level (exclusive) to
+            build the feature pyramid. Default: -1, which means the last level.
+        add_extra_convs (bool): It decides whether to add conv
+            layers on top of the original feature maps. Default to False.
+            If True, its actual mode is specified by `extra_convs_on_inputs`.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 num_outs,
+                 stack_times,
+                 start_level=0,
+                 end_level=-1,
+                 add_extra_convs=False,
+                 norm_cfg=None):
+        super(NASFPN, self).__init__()
+        assert isinstance(in_channels, list)
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.num_ins = len(in_channels)  # num of input feature levels
+        self.num_outs = num_outs  # num of output feature levels
+        self.stack_times = stack_times
+        self.norm_cfg = norm_cfg
+
+        if end_level == -1:
+            self.backbone_end_level = self.num_ins
+            assert num_outs >= self.num_ins - start_level
+        else:
+            # if end_level < inputs, no extra level is allowed
+            self.backbone_end_level = end_level
+            assert end_level <= len(in_channels)
+            assert num_outs == end_level - start_level
+        self.start_level = start_level
+        self.end_level = end_level
+        self.add_extra_convs = add_extra_convs
+
+        # add lateral connections
+        self.lateral_convs = nn.ModuleList()
+        for i in range(self.start_level, self.backbone_end_level):
+            l_conv = ConvModule(
+                in_channels[i],
+                out_channels,
+                1,
+                norm_cfg=norm_cfg,
+                act_cfg=None)
+            self.lateral_convs.append(l_conv)
+
+        # add extra downsample layers (stride-2 pooling or conv)
+        extra_levels = num_outs - self.backbone_end_level + self.start_level
+        self.extra_downsamples = nn.ModuleList()
+        for i in range(extra_levels):
+            extra_conv = ConvModule(
+                out_channels, out_channels, 1, norm_cfg=norm_cfg, act_cfg=None)
+            self.extra_downsamples.append(
+                nn.Sequential(extra_conv, nn.MaxPool2d(2, 2)))
+
+        # add NAS FPN connections
+        self.fpn_stages = nn.ModuleList()
+        for _ in range(self.stack_times):
+            stage = nn.ModuleDict()
+            # gp(p6, p4) -> p4_1
+            stage['gp_64_4'] = GlobalPoolingCell(
+                in_channels=out_channels,
+                out_channels=out_channels,
+                out_norm_cfg=norm_cfg)
+            # sum(p4_1, p4) -> p4_2
+            stage['sum_44_4'] = SumCell(
+                in_channels=out_channels,
+                out_channels=out_channels,
+                out_norm_cfg=norm_cfg)
+            # sum(p4_2, p3) -> p3_out
+            stage['sum_43_3'] = SumCell(
+                in_channels=out_channels,
+                out_channels=out_channels,
+                out_norm_cfg=norm_cfg)
+            # sum(p3_out, p4_2) -> p4_out
+            stage['sum_34_4'] = SumCell(
+                in_channels=out_channels,
+                out_channels=out_channels,
+                out_norm_cfg=norm_cfg)
+            # sum(p5, gp(p4_out, p3_out)) -> p5_out
+            stage['gp_43_5'] = GlobalPoolingCell(with_out_conv=False)
+            stage['sum_55_5'] = SumCell(
+                in_channels=out_channels,
+                out_channels=out_channels,
+                out_norm_cfg=norm_cfg)
+            # sum(p7, gp(p5_out, p4_2)) -> p7_out
+            stage['gp_54_7'] = GlobalPoolingCell(with_out_conv=False)
+            stage['sum_77_7'] = SumCell(
+                in_channels=out_channels,
+                out_channels=out_channels,
+                out_norm_cfg=norm_cfg)
+            # gp(p7_out, p5_out) -> p6_out
+            stage['gp_75_6'] = GlobalPoolingCell(
+                in_channels=out_channels,
+                out_channels=out_channels,
+                out_norm_cfg=norm_cfg)
+            self.fpn_stages.append(stage)
+
+    def init_weights(self):
+        """Initialize the weights of module."""
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                caffe2_xavier_init(m)
+
+    def forward(self, inputs):
+        """Forward function."""
+        # build P3-P5
+        feats = [
+            lateral_conv(inputs[i + self.start_level])
+            for i, lateral_conv in enumerate(self.lateral_convs)
+        ]
+        # build P6-P7 on top of P5
+        for downsample in self.extra_downsamples:
+            feats.append(downsample(feats[-1]))
+
+        p3, p4, p5, p6, p7 = feats
+
+        for stage in self.fpn_stages:
+            # gp(p6, p4) -> p4_1
+            p4_1 = stage['gp_64_4'](p6, p4, out_size=p4.shape[-2:])
+            # sum(p4_1, p4) -> p4_2
+            p4_2 = stage['sum_44_4'](p4_1, p4, out_size=p4.shape[-2:])
+            # sum(p4_2, p3) -> p3_out
+            p3 = stage['sum_43_3'](p4_2, p3, out_size=p3.shape[-2:])
+            # sum(p3_out, p4_2) -> p4_out
+            p4 = stage['sum_34_4'](p3, p4_2, out_size=p4.shape[-2:])
+            # sum(p5, gp(p4_out, p3_out)) -> p5_out
+            p5_tmp = stage['gp_43_5'](p4, p3, out_size=p5.shape[-2:])
+            p5 = stage['sum_55_5'](p5, p5_tmp, out_size=p5.shape[-2:])
+            # sum(p7, gp(p5_out, p4_2)) -> p7_out
+            p7_tmp = stage['gp_54_7'](p5, p4_2, out_size=p7.shape[-2:])
+            p7 = stage['sum_77_7'](p7, p7_tmp, out_size=p7.shape[-2:])
+            # gp(p7_out, p5_out) -> p6_out
+            p6 = stage['gp_75_6'](p7, p5, out_size=p6.shape[-2:])
+
+        return p3, p4, p5, p6, p7
diff --git a/python/autoCut/models/swin/mmdet/models/necks/nasfcos_fpn.py b/python/autoCut/models/swin/mmdet/models/necks/nasfcos_fpn.py
new file mode 100644
index 0000000..2daf79e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/nasfcos_fpn.py
@@ -0,0 +1,161 @@
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, caffe2_xavier_init
+from mmcv.ops.merge_cells import ConcatCell
+
+from ..builder import NECKS
+
+
+@NECKS.register_module()
+class NASFCOS_FPN(nn.Module):
+    """FPN structure in NASFPN.
+
+    Implementation of paper `NAS-FCOS: Fast Neural Architecture Search for
+    Object Detection <https://arxiv.org/abs/1906.04423>`_
+
+    Args:
+        in_channels (List[int]): Number of input channels per scale.
+        out_channels (int): Number of output channels (used at each scale)
+        num_outs (int): Number of output scales.
+        start_level (int): Index of the start input backbone level used to
+            build the feature pyramid. Default: 0.
+        end_level (int): Index of the end input backbone level (exclusive) to
+            build the feature pyramid. Default: -1, which means the last level.
+        add_extra_convs (bool): It decides whether to add conv
+            layers on top of the original feature maps. Default to False.
+            If True, its actual mode is specified by `extra_convs_on_inputs`.
+        conv_cfg (dict): dictionary to construct and config conv layer.
+        norm_cfg (dict): dictionary to construct and config norm layer.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 num_outs,
+                 start_level=1,
+                 end_level=-1,
+                 add_extra_convs=False,
+                 conv_cfg=None,
+                 norm_cfg=None):
+        super(NASFCOS_FPN, self).__init__()
+        assert isinstance(in_channels, list)
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        self.num_ins = len(in_channels)
+        self.num_outs = num_outs
+        self.norm_cfg = norm_cfg
+        self.conv_cfg = conv_cfg
+
+        if end_level == -1:
+            self.backbone_end_level = self.num_ins
+            assert num_outs >= self.num_ins - start_level
+        else:
+            self.backbone_end_level = end_level
+            assert end_level <= len(in_channels)
+            assert num_outs == end_level - start_level
+        self.start_level = start_level
+        self.end_level = end_level
+        self.add_extra_convs = add_extra_convs
+
+        self.adapt_convs = nn.ModuleList()
+        for i in range(self.start_level, self.backbone_end_level):
+            adapt_conv = ConvModule(
+                in_channels[i],
+                out_channels,
+                1,
+                stride=1,
+                padding=0,
+                bias=False,
+                norm_cfg=dict(type='BN'),
+                act_cfg=dict(type='ReLU', inplace=False))
+            self.adapt_convs.append(adapt_conv)
+
+        # C2 is omitted according to the paper
+        extra_levels = num_outs - self.backbone_end_level + self.start_level
+
+        def build_concat_cell(with_input1_conv, with_input2_conv):
+            cell_conv_cfg = dict(
+                kernel_size=1, padding=0, bias=False, groups=out_channels)
+            return ConcatCell(
+                in_channels=out_channels,
+                out_channels=out_channels,
+                with_out_conv=True,
+                out_conv_cfg=cell_conv_cfg,
+                out_norm_cfg=dict(type='BN'),
+                out_conv_order=('norm', 'act', 'conv'),
+                with_input1_conv=with_input1_conv,
+                with_input2_conv=with_input2_conv,
+                input_conv_cfg=conv_cfg,
+                input_norm_cfg=norm_cfg,
+                upsample_mode='nearest')
+
+        # Denote c3=f0, c4=f1, c5=f2 for convince
+        self.fpn = nn.ModuleDict()
+        self.fpn['c22_1'] = build_concat_cell(True, True)
+        self.fpn['c22_2'] = build_concat_cell(True, True)
+        self.fpn['c32'] = build_concat_cell(True, False)
+        self.fpn['c02'] = build_concat_cell(True, False)
+        self.fpn['c42'] = build_concat_cell(True, True)
+        self.fpn['c36'] = build_concat_cell(True, True)
+        self.fpn['c61'] = build_concat_cell(True, True)  # f9
+        self.extra_downsamples = nn.ModuleList()
+        for i in range(extra_levels):
+            extra_act_cfg = None if i == 0 \
+                else dict(type='ReLU', inplace=False)
+            self.extra_downsamples.append(
+                ConvModule(
+                    out_channels,
+                    out_channels,
+                    3,
+                    stride=2,
+                    padding=1,
+                    act_cfg=extra_act_cfg,
+                    order=('act', 'norm', 'conv')))
+
+    def forward(self, inputs):
+        """Forward function."""
+        feats = [
+            adapt_conv(inputs[i + self.start_level])
+            for i, adapt_conv in enumerate(self.adapt_convs)
+        ]
+
+        for (i, module_name) in enumerate(self.fpn):
+            idx_1, idx_2 = int(module_name[1]), int(module_name[2])
+            res = self.fpn[module_name](feats[idx_1], feats[idx_2])
+            feats.append(res)
+
+        ret = []
+        for (idx, input_idx) in zip([9, 8, 7], [1, 2, 3]):  # add P3, P4, P5
+            feats1, feats2 = feats[idx], feats[5]
+            feats2_resize = F.interpolate(
+                feats2,
+                size=feats1.size()[2:],
+                mode='bilinear',
+                align_corners=False)
+
+            feats_sum = feats1 + feats2_resize
+            ret.append(
+                F.interpolate(
+                    feats_sum,
+                    size=inputs[input_idx].size()[2:],
+                    mode='bilinear',
+                    align_corners=False))
+
+        for submodule in self.extra_downsamples:
+            ret.append(submodule(ret[-1]))
+
+        return tuple(ret)
+
+    def init_weights(self):
+        """Initialize the weights of module."""
+        for module in self.fpn.values():
+            if hasattr(module, 'conv_out'):
+                caffe2_xavier_init(module.out_conv.conv)
+
+        for modules in [
+                self.adapt_convs.modules(),
+                self.extra_downsamples.modules()
+        ]:
+            for module in modules:
+                if isinstance(module, nn.Conv2d):
+                    caffe2_xavier_init(module)
diff --git a/python/autoCut/models/swin/mmdet/models/necks/pafpn.py b/python/autoCut/models/swin/mmdet/models/necks/pafpn.py
new file mode 100644
index 0000000..d7c0b50
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/pafpn.py
@@ -0,0 +1,142 @@
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule
+from mmcv.runner import auto_fp16
+
+from ..builder import NECKS
+from .fpn import FPN
+
+
+@NECKS.register_module()
+class PAFPN(FPN):
+    """Path Aggregation Network for Instance Segmentation.
+
+    This is an implementation of the `PAFPN in Path Aggregation Network
+    <https://arxiv.org/abs/1803.01534>`_.
+
+    Args:
+        in_channels (List[int]): Number of input channels per scale.
+        out_channels (int): Number of output channels (used at each scale)
+        num_outs (int): Number of output scales.
+        start_level (int): Index of the start input backbone level used to
+            build the feature pyramid. Default: 0.
+        end_level (int): Index of the end input backbone level (exclusive) to
+            build the feature pyramid. Default: -1, which means the last level.
+        add_extra_convs (bool): Whether to add conv layers on top of the
+            original feature maps. Default: False.
+        extra_convs_on_inputs (bool): Whether to apply extra conv on
+            the original feature from the backbone. Default: False.
+        relu_before_extra_convs (bool): Whether to apply relu before the extra
+            conv. Default: False.
+        no_norm_on_lateral (bool): Whether to apply norm on lateral.
+            Default: False.
+        conv_cfg (dict): Config dict for convolution layer. Default: None.
+        norm_cfg (dict): Config dict for normalization layer. Default: None.
+        act_cfg (str): Config dict for activation layer in ConvModule.
+            Default: None.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 num_outs,
+                 start_level=0,
+                 end_level=-1,
+                 add_extra_convs=False,
+                 extra_convs_on_inputs=True,
+                 relu_before_extra_convs=False,
+                 no_norm_on_lateral=False,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 act_cfg=None):
+        super(PAFPN,
+              self).__init__(in_channels, out_channels, num_outs, start_level,
+                             end_level, add_extra_convs, extra_convs_on_inputs,
+                             relu_before_extra_convs, no_norm_on_lateral,
+                             conv_cfg, norm_cfg, act_cfg)
+        # add extra bottom up pathway
+        self.downsample_convs = nn.ModuleList()
+        self.pafpn_convs = nn.ModuleList()
+        for i in range(self.start_level + 1, self.backbone_end_level):
+            d_conv = ConvModule(
+                out_channels,
+                out_channels,
+                3,
+                stride=2,
+                padding=1,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                act_cfg=act_cfg,
+                inplace=False)
+            pafpn_conv = ConvModule(
+                out_channels,
+                out_channels,
+                3,
+                padding=1,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                act_cfg=act_cfg,
+                inplace=False)
+            self.downsample_convs.append(d_conv)
+            self.pafpn_convs.append(pafpn_conv)
+
+    @auto_fp16()
+    def forward(self, inputs):
+        """Forward function."""
+        assert len(inputs) == len(self.in_channels)
+
+        # build laterals
+        laterals = [
+            lateral_conv(inputs[i + self.start_level])
+            for i, lateral_conv in enumerate(self.lateral_convs)
+        ]
+
+        # build top-down path
+        used_backbone_levels = len(laterals)
+        for i in range(used_backbone_levels - 1, 0, -1):
+            prev_shape = laterals[i - 1].shape[2:]
+            laterals[i - 1] += F.interpolate(
+                laterals[i], size=prev_shape, mode='nearest')
+
+        # build outputs
+        # part 1: from original levels
+        inter_outs = [
+            self.fpn_convs[i](laterals[i]) for i in range(used_backbone_levels)
+        ]
+
+        # part 2: add bottom-up path
+        for i in range(0, used_backbone_levels - 1):
+            inter_outs[i + 1] += self.downsample_convs[i](inter_outs[i])
+
+        outs = []
+        outs.append(inter_outs[0])
+        outs.extend([
+            self.pafpn_convs[i - 1](inter_outs[i])
+            for i in range(1, used_backbone_levels)
+        ])
+
+        # part 3: add extra levels
+        if self.num_outs > len(outs):
+            # use max pool to get more levels on top of outputs
+            # (e.g., Faster R-CNN, Mask R-CNN)
+            if not self.add_extra_convs:
+                for i in range(self.num_outs - used_backbone_levels):
+                    outs.append(F.max_pool2d(outs[-1], 1, stride=2))
+            # add conv layers on top of original feature maps (RetinaNet)
+            else:
+                if self.add_extra_convs == 'on_input':
+                    orig = inputs[self.backbone_end_level - 1]
+                    outs.append(self.fpn_convs[used_backbone_levels](orig))
+                elif self.add_extra_convs == 'on_lateral':
+                    outs.append(self.fpn_convs[used_backbone_levels](
+                        laterals[-1]))
+                elif self.add_extra_convs == 'on_output':
+                    outs.append(self.fpn_convs[used_backbone_levels](outs[-1]))
+                else:
+                    raise NotImplementedError
+                for i in range(used_backbone_levels + 1, self.num_outs):
+                    if self.relu_before_extra_convs:
+                        outs.append(self.fpn_convs[i](F.relu(outs[-1])))
+                    else:
+                        outs.append(self.fpn_convs[i](outs[-1]))
+        return tuple(outs)
diff --git a/python/autoCut/models/swin/mmdet/models/necks/rfp.py b/python/autoCut/models/swin/mmdet/models/necks/rfp.py
new file mode 100644
index 0000000..8a63e63
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/rfp.py
@@ -0,0 +1,128 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import constant_init, kaiming_init, xavier_init
+
+from ..builder import NECKS, build_backbone
+from .fpn import FPN
+
+
+class ASPP(nn.Module):
+    """ASPP (Atrous Spatial Pyramid Pooling)
+
+    This is an implementation of the ASPP module used in DetectoRS
+    (https://arxiv.org/pdf/2006.02334.pdf)
+
+    Args:
+        in_channels (int): Number of input channels.
+        out_channels (int): Number of channels produced by this module
+        dilations (tuple[int]): Dilations of the four branches.
+            Default: (1, 3, 6, 1)
+    """
+
+    def __init__(self, in_channels, out_channels, dilations=(1, 3, 6, 1)):
+        super().__init__()
+        assert dilations[-1] == 1
+        self.aspp = nn.ModuleList()
+        for dilation in dilations:
+            kernel_size = 3 if dilation > 1 else 1
+            padding = dilation if dilation > 1 else 0
+            conv = nn.Conv2d(
+                in_channels,
+                out_channels,
+                kernel_size=kernel_size,
+                stride=1,
+                dilation=dilation,
+                padding=padding,
+                bias=True)
+            self.aspp.append(conv)
+        self.gap = nn.AdaptiveAvgPool2d(1)
+        self.init_weights()
+
+    def init_weights(self):
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d):
+                kaiming_init(m)
+
+    def forward(self, x):
+        avg_x = self.gap(x)
+        out = []
+        for aspp_idx in range(len(self.aspp)):
+            inp = avg_x if (aspp_idx == len(self.aspp) - 1) else x
+            out.append(F.relu_(self.aspp[aspp_idx](inp)))
+        out[-1] = out[-1].expand_as(out[-2])
+        out = torch.cat(out, dim=1)
+        return out
+
+
+@NECKS.register_module()
+class RFP(FPN):
+    """RFP (Recursive Feature Pyramid)
+
+    This is an implementation of RFP in `DetectoRS
+    <https://arxiv.org/pdf/2006.02334.pdf>`_. Different from standard FPN, the
+    input of RFP should be multi level features along with origin input image
+    of backbone.
+
+    Args:
+        rfp_steps (int): Number of unrolled steps of RFP.
+        rfp_backbone (dict): Configuration of the backbone for RFP.
+        aspp_out_channels (int): Number of output channels of ASPP module.
+        aspp_dilations (tuple[int]): Dilation rates of four branches.
+            Default: (1, 3, 6, 1)
+    """
+
+    def __init__(self,
+                 rfp_steps,
+                 rfp_backbone,
+                 aspp_out_channels,
+                 aspp_dilations=(1, 3, 6, 1),
+                 **kwargs):
+        super().__init__(**kwargs)
+        self.rfp_steps = rfp_steps
+        self.rfp_modules = nn.ModuleList()
+        for rfp_idx in range(1, rfp_steps):
+            rfp_module = build_backbone(rfp_backbone)
+            self.rfp_modules.append(rfp_module)
+        self.rfp_aspp = ASPP(self.out_channels, aspp_out_channels,
+                             aspp_dilations)
+        self.rfp_weight = nn.Conv2d(
+            self.out_channels,
+            1,
+            kernel_size=1,
+            stride=1,
+            padding=0,
+            bias=True)
+
+    def init_weights(self):
+        # Avoid using super().init_weights(), which may alter the default
+        # initialization of the modules in self.rfp_modules that have missing
+        # keys in the pretrained checkpoint.
+        for convs in [self.lateral_convs, self.fpn_convs]:
+            for m in convs.modules():
+                if isinstance(m, nn.Conv2d):
+                    xavier_init(m, distribution='uniform')
+        for rfp_idx in range(self.rfp_steps - 1):
+            self.rfp_modules[rfp_idx].init_weights(
+                self.rfp_modules[rfp_idx].pretrained)
+        constant_init(self.rfp_weight, 0)
+
+    def forward(self, inputs):
+        inputs = list(inputs)
+        assert len(inputs) == len(self.in_channels) + 1  # +1 for input image
+        img = inputs.pop(0)
+        # FPN forward
+        x = super().forward(tuple(inputs))
+        for rfp_idx in range(self.rfp_steps - 1):
+            rfp_feats = [x[0]] + list(
+                self.rfp_aspp(x[i]) for i in range(1, len(x)))
+            x_idx = self.rfp_modules[rfp_idx].rfp_forward(img, rfp_feats)
+            # FPN forward
+            x_idx = super().forward(x_idx)
+            x_new = []
+            for ft_idx in range(len(x_idx)):
+                add_weight = torch.sigmoid(self.rfp_weight(x_idx[ft_idx]))
+                x_new.append(add_weight * x_idx[ft_idx] +
+                             (1 - add_weight) * x[ft_idx])
+            x = x_new
+        return x
diff --git a/python/autoCut/models/swin/mmdet/models/necks/yolo_neck.py b/python/autoCut/models/swin/mmdet/models/necks/yolo_neck.py
new file mode 100644
index 0000000..c2f9b9e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/necks/yolo_neck.py
@@ -0,0 +1,136 @@
+# Copyright (c) 2019 Western Digital Corporation or its affiliates.
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule
+
+from ..builder import NECKS
+
+
+class DetectionBlock(nn.Module):
+    """Detection block in YOLO neck.
+
+    Let out_channels = n, the DetectionBlock contains:
+    Six ConvLayers, 1 Conv2D Layer and 1 YoloLayer.
+    The first 6 ConvLayers are formed the following way:
+        1x1xn, 3x3x2n, 1x1xn, 3x3x2n, 1x1xn, 3x3x2n.
+    The Conv2D layer is 1x1x255.
+    Some block will have branch after the fifth ConvLayer.
+    The input channel is arbitrary (in_channels)
+
+    Args:
+        in_channels (int): The number of input channels.
+        out_channels (int): The number of output channels.
+        conv_cfg (dict): Config dict for convolution layer. Default: None.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+            Default: dict(type='BN', requires_grad=True)
+        act_cfg (dict): Config dict for activation layer.
+            Default: dict(type='LeakyReLU', negative_slope=0.1).
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN', requires_grad=True),
+                 act_cfg=dict(type='LeakyReLU', negative_slope=0.1)):
+        super(DetectionBlock, self).__init__()
+        double_out_channels = out_channels * 2
+
+        # shortcut
+        cfg = dict(conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=act_cfg)
+        self.conv1 = ConvModule(in_channels, out_channels, 1, **cfg)
+        self.conv2 = ConvModule(
+            out_channels, double_out_channels, 3, padding=1, **cfg)
+        self.conv3 = ConvModule(double_out_channels, out_channels, 1, **cfg)
+        self.conv4 = ConvModule(
+            out_channels, double_out_channels, 3, padding=1, **cfg)
+        self.conv5 = ConvModule(double_out_channels, out_channels, 1, **cfg)
+
+    def forward(self, x):
+        tmp = self.conv1(x)
+        tmp = self.conv2(tmp)
+        tmp = self.conv3(tmp)
+        tmp = self.conv4(tmp)
+        out = self.conv5(tmp)
+        return out
+
+
+@NECKS.register_module()
+class YOLOV3Neck(nn.Module):
+    """The neck of YOLOV3.
+
+    It can be treated as a simplified version of FPN. It
+    will take the result from Darknet backbone and do some upsampling and
+    concatenation. It will finally output the detection result.
+
+    Note:
+        The input feats should be from top to bottom.
+            i.e., from high-lvl to low-lvl
+        But YOLOV3Neck will process them in reversed order.
+            i.e., from bottom (high-lvl) to top (low-lvl)
+
+    Args:
+        num_scales (int): The number of scales / stages.
+        in_channels (int): The number of input channels.
+        out_channels (int): The number of output channels.
+        conv_cfg (dict): Config dict for convolution layer. Default: None.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+            Default: dict(type='BN', requires_grad=True)
+        act_cfg (dict): Config dict for activation layer.
+            Default: dict(type='LeakyReLU', negative_slope=0.1).
+    """
+
+    def __init__(self,
+                 num_scales,
+                 in_channels,
+                 out_channels,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN', requires_grad=True),
+                 act_cfg=dict(type='LeakyReLU', negative_slope=0.1)):
+        super(YOLOV3Neck, self).__init__()
+        assert (num_scales == len(in_channels) == len(out_channels))
+        self.num_scales = num_scales
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+
+        # shortcut
+        cfg = dict(conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=act_cfg)
+
+        # To support arbitrary scales, the code looks awful, but it works.
+        # Better solution is welcomed.
+        self.detect1 = DetectionBlock(in_channels[0], out_channels[0], **cfg)
+        for i in range(1, self.num_scales):
+            in_c, out_c = self.in_channels[i], self.out_channels[i]
+            self.add_module(f'conv{i}', ConvModule(in_c, out_c, 1, **cfg))
+            # in_c + out_c : High-lvl feats will be cat with low-lvl feats
+            self.add_module(f'detect{i+1}',
+                            DetectionBlock(in_c + out_c, out_c, **cfg))
+
+    def forward(self, feats):
+        assert len(feats) == self.num_scales
+
+        # processed from bottom (high-lvl) to top (low-lvl)
+        outs = []
+        out = self.detect1(feats[-1])
+        outs.append(out)
+
+        for i, x in enumerate(reversed(feats[:-1])):
+            conv = getattr(self, f'conv{i+1}')
+            tmp = conv(out)
+
+            # Cat with low-lvl feats
+            tmp = F.interpolate(tmp, scale_factor=2)
+            tmp = torch.cat((tmp, x), 1)
+
+            detect = getattr(self, f'detect{i+2}')
+            out = detect(tmp)
+            outs.append(out)
+
+        return tuple(outs)
+
+    def init_weights(self):
+        """Initialize the weights of module."""
+        # init is done in ConvModule
+        pass
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/__init__.py b/python/autoCut/models/swin/mmdet/models/roi_heads/__init__.py
new file mode 100644
index 0000000..4c1a5b3
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/__init__.py
@@ -0,0 +1,34 @@
+from .base_roi_head import BaseRoIHead
+from .bbox_heads import (BBoxHead, ConvFCBBoxHead, DoubleConvFCBBoxHead,
+                         SCNetBBoxHead, Shared2FCBBoxHead,
+                         Shared4Conv1FCBBoxHead)
+from .cascade_roi_head import CascadeRoIHead
+from .double_roi_head import DoubleHeadRoIHead
+from .dynamic_roi_head import DynamicRoIHead
+from .grid_roi_head import GridRoIHead
+from .htc_roi_head import HybridTaskCascadeRoIHead
+from .mask_heads import (CoarseMaskHead, FCNMaskHead, FeatureRelayHead,
+                         FusedSemanticHead, GlobalContextHead, GridHead,
+                         HTCMaskHead, MaskIoUHead, MaskPointHead,
+                         SCNetMaskHead, SCNetSemanticHead, CondConvMaskHead)
+from .mask_scoring_roi_head import MaskScoringRoIHead
+from .pisa_roi_head import PISARoIHead
+from .point_rend_roi_head import PointRendRoIHead
+from .roi_extractors import SingleRoIExtractor
+from .scnet_roi_head import SCNetRoIHead
+from .shared_heads import ResLayer
+from .sparse_roi_head import SparseRoIHead
+from .standard_roi_head import StandardRoIHead
+from .trident_roi_head import TridentRoIHead
+
+__all__ = [
+    'BaseRoIHead', 'CascadeRoIHead', 'DoubleHeadRoIHead', 'MaskScoringRoIHead',
+    'HybridTaskCascadeRoIHead', 'GridRoIHead', 'ResLayer', 'BBoxHead',
+    'ConvFCBBoxHead', 'Shared2FCBBoxHead', 'StandardRoIHead',
+    'Shared4Conv1FCBBoxHead', 'DoubleConvFCBBoxHead', 'FCNMaskHead',
+    'HTCMaskHead', 'FusedSemanticHead', 'GridHead', 'MaskIoUHead',
+    'SingleRoIExtractor', 'PISARoIHead', 'PointRendRoIHead', 'MaskPointHead',
+    'CoarseMaskHead', 'DynamicRoIHead', 'SparseRoIHead', 'TridentRoIHead',
+    'SCNetRoIHead', 'SCNetMaskHead', 'SCNetSemanticHead', 'SCNetBBoxHead',
+    'FeatureRelayHead', 'GlobalContextHead', 'CondConvMaskHead'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/base_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/base_roi_head.py
new file mode 100644
index 0000000..2d61cc0
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/base_roi_head.py
@@ -0,0 +1,103 @@
+from abc import ABCMeta, abstractmethod
+
+import torch.nn as nn
+
+from ..builder import build_shared_head
+
+
+class BaseRoIHead(nn.Module, metaclass=ABCMeta):
+    """Base class for RoIHeads."""
+
+    def __init__(self,
+                 bbox_roi_extractor=None,
+                 bbox_head=None,
+                 mask_roi_extractor=None,
+                 mask_head=None,
+                 shared_head=None,
+                 train_cfg=None,
+                 test_cfg=None):
+        super(BaseRoIHead, self).__init__()
+        self.train_cfg = train_cfg
+        self.test_cfg = test_cfg
+        if shared_head is not None:
+            self.shared_head = build_shared_head(shared_head)
+
+        if bbox_head is not None:
+            self.init_bbox_head(bbox_roi_extractor, bbox_head)
+
+        if mask_head is not None:
+            self.init_mask_head(mask_roi_extractor, mask_head)
+
+        self.init_assigner_sampler()
+
+    @property
+    def with_bbox(self):
+        """bool: whether the RoI head contains a `bbox_head`"""
+        return hasattr(self, 'bbox_head') and self.bbox_head is not None
+
+    @property
+    def with_mask(self):
+        """bool: whether the RoI head contains a `mask_head`"""
+        return hasattr(self, 'mask_head') and self.mask_head is not None
+
+    @property
+    def with_shared_head(self):
+        """bool: whether the RoI head contains a `shared_head`"""
+        return hasattr(self, 'shared_head') and self.shared_head is not None
+
+    @abstractmethod
+    def init_weights(self, pretrained):
+        """Initialize the weights in head.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        pass
+
+    @abstractmethod
+    def init_bbox_head(self):
+        """Initialize ``bbox_head``"""
+        pass
+
+    @abstractmethod
+    def init_mask_head(self):
+        """Initialize ``mask_head``"""
+        pass
+
+    @abstractmethod
+    def init_assigner_sampler(self):
+        """Initialize assigner and sampler."""
+        pass
+
+    @abstractmethod
+    def forward_train(self,
+                      x,
+                      img_meta,
+                      proposal_list,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None,
+                      **kwargs):
+        """Forward function during training."""
+
+    async def async_simple_test(self, x, img_meta, **kwargs):
+        """Asynchronized test function."""
+        raise NotImplementedError
+
+    def simple_test(self,
+                    x,
+                    proposal_list,
+                    img_meta,
+                    proposals=None,
+                    rescale=False,
+                    **kwargs):
+        """Test without augmentation."""
+
+    def aug_test(self, x, proposal_list, img_metas, rescale=False, **kwargs):
+        """Test with augmentations.
+
+        If rescale is False, then returned bboxes and masks will fit the scale
+        of imgs[0].
+        """
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/__init__.py b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/__init__.py
new file mode 100644
index 0000000..bc5d29e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/__init__.py
@@ -0,0 +1,13 @@
+from .bbox_head import BBoxHead
+from .convfc_bbox_head import (ConvFCBBoxHead, Shared2FCBBoxHead,
+                               Shared4Conv1FCBBoxHead)
+from .dii_head import DIIHead
+from .double_bbox_head import DoubleConvFCBBoxHead
+from .sabl_head import SABLHead
+from .scnet_bbox_head import SCNetBBoxHead
+
+__all__ = [
+    'BBoxHead', 'ConvFCBBoxHead', 'Shared2FCBBoxHead',
+    'Shared4Conv1FCBBoxHead', 'DoubleConvFCBBoxHead', 'SABLHead', 'DIIHead',
+    'SCNetBBoxHead'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/bbox_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/bbox_head.py
new file mode 100644
index 0000000..408abef
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/bbox_head.py
@@ -0,0 +1,483 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.runner import auto_fp16, force_fp32
+from torch.nn.modules.utils import _pair
+
+from mmdet.core import build_bbox_coder, multi_apply, multiclass_nms
+from mmdet.models.builder import HEADS, build_loss
+from mmdet.models.losses import accuracy
+
+
+@HEADS.register_module()
+class BBoxHead(nn.Module):
+    """Simplest RoI head, with only two fc layers for classification and
+    regression respectively."""
+
+    def __init__(self,
+                 with_avg_pool=False,
+                 with_cls=True,
+                 with_reg=True,
+                 roi_feat_size=7,
+                 in_channels=256,
+                 num_classes=80,
+                 bbox_coder=dict(
+                     type='DeltaXYWHBBoxCoder',
+                     clip_border=True,
+                     target_means=[0., 0., 0., 0.],
+                     target_stds=[0.1, 0.1, 0.2, 0.2]),
+                 reg_class_agnostic=False,
+                 reg_decoded_bbox=False,
+                 loss_cls=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=False,
+                     loss_weight=1.0),
+                 loss_bbox=dict(
+                     type='SmoothL1Loss', beta=1.0, loss_weight=1.0)):
+        super(BBoxHead, self).__init__()
+        assert with_cls or with_reg
+        self.with_avg_pool = with_avg_pool
+        self.with_cls = with_cls
+        self.with_reg = with_reg
+        self.roi_feat_size = _pair(roi_feat_size)
+        self.roi_feat_area = self.roi_feat_size[0] * self.roi_feat_size[1]
+        self.in_channels = in_channels
+        self.num_classes = num_classes
+        self.reg_class_agnostic = reg_class_agnostic
+        self.reg_decoded_bbox = reg_decoded_bbox
+        self.fp16_enabled = False
+
+        self.bbox_coder = build_bbox_coder(bbox_coder)
+        self.loss_cls = build_loss(loss_cls)
+        self.loss_bbox = build_loss(loss_bbox)
+
+        in_channels = self.in_channels
+        if self.with_avg_pool:
+            self.avg_pool = nn.AvgPool2d(self.roi_feat_size)
+        else:
+            in_channels *= self.roi_feat_area
+        if self.with_cls:
+            # need to add background class
+            self.fc_cls = nn.Linear(in_channels, num_classes + 1)
+        if self.with_reg:
+            out_dim_reg = 4 if reg_class_agnostic else 4 * num_classes
+            self.fc_reg = nn.Linear(in_channels, out_dim_reg)
+        self.debug_imgs = None
+
+    def init_weights(self):
+        # conv layers are already initialized by ConvModule
+        if self.with_cls:
+            nn.init.normal_(self.fc_cls.weight, 0, 0.01)
+            nn.init.constant_(self.fc_cls.bias, 0)
+        if self.with_reg:
+            nn.init.normal_(self.fc_reg.weight, 0, 0.001)
+            nn.init.constant_(self.fc_reg.bias, 0)
+
+    @auto_fp16()
+    def forward(self, x):
+        if self.with_avg_pool:
+            x = self.avg_pool(x)
+        x = x.view(x.size(0), -1)
+        cls_score = self.fc_cls(x) if self.with_cls else None
+        bbox_pred = self.fc_reg(x) if self.with_reg else None
+        return cls_score, bbox_pred
+
+    def _get_target_single(self, pos_bboxes, neg_bboxes, pos_gt_bboxes,
+                           pos_gt_labels, cfg):
+        """Calculate the ground truth for proposals in the single image
+        according to the sampling results.
+
+        Args:
+            pos_bboxes (Tensor): Contains all the positive boxes,
+                has shape (num_pos, 4), the last dimension 4
+                represents [tl_x, tl_y, br_x, br_y].
+            neg_bboxes (Tensor): Contains all the negative boxes,
+                has shape (num_neg, 4), the last dimension 4
+                represents [tl_x, tl_y, br_x, br_y].
+            pos_gt_bboxes (Tensor): Contains all the gt_boxes,
+                has shape (num_gt, 4), the last dimension 4
+                represents [tl_x, tl_y, br_x, br_y].
+            pos_gt_labels (Tensor): Contains all the gt_labels,
+                has shape (num_gt).
+            cfg (obj:`ConfigDict`): `train_cfg` of R-CNN.
+
+        Returns:
+            Tuple[Tensor]: Ground truth for proposals
+            in a single image. Containing the following Tensors:
+
+                - labels(Tensor): Gt_labels for all proposals, has
+                  shape (num_proposals,).
+                - label_weights(Tensor): Labels_weights for all
+                  proposals, has shape (num_proposals,).
+                - bbox_targets(Tensor):Regression target for all
+                  proposals, has shape (num_proposals, 4), the
+                  last dimension 4 represents [tl_x, tl_y, br_x, br_y].
+                - bbox_weights(Tensor):Regression weights for all
+                  proposals, has shape (num_proposals, 4).
+        """
+        num_pos = pos_bboxes.size(0)
+        num_neg = neg_bboxes.size(0)
+        num_samples = num_pos + num_neg
+
+        # original implementation uses new_zeros since BG are set to be 0
+        # now use empty & fill because BG cat_id = num_classes,
+        # FG cat_id = [0, num_classes-1]
+        labels = pos_bboxes.new_full((num_samples, ),
+                                     self.num_classes,
+                                     dtype=torch.long)
+        label_weights = pos_bboxes.new_zeros(num_samples)
+        bbox_targets = pos_bboxes.new_zeros(num_samples, 4)
+        bbox_weights = pos_bboxes.new_zeros(num_samples, 4)
+        if num_pos > 0:
+            labels[:num_pos] = pos_gt_labels
+            pos_weight = 1.0 if cfg.pos_weight <= 0 else cfg.pos_weight
+            label_weights[:num_pos] = pos_weight
+            if not self.reg_decoded_bbox:
+                pos_bbox_targets = self.bbox_coder.encode(
+                    pos_bboxes, pos_gt_bboxes)
+            else:
+                # When the regression loss (e.g. `IouLoss`, `GIouLoss`)
+                # is applied directly on the decoded bounding boxes, both
+                # the predicted boxes and regression targets should be with
+                # absolute coordinate format.
+                pos_bbox_targets = pos_gt_bboxes
+            bbox_targets[:num_pos, :] = pos_bbox_targets
+            bbox_weights[:num_pos, :] = 1
+        if num_neg > 0:
+            label_weights[-num_neg:] = 1.0
+
+        return labels, label_weights, bbox_targets, bbox_weights
+
+    def get_targets(self,
+                    sampling_results,
+                    gt_bboxes,
+                    gt_labels,
+                    rcnn_train_cfg,
+                    concat=True):
+        """Calculate the ground truth for all samples in a batch according to
+        the sampling_results.
+
+        Almost the same as the implementation in bbox_head, we passed
+        additional parameters pos_inds_list and neg_inds_list to
+        `_get_target_single` function.
+
+        Args:
+            sampling_results (List[obj:SamplingResults]): Assign results of
+                all images in a batch after sampling.
+            gt_bboxes (list[Tensor]): Gt_bboxes of all images in a batch,
+                each tensor has shape (num_gt, 4),  the last dimension 4
+                represents [tl_x, tl_y, br_x, br_y].
+            gt_labels (list[Tensor]): Gt_labels of all images in a batch,
+                each tensor has shape (num_gt,).
+            rcnn_train_cfg (obj:ConfigDict): `train_cfg` of RCNN.
+            concat (bool): Whether to concatenate the results of all
+                the images in a single batch.
+
+        Returns:
+            Tuple[Tensor]: Ground truth for proposals in a single image.
+            Containing the following list of Tensors:
+
+                - labels (list[Tensor],Tensor): Gt_labels for all
+                  proposals in a batch, each tensor in list has
+                  shape (num_proposals,) when `concat=False`, otherwise
+                  just a single tensor has shape (num_all_proposals,).
+                - label_weights (list[Tensor]): Labels_weights for
+                  all proposals in a batch, each tensor in list has
+                  shape (num_proposals,) when `concat=False`, otherwise
+                  just a single tensor has shape (num_all_proposals,).
+                - bbox_targets (list[Tensor],Tensor): Regression target
+                  for all proposals in a batch, each tensor in list
+                  has shape (num_proposals, 4) when `concat=False`,
+                  otherwise just a single tensor has shape
+                  (num_all_proposals, 4), the last dimension 4 represents
+                  [tl_x, tl_y, br_x, br_y].
+                - bbox_weights (list[tensor],Tensor): Regression weights for
+                  all proposals in a batch, each tensor in list has shape
+                  (num_proposals, 4) when `concat=False`, otherwise just a
+                  single tensor has shape (num_all_proposals, 4).
+        """
+        pos_bboxes_list = [res.pos_bboxes for res in sampling_results]
+        neg_bboxes_list = [res.neg_bboxes for res in sampling_results]
+        pos_gt_bboxes_list = [res.pos_gt_bboxes for res in sampling_results]
+        pos_gt_labels_list = [res.pos_gt_labels for res in sampling_results]
+        labels, label_weights, bbox_targets, bbox_weights = multi_apply(
+            self._get_target_single,
+            pos_bboxes_list,
+            neg_bboxes_list,
+            pos_gt_bboxes_list,
+            pos_gt_labels_list,
+            cfg=rcnn_train_cfg)
+
+        if concat:
+            labels = torch.cat(labels, 0)
+            label_weights = torch.cat(label_weights, 0)
+            bbox_targets = torch.cat(bbox_targets, 0)
+            bbox_weights = torch.cat(bbox_weights, 0)
+        return labels, label_weights, bbox_targets, bbox_weights
+
+    @force_fp32(apply_to=('cls_score', 'bbox_pred'))
+    def loss(self,
+             cls_score,
+             bbox_pred,
+             rois,
+             labels,
+             label_weights,
+             bbox_targets,
+             bbox_weights,
+             reduction_override=None):
+        losses = dict()
+        if cls_score is not None:
+            avg_factor = max(torch.sum(label_weights > 0).float().item(), 1.)
+            if cls_score.numel() > 0:
+                losses['loss_cls'] = self.loss_cls(
+                    cls_score,
+                    labels,
+                    label_weights,
+                    avg_factor=avg_factor,
+                    reduction_override=reduction_override)
+                losses['acc'] = accuracy(cls_score, labels)
+        if bbox_pred is not None:
+            bg_class_ind = self.num_classes
+            # 0~self.num_classes-1 are FG, self.num_classes is BG
+            pos_inds = (labels >= 0) & (labels < bg_class_ind)
+            # do not perform bounding box regression for BG anymore.
+            if pos_inds.any():
+                if self.reg_decoded_bbox:
+                    # When the regression loss (e.g. `IouLoss`,
+                    # `GIouLoss`, `DIouLoss`) is applied directly on
+                    # the decoded bounding boxes, it decodes the
+                    # already encoded coordinates to absolute format.
+                    bbox_pred = self.bbox_coder.decode(rois[:, 1:], bbox_pred)
+                if self.reg_class_agnostic:
+                    pos_bbox_pred = bbox_pred.view(
+                        bbox_pred.size(0), 4)[pos_inds.type(torch.bool)]
+                else:
+                    pos_bbox_pred = bbox_pred.view(
+                        bbox_pred.size(0), -1,
+                        4)[pos_inds.type(torch.bool),
+                           labels[pos_inds.type(torch.bool)]]
+                losses['loss_bbox'] = self.loss_bbox(
+                    pos_bbox_pred,
+                    bbox_targets[pos_inds.type(torch.bool)],
+                    bbox_weights[pos_inds.type(torch.bool)],
+                    avg_factor=bbox_targets.size(0),
+                    reduction_override=reduction_override)
+            else:
+                losses['loss_bbox'] = bbox_pred[pos_inds].sum()
+        return losses
+
+    @force_fp32(apply_to=('cls_score', 'bbox_pred'))
+    def get_bboxes(self,
+                   rois,
+                   cls_score,
+                   bbox_pred,
+                   img_shape,
+                   scale_factor,
+                   rescale=False,
+                   cfg=None):
+        """Transform network output for a batch into bbox predictions.
+
+        If the input rois has batch dimension, the function would be in
+        `batch_mode` and return is a tuple[list[Tensor], list[Tensor]],
+        otherwise, the return is a tuple[Tensor, Tensor].
+
+        Args:
+            rois (Tensor): Boxes to be transformed. Has shape (num_boxes, 5)
+               or (B, num_boxes, 5)
+            cls_score (list[Tensor] or Tensor): Box scores for
+               each scale level, each is a 4D-tensor, the channel number is
+               num_points * num_classes.
+            bbox_pred (Tensor, optional): Box energies / deltas for each scale
+                level, each is a 4D-tensor, the channel number is
+                num_classes * 4.
+            img_shape (Sequence[int] or torch.Tensor or Sequence[
+                Sequence[int]], optional): Maximum bounds for boxes, specifies
+                (H, W, C) or (H, W). If rois shape is (B, num_boxes, 4), then
+                the max_shape should be a Sequence[Sequence[int]]
+                and the length of max_shape should also be B.
+            scale_factor (tuple[ndarray] or ndarray): Scale factor of the
+               image arange as (w_scale, h_scale, w_scale, h_scale). In
+               `batch_mode`, the scale_factor shape is tuple[ndarray].
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+            cfg (obj:`ConfigDict`): `test_cfg` of Bbox Head. Default: None
+
+        Returns:
+            tuple[list[Tensor], list[Tensor]] or tuple[Tensor, Tensor]:
+                If the input has a batch dimension, the return value is
+                a tuple of the list. The first list contains the boxes of
+                the corresponding image in a batch, each tensor has the
+                shape (num_boxes, 5) and last dimension 5 represent
+                (tl_x, tl_y, br_x, br_y, score). Each Tensor in the second
+                list is the labels with shape (num_boxes, ). The length of
+                both lists should be equal to batch_size. Otherwise return
+                value is a tuple of two tensors, the first tensor is the
+                boxes with scores, the second tensor is the labels, both
+                have the same shape as the first case.
+        """
+        if isinstance(cls_score, list):
+            cls_score = sum(cls_score) / float(len(cls_score))
+
+        scores = F.softmax(
+            cls_score, dim=-1) if cls_score is not None else None
+
+        batch_mode = True
+        if rois.ndim == 2:
+            # e.g. AugTest, Cascade R-CNN, HTC, SCNet...
+            batch_mode = False
+
+            # add batch dimension
+            if scores is not None:
+                scores = scores.unsqueeze(0)
+            if bbox_pred is not None:
+                bbox_pred = bbox_pred.unsqueeze(0)
+            rois = rois.unsqueeze(0)
+
+        if bbox_pred is not None:
+            bboxes = self.bbox_coder.decode(
+                rois[..., 1:], bbox_pred, max_shape=img_shape)
+        else:
+            bboxes = rois[..., 1:].clone()
+            if img_shape is not None:
+                max_shape = bboxes.new_tensor(img_shape)[..., :2]
+                min_xy = bboxes.new_tensor(0)
+                max_xy = torch.cat(
+                    [max_shape] * 2, dim=-1).flip(-1).unsqueeze(-2)
+                bboxes = torch.where(bboxes < min_xy, min_xy, bboxes)
+                bboxes = torch.where(bboxes > max_xy, max_xy, bboxes)
+
+        if rescale and bboxes.size(-2) > 0:
+            if not isinstance(scale_factor, tuple):
+                scale_factor = tuple([scale_factor])
+            # B, 1, bboxes.size(-1)
+            scale_factor = bboxes.new_tensor(scale_factor).unsqueeze(1).repeat(
+                1, 1,
+                bboxes.size(-1) // 4)
+            bboxes /= scale_factor
+
+        det_bboxes = []
+        det_labels = []
+        for (bbox, score) in zip(bboxes, scores):
+            if cfg is not None:
+                det_bbox, det_label = multiclass_nms(bbox, score,
+                                                     cfg.score_thr, cfg.nms,
+                                                     cfg.max_per_img)
+            else:
+                det_bbox, det_label = bbox, score
+            det_bboxes.append(det_bbox)
+            det_labels.append(det_label)
+
+        if not batch_mode:
+            det_bboxes = det_bboxes[0]
+            det_labels = det_labels[0]
+        return det_bboxes, det_labels
+
+    @force_fp32(apply_to=('bbox_preds', ))
+    def refine_bboxes(self, rois, labels, bbox_preds, pos_is_gts, img_metas):
+        """Refine bboxes during training.
+
+        Args:
+            rois (Tensor): Shape (n*bs, 5), where n is image number per GPU,
+                and bs is the sampled RoIs per image. The first column is
+                the image id and the next 4 columns are x1, y1, x2, y2.
+            labels (Tensor): Shape (n*bs, ).
+            bbox_preds (Tensor): Shape (n*bs, 4) or (n*bs, 4*#class).
+            pos_is_gts (list[Tensor]): Flags indicating if each positive bbox
+                is a gt bbox.
+            img_metas (list[dict]): Meta info of each image.
+
+        Returns:
+            list[Tensor]: Refined bboxes of each image in a mini-batch.
+
+        Example:
+            >>> # xdoctest: +REQUIRES(module:kwarray)
+            >>> import kwarray
+            >>> import numpy as np
+            >>> from mmdet.core.bbox.demodata import random_boxes
+            >>> self = BBoxHead(reg_class_agnostic=True)
+            >>> n_roi = 2
+            >>> n_img = 4
+            >>> scale = 512
+            >>> rng = np.random.RandomState(0)
+            >>> img_metas = [{'img_shape': (scale, scale)}
+            ...              for _ in range(n_img)]
+            >>> # Create rois in the expected format
+            >>> roi_boxes = random_boxes(n_roi, scale=scale, rng=rng)
+            >>> img_ids = torch.randint(0, n_img, (n_roi,))
+            >>> img_ids = img_ids.float()
+            >>> rois = torch.cat([img_ids[:, None], roi_boxes], dim=1)
+            >>> # Create other args
+            >>> labels = torch.randint(0, 2, (n_roi,)).long()
+            >>> bbox_preds = random_boxes(n_roi, scale=scale, rng=rng)
+            >>> # For each image, pretend random positive boxes are gts
+            >>> is_label_pos = (labels.numpy() > 0).astype(np.int)
+            >>> lbl_per_img = kwarray.group_items(is_label_pos,
+            ...                                   img_ids.numpy())
+            >>> pos_per_img = [sum(lbl_per_img.get(gid, []))
+            ...                for gid in range(n_img)]
+            >>> pos_is_gts = [
+            >>>     torch.randint(0, 2, (npos,)).byte().sort(
+            >>>         descending=True)[0]
+            >>>     for npos in pos_per_img
+            >>> ]
+            >>> bboxes_list = self.refine_bboxes(rois, labels, bbox_preds,
+            >>>                    pos_is_gts, img_metas)
+            >>> print(bboxes_list)
+        """
+        img_ids = rois[:, 0].long().unique(sorted=True)
+        assert img_ids.numel() <= len(img_metas)
+
+        bboxes_list = []
+        for i in range(len(img_metas)):
+            inds = torch.nonzero(
+                rois[:, 0] == i, as_tuple=False).squeeze(dim=1)
+            num_rois = inds.numel()
+
+            bboxes_ = rois[inds, 1:]
+            label_ = labels[inds]
+            bbox_pred_ = bbox_preds[inds]
+            img_meta_ = img_metas[i]
+            pos_is_gts_ = pos_is_gts[i]
+
+            bboxes = self.regress_by_class(bboxes_, label_, bbox_pred_,
+                                           img_meta_)
+
+            # filter gt bboxes
+            pos_keep = 1 - pos_is_gts_
+            keep_inds = pos_is_gts_.new_ones(num_rois)
+            keep_inds[:len(pos_is_gts_)] = pos_keep
+
+            bboxes_list.append(bboxes[keep_inds.type(torch.bool)])
+
+        return bboxes_list
+
+    @force_fp32(apply_to=('bbox_pred', ))
+    def regress_by_class(self, rois, label, bbox_pred, img_meta):
+        """Regress the bbox for the predicted class. Used in Cascade R-CNN.
+
+        Args:
+            rois (Tensor): shape (n, 4) or (n, 5)
+            label (Tensor): shape (n, )
+            bbox_pred (Tensor): shape (n, 4*(#class)) or (n, 4)
+            img_meta (dict): Image meta info.
+
+        Returns:
+            Tensor: Regressed bboxes, the same shape as input rois.
+        """
+        assert rois.size(1) == 4 or rois.size(1) == 5, repr(rois.shape)
+
+        if not self.reg_class_agnostic:
+            label = label * 4
+            inds = torch.stack((label, label + 1, label + 2, label + 3), 1)
+            bbox_pred = torch.gather(bbox_pred, 1, inds)
+        assert bbox_pred.size(1) == 4
+
+        if rois.size(1) == 4:
+            new_rois = self.bbox_coder.decode(
+                rois, bbox_pred, max_shape=img_meta['img_shape'])
+        else:
+            bboxes = self.bbox_coder.decode(
+                rois[:, 1:], bbox_pred, max_shape=img_meta['img_shape'])
+            new_rois = torch.cat((rois[:, [0]], bboxes), dim=1)
+
+        return new_rois
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/convfc_bbox_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/convfc_bbox_head.py
new file mode 100644
index 0000000..0e86d2e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/convfc_bbox_head.py
@@ -0,0 +1,205 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule
+
+from mmdet.models.builder import HEADS
+from .bbox_head import BBoxHead
+
+
+@HEADS.register_module()
+class ConvFCBBoxHead(BBoxHead):
+    r"""More general bbox head, with shared conv and fc layers and two optional
+    separated branches.
+
+    .. code-block:: none
+
+                                    /-> cls convs -> cls fcs -> cls
+        shared convs -> shared fcs
+                                    \-> reg convs -> reg fcs -> reg
+    """  # noqa: W605
+
+    def __init__(self,
+                 num_shared_convs=0,
+                 num_shared_fcs=0,
+                 num_cls_convs=0,
+                 num_cls_fcs=0,
+                 num_reg_convs=0,
+                 num_reg_fcs=0,
+                 conv_out_channels=256,
+                 fc_out_channels=1024,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 *args,
+                 **kwargs):
+        super(ConvFCBBoxHead, self).__init__(*args, **kwargs)
+        assert (num_shared_convs + num_shared_fcs + num_cls_convs +
+                num_cls_fcs + num_reg_convs + num_reg_fcs > 0)
+        if num_cls_convs > 0 or num_reg_convs > 0:
+            assert num_shared_fcs == 0
+        if not self.with_cls:
+            assert num_cls_convs == 0 and num_cls_fcs == 0
+        if not self.with_reg:
+            assert num_reg_convs == 0 and num_reg_fcs == 0
+        self.num_shared_convs = num_shared_convs
+        self.num_shared_fcs = num_shared_fcs
+        self.num_cls_convs = num_cls_convs
+        self.num_cls_fcs = num_cls_fcs
+        self.num_reg_convs = num_reg_convs
+        self.num_reg_fcs = num_reg_fcs
+        self.conv_out_channels = conv_out_channels
+        self.fc_out_channels = fc_out_channels
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+
+        # add shared convs and fcs
+        self.shared_convs, self.shared_fcs, last_layer_dim = \
+            self._add_conv_fc_branch(
+                self.num_shared_convs, self.num_shared_fcs, self.in_channels,
+                True)
+        self.shared_out_channels = last_layer_dim
+
+        # add cls specific branch
+        self.cls_convs, self.cls_fcs, self.cls_last_dim = \
+            self._add_conv_fc_branch(
+                self.num_cls_convs, self.num_cls_fcs, self.shared_out_channels)
+
+        # add reg specific branch
+        self.reg_convs, self.reg_fcs, self.reg_last_dim = \
+            self._add_conv_fc_branch(
+                self.num_reg_convs, self.num_reg_fcs, self.shared_out_channels)
+
+        if self.num_shared_fcs == 0 and not self.with_avg_pool:
+            if self.num_cls_fcs == 0:
+                self.cls_last_dim *= self.roi_feat_area
+            if self.num_reg_fcs == 0:
+                self.reg_last_dim *= self.roi_feat_area
+
+        self.relu = nn.ReLU(inplace=True)
+        # reconstruct fc_cls and fc_reg since input channels are changed
+        if self.with_cls:
+            self.fc_cls = nn.Linear(self.cls_last_dim, self.num_classes + 1)
+        if self.with_reg:
+            out_dim_reg = (4 if self.reg_class_agnostic else 4 *
+                           self.num_classes)
+            self.fc_reg = nn.Linear(self.reg_last_dim, out_dim_reg)
+
+    def _add_conv_fc_branch(self,
+                            num_branch_convs,
+                            num_branch_fcs,
+                            in_channels,
+                            is_shared=False):
+        """Add shared or separable branch.
+
+        convs -> avg pool (optional) -> fcs
+        """
+        last_layer_dim = in_channels
+        # add branch specific conv layers
+        branch_convs = nn.ModuleList()
+        if num_branch_convs > 0:
+            for i in range(num_branch_convs):
+                conv_in_channels = (
+                    last_layer_dim if i == 0 else self.conv_out_channels)
+                branch_convs.append(
+                    ConvModule(
+                        conv_in_channels,
+                        self.conv_out_channels,
+                        3,
+                        padding=1,
+                        conv_cfg=self.conv_cfg,
+                        norm_cfg=self.norm_cfg))
+            last_layer_dim = self.conv_out_channels
+        # add branch specific fc layers
+        branch_fcs = nn.ModuleList()
+        if num_branch_fcs > 0:
+            # for shared branch, only consider self.with_avg_pool
+            # for separated branches, also consider self.num_shared_fcs
+            if (is_shared
+                    or self.num_shared_fcs == 0) and not self.with_avg_pool:
+                last_layer_dim *= self.roi_feat_area
+            for i in range(num_branch_fcs):
+                fc_in_channels = (
+                    last_layer_dim if i == 0 else self.fc_out_channels)
+                branch_fcs.append(
+                    nn.Linear(fc_in_channels, self.fc_out_channels))
+            last_layer_dim = self.fc_out_channels
+        return branch_convs, branch_fcs, last_layer_dim
+
+    def init_weights(self):
+        super(ConvFCBBoxHead, self).init_weights()
+        # conv layers are already initialized by ConvModule
+        for module_list in [self.shared_fcs, self.cls_fcs, self.reg_fcs]:
+            for m in module_list.modules():
+                if isinstance(m, nn.Linear):
+                    nn.init.xavier_uniform_(m.weight)
+                    nn.init.constant_(m.bias, 0)
+
+    def forward(self, x):
+        # shared part
+        if self.num_shared_convs > 0:
+            for conv in self.shared_convs:
+                x = conv(x)
+
+        if self.num_shared_fcs > 0:
+            if self.with_avg_pool:
+                x = self.avg_pool(x)
+
+            x = x.flatten(1)
+
+            for fc in self.shared_fcs:
+                x = self.relu(fc(x))
+        # separate branches
+        x_cls = x
+        x_reg = x
+
+        for conv in self.cls_convs:
+            x_cls = conv(x_cls)
+        if x_cls.dim() > 2:
+            if self.with_avg_pool:
+                x_cls = self.avg_pool(x_cls)
+            x_cls = x_cls.flatten(1)
+        for fc in self.cls_fcs:
+            x_cls = self.relu(fc(x_cls))
+
+        for conv in self.reg_convs:
+            x_reg = conv(x_reg)
+        if x_reg.dim() > 2:
+            if self.with_avg_pool:
+                x_reg = self.avg_pool(x_reg)
+            x_reg = x_reg.flatten(1)
+        for fc in self.reg_fcs:
+            x_reg = self.relu(fc(x_reg))
+
+        cls_score = self.fc_cls(x_cls) if self.with_cls else None
+        bbox_pred = self.fc_reg(x_reg) if self.with_reg else None
+        return cls_score, bbox_pred
+
+
+@HEADS.register_module()
+class Shared2FCBBoxHead(ConvFCBBoxHead):
+
+    def __init__(self, fc_out_channels=1024, *args, **kwargs):
+        super(Shared2FCBBoxHead, self).__init__(
+            num_shared_convs=0,
+            num_shared_fcs=2,
+            num_cls_convs=0,
+            num_cls_fcs=0,
+            num_reg_convs=0,
+            num_reg_fcs=0,
+            fc_out_channels=fc_out_channels,
+            *args,
+            **kwargs)
+
+
+@HEADS.register_module()
+class Shared4Conv1FCBBoxHead(ConvFCBBoxHead):
+
+    def __init__(self, fc_out_channels=1024, *args, **kwargs):
+        super(Shared4Conv1FCBBoxHead, self).__init__(
+            num_shared_convs=4,
+            num_shared_fcs=1,
+            num_cls_convs=0,
+            num_cls_fcs=0,
+            num_reg_convs=0,
+            num_reg_fcs=0,
+            fc_out_channels=fc_out_channels,
+            *args,
+            **kwargs)
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/dii_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/dii_head.py
new file mode 100644
index 0000000..8c970a7
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/dii_head.py
@@ -0,0 +1,415 @@
+import torch
+import torch.nn as nn
+from mmcv.cnn import (bias_init_with_prob, build_activation_layer,
+                      build_norm_layer)
+from mmcv.runner import auto_fp16, force_fp32
+
+from mmdet.core import multi_apply
+from mmdet.models.builder import HEADS, build_loss
+from mmdet.models.dense_heads.atss_head import reduce_mean
+from mmdet.models.losses import accuracy
+from mmdet.models.utils import FFN, MultiheadAttention, build_transformer
+from .bbox_head import BBoxHead
+
+
+@HEADS.register_module()
+class DIIHead(BBoxHead):
+    r"""Dynamic Instance Interactive Head for `Sparse R-CNN: End-to-End Object
+    Detection with Learnable Proposals <https://arxiv.org/abs/2011.12450>`_
+
+    Args:
+        num_classes (int): Number of class in dataset.
+            Defaults to 80.
+        num_ffn_fcs (int): The number of fully-connected
+            layers in FFNs. Defaults to 2.
+        num_heads (int): The hidden dimension of FFNs.
+            Defaults to 8.
+        num_cls_fcs (int): The number of fully-connected
+            layers in classification subnet. Defaults to 1.
+        num_reg_fcs (int): The number of fully-connected
+            layers in regression subnet. Defaults to 3.
+        feedforward_channels (int): The hidden dimension
+            of FFNs. Defaults to 2048
+        in_channels (int): Hidden_channels of MultiheadAttention.
+            Defaults to 256.
+        dropout (float): Probability of drop the channel.
+            Defaults to 0.0
+        ffn_act_cfg (dict): The activation config for FFNs.
+        dynamic_conv_cfg (dict): The convolution config
+            for DynamicConv.
+        loss_iou (dict): The config for iou or giou loss.
+
+    """
+
+    def __init__(self,
+                 num_classes=80,
+                 num_ffn_fcs=2,
+                 num_heads=8,
+                 num_cls_fcs=1,
+                 num_reg_fcs=3,
+                 feedforward_channels=2048,
+                 in_channels=256,
+                 dropout=0.0,
+                 ffn_act_cfg=dict(type='ReLU', inplace=True),
+                 dynamic_conv_cfg=dict(
+                     type='DynamicConv',
+                     in_channels=256,
+                     feat_channels=64,
+                     out_channels=256,
+                     input_feat_shape=7,
+                     act_cfg=dict(type='ReLU', inplace=True),
+                     norm_cfg=dict(type='LN')),
+                 loss_iou=dict(type='GIoULoss', loss_weight=2.0),
+                 **kwargs):
+        super(DIIHead, self).__init__(
+            num_classes=num_classes,
+            reg_decoded_bbox=True,
+            reg_class_agnostic=True,
+            **kwargs)
+        self.loss_iou = build_loss(loss_iou)
+        self.in_channels = in_channels
+        self.fp16_enabled = False
+        self.attention = MultiheadAttention(in_channels, num_heads, dropout)
+        self.attention_norm = build_norm_layer(dict(type='LN'), in_channels)[1]
+
+        self.instance_interactive_conv = build_transformer(dynamic_conv_cfg)
+        self.instance_interactive_conv_dropout = nn.Dropout(dropout)
+        self.instance_interactive_conv_norm = build_norm_layer(
+            dict(type='LN'), in_channels)[1]
+
+        self.ffn = FFN(
+            in_channels,
+            feedforward_channels,
+            num_ffn_fcs,
+            act_cfg=ffn_act_cfg,
+            dropout=dropout)
+        self.ffn_norm = build_norm_layer(dict(type='LN'), in_channels)[1]
+
+        self.cls_fcs = nn.ModuleList()
+        for _ in range(num_cls_fcs):
+            self.cls_fcs.append(
+                nn.Linear(in_channels, in_channels, bias=False))
+            self.cls_fcs.append(
+                build_norm_layer(dict(type='LN'), in_channels)[1])
+            self.cls_fcs.append(
+                build_activation_layer(dict(type='ReLU', inplace=True)))
+
+        # over load the self.fc_cls in BBoxHead
+        if self.loss_cls.use_sigmoid:
+            self.fc_cls = nn.Linear(in_channels, self.num_classes)
+        else:
+            self.fc_cls = nn.Linear(in_channels, self.num_classes + 1)
+
+        self.reg_fcs = nn.ModuleList()
+        for _ in range(num_reg_fcs):
+            self.reg_fcs.append(
+                nn.Linear(in_channels, in_channels, bias=False))
+            self.reg_fcs.append(
+                build_norm_layer(dict(type='LN'), in_channels)[1])
+            self.reg_fcs.append(
+                build_activation_layer(dict(type='ReLU', inplace=True)))
+        # over load the self.fc_cls in BBoxHead
+        self.fc_reg = nn.Linear(in_channels, 4)
+
+        assert self.reg_class_agnostic, 'DIIHead only ' \
+            'suppport `reg_class_agnostic=True` '
+        assert self.reg_decoded_bbox, 'DIIHead only ' \
+            'suppport `reg_decoded_bbox=True`'
+
+    def init_weights(self):
+        """Use xavier initialization for all weight parameter and set
+        classification head bias as a specific value when use focal loss."""
+        for p in self.parameters():
+            if p.dim() > 1:
+                nn.init.xavier_uniform_(p)
+            else:
+                # adopt the default initialization for
+                # the weight and bias of the layer norm
+                pass
+        if self.loss_cls.use_sigmoid:
+            bias_init = bias_init_with_prob(0.01)
+            nn.init.constant_(self.fc_cls.bias, bias_init)
+
+    @auto_fp16()
+    def forward(self, roi_feat, proposal_feat):
+        """Forward function of Dynamic Instance Interactive Head.
+
+        Args:
+            roi_feat (Tensor): Roi-pooling features with shape
+                (batch_size*num_proposals, feature_dimensions,
+                pooling_h , pooling_w).
+            proposal_feat (Tensor): Intermediate feature get from
+                diihead in last stage, has shape
+                (batch_size, num_proposals, feature_dimensions)
+
+          Returns:
+                tuple[Tensor]: Usually a tuple of classification scores
+                and bbox prediction and a intermediate feature.
+
+                    - cls_scores (Tensor): Classification scores for
+                      all proposals, has shape
+                      (batch_size, num_proposals, num_classes).
+                    - bbox_preds (Tensor): Box energies / deltas for
+                      all proposals, has shape
+                      (batch_size, num_proposals, 4).
+                    - obj_feat (Tensor): Object feature before classification
+                      and regression subnet, has shape
+                      (batch_size, num_proposal, feature_dimensions).
+        """
+        N, num_proposals = proposal_feat.shape[:2]
+
+        # Self attention
+        proposal_feat = proposal_feat.permute(1, 0, 2)
+        proposal_feat = self.attention_norm(self.attention(proposal_feat))
+
+        # instance interactive
+        proposal_feat = proposal_feat.permute(1, 0,
+                                              2).reshape(-1, self.in_channels)
+        proposal_feat_iic = self.instance_interactive_conv(
+            proposal_feat, roi_feat)
+        proposal_feat = proposal_feat + self.instance_interactive_conv_dropout(
+            proposal_feat_iic)
+        obj_feat = self.instance_interactive_conv_norm(proposal_feat)
+
+        # FFN
+        obj_feat = self.ffn_norm(self.ffn(obj_feat))
+
+        cls_feat = obj_feat
+        reg_feat = obj_feat
+
+        for cls_layer in self.cls_fcs:
+            cls_feat = cls_layer(cls_feat)
+        for reg_layer in self.reg_fcs:
+            reg_feat = reg_layer(reg_feat)
+
+        cls_score = self.fc_cls(cls_feat).view(N, num_proposals, -1)
+        bbox_delta = self.fc_reg(reg_feat).view(N, num_proposals, -1)
+
+        return cls_score, bbox_delta, obj_feat.view(N, num_proposals, -1)
+
+    @force_fp32(apply_to=('cls_score', 'bbox_pred'))
+    def loss(self,
+             cls_score,
+             bbox_pred,
+             labels,
+             label_weights,
+             bbox_targets,
+             bbox_weights,
+             imgs_whwh=None,
+             reduction_override=None,
+             **kwargs):
+        """"Loss function of DIIHead, get loss of all images.
+
+        Args:
+            cls_score (Tensor): Classification prediction
+                results of all class, has shape
+                (batch_size * num_proposals_single_image, num_classes)
+            bbox_pred (Tensor): Regression prediction results,
+                has shape
+                (batch_size * num_proposals_single_image, 4), the last
+                dimension 4 represents [tl_x, tl_y, br_x, br_y].
+            labels (Tensor): Label of each proposals, has shape
+                (batch_size * num_proposals_single_image
+            label_weights (Tensor): Classification loss
+                weight of each proposals, has shape
+                (batch_size * num_proposals_single_image
+            bbox_targets (Tensor): Regression targets of each
+                proposals, has shape
+                (batch_size * num_proposals_single_image, 4),
+                the last dimension 4 represents
+                [tl_x, tl_y, br_x, br_y].
+            bbox_weights (Tensor): Regression loss weight of each
+                proposals's coordinate, has shape
+                (batch_size * num_proposals_single_image, 4),
+            imgs_whwh (Tensor): imgs_whwh (Tensor): Tensor with\
+                shape (batch_size, num_proposals, 4), the last
+                dimension means
+                [img_width,img_height, img_width, img_height].
+            reduction_override (str, optional): The reduction
+                method used to override the original reduction
+                method of the loss. Options are "none",
+                "mean" and "sum". Defaults to None,
+
+            Returns:
+                dict[str, Tensor]: Dictionary of loss components
+        """
+        losses = dict()
+        bg_class_ind = self.num_classes
+        # note in spare rcnn num_gt == num_pos
+        pos_inds = (labels >= 0) & (labels < bg_class_ind)
+        num_pos = pos_inds.sum().float()
+        avg_factor = reduce_mean(num_pos)
+        if cls_score is not None:
+            if cls_score.numel() > 0:
+                losses['loss_cls'] = self.loss_cls(
+                    cls_score,
+                    labels,
+                    label_weights,
+                    avg_factor=avg_factor,
+                    reduction_override=reduction_override)
+                losses['pos_acc'] = accuracy(cls_score[pos_inds],
+                                             labels[pos_inds])
+        if bbox_pred is not None:
+            # 0~self.num_classes-1 are FG, self.num_classes is BG
+            # do not perform bounding box regression for BG anymore.
+            if pos_inds.any():
+                pos_bbox_pred = bbox_pred.reshape(bbox_pred.size(0),
+                                                  4)[pos_inds.type(torch.bool)]
+                imgs_whwh = imgs_whwh.reshape(bbox_pred.size(0),
+                                              4)[pos_inds.type(torch.bool)]
+                losses['loss_bbox'] = self.loss_bbox(
+                    pos_bbox_pred / imgs_whwh,
+                    bbox_targets[pos_inds.type(torch.bool)] / imgs_whwh,
+                    bbox_weights[pos_inds.type(torch.bool)],
+                    avg_factor=avg_factor)
+                losses['loss_iou'] = self.loss_iou(
+                    pos_bbox_pred,
+                    bbox_targets[pos_inds.type(torch.bool)],
+                    bbox_weights[pos_inds.type(torch.bool)],
+                    avg_factor=avg_factor)
+            else:
+                losses['loss_bbox'] = bbox_pred.sum() * 0
+                losses['loss_iou'] = bbox_pred.sum() * 0
+        return losses
+
+    def _get_target_single(self, pos_inds, neg_inds, pos_bboxes, neg_bboxes,
+                           pos_gt_bboxes, pos_gt_labels, cfg):
+        """Calculate the ground truth for proposals in the single image
+        according to the sampling results.
+
+        Almost the same as the implementation in `bbox_head`,
+        we add pos_inds and neg_inds to select positive and
+        negative samples instead of selecting the first num_pos
+        as positive samples.
+
+        Args:
+            pos_inds (Tensor): The length is equal to the
+                positive sample numbers contain all index
+                of the positive sample in the origin proposal set.
+            neg_inds (Tensor): The length is equal to the
+                negative sample numbers contain all index
+                of the negative sample in the origin proposal set.
+            pos_bboxes (Tensor): Contains all the positive boxes,
+                has shape (num_pos, 4), the last dimension 4
+                represents [tl_x, tl_y, br_x, br_y].
+            neg_bboxes (Tensor): Contains all the negative boxes,
+                has shape (num_neg, 4), the last dimension 4
+                represents [tl_x, tl_y, br_x, br_y].
+            pos_gt_bboxes (Tensor): Contains all the gt_boxes,
+                has shape (num_gt, 4), the last dimension 4
+                represents [tl_x, tl_y, br_x, br_y].
+            pos_gt_labels (Tensor): Contains all the gt_labels,
+                has shape (num_gt).
+            cfg (obj:`ConfigDict`): `train_cfg` of R-CNN.
+
+        Returns:
+            Tuple[Tensor]: Ground truth for proposals in a single image.
+            Containing the following Tensors:
+
+                - labels(Tensor): Gt_labels for all proposals, has
+                  shape (num_proposals,).
+                - label_weights(Tensor): Labels_weights for all proposals, has
+                  shape (num_proposals,).
+                - bbox_targets(Tensor):Regression target for all proposals, has
+                  shape (num_proposals, 4), the last dimension 4
+                  represents [tl_x, tl_y, br_x, br_y].
+                - bbox_weights(Tensor):Regression weights for all proposals,
+                  has shape (num_proposals, 4).
+        """
+        num_pos = pos_bboxes.size(0)
+        num_neg = neg_bboxes.size(0)
+        num_samples = num_pos + num_neg
+
+        # original implementation uses new_zeros since BG are set to be 0
+        # now use empty & fill because BG cat_id = num_classes,
+        # FG cat_id = [0, num_classes-1]
+        labels = pos_bboxes.new_full((num_samples, ),
+                                     self.num_classes,
+                                     dtype=torch.long)
+        label_weights = pos_bboxes.new_zeros(num_samples)
+        bbox_targets = pos_bboxes.new_zeros(num_samples, 4)
+        bbox_weights = pos_bboxes.new_zeros(num_samples, 4)
+        if num_pos > 0:
+            labels[pos_inds] = pos_gt_labels
+            pos_weight = 1.0 if cfg.pos_weight <= 0 else cfg.pos_weight
+            label_weights[pos_inds] = pos_weight
+            if not self.reg_decoded_bbox:
+                pos_bbox_targets = self.bbox_coder.encode(
+                    pos_bboxes, pos_gt_bboxes)
+            else:
+                pos_bbox_targets = pos_gt_bboxes
+            bbox_targets[pos_inds, :] = pos_bbox_targets
+            bbox_weights[pos_inds, :] = 1
+        if num_neg > 0:
+            label_weights[neg_inds] = 1.0
+
+        return labels, label_weights, bbox_targets, bbox_weights
+
+    def get_targets(self,
+                    sampling_results,
+                    gt_bboxes,
+                    gt_labels,
+                    rcnn_train_cfg,
+                    concat=True):
+        """Calculate the ground truth for all samples in a batch according to
+        the sampling_results.
+
+        Almost the same as the implementation in bbox_head, we passed
+        additional parameters pos_inds_list and neg_inds_list to
+        `_get_target_single` function.
+
+        Args:
+            sampling_results (List[obj:SamplingResults]): Assign results of
+                all images in a batch after sampling.
+            gt_bboxes (list[Tensor]): Gt_bboxes of all images in a batch,
+                each tensor has shape (num_gt, 4),  the last dimension 4
+                represents [tl_x, tl_y, br_x, br_y].
+            gt_labels (list[Tensor]): Gt_labels of all images in a batch,
+                each tensor has shape (num_gt,).
+            rcnn_train_cfg (obj:`ConfigDict`): `train_cfg` of RCNN.
+            concat (bool): Whether to concatenate the results of all
+                the images in a single batch.
+
+        Returns:
+            Tuple[Tensor]: Ground truth for proposals in a single image.
+            Containing the following list of Tensors:
+
+                - labels (list[Tensor],Tensor): Gt_labels for all
+                  proposals in a batch, each tensor in list has
+                  shape (num_proposals,) when `concat=False`, otherwise just
+                  a single tensor has shape (num_all_proposals,).
+                - label_weights (list[Tensor]): Labels_weights for
+                  all proposals in a batch, each tensor in list has shape
+                  (num_proposals,) when `concat=False`, otherwise just a
+                  single tensor has shape (num_all_proposals,).
+                - bbox_targets (list[Tensor],Tensor): Regression target
+                  for all proposals in a batch, each tensor in list has
+                  shape (num_proposals, 4) when `concat=False`, otherwise
+                  just a single tensor has shape (num_all_proposals, 4),
+                  the last dimension 4 represents [tl_x, tl_y, br_x, br_y].
+                - bbox_weights (list[tensor],Tensor): Regression weights for
+                  all proposals in a batch, each tensor in list has shape
+                  (num_proposals, 4) when `concat=False`, otherwise just a
+                  single tensor has shape (num_all_proposals, 4).
+        """
+        pos_inds_list = [res.pos_inds for res in sampling_results]
+        neg_inds_list = [res.neg_inds for res in sampling_results]
+        pos_bboxes_list = [res.pos_bboxes for res in sampling_results]
+        neg_bboxes_list = [res.neg_bboxes for res in sampling_results]
+        pos_gt_bboxes_list = [res.pos_gt_bboxes for res in sampling_results]
+        pos_gt_labels_list = [res.pos_gt_labels for res in sampling_results]
+        labels, label_weights, bbox_targets, bbox_weights = multi_apply(
+            self._get_target_single,
+            pos_inds_list,
+            neg_inds_list,
+            pos_bboxes_list,
+            neg_bboxes_list,
+            pos_gt_bboxes_list,
+            pos_gt_labels_list,
+            cfg=rcnn_train_cfg)
+        if concat:
+            labels = torch.cat(labels, 0)
+            label_weights = torch.cat(label_weights, 0)
+            bbox_targets = torch.cat(bbox_targets, 0)
+            bbox_weights = torch.cat(bbox_weights, 0)
+        return labels, label_weights, bbox_targets, bbox_weights
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/double_bbox_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/double_bbox_head.py
new file mode 100644
index 0000000..6c154cb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/double_bbox_head.py
@@ -0,0 +1,172 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, normal_init, xavier_init
+
+from mmdet.models.backbones.resnet import Bottleneck
+from mmdet.models.builder import HEADS
+from .bbox_head import BBoxHead
+
+
+class BasicResBlock(nn.Module):
+    """Basic residual block.
+
+    This block is a little different from the block in the ResNet backbone.
+    The kernel size of conv1 is 1 in this block while 3 in ResNet BasicBlock.
+
+    Args:
+        in_channels (int): Channels of the input feature map.
+        out_channels (int): Channels of the output feature map.
+        conv_cfg (dict): The config dict for convolution layers.
+        norm_cfg (dict): The config dict for normalization layers.
+    """
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN')):
+        super(BasicResBlock, self).__init__()
+
+        # main path
+        self.conv1 = ConvModule(
+            in_channels,
+            in_channels,
+            kernel_size=3,
+            padding=1,
+            bias=False,
+            conv_cfg=conv_cfg,
+            norm_cfg=norm_cfg)
+        self.conv2 = ConvModule(
+            in_channels,
+            out_channels,
+            kernel_size=1,
+            bias=False,
+            conv_cfg=conv_cfg,
+            norm_cfg=norm_cfg,
+            act_cfg=None)
+
+        # identity path
+        self.conv_identity = ConvModule(
+            in_channels,
+            out_channels,
+            kernel_size=1,
+            conv_cfg=conv_cfg,
+            norm_cfg=norm_cfg,
+            act_cfg=None)
+
+        self.relu = nn.ReLU(inplace=True)
+
+    def forward(self, x):
+        identity = x
+
+        x = self.conv1(x)
+        x = self.conv2(x)
+
+        identity = self.conv_identity(identity)
+        out = x + identity
+
+        out = self.relu(out)
+        return out
+
+
+@HEADS.register_module()
+class DoubleConvFCBBoxHead(BBoxHead):
+    r"""Bbox head used in Double-Head R-CNN
+
+    .. code-block:: none
+
+                                          /-> cls
+                      /-> shared convs ->
+                                          \-> reg
+        roi features
+                                          /-> cls
+                      \-> shared fc    ->
+                                          \-> reg
+    """  # noqa: W605
+
+    def __init__(self,
+                 num_convs=0,
+                 num_fcs=0,
+                 conv_out_channels=1024,
+                 fc_out_channels=1024,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 **kwargs):
+        kwargs.setdefault('with_avg_pool', True)
+        super(DoubleConvFCBBoxHead, self).__init__(**kwargs)
+        assert self.with_avg_pool
+        assert num_convs > 0
+        assert num_fcs > 0
+        self.num_convs = num_convs
+        self.num_fcs = num_fcs
+        self.conv_out_channels = conv_out_channels
+        self.fc_out_channels = fc_out_channels
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+
+        # increase the channel of input features
+        self.res_block = BasicResBlock(self.in_channels,
+                                       self.conv_out_channels)
+
+        # add conv heads
+        self.conv_branch = self._add_conv_branch()
+        # add fc heads
+        self.fc_branch = self._add_fc_branch()
+
+        out_dim_reg = 4 if self.reg_class_agnostic else 4 * self.num_classes
+        self.fc_reg = nn.Linear(self.conv_out_channels, out_dim_reg)
+
+        self.fc_cls = nn.Linear(self.fc_out_channels, self.num_classes + 1)
+        self.relu = nn.ReLU(inplace=True)
+
+    def _add_conv_branch(self):
+        """Add the fc branch which consists of a sequential of conv layers."""
+        branch_convs = nn.ModuleList()
+        for i in range(self.num_convs):
+            branch_convs.append(
+                Bottleneck(
+                    inplanes=self.conv_out_channels,
+                    planes=self.conv_out_channels // 4,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+        return branch_convs
+
+    def _add_fc_branch(self):
+        """Add the fc branch which consists of a sequential of fc layers."""
+        branch_fcs = nn.ModuleList()
+        for i in range(self.num_fcs):
+            fc_in_channels = (
+                self.in_channels *
+                self.roi_feat_area if i == 0 else self.fc_out_channels)
+            branch_fcs.append(nn.Linear(fc_in_channels, self.fc_out_channels))
+        return branch_fcs
+
+    def init_weights(self):
+        # conv layers are already initialized by ConvModule
+        normal_init(self.fc_cls, std=0.01)
+        normal_init(self.fc_reg, std=0.001)
+
+        for m in self.fc_branch.modules():
+            if isinstance(m, nn.Linear):
+                xavier_init(m, distribution='uniform')
+
+    def forward(self, x_cls, x_reg):
+        # conv head
+        x_conv = self.res_block(x_reg)
+
+        for conv in self.conv_branch:
+            x_conv = conv(x_conv)
+
+        if self.with_avg_pool:
+            x_conv = self.avg_pool(x_conv)
+
+        x_conv = x_conv.view(x_conv.size(0), -1)
+        bbox_pred = self.fc_reg(x_conv)
+
+        # fc head
+        x_fc = x_cls.view(x_cls.size(0), -1)
+        for fc in self.fc_branch:
+            x_fc = self.relu(fc(x_fc))
+
+        cls_score = self.fc_cls(x_fc)
+
+        return cls_score, bbox_pred
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/sabl_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/sabl_head.py
new file mode 100644
index 0000000..5153996
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/sabl_head.py
@@ -0,0 +1,572 @@
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, kaiming_init, normal_init, xavier_init
+from mmcv.runner import force_fp32
+
+from mmdet.core import build_bbox_coder, multi_apply, multiclass_nms
+from mmdet.models.builder import HEADS, build_loss
+from mmdet.models.losses import accuracy
+
+
+@HEADS.register_module()
+class SABLHead(nn.Module):
+    """Side-Aware Boundary Localization (SABL) for RoI-Head.
+
+    Side-Aware features are extracted by conv layers
+    with an attention mechanism.
+    Boundary Localization with Bucketing and Bucketing Guided Rescoring
+    are implemented in BucketingBBoxCoder.
+
+    Please refer to https://arxiv.org/abs/1912.04260 for more details.
+
+    Args:
+        cls_in_channels (int): Input channels of cls RoI feature. \
+            Defaults to 256.
+        reg_in_channels (int): Input channels of reg RoI feature. \
+            Defaults to 256.
+        roi_feat_size (int): Size of RoI features. Defaults to 7.
+        reg_feat_up_ratio (int): Upsample ratio of reg features. \
+            Defaults to 2.
+        reg_pre_kernel (int): Kernel of 2D conv layers before \
+            attention pooling. Defaults to 3.
+        reg_post_kernel (int): Kernel of 1D conv layers after \
+            attention pooling. Defaults to 3.
+        reg_pre_num (int): Number of pre convs. Defaults to 2.
+        reg_post_num (int): Number of post convs. Defaults to 1.
+        num_classes (int): Number of classes in dataset. Defaults to 80.
+        cls_out_channels (int): Hidden channels in cls fcs. Defaults to 1024.
+        reg_offset_out_channels (int): Hidden and output channel \
+            of reg offset branch. Defaults to 256.
+        reg_cls_out_channels (int): Hidden and output channel \
+            of reg cls branch. Defaults to 256.
+        num_cls_fcs (int): Number of fcs for cls branch. Defaults to 1.
+        num_reg_fcs (int): Number of fcs for reg branch.. Defaults to 0.
+        reg_class_agnostic (bool): Class agnostic regresion or not. \
+            Defaults to True.
+        norm_cfg (dict): Config of norm layers. Defaults to None.
+        bbox_coder (dict): Config of bbox coder. Defaults 'BucketingBBoxCoder'.
+        loss_cls (dict): Config of classification loss.
+        loss_bbox_cls (dict): Config of classification loss for bbox branch.
+        loss_bbox_reg (dict): Config of regression loss for bbox branch.
+    """
+
+    def __init__(self,
+                 num_classes,
+                 cls_in_channels=256,
+                 reg_in_channels=256,
+                 roi_feat_size=7,
+                 reg_feat_up_ratio=2,
+                 reg_pre_kernel=3,
+                 reg_post_kernel=3,
+                 reg_pre_num=2,
+                 reg_post_num=1,
+                 cls_out_channels=1024,
+                 reg_offset_out_channels=256,
+                 reg_cls_out_channels=256,
+                 num_cls_fcs=1,
+                 num_reg_fcs=0,
+                 reg_class_agnostic=True,
+                 norm_cfg=None,
+                 bbox_coder=dict(
+                     type='BucketingBBoxCoder',
+                     num_buckets=14,
+                     scale_factor=1.7),
+                 loss_cls=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=False,
+                     loss_weight=1.0),
+                 loss_bbox_cls=dict(
+                     type='CrossEntropyLoss',
+                     use_sigmoid=True,
+                     loss_weight=1.0),
+                 loss_bbox_reg=dict(
+                     type='SmoothL1Loss', beta=0.1, loss_weight=1.0)):
+        super(SABLHead, self).__init__()
+        self.cls_in_channels = cls_in_channels
+        self.reg_in_channels = reg_in_channels
+        self.roi_feat_size = roi_feat_size
+        self.reg_feat_up_ratio = int(reg_feat_up_ratio)
+        self.num_buckets = bbox_coder['num_buckets']
+        assert self.reg_feat_up_ratio // 2 >= 1
+        self.up_reg_feat_size = roi_feat_size * self.reg_feat_up_ratio
+        assert self.up_reg_feat_size == bbox_coder['num_buckets']
+        self.reg_pre_kernel = reg_pre_kernel
+        self.reg_post_kernel = reg_post_kernel
+        self.reg_pre_num = reg_pre_num
+        self.reg_post_num = reg_post_num
+        self.num_classes = num_classes
+        self.cls_out_channels = cls_out_channels
+        self.reg_offset_out_channels = reg_offset_out_channels
+        self.reg_cls_out_channels = reg_cls_out_channels
+        self.num_cls_fcs = num_cls_fcs
+        self.num_reg_fcs = num_reg_fcs
+        self.reg_class_agnostic = reg_class_agnostic
+        assert self.reg_class_agnostic
+        self.norm_cfg = norm_cfg
+
+        self.bbox_coder = build_bbox_coder(bbox_coder)
+        self.loss_cls = build_loss(loss_cls)
+        self.loss_bbox_cls = build_loss(loss_bbox_cls)
+        self.loss_bbox_reg = build_loss(loss_bbox_reg)
+
+        self.cls_fcs = self._add_fc_branch(self.num_cls_fcs,
+                                           self.cls_in_channels,
+                                           self.roi_feat_size,
+                                           self.cls_out_channels)
+
+        self.side_num = int(np.ceil(self.num_buckets / 2))
+
+        if self.reg_feat_up_ratio > 1:
+            self.upsample_x = nn.ConvTranspose1d(
+                reg_in_channels,
+                reg_in_channels,
+                self.reg_feat_up_ratio,
+                stride=self.reg_feat_up_ratio)
+            self.upsample_y = nn.ConvTranspose1d(
+                reg_in_channels,
+                reg_in_channels,
+                self.reg_feat_up_ratio,
+                stride=self.reg_feat_up_ratio)
+
+        self.reg_pre_convs = nn.ModuleList()
+        for i in range(self.reg_pre_num):
+            reg_pre_conv = ConvModule(
+                reg_in_channels,
+                reg_in_channels,
+                kernel_size=reg_pre_kernel,
+                padding=reg_pre_kernel // 2,
+                norm_cfg=norm_cfg,
+                act_cfg=dict(type='ReLU'))
+            self.reg_pre_convs.append(reg_pre_conv)
+
+        self.reg_post_conv_xs = nn.ModuleList()
+        for i in range(self.reg_post_num):
+            reg_post_conv_x = ConvModule(
+                reg_in_channels,
+                reg_in_channels,
+                kernel_size=(1, reg_post_kernel),
+                padding=(0, reg_post_kernel // 2),
+                norm_cfg=norm_cfg,
+                act_cfg=dict(type='ReLU'))
+            self.reg_post_conv_xs.append(reg_post_conv_x)
+        self.reg_post_conv_ys = nn.ModuleList()
+        for i in range(self.reg_post_num):
+            reg_post_conv_y = ConvModule(
+                reg_in_channels,
+                reg_in_channels,
+                kernel_size=(reg_post_kernel, 1),
+                padding=(reg_post_kernel // 2, 0),
+                norm_cfg=norm_cfg,
+                act_cfg=dict(type='ReLU'))
+            self.reg_post_conv_ys.append(reg_post_conv_y)
+
+        self.reg_conv_att_x = nn.Conv2d(reg_in_channels, 1, 1)
+        self.reg_conv_att_y = nn.Conv2d(reg_in_channels, 1, 1)
+
+        self.fc_cls = nn.Linear(self.cls_out_channels, self.num_classes + 1)
+        self.relu = nn.ReLU(inplace=True)
+
+        self.reg_cls_fcs = self._add_fc_branch(self.num_reg_fcs,
+                                               self.reg_in_channels, 1,
+                                               self.reg_cls_out_channels)
+        self.reg_offset_fcs = self._add_fc_branch(self.num_reg_fcs,
+                                                  self.reg_in_channels, 1,
+                                                  self.reg_offset_out_channels)
+        self.fc_reg_cls = nn.Linear(self.reg_cls_out_channels, 1)
+        self.fc_reg_offset = nn.Linear(self.reg_offset_out_channels, 1)
+
+    def _add_fc_branch(self, num_branch_fcs, in_channels, roi_feat_size,
+                       fc_out_channels):
+        in_channels = in_channels * roi_feat_size * roi_feat_size
+        branch_fcs = nn.ModuleList()
+        for i in range(num_branch_fcs):
+            fc_in_channels = (in_channels if i == 0 else fc_out_channels)
+            branch_fcs.append(nn.Linear(fc_in_channels, fc_out_channels))
+        return branch_fcs
+
+    def init_weights(self):
+        for module_list in [
+                self.reg_cls_fcs, self.reg_offset_fcs, self.cls_fcs
+        ]:
+            for m in module_list.modules():
+                if isinstance(m, nn.Linear):
+                    xavier_init(m, distribution='uniform')
+        if self.reg_feat_up_ratio > 1:
+            kaiming_init(self.upsample_x, distribution='normal')
+            kaiming_init(self.upsample_y, distribution='normal')
+
+        normal_init(self.reg_conv_att_x, 0, 0.01)
+        normal_init(self.reg_conv_att_y, 0, 0.01)
+        normal_init(self.fc_reg_offset, 0, 0.001)
+        normal_init(self.fc_reg_cls, 0, 0.01)
+        normal_init(self.fc_cls, 0, 0.01)
+
+    def cls_forward(self, cls_x):
+        cls_x = cls_x.view(cls_x.size(0), -1)
+        for fc in self.cls_fcs:
+            cls_x = self.relu(fc(cls_x))
+        cls_score = self.fc_cls(cls_x)
+        return cls_score
+
+    def attention_pool(self, reg_x):
+        """Extract direction-specific features fx and fy with attention
+        methanism."""
+        reg_fx = reg_x
+        reg_fy = reg_x
+        reg_fx_att = self.reg_conv_att_x(reg_fx).sigmoid()
+        reg_fy_att = self.reg_conv_att_y(reg_fy).sigmoid()
+        reg_fx_att = reg_fx_att / reg_fx_att.sum(dim=2).unsqueeze(2)
+        reg_fy_att = reg_fy_att / reg_fy_att.sum(dim=3).unsqueeze(3)
+        reg_fx = (reg_fx * reg_fx_att).sum(dim=2)
+        reg_fy = (reg_fy * reg_fy_att).sum(dim=3)
+        return reg_fx, reg_fy
+
+    def side_aware_feature_extractor(self, reg_x):
+        """Refine and extract side-aware features without split them."""
+        for reg_pre_conv in self.reg_pre_convs:
+            reg_x = reg_pre_conv(reg_x)
+        reg_fx, reg_fy = self.attention_pool(reg_x)
+
+        if self.reg_post_num > 0:
+            reg_fx = reg_fx.unsqueeze(2)
+            reg_fy = reg_fy.unsqueeze(3)
+            for i in range(self.reg_post_num):
+                reg_fx = self.reg_post_conv_xs[i](reg_fx)
+                reg_fy = self.reg_post_conv_ys[i](reg_fy)
+            reg_fx = reg_fx.squeeze(2)
+            reg_fy = reg_fy.squeeze(3)
+        if self.reg_feat_up_ratio > 1:
+            reg_fx = self.relu(self.upsample_x(reg_fx))
+            reg_fy = self.relu(self.upsample_y(reg_fy))
+        reg_fx = torch.transpose(reg_fx, 1, 2)
+        reg_fy = torch.transpose(reg_fy, 1, 2)
+        return reg_fx.contiguous(), reg_fy.contiguous()
+
+    def reg_pred(self, x, offset_fcs, cls_fcs):
+        """Predict bucketing estimation (cls_pred) and fine regression (offset
+        pred) with side-aware features."""
+        x_offset = x.view(-1, self.reg_in_channels)
+        x_cls = x.view(-1, self.reg_in_channels)
+
+        for fc in offset_fcs:
+            x_offset = self.relu(fc(x_offset))
+        for fc in cls_fcs:
+            x_cls = self.relu(fc(x_cls))
+        offset_pred = self.fc_reg_offset(x_offset)
+        cls_pred = self.fc_reg_cls(x_cls)
+
+        offset_pred = offset_pred.view(x.size(0), -1)
+        cls_pred = cls_pred.view(x.size(0), -1)
+
+        return offset_pred, cls_pred
+
+    def side_aware_split(self, feat):
+        """Split side-aware features aligned with orders of bucketing
+        targets."""
+        l_end = int(np.ceil(self.up_reg_feat_size / 2))
+        r_start = int(np.floor(self.up_reg_feat_size / 2))
+        feat_fl = feat[:, :l_end]
+        feat_fr = feat[:, r_start:].flip(dims=(1, ))
+        feat_fl = feat_fl.contiguous()
+        feat_fr = feat_fr.contiguous()
+        feat = torch.cat([feat_fl, feat_fr], dim=-1)
+        return feat
+
+    def bbox_pred_split(self, bbox_pred, num_proposals_per_img):
+        """Split batch bbox prediction back to each image."""
+        bucket_cls_preds, bucket_offset_preds = bbox_pred
+        bucket_cls_preds = bucket_cls_preds.split(num_proposals_per_img, 0)
+        bucket_offset_preds = bucket_offset_preds.split(
+            num_proposals_per_img, 0)
+        bbox_pred = tuple(zip(bucket_cls_preds, bucket_offset_preds))
+        return bbox_pred
+
+    def reg_forward(self, reg_x):
+        outs = self.side_aware_feature_extractor(reg_x)
+        edge_offset_preds = []
+        edge_cls_preds = []
+        reg_fx = outs[0]
+        reg_fy = outs[1]
+        offset_pred_x, cls_pred_x = self.reg_pred(reg_fx, self.reg_offset_fcs,
+                                                  self.reg_cls_fcs)
+        offset_pred_y, cls_pred_y = self.reg_pred(reg_fy, self.reg_offset_fcs,
+                                                  self.reg_cls_fcs)
+        offset_pred_x = self.side_aware_split(offset_pred_x)
+        offset_pred_y = self.side_aware_split(offset_pred_y)
+        cls_pred_x = self.side_aware_split(cls_pred_x)
+        cls_pred_y = self.side_aware_split(cls_pred_y)
+        edge_offset_preds = torch.cat([offset_pred_x, offset_pred_y], dim=-1)
+        edge_cls_preds = torch.cat([cls_pred_x, cls_pred_y], dim=-1)
+
+        return (edge_cls_preds, edge_offset_preds)
+
+    def forward(self, x):
+
+        bbox_pred = self.reg_forward(x)
+        cls_score = self.cls_forward(x)
+
+        return cls_score, bbox_pred
+
+    def get_targets(self, sampling_results, gt_bboxes, gt_labels,
+                    rcnn_train_cfg):
+        pos_proposals = [res.pos_bboxes for res in sampling_results]
+        neg_proposals = [res.neg_bboxes for res in sampling_results]
+        pos_gt_bboxes = [res.pos_gt_bboxes for res in sampling_results]
+        pos_gt_labels = [res.pos_gt_labels for res in sampling_results]
+        cls_reg_targets = self.bucket_target(pos_proposals, neg_proposals,
+                                             pos_gt_bboxes, pos_gt_labels,
+                                             rcnn_train_cfg)
+        (labels, label_weights, bucket_cls_targets, bucket_cls_weights,
+         bucket_offset_targets, bucket_offset_weights) = cls_reg_targets
+        return (labels, label_weights, (bucket_cls_targets,
+                                        bucket_offset_targets),
+                (bucket_cls_weights, bucket_offset_weights))
+
+    def bucket_target(self,
+                      pos_proposals_list,
+                      neg_proposals_list,
+                      pos_gt_bboxes_list,
+                      pos_gt_labels_list,
+                      rcnn_train_cfg,
+                      concat=True):
+        (labels, label_weights, bucket_cls_targets, bucket_cls_weights,
+         bucket_offset_targets, bucket_offset_weights) = multi_apply(
+             self._bucket_target_single,
+             pos_proposals_list,
+             neg_proposals_list,
+             pos_gt_bboxes_list,
+             pos_gt_labels_list,
+             cfg=rcnn_train_cfg)
+
+        if concat:
+            labels = torch.cat(labels, 0)
+            label_weights = torch.cat(label_weights, 0)
+            bucket_cls_targets = torch.cat(bucket_cls_targets, 0)
+            bucket_cls_weights = torch.cat(bucket_cls_weights, 0)
+            bucket_offset_targets = torch.cat(bucket_offset_targets, 0)
+            bucket_offset_weights = torch.cat(bucket_offset_weights, 0)
+        return (labels, label_weights, bucket_cls_targets, bucket_cls_weights,
+                bucket_offset_targets, bucket_offset_weights)
+
+    def _bucket_target_single(self, pos_proposals, neg_proposals,
+                              pos_gt_bboxes, pos_gt_labels, cfg):
+        """Compute bucketing estimation targets and fine regression targets for
+        a single image.
+
+        Args:
+            pos_proposals (Tensor): positive proposals of a single image,
+                 Shape (n_pos, 4)
+            neg_proposals (Tensor): negative proposals of a single image,
+                 Shape (n_neg, 4).
+            pos_gt_bboxes (Tensor): gt bboxes assigned to positive proposals
+                 of a single image, Shape (n_pos, 4).
+            pos_gt_labels (Tensor): gt labels assigned to positive proposals
+                 of a single image, Shape (n_pos, ).
+            cfg (dict): Config of calculating targets
+
+        Returns:
+            tuple:
+
+                - labels (Tensor): Labels in a single image. \
+                    Shape (n,).
+                - label_weights (Tensor): Label weights in a single image.\
+                    Shape (n,)
+                - bucket_cls_targets (Tensor): Bucket cls targets in \
+                    a single image. Shape (n, num_buckets*2).
+                - bucket_cls_weights (Tensor): Bucket cls weights in \
+                    a single image. Shape (n, num_buckets*2).
+                - bucket_offset_targets (Tensor): Bucket offset targets \
+                    in a single image. Shape (n, num_buckets*2).
+                - bucket_offset_targets (Tensor): Bucket offset weights \
+                    in a single image. Shape (n, num_buckets*2).
+        """
+        num_pos = pos_proposals.size(0)
+        num_neg = neg_proposals.size(0)
+        num_samples = num_pos + num_neg
+        labels = pos_gt_bboxes.new_full((num_samples, ),
+                                        self.num_classes,
+                                        dtype=torch.long)
+        label_weights = pos_proposals.new_zeros(num_samples)
+        bucket_cls_targets = pos_proposals.new_zeros(num_samples,
+                                                     4 * self.side_num)
+        bucket_cls_weights = pos_proposals.new_zeros(num_samples,
+                                                     4 * self.side_num)
+        bucket_offset_targets = pos_proposals.new_zeros(
+            num_samples, 4 * self.side_num)
+        bucket_offset_weights = pos_proposals.new_zeros(
+            num_samples, 4 * self.side_num)
+        if num_pos > 0:
+            labels[:num_pos] = pos_gt_labels
+            label_weights[:num_pos] = 1.0
+            (pos_bucket_offset_targets, pos_bucket_offset_weights,
+             pos_bucket_cls_targets,
+             pos_bucket_cls_weights) = self.bbox_coder.encode(
+                 pos_proposals, pos_gt_bboxes)
+            bucket_cls_targets[:num_pos, :] = pos_bucket_cls_targets
+            bucket_cls_weights[:num_pos, :] = pos_bucket_cls_weights
+            bucket_offset_targets[:num_pos, :] = pos_bucket_offset_targets
+            bucket_offset_weights[:num_pos, :] = pos_bucket_offset_weights
+        if num_neg > 0:
+            label_weights[-num_neg:] = 1.0
+        return (labels, label_weights, bucket_cls_targets, bucket_cls_weights,
+                bucket_offset_targets, bucket_offset_weights)
+
+    def loss(self,
+             cls_score,
+             bbox_pred,
+             rois,
+             labels,
+             label_weights,
+             bbox_targets,
+             bbox_weights,
+             reduction_override=None):
+        losses = dict()
+        if cls_score is not None:
+            avg_factor = max(torch.sum(label_weights > 0).float().item(), 1.)
+            losses['loss_cls'] = self.loss_cls(
+                cls_score,
+                labels,
+                label_weights,
+                avg_factor=avg_factor,
+                reduction_override=reduction_override)
+            losses['acc'] = accuracy(cls_score, labels)
+
+        if bbox_pred is not None:
+            bucket_cls_preds, bucket_offset_preds = bbox_pred
+            bucket_cls_targets, bucket_offset_targets = bbox_targets
+            bucket_cls_weights, bucket_offset_weights = bbox_weights
+            # edge cls
+            bucket_cls_preds = bucket_cls_preds.view(-1, self.side_num)
+            bucket_cls_targets = bucket_cls_targets.view(-1, self.side_num)
+            bucket_cls_weights = bucket_cls_weights.view(-1, self.side_num)
+            losses['loss_bbox_cls'] = self.loss_bbox_cls(
+                bucket_cls_preds,
+                bucket_cls_targets,
+                bucket_cls_weights,
+                avg_factor=bucket_cls_targets.size(0),
+                reduction_override=reduction_override)
+
+            losses['loss_bbox_reg'] = self.loss_bbox_reg(
+                bucket_offset_preds,
+                bucket_offset_targets,
+                bucket_offset_weights,
+                avg_factor=bucket_offset_targets.size(0),
+                reduction_override=reduction_override)
+
+        return losses
+
+    @force_fp32(apply_to=('cls_score', 'bbox_pred'))
+    def get_bboxes(self,
+                   rois,
+                   cls_score,
+                   bbox_pred,
+                   img_shape,
+                   scale_factor,
+                   rescale=False,
+                   cfg=None):
+        if isinstance(cls_score, list):
+            cls_score = sum(cls_score) / float(len(cls_score))
+        scores = F.softmax(cls_score, dim=1) if cls_score is not None else None
+
+        if bbox_pred is not None:
+            bboxes, confids = self.bbox_coder.decode(rois[:, 1:], bbox_pred,
+                                                     img_shape)
+        else:
+            bboxes = rois[:, 1:].clone()
+            confids = None
+            if img_shape is not None:
+                bboxes[:, [0, 2]].clamp_(min=0, max=img_shape[1] - 1)
+                bboxes[:, [1, 3]].clamp_(min=0, max=img_shape[0] - 1)
+
+        if rescale and bboxes.size(0) > 0:
+            if isinstance(scale_factor, float):
+                bboxes /= scale_factor
+            else:
+                bboxes /= torch.from_numpy(scale_factor).to(bboxes.device)
+
+        if cfg is None:
+            return bboxes, scores
+        else:
+            det_bboxes, det_labels = multiclass_nms(
+                bboxes,
+                scores,
+                cfg.score_thr,
+                cfg.nms,
+                cfg.max_per_img,
+                score_factors=confids)
+
+            return det_bboxes, det_labels
+
+    @force_fp32(apply_to=('bbox_preds', ))
+    def refine_bboxes(self, rois, labels, bbox_preds, pos_is_gts, img_metas):
+        """Refine bboxes during training.
+
+        Args:
+            rois (Tensor): Shape (n*bs, 5), where n is image number per GPU,
+                and bs is the sampled RoIs per image.
+            labels (Tensor): Shape (n*bs, ).
+            bbox_preds (list[Tensor]): Shape [(n*bs, num_buckets*2), \
+                (n*bs, num_buckets*2)].
+            pos_is_gts (list[Tensor]): Flags indicating if each positive bbox
+                is a gt bbox.
+            img_metas (list[dict]): Meta info of each image.
+
+        Returns:
+            list[Tensor]: Refined bboxes of each image in a mini-batch.
+        """
+        img_ids = rois[:, 0].long().unique(sorted=True)
+        assert img_ids.numel() == len(img_metas)
+
+        bboxes_list = []
+        for i in range(len(img_metas)):
+            inds = torch.nonzero(
+                rois[:, 0] == i, as_tuple=False).squeeze(dim=1)
+            num_rois = inds.numel()
+
+            bboxes_ = rois[inds, 1:]
+            label_ = labels[inds]
+            edge_cls_preds, edge_offset_preds = bbox_preds
+            edge_cls_preds_ = edge_cls_preds[inds]
+            edge_offset_preds_ = edge_offset_preds[inds]
+            bbox_pred_ = [edge_cls_preds_, edge_offset_preds_]
+            img_meta_ = img_metas[i]
+            pos_is_gts_ = pos_is_gts[i]
+
+            bboxes = self.regress_by_class(bboxes_, label_, bbox_pred_,
+                                           img_meta_)
+            # filter gt bboxes
+            pos_keep = 1 - pos_is_gts_
+            keep_inds = pos_is_gts_.new_ones(num_rois)
+            keep_inds[:len(pos_is_gts_)] = pos_keep
+
+            bboxes_list.append(bboxes[keep_inds.type(torch.bool)])
+
+        return bboxes_list
+
+    @force_fp32(apply_to=('bbox_pred', ))
+    def regress_by_class(self, rois, label, bbox_pred, img_meta):
+        """Regress the bbox for the predicted class. Used in Cascade R-CNN.
+
+        Args:
+            rois (Tensor): shape (n, 4) or (n, 5)
+            label (Tensor): shape (n, )
+            bbox_pred (list[Tensor]): shape [(n, num_buckets *2), \
+                (n, num_buckets *2)]
+            img_meta (dict): Image meta info.
+
+        Returns:
+            Tensor: Regressed bboxes, the same shape as input rois.
+        """
+        assert rois.size(1) == 4 or rois.size(1) == 5
+
+        if rois.size(1) == 4:
+            new_rois, _ = self.bbox_coder.decode(rois, bbox_pred,
+                                                 img_meta['img_shape'])
+        else:
+            bboxes, _ = self.bbox_coder.decode(rois[:, 1:], bbox_pred,
+                                               img_meta['img_shape'])
+            new_rois = torch.cat((rois[:, [0]], bboxes), dim=1)
+
+        return new_rois
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/scnet_bbox_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/scnet_bbox_head.py
new file mode 100644
index 0000000..35758f4
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/bbox_heads/scnet_bbox_head.py
@@ -0,0 +1,76 @@
+from mmdet.models.builder import HEADS
+from .convfc_bbox_head import ConvFCBBoxHead
+
+
+@HEADS.register_module()
+class SCNetBBoxHead(ConvFCBBoxHead):
+    """BBox head for `SCNet <https://arxiv.org/abs/2012.10150>`_.
+
+    This inherits ``ConvFCBBoxHead`` with modified forward() function, allow us
+    to get intermediate shared feature.
+    """
+
+    def _forward_shared(self, x):
+        """Forward function for shared part."""
+        if self.num_shared_convs > 0:
+            for conv in self.shared_convs:
+                x = conv(x)
+
+        if self.num_shared_fcs > 0:
+            if self.with_avg_pool:
+                x = self.avg_pool(x)
+
+            x = x.flatten(1)
+
+            for fc in self.shared_fcs:
+                x = self.relu(fc(x))
+
+        return x
+
+    def _forward_cls_reg(self, x):
+        """Forward function for classification and regression parts."""
+        x_cls = x
+        x_reg = x
+
+        for conv in self.cls_convs:
+            x_cls = conv(x_cls)
+        if x_cls.dim() > 2:
+            if self.with_avg_pool:
+                x_cls = self.avg_pool(x_cls)
+            x_cls = x_cls.flatten(1)
+        for fc in self.cls_fcs:
+            x_cls = self.relu(fc(x_cls))
+
+        for conv in self.reg_convs:
+            x_reg = conv(x_reg)
+        if x_reg.dim() > 2:
+            if self.with_avg_pool:
+                x_reg = self.avg_pool(x_reg)
+            x_reg = x_reg.flatten(1)
+        for fc in self.reg_fcs:
+            x_reg = self.relu(fc(x_reg))
+
+        cls_score = self.fc_cls(x_cls) if self.with_cls else None
+        bbox_pred = self.fc_reg(x_reg) if self.with_reg else None
+
+        return cls_score, bbox_pred
+
+    def forward(self, x, return_shared_feat=False):
+        """Forward function.
+
+        Args:
+            x (Tensor): input features
+            return_shared_feat (bool): If True, return cls-reg-shared feature.
+
+        Return:
+            out (tuple[Tensor]): contain ``cls_score`` and ``bbox_pred``,
+                if  ``return_shared_feat`` is True, append ``x_shared`` to the
+                returned tuple.
+        """
+        x_shared = self._forward_shared(x)
+        out = self._forward_cls_reg(x_shared)
+
+        if return_shared_feat:
+            out += (x_shared, )
+
+        return out
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/cascade_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/cascade_roi_head.py
new file mode 100644
index 0000000..45b6f36
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/cascade_roi_head.py
@@ -0,0 +1,507 @@
+import torch
+import torch.nn as nn
+
+from mmdet.core import (bbox2result, bbox2roi, bbox_mapping, build_assigner,
+                        build_sampler, merge_aug_bboxes, merge_aug_masks,
+                        multiclass_nms)
+from ..builder import HEADS, build_head, build_roi_extractor
+from .base_roi_head import BaseRoIHead
+from .test_mixins import BBoxTestMixin, MaskTestMixin
+
+
+@HEADS.register_module()
+class CascadeRoIHead(BaseRoIHead, BBoxTestMixin, MaskTestMixin):
+    """Cascade roi head including one bbox head and one mask head.
+
+    https://arxiv.org/abs/1712.00726
+    """
+
+    def __init__(self,
+                 num_stages,
+                 stage_loss_weights,
+                 bbox_roi_extractor=None,
+                 bbox_head=None,
+                 mask_roi_extractor=None,
+                 mask_head=None,
+                 shared_head=None,
+                 train_cfg=None,
+                 test_cfg=None):
+        assert bbox_roi_extractor is not None
+        assert bbox_head is not None
+        assert shared_head is None, \
+            'Shared head is not supported in Cascade RCNN anymore'
+        self.num_stages = num_stages
+        self.stage_loss_weights = stage_loss_weights
+        super(CascadeRoIHead, self).__init__(
+            bbox_roi_extractor=bbox_roi_extractor,
+            bbox_head=bbox_head,
+            mask_roi_extractor=mask_roi_extractor,
+            mask_head=mask_head,
+            shared_head=shared_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg)
+
+    def init_bbox_head(self, bbox_roi_extractor, bbox_head):
+        """Initialize box head and box roi extractor.
+
+        Args:
+            bbox_roi_extractor (dict): Config of box roi extractor.
+            bbox_head (dict): Config of box in box head.
+        """
+        self.bbox_roi_extractor = nn.ModuleList()
+        self.bbox_head = nn.ModuleList()
+        if not isinstance(bbox_roi_extractor, list):
+            bbox_roi_extractor = [
+                bbox_roi_extractor for _ in range(self.num_stages)
+            ]
+        if not isinstance(bbox_head, list):
+            bbox_head = [bbox_head for _ in range(self.num_stages)]
+        assert len(bbox_roi_extractor) == len(bbox_head) == self.num_stages
+        for roi_extractor, head in zip(bbox_roi_extractor, bbox_head):
+            self.bbox_roi_extractor.append(build_roi_extractor(roi_extractor))
+            self.bbox_head.append(build_head(head))
+
+    def init_mask_head(self, mask_roi_extractor, mask_head):
+        """Initialize mask head and mask roi extractor.
+
+        Args:
+            mask_roi_extractor (dict): Config of mask roi extractor.
+            mask_head (dict): Config of mask in mask head.
+        """
+        self.mask_head = nn.ModuleList()
+        if not isinstance(mask_head, list):
+            mask_head = [mask_head for _ in range(self.num_stages)]
+        assert len(mask_head) == self.num_stages
+        for head in mask_head:
+            self.mask_head.append(build_head(head))
+        if mask_roi_extractor is not None:
+            self.share_roi_extractor = False
+            self.mask_roi_extractor = nn.ModuleList()
+            if not isinstance(mask_roi_extractor, list):
+                mask_roi_extractor = [
+                    mask_roi_extractor for _ in range(self.num_stages)
+                ]
+            assert len(mask_roi_extractor) == self.num_stages
+            for roi_extractor in mask_roi_extractor:
+                self.mask_roi_extractor.append(
+                    build_roi_extractor(roi_extractor))
+        else:
+            self.share_roi_extractor = True
+            self.mask_roi_extractor = self.bbox_roi_extractor
+
+    def init_assigner_sampler(self):
+        """Initialize assigner and sampler for each stage."""
+        self.bbox_assigner = []
+        self.bbox_sampler = []
+        if self.train_cfg is not None:
+            for idx, rcnn_train_cfg in enumerate(self.train_cfg):
+                self.bbox_assigner.append(
+                    build_assigner(rcnn_train_cfg.assigner))
+                self.current_stage = idx
+                self.bbox_sampler.append(
+                    build_sampler(rcnn_train_cfg.sampler, context=self))
+
+    def init_weights(self, pretrained):
+        """Initialize the weights in head.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        if self.with_shared_head:
+            self.shared_head.init_weights(pretrained=pretrained)
+        for i in range(self.num_stages):
+            if self.with_bbox:
+                self.bbox_roi_extractor[i].init_weights()
+                self.bbox_head[i].init_weights()
+            if self.with_mask:
+                if not self.share_roi_extractor:
+                    self.mask_roi_extractor[i].init_weights()
+                self.mask_head[i].init_weights()
+
+    def forward_dummy(self, x, proposals):
+        """Dummy forward function."""
+        # bbox head
+        outs = ()
+        rois = bbox2roi([proposals])
+        if self.with_bbox:
+            for i in range(self.num_stages):
+                bbox_results = self._bbox_forward(i, x, rois)
+                outs = outs + (bbox_results['cls_score'],
+                               bbox_results['bbox_pred'])
+        # mask heads
+        if self.with_mask:
+            mask_rois = rois[:100]
+            for i in range(self.num_stages):
+                mask_results = self._mask_forward(i, x, mask_rois)
+                outs = outs + (mask_results['mask_pred'], )
+        return outs
+
+    def _bbox_forward(self, stage, x, rois):
+        """Box head forward function used in both training and testing."""
+        bbox_roi_extractor = self.bbox_roi_extractor[stage]
+        bbox_head = self.bbox_head[stage]
+        bbox_feats = bbox_roi_extractor(x[:bbox_roi_extractor.num_inputs],
+                                        rois)
+        # do not support caffe_c4 model anymore
+        cls_score, bbox_pred = bbox_head(bbox_feats)
+
+        bbox_results = dict(
+            cls_score=cls_score, bbox_pred=bbox_pred, bbox_feats=bbox_feats)
+        return bbox_results
+
+    def _bbox_forward_train(self, stage, x, sampling_results, gt_bboxes,
+                            gt_labels, rcnn_train_cfg):
+        """Run forward function and calculate loss for box head in training."""
+        rois = bbox2roi([res.bboxes for res in sampling_results])
+        bbox_results = self._bbox_forward(stage, x, rois)
+        bbox_targets = self.bbox_head[stage].get_targets(
+            sampling_results, gt_bboxes, gt_labels, rcnn_train_cfg)
+        loss_bbox = self.bbox_head[stage].loss(bbox_results['cls_score'],
+                                               bbox_results['bbox_pred'], rois,
+                                               *bbox_targets)
+
+        bbox_results.update(
+            loss_bbox=loss_bbox, rois=rois, bbox_targets=bbox_targets)
+        return bbox_results
+
+    def _mask_forward(self, stage, x, rois):
+        """Mask head forward function used in both training and testing."""
+        mask_roi_extractor = self.mask_roi_extractor[stage]
+        mask_head = self.mask_head[stage]
+        mask_feats = mask_roi_extractor(x[:mask_roi_extractor.num_inputs],
+                                        rois)
+        # do not support caffe_c4 model anymore
+        mask_pred = mask_head(mask_feats)
+
+        mask_results = dict(mask_pred=mask_pred)
+        return mask_results
+
+    def _mask_forward_train(self,
+                            stage,
+                            x,
+                            sampling_results,
+                            gt_masks,
+                            rcnn_train_cfg,
+                            bbox_feats=None):
+        """Run forward function and calculate loss for mask head in
+        training."""
+        pos_rois = bbox2roi([res.pos_bboxes for res in sampling_results])
+        mask_results = self._mask_forward(stage, x, pos_rois)
+
+        mask_targets = self.mask_head[stage].get_targets(
+            sampling_results, gt_masks, rcnn_train_cfg)
+        pos_labels = torch.cat([res.pos_gt_labels for res in sampling_results])
+        loss_mask = self.mask_head[stage].loss(mask_results['mask_pred'],
+                                               mask_targets, pos_labels)
+
+        mask_results.update(loss_mask=loss_mask)
+        return mask_results
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      proposal_list,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None):
+        """
+        Args:
+            x (list[Tensor]): list of multi-level img features.
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+            proposals (list[Tensors]): list of region proposals.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+            gt_masks (None | Tensor) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        losses = dict()
+        for i in range(self.num_stages):
+            self.current_stage = i
+            rcnn_train_cfg = self.train_cfg[i]
+            lw = self.stage_loss_weights[i]
+
+            # assign gts and sample proposals
+            sampling_results = []
+            if self.with_bbox or self.with_mask:
+                bbox_assigner = self.bbox_assigner[i]
+                bbox_sampler = self.bbox_sampler[i]
+                num_imgs = len(img_metas)
+                if gt_bboxes_ignore is None:
+                    gt_bboxes_ignore = [None for _ in range(num_imgs)]
+
+                for j in range(num_imgs):
+                    assign_result = bbox_assigner.assign(
+                        proposal_list[j], gt_bboxes[j], gt_bboxes_ignore[j],
+                        gt_labels[j])
+                    sampling_result = bbox_sampler.sample(
+                        assign_result,
+                        proposal_list[j],
+                        gt_bboxes[j],
+                        gt_labels[j],
+                        feats=[lvl_feat[j][None] for lvl_feat in x])
+                    sampling_results.append(sampling_result)
+
+            # bbox head forward and loss
+            bbox_results = self._bbox_forward_train(i, x, sampling_results,
+                                                    gt_bboxes, gt_labels,
+                                                    rcnn_train_cfg)
+
+            for name, value in bbox_results['loss_bbox'].items():
+                losses[f's{i}.{name}'] = (
+                    value * lw if 'loss' in name else value)
+
+            # mask head forward and loss
+            if self.with_mask:
+                mask_results = self._mask_forward_train(
+                    i, x, sampling_results, gt_masks, rcnn_train_cfg,
+                    bbox_results['bbox_feats'])
+                for name, value in mask_results['loss_mask'].items():
+                    losses[f's{i}.{name}'] = (
+                        value * lw if 'loss' in name else value)
+
+            # refine bboxes
+            if i < self.num_stages - 1:
+                pos_is_gts = [res.pos_is_gt for res in sampling_results]
+                # bbox_targets is a tuple
+                roi_labels = bbox_results['bbox_targets'][0]
+                with torch.no_grad():
+                    roi_labels = torch.where(
+                        roi_labels == self.bbox_head[i].num_classes,
+                        bbox_results['cls_score'][:, :-1].argmax(1),
+                        roi_labels)
+                    proposal_list = self.bbox_head[i].refine_bboxes(
+                        bbox_results['rois'], roi_labels,
+                        bbox_results['bbox_pred'], pos_is_gts, img_metas)
+
+        return losses
+
+    def simple_test(self, x, proposal_list, img_metas, rescale=False):
+        """Test without augmentation."""
+        assert self.with_bbox, 'Bbox head must be implemented.'
+        num_imgs = len(proposal_list)
+        img_shapes = tuple(meta['img_shape'] for meta in img_metas)
+        ori_shapes = tuple(meta['ori_shape'] for meta in img_metas)
+        scale_factors = tuple(meta['scale_factor'] for meta in img_metas)
+
+        # "ms" in variable names means multi-stage
+        ms_bbox_result = {}
+        ms_segm_result = {}
+        ms_scores = []
+        rcnn_test_cfg = self.test_cfg
+
+        rois = bbox2roi(proposal_list)
+        for i in range(self.num_stages):
+            bbox_results = self._bbox_forward(i, x, rois)
+
+            # split batch bbox prediction back to each image
+            cls_score = bbox_results['cls_score']
+            bbox_pred = bbox_results['bbox_pred']
+            num_proposals_per_img = tuple(
+                len(proposals) for proposals in proposal_list)
+            rois = rois.split(num_proposals_per_img, 0)
+            cls_score = cls_score.split(num_proposals_per_img, 0)
+            if isinstance(bbox_pred, torch.Tensor):
+                bbox_pred = bbox_pred.split(num_proposals_per_img, 0)
+            else:
+                bbox_pred = self.bbox_head[i].bbox_pred_split(
+                    bbox_pred, num_proposals_per_img)
+            ms_scores.append(cls_score)
+
+            if i < self.num_stages - 1:
+                bbox_label = [s[:, :-1].argmax(dim=1) for s in cls_score]
+                rois = torch.cat([
+                    self.bbox_head[i].regress_by_class(rois[j], bbox_label[j],
+                                                       bbox_pred[j],
+                                                       img_metas[j])
+                    for j in range(num_imgs)
+                ])
+
+        # average scores of each image by stages
+        cls_score = [
+            sum([score[i] for score in ms_scores]) / float(len(ms_scores))
+            for i in range(num_imgs)
+        ]
+
+        # apply bbox post-processing to each image individually
+        det_bboxes = []
+        det_labels = []
+        for i in range(num_imgs):
+            det_bbox, det_label = self.bbox_head[-1].get_bboxes(
+                rois[i],
+                cls_score[i],
+                bbox_pred[i],
+                img_shapes[i],
+                scale_factors[i],
+                rescale=rescale,
+                cfg=rcnn_test_cfg)
+            det_bboxes.append(det_bbox)
+            det_labels.append(det_label)
+
+        if torch.onnx.is_in_onnx_export():
+            return det_bboxes, det_labels
+        bbox_results = [
+            bbox2result(det_bboxes[i], det_labels[i],
+                        self.bbox_head[-1].num_classes)
+            for i in range(num_imgs)
+        ]
+        ms_bbox_result['ensemble'] = bbox_results
+
+        if self.with_mask:
+            if all(det_bbox.shape[0] == 0 for det_bbox in det_bboxes):
+                mask_classes = self.mask_head[-1].num_classes
+                segm_results = [[[] for _ in range(mask_classes)]
+                                for _ in range(num_imgs)]
+            else:
+                if rescale and not isinstance(scale_factors[0], float):
+                    scale_factors = [
+                        torch.from_numpy(scale_factor).to(det_bboxes[0].device)
+                        for scale_factor in scale_factors
+                    ]
+                _bboxes = [
+                    det_bboxes[i][:, :4] *
+                    scale_factors[i] if rescale else det_bboxes[i][:, :4]
+                    for i in range(len(det_bboxes))
+                ]
+                mask_rois = bbox2roi(_bboxes)
+                num_mask_rois_per_img = tuple(
+                    _bbox.size(0) for _bbox in _bboxes)
+                aug_masks = []
+                for i in range(self.num_stages):
+                    mask_results = self._mask_forward(i, x, mask_rois)
+                    mask_pred = mask_results['mask_pred']
+                    # split batch mask prediction back to each image
+                    mask_pred = mask_pred.split(num_mask_rois_per_img, 0)
+                    aug_masks.append(
+                        [m.sigmoid().cpu().numpy() for m in mask_pred])
+
+                # apply mask post-processing to each image individually
+                segm_results = []
+                for i in range(num_imgs):
+                    if det_bboxes[i].shape[0] == 0:
+                        segm_results.append(
+                            [[]
+                             for _ in range(self.mask_head[-1].num_classes)])
+                    else:
+                        aug_mask = [mask[i] for mask in aug_masks]
+                        merged_masks = merge_aug_masks(
+                            aug_mask, [[img_metas[i]]] * self.num_stages,
+                            rcnn_test_cfg)
+                        segm_result = self.mask_head[-1].get_seg_masks(
+                            merged_masks, _bboxes[i], det_labels[i],
+                            rcnn_test_cfg, ori_shapes[i], scale_factors[i],
+                            rescale)
+                        segm_results.append(segm_result)
+            ms_segm_result['ensemble'] = segm_results
+
+        if self.with_mask:
+            results = list(
+                zip(ms_bbox_result['ensemble'], ms_segm_result['ensemble']))
+        else:
+            results = ms_bbox_result['ensemble']
+
+        return results
+
+    def aug_test(self, features, proposal_list, img_metas, rescale=False):
+        """Test with augmentations.
+
+        If rescale is False, then returned bboxes and masks will fit the scale
+        of imgs[0].
+        """
+        rcnn_test_cfg = self.test_cfg
+        aug_bboxes = []
+        aug_scores = []
+        for x, img_meta in zip(features, img_metas):
+            # only one image in the batch
+            img_shape = img_meta[0]['img_shape']
+            scale_factor = img_meta[0]['scale_factor']
+            flip = img_meta[0]['flip']
+            flip_direction = img_meta[0]['flip_direction']
+
+            proposals = bbox_mapping(proposal_list[0][:, :4], img_shape,
+                                     scale_factor, flip, flip_direction)
+            # "ms" in variable names means multi-stage
+            ms_scores = []
+
+            rois = bbox2roi([proposals])
+            for i in range(self.num_stages):
+                bbox_results = self._bbox_forward(i, x, rois)
+                ms_scores.append(bbox_results['cls_score'])
+
+                if i < self.num_stages - 1:
+                    bbox_label = bbox_results['cls_score'][:, :-1].argmax(
+                        dim=1)
+                    rois = self.bbox_head[i].regress_by_class(
+                        rois, bbox_label, bbox_results['bbox_pred'],
+                        img_meta[0])
+
+            cls_score = sum(ms_scores) / float(len(ms_scores))
+            bboxes, scores = self.bbox_head[-1].get_bboxes(
+                rois,
+                cls_score,
+                bbox_results['bbox_pred'],
+                img_shape,
+                scale_factor,
+                rescale=False,
+                cfg=None)
+            aug_bboxes.append(bboxes)
+            aug_scores.append(scores)
+
+        # after merging, bboxes will be rescaled to the original image size
+        merged_bboxes, merged_scores = merge_aug_bboxes(
+            aug_bboxes, aug_scores, img_metas, rcnn_test_cfg)
+        det_bboxes, det_labels = multiclass_nms(merged_bboxes, merged_scores,
+                                                rcnn_test_cfg.score_thr,
+                                                rcnn_test_cfg.nms,
+                                                rcnn_test_cfg.max_per_img)
+
+        bbox_result = bbox2result(det_bboxes, det_labels,
+                                  self.bbox_head[-1].num_classes)
+
+        if self.with_mask:
+            if det_bboxes.shape[0] == 0:
+                segm_result = [[[]
+                                for _ in range(self.mask_head[-1].num_classes)]
+                               ]
+            else:
+                aug_masks = []
+                aug_img_metas = []
+                for x, img_meta in zip(features, img_metas):
+                    img_shape = img_meta[0]['img_shape']
+                    scale_factor = img_meta[0]['scale_factor']
+                    flip = img_meta[0]['flip']
+                    flip_direction = img_meta[0]['flip_direction']
+                    _bboxes = bbox_mapping(det_bboxes[:, :4], img_shape,
+                                           scale_factor, flip, flip_direction)
+                    mask_rois = bbox2roi([_bboxes])
+                    for i in range(self.num_stages):
+                        mask_results = self._mask_forward(i, x, mask_rois)
+                        aug_masks.append(
+                            mask_results['mask_pred'].sigmoid().cpu().numpy())
+                        aug_img_metas.append(img_meta)
+                merged_masks = merge_aug_masks(aug_masks, aug_img_metas,
+                                               self.test_cfg)
+
+                ori_shape = img_metas[0][0]['ori_shape']
+                segm_result = self.mask_head[-1].get_seg_masks(
+                    merged_masks,
+                    det_bboxes,
+                    det_labels,
+                    rcnn_test_cfg,
+                    ori_shape,
+                    scale_factor=1.0,
+                    rescale=False)
+            return [(bbox_result, segm_result)]
+        else:
+            return [bbox_result]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/double_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/double_roi_head.py
new file mode 100644
index 0000000..a1aa6c8
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/double_roi_head.py
@@ -0,0 +1,33 @@
+from ..builder import HEADS
+from .standard_roi_head import StandardRoIHead
+
+
+@HEADS.register_module()
+class DoubleHeadRoIHead(StandardRoIHead):
+    """RoI head for Double Head RCNN.
+
+    https://arxiv.org/abs/1904.06493
+    """
+
+    def __init__(self, reg_roi_scale_factor, **kwargs):
+        super(DoubleHeadRoIHead, self).__init__(**kwargs)
+        self.reg_roi_scale_factor = reg_roi_scale_factor
+
+    def _bbox_forward(self, x, rois):
+        """Box head forward function used in both training and testing time."""
+        bbox_cls_feats = self.bbox_roi_extractor(
+            x[:self.bbox_roi_extractor.num_inputs], rois)
+        bbox_reg_feats = self.bbox_roi_extractor(
+            x[:self.bbox_roi_extractor.num_inputs],
+            rois,
+            roi_scale_factor=self.reg_roi_scale_factor)
+        if self.with_shared_head:
+            bbox_cls_feats = self.shared_head(bbox_cls_feats)
+            bbox_reg_feats = self.shared_head(bbox_reg_feats)
+        cls_score, bbox_pred = self.bbox_head(bbox_cls_feats, bbox_reg_feats)
+
+        bbox_results = dict(
+            cls_score=cls_score,
+            bbox_pred=bbox_pred,
+            bbox_feats=bbox_cls_feats)
+        return bbox_results
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/dynamic_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/dynamic_roi_head.py
new file mode 100644
index 0000000..89427a9
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/dynamic_roi_head.py
@@ -0,0 +1,154 @@
+import numpy as np
+import torch
+
+from mmdet.core import bbox2roi
+from mmdet.models.losses import SmoothL1Loss
+from ..builder import HEADS
+from .standard_roi_head import StandardRoIHead
+
+EPS = 1e-15
+
+
+@HEADS.register_module()
+class DynamicRoIHead(StandardRoIHead):
+    """RoI head for `Dynamic R-CNN <https://arxiv.org/abs/2004.06002>`_."""
+
+    def __init__(self, **kwargs):
+        super(DynamicRoIHead, self).__init__(**kwargs)
+        assert isinstance(self.bbox_head.loss_bbox, SmoothL1Loss)
+        # the IoU history of the past `update_iter_interval` iterations
+        self.iou_history = []
+        # the beta history of the past `update_iter_interval` iterations
+        self.beta_history = []
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      proposal_list,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None):
+        """Forward function for training.
+
+        Args:
+            x (list[Tensor]): list of multi-level img features.
+
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+
+            proposals (list[Tensors]): list of region proposals.
+
+            gt_bboxes (list[Tensor]): each item are the truth boxes for each
+                image in [tl_x, tl_y, br_x, br_y] format.
+
+            gt_labels (list[Tensor]): class indices corresponding to each box
+
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+            gt_masks (None | Tensor) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        # assign gts and sample proposals
+        if self.with_bbox or self.with_mask:
+            num_imgs = len(img_metas)
+            if gt_bboxes_ignore is None:
+                gt_bboxes_ignore = [None for _ in range(num_imgs)]
+            sampling_results = []
+            cur_iou = []
+            for i in range(num_imgs):
+                assign_result = self.bbox_assigner.assign(
+                    proposal_list[i], gt_bboxes[i], gt_bboxes_ignore[i],
+                    gt_labels[i])
+                sampling_result = self.bbox_sampler.sample(
+                    assign_result,
+                    proposal_list[i],
+                    gt_bboxes[i],
+                    gt_labels[i],
+                    feats=[lvl_feat[i][None] for lvl_feat in x])
+                # record the `iou_topk`-th largest IoU in an image
+                iou_topk = min(self.train_cfg.dynamic_rcnn.iou_topk,
+                               len(assign_result.max_overlaps))
+                ious, _ = torch.topk(assign_result.max_overlaps, iou_topk)
+                cur_iou.append(ious[-1].item())
+                sampling_results.append(sampling_result)
+            # average the current IoUs over images
+            cur_iou = np.mean(cur_iou)
+            self.iou_history.append(cur_iou)
+
+        losses = dict()
+        # bbox head forward and loss
+        if self.with_bbox:
+            bbox_results = self._bbox_forward_train(x, sampling_results,
+                                                    gt_bboxes, gt_labels,
+                                                    img_metas)
+            losses.update(bbox_results['loss_bbox'])
+
+        # mask head forward and loss
+        if self.with_mask:
+            mask_results = self._mask_forward_train(x, sampling_results,
+                                                    bbox_results['bbox_feats'],
+                                                    gt_masks, img_metas)
+            losses.update(mask_results['loss_mask'])
+
+        # update IoU threshold and SmoothL1 beta
+        update_iter_interval = self.train_cfg.dynamic_rcnn.update_iter_interval
+        if len(self.iou_history) % update_iter_interval == 0:
+            new_iou_thr, new_beta = self.update_hyperparameters()
+
+        return losses
+
+    def _bbox_forward_train(self, x, sampling_results, gt_bboxes, gt_labels,
+                            img_metas):
+        num_imgs = len(img_metas)
+        rois = bbox2roi([res.bboxes for res in sampling_results])
+        bbox_results = self._bbox_forward(x, rois)
+
+        bbox_targets = self.bbox_head.get_targets(sampling_results, gt_bboxes,
+                                                  gt_labels, self.train_cfg)
+        # record the `beta_topk`-th smallest target
+        # `bbox_targets[2]` and `bbox_targets[3]` stand for bbox_targets
+        # and bbox_weights, respectively
+        pos_inds = bbox_targets[3][:, 0].nonzero().squeeze(1)
+        num_pos = len(pos_inds)
+        cur_target = bbox_targets[2][pos_inds, :2].abs().mean(dim=1)
+        beta_topk = min(self.train_cfg.dynamic_rcnn.beta_topk * num_imgs,
+                        num_pos)
+        cur_target = torch.kthvalue(cur_target, beta_topk)[0].item()
+        self.beta_history.append(cur_target)
+        loss_bbox = self.bbox_head.loss(bbox_results['cls_score'],
+                                        bbox_results['bbox_pred'], rois,
+                                        *bbox_targets)
+
+        bbox_results.update(loss_bbox=loss_bbox)
+        return bbox_results
+
+    def update_hyperparameters(self):
+        """Update hyperparameters like IoU thresholds for assigner and beta for
+        SmoothL1 loss based on the training statistics.
+
+        Returns:
+            tuple[float]: the updated ``iou_thr`` and ``beta``.
+        """
+        new_iou_thr = max(self.train_cfg.dynamic_rcnn.initial_iou,
+                          np.mean(self.iou_history))
+        self.iou_history = []
+        self.bbox_assigner.pos_iou_thr = new_iou_thr
+        self.bbox_assigner.neg_iou_thr = new_iou_thr
+        self.bbox_assigner.min_pos_iou = new_iou_thr
+        if (np.median(self.beta_history) < EPS):
+            # avoid 0 or too small value for new_beta
+            new_beta = self.bbox_head.loss_bbox.beta
+        else:
+            new_beta = min(self.train_cfg.dynamic_rcnn.initial_beta,
+                           np.median(self.beta_history))
+        self.beta_history = []
+        self.bbox_head.loss_bbox.beta = new_beta
+        return new_iou_thr, new_beta
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/grid_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/grid_roi_head.py
new file mode 100644
index 0000000..4c52c79
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/grid_roi_head.py
@@ -0,0 +1,176 @@
+import torch
+
+from mmdet.core import bbox2result, bbox2roi
+from ..builder import HEADS, build_head, build_roi_extractor
+from .standard_roi_head import StandardRoIHead
+
+
+@HEADS.register_module()
+class GridRoIHead(StandardRoIHead):
+    """Grid roi head for Grid R-CNN.
+
+    https://arxiv.org/abs/1811.12030
+    """
+
+    def __init__(self, grid_roi_extractor, grid_head, **kwargs):
+        assert grid_head is not None
+        super(GridRoIHead, self).__init__(**kwargs)
+        if grid_roi_extractor is not None:
+            self.grid_roi_extractor = build_roi_extractor(grid_roi_extractor)
+            self.share_roi_extractor = False
+        else:
+            self.share_roi_extractor = True
+            self.grid_roi_extractor = self.bbox_roi_extractor
+        self.grid_head = build_head(grid_head)
+
+    def init_weights(self, pretrained):
+        """Initialize the weights in head.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        super(GridRoIHead, self).init_weights(pretrained)
+        self.grid_head.init_weights()
+        if not self.share_roi_extractor:
+            self.grid_roi_extractor.init_weights()
+
+    def _random_jitter(self, sampling_results, img_metas, amplitude=0.15):
+        """Ramdom jitter positive proposals for training."""
+        for sampling_result, img_meta in zip(sampling_results, img_metas):
+            bboxes = sampling_result.pos_bboxes
+            random_offsets = bboxes.new_empty(bboxes.shape[0], 4).uniform_(
+                -amplitude, amplitude)
+            # before jittering
+            cxcy = (bboxes[:, 2:4] + bboxes[:, :2]) / 2
+            wh = (bboxes[:, 2:4] - bboxes[:, :2]).abs()
+            # after jittering
+            new_cxcy = cxcy + wh * random_offsets[:, :2]
+            new_wh = wh * (1 + random_offsets[:, 2:])
+            # xywh to xyxy
+            new_x1y1 = (new_cxcy - new_wh / 2)
+            new_x2y2 = (new_cxcy + new_wh / 2)
+            new_bboxes = torch.cat([new_x1y1, new_x2y2], dim=1)
+            # clip bboxes
+            max_shape = img_meta['img_shape']
+            if max_shape is not None:
+                new_bboxes[:, 0::2].clamp_(min=0, max=max_shape[1] - 1)
+                new_bboxes[:, 1::2].clamp_(min=0, max=max_shape[0] - 1)
+
+            sampling_result.pos_bboxes = new_bboxes
+        return sampling_results
+
+    def forward_dummy(self, x, proposals):
+        """Dummy forward function."""
+        # bbox head
+        outs = ()
+        rois = bbox2roi([proposals])
+        if self.with_bbox:
+            bbox_results = self._bbox_forward(x, rois)
+            outs = outs + (bbox_results['cls_score'],
+                           bbox_results['bbox_pred'])
+
+        # grid head
+        grid_rois = rois[:100]
+        grid_feats = self.grid_roi_extractor(
+            x[:self.grid_roi_extractor.num_inputs], grid_rois)
+        if self.with_shared_head:
+            grid_feats = self.shared_head(grid_feats)
+        grid_pred = self.grid_head(grid_feats)
+        outs = outs + (grid_pred, )
+
+        # mask head
+        if self.with_mask:
+            mask_rois = rois[:100]
+            mask_results = self._mask_forward(x, mask_rois)
+            outs = outs + (mask_results['mask_pred'], )
+        return outs
+
+    def _bbox_forward_train(self, x, sampling_results, gt_bboxes, gt_labels,
+                            img_metas):
+        """Run forward function and calculate loss for box head in training."""
+        bbox_results = super(GridRoIHead,
+                             self)._bbox_forward_train(x, sampling_results,
+                                                       gt_bboxes, gt_labels,
+                                                       img_metas)
+
+        # Grid head forward and loss
+        sampling_results = self._random_jitter(sampling_results, img_metas)
+        pos_rois = bbox2roi([res.pos_bboxes for res in sampling_results])
+
+        # GN in head does not support zero shape input
+        if pos_rois.shape[0] == 0:
+            return bbox_results
+
+        grid_feats = self.grid_roi_extractor(
+            x[:self.grid_roi_extractor.num_inputs], pos_rois)
+        if self.with_shared_head:
+            grid_feats = self.shared_head(grid_feats)
+        # Accelerate training
+        max_sample_num_grid = self.train_cfg.get('max_num_grid', 192)
+        sample_idx = torch.randperm(
+            grid_feats.shape[0])[:min(grid_feats.shape[0], max_sample_num_grid
+                                      )]
+        grid_feats = grid_feats[sample_idx]
+
+        grid_pred = self.grid_head(grid_feats)
+
+        grid_targets = self.grid_head.get_targets(sampling_results,
+                                                  self.train_cfg)
+        grid_targets = grid_targets[sample_idx]
+
+        loss_grid = self.grid_head.loss(grid_pred, grid_targets)
+
+        bbox_results['loss_bbox'].update(loss_grid)
+        return bbox_results
+
+    def simple_test(self,
+                    x,
+                    proposal_list,
+                    img_metas,
+                    proposals=None,
+                    rescale=False):
+        """Test without augmentation."""
+        assert self.with_bbox, 'Bbox head must be implemented.'
+
+        det_bboxes, det_labels = self.simple_test_bboxes(
+            x, img_metas, proposal_list, self.test_cfg, rescale=False)
+        # pack rois into bboxes
+        grid_rois = bbox2roi([det_bbox[:, :4] for det_bbox in det_bboxes])
+        if grid_rois.shape[0] != 0:
+            grid_feats = self.grid_roi_extractor(
+                x[:len(self.grid_roi_extractor.featmap_strides)], grid_rois)
+            self.grid_head.test_mode = True
+            grid_pred = self.grid_head(grid_feats)
+            # split batch grid head prediction back to each image
+            num_roi_per_img = tuple(len(det_bbox) for det_bbox in det_bboxes)
+            grid_pred = {
+                k: v.split(num_roi_per_img, 0)
+                for k, v in grid_pred.items()
+            }
+
+            # apply bbox post-processing to each image individually
+            bbox_results = []
+            num_imgs = len(det_bboxes)
+            for i in range(num_imgs):
+                if det_bboxes[i].shape[0] == 0:
+                    bbox_results.append(grid_rois.new_tensor([]))
+                else:
+                    det_bbox = self.grid_head.get_bboxes(
+                        det_bboxes[i], grid_pred['fused'][i], [img_metas[i]])
+                    if rescale:
+                        det_bbox[:, :4] /= img_metas[i]['scale_factor']
+                    bbox_results.append(
+                        bbox2result(det_bbox, det_labels[i],
+                                    self.bbox_head.num_classes))
+        else:
+            bbox_results = [
+                grid_rois.new_tensor([]) for _ in range(len(det_bboxes))
+            ]
+
+        if not self.with_mask:
+            return bbox_results
+        else:
+            segm_results = self.simple_test_mask(
+                x, img_metas, det_bboxes, det_labels, rescale=rescale)
+            return list(zip(bbox_results, segm_results))
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/htc_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/htc_roi_head.py
new file mode 100644
index 0000000..5b5c2ec
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/htc_roi_head.py
@@ -0,0 +1,589 @@
+import torch
+import torch.nn.functional as F
+
+from mmdet.core import (bbox2result, bbox2roi, bbox_mapping, merge_aug_bboxes,
+                        merge_aug_masks, multiclass_nms)
+from ..builder import HEADS, build_head, build_roi_extractor
+from .cascade_roi_head import CascadeRoIHead
+
+
+@HEADS.register_module()
+class HybridTaskCascadeRoIHead(CascadeRoIHead):
+    """Hybrid task cascade roi head including one bbox head and one mask head.
+
+    https://arxiv.org/abs/1901.07518
+    """
+
+    def __init__(self,
+                 num_stages,
+                 stage_loss_weights,
+                 semantic_roi_extractor=None,
+                 semantic_head=None,
+                 semantic_fusion=('bbox', 'mask'),
+                 interleaved=True,
+                 mask_info_flow=True,
+                 **kwargs):
+        super(HybridTaskCascadeRoIHead,
+              self).__init__(num_stages, stage_loss_weights, **kwargs)
+        assert self.with_bbox and self.with_mask
+        assert not self.with_shared_head  # shared head is not supported
+
+        if semantic_head is not None:
+            self.semantic_roi_extractor = build_roi_extractor(
+                semantic_roi_extractor)
+            self.semantic_head = build_head(semantic_head)
+
+        self.semantic_fusion = semantic_fusion
+        self.interleaved = interleaved
+        self.mask_info_flow = mask_info_flow
+
+    def init_weights(self, pretrained):
+        """Initialize the weights in head.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        super(HybridTaskCascadeRoIHead, self).init_weights(pretrained)
+        if self.with_semantic:
+            self.semantic_head.init_weights()
+
+    @property
+    def with_semantic(self):
+        """bool: whether the head has semantic head"""
+        if hasattr(self, 'semantic_head') and self.semantic_head is not None:
+            return True
+        else:
+            return False
+
+    def forward_dummy(self, x, proposals):
+        """Dummy forward function."""
+        outs = ()
+        # semantic head
+        if self.with_semantic:
+            _, semantic_feat = self.semantic_head(x)
+        else:
+            semantic_feat = None
+        # bbox heads
+        rois = bbox2roi([proposals])
+        for i in range(self.num_stages):
+            bbox_results = self._bbox_forward(
+                i, x, rois, semantic_feat=semantic_feat)
+            outs = outs + (bbox_results['cls_score'],
+                           bbox_results['bbox_pred'])
+        # mask heads
+        if self.with_mask:
+            mask_rois = rois[:100]
+            mask_roi_extractor = self.mask_roi_extractor[-1]
+            mask_feats = mask_roi_extractor(
+                x[:len(mask_roi_extractor.featmap_strides)], mask_rois)
+            if self.with_semantic and 'mask' in self.semantic_fusion:
+                mask_semantic_feat = self.semantic_roi_extractor(
+                    [semantic_feat], mask_rois)
+                mask_feats += mask_semantic_feat
+            last_feat = None
+            for i in range(self.num_stages):
+                mask_head = self.mask_head[i]
+                if self.mask_info_flow:
+                    mask_pred, last_feat = mask_head(mask_feats, last_feat)
+                else:
+                    mask_pred = mask_head(mask_feats)
+                outs = outs + (mask_pred, )
+        return outs
+
+    def _bbox_forward_train(self,
+                            stage,
+                            x,
+                            sampling_results,
+                            gt_bboxes,
+                            gt_labels,
+                            rcnn_train_cfg,
+                            semantic_feat=None):
+        """Run forward function and calculate loss for box head in training."""
+        bbox_head = self.bbox_head[stage]
+        rois = bbox2roi([res.bboxes for res in sampling_results])
+        bbox_results = self._bbox_forward(
+            stage, x, rois, semantic_feat=semantic_feat)
+
+        bbox_targets = bbox_head.get_targets(sampling_results, gt_bboxes,
+                                             gt_labels, rcnn_train_cfg)
+        loss_bbox = bbox_head.loss(bbox_results['cls_score'],
+                                   bbox_results['bbox_pred'], rois,
+                                   *bbox_targets)
+
+        bbox_results.update(
+            loss_bbox=loss_bbox,
+            rois=rois,
+            bbox_targets=bbox_targets,
+        )
+        return bbox_results
+
+    def _mask_forward_train(self,
+                            stage,
+                            x,
+                            sampling_results,
+                            gt_masks,
+                            rcnn_train_cfg,
+                            semantic_feat=None):
+        """Run forward function and calculate loss for mask head in
+        training."""
+        mask_roi_extractor = self.mask_roi_extractor[stage]
+        mask_head = self.mask_head[stage]
+        pos_rois = bbox2roi([res.pos_bboxes for res in sampling_results])
+        mask_feats = mask_roi_extractor(x[:mask_roi_extractor.num_inputs],
+                                        pos_rois)
+
+        # semantic feature fusion
+        # element-wise sum for original features and pooled semantic features
+        if self.with_semantic and 'mask' in self.semantic_fusion:
+            mask_semantic_feat = self.semantic_roi_extractor([semantic_feat],
+                                                             pos_rois)
+            if mask_semantic_feat.shape[-2:] != mask_feats.shape[-2:]:
+                mask_semantic_feat = F.adaptive_avg_pool2d(
+                    mask_semantic_feat, mask_feats.shape[-2:])
+            mask_feats += mask_semantic_feat
+
+        # mask information flow
+        # forward all previous mask heads to obtain last_feat, and fuse it
+        # with the normal mask feature
+        if self.mask_info_flow:
+            last_feat = None
+            for i in range(stage):
+                last_feat = self.mask_head[i](
+                    mask_feats, last_feat, return_logits=False)
+            mask_pred = mask_head(mask_feats, last_feat, return_feat=False)
+        else:
+            mask_pred = mask_head(mask_feats, return_feat=False)
+
+        mask_targets = mask_head.get_targets(sampling_results, gt_masks,
+                                             rcnn_train_cfg)
+        pos_labels = torch.cat([res.pos_gt_labels for res in sampling_results])
+        loss_mask = mask_head.loss(mask_pred, mask_targets, pos_labels)
+
+        mask_results = dict(loss_mask=loss_mask)
+        return mask_results
+
+    def _bbox_forward(self, stage, x, rois, semantic_feat=None):
+        """Box head forward function used in both training and testing."""
+        bbox_roi_extractor = self.bbox_roi_extractor[stage]
+        bbox_head = self.bbox_head[stage]
+        bbox_feats = bbox_roi_extractor(
+            x[:len(bbox_roi_extractor.featmap_strides)], rois)
+        if self.with_semantic and 'bbox' in self.semantic_fusion:
+            bbox_semantic_feat = self.semantic_roi_extractor([semantic_feat],
+                                                             rois)
+            if bbox_semantic_feat.shape[-2:] != bbox_feats.shape[-2:]:
+                bbox_semantic_feat = F.adaptive_avg_pool2d(
+                    bbox_semantic_feat, bbox_feats.shape[-2:])
+            bbox_feats += bbox_semantic_feat
+        cls_score, bbox_pred = bbox_head(bbox_feats)
+
+        bbox_results = dict(cls_score=cls_score, bbox_pred=bbox_pred)
+        return bbox_results
+
+    def _mask_forward_test(self, stage, x, bboxes, semantic_feat=None):
+        """Mask head forward function for testing."""
+        mask_roi_extractor = self.mask_roi_extractor[stage]
+        mask_head = self.mask_head[stage]
+        mask_rois = bbox2roi([bboxes])
+        mask_feats = mask_roi_extractor(
+            x[:len(mask_roi_extractor.featmap_strides)], mask_rois)
+        if self.with_semantic and 'mask' in self.semantic_fusion:
+            mask_semantic_feat = self.semantic_roi_extractor([semantic_feat],
+                                                             mask_rois)
+            if mask_semantic_feat.shape[-2:] != mask_feats.shape[-2:]:
+                mask_semantic_feat = F.adaptive_avg_pool2d(
+                    mask_semantic_feat, mask_feats.shape[-2:])
+            mask_feats += mask_semantic_feat
+        if self.mask_info_flow:
+            last_feat = None
+            last_pred = None
+            for i in range(stage):
+                mask_pred, last_feat = self.mask_head[i](mask_feats, last_feat)
+                if last_pred is not None:
+                    mask_pred = mask_pred + last_pred
+                last_pred = mask_pred
+            mask_pred = mask_head(mask_feats, last_feat, return_feat=False)
+            if last_pred is not None:
+                mask_pred = mask_pred + last_pred
+        else:
+            mask_pred = mask_head(mask_feats)
+        return mask_pred
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      proposal_list,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None,
+                      gt_semantic_seg=None):
+        """
+        Args:
+            x (list[Tensor]): list of multi-level img features.
+
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+
+            proposal_list (list[Tensors]): list of region proposals.
+
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+
+            gt_labels (list[Tensor]): class indices corresponding to each box
+
+            gt_bboxes_ignore (None, list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+            gt_masks (None, Tensor) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+            gt_semantic_seg (None, list[Tensor]): semantic segmentation masks
+                used if the architecture supports semantic segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        # semantic segmentation part
+        # 2 outputs: segmentation prediction and embedded features
+        losses = dict()
+        if self.with_semantic:
+            semantic_pred, semantic_feat = self.semantic_head(x)
+            loss_seg = self.semantic_head.loss(semantic_pred, gt_semantic_seg)
+            losses['loss_semantic_seg'] = loss_seg
+        else:
+            semantic_feat = None
+
+        for i in range(self.num_stages):
+            self.current_stage = i
+            rcnn_train_cfg = self.train_cfg[i]
+            lw = self.stage_loss_weights[i]
+
+            # assign gts and sample proposals
+            sampling_results = []
+            bbox_assigner = self.bbox_assigner[i]
+            bbox_sampler = self.bbox_sampler[i]
+            num_imgs = len(img_metas)
+            if gt_bboxes_ignore is None:
+                gt_bboxes_ignore = [None for _ in range(num_imgs)]
+
+            for j in range(num_imgs):
+                assign_result = bbox_assigner.assign(proposal_list[j],
+                                                     gt_bboxes[j],
+                                                     gt_bboxes_ignore[j],
+                                                     gt_labels[j])
+                sampling_result = bbox_sampler.sample(
+                    assign_result,
+                    proposal_list[j],
+                    gt_bboxes[j],
+                    gt_labels[j],
+                    feats=[lvl_feat[j][None] for lvl_feat in x])
+                sampling_results.append(sampling_result)
+
+            # bbox head forward and loss
+            bbox_results = \
+                self._bbox_forward_train(
+                    i, x, sampling_results, gt_bboxes, gt_labels,
+                    rcnn_train_cfg, semantic_feat)
+            roi_labels = bbox_results['bbox_targets'][0]
+
+            for name, value in bbox_results['loss_bbox'].items():
+                losses[f's{i}.{name}'] = (
+                    value * lw if 'loss' in name else value)
+
+            # mask head forward and loss
+            if self.with_mask:
+                # interleaved execution: use regressed bboxes by the box branch
+                # to train the mask branch
+                if self.interleaved:
+                    pos_is_gts = [res.pos_is_gt for res in sampling_results]
+                    with torch.no_grad():
+                        proposal_list = self.bbox_head[i].refine_bboxes(
+                            bbox_results['rois'], roi_labels,
+                            bbox_results['bbox_pred'], pos_is_gts, img_metas)
+                        # re-assign and sample 512 RoIs from 512 RoIs
+                        sampling_results = []
+                        for j in range(num_imgs):
+                            assign_result = bbox_assigner.assign(
+                                proposal_list[j], gt_bboxes[j],
+                                gt_bboxes_ignore[j], gt_labels[j])
+                            sampling_result = bbox_sampler.sample(
+                                assign_result,
+                                proposal_list[j],
+                                gt_bboxes[j],
+                                gt_labels[j],
+                                feats=[lvl_feat[j][None] for lvl_feat in x])
+                            sampling_results.append(sampling_result)
+                mask_results = self._mask_forward_train(
+                    i, x, sampling_results, gt_masks, rcnn_train_cfg,
+                    semantic_feat)
+                for name, value in mask_results['loss_mask'].items():
+                    losses[f's{i}.{name}'] = (
+                        value * lw if 'loss' in name else value)
+
+            # refine bboxes (same as Cascade R-CNN)
+            if i < self.num_stages - 1 and not self.interleaved:
+                pos_is_gts = [res.pos_is_gt for res in sampling_results]
+                with torch.no_grad():
+                    proposal_list = self.bbox_head[i].refine_bboxes(
+                        bbox_results['rois'], roi_labels,
+                        bbox_results['bbox_pred'], pos_is_gts, img_metas)
+
+        return losses
+
+    def simple_test(self, x, proposal_list, img_metas, rescale=False):
+        """Test without augmentation."""
+        if self.with_semantic:
+            _, semantic_feat = self.semantic_head(x)
+        else:
+            semantic_feat = None
+
+        num_imgs = len(proposal_list)
+        img_shapes = tuple(meta['img_shape'] for meta in img_metas)
+        ori_shapes = tuple(meta['ori_shape'] for meta in img_metas)
+        scale_factors = tuple(meta['scale_factor'] for meta in img_metas)
+
+        # "ms" in variable names means multi-stage
+        ms_bbox_result = {}
+        ms_segm_result = {}
+        ms_scores = []
+        rcnn_test_cfg = self.test_cfg
+
+        rois = bbox2roi(proposal_list)
+        for i in range(self.num_stages):
+            bbox_head = self.bbox_head[i]
+            bbox_results = self._bbox_forward(
+                i, x, rois, semantic_feat=semantic_feat)
+            # split batch bbox prediction back to each image
+            cls_score = bbox_results['cls_score']
+            bbox_pred = bbox_results['bbox_pred']
+            num_proposals_per_img = tuple(len(p) for p in proposal_list)
+            rois = rois.split(num_proposals_per_img, 0)
+            cls_score = cls_score.split(num_proposals_per_img, 0)
+            bbox_pred = bbox_pred.split(num_proposals_per_img, 0)
+            ms_scores.append(cls_score)
+
+            if i < self.num_stages - 1:
+                bbox_label = [s[:, :-1].argmax(dim=1) for s in cls_score]
+                rois = torch.cat([
+                    bbox_head.regress_by_class(rois[i], bbox_label[i],
+                                               bbox_pred[i], img_metas[i])
+                    for i in range(num_imgs)
+                ])
+
+        # average scores of each image by stages
+        cls_score = [
+            sum([score[i] for score in ms_scores]) / float(len(ms_scores))
+            for i in range(num_imgs)
+        ]
+
+        # apply bbox post-processing to each image individually
+        det_bboxes = []
+        det_labels = []
+        for i in range(num_imgs):
+            det_bbox, det_label = self.bbox_head[-1].get_bboxes(
+                rois[i],
+                cls_score[i],
+                bbox_pred[i],
+                img_shapes[i],
+                scale_factors[i],
+                rescale=rescale,
+                cfg=rcnn_test_cfg)
+            det_bboxes.append(det_bbox)
+            det_labels.append(det_label)
+        bbox_result = [
+            bbox2result(det_bboxes[i], det_labels[i],
+                        self.bbox_head[-1].num_classes)
+            for i in range(num_imgs)
+        ]
+        ms_bbox_result['ensemble'] = bbox_result
+
+        if self.with_mask:
+            if all(det_bbox.shape[0] == 0 for det_bbox in det_bboxes):
+                mask_classes = self.mask_head[-1].num_classes
+                segm_results = [[[] for _ in range(mask_classes)]
+                                for _ in range(num_imgs)]
+            else:
+                if rescale and not isinstance(scale_factors[0], float):
+                    scale_factors = [
+                        torch.from_numpy(scale_factor).to(det_bboxes[0].device)
+                        for scale_factor in scale_factors
+                    ]
+                _bboxes = [
+                    det_bboxes[i][:, :4] *
+                    scale_factors[i] if rescale else det_bboxes[i]
+                    for i in range(num_imgs)
+                ]
+                mask_rois = bbox2roi(_bboxes)
+                aug_masks = []
+                mask_roi_extractor = self.mask_roi_extractor[-1]
+                mask_feats = mask_roi_extractor(
+                    x[:len(mask_roi_extractor.featmap_strides)], mask_rois)
+                if self.with_semantic and 'mask' in self.semantic_fusion:
+                    mask_semantic_feat = self.semantic_roi_extractor(
+                        [semantic_feat], mask_rois)
+                    mask_feats += mask_semantic_feat
+                last_feat = None
+
+                num_bbox_per_img = tuple(len(_bbox) for _bbox in _bboxes)
+                for i in range(self.num_stages):
+                    mask_head = self.mask_head[i]
+                    if self.mask_info_flow:
+                        mask_pred, last_feat = mask_head(mask_feats, last_feat)
+                    else:
+                        mask_pred = mask_head(mask_feats)
+
+                    # split batch mask prediction back to each image
+                    mask_pred = mask_pred.split(num_bbox_per_img, 0)
+                    aug_masks.append(
+                        [mask.sigmoid().cpu().numpy() for mask in mask_pred])
+
+                # apply mask post-processing to each image individually
+                segm_results = []
+                for i in range(num_imgs):
+                    if det_bboxes[i].shape[0] == 0:
+                        segm_results.append(
+                            [[]
+                             for _ in range(self.mask_head[-1].num_classes)])
+                    else:
+                        aug_mask = [mask[i] for mask in aug_masks]
+                        merged_mask = merge_aug_masks(
+                            aug_mask, [[img_metas[i]]] * self.num_stages,
+                            rcnn_test_cfg)
+                        segm_result = self.mask_head[-1].get_seg_masks(
+                            merged_mask, _bboxes[i], det_labels[i],
+                            rcnn_test_cfg, ori_shapes[i], scale_factors[i],
+                            rescale)
+                        segm_results.append(segm_result)
+            ms_segm_result['ensemble'] = segm_results
+
+        if self.with_mask:
+            results = list(
+                zip(ms_bbox_result['ensemble'], ms_segm_result['ensemble']))
+        else:
+            results = ms_bbox_result['ensemble']
+
+        return results
+
+    def aug_test(self, img_feats, proposal_list, img_metas, rescale=False):
+        """Test with augmentations.
+
+        If rescale is False, then returned bboxes and masks will fit the scale
+        of imgs[0].
+        """
+        if self.with_semantic:
+            semantic_feats = [
+                self.semantic_head(feat)[1] for feat in img_feats
+            ]
+        else:
+            semantic_feats = [None] * len(img_metas)
+
+        rcnn_test_cfg = self.test_cfg
+        aug_bboxes = []
+        aug_scores = []
+        for x, img_meta, semantic in zip(img_feats, img_metas, semantic_feats):
+            # only one image in the batch
+            img_shape = img_meta[0]['img_shape']
+            scale_factor = img_meta[0]['scale_factor']
+            flip = img_meta[0]['flip']
+            flip_direction = img_meta[0]['flip_direction']
+
+            proposals = bbox_mapping(proposal_list[0][:, :4], img_shape,
+                                     scale_factor, flip, flip_direction)
+            # "ms" in variable names means multi-stage
+            ms_scores = []
+
+            rois = bbox2roi([proposals])
+            for i in range(self.num_stages):
+                bbox_head = self.bbox_head[i]
+                bbox_results = self._bbox_forward(
+                    i, x, rois, semantic_feat=semantic)
+                ms_scores.append(bbox_results['cls_score'])
+
+                if i < self.num_stages - 1:
+                    bbox_label = bbox_results['cls_score'].argmax(dim=1)
+                    rois = bbox_head.regress_by_class(
+                        rois, bbox_label, bbox_results['bbox_pred'],
+                        img_meta[0])
+
+            cls_score = sum(ms_scores) / float(len(ms_scores))
+            bboxes, scores = self.bbox_head[-1].get_bboxes(
+                rois,
+                cls_score,
+                bbox_results['bbox_pred'],
+                img_shape,
+                scale_factor,
+                rescale=False,
+                cfg=None)
+            aug_bboxes.append(bboxes)
+            aug_scores.append(scores)
+
+        # after merging, bboxes will be rescaled to the original image size
+        merged_bboxes, merged_scores = merge_aug_bboxes(
+            aug_bboxes, aug_scores, img_metas, rcnn_test_cfg)
+        det_bboxes, det_labels = multiclass_nms(merged_bboxes, merged_scores,
+                                                rcnn_test_cfg.score_thr,
+                                                rcnn_test_cfg.nms,
+                                                rcnn_test_cfg.max_per_img)
+
+        bbox_result = bbox2result(det_bboxes, det_labels,
+                                  self.bbox_head[-1].num_classes)
+
+        if self.with_mask:
+            if det_bboxes.shape[0] == 0:
+                segm_result = [[[]
+                                for _ in range(self.mask_head[-1].num_classes)]
+                               ]
+            else:
+                aug_masks = []
+                aug_img_metas = []
+                for x, img_meta, semantic in zip(img_feats, img_metas,
+                                                 semantic_feats):
+                    img_shape = img_meta[0]['img_shape']
+                    scale_factor = img_meta[0]['scale_factor']
+                    flip = img_meta[0]['flip']
+                    flip_direction = img_meta[0]['flip_direction']
+                    _bboxes = bbox_mapping(det_bboxes[:, :4], img_shape,
+                                           scale_factor, flip, flip_direction)
+                    mask_rois = bbox2roi([_bboxes])
+                    mask_feats = self.mask_roi_extractor[-1](
+                        x[:len(self.mask_roi_extractor[-1].featmap_strides)],
+                        mask_rois)
+                    if self.with_semantic:
+                        semantic_feat = semantic
+                        mask_semantic_feat = self.semantic_roi_extractor(
+                            [semantic_feat], mask_rois)
+                        if mask_semantic_feat.shape[-2:] != mask_feats.shape[
+                                -2:]:
+                            mask_semantic_feat = F.adaptive_avg_pool2d(
+                                mask_semantic_feat, mask_feats.shape[-2:])
+                        mask_feats += mask_semantic_feat
+                    last_feat = None
+                    for i in range(self.num_stages):
+                        mask_head = self.mask_head[i]
+                        if self.mask_info_flow:
+                            mask_pred, last_feat = mask_head(
+                                mask_feats, last_feat)
+                        else:
+                            mask_pred = mask_head(mask_feats)
+                        aug_masks.append(mask_pred.sigmoid().cpu().numpy())
+                        aug_img_metas.append(img_meta)
+                merged_masks = merge_aug_masks(aug_masks, aug_img_metas,
+                                               self.test_cfg)
+
+                ori_shape = img_metas[0][0]['ori_shape']
+                segm_result = self.mask_head[-1].get_seg_masks(
+                    merged_masks,
+                    det_bboxes,
+                    det_labels,
+                    rcnn_test_cfg,
+                    ori_shape,
+                    scale_factor=1.0,
+                    rescale=False)
+            return [(bbox_result, segm_result)]
+        else:
+            return [bbox_result]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/__init__.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/__init__.py
new file mode 100644
index 0000000..ebe1184
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/__init__.py
@@ -0,0 +1,18 @@
+from .coarse_mask_head import CoarseMaskHead
+from .fcn_mask_head import FCNMaskHead
+from .feature_relay_head import FeatureRelayHead
+from .fused_semantic_head import FusedSemanticHead
+from .global_context_head import GlobalContextHead
+from .grid_head import GridHead
+from .htc_mask_head import HTCMaskHead
+from .mask_point_head import MaskPointHead
+from .maskiou_head import MaskIoUHead
+from .scnet_mask_head import SCNetMaskHead
+from .scnet_semantic_head import SCNetSemanticHead
+from .condconv_mask_head import CondConvMaskHead
+
+__all__ = [
+    'FCNMaskHead', 'HTCMaskHead', 'FusedSemanticHead', 'GridHead',
+    'MaskIoUHead', 'CoarseMaskHead', 'MaskPointHead', 'SCNetMaskHead',
+    'SCNetSemanticHead', 'GlobalContextHead', 'FeatureRelayHead', 'CondConvMaskHead'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/coarse_mask_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/coarse_mask_head.py
new file mode 100644
index 0000000..d665dff
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/coarse_mask_head.py
@@ -0,0 +1,91 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule, Linear, constant_init, xavier_init
+from mmcv.runner import auto_fp16
+
+from mmdet.models.builder import HEADS
+from .fcn_mask_head import FCNMaskHead
+
+
+@HEADS.register_module()
+class CoarseMaskHead(FCNMaskHead):
+    """Coarse mask head used in PointRend.
+
+    Compared with standard ``FCNMaskHead``, ``CoarseMaskHead`` will downsample
+    the input feature map instead of upsample it.
+
+    Args:
+        num_convs (int): Number of conv layers in the head. Default: 0.
+        num_fcs (int): Number of fc layers in the head. Default: 2.
+        fc_out_channels (int): Number of output channels of fc layer.
+            Default: 1024.
+        downsample_factor (int): The factor that feature map is downsampled by.
+            Default: 2.
+    """
+
+    def __init__(self,
+                 num_convs=0,
+                 num_fcs=2,
+                 fc_out_channels=1024,
+                 downsample_factor=2,
+                 *arg,
+                 **kwarg):
+        super(CoarseMaskHead, self).__init__(
+            *arg, num_convs=num_convs, upsample_cfg=dict(type=None), **kwarg)
+        self.num_fcs = num_fcs
+        assert self.num_fcs > 0
+        self.fc_out_channels = fc_out_channels
+        self.downsample_factor = downsample_factor
+        assert self.downsample_factor >= 1
+        # remove conv_logit
+        delattr(self, 'conv_logits')
+
+        if downsample_factor > 1:
+            downsample_in_channels = (
+                self.conv_out_channels
+                if self.num_convs > 0 else self.in_channels)
+            self.downsample_conv = ConvModule(
+                downsample_in_channels,
+                self.conv_out_channels,
+                kernel_size=downsample_factor,
+                stride=downsample_factor,
+                padding=0,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg)
+        else:
+            self.downsample_conv = None
+
+        self.output_size = (self.roi_feat_size[0] // downsample_factor,
+                            self.roi_feat_size[1] // downsample_factor)
+        self.output_area = self.output_size[0] * self.output_size[1]
+
+        last_layer_dim = self.conv_out_channels * self.output_area
+
+        self.fcs = nn.ModuleList()
+        for i in range(num_fcs):
+            fc_in_channels = (
+                last_layer_dim if i == 0 else self.fc_out_channels)
+            self.fcs.append(Linear(fc_in_channels, self.fc_out_channels))
+        last_layer_dim = self.fc_out_channels
+        output_channels = self.num_classes * self.output_area
+        self.fc_logits = Linear(last_layer_dim, output_channels)
+
+    def init_weights(self):
+        for m in self.fcs.modules():
+            if isinstance(m, nn.Linear):
+                xavier_init(m)
+        constant_init(self.fc_logits, 0.001)
+
+    @auto_fp16()
+    def forward(self, x):
+        for conv in self.convs:
+            x = conv(x)
+
+        if self.downsample_conv is not None:
+            x = self.downsample_conv(x)
+
+        x = x.flatten(1)
+        for fc in self.fcs:
+            x = self.relu(fc(x))
+        mask_pred = self.fc_logits(x).view(
+            x.size(0), self.num_classes, *self.output_size)
+        return mask_pred
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/condconv_mask_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/condconv_mask_head.py
new file mode 100644
index 0000000..82c5aae
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/condconv_mask_head.py
@@ -0,0 +1,517 @@
+from typing import Dict
+import math
+
+import torch
+from torch import nn
+import torch.functional as F
+
+from mmcv.cnn import ConvModule
+
+from functools import partial
+
+from mmdet.models.builder import HEADS, build_loss
+from mmdet.utils.common import compute_locations
+import torch.nn.functional as F
+
+INF = 100000000
+
+@HEADS.register_module()
+class CondConvMaskHead(nn.Module):
+    def __init__(self, branch_cfg, head_cfg):
+        super().__init__()
+        self.branch_cfg = branch_cfg
+        self.head_cfg = head_cfg
+        self.branch = build_mask_branch(branch_cfg)
+        head_cfg['in_channels'] = branch_cfg['out_channels']
+        self.head = build_dynamic_mask_head(head_cfg)
+        self.max_proposals = head_cfg.max_proposals
+    def forward(self, features, pred_instances, gt_masks=None, gt_labels=None): # gt_labels for sem_loss_on
+        mask_feats, losses = self.branch(features, gt_masks, gt_labels)
+
+        if self.training:
+            if 0 <= self.max_proposals < len(pred_instances):
+                inds = torch.randperm(len(pred_instances), device=mask_feats.device).long()
+                print("clipping proposals from {} to {}".format(
+                    len(pred_instances), self.max_proposals
+                ))
+                pred_instances = pred_instances[inds[:self.max_proposals]]
+
+            loss_mask = self.head(mask_feats, self.branch.out_stride, pred_instances, gt_masks)
+            losses.update(loss_mask)
+            return losses
+        else:
+            pred_mask = self.head(mask_feats, self.branch.out_stride, pred_instances)
+            return pred_mask
+
+
+def build_mask_branch(cfg):
+    return MaskBranch(cfg)
+            
+# modified from https://github.com/aim-uofa/AdelaiDet/blob/0157227f966eda93c1299a402537b616207ba226/adet/modeling/condinst/
+
+def aligned_bilinear(tensor, factor):
+    assert tensor.dim() == 4
+    assert factor >= 1
+    assert int(factor) == factor
+
+    if factor == 1:
+        return tensor
+
+    h, w = tensor.size()[2:]
+    tensor = F.pad(tensor, pad=(0, 1, 0, 1), mode="replicate")
+    oh = factor * h + 1
+    ow = factor * w + 1
+    tensor = F.interpolate(
+        tensor, size=(oh, ow),
+        mode='bilinear',
+        align_corners=True
+    )
+    tensor = F.pad(
+        tensor, pad=(factor // 2, 0, factor // 2, 0),
+        mode="replicate"
+    )
+
+    return tensor[:, :, :oh - 1, :ow - 1]
+
+def downsample_mask(mask, stride):
+    assert stride >= 1
+    assert int(stride) == stride
+    mask = mask[:, None, stride // 2::stride,
+                        stride // 2::stride]
+    return mask
+
+
+class MaskBranch(nn.Module):
+    def __init__(self, cfg):
+        super().__init__()
+        self.in_features = cfg.in_features
+        self.sem_loss_on = cfg.semantic_loss_on
+        self.num_outputs = cfg.out_channels
+        norm = cfg.norm
+        num_convs = cfg.num_convs
+        channels = cfg.channels
+        self.out_stride = cfg.out_stride[0] # select the highest resolution
+        self.in_channels = cfg.in_channels
+
+
+        self.refine = nn.ModuleList()
+        for in_feature in self.in_features:
+            self.refine.append(ConvModule(
+                self.in_channels,
+                channels, 3, 1, padding=1, norm_cfg=norm
+            ))
+
+        tower = []
+        for i in range(num_convs):
+            tower.append(ConvModule(
+                channels, channels, 3, 1, padding=1, norm_cfg=norm
+            ))
+        tower.append(nn.Conv2d(
+            channels, max(self.num_outputs, 1), 1
+        ))
+        self.add_module('tower', nn.Sequential(*tower))
+
+        # TODO: add sem_loss_on support
+        if self.sem_loss_on:
+            prior_prob = cfg.loss_sem.pop('prior_prob')
+            self.loss = build_loss(cfg.loss_sem)
+
+            in_channels = self.in_channels
+            self.seg_head = nn.Sequential(
+                ConvModule(in_channels, channels, kernel_size=3, stride=1, padding=1, norm_cfg=norm),
+                ConvModule(channels, channels, kernel_size=3, stride=1, padding=1, norm_cfg=norm)
+            )
+
+            self.num_classes = cfg.num_classes
+            self.logits = nn.Conv2d(channels, self.num_classes, kernel_size=1, stride=1)
+
+            bias_value = -math.log((1 - prior_prob) / prior_prob)
+            torch.nn.init.constant_(self.logits.bias, bias_value)
+
+    def forward(self, features, gt_masks=None, gt_labels=None):
+        for i, f in enumerate(self.in_features):
+            if i == 0:
+                x = self.refine[i](features[f])
+            else:
+                x_p = self.refine[i](features[f])
+
+                target_h, target_w = x.size()[2:]
+                h, w = x_p.size()[2:]
+                assert target_h % h == 0
+                assert target_w % w == 0
+                factor_h, factor_w = target_h // h, target_w // w
+                assert factor_h == factor_w
+                x_p = aligned_bilinear(x_p, factor_h)
+                x = x + x_p
+
+        mask_feats = self.tower(x)
+
+        if self.num_outputs == 0:
+            mask_feats = mask_feats[:, :self.num_outputs]
+
+        losses = {}
+        # auxiliary thing semantic loss
+        if self.training and self.sem_loss_on:
+            logits_pred = self.logits(self.seg_head(
+                features[self.in_features[0]]
+            ))
+
+            # compute semantic targets
+            semantic_targets = []
+            for gt_mask, gt_label in zip(gt_masks,gt_labels):
+                h, w = gt_mask.size()[-2:]
+                areas = gt_mask.sum(dim=-1).sum(dim=-1)
+                areas = areas[:, None, None].repeat(1, h, w)
+                areas[gt_mask == 0] = INF
+                areas = areas.permute(1, 2, 0).reshape(h * w, -1)
+                min_areas, inds = areas.min(dim=1)
+                per_im_sematic_targets = gt_label[inds]
+                per_im_sematic_targets[min_areas == INF] = self.num_classes
+                per_im_sematic_targets = per_im_sematic_targets.reshape(h, w)
+                semantic_targets.append(per_im_sematic_targets)
+
+            semantic_targets = torch.stack(semantic_targets, dim=0)
+
+            # resize target to reduce memory
+            semantic_targets = downsample_mask(semantic_targets, self.out_stride)
+
+            num_pos = (semantic_targets < self.num_classes).sum().float().clamp(min=1.0)
+            
+            loss_sem = self.loss(logits_pred.permute(0,2,3,1).reshape(-1, self.num_classes), semantic_targets.reshape(-1).to(torch.long), avg_factor=num_pos)
+            losses['loss_sem'] = loss_sem
+
+        return mask_feats, losses
+
+
+# dice loss
+def dice_coefficient(x, target):
+    eps = 1e-5
+    n_inst = x.size(0)
+    x = x.reshape(n_inst, -1)
+    target = target.reshape(n_inst, -1)
+    intersection = (x * target).sum(dim=1)
+    union = (x ** 2.0).sum(dim=1) + (target ** 2.0).sum(dim=1) + eps
+    loss = 1. - (2 * intersection / union)
+    return loss
+
+def parse_attn_params(params, channels, query_nums):
+    assert params.dim() == 2
+    assert params.size(1) == sum(query_nums)
+
+    num_insts = params.size(0)
+    num_layers = len(query_nums)
+
+    query_splits = list(torch.split_with_sizes(
+        params, query_nums, dim=1
+    ))
+
+    for l in range(num_layers):
+        if l < num_layers - 1:
+            # N x dq x nhead
+            query_splits[l] = query_splits[l].reshape(num_insts, -1, channels)
+        else:
+            # N x dq x nhead
+            query_splits[l] = query_splits[l].reshape(num_insts, -1, 1)
+
+    return query_splits
+
+def parse_dynamic_params(params, channels, weight_nums, bias_nums):
+    assert params.dim() == 2
+    assert len(weight_nums) == len(bias_nums)
+    assert params.size(1) == sum(weight_nums) + sum(bias_nums)
+
+    num_insts = params.size(0)
+    num_layers = len(weight_nums)
+
+    params_splits = list(torch.split_with_sizes(
+        params, weight_nums + bias_nums, dim=1
+    ))
+
+    weight_splits = params_splits[:num_layers]
+    bias_splits = params_splits[num_layers:]
+
+    for l in range(num_layers):
+        if l < num_layers - 1:
+            # out_channels x in_channels x 1 x 1
+            weight_splits[l] = weight_splits[l].reshape(num_insts * channels, -1, 1, 1)
+            bias_splits[l] = bias_splits[l].reshape(num_insts * channels)
+        else:
+            # out_channels x in_channels x 1 x 1
+            weight_splits[l] = weight_splits[l].reshape(num_insts * 1, -1, 1, 1)
+            bias_splits[l] = bias_splits[l].reshape(num_insts)
+
+    return weight_splits, bias_splits
+
+def parse_dynamic_params_with_aux(params, channels, weight_nums, bias_nums, aux_weight_nums, aux_bias_nums):
+    assert params.size(1) == sum(weight_nums) + sum(bias_nums) + sum(aux_weight_nums) + sum(aux_bias_nums)
+
+    num_insts = params.size(0)
+    num_layers = len(weight_nums)
+
+    params_splits = list(torch.split_with_sizes(
+        params, weight_nums + bias_nums + aux_weight_nums + aux_bias_nums, dim=1
+    ))
+
+    weight_splits = params_splits[:num_layers]
+    bias_splits = params_splits[num_layers:len(weight_nums)+len(bias_nums)]
+    aux_weight_splits = params_splits[len(weight_nums)+len(bias_nums):len(weight_nums)+len(bias_nums)+len(aux_weight_nums)]
+    aux_bias_splits = params_splits[len(weight_nums)+len(bias_nums)+len(aux_weight_nums):]
+
+    for l in range(num_layers):
+        if l < num_layers - 1:
+            # out_channels x in_channels x 1 x 1
+            weight_splits[l] = weight_splits[l].reshape(num_insts * channels, -1, 1, 1)
+            bias_splits[l] = bias_splits[l].reshape(num_insts * channels)
+        else:
+            # out_channels x in_channels x 1 x 1
+            weight_splits[l] = weight_splits[l].reshape(num_insts * 1, -1, 1, 1)
+            bias_splits[l] = bias_splits[l].reshape(num_insts)
+
+    for l in range(num_layers-1):
+        aux_weight_splits[l] = aux_weight_splits[l].reshape(num_insts * 1, -1, 1, 1)
+        aux_bias_splits[l] = aux_bias_splits[l].reshape(num_insts)            
+
+    return weight_splits, bias_splits, aux_weight_splits, aux_bias_splits
+
+
+def build_dynamic_mask_head(cfg):
+    return DynamicMaskHead(cfg)
+
+
+class DynamicMaskHead(nn.Module):
+    def __init__(self, cfg):
+        super().__init__()
+        self.num_layers = cfg.num_layers
+        self.channels = cfg.channels
+        self.in_channels = cfg.in_channels
+        self.mask_out_stride = cfg.mask_out_stride
+        self.disable_rel_coords = cfg.disable_rel_coords
+        self.rel_num = cfg.rel_num
+        self.rel_abs = cfg.get('rel_abs', False)
+        self.coord_conv = cfg.get('coord_conv', False)
+        self.aux_loss = cfg.get('aux_loss', False)
+        self.mask_loss_weight = cfg.get('mask_loss_weight', [1.]*self.num_layers if self.aux_loss else [1.])
+        self.norm_rel = cfg.get('norm_rel', None)
+
+        soi = cfg.sizes_of_interest
+        self.register_buffer("sizes_of_interest", torch.tensor(soi))
+
+        # use condconv or attention map
+        self.use_att_map = cfg.get('use_att_map', False)
+        self.att_fn = cfg.get('att_fn', 'relu')
+        
+        if self.use_att_map:
+            query_nums = []
+            for l in range(self.num_layers):
+                if l == 0:
+                    if not self.disable_rel_coords:
+                        query_nums.append((self.in_channels + self.rel_num * 2) * self.channels)
+                    else:
+                        query_nums.append(self.in_channels * self.channels)
+                elif l == self.num_layers - 1:
+                    query_nums.append(self.channels * 1)
+                else:
+                    query_nums.append(self.channels * self.channels)
+
+            self.query_nums = query_nums
+            self.num_gen_params = sum(query_nums)
+        else:
+            weight_nums, bias_nums = [], []
+            for l in range(self.num_layers):
+                if l == 0:
+                    if not self.disable_rel_coords:
+                        weight_nums.append((self.in_channels + self.rel_num * 2) * self.channels)
+                    else:
+                        weight_nums.append(self.in_channels * self.channels)
+                    bias_nums.append(self.channels)
+                elif l == self.num_layers - 1:
+                    weight_nums.append(self.channels * 1)
+                    bias_nums.append(1)
+                else:
+                    if self.coord_conv:
+                        weight_nums.append((self.channels+self.rel_num*2) * self.channels)
+                        bias_nums.append(self.channels)
+                    else:
+                        weight_nums.append(self.channels * self.channels)
+                        bias_nums.append(self.channels)
+
+            if self.aux_loss:
+                aux_weight_nums, aux_bias_nums = [], []
+                for l in range(self.num_layers-1):
+                    if l == 0:
+                        if not self.disable_rel_coords:
+                            aux_weight_nums.append(self.in_channels + self.rel_num * 2)
+                        else:
+                            aux_weight_nums.append(self.in_channels)
+                        aux_bias_nums.append(1)
+                    else:
+                        if self.coord_conv:
+                            aux_weight_nums.append((self.channels+self.rel_num*2) * 1)
+                            aux_bias_nums.append(1)
+                        else:
+                            aux_weight_nums.append(self.channels * 1)
+                            aux_bias_nums.append(1)
+                self.aux_weight_nums = aux_weight_nums
+                self.aux_bias_nums = aux_bias_nums
+            self.weight_nums = weight_nums
+            self.bias_nums = bias_nums
+            self.num_gen_params = sum(weight_nums) + sum(bias_nums)
+            if self.aux_loss:
+                self.num_gen_params = self.num_gen_params + sum(self.aux_weight_nums) + sum(self.aux_bias_nums)
+
+    def mask_heads_forward(self, features, weights, biases, num_insts, aux_weights=[], aux_biases=[]):
+        '''
+        :param features
+        :param weights: [w0, w1, ...]
+        :param bias: [b0, b1, ...]
+        :return:
+        '''
+        assert features.dim() == 4
+        n_layers = len(weights)
+        x = features
+        x_list = []
+        n_inst, _, H, W = self.relative_coords.shape
+        for i, (w, b) in enumerate(zip(weights, biases)):
+            if self.coord_conv and i == 1:
+                x = x.reshape(n_inst, -1, H, W)
+                x = torch.cat([self.relative_coords, x], dim=1)
+                x = x.reshape(1, -1, H, W)
+            if len(aux_weights)>0 and len(aux_biases)>0:
+                aux_weight = aux_weights.pop(0)
+                aux_bias = aux_biases.pop(0)
+                aux_x = F.conv2d(
+                    x, aux_weight, bias=aux_bias,
+                    stride=1, padding=0,
+                    groups=num_insts
+                )
+                x_list.append(aux_x)
+            x = F.conv2d(
+                x, w, bias=b,
+                stride=1, padding=0,
+                groups=num_insts
+            )
+            if i < n_layers - 1:
+                x = F.relu(x)
+        x_list.append(x)
+        return x_list
+
+    def attn_forward(self, mask_feats, query, n_inst):
+        # N, C, HW
+        x = mask_feats
+        # N, HW, C
+        x = x.permute(0, 2, 1)
+        n_layers = len(query)
+        for i, q in enumerate(query):
+            x = torch.bmm(x, q)
+            if i < n_layers - 1:
+                if self.att_fn == 'relu':
+                    x = F.relu(x)
+                elif self.att_fn == 'softmax':
+                    x = F.softmax(x, dim=1)
+        # N, C, HW
+        x = x.permute(0, 2, 1)
+        return x
+
+    def mask_heads_forward_with_coords(
+            self, mask_feats, mask_feat_stride, instances
+    ):
+        locations = compute_locations(
+            mask_feats.size(2), mask_feats.size(3),
+            stride=mask_feat_stride, device=mask_feats.device
+        ).repeat(1, self.rel_num)
+        n_inst = len(instances)
+
+        im_inds = instances.im_inds
+        mask_head_params = instances.mask_head_params
+
+        N, _, H, W = mask_feats.size()
+
+        if not self.disable_rel_coords:
+            instance_locations = instances.locations
+            relative_coords = instance_locations.reshape(-1, 1, self.rel_num*2) - locations.reshape(1, -1, self.rel_num*2)
+            relative_coords = relative_coords.permute(0, 2, 1).float()
+            if self.norm_rel == 'boxgt': # norm by gt when training
+                relative_coords = relative_coords / instances.boxsz.unsqueeze(2)
+            elif self.norm_rel == 'one': # max - min = one
+                relative_coords = relative_coords / torch.max(locations, dim=0)[0].reshape(1, -1, 1)
+            else:
+                soi = self.sizes_of_interest.float()[instances.fpn_levels]
+                relative_coords = relative_coords / soi.reshape(-1, 1, 1)
+            
+            relative_coords = relative_coords.to(dtype=mask_feats.dtype)
+            self.relative_coords = relative_coords.view(n_inst, self.rel_num*2, H, W)
+            if self.rel_abs:
+                relative_coords.abs_()
+
+            mask_head_inputs = torch.cat([
+                relative_coords, mask_feats[im_inds].reshape(n_inst, self.in_channels, H * W)
+            ], dim=1)
+        else:
+            mask_head_inputs = mask_feats[im_inds].reshape(n_inst, self.in_channels, H * W)
+
+
+        if self.use_att_map:
+            query = parse_attn_params(
+                mask_head_params, self.channels,
+                self.query_nums
+            )
+            mask_logits = self.attn_forward(mask_head_inputs, query, n_inst)
+            mask_logits_list = [mask_logits]
+        else:
+            mask_head_inputs = mask_head_inputs.reshape(1, -1, H, W)
+
+            if self.aux_loss:
+                weights, biases, aux_weights, aux_biases = parse_dynamic_params_with_aux(
+                    mask_head_params, self.channels,
+                    self.weight_nums, self.bias_nums,
+                    self.aux_weight_nums, self.aux_bias_nums
+                )
+                mask_logits_list = self.mask_heads_forward(mask_head_inputs, weights, biases, n_inst, aux_weights, aux_biases)
+            else:
+                weights, biases = parse_dynamic_params(
+                    mask_head_params, self.channels,
+                    self.weight_nums, self.bias_nums
+                )
+                mask_logits_list = self.mask_heads_forward(mask_head_inputs, weights, biases, n_inst)
+
+        out_list = []
+        for mask_logits in mask_logits_list:
+            mask_logits = mask_logits.reshape(-1, 1, H, W)
+
+            assert mask_feat_stride >= self.mask_out_stride
+            assert mask_feat_stride % self.mask_out_stride == 0
+            mask_logits = aligned_bilinear(mask_logits, int(mask_feat_stride / self.mask_out_stride))
+            out_list.append(mask_logits.sigmoid())
+
+        return out_list
+        
+    def __call__(self, mask_feats, mask_feat_stride, pred_instances, gt_masks=None):
+        if self.training:
+
+            if len(pred_instances) == 0:
+                loss_mask = dict()
+                for i in range(self.mask_level_numbers):
+                    loss_mask[f'loss_mask_{i}'] = mask_feats.sum() * 0 + pred_instances.mask_head_params.sum() * 0
+            else:
+                mask_scores = self.mask_heads_forward_with_coords(
+                    mask_feats, mask_feat_stride, pred_instances
+                )
+                self.mask_level_numbers = len(mask_scores)
+                gt_inds = pred_instances.gt_inds
+                gt_bitmasks = torch.cat([downsample_mask(gt_mask, self.mask_out_stride) for gt_mask in gt_masks])
+                gt_bitmasks = gt_bitmasks[gt_inds].to(dtype=mask_feats.dtype)
+                loss_mask = dict()
+                for i, mask_score in enumerate(mask_scores):
+                    mask_losses = dice_coefficient(mask_score, gt_bitmasks)
+                    loss_mask[f'loss_mask_{i}'] = mask_losses.mean() * self.mask_loss_weight[i]
+
+            return loss_mask
+        else:
+            if len(pred_instances) > 0:
+                mask_scores = self.mask_heads_forward_with_coords(
+                    mask_feats, mask_feat_stride, pred_instances
+                )[-1].float() # select the last output value
+            else:
+                mask_scores = None
+            return mask_scores
+
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/fcn_mask_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/fcn_mask_head.py
new file mode 100644
index 0000000..be6772f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/fcn_mask_head.py
@@ -0,0 +1,377 @@
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import Conv2d, ConvModule, build_upsample_layer
+from mmcv.ops.carafe import CARAFEPack
+from mmcv.runner import auto_fp16, force_fp32
+from torch.nn.modules.utils import _pair
+
+from mmdet.core import mask_target
+from mmdet.models.builder import HEADS, build_loss
+
+BYTES_PER_FLOAT = 4
+# TODO: This memory limit may be too much or too little. It would be better to
+# determine it based on available resources.
+GPU_MEM_LIMIT = 1024**3  # 1 GB memory limit
+
+
+@HEADS.register_module()
+class FCNMaskHead(nn.Module):
+
+    def __init__(self,
+                 num_convs=4,
+                 roi_feat_size=14,
+                 in_channels=256,
+                 conv_kernel_size=3,
+                 conv_out_channels=256,
+                 num_classes=80,
+                 class_agnostic=False,
+                 upsample_cfg=dict(type='deconv', scale_factor=2),
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 loss_mask=dict(
+                     type='CrossEntropyLoss', use_mask=True, loss_weight=1.0)):
+        super(FCNMaskHead, self).__init__()
+        self.upsample_cfg = upsample_cfg.copy()
+        if self.upsample_cfg['type'] not in [
+                None, 'deconv', 'nearest', 'bilinear', 'carafe'
+        ]:
+            raise ValueError(
+                f'Invalid upsample method {self.upsample_cfg["type"]}, '
+                'accepted methods are "deconv", "nearest", "bilinear", '
+                '"carafe"')
+        self.num_convs = num_convs
+        # WARN: roi_feat_size is reserved and not used
+        self.roi_feat_size = _pair(roi_feat_size)
+        self.in_channels = in_channels
+        self.conv_kernel_size = conv_kernel_size
+        self.conv_out_channels = conv_out_channels
+        self.upsample_method = self.upsample_cfg.get('type')
+        self.scale_factor = self.upsample_cfg.pop('scale_factor', None)
+        self.num_classes = num_classes
+        self.class_agnostic = class_agnostic
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.fp16_enabled = False
+        self.loss_mask = build_loss(loss_mask)
+
+        self.convs = nn.ModuleList()
+        for i in range(self.num_convs):
+            in_channels = (
+                self.in_channels if i == 0 else self.conv_out_channels)
+            padding = (self.conv_kernel_size - 1) // 2
+            self.convs.append(
+                ConvModule(
+                    in_channels,
+                    self.conv_out_channels,
+                    self.conv_kernel_size,
+                    padding=padding,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg))
+        upsample_in_channels = (
+            self.conv_out_channels if self.num_convs > 0 else in_channels)
+        upsample_cfg_ = self.upsample_cfg.copy()
+        if self.upsample_method is None:
+            self.upsample = None
+        elif self.upsample_method == 'deconv':
+            upsample_cfg_.update(
+                in_channels=upsample_in_channels,
+                out_channels=self.conv_out_channels,
+                kernel_size=self.scale_factor,
+                stride=self.scale_factor)
+            self.upsample = build_upsample_layer(upsample_cfg_)
+        elif self.upsample_method == 'carafe':
+            upsample_cfg_.update(
+                channels=upsample_in_channels, scale_factor=self.scale_factor)
+            self.upsample = build_upsample_layer(upsample_cfg_)
+        else:
+            # suppress warnings
+            align_corners = (None
+                             if self.upsample_method == 'nearest' else False)
+            upsample_cfg_.update(
+                scale_factor=self.scale_factor,
+                mode=self.upsample_method,
+                align_corners=align_corners)
+            self.upsample = build_upsample_layer(upsample_cfg_)
+
+        out_channels = 1 if self.class_agnostic else self.num_classes
+        logits_in_channel = (
+            self.conv_out_channels
+            if self.upsample_method == 'deconv' else upsample_in_channels)
+        self.conv_logits = Conv2d(logits_in_channel, out_channels, 1)
+        self.relu = nn.ReLU(inplace=True)
+        self.debug_imgs = None
+
+    def init_weights(self):
+        for m in [self.upsample, self.conv_logits]:
+            if m is None:
+                continue
+            elif isinstance(m, CARAFEPack):
+                m.init_weights()
+            else:
+                nn.init.kaiming_normal_(
+                    m.weight, mode='fan_out', nonlinearity='relu')
+                nn.init.constant_(m.bias, 0)
+
+    @auto_fp16()
+    def forward(self, x):
+        for conv in self.convs:
+            x = conv(x)
+        if self.upsample is not None:
+            x = self.upsample(x)
+            if self.upsample_method == 'deconv':
+                x = self.relu(x)
+        mask_pred = self.conv_logits(x)
+        return mask_pred
+
+    def get_targets(self, sampling_results, gt_masks, rcnn_train_cfg):
+        pos_proposals = [res.pos_bboxes for res in sampling_results]
+        pos_assigned_gt_inds = [
+            res.pos_assigned_gt_inds for res in sampling_results
+        ]
+        mask_targets = mask_target(pos_proposals, pos_assigned_gt_inds,
+                                   gt_masks, rcnn_train_cfg)
+        return mask_targets
+
+    @force_fp32(apply_to=('mask_pred', ))
+    def loss(self, mask_pred, mask_targets, labels):
+        """
+        Example:
+            >>> from mmdet.models.roi_heads.mask_heads.fcn_mask_head import *  # NOQA
+            >>> N = 7  # N = number of extracted ROIs
+            >>> C, H, W = 11, 32, 32
+            >>> # Create example instance of FCN Mask Head.
+            >>> # There are lots of variations depending on the configuration
+            >>> self = FCNMaskHead(num_classes=C, num_convs=1)
+            >>> inputs = torch.rand(N, self.in_channels, H, W)
+            >>> mask_pred = self.forward(inputs)
+            >>> sf = self.scale_factor
+            >>> labels = torch.randint(0, C, size=(N,))
+            >>> # With the default properties the mask targets should indicate
+            >>> # a (potentially soft) single-class label
+            >>> mask_targets = torch.rand(N, H * sf, W * sf)
+            >>> loss = self.loss(mask_pred, mask_targets, labels)
+            >>> print('loss = {!r}'.format(loss))
+        """
+        loss = dict()
+        if mask_pred.size(0) == 0:
+            loss_mask = mask_pred.sum()
+        else:
+            if self.class_agnostic:
+                loss_mask = self.loss_mask(mask_pred, mask_targets,
+                                           torch.zeros_like(labels))
+            else:
+                loss_mask = self.loss_mask(mask_pred, mask_targets, labels)
+        loss['loss_mask'] = loss_mask
+        return loss
+
+    def get_seg_masks(self, mask_pred, det_bboxes, det_labels, rcnn_test_cfg,
+                      ori_shape, scale_factor, rescale):
+        """Get segmentation masks from mask_pred and bboxes.
+
+        Args:
+            mask_pred (Tensor or ndarray): shape (n, #class, h, w).
+                For single-scale testing, mask_pred is the direct output of
+                model, whose type is Tensor, while for multi-scale testing,
+                it will be converted to numpy array outside of this method.
+            det_bboxes (Tensor): shape (n, 4/5)
+            det_labels (Tensor): shape (n, )
+            rcnn_test_cfg (dict): rcnn testing config
+            ori_shape (Tuple): original image height and width, shape (2,)
+            scale_factor(float | Tensor): If ``rescale is True``, box
+                coordinates are divided by this scale factor to fit
+                ``ori_shape``.
+            rescale (bool): If True, the resulting masks will be rescaled to
+                ``ori_shape``.
+
+        Returns:
+            list[list]: encoded masks. The c-th item in the outer list
+                corresponds to the c-th class. Given the c-th outer list, the
+                i-th item in that inner list is the mask for the i-th box with
+                class label c.
+
+        Example:
+            >>> import mmcv
+            >>> from mmdet.models.roi_heads.mask_heads.fcn_mask_head import *  # NOQA
+            >>> N = 7  # N = number of extracted ROIs
+            >>> C, H, W = 11, 32, 32
+            >>> # Create example instance of FCN Mask Head.
+            >>> self = FCNMaskHead(num_classes=C, num_convs=0)
+            >>> inputs = torch.rand(N, self.in_channels, H, W)
+            >>> mask_pred = self.forward(inputs)
+            >>> # Each input is associated with some bounding box
+            >>> det_bboxes = torch.Tensor([[1, 1, 42, 42 ]] * N)
+            >>> det_labels = torch.randint(0, C, size=(N,))
+            >>> rcnn_test_cfg = mmcv.Config({'mask_thr_binary': 0, })
+            >>> ori_shape = (H * 4, W * 4)
+            >>> scale_factor = torch.FloatTensor((1, 1))
+            >>> rescale = False
+            >>> # Encoded masks are a list for each category.
+            >>> encoded_masks = self.get_seg_masks(
+            >>>     mask_pred, det_bboxes, det_labels, rcnn_test_cfg, ori_shape,
+            >>>     scale_factor, rescale
+            >>> )
+            >>> assert len(encoded_masks) == C
+            >>> assert sum(list(map(len, encoded_masks))) == N
+        """
+        if isinstance(mask_pred, torch.Tensor):
+            mask_pred = mask_pred.sigmoid()
+        else:
+            mask_pred = det_bboxes.new_tensor(mask_pred)
+
+        device = mask_pred.device
+        cls_segms = [[] for _ in range(self.num_classes)
+                     ]  # BG is not included in num_classes
+        bboxes = det_bboxes[:, :4]
+        labels = det_labels
+
+        if rescale:
+            img_h, img_w = ori_shape[:2]
+        else:
+            if isinstance(scale_factor, float):
+                img_h = np.round(ori_shape[0] * scale_factor).astype(np.int32)
+                img_w = np.round(ori_shape[1] * scale_factor).astype(np.int32)
+            else:
+                w_scale, h_scale = scale_factor[0], scale_factor[1]
+                img_h = np.round(ori_shape[0] * h_scale.item()).astype(
+                    np.int32)
+                img_w = np.round(ori_shape[1] * w_scale.item()).astype(
+                    np.int32)
+            scale_factor = 1.0
+
+        if not isinstance(scale_factor, (float, torch.Tensor)):
+            scale_factor = bboxes.new_tensor(scale_factor)
+        bboxes = bboxes / scale_factor
+
+        if torch.onnx.is_in_onnx_export():
+            # TODO: Remove after F.grid_sample is supported.
+            from torchvision.models.detection.roi_heads \
+                import paste_masks_in_image
+            masks = paste_masks_in_image(mask_pred, bboxes, ori_shape[:2])
+            thr = rcnn_test_cfg.get('mask_thr_binary', 0)
+            if thr > 0:
+                masks = masks >= thr
+            return masks
+
+        N = len(mask_pred)
+        # The actual implementation split the input into chunks,
+        # and paste them chunk by chunk.
+        if device.type == 'cpu':
+            # CPU is most efficient when they are pasted one by one with
+            # skip_empty=True, so that it performs minimal number of
+            # operations.
+            num_chunks = N
+        else:
+            # GPU benefits from parallelism for larger chunks,
+            # but may have memory issue
+            num_chunks = int(
+                np.ceil(N * img_h * img_w * BYTES_PER_FLOAT / GPU_MEM_LIMIT))
+            assert (num_chunks <=
+                    N), 'Default GPU_MEM_LIMIT is too small; try increasing it'
+        chunks = torch.chunk(torch.arange(N, device=device), num_chunks)
+
+        threshold = rcnn_test_cfg.mask_thr_binary
+        im_mask = torch.zeros(
+            N,
+            img_h,
+            img_w,
+            device=device,
+            dtype=torch.bool if threshold >= 0 else torch.uint8)
+
+        if not self.class_agnostic:
+            mask_pred = mask_pred[range(N), labels][:, None]
+
+        for inds in chunks:
+            masks_chunk, spatial_inds = _do_paste_mask(
+                mask_pred[inds],
+                bboxes[inds],
+                img_h,
+                img_w,
+                skip_empty=device.type == 'cpu')
+
+            if threshold >= 0:
+                masks_chunk = (masks_chunk >= threshold).to(dtype=torch.bool)
+            else:
+                # for visualization and debugging
+                masks_chunk = (masks_chunk * 255).to(dtype=torch.uint8)
+
+            im_mask[(inds, ) + spatial_inds] = masks_chunk
+
+        for i in range(N):
+            cls_segms[labels[i]].append(im_mask[i].detach().cpu().numpy())
+        return cls_segms
+
+
+def _do_paste_mask(masks, boxes, img_h, img_w, skip_empty=True):
+    """Paste instance masks according to boxes.
+
+    This implementation is modified from
+    https://github.com/facebookresearch/detectron2/
+
+    Args:
+        masks (Tensor): N, 1, H, W
+        boxes (Tensor): N, 4
+        img_h (int): Height of the image to be pasted.
+        img_w (int): Width of the image to be pasted.
+        skip_empty (bool): Only paste masks within the region that
+            tightly bound all boxes, and returns the results this region only.
+            An important optimization for CPU.
+
+    Returns:
+        tuple: (Tensor, tuple). The first item is mask tensor, the second one
+            is the slice object.
+        If skip_empty == False, the whole image will be pasted. It will
+            return a mask of shape (N, img_h, img_w) and an empty tuple.
+        If skip_empty == True, only area around the mask will be pasted.
+            A mask of shape (N, h', w') and its start and end coordinates
+            in the original image will be returned.
+    """
+    # On GPU, paste all masks together (up to chunk size)
+    # by using the entire image to sample the masks
+    # Compared to pasting them one by one,
+    # this has more operations but is faster on COCO-scale dataset.
+    device = masks.device
+    if skip_empty:
+        x0_int, y0_int = torch.clamp(
+            boxes.min(dim=0).values.floor()[:2] - 1,
+            min=0).to(dtype=torch.int32)
+        x1_int = torch.clamp(
+            boxes[:, 2].max().ceil() + 1, max=img_w).to(dtype=torch.int32)
+        y1_int = torch.clamp(
+            boxes[:, 3].max().ceil() + 1, max=img_h).to(dtype=torch.int32)
+    else:
+        x0_int, y0_int = 0, 0
+        x1_int, y1_int = img_w, img_h
+    x0, y0, x1, y1 = torch.split(boxes, 1, dim=1)  # each is Nx1
+
+    N = masks.shape[0]
+
+    img_y = torch.arange(
+        y0_int, y1_int, device=device, dtype=torch.float32) + 0.5
+    img_x = torch.arange(
+        x0_int, x1_int, device=device, dtype=torch.float32) + 0.5
+    img_y = (img_y - y0) / (y1 - y0) * 2 - 1
+    img_x = (img_x - x0) / (x1 - x0) * 2 - 1
+    # img_x, img_y have shapes (N, w), (N, h)
+    if torch.isinf(img_x).any():
+        inds = torch.where(torch.isinf(img_x))
+        img_x[inds] = 0
+    if torch.isinf(img_y).any():
+        inds = torch.where(torch.isinf(img_y))
+        img_y[inds] = 0
+
+    gx = img_x[:, None, :].expand(N, img_y.size(1), img_x.size(1))
+    gy = img_y[:, :, None].expand(N, img_y.size(1), img_x.size(1))
+    grid = torch.stack([gx, gy], dim=3)
+
+    if torch.onnx.is_in_onnx_export():
+        raise RuntimeError(
+            'Exporting F.grid_sample from Pytorch to ONNX is not supported.')
+    img_masks = F.grid_sample(
+        masks.to(dtype=torch.float32), grid, align_corners=False)
+
+    if skip_empty:
+        return img_masks[:, 0], (slice(y0_int, y1_int), slice(x0_int, x1_int))
+    else:
+        return img_masks[:, 0], ()
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/feature_relay_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/feature_relay_head.py
new file mode 100644
index 0000000..a1cfb2c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/feature_relay_head.py
@@ -0,0 +1,55 @@
+import torch.nn as nn
+from mmcv.cnn import kaiming_init
+from mmcv.runner import auto_fp16
+
+from mmdet.models.builder import HEADS
+
+
+@HEADS.register_module()
+class FeatureRelayHead(nn.Module):
+    """Feature Relay Head used in `SCNet <https://arxiv.org/abs/2012.10150>`_.
+
+    Args:
+        in_channels (int, optional): number of input channels. Default: 256.
+        conv_out_channels (int, optional): number of output channels before
+            classification layer. Default: 256.
+        roi_feat_size (int, optional): roi feat size at box head. Default: 7.
+        scale_factor (int, optional): scale factor to match roi feat size
+            at mask head. Default: 2.
+    """
+
+    def __init__(self,
+                 in_channels=1024,
+                 out_conv_channels=256,
+                 roi_feat_size=7,
+                 scale_factor=2):
+        super(FeatureRelayHead, self).__init__()
+        assert isinstance(roi_feat_size, int)
+
+        self.in_channels = in_channels
+        self.out_conv_channels = out_conv_channels
+        self.roi_feat_size = roi_feat_size
+        self.out_channels = (roi_feat_size**2) * out_conv_channels
+        self.scale_factor = scale_factor
+        self.fp16_enabled = False
+
+        self.fc = nn.Linear(self.in_channels, self.out_channels)
+        self.upsample = nn.Upsample(
+            scale_factor=scale_factor, mode='bilinear', align_corners=True)
+
+    def init_weights(self):
+        """Init weights for the head."""
+        kaiming_init(self.fc)
+
+    @auto_fp16()
+    def forward(self, x):
+        """Forward function."""
+        N, in_C = x.shape
+        if N > 0:
+            out_C = self.out_conv_channels
+            out_HW = self.roi_feat_size
+            x = self.fc(x)
+            x = x.reshape(N, out_C, out_HW, out_HW)
+            x = self.upsample(x)
+            return x
+        return None
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/fused_semantic_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/fused_semantic_head.py
new file mode 100644
index 0000000..2aa6033
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/fused_semantic_head.py
@@ -0,0 +1,107 @@
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, kaiming_init
+from mmcv.runner import auto_fp16, force_fp32
+
+from mmdet.models.builder import HEADS
+
+
+@HEADS.register_module()
+class FusedSemanticHead(nn.Module):
+    r"""Multi-level fused semantic segmentation head.
+
+    .. code-block:: none
+
+        in_1 -> 1x1 conv ---
+                            |
+        in_2 -> 1x1 conv -- |
+                           ||
+        in_3 -> 1x1 conv - ||
+                          |||                  /-> 1x1 conv (mask prediction)
+        in_4 -> 1x1 conv -----> 3x3 convs (*4)
+                            |                  \-> 1x1 conv (feature)
+        in_5 -> 1x1 conv ---
+    """  # noqa: W605
+
+    def __init__(self,
+                 num_ins,
+                 fusion_level,
+                 num_convs=4,
+                 in_channels=256,
+                 conv_out_channels=256,
+                 num_classes=183,
+                 ignore_label=255,
+                 loss_weight=0.2,
+                 conv_cfg=None,
+                 norm_cfg=None):
+        super(FusedSemanticHead, self).__init__()
+        self.num_ins = num_ins
+        self.fusion_level = fusion_level
+        self.num_convs = num_convs
+        self.in_channels = in_channels
+        self.conv_out_channels = conv_out_channels
+        self.num_classes = num_classes
+        self.ignore_label = ignore_label
+        self.loss_weight = loss_weight
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.fp16_enabled = False
+
+        self.lateral_convs = nn.ModuleList()
+        for i in range(self.num_ins):
+            self.lateral_convs.append(
+                ConvModule(
+                    self.in_channels,
+                    self.in_channels,
+                    1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg,
+                    inplace=False))
+
+        self.convs = nn.ModuleList()
+        for i in range(self.num_convs):
+            in_channels = self.in_channels if i == 0 else conv_out_channels
+            self.convs.append(
+                ConvModule(
+                    in_channels,
+                    conv_out_channels,
+                    3,
+                    padding=1,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg))
+        self.conv_embedding = ConvModule(
+            conv_out_channels,
+            conv_out_channels,
+            1,
+            conv_cfg=self.conv_cfg,
+            norm_cfg=self.norm_cfg)
+        self.conv_logits = nn.Conv2d(conv_out_channels, self.num_classes, 1)
+
+        self.criterion = nn.CrossEntropyLoss(ignore_index=ignore_label)
+
+    def init_weights(self):
+        kaiming_init(self.conv_logits)
+
+    @auto_fp16()
+    def forward(self, feats):
+        x = self.lateral_convs[self.fusion_level](feats[self.fusion_level])
+        fused_size = tuple(x.shape[-2:])
+        for i, feat in enumerate(feats):
+            if i != self.fusion_level:
+                feat = F.interpolate(
+                    feat, size=fused_size, mode='bilinear', align_corners=True)
+                x += self.lateral_convs[i](feat)
+
+        for i in range(self.num_convs):
+            x = self.convs[i](x)
+
+        mask_pred = self.conv_logits(x)
+        x = self.conv_embedding(x)
+        return mask_pred, x
+
+    @force_fp32(apply_to=('mask_pred', ))
+    def loss(self, mask_pred, labels):
+        labels = labels.squeeze(1).long()
+        loss_semantic_seg = self.criterion(mask_pred, labels)
+        loss_semantic_seg *= self.loss_weight
+        return loss_semantic_seg
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/global_context_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/global_context_head.py
new file mode 100644
index 0000000..d8e8cbc
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/global_context_head.py
@@ -0,0 +1,102 @@
+import torch.nn as nn
+from mmcv.cnn import ConvModule
+from mmcv.runner import auto_fp16, force_fp32
+
+from mmdet.models.builder import HEADS
+from mmdet.models.utils import ResLayer, SimplifiedBasicBlock
+
+
+@HEADS.register_module()
+class GlobalContextHead(nn.Module):
+    """Global context head used in `SCNet <https://arxiv.org/abs/2012.10150>`_.
+
+    Args:
+        num_convs (int, optional): number of convolutional layer in GlbCtxHead.
+            Default: 4.
+        in_channels (int, optional): number of input channels. Default: 256.
+        conv_out_channels (int, optional): number of output channels before
+            classification layer. Default: 256.
+        num_classes (int, optional): number of classes. Default: 80.
+        loss_weight (float, optional): global context loss weight. Default: 1.
+        conv_cfg (dict, optional): config to init conv layer. Default: None.
+        norm_cfg (dict, optional): config to init norm layer. Default: None.
+        conv_to_res (bool, optional): if True, 2 convs will be grouped into
+            1 `SimplifiedBasicBlock` using a skip connection. Default: False.
+    """
+
+    def __init__(self,
+                 num_convs=4,
+                 in_channels=256,
+                 conv_out_channels=256,
+                 num_classes=80,
+                 loss_weight=1.0,
+                 conv_cfg=None,
+                 norm_cfg=None,
+                 conv_to_res=False):
+        super(GlobalContextHead, self).__init__()
+        self.num_convs = num_convs
+        self.in_channels = in_channels
+        self.conv_out_channels = conv_out_channels
+        self.num_classes = num_classes
+        self.loss_weight = loss_weight
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.conv_to_res = conv_to_res
+        self.fp16_enabled = False
+
+        if self.conv_to_res:
+            num_res_blocks = num_convs // 2
+            self.convs = ResLayer(
+                SimplifiedBasicBlock,
+                in_channels,
+                self.conv_out_channels,
+                num_res_blocks,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg)
+            self.num_convs = num_res_blocks
+        else:
+            self.convs = nn.ModuleList()
+            for i in range(self.num_convs):
+                in_channels = self.in_channels if i == 0 else conv_out_channels
+                self.convs.append(
+                    ConvModule(
+                        in_channels,
+                        conv_out_channels,
+                        3,
+                        padding=1,
+                        conv_cfg=self.conv_cfg,
+                        norm_cfg=self.norm_cfg))
+
+        self.pool = nn.AdaptiveAvgPool2d(1)
+        self.fc = nn.Linear(conv_out_channels, num_classes)
+
+        self.criterion = nn.BCEWithLogitsLoss()
+
+    def init_weights(self):
+        """Init weights for the head."""
+        nn.init.normal_(self.fc.weight, 0, 0.01)
+        nn.init.constant_(self.fc.bias, 0)
+
+    @auto_fp16()
+    def forward(self, feats):
+        """Forward function."""
+        x = feats[-1]
+        for i in range(self.num_convs):
+            x = self.convs[i](x)
+        x = self.pool(x)
+
+        # multi-class prediction
+        mc_pred = x.reshape(x.size(0), -1)
+        mc_pred = self.fc(mc_pred)
+
+        return mc_pred, x
+
+    @force_fp32(apply_to=('pred', ))
+    def loss(self, pred, labels):
+        """Loss function."""
+        labels = [lbl.unique() for lbl in labels]
+        targets = pred.new_zeros(pred.size())
+        for i, label in enumerate(labels):
+            targets[i, label] = 1.0
+        loss = self.loss_weight * self.criterion(pred, targets)
+        return loss
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/grid_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/grid_head.py
new file mode 100644
index 0000000..83058cb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/grid_head.py
@@ -0,0 +1,359 @@
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from mmcv.cnn import ConvModule, kaiming_init, normal_init
+
+from mmdet.models.builder import HEADS, build_loss
+
+
+@HEADS.register_module()
+class GridHead(nn.Module):
+
+    def __init__(self,
+                 grid_points=9,
+                 num_convs=8,
+                 roi_feat_size=14,
+                 in_channels=256,
+                 conv_kernel_size=3,
+                 point_feat_channels=64,
+                 deconv_kernel_size=4,
+                 class_agnostic=False,
+                 loss_grid=dict(
+                     type='CrossEntropyLoss', use_sigmoid=True,
+                     loss_weight=15),
+                 conv_cfg=None,
+                 norm_cfg=dict(type='GN', num_groups=36)):
+        super(GridHead, self).__init__()
+        self.grid_points = grid_points
+        self.num_convs = num_convs
+        self.roi_feat_size = roi_feat_size
+        self.in_channels = in_channels
+        self.conv_kernel_size = conv_kernel_size
+        self.point_feat_channels = point_feat_channels
+        self.conv_out_channels = self.point_feat_channels * self.grid_points
+        self.class_agnostic = class_agnostic
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        if isinstance(norm_cfg, dict) and norm_cfg['type'] == 'GN':
+            assert self.conv_out_channels % norm_cfg['num_groups'] == 0
+
+        assert self.grid_points >= 4
+        self.grid_size = int(np.sqrt(self.grid_points))
+        if self.grid_size * self.grid_size != self.grid_points:
+            raise ValueError('grid_points must be a square number')
+
+        # the predicted heatmap is half of whole_map_size
+        if not isinstance(self.roi_feat_size, int):
+            raise ValueError('Only square RoIs are supporeted in Grid R-CNN')
+        self.whole_map_size = self.roi_feat_size * 4
+
+        # compute point-wise sub-regions
+        self.sub_regions = self.calc_sub_regions()
+
+        self.convs = []
+        for i in range(self.num_convs):
+            in_channels = (
+                self.in_channels if i == 0 else self.conv_out_channels)
+            stride = 2 if i == 0 else 1
+            padding = (self.conv_kernel_size - 1) // 2
+            self.convs.append(
+                ConvModule(
+                    in_channels,
+                    self.conv_out_channels,
+                    self.conv_kernel_size,
+                    stride=stride,
+                    padding=padding,
+                    conv_cfg=self.conv_cfg,
+                    norm_cfg=self.norm_cfg,
+                    bias=True))
+        self.convs = nn.Sequential(*self.convs)
+
+        self.deconv1 = nn.ConvTranspose2d(
+            self.conv_out_channels,
+            self.conv_out_channels,
+            kernel_size=deconv_kernel_size,
+            stride=2,
+            padding=(deconv_kernel_size - 2) // 2,
+            groups=grid_points)
+        self.norm1 = nn.GroupNorm(grid_points, self.conv_out_channels)
+        self.deconv2 = nn.ConvTranspose2d(
+            self.conv_out_channels,
+            grid_points,
+            kernel_size=deconv_kernel_size,
+            stride=2,
+            padding=(deconv_kernel_size - 2) // 2,
+            groups=grid_points)
+
+        # find the 4-neighbor of each grid point
+        self.neighbor_points = []
+        grid_size = self.grid_size
+        for i in range(grid_size):  # i-th column
+            for j in range(grid_size):  # j-th row
+                neighbors = []
+                if i > 0:  # left: (i - 1, j)
+                    neighbors.append((i - 1) * grid_size + j)
+                if j > 0:  # up: (i, j - 1)
+                    neighbors.append(i * grid_size + j - 1)
+                if j < grid_size - 1:  # down: (i, j + 1)
+                    neighbors.append(i * grid_size + j + 1)
+                if i < grid_size - 1:  # right: (i + 1, j)
+                    neighbors.append((i + 1) * grid_size + j)
+                self.neighbor_points.append(tuple(neighbors))
+        # total edges in the grid
+        self.num_edges = sum([len(p) for p in self.neighbor_points])
+
+        self.forder_trans = nn.ModuleList()  # first-order feature transition
+        self.sorder_trans = nn.ModuleList()  # second-order feature transition
+        for neighbors in self.neighbor_points:
+            fo_trans = nn.ModuleList()
+            so_trans = nn.ModuleList()
+            for _ in range(len(neighbors)):
+                # each transition module consists of a 5x5 depth-wise conv and
+                # 1x1 conv.
+                fo_trans.append(
+                    nn.Sequential(
+                        nn.Conv2d(
+                            self.point_feat_channels,
+                            self.point_feat_channels,
+                            5,
+                            stride=1,
+                            padding=2,
+                            groups=self.point_feat_channels),
+                        nn.Conv2d(self.point_feat_channels,
+                                  self.point_feat_channels, 1)))
+                so_trans.append(
+                    nn.Sequential(
+                        nn.Conv2d(
+                            self.point_feat_channels,
+                            self.point_feat_channels,
+                            5,
+                            1,
+                            2,
+                            groups=self.point_feat_channels),
+                        nn.Conv2d(self.point_feat_channels,
+                                  self.point_feat_channels, 1)))
+            self.forder_trans.append(fo_trans)
+            self.sorder_trans.append(so_trans)
+
+        self.loss_grid = build_loss(loss_grid)
+
+    def init_weights(self):
+        for m in self.modules():
+            if isinstance(m, nn.Conv2d) or isinstance(m, nn.Linear):
+                # TODO: compare mode = "fan_in" or "fan_out"
+                kaiming_init(m)
+        for m in self.modules():
+            if isinstance(m, nn.ConvTranspose2d):
+                normal_init(m, std=0.001)
+        nn.init.constant_(self.deconv2.bias, -np.log(0.99 / 0.01))
+
+    def forward(self, x):
+        assert x.shape[-1] == x.shape[-2] == self.roi_feat_size
+        # RoI feature transformation, downsample 2x
+        x = self.convs(x)
+
+        c = self.point_feat_channels
+        # first-order fusion
+        x_fo = [None for _ in range(self.grid_points)]
+        for i, points in enumerate(self.neighbor_points):
+            x_fo[i] = x[:, i * c:(i + 1) * c]
+            for j, point_idx in enumerate(points):
+                x_fo[i] = x_fo[i] + self.forder_trans[i][j](
+                    x[:, point_idx * c:(point_idx + 1) * c])
+
+        # second-order fusion
+        x_so = [None for _ in range(self.grid_points)]
+        for i, points in enumerate(self.neighbor_points):
+            x_so[i] = x[:, i * c:(i + 1) * c]
+            for j, point_idx in enumerate(points):
+                x_so[i] = x_so[i] + self.sorder_trans[i][j](x_fo[point_idx])
+
+        # predicted heatmap with fused features
+        x2 = torch.cat(x_so, dim=1)
+        x2 = self.deconv1(x2)
+        x2 = F.relu(self.norm1(x2), inplace=True)
+        heatmap = self.deconv2(x2)
+
+        # predicted heatmap with original features (applicable during training)
+        if self.training:
+            x1 = x
+            x1 = self.deconv1(x1)
+            x1 = F.relu(self.norm1(x1), inplace=True)
+            heatmap_unfused = self.deconv2(x1)
+        else:
+            heatmap_unfused = heatmap
+
+        return dict(fused=heatmap, unfused=heatmap_unfused)
+
+    def calc_sub_regions(self):
+        """Compute point specific representation regions.
+
+        See Grid R-CNN Plus (https://arxiv.org/abs/1906.05688) for details.
+        """
+        # to make it consistent with the original implementation, half_size
+        # is computed as 2 * quarter_size, which is smaller
+        half_size = self.whole_map_size // 4 * 2
+        sub_regions = []
+        for i in range(self.grid_points):
+            x_idx = i // self.grid_size
+            y_idx = i % self.grid_size
+            if x_idx == 0:
+                sub_x1 = 0
+            elif x_idx == self.grid_size - 1:
+                sub_x1 = half_size
+            else:
+                ratio = x_idx / (self.grid_size - 1) - 0.25
+                sub_x1 = max(int(ratio * self.whole_map_size), 0)
+
+            if y_idx == 0:
+                sub_y1 = 0
+            elif y_idx == self.grid_size - 1:
+                sub_y1 = half_size
+            else:
+                ratio = y_idx / (self.grid_size - 1) - 0.25
+                sub_y1 = max(int(ratio * self.whole_map_size), 0)
+            sub_regions.append(
+                (sub_x1, sub_y1, sub_x1 + half_size, sub_y1 + half_size))
+        return sub_regions
+
+    def get_targets(self, sampling_results, rcnn_train_cfg):
+        # mix all samples (across images) together.
+        pos_bboxes = torch.cat([res.pos_bboxes for res in sampling_results],
+                               dim=0).cpu()
+        pos_gt_bboxes = torch.cat(
+            [res.pos_gt_bboxes for res in sampling_results], dim=0).cpu()
+        assert pos_bboxes.shape == pos_gt_bboxes.shape
+
+        # expand pos_bboxes to 2x of original size
+        x1 = pos_bboxes[:, 0] - (pos_bboxes[:, 2] - pos_bboxes[:, 0]) / 2
+        y1 = pos_bboxes[:, 1] - (pos_bboxes[:, 3] - pos_bboxes[:, 1]) / 2
+        x2 = pos_bboxes[:, 2] + (pos_bboxes[:, 2] - pos_bboxes[:, 0]) / 2
+        y2 = pos_bboxes[:, 3] + (pos_bboxes[:, 3] - pos_bboxes[:, 1]) / 2
+        pos_bboxes = torch.stack([x1, y1, x2, y2], dim=-1)
+        pos_bbox_ws = (pos_bboxes[:, 2] - pos_bboxes[:, 0]).unsqueeze(-1)
+        pos_bbox_hs = (pos_bboxes[:, 3] - pos_bboxes[:, 1]).unsqueeze(-1)
+
+        num_rois = pos_bboxes.shape[0]
+        map_size = self.whole_map_size
+        # this is not the final target shape
+        targets = torch.zeros((num_rois, self.grid_points, map_size, map_size),
+                              dtype=torch.float)
+
+        # pre-compute interpolation factors for all grid points.
+        # the first item is the factor of x-dim, and the second is y-dim.
+        # for a 9-point grid, factors are like (1, 0), (0.5, 0.5), (0, 1)
+        factors = []
+        for j in range(self.grid_points):
+            x_idx = j // self.grid_size
+            y_idx = j % self.grid_size
+            factors.append((1 - x_idx / (self.grid_size - 1),
+                            1 - y_idx / (self.grid_size - 1)))
+
+        radius = rcnn_train_cfg.pos_radius
+        radius2 = radius**2
+        for i in range(num_rois):
+            # ignore small bboxes
+            if (pos_bbox_ws[i] <= self.grid_size
+                    or pos_bbox_hs[i] <= self.grid_size):
+                continue
+            # for each grid point, mark a small circle as positive
+            for j in range(self.grid_points):
+                factor_x, factor_y = factors[j]
+                gridpoint_x = factor_x * pos_gt_bboxes[i, 0] + (
+                    1 - factor_x) * pos_gt_bboxes[i, 2]
+                gridpoint_y = factor_y * pos_gt_bboxes[i, 1] + (
+                    1 - factor_y) * pos_gt_bboxes[i, 3]
+
+                cx = int((gridpoint_x - pos_bboxes[i, 0]) / pos_bbox_ws[i] *
+                         map_size)
+                cy = int((gridpoint_y - pos_bboxes[i, 1]) / pos_bbox_hs[i] *
+                         map_size)
+
+                for x in range(cx - radius, cx + radius + 1):
+                    for y in range(cy - radius, cy + radius + 1):
+                        if x >= 0 and x < map_size and y >= 0 and y < map_size:
+                            if (x - cx)**2 + (y - cy)**2 <= radius2:
+                                targets[i, j, y, x] = 1
+        # reduce the target heatmap size by a half
+        # proposed in Grid R-CNN Plus (https://arxiv.org/abs/1906.05688).
+        sub_targets = []
+        for i in range(self.grid_points):
+            sub_x1, sub_y1, sub_x2, sub_y2 = self.sub_regions[i]
+            sub_targets.append(targets[:, [i], sub_y1:sub_y2, sub_x1:sub_x2])
+        sub_targets = torch.cat(sub_targets, dim=1)
+        sub_targets = sub_targets.to(sampling_results[0].pos_bboxes.device)
+        return sub_targets
+
+    def loss(self, grid_pred, grid_targets):
+        loss_fused = self.loss_grid(grid_pred['fused'], grid_targets)
+        loss_unfused = self.loss_grid(grid_pred['unfused'], grid_targets)
+        loss_grid = loss_fused + loss_unfused
+        return dict(loss_grid=loss_grid)
+
+    def get_bboxes(self, det_bboxes, grid_pred, img_metas):
+        # TODO: refactoring
+        assert det_bboxes.shape[0] == grid_pred.shape[0]
+        det_bboxes = det_bboxes.cpu()
+        cls_scores = det_bboxes[:, [4]]
+        det_bboxes = det_bboxes[:, :4]
+        grid_pred = grid_pred.sigmoid().cpu()
+
+        R, c, h, w = grid_pred.shape
+        half_size = self.whole_map_size // 4 * 2
+        assert h == w == half_size
+        assert c == self.grid_points
+
+        # find the point with max scores in the half-sized heatmap
+        grid_pred = grid_pred.view(R * c, h * w)
+        pred_scores, pred_position = grid_pred.max(dim=1)
+        xs = pred_position % w
+        ys = pred_position // w
+
+        # get the position in the whole heatmap instead of half-sized heatmap
+        for i in range(self.grid_points):
+            xs[i::self.grid_points] += self.sub_regions[i][0]
+            ys[i::self.grid_points] += self.sub_regions[i][1]
+
+        # reshape to (num_rois, grid_points)
+        pred_scores, xs, ys = tuple(
+            map(lambda x: x.view(R, c), [pred_scores, xs, ys]))
+
+        # get expanded pos_bboxes
+        widths = (det_bboxes[:, 2] - det_bboxes[:, 0]).unsqueeze(-1)
+        heights = (det_bboxes[:, 3] - det_bboxes[:, 1]).unsqueeze(-1)
+        x1 = (det_bboxes[:, 0, None] - widths / 2)
+        y1 = (det_bboxes[:, 1, None] - heights / 2)
+        # map the grid point to the absolute coordinates
+        abs_xs = (xs.float() + 0.5) / w * widths + x1
+        abs_ys = (ys.float() + 0.5) / h * heights + y1
+
+        # get the grid points indices that fall on the bbox boundaries
+        x1_inds = [i for i in range(self.grid_size)]
+        y1_inds = [i * self.grid_size for i in range(self.grid_size)]
+        x2_inds = [
+            self.grid_points - self.grid_size + i
+            for i in range(self.grid_size)
+        ]
+        y2_inds = [(i + 1) * self.grid_size - 1 for i in range(self.grid_size)]
+
+        # voting of all grid points on some boundary
+        bboxes_x1 = (abs_xs[:, x1_inds] * pred_scores[:, x1_inds]).sum(
+            dim=1, keepdim=True) / (
+                pred_scores[:, x1_inds].sum(dim=1, keepdim=True))
+        bboxes_y1 = (abs_ys[:, y1_inds] * pred_scores[:, y1_inds]).sum(
+            dim=1, keepdim=True) / (
+                pred_scores[:, y1_inds].sum(dim=1, keepdim=True))
+        bboxes_x2 = (abs_xs[:, x2_inds] * pred_scores[:, x2_inds]).sum(
+            dim=1, keepdim=True) / (
+                pred_scores[:, x2_inds].sum(dim=1, keepdim=True))
+        bboxes_y2 = (abs_ys[:, y2_inds] * pred_scores[:, y2_inds]).sum(
+            dim=1, keepdim=True) / (
+                pred_scores[:, y2_inds].sum(dim=1, keepdim=True))
+
+        bbox_res = torch.cat(
+            [bboxes_x1, bboxes_y1, bboxes_x2, bboxes_y2, cls_scores], dim=1)
+        bbox_res[:, [0, 2]].clamp_(min=0, max=img_metas[0]['img_shape'][1])
+        bbox_res[:, [1, 3]].clamp_(min=0, max=img_metas[0]['img_shape'][0])
+
+        return bbox_res
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/htc_mask_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/htc_mask_head.py
new file mode 100644
index 0000000..330b778
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/htc_mask_head.py
@@ -0,0 +1,43 @@
+from mmcv.cnn import ConvModule
+
+from mmdet.models.builder import HEADS
+from .fcn_mask_head import FCNMaskHead
+
+
+@HEADS.register_module()
+class HTCMaskHead(FCNMaskHead):
+
+    def __init__(self, with_conv_res=True, *args, **kwargs):
+        super(HTCMaskHead, self).__init__(*args, **kwargs)
+        self.with_conv_res = with_conv_res
+        if self.with_conv_res:
+            self.conv_res = ConvModule(
+                self.conv_out_channels,
+                self.conv_out_channels,
+                1,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg)
+
+    def init_weights(self):
+        super(HTCMaskHead, self).init_weights()
+        if self.with_conv_res:
+            self.conv_res.init_weights()
+
+    def forward(self, x, res_feat=None, return_logits=True, return_feat=True):
+        if res_feat is not None:
+            assert self.with_conv_res
+            res_feat = self.conv_res(res_feat)
+            x = x + res_feat
+        for conv in self.convs:
+            x = conv(x)
+        res_feat = x
+        outs = []
+        if return_logits:
+            x = self.upsample(x)
+            if self.upsample_method == 'deconv':
+                x = self.relu(x)
+            mask_pred = self.conv_logits(x)
+            outs.append(mask_pred)
+        if return_feat:
+            outs.append(res_feat)
+        return outs if len(outs) > 1 else outs[0]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/mask_point_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/mask_point_head.py
new file mode 100644
index 0000000..fb92903
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/mask_point_head.py
@@ -0,0 +1,300 @@
+# Modified from https://github.com/facebookresearch/detectron2/tree/master/projects/PointRend/point_head/point_head.py  # noqa
+
+import torch
+import torch.nn as nn
+from mmcv.cnn import ConvModule, normal_init
+from mmcv.ops import point_sample, rel_roi_point_to_rel_img_point
+
+from mmdet.models.builder import HEADS, build_loss
+
+
+@HEADS.register_module()
+class MaskPointHead(nn.Module):
+    """A mask point head use in PointRend.
+
+    ``MaskPointHead`` use shared multi-layer perceptron (equivalent to
+    nn.Conv1d) to predict the logit of input points. The fine-grained feature
+    and coarse feature will be concatenate together for predication.
+
+    Args:
+        num_fcs (int): Number of fc layers in the head. Default: 3.
+        in_channels (int): Number of input channels. Default: 256.
+        fc_channels (int): Number of fc channels. Default: 256.
+        num_classes (int): Number of classes for logits. Default: 80.
+        class_agnostic (bool): Whether use class agnostic classification.
+            If so, the output channels of logits will be 1. Default: False.
+        coarse_pred_each_layer (bool): Whether concatenate coarse feature with
+            the output of each fc layer. Default: True.
+        conv_cfg (dict | None): Dictionary to construct and config conv layer.
+            Default: dict(type='Conv1d'))
+        norm_cfg (dict | None): Dictionary to construct and config norm layer.
+            Default: None.
+        loss_point (dict): Dictionary to construct and config loss layer of
+            point head. Default: dict(type='CrossEntropyLoss', use_mask=True,
+            loss_weight=1.0).
+    """
+
+    def __init__(self,
+                 num_classes,
+                 num_fcs=3,
+                 in_channels=256,
+                 fc_channels=256,
+                 class_agnostic=False,
+                 coarse_pred_each_layer=True,
+                 conv_cfg=dict(type='Conv1d'),
+                 norm_cfg=None,
+                 act_cfg=dict(type='ReLU'),
+                 loss_point=dict(
+                     type='CrossEntropyLoss', use_mask=True, loss_weight=1.0)):
+        super().__init__()
+        self.num_fcs = num_fcs
+        self.in_channels = in_channels
+        self.fc_channels = fc_channels
+        self.num_classes = num_classes
+        self.class_agnostic = class_agnostic
+        self.coarse_pred_each_layer = coarse_pred_each_layer
+        self.conv_cfg = conv_cfg
+        self.norm_cfg = norm_cfg
+        self.loss_point = build_loss(loss_point)
+
+        fc_in_channels = in_channels + num_classes
+        self.fcs = nn.ModuleList()
+        for _ in range(num_fcs):
+            fc = ConvModule(
+                fc_in_channels,
+                fc_channels,
+                kernel_size=1,
+                stride=1,
+                padding=0,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                act_cfg=act_cfg)
+            self.fcs.append(fc)
+            fc_in_channels = fc_channels
+            fc_in_channels += num_classes if self.coarse_pred_each_layer else 0
+
+        out_channels = 1 if self.class_agnostic else self.num_classes
+        self.fc_logits = nn.Conv1d(
+            fc_in_channels, out_channels, kernel_size=1, stride=1, padding=0)
+
+    def init_weights(self):
+        """Initialize last classification layer of MaskPointHead, conv layers
+        are already initialized by ConvModule."""
+        normal_init(self.fc_logits, std=0.001)
+
+    def forward(self, fine_grained_feats, coarse_feats):
+        """Classify each point base on fine grained and coarse feats.
+
+        Args:
+            fine_grained_feats (Tensor): Fine grained feature sampled from FPN,
+                shape (num_rois, in_channels, num_points).
+            coarse_feats (Tensor): Coarse feature sampled from CoarseMaskHead,
+                shape (num_rois, num_classes, num_points).
+
+        Returns:
+            Tensor: Point classification results,
+                shape (num_rois, num_class, num_points).
+        """
+
+        x = torch.cat([fine_grained_feats, coarse_feats], dim=1)
+        for fc in self.fcs:
+            x = fc(x)
+            if self.coarse_pred_each_layer:
+                x = torch.cat((x, coarse_feats), dim=1)
+        return self.fc_logits(x)
+
+    def get_targets(self, rois, rel_roi_points, sampling_results, gt_masks,
+                    cfg):
+        """Get training targets of MaskPointHead for all images.
+
+        Args:
+            rois (Tensor): Region of Interest, shape (num_rois, 5).
+            rel_roi_points: Points coordinates relative to RoI, shape
+                (num_rois, num_points, 2).
+            sampling_results (:obj:`SamplingResult`): Sampling result after
+                sampling and assignment.
+            gt_masks (Tensor) : Ground truth segmentation masks of
+                corresponding boxes, shape (num_rois, height, width).
+            cfg (dict): Training cfg.
+
+        Returns:
+            Tensor: Point target, shape (num_rois, num_points).
+        """
+
+        num_imgs = len(sampling_results)
+        rois_list = []
+        rel_roi_points_list = []
+        for batch_ind in range(num_imgs):
+            inds = (rois[:, 0] == batch_ind)
+            rois_list.append(rois[inds])
+            rel_roi_points_list.append(rel_roi_points[inds])
+        pos_assigned_gt_inds_list = [
+            res.pos_assigned_gt_inds for res in sampling_results
+        ]
+        cfg_list = [cfg for _ in range(num_imgs)]
+
+        point_targets = map(self._get_target_single, rois_list,
+                            rel_roi_points_list, pos_assigned_gt_inds_list,
+                            gt_masks, cfg_list)
+        point_targets = list(point_targets)
+
+        if len(point_targets) > 0:
+            point_targets = torch.cat(point_targets)
+
+        return point_targets
+
+    def _get_target_single(self, rois, rel_roi_points, pos_assigned_gt_inds,
+                           gt_masks, cfg):
+        """Get training target of MaskPointHead for each image."""
+        num_pos = rois.size(0)
+        num_points = cfg.num_points
+        if num_pos > 0:
+            gt_masks_th = (
+                gt_masks.to_tensor(rois.dtype, rois.device).index_select(
+                    0, pos_assigned_gt_inds))
+            gt_masks_th = gt_masks_th.unsqueeze(1)
+            rel_img_points = rel_roi_point_to_rel_img_point(
+                rois, rel_roi_points, gt_masks_th.shape[2:])
+            point_targets = point_sample(gt_masks_th,
+                                         rel_img_points).squeeze(1)
+        else:
+            point_targets = rois.new_zeros((0, num_points))
+        return point_targets
+
+    def loss(self, point_pred, point_targets, labels):
+        """Calculate loss for MaskPointHead.
+
+        Args:
+            point_pred (Tensor): Point predication result, shape
+                (num_rois, num_classes, num_points).
+            point_targets (Tensor): Point targets, shape (num_roi, num_points).
+            labels (Tensor): Class label of corresponding boxes,
+                shape (num_rois, )
+
+        Returns:
+            dict[str, Tensor]: a dictionary of point loss components
+        """
+
+        loss = dict()
+        if self.class_agnostic:
+            loss_point = self.loss_point(point_pred, point_targets,
+                                         torch.zeros_like(labels))
+        else:
+            loss_point = self.loss_point(point_pred, point_targets, labels)
+        loss['loss_point'] = loss_point
+        return loss
+
+    def _get_uncertainty(self, mask_pred, labels):
+        """Estimate uncertainty based on pred logits.
+
+        We estimate uncertainty as L1 distance between 0.0 and the logits
+        prediction in 'mask_pred' for the foreground class in `classes`.
+
+        Args:
+            mask_pred (Tensor): mask predication logits, shape (num_rois,
+                num_classes, mask_height, mask_width).
+
+            labels (list[Tensor]): Either predicted or ground truth label for
+                each predicted mask, of length num_rois.
+
+        Returns:
+            scores (Tensor): Uncertainty scores with the most uncertain
+                locations having the highest uncertainty score,
+                shape (num_rois, 1, mask_height, mask_width)
+        """
+        if mask_pred.shape[1] == 1:
+            gt_class_logits = mask_pred.clone()
+        else:
+            inds = torch.arange(mask_pred.shape[0], device=mask_pred.device)
+            gt_class_logits = mask_pred[inds, labels].unsqueeze(1)
+        return -torch.abs(gt_class_logits)
+
+    def get_roi_rel_points_train(self, mask_pred, labels, cfg):
+        """Get ``num_points`` most uncertain points with random points during
+        train.
+
+        Sample points in [0, 1] x [0, 1] coordinate space based on their
+        uncertainty. The uncertainties are calculated for each point using
+        '_get_uncertainty()' function that takes point's logit prediction as
+        input.
+
+        Args:
+            mask_pred (Tensor): A tensor of shape (num_rois, num_classes,
+                mask_height, mask_width) for class-specific or class-agnostic
+                prediction.
+            labels (list): The ground truth class for each instance.
+            cfg (dict): Training config of point head.
+
+        Returns:
+            point_coords (Tensor): A tensor of shape (num_rois, num_points, 2)
+                that contains the coordinates sampled points.
+        """
+        num_points = cfg.num_points
+        oversample_ratio = cfg.oversample_ratio
+        importance_sample_ratio = cfg.importance_sample_ratio
+        assert oversample_ratio >= 1
+        assert 0 <= importance_sample_ratio <= 1
+        batch_size = mask_pred.shape[0]
+        num_sampled = int(num_points * oversample_ratio)
+        point_coords = torch.rand(
+            batch_size, num_sampled, 2, device=mask_pred.device)
+        point_logits = point_sample(mask_pred, point_coords)
+        # It is crucial to calculate uncertainty based on the sampled
+        # prediction value for the points. Calculating uncertainties of the
+        # coarse predictions first and sampling them for points leads to
+        # incorrect results.  To illustrate this: assume uncertainty func(
+        # logits)=-abs(logits), a sampled point between two coarse
+        # predictions with -1 and 1 logits has 0 logits, and therefore 0
+        # uncertainty value. However, if we calculate uncertainties for the
+        # coarse predictions first, both will have -1 uncertainty,
+        # and sampled point will get -1 uncertainty.
+        point_uncertainties = self._get_uncertainty(point_logits, labels)
+        num_uncertain_points = int(importance_sample_ratio * num_points)
+        num_random_points = num_points - num_uncertain_points
+        idx = torch.topk(
+            point_uncertainties[:, 0, :], k=num_uncertain_points, dim=1)[1]
+        shift = num_sampled * torch.arange(
+            batch_size, dtype=torch.long, device=mask_pred.device)
+        idx += shift[:, None]
+        point_coords = point_coords.view(-1, 2)[idx.view(-1), :].view(
+            batch_size, num_uncertain_points, 2)
+        if num_random_points > 0:
+            rand_roi_coords = torch.rand(
+                batch_size, num_random_points, 2, device=mask_pred.device)
+            point_coords = torch.cat((point_coords, rand_roi_coords), dim=1)
+        return point_coords
+
+    def get_roi_rel_points_test(self, mask_pred, pred_label, cfg):
+        """Get ``num_points`` most uncertain points during test.
+
+        Args:
+            mask_pred (Tensor): A tensor of shape (num_rois, num_classes,
+                mask_height, mask_width) for class-specific or class-agnostic
+                prediction.
+            pred_label (list): The predication class for each instance.
+            cfg (dict): Testing config of point head.
+
+        Returns:
+            point_indices (Tensor): A tensor of shape (num_rois, num_points)
+                that contains indices from [0, mask_height x mask_width) of the
+                most uncertain points.
+            point_coords (Tensor): A tensor of shape (num_rois, num_points, 2)
+                that contains [0, 1] x [0, 1] normalized coordinates of the
+                most uncertain points from the [mask_height, mask_width] grid .
+        """
+        num_points = cfg.subdivision_num_points
+        uncertainty_map = self._get_uncertainty(mask_pred, pred_label)
+        num_rois, _, mask_height, mask_width = uncertainty_map.shape
+        h_step = 1.0 / mask_height
+        w_step = 1.0 / mask_width
+
+        uncertainty_map = uncertainty_map.view(num_rois,
+                                               mask_height * mask_width)
+        num_points = min(mask_height * mask_width, num_points)
+        point_indices = uncertainty_map.topk(num_points, dim=1)[1]
+        point_coords = uncertainty_map.new_zeros(num_rois, num_points, 2)
+        point_coords[:, :, 0] = w_step / 2.0 + (point_indices %
+                                                mask_width).float() * w_step
+        point_coords[:, :, 1] = h_step / 2.0 + (point_indices //
+                                                mask_width).float() * h_step
+        return point_indices, point_coords
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/maskiou_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/maskiou_head.py
new file mode 100644
index 0000000..39bcd6a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/maskiou_head.py
@@ -0,0 +1,186 @@
+import numpy as np
+import torch
+import torch.nn as nn
+from mmcv.cnn import Conv2d, Linear, MaxPool2d, kaiming_init, normal_init
+from mmcv.runner import force_fp32
+from torch.nn.modules.utils import _pair
+
+from mmdet.models.builder import HEADS, build_loss
+
+
+@HEADS.register_module()
+class MaskIoUHead(nn.Module):
+    """Mask IoU Head.
+
+    This head predicts the IoU of predicted masks and corresponding gt masks.
+    """
+
+    def __init__(self,
+                 num_convs=4,
+                 num_fcs=2,
+                 roi_feat_size=14,
+                 in_channels=256,
+                 conv_out_channels=256,
+                 fc_out_channels=1024,
+                 num_classes=80,
+                 loss_iou=dict(type='MSELoss', loss_weight=0.5)):
+        super(MaskIoUHead, self).__init__()
+        self.in_channels = in_channels
+        self.conv_out_channels = conv_out_channels
+        self.fc_out_channels = fc_out_channels
+        self.num_classes = num_classes
+        self.fp16_enabled = False
+
+        self.convs = nn.ModuleList()
+        for i in range(num_convs):
+            if i == 0:
+                # concatenation of mask feature and mask prediction
+                in_channels = self.in_channels + 1
+            else:
+                in_channels = self.conv_out_channels
+            stride = 2 if i == num_convs - 1 else 1
+            self.convs.append(
+                Conv2d(
+                    in_channels,
+                    self.conv_out_channels,
+                    3,
+                    stride=stride,
+                    padding=1))
+
+        roi_feat_size = _pair(roi_feat_size)
+        pooled_area = (roi_feat_size[0] // 2) * (roi_feat_size[1] // 2)
+        self.fcs = nn.ModuleList()
+        for i in range(num_fcs):
+            in_channels = (
+                self.conv_out_channels *
+                pooled_area if i == 0 else self.fc_out_channels)
+            self.fcs.append(Linear(in_channels, self.fc_out_channels))
+
+        self.fc_mask_iou = Linear(self.fc_out_channels, self.num_classes)
+        self.relu = nn.ReLU()
+        self.max_pool = MaxPool2d(2, 2)
+        self.loss_iou = build_loss(loss_iou)
+
+    def init_weights(self):
+        for conv in self.convs:
+            kaiming_init(conv)
+        for fc in self.fcs:
+            kaiming_init(
+                fc,
+                a=1,
+                mode='fan_in',
+                nonlinearity='leaky_relu',
+                distribution='uniform')
+        normal_init(self.fc_mask_iou, std=0.01)
+
+    def forward(self, mask_feat, mask_pred):
+        mask_pred = mask_pred.sigmoid()
+        mask_pred_pooled = self.max_pool(mask_pred.unsqueeze(1))
+
+        x = torch.cat((mask_feat, mask_pred_pooled), 1)
+
+        for conv in self.convs:
+            x = self.relu(conv(x))
+        x = x.flatten(1)
+        for fc in self.fcs:
+            x = self.relu(fc(x))
+        mask_iou = self.fc_mask_iou(x)
+        return mask_iou
+
+    @force_fp32(apply_to=('mask_iou_pred', ))
+    def loss(self, mask_iou_pred, mask_iou_targets):
+        pos_inds = mask_iou_targets > 0
+        if pos_inds.sum() > 0:
+            loss_mask_iou = self.loss_iou(mask_iou_pred[pos_inds],
+                                          mask_iou_targets[pos_inds])
+        else:
+            loss_mask_iou = mask_iou_pred.sum() * 0
+        return dict(loss_mask_iou=loss_mask_iou)
+
+    @force_fp32(apply_to=('mask_pred', ))
+    def get_targets(self, sampling_results, gt_masks, mask_pred, mask_targets,
+                    rcnn_train_cfg):
+        """Compute target of mask IoU.
+
+        Mask IoU target is the IoU of the predicted mask (inside a bbox) and
+        the gt mask of corresponding gt mask (the whole instance).
+        The intersection area is computed inside the bbox, and the gt mask area
+        is computed with two steps, firstly we compute the gt area inside the
+        bbox, then divide it by the area ratio of gt area inside the bbox and
+        the gt area of the whole instance.
+
+        Args:
+            sampling_results (list[:obj:`SamplingResult`]): sampling results.
+            gt_masks (BitmapMask | PolygonMask): Gt masks (the whole instance)
+                of each image, with the same shape of the input image.
+            mask_pred (Tensor): Predicted masks of each positive proposal,
+                shape (num_pos, h, w).
+            mask_targets (Tensor): Gt mask of each positive proposal,
+                binary map of the shape (num_pos, h, w).
+            rcnn_train_cfg (dict): Training config for R-CNN part.
+
+        Returns:
+            Tensor: mask iou target (length == num positive).
+        """
+        pos_proposals = [res.pos_bboxes for res in sampling_results]
+        pos_assigned_gt_inds = [
+            res.pos_assigned_gt_inds for res in sampling_results
+        ]
+
+        # compute the area ratio of gt areas inside the proposals and
+        # the whole instance
+        area_ratios = map(self._get_area_ratio, pos_proposals,
+                          pos_assigned_gt_inds, gt_masks)
+        area_ratios = torch.cat(list(area_ratios))
+        assert mask_targets.size(0) == area_ratios.size(0)
+
+        mask_pred = (mask_pred > rcnn_train_cfg.mask_thr_binary).float()
+        mask_pred_areas = mask_pred.sum((-1, -2))
+
+        # mask_pred and mask_targets are binary maps
+        overlap_areas = (mask_pred * mask_targets).sum((-1, -2))
+
+        # compute the mask area of the whole instance
+        gt_full_areas = mask_targets.sum((-1, -2)) / (area_ratios + 1e-7)
+
+        mask_iou_targets = overlap_areas / (
+            mask_pred_areas + gt_full_areas - overlap_areas)
+        return mask_iou_targets
+
+    def _get_area_ratio(self, pos_proposals, pos_assigned_gt_inds, gt_masks):
+        """Compute area ratio of the gt mask inside the proposal and the gt
+        mask of the corresponding instance."""
+        num_pos = pos_proposals.size(0)
+        if num_pos > 0:
+            area_ratios = []
+            proposals_np = pos_proposals.cpu().numpy()
+            pos_assigned_gt_inds = pos_assigned_gt_inds.cpu().numpy()
+            # compute mask areas of gt instances (batch processing for speedup)
+            gt_instance_mask_area = gt_masks.areas
+            for i in range(num_pos):
+                gt_mask = gt_masks[pos_assigned_gt_inds[i]]
+
+                # crop the gt mask inside the proposal
+                bbox = proposals_np[i, :].astype(np.int32)
+                gt_mask_in_proposal = gt_mask.crop(bbox)
+
+                ratio = gt_mask_in_proposal.areas[0] / (
+                    gt_instance_mask_area[pos_assigned_gt_inds[i]] + 1e-7)
+                area_ratios.append(ratio)
+            area_ratios = torch.from_numpy(np.stack(area_ratios)).float().to(
+                pos_proposals.device)
+        else:
+            area_ratios = pos_proposals.new_zeros((0, ))
+        return area_ratios
+
+    @force_fp32(apply_to=('mask_iou_pred', ))
+    def get_mask_scores(self, mask_iou_pred, det_bboxes, det_labels):
+        """Get the mask scores.
+
+        mask_score = bbox_score * mask_iou
+        """
+        inds = range(det_labels.size(0))
+        mask_scores = mask_iou_pred[inds, det_labels] * det_bboxes[inds, -1]
+        mask_scores = mask_scores.cpu().numpy()
+        det_labels = det_labels.cpu().numpy()
+        return [mask_scores[det_labels == i] for i in range(self.num_classes)]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/scnet_mask_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/scnet_mask_head.py
new file mode 100644
index 0000000..983a2d9
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/scnet_mask_head.py
@@ -0,0 +1,27 @@
+from mmdet.models.builder import HEADS
+from mmdet.models.utils import ResLayer, SimplifiedBasicBlock
+from .fcn_mask_head import FCNMaskHead
+
+
+@HEADS.register_module()
+class SCNetMaskHead(FCNMaskHead):
+    """Mask head for `SCNet <https://arxiv.org/abs/2012.10150>`_.
+
+    Args:
+        conv_to_res (bool, optional): if True, change the conv layers to
+            ``SimplifiedBasicBlock``.
+    """
+
+    def __init__(self, conv_to_res=True, **kwargs):
+        super(SCNetMaskHead, self).__init__(**kwargs)
+        self.conv_to_res = conv_to_res
+        if conv_to_res:
+            assert self.conv_kernel_size == 3
+            self.num_res_blocks = self.num_convs // 2
+            self.convs = ResLayer(
+                SimplifiedBasicBlock,
+                self.in_channels,
+                self.conv_out_channels,
+                self.num_res_blocks,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg)
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/scnet_semantic_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/scnet_semantic_head.py
new file mode 100644
index 0000000..df85a01
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_heads/scnet_semantic_head.py
@@ -0,0 +1,27 @@
+from mmdet.models.builder import HEADS
+from mmdet.models.utils import ResLayer, SimplifiedBasicBlock
+from .fused_semantic_head import FusedSemanticHead
+
+
+@HEADS.register_module()
+class SCNetSemanticHead(FusedSemanticHead):
+    """Mask head for `SCNet <https://arxiv.org/abs/2012.10150>`_.
+
+    Args:
+        conv_to_res (bool, optional): if True, change the conv layers to
+            ``SimplifiedBasicBlock``.
+    """
+
+    def __init__(self, conv_to_res=True, **kwargs):
+        super(SCNetSemanticHead, self).__init__(**kwargs)
+        self.conv_to_res = conv_to_res
+        if self.conv_to_res:
+            num_res_blocks = self.num_convs // 2
+            self.convs = ResLayer(
+                SimplifiedBasicBlock,
+                self.in_channels,
+                self.conv_out_channels,
+                num_res_blocks,
+                conv_cfg=self.conv_cfg,
+                norm_cfg=self.norm_cfg)
+            self.num_convs = num_res_blocks
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/mask_scoring_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_scoring_roi_head.py
new file mode 100644
index 0000000..c6e55c7
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/mask_scoring_roi_head.py
@@ -0,0 +1,122 @@
+import torch
+
+from mmdet.core import bbox2roi
+from ..builder import HEADS, build_head
+from .standard_roi_head import StandardRoIHead
+
+
+@HEADS.register_module()
+class MaskScoringRoIHead(StandardRoIHead):
+    """Mask Scoring RoIHead for Mask Scoring RCNN.
+
+    https://arxiv.org/abs/1903.00241
+    """
+
+    def __init__(self, mask_iou_head, **kwargs):
+        assert mask_iou_head is not None
+        super(MaskScoringRoIHead, self).__init__(**kwargs)
+        self.mask_iou_head = build_head(mask_iou_head)
+
+    def init_weights(self, pretrained):
+        """Initialize the weights in head.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        super(MaskScoringRoIHead, self).init_weights(pretrained)
+        self.mask_iou_head.init_weights()
+
+    def _mask_forward_train(self, x, sampling_results, bbox_feats, gt_masks,
+                            img_metas):
+        """Run forward function and calculate loss for Mask head in
+        training."""
+        pos_labels = torch.cat([res.pos_gt_labels for res in sampling_results])
+        mask_results = super(MaskScoringRoIHead,
+                             self)._mask_forward_train(x, sampling_results,
+                                                       bbox_feats, gt_masks,
+                                                       img_metas)
+        if mask_results['loss_mask'] is None:
+            return mask_results
+
+        # mask iou head forward and loss
+        pos_mask_pred = mask_results['mask_pred'][
+            range(mask_results['mask_pred'].size(0)), pos_labels]
+        mask_iou_pred = self.mask_iou_head(mask_results['mask_feats'],
+                                           pos_mask_pred)
+        pos_mask_iou_pred = mask_iou_pred[range(mask_iou_pred.size(0)),
+                                          pos_labels]
+
+        mask_iou_targets = self.mask_iou_head.get_targets(
+            sampling_results, gt_masks, pos_mask_pred,
+            mask_results['mask_targets'], self.train_cfg)
+        loss_mask_iou = self.mask_iou_head.loss(pos_mask_iou_pred,
+                                                mask_iou_targets)
+        mask_results['loss_mask'].update(loss_mask_iou)
+        return mask_results
+
+    def simple_test_mask(self,
+                         x,
+                         img_metas,
+                         det_bboxes,
+                         det_labels,
+                         rescale=False):
+        """Obtain mask prediction without augmentation."""
+        # image shapes of images in the batch
+        ori_shapes = tuple(meta['ori_shape'] for meta in img_metas)
+        scale_factors = tuple(meta['scale_factor'] for meta in img_metas)
+
+        num_imgs = len(det_bboxes)
+        if all(det_bbox.shape[0] == 0 for det_bbox in det_bboxes):
+            num_classes = self.mask_head.num_classes
+            segm_results = [[[] for _ in range(num_classes)]
+                            for _ in range(num_imgs)]
+            mask_scores = [[[] for _ in range(num_classes)]
+                           for _ in range(num_imgs)]
+        else:
+            # if det_bboxes is rescaled to the original image size, we need to
+            # rescale it back to the testing scale to obtain RoIs.
+            if rescale and not isinstance(scale_factors[0], float):
+                scale_factors = [
+                    torch.from_numpy(scale_factor).to(det_bboxes[0].device)
+                    for scale_factor in scale_factors
+                ]
+            _bboxes = [
+                det_bboxes[i][:, :4] *
+                scale_factors[i] if rescale else det_bboxes[i]
+                for i in range(num_imgs)
+            ]
+            mask_rois = bbox2roi(_bboxes)
+            mask_results = self._mask_forward(x, mask_rois)
+            concat_det_labels = torch.cat(det_labels)
+            # get mask scores with mask iou head
+            mask_feats = mask_results['mask_feats']
+            mask_pred = mask_results['mask_pred']
+            mask_iou_pred = self.mask_iou_head(
+                mask_feats, mask_pred[range(concat_det_labels.size(0)),
+                                      concat_det_labels])
+            # split batch mask prediction back to each image
+            num_bboxes_per_img = tuple(len(_bbox) for _bbox in _bboxes)
+            mask_preds = mask_pred.split(num_bboxes_per_img, 0)
+            mask_iou_preds = mask_iou_pred.split(num_bboxes_per_img, 0)
+
+            # apply mask post-processing to each image individually
+            segm_results = []
+            mask_scores = []
+            for i in range(num_imgs):
+                if det_bboxes[i].shape[0] == 0:
+                    segm_results.append(
+                        [[] for _ in range(self.mask_head.num_classes)])
+                    mask_scores.append(
+                        [[] for _ in range(self.mask_head.num_classes)])
+                else:
+                    segm_result = self.mask_head.get_seg_masks(
+                        mask_preds[i], _bboxes[i], det_labels[i],
+                        self.test_cfg, ori_shapes[i], scale_factors[i],
+                        rescale)
+                    # get mask scores with mask iou head
+                    mask_score = self.mask_iou_head.get_mask_scores(
+                        mask_iou_preds[i], det_bboxes[i], det_labels[i])
+                    segm_results.append(segm_result)
+                    mask_scores.append(mask_score)
+        return list(zip(segm_results, mask_scores))
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/pisa_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/pisa_roi_head.py
new file mode 100644
index 0000000..e011136
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/pisa_roi_head.py
@@ -0,0 +1,159 @@
+from mmdet.core import bbox2roi
+from ..builder import HEADS
+from ..losses.pisa_loss import carl_loss, isr_p
+from .standard_roi_head import StandardRoIHead
+
+
+@HEADS.register_module()
+class PISARoIHead(StandardRoIHead):
+    r"""The RoI head for `Prime Sample Attention in Object Detection
+    <https://arxiv.org/abs/1904.04821>`_."""
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      proposal_list,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None):
+        """Forward function for training.
+
+        Args:
+            x (list[Tensor]): List of multi-level img features.
+            img_metas (list[dict]): List of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+            proposals (list[Tensors]): List of region proposals.
+            gt_bboxes (list[Tensor]): Each item are the truth boxes for each
+                image in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): Class indices corresponding to each box
+            gt_bboxes_ignore (list[Tensor], optional): Specify which bounding
+                boxes can be ignored when computing the loss.
+            gt_masks (None | Tensor) : True segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        # assign gts and sample proposals
+        if self.with_bbox or self.with_mask:
+            num_imgs = len(img_metas)
+            if gt_bboxes_ignore is None:
+                gt_bboxes_ignore = [None for _ in range(num_imgs)]
+            sampling_results = []
+            neg_label_weights = []
+            for i in range(num_imgs):
+                assign_result = self.bbox_assigner.assign(
+                    proposal_list[i], gt_bboxes[i], gt_bboxes_ignore[i],
+                    gt_labels[i])
+                sampling_result = self.bbox_sampler.sample(
+                    assign_result,
+                    proposal_list[i],
+                    gt_bboxes[i],
+                    gt_labels[i],
+                    feats=[lvl_feat[i][None] for lvl_feat in x])
+                # neg label weight is obtained by sampling when using ISR-N
+                neg_label_weight = None
+                if isinstance(sampling_result, tuple):
+                    sampling_result, neg_label_weight = sampling_result
+                sampling_results.append(sampling_result)
+                neg_label_weights.append(neg_label_weight)
+
+        losses = dict()
+        # bbox head forward and loss
+        if self.with_bbox:
+            bbox_results = self._bbox_forward_train(
+                x,
+                sampling_results,
+                gt_bboxes,
+                gt_labels,
+                img_metas,
+                neg_label_weights=neg_label_weights)
+            losses.update(bbox_results['loss_bbox'])
+
+        # mask head forward and loss
+        if self.with_mask:
+            mask_results = self._mask_forward_train(x, sampling_results,
+                                                    bbox_results['bbox_feats'],
+                                                    gt_masks, img_metas)
+            losses.update(mask_results['loss_mask'])
+
+        return losses
+
+    def _bbox_forward(self, x, rois):
+        """Box forward function used in both training and testing."""
+        # TODO: a more flexible way to decide which feature maps to use
+        bbox_feats = self.bbox_roi_extractor(
+            x[:self.bbox_roi_extractor.num_inputs], rois)
+        if self.with_shared_head:
+            bbox_feats = self.shared_head(bbox_feats)
+        cls_score, bbox_pred = self.bbox_head(bbox_feats)
+
+        bbox_results = dict(
+            cls_score=cls_score, bbox_pred=bbox_pred, bbox_feats=bbox_feats)
+        return bbox_results
+
+    def _bbox_forward_train(self,
+                            x,
+                            sampling_results,
+                            gt_bboxes,
+                            gt_labels,
+                            img_metas,
+                            neg_label_weights=None):
+        """Run forward function and calculate loss for box head in training."""
+        rois = bbox2roi([res.bboxes for res in sampling_results])
+
+        bbox_results = self._bbox_forward(x, rois)
+
+        bbox_targets = self.bbox_head.get_targets(sampling_results, gt_bboxes,
+                                                  gt_labels, self.train_cfg)
+
+        # neg_label_weights obtained by sampler is image-wise, mapping back to
+        # the corresponding location in label weights
+        if neg_label_weights[0] is not None:
+            label_weights = bbox_targets[1]
+            cur_num_rois = 0
+            for i in range(len(sampling_results)):
+                num_pos = sampling_results[i].pos_inds.size(0)
+                num_neg = sampling_results[i].neg_inds.size(0)
+                label_weights[cur_num_rois + num_pos:cur_num_rois + num_pos +
+                              num_neg] = neg_label_weights[i]
+                cur_num_rois += num_pos + num_neg
+
+        cls_score = bbox_results['cls_score']
+        bbox_pred = bbox_results['bbox_pred']
+
+        # Apply ISR-P
+        isr_cfg = self.train_cfg.get('isr', None)
+        if isr_cfg is not None:
+            bbox_targets = isr_p(
+                cls_score,
+                bbox_pred,
+                bbox_targets,
+                rois,
+                sampling_results,
+                self.bbox_head.loss_cls,
+                self.bbox_head.bbox_coder,
+                **isr_cfg,
+                num_class=self.bbox_head.num_classes)
+        loss_bbox = self.bbox_head.loss(cls_score, bbox_pred, rois,
+                                        *bbox_targets)
+
+        # Add CARL Loss
+        carl_cfg = self.train_cfg.get('carl', None)
+        if carl_cfg is not None:
+            loss_carl = carl_loss(
+                cls_score,
+                bbox_targets[0],
+                bbox_pred,
+                bbox_targets[2],
+                self.bbox_head.loss_bbox,
+                **carl_cfg,
+                num_class=self.bbox_head.num_classes)
+            loss_bbox.update(loss_carl)
+
+        bbox_results.update(loss_bbox=loss_bbox)
+        return bbox_results
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/point_rend_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/point_rend_roi_head.py
new file mode 100644
index 0000000..478cdf5
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/point_rend_roi_head.py
@@ -0,0 +1,218 @@
+# Modified from https://github.com/facebookresearch/detectron2/tree/master/projects/PointRend  # noqa
+
+import torch
+import torch.nn.functional as F
+from mmcv.ops import point_sample, rel_roi_point_to_rel_img_point
+
+from mmdet.core import bbox2roi, bbox_mapping, merge_aug_masks
+from .. import builder
+from ..builder import HEADS
+from .standard_roi_head import StandardRoIHead
+
+
+@HEADS.register_module()
+class PointRendRoIHead(StandardRoIHead):
+    """`PointRend <https://arxiv.org/abs/1912.08193>`_."""
+
+    def __init__(self, point_head, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        assert self.with_bbox and self.with_mask
+        self.init_point_head(point_head)
+
+    def init_point_head(self, point_head):
+        """Initialize ``point_head``"""
+        self.point_head = builder.build_head(point_head)
+
+    def init_weights(self, pretrained):
+        """Initialize the weights in head.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+        """
+        super().init_weights(pretrained)
+        self.point_head.init_weights()
+
+    def _mask_forward_train(self, x, sampling_results, bbox_feats, gt_masks,
+                            img_metas):
+        """Run forward function and calculate loss for mask head and point head
+        in training."""
+        mask_results = super()._mask_forward_train(x, sampling_results,
+                                                   bbox_feats, gt_masks,
+                                                   img_metas)
+        if mask_results['loss_mask'] is not None:
+            loss_point = self._mask_point_forward_train(
+                x, sampling_results, mask_results['mask_pred'], gt_masks,
+                img_metas)
+            mask_results['loss_mask'].update(loss_point)
+
+        return mask_results
+
+    def _mask_point_forward_train(self, x, sampling_results, mask_pred,
+                                  gt_masks, img_metas):
+        """Run forward function and calculate loss for point head in
+        training."""
+        pos_labels = torch.cat([res.pos_gt_labels for res in sampling_results])
+        rel_roi_points = self.point_head.get_roi_rel_points_train(
+            mask_pred, pos_labels, cfg=self.train_cfg)
+        rois = bbox2roi([res.pos_bboxes for res in sampling_results])
+
+        fine_grained_point_feats = self._get_fine_grained_point_feats(
+            x, rois, rel_roi_points, img_metas)
+        coarse_point_feats = point_sample(mask_pred, rel_roi_points)
+        mask_point_pred = self.point_head(fine_grained_point_feats,
+                                          coarse_point_feats)
+        mask_point_target = self.point_head.get_targets(
+            rois, rel_roi_points, sampling_results, gt_masks, self.train_cfg)
+        loss_mask_point = self.point_head.loss(mask_point_pred,
+                                               mask_point_target, pos_labels)
+
+        return loss_mask_point
+
+    def _get_fine_grained_point_feats(self, x, rois, rel_roi_points,
+                                      img_metas):
+        """Sample fine grained feats from each level feature map and
+        concatenate them together."""
+        num_imgs = len(img_metas)
+        fine_grained_feats = []
+        for idx in range(self.mask_roi_extractor.num_inputs):
+            feats = x[idx]
+            spatial_scale = 1. / float(
+                self.mask_roi_extractor.featmap_strides[idx])
+            point_feats = []
+            for batch_ind in range(num_imgs):
+                # unravel batch dim
+                feat = feats[batch_ind].unsqueeze(0)
+                inds = (rois[:, 0].long() == batch_ind)
+                if inds.any():
+                    rel_img_points = rel_roi_point_to_rel_img_point(
+                        rois[inds], rel_roi_points[inds], feat.shape[2:],
+                        spatial_scale).unsqueeze(0)
+                    point_feat = point_sample(feat, rel_img_points)
+                    point_feat = point_feat.squeeze(0).transpose(0, 1)
+                    point_feats.append(point_feat)
+            fine_grained_feats.append(torch.cat(point_feats, dim=0))
+        return torch.cat(fine_grained_feats, dim=1)
+
+    def _mask_point_forward_test(self, x, rois, label_pred, mask_pred,
+                                 img_metas):
+        """Mask refining process with point head in testing."""
+        refined_mask_pred = mask_pred.clone()
+        for subdivision_step in range(self.test_cfg.subdivision_steps):
+            refined_mask_pred = F.interpolate(
+                refined_mask_pred,
+                scale_factor=self.test_cfg.scale_factor,
+                mode='bilinear',
+                align_corners=False)
+            # If `subdivision_num_points` is larger or equal to the
+            # resolution of the next step, then we can skip this step
+            num_rois, channels, mask_height, mask_width = \
+                refined_mask_pred.shape
+            if (self.test_cfg.subdivision_num_points >=
+                    self.test_cfg.scale_factor**2 * mask_height * mask_width
+                    and
+                    subdivision_step < self.test_cfg.subdivision_steps - 1):
+                continue
+            point_indices, rel_roi_points = \
+                self.point_head.get_roi_rel_points_test(
+                    refined_mask_pred, label_pred, cfg=self.test_cfg)
+            fine_grained_point_feats = self._get_fine_grained_point_feats(
+                x, rois, rel_roi_points, img_metas)
+            coarse_point_feats = point_sample(mask_pred, rel_roi_points)
+            mask_point_pred = self.point_head(fine_grained_point_feats,
+                                              coarse_point_feats)
+
+            point_indices = point_indices.unsqueeze(1).expand(-1, channels, -1)
+            refined_mask_pred = refined_mask_pred.reshape(
+                num_rois, channels, mask_height * mask_width)
+            refined_mask_pred = refined_mask_pred.scatter_(
+                2, point_indices, mask_point_pred)
+            refined_mask_pred = refined_mask_pred.view(num_rois, channels,
+                                                       mask_height, mask_width)
+
+        return refined_mask_pred
+
+    def simple_test_mask(self,
+                         x,
+                         img_metas,
+                         det_bboxes,
+                         det_labels,
+                         rescale=False):
+        """Obtain mask prediction without augmentation."""
+        ori_shapes = tuple(meta['ori_shape'] for meta in img_metas)
+        scale_factors = tuple(meta['scale_factor'] for meta in img_metas)
+        num_imgs = len(det_bboxes)
+        if all(det_bbox.shape[0] == 0 for det_bbox in det_bboxes):
+            segm_results = [[[] for _ in range(self.mask_head.num_classes)]
+                            for _ in range(num_imgs)]
+        else:
+            # if det_bboxes is rescaled to the original image size, we need to
+            # rescale it back to the testing scale to obtain RoIs.
+            if rescale and not isinstance(scale_factors[0], float):
+                scale_factors = [
+                    torch.from_numpy(scale_factor).to(det_bboxes[0].device)
+                    for scale_factor in scale_factors
+                ]
+            _bboxes = [
+                det_bboxes[i][:, :4] *
+                scale_factors[i] if rescale else det_bboxes[i][:, :4]
+                for i in range(len(det_bboxes))
+            ]
+            mask_rois = bbox2roi(_bboxes)
+            mask_results = self._mask_forward(x, mask_rois)
+            # split batch mask prediction back to each image
+            mask_pred = mask_results['mask_pred']
+            num_mask_roi_per_img = [len(det_bbox) for det_bbox in det_bboxes]
+            mask_preds = mask_pred.split(num_mask_roi_per_img, 0)
+            mask_rois = mask_rois.split(num_mask_roi_per_img, 0)
+
+            # apply mask post-processing to each image individually
+            segm_results = []
+            for i in range(num_imgs):
+                if det_bboxes[i].shape[0] == 0:
+                    segm_results.append(
+                        [[] for _ in range(self.mask_head.num_classes)])
+                else:
+                    x_i = [xx[[i]] for xx in x]
+                    mask_rois_i = mask_rois[i]
+                    mask_rois_i[:, 0] = 0  # TODO: remove this hack
+                    mask_pred_i = self._mask_point_forward_test(
+                        x_i, mask_rois_i, det_labels[i], mask_preds[i],
+                        [img_metas])
+                    segm_result = self.mask_head.get_seg_masks(
+                        mask_pred_i, _bboxes[i], det_labels[i], self.test_cfg,
+                        ori_shapes[i], scale_factors[i], rescale)
+                    segm_results.append(segm_result)
+        return segm_results
+
+    def aug_test_mask(self, feats, img_metas, det_bboxes, det_labels):
+        """Test for mask head with test time augmentation."""
+        if det_bboxes.shape[0] == 0:
+            segm_result = [[] for _ in range(self.mask_head.num_classes)]
+        else:
+            aug_masks = []
+            for x, img_meta in zip(feats, img_metas):
+                img_shape = img_meta[0]['img_shape']
+                scale_factor = img_meta[0]['scale_factor']
+                flip = img_meta[0]['flip']
+                _bboxes = bbox_mapping(det_bboxes[:, :4], img_shape,
+                                       scale_factor, flip)
+                mask_rois = bbox2roi([_bboxes])
+                mask_results = self._mask_forward(x, mask_rois)
+                mask_results['mask_pred'] = self._mask_point_forward_test(
+                    x, mask_rois, det_labels, mask_results['mask_pred'],
+                    img_metas)
+                # convert to numpy array to save memory
+                aug_masks.append(
+                    mask_results['mask_pred'].sigmoid().cpu().numpy())
+            merged_masks = merge_aug_masks(aug_masks, img_metas, self.test_cfg)
+
+            ori_shape = img_metas[0][0]['ori_shape']
+            segm_result = self.mask_head.get_seg_masks(
+                merged_masks,
+                det_bboxes,
+                det_labels,
+                self.test_cfg,
+                ori_shape,
+                scale_factor=1.0,
+                rescale=False)
+        return segm_result
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/__init__.py b/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/__init__.py
new file mode 100644
index 0000000..a6ec0ec
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/__init__.py
@@ -0,0 +1,7 @@
+from .generic_roi_extractor import GenericRoIExtractor
+from .single_level_roi_extractor import SingleRoIExtractor
+
+__all__ = [
+    'SingleRoIExtractor',
+    'GenericRoIExtractor',
+]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/base_roi_extractor.py b/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/base_roi_extractor.py
new file mode 100644
index 0000000..8479325
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/base_roi_extractor.py
@@ -0,0 +1,83 @@
+from abc import ABCMeta, abstractmethod
+
+import torch
+import torch.nn as nn
+from mmcv import ops
+
+
+class BaseRoIExtractor(nn.Module, metaclass=ABCMeta):
+    """Base class for RoI extractor.
+
+    Args:
+        roi_layer (dict): Specify RoI layer type and arguments.
+        out_channels (int): Output channels of RoI layers.
+        featmap_strides (List[int]): Strides of input feature maps.
+    """
+
+    def __init__(self, roi_layer, out_channels, featmap_strides):
+        super(BaseRoIExtractor, self).__init__()
+        self.roi_layers = self.build_roi_layers(roi_layer, featmap_strides)
+        self.out_channels = out_channels
+        self.featmap_strides = featmap_strides
+        self.fp16_enabled = False
+
+    @property
+    def num_inputs(self):
+        """int: Number of input feature maps."""
+        return len(self.featmap_strides)
+
+    def init_weights(self):
+        pass
+
+    def build_roi_layers(self, layer_cfg, featmap_strides):
+        """Build RoI operator to extract feature from each level feature map.
+
+        Args:
+            layer_cfg (dict): Dictionary to construct and config RoI layer
+                operation. Options are modules under ``mmcv/ops`` such as
+                ``RoIAlign``.
+            featmap_strides (List[int]): The stride of input feature map w.r.t
+                to the original image size, which would be used to scale RoI
+                coordinate (original image coordinate system) to feature
+                coordinate system.
+
+        Returns:
+            nn.ModuleList: The RoI extractor modules for each level feature
+                map.
+        """
+
+        cfg = layer_cfg.copy()
+        layer_type = cfg.pop('type')
+        assert hasattr(ops, layer_type)
+        layer_cls = getattr(ops, layer_type)
+        roi_layers = nn.ModuleList(
+            [layer_cls(spatial_scale=1 / s, **cfg) for s in featmap_strides])
+        return roi_layers
+
+    def roi_rescale(self, rois, scale_factor):
+        """Scale RoI coordinates by scale factor.
+
+        Args:
+            rois (torch.Tensor): RoI (Region of Interest), shape (n, 5)
+            scale_factor (float): Scale factor that RoI will be multiplied by.
+
+        Returns:
+            torch.Tensor: Scaled RoI.
+        """
+
+        cx = (rois[:, 1] + rois[:, 3]) * 0.5
+        cy = (rois[:, 2] + rois[:, 4]) * 0.5
+        w = rois[:, 3] - rois[:, 1]
+        h = rois[:, 4] - rois[:, 2]
+        new_w = w * scale_factor
+        new_h = h * scale_factor
+        x1 = cx - new_w * 0.5
+        x2 = cx + new_w * 0.5
+        y1 = cy - new_h * 0.5
+        y2 = cy + new_h * 0.5
+        new_rois = torch.stack((rois[:, 0], x1, y1, x2, y2), dim=-1)
+        return new_rois
+
+    @abstractmethod
+    def forward(self, feats, rois, roi_scale_factor=None):
+        pass
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/generic_roi_extractor.py b/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/generic_roi_extractor.py
new file mode 100644
index 0000000..80c25bb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/generic_roi_extractor.py
@@ -0,0 +1,83 @@
+from mmcv.cnn.bricks import build_plugin_layer
+from mmcv.runner import force_fp32
+
+from mmdet.models.builder import ROI_EXTRACTORS
+from .base_roi_extractor import BaseRoIExtractor
+
+
+@ROI_EXTRACTORS.register_module()
+class GenericRoIExtractor(BaseRoIExtractor):
+    """Extract RoI features from all level feature maps levels.
+
+    This is the implementation of `A novel Region of Interest Extraction Layer
+    for Instance Segmentation <https://arxiv.org/abs/2004.13665>`_.
+
+    Args:
+        aggregation (str): The method to aggregate multiple feature maps.
+            Options are 'sum', 'concat'. Default: 'sum'.
+        pre_cfg (dict | None): Specify pre-processing modules. Default: None.
+        post_cfg (dict | None): Specify post-processing modules. Default: None.
+        kwargs (keyword arguments): Arguments that are the same
+            as :class:`BaseRoIExtractor`.
+    """
+
+    def __init__(self,
+                 aggregation='sum',
+                 pre_cfg=None,
+                 post_cfg=None,
+                 **kwargs):
+        super(GenericRoIExtractor, self).__init__(**kwargs)
+
+        assert aggregation in ['sum', 'concat']
+
+        self.aggregation = aggregation
+        self.with_post = post_cfg is not None
+        self.with_pre = pre_cfg is not None
+        # build pre/post processing modules
+        if self.with_post:
+            self.post_module = build_plugin_layer(post_cfg, '_post_module')[1]
+        if self.with_pre:
+            self.pre_module = build_plugin_layer(pre_cfg, '_pre_module')[1]
+
+    @force_fp32(apply_to=('feats', ), out_fp16=True)
+    def forward(self, feats, rois, roi_scale_factor=None):
+        """Forward function."""
+        if len(feats) == 1:
+            return self.roi_layers[0](feats[0], rois)
+
+        out_size = self.roi_layers[0].output_size
+        num_levels = len(feats)
+        roi_feats = feats[0].new_zeros(
+            rois.size(0), self.out_channels, *out_size)
+
+        # some times rois is an empty tensor
+        if roi_feats.shape[0] == 0:
+            return roi_feats
+
+        if roi_scale_factor is not None:
+            rois = self.roi_rescale(rois, roi_scale_factor)
+
+        # mark the starting channels for concat mode
+        start_channels = 0
+        for i in range(num_levels):
+            roi_feats_t = self.roi_layers[i](feats[i], rois)
+            end_channels = start_channels + roi_feats_t.size(1)
+            if self.with_pre:
+                # apply pre-processing to a RoI extracted from each layer
+                roi_feats_t = self.pre_module(roi_feats_t)
+            if self.aggregation == 'sum':
+                # and sum them all
+                roi_feats += roi_feats_t
+            else:
+                # and concat them along channel dimension
+                roi_feats[:, start_channels:end_channels] = roi_feats_t
+            # update channels starting position
+            start_channels = end_channels
+        # check if concat channels match at the end
+        if self.aggregation == 'concat':
+            assert start_channels == self.out_channels
+
+        if self.with_post:
+            # apply post-processing before return the result
+            roi_feats = self.post_module(roi_feats)
+        return roi_feats
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/single_level_roi_extractor.py b/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/single_level_roi_extractor.py
new file mode 100644
index 0000000..cfc838f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/roi_extractors/single_level_roi_extractor.py
@@ -0,0 +1,108 @@
+import torch
+from mmcv.runner import force_fp32
+
+from mmdet.models.builder import ROI_EXTRACTORS
+from .base_roi_extractor import BaseRoIExtractor
+
+
+@ROI_EXTRACTORS.register_module()
+class SingleRoIExtractor(BaseRoIExtractor):
+    """Extract RoI features from a single level feature map.
+
+    If there are multiple input feature levels, each RoI is mapped to a level
+    according to its scale. The mapping rule is proposed in
+    `FPN <https://arxiv.org/abs/1612.03144>`_.
+
+    Args:
+        roi_layer (dict): Specify RoI layer type and arguments.
+        out_channels (int): Output channels of RoI layers.
+        featmap_strides (List[int]): Strides of input feature maps.
+        finest_scale (int): Scale threshold of mapping to level 0. Default: 56.
+    """
+
+    def __init__(self,
+                 roi_layer,
+                 out_channels,
+                 featmap_strides,
+                 finest_scale=56):
+        super(SingleRoIExtractor, self).__init__(roi_layer, out_channels,
+                                                 featmap_strides)
+        self.finest_scale = finest_scale
+
+    def map_roi_levels(self, rois, num_levels):
+        """Map rois to corresponding feature levels by scales.
+
+        - scale < finest_scale * 2: level 0
+        - finest_scale * 2 <= scale < finest_scale * 4: level 1
+        - finest_scale * 4 <= scale < finest_scale * 8: level 2
+        - scale >= finest_scale * 8: level 3
+
+        Args:
+            rois (Tensor): Input RoIs, shape (k, 5).
+            num_levels (int): Total level number.
+
+        Returns:
+            Tensor: Level index (0-based) of each RoI, shape (k, )
+        """
+        scale = torch.sqrt(
+            (rois[:, 3] - rois[:, 1]) * (rois[:, 4] - rois[:, 2]))
+        target_lvls = torch.floor(torch.log2(scale / self.finest_scale + 1e-6))
+        target_lvls = target_lvls.clamp(min=0, max=num_levels - 1).long()
+        return target_lvls
+
+    @force_fp32(apply_to=('feats', ), out_fp16=True)
+    def forward(self, feats, rois, roi_scale_factor=None):
+        """Forward function."""
+        out_size = self.roi_layers[0].output_size
+        num_levels = len(feats)
+        expand_dims = (-1, self.out_channels * out_size[0] * out_size[1])
+        if torch.onnx.is_in_onnx_export():
+            # Work around to export mask-rcnn to onnx
+            roi_feats = rois[:, :1].clone().detach()
+            roi_feats = roi_feats.expand(*expand_dims)
+            roi_feats = roi_feats.reshape(-1, self.out_channels, *out_size)
+            roi_feats = roi_feats * 0
+        else:
+            roi_feats = feats[0].new_zeros(
+                rois.size(0), self.out_channels, *out_size)
+        # TODO: remove this when parrots supports
+        if torch.__version__ == 'parrots':
+            roi_feats.requires_grad = True
+
+        if num_levels == 1:
+            if len(rois) == 0:
+                return roi_feats
+            return self.roi_layers[0](feats[0], rois)
+
+        target_lvls = self.map_roi_levels(rois, num_levels)
+
+        if roi_scale_factor is not None:
+            rois = self.roi_rescale(rois, roi_scale_factor)
+
+        for i in range(num_levels):
+            mask = target_lvls == i
+            if torch.onnx.is_in_onnx_export():
+                # To keep all roi_align nodes exported to onnx
+                # and skip nonzero op
+                mask = mask.float().unsqueeze(-1).expand(*expand_dims).reshape(
+                    roi_feats.shape)
+                roi_feats_t = self.roi_layers[i](feats[i], rois)
+                roi_feats_t *= mask
+                roi_feats += roi_feats_t
+                continue
+            inds = mask.nonzero(as_tuple=False).squeeze(1)
+            if inds.numel() > 0:
+                rois_ = rois[inds]
+                roi_feats_t = self.roi_layers[i](feats[i], rois_)
+                roi_feats[inds] = roi_feats_t
+            else:
+                # Sometimes some pyramid levels will not be used for RoI
+                # feature extraction and this will cause an incomplete
+                # computation graph in one GPU, which is different from those
+                # in other GPUs and will cause a hanging error.
+                # Therefore, we add it to ensure each feature pyramid is
+                # included in the computation graph to avoid runtime bugs.
+                roi_feats += sum(
+                    x.view(-1)[0]
+                    for x in self.parameters()) * 0. + feats[i].sum() * 0.
+        return roi_feats
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/scnet_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/scnet_roi_head.py
new file mode 100644
index 0000000..85aaa2f
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/scnet_roi_head.py
@@ -0,0 +1,582 @@
+import torch
+import torch.nn.functional as F
+
+from mmdet.core import (bbox2result, bbox2roi, bbox_mapping, merge_aug_bboxes,
+                        merge_aug_masks, multiclass_nms)
+from ..builder import HEADS, build_head, build_roi_extractor
+from .cascade_roi_head import CascadeRoIHead
+
+
+@HEADS.register_module()
+class SCNetRoIHead(CascadeRoIHead):
+    """RoIHead for `SCNet <https://arxiv.org/abs/2012.10150>`_.
+
+    Args:
+        num_stages (int): number of cascade stages.
+        stage_loss_weights (list): loss weight of cascade stages.
+        semantic_roi_extractor (dict): config to init semantic roi extractor.
+        semantic_head (dict): config to init semantic head.
+        feat_relay_head (dict): config to init feature_relay_head.
+        glbctx_head (dict): config to init global context head.
+    """
+
+    def __init__(self,
+                 num_stages,
+                 stage_loss_weights,
+                 semantic_roi_extractor=None,
+                 semantic_head=None,
+                 feat_relay_head=None,
+                 glbctx_head=None,
+                 **kwargs):
+        super(SCNetRoIHead, self).__init__(num_stages, stage_loss_weights,
+                                           **kwargs)
+        assert self.with_bbox and self.with_mask
+        assert not self.with_shared_head  # shared head is not supported
+
+        if semantic_head is not None:
+            self.semantic_roi_extractor = build_roi_extractor(
+                semantic_roi_extractor)
+            self.semantic_head = build_head(semantic_head)
+
+        if feat_relay_head is not None:
+            self.feat_relay_head = build_head(feat_relay_head)
+
+        if glbctx_head is not None:
+            self.glbctx_head = build_head(glbctx_head)
+
+    def init_mask_head(self, mask_roi_extractor, mask_head):
+        """Initialize ``mask_head``"""
+        if mask_roi_extractor is not None:
+            self.mask_roi_extractor = build_roi_extractor(mask_roi_extractor)
+            self.mask_head = build_head(mask_head)
+
+    def init_weights(self, pretrained):
+        """Initialize the weights in head.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        for i in range(self.num_stages):
+            if self.with_bbox:
+                self.bbox_roi_extractor[i].init_weights()
+                self.bbox_head[i].init_weights()
+        if self.with_mask:
+            self.mask_roi_extractor.init_weights()
+            self.mask_head.init_weights()
+        if self.with_semantic:
+            self.semantic_head.init_weights()
+        if self.with_glbctx:
+            self.glbctx_head.init_weights()
+        if self.with_feat_relay:
+            self.feat_relay_head.init_weights()
+
+    @property
+    def with_semantic(self):
+        """bool: whether the head has semantic head"""
+        return hasattr(self,
+                       'semantic_head') and self.semantic_head is not None
+
+    @property
+    def with_feat_relay(self):
+        """bool: whether the head has feature relay head"""
+        return (hasattr(self, 'feat_relay_head')
+                and self.feat_relay_head is not None)
+
+    @property
+    def with_glbctx(self):
+        """bool: whether the head has global context head"""
+        return hasattr(self, 'glbctx_head') and self.glbctx_head is not None
+
+    def _fuse_glbctx(self, roi_feats, glbctx_feat, rois):
+        """Fuse global context feats with roi feats."""
+        assert roi_feats.size(0) == rois.size(0)
+        img_inds = torch.unique(rois[:, 0].cpu(), sorted=True).long()
+        fused_feats = torch.zeros_like(roi_feats)
+        for img_id in img_inds:
+            inds = (rois[:, 0] == img_id.item())
+            fused_feats[inds] = roi_feats[inds] + glbctx_feat[img_id]
+        return fused_feats
+
+    def _slice_pos_feats(self, feats, sampling_results):
+        """Get features from pos rois."""
+        num_rois = [res.bboxes.size(0) for res in sampling_results]
+        num_pos_rois = [res.pos_bboxes.size(0) for res in sampling_results]
+        inds = torch.zeros(sum(num_rois), dtype=torch.bool)
+        start = 0
+        for i in range(len(num_rois)):
+            start = 0 if i == 0 else start + num_rois[i - 1]
+            stop = start + num_pos_rois[i]
+            inds[start:stop] = 1
+        sliced_feats = feats[inds]
+        return sliced_feats
+
+    def _bbox_forward(self,
+                      stage,
+                      x,
+                      rois,
+                      semantic_feat=None,
+                      glbctx_feat=None):
+        """Box head forward function used in both training and testing."""
+        bbox_roi_extractor = self.bbox_roi_extractor[stage]
+        bbox_head = self.bbox_head[stage]
+        bbox_feats = bbox_roi_extractor(
+            x[:len(bbox_roi_extractor.featmap_strides)], rois)
+        if self.with_semantic and semantic_feat is not None:
+            bbox_semantic_feat = self.semantic_roi_extractor([semantic_feat],
+                                                             rois)
+            if bbox_semantic_feat.shape[-2:] != bbox_feats.shape[-2:]:
+                bbox_semantic_feat = F.adaptive_avg_pool2d(
+                    bbox_semantic_feat, bbox_feats.shape[-2:])
+            bbox_feats += bbox_semantic_feat
+        if self.with_glbctx and glbctx_feat is not None:
+            bbox_feats = self._fuse_glbctx(bbox_feats, glbctx_feat, rois)
+        cls_score, bbox_pred, relayed_feat = bbox_head(
+            bbox_feats, return_shared_feat=True)
+
+        bbox_results = dict(
+            cls_score=cls_score,
+            bbox_pred=bbox_pred,
+            relayed_feat=relayed_feat)
+        return bbox_results
+
+    def _mask_forward(self,
+                      x,
+                      rois,
+                      semantic_feat=None,
+                      glbctx_feat=None,
+                      relayed_feat=None):
+        """Mask head forward function used in both training and testing."""
+        mask_feats = self.mask_roi_extractor(
+            x[:self.mask_roi_extractor.num_inputs], rois)
+        if self.with_semantic and semantic_feat is not None:
+            mask_semantic_feat = self.semantic_roi_extractor([semantic_feat],
+                                                             rois)
+            if mask_semantic_feat.shape[-2:] != mask_feats.shape[-2:]:
+                mask_semantic_feat = F.adaptive_avg_pool2d(
+                    mask_semantic_feat, mask_feats.shape[-2:])
+            mask_feats += mask_semantic_feat
+        if self.with_glbctx and glbctx_feat is not None:
+            mask_feats = self._fuse_glbctx(mask_feats, glbctx_feat, rois)
+        if self.with_feat_relay and relayed_feat is not None:
+            mask_feats = mask_feats + relayed_feat
+        mask_pred = self.mask_head(mask_feats)
+        mask_results = dict(mask_pred=mask_pred)
+
+        return mask_results
+
+    def _bbox_forward_train(self,
+                            stage,
+                            x,
+                            sampling_results,
+                            gt_bboxes,
+                            gt_labels,
+                            rcnn_train_cfg,
+                            semantic_feat=None,
+                            glbctx_feat=None):
+        """Run forward function and calculate loss for box head in training."""
+        bbox_head = self.bbox_head[stage]
+        rois = bbox2roi([res.bboxes for res in sampling_results])
+        bbox_results = self._bbox_forward(
+            stage,
+            x,
+            rois,
+            semantic_feat=semantic_feat,
+            glbctx_feat=glbctx_feat)
+
+        bbox_targets = bbox_head.get_targets(sampling_results, gt_bboxes,
+                                             gt_labels, rcnn_train_cfg)
+        loss_bbox = bbox_head.loss(bbox_results['cls_score'],
+                                   bbox_results['bbox_pred'], rois,
+                                   *bbox_targets)
+
+        bbox_results.update(
+            loss_bbox=loss_bbox, rois=rois, bbox_targets=bbox_targets)
+        return bbox_results
+
+    def _mask_forward_train(self,
+                            x,
+                            sampling_results,
+                            gt_masks,
+                            rcnn_train_cfg,
+                            semantic_feat=None,
+                            glbctx_feat=None,
+                            relayed_feat=None):
+        """Run forward function and calculate loss for mask head in
+        training."""
+        pos_rois = bbox2roi([res.pos_bboxes for res in sampling_results])
+        mask_results = self._mask_forward(
+            x,
+            pos_rois,
+            semantic_feat=semantic_feat,
+            glbctx_feat=glbctx_feat,
+            relayed_feat=relayed_feat)
+
+        mask_targets = self.mask_head.get_targets(sampling_results, gt_masks,
+                                                  rcnn_train_cfg)
+        pos_labels = torch.cat([res.pos_gt_labels for res in sampling_results])
+        loss_mask = self.mask_head.loss(mask_results['mask_pred'],
+                                        mask_targets, pos_labels)
+
+        mask_results = loss_mask
+        return mask_results
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      proposal_list,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None,
+                      gt_semantic_seg=None):
+        """
+        Args:
+            x (list[Tensor]): list of multi-level img features.
+
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+
+            proposal_list (list[Tensors]): list of region proposals.
+
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+
+            gt_labels (list[Tensor]): class indices corresponding to each box
+
+            gt_bboxes_ignore (None, list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+
+            gt_masks (None, Tensor) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+            gt_semantic_seg (None, list[Tensor]): semantic segmentation masks
+                used if the architecture supports semantic segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        losses = dict()
+
+        # semantic segmentation branch
+        if self.with_semantic:
+            semantic_pred, semantic_feat = self.semantic_head(x)
+            loss_seg = self.semantic_head.loss(semantic_pred, gt_semantic_seg)
+            losses['loss_semantic_seg'] = loss_seg
+        else:
+            semantic_feat = None
+
+        # global context branch
+        if self.with_glbctx:
+            mc_pred, glbctx_feat = self.glbctx_head(x)
+            loss_glbctx = self.glbctx_head.loss(mc_pred, gt_labels)
+            losses['loss_glbctx'] = loss_glbctx
+        else:
+            glbctx_feat = None
+
+        for i in range(self.num_stages):
+            self.current_stage = i
+            rcnn_train_cfg = self.train_cfg[i]
+            lw = self.stage_loss_weights[i]
+
+            # assign gts and sample proposals
+            sampling_results = []
+            bbox_assigner = self.bbox_assigner[i]
+            bbox_sampler = self.bbox_sampler[i]
+            num_imgs = len(img_metas)
+            if gt_bboxes_ignore is None:
+                gt_bboxes_ignore = [None for _ in range(num_imgs)]
+
+            for j in range(num_imgs):
+                assign_result = bbox_assigner.assign(proposal_list[j],
+                                                     gt_bboxes[j],
+                                                     gt_bboxes_ignore[j],
+                                                     gt_labels[j])
+                sampling_result = bbox_sampler.sample(
+                    assign_result,
+                    proposal_list[j],
+                    gt_bboxes[j],
+                    gt_labels[j],
+                    feats=[lvl_feat[j][None] for lvl_feat in x])
+                sampling_results.append(sampling_result)
+
+            bbox_results = \
+                self._bbox_forward_train(
+                    i, x, sampling_results, gt_bboxes, gt_labels,
+                    rcnn_train_cfg, semantic_feat, glbctx_feat)
+            roi_labels = bbox_results['bbox_targets'][0]
+
+            for name, value in bbox_results['loss_bbox'].items():
+                losses[f's{i}.{name}'] = (
+                    value * lw if 'loss' in name else value)
+
+            # refine boxes
+            if i < self.num_stages - 1:
+                pos_is_gts = [res.pos_is_gt for res in sampling_results]
+                with torch.no_grad():
+                    proposal_list = self.bbox_head[i].refine_bboxes(
+                        bbox_results['rois'], roi_labels,
+                        bbox_results['bbox_pred'], pos_is_gts, img_metas)
+
+        if self.with_feat_relay:
+            relayed_feat = self._slice_pos_feats(bbox_results['relayed_feat'],
+                                                 sampling_results)
+            relayed_feat = self.feat_relay_head(relayed_feat)
+        else:
+            relayed_feat = None
+
+        mask_results = self._mask_forward_train(x, sampling_results, gt_masks,
+                                                rcnn_train_cfg, semantic_feat,
+                                                glbctx_feat, relayed_feat)
+        mask_lw = sum(self.stage_loss_weights)
+        losses['loss_mask'] = mask_lw * mask_results['loss_mask']
+
+        return losses
+
+    def simple_test(self, x, proposal_list, img_metas, rescale=False):
+        """Test without augmentation."""
+        if self.with_semantic:
+            _, semantic_feat = self.semantic_head(x)
+        else:
+            semantic_feat = None
+
+        if self.with_glbctx:
+            mc_pred, glbctx_feat = self.glbctx_head(x)
+        else:
+            glbctx_feat = None
+
+        num_imgs = len(proposal_list)
+        img_shapes = tuple(meta['img_shape'] for meta in img_metas)
+        ori_shapes = tuple(meta['ori_shape'] for meta in img_metas)
+        scale_factors = tuple(meta['scale_factor'] for meta in img_metas)
+
+        # "ms" in variable names means multi-stage
+        ms_scores = []
+        rcnn_test_cfg = self.test_cfg
+
+        rois = bbox2roi(proposal_list)
+        for i in range(self.num_stages):
+            bbox_head = self.bbox_head[i]
+            bbox_results = self._bbox_forward(
+                i,
+                x,
+                rois,
+                semantic_feat=semantic_feat,
+                glbctx_feat=glbctx_feat)
+            # split batch bbox prediction back to each image
+            cls_score = bbox_results['cls_score']
+            bbox_pred = bbox_results['bbox_pred']
+            num_proposals_per_img = tuple(len(p) for p in proposal_list)
+            rois = rois.split(num_proposals_per_img, 0)
+            cls_score = cls_score.split(num_proposals_per_img, 0)
+            bbox_pred = bbox_pred.split(num_proposals_per_img, 0)
+            ms_scores.append(cls_score)
+
+            if i < self.num_stages - 1:
+                bbox_label = [s[:, :-1].argmax(dim=1) for s in cls_score]
+                rois = torch.cat([
+                    bbox_head.regress_by_class(rois[i], bbox_label[i],
+                                               bbox_pred[i], img_metas[i])
+                    for i in range(num_imgs)
+                ])
+
+        # average scores of each image by stages
+        cls_score = [
+            sum([score[i] for score in ms_scores]) / float(len(ms_scores))
+            for i in range(num_imgs)
+        ]
+
+        # apply bbox post-processing to each image individually
+        det_bboxes = []
+        det_labels = []
+        for i in range(num_imgs):
+            det_bbox, det_label = self.bbox_head[-1].get_bboxes(
+                rois[i],
+                cls_score[i],
+                bbox_pred[i],
+                img_shapes[i],
+                scale_factors[i],
+                rescale=rescale,
+                cfg=rcnn_test_cfg)
+            det_bboxes.append(det_bbox)
+            det_labels.append(det_label)
+        det_bbox_results = [
+            bbox2result(det_bboxes[i], det_labels[i],
+                        self.bbox_head[-1].num_classes)
+            for i in range(num_imgs)
+        ]
+
+        if self.with_mask:
+            if all(det_bbox.shape[0] == 0 for det_bbox in det_bboxes):
+                mask_classes = self.mask_head.num_classes
+                det_segm_results = [[[] for _ in range(mask_classes)]
+                                    for _ in range(num_imgs)]
+            else:
+                if rescale and not isinstance(scale_factors[0], float):
+                    scale_factors = [
+                        torch.from_numpy(scale_factor).to(det_bboxes[0].device)
+                        for scale_factor in scale_factors
+                    ]
+                _bboxes = [
+                    det_bboxes[i][:, :4] *
+                    scale_factors[i] if rescale else det_bboxes[i]
+                    for i in range(num_imgs)
+                ]
+                mask_rois = bbox2roi(_bboxes)
+
+                # get relay feature on mask_rois
+                bbox_results = self._bbox_forward(
+                    -1,
+                    x,
+                    mask_rois,
+                    semantic_feat=semantic_feat,
+                    glbctx_feat=glbctx_feat)
+                relayed_feat = bbox_results['relayed_feat']
+                relayed_feat = self.feat_relay_head(relayed_feat)
+
+                mask_results = self._mask_forward(
+                    x,
+                    mask_rois,
+                    semantic_feat=semantic_feat,
+                    glbctx_feat=glbctx_feat,
+                    relayed_feat=relayed_feat)
+                mask_pred = mask_results['mask_pred']
+
+                # split batch mask prediction back to each image
+                num_bbox_per_img = tuple(len(_bbox) for _bbox in _bboxes)
+                mask_preds = mask_pred.split(num_bbox_per_img, 0)
+
+                # apply mask post-processing to each image individually
+                det_segm_results = []
+                for i in range(num_imgs):
+                    if det_bboxes[i].shape[0] == 0:
+                        det_segm_results.append(
+                            [[] for _ in range(self.mask_head.num_classes)])
+                    else:
+                        segm_result = self.mask_head.get_seg_masks(
+                            mask_preds[i], _bboxes[i], det_labels[i],
+                            self.test_cfg, ori_shapes[i], scale_factors[i],
+                            rescale)
+                        det_segm_results.append(segm_result)
+
+        # return results
+        if self.with_mask:
+            return list(zip(det_bbox_results, det_segm_results))
+        else:
+            return det_bbox_results
+
+    def aug_test(self, img_feats, proposal_list, img_metas, rescale=False):
+        if self.with_semantic:
+            semantic_feats = [
+                self.semantic_head(feat)[1] for feat in img_feats
+            ]
+        else:
+            semantic_feats = [None] * len(img_metas)
+
+        if self.with_glbctx:
+            glbctx_feats = [self.glbctx_head(feat)[1] for feat in img_feats]
+        else:
+            glbctx_feats = [None] * len(img_metas)
+
+        rcnn_test_cfg = self.test_cfg
+        aug_bboxes = []
+        aug_scores = []
+        for x, img_meta, semantic_feat, glbctx_feat in zip(
+                img_feats, img_metas, semantic_feats, glbctx_feats):
+            # only one image in the batch
+            img_shape = img_meta[0]['img_shape']
+            scale_factor = img_meta[0]['scale_factor']
+            flip = img_meta[0]['flip']
+
+            proposals = bbox_mapping(proposal_list[0][:, :4], img_shape,
+                                     scale_factor, flip)
+            # "ms" in variable names means multi-stage
+            ms_scores = []
+
+            rois = bbox2roi([proposals])
+            for i in range(self.num_stages):
+                bbox_head = self.bbox_head[i]
+                bbox_results = self._bbox_forward(
+                    i,
+                    x,
+                    rois,
+                    semantic_feat=semantic_feat,
+                    glbctx_feat=glbctx_feat)
+                ms_scores.append(bbox_results['cls_score'])
+                if i < self.num_stages - 1:
+                    bbox_label = bbox_results['cls_score'].argmax(dim=1)
+                    rois = bbox_head.regress_by_class(
+                        rois, bbox_label, bbox_results['bbox_pred'],
+                        img_meta[0])
+
+            cls_score = sum(ms_scores) / float(len(ms_scores))
+            bboxes, scores = self.bbox_head[-1].get_bboxes(
+                rois,
+                cls_score,
+                bbox_results['bbox_pred'],
+                img_shape,
+                scale_factor,
+                rescale=False,
+                cfg=None)
+            aug_bboxes.append(bboxes)
+            aug_scores.append(scores)
+
+        # after merging, bboxes will be rescaled to the original image size
+        merged_bboxes, merged_scores = merge_aug_bboxes(
+            aug_bboxes, aug_scores, img_metas, rcnn_test_cfg)
+        det_bboxes, det_labels = multiclass_nms(merged_bboxes, merged_scores,
+                                                rcnn_test_cfg.score_thr,
+                                                rcnn_test_cfg.nms,
+                                                rcnn_test_cfg.max_per_img)
+
+        det_bbox_results = bbox2result(det_bboxes, det_labels,
+                                       self.bbox_head[-1].num_classes)
+
+        if self.with_mask:
+            if det_bboxes.shape[0] == 0:
+                det_segm_results = [[]
+                                    for _ in range(self.mask_head.num_classes)]
+            else:
+                aug_masks = []
+                for x, img_meta, semantic_feat, glbctx_feat in zip(
+                        img_feats, img_metas, semantic_feats, glbctx_feats):
+                    img_shape = img_meta[0]['img_shape']
+                    scale_factor = img_meta[0]['scale_factor']
+                    flip = img_meta[0]['flip']
+                    _bboxes = bbox_mapping(det_bboxes[:, :4], img_shape,
+                                           scale_factor, flip)
+                    mask_rois = bbox2roi([_bboxes])
+                    # get relay feature on mask_rois
+                    bbox_results = self._bbox_forward(
+                        -1,
+                        x,
+                        mask_rois,
+                        semantic_feat=semantic_feat,
+                        glbctx_feat=glbctx_feat)
+                    relayed_feat = bbox_results['relayed_feat']
+                    relayed_feat = self.feat_relay_head(relayed_feat)
+                    mask_results = self._mask_forward(
+                        x,
+                        mask_rois,
+                        semantic_feat=semantic_feat,
+                        glbctx_feat=glbctx_feat,
+                        relayed_feat=relayed_feat)
+                    mask_pred = mask_results['mask_pred']
+                    aug_masks.append(mask_pred.sigmoid().cpu().numpy())
+                merged_masks = merge_aug_masks(aug_masks, img_metas,
+                                               self.test_cfg)
+                ori_shape = img_metas[0][0]['ori_shape']
+                det_segm_results = self.mask_head.get_seg_masks(
+                    merged_masks,
+                    det_bboxes,
+                    det_labels,
+                    rcnn_test_cfg,
+                    ori_shape,
+                    scale_factor=1.0,
+                    rescale=False)
+            return [(det_bbox_results, det_segm_results)]
+        else:
+            return [det_bbox_results]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/shared_heads/__init__.py b/python/autoCut/models/swin/mmdet/models/roi_heads/shared_heads/__init__.py
new file mode 100644
index 0000000..bbe7014
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/shared_heads/__init__.py
@@ -0,0 +1,3 @@
+from .res_layer import ResLayer
+
+__all__ = ['ResLayer']
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/shared_heads/res_layer.py b/python/autoCut/models/swin/mmdet/models/roi_heads/shared_heads/res_layer.py
new file mode 100644
index 0000000..b5c3432
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/shared_heads/res_layer.py
@@ -0,0 +1,77 @@
+import torch.nn as nn
+from mmcv.cnn import constant_init, kaiming_init
+from mmcv.runner import auto_fp16, load_checkpoint
+
+from mmdet.models.backbones import ResNet
+from mmdet.models.builder import SHARED_HEADS
+from mmdet.models.utils import ResLayer as _ResLayer
+from mmdet.utils import get_root_logger
+
+
+@SHARED_HEADS.register_module()
+class ResLayer(nn.Module):
+
+    def __init__(self,
+                 depth,
+                 stage=3,
+                 stride=2,
+                 dilation=1,
+                 style='pytorch',
+                 norm_cfg=dict(type='BN', requires_grad=True),
+                 norm_eval=True,
+                 with_cp=False,
+                 dcn=None):
+        super(ResLayer, self).__init__()
+        self.norm_eval = norm_eval
+        self.norm_cfg = norm_cfg
+        self.stage = stage
+        self.fp16_enabled = False
+        block, stage_blocks = ResNet.arch_settings[depth]
+        stage_block = stage_blocks[stage]
+        planes = 64 * 2**stage
+        inplanes = 64 * 2**(stage - 1) * block.expansion
+
+        res_layer = _ResLayer(
+            block,
+            inplanes,
+            planes,
+            stage_block,
+            stride=stride,
+            dilation=dilation,
+            style=style,
+            with_cp=with_cp,
+            norm_cfg=self.norm_cfg,
+            dcn=dcn)
+        self.add_module(f'layer{stage + 1}', res_layer)
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in the module.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        if isinstance(pretrained, str):
+            logger = get_root_logger()
+            load_checkpoint(self, pretrained, strict=False, logger=logger)
+        elif pretrained is None:
+            for m in self.modules():
+                if isinstance(m, nn.Conv2d):
+                    kaiming_init(m)
+                elif isinstance(m, nn.BatchNorm2d):
+                    constant_init(m, 1)
+        else:
+            raise TypeError('pretrained must be a str or None')
+
+    @auto_fp16()
+    def forward(self, x):
+        res_layer = getattr(self, f'layer{self.stage + 1}')
+        out = res_layer(x)
+        return out
+
+    def train(self, mode=True):
+        super(ResLayer, self).train(mode)
+        if self.norm_eval:
+            for m in self.modules():
+                if isinstance(m, nn.BatchNorm2d):
+                    m.eval()
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/sparse_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/sparse_roi_head.py
new file mode 100644
index 0000000..8d85ebc
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/sparse_roi_head.py
@@ -0,0 +1,311 @@
+import torch
+
+from mmdet.core import bbox2result, bbox2roi, bbox_xyxy_to_cxcywh
+from mmdet.core.bbox.samplers import PseudoSampler
+from ..builder import HEADS
+from .cascade_roi_head import CascadeRoIHead
+
+
+@HEADS.register_module()
+class SparseRoIHead(CascadeRoIHead):
+    r"""The RoIHead for `Sparse R-CNN: End-to-End Object Detection with
+    Learnable Proposals <https://arxiv.org/abs/2011.12450>`_
+
+    Args:
+        num_stages (int): Number of stage whole iterative process.
+            Defaults to 6.
+        stage_loss_weights (Tuple[float]): The loss
+            weight of each stage. By default all stages have
+            the same weight 1.
+        bbox_roi_extractor (dict): Config of box roi extractor.
+        bbox_head (dict): Config of box head.
+        train_cfg (dict, optional): Configuration information in train stage.
+            Defaults to None.
+        test_cfg (dict, optional): Configuration information in test stage.
+            Defaults to None.
+
+    """
+
+    def __init__(self,
+                 num_stages=6,
+                 stage_loss_weights=(1, 1, 1, 1, 1, 1),
+                 proposal_feature_channel=256,
+                 bbox_roi_extractor=dict(
+                     type='SingleRoIExtractor',
+                     roi_layer=dict(
+                         type='RoIAlign', output_size=7, sampling_ratio=2),
+                     out_channels=256,
+                     featmap_strides=[4, 8, 16, 32]),
+                 bbox_head=dict(
+                     type='DIIHead',
+                     num_classes=80,
+                     num_fcs=2,
+                     num_heads=8,
+                     num_cls_fcs=1,
+                     num_reg_fcs=3,
+                     feedforward_channels=2048,
+                     hidden_channels=256,
+                     dropout=0.0,
+                     roi_feat_size=7,
+                     ffn_act_cfg=dict(type='ReLU', inplace=True)),
+                 train_cfg=None,
+                 test_cfg=None):
+        assert bbox_roi_extractor is not None
+        assert bbox_head is not None
+        assert len(stage_loss_weights) == num_stages
+        self.num_stages = num_stages
+        self.stage_loss_weights = stage_loss_weights
+        self.proposal_feature_channel = proposal_feature_channel
+        super(SparseRoIHead, self).__init__(
+            num_stages,
+            stage_loss_weights,
+            bbox_roi_extractor=bbox_roi_extractor,
+            bbox_head=bbox_head,
+            train_cfg=train_cfg,
+            test_cfg=test_cfg)
+        # train_cfg would be None when run the test.py
+        if train_cfg is not None:
+            for stage in range(num_stages):
+                assert isinstance(self.bbox_sampler[stage], PseudoSampler), \
+                    'Sparse R-CNN only support `PseudoSampler`'
+
+    def _bbox_forward(self, stage, x, rois, object_feats, img_metas):
+        """Box head forward function used in both training and testing. Returns
+        all regression, classification results and a intermediate feature.
+
+        Args:
+            stage (int): The index of current stage in
+                iterative process.
+            x (List[Tensor]): List of FPN features
+            rois (Tensor): Rois in total batch. With shape (num_proposal, 5).
+                the last dimension 5 represents (img_index, x1, y1, x2, y2).
+            object_feats (Tensor): The object feature extracted from
+                the previous stage.
+            img_metas (dict): meta information of images.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of bbox head outputs,
+                Containing the following results:
+
+                    - cls_score (Tensor): The score of each class, has
+                      shape (batch_size, num_proposals, num_classes)
+                      when use focal loss or
+                      (batch_size, num_proposals, num_classes+1)
+                      otherwise.
+                    - decode_bbox_pred (Tensor): The regression results
+                      with shape (batch_size, num_proposal, 4).
+                      The last dimension 4 represents
+                      [tl_x, tl_y, br_x, br_y].
+                    - object_feats (Tensor): The object feature extracted
+                      from current stage
+                    - detach_cls_score_list (list[Tensor]): The detached
+                      classification results, length is batch_size, and
+                      each tensor has shape (num_proposal, num_classes).
+                    - detach_proposal_list (list[tensor]): The detached
+                      regression results, length is batch_size, and each
+                      tensor has shape (num_proposal, 4). The last
+                      dimension 4 represents [tl_x, tl_y, br_x, br_y].
+        """
+        num_imgs = len(img_metas)
+        bbox_roi_extractor = self.bbox_roi_extractor[stage]
+        bbox_head = self.bbox_head[stage]
+        bbox_feats = bbox_roi_extractor(x[:bbox_roi_extractor.num_inputs],
+                                        rois)
+        cls_score, bbox_pred, object_feats = bbox_head(bbox_feats,
+                                                       object_feats)
+        proposal_list = self.bbox_head[stage].refine_bboxes(
+            rois,
+            rois.new_zeros(len(rois)),  # dummy arg
+            bbox_pred.view(-1, bbox_pred.size(-1)),
+            [rois.new_zeros(object_feats.size(1)) for _ in range(num_imgs)],
+            img_metas)
+        bbox_results = dict(
+            cls_score=cls_score,
+            decode_bbox_pred=torch.cat(proposal_list),
+            object_feats=object_feats,
+            # detach then use it in label assign
+            detach_cls_score_list=[
+                cls_score[i].detach() for i in range(num_imgs)
+            ],
+            detach_proposal_list=[item.detach() for item in proposal_list])
+
+        return bbox_results
+
+    def forward_train(self,
+                      x,
+                      proposal_boxes,
+                      proposal_features,
+                      img_metas,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      imgs_whwh=None,
+                      gt_masks=None):
+        """Forward function in training stage.
+
+        Args:
+            x (list[Tensor]): list of multi-level img features.
+            proposals (Tensor): Decoded proposal bboxes, has shape
+                (batch_size, num_proposals, 4)
+            proposal_features (Tensor): Expanded proposal
+                features, has shape
+                (batch_size, num_proposals, proposal_feature_channel)
+            img_metas (list[dict]): list of image info dict where
+                each dict has: 'img_shape', 'scale_factor', 'flip',
+                and may also contain 'filename', 'ori_shape',
+                'pad_shape', and 'img_norm_cfg'. For details on the
+                values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+            imgs_whwh (Tensor): Tensor with shape (batch_size, 4),
+                    the dimension means
+                    [img_width,img_height, img_width, img_height].
+            gt_masks (None | Tensor) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components of all stage.
+        """
+
+        num_imgs = len(img_metas)
+        num_proposals = proposal_boxes.size(1)
+        imgs_whwh = imgs_whwh.repeat(1, num_proposals, 1)
+        all_stage_bbox_results = []
+        proposal_list = [proposal_boxes[i] for i in range(len(proposal_boxes))]
+        object_feats = proposal_features
+        all_stage_loss = {}
+        for stage in range(self.num_stages):
+            rois = bbox2roi(proposal_list)
+            bbox_results = self._bbox_forward(stage, x, rois, object_feats,
+                                              img_metas)
+            all_stage_bbox_results.append(bbox_results)
+            if gt_bboxes_ignore is None:
+                # TODO support ignore
+                gt_bboxes_ignore = [None for _ in range(num_imgs)]
+            sampling_results = []
+            cls_pred_list = bbox_results['detach_cls_score_list']
+            proposal_list = bbox_results['detach_proposal_list']
+            for i in range(num_imgs):
+                normalize_bbox_ccwh = bbox_xyxy_to_cxcywh(proposal_list[i] /
+                                                          imgs_whwh[i])
+                assign_result = self.bbox_assigner[stage].assign(
+                    normalize_bbox_ccwh, cls_pred_list[i], gt_bboxes[i],
+                    gt_labels[i], img_metas[i])
+                sampling_result = self.bbox_sampler[stage].sample(
+                    assign_result, proposal_list[i], gt_bboxes[i])
+                sampling_results.append(sampling_result)
+            bbox_targets = self.bbox_head[stage].get_targets(
+                sampling_results, gt_bboxes, gt_labels, self.train_cfg[stage],
+                True)
+            cls_score = bbox_results['cls_score']
+            decode_bbox_pred = bbox_results['decode_bbox_pred']
+
+            single_stage_loss = self.bbox_head[stage].loss(
+                cls_score.view(-1, cls_score.size(-1)),
+                decode_bbox_pred.view(-1, 4),
+                *bbox_targets,
+                imgs_whwh=imgs_whwh)
+            for key, value in single_stage_loss.items():
+                all_stage_loss[f'stage{stage}_{key}'] = value * \
+                                    self.stage_loss_weights[stage]
+            object_feats = bbox_results['object_feats']
+
+        return all_stage_loss
+
+    def simple_test(self,
+                    x,
+                    proposal_boxes,
+                    proposal_features,
+                    img_metas,
+                    imgs_whwh,
+                    rescale=False):
+        """Test without augmentation.
+
+        Args:
+            x (list[Tensor]): list of multi-level img features.
+            proposal_boxes (Tensor): Decoded proposal bboxes, has shape
+                (batch_size, num_proposals, 4)
+            proposal_features (Tensor): Expanded proposal
+                features, has shape
+                (batch_size, num_proposals, proposal_feature_channel)
+            img_metas (dict): meta information of images.
+            imgs_whwh (Tensor): Tensor with shape (batch_size, 4),
+                    the dimension means
+                    [img_width,img_height, img_width, img_height].
+            rescale (bool): If True, return boxes in original image
+                space. Defaults to False.
+
+        Returns:
+            bbox_results (list[tuple[np.ndarray]]): \
+                [[cls1_det, cls2_det, ...], ...]. \
+                The outer list indicates images, and the inner \
+                list indicates per-class detected bboxes. The \
+                np.ndarray has shape (num_det, 5) and the last \
+                dimension 5 represents (x1, y1, x2, y2, score).
+        """
+        assert self.with_bbox, 'Bbox head must be implemented.'
+        # Decode initial proposals
+        num_imgs = len(img_metas)
+        proposal_list = [proposal_boxes[i] for i in range(num_imgs)]
+        object_feats = proposal_features
+        for stage in range(self.num_stages):
+            rois = bbox2roi(proposal_list)
+            bbox_results = self._bbox_forward(stage, x, rois, object_feats,
+                                              img_metas)
+            object_feats = bbox_results['object_feats']
+            cls_score = bbox_results['cls_score']
+            proposal_list = bbox_results['detach_proposal_list']
+
+        num_classes = self.bbox_head[-1].num_classes
+        det_bboxes = []
+        det_labels = []
+
+        if self.bbox_head[-1].loss_cls.use_sigmoid:
+            cls_score = cls_score.sigmoid()
+        else:
+            cls_score = cls_score.softmax(-1)[..., :-1]
+
+        for img_id in range(num_imgs):
+            cls_score_per_img = cls_score[img_id]
+            scores_per_img, topk_indices = cls_score_per_img.flatten(
+                0, 1).topk(
+                    self.test_cfg.max_per_img, sorted=False)
+            labels_per_img = topk_indices % num_classes
+            bbox_pred_per_img = proposal_list[img_id][topk_indices //
+                                                      num_classes]
+            if rescale:
+                scale_factor = img_metas[img_id]['scale_factor']
+                bbox_pred_per_img /= bbox_pred_per_img.new_tensor(scale_factor)
+            det_bboxes.append(
+                torch.cat([bbox_pred_per_img, scores_per_img[:, None]], dim=1))
+            det_labels.append(labels_per_img)
+
+        bbox_results = [
+            bbox2result(det_bboxes[i], det_labels[i], num_classes)
+            for i in range(num_imgs)
+        ]
+
+        return bbox_results
+
+    def aug_test(self, features, proposal_list, img_metas, rescale=False):
+        raise NotImplementedError('Sparse R-CNN does not support `aug_test`')
+
+    def forward_dummy(self, x, proposal_boxes, proposal_features, img_metas):
+        """Dummy forward function when do the flops computing."""
+        all_stage_bbox_results = []
+        proposal_list = [proposal_boxes[i] for i in range(len(proposal_boxes))]
+        object_feats = proposal_features
+        if self.with_bbox:
+            for stage in range(self.num_stages):
+                rois = bbox2roi(proposal_list)
+                bbox_results = self._bbox_forward(stage, x, rois, object_feats,
+                                                  img_metas)
+
+                all_stage_bbox_results.append(bbox_results)
+                proposal_list = bbox_results['detach_proposal_list']
+                object_feats = bbox_results['object_feats']
+        return all_stage_bbox_results
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/standard_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/standard_roi_head.py
new file mode 100644
index 0000000..c530f2a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/standard_roi_head.py
@@ -0,0 +1,295 @@
+import torch
+
+from mmdet.core import bbox2result, bbox2roi, build_assigner, build_sampler
+from ..builder import HEADS, build_head, build_roi_extractor
+from .base_roi_head import BaseRoIHead
+from .test_mixins import BBoxTestMixin, MaskTestMixin
+
+
+@HEADS.register_module()
+class StandardRoIHead(BaseRoIHead, BBoxTestMixin, MaskTestMixin):
+    """Simplest base roi head including one bbox head and one mask head."""
+
+    def init_assigner_sampler(self):
+        """Initialize assigner and sampler."""
+        self.bbox_assigner = None
+        self.bbox_sampler = None
+        if self.train_cfg:
+            self.bbox_assigner = build_assigner(self.train_cfg.assigner)
+            self.bbox_sampler = build_sampler(
+                self.train_cfg.sampler, context=self)
+
+    def init_bbox_head(self, bbox_roi_extractor, bbox_head):
+        """Initialize ``bbox_head``"""
+        self.bbox_roi_extractor = build_roi_extractor(bbox_roi_extractor)
+        self.bbox_head = build_head(bbox_head)
+
+    def init_mask_head(self, mask_roi_extractor, mask_head):
+        """Initialize ``mask_head``"""
+        if mask_roi_extractor is not None:
+            self.mask_roi_extractor = build_roi_extractor(mask_roi_extractor)
+            self.share_roi_extractor = False
+        else:
+            self.share_roi_extractor = True
+            self.mask_roi_extractor = self.bbox_roi_extractor
+        self.mask_head = build_head(mask_head)
+
+    def init_weights(self, pretrained):
+        """Initialize the weights in head.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+        if self.with_shared_head:
+            self.shared_head.init_weights(pretrained=pretrained)
+        if self.with_bbox:
+            self.bbox_roi_extractor.init_weights()
+            self.bbox_head.init_weights()
+        if self.with_mask:
+            self.mask_head.init_weights()
+            if not self.share_roi_extractor:
+                self.mask_roi_extractor.init_weights()
+
+    def forward_dummy(self, x, proposals):
+        """Dummy forward function."""
+        # bbox head
+        outs = ()
+        rois = bbox2roi([proposals])
+        if self.with_bbox:
+            bbox_results = self._bbox_forward(x, rois)
+            outs = outs + (bbox_results['cls_score'],
+                           bbox_results['bbox_pred'])
+        # mask head
+        if self.with_mask:
+            mask_rois = rois[:100]
+            mask_results = self._mask_forward(x, mask_rois)
+            outs = outs + (mask_results['mask_pred'], )
+        return outs
+
+    def forward_train(self,
+                      x,
+                      img_metas,
+                      proposal_list,
+                      gt_bboxes,
+                      gt_labels,
+                      gt_bboxes_ignore=None,
+                      gt_masks=None):
+        """
+        Args:
+            x (list[Tensor]): list of multi-level img features.
+            img_metas (list[dict]): list of image info dict where each dict
+                has: 'img_shape', 'scale_factor', 'flip', and may also contain
+                'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'.
+                For details on the values of these keys see
+                `mmdet/datasets/pipelines/formatting.py:Collect`.
+            proposals (list[Tensors]): list of region proposals.
+            gt_bboxes (list[Tensor]): Ground truth bboxes for each image with
+                shape (num_gts, 4) in [tl_x, tl_y, br_x, br_y] format.
+            gt_labels (list[Tensor]): class indices corresponding to each box
+            gt_bboxes_ignore (None | list[Tensor]): specify which bounding
+                boxes can be ignored when computing the loss.
+            gt_masks (None | Tensor) : true segmentation masks for each box
+                used if the architecture supports a segmentation task.
+
+        Returns:
+            dict[str, Tensor]: a dictionary of loss components
+        """
+        # assign gts and sample proposals
+        if self.with_bbox or self.with_mask:
+            num_imgs = len(img_metas)
+            if gt_bboxes_ignore is None:
+                gt_bboxes_ignore = [None for _ in range(num_imgs)]
+            sampling_results = []
+            for i in range(num_imgs):
+                assign_result = self.bbox_assigner.assign(
+                    proposal_list[i], gt_bboxes[i], gt_bboxes_ignore[i],
+                    gt_labels[i])
+                sampling_result = self.bbox_sampler.sample(
+                    assign_result,
+                    proposal_list[i],
+                    gt_bboxes[i],
+                    gt_labels[i],
+                    feats=[lvl_feat[i][None] for lvl_feat in x])
+                sampling_results.append(sampling_result)
+
+        losses = dict()
+        # bbox head forward and loss
+        if self.with_bbox:
+            bbox_results = self._bbox_forward_train(x, sampling_results,
+                                                    gt_bboxes, gt_labels,
+                                                    img_metas)
+            losses.update(bbox_results['loss_bbox'])
+
+        # mask head forward and loss
+        if self.with_mask:
+            mask_results = self._mask_forward_train(x, sampling_results,
+                                                    bbox_results['bbox_feats'],
+                                                    gt_masks, img_metas)
+            losses.update(mask_results['loss_mask'])
+
+        return losses
+
+    def _bbox_forward(self, x, rois):
+        """Box head forward function used in both training and testing."""
+        # TODO: a more flexible way to decide which feature maps to use
+        bbox_feats = self.bbox_roi_extractor(
+            x[:self.bbox_roi_extractor.num_inputs], rois)
+        if self.with_shared_head:
+            bbox_feats = self.shared_head(bbox_feats)
+        cls_score, bbox_pred = self.bbox_head(bbox_feats)
+
+        bbox_results = dict(
+            cls_score=cls_score, bbox_pred=bbox_pred, bbox_feats=bbox_feats)
+        return bbox_results
+
+    def _bbox_forward_train(self, x, sampling_results, gt_bboxes, gt_labels,
+                            img_metas):
+        """Run forward function and calculate loss for box head in training."""
+        rois = bbox2roi([res.bboxes for res in sampling_results])
+        bbox_results = self._bbox_forward(x, rois)
+
+        bbox_targets = self.bbox_head.get_targets(sampling_results, gt_bboxes,
+                                                  gt_labels, self.train_cfg)
+        loss_bbox = self.bbox_head.loss(bbox_results['cls_score'],
+                                        bbox_results['bbox_pred'], rois,
+                                        *bbox_targets)
+
+        bbox_results.update(loss_bbox=loss_bbox)
+        return bbox_results
+
+    def _mask_forward_train(self, x, sampling_results, bbox_feats, gt_masks,
+                            img_metas):
+        """Run forward function and calculate loss for mask head in
+        training."""
+        if not self.share_roi_extractor:
+            pos_rois = bbox2roi([res.pos_bboxes for res in sampling_results])
+            mask_results = self._mask_forward(x, pos_rois)
+        else:
+            pos_inds = []
+            device = bbox_feats.device
+            for res in sampling_results:
+                pos_inds.append(
+                    torch.ones(
+                        res.pos_bboxes.shape[0],
+                        device=device,
+                        dtype=torch.uint8))
+                pos_inds.append(
+                    torch.zeros(
+                        res.neg_bboxes.shape[0],
+                        device=device,
+                        dtype=torch.uint8))
+            pos_inds = torch.cat(pos_inds)
+
+            mask_results = self._mask_forward(
+                x, pos_inds=pos_inds, bbox_feats=bbox_feats)
+
+        mask_targets = self.mask_head.get_targets(sampling_results, gt_masks,
+                                                  self.train_cfg)
+        pos_labels = torch.cat([res.pos_gt_labels for res in sampling_results])
+        loss_mask = self.mask_head.loss(mask_results['mask_pred'],
+                                        mask_targets, pos_labels)
+
+        mask_results.update(loss_mask=loss_mask, mask_targets=mask_targets)
+        return mask_results
+
+    def _mask_forward(self, x, rois=None, pos_inds=None, bbox_feats=None):
+        """Mask head forward function used in both training and testing."""
+        assert ((rois is not None) ^
+                (pos_inds is not None and bbox_feats is not None))
+        if rois is not None:
+            mask_feats = self.mask_roi_extractor(
+                x[:self.mask_roi_extractor.num_inputs], rois)
+            if self.with_shared_head:
+                mask_feats = self.shared_head(mask_feats)
+        else:
+            assert bbox_feats is not None
+            mask_feats = bbox_feats[pos_inds]
+
+        mask_pred = self.mask_head(mask_feats)
+        mask_results = dict(mask_pred=mask_pred, mask_feats=mask_feats)
+        return mask_results
+
+    async def async_simple_test(self,
+                                x,
+                                proposal_list,
+                                img_metas,
+                                proposals=None,
+                                rescale=False):
+        """Async test without augmentation."""
+        assert self.with_bbox, 'Bbox head must be implemented.'
+
+        det_bboxes, det_labels = await self.async_test_bboxes(
+            x, img_metas, proposal_list, self.test_cfg, rescale=rescale)
+        bbox_results = bbox2result(det_bboxes, det_labels,
+                                   self.bbox_head.num_classes)
+        if not self.with_mask:
+            return bbox_results
+        else:
+            segm_results = await self.async_test_mask(
+                x,
+                img_metas,
+                det_bboxes,
+                det_labels,
+                rescale=rescale,
+                mask_test_cfg=self.test_cfg.get('mask'))
+            return bbox_results, segm_results
+
+    def simple_test(self,
+                    x,
+                    proposal_list,
+                    img_metas,
+                    proposals=None,
+                    rescale=False):
+        """Test without augmentation."""
+        assert self.with_bbox, 'Bbox head must be implemented.'
+
+        det_bboxes, det_labels = self.simple_test_bboxes(
+            x, img_metas, proposal_list, self.test_cfg, rescale=rescale)
+        if torch.onnx.is_in_onnx_export():
+            if self.with_mask:
+                segm_results = self.simple_test_mask(
+                    x, img_metas, det_bboxes, det_labels, rescale=rescale)
+                return det_bboxes, det_labels, segm_results
+            else:
+                return det_bboxes, det_labels
+
+        bbox_results = [
+            bbox2result(det_bboxes[i], det_labels[i],
+                        self.bbox_head.num_classes)
+            for i in range(len(det_bboxes))
+        ]
+
+        if not self.with_mask:
+            return bbox_results
+        else:
+            segm_results = self.simple_test_mask(
+                x, img_metas, det_bboxes, det_labels, rescale=rescale)
+            return list(zip(bbox_results, segm_results))
+
+    def aug_test(self, x, proposal_list, img_metas, rescale=False):
+        """Test with augmentations.
+
+        If rescale is False, then returned bboxes and masks will fit the scale
+        of imgs[0].
+        """
+        det_bboxes, det_labels = self.aug_test_bboxes(x, img_metas,
+                                                      proposal_list,
+                                                      self.test_cfg)
+
+        if rescale:
+            _det_bboxes = det_bboxes
+        else:
+            _det_bboxes = det_bboxes.clone()
+            _det_bboxes[:, :4] *= det_bboxes.new_tensor(
+                img_metas[0][0]['scale_factor'])
+        bbox_results = bbox2result(_det_bboxes, det_labels,
+                                   self.bbox_head.num_classes)
+
+        # det_bboxes always keep the original scale
+        if self.with_mask:
+            segm_results = self.aug_test_mask(x, img_metas, det_bboxes,
+                                              det_labels)
+            return [(bbox_results, segm_results)]
+        else:
+            return [bbox_results]
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/test_mixins.py b/python/autoCut/models/swin/mmdet/models/roi_heads/test_mixins.py
new file mode 100644
index 0000000..78a092a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/test_mixins.py
@@ -0,0 +1,348 @@
+import logging
+import sys
+
+import torch
+
+from mmdet.core import (bbox2roi, bbox_mapping, merge_aug_bboxes,
+                        merge_aug_masks, multiclass_nms)
+
+logger = logging.getLogger(__name__)
+
+if sys.version_info >= (3, 7):
+    from mmdet.utils.contextmanagers import completed
+
+
+class BBoxTestMixin(object):
+
+    if sys.version_info >= (3, 7):
+
+        async def async_test_bboxes(self,
+                                    x,
+                                    img_metas,
+                                    proposals,
+                                    rcnn_test_cfg,
+                                    rescale=False,
+                                    bbox_semaphore=None,
+                                    global_lock=None):
+            """Asynchronized test for box head without augmentation."""
+            rois = bbox2roi(proposals)
+            roi_feats = self.bbox_roi_extractor(
+                x[:len(self.bbox_roi_extractor.featmap_strides)], rois)
+            if self.with_shared_head:
+                roi_feats = self.shared_head(roi_feats)
+            sleep_interval = rcnn_test_cfg.get('async_sleep_interval', 0.017)
+
+            async with completed(
+                    __name__, 'bbox_head_forward',
+                    sleep_interval=sleep_interval):
+                cls_score, bbox_pred = self.bbox_head(roi_feats)
+
+            img_shape = img_metas[0]['img_shape']
+            scale_factor = img_metas[0]['scale_factor']
+            det_bboxes, det_labels = self.bbox_head.get_bboxes(
+                rois,
+                cls_score,
+                bbox_pred,
+                img_shape,
+                scale_factor,
+                rescale=rescale,
+                cfg=rcnn_test_cfg)
+            return det_bboxes, det_labels
+
+    def simple_test_bboxes(self,
+                           x,
+                           img_metas,
+                           proposals,
+                           rcnn_test_cfg,
+                           rescale=False):
+        """Test only det bboxes without augmentation.
+
+        Args:
+            x (tuple[Tensor]): Feature maps of all scale level.
+            img_metas (list[dict]): Image meta info.
+            proposals (Tensor or List[Tensor]): Region proposals.
+            rcnn_test_cfg (obj:`ConfigDict`): `test_cfg` of R-CNN.
+            rescale (bool): If True, return boxes in original image space.
+                Default: False.
+
+        Returns:
+            tuple[list[Tensor], list[Tensor]]: The first list contains
+                the boxes of the corresponding image in a batch, each
+                tensor has the shape (num_boxes, 5) and last dimension
+                5 represent (tl_x, tl_y, br_x, br_y, score). Each Tensor
+                in the second list is the labels with shape (num_boxes, ).
+                The length of both lists should be equal to batch_size.
+        """
+        # get origin input shape to support onnx dynamic input shape
+        if torch.onnx.is_in_onnx_export():
+            assert len(
+                img_metas
+            ) == 1, 'Only support one input image while in exporting to ONNX'
+            img_shapes = img_metas[0]['img_shape_for_onnx']
+        else:
+            img_shapes = tuple(meta['img_shape'] for meta in img_metas)
+        scale_factors = tuple(meta['scale_factor'] for meta in img_metas)
+
+        # The length of proposals of different batches may be different.
+        # In order to form a batch, a padding operation is required.
+        if isinstance(proposals, list):
+            # padding to form a batch
+            max_size = max([proposal.size(0) for proposal in proposals])
+            for i, proposal in enumerate(proposals):
+                supplement = proposal.new_full(
+                    (max_size - proposal.size(0), proposal.size(1)), 0)
+                proposals[i] = torch.cat((supplement, proposal), dim=0)
+            rois = torch.stack(proposals, dim=0)
+        else:
+            rois = proposals
+
+        batch_index = torch.arange(
+            rois.size(0), device=rois.device).float().view(-1, 1, 1).expand(
+                rois.size(0), rois.size(1), 1)
+        rois = torch.cat([batch_index, rois[..., :4]], dim=-1)
+        batch_size = rois.shape[0]
+        num_proposals_per_img = rois.shape[1]
+
+        # Eliminate the batch dimension
+        rois = rois.view(-1, 5)
+        bbox_results = self._bbox_forward(x, rois)
+        cls_score = bbox_results['cls_score']
+        bbox_pred = bbox_results['bbox_pred']
+
+        # Recover the batch dimension
+        rois = rois.reshape(batch_size, num_proposals_per_img, -1)
+        cls_score = cls_score.reshape(batch_size, num_proposals_per_img, -1)
+
+        if not torch.onnx.is_in_onnx_export():
+            # remove padding
+            supplement_mask = rois[..., -1] == 0
+            cls_score[supplement_mask, :] = 0
+
+        # bbox_pred would be None in some detector when with_reg is False,
+        # e.g. Grid R-CNN.
+        if bbox_pred is not None:
+            # the bbox prediction of some detectors like SABL is not Tensor
+            if isinstance(bbox_pred, torch.Tensor):
+                bbox_pred = bbox_pred.reshape(batch_size,
+                                              num_proposals_per_img, -1)
+                if not torch.onnx.is_in_onnx_export():
+                    bbox_pred[supplement_mask, :] = 0
+            else:
+                # TODO: Looking forward to a better way
+                # For SABL
+                bbox_preds = self.bbox_head.bbox_pred_split(
+                    bbox_pred, num_proposals_per_img)
+                # apply bbox post-processing to each image individually
+                det_bboxes = []
+                det_labels = []
+                for i in range(len(proposals)):
+                    # remove padding
+                    supplement_mask = proposals[i][..., -1] == 0
+                    for bbox in bbox_preds[i]:
+                        bbox[supplement_mask] = 0
+                    det_bbox, det_label = self.bbox_head.get_bboxes(
+                        rois[i],
+                        cls_score[i],
+                        bbox_preds[i],
+                        img_shapes[i],
+                        scale_factors[i],
+                        rescale=rescale,
+                        cfg=rcnn_test_cfg)
+                    det_bboxes.append(det_bbox)
+                    det_labels.append(det_label)
+                return det_bboxes, det_labels
+        else:
+            bbox_pred = None
+
+        return self.bbox_head.get_bboxes(
+            rois,
+            cls_score,
+            bbox_pred,
+            img_shapes,
+            scale_factors,
+            rescale=rescale,
+            cfg=rcnn_test_cfg)
+
+    def aug_test_bboxes(self, feats, img_metas, proposal_list, rcnn_test_cfg):
+        """Test det bboxes with test time augmentation."""
+        aug_bboxes = []
+        aug_scores = []
+        for x, img_meta in zip(feats, img_metas):
+            # only one image in the batch
+            img_shape = img_meta[0]['img_shape']
+            scale_factor = img_meta[0]['scale_factor']
+            flip = img_meta[0]['flip']
+            flip_direction = img_meta[0]['flip_direction']
+            # TODO more flexible
+            proposals = bbox_mapping(proposal_list[0][:, :4], img_shape,
+                                     scale_factor, flip, flip_direction)
+            rois = bbox2roi([proposals])
+            bbox_results = self._bbox_forward(x, rois)
+            bboxes, scores = self.bbox_head.get_bboxes(
+                rois,
+                bbox_results['cls_score'],
+                bbox_results['bbox_pred'],
+                img_shape,
+                scale_factor,
+                rescale=False,
+                cfg=None)
+            aug_bboxes.append(bboxes)
+            aug_scores.append(scores)
+        # after merging, bboxes will be rescaled to the original image size
+        merged_bboxes, merged_scores = merge_aug_bboxes(
+            aug_bboxes, aug_scores, img_metas, rcnn_test_cfg)
+        det_bboxes, det_labels = multiclass_nms(merged_bboxes, merged_scores,
+                                                rcnn_test_cfg.score_thr,
+                                                rcnn_test_cfg.nms,
+                                                rcnn_test_cfg.max_per_img)
+        return det_bboxes, det_labels
+
+
+class MaskTestMixin(object):
+
+    if sys.version_info >= (3, 7):
+
+        async def async_test_mask(self,
+                                  x,
+                                  img_metas,
+                                  det_bboxes,
+                                  det_labels,
+                                  rescale=False,
+                                  mask_test_cfg=None):
+            """Asynchronized test for mask head without augmentation."""
+            # image shape of the first image in the batch (only one)
+            ori_shape = img_metas[0]['ori_shape']
+            scale_factor = img_metas[0]['scale_factor']
+            if det_bboxes.shape[0] == 0:
+                segm_result = [[] for _ in range(self.mask_head.num_classes)]
+            else:
+                if rescale and not isinstance(scale_factor,
+                                              (float, torch.Tensor)):
+                    scale_factor = det_bboxes.new_tensor(scale_factor)
+                _bboxes = (
+                    det_bboxes[:, :4] *
+                    scale_factor if rescale else det_bboxes)
+                mask_rois = bbox2roi([_bboxes])
+                mask_feats = self.mask_roi_extractor(
+                    x[:len(self.mask_roi_extractor.featmap_strides)],
+                    mask_rois)
+
+                if self.with_shared_head:
+                    mask_feats = self.shared_head(mask_feats)
+                if mask_test_cfg and mask_test_cfg.get('async_sleep_interval'):
+                    sleep_interval = mask_test_cfg['async_sleep_interval']
+                else:
+                    sleep_interval = 0.035
+                async with completed(
+                        __name__,
+                        'mask_head_forward',
+                        sleep_interval=sleep_interval):
+                    mask_pred = self.mask_head(mask_feats)
+                segm_result = self.mask_head.get_seg_masks(
+                    mask_pred, _bboxes, det_labels, self.test_cfg, ori_shape,
+                    scale_factor, rescale)
+            return segm_result
+
+    def simple_test_mask(self,
+                         x,
+                         img_metas,
+                         det_bboxes,
+                         det_labels,
+                         rescale=False):
+        """Simple test for mask head without augmentation."""
+        # image shapes of images in the batch
+        ori_shapes = tuple(meta['ori_shape'] for meta in img_metas)
+        scale_factors = tuple(meta['scale_factor'] for meta in img_metas)
+
+        # The length of proposals of different batches may be different.
+        # In order to form a batch, a padding operation is required.
+        if isinstance(det_bboxes, list):
+            # padding to form a batch
+            max_size = max([bboxes.size(0) for bboxes in det_bboxes])
+            for i, (bbox, label) in enumerate(zip(det_bboxes, det_labels)):
+                supplement_bbox = bbox.new_full(
+                    (max_size - bbox.size(0), bbox.size(1)), 0)
+                supplement_label = label.new_full((max_size - label.size(0), ),
+                                                  0)
+                det_bboxes[i] = torch.cat((supplement_bbox, bbox), dim=0)
+                det_labels[i] = torch.cat((supplement_label, label), dim=0)
+            det_bboxes = torch.stack(det_bboxes, dim=0)
+            det_labels = torch.stack(det_labels, dim=0)
+
+        batch_size = det_bboxes.size(0)
+        num_proposals_per_img = det_bboxes.shape[1]
+
+        # if det_bboxes is rescaled to the original image size, we need to
+        # rescale it back to the testing scale to obtain RoIs.
+        det_bboxes = det_bboxes[..., :4]
+        if rescale:
+            if not isinstance(scale_factors[0], float):
+                scale_factors = det_bboxes.new_tensor(scale_factors)
+            det_bboxes = det_bboxes * scale_factors.unsqueeze(1)
+
+        batch_index = torch.arange(
+            det_bboxes.size(0), device=det_bboxes.device).float().view(
+                -1, 1, 1).expand(det_bboxes.size(0), det_bboxes.size(1), 1)
+        mask_rois = torch.cat([batch_index, det_bboxes], dim=-1)
+        mask_rois = mask_rois.view(-1, 5)
+        mask_results = self._mask_forward(x, mask_rois)
+        mask_pred = mask_results['mask_pred']
+
+        # Recover the batch dimension
+        mask_preds = mask_pred.reshape(batch_size, num_proposals_per_img,
+                                       *mask_pred.shape[1:])
+
+        # apply mask post-processing to each image individually
+        segm_results = []
+        for i in range(batch_size):
+            mask_pred = mask_preds[i]
+            det_bbox = det_bboxes[i]
+            det_label = det_labels[i]
+
+            # remove padding
+            supplement_mask = det_bbox[..., -1] != 0
+            mask_pred = mask_pred[supplement_mask]
+            det_bbox = det_bbox[supplement_mask]
+            det_label = det_label[supplement_mask]
+
+            if det_label.shape[0] == 0:
+                segm_results.append([[]
+                                     for _ in range(self.mask_head.num_classes)
+                                     ])
+            else:
+                segm_result = self.mask_head.get_seg_masks(
+                    mask_pred, det_bbox, det_label, self.test_cfg,
+                    ori_shapes[i], scale_factors[i], rescale)
+                segm_results.append(segm_result)
+        return segm_results
+
+    def aug_test_mask(self, feats, img_metas, det_bboxes, det_labels):
+        """Test for mask head with test time augmentation."""
+        if det_bboxes.shape[0] == 0:
+            segm_result = [[] for _ in range(self.mask_head.num_classes)]
+        else:
+            aug_masks = []
+            for x, img_meta in zip(feats, img_metas):
+                img_shape = img_meta[0]['img_shape']
+                scale_factor = img_meta[0]['scale_factor']
+                flip = img_meta[0]['flip']
+                flip_direction = img_meta[0]['flip_direction']
+                _bboxes = bbox_mapping(det_bboxes[:, :4], img_shape,
+                                       scale_factor, flip, flip_direction)
+                mask_rois = bbox2roi([_bboxes])
+                mask_results = self._mask_forward(x, mask_rois)
+                # convert to numpy array to save memory
+                aug_masks.append(
+                    mask_results['mask_pred'].sigmoid().cpu().numpy())
+            merged_masks = merge_aug_masks(aug_masks, img_metas, self.test_cfg)
+
+            ori_shape = img_metas[0][0]['ori_shape']
+            segm_result = self.mask_head.get_seg_masks(
+                merged_masks,
+                det_bboxes,
+                det_labels,
+                self.test_cfg,
+                ori_shape,
+                scale_factor=1.0,
+                rescale=False)
+        return segm_result
diff --git a/python/autoCut/models/swin/mmdet/models/roi_heads/trident_roi_head.py b/python/autoCut/models/swin/mmdet/models/roi_heads/trident_roi_head.py
new file mode 100644
index 0000000..245569e
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/roi_heads/trident_roi_head.py
@@ -0,0 +1,119 @@
+import torch
+from mmcv.ops import batched_nms
+
+from mmdet.core import (bbox2result, bbox2roi, bbox_mapping, merge_aug_bboxes,
+                        multiclass_nms)
+from mmdet.models.roi_heads.standard_roi_head import StandardRoIHead
+from ..builder import HEADS
+
+
+@HEADS.register_module()
+class TridentRoIHead(StandardRoIHead):
+    """Trident roi head.
+
+    Args:
+        num_branch (int): Number of branches in TridentNet.
+        test_branch_idx (int): In inference, all 3 branches will be used
+            if `test_branch_idx==-1`, otherwise only branch with index
+            `test_branch_idx` will be used.
+    """
+
+    def __init__(self, num_branch, test_branch_idx, **kwargs):
+        self.num_branch = num_branch
+        self.test_branch_idx = test_branch_idx
+        super(TridentRoIHead, self).__init__(**kwargs)
+
+    def merge_trident_bboxes(self, trident_det_bboxes, trident_det_labels):
+        """Merge bbox predictions of each branch."""
+        if trident_det_bboxes.numel() == 0:
+            det_bboxes = trident_det_bboxes.new_zeros((0, 5))
+            det_labels = trident_det_bboxes.new_zeros((0, ), dtype=torch.long)
+        else:
+            nms_bboxes = trident_det_bboxes[:, :4]
+            nms_scores = trident_det_bboxes[:, 4].contiguous()
+            nms_inds = trident_det_labels
+            nms_cfg = self.test_cfg['nms']
+            det_bboxes, keep = batched_nms(nms_bboxes, nms_scores, nms_inds,
+                                           nms_cfg)
+            det_labels = trident_det_labels[keep]
+            if self.test_cfg['max_per_img'] > 0:
+                det_labels = det_labels[:self.test_cfg['max_per_img']]
+                det_bboxes = det_bboxes[:self.test_cfg['max_per_img']]
+
+        return det_bboxes, det_labels
+
+    def simple_test(self,
+                    x,
+                    proposal_list,
+                    img_metas,
+                    proposals=None,
+                    rescale=False):
+        """Test without augmentation as follows:
+
+        1. Compute prediction bbox and label per branch.
+        2. Merge predictions of each branch according to scores of
+           bboxes, i.e., bboxes with higher score are kept to give
+           top-k prediction.
+        """
+        assert self.with_bbox, 'Bbox head must be implemented.'
+        det_bboxes_list, det_labels_list = self.simple_test_bboxes(
+            x, img_metas, proposal_list, self.test_cfg, rescale=rescale)
+        num_branch = self.num_branch if self.test_branch_idx == -1 else 1
+        for _ in range(len(det_bboxes_list)):
+            if det_bboxes_list[_].shape[0] == 0:
+                det_bboxes_list[_] = det_bboxes_list[_].new_empty((0, 5))
+        det_bboxes, det_labels = [], []
+        for i in range(len(img_metas) // num_branch):
+            det_result = self.merge_trident_bboxes(
+                torch.cat(det_bboxes_list[i * num_branch:(i + 1) *
+                                          num_branch]),
+                torch.cat(det_labels_list[i * num_branch:(i + 1) *
+                                          num_branch]))
+            det_bboxes.append(det_result[0])
+            det_labels.append(det_result[1])
+
+        bbox_results = [
+            bbox2result(det_bboxes[i], det_labels[i],
+                        self.bbox_head.num_classes)
+            for i in range(len(det_bboxes))
+        ]
+        return bbox_results
+
+    def aug_test_bboxes(self, feats, img_metas, proposal_list, rcnn_test_cfg):
+        """Test det bboxes with test time augmentation."""
+        aug_bboxes = []
+        aug_scores = []
+        for x, img_meta in zip(feats, img_metas):
+            # only one image in the batch
+            img_shape = img_meta[0]['img_shape']
+            scale_factor = img_meta[0]['scale_factor']
+            flip = img_meta[0]['flip']
+            flip_direction = img_meta[0]['flip_direction']
+
+            trident_bboxes, trident_scores = [], []
+            for branch_idx in range(len(proposal_list)):
+                proposals = bbox_mapping(proposal_list[0][:, :4], img_shape,
+                                         scale_factor, flip, flip_direction)
+                rois = bbox2roi([proposals])
+                bbox_results = self._bbox_forward(x, rois)
+                bboxes, scores = self.bbox_head.get_bboxes(
+                    rois,
+                    bbox_results['cls_score'],
+                    bbox_results['bbox_pred'],
+                    img_shape,
+                    scale_factor,
+                    rescale=False,
+                    cfg=None)
+                trident_bboxes.append(bboxes)
+                trident_scores.append(scores)
+
+            aug_bboxes.append(torch.cat(trident_bboxes, 0))
+            aug_scores.append(torch.cat(trident_scores, 0))
+        # after merging, bboxes will be rescaled to the original image size
+        merged_bboxes, merged_scores = merge_aug_bboxes(
+            aug_bboxes, aug_scores, img_metas, rcnn_test_cfg)
+        det_bboxes, det_labels = multiclass_nms(merged_bboxes, merged_scores,
+                                                rcnn_test_cfg.score_thr,
+                                                rcnn_test_cfg.nms,
+                                                rcnn_test_cfg.max_per_img)
+        return det_bboxes, det_labels
diff --git a/python/autoCut/models/swin/mmdet/models/utils/__init__.py b/python/autoCut/models/swin/mmdet/models/utils/__init__.py
new file mode 100644
index 0000000..5165b22
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/utils/__init__.py
@@ -0,0 +1,16 @@
+from .builder import build_positional_encoding, build_transformer
+from .gaussian_target import gaussian_radius, gen_gaussian_target
+from .positional_encoding import (LearnedPositionalEncoding,
+                                  SinePositionalEncoding)
+from .res_layer import ResLayer, SimplifiedBasicBlock
+from .transformer import (FFN, DynamicConv, MultiheadAttention, Transformer,
+                          TransformerDecoder, TransformerDecoderLayer,
+                          TransformerEncoder, TransformerEncoderLayer)
+
+__all__ = [
+    'ResLayer', 'gaussian_radius', 'gen_gaussian_target', 'MultiheadAttention',
+    'FFN', 'TransformerEncoderLayer', 'TransformerEncoder',
+    'TransformerDecoderLayer', 'TransformerDecoder', 'Transformer',
+    'build_transformer', 'build_positional_encoding', 'SinePositionalEncoding',
+    'LearnedPositionalEncoding', 'DynamicConv', 'SimplifiedBasicBlock'
+]
diff --git a/python/autoCut/models/swin/mmdet/models/utils/builder.py b/python/autoCut/models/swin/mmdet/models/utils/builder.py
new file mode 100644
index 0000000..f362d1c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/utils/builder.py
@@ -0,0 +1,14 @@
+from mmcv.utils import Registry, build_from_cfg
+
+TRANSFORMER = Registry('Transformer')
+POSITIONAL_ENCODING = Registry('Position encoding')
+
+
+def build_transformer(cfg, default_args=None):
+    """Builder for Transformer."""
+    return build_from_cfg(cfg, TRANSFORMER, default_args)
+
+
+def build_positional_encoding(cfg, default_args=None):
+    """Builder for Position Encoding."""
+    return build_from_cfg(cfg, POSITIONAL_ENCODING, default_args)
diff --git a/python/autoCut/models/swin/mmdet/models/utils/gaussian_target.py b/python/autoCut/models/swin/mmdet/models/utils/gaussian_target.py
new file mode 100644
index 0000000..7bb7160
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/utils/gaussian_target.py
@@ -0,0 +1,185 @@
+from math import sqrt
+
+import torch
+
+
+def gaussian2D(radius, sigma=1, dtype=torch.float32, device='cpu'):
+    """Generate 2D gaussian kernel.
+
+    Args:
+        radius (int): Radius of gaussian kernel.
+        sigma (int): Sigma of gaussian function. Default: 1.
+        dtype (torch.dtype): Dtype of gaussian tensor. Default: torch.float32.
+        device (str): Device of gaussian tensor. Default: 'cpu'.
+
+    Returns:
+        h (Tensor): Gaussian kernel with a
+            ``(2 * radius + 1) * (2 * radius + 1)`` shape.
+    """
+    x = torch.arange(
+        -radius, radius + 1, dtype=dtype, device=device).view(1, -1)
+    y = torch.arange(
+        -radius, radius + 1, dtype=dtype, device=device).view(-1, 1)
+
+    h = (-(x * x + y * y) / (2 * sigma * sigma)).exp()
+
+    h[h < torch.finfo(h.dtype).eps * h.max()] = 0
+    return h
+
+
+def gen_gaussian_target(heatmap, center, radius, k=1):
+    """Generate 2D gaussian heatmap.
+
+    Args:
+        heatmap (Tensor): Input heatmap, the gaussian kernel will cover on
+            it and maintain the max value.
+        center (list[int]): Coord of gaussian kernel's center.
+        radius (int): Radius of gaussian kernel.
+        k (int): Coefficient of gaussian kernel. Default: 1.
+
+    Returns:
+        out_heatmap (Tensor): Updated heatmap covered by gaussian kernel.
+    """
+    diameter = 2 * radius + 1
+    gaussian_kernel = gaussian2D(
+        radius, sigma=diameter / 6, dtype=heatmap.dtype, device=heatmap.device)
+
+    x, y = center
+
+    height, width = heatmap.shape[:2]
+
+    left, right = min(x, radius), min(width - x, radius + 1)
+    top, bottom = min(y, radius), min(height - y, radius + 1)
+
+    masked_heatmap = heatmap[y - top:y + bottom, x - left:x + right]
+    masked_gaussian = gaussian_kernel[radius - top:radius + bottom,
+                                      radius - left:radius + right]
+    out_heatmap = heatmap
+    torch.max(
+        masked_heatmap,
+        masked_gaussian * k,
+        out=out_heatmap[y - top:y + bottom, x - left:x + right])
+
+    return out_heatmap
+
+
+def gaussian_radius(det_size, min_overlap):
+    r"""Generate 2D gaussian radius.
+
+    This function is modified from the `official github repo
+    <https://github.com/princeton-vl/CornerNet-Lite/blob/master/core/sample/
+    utils.py#L65>`_.
+
+    Given ``min_overlap``, radius could computed by a quadratic equation
+    according to Vieta's formulas.
+
+    There are 3 cases for computing gaussian radius, details are following:
+
+    - Explanation of figure: ``lt`` and ``br`` indicates the left-top and
+      bottom-right corner of ground truth box. ``x`` indicates the
+      generated corner at the limited position when ``radius=r``.
+
+    - Case1: one corner is inside the gt box and the other is outside.
+
+    .. code:: text
+
+        |<   width   >|
+
+        lt-+----------+         -
+        |  |          |         ^
+        +--x----------+--+
+        |  |          |  |
+        |  |          |  |    height
+        |  | overlap  |  |
+        |  |          |  |
+        |  |          |  |      v
+        +--+---------br--+      -
+           |          |  |
+           +----------+--x
+
+    To ensure IoU of generated box and gt box is larger than ``min_overlap``:
+
+    .. math::
+        \cfrac{(w-r)*(h-r)}{w*h+(w+h)r-r^2} \ge {iou} \quad\Rightarrow\quad
+        {r^2-(w+h)r+\cfrac{1-iou}{1+iou}*w*h} \ge 0 \\
+        {a} = 1,\quad{b} = {-(w+h)},\quad{c} = {\cfrac{1-iou}{1+iou}*w*h}
+        {r} \le \cfrac{-b-\sqrt{b^2-4*a*c}}{2*a}
+
+    - Case2: both two corners are inside the gt box.
+
+    .. code:: text
+
+        |<   width   >|
+
+        lt-+----------+         -
+        |  |          |         ^
+        +--x-------+  |
+        |  |       |  |
+        |  |overlap|  |       height
+        |  |       |  |
+        |  +-------x--+
+        |          |  |         v
+        +----------+-br         -
+
+    To ensure IoU of generated box and gt box is larger than ``min_overlap``:
+
+    .. math::
+        \cfrac{(w-2*r)*(h-2*r)}{w*h} \ge {iou} \quad\Rightarrow\quad
+        {4r^2-2(w+h)r+(1-iou)*w*h} \ge 0 \\
+        {a} = 4,\quad {b} = {-2(w+h)},\quad {c} = {(1-iou)*w*h}
+        {r} \le \cfrac{-b-\sqrt{b^2-4*a*c}}{2*a}
+
+    - Case3: both two corners are outside the gt box.
+
+    .. code:: text
+
+           |<   width   >|
+
+        x--+----------------+
+        |  |                |
+        +-lt-------------+  |   -
+        |  |             |  |   ^
+        |  |             |  |
+        |  |   overlap   |  | height
+        |  |             |  |
+        |  |             |  |   v
+        |  +------------br--+   -
+        |                |  |
+        +----------------+--x
+
+    To ensure IoU of generated box and gt box is larger than ``min_overlap``:
+
+    .. math::
+        \cfrac{w*h}{(w+2*r)*(h+2*r)} \ge {iou} \quad\Rightarrow\quad
+        {4*iou*r^2+2*iou*(w+h)r+(iou-1)*w*h} \le 0 \\
+        {a} = {4*iou},\quad {b} = {2*iou*(w+h)},\quad {c} = {(iou-1)*w*h} \\
+        {r} \le \cfrac{-b+\sqrt{b^2-4*a*c}}{2*a}
+
+    Args:
+        det_size (list[int]): Shape of object.
+        min_overlap (float): Min IoU with ground truth for boxes generated by
+            keypoints inside the gaussian kernel.
+
+    Returns:
+        radius (int): Radius of gaussian kernel.
+    """
+    height, width = det_size
+
+    a1 = 1
+    b1 = (height + width)
+    c1 = width * height * (1 - min_overlap) / (1 + min_overlap)
+    sq1 = sqrt(b1**2 - 4 * a1 * c1)
+    r1 = (b1 - sq1) / (2 * a1)
+
+    a2 = 4
+    b2 = 2 * (height + width)
+    c2 = (1 - min_overlap) * width * height
+    sq2 = sqrt(b2**2 - 4 * a2 * c2)
+    r2 = (b2 - sq2) / (2 * a2)
+
+    a3 = 4 * min_overlap
+    b3 = -2 * min_overlap * (height + width)
+    c3 = (min_overlap - 1) * width * height
+    sq3 = sqrt(b3**2 - 4 * a3 * c3)
+    r3 = (b3 + sq3) / (2 * a3)
+    return min(r1, r2, r3)
diff --git a/python/autoCut/models/swin/mmdet/models/utils/positional_encoding.py b/python/autoCut/models/swin/mmdet/models/utils/positional_encoding.py
new file mode 100644
index 0000000..9bda2bb
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/utils/positional_encoding.py
@@ -0,0 +1,150 @@
+import math
+
+import torch
+import torch.nn as nn
+from mmcv.cnn import uniform_init
+
+from .builder import POSITIONAL_ENCODING
+
+
+@POSITIONAL_ENCODING.register_module()
+class SinePositionalEncoding(nn.Module):
+    """Position encoding with sine and cosine functions.
+
+    See `End-to-End Object Detection with Transformers
+    <https://arxiv.org/pdf/2005.12872>`_ for details.
+
+    Args:
+        num_feats (int): The feature dimension for each position
+            along x-axis or y-axis. Note the final returned dimension
+            for each position is 2 times of this value.
+        temperature (int, optional): The temperature used for scaling
+            the position embedding. Default 10000.
+        normalize (bool, optional): Whether to normalize the position
+            embedding. Default False.
+        scale (float, optional): A scale factor that scales the position
+            embedding. The scale will be used only when `normalize` is True.
+            Default 2*pi.
+        eps (float, optional): A value added to the denominator for
+            numerical stability. Default 1e-6.
+    """
+
+    def __init__(self,
+                 num_feats,
+                 temperature=10000,
+                 normalize=False,
+                 scale=2 * math.pi,
+                 eps=1e-6):
+        super(SinePositionalEncoding, self).__init__()
+        if normalize:
+            assert isinstance(scale, (float, int)), 'when normalize is set,' \
+                'scale should be provided and in float or int type, ' \
+                f'found {type(scale)}'
+        self.num_feats = num_feats
+        self.temperature = temperature
+        self.normalize = normalize
+        self.scale = scale
+        self.eps = eps
+
+    def forward(self, mask):
+        """Forward function for `SinePositionalEncoding`.
+
+        Args:
+            mask (Tensor): ByteTensor mask. Non-zero values representing
+                ignored positions, while zero values means valid positions
+                for this image. Shape [bs, h, w].
+
+        Returns:
+            pos (Tensor): Returned position embedding with shape
+                [bs, num_feats*2, h, w].
+        """
+        not_mask = ~mask
+        y_embed = not_mask.cumsum(1, dtype=torch.float32)
+        x_embed = not_mask.cumsum(2, dtype=torch.float32)
+        if self.normalize:
+            y_embed = y_embed / (y_embed[:, -1:, :] + self.eps) * self.scale
+            x_embed = x_embed / (x_embed[:, :, -1:] + self.eps) * self.scale
+        dim_t = torch.arange(
+            self.num_feats, dtype=torch.float32, device=mask.device)
+        dim_t = self.temperature**(2 * (dim_t // 2) / self.num_feats)
+        pos_x = x_embed[:, :, :, None] / dim_t
+        pos_y = y_embed[:, :, :, None] / dim_t
+        pos_x = torch.stack(
+            (pos_x[:, :, :, 0::2].sin(), pos_x[:, :, :, 1::2].cos()),
+            dim=4).flatten(3)
+        pos_y = torch.stack(
+            (pos_y[:, :, :, 0::2].sin(), pos_y[:, :, :, 1::2].cos()),
+            dim=4).flatten(3)
+        pos = torch.cat((pos_y, pos_x), dim=3).permute(0, 3, 1, 2)
+        return pos
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(num_feats={self.num_feats}, '
+        repr_str += f'temperature={self.temperature}, '
+        repr_str += f'normalize={self.normalize}, '
+        repr_str += f'scale={self.scale}, '
+        repr_str += f'eps={self.eps})'
+        return repr_str
+
+
+@POSITIONAL_ENCODING.register_module()
+class LearnedPositionalEncoding(nn.Module):
+    """Position embedding with learnable embedding weights.
+
+    Args:
+        num_feats (int): The feature dimension for each position
+            along x-axis or y-axis. The final returned dimension for
+            each position is 2 times of this value.
+        row_num_embed (int, optional): The dictionary size of row embeddings.
+            Default 50.
+        col_num_embed (int, optional): The dictionary size of col embeddings.
+            Default 50.
+    """
+
+    def __init__(self, num_feats, row_num_embed=50, col_num_embed=50):
+        super(LearnedPositionalEncoding, self).__init__()
+        self.row_embed = nn.Embedding(row_num_embed, num_feats)
+        self.col_embed = nn.Embedding(col_num_embed, num_feats)
+        self.num_feats = num_feats
+        self.row_num_embed = row_num_embed
+        self.col_num_embed = col_num_embed
+        self.init_weights()
+
+    def init_weights(self):
+        """Initialize the learnable weights."""
+        uniform_init(self.row_embed)
+        uniform_init(self.col_embed)
+
+    def forward(self, mask):
+        """Forward function for `LearnedPositionalEncoding`.
+
+        Args:
+            mask (Tensor): ByteTensor mask. Non-zero values representing
+                ignored positions, while zero values means valid positions
+                for this image. Shape [bs, h, w].
+
+        Returns:
+            pos (Tensor): Returned position embedding with shape
+                [bs, num_feats*2, h, w].
+        """
+        h, w = mask.shape[-2:]
+        x = torch.arange(w, device=mask.device)
+        y = torch.arange(h, device=mask.device)
+        x_embed = self.col_embed(x)
+        y_embed = self.row_embed(y)
+        pos = torch.cat(
+            (x_embed.unsqueeze(0).repeat(h, 1, 1), y_embed.unsqueeze(1).repeat(
+                1, w, 1)),
+            dim=-1).permute(2, 0,
+                            1).unsqueeze(0).repeat(mask.shape[0], 1, 1, 1)
+        return pos
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(num_feats={self.num_feats}, '
+        repr_str += f'row_num_embed={self.row_num_embed}, '
+        repr_str += f'col_num_embed={self.col_num_embed})'
+        return repr_str
diff --git a/python/autoCut/models/swin/mmdet/models/utils/res_layer.py b/python/autoCut/models/swin/mmdet/models/utils/res_layer.py
new file mode 100644
index 0000000..4a4efd3
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/utils/res_layer.py
@@ -0,0 +1,187 @@
+from mmcv.cnn import build_conv_layer, build_norm_layer
+from torch import nn as nn
+
+
+class ResLayer(nn.Sequential):
+    """ResLayer to build ResNet style backbone.
+
+    Args:
+        block (nn.Module): block used to build ResLayer.
+        inplanes (int): inplanes of block.
+        planes (int): planes of block.
+        num_blocks (int): number of blocks.
+        stride (int): stride of the first block. Default: 1
+        avg_down (bool): Use AvgPool instead of stride conv when
+            downsampling in the bottleneck. Default: False
+        conv_cfg (dict): dictionary to construct and config conv layer.
+            Default: None
+        norm_cfg (dict): dictionary to construct and config norm layer.
+            Default: dict(type='BN')
+        downsample_first (bool): Downsample at the first block or last block.
+            False for Hourglass, True for ResNet. Default: True
+    """
+
+    def __init__(self,
+                 block,
+                 inplanes,
+                 planes,
+                 num_blocks,
+                 stride=1,
+                 avg_down=False,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 downsample_first=True,
+                 **kwargs):
+        self.block = block
+
+        downsample = None
+        if stride != 1 or inplanes != planes * block.expansion:
+            downsample = []
+            conv_stride = stride
+            if avg_down:
+                conv_stride = 1
+                downsample.append(
+                    nn.AvgPool2d(
+                        kernel_size=stride,
+                        stride=stride,
+                        ceil_mode=True,
+                        count_include_pad=False))
+            downsample.extend([
+                build_conv_layer(
+                    conv_cfg,
+                    inplanes,
+                    planes * block.expansion,
+                    kernel_size=1,
+                    stride=conv_stride,
+                    bias=False),
+                build_norm_layer(norm_cfg, planes * block.expansion)[1]
+            ])
+            downsample = nn.Sequential(*downsample)
+
+        layers = []
+        if downsample_first:
+            layers.append(
+                block(
+                    inplanes=inplanes,
+                    planes=planes,
+                    stride=stride,
+                    downsample=downsample,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg,
+                    **kwargs))
+            inplanes = planes * block.expansion
+            for _ in range(1, num_blocks):
+                layers.append(
+                    block(
+                        inplanes=inplanes,
+                        planes=planes,
+                        stride=1,
+                        conv_cfg=conv_cfg,
+                        norm_cfg=norm_cfg,
+                        **kwargs))
+
+        else:  # downsample_first=False is for HourglassModule
+            for _ in range(num_blocks - 1):
+                layers.append(
+                    block(
+                        inplanes=inplanes,
+                        planes=inplanes,
+                        stride=1,
+                        conv_cfg=conv_cfg,
+                        norm_cfg=norm_cfg,
+                        **kwargs))
+            layers.append(
+                block(
+                    inplanes=inplanes,
+                    planes=planes,
+                    stride=stride,
+                    downsample=downsample,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg,
+                    **kwargs))
+        super(ResLayer, self).__init__(*layers)
+
+
+class SimplifiedBasicBlock(nn.Module):
+    """Simplified version of original basic residual block. This is used in
+    `SCNet <https://arxiv.org/abs/2012.10150>`_.
+
+    - Norm layer is now optional
+    - Last ReLU in forward function is removed
+    """
+    expansion = 1
+
+    def __init__(self,
+                 inplanes,
+                 planes,
+                 stride=1,
+                 dilation=1,
+                 downsample=None,
+                 style='pytorch',
+                 with_cp=False,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 dcn=None,
+                 plugins=None):
+        super(SimplifiedBasicBlock, self).__init__()
+        assert dcn is None, 'Not implemented yet.'
+        assert plugins is None, 'Not implemented yet.'
+        assert not with_cp, 'Not implemented yet.'
+        self.with_norm = norm_cfg is not None
+        with_bias = True if norm_cfg is None else False
+        self.conv1 = build_conv_layer(
+            conv_cfg,
+            inplanes,
+            planes,
+            3,
+            stride=stride,
+            padding=dilation,
+            dilation=dilation,
+            bias=with_bias)
+        if self.with_norm:
+            self.norm1_name, norm1 = build_norm_layer(
+                norm_cfg, planes, postfix=1)
+            self.add_module(self.norm1_name, norm1)
+        self.conv2 = build_conv_layer(
+            conv_cfg, planes, planes, 3, padding=1, bias=with_bias)
+        if self.with_norm:
+            self.norm2_name, norm2 = build_norm_layer(
+                norm_cfg, planes, postfix=2)
+            self.add_module(self.norm2_name, norm2)
+
+        self.relu = nn.ReLU(inplace=True)
+        self.downsample = downsample
+        self.stride = stride
+        self.dilation = dilation
+        self.with_cp = with_cp
+
+    @property
+    def norm1(self):
+        """nn.Module: normalization layer after the first convolution layer"""
+        return getattr(self, self.norm1_name) if self.with_norm else None
+
+    @property
+    def norm2(self):
+        """nn.Module: normalization layer after the second convolution layer"""
+        return getattr(self, self.norm2_name) if self.with_norm else None
+
+    def forward(self, x):
+        """Forward function."""
+
+        identity = x
+
+        out = self.conv1(x)
+        if self.with_norm:
+            out = self.norm1(out)
+        out = self.relu(out)
+
+        out = self.conv2(out)
+        if self.with_norm:
+            out = self.norm2(out)
+
+        if self.downsample is not None:
+            identity = self.downsample(x)
+
+        out += identity
+
+        return out
diff --git a/python/autoCut/models/swin/mmdet/models/utils/transformer.py b/python/autoCut/models/swin/mmdet/models/utils/transformer.py
new file mode 100644
index 0000000..83870ee
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/models/utils/transformer.py
@@ -0,0 +1,860 @@
+import torch
+import torch.nn as nn
+from mmcv.cnn import (Linear, build_activation_layer, build_norm_layer,
+                      xavier_init)
+
+from .builder import TRANSFORMER
+
+
+class MultiheadAttention(nn.Module):
+    """A warpper for torch.nn.MultiheadAttention.
+
+    This module implements MultiheadAttention with residual connection,
+    and positional encoding used in DETR is also passed as input.
+
+    Args:
+        embed_dims (int): The embedding dimension.
+        num_heads (int): Parallel attention heads. Same as
+            `nn.MultiheadAttention`.
+        dropout (float): A Dropout layer on attn_output_weights. Default 0.0.
+    """
+
+    def __init__(self, embed_dims, num_heads, dropout=0.0):
+        super(MultiheadAttention, self).__init__()
+        assert embed_dims % num_heads == 0, 'embed_dims must be ' \
+            f'divisible by num_heads. got {embed_dims} and {num_heads}.'
+        self.embed_dims = embed_dims
+        self.num_heads = num_heads
+        self.dropout = dropout
+        self.attn = nn.MultiheadAttention(embed_dims, num_heads, dropout)
+        self.dropout = nn.Dropout(dropout)
+
+    def forward(self,
+                x,
+                key=None,
+                value=None,
+                residual=None,
+                query_pos=None,
+                key_pos=None,
+                attn_mask=None,
+                key_padding_mask=None):
+        """Forward function for `MultiheadAttention`.
+
+        Args:
+            x (Tensor): The input query with shape [num_query, bs,
+                embed_dims]. Same in `nn.MultiheadAttention.forward`.
+            key (Tensor): The key tensor with shape [num_key, bs,
+                embed_dims]. Same in `nn.MultiheadAttention.forward`.
+                Default None. If None, the `query` will be used.
+            value (Tensor): The value tensor with same shape as `key`.
+                Same in `nn.MultiheadAttention.forward`. Default None.
+                If None, the `key` will be used.
+            residual (Tensor): The tensor used for addition, with the
+                same shape as `x`. Default None. If None, `x` will be used.
+            query_pos (Tensor): The positional encoding for query, with
+                the same shape as `x`. Default None. If not None, it will
+                be added to `x` before forward function.
+            key_pos (Tensor): The positional encoding for `key`, with the
+                same shape as `key`. Default None. If not None, it will
+                be added to `key` before forward function. If None, and
+                `query_pos` has the same shape as `key`, then `query_pos`
+                will be used for `key_pos`.
+            attn_mask (Tensor): ByteTensor mask with shape [num_query,
+                num_key]. Same in `nn.MultiheadAttention.forward`.
+                Default None.
+            key_padding_mask (Tensor): ByteTensor with shape [bs, num_key].
+                Same in `nn.MultiheadAttention.forward`. Default None.
+
+        Returns:
+            Tensor: forwarded results with shape [num_query, bs, embed_dims].
+        """
+        query = x
+        if key is None:
+            key = query
+        if value is None:
+            value = key
+        if residual is None:
+            residual = x
+        if key_pos is None:
+            if query_pos is not None and key is not None:
+                if query_pos.shape == key.shape:
+                    key_pos = query_pos
+        if query_pos is not None:
+            query = query + query_pos
+        if key_pos is not None:
+            key = key + key_pos
+        out = self.attn(
+            query,
+            key,
+            value=value,
+            attn_mask=attn_mask,
+            key_padding_mask=key_padding_mask)[0]
+
+        return residual + self.dropout(out)
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(embed_dims={self.embed_dims}, '
+        repr_str += f'num_heads={self.num_heads}, '
+        repr_str += f'dropout={self.dropout})'
+        return repr_str
+
+
+class FFN(nn.Module):
+    """Implements feed-forward networks (FFNs) with residual connection.
+
+    Args:
+        embed_dims (int): The feature dimension. Same as
+            `MultiheadAttention`.
+        feedforward_channels (int): The hidden dimension of FFNs.
+        num_fcs (int, optional): The number of fully-connected layers in
+            FFNs. Defaults to 2.
+        act_cfg (dict, optional): The activation config for FFNs.
+        dropout (float, optional): Probability of an element to be
+            zeroed. Default 0.0.
+        add_residual (bool, optional): Add resudual connection.
+            Defaults to True.
+    """
+
+    def __init__(self,
+                 embed_dims,
+                 feedforward_channels,
+                 num_fcs=2,
+                 act_cfg=dict(type='ReLU', inplace=True),
+                 dropout=0.0,
+                 add_residual=True):
+        super(FFN, self).__init__()
+        assert num_fcs >= 2, 'num_fcs should be no less ' \
+            f'than 2. got {num_fcs}.'
+        self.embed_dims = embed_dims
+        self.feedforward_channels = feedforward_channels
+        self.num_fcs = num_fcs
+        self.act_cfg = act_cfg
+        self.dropout = dropout
+        self.activate = build_activation_layer(act_cfg)
+
+        layers = nn.ModuleList()
+        in_channels = embed_dims
+        for _ in range(num_fcs - 1):
+            layers.append(
+                nn.Sequential(
+                    Linear(in_channels, feedforward_channels), self.activate,
+                    nn.Dropout(dropout)))
+            in_channels = feedforward_channels
+        layers.append(Linear(feedforward_channels, embed_dims))
+        self.layers = nn.Sequential(*layers)
+        self.dropout = nn.Dropout(dropout)
+        self.add_residual = add_residual
+
+    def forward(self, x, residual=None):
+        """Forward function for `FFN`."""
+        out = self.layers(x)
+        if not self.add_residual:
+            return out
+        if residual is None:
+            residual = x
+        return residual + self.dropout(out)
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(embed_dims={self.embed_dims}, '
+        repr_str += f'feedforward_channels={self.feedforward_channels}, '
+        repr_str += f'num_fcs={self.num_fcs}, '
+        repr_str += f'act_cfg={self.act_cfg}, '
+        repr_str += f'dropout={self.dropout}, '
+        repr_str += f'add_residual={self.add_residual})'
+        return repr_str
+
+
+class TransformerEncoderLayer(nn.Module):
+    """Implements one encoder layer in DETR transformer.
+
+    Args:
+        embed_dims (int): The feature dimension. Same as `FFN`.
+        num_heads (int): Parallel attention heads.
+        feedforward_channels (int): The hidden dimension for FFNs.
+        dropout (float): Probability of an element to be zeroed. Default 0.0.
+        order (tuple[str]): The order for encoder layer. Valid examples are
+            ('selfattn', 'norm', 'ffn', 'norm') and ('norm', 'selfattn',
+            'norm', 'ffn'). Default ('selfattn', 'norm', 'ffn', 'norm').
+        act_cfg (dict): The activation config for FFNs. Default ReLU.
+        norm_cfg (dict): Config dict for normalization layer. Default
+            layer normalization.
+        num_fcs (int): The number of fully-connected layers for FFNs.
+            Default 2.
+    """
+
+    def __init__(self,
+                 embed_dims,
+                 num_heads,
+                 feedforward_channels,
+                 dropout=0.0,
+                 order=('selfattn', 'norm', 'ffn', 'norm'),
+                 act_cfg=dict(type='ReLU', inplace=True),
+                 norm_cfg=dict(type='LN'),
+                 num_fcs=2):
+        super(TransformerEncoderLayer, self).__init__()
+        assert isinstance(order, tuple) and len(order) == 4
+        assert set(order) == set(['selfattn', 'norm', 'ffn'])
+        self.embed_dims = embed_dims
+        self.num_heads = num_heads
+        self.feedforward_channels = feedforward_channels
+        self.dropout = dropout
+        self.order = order
+        self.act_cfg = act_cfg
+        self.norm_cfg = norm_cfg
+        self.num_fcs = num_fcs
+        self.pre_norm = order[0] == 'norm'
+        self.self_attn = MultiheadAttention(embed_dims, num_heads, dropout)
+        self.ffn = FFN(embed_dims, feedforward_channels, num_fcs, act_cfg,
+                       dropout)
+        self.norms = nn.ModuleList()
+        self.norms.append(build_norm_layer(norm_cfg, embed_dims)[1])
+        self.norms.append(build_norm_layer(norm_cfg, embed_dims)[1])
+
+    def forward(self, x, pos=None, attn_mask=None, key_padding_mask=None):
+        """Forward function for `TransformerEncoderLayer`.
+
+        Args:
+            x (Tensor): The input query with shape [num_key, bs,
+                embed_dims]. Same in `MultiheadAttention.forward`.
+            pos (Tensor): The positional encoding for query. Default None.
+                Same as `query_pos` in `MultiheadAttention.forward`.
+            attn_mask (Tensor): ByteTensor mask with shape [num_key,
+                num_key]. Same in `MultiheadAttention.forward`. Default None.
+            key_padding_mask (Tensor): ByteTensor with shape [bs, num_key].
+                Same in `MultiheadAttention.forward`. Default None.
+
+        Returns:
+            Tensor: forwarded results with shape [num_key, bs, embed_dims].
+        """
+        norm_cnt = 0
+        inp_residual = x
+        for layer in self.order:
+            if layer == 'selfattn':
+                # self attention
+                query = key = value = x
+                x = self.self_attn(
+                    query,
+                    key,
+                    value,
+                    inp_residual if self.pre_norm else None,
+                    query_pos=pos,
+                    key_pos=pos,
+                    attn_mask=attn_mask,
+                    key_padding_mask=key_padding_mask)
+                inp_residual = x
+            elif layer == 'norm':
+                x = self.norms[norm_cnt](x)
+                norm_cnt += 1
+            elif layer == 'ffn':
+                x = self.ffn(x, inp_residual if self.pre_norm else None)
+        return x
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(embed_dims={self.embed_dims}, '
+        repr_str += f'num_heads={self.num_heads}, '
+        repr_str += f'feedforward_channels={self.feedforward_channels}, '
+        repr_str += f'dropout={self.dropout}, '
+        repr_str += f'order={self.order}, '
+        repr_str += f'act_cfg={self.act_cfg}, '
+        repr_str += f'norm_cfg={self.norm_cfg}, '
+        repr_str += f'num_fcs={self.num_fcs})'
+        return repr_str
+
+
+class TransformerDecoderLayer(nn.Module):
+    """Implements one decoder layer in DETR transformer.
+
+    Args:
+        embed_dims (int): The feature dimension. Same as
+            `TransformerEncoderLayer`.
+        num_heads (int): Parallel attention heads.
+        feedforward_channels (int): Same as `TransformerEncoderLayer`.
+        dropout (float): Same as `TransformerEncoderLayer`. Default 0.0.
+        order (tuple[str]): The order for decoder layer. Valid examples are
+            ('selfattn', 'norm', 'multiheadattn', 'norm', 'ffn', 'norm') and
+            ('norm', 'selfattn', 'norm', 'multiheadattn', 'norm', 'ffn').
+            Default the former.
+        act_cfg (dict): Same as `TransformerEncoderLayer`. Default ReLU.
+        norm_cfg (dict): Config dict for normalization layer. Default
+            layer normalization.
+        num_fcs (int): The number of fully-connected layers in FFNs.
+    """
+
+    def __init__(self,
+                 embed_dims,
+                 num_heads,
+                 feedforward_channels,
+                 dropout=0.0,
+                 order=('selfattn', 'norm', 'multiheadattn', 'norm', 'ffn',
+                        'norm'),
+                 act_cfg=dict(type='ReLU', inplace=True),
+                 norm_cfg=dict(type='LN'),
+                 num_fcs=2):
+        super(TransformerDecoderLayer, self).__init__()
+        assert isinstance(order, tuple) and len(order) == 6
+        assert set(order) == set(['selfattn', 'norm', 'multiheadattn', 'ffn'])
+        self.embed_dims = embed_dims
+        self.num_heads = num_heads
+        self.feedforward_channels = feedforward_channels
+        self.dropout = dropout
+        self.order = order
+        self.act_cfg = act_cfg
+        self.norm_cfg = norm_cfg
+        self.num_fcs = num_fcs
+        self.pre_norm = order[0] == 'norm'
+        self.self_attn = MultiheadAttention(embed_dims, num_heads, dropout)
+        self.multihead_attn = MultiheadAttention(embed_dims, num_heads,
+                                                 dropout)
+        self.ffn = FFN(embed_dims, feedforward_channels, num_fcs, act_cfg,
+                       dropout)
+        self.norms = nn.ModuleList()
+        # 3 norm layers in official DETR's TransformerDecoderLayer
+        for _ in range(3):
+            self.norms.append(build_norm_layer(norm_cfg, embed_dims)[1])
+
+    def forward(self,
+                x,
+                memory,
+                memory_pos=None,
+                query_pos=None,
+                memory_attn_mask=None,
+                target_attn_mask=None,
+                memory_key_padding_mask=None,
+                target_key_padding_mask=None):
+        """Forward function for `TransformerDecoderLayer`.
+
+        Args:
+            x (Tensor): Input query with shape [num_query, bs, embed_dims].
+            memory (Tensor): Tensor got from `TransformerEncoder`, with shape
+                [num_key, bs, embed_dims].
+            memory_pos (Tensor): The positional encoding for `memory`. Default
+                None. Same as `key_pos` in `MultiheadAttention.forward`.
+            query_pos (Tensor): The positional encoding for `query`. Default
+                None. Same as `query_pos` in `MultiheadAttention.forward`.
+            memory_attn_mask (Tensor): ByteTensor mask for `memory`, with
+                shape [num_key, num_key]. Same as `attn_mask` in
+                `MultiheadAttention.forward`. Default None.
+            target_attn_mask (Tensor): ByteTensor mask for `x`, with shape
+                [num_query, num_query]. Same as `attn_mask` in
+                `MultiheadAttention.forward`. Default None.
+            memory_key_padding_mask (Tensor): ByteTensor for `memory`, with
+                shape [bs, num_key]. Same as `key_padding_mask` in
+                `MultiheadAttention.forward`. Default None.
+            target_key_padding_mask (Tensor): ByteTensor for `x`, with shape
+                [bs, num_query]. Same as `key_padding_mask` in
+                `MultiheadAttention.forward`. Default None.
+
+        Returns:
+            Tensor: forwarded results with shape [num_query, bs, embed_dims].
+        """
+        norm_cnt = 0
+        inp_residual = x
+        for layer in self.order:
+            if layer == 'selfattn':
+                query = key = value = x
+                x = self.self_attn(
+                    query,
+                    key,
+                    value,
+                    inp_residual if self.pre_norm else None,
+                    query_pos,
+                    key_pos=query_pos,
+                    attn_mask=target_attn_mask,
+                    key_padding_mask=target_key_padding_mask)
+                inp_residual = x
+            elif layer == 'norm':
+                x = self.norms[norm_cnt](x)
+                norm_cnt += 1
+            elif layer == 'multiheadattn':
+                query = x
+                key = value = memory
+                x = self.multihead_attn(
+                    query,
+                    key,
+                    value,
+                    inp_residual if self.pre_norm else None,
+                    query_pos,
+                    key_pos=memory_pos,
+                    attn_mask=memory_attn_mask,
+                    key_padding_mask=memory_key_padding_mask)
+                inp_residual = x
+            elif layer == 'ffn':
+                x = self.ffn(x, inp_residual if self.pre_norm else None)
+        return x
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(embed_dims={self.embed_dims}, '
+        repr_str += f'num_heads={self.num_heads}, '
+        repr_str += f'feedforward_channels={self.feedforward_channels}, '
+        repr_str += f'dropout={self.dropout}, '
+        repr_str += f'order={self.order}, '
+        repr_str += f'act_cfg={self.act_cfg}, '
+        repr_str += f'norm_cfg={self.norm_cfg}, '
+        repr_str += f'num_fcs={self.num_fcs})'
+        return repr_str
+
+
+class TransformerEncoder(nn.Module):
+    """Implements the encoder in DETR transformer.
+
+    Args:
+        num_layers (int): The number of `TransformerEncoderLayer`.
+        embed_dims (int): Same as `TransformerEncoderLayer`.
+        num_heads (int): Same as `TransformerEncoderLayer`.
+        feedforward_channels (int): Same as `TransformerEncoderLayer`.
+        dropout (float): Same as `TransformerEncoderLayer`. Default 0.0.
+        order (tuple[str]): Same as `TransformerEncoderLayer`.
+        act_cfg (dict): Same as `TransformerEncoderLayer`. Default ReLU.
+        norm_cfg (dict): Same as `TransformerEncoderLayer`. Default
+            layer normalization.
+        num_fcs (int): Same as `TransformerEncoderLayer`. Default 2.
+    """
+
+    def __init__(self,
+                 num_layers,
+                 embed_dims,
+                 num_heads,
+                 feedforward_channels,
+                 dropout=0.0,
+                 order=('selfattn', 'norm', 'ffn', 'norm'),
+                 act_cfg=dict(type='ReLU', inplace=True),
+                 norm_cfg=dict(type='LN'),
+                 num_fcs=2):
+        super(TransformerEncoder, self).__init__()
+        assert isinstance(order, tuple) and len(order) == 4
+        assert set(order) == set(['selfattn', 'norm', 'ffn'])
+        self.num_layers = num_layers
+        self.embed_dims = embed_dims
+        self.num_heads = num_heads
+        self.feedforward_channels = feedforward_channels
+        self.dropout = dropout
+        self.order = order
+        self.act_cfg = act_cfg
+        self.norm_cfg = norm_cfg
+        self.num_fcs = num_fcs
+        self.pre_norm = order[0] == 'norm'
+        self.layers = nn.ModuleList()
+        for _ in range(num_layers):
+            self.layers.append(
+                TransformerEncoderLayer(embed_dims, num_heads,
+                                        feedforward_channels, dropout, order,
+                                        act_cfg, norm_cfg, num_fcs))
+        self.norm = build_norm_layer(norm_cfg,
+                                     embed_dims)[1] if self.pre_norm else None
+
+    def forward(self, x, pos=None, attn_mask=None, key_padding_mask=None):
+        """Forward function for `TransformerEncoder`.
+
+        Args:
+            x (Tensor): Input query. Same in `TransformerEncoderLayer.forward`.
+            pos (Tensor): Positional encoding for query. Default None.
+                Same in `TransformerEncoderLayer.forward`.
+            attn_mask (Tensor): ByteTensor attention mask. Default None.
+                Same in `TransformerEncoderLayer.forward`.
+            key_padding_mask (Tensor): Same in
+                `TransformerEncoderLayer.forward`. Default None.
+
+        Returns:
+            Tensor: Results with shape [num_key, bs, embed_dims].
+        """
+        for layer in self.layers:
+            x = layer(x, pos, attn_mask, key_padding_mask)
+        if self.norm is not None:
+            x = self.norm(x)
+        return x
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(num_layers={self.num_layers}, '
+        repr_str += f'embed_dims={self.embed_dims}, '
+        repr_str += f'num_heads={self.num_heads}, '
+        repr_str += f'feedforward_channels={self.feedforward_channels}, '
+        repr_str += f'dropout={self.dropout}, '
+        repr_str += f'order={self.order}, '
+        repr_str += f'act_cfg={self.act_cfg}, '
+        repr_str += f'norm_cfg={self.norm_cfg}, '
+        repr_str += f'num_fcs={self.num_fcs})'
+        return repr_str
+
+
+class TransformerDecoder(nn.Module):
+    """Implements the decoder in DETR transformer.
+
+    Args:
+        num_layers (int): The number of `TransformerDecoderLayer`.
+        embed_dims (int): Same as `TransformerDecoderLayer`.
+        num_heads (int): Same as `TransformerDecoderLayer`.
+        feedforward_channels (int): Same as `TransformerDecoderLayer`.
+        dropout (float): Same as `TransformerDecoderLayer`. Default 0.0.
+        order (tuple[str]): Same as `TransformerDecoderLayer`.
+        act_cfg (dict): Same as `TransformerDecoderLayer`. Default ReLU.
+        norm_cfg (dict): Same as `TransformerDecoderLayer`. Default
+            layer normalization.
+        num_fcs (int): Same as `TransformerDecoderLayer`. Default 2.
+    """
+
+    def __init__(self,
+                 num_layers,
+                 embed_dims,
+                 num_heads,
+                 feedforward_channels,
+                 dropout=0.0,
+                 order=('selfattn', 'norm', 'multiheadattn', 'norm', 'ffn',
+                        'norm'),
+                 act_cfg=dict(type='ReLU', inplace=True),
+                 norm_cfg=dict(type='LN'),
+                 num_fcs=2,
+                 return_intermediate=False):
+        super(TransformerDecoder, self).__init__()
+        assert isinstance(order, tuple) and len(order) == 6
+        assert set(order) == set(['selfattn', 'norm', 'multiheadattn', 'ffn'])
+        self.num_layers = num_layers
+        self.embed_dims = embed_dims
+        self.num_heads = num_heads
+        self.feedforward_channels = feedforward_channels
+        self.dropout = dropout
+        self.order = order
+        self.act_cfg = act_cfg
+        self.norm_cfg = norm_cfg
+        self.num_fcs = num_fcs
+        self.return_intermediate = return_intermediate
+        self.layers = nn.ModuleList()
+        for _ in range(num_layers):
+            self.layers.append(
+                TransformerDecoderLayer(embed_dims, num_heads,
+                                        feedforward_channels, dropout, order,
+                                        act_cfg, norm_cfg, num_fcs))
+        self.norm = build_norm_layer(norm_cfg, embed_dims)[1]
+
+    def forward(self,
+                x,
+                memory,
+                memory_pos=None,
+                query_pos=None,
+                memory_attn_mask=None,
+                target_attn_mask=None,
+                memory_key_padding_mask=None,
+                target_key_padding_mask=None):
+        """Forward function for `TransformerDecoder`.
+
+        Args:
+            x (Tensor): Input query. Same in `TransformerDecoderLayer.forward`.
+            memory (Tensor): Same in `TransformerDecoderLayer.forward`.
+            memory_pos (Tensor): Same in `TransformerDecoderLayer.forward`.
+                Default None.
+            query_pos (Tensor): Same in `TransformerDecoderLayer.forward`.
+                Default None.
+            memory_attn_mask (Tensor): Same in
+                `TransformerDecoderLayer.forward`. Default None.
+            target_attn_mask (Tensor): Same in
+                `TransformerDecoderLayer.forward`. Default None.
+            memory_key_padding_mask (Tensor): Same in
+                `TransformerDecoderLayer.forward`. Default None.
+            target_key_padding_mask (Tensor): Same in
+                `TransformerDecoderLayer.forward`. Default None.
+
+        Returns:
+            Tensor: Results with shape [num_query, bs, embed_dims].
+        """
+        intermediate = []
+        for layer in self.layers:
+            x = layer(x, memory, memory_pos, query_pos, memory_attn_mask,
+                      target_attn_mask, memory_key_padding_mask,
+                      target_key_padding_mask)
+            if self.return_intermediate:
+                intermediate.append(self.norm(x))
+        if self.norm is not None:
+            x = self.norm(x)
+            if self.return_intermediate:
+                intermediate.pop()
+                intermediate.append(x)
+        if self.return_intermediate:
+            return torch.stack(intermediate)
+        return x.unsqueeze(0)
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(num_layers={self.num_layers}, '
+        repr_str += f'embed_dims={self.embed_dims}, '
+        repr_str += f'num_heads={self.num_heads}, '
+        repr_str += f'feedforward_channels={self.feedforward_channels}, '
+        repr_str += f'dropout={self.dropout}, '
+        repr_str += f'order={self.order}, '
+        repr_str += f'act_cfg={self.act_cfg}, '
+        repr_str += f'norm_cfg={self.norm_cfg}, '
+        repr_str += f'num_fcs={self.num_fcs}, '
+        repr_str += f'return_intermediate={self.return_intermediate})'
+        return repr_str
+
+
+@TRANSFORMER.register_module()
+class Transformer(nn.Module):
+    """Implements the DETR transformer.
+
+    Following the official DETR implementation, this module copy-paste
+    from torch.nn.Transformer with modifications:
+
+        * positional encodings are passed in MultiheadAttention
+        * extra LN at the end of encoder is removed
+        * decoder returns a stack of activations from all decoding layers
+
+    See `paper: End-to-End Object Detection with Transformers
+    <https://arxiv.org/pdf/2005.12872>`_ for details.
+
+    Args:
+        embed_dims (int): The feature dimension.
+        num_heads (int): Parallel attention heads. Same as
+            `nn.MultiheadAttention`.
+        num_encoder_layers (int): Number of `TransformerEncoderLayer`.
+        num_decoder_layers (int): Number of `TransformerDecoderLayer`.
+        feedforward_channels (int): The hidden dimension for FFNs used in both
+            encoder and decoder.
+        dropout (float): Probability of an element to be zeroed. Default 0.0.
+        act_cfg (dict): Activation config for FFNs used in both encoder
+            and decoder. Default ReLU.
+        norm_cfg (dict): Config dict for normalization used in both encoder
+            and decoder. Default layer normalization.
+        num_fcs (int): The number of fully-connected layers in FFNs, which is
+            used for both encoder and decoder.
+        pre_norm (bool): Whether the normalization layer is ordered
+            first in the encoder and decoder. Default False.
+        return_intermediate_dec (bool): Whether to return the intermediate
+            output from each TransformerDecoderLayer or only the last
+            TransformerDecoderLayer. Default False. If False, the returned
+            `hs` has shape [num_decoder_layers, bs, num_query, embed_dims].
+            If True, the returned `hs` will have shape [1, bs, num_query,
+            embed_dims].
+    """
+
+    def __init__(self,
+                 embed_dims=512,
+                 num_heads=8,
+                 num_encoder_layers=6,
+                 num_decoder_layers=6,
+                 feedforward_channels=2048,
+                 dropout=0.0,
+                 act_cfg=dict(type='ReLU', inplace=True),
+                 norm_cfg=dict(type='LN'),
+                 num_fcs=2,
+                 pre_norm=False,
+                 return_intermediate_dec=False):
+        super(Transformer, self).__init__()
+        self.embed_dims = embed_dims
+        self.num_heads = num_heads
+        self.num_encoder_layers = num_encoder_layers
+        self.num_decoder_layers = num_decoder_layers
+        self.feedforward_channels = feedforward_channels
+        self.dropout = dropout
+        self.act_cfg = act_cfg
+        self.norm_cfg = norm_cfg
+        self.num_fcs = num_fcs
+        self.pre_norm = pre_norm
+        self.return_intermediate_dec = return_intermediate_dec
+        if self.pre_norm:
+            encoder_order = ('norm', 'selfattn', 'norm', 'ffn')
+            decoder_order = ('norm', 'selfattn', 'norm', 'multiheadattn',
+                             'norm', 'ffn')
+        else:
+            encoder_order = ('selfattn', 'norm', 'ffn', 'norm')
+            decoder_order = ('selfattn', 'norm', 'multiheadattn', 'norm',
+                             'ffn', 'norm')
+        self.encoder = TransformerEncoder(num_encoder_layers, embed_dims,
+                                          num_heads, feedforward_channels,
+                                          dropout, encoder_order, act_cfg,
+                                          norm_cfg, num_fcs)
+        self.decoder = TransformerDecoder(num_decoder_layers, embed_dims,
+                                          num_heads, feedforward_channels,
+                                          dropout, decoder_order, act_cfg,
+                                          norm_cfg, num_fcs,
+                                          return_intermediate_dec)
+
+    def init_weights(self, distribution='uniform'):
+        """Initialize the transformer weights."""
+        # follow the official DETR to init parameters
+        for m in self.modules():
+            if hasattr(m, 'weight') and m.weight.dim() > 1:
+                xavier_init(m, distribution=distribution)
+
+    def forward(self, x, mask, query_embed, pos_embed):
+        """Forward function for `Transformer`.
+
+        Args:
+            x (Tensor): Input query with shape [bs, c, h, w] where
+                c = embed_dims.
+            mask (Tensor): The key_padding_mask used for encoder and decoder,
+                with shape [bs, h, w].
+            query_embed (Tensor): The query embedding for decoder, with shape
+                [num_query, c].
+            pos_embed (Tensor): The positional encoding for encoder and
+                decoder, with the same shape as `x`.
+
+        Returns:
+            tuple[Tensor]: results of decoder containing the following tensor.
+
+                - out_dec: Output from decoder. If return_intermediate_dec \
+                      is True output has shape [num_dec_layers, bs,
+                      num_query, embed_dims], else has shape [1, bs, \
+                      num_query, embed_dims].
+                - memory: Output results from encoder, with shape \
+                      [bs, embed_dims, h, w].
+        """
+        bs, c, h, w = x.shape
+        x = x.flatten(2).permute(2, 0, 1)  # [bs, c, h, w] -> [h*w, bs, c]
+        pos_embed = pos_embed.flatten(2).permute(2, 0, 1)
+        query_embed = query_embed.unsqueeze(1).repeat(
+            1, bs, 1)  # [num_query, dim] -> [num_query, bs, dim]
+        mask = mask.flatten(1)  # [bs, h, w] -> [bs, h*w]
+        memory = self.encoder(
+            x, pos=pos_embed, attn_mask=None, key_padding_mask=mask)
+        target = torch.zeros_like(query_embed)
+        # out_dec: [num_layers, num_query, bs, dim]
+        out_dec = self.decoder(
+            target,
+            memory,
+            memory_pos=pos_embed,
+            query_pos=query_embed,
+            memory_attn_mask=None,
+            target_attn_mask=None,
+            memory_key_padding_mask=mask,
+            target_key_padding_mask=None)
+        out_dec = out_dec.transpose(1, 2)
+        memory = memory.permute(1, 2, 0).reshape(bs, c, h, w)
+        return out_dec, memory
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(embed_dims={self.embed_dims}, '
+        repr_str += f'num_heads={self.num_heads}, '
+        repr_str += f'num_encoder_layers={self.num_encoder_layers}, '
+        repr_str += f'num_decoder_layers={self.num_decoder_layers}, '
+        repr_str += f'feedforward_channels={self.feedforward_channels}, '
+        repr_str += f'dropout={self.dropout}, '
+        repr_str += f'act_cfg={self.act_cfg}, '
+        repr_str += f'norm_cfg={self.norm_cfg}, '
+        repr_str += f'num_fcs={self.num_fcs}, '
+        repr_str += f'pre_norm={self.pre_norm}, '
+        repr_str += f'return_intermediate_dec={self.return_intermediate_dec})'
+        return repr_str
+
+
+@TRANSFORMER.register_module()
+class DynamicConv(nn.Module):
+    """Implements Dynamic Convolution.
+
+    This module generate parameters for each sample and
+    use bmm to implement 1*1 convolution. Code is modified
+    from the `official github repo <https://github.com/PeizeSun/
+    SparseR-CNN/blob/main/projects/SparseRCNN/sparsercnn/head.py#L258>`_ .
+
+    Args:
+        in_channels (int): The input feature channel.
+            Defaults to 256.
+        feat_channels (int): The inner feature channel.
+            Defaults to 64.
+        out_channels (int, optional): The output feature channel.
+            When not specified, it will be set to `in_channels`
+            by default
+        input_feat_shape (int): The shape of input feature.
+            Defaults to 7.
+        act_cfg (dict): The activation config for DynamicConv.
+        norm_cfg (dict): Config dict for normalization layer. Default
+            layer normalization.
+    """
+
+    def __init__(self,
+                 in_channels=256,
+                 feat_channels=64,
+                 out_channels=None,
+                 input_feat_shape=7,
+                 act_cfg=dict(type='ReLU', inplace=True),
+                 norm_cfg=dict(type='LN')):
+        super(DynamicConv, self).__init__()
+        self.in_channels = in_channels
+        self.feat_channels = feat_channels
+        self.out_channels_raw = out_channels
+        self.input_feat_shape = input_feat_shape
+        self.act_cfg = act_cfg
+        self.norm_cfg = norm_cfg
+        self.out_channels = out_channels if out_channels else in_channels
+
+        self.num_params_in = self.in_channels * self.feat_channels
+        self.num_params_out = self.out_channels * self.feat_channels
+        self.dynamic_layer = nn.Linear(
+            self.in_channels, self.num_params_in + self.num_params_out)
+
+        self.norm_in = build_norm_layer(norm_cfg, self.feat_channels)[1]
+        self.norm_out = build_norm_layer(norm_cfg, self.out_channels)[1]
+
+        self.activation = build_activation_layer(act_cfg)
+
+        num_output = self.out_channels * input_feat_shape**2
+        self.fc_layer = nn.Linear(num_output, self.out_channels)
+        self.fc_norm = build_norm_layer(norm_cfg, self.out_channels)[1]
+
+    def forward(self, param_feature, input_feature):
+        """Forward function for `DynamicConv`.
+
+        Args:
+            param_feature (Tensor): The feature can be used
+                to generate the parameter, has shape
+                (num_all_proposals, in_channels).
+            input_feature (Tensor): Feature that
+                interact with parameters, has shape
+                (num_all_proposals, in_channels, H, W).
+
+        Returns:
+            Tensor: The output feature has shape
+            (num_all_proposals, out_channels).
+        """
+        num_proposals = param_feature.size(0)
+        input_feature = input_feature.view(num_proposals, self.in_channels,
+                                           -1).permute(2, 0, 1)
+
+        input_feature = input_feature.permute(1, 0, 2)
+        parameters = self.dynamic_layer(param_feature)
+
+        param_in = parameters[:, :self.num_params_in].view(
+            -1, self.in_channels, self.feat_channels)
+        param_out = parameters[:, -self.num_params_out:].view(
+            -1, self.feat_channels, self.out_channels)
+
+        # input_feature has shape (num_all_proposals, H*W, in_channels)
+        # param_in has shape (num_all_proposals, in_channels, feat_channels)
+        # feature has shape (num_all_proposals, H*W, feat_channels)
+        features = torch.bmm(input_feature, param_in)
+        features = self.norm_in(features)
+        features = self.activation(features)
+
+        # param_out has shape (batch_size, feat_channels, out_channels)
+        features = torch.bmm(features, param_out)
+        features = self.norm_out(features)
+        features = self.activation(features)
+
+        features = features.flatten(1)
+        features = self.fc_layer(features)
+        features = self.fc_norm(features)
+        features = self.activation(features)
+
+        return features
+
+    def __repr__(self):
+        """str: a string that describes the module"""
+        repr_str = self.__class__.__name__
+        repr_str += f'(in_channels={self.in_channels}, '
+        repr_str += f'feat_channels={self.feat_channels}, '
+        repr_str += f'out_channels={self.out_channels_raw}, '
+        repr_str += f'input_feat_shape={self.input_feat_shape}, '
+        repr_str += f'act_cfg={self.act_cfg}, '
+        repr_str += f'norm_cfg={self.norm_cfg})'
+        return repr_str
diff --git a/python/autoCut/models/swin/mmdet/utils/__init__.py b/python/autoCut/models/swin/mmdet/utils/__init__.py
new file mode 100644
index 0000000..e79ad8c
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/__init__.py
@@ -0,0 +1,5 @@
+from .collect_env import collect_env
+from .logger import get_root_logger
+from .optimizer import DistOptimizerHook
+
+__all__ = ['get_root_logger', 'collect_env', 'DistOptimizerHook']
diff --git a/python/autoCut/models/swin/mmdet/utils/collect_env.py b/python/autoCut/models/swin/mmdet/utils/collect_env.py
new file mode 100644
index 0000000..89c064a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/collect_env.py
@@ -0,0 +1,16 @@
+from mmcv.utils import collect_env as collect_base_env
+from mmcv.utils import get_git_hash
+
+import mmdet
+
+
+def collect_env():
+    """Collect the information of the running environments."""
+    env_info = collect_base_env()
+    env_info['MMDetection'] = mmdet.__version__ + '+' + get_git_hash()[:7]
+    return env_info
+
+
+if __name__ == '__main__':
+    for name, val in collect_env().items():
+        print(f'{name}: {val}')
diff --git a/python/autoCut/models/swin/mmdet/utils/common.py b/python/autoCut/models/swin/mmdet/utils/common.py
new file mode 100644
index 0000000..879a210
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/common.py
@@ -0,0 +1,15 @@
+import torch
+def compute_locations(h, w, stride, device):
+    shifts_x = torch.arange(
+        0, w * stride, step=stride,
+        dtype=torch.float32, device=device
+    )
+    shifts_y = torch.arange(
+        0, h * stride, step=stride,
+        dtype=torch.float32, device=device
+    )
+    shift_y, shift_x = torch.meshgrid(shifts_y, shifts_x)
+    shift_x = shift_x.reshape(-1)
+    shift_y = shift_y.reshape(-1)
+    locations = torch.stack((shift_x, shift_y), dim=1) + stride // 2
+    return locations
diff --git a/python/autoCut/models/swin/mmdet/utils/contextmanagers.py b/python/autoCut/models/swin/mmdet/utils/contextmanagers.py
new file mode 100644
index 0000000..38a6392
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/contextmanagers.py
@@ -0,0 +1,121 @@
+import asyncio
+import contextlib
+import logging
+import os
+import time
+from typing import List
+
+import torch
+
+logger = logging.getLogger(__name__)
+
+DEBUG_COMPLETED_TIME = bool(os.environ.get('DEBUG_COMPLETED_TIME', False))
+
+
+@contextlib.asynccontextmanager
+async def completed(trace_name='',
+                    name='',
+                    sleep_interval=0.05,
+                    streams: List[torch.cuda.Stream] = None):
+    """Async context manager that waits for work to complete on given CUDA
+    streams."""
+    if not torch.cuda.is_available():
+        yield
+        return
+
+    stream_before_context_switch = torch.cuda.current_stream()
+    if not streams:
+        streams = [stream_before_context_switch]
+    else:
+        streams = [s if s else stream_before_context_switch for s in streams]
+
+    end_events = [
+        torch.cuda.Event(enable_timing=DEBUG_COMPLETED_TIME) for _ in streams
+    ]
+
+    if DEBUG_COMPLETED_TIME:
+        start = torch.cuda.Event(enable_timing=True)
+        stream_before_context_switch.record_event(start)
+
+        cpu_start = time.monotonic()
+    logger.debug('%s %s starting, streams: %s', trace_name, name, streams)
+    grad_enabled_before = torch.is_grad_enabled()
+    try:
+        yield
+    finally:
+        current_stream = torch.cuda.current_stream()
+        assert current_stream == stream_before_context_switch
+
+        if DEBUG_COMPLETED_TIME:
+            cpu_end = time.monotonic()
+        for i, stream in enumerate(streams):
+            event = end_events[i]
+            stream.record_event(event)
+
+        grad_enabled_after = torch.is_grad_enabled()
+
+        # observed change of torch.is_grad_enabled() during concurrent run of
+        # async_test_bboxes code
+        assert (grad_enabled_before == grad_enabled_after
+                ), 'Unexpected is_grad_enabled() value change'
+
+        are_done = [e.query() for e in end_events]
+        logger.debug('%s %s completed: %s streams: %s', trace_name, name,
+                     are_done, streams)
+        with torch.cuda.stream(stream_before_context_switch):
+            while not all(are_done):
+                await asyncio.sleep(sleep_interval)
+                are_done = [e.query() for e in end_events]
+                logger.debug(
+                    '%s %s completed: %s streams: %s',
+                    trace_name,
+                    name,
+                    are_done,
+                    streams,
+                )
+
+        current_stream = torch.cuda.current_stream()
+        assert current_stream == stream_before_context_switch
+
+        if DEBUG_COMPLETED_TIME:
+            cpu_time = (cpu_end - cpu_start) * 1000
+            stream_times_ms = ''
+            for i, stream in enumerate(streams):
+                elapsed_time = start.elapsed_time(end_events[i])
+                stream_times_ms += f' {stream} {elapsed_time:.2f} ms'
+            logger.info('%s %s %.2f ms %s', trace_name, name, cpu_time,
+                        stream_times_ms)
+
+
+@contextlib.asynccontextmanager
+async def concurrent(streamqueue: asyncio.Queue,
+                     trace_name='concurrent',
+                     name='stream'):
+    """Run code concurrently in different streams.
+
+    :param streamqueue: asyncio.Queue instance.
+
+    Queue tasks define the pool of streams used for concurrent execution.
+    """
+    if not torch.cuda.is_available():
+        yield
+        return
+
+    initial_stream = torch.cuda.current_stream()
+
+    with torch.cuda.stream(initial_stream):
+        stream = await streamqueue.get()
+        assert isinstance(stream, torch.cuda.Stream)
+
+        try:
+            with torch.cuda.stream(stream):
+                logger.debug('%s %s is starting, stream: %s', trace_name, name,
+                             stream)
+                yield
+                current = torch.cuda.current_stream()
+                assert current == stream
+                logger.debug('%s %s has finished, stream: %s', trace_name,
+                             name, stream)
+        finally:
+            streamqueue.task_done()
+            streamqueue.put_nowait(stream)
diff --git a/python/autoCut/models/swin/mmdet/utils/instances.py b/python/autoCut/models/swin/mmdet/utils/instances.py
new file mode 100644
index 0000000..a7818ac
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/instances.py
@@ -0,0 +1,185 @@
+# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved
+import itertools
+from typing import Any, Dict, List, Tuple, Union
+import torch
+
+
+class Instances:
+    """
+    This class represents a list of instances in an image.
+    It stores the attributes of instances (e.g., boxes, masks, labels, scores) as "fields".
+    All fields must have the same ``__len__`` which is the number of instances.
+
+    All other (non-field) attributes of this class are considered private:
+    they must start with '_' and are not modifiable by a user.
+
+    Some basic usage:
+
+    1. Set/get/check a field:
+
+       .. code-block:: python
+
+          instances.gt_boxes = Boxes(...)
+          print(instances.pred_masks)  # a tensor of shape (N, H, W)
+          print('gt_masks' in instances)
+
+    2. ``len(instances)`` returns the number of instances
+    3. Indexing: ``instances[indices]`` will apply the indexing on all the fields
+       and returns a new :class:`Instances`.
+       Typically, ``indices`` is a integer vector of indices,
+       or a binary mask of length ``num_instances``
+    """
+
+    def __init__(self, image_size: Tuple[int, int], **kwargs: Any):
+        """
+        Args:
+            image_size (height, width): the spatial size of the image.
+            kwargs: fields to add to this `Instances`.
+        """
+        self._image_size = image_size
+        self._fields: Dict[str, Any] = {}
+        for k, v in kwargs.items():
+            self.set(k, v)
+
+    @property
+    def image_size(self) -> Tuple[int, int]:
+        """
+        Returns:
+            tuple: height, width
+        """
+        return self._image_size
+
+    def __setattr__(self, name: str, val: Any) -> None:
+        if name.startswith("_"):
+            super().__setattr__(name, val)
+        else:
+            self.set(name, val)
+
+    def __getattr__(self, name: str) -> Any:
+        if name == "_fields" or name not in self._fields:
+            raise AttributeError("Cannot find field '{}' in the given Instances!".format(name))
+        return self._fields[name]
+
+    def set(self, name: str, value: Any) -> None:
+        """
+        Set the field named `name` to `value`.
+        The length of `value` must be the number of instances,
+        and must agree with other existing fields in this object.
+        """
+        data_len = len(value)
+        if len(self._fields):
+            assert (
+                len(self) == data_len
+            ), "Adding a field of length {} to a Instances of length {}".format(data_len, len(self))
+        self._fields[name] = value
+
+    def has(self, name: str) -> bool:
+        """
+        Returns:
+            bool: whether the field called `name` exists.
+        """
+        return name in self._fields
+
+    def remove(self, name: str) -> None:
+        """
+        Remove the field called `name`.
+        """
+        del self._fields[name]
+
+    def get(self, name: str) -> Any:
+        """
+        Returns the field called `name`.
+        """
+        return self._fields[name]
+
+    def get_fields(self) -> Dict[str, Any]:
+        """
+        Returns:
+            dict: a dict which maps names (str) to data of the fields
+
+        Modifying the returned dict will modify this instance.
+        """
+        return self._fields
+
+    # Tensor-like methods
+    def to(self, *args: Any, **kwargs: Any) -> "Instances":
+        """
+        Returns:
+            Instances: all fields are called with a `to(device)`, if the field has this method.
+        """
+        ret = Instances(self._image_size)
+        for k, v in self._fields.items():
+            if hasattr(v, "to"):
+                v = v.to(*args, **kwargs)
+            ret.set(k, v)
+        return ret
+
+    def __getitem__(self, item: Union[int, slice, torch.BoolTensor]) -> "Instances":
+        """
+        Args:
+            item: an index-like object and will be used to index all the fields.
+
+        Returns:
+            If `item` is a string, return the data in the corresponding field.
+            Otherwise, returns an `Instances` where all fields are indexed by `item`.
+        """
+        if type(item) == int:
+            if item >= len(self) or item < -len(self):
+                raise IndexError("Instances index out of range!")
+            else:
+                item = slice(item, None, len(self))
+
+        ret = Instances(self._image_size)
+        for k, v in self._fields.items():
+            ret.set(k, v[item])
+        return ret
+
+    def __len__(self) -> int:
+        for v in self._fields.values():
+            return len(v)
+        raise NotImplementedError("Empty Instances does not support __len__!")
+
+    def __iter__(self):
+        raise NotImplementedError("`Instances` object is not iterable!")
+
+    @staticmethod
+    def cat(instance_lists: List["Instances"]) -> "Instances":
+        """
+        Args:
+            instance_lists (list[Instances])
+
+        Returns:
+            Instances
+        """
+        assert all(isinstance(i, Instances) for i in instance_lists)
+        assert len(instance_lists) > 0
+        if len(instance_lists) == 1:
+            return instance_lists[0]
+
+        image_size = instance_lists[0].image_size
+        for i in instance_lists[1:]:
+            assert i.image_size == image_size
+        ret = Instances(image_size)
+        for k in instance_lists[0]._fields.keys():
+            values = [i.get(k) for i in instance_lists]
+            v0 = values[0]
+            if isinstance(v0, torch.Tensor):
+                values = torch.cat(values, dim=0)
+            elif isinstance(v0, list):
+                values = list(itertools.chain(*values))
+            elif hasattr(type(v0), "cat"):
+                values = type(v0).cat(values)
+            else:
+                raise ValueError("Unsupported type {} for concatenation".format(type(v0)))
+            ret.set(k, values)
+        return ret
+
+    def __str__(self) -> str:
+        s = self.__class__.__name__ + "("
+        s += "num_instances={}, ".format(len(self))
+        s += "image_height={}, ".format(self._image_size[0])
+        s += "image_width={}, ".format(self._image_size[1])
+        s += "fields=[{}])".format(", ".join((f"{k}: {v}" for k, v in self._fields.items())))
+        return s
+
+    __repr__ = __str__
diff --git a/python/autoCut/models/swin/mmdet/utils/logger.py b/python/autoCut/models/swin/mmdet/utils/logger.py
new file mode 100644
index 0000000..6fc6e6b
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/logger.py
@@ -0,0 +1,19 @@
+import logging
+
+from mmcv.utils import get_logger
+
+
+def get_root_logger(log_file=None, log_level=logging.INFO):
+    """Get root logger.
+
+    Args:
+        log_file (str, optional): File path of log. Defaults to None.
+        log_level (int, optional): The level of logger.
+            Defaults to logging.INFO.
+
+    Returns:
+        :obj:`logging.Logger`: The obtained logger
+    """
+    logger = get_logger(name='mmdet', log_file=log_file, log_level=log_level)
+
+    return logger
diff --git a/python/autoCut/models/swin/mmdet/utils/optimizer.py b/python/autoCut/models/swin/mmdet/utils/optimizer.py
new file mode 100644
index 0000000..2c924c5
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/optimizer.py
@@ -0,0 +1,33 @@
+from mmcv.runner import OptimizerHook, HOOKS
+try:
+    import apex
+except:
+    pass
+
+
+@HOOKS.register_module()
+class DistOptimizerHook(OptimizerHook):
+    """Optimizer hook for distributed training."""
+
+    def __init__(self, update_interval=1, grad_clip=None, coalesce=True, bucket_size_mb=-1, use_fp16=False):
+        self.grad_clip = grad_clip
+        self.coalesce = coalesce
+        self.bucket_size_mb = bucket_size_mb
+        self.update_interval = update_interval
+        self.use_fp16 = use_fp16
+
+    def before_run(self, runner):
+        runner.optimizer.zero_grad()
+
+    def after_train_iter(self, runner):
+        runner.outputs['loss'] /= self.update_interval
+        if self.use_fp16:
+            with apex.amp.scale_loss(runner.outputs['loss'], runner.optimizer) as scaled_loss:
+                scaled_loss.backward()
+        else:
+            runner.outputs['loss'].backward()
+        if self.every_n_iters(runner, self.update_interval):
+            if self.grad_clip is not None:
+                self.clip_grads(runner.model.parameters())
+            runner.optimizer.step()
+            runner.optimizer.zero_grad()
diff --git a/python/autoCut/models/swin/mmdet/utils/profiling.py b/python/autoCut/models/swin/mmdet/utils/profiling.py
new file mode 100644
index 0000000..4be9222
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/profiling.py
@@ -0,0 +1,39 @@
+import contextlib
+import sys
+import time
+
+import torch
+
+if sys.version_info >= (3, 7):
+
+    @contextlib.contextmanager
+    def profile_time(trace_name,
+                     name,
+                     enabled=True,
+                     stream=None,
+                     end_stream=None):
+        """Print time spent by CPU and GPU.
+
+        Useful as a temporary context manager to find sweet spots of code
+        suitable for async implementation.
+        """
+        if (not enabled) or not torch.cuda.is_available():
+            yield
+            return
+        stream = stream if stream else torch.cuda.current_stream()
+        end_stream = end_stream if end_stream else stream
+        start = torch.cuda.Event(enable_timing=True)
+        end = torch.cuda.Event(enable_timing=True)
+        stream.record_event(start)
+        try:
+            cpu_start = time.monotonic()
+            yield
+        finally:
+            cpu_end = time.monotonic()
+            end_stream.record_event(end)
+            end.synchronize()
+            cpu_time = (cpu_end - cpu_start) * 1000
+            gpu_time = start.elapsed_time(end)
+            msg = f'{trace_name} {name} cpu_time {cpu_time:.2f} ms '
+            msg += f'gpu_time {gpu_time:.2f} ms stream {stream}'
+            print(msg, end_stream)
diff --git a/python/autoCut/models/swin/mmdet/utils/util_mixins.py b/python/autoCut/models/swin/mmdet/utils/util_mixins.py
new file mode 100644
index 0000000..69669a3
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/util_mixins.py
@@ -0,0 +1,104 @@
+"""This module defines the :class:`NiceRepr` mixin class, which defines a
+``__repr__`` and ``__str__`` method that only depend on a custom ``__nice__``
+method, which you must define. This means you only have to overload one
+function instead of two.  Furthermore, if the object defines a ``__len__``
+method, then the ``__nice__`` method defaults to something sensible, otherwise
+it is treated as abstract and raises ``NotImplementedError``.
+
+To use simply have your object inherit from :class:`NiceRepr`
+(multi-inheritance should be ok).
+
+This code was copied from the ubelt library: https://github.com/Erotemic/ubelt
+
+Example:
+    >>> # Objects that define __nice__ have a default __str__ and __repr__
+    >>> class Student(NiceRepr):
+    ...    def __init__(self, name):
+    ...        self.name = name
+    ...    def __nice__(self):
+    ...        return self.name
+    >>> s1 = Student('Alice')
+    >>> s2 = Student('Bob')
+    >>> print(f's1 = {s1}')
+    >>> print(f's2 = {s2}')
+    s1 = <Student(Alice)>
+    s2 = <Student(Bob)>
+
+Example:
+    >>> # Objects that define __len__ have a default __nice__
+    >>> class Group(NiceRepr):
+    ...    def __init__(self, data):
+    ...        self.data = data
+    ...    def __len__(self):
+    ...        return len(self.data)
+    >>> g = Group([1, 2, 3])
+    >>> print(f'g = {g}')
+    g = <Group(3)>
+"""
+import warnings
+
+
+class NiceRepr(object):
+    """Inherit from this class and define ``__nice__`` to "nicely" print your
+    objects.
+
+    Defines ``__str__`` and ``__repr__`` in terms of ``__nice__`` function
+    Classes that inherit from :class:`NiceRepr` should redefine ``__nice__``.
+    If the inheriting class has a ``__len__``, method then the default
+    ``__nice__`` method will return its length.
+
+    Example:
+        >>> class Foo(NiceRepr):
+        ...    def __nice__(self):
+        ...        return 'info'
+        >>> foo = Foo()
+        >>> assert str(foo) == '<Foo(info)>'
+        >>> assert repr(foo).startswith('<Foo(info) at ')
+
+    Example:
+        >>> class Bar(NiceRepr):
+        ...    pass
+        >>> bar = Bar()
+        >>> import pytest
+        >>> with pytest.warns(None) as record:
+        >>>     assert 'object at' in str(bar)
+        >>>     assert 'object at' in repr(bar)
+
+    Example:
+        >>> class Baz(NiceRepr):
+        ...    def __len__(self):
+        ...        return 5
+        >>> baz = Baz()
+        >>> assert str(baz) == '<Baz(5)>'
+    """
+
+    def __nice__(self):
+        """str: a "nice" summary string describing this module"""
+        if hasattr(self, '__len__'):
+            # It is a common pattern for objects to use __len__ in __nice__
+            # As a convenience we define a default __nice__ for these objects
+            return str(len(self))
+        else:
+            # In all other cases force the subclass to overload __nice__
+            raise NotImplementedError(
+                f'Define the __nice__ method for {self.__class__!r}')
+
+    def __repr__(self):
+        """str: the string of the module"""
+        try:
+            nice = self.__nice__()
+            classname = self.__class__.__name__
+            return f'<{classname}({nice}) at {hex(id(self))}>'
+        except NotImplementedError as ex:
+            warnings.warn(str(ex), category=RuntimeWarning)
+            return object.__repr__(self)
+
+    def __str__(self):
+        """str: the string of the module"""
+        try:
+            classname = self.__class__.__name__
+            nice = self.__nice__()
+            return f'<{classname}({nice})>'
+        except NotImplementedError as ex:
+            warnings.warn(str(ex), category=RuntimeWarning)
+            return object.__repr__(self)
diff --git a/python/autoCut/models/swin/mmdet/utils/util_random.py b/python/autoCut/models/swin/mmdet/utils/util_random.py
new file mode 100644
index 0000000..e313e99
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/utils/util_random.py
@@ -0,0 +1,33 @@
+"""Helpers for random number generators."""
+import numpy as np
+
+
+def ensure_rng(rng=None):
+    """Coerces input into a random number generator.
+
+    If the input is None, then a global random state is returned.
+
+    If the input is a numeric value, then that is used as a seed to construct a
+    random state. Otherwise the input is returned as-is.
+
+    Adapted from [1]_.
+
+    Args:
+        rng (int | numpy.random.RandomState | None):
+            if None, then defaults to the global rng. Otherwise this can be an
+            integer or a RandomState class
+    Returns:
+        (numpy.random.RandomState) : rng -
+            a numpy random number generator
+
+    References:
+        .. [1] https://gitlab.kitware.com/computer-vision/kwarray/blob/master/kwarray/util_random.py#L270  # noqa: E501
+    """
+
+    if rng is None:
+        rng = np.random.mtrand._rand
+    elif isinstance(rng, int):
+        rng = np.random.RandomState(rng)
+    else:
+        rng = rng
+    return rng
diff --git a/python/autoCut/models/swin/mmdet/version.py b/python/autoCut/models/swin/mmdet/version.py
new file mode 100644
index 0000000..a3b741a
--- /dev/null
+++ b/python/autoCut/models/swin/mmdet/version.py
@@ -0,0 +1,19 @@
+# Copyright (c) Open-MMLab. All rights reserved.
+
+__version__ = '2.11.0'
+short_version = __version__
+
+
+def parse_version_info(version_str):
+    version_info = []
+    for x in version_str.split('.'):
+        if x.isdigit():
+            version_info.append(int(x))
+        elif x.find('rc') != -1:
+            patch_version = x.split('rc')
+            version_info.append(int(patch_version[0]))
+            version_info.append(f'rc{patch_version[1]}')
+    return tuple(version_info)
+
+
+version_info = parse_version_info(__version__)
diff --git a/python/autoCut/models/swin/requirements.txt b/python/autoCut/models/swin/requirements.txt
new file mode 100644
index 0000000..6981bd7
--- /dev/null
+++ b/python/autoCut/models/swin/requirements.txt
@@ -0,0 +1,4 @@
+-r requirements/build.txt
+-r requirements/optional.txt
+-r requirements/runtime.txt
+-r requirements/tests.txt
diff --git a/python/autoCut/models/swin/requirements/build.txt b/python/autoCut/models/swin/requirements/build.txt
new file mode 100644
index 0000000..8155829
--- /dev/null
+++ b/python/autoCut/models/swin/requirements/build.txt
@@ -0,0 +1,3 @@
+# These must be installed before building mmdetection
+cython
+numpy
diff --git a/python/autoCut/models/swin/requirements/docs.txt b/python/autoCut/models/swin/requirements/docs.txt
new file mode 100644
index 0000000..89fbf86
--- /dev/null
+++ b/python/autoCut/models/swin/requirements/docs.txt
@@ -0,0 +1,4 @@
+recommonmark
+sphinx
+sphinx_markdown_tables
+sphinx_rtd_theme
diff --git a/python/autoCut/models/swin/requirements/optional.txt b/python/autoCut/models/swin/requirements/optional.txt
new file mode 100644
index 0000000..35b5242
--- /dev/null
+++ b/python/autoCut/models/swin/requirements/optional.txt
@@ -0,0 +1,6 @@
+albumentations>=0.3.2
+cityscapesscripts
+imagecorruptions
+mmlvis
+scipy
+sklearn
diff --git a/python/autoCut/models/swin/requirements/readthedocs.txt b/python/autoCut/models/swin/requirements/readthedocs.txt
new file mode 100644
index 0000000..0542bfc
--- /dev/null
+++ b/python/autoCut/models/swin/requirements/readthedocs.txt
@@ -0,0 +1,3 @@
+mmcv
+torch
+torchvision
diff --git a/python/autoCut/models/swin/requirements/runtime.txt b/python/autoCut/models/swin/requirements/runtime.txt
new file mode 100644
index 0000000..cf106f0
--- /dev/null
+++ b/python/autoCut/models/swin/requirements/runtime.txt
@@ -0,0 +1,6 @@
+matplotlib
+pycocotools
+numpy
+six
+terminaltables
+timm
diff --git a/python/autoCut/models/swin/requirements/tests.txt b/python/autoCut/models/swin/requirements/tests.txt
new file mode 100644
index 0000000..5f3de01
--- /dev/null
+++ b/python/autoCut/models/swin/requirements/tests.txt
@@ -0,0 +1,13 @@
+asynctest
+codecov
+flake8
+interrogate
+isort==4.3.21
+# Note: used for kwarray.group_items, this may be ported to mmcv in the future.
+kwarray
+onnx==1.7.0
+onnxruntime==1.5.1
+pytest
+ubelt
+xdoctest>=0.10.0
+yapf
diff --git a/python/autoCut/models/swin/setup.cfg b/python/autoCut/models/swin/setup.cfg
new file mode 100644
index 0000000..78eb65e
--- /dev/null
+++ b/python/autoCut/models/swin/setup.cfg
@@ -0,0 +1,13 @@
+[isort]
+line_length = 79
+multi_line_output = 0
+known_standard_library = setuptools
+known_first_party = mmdet
+known_third_party = PIL,asynctest,cityscapesscripts,cv2,gather_models,matplotlib,mmcv,numpy,onnx,onnxruntime,pycocotools,pytest,seaborn,six,terminaltables,torch,ts
+no_lines_before = STDLIB,LOCALFOLDER
+default_section = THIRDPARTY
+
+[yapf]
+BASED_ON_STYLE = pep8
+BLANK_LINE_BEFORE_NESTED_CLASS_OR_DEF = true
+SPLIT_BEFORE_EXPRESSION_AFTER_OPENING_PAREN = true
diff --git a/python/autoCut/models/swin/setup.py b/python/autoCut/models/swin/setup.py
new file mode 100755
index 0000000..55eea6b
--- /dev/null
+++ b/python/autoCut/models/swin/setup.py
@@ -0,0 +1,161 @@
+#!/usr/bin/env python
+import os
+from setuptools import find_packages, setup
+
+import torch
+from torch.utils.cpp_extension import (BuildExtension, CppExtension,
+                                       CUDAExtension)
+
+
+def readme():
+    with open('README.md', encoding='utf-8') as f:
+        content = f.read()
+    return content
+
+
+version_file = 'mmdet/version.py'
+
+
+def get_version():
+    with open(version_file, 'r') as f:
+        exec(compile(f.read(), version_file, 'exec'))
+    return locals()['__version__']
+
+
+def make_cuda_ext(name, module, sources, sources_cuda=[]):
+
+    define_macros = []
+    extra_compile_args = {'cxx': []}
+
+    if torch.cuda.is_available() or os.getenv('FORCE_CUDA', '0') == '1':
+        define_macros += [('WITH_CUDA', None)]
+        extension = CUDAExtension
+        extra_compile_args['nvcc'] = [
+            '-D__CUDA_NO_HALF_OPERATORS__',
+            '-D__CUDA_NO_HALF_CONVERSIONS__',
+            '-D__CUDA_NO_HALF2_OPERATORS__',
+        ]
+        sources += sources_cuda
+    else:
+        print(f'Compiling {name} without CUDA')
+        extension = CppExtension
+
+    return extension(
+        name=f'{module}.{name}',
+        sources=[os.path.join(*module.split('.'), p) for p in sources],
+        define_macros=define_macros,
+        extra_compile_args=extra_compile_args)
+
+
+def parse_requirements(fname='requirements.txt', with_version=True):
+    """Parse the package dependencies listed in a requirements file but strips
+    specific versioning information.
+
+    Args:
+        fname (str): path to requirements file
+        with_version (bool, default=False): if True include version specs
+
+    Returns:
+        List[str]: list of requirements items
+
+    CommandLine:
+        python -c "import setup; print(setup.parse_requirements())"
+    """
+    import sys
+    from os.path import exists
+    import re
+    require_fpath = fname
+
+    def parse_line(line):
+        """Parse information from a line in a requirements text file."""
+        if line.startswith('-r '):
+            # Allow specifying requirements in other files
+            target = line.split(' ')[1]
+            for info in parse_require_file(target):
+                yield info
+        else:
+            info = {'line': line}
+            if line.startswith('-e '):
+                info['package'] = line.split('#egg=')[1]
+            elif '@git+' in line:
+                info['package'] = line
+            else:
+                # Remove versioning from the package
+                pat = '(' + '|'.join(['>=', '==', '>']) + ')'
+                parts = re.split(pat, line, maxsplit=1)
+                parts = [p.strip() for p in parts]
+
+                info['package'] = parts[0]
+                if len(parts) > 1:
+                    op, rest = parts[1:]
+                    if ';' in rest:
+                        # Handle platform specific dependencies
+                        # http://setuptools.readthedocs.io/en/latest/setuptools.html#declaring-platform-specific-dependencies
+                        version, platform_deps = map(str.strip,
+                                                     rest.split(';'))
+                        info['platform_deps'] = platform_deps
+                    else:
+                        version = rest  # NOQA
+                    info['version'] = (op, version)
+            yield info
+
+    def parse_require_file(fpath):
+        with open(fpath, 'r') as f:
+            for line in f.readlines():
+                line = line.strip()
+                if line and not line.startswith('#'):
+                    for info in parse_line(line):
+                        yield info
+
+    def gen_packages_items():
+        if exists(require_fpath):
+            for info in parse_require_file(require_fpath):
+                parts = [info['package']]
+                if with_version and 'version' in info:
+                    parts.extend(info['version'])
+                if not sys.version.startswith('3.4'):
+                    # apparently package_deps are broken in 3.4
+                    platform_deps = info.get('platform_deps')
+                    if platform_deps is not None:
+                        parts.append(';' + platform_deps)
+                item = ''.join(parts)
+                yield item
+
+    packages = list(gen_packages_items())
+    return packages
+
+
+if __name__ == '__main__':
+    setup(
+        name='mmdet',
+        version=get_version(),
+        description='OpenMMLab Detection Toolbox and Benchmark',
+        long_description=readme(),
+        long_description_content_type='text/markdown',
+        author='OpenMMLab',
+        author_email='openmmlab@gmail.com',
+        keywords='computer vision, object detection',
+        url='https://github.com/open-mmlab/mmdetection',
+        packages=find_packages(exclude=('configs', 'tools', 'demo')),
+        classifiers=[
+            'Development Status :: 5 - Production/Stable',
+            'License :: OSI Approved :: Apache Software License',
+            'Operating System :: OS Independent',
+            'Programming Language :: Python :: 3',
+            'Programming Language :: Python :: 3.6',
+            'Programming Language :: Python :: 3.7',
+            'Programming Language :: Python :: 3.8',
+        ],
+        license='Apache License 2.0',
+        setup_requires=parse_requirements('requirements/build.txt'),
+        tests_require=parse_requirements('requirements/tests.txt'),
+        install_requires=parse_requirements('requirements/runtime.txt'),
+        extras_require={
+            'all': parse_requirements('requirements.txt'),
+            'tests': parse_requirements('requirements/tests.txt'),
+            'build': parse_requirements('requirements/build.txt'),
+            'optional': parse_requirements('requirements/optional.txt'),
+        },
+        ext_modules=[],
+        cmdclass={'build_ext': BuildExtension},
+        zip_safe=False)
diff --git a/python/autoCut/utils.py b/python/autoCut/utils.py
new file mode 100644
index 0000000..b3cdd8a
--- /dev/null
+++ b/python/autoCut/utils.py
@@ -0,0 +1,12 @@
+import torch 
+from global_settings import DEBUG
+from logger import logger
+
+
+def find_device(cpu: bool = False):
+    if cpu:
+        logger.info("Using CPU")
+        return torch.device("cpu")
+    else:
+        logger.info(("Using GPU %s", torch.cuda.get_device_properties()) if torch.cuda.is_available() else "Using CPU")
+        return torch.device("cuda" if torch.cuda.is_available() else "cpu")

From 9b68653b13279e70d367a1e2bf2b09dca249d0ea Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Wed, 19 Mar 2025 09:32:05 +0000
Subject: [PATCH 07/33] auto cut

---
 .gitignore                        |   1 +
 PretextView.cpp                   | 496 ++++++++++++++++++++----------
 python/autoCut/global_settings.py |   6 +-
 src/ai_cut_json_parse.h           |   4 +-
 src/auto_curation_state.h         |   4 +-
 src/copy_texture.cpp              |  45 ++-
 src/copy_texture.h                |  10 +-
 src/frag_cut_calculation.h        | 305 ++++++++++++++++++
 src/genomeData.h                  |  24 ++
 src/hic_figure.h                  |  91 ++----
 src/utilsPretextView.h            |  24 ++
 11 files changed, 756 insertions(+), 254 deletions(-)
 create mode 100644 src/frag_cut_calculation.h

diff --git a/.gitignore b/.gitignore
index 4deba74..a3fcf7c 100644
--- a/.gitignore
+++ b/.gitignore
@@ -75,3 +75,4 @@ subprojects/libtorch(not needed in this version)/
 hic_figures/
 auto_cut_output/
 python/autoCut/models/swin/README.md
+python/autoCut/models/swin/mmcv_custom/
diff --git a/PretextView.cpp b/PretextView.cpp
index 4e151b6..b9f8980 100644
--- a/PretextView.cpp
+++ b/PretextView.cpp
@@ -105,6 +105,7 @@ SOFTWARE.
 #include "showWindowData.h"
 #include "hic_figure.h"
 #include "ai_cut_json_parse.h"
+#include "frag_cut_calculation.h"
 
 #include "shaderSource.h"
 /*
@@ -662,10 +663,10 @@ Yahs_sorting = 0;
 
 global_variable
 u32 
-AI_cutting = 0;
+Auto_cutting = 0;
 
-global_variable s32 ai_cutting_button = 0; 
-global_variable s32 yahs_sort_button = 0;
+global_variable s32 auto_cut_button = 0; 
+global_variable s32 auto_sort_button = 0;
 
 
 global_function
@@ -731,8 +732,6 @@ void
 restore_settings_after_copy(const f32 *original_control_points);
 
 
-#define Max_Number_of_Contigs 4096
-
 
 global_variable
 meta_data *
@@ -799,7 +798,7 @@ global_variable auto auto_curation_state = AutoCurationState();
 
 global_function
 void
-UpdateContigsFromMapState()  // todo reading 从map的状态更新contigs
+UpdateContigsFromMapState()  //  reading 从map的状态更新contigs
 {
     u32 lastScaffID = Map_State->scaffIds[0];       // 第一个scaff的编号
     u32 scaffId = lastScaffID ? 1 : 0;              // 
@@ -821,13 +820,17 @@ UpdateContigsFromMapState()  // todo reading 从map的状态更新contigs
         ++length; // current fragment length
 
         pixelIdx = index + 1;   // 像素点编号， 加一因为第一个已经用来初始化了
-        u32 id = Map_State->originalContigIds[pixelIdx];  // 像素点的 contig id
+        u32 id = Map_State->originalContigIds[pixelIdx];  // 像素点的 original contig id, 这里 不使用 % Max_Number_of_Contigs的值是为了区分后面cut之后的片段
         u32 coord = Map_State->contigRelCoords[pixelIdx]; // 像素点的局部坐标
         
-        if (id != lastId || (inverted && coord != (lastCoord - 1)) || (!inverted && coord != (lastCoord + 1))) // 如果不是一个连续片段
+        if (
+            id != lastId || 
+            (inverted && coord != (lastCoord - 1)) || 
+            (!inverted && coord != (lastCoord + 1)) 
+        ) // 如果不是一个连续片段
         {   
-            Original_Contigs[lastId].contigMapPixels[Original_Contigs[lastId].nContigs] = pixelIdx - 1 - (length >> 1);   // update Original_Contigs: contigMapPixels
-            Original_Contigs[lastId].nContigs++; // update Original_Contigs: nContigs, contigMapPixels
+            Original_Contigs[lastId % Max_Number_of_Contigs].contigMapPixels[Original_Contigs[lastId%Max_Number_of_Contigs].nContigs] = pixelIdx - 1 - (length >> 1);   // update Original_Contigs: contigMapPixels
+            Original_Contigs[lastId % Max_Number_of_Contigs].nContigs++; // update Original_Contigs: nContigs, contigMapPixels
 
             contig *last_cont = Contigs->contigs_arr + contigPtr; // 获取上一个contig的指针， 并且给contigPtr + 1
             contigPtr++;
@@ -865,7 +868,7 @@ UpdateContigsFromMapState()  // todo reading 从map的状态更新contigs
     if (contigPtr < Max_Number_of_Contigs) //  contigptr 小于 Number_of_Original_Contigs
     // 更新最后一个contig的最后一个片段信息
     {
-        (Original_Contigs + lastId)->contigMapPixels[(Original_Contigs + lastId)->nContigs++] = pixelIdx - 1 - (length >> 1); 
+        (Original_Contigs + lastId % Max_Number_of_Contigs)->contigMapPixels[(Original_Contigs + lastId % Max_Number_of_Contigs)->nContigs++] = pixelIdx - 1 - (length >> 1); 
 
         ++length;
         contig *cont = Contigs->contigs_arr + contigPtr++;
@@ -901,7 +904,7 @@ RebuildContig(u32 pixel)
     for (;;)
     {
         u32 contigId = Map_State->contigIds[pixel];
-        u32 origContigId = Map_State->originalContigIds[pixel];
+        u32 origContigId = Map_State->get_original_contig_id(pixel);
 
         u32 top = (u32)pixel;
         while (top && (Map_State->contigIds[top - 1] == contigId)) --top;
@@ -919,7 +922,7 @@ RebuildContig(u32 pixel)
         u08 fragmented = 0;
         ForLoop(Number_of_Pixels_1D)
         {
-            if ((Map_State->contigIds[index] != contigId) && (Map_State->originalContigIds[index] == origContigId))
+            if ((Map_State->contigIds[index] != contigId) && (Map_State->get_original_contig_id(index) == origContigId))
             {
                 fragmented = 1;
                 break;
@@ -934,7 +937,7 @@ RebuildContig(u32 pixel)
                 u32 otherPixel = 0;
                 ForLoop(Number_of_Pixels_1D)
                 {
-                    if ((Map_State->originalContigIds[index] == origContigId) && (Map_State->contigRelCoords[index] == (contigTopCoord - 1)))
+                    if ((Map_State->get_original_contig_id(index) == origContigId) && (Map_State->contigRelCoords[index] == (contigTopCoord - 1)))
                     {
                         otherPixel = index;
                         break;
@@ -968,7 +971,7 @@ RebuildContig(u32 pixel)
                 u32 otherPixel = 0;
                 ForLoop(Number_of_Pixels_1D)
                 {
-                    if ((Map_State->originalContigIds[index] == origContigId) && (Map_State->contigRelCoords[index] == (contigBottomCoord + 1)))
+                    if ((Map_State->get_original_contig_id(index) == origContigId) && (Map_State->contigRelCoords[index] == (contigBottomCoord + 1)))
                     {
                         otherPixel = index;
                         break;
@@ -3520,7 +3523,7 @@ Render() {
                     (char *)"K: exit",
                     (char *)"S: clear select area",
                     (char *)"Q/W: quit/redo edit",
-                    (char *)"Space: YaHS sort",
+                    (char *)"Space: Auto sort",
                     (char *)"Left Click: select/unselect the fragment",
                     (char *)selectedFragmentsStr.c_str(),
                     (char *)start_and_end_pixel_str.c_str()
@@ -3834,8 +3837,8 @@ Render() {
             textBoxHeight *= (3.0f + (f32)nExtra);
             textBoxHeight += (2.0f + (f32)nExtra);
 
-            u32 id1 = Map_State->originalContigIds[Tool_Tip_Move.pixels.x];
-            u32 id2 = Map_State->originalContigIds[Tool_Tip_Move.pixels.y];
+            u32 id1 = Map_State->get_original_contig_id(Tool_Tip_Move.pixels.x);
+            u32 id2 = Map_State->get_original_contig_id(Tool_Tip_Move.pixels.y);
             u32 coord1 = Map_State->contigRelCoords[Tool_Tip_Move.pixels.x];
             u32 coord2 = Map_State->contigRelCoords[Tool_Tip_Move.pixels.y];
             
@@ -4097,7 +4100,7 @@ Render() {
                     pix1 = pix1 ? pix1 - 1 : (pix2 < (Number_of_Pixels_1D - 1) ? pix2 + 1 : pix2);
                 }
 
-                original_contig *cont = Original_Contigs + Map_State->originalContigIds[pix1];
+                original_contig *cont = Original_Contigs + Map_State->get_original_contig_id(pix1);
 
                 u32 nPixels = Number_of_Pixels_1D;
                 f64 bpPerPixel = (f64)Total_Genome_Length / (f64)nPixels;
@@ -4116,7 +4119,7 @@ Render() {
                 if (!line1Done)
                 {
                     f64 bpEnd = bpPerPixel * (f64)Map_State->contigRelCoords[pix2];
-                    original_contig *cont2 = Original_Contigs + Map_State->originalContigIds[pix2];
+                    original_contig *cont2 = Original_Contigs + Map_State->get_original_contig_id(pix2);
                     stbsp_snprintf(line1, 64, "%s[%$.2fbp] to %s[%$.2fbp]", cont->name, bpStart, cont2->name, bpEnd);
                     if (Edit_Pixels.editing)
                     {
@@ -4336,12 +4339,12 @@ Render() {
             glUseProgram(UI_Shader->shaderProgram);
             glUniformMatrix4fv(Flat_Shader->matLocation, 1, GL_FALSE, textNormalMat);
             glViewport(0, 0, (s32)width, (s32)height);
-            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "YaHS Sorting...", 0);
+            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "Auto Sorting...", 0);
 
             ChangeSize((s32)width, (s32)height);
         }
 
-        if (AI_cutting)
+        if (Auto_cutting)
         {
             u32 colour = glfonsRGBA(Theme_Colour.r, Theme_Colour.g, Theme_Colour.b, Theme_Colour.a);
 
@@ -4354,7 +4357,7 @@ Render() {
             glUseProgram(UI_Shader->shaderProgram);
             glUniformMatrix4fv(Flat_Shader->matLocation, 1, GL_FALSE, textNormalMat);
             glViewport(0, 0, (s32)width, (s32)height);
-            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "AI cutting...", 0);
+            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "Auto cutting...", 0);
 
             ChangeSize((s32)width, (s32)height);
         }
@@ -4588,6 +4591,142 @@ TestFile(const char *fileName, u64 *fileSize = 0)
     return(file); // 仅仅返回正确的file指针，否则返回空指针
 }
 
+
+
+global_function
+void 
+add_graph_to_extensions(
+    memory_arena *arena, 
+    const u32* graph_name, 
+    const u32* graph_data)
+{
+    
+    graph *gph = PushStructP(arena, graph);   // create the size to store the extension data
+    extension_node *node = PushStructP(arena, extension_node);
+    u08 *dataPlusName = PushArrayP(arena, u08, ((sizeof(u32) * Number_of_Pixels_1D) + sizeof(gph->name) )); 
+    gph->data = (u32 *)(dataPlusName + sizeof(gph->name)); 
+    u32 *namePtr = (u32 *)graph_name;  
+    for (u32 i = 0; i < ArrayCount(gph->name); i++ )  // get the graph name
+    {
+        gph->name[i] = *(namePtr + i);
+    }
+    for (u32 i = 0; i < Number_of_Pixels_1D; i++)  // get the graph data
+    {
+        gph->data[i] = *(graph_data + i);
+    }
+    node->type = extension_graph;      // assign the gph to node
+    node->extension = gph;
+    AddExtension(node);     // add node to extension
+    
+    return ;
+}
+
+
+
+global_function
+void 
+push_extensions_to_opengl(memory_arena *arena, u32 added_index = 0)
+{   
+    u32 exIndex = 0;
+    // TraverseLinkedList(Extensions.head, extension_node)
+
+    extension_node* node = Extensions.head ;
+    while (added_index--)
+    {
+        node = node->next;
+        exIndex ++ ;
+    }
+    for ( ; node; node = node->next) {
+        
+        switch (node->type) 
+        {
+            case extension_graph:
+                {
+                    graph *gph = (graph *)node->extension; // get the graph ptr and assigned to gph
+#define DefaultGraphScale 0.2f
+#define DefaultGraphBase 32.0f
+#define DefaultGraphLineSize 1.0f
+#define DefaultGraphColour {0.1f, 0.8f, 0.7f, 1.0f}
+                    gph->scale = DefaultGraphScale;
+                    gph->base = DefaultGraphBase;
+                    gph->lineSize = DefaultGraphLineSize;
+                    gph->colour = DefaultGraphColour;
+                    gph->on = 0;
+
+                    gph->shader = PushStructP(arena, editable_plot_shader);
+                    gph->shader->shaderProgram = CreateShader(FragmentSource_EditablePlot.c_str(), VertexSource_EditablePlot.c_str(), GeometrySource_EditablePlot.c_str());
+
+                    glUseProgram(gph->shader->shaderProgram);
+                    glBindFragDataLocation(gph->shader->shaderProgram, 0, "outColor");
+                    gph->shader->matLocation = glGetUniformLocation(gph->shader->shaderProgram, "matrix");
+                    gph->shader->colorLocation = glGetUniformLocation(gph->shader->shaderProgram, "color");
+                    gph->shader->yScaleLocation = glGetUniformLocation(gph->shader->shaderProgram, "yscale");
+                    gph->shader->yTopLocation = glGetUniformLocation(gph->shader->shaderProgram, "ytop");
+                    gph->shader->lineSizeLocation = glGetUniformLocation(gph->shader->shaderProgram, "linewidth");
+
+                    glUniform1i(glGetUniformLocation(gph->shader->shaderProgram, "pixrearrangelookup"), 3);
+                    glUniform1i(glGetUniformLocation(gph->shader->shaderProgram, "yvalues"), 4 + (s32)exIndex);
+
+                    u32 nValues = Number_of_Pixels_1D;
+                    auto* xValues = new f32[nValues]; // allocate memory for xValues, actually, x is the coordinate of the pixel
+                    auto* yValues = new f32[nValues];
+                    
+                    u32 max = 0;
+                    ForLoop(Number_of_Pixels_1D)
+                    {
+                        max = my_Max(max, gph->data[index]);
+                    }
+
+                    ForLoop(Number_of_Pixels_1D)
+                    {
+                        xValues[index] = (f32)index;
+                        yValues[index] = (f32)gph->data[index] / (f32)max ;   // normalise the data
+                    }
+
+                    glActiveTexture(GL_TEXTURE4 + exIndex++);
+
+                    GLuint yVal, yValTex;
+
+                    // generate a buffer object named yVal and save data to yVal
+                    glGenBuffers(1, &yVal);
+                    glBindBuffer(GL_TEXTURE_BUFFER, yVal);
+                    glBufferData(GL_TEXTURE_BUFFER, sizeof(f32) * nValues, yValues, GL_STATIC_DRAW);
+
+                    // add texture and link to the buffer object
+                    glGenTextures(1, &yValTex);
+                    glBindTexture(GL_TEXTURE_BUFFER, yValTex);
+                    glTexBuffer(GL_TEXTURE_BUFFER, GL_R32F, yVal);
+
+                    gph->shader->yValuesBuffer = yVal;
+                    gph->shader->yValuesBufferTex = yValTex;
+                    
+                    // add the vertext data into buffer
+                    glGenBuffers(1, &gph->vbo);
+                    glBindBuffer(GL_ARRAY_BUFFER, gph->vbo);
+                    glBufferData(GL_ARRAY_BUFFER, sizeof(f32) * nValues, xValues, GL_STATIC_DRAW);
+
+                    // gen the vertex array object
+                    glGenVertexArrays(1, &gph->vao);
+                    glBindVertexArray(gph->vao);
+
+                    GLuint posAttrib = (GLuint)glGetAttribLocation(gph->shader->shaderProgram, "position");
+                    glEnableVertexAttribArray(posAttrib);
+                    glVertexAttribPointer(posAttrib, 1, GL_FLOAT, GL_FALSE, 0, 0);
+
+                    delete[] xValues;
+                    delete[] yValues;
+
+                    glActiveTexture(GL_TEXTURE0);
+                }
+                break;
+        }
+    }
+
+    return ;
+}
+
+
+
 global_function
 u08
 LoadState(u64 headerHash, char *path = 0);
@@ -5225,93 +5364,7 @@ LoadFile(const char *filePath, memory_arena *arena, char **fileName, u64 *header
 
     // Extensions
     {
-        u32 exIndex = 0;
-        // TraverseLinkedList(Extensions.head, extension_node)
-        for (extension_node* node = Extensions.head; node; node = node->next) {
-            
-            switch (node->type)
-            {
-                case extension_graph:
-                    {
-                        graph *gph = (graph *)node->extension; // get the graph ptr and assigned to gph
-#define DefaultGraphScale 0.2f
-#define DefaultGraphBase 32.0f
-#define DefaultGraphLineSize 1.0f
-#define DefaultGraphColour {0.1f, 0.8f, 0.7f, 1.0f}
-                        gph->scale = DefaultGraphScale;
-                        gph->base = DefaultGraphBase;
-                        gph->lineSize = DefaultGraphLineSize;
-                        gph->colour = DefaultGraphColour;
-                        gph->on = 0;
-
-                        gph->shader = PushStructP(arena, editable_plot_shader);
-                        gph->shader->shaderProgram = CreateShader(FragmentSource_EditablePlot.c_str(), VertexSource_EditablePlot.c_str(), GeometrySource_EditablePlot.c_str());
-
-                        glUseProgram(gph->shader->shaderProgram);
-                        glBindFragDataLocation(gph->shader->shaderProgram, 0, "outColor");
-                        gph->shader->matLocation = glGetUniformLocation(gph->shader->shaderProgram, "matrix");
-                        gph->shader->colorLocation = glGetUniformLocation(gph->shader->shaderProgram, "color");
-                        gph->shader->yScaleLocation = glGetUniformLocation(gph->shader->shaderProgram, "yscale");
-                        gph->shader->yTopLocation = glGetUniformLocation(gph->shader->shaderProgram, "ytop");
-                        gph->shader->lineSizeLocation = glGetUniformLocation(gph->shader->shaderProgram, "linewidth");
-
-                        glUniform1i(glGetUniformLocation(gph->shader->shaderProgram, "pixrearrangelookup"), 3);
-                        glUniform1i(glGetUniformLocation(gph->shader->shaderProgram, "yvalues"), 4 + (s32)exIndex);
-
-                        u32 nValues = Number_of_Pixels_1D;
-                        auto* xValues = new f32[nValues]; // allocate memory for xValues, actually, x is the coordinate of the pixel
-                        auto* yValues = new f32[nValues];
-                        
-                        u32 max = 0;
-                        ForLoop(Number_of_Pixels_1D)
-                        {
-                            max = my_Max(max, gph->data[index]);
-                        }
-
-                        ForLoop(Number_of_Pixels_1D)
-                        {
-                            xValues[index] = (f32)index;
-                            yValues[index] = (f32)gph->data[index] / (f32)max ;   // normalise the data
-                        }
-
-                        glActiveTexture(GL_TEXTURE4 + exIndex++);
-
-                        GLuint yVal, yValTex;
-
-                        // generate a buffer object named yVal and save data to yVal
-                        glGenBuffers(1, &yVal);
-                        glBindBuffer(GL_TEXTURE_BUFFER, yVal);
-                        glBufferData(GL_TEXTURE_BUFFER, sizeof(f32) * nValues, yValues, GL_STATIC_DRAW);
-
-                        // add texture and link to the buffer object
-                        glGenTextures(1, &yValTex);
-                        glBindTexture(GL_TEXTURE_BUFFER, yValTex);
-                        glTexBuffer(GL_TEXTURE_BUFFER, GL_R32F, yVal);
-
-                        gph->shader->yValuesBuffer = yVal;
-                        gph->shader->yValuesBufferTex = yValTex;
-                        
-                        // add the vertext data into buffer
-                        glGenBuffers(1, &gph->vbo);
-                        glBindBuffer(GL_ARRAY_BUFFER, gph->vbo);
-                        glBufferData(GL_ARRAY_BUFFER, sizeof(f32) * nValues, xValues, GL_STATIC_DRAW);
-
-                        // gen the vertex array object
-                        glGenVertexArrays(1, &gph->vao);
-                        glBindVertexArray(gph->vao);
-
-                        GLuint posAttrib = (GLuint)glGetAttribLocation(gph->shader->shaderProgram, "position");
-                        glEnableVertexAttribArray(posAttrib);
-                        glVertexAttribPointer(posAttrib, 1, GL_FLOAT, GL_FALSE, 0, 0);
-
-                        delete[] xValues;
-                        delete[] yValues;
-
-                        glActiveTexture(GL_TEXTURE0);
-                    }
-                    break;
-            }
-        }
+        push_extensions_to_opengl(arena, 0);
     }
 
 #ifdef Internal
@@ -6195,6 +6248,69 @@ RearrangeMap(       // NOTE: VERY IMPORTANT
 }
 
 
+
+global_function
+void 
+BreakMap(
+    const u32& loc, 
+    const u32 ignore_len = 10 // cut点到开头或者结尾的长度不足ignore_len的contig不会被切断
+)
+{  
+    if (loc >= Number_of_Pixels_1D)
+    {
+        char buff[512];
+        snprintf(buff, sizeof(buff), "Error: loc (%d) should be smaller than number of pixels (%d)", loc, Number_of_Pixels_1D);
+        MY_CHECK(buff);
+        assert(0);
+    }
+    u32 original_contig_id = Map_State->originalContigIds[loc];
+    u32 contig_id = Map_State->contigIds[loc];
+    u32 contig_rel_coord = Map_State->contigRelCoords[loc];
+
+    s32 ptr_left = (s32)loc, ptr_right = (s32)loc;
+    u08 inversed = IsContigInverted(contig_id);
+    while (
+        --ptr_left >= 0 && 
+        Map_State->contigIds[ptr_left] == contig_id && 
+        (Map_State->contigRelCoords[ptr_left] ==  Map_State->contigRelCoords[ptr_left+1]+ (inversed ? +1 : -1)) ) {};
+    while (
+        ++ptr_right < Number_of_Pixels_1D && 
+        Map_State->contigIds[ptr_right] == contig_id && 
+        (Map_State->contigRelCoords[ptr_right - 1] ==  Map_State->contigRelCoords[ptr_right]+ (inversed ? +1 : -1)) ) {};
+    
+    if ((loc - ptr_left) < ignore_len || (ptr_right - loc) < ignore_len )  
+    {   
+        char buff[256];
+        snprintf(buff, sizeof(buff), "Error: original_contig_id %d current_contig_id %d, pixel range: [%d, cut(%d), %d], left(%d), right(%d), smaller than ignore_len (%d) \n", original_contig_id>Max_Number_of_Contigs? original_contig_id%Max_Number_of_Contigs:original_contig_id, contig_id, ptr_left, loc, ptr_right, loc - ptr_left, ptr_right - loc, ignore_len);
+        std::cout << buff << std::endl;
+        return;
+    } 
+
+    ptr_left++; ptr_right--;
+    // if ((inversed && Map_State->contigRelCoords[ptr_right] != 0 ) || 
+    //     (!inversed && Map_State->contigRelCoords[ptr_left] != 0))
+    // {
+    //     char buff[256];
+    //     snprintf(buff, sizeof(buff), "Error: contig %d, pixel range: [%d, %d] %s inversed, but the start real coor is not 0 (%d instead)\n", contig_id, ptr_left, ptr_right, inversed ? "(is)" : "(isn\'t)", inversed ? Map_State->contigRelCoords[ptr_right] : Map_State->contigRelCoords[ptr_left]);
+    //     MY_CHECK(buff);
+    //     assert(0);
+    // }
+
+    // cut the contig by amending the relCoords..
+    for (u32 tmp = loc; tmp <= ptr_right; tmp++) // left side
+    {
+        Map_State->originalContigIds[tmp] += Max_Number_of_Contigs;
+    }
+
+    char buff[256];
+    snprintf(buff, sizeof(buff), "Original contig_id (%d), current_id (%d), pixel range: [%d, %d] %s inversed, cut at %d\n", original_contig_id, contig_id, ptr_left, ptr_right, inversed ? "(is)" : "(isn\'t)", loc);
+    std::cout << buff;
+
+    return ; 
+
+}
+
+
 global_function
 std::vector<s32>
 get_exclude_metaData_idx(const std::vector<std::string>& exclude_tags)
@@ -6239,23 +6355,25 @@ void EraseAllEdits(map_editor* map_editor_, u32 max_edit_recorded=Edits_Stack_Si
 void run_ai_detection()
 {
     std::cout << "Running AI detection..." << std::endl;
-    int status = std::system("/Users/sg35/miniconda3/envs/auto_cut/bin/python python/autoCut/auto_cut.py");  
+    int status = std::system("/Users/sg35/miniconda3/envs/auto_cut/bin/python /Users/sg35/PretextView/python/autoCut/auto_cut.py");  
 
     // 检查命令执行状态
     if (status == 0) {
-        std::cout << "命令执行成功!" << std::endl;
+        std::cout << "Command successful!" << std::endl;
     } else {
-        std::cerr << "命令执行失败，错误码: " << status << std::endl;
+        std::cerr << "Command failed, code: " << status << std::endl;
     }
 }
 
-void cut_frags(std::vector<std::vector<u32>> & problem_locs)
+void cut_frags(const std::vector<u32>& problem_locs)
 {
     for (auto & i : problem_locs)
     {
-        InvertMap(i[0], i[1]);
-        AddMapEdit(0, {i[0], i[1]}, true);
+        // cut the fragment
+        BreakMap(i);
+        UpdateContigsFromMapState();
     }
+    Redisplay = 1;
 }
 
 
@@ -6389,16 +6507,17 @@ void AutoCurationFromFragsOrder(
 
 global_function
 void
-ai_cutting_func(
+Auto_cutting_func(
     char * currFileName,
-    std::string file_save_path = std::string("./hic_figures")
+    memory_arena* arena = nullptr,
+    std::string file_save_name = std::string("/auto_curation_tmp")
 )
 {   
-    if (!currFileName || !AI_cutting) return;
+    if (!currFileName || !Auto_cutting) return;
 
     fprintf(stdout, "========================\n");
     // copy textures to cpu
-    fprintf(stdout, "[AI cutting] coping textures...\n");
+    fprintf(stdout, "[Auto cutting] coping textures...\n");
     auto texture_array_4_ai = TexturesArray4AI(Number_of_Textures_1D, Texture_Resolution, (char*)currFileName, Contigs);
     // prepare before reading textures
     f32 original_control_points[3];
@@ -6406,27 +6525,58 @@ ai_cutting_func(
     texture_array_4_ai.copy_buffer_to_textures_dynamic(Contact_Matrix, false); 
     restore_settings_after_copy(original_control_points);
     
-    // generate figures
+    // classic cut 
     if (1)
     {
-        fprintf(stdout, "[AI cutting] generating hic figures...\n");
+        FragCutCal frag_cut_cal(&texture_array_4_ai, Contigs, Number_of_Pixels_1D); 
+        std::vector<u32> cut_locs_pixel = frag_cut_cal.get_cut_locs_pixel();
+        #ifdef DEBUG
+            std::cout << "Cut locs in pixel: [";
+            for (auto & i : cut_locs_pixel) std::cout << i << " ";
+            std::cout << "]" << std::endl;
+        #endif
+        
+        // add the cut locs to the extensions
+        if (arena)
+        {   
+            u32 added_num = Extensions.get_num_extensions();
+            std::string graph_name = "hic_pixel_density";
+            u32* graph_data = new u32[Number_of_Pixels_1D];
+            for (u32 i=0; i < Number_of_Pixels_1D; i ++)
+            {
+                graph_data[i] = (u32)(frag_cut_cal.hic_density[i] * 100.f);
+            }
+            add_graph_to_extensions(arena, (u32*)graph_name.c_str(), graph_data);
+            
+            push_extensions_to_opengl(arena, added_num);
+        }
+        
+        // cut the contigs
+        cut_frags(cut_locs_pixel);
+    }
+
+    // generate figures
+    if (0)
+    {
+        fprintf(stdout, "[Auto cutting] generating hic figures...\n");
         Hic_Figure hic_figure(
             Total_Genome_Length, 
-            file_save_path,
+            file_save_name,
             &texture_array_4_ai
         );
         hic_figure.generate_hic_figures();
-    }
 
-    // run ai model with python script
-    run_ai_detection();
+        // run ai model with python script
+        run_ai_detection();
+
+        // restore the error info via reading the error json
+        HIC_Problems hic_problems("/auto_curation_tmp/auto_cut_output/infer_result/infer_result.json", (f32) Total_Genome_Length / (f32)Number_of_Pixels_1D ); 
+        std::vector<std::vector<u32>> problem_loc = hic_problems.get_problem_loci();
+        }
 
-    // restore the error info via reading the error json
-    HIC_Problems hic_problems("auto_cut_output/infer_result/infer_result.json", (f32) Total_Genome_Length / (f32)Number_of_Pixels_1D ); 
-    std::vector<std::vector<u32>> problem_loc = hic_problems.get_problem_loci();
 
     // cut the contigs 
-    cut_frags(problem_loc);
+    // cut_frags();
 
     return ;
 }
@@ -6434,15 +6584,15 @@ ai_cutting_func(
 
 global_function
 void
-Sort_yahs(char* currFileName)
+Sort_auto_sort(char* currFileName)
 {   
     if (!currFileName || !Yahs_sorting) return;
     
     fprintf(stdout, "========================\n");
-    fprintf(stdout, "[YaHS Sort] start...\n");
-    fprintf(stdout, "[YaHS Sort] smallest_frag_size_in_pixel: %d\n", auto_curation_state.smallest_frag_size_in_pixel);
-    fprintf(stdout, "[YaHS Sort] link_score_threshold:        %.3f\n", auto_curation_state.link_score_threshold);
-    fprintf(stdout, "[YaHS Sort] Sort mode:                   %s\n", auto_curation_state.get_sort_mode_name().c_str());
+    fprintf(stdout, "[Auto Sort] start...\n");
+    fprintf(stdout, "[Auto Sort] smallest_frag_size_in_pixel: %d\n", auto_curation_state.smallest_frag_size_in_pixel);
+    fprintf(stdout, "[Auto Sort] link_score_threshold:        %.3f\n", auto_curation_state.link_score_threshold);
+    fprintf(stdout, "[Auto Sort] Sort mode:                   %s\n", auto_curation_state.get_sort_mode_name().c_str());
     // compress the HiC data
     auto texture_array_4_ai = TexturesArray4AI(Number_of_Textures_1D, Texture_Resolution, (char*)currFileName, Contigs);
 
@@ -6512,7 +6662,7 @@ Sort_yahs(char* currFileName)
     }
     else 
     {
-        fprintf(stderr, "[YaHS Sort] Error: Unknown sort mode (%d)\n", auto_curation_state.sort_mode);
+        fprintf(stderr, "[Auto Sort] Error: Unknown sort mode (%d)\n", auto_curation_state.sort_mode);
         assert(0);
     }
     AutoCurationFromFragsOrder(
@@ -6951,7 +7101,7 @@ global_function
 void
 KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
 {
-    if (!Loading && !Yahs_sorting && !AI_cutting && (action != GLFW_RELEASE || key == GLFW_KEY_SPACE || key == GLFW_KEY_A || key == GLFW_KEY_LEFT_SHIFT))
+    if (!Loading && !Yahs_sorting && !Auto_cutting && (action != GLFW_RELEASE || key == GLFW_KEY_SPACE || key == GLFW_KEY_A || key == GLFW_KEY_LEFT_SHIFT))
     {
         if (UI_On)
         {
@@ -9731,6 +9881,12 @@ restore_initial_state()
         u32 nEdits  = my_Min(Edits_Stack_Size, Map_Editor->nEdits);
         ForLoop(nEdits) UndoMapEdit();
     }
+    
+    // restore all the splited contigs
+    {
+        Map_State->link_splited_contigs(Number_of_Pixels_1D);
+    }
+
 
     // waypoints
     {   
@@ -10153,9 +10309,9 @@ CopyEditsToClipBoard(GLFWwindow *window)
 
             u32 oldFrom = Map_State->contigRelCoords[start];
             u32 oldTo = Map_State->contigRelCoords[end];
-            u32 *name1 = (Original_Contigs + Map_State->originalContigIds[start])->name;
-            u32 *name2 = (Original_Contigs + Map_State->originalContigIds[end])->name;
-            u32 *newName = (Original_Contigs + Map_State->originalContigIds[to])->name;
+            u32 *name1 = (Original_Contigs + Map_State->get_original_contig_id(start))->name;
+            u32 *name2 = (Original_Contigs + Map_State->get_original_contig_id(end))->name;
+            u32 *newName = (Original_Contigs + Map_State->get_original_contig_id(to))->name;
 
             bufferPtr += (u32)stbsp_snprintf((char *)buffer + bufferPtr, (s32)(bufferSize - bufferPtr), "Edit %d:\n       %s[%$.2fbp] to %s[%$.2fbp]\n%s\n       %s[%$.2fbp]\n",
                     index + 1, (char *)name1, (f64)oldFrom * bpPerPixel, (char *)name2, (f64)oldTo * bpPerPixel,
@@ -10552,16 +10708,16 @@ MainArgs {
         if (Yahs_sorting)
         {
             if (currFileName) SaveState(headerHash);
-            Sort_yahs((char*)currFileName);
+            Sort_auto_sort((char*)currFileName);
             Yahs_sorting = 0;
             Redisplay = 1;
         }
 
-        if (AI_cutting)
+        if (Auto_cutting)
         {   
             if (currFileName) SaveState(headerHash);
-            ai_cutting_func((char*)currFileName);
-            AI_cutting = 0;
+            Auto_cutting_func((char*)currFileName, Loading_Arena);
+            Auto_cutting = 0;
             Redisplay = 1;
         }
 
@@ -10667,15 +10823,15 @@ MainArgs {
                     showUserProfileScreen = nk_button_label(NK_Context, "User Profile");
                     showAboutScreen = nk_button_label(NK_Context, "About");
 
-                    // AI & YaHS sort button 
+                    // AI & Auto sort button 
                     struct nk_color sort_button_normal = nk_rgb(153, 50, 204); 
                     struct nk_color sort_button_hover  = nk_rgb(186, 85, 211); 
                     struct nk_color sort_button_active = nk_rgb(128, 0, 128);  
                     push_nk_style(NK_Context, sort_button_normal, sort_button_hover, sort_button_active);
                     nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 3);
-                    ai_cutting_button = nk_button_label(NK_Context, "AI cut");
+                    auto_cut_button = nk_button_label(NK_Context, "Auto cut");
                     bounds = nk_widget_bounds(NK_Context);
-                    yahs_sort_button = nk_button_label(NK_Context, "YaHS Sort");
+                    auto_sort_button = nk_button_label(NK_Context, "Auto Sort");
 
                     // AI sort button
                     /*
@@ -10719,17 +10875,17 @@ MainArgs {
                         }
                     }
                     */
-                    // ai cut
+                    // auto cut
                     {
-                        if (ai_cutting_button && currFileName)
+                        if (auto_cut_button && currFileName)
                         {
-                            AI_cutting = 1;
+                            Auto_cutting = 1;
                         }
                     }
 
-                    // YaHS sort button
+                    // Auto sort button
                     {   
-                        if (yahs_sort_button) 
+                        if (auto_sort_button) 
                         {
                             Yahs_sorting = 1;
                             auto_curation_state.clear();
@@ -10773,7 +10929,7 @@ MainArgs {
                             nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 2);
                             if (nk_button_label(NK_Context, "Cancel")) 
                             {   
-                                printf("[YaHS Sort] Cancel button clicked\n");
+                                printf("[Auto Sort] Cancel button clicked\n");
                                 nk_contextual_close(NK_Context);
                             }
                             if (nk_button_label(NK_Context, "Apply")) 
@@ -10785,21 +10941,21 @@ MainArgs {
                                 nk_contextual_close(NK_Context);
                                 if (auto_curation_state.link_score_threshold > 1.0f || auto_curation_state.link_score_threshold < 0.0f) 
                                 {   
-                                    printf("[YaHS Sort] Warning: link Score Threshold should be in the range of [0, 1]\n");
+                                    printf("[Auto Sort] Warning: link Score Threshold should be in the range of [0, 1]\n");
                                     auto_curation_state.link_score_threshold = std::max(0.0f, std::min(1.0f, auto_curation_state.link_score_threshold));
                                 }
                                 auto_curation_state.set_buf();
-                                printf("[YaHS Sort] smallest_frag_size_in_pixel: %d\n", auto_curation_state.smallest_frag_size_in_pixel);
-                                printf("[YaHS Sort] link_score_threshold:        %.3f\n", auto_curation_state.link_score_threshold);
-                                printf("[YaHS Sort] Sort mode:                   %s\n", auto_curation_state.get_sort_mode_name().c_str());
+                                printf("[Auto Sort] smallest_frag_size_in_pixel: %d\n", auto_curation_state.smallest_frag_size_in_pixel);
+                                printf("[Auto Sort] link_score_threshold:        %.3f\n", auto_curation_state.link_score_threshold);
+                                printf("[Auto Sort] Sort mode:                   %s\n", auto_curation_state.get_sort_mode_name().c_str());
                             }
                             // redo all the changes
-                            if (!auto_curation_state.show_yahs_redo_confirm_popup)
+                            if (!auto_curation_state.show_autoSort_redo_confirm_popup)
                             {
                                 nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 1);
                                 if (nk_button_label(NK_Context, "Redo All (Careful!)")) 
                                 {   
-                                    auto_curation_state.show_yahs_redo_confirm_popup=true;
+                                    auto_curation_state.show_autoSort_redo_confirm_popup=true;
                                 }
                             } else
                             {
@@ -10808,22 +10964,22 @@ MainArgs {
                                 nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 2);
                                 if (nk_button_label(NK_Context, "Yes")) 
                                 {
-                                    auto_curation_state.show_yahs_redo_confirm_popup=false;
-                                    printf("[YaHS Sort] Redo all edits.\n");
+                                    auto_curation_state.show_autoSort_redo_confirm_popup=false;
+                                    printf("[Auto Sort] Redo all edits.\n");
                                     RedoAllEdits(Map_Editor);
                                 }
                                 if (nk_button_label(NK_Context, "No")) 
                                 {
-                                    auto_curation_state.show_yahs_redo_confirm_popup=false;
+                                    auto_curation_state.show_autoSort_redo_confirm_popup=false;
                                 }
                             }
                             // reject all the edits
-                            if (!auto_curation_state.show_yahs_erase_confirm_popup)
+                            if (!auto_curation_state.show_autoSort_erase_confirm_popup)
                             {   
                                 nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 1);
                                 if (nk_button_label(NK_Context, "Erase All (Careful!)")) 
                                 {   
-                                    auto_curation_state.show_yahs_erase_confirm_popup=true;
+                                    auto_curation_state.show_autoSort_erase_confirm_popup=true;
                                 }
                             } else
                             {
@@ -10832,13 +10988,13 @@ MainArgs {
                                 nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 2);
                                 if (nk_button_label(NK_Context, "Yes")) 
                                 {
-                                    auto_curation_state.show_yahs_erase_confirm_popup=false;
-                                    printf("[YaHS Sort] Erase all edits.\n");
+                                    auto_curation_state.show_autoSort_erase_confirm_popup=false;
+                                    printf("[Auto Sort] Erase all edits.\n");
                                     EraseAllEdits(Map_Editor);
                                 }
                                 if (nk_button_label(NK_Context, "No")) 
                                 {
-                                    auto_curation_state.show_yahs_erase_confirm_popup=false;
+                                    auto_curation_state.show_autoSort_erase_confirm_popup=false;
                                 }
                             }
                             nk_contextual_end(NK_Context);
@@ -11330,9 +11486,9 @@ MainArgs {
 
                                     u32 oldFrom = Map_State->contigRelCoords[start];
                                     u32 oldTo = Map_State->contigRelCoords[end];
-                                    u32 *name1 = (Original_Contigs + Map_State->originalContigIds[start])->name;
-                                    u32 *name2 = (Original_Contigs + Map_State->originalContigIds[end])->name;
-                                    u32 *newName = (Original_Contigs + Map_State->originalContigIds[to])->name;
+                                    u32 *name1 = (Original_Contigs + Map_State->get_original_contig_id(start))->name;
+                                    u32 *name2 = (Original_Contigs + Map_State->get_original_contig_id(end))->name;
+                                    u32 *newName = (Original_Contigs + Map_State->get_original_contig_id(to))->name;
                                     
                                     stbsp_snprintf((char *)buff, sizeof(buff), "Edit %d:", index + 1);
                                     nk_label(NK_Context, (const char *)buff, NK_TEXT_LEFT);
diff --git a/python/autoCut/global_settings.py b/python/autoCut/global_settings.py
index e887d8e..4b4683d 100644
--- a/python/autoCut/global_settings.py
+++ b/python/autoCut/global_settings.py
@@ -7,12 +7,12 @@
 SKIP_HIC_IMG_GEN=True    # Skip the hic figure generation (hicstraw)
 
 
-AUTO_CUT_DIR = "python/autoCut"
+AUTO_CUT_DIR = "/Users/sg35/PretextView/python/autoCut"
 MODEL_CONFIG = os.path.join(AUTO_CUT_DIR, "models/cfgs/error_model.py")
 PRETRAINED_MODEL = os.path.join(AUTO_CUT_DIR, "models/cfgs/error_model.pth")
 # path to hic figures
-HIC_FIGURE_DIR = "hic_figures"
-AUTO_CUT_OUTPUT_DIR = "auto_cut_output"
+HIC_FIGURE_DIR = "/Users/sg35/PretextView/build_cmake/PretextViewAI.app/Contents/Resources/auto_curation_tmp/hic_figures"
+AUTO_CUT_OUTPUT_DIR = "/Users/sg35/PretextView/build_cmake/PretextViewAI.app/Contents/Resources/auto_curation_tmp/auto_cut_output"
 ERROR_FILTER_IOU_SCORE=0.8 # used to filter out the errors with overlap area more than 0.8
 ERROR_SCORE_MIN = 0.2
 
diff --git a/src/ai_cut_json_parse.h b/src/ai_cut_json_parse.h
index 3f2e890..68a6a82 100644
--- a/src/ai_cut_json_parse.h
+++ b/src/ai_cut_json_parse.h
@@ -54,14 +54,14 @@ class HIC_Problems
 
     HIC_Problems(std::string file_path, f32 bp_per_pixel_)
     : bp_per_pixel(bp_per_pixel_)
-    {
+    {   
+        file_path = std::filesystem::current_path().string() + file_path;
         std::ifstream file(file_path, std::ios::in);
         if (!file.is_open())
         {   
             throw std::runtime_error("Cannot open the file: " + file_path);
         }
         
-        // 解析 JSON
         json data = json::parse(file);
 
         for (const auto& i : this->types)
diff --git a/src/auto_curation_state.h b/src/auto_curation_state.h
index d19ef5e..074aca3 100644
--- a/src/auto_curation_state.h
+++ b/src/auto_curation_state.h
@@ -35,8 +35,8 @@ struct AutoCurationState
     f32 link_score_threshold = 0.4f;
 
     // Variables for the editing UI state
-    bool show_yahs_erase_confirm_popup = false;
-    bool show_yahs_redo_confirm_popup = false;
+    bool show_autoSort_erase_confirm_popup = false;
+    bool show_autoSort_redo_confirm_popup = false;
     //sorting mode
     u32 sort_mode = 0; // 0: union find, 1: fuse union find, 2 deep fuse
     std::vector<std::string> sort_mode_names = {"Union Find", "Fuse Union Find", "Deep Fuse"};
diff --git a/src/copy_texture.cpp b/src/copy_texture.cpp
index d67e755..a302bb4 100644
--- a/src/copy_texture.cpp
+++ b/src/copy_texture.cpp
@@ -766,6 +766,30 @@ void TexturesArray4AI::show_collected_textures()
 }
 
 
+
+void TexturesArray4AI::cal_maximum_number_of_shift(
+    u32& D,
+    std::vector<f32>& norm_diag_mean,
+    const contigs* Contigs,
+    f32 D_hic_ratio,
+    u32 maximum_D,
+    f32 min_hic_density)
+{   
+    f32 global_shift0_diag_mean = this->cal_diagonal_mean_within_fragments(D, Contigs);
+    norm_diag_mean.push_back(global_shift0_diag_mean);
+    f32 tmp_diag_mean = this->cal_diagonal_mean_within_fragments(++D, Contigs);
+    while (
+        tmp_diag_mean > min_hic_density && 
+        tmp_diag_mean > D_hic_ratio * global_shift0_diag_mean && 
+        D < maximum_D
+    ) {
+        norm_diag_mean.push_back(tmp_diag_mean);
+        tmp_diag_mean = this->cal_diagonal_mean_within_fragments(++D, Contigs); 
+    }
+    return ;
+}
+
+
 void TexturesArray4AI::cal_compressed_hic(
     const contigs* Contigs, 
     const extension_sentinel& Extensions,
@@ -846,18 +870,15 @@ void TexturesArray4AI::cal_compressed_hic(
     }
 
     // cal the maximum number of shift
-    u32 D = 0;
-    f32 global_shift0_diag_mean = this->cal_diagonal_mean_within_fragments(D, Contigs);
-    std::vector<f32> norm_diag_mean = {global_shift0_diag_mean};
-    f32 tmp_diag_mean = this->cal_diagonal_mean_within_fragments(++D, Contigs);
-    while (
-        tmp_diag_mean > min_hic_density && 
-        tmp_diag_mean > D_hic_ratio * global_shift0_diag_mean && 
-        D < maximum_D
-    ) {
-        norm_diag_mean.push_back(tmp_diag_mean);
-        tmp_diag_mean = this->cal_diagonal_mean_within_fragments(++D, Contigs); 
-    }
+    u32 D=0;
+    std::vector<f32> norm_diag_mean = {};
+    this->cal_maximum_number_of_shift(
+        D, 
+        norm_diag_mean,
+        Contigs,
+        D_hic_ratio,
+        maximum_D,
+        min_hic_density);
 
     Values_on_Channel buffer_values_on_channel;
     u32 cnt = 0;
diff --git a/src/copy_texture.h b/src/copy_texture.h
index c72b4e5..7fe1ae9 100644
--- a/src/copy_texture.h
+++ b/src/copy_texture.h
@@ -575,7 +575,6 @@ class GraphData
 };
 
 
-
 class TexturesArray4AI
 {
 private:
@@ -658,6 +657,15 @@ class TexturesArray4AI
         f32 D_hic_ratio=0.05f, 
         u32 maximum_D=5000, 
         f32 min_hic_density = 30.f);
+
+    void cal_maximum_number_of_shift(
+        u32& D,
+        std::vector<f32>& norm_diag_mean,
+        const contigs* Contigs,
+        f32 D_hic_ratio,
+        u32 maximum_D,
+        f32 min_hic_density);
+
     void cal_compressed_extension(const extension_sentinel &Extensions);
 
     Sum_and_Number get_fragement_diag_mean_square(
diff --git a/src/frag_cut_calculation.h b/src/frag_cut_calculation.h
new file mode 100644
index 0000000..4a30f7b
--- /dev/null
+++ b/src/frag_cut_calculation.h
@@ -0,0 +1,305 @@
+
+#ifndef FRAG_CUT_CALCULATION_H
+#define FRAG_CUT_CALCULATION_H
+
+
+#include <iostream>
+#include <fstream>
+#include <filesystem>
+#include <sstream>
+#include <vector>
+#include <string>
+#include <limits>
+#include "genomeData.h"
+#include "utilsPretextView.h"
+#include "copy_texture.h"
+
+
+
+void cal_prefix_sum(
+    const std::vector<f32>& arr, std::vector<f32>& prefix_sum) 
+{
+    int n = arr.size();
+    for (int i = 0; i < n; ++i) 
+    {
+        prefix_sum[i + 1] = prefix_sum[i] + arr[i];
+    }
+    return ;
+}
+
+
+
+
+class FragCutCal
+{
+public: 
+    TexturesArray4AI *textures_array = nullptr;
+    const Frag4compress* frags = nullptr;
+    contigs* Contigs = nullptr;
+    std::string file_save_dir = std::filesystem::current_path().string() + "/hic_density_curve";
+    std::vector<f32> hic_density={};
+
+    u32 D = 0;
+    std::vector<f32> norm_diag_mean={};
+
+    f32 D_hic_ratio=0.05f;
+    u32 maximum_D=5000;
+    f32 min_hic_density=30.f;
+
+    FragCutCal(TexturesArray4AI* textures_array_, contigs* Contigs_, const u32& num_pixels_1d)
+        : textures_array(textures_array_), Contigs(Contigs_)
+    {   
+        if (textures_array == nullptr)
+        {
+            MY_CHECK("The textures_array is not initialized");
+            assert(0);
+        }
+
+        #ifdef DEBUG // create folder to save the hic_density_curve
+            if (std::filesystem::exists(this->file_save_dir))
+            {
+                std::filesystem::remove_all(this->file_save_dir);
+            }
+            std::filesystem::create_directory(this->file_save_dir);
+        #endif
+
+        this->frags = textures_array->get_frags();
+        textures_array->cal_maximum_number_of_shift(
+            this->D, 
+            this->norm_diag_mean, 
+            this->Contigs,
+            this->D_hic_ratio,
+            this->maximum_D,
+            this->min_hic_density);
+        this->hic_density.resize(num_pixels_1d, 0.f);
+        
+        for (u32 frag_id = 0; frag_id < this->frags->num; frag_id++)
+        {
+            this->get_single_fragment_hic_density(
+                this->frags->startCoord[frag_id],
+                this->frags->length[frag_id],
+                this->hic_density.data() + this->frags->startCoord[frag_id]);
+        }
+
+    }
+
+    std::vector<u32> get_cut_locs_pixel()
+    {
+        std::vector<u32> cut_locs_pixel;
+        for (u32 i = 0; i < this->frags->num; i++)
+        {
+            this->get_single_fragment_cut_locs(i, cut_locs_pixel);
+        }
+        return cut_locs_pixel;
+    }
+
+    /* 全局坐标存储在 break_points 数组中 */
+    void find_break_points(
+        const u32 start_pixel,
+        const std::vector<f32>& arr,
+        const std::vector<f32>& order1_derivative,
+        const std::vector<f32>& order2_derivative,
+        std::vector<u32>& break_points,
+        const u32 windows_size=8,  
+        const f32 threshold=0.05)
+    {
+        int n = arr.size();
+        if (n < 2 * windows_size + 1) 
+        {
+            return ; 
+        }
+
+        if (1)  // detect the break points according to the value
+        {
+            std::vector<f32> prefix_sum(n + 1, 0.f);
+            cal_prefix_sum(arr, prefix_sum);
+            for (int i = windows_size; i < n - windows_size; ++i) 
+            {
+                f32 avgBefore = (prefix_sum[i] - prefix_sum[i - windows_size]);
+                
+                f32 avgAfter = (prefix_sum[i + windows_size + 1] - prefix_sum[i + 1]);
+                
+                f32 current_window = (prefix_sum[i + windows_size/2 + 1] - prefix_sum[i - windows_size/2 + 1]);
+                
+                if (
+                    (avgBefore - current_window > avgBefore * threshold && avgAfter - current_window > avgAfter * threshold) ||
+                    (current_window - avgBefore > avgBefore * threshold && current_window - avgAfter  > avgAfter * threshold)
+                ) 
+                {   
+                    u08 min_flag = avgBefore > current_window;  
+                    u32 min_i = i;
+                    // find the minimum value in the window
+                    for (u32 tmp_i = i;
+                        tmp_i < n - windows_size && tmp_i < i + windows_size+1;
+                        tmp_i++ )
+                    {
+                        f32 current_next = (prefix_sum[tmp_i + windows_size/2 + 2] - prefix_sum[tmp_i - windows_size/2 + 2]);
+                        if (
+                            (min_flag && current_window > current_next) ||
+                            (!min_flag && current_window < current_next) )
+                        {
+                            current_window = current_next;
+                            min_i = i;
+                        }
+                    }
+                    break_points.push_back(min_i + start_pixel);
+                    i += windows_size ;
+                }
+            }
+        }
+
+        if (0) // detect break points according to derivatives
+        {   
+            u08 increase_flag = 1;
+            u32 len_count = 0;
+            for (u32 i = 1; i < n; i ++ )
+            {
+                if ((order1_derivative[i] > 0) == increase_flag)
+                {
+                    len_count++;
+                }
+                else
+                {
+                    if (len_count > windows_size * 2)
+                    {
+                        break_points.push_back(i + start_pixel);
+                    }
+                    len_count = 0;
+                    increase_flag = !increase_flag;
+                }
+            }
+        }
+
+        return ;
+    }
+
+    void get_single_fragment_cut_locs(
+        const u32& frag_id, std::vector<u32>& cut_locs)
+    {   
+        // copy orginal to local
+        std::vector<f32> hic_density_tmp(this->hic_density.begin() + this->frags->startCoord[frag_id], 
+            this->hic_density.begin() + this->frags->startCoord[frag_id] + this->frags->length[frag_id]);
+
+        u32 n = hic_density_tmp.size();
+        std::vector<f32> order1_derivative(n, 0.f);
+        std::vector<f32> order2_derivative(n, 0.f);
+        get_smoothed_1st_order_derivative(hic_density_tmp, order1_derivative);
+        get_smoothed_1st_order_derivative(order1_derivative, order2_derivative);
+        
+        #ifdef DEBUG // output the hic_density to text file
+            std::string filename = this->file_save_dir +  "/current_id_" + std::to_string(frag_id) + ".txt";
+            std::cout << "[Auto Cut]: (classic) output hic_density to " << filename << std::endl;
+            std::ofstream out(filename);
+            for (u32 i = 0; i < hic_density_tmp.size(); i++)
+            {
+                out << hic_density_tmp[i] << ", " << order1_derivative[i] << ", " << order2_derivative[i] << std::endl;
+            }
+            out.close();
+        #endif
+
+        std::vector<u32> break_points(0);
+        this->find_break_points(
+            this->frags->startCoord[frag_id],
+            hic_density_tmp,
+            order1_derivative,
+            order2_derivative,
+            break_points);
+        if (break_points.size() > 0)
+        {
+            cut_locs.insert(cut_locs.end(), break_points.begin(), break_points.end());
+        }
+        
+    }
+
+    
+    void get_single_fragment_hic_density(
+        const u32& start_pixel,
+        const u32& len_pixel,
+        f32* hic_density
+    )
+    {   
+        for (u32 i=0; i < len_pixel; i++)
+        {
+            hic_density[i] = this->pixel_mean(start_pixel, i, len_pixel);
+        }
+        if (len_pixel > 4)
+        {
+            hic_density[0] = hic_density[1] = hic_density[2];
+            hic_density[len_pixel-1] = hic_density[len_pixel-2] = hic_density[len_pixel-3];
+        }
+        else 
+        {
+            for (u32 i = 0; i < len_pixel; i++)
+            {
+                hic_density[i] = 1.0f;
+            }
+        }
+        return ;
+    }
+
+    f32 pixel_mean(
+        const u32& frag_start_pixel,
+        const u32& relative_index, 
+        const u32& len_pixel, 
+        const u08 diag_flag=1) 
+    {   
+        u32 num = 0, pixel_index=frag_start_pixel+relative_index;
+        f32 sum = 0.f;
+
+        for (u32 i = 1; i < 8 ; i++) 
+        {   
+            if (diag_flag)
+            {   
+                if (i < relative_index && i + relative_index < len_pixel-1)
+                {
+                    sum += ( (f32)((*(this->textures_array))(
+                        pixel_index - i,
+                        pixel_index + i)) ); // / this->norm_diag_mean[i]
+                    num++;
+                }
+            } else 
+            {
+                if ( i+relative_index < len_pixel)
+                {
+                    sum += ((*(this->textures_array))(
+                        pixel_index, 
+                        pixel_index + i) / this->norm_diag_mean[i]);
+                    num++;
+                }
+                if (relative_index >= i)
+                {
+                    sum += ((*(this->textures_array))(
+                        pixel_index, 
+                        pixel_index - i) / this->norm_diag_mean[i]);
+                    num++;
+                }
+            }
+        }
+        return num==0 ? 0.f : (f32)sum / (f32)num;
+    }
+
+    /* 
+    计算导数, 参数：
+        arr：        原向量
+        derivative： 倒数向量存储位置
+    */
+    void get_smoothed_1st_order_derivative(
+        const std::vector<f32>& arr, 
+        std::vector<f32>& derivative, 
+        u32 window_size=10)
+    {   
+        u32 dist = window_size / 2;
+        for (u32 i = 0; i < arr.size(); i++)
+        {
+            derivative[i] = (arr[my_Min(i + dist, arr.size() - 1)] - arr[i > dist? i - dist:0]) / (f32)(dist * 2);
+        }
+    }
+
+
+};
+
+
+
+
+
+#endif // FRAG_CUT_CALCULATION_H
\ No newline at end of file
diff --git a/src/genomeData.h b/src/genomeData.h
index f4e8067..0942675 100644
--- a/src/genomeData.h
+++ b/src/genomeData.h
@@ -29,6 +29,7 @@ SOFTWARE.
 #include "showWindowData.h"
 
 
+#define Max_Number_of_Contigs 4096
 
 struct
 file_atlas_entry
@@ -97,6 +98,19 @@ struct map_state
     u32 *contigRelCoords = nullptr;   // local coordinate on the original contig [num_pixels_1d]
     u32 *scaffIds = nullptr;          // [num_pixels_1d]
     u64 *metaDataFlags = nullptr;     // [num_pixels_1d], 256kb for 32768 pixels
+
+    u32 get_original_contig_id(u32 pixel) const
+    {
+        return originalContigIds[pixel] % Max_Number_of_Contigs;
+    }
+
+    void link_splited_contigs(const u32 num_pixel_1d)
+    {
+        for (u32 i = 0; i < num_pixel_1d; i++)
+        {
+            originalContigIds[i] = originalContigIds[i] % Max_Number_of_Contigs;
+        }
+    }
 };
 
 
@@ -142,6 +156,16 @@ extension_sentinel
 {
     extension_node *head = nullptr;  // point the first node
     extension_node *tail = nullptr;  // point the last node
+
+    u32 get_num_extensions()
+    {
+        u32 added_index = 0;
+        TraverseLinkedList(this->head, extension_node)
+        {
+            added_index++;
+        }
+        return added_index;
+    }
 };
 
 
diff --git a/src/hic_figure.h b/src/hic_figure.h
index 9658569..de1dd68 100644
--- a/src/hic_figure.h
+++ b/src/hic_figure.h
@@ -35,29 +35,6 @@
 using Dict_json = std::map<std::string, std::variant<u64, f32, std::string>>;
 
 
-/*计算数组的 95% 分位数*/
-template <typename T>
-T percentile_cal(T* data, u32 size, f32 percentile=0.95)
-{   
-    u32 heap_size = (u32)((f32)size * (1-percentile));
-    std::priority_queue<T, std::vector<T>, std::greater<T>> min_heap;
-    for (auto i = 0; i < size; i++)
-    {
-        if (min_heap.size() < heap_size)
-        {
-            min_heap.push(data[i]);
-        }
-        else if (data[i] > min_heap.top())
-        {
-            min_heap.pop();
-            min_heap.push(data[i]);
-        }
-    }
-    if (min_heap.size() > 0) 
-        return min_heap.top();
-    else return 0;
-}
-
 
 
 
@@ -210,9 +187,11 @@ class Hic_Figure
     TexturesArray4AI* textures_array;
     f32 bp_per_pixel_hic;
     std::string file_save_dir;
+    std::string hic_figure_save_dir;
     std::string auto_cut_output_dir;
+    std::string info_save_file_full_path;
     #ifdef DEBUG
-        std::vector<f32> resolution_mutiplier = {2, 4};
+        std::vector<f32> resolution_mutiplier = {4};
     #else
         std::vector<f32> resolution_mutiplier = {
             0.5, 1, 2, 4, 8, 16, 32};
@@ -224,44 +203,42 @@ class Hic_Figure
         u64 genome_bp_,
         std::string file_save_dir_, 
         TexturesArray4AI* textures_array_,
-        std::string auto_cut_output_dir_ = "auto_cut_output"
+        std::string hic_figure_dir_name = "/hic_figures",
+        std::string auto_cut_output_name = "/auto_cut_output",
+        std::string info_file_name = "/info.json"
     ) 
         : 
         genome_bp(genome_bp_),
-        hic_pixel_1d(textures_array_->get_num_pixels_1d()),
-        file_save_dir(file_save_dir_),
-        auto_cut_output_dir(auto_cut_output_dir_)
+        hic_pixel_1d(textures_array_->get_num_pixels_1d())
     {   
         this->textures_array = textures_array_;
         this->bp_per_pixel_hic = (f32)this->genome_bp / (f32)this->hic_pixel_1d;
-        if (std::filesystem::exists(this->auto_cut_output_dir))
-        {
-            std::cout << "Remove folder: " 
-                      << (std::filesystem::current_path().string() + this->auto_cut_output_dir) 
-                      << std::endl;
-            std::filesystem::remove_all(this->auto_cut_output_dir);
-            std::filesystem::create_directories(this->auto_cut_output_dir);
-        }
         // if the folder does not exist, create it
-        if (std::filesystem::exists(file_save_dir))
+        this->file_save_dir = std::filesystem::current_path().string() + file_save_dir_;
+        if (std::filesystem::exists(this->file_save_dir))
         {
             // remove this one 
             std::cout << "Remove folder: " 
-                      << (std::filesystem::current_path().string() + file_save_dir) 
+                      << this->file_save_dir
                       << std::endl;
-            std::filesystem::remove_all(file_save_dir);
+            std::filesystem::remove_all(this->file_save_dir);
         }
         std::cout << "Creat folder: " 
-                  << (std::filesystem::current_path().string() + file_save_dir) 
+                  << (std::filesystem::current_path().string() + this->file_save_dir) 
                   << std::endl;
-        std::filesystem::create_directories(file_save_dir);
+        std::filesystem::create_directories(this->file_save_dir);
+        // clear save pics
+        this->auto_cut_output_dir = this->file_save_dir + auto_cut_output_name;
+        std::filesystem::create_directories(this->auto_cut_output_dir);
+        this->hic_figure_save_dir = this->file_save_dir + hic_figure_dir_name;
+        std::filesystem::create_directories(this->hic_figure_save_dir);
         // clear the saved info
-        std::string info_save_file_full_path = this->file_save_dir+"/info.json";
-        std::ofstream out_file(info_save_file_full_path);
+        this->info_save_file_full_path = this->hic_figure_save_dir+info_file_name;
+        std::ofstream out_file(this->info_save_file_full_path);
         if (!out_file.is_open())
         {   
             std::stringstream ss;
-            ss << "Cannot open: " << info_save_file_full_path << std::endl;
+            ss << "Cannot open: " << this->info_save_file_full_path << std::endl;
             MY_CHECK(ss.str().c_str());
             assert(0);
         }
@@ -283,7 +260,7 @@ class Hic_Figure
         u32 end_bp = (u32) ((start_pixel_figure + this->figure_pixel_1d) * bp_per_pixel_figure);
         
         std::string img_id = generate_random_string(16);
-        std::string file_save_path = this->file_save_dir + "/" + img_id;
+        std::string file_save_path = this->hic_figure_save_dir + "/" + img_id;
 
         if (this->figure_info)
         {
@@ -314,12 +291,11 @@ class Hic_Figure
             std::cerr << "The info is empty" << std::endl;
             assert(0);
         }
-        std::string info_save_file_full_path = this->file_save_dir+"/info.json";
-        std::fstream out_file = std::fstream(info_save_file_full_path, std::ios::in | std::ios::out);
+        std::fstream out_file = std::fstream(this->info_save_file_full_path, std::ios::in | std::ios::out);
         if (!out_file.is_open())
         {   
             std::stringstream ss;
-            ss << "Cannot open: " << info_save_file_full_path << std::endl;
+            ss << "Cannot open: " << this->info_save_file_full_path << std::endl;
             MY_CHECK(ss.str().c_str());
             assert(0);
         }
@@ -343,7 +319,7 @@ class Hic_Figure
                     ptr += 2 ;
                 }
                 out_file.seekp(ptr);
-                out_file << "\"" << this->file_save_dir << "/" ; 
+                out_file << "\"" << this->hic_figure_save_dir << "/" ; 
                 std::visit(
                     [&out_file](const auto& v) 
                     {
@@ -403,20 +379,6 @@ class Hic_Figure
             0,                            // output stride
             1                             // number of channels
         );
-        char buff[512];
-        snprintf(
-            buff, sizeof(buff), 
-            "%s/%s.png", 
-            this->file_save_dir.c_str(),  
-            this->figure_info->img_id.c_str()
-            // "%s/multiplier%.2f_%.2f-%.2f_%s.png", 
-            // this->file_save_dir.c_str(), 
-            // this->figure_info->multiplier, 
-            // (f32)this->figure_info->start_bp / (f32)this->figure_info->genome_bp,
-            // (f32)this->figure_info->end_bp / (f32)this->figure_info->genome_bp, 
-            // this->figure_info->img_id.c_str()
-        );
-        // save the figure
 
         // 95 percentile
         u08 percentile_95 = percentile_cal(data_out, num_pixels, 0.95);
@@ -441,10 +403,11 @@ class Hic_Figure
         delete[] data_out_f32;
 
         // 生成保存路径（添加_red后缀）
+        char buff[512];
         snprintf(
             buff, sizeof(buff), 
             "%s/%s.png",  
-            this->file_save_dir.c_str(),  
+            this->hic_figure_save_dir.c_str(),  
             this->figure_info->img_id.c_str()
         );
 
diff --git a/src/utilsPretextView.h b/src/utilsPretextView.h
index 06058aa..7e57ba1 100644
--- a/src/utilsPretextView.h
+++ b/src/utilsPretextView.h
@@ -231,4 +231,28 @@ void my_code_position_handler(const char* file, int line, const char* message=nu
 
 f64 GetTime(); // return current time in seconds
 
+
+/* 计算数组的 95% 分位数 */
+template <typename T>
+T percentile_cal(T* data, u32 size, f32 percentile=0.95)
+{   
+    u32 heap_size = (u32)((f32)size * (1-percentile));
+    std::priority_queue<T, std::vector<T>, std::greater<T>> min_heap;
+    for (auto i = 0; i < size; i++)
+    {
+        if (min_heap.size() < heap_size)
+        {
+            min_heap.push(data[i]);
+        }
+        else if (data[i] > min_heap.top())
+        {
+            min_heap.pop();
+            min_heap.push(data[i]);
+        }
+    }
+    if (min_heap.size() > 0) 
+        return min_heap.top();
+    else return 0;
+}
+
 #endif // UTILS_H
\ No newline at end of file

From 8eb4f63b37cfd8b1fc1258d9b435a1cab96e6217 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 24 Mar 2025 14:30:38 +0000
Subject: [PATCH 08/33] add fmt lib

---
 .gitmodules     | 3 +++
 subprojects/fmt | 1 +
 2 files changed, 4 insertions(+)
 create mode 160000 subprojects/fmt

diff --git a/.gitmodules b/.gitmodules
index 1411617..cfbf012 100644
--- a/.gitmodules
+++ b/.gitmodules
@@ -15,3 +15,6 @@
 	path = subprojects/pytorch_scatter
 	url = https://github.com/guanshaoheng/pytorch_scatter.git 
 
+[submodule "subprojects/fmt"]
+	path = subprojects/fmt
+	url = https://github.com/fmtlib/fmt.git
diff --git a/subprojects/fmt b/subprojects/fmt
new file mode 160000
index 0000000..6fdf225
--- /dev/null
+++ b/subprojects/fmt
@@ -0,0 +1 @@
+Subproject commit 6fdf225a32625389c53e99e4d2b928df24a0b124

From 764d2467cc3d083508e567741360e77bc24cf891 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 24 Mar 2025 14:31:54 +0000
Subject: [PATCH 09/33] add select sort area mode description

---
 README.md                  |  29 ++++++++++++++++++++++-------
 doc/image.png              | Bin 397950 -> 0 bytes
 doc/image_user_profile.png | Bin 0 -> 80079 bytes
 3 files changed, 22 insertions(+), 7 deletions(-)
 delete mode 100644 doc/image.png
 create mode 100644 doc/image_user_profile.png

diff --git a/README.md b/README.md
index 6654fd4..2b7d25c 100644
--- a/README.md
+++ b/README.md
@@ -12,7 +12,7 @@ PretextView is a desktop application for viewing pretext contact maps.<br/>
 - `right mouse button`: holding down and move to drag the Hi-C figure move within the window. 
 - `Mouse scroll` to zoom. A three button mouse is recomended.
 - `E`: enter the Edit mode.
-- `K`: enter the Select sort area mode.
+- `F`: enter the Select sort area mode.
 - `M`: enter the Meta Tag edit mode.
 - `S`: enter the Scaffold painting mode.
 - `X`: enter the extension edit mode.
@@ -48,11 +48,11 @@ PretextView is a desktop application for viewing pretext contact maps.<br/>
 Enter scaffolding mode with the 's' key.
 - Use the GUI to see a list of scaffolds.
 
-## Select sort area model (`K`)
+## Select sort area mode (`F`)
 
-After click `YaHS sort` button in the main UI, it will defaultly run sort globally. If enter the the `select sort area mode`(by pressing `K`), it can sort only the selected area with pressing `Space` after selecting at least 3 contigs. 
+After click `Auto Sort` button in the main UI, it will defaultly run sort globally. If enter the the `select sort area mode`(by pressing `F`), it can sort only the selected area with pressing `Space` after selecting at least 2 contigs. 
 
-Enter the select sort area mode by pressing `K`.
+Enter the select sort area mode by pressing `F`.
 - `Left mouse click`: select / un-select area for sorting.
 - `S`: clear all the select area.
 - `Space`: call `YaHS sort` to sort the select area **(NOTE: only work if select at least 3 fragments)**.
@@ -61,7 +61,7 @@ Enter the select sort area mode by pressing `K`.
 ## Sort fragments according to link score
 
 Usage, first endter the main GUI: 
-- **Right click** the `YaHS Sort` button to open sort settings, 
+- **Right click** the `Auto Sort` button to open sort settings, 
   - `Smallest Frag Size (pixels)` (defaul: 2) which represents the smallest fragment size in unit of pixels to consider during the sorting
   -  `Link Score Threshold`: $S_{thresh}$ (defaul: 0.4).
   -  Select the sort mode: 
@@ -71,13 +71,27 @@ Usage, first endter the main GUI:
 - **Redo all edits**
 - **Erase all edits**
 
+## `Auto Cut` according to the continuity of HiC density
+Usage:
+- In the main GUI:
+  - **Right click** the `Auto Cut` button to open cut settings,
+    - `Cut threshold`: the smaller, the more likely to break one contig at the discontinuous locations.
+    - `Pixel_mean window size`: length of range from diagonl line, which is considered to calculate the HiC density.
+    - `Smallest frag size`: the smallest fragment length that can be generated.
+  - **Left click** the `Auto Cut` button: cut all the scaffolds/contigs globally.
+- Under `Select sort area mode`:
+  - `C`: cut the scaffolds/contigs within the selected area.
+  - `V`: glue the splitted fragments within the selected area. (NOTE: only link the continuous fragments.)
+
+
+
 ## User Profile
 
 Click the `User Profile` button in main GUI.
 
-<img src="doc/image.png" alt="" width="600">
+<img src="doc/image_user_profile.png" alt="" width="600">
 
-Set the `Meta tags`, `Background Color`, `Color Map`(HiC figure), `Extension settings`.
+Set the `Meta tags`, `Background Color`, `Color Map`(HiC figure), `Extension settings` and the `Colour Selected Area`.
 
 
 # Saving
@@ -120,6 +134,7 @@ PretextView uses the following third-party libraries:
 * [stb_sprintf](https://github.com/nothings/stb/blob/master/stb_sprintf.h)
 * Fonts from the 'Droid Serif' font set
 * [Icons from open-iconic](https://github.com/iconic/open-iconic)
+* [fmt](https://github.com/fmtlib/fmt)
 
 # Installation 
 
diff --git a/doc/image.png b/doc/image.png
deleted file mode 100644
index a6c47f753fb6337208de13a52cc77c7da21b23e4..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 397950
zcmaI7byyt1vo?wc5+q1)2?2t;+hW0jLxQ^#To+kfHUwDQJ;*|E3GVLhwz#`Hm;BB-
z_ufCw_ia7B-8<VePfg8R)!lFPgeogaW4tAP3kL^>AuIDm6%G!?6Alg`0Tt!-h~1Sm
z5DpIBz)C_wSyn>gqq38|xs{C>92{M=Ym|tLxC&m6$Y7LwM|O7h7&F2<52C**l1P~O
zl0?YDW%>TTcB2I2_|@EBzHkXh9YxDYNN`E}W*27XF~8USst=p|$}Er=d9ksP<~1rT
z8YU6XVT&Bqzpw4?G$c%+=dh@6<WF|w(4P0J=@Kk1mar<wD}Cob9li*EQ5xGk`OvXk
zsr;K4*`l!a!rohLZX>kG@Hq=4{zJg<EJlK)fy*OoDbJDa+`qq{m?D+5=*;7<`DfCP
z2C?a%@_PD5O_hq<m3SOJH?y>uvq6E>l~vf=c7NV%B#B1#Mxf$_E5FUhzJ>N`9HD)D
z6RKdTDTzbBdw4w=Ekh-iA&!p_cpCM=z0k;cwiT#(Gqd71?hW{>cawnc&-py)`2FVI
z?KWL+^<45#%6*lV+<b}wk?M~n6{H1Bj2x~vbNBuOIYvzc7&%Tz)c|;cEv094v)@|;
z=_Icfw=vU{HCIr8V|Z<&!Xd#E!y&)6;9m;>p5*^+OTp8@A^zt$0vsIJ3J&SN`zXHF
z|2|Q#<=;8~RU>2}{7(-Q&n(3MX(J^3>uQ`UL-AUnIml=^!@=Rx{44OXs??`&aH4Rs
zU&Ow-!yh)IC7DUPW)W%Ce=}28H)CEHYYFILp+?6Eo5N0R@2P7yBw?%Rwezy6O{}o!
zv8!Fo7=^hDEl&o%v0*bzRX{;EC-_K-gi47PDk&q*P+0As8x!}}OnocE`S~YsC4z)^
zz4<}EB{BK?FSV=l6%xnwu25WT+63IewqTK~G_$HE8x$U!D6r5Ql-uf$W%!q=SWCAv
zP|5ea9a`^|NYX_3JVrmi{Yj3o#A`#y78W*mjYfel*}1?9uA=7q3hc~6x^y@Vgt-3b
zYNoE{;rg`Q+z%bYn=;qVu5St=rEU^3@;AI6#=TF7dt5^!XFdPn@xw<4j52xtacOK7
zrK+yll<bY<Oc;*;fR0%<Wy<3^<;#`M(-oSqB{SLtN^fukA}{hf5gxhi(s}g>HG~50
zrK2na2<hO&W^@;#Bm3~)C?2o+13rE*Z$CwWsLq|Tufj`IZ_)i<&ls>=SeUt`W#4c5
z%mI;S(XZw+RcCL=xeZWzyikAPHe8gPN8WWW6%a*CsGjRbfoy2rVC?KThl;+T_&SJ%
zvmJ2)J;~*!q3SMtxI8djW-B>0w)zq5%XJ%Ven~S8U=l1juZlOIBWP_Z$t~_(ul5W=
zm9!{Wxw-<MPyBhNSM)JQE2c#&?&TmUYCT#=zoMRuKgW?1;hOz(q^YTCd2PgHOHV!x
z^K}b208id1c03&Ug_NoF2!=}Rm4a!7Wsdq;n8s#FZ3bt)PAr6R5YBAbJ!n?cG}x{$
zevAV6wb(PZbU5Mr)Q0Bg0gK$ZKb}ROQ~(;*i#O-7=1YSz)s4L9AJor^^9<5CDiR*A
zsO+W}g@{TYoROP<mQ6ZMzFVGs*TNa`H@3~Bg1j`n`k=Qd2B7eyIYW`ZpIfek3}xpP
zLB5ci=Klg~oxEAYkTbV)n?VuWi;LAky74M`ZhXg4+hi$)(&2Mh>EW_pp1!fdAz&QB
z_*}Rve!m7}(B(ZgTZmL7;}Goa{S*Wex*=C`rna2yhmN}(&JRv?q!CdNTaCTjSSJ~a
z20#3`s+BKmYqGc@yWSvyHMS3>N^;tum|>C|Ru1=UWC$2U#=0ifId1gUUfd9>l9Sxt
zYNnyEz-H*El-lkZPru^gK<te8EQy`6Oe6$I3G_+1^haU8+f6%1iQS0(y{WL=S%xK>
zSo{+LyuRO$6z&;q?8p(R`##OorRs*Mh$7SSEv(IQd06sf-|9S~=y}gVrJ!dfn^}Qo
z*<zVJu@r4p6WM_pckK0l*5Xml3f8)C{6fZR&)&6crJf?Jp6x45{Gg4Qi-&a9h^6@U
zVZ||Eyv;)FFxSP9-wIQ*Xmi|ieLMfFIN9e>f+}*tHp&_>RUT+V+;4SHk41QEZ0PX=
zX?=NzJtR`+2ON>&RG{^R(k?<on(e@ZfZLt4)m{5j)5Y==&@}h_e7VsB>!KGLJ-|rO
z>kF07-~8il)Qr~p;Y7wW#{8E-u}jUssJf5Q{omvD()-RtGLQHHRuLs*M{Um-xW@t$
zcG64noL|Ak-&?#}7&4?Q0{3ro_(F$sr^hQ4Hdzd+nfE#B-hGaD;&wj@k}hfWNz!cG
zcqp6kV${03YKUz*>x_nI-aFo0;lbYe+^Rp)`#_cS3f94LXQ+@-ELe%dnf;sepqoYb
z-I;i4TF2^*+RW#7l@5JMu{BorrVcPn^_Uqc-ex^K=BLhvr9Y7$sB&EyXE1wDrz2Wg
z#Q!k=hFA$-+^K??7yHD_Fck*Kd{TH*u!1hm+P;&nn5_1aT=YiV#wk%i$!sZRcV>Y*
z%ZVlVWH$m<<B#yZ+#icCO4ctYf4iqZ@0-apT^|yPy1his`Is7Wlbb3`>P6v<=C3Z4
z_o8N7yy!?us<In>z-HZL_S&r<c(1<ct3KCJ4XT4`-Zswov?wZ^sY;_1mLCiY(}f&F
zG(&F3L1zb*FDuv;Jbh2U7e6n2@F)VQf5{#7cZDz=s$ts>m5d%veG*}Q*ZbB`T}I9S
zz^SV*u+h>Wd3KSV=ZCqem>Rs#Tt|Fu;Dz&cbRc*DB5vVyUtXuAb0Zx4rppg+eRXRW
z(B#X23OlY^rA_OM1iMrjfA36S2jlOgim}DviITGDuFX8#!ooQ_{`PX2*Hsr%hBH1Y
zs<>2&u53tj(&y$2N|B4A3&+g+=f~G9DU^RW|0{(JID!xz9J|%Np9-OUF9J`Z|M`S%
zYOWuYw$&R%4jUO>TY%$<y4abn*ucTWSpT+4CNQIrz9pMB!$y7L=G3xgK@m~le&d@`
zhEb!}6kzB<4e5bUwyeQ=FsMKSwRvVQKXq;+Vn2?bIXcnxMszTaTaI1r&6VTuA|joJ
z4!mjE<GV|zHIqIwOc%rC6BsM!I|oBCSgUPyBM4DVJ~&C(eGRkuhz`e0LLRc!I#F@P
zbv|@2*W~tb617i|57b}~-MVh`wAy{{CxPBIxsFOsn01CRS9hp$_Quy0i-dTv_Tx8{
zeH@fuxkPF2bstCKfj;a_Y`Xh8cWS=pLwPcK0=^+u!2tt>xIfqh(t}QrYQO&}s#hi#
zg07D{MYb|~>AmF{%J~K=+A3{`dQ@4t+`BLwtTS|Z>%;e;C>-rnxaHsUBi(>Sv7^V&
zrY%xnJ8%N37$fy`?brbniIix~xoaFjdX3Q(qxJF1hXAIAuy!vIk3ZFjdgYB7D2UoO
z1rqnAvVYXd@M<$h)oo)8$__Ro@+YxoF;4z@kgG(aUXo_ZSmHO}xe83Jx}!;Pt$V4n
zyrOP5vs=_7CxL;?yXSN^a1mv`|7fS8WwAzKV&2vp(yuV76C?X^P<@RH)Mz0ScrrE&
zRe{>w-QFLwKJQgCPdJ8BupBBWV{wacQM6&5;Yr`^Ppe&oLZAM|QhmsgLNuFP5k(*G
ztd6pGlV6niRs=fy-eBVL4SydBy1j$?v^;;S6;R3}TEEVj#u`*6v|w#LIEC4gp<hVe
z^$0V`xnbt}`z?sAFgO>5G<C?dHr>s;S9mO2`evORctj*NX=HRj6F#`ohUZcG9;IOl
zX*EDnwv;w-Nw@yqb+O(W)fwPRj8+P<r)a)NX?9qFnhvmd{p=~Yim$A@Iva_jPc>Hc
z_I7baV8UHDoeZVvK-I{uSP+JXLe_N`QH%GAmw2@p^>-|9qHB(phUx~{su|u^7VG*M
zcUb}tjoI?THTNGG7nd7!Ywl(fz}cI>aYX~t9{r<s)P5uMmm`N0;?5PI@P+ofj=Hsu
z`f@@6DXF=G3Y(B4E2+UAA`u%o<zJoSdOSKI(nIVzb1FG!jXRLxob{IZhDf{lsX5oc
z%sA(H1uiNfMP@z#B%LtPNtfxpQ}wa|y6j{CxB}IosWUQ%yK<a}W%Z(*&^GRKon*aR
z9c3ga@qs2qk;>7S&e_Tzzk0dWnUN&8&7rOEJK%8MOBwJTGQyVsOu}2IRsiz73KMzf
z1M+oWtni3UbW<pc{O&_hcbZJFxE;!+jfl<)@6A1w`#4BF=<-<pOFHCspxGlx3sH=7
zX#FSY*_%f;jBht%PRdTjXX8eEi(o8MHL3)<ue{Z$8xkbm$2o4-D^e^URu)$m`@BhC
zn%lL4&D$z6uW2B2xNl(0qV_(Ya^5R#;Wt=kX%fPeU>@QUXYr*C5AGE}jGc!<5;O14
zEVUmCb+cB~1c`}5D6$tZLK0cKpYHYj&l^-dn}1iS0t7H%%rMsudI(@`*cYo+qE+20
zbZy4<-9|+B#-yt2{GX+}TXxSU2c0m!hiHhR<u&7Q`q<T#b~w;QPkwYB;P(Y@(wbes
zdT-_{dC8-ChcmCsK~Q1f#>Fyy3&o7@2@M&su)#~0W$zNmXTM3J6a4L^qnBS5wig(T
zOK2UR5WlU*Aap!<3H{1=Hc@B;T~wvixt?$0h-Xm3mwdi`9<G#Wx4+GsKr`5fDsBCh
zg|4mks&t_3C*A+`V({L*KOtt>lwR(%c*~l*xAE6&wwyTee0Q*IoonKFa+0kAy1s`?
z`Xc<;gLS&$%F=BRhTD&rS^A<Mq}ytLkPIWVH;pxI|05n3iNA#mYGWPzDv^nqa~q~l
z5)~?%LCHo`+LJoQI%9FMfpvBO>g`~J-3LKlN?{2;xe(2p-qlQHsiQAr^ge=~(!YIK
z&Bo-k83yCW&^vqf-G1hrU7yzzJGR;XX$`@ywRukmWl4)PX(9gmE9CHG+?H0WbZ}``
zqyDPhFhHf}%~B5-oa4FFdjVs2!0U~;13Y{2+rfH$fkGHz{>JUcQlXP>)pH)4JbjRO
zGfq}-2VwOQBDVx8p<tmP7=ZS;nbg+{lqt?%d=l$<5)RmdKL1T@rhrv)hbyf3M!G)k
z-?Z39+HRQ0VD?6Xt)C1fVMo>*k*<juVgs<=t>s=u@OSyV>07%AmPd~&BQp4dwOt0t
zjl8EjU&X9qr|U0@Ftz1|>}dtUh_YCVMtAX-Q$hU1$c<J;yCH=q_xopUeGo4NFGy)E
zZ>jTE-1E71Vhnm84BICMn7FU=e2fP<I;yLm8#z)~X6#T9%X&M#Ge$FriI<l{bb=*3
zsv{de0&yM7c3p69zIM0_0@EFS<tW`=!a7gJ%Dy)k*SRitxSO1HS$3aJm8gL*YGZb^
zyXtX#ys&EwtV(5PXRRfoOM2O<uskC^r2IzLP%10=HdC@WK44FC<|a|Am?XdlfTgcL
z7(LziOoWa@>hB(1?pLfm(|ce_puIpx$)yJW`16>ztUC#_3KoLP__5!my=(9yzw&LD
z*{!K{-i|6dJUTR4&y6)7*HWzbo&>oOjzlir99-({S@L;@950{lPoBd}s>OXH3^M6`
zVavX=4hMJj8FzZDO?kI6^|7%1L{CpMgl2R6okPtyK&l8fG3U8M?$mh<DVToL|68WQ
z#?|`XE;oAFE5wpsO1n(#_HQ&nZVZu%XOU9L*QzMZxp|fPXv3;=Unt6ga-y$q>)$$d
z-CGkn7W-@8*8*;(${fOfV5|Y0GFRB0x;{U%3$RJ=Qfwi&BA>F*N(4(oe=<r(+Ada)
z%WF|7&$E5G)dbbe`i~K=^93l?R^??K-8#SZYCB964E@Z*z=}B2hsYdLEs6d;7kP=F
z9d7T;=J2sSlGb)&yKhOn&ly_;u=R-P*m>eKH^zItOyq5``D?s=xW7n<;4nAL(Y_K}
z{uOS{mWC1!&MU-g=TveZ<qYF#%Bg!YFc3>|a9wS_e>m1XW(u<zZX2V^B>TfmLjBT@
z)AuTxAJDR9jR_{e(H714U#Wd9&`i;aEztOud04}om*xgCsZZ}l8io_4d`AEE9>pnG
zv@lUoBo8ayNgHZ1d+l9-1Zob!w&3TWrQo!*zV!ik2PzopI^aGLvmTQI)irD040Zh^
z{z{s_*-#MuMS?|#3d^WBa9QM+$|3p$pDW7*n$=4hp_vp9&-yho)Lq%hs8*W_1GdUM
z=*|BwLQhbVZ#{JmwHaZ1YnL}%Xl{ki*w(pIIaewJp~?o;=@{A805CFo)-C9WzSw<6
z+2TV`$6`~|`d!!b@YYV!zCOCejfY4&M~I-<<y{NF^CWVJM9DkAbxdO8E$lIYCW@xq
zj+%|FZk&mkw5AB1uapJ3I7^jE=`C&JKU2IK<ISsq5=!(*;xKh*tPd!aIIy%3v=A5*
zMY9)|usaK71C}gw>Wevpnf#gjr=vx)XXm!+(wZ8X4*pK}N0Q1Vh$gfR(5xPCdrhN4
zrMXHKPz#;QVpcv7-d}A9P6N?5h^u<Ad`b0ufQ{ZDUb;Ga?lhq$-vI|-Mfj!u9H}=L
zgSZkQh!)D{7$o~UCu|HIPOU=)!ZEgOq$wGkCOYvgIT~G1c0y}EBJyh`I#o|hi_t(V
zEl7Tz!XL)6fCeu)=AO4_>gU3GoTu>gCZq&<ie~($j<lq-m?;E~L%eC8Zk$o9&G{xi
z|L~wGf>XW-vjX9iQ%jNaf8k$CO|L?#e3hAOB)%D{gTJmx?8$lm?(~D~$Ag*=p{#F4
z{+$NKvE2Q8I~B8pk=pzFfNhdP?w^gAX~yzf^NrLzm7X+qw6=cl<2X&c4J`U*s-k*f
zS=*thMu?k6;=W%b==zqF8{SIyo=N;ROd#wwPUDn2yunKIO3l|5=)V6a8}ewu?>&8K
z`Rhunn~<wJ&4K6g+g7aJt6K3$b@sg)Wt7fo^bJeK-$6e!TmM^OZ3L~d5?9R4+ke0B
zDgBEV?<uF{IsbVEGfkGK?br(0VAo1zGncl<fQd@{msftm6pfS~N4eSj&Xzqf-_zbZ
zuP9vw&aPdh>C|e-h+pa`i8+;|zr#I?uV*DV2Za!#vj}vr@$c;%-+KPIf;Hr1pp1==
zjxo5Hk3srr`;pi79C&eS^aymPzd$6*%zxm~JK0uo@+q#n>)p0`WeIb`<0H*<Jc{SX
zs+@?--~O}M|A^jKLnHQPlM1;Qe$7_^y+SBO*3Y@y|1U9H!`ubz-veOsf0pq>qkpL*
zo%VPNWmX%0eGYv%yH!XXL@xkojm*c&d6BT_yrESiHx*AoqigaEt91vkqCds)vjZ#^
zY(DfU@sC!+p!w)Zpn@l%gE>LgiD4CKV~zLJa`weohWB<GX$YBczQQre+v8dKb)Jw1
z;B7`)hzJ#MrXU+!=WU|c2bD||8SKf6IH^QW*le6OcF$WC*-LarqJPg>a(k3QQM&BK
z`}AbjQN7W5GYY)Co#;FHpBwyF7cux}w>NE@en-xr8vLhw5t56MWi4V)3rAw;e>vZ!
zCW87-{LZQ;sl?~RS6_{y8B;<l)<Dh=iR19iaZ8k^#IpG5*b@_qZ4u|#64cpjBQwul
zBfAv0?Z<ETfDdNzU0w!&erT@f-EKrKuCT4(*)<jV-nN*zcL^NL$%6oNBKMb1m)>pi
zVdJt9@>_gq*;4EWfQM;T3ymUOA>BlbFlZCgfXemL9Cw0+hQI(MJncVf{8QzBlhe5<
zxC_k&907Sp6MvW8wEsrqD*`{khqFE6fv5sccEV}wEF#wTW7&N8cF23D*?J1j@*U1#
zakSolD9(cHdgPhQH)t_Z!6Bw^u%)|*tS@CvW-he)SyI%Ec)3ml{`7*4E2z+@uzJXM
zYkx^}#$|p_{^Pqco)gnI`d`TSKY!;VX%-75X;!2GFS)wo4F0(`<O^D{m$j>96k2QY
zsRnOGFnkst3v)cXk0v_5XH&|A^E2sujceD@n9A4JmG^J}Dg2xKcTl4?x=H3+#>XD;
zz@@QK4TR5!?Y<b#g+PJVMLRg}mO7<3L1nL5=n8mQ9b?7+rzr=@6!{qczuN5ov*TD4
zToHuxU)23i<jqG)TH4tOY~a6nsyb#N?Ox>UPZ#t_nrm#pND_W%(>RUb?jhCIjS_Uj
z9Bj6^wI;kFjeYRKCrCiQIco57pQ700Ad_Cc0kUkh`VfMPZXQU}fP!8Fx69tHEzbFk
z?w%9_iAvV{^JA0RfO+26$nbxbRdz;rJ=Y<cIX?hJ6+%DSzrpYov5iQ^u@7~7qe-C=
z$NJ>q9BsS!$6ezNyVl;l8n7#?A;-mEf}U-s@FpX#t0nWL<4ekLVr$Zw>v08hQ%hxs
z4!#I}wiAfCHIOuoA~VO}RDP-MFMy*|p7JUWB}2W@c_H|_Z(cW0CQ=fcbZ~pavToYb
zTw?TxtUzz<|J9TZL0=6eOnxh89z0LwP<<OoKKf6<kZU6@ZsZ3Ry%f(AVh)~J20P}v
z=9&bO+Sk}@o2R$;5-Ge5RKzBh_(g2<c!Q90pYMVqU(whzo)mwZPw-RCN?@s^^VqTm
zdHyPPJ?{t7r{bPG-P4llY36O!+wK+d!0^b)5xNa9^6bm?-s#A1REU{%@CMj1gc4@9
zb@IQ2x&MDO_^-C~U!;z7p#yiwWx)tIt?r-J@&B|g;t7h!J;n}JnO9LmCd5o8w<OKv
zF?54ey+a{ntCD-?^r-EAs5O+dE*|%Se<sf=YBDe1KY3KzUdGl}*XZ}lzSn>XG&%1*
z%F4Q6nI-wqKJGAmNf<hKgj(6BuY0^0si$WOwMt~f+9x*qS%|w7P<5HgiJtdV1n9#a
zgFc+Um`g;mMm%QFltXC7N+z^>We1A?O@99GQmlhkd<04pDU968wGxg7?(9I6)Xvsw
zY<TPlxky{_#VcZapt!~#V5#6aX49cMA@(u8(~YaU4ZS_gq@z?mZM(L^Oy`i)HGeS@
zQf)uM4283j*e6Z^*C!+Mci`3GNR?(!+aQ&OgM-4#35$3^+W9|TZrErx4!nwg<AO^L
zD{y8wzFiOf2dn?^`d_m91)<(xCbXKv?mt7zzu`sHp8CBGbu%5rgEz8IV|!s=o`r>Z
zOEzaEuBVqRUU%j`_HOU)TovTH??Ep*Dgr|M)dE&BxgHtfiC}~vcheIRS3587XW;mq
z=r=`TVGN$xI-C3W^dAx~T4M0wQjt#9*Z@D+l98}W%p`4e$djjhXi_jY?pmh#=D==<
zk`sXn;`dW>yLP}!S>(<I1k)Ef^xHG|+w7{y1}V|UF4lx6XSdI!W?Yl=nQjv-%~VBw
zc^<*p%|n`LG1@cx;VB=+lWU3PDK8XzAcMfqDuAOQ^>12IK<~fAT0~zEo<O54flyyk
zN|_rTLo*hPM2_excd|W>){x`&IoDjDLrTPa(S5L^1B87cMWn8Rz<pu=fnVG48|hiK
zv$3<BK$Y`5^A|j6BeMc#|ALBa)ROI}4m0Kzx-md*OJa9lmF)Yces+r4yw5d*y-@w8
zUN!qjdlEOFk-m*}x)TPcZ*1q^+Jhq~6f{l;OEb{Z<qR!N>&mZ&9Q|?5M+|A&{)FYh
z*IO3$jASTEhOiLX*?hMqaMR-GCXU_Tju;~+G8WWT5mR;4T)2?7xUu<CJU~n8o5sJj
zi939IYtG5XN2k8yQRVvre-ZnJoWqOUXKrY?-WD8aJSKQB#5t5~k!CLW_h4jsyL9;`
zDUXH)lhH8NMsTT`z4{BmbCUvvfXxPdBfzaSBsVS6p4^(=ue)2)+2MykDCfry+tq4H
zJdq+hOq|pL2hR__Hf|>^iwO?aa5)8oe>nCJJJw^w63>S2ndm20+<z{;QuZ?BSQxhD
zMe}_{X6NEQoA3?RVm{}|lya-=%<!$P-!<OX8>lCI6L~P(O8Uh*Yd})%&2vd6Fv0`)
zJ%w+ZTD5zrYv=qnd#m<}3ZA}H9z6mbYk}5;c-9^Vi@vjZSQp`7;Q(1D-x+50WNNUT
z%ZY|AXW(L()g<F=r8Q97jM~7dZCl}`TN{!GsWh%<VGDU<(@5t!=W|t2rZ5Sd3dqH1
z{+U4<a~Ghj#F4LM6K+K~o5d?38Y$j+jV(U<xC}&727pbU!`KUlMXU~N^kwwfSzM2n
zekaFzlu_M%A-ApK{h1;7=fD&uC;wioSk;SeS7&<EQfDgL$IG(l@z*D(wD;`5-8f*=
zDnVdUBg5Y$gCT{z^;Z(n?w%ary%fMA=(a%QQfDC(p*Cf-n26C2ytQ@p50V}K)s5?l
zOSFsd6I07Fj6AH2ROX`eP@vj2On=IBi0UD~p{1!n{+jL<$Kyzlu81Y~(Kar0t0a~8
zs&0*(4cAFloeVi*C0U5gR*=w@dhiQUWAwhCo*sv<O9=}RIXMq2sNH!u%-~eWdC9T&
zOSfoXtHmhQw#m!2)y4A3vdwcRiN1yY`R^Ux3qr(EjndncyWL)Znc$nRn-hu6b4}OW
ziTf6pcPlsEW1tJ2S;;5zFbJW=^Qp*FZ-Q*=Nee})$je}BA>j<y7^D8P@x^_n$6uGF
zrPYYXhvyq#L%?~oFY{L9x?IWMSo}$_#qc4B$B@8jZ#;EZ@^cBVJ#lou60qk7;{ZA&
z-(q_RJ!INlXNtl7dJu<?ftrnd#Y5!N&0tgUiC61TSeVARa}Ei3c&-Oqi%vCzkIAIL
zxol5KK{l1sH;yHP9?I`q0z^aaqbAq_PWCx0PNF?lD?_k`M~KvscCHr^<$VlJa17;}
zT|x5bDd+3(Qu00H%?zWAe{?1Y6eHx4tlQ`oG1#z!*u#9TiyxB&cL~_~=gZJae9qEx
z86|@qhs?G5`m08o&_n)qe%g?jSSd<#^m2h!bF+mH4qPcvb&KztF3soA;<Oc1dAQ`H
zmu)nYMOip=@Uk&W>s)oiEkXQ0+;d8s4KA$u%`=#lq)M7HnywX~uZ)Nu90<L53!~+*
zohxHmA}!2<k8fvDr1lP~1b+lEgbPs~1#Hh5W^1U0BqKDg=m-k1`xe~<Qr}d{2Zf{)
zVr@+!4{-^BUAnnp@)`Mty^td}i!TEr+bzHoALkSA8{X!*?~XIcP~KKG)>y9_qR&3F
z$UkKW4~HO3bM2}p?lFo1=()-aE=<~F*wRo$lgVsJpvQMXa-Lt=IUk}1FDBY5aN^PC
za?p4N>{1PUds&k&GN0qjE9qu$zIrSO@t8yS&B3nH$GsH3#obS-t=`c3O%DRz2%(RL
z%5|G}1gT}9tAp(|Foo4bN=8PFV1Nr)dD7&Aezi%7QbeWX-|unU-Xr~a)>C^GDtosJ
zwz=sQor=X8t><P6R9M;ltx&)f{q`@$GM>hh7HEaGYr&KlRMALYm`&Kk-2AGeZe)FA
z4jb(!fH)EJDv8houfW8agFgt69UIYdE+{V_!{OI`GPZW@i4Gzw;B^0;u8oz|n6SEK
zIhA=J)+5b{1(w88VI5N_LC<x4Fq5t-*3K`nf2#bX{`vg#7mGVYOBZaxV)q6irds~Z
zJ)UIK%{uhY*bDwT%}3>Z&by`jKYeOr^)RGnFbu3&l12zMCw-DP*9r)MNw;rln=OmD
zmFR8bOcr#|CXmj3jV-50$y0dWU=3f|I5iQ8`DlsYpzc33?PhOXWq}*w-rs9KKmce1
zd|RH~QttZ+gi7phUG!>Ab;S)ns@IyE6ENv5`|m)-V99X5OR`2L6!(zcE4k&I>O?*B
zUuvZ3M{v0(a7(RvAbH1tbxePQObp}1d|CTsacdOvAzW^^mJ_JG;=GDVMk?iHCup3m
z?o!*~+7ui4)#uL0I8jYTlp}MyGF9Ys+&wvmt77TnBWG$+e~d5E8Y-3Ut)KqT5*xH;
z$c4pm*WY}()DIz-@e3G-9|IdFjxOv=Y!TU^3=PE$Y1lh#8GUuTfeVK3_E``)Tjm@K
zQ1s&a=<KBv*y#3anFo>>@auu)F@6Uj=P_YBkgIz_4HNcKrYRDrDS@a^#z(&!QfW)p
zsl5DwMagHqe%rS7F@I(X0cNk~)qDr`eN4<hD1@e&`rOp|Hs7Ut-feTwCf~HY1k#^y
zI~w4bEy!i5vC$-|!tcf(>F_2JXS1;gP!^Y4>%0W1&l#avvF~{+e{PMGGOUfT-=vJ-
zUAgqWDD`!3ZAZP7Zhcw2MrBhFZ1i<IUNWk^|C0*fce2TG`3SnLJUp)#4;N8#f1IC@
z;bO0tzBxW<<e|Qw>a%H_T3j#8^*Y}qY=XVyh;ZuE+Odl~G-zn$XwwF<Upk;0pvot)
z4Az*HNXw@f&5esiu15|cJ9^NDel^(|9qa4N#l}r-87B<v`(pZ?29fGgX%J0e+=wuk
zFb>g8w2AyOe#^SL5Q^3RN1oY>G4V(Ksb-X)f&QGr5eYWe(-gGfi38(8)BC!?U?whS
z>gx4^Z*A@sT8@90{dINru;4jjw;kE^pJ}9Q>!dSXv=4P}W|GNzy1WywT}&raSldh#
zz!?LT1TLzP(rd`;au6LMqGVWUnTi?ov39P8%kY1e=W_@~2^rQ(28}qleD&)eJz7C~
zf4wrFM|H_(PycYJALj4gKj-dlhTEVw!>qmpmZH?D%~e2iWrV(`vxz5btGIkG)q6_l
zsi`U1x)d1*TK>F2S-ZQ6I;`~JOJ+u3B}steFMGI;3@+3S6BLslEqZAB7T)yG=Wvvw
z5tJJ-Nsiuotb+$ngrh`%17hq6`L#fEX`Hya`%869d{>L++t`*5kWB#ZR&f&2vA-@w
zC-r0HxFUMk1QbCeE!LXAF7G*Z5<HAa1mqj*7ct7M(CsDia7YaPt5&tm=VhPtV24y0
znN(r=N{VdP9qE9$PMWhS%A4lc&I@#?S|ps!ZRDCTLZVtpda2UfX%q&xvvli?PT+qS
zM)ai_zTKZed_ir&E3sSd<nJR)7XXF0{IcIvrQqR;H%w*261t75-?xTw&4La*ombo1
zsd)q?lQiGl%q$+$tyOfsxv>XbEN(|j)5`f2-?y#_!d%yUzF}&J>sDdCb@YYH!IGw-
z3kqag;%DVK45w>(L6SZ*c4;fAnwWxm<TlyET84#Q>IuHjHLvxY%wsN6)(h^Cz?y0{
zePK@`?6pspiOXkd=TTU0@s3S8s5Z&XS2c^nvxwT1`p+zKaLs3Xqo8rpgm}gDHa{&6
z;EIO@P2c7wz$-Gwv;33v8<G8<*my@Jz8#y3P1<y8U!B_049Ah34NLc(!;SH_TYeg{
zAHJer1m@rRiV9He@=)x4tE#O05M{92c!jb%?bcMro53k^pn>4_p4e#g8acodS8PyG
z9pMQZ4;P^Y8=~l8xk>5ydFe5fe;DsG@RwqJLqaBDr~uKo1c{py>*n+6Ry+#HIJ3uI
zQ|hhRH$4bmX4Cv(xW*d+{PevKP}A=X6a{DcH!R~aV}~wCG9S%l_z>nGl|M(!f&qRb
zm_|9bt|iVI%C*<B%fH~mCbh-V^-u`$G=<{#?d^>y1FEH;f{=-DDXc{K4aEXS66s!;
zm`uhc`SQg^2!}O);vzJ9a}>8AvDt;#k_oz79@GUwr?Nr?_=KpWkNQ<?DV0&~btF0n
zFK)&^`irT2vQ}|=ONFJHyP5yB49|01r^GQ!MRcJLQ4xD|%x_=F)p^vU#>Vcw8Zhsu
zg#xg;(BU2JS&*~^B{T*vi=W84CB6&i+F&s`pSD}dqx`gtc%>b0p$8bZa5YqSbIi+X
z;-U<uIR+j!3_gqU?{)PizufQmGLhx8QAON!(TbR`yqs&~#IfN&O}=5R@>w*c?J{_o
z&Ro#L=X(ubzMKY)_bHZjy#Z83M|hh2#60>xQ{t)LIjMYS9Gm_;X3{}d)7TqX^r3`m
z_#~}rlq;22<K(mFZnp4n?vlss^0pwa_ozHs8qXnDz3lPM8ea|Y>FHtETK~z(c|^-X
zduj2cZW46woJfKmMejbAliIPD=jCpy!)jHAUxN4sD<wIZh?Ak{$9~7><m$S^%E;8j
zrp@5(;;NC##xl2MkI1`Ai&jgvl{?X8G4^{4U3+p0Uf)x%R{FKH2CA~M#<kBPnm%Rp
z@Ru=V*F^HpQ_|>TBBb6~g3fJX!@-CJs+!}@sb=q5D5U#+O8`s;B+Ygs+UzkK!rIhP
zKfr3o(sB*6-U7L$ktl*P<60GHP7@vbu+zEs4kow}DQ=e}(l5olhT>x)G0x3riFa1a
z@C!S0UBbADN5_<H6rE=7WnBTas3oq4^EWcFaPKsXx3*;Dz~=|yQ@?eQ*?u^=WIs3&
zhQA9jBaJYsQZw{dlUFCy+@C5*wp{*olOE6819J5bJ@VMIcaBTKI-HPT4H8rYrNy^U
zgf^<z{1g+?t?QnP4z`x9kaKg`1PF?P4AWwbTxr_G(=-jXRxfQT%TU0C=RLV6TaoO0
zu<0_ZqSO)6#rv4H`!sOQ<qtV1P4bDE^Ae^n5E7{of6(qRZaY(e)2UW#Rxf2L=C)ne
z{Y5f~K^SJH4bj;+&Z=Oys3wY8jsxIHH{6wO-H`dOFh96mf1e$jAevq=d2mifb{!7u
zFf-XGXz9pvFS6aG4<NkQ$P_XJUUflA8WeyVGUS-2WzmV{-AtKY#8vh$@SR`ote_dj
zvQ<UD@iiL)=!2e&J2rn#luFWm`7!plS1dF$-KGMG>=SNW5%S^dhE80!j5)WF;O%~V
z*$FsIHrBwJhJ?17z5T=?HEe4T#!JJU=RiNL6#a<PrXHdj%)88E=jA5yejaoj?dAS_
z_C)c+XVz1zn*(0Ny22?dfmA^BH@PcA4L19jyn;fqlJ`C*&*x()4;R=sA$7_TKMR+f
zMQ(9nz)qTTid2UC$I<VD`ej#P1d^@#C8bxlZk5O4tIUK`(=iHo9GrQi{qK^gRVA=2
z({$2P*d()^XPj)<mcz?p6Q%U;-{Fupo+l<Ye01S+QjFMj4Wv2yWM<QSq`^^n7)`S(
z+m_Jm(mK?6c*w1qPg%b1w4BwL^2C7xw}zz5=;{wJ`K{zi1*>Qz_?ET1H%A?N54^AT
zG(d>IO#AMbzQ^yy6@i|+ju@s`A3*ug+MgJo&4qA0kK9a}n}<Zm>cFIcdI{b?Ue8Q4
zSpZ33g;01M+R9W<0~uhpJ+*J=vMJ8u?gfSEbSM>!=t<wX^_h__he4^x=A)Ro<qm|g
zUB2YBV{ah~<3Tt#&}SWNPxCzsv|wK8C-mynK1YJT1LwCofORH;=gWoJz0)Osj50jU
zfS9@N=SyCywL+d18j9c0Foajff!(LmDy_Wu{&n3A58(=0bLHAv>4qgWoPveb#kc-!
z2wf998$k_mWB@R=u1NdLmHxiKG!?nR?TJ+6)Yn=l2^|&VOrf}a7-=L4f=9bbPD$JK
zY!mX&%^NX@T(O!vmD7Q&r{P>|^m*&VBi*!2Bw#DlWwBevb)p14cr;1q8&+|U9pNa0
zj~YPxZ|H$9xgtj%f+@<enTT88ZBOi2Um!NPl7j^$rQS>nmHLAJ;zEumoY>r@HbSbf
zcOdxY#m0`Qx@QhCoY(C&yUb0x2fRh!3x19aw%;$}RF1!hu6B;ouNeHp6j6#U9JP9z
zmy``Ad+$P?ykycQL#g8MRvoXxMGA9}p^r#_9D~n=uI2B%=R1tuaV`4Y1i^-cn&O@@
zf7zo!>R=wSR)xxdSWL<fOZ1|*y46Lzgk81T>4jFi)3J4y#T^jT^Pt|Teac6d0AuwL
zF5Pjz=F*`&BQpZz!1lNyE!XVW=IrpLx5Tya36&kk7!%E66AW6EPCFCM{qKkzRLxZ#
zMS-9yJ0d3I>}=il2N&aTTnlCw+tEo3{mKlQ3Wn$Xvy@0C!NeAf&|?cjQ5wl!hqSh+
zse{A;#jJ#neN!LQ5skn7CZ9(``pM-d5DKYikQ2D7(4*glIOHXUL<$#j?lKy>#FBoP
zvP3nhE1n7~?ZYJ}np3eaFd^9-`4mrJT;ux2o3UpkuF3eUdnv0Z(EW;bn6zq}Lb%XJ
z$dNX^FYH4Edwq0oAE_`lXFi+%z_J`5>z#3+t9$jGj1DD8Y|)_JU!Ko1T2H2`&%hM-
zX!RXg2Y(9>53Qp&`Kqf!69}EmNfG0b%)T>0WJ}?9h3H*nAf8KM9P5B~2$O$?jObgU
z27kA|yCNUrI(@W?cxq5->@gaobz0HWrSz_hem?)KgYMHDAv7cigzFcsq6uH3om<57
zd(t4OhlB4U<=jjJaP_~9X8Uec<)(hW$?;8Ys$_(iFyyrp2qb(jWp5c9^fAEWm@&+(
ztHntdWH5{HGRWFFFxV8=k!l2|t#V4(Q;PzEx9|t5WnIM<KjKeDNxMxHo0E^z=Af(f
z^aGC=-!IkFb3jtEz2i8Dtexot*DA^ZFt@b!pg(b1ZghSJi5mC!alp^jYPRJ{Wp^x>
zmXUw4wjy-~uPjGqP6$s*?(K6aI$1sAJ&T@5d-yg9Y(-p2TmFj7ZI+yM)L}0Bh1XTZ
z2yE?eThWecUnw!u?+Z3aSgOD}1(VyvM8uT*9ec&<m3XFEL?fI0YJaqtAOXGGObz6y
zwAW(m__YbMmbCS#93RY(MDFKLb;rluKY!G#-B&s-%eAAK`@=Vh(%qI|bw=mJ9rArr
zVIT@?tSj`K(EZNxU|%@T!ZLON{wlob3$#%2{EuAY>eZ8E(v;eQ{s8yu2K<9<ZenLE
z*WPzHbt8O<j-N~{?=dP`-k_~w&ZrB<q;7DA#<svUd>{*15m)bnI8q9fk<3#Pq#v9k
z|F|fVqJ&~05H}Ql7sytqH@|r=vcPlReljTvzGP#y=)Ejtr|A;6+(bLlnzIYo9~Bf$
zb()o>QJ-IU#A#L(eDm^llhrH<wEKnP6YhB&sQgJO^0UioEWb=o<v?Ih{YvJ7z19f`
z|4{`YOo(}@v-Z8c)_M8dQ(*`R!N8i`k80=`MM6CdXPr$5b=gjLf$03DBX`ZX!bows
z%78X0SK4S*19k^c<LzL6){eVc=IQ*8jOus;8H<vZ=%RGra5mBAqJY@tk9QmGzqPm~
zbgEhONIZo+^JdZNg&Z79L1{6cvJ9no`nvbsl=(lsmG+)c!s#KZY8XI=AKRTDDINRW
z?)=rQ-RAw)aPCeLUDgwKjT-02t&ZhklbnQ=(^H4o8BnLPcvDU6BGZ(WzRn$AizNMo
zvJA1Mp{4wqFp6DjZ58N`NaY-z5DU|{Q9|T5d2;-spXuz?dvVLSQ)IMd#Tft*fwX51
z?QFPcC4P@fe0<dBCLh-Pihm_xZ=`ccRZ5Q9bICZTB*@RPHukTlFK!*<seBBJi7b{6
z&~UrKP&E%iiJ+8L&ixz@rhbzC6eU76K}3wFH8E$xL=zDwMxkBAOHuXQXu!hzOC`vi
zpqq_{HvN}?84W(JMMI<^U;LvV<{qUIya0aOL&GW--xC|*x9fh7{tCs>a9z?N`zpd`
zpSxjuE<xb&X<z~@55)d6XYy;k61=OMvOsN4K&`iNV6ly1ca}We*j>!!DWd8YDL>e%
zBAs%hV_fTPr_)s0q%%oZIly{ya|i7z^5XWXks9<Y8%g$v37PZfv%sxg!rpxYX<FQ-
zRz<CM2W6&(lBz`$ni!)H(O^Xi*4nYS0;ZKR(G^jo2O^NSxYAmFa|%*=GV;JFYj=S~
zL|g=M8ON)JB{PpKdo=<f6Wje?!s47@{Vi%`T5#Bk0(76Y-jVGrlwk|9Po0EfM+k~U
z>Z4^*(E%?pRNr3@CUrqyuqR8t26}GZ3{DAyc&y~+kE;#v_qla|d7OiG!}?c);y!nf
zJ@$!qB+DOAOvEj9Kg>mb6}_M9U?Tjd?iM|YERr9%3uuiv6#mq3tk8$$arv?2ZF!rb
zt}xNntQ~Fm2F4%>&ZRVh_pyBa1a9#k%LrFTuo2^Wk*$CSq+WPA&bf`<19M*|zW7Xv
zVEVi3rsCaqwbtXB!a$%$d@RaR99}{FWm9)<nBzE4+POb=2<WC4bu3>2^Fg$Z@s;|i
z3<E^kyI(#XnWn`tLk!jdA#Q$4X+BI8vPElOuVgj{yFOHRK89Vv;!sBC`D+Qy7Zdi=
zKUP1H<U-iD7ivK=_fPr}1kZgM$G<;5(U(8D6ER_M-+S&ld}ixp_^P+UB2<mgvjZ5j
zPv<b+JtOm|ZIXp)K^AB*C|8XM%w#iS%P!}Qftf*T4Nv*<bObd|IO3Pt9iB#u)Us42
zK?0%^;>1BZ@eSX>Vpz0HiTl%k4(Sw^p#yboP*cO8f_pr^gPq|TvFPXb_XRV?e6q)z
z&wDMOe^#KIENL#vb#2o6JW5I4D27tvFKd?MV(`#d??aq<(Zy2PH_Fo=evRj~)J{yx
z7fc*y)&87)vya=bLS95iak@fpX!;XgARkk{|IE%64xnY4L|iKts9JVJ!a=$GV3xsP
zMrhxoL#1T1xVq<xZ(Dv#n>?zpejLmZ0ieZTClkHz$sNvGHbtkX>4)ckKocJLY-D11
zRm3|wxST-7-1HM$Pm*_FCmOkgP+bI?qt|e$ho8;`G3fL*cT<6U&(Z&26w=)2Fm^%}
zvC+;KQUccvWXZuu@tH$N<UL&zyHn=P02H)(X-mXjvGneMSJQde;=v`4;sVxJ3cif$
z8f(yno<HHgbk$m>`#OQ1mt}K|?v+l1rHc4HU`hJ&UkEgCr@61ojh@_WLx*zN&zhO{
zQ_v(@;l0r9*Buv}IyE-fIsH+le~JO`yBq9b0B4x@O)`mH-T2{(cWwG<s_D53o*p_%
zo<C>?S&4JB>6B=@dQy@5l{mxh-nun1_8iRsZz=v39Cn4Z=m#iiht(>9!|N_55|c}C
z7lj7R{o66~eD4zn4^8j-X@y6c*0)&MchbEer@|~&FvI-rmw!30QP8zhXEQrx)Ms@_
z5MybE!$|V>`i?H???rpOlO2&8<cm|PSk;9$L%5mlrT-}FuqFua#zyLq>1BJM+V;$%
z9j}J|6$XF6i5R0uCgGmvV5+`7M9>X-VgpOg-RWx{1YFAsS~(BU(<aNYcx3Qj3~?Uj
z3K&75Sjl?8ql5=F7GB^WyH%S}vT0knWG_5<VUVEPA=fBq8=2>@qL<wI#AduhSWwV<
ztJUkq&Ux9DH%Xe@qf1@CW5UpJoV+ROrS&o)Gx6eV`Nzwol3;fa!Sh|W!L0|P0!5wl
zKGBZIG^v+7clxhh>Sr+PLl7y+Os3*8glv(<mxV(q1|)FiNIS;?K~EZju{v!h)!TbN
zpu8r1d}f*Ym*qe4o7r|Ng{Kh6yryi{?zV?G6|1&l&Xv4VqR}WJ7EO2H+LWT!XXYv(
z>{<KF_LXI7y6jF&f&#W=x06iVWOj@+(ec2;5?<u&vtI>DpL_mZ`;ZUFkhxDY>i-zw
zd-}GBmOT_CN_0CWEwVP?)fwB+{baBMSR=jI*JvIRIu6?CG~C`|^@@&bw$KC$sJg5J
zUhY^^n!Rp#9QW7+a{4Og>%BXK9<&HjXoc3!w!hS8+Su3>7c#Q-2n*EDF45lz>p34`
zY-e~K*SO*b7LRf|d5<`=a9E@{hUnc5^Co!RorvlK?#nXSE}W~j-3z@&eV;FEp1uqB
zVtzf}e3xioWOsjzHH5PFXqy=JrE>(Ij6r>3EM%fx-Phg9typlEjt99TY|ONK6=kpM
zuFcRy+?}89hGv?W?#_Q}xfI=MuJNz;sY(>~`MJM;5LRwvD2_teJvIcwUR1I=*)Qsy
zZe$1Yh;6oPc5t<CUVZnK6xV;{Ev*EhXtY7y0}%8vup~g~P4YFTM1m+Uj$yEHA_my?
zw_a<wXSGph@xeu0kyb3mQ6e`PF{?a_YAOBX6t}M*It8Jz&efphx@wR0GW&*04T((+
zXL-=$^qu-L>5NADp+sixeLVerLD}|4?Fqt(&uP=V{!J3X;1ycmF(BY0$!Gbj_8{o_
zTg*E7o&enaSi9Q*$&O&rv<(jU)hZP7Slnu;0Dazb9LDFC8?sGEm?timQJVCA?QrW)
z^hO1&%IW7SrND0Lil%c%^-GrI{@h4|m8{;m;L09F4oIvyI<)Nzb`Q$^*{}d*hJ4ic
zHoF!+u?L!*2PHZ8c~ogRxu5i1$DB;ET)L0p8Iq&;t)XKuR-b%tNf+3Lx#e*!^*UK@
z{)y$M+(FFMjcdk4FpAZ$J9<nj_+H-2a)i|Cpgn5aUBG@LEAAk4j1v86V-6{GfHb9o
zBj<TQWA3K%m<O^?&0;kEatdaj&5V*nDQ!^#PAh%$F;<%R9Y(Y&W`;@u0zZ8nFx8V#
zYOw1s^&_8I(`@kE!XNJZ@E)o+AxoR_eDk)i^s91W=5#^9%>g{y`I!B?!Qy3PMu8ZC
zFeAOEJ-IQ9_2hl5im+jbcL0*Ah)Y^nyUf9R%dS8>=}5mF#L5oKu$FVz#`Ub)EaYKz
z!>dP8@6O;n8rO88a51F|<#n;r?2-HtB{_=s$x#~FKA#nIB<+}ZZWe#yw4yzsy$mH*
zY@m`!0vKi7=GcY#j6@V2J7|qG`$YA%^=M${tq8f%O?LfNLBDM=A2P>W1$^6yn;st`
z+5Fgf;g;3k0T|KEz?d|m5_p4Jebwi&jfePKBMe%mEuG}Snmt=}T<5^BA&a|Jhyh}v
z0a1PzDi=%0IzwCL#A~o?{H2#G?P%sx?>MT2uCr`xQY7HB+O)-NUs;3PZGQv0X@pTZ
zyN+`2^plp>fGgNu9$6`PCY_$XkY$bJ`aSD&dzYB*(tT(?!P{b4){@--hG?DqT)bvZ
zmE4=6{YpzJ(Q#n#!pY&GD?ea@x_11*P5e~wxX$G0<nbcYuC#8w=vo>PPiMLt>7ndp
zfOTM*%tG43h_|DkHq+F1wEp&!b9cM}w}$W0fg8fg&f_b0$OTPvsyvwjS<E8zAi42A
zuADrEB)^x4!P40bxP|ry@%4nZ-WIiTJ9BQmZY?1|q!2iJOUP4mGH1@>-ZME>c>cIH
zEW+hQp3Y@|7tEVPAoOBfwEDqnqbFkKqjSJ($`~|WHPq_6{cbzVph1;^1GyU>qr+nd
zup*?0j>VqzfohI`rFBByX3==%3{tD*2_O$CzR0}Bb8c8_z9e|*h5drJYqfoEg*UJ0
zTP&Yq;JGI^EQur16GGTV>$C+NmmKAp)%QxjTPJ9x4&<5tOB2YX9y;d0Rv3xSK9T2l
zDRd{hqu%t7{Y)L1RE*f6fKSqWxuvZq$gd3{zx6o_%dppf*q4*^nVxpNHQ`zVv;UY)
z2PR{IRx`VAuViS*mif{=2W0={(B9P?2CqfW7NFyd0uIn0y!OHj44<zyUhG;Nhie(P
zU|3<{u(O)lZxH7$x`2uhSe4H{!o{)gtzU9VDXo&b`~76bp|qMObV+p<3WXl-OMZqa
z3Y;Nb5WL()%DJz-31RkmOE}YVKV8%{wxO$>AIHicGbN<s{(P6|RWe^=D?To#G;oS$
z?7j^=lN2%YaUZnZetJ1qp%(-*FeZ=DD@C8D1X!0guPnXJ7PgP<eT%Xo_iUdVF-Q3y
z<h`KtFITYV5uwL1#CWtN<(&yFQ}JM-?`^jddh%r0zugOI(}y%^7Zq3f1{^M#alO_|
z#eR|hk=E#C5A_)8az~C=q{bC35smbH=a`Q<%Do6T5%G+wku<JYF22mGs!|ijDsWNp
z{lv}nd+M}hmo!wwCXGm!U31K5;x``8qW|j)Q=}W{Kiti;;wj~_sL*8Ke+vamWhhJM
z-kZt4Wt06%L5;GuNykSR^+?9LJMM7B_xSqHSlei(O&f+XnO}>cP-j=S`n)g%K4%sV
zHpQ<#4;J+eujb4~P)f*rPk*jJav`&!&yO-l=p#{H<g`fWmMnTU<q_NEUuX>F6bSV9
zDu5<)`B_Z!BYjREAh0q~{fOjZ-+29-e{Bs}isx`?tUOM*ZMOAwUxnH=c~U8}9P6*g
zg(x_Jp<cKReGy771{?z0m*+wW1*&&<Ek=gbrcGj4LAY|&tzfeWe-mFz`%W~T#hb*M
zH~^|=h(NcoSa{f40N+09R=6iN@)?Zj$hx3+%;O;lc+{HpGG^35E-9jtn4f2#Et>xS
z0H;7$zvDM2i)`TC?{k5J!Gnnabc@ypy8}h|C_$i5Vw2MJP5hlP7)j9!K94T_8-<_p
zxDo@1)^~J#On)+NCoQeOPJ?K{0RR9%07*naRN~6QO&_%1tVR}F--Qu$x&ryMMf$SH
z*w;hq-|Z7QY5Iv583Bw5NDDgG2lAcrr1|Xoh3{)sw?aPZe3PaaV0d*BiTQTshG>zS
zU*}p)a&<<%|JY%>s4@0iS86=1p440VZDBsLfX44z(Qdn48U;1#v+AM3)xELUEG!vL
zPZv8Y=ND%xrz@x)wR3qMQNm(upg$4ncF<J66DRc%oxMr)rq)r9RL8Hh3*mSQBDP#M
z>V3yVBLU!@(`Z7V@23Ve<P9&#-Up05@R?d@fd2IRL4>vXQ*^J6_ZAjs?E-mg4D*Pi
zPwN_bG!FQs43VF-JL>L{Z7xJe@yx<8+H7d}V0VvtyJyJ)PWey(pE2;{!gMWttKssq
zDuMm7V12^2yG16Hk6a@Rnts4c%EH?n($?;oJZQ$YZt{#TE-Xfq!_tR2M&{%lxNwA@
zls^9s>6eEF8{(RoSr9+QFLH{L+~MPlKm+FC{Zr;xIkkZMe9GYH7Dgl2Z`dpItJ`I+
z(TABOIY#iX7>dSZmzlSl$l)v-z2ArLYPx+&-Mv5g$xoCMfC@W9J1X?^PL;Xt3!ZU)
zOVp2j>=U&OJuL}UJd1Kr@$wu7PdxebKli?CzrI8!s@WwH!Z*Kpjh!H;VF*_}+RlMD
z0KM?S3+!YzG`{eqE6kis5QcD;OvwJ)uf9zt^W|7C0e0Jb=5v>u4}Ik0=1X6>%6$I^
zx7ghf+Zz+H2>y&SPL~z-cgaqKea%iaVF7k<{4am-+vZd;BKQhx*RDHdviEB9otwTd
ztM*TjL4ccpXP3h{XP;#-KHPD~E=NTE63_nq`^~-gJz$O(BNLP3K46vm$pa6XS6=!O
z%b)M(zi$${ci(fbOycip4$Nshkrn@ZLk|Xk!%}(gyZ<M)Q6OIU@ZsRWgHrHLwL2>~
zE^Bgf(q3c_|Abed0XtWAC~(O=J4bxL^0-&RVLT(84-BmC|M3In)vtV+_4ob!cjU24
zb~!xoz(Y-EWclmA_BM0j1<#NMzi2?&N%7?`UnynaN^|?JyX<a|t+Hba0FND4Op^cL
z2R{_K$Z>9x4mSZcU?<{fAZJ7Y(mC?!YQ*efIy}Sbc#gkvzZ-A*p1JwvTP6SQx0CoW
zV-bcWc0OV7i2()^3^5_xC}rrr`+pod+i&vg^*{4!^D{sFN|9|bP-5566<__j{jsBl
z-BOyj7$(m@|7qrCDeEyK$+Aom+x1ly#aLwl1IO?E&;OI`@;bKZ$)Jc`BGkbj-M-7#
z4b69qs84(9ldRm>wGx9<#aQ+`zx8W&SCH?|0QByA?z82U!P4U%cf7&iGcV;E!-x+D
z4jeRBU42b+fQH|s?35w=7HNnudd1+n>#i3g>5M%GnEW4=di#v0Kh^HwLM}05;vuO+
z{Mh9+Cp%~`7=PzGH;Tc?8K9W3oRwo8UU}(@%`d$1=VS*@F)2rwYp%K0Ty@pe=9XJ-
zmE&s$Z2c#%_)%w`@wBI!o4$9mbn>{!uBe}s!v`LG&}<ps5;aa95Px<<J@LFJ#FA>=
zDBl>i?z^{KXYmYmE;@wHX4qM!4l|k34J?UFxzEf?R+9{A#o)uF1|52Eshf2HV@<3B
z${-;%;q22UUVi{U)O1397T{4=1Z!B_h=X4YY`OC4s^ZYekFA(hf1d1o2qzgBP#y?Y
z1mgx@{s0fY_|-|fALkMCF6D!H3BO(PdGAE8`QeU~cP1QSMvXe<s$Cj4ZLAI?2@~oF
z+`@htZfGs^MoLe>dpKco$R68TOb}A@IiH;}9D+`EmfQu6fn>O%3H~hKo+{BvnQ#!S
z^cV;eZY2=bTct%E#ZUDOy!bjs@lsE8a;SsA)d`VQICT$oB=LvGpAbiwKk&_wMmh<q
z^z>=VV=y6#A9<yd4~koQO1?2UMVvmc*vX#^#}6IL&VIj0z!&XAYJ$P1dBnW0vTmbH
zgz`sLv>ol*sBUFS9<o~3Z*!r=#pLv#d?0_RFFwF8yv6D%x+g5ZHmuM=7f@EHbCgT;
zr9V#6IJmFyP-IMrKZZ#R4t_broyk<c91;ipl2Fgkhl&|`!MA!r!uX&%5Z+k~L%!ff
ze@TI`ECBK40Zu44?iXa$w+~F@!jlj19ph1PNRuDGy0_JxHs56tL6AmtOSrJo_G_r`
zN?e}w4d*5QREK+BUB6<q)AX}|2E&E|dCcl~ACN2Y`+4pA`|lX_ViVMq-G0?Ajh}KF
zYb#QHVsK;ihPz98!fQjXh2w0bO&%W!eKvGuG(Sa;sF70Tzmg~Aj<~e$D=q4GYN8(e
zu)c#I9xu(pFzT=0S0X-iajb7e8yl7*%0r0{$wc%)C>!|anDP{WAL>VN35S6Jor+vp
zco>cy(NCS0OLz?HGx0GuWRI)lqZc6ZL}tiNfpX56($~=x{ILBYFBW9sLUgCH_}LR4
zJi5?Ff0ptCM<EpULy=>c$H<!9Q(;+ibTJaTUFh)5{06!*>@UDGi!EqRV?f`WcE$(v
zkTpPhj_~LDctWrN@J0KUqOEvq_VoH|PL~zDcf!+>oTSuC7#v{Hnahu@a#GDES>V>t
zw2i0iZnGz>IIdmXj)&34LV6Z+=#gtaZ(O6oaoUa^J_YZL+x%~Q0pfCg?%Um)<HEV!
zXS^459G3egzKzW>f|OU$Um*-}G`GZE+&5EAmxgiamOfJtPh4U#J#+D#g0*Q=c}`#B
zB|pCMl`HKW;Qcbs%e)?Ql_#HkqP?<0@t7;T`s#0pUEyoy?z``4=A-&|^VIsiSvHCk
z;LAwH`1|BUFM+rJ#;-L|s*-Mcb^k);Pj%u@?4eR5CRS*Mzy#1MUv+6o;0W%Emw%<H
z=n(M#^Uj}`zxcC1Zcd&Ojw6=t*mYNlBn{Hr-}>gtaa~N<e&|CVl~v=1?aJ|8ciwIO
z*FXMUQLx2C{Bxgsu{{#*hd;c<(qypwqaWRF_Fn&88_ys7r~g=riw|~eU_f|8cGxia
zKOx4dXI}X9%EVT*F_7@MGmz1S7sCU_r5j~Z`fG9w)}Q>*?^bqYJmVQpm%CAhWkUHO
z^Pvx4CM*A)q2=_`Pcv_p`%2Vc27l0f?Q33PzW4nf+T*8iBVFJA&Ua;J!mZ|hF(|$5
zZNG?f34Zo<ud?ZSRMLg)*zNG-CqJpu$Sej8263HtZg~>>z<~qiUGI92x&MJZ&B4oc
z*WYOV_z!=_aO9l=$7a3!<u5g#``j03`|Jc6lL&wMU;bKA%&)xd&E~lmKdbr83+egG
zzy80IyJ2p=<u=3ap8x#6{Xu0nN{LX9KG>nc(P94NCWfTHd-wayZ9lrh#&UzmAHxVc
zU=+4*-)?^K7vE(5+kbzjl#B8png+BEV7z1}PiWM_nDt>f0*#Im@Zduao8S74x65GI
z)k%2$+1I|(?2`Hv9y3T8`uW$twz4w`gY0|X_d#>Rjo-6*z|o6u5WE*&@|;RJN%2no
z<@e2RT_^SXk6IcQ>5S1yo|!sm{_M~HN;;Uc=DBiI+AqK5KUB#7lkB)5&n-w<;w~$6
zlI97Wp(j1@3FdWTjMKUYE(-p#%RX)j9bn8=QV-w!rZ?Cl5*<M?Cd<)l=$Uu_{XdFR
zVLZ53j5GI$QTN-TGyd#<|3m4}myM_wU-V4t#<7n-_`%J#156oUhoKK)Gm0M_sF==W
z7hO1Tpo7P+&LG#1mv_Q2koH&Xgg6W~kSCpF;y)zyl+K@6Bl`x_0aov_vWmemD-5ao
zl9eSWb<jN@zccSSujmy$<*#N{VzV;nsnl8FlgS^>*!KZLz?78jaM6U1li#xhLJ7n!
z8}77UhK^tWKPLmw5bQbO;z;?lolhz2+I}c5^^afJ@%JYu6qhINI$7ttV*rWOTl$jz
zk%xYmmA*rp359TS#HS0t=o{Pah>U1+{9S=QZm-_y4>IXfT%^u~WuW9FvY5Q!0X^3k
zH};Ap+m=Kpj75eaSc8r*5KcxS=WvoZ44c2x5(tegis|Z~e)aar1XMWLPYjjGDsiKK
zLf(gbLL)Z$lj566!LZ-L@o?z2aN>%x&TmO3uR@@_(LVb?J)#W`!MaI3_W`<;`PlJa
zDc+fE_tQ$>XtQzcJ+jf`fc$vCVImEEu6i@nv*5?-JAEKGF;GpvA4jL}jMYO~Lj#=|
z0{H^J%Deir5FsXO_3z|q^P(|Up%3iK7&?@hSlP2WM~sr~`lUL#67rPF=MZnzzNI|J
z7LL#a_jCf>OORcTD)e1l3M!O`x1L_tZGMXQ!m%21LkCpqx5%7RZNeRi=p?K2f<6I;
zfhjqTmq}C|=f=j1)RPd<*~lza-h-%`!Qt5*3+J4yE)e>ImX;gu*i#LfA?m#h1ecC$
zs4C~`)8$hL(%0ey85SO(Pf-?OUWDHoTdZxwkG@Y#f6~tlfws*D@}SeYY<C~|!jX4$
zb!b2$o)}-aQ>J40R_FV)L6&cHaEQ-XICf7YK{C{pKqzd<OW6gpsbcJ@M{k-wh;2LN
zuB^r}ZIoM%a&r*V>B@r(6Yk(SD7%`s4EysK1%7!fh`Ud}{0%>4+GT%DVXSRezfPau
z|NajQzpsD&o94HE>o@Ffosh<lfBX~kv5);*)A_Eub{l@2IL7r10LGtpzVltyy@}Z7
zrI+alqbe`K3w^u%wjC|a+kghjUUXFjVgp4cq#rNG75Q-cuHBUm;(pl;@IA>K1scHn
z?z_Lzk-tEUAm={eY(=wAl=Npm`*}Ngt{p^nd3;w^<oiHF_mY>qu+ee$eT?08z^;H0
z;D2)OKJ$P6?f+{&a@oI`e|Yad*_Hl29tC2z+|?5U<P=l>Og^)6hMhnmjB;O%1bqKL
zeMBae?`wKvQM~-}FE_vS1s<Oh8b?GX9M|!^TYe;yJk`!SVIF?uQS+{MzsJh%!(veS
zQNu`~bh*pt?t6>)*_HD0m%dQFn@@H|V7y}Wrvi7leDdEvr=QIymFU=GkFnq3L_rg&
zi<iBUN2wh+U64x#VRs2evT!o%`tRJ-H1ZG_g93N4s3l(so_+RNVt_bLKbub}nzi7P
z=U-eho)L)h_h~u8&mTPg;O1K@ySy+qyz&(<sl;=-?6wN)6iDB^=DMblhi`T%UG}k0
zS>qw!odA~c+iu@gD+scnL%|V|ttRQ3#H5bV5qeaPj(PNe>|{GQZH>q_pBjTl#brmI
z8hLOdF7(GI|NZl_>r2k#_28^CPq(|UJSY1~nEm_riPgP&?-7PL^8BCw`9tR8AODm+
z0`5y+xvE40Bi~(jl?~V+pjTr9*3`r!M@n`Jtdh0&R7CW+Pm{DV@Z7d_%pG4Q9f*Dc
z`kM<SAKgD~CM7>P9wZ!`FffflH~|&IsqNNCCWC`;f&&^U!+vbwMF%@IV5rScMtn?3
zfHD)dF$kcWepzhd2wE5(=_5QM+w~{Wn^NCm6L+!NhF*xlFb3L)pZ3ebFm#MDeDno`
zancu?ybR-`Z-`-=Bb)3|NRc#zVOH|Z<TA%fs-A<-5LQLcH6o2I(xUHb+oURE(FF-i
zFr+5nLK6AADuGn{MLzTkVyn365BYsbjT=5<6JoX>-$8wejfZH`!wE5sH!2s2FUC8%
zBqlHT4*L<%@3ap!!$h8!AH>H*C&#c+pRJ+O^S5b`Z`|q6JSw`D{;3+-HH|#OV7Q@6
zWBL@ol|VRw2d$(o7RI9a&hDcQ^1T(m<Rxt$z=U>~hwv8bbJ0e|`e$KW(J~}RVLV~D
zRsELavtZ3%n5OYFzKQiM)xX;(Z48s&$j42BuQQn>5~gLP;vK6*6{P+-4B3S))hW5q
zu&o>{G!|y+H&5D7#vK)qw6WmMALm*FK!6x4)7|P33{<5BYQh6M5iwjdPP50OcOmn1
zee8l*kpA!!*>i`W_O<N2a)FRyZ3A>t{h3sMDV*<t2D;1YR%wIiixPh)yUwWhu|6{G
zWsnS_v%6RAOYnzwI+mW8UL|O(Y@(Y|c^5>NKcf<$eACY9&mnEvz+6t*&H1?caCniq
z)p*@-N1+jR*&fq|fo5z_ck|#bJJ`hFQiS(;a61SslKta&D2zNBXN))5iQ~=rIvOrM
zrRR=~E8Gfvr3(#rA1rkQ(5KSq_Q{+uh8i}Kguo^ePMGq69j<4c<!lwqSHA0A|Im?<
z2U=P;q!;RwuJK;TD1}^YRZ^H8xW_vB8XUzR%ZylupG(zsSAmA-xG`6#)x|v@lw;C3
z{t3(4M?d<n=3X^2crtj=3!YmJBwwtuXJ-oFcZRrYBZL!AJkf6H3f&1u=EmU5GEw{R
z!;e}o((M~Wxa1y#441j5AUd!SWG1mwm)djORk-s4`nO1$nAi{D8RtKxxyr-$V@D1<
zGh)E~LsSNU{NBBLcN2}u@xBjy*xt{hj+Z>|;!0i;?q<1Dh=wv83}9^Fcr72OOx)Q6
z*bcgvw?1As?=eGNH*{Ks!60&yC<z}(`?s#W!FRIn?8pk03=Qp1J?p|IuGS+V(H1b&
zgu9&x#EvG84D*3}x%IX?d?)*Up47eY2sQ4Ds%+yIhOW8xdYi945y}#~bGm_9F%)$I
zbWUT&<0K7{AuB)W*s=Q!16By%|NhOPll?~T-1eiYA?lLnKPTFm+9U?QL$Zr22E>6;
zmPu1|<*m2g-b9<0oj#wCJEPbY6oRk2d_OHCz8mkc0bz)f{1#a;Xuyz?nuPVkQcYQ@
z7M^wQ<Mi?!xzV|`1Ed;tqnB|*OVtwR=Eoq_KBfi?96^&IOQ5a<+@yIGJ5}z$?4(~-
zqAc|bV={HEPCL{LtCghbz-q#HeM#vy`om$r!6UcgSJJ~Iti2CI+QfE1`^*lXZF2t(
z$0qd~NIQBiHjwXk+&=x(qMKCS2-^{mslxt?EE4Zg;?$+=nDp%A*OZpT7i-_>Z$z_K
z0>;`m`thBxp3-5dvZfQzO}^!Y0d;JGpQC1DKzd@jEk+|2Ufse;F?i-w0(L;9_}8$i
z>(H~Y9ZF5SRd|W5`X+B<`3pO-J}CMzmNy-<FNMcn&=nmoO9qBV9T5-uCT6s%@Q^5=
zsdBF2JE>Pm<AV~5JmH_8y&p}ERpR)%5Lo@+uT}J#e}4YLPloO!5H-*UsZ{?bp{}C3
zt;vxTS{>)M)j7>T-M!fHR}d8A5{Bzmi;jXI>*BL#V@jzXENBXWRx_3#xIHR&J7VPL
z#EO_qLt6ef;bUzVZn2hH>$FdU_f#J*wwoM$<Q0Q%^jwV2W77@gs&dsQlHlF_xHE@$
zXPj}$m^*KZSqVR!I5z+z%B1XGBdk90;vO6aj+Vs}L#=~lyYuIO?BH?qC9KTZy7014
z&bouCaXKE=-yQGk+`w{i72mk|VVzM5f5!{HdS{OsN%V<d2!1;i`o+GP-+s?~-Y14t
zm4U;;Ch?AUysdPl!%JWKB73hk{nHQ0>0tgfFEQGooAO3|>a0iXQ$jXz=zdvuoSt=6
zrc)?Z2?PV-n6)_8aeTog&#}gbFJEz$7y_nkXOXZs+;F44XCoE{KDhLr%gjSy8y0|w
zd`0_^9A6cKGtWHTJSg{U)B$!DaV(k-7+^m2>Cc;o<San$7vh+vR>0tZfqXbI%OrMD
z+{+YhzvE6xOLfpfTo1`nZd<na=ldCR+_A@+agk3r;a{d&d@*5v$L)7CEs*%IE9pyL
z`idB}9x>NkcT>qoL;i3SRtOlK_RG#6*XWdeN%#GdH;+H(tdhrha(~IEKmA#|x`0M0
z)>xGStfLR^UkS@OJ6#^yQ!Q)kq?+8jzqx}+lX3cKr&+^FZ1soY#Gc&ebixVcdvU<?
zcYXQc9(rhx|7qX3cj>fKPHr0ba6J30Gb(u8H}qg6JcxJ5JwKF9-9BCkyp|49p~#=7
zJo(&8J|XW19(W+i9Xf}5uFik@Q%e5Gd}7BAb6~GCB&(-8T1>ynPc?k%O7e%KK7IPL
zUle2XUh{(=-YR!=P20<>%b_^>Qnu_z26|z6({O?MRp*YLtSs8QaLDeU;%JdnXHbcv
z!?tZ1mDT^nWyk8*^%Er+wr}L4I8(;#U|aGOImrLkgP^XNC<oMzb*Qryi}R&pD~eMS
zN2xqidij|p;7cH-Tf6C()f$*dy|R<1jTL*;Gumd#ZL5yA&ol3_xWax}7^YpCLi~=u
zg#jjFnLy;&6L#*T>b26QZcoTXru%|>xYWOA&xV}2Bo{e59q061Y}IJb;`IrkQzhSH
zz>#(_7>?j;nW-xC7=`Q>N<iD3&>gw9f=(A-?Ad$JW)bx}R$nQD9CZ_e6n+dh<(8`N
zsr1L_z(-6jw7;n%X=C`Y^u*%B4Zg!}N(wjWPJI_f*x}3K57vK>5AYT<x+UqXz=ODA
z?F)Q_7m{l<v3?Es{9mb1psoE*b`Yi(9?i-`*0^Lqn2#8V?9PT3nUgA8AuUa>KD(`3
z8ot$&Qwv6h$8CCIZ7dT;>=uh2t<;KeO}|AE3P*4O#P5ICB(Y{_sr<C%am~0&ADe|S
zQp>o{UP}MO%7Y({bb>>8^oUd;0jEb|M%+@^4n9%_st!m*QB!k#=*GT*kcN6}Oxfu|
zyqNw(4@S!+?kh-CN>-L8e`<;>RW9fc#~`(6tFL(!W8|oFlY6hk_);9L#*UkL!)`!!
z_OP3VZ&#Ku_{=vJow>|DB)fa)i*O_w@8Dr9(I4X4A!8Q|I=H!b2JFrmzUg}~f8}@>
zZWxqgr=IY1SWX%kl6mQwgEQvf2s^;!^cvX!G0vDt=wkRG{0_O#6Brg8zW_R(!vd+_
z`m{l#e>y16(eR-oa@w`)PB~rYlje<Ye1oK?;Nhf`PBd?Q>o1vizxzGryz|aAuYK)n
zZ169C`3m!`Z+*K&3Oa<<(C9__Ex3oenkewMebi=%5;^;UL9f_3RrL>D`eNvD85TxW
ze8;-HBp5-ESqhVr`|OcO_BgVJBTDkpm%rFt^y~}Gb=TcsZo28acH)`I<5uyqLO-Tp
zIX29Kjkt)N-6SYUKXbv`xpRlZukrVg9Oo2H;@>3(5q@=nG%?9A7Tn!HL5ZaYx+x=-
z^V?%M0?Y>{@VUFEY3PxRVBo-@Crm%#*hw@ubI2A(rE||c_lVJe3#XQ3wSRRd&M7CK
zq^Qm3-o0)olM{U*dk~O4JE1~D&$eyb?43Zmo2W_b2zWX3jMFMya^%-cVs}T(2)KEC
zv)tb^EOK6KDhXGgs6A;Q?pR)Ov|tR_JrV=%pHW~Z&#`hYWe6DWxOXT9_+cnaU33-;
z)5?uY204S9j<BbPTLr23p;R$c(4nS-a?V**gD{3icJ8DA17`?W{%H{QPfk{Ni>>%Z
zhhcc84(WiE-G&?ur;wUFV{%c(thJ$Nc4`NSBTBX9C?PJA#Ms1zkW4bB@&diZgn+%k
zv0UZj^xL?W+6?(%0E{68{goOBwISIv@30Y9!l}-&`iIYyEJM8foh5(-Qsu5y|A=sP
z`(@zLS!3{r-lX2ACi^XJ%|B6{WOYPZQKSa4HKVMJ#^NNq8dZ+1q@8GI*hU{y&--!j
z4EmsGuh#y^GjwnW@R(|=Lfqzej8M}vzqFEoU8KgiIe=K7&quL?P*DsiouRns4se@b
z^QxL*4CFh>ch|N-$D`ehr268sQ1z|nl9CC1$k>ExOR9BQCwW~m@A{{1aR~j>uKNRZ
zd2N&6YOXT@+IhbBf*o|R$^&{_7@h*!x3G<&KT12+YNdRtZndj3UDvHt0z7xhu-Nx4
zXt5$b#wpAfa37Sqd%=403SBJ&95cfTWHtKqB!WH@={7(6%$Ag6m1;RAf5%$cC;pWn
zV>U>E8I_Ygs56xN6keQa;YzqVWviY>FX+~NE44#jpl;LSk*K~(S~<R$JfLmp#h5ad
z%l(p7oqVb$y6TcDo1L`vP!08OHKy#+P6F;|wE~Mgo|ieJO^x+d<1+uUj}vf)Wzo&h
zh}k3epvz^ZHh>?N_et5Eg8`>79JrE*n-BU;-Q%b8(a<OCsL+~ToHHB}HzebprV+wg
z7@#@Oa4SlBxEqYSg4BTI$6bC8YfnM1j92|?*ItJq6#vRAuQF$xak_c-vwy0Da{l>G
zGjDmzo8*dyr`yx2xIX5SpZs(QhYzKr9_U^*_NW^;$kpQ3P-IQ{DgpZpWm9<e#n98P
z(@w#R0m{T5xKGLI_3P!*^yka5Um;9POqiFx^u^|dFT6x9Y`@c7D@Q`@+O<nMIux}k
zoN)ZHRY!3-^UO14x6fwVaLvdBv~GtTl@4P}xs8sFMd`#GnHY}ktcl^aa${9aOrfSG
zS3*OJPx}CsJ_1|hP7@1KR#=h8L_(^A0gWk95FE!87UBqL*-}0x7s{6v2Y6FcjU&UP
zK*Wl~{s^xaw|2-biM_Jh#r8I10b@+8u^W^A+qO+qxEK*R|MY7vjc#>4K2l~!#K?o9
z@PRyt+&Al^M0fLmaKZUcHD^SQDciAgdnvqrurp2pgA+%>g@BPLwc}`B%5v&F^^i`T
zZ+6Xb-Y=an43ixV)TOLr7#NqjuH~4H+=Lt%8N$-iV&xtlABTIwVsOmPomQ!bD`GT0
zKg`U`jBFE~vjerXj`Twm?+jkzwWz@H-%_1rAE%>tsfQTc=H?H(qa;L!#CEMf6PZsP
zF6;V^;_{HS`ZI``kd-2_{6{vN8H@pqHZ-NzJhDA`O|{pqgQJ|ckIHWo+hb>p$GQYk
z<*rx#GTL43BZP5LRw>`;c<LQ>t?uABq*W&$Q+l#a*luu$uP|)WcF0A&ksT_~Chhj(
z-1=3$Elb0)Vs5MGIVRLY@a2wwO8<u8R`LxOv3@d1?zr#ULpiM^O-Eu_w|cs2>1kK@
z*j-D#qracp&B9KJ{c=pJf8SrV^f{?fKT=1aMdfPqe0^j*eWFg<TJ#%osKb37oS1x=
zl*M=v=A+8l9_1z<=%85NCq{FnZZSQr{M4o`0wK(bK!or2<roi-CUkKs9?}AhsfF~>
z(5qdyQoQ=^7<2TF70Rr(yR{0VPpFG`KmRfE9BI&Pe)c<^zFhsdRy?+AlyUTXIPSFd
zR>nr?cy?v6XeZU5@p)Rycb9)v4m#agQ>9RLLFdxfiX8z@Tfulz?;)(ac#ql=>o2Wb
z8J2nP^;eW<i6oU5srWmYqe#(MOLx*6&cBf+nzGf?g~aCcE<1LBEixD9!?sPuf`{XF
zY!NSQY<+#rMg~8W;u!O&^!Ikk5!4!whQ~NGHV9t9gQKu)j*&U!eKLpT!c_g;E+I?~
zp&t~Sn=?eoeN-~v#+;42rw4xk93#f5S+k?Lpu;7A9B5}m)H$zE2U?=uAMiq)4}S0?
zcG31Ja-xwB&wJi;>?f!Fy!X8yNX~ouVR)wu5fAcLzKJtdzQVBaZzr1~zZ}+rSu|qS
zmtS4*3xv{4hV_xFNae!i{2>{3vd6@baOD+O+9TH<_qY?Pgv5mj^z+Vrg8e-p_w;bA
z*Y4eSS80i{AqMB3`*_2zE)0uxzb?!r?(L}?Uh!y5_=OE{sBjcNwVML})LlX&vgHrh
zylG?9sAN4Ge-6o(a4rO8B7q6kcCaaNga?!Kt$^dpV&HaoNr`D&4r8I$#*INp?DE!e
zF*08IvKQ56id{6Z-5~7<jEy>reu($1pSmD)YWc0@1SHZF-p_-3>Ru=7Cx(oa5eWZo
zfgLbhR7i&l!%h2ADGMH?Q@cf0I{5mC4o%E>MW?P6->EcX<mBj|)Fg_ZUX0Xq>SO1L
zE1t9sQDBEmt8mdkD|JN35q_9_;MgJt&|4zA0+BHjkqeU66@!o<|6)cSWX1X9TmY(a
zW81MmIk4jRTM*DVLlg{?l&1_?0xOn)ujm@pNjtPv+CFl`H{uP+uwS-W+Vj}L6DFTn
z0MeWg7J9aiP)i#VPn9R<0%LJcUw5ZMugG0x5=JlI+@rSlfb6ywqc$=Lf%cJ3e6I$#
zOgv1;;42)2&}LZuPk*Yx=jf*R@sJ^pEBz)qS3Xowjv_mt9^E&+Qt~-`O$p3P-yt^H
zlo})WWT|&d8nbXK9AkiYSaz{B>vSVvqnP<*>^ppLdG(P9^b@!<rz0pA`;LOPL#D67
zL}RGqm<%{5cZbbMpI7TU_p`J$^rSMa9V3B1WWr=)*vDwNnm~HByrKi=g@;xE&y2xB
zd9q@EYCsG<3HhwJZb|*=q;9oJgZkO@%EGF(0Q80CzwYQ+3qHDyld)pAy0Y6i>x4`7
zf4B>^)sC}vwB0zDh0A43=9K6+*wZ5#Wq*qt+oW}>jJ<xo*tKTMV2~ctM|;<kv6awD
z8F^?cITlg#i+)+GJjEo2m}C8=O6J4@uc2Ew27ndPKe>OpDTg&=Z`GoX^{43L2Su^(
z{BX|=N1$;8o4c2%h;qNQHxnD(@wwd9bLil#Ic~cedm0tpay($R%V|yulhUVvCgYV2
zl?4+nh3vSQn_n{9WS$t{8;Tky1}1k05BT`-bYa~%$i0hW<*vagnHwCJd1wqd+oeCl
zQFLV(u(aWES`1ve%Ltr0IS`JU%Cw8G7#{QHfB*OIHGlFae`q&`6xi<VV;}pZoXpgD
zzc=JNR=&cxFvx5EJ3p0kh$;3FduSfdn!u_e8Vr;i6PK+ZY*0enm~|LMqD2sVr#wgO
zD*E@m?*ryFuX&Yu@{^w=?&aMKK5dLazw#?@HJ4xh6?65~U-!M_sHph4(d;TwgF~wo
zi15sBdb<AL)lW}}nNB~15R<q=qvgqjl$=w>9HLE4Ru5scp}~>~g7(l^$_9gNfJ(n9
zZ)7WuA=MPG><-dKR%xyGC*?MF*&<~xwaZG=M9pyO(d~kuliw+%%KyH7`;%oN7KeYN
zn>*jPP~3{&s-s0GZK=aaZpvj$b{b#vDPrNM9O`N6ocJcLykOJ}>kyKS8LyUgtDjH6
zdry_sjj~df0eNhdfET>^O6R-PfHgIVi;PwSh=)NdE5Df7O&(1of~1{*D+UiC#M*!m
zr|vhC$+1a}SveE-P^S|&btA2Tm&&WK%+(;3QL+T05{UK7VuLaQw<~wCI+^O1h0L&<
z<bVtmQkSrX__6PD571$><2gpB7jAZSxSjYOw8moYBZ~q35C;7W4#L4cy&CxMRbK_;
zT5J%4F?-dG{GKpqGeaQVy-a&ZKYy1cz;4<Y=uuyF${_u=*d(9Zff`V^na-bbzr*Sf
zoMMF{AhA}TN&VPjY?<V>kLir{PvJw4D^Bv%R}azB)oho#PKB3<`zVhvS^k0@jFD}P
zbi&`ZaXDQ=c18MtuBShnszWLF+I350f^n{P-SU}Rhwqxzu(o=u*n1}kJr+BpSUW$%
zfIa$7>7`hk-$|4)Jwi3LMxI2w(TaYZwxm_~mHH7shUM6p54TqH4|U^4AB1{>ux>F^
zdiP%)`Y{zromy@ES%+7}iH&8g9*{q&a;stUvzHQZrDtrkn5T8Sc_=^*CS=ZuJ99Wr
z4MWiE$U@T?G%h=T+-$4E)g#;NQD8#Sg&PVCK(c6e(~#`sad-BN$>RRmr3JH53^8^G
zj@O(X5M#}Vy?Y1x1y3$E_xFIDoV8U>0xEXpI9!Z8fOGe23GeQnAa}11_X?Fw7nD!e
zE_p|wS!%ue!@cF#U;iERf)_l$LiBa7dyTpM_B)arDJlWP!Po07m}vZOdkiYaq;ii^
z6`Y)W_j%dT!wpGgY55rbyW|&J$YBg&wu5$CDE$N~t>&bE!+{mwI(f+vXdnLYW#)-b
zJkLDq+0Qa(oq4+5$)jQYCp&yz{_>Z|Jv9%RTW-Cr>6J2maHbv?pabj(NDb(>%1TO(
z99k=2(sE1;JvGGN(ipa3gI6oP#L#fk;f>{uhgN#6!wUgYyB%~w-BwWf6E+y+WejFy
zVuMMc*u+o_e?7QioI%DR{P-sinE&#Ze-k>FzL%He60-rh{BkfZj}mg?V^#)2At3Mn
z@ptbryYId?^s4pE5glQ@C7|YOj0kl)HlY~9b+lyFfoutlb?Q=KFh<k4Z3XZU8+a{A
z9?gjLK0IVERc|zoR9QVNTUjw|vHIH|m}<E2X?L;tZCrrBfH4LPQsku8m{-r-$h&P@
z#+q_-6F{xwJj}6^qPHq56?!;YNF7nsZb3rN1d||y(Jlt$)k;?A#b{=?$Djlf{jyZK
z>!x284x0UbnJ;nl+e(Z)(27ndi#vKJ(sbJBTP^&t2wnW)cr<_K%*@;d(X7?ByGdH~
z^&#n#px^!evF)L>(B;kWHD8i?P8I>J8A$plZ`HPxU7K?EMM4Lz8NWwq#PlzzgJBt=
zJmR(@Ltvz>rfd`EV6dZ~JZD_c!(!j_9D{F?<kwXq;X>o}0`L}F+=PzQg^50|$b2;b
zxzXnz491b@XvU7Q3Bi8w#i<5b*5Qc>*tTHt=A#q1-0PS!==7v+N!(G0)U7DqI&(}V
zXPvPbWrscu#-4CY*%MtEjj^6%s;#u+=`IX1#A?U6vNJFQl|kr!q<nLoL70E=&hETk
zjQdo#gz|{d4%?A1O!#f*OoAN~OunzC{tS6)_ie`kg>9jUPQPzXj@inv8WKS3x|Y7Y
z+eLd^+6xdvjOYstJ>c*3u3HGXwDS-)Umb0y*OiXLc18u?ZP^-ml$O7{W~&%$_C-&!
zT5cZ0hLHyVZ(C#Mk<u!j8eh@Uu!9F9kpuh-dHD0z{(C`h<!@{>dG2$o*T_Km)KgC}
zZ+`O|&HLZ~&y{)XIx*W`vnIJtfJQlRs2Wh81p$maK(z%ZaGGy)-c@+oFNsZBy8tr4
z3zS%ski4WOdTk?D<f$}JW!9LK{^1XAG4K4lzc+vJSAT21e8txc7rTeX0peoP^_8!@
z)HZS&_w;mpTe*MdNOiE<2rcqk>p=kWsXI9sJ9<jx5W6b}zd11kXdWwWt0%a1j}V(V
z%(;zKP(GnWxd?&X5FwD?++|{qlJLq)h_~Lp+3^t@ppD8_$CQzvmuZ-a>_npz=$HRa
z<D>Dn9e^NUEZ-D`v{tu;s&Dl<E?Y}uWuuDexk$N4>nDdt=PYJ$!jKXxukgT~L@7w6
z`Ou-mwmFJ*5JOt%)^-C{-L@WvwVmE-6$;U`pHOwL;ldF`zK*ndzFnL)FUkfFs%>gg
zw+0&(6egcTAPxrTAyf=bAwmzo$*UN|1}HJSQd=HH$c~R?3B>wk$b-H@Y{IP{{W6t*
z%;1lH?W9kU+7;>R9F5ODrP}wrUnVr!4eIlqjNfAhT@&6-LSKb{b)l-o${}OzAo(0K
zu&t(^i>;8Rea(8VoPYiPVu4ExMmbinUEzuaY~?1!I=1teGTKVdlt}iPrxNk>2X!X4
zV1{~LYfQ5$e8+XC)hQ8`vexw-eOX6O+Fa>@{8(|X{XX(CHAx<p0t}w9IRhq$Tjl+#
z)~%4~b@t6TB?fj?a?E2Ly7e+Y(RUoPxKfC<qtK+R*MdmBvg6!{^0#B7NItHn9;ub4
z1QO%vSRWSyP3`taX^Y%g$nHa|uZ4ds&W;uv(QZb@h2)c4tm$(Rd5`3A1gqyxF`kJe
zb{#=9@-^Z+%1b-PNWEwmWOe*gLWE)8Z&{93W5-945RPzT_f9w-(WBcA%0>e4$LwPA
zBAiQ?Pn&oDo82{TUTjj1fdrQ1=>Gk4{;s!>KZ|C^Ee{{?PssA)E#BwtX*$btkKUYF
zum*2As#?;b5UZ2w*Q+O|g}vh)Z?$u^H{N)Y`ObH4tdPFwqGy>GzVMQYdw(5}Qz5Ip
ziMjS#ty6gmqa&I+!fc~h=!R+Z%4MOMR^1+p1GowY4o>N7<LTl-iSFFF!<>EgS=OCQ
zz3=`Xn-72V6Xs9;^smf^KJsyE@CgE&9XoaycbALvdEkMELKORb+;PV>-BSZoH+V@T
zs2RG?FCq?h_0)kaTgH=UTefgkV-+w2q&jd_Y<Cu_KN%9kQMW^KcHzNF{>qU~9-qT@
zM}mfM&xx&7Ug@Mi^2j66#<CoGeSIy;+2xaydn-(B-8yc@$2V8pd+qSRgYoizqA0L#
zVP48(Y^NOe4yBA&7`9p&OhR5d?zA-sblYLY0P&z~(F}oARbz6WPYNcsZjE{&cgifs
zx5bQNVIVG`YX#I%R_=ztAUjrvLcGlwYq}higp;bom+BDv@rHR}kMD~q8>jO9p!U68
z9MNFn7;}%UlC4A-IK)^5r5#@w8)Fj?Oe(YsOUFNU33I?&{(KBeAl5Il;*f$L(-C$1
zWxhyfEBEBZYE`_hrFQ@L^!@jF86d>!+G_U8=;ej_%C7Eg)BTGXwd@M<h`1t7Rt&8s
z8ZAO9pf5z;h1Xh>20^%R5UAembk&^n17i8c5ud4XK$EuY2tm)a!&Cet=^2wrKl!JR
z5!<=LL?LZXuVE&!lWa7yKr4)!U2|jLwW(KOxSo7VJ&MVRa$ZMH9MKp~47TH|yJXg^
z@It(sesZe5x6?x`Mv8G(r@qtow7c!Q6NnSDg!OmF+r?2ah9a55{DHmODDWfiTTQ6j
zx)ten_p5FN?5(>`j*kH_?ldRbYUaUuNDO5rHhE2c`emyfuiR0*O-_5Mqx0I)!eANY
zNpxknbFxB9vV(hI>)e@K3EOZWOS{n{DSG{Mr)`ZL%k9QZGOy^0XL}X|kN$@}Ufd7U
zL0PUd?)Djp=^y;-v?;Vbw4JLB^kriUCpqP)p$d{L(VigVc6dnwtL@>JgGAhGYt$|m
zPBqdLtojLXxZ^{ds>1PS#W8N+V>|)E<KL`f<B!Af<17p<g~4UHc@!HvfF>Hdb{q-E
z5BK>1gyD!c`U1ypW$qK&K7FtT?EODHhwY{hxsE~7FtX8={UJ$<tK>yEeQUNykO`ob
zUx3>D93LMyzxkWLZtv##@sEFOKJbAL+TT6*+}p&_m!JE&pEc*4bGCjqpRM9e^~0Kh
zTfV3q9FiLD!XWJpJwnHC_;m0xxp!Yk%YI{&VW6DqG}+FeU}C_*7y}#@LjdQT`+#8B
zeS+YA|M!2_yyY!#vJ<WTsv*K%an;w&fBzqUXC8fY(hqJMEPs3XU3cFjJ7TIUtj{>(
z^ybR`5q2Bc0#qVpMy1fEtoAYda_E|0o)%0QP~2*y_VkWF?${*KvBw@$@#Cnf{reAA
zR-{%CJ%{97K3+N$oz=c~-gUQrHlI6p?rh3F93(Jta@_INbnM-`SMIc#s%`XYMz+Q;
zD4|_<?yhiwG4j+?PO7-a9AZ0AVt6sPJMX;P9=Q<uJ^t}kr4-`u9gCY4*n1v+w1Pe{
zu~m$$u`?u%*}i?7-0@Nvcq@SehdRcQAc*Y<Q7=_jh5n?)t&G&Bp}=4O6U!$ISKGF2
zD-auRTHKF6zAzXo(t<Rq_uO;8`;@3s6C;Xm6@HZ^^=>{7$?m%t8XbM*r-`gDqQ|MT
zTpHf(pd?%cOI+;+?KI}$#3}s85<r?_4WeCXOM8}pSQ&x0aL_`T@Y^z<T&#Snf2yq5
z$<!8nQT{kPJ_Pu{;FcPU!27s#jOo;;jt~fG=I<j<0vy-SHAqmuEQh*|6|13Joqn0m
z4@SyxP#UYBgz@`jmWJ@hoicHVr`t|+M9yey)h|oc+iv@1=%7%~kv3n}u{2RneO}NR
zvAR!K{9<4YT5-`>^pGzc=_`h5MK6_&fJJORkA<syiw)%QNg3KSJc}b@i;xI6Dt-7A
z!1sh)tV-XnQhri_jv#mRxYn0dnHZ2_bwE~C#>&Z78GQG8f(T=BY#{@6rBz(?3E&6)
zkSgwBnbvGijg8ble3Az-I4nniXx{k#^p8?;!z`0=?D(_0r2K&TE;zC9mX8!O8n*Ow
zN8^>8<b(8wWH3&p>QAfs9@_~>e5tX!^1@_vY(jdY?09e0U!gz71?OYZ5A8G`HEy1y
z)sH8}i;rJ=mpYO<p55NDmMAauSGj|3wd$7goa!TlWkB&)<LSrEku#c?JQGdZYOvK=
z9F6>pwZ%+Ihw~8?;!;j1gFEC{amsa!4s|b9|6*>IzVH<5Gl0LD@pP(eiaR<9!}%tu
zgFcWSG2ZNMxPTsdTX)I%G<Z+)EPkmlS}sCbz4~Sx8C%dqzR<q3Gguc!I+2;k$%j}y
z$3Gm)QWvb=@B=sAQ!-2GPmCrhJx<<l+DO~u#)V;Y#1+#ODIUV``b~ObKpEl~_7Euh
zwx6E}2feM0v2443WJr5Gf4c)}*N#Iq(7*xr>r{<B_>d_Ba_^2a?yy^j{0Q^d(b5cb
zVSM8b1{Oz477#8k+PLhUK13nF(QO!z6lUyEegjP-lDgabe59gbRMOa-pNo_CB3<4B
z0*mc-^iam3zu0VOT>C>`4l1!eor}Zyb*htC82qch`ZjaI3CEkMscG}>cfVI`6^F#g
zeAv9}UGJ%EG8q=#^Q*u5_U2q+D}I!(SiMz0o*@ms)oAA9Rg{?L+1k(}lA{v)6;$=%
zt+(Fpkttr!JMUck!9-Q5!z;AtB&*e$31+<@-WmavaDx<+cNaeMLUX|d=SxF6C<BKP
zGchq?8&?Jj<SL4siN@)vDf5kQTvIS1Z%nc>5v_`odtH8T^DQMDK2AF6M04VaCt7#-
zr6X!rPe5O4gq<9fJ0@Q7ikF+?k3YT=-lj)zZ7Hkx)^w2Y{%UzEZ)gTU<J0a)DuSGI
z_8C?-nsz0_4l1^wh49EDj~aGpF<=jmR0@>>VQK6*)3{mJJc3lDFkZZ`|ISSwmV3n*
zKm{V}Ni&@l8=s_olf0iIc@x6Tl85t>x3K|ch`ZjtZ@lUIwoHb6J@?{g*|HY8``@RZ
zewumhPrq8~RO>41hw<O{Jp70l9q#j;?E87=o@0(ZW@p9ScR+c)`At7(ch>ptyx+0w
zuF%OI2Y1qm|37>00j$|^-FMy>CI<im5CjvL0|;gklt?j%q@<NZij+iUTed`bwNlxx
z+VYlnRb{!Xl9QZWYnQinEz`=DMaz~g(vnC~%ozj;FoGZn5Dds^^2|$ne%<%~edqSq
z_X}^ndGp@P>6!O+_vv#^pZ?#zef!>XyYFL0=n$T1=Y4hZQR(mS*Cuk%`+^=KP`+vx
zpju!B6@9mMgUH0?i?`ltE0wwr=)wJoPrP<S9sQQ}@ZGR7eDhm(cUA|nudvRUDh9<E
zSn5-r8`V>lZCbE}Z|G#3jN35feBz-C57g3|(?+gM{H1$bTss!r5cN5GbYq{+=~giq
z)ZV707`tjt8^#;nxZ>C{Micu7nKir6$Bj>#>>CD*>3nCc??Ru(te!r|g|!?P(T@z6
zdb`R8T&HY{f!f~h+#GN-0!HAI{;X|WPdJ9UdD!f+!oCRG8v**wrh@gCzImDSvCV7k
zp~0kindBL#_*n1A4Y1A2YB#~$T&MQXm)m*StaAkFs*P)X^D;1Cyr?-S2J<pCXxg~v
zx_}d%8|{&Kp0$^;h`Eab!#E!X@7rp(Z~ILi*wQA(1Rg>#j{82+JI<L*vde5i=CXr(
z2iF<dq?q!)z?=%(%fJI%*~K`@80YOm==jMq)iIL(cEIM(>0W@oVSl8VS6|UytB}7~
z$2nbZw(7wE5APp7c5380%6x+l(I2pWtoxhtIM1B7<Hi`FIgR8w#&Y24`kRis^&R>G
z*Eg=|yx3CfZ^-m*2~%EQRcBw^*WcIIHb!19L<i>{_A#Dev*u9rcdqr^AA)6{lO(Nj
zF18&<FEYDK_p|{_OD8+B>D*~oLSf%-n=fEjc??ufU;8PqZM_GbIUVy4`iai>&V7g3
zcHEA8`2YfYnA31>m*+d*Y2<uo&V;SC<4)ZZgVn)HGSsEuk|K?ddj#ln<5qNW%paL>
zZlx6!J^IHRRWI#3e#?9#X^!>OYoDF!kIU@RTPwlb;SqD*mDe8iW5!GRGUq=1nLoxK
z&JFtV<%jHk!f1rrwRtdx6wPAn>s&uwIWGg&?vHQ+Fs34xFt%zLCprJIlj}2~JSS;e
z-+gX+c?CW8$9#!v!>r>7vVDCt(XITv1sg<gzSi1D+m0Nz>z$1i95c>KjurUP*26YO
zy3%%2;TRFl@}SxcN_|_XbL`VGw{{VydrHQ%N+Z*FUu(ZHAGa6VORTv@@NLLJzcE?i
z%eZ<dU3===o-+r9w+xK8>Z4^@El=O<<3DLO*Pgq9G4o%p71)se3>&T^SJ;Y;G#9tZ
z%J$ak585;c<kc;nGdo`B0HZx|pE-Esm^onI?a7XtEKNG+yK>&S<LlYIur$yGoj%Lf
zZQ8yKn!X?8>w%sFHrfg=d!8?wR04e{L)ts7vmEuHf5<nKCdGVPbM9`*>^a)@qd)rQ
z=Go7_E?pyj>$iTlx!ZQ=(Omw42OenN`qtkVd8W(D>^O(gXDT=uEX-;X<G9B1QkJ}&
zX5Cihy1KNSTjpZb7dYve`1+uW;-_!<Y{T~_%I%C^@Pg+xpZLU$%^kLahLfDFKe!Q3
z18nQbV->^qec$&E)PC#}qx;=4C|uF5_<<46+KXQF!sf2K?oJ<P=V1c_5{>5J;loFd
z4fK&<h0j^rCy<LH<vehE=byi)dH(aRH@R+`-uTAXH~-;3{72)@A8I^e7XVhFa3SUG
zNIt;ys#m?j@=MJ--u{kGJzP)@+jdm@_wDb{$&CB{3CBOZHDY{R5dDKo-ngf2JSaGA
z+hTB2l7dA{0{SkjnW<@>_SB~|&oXDv1NYx2Qi^e|;Vdh|r#}6eL1*Pb3lB92tf=za
z>q(9W<s4TouzGl`GTOz>5anR5hcKVG@zc!@z47~-tDIAT6NzJb$AA0dG+;1U0Z)$Y
z%Wc0yeZxn(ao&C8BOgl_$;x=v9Clm5NhS?Mwv|}Z+%M8d3{P$QG#ngqMve#>!EZh1
z%WJwQ$D#S9+rH9V_pE1l_U45zcy9B8=Rc?6IMf2pu($l|Uo}U|mCgI!|AFQ~Th-*f
zz|&kn!S3Df`9Sk?Pr5c?t$Oy`x%lQE{(<J#-}*brIV!qhwpn}g551vz-D_Ui+;0Qh
zpZ?ijs3^ssyy?@;_kPc7(xB!GUipe|Z*Kk4m(u_#`u_d<8=P9TK-nH!Id{swzvKgN
zk)%mzI)=U^Uu3R2a(Q#_z4xYbnyra=C`BJ6;DN7&r$6l}35yfv{LK@j?f|)R`c!kv
zEuU#_xZ%Ye@fW<{`OSwu^pU~2LLrl!$Jp_C#?znD;bols3&Y!gyMlhr%{4cZwLY@P
zoYv{Xn;lK{3=>7hh%{MuEQ}3T*aVmJ4jbU1e=E>e7!NlSnEWE^>qIjTZ@>|U299Xz
zWZcZlx`%Zmd(0;0wR1Pw)7n-W->B2q%h+>vC4YB~2O`w%$58Z<ztXNjTrjzo5qK~~
ze=2Y2sgp@IH*Abs9UHOBXc`xjqV|W0KYt3yKfH2E*`(*_iBm~;*)FvyJCW>I*#jbX
zsM0uCd)PBn9;uy%K<X=YN^pMQ2&o17XL3B&3XL{yUgkw=V+4Bp=4Fy)jNm+}z4=Vp
z$<r%#-5)wOkYgTpg}vQc&dbn|w#ATgi|ONjUZyz$EVVHPI_I*Vmr)<%8#nskr3W4G
z5;(V*=kZ-U=1jF4)XX{hd&1^GoU8tPr9S30dv@n7UwBBwae#-1EgI#(hwDmOwK%Fe
zQQ%=L^y7wDxncHBBUWY|U(KuCPuduyxuN)pX}iWV=Ie}i4<9?3#vI0CA2fzCPL&Uo
zkUMKwfu|1~FUG!u&G4AB0)MsL_5ogw;nCx#(t{8GhR9O5R{#J&07*naR4=kzBlp|2
zEq}u(&)K0X@ZwJT=AsA}No<Gtf&}{JLw5XmIHU(z7?O^K^%Y)#;C$eCaNMAC?df~?
zIIbO>-^c7ekoy2X&!FE?Kj&z!OnrA$lTXmDiUKM{ML;P*ML=nx_oAX8(p02(=_M4Y
z0RkcdDov$#k=}bEEr68J5=wvop@$^&76=_~{C)Sk=l=DcyxE+yvzeLQnSJINZe<73
z$J{ewJn2cyXuz)WY(x_sFu{GsD)@tqJb0i$;#4@AN=offOI){irWgg!$%!8CgNXWA
zEANM-pI>`tgvxfmM3fzHKYJY18O7D@(QI#?4B(Ba%JSqA44htCFzv_O^z29m9fo*k
zTE{n)zP<$9K=iU|w|)o5a}ft)V?OkEf&Io@^yg;c>p249sE5vww*0q6(Ysf-*ncM>
z4wTeVc9TR;zaHBjdDbt`+VAY#nqBLjttwMI$a6z$$M}M#@{|}vA>>`*M)Nzu%k<=h
zgO3x3N-zgLsQZ+Jd6M5CpgbEK^immgY-8p2SUOh0ds}tS@>ohkS4e8JwsmhK+Z)9h
zu-v;v4v|#@>&*8h*qIL1U!e{F1=?ng`mW8j$+jCl30{fPpnm9C5q1S|;A_!qDL7ZL
z85FWzOmKhCaaV@XrnYl;PlBBCG>Uut!&KK!Uq5um@B_UL?hBIr4v9okkY`;F6htuU
z{WumKs2VE`9)~3ijI_k^oFO}EQ|h%d)4Yj7<9*G`A-wW!N4w%dVg}@kf`61kV2rhU
zUm6Q8^Tn|4?)ViXQPXL}sMkkcxg!>F$spD{XvorU-D--2(IYsl0J}p(fvmgJ8utbw
zuRET=a8H_X6t`uI=u>W)YYOfl_)YO8?epz&6Y`I=ZNb3nnGhgLR1NW2{PNSYu)+Xw
z?@V(GNuru>X*Z+DcyaWAK&h14nz0pkZP~`WoTKMX3{#kKnfQckZ|^->%kM&m4M%=<
zAubHd?22BuuDi^t`VA6jBlfX=4h#<_au04|)mOw|jf^G@-mvd+AG{J{xM~Na-Cb`=
zWP#V6TWaQGw?->wW&&4qv$P$9I!K2D!BvL`2sf7lrC7!R3KtDheYYLCd47f+<`_<T
zj7%QSxLagX82qv*@3W4uyene$7c#5Q($@evT$2f&t`PY%awfCf*V_5m?6u3XytKnW
z{h3il$d|7po|&Yz-PRrW@rDM~rO<n<mz4}rCfmw;y2^F}8awRgvtkJ3=E)59C#t%f
zn{c;XG^dqnlTz}<PLK6<yr9XdkrsT`&Di2tx$v>)kWvxUnC3PP3!KYw<sIFhIsPbA
zBsZln0q{9)P1Gz+o(mD3<;bR2m<wL0!$`ULweL$)fSP@8*_DqEfe&f)zxI-{_NNHo
zfq)o;l;K9CGZ@FGJiT?ACMM!1e!=tpJarolwY$SO@_EelAD5^eQ}HY41Wf9t@hKmw
z^S=@WA<p=ewy%!KZdQ^-91Sb=G-4fRP6%88kNyl9aP{H(EW|`b(^4cn>j+#4Io-qX
zG@Q5T9=63!Im}MW{@_KPJ-a-6fds91g8SqFn?D{-YLZD%;kIPu8^XXn3>pkST(nxP
z=Q`i!y6Yj<Ov>1FRXW*Db6F7S;C6N#E{>NzZQajP;!S-II0&APggutOrMyvvk*GL7
z+mvc*WSnW;4x!<487G(0?c%utar@pJ1`AI&x^!lyJZN8cRn?=9HtD6CDsdg3&cOo^
zpVS*w4ZY)QLbn4}JK)DVsRk{-4@;C!nUJHg{_iFiwx)x(;dY(wdsgj7&)LL=Emt%s
z@YD@>Bls_lD~W>%{+V2JiD*J7v*WprqCPs`!zk7qF1PIFAYm;M@(N1ZUv(rc8cFQt
zwewhGooaY^O*;+zw0-6@efB!}#(XcgfkZGKc%CeKKWl2rN+41gwmu!$VkHl&m<m#}
zzY?{(E@eN$cs{p^%n#hYk&&z??>plW!6_fmH@-$_dg&!^5AZv1AT?l`_%j$&QiCh1
zKx6KKRafR*<zr%D3Dz>^F`VpjkrR=m`T>Y`fi{Jm&Ull8;BA|W)>1xXQXW2VgS_b@
zf;w^7v(@t@&3K!3@&2LR>LkeV7qg}}FO_ZgMkPW-qv1cX)0_#d`qsH#+cy%+%y_lP
z9Tt(tD>iSDqL$TtDr#MUl#(+rVvXBl%H2?nlYK5v5&yjU_JW-nlX)*{a>vG;`jOk%
zXDVj#6g6sN*$_1vZ!yR6vd^s{VQ1<-+tGPD7x@viQ!>xGud}W^6mgszIAgrTDB$-+
z(S3*UmGJM+(I%Rrcg~%xCOyh{8_gVLtm+?Y5Y=LT?1oJSYD%D1#T?A_hnIad*OoCW
zzDt<&{@9f6{2+Vl%ZyKsE)g#^yfQkf=;BzLe|nGjm~reK%!`3QJ01}tY$*<TpB`i*
zktq6<IREPFD?nT#NL&0A(GYic{qUML{_s^M_%4pagJ1eS%7dUiwd-ju`S|1gojU7f
zTUmhrmyU6DPK6~wd^d?^Mket`PK$dwWbV#WoNBuDJL<(dEgGjhT%@;tkeB;ce2_H0
zXN4N~_vnweaU99pH!k<Tw0JjvDHb#bo~+bU-eNSDVNn>**&Jq-8CViH<G>q!003$d
zjPV}Av2lQfUlp52tKQb^nDL!?V;|a>2{|Nx7g@q*!t^nc&)8Zfsh1|__SY<$xPtzt
zX4I2i%~$WlLjY;nU~F3$jTc_3sc8b4{+E>K{U&@p-D2*Y{CN}I{`LNeWY!Ne2OrG_
zTQf-+$sQrNrO=hnUE6{xfQbIb=1ags2O#vI74q2aZ_SIN2d4<;-`e9}GmmnkWlO13
zB}MlCR=ptucP-@+0MmMw5duu%M(5qDs%Q@G7+wGQg7*$i2(SW7Se%a=!|HO!*HC8G
zK|FX)Hw~uWuPi-mX)2C2)}ouvm1U6kKy}F?Ar}#n1F|F4+pqCetzDu4RVpD$6$!<-
zm<&otR>g%F4CQnQyQU*6tjmBp0iS|7Di0Gg;%r;BQIS(w4Is44-~F4CF%E${{x(4C
zTcn#ud<?|N!$-1CzhlLua^kib4oNpD&G4xF4xOj}G>D<d9IahCBK%aCa4#+K45g1(
z{@nb-vea;i*R~bV<+gqBf$)I?7;_zgrWt)&`aNcQ8|_#c5l$b-<uojFkqhwz!C<tT
z&k0NqTDrU~9+POTBsY9Awdkk4*r}cbVR|=2`eNmxf*{j}dLYf?dv(zH&;+TK?9-#J
zEkCU?OFG=}rN;C{Fxe`=&ZjF~u20Ihpi5z8(v-rE9{^=`ZZBWldSd}S3`dHyHD39M
z^0AOR^?}9QPk(O2rD~{Lb>_rPVa3;AtItI^^yImnI>vGy$LaHxgz1;f@YFcMkX2-3
z<@PcS@Z?{lUzOb`f4Gi6%GjR^BhCM+#b+9Nc*WwJiSa0bg>h=o)1<8}-H040ojv5E
z+9j7EeP-wn*!5G|Ne<z#o-DCA?cJp_9^i$a;437grzIK$nBO0WmDkwk+4@m1P4O|)
z1}**pa{H<)HP@ukCWD@pHHG)<1g*ll)BvcgDsUh0EKsaRwZm>+-r~5~U5oS1xWuDN
zy|L#@j8e0w4F=e`K~B0iev>mOW&d3V|C-E%`w@&>)p|D~sJ#_Wwz!eXr>y7D<>S{1
zCa+GXTS&nY5exv1?yZXl>~2jx`f5|IsIO?`&12rdzLwMQ-CxF<(z7Qi1nJ;o<-Cac
z-&G0JO#b>`Pfi9-y-2@VmD~pJcBu5Xo9mM~Dw!__Ng|9qT~0$Sj+2m~<a?PR-^uU=
zdGkbQko61I_4^V>8_oM?1naDW+-1d?$^<mA#a&=g;)$Kx*?|i3;V;i^DW4{ySC@3t
zSwPz6;Rc2mas*!dk%Q!;%fd91N2OZJrS_9O!`!1m@)4v~nymx3;9W9dJ$Q#mB^J8x
z+(Kero<EJW*gd&vA9-gJemq0KK{>H^<-2^b?I+l^=Z``~xCW7A=@b<_z1x9OS~3Hu
zt)9Hiu54RUw;>vV{6Pp$Y8nJIF~-WWCQ*2wx(H=0djT2h3ME?RwgPH~{QJY~<|B<)
z?o#oT*t$O~W#hFt?}80GF>&0%m>(Yyu$tsAvJadt+7q;-r>Fic<$my7yW?u(YDGY}
ztX;W|c4R7DE#(|RCL|67l^^Ez1ochgsZQ}i$|PN6TQ5H|hmab`dr~+}f<ufLYB9e(
z&lp^L7(IK!S;3!&ahWrzx!(VGVd;5}4T0mNGwVEsU^EcuZxoMWJYtj%j#yXdOv$@&
zYH&WJ>;?#Q-^H3@Og+6RmE~nEKVm`;n9-Yu)FXK4&!{uC^Yh@~Q&_6SA-+ZNZ*Xt=
zMQO+i!7H<Ab$U}ndy!IZ^_9R8m<>vkIilxS`Q;Wk!%+#FJ(FDB&k{pSsVo!X)Y(F3
zEY8DV61zC{dCI-98GX$2=|`Tf3TN9o4nbu5^ADb#_;x+_K{rZco?T9^nZ=+D&vG5q
ztBUg&`3;vdWXrL)33#+~g~FBUNa?|&c)yN@&s>P3Zkx!>)jZ@LR0!_wQsd~X^)Q`_
zR_7g7u{VH!ycgE2c?0~bYxc4pu6Q~A4d322wQ!tOSBI&y60d^0&4a!>@iTH9v!7#J
zb>lfz!-2ST-Zs;@JJ)5Qno{z(2W$`R4HvJTKZ9>e2f2R2-|HGXZBidClf)N>I-91-
z#^ZTgRh)T|)sc1{fvw3^m!c5O8K@od?(1InxafE7m%pv@gESDm3&r$#TJkb2Rj!O-
zpa+dMah=A=wx%ULZfb1Oh}6$0w30HgWO<#FyL5AyD2?9q^Xt{%TecfVp}hd(e(fr8
zyRU!5wWIDL4V-+U`)_hExgD&Auv&lY5=Sx4KmjsO%NS;MiB08K03suRvr#_{Hj^L`
z=m&8jL|j+B1i&aDjMKVRkSK_F7|j=TB`T{tddiaYBn;{V7s5s%q48CE@peqswB%MQ
zLX_tM!Vcg*b_sJZ*-qovcC*p%$(zXpM7_cp8J^%q9^w;oRpDI9j~$c(^lpi0Sc}=R
z>W=GMr*Sft8|>1QPvkB;H<W)^uWJ8UZq=VqZ@v!#%8>6EDUM{5uAH*1Heg&$Mt_@K
zOaH3~<qtpXkO&Ei=XgPPjU_qqjhn?>`b*3nfErcgWhTXm52VJ*S{$L&-c?K9{iDsE
z+!*f$-tWmL4WWF~_kw~_2YdtktU=tC!RuF8dXv+(R3SnlUE71emPhRle+|=!+`33k
zEKd8BBN(8t{AV@wJHbUJ@(#E3O$TYr5A}<hPQb?x)n4PDx@0M><+r$-N6{1Krxxm;
z%+{3-B#yLr$Y@r>blT!~C;xOEKAj7kUG|jfx`?9cx+NRD+^4#c&XlfWzjjpW3A|{j
z++Fk6W%7a#Uw_IB>k>^mIx2GdIBdSzY=tzo$;TPs<-vd_Da7hcaV>Nxf=JYRd+)O9
zT3C`}w;o_iiA)W=$T!Y=NKM?qgkjrM*x^l3xp{<(LIEcef*Y-7VF}<rLGRt0)+LyM
zDu3WmW4`qTUS*jQiTAD(#`_Os*WB8z)7^nFq9iqa7pKzoF}YN9RnBL*dYUl~%Xs;g
zC{foSYJjDR#T>ZM%VT}+2qhT@?-^GvavNKjeU-|Eac2-I$&_+9dwy&%R**%dnG=lp
zJxw!C690?K-1@V)*p2n^&NHHK7hQLC?E2I69bVF&=)CpFVU$5TJymJ@CZoaRc%%{G
zm#j5{2@Ee#ft23*JFjB)8K*z+RKR|Z@hOI8xZWHtS|j(iVl58T?(Z%5A%J4!b$jLk
zPc!63&hyp#gnPc*Go!@U>z8UW_xj_*^IIe5xlWf5=$QAL{+LEIgiao!IDa)(rvyBn
zR|*W=D}hO8)!DZN+YPXK3m*FwWTWBE{8^Si)-bLuLOjZTr08g5EjxRCY8Y!U);AM9
z^6@Ew2N?8}MBoZkAr-2brD*%<jkA)rHpR1!o5~gsfl#@EW0BhU`80qs$sp)K*-o;=
zuk9y?&6}r|B(>lZK1bVUMH*m0X7J{I6Z9~1^9Z#Vc&ix*A4r#PIc*HKoIN=F`|%22
zd&`l1aPS)0g^D1xogEvf0Q#XfES+|K6Z>IxZ+3s0(*ue=4hTNalA7I{G{{rle6$r+
z3O$HZhWL_{XU+=9mX<Pygn6ZPHjGCzV0aU99+5hGvVLy*Q;p;UuU^}mAe;s^%N89F
zhH|Z6o%=<|{ytlvlr8Wj8GNJKvN)@;SeJ<o1X~<Y!l*NXZ}2%|T8lj@oG&%UesU}L
zG{Zu&U_m2|64JS2+tjPwPNQd?mA0~DV=Akz5!5I&Kf6da;bh#U7({H)nl+N^Y(7}0
z#Q)Wd5g$fIWY*!J<wy@$!)%^o!r*_Z`PMz^l(2b1tbNv>bhyh@0eMEk%<Lrn#ObzI
z1JxKyOwUeDg5@EtYZYgBE;l%7bavi|+R^X4>-<aud480brEM}ve~tIy*|zc_2AKpu
z8D+pE7?=;8fIOd7y>c-<s0iMTA;ivZ+aDU^RXa*5;KheMT<7xySxP=#>pNn(wd&N$
zbND11^r9PV7;bSI&|)J{3^4f_K!Rb(N#}E1=Ow7ai9S;n_Zz12y$htDn2;>r<}^^B
z9X|hxz4;L_&)&EZb~P)@6r-8yZCbKSJ9+0aVPca|RARStIA&2&6YB$ak+twYo)i^=
z0DQ7TMK+yP#v0axd^^UW%m^C*g!qhedfnkze_SF~j*yLe#<>q3(Kx=!I^p;4_Gf<S
zg}gmh`xFd-_N%CE%**3^&DS%MGSv$~dWvyt$$1=USLNONNx0Om>Z#<RSr0tPXID^Y
zQbU^XnMVPYe(mm5*x2EMMc@T5&I^A$yRj8Lj;r#ck1A{$s&)wnYvIjJ&OqmQO1{N5
zezli@Q?@A(D&l_otGv8VVA`V}3?7fewGAC!sp;hFoC_cZb<P%J|5cLuYTrHyvp7gw
z?o0{ZaX$amkfWRR$>J!_LOUz|P5p{qP^zJr0jvBYdRyt#is>_aeCJTkFTk&peo;nS
za9k!cQ{eHZvMIp%A4D>mSX|qC1t>WqGBegfoH<Udm`vN3h8!gl(q@~SiyID5h>B#~
zLj3?3dTN%Wzft)qboAm#w7GuL)|&&J@oP-_gPPZx>c}z5U<1+_V^A(Q1XNF6Tg|j$
zEaZyQ|6vJqbFQ3RVkdt*rdaAT&pvx4=>?g;vY(nP_C$(iC|y;Z;|x*t=?V9mI=8ZB
z$!gHydHnzif|&4F2mG^Hv{InC%427wh;W_UdfE(?3oD!MZDnWJJZ@E6qpv?o`^BSt
z2%H{+{4vLE*8(BSQdoxd5`rQ`m9$IYT5S5Q^OAu^ACgg(pzBiSL&UTU1#U;eB(vZX
z@GG`bm`kYQefXw|Ble>)m@9g7@x~IYlrD{b{LeEB+`PPYfnz(&rLor7^0+#9yW-r;
znE81NVaf_m%OEv#<K^j5W&pbPSDKQyuID}sZtE7GPFhaYlm)p3orcIp&zxrNOAo-@
zs1&jAyB*)|LV-9TUU%CUR}piG1A*|v^gV9WyvMw=Z;+xak0DRze)xC;g`&RMQKL0r
zqrT#>8=+)X;;m%urf&~C5183+iz%30B{bzLq2%Fe5+J{Eea?h=$~~0R*-*u4384-e
z#y&e#->ZLhqh@`~Fq=l%q;F_|LU873^bQlNe-qbpe(H7(Y4?dOUGNXz!*G+$?!iH|
zP}k$NOqq#6rLH0!Ba3FYeyB4S%<adhK0HB6Ao%>0F-AsOPfApHQuIeKfR0##Lat1g
zPSqOrYn(UjYC8LFUAc5mWFSad$E+^L6TXL8S<A@)nxK+q;69ofSrKNWLD8~iY_i*w
z$|oV|0D+hu{SgoNM{Eq#Eli74Xbq4<VLYa<H76M!+k>fs1m>~=II+uqsP;tgQnGwE
z@y&eGq&!eDF10~#PqW%OaVNupL}Bp-1E&mWS|@ruPiAP+RAN_ygzu*%fxSE<RCp^U
z=O?mzC&TG>elsyGUcRNMduVHW!=#p8g|IzKS~lV7*!Ml4{(jEz+90-!+R5PRcJtk}
z(PGXevY~sW)8pV+yQSO4FydGwlr5%}gr4kecxQMxRGdix0qt~z@q_nROi;k_Ygw_;
zYYO;m0anqisZpBIRmTsDH<(^!ie=fJdJI?f@f;}<nHArTu%ZnCQz&`b4)Cgy1sT<N
zx>QoObTPK&Sm0a+6r3mL$i(ouRuiMoX<BIs`Od5;ShjbOI83FclN;wbgFs_yn1<`l
zh_h0Kbyk^fZ$?)tMrG+inGN|xZpNA5aY;r_^dx&=NNeYpR{a=fR_>so{3rKrjAzn{
zK!qlu@qLOwV!tN}s(6#O<Wk6!x%Vo+(rXeJN8g(^j>q`b^4<7zaB~YcS>F^TzDc2F
z8v}j#>t)d5vMZw_<2rM-bsyx?=NzCaNb~Bn+s9Tc*5S|y?pxc<qJO*wb=&LyLB$Dc
z5AxjdBPOTIN=AHlSSZ3qs5o>^erM&@Z2YaJ9j}y8Gsw4``cB*IY*LoK1y(_Mi{SM<
zD8oJ!&*Hf>tp9ZY6MCGH54(f}-`~FApR^vkL}i$v03MEm+8zaRqCpemfE-~7YEo%%
zW`qCS<ZUp(2dRR{k7CTG$Vq#dA{0g3G1`&YGVB;{gvY;+v{5fwTVV2iX=v*NRjdCQ
z)q3Z0xcDmwLbbm%v53Z|+LM}Z=nNI9rs8klW2lp)Et~{a`{80$$M|%h>p9n0);Ai4
z)NCOYwJ<pA;+}m`KR3<ybcJ}W)rUnD;XRY~ry04<Z)Numorw}BHIqw+jT!=h>bbsx
zmt3ZyoK*TLTuu^qs#F2Cbg%y3B>(SR|M%g-d<n0p;MS`4-;24o&p!^iId`?r%k(*=
z^gKOjFDT(#i=!?5(-Q)k`@jY<DzI^ybhiCXk85uvJqyk0Q60*QZCbmppE-X1m5bl&
zlV1YLEB&ENK>k(qj@Gbbt>Z0XVBaeF`oJ5yvvnbuga9(Ntnk|nT9=!b`k)aDzlGKX
zsQCD8Z~nJ{|2rwla%MzbfSEOj>Ndq;nV>ZPzJ2{*4u*f+_kio)sMgM?3$YV0JG%au
zn;c@(%nE~^cTqRsNuCO_M!zJHp?rs?YTCew2NzoWA67I>IQ(KXn7hYQ5c1xkIJEZJ
znceqYCS@=F;7Ee|q|vn$Q{Tkb#{L=^3j_GrbVcjo-$Jz|Ax))g$--n%miGHEpR?NC
z_J%*#?d{pPPM-xSoWY1@vujR?sU)~Jw$_q*c8!)}yrqSoLjP@Ut2FyuUTuHPSbG*f
z0kik@e!=KQYvyBV;knXpbliIC@9E=RW@MpDRGg*W;5ZG^L1b8a8?ya>qPOOSUVs%)
zWG@}lIlT0HoSWgFdRdxqS4z6oxjEDR{qs-P=4oA`(RQSJwtS2=8eD$nsg5+0ZkR%^
zz1p&~i59+uO|x3HGw#jFV`kCe5<_E_pCnp)8l{N7{c0zg%xDQj=Y7u!*OhUH@^P5K
zv*zudx>3UJ@+jep;y%)<R%vO;w94&#%?TD3Fn7>n=ZJZ`n*6@8LC~OF3@Jw>3H0^K
z1{QI;@NUP8p$SPWY7D7N31KlKydFC@y3}E%H$iHb$(ys7e<Yk`spKi5sXar&S*QiG
zoX&HFtO)WZh+z$jj>5i;5%Xifer*YmJ&CPrzgb?v#{cyC{}|JM3u0-1uTm?YC{I;d
zDMqm}D9@#t`kHC3@2(mb+6(`!f<EnmD9Ge)V$^)SaPV|N$mFN8+Wr0pS>1A(!Cnbu
zkEF?Zja<Vzc1hYJ*YSYk@K@QOtV>K%b{rpR&ahV-u%m)?+IuN5yRhuL^Ci;HdT#N)
zFTU**i4P#xmc)ibK@qG4gyltbhPtoOE98q?%h$J`H#4SyUIkMK{D(~g)^TY`5UfTj
zLWz)CoKO*f6kyjs4&2s&N$yYYKk-~mf}7&8YzwNMcDCNUh`7WeP&7ooATjynX1X<V
zBK7~%`qsR!tsTd;8cc}}-V{?^v%KurlnYD3WqQ$Yb&DucWLLd0n4B-_`q|3l1|^hY
zJYL0?Pq@B{JbFkSc%7ahaakqtF5Y08zemc!JGyWJJW@a<oIsEFV2GB>=)fe$Ips~m
zwdxlA0NN~yzJM10e3Q-k#jd>yIN8g#7dPmpCF6foHIc~kFP$eXP3a|T4Kgt{c7o&Z
zK&{}c!!=gn!kmb-X=}c5Zn3}h_b$YwFUs_L-<M;B+-Lv)boa?U_n;e=jb^U8)+nE#
zE&4Kw;a1ksZuu2$KD}*i_W7S%yzIk&sdUK-J|^8YRK7vAKQ&CMy4AhfKTELclYqC^
zOfCvd?nbBcLSPXIiZYNn9)~n(2qr=ZJvJsi$t?QSwqIuqx#B5H!KPP8mN*{<h_bEj
zkMEBTl$Y@U^yzN4MogzyN4wjZsYHx~Pb7o#=x7*Yr6cg_T0eEKO|y#_9ml2#3Pjz0
z3$c5I$)BtW(edj1sn=xL_)3fY$V*dz(PE;dNqbl_^DVEa);3d2=Hn-m+JY8mE1cpf
zo|KcDEK{(Eo5f|RfS+!&EMd0)x%&UBv!<rB?xFo<#9de{tHE)OB;`Bz?fb3bg_6fc
z)fa^oztlEuXjzvw9>n<8KDzNI{zft(k0aU_z`8i5V8!6*7U}!xU~|icVk{H(Vn0b#
zYB6w&I8N?`i58pKle%ui!|>U$B6bRPLRtHFjle2$V>%L<jbC$+P2dnQ0-Q2tDok@!
z=|ob<ocZG$!og^;&$JCWmCBh%HX<f1=b9Qyr9hhzei*0EEo4jc#dn8pwdXNmy6%^C
z{-4Ed!z5cYp}8hsCp<x=PETb<I7@iX0YBva;$4quDx=!KBVKIk4$8N=08G5ivJ(w}
z6xdfCd`HV!pubh-+_2mlPNSBmWOJnccRT#278<l6<h-V2nAtOgLMj`r=^Cv&@LOEO
zcR<XZf??yh#~1PyiZjoPcyh=gTsG;0+91W!QVBCKDsLTmAwQKzL0>ky?7Qnl<-W^r
ziuDCfd}%>&oD#Nrta2vov27VH`osJWI9>(#*uy^UZe{$^!zFpXLZ#1N%j-&4$RC+}
z2*`YmTdV&I$Cm%nQ@3}~8~Uc$E)4Tyx^y;LJiPt8>HTshk17-x<vDmS)CJkY4MC4f
z#v4TPOmhLrtS@@BRX_5$%6K^i%h#JF{a(NBBY3xI(SzJ4%LKTtUwsqRs8QwiJdG+O
z<J-rsI;CUq)BBW%*Czl)bS%row3_979g>ura})>={gSeYwZ=SY?W*&j4DwidZbUH!
z7q!bL%}d>+y1#QDl#()$bupWtyf=)}`Es-Yde$bXc^^0~e7N*T2=N-w5}|!-Tb%}5
zyY93{=U9(A>Ep5d!$xx$-R9%&f=GlJu=nm{{F-NRwM+SGY0oM)k%DKyI%S4B>iygu
z7k_PS)hD-0zdnFZb&ey&+|em0?N>>YdbuefH;2!%d53s7p^jI9a3BH&^BP$y#@WQr
z{toG?@HvFE`aH92*M4`r@z&yHs0zhP%w6^WpPBPUIXhOf;U;|)K3M5eUPV2%%(i!7
zS9)IAICclyb??rLf5tz3!_w5`{c+ZV$^;Vz(Q@X&=l>K0Dt~D`Fq&pLkK0~8C6kjw
z=7p1$Rma8Q2D>ODv2Q>2QJ11;2b%#yHB;XpS;YPny=1wk3||szoT2pW?BgIAN}yVO
z1?w7+DeE;>XzHDvD&dtNl<Rh26Y&GC5r0^e7L5Dr-dNMm6VdB#J1nzYPAA(F8z;xb
z@MHxk72slX<&Kxfxb5-oHfs#P>bA$dO8T4QhVp2^T9eH^dOOE7CDMm>_WS>7@_Nh@
z0OlwIGtNa3Rr#Bh1x3_&+FWpQS@jsQ0*F!hd<Uz*|Kr~#y&lv0M>Y-mA<_fX$kka1
zr2JmtQjjU`>Cu)g$3R`i@R^m8Bj6~&7`pdE{180X%*e@gCwLfM5NB}uWCwLg_MMo(
z?8IF;6S3&o+G~iQ@gUG0x|%xE^B+`4Qn9{W-BP2lp*ayYt_E`JYKA~h%-Ge-8Nt^w
zb!z8*?0+x%IeI+82+6|IU6umuBFU4dZMnuZbbc$0UvJI65$a=KJ^L6fcA4kj@&iI@
zZGTX9{kI35_!H^`YH>xuEqx*Ofns_%dr5(dgj9hJ;HsiVF4aHp_U9Mvfk(8c^V=^T
zpj*%0{HNU)FDgSPZmprBACtE)v<*zZcfm5K3Ay~Uc@N5H&~SP*%^T73Xch6^e}@<O
zdm?>*pqy)EvgQpw+pQKck<o-|#lvGM0}GGArTVB*3G`>~<9#`3G+}YlDRiw`F?OoX
z`9O$M)U3*_8)(lsyLxw$s$k8;-K4(9ycd$H@Kg-E3f1hx_B~eA>i1ShB{?1tJ9F%z
zl9042zMfqBM#)lDXDh6f{;nX9?fY60BJB><Ce!#WP1bOy0mbd^gPUjicJ40S8m!AR
zA_~cemIAS@@2*9~h6l+0P|#{PdUA<IrDN;P)wi+6`!^J=EY~>McK$M?@#C-dvFuk1
zdtJj-M-@905(3eS-c$2PM0LfOV=X<HnhZ-m^ba=$731GDcI~kI{TWhPyYSTawd9C(
zb_c_m@ym$8p51M*k$iuVu@2V7e+Y0};(A|3kfb`Mdc92D(89zymiX{%WX{Pm`UUha
z;_;Bg39G<>{(Pn0o=d04B<j?^W<AyeRMQ<cmG`Mt`%b&hr)ZOD3*KnZOb<6!IA7Le
z_bTV;7pjm*nh<qds#_7j+)R`5w4nT%8dpLhPyJ`4%SQ>0SX}a|Ro?s3x5ib18S$*M
z65G!{?6lTn%^%D*JD+HCs6>0S1LV9LdSbGx10Tt1ezkW{5oUsP%L?bM4$EX_n2g0<
z(FXk8xiot=7X@uqa($d~YrWkwhetA{v+4Z}<IYt-ksG&k(?F!2l#7f3Z)#5C55Axu
z4R<}t4m1lCDePZbUOxSkXU?0o=78@`)?^W{x2YJa<jymh4(tGYxK+dTp8Ja;zKUx4
za?yY+_HCb^mW(9N)yElk-^#x4n&W(07@C21-TL+Mj^g8J8J}F4A20J-Z!Oo>R!SP|
zzH7Gq$q*S=dMz8~cpxRi+sj_iCh`j`-x(KQfAlj(Kg@+=%lQ5Du8M^V_Hqd<R?O(0
zS&^9-Tu7%uNYNqcDo>CD3$=f`kX$z2Wx?uk6y3}k`cG5sINd%I%D;Av<L;Hgh*+`3
zqr|hXlwO7*NVsJr8V4uVB;?1tRc2N+6g01;R~)Ih-E&vE@n`Tfb^E(USQmmc)Me!5
zn&bguNelH2qFwTjQq9WD3omQn?a7ZEGSa|8RT?WpWHCN>v%LDHnJkv7KsCfTw^ap{
zP=)m*7zVpPeD`83PTx{2_N<24N0rkgvNa5VfkNxvnA{WLx}1#NDlLIrli|t}cls&{
zQEi3F9#<vQAfKApinl09E`KfPWw!rFAqmt-WA(z$!)?q#DFuLl4vWi*Mf&~QOV=0P
znBv$C^l=r#U3aZ6jmkNR<wQxAHz?WPacU|yW;524^%D4{`C%r&YO#zSqUt*QO}0d!
ztgx6y&C2Ln?cqZj6w5FWVL55}0qvoKkObAWvHtT2tzpkC|5p0n*NSWaT3z<hY?%@b
z-)KeM9(OGXWr?sh;fxWI%x|un9G1(e{_lkN-??~6?Lq|SQ()V^=i7xWt0ONX)u7=h
z9sIW+_kS5kSg2XZU1e6zggu(H@i}9YcI>9y+WFmg`{^b$@|^x)nYFh-&g3$6ZQ}iH
zw$OzkT+)802^Yr*^vbO#74gLk-1xnFW$&N!D#fc=u~Q#u-Y>CzI}M_Aaa2rk?aT;F
zyE|x{k+}ccarC-Woe$6Y$wPS4s9E`trQQ9vD7u3h2lDuJ7yZ-M>c$f}%+h73v2Qbv
zBquuLv&(Aq<X!jz)mz`Mq~yoE^)2`<h4ULRt*ZLuVAWH&W%=5;Gk4g(-z3$~zT~Nm
zZc)S5TM`R<@S)7ONelFrT1oL3R!^2~m-^11Ym3WcJ=qUjjOt<|s@Ps;le^LvJ3WgK
ztRJqrG(~?X+J`%w*FR|W%h%|vU%7VpWX|;FmVk7AqvPE&kCF;?6SG-|NTB(vzxqtA
zpb-kbYk!M4W|y6gIKgBA(WSp<OUtl&uI3N~?hi9)dgTcETk8!1^}^DeID~p}PjI$}
z*2O1Cen4u#T3bXX;2B2dvfra#8^i7I`ymlA(gs|zYw$>MLG~<uK8%vGu};}WY0KiC
zf4(}Vl%v-Sg3NKIdn$feg3kgVBQVJ)C&R(Ao*g);-(P8SECgNnl>Z%iCZ?3x&Yk!k
zFw_21tp3wtQ)LXylCj%UE`YfP-g$B_OX*iD;C;q-E(zO-+$P;r@lxoo{!hXDw##ca
zkG-I!M!ydBc3#{>zJ2MhNBsEmZ#*B|RK2?2U7ehE2^6Us^g4m(t^bqmYr4=f&kma<
zRZXv9A+DcPvOs6Y3Ka!Mtjx7X2B&2+BQsFC=Fc}dZ<tNnTvbmJHTd=u$Fuzqb&^xe
z>A1A$2zsdyrd?qHT_$j;X*y*+4vS6es^u6PHhmta(yh0|DZQ05V6oH`M^go7XRn}v
z)3Q=u3|<qCy7I=Ea!#P={*wk4cDhtA|L5<En$@_6S*b2zZ8x%66P<`=UrG8X6IgY5
zb;cs_`6+s=-t0&_t9iG`@kp0`KjV4hyUF@dgiQ&D2o|p}=@gg)|F;+4zB$M=I7P;l
zK((+t8x(KWeV^uK(g1U9k2?Nm;r6J**8Wglq@j*y!8~&daa7l&qDC`G!4|%lrFIrL
zCfdBQc15)5#{4||eGL6_ZSf}w+oDR4g3gRdrWc0^wc&?uRFojDQ1xYr{;ifG5zg)c
z!{5T%Kl$I8(%Kj>G7TD?7s_ELQh%)dN^P@{sz}?%Lmbia4vmtZeCFoVQQ`0IGW>e3
zIg3zVhL(>Ic5`~oT+~yuE-Ad?%kXLHf6VvBuR2Z;@s+_sAc^yJS&Tm5F3S<!?z6%x
z7FY2<4K%@h;xaB<cj*&D&Juw}LPi$a4fN_}C7$<`AN3)s^rTDeU%P(QSC$riT>EAP
zr+QD{_z{2jY{5{7Ox1bl7lz&gWFY9pqsN!<N+Y0d4tNj{hc*MjbDhs68XAU8(&8ZB
z?28~A)aHzVOPd@miEZd+O^XJF7)T)Lmrij>M?Z-Y1C68ksG2KWeww?XADZ_%@AVhA
zV|3BPK3Kq!>7+SyJ<xi{R|!=^Z=2SIA2sj9`4tH1vcrnC8nmru?XZ-FrI{@i+I~%-
zaYSevJR-EFss2%vXWAhrwHx;PHK{|kb;mpi+$GV=BBN$0IUEBEU<(=$k9L`@+Vz);
z6W)FkB)+HaBkt3$n?K;AlT?Pbj0G(K*LUR}$HA3?fQ+%IMu`HBmyD~VBk!PRDr4tG
z=F9sZG6@3$e!Dsf-%d5&4&zrtffG|Rx4-fY(HPhG)M=e?niEJuEjC}7M}*%43nZ$Q
zzLY164fl}AGa*gDB!*c+Im&f<fi=<GpnO1s5#19$mRNozr2zXg+th-8I!EtX#A)tZ
z8T+>rWZ4p~676-4zFMklVe`K%hSVlWy%gNcoF2Rjcl>jBt0bm8Y5<?a$Ai<b(bb9|
za%ZgfjTDz>N+tT|_vFlCh_<gx=F8@J?+#d~$#{-UKIBva7$gx(@&;Z&^EaOj5{AM9
zIU2siqWeU55mgn_IKM%Q#ML7nNW5=ZFXnjyQ$P$`3sJA3)L_?mZ7&Ir>wRU8?<0Rv
z(rQXM>RQeyP@0s~f#B9qwq_3vq{mdB`vCF4llK@(g3IQ<<|^3@dG?_@kr55O{+6Ff
z5ks}U+u-HZ_kA6j0XIWu$<CvKi?1Jv(1Ah|SGf}XQ#R=E2P?1o2&?-)>vlgmv*L^0
z9IUmjvBlJKJVe{6{VvboM<HBH75N4Bh=ED~@9E-pM9)y%IJ$E-{^9Aj=+U2{J0CuM
z_Y)s@Qsi*sgo%PRbI1d5m;B+5yZrNY8b7=0%tNF1ac7fW5vm0wg$2;Ex_DijeVpv@
z7OZL~CXp3b{{hD-7xh7|XivG4iMsej_adr7fZs%6&Qsy5-C@V|=5V*L;3u!dy6J0u
z(J@m=xvuBLLw>@+LOfY7_BQl0150&%I~~md@w*qpZP?dh(_7AY4MQ?mJ)OT=d$k=L
z6BR><<t59RGXL+oipB3=zgG96nHCOB+FusNkh^=}RpWcf{7wGfXlxFCo<YTcNiW6T
z-xK8(y}o%zQkPk)(Ajj`Zfh+rJJZ8E8q^o{NxV9?EcNeVe`dG0qsT7j>|tHdjN0au
zV4UQ)0WFh6vNb|sS8GHx_L}{b*6qf{!lE$H3!nJlSKlx=G_Ip0P;Ws-h!whK+Ml87
zvsR_Rs5RTe_~y8)mmwTv$F|1Q#5sR|E@v@O%vj`hah^Zjwo}I?Ek_DDvAMQ_Eu9XR
zekLrZPb{NB?k-R-H{(_khe-6G@!!|Y$W%`By>*92y82ZG%Y+Cq3#C%3jiD#?L@xJT
zTJrarpU|Dsijq15*Vk6kLkSy+KhpGO3&VTj+dPu{**Fwwqb$cgEol~|M7!mie|yJ5
z@?%EOz20E)D%G?)74?Ys_JrhDs(*NIg&&Rj^S%8(C-jnv@H3OdBU!_^%JId3$$Oc9
zm3G_IMlIg#-ZP7=v`!B9M^$%9U^ba~)s*NxOnK{T(cCVcxlzre`YJB{<(ozUs}o?<
za{QQhjRxIbRE=%7B*aOayzk7lKT>AlMB*`fXvAX#IMzFUcO}G!khEOg|9kyKtxd4(
zlelsY-##(c+O;pNE$<!gD=R!3$^Mxg^vJB<%8b8;^&)k(3fto_*ZRVaC0@e-8o6*O
zeD%<5lrH8hJ0lJsvG+TGL)o@{C$PW4?_ejJ-YLmt09rm?@I2hA(D&tPej)~gWOoR3
zfxYm}(|wqw?tGgwOz(jsyC4o*=|3!OV0_k*xPqe26V_=49X^@axeWd6tXH5=`uHK|
z5=%BuP5~!)C;@H8$r<*AC$H?8+>_bvb;WEdpIcuH)36P?bkrdBF?|~S0^E=FSMRCn
z)BE)Lg7_p`)rbzoIdvE5FqmpiDn(U|z1qcU_2m4Br1--I8N17=^1oDSG%xxtXf%b6
z99-T@gkAt&3+*ihc+whxO~yQID|0{G(&jVbn!6u0PBql~s66q_4AY?O3UI*xbIJ2&
z*Mz46-d}iS1M$^bpjP~!aesU#W?%nT?^o7RIKBMWu|B3E+a1uhV}*5Nu5H$PH-oR)
zz%)rAGl*G<esCGXL2iluh9i0hd^hih-uWl}mEweepm5-x_;UK~*tisytZIgy1aeYf
zZ=?lyfeo3opI20K=l!uEObAi~s*rTZZ0O?DDYs(7<lh}|f}$dQlvc7-`7pljTOnU}
z1xri<70p@Q)()%1Hdh5Tmu~Bcqij|Rm$i3CE*7)h@<@dyHlU-!M>nzgY`KgnZ709*
zd$vFI1i_uQ8NZV}iV<HV6=$?Zqno@05>XvG`30a|W@o)R;bj@jpD~6CT`2*wEp9k(
zMod{7D#*?!-y<B%@v6BufOD>_oFU^}PWo|4m<M_EKwd&`?3S%YkuI^`CB&?W6Oy=H
ztwf)-^G%n_6O;=``5d%m4Zl~s_>S00&6`CzJ(b<+)phr#vw8HYA)>%_Wn1jmD3N>O
zBE9+ikI5miBkN3*Xo1ejaLS}JKKXAm8$M4#qRvEgM@b0sjRthPZGnxR^LN2+?sl9M
zW6W!P)M<|M?-CDku1kv92c=uZEehJCp`dN2?DM4NCwGcl3f+<(9qd%Qe0Lkr3wABk
z-1|;E*q%jqB%wN`aD6m;Izy53dR#xu-|NIFStx-fSAS??y=-;Y)(xJPnNRi~Ogt`m
zbh}(GMpC{|ZT9-cQryL)*QV{kV+4!{M}e<tN=0M70is4fytDwbl5GmP9HcxGVAShR
zRgiC1(C6_x>HN}dT0V=tk#*8DlXz`Xgix`70gfg%+HT-MiSG#eho)f1q#?1{nNIxG
zu{H#MLV{1D^zx^MA4|i*U4X?*Nxt@MGq8zFZ){m<?zWamC#Q+5Gk!$pi2>C5<$%QE
zAeGrg*&%UIj!~6;N+Zk8HJ)uN0Z+*Z4|k(IuuI2uu3U$Fu2ksydbrQg+xhR<WbDIK
zV&74c%axa1%_fPSX7!4r*2DN^r)8+~7oaI_u{zN<KWYoF(mXg9(R2TLL8)0@azOqQ
zO(ol?f-;rPuXDRRM<2Pevo+t&2OFDZJ_I``t%C`PfN>C;t><0cl%c>oP{kzg9TAq3
z?_xS7YI_Wq!+`16gg~v_{_OihkINHA9m0BxswEXAjk7q5b9L+{cRWWc{CoY<&CrRX
z2Tl^0@GJ83`WV!>balUSbFFEX1U6)kwN;P(WDFVCVmaF|dFOCOZ=y2t-fYCYU8>K0
zuMCZhytR5lL0f65HCu`!7*(*L#4h|i0|GU5*#)@V#f-U_npAp#z#_6~ZaA=p{lxD6
z6=gf}_Q;ZRev(}@+ANM#3s+ommiB4L-rtEEw#D`~CFyuJY`p7Yjm_0Cvo*go6(x36
zXSP$I{LNN8h*_ERP%9yqy?-c_&hNnOMX}Pdik)KfvZXE{VD*Yxd8(QTX7cL|iJo67
z(uv&_DI82>kEhU4za=gWP}$k-@0#V@u&@4(^7j?FefNs9cBh}ejY%j{3Q3gE+UGXa
z8<2VIcKoq|=D~7UMLsr((JRCHHzGbnKe&{p<3YC6GxE``>6@n2gChM$IHg?etR1N4
z&lTLBcB~R$IX|{+0x2Prx3R=llyo$=e0_;)7xyVz>Eq0EAKPkmGhd{u!Xcn^B}6h7
z%E8D-`$u>+cKvljKv1Nb_b`K3PS>jT;n=<6Y^@$21Fq`2U!tGI@`@@r+Os_va@YI7
z;*2ioXYcGTbJ9$-Nvk!@3{w|8U``-r4y<zQ*4p@D%}33~-s~jdN!_|?PEG9BiP^sT
zoF+qKMxa|QM<F}kAQs{II8#)_a3qBa=q^w$Bxni-@&dFadF`y*#t9t*AnB`>D>{f;
ze9p#0-7z<>W$Ce}7JO#ZE<&rHX&N7G5$kP=^nib3oVib$qSlCam(wNjsmk#?b@1O+
zRFHb)^>86d?Vm|l?Cy%p*kp*;Ve_0mUh_?s>*L1=p(R`8TbtU^&cq+eu8{TA@PhCx
zkpcS)MEeayzuQ%FpriK9y|7dvT(8r~;F9m5Oap4Xy_8f`;HSIDM#y{MzGE1V$KKbG
zr<Kb$u|<dSX-ldxta*<L)oRnp%7r8L+T!eQ&HgwI5CjZoxW7{JTkxv!-K>^X^vM(I
z3D;)&Rx?$^WS<&+%8s0<DM3!Aj1mTK%KEcVqgeiI%KMAaci1QCrwK4_gpN?Bi;e#o
zoI0KxBCK#n&`_GAX#!d!z|ZeWH$s&Ik2GuC_J8i-@?}HX5obL;@dD1k%jHNj#7lyl
zvjCS5AI<P1WsWm>wd7V7KS)e^InpTjFyI3yPc~@R%X5!8A~k?ZZfs4tmd`3v6m{yD
zte^H`U~;^AeW{&~DTuoEW~af}7!n8+EDrU${@`E6hLz=__B?&SqXsy1Oi<XqqN(Ne
zQDpzhic@O0upqI`y8(Uv7YcM@q{O!xOc%Q(yGrpsrfGcF;y6Tu=rW+@t}(PxB~u-F
z%&5ia*7n!arETIHym{%he`i?<9xAg1>~aIcT@!NJP+sZQ1om-TlMOefH^I?t1p_$Z
zp+ZIka3oCxan}1#dZ^T_#Exn`)Ie+rPe@>1;Z0}AXD&DtW6XaEi_Lyp%I?^++65~$
zSl9G0;c_Z6x!62hAjP8?ru;UrF6g)appk%qSx21f#mbA2xZRh0jO&B4%TsT$-*t~6
z7TfsYb?drWG;dy*GglVIDBux>kGNdFmoC2X%bz-lT9;Fxv;0q-rhjCpWMQ9w6pXWN
zbX1aEJi<--lMp9tEw1RNK2YxEU)-^n)<yXkA?-w)lA&BC&O*d?io=ovRwuL8CCuQd
z?c!4bOr}O${SBd70?%|&kgvI2%!zzq`{j2nK)z^UK;e6U_iGPIztwo{p0=ts;kpt@
zi5~8VOa}9!@KTxphi4{QCeM^GoqLuUq{-?cv!Wp*E|#S!9NQ<rr?*A@2j1qq>_1t@
zlQk{={;m_c3RN!Rd>amzDzm;1sa&1JJE}+bJUs0&FA!wb26!4noj%VvaS8oF*NYo_
zFQn?LWsBL$O8xz#Z$A8)9`~XlhA!q^PyOB@TTC6<u~Z1*!Wrxa9-atoq(pKSpS$}d
zP3F|SKGs}Fq&glE)$Li#3RLw7TKwf%>LZ9<k+c4}7CI|$JSbIJG+I_DwaJE|angF#
zI5#d5eOib9+K?@tIV80_c*t+^Whk2*aCE4N5qWexB)FH=S1(_lX;e<kqhDdi?IzjQ
z;dFPt?;j-W6>kccw6g`zdY6r?AKE!KV#L=yM(!9+Zh9kAGg&-mQAVrd*UlWVqHZQJ
z99yWkQpl3cphL>Cj9(14IA!Bn%9}`r&Dm)|n1EdaL)`sMb6j?&URpr;dR!b|60UXd
z+HW;ZB};~zc%CdZpgxty83hB7UVWlu67ALH!^~dpK2f<zD3t^=zjckR?Am@d+Z?!=
zxeLmm8fATPtomwc`iF!(xU!?qVGrd>R9J}GEAbAODJh$I{JG2cLe8l{?HNnSTcQDB
zPwJ(=L*%0X`if?upqY)eS|>NB&Z#Z|^0&W#VJ(D!aD7&~AM_N90)7103`&8~m(o*D
zC_$ZT<o|#yU6TJxCbizC)f6q)%C=<^oW45Cdt2s*kvFn;vv~5Im@?ZC62Hc82ZL>1
zd=)&69v!NX?j!D)Y8d1M434;UB>yajDs`!o_j-^?N*QG}qef_>9Ay69FpkCUTiBBn
z)>G33Qg%(P{fu|~Q8&{`lYzt#54V3b2L1EdzHXdnCoy@Ai>c{-U5MJv4=?E)Q_xtW
z#I`rXy-gKvnagFx_WEhC60;hpwfTvP!U*Ea)f~@I&CN_-xLeHR!;GBM*G=t5FH(e2
zP#eb#gn5z<qhv3j^VEmlHlD5b%L2|Hl9{7v4|#2??6~}fPUzXa%2C?g{g!QFn?AN)
ze1!@LrdEyGbs1Q|xA{06+H%OY6Ts=bvSN>kVQkIMpP1CW()B7^%<Ydjd4O`KOMhpe
z<->9psJyw`w0?nIm@x;lP>wG)5Y5AHEBg;f#|_fB!K>POYWm2@hORzUr2^0X-D#R!
z)}L5;6pa15*_kmbRRiM(bq=r{j|loIu4AE!R)+s&^rXtGB=J83R3HLRCf3s2aV{j7
zhLTxtdTc|Hia&GG-q_9pkKIujJB@z(7W6)FT>R_9*V_Co8P|DemaZA;<Zo5K_kTJo
zekZ4b$*eagt-kzIz=Y_JM?1yc;u9}us#=5}d^Deo>(#sk8P1kt4lwK1X8+FZ*Ep!2
z*`R0fo<Hav!J8I1B$2XzbX}6EhSKvxu+K&}!}p75;8b#9%gBMV#EQ6~oZskylm>cx
zcW1mfXkppbxEi45^77XIasg0fkxvG%qB)Y3IE_@!^rkaa8x)jm>v!M--i{N`GBk?S
zPG2>342LvNOVfMaRharL2B1UL#}7b4NIJG#M^EHsj(a^vNQ*~h;Vkc59(J?L7fLM-
z4_Rx}AK8MZnoE0cDGSy3L%<E&(4%JIyf`cCpj>-84gMX()AcwyTin(yF45|*v+pMg
zJ)?K3vtm{a@;JK%UfLJw;E8OpB>`;HHNCMoR@F>IK|z(hxI(e@Tdo}l0u!oWdl|*K
z{3hU0&O4&s`}73_*KsLiyrLV;cRN;>CU#B5?k!h=xdggGeOMyezTU<mLPY(6eqj&=
z5IIY5QORU|$tarba{Z1N`B3(9;9%ADpwY7AonZ!tYjVQ90kT6*Rmu!?JK(`etaX7z
z!()MpQ{3d&{@l6Mv<~^r4;ifHZeW)tNUp3t90)c_aOTMG#~A=x;BdTmBKbfpZ~o2e
ztam}C3fjK(X&wzny$V4Pg-j)Z^V;fRf!q_1Z8_dNgq4q!5I!T<>S)I-auu0^E{JQj
z2JDY(TdbgWzCcLfRV%NL6$W8x34Du;1KURPqUfGiztsKSV{D=PUk==lL4fyMF{$Oq
z@kH<Hy{T6<SU3n?V3wT=J(eT<P;&i0jD2@hQ(M!wA_^9Yq99VD(nUZK=|x0AL_kD(
zvw*aO-bp}2L_~;msi7k^)X<v<2neA=2oQQtLJw*1<v#Dz?)83aJ%44LtmN!{rv7Hm
z%<RH==3M;M-rV9hx)FlKHk=2gi}RQ2$y8~#UCLo<wX}*W_D)QJG2PV%w7N2c2p!lt
zbX``cktGRuYinXg=k6bhKu@PYA0$;0@!3J)$HC~6h_0c?rq8Ho6i}<NVX%%peF(D~
z#Q}X~N8iP|(-M--r8nr8KG&1c)&-DhP;LqRyzq?^a)y6a?_4AJEuXIsSfSkW(ZW4m
zi}w?4*8{V}{2%i~k45;~7e0}gf1Xe6r_}X5dFUK=zI-KyyKHb;#||R1QQm3%Y}1Q&
z(#2q+Txj#=fu!@$7sszxsszli>sQ>y^oR2hqjF1HIY&=nk?z{{)tJ25V$>O^O78G+
z?b~i=`d9gEHnUAlrKChH)`WLbP6#;1^R7MhSA<PO1FV-q<10gyPXg@e$_MGi_89o?
z_juYC&E}-2@P|U~tmmd&of?E8$BNXv+$OCVq(+i2jzfV$u=Q?4Zi9dUpW9}1=n0EK
z?MY@nF|nQRomOMYd^PEAs7L4b77(%>(juuh&RZeG^mFdaI`Vt$2q}Dwi|3sy()ww_
zB_;yyUk#pEw;#%OTz1Z{*F{Q!)z-<C3$O3nYh!M1Q|~TZh={Pv)76bO^6k}$8+kR7
z(PTE?CFRO2Hw=9Sb0W)A>9{;_(MX6uPaGs0X-WF*@ci>B497!))b2na<Sh$V=#|hM
z%XlfE!wGY44rFkA#^EecOryBqgYI$D(fmR4S0g!Zn~JK<bgp^m$*B*Da)sHzJX(;L
z{yy{OgCb$n{7ye(f#`BkfPR7e!({J`J+c(yOJ0QXm-T1hRiJ%Ug0Ve3Lp`ejis|_T
z&yFL`>hZ*Xvz*YxU<-8@zyFy3zBO|B2IR%*h@Gd)*Q>s-lg~<ZeGxw)y^0(&*^ZO%
zer>XijF#}!V)rL?|Cpb5-S=-YH7&0qJy$RC$Uh+$$o(9*w^*hnKx`;UtyhaA=DCyz
z`=B4;%bfGNx=tW(yBD!Wn*JhqJ49aO<{0`=u1xyUSuuQ4B~aG)vB}!Sh|B0$f{m=K
zVrseMi<8(lTGHGqXCQ8!zeE<)x|Vm8Bc2pPW3l3%S#9(DscB0#4)znFF&gEY2bBp?
zF$|`%eu|Q`Vh);SNXJJ$DZS+xulg!kTI|=EKi=iM=wV&a(GoH$42yjwp~n^ISg9%5
z1b#NLE@Uy$e&Q@G^kDlTkUHr>;^NE^F*5ZNn&Mv*G^7-$-S!atr5Jl{KW4dHd1FCa
zFyy7f?v-zn9<kC($=HmO3nStaCHJ10C`moz<z1UeGCw<~;cU`;V3YM~0J1USudEwt
z<#jf#oSRodw`oDo68vel_@t6-ta(HpRN92_)a?TPi~GSbbt#2PKCEHlD-?Xb@(l*_
z?Ar*FN~eL_V^MJ$#yh-rOukI3awKkqe{TH-^292Ce9AJw^v=SMG+Hy^8gvaX48NLG
zmUE2x=Svy>ZBmTRs}OUIjt7_D5P_HccJmzUgFUhs{6=MmMJ;}wfj&>Tfw;`8>Ek=0
z70_!~r)AhD{4(guDm(9HW{lqf-=k)LZtJ;Zst>)hLuf^qV&Iv<97Wf|c;o5^puCiw
zUj$84Mc`d-l)BVP%EgGU{UgdrNb@PHy>M|)sZgiLJm<U`MQ!O8vpK;-`&X-aiV>fe
zx7_4$P2l)SXDe6=91hLP9%5Pl$SjA($?+R^6Szpyo`nGzTgAhl{S56wAr8r;MhOi|
z-?<AKeLM>}9A;gGx~}4xYSINduTs4;w~DzUeDh1ru#iq$!ho3{RiN5E%HPhSC0Rga
zzK_UC@?9_3>P|bbN%nud^rF?U>e`O=f=Ep2W=ulB_WR>bj%vl5#B_N@P~o6sUx|CM
z`}lXKqb=;-s^(s+0#oyH>#4ve&qMo^F!%;;aLn2cHpVirAxd{oR9bD=d4>Sd*&RG}
zaM58;@6=R6#LzX1PfF5P7@ix3cdqgp84leI%auNV$FihyN4P7$$$leV+&XH9_zbqK
z{D!;#_V%uCcB7aBo1ZSMZ%2JFO-DWulWUJtaQ!s5^rLJjU%M!6!1JD87(fvsSgccv
zIdJ`QuO5G(`9OchOnxr(>8N*$A>wJIqN;C-9?Q4`e{zUGU4Y9$dhpk)kRL00X_6sH
z?vHJLkS9y7%2qBszi*@skd$o%8>5W1Fo!`Oc=AR=Bd@~S6g=K*xn)L}%<50*ISTn`
z!hrYshO9$hbsM)2bBxzTutugSzo*s00fp!(DBQ;20xIe{=RaPFUhNNl>$KqaN#keI
za}nN`UxWH(EZ2h;&J~<0%CP{}FRAPnxV1BOGY`}>taKRKbQfL+KIfF`Do7bOKINy_
zD3R)HSJ9}HP{aS&g2>lzqF?MZqMF!U*2%D@&*>2uyo6RyoWMKxXRv{J5+x~%s`2;p
z+3FGmEWQd4=qL`4U-OpqyF79Yk!Zj0@)B%L*}~%Dow01Mr+paNDZA}OWo=Z=)3qh;
zuZFzJhr8!(NOjgm*yCf_h?fZyMiaB-%ktFVAdC9=;xu618e8)#57CdOBp~6+1vo+9
ztqaQ}x3CWSK;Ez=n-S(N{oFc-`B-ZRr_}DjRb|Yn!OY<BKDmJ}p;@{KDr2&bP5St6
zmRXzmxZnzy2T`+kDsK$AHF>3l=U&#?$mT6otKP1~6@^>D5~!IX5|p@vln$m{2?_Lm
zE_-mqA(LZ%-i$&0I`@2jrG3)dRLI`&^Okk8l@$*!R5{Il6;hWAdTpgtwZFF&zpI=)
zlLZeNs(C89Cod=7-L{T4ox|_YO2er`=s9A3Tk;yx%39x7G=)@hzF&-IZP~JUaHiBw
zzUBVw@Qn*Bu9{kqa!~2U`Q}^K;8NS!`N8tFr+iYU)n_x@+E@4v=G7A=2=;3`zSgJv
z;}aYN@@1(9@*jqGU<D|FbGF4()%^tR)vd+M^MnB<WTm|a>KfqmccU#vHLY(SnT<!R
z80fWP5u&<9mCG;zZ3Tz@CxaG_&1-u$YlIv1{bifib>`pu_v1n96=udncP_c@HqNf6
zF$bZs#hUr4HXx{cO*4PQ0>JZyRfU62yb{B<v{hbeo4Q{Zbfb6Ch_gB6mOnw9WfyB^
z9XY|o3%#>+HG|ui&i>FC@?93LD49)l;Og;uXO;gzH4%TS{#>EY#m5ZMD{g+mP%-6n
zH66+BeF@qC_PISXgr-LA$y6Wmt1hD3=C()9rM7ifAzf6)_m?Lvg77=6-v{^}T&KB@
zJ9J*^?$5|EyzY)M+1Sa?rDFdG;Le8K16>ak-;b3!yV!OTUhmwn^v32+5LdVB-rnjD
zW)=Hvuc`5F^~IbM0o}uv5)!<3+b}DyVAem3Iv_H}0#=Fesws4cQj^nQ&N*VVx{^a+
zaZ#n3?@00G2Endih}cyjrK;`TK|xf7Nh1XNMY)@(003n&W0rri<rWwI_yL0|&a63d
zxl!_=`4PiYLi!VxsH`WLOlW})+(_E@iu8F0*~MszRxK6qLc`MYG)L7~4^e7wzY4SC
zJDIPs%xxd%$~@+wx+CiGrMSvyuN7*!dS4}Q-uk&Z`E<F+vU*^SnLs%_o3QxM2(ofs
z4B)u<yzSP0<8xVNtvszgM|h9F!aBFb8b9TSQXx-tVf)Uo!B8{j1hKSheWB(H?3A-a
z9^xB@_IIIpMb9{@j?ItQIBL$XEL~~$k{bt2Vo7t?jB*d|nC?Hy+Xxz37lNimpW@-_
zkx$n>GmH{i#@>T&Buka(6u1}m^mqY1sT`TgJcUgN_j4ooQeDbC5W6lkZw*fD^VUrk
zv`{$8jY{S~)*{Tjd>S5dajcKW)kMKxUO%*tT&(e{t^=ZYw*aULp@@OIq1rGdSst)3
zG-k8KzK!5wJdoKEZ`V((pG0t0jDen=mt2YBQ)dRW--;9U2B+cA?wy2ZK{a%sP20Ww
zy1z;nc}S>tIP-Td1R4OE?}=sBfPVQX%Q_y!_{2RL-|{A;>mPOGDID}ZF0vxOBunCD
zPE|~MJSoq)NkDAPjEpT0<(MZD%xYq7sv0~B8+p~8*^suo_r7xRE%583A0Id2E&Uiu
zsg!wM-i2PrcCSWME}l^w*6ZG6J#A{$EKD>o>Q!GeGpi!3bSRb`4`#L$)Nd0&Wq03j
zEj~PhX3U2xr468k>heZ-KZZF!+khhZr%k#?mLDhDyCj`r+x0v5ldls%b(rw--FQr^
zl3v56V;;faR?@9zYbnEqU5~Bo`Q$GdW2MeTf8`A?5mWP3DPhX56`Uv>D#U+nLn-yB
zL#Ie<rpowGrj_i{-Z7>~3x9=;k#k-pb=YH`MGn92U*OHrYMMAW$+BlhxoJc4hy$BN
z!7erL*vF9iiAD3^RdM91sS-C~g4pT_E1$Tl;RkLG9b0Z!+kocj`tuG}?5eI_Gi-!F
z6jr^uH6Krpt=nfNNI4YFo;<4AF(iG2Z0;+tvGx4ia^&o&8d`7qI#{(+`omG7s(M2v
z2gI%B+<c%BS<99U^XD`g*3?-Gy``I_@wjtqP2l8K8q|21K<uB&bEl;_yH<?%w?HQw
zD+31$66yiU(gQ`$Y_-!n`0^qQD_h#mux;PisiSP|FAgg8y>#zS(FLB=*hJ?=5$b-z
z@aX|9(w-Ic_?jejotl=yN{k^(op*T@Jj7*p_~~iWKPr*?!*JR=C-U+%<o{tav@Qj?
z$`}ky-UKbXR=wfE?u{|P3;~-`gAnwUk5;Z&^@bwdh98y^Y_EKB+7dKB?AwWQ*e&d}
zJd_sAO}T|Th;LYV*4XsKUdcRvHp#K@$f4XSUK}NT0VtO>L;S{O_PhYkGrSQ4wz}C^
zA-_S|^_Aa`yWI2UzysXyBvE{^XWMtV5i-c6EgNPSKEmN(lboZO_iFcT*9VXZbA$>?
zh!<#ZB>*O<J9E5vqT5$rf&IHBslxj3VP3aVql879?vxptm_GCbFfLl(^Hu(V=z2_F
zv}Jd{<<{s#MpS8K{f~E`9S6h;DgCzAKBKL*gWH^biqD2^7vkNsFU*TQmpKs$LtA1!
zj=t7*cf9-+tySt=ei9{D^9P<wi|Amv+NAQqpI729=Cj1bff(wAP^B5cW>@gr+yqmV
zIYn)olBatjUJeEFU^#n<<)Jx!gV>R7N2#t0ggyNQJC@_TRxDDBnrzRZdEvRPeD2JF
zA`+b!9#gMvl0HQ5k%R#Qwr8wE6JXw^p1ZTPiKJ|Ib`yyqZ{EI4-3Cjz9mn;3s;ad1
z6>?tLkh8(W^0z%t-9fp?F=Us}Dy^b^eqsZiF@FkV$|(`3JUC99N@j%dNj5mjl2FNB
zcy#mdnQhl2UjWows&%J7@lIpIa3?_2Y+-my=Sq%E!YejSdH7^~f;B0ZwXFM+T8!l&
zy;Z!(@}UB9%RS{TulrzSJM{R^H06<_2;T0~ZJc&}oz8!xvGnOZGM)8zz8C$YmQp&T
zr9tR5pUhxifvZg<2rcMnJdo@5%vk!xGv6*a7^^-x)_H;jZstadDOR8QCN<m`>KHWY
zcv6ZDT%MKQrieUJ_XUt8mHJ1r8;9o?X+suHwcq6&k>GX8jr(A=c-KT%O`RpFZ;Xjp
zn<lj5Q#y+_*Ye+RSR5M6ZHJgbHfh3Q_oAS9MiG)EwM{Q~Qqx~<`^<2g?{_QCFNu5O
zhQqNAF=FL;66%j^_|9~&I!KN$pjFe9m+dWWikVgx73VTI7YDD&rM_OUIZ-y)pJP;J
zw|Ub5K^5hxh>o#@b6LdDQ7V%ejx}Kee&rx(f^pNn{NbX_Mvo9mF27;8(PEop&}mko
z>l8!_c;~CV`O<j8o_0YsT$v^f0xC9D+gG#^8Ce^7{lWy&uz0rviL^4PS9pT6m+cXG
z1}ULSSXEX4qi1L@87<*>^Pz$)cyaQ8L~6xPo#@2p(?(-OZbRH{O?A0BJ%a5Jx0vpz
z9mN3N{ycX>Ka_-RT=UlhrBO6fg0g=@FQoH(&*%Z9)??W2#!jT5U6q_|CYohW$*$1N
z81+hR`f81JrU?A<tJRjET(|cAg`3y4H*FFnr|>n>x_K|%&F9y0#7OMXl`5po;{pn0
zM}h;J;aWWLU5OEURd1&DXUweR?j*{|{t>4RF#?O=XpIr+q)UXTN2gBckUaPt;2%qz
z#~--frXZ3}$<fnF`(J3Cr!{Aom4d1eK9jGmuL&hWBGTgFx&flA<_sNEm85NZ^rB6!
zeRzo`q~V6q=YYsy*vG;mU54Gk0GLn&GxVg?$%H404QEHjbOs!~9>ZR32~sJQ*b5f7
zDWI)K3bf|bB56jM)nsAY&zonG?H99UW?yhdCOWe$9$x+!Xh@9~ou`#Qg3SX=xn@pq
ztB4O<+m=1t{?v(*(@HE=hgYM}OGBkmLM|Huy3;#decRv+*y)7hEK+2y1v6?@(@Wq;
z|EFw84_~vZehO?c^HDe5-Uso>B?lNRJsRQPy9(dA|3u(TzddY?9kBN;VBkm0M1Obw
zXXAOD@&PV~ck{Q3UcOTX#wE3uIe6;fP9R2rC0qB{nk4`l5<~~0a_U`6I~&Z(d~(Qi
z=%SQ-U#OBzW6Z5*zMTF_H+J?v2{tt~bZRJm5-6;JmTYU5Rl}jx3mG~m_aFnYM!xn|
z`Grsx`G(bo>T&3KUciBa=hZxSrT_`(6VAsxHojH?rnUCO=PZ=>nm%tz1~B$(>uqJR
zVmtEU->WOnJ4REB(>HZ1RbTo8#qfLQe9p+fT2m-CeH8puv~cU%TP0rO*|$9Tpo?Qy
zLV!k6)fG8$HtIO0^6HLT;c3dgD@xkm!uXS1R&2)S-r2jbB9oLtlWmQc562Pj1SbQ>
zWC{xmyiLYwj|t}GeCOQG=|9?R>ng%gW95S^GC>UvzP0k4)-}1tfOnc#a@-|r0~T$~
z8l&CU^tUU!Htuqs+`W)qI9;*N9W>_Ql`Ouo4_X-#<w>!sLg?Xa;w%wjtISre@8{gl
z6xnT~YkZUKPnZoAo79V=tMXoV-i#wTUjTIvxbLq7_pYt#_kK?c)~-03<WdUXYW=Tf
zLIprjRVwDao$5a_lK+55f8=2Pb>-5>$j^bq6)go-EEhQqor#1-Dc)WZ40uqoc*$bx
z@}4&AP?;oaQB`ga{4)7HET&fpQ&WrbM%nac2&q*0E^H(&ZBO|>EAyV_HnY$a#S_HE
zSyEoy3JV>{D|-5U;C<WjF3I5poW(Tb)4ay<P*$@=U!kb(j&_UQd0Dh%ilQS|c<-FP
zmE1}1(*NGpKV0^|H~HtQ(mSliK9$A7q$Ta%DXaP{-(PGnsaq2YR?K4fEB%oBiVO)}
zcexNxHFZ)TAu*xbO;4TT#yXc*E3kXkW)F(flm~7q9zL|x_`kNC{|lx6yux3uYGhe`
zCc~3nv#1Og6_uXGFTG1qta!v%!C9NgtyGw&D}9!sB&oe*=8xA-|DQeo>5eZnu9cQ8
zKXu1b3BB!a5FQG!mx3C|50GQ|sH!9I1Y6xd`i1_Njg>NP)Hj$I19pNkdDnHEvn3?%
z=55p~+RUJp!SFnZ%?MsP3_<VPs$Sw;V)FoBE3Kd9f7u|dKzf+9<=Rkn`-Gp_BYx#A
zCm>Mj>{;r%^9;H)F#N*TV;c@JJbKhCLc$;C{<j-4(u*6?dr9=+?xUK&j*YR}c4`lC
zn@*=*m2YtLuCRS+f76@*lU14tb#*R;3%=*GuQhRaQ(c}bX7@ktn9=+K<IJl>@h7G!
z?)Wk``$cFas0c2aIOn=u|Eo)lQA79LwSfPPKUZ15er~ys?*N(Xn-Nr^FC3aF<T!p5
zn}{q)1C=Oyq-RA?imq`hRa|GM{I8qay}CJGD6S=tfHDNRZN`eayRf5#{BxP-jd$L-
z>KJT4Fo`&OcK>y>Pn?_0{IKlFsB{0LEX`a{KL$P1$0O$xOZYOwm?Zs3H$p|u&t+|6
zs<DvPId4?aU=6^T{FsVqciAl}E|#%18cPb*;c+x{G=0bZKgif0KWP#8c&|51U;a%j
zp#eDM<>Nby($Dg*y)f54RUe;RD6TMU9^evKVVeM{v%QU9LJX8}exD8Au{}zg+OIB*
zg@XTOZT_SNfAb}W7@e$1AFxQe^Ve-P*~V#$3&Cs*yzk~rlbi`Yp~#HUF88q$&~poq
z#!^%?eUZrR^UAwz(jayh9Jh*Rt{lWmASWT7J-LlcxLY5eVCtIK|Ay}$Dd)dg>#wYv
zxvqW;`7<qK@r^F|c1mM7pH(}jtV;t&dtQ(R1}E2qi0(Hv!`K=2mE48aN(K%szaK~d
zwe|+*A*D3-ahi%G;^-8Mls_}p2tlkM<LoTS--M(P;5?J~n9#v_K`>=__4xb*$!}+7
zfUVcSh#MHszWo@c`jY>D70R0Ttg>~l(8ax*ARg8ERAri(%eka@wy4aNE1%~UDJVGd
z>dSC|cW+#8n=SI3FvdnFO@OSi;^S<HRZz4s#uz*6z`CY$%5QG}_0g}oxbpRjA2!!;
zaTfpg;DCQp?FAj3X~>=8m{74v2Xsx?5|2tjQn|*js2|<DbCTvw{x#hiI=nve>){8M
zENDdBaR$~x|AX+~dCeh22Sia8IxR;O*EON`C5@y<?k|pb$(GBn-1j#ec7BHG_sZb?
zYlw3^-gmvHcV1LH^VQs#dmvzX>2CyesPqiYp*-o;%W|75H}`E%GqUd%uNb+n+1&Ik
z>-%F0rfDy3DU_EyT#~iiVyWLBkQZt^yO3DH6eyb48(isM!;sSbe?uYuBpj*XEpv1C
zcR*t*Im93%#b#g1h|v?^PZ=EK8?lW5F0Wl3g`=o#5B6XD&7NLpTz4523y`a)O&B*$
z!;XwMF@^rBY5A05p>?)bgnSAa=TCUE9C5l-@}u((>$vLd=Jr!0XW{(k#^-qkY75@z
zLnlbzwG_qoeV=&2PZ$VT&^{J5l_HhbC*qF#aU<=$KIN}=fkS|<M@y?0*|>G7w+BY3
zZ0$WkI1VGPRVR@F%G2ha#MUKWU4L-fTHwhP)xe!j%P~p^T4lI_kZblZQucQy1%;mp
znhI}&V`f}XAc8KiP`azj-D3~<8b5TG%;9~POkg&mJI)-LxeMaeO2GH(b*5Rg<tX~!
zwAvm%#yeeHJ1B7Hu>Y{JR~20_yU}gxKDV2=^cgh9`hR`E|LrZr#pwg~nIxQ>+ENtH
zcu(%#+i{-W$1OF5MurV5&QEB_1g;A;cRD@&;G85jQE01jXdxp7Xj<&)MU8iW-WvR_
zI5TsqGub=gGcGS}m?D!jTj7IIDTa%}>Y=)0+YqT0<(nl%#YC`B<BO4vG8sZ7xBQps
zg|_1t|7K^*^l6Two|apbJnt(z>qk^7j}S|URozb7QnB;Vi%{sVAc%hwUjh$z2E*_F
zrV9!`9z?C*G37*<iKa>pA)F0s$+~F;1-#Yewp*=PLO^IuTtT`Ju3SKq{7FE)NpbKu
zjV!HTJ-~65R48SHR`WnpWJe}_k-;PClLsLRG{(wJi@dI1CIfg&D61FM1^<SNJdrJl
zdks&QX5Jep-WB?lzqjhG8KbxrM<zt1pMv--$A@FJR8BOtdy;+n+d6fmSWuY6I0H0Z
z`v!wzFuf@F`d{UYHY_Hnr2<A7*HIfC8&&XynN&kV0;ZeLSuPWn>}2SEMB2!0$vJ_J
zj33?s`#4k_BCP4V7t;(m-QOGUz+O%4J6<1;&5@}fmj{!hu?nqkppf4Y*v#blnR*4A
z&FPD<56<miiT9y>DfTX;r{Za^bJwHdDuNpnC4M~SD?FxMo=L@ltQ;0LIKf$=Tz3x-
zE{pm*yPrT*vYJ($uKW$?gr5!yMqsZ*asWzX!RfW6Pino5II?0B<*fIljZR|75{@BK
z8bJJIltdORT6$U5{~BJG>AO`?0&=;bv|rFO)c66OV4%7}Op%M4CghtX*W&_z(8k9D
z&lB8LHb0&B1I*6Dnj48Lw#`YOtO(KTN$apdPdndQbh9T->i`X?m}g0#!l-%jL*2md
zS;;*tSS`Zii)mFH3pl@-sHL51%t*s>hlCVC(Z^#U*YUwi*5CZXi_UAMwyyRsDvofG
zAKkiz*?IDQmF$NgME5r|t?lL%DpQ&RKUFN3TeLZ6`0=i;g$FLJjxl|TuMILWr40iX
zeQJbFR?-0Ve0VLSi#BttPYxy%-nZEvGx|-F!cQDG+^KlX&>6v^vhx9Nci4_^tS~7o
z8YcFOSvegf3*JmhSl+{IH7sev{aQb&+4D7iLPQx7H=|K(@LEY~<5D})ApoBWaGZ8#
zhW(0hoB)YBSh`Kc<mO9nfp?lh85(rKzv5oo)N<TtJMZp8t&(XhubocO4{%i2n*A2t
z3-|A;NU9Fpj9<b6>U`Hmv<KN-|752Bup=Ai=tq<G-sE#Odea6~))y_lTBe)wEh|j&
z(jC&7aNGQtm+Nhg8Vq-?nKjzSO;2qKT5uW~jwKD*8V;`UIt}Z<6bDi)t(@5Rt@O&(
zm4DMBFIGBPUrTDHZT2PU>5W3T?4&t)MKsy0*~{B>3*YlbQ)M|H>NvgKVI&3gUupwc
zHCGdhkBGBNCtyutkQ2MN5#1Wcr<aqsJ<0gv#mQsKhoMlJJR`%TPD@|nNB1+nUP1=|
zKuxVLL8z&}bo(R67$Xah0t&p^=YItU;tkw8?>@=zW?HptbvPH(E3|=D16MzvE$Cs9
zZr(L&YWD4Z4H$9VPTvU(Sh>@i*W_Vqn)u_BFkNKkdK%^HQp>bbub34vclZJKZ=Onl
zCrF<~hMX{(L=9z@=`=FB(p;AZw|5(V+GLI6Uc98HQeQTGv${UN*x0YS_{z2pH}Y&C
zC$3r4IiDJDMIovVxNUQb0i}vIUf$K^90dYu+4B0NLRmfs9Q(Eq@jKkx;HJmbtOqT1
z_b^9ZrexY$q7|`QP=~|aEEs5-++~(L@1=6f(Wj$oEZhH_L}=R4a&wcA94H;*I!5hP
zp|oq<HR7I7dC`X2p-Kcg?L2hf7cMoJ@OQ2oL3oj$?eu%6K3Tk>Pl<D$9HS5j5C`kC
zch_cEY>{V+Ex=X`dB`Jz23||5i_KkkH(sV8(|0$jEp2iVj2=r4+@VG1eM#K2J(!wm
z;vx*9Msk8)w(@TC>@D`5o&s;USBKLkTl`fv%=sCI5un1my$Zl?H#wL6o9C=PDAZII
zwh8j;77c*qRx3vtZIJ=!++vB%Q~w;;^WXYhQ^R^t0fm&TCa>>gjzJr7AC|@rNy*7D
zl2uc2Esaj{qUM6vwj%ri<=C-0!g;IP45vliFS<}BeF-PJndgMi4p)l5a<YmAHF+oX
z3h6-Q(+(n$ugdPeOHir|bguB13yfI_N60VtAj-d;lYLhb*gY?Ay-C!?lMY(C)Mm%<
zzNNLR;a2OLW$kD#E8@HJ%Z8yMlE1kjm#(WWuRu2aqe&+|*5}}OhBmpBPpE1QTSMel
zTXWquBCsl-Zm2sM3;!M5<DipOJp5&djg{eOijxI55zw$9N}JU|#uB`lc(=D|Y<%A~
zEkg)0&q1>=GgC6s-YT1q+6ZqAvB@d{3esUpuP_W^CkOz4f(Hi3#)2+_;6!~3=>|Rm
z2;X`%4CfBq5Jp3UkJHZRjuCEiu(O6)z2^SD^Yqvx<cN<(BxHc6M>0Z4k@OMd?`_?}
z7WUuZ;~(6pkz(yv8lmD2Kln3GDd?3*eH6yq-(L$F^5t@%om;vi3hD?`3x!6e$v3`o
zE-cQiUW^AduR;)mw!Zc?v36H-TG@WrtrJ0*z%ild3=9oKLM<Y`$2&+g9(MJHr+F5G
zvi5V9KubU5Jm2gpz#R6H<h-W0GR%@xh?};5z==nMOCxyNz#b=nwncr%WH5Mh37y21
zM$PU!!)N`6{oGJcg2kT>{_3B~RrGoD=azo&vJ=4b;;i|r`dUda7ROMili|;+n1oC0
z$F|*pd1$-WgD!FW4pC^(YawN;SRUbrkgHTQ7j};$(?{l!>pvv%!gsnG{TU*!>H-oh
zcQK2aW;H0LzC{KCS{cQVq(&w%r6>YpTBgT;8#Z^32EX3=9tQd?cMi$eHF+Mjv2k^1
zy};ODvi?vkh`YfeSviMp`qk@tX@r9ROw=p_G&e7pcV#m5i<aQ11Q=Vgq+ES6Wx@4u
z*z(!yD&#ymP8vq+mPwu$(!hN2pSx?6wDF?tL_W}1_4!$$-Yy$$uG6#w0jW($Mzz23
zmKxWePABv;%DEo~?oU&V5DRs9^m2+al-zWO(D&sIw;R8*fPQQ-p-u10qeyzrsJAU{
z{OBO=ADVBE9p6^lRtq?DZXuL%f+)s#_1a;pCO20YpULg(kIJUYADCqFM_(6s%-?=Z
z&if5+a@iMFq-zjZZv-!CtS3*pV+wIf76yf*g-!Kd_^B$dkw}nL{aq?{=m2$VHLwud
z(MY~i7%2ElDI^&8jcY5*(cjTJFnND?(hoIes++Twwb5@W+uifn*AyzoQD^T`paF_z
z3wm(%{3H4c|NfY<SDT4}QC|^41NY-_WR<Oq$Pp#Dsf)Ezk>M8PRdV?e2~T}Qz#Jiu
z8u(Uj$9C%+-@k{_GLk^GSu5B$u3&rhQ1T>C^cg3;xhVyW&Q3N&Bee}zeRObks`+TF
z+4FwZ=3#=BaG>kM-fGD1iT!}II;~b`@S3iz^>bfjc!I4VW7uKY7-opNu<xBO--R>@
zG~KM67Lz*PegFCwMt-`md;j@RG+^9-28M3%0=yK~vv5<vR9yEG_{)33mihSjO|0S*
z?@I6SsY%Rw2@r&&2Yta*Sd3WKc}%Z3^7tVWEd?gL5gHceA|5@4%od0^wc?ikBjTcj
zM;4po#FUC~;Inw&|E%@Tl@$XSOWz$Os1Q*&qYWNu3@K`?yWUrsbV%^viii_WZ+n~w
zAO|&1P~k_;Q-@;Ly?PBP6a~y-S>(MQ(J{CO5bl*UeC8V^_ardj2i-GEq2n2NVmwP9
zFtx+qcYH8ncc~XWwcJFhdKL-b@IvtA57T%sCVv>OJq@$5iI(^ixBlQ=+7o)_&LG}*
zirGwLh-zuby>mevImKJfg75A*k4qyR4Q*9>i&-z%1$x63lofH2;;9~dXK58LEOlRw
zlCbYfowgIib@qA;&E?+z(rws(ITR=I+Zq>+haM}m4L}+mY!c3&-K8vG4#az`oJxF%
z)y>$fMTpYIsq_agmdH(uYEyx`v`+H&OX8vXI9`O>YticDYP=^P0K16WGb+)nOmyr#
zjycGIH|x{M(`SN7`%p+a*|wR|RmT-{w*jBSQ8xz6VF^3j!12VoQD*XM4eLC;y>rBh
zoMjLpj;q?HzjK7hSh}WO=x7XbN73D;H@oPg$0A)%dWtq9c-I>NtB-B>GSSH;zY-z(
zgA#(_m?kuqhsTrbsF5?p;A~v0_GMYXzp*d#$93MEeK=>;o3y_O89yEJBpto}_^YeA
zNXV+=)X`>z3WoZnF<}``+4M%U0W0ujFGUd?5>_itH3U3(XdUixG}9UQ;!PEkUo&a6
zehNvMgnPmcci=ie>Mn+I7!s5gZ`B`wC8%K9@3EMbLN*~1i4ii7jhd*h;T$TA52&L6
zs>#N5c9SEzwQ5j>6oOd^*kK_rno4SF{g<--pX%B&9B&NOFhfjvql@je?(XeR)TTar
zE?$R(VrQ$}xw863DL!r@3CE(%y6$w&Fw$qlz6s0;OZpqp-eF`-Cx!tik+^2+DgfI=
zm<McNB}Bo8i_TO%f^Ohyj5hDIQPvOL2*6QLvoyzoYUdx1<{d~0kr7no?Yh}bb-0G?
zIg_eDASmV+$&)P6O%czlLH>=g&D%@8I3&iWFxCv<9n8DM`m;OscRCBv`SnF;%o(H9
z<V){GL>umJM)Dd`TH3ro1r-s=ym@TY$q+2AZ&r0SC!%E_IQQJHqVPKr*k5j>7{XY(
zy$Cq8Rk3_08Mw%0g`IZ9_nX8XwmX9hY9ml-W=s{0Rj9_pX^!RRjl{@CK&K%?yVMoD
z4bN&*0JsRIo;KkrgP&Yi-XlsNFBEquczPjZdJrIdeLz*^5zh87Q$bcSY5vkIlNo5(
zRBpGRF;rOpce#Wy0^!JJ+gL36Rb6SB$u!)3U>dDPuETz9QFFmpVJu9&Pw*%s>YAh`
zDx2$=%)z?4!`jqoS0?_diFBLvJeI_4p+A{Bnk3$$JDO#I7<rFi$&?_%F9zFJ1mEVq
z1Ps{r!#V7f0zs&9z%Gk#oQ(}%XMfVP#G>}HiBZ8(0SLoxL@}@TD<gLpm3U*eTa6Sb
za*}{RGPL>DH-ZGP3VGNHzWUMoVwt3!o<pOCUDOlAU)K(6iHT@Y^8XeQCtuWY&Ay#k
z>VjEyRON(5GdFHpMff<+$jG6mzwT;kG9lt9y=dXuTlaY<s?tb6L-98W{Pw`8Yuq!H
zz<phnAs!D6&XuinYq`G|8XrqmpS>K^B}EK{2_9>3E1ZP%a#n_POovo>yWF}~b|be+
zvZGn}`7Z@_R=>rGB7noXAhJ<gq<GM_`CtUMj^KHxOZiHQXK8|yM_HPizZ_8_xOmK}
znR2G6!8jJ>qZ~PQx)%WOUXEOy&<nitQNOnrX0!XQZlBxl+*@QULaA95wV+B<?wu=i
zK>&^D?D#=bxrBq~4peS?j7(wq_t46wve5{H#?(WzfwsqKJL}0)duNv)b8(;HZmf`*
z$|OaC&dNv|9Z)(kNBuIMN1Yr;QYVynV{iVIw#c(|RuD&NS-0uwik&PBA*>Zw(T=`U
zedgoaJ|vX{`1O;Uo7;=9zorjrnr+wO0OlZn#?;{t4`jTGn(TAjBMel<njjTpro}rv
z7>XZ(B)fTE0nca}dkFAn(F@1p4}TIo=yOcN06-N=BLIK6L<r0zx8f;F>{^5W@&5nB
z0y82HXHu}N_gI%RLAALrdeg9%U#Ns6Y2wk>eAKRL0D&m=D-omh3E1kWkWY`PONUbF
zCsdE8av-XO%Exh9MN338z`@f~exzfM{r*DO&6|Ave$~BDsu^a2`t9UR>B$(|?{Cw)
z0!i~Kw}`XHsT_gJLsI}M%IL?lZHPZh-*VIL9_N*~b@qPZI?dTS4Fw3Fmm-!lZ%0gF
zh=itkg$-ZlzEDdp%ZKw!wDa-~O+qbTmQ~(6=8Dyz!U$%CX=KF{QjA1EF}XdEx++H#
zi+wi&hDL+cpIZ>ep|pmImhn#I&1qaS1+6kpodO)DZD-+eR20D8{ZsT`0S7;w?j<?~
zq0Si0J+65JIbcTaRBDv~*KwZ1-6ozDiik^=j@8rzjuB_qd16hOr<M?-ZUW+1Cd$-D
zL^h)|))Y13e}`O&4`+_ZGJ%7TPMJV8LKsV<^I+#dhx+LzFj`|816h)2nxN)@y!XH7
z?AT~Y(%qrAe^lg0T*Od;1Zco)wP1PErhD@~!$Y{$$u<RG07;`6PCjo&Eq@&pU5Ge2
z$tC0Db=1>*5P@qZbv56vezoLGt;Y;Bjv+kBsk%qL1bZ~Elkn^^jJlXSB}21{atNeZ
z|7vft_tFaF{SlTUu+va1b-f8qTF2Dy46rohk@1JE&^W^RGJ~m{9fhe{?5dWriOH^6
z;BJU=m8R}*9x9JDsJCkGLd&(4a%WNBzTz<*^GX9{eh?}*{Tv?s@-B2G&VPSrv)_<D
zQC5;Zl>e;JIm$hKhsG=o5EmS>^7`m$+(Tm7!=y-BS9sNq#LZW|`+tG0SI_&rJO2WG
z_nk1M+^4xQR9%(ZyeVMs89{r$kM|u9>|oO|C)Gf?lYNFbH#Ut6SUNh~!*S4jYBGLK
z^^nKEb;Yb2=LPUR@_$1oORsbBDsj*Vvv$~dL`|rZ057T&Kmg)B!dx86`)CFkI!$<H
zgh}188D<tl=pF1s8<jhEkb!#yh3SC`;PsDtT}W`*_(={x(S3%!vxsLu=RMo#G>?Xk
z_C(paxtJe`t*dPd{Kc?`wXnTyp(amFbOm@5yW?<ajY6Rw!#rq=X0^Y^I0!z*Fq@81
zP+pN9ldkjMKQ``5?%iBU^`AP~quKD<*72AB{4{m@9xCkQXJ%~<iUL=~x+^1lMwY~s
zPZq9V6Gq`UBEGjVY<r&)@;l{5&%e0TtwPMfU}=v2$dxcgL$kOn_OI7!RBwGCta6*(
zS~<i85Jv00>yd%Pb()Hxth|8E>R;^iNP8*=V3V*NKFrP*E*Nf!w+z^FJFm)dJU4L9
zABM#xISnF@I`ZnEfqNq~KntVJ8S&+t94**3k2k()6*3eM8A<Q8`c~%=If*<X<<(gc
zWniK;;K^z2-HD!=XD1XH_W3c|XWu|p*PXp#B#M<+^HHrtTucApsp#8(@g6iNe1=|~
z%h9I27eVc-aYjFMM*Djk2es9Efv<mQgJ-)dWR=?_Pk8AvgEZd9I5|(<f3UCM0BPe*
zufVYNM`{m{d{|_y2kHaehOu8Gm3pehTq=cm%u2ug48XG-jn7%xzXmaQPh**@OS~Lu
zr)owW0{=$VFCPb)A)Y+7O>`<x-Mg@rVe?$ckk>^{^a-z{#-Pt%%;6WtbF^OeeVF$c
zV!v)3K8!8#UDlb5);@k^Q5ZaAxR5g5_FBTKuD_kIwrTdR7%hGC>4VRqsi!NMxXZme
zq#MLm4G$A_4;_u`t*BIth#a6c)0XZb`)?Ng9%G8<t$7p1?YF%q^Vb{h{Jw7?uAru?
zES42lG*$)tHK5PH@IIprY9_qN!=Tr_7APF5S{H{r&H2&`q8H9*|KCB_pF^@gUn$x$
zmVQzX!wx^U&FK>(o83)qmSmzUzM=3#MV_<yf!#;@!8bfzzazoFzkf?bWXqvzL>TJ~
z$_@W3Ow{w@F~9ka`l-S(c<fu&p!A>KHje(MBo~bHv%tBoVV@e0|3rY$rQX!9pCa!1
zpZ>DAe<j_uq;C5$!1PMx`(A6!-z?Q{%JGlbcqaee%*tkVy$!zin)G+Fd+c1j_eBLx
zkaK)`?;p`kGS_~9?rGzIql40V*JaC$sfuk2v-j@K4Xhg5X#hQ-=m0zV^!Arg85xZ!
zJ8J_ndUA4OR9d<Sc>d~P(G%v>{kvz++C2VWH)EplV(Q)91P!w(p0ht$tzWh?i<Au4
zDlwmG1g}yKQsssA1CS;4wbe!kMA>TP8#eFcmy5wKUsq66KOax|ZU-1QysvZo+5j`~
zAxr)Pr^rszpQE4kAB(_IDped|l>bf!{5du7=c^e`jklLiuV$XK87=;f0A|-@E@z&!
zzG##|UDV>JQ#)JKP-;=_ppWkB`?i^1GRffbKn<!ZRwp(fa&ujS-Egq^wJ6(oVy>wn
zS&Ck2@#nMO6ywi=^zSR=6*~0#-7oi>8TW<9X@{lP4StZmE}~rio$RZ=?r0;xisiR9
zd9KDK`j(MBDkfG#DQ7_035V}p;jpMp++MnEXc3~!v^gB7R8jKy;nLsB*?(mi{<*HA
z8RG^mX!IA-ScJ$2Ttb#KzU~w{`St<2^ygVoKSpL^wire{Lb|?ZzndH3t%escJD)wn
zLODZugdZucll-|(LLmcl$$5c9;{7XeFCtwN#LykjVoj#hE~F|N?Y>pb{d>v#GtFsU
zZuI>4^!N&b<^{xhPy!Lh&f$9p$X*{a?JhAdR^z7eC04RyQ}u!UjUHPRdS0K7*#z7)
zb_zEa&gT#B^E26NtM_{ZJW{W_J*ry!-bPqz5OcICbLqzIG@21wdt}kdcbM6ty!HuJ
zhtL2e{uQPE|ENU+buQ#-OYbT9M{h%oyO@u|TvH8X&-^I;Zfgs9{!XzHj_T$_*s-0!
z6JEOZ&Stcm((o&m8tgP>NG)wlrn|TDojK@wj)De5jDhNz8~NID0TDXD)$?JEv$5uS
z+0pa&{zvlAT!1INcq-g1x9n4BHe;CE9_%Wzc*_A`H+Iw5y>spT^GXXY$!eYbdZQ96
zku5;{*rdGdN*`Qz&%pq?+fj^p043V3yIkS<6j9dH6fnp$+05Bx^1HG8duGifs=mO^
zSjxB+I<9;*+!WrNSLitd5IFcoe#<KLPT6_7cFdyLr_W>LlG4vda5uZ#OjQwHb>UIX
zv~z(by>W+{&i>N>mh=A~t^bKgH!JyE+?Lh7iCjl(*<AD>M#PtO3#FJ2S5?K#>H|u>
zKeNfUzIU+bb;~CB&8cpddOm+N@{BzMG@OdAn@j~035H)DkN*90{8i{^RAr|B7-SwW
z+wKIR=<qg-e!8;Q&)6ZXYJ`rcCuk)b5`C28x+KtN0I+df#O^4rL`awJ$(<jrHr7#@
zdWA1PdhJmhExhqPQ&bhI(bPo9`c&5_@!CG{_hiQ3?`S4W<3)tg*v8Z<Lc(&*(wEBh
zshQrIai&yW16`K-IqUA34O)W*mRvvOiQWeNdggX8tuzj6fOi($Y317DMV*0h5pdpP
zeHB$`seH)eM6qJ-71fj%$T)Yn2Z4aWM&hQDN4{45Yw?-W_hw4b_{dR~y1l(R9|MjD
z;y=uVb#)@^_vZDgQ6mX!O2@_XX#GO}GFqFgtTSmdF4%PRxUqyp+X&0jUC~ng4nMvG
zuF7yLDL0(x0}b6LgX=^^XkLPUOc`@rl#ps9$v$ZkY0P=l$h~K{Ej==wnXg?-%6!5u
z=l!o)0HjRwrY}RsCrzwv^~l$gdn#e09R<`*4q%iY^Ft7KU}iJ->%}|8L$|N|%(?M7
zpOIh#^s{&vdVWq;R6YX@^YFj2IHwS@rK%o^uP+T2RJxeWl60Yg-)?|oHDf9vr*mi~
z?DWW-SnWtwgs|GwdK@@hSL#4B2VNyh3^;G$71H&!SSCg%b!_`|VE;8g7~K}33Vz>e
zbopB9N(*7T(Z)vQ6lu3p`G;TSn7m>Cqwme5xv-q6lqlz+cvA<Wpo%pV49sODzMfx<
z;>%C<bDT6PAtHZDkkj;=e3J;RL!sSbN^<~*;Jn2IYoxr@f>z~n*~fqm>8|^PRqHpq
zr1+t@G2kvYfQNln)CkUw_3zY}+9#=V8j&nffT7_JZ^8OpmDKsyWM*1l$>igUE0#`3
z9|Wk$ny_N`eVHr~;C7*DS3n|^X|~Quh$rCcSxAo?>AFg(_S|GFt=cz@Nx?*VY&87b
zp7xG!8+`-jq2|G+a!Bs<jT%tmVQS%6ed-R}yMDg0AIAgmSRRu(J8GKhu0Rib>W5!R
zuP-e_m0e$^W!o-Xd?(J?-2e-vXMVid#&^ZXy!xu{69=fri0-8oRUZ&2i*|fq=C1X>
zGmDCPj2m9wb<_c>lazzi8Z!$^eQ#ST&fI%7Y+E7qzhFtl!+QjJM`={+k7JrDhh3VT
z(pr^oKtlU0yaRi|N%Fczk-+jG$ryHi!CS(TFFOy{91_f5UJ2~)vpxwjsGA5m#zYPc
z?exOmzvob(vKC!5rp>8*=XRpvXKYg>^Qr-H^Ctzv>e*fFrT5$@@S*eKoB^+IdT^m9
z#0{tjU5Fi$wl)HHo7Iu*#5j@U(w^{*LJQwrp+P9{oc&ECV8moqSWC1sWb3iR%jxK#
zYN(FclPC_DW{Qy*%yHdgLmg7=S4k@gPpuY9yk<Haiu5`AyfTQ(hLld9DwQ)#3I@BC
zydFRYWzOsPqArpvize+4NWi^;tcZu;;zO<6SEGlK??Zpo%=D)&mK7u|t1OOzd)hJ{
z_?I1~8&bVHg4g6DC?<QhCQ9T^(MBelmkH9Nk#%j2jo-bgafG<rmmy)EU(KXAxg6M_
zW3D~#m=eUs`=i++MudTPmXPO*U3_f=;R=%~Yt!Cal+Vmve#tk6)VE3LWmi7jJFV<$
zc|$gH<krGi=W(`!+CVdaV^p8aG&s+HFNfm|xRo6{!0Rzk>O0_a6+1EN$RPpr@@ME|
zXUJFF-B}y;YE}K(YF3uWbc%S4MUkH<<ln!hv@#qjn%txEg#v9pK;eXxH(#O=px0I2
zyDwz_oj#-)_VlysL(|Y*mH)&y@ef+W51({M4s=3}sgwxGg+%AZ<W}zFg{e|6OKf*o
z33;vS7d#P<?P}l3mW{E9F7#U$idmv}$qo(+UsGe^effIICmmqcq9L*t?Yv2pq&PX5
z%-PKOKI7yV1{A~=GBqS>Z;$8;Xaz0_<fd2}1{K<`SH4e^HHo+xSf0ZzauSexC0Afe
z1ctag`kL05PXwcOcH;)Bg<J0*1ECh=IQ5mgR2Cgh2tCU_tN-q+o>AO4*iV1`a?_5?
zGMA0_wG&@rm2}xGl%v@AKWXxBJYx^_h}QK<DV=zDfn}Lnvv#5@>$6?4`$M<ELive@
zJEZh2ix%%=i|?`^->le(-;IKAWk$_A_t;KYL(89q8v;MIy772x+;~XpPXNP6K8lS<
zTjY>Y<e_@gHI-S51ySm9%FgTQo?RaAt^Sg70_dj%lwonyb-QkuR#h6h>yI+@<urkY
zdi-o!ThKEq@}D+dQkZ|qKRc3VhMgwe2q3Itvs7+QffIzMsz><vx4k=B0UPC}{DO@=
zCrQcdu~)IQH%^2N-EBhgle*X2MNfSFz2WtaHi%Q??y>2#%GKFp8{8hAIokG>EDdTO
z%+tDMw#-ozYnC-N`UP_tZtjx|TNxUs%?=9RL)#-S8h7F``NwZ{@pyDD)@27^Vcgp9
z6c6671Pzo})*BCBUxp?>?-BIWt(Tg3uJ7$(>{?T<1PNtk+53puv7C5Nf;i3Pb_*JI
zkKLZ|{<Tf#;`HHzyyZrF)trZ2?APM1vVd*50?e)D(Em<7b-q8w;bk$&L2dDQBA$!-
zm*1n|%SVjE{%=>O?uno@a&Bpx^^j)8WgL1cwfDNLuf2=+aMs%$FN!(H=19v#Ggf)t
zd~`z7;B`TH+|#`R*X}b94WH*49QvQOD!Gyi8$zI>9ki`F=gJ>m3ZO<LYY76hvYapP
z`145VdJYd(Dqo6}9E!I;U28LRqrr0dOUgt6yg)mLFGRlK4kY}^YuBheU$HHbiLX`G
zA<|B)(B<$>W96R!pNBhsBFMG;;t_R_Dq&&a<M)RFKpun~#2^${Xt6voJn<ay_!nv9
zmf53~OMwd_0Sd1cQ*#Uh7cSMUsr(ogq{guGJrP*yi7w6+uyCvb#V>h=^~?)Q6g{sh
zL}Zw|b+>b^XR16BZ;sPS07k9rf*`~&rf8I_<l<*io>(sC+<_NrU$Lpk3#Py16Z1eK
z&ecL{v8vahf<7!ZMUbk>T9_AxL#IcbfnB*_zp4w>(E3@3o!?WWbp~|SE%V$s>x>*)
z4qw)jo7gL_T-BviYH(Ji8tAmJaxu`0VOv>He4FBd4~E+4hoQ{OS_(mpYROdEjPQvl
z4Vl)nK9SJRP70}CC?_~`dg-l$Q&d(muetj?4ZeB?QI`3}I^z^l%>?}O%bRyuudN(=
zq>ec)ME{KJY-LxHZIO#?UK@)7^++Taea`(1Po6Jcni+A;(r&oCX)}`dlDk6tV*3RM
zn?#{=ZhpM|j_9XhTerw4>p1sw!b3ubk)NS;M{TF5Cb2$dt+TSX#-@>PqCWg5|B)t1
zPs#Zj8o_QQP8+70TmK(y)E|DVxr?~9QSB6**hMFx)U}4^{K3N1H}DHezR|_Dkk(*t
z1{cvxD6+Gm#|~=XVBaMc2Rr~^4Ex@*&JUCa9Bjghpn`>-%yw?B^?49FH(I|j3ZSqV
zW>mXeC_c=Nao&2QF6R+vE2PI?1{PP+``#*Pe?g+Aijs2c!+os;<LnsH&uOMhG0u5n
zqZxa|Ip4&Fx$!$051lN&etvi$B_e2H$<D<ek!qTJw2S?@KHj#vm;VH3BglO1mX6p)
zH-DDCa#q`dtU=Vya#BZFM-4tt-=L^KN9R6l%pl*hJVvugmN!#nO=3B6Q$oIDIch`G
zxwYMso0G@{-4%f@E0G#l4B~2RY?8?ARv-9A^PW}EEU8f;!K^aF-1rMT$=Zx~-S>dv
zy?H%=cnYSV7~_gVXO*n5d3L0)bT@KW%HrnA_9?MZFEfqJ93R8ow#a@r-G6=99P=T#
z4JyPEwpeBbA$roI?aR<`J7vOhNS=flt|A-^oL7h!Av08=wtETsXh@j_DopG`D+_{g
z2k#OUS%^l*Y?0Cm6V6EBKPK`US(`}@=7isE26t_1ah%Zx{SmD+TGKe=deh{5+vQe3
z@$(GG(dbmk9m^GGJkHC=3jx|GU6pJ3lo76we24Vq^}37z%D}>`64eu{l^L*@)2(7(
z5pQu8$S5?OnzcI^;(k&TZq%t)tHSXhM)7!z_IgOiW>O>9M~^l`EsHx-*=~(Da#k-#
z@h+~<`>R3)t^?geoh6{I&9cgd305huBn>U@;EVJY5|)YiEMVZX)aL)k)LX|j{kCu1
zUlB!=5cv{PlMqB=5`xqa5D@8(krIM1kj{-#ksRGHq`M@fQwEG~HhOgDfDQ2BeqQ(Q
zxu5?(|6QN!y3Xr3kK=t}lZCBW|946L|3xn^k;mpAzxU~Ketxe~{HjCm#xO<X`fUZ%
zFGT8G`sE;mNQK^B7OMBrD41*8COaUMS{<o4$xR3a*yUt^bQa~y#30A6b!(@urGK*W
z9c!KyusT_6UE2gD3LNHgL~=(?Z_*(H1l~4rGFt+kKI*Cd;-@e#ccvTFa<x6K*wo|Q
zH)w|M_NI!%uY(W&0E7<GfH}%odU5|T#WeL9E^gaZ48Fe%=0hmM(GKpMUkCC3l^X8i
z^-Tv%L{-vvUduq<T~5&#Yv=w-KPE3e__c}upaIouoVxOK#5k@oj*X7(*gZV<LK`OP
zKRrOroI%ehb-WDW^#w>5NxF}OoYIq)s?9kmd;hP`JMQ%J&2{Ja^^&ZTu1D8Qf{P-g
z%bT82a=1N-(1>2U_lTFf<zO;NXV-LYmRp8}Y9?q8f9X0ugtz5<lj}jm0<z^07dvRw
zWb_NcBHc(m=Bq`XcvbEh@?1Wn-1AwC@(TIF)+_g9*tTS<0vxR<Tw!+ezP`?`$;NBb
zV3Z^w#v0Y|qXnI(=)Q~R07wod35{S!d;i;RYZr<P;7=TO<ND?Lzw_c%$UkxJsk|$S
z)vy<G<K3$1%}1_-t-OY|4^gj1ryGZ(RUK@z_57@5XE#?m?A-Ja+XTNC1P^xyNjfPz
zxgo*}5CZr)$;@=U#(GO@VN4FA;j6qZV~d&@<7lfN?_8{}&@$~IIqfLuv+Rk|Qa0V=
zlYYIPi`t3d_YxKvxnlxZ9SbbojHi3P^^r)dzfXTOLEo&@2Kv1R;bpCUqjI`q^|&i&
zZpVE@Eq4T{viw{gkTu%lc^`5W__M!%nSGSg<hISZ9Nwp$OU?54nuBFcGo(TpDcEdU
z{*7EbSFypWay-c|4p@Id3E|;>;ur4q$nqjug#Z;evD92Cvy@5e3A&(NLlLX2T-a{u
zNXOTE(T+CrT4zXvuMRbu@El|>Owor+PF&A8)e>NM9H?<)OJaY;P$9$%ARi^&soJWk
zZzyx0=(%57_&${V)lXf{QkW*&k04_*<IR>gHQY)-K@Os)9HZC1)+5N`oMeMYY_LLW
z28}eMqx-59CN`@3HQIme;2UJ{kR1<a5++hVrb7pFSje=m$AUAys<kz~M}{5TOIT*T
zxQDx$muY2Yi9IZ>#dDWR4JNtnh`<&8uf>&VN9=awVDbvF74>Q3f4kFMkQ=qXXTN5X
zwea*JltudIDi^izYzV#9>PYX-s^K*i(8O9W`*3qVM@@s??${DRyGx^QTdA|BqM!I)
z9g+cKLv<HlCHQ|?3f`O;rEI(UrjY70RW2t#MbH<c^Z($PA&kgaiWZhcZR-Iz7zr<K
zjoygwUbf+lQi<VKbhW@8Ie-~}xvA`#ZutF*X{@&}3jrN-Mp}yRDMYFwQv^_NCVrUv
zwR8EJ_F#^zk6xTZ<gAC`QfB8ar;0cIKzPhXUPNZt-P)}N9>c<$t6a5C1r$kN(V)lf
z%jS_7Ryw!~`x$4x4<SmkLPI?`if(YUu45$acFle7oT-ADaYx=MKV6ht%>_2RC)t|8
z;zyAaV}{^-MECyp%IM6W0?%0x>U=85z=GQT(Z#KlsTI@6x`&+M!pMU;`Ku}^oc11H
zV^50LoBc9NPDD0EYZtwu+$|kce{iSeT>{&SpL{L&Q?ErGz39NVp|+Z)Jg#geB|2^n
zb_J!7rK{ij<9zsAO)?|h;YWGnFjjz-p51IIhZoWQPA0J>E~k(7%V#bxVh@3{GqK~^
z?S~B1=A(MfV@Qdh9*HKJ(HXie%bh=9WUU!-h;Zwhb>g4D6<*cl(lvkvR)gJ;w_4`A
zch&!W=N=BQk9dyytxIGwtNc&g?t_1V83-AUi&~5~nk^}cN%z=cL&U}WQ>~&wsx$9N
zn#+iFn_EpleNj#)Yl)_PVd^0B@ah}ePs$mI6zKP{Wwn#iIlvP8!Y89}-BXG)Aik^W
zM<Yg&{{v{@mT_A1!K|eFssM>jT8^-!DP8y8zFb1qdHs7h;=0aa@2A*<-6LSPMt3&#
z&uUkkmrMu9naxHf3m8ofdq43E#TUgjtNsiy#jgg-AIgf>#(jlcvS!=!b4*J0EO7Pl
z#ZmV4MIEz3Z@lDNSs7O*-{{V7K##+(!uFs(VP(Az-S_RHz7NT?l(REnql618thKLN
z2jQBvaf)Wm5gR9Neeb!Do%OyA*8Q6&8|~2%4(CT?ktLv~>SlR>(Igjsg>3F5b=bhz
z;xS)sn9RXuP_^0R1l_9LZ}h|mji3NUp3TvJU$Nc!_okV=-{{`>Lp6Lr#PrVD<7Zl5
z^6zEThfuh!h%Tng;2N7F_g*jD7e?SKsXw7eYGaVN7Q*W<ZaX^%K6P3+>|I$DU~}_d
zy$y%glpR7YK_bMS|29*p#j=`lUJq=5OlEdfKbYM2NtEIB{_n=k<R)JELrB>+_Tc>A
zmAn4fJUehlnz~vgznk>Yb_}h5JG<Fsv5)G$>^R)F1tAmnu5lwvEk)_M%jj)<g076Z
zF(GfV-k0TGF$WeSoI(<*!}*K%ym_R9Z98NkDf{X%!tb$2j6O)lELXoVmz{-i6ny#e
z+WrW4*%n60*Y*xP;dw;9=0T65kGQFKvyZ4dm4|I-eus_T3jG(d#fI0b7xGdG_V`do
zPOaD85oE#{B(|}tNAJZmn|fVfZvnH9s$wD~Cw~=pBpsz&?p&W)N6J^I#}PIIljZ=A
zwo5rUE}gM^ctO7^in4^}Ry_$=?}lE}?hB_QUbz$cv!h1^tvFg}j9ECkMU9bGND2A*
zyKRE5o5HM~3EpY}p3w_5-Gq_JgUrqat7V!B!p+3RXj`G9Z(^H=^HAd~iefC&x_$-|
zB+@owj#rlf125}nkMP*hdZ{Go)|WhRJkTM6w*29-U@6ft&xeV8n%Bf!Ms*)F6b2&G
zpMCGpi#@Omc9&%{2nNnXftI;KX`~)Tylc6bJ*H{seKc4w(+YL`B}_P{3YErl`qYpk
zh-@zS7)AFNM5TGoZx)tS%>D>X(9IIy(#L~&y@YRqBkVeY|5dlW93PL!pEPWJ@euys
z`t}nGN*AVZ*sP4J!(5!&0m3ky={h_X>{mmbJo+^mmR{mGm{E687_0GF7P;&N^yX2^
z+()2`)fIPG<f&(?`K0=vE1F(u1+VqKhqJ3+;e-<iyCdbHIHSbwFALRx=gH=b)jOBi
z?m~p$iJzQxrO(K!yW|f_(_ujazGJXBTV}TJ<pl+m)rBQfxPPy7&u!q*F5Tz)a|mZd
z|ElfsdHRX<g)o2bXx@nmv_zz&;8?rCOW3>Q#$2g2tlp`b*7q3WJ2S>-vZ9?~d{tus
zwtTZ)mw%kWZn*V%jW4rs7*%@#B^V}3P4TrH%Zq`ZW^iVx6&o2*3T0jF!sD)}R2w!?
zt8#)Y?2ESq6JT{Pc@MdrVM;${SV}l5>?g^^X8o<lC8IFtLDcDC*wj-fS4J(Bv<JUY
zKn{><ZfTyMLA`&OZi@K*Zx8JdL|et2SG^_XRaq=m8_{DvsY6>`4Zrh92lBEk&1VW&
z8t8|bK55&r0>Qn9UVVywF#umh#rYxIoYTFG7O5gWG3e2C44;5Ec;usr->R?YCwD>}
zNX&sC@`dg`xbBJ4R?xi|M*vGa$N^HnSWZn~n{90_%Y#d7{a<a>E))>Zwh6jB-^c&I
zHbeWv2ezFFfMJzVTP=e)-c+U~SfF&R5EZ}SfQ(}PvAQ%5fN4DRyI`PH>^fQx<1PDp
zYPo|&cQ*ZVp`_mj<&<uV^?w>t6GRSja&sxK%W^$x`zW`Ow-2o&<rpc`nm^*g2wNoJ
zau@Rg#ZSHbG8{=5i9O$oecBHe|9%06P!`J__k%^)Aez#4(?^OT$4H+EfOCP#%^z$K
z|L_5743qtn(lFBfC>ABXHX{5Td|P0`Uh1bMg+n&!B3)E&By||C#`sd}@2GABZEoL%
z2o$bDD#sKVA1l#5HOp6HIBs$ofQ3opL8vp^&&TqYwpSU|#qNC-FG-{~l!&Pu*bKe6
z4}RrY?qi;hf2LjxKE1<yY>4**<@y#5?ikcDIfuJ+n)mLrF>7wvZO>?gwJx#i7fyM`
zS@llRsrx3cZ5B9Ct&W5x(t2n{HR@#;H<$)1ZtyM@wUh3lwa6I9w$3Ua#?KJ@*eXt;
zp3&`4r(q{Wi@QoVQe7POF`wy(SQnfw@(@{LdvVv<vNd~yc1vIH2bud+Xjs%x&T2@l
zx3*P53PBtjn3?>&&0NPjbX!bt-MA=iW&XQTd^(xh48Uk3C*nx>15?4UL=_QOQ!R~R
zVg_pd=!#ICSo6jji~3}V*#%2)s<-^C_99fdTR3zr`0Ms-7xCVDP})zv2ekbLIYEbS
zKrmtv54{taX`oUWX50DvS#+=KWtB-AiW1$BNWwMawBk5lUSGoDx6CjgCBG<zj0Tmj
z$+w0mi8mzO0c_*!=x3k-^LNvk&~V{bRmU&SUrVj8_X^<}=!iI!;r%!1a!UW{EsGJq
z;kh*I@>Z&@NKpFtBqUc!d`;tZN7p+I6B_ezSv>`SAF;<$jwz}%s*qYSdwe}^bL2(b
z(y}^k<hAf+xAvddpAo8;-&&Keii2iocNP|prVTBq_*V+}{pC-@sBPcwxL<f!U4Q-U
z)J3wKJ6Df9js_fN2d@(JxYco=5N(}bJr46&;FQA5=kzs(`O}(a2;ipCSAS(a<h$za
zxSoi5;3X-1_Fq0+9-QUU=*-$`?RMT$tB<&~+pRto7xzXnK0nWHv6;^-aUk1o^t_kt
z(i)R=1~F{4-pjm)=t*<yQ;bv|wJf*Xztq_$n1+iN?)>JVgS!>&U%zC0lKM8D>L90g
zxD%H%B48DK;<PEw!{jb`S*~lY2ceYEH~*%~Dq!x631Ud%y^#4h8D>^x2n(i%`da1~
zlTg6ns3voR=UvHMd#OXroX&jQ7dhvD0vH{ti##k?pAzxcB=Di1QE)_Bti`yyUa?Q_
zbeUZew*eKsw*`8!-d3KrDqgVuw2gJH{v$hKQ?>^%-|yJEwaIB!M%`bul_d%4Oa*0;
zXU9=2rzoy|j_q7S@`U{1pLH|^*gV0LPgH@Ck3<51LzGF^X?5|0n`0?c>FFt<+~vda
z>sQ_ZZz%)w29<H&njkmB?yQe@OU0U<z+7{6)V+TjZv{%s|MvtCk~QD?_Qk{~>)iG{
z7vuk_UwG`y@x$jt+(M0=^dm>fcq<cwS^%*D&M{ftgPD{5SMk-+jZ*#jl%G>SC4W=M
zQ<F;Y_ByqucJt-aubFq7Na!Cj41%``%;S6uM=#fQ0V&sAV4J^p_I>d!eTlc~CcSLC
zy4s-}+TYr?lfVCjd<D}?afDZ5>p#0(LKizd+pX>Q3ki+-ikLa^DXE)Pzmf&^5!N7+
zQ+U$hj%CeyJ8;Iu1QN&sQ7rHokOH-q!53Dxurwn5Ee^TfWReIVlGJ2TM91{&X!LG1
zWc2%Xe@^DiUxN?;!|&*G3zdH!Fjg@$^zA?{WuFB)@2*KLX1|p~iJqQ4f6X1@6hpqH
zh4?tX6MNAAXUBE;V5@c4+szc!>J+}=y?WH#x(g|qI~LofI}p6=2!xt_Em(^Q_t4RA
zjo2qy6H6##`KRNDz`JZR#>rLza)dfeOW)e*Lu=36j%@<g^LDsln~_U)U*`50-BwM#
zZ7i76sZUWFEQ#90Nq+iy7rMz>I|R3tYt~&^hHhVYJ(`*7r_uY?o%@V3km{b5<Zv#K
zEb~YeP8t6=T_cuxvhVf!i+Uolk!JWI`Ew#RVufbYVPNC=AA+*^L&TpAYcA+nw}xY1
z$5<>4#Md5|NeVllcE5*uP|U5AIme}EReSZ=zhn@lP3x4aeSTh7hRc;wd=N2I8#4a=
zo}->pse(lH+lSQJVNXQYAXc4s4}w)Z4y{gc?iiFO@(SR22O4tj!1m{+Kdx!v4%R~O
zJ83@dYxE~rWB;1n$1cl^w#%%O@yV1^k^i@Z`Sfp~b%n2FM7w)zf*42>DDD&V#^Wy-
z4cP7?^B~_~2nU1*DH+EtpPb(_PJ2Xtd5_ZMaWZ%F9g<heKP$@Y?QxaJ(^USsemM?K
z#&P_VQT=tcREy{R>AsJPZbw-W``i=tpQVlJmG1t64}Y@oHzio+&Ro-!H1|vWm=AdX
zM{a7SsuAC(qW|mgLOStGwZp2k|3QhLtx!NsOz&$}g9kX7M{#Q(;{O0scV!BA9Y_D@
zn=zW%Yle8m=g1w8g*YwH9leC#+UhB18OFF;pbb~s_H<A~(1reT0?Y-EPMtWBgjz38
zOv7m1$XW47=Xlk%g4$7W7>rD)fu75%>f8Ja#+?Yp8{LJXknQjn)uQSMDg6okwOF(G
z`Y0Ze9I354mVO1&RqEBM<tMUY$Gy0`AjMI!gm3y6+1mKLyV5PN%^1bz@Z<6R+S9RN
zy=amJTSfOZK4|vY5!tf%Q)#WSLm5>n<<UNxuSdcOPvu}t^3s+jE~}<$){B6?m*>@T
zYr2;?<u8B;Hs%Ie@<yu{lzAX;;4d=xhEu|@EQ!x(7hw8ae(P|l@{oVsIM?p?yg!{c
z_&hQB)mb2wp$w9(*=TQ}4*1LHo38wa&gi-ENi}}&Tie#V8o|QHJ|GUoUsl|NXQQUc
zyO6VcDb`xbdNnQ3y5MBay6-#6qE7P%Q9Uh*k~bw0>A?DJXk${Ix|}EK1IMCJ=s$7f
zSQ;1CYs-WIG=3$No|I(|9T|2VBjTavgP583PYO-z_e85-Ei`^E2$j(0^g%9?g{%|S
zbx*!cg7{gLqc;yT*+FZ6M*ra73C%A_I8A4+XZrmDtwEK~|KB!?m)_%<HYN34w3|0R
zgPG9N{m$@O(AUgxhT<R$Q(k4fI2X)Nsx|9NSoS~1E_O6cQtnfEEoHj{oNGi)BdxAL
zN4|M0=1GU61~8nU$qZ>C)68~~9$R#~9HWDE!MnE)U;s%5lYe%y=88*+7kn`s!A@hf
zFMv2w`i7=0Zvt1{Sg*r9f@IOaYhM6i`2EV8$8WXKzNmfX9VhXa+oC`70r8u)*IH1+
zp(0gxmpWN06Y9E=6ph>=&usrE^)B^o#vH)g(?MWbhx36+uon;VR2$Y+EC@QB?*w$P
zjTUH)S4e#N*jeA3IQ`4Pkh<I)3#nTTL0g*Py$G9pzUmB-pwB_curEx*FSlkMDmtMs
z8chXCdcsj0ejLW9d&i4ApU5t-0KzxF^f{er@J)^FpOl`EE6q5o1WWL(=#sQpOn^~b
za)C}=)ILZ<iAg3$OI*&hjd;qdF-gjG71z9*+VCM+%tz+Y%!8aL>s33uNdep11c(Io
z8Ck>vzx5bflWG13GSn-UY{aJdKw-t$ct04ow3Acs#;eS@ErENVi6p#gP`IDTHgR!S
z9l35gX1AvC=JP-MguBD*>$dkXBvm1;t=-8Q^zxu+#JRLbDL<jV1*9HFu?W(^9sIRB
zqEeCd=c(0W?BL2Ux)a<Sq_v<Q#G0wtLDecMpJY-8vOF8}*g+N~%kD$I<$s$DCg6hC
zuiWktgS(UfVoC;W8)4V&kCd*@tb$BpqA1kQlf4Z4tGV#fgnop_-;NT%%Sq!=cAd-<
zKby%Yh2k!&$#c^8@u5uodS$2W;QX21h31a{$FE6TFw(g<X(~9TnOvg>INGGCxm67m
zvLMXnO}F`!_L{$V+5<@|;<@RdzhPdzdfHI75%@otklJ$9_mv;=D+cW&rpJPB)&Fbr
z^8w1D0rx~*(J~!9_S!VgT{&pvb=m{Nz`e1ngG`qeN*R_5(g><oVRnP2M9oX}uzypE
zs0)u|SfcdF1X(q9G1~>nQ{LQ7O(CIs>fq~|^DuR=^!Q={{T&JN6t|B-s+5OX-`A;M
zYucv9BUL_E1g3~Pqr&dTW|FNpSs!QebBXc~S-tNEGEhKDa=9U&7atu>vPx%ZPo1Z<
z>cx+jPbkW@0{art>`vY4eyb`YylL5{!07^c<md@S0_ASeE{W=bx*1I4mNROsP;uUs
z^qeQv;RRjGt@U0x7M&|6=9=AmLZS{pT%`P#WC6UG{i*7*8bNLF0wwCzOm1Z<AvqH7
zCpJdg7iJ~~vy@ir8RwTc)^+cTRTeoe<Yxi?`Y^jAQ@$2_TIQ4ae&1r%uK<6du~;R3
zGn4KIk79mP_IC`s#;#TzP8Y`l?r#No*u^S&hATF54U~`I9eL}F!qU^-I=@2>L>=T-
zV%@hn<<Z~KToIA7HCsCmk9<>gYIbWcy8l+X?BAt*?#%Rb^J+`VKYUdW;$)%@w^`sc
zfX98&TBv46B-DGcMe_Z5z(4m}8}U7fS2)(Czh>)TZ@%Ri_H0vN?_ki>bGA{nk1*^7
zb9JB-@5l(pdHQ?y1H~4cetkZ!i`HFi&a~EJ2sOj+!}Y;bBSz1;drx}62WYDDg+p%f
z#`d=`exu1+Pp_HlIG}&)?@DoTd0y#T4ani^ww-y;3-!gGT-<vz1I5YaMiI!*YQyj3
z;212If$iFJR9W{bC3v8_Tf}wv14BmrqbA?t3&Sma96Qw;9^&8huI!?JRln%bYN@kT
zvJr{UQ##fwS~aw8QMkM<`@67V5pvYB9?;!ow@rwgP(+*rH4zk6@bfC)Rp(rA7wtm_
zEH~9YqRDT>wl|#6ZtwI?x_xt435t=EUC5Aijl;eaui~(XE^-X5s5aXn2m6w`Xr!L1
z?6PQ_=!ewq<?=hGttdB2nGET_k1J0%`^=BB)Z^Z7nrl(;jI4a65THWOyd?GJOtxMu
z79s`Bnm@^?ZgtwazjwubDcH=RUJ`$!E<MZV{d)`U6wXi#H)V5~cs4ER?*8!FG+rSG
z2hRhI(n5cWhD#c>N}=;vxZLw1E!8Ek{e)?*yBWCxgSkT3r;Sd*A&osR#oomEf)Wx4
zN92*xR>FFPU2<cn&Twc>2NE$vgYN=VuP)T-DzD6aLwG!&?IyFX)h$U-NV|HgX8Hd2
z`y>~KxLjp7Gdeh0hs5$Siq<a?;%~f$SWOyip)|ceFJ70)-zEu}6$I5?on9X{w(CM)
z-?pDv1UpDeFA71MzB=Nz5~{JuNyt|9VuiFrTS>1Q16<=IC)3EhXks^`>*ZaOH;?RH
z&|LVkv17(ry;x}|gb&Y_dQNA0S&W_UR20uC-7M&%+ln>Q6h#^Y+AQX~(lA)#gUR*I
zJK?RMjNWvSpI{~YG}6hKH96Zn2pzBd-I$@5bl_r%Ox#E?HiS;4+^nWgUU@7p;I0_R
zi`i+uWf3yd<yyu8GFqH1iK`_jKD2j>TY4;R_>;XUw#iO#8knBek!DUN{{EH_fqHY5
zNY((P>k%Lq7yMkHYJGb-+yT@=GJo}bQg+(g_i)CD=z&M2gRG#WQnvt*Q+rH<ez^&E
zNhQJRTdCRPHxQm`{5J2pEc<zJ>)n2-^!S5A;`~<i8EHfe3!;Bw$`K`La#h&=5pdDa
z{o_Dst>_~rUp_QoYPlGUoHgVu?Vpvbn-;C&02lPP@M&x8^Z4THEes-duZdG`A3Emr
zcl=L;2U5yH+jC?B%GA6oxk10%U)xymbd%K6y=DP8y!rhrby(;?WBEz8J%jh~na%4*
zlyJu8b-f+Ri-F>wmSAW0^ajUs2x)!9G_i=yA^4e&-kUV&<-qmC&$3M;?pcA)3&D~R
z*fmIG#%|N}Pj!bJi6cm!+H+U7!_CL$5y!g74~#BP;xv$4wRMg(s_V9|qd4<-taR5b
z$4JOGgWC=#^0VW#V_*8!cr!Or9tqfajfRk0J^%Da@aH?0xB&syu5RigwHEz0w!huO
zouWM*0!j#WK3!v*Sk2$(-2Z~)Ko7<XE$7Z;acjK=;;l7WlnqQs+BRFL%V)Fr^Rq_{
zg=s$BU`u)E^zQ@{t4*Occ4g2M&wPYpLX2huUe`W*bf_v;^>W|+hdxqA{3uPpz$$_i
zWH=>u2WWj7cc_nl`12@G&B~3USJc2RuGyt1OVw@)mZ!k<XNT~dp`kL1$y+C>;#eNU
zj)j+?e^Neoa{Zy+m}7eRngcUU>mI6Jr=xF*+|UWL#3Uu|=_9VlS}YW?j^CY(HO&<_
z;*NEp$JZy=5^JHc=CaM!?M}n9Ld!ys+njn15FJZ_KjL_PNdp}d&@1<V#{rXIDS~y1
zd*j^9l&Q*X$8?_STyO2hMQo}I1m3AFb#iSzGw8rs3dqA2`K3kp3SBWlWDgx4|6sj*
z`+YTf7F}*Zl0g?69I`~AwAfeu)6dtn#jmZCJ=N@p?b4Z-td7*o%$rU7cfr*drJ;x<
zIlh)TO;TE37QTp9!{@x59T$k$EAHBG{(7SgxG?ZEcOpUWiZqg)J}BCFbAAg)^IV@P
zOp#toQK?o6rQwQU;Vdm9r6tnxk=CdgB)+FE5%-Hm#qe6YwTS&wAtMX1)&tlP(Z)yp
zPgzy8P~5*qM;G(&pwz4vJ)mo7XaO`_JwI(yO!6seRq0YzSbEgg_9n^F-2EyG+U2Df
zSzv3oF;5N>Ncnx-9jhO`_FwbRFZuS?VM>TD<w0two!cnRuI}KvkJ-`u1?A_6X`7^-
zG$<1I!FuhR$?)&;ve_w%H3qC|nivrq68DPuoEUEZZKM-6a|X;mQ!O7o69-@l0hlgc
zBq#E-DSpbc#j#Jp=!B(v*VFzQSt6-o$xED}S2c`tn+tZP6s#W$$I3spzdC%_nAW%J
zXi`P%U&@@*CcPxXu&@6zGofg1Cz(7b^v`{hH5Ou1#sWS`FOo5`o^^YBdx;GNTI#@%
zsJHfGkSEUsc2qVm`abseUTAYoIUDVX3~*UXixE?Xa#f||%7t)>WSy7{P~X6g>BJ@Z
zq-tWWR^yxFfeAS<ALcTTVa!XRJKK<2FT8C*6`gy=HwrgAmKq`%g%8G|0Mnqq*U504
z-@N7B3<1)})jbW`3=yx_<wZP~Cl6o;B$DC>|NaxhBqqKL7tiuF`}nq}UQ>Z3Wkz)P
z)8OFOx$3?XAFyewwe}kyZ=jG4C~_dWl?1Z0Q$NkavQ{H+JXC#q4CAi6#?Mo)?4Zce
zx=mmDmK!G9*>6a$MRkYAzF&L1>EfoDF*^DH{C4=}XXz5Zr({F@Q^DIui$dXUuPKJv
zViM3`JhlqL?QqYxdpg$Xp7&D7c~qH*W_w72;Q9W(ClKTN=`a%KdiCn*-y5rv*R3HH
zhLs!bdb8<_<B^cqnZ~_tBJmKc?i=BEh+E(S9U6Ro;5j`!*-quxM1q|6CX~sTZ#Un|
z{|d)5o;}NPLyPpqDI@P-h9~9K;%xNRYKTN)Q(k@V%=Va>9DXKV^bw4-m8~I?VbQwp
z^N?)#K=C@pG=!>yvTTO6V2~@50o2@8=_mMvt}FS0*Q=N@^f&^v^=DYKykf|c{@@PV
zfwGF@p<}xEtVJCC0eaQI@=-%)z0j9_qmZawdI$77mAR<zLK)bqKC3Y3{Gn-xs%bmK
z!#D0V<ejgGQrv$b2uppkS*6=mOLEOqhb@au#mH7{in;zuW9170@3y|};i8%5AGK5}
zOV<LDZtYVp&6ev*%0hR(R=M5-`Z}pCe#o$XWmJbK^Nse&xc{pz5!8mB^GM#k`DoP<
zJJrQR{WOZMt)92H`$($NlWHq4niNj+W_2wyYh$M3<W<BxmTJ|)U5$x-jiqP*u6O<N
zLS%pa`~x+gG~#@l*!iG%&-a6mf5kqK^dbJuh|3XDdw<yoL}>$gQe76OQYG1XsrT5!
z(>X7iRm~E~*EId@W3`$=aGd4*NpW+j<5B&w^={4TQ0zK<SkMPhFwBSHLubPtzvV>^
z-#6H81ISg8w~Qo{6+Fg3O?b$nSy{8IlvDQZSw7a@6EI-+_G6#P0iiOt%i3R5gZl0Z
z_i>;5RWU&>aq`E#D~#7hZVSCbFY!_b<9d4j3BY52s;Tf*@LriEo4*mx_5@p)M{UHk
zm|2l7rx3koOTy?bRC4M1uatP(cOuaC-K$)KiV2wb9_Rt*jbetP1IxE->!^w=FQyxd
zG{~^&OkJ9r@&ZXehxE-=6}-8%?1P$g-lMSHJv70Yf-l*X&^g6X=sa9@pP_29lv^Ii
zqcU=|f0GH_@XXMoY!P0#KSc6np^k4vWKT*0ru$G_MDbTlHXRaLm!Bd~rYlCAZPbz2
zt3F@A%4FnMwlne4<J#F^s_vXUk>C*SL@Tc;<nQ=6*>JF{4uaLm6xX#*PYOi@1}K{Q
zZ|dI5vmu)`?9v{!-f6g{-Wde`I$QCJ4&htGIiVLNRA5hSv}ZoN%AcVeHT7F9+GHSh
zAsX`Le0S)}-;4EnQBK|b(+RiM_fFrwTAA&${yvmO{r^%*aKIhCvl=bZb>Z?))(Kk|
z&dduk)6JaX-;Bpafba@H)KG2c&g=pPR~gtu`GU76ii}cF(yqvuzz`kDUz#%Pk>c0u
z`5Wrx+HBBwoP~?ofU({8Hc*wT{A#HH!jd3SyY+he|4g<@{djT!b~CP0j1CC_+rX4U
zEz(_q26&r?bo1cHWYqBEEdG+vc6Rx!2&G@WR*9So$tsy+kO3nJe-1M}6EVnUA>4dY
z5VE;mq<h-KEny21XyA~lEt^^{a=S{$Z(HZ}ZHxg^w@ss}GuDKZ-2c|>kp=7KYM&a=
zNv&V}1N5)PdN;F};UsTa5#qOwp`_6rsN-^ItFAkL7bU!2m6<Ri>``W5yUKz7=9xD!
zEUv_)XbCx?PKUzmo~VnrIL(~g{0K3|(|4)J_KTJzicH$kRk+}v>i^>{acH*^DF}(x
zje_Ob*@f{453`%PjJlH<9cP;D^_MvPLguXHT;zo}I)r&7Hq>vRK3N1~f4>NoyJWlh
zanA+SkK3{#GExepo3ySNRzu-N3}k?UGckPIGpzJ8)#KSueU*GP0v`GmD9O1rX3Dq7
z2_6==0O5EYy0&c)SK;JmI@N|l{rMd)dTVI~mhU%u(v{wm_aw7Tf>`JnF_rT?7(o^1
z_=DaP82#TUy4^9;n_XpqFNw&q*Ug-b=cN>IfLQyD$OmFXIr=kG-b%?N>G2N~h6(`A
z;_hX~)R@ecHr|F}KFf<=)jnLTRb{$EdGrj<hmM9kAKpRnnW{^ORqNiBB9#hv3q-MH
zCgNXUcUmKOF?HNhtB4N4xXZY(fJMxrllMAm!4p7#e2UcE$EYd9u{(f!x~J^+6T>KQ
zX_TKH#Mu!JX1>kV`Q2g99D2Sf6t~N}P#s!zVI>kL&}@I_aw;t6nB!I4m&ATTFz*jx
zp+I(aOV?60^fg+ko^U1tsIH-5x!9$|{HIX<PuM6^lGuRMrx{SrwVtnc_72Wf;;mwp
z0?|dCy?QBNSZ-l0A2TcMHSH){aQ=gOKc+0!d+Z4;(OkVtAV>-`&HMIG|6d<N$QLPk
zyAe=FB1x4K=vOQki%^gx@v4Ev9|877U;7)D0N<vDvX6TPt_Ph^l+5t9Ly+X$c|2rp
zmnHW#%QuDtsfp`*=L%&tPOV`Dt8*2<NfQ%QZqW!Kq5n`%sVE4h$Q$-4*pT90QR#DL
zAmtcnk$U|qFGDG$t0(&GEM0uMGPSI=Ix+3E5Z%A0=`ju&6Xk)-vjO(Dx^Js2Bx}{j
z3x-COj;Yxb#%|(8*-x5)7F}Gkhc$Xn0okP+Cq6HlF4>1gv@_4F^6l2W2J0lf%D!=d
z2q0cd>!&xp+Itgkvp6HF)`D5xC0{jeXfuVMFhL$#9(|Rz&sRA{Ql~8dm>Cc5N@hj@
zx8cLfF-+FGha{2CtP-`R`rNvI^EZB0skxcciEXhe0!hWTUnX^41dC|A)4Mw6iBjJR
zpmHz9mPCEEsw#IGur_m>9DX@nb%KXB3`gM?r}wv(T3*z?(W?q<ewA3JyZZdsizI}5
z`ovL4#HOh!(oopzt7y%O9FGN3MZ-PuBuVni<n9XNJ=Dr87TlAxK6l5V$7N3-XSIr;
zZI*D5n0gXLkkQL}^klwsLF0x=elxElYE%3X%Od4LT7kA6hVPL>w9c)q`g&=zyUs#_
z7n2!Uel^0DX+(2(aA(nYgZv$CC?i-7tXj8V!UXYnZEMxWxQFb;U6|fUo8hXI4Y<d0
zK!Pc|<S>(hEW$L3q(RTR>#QQ~Y|U`{>g)GG2+k3I{R%#&m958bQXh1({!@dZWJq#M
zlWs+n+gUiDFj1qX%swg0p7LzB2<<l=2uWtM?-0VBi0eOw`9SBBVk|Ktua?dJ&KuW6
zpM9bTte!+(*qOry2z?pVkzg$HVouK-PDSIl`v|v0+bTHv>x|8sy<5nl{hS<g^R@3a
z;`fWXIynoMsP+LBL9oi0a9kA&MgnRSN=jRY#P~OH{jid?UdiKI&`PQKZ*5&b9N5a&
zwo;s_tA8?beQcT#aCcfm{nL|Hcc>9|UUWH?UPlUFY(hdHB74)-^kXK-VdMEDAA_vz
zfon5!#3K8jo3qJ1c1Hc<5iTO(8&h(Iv#Gkqh?G`PJ&TBpKIw5~Cn`vh_$zMLlmm?w
z)xK~4mS=oC;-5IfxoM$0x?I!wag!?PLQ2XXWvN(Hbz;Wfm6MvO3U*W|d8)qaMDg98
z`<w$xu;)bXZnlCME=7#Y(6M`FD%#S;T^GcVkHf@;()}Fvog|pkoMG~dKgxzP?9)j-
zAq5rQ-&qJi!%MNQ-Itm}uLu97+e)#ZNbo9^Ju#j%PbY{m5`K}vT*D#*fMr#B!y%vL
zEls?YDbqXOwRVI!W&mQx-zN-b0$4I0C72nNz|e;cc&<oojD{_VkXNp<NFfT)53=)l
z+GYdJ$7(NZ8uLf=@S4=1>ZT$5gB-;$rVE8ljtoF6B%%7Wl#RWOgS|c>eE;O9$grZ}
zZ?+tl)vc08mbDcxfntbyKSf~OwsYwmAE<}a4eok6ipf@y*m$RGwi@Z)_ljSq2Uq=O
z1fiI)y4Y2@aRhsO%|dbDWcCzn?vKWL8kyxtAA?Ii>LkIoYp8NZZ{fH<6JEh|qpac&
zX18gG7<$LmgJ_5&-(zrhm0NpzI=oVtB#-a1=sk_$$(msBO~V3sdONu0_4bA^Uu%)G
z^_uHVssCs!dajB{7aAXzWEMBu8&}WNTN{!6(>L{eCr&+EZd}U34&cA&yU=-eF}cuK
zvWkLP<svDN<>x8WI?SCy-MdG){#&@f7nU9_m7F>D5=^fRPmMe^1!LWF1`NVMX!aH`
zh{tU?#{;vbaCQRpB2=Q%Mm$+w{%XdP>?t6`Pl6|FywV%b`>rZ@PNF=CliOc)Q!!w?
zFGK74b{y>k$@|2~0SMnO?N>|aor>8~1Ma^2zuIib8V2M@<H}llxH|eS^5;>t<?3ra
zejT?XpAM~wXiF<iPU~$Bh>cz+B=AM*YM4^a9GPZA9$g$#G#?$14!dj%Qu>pX`N#;|
zDF}tu#e><*>qcj85(xf!qcvLIK=X@_*yP&Ne=pm2+>h1r`{>*AyOOe&?vx~xthSWj
zLU+^|i15sV-FFt>sGJHK=903zpnIRL$0-@}*t5xW0Vw)DEU3R9@!Z7Q<QQ}17056j
zDT@knZ1<S@6`d8qzgaBhsHj9?s#<J2HH#;JlYL;}#f243|JUA>>Up{<Ig*e}(K;Yj
zE?$#)cIK9rg;<I7jt}3P7_EJ;uB;~96TF&GQkKvata>IRd7I1fi_L2*1NpXf^FUWe
zE-ZO@loYUF%f*u)qU$ROK-+I3lco@v6Uog$PRwV0DfCaZG`IE&9j^TX_sw6mTh2+h
zO+Zg`_4b@_Qpw|)4Ll32VX@EEn?sLP%*ds{<b+vlCM@cb)&ff)p>sdTNn-kZ;ulYF
ztJ%!dt7MLbdlV85^edE`WAYo1t?|#(b6scW;FPm+Q%Tn|xvF*abmPana(8-tnR{a@
zKm8pRu65`G9&gpprKDyMe;(E<xyo06>3bbSv3u>k<IPgT%ecwfo-KaxNAs9R`jDFC
zM$)9YrUbZJ`jvH*<Op@&t)E!PKvV8zi6yVW&wyC7Z(8)bcVR)oSGZ?O9p~xw+)uaS
z+hPk~;|n`?5v$GWBZVd8j10(9bd1wxxXTDW)+@y2lApOV0{?C(ee2ObeV%=UIVW$=
zQHy==X#F-RW-=o;ubEE1Tf9=)Mv+5Kwi8CAW8|Uf$U0YAH{rJn5$mot0PzNLDfc!B
zqV(W`>|RN7yq#({Ss9CX3hZDg*`|%4J62Ujma0<u%yyAQvIw=xg!vU;Hhmgd9k#Nn
zdh1;Yx^=62Gpq%8X;Hmtp2w+lFqfM(e&;)J%nTJN=c=33pB+0<ZOWgkv=6`ca$Re<
zuVav4_k>dNOIy_@&@rL$c7)7Xz@84a82$a#4YeEFl6Ku|32T2FowV!fjGb!StX5Qr
zn#1ImgZYNbKX+dWI3t8y{g>#o0HSqs*uXeltx@V*_O)sB66Lv3lZegt{cn0z3C*GH
z*tY=(cQ~a`#E}VNa)tJ6I@V~VwR^{U{OXEP)_=y6*twvoOR{NhEJgttp^CO~(maoK
z#=7NG+J%fV-WT6|;KD2Q#wPy`M&ve6;ia+Dk}G2yd6Il+I>bgG=lp7`F@defgT=kJ
z;Ux_dxUcQgdwhJ?0gBJt7BYA;huTa{O4U+hhpWVg*Ua#TX9s8CU!;2J4O)Y-7}QR3
z$sn9{mmuzv?YwYQBj&#uqAtF4au)iRbN5VB4-jJ*-wgB0d&GQY!L=Rk?&jba6;=`7
zg5Y!%=Uh;8^Ja#6obuDTMK`PDo%)CAUvF&~w_2maZynSk!+fB%R=z|j%ImCsyO_o4
zNBKRxOkukQ2ErGtcDMJ@fL?2ot{y)N!gZiUZ8872!y?0~pj?3MX$2-}OK8!4C5|7_
z1qVIxi*Kgb<nZ$QQltgg9mjmWq_0rs%N+L_*<Q3et+C2>oJ?THqWG8#^1NG~dwmRR
zS1EBBDG4C8hI5@VdLA&AY2_(PuHH7i@)FvK{ByY%h2Ls3>|!%oTJ|Fu>+0*DuaXvD
zKqj)FXX|@odUZy==dk)u(jS^t%Qj^~VjW%vxc|BR=NG3PXW^61AHkycG`fK;5BDGc
z&|>`9$}1HsgLvY%@;n4Ny`00pINfvCa3&n_HkxwgnPRMJzf-ffYd*zZ+5BzLMisFX
z2SI<thb{kIpZBHyJJQlJ`gipi__uWySN28vgm+anr*85yJ5`ojB;C~mW=rilCq{-)
zIKA8z#{(C)wq=JdIwBAT7^Qz+gNA4oCtTq~s#^R$2@lu{@*kg1f^U;p0t2hYi+ViJ
z#@PQed7rYjyJyvAC)(3dQf|)ks661m-4C=x#=k0s+)V8IGQQ{Ej1>HAn}YTV;-cAS
zwDfpX{iV;tjwQUt7ebH^+2={l1c9Tkst)u!$$BkZKejmozxhjkc{1K-f#sA_nVK-l
zm^@z!>;7o!KrFOo^40^O#d1Rm(FE;;wvT`RM@}5Xw-@}XSuXHhe&3XQL7;RQ+`-8&
zj}&TL=;I(i>Lti17{~D;{(`#5I*Ifx5K9PJXcev<Lpu93x}p4qL)U{}p^Ty(R;BIa
z>m%(prQ|m01D>ZE6vL+QMharFSI#Qe?*!`H;}d+n%U}WDc`t-gq|nJ~iP{Is402yI
zqLA5UF57gCF%l1{Cr7fIPhA%T%CBei81e60S=^N>He?h|4f0`|oJ&PKqfX09XSIhK
zO)hR1f*aLD{H9!z7o4VRy>TKcBjbJzE~jvNFKF8`+<nU+9-&vBr%Ng_^?xjYdZ(FM
z{Q4SURT2)&QTNYOGpviZA1LC7?W1U)S6W1bzlKdOi*8`oZDDv_K`Jw|jTWaB``fr%
z$VGN((#cvc-A>7&g5l+c`-RJ@zR5<!Hl>7FPeI6D@aZF=pt~1;J^;#uHPeoF=!Rwu
zG|PAneviKVw-9I|DQ!nXCvNXw(D9+RrLtwd#pU`#OkmF~mFeaM^m;18(XwWA{<rIJ
zxd7L^Z-vR${vNZ3ng6+(Hz4w~-5+Z@uka8(U7GRkA7!7Swhck`E?o(5WWmAm^9$@5
zZD1hf(Xk_<l_RqiXDQJJebPoG(;8Wg=5E7}O{R#~*h@&zzB6imX+8|AY5RQ9=4R12
zf4A{qPyGNmx)E2#v&SI)TuM0+p-X&KoIX_t$y@ZO17Nx%%r5U{C_N;Shw{K!NNiM$
z_7-aZayVIY&W?WV(wp5dHABV{-~Q_DN{ESlP~MN>OpK-_6wN+kT+evR$QU9Oe(>CU
zB8%%kA|MOX9mmSEyE~-_8=o%+T5@c!<OBG`QB7<4xmu~li=`fJ_1Dn~kL~pL#&QE>
zNj)NV=Dy(|VradZ<o@EE9N+NEx{6s2Ez9Sqk4o4|LPB*U<2kPfU;tNg_xxDi8oH5H
zrc>n`qCM=>E4E8bRz;#V!{*VYN5S`yjfS8<0HW-hG}D&{BNiXswTSeMcZt}`gv|Pe
zTVGD>4NwObP~A}z-5*j~Z3Mm`wpBPif0VuRXSO&eBtYV7Rb>VLHEPrrubmbO%B;2)
zjH&B%1o%8;G<1_KFmWr=r9eCpuL@`CJHO|jS5ZgdU*0KETGXZ+_^Vg*$b3@4VXMN|
zfzYrHPOPd5V&u7$JQ#HO_p<A&G`-6l>bxm*{5}lVHgZL8UJduy_g-GCO#1AaeqE*a
z>Rv|fQ@@_?F?42Y#R#m8dg^BWAjS~THs|}jp+GR#bJu1o?to_|dSXFq?+h;}SVrpQ
zR)aUG6uh^MI8GdVk2P2pMRpZvoT(}&e2M}`|7f(7syx?AN&w%7v-<loSAN%BGirS+
zfyzCw==rT>);>$Pn_0#0eGF>0c6w?$oKrnSv%T;g7+w!N!0zSbAkPyTH_oilMty|&
z&(ij2K~P5RpgjHoMf4lW8V0VFZTFFNeWJ^{nKpwE{YE?Ol=LO|7%B<qQQCjxQJ#U>
zQX9Ixic8LJgK<xGOj|z!HJG`!^yW9LB}BEmUOQrlGGVVr=gp^4wfzpB`82X+I99YK
z)*>Krk=VxQPM&R=cuBAgRq9Fv@G!cLFPau|)UII0FdLS+HU9~qKZXz^{+wka{^|kx
zv`o@)=jmJGu*}@IaEd1-rno&Ph;XKZMfE42K^hFlMiIJUZ!CXA<>JNR+E~&Da0#^C
zLwPznlGgt}6i0J!VjI|FG_dq$<5}W#N_uE}*V}*u5n}R{U|Muoxc$ArNL^|fVj@=J
zr&HdAMH@c#8bFsa`vuvK5Y*eOpt6GpK~yg=m1ScFq?1YQ7M5uV3Q(q!(JJ-m{l^3k
z#Jl!SBOow3Awf-f#xTJ1;JWJ}^0z+n$^LU?w!@ldSrrwn+IJiG>AJhN=++8_43!z<
z_??ec#Y-D<#bw0ETPH}=R36J?bF-~R%R()VE{x+cV^1LQ;L}YmziohZ9<3@wfDYJ5
zYD1M93#nzzS<;OU0Yi@<r}L<Qlx^qiIn<x(bJjX3yR)gI9{<=^8RoA!UzyfuMjez~
zw3kNJ7bE@qGrgc`J**8;e0!U5z)S!}@?DW9*g{*Jp3WJqF24H9??~LzP;gAyHy$vC
z#c+Ib9J*(kZ6b}^BqcTbT%^MV=zJJYRxB9u%&%pUI*{g3nJYcbwz`fYf-qFo+oaMm
z+-lYFjfRg{g7)acy2Uyb9W=yTsLosZBiPK$&0*(vA-izDUM6}SGNA3Z<6tOmJzPJg
zAd9$&Px>xkveqiu&)c;zTm*t`mYll3S2ts9MIJP?9)O_1TgD1p!jrnRX7cV~5Cr8X
znF7EEl~VvZ>}GH9%<*99H>Rlx6X_w+loOn@f74@;vcd6>vx}G234)f4egiqx>R5v8
ztq1PMlMwH>Kr!onqD|Dk@R=i4jJS?6*DVFH?m2j|4#2Mms`gxSNSAPGRQa2Kgjs!g
z#kc`fY2`Xv{&H@<^5|y4P^e)r((&&COcY(cH_>y_OE(1ZIB3;IT*SfY%F}oD(X>tz
z!)7xtt!+6rfo4r}^M4eT&kxYstOT}iPK#HVs>?|N&Xnaz$Mb#?>^TKZHq^mrWOCfk
z<ciqOF^!zr)#|nB3@y2L549h$Q@Jskn<qYN&p@KJvflxT|A!pr)ko>295Y%o$av<8
zTcY&AMeQQ9pF30vhT9Y09PM*3`%-6^lLN7D7r}I8!yHF1>2-c!;r>^Dr2F2&NmjkN
zlM24hi$c=h8{H@2Oq-h8tFe`oX^gn)TIU9mwd!(#y@*nwmtcLmptla+u==l^t269m
z5kuWVy6OG@*+ewNS9bJH#QLLgx21K?+ko?pi^=*0&I3pA86H)ONGL4}0X)!P!cdYP
zpQ$#@9?W)55Dqz#I@U*E+DuP)a~6|I7o|$hNc}Q&vlNs6CD8t`d_CTeQ}?A@X<J(T
zvGxh3j&=JaK;^|iwS*tu>LjM1wz<k$-*`=3nfmyf9Ch``dwdMvjVu{J@~}_L(*ct-
zp?m$>R%cJwW;X}*SJGVepH$9)WQ})}l9~I1N(FxTz%g<_HOJ8;170U}ovjbVm#wJ~
z;-@oTFUPJs|BG*;d7+aYw5oloi+sCVMIqJyV|rtWZPxmtr4)F6#=aflUXe4&$>5C|
zHhgzGn8|D2UB|g}`{D?47}rh&UCYo&M6o~wycr`DMc~C)NLbS=Zl6e{4bsM3MwZz6
zGUq#O(x7jnplFSeA0AsY8BEukdW>Rj%g;?;L<n43@Cy0e0%^dcQn!Jq13I@|?<NI7
zqcb*~Y6sEQ+o6J&Yuvi^FEkOGt9*u=)$*6TT91H>{tmYC%Hc^Kw*iov4Efk0T27?M
z&qha?7f0s9Z~>>qlN!9NL^a6!DYKJHIS-8Ki2o(Ca8yi9O_ad3hD4q4{-pF;O%ZR3
z!L;{tDVW~~P0vC0s$p~U>LH~#bh)YLBs?S?^he}OXl2R!57wHy-p_8d(j;y=KzjVd
z+v8XB>;bJ6SFDRo>5<H@P*d&hBQNKI^2~<zgJkEF1l~>q9y^1}Y)JQ48gQKn(@~CN
zneTE!vEsgu#Id1EaomY%8iDq$tmmZu;*P7m{b%_3g^J$gZ(WH^_T-~5$Oi5Bn!PTi
zTmP|{TM(dhp~26+KLm8dS|Hc~cmmY|41FWnhnlP@-%Dc?KVM!9?0T3{fu?moa4(4#
z@m7qw2EIXuxb?gx3llx-VZvgn+cjbY@OAfk9uOtqmMWNtT!|F6>D>B_Qr_Jp2>gnC
zI9@w?JwXCW4a^ka9M)aGD{&?zNtSR^1F8l#zmfTv<2ui&o;VwoXLj<m?n|9;V@K{}
z;5|PtL7$W2`PJ8%yVkGcC{GW+p;g4$(<1M7gP?cWX8%KM`~(k>{?;&BKCGSIk};9|
zApIVZWM%tZ<+7crBRBiiJQe2*Ott;e>W~YsZh6cBI*hi!>7@Glajp5VP|PkKb(WFP
zbeoC~0m&jachITVS1|hN&!7BnKVhk)krZ%-Po`Y=NvR+JC)vM>eb9gpZ8m(L_^7;!
zAKz{pI}tJK(k#UMIO$KNoq=x67}qN|;{x6%k}-Wa1IF}e{!8?P!r2{f)tjP7Quzvf
zw*r@yHm5k2i?E7BUrek%xYXC=Gdob6Of=7~Nb5g=s+sNaoy!-LTcORxKz6LN;diq}
z=une0zV19s+;V=$uJ%ah75Hl{`ZkMw^g>^PUOmd*L73B`a<?klbYU4=uM*p=qOxrR
zKzEaH)8CG0QBN|+3m3mKdODG98S*bCOKE5)06>;imdZg*n5vF+y!f-@Z5!SZ1e!m5
z85+_!&BeIlo8E9<wHD&1yMGCNYIsz0%BvhM4t81T$Q+^^F7U90f;zBp(&kt>1vt$F
z*0D)xM(!W9ixSu{%&cLWtEqof{aErp{ROqf;*+~Z6V%9pByb*Y%TjT`JLT4`#bix!
zjty_9;oheh<qDUIU-Bn=lsTpCsuN{PQJWEP+6uk+AMmTPrdy6JRxS)11`5khKVNk)
zdi(!S_2%(Vx9$7*UENU;bt_AjDUxIh*~h4aN=VASW#7gavKtf1l4VG?22)g%glsd`
z8N%4f&J4yj_QA|Bma+Ui&*$@ee$V&c_iO%|^E%G+xX$Z5jx%y@D>(EV{lZ|i*NJQG
z&$L`45_QgoFQu-ku}(ZuyFt5WHU<zpUR=Hu>)#jSd)hK-d*z&1Z+wv)e7A7Iv!p)o
zzU>g<`k6b!7btJ5*E2d&GOi+?8I3>e&eHNcvvDW@9v~o&>ak$>4p!!w#zregl2yU_
zqgmc{?ayOpT4Rd%G8+^)uF%dq)^Ua$3D*uPn_lhKoEV?F_tgd+V&d*ufnD9Y)gFD^
zf(&vxbH{K*&qd={oq)I!aglUYg-vrL5HM#!EMbBh`)6ER+B)(LLLYjuK5Dkf|2h}v
z2@PIs!aPFiPxUvrB$eFTV1|;{U42IvY=C;aJ|Fc1BtVY1#>lQenr-%>2%MVMu_rW8
z;>^Z}+xUo}Orqvf&0<nup3;2@<8R-?=B;Ifw)QXE7?n1yr)~ycajv?)BAy&0Tf4y}
zzUkOcN}Ei#_KM$1?-Kopr3Rn6gmZsp8}0f#?A`Al2aMMH^Lvp%4|sCjmPlrLZxQtR
zIL$)!-3^q3aE;r)q#>J<f4$Y(n2GWY5+~EmYUa$y{?dkvk9!tM*R`2-*&Q390v+}@
zj#^YcgzjDg5iRRR>C7;c<^E*!!ev-W!cN=nWN7vy-1|^YOHdqki23P`Pr_6ClDA2%
z)}YFrIg?dB1*n%)TU0`^&U*-@tgx;{@c{$&8j_Ox`SARhz<UjwApbT?O<Qf{n6DY0
zTYF+j37>=NG6rg{@JX4bKcZFrFa(%W8q;4sK1cucHfv{4H+(8cXFH6jYybu=eE#dC
z*j7&go5Pt)1%M(jAz2R}<itUiB$p29Uw0bPwb=KNdNS<@scS*Dy*IO|)*4RvcDM7<
znaziM_i~8~w<K$PmrTN2!vK$siz&!Cv698PV-tBdAUnnQ{A7Brh~l*T9bZDw?X!dD
z%)LhSQx{D4fVVio&l#_zjc#THp8p`>8<V(_!|rE&B5DJ6`z4CwjDe-c=b>;1t7)hI
zGhY?qJNe{ih3TS?jn<3g1*re^QD^--CxKqw9jBx@*7a6>$HPi4W-OUl|A;?Wks@NL
zJM+(ebwlG)FX*`R35VtjYbKIC{ba9SyLc?9{pFVj9O~Vk*)f8Z_={Qgn{+?vzc*cu
z`7ocwedBfJ#<dZr<e(QBI92)UARh4Yh{=2f_F2}=3O3xrZhY!HnXJWAz#U1!j(jgh
zXuEvyn>wzoiPE0SW9q$~UGfQMU0VjwcCUo_h5)W>tnuId0Kcbi?C!KbTi9{Weisw2
zYuo;3!HwJ2O+9AHxZfNxdM`s{?0teS#H1oKUt4)<TLG6~{Zr3IojvSGtXE-cF=q3f
zA(z3SPf`%)Vk$AO{vpaEi}FBA>MO4}EaL%a-Or1$)xrzh_8D`$-DP1_54N+_Ri?X1
z>uN&BMOy<so}BOJ<RG{Ie)a98H7&sZ_)J;ekt2CP`UdSPW%^6M!~1y{*Q#H@Vr#7Y
zSI>`mwQu(A-KOn10@CXznPb-J#h0Zvws*($^azs#6Aay9r8IMtw@6&yK6;L|?rh;c
zkYVorz}_i<RJ&CDg-WmtfZ6rQvAm`UOlaM#?mfV!*BCSY@j1fN?V2UWCtjg^sb0Y$
zUy8saM0f6+gEn%Ly2a(MpEC9-=6$`HdA2k?oH;6w$Exh`iRcveyhVsnh@gR95!(Lz
zwHxXpdJfOfb%Cy4bzuM*`HvQ3JSLp0zjAD(&TAHtcGjeCSu!naQq%vlc>;Lp?^UkS
zEyqe=LtmBEYHhCOC9+}R;9f;=R+7PGPMiFemQ#Qy0lUqb{#SSB11y*}p0n6Do<%Lb
zM<NM2teqH?e45OpsIfmoZ~fH8xVp`elY0LiBTZ=7(P)DCcikKK_$DD0XdA0+FdAo+
z`r}8uT}{)--Q8P^J3HTJ(T87}7DBe~6>8a{GE%&a29zQnb)?JolFW`R6Y7&bKQdbS
zEg>fs($)J;Vj1>+rE^mtK`r2-o?Icc14@6TWq4<vcRHf(m3``^){tvle~x7~Ho3#M
zI-{}V_jX^5Kv*x%KA>sY3iy%S2y}N55x2jsT(^J0rbd(7d9mp7%aoE5*?L&`LvQ2S
z9g%==({Sm0R{;^SU6Ka*;P(g2aAwLAFP@2oT-O)yI*~Lh^}6%N>Pd?Q89)z_aBFXG
zKlqaQ&ly($1a<T?D2@~vEG(mT8Zc$`_2}4z$ZY3WHwlSvNEqP*lYdd?EYAu=K6hrR
zasZYNR8mxy<Ec2q+a0VKlBF|EWPJ_uk=^~hu*3Uxu@zLLR|+63t+^z`4%a*wgmUXg
zDgW(~8&qkG?MSVI-Ma=*4fzoCvzPsAZQ@>d={*%o9&?%v`;&0t^<Gr^3MR5Z&RwG}
zE5PKUe{TiqMa!y>qkKl!1g(DC+T@@A9bNdBpmnvT&+<UAbvyH{HEHG+F%cZgAyi&F
z8DukepeZ#$L&klBiG*3$<9CFDI->3zUh$?3Hz~C-PnTC=-770MrSC32@9qx0Fx6%!
zfc~xcR^<H21gjgZj}9YtOznT8^3vb|p(<`NYKx$kda3kW!{cU$(aw+9)vvN5yXeEG
zW`YSGa`!a>8&Y?N9NZ0_g@TS9#ty!q(JZ>?5pQLvO;K6KHifsuO(@DaSe$W_o^E8{
zRhr+8a9%e87tVI9{DF#@S*Qa6H8cF%)DH)MeO%67-lbjEesWIw@zUECu`6}UsdIZ?
zk6mtNaOoy>%s>*+I+*KXHZgiXVTF&v^^fwbePLlm0d93#9%B{21-hf-<s2C;l4@lo
zrl8z1CTQkB{JlEbrLw~FuJD%L?&#$3^Q}MIK|7Z8<0j4al5Ls@%>2b6l%~ZgZAjuv
zvO;O8f9r4-D^9*Rv9P%58}6KdT?kG_Y_0J1@*M?fA&hEgB}S(709M&4dHci6SH2nc
zsiv49z)U(6CKbQN4Rn{QwS_?4;<%5z3&lq{I+KNqyd#axICuHB@J>6JXYogY-gskk
zhJ-mssp|*jn%h)!i<bT?0w(EM&&ap<m}mfz4}a4=af263nDL<mg}<;;sj)YTI-lrP
z-mVM*$)EklY;&V~fH?UTTe1kQv)^rt)Pa!|i(d1DpWj;-Mo<WsU%5TI`Oa1If6Nr7
z+O<Dz<(CGjg-Qwrj2F8*Zy8SA8j1Dkc1rUmZ@Krm;9)psUzCPR724#y0#9D45zwb}
zF#=I_PCz(jAm+=w^|II{*0jjGn0v}DaB^21b`_u`;-H~`ayZ&~dx_%wNvE|neNB`n
zsdh8_P~^cC7skBy?Q&@g{l7&xPEs~y$TI%E<LvU55oeE-nw1|t`Jo@}>#6YJk%c(S
zW<!Y&ujN*#?um42Q|cpCb<276OR>%k2Eku$ZmI7|wCB0>Pr9OgE45seUWmA{$FwS1
zJ+9oTpj<lf!?ad9<VeR#e4CB~eu;wql3q{F9l%8Btprtsozz;~cemE9O}MyumNB}v
z(tpir=eHeD<<44i8jO;ck_v=;?SglRIDWt{#$2B=d1`Wp(*HT6tMRXi>p@FA{_Q%*
z9h++Of-xYPUa(lfAaAQV#2mPBd^BpC;)5^F3)_F7VZZiv63X`YmIjxM3Wcw%k1y)~
z)}bC`&t%RY=Z1eI!hXM~ctQ7-CbZUy$Jl6zSORmE=Q>{tlsG&~;{VJ4i$P}5(=h*$
z%Vgfp#n0=DPF+&F@{%8Wg$pH6k1SMCab?dHw(Fba*R2G^s`hC^%Q>IrCDeMGnC{r|
z+NfW_O+~6HoR|Ra<u6pL?PY}U>G#Lk^5$aWyH8~@$Fm@dxWohg)f5>NiX(LZ*IOnP
ze)93TTXEm4F_*H{sSl`$P%=4)HGKWg+w*%p0BQ{>_OLwz8)C{+xr1l)OX>VvK<7Sp
zIsAFG;rd-p-=F;|wT7GDQtGA9>pz-#zQM|h&_<z&Z=!ifkurvn8?awnkz+`<HxkdD
z*lrjC;FIMx9gSc|n(L8y<d9*J#LGT;#fuJA3DOPATX`qRK{P+9Yh<kYVnwPb?1VT5
zV}VUGw2wq2MB!%p=t}?b_(!mxD85{@A~4zP*^?Zzzu{H%Oy`J=`^=bs!pg~!m@;8Y
zMA<#$BLA44*!hmc-6dCu-2&E3c7$q*QZCt;#6|NglUpnR+z`^=hO><+(a{#qxW6GO
z^%3PizedAC^+dV9?7Z?R1mh>QyU5uIT}QDJ{)_VGw~FMh5<H5tLbJbDpeF;l%U-S<
z&4?6?a}c3}2M-Y+uS=Xna7Luk62|l5D$}*Y>=ecpBzlJpjI)EBpP4>#y{Xp7|H>!Y
zGcv0vqeDRJn?7G~TFSSvY>dK*Cwit2`vE#n6_2`2s`nPnzX6W4LLp^RkIKF*<s@7a
zYt?u@;cywrp?r5UQVMrc<?kJ?xk!XA=%L&hAHB%Ceobs=WItYghuz}}hB&)85i$pj
zhCo;{<$D%O4JsOQbKp$ii_5zdK4baHMgv5(D=1kLMqk`adp0Vn|70b1v$@>spJt@9
zRJ1`<;Pd^_Lj`NPzk7Me?BHq8u}k8ErqfA++dk`Q>fRf5%K^DCXY&cd-k8<x3y$4k
zqaX7fT(;8J-xZBlqJkJYhF$Tq7`e$VQTSlEStS0iUTO*$>>;DOHcm+D4QhFwRLk6+
zAInAPGZtl`4R7UIGciegSu}GA@4_5b>n{z2P&lKmWjf$d@jC+>VfKUNCi&-Hl*@X9
z*=DO&X265N%ITbruLrk3Ng!i9<^sZQ_P`~9UK!TO_AF7q&&|O7HO9`pfVXCYhu#L>
zN5>WqjU(vYiVK-|FbFntom#rP)jBJc5xzYJl;sO*z_|h1reH(ih0$dcyc7w^A<rye
zIdqhrd0QWU$G0^30f|p}ja1Rn$J~aQXD<hIrn1GIz`kgP?Y|n&KHTr5ofxVT&=`jc
zR`*4J1bZG0AX?WY$H??7`dI72Ab@HY#8wqI06F<V!Z5k?=1i)3xr8}?jnN6J%s-mm
z@Rq$*2&p|D?MOU%Rxc-9EN^I6LtQBifE83I!Fvu=Qai`5-%fvr+^Ro&QqQ)i>@|Hg
zP%ItCKbWvP88*|4d^kQDy!)cIQDIkPf34*0)^bVWo-xZaZ>Iy(gHK4~zCJ4#+sp0V
zW_PfqjD>^uzr*Ugh68s~<-(ZA8;LP^LJ@~$dD!b&m&fxbg-Fw5<OgAi4A#-;B=IO;
zNMI-9!wK0ulI>XF#cx^wLUDD$hMc5ct2um+^B4agNCmcuIT@TG$jX)iYFVqB`6R7#
zT!pIJR~X@iRA$EHH#)0nEhu<ZFw^Sqh25Zcy{**!L0n#@R`gVqquou9+dYtjI=FX#
zD3G>SLG?NaBL(0__KnR7t9QwKLR0PwEk&oJLBsPIqd~hb%Jx6Ag)_#vT3cn9;hMhz
zZ|(s1tQMB%k6u{KRrd}6xk_1(g;CFa#h$u!`+aG;7YI6#eY?kuyziquDKob}iwDno
zju<8-?$5^r{VKuNYhrNq#MB8dbWwGSImmm26SAYOQQS>bk@k|A;pg$kM|<uT`~8KI
z#wlf>>5-l#7@H(trzvztLzRfzq#9SH+D8>KadPnTA@jU^-U8c;sBOWMr=nacb%tF6
z@TDBuzVimnjZKwjMy}pY61CNGiw?1`if|Jm{v6gu3vm(JMsMmXsds8%Bx6YO_C^K#
zGu<V!k2GnCNO7sc5Kq;_IE;?wuGlTs3kbYKS+D&b(!r@f@FO)4{Ge;6-M`+rnA?-;
z344SJ`gO@cfeUjP{En0e{(_S-gg7P*1KM@ZH)SGb&%Rdomf!ni7Y}OJA>87!ptQK;
zXgG4lIU1^MPhZp*+V7kRxXtydqj?L8S<Ixa!P|t)i2WiWx|7XqIumM^(q57B@RTH2
zgt?^F{x3o`rsl`OS&O|PL$AH=V3%C}S^mD!fT&&X^?7qU-zEp4K*zcSLqHy_@(cBY
z;H$Xqu&>rh7CDamU}?RVB--1cE_7-42-Sd0jX2HdbF|-@U`@w@{c$!lzSqM8SkR^x
zsb{Y|S_bm8mYQs``9a=ZL#m_nC8+fLl%e6l@D<SZfPjhuf@&2q=WjRB<h^8D9|B`+
zE**G7Tyy6Fe=>{W{6F}I-wclZXf7?#6j(wYuX1a^I9z*O-Nd7O7;v6L8CCg~zf<M7
zZzG+zbG@fknqQe-Vp~_W+g<KD;j%@n%4qp#{&#sRg0()^)}Ghu3Z_L$c<t09r5COL
zUY5rkW>8kflT9}C!d6W-HduGsc5JRWUS#N4U3iK2p`8xh*E;-8`7)MrYW`Zh<X)2z
zs`qt<gV#nLoaiNxFa8k?J0l!GmYsdbxvR{s7T}dzEAzWSCtpv-yE!Q<$l5EE6ge-o
zVSZ+L76SCP3zA8SAkvrT4LZr@<~8Kps{_VItBSY1ySkd@qHyrt4uc4*uidzM{F(s}
zu(qrJ$LxvT>I~t*Oj-B&)6gA!=Y4MA_|@RPfH#MOkGq__H@#$TLroF{ub1H6*)vMb
zv#;e;_ThK@;%S`$d5$Ms#vWCxDX~^85_UNpbLX-q_@{My({0pys0yJ!ZkA$bBakzd
z$beAivb61~n6<VB_S6$hJFQE5Dj3LiU<S9#YU|x_HG!!(L`ZoUFHEbqC3sknp(w4;
zb5V{2xax!Z$#E+?UQ)WbZohL!I|^GujLFy$i{Wx(aV!k004h(vqaf6a>vcD>fxI3>
zJzRjo47W&e>AIs$?we_$mHXV*a)*P2O-h;1MV4Q#Ug#=Kx7Ze|lQ$NuZqJ9DD`Y{8
z_J+yq#7#1xd!)vef`8d~u<=nIL&bPWBstHQvpq7b;+-lad#43VMJ+s|g>VVQ8=FxI
zTF#~GVAS#6yPl)}%6gK=|8QOr>;7cd6=;z<uD=aGsdp}2vQBi^^s$sHbA;(p_yj(_
zD4=pw=W7AcMQRkaF7SkEBUR3|Sgpk49BI0avL~|@xqb=_7NB$LmM12*QkCs*c2r=s
zDhffR%O0AuuX|P1C+&YgH?db04=sv?qx_$ulw9{@$PRsi>`r6zo>@UPUGJ^CQpa%)
z-%B^xF!YM<fMs&s^h$HI*8GKqfc*<GMqF+yX1bqUO?hr8n_ESFjeXlB*t5#x&^Xua
zxGEGdFD5_w^iQY>_U+gjdj)A9%-OD)ELPlC%sNP}A>aO%(juxy^w%~_>Sq7Pdr@9D
zn9bwR7V=GBg8+t98fX>1s!a>7?Z$uuA>IFcBIVJYILFvrf4;~SQe!u<B`xfxLI?L6
zZ;&>1`o9f=j%}m;iogdSa#)~NN23(%@Z$UU!=KmLl155<-(?^VW0#b%nU>swUlo#w
zYE`){JzHY~kd8GXZe5f+T)}8B2jMY+R-3Fx+?PI-sga-R&^ZU2AGkeF)M1yGPJTM~
zC@BMC%hbso4D<=r&u(&}#VE&LfLrj!Hd~F57f<v$^lX#yI9gy>^D`b4htOsmLo57X
ze|{Y|clHxv&*3bFj7``WOfG#*Att>2=N!ki#a|1OB6!gr==X9$w1cfK<miCf$>a<A
z>V(@8e~?3JAAg=L{~CSJL=WZ`*(=u(@wXTVNQns;7eR-9;FdWBl=q`)Zx4SW<!<b$
z)i!Ot6+HQ*O@ow3rqoRKMb4eg<AZzgRf0dq=@cT&eP$w4uilgi<MpaC?oV<)kV}!1
z+2CxV)Q^?>G~K_ne^3%b8X3s;JqOo#S@y6cv82A`mapGPUAAy{&&k?}m$Y|CKck&^
zABWAWeGXwHLNAYD^A67NF;-0dx*l7&t-!{w4aU3W*!i}Gk0($MEX2()MCjqOX@r>h
z&FM!rJz3#HALHeAYi|QjY<|yCaiP4*^<36K4X6U-Dt77%VX$+2Ng42$(o~@d)JKaR
zc=5c#>v;y%H+TNy^pCgkV+|5zYe$7GS|exlZCOeu*|U{RkDJn+qA!G--Y}mu2RPaW
zo%h`fshnCE?+uuCmtutK-d_+Qtqeo>Cq->e@3>BThUHLw&nv{Gc&7XdVFikR_?{7C
zPJ92Bjrp>_=;Prhc0$u6OVEEj!2vLk6i|Qf1*FYZ(G>PG{;>?kCDz)8GWOmq8#aqI
zD&ltdHrHJ&fQwq&VAFDj2%Ul+Z5Ii)+@@<hIxKE)kGhxhH&)Q>uAwe>3Yy2l)0$q7
z4urN)t70bNYkP`hMEHVZ@s7No7^=a(cD{9VmkCz~o|}Fbn`=5)xGpX=$@$MeK9sM6
zK8Cf32JJiV2epBYovxnTo&SA_YB~3GjoGp2SF4ImnMUGT<VTG(oZ1U)Do>cTzVR|G
zZWOTD%Z2)no&4DHCcG2iIK8+C%>a>Z!I-laJjtn&CaHK2yfFQkgk3*o+QkVfVc?fg
z3I^2zn(4k7wR69mK6H4Rp{T@K6=&)1DgMssInb@uuKIk1RNcKhe*sml_MQY*c4cC~
z{wr0>GeIpASR?7m#Q}+jK_6pUIOj01-wW<6;zQ=y+=P3G^VzbhJAYr;bufM-5nx++
zdfocz)Y7H-Rg3FI+yD*`@bnv*Z2RfF3n7r!>v_AhatRf(9N^X=eF&?9Ittmnh5h#6
zXp5Vw{Xl9I&Kw&!RPvW2?viE?XC)9U)zMadtYb@b$$hbKua}{HywSMu#u2>=bP5O_
zz?dQNIod(?us)cBLm9mA;K_sO7hyLA{+?K)u9#8B8@dPOf2n|fYE^K*I_ib*z(bP}
zD3B*i2s7JTZsZL*>ULEj?@yPg0FL$;%<4_#{DMDgoatU0xc_A$Mi>Nl!^!VD2(UUc
zsP117fNd?WAlgJHQsR7P_TEp*@^1WyruWRzmzRfoLky&Z!C^bv2!cCU2~<HHEDs3g
zb~ucCK|1jVv+zX@`y<8?4WR-!z#eVE$tHR9?+DJY+XD7@pvyLUxf&uuNTAA-4Ts;I
zpQog*d*=ePjtrVsMyMCpM^NKoAp;_3RXb85#7+TB*}|7SUvE>7U<|l<`80obN(7Vd
zGq&XHPG&>#D=^LFj}7&`ot5XS3dyx$!iUU=sso&JsG!>QIKHNCcF5q(LM0XRx}|6S
zo85`(j(Wv8v;R1rkB+qU;**9f_kp3;!wbPrtoAf_VO}fU+nVA2+q8VD?Ll$fXn9s$
zBZ%>0;jRO7e|_n_7%3F@j~U6`tn~pIxkZc!oE%c!#>fR!U%dwKr&OC}Q~?S#<Nft@
zpRAX`y%Qx5oI^R_kJe85%*<8lW`Ii-vuZq2{bb~=sI@+3TwqtfXF1AUP(I0J<_~l5
z<k#Zh?+1a?hiap^Fg0FDkz46#06dIAXVEs>;D;qgTOZP?4+=<-hpj9dtg~-f@^4x$
z+UR`@h6A{yUl*^9!iLHrNcVZ(zWr(jE{$sWoM49Bd&A+&_qCs$Lmfp}w;L>Z%+r~g
zJz_1n)9gyPt~Z=?1b`o*k?`ZAr~<TvmP$h_&(AYjkFAROjy)$-t0bj!VY?qBLSkJ+
z@X%LL^kWyC?W0Ql!wmQ~zS0#%6g8Dvp&PA>K=jx^pE#=)!tQFTM;{aUwuV~gDOKq@
zt3$38u~Ya=&iz3R7K?d=+l9*u?w#3eYnV`MNW2kn4fKpP((ua(8J<2)Sx!9RKTLE6
zpe&}I5)xH9cQUm?V$Bj`8CU9P<|whA(1RP~>hC;L=sgA8In)+I?pvQ=ye~`9%4y0a
z9=QCc$=LA$99(^=NyLa~E7&xd_i#%b9ljdx=Y0zuAQELPiyQU|RP{xEgSEH<yk{JM
z?kIke6;xFgI_TT|Dh^rcSM0T$J@&OrUNXNIs3<$vqanYvB1=-Ct}EKEZspLvS_8N(
zdL8ObO=qu$w?6d`J=n|ud#yDstSf#sLR7Y+{uN<$2V-u4U+aPClOmInBL1Dx;kuA8
zeJj{7v1k5^`i3TfV*b&$+C{;CyJ*EAkWSaXIrnbFbZh`TW-W$S)|fJ1wV!r2eZTP0
zywS7iemuu*Ynl2vDNbIP&6@}-0VC3%fE0GY+)I0Pnnj}Q3LVvmYbM?ZT4{rJyN<|K
z`5p@^Hk>8v+@;@&q!Yud!!B!!sL+?t@37oAU^_3omy6;e4op`~pA$mdt;(7)vgJ#Z
zV^|6as4DpUD{So9Cyd?$HZCAO3;*ZSzaS_m(57=4blb1Dox8_z*H!6xPq%6@V7Z%8
zH%+YOHKCe*gAO1-hn0-z43*T%*>56mN7q6o?*Gi}9EzJUKwVpwpN=sjiVx%eGOCU7
zaJ7|#PAUP?<XX-AN0?-#=;fG9kF`Q>$kcJ5CZFSouDFOa?Rut_nsJgG<l#d0Q@h>8
z{DEkY{KjL!>g2CNVN@v|{U;7xXZhdJ`!wJqu8$jCe|G6@_jRO&PyZ>NnfkuVwvIEf
z0#wK(>3}?cE%QhDu72n&HAClAWl8&ohCl}irEm4%XH&yQDyB(vEEsZ0D(`DW#$1PT
zhdMl_Wm(IDg6uv){ocDBJD6ZyS`^&an*Fhx>{m#t7h;#6SKCnDdnvv%DLAVg+$rx3
zwT&vkR2vLU#ORScm*;vhn6`xY7JfM5#ulciWdH>gC0pK};h4(J4d*#Iw^O4U9CYO^
z+Q*vzN-QlY`!{3>C?#}Celn5UFn9>>ZD)irYqsj;eKXjuZ6~m0MpcQybjErIoS)nI
zl_nMp?yYJkF_?qg{P<`DGrc?4n*ay)DRy>T)A*klrbKXqPVK&2i2|(3H;A3mV@vwq
z_<n-NuBr4`%I@s+7|1~s1~)&=Qo3uJ`=VpG$rXiM#UJ|CzX}AE_XyVe{9&?XAAVf?
zA^HSYKo<UH?KW1{jdwBlF+w#|IowtMQdHI=dHCLTTzgRKTZ!k4cp$+kF^BYuFdbkx
zkz2$#ABLekvQkUzR2}W`@;tY`Tx-8Hz{`zh@T-LfcZYp0@ny+=iM>~xL^S_nH8`Nh
ztw)gUxe`WD--{BjCeZA+5cL6ZO|HyFi{^n|aD)LbwTOqqFqmjPkUQ4x<oj9w06Bg&
z@^Q}2a|%E9b{zbBz~01+6KS~Abai;Yg$vMATHyyEy{816?kxPqzvf9u29??P<$!qf
z1drBd3h6J%FOHrb$Y`mhd)_8QP4J4S?WO<Ne%24c_$#MQh+m>smKJm9?&TA7g<u3(
zBO9-i3!M_MU!Uh3+5ouVa6+7jVepdKnzeZWhrW9ZIQ@sdRt21dGxaX~^-yOzFjRrP
z%{R<<Q403D`=pVEFs3A6wg6>ZBYjUFiT-$#9%na*94h@&p7l{1`Y+V|yUKa60UN`=
z&uS3-|MTfzksI&%OQm|R`{!O9EMh*fm}a)ifm<4Qqio(ud{(yH@{DzA^;6ySJbNH(
zeDM@g<?T}MIrG&%KS3KowZiTMSI>)RBGD*GqWc<7q!mGo&o}hyC_jJ0FRO-CQ9=s_
zOnQRXX2{g%;14T`TcPS3FA!eoWb3I@+Kt?Ds@YaWWAj6L2YR{gRxa5~8jiy*rjBBK
zl>wSbZ*MzerXRzt##ec|OjfO*oYuF)Z%VcWHR|WQJs3EUqEu>@Urx<d8&*$hcGoW&
zSN9JDkCdY8_lKA1J6~6cMET*j37%A}5B=L@RRLsW=v&K2t1o`d7$HEHd^}TRAk9Ro
zaMu-p?3On-n|G?QLyjhtC|qdL-ith%7VsR**Ag61ay(&zZAs4o{b`-gWq)Plo~VOT
zRL?#4*{QCAoGz0sG&YRsSY(~ljxe*3ExW47<-_ZEAJX2EjAzZ1^Pu-E;S;=O=8IFO
zx#NxL)ql}pEDsfu`%y-QoS`{|*pD>;TaBAiGhO!dg`}TFOk+c*UP^aNrIM})i2E6k
zg9dnmWm@kWb+`3t{U7cWPUXI?e*HtDkK4DQ_0xZF_wNXsusPVwU!T2kNkybudFqCs
z950A|L&d&WurA;)4nA{lp@D5S1u#d*x%G<5_20}J&&a+$r_NrPqrb3HjXr4jw7|PN
zY4ChwuOVD!#q32<8Fg4XJ&QB?;)rqvzh5Py&an0}$MAd@qCI>4CG*wLQI~7l3uFgk
zvg(Hn>5@PQsq2)p4<#>&=|f~AgF`gG{OX@cqsgn##3aIpxI0}gC`)Ru9UIfO`Ldd1
z(BNf@dp{ssqN#Q^MO!O7eNASLZ_4RQ>`3XjQ(|1fN!nz0&)TH3YVwj;Ui^!d&~JEx
zRyb4Dr7l#VYL5oy9(<M6bgmoy4&$JxHSEhd=_jdhU9t#9<;m|d$2wfh*y2Qv7xV`B
zdR59zV+m|1`8nVOJ(aEx;$h*4gxq}rGYpS*C`PVpLlKa+V#5;P;NVw85txO{A#gd}
z2!6QdMbqY@{<`O4S8pvne$^zm3X)MG$iorca)wW&pDRIV#<L!MM1K@I^Yz43E5uhL
z0QaLvHhOAS1)!zD=#EUr331~EoArYeWmpJH>B_dKePkl}aep)#zQI_f9m_{ipqY4@
zZtx-Vh27txh|tr(lCY0%(OSGAE*5G||H-Kvjm`}`?zVSfK2bonSO1OMRhQQ*ubt0-
zBfpv>INA#OoYFY5RPn#C@IDb-EMGNkR;rX7qAj4QJQ^}9lps%~K^3_K<p%`<w+s>-
zOjNuV&~pK-qCgy^Q1)@%Va=KSi+_zYD|GhNzENhVrP|P6S`W`GHR*MSed-22WlO?*
z#Z+6Ip3K<y6&K<1+4~}w8_;r#;i3v##l8cXa37^31k%bXBDjM)nm_xkb}SC`9B6<`
zFY0S>sQO1ax~155ri5X(={LR&6O1W=OIm-&9L|-Q+OB*=4=$MKWCZmm1(21&ET?;f
zx<c^<uOVl<PHQ3K8%a&B^v|u3j^=0i8NKRTq14U#mXd6Bh-s%rXH1Pn4yZ7E?Q`Cw
zk=ASnQ)%)a2(H4@oirZcAd(kwA_{%C%*jhKFl7ph-(4M^(X***^p202xQ2TFXq=Dc
z@3<5iuxg63?kFzKIBU5bm9TWB*oYXJeA>!3BV6N5ospF0;j6HEq?&HVTQ#Uba7elA
z+OLP=N3Mw6$z)Q^>-=y+eMXeyXV>&n?STJ|%>Qwne`nn%Y&W*)d!xm>VX_1w_vkDe
zwDv8jZ5ItXTHXgnj5Du%=fO-|YZF8+*2|kiZi|}{zhpDhZ6>Arw01OQH%BI$|MGkk
z?pz)sWjbf29$ZX=p4Y{p9AT2|i*kv^T=Op+%57TLg!F8ol_k`4YuIqG;@Hkc-1x3r
zqVPE3`G|07iGr6$U$1F)sL%_E4A3)ft&(>aJ;6d&s;Ye!o`D7FeB&b9HbJs40gq9o
z66oNES>~Spl_hl}hItois6UM<<J!5cJ-6#CA)k1pPz!Mt`Q6rhCjRBeV&E$o^pE*M
zr}h^cZ=rL)<VX7skBns9R)2wCc<H%%Z@b*>qg<4`y9VXy80+VXvcq(<gj>H{6(99%
z+PZE1BPfLRcfz?`pMtj~kMn<xPKlbdbnP|&7ClYM4G8xaD(h(ZeYvVc4p<*q#3+z`
zeG}DEE<W&WcqWb1yO;0$HK+ROn$FxtCuQlD|5iJ?sI<l=C~u+mtr3^U=DgHw^&p5h
zMAmun58%If=6`VYZ!RhaIyL#aGhz^Vvw(35%fDLPo*LM?>I=9lq_RZ6kxs3PD>`<=
zY>b!lZjOkE7d*Gcma3fLp?Z@bXR8KtG{~y5)C!TV@xZV~%rB6K`AY<jlt9IeUWgg=
zh&2&3x75#1F<*Nz;r`U0%YD_rku@lwM_ht)XfPuNbFs9v@dWMgg+EG;e#PQuA@3b2
zn78=SjYZs(uEK!D)Ito!uo~J$KgA{ZYPE-u3rdwD(M1Pjw<=s|&pClf+gI>IlZQ55
ziUHJuc5R#fOP(?~X4-bhKMB>pKuy>r(&vt)UQAB>opzP@vlb5T?ZUjZoa6EnvtLFy
z7kj0FGT!;VPl{AsU;7hiKLknS-pE}WU%{_VE`Tq7<>du=6z@)CxB@hBG_21L1194v
zAH`1E&WI4%WHW!-PR~UET9i^Gb5R8?VFn5=138y_^ji6+JftS@*?=^g9iH)mJ#sRM
z%#XD0vLFKH&mjegy&T>zzQ*e&Iu$2^joTk?e#%#K<C1Hh<L3owVZDX_;VA!4c>V>S
z<7;0tXOKxT%o}v@wQe?hVss-4lNnxKdZH8?Bz6Z9=BM3{G}i`s|1-e+s;N{5yI3%B
z!`xBOu%Y){{|_-<Cjz*rzO)~GJ|_N&Lo%)GOJdp2Mi(I_<v5*y<BY5}E4#e@F#%8g
zC`-UEzn3Cye6d-cG((Pl0|iDJzQ8EzenXHlp_J^q;RcTcH#S)Yl$Oyd=llRtlJyS(
zwoQ5q?in^a+kCFudv8?zM(b;Uox^umaag{Q-DaFHr_yQLbI=PC2Zs>a*e=nE;AA~K
z%AKCZji1mx{plI+h|RgK%yX;vYRYc|lj`Hi9vTcq!PlxD^ydqWT{oPGV&$4vXN)s2
zBbb%ZUw+7u;1_CgyE^ORg%~E~HkaqtVAnP1Z~-1-B};I(l4Y%>CN*6d&ph(PRlqVV
z@$0{7CPbI;1Tj82esU{fgtW6;={;rrMV=dI@t*1H!EPvsqEAb;gq@_caF?(KREW;q
z1KK|DqY&V|xp7vpX3mURQls60Ac<jyMcsN2;1nV%aOulVwY^P3J4nM}TX`z5$I)EY
zbwYx`x~ac1<U3gvlq-`Gy}8+5lxg!kBXA-gz~@g4>k&&+gyAK<yi~5ap95&!9*XDF
z94<Lpl2bX_VQJ;;{7MCoEUN3?D3`1MU*S6jDBt@CD6O5HM9;9#ZG>N&Rs9Y84-fj+
z!I+demAzLj(8#{o3hn%Wlo}a-FsPi*uTszp(QOi6s8Rz;SDjegI%9na0we-dOsP2x
zu6P<(9w8f(qy(;WkMBCB2c)00>j3cVpJALTVf7@k?B`H7u*=Ck-WK2HCp{yCKCs9U
zT66`DOZ!`OtN!@uYYL+(rVGG3rp9ZgZTwEa5igCOI=>7LKuJo2vBKsU6W_U=uF}$o
zavA5DRK0q&?V21M>nSlZL$Eqv&8&R)OTGUVWS`)0J~t0RT<f4vA0@wT9p0C}fe{N&
z5uCj((08>2s?_JM^;L5~j4WkUr-)>2^ai<<)i!^wr#(fQI6E+xn~6Q-oj%i?1N|*8
zX*IoZlk9(lG=7!J9p&1=tgi(6&H*(@f&YAAd^H?eQPCO)RyPeGiNnx)iAnNKw#lw!
zSY#F82mgT0&W2s9{o(p=#szb_Nru*bh}@0<fJ2_JFaX+Y%nD?@K8&G6An|o!P{!UC
z35h>gHTfQrg5av-MO_VM_N5{bW~IxvjLkpgXaq#&Z_UunDzK~Q71}V`H=(Gbwvo=p
zc3^LT$od|4ewTq!?Qb0+N{4M`W+?r#(-OuPoSJ-gpr*T-a4_J1@Fewc^I}df2!HT=
z=<MJX-`b*}ns<5G${ddVt0+z3_N~J827wGC7uYW9KmtvbLG4euIxBn-SWW(~HiDfC
zTb0jjUi-%YeB0&|R;SSL#jXEIBL7R2TRDA}w+Z)#4{GeE@2|xpo$IPDzWK>agKF~$
z!`^KzCL)FH>jZN69s1KfFMH8qR-efi3wQ@)r6ynswv@^vtPZ*mdmd7Fl}haKYS5Xv
zhZf-Vu_%Ki8LtjSz*0Q&eW2t>gCa&ig_IF2#?_NuxcBfNfnC6Q^7S`vZaev}GG&(I
z0m4=v_H;Q_#2p8C)ff`AwjXwT!XvNg=4Pa<Y#2N0a=e`La9f!wzNk+;eKcW2y|i7Y
zc0Gf`N1B;Hjpdr0)78ytPhLR;$SB&Ri9?27;D~rB5Va6YNTF7IwxJHDLVTC3O>nNM
z{G?Hd7}CWx&uSN1r=sdGy3<~`<l2g1I<`V!)~a&vuD2kVfXma1hczgW^!ia%KP~2}
z-#ps?syxkKQmMQNH}}nLJ-4Ohk`p`t`SPsK$dpwP>lcT9@IgRADP9<;l=qWuABjLs
z>=(<>KLz3>2x#N5yuXEwb{C~eF>&qH-&R>F5O3Rx5!*&Xc$~K3=XsSVex8<=qDgm1
zLZgl>Q8wv2x$t-{r<6H8tJ8OMeDekUP0I}~E6Ta8Ca@AuX_xRHG=dC^3xE~nMpNM^
z<|9%)v!f<P{ndXJ3cmD+;|AKxdB8T|u1KYxn;khD9=Ias6{}bEU%zT=xGb>o@iljM
zcqXP;1(?i={D|ZqKXXN4awLyh)d(zB-)KB23}Hm!hNntTTS0z88S(I`z9B)U!a&Of
zl@loqR%+CIeH^s>eBmP@?XaOx;gwiqUlFAa+hK>j8tdr!b3Ulj+$t?KJgsOEmUH~b
za1c|hNLcO!i+wYQ0Oh_bbli+~PrR(5menLHAX&I$K{o%RhDfFa&~)7iX#p!;cAA6o
zeSv^SxkC=%0<7sNy|jXIT|4<^beAR2F?xjNz58^1JP<i&usX+<0Xr~8TTc$gyI(?6
z3v+kI4cmOJ^2@K^AT1a0$o29nQ-6Sm)IckPir=Kak`OJD$d>#kLR&=d)>bCz3UiAR
z;M?VXDw<{{IJeiKW>`y)FyK4@diMut-FLm`Exj|7c{i&iQm9hC4}CjkuW-VxxF&*s
z-N)rgL1>0DXvJ%NVNjW$Uc)y~d;A0ip1(s1qy+Qrca@N`!JP(2`^%Wk7F?<k<-4CD
zsg>OvW`Xc)HopTcot|aB>Pgqyv8_ZuAM9*8`9!wvCi}Ws&c)Q(`&`j2|2PV7{hFHy
zYOuB1n|U1HJ2NuWx7{x2KPY#qGg8243J9O%wIHTyLx`zQFY&#KjT5q*lx2T7`zoT1
z7RDM`_ujdpYB1HvUSb)f{#>XAf<HW*(^oTc?ZcpiWQgXV$K8FP|G$LbH#qBRb`R%i
zsavfl{8Y0q?Mmr-?+a1>n~LoGQv6eLmdMS<`mCX+ky^0*Ekh-bt}7HU-o=Q{Rc>Rj
zvH2EjOl>elRVMnmv_=y{PfJPo9*a+l8NT7#f7v7!OrB{Q^)eF-ie13Ht?CvQYDTH(
zywBk0pN+kyw4m~=M~N{xY(P`n^X~uH>RIF(&>RQ!KXx%L;q&F?n8XM^Nzxa@CNe7t
zbv;=W(DRJiiD{+u7D~nA1f1`o#{+9ZtQU2JRM+$yS*J8yL_1#yowW@22ogeN{!BdH
zNL_#k9sRD|ldUvDM0W)>;=p!(I(?H@fV)bq9TVsNs`!r=fQIViEy&DNZ~r#xl2^s_
z)agLFxHYGO{FEp3r<Coc2fF*6v9C>$*(yy1m=Gw?{09q8ObE_|Ii+V~+xCE-n+Rg6
zl>4l?A1NO)@P%(vSLOP``u9|B85rYni2K6ssVC_zL+;2ghTf_Qy7EL~NhND}R>CL*
z_ovru(}d~?cd?SMl-q8j{P{?JcOcJF<8^F?V-h7w$4rkd`0uso!7Ekx_+3EO)_2i~
zs(TiTvrW~RA^uZ-*P4S9vS%XCa$F-<wlrz&(~twA=i@njE%DxFpi8`3A~IPzW>~Fw
zZ8cu62B$ik=xkl7;YtHkiQ4Q6Wo~n&&s@3_7)5FSiVz(B9VvhjlKfRZcQReUVaGqz
z^4)J7ND6;Qbw%r_ZuqaNafSCvxF4=TaH__l?41J^ADaE|g@zJwnL9ywtzJU^S8R?S
zp?2oyji{pGnieAw&}UAc7mQh*?@9!jH~*iM6q9wLhhfdLf1CASJbKmOwk`^QSqe3I
zCn&t-awp_wJ#9WQRi*oT&B@vBf4qef;7zt_(1u1C=xX&vd!WNc*6FaVw;<;VVKLzH
z_uwU?U>?n&#**EFxg_vBe6;IM)K?GRu)P9)N@%6L_<S+2>lDVWG9Qv%kl$^0S}Na#
z%ctlThr(7-@GSyT-C3IVnnXs|DVXCE4VJ`jH`T!D3&`-3MC4-FwF{mT_c{gT1)_YJ
zYTHWS?`aWDH#lrE-uEb-df_~GLU7@_QaJWqu%fMx5h}sT-M1u^*HKsf4rokqHb7{j
z<5bU6en#7FRpWXo8)wjJkcAtuQd#b@Vo=&-G5faq`!V?kg`0(|vP#VPyIMmcyH6HW
zdE|QJuh^}th%8I*n3_>G75NI#McjO}ly2XoI^OwgMo;WdsdUE^Mp{=`yj&Xc?ecVN
z4<cJXuJ&8~LM;(#g*^+4r&`oG{G5<j)p{;Goqe<2Pv&96kF2p5NY*B$l^0NF{>vvK
zf?0@#^zu6RsxzKHCZvPa3U(9uUP=4x@3Ph}`UA)qv{{gQ>w9h4*RTDiquwUl66c)4
z|Ivu&BwE5*^cQw#PBJD6RV2dU5H_w-HQwAyW!zZ)aAYepK1h^+cAM(VL(Sbi^0(Tc
zc`n&jp!aXuOIMI(>E!+8R9Vv!_?rsq@d(79;66fkWU4uVSp(Lyh3_+KFLu(E{O5;)
zP<Rh`rZlaYPkxX-$3jAWtS<i7b=J;_ji`+|kjz04+ZpZj*?`bfCm1yctWG2u%aLkC
z{Mt?^Ilq=7RP*0G`YG3ph`Jj=gNN_4dfUhG(>H`vau)r>(gX!2pFEvUlz=>tS3M>m
zH+>Uro0P;Le{!9=?`>D^=z!M56A|<!JweAZl#=P>Yop+;LfN&8n}PzH^>KG^{K%;p
z0ttLb+E3voC<r#atFtR4_yf{~I>RnuEdUO{asUiZU9x@#Uc`GCZ31Bm2y6EwS~j#x
zZ}So@KX4*6XMM<{7i+Fs+!z61ECzhBf;M!XZ?0Eh6tR0Ql!X7H&EFX>KgfVRXF>ho
zQ~`t#{TBI$UyoxslQLNr7PQ+nJBceL)Brf@tbpa<(}JYTFgugtH~v4yW!<rtTa^w|
z{kYe&Nlo8LM7ck%<C`UM2Sonq(AcueCf$v^YqF#xn^{U>xrecX%@+3%dB*i(Q%h8C
zywCUX(}_=Qyy$ZsXjB3V04$gl<vPTf^G0DXXXkeI7qN~hZk<_1s86n?P2}OMX`4Uy
zTW2~4n6Eg*Degy);>qv%d41>w*X0L0ijGh3@#atxZ^2DDQe>Be%!hGMxU&trG{igp
zLtps#n!#7Z;pNaoghXZd_AdJJBum1s=%vWFhPaLPP6@I>Gb>MDW9$)LMeL8v*Cxvi
z360Bx29kHUS6xR>`}}^1b)s-$H;fC`sZ6${NCAVe9H%}uZHq;tQvb0?laUB&chxEx
zZP+un0{@&E=RRlO;#C<l!7m0MvhDO3DK{Rf$2;`E_ROTBuw#fLSLlo1N#L{Vla@f=
z?AgdO$|`@<76W=|5!m`qzMQd(2&}2xT8&$&BII>~angw;&ddr0Q;dM1xg4F>z>%M?
zj{aSOohT}Ja82VhHo<vEB>e7A(m#b4<M$V9u4^~D75=<r6<bI_586##8u1*}|5-4y
z>;A+u&roe3oBQ`E?A#{-mYKAeYj@B-#NRU^ag5=97@|6NLG<{ojL!Y3UD2gAB2(0q
zZ`dZ7ZTUSF+tQl_`vhrU|1AtuhR{AIrv?D{6N9tHdY$}MuGRA%E#_ky7>i4z0o||0
ztZT0^Pk+WtG3K{cFc0q-&WDz+E)^&GA!fA)4zhiHN6Z_@vgcZx?YQ``FSSbNiXCCG
zKI^lLD@d8HtU53+rfZ0lAB8k1)Ku1bz*S69jfcs9Y%ivS0;+18Po@7>E4PDGe7RK{
zg5-R+&BA$YZ?DWfW)_)a=6Hx!q(v?aPtEp?n#&8Jj8}%)BY%5T!3XeA_x8&wwQiPf
zVvHM3-##$Cpr@@iy8M$Qe80%>7q9f0{0W{-Lgsg-oEdVzR}e@fRR=trrs>*DTKfY9
z7N!7)V|oW&rG(%9zLQTUMO+ix`RTY(X<MoxJY_O_7F9<(d)rSeok-*NTy~r7d?xR$
zy)_tGDxL(+3O1Oi1Y;tqaZa<U%lyfHQEnv{7IFbIul1g6m9`dnr|UCLp{PasJ*$*&
z!yxBS>T=93HjIe;*iDd*+3LDY?Lv_%dr1Q#-C-hbJ9AH10{Ai^JnKVE?O?kj=lbtJ
zKC*qumX23&*F!BUHPUZ1wGxb-773S1@^H*G0i_lg*<gm&hlDcLtTF|=UP>~<7JxM+
zcn`E4Wye+Hwkv(-DRc(l%jyh9eKa5AvVu;=Fy0%;(z|jk3rbhPq|F!5>es(j@$T^#
zP!%PO3zEY6mDfZNiT37IJ2Q`J`38>MD~Hu${H)XX;$5kaR<#ajD@dcGHI)O&tFujd
zvde8fX6+VJp+h)@wHBlbEcc>Vyg2ZDOLo5Q`~vU1sLoiYSvM284zY!AQ@gber}L4$
z;Rj>gC}OIS(GF3Mt#3-swhkA(^(eIFhfFMuq-{;E3+v1M?vZJv6&}X0sFg^2t0T)7
ziXr?ZA`G3QZ8}?R?!BtMlu0>P3*X*Hc^?+`9_G$>r<#U1a;zFAO>$_zZ<u<*_V?ev
z|30<;K>Ih>SuR#74?<aaaVbI%Ew?Pa*dkhD9yaTkAh;QeNY5U9VY_;>DIoFm`o6On
zzz>eE7vq+jy4jSTp7Aot-sx@Zhu2rcTfcOL=&Oj@9vu9+F`y^Y9!K_QnwWJdvH5-N
zv^8*eM~M7OK(ni1Aanq9?Oq>D%<XNq@eZ^$$tsBu2B+qKow(>l7T?|s#fozY9TZ#6
zL~#SOHudIBHRj@LC#}i9e>q+$Ww^(^Y)MLYc&YEEnT)tHF}j<fYV}$(H=UC=gB9MO
z?Dl-N)^-G}9U038!lbG#mWiM7)y3Termr8qehGht7rsugZvFxFH#%RLJeBc{d~p@?
zfa?k<%w&TM!EH`XO!{ioASrjrL*W9G?L+*Hqio)CX%0ncCTP3<1`L0-z;_Q5R=Yn6
z6E=Kou}1pc<pm6o=k@^2bwN#&^kELuRc)opq|rWfn-Mv{%$lZ~FsjN$U2m#fFtlCK
zl&&cvGpe03LKvtr?$0oH>s32{*+Ze$iQa8=)V}{khtyU>*_ls^BO*mMM|;d?#l;Nu
zr%EMOtXTSHqV%>S{}^AxAdGtg#u0wgH*`Fc8Yn>iRdbLF&8O~D!I&}T&M#!kC%uLH
zTj+4pP7BXp)kTA1|02@&S1(~&Otam!S(p4)e&MN!_*6MaGhyp5QuO;ejxT2IdO>05
z<CJF_oVt9C9h;O=PtEL|&QJq67)Ki{Z_+@X4BOU93gffG&(@3k&bWc2bf7%O4(r(_
z0p+Rk20*oR>(dIe!fuEWWOiXQCDxmgUDer59p5E;+JWe5gFoF#(1~yL?xeU#DOS6)
zm1ZT>wJuGCR-yULlLNF+6*8CI>oI-v&1kRhVrdEj)S9gRoAgieWhE!v2>ydeFito%
zXjzk*+XK$b^g<(u^H0;&e(rhh%^zI(38hs3tuZ$jT%sz0z1`esogZZL>#dG0Rtn}<
zAY5h<E~Ga@;J>x9p>6K7y^Fa!_sP>(c_rU5?7)kte!0+KSfHC1<d?P}yM1l-eJkt%
zE&)I;KbXjLseT}#x7X)lx{$4yB(P>_(WB@1-EZ2mxOqZC`9Q)59QD!=3M4U0ZP^{|
zupB}kD##2OSp>>UHN3zHW7A>=*rcU0xUyPSVz1~k-63h&g7ykY9*Xg`4wyXwZvOc{
zH^tU_B=@&rrdz!~i)U}wWgnqG7laup84a5UIn32E{hh1bKFQ9u@0807>+pg{MLOJS
zuQs}#z;X4IKpbfjD9a@hmip>wzo`=x5&Ss%bUIe6j*5roYwb6nIPR>BS7`zo;pFh6
zg%LO-7`?a<J1mNsTkN28mjLrwt4rqDAy-D)Y;a+fTJbl^>!hE9zkaC%9n8%GNca2j
zd$nX{OGcQ&SyGtjce&*|Jw<XcFDJu*yA}~+=08^3^RZNHGGYDktct#@cXeA-?$eNK
zFGEk)x2I(br$+sCv}3OlK<v|7GrS7g?ekL*THKn6Iwcbxtx04={;icS(j~NcbMNyS
z;EvUKzU=ZybE_>;jn+70s^$Mj)49Jh{l|U0@{P)=QVuy)aww-HXQonAVwD_+$sBUn
zmNPLbAvuhk&q<LSX4po~r_EV1#^yZCX_&(t=5V|3>$?Ah&kvvLdR_15^YM5_fh6-(
zmOa1U>Ga9`_IWI06=xvRx0tIq$Hln$9dLL^dqb-}G1q(?{#aC<FZ&yv05*gf0+B!s
z6%A62NL|B>lHcQXx7V+yp&%i8W{+L#m|$|`y@rOXp+$9?t55q2pEu>&2>9yi51rNj
zmX%w<B@wDI7{z0aoCs#yJbWonoexg&MWS9;r3x#>+!JCuZCSR}d*^$pZruma`&ToU
z^t=K+NuO-|?uc;HkoM}DHi1iY{=UzF-Okvp#>(`}P&Y+vAA*xbMVwQ+Dgz0ba2}2_
zUWJpJ0GuqW=HefeFwPVGb`lA28QI`w);|gvN|IJQlO=K4gr$=rZ>PTG^jglwHs*T>
z-6AFiasx_VwD_B6B3`PR5q;7aQWr&+T0s8V`PmkjL%faGcGx+n1X+u)<aJl~i9+tg
zjdYP`$SflTlh_nU3?WN{YMowf{E76Ub8Ug~ObCPi0e8Ug@(XD~<19VpI!g;@*G0@w
zB{S`A?2bOmZzW3!zu97vtwLs6hh4VoKiVhQT-6yGy3Io`tZ<lMI9kAp_-OWW;hGws
zvpjwS!B`XLKoKtk_i6;DIzPF{UN+}IDc>Vg&8b2nMTPBDsJw<ohk*0VqCZ#qYBpow
z_^B$VKer?%dbz#VI_b=ZeNn;o6-H8?wRZF_hz;ahNH6Akt(1)<&MRA!lEgbv8&^~@
zB&BqdkLKUbs~EK|lEaD<benaP>$u@Mmh#nWkf$y<KFY)?OLc@F)vb}>p#=NRm*M4z
zhiA{YE&#OjmpDOXL}f7#;!a<A7*K&a*jqI(4JLo?+4~AWr;vs8+MJ%5`ayLy%Iln0
zl@Poe0Pn$73<K8ov5s|_(!HCqBQM^Gu22to84kloSzBk5;!XZ4jSmFDp$<Efk485G
z1~u&6jeOk3^LY-68>`JH+m&L}Ih0pbcsDm>Ry6<`|8kf&yReYk#!Ic;<WaV+okrql
z&ob7*2LPKYjn|SY8>z?hcdKZe(CcOmd&yH|a2#Z>Ryu$|{0jMAhoKkQFc~$D9L_PE
zl*ImCz`HJW1II+yp?Bb#X{MF(kvm_tm+=U8oO+<K=mbHCdz9a^pK{?1CbfkLuDimT
z){-1slhBSCr>}qUjf6%Zm9^fER(Qlm1S_onr7d;&qtzLXpqX54s9E*=M=o8TlPyZ$
zaO(VP-JBtck<xGDEL%tV(nTySE)_ZUVeNJIcJ>QJwpyEgpMd9WF_O0Dz@aIpbJF;{
zl%gaB;FCLgfxkPN;TIjnzVdY+ZH64Ncy#19|4KwkmAg?uI)Cpi{`ZE^YFnITZ~Kf(
zo_*=l>#H3a{GahL7iMaArXGZ}mh;xu1$<xi*(l;&<Efk=;w@RLV5`QKu+!4&(}14&
zA&2Q=bhPlakv8m$*VooJ3)HrZIh*0jp}oO0xahP{a#W4)QJ2TZ1<0G;LRLprM;T)<
zfd8X{8>5SxVbSg)z{m9s32i0w8)?B1o(>1+M>hK1n=;K#TO}|BDP@l;t5&G9>?r{!
z3`I&7>lp);kXpmi^tn$byLuX66Pm-8_^Pv?Qzy?zUJXh+vQ?Y3ZK~xhegAGco!$|^
zcHLw=0o!_}Z5qFGJ1^#A`k_DOvyd)F<|Fp#N67<4Z9RgOpkFK|TbHl*$x@b!#fHxx
zbk*Ch&h9dZYtB<D26_X{?w-?EQkAk9H~QXmc{EcjTp{jvodvW{`}(LIbtdli9-i@&
zN96o!{<oT1e!b8CM+Lr{n(Cu#TuD!CgH6Qnhn2V1A+hN4G5Ha*WQ=Ni^YY!`nP^Am
z{c|?*GvRc!DX~!vP}?0Z7OI@Dv2Jap_^W?tHvq=yn(yJ!Ylv>|%X0|4lOvaHLZ2u#
zUn5(+klWKNFk7GxsiacOXkQ~H-^U|<(B31yiBg30bn`1C$$gyMdRZI3b@}SZg_nw6
zbppDC`$5aEH``X|O1E*5cQlI8hMC$ai<Q>^on~T0Y3Ep-*~*xhRH!t*?P$g9UJ;|A
z&5%nZ<UL;vx|HJGo|v<*!bbQ4TiQA6w4|Py*(yEJLHO}OJc3<Oq{GChHD3hmqTF&L
z+t<tM)`8XA5v=sozxQ1Nsyic{Mwq&MMZY#%<#5*lR@?2qM<zjhbJq6mGcDD%E5J~}
z;tg5)P)Sv<n;zrJaH1DCN#nmt!7@MHd(M<6m;S;y!v%p=z3ABUHWLD5i#@?BCl3iq
z&@#t;3)O^rgxB#lNrRXdM{J*F+0KLR{T@JY0OqLRBd@>A&eUiNniRyltC^Dq{|2l-
zDfdWhG1gG`qg9OhE1LwC;XRgMuf0onET=VR0W#8qQFF_(T^buJoP)38s%|0ddOB3&
zlY$DpR(=_z?EL}eE8x7=Zy3NN5Z1rZ&c&p}Oq02XTJ~SttVA&i#Hj(S&$Ke>W)237
zLofhssVfi=R4@dVSEGM-nG*bu3x)9!vv7U}V(HtvFCqW*-d7@O)c5P^l3wQs?g@e5
zYS##g*oHrvDU<cs#@D07HBDdq11+5LfzMnFBI`gDWQ%G4S-T?JYl2_H9^^`3sOTZp
zrNa_gZmnsdbZsX|RVq*eO6UgI?ZaAhj(kKq+>C{UQ6=Sh<U1N2=-FX6k)hbZJ%{I4
zD$q7x3^KYapgwIqm7oK`8arC_M$L4SOI`@Hkbo#mEQTU|-oo3s3N)9JUW-I+g-F4-
zyKFJy-)Urap4;vb=5&rZAWD@GpIday4v+e|^sR>-B#&xOv=La%<qi7`3L`<bvg5Kx
zMe4mw{L`N4vznZoqSeo5rq6M5hwRS<M0}uaZRsL@jtCK+R85xm!Fz%Z=usKRK;Qhd
zE)J!}qWVp2p)~ji=Gm5qRMLbmY=8E9!aKr)oJT)X=u7=F3ekV(ezltDgCqs=_TW?Q
zCkD)I%KB#_m=AepcS(w{`lG3>k*Z;yEu2@}i?11n8*u$#{w<deG9HVb4EB)$AV|zM
zR()M9a<XUi-_hz}REVC>q1}sy=Z<skiClkLl7GDk<U4ag+n)6#py?Ww?LI0an#oJA
zLGy2Q<Zn|lL{7wA5EA&?B7Xn6sHZFB9+wnj^BzRE^NPeltSOs5bx1^t+QJ?uruII6
zjqR9q=V^NRr!UsDe3(&ZS0ZQXH^M5wUBF+ww>=mfY=4t|WZ4D#(By@uTmhY!$m8NK
zYB6Cxe*dr6*=s2A$+=T#%SDIA@IM%n5AwqUd>Ui1C&Fpyd;em{lsF&QBjPYh%IGhf
zOVY_-)-x`F=kDl}Gx}Hkp|c`4beP$3A67l?_9`{ZyzznaP?TDIV&$H7#R)rkD80>N
z6NkOC+Viiz=k6FF94xf{QT)e&?6hXzzg4N%dJ}B`kqpWJ^GuCoaY8FVVPwuvez_vx
zT4|ktFru5!YG!417^&@|<9e7rS)$eTecB?y8yj!3E}KM*zeN$<<g4x3?{g3$rv)!O
zk5|Lr?FuAQk_%|Cg>Ee6Yjd_NC8_1GxIU5Ac{T#wvm<8Zd$XRkr=^uVo-*V7hUW1`
zwsI>~*+APW__0T+=)VZE6t4<0Gmi+_r}`@&+S>cSww@Mja|e@!H=f}H8Yr)MAzZCn
z{^jO_yk7#XD|VO;bJI4spR^N@C5qkjPJe>O3nFq?dhJ9y$GZ`!h<h4}Le44j%xqnD
z!&<dE{~sAuw+nkJ`%-Tyunv$a{2v0I93aS$eNv9vXamM8aK`@*56z-;g<`APCBGrR
zD{pt98w-MQk}vfYUe?9;Zjn!XH&>M-j9#tp2Cje0Yp@mO?ix58%aj(AN9_KAE;WV5
zFcfqB?)WTUtGuZZvZv~L&YP^n^93SYw^c{DT190~*6H-mofJ$vNO~P}IxY1=(_3S%
zz1)~eF=q}>M|vu|@os0`o=1>{l<yFD1mq3vCT9nkesp>d+?t6)Ih2OcXH~By;THe{
zxLHrz53>Jk3VZRBBImtjYtIW7SS-Dl(cHD$pP5oP4Iub*^lHEF!usN!N-{NZ+y^2#
z{!6U`^{=#<^r^?3`$Kjyk<4NuU8Uv6a<tav{@#Eq(<T*!;m=NJ$KczKL!>|cd;vin
zw00TlP)Z|>a>#`ZL=FPh{sMqDnhl(m5=a8&#CCbzw#L8Vkqu$a_R_zK?4gZ84#QLw
zr4q$@f#Qm04CllAHYk)G7j^=6mgM1GvmP#A?T-6ktZ|2Z$^KsrsJ^v1BeSzYb>;ZL
zVwhzi+^cD?qlaRX-em^%)qSo=Mry`a2v+uUsCE{|_dH$yDPE?L;*7nVqpI8x%x{>p
zK8J=mah8H*71@=!sI=giN4jGx9dS9mAM7abh8g`qz{VL;cM4$TB`Q7(1nua1p^I2G
z&xt8;1i%|)-C?>hs8f*pghw$V1JX#biQ23n9_RaWU128sDIrG`!1G}E+`u~Pj?-jK
zeo=$Y#&UwRS`Ien(+66VGxcohXe?M;m~cZK0D~S3T8UfgLK}QcIL$#oictzydl;}}
zT#EVK^iK`0yhZB*l3$HgC<5r$0bc740~^SlM>Ms93Hj>{m#TGbVeE*D6!rsD@YY#V
zaV<{$9NdJ`@3x><y9otI*ziildJ!;teE-^|nLKc4xO?VKXqH{T3w_~)S$6F^Nt{my
z7*jGO=B%bc3x1G$?cInP34!nZ;_s?4;)n~6K~J6wMDyNnsd;y=XiG?sj4I<<oysq+
z_|X)JRe5S@*H#<)#}y$Q&RJiU`A6hx#a}(*EbX^r*MC<;#w3qihuN2WEq;g=-50z}
z={;Sj*wb<;>1PZ72LWB2%!Nb`9u~~zRY(2s<^GbBDgXVDg)~QQRI*^1%tBz*1&eUp
z4a=~NvMp8>e*jf$wzh-y<h*Ij*x^gBQ*_3^KGFlq%@cB(ko-VsJkzSSs^Z|u>!O&G
z{c=1G!>$e3GK;5dX^M}i7Rv8J{c0upUWI-Gc36ZvAUGsoN#k{xwvF?n50&++MM)~b
z?eW@kIzofoAtQ!F-@v|)JB!%mXMy^{rKLhk-QIAq?zNi~PA|>UyN6-eV~jxfxF7=B
ztE+vzp+Iq%^ODNw*%IQh#CPqm1VLlGs3=d>mB7A^Ge10s`%+H@4|zx1Ia8ogS`z*d
zu^S<>##<@Y{ybH8VY>eXYO|r>8}Cu>MFz$-d&8Z)!<9O#jo4)6(OU{qLZ<JZDp!}H
zf3f1JvX#yng@4EmD$LW*5iYnSy)BSZVaR9A4PQJ?=1zFGEPdHrj8>-Mu`)tPQ60Le
zoSbGXw3`P72O`z<aBqZ%{rONvFisB}Tsjg0Vi*JmeJGebl7koYme1d{(-E6@X2`|a
zHqlZO6TtN1reS_(=owQW*s=tJF@rY@rKx%2^LSmErvt*Xg=3Oj?Ecg*O>Z=PY)jd<
z0Lt_}7CqWc)(BnpHDgiAMjM<xmPb01H-IC>#)t0a%PIip0F-vh-d^<=!?$rOG1u&z
zd(QTx9lj(4q-&3eCAb_*ZYdSciopkh$b%UZ=a`k33eJGpA6WSU%sOmhJkzV)IkU(n
z${G^py!E^@o)MV%)N)JhHg!ji14CCfFG;YTZ++>Nr^WKL2$jC?e4ro#e+yqY<y-Vb
zL0N$kXl=*bc-LTY2G$2>&+!nh*pug(9h!F282~yORcY1nPgh=s)2&ODmbWlZY#B4B
z5Q9@FIPSE?7FojFxg_?A<s67>yQ<2TCK>*xE4}F%N6sTL76=niw8ejzBDK9VkX<Rf
zTPsuD-ejK9bIvlPa&j;s&z$}wZO{8NkhSJ8Xg2N)gzCUJpWqF8mRzEG*O&8r;-{C@
zys(*XhFy=|Qs}>bA6?in4NT)-+Bf~0nP#{}>`M;IN0J2@9jjdFp=jNh`|62GU41xi
z?dsz~{xVXj$?@6dR9nIBqh!gAXptki_c|XRmml8HE3h&GhCqiNE`PejSZUA7tBW*6
z3X5S{gpzswBxG9j>VC89>TuV|EmZSqdB+v7uacQQWdNW472#(^wg0rO()r;p%`Sm+
z%uGm*rfp{qr$OVsquUIa14zizk+K(e0B536E3Vox?Mi4HV}_zL`S6!iNcz9d9jmf%
zyi<G!d1>;i7cmCo(Q<uajj}KFcAJF9ZbqQO;p?T%*ogirt8WTck!rE~MIZS}Kh7&E
z`fA%27&!dl9r<oM&H2O)JG}QH{=<eV_(K1BvtT6klHbzyd2)t_pUtY9ruw^JD--RY
zp5LqoLpkLx_#lL^g~MU}g|kC&<jF&)4r=M8`V)mL(J=)!QZ09c(a6y?SCp9EfX{yq
zSJCbVZKpZrJ@&7D31HknCi`+H0lWf>8l&qAzGGyG!L#k38$X)>0<A5^%;Y?O1dDK*
ztlx&>O$c{4bT{VIwXvc1&FBP?E1F6~(j~+V{=+L?i_jW*df(!eB`)Oz8)}G?-xFrN
zzv3=pVsE5ghk{+bNRacZI9Rm_8f_a=qA}idk8My*dhg@o4r+D`YusC2u#|1Rl2{ca
zM`i8uKiqC1T`<EHw7sD&1%KI)Yi`t|M*f!XkU&O296@AuYvpf=1j+)@?gq|0s2N&o
z%5`3n&rS=LD$;P`xeX&Igv=R?sHl(xk{1nki)~As$l~BRrUY+o^y*7n!43bIKJJS4
z*{7m+Y^z11?`QgUb5j;w$FA0`O~Lt<nVEVz?2~GQNc3S-rvZKRVUA#5$6BL-Sh($_
z;w4&Je@26|oQ9dYf$mSQ*tAQs@;w-YtHVuk4XZSwWnSTKQp70ssICnBMh^;z@U1QK
z%OFC|t#@;pk#8WO)(HWw)yZ!)Z%`Ir%nWt<s=30+X!&vnu{}YP>Pk$>t>f-`S>PXH
zhL*#n<)^7OE0kzIUeuUMDxdGkiDRn?t-P4xnbZxcQ2Mahd4t-n>h(Uw>z2<*xHy;3
zUbzkJmsb%?*uK17F=Cwz)X-cm9<+3mRa?)@y9pE~@cLp!TQwC9uBHqiAo(CV$t$e9
zjtuGGYGK?!Kwrn{LPNHoLCV&hf_;+XhI=y=HV1v`-9J@HLn0LkiWE<x#ux1aN02ZL
z<pti224`UeX<VK`-TQBpi`q=OZk95WCVqzNiMteZYx?l^I~~^dO#^Jkh{5=^^SjUL
zx5rAs1pM%#;CfMytG0^3=FuSu{K{f=!^idQF2;2<FlaDmP;4zNM^S7dSo7ys;Z+)E
za=rCpSqCc3FqTGlTf7)is^EDecxY%m$ilo>ea3zqi+XFal3;=d)3^8DzO9WfvS$?Z
z?nPmB%4L9Q&}`Lr(-$Xi?)c7hi@(xec@!Ehq8>1NmJlp9`nLDJm?+y6ULE0>*&W9_
z^7#ZScGV+#?a6%tR+up#v?WZ-LY^Ts1P|0sMKthh)w7!C%4Pn^pvBR4%6K&gd9hxR
zl6m+(fA4(gN<(0ItC9Gqi|iZHr_YRjr4pTbS<HQq-0-HfD`<==FVadpkN`i$HI=rX
zNqKZ$Th$MtJ;txw)lm9I8-f1hFtEY~?~j2Q_=dradR-;k+D_yDSu#ef))xfiPC6As
z7`blgs3{ax%cli<vg*XlaC8T>pQW?{Ylw#2MsD{G;&SXxi9#NaUhIKh_$v0FFSO|w
z@282suTQCrd|!fGlr~)D^^5je@TI?xP2CH6(9o@XVn!)yqDYowPYGrID#tvAnwW=u
zgl`^hzxnt!Vg~`aV3HwWq_EoQYtv=AWH`UkKdMhtdO&ZFtvh(9$P<*OcymnJom`Td
z<Q3#kY8Dj#7@=O9+L3Di^rt_r$FGrP({(m8#>}z=HR3zGJjL|pjoT9;hhnzvgKtYW
zoVZSFVR3EcGBf-l<EfqFsyXxo<&Z#;K+US(!M<{tv#!3~jt!*4Yq)Dlo#xc6S%{&O
z!J>w(<C8=80%&JXsK)nI6|vgIfe8AD$ICG@giEg-VLMk@0fHT$erQYe=F%8)?Mp?N
zW^H`yT!4F|x1{wvlLE%}VgIC!vA!<UtQYK;LK3@zq*62;YQNUNfb>>Zt)e!wtbJ<^
zPo0N3+NgC`rRlHmE;Y}cTKrd}l8m7fGsBQ#p5y6)+y{_Uh9O;Us!lUI|HKbOOqQ~}
z?c;hn1-Gk)q=sf@t%O&sW|ng<^Z%qi3&LUwtvA~&B{nMVFS=zM+CHgoon?-Gk%jUh
zwS77a1A!@fn;H9duDbL$N25<gl^K12f@?r`tk<)OgWFrZ+(@l7)JovV7GQQEE=UrU
zI;vXOhqZ7Vtw>ASEQ!G!r>eH7%dJ)rNank2#%G#3wm`tIkU4t7xB=(xD}tu`5^=v)
zu2;1NJ_(6IMNB@{?MR`^KkIZMK^uK)Sz%t_4Ef+KqJ6DY#==W8N=4rV&o4h$wb!@3
zi8dhfUN<`PtxAgWG9Cs99XG_QcBz*N(NqLA6h+NfE^o%h2Zs>>o(_IOq1KiSG7TP7
z_=(+DNR{od96aBBg%cfHtcD;&%hJn2hLbKu(Khpl*;_=CV3Rv9x$?Ck?a+$VT4x0a
z4JOym@$9D1)#Bt-B06!dIAP9*3la>w;G=>D-}uytqum_7au?3_Hq?Ls*&}zs_)LJ#
z`?uHU!g}unXZqK@exJ$qJXnT(m1-u^>OU1FJn)|az-P?B<XvMJ=_5~uG?G*hVBz?w
zh)yl|my`O9<S;zRx77jWyGr|h#)h+Kn{<JvoH72sq`eQ^=h!r7eM)-y=NT*4-HU0^
zVuR#!K%{rHff0DVEKi*EQ>v0j>3zO4SFEV0#!qdRiATc%Rr8Aezpabk7jO7*+PNX6
z6crT$vnd%1FB+^eN98#0$J|XE%Z3`LcjrDL3246I!!|2@lKfr#py|7D$m?ZJg!Gl~
zus_3hA0wL-q!Qbl!Ovap=-r#OaLGwgIs71ma0hN|kAB<Oo9Ux6$0n+D<^NOBJMeur
zP%@R3bApo&{O8z^CrF(yS^Xu@{={OtF5!AvMbyt$8%k;Y08DnSbw>3+T^*?(O-aYU
zN<TeV@==72%SOSRms@whGg1V>h!v<p1;TdY(z=LUm;yfK3KN|91c}F(N~RtadavFo
zjrbhUb(Fk+EoeL}PgGKEYU__PpT(u|yE4-i3e*XvwzqXoYQbLa1JEYOdhwr9W+DS{
z*98n+S5H!H@k9W#098^!gTc>|UZModyD+CY2um|BGu8_Mct=>}={`dZLPmE<RWasC
z8Wi=K4iZio9h%3vSBk2T%Bgp(6}CojJ^tFhAD7l=Cs!>nD2#|{uf|mk6JA>Kado)v
zpcqqvzR8jCC$X!7py*VT&49ix3QDsz4MKza3a`|5sm~1mmQ_Q0J?li?N~6u3>_UfI
zceK>2Y4N86-GZ0j6?u2J-dqn_kVa9WCAP02BPK^x$umlZIzFtB=8MkA!a=e`(P^&3
zrq2bXh~*grU}=o7sGgeB9>&4dQ_BOvfvm@P6`PhGN6Wt4s`S$8*c&T4<HY`6k9?FR
z2(<Jitz!x(K2pjYaSlARcw{~_2=9#4rG;AP(j?8sY+UI*2^cR@edaVBbbQcwPnWc<
z4uVL)RtIZxjrcI3uiKn#T8D6}{^d-#%PeU9Hy6g+mxYb@;-1+;TEt#nN$J+{EQmoj
zhlvOip0Hk}uNO~TI<5$QQjGaRsVEtK;DdsBlBERv<Wq@=)*z_IYbuu9EaMzlWt3<s
zT#8vQuFE&UMqJj*%At33nQ;}kMLTlTChH;sdV!W$RL^DA){g!9SFoHxGfrlT8?8CJ
zNK#Ukak)eX&^=pB**drA_i^6*B5SDqs9MKIIjW)apFUv8nZ7i5@u;Sc0Pp_=%(}A(
zZhOFS&~Vg~qIUGx7W63l&%Mg9KJ~JIf!HFs2Tc;BZTj|ScTh_TT_^<Kn4s6)Q;CLh
zM{Hvoj&&wk|1v)Y4CG8&APMc27^pY|>Qb=34MrgT93A=yZ-ngO8rc7FR)y?jfXDWy
zGH6jVo?E(y;3Jwvx>v*;UPBEzGG+02s<ET#k|)nS;z~5p`Vy0LodcMYAapd3JEDFv
zkT}PTe#K+Fx!*F;slmk_9+rh49A6JD<On}hI&CNW%6~~=vgN{0(`YvFS{ZUhq`-#B
zGXPX+=QB=8MGGkNwtA{2C0`HjL9V*`p;rKH=awS~JZe4ZdODiH13MuhFfS+v5M1@2
zc4&W>bGbbs$emYxubK0`nI_M(zzb}ejt9+<sg=_7-bcdsb|BY&(c&5BWsU?gR=#nJ
zSVh&dT=wdRNe-K6^d{i@F;RaPxa*3|YUpT-P1XDb5eQu|8K<a(O6hh@5{iY67b387
zC2(tQ|2sOMou@%=%EsWy=(MmC<Y>nkH+7Q<7c1vzAFiKE$(iU>2`ufGu-xu4rJr8=
zBbQ+Nv8FumPg&JPUHx)iSWCU>KOUb4dm=TC=yQ9`PbG%;{+3-I@pMBQL50BxT?@O;
z1sFzPzR-%+aeff+vZ{27P45|T?gl;aSfO(;>{c)s$;~Kqfz3YAhU$LH<T8$V$dh}~
zmGWj%O2Kw5K;)6Z{nenAkN(P~u52MV&Sf(b$`uk<jt6|9DHSI)_jUOz(vUgtEHdY6
z7{a|;WB2c<W^Mlx)_VK7t1T`zWEwm7;hlO<)W-R=M<T(kySWcUE!+oD+?M)?e>epX
z`GWq^#3E$D$(nkXe^R>c`63^K$A;2+q20W@R04rf5$m9o`f6=^=~txjZbt!$jngbg
zMW66_RFyW-U9u9Fz-B8FICH&k*j6*;vl?L`4=k}Hvp9O_4%vS`bER%0cu}3yGKE#<
z9d44$K`hPOOAga^PfC+QxaSD>5>7PuDjRPq2Z)MSS$5R^&S4bRyjuO?Kq4xg`VQ44
zpbZ+ef7}l7cR_Y5<#TBqKMwpN?Ayh@fpCjO`{O&bZ<yoT>t&ton;KmL$cN{<zp9h%
z#v<ArV(J^nl*R!NId6njJLMImD$4pn!5=6iEhqc)@_Z-mN^KF_gZc|5uVla|l6p>H
zvA6AsY$_p@BZ}JL&MW;*_g{l3Vqg0sID!qukM6`J7U4>RnPr2DBE34hf8?}EjO0MI
zZ}SyUxslMhioe@1&tJe5d9fK44l;=x@M4y<iQ9*dXo@X;G2G?%JKCpk@nvYItjoeh
zaqF30os+1;`MiQ&$OqNw65-{_v9_xjNx!9~Y^#mNXie+Q$of`~;>O+wZX&>lhcOuS
zo%Pi4(bS8E83L>-n(ltT{$b@oRV@L4PV;*ZYA)=$7a-=FIcy>b(p5mhLJ0agry%ss
z4v_te0o0uy?9Ae-NmrUjLW-e>Ie6C~aj`oEv1l#Rgrhv|-hJVlsr>gs0_CBcQZr4k
zco04~s+S>Z?(DC@XcM^>Zgt0eQTAN4B7J}-jTW?pDXsp!x6BujYo;4XsRIsuRWKlT
zlYMHHyCp60XTeOxk;7MeS0oVG*tA)QRFOxQv&|NB_T3?Ku*dxPs`31(?Gb1_jd<`_
zYcT{CC7;}R|2TJ1Bq!XJJ8Wmmp5X5U+*vaX*EYnjIUZ?NZsdJn4&gV;wGVzMhWK#=
z)vw)*P>C|#a$A@=ikJu<X?PI@fS4)piiB*pIeVjS9kn#<DiVT|wC>)py(^q3uKJ9P
z<34%#r@Sp(+@0{@g|S4aaBT00{`6TgXM~6~&V~Kp>fgos+k>pKX#1ts!tduj3L}4g
z28pakC%<!BxV^08$1`F~=HtDn;Rd4Ns?5_jk{f?Y@y_h_dzvgVxo46$g8#b5tCr2h
z&G}YA)fB@21x(L&B=FNg+MIC#*OReX<M}nU|0X8&n!x9Ht}6uV7u6~L$h;lbb(#b6
zlPf1`<Mn6k60DAK7pPbWCOco(qz{UC#>Wdd?Dc4?!+x<o8gKXa^g(#UyZ#ze)wcvF
zCNQfUC_D%2)sz{N9`GLhq|A037q*x6JIUu~<ZGX6EEV<bjmsKVoqg+O_YJ}Brpo3v
z-%9^1Lx4f8l|6%=-TPKt8thhy{kS?;4#l0tP=GqpW7s5-veN9&*YIpLL4V@h9rjq>
z!ah}XTM+}%hEdNO0Lq{_Epqgm01yhQ2_AJtF9;wS&fA`V)PJ#EYDN8{5Q>O{SejJO
zMb8*{hG~kpVCx5^W0~Jeb8$cP<-kblSKDmyJXL<$Fmh_69uTcUM<v;mKdGY&Z9gN3
zNQF#h^-T{O+~VCZ5&1)ua1Lqdg>YF%##8qGVf9v3*y^YLI8@B?Y7p3W-jH}P0nd!Z
z?PmXL8P?DG#cuV5wN`aUq&YidNCYckFzt-lAzr`*VD#a^Z}{Rv)FIwO^Uh%f#SG`s
z%hcg`3bxdOm9jY<9YZacwF0f#DRFzXnqT(x7XZhN6Z%Y>mfs&|Tq8MY0Whrvel~l*
zl_E%u$&7bUukIf0v(FL9j5Qh(*P*hnTBF)Dxqdo)I@ky#OkNW*FX;!(44CQmI&b1{
z0rwK6-8^ufBli}Ii$vFn7tEP_sNP{6!21`;{*+k4&ZIa=N5Z)s(}yq|DId|Y5i^{@
z^+KZVxCXpb6_yCT#Z)c5ZZ;GxT1dg-gm<B(gUo-2S=bR<A-wN#4jeMiOg@TOt7x-X
z#k>Ni;OpA768an#kky7Bw(r=qX->TtcZ+AAKBJkSiSx2j+t+sf70k6t8w;_*34)G9
z{3mg8v?Stsk+l&q3Om$UGMno(ao$<iqp+eR`_IHi>dkaxv=6yITj!^P`(+kn(=BrZ
z<<XDYozUzZ9IkUJUdMpzxMNr^@sC({COX}t4t!Ntk3FJ?FSC>)OGNB~3X5og?t>j0
z=eOkQg;rT>B{^<YZkZTfQglJ*ym^|3wzmzW_#myp%|4WNMS)*An6*68VFb9Gw^p<Y
z-gMgQ!B)9uG<ZPd$mpt%-6&%{1$}_41L#j{4_UdsHP_i=p32-DpfXv`$%ej@A`z-`
zW-)C)GyWGUJ;o*0k{Ucas?E1+M#o#HuO{CRd!})JD}xecz$2J_Zt00qmWL&WxFQcR
z)$2qX|H_Mk)5Fp4@2eKAgapOSqZ_N+8(T8W^i+2oS8LC{?3b%C*1c_Ny`-qOYS!}o
zqyG8FO;-|pMSCdr_$s1`3v#Np#x9XFXeRxk$!1i>X##d3D5gwt`EIrKCd2Einw(l$
z!0U0!g~x<-{g0q24@R&OZs!zmfoX&X_f8(wRsxjdhbv6?LBsr@yc<uewSLGqKF>1@
zuKHa|Ae~a*@H*E%TnzluuP!f>dpz>tt$kJJ#aL~fJd(DRp=^V-z@Xo?us;zu30%^!
zIOQPHI!^5yy}By=DkM};#4$Q#rXr_UK)Z5hxgqaAI3X4~z|8_tdZS`XGi}?8t5~@9
z;h)WKn*yE{`cp0c{Toa&&l>K04L<k^Xe+SN3BtZG3)vi#o4f^P4)EI&G&ZKxxz1!)
zny*CMTw5$@`Vq2MnAc$1$e2DwWOVrm@J!ohdd>qCSeTXime>Q^hTcBcJY**{8-wx%
zA?zCr3k7!k?lwdfZHJds2f||&&L0u0siXd+a=$D#lFCNMk{oT_26Db3se!fD95$xH
zpoVn$;o0>?(y;P#3L_IW)u)lxgAw&2lwa^X?eWS>qnfcz_ig>tK!`<i-JzR=i~tP3
z^O#c9S7uFeji*B`X(u_vO`GJ|?!|4}Etvz8U=X(%2~3O)T;22lv8Muql&%ZmEPR+9
zNs2z0(tWa(`F?TEaHm>JfgBW-6+!C$P=k+G>X%YL$vSB|*qS)UiuR3L+`-l!KBy;G
zs3Us6^GpP4JhnzA;LLspJ%%?>w=b$5eJi=M+-8Qx94TNuvz(0<YU}UzRu2Ld$Al#y
z<N>~eq0#XxLP1FbORv4&=;Ct!%#>(@1U=>hswb}E<P~YKD9vef$;wM^Zi1hCXnBUs
z^quiVJ+Z~~jA$*7q?URh2f1KK!Lh;@@pM4}t6U$EIH=UUm$Dvm<o_GTQifXIQFGeR
z>l<>>P54=l-CkRC(s*UQ%iOC#-#B((EhB>50t<s>R)l~_{it$7hRzSbMivyY+=eX}
z2sU#wZ}`K#JP~w<|6}XbE>*-+zze+wh5f4ov8d#{d0>D*GET?l!+5n@a88i+banFk
zR9EY_9*Way;uF0j{?#K=FsfkP{k6&A8RyJl?XxlBq2eKfjWK6|L6gUP`taQEvsD)?
zvwSoC3^KSm1WGLlZIAA&Y53a~E|99uq+IObTj2^h^%cuR2}ifxP{@zpQg6)Nz}$@x
zyyVf!J$9cstJ6HRN)&va+*@YpqbG8~p7g05(Z2|tp<kZAA5(s9Pm=KQq`|d13&l9J
z;DeBknyA~Nyl<|pa^5V7`K73NHwcacP#XF^m~FRRTgR@wUI|qklV%RfRYQ9|&_geM
z0LdA=m=)CYTYBy%%21OLzaGSGc*+%(|DWHI9{wYEw9=IkV%50N-%?v^HWR>k@N$g<
zC?FrC^Ii%M23~@$(e;EN7M@{0GW;AX=%FA&M}pc|eMjL7lxki{aTgn#f3nH`eLb!j
zjzEp!UWAr>fS8!r3gYaMYjbvaQ9rFQ;!lTtx5z~8?r-U7y`hZ}GRzg)`XU43czj=t
z*#ej_&*r5EaSYpa(+!*e%6dozNWRj5a6KN4jmiZED#Flvm^tn5JlUH`p5Ch!`>FL}
zt|Dng5!1D@KniBhr&Y3MJ8eQc-1ZBbju)Icb##T&+K5(Bn5oUAxj1V<tFPfA2J=@o
zItvqkyOL>YkxuFw8VVPLQ){}bJD-5XszepnW1;XhW&AiCN;7A=uv~liuoQU?#yiO6
z%<}GQhe49k{^v9ieoZsnTwxZ0P>8qM#D~|@;)xE{gY90&Lc5@wd8)~Y{p^88C)GAE
zj&59}bvUE8sCI9as@@aU1fa-OWJm_0U+F3rZxk;ojO5qDLPw=%N(5&~L&x%wkekL6
zy8-gx>Cf^sWtL`8yRvS((`iQd;<H%9Y@>02rvoSl-|q8z%&(CT-nn~_yQifiMwrap
zofKU1P_1qx4aZxyn|OwnFzs=R+G<YjBgx8+9XLsjiUPyfn#zQMVeB#=OCeRBvVBvU
zIa<Hqrs18{gDgS66M3vq2J-jp<P2&%ex-NNQc}()Lh^KxujjX7dJ*U4B$2$EZU4q)
zD-a9bIWk;sL7RxP5~Y^?$$#*5j>^>I5Xf2m0Q!2cfc`A`9a*Gh(WBUY8Qdbxv{~Gq
z<rp7TTj7I{wQ^Qlykd;@(x=imS2j#Tazt#?ae;dVbkv|?8wj$GX02$=;P+?H4@Qup
zq~wN@iTROc@K+LJFqj;|2gZeq$bk)=YSfa|_&w`gfvERs9z$U{m6wkSRvUC9ldiR6
z4{i-S4^WWnKRi0z?>mr$nI*V&M!^XthLImqP&rh&LL`yI#KCPbUSq=`r=7Ir4RFqp
z@5|ZZ14dT{9Uvv$q1-%|B22)%o$Ap2d+@FQ_Hfn+7VyVIFElK<=@b}j@#f8bQ(JE<
z_nWJue?`>R`rr<b5V5g6Q;(MIs$aIFck}zFMe79W{X0Z4g~pB#-~W`OZcAuhz33rE
z=m@GfIj(-9qx3_72~1pvKB#*4qkxV@XCv-H=Ed=4VQhrBsvU4F{}Pgm=irMlr^05L
zmL;C_iQ5KTtFFODb35kJqG2JcZ{rN}HgFmqcc7PS3}POw;sjFyS|;Bk!>(WF=yMBL
z?%2C~r%KDDh}!=P)*&ns>in7CeO2aIay?@hI8zv77uheY)boO^M0-A9Z?3W`DWUwP
z5O3S<z#7AJS?dTmc{*`GTDPd{!oZcj2TTd`Pd!~3r$8$@P5HJU$4z<VT7kQh-P%=^
zCdNJJTb%0f#cVe-0Aucmh(i<KE(j=4&Mu{XZ--%vve4U9XBoHA-%jH8f9bow_DZ7e
z64W-bD202cP(8mm9;87k&5W%v%L})d>A^8uS2Dm?=E%J8=q0P=@!(#7S5MEsi1FlV
zo5kRaKbDkgwSj9JgjU?<&50`lp0<JY$I1yAv`_XNAibKO=0bqX$^p>x|KRq{P?U5e
zx0e4`PT|`X-%G4EUK|Y2+B}d-xApg;S^JN5`4{^?Y-m5uv4?G@5N>XcyI+ow^WsEW
zd%l06Px)G%9BRT#Kjct91+t;fvLkzSPuvC=we<_+ftNZ<pe3sxTT&n&o%{H+P>Tt{
zv9I$Rik7Vab7B;sS^~#nnhs?4k+H^*2E1|ajxj!61Y6SNsKtUQJMbe$3h?8kwAz?R
z4;O0Y<8Bow6(Yw^*-e@51w#6bGwRZ*k5-(e-Ti*x5;o1)9zF`le4;_%z{v(&!^0@!
z(MS9NXBGhwogRxc_kAOaql*(%RV?{$@3gA2KH|NH^)%(~at?e+;{&7C@QD{0Z<n>V
zS@Hk>{_G9jOH|>y4L|hbwj7BSacVSN$}l8lGji5H&dWfQ>SJ8dlO}E>WtTVSe*&fm
zUMyA)`8xzC*B0o$y``Qgve%@5<gf+43SeT4<lcSx(=WUVVj1AsJckVMMX{HI<agLX
zNKqj^ZDx_~ZaW8o&9)*SG#U!bsvgLGhu2B?$qkKty`^rZ#6(T;8R`fdF5<|foiqgh
z$cDk;%$8v7YzR=nz+QpzbG$l^34+dPVd+Pf=q~GDtjtEs%*7RxB`>F%0*PRp;aqO^
zl2a|OyqB(|+MG1<a;F%16%(X_6k#dn>neh?q`+<9#iIdm{R-iFh~8WPDDEvo-*`!J
z4^HVuqVPX-`91yLovzmOCw6HR@u2{YBl^H@Q#1N-zDDW9{#-@I1eR31Q={D*s+fE_
zrSgbot_a)N(*VTu@;BZ>d1Xv!e<x~pzpAd4_<t9`bRP;=6f`Jy@PC%&GNzDMkj2nv
zwg;M%SNnm-@B7(oT#DybEyxD>XGK=D+xJcustC$xsR-O|Z+(-YrF|!%7z}iJdnG}|
zzw6yO=+_JL9h1*+f<#^mSsHl}ja>=AZdxo28XJ+vhQ5$Xg*I%SZlkB>Gu<BI6KpP-
zP2Pl0tY>IQK%If8O0G>5j~?k;aGeMz57~~u)KUDe#Z6|IC-RB-__uyB&P};jTWz;u
zz1!2jD|&Bc!+rji<ivi;S~M<?Q+W-(^fmNdf6kiR-)LTBG_6v7$VhPDD{eYDE~QM%
zL>!0{YF%6$?$2%p-$UebKU#8Adn;j3Q1JP@t;`HlLGr<hlTXg0GvTQl`R}E-V3oDi
zx&>Sy#5WwbQ~35rDG$2}c(i>#=pw<RPu&F2a;xh~Mr9XP-9e0$Xi3$DdQyI9Q^>Hv
zNX)V3Q7b6m`qM1JcA|@NLatX)u7C*=?P+b&5Jgwnw~rn?jIaQ+ree_GsGb9DFMQ_S
zd9TjZj|YiA-}#A`C`D5JTLAy{QYrk{PI{N)L^{NCkAt8gchDV68J3`FH|-K3m9@n6
zY!8gm=SRF8ic}L)z$GacH;JLrUofAT`(cC`Zr9h&;*|-ZOi9pYKsRMwgUS%`i^(x5
z+kP{XstzB+b>rWHOKXPF2RCbl^{_Rb0g&R_c6q`d*P=Srf+b9IN&OI6@mf~lR@mAn
zH~I*3R{0#NQbVGqJTFJb-~J7YTro1Bl+A3|cn&H@xFpZgo-9{qUlYJP3ENbCqj2sl
zLc)vg?CRh<)rzKcwe&H;y#nr&8x?n`D5`z7gKcGQJ+Kgx9IDRjv^ku7jg<EB(@0JY
zHS~}=QlV=zSEjadQqz{N9JmXY0iQ4KENj6%Z3`4BYqzDqnK@H&1|pa{t_OYV@sual
z>%&ggZ+ifwkNx8GQN_vajWd5NM+T%<yP5Unb?D|uFaoqBdtBK8KPwr*)-vY?vrE?z
z^9Li&@pb$D?v8<Rw7O#*PTT52>+&Eg8xFeY$Gu%Dg0oyzC}i1*$wn8fC@<4A3Y1r2
z$QlFB5Kii-y?;a(MZVK$x7!>~(5BAms&FdL?=Rkc!iRJp@_WH?pu0cU70Xa_0EFyU
zhDhm|M_CUYQ~ta1#hTPmY7Gtre6@JteMi`D5UXFlDRPeZJ~Z{1X4MHieJfZtqS?F|
zu{Yn*BTZM4-B;jk3oD`Z5fXIBd4Xj97Wui};Mf;wQL6iM7xskoi?>M7XRQYW8ny7|
zED@{-6#lu~ushbUGG?&+dD;J&i~{j(a<w;Mc2eDz%Eoimy~#{?X1RFyT{(M%`*Lek
zgqep&Glx{_=O#lpb&ainJ+=gEqcMiq8M_(w`=}QRr?X%Dt-Sti(36+(ef8UH2}1+>
z!A40(v7`oH<8O6!(!BCllJ!|ZTenQMyl1htidov<P*snfJfMW^yjwP`QGCxe`1biY
zaL1|pkzf=X#5-@oC53KO1FtgM6$W$wGX%9@Vb~#=EVsDK+lyvRlj>WMq75ZgXCBhK
zGM{e?Bpk1P?hyakM{d1g`r*}YW-ZUS-zaa1i^+Hai2NkD^SP%)^FKY+wd^ouF~UG0
zB?IAVTi_uFnBPXksjYSSc2*B*@P8k;ot*2P#|aM&5cZC|1^m}h{we6<AmEGP=yscb
z^xs{x=*1EZg&}nv_75FVsnI$V&ucBJJ_l|(WYx-HZ{bPZgPR<R%p>PjoyJ+s<3URH
zmAdv1SMSByosdXXZoA?1<kff2<e#QxlFkbwhndWZZOa<w3IG%guQUD%Epe9}F8BO)
z<7MLGFB%DU?2RWrUOjo;;8fK~|E(ou6_bGh=ecg~#?uz3yS@E<OntVGHZ_!SZoff3
z1BrX#vz1sG%Daq$WUQQVwW4!alD+lJ0K*0G<ysIXCAMYUynR7DF1@+Jfq~v<By&ul
zEzSt_st>s7hzQfY^(lwHD8l*oOT`<`J1Bg$*Ecdvi?~JMsVNT_H<mUh70UaZm|5w`
ztP=q?2$MA_rE`%Mg=8L2Z4|x2fG(?AD4#UQd}V90qYZZzkLrwd*^hRgu2lR19d`B%
z>>0AkMrh_&szMEp)irx{T7yDGo_o<wrMQgMJ@?go-39-eeyFr?LyFUX%g%mQS+Qy*
ztJut5(kn0xZW??mMK)!myJOD(ozC5!jvv}lppktUY6?QtvoCoocd;&_zP_@LY`$kQ
z#4b3O?cpksLS|KVBI1JkrUM#-a-mwVdTPN01cPvI^&kq)FYLSuuUxi-IKB}Mys5JN
z^M%`P9wRiuJNQ*oS%1b{CLQEu<R=+uWa?GvUZi6*2!oBgWOe2j9{TiKmRfk71le{U
z89Sg1uYC9dt4t8QsobUA{W|>@A@J(q<A*cozJ1&VnR(WR)lCifPIpTulgdkQOsK7+
zSK)yi8tYlEVXGNCHQcDMUE0=&O~Srj`aKsxLH%adBnfw<>O<tJX%0BZJzLgGW)SMw
z-w7rLU)C~nQh&FwU-T}A;=8R6@PntssLgui(B)mV+t%e>%^!Ox{6O@$g_XS<z`nbz
zSqS;~)NLtFX!o(e%-!o@`_~2vPA8et;>qZPmcp+R4#XE>*|EHqSk1rydUzufN-VTl
z^h{)NVLTV#d@#^*O`wy=7q)6R!oPoxnNL$iqn?x`h+ejB$QzC=9q^CKxyd8<X1sDp
zespvHyG&fIQPqA7=^?Au^0{g6dO+Fw?FYsi?LXhmpC|`ZeF+Pxe_rN4u|J!}!KRjV
zBfeN{*ZxAy<Lag|`Q@G5@Q;Y$B|Uw(eccnHzWdbx*?R4}<XkJ(LtAK^Vji)6`aZZr
z;3jyqLq%rLjd*uc4q3}A>E@3*Qu_DP*YuU~ZlO-6i*94eQZB+=^rOkNN+Rcdop^1X
z>9%On*{Li2{10^(`vJpzlJ2tCU+C|+@>ynagcWbrv!^8&A`hYxnuvxPv1l_Sojo%C
z&Es23r1#pd+`h>W$4z@XAKPBR(O2UUZcCzOJu*_%p(ipyZ@<Cq2H!?s{;dfhE@9nN
zzdu=DN6yPQ1kMWlm^8Be?Fw(hd^Y}(-ZU(wSCqsF3QJI3e1vZwznw$wWVp*_EtV`>
z%wl+%mIF|bt3&~*6n1yUOI4HMp5s2gn8YZFo~wyB&APQ<F;eo6`KDu@Hu`2n>+kY%
zjN52wjIR$%YiKMs26+?k)MRnU_;-U^=jMg)ELHKeG~8%vCqCyzU-`=n7mXr^)UnRI
z-<m<&gR8apnW`?=5EVMHDc&~0__$cj`kGucJb$A?4WEnnb`&?>SzWEf50SHyANOK<
zX%{JvZ)k{M`raWVn3xZ@4sYeTs7;(@N;$cJT#k{xh4DvrRUsK~55AQLSP)}-YO=Iu
zgxGR`$-3$Ce#&O_(0oWgulw9vak?r+CqduK8ucG&kBfv}#9R;aBvt%!IV$xnPMjU7
z?VRzq!s!<HTB|=QQV(Hn`9}mIx3+P97rnJgpCm|y*<id}VgYlhu))0AwuY<^zdh;h
z-X`WlwnE|*btLiTpmAwNFqc7c#XGWw*&p;skGi_yXJ3j=N_NoX`0ze7YPHBuuXlcm
zt$OFVlc!r6es}h?_|RISj6?zV6V+15A~B!%#r@{N_hLU><k~YjO-g)zq#MMB!OqzB
zM9&DXlLhD@{r0aS|5)ks5MnjIO`rO6MXnWMxCIzY;dM=Va*^e@7ynA_vM}QB_&ULf
zjUe^<=A}26t;vy4q1?)1Q!b8($it|B=#DB#y{}S-?BxTQy~|m5Ri0hB7-mE%`uy|Z
zL95%#3L(+^{X#d?yTo8vWNNEe$l9eJH*4`K`LTCmFJ5jt`O=-!K)*Dn614UuC>qer
zE^GGcOdI&Z(m6Btd(Deqep`Sy^~+UuubY0bnn+)KXtw+&=ejXw?(Ks|7f(#CiruQM
zI}P-{ltVsl&0n}|#5wz2WWyd9($VVEtnuTMu}4fCYQ$;TJ9h}`Fuf%y@$hJxKM3P#
zR9q$Y;=6w+9PC~5jGsH57#Mcv5=}2YX7x9G)?MtHd{OZWMJU)OuB1tL@XbVWiTuv|
zHMS2?7&cOJ{FAP=Z{*9!-O7nn9O8M6t87w%F40nytEFh^^Vc~@9tLLyim`$U#4^<|
zxqCmH&Z$H{Hg-%LDSc*Rm4hwx-PC-Oc4^t(>IP$M+8+9o{wp&d|Iko%(5Or1Re6gK
zUsF?4?(N&R1@~ASH$>wq8<<}`XSly~+a;E`*8Ci3v75oGa4Eld1a+|*<T)Os7tcD5
z8B44Ou@p?MS2r26Tgk0Oi>q)<8cf_9tqs>rcc4i@DdkuQt5EypECt|ZyVMBWATDrb
zBVPAi^l>#DsI=`O7Rmen&R8!IU5~zO`MK}jx?L$LGyOZZik}=lRHWd-=o^kQi&7h`
zSSLOo0?b!XeqkI;C(Ii2kYs64!LE$NgZ4sgAG-G|0a4qhs&J^wfknPc2o&r|UnF6z
zZPc{M{jBTIgr(&5%Ho^qmj7OM8n9>go}D>2f~Tq7`3<x^Yr9wby!tok2_bblx}Khb
z&8z)R|1M4TE6Vd#G1YlV&%>OxZvU;Fe3=#=)Gsz$DUU(^7jOn~+-LVWH^N?jt3dUY
zw7y;EHtl<H-RJxzY+})Y)S&Mdz}?QESQj0i_JxVNo7dJKext1ZN9NcZ<k4Q%Ba71^
z4nyOw%(q^dZ9O*4&%em$ujxFk9@lmD9GG<<GbfpW;nX$Y$5|sf2N)LhzCaU@dGe~h
zDkgqFzJ{GYg=kz*4vp|I!D??}id{;@=gDqgxYqZYjbn*YKJ&kzMqZk0#bWeZ>@z-D
zr!^!+L3!3QuYf2G-zmn^fW#ye^Q9#QRcQbH=6YQItpaNQrTVtUPEbAbg6(g^jJ0Bg
z0NYy`Lj$nbIFrERPjK{grxcFA?aW&4M@~KeAI82iF3N8U_phL!2nZ-hBZ#1s^w26P
zDIG&d3^~Lwbc%F~Gz^1uHw+EZJxIq;Lk%_3U6*t2y&sR~{N}^^VSaDyz1P~$TI*SB
zZDp>Wftk{`iDNaQw5iDFTT2kzdZh5JBerb?<UA6b2kqRlqPsk3Ri34Lf+e1idDU;N
zi%Cl`1u>SnfDQV`gw?QRX_&n+fhAq}Z4UAaU8}Vo#`KenRDoIJZ?s)0qS((ed9ic{
z^FL||Osy&2{FqeLx`L3FrzOkG{-$ty>J=XzL1e#S94v0ts`k95#OuNYG%(Ay(j(rV
zLMz}<y-_bT&2^!wvslwp+-kj2<CHhUgVdVw(Y^c0+HMN~?IAT@sZ$)=8VNj@M>$ya
zvMB}UavXssHTSLljf<LrjsS(ly>&4!2Hp%gZne4Tq(YATDfVY%O=pZTJ|PssPCw6{
z<DEEV7~MKZ28rI8NivkWKIa*+vs?VL*%seJpJi&$L-v1~`K9AM6Bu{N(Dh6VfSr6u
zi9FDmbm`+I>drrTBb$>yY-SUovQ{E4su@BoNX4Ix$eO1zI_6N(td8RLd$b-lWU7%_
ze=Cg|f9}oUD9uvrT~j2=Ay`bVBSmBVMJianT7SB`4JRt$MOEb651Jrr^Gvt6n0t^L
z$)+3q7ZriL9sN*)0=2wP>_EwgyM6%_CHX&WSczC}Hr~GAU#X^vv@spwnR9F(dltzE
zWLoR0p!mmfH6~0dZT7uW@%Z!NkFu8~rt0;8#5TWhPATt*pFcF5O{7bryl5$`-gC2;
zv(8GjmglCxdBp1Ig!HX>&2$(xp+j$@JK;6F3Fk2<+=74fechEY;<DCbVtJX-OW^qh
zvjas+DyoQ6<5Mho)D#>Jzu+zPq%P&w$E{mL4OO70OFV`(;)_Ou3r}_53R10bRjGDc
z%y|*5In-Vkg{de<A#CVtl!hI54#6tv{9lSuy{qu+@w@<z9gTRE;~M=Svg7)S0{CR2
zae-q&WVQb5yenm^<YpHSZ1<e9{VQjXZNk1S(8W;i#u8I>l!Fo&Yte2tWQr*&wCVYA
zGB%lWj=rKw65kgyt&6d@x~(oQz^+(&vn<(HoSkO8zzAg>GEL>$`tlvjCe*|q1VPj8
zdR+HMPx)-Ropv9OA-W}JH$AG0c5^vzU8N4i_((-Z&+Tn%w)5f#)=KOe1sogR4A07M
zv+fQ<%f_Z1S4#S+*-%1FjrHBVZ`WFQZRT$hT>RTwUZCVI1!qab#Js(Y9i~e0rE5vV
zT<lOyT{U+nWpM4%eXp{)(7Rbci#+yy;5@1=lKnQkqsK1Q9MpeMScLhbm{1d&cF>_4
z`RC1;O;69J>G0}^fTJ*&nqv6v5BZkZ3Ike;?KoBIMnz=^dr_?yjri=Mi?6RalgClC
zQ44++4IW$aTmZ(8OcEj;Lac3eEKTcc*5iRTrGSfPe=owU&<`Nk<11RGQw=z>{FN?c
zIS4kdAV`8ty2iM(%t-b?8K5++;}K<8CI1KGaCTEJng`WGjKUKQr*A<MOyP>$eP{~j
zaf+n92U#O|mSwp#&LfFa-=(ZAeaEF<0;@|zHC~o5MSH_UYwX$zPh~HEZ|IIlU<F;4
zeSe(2WtU|-$i-Sx1y*p`e)INvNi*~eu(bQrP?*F>A8(Z*&8RC`!oga=?CI98&jvJP
zB%YrjIt&W*1<~G-E;ehH@^ACw5)GL&Lj<H{#%-jCQfqMiWB6>Re?(1eDSi`V)9aAI
zKDW*Hc?sNB_6BdT+Ev&ylJ~s+)@$9mtxTs#po$k2v%Ew=2by?EtT>XP-cc3hZR<t8
zVwqdtnDj!EO5yOaY$H(~)<$=C4aNzb|G_cA8C+5d8|L85J^=$?C=blOFvSd*MzL!j
z_S(?VrVh`Pi(Xi+@v9WNW5W|leC*~Ba8ZxV<5eo|U6kr<txKh;YEQGwDst)P)#I#H
z4<|cq_;q9GKU>iw5*@W;^*Lzy)zWPT+l;*QDQi={9pWjOJfU1~MfK~24}}&-`HB+-
zzwJD>-S{+7AglPxh)WF#{|UP*OAcPW7<!4uh+=W#elV|H$(T>R@U?%>;n!WR?1CpC
zkzmK+dpl6tt;HfKJTZq+f$H>j&e?nez^YqL+(F?N_^jN(v<CaAa6fiEWD;|PeW6Be
zrHiWZ3NM?)sLfV*2vNErB;p1x6b+(&?*=&@fKELwlGANucT@EZSPV<KK*AvJ&2AYa
zVk_3MxxWPm()uQ6DqbjKB8Aarm%D#_!j3u2dD*AR<2Xz&ckJ5CdFG%i!FFLbu*^9k
zPo27;lq0V}Z}Qp9$c^&5R|GGlh`MZEpRN6&M#-j5dFu4o-#LC*UedJcYo{m%Q&g4v
zmYTJ8nQd+CC+wZ2t*KdC2%D|lba2ALRDmFurR^FGU8Rz!A$9v*L2Bue2G#C`z1;S#
zx4?AD>MX9cf{0uzqdK}bDxgot9=dSDTbG6ccq#uES=7o5p-(}vt-@hp)*2t^CXYu?
z>r(TbD(z0Jwn4u~v@YeDX5Vc8=IT0~<9xOD>(qlcc}@E1afz{-bXGIdqgKVj)~<Z&
zH6?ElwuY$D^8t&D1r^X~#n$xrtl69}SFvJBPs>D=yV=;VwOhfHrZ^{M%#pfozb1Hc
zs@n14$^9~Q`l6*~iUwW#%rV!`w9&cF(tUCBg1ro7BjN?637>>qAByzHf?w%2M~H6m
z3Fl~OrsUJYL<^dEnByfA@e`Gb1OsmT=8c2BWZP%d17BD~y=KASj>u4=S1b5FMg6vv
z-TM9}cBIPtR|Q!CJBA8<m#s5mU9l$h%I;5HaHeR%G<TtngFUOA$NsrY8#L-G(-J=5
z{(rf=Gr@(=mAiBbsNrzsMqK}F!sa|TICU)Uc0Z`cxXx2N3o<M6#h2A$=^1L;RL^3!
z2Y3caa_BX0jTYbbSCX8R!skQ7OR1EBr7QpaW&i%3Uv;uyPAgO8FO!5FoVEoulUQXf
zv*Wg=J%F_?r~&kM$}zBW&z01Z_W_ioTSXw}GrnLimjkiqh^u)L2V9zavcLY@#ZUYG
z{mn`Ht^5kQ=cy{(?p7<4!V1n4+LD1oQxBM{fHxngTA3(?=w+J}VYE11*9K(R%Ings
zZ>>WAZMOd3ggSg>EHCiYrb!mIC8dxb%STNVFz__34<~07dlcIKDnD=dYWBKcBHwkn
zQwD$Jx1HybS8-YEvEt67dWWoruNMEyApSK@w?Kx^>tBt}uv_^(ZCN~ITSC?id7)5q
z`e$KN;LxqMP2QIRzX#|XN5eeQUaP>LnGYq+coUSO|JRl9iTJ1GhyQCOJF2!)G0Ee~
zA*Q01_=c~2pYxRAJoP*++++_wj`%RaWCxYud~cSvM;=Zil@9&=C5<gO*!!psn*J9F
z``4V|@!a2Ht1yL$X5<-GyPH7|-WEFppoVl*pIy`od6WjTe^Z2=L}szzX1e^$5DP4`
zj6Q?8OwGxNboN!*UY>MDP|*Gl>;CU4TVwbxgDReHVcMA%o5uZx)Dw!AmU9O_yy!TB
zx4afB8LKYfdE>P<!^~ZnH+A6Y>49Uh{{&0-*nbP1TrPX?;f$<vQ@yLYX2boz3)=tv
zCL$W#QOi`xToqf&xwz24OXhg#&+b*1VN2AaT>WFuga-0pqruT)Zz@w5J(d*Hc!Z96
z6KtE8r~Rl>YeTDK&y@d*ivB;L``2X%A3R;p#xwaTfLMdy={3gPFL6b-qoy{4K~?x8
zM-&}$^FtcZ(=Pwk;Ae%i1*|*v|Ldb^{=wJSkEUVw<NMvD7ZB|{4o<M0Rg<^P1+m>O
zr$lmyr@&Lf^SVYlCT+<)JyJxaXDO9RAj2a6tHS(bEO#!K*ejG53XZfJ3yQRr#2JAf
z*MsR;GyVE!0bVz-*R7HLgaH|oA#`u^6gf`<$9Ub3>fjm1u#v-Ll@>7?k3DwVG4=m-
zO-`tB%}<mcSFJu+V}v^5s9F#!(+&ZJt~e*%`HJivr`4b*zhw$z6DG8bLLg!#6Aj*}
zk1L_n@EkO31UiR4Y`2g&UfA=b_sJ6RdiGi6f0Mp1t&+cF*5jP%d>?bs2=IDb=rKZ;
zhPG0lIWWADPJ7f8{GM7`*=AsP`#ce(5)dllQWJ!u1bfG=xMgG8WL$YUN7+M<CQAQf
z3?=;cI1OX)j}&1ql=a;#-Nw2sI;S$Th~f{I^<1dErA=1+UH-BePP^iR`(QO$%`==M
zA_)%D;5Rb|gtExVHgosPAU#QFahD7o9`R9ybY8%MLE}Zoc%M2|P0gIyOpC4dsPTMz
ziF6%7kgyk>`$gbh4*fid;cR&hfh~pezh1fYOZv+<x1G}`rs0D)gfXGtIxfq^C7N#9
z$L_m_lt}q{46mSzrWUa;Xe}w`dj2yk{PGF-Ck6oGGhlRPfbmnsWC2^c9v=n`4laB&
zoY5ktBl!RbpT3D?W!T2O_?=0XHLIP-^{81=(%;y@lJ)51D8D4><df7e-DViONYKmu
zy`$b6+pKe&t?NECw0mDR2u|0`uJn%oxlX$GSwXx3$49v1;J0_~YWA^d`gf0Tb<=>E
zmhq+Vds_qkgt!|JQWv}IP~nTL)wrXG`c4OE7Uts#DeR=bv^C9`BCeP1+`Hek9_yD+
zP7;eErQCDrxOaa#^j8E2n}Wg>)-P6D7Jf?rULba-pKFz%XyAJFuh~6)lwcirbobYo
zQp8NH-G<y*rqV>^miOwH=VIFlmv{E*@#9P}{C`u@tl(b_Sdn&7-P$U%G;V;2!&MV|
z)C=bW2Ak$bS_`ukJbM!f@gug&4hQeGf~(vgwtqgFKN{^Vy)npe-Ku8Ix36={C7oab
zHlkPkE<8j;UTL&iH%I`yd=8tnrP&K3Hhr%Kh&y73OG+83%#~9~hdH;fn@zMM30RR1
z`<pdwwi&VS0m?s0QpJy3rieV{aQr1JzqGtzkMBAF((6&dfz9Ge8V*}@egjBx?xwXL
zyp7t$KP!rS8N4oTFK5}>Gd_HMjQZZZ-4E^fBuhUz;@1M%9zZuV(h&;IUAsB;Yk!5d
zQlP6|U+>@ZU)B`@sU4QloLshD{PQ=A?B=*X*_TB`y$v4OfIIc+YX6K`%V~J>EEuWy
zs5+UIKcP&S)~t{e2`6S3<(3*&^7Zvi5`SJ6hc)>w7uz>7;;%}z6Kj+j{XEhlc$kBj
zGh9_LnBMWTj??PZuM>HY)5vCzw#yOB^I!ymt{fx>0yw;Ha?S>B5?xOcLnA4DFUI@0
zN54J{dL?fj3MfBu>Fk<S0(L~6N>JwZFaNRXp?#@JT!MP?ZA>8a>n}gaI9!5GDVguP
z+t^10#HrLq+G(z8ZdWs!=xvE+L1C}%Cv}{JT%t!y?-Uqc{H&qAJGaU*u%+|mqXm(!
z50X+0BjBgd*O<Cj9|Pss@O!w2NlR7JR^g)@4O>k(?>eehlqk~a7Q83iA$B}cu$X~W
zaie;3(<EV#8elts$<!&D$QL<ig~rh~dP+W+MtgHUi)svep+U{0w0WADwLUOo+r<y8
z@Q|ElY&Fl94^~(mZL7+Y-B((!GbzSejv*2|4m7u4rR;uxU4b#0eW6vhIM;dj?XSpR
zUiS1eOO#z(ZdB^H|Ca%Pll4GBsI2)})<%xsd}YbRS1`qR;tpO<Ss{z^PRvEd^q1Wi
zQb*Ya^wC7=Wrn`m&8e-I8G-@4pqGdUL+Abx^ngT5)u@&+1<X+~Ik^`@lht~)b!-dj
zt~}r`6BcSZ<u40UD$l<}`5x07*Rsp+D(SUYd!JG0&IKvQv4%)oZKMKe9)ZVBjBe14
z#vKV*Zef$z@lSq%YljaO2bdp&`d1n6e=WgFHtHEz-fz8LwF}+TMW~Puarm&qZr44H
z$cSR4y*6f1<iTaPJK}6Sb_dPNeCqdYN(rLmwgav@qMP2p-U4HCjszFabe$^Oc1eOq
z+NyGUHhvbFd2-^j3<v0~eler@X%C=YJe@tvBKz^u(dc?e&@Mt*gZ7}w-#lRPd}+HQ
z<`BcmChp56W}|v`j-${%##LN$c&hc#VvOD=8?rls@uc@fs2$F^wjLn{GPCdlQRG|%
zw`cRf*{u`eCqb>3lS*i8Vmj@vCykPVRx#0H<JwCh5xeq|Qf=YCBDJaCSLeyNnKDNI
z()cQ6dgo{HGsF9>2oo!k5y>@Wp`KUHN59fv#XJ*nt*kSWlK+LG&L&Fdr>+ec_F21t
zkge^lLW{Pd;3bu5vxBeN1E2yr8$PR>@wUc{2YyvTKTEj6X=l1Qm>Dfko|D8w7Cy=D
z7OhudOZ6MMO}a9nln`>_DNx_Z1Z4jz0WUGg$n#VajUU>rmhxJajNW&EnOhM{K7+J1
z8m%B$4?KsOV+S}j>>7t*E5}3ByTrsCo-LK7&tC~g^yb`2k1N0mZR{r}wDJD*aOszO
zG(b6&PyOO#h*&_#ed(^}630~uKu!EGw00sbZm*t&A_NC&1n@v^KlPTo?cew=#M=RJ
ze5hL6Nps$QYKPStTKg;b`*DBxGVl}DGbU2dcXF}{hL5YpCZxvM1a?~+SK+X-0K3D3
zJD&8BD>j;4BE*irW(1wBev<A9KXeDj5++v+R^Z>vs5A>+t>rC>nB%p%j@_n_ie)-2
z#t!hK;tZWGAWCccXnjcL2Dvsdr)~_HRaji_<A|4s-_)Z^>`?+iLr$QSlGeLKWoo_s
z(V)H7+m)bSLM#a(JZF2<T#s4$Dc@g)O>6z3)rO5#@3od)sO1=l@dn!Ljra3wYRvXO
zt@$AJ^cj=KD{E6xvJvgEu20ibQz783+*XPKL=<mlTM=#)2sx*@Ef^S_ZtX(8!r-Zj
zOr1pqkX;OwiJR&SEsuNT4MH;O4Ev;{M$KkipR+(49xSs_QqVX4=Plxr?0!LPaEr>W
zmAHzq0a|fDvpXEhD!T`DKIZiR@Z!Yy`P~!p&zNDFK#j3kj4D3yG({owG4XkKLU+Kn
zdna@a5E66;!lY=m2_K0p?<`xqT?-boO(Ug%c_?b=Rv9bkw0L;e%~8GhK!^8`2-lFk
zvNCBQ(mx>CA(Q6mZQGR5^a*=rp#e<J*DNNnrTjA|wNsbveLS-fkI&={r`^dCMN{^V
z7O&6!(WazbruOUzC#Lr0;nur`YY$c9x&7E)K!=}V1zuss%D*fUc30z&vwuzZ*!lK$
z$R*C|$R_isNbsa6-Dl!ra^(Gkd$V858+DPNDG8T-qFzztj!}vMWFP|EiGW2syyLG+
z^gXQJ{bH+yM=>_spSlH-x8IN$1E7~6HU%WZH}jAze;enPxhkn0NcSMIodJ!H%OaWP
z;1oz%q|{wOq<D8h9R}2(C<FlxLrSQ*0NCkXG);E;9k{xkbU7DTT2W%|ueA5wH@wE_
z5(R-_HkyIvimB|VQUj`WBwSUx=~K|Q^YMecz_(Gmue-Q8H>xa2fw7Ne8dstAiC;R%
zbLUBi$^Hm}_eEzWBwk?sYD@+#raTRyh5!RZBuv0v?P|wlySIy(1ac;lj9Hpu0_Gjy
zJb`%)-V?(VqI??u<|F=Ekp21`Fd1F=y>#-WQLcJ-kIyxq(-Z6uEJ{t5_+#v9V(*2G
zPygkQ^#OtRQfCnq!Nc>>Vq^=Db0MYgdqr5@C<z|X^40p+E#x7O0VN3Ay9z({;YPGZ
z-N*h1h72ep$U<L14?F%LF<et(1OA}>*b&}8M(PKtg6&o~bt%jo`semC3If@iFCtbz
zkdyI^ka8uJ_y_$)i5$O+O@kyyT4RJ%9?JH9T4opa98pEDni7SFcW&6=k`ep_nt9e}
zlBVr_mEwh%_aJB+!O$#t>-I=q2arpQt=v0-8F`J?%N`-q3bH*xidh-!=ZoF!NiMs1
z10IZX`(@pW&<$u^iOK6=A`SKHU|=R+u;|nB?dccLvTr$i(v~~X2Zh7foGh0CP3S*8
z?Xk(WS0UBE3fwLrvQnvlSGm2Jyx8}EpOf|rbZ~cLrMhc$Uc2dVi;k;dws(<QcR{;=
zKQ*Wr7S2a?G#WHClFJ#ECy<<o!9M}7xwXd}W@SB~3qO8(=-DxeHSXFXL@T$!Jxlvd
z02g3t*Bn@KN3s%tC#iPO9yx?Qcbv)^?hHva^66PaZ*;6rHg=zX+-<$Ma!8!JM6n@(
z>pl%bOv*{7An_8t+Il;bxmE~Hl6Wzynt$$e70S~(8Fu(fe)Vw9CK2xZ5rF9oL~lHR
z;v6Q{Kki1k9tC3z_2}uIdJaDFZ26(rA>{cxq5%CYIo&>%*HKMYz)|W<2M<XH3G>G0
zodz6NVrdQ?&>48aQ$|jDJN~m1aqhPTr_Ea7+|7~~?KLY`nUZVgnlb>&R*!iD2r~Z(
zP{`*@iNdCW8AMQ5Ht(R)#Putng`TNyY01A%CFH8{{Dy_;wdOkU6d5oP-pq=O^fy-w
zhyJ+q8JnctpK&P?ow}Q2avw=}D%1?g%8?`};&n7sM$~#*yBKn)eXic6>eOb{6R6!M
z=H2-c+WvtzuY1^lQ;6u2&}6$H-Ui1d6ql{hVjQ0LH#^_d|DSFCt}Qc*%LZ6tKN~&4
z_<p}Fp{)1wtqEo0d?TNk!Q~jvXGMPEkfWnsu7n=4$bTz?I2&G2`k14!E-nZ3sNA6h
z54Ex(q)hG^tiTRv(I%HQa=m%O9mZ1AQFh)IHYL5>>He8qz4go=m6_Gm@3T5YF|fSH
zIAGmRHsq(?re=4^3Dc)(LYjbtqn+&9(m4qFnsko5OOCF0?^^b0k32@uYaRezJFGXY
zYqAK2RxSBfEUB$;wAJ4g!Zq+LPi3<{fZb>R3FXj*tvhE(eVJnl9>1k;o3=X)<|GP}
zxQ#sR#HOkN^~_bp40J41#hH<Q?f3k~UNU1TrgqBftCquUq9w5k5)HhZRC&Uh{kMDA
z)V0u#Sd*iTg`ULVOE_zHBBwsRkgxJt<f%i;C@kTjbHw1TBvA^b5vCvJsZqL*$xg@8
z1hen?ZV*xgu_46@{uF231M>)zW8KlX5vpSCog*?*T?IWfbQDM7`K4Q^7~?#nOffZ&
zVNFaROXs~rJC2aKs-nrROU4N+ZSlhZrViOMcjr!~_>~r>k9a+=xoGGH$7Goudq-oI
z`EII=yz@XctB|ODwf46CYaspO)K{br=3OwBLu7eWTnfWY<yv86%jf2Rj{}w#x9`u^
zBTqxK@?N%iJGTJj2JpeHt^<<$l;@>I$VehFwgK;U9XBGbA0;%mcV>Dj%{#3GQRM8M
z+C4`<X`5Xj)Hj8hK`m6GxftD+{0K)x0KcMQU4AD=NKRTF(p+`b>Qn7%+7jbEBl#F>
zR`i~+5|AD`#cTn8+TuCjkG|x2y<QyqpHVsL*TRP{J%8HXIh)u|*I2w$O#Z!KRH<&3
z7g9XBeotso)<u2n_|>Qrh^3YvwuKq>d-tppN>*f|@PeIOnYmCFxs}XQ4~fE6yEEm(
zPgFf=ny@y(!vr>hLGcGRds(X#I|)ef{rtM5o{t9dn1OS_r@^IU4|w!tiNl9lUn}F+
z<Y9~B@xKD;^j`yUZHW>?4%gXrzXv3V!U#Lt&wOt&^g)Tt-xNY0j1Qq)IvGj+q4Lf;
zW^X48G_^tvl&Avi8+VEmq_v%VZ<a=1VA~lt8$l41Ft5Y+ryNH{t$Wu9f+GYLy_#68
zGrzpihGf+JZBdaJsHt$(eiLio+SJu2`K+hbu0cYSMfH`GSd&2}-S6N!70x)LHUEi3
zU#~oVirO!XdI2tVF6<3?oz8NYC#p{p0GPGAn~U#mPG$erTIZs!AK){<Y%w`3)ArK&
ze1@glwx`8?<$6+o&C6>qqOIL=Ybir=uvI0b<DK)t&X3-p-MOpxIww~tFS?EvrdkT=
zn!B&e_pg~qIri)A7vmy*&=*ZAP9Z*3PU!%BU&a`9fL|Sa@DaW&!_+p)u{lwWZ?3H6
zFzkh~)8%qrN~U~qi9Q^&tyjMoP0^FV%sA+CDrt{zbaqo>TrWy68r&F<jk8FtHA*-y
zjCtEjw(o@M7-%OO&<^fc6>pST4?t>WCiz_AXa~2RcSoW#aw4Ja8%5#-4m?j^#*Ln&
zT|)*$*H#9Xx_Zb|jRpnF*L_q%H{$ph!HI*O>-R{$$%gpvI)=u5N-rp&<*<*l#ckbd
z2AUo*C<Fg+DDbX{H?9t@tT^WwRTDqlv`Eni99VTfkB}OgV;vC_GBLdA7b%M%G1pUR
z_4U{#+dp?Hm!fFC-6Gf;ezHC+5SrQE)}YEbd^5n+!Y`xmV}!WrfZ(I2FR;W-rBS@6
zx2K&}F0BB!^ZhR4<tq?;8-ptl%INKQUI@-kKN<L2=+G~Q>6ZuXcGOd4x?W*-uy=8B
zvyy-^olY#s;1f2V21N6|%BE*CZbt2#Pm_iC%8-Ad<89(+<VY-Mh;$p|1i?is4o1gu
z${(;DR#9*~0b|WWrN_q1ia0h}SQ4WxFVm)s-1Nd@gKcYWmIoqvBc`da(c&&j`ikQ|
zzo&?s_cQlY78|8?AuoS)$&r7&-8J&Cxmt5?R^wq`xIy0DM3Y;D#2Y_Mb8^~tF_V(K
z+9ux3x3>uL-ypsjOEt@=$%w29mY${l<KEOpvqnSolpHj6KFGCap@54AG}BGtg?hM9
zzoB9F5Fe<-7?JkOM7*Fiah?(0mUR^+t+i7VClPPGtzC?eX2ic)C9@wZ^te!`yq)4j
zp(h@^>X`JZmxKE@-jvo&N3dU;_$A>pDU=EA#%=tTOLNgO7hSDanRa2%p<lRh=zK12
z`s!^2$2mKRyUXfJykN`}xO^u`V(YgMhcX?4;i_kMfz+Y$!(gn$^@MA{rut)98s&S)
z;tPADs}D=y_}jD1W@G<3i}_K<(BBoPv;y%{rqc}$Vh+No`ST<|0K(;3|BJHMPKrVb
zOq6HB)Hhk(OYS{DN1P!Fzxnh9HBpVk#Z~P#GA-UlAV}q|DGm9W(Pg2!Nx*0JB*D!4
zyYam>G}rr>R<fDFyTe1U#ML2B_a80@c^oBE+~<O)-eP0z6R`uiYn>@>bJ|~e63pjN
zQcXH0C{?&?ueWNZGn|y=B}${^E|#@LBo;4w>s;x~j8N<0;sNxls37&0*>lSJn`MuX
zV~ibD2gKqBxeXnm(6j}P7^zzZ)0^$GTg4rs$+UHV?ZRop;T?rNd$ykz9bjV*R1_dZ
zag};)2JPjw9z+S#ZE(b4aiKur!oS-LOtQb&ja~V=sr2tEkp5HzIUMA}x2$LS*W3^-
zn|Z@9rFX31pS!rKN_1j|uMb)@trs1~(T9oYK1MDvKI|T@r>4W^6z0(I28>)bu?6%!
z7|5^BSm=+ltlsdaRCdL9u^%(F4WLn5W~ZaMm-R^zL17jHVD?G2!#u+nm%m6M1=2<`
z+oCGeiRA4g%fm+a)qz>$8!P6azbZOSARZwXD+dD?aSj^gSd?YV!s>KKaw42iGWmfc
zbLj)%w;ZWwBisc2wcYY-r<%#zoilF}+{C#P*~Q?LOh^J`Q2)pSQYG2uoUQ?_$u!Zu
zDf^hJ&z+QCLhn4pv0%AGG{C*Q>s*QIjj3%c%VUf)CY|c05s(x@#hA$&srbnmoNRG#
z7*_$%VAr(U*AN@rkiKwPCzZsv=qWYtz&YiXsZ=>X3w4xH73{>2AzNuGlxDS8!Su-c
zdQ}hwH^o2kt?_cXa|!OE6cAAizJ=2ktNEo8OjKSWhnTuk$aGVaXV*awVjk^Wxr{6P
z2@zfuM=dz~lEzFPkP(Lc6Sq?EH=u2K|9;Arx=2!*E6ylKr2wqgEfZy7J4GudSt+Ll
zp6b#iEde@a@Vx^F<CUX`A~b#Mg}o-h(?7(Ff^4$^Roy@As=7@ee<JQuRych_r6Lq#
zO1jw5iuLM5_9~%W4vkMgiCLts(x?{L)q*l<?P6?a1^Qg%!9p_W)0xQ8uk6da?@--V
z^K-gG7#CLG>d}qfXBHo#98v1%4fGBJ!Yiwlz0_{8UcoSuW4Bxr;ii)P+j0cAiuQT<
z7)-L`>g7j&j-LoQBiGj`++5|iSOTC+q-*<!bJxYDPvomcPq+vcz7TVD33F`M^>^o^
zj0$Kr!?#n?Vx`)P^=fCrIyvS><)U8uJ$?YAkn&k&l-BUBl)AsHc$O7yM3(gtspq62
z8%#_TaCF7KGtWGHfBQ?IEJ^`H&=5CmS5DodRwzNWHtzSs&R8dhkfuV2QNCYT{ySQ{
z9UV)W?hjJlTj9ark=!#6wZK}RE0za9Z{xCnY0)r{RA)@i81pi3*EbKwk|ar~{f%q;
zW6|Beu%+epPp5gN=@Ep40VO}Qn9baCh!~IJ^Q@kyHsh}lj+UT4n_MCa)N_B>GLmnc
z3aSMUQ-u<p?Pya0QzB0Q(=Czy2#$)n9xaL5K7sd7qmp^8RoSFoN4)wNW{Xxc8A%QN
z6$eOyYkL@W_az7RXw;&3kR(WI;~l8k_G@JQJx>M68k~NPXwgd3PcZFoRV=bJQSdD0
z%KUzJ8QwJp=RH$y{n|6NbDz*}1FA^(xY(lh*tC>ySRs`3J|pATmWt=vTkkjMEyYo?
zD9XEz7HKiLQUX`!iM})dUTZ`|?4b#w@2WBmZj|j@&s{nH9HBPLX_22DAjy^E?ueL>
zdxRk%xTH@g=~QDWuNfi_n@gAS6ulbne-U}0wk-PD9T{p;)EhVbmQxn2nMgYPX`O^<
z)?3<($oUXGp(Fu`%09)rfW^2Jzy@N~{#>zVmO6Ci+#OPH)k!N1rbtxV`kHssf2bTr
zG)F&wUOsr*E!ydKbBLGNi;=PJ8u+WX6#DhzGno1WoTl3xiJj7raURv!wg`zT4SqS+
zYQd@q|4oJn@38bxsdy;ut!Z)%cskUf5Xy^t^wV-J=8p=A7-Df1;-sjOX`_}L6So_K
zMFBcRnW_1qh$~?uVtFo+MaGHYc-`{#e|F;l120uPU%kz&_>#|ah*_;=8NG<fX8%?7
zuGwqi;zi5|Si!+M@T~_8L?`08A5aChU&1jhy);+03Hj2@&^l*<swc-Yuwx(Xv?K}?
zU%RN0MmTvGW4q66ZWDY!9};{bnR`oqM<sP+o`-J>bk&bLMuYy;XSQh_ord^xQ+$$f
z0?Kcm*p1g#r@hr=Upc!@zrD!%9$u}W<WLf_b-oha$_(?^L0dVqLl@MeMzI|luQ5qK
z)81mx`H|e8fLt^A{IT%XoS<|4-jVDDt)5*5RS;#K9myjUm|4)*!68q|jD^G|!Qpl5
z+9+>F`gOlZhe(;I;2vy!rt69=?a$;_%-;f_GSBZBAS#PQs#d|Y3ca7pcnV{4wSq;D
zo9+myt<}&6hq*3<Y)=BZiLY0z$wx%Bz8l5vPFFA;^>ZZ&q5d%;ahF4c%r0(Fo_C!j
z`nY*A&?=a;*(HNRozzH~JC@m7<9bwoq7eH<?nuZBPKcZrC)L-G*wcZw1&&vV?^x2q
z?JP%V@8RT4DhiPW*;fDh${H@N`<{8|+QbJ}NmGsaM<&<I>rc<I2`YoW+h<Zifex9k
ziQxPkoFV?SAaW;)hyz{OS(enw-hnjBUG=R^wipu{FW=u0R@^3m8?P7(qlBGG`70LJ
z33fbJ&O%66*r~eU5N+o2j`Jj8#Sn7&P`QFQ$C!5oiW58|^1d%gf-L$^qU`H7j@)ik
zkj6hx%!iYx#E%!I0&X!15_Uq3x!}}_DY&ov+#h9IOFb7Eb<W4~&L{FI%70YB<1*bk
zRaQ>$5{1J#e;WaC<l*nhwRQQZ#8e4D${)AY0ypv4noschp&14H=Xt|-R(sE3JY77i
z_;#GKbJNaWwy}2?_~qcPSupy|o1H4`tRh)Fj_c<Wvn_%WiMJLug4D-jLp(aOF3P7i
zYzaXZleW^QR0Mt3IM<bO(`PvqR2ucyFRuqByzK{RMal6H_6oFO><cYhai;BQj!!R>
zd*~^<W)K{;py-+j&UBPGzB|#=*7YB}r~-ir^ph91CaQ3bol?=$SGu2#o&sEhjk2Qg
z!jSi7ZrGS%NlKVMNq?4$8{QMCwVp++yb9a;o`AEf4m8z#P1WV$v8>CPM^C{SH^oh`
zE?=#;onU4<k=P8PV3`$S;e3(^bN>jq={@HF5)<XopNdt8B9+z)NG+@ke5bOUS*WnN
zsyyKH9xeB*Pog+<dKX?gP2ik4-oy57E`SokLqBPrF2V)E>2nt=kg<r4<a`^h(RA@m
zy>ZKqv$B8<hv^y8TQ2gqnSpmTjA<~P3?w3^0N8f;#flymH@W$s7x6|_L}9JCEY9u3
zYr7jSGix_rl32OOloF30=yTv<$oRO`Yu>=P4_YMJz!ILo`=9mbg*Pr0TPPt4U>BFq
z(vZtlT`I<8nkEYH9{(VfcHFpn0PK|IMcp0fa&hn{uqSw-nvrAdg+@1*b<fo0g3I2<
z#;&)QyQn9V*f<P}nif_X_N!M}mH7zM{K4dtW_U7i4D<5%?5UhfO-Hc0>xn)0ll#Ue
zd`HD<5#xomvbUPx#Zp!>Gzf0~uCJ=ZL)yn!77d|>W?t1ix1h0TJt<`aA`ML-TD6Wq
zCKATjSKro`$D--SGRn(Uj)<K874s@<y_64~wlaK?YXn48b<K5$@JIO_cJ<XskfK8T
z#4y7;%cI~#sKRuAB>tm*Z5TB3xmBvK$$4QNXx7ScshQMP-{U0Hp%&bV-wt>aog^V!
zWZK|1bBIo+nB?1+Df;E?>9IDuS067J++h~ga(w`856q*TawHyEcqB_p6(#DVtaC?U
zHlM>*C{m>;<<*j91}N3)+oZmWN4qv0XXNakcUOJbJg8E6a>t^O{EM6yH(FpumE>Tc
z^A5u+`wJShsg>~y7Ofu5fPKy%`NO7Lr}gn8qNi88-W!Byl%UvCAE%xUXH52_=hj~|
z?ap9eg&tVikGGo8N{zp1Wly&wvC#evPCX8xBmdz{Ajn^e(r#O%w=@!Rdq#~^Fn)RN
zR28B)|A~OOC_TLWqbQz#bU;738U1&;jA364hKXB@alY2jBLOVwtQQ`MmspB#6mOKV
zvf<}&zruTR!R3fuk%nn@MAPvE8I^{k_XO~3x_e6yR8*WrY_o6^Q5D?cd5Gf|h%&(O
zSH<9l3Q*d}5)+c#Ye#_6eHLvxzPVRzUo%$~`M&d7>Aq}NI+f5}oDIeQdVNQdcf7-_
zq8qL@PGs9_YPq6<CwTvyJO1((jG|+N*2(&8<7ms`(WwZmudpn*=sJ}zDW-~edR+VD
z5F&K1FgEtv{Z`)$MKWor<FhLN#ism=r))Qf7AF%wKN)xxR^Uz`ba6+SxGP1`r3G92
z$JHMM``(rSxY!pDnIymVwzgi(s&wq$RW6~w`<u&{g2T@*Z#N?=OHkx9cPTwzQpK2I
z+JL96zT1-h{kf@#78>!>lGN1(t_V~e34f>ecc!ylpP_p6IX`uKhf2v{C}8+6d4pfL
zn998^e;#?!He?c|!14kO>RTEAe=h*xfm+Z&kFs9S6dS5*g2`h%+4O;KL^UX`n;P=@
z(qgSRyh7YRZ>vuCrxL9g*ucdRy938d5T&4fq={Wy-ErqZ_}GgmgHck^O`KJl;}=7y
z4vD^Y=a~%5WVg<G@&LFZE~_iKYxGb>2+`H&{|6hMPAc2cA;jw$cxpY<&F*~gd+#o}
zfW*h~)YLPF*mS<vY8h9N$=x#67U^If87E+mdWlNEX4OC3;_a%3x;Sl>$5Sga<-CO8
z+fFx#&3WDa*o*a~qkbOzNR{s25qRi`>T<p=ean8;m5u|^(+>e1n6X`715x`=m(nQ<
z7*QN`6KaIjC>_!Jyb<ZIA@##oW2YTly#gL!Hhhw<`Y(p2jRLTJ7@l()_>miG85ED$
z>fG-^z`bN>_-~gSlBEg_+-5750V$IQLW3@L^mP~C>OlHPV?YCwi_PB*yR@fC3uD@E
zN0EL0YFqvzmCk2zRM8h?oipl8nXcnA9S{bZqD)osx4iJt{q-ePuulHvMyip)g$)?t
zO8ur{f;fsSQ-qxsB4?U0a&Rg`hq){7nf1P-qVEC_Fiy|QIe5o|R6gP0n~Dc0T4D)l
zM5uD2WfBd<!9#viz}f2BG9h%dZ{+Tu`F3+|H{Wc>9a)PAp6hnX9$HGyGTU>P0e^)R
z1Q1*T0Mv;frzWTcf+?6si`YF%f%=exD<oc7X*zmB9X<CJiR&SQz!4FeX|d}m?VhnZ
z^0(b_R90NSk_mkVp)$aA^D*%1-rv7sW$e7$S(qx776c%@$K9;0wg-ytFYP$$q-jJO
zMFfyl4huV!+#fe8EQ}&zOxaf1=`tB+GO<5+qlxbVo;$x>f)X4QNtLS0(l!t(aM%Bs
z6Sfo|_C8s?8$MrUdio_vxWb_?{}8XqVQ2Md0zF2gz9Z=@OPp{L^18p_VL=o<E|jos
z5yfXnD;kD#?Rgg<S!Voszz6#9XMc^$8Z!L`#RJLh(G5>w(zvsx#Aq19eQj)j*)=xr
zSyV)l*Beeo40!X_hf%t<R=_ICH;`O8<ePCufcz8klf|no-uCR4ZYS+AfShAE8fwnZ
zRmeD;$vMLCW++~I)~;`%S1YmihYO#@`cZ;4j3gC=`V|*6RyN+<!a5w|4kJ~@L^`f3
z%6Cw%fE;uCdrIAFz%!FfQN0xH6uu5X7wVeol9;$Mz-ebgES?C#Ix-5_&7IdD(o08v
zRg#4VxAV*Q7=Rm*Ex5%bP;K_fa^Xjdjh(x9z&h*%$Bk1E{1aADwVDe(@pEar?nCbA
z6a{2N%$KRfW1X`EsbuG1L(`|RYsp7<{RSxdD2*B)K5XB;ITrjD`MbT~bj&jypZHV2
z6ur-?5~=effZ8%siiD(oZS`-E;rjGNRCKoBGl_0lSNWknNqgt#I#co0u&8YHR1akp
z@Awd-q(sUp!aPE32~W5YIPuWNcRwCdr39Rro@wEbrI{W|_?-3ORCHnXAwfRPlZ8;E
z$oYobPdDYL2giOH?F5$Fp?b-Ts*gaO=WjIM+>nZe%2;n8QUeJK>>Mq)^j;9xsyM?F
zrH^*!55DwAtsQJ<*BVM<lR7Z`j|Ha8S^bw`eH^8EC`VoBDxPbtVXB7AKo8A=SIhT<
z9#6d>Z?=P%G(eti9$8#flGeQof~;6PnmfiUp_wQvas+J-<N}XHuS6#;R#3uf69huD
z_N_t>Dvs&^&=5t1&CIcA0s!eO36S$<6v#!W-g%yqb<=h=X%C7rurd7%XPOZ*^w<Sr
zjn`3g(2nZqY?mw5!;Vx=@hqR9v4$aFFn0Cs1W`=dE7Unhn6gBvrblxPgk~KO&+cb`
z`~Dc?H*x$o2KxOeel^hJZISv7qV1;mR(gBDW2_Wz37-ss^C(;<8A<Kwb0O1eF6&sN
z54|hJqDRO3&bw4;gkh0wTAvgCcwgxPi!q)a2Z?MGWo&W}Ek+y5gv&t2d&d;;*TKFL
zWLl*&mR;ND(sxFR_)UVsiqMBEHBs4mX;boV6m%{QqGXJDvEfc1{*>288vRu7raZm~
zSB}lEnB;~crPBd%7hR8jFu|1O!Lc#~dA;sjcX>YwK*Z=B3fwZq(11{MB>a%AilOG8
zSobQ6po#6;akdcgdL8m{s0P9<Y6p4aSI4141#;T?R$oyD0tS`L*KyFep1o8Wf&eb(
zX2x*v8rb#+Zne|ABFIxmU9PQ3cPfAJgDwkca#jZWSh*--g0PazdPBlO(}=AwsES%m
zZZ$YUme#{z?vH$+km(wO5KbUed$daV0`DPnL#EN69768O^|iDiU~WO=+s&=x)YQ(T
zZTtG8!M*XR$vF6obT#Q|Q4aO4C*#gE{})}q5}FyGC{{9sU^-E@gO<ZXyecA!y~|Y;
z7cmB2{cetFI5(KrLeqABVe5TnBpC%AmCtK$D_JSFW_Is1POQ!{)!vR`k1@zibQK1d
zGLAO@9$6@*S`UpMs^_hyPdNL%!VwVCyUXj#WleE4gxEF7*sO}eU(04c3;B)Ky_mb>
zy&t&K?fy8A8#N}K$?2SSxVXAh`J<(!e&JYo*JXwfxB|`o)JyO;%vvG!OKv_t&W&->
z%5pDPHHrttgatyB3g!~|7HCAnx>SRaBVfS!Mn}7|tN3i?`$5K=<vchs#S4{t;hnJ7
z?^#T1OIVsUWSpuH&AnSMEEZiWzSrRPwC+In{q;3*E4LRm$aD4mX*{=Zuw{wvSY8kt
zNcE})lYUowpgYZ(Wnri@wft&{CX~1jy>fdK>uILyrS4>cR&-ND6l8JHh_-w{Ov?Ml
z_?(0iWSM^B$am1M2&AfeIx@;e^jHCm>27DK6f!}3u2)}?L*7P9{?m9!ltNPaoma>3
zE_$wq0Iy8#`<@O_ckpt>i4f}^$w<tbZ$;iE9VV3NQBc@!5)-VA<tj!&8JrDitG*a+
zp6G)q@C;ot61mIqZ$5+q>Njs#Krfpl3<Z5QZ#xe(AtB))0+v0~mvf`3%>DoLJfHgZ
z;Ks<k=Q(6fN+a78pvCyzfzOF75QKU=Pz69v`U%|+4^9wIFS15McJ-N*zdswEGH|eN
z3a(2`8g6BcVI36gZgXQt$93LI5FVVqAq<Y1m08~*iw@%|N?(VVEi}04)}0`_i~&US
zrSwt6+xe^dY0<4~Klbb^%hPAg^wZr?AmfvWpcRulR}GjcT%Eqy>RfS`CZPdv^jSHV
zqP1vrFeq)ktSHYxU>TGJ&FU#MeLO?SY|giib-j$7`?*zb??J4gM1(}cY0Km8%B%5i
zM0a3ycVtlDPETq5tE)4lY$r)2+rn5?u8Nf-A4ihsXD9X|4yCqT8u#42C-PZ=bqOE~
zor=hL!c$RBZ(GNn{F#KKGbW-?Oi0ebwxGTJMFcMJ5@ye>VY&op`B2&x>1c!!sz1nC
z$GjWf2)?U!0{($?k3Glm!BQe<Wr1NIbT5(r*%<gm^1>50Xa)Gmxw<$;crtm+4wsD6
zcZDCaW+O))wUaC=#5~!HDE1In)OXS>6&|=8O}|o0FzTj>CW03sMT9u1=wB0TtYYi&
z*_*g@_to5V)V6<qd3iLp>uKwq<SKfladzeEW5`tn8U<N!J(lhD+f0Pl=lA74uK4Ob
zcCGO0n^~0<K4C!mK-r19$^IpUp|2l1gP@4gchmhOqqwRUp=4d0XSY2nyMz)=*XVJ5
z_9~rQi`E08^9gTHE+!J<*5lR|t>`fKSaurUC1;FajsFLqKIOQF;ZsLzNZE8Eg>=t1
zaHSZRFYyt_cA;7(B(Yt`l3mG8xiIN~suf0|f4nmQAJt4U&Sxm5Le*_*rEre}k}EAm
zwP!1*9?>F%O^h$Lc}|XCv3E`Hb@^^^gvgi-vSsZ4l?_YOcJg$SnxF)r!zr>I;4ll3
z<x6hM_;vNmDq!Wc0pA^0j|p%!aJP$9E`6VVyJaKT`8JdC=4{2jkMr9KU^&I%+TpF&
z4N6d9yjQnUS_g{z^xEx9A9OiFpOWc=Or_2sKZx!9I{nlpU+B+%5@oh^9$*f{Ae7+^
z6HMSX5%6x9+(Jrz5W_(VYCrJ>(BEP&8dvyDQh5~ad%Nk=6tSn=uKwxSR_1VW_-4Qk
zwE_fGt@-xV$kFnqfg!uimxa)k<IgOyTT4!$sN(Ov#ISdOST0d)U4we;_TgDoQTw9r
zcJ2<jt#z<PU*6(vgWFB?o#7|a6v<=1PrUr9<4tBKd5|w7%EZw(TdLJP3KA(YQfu$S
z;H8}z%#N-*s{^c(wq`#Mb%&Lv-(^s)6{oD+C$bx@JWSlcc&ms~?7Gvi@a8EIm5KXb
z?(Vt4RZBV6MTt5=2gTjSc_bkiIleJ_-=(dD8;0R}fC3!&e7IK?^JF;V_G&~hc;$y@
zVb}XdD*{%jm_R#~-sX#Bm8KG(64e#D7TbMw<IO{^F41(6#NNZT5J@WKrc1aEojA*B
z?MYF#kj3AMRU-cJn;)=y%#!ZuGvB`Mck`~1tJS<k6JgF4J{yzMUnh?Em(^^I+$^B}
z!*jTbb1D;wB44B^HEI-XLVq{?I6T4vJ4<elf{nxiUl#Fyv@nb0+nH-%R!s5)N1}w>
zEdm3<#Z9lGzP4WVJEPgA=AVbSyihqxm$=BS``x!pN6?iJ()-0P#wZDeA5)Wx2)bj8
zt-U`S1MdP44koN5gdO*pVu%Wq;*CEf37oIpK9apd!aKJ?PgBK;jHQR<K;dSl@~cgn
z5c~S_k}*W>9GvJ>0&pvB_O1ZD*)mVzVK=D)O>8dL$uTh(GgzGKvtVR4OjaB>@kpcz
zKpB%;`)pq1iJxX1qvWy)fru{rtnm-hokYRIRo_Ebmd8yg0#sU74mPAJMhFq~5rn}w
zs-oF|k!R}UU_tvulB`cHM=K(xxLUeL-)e+jd9ZI)Db2!k$F|O1`d0<|RPv7{f?6uw
z!g0b`A1NH!_ze=n)CK1br4&{vMo9IIVB}cK9KiiQ?mIKA-*E7HmhVQ#I6`++lo4tg
zg)MdX=)I3u8|?X@V`lkCm><&eyR4(@P5GCF9fuWfl_=J?UvF+N8q30Bo?*M&^gFE{
z%7A$jp?X_#4q(xjQH1LxQyLh~<3UiSgbwTNdgK8^EbAkp9F?rMzXEUAe_XEL6(f|0
zcqAQ0ho}eMIPA{27h-D_Hi@U+GZoMF<vOeMdoePap);T>-kB@Q3;$5VI;4uezFENy
zc3ysQi5#YX8aydRTf5o*$YpV5pld{kpXT+{WK`s(pHRQTIp7Cn52nSr3Se^xtmY2N
zGtyxqQ6X5VdPN({A@O#^TLEOJXF~7^$s+8v0L!hn2KAPa>Dzf*>PCETT&gxTl~AYT
zfHsFpF?!$y@2UP^x!Ammh9n5b9T|2~77df{K#29dJc`D+`W5Cw35dgE-d{xE@WQh3
zhEqs(Up5EL)+)b&rm)h9b1}ecL_6wl#`2UyU!Z@!-ikW^+7T_zE6b3XZTGZ?IXxBx
z)R+0-3;mA{<Q-;k-}e+?Cq5+S^O=)iCIh#TiYQb&>N#(u{%+d-?(#m-VPH#BIdNOl
z@Z74wPO}0!&2=-&6-<^(|Ey?e(snjSqlDF!2R^km=q&@oF}grw9Ul%iAlonHme=tT
z;^1PL1|H%w-+}|%?y~G#9I~$-dq!WGl8lqaC`hmgG4=FY!5_5nFj9TpctDP;lGrs&
zw)*{4mp4}`7&nRtx#y-zPOfX&a@&V7(U4B&SR?cMn0>Fc*CONO%S>)W7ZhR99(p`i
zc+CB{O-7*npBv54@;V@lR7$+!g=abY7FlGq_56C>cZEy#goC1E`d$}xfVpD;^daLx
zE;THvYubDym15-cKu+0S12ZD)lK%FBNl0)dpzfdH<#Rqr?ibg-u!)gRANg=Q>&Wqy
z>o0MypsKuI&b4oSB<u7MAszSHKOR2`RsKg7DbvPXr!nj9uo;$H3N^Sn3ZLpOfjQi-
z)QLQ_3jfmGE{05{n0ijoryuTbt~`7~5Rsh9^b!V{*QcK~xp895%oV?!*k1m4dS1Ci
zh>)@a$Ck7>h`#I2vjOgkT^z}WHn{0`Plt)0z9cbv8h!xYJfTWXm5N*~{gmgUs>1i%
z)!sedt~QK_XD%F;<3C1(p(Zl9;FT>Mel|1t{bu*OC#WhS4NI0DXZb@G?vqhePhfW!
zyi4z~6hmZ-NW*dgd3>4CRUN<RG4Bo^_b9T^OmBTZKWQLPe#8<c`ORHenx^3zN0PfC
z(qe#Yz-Z8Js7L>G<qn!xE9g*JNLYB$L2oI4#f$yVFle}p#2|TPd75!_KzI;kJTSO6
zQ)#E}3jLhUF>MaUu5*n@f_WdI*>O~mBW<cwkd=&SW|B1MF4-{8oSbOl?SO=rHsf;C
zV~hSn>O8SwX;@6}Q^KCZSbTb<`%3-l-`wtZ;!l51WLRZAEyzLy@m4-cY(b}p;TCMv
zKQ9(D{{GCcAq|yFA#1UXxNd9XrFq{IOGJQ5BZm~7)Sj(1)IG2ckDcF$)S@p$2qicc
zZ@Ywl5WJ@R!$0PUNJX3o&Y(>?03+Vvu*<`!f#Cw#88nzdc<YVsdlljaS$IAN4G$o&
z?vFfR<7R2M^*LbIg3s#J%Jw)hSTf6*D=NE;X#~rMtSDr89CjGT8IKik-WG%P0G74n
z%;-?+dLQBchq1Q~Yx7&eMXNwdf#MD=w8cv)E<uVHEACKSgA*J|afjlCKyh~x+=~S(
z1Wj<a;2P}m+vlEp_POWo-u-8u^*n3kn{Q^mnOXC`gCzl_O=s7w(C%zFmCn3V1Q3i(
z&Rkm>a@rt||CXwN@xSgs>4$|%T;!m`Bu50>0sRwu2u`ARN|sbjvGG;9byhxrY-#eJ
zKV22fqCRiqW+?kt{fHaY8PBU)LQs?&<UQPboK@v=6p6Lr^R0W{&gqTwR&G#8up2B@
zrcwQs;;vuY5)FUbMisQ{Yg5O<?c&Y~j}dd<BX6bZpq$C5H`M(X0uVt2RNC17O(VF6
zU*j~rphphOE)jlnI^{=_LxF@`g41YU_#K7SVL&g`jB$&p*Ve{v@C#=7(l`EN6{Y<)
z@`1E>WS|Q09^rZ387(&?NSsF@`2|;?D8N#bCvpcnNVQONibZ{nXDpllJAOle#goN`
z;`<hjNUVoYE;R<c;K_j*UKDh9tY_LYl5Lz1Pb#dYh_4H_l^F0qmqlmeh3*@IA$Zs{
zG!A%FniEn+8}`oG>2JRB4vUJ*>L^rTX4Sr53&F$phlqpA<CBeTG~$oiP+@$C(|+uA
zH5fH>A8Ku%kVhGowY{+@_1;V4jo#sOCnEa?a=O?q@y*$c)lmb-LQ*}e3)fD<<US~p
zPfuGufr-5WiwK%twHD*^PWC@GR;tib_1J51dU+Q2pOHQei?4s-^KIM9QgcVv>$iU?
zE;wUVp_t%QrWTrfg;X7747Yw^#v<N8sE{eZe?4#LHwP91e>?KL2cwdyKDkFbYz=||
z^fc}ZMWbCC)I`HVOMQ)Gx~(_Y9eJnK76#<^;ADh14ZO1p7dQS&$4@^l+Y#Xy5<<T*
zQdBc@gqrLg*ftya8zTW#J=UVh%y=6MT4%i9<c2NMTVo-Y<bfIn^@I&RL9l>FVOrs%
z#vonOaGo9&aSf?cL>|&@*IE<h`Y)y5C4RE!W9;#jk|(by{;HE<-d8(C#6ly55kD$v
zI@<=8+v;!@Cf$E2q!}{9nLk;wPblaHK+Cu6rfOn5avR{T!xzF{9R&Z=kjKaSV~cPD
zUdt4}DxXQ%W#78NomEjc%|LuZJ9p@y)oMM*@|cS@?=L+*qB3>9HU?pXf=(*rK^$wm
zz`6WSRzI2ZP7VHr+x_Y%_RU$P)$P77_a2)===qN176p$Hq`Mw5^h;)KuE0xsV;q30
ze<z#JbKg*o0ksjdT(3nRI1tReiz72^*ekDWT28WeH5=JI$6B0SNiWdMbo9O{&0rY;
zFzMab-W_jk=b`i`c(_uSOH`@665sJhi+^kQi_`-h>op7{1_l1kb&wM#y3t}GRl{Ak
z_K(rE3yXfU4Bua9<SW&y;b=asuIW@h#*2<4k<euou=a&YZ#FmAhzQ;8Ml3P7iO?jK
zYJn4-*Zrd&h<Gjv6|j5nPu|}@%PryD7Y)`2Qk4Jz!h6JVd?!;cYu}@G9zR{^K2}7s
zFl~9$&gelWLP49~;y-KgA{Ny&R_Q_m=OFl_nLOsqKum;o>=-J7>j$PVHW+qJ`#NUJ
z)nVmcBR=$C9SAku#{->@<un4jOn1sm@G=?V1Z$ShbmOnkZ&c_5@2|x;sx44GY>q~^
zm&rVVz}Ba0WQM7SKm(C0hqGDtyZFP%cf)i1<{h&i4A(~ntyGsi0D~1^f`%FsO+T#`
zBRO|vDnXWe{6gru<KO&&yBS@j$FH`%RU@$M4g0USq>@*;HtI#T?7?1R;uwLLnKjG_
z6#Q%M|GHdkzYTeSeN-|D46nGUIG<nfiF$yWxuH=JQ{gj5dLM89XBv#Jw)oS+TrY!7
z0aZ%T0n?S{&MEwE=9{S0mprYt{4e%A`Q?A|3%+Q%c1jVl0Hu7)<E5eOK*tfhPhtzW
zsbVzmd^7XbTy)!uK?*N2Tvn&vXOX}=V8lFxiui$>#(r+4x@~NxnqbnW>ZOHyFg9nw
zNgPhAt8QS&!r$P!|GsAV>O)I`LvB0&a%ID5xE1z#*h=1ddfZ$<0$l+gou2k-f`Fy9
ztdFv8GrgcHmC$FNie@|-=O5<cPvc#+|LQ#c-LV{mOMe|`7sbXvfu!V2+`pLH)KRVN
z7UnQA>_=x%F%SPO%KdlY{tEtfk7E5+iK9FGI_K1ED|)Uw$##5X&qe~J<zcD<gPaZ}
zDoJv<o#a}2%>3oG{5xfdKgA;nxPwM?RB@_6)p0)`v%Ts>K*2T9rkL}vl8(gy5=5jd
zxwlzW=sy2n?)Xqll$wJ=A;(JHA+`=#qwSM61sb>8lcy=G+3F<$6GVXZ6*Wqd7Fu@m
zUrcHLEk5}7H7YVJEem-8x^&KQo1c?5{B?jdgo5hC77q$B>ilO(THVg%9z?&H=(8Ox
z{;kGqB;yaKScm`>OT`QD>s~88{Y|@y5SW|itIbz#5sSZ#=1Bs?y*e}I#En0{aIo*6
zwYtcIJWYFtqfC_eY|J09nQcIr`ayZm751sByiX<ch%nFB(&5Vyou-W-k70a7U1KvU
z?xECGE<zIWwrH|r3_U*|X5`@D*76?lBYMX77cc%5L~)OVH8x;GBvc?RL|WU`7Cclk
z8!L3|X+0t|k%gNbgrO(tb)d@c{>6IXXj}zhh+<(fSs{lhW;NvzE94v!TnX8y8oqRn
zY#`?#qyZ}9RTPpw>^E*A!`c5@42UsSHppDrqzf8aik)XEw%LCQQP>}E1qOLud{sG7
zE8{B1p)xHs5QjNeJb8kFfgx^aIAs|1bN!O#Ap>m={RgTi?!tH*mp_HCkdQLqfim}C
z%umqzjw?bDZ1i#NWU4UgbCTn+hr->f8@&xqeXFluWip%rRfU;3>gaUXuaX7v6T$CL
zj&v@_LRnPYwi|L@3)Z`D-<8tt2|%6>!D0)@IRS_w=r~TF+rIE6uJ9z+51X{@Xj32K
z(q4YOt1Yq+8w;V$!(D4$TGTZ)-VNa8Rcx1g?K&Cg)$RO-y?ME-H?FMCSziq>UK*1H
z?FPODGnHCKj5Jm6KWM1RvOxZNp-qFDf%GdLb5W3m*n{M+vc0YvZ)Q`r2u(NQ5_2*p
z9$o=H?-Wr|FhXG}^D^*%@tX>|NpVWkIuFVWq11io)SVO~?F6gDdi)L&n5ZZI>8vCN
z+;i$(p>J2o^J?|cd;qv~8EuJ+d|T?)cMD!HyIsxl<JMjCot_Msu_jPAV_g>RsmuH0
zT5D!Pcq=<U37O|z51}tE3mvM1Q!45UB2nAHa1BB55>!3k!hRQoI_mh^_90f&dNl#=
zQ%tKfIr$sSkl=8>OOgYH>QmD7wj0x~jd+X|7F#X->IkbXfclx<4E)^G4#7ShC(h4~
zn|9zTBIMHOd{0Hfhi6<QlkME|(R#N}MHpyJ%QfM9V1$FmfcN{+)4N}DW0<$E<5iSZ
z$BB=eaJr4uUK(mq3KUlqA*^Lc`L{={Eqj!u53{2_!6!><A#-{aW&Q3hdXR*l>8wKQ
zFJi|F!VX*v#*&7aY$ZMh)0sbFCM%($U%iw-khx{SyrQQH*_@cg!$zLy1{at$elnou
zMsVj7dApED90d`}a*V58@=WlsjvQmYdqn6J(LG`()q8Ep>sMNarXT)ze(XJSqueus
zqnXKRrO!Z#E>&((6J{$(&j;Tp?Wzb$ys64GU-}YuYD}wea2d<M&h)^x;Nj^dQsJ&D
z^~&qlU@&yFh+W6fTzUtY$N!=#dB=V@+E;jOyK!lT;~=%d>}Iewb1O(c(b9WfbIX}o
z)!UGfL}aYd4qZ`gO%JwMl#fBq`HdH~BH*<gI%~B`O^Llf&*Fs8a|oF=_(zbSvO2XO
zZQl+c2mKRd9piavotS55Yl`z`u-Yaq6hgEmjlN`6HT%u={2lX9TrKtbjHNQVLC`ZP
zf(CbrMisV=5K3|#))w~Al`7Yzh;H3p6rzgBy<yo&11k-URB8;(0xa`vg@GfJMPL_Z
z`W*Hv@^+!JolyCzJ@sq%_y*&tjz3eESw)pU`-JaogAk*wW0*kFhFZ3gwDk$8pZQKY
z8vJaU3MJb`R7KRP+Jel68ksB3;4zz@Fr0#$rIWWn%>jXmlpS<-^axh>TcFa_U~`pg
zBr#=i0Wea^8oAXo?{GTC1ATE;Pzp6e+HJ2^xsu5am<3~AT^f0c&7EPg?M;nl7yIP5
z=rwu4`;Q^Sa@$=Q(+a&iI>ej;f9w!nz~KsNBwcpaO{y~#?39na?`L?*zXLKnk6G5`
zJsS|7treWEKC?A2v(@1-XHiYEw`kJWPi<ko*7FNNYNTfPDsPU{3X_$V4w>j)6(7r?
z5`7#;Jrzy^de3UIN@ls#foY$W-dWa0_es;|R@#j6kLgY)><klPd}-nxR}PqX1^~qA
zr^LlfB61_+xI0<Xl{-nj_;9kH_4CGkh>I8%QRyY!wNwgx7hWTISrxBbb!L&#x16$M
zc*v4%-X$Z6iO*zW9Q)~K&-Z3!%tJ*l<qD1WrKI~$+Z^5#QCM=T8bzG_)+hr&>^CH`
zCMvT(O0qw1J>rs-zA9&G4t5#kMKIAhKDWSij%F7>kv^>~^0?3=H5jbvG%(Vt;1>E*
zV|X3A76iO9ll#4GcH%N@Q(WC=q<V=K5)a`x2uHnUR6B@hMue^q@!!54sgI4f-AaZS
zlMsR2rRTTWB}6mR<|3mnvz)8qGv?F4>j}e$H%8@oCdw~h7drN;dy;Pj*D*A0(%>ha
zc}G?=h#nQ41=|*ggqLU8fD4>Qrz|pEbrLlEvARgeM{T!pt0EdOrAVAHY=g`Mp;QYh
z#(kwJHD0Ontc4p(<;0blSYr4CdaSf?g5F?e&|4V}X;XSbN7+$nWvZ8>R*=f1x&_Ok
zG^)&nF9IWE*r-@*eMEUQFNMs!%1mx!o|byX{I(l5ioU0K8yOsuV{!`7?`9|~QJECB
z3|F+H(KSH*m>3GcJe`zksaBB6!~seK>nF}fp#iRZ7_Pf>(G{O{t&y;?=p$K_MP6db
zkEw>@>TWl4wl!6tpdm6tkRY<Od<dSM#I2seR=2O+3gSZ-=lBh1bd<>FM`rE=xy#Qg
z<aIU(ox6m)#x|_97wYd^4b~7UEqdS<JP*6WPtk2%{k+0>izE`!$*{?3g|N>yNQFd|
zFHKf0+eJb^#-dWAiH$7SwA$L(IBdEZrfpCl<OAEXG8~=qRxAhKfR3y=CoJCg-+884
z0oFhuSDHL0l#3HE;c%08Dk0%g!B3WrR=+7)R2wO+cpS$bbRz68LSgzpt<3yX@w2vq
znOExzViZDQXUr*n>Bpz0+Y<yOM8Yq3cj=@djVD!C-ExZNQp|@gB}t_6Ah%Hhkj+4B
zt*N*(t^=E52X;Ng)Ge+!L*`i_QKlu6+G$%M-zPbETKfB04B^E<?S|QURM!YB3G&!*
zhf5L<*=!5&PE>A%y^!tF2}yrDFy;L-slQ8ehO1(wz`)Vsn&k2JiP>qKaw{T>=DYm{
zotcd4J&-X$`gxHpE@}55tf)C38r}Ssx&&rzcq*3yJIaHpsMM;nc>r@b2x=jDB#uVo
zrBjrNA1^O{<YH(Xnw4o?PqEvL8g(%UYkX1GjZ-}^22U6IiIIrutjf^@3RG;?hb?R~
z61hsuj|W+sh$>#G5tbm|2d??v&pg5)v4voa4j7Y*wLVEHq(WMh#XMifX@wav<bA4u
z<j)uY<;f#;xNTOVab}vK<E8jUCH*`Z>fNMi+4~T7ZsTyr4-4V^(nS|jWU7b}sS{l2
z<y(&WY$f*CVD>u`Z4?oT%e{~8ah#<Hx8(f%2rTA2y*4CzviIzR8a?{g*%s*y{7O_+
zVgkd80oHGMwMoM0TJ7|8a-*lbZu|Dba@-jtUig?ZMwwG{!`-f&DH=4~VFUf8Qs!6t
zs~04wAv{EHVu)i?hSJhUg&a9_;+czP8JLv<j~Y0V`|}$cs)5+ZNIBF@GbmmFEuZRn
zmHq{wVxW=f%_{d)K&K}fA&Ob;`Ly6vfXPsGVxn3BIWa1Hx;3pX5>hugGY&O6?z3!Q
zRTmVffXE}tef&8lS$z8nfC-2%Vo#(*&CE9mr=^N<vXxnd9~Y@+<OhAaZMmPYq@T%K
z8wG$AS!HRQ$W?gFtBH3Rr^TxaTE<Q_lz6Yj<v(HRiZ%wln2tQNK0BRP)<<GEKk`+S
zLfUUm?8%lf%1)3>6x&FeEU<R1o?eAfD-A}xR_T^RS%*J6>a)!7%{?98y2$yujze8i
zK*vwxieMNjNk5ZRMXO5bz^9ph#oBsr*zOS?701`4dhx8TCRhNi=A-4-oJL4*D8<cW
zCW6=2j=~kn%XJM9ys!q}LaL&5psrp)LS_?asHUXy>r~%(Fe>24*30ZtFU_o4{qp2#
z6M{iubV_JS;`pAlQx!Ia9+jyY@_g|7IxiDJ8X<|m#2-)zcda#~OdBczg7?R*gl)7^
zisL8Y?_5DNA24>csZo?bF#*LBHBt`)oYeKpa0Qgb^%W;HXDmp6KB7&G(k$_h+LdG~
zsSjLWNcDZ%<7rI*e#v>5P_7Bv=O`Y=oC2Vg&wI%pV}E&$+h7p=l=bT}_{XAkxNkA>
zK%K`nyn5RuKA)#nBr=&J|I}KyA}tB7_FfJ=ZVn!Ottx3S<o3BCS-poR?Z~uV;XE{P
zj`FgaLzh85Ri0b~<tE6VZ3~^G^NUs0R+u2U5i18rPR7~cUuF+_8EN+BqPoL%z%WK7
zGl=a@RmvpXK>K$rj08|cl6GtbxRI#4WJi5sNDL}*mfPW)(3sA^ogePVYAl_M%RkcL
zT{_}5P}Lw!tTj^qVglQUS!!95WwBW}0*IVpeb7EZxFHcTE=|B4(WDJ=wKg?ZLB-bC
zJ(?e1`Yx=PH9y)d1Sg&pCkc~;zcUGp>H9^(ZVONyKi<-0SY6d5B<SaSgI9<SZ-s?2
z$BAzL>e$MO1_LCv7gO-Hf1k=dxFj3R@qR2bPMj7b)DscCNh;lXaT?m}%deRxb2zH#
z<APP$P32RB^R#L7*&8#hO{1SXVxLompiK0HZCG+rO_+F{`O)}*o5v4_Iu$t#e4xg>
zzD9ZW-xCpVpPnLBOv)9N`{_IENf+4@L$u>UD-#$d0lR#)>@dw4My(|Upf&($pVq6F
zIFg=@d$!d!8TUyK`-A33Z8M0%gn{ns3UA3*A4r^ySZ$+Ej3ax6qg>=S&(GcJ)H^I|
z`3Wd@ms;A!=2OJGfH}n(i8YruOFySMJx{y>&jhLE3O;%k$&<79j%?ALe`aVspV`-3
zA&z-?X|SN_H~^o81Y<(3_7dSNwZ$z33tn)3!y(wFTw;_B-wm7t+Er89!c4DvrEaQk
z?oHp0=fj7D&H?IprDBJVLC(A=rNoUDnRPEjMshEW|IMXm)Y$@Fj)#<zx?Tw^>PDVX
zrw=^JIgi!~mP9DA+34gauh`Q}zqO!Iuq##(%oTLQ3ue=G3yx9}!8fzh^)}Rp(u}HZ
zq{z=;ep{>gMMlaQHLqm>yfT?fLfT7xeKQ>+QP)6<X&#nFygX35j;(lz+eUquS3p?A
z`RgM7Gs(fqq>XKxIX4Sgr`Q{TQ3JiV(wGdAMpzs2E`NT16e5|Hl!En1ggkRiD2tbd
zZ9gcgXdkF|tS`j<@G_B*-cv$YY7>k7_LH=zX!*lLH5oCm(22jfhj|CRz3h&xW1(o@
z0W;K-exbDBbW9~72Uz};X}ySnbmwhU8xcOuO=M#y+(xcx)Jk=V#n#CG);hCtB*H0c
zAjRrogU;2D7zMfFGi)SGvhfv{x|B!mfyte6Tr%uSE<ehC^rO&F*tgf{@ba@X-$FLE
zN=<gTjA~f1ENf1IA}f<rwI@s**TdjT^DP<i7&Q9@5c_dPs~W>|O?Pp7zg)ZC29an-
zJ|FXWoMfqwq%eef8dTS;x(jvQI$2|aobow5OT(9`h!VC}XJ$h-Sh+5hl&+OfE5HRN
zM$ybg5Th$qNmBBb4T1h-pLy3Y${jQjXM}VM@N!YLVO}SpW-@POUrmZVk!fY^!n^05
zim5+ov!JzfP-~4ixvF|$xSaalFX&ELMYBNrtmd~jj^<A2Z>i%re$*Ikgq>k#td%g8
zD77Rt2>;ZaQ7rOC+L44A+NNpes9R6*Mu<4cRv42m?f_(kTUyP|CP<3bbHt9#FX!$R
zOyp&a&-P^bmI1O9m{YdMSH!wwPu^x-2fzHHJ_0)pi!X5B)3LM3XeBwfvH`D63qzW2
z`>cHAhkOd_r}>M8kV)%N#nn10H<60_O&}15aWCAs7F((S)c8SNwpP_Ek|j}g0P%%9
z)3YU5-E`a_aopgF`3PldyFgagk`EFt#L=y2KFmOHtfSG(jYDzZF7AS~c-1C}{CkL2
zDnG2^N4_$jx=U=){w!R~s7az}m65k6eOBwB{RiXgs7+;CZdAU4>rv@^J%mOq4<u8d
zxQi1qoH1(p!^Rc0ZLx|Kj606|X8Xd`RqsRDOx9qD?so2IQLD8Vcz4<}t-{)mYqhv`
zKpYxAYk5+jOL_fCd_#0XSmW}K%#}Yd|L3-~k52Fcs<{+l59CZ9<L!Q}Ss?^0afCy!
zWeFrv$+i<&pN`Ra(gW5D`Q4>)Nf*Hde$QiL%-+0Js#WC52UL@A8BVt|sn!t#i4;=-
z1UOxC^Ut3unAqo$a?uv`y#%Y;E@+gze3<xpO^^C5)tfW%#%W(1k91BIhU)2Xc5=)j
z-<ikS)Wkxzq|H5oLNdruGxWFSgNXA_F>ZOmW@{2$CJ(9KsFS9lKZ%xEK4*u@^?Vr(
z$qraG(GK@S;X^#fZB|uw1yL}Oi$R6$dc#`%qa<1p(^0VobIMems>SZ}^{X#tGFEc$
z#kxA)UZQL9^|w8?t|wa=Y4t-Te(Nw6gix`&kl{^H@T2fN45p=`p&u05ioZx!l@`Tq
zn&aTcYX`WdAg+>iCU#&Jn#a2|EsmhZy+P$Gk2qNGrQuHWJR!={;uJ_jy>|ryLTcpZ
zEJwiUEwv+H^f(aUY+Q=GZxW|Zv=BOL)PQK_KVC4P+?6h9;23mPbBsLVehQJLK2{KU
z`~{LJ8vq#rL5$={XU4)w`ujwi=Pj6A8Z}jp2O~fUCDF|~Pcu2bD87f3R5OWamK%?3
z6KA?UEr<>wj`ejyy|SSMWCT2Z9H7w%r15cBaVApn`YDy~A$Qq*UL(a`RFHc?lsVCD
z0p(B~flil9)tal!XoM{{eTFHWjyW+e0JTxvUmi%W@l^KBOz_Xg+9eh1C%Lodp4-=$
z=9pfMF0MNtHu<VZ8Vbz(S~v-@czWd7GND$iM9vScW43^Qd{O%1@}f=-MwRD;u;-Rr
zs8=dJ!6!GEe#B6c!zZJjx6!g~9ZqAq@kvKhhO4bEzCWq~>|L{V_7(zHpe%;ZJJ|qJ
z4anOjJ_W?@utUz!+_@1prI~3-nd>$&sDwF{y>W}Z#^o7gHU;@?%t_*<<~3mQVi!tr
zcsmgisaD%0Qf0P-@{!SHT61U4aWz}XH5BMQl`hms_Nt29t*N_Dp(@Nf<gc{erLAcC
zNKH3<k7l!>Rk<1jzplyn<WnBAxpBdE7AV$)-ktRLt_OEjB$k7rDrCb{o*~aNJuL2d
z+Xee_)q+<X)>K;Jl7PuuHi(6;jiL9z&h2t!YmS;COX%GWHb7@E-=?l5eQjjdVO~i?
zHU+(6uG^-N&d_wm<tGB3hg(Q)dk9KFh2g;I?MM2iDz+tsPUU|re=BT5P~cfg&?&DV
zvQVxL+~kWoaxa^Ansx+J14wdbe~lSxsw#T$w<JszffY~H*+m%yq4VqTTiiOmyi9e(
z+Cv>lyZRfdFq$@rG5$a0F(lYGqLvJqXU}F%<Q{*t9)Q7V=RmNdfqTE2Gl=1%N(D1r
zw9yR@x%t2hR|V_&TnQ9KJwi~B^pzy#dK=YDG)U3mfq#BQ$r;F7Uw%?BdoKO?wIU@t
zizPzhD2%d<3{V}kQV>^+Vq_O^+<_trxSeq+n~ihjN8LGyj7VI&n|#|<zb_}xA-p;s
zMrFt7L_2XyqB|mXC@&#SwGyq9I?B3z+jFqladwH(P%PpjZWGCYVhXlyo;ReV%5+!|
zYm|(BW1kG^_U~j}Sm@ZJ%h?-$lt7#LH>dG3?VLeCu3@?n=lcu9)lch$Fd?H&+9PlB
z)tKjq<<`OlN^bIr?)TUVE;?o>*tUsYWP*|+5r@W)X(A#7-XXsH*c%{iH-4fArz?O=
zz^6B!FLyx<J5C<HHQwl&ryv`_Uze#<Yb$6{=Q<J^VdYK^|4B8nJ>8rL@i46yIhj|X
z*7u$|nwNK;5<neWo?E;*M_Zno)$aPI8ff06ml*51s>~!9f)o-)nm->WEFo=5K}9=d
z;IIPEpVo$d6#5Zk0&Co|yu@nN{0fpEhq#%vJZ@&c6~y59`2HT*c{eoq9R{7Ob+T<L
zgA7sa9$AZ!NS0o5vTuhi2Zo{Ue$}yUUz|4ipQ)>|$QfJ;NH_FAD|8HrX%nYFW7S7M
zWRm9FOoaTvoru@|!*^4T;IPTTlhuw_gmE%Y+S3D4WWNwFg=I^GWwXuvU>jgT{-6T2
zjLeVBLp0AYL@IVq>&^=5oU|%f`+t<l$0aa^i{++prpSJKgB6P<fg8}zxqt1tza^}%
ze-1i-gZu67!JNA5_IQ!%U71%?fnsLyecyNg?~|>$Ko)gu5)SH3_`En;pLryyzm#RD
zQH9-|rX2(p$@_Vf{$42Ac|cB^Z1bjQ41FGyw4-3MQkef3x~28n1_(j7Pm!*QW|bzL
zYX6k!UQpLim6^Hy0(E8VEhHhYEio$(MvHsd`WUoT33!vn9F`52dTTVX`X<pFoL-W|
zhiyF`nJ=jVEd~K=DzVo}g}(+~v0cwwneDo&H5nfj2H<FeOmBcsZkBvfENUPM$D5ob
zQ7ycsk?San=xvt#xQ=h?BW^dP8`=`CRl24Jr7DyI%()L(BSsu@uuS?XDz?Azoou_k
zKTv<Fk~XqU3lu%t@c01}L{Bt)=g%y)N}y6&REH;9vXqrr&g4V3D}k|^%_s8a2do;5
z7RqHllg!BzBN7KjU0SF<rwemQnCXmQTM2zhvH7i-W8W_2=14*aybzKVtDUJn`gZtN
z>X6C8o@eQFF+lK)c5F9A<wWE=r0QFHh<N`P_pv5Xgqh4zYZ*_hk8T|iC8zk4&<FUh
z9!Y$|dijt}^hs#@INi%T)a*E-eg1!6cz~Zm_GI_`iu8Opoh*-E2edKXLamO3D=v86
zC*fHm@x%OW7)3w&96Khh^M3B8qs_*7oo6Ez*-NimmS3jLsrZ7F3d;uX+4js6DW^&Y
zis>@{?<4=W^Z$<@(p-<|DffmUhSVGyGGOj1fdf(I_=L}9z#5+gu{wvz$ZS1s2E3C5
z@F*eZ7d6M*|Me*U$5#LAA*;gMD_Is)*hVr5N*=TKXbM5f^Vs+!WT?dp>pT*bE@Tx_
zb9$eDQ2Kuf=l^p@e-rA5^{Sr1!5@wKoPL{f_r+(zC)cS3w=ev0)9M5qkCuK>COJ=2
zP#&%uYP=fc?e9n*8VX}zDm6*{?*zc|Xf7yeFpHXkg0W=4dG!}=bei&IonK}1oWGiq
zieY|r_Q47`ZI%sb=sHy1Pi#Jtp7=bn5pZ?ZG&*{8Ko6WlAeIqG#M~+^W**?<e<vVh
zzym6FhXX325P1`mfE{aVP>3{!zO)GeQBHI!d5}u1{C;1ojSXvkD;Eyi)Fze0N}wrN
z)~HM%w07uT$t)*{!}*mFh_Pif`sy=lJS7DM$mj7*a2M;p>)QXPIRDYYXvHVvhhuYk
z;p@r_c=zLCK_{r?)rTE3s!^o|i8#~(Fg$7znDjk5A1@nJZ7C|G0RRBBCsqmnrNRGk
z*Z-l<i-qE^>{+1J0dw%Fjgi@zYL60R)m<!Uh_LHsU}Kp$!z_Yrl5}{J$(f1x=X2cq
z*#j{zH2Z7{h>1=AhlTwA?awILE?{UZ9(k#hz*|r@#LMqHvuvz7QSh-l{1qo$g5aro
zH{JdA5u0KPaxW(WqD2voD$9cZom_mMj4Nf+;|yfoJ&EI6JdMG%iVt^DGxo&oCl_7G
z+iz{C8XFC8_AJ$9;BsgNR|EdLU5TE25(!HZ4~dDErgzk#gz>1ut=%l$zoZKV*hys_
z@qt;v-91+8hCFSn&a?*AjrV7iA)g23SK<u(KPL5esW3g-J}4rMEze*7$T0Gda7MLc
zIOe@z)ikhs5?E*jv|@W6$$NaM<oAay1r9Law*Bu6&niR|!bml~(eENn=e13q%jd~F
z5rV9n$oY_<4{^`9>a?HR^xa?C&+1}$5@Pm`cl-OllvoUv`lUmJyM0)E(lai+n^I%>
z2gmDaQ0wi(sYxB04)~uxi#k8k=9X@Kr^crjaSB_S>Q}cISPyYCEet&1K5HR(_0(Rt
zzn}JlC$Va|NcE(3*INB2H(PJTrg%SCY7Y|n>pf`{MCzHCaJ5%X?|oIuq@*EdN{UH*
zk<rW9gm4ZK6cSp-rP&)uG<y5jS7%TA=n4}vJ!6iOgvEF`*F2f3i#|Hz_PI2kfE5rJ
z*Nygn=6Ix)vvp8qN#$`Fp(N7bv)w4?jB_`sAJdTL2Nl)~PLvSEXXN5;6uDl7*c~=&
z!@~3LPHF8O+3jcW%hFia-CK4EjWliKI&Mz)eOWapc95VZ5p>i32CVYO-~5(8`lTq5
z!sgeoLtI=OSgPQV<eCPC$IEmf*gmZS-MxAdNGs6AHJ5l#S0swvW&kRhSk_t3cRLRb
zVks$TKtcIVK|=(i@HcIhv=rnZg`)_kFXH02?0YR=qU@!AbZwjvCN<{gTUWk&OiNDg
z8X(?f_tl{fwZG*Wf_vWNpY<B^c2u;5pseP6(TF!E2Vy)-g|o6kmQUla+`xTCv*|cT
z4J2TjWR3$s=4=S9tm667)(1CT3d(hMqYXbuue+s*C*WuJ-_$4Loi+lS&Kq#j7M#+c
z`=A?DvML^CG5Mx5Y9}zNth@zsV#E=*Qt<w-dc}<2{`qbC$F6D_URIlHeWxT`cM5QT
z`kg7a^Z2opS(USfoK31T*6B92j47D%W#)^M)5h2A>~sd}qSVSF4oBAm>&bJ{$z`rj
z88>M+-Ga_=O(OzvkUBa)=PdoKN^;k~ld{9)CDvseF0R+3c1-=b{hF#-S<#rmv+LAN
z)X>ta8CCYCxht<~i5RTgu922)nz*SW$>Z4#;p*s(_tf~BKu^jr)b#CjNu5XUlD18Z
zTjuSenMv!yHZ@ar*swv<V9Wk8{3czhfudu}sW)e2VCU)npR#u%^XS_SvGb<89mAV5
zAg$NA;B2==`Uiu=zuT@T4pzI^<AN838-KppY}+|JC431N{K(kxF$Y(~d_O4^4Dea&
zWhoSw0Rv>*PR|<oQBDa#7he~rl|O;)P{E)-7Stkd3slA)Z5+Gw+lkiM`b}`N85TC%
z`29~VfR?jk`szQ3`N@2AIMY1~=Mxp{L_;L)Yz>loP`{|Ocq^n_sG#`86OzF#kY?M8
zS{?f0>0To5yc3`$&(}ufL)a%yLBUx{d(<hZ%~_;VND2r$1lP7FoBKU8-NTnyK|w1G
zoK3U+EF=vq!@{GlwTxCM1tl?(*ZT5&LU6Ab-?9#doLt-J&JCm$fFjENU?jx|J&{_?
zm~OPm)P;U6*Y%v;4SzU$mCN_}?U6;c<<e~Ga%nS6&||$1P#h-_s^g`f4dQ`0TkiXJ
zm8)h?=|sj;TBw2FSwtI^G%y$-gv>@g-Fw>eZQcthWU#sKCS&1CyNTQyhjmc>)nlxD
z`sy<Bj@NILV2Zsd!>G5s{ETF3etH*o;Jp=FkkrhxvQbFiq)l0r0&XiU-#3i6M+FM^
zgER}|@A_iL-wwYn-R$t^EPxD2Kf?(vtq=9S`txP;Tk}L7%l9TI@N_xzcF?{aqF-|}
z5VF}m%E{=}$DIaPeHDh4A;sY&fst*NnfkkKi)`t5M8vl*h77yI?_yx85i&rc_UfDA
z#oZ2dOylJG_q5l#!r*GBs;1kcnz?3eeMxQGrjxjm;do#c0=&(tLF1o#Z<4hB8t~Pv
z==Nv;=DWGy6LK-f_*M>`%pI~`Qzq<<w8952+N;!iOCxg*iGoI3-o&R{4Gd)t8{p<n
zz{Mbo*>ykb$;oNc;}Ea8((?K)kB?XRc7L&92Ufa2Y)Hu4_o=2QWm_D4X}mztfF|ql
zAc?zY;K}(6pevxfq)402MF(KGz-3%eCZ@-^+(2$sO7rp~qi%0UR)Tc5PdNHY?)4|=
z!+UBTnP$Zfeqjx}?Bnf&%ZP<mM}#dNMg=wW>i~WFhJP9OrO$!pLj4>LUNW>5{*Q(x
z6xPV>q!Rd>ko)&XgTzl3_ph&;ymP_Jk;3+k8%*aABn`9k(Jz|w<2&>^wVl#Sg}{(A
zSog77_juk!=8dWawin%Bg5ur>kEYGp!m1qSw;uII2zXTLxyd}T5)BX!6Zxf@E(9XS
zixc{#e?)IL(2Fndc@Zfvq<Mdu&*&)UE9%$sjJcCZtF}K^y)#q_vSv*aW76hOtp~ci
zja$nHTU%@kChYdJx~5Ha1w6JzFFT-e$zTZdyHd{u{sSm*IsNyU4kW$q<1I~*Q8s>|
z_j3h3yrH0X=lwS9hZ}QRWzz&Gjj+{zr;jK>cRq@S05)pbzs4I>x>!^{LaCcY4}Nmx
zX9?wW^>4p(56Qc2|FXQ;5%)fTfRsxKbcOJ!EFgPr3V7eQ0x2oboM&OOpBuNI3x21y
zNR=Sorn~x$RxT+2dmtru<9Gg?&PgS818QM=6kxOU0D_!Vv)$*yH$gU=#0Z|}t`Qd}
zv8U^H3>bfPq_DVMoS+L~!5LUmdVK#?!m5z@{Us@bjCMHo`{q4u$Cs9XF-*FYeC#D*
zE1+Ex>PZ?Ew7rX;`7r_Dsh6uGjYa<K#qja9hm6CSZTzvp?UpoNBPlPi#D>g)B<p_Q
z!^0IZxDvP7^fI1BBiC!<G|b>&Z&Om+Zk{k{I&NN*Xd+XgAfofv9tLl81ttY)gidtU
z&|Cw;Nhv1nPp4Fh{9bv8T6@sJu@$wq6!jtXpVq+DW|ZZQG=Qt^v^|8}@Bv+l^GIvs
z1mI}a#zP_8HAVH7CjJVCjBX6?W+{kmWgd6Tt4*7ZJMH&-y#Uj4ynetbNQ<jg*m1##
zHB^W??nWV`^6eJc#^wd9k<EX!Cx4TN4{e9?$K%#FT%jxjGld!C)i;z>fL|--E+l4^
zFD$?S#uD5C>43@?87lpw^i;MBBm$nVLc6z%W}ymE5qZ5s8B6ru(L^CJ`t8t_tsjr)
zk1fqLoA^RC>fxr!p556VPJdKOSfjhPlJpceyE{%O>zdd0blW;wPS7Cj*;1e@33he~
z$%va~hF7+3%PJqQ;fElT$AM3G146F4<Dd%w7UcNc)i}lsZ2L<_QG=_4qR%yeR72uR
zV3YyTmA&6ZLo5o_$<LUaQH7vx`x?DNqPNJ{{^@dFJY+77K-nL7<cY?Zwf9|Rws6#U
znzcP=*>C%O(?Qx6KmCE;x4b{&2C%nNC$ihyYTfUB-Ws8JuDdg|X?g6Xq-5Xb<#CLA
z-r|+vqWjyClwL16$hm2b*rLLf;l1uANf`VYF4KXdHW~wCc2fUp4!Z*AdLG(h-_Org
zRr2&wN#Ct#Cm}SKwKMfYOlxJorW>sJB5)CVLyQD)w!E|FU%;EB?zb6UGpEwQ8PR$U
zXQ0Yf|9~grG|{m4Pf*1_*Z&v`^tk3el<k_kt~s^aXbsS|hFv(YHm{7U?~QRz@$m@=
zt{!~U>-BM>&h#-W>EhRHCv{iC$hvdOaB~XU79U%1R)?BO(AzFyT*V5XEx+4e?a+ww
z;L=^6&9f+TEGXBbJ9dNWO=<9-vo?f`k{my9-kC2j09Du{^^XDRQx*o)AHL=6bix%K
zQg)ozes0hDcWo^$3ETG)vI9YQfU9t~!k&{QgWpuJ9QFnDH^JCj^kS{L+^qTrIDuhr
zU+ini6AM&9MVnp(F;j+zyLlhm%0D5D{!O&%4d<itCsf6lp6FRscKf)ivuAa8U#>M@
z5`XT0^%wVu@laxQF_K13i#CXm;myIR5V$q&^%R0}=dx2UMe24imgmN4e)!uH?XT2i
z1Azeq4|6xpL=CH*Qf!y<Jr1%cO+6UzTt>{@*3!+&>jhr;KMdRK_#`4wHO?BHkzdE3
z-HQmgj$1xB9RB{tor!i>dr;VU=x}erd$GFE`N=R9pFkyFljWZLU86+$D>vb@?^ivj
zm_GKS;~av@yczbEePiuSef1X%s>d(&{UtZ|hI#IEoUb39XGADE8{UkH&!^@(o#v)|
zE&x~NZu84kJ#}XN;dd77y7zkxd!X_6cP_e@>qEOo?2Av4huF6a2A=vP6-~28lCr-O
z1auV*Te#rC296H%XMZ@C4}R3>`tE^o?~$;L-PqXoReVRq>%#?~1!|IlA9)%LBqHbK
z$8*FR>6<w`w$w-Mzgi&O;8)l??k!i}Rp(GKLL2|+T+Y<aAh+B<UtZPcXr3@}(PzqZ
z7Mb;fJDLV5b_+hQOg-e4w6)1+vcHeWo_x-fn7r;)GSCGj;cOqeqzKl)^_f-xa+P-j
z6$^U?3kKD)uG2&*w>=@{4i=%!K8;-AH^&x34m-cM+?T9aZpir^n&mN;&>ao<1nF#7
zoSm=BShKP2R+3vSv{OC!_?F!3_o_Kt;(6YMdOQw!W{@I?Cvsj)duPhuNHUVS1`f4J
zdVk>jvGy5{$^4KGe91wn#V|X$l@8x=haRZ}iXbF9m!P>BP@=)NkfY<U2WhHb?nls1
z5bPZuJEOIjqh~Ywd|cM}8TvU<bhL+sD9X3Rn9;I)$t0nV&qmsg=*x^h|0O52mTE6I
zv!-8rJ4||5kjs*>SYmJX#&t;4Ty3)42a!-}ME~J|O^*!l0K+t)Hm0=tC?rdw(CTs<
zl6l^6QdXr5ZGDz5p!L8ufz)z;n7Z6(52n`Wd6dO6QuLk+_ZjOe6R}f*FoO2)tEOx}
zio>)Rd+01>D51L!BmEkRuU}i6_g~MvMgoqm-<_ZI@G@V7{<^qPYL=#A%X`U=wG{c1
zweempYH+ca7-Zf=D^lb-ygKOznOD+BEK;HUd=DwyCUvxJd*wfGJ)J@9>beQ9`@U&h
zO6erA{2Y41FP{D;bp9Jo(#vOon~|{@$%VucOy52X86X?sca6Sfe9^XTDZeXP8{jvK
z7D}=w?Ii|1ygzs)VsGYq7giQ+RlXA*O=4y^-qGQfZSVq5U4aDLkH619J}SAnyNp7%
zz_H`#>)Sq*D=Y_!<~|_V=HQW#((WX~S8!ZW-)QitX|z?>rg6rJ<08Sd6+5#RLEG)~
zb@^lmwpKda#7k$mRrFo5PkVj=`-T>aj*k#%yZh;719yia2dcdcW03S$T%8Xe%*1B+
z0n@y!jzJ?*tB=qKHhdnrx5jTsJVI;^)Ab1Dyl|;Nx^8>Uy+d1P%I~1K9une~t2of(
zBitS?f^OKOyZh)@CSgp@Y^CxeS(k-y^RTpjg1P=j-j{3`SCzyIw%RGZdR-6!CWk#S
zTDrQI-8>qNn<s=+RKBm5guRc}hF*IOn(@i`ZMBd#?7Ro4LMd5?xFzbo-rjC@SX|c_
z2lJ_Igw_zG{2Kb$Fkeurw+c?==w*1(N!28k;U=&@AtO;bdvg(?Qz~=lec{uP`Qh$#
zviGLepeI@ea`l$kp<;HSKV2^S>chg4rPy!v&AtGiP;>ww1z8LSMOsoyAp94pBm>EM
z8omw>7GCRD`((J*ceLKuKf?9<TfXkIZwA8Ac~fo|g|xOTDU!~Of^T$ogNk*J%^(c_
zx$?U>E%8sDQfJ!#QN+E}83|GNpiGlM|K#WDCS{q;XXl{cH<-R9rLN~OeH^%S5w5nj
zE0c7-WX*`RO{%wj#WV_#;ru4(Tbq{5i@}P<3F@%@tlYPWyUA?%2%Q0XirwXAbw2PB
zoTC#Hdn(i&YpYTscGmL_{u-Hl$Z}eK1`Th#^BTD6?zZG}eib6=C^sc@GDIB%*hgb7
zBW33*_o&(!?U7$uirdgVO5P_J7k3WumTL9c5TD<{unv5u;jn>jqZ&DF?gFjSxmU=6
z7AD1~tY3C`ltW1_*QsL~n%)rf3kawk=2?`V4YRP(``k}BTxecV9`RI9mv+FtzA~(D
zWPH?I(Y8!=o=7Xa*m^KhCwucv@D1{h|I4j_sWicJG-H-=4%cy23H!k?1l4rzl}_5X
z%uQ$3*<e4y_G#{(e^1+e9UDq~ZEx5L8GL=xcU;jRlCVi8VZ;PxiP)sA-(@}5pLwSs
zyGX#WXmcBMth>nCN1ty0n1S2?;B&JV?7FcYM!K^Y*|kGHQ@z`EjXsrZHGhwPT~!$k
zT(*mA2;T{<$uCbvHFiFAvyBV!h-O<hc4<wIaE9`cig!+aO__4|@<?vbLgL2z;Hc))
z=rbnWf-H4DnvJ-LoXdZ5X6e-FhR4#qcL`T7q7~js99%;sVzoC2p=#H^z1Ts9=y|0Z
zD2Q1wg9s^ni#XL(<BO>ZT1>WYS&;4$6M@=vk%U*uR&eay#27kfd5LEbR{Xr8RV-<u
z^^>KNpq$HuT~`e+rIgh(%p+Kis)H73eYOQWdelU6Uf@dE%7u~c1)al^!jrzP+?POx
zjUwuzmgR%G(5u7|v)g&S+EuH`_{%Tf;{WGTvVRp_?3%k3eh4n3%afoZkQWc+p`$RS
zSu%uth%F-z$fe#;s&;0?qM;s${bFhHB2tor^-R(2o_gBw`fS9^Zi0Kpr2WyG-3thw
zfrg8%Wo9TJIwZ|^{mfM2-B+<%!}sNHS&N0)BT_!szt9?gVU;BDc!~6qqmxH!=YCpB
z_Vc#J5V+*Cwst>_Y`0GoBOsj%$eq`Fp|a4YR9t??LrvGItQ-t@=vFOg4)!zbl_ZDY
zFP2kgPBhQ6sX@=9@0;3z25~_kR*H~7tYTo(S%b(`0WF&C;C5Zt?b11>l!{Ja*Mo~M
zextZAm7P~E@JAHdA6G(e)@5H+1AIPgd`J<e%u%iAujSJhAYNvGIMTS5ry4Zcyb7Vc
z9?X;ruzWFnw-UlUqTCs<aWat@O*Q=pplRsvIjHM}yY~$=q$M8H$sPKNZh!r1GKX&|
zP=-LSf-Go?`JE~Yp7J|~`79AR2aUifLS>GH)+<L|qQTz#lx0Z~Kcrg%e|<CC<1Hrr
zN$26oTvH~afJZHRx)q@m#j<bXA$VheTdhdtx%5gBwU=BZGykkt8gyd+gVd^Gq{D*o
z{9>!*$-Lq9am)bS<53KJSNHYUzm$VTTd)=*GjtfMU6N*Y+ZVspyAw>^*Dx)0mApzW
zdIfneje~M9TMPXCFUZLQ#HELgXVe6}U!uMlcHZsvmV^AROQ!AxyNg4xRh1jcpUxE8
zc)@<>9^oYqy@S#5F}9w}d3m3_Kr9YY+ZGXJw8XpC^9>h-;X{`?*%#O@yHESh9MjgO
zu96mKvB0d*R-M~rxaL*O=U2S1B9M0rZhL3C9cvKX26qF=Qw_)8THnP9HcGTQHMAM}
zp_bZ#lBX#-m`9;{4<}!3PNGH~1|9~qC0jCh36BWyQ@)wi!HGCmk{sspA3m)PrFqY1
zLfPQzBrF&CX(lCd?R3Dq2uaQoFo1aXVk0w9tHTICXT=OzPqySu1Yqx<>`MJww>8#?
zxkgh(#^Tq^1kJEbyyv<P<mXoo?Bh*0BqGSJH4bEZTqjGE`c~pm3(GQ9%}2;OzL?wP
zOd^t{l~vh)`Ul&a!A$9ll3L;o(=shJ52b9MOo`zM@zBNM<d;v$>1PWZ4mO3lXO*Q1
z();@Ez_WF4{$ob@knRJ?lSZ>Nhz^!bALQk#&hzPQHvSogkq*^}VVm`egSjN&TyXH0
zNEGt^60GH$4`q7$Q{@u3kN5Q#vD`t@R#2MAi0MbPE~Bo~QljK^^(N+dd9qB*wngM7
zj|m@HOG(azJ%Ec9Nv2Ja&{snCI{TFrLs_AV@9vLdWnAjLl9wTb&ye3<)2%<7iQWP6
zTjXaHVoFdkfqgp({+0VlgT5;B>^Qh3N<R-#+Z~L`n1^*6Op~i<1j;j73~0Ub<>51V
zX0i13ijnnq7l#pP;SwL8_OR-+XK^~(iuDE!6)(91oikX$5K!t>KC6ZjLnn!A6=wky
zh1jPSKiegrjzV=9*Q%sHu5u2j9{zz~kLF3|CvraNKAD2q5Nyc0^Tb0X)nND5<0Uiw
z$>D0xDI^#6W;k??IL>|A95;m~ayAnIO>*&1?+D~{`2<4-<0I*UpM<>YNt$|Z%SmiD
znD~by3WZ@9>qR<hozmsL_cz07H>Hc+sI0pAWVg_2pk1@$lo#()D#2t}qFBr5d}7&U
z6h<fZM7Cbk>y6||!lI0m!-b1EpV+6Fs)Q;?;_VNrdZUtBGwcnb9LsqoB!j+Z4~5?u
zjMvVu-gtoHtNk_o{FCct;j<DiH64p^(qWGP)koNSuRA<AS+UWt0q-qlN*C(i9+zf*
z5oX=dP}3^C8^qbUP6CD!wNDzJ!x~;qVBwRb+><i>Rv`lfG@5q^<W{ln(*3|x#E_e&
zVTVf6nChgZWpDhrAN-Ta`u*6nj&JIdG~W7pDLJ((KfD^9!YG{GZ>)$BNYu$cEl?xm
z#M-Q3jXAs@hClo%9G{9dG~=hGG#<LzV*NLoh8+)WqRfuzg<ds)MH_qA*yj;qDQ0vZ
zTd3R8YE?un>A}xJsRrrzm>q{TH_3E&=+`DKxV>~Q(gqvEtRJLr|5fRgu`IvI%V$nn
zQN%y2@Kc*aQV_TecAhZCSOcqFv~PdKySng~6l`uP=lwi8J8LwrnZPI6UV0!vuyyR`
zr#d6H9M6G~PIy@1CZv<^IbZtzylCgWYeAlFz0nepMc-oOLEGgd7<V{}uAk0vv!#Qd
zHq<VaC1RU%?HW7`)nFoBZ6um~=49zrHXe;dZbe8iTo(pB4cg55@SV{%O|wViYv`MH
z4G_iFLxx^Bx$e)P#>(j!tkSpDe}K<69A7=}CvbmlCaDzY1Z_4<x8AJN*tWGVagfMD
zJ=7i_$(aV3IlLGq?)?f-R`u|am(X`VQ&uhzB|>hq4TOEvrP6vGB@-x_?~5<>WKe>z
zoGqPHfVS-RgFye}M0`k12`9nKIEy09oDOR~o`plVrN6{-^9lM)OMR%>!(UvHvM1eL
zC&Y_3&M7O0frLC4IA5+-r{mp(wrKm$^+josM-;DW42LIEnJl&$i`bu<B|h8nln%ab
z<eX$Mvl3BNL@vqyGJiH}Y0rAm9{Z&NW`(Y(0J_i`dKmYQ24Gl|jXmU`U|+N9qGtj>
z$5f)B`R69d@On~zZ^ZAIepuj*!w0m*+p)EnyS@DI8n<6~A(@7rJ=Sg=MJ#jGPo4EA
zLAD+~$VQS>u_%W(PjlmVA9~GplZ1gT!I^%^ZeNP3-_4-6k}Fz%-%Q<k$?av=!S1=^
zf3h|#+p?igBDv+djWiJ&&vlVO$Tg#fUO7~)2JGGGq#go`4iU<bV*?Bf7I2=!CZsNj
z<|Rjq>QGu?!{~cw{x=eFMsm%p#pkx!tbimu>6P}cLdEY~ET=)It7)qbsUid(CwVCT
zzU@mFZ3ClLG6}VnBR!6ZpgmNo?zkOUq|Ufg0dNUer-$_4iqNX}<86<Ai8L+Ga|&$=
zq`{^>E$RAqnJf9VrNMPPPJ+;;zvON{@DW-jUw>$BsaeDggIfeAKyaO2yF%aXj7LE~
z5hEpKZCHtDo;|@FVZZ6WOFd!C(5c*vt1--Qw{i~JRGAsu{)w$rw#aVCH+9orAhPMZ
z<wM#OpTy2A;NhU~^6+q%#lA69`1Xg_(b{J)F-#?RVSdn{FF}M$)<RRy?{Wrr1^XG&
zvE^=zQyf%#O~_BmD73S_jjiZfB7sTwby4(Uvx(io;fGe^>1~0TnBR+tHdn>HPnBiq
zh_Cz9YcpIdD%RQQIM9l%So`;C0>2j-`?{<?lI&N37wq^&n505e(y+dBa;L}qN#oTR
zw-|m#sQj#<>SgxbCxVL-0UKi^ZD$y<Gg%0_pgOYb$=?d$UzzdUSE8ti=YkYzN_Fre
zajzyy@?CHQ?cWZRJB`S>Reg`gG~E&EYWA@_eAcQ~^*mHcJHo@`dDLFr|Har_hQ-w_
z>)IiZKtgZ=1PuwU!J%=t;O@bKHg1gucemggTpMWIgS$iH4vo8$)9?EBT4$egzV)vC
zdtMjKm~+e;v+B9)DaHpE$x{<Fz5~8ESqe+!s5h8+?O0bAe<jGF0s9Opf#qdw_79(r
z+pdVhVlna4!IK%SE?OzN>V}v)$Kv0W4n5NtNUlz+z24)6ogMtKW76|D+drnYS61Vl
zCW+<oxw~3iZvPWl`UG*_Ylk1KPZNa|j*HtbuHl6%xLR9oj(6I8Q;c>DLC#`4r($D^
z!ID;tDT?9?6Zc%3{3?Mfn@OUghVR(avlCc{9Lo|f46p90ElEtk`Z)b@gnwSJ6w1||
z{oX-3+$SOUgXiKE3p~YK3U+Khr}@s7>uH3Z&BtH%`EROylBh4|gV3-5j(B6m1j<>{
zb<!EXl7gsC^7rL^#z`(Jf4iw8&qk5w6_<-XN9k2IyN>7_=azb@gMJHQ!rXO+_#H-v
zox;{aePf+pBSY+J%+ktuZ+^dp^GQ#8zGlk`S#HI9TcKC6ef}z+@YH@Bbsg4?PAMsn
z^gi_1t{#zk?j(kN=ceD&Sn-j*9L`pBn}X;a3qBp)ENwEk>6dlk*>N{ZX7FixdV4+C
z-26IO-h4jW$rA6=vgY&sklu`)r_O^x8*LQJn+Gn}H1N?i_==AQGY;58<IOT+RT()5
zUqhmPIa)K2OCX^IzB=WyF21bIdkPc3a5yTi!w~E5f#i2<8b{&Gq;Z(JyZ4>!4FFJS
zF^t71dqyt-3=^SOinF@I2hF5U34ccssf2cXfed27%*%8UX8~5xOq-njbe^r=c!|vW
zRjIOfdchA~IZHjNCCodjO0g_*tahiK>sIx0;YSN4!mJ-hLkP&BuflWOcY5G#MPz6r
za^`|&=UCXwB23)Z>DZcf82EF$POkR`*9w=Fm(@p5V`l4fv+Z<<r8LVYNnm%^{R1;k
zx7SUf0u6ntQRm%x&FnqRJ8mbN`vsNbn6FECTHJkZ=+e_;M7%eR^%PyyJN;pF86T;3
z$nYkbDclDphbvrk%KFC=90NtV=3S0?3BYW9xV$&7BeAyGm(3MSn%Q+VoTqyZGs^6(
zV?r`XST(v(NX5qg5(7N0MkenRSei@D?C{2ppMvRjc!^6IZi0|EcCk%I{t_=W4#1XA
zX%0TW&A$^qe<&4ICCoh+fkJpI=nqz`hV0y-7lT~_npA(YN2kybxb@bks}H^*PhB{o
z4=nKMbuRE#`BbJ4vB{2ZEBy&8C@KxS^vtvblj%Wb!^J+kmvDIV9&q^Wvi;e#k_?DP
z)^x7x!(YP)n=%s1eUb2etjOu*<q(K%zLz1tvBpwR1Ig<*UUhUD+vlciqU*$~p+`me
zA;7wQ!8q56=T~AaZ55eSYMq91ZPxbRC=MaocksO3fvfF;FBm=?(j>f@Z0RAP`Jftq
z35Y?7B_A!Fj{)hKdVcvnUi&guY|%yI5`kI-Pd^0qldRE(_nI_B-`LutKwfB3Y1R{9
zGy9I7lkwcPrH;!IId{PbiCS?lBy=D$5P5g4RiRaP0(WkE89!aQv$9vS^^!ZZgT5WZ
zAW96aGj?hs>9Fv_RHL2E_N;xw?s@_tbL*$k(s>sLt=qhI9#$%D6;bG|#1a->I8<bj
zsirO$*5$>|?0{j;ukv@<#WLoP_}8*a%lj!jRh9H32Mc^8zBOj%b`5)5zK;m@?`Q*u
zwrLJzO@o3&i}&^!==7mYjCk6r38}ABes^9Vx>CRgNkFzy>2D52^%@d!A>Jj4>N2Wd
z`7f7Bb?ARrcX-W0H|cMae2el7T6Rd#Cj^ai9<o-}GQF-w!V2JCO0%5uAxKTIrTFk8
zm*esXH(GUhu>$#II%2rWbE+*q29Ms{T*tpVK8T|3j+B^RG>8fIpy-%(=HOxzn;1Ib
z;f;?8*n?>9A<1xyqxU4;p+k5Qz-~SrZt5t5g?JWd6X1e$;U&wC5YL~j_s7E=cQ<}B
zQC6qCf`6D?C-w&n&9sAouSNUbR?H0coITY2XzFi$+$cCgA8R?Ae1?l&vT5}iMOu!J
zes42bXNi5%zlYQAtly*Qgqfx-_!Q0Ih2P;MYg|{GA2ZZ*J7C}vnVNuY#|~59shRu@
zujku@)j^lwJWEXQm6$7}1%V`~pH%k3KO^Hi0d=h2L|3CAjSlH$EhjY#ph)AbX+m@@
z!NgQR;)aFPHhShQ^I+lRVcAg#*t%m@A8u2fn@bqw=aI<}%@o4JroFwCxWG1;!63Uc
z=3-VBWqUC@tKBI05zr2wUNI49EH-N*PPE3C9<A$K5$(dQfY&8D2=yL1m)v?xsa(F!
zK8Q>guJK&gy+PGz@9s>&oz%8R(!6HY_40-1`v;x@7`dKdm8GMOb9(Lt%^T@BRsgG4
zx4+Y6Gk}RYYJr@A+sIj<nLwTR_o=B%;)UgJ#l7xC)`MfQJGNH}ImNbJk9fE0gHh?L
z6}@4DJ*I~j3i>JgUqd1_h(7yKLF6<%<dMQ4br!QO4k7Ac;_;oa_~C~QkB4<kR_J}7
zR?5)owlFR)8-1FteBbWk5J$JWJZ&Ez)(z!eC^>gqI4)>eEuHv0@_UDKis@<jU2r<B
z<vz!;sY_mE)_Ufo2{!X#quvZ>MDvqH?sO$ZR$C!b8*S@Rtmo$*^kML8<e=WHfI|=o
zZytLJD<Puiclew@M;({L;x+$j&_m|ir!@mNSP7W%KAY&qaxXuCf##s>)3hnjc5Wt<
zgf`ISKV&rXA(<&715F74J9aR2Y2`eW#AK03w(vmn13z;Atl9qS_J7D58VE>;TBd_@
zxK~fe7!5&<^oeR5PveQZs)?a*Iyx^_+aE>kvH7;H2l2j51y(ffvDce;o2tIXbLFsn
zt>bFAPN1Xs1cggRve9XJhR^<EEgSWpYu;<xMT@`xw*JsK2e}X2fD6s`S<Utt2I;oe
z<_wNOLK>A~zH6Suy7aP5AJLF~m$l!?r_7umJTvlM-`9BB2_WC|>Jdw~KemNGz2w+B
z@qQGq2WspOV~Qbqt(~*w8T+#xB79gKx5JL~b9Bbsw!42LTP7{gZZx60x;j*8>)H*b
zDMvaS6|hgi<In8Civ4KHtVJW}JY(b<)nCDTL-AK0;ANnhfL_nvz0|f9!J!n9hWhGe
zyrxouL|Oc^-@D#mHmx<Jn$Z$dRw^+QO=?W*Khw_1fCHKUuhqEnqfF3tq>!WX?<A*r
zBl?KfomU+{tG`8|<kJ7<ap+Fm9R$g;iWJw(;ivxKuaLA3Z)Tfia8Ax>EXeiYCcvu3
zwIM=TKH>Yb1%Bnp-uvnKFFn@KDfnF{IqSZkih9yQ-~;0B#@Z6GSojP4^u9|&kp2s<
z`L{%aQcnD5lLr&_2?J`hYkU;ko_5R~Hk6macm6zYf=R49+{~P6dATJ<ae^6-L+MR3
zI}MSW+a2dNZG04Hb6;&PGdP}*6qe(r#66*!cyOju2D1H1SZn#=Nb9uKh2YuunOSwI
z+uXHn4JFu6s_F@3-+Q>Silg3G|9b%<l$-yf9{YGzeB+C!az%<nq#=6yMN9lLTrAEv
z66n!Vh1X-3-#ZEiqW5z;{uJd7@5hGN=rmjV_SxO}H6Gv7k0ybty3hta#r}}DHab=c
z&e3lf>l=8SrI=oqzPnti_ug#S4S*T{orZ_&JHel})FS?i1DM`7HTjbm^dlrwm}?{T
zSA?^f*b4iQdn>kjlZWK%z~ufsVEf@-VlA1o*XM2<JEz}Q1B8xNB{;wdo8-gEsgz8g
zNY+dLm{n@CKbMC~>qDj9&xAMbCkXYmtwS8T&zV+mV|!IN*{`F!nd+>zt+q!0=T3t|
zN98T@7=GKeF^|vgQS^$E|GH2|E`XR%Bi+jgoL!M<5oVr#<bf@Tf>&-PZH@=Y0izCJ
zlEjIRdSrP@6Ki7NlaQCZMg8S-Z$%c7sq`|qYi&G)SKj*xqj1E)=Rp4E%!<h=9I<e!
z>T8l@*6LQ~N`rW=n?9|h%eT(Uxytq~X58S>7Yvr0UrB2FgpV{sG`e``eeQj1T)4v#
zC&cED(V||9nVMkXHLu&b{{=s^a2Hp7d`Q7C*vFqV24cQH{Y$^lIo3JB2l}pOdEO&>
zvjC_JB3=9Pm>Vje_k;z#8KCX5={Q5C2xRgpZ;?ex-feoSTK5t!)dG|2TeSr{u#}Xz
z%6K)>5q}YV7>$%P<y+i(mjiNe8Rh-$izxx<1YN6zx`<4R<E`I6pA}XNMKn#ijGsl<
zFYV&PMEfvJs*dn-SxLQ*IZo2M`~=b!=G__SbFGqxdAKRK!q|2OlDQr#*52KoWIfqA
zJur%W6!w1V<w9ycvP^BiZol`HU8dZ4Hy#|P-Zk(-lF3@=QYTA{`zg9|nF>Af{c$UP
zMum|;pz%AiH9qD=%ng8P#vhegnnJr#>7ixn%GUYSQY|C9rH;LEo3+Ts`=e&EXwE><
z9gUu9v+5pA-%TrU21)PR-Q>s}VGV^L=lj*wr()NrVxuKqZc&smGHwv~lfhboe~m*a
zcl>ceO&LRH=p#>P2rr`@qze@OQ0w}wqg=m9C_P(tDE&(V{_}BxN2_Q*4E5q9(PJ&|
z1kgU5f<Ku3j(SDI`}W7fNVM*D6)=NRtbV!O+kcUcbA9E<h(>zNY`OH&+bk_8-o<^2
z#){L^xu@Th)5Gyu7BIgl1^eGhDa6ev5!MZh8BU{j;xDS4(<H7sCj77XA`kbmB;^Ta
zrl?1=AnZnx^<GogF@33gNkMACF10Ge{4NLePN9<;CvjvMC+_=NS$~#w+SdV;=9{T^
z^n;(>6~k-X*#Vt9)#IV(9XSJ`dkf@IaqTg$R`&5a5x+jKC9l#iRx;kizSjzm)Jzb=
zeXGx%O=3!BNPbjJ6JgN0Id@^ucIB=0+mrSl)k!Sm8@Wfh*!(c2S-76=HR?K;AIDn{
zK8?9ugP6zm*84OZB-`Mk*^$NT7}O1-_=zN=SsR3AxARe?6X5uTi959_Fa=O3A%?Cb
zWAG@yKlj+Nj&sr~MJ%6)+;1f)g;j2M9wzR67;@S1=}86L@73Vb@rRYnONW6xy8Fo1
zYt14Wjo66$1ZlU9A)yE_Jn)E`KAKIHGbDbWF-0Rp@!plpqh6=Z%{zsS%6gIUXZLsO
zaAZj|ylAVONQV(jVW$=`#^DY;Z9ht4fBLP%_Y7s-cn7+EqFV+hZqsmW{a}dLzPhc-
zE&6CIyH6rSa{<Br;}U~>a2DUAE|#lKG~XziUz8bF+x`mFwd3iSS$DH9wJs75`fZfl
z{Vl5<rtcSsrWgjSrA18BFbe9JVt3?gS})f7u|(_ywbmC0@#gIgVy^Y`b3xYKUR`%9
z;xV{p4mV3YMtt>RYo&_=TI-eU;k9T})%VrT;p60+YeQbXh^bL&nk1&79a0l4j*G}F
zzBud9qzT@1#<6zD2xF??)JSYOeh$8IU>+{I@=`rPJ$A(f2ANy0mt1<h7vlC}r{O{C
zUr@WZK&M5D_)lZ%e%<Z`?)s%0rxb$puXa(sr=H^;WTSmJUaCvjWc=!$*F&w*G|kLj
zC#0eBi#v<ktw3ng;uya*av?wj#9exJ*u}Q}a$y$jCXI3}+T+#Y!qZ-k+4jnbOBPNr
z=8lb<o^|uxhE(9@hl!XFOrlD}aoG3Fe@n=(&5M7G*sG4^DhJEYeV>Je#R_b^BLos$
zO#qsXpq<mg0ed7swvQ8hcIQ75g84K>CrSCXs+nx&dU+{?jOHNaUCzg-f7UUUo)+ty
zYj%bmd#ccYCSx#R{X28>1I0`fcQx<j=lAyBA2TX0RqTJ7FoOH6gFgQ<$hA2|y~qF;
zMR%`OGb;I1{UEXm?>P{>3npuD2765wgFm)N;qmCCragYTbp&gA7d}ps=m_~%NdfC$
z*4(=cG4I+RiTDAzmLC$hef6&=I5h2gHyjJooqYq%8Bn^he2e`1^uK;`sA?#`q{@9o
zp<SR3J?+&ISK_9q`sr0^Ef*;D{xZCXnnBU{Oz`YmlsoTQLPsN2h?~$K5Nc}=y5I0G
zhorZ}=%Uzf0<Gq)ZCZqCb5VkI%8{JbhW*$O<v%fO8FvycmKX|?U}<XhuTsoiztVd2
zvFzSyf`o_$;6Y|R3Wq~gOqyA^d+emMaa>lYl`un^5Xf^8QTMKQZr<-NG7Y0zc$~m_
z_-ppZvDv=!N?T?g)i3Jla_q9!{h#W3W)c0FWq(gRDcHuaMabJ(nXGimbK_ub`yK><
zd>fHP#0b(g*r1DLsYc!hqm!I|;yVPa^OorHT*I>5ge)g(D87hDyDRL{4V1^*UK`@b
z=yu8->SCm7pT#Jj7X+UaiSfft=zmk%L~`4i$>x(0M>mXNGF}An;a82U#)B!B1*$0}
z=G`kZJ#4-eEh*&#?9tj7bw-x&_QF@}z0Sh|p9L=EEk2l4avRz+t}zIkyvsd1Y^`)R
zqD`MRNUDimlj`-VS58j{$Xh4gEYa&MFDtZYE<sPe+dBlwTh8#(G<{)SyDc?aG9NQ1
z6_2I;x4d|V*ry-PO3gbZ+wh((8+F-pf1A_$halbhGK8ptuC;w0f#v;1fi?1b>g5?Z
z8A_3wPG3VVX!>R0lTg=AT@Q1XUEJ7?%`%=9w~|gh$s2UiR|iM(_V)ISRB3Bpi=7Q!
z!PdSt`!(R6X%N^sd4+qP)#fRPtEJ{$_)HR}S3r+r`mVneova<+V<EHG7`T8>@WW5}
zpix?63Bctj?-@X@XNh#^_VF<-z{3cj-#$=f!Ri#X4FIG4+ZDaQB0zn)J_}?ac`!xg
zYQ7?bsHfo{n&3ccyz$eQrzqvghEzb-7|6Sdnl~Fb(eAMivw{V=6+e#V!JnL+a$0Hl
z5k3IiV953*cwMVD)EF_1J6$SIcMM_JxwLlQr-JCPI9Ho@rEgXRE^@#Z>X@BuYVxdg
zTwkdXrv}D9K>6Qq?+c*&tdc)k_*np=JXoz$-(}!<(iTPi+c@z9B?1*8a#D#>OP<tu
zPiXXB&n_S|eu|Jm!MXTYGh@(PyGkBz=$S8q6Fb?-hDAgA1Q?BT<YZXa&9amy?KdW3
zv-aS!l9Q~y^IovA+DP6HhozAJO;zbH;ltPX1p<+t|9a49vcIT5SX`%Os(i`i=DM*0
z@_jD#cr3_Rmi^vrZe8C;7d474G{d#Exl$pD%Yw${%^WZ5)@4f;CXCtEO*l%?TxN9d
zL!64w+eysc;u=Z({_@U2)Wq#X*F*lj=bQSI)LFCBt%b3sV(}Md=bC^eu5aOuAY!OZ
zai*h-HSbR7f9TWtfkN|=v+1=r_Enxae>&es{z@~kPtMZ=h~BO_uR;q!Bs-6ep*wG!
zM}YK#^~Ca-iW1IV4kH5sizo=(pak|{)r%q=14S*Z?F(h%^Cu6VCpiG<T*yf8!%OaW
zE0ZOZ7bJXXGbm=4#AfMY+(WzYYIiAz@qccn^Zkm1{*JN}W$~KW4&E-=nq|^IkYpwD
zlsAxa{*>3<7EwgfUtdc#n&KZv_SLlYM@uL56!j6hY(Dzd5sV8q{<@fwKJQUD`7+eO
z{tdr%NR0<WB~6jGrId&JgH=o}3Kk9dDLmgxdK;(HW>@hr57!&m0E4YRF-0_q(nB>e
zT)_{A!E-#s&*&lqORnDc0gG)O?aUXEf3wLj|M79wp@}(eo(^ZU{_5~-oH`_IEqTuQ
zy`DG0%a*3&d5Qk%N0m0e=qJ!eD=$X#`B_Cwy;sN}S#y8m?s8*l>(cWPhlXf?$J@bL
zw9Xo{o1%9rpS~=hEM47(A5W{v5`S`XlGk@!xA+SoZkVe-ae08ZR^_Z4&dp99es^_k
zqP#F`yq=vL8J`o}YjeEWJnq8~<XR|D7W|vn`5(wGF(Qh&YBa!W&9n3V%D0?%d3alv
z;w?2#o|%F{JGZ68^67Zh0nGX=ioABl<;UA2_aMqq<WA1qBA<_;!xk16=g$5ce9)6P
znorz*P8q~N#^si$_pV1>YS_%|ap@CziI&8J^Cf6Xe1oC^7#V_LxKjfn-!!k89@A`>
z_g9>sX^-LZkG<n}m%&bzNCAi4gv9W?<*0s>WuLt&kBwg=$C4NVn=5gSIa2>#R{ZCc
zUjG-5pM!rZ;oM71-%Yl<wP{qET@G7VIX}F<UrxUs&GmmRpQ*o4P%ih$Z5$oIo0=Ma
zg?+9f5lo1zmpmPu-$uB7BEY>yh%aI+IxqpZnj#-q8zwz*meKi!c?;LxRD55g^9y?K
zX}Rv^jWwI1?lhQ&?=-!tY1Y$_EVZ&=;^ua_M>fF_J&S@<xE;Xdx-npqc^1N1T-@JK
zGXM3OenonD8usm>&N3h*vqEH}C1qw3mcax2`R22Gwrne-b|JXv0GAqL(C*7ZTOFE-
zP?oMEm?+NC{Vh09$Nqqm-yd0?R56~_bAWg!J}Cv$cR=Es6tO_1&GNQE{4)O4-;@de
zE46?Jcftwd#S-IQo5dja?a!6YNrj+;Enn-1wJiin>enCMSWn<#=Xwbxa=ghp`R6VF
z+l6rxV(rD3&l$my2&4Q_!g7CG@v%Du=F2+s543Q(KX&jpjE<N;5-gV%$2q;0sT?Wu
z<?l3+fBk01MEHx*xzN~qMfS*nAIPg=5+}b+rf>XkcW<WQG1p6HVj?e=uLLv9nsnQ;
zS0qU6Zvf$cEx<RA1mP-YjW0vn{Lk7C-UO^_|LfDvqESTyyml>!Ej_^9nj)I6DvD&e
zP43|5zfZ{izi9A}r@E#o`OzhFv@_%aS*O)m=64hT2(R&Z6P&tUT_&C3>d+VR$+hMo
zzf@<T^FrQfA+5<vgP8tZCFFnqv&Ps62cFe;`aDy803Yj&dn4R0jHSaZqu06Z5~hlU
z4ujoEz*R5DJm&kz?Xr&3(;|9yj$4U1XLkXNd`acLvHqZDHT=IT*$e&ao)J%>XbQ@D
zxEGAA#|qE!emUHM)Pbxq%Mm5zBdi~)1k{5W7X?~O>hQ0kNYlrr5`qrd3CE(AN9Ep+
zXx7%!5V-__-IHQUjHR4Q`EJ#@?j^Mz6dSdtG(AY-;Z6;H=}|eEoOGp-0ONYVd!h)C
zx%qF7FyoO;^7lpE9-lD0b3f?pu4DIn5h2mmOvm9L;0QTeuNj&w{ArP@xu&bBE^r^V
zV0veXvE{^IP6nCcOckM$(7F=MZTXnJ@NDe?EaWWW4`YMbAK|i*)Ie&QEJuO?Ru;kD
z_L3b;b;66CE_E89y8J4{fHPrN)}c$nc?SozS^*0+cZK6=<?>d?&KdVr{D@ZrCRW&a
z!NHOg^!M>h#5?3Gg!wjNd_!^$%n^G%!o1VrhefA$Swf_sBE^?u2LV;8*(=y4I3MgL
z6rDIY-F3t?HL>SW5MnBsY_nA7Z}Kl*Pax>!P2xR02-q1?`157{A_#}7m(`c3a+2ue
zQh-Qb&_@TSZcZW57roGz>3H28%t(L~6krx65Ue_COzpCsrB9##al5!|8>DG1PIuBr
zcG{gg6V4~~Fxz4_n6~0tMRaX^OrL}9X;((fP;dN$Z_qq?C0n$*3L=D_&&Mk<lO=W9
zs*coUrdEDLrDDTsEprTRm~ngTpO!Uq*Zt0*yJ*H;csBn5btX)%+@T7=#qW|-BFy6}
zOZDcDbVjpnax>{|c5(O!%mqz7?#wD;NVRci4Min$Z4YsXUOa;U^tbhkRqXp$k!KFn
zk<@psN%mjCZ<MZPDzs{qNQM=CR_Y*@HK!sTF)Lis#SqtFjbZFzoLUT^f0CHb`4fee
zz=BI^{WXjr7CNT8R~eX%lQ83>tn<(eHP<PMoV6*(*aj{P>UgW`JI|(yU{t6aHU|mV
zswN+)fI!8A#_#U=X%tAE!{Q-z=WBPgl*Iybs$5%&!JzL)vsQ#t!JJodu!-80>l{yh
zcU+<w!`+~BD&F$%%+_72NF#v~KJDG4t*>V}OC>L^o3s|bV1O{6M>|^1?=Be`-St<}
zU?B?y4VG`060^yG35m`pLi&qVi<$t2!85L6n97Hv&)>5K=W&l}OzO~!jnXNtoYjX)
z_34pGk40Mr>zLA;2g$6oQ6X1*QIWqguZ?3z4%jM3uhrz_E4iRF-v>P9pS5S&rLPu3
z#}E?zZ4}RMJkRSryQFt~xP!;hv#uMibM$D8D-NC(PzF@d0it?P1@#h6=zW!fPH;<x
zKY;U+$jZ!Alz%FRn5Z4f@kMb_i-FF75olpy|K#PRZ+#c`Uf?K}O}DCbP8EAJvRWt1
zVinosq54=P950;qllpXulF922)_ULbwEZj%0=~+s#1j#JICCV>Nn7ACzOYNJ<w~;k
zP#;@@G#NpD)I0Z)*1L74m%&o+pc=vRixcWiNjdSzW)G^}MM?5R6Xc}qGWXd+JdYhO
z%-sk+*k{$7&LC>>^^1U}5Acx^uP^|NNX7Nc+v}YlK3`(#E1n8LDOOlP<C+;UDteHU
zmTS6TTw_s>%SvNpGQPwqcRtT*g0kywlW$STByM;<TN6wam7;VXF+|8dCaRSISBqqO
zTCHyjn%HODzrB%^ks2@5l_SsN3;^jRydp!&fXXkLr%ENXi_TudzYJnmA$|Tl`ItJ1
zLcHG!OnW+DsLnZM=hc8(HU8>*0*rG@<9YSFYL_H@C=xFpu5ymS$B!7K6)(P5w8qD1
z?6pMLCVLtN+U&_X4U!QQ8{NtYg}e*+YM=WnZ0+)N6vHA^Q0DRKC9|`3Q9LqyLb;2S
z`0x1SWZePMa#6bLFh>fkuaWQlA2nk|%t|oiaSbW1e?cgZ;&$#%yiF$gCPIBhUZtz;
z8O*!~(NW8^KE)u94N<x1Fl2udg9&%AAb?HihSiBbZa&EM?F{+7zeE@^8Sc9BDj@KU
z_UW23D;bl`avxE8og)w{d#~Cy8m~G$hS-0QE<_6U_RjapNx#;es83cgwOC{&QKtU>
z;4;oegYLlhFzyfC+x~?NEQI)N?<VH&ktNKBy2h-1`lGtC@<-Zh+avc^oKKnK<pmW+
zaB(mC&5#?-_sY@qpZa`R7cgU><+@hJ?N_0NLJRTXQNGp~3_%@Lfoaq2E&woZ4Ev`)
z{*E8ioJD#>?&LSN`T$vJ38SFg@l$>|6bDvESSt^5;TuF%S?m=c5^2=rIbb6kbFQ0%
zAdW#3C=t`j^m5R3Aa{ycaMR`mOVh+5*(2`rt%BUqC~8!$dBbh0!HO~-MJoey+mkf+
zLlLsu(ExlD>3QZh)P4T9DOIE`wT+ciEv;<;iG28AnRR2RH?m~LY^nUE8gntJ+(h&r
zrjKE&g35JD50g!rK6L(o2uah<deF4h2av3mCp|n+8{)B@j^kCH@`lu*I*>02#Lk)7
z7YSOM4Ux-3(4X@3;iqpCj`b-C-8y-X?7$8SMp-$XKkj20V7io4u9FC3pKTe|4>LIV
zoH%pRu1#uN@?@D>$eb_%u9L3x$t2YYzAMNoPO`n&lVp+gZb_QnaDPEo*A01d5hE;k
zD^PJmXHr|f;+P6c_xC1=KT&|rdr;^mS2>SkbS-tu&JkToj*KHOi~4taM-?-+DM8Wi
zVk%8*$3B&dhOt=iBw!p#w!-ScgW>xKV=GxurrxQh>U{UrIjA~tG;WS3QOy9gU7O4j
zH3#LNUH~~!1m19ew`(?3XG1c&!XxXj1B=~kqae^b0e$lml>zhCf})E`)mhkUL5kX`
z1LE1d2r@zZ2eelaV7rKdlAmfs;qL~viY|>TiQdMjf#g~h?~|V182&oHSg#vOuBhbX
ze7<LoP-xPpR&bN+c3_-!N^vwU<4fm*n{$NS1i>nHwtS<si3#+Iao#+24<LdaB`)^x
zFCMD6^Vdw2dFBg{3KJ;n)IcS&Lx!#A2hyS$Q1!c+`|&fYW=(B5Wa#Y!#tSVv=0wZO
z?&aCfm8MQbpxVing0&u+Lo(L!U-z}G*JM+%Ju1s3pA%38P0EPRwP#fqha6j^N)Rn*
zg2CRxN<6?lAS#g*<!#Ii6Z?z{CmKsNEM^WoV&zdj1(rLyO~7dOE6!CjeZozl@@ENn
zvRF9a0MCG-@>2|;@*D0#k+uE#T2eK`m87YHM(Nh`#zDV9yA`CFlpemcL8qNJVJ<qr
zi-rQyp@jX)>a8>3a)VsiUP=*ts;BATh%ALJhBOVs1$t?ysJubwv1(<l(yp3Pa-|Fm
zN9z*ZuN$#=mX?e!<~&~ckpyVXJFu8*i=-c_Ra^S?%YLcm?soh`d%u6nsnRy67}S68
z6b)EqucBl5Y!U9MRA7?yrB#(K6&DvFKUH$~7i)z>an4T62D|~I($rk4wkQHJ4d&`T
zs4|$9(WG$jUedP!cVyqk&n7m>Jc@y*=`A~Oq9_ECP3qoE0Win!66YuyUxKg~oHZ$E
z+FcgFzX%!V5@fD+ruBy<_1;a4In8jOUzHdQ7efmz${|IbWt-L0`!!Nm+xv<i(l(NV
zk`xD<Y}U=1qO?>>kpbXKA#~^aui&E?4KEk#%)<AE{wzsDl9_zLB5FBb`L7Ih7_Ca0
zN)pwgERcK&cf~^N#iklGwH)~mxjfZ_uVzD90Ex24dHE*WXA2}Nl@NhiL_%<HwikI|
z6gV08axoA#H$4<i-XLbmr%Lda*)Wp<X5Cs6dR99ML#A5iRx32J*jAJL3Huq=6kwS3
zYg{WlQ_+@jd%kYSYS%|DlRVpatSHmY<q**0wl{x{lTMANG*d}?|IT9^AV-ck2`*SO
zv{_(vlS#6z%zBDY(MroDIX2CLhOygTNtE!*nznc;CMnb6lVQM)*pNm#rd$k3dVe{`
zyrTS>fVTkT-!V0g7X&wbg);({+e(o6oOg69{&5(US@>UU;cFptS8Y#31%83S2DN!7
z#;>tvwR%9#>Z3cQ3%cbiu)#rvfGEX!ub|wDOo2=HAYiSvN~;RoaZ*QBerQuVX9!jU
znjzY>&?~pFCa!~I(D3v)<zA8uxOZ6*CiSs#rXx;GEUBhzsNZz#*`>hj^e;n2Ur$Jb
z^5u#8FQci@(|_{acU@q&?^|E<9#JS)$rZOM4*Ht`)jp(297NjChkJ6bz;lDB>sXjy
zTbX+)y~1~?3XFaqu$HZr8;hkHN@E%Nh10#u2ed1Hd0;>0#f+BMe06c=IeV#TqpBvI
zflXXo(L~_R+l5-x=|vWkh1UY!H^pFbboF*vNaUy98er*+=<HxNbKL-}8boh<p>^~z
z&$!B+ik0kI&-;kZ1MH$PE6+qd>&oF#!w|NAzFPh&W?olZu(~Z9nZTV`<xBXslRW-V
zzZBwAUT1ir56Nvy{{!=q@~J^kK2vqi+|G)mcRJt&8_w19c<t>RiV1d|WvfPP%X@km
zKGQpa4s{7>ip+QfkE*jE&$v&t&;*p*9Sdja#Qokm3!b_>-C%VY$LX|O4U|kysRPC7
z1HG?PwuhL$buyeM?F_;~`U27yG>Jo$<x1rN?(c8F%2ijX_3pLh1%P)6<R`lYoL{Gm
zJhrWjX_KoJX^ivlmG6G9h|#!U6hS|;FdDH__A1!e3docnGkq~Qu)>YMj&<74fA?pt
z&e2)s+vq;U&z+Rd(UB~Yz?qV&RZBFH;jn1EsV`1N7Vq2Gik_sd8edagT9RzpABA6)
zbmW5^MV%K;Xa?En;F~GEEW??004r!z(N=TRLV&TP0J>twB#(qreeea#U6h1e9ptzS
zDOVeX2)E_7tA1>WRz<&99=%g-`T6@0#rb`pBWqh$dQ`~*!c5&LB8iSU`mL68j`trm
z#zY7Zo8(O)o`+L^*az*Pu}SUAJ<Vx6xx|yB`d0&1SqQu}1kglU?)UHC>W*tzFR3FL
zS5O&mU%<yC_1zfu8iP%j^8^>EkNLM3#<<zR4kZ+_m9ul{VW0L&3KW1ssnkSJ(ub8F
z*KnJ25-#iBmfxH<EMbnDj;o{ZmK-hN9Mw&R`lmO|@;jRDzzE%<!WHm|2B8*jPPc%6
zH(*h+jEQ3`Y$kc^WkXn!{f&<JAI&Vasv*b-c<S6MjqZ9{1?Zfor>U-I9H5`3pqXmA
z8sKb~Kwt8zUE2f}SZv_(t5fR>OP7mPQdcAAN8kfR7<Pi`Vu#{@RzMrr{F^2*|M(}2
zWs!~9)7)9kw6RH*$kmu<f{#>l%$zi)k&*`-kpsS5ALq5a=(3W7n$`kT^btOIAg~Kk
zWaIe>7nc=4RQ?=T6^D3fZfI1YkPPRRrIt}ad`3i(t*;je-b`H=vhI<A6gkqci_cxd
zr5pRW(b8%aZfa@oFBYi>zQvGrsed(MIJ_dOn0!5tP;YWZgLa3@!tGH-fZj;XjtAcC
zBB(kIyx=*-I44v`(ffHz<MC7s<je}>y_Xgk$9dxvNo?JhLjLg*tqr<a*`A;jF7Dm)
z*5z9>_AsUAU6<Rm#0<S$OPk%XrVe?YNnb&WyYi7>sT^xp8MJuUBe<WSDKb)&2_X}%
zz$hiQSqIpgI(r)*ZzRkA)DegVMObBx@`!D(HTliK&lzboCt>7%Z{OdCLIwj1Y`HjO
zNUUU&#rxYFCnVzIpWU6USmy&+X+jD71d8V*+U9;cTy7F>jhK1jyaS9B=@rfSrg_>|
zQqv6N)iT~|C=HbI|Ef?*i6HSSQh<#Ys#@qA$<VQUKW()vm*kCDy@>bJ;)jY>fBB`Z
z?<VOJYGjZI>%qYJv20#~!<nNHvH^a9w>429WkGr$x2vqI(lV?3)$B)Q2b)M9<k3{r
zLl=i!EnHyyO_|#HBEfvrFVWh*n%B-!)y`@4_|E1>&fPmE!{!1Eb6B*8YENhuf1vov
zbz~RwI!bUiI~GAS%yAy=O4^ysyPURpT$<alw%XShBbYX&p#@Gd_%l#e0V0<AGdynv
zdd`@L!TsIZ+5KHE0mnY0raEzFXstHCN&G{_<3s4oy5OJB9UEFNAkawFKS-xBA>W@k
zs0JIZMg8Ihc&*9$7aLh7JgkGgav02P<q0_O;+`^&uGDm`saf<H=lWqqNplCva!i$^
zb2N{nKdA+koMSE2YhW7<pFc`!avYSaeK{2UlT_uYYZY_;Scmbtd<=;<*~k&e;eKiS
z<Sz93(G^&zY^ieS{$Q3ySc*o0cBDjtr8HH-mv-g)D*vUFXLpstmzsM-vI~S&9wU=0
zQL!xketH1{xT!9XjzZ?LG0}|cUR4i*QAgez9*KHx*;G){Jxhz>)y4FSAwyd*(<#@v
z`*7a*ZD0AV&SKwnu>}VO=<?TCOO?T6M#9YksU>O#QT3glfQ*LhPDBd9paw6yxZwD!
zmcttI!_}mmAqaD~BCg>`MJiJ%Vw2IPaF#l{slsINW9Zt05Wh8-D5E#N5Q(+F_wcJn
zJ(UACxf8BcgZu9wh#<XxOI_ba0ANw@P=%k=|8fDfu{!43^eBL|x%b<JA(<iXbm^$#
zY|_zB1~Sk$tY!?+#wuOecEhx)bdExi#<PT)!|4Z|N?UcZpNLujFf`!Ykf&6y^|5*$
z@1Ps1F;t@_NN?QKQ5s~S-n`$7+{OZ(nL0QGG2sa&t}>cE3<Is`1;~t9)yJX`gO~h{
znSlC9$9zunBPTGL&7ABk9$pXh=A1TRWXb+6Ew&cjI>w!*>i0VCd}!7+j?!Z%9})zU
zplPW3WRTEBXU*Q_iQ}RBASp6y#9$z_Wwf>nrm}EWkk@;L3#0pKo`re9pd-SwhlHUs
zLu*<$<1QzvXJ_5HfJoO&mW|1*c!(NAClbb-lw{|KhtCZR+?sepTBH>~nDHf5M13)U
zaR$AAp3k}$Io8`(QBPgsBF0zWOBi43;j7Fq*M6<9PWG7h=K}fP5YZ(Ai0=dPeiz0C
z3vAW3nuL=ztnDK+wM7V{Gs~`t!fQ3Af&nuE^tOKpsvgVus)VAU^mt9|m7Eqj^FN+v
zzQZF=;<>cd-m6R+@>deqsD-VsISpX+ALyalA`4_Pwkffm^XQ)-S(;vAa<7J9MMY(?
zIk&Skne=!>$UBK`wbjf+Gc<a7y2)E)r()==D>*+4pm@cziJ7Fopm2ATU^B$iP7$Dz
z!Uj+ZEe<LERgH*AxxZ8fFJ(Ea*_RZY2{Al5l{lN0C@~>ELc3D}YK(L5OgDjqY|>(&
z8E(nOm#a^cDJP29Zc+i?Y&T*MY{u#D1D!lE`;d+NBMf}Vv$=)QuLgdyV2WB`(ard7
zIKL4ksiH&kis{xdLNvkV#imY?8Eg%BAXL-$iIddPb;UeyV0X8DNFi4fkTd-lT$t{P
zB16g;e~`{+)*6-X5t6ho+xv;6-zUb0$|$p+U|Rk1gt;;Qr{utu?AB9F-&jU|gzIIY
zBiiQdt~G0JT>6M`w0Dq}H$zq`VJUzL1Zt3{_~=e@Fp<J-4zo~->uE3=yE@xruoxQ)
zbR15)@`&<=e78_;tkM^3FCSH;s!X3jzcH7>O2VHwp-&Au>ZpGz%3M>gyk)YyGWQ^<
zET3wD&78+3Y<PEAF{s*>T5?BDM6gahJu#-c1O4OXmrB(zj_rBd=J^f2n`@tDi3?Mg
zP=D)7_0YF|E{z(N;CVlwX_BL&u5u>wVYS*2<8mXAPvCRnM}JlUMDkML;mwy*p%_t>
z{{3CHbMg;Y(rO(-N>@Vlu;40aEc@3A-~^SaOqk8t<h3QQSVD4bsb`cGKlI!2{9T^2
zlwn>{4!*;AyHtOkzO+3W3$1jK38=&eF^B)T0&zXF#Kni&J)T9AcVZ>#jj1fboLEdo
zI|v|}4m;viaBXe+SUFd$y-4u9tq^WWcz$q|f?RCrh{WnrvEX@ENJ$|(w+chsHT{&b
z6dqQcD9K-8{|*^QG|D!Uch!)Ucsh*YykIYcZVJqSbi6(-FH2N;l5z&Vaw#S~%|Qj$
z`b?8!q}67f^1k2dJ`>hpGy&F~H&oX2U)%O%qmCJS>F2aU(|@5}&6spSv@Sj;h&HM(
zyL`<dW#i~78FqL#rwy!@IjWJpV_9a&91NzdNNKe(Y)xXM0?iG;O8U@`%vvl+D`q(_
z5g#5gq%H{6Sc;O&O1;&!)qd97y&#dp-6%D#ktf}@C|S(t96%hL^aYes4-Fvl9BjS8
z65wOq+i`Vi5O*}g%iuHj2QG+cof`9<mS?zoM<3y{!fSp_e>qbs)@Wq~F~EmF1>RSW
zNAP5%jGT{Bf{lk%W0VXRXL&RKcwTtvTN%$|%2<KMgUM8BHUo*434`=c$!!JqyvFv^
z29xpT;{@4Ht2*%U%j<#G0=28NISREH(nq3U0=4h7yiTp?`M6><KFpZ*<vE*^T_f{6
z8i`4p))5cVx6jt*H+m`?69e!fFdZ5MEW@B!2ZFJ3(9c;j!O#%*88c9-q@{BQQ=+70
zbpQj5JX-ajwzVKeg`}U}u=<0^cCGxBCU4JM)@pZPsYMZ)X&n)(c4aXn?Qr@!nIkGD
z+5qEwoa+8eS2@zaBqOB(z*Gd(7o>U%|AN@s7Dze>0jz!g)7mD6--i41wK5G8gODOh
z_?34jsfD{~3wrUaw-SWx(PAo}|Nf$6Ja(94390I>K#B=7RaK{qL1<0yg{M!Oz4E(N
zoWvk{@W79NWEx@;+9yw<>k?od3Wr>^|I5cF<beC%<OwC`Ex%v)yw<xlt3?sTdo*ke
zf{pv4t?fH5@%AWQ=Kw57?yspt<N?0&1Gk)Uk|aB?4u#)h&F}SnY!G5<{1gi8hZj>O
z54NsQF@P;ms)aCr^?nhieT2>Pf?KW#^5P52a8?@M=*$W9KbJf~p0U}o`U4HvATI$5
zb$m*dG#ouj8<){bW!$b@?N%gyQlF_W1v;f(Vnuj<?4^3#v)NkTv0P^Ed$b-<fGB-Q
z3k&S!9ZM5q*%p>Y6J?klsx;5lO-9`Swi1=8bMLrJCxPtIo$iWkQDDs-yita()HL5M
z|6rHv{MvoJ@kOhf{`z~XXfq@l_2g+{{*!8!UbkwjUA?}akE$1sk{*#6M#C;k&&{D0
zr56KLxmxsvx8`B5A<>Gyv-@Vr<@ErD_N105%DWt~^G&Wg9Cza;rRWEV6Eeo+XC`{D
zb~U|F&s;Rnw%LOD0n2xic3Ab5WiU+S!<Ad{lX+HuBCL#rE?RX_lx#4S${AoY##CAZ
z)<)H{M0D#fCAOXf%o`QABs`XvU6}`uX)lsGyV#hMK_yi7dYmU&w0SrX+=chLefng>
zOi$=kddsRYqZnu3vCcw%zRR4M%J>pA$-tn+i&daQd=;$qi!?5Ml>XA}?Dtju*3nlG
zWDw$uu9sEiI5Mb{(Ju<GUF*%)pL5i$md$vVi^AKKO^1;^1xkT2HDFHLD??qkVm+y+
z-S#iWPm7RLNe{{yV~{rfW)3>1*+L<(o4r65H2!M$#i54Bc?FJBUR&x{bN@P?1FamO
z9gr&3Qpo~1{d_=D2tl1v<T_L$JqO@-yJ&4&`iRm7Rjo#mfV@cQ434V20<Qy1YiF(@
zdbxsk-qv@ibDUReZy&x0Zu`WlD#R*L;&?!U<hodLNuOYVm<>Eg&9<eBT4{Y#fr`M`
zIwld0TS(9wS}N2Jjus@j)=(nZ3F@LbmykVQhFr(%>zfDME9mo&Q{#YZ!wf)W_D~#5
znY0%RG6Ns~!YkUAWb9tDJOp)NptX)(FaZ%%GA?=>Jqk7+E8lFRF;isWU!zh9bX@s=
zabYXoD8u+M?~elPLVZd4P6brnhpcH-<$Wig&QUEaB)QU|gz;n^j^CRHg4b(Aw1C1|
zN5YXi<azmgjGLW)*KV{nF&NvbD=-n#BSvP=`JB;N(deW)3^}Dj6Vq?W-olD)QCdUh
z$o%eJu^O#EsZ)c~AxfD^O33<}uiMp?GqfzeD$ExqRzoz~c1ma9Ge`{$v|Q*67Ud)?
zf%Q*X-qi+%L7q?kFI|Iy(VQ`I$EmqNM!me4<5U@{Qq>_@lHL<!@p>rW=VaA_Q04=^
zDO(|nx4f7G>){Srv^j?vn8leCw6s^}18D}AlFjV>RBAb3sT>VuQ&6a~frUT25@<V#
zLnftIAmR6Gu_uZGE5WXsk^D(s73*XwjEL5n!CL8bjW++Sdtu2E+@{G2FTrnCDtlco
zeFrIAKxZ3i?Gk>C`WpRvH3)Ove{cfVgk29CxAD`(kWuEwBsXzAXvV(L2`dAqp;hm^
zBa$z3%dz#O-z2F+s?STt&Sg?db`P0rOJu3B_fU-?xVXT?$DCxnG(DAq+?|()lgy)9
zk+^K-7?nQUoQWR1Q{y{)I$yjqPCyRvN*duAGbXry=9h|QIjz2ysKAf;8qszIle(D`
z32h7#%e0YhPQIqHNO+nLAB{Fs$~TYLDor>#Dn<wq)pst9`ph`PvbzB6D>_dyTeNK5
zj-(V+a_^mVRVRwzv0BmsmYhvR>oOPkVD4PyIL{oXSU%-?=d!4KS>{18le%B6j^oYb
z9o3N<2UTdcCeE}B{xUG*TZBmR>x=q3JeD<LLglYMinnfyQoCKh;yJAoq>pZ5czZSw
z2Ol)!S~~@g0Lg<zBwF{i28(!Lc!Q1C-s(*bQ{tOD!^Tv`oEFC5>jl8pS($nXRjI=i
zB}aimi8M620K(>()yFq`!k{j3Hm$m7;S@evSEawr9thu4`gN+#qD%^c*7X5n-=7{U
zW@jPYs|5g;nGg_$-k=6(8eVr3%NX>nbkql##|N|Y`no{5F16m?tv8gHU0Z<SQ|9b*
zqRY$zuWe5>wj;SRU}xGM@duVXzd$dTtWkHR8=Xw0vknqD23mgzz}w9`;BR`*+_@-!
z4nI6oVZ3$I@Bc$%T9pR|4Tnw|KcMuPA6SlP)X<j-w76Ow*@Cx97W{#82}Z^*ZK)ti
zIN+v^IdDP9WM`6{_H|?g@t`4D$Y{zk6xKxXDhJ4a<n<Zk|5+@n8?N$1AdXW18pqrI
zJW?v!I^)kubD%G68GmDEy%vN{msk(3BH=Xj0JK*my$y5;GM!26`LmT8zrFN{oB?-Z
zJ*slWjV8zOp(^+tsfgR&$!gOpaqRz_cd|u%`AcwpC1`Zv|8WDy3H$Nn{bwUH!+5sU
zW0-Lht<Offv8{Z&2du2V3}G!UW}bgm0PK36F#E>8@{VUP3kNicSQ($MUppD5xeqq^
zN4j2S*Nls%YS`%3RAjc%$o`9&_-_!%KZip66UX!Y{Mt7b&DBF^MhmCF<)e0C3t!>5
zuM;Yl7zR1vAlzjnUc#W5NzNExpDq%QLQzOiY?6o@_CKJX{sD*BBb8nB%5?OxeqQck
z+-VqaIM*bp_B8IFZv7=HFsRkrV>0k}z1#mnY+hjXuind@35zL^E__QD=F+`f=&$D8
zK2DqLTuFqTqeqk`PzC=-Cho5__c6VMwJ0fk{i$uz6+(J2?^`Ohm@8Lf4ZW^IV}96)
ztoc6bL~?YbEG4Fj`F3yqjNt4)vvz;|-3^=KjE_VBTT9yvH(+gh5G$8L?l%5h+0}<P
zYZgvNz$2UQH}4GDaU%n|CA++^_b*#pRMmT4{@?rKzpeNmJlsE?T$Ck}kvFz;ZvbLF
zSbb;wG@F-OwNWlDQAw}lb6NK+P^O2?{jJ<9``g;Yy8u;TC2qS%nyi1%0{`>T|3z9w
zxgd!IqxKTrD;p7Lrc6vIaeWFC(Nm`?=st&cZw%sFi>s)_uU3@PI`3uUpyYN-01b)$
zr!zE=5E^8rU0IXJ^Nu1!!fnGm%W)ki%pZ)zFcQ$B-|^a0yq87`at-|-;ZSf(AOWQ5
zO>D=N2ysWaQ87E;;r;!+9y!Xgu%sM<EdxpQ-;u)q?N&M=^=@Vh>$FTSsGy-HPE8LQ
zKX58;uB~Ms@OaK23Azd8f+S4RAaUI1+Js4-qM{SbCeAGXv)KiC=m@t`r%*xeHI)(L
zY>51j7OhEQO#LWnc}?+7O}dP~%{<z`!4Q$YT_iF(iIdlpmyyzB84G_FGRdbb8VfsM
zdAZH(uuWaw&gDhcbGFWx!e<u_S0Q2eWBM0q^<Q+{|6b5PuY7&^@-%jhqo%Q^FjZ0m
zf=-H;AY82qDiUO8UqkONHFLx4VFk??Z<j~dezXnQq3u72w~}{>guR~}#Ojs;qgAw)
zk2t%!gl!f+hh(aS?=~_&&PE>J2t1PlJaJ75{(tx7KY#lln>3^Y%l`+!^jnpEp4$UP
zGzB`*wv{DvAzyw*;P*Mo%BTWAd|x;xL&8@gdQcaEi^ZSq$LWW8KJXp*2LEJa<kj57
zO46(hs|4UFfO2L1Vi<1SQTcd^P;Wfu%VKF%<Bqj=Z0#D!v&#?}5zzoQ_N6SM=Tw6o
zMV^bn`!C?eqziQr_T86%0jKw`<B)wSKGOd!a{QAMWzCI<@~YJQgIbc=1zjFrep7Pl
zY_ZMN)DPPW8Y_>jM9c|9nI-W!rJMq%1BjTkKQ{;}$Cw%s9Yq<8k<C8m?b#7o?d(r*
z_Q(79_1w%#YO&#%FX;_g+bwWT@Jj-1_uEfj)KDm*vx7cQp1E~5IWEtfaC6|7q@C!T
zv^nn@H?o#dy!@N{`punuT+e)fy{LxvPv2^F&}?j8Q>Th@R%Ft}q@UVrVMbBu97Vf8
zb|)Hz0ZmvGRWL@bDJF;G;9!r<b}lH+){$*%M%UZ{(ZBnzsT7NCX6=i&y*hATe3zz`
zXWwce&K*v;@%s5LV_iL^2TGLI1lEe(ca31U1DTCn-aN$0M^DKNZt=Jb4s^7fjO4L9
ztx69UcaV#WwE3>#a8jo4uIq5(R6F2UI6(qCaz)-+0XgpK|K%(Cctt=gBZ5^(?^Ff)
zlcPUJenWWd^Z%TL|HsND$zH-Fi#b{8gV{i9;|R9joaV{H3iGN7!Z+1sHkYcslC5G>
z4SDhlV{9&7NsgIJ#U7qoB#isVt1~#j?GL}!@7%uPU_2z<+jqX<y90Mu-d6`s!FvIW
zAM)Ct+Ta$-r1N)~s0xlD=RA1(wQAu@l|FlesX@=vL5#<hZD;YRc`^ca3tGn|7xE?2
zHwNBMeSfttSjO>d2d)wG+b@W?KD{@4@Ia2r;Jd$b&0t2)@;Zi_mAYUEKBN*6W@e&i
zovdEuH}9_$E=_KJZ#HOucCde*+0h&(;yw63-;)0xeBoDq3ZTFT@{B~J?~z(rtitE-
zC}i*u*7?ue!7WO(3Kb~I*-(fXtx}~DZ?z#7XAz`lCsNar=Q2oyy5Z?#Y-&#VJf4_!
z7jQr7=?upWFl*ny4>97FyIK8ut5x!oTdg`IEP?rFCi1Py=F6|HVca2q4TX?f&?4c-
zwm8jb&az9+Z0sG02?;fhg`5bi9lJk@&rLu#<m1`$&LK6t_6sk$JS2j<AucmTb-&g|
zHCx#=-YJoewKmOMp`kY|Ka?H}tGNVdbnV|3-xzHCFT%bvs>v;A_t^C)L{UVgS%L@%
z2+~_rkQR!F)JX4;fb<?z1Ox;mR0%~uM0yW3L8KWVlu$z;k=_yl1VRbra_+j{y1%}2
ze&t8j%9}lV_UxJGdES{`MgOokiNUE)AN4(IC!~qKwC(Noi7#fZkuqAQbF!xwY^N3K
zf1#QC`V(aM5l_UAq9ye=rLKgE&V#iUhe@}f#i%96yRd)jDx_lh>HN_0?7IN(4M(Hp
zri$zvQUKXw-W}E|zH?N$YqvdM=VEqZ>ZweM!>dCUu;}m52vruA-R8<s*8Ad$Y+Ong
z=e-(_ve4UqRZ;3z_eBC4^7#QCK=CIkY`332;(zqfWIF0wGUB1;D^sdzvb5jaaIFgs
zjxx#uIZ>1H-1rElWe*}NtFJweLflY^I{i5E37hP>yZKL3-$qV8R%$L}hR~V<ye$@2
zPn>NN_=7H#t3Rw=MZ2vo6*6@}J0tVoY*=Y5leb+;>()!>uUZ_s2D}hH3#joQ3mhQu
zM<*^nh9L`UG5h<~l@4#+FUtpR3x8{Cr%MiaE}V*!40-?zIBeJ5e=#mse0J9qeAc~3
zq_(ZU&SPGqQ%cR%oGj3>oIu{Ai*5$_{;dSK0d?OdAGOwwT@O$Ld`g}j%?e^bP|u0M
zC{u@!xXyY8rrUhHY5M)+fD6Nqb>zP98u11=(22av99Y<nqb}oxi^4`5N?G{ZNB+y_
zPoI4E<o5seksSH+Syl5Fn8&J0Boc3{LAHU^;HskX*{@l67g(%}+G=dPfS${Prkw_4
zsZ`%;@Qd8J6e2K1=Y`LmVgUpKl72NxA4X&a5uYF;#xQ$m2$Fm>n{p1J&~Ktamk(Lb
zn}<W@^sn<6gJ!}Am-FfuraF-sjr2x@4{dUEh;*>ALD!}6ai~6JRQlLyreu`J`$z+L
za*A|SV_fq5o2uqhKx^N#fqe-?U#aH9cwK2SdMooC5OAoGyno2wj6lh6u9uH5X9ceb
zvlPKuhktctKKPgA{$q-@(06~)#<Cw^l_8Ow<Cu>aZ;KHt>3i%5CI5RU*I?|Psgxr1
zUM)GjflBmK5aJ~F>AaD|Jsk)o$umA`<J=B?c^z59J^<eA56PQLCi+<;QJ{l;b{w41
zTy)J4AGBV0sm_7Cy?7PFNp-k<f=xb}+I0oxFsRAP`^yBiyu8?|#?LlpNXhXbq98Q?
zZ~g(>JZ^VuwnZLtRnih~{>`|}MuI`Ezb_C4Dj#lol3jTMog4W9)v{%W`xlK^vE?k&
zeVB!sjH1f3pA<MV_U8T&J_3Wz{@X5}eD{;>O23T~+D7d_Zys3X<m$~vhtyF8NFWh{
z9^=pOEPS3qN0=^<ppTR1WD^hXqO;a^E3#F(W&SdZ3PGW3=Y~a<Tz*d&K_&ZE6`A3S
znQvOn06VUpW$4x!y3vHhla=K)I2$RY+6P`rXtK&c(;c*{I>u+rEy;Fne2Yc)_Gywu
z3lv3Z7knesc<GyOLGg!RH=B+iTpvzyWwa;Xy-M6R?;>$EDa@DFG#i$mIQveis4~_$
z;i(Zh$oBxs*`He6dZQ&pjMs{o!A-I0U2o&hGANQ0y_j&uB7VHd3P$fWGLvbWLTeb2
zHxBLuXjpCyr!;%IXt;_f%(Tr<bFjKb%K#<9e|A0ZE!&cc7%lw*rfPUj^jslrso*1L
z=8T>q{?NbdiCKA7${xwqevC`f>WK_>FP)=b0@0UDJs<T}^*%j&?|(Z<|JVL9Dpww4
zu_qsxjv)vmrHfBhBX6;(M${BRjo0s)i0Y%30%i9EnQ6}<b`m*t!z%Z_Wtxmss6n-a
zf_Q>G0_jNJgV*EZOJAb?M3}ojW$!UXeEk$@a1Wrsux)KrFKwJToL?Cwk%)4|i3VBc
zegl)up+c%UvWLIzbNLK5FT=}9)Ti3Fe8=Q9%e{%{IEiQ`lO`1(lT@s6HhuTKP1>lr
zmyhWe_fOmxc|JdG@%xIC@&=ZdHY&!sF?%NswiT`&EgcgRU7>!sl5HuADlr1){nF$^
zH#or`7s*fp@_5k?laEiRq>;(Xxw2*-*apGL@$Zf;AFWK3?MbDe`px~pwC3jJ@4dsO
zDd~Zx0d;-GZik1*Sjw^#O{ZaYLLYgtSEfL=m1I@bvIk^KRIp@cqc4=6<8{GU?M(P>
zhYr`3(-Yi6?5Y~>-SMgC(<*JIA6_F1baE4#cJcHyPaRnj;-kx;p|`vb)pa$%5}alP
ze9&lEsQviKd`}bKv#yO2&eilWrAk@|z58tMQ^f7`6mtDsW~S5pk|<_-3kU7oq1C{c
zN?XG&i2XCk4+5A%qFO|3d^NBhS7AHR<Q6>~VArx>;=|K4ho>VwbrhBO{AE-ZClC6Q
z=QoPnoJwnaSt<~5HmWb}1;S>RRxZ^>3lsbbGmcg$0dIM&CtB+5XX?f~4sa{0jy?u#
zI3N_xh_E0o?S;$ps?)w&D?mXzhRAdL+<_{sf$hdr6m`k$f>HUhT#9_mxqo*qKc3r9
zHdHU%s4;lFkE2HMjBD<dcR8pvvZOEEo4M$N7WNxg>g~o!7KL9`yol09C~jA6Z%Hu9
z`^29X8N#<d0IUnB4suh<9@BXX>2lXrQTe{aK>v<ruZ(k;W=3Py4Sd-ZzqP$(jts(n
zYwhcCF4vG23n7Iz(9|MsiJlhS>+I*qv^aJ_mJ4~Ida+}5nBBEN^8|#Brh*@+U;WYR
zH+zvw$EUFQ$8K3jhyk4?@Vr^+C7Zd{Jcea`5mc(U*)@O`fQv?MoB3#n4!lz+x-_Ok
z<t_`*Go4&(4B4F%d8|K>HHNCHr)EYVQ_>&Ty<hlrHO?uvih7DSAZX>YBH}ei*1>y~
z=3%cC`lJ1(L7np1Hcm>$<CTd0&nl-ZI^g?yNc=()N2V1vQ$7u}*CVa}tVQU`C>X}|
z=d}><f@|H%;XI3|Z2m5`7A?+~(8P7&tYJTT|NS5*A~j@3n(XLz)c^A0g~riWhQGu_
z)nDkCW>(v%cvMa3(85so)epH++p$*`01JW|m8P;pR2-91>soH(>p_n@Y~f@p4X$s+
z%c@x~f3t~oJWM&-c00@V<I|j)`4M7XQA)a7g)LGXib%EF8pv;UpF~_$ECw?B<t+??
z&s~`TFQoCE*7zH<GA!zP_($J@TT`Mm6C$~xAkSEf!KE|@7V&lUy21BL%G4tbrXJaN
z#%%M(tH_Vf!dLp5ZyOH%$EDYQ_bU)!R%4({tPjaHkxt0c-}~093K>4`^Gx(y99~&K
z=8hC=K;15$EPBWD@EX8a?nkZKpdfg$*fhxqu=w=wJxtN-2M1r!VQU%Jgz#i<MFh5{
z1ixCv{!qxo?;|sF6jQgm%j-XRoClOlDbaB&gQrHx%(fJyi(7M9%9AO-^*6ht@VyM&
zx|*vg_6?$Brf0tSa7}(t8H!`Mq3~_t1pR2&AxpQ2Hz+4Gf&@BhEvu?{oyZoU^uX}%
zuiL{et}AH%PP%U8u>7Y5G~>x6lP+G}2RiuHYao<I0Uo#+q1^7(Z@h%q=?mo(3psh-
z>z6j{G6fn<6FAju=0-%v{!SFfR!t&40%^w$Q$UmZ!$!^0Z`tdbR;R0w$A7$Cd0v)v
zv4Sn~w2jF}WRf~>z(NsH7JRl`bCDF0b;HyuJkH3>&z?Tz+iY(_Yq1LA=Lz<^8P4~D
zRbnt5RBT*?9*tsG00zEdAh~9aD7B?!_kXBodMF(w^1W1r>7LbzASy8CN<={*EA#vQ
zW2UhNPl}hPUk49%_rgf7kYV9#OjrQKjA~?p*#0N!IZkqf2qH(l!Ym6)LeRyc&pUe@
zpk37bugAJ<0>ko@V7+vQMSuSb+-_Jgwklf4VE0*s!gP@mt$HS<Nl-&7hxK0*zs$Jv
z3R`W1XakwXhfBHavc*H$<=&jU?1x`7fN?uHafL@ay%2g-=9e!9xDFQ1v|!Y5|Fkuz
zzRH)3vCsH21b5H<q)1fY3t}Bfc5TYaH#7`5gNvqEQ3J_q_;=@46doR;b>a29nJ}mC
z9&<6i`RCTkLa-N~YtlsSTCH)0Y}DJwad#&@P&j`S#_Jz!IyFU&cif}K1y`h~*Uh7t
z_++o6Wo6o83D9b*734Pq#{0OAPK?B54BxS6ocAH#qQoZtw%Yn2x&rMkuqapIh?L|>
zCp!@b=cp^O8`c5?!yyxjKTTgQbTsm46?qs7PhBLM2KO4iO@z##d)S_QiWgpSt%MCF
zn;I*6CZZ$qABODh3gJ<&``wyp@zu_A>>TN0w|&0wT{~Qj;vAHMSeM(4)w<OQ1w#k4
zls$fNusMOYGe^JMY^!A*hL?=j%g0dFqo-2fn=!b06!#a=vhrc&V#;nQG9(jJiYt8Z
zMzEWw5wkOEi7ng8D*yJX<n!Wg*S0J8q2ZGX949Zr4rDJh)3p(eLPR9lY>zP<Mv2Zc
zs4jUjf(P?P%<S|}6qk(#nith(Dw`aEg2E1SOZ@H-b?%7m>5_a2msd-^#Is{lvMOLK
z!6&0GAI?^ht3PNTaX;sp0gOrc2o|p~)lhdSqwsCfL6mF%HWdQxt0E58Z`>!I9N;ii
zIKanY?(l&%#z_IRRz`1b0$$3;ob^a(z4VUT<MDruE<fof-fbaAzZ=MW@{TE?nAm{b
z{HA{ozH~O+eNpt09}`k2m}zU!*N2mM?8!hdaY!jsYpbN2oA;?xnFn_JB=i9kaSbqM
zdLb*xe2RXO*im4ZIbKnvYR(n|+ZZb>el6=Y$ql#!^nav-pAr>4ibW%B)^^f^H{6%d
zO5nf0=6Egb=1Y%lhDBhY-G-&t>sQ`&zTD{Ot{i_aytp6gvfhYVtqcP$q4}Xq1}B5(
zdtC7P@_g;#3K>u$Gu5*B%&t9(q6?|Q|6pWh=*NKsVet!8vCL>_pm0|S<8u%aE$vt^
zho$On^YnI~#oj~B=uh=DEv-)Z{#9W*zj^iCRZD!fJ-EK|QJ){6n;b+<n@5h431G|k
z>iyPcZqoOdBt_6>)J3(@Qo-PzL-*vSA4&Nr-8jP4aD$Q|!%KI3k$k_qN={`b%-->5
ztnlu7(Z!J6N1aF8yV={plMY4c>emX&-xq3~v}X1_yps<ERK_V(2v4bt+mCO+<&Q>t
z0V2evg(+cFyQz*fmb8EgZtV~BH-6q37~1tLCtTK42&#c~u0A@50}bwnWjGY~+PHu$
zeikT<DuVufYoFYHq4ss6BC3rgE1zRqQl8rCNoLZx%~c+y7GP1qcz%6K92(g^e~_O}
z=z)%x5Aggky@l;kVqRpnczh+)yBKNiTsodvM$+UDk}5`)u?F;!mU!n4S#GXV65?Pw
z@>7d<7zWaliYlQ`5Cx3eN5YJk-fRzr^va)4x2G_9{VT25l}lbb34ywr4;{{25K3|m
z6E1TOOX)ruHrgX5RiJMG`URM<?lK58ukW8KS#sEJptL;c(VT|Lbc}%_Cf&7j$(Se}
z?XiXP1uXdy+^PJvG%bk7q?J(04QZh|<E6@cKT+1mAEvU-`mwnk_pDwC0nhsbUYhWX
zZ@k(<%Q_Gy6e>3kzrJi9v2dg97bJNENq#|$cw=A>46QT(3f;R8NS)0pIT99hb=%MV
zUq3hb|Eb7NBi9;unYK&XEDZ83WC#wMquWV0y}*25IyAH8h1X@Zt)3g~#y;3=_Mgm9
z#tKfOacmiORzzu3$V2d`KWdZ`G`&G%pnPm%Cyy2rG&1OvSSF9QMiJh8^VGi{#wV5v
zCpjIMG`($I>PT6d%gz(|-NdVbeJOm(C5#86p18ucRdh*EA4ySQpjBC~Qj77aZ@H<^
zfjOv49C2L<ypXLI4v@@bQ|g|?$FN_nsdPI-wrPtV?slkdg3OR&xBpsC+4EJb%d4+5
zGs)Q*J6zhymCB14&JoU=FFVkcUHyH>fX8~}gdeu4kB5zZIRIWNU4u9;rc<@Wp(JB3
zs1W4fB7y=k50xQ!DSd^H=1ae#cnlZH<$k4rp-5Eo!JfCkg;N<3hYI@(C*&X);o9<m
z2taC;^9Rqkh&7B?$z>B_y!v{B9kZnq<x~3Xu6v82Kc-yrhK4#D=U{TFXFeAh=+oF+
zE5d6grBNDhOb^@L{mS=<H;96*&OlnU|LPJzxe8sRU9uMXRr$z*z)C&NBOjP#92}P>
z6%89PyXHaw5T-=HM^S4*u|lVTk95ihqvVKHWG@vU*D)C3CD;Akt}ZB#){A8P_^-W?
zIH`bTL?SCKW;DXnrtQS}LP9nf#5BUeIfitj56#tw7~LrF5-%jVDEg4%f^sF2iY;nI
zunHW_>N{Q~4^wRw=GclbIvh>Gks78zzg`(9KS=-pRU13TvnrNyDV`U_E<Z_%{#E#4
z?ixa&MtBsZFs%DCkzw+*a!|=GXQ3y-Pq(|Lt*8~_cA>a3XlWS<#z$=vP|M+sJdWn{
z>8i8P=09WiZYe2volA)B>m{NpB*$LX(ro6=TsUE2tT#~O;qU_QqD_WxVqqCVW<2F8
zY9)^ji(tqycA@q${t9BAIMu|jna?rk?m2R9hw8K&ZW+!ntMDm-h5|YU6CYc=ff_We
zC|c^)jzY`Y(&d6?#v`vP)&LPC?^&v<{J&Py=T9fcz72t6J<Z9YB~eBrdFgu~Pli8^
zq7SJhw9X$2QH^TAnwu|7IvBKC>Ng*)qtSg226W_T6`+)4h<V6=mhFZbjE8jBW$U&1
ztr{-kB9A?BfxE(8CLZM|6wrV&d*<fS*=Xf=CrCK{ZSS|a2gI~u6w8nl(GdEI;EavB
z2jvvB`01Kt2cQl=yUI?2(4k2}P-eXm2gI_1os_qbmCUBtRJAa6kG>pBGoH>B+I=Jf
z*Qt4#@?kn{w>})!#7Zs~L4zIOThP|1v;?7|qK@6gY1%^)b#^XQ-;}Vh)K;B{mke33
zKW7xIZ5`-c@ls1p*B5%?0o<f2$06+{2C3FSoRAMv(VvubUGLue5bJvry*0<NX^8O8
zc>K`cRH){|D{a>C1mgGB>~5X@|8U5!aX(j2UBsoeZqWP0D{|Ap76ElN-egkZrE!TA
zi)g)h7tI(#Fhl9HN#I{`w>8d?)b4h^G%pvf{A{(==UGs!@?LF*8t_>Gr))qd94`Kv
zR#tq&Sb{>ploU~@V{);>V67Y6ts)0{o8H`z_ujBO_aTAXEUvO=GiIN|IffqT24F~4
zA2<ks2XobK=!N)jl_y`u{FY?%rLy*xC7p;;Y@jCcTmaD(wv-q0L4Cey<JV)Pk@=f&
zSiH|G&9#S<tA5&O<oEm7bvsDY7lY80^t|4!((Th?^G;wRYwym|0c<C}k0d}Y8J_bU
zQka~?9OZ8wvo~^N5~J&4!>htgf<eMegTo>kYLM{s(!XcOqvszuhzp$8RjRKdlsjox
zxl4ia)pVu08tgqik$<kZA^G5%x0-@+)~$g756cpG=l9eaoQskxR;k@UcgG>2FMkF;
z^7varwt_ip43zL+3Q<X1o<-qC=e*JC<|vKcVa<S#y(eC)evdE!)`+i8%GCG<Jn~h)
ztFuRGe7*ECdd70Vhv1yDvRc>dKLhT&`Nn;E0Lh8^8mPh40g~3msa>B>gi>2%mr`4i
zA^Vu_a8g7Y%C-20m5?OKNM&OcPuJy@sAwZ)b_LqmM$JH61<sYZNXBe7EUU?(9tn{{
zhw^Ak#U6$1B%{aLUl`RtGLSn(IPZ2b!3)5f7NcxVToIkMI<*ZYD~hV5l<~po+vxuY
zcneogDmarpEh|jQg^NbYZo%6_%l-UTV&~tY0}uR&==`3t9~L73V$sVnla;#TF&txa
zH_e5yr{aJByX(2<UzSy#Teo}H>C6+Qc%=BPyG4D?@n<cu(YWi}(c0Ku7~kIjFBd>~
z(#MqzF^C519i$=jA>GfrWbd}s<HDgGh06|GW1e}h{bwd|D;68iAYtDIW2sxLF7VCD
z6h^Ep&8vV<<v#bl3rSlp3rLIIidqSNV>WDxB5Hb86yxRXx$En?ga-dpVVPE_#POzV
zg9AHyeWfWpjY0X^abn-yM^DA^rFX&p8tpAZ%?=0Y2aal`U`yMUVcSX*d6joJOY7xb
zgc=+EF2Z~lsF~#y+B=-1?YIB_fR|8uCiNeo?!MZS`l}N2$8z}|Uxm1oc7v+NI582c
zD$gT~{iK$!^?7C|Z>Gi7%3v!u(-T0*P~PE8<BB{dZtsbSVvg_TZ{s5|=?mF*_N=yE
zXIxNWd$>5bD}nasCI_Cnt(Zp2f2s5rh+zJr&tjU!Q}N6?9kv6!<YPoairWC9O-!mr
zE8o`OYYx~UIN@zaZ@2nkMX^YCxNB;klwr3<)KCZfky9=C=wr#Zu>r2ZOxq3bUV+a1
z`<+MwP`7ZI|GJ5?1`iaDWmCv%p6-}SqYd%a$T{@RLrC^+FM_v8=+F-r0_A;Hos8Y}
z?^mf?XL?blaK}~?kLP3KFNO<t!@+kw;5<DqiLsYMpc-Qa4ZS##UGB#4bv~!AWLIKm
z6Gu_E{eML_%cb)>`@eyP^9$$5-7^UaN*q--15##Hpp6lEzjB<En_WD{E6%j)yMOWJ
zy`2N!G5Tw@hY1R?Nog}XdHXGtgb$Wy`ap*<aXgzqz>MF{dI>HAa0#T9Srai`Ty5-X
z<*f!|3h-=JF6S0`!9uqBlk<}^jQxLiye58!(f4Z1rht3jSzN+*po$%B!KK#>hhBRU
zgEOW`PDQUW$D}5cO(kv2JQt9olVU(~h=3$e;*QnK-joe+IA&$!n}OnT5!{(6BAH)k
zg4?2Uu5#h6!ukmOywWPPyO%v-XC@~Bu1>3gvymb-?AZy91vrx4!mkIPT%Nl3j&wz;
z0r+d3J)U0t*O+$lo${#@S-u6&t6NYZHu9*KvD7K&yhOS<Eu_T+_lXC?j7W9X7aiBU
z8s%6(e+X9`yubq*DrLXk^bq}eX=clrdKG5(3){&p`?CRuiV*5vCJ?^6Gz!|rfjr)H
zi#kFpy!%?aL)LJuo+Kmv>kJNxAoAgb>R!~WVD}^dR`E^mAoy7xB^%$o@x8GqQe>Lh
zD_#p5yb(7hv)_9&mW;}eND$1Oe@D?a<@Ig*Ud?W%efS}ECh#a1oTYEEOM;<=7n}Nw
zUcU)&;4+V8<Sk@K0lU4IE*g<uMV0!6ti+n3rd`ROCq0r795U-auxt|J_*m*4m$`I(
zx?tt4zCqj0!5r>?r^HTnD_gE$rW|I?u1m8Qc?6l~X^s4(2aOlO0_&W*pH+?ZEOI}R
z(Zj2;7MQf3GOKpVC{7I0S=--wjC!AcU7BV)6@1UmO1^B=Z2uQxf&iNT7=r^b|3Hw^
z76GKUmxR>2wl~)>|71ozZ`0+P9P&CEj|TXWYMEn#6?~&WE{rb`t-P0?<>%GPFt5S4
zGkbB{Lv8swpABNTfYXyc#k|2D#LwQb12sbBFwSre4t`zu@qDWR<%)+%PjcK3#-~1T
zG0ewYbr)cwW6UJXQU6S)gaK0na<JK6<i>-So!iMxqwC6BeKcnq!>m<;dxf#MjBU{P
zYyZLRh{e0bL*6bCCAW)lXAm__E<YZoBNOgBR=<oc`O778c(iG!=lhD;>y5@1!-Io;
zK(h#p{vfT)=<jp?i4sWIh5r1XUI6qHCk#n1Py~6eu8nF~3u@@sWBH?r-fb65!+wtx
zLy)!a6~1%O(RTn`A3DIRwRVZ$gf+ZR1bLD1;_zs<sLYVuNgt#IdDF+Edr`t=XK!;q
zs_8;mH7-&TkjF40IRW)`+~LA0i(~36(9?S^d#lLY%RbcA#MngNHMsp`@joQRvgWnJ
z>1GrQ7of1woa<;;$ko~JH?4ah!p*UF#ZTtw_@{4~db^jETsp=P0RdOhjCqz;AydnZ
zKYNzfJHTu)&+wW6B3hRey*UYqxM~J`%M;=p!J!S~SsgVr=pkHeXjq#|O$qbs8kX-g
z^;lzr47?O_p>pdDUPn1JZZ-`@46OyW+~j8+pdAw#Re9p`cUo9a5*Bq|A5P9IfuHT2
zwC4>NEmGrgdlcNaBI0UpNo(B*+kSDdpc@Qxga_P2>9(qwe9y2bR&6ZwaOJ6RM`sL*
zKEOdLDRtrB97+<4>aOu(zm=JLGCy}+KzS48ErO?hJ||yv*Eu={dj=3V2gvlcao=){
zw2@(KOZxSQ%ofLgj~;!GPj?N=T?#?NTBHgufxd9@O)gAzmob(+&q5gAvs6;2Y6PHt
zD6-O!b8GwOkdi2rJ2>s!?#c_zwViAwN7}&7WOGx{qA1N@<|vvL9851d!V%M?(Ugh#
z#F;oKYi&f)F*=-OsQ%gqD~oGB{NgKlwBJb{Uo+=s6Ca)2eJh<oDxj}-m$~wg<#tv%
z3_Y4=k~mlu%FSIw-yLCVo2?z~a4fzga?r}&)OPlt#n;4>QK?<;OG<s)gUG1Y%tkk|
zaiIk0P`S)8w!EU?k(Gsg<as4bx!_`M)<Lr3CtmaVllG#^e$7{k?ld6okq%!HfkoA9
zGjG6B-sp34f{^xhdHbN<Ug6`$OYMV>3BVKGmC#Dt)+@+xceT}VB2Cvbd&s}D$YQ7%
zQYFVtPKcGw*_{J4)qlxCmSB9BHWOkqZwYm59!i@bV&yyE*`A_rw9??(8!I!imI+&Q
zBa+OA(3CIV(1A3}JScb{j2ef&NxBAUKUmg9=gJb;WHkeR^JskH4WxHJ9Po@+FCaN-
zWRj2CQaF1%OP{@8oUYO>xKCF9H56-*BD{Evl%nPfsPgAje&2By`OJm1P$U*td?VH;
zpSJU6xdB=mv$VcA=c%h`^=D#gyl4X}8X}DbXax_0_Bz06fHdBDVftSdz?EYHgG1_Q
zFoHbUv{Rf8J(*JZ?X+jNsDH6)F`q2Be(~JDlFxm<r+uh2(TN8}6CZ2QW?Q*wFtQ)5
z*POYTby5n)3c2v12jv5iS^rv~=!E!KLSzj*Ad-_dcLdqNw<0sCfJt&sdN84_b2ILU
z+Nb?6`u2rP^8O-O(pPbZQcAkXArT^bq)XcFm``n$j=eT_#~+r|Ia-EREMYAaNSNJm
zBl{AbytHq&n<#Bwc{h&=<Z(45hLXU&ir$W4r-)q#7Q9UBRw)xfg^%{#LWnbE4Qrq5
zLlwD7*-0Re_84EuZ4l!XeVfsHHEZJbl{HVspIgD8>ufKNdutn2!8%tH$L?Z7i@VGr
z6`KQn!Hv5czHaOS!)h?|EJIL*;K#UHfu}rh3pa?#r1eZYD>jLLq7Epnjbpw#=B2-F
zJ~~X2x_%T3?D)0ATfH7K(;y!(vv6nUXfK)u({VORxZK@*c%aJ)6f)oVrH@~Z0r!C>
z-zYWRV!Kc!<k}=36GmODma`|ZojRItr8(==2={dTa-V=58sn66C`X60Mo_!X=(MUR
z1J0fp8Yw`xb;P3SVuWnIN6Qkw)bVUkQnqW^;Z6!>u!&hCs^hsdB@EG@-jD&;4H^2C
z+f9X3&dy5^d$za?ZUNiiy9X(f4(VP!KPwzQ)lM?sXp$&|#?OdbJpdAOwmbK@lK`A{
zHHC<RZMmFI#k8ZIw;h7}Wb+<h2v_QSPsO{|_;++wxrZ4%2DWyU45gOox)Plt3%~o4
z4%<^oU0306qWMC|N{@7cK2xSsCtYaxyJ9N|ip8(@@_B=zxwo~Ir`}${-tTpJv%Wx7
zMAeX*4?=afpcNh;h570;Upq$r829gZUBAA14vD0knuqe5K<uQhsbBtA?|jGm<}`y0
z_xD@{iw2LVn#?;~`z=gH3raS_sl#_AtDFg*t@bTEH~LUSlcy03tt2%+KbH_{NQ&Z(
zV3BJimJK(`yCbB9k2zfm0&6Xd84n#o9Fw|}I0uZs8NJ?(+%K!eBX_i?rZl<VMzr|C
zb`{7uW>2}xEDBPxBiL!AxxMxcY;9#uH7|mC3|*Uw_x>UiO>BhpNNae{mxw<bv<W=e
z>vn4)7pj<v;YRvm=AW|&Sv~+lN$Z6id8ni5>@73^MB3lwDAfdfR*aK^=%HzqV9kvu
zSxMYb&(4UAw|th#$VJ&!L-+qHS^1;J-SLbq?A#g|^&yVUOu@K$9HFJTxt5u>@l>X$
zMg(Qs_roKtq-$Dv%1l&qlTu+IQmnPxzKv<wQDf=3Q<8djcULWI*IxNo*s!QwPDK4v
zM9k@yQRqsUh9T(~uDps<J7jR;BB0$*Nu&BgT%9h|mZOlvm9i3J#5qu$+vLDLGcM@+
z)Z}9*)~A{ee&ZB<KM1U~E)X4thmz}ytc0GT?sH`>1jl$wrGP9w%^~zH*-&AUTb}Rh
z_|NIU0lIf<N|HT@F%K4Y8Yi27-tFSbdE}tP#H}Fvf>*+UVC&_&d*>3QHf4QRE9f8E
zazvYKR>((F`A7ifxJ69eo-ON*io#Bg`DJZi(9nKL^RB5V_EeV2l_HcXd#NrEwEF@z
zRw1yk{ovrwnSXblpN~#z9`golabu6`Y$ztvTF7^Vjm^?FrSZdZ$i|S}m9lSD!4T1b
zQ-l2enpS(~?yW0ZFT{ZkVivUCcg{R}Kj+r$ZWe&uzsb*yMBSOOQiKRrc!Wn(7XNjp
zzEyb^f;svO8}_znZItOqiV&V!<2|BWWWh#8oBMgRUI#-@b^6#eiJs;>Ml2;u{HzRq
z*)Sf}7X=L)`j<}*0_Wap^HEaH+`<YN-mOUcm~_4P<A6+y^-638jvJ*ptqSq?nlS?~
z5PW$h{Lrl)j<SY&U>rqM8XLNP%rJ4T_#!i2&F$QL4EZQ_R90KEZv1qnTM~6(T0fb&
zkU|yK-7<RJaxZeL`Cy7H<yH=>GY=ZUCZ8v!8**Ri#z{F~Y$K%&Z|RjDQzg|@BkJo3
z=MAoAjlXB^2(&c?Rh7W4R?56x<_Dt^(ah_dh7rN^J^5RxQlU=y$OCurrtQdKwPJhO
zgEkxdbq|N}ztgo*!4s;{G1QLKnRI!8@$`%-q{NGPta8z%@r0~WDm+)CjmQesEE7Lc
zSCN*IfW5EXe@o79yHJ8(0o9GO+v-^lfy&JrXvm=dPv@HmrI4*88VldTHSlN?9B)Y7
z&*b|{wo`pI(*awSOP$V|&*pPbzl?Zd-QK@VQG9t^{QmC0{#KW;D?{~{T9(U6&HWF8
zweP$a-)n{Y0q~-6dTiA~<Tb*b7J8~;_lh8*i15sz)MS{8AxCV`KKbB9P?nU?X+uKm
z^TS^u#J^#ajMqAAz|Z%0rpx9Iq)e~=_{1)IL#k7b``()e9TN9vl`~=b4vQ@ZFW;|@
zH^i_%%xbn?n*=vmgrDvf3G9m|YeEI!DMj?xNd3N7O*`4eo8ETyokIUX9Pmn<POkDs
zA|5y8I?yMU!+rOl>1G+E;m09kg`2TFKFe15Us+U*bk1J#&+l_i$o{zMDGgtdGNiw-
zM}HlUw6R!Dn9WWU)zobhXwC#&;Cbrth^K?+5;DrSU@QB^;LA>rABz<RgqDXzlZ}b9
z4z(q0vah;4VLFQX>q`1?)+GPS?*EZ|+MOOB(yYQFWRIb4!C8p<;m$7`E8DKl)~;bc
zlkwvDY%Nrl5t|ullrGRxrSR>U{AHP)^IY?!cJDQGSj7LXf&70b9RB9AWmqgidhUnM
zs)geSZgvQ~d+q@ECb3$8bQ#eDoowVs2XYbZD)TpU58(-1nwx%VY(pckWnF!cH7$Mz
z9T&1ByUUvC)N!-&-zl2^QDd%!hCLB>z$eXG&1!D++Y_ZM5@!kJ&ZV;b788S4zf^$w
zQ(#y|DtF9HF!vUl;sashznw-dB}H<>CP%FD6MCXskLKJ$iW7K09WPF%?tF2TXSv}>
zS;3%7<$vM;PyXyiy=4;#AjfvT+g{qCj<pFTD9<gYQ&D%%H#TKO+~qaJ=EAKT%7o>8
z>&;z7#sUNERobs;@Hq4NIHsIXha>iu%mVLUx%lOuXP-Y`-Z|6mkoCGIXtkhIn(m2K
za5K37@juO=K)xc=Z?91;rzU$9rB%e|uRi^t>aGc!^xj*3+Ppx5Lcf67F|fpqc7-7Z
zI5aX%0(Et18ls}RN<BQ6Xtk!1nIfU$Y~IHP<q9~@71-7>KQXA|Xzn$CF%8@&Uvlfu
zaa<kTr)TYC&v$XsT>qh!>lqAYT3Gvm0pa$4XeN)PqCw9sS4e2w{4gEI3jXu?s~+it
z@Sr;82S8~ve*FT^;x)D>YY*@RC2X1}kLM}+UDx2z6dh1TMC1T<m&a008VCpcBg}c@
zc=ye5p(n`;-*)LZ>zm0Jx^sNbHSP|RckB29tj~~)`=9-fPyBBvVW!df&$D*|bAU0t
zL0^LOvKo2{)OpR#w;u1a^13ySzysVZVkm5iv9PjbCn!hls~}4qVA?;={&;t7F@czN
z{8C2X^r8=y;6Ph2O*d9L{U1!h?aO|c+h4h`)iclI+Isd3JQWQ}{+OVhJA8~>(k7W4
zX_Al8TQV^_ieWiZu*jM?DzU4Wt;febZZ=>m2exvZ?|31su+y$hD=p!DK)d^PUi3U?
zd(|?ETt8(&@ajKb|NG7Q#M1gSiok#($Ek!Df>$30^z|*bKCHKiv+D*Tb{0oGr27rj
z*Lnyeu%Jy_>GjRnSo}_o;%2*JeF^fC>OUAypZ00>MGCpOc}E4YQ135sZ}~rBfS=Tr
zH4PqKwcWNPF0C&S@%uKqG))}BV*GYXZVvZ2^h^J{??)U5*0m=0Ak+G{s9+VL|6a!b
z+t{yKzu&4G!SjL^ScTqS-xk><P0b8YDl2}Dz}V84#>!acSd(A<$CdK`4ff|8&j~hJ
zGYyWjpeX(~iI)UxeQcihlXlzfbBTxAazps5+Qix|lxwt+$N9AX0PX+3QEsSMzDB6w
zAwKie(V=`&ux0kB_C;s_x}{@aWkgDsIyrc6Z&v4viZv4zM-GVyQR{nW-%_a<)OC_&
z<57O%C-{Fk@&7CU<aZuQtgV&%!5z(HGoy`~ydKL!NH%VZO6XDugOmxkAaJ<?#>Ka*
ziXtI*Tu#)MjHAZn{ePpOlEk1z+A$+AiJ`7fHb(3>ADwYvah)6qw9>sJ=W>ZOl`FkP
zX+GF^cDLa_gU|nS!2LgW%l%!wz29&e;`sK&Rz|<?)oHH;dbrHfIla^wJ-FU)C6gn#
zO;Vu^vD$7RlL{@&rx~qt)!1v(jpGL0HcM{CB*_;&ESO=x5>{&4H^rZsUI{i=6$<`3
zz)&ze8UKXMOwz}|*5bg)nWER!r-damA9|kkOH9i$Rgg4Ap_*uElee_Z|HNBDBwN4;
zRe}i==J(~+^$oDF$v$&iIx_Z>0uzL8)ez5Yb|XJ~EG><M-m&Nfh^`QS7kUlw&w)#e
zN~d1zZ5~uosR6|jyyzy|>^oqAIN{A!ieDUFH;T&9d7Z*X^S9%St5f`GRP^@fxxB_Y
z);~(o<hWJ%bZGeBqkq{}B#f6pD;*>r4<Mz}0Dr*QS&~Gb`?VKr6BEXMa=6=S+0|;x
z1e#Ecv)ucyl@{D{`OTRi_VVMs`0*PbLF=CugV_RZB@`vVe|7tXpR40?>`dJg7L}cj
zu4~ghWOno?9T*hZr?s~q5bSl8R`~f9L+AUJYDD(ntdu@P#IZK$IvEb-JLB;7e)TK-
z*AlW%fw<O}{shncZ2M<bc__xCyJjqS;Lnc;V}|!w&!-yq-p+n$IGdZ;=L^^DFXfu2
zKGM1j^`^W;+MHwo-5-qNGf|kic<XpTp6_~zR6`F<k55fr)7pdd0FvvM5a>4#>|Svq
zY|pI2c&#`lyS-{sa*1|T(U?~j4ZOaDFr_Y(U!jYS>N>EmCY>B8xcFi(abf|&c(t&y
zQU0yjNFC+IS1VJKLG<by2!656yYs7{TSvGQ&(V-hYby#&UQJnh^+)Ke<1bb5IrZmi
z*Lf<IGmX0L=2n(F!VTS1u&L7#b>;kNo6i$Uod@_^)Z_~DAekxCyv5@)OPvLN$Gtg2
zD9N4Ic@uV4@LOWbH~iAZ6Ly}UR)U^nz3e?V;LRGKP}T7#=wT;0FAFY|Di#9Yj7178
zrH6H(2bt~n7jAC5F78kcNY31v7Gj!b1Fw0_A$LufwLiQH27Yn0*FLt}X=JgP4Z7j2
z(5@eIZv6IBiX6|7<<o(BZAb;)FkaYMb5Bu+c<8X-A9f+=o3qcK7(q$&Em&d<f<Jh8
zdN6-z`kEV8^LOMYl*LBDINdR2;T1NYoVQ}vXuI!uMzooYDw37(x~h9O$4^3fX(j!-
z?%@wUqv<k#LGEF^L32@Z`f?ijR@}_*=202Gj)VK1wa^EwTWYv-whwmx+~Uj7!jUCS
z@$S#~?5CqwVO}*E^^r)@NSh}3^23Ave8-@A>AlA4cV014U0u~1dBRKj`a%2CPq#gX
z2B(Dc9d(#14HWGa2UGvuDc{B716JWJZ*o$4r)ssHB;&#0vR%nO+OqGdJ(awd8wHY*
ztiRRF<)ZCYAVDBC)?#vO3Ve*p%qJ;MMo#Gg<4(%<vqqN;6QPAbPEA>Ck-bnku4rM<
zJ)zhcyOWm0hi(gXsFNzv^Gmtp6J+I@h+4YWC-109$~+Z~TMcWMDCLK3b;r}`8Y9j5
zXSq9cw9Lu1!I#WnqBNq}Moyp!nKUAHf77^PhTKx+V_!<B^mOpFbnT5jwJ`tAeqs98
z+e-n(i?R*cHwI{e$C=r)Wk<uOdcKqD&qo2<8$d=dSx3k8WD;EE$@X(bwnn(0*UVOB
zjnC56o0DS>6l>L`TNc{2stGL1A{g6-9W$>@XFE23?&f3Z&n=OlAtuX(L$N6_(cpVv
zf}8XyKI{knhxc!LzT+IusVOkn6MhizqU_wJO223gs^gx(d}zpD_Zpntl&hVVZ?WcS
zlWEQ~2{Mu7PRg-fQX-~+tE~JQ`n#xnA%FZ+DeP7=KF-t{*D2l^!__&;VLd0KzajbO
zNt*7aB*%}u=P_o^+AINMRy<dpW$Az{hCWRQ{r(!7PQmLmEI}B|h8}G~7bD&kT$#{!
z;*#Lh=Q<rNrEH-N%IL#dDRr4gn9NH6McdZe3ahMxQxX?UCrVgK^>(x)&v|{WmE+^P
z^oX|zp(>-L1mV0->TpQenmq|mUf-Gfh<~8B#NK^ziw)Ro8%GBH9h=Ct)j1=Bl@9%B
znD&?rd{0?i41c5Q=1E5}$(vI`NP#;au3!EAcj~tfI+ZRnPIO3Q1ldsJqQk<4zT(W^
z*bLc)^ix-uiUUGFkkP&3?<K#5cI$0`=U8L|ST9SA=%uRL3$&WVu2vr=PYd-GVeC0r
zgCnPYwCS5^zFA#Vdbu&Ao3Cc|R>Z3}?4?CcAs9pPn5&$C2_o2az5UIX%+EKEtX76L
z`4wg-oeBBVGxkxpvM_!duHVdp?|xz=u#|_H?J@=Y&bCo*2HEP?><nEy`m8H%m%*nN
z0#<{ofy=mzE<E3*6&m&0D9vvq^fh9(f-IJ|!Ae|So!p~(L7QKEueHYqhLY3W%%4de
z7a9g&R+k)X<-#s1MY!(__dUYT`}7Zs`?4!@O=Pc0$9wCe5ndrFdPd-mzUNbOPC9m9
zmx2dH$^vU56z6Dv-%KoF6Iw55*Vis~4pf@iu<1Z`gr&y~u3Q8-?#U)Vvxkz$W!gCn
zT!PG1C9L&{^v;?*WojodKs)3bVAnl}keKc~V8GlJ9DnId87*bWg$lFl5-7J4K>P|M
zeC~*B#;UiyI!dDU^VI&94dPNwR@^_AV4hIDDYZ`-J4%UMfW6_#Yj!UcLt7F>+PmP-
z^CIFw4&Ztbc{8u2=)Ux|ZLWcMI&4W>U(_h%rA1y=B`7lEr5fFf&CGVcevX*1d@ujT
zkBX^izy5u$EiX~>@W^_pS%i3rYlRMV1J4VtRmqy^Q$Chk{i?~H^)hpD-s{&q5UCew
z2q6vd&BSSIcUaKcvVbCS@td!hF5evs`74A8NyXsin&bD@H0#B~TZOT??~8kd`CQuc
z6-BV2>1<+><bD(+$Ht}+AEtC%L<oS6t8RO^tHE^qZ1!)T<uML9;CW66b$TlpQkm;~
z9655m{`i}aS8rXFQ^G1|dGv&P*ISX(BYQe4wGgvUBev;3hDw#ES%&1+?gngEcLO}k
zzmG&1Cg&~sahhkL&Y|xw%qcdHbQn4T4)CgkGQxYq1R<vlBfQ-pJw85t`DHA|4`T?~
zC}eKJh1SPK65{P!s9WF$%80WlahNCHXh!!L{SXMy<{fx7y4dff0F>5<$rQCC)19e$
z8#8{HG-8eWM?l#u;7H*{vL2-}*K{+fr2D+lNB}Nvdyyz<;)o70u70w)wkT=(rBXCB
zk`&+}Zk<n%*SH-V##jBOIL@f-?CcpqD!kk>S68=OlF(Yxd86ZHgT_eM-2h$5_4L0!
z<q#9htaCwye7AN^ypm6RWP)vtDRF1)j|B(HM)>zHI2QfBtJi3+d$sjz-;{gGKHewB
zA~z`jytJ^cQz|PMdkbJ1V@_Y33u?T7u>@kj>AqT7do95CQttFh;Km+iGUdjkX%y?C
zpCXxuTGE%H;_&U0n9=fsBzO?HmN~F60&LO~&r{sX+t-ZI^C_o+YHnzz%{~_`a~m5`
z6VHoXoN?Z4i0b&Wo8!HdR1sh{j_<B1wi8&_%+O$lj@%y>oHOjublQ4B9fgOAY|HFa
zbR|3YbyKaH8ts_lzlUazHD9BPwF<g0=bHFe7LpQo4alQw_XQSzs2``2%lKF?MQ+nb
zC^m&>t9t!+@+e(`a9#WOi@C*VheXW603&!QAkD4S!c`y+=Rs9O?(uMr*I@Fm6w#Or
z;LKt}r=PrJP`3ll9M0A}*fnIGI?^GtTPpTej{|_@E2G5G<%L587<su3*L`nBq=e<$
zd_F1<e31rPT;Fg`+4_2YY-{WDmJ4bz@ZhOQS&D3M6dix-$3}W;+D86imDxqlc0JC{
zy41!U#bWHZ&xNVFOTW5XXRLolpll6tbp@DT<aUH_E^oeY{j7K`cb4yaG~3OGeRP=Z
za7qx2BAK@TqIIFHjRQ6py;q4mzeYTMCqX}@GV?wfS}(E!(=d`U0N+2bOTB{q5WV11
z6R#E?XQ=#i#o;&Ig6RQ$fR^(OB!4McjHmMiSQp$800Q%7rf0TNe`qJ0Rgwy!&*(7G
z86sQ>d@gNzy|Ow%0mFT-<VLxYn67DgYbQrsXufZant##y#+TIt*1f()6?<)WFD|^l
z-3++s`NP{9dVv;^aj!YOj?Oxha1VSWUCsP~#KLcao%np)n@*N$SHwF%cLy)-i4JZ!
zX98(lc^IcFS%)iu-6dR&OGqB-wc7e6Eb;Be9f@~RTpURQ;=KM*9c4b_;(XrX`t<Vv
zla69OzVtNq-;pZhan&dJqH)*rlJb%g*~GqdmF(jz`Y8@iaZm`6_as+;r(-{0jFEDf
z@V@`an+zoe2Rjv?)>fII4`;{hM`aq$r<NKnY_?Xv%hT!j>iv0WASsB)GZ`0hLFKk#
zy6dm|QwcX7IT=sYjrm);Gsa$&JYKhW0+e4LR{nX)sqHYhmA55x^Ugp&G*u#%<W=^-
z(j`%G{<wdZ&^VZX$uufVV_qa*O`A0hWMA{~TJv@!*NkfOw^8U5x<@5;b1GA*RO>cE
z>+)5;7=ak!?OhSr4ra3{ea9eEOc719tK*ZU6evC%jsx608l6CtaD&#rmq?;S(0cB%
z)7VVVNRGC??zBy!V)=#?B?@%gyp=6dW15FA!BzQGW-@7g-6qWQdF}&$5?H$Ii0LzQ
z!@0>7N%CwikzpnVfEVBmqQt+}aWXD>stotxSb@<KA;EN;*38-^VqVC;pGB@MT10Mx
z=H0bhFgDjZI_=-s=K~Sj6%&^Ho>X4(iY%!ZQ_?|~!0~Yl^&wFi@TxRK?VNi-d*XpZ
z0YcQw_BvFaS-OJQHtL}GJL?z4C@`L1&Da%5TexQ-Uz^9TB(8L*pi?(_#((Md8J<ud
z#abdOSDBIKFV6Oq6rh6zLi=DOH6M#{JcbZ;-}gNTB-Ku*G%XB>#PLdqJ{CzX^*7y1
z3bA1NPc02r2Nw@6#w29Sl8*=7z~b0p9Se4AOLhuw*UYEAGw5?-4?9sI12AtZEn02j
zuy@h7K_f0n((RDM2odViA}KVks7c4KNHUXh9RD5aVFX&GK3#}=cyZ4MXUJheOOEe7
z7KWl=<CqhHc;OXsR6dMT93_-{d&Q`_pKL`;*ZZ-4hk#~iX(*r^W~y3u>`cZQKhKYu
zH%o#AX7{2-<K>WZ#&Tz3#9}B9<R>IMGmv>@rD9v8f!5H8&3i=PmE;p;=d-Y-yfdc-
zWX;M=oWJo_;-813cviDV4e-heHkur=yK53lt=$~MJqBDG0%lnK=oI39g6PrZ_~QK+
zfoYqsa{G)O#@SyY_<2BvZlX*|oUX#+HNtgnW{zFtp7hoC7weOGG$S}ep^NwR?%xFp
zb=z5ZIs0M$7^MaVI76On<gKgU@5`2Q0_GG|m~^~jaom~xVI{vc=u4D(b~7%>g%n`6
zel7pmTXTF{Zf}M0K2>X&O<p|z@WxL4(w5Uz`OdQtx!SM#e*{n}j?F4YF;a&=y4}I5
zdx%_7o*Jb#gx}@jZ(^oM>Y)ekC#J*2Az(rsL5HZtt^j*_M{())lb)QxeC?ZdkHNr-
z>@o$>Js-Z?+?2w_*DuJ-@Nm}7>kWQszDq%J4+jTtXr)r4D!F!M_hpq#)#p0=7+Gf!
zq|U*!-}1WrnG(?;ZTMs7_p{m?HEqSpcTsJ>P+K1S1mOEbnE@3Vb{aEWTNmHLdLfo8
zYuLWZ$g#HKi>Zi*mgp{QiS5SMn!6<#?_fg+%<}|mi4#l^%*??%Dxv~#I#jsz#KTD9
zSgJOkckww;%$fR;+Sb3sppH$I!*N0@HjiyGa3BthcZrEk@gMAzS^UYd(#6A5SWmA@
zinszr7n}wo!rZkdE!QmcxSGuD-JE}=H)VcC80URONRl45!cqCU%<E_!g-%->YOr5c
z(j)QmMPop#q)fMNRx!Qu{2jFLlHV;xz0?Tm>zc=67p6maf-4An<{*h=BY#urqXlKE
z<Yg?x%fc7?*1=x^ZEgf%NR<72;dwO;TG$aguNv9BscDgb0pt$xOR@7tJ7Pv;&N`}Y
zx&zjVb9eX>_6Z!GGBuE!<A@}dc0t*J;S>cAvxU=M>`Ma?l(>kLb%jO5%*4;Lo#DSk
zPZQK4WNsK>I2KYb&%OSoil}vq0oPD$PF)?ci-$5YW_|*ibG|8;jD4t`z8n5<v^LfP
ziw)lo;F(!<v5QpT>G<upA<4!m_}k&jp_?n=MY{3aUfLV6`##s=YIDcRQ_6HhJuF+@
z*BHYLpAXrN;1c#YXUfoc8~6$wv0h;p$tAz>$-+aU{$M$$g~4d*{YJ(65IS{Y=Bce^
z+#pfqMg>eJc`i3PbAj|nQR%7c+ZQW5;Q*`Xknyo^iG6pz=HH<Dhjp?y%*#rK`4d59
z;;O&?;+=c_yqow+3!BRPRqD5>E*glFy3rV=s%-p=gDi+QvOvXu)?WX?C(64Vn3<BI
zs4GMy0pJ7KixavT(_lYk9uDQ~@s$8H&%mu;*ums+!}Q>`D!yF?y?owgr~<)K#($iQ
z<pw+%`SAIwZcR$|<a5zE)h#p&c@=Lp5ddHn&(k5OsT#NBKHA>p^oZrxk8+~Lm6YyE
zLI}?euO}B)Q9p6ReC1f$FJx^)<Al=aQE{nW=+C9wqG@F=`|dXMJv(Gb_xkEgnwhN~
zBxuqltfn^N(`>iP$p9`PZcte*0=Ap~{c=Lw#WY(-rg&6r`lB^DC=u0y(x)qpa7ht+
zh+ek|Ilf#giVBjPB%Vla!iLhE8nL3_zI{wm+V*5|<vEz9Qj(2kCvJ9LSk*#Pc6~oE
zd^TSyJPjUmzcFF#Nk|soH$x)IOR#cj#Y0F5O7Fdks>&S-;!|bTSc*>WOxtYN=mH@*
z&5kyfW9RS8^4`>P#`?*cB_b@Dd>H3GeF(?jgbyOSx3mM+KJuyEFC0xuvf@d^qf)N_
z=^sjtEpyTb&gkyoc|7=HD(LY_HGfxbh?<%>m>uqw^u+_L-|1%A<c?iVs<UN|AMgt0
zScOiD&Jx0lZLwP2R4-p=<g5P!sz6o0T=-R(Yx=PzT3Hj9bQoLyfK#@UdHEkx=t>r@
z+R#<9Pxbo80eOwD&mG;$5%e(xmwZO_?GnGnD0y$y<L=KrWIT_)^hH+<7Wud$HCQAg
zs*aud-X^kO#wX!qefT!4{<y5o{DGf%k}(P5z=kluj>oi*jtRBJk6FhQmg36o+{?CG
z#kPXx3$aqXEy+-OI^z%zogClbb%X?$V)6P=-BixW6%$Mr&2Hk`H}ND3&KI~O8cwYY
zzOx}mq@)E-l8x3G?S9baM|308OkMj*$sViaC((&69aw_Ot5}jf3?67l&tS>SsiwW(
zzBuvt#hLo8j8%iB4UTM;mF&|ZFx})g6WaMCP1CU;kjZ04G4mK(qEolJ812a+mUvk{
zeJ69-CVj@2G0Rw`bJ1w2$R-`pteJhaK_tOL`>Q;smlCZT{!-KjbxP<w&Q21aE9%$T
zUQTn|OF+A9m&5kgzEUTjcO|vo7cObfVI#|y=t%l&r!>>)t&XdGZ23Nw_0>ky+jgow
zI(+24j<;Ld@V_&>J>Qm3Y)W1<>QVO5sU>;5ecE7`uFKf7_M-iDqF(;)Wswukp8i~F
z3%gM<cgYC<QM&NDj!C|QWNo`jrv}G}7Pk+t=jTw?$lsdNUKY-*ttOqMNMEvktiCP2
zugCGdEb}D)R_ap~vkf^FDn7`0+mchP@W4ZbR==0eQ}tEXWTht=j)%EU*n4@}9P9i$
zr0*M9bcuzJ;^Uk_p1$;GBP$unbKLo33XZqu<v0d$AP%rVwqLRzY!e@a8ocS^2c8NN
znrnGHSqFI}^Kt6ik{)&_Sx(d}haLD>%45v3@~b+_<gqdaeTY+dA=Af7ntgw?L{&_x
zy_96~c=8xMpLW6?=WQ9pRn);+GHCIkFJG$BsI`fck9ay>$g98;zcwx}r^popdNrHU
zDeHXv2dUVUe{sq_^cqd_;+OF|sC%Nqvqx}^E+3=l2Jb+j52>Ux-zdxv3Rpl9T{LQn
zr_;~cxz=&j`RjSeQqYH9VggSzYT=<{3m?0Z!G`x!^M$NFjE}bDu;<(K{n$#Lmlsd9
z=h0ToiWOQ{!G=D_C|;fW((}H$;*`uJb6>KuuUfdPKWVKod-)P?@#MRvBN=MRmF-#|
z8sjE@j&;p;e~u4dAG$D83^jZBoTSJ${?)GOdF=K36nVvf4l?>Tt;Un-IYMNaADTz-
ztec9L7tPy7Xp(VSNWJ3qeOm%fRHUJrSj#yLZ7GE#FP}@ZkiCxkw_Isk=c{V?`e0G=
z<+wywoxQL8x3^!mNrF(IgPlImC|*8O`rn=>+~JaN5l?oZ(<T8y!bS3uQ(tn_DzZT>
zd3ekl^)(Lccsbs6!-wCO6cSC+Nd&%z2$|DCZv1<OkYSY&NvE>Z^lNgZp6{17=+C(%
z53<nwxae=)y9ZV(B&NVeNgmNh)wA}Vp0^!y+F+sw1qI{}7<4W$&^%r3vewPpUzD)X
zRnRX_o-Wxgy5cVRibXMcJ~Te%*d)sZT5WG%G}%($%ZTUe8Xp>C=Oe!Aak}E8%gLe(
zl<3roO|@)uqfy(-(}u3&AsO{03te)u4NvqeC+*zxHv-9KPXD`{GfI>&Ha`8a>c+x9
z%5nsmmu;2t$^#LJhZY}CL(Y#!eCWO~7VMxaJBq{W!>6duz5chSd%eCyZh1M+>!Kqw
z%AU88;<qnGgFTKxe-%&0T^D2jBS-cbH+soV;Z?I**x=W<;Y%N0Sv0<EJIC0U5jRtG
z6X^4EOxcI0=C<OeM+}Oq-F>X5ps|UdG4=G?n@_Uff9aWkWb*@%)}vK6b?jsR7&3YN
z=*!sVG1(zV4>zcjAJrzGbu(Q&Y(Y<Yk^$Q`9{I@0ldPL<XaAyW)eQD}TQ_1j6@q93
zNqzIlUkCDHM*|Hp`vaQvvR7W}xN6m^)aEhWDwENR9<jz>+R!Ip;>a$n57P<>FiOug
ze%nqj*(0VINXe~s+!}K-U|(6@u)tR<<86zNxoolxLI$3t+9Uleg)qI_A?St(J)6(Q
z_?`)$`0)3-(kJtMlA?d8=zdkxshMg|`I>mTwu@2IE$_dQr7q>WkF3|~ZhL#BEqEoZ
z+tP4slI1vXDM|M)YtPN*Ym)e?!FUwTk~@@Z;i;NfeY@7zcqYNRc1-&A`>hg3@b~un
zk~zL|&UsnlEBlIhl3Kd{d_=E|Q8B4Tro9Ex94MMA@V&ej620`Rc)aYWw(N|PDbIQC
zV|m^WYZ8Ngmta5j?W;TvXqNGpePk%;Vl%BuE>OvF;AODLuW(k4l;SI~_Wsd_&lsSK
z2Mt@k#wNv&5xy_70haK*EOK6&-)gdtFHBL}FN^^>+4s7l`QOsA&sg}VcueqD#*7^D
zIhKqCG_Oa$R=f=H?YBSb!B#o$Qn!q|<WJJ~V?<u~_?T*A?Xu0WlMQ4zhQ1FCpYp3(
zbn3Do(}hL)<E2tziG5=AerV&n#96-^hQ`$nMtG@~9m0Ypz0xjxojbI<itN_-vMnAV
z`D59JDE7!&veas$Fv`CSmT`z)tG(?Se-heKZH&_!cRa@FPZE16Q^QVvR8OK)%1qPG
z*WmtqB_@1p?rRP(x73#XvM;*o+W5q))&8C;+tg~;bo`tbzdV-6=((0`mH5_T<$RVr
zw!I8IKDFzf;_!7XPS5vk@5j^2Hge@SN|{OfllmNIZz*;(2Gunmo@Y~HqU&eP&N!O)
zBiY)RNP?)q*9RIl^iiL>R=n-fx93ThWYEt2{cY5c(UxnH@iqL?uHV8<=XuRnvePn3
z8}xZ*#A+83BEG;YX`W9TnT@k^DY;ehd|!S&-kwieG9}%|;1fEML(i|^p?~=5d9izm
zDQ~C8v<C8Bz6aMhIV}N=B5i+xK}>w#E@|Zj4p}Zn$Z?@t*tclQ0#^;#DlcZ}aPT>q
zmQoZy>ZNUH<+$)6SK`SAJc{_VvB`r%9?Bp`@%&8b{w{5+uUeXAUv$;-sT&A%>1Z61
zrLG+t&!>$ZHYl>6d-+>8LXM4mdH$>&+i6P~Yt65x<r|i#OzyEu%hPGs##YMURe1R^
zcwOJFMd@wU_%+(7<5;uDvC3PL4IWcs%YQfP{}KyoF0!Sum43xXntw>Zm$v-DlRj+<
z@#vo**x+2x|IBf7)QFnbl25YeyP2t_r1<8%^f6)=S#0Y#^=%Lby|&o2)flMZ!S%E}
zKCWf<!>^s@jF*H4o>cJ196exks69;ViAahiT{m*9>7UsqlouN#o#Vw8J$=)XE^QUg
z$CNy22@Hv3&HnZsV%U&>4L9LJ$NuL!mVZ$r*)|n*Qchoqm%glfSlVWdXzbyN4RnS*
z?5ZW(9AKH>#f4r(=Gu;p=@ZT(D~YTSUn*@44;dvQ0k?N7680Q}nY1+?hLF$mFB{+=
z=$aSl*jU^;u7E(|O-Uv+Dp9uq<!X|xz6KPg=-_*j7^X0}VxsNg7{W)`%27Het;1OM
z+5P@7=7)7SE~%Iy<}9l$Qf&tu8?d3FFVR+GY&8nAX>G~bD4UW|UHT~dl3R(oX4@be
zD8B2jzoPrGdH#SA>qs8D*D2+^4X;zmiC6wELk`|3AKq3eC%dvESx*=32dS=&ucVdz
zX~tjE6|Fq)WmkOq94CsW(f0Et2^3%D@%8L+ytGAXc|w>0uSc7E0C6a`+MM)$)t0<!
z=;&gpca4*<)%c=FzKlb3WF$+wt4JSTY9q^dDA=nV8z9M1Cb1+}9tR&wmdf_Ju*opj
zw6eC3&)ZYm^J~~)*~_Ti?NOHdbved9r+-Z3yOfnbYA~f%Q4G*^FB6Yix}H`$F0v!J
zF093;md+%yUQcc5^V^MfEiTDYmwbLUc46i9)UIJ6-l%%q_&mRI%J*<8eQ5m9eQcw!
z&>Zo$_&v$d@HRZZtF3cUe6VCJCC>6$+Sgch9K=W8j}80C@b3ZsI7@oaOF3-tE+2kN
zBd2(f^RmcMJgw}PyreMV3we!UgxJ(9*;GBs&!jO=DyLv;zC|l#)UVZ}{AwQ7^rofB
zkLohkQTC>hm7SUnzBTtHM=e>!=6Ng8&gW&G+w!#(yV9lBzcp(y`)erYsm3f>&sQ9>
zUGhYi9CeMR-#uQ=w`)3{U$zx%NvA)pt(uKVX`|v8XG6NGYx1LbHTi={tKsTt#OnJ#
zmT}`ikJT&l_K=Hh#=JZ{%G#N^B*QjuyZ0|78p(*KpM1JG;n*Wm-|H!!#IP^*saa+7
z;Fhh;2}f`4`O*p&{Ix%8d*_C?v3A|+wAy99c<wB_lC9OqOj0J1>yq&+h<rk)^s8g#
z?TE(sy&ZJL^YW$5p+t|Ayn#|leJ@*+uU$m^1&Jh&FWB>Z@pQpc-`C?}VDOTu+17>E
z+br9o08BR92aNJn>eE*Y;_<`D7ayCKrOiJacp0z%A&aj1US2i_t&7^Pr3=Nky)93d
zrswO2VwbPl@s}N6dz+q4n;V||h(}!fhy_CCMtOX697?+MYqcM18BfhlDI=Sbh0j6f
zfwYdTc#_pPL{B=#r0?UAe61cQSH>yKc&N^!bor*m55#^9dVrq2ax7IC5B<I_hBzMC
zn<TxC9(?$CY0E~|%;$;xczq5w2^UX8SH6geK4ZczWmq=T;ro9?C2DI(2qat1j^dO<
zY|937ltLeRPJUz>-{R+l5yB)C+C+vKOyjaTxwEeWW=A7S$rp^Ik{!V&VZlQzN)opk
z{r%)wuwm_iB#-HEe3Gnv<y9=`VhkPo^h(=lD`c>j-ksBvvxaT?c3=_ZY=C&usw%x~
zDA6&RlyR2+;x|!p=-GebXT#{H-QSSKtQ9KYQp{x%Xwe0qR`#?@vXD=n{X2h*ApK<A
zB+yBJ&1yX1I%C@fweGW6$G28pbg-92PO&e`!ogGS!Uh~??BRe4ePLAI?~_bxNC`?^
zE8&SITa(sZu}!LvoWD-dA7^tM9~%^oX<53Ca~VUaSL=7#Le}$ijns9sW~<ckGF{{K
zcD+BZTiVri)XR+0^>Vb!SiCN6ie!inf9hv${#NXs*37OM=sxS6h23LAsEw^lNm!}w
zk`sMWIm!E4aZPJOdft}W<9J+qYRA><(JtqpzYfWl>N0-yb)T>45(D&fA2mb%-X<P#
zsEsclBYg@!^n2L1;dyyxxa8nbylwjV_kh=eI#?JRziIrKd|T4cO8ereE<Fi1>FM_x
zwo2KW4P;6i$oY38NN*)-#YA0VkNy1Hum8J6vB8tBXw=LF=7D&~K^NY%Ygi*!E0RUe
z+pNisqD`yIJkv3h-BC8+mGaUXRcm}TnQ5^i26W0&#$EO$pMICO|Lzw}@p+x491C=6
z_bAy_Ex)Qc4$K{0%aAMmBg1doviRd6ojE6`v6XuAA-Z(cFY6k=M#HWO{Bm64`Pz>G
zy0`1w-nNflKE(Gr-j8pWGCl^b5#{+U+t^TBwXP}Zdq2~(%VQ<G)7X$~Ilf6^lg=cv
z@}qhhooQrcf7%)!jK>#$j6J=ptvIP`=e6*n*0tZ?yP?4=k1PHZA2RCa=M1TvzGPpr
ziidHOMY5txR(<MH#`8wCYqo0g-FN?*bf%TZWT`QBdl=6dMsvK*I0=tg*O<JF?A2)S
zw%hh(%%*Zh%2T#Qm|s2OeU-A_hsLaT`K+CpZFX6*XY<@#v+3CW)^E~LO?=*m=+I8I
zB#$q8Ck;AEkkn#&ddHsT!202_d0Wp`yBw=<^T$bTYLaVQx0TcpPkld*J%;xh!@cz3
zWqNn~(s8L&UKBhZp7b$JpCJ8($@6^xzOBJ{+9Y~jre>q0d3i3p>Z|tteR~`qS$x&x
zN;=r^!M}c(Ngr8ZN}D7P#p9_yzA|_4#DOL{vMR{?_-L2@b*w!$(y!T+FOIoxIFhIK
z{ybgn(x#U;cm|4MgY0jHUdZcvI_)l=u!WE9Qciu<*isQqHFRtcyX;DqT5>fWdXyS3
zb0+`BPi-v1P?mUzE-cf0Y-bL~GU=Nk9~Z|tmrPB`f|jN$0X`Um^aeH5%UJZx^~pUw
zJ#Lo__DY*l!)J|G($r5$2`RU=c%BRm|HPt|B>CLv4Z8Q>rgw0A_DoyBw{~DzsHsrX
z#D_l*VJfk~JW4;_+LM7~1f}gJTe9bUB_aFL3Y8MMtV00y$tDNl%KCOhIp7LZf{AH7
zSdPZSqXRaf5}mdsU2@y%9vyR84l{{wJdKB0lae&%mpUCM6A~YJUqn|8&$lI?Xze(=
zWKuA#@xo0@eRiuuF{Wo1k%QyCJQpYRqT{2w)N|9X$MQ7kQFq+~B-1nC8s3)3lJ`7p
zciETTQfp!7{gt*n&$lI4qls768lUK1m$v%CRDH>-K3I6j{<Be+*wE)bso!bBLiM;9
zCz10$eOtD@4fy@v6OtILYc>WYx_e!^UG#2CeU6VGs`{#-mG(8axOMBdmSo`hHu_yf
zy6}Cwl!vF=ws?v`{Yh%Zr1N={e0dBt9@V987ad<+JZyNLZ+rXwcO&$vU9x?N+|qAj
zYX2$aguU!i6T9%DPg_ND>Swt@8F<ijoM@N0A~UXdzmhHei(adBZScOk^hH<9vL`xy
za8zHhmGaV?wDz%~BfFF8Pa~(;y^elUdRiG1yh-Eoa<ut<JV`-H$1BTwOWFUvefh7f
zWlwY5d+_I7@-Jam%a7NCJ?JqjfQ96!)vn2Ajr{Fpp-9hPgG3WwI+79Jk7JZy#s%EW
zNiXNyIlBBUfAqWzbU&}TMu-njxT8mb=8LW=o<_Trqp!K<d7_IaJ6(2V+uM*%jV5`?
z)cVjg=5cbKe<-&Zj~|mi#$9oD;pgp0S2SwPVezm7t>#lc(1XthS^DB(NA{tq4ZXDK
z`CcDcU%U-ZKiKV2F_mLcETeQJGf6#(or8^wx374p)n*)B?~ai!+mn?^67;q61~w>?
z&%N$%<n)mg$sUTgQN~H%^VL?J476XK=G$I|He)EqR?3X(dmNF|`+~Z3JipW}b;`Ku
zmo%@#mUv2Pjk8UXhce1uFVRy9m3^H^`qAcphrmN#I`qq%i+Cxb;%jDRRkLP(O|y2*
z+Ge#a-RfH^f5a4uZP(k*H&gR(%h1y{LiXq3YO7}JpPBS$<M||!zUsP{9F;&XZ`$a<
z<I`oQMu#RkH)z=vrp6vwq>GSm(<helrZ0c`NuxgUDmvCRo@7+xgF@VP*wc+BQ}TDB
zdxEuz2|ei+?13My#NZ3DDL&{WU$h!sJnF6@J^iSaoanW_c&a7SReQOvc9(8RFMk+#
zzHF<8$1$RSea4Y<C6SD0|Hy{n6zFm;<A;y>=<?4u`ss&q+s>A<ZOvt9`llZ9qGQ9)
zLwG7`OHcG1oB~vxxr(nZt5*NvLDtA#`w5U7pT>ih{L4b7WvXKC(ab2c6zgC}*%EyF
zz?WR=n>n^d$h?+sos?KHP3Zk%+U+?8E!i=BWJ)>dAd`<NabpY{sgg*pD?YPLjN7%6
z7+Ix}oVBW4KlMqZEGs{K1bu+!-%kGC_?UPpEW@Rpgo9&g!(##_-w3NPwoFcAGIAbU
z9;0YTXE}5SvCYXd)3ca*eyyRl_lBK=RX8NIQuAM<)pSdm&bgAl^8F%0fw}UI1}t`A
zI3P4drx?Y=n$lBUQ!IZQc^T=chUdq|z0>o&ZdV(5Z=)vH#h=ud^!>fL=BGxJ?zB9~
zPa5MWxo7VDP}I*(4F(*eQNJ6O8=1%1w`+FckGn*Svr)4vT9<A~FZ<G;re+*Ew<odZ
zWqf-YpIx%dXWcftWJaBLUWeaz<IWGe>keF(#*GKQeq-yJk$(^7>zjB!4%*%>ZQn0t
zbu3D~QjWf`AO_mh{!{8mo;dwAS@be4^;LU&qp<a3@H)ONpPp8>@jFgVzI1I+A6q_d
zAFt@1R?6$RLLV2yq%uA>bPi_G7!QVz5)but4CGVvvQKQKEHaX(o|_xoxP{|E;ahUF
zGZoHYJ5&08yq<WfWutYG$g(Y8l2t7o`jQn-wGv<XeCu11ylUBygtskQWu4}y36^sh
zxdl${qIh4PC10|m8X0)Pojz72Pk&S?<>+hvd!DC{vgdi|cpKu0rn;olU(Oh0?@+IG
z^!>4%BxaABY>d)F1|MDz`LgfhfQMeqzSk39^4L}38G%Z_HSAY1D_JY9(k8zpYQ@J_
z<3rQwDp}QC@{`iN9r-}F9_gf3moBGxsMW?=S3IKg{+#k1FX)pX@piC9DP=Qt`tLH%
znlEqF`(?{G6(v_ZU%b`xgZDYnC#}nv#wCqBzto-7=O{VZR6R<zi?`4$Ce{Elnhf4B
zG~efDm1ssRdES}l`L>s#O=4%mhV|y7?`2_eq1m>5hdsEnJ!1VYpU#s7002M$Nkl<Z
zhEm!3eXYzyes*@Y*}QRMMY>N~TwH9P+O{qE;DH~7i)hWCk6-?3a`3urBnQP8b}6#S
z&6`WT{He}T-L~bE3y^*Q$Zi*W`C=;1@{nu!RV}Xk(1%Bng&H34Yjk3g9x^JjsUNQN
z?T`$NNwTVS(e>jNZ#lJMnAV2IP#a&DEp(=l_jdT<Tf^APlx^faZu@NuE*0hlm~%dY
z0UY8h{V*5cLq`fdox{n7=_oE_vrM0czclXFfiR#(4;{%+>v&*$n&M+qT&?E;OFs5t
z`#DH?E@QB8%~s~pR%wTp6Adj!$o1*D)$1c;+l;a08vlVUC2vQ@F;N8BU-H|M&`Ajt
zBPx-im@uR5<ObrilLp^f*_2iawLbBNm+adt^<0|{K8zq+iI>FLkBx>(a-2H|_iRAQ
zHj^~09uFO-V%?O8u~;=OjkT|n<w@Kz7S)-Wzo8RCFS&y<er)dNapVB-LGtaHEbVX)
zs!N@V2g?cQF%F+RO}=`Q<6wEtc1sKukG%SOmuQpLrL9TzCY2i(n;#qZxuwM8ZPenb
z@u%@4ea50|b<IzWmQF<(C<GuA2gQp_6I<4<f6LCnK0oT-4?fjnhBuA3Y%eF0+VSbK
zC;zHb9M;jjq)YE$(|tUX_?lLJl+AJHf0yp47?{ta_}I*M|1Rr&`XuKDw)nk)&+@lo
z-_l@9cG00QSA~WCO_mt$OMFOPPqv(Bujz5<ici-WoiE-`&yZ_KuJkLq6CG;$cwmk?
z4*X8%nm?+HV*7C*=lj?!V;SWK9`Sg8im$|>%Z_x*IJ~@MripDDS?_D9ZLU`<;ZtK(
z9O`>~<j2z_Cq32q9LzJy%gY9Go-RA$tCqc5?PYx1%gU~3-e;GNWXl>`>E9eyPdm?K
z%ljA2*Z8Q}_V*>`I{4P)ye;vmb$-BS+>{cVOwH$(Ki5ezX*U<iJlnO$k#TUXFMsdp
z_<GqgZplnaFXdLE-xY(m%@~o>G4;Gr?MZB-@8u@(AsOir-zc2%;qS-g+{<_Dl8RO?
z@}vGr)-|uZ^Hw{SEX9}3pK)=I<Cng@ynG;28$+qPQhkyfN-p?#d|Uj|r{Y|{YR=vl
zJ;?R?gPi`>VbDj(m3p<l^r^SmJAC{$WPWVudY-qH9w!*VUzf$#7V+w<W|b2wWDdpR
zG0`y+jd;~A@zGdFcy(cg%xsen_U!!X`DWk3LV8ER!os2@Nd_Ns$vH}L>S@rqA+eu6
zblsf13~lKm@5|aXYfR5eTX3spEdeyw?6G|VCoKc^Z8q?3fc#Kz&~|Ch-=uUAMaJ**
zJlevbMQ=l#f!IW}T6TpQEYZp~{nEDdME4gU<b5Fq72f&71I?M)?9JZ@Btz|Ep)Wr*
z9<=3_(jR@D&=#)Eq2V7{U1RRTe3}?#Lp8dK_99Dd4ME@Y=u^sZWNQE0>-ctGV*Jh9
z%gTmo&r@4GE|lqqfdMC)eykBU=L5BP=+oD+*NbM=r)D1J2jnD&(Cyps@0VihGtF7q
zRiC&x*2Kn#g6)0wLeQM}$h?-7WJxxpI0>*V@QH!>0$E$FA&Gq=W4z?H25coueLgs(
z<bYickx#ZbXGo~z@RA?g$`|%iJNan?vI==6TA-qyZP-{X;l8R(?4%b*<H!m=Dd`;p
zb}dVz!GY1!%T(Km-;xpy`FvDq<*{hot+~vw2ztp7GpRPNi9ZJnUS^E|)971Os^pX{
z+1g+k5wxBh?Hqp|Xxm0tg@CZHvFVLb&=7C8d`sfc?7Yi9VA-gWNxrf#{X-CWC7}jo
zpeJs&V>T0x$)AliP24=@@;x(Xl|`OQ`t!A~nUE4~HlJ1Wmze55Q*}Rr>`~XNrT8yx
zEycztopG_1b8V^dly*kN1W$fy|B&pmHEt|5pLPn4EMj|~qIpH`CHTjlm-RAbTXrUu
zFZB<_{+#`Gp0?!eeGW8pwzWg=z)xX>YjV1;_O9C?nv2I)c*7CBy_L1n-PKaYm99T_
z55sTo`d@M3<EpTSNTs~Sf}noW3@LRGy3QHy#iOt(V<^pc@fDx9?c*=)_}^pVmGYc(
z`Bz{54O#NhtI?qQIE96$Awxl5eEOvfd`d|#``$*@=*Pxb`RzAKMmju{@_!Uk`jH&9
zAG_ofuRjjRavtg5r6mvkG=*_^-(_F1Okxu|Ii9}OQsTo-iLZ}Q?Y@`s(M=;OYg`YQ
zw@WEWw@_RP7KMpkSH~X$Aqq2L(-U+Jt{2GqAxU1e^zSJGZ3!hm|8|yif8OffeaKTv
zy-dyDlA$L{<#@dOG;R5vMh98VH^oL=vZcCaQ+n{qxnGknX?{)!3uL^1>3CUaOHJ~W
zc4U94weaD3S3XSj$5r+<PLG{<swG3;pNFFR<0PGxuH~!CuEsr1u1g=fY;?)Uw(4>6
zvLku=qi`O_uN^=6m^2p2kNdqMxluLq3p*;)#vnc6GmXBN^=)j*p6r+PIDRR=l)g{m
zcz?b<iQQ7Z?0Xv}4&p(ljMB{#{cp)qm&pk~w!X%2OL{eV-qEvR^YFiY4y7bdiKoXz
zqZN(5+O_dwOL*+Fl{<57Qb%@tJ-cE3`qXC?&92>hn!S5@2@nsw@()ffBKn6uPBE?v
zGkti-LZfhVl6*O+EaAWTM>$r8i>HsWQ{(Nm#MR?ZJ!Q$KnP%PEwYDwX{IG+Tl71Yy
z5NKjk5A(f!{n(MQG<G`_y0L29@cl(k=CK8~{=y8ceCf38zynw8)0Q1;rUKT`DT<r6
zbZApbT`ulQ^1%1;&@cT!Q&DW{m$pS$EqTdNr@7*sxTe$+PPz%9lQGVcvWc+l_7vMj
z&1NYxs_*rETk`^4e~cBAmqn(GuWP<ZuavLZfTqGfuECSjNH)atwRp(Q+Lz+ZW5|Db
zBpZr>T7`M*2ZJ6-1PaILK${<iPQex`%x!$BKv#i^Zq~_LV2CF=NeKSaKYr0k$qPoC
zY}xfEH_~_VYcldfa;YaT-8KfvBn;cJBw6F9x*Zd9FU8u%LeeIWC>x?r(tub~5<-Kp
zIz0$5J_*A7P{4Gva$YTCgU?+XN3u`wDbQqV(GnrZr$mv}x!re=5*$X<Sj0)OB+Jep
z<4Cshc%-9;@x+l<@Q_!sl4NaL&HRtQ+=rO@MsHzHab}B2%f8GJ#xYQq9<3du2Akwk
z1g*Vnqi4o2r1bi^m%jPZKTwslf|kZ-+skoGY9-rh?_Vb4`!)HMpyd%}OTIm+ZFI_e
z$fUZ@WjS51o(p5>I<7i)%zWM3{CG+|^^x~Q)0rzrdGAI}dR|9ubag$0R-T2WFJ$mB
z$|f?tOna;p12$^@WxI=~*!Vrf`26_+FI~q>xu!4O8c%)?^;%<4oEz6~uxr_LBYMJq
zFN4#%b!+Xn@4!NTnVqQSI?TT%_U_x)?AgP=CF~EbOj9vxZuscL^EJ+-r}k3Dg)Sw>
zwb<-SaqUTQWzNcc`|&LG0bLmh=jkLxzm2@Ca2BTUL{ofT4_D%QJGA+Iq-%yk(m92F
z+0l8YYfRSVJ0Jht>VNy#e7i5vZ}raW78K$@hd6|jV(V*lkIBo>_T$H{FZ4Awk5T#~
z918$^bh^RUwO#S>??xOz7fm+lCkxidZ9MmX?|Of5CN2ILpo2g8Rz0aL$#v=Bi=yKo
zUoE-k7a5NaeHGbO+_E8lt(Ht#*JMSLEPWj_@y69&7um8uPQRuj8@@&kJeW89p7n(u
zvO50;xw4WqYQGKmLtf@GMaP1kXr+$q`}+y9n&Yw|IUhmxl3exX0LNGV;(^1z&qf^w
z#g(lLKdTU3wVyvVjD#QZv=zU{0D#M9GO{Z_)HND(YzdEAzl#Q)QtD@F_xH5&-InSc
zjm6J_n&eXBs2zX#toeDSXv>KKUFKGKJ`uZc@_cB_*Wz?Xyjqaw`T%x17ZtDUR~1`g
zDMR0%FZ9P1*-%}^E*r!pnQ=DZ=j($;Gl@oDMt-`^&1|2sf?fW14K%!YkHP{TMRs$q
z-{%<Wu3nB`vMX78Ol8-{rnYKie0$G<#WePLd#?-QWkrbI1*26q$WzqMz5chiM|+2@
zm;rB+4%$);n*(^794l)`SZH<1sMRR)p}OzAyTvO8%7Q+njK%k}#2`+ht#n=UI|>K<
zYYeo#zHckZBm9x|rH2gvxZ7w+pAGBRB|bOU+C>WEOlVF73Nq;jEHvxqiJw?QLBVH2
z^vi+t55j&ww>Sy+F6`@({eXJJqJa#)DLUcU`fb;)U1?$`M&%1xbouB<1#Jq})TYnP
z--pk`U3B<}PTh5pVO$<lVI|-66@z@svX6sUeDOTm{>Dix(6a5UC7a#`ZG0l<Wuepe
zMRGcpo-dx{brICKa;({2idBL`b(VzOZ)G8mVaXdZV}!xz2PGJT#sSC2KTgkE(#_V>
zYHiK88V_1+JT;ov_s6?D7PT?tIWcIJxHONcvrOL$$I<`z&EtUD+OdjxrT?Vqrw+07
z<WpPcsyNveGMX<k@5ceZ^do=p_)i5n-xr^_d;&l+T{ZSH&W4SFgPdZJ4sJ>2q=bzL
zBaaN%iCpqHk>f`Od{*;Nut7_bwBhrWYHLzNqs<udmSiSq439PRR9by+V{q(!y_BSC
z^^oV2e-LF~eR~9EG?V3%R>Sn7ZxvHNiFJHgd<5bHYbBGtmYzhIY^}XKM)`E=?Qk<5
za+%|#r+=iRW7A_Bw#l~ft<JWSv?Qjk{rM+lXoSaKQuia$h$J;!w?pbN=%eqie?8^|
zjndu(D$|Sj0k+yJ;diNHo^$}U<2I>oDOZ6Yi9qKp@4}<hEBmD%k4byl_WYEx^s$$(
zS-Guito*2!uE#_???-Z7<Ehc1dAja7*isw5_YF@4|6U)uV$0`Q=1K-0W0nl^-nOU1
zd#)AmD96A`cs?XtMisU%Ms9anIS*p?yL<1xX7`>wwvv|L-$HTF3V4#h;lurb|5Uin
zvMOl(+I7tln>VLA-RoKbKWBS7ZA0UruC-IF7TZEvWh(8#TCvsRnoWQBb-2Z~#&#0u
zd+8c!w$1APNy=nWlUSNgrskLHOgf(`Oy*2J2Qtct(9i##$-bqQ&$ryu`H6&%c|KP8
z-bve^TUn0pjI@qvrg-#o+w1mYlyu(4DErGa+U@N-5pt|KhBO`vyPR)|LB2Fj48X$%
zwWq`DD&UN~?8=||=;?Q}w@VvY{@nnUk_Y4b(?=AG#ez{hiW|G5ijEI@$*&BIQd%M}
z$J<v{<kV`5%a}`=c+`s9)5^9mkR8>fT&+KiEc(m`l~T9tD=x{`>QYC2)#%X1FY|`B
zK05#CgA;vx5+XI%S>XrG$8OzJ&{auu4?QE~<I?}5ysV#nWO*oxh4Vf8*8I;h{o|eD
zX^9>e<_do0FFBWTQ4_ypk*8?9g=nv<>jCjm+A3jzz5HQZSw-;+1N`WGMMOHXo3-+H
zjn<m=V$~t$(F}<7uALm$u2SQawmjb6PN`e=yYytk;{wg=gOe~HXGiv?sdc<2)$y@)
z>3A96mOQlopS?F*mg7j$G!KZCSjfz-${OjBnZBHUrFz7Eow*w6+ny0A>8h+slGsUr
z82wcJbJwsi_lR=<WRlfoKpZnu{q<M1#m%>H59Qr|FN>Y!&e>77H(pF}{n3Znr(sNP
zpBfV!d|?A0xzC6?I<O;uh)?X{65~{}-gC>?lg4(dHy<&`)tQ5tC1N-AJ%n~$4{oe$
z>$bBj7ps5uX&a5!zQ@=xeJqe0hZyUuPn0)C5nm>E++~}^Yb;<Wzxj2$cCH=m<?}dx
z@8-#JlSiGLJ{@A+8$caL8ND|Y-iEIy5ALrwd9WF8@@V5zDeIhrXi3~ua!e3WfBWnC
z`3ruzVWf_mMkcT2SSPnFT?Y3owQy$Yi?Vl<cxOiTow~>E1{J@dG$z|%e7&DWtYbLJ
z&%V0HtK1$x`X#>kwskxM!z+;Tz+hxp>R_1cEIhM7l=Yn>vNy|(m`9HuB|j&huS3S`
z<lKn5OJ6&ntT(<dtomWY=hJoC_MBKoOy>q`UAu?FYwm1&ZS)%&Zfag^)IR%cfq@y@
zs|-9F{21T(gpnT2v9li-l;!u%SDw7^7+`s^vy822lqLVbM|{Lj>UjI2%f!iVPJ4@?
z;-*d?<)Rz8^NaBuIkfHlq)blz^nC4~MF_UXDu*`KXLe$Ge&ky&G1l!%cC7kQXRZ$}
zY<sf+S&v;_vE=k;jR`dnRb$tWZSlv>D^&V24lh)H2ok>)Di`CyxWKrLO=yE8pR!nV
z@RCkAQ`QbX(sSXk*M@Q)OfDSkVyW}HHnfGm#<NlcR+$G7)QK^73}w1v5{sYU#NW-5
z69X}mXB?}8;^&oaq<rlc8ya{FTFPo%97_^-Lh#APEq=0n40`&GHU*~QT)4zX5UU*5
zx(a>uM9Hwvr+dJM&V%tati=|t^hJG0*$>EworcB#Rfaeo>8}75GL*&{n!~gbl`S#A
zU%nR0d|ZqMF8q<(F9nTnjo*(ajO>F2M_bX8)^I{r(&~@(#t_&56HfJ!mSunK-@g56
zqs(na`K5e}V60!VAN8MWKk`lMbl}?SQH*N$J7}jAM?%&hwq4i7BkXN-j?rywz}~h!
z#C_#;<7p`AFF5Gaz8}1{KTg^?1V2<y{HxOs{*_Vg$@RZuz>ddZ?yKd=HF+T-PUIwc
zvK*DsAM%kKnF%g^Au~Sp<i^ltzlXBp(PMr6=FL*Qck|tg=k+?a4v6x{U+Xt2tL2;W
z?Yp;Y<=NAx@iaU)L{1Y=#rzn*p;^9TTZ{s~i^-Go`Rca?&i;m1GxKBMuU@~-8#J@A
zJ7wRm(XJY&dpF;I_hS0w(c8E0;>`rKfw;a&-`_s@HZI<+-)YPJAUC4EiKji^ysf@E
zimN}xO>}#6rE|<M*UIbx@TShKtm`d{%j1{?azVC^3*Gh|H?|4-2^(}g<<Gtkd`d$h
z7Am(%k57@23PZnyLD*7<`BHM5WKH)XqBg|QWp1;bm+<v%Otz)!tP2|kZhhJp;7#~I
z?+WAkDkmE4v?135yyfv+uj9Lo@#qTzScnBKQWV!@KdgeGpnv1k)P5S^XI1Ub6pH}8
zYwa3e$SjX`jIGB#m$<tvanJC`Y+J`uwyj-0<5=~F%0l(NU+pO5v2d*^($to;p_C)x
zW^uiudDG8IwqfSCkNwyOV?BA|Pme)=Vhd9q!PR!_m+Yc75kzUx#kLV+Nn90&8d*<s
zeIV`SZgpu+qG$VMdmuh}C|^pwufU%A+M-YnXJ1dS0W}*3bSCUJwkr4d)RVj2U`yFG
z0(<ntOl~`)c$UDrR<L$ZKWpnrL>onq6RCcckA7X7HijyRH%D6iSQ6st+kLl(6MeVM
zMxO|}MU@?=bIn{bquaY)2E&N2xxSre^m#zpah-cF>0bY#b-XFt@10W_`tn^7j&XJ#
zbUjXNP0kT8meF<mr9Qr-hwmB;xG_1WiErDBZm6-{V_l^E*0(iv=Dq#O#(a-|iVHAL
zCVq`Xd(LlnY@WDj@F;E+d>c1!NX&<)-<-a_>?ksB3^?!p<<*;;KRvGpJ(MNs@&_3=
zcHaH`>&<_9Qg6xXzPQ2judm<U{Qa|U)1v=2PUrJx#Cqc=(6gILwrO#nadN%p`}T$|
zIlw%6IQ;mKhC{CZL42obZx=Vv_EN{`-~Qq0Dign<G)sSO{rYclj>ECU_i?X=#Q1jp
z6b^slNCn7g;`eb#eS)xY<j!f4*T#7MEoByXlO<$ubv`gv|CA?YKmPRN&A<Kl!_7~x
zUdC&VwwpX~_`$}5M#?08Snl-%?EC>od*rT*dj;+slzZ%B8;WO(5q)1s(DT|Rb39Py
zLt^6Q4k-FJ!ikuwPwBIaFY4&wWXI8yCqhci!C5BfGd=<i4}Iyvhz-7Hsml#P-{Og{
z_fw7e)>}K&nr99g9I?=55<3p$<o2t1=LAEtJ(QJE?nygybiLfMmeqGoWXjFOlhcI0
zJ#^iyEVpgwJ56==zxK;L1*<3dEG7VO*oT=%z0dy`%9F8VHp!LazEpaAj+Y0o&OxB+
z@J$(6w<R9hhn!jDVW3{g=@W7={Mj(4V_tyb5}UHDAdf=)XWLqeC0l$i^kk<A`}p8D
zZ|Y>Z;zn+={nWY8RC#P?u`{NxxuH^hz8TTqB*q|rk0|Pnhjr}W&Kn_%DP@|G?Z`qD
znRMV54BS~xS**iL$0Ii>vyShwZy=O$VQiWpLzkqBjR8b|9j7JGr4zHA#F%{eTiR01
zI-%nCeez}+*zlKOWDIZ?uVG{@bkyxz$Px}3z*13${%a=Qm(F(bw@$o!yQkZ%S>3M6
z(Z^jFLrFVLmu7>w6Ap!uv(H+laFf<rdm{ME%AWs`_S*JzyGu6uHXm(s?Fyk&XNFVe
zKFBduPTey6lzUuN-y~ZJbfuXLO<(U9XmY#Hao=EF9`kM^7u$Jm>nbngwpW<@RO+Lj
zJ(;$@^}LVcX{?{(3tRKkM1GI6M*Y;H*Y<3|Y};V?2rgdv{8QZg`RVnmm`kS-<=X<D
zzlaO>rzc{ksP1{cTw_PO&cFF~YQ=o{m`^8hPBCliw{f!SYwBjP%w+f=hEX@pH|H2Q
zu`@eld;0)%exkVYP$sEkZd)!*{YA<@w=?4qQ^!=ri!M19F7}aZd%ay^^d}Fw=PHc-
z+MY0glQqa~9xOb3`ZkH&I4PsYZcMh5oH?j3Y?0|lZsU;)Ic2HM^#c#G;^*%?j?Hgi
z1GaK5++)8bQT!=J@YT0loQ2@{+I|qL?HKDQH*Ru2_J>$<bzNqeeG2-$L`=rq<8zzG
z*1n-kdoDc@H}i+A^G9_}dw2k&ZjWPzT^+Q-u6;We7PGC1MWCD-H;a3{xtJC`M(yok
zPFXj{xX(3wsN3^5;*&Rw+}ASyXiK+eesc|ha}aUJocD0_a&HvzU_<V_r%W=h+N>US
z0pjDl^7u!aS*|~IM}E-JR@!_Gx18joA7u6$@mfYsG7i@(_2l@rZxg)yp<jD)ePBBr
zR0(4}GR^e%qnN$i?Hmq1ecsICln3|Ybvku$w(wFsixJt_ZaMCnkJqgBc<^28@P=@E
zwC`xcEbh(XGGOEw`{jgdsNkD*m2ud!>+$vR>UhhJZMPX*UaV?sJ{s6}qnx_i!%X)%
zvb|W2IIpn*!~DDkd*tgjx|uI-8vNYO-yP49cV;UF8}!`d5x2*~jh#O(H+jJNf$Pt4
zqv-FJn>@((ej5i}f9mMv<O!Sc#t;0&^l`bC07p;MSwF|8WyJhz+#DhGCxF!L<-F$Z
zQnH`8X=CbTW#&C@`$rBi{}w-=)022_o}v3s-#)qd`Td9N4{go6Z7+ZGK3>%@CSLhq
z$3ftZrC-F<h6SAs*h3H&NOzVx7o*KA$Dergmd9LMJ3zfQ)YXZdNx-?dSl;kel4oy&
zHf9bG;NU}FkoV+4p5F`<OTW-x?iAtSL!H>zDN+Z6jXU1XO&J}@+?}cMDKAw09ovLI
ze9cm*GQme|_6xb?(bP5>^QebVTgoPI*wAKouIkLSn{kSn<=FM{>Gr*>tupL-8GX+A
zP^@x;T8z^slx{nVIPEO2whCIVu8FefSd8*nea3;k^PU*lR`UkJQqKB7=Ajw3%p2yh
z?Jq7*3<`<PRP(ALL7FEEDkI<G!4}+R3k*H;0l(=BNYSBAnf?$Le{Da74{7{oF1bd5
zQAch<PYkI$$%!L0+)s#q`6@Oz!!V=WDAE^w=m&1sP5CpA80zLtc~8I5uXG)B+~Dy>
zPSS@wbx|Q?Z-`K)ZE~frwH~Yi$n8ku$Z3C);2M8Xg&YoGA>mKiWod(hL7)8*hUCeI
zls54gYP(634i=96to*@?e{SGZBDkfP?TBh{>+@dRNUL}<{JaocYevPWzieMr=ugY&
zvZ)E89S9bE2c`b}k71o_u9xscsO#hPR)zGy{I+&8{P+W_uH7^2MOk#rN#8`Dj!(W`
z<OskCM>C+^+qjj2dQQuR``qH5|Ig)Re)lzJ#;0%cv8^4?c0dw;qFXBT3*$MCDfJ%5
z$U%M9r>?covnJ2d=&NI}Va~%p`UKZ06T1YF$3Tl+=d;|~0ta~PS%*pd9pf!cs(;si
zYR#KBVNb_vjc2L%$(}8CeIC%)mEs0W-0*q*W_v1VQN(zB6R(Bl_dsKfUxZqTxlQi4
z9@(Uy_`^-DpQksx?%h0o{3t2Lw$i?3>P%gXvCrl_{=s!DW!o?Dwflh1qwQYK_Q0_4
z!H&5<+V}X>k*94K(k4D*y!ce+HdWhSeIaMvzJm-LG4wX#gOAC+>oVpPc1H<rd4V!!
z`qqv*pUxH46aQk(;j(>jVmtPqez5@$CVfSL+%~DN6^~javDQcZX}=Z<KeME~=al*2
zN4d7@toNM0xlIDzw?*pA1=k1ap7-wYc9+CU-*;KJLq5|fGw(Kg<=eKk^BmbacRGG1
zL%%1l?P+T(IOde?>+4mYYg#ZS*nlVZgIsPg$s0h9CqjLa9LGFo2@QN4@4H<v?~&F1
zxyr^P6|rKIw&4T+`a#BA7fZb-zA29_Hp~r@Ic4m$wN9=-btA9-!J-3KZK~r2<kzHa
z-*=AgapHqE^f>HWEx|iB%E;;Cz0`foFf-F`Gt-YeVE8z$jF;_pof>VZ+xIhFmocm(
z7Jo<W?)G1@OdR+g<0<LsdJ1k#*fFl;v#}Jft-11^!(sm#PoFb<kF@s@bDC{h8@Z`F
zb|b%h%cj-U)3R~Ql=##4sK>6)yHRhf$WOPD5qVwRzVE!(Svu0E=&{}K()aW^I_i6E
zKV`am4IlXGHz26HEz0~$zJJTwX%5~*xz~~Fqhl<OI_i9U722c6+mk8Un1kzy8%D2|
z*Obj}>VTca=M9gMk2()j>~Z~F+#DhGCxB9L@Q2sjU7NgV1HPBFH(qn)w9!|-QT8U9
z`Vo&D*+0DCT&mKnpVXOu^C4c5*3-{%ca0xW<qohLi;D@12+6y{UTi`PmJOL83Kr$O
zC>N_yNXL!5m|&~a8w9V5<%k>c=-I&V&(FScuu@UfMES5@G}Lm-c^%y>**;>b_MGF&
zZ!N0$Ctmx*#S`wdUz}hg6Mc!F{8Wc!JCNtatVOn%9%ga^G6oHd#r`!0{3rv9JKAxS
z#2X`6Wcui~v@3?<EE%6|$E7mB)Zr&N|0ok1@-449g748k^<uzsOZa3Sj`;3hwDw~Y
zdVSk2YSGu092_4p@bMY9gAkpJ19h>KNyb9n{im|&_v*^Y=~@oFb9|ysA5$kTx8dsJ
zZCvS1`wpz}5Da6IvJl!<+eltx#$2n&6Tbw6pD<5*wnhHs$;RRV8}!%&)q*YKLe~wI
zg)XhYl87mNEp^ugI7_}^j7^@@2iukxcA=^#dKR$b0f%M!(ND$_T-i3-6aP|1yLi$s
z^0JI_TzjJrLBPdL4C?d~iTfv{Y+Kl-an$o>hrxw4^}0zEZ3tuP*{_rtlgNGY2S;hT
zbZ}vVGvt)>gbBU?Es!i5N5-=FNS{%zM9znnTuS1!U4oJV^Yl*~{4PX2$&+o!;I&)&
zBH+Cp#z0$oqL%%V9BX%W$ZmP_xC{BJirIntT^;r(Zs0g-+ib18<ac~x=n`K2G>#an
zWw?ZWk$mdwfY~-_XMSLk{mMEw?tip_qkojM_R{a_kmo*b=m>DEStd&B<``u-YE3Hy
zn?vW+j(lV3(ry2Re%PE&feXj!7RUB>&JCCC>v&rHp?!m^$F4KlOkZ>z9P+SzlfV&g
z?|U)m_u7l@x1RbKe{QVw34L+YALzEdU3sK`O}viz=gAW{50}@w;luI9X7sl)m9BH^
z<C=VHOU#j--}Sj4zfGFkQX`}(9J?%2r)g?G%6Sb>>+<TinqT#*8iCHS&A*z;k(<DZ
zS61!ke1`1>A9W6;zURHKN?tKCLcA6>_S*T6{WN_}vOo8pkORi{<QU~XK4iwNJzR2t
z2Z1dmlQz_QnL1e7j-RgU_TYL_M~=_lcOG}bYwCK%`l^@<IqG^F;h+yplsd2Sqr6t{
zLiBOK4{_ALjln)2+i2_?vP!f$vAyvPWV~A|nz9w)-}3=%w1>AJ>w>{Pw;w8z*jLid
zBi=%u@*A_do{w>K-ClX@nl_|+ChN8`0%o@IU(?Rc*V|$lF6Rld?hhS(E>;|J=8^N7
z92xOoZ_eDRc~@%zW%TE>aPx~Tev&hCx3w_K1V>NQkntf_`-Cxt!}$%4{j&V0EL(fh
zraM^5*mZ8o`Ua9Q;Ezg6tnI{tTpdx;41FmFH=Bzo4+HSFRU2J(JL6Pmjy(x1Wye!}
z&*dl1y)@JHIA(U2^w<y=Zs|KyFDoMl+hA)ScZ_A4-?7Ta9N)d!_S=XX>x|8{$vDV+
zpU_Tdx@^X&JwEsP2D@t_g3DuU_p|cTc4RgUK0ef)?R_5ga`(-)nKp}1KOGaaYtsKJ
zh(gEG<{4%t>l~ErTV8)pg12+0@WLPQdvl}r!$^g#F~~Dw#=2`4a>r`Kd`^9ggPSVj
z4G51-yy5aDzSKm2|2e*Jg>2;L@qkm$>%l1h^G~l5LU|lFenW=WvuQ)ltK&%AXyK_I
zcsegSaN2+&y^E((juIUD$Cjsce#qYwEi3ut*n2f<_bv9fl)uZjD^x`G&0c){H++0@
z$bOnVt@Poho=SSa51aDCs~=<I;`Iylx`ukME*-=Mb3aZr_%>c&xLBM`JOT9fT|FAb
zJG!68qg|ZTv51i1;8hE6;_)e7r|>9$2k7)&bntclxA6$mhlmR|ESj?mTbyX|qpG}W
z?wgow?5OJ?h98;bhSmGH;1~YGewm9IkAJXWQ6x(Jcy}}xgb?Ok^Vov>gtKEg*ujq2
zv%G-B;s?ETbGYk3jkt=NKH{X2e=bsT(<bABb4>Cg@BM__12Szo*4Q!+Mc0WZS1vv|
z6^5QV3#8+Xig?+9Xe`|LY!47+QC#NR)~_s{U`X32RPEO|fyj%ufx$m|cXA-?XE;ds
zaUrMP7k$d;@wN89ju%^O5J5*<wSm8E>cTv@qa??OQNDz`su_PTScySNnKa_ZdZD(z
z^t}f^a%|cvMf0*hM>+lNK@5I!%%gzJ*vOp|-9F<AW#Xbz{Z@$F9-^J3Z5)?+1ml<#
zy8dF*$5Vbfq%4!Hn^ZZJ^}9Y+Ywl3%^N0Et3$@dF`YUnT_4*<vlw2ROun6VkW`v1&
zS*zB#RYVRf<-{GIyorGxdGtGDMM@pzg{0=pQc1+BT59K6=YyyEy>To~Jh+gdp28uZ
zh$j|JU>R5zL>=Xu58C&M8(G<|;>5DPg}yZ~ak<Ehu3{8Z9zC5L8BFd6?5pC#M*=cl
z=61HLf<FrO);iV-`*wVS=GJ7lFsPQ7K`(cT@rj%2k&f8PeXZW4t#sCR@$Kf>&bOKD
z6_4WRHsHp#jrES7m9O=MZQpNbN3)qqM}EPwF3#36c;%vON}FMwVBJe>{JG>0z5A)!
zC$rXVsh2nDJeE<<+MCA$T@ht`lHl&~Sx;sA<MGWHOYRdmehU)1V4O6a`#;nF`^P(b
zm!eW3hgC#M*7VNVREp$OybC!@=8)KiIgK4EIpr{N&Z!bQIOH_v&2e)+GdAZrGlwy3
zn6J+d-~Zrsd);2w^?E)ZkH>Yt{O7eUY-D3=Or-J#VDq=PP_<Hoz`yD;66=<?x1e%6
zzDunxcBPQK0*M_0rACdb5Lt~m?zy!m-El@m!AY+RUtVUDEnmu|Pr1ev6y<wi>=G&X
zt5m1WT9c~`@qjVGX`dk9_JcsHtaCJoT6U)v@u^24P4Zxf!Fa~Jd!!OoD`d#;BN$i`
zfc)vxs(N9Mj1U{akEPaYxwg5()2bt4wjlIgbB_Q_RYUVFmWEA(;5Obpr%m<TtnV0F
z2on4CTTFwJ$({t}($uPtf&tfe+ta7z=#sc1MR4SJr~EXx;D&CF_=u4t%S*!Q&^V=!
zD$5=Eb~H>g)Ps68BcjR7Um`f?;(M7$bgUFz-z}yMnHzRRW$ATlJx+@-V(Bvdk1y?Q
zDOutW757&fefba1e?$>-sNB;v%AF6eUbNGIZ{EXg{eicM@|6_|isziWz?Q-Z`F@(q
zaOokON$Fq<(T7hADdp?CZSI+^gh|Fb*m+fI>zH|pRkg?gm0Y#Q#f4pO!3JXg&Qn0P
z+|944!wu)PCgMW9zGaJtrFOJK7mkOEf`<`40xjGH*1fDd<WblWRWDjyO1GvT%oVWi
zF^mm=EWmo_@I6-_wCKNn`8wUwjv>hCW_F43JABT#ui@vO!Vg;EW9ndG3e!;W@#!Dv
zKAnvpCAlx!N&ahh;!^=u`O2UEgk=D?^Yj`p34J?l+e}&VmvT+$iTV{0JhG!a^h~<D
zbz0+_^axlDC0nJd3+F*jmn=dr?TiQr34a>(@x*&`5BtCt25pD~HYpf&$(Ua&sxFZr
zpK{*xf?5j!5o=0$X@qKs@S2{9j9~-gbPtxO=f7oq02@b~?ftXaawpyKaDXUPcSW-T
zx2}crMXfJ!82vLJ|Gw~jbz*s+Sve~=t#gF8XDIq{H>j=1iC*Gww_556-es-L7z5L0
z+S4p>gTXw3t}(6LF_xaw8s8Y$UC2ioE;%~fnXChDJi6R2=pHjC=rDokdsaUNoY?Q*
zgTT{(B;@)gRZvt+xzbO}cT#RyBqdjuoMVkDAAzX06)Zap#dgbXfWVxH^s20jiaSrM
zIK7)<U`CAghD4R7W(ZhXbKJ2755!r7QW~4j1jXRpgQX|xjRhxlI(Jk={e0IaQNw3n
z4V|hbudBe9SDEDi99IswtMRGXHx`{KekM)g(A1gC)fci4#Y`omKBgZ$cau<hLcWcz
zZU@A)z~Tj^@Si1)(Bh$u_n&4Qt}*Vnsy$pgY0evz6u#IHKatrssh=dICk!r&hQ$2U
z{Lu4KFdpxg$|t@a5Uc*Bsl=--H{cuKXTtH%PZ<KnsLm40J)l+K1F_+?g^ZhV#XR_V
zL8XcTPI#Mp4PSdtf{3waB?hSO*yLw%nlb%Y_WOtQ;>YG*`!|&%Ul0v$u>YhyOn1F0
z;4!{IqAN19hD-hB*t+y6QR*NfXyVK-hfjUtC)H;}_ur@zx#9z6L$1irLN=>o!~CuC
zW&mJgSAV16QmX&_>#Nc;tn#_LJhd{V6C`DmO_z^D0~?Y`(&Y=ETfN_na)jTA%BPk8
zP3H^vmE*e_=iPmy)P-gzHuvZ4k;A$lTLUGzHDbr-j~S`Nn^nDzZ5w*f_#S`=vLAay
zk)?b#qXb$18=f5&GP`j?K%?5hmR|~3@Jzrgb`ocMDTX~GDDdX~`$GaJ1Js<!o2KGl
zrIk8wB!EMR<H_w^_68>nvYTJHL&bA#?0%`lE5+S`!KB=0UA8}bo4a#bpsqYB?_J8B
zDJ5+!OknXNO^0Pkf7Gebb2i9_g<oD~Bl>~Ej=;+kUrYU6>E@8570_=xzu7B&0qq)b
z9jUcmRA~}Lv@aq8!)-FOT28L1K+x5ne|8B!sE-7f1`}l?B&e?wD*ilBX8|O7p805T
zZ1Y>b``Py|n7EogC1T@C8aJ7H!v=c`S;~dQit3-=d+@6iyYkCxRf<6WHl{z>;_@xZ
z3*2$5toDv%I#Ouz!SKHmS&u1rVntqp?(lPKbruZb|F)J`gVl!8Zy+OC2e|JdrLT7z
zb3Up1?ayFSo@s@O-A>_KdlX~&9o?o07hz@wlfWTjnk(xm)2MBYh>wh~{<Q#-+yx>Q
zJ%mHfjkfoyyvCl_dZsU*g>|hHK?_k@(41xH@Ed>5!VB<3gV<T})a4I4uoYeVy=@8q
z9YOJ5>DgWI&uZM{J!Lzp0UK()KL@-HJ(k_seFCK=ql>W0Xum#|r&+>GmF)E<ZT9E8
zA6-pO+*&Q%ugRx$k8AP>f+0eHmfEH3(0f_kc{}HBQJ-U~nga}1oEB({Vna6P)fd!d
znLe5T_dX_G3YUtQnyYYcYb-otWgx_$&2&(0%eW%Xs7Lj;o;(h}Tdp)-+gA4oKK3r#
z2%c`|qFTy6R3p`)VenGaLB#+%ggoYIXuH~9F!z&@z%l&t`l%^yJ~&P}9F{DyJ$|KI
z;Zxf=dxFr&n_$=Y6v+0+R&-Eq{w2g04cx1xScPvE);sm)&qJC;a&s=d5NYfvIUL_F
zWlh1SSS#PcDPTKC*TgW9m=%qvZ2U@yL^<3vA*`G-K91+m7*)kOUzLqG+(e<yGMOu4
zG2$&F13&R<yJw3QKkTx1Y7Vp(*fanM1x!Oqv8ol|%_4IV;Lhu;v6zV@rDTQ2cKcR!
z%|3MWzzIZ94I5oPahG)-pTFg|7ObKWA$Ag~=!o3#=*=672smhsQ?tG*AH~4ZZiq=a
z0Q~;0_VCnIloW;dU;yKGG&K2wM(Pp4GQgz12`(IJE~I=T%0GtBA3m_l-OEnaYH)X(
zC7;-fkrhNpuGy0V8w}GxwYql9-H8JjCvmTGMc-|;bj(f2QN2CdRFZMg7+zUaUEdDe
z7}-2;wiT8ump1pSHC43Z!z8>|_?YJ8ww#fGWJ73O?{=02#P!PLqsa&8j3rKQcK1(B
zvx2|JgZf#ps>_i8w`s9AXva`Qb3sARTEjz@eT0}!r>rvS{`$xB1l#QslE&BC;Tttg
z57hR<%Uk*j*u};i<JmR76cPIktJ23hi!RZ>0yaL!Nld4y0oKZ-eo`whd`=}5M202v
zT~Bk}82n8kBsyWEZ(I8p-lz)$`Zlw*;cc;c19HIcCB}80C6Dny`{Ga!(m&q}#9>!!
zLj{)##Z`?{G3!^1^3Hu|CXIK*D7rS*M&vT<eskfAF7_-RaSxNVT637i!;_o=A8rO<
z$no#hC(uHzDd_Wj)lmEO{j0QA<)@+bs)f(4CrlTr@xl<Q_u=a2GiXbpf|{4nQ58VD
zOng*%^}E^=NrUBDQMRHZPg&23R+zUPCFgJ0oQ>38xkaCdj})*Wi|@xx+GFItF`L64
ztggbWJ%lzR5nRE>eD{D4ipsCMq6>+|_W&=OgHp&*d!inHRcGU6uY!+5Hl6()-f#ky
zyY$CVQqumIMXJ?ExoC0I%Ht$l`lUSiUcrCmKcimm#wc6meC!Mdt@0y0Vo@J{u3wne
z9N2k6M=T#};&mjih;(a3XX8Vp%Cs6gD#m<&iz$NY<B4(|pE71fl&)MP57_+Gh;F&#
z8!eX?mm3;j+cR%v-cJg9Si<?YFVoEQ&Cu=zRQb%wdI$T|Do?HDqWpsD+l~Xv(HRuQ
z1-PjeD~`UPn7^ibdCi)b3tL<QW~7qxlyeTpx<4*QzWojHEMi?+oaZ5mOW_I$Y|siX
z@^A7!bMT!~!CpCVeiH-G4zLEVj$uqFdC05vX(i`#b{wa&7A%XS)lhm^WgN2-WII5y
zC|rq;m$i>fv7mqkr|7(E^QJ@~rS~>ovcl|zzmT%HBsu2#+H~Pem+$>t==noe&=rzx
z_w>{WaR~1BvGBz+Ck+2yP|f~TDpFH;z2mmc0gQXppr*VvhQeN6c{toE>t0jk4?(Eh
zPs51gY^t#mv(MJ1+TBb~oX-Ei*j=zl(S8vYohExH=NJ1^oPX%73_jG%orKy_O20Gd
z6J2J(hE=6k`6+G&b_;f<fh#$3@JrJPZ|eS*b}=6l)54H@3(z`V99D+y-k<gF+|wRd
zuMBUa<aKc`an?3Gp)h4i+OE!;`h3m&qN<*zgR3?2C**0~Z3!pwrlngQW*f5cn*(JV
zjAte2==t5a*|)%H?r|!k+}`J;I$`26_zq3g#vfZzmB41G$pTpCs9v-rz<h&M%=)pB
z#HDrmwp;T1R_`B}-Fc0<Cr#VWoR--znMaH040#C9CmGx+7t@tYFwqlNM}Ju4F7ta9
zz(BUGg#0sb`7nvaMpGw+?e*>I??b!}?xBuR;+g8W4W~#)@%lWmy{;3VVcBmjO}cqq
z)mAI-w6^dO=KkxB1j~OrL~XO~5@x&|)#65E@{HWf)K$sm_=OCk$jT`XRox<qtN#f&
zAuNH5E62nR%|4~+Xu>X?9SYaZ134+w2A|P`OmQFZDc~^=Cx)3y)u#o$%75ZQ_%9}4
zA!n?Nr6?nwy7H;+q+x1*WFf|1GK!9%F<*B>CSw?RsziKG2@kd4=*nkYv3s7nzps!t
z@4x;S!usOUdN6KTFQ!)3-0Kd0kRhMfG)tohsY$0lT5~+7g7B8M<z!$ILK!okH|MC^
z@niiH_YmMlMvPHvu*d9J%NCrse(@lGw3Q_?t?$eLN_$~lLaD2w>Wb+~pN&(!SkZqL
zH(T6yMYaBs04Tp%epF;SP1mp*e_-69ykSGa#K@kzTexDQ(HV0Q!XPh>Z}w!X=y9)~
z_#gzZk-eiYe&KJyjM1%@hUbMUhr?pNttEdD7uelqz0d${u!53Jw05_lX|8Lx2}8Zw
zo|l*(0b1y{hi_Yim<CCzsutE!1CKg17+Y9N#>y+U&_Y3_7~Av0N~f2y5W=61)-%3q
zo=ebExKi>=)Oc+uh}DnzPkp)@_pg^>LyTOkqy`M#=pa_J$@?MTgi$GYFD0|slR&u0
zQ~N>ZS{rRY3940oNKK3a0`B;qv}Cl{9{EQrl(gEcZ6U$!PT+gHsa3VXA4;OXcWaj@
zZ@9=R9xGLfUs;Yx){sWsFq^V+<CMimL!LHtXe=O`b%#qbZ==QW{aKPldLpnZlC%YX
zYiG1w30@p-cU-qFA9Ed;IHgxvmnE>XswawEQgWU6*(!_Up3)$8AwOI(k}y|^;1lrV
zSg1G)vzl7K2(U0D&GS)f&(`^5f?LdZr^P`TDcH}IU5Q0;f5u?3F-Iit#ByS-MOL*w
zJfQ46GqdD|2(H~`6`zsE!}38v0+F$~O0_~N;O6(3KiPtw%m7IN?iG3_<j(7T(2R7X
zui~r_DI!Yc?=}SNY_Zub)cyQ@r7%p;lDK#I!7(c3E#p-%&KbI-UE^Rmml2*!xDB&$
zYbp|)UtDztTpaa|(?lKyuO@tshScrlZsZAWm%8%qK_`#6ZE53{Gf6rCNzPapKMEgF
z!nnaM`9j#*ga&&h6k%Fb>}q4W!YsBdf6XsGIoWV~!RpUcV?wA&o^EuJ%)fC_ZOUPW
zIpMG$zJ%@j^N1o9;A<b6nZxl64{7d)T8yNJH`Ur~;49$~q&aY-k2ib-En!BxD@I+(
zwD)(GfIbCqhFk(PAHd6@{#rh&4xWOOyI*Aq=TLK|#QRD-UbZj%W!J1G`|?dP^U*+U
z>5W<NO*qZ>biW~LvdPTIM|FB9MkIQFgZpCmhaYZT0si0K1q6R$>Z}{I(Izwhw?z7+
z)$CsTqB6o8GebZ7eZ))Pg|f@4sthGDS;trs0g27X1xkT!Vp8)bBa^}TDo$Zyrkd_V
zUBE0aFBj&OQ=AK=l^LnXS_wJ}7rqked+YV(&%G6BGWO`Q2Kl@N|Lx}8+p4NJfmv#f
zPtEgSPm+RVxF@QvdBCFR**Rei&v~8DcR+mh4S7AI-y4lTb<-&S&Lwm-1btr=RGM4j
zpV3mY?>934O`>3*=WANoZqMP%#y3q}OLzDa(B^f<*NRd;Om}_<tEPRoZs^)+QygxH
zlmJYp#;|Sx79|j;tsPaZMpXMj%fHGoZo;&J>dgSjHyO+y>GE5!6FHz~%gmz)tCo`7
zwHrEKw$WRUE}j&b5&Ff`NoB0W_5lXRWjhHUdqn<b8mMoWy<`igdM_telE8dx^WZe(
z1!z-YoY82DwPE?X(%`FN(#u+tD=$rfK<n7wbDzAdUso;@E6<EGTd9RF;pjmR8v_O7
zbl?cWNo3D6b=X>F>qc^OZTR}}A<0Xr&PsfF1-(4|Stz6nv6+vd?-V9P1a2o4xxM})
z!Fo$4oH3;aZ~CiRwO!|`bz-D|MB3)Tnc=|K!gW~hiU8?Luas(LM6{$if4p~vg=m`J
zEYCQh$%o(c;l1uKM6ODOj`Rw@T|Bxt)9@OvmhSI5*~D(#Wxskt#myJ;bZpY(K{_om
zIHayRXaUlY@UE;`n>7=rwytT(cJsDr8MW<Jvbmp<|2m@sv9#OD64UgE+FxYR>|moj
zH4@G?nQJJyw6?R!+;mig-rq9b-xV$_`T@My$Jqe?=M#(9@E<(4^`>0z@zMf0JHl^l
zd2P9QpT57n^5<sffHWgWV48Se<Pb{oM`Q2tqJ{`zqi5S$gUKr8+j8&<ba<G%=Ct7V
z!WTKAVQxqw_SzGxS8Ik}bXx;g;$%*(FyyYnb-Eq;(~R9s*h<iavW!CJLB>^YQ)rs#
z_FaZZEPMZ02U%$ylBE}3PS{XY1T#uABL4U{iFxT)xh47mZb|`!AMIuIDm$DX<`%a%
ztwF>43>6X=mhQT;?{cZ}R_d(_#hIh@pID%<zWNi^jTP^Vj>mu{yEIHUZ0KaKQ+4Y0
z5vg=F%vCe0`-1;yLDL%L*doYt*Cr{EyE!djtpVab(DCQE-4#iYw$s~>&WBmKZ5$3w
z5rnqMlmAU0d;)VnVm&UWDF==Fued;EhRZTL#?Prs-mc}V72>bhwfB4^^5F?NJuXM=
zmS6Zy?Q|u%S4$aI-Gt{SMfn_9P)hrG^1CY8v{xVpSJju-C#J5a!$XrVSuULuGoF`s
zobq^UtVtKM5|5gS0ljZk{i=S_6D|NQc;7kr!MWCPYAsl`;0vX(Xc7_{5L)P!vu&<y
zW&}5Jl4QMzFetBkYxzp(!U1_GPB!f3KrV@O<k_5fSfe})7lw14M}P><27D3=bB0V}
z)L+s9?1#oet7P&{SU2zIsPbY*;2HLFB@gE7#Lk>{Sdt3E18_#x924!rm&K^Styw#Y
z?%&0X;kU3%hb2E!O(@n15gpUz1u617lE2-{m~$|w#NYafMMjNog#(#uNZ4N}Ae`NC
zmmxWOo<1_#yy7d)?*oXD8j~VW?!_m$X!P%oOnLPTPodac!v+=7x%Tp>%zMfMySZY-
zkV6R1e;4bPLa%fLsu+JWEF!Fa+?0P!iO3)|EasN0)|X!Z`%j98B~Ii5|H33$`lnk@
z1Qid2`lH}|ln=7G!-AV7?MmnF(N*Q>wn<NafCuJfxwVX{hbg@4()f$QzlGir<GW1x
z0LXG#gu+c3VQtG1fg8#41-Q~2qZEy)Kk_%pwo9x^oBrduOgqV{Hkj&04695|ccFq^
zraI;JJp&O6onH{r*d#wi8PLR(F#)H=(6vmb;t~OGU~MDi80l9++_xFVt#Mb=bS=B)
z2?Bz-o<}0&09J1|++81ty*pxkY_m4HQ**SN?4CT)&4wL6%^ro@lQYBtZ5GqET_EKs
zkZIq02}-D%$LvA)v!VQIg+&Vp_7qhVaaFg!?^Yn@oq(9{4XKsp%kCB(%sUH&YR2ub
zOd|^$*T|)sn1*tNaHR>J{5h#m(qxG)xoS+^@gCT1)2E=HTM&12_4sG>%rlc^FMBeA
z9u@64*yD;xD2sM0xae;PKTWj-zV_1!d#sZf`3stjZL9484E7I?aE}U5Ro&1}T7E0-
zFP;vLLDgxDExD+Fixlk5L4>5YKS|Z7wvLqj&st%{CenHf*b&thm%PGBoRE-$wKu}6
zzt4E&;mWPEH|5PB&&wQ_Hhm+fxQBXM$)8eW^?lVEJT6Ukjx{2dZR@ic{XSrOs%<fo
z^T)S>+=uuazT1al(#Ev%&8nQWm(R$;(y*rao{J(a_{odSw!U`RvQXufiWOcS7Wrh}
z%^~9dSpY+)0|acq#6oyba&x$7q{VmR1@81O0{!{&aP@F*)c$oFBvq(Tln2;v#aX}3
zhO}BXT25#rYSMhoAj9TEMmNCIi%0DzvpQRSKRBIst`8e#Nou&SqIN@kj&A-uHI{yQ
z(l*Nk9i{+3gb9^qND7}rx!t@pR5p&Px5e|Ru64R@xpGg+dGA5x2XC%uD7-x=q+MFv
zsf6JGh2P^9>BkEtRV>n{N3^DG(kFC9i3hq7y?(iIMWtrtJ1|)N$>K&Z<^yRNsRVQc
zi3M{`_c|<Fg@0R&Y+<)&f4Z}OdC9IJ@z%BMT+~&K7RF|a@3-uH#o1F2^@}m#MR<_g
zhIJ6C-FYAWu_r?hHYoi2jtfLYXR2k8<9}#8@#EU(PlEDl9xsLf`k6a7!%Q|jwxrUM
zL3T~$q8}VaFIZeN;GB1*O{dT%WxJ}Jg@VG5PV?GJmXiNobvU7LXZ{weKmNq^pNjbl
zwlUuNuPa|4b!tfVelDOZF0!Bn>*8ZpPfPp~)$d`n3VxWpDf(ie=z@o*-a%}k*n0OG
zdFr&q1DCw`^f!02{)CAn{j?uDXCe)GF<W(_a)N)olw0}HcM=P03H2l?G6r>V#30^Y
zwX}lo7czE%U0M8Qcx5qo7s(AV_j66%`ck-a`)}z)fuL1uftgNuwnAw*5WR){7LK(J
z%Q$T$wtm*3;y+>K)`I>~bkL4w3mM~t&|aY{?7PHTo9aT^a5G1s?ZJpU>N0))=K~B<
z2P2DiRnpv0Yr814&G7Ep&I+7&7GQJH>6%Q(MPF8W>8V5=Bb#=z7~cZ<>Eqgn3!jl9
z^S)9thu4CsSrI7<o804?v(M;yL-WEDWt;{?<d@ZMp?sVeMDFqFZN^DJN9|~%-sIx6
zM)<kJ#>AbPK#^(@`UvKgLP`@Xrwi)yuWtA2A&$Rjqy}N5?0Bs-yQI0i=k;CDs;<aB
zO&e6@>CuX>FDQ9-c_X7CflAZ@71tf-#eUV9&saEn`o{=vg<(Usq%cuigZ@j4n!5(r
zi1efp10}-ipmJ$sZso^CKiAE?i+9Y1z!hnJiI0AVqz3oR;Xl8$QTdx$L^2m#rT03+
zdT&s@dsE^YRMQTIZ%=3rIMfgbh8BIU<r8%vvXN`TBt4M;e(0TRy#*>tYW!immKJl(
zN8;T2z4dCF$3$ywb^KTQb8lA|-ftZl)QeK?9ee3`nkK=ygp>nN@)$6egy&$E|Bp)h
zKj}q#c5VKsp_h*I9+6wV_M=VjEyGm{(>A9hw1>>|;(SuT+|4d_sT`u(qdJZU3*fL%
zLjZaP=<pZ#sw1ujQjUFiHhr>W%QwDkGwt+weI5{43LR9?x~YCiBk<9J@zS%-L7&Ta
z{JBmg-|px=YoBv04CRD|i0SGgX#}rjo1h#h8Y98-_10pDugyJ~(_<U`lpg^6GbD=V
zpOuWxN?F}KZ#-)Y+)VJTKAcaRI2H(ijMVins00T$#v6IOXOsv+TZEEUAJpL>H6hNx
z?hfLWaM<V~|F)lcqx;==jh8y2^~De(&$405gPM1=;e0LXjHkmaKFv*!tKPDKk%i)9
zLnDuz-`+ebR*M>IPrmI9<Vn`gL9Ay*Knp~B@Oo{d>v`x~EfL6y%bhDs<ibMjLF5Wx
z>wg|P>Q#FCu58&)5Aww$c7?Ykq7+Ymukm?RD~jVnnrfn5@9qTv(X4JV#|GH(D!ldI
zojBM%hLwXNg0ASdBV=Qq;?v;!r7LKE+|YnU-7eSm(0tG-Qjhq4LE&@%;E~U4JI5Tk
ztG6m`@>yg)qe`0~q<%kffsdTI1iQ=r7<{h4%e2Kp@G)7k1j6775<(vv?bKaL!i36C
zUoFDah-mMGZub-k+{$5pRQ#HBK)W@`G<azCoc~jp{nd}@4J!hTsoxuIYhz{h?NZOQ
zXxV0T1m^my{6r|aM@yT2@v+#D9lN0`ihE;BE{Z*u&-O)+sG$g%!7;Sndso)Nc3m7}
z?!cML`fkIt`Ow-i@P`<;SJM-$+b)lwvC!(IT>7&Fb|Kw7K_V{cy)I9Oey%bv5QMS-
z;I}KVQ?&}Ezn+EEE32qB8;(e7=vAvRe}?JV?(|NN7duDQPhPL$rrp)yJ~pzW?{CEq
z2cze`hZ0h$sGn_niC)OfXmR1P^!YJc?@?ECeJV_9_r#g>>W0PM(JG8E(KtO8v^koW
z&ImdS!0vQFA2?<whI=`6Okq7w+PU`HxDG#=RMwg1qZrD;pGeKdxx!!D>QJ|**ob_F
zUglP4;J0f$P}=#00w{e}XKUX5AIfoQPdIS0P3Z6Ln#z%)Y9$w^QKmntuPArssy5>U
zR1r)D?Ex3B_<6-nU)MDP^m#*cEt;NSV1I89M=c>Qrl4c>M4hH+Kir}%Wpo80*w->o
zR`YWG%1#Djy|CYR&Yy($h&bsXwo20hUPPOqbLmZO(GktyF7Jj{$}5|9EvR@0OgPJC
zB@u!Rbvdnc<y8T=U3gKO@afRQj)YVXnB@_>N{x}<5&yEfy00cR+6FiP+JqAqd@*4`
z>~F-uhG98@nZQu5xL2^FUcCBSpY{}mRCc2p%5<PEz4rE&QAIhA%yRc{r^IJQ=O;Ql
z&L?%Yi7qbjr;HJTQD1Rt)~GG#q_Kk}deY!BRZ)q18(=Asij~{?sOi4-h-!w5KSDZ?
z_y*riizli)_)`xEtyp^|tK+qCd9&%Qk*nHYs9hQuUB)~*eS$Bc{fdux?E7y^Jh-E}
zU4bY)AcygD>p$4<H)JfctTG;thO|2GL+ZBnj$oZHBF+4F+%Cm19$*t1?w!K);4$h^
zf^g`4nY3Y%=$P>Aw`?DoMeade7|@dO+yWcA6Ue&IeYEz0<*@LOAWh22G`W$w*E;J>
z^CAVZGPfhTy)7y5*k4jwJ)w-|SlO!Qe@6bR>qu6ok~g!3JM&nwWA_TqF@m=M@MX=r
zv{ux<M7k*V6={&`ug==V?XPH!>;m>@aRk3aKtu$HCVUnozyn8-{EKJr?;Rx`SIUA3
zpW`23-6ov7JFs5j3PNknr8C5k>661h_?io+BE6^Pyo`^GjQ{Rsd|#iEj!ZU~ROfRr
zn(J~ubT7*tB*Hd-C(4J;31h(htOvnc$G2|uYq8vGFCHuh(G-cXB0HugzENYPP0{R@
zX()^3ynA3yMPz^wyTwUdw?Z_3TR*In7^yP#WqPm{q<;1}wS0cvCK#C`m3)2I_p>S^
zY3PT%v7%Bg?4{y*XX1vtQ_4TF;ZuhR(Cic{ehup*;M&CL_>JV3;0qoc6o7V1E~B5`
zM8D5~gw_XK+>n9!;+OU|Umco)E)6E%@DMl_)DW566R?^PJTa%O3+RP!3|>$(G#RBF
z68AHzVKs}Z)}P&3$Uk(<I9^H?^^eYt2?6#3U&mBE%Ji9ZCh1V>(7F0T$0DVA?Q04~
z3yA`5Dldz{RPq(pHHAWrB&(E6joh0u=$7PY`8Dd}liPKNub6PNoiYYb3&jZcK9o{=
zb_)^x<dCVf`DZy7epGDl>%CFq1&(3rWP5gai-j8{Je-#!8$sYl<?M0WpG^wdyc~U;
zv~$v^NA&Mikb&llN7f!&tXU81LedYdPrI9h*-NvZrSR$bJ3TzEpSEYt8R!Do$#2z8
zXR!e6;X^Ml81Yi07XQSfxySV%Y7yivt_8!Y=$4X56(_xvih{hV*`n<Vo{0rO3hzDF
zU$9w3`rBFaMoaXi9JaR{Ce>ua`|^P9G{T3W&Gye5(4xwXX}_JW(`@UEXa6H6AhoBU
zRiioQXZhD+^@qFe#mnN;8!Qh?&Yrno17e~A=u0`)DcbU5Qu;Je$4i|hDzNqWZz8pD
zN<@~$5z|?#>{lIlqW{`8lKB=mZFGX)I|a}<@j#}$*`HKsy`7mtgxMbRmeF?d7;Vl+
zxAn)k+_n3;<=f{%UzDPV$eL4L=9CZ8mcy)*Fuj;aWz_hx-@$3;bg1^HHIWa{HzpH{
z2!!%^sx;$iK!6tcFW-aCH5)C*6aLL`a`2g71gFx!NCo#zeomTJfwl1vavOe&Y=yPE
zCZ~Mhj`%3aByU#r0n?B%+55}XG(7T>w9Z?b8}QL7qI1fPZ$bO^ylJr`N-`ofs}#ne
zj?QH*753qe<?V+-voG%0>V~Fgd7cSbXjbY(2_W4X$(*#c8_?;6(OBUv0<p&0S-W*)
zw!H9R9AZi7PoQI49cVq1rd%{A^7~Jz+MHzllaWOM(`D`N2%SnBC-uApfs5KyhnxpG
zxaJ{CJ<k$>|58Q>6li&;slCJ)RoIgN7}Fe*%1d1kH(i`^7Q<PY7BdqItY9~%T-8%J
zqN+siJ(|d{KzQ)Wp;@;n&JT6l2l{YGlHfwgDM3AP#uFT#VeylzZj=4sZSr_QPD+(5
zX0B*4Pp%bbxkLP9RalZ9L1F5uBc~ebFE6fiyVUP$w_*$M4tW&V@OYu|bR5HFT9kG_
za%pT*8HOGi`r0v8GoP;;6Oc&pK=6@L15y?L#;?`3AICyxforcOaN9gir~t-z#5loJ
zDY9d~b1k-0tKrO7i}Pd9i4#peYo5jln!d%F$9l@0QkuxwA9So{CB2{+^!Ei?vIeV-
z76QMNp^Q3k@91s48pCs(vz!1e>^`4+?=ED&PEgeShL75uRGGr1_+p=Dz>MkS+j|3_
zg#&fLh&E4uZ|@d$qC(r+?=qEb5v+8)i|?Z`VLV*Fe5XuDxn63KPftqFiyQT$8l^{S
z8|y&H$xN1CK;nNReK<Dd0!AGr9huRH93TBlP;voqNnt^P+Nz6|P`-r+y%KJ%k`Vl$
zz^M=7%pH!lni}pB@wsnJG8+|e<nM-FaqJZ#u|O2!7Rb-eTis!{^R$sxSkdSAIej~i
zNthimj(?j+)Gb+`-O$J7gEm{o_Kwy(LER~iP0}Jh%PbZ<DqdpL=iM?Kc#Cqgr5Hp_
zZ+LM!Wj9CePVt}Caj9Cv?>bsnIz7hqUlRE<E|#1TJSkW-)gBuqtJ}9Iqtkbd$aFMD
z<p`a95`IWWg>nq#Mh=ziig7##J#{nN_aUD}lllH0;p)<2dz_Hlq0XL%QD?$OYsXgU
zaEA}K%}u79b!wIk&#C|?2GMNEX}xO?J#Zyh(Yxau!#2l+Jl-i>$S$(y`i?$VYAhp{
zPK`FCf@F0f^Y$-K1DPvdFjM!yS`BL_NoM1ZV_ay<Il*Hw7X~KX&jh*vgOXB|nqlHn
zdDKQ5BujR{iU%^o16rkKl-QChZ?f}N{MVXkWZtKC-hs{C#IgKESDGw0ge<c;X}e!G
zX{4c2^%Lb<htE1U8_mEGf{88!EjMX+cok+#sdbU3K1jl?p)rACJBhM*fbQOds#CdG
zeb0@gV~goDC$y0AbC%AV!C6w16GFti9d6NtSW-^bXVmn9jngzDQHx-=yV#A@*SkTT
z^@E*zUzAIH9V{`<osz&D`ns#;;2`1io~In4{rgkDe#u|cq6zD7roN4?d2GOZtssrz
zoN&vb9IXn7yJi9|DD0O<e8o@tGe@hC*RHLi%~$490^a8Klrpfpmn-9J<=$wIS?W9<
zLS+JIeWnyZb7MjK7x{>Czjb{-cb`@V;9@UrL}ceNS^qQxJ+y1qyxwa_1S#gL#NP=8
znT;Yl=WA?-p#2?I>X8NOt*^&}Q(f?o-8cBERzC}|by%(z-neac2HFPQJHLOuin(YU
zQll!FY&EI+2`bK4KovK=3{ZkoW21|*Hp+&1-D5Jf4<<#vA=XnX&qDTr>ploA&kM%X
zZe@(+0bSQBp|fSPTFKQ+nu6@hhI6jfHtWQ`q=Wn1>m--45>EkP&AE4>J%6d}KY6oW
z#L@1|5~f6@nKx+&W}9LS{gW3Clq^kF8|+3vFEGr#dDZ4%T`(6?r&it^Q}D8Ee!;jE
zF}}-T)00UMY&Jzi4Fb%dF8?&#C3l=jzaAk@{@~Z=e7Bc5GS#i4u)cQUA1um0QBRz0
z^K!qm!wG}FkiB<`>@GpSLqo<;%t}+!ugVX&x}t@`B@Tu88E2RD^flJd4!3R4@WMAV
z-aR8Fm`Rbd<cY}8-vEsGIdTa_JqFMyXj9E;s=d}5bI(paEdOW%Yj7-NI<dz1(Kt&8
zC!*9Mt->+aCEJC$hAVq@2Sq*}=rsslnkssyb}7ADu%Xj#j%Oa8bju^6+<dz4c5v&x
zl1Un%X-SMteqt>uy|g<ab;*_7M}J=d{k!(=O)-+MPt+gPu*JQnLhohX@Ua9HzaOj=
zkXKg~lrKJRs+imDN(lEcX1o`AdF9bz>W|iV#uuZKc>i>%C2-MxKnS|(T+#VQ@7JGf
zh+FkYD!(2eMXVG@jHD^Liy*09N&h6ZHTuY_tu0V%UOu=O;#2*+*ArEr5u>0+qnqZN
zZWcwf>DZ0g%ZHWG>lZ8huZ`Vk@w%y<h`$=xD=qmz^QmVT@0cTW>9+d}zi!dIa^}=-
zu|~Dud87}{2>BhhA<G&wnKlVe*ud<hm$nw%xV=7pel}%xEWHnq34SZO_}@zYX!5cY
z8TwNcGJ9WZAleJQpsTlZPZnp_zU5m<OsEyh=SWguQyWY|yxEQ>PlHApCQ}fH)i<YY
z4E)P9Ig-Z9P~l7h=h6HW?C?pKfKxWf(WF|P{zg9=`LG9p^fAH&I!hMYIfJTNa0MR8
zK^+N^wcma?3?S|-&s1(R(}g?J#x!u7OEdrD0C+yDzFV~(IqgC(ST++rVy9I&?elTl
zFkR|`@jiBEN{{!lyC{2&QB~iO3&5)ih~p$U*EJ4iU!c`L=#UNn#H1=jwwQjLqpv-@
zp279~bz6$W?(Q@fR-U;#Y@ietws*^%avg6GjH#PjQSxGxRRKt8FFN3_cdfmDX)FrD
z5@USybR9bqDpxxQQBsf*$~_fYYkMNG)6S3qh4>bcdCIdIKM+T^7r_%79~=@)gOaOS
zBP1%HM*=(j`;w43_YKJX<wL~ABnG9hoLHIHDqlG=9#(j;kzIN-ubkP1&ABz!iU>=f
zpx!3O#&p_S<6z~T!Z63-7Bc%N3Nco;=fdX$G%dx%>aB+F{OU;k3fu@*MHyaxt(let
zkBtC#b#A&RELDs}$Eb0o1eonRDp%LVwG~*TmujDqMZ!5sYr_L%kwRd6=r7lqEyC{I
zNU5PYdW9;X!B*V5!!RwaBCl1>MJrx+IJ2U_lc=wx>uz260yl-g<BSz%OH}hsDq)JF
zQ~*k1U4y;7L{&((g!SxiYVmenC5_HaG|UA9a21?V0JiDzVCv78C!DKiC0Y;Hlto3#
z-W~CN33eH5<A~11&J)0uVIXW`e@wlhy8jpEzX#va^g^M#?C<41eLY=KG$&Lm$23;Y
zt?;KQB}kARD=3Tm_B(9!g7@btOn5XKAM9bbHbmF4-YvnLuu&(=02_dN`5ISbL$P;C
z?8x77lO|M(RkfOc&Gh=U=;5hqiE&q8LDE#=D3YeVux)ecPdNWgf3#&-%O`J8-xT4n
zrGrPW0(&}}IP@?H&v*3WCryylV(&kqYIH!OL6cvdj=+ow=eFfG$}SZ3AIQuq{P3=5
z*!vT5UeFpK^9dH+Kw<vtz#k7gB!iwfYU42|4zZe|^nZTGk^i;*yHpPEEK?6Ah0s?q
zRfWTz&7xj3qpF|2Y(8<cP#|^z%j)-ge;h$0Zd)qS#Qz(LzgQCWbv*!MmmI)r2=`MR
zNWSr&qL98Y{vARK_#`-Oq9Z9h-t(YXUL7KnD0`AVb!x<p@A~5<(^zHvswn1yIo4^y
zX49<UBmUUMG~b1W@4E`WLmB7a-<*=Oml1l7Y@ty*w{|5OpZu4wksyU9Ry)wb*zy0G
zmeF_r<6I2XtsXZGR6S#Pdaw*H=~cJ%M+*Bl97+lIx|8EoYS<wtztU4m9jwU(tkcP=
zlN0uuv*ZlSAc)$~zeZ|gF?EqHjWAbT*D*>j9$i*j6A^1MdC`GWG5uIoOc1T&q>H=T
z@thoJfIsJp5zB{I;!HE%1BP)?##{eBzZ8wz-}6D;4L-;(1j^dYZh81V&BFS77K3xQ
zm`f&>T#X;DVn(7abe;W~y~<<&<JS+rS5A*$UXik8&g{Lg6+73rR)c|;E%9O6CrZ_!
z?@k&}n)Afv9ax}L{ZrkX%81G>h2(aj4Q}7cTq~-t{6-jHdceuAyuC4!^G1E3cB`O-
zG`$w4*f%2kGkH2;@&Gj2tl-{GE2K<=HU^2Jl#uxp<{d(qaSL>KTV+ywP1auc)@)mh
zGmcy*4ldOQ2>jad%iK~FXM*ik7`RtzEbyzdcffHk%94zO<Qvb}z8hRtBd)v5et93c
zZVje&Wyk8xy!eB>;qLZJIeq9Xkm37p{C3Lqsm&3C{YLxh!$Lh4XzQ6*+2NR3*WBI+
z)jU57>Rn4O-Ev-I_;3;EzRtoReVdQI=Pk)7GaN0~sZ4?o73;hq_jRW3qySLl#O_-e
zu7GH|%?8}qsYCk?R|GWJ3KP&uU$(3~I1-2LxJ(x@d&;)#$_(3<eKyO;s&4^`9T>xg
z(XtJ#Wd6Cn6;4%{)fml$(s^UWuN4ZWwZWea6V7<Naf(E=475`c|E{<*CyLc8K-Xqy
zRXY-}@OpIL7MC%tK_T9!w%VaahFb-*mA%tOa6kQ&Wly*oj)75bNIEuA@@5~D2S6U;
z{ut;kQRO=%|06r~#XVt?zEyU>IYPF*a|hm2GLwHnsyBT9pBketj~fH)8>{GZ`(F?V
zT^c)NqV9=cjl0VV8R(BMoWS!vKj<-w{MFE$wqXM8(iB%nd+)Pg!4xg|ZmCAsl1l4j
z&SGm~__=BaS8fMHOm)h>O5Lb>y{4kvyN_)=1j;h_t|_BG)6(ozn3t_V*45mEcF%e}
zj6^`?^dk5~XzwMp+7KSZ_YdKaTBqG6?&E)biU)tdMV<F9;u<sLe^PxvDNkCW9org+
z4Hqrq`=R=-GE-^t0oS`v83V(UWs9H2OP&vwP`c%K^@^j(mg#wJu4|4kTzjBButx8^
z)ozvmbHY?kK=G2OyMtxpBMFHax7F2?YtEfg2g@X9jQ{DbeXsG@S;WSqLpeF4|MZos
zCuGYE&MJ<6vqTKW%3!6k;Y~mChGRPw4REVM5Solb|6Qdr!+0vff{`%toReR3Y)hQd
z05lr8+?7S)TU(vj3cr?T+={<FN6g_rN&v3{vhR=H?C!_4_wBLGcL<x;8Ql~$(sjYc
z%&aZ*dpjm5h%xEH>^%)#SY%S-1!!f#M8Kty@T02Syjz5`0>7ZoBD|<`wM)MbR+2p*
zTW<V#M!);{g1h8{$_~Y0DJ?aXd3A;Oh&>~DUD$(t|0mUBaIuwxHj#tD8_!(8=<#f4
zbvl87vhJeCD*tN}@a16Tu*IhhHJ^FzC==vnveF$*ESmeD0OXYKxqLq{oM5c#{sMb3
znY6pg_#Yo-v<kbf2D#`X_XS)^CW^t-;}rq0R|%O0d%<P%8w_Ku>065mBdg8XiIdLK
zW=9dW^Yy+PkN3-s-if-8*WS_DRCad;LOOQ_1DIaSDEvvl5^38u!@9BL070OqaPxpF
z>L!w{!U9wMPPeGzy&KE+9*@Gh!@V3>5muY3WiFpJ{^zPrU556+UHQv2mt5&Vd;&Hl
z%OY=a4VgIBkbAg$m>A<fUI|O(Q;G>)C)R%tbh*&lFD(%6Dpoioe)Y(7(Houpjp*U(
z(>P<hCV$c<;k_^wCbwn2fOdP7Lj7}lI)6t)H~tuUOIzN5%~OCHqd>pqg9Rnn-3@~O
zIW0ip7ZoAD5tntNW>up?F!%>x2Td8T5yt@|=*qUmQYyU=c5qB<G*6!<k9g^N>sEBU
zE??30I~a`j)O3I~&nM?mg;jS%yk@Mm;%jT1jVvIx>bk&#No$B=x4_P_J}VNYGT9+=
zxn?gA>6nsHEtu4cItV&Kv-jOQdI!MX?Mr_yL@<!eP9%0uxf~s`JvsAGyga=uPFo%M
zV4yuN*#&_K+Trx-_AC1<>x3Z1??yPLEchAGUDK6jB$VU#FlnRlHkUdVXKXae&`{kf
zVlzkc+Ia`=`20oX5Z)STm%1AvX)~cz-Cf^&MBPoTQ*VS%Jh)Wv#4wEep>NtpPK?o}
zgy}qz_+DFyfXu-=&JA^0tAsSNLf8V>*(Hb7k8j78Qa?siRfAbx*1#==c?ZyH+mw9b
zT|iSs3ZX}skh%U!iQ0jVwFIo6Gp<sYlZvlkf@CO1t%0VpwfTmpXzA6UNSW*r5zD*Q
zJ0Hz=hVN~{>8+{tEM@nJw}M5o)Od~qBrzicRtz=RJf}L0eWN|m`j6^$4`O*V_nYk6
zbF9(qN%sqYlyiWuPnvF+<l`KSCA6SVN#BaBr`oKX^64)zF4-ch4Sb@!&r*Bwf(ut1
zd8@`&Gx4VH6kOMmRG(}!<izR66?R-uiou9Q_wJNzuOqKOp@Ny9EQN9<(ZzEo0i4mH
zzkH=1&6HmsI%MExwGo1f6=|5D;+4JL9knGUnIteDzVL8B)soaR+lO!t?JE;!ji+N<
z_p8~HS0b*=fda)LqCQk9VX@=tuP@4}M`(M^h_2M(zN_Tt<aN9tsp{eTnG^bazwlu4
zcGDWHLnY5G)nyWA*7#=fi=V~ow)<hZ#h<bY4gGCCoY8`kI?yhdA42u!5qf|eOWpY=
z^!WF2N6dtL{$O8<U~C1qnzSYztxDONjP2ZJqGR%u%*vii5|g-cYe{UAq+wA{{4b{h
zWQL>ObBS=53(tr5LPol}>4pdQ1zx2xVzLxBO2K!JVzU<J*IBpgkdm}bHLUa&P>big
zt_auJJ3l7AVea&3{p08Sl$aRU2ybCqZD#)SO&?bnvT>^LAOO7)P(si4!{6!-H4Sci
zB;R%%ygkp%Sk{mm@$PkOIez0j?$3X!=2~-xIX+F+`^7PI@hY_SK&!`>jNTJN*K@q7
zQ@U0OOa2D_=1|fJb&u7Xrn4V5r{Z<d`RW7K-i90`H)DO>b31d2*Q;bxCzgz%m4?A9
zOLIR(2V<v>{0Ke2PE~b9FF0`=HGAQ7mQ|(RPGV1<&Db;gur81+Lbg(Or*(lHXdx2F
zjkZh|%9GLPS^fT%x)om=q55#Z&~vSJ*A1^9LPZH<WN#%i1XL?*!X!CWC?j2}GAZ4~
zAtCZt{<Lb{h>Ul^jBeGORl*xep}!_bb#lI<GLJc8LGC){5eXf-BKtUBP;R(L+sJbN
zGSo?ag&ra308#VN@T)AZUA4uw+Aa|3E!6i1Mv*1fbA>m-J(joAvrJGe)qXQp3DFuJ
z*!VNS%qYM>_e>7;W?U5fs%`-UL|;x!EXdncP^HKhReF~^?~r;_iZgeYY-~~-=)C<_
z+YEcr#%)S>Tgo~^<(4EHMP;*|6wX#|IZ~c2du@}s1_Ml;U2_UCJBin0-s`B9`X_;z
zAc~`~3zZV=4;oddwItef8T3=CbXe7~Ft5}VFn_c>Wh7n2i&iA@$r4H&;7sZ(Alus~
zsKM7gGsk$J5@gr!h)?&yI^~ov`6&u^KRL8yS7HAnZTQ7Yi^s3Wt}*%a#eOFc6u-T$
zJOMA)mj81!4dyiZE8ON2d^P5*U!GU{DeD}5<mLRrG&A(%2KC(Y(R1lvr4_ZFznVWR
z<<gLA{-(g`R5MR%|AXeZ_x5o)oauiZG9MNk7BE52m%TQ!G|RalO6OpH#tIzyW3dbr
z*{w_2WWAw!Rm<m1u6(%(5u1GjTv26MdE};-vk|7;dvxhi&fwAECyDKG={Vl?wB4N2
z<KarW(4gffxQ3GD%FTa*bTrT^`vW68Tl7UC9UQLGPeb*hGTc|f@*cxf!(oFrlbrx$
z$fYN&y;ZdXBUE5Z?FTzOKU{!;31?8HOs>MpbYyv!Mv;U=5$CvaStN5>pE%Knb;UI+
z6GjD-J=eP&s;jQPwbzmG0=iCSP(XI-)SLt9rOl1n<S4$12T8iOtLW_BOYEMZRqVOn
z#tQp9j#{iJ()alN7Z_*bQ9H2`;?8U7E<Lub52@~Ah@p!2P4-GC`GEB#EeB{`dm{Cc
zl?hBIoPD61L9USos3jN&gX-!QUiCC45BrN?tuC2Cs-#6|;KI1pcGQ`ks;Yl=4pik4
zFD@v30@!kiQR%VatH3=~16!lIEx(quTA0(Uk3yz8RSpyh?)ET~*4pj;&yAd@q`2$l
z;j4kl)`k1xn(LlQW)aS0&H7HarYvPY6VRI5nS!w^=hO$zJgR0dvbL3;rbtK5C+{k4
zOOGF75qsB`m);=u6hISR%N7Hbvzlo(V7nYMaR0%(ThkG(04+zZOfRn}xwJ_JqZ2m#
z%cRsvf!~-1e;C#K9q8xY;1++R_tb)yV8#5HJP<{B7DU=v^s}REE#MA+GKL>yqJ#Sc
zSu8Jy>m8&}7q75t{6_!Y2PixwKz6qO;ka_hU(n1w@|)BY@;_NCR<Tg+<O)QKoNnnh
zv?EjFSi|e5^q)ev^v+<87`YAv2urq%H(VlLMzL??n|;@6Ma=)fHBP$gv#xzRPADBz
zB2CU6)#|5TG;I_Dk%bELEXP0Gxt0O=68zX^{X4C)DpO|1YuZ?B_wj$vH;X)<oL_%c
zcF#e#+oIw5Ui{#pAFYAe*P`=>7`A%PUOa8hG^|Y)VRf-&!JG?GtFFvfp><C`)_xBe
zHaG3LBZqLDz5zo;B+I?Vd`BGSPnyxW^MV)b#(e)J+2tOUztTU3mv8!;o%@n6OPc+^
zyV|!ZLTdX{PQtR;HhDX5NFyX07HawRx#G*fn3+GX*1H{Fh+nYU5xnA5ApUw^+Xo}K
z>M3j7n38S0R;jdN(w3-}7r$d=%9rQ?R}TBT*o8ldS*~<(5-s@>!RQU56^zXjRh>kH
zEOY4ATU!2aos0KU{I>3~G(xENcPDmM)_r`^G{7q>_%+KXKuC3mEJnwY!02FZKGm9%
z<^I@uP1z)V_OWVJvuR80^wnvzN%gf3Tb&;CtfJT)Xhp-+t|3|<o0<bk*k!U0ldAO!
zA^G#2Qf+DJAh2p%C6;_RvLKNiM@Vf&Wx?a`t`K!9_!ra896(*=m?d6#A+YPxtme+J
zcfiN>I|byr-KAX&@x!;gob|0jtD-98mf~Q+C^JH+?VM5afd|~!?m_nvOBTf!>T4uy
zu&k2Dym|Cd7mice(DoR?y0o!#a!$M&ZHSh@h5BcabiA7>_rxJWNcwi<Dl{FCyH<R%
ziq4rWelD<1cN>yir^EHPdp4D@`(FlAh25<$)>W2ac`}jwB+rUIC+o3Rh*~7Dmhz{A
z_X;!_EJ*JttHy*9n%uEE88RLL&pZvUD4eiE*+)jSS6KP9qexqcWmnYv0V-=+zT37R
z)B9=@$s#+I_UZz=<_XZY#^^290ZPL*loGa(G0JUfzLi>a|Fn+3n|#V#MR-9p)RhGG
z1E&q!{=rb52O+%MB5|&HYX`)+`B&;|np+?y?~N>Gd?#vx$b_b?_Fi+Hn67CRPi@93
z@`ja$ZI67ERNwPIBs-cIMlxsX!_d;;pII1J@y4?_;lo&62}}Tq_xs|CuUNKD!-sH{
zKeumrH9XI~fkEjU5lFI{3d|Xlv!uKY$D6O}wvY~A5pYAl#b&SW+W*@?IKK1LL4Sz%
z@&9N#?|&-%_l?)Pp%N0=BgrTZk-bhOBrAz<aO`z(>~#(*iR_g<LPqvHwo}%zH-~eq
z931Ny=UB%vKE6ME|A^Q9c-_~1UC)cEy7l9eL^)kX-sltbO68*;rV@VZFHpy*&6y&P
zvfHKz;Q6~lAJ4aD&LR}zIt1fL^9iure`nZx_=7S^B`&}JPGUm{-N>uH8<m|EMP;96
zmp506<THQhbH3We7=o7`rra$;JxKYQmq$NQ7gPM5!DUT+Tln4gi#}#^j3pg7@^vTD
z>=^W$`OK1-FP?h>mRvD8<&-)Xu`194y8K+<-kIRFDyX_$HP6HCWqt{#{PKBIFzT4s
z32lLI5=QZXr-SKrS_?<#ZC)!h+7N0}*E=K0<R+y@B1hQ9E~t+9;pBo+J4{+9JCNOj
zayqd@Xg-qWPQv+gPq@`>ndH%IIDxq+!mqqw&ML!g#vtX?Ox{0#WxIOIRXRAS<}QLM
zC_J+~W9Sm~HZv9`J)Er=g-hPBGi#OOk8YJ+4Pri13Vi9J5RFa}>(SZg4p{S^z|Nv>
z%P-l;AIjV{Lg{7K{U%k~9#YGf=j3JLMO@Q%Cx`uF*%MN-@<`(=iOp~Y0|ue)Hr>GF
zmZQ&yyJ)wEd#<Og2mDQ9ti#_h_m<>CjlHINSDx)O8I>*>hs`~v=-myUInHEx9pZ;F
zxw7du6E;&ajFCvN80NvHsOonA{pOietnb((+lJ)-zBK#waVVlD&_FF(@oez)8if0=
za(0eD;k<I8(h1w<)!sIAxkuh}@2w_oPZwk!jU4uCly}e3`=dNjqEQQsqdsojEbzHw
zk`r1pbii;D%$_IXL9oN8TJ``UxG$BU9Pj5!)(A)e3eo>$hoJZ8@HP0z>nAky<oZJY
z+&r;}dvj@EZoYJZ54nsBt2sh*(cZ-lOHQ+*@{j$IR>QNl3byB}tdwJoUV;g>xmJai
zq@mOtcy~0Qb}x*Ra+MsmeqffQoUUk!nyXWYsPXvMOxdw<>+3<Dt^Pfcm+reDo(H(U
z`)DhI(p9oV$sqzv1c#rgDNS%Xv&Lji>`Z%Zb1$tQnuX#CyN-zR`dQrMQT}NmU476J
zH?W>SXDKlxR-i~G&%CUWxTO+10Ts*e38uo-WG1qwD=dqbGGm*>`)aP~J#0C#keZGm
zu{&2Yv2y5eWd@DYIGe3qoA)nxknjo7W1n?r3t-aof8*dr&4W$LLV|r^M#d#rO~*1>
zl_w9_b0&V@ru*9|W<qq=GLJCrvli?2LXIEv)sUUGQaEOM_&2xPVI>QGre%&qxICuG
zYr~}aV9jx2{JT9iCR26kE9PtJ-KH(%y9+j4CiK!aI1~_G8z<q?%$)>fOG2FgHz)v1
znlWdWv)wUgZLX!O6f?I29t1;U{YVz6A3BeDvR)g#((*}65WxINS&a(6H>MhwsJ=Xn
zi*m1%J$a5vIs5YM+{jWQY1vq2Eoc)Wt7uc(psx7xzW3m1?7zFTk-bZKA~$voUKn*r
zVXJ%M6=ic7WiN=pm0aQG<Nm8Mk3_iMt=50ORbW|^E@VW@8mO6A<7}bKgZ3q^p%sc^
z+)q@IzpB9}PV`3$4lj4wzAzPbn8aTB2%b{Pxe-6B@l0;RPcDR-Cn!nBc29u4isiZ#
zIZ81dv2Sz(c?D^r%7Xq?#$568Lg`rz)OdlPh}6$Ek+*FHATG_HR(62i-z<wyv#U!s
zO$x)t1cPZ@M2kF|jmRBCmpl(d|32C#X%9+?)jj<6)hk}|L#u2<R&0qpF`p&_y|_H#
zGB9~`#?P|CdtPs{b0!vryk@M7ki2xBElH;6C0w$Y11Z-21B1}+mhTI(>tf<TqJtLg
z`2T49pl_XZ3+GYBoZ~;#hCo~ZpuK-M#gn9@?WsOMwM3rII%bRjX75+g9O3wf^D;%_
z584<+lNF{0>tvk^F@5XQ(SN7)57HgzaO+%V3%12p4NZRWunV4BE-qD|7XC)l9gFYD
z4JTq;D}klA3Zm=6Z8(nhj^$(2m_ii?0?lp-1diC<)d())-PwD=M`x(?*haquxj~ev
z7D^)8BgyjM1V!77k*)AWd~A=*fY_SN?byNPjb1xuS<x9!ipGj!x$Zm)y1Zre7xCgE
zhGbcrGf(K2le2uL7L@4UGJUVZm-cHb{_j%q@8pTrj4zo$BHMz@$LYY+&Q-7YN;jm4
z4%Izrw}~lTHmAssm-6}pgjcs9?cVdzi<-7^T43wlDKXx*w!zH=Zv&^d2)MoN5NK6I
z>SSi5&KAUtPGk*SMkgtxxl1L#S2~yl)QgL4k-aV+HQ{W}9!Hx9H?EWe0o-*&Ms-3M
zT|xQIfGKSZsnRO;=xcOiSIdlsxd_4I(1~s(frcbFrCcsAcx0@3SCJOMU1b`rvAOz9
zZ0bjhUpaNAz+G0l)64tqGj~HD7&~F40x^n*7<EvSlUXU_MNZl8r)1C|n{IV&tO<CM
z;jkf-xB@~CQ4cmH_#-m+eIxxIQ;oZ=sg)i8yX!pqm#2-mWf#|PpCQ`<o0mo%H2iEA
z$ba`A2PW}S3WLa19cb|;@cc#f6hI+xQ7Ee3uwiCWFQz_Oel?BgL${$iewT8>djx3H
ze54BG#hIOkO8XC~Z+CqkJUyGFTN~+J8dPmY{%e=s$xw^g)w4JXo`~TFcE>bo-;~(B
z)Sjz5|ESNL9d~1E8jAdYQbVM|Ptzl^vKSGyHA3wq*wIznsKguKgF0}>zg7~tv!5lu
z5FPL*WS-iAPAq{pz1L0vG*>2dDp+W903&<|(h|YrT^%6NV4p}LuSfJ{T7FD$)0^z6
zVjk<|C)(kb$l|lNy7<IrA<;uqhQ;|s+g`Y~6L5rXPu;%~?oqolTMb;I^V2bB{5Pz0
zpTS~T#Rma(Q>3-_8bz*pn>x+7IJv#(G?`*D!`fu@-TC-tX*ngvJ3^r17<W^*Hf}mH
zBB5!5`Db^KTi<s^IVFhfBVzXX=s0(luTAsfts56>9%`TDY_KBC_iTDK6sASps?Ldf
zU!JqL^!Nt%L<r3P`pvTO^;y4Xpxoc;_d~;7I^9&jOl9L!zz;b;4gFg8-qFv}s+|#D
zy>|>5`xll4q9<<nw?5ZG4ve)oxT|mqE+slpF=sEin*8=#GD{;NIgSCg_&(6vL*Gc^
zebf5F0q;qh<gJUe6R$VhYqdwcy6iEYd7Ja#<So@R!=2eW8nk-*-TF+H!P?CA+DJ|a
zy)5L7OY~I>2A;<9a=%4?%Zn4)+Xy9ESKS@)zO7K#vxlcu--@MC<pRra|JuGz8*H+A
zryez@knMn)_Wl{meZ}89(i%j{Ii%TN=rBv>U&FEI^=AF3-ZPm3WFianQ0H{DKZ1D}
znx#ZO9QeBM-R~0m(F5U<SDXnt!i_5-l6qTRi|)o`K)3i@Lu`;IT9Kr)bmrk~e4AUX
zA2YJKLrd@^n*6Ea#5A3rb>UkiDZ<gYlT+HyOd$v%&rP1=pd5Lc6juz*K$qawHL2@y
zqhKeH(|u$U_^vGF%x>@TkINAI_p^RkgP^6*LYPdb``I)pKO?X)F{}h<Nd0`sc6wNz
zLvJ!Nc~d}q`1bB>>yfz3R;*ZIZK1aCnm?QKXjFBTQb%OQ^kRDJ?sR$J>^Q-H#;dvi
zgki3f?6oyb5At&oV6#EkMV44e6^zGo6pr}~IvW$R%J#yZGegTy(>R7)>^M`;2VmeG
zEy&RW42$y7U$I(~9#+_1=Eq15m@HzzFF-&fZ2H@@w-w148a|GmLEKxNHI&Pi{L4~_
zDu5>R&YXXk<#WYe1(Tewd28oS54}f?Wwi}P==fbnS8b@WlB9tBVgp4xgiQ~){8W4+
zarCmdDyf3}@t&-+wZe&Sv9_saLnDzZSlk0`y}7zye<rc{@f%vqPRUDCv>y%gE@1Xa
z9M_qSK9ihUStEJVUGpwrr!Z@Be++=T3?U@=aj6I6no^h}IbY|Dpi3dB>;5(Fjc#7=
ziZ%gsIUdLMwd^pU?dc{$&BnQ|Mscni7bnvEA*lk_E%md5J3G|{!qfA1w=Z+W=Ip~l
zqdmw>+Q_NaLw@p`=wM(}<I@Yt#;VVC83a(|57VXlYD(<=sdu*)fbzgqbaKX2Z}XO?
zB-Yc?Uz;3sG0rVFsZQ=4Z%Y12(<jic#c8?YPGjf%{;1IpS-<;`c5Sx;NHp7S(g?h`
z1kN*Hqc1#WJvY}ow9Y05*UQRYXDsAcTdjOg4TT2-pnp7zpDx6(3oW$z>gvpP4uwG@
z{eo>~@5zvte>k3+Z_a-gl(tyJeO<<92Sszb+!*NeJzFIV1~Z>taR4{qe$X!j!bp$j
zX@<IbF%BN+!{qerrN^Y0g@B-Mc-4CZ@;&L8V*@bqg(J0Hf-<Kct>JNAOy?sjydr|V
zk;MsSlqu`4cd4!UJIk<y9Xw$;j{}EaeMW7QItdn*-Q8^4k6We+lB2zeTjxI}{khLQ
zr^-PLhYazQ<DYZwTgk7nV+(28y4yAoFt-hEtQ<>lmG)^*u#{er&XQ~CN%S+yjfpR&
znjht@0NqQtD-i;cnw-xOw?81{?_7p3O(3rwxc}=D{?HsyUHehHJ5?$!1vq_x;IUJ>
z$8bAfkj7PVs`DFLV-={Ii9>i-FZ6cYbk~_2^iQF>`N$*_eTZfVfA^BIp*`j^K&-vh
z@}6pZq2&mqI4PYiai#2(Q6Fb6|3Y-`mD@@o(6PLlpIMNCx;6#tddu;vvpe1p{H6f;
zitD10E79ZeVyW8+#{XVcQOK9z^mCZv{1tcSPX6Z4Kx$o^2O2l4&w0KTTO$(Vel;?D
zlXvN{<nl-m$W~ZVFHtGab~UZ`{wze}{4QSDjfa{t8N?VE1$`4=4Crqta9(SNc7IJ(
z{ay~_nvfdJj3{H@?i?jN8AgoC(+AP#ad8HaRo)~lhi$x(Ry7Y2J9DIAqXz{nb8fP`
z3lRn%thiL9tCVGS$3wK8Vy5b@?KEl4c=xDfYW)H$P)QfWAvdeIu`*Ngfd>4lR_W_M
z=&9CkX<(_nxRMo;Emd{0Ju@`u0^nL4Cnt)_93{^Mj?Yq1@~BM?ACJ?`l8N~hJwW!7
z@3HPcb3-KyYN}=-Am?nhb8$L@-lEJpdo)|5DyD({FQj9>4d>(M$Vf4{iN+g<_UGJ%
zEajNae!@_`Oyf{7N$mW8IjgEJiqYx49|pEyva5KLFP$8_DY0$GhkFVC;5nxr$`J_H
zwvNE%OI31{+J@l3c3ixW-KE7l1fEwDp3I*8iHXQehh~xfL=h?y79NLWv&e{^Am$8%
zkD8f}p!1=?A&2cyo#C1-2VXGqJ42TZ_uJ-~^OM7OTkF<y2i2LCrua6on;i+$gNW|>
za}kuZ?qBV;ae6`7pcTSq)!sAha=Oktp%i=@)0bv)C%m-EI*ZZA7Q%7)`|D<?H7qO>
zzjvZsUsr?qus4};RYQEJcoMwI_?whuOM#TWS<D9vsbJIgocRU#8oNeBPxPf$1+L3K
z@bj5?!bN!r#26=EYw>u++URRDk(u5X?M^_MT&!MZzwOd1k-MWu*%+Q=dHjvdw~+u3
zHI<>iMPYemAewI^iRd3~IbLOn!%f&+F__#tcf?Bg^;YNfgZ!4)lWFvl5{37l6|g4D
z5nIKWYKGxTtLEyjzk;zX%sI1%I4%)&X6<hJ+3MfLZ&MXYT%UaM5lP*Ib-t$p2MK2I
z6XN){KLk{Zcx`cIjrz<F9{(x?YVDU8Vl7NHhf!kQF5UyWM%)3KF}q-8G<F1A+|n;H
z=(n_a3^~ceRoOLjqL|5&z2_MgD;aeb=~o|I&<{s8_je!u#9{-R<_x{F(JP3Y%RRxX
zdj}z0fYrSz9Zhf8uvOSPoyxr%jK9XbDNIId?rxF>E0D7aXSU}fUHL4_SC2%@tX2nO
zT|&9z8&_P?KY>nJop;8*X-J9U4jhCb3jn%U79^QvTl3rsE=HhK23@S<264yn(<Q!a
zc=B0)0H@lf^q$Gth9#xgDueDioqEt}>&1_MJ8oMOsdhU`@Bu@EF~QSy@_1?V*==U&
zx$~A!7dz4yq}2;DDMtF@7<^ulexOPxV1qm*@&pP?&f__j6|4R!Mx+aZQcRnFkZumu
ztB_9!>AnG|(DnAN!5@KNs>L#rj=!;IgK>Q;sB9sAx!?qDmwDG*xi%kp9hqdV%Uz}9
zSnR!O7Ia<ldqRDdJsRg*$1dsv<6hV*(L(T+1EBKi!mbO<+(r$8KoOA*4r`AU^fG4u
znmt{ZDy(XCZP<!{v6M;95SL$<i|%?SYu}%TbQw2NGeHmb!4x>D3rLRRHk=FnmT`^G
z*G@JrqQe_UiV@uGsN&yN^RLU!1)R}5DV(<=0~9ciM|TCX*Ipzo%CUYz8PcO<NrQ^#
zudB;CcQ~}QPI>rmCPi{u8BKt9^@}s4Jjz;kUR2=&Z-}2&{t@F^oBdtQuJth|7=K~6
zqm{(hTP4xDsrOn{)z&7;v@|}+wxWR`y_oh0epTr)cW`sTyRY&t4LR?Me;H2xNR3l|
zecVIJ3<e(vid^B;94v37gis`}*c}gFS-YiIH;b0aZ<sD04REdt8t+-gsRsIXyqqDh
z%18?{N>|-sR^O`PY>b&wL(G?((5eVUM|<coH8k2k%zu-_rWt-bLuEG|ij}l*f59UH
zKQi3%egBZ-8q)AF*}e`1bw1{SJj^Fh%l_A?lku<P_%nAM%wVgDg|_J><DF}154h=M
zuYL6QHNa#W22``_FuY{B(h%7j_-uh@z^gU|C&mBbS6PbbsLh|=MU-W9Ka1$Jfjsck
zD|cO)clFP6eekVUDEsp$pTw;Gth=2wo#*w0BXd*Gl=LKoUHw>T;XP2gMzqz6{^a1C
zDzdcSiEHL%TPR(H6ORPlkKKGHkkBn&VD(0t4g_vt)0k0H7I+Ikzqx~?*vwQ!DXv5q
zoazQ5RIWt>6zaw(#2+JYo+5aMvBcz-gV8%YJY6O0bmDvTPCVx%o6w-m324dk>Nljf
z*n+TXn?m=_4Yz>Y>Xd8z9Sm=f)dJp*!e{)nZf?$hU$OdDTr#KJo#}PPDn~lpl0|dN
z9+;9I9<A#HRl6CPMcAAHFOCypYM2fe-4ypR-AH|3_IA2&WL-Cq`(DH=GfyrRF!`v1
z_qxz{8!E-0=VpcYKTtV$f1~BekS<5AvRAv%oC@D;KHz&#!Zg%D>Cd3mluM`x$t7Cw
zFlLRKwfXYc*PAjzR0>99i+l*rfKK7-16^!SeiWtutzq+N(~{4I)q8S240v-BhM9Ve
zQ@DqmPy^I643zV?;VSxO(@4TxUeS`{AcEWlAXPH$R5x=amFbuP|0(r(J;?v~U<knn
zQqNxX?MDNu9Jj5+uw#i2oMnTRo91mhAmw(N{Z~fxoo0YpQaxDzpIvpP!WdDMgezXB
zaye#oW?)Xb+bG8SP)}jyQ{Y?@_inXl&-Y}n*#3MQQBr-aO}&kp(!1wo0NB{hjN`vU
z>m0nSWITlY<&+6_Fl&&(3riJXKmM~KsCQ^T0|9w&u!v_mU(ZhDom6waEss9%7GyV4
zB^VoclJoX|iz?OG1K|6_QwHrR<gX~_J}#3X<+_DC1VVOo$GOL5#`{3)5QUYD7k1I&
zV3`FAZrK?zE@v_p2)edX{V4rsB$J=J!l{eq+1si}JyD|dTvEN!M>rsIgnL~hxlr3M
zmZj_AR@3(Sr$O0EMo7PTG82=ZpmD}bbx$zdY&i#+FL)sfO;DuXz)Bx@bw$1%Fw!xM
zKNdxvQbI2-x5xSB3O^_UT)I1CODX&t`%IGE42~0R07<=Oil}B_-2os5wu#>}gsYr3
z9kMi69uba!r;E&b!W?_fGE!1-oawen12L-p5@H=mmSzHt-hKeZ<}*``J#-AsP46q?
zd|G|}Hg|sX#%KAbttp$-tS03?;dg>(mxGT5DPs<kieTevyAcTn)$^0C#2d+?8`XrQ
zVcwSc&k9xu?M45?(ak|+RB47TGg4rITGzKNapM1305YC@RNu5sQW<h}npBbkMe2}E
z$JVsE#~IS}Tv$5!DrfeJ{ma1H^dS>z_yiL}?oGSyK@flW2TF@w`^g%*TSPA45G!sF
zZVGuUzw390A82%gX$NXL_Md}l<hermlKB&b`yx^Squqw3$zNZuZC;*_JYpFL2^lwq
z?-+djm-#9k)1}a)lxxdK?-xSXW?&`7!h2XZvWQ9mj`}Fr-~4*_KT$)O9QB0=n<n$G
zHJ)3Fvr>OZ3@V0&hMeuA2)_E>C-~%;mL8nysA!N&GNNmlVF0sDHzIMbt1y1{KfhY{
zhP*zigRFL2f@6Rt`gY}V`9GZGvy~lINy<Hhn{D=?hnQ%69s{q9M(K2;w-E`)q~sK(
zi~3zg%x>nuHywD_v+2!<dd#1muVnS>ACmW0t}y{+c48J9Mk6x;!>GnoL}4AU+;G|=
z@2>!p+;K#iy+dQxa7)yp5Ax#g!o6J@WRR7>wdF*d0JZ#aZlSa983+oSs-u<-hs5U3
z>FP8Er25K*-jK}dTyo$srBV+S?QN|W+mq%^S%{c+=AHC5&Q4a(ZW_{BUDxL)$%0n1
z7Lt&!lAy&9xqwXMh64Rv%mfH;z1c*S?)9CNKBF=BF8hf_6ccCK*1zUx)^MxuBiobJ
zd3(;hTX|DtHF?z3(oT(QqXatFFU#0bRFL_Jj<<)IXKTBm4UjLxX6EV2-*_&+&;qhZ
zz@*pg*s#az>9MOC_-FCVIsucpCFN%$2ahOv@ZZih9MN;TRi?uZyHNy|rogF2rWQ74
zpbT{fKV|Ra8^CbcYg6f}!`pt*wE<}--6DT?#U`QowLC7~)IlpA<C1cZx%l4>)xS9=
zPs`4D5$MXgh07K^Vs1)x)~ap6=#Z4Z?^thjt}a9KZYp%(!q_`8d%CURg5I$B?~sa=
zc$o`B^C3M^#i|(^9ivcQCPGoXNPgdM)^ALxSQ;Tm-|NYKTgH2m>Sgv}{olDcxO(Y3
z)6(1RM&HcrCNMsCwccsf>c_sCe-qU{^ibk#0kionCJEDx&E6uIu1039%6<V^as8Fa
z&1)e{&J`b>3<C#&zb5o7&Cl&)C>>C~T$PT0Tv=evHs4cu`L$Z_#1U+l8Pa?q2^S-d
zVcmNg3_P33+J2&<RKnD%f3#F?<CUA}<7@QI2aUFgI5V~@qvOv{zF0&6^RcU#+aG6&
z`9f*DT^mJBiT7-jV#ClSqwm^q53SF>0K?V%6=&%cH^;gOEhNWv|5=)x0FsaD)Za#F
zpV&@OS{aVlHgffLwQfQ#|8>b1u*#ne<kK2+l=fZ=;A~mLr|@lBncP&Bcg^}8Yi7Y}
z2KP`4>{s?&)kv!RPJ5V}*l6P(Xz{Rm(ruG*Cf1fRo{?8J`yxVCBJaOEg|Afqep?wO
zq(OiC&)#YGWRyN(<mr~*9Tl_m%g0#8q1IM3Dst?;v@HDSdPwninyw0)ZL_>l_&B(a
zSuk8Q<-1$1ez<f8)Y{K;f~4!N060l$5S44@grsv}>rTgIzVzMD6)Cs7?&%usMIXL4
zFnoJ>Ezat?=HJveYN^jspi>rcc4KGPU9ayHn1It%eGK-rN^~q?QM?lEB4k%?9+>7!
z)X!nyu4!5$(HBS<*#As6yXw%nvu^q}=uSFB31yO|m=1!ZVaez}vo4B{)vU_dsd|>J
z_|lxRM?O$taKo@LF<1z&H6=$c`gpeEZl_hY924jG`DyLBlg>Cd07!;Rbz>rRm84lT
zhG%(6FXg67ahB3U^UYk~K_2+>aYvL&CU|gO7$gFDo}|ne5*{K%{EIZ1HXzZ^faa<p
z5x#P75OIZ^cxeOd7mc>_qP-24MO?xmr%{^=@;QY&&{$+>r2~<zmvUdKy|9N}re`Ln
z>X=nJCBJ0W@osFglJowDpCuX~6^VnI4EyN>z~~1%Kg?yb{KxNm5gxfyZHzM+xLGFE
zC^}9yu{8{BTKH<>>5K9M`FlM=!(*l;Rnjve3Dpli(y28mo~Q2nM}n#z)u0Ojd$t3!
zE?tGr&0ScHe*9rI2wG95FyT#v50RdXkN)=9qJ+E}P=M#WdyWFhmL5GZ;x=`%5`!-I
z!Fq{`axzbiZl*0UBFjT6SR4bl`_(L-BElYO7+{f%5n<R`I#WFn$)kfrQeUT1EX`OM
zi*5<t;@iexuYY8<cEw!HLPEv*^mW{Pcz4B0@@m0`alsF@!9ioGBbcvI0h&pPq(wps
zfW~_-mY&Q$ibW;CdsNG^DV^oBoO?Fi;V>2ug8S6khtCMJg(tH=KrXyH<+C%&VEK8A
zA!ERfeT@8;3%Xx!^a2d(EMorRNamnwn>xc>NGcJLft)&jtmopg%oMvH(TPP_|Aj*;
z-cbr#PD<^racpl+?Aek51W7Q$oa2YXO0J6h7vSKEg1QIj=i5(?qrX8CQa4>2dfwB!
z%*%A@wIcHFe_QrY#IQ~(6(hZ{iEu=yXw@j{`w!nvbo|2GYlYD@+x-S5-P-*I0Qs1K
zfjAc(uxGD&&GzvANNm<xtFimUvaY<vD0<)jlwQ$ld`fk+FUECaZ$tuR=iAO<aZ5vz
zWexQmQRD$Y5(j-uJBQzUXebRbW?mm6Jpi3{*uj4w<ET``c;Bc>?@EsU?n=+MQ!qXr
z&a0>c%idhFJM2cXst5k7Xjpq9P_tYeVs^~rzhudW)dx-Yyv(bJ_Zy{+uB==+7o<1P
zpo+g!+wkrKC$<EBD`BeUwNb-r$eAul_YTVp5EGv;wp8aiM)MFXLHvNrktOupFw5QD
zd%irNw#h3Gh)p#YPP;zG{8{Uk?5d!`u-13p$k$h*afbBCOtEo$GSoxX9CfmP?1Se`
zmfilwKiOZlyY0Hl_>O+|m@%QKEDfE32oKp1tM|<oib_EwDX7(N%$CeXS2FYL6IZZb
z)Z2oF-}@#`F94vgF2^DSS^06?zsYXwd!hc%BVPbnQ!jc02T3x5098b$;7xqUp}8cu
zQ(Ui@{k$z^hQJZJs)G1ig;)t)%>ru}ns4N#SEv!apRzihFRmLu-u(13PC@D2WsRQl
zILr<!9YEP(kqf2iJ<uWZQ6Ovp+0(R(trQc5?zs$U^01hx^ZI0?KAdT0iQ_>N@pFY+
zG~(Bjo6DMT-D;p^EV9Sp&A6&G<u%y*xl0m#*792Lq)?s9)aLrb8verB!)K0iTwi`S
zUO^tx7@LMgc_c^%jr31T4)L9tm4e`X-u0d7VlcE;`#|-`<BFLV+Y<%=R*RynY`JWd
znOyu;X9;vHq$?~NV-}*1Ry0|^U0(++BE7R3&wf6-q%{An=EHFPpTT&XRHHzlY=vK>
zDh6a(dH|El^L4y?(ibq_7RG~HMm#f?*faKmx~#${{af7z+iWw<v}0NK#8oS_S`HWL
zJF?EQw!>(X&Q;x8;Wac-?t%jV@%UDBS|g?}fD1H;2Z(LO5&l&TZ*VF*W^M^6tk$pU
z1JWr?K_N)Vo3(KPzGX6?S|`$&RNz$GREj?#Weq}%Z~mp!#QC$n3mG{){x(rFZ=s!>
za5gmvxKn|W&s%2M&0N?ZD+|oG9u`!H;MLQxu&aC(Q9MBmj?^x+28NCEU*cBKm^>PP
zEoSft3NML7S2_lA+Rat0#O%i$n?TKZn_=G5ycr;3m=4vYsbOQUha^b)TEz!#?>Q-(
zE=*sdSsfdy4sFmfiixTEOM9REjw<sSpX?tW9@kd|*NU<~vA+x?F5jXubp5%Ptdna(
zhcoaES8*Fdab5gb1dl9zigL~z2oL=@7Ftl{-bK_MUt#wW`TU#yPWn;@b>MQS)Jb~^
z$Rm2hi~2<pu$MtVrID!pMs55-sl$~THGb&dreb>5YxCOtZSbd8#?!*gvRBUYcne!D
zo!5<PQZ850?$y=jfwqjUtI*jBZ(X;DuXhoCJ9>+}i(M9&-NiUF!{WOx3|sE9o_tnQ
z?$+<nck9Rx-Y(^S?YRYg1nRIjwF<AhM@{>b6m^{1;MD5({H$f}OHO0s!@$){MX|2a
z2Gc)c@m4U9b@wewLm9l6s=!j=?8tt~SxQ_BpZmErjsF?b6li6xd4=H$Ahj^pFS?dj
zZt=A!^NSVS*wwY+S1KqQS@}K`S8CSQREBNiVQFH>w3LkP+j1rDR6Z17pTulMJb2TZ
zxnW3SxoX4d8)%ZRdq*nK<JtUl0%ia2r~0Iz^*f?iCTD0F=}}CO+_#(IzH$Pe!()1;
z`1XUt!KeSA=kHAkP)?mFBfPRV>GVo4klgU?yNQnKUSCX>2_bnI<P|j&Z<M#<3o#<c
z<E;VX8?d3R2Z4PKNxuWiQGos_XHVGq;hJQ(G4inDfiktSf0_^=l}U@(C)ub6s<Spe
zfz<wC7_(Z}h38ex#6x@c%ZAV2Na3sMquyvVjL~QM{}O5yUVuOps-x`{Ga(~tdy!i!
zVvoN?CScp`M>`>tbqe4&YneOh$SL|({rYZ15~${5KVGRVmKPzmawq*%On_`PY7EZk
zX{v>tf3}Hbuq`XRgf6aq3>^poIE(Uozg_Pd{g0dQJh47$KCECi#`<u#XE-b6(7fK7
z6>_IBr?(;eW~;u0_wnxxM-%8etiN6}0h&8`CMl<LY*JI%lVh@7H$DdGLP1vqX%(x+
z$5syd8sQfkTH9u2rWZHISij9rW1Wv7UD!T)sT{lc=1?+QP+Y2T?P{sDpD6}!P>pIl
zYwX%%JV!`XW^wlfx{$l@HFWT0JDvHBjL85X{nc9C>dcczH#oG-8^}Zd-3c<mjKn0A
zH;nWOm@RgFmbZ?ryD(Vo;>hCoToCbodhzpcZy~p@Zq=7Sku6{t)!lo7a*InemwI#6
zw5w*vEzoon%LqfYM_d(fZaFT){Xg2;pLu2X{c`Z)--<Dv(hp82fsC`+OQxObTuZMd
zpEc&Fa2WK3rY|{#w0YT|=)D+Ht30+*cvh+QPn|U|Pk1TvIAFBvaQ{3ps#n$d@J-9r
zX1Yk`-kU-_HgUbNTG2q4C784Lm~t}5`?o8E9(gj2|1u_5`b&S~+=xfs2;o5L>p!^1
z*S7H&olj)(Ot16wd~?xt*DOm;ACwv<zL4H50S60H{^8;eXKU`pwwWlPQ=hp&R3r0k
zU7TH32g(5AHT$!eyn6n0vQ*97Er~9W?dn8!YeT6%QqR~PdJ{+#Of>^!#}Hxu^*=p?
zPuTh~DP=3xzmSrJM%o@=lG&k0HcrGngU+=m*NIipLQ@H9*nsy<J)?#YbjbfPvlDOo
zNAZ}kue)zBTHWB|u8Psw;4P3M59Y53)k2`+)gV0WLT7*AwzJTDt3NOK`Lpjnvj16B
z6|BQqTm~WJTP{M-;}#WU$oKN0%r0i8Lm97@+6h_bu7~V18LM`ZnSA~~(mXb{uBmRE
zUUzoj;#{ZC#uO)y5xDXq$`IP`K-l<YeQMKcuk)^trkgLp<E-o9^>y8W{h?(ipyl!?
zWLIdOIGpD*sdY-hq@xrSr(oaAwthGyswTJV@7GOto|1677a|wnmfKw#FmyWF#@l~@
zk@Vh%z}DBd^X;}I=X6aLO_o4VP5=EZA@TT*P(?8t4h#f!e<m)={ng+s`t0eRKH7go
z3$A3q^mz-mbM8&8;XPFPApP0qIa_OByBg7MGFilZS6;EPb_qA>3~_tL#VrRl>ODT1
z;2PvH(?|XHQKR4o@6wknKj^Vm*n1|h#?Y(#Ym@PV3tJn#^X#T=xQ{LiW5c66x97%U
zAXrR1n4iQXOYZypa3@mbVivOU*`!EIvurjDf&UXt8)Wk9_&-hcCD?PR>QI`qzz9*c
z#l+->>BVyeA3R<_{Z4XT5zQ0~<h!e8Q?|)$C8K+XLWj7a{a$5AY*3(V#d7RQ!x^vQ
z73PJ8x)-x**Kk<X@$W?kxbAMaGG1|LItVWG+x6nu@=A3(-(4ckZTh=}<91093$MA!
zJhN^ebYHMrSSRsyY&QH@f)I1H^Zv1Li{X4<$#{54cejlDp`CQ)+WW6LT`T~>z5OdP
zJ(O2*B?%f3MjLR?->j~(5SEx3mG&!V4ppeRTHQZ4Q0YIfvnsJBEX+OO#N`;?QJ|`W
zSGM60xYvVce^%nD(XP_7m0o2c-QbqmV4l6(SInk7R=TN9L?V2rpgo8;MZo0AyS`~3
z^KCy@@RX6V%D3!<)nQAvdegi>yQ7S$G=gN{;gUg>aP*Ct&vl2Cm7SW>?+Zdf!zs0v
z4IhwRZP@k!3E<8%J5QxbuX5|VoKzq&Le7Dc9f(!5-vjqj%sTZsMv+n_BQ;;)x;0<T
zyFMe+5oF=ysj?K|ujLbkyiy_u^E&FQcTaaXXz6X~o76I0ufL?HURi%Da^!U`Y*v1>
z`grpYGlRhJavRHJY@W#4M>M{IbS<Or3JE9>^>_Ch%h%2j3Cl$`9suoL8&gbmS<`C8
zLCl5?79ynzpV{e>9-He+hyoDnxXEe$5$6V}n4z@!B){qQSk6g6s)mEP$Lv40%ebEg
zqC?}%XdG!EQXbOGG#s{&MuD9=t;cK~_(CZ4lJWy3rz^K+9M)~vj$DYX5BL3#>T9xg
zq52h3y=)f-fi6`f5;Vxyw{l^F^~cl>X@VGu@~rr6JMR$G&w+VV!^EX?nAq$J$pwob
z>y#T*YG?nnEmxsI$G*R5yk$Ur^Sugk_nhh)yDFYrA08)=VzH3Gt*XB>N&>&0I=WY@
z<N!VJ2+!RJYKC>KKnz^uUiq;1iUMUfBIQv^RUEy`^y-rb#vT^zqf&jnJkrND<IJB+
zD_&81S?*CAfga~faVN_ysSEtMJL9=^@!>R`r=VZ`xe_t-iS{Gwc!ah=4tc5f>pRKt
zbgtm4S{9YEM_d7zs-o`pcTZh7jC(^4taA9R?eULf?X5*Bm)OVXMJN7%gq2bs%DOoP
zzn)@_`{cu&OY75K(BsnoCPN|z7i2N5uPowe&2VbGFzcQ>-bz#b022Xq5OBp)Ne$ZW
z{Q+z^jX2yW{JmfMS_3!Oms*el0xORyg&WQQqr;&IDlaH8ai<1clt#044}UiNs2ZYv
zAUXq!DJjFh6F!b1>fw}*h6i8ld|?$TBVX0RQ?b?8HO3brp9Neb9FeFx`Vb+CU;2o!
z`aQrgFLudIDc<8h#cZg$i%t~Ojx5l;xB`D!;T(B*wNB(jcAJSi6~MP?JCNm)w9`AW
z4`{gL<(!Qk`aY35@jH!PUJj%4;;LI%v#L|<Z#(kznolExhP@o&ZINB4X7Wh^f)y5J
zxB;a~zPh}$d`HT2ym#3_)jFk`2@EEgXcMTwcUwF!FvM652ExUM`sLT4?5puk(86#?
z1c1kxQ>U`=wJ8^ajjsKXE=BU<<wtiVcaZB5<DPv+cKh(Z07kybiLwBmy;{W8Lzv$}
za=rdq2LCFgrf*`-bC@k5Bckxm_KFWqB}*}|wP_LX0Jdd46Lf~l_V5`!U&j%BGB(d&
z`KsYBC>`0}Z-5HB*27)Yibv0kyj4tZi_*zFAob0yI|K`1WvK&q472N$F}Mr8X%2Gz
z>hOF9Vj1mZeIetp{HfsJ3Z{wfKg8PDKl7fyjI7#*GGXo(k2sfq<pn>cazqI276?G^
zn793T-JQMsH?*CZ$rQNAu8b9=f-c))tB${LKG-YT<5XaB{@Onn;2W;4L--o_aBq;-
z3Abt8%%a-f>Ka*n%Z9K!km_!eYoP--FZi$HjsiS!ptJs_tQRedijkjp<M*4|jFtyp
zvGFWxYq^Eg6TY{)!*YebA6s1MV<66Zv&sBa*-N^H%4Ph}B>YZn_?SAaE41yb@ye?l
zlVhGx1;0?fxkZ+)sv>Ip@80fIim2DJwN2RyXI3F0!l-n{`u{xRD!Lzf)jwrUEp{J<
zTt|M(J-c4Dfo;@(YA{}@-zKf(w-7FN==M~Rb5FD^;Yn!UnT9idb}DdNHnf<eXQg<A
z_Q_lpj3Kbfmf)$5M`ZZj;ZqG|GlP^KRy}(=UpAKKhq5VwSst_#U)!ZRz(cnO;BC)2
z5BbZydGVRk!_ml)kK?M8Pf2L6SwL!QEyR4Oqt5lHI~2~f$$K(3dYl;|Flzu{_u?X5
zekl)Dk#OjG*wmP5MzThwF8X*s0Jh?MmBug5vy@jmno!2x1A-S3s4tT04%h2z{is0@
zkL@tJ$oj*eVl^4i>Zg5}$7jEWt|V8MVgGH%z*o#Vb^Api?E6)qzI@PaJ*C=HW>K_X
zAl^Gkwg!Owu=98R`4G&N57sHLUOsuC#7#C(sJgsAi5M^T;`+V+5Uvc#>LZR@RaAX-
z^!wE&i8vo<kb#FT+MU{Va)uZ5+y7{wE0t~szwpaprs*bOC-9(OWJ}ac#4P3T-BEM6
zP~^q#=FZ2*Ts1y`{o0#e^TJ%gK)troyYS0S`IE!;sZmE~s<q3lK(znBJ<X;XY{Zoc
ze=gczRyWGW+|+x|&|-2Yi1||`GqrTBY?%AcUx2wsO-15a@a~Uf_9WW}qa6aH$~BU<
z7V^i$YqCr#{7$S{Pu#!@Hm{xrYN#F-2@o7CggmP9864CtYd$VomwTFIy-tpIbFj<2
zm-?iZcnz>5$mBSm(w%#kT5?}el<#uVVhwuMIZ&C}+p<FwqBO4Xe?5$?rVlObN_~y{
z!cIlyxPq^k$|UBXEZ^_XNekTf%=LROEX8zxq#k&r@>d2)%F%zuSM{eUn0;ShkM{Jr
z-_VP*uQLNi2&JQpBrOPi;_%)LK3hYTw5A?CV@mVl7RAZ*Izf3_1h6nVvS-nJ81N==
z+haFUc-3v|-+{IJxo1LQ=xDA65Lb_BdIz<TaIgW2o)&0^3UW@?r64F6fhFdODWgrH
zRgJ@IV_ncX^w!}qu@Hs4**b(bWA7|Ig=~G;=`v2B`iPjaJ|b3xU`e<&!=qKh?q-AY
zs;k~n+0I8xWnTX7lBR{Zl4tROI}J0$I4fAdHP*(_x-dB8-*@R^(K&v#X`SAFau0!5
z$+S4Ab2V+Kr;rQUkI=)GdDR|YYh9ge+J{kMleQ!sh<Pg>5mUtcWOULNYi(Iz?-VQ6
zJ$kg_;)aVP3NSOh3Hl;i_PdLltQiH~uzEucEZZA)?<t;Db|R{H5fV;{`Hqqo;3cd7
zn!0DH_fOJSEZ=Bc*ioXg04p6j(E2b_t<}&Y9Y_%4bfS&cY25Do%egPI)kl1!CGWkY
z@#>!TQ4_Ao%@r<%zNE%Kw0iQ>BbcK`lz}Q&mjco5SMr`(mB*(NXys(XnA+5iu7K;h
z27eWoWZYF4#BIh>k5|TOcVEN5{t2?s^5TGok6VK}(!cG`V!c`AkhEGy0z5Wj_diE>
z2G#(NFH%ac|AC5D9_I{jz9W!%>Lauba|-g!9?C4dX)Dt84S&9?%*95!vH(n7;Gr|F
z<-Vg;XKlv9dHpme`m(AR<{<z0CwYfpc)mNzT+s8qPTyqM^5%_dNvqG_dcV#+Tcegy
zQj5$YP@}GvJ|E=0W|b2nAaq(FFLgH}flYg8HmO%Rc-T_iz2Ab5ikHdx)l_S?@-2Af
zpFhwSW#$?YitPnDSmr#FA@NVzPhQqXzfn<zf22Kr_2^34VbxC^wtm$;K-e0lZ~Aq@
zXnwfAtJq&KU?x|62U5Cuy6;}8UtSm_6|f`YXx2}62-&i7H^K04{ubt(#7GwhmH9&>
zXJX>Yqgjp42N1;R_OGZwH_r^{`b(?QHu7pZ7mT?zIh!kh`8a$R;B;xk^axN=yVX0I
z!ZS)t#aTl*U?ggC3!$^|hFOT+406};eB#iU!oxXP(_I~+GnQWf9E%ecBr813Ls$;O
z>z*fJs-x*-1;%ocGz)$MI1YO0^6`3^6l1W=C0rVj1AheU<2^O`8kIdD8@{hkb<m>Y
zx_NfN<VnWFKM>8lU)NdwD#7_PJ`gIuK1gj*WUJXTX)jz^IF`__#1mmMmC&|fY*yV~
zL0+4zZ30`Nv3gm<2Zua!CXnpk91#NClRJ|=4qQ`vHkpzwW*Mz#F5OBsb0nN5yZH)y
z`H)^7bWp(xQlx{q^-xVJHfk>LMwn&BG>JN`!gpl&SZ7=&<8!uWdc^Y%rvhef;oftr
zPD_gJh$5!>g36qY{L1!VXHJ(i`C$?+$y-(NlZYa(PYA#2VY>G`+Lr6iL!6K}nJaj$
z$DYe}PBNxim<26dSGIsjkUc(4mv$o%QFJ?8fd?Cd%dY^Us-gu<t)ZiN{i-EHa?pl<
z_3!5sr7K%t$MzLG><a1chjc$KrfqHM>6wRwor7EEg9y_#*t`q-ieMPzh<M)Koi=c=
zyu1yg1Z1IGnrL43`8O$V&nvAUPse(HoTzQEc?hX$h>q`S1<>-hW@zwZ3B9(d2Cx#d
zep~7oX9uAI?-W2r+Eg`iQaj44Ht4sad&kxCtYAeG>wh+6zi+P@%Ub@>bNbaiSFE*@
zJt1BrkJLaFlbB3SUwm`RTQz_4PcKj$svOjPMMGN1`kEq!q|VS<DK{Q_=d_o*rWY7a
z{1U_Z5#GZ1lz3bcwBN<xyzb*092r>2Dq)>p$X5F2=L6!t{wu0=*5{te?c^kB{IKKv
zh5gRl6!0$Zs@nrUXXSFvdrGKw+r`a<ko!RXv~rw`nL|g7QiO={@@++Cq6!m}y;;&V
zD62o3UI{*W`HiNLpj^X}rwR<wVrSOH?@YR<&DV5v{%bV(D<E}+my4a@8K9E+u<wPm
z#9m~#s<o5VYXV^;*(n8@`7ni6Ti4y(!6}%0%utrwXpXR_{w&&tyY|evc_41M%83=F
zYki;;&0$}>0>7nPcj&iaH&cTxW2rGEk5yyGoO?}!yZIM{mJKNb!U?#({n`0P2`g%Z
ztcdz)4u6E&`Sw4~-f{bZN>fTCjr@O_E(fcYhkW(<)0ovx?0a;|Mv05XqGS+6d~DJ$
z=>$^eT8LqD9A;(U8!;x6a89V{NQVEmhK(r_zdcS}X6^yeUM*uK&Xr}>z%YAO{hw2w
zz4k-5ffQE)0v?;$M3*_eVe-bzt$zRUc=6?q_skm9GK=Q8=NzL51lQ6hmhb9AG8Reb
zej1|UUEi|z{V^xuzmpvPJrZj4hxL&fy)+&Fy;L?(s<l>CSh=T2o^1b=<Tg>@vi@WI
zgli&}*0y{2EY7l@ZaGA9tX0kIFn!ux`}xlMg-yTd{=Btf8=qwjfKa9e1o1=CX%8eR
zarUoY@Y@z7X&jN@5br$N3EMEHEGKaTkUx+~a4pC>?|2^enVCrna*>46dpTp^V)Z6~
zQ)nr}B1@xAJ|1lr?eW!EHDGeO*QUE?Vza)gA51`lXA)$Mm;X{T`Blf%DTqJKD||Qg
zY*{0TVY_JP{M5f=si&Fv`aEVQ8R@5o_1!x;c(k1c+~^muoJmWF6gboDCje)bKqdS8
z{`)&yUarE#qW$I03-Cqu1vXudnJJ)dYRFaVo@~bSQ{*=^uK(Lst#%RdYF)R3)>z&0
zB+EwqzPNM$-@B!yyjx`lf(FkL3)!G0gL4c8QWNV(TM(*wT}V94@V;4p-15{0Td4B=
zTJp1+Z};GwMIfv32o+DK-!y3$^SwePg(GSD_v0%QAERhP#LS!9bu6aD_h*ufu?#G`
zBm~(IC};$ho5;P?a#Z?{Rkt_(==Gko_bOl{GXu5KJ@4MrJZH)moDg_rZ1_^DyINQ(
zRiV0-9J9sC{_S7HBXQx+mLKG(t7k_3ixNdJ@;@a?X4Q&0(}!~cIv-bq3V?Hq_1=kB
zzJ0u~QHpqHra5a>WqgkwwAV2|y(&%X<nri+lSh@5CuJue&TD<gp6PC5h>NXKeT?l3
zN^NqB@-8*NuiN>yE>Y9%FmEfSE&u}KfsQ<GzHy|P>h#gdm;b!3Y8Qxl2m_?~7Q@6N
zC+FX1vk#n5s%+;9AuGQAer$gA(xKBMIag-qO$OQ|Y~DyyWklsC{mKqt2G_wK12q|X
z-)q=fvt_PSL6VrJG`}ag>MIXA6ZB>ab2%tK*7jZh#h05@mtrRTlt#MwrM}chwsud*
z(95d?{8PT>G30#fe%{j8M}DJgcC4;~?J@JmH4V$J+K1Ty#u1{~i%)<?+61^ZpK<wf
z?Sdl=Yy~C*T89y$mN(rtoS8uZ{qBAy&ojLTH_0jMn0#yFsOEeyHfiXVPiM)28c7mY
zRAwwwJi|ni67?^cI#*M87E*#ZiJ#E}=QdAjN7qTKf~p>Wq?DbppQZ_978ar5tb@N5
z&$oLwxs2e35ibL6WeGvzWqbmdqDu?jt<C#pLP=a^8cfFXOuw-4As?ZX_yxV^w3n}9
zT>A1=1oz5bmzbkir1xPTh&R3oYJ?~9adFcn&7wT0xrf4<5qd`#<`p*+E+i|q(eABe
z8UH)^-G()~q~VQzIyfhw-CU<y%_)I*Urko_i~QKXb)g@1)3M4ox$C>h$kW1$zyH8O
zGCMR%kLzc*PVThKhvlL<E~~vl))FR^Jl*yCdjuBN*uVMvFG_Smjx81fG-NjNB-(B9
zw~;I9R*Vn#)}wcS1;aHAB%-L#O`9W6=Nu_)=T8ek0c*XJG6TvnWb?%|lp<_I_UGD9
zRj`shJV%!woXajjqDEI|_Uik#lxy)FkXVGh5?a<BK+Inq(PE=1q#WL$e1-`=uGX}+
zO6Y*h$ng7AQOX35nBJPkl%L8cj#_KGL6qSxoMWWyM~$Te5OMGB`uzlyWactU(FI;r
zRqTa6eC1aNJib3hVc*g@SsBGLU+5gvL^$uQl)NeTJe=jqE9J%m+pAYaeM2@T3zuXZ
zqm5n{t&u)QCQuJQ=%-S8|LHcjaCFRJ#t*(6uUAtAGkwFkjLPojOE)snyH)R2{G{TU
z;=)l0`yN&qSI0o&bBw}zdA-g2x;TF=c1bBpef9c8sR|fdx$6H6FC311*YqC0rBF5)
zDikvMwN*lD|2zIos#?v|w#Q>X9oa-3#ftiqv|*mFeguEHn`xxF?0r_SbA?}eaOjsq
z)7_tO!D)EBguc5;x3wB2_(oRmdt;KmVKqQ6mszW1)yk-raBSP&o_P8J)PbvJV`~U!
z<`h-^S{^t=4_~XXU+y{B?YePCDljo{7gwXs@Y4aViPu|>{{m$E{`*@}KnNG*qm02X
z`@tt%r%$iQ6^<B|RIseWPWfZrp1#Z*SnLT(Y4IosdJ!aQ6%xrLGghFbyMSf7jz48%
zXEfx97a@IYysE|jD$DI&u`=*vm+;=aTZJuiy=8W9T$1WP)sE<iqAEN>MI!A#H>16%
zt-{y>I>s9hL{-pQ4~LQ6WZR^C#g%SMyC*Bvom>jO^YOCOTd=eB^)JAus{P(e{W_Ma
zrcmxwESA}tAs#x~2`YhbZ=9!{hh^iqd-LhN#At#&CAcS$C^=M<6qgsglQ(Tct~(<*
zOS7z|E@(BS-YF{tnS^#~=7V|9SGSt6rza)>|9&5q3{1*DW2|zjORo>X>@)>U%@(<B
zDj`4jl!zQ1CyA-z^I)jx-jOEJctYmr`Qc<q$FpWDjKzA|bi<x8c%tr5^royYY3}q_
zKkbpgD8S7rF7=8Yj(uKL???S1R(XN{Bx#Q?aKFv3DnJBY>%-K#Eo+vY@<y9k^~Z3`
zP*m;J;_{WF{W%1Oy&wYKcTV?#-5#ZW@Y1}N#4k$JaHkHiD$XX4#U(GvwyupNxg_}*
z^SsfKK^K9Poa*xSnmx~Jmj+~KHgq(!WvX+;^k>KD!2y%;477QxI{I(^Y@~kFzWM&u
zNVZ71y`7?-r&-SDjvH1P=0owdGBPNwgL{V&3e2}%;5T(vGdUk$EQ6Gy^;#AmWfI=I
zTS&w2o0<Ii_O2xhKs&of@NWLY$2}fmW`CP5?|(F%cQl)S*v30`sOqF<T8b*6_LeTR
zwc4N7CN-mil-NmVYu1X<ikTEOT6-lZqV^6oV#kP`*gM4Qeb0ISe$Mm9bI$!e&-dKd
zb$za=4~8)ajVnDpuIN2)WSSTt=JIe4+>RW;?Wc`zzN8;7&2(RM8j9IF^(^*o2gKXz
ztDz`|nC#!ZIQhE^=aZ$Kq^nHreV)l%Bia74w83G`Mf1}M?=zC>tWF7jDW=Q0`JE0o
z7cKa5(M^wVa{7fXzI3*w6G@^v!7_in?rq*t-_W>Z;-8r^z|VedBfnL#y0_-B!JlXW
zV~?ED$1Im7M9g&3y`!p|E+|V$n;=f2+#N{ZrrZ`_-r=Q{-cqs~p)(|_n<E(#Aq00w
z>T7bq8V%!iSLrpCu(_LZ!(@AmtDAgG#^boFM>^R=!xERB+`(MeWj`Qg31S9h4$IPF
zygZdWoHl}z(#OKOr6MHXgi0JG-tbLUKVS0FTh#sY(wP^oz|afIQGKqzpZVF&2GF74
z?C>P5GnzR^c8K@9q7k|^fG4#^ulv?pb~jZ`5>hgMt)CCT)BYI8erK0}0Qy8<_QHK*
zAO$_&3YA-4ehk$w_!N2N8;xDsN8%qzXvXyxJn-o`834IAKnS^eovT!Z&xVzyBo02<
zcwv_R3T+`=b4xV6&v2})%^*He`BA>D+Chrf)9@38oob-&h<v+jultfh#V{OvB{&}q
zCWmjFvU4GSfrCsl!M8Zd&P3LbgCqe~y~0Ib9>d<0(7sslxusEDTr1C|&#8hin$bp4
zj$>M=&s+`jA>@1CTi!l-0DI}&Z72N2;NXDof|i`A5J&S3l<aZP^;hd{stIr_*@kwd
zWZL)OO&y(4co%8ty%dKsXx@BSLYZTB2EpLdQ7K@B#G^NT3%-VF8R(sQ&mhZ}pDK$g
z`Z2Q_FWqEUPgP&GtL7G49;n8m|2?tv<gj_ar6qs!$$<y3`u^oMXWJyFFkxO!J%F5r
zZ|Dn*XI3t5KmE;oEQv4h4LR1U`tmK47r%uNKTn{47|Fi<+3OK8di)gaqe#>duUB&i
zJs$M#vY9}3UxD4`f97r)TvK_oGS~S1qaHyQtBJq1@gs$^rp1d1p^@SwWO@;fpAY1&
z^SgN2??+QM%{K9ub5u^MhlDdkhc5&)#3#uZxQ&+TPN3bfuEBx)ye9`NzwK)JI7_$w
z7Y$8ijTWg*_#!S81(R}Hc%|iOo`ZZJCy@9Q`*1H&3+KOAaEFR>Xcf#>_x!rdDGBD<
zp88B#j=35f%{Wp)v^SaiF+xC~!@(UyA#MJ`9mcBD8)`N?U<|ivStI|m_2#)I!6c`u
zB2-}ZbslGD6W;MQPLFOW?|Yo;pGBM=O^M;w9z;OO>I-4c>nvB+xhioq4-X1)4)gVK
zDJ5EtYdtku{hL7yb9)BIUkePkl@e=w>2Xye=1<qhjE-6*o5Eu)zE0P9d2bKUXsc@Z
zJ8iwADh`~fYHz=hMIQpyiyqv}DtexCWb9T*`4fa1{$a@f_Ga3iN0}TSpC_TeCf38D
zv==J!Ne`bqTHkS?t@6?SvWgy`Y^HYVwoHvn>Ctzyor$Lg9D%)?XudJ=n<_eMnYg1o
zx+Yj<{1#ejp~>$a^ra>T#J%lzL3uqcG}blYddYpdycSo?g>nwg0w(qAV&B!uy(^>T
z!EHZmsQSJ-ENX%WpIeuEo<vQDaI^e*IQ=Xu@<soHV9@thOV+n^b{iM#Sq(m)Cc^?<
zXRrOdB<OD=jm2MzH`h?*+-3iA#fs<bLz5?e71`4|1MdjV(7LsPS3+gm-X-WJl9Ie-
zrS$J?aY@KzI`Vc6Xr*yID~Yv&`y8g?GPB%&?2KqfdP*+ig}#_}3>?L~01bA{lp+6y
z4sHC8hDsj%wo}-J^Z1Qpx=m4UVQgOP$Kn8qCM(=LV{NI~fv>|-rTj1hm3zrSyU){b
zwDag)wLr?O+aLU@Yw}fTImyu=@<XnZ*F%KIld_qu5(l^1G+~D(9dqi=!v<=a&{3C^
ztBN|!;7RU`owI&fKq|ZEpmqC>FFk%yoRtX4j)z*=jad4)qZe$G^F1Z8@lILvZ98Wk
z(6rXzX)xJzjMk=bBZlE{W+PF3bI4C&_(W=WJDn5Ep`ozWzdC@H18BR^;yBHyGGD0g
zAeN*S@#N8kxiPyu@oY~EK}l16f>7T-C9?uM!TIbal2j{TR^+6zCcm_jh`)F1?l2_+
zarW<<L(Bo<7n7|X<YfI#;WOS=njUfZNrZPmkTE^0%s5xG0x@xV{Z@;ekA7a`16j^<
zoMbxvD;P6svXkf840#ifuB_GSDoV-zS08DMV1Bx;efM6B6xrQFJg-=8MGkauJ*Mbg
zKpMXGVVWN=1VU=WSRASM0_cw!Mif7>5;A2J)3Jx*-H-K{6)9`M2rF$?M&f!!THd#k
zRC!Eys0mwkfGX<O|B%Q7srBfL&cx7KbqMIohL<FfGq*G4mx{v1XK2=rnIcNxuJ9Oh
z;1|9}7#b3QzsyrRoXz_9HK?bH!d=Nr=MA)^aas7hPW6F0cbT|vV?2HgsK2<N<Xvvj
zW+2$1d(5r&@mXx@MFq-!#sZH*|LZyBy$A#K!bTPq`~BPD8m05fcY^R%@hr;UzN(j}
zILWVxjl_zkh}subq;35*{0kZ(BxQkWQ3@}EIB<>a#JU+7>T`@v!3T1aWI_E=l3nO$
zX6RH(BB$TC^IV*G)|a)3?=MJVUioNxnskO;Y8tBlsv&|F{MtPI^$idsG%noX-DQ)b
zaeJ>dH@KjQ6SP*P_kPjK!+gSBHY*OK`H^=wh@D5fVTN?{;4<@h%=!;2N#-<<sFXa%
zPcg-7pM72TMPGVKpL_D0zx*L+j`QFBR|yenA$<`dJoivL5WKtm4r-DwyeDK;{Y*tI
zr#xlZyUkShPpu#Wm@+lPbw`afj9DD&i(Bj^EV(*Sd8$cvW-JM#`<L*}rrN)NHUWGp
zPC9l!;~LHR$E(U5%5_%A8(ZsXSrEPvK@fDfirus0%QSS@p69Ey+;fLP;Gya3`S7Q2
zEjqSO>QxW1erSmr=pdmam9DD(pSeu^X4DgI{Uu(K$pkAvwOsB(X{3`1AtnZ?i}9l)
zfEY$)WbHwCcEY*pG((*m`HXbHFH)nj_3?zeOCs7y8%{k$+KQo0%w%!AHd~d;yipoX
zA^GP0WmUuhueO1M<q1Aabl1$qQhW96#9#A-?h~Kt+tqE^mOt@YPE_LlSeElU?~*4#
zQ|)JNz=`}1(EY>N4h4eeuMe;;Qr5rQUnxY#c_D^U(pf>Wbn}aXrd$s*^)f%n(4gFS
zr;cC;C|Kb8n4^VC&F{6KOJf4en8?ApXEPS@-#92-0qA%=(ghb}|AyMIQ^{akH3C}L
zl#vYjT{z`W50Ef)G}~MITdODR<Ct|uVV~=ls;Te82euf&`p*XLj$5{I@}6!u{J~V5
zQ<K=^5?s~asT=D4PdQbr;*>e8>!=eoue`a6UvR63vq^$SM?4P@cDV;bt>;yli&8Ht
z*0V{`TEfUs^*dX$m`97bJ^uh@MYj%m77ZLWcZzi2!KhYl;=7rqi*s%tSOijwc*3Ek
zq4(Z|R;><tR&lL2$>yd)09gHl^BDDT`K8>nCO{#dS-bt;K%2k9pQ2sOOQrW`uD3^-
zfuk}T5^KL#pSq<uM^u<hhSh>tL(1!@CZ8f_m|B1C4+1RMoHCLggs5(}@$a*#4q0ZM
z3tcmZk?N!|Nh;6-?ijgB$skG?LF12obw|vI{o)nCa5G7KK0O~1Pi!rcIGTOs)TSbR
z2qayKR~r6O+P|>75mbwlaetoJvabpxWd2M4dO^PLP9po@Ya(_e%(-m5-*c(fDQan?
zfM!Ju$=-~<E>W#`m5?c0pp1cP=Nrrx-oK4@>@_RDO5V_ruBkw{FOD1V&3kF13l{y=
zu4;FTBI@4&yRfTjRa;BSGrz~%-{oT%h*{1m>7>vM`u(rYj)Q9+pa2KJSs<ZQ#|2$f
zGL}g<8&;ZzU|~={m$pSiz@St4N6j5LZ|RD>*rZEbO3bq1sNaj?RsMa6ylTEL*)}Jf
z_+^{Fh7>}3TQ}EsqWq?kE=ge{Yv#TF9LJU_03U(s$VOV02%(3w)}(_S`s)q!h9Tj?
zl@TIJ-iZ>!h~NH<O%kh@pZBq6FY(>{LIs(5DVrZ>>aye|{|V?bWIcn%*SUGN)s`do
zp4nXSw_`Ob&vP5tEMzR<vGp=X>M}n*8%1KuHvfq4`eKU*-OV_Cb!ye_K;>6b3wd9I
zG-yiXy|sGm#kQb;_!>=L*QqF85kuZJ#h9nhT*W;g;_FIl`+a=EIe?aDS{I(nP79;n
zXI#+td^D$zdMy}U!dX83lbupeGQpmJ2Rbw2qaFp`a{JhzxwXzS+VfzuZB8fPt1~q;
zKy}~xWwANqE>Ha}iO|?o<OBT7?IhLH?>Ow}-=VDs36mH$?vu9a>*qf8MPG6{w+}mR
zKMQ<ea6(v$UdpgCfa<(e>9YNGTg>~`c+Rc{&&?k&>&P2JkLuQ~BbEG1cyw$7UjkJ$
z`i|DJ6SuY^#S#n#cL57_s-2uIy8|WDA9uQS*Z%^LxgzsL)HAY48m7RxP;#t)7HXKQ
zXKU<MX4aDM)bb^@?>t(>*|CmBgQnh1$a8ef{5;halmvY9LUaBKP`0hVS{vR}Y&tI>
zCx1jUaM8wV+?4D-yO6kBbLxxOcNdG%Ukr<;`c?%&M-%hWaB_Xlhb5BFwb>L?sbBnQ
zud;SBIla*8C9e9_Kqh5*|0QXcAEiH`#Hm^uG1WXKb7b7mj@>J-PpUWShKvr$P~7}*
z0v>`4#B>e{Ff~|G!)&eH=};TzIC83vkCcp&ZyyEDboue7f)Tq`!@|XG6BV~JW^E#~
z54RVlK^_(-y~W$Itw5yH>{jS_jp9Mq?1-TxJAxEr{C+ssfbrihqW0wqxsdn<Zd^FK
zR3_pZY)N|`+d99oZTsH&i(f%vg@rjvJO3t7BDiPraVv<!!tD*GZ|zHg%7gdwte05o
zz6eMk4qPz#xS^3isgn{hC+r2V4@H60&t5QhYOnpQui^QYk>1-FsW~Q|&KCyv=eUAW
z{F|y{+@L0UuD9V&kCss(Z$t3$3+U%~RSRsvwEbFH_j{d_a50+KHKX3H{<oxUPoRDQ
z@1IXk`;6mz%!AM0+<!FR<RrzVb8qQmbWrSCi+k3&2C*pR(&ySd0lz+FHVL$rjb{tY
zvnsw+L!-kLYpsHBo`zXq`+ckm^H^~LA})ite_s`;nmH2lMr$R7l|Ws`1sO?Xh-Hc}
zC>8V*t1WoeP8`ymLP^}2EpPuuxidtLiA%YH&uNknmRYSvHLfCQom0bDpcg0^kG(?7
zY`{UXb|4F_ho+_Vhhkb%oX%Y>8T-!rW}_X<dz2C!`|S~sEl#F6XT0yjFz=MpRL#7W
zk&CgSv&w&0fAdL|o6K|16c2kp`$#z<1Bi*g*vx}Ev$MfTKULc5)LeKVS@x^d6Ho??
zUa{jfQL^K^W6adaYgr&<rlju5R1-D*Fc1kw4KfT)gHo;hso*S46seeVHk{1Bnkl*Q
zDyH{s?#6GS5-UGLAVRUJ(}R~v<izzBqOjh$MG`-#yX=pgeyj>qJZRS()cu(E76Fo!
zyq=bkZ7|Kkd*LBUP<mu2$yIC9@`94ykbPvGhJM8`0wZ(2t#rE6Jm}0-R-UyzYVax9
z_7ft8n<ZgTnQLw68onN#yjbq^UZ9X${bQ@mO|h`k*pb*4f8eW^(qp1{nk(<rW>SUu
zebOrZqJ9aEb}9Lw+%2pRZus%<=AmR!v9o8#&Brc%Ba$7^rFv5;)8HUeiR1Gj!T$!w
z3#=qUx|tKEs?r4a6Pjo!V{iVc6L@$o3OV?+JNyP`VJa)XjYCz;R$Yef%4gE9UPLof
zcW>dYJx=&fC$N7qIa<#9OV>8Xrx$N{8yyvB-+rWNZ&8J8)%_+Hf1}v}6q5{t+Wnm_
zS@89Br^G6(VNVmt_1Vm}`YqDufxc%iDl+~vPr<)A2qJm^msM%tmaR0aI{d(fdU@f_
z)y+%bv2&=nb@Ca_GhLDg-5zrLT<*=imi{lS!>Z9ij8EV!tv}zqRwpp=>pIMF6@I~=
zAAI~&1OJn`u6ggH0~9b>g*5*4G`abdTGTEUa<i+&uRa#LBUP(**1>O+nxGKwJhGwT
z=Zy?TbH@Cc&hD-I7gxHlbD4XASmJX~ep>+lc07+NMeNU+>*h~UJ?=>+HOR8={7|c*
z#qXo=CoF3TU~$ul<ns%F*rp7_uGAKjNtXY)^=~83DeSyw0L$<f))3zJa5a~beY~gU
z{5o<c;y3RomEDHx7uwlG?kJ}!<hS|h9s(I=ku9c{`1j?Hd^djxiE*=fnYEOK6w<B4
zqMX1PQJdM=Y;w_Fgb$1>r|=1B2e#~q>|aSN^<sn4hw8Qi3l^8W=ipUI@Kk2`rX`{#
zo`*`_Ftc%41apfpY>-C7H$rUDgXGqRH(;#k>WOT$3Sdsb{NPDEJI}5p7pV`!;0E4&
zBO`Ir$)`Zi?)ee3-+2PJ#;gOBp!1b>kk7z&4il>TQ!NB+*wKb{g&$;#xoPV1y4<CT
zVr@>IHdU!yT~AvUrX)oQ9;$Jk(OM%cMBpz)!dNe;HLCyuy-TQF%d;oSSa1DpXI=eG
z*-?xh79OUrx&}8>I92iaaIR}P#7HzfR7$n}YOgfFn0lQB{+xc$Z<}=?SRHa)&%mZo
z86Hvmzs8KrJi1`pX;SlY1;FtiHKn^-O~zU{^8p(iyg{JX{PM~<kR0;f-0X6tz7D+B
zK2Ug~6L0vp=eblEe$$8*`K<)DWze?Uzze;b+;1!PqLVxQQ6Ww9Q0q_cV0G_8Tujpg
zhD_oExP)<CT17$IhT?h)T84sit1r5J71VlqG5pb!;2C!?uUKUopK^b(F{i$Z?kjxg
ze{+E`(U8;6i0h5#wZ-Mr`I_$Cl;8}F1xxq^q9xlSwepG{P{YHrR;3V@PT$16nphQX
zC`HahNft~dyI#|}sy@mgYu;AXS|_&Xrus+~G8oX!>~htnoD#k=hiHCgDUX`!$Sc`<
z{ueF2EUWNd`F)l7^BJ2rNI1{nVC^Nf!%M{Jl+#?(kEA34P7}xy^QSlUJ+3RAK%rBw
z*4}*743zYvoy%!VRB6vQ{(uqqmt|U;e~I9Pcc{0@6YUHq6>rT!ve46@8=DNbYONCl
zDV-gut;168G3%-of|8xTvTI&K+JO7eZpkg>!`~`(j3kBhY%oPSc%8{IJd6ld1NF^+
zMIO2z07>ViS+@AEy7S!(EgtMhsFAM~Ga5dTtK20sA&3aMOgXJDGECmC@7H0llyZNi
zV44*gYgacSfbnoZN4eGuN?f+-U<Oe(ep0>Rqlp6-4hPe_AK<<iH^h(in!->1)6j3p
zK>RmQ+u#wfSesx3@kdDa5Zdh`fXo3YyCXwHpig5<{7F_cYj55y>B&{l5SyuUh(~oN
z#MV9aM@k%}swf*Rvo+Z>Z$8R?>c;&0ALmStHom8Zj9v|)H^kYyNp>7v2e~C6JL2ya
zohyryYHdeX0|m%QH^VOY6_#!JmH?}^-a!&+h%AuZpV@hSf26qXlX(<(vf7<pf!q&_
zZ%mqnq7pnNq=Mpiu(q}t!p=@CAtyvOZ-5SGz=$N5;?`0H5dt@HmA>tkU$T7J;Tx61
zTX(d?AlEBOt~v3@Y(#l!*F<G{@ar!QDG5GuMzLMZ!P9%2n)IGg{wJv<LtH?`;fGmg
zDoM^;YyN|FVQnwCh1hazN?1V`ZG@Jec#)UtQhmx@pz03dy`$4J9Er3BgEBbNmpNCg
znbm#UZ~2R)dx7QuvjBdC^o9ZF4Gw8&hh0m_naM5lH_>cjOPZ4EFB9l@H~J){a>LH3
zup2}KaT%g?Ca@nj!dVxXOPmgqYs#Pulbg(NOko(mxI4Y1i5zS1(^k}KU@oQfoNYNn
z&BI0)tW&}gE?+Uj`q<?vUn%5*wO`>S)aSb`0*^{QIs9!o_n6&0do*M>I?dTchVK!|
zwc|mQL6M*iLhhcjOV7EywVv-jQsa-Fq})V$4Z%4I18j@`xi&;L#mQ8(OLc}@&1^20
zh3ZC441KwG$Gkd-=o;5-(IPCxN&CQi`LtK3hM4Jnu`kiDdgZqtj_V9|A-g2oUWAyI
zRmJPr8rF0y57utF_1j#`sOEiwRDG!R`l(iJ{=Kr3a{)|I#aE8U)^H}xeW2@S6*dtS
z2FV|i+dZu#I!D5uxLXQGjO=F93TNwg)r8b{!f9X+i7LCOIb`JY*h7W-9G|fKkQ!;H
ztJx@r$W(?J??Sor{<m!{dLE8$d+_4;LE(;pf+Bi4a;`np>qCkXWO$d~LVBN|t}$x5
z>%sv!;>C1e0wKOG5R39D^2|BdH&R~ij*v%3?V5XrY>VsSI?gqjs3^bcJKGXJGz%{5
zX71SjfPL?{17%*nFdtHrtsosgBG@ShDAsBl4RGU1nZXGX@ju*LkfNH-*|&LQX$#u=
z<xvMy+P1I!wl|Yy9QR-wgi}F^<f^L+&H&F#R_ZKsXOuktlj-WxVMEj>=EFMRpuvcK
zm^RoW9!ayCQ`i09>DTwQ5ser$zqLBCgt&w2t#9M1G;i3OYZChtMFOw7p#SkOF7M^H
zKM;W&6vZd!<4=AnPUq=uA4JAyC})rI8u8XXFs(?k3V775i_L;F3J4~?H9eVeA}`r1
z^OMiBU+Z6di8#!j+4A5KBEmNuZD#9odo=#2m*eEzr{~tbGxu_CQLvXz&sX*_vt(_#
zDQSMr+cio0ftn-3K;c{fi%V_Xg(x+t8{yB^%!eG#Hw=qsy%65J`sVq*eNw-5Z)U<{
zLhGIeo5acO_}gcUlla<uvj-Z#eh9RH)J~s$AZz)onUKxLY5@jbvS{hPfZF4~X>V=y
zLP5wpBVq2&FUq+o(f-~hg_cMCs+w_`UDIst9LkH(zcNuZ5{q*cYtsWEZsBBu8U&!l
zp~jJ<>XB#sTH2ytB*3%puY2_PreiKp;itXi_&~DX&ovjvU8q!ii#&n2wMS??DDSK(
zH>aYol)WLJ@sE*}Wgbq0)_;Ow(1g9R0e`Ffw!-zG((D4b*P?Er8h-VOVN0(vJCL$N
zC5B)KrSXyc`61*uP_lZ7f4)-2W)5#hD_Y!Z)p=S`nKP<!u-z<tuwg=G7jZ|eD)ryW
zgwamXb6C@0`b(zhvW82(frN=LhU4&S)kFo^o2+(Rav$4bg;FJSugkbqp0^3_^nJ^?
zkDq&=JEp#T4`nzO?lm^KhWRh5+0D?}%xCOR>)GaBMbn+t7|#z#%o3pMr8C9{ZoAU2
zsC6{_dtL!SxNPJ`*eRE$AOn><RLlK`F{L9~bN&3I<o02ciuyM-BJyo3gWEsR<%f&W
zB?~iVeki%gX0`^ay%&g!m`s@aNg!C`tFgn>dW`qeV0Bh$J*^U_iu5*KIsd(xa4Ex#
zj3!ot)G+stLHLf3?;uSc<3_$uGG>|gU~b|r#gAwFhor87Pjbn7r@TsL4sMsaKNCjU
zUq9b>K<BP}LU+B**Sq*z_h<q;v_4`uVN889#glW(ajyg98*>&mbb(cVFH=Lf*MibS
z5BSLyONlO##|N-@)Ku(n{Jp&N=~YlsO|qu{pMU3L2@&)&3agsi+HFQ+QL4Lc43}6~
zRN3Lmeh97FVJt3bu|7WP-~|;gYbxvQ>tTYHL=B8+{r3Mk;g}nQ`s%`3s4=R78>Eu=
zfPg^<CHBgiERkCGx8Z*IzJ+|$-KExD)76N>($s;960hZh9Dvrg3m=Y#a$&ZwR`EB*
z_*7chIMK;>Ae~1sbm$D)31S(ga<bh{|LXP1La3giTyzHMfit9mN=Lc&Oj~AOj182*
zM7CtIn=L;3T{DB9m^ERdirm@Ilo)Xi-yOvW12-q&C+$7+qJ9NS{A9WH=@5{`bLYR6
zStnJaiJxN(c{%Di4NTYRmsUd6;qUb?n{BR!jvJa|rfFZ#GQyLKfFUep+Nk_AIkee+
zz8(C8OS4%hf|FX!tnpG^ZCV{cPZ;;RK~YmV-5X}RGr;kPJ{NSkF>S3Rht+wc^-9C&
z`^=c{QIa{0)>^7Z)~xEnmsOZw2t7e0W`}Lr9ozgSPc{%ISUg5IcRhF-A87Zyk}q6o
zD&UWE#F`j~SJ_@5_`QZiAh+)d_+DGtnYiPd<ByCVtvAU)mpvSFY%uDa#%sgzxA|N2
z+uX2siMm~^RrPBlIjvCF%RU&<a0?Bf<-jKi9Wm8+w+aKp&&z)ERyI;?y8drnxpLxs
znytA{OAAL0=>rJ!jB>*)o@md2w_2qZ&JOx@(NvoU@fSr10B!ZGepl6aZ=GJxSw_GO
z#oPYVYJ<v{k`|clF4Dk!$c|k4(Y#v6srL&YMIDsUXn;1e!z961Zyag6L)g2a&T0h3
z2I~nIH-0xnF5hAf|6*!!jZ<+P>aky8)K)JKeVHb%k#4zsC-rN)+(|Hu6Oxy@D`HYh
z9(?ME2KRa7OETne%zuoQ8wb;`)Lfvx=otJP&B}@3p7@dFlUt&agkUa`pk|yMYs1v^
z)N&1N@y@Mr@rZBTygxTru4#k!dD#KSrcRtTgb_^S<|Y1bn#jKiG;+`|Hv$`flgKG?
zsCw-7XSOfYlaisnap+m?>pxVQ=gKqdE@nt-Ht6*~UU4RuO|M)|sk<}l&VFd3Xy!CB
z9v|Yre*_?eTE3gZF=r`!;2GIft8N*@-hi{)lB&_ji)zcvD)v5(vBv{8Q#{XM0&ZIa
zA0V?c!cQjBVrn+u7ajti@jHjmWTAfJ#LU8SVAqVHpIPnl_jAB~Zk65LcF4FHp-@d-
zsgR(WV=0J6xco?Urr?tHfAp5<xIrZRs6s26-joHbtzA`X@jFM4Q+&rYyk^gfC<#7R
z`dxJyJ52zD9h#rlw+;Hw>9f_BxmOMH>BbI;f+>4V;;grCal9H^eR=C3qD7Y-=+~;|
zHL!K-FY9ql+NC<!H$`?_LuaQBaINfdpbEE2P4~CtJ5Q3IR6O==>eE}XLu#wRI=0t0
zGmrSuwo<@a$I1r3z<!^J(8Q}vx-AB;_N)t8cFyySRJA>g*sD7)sIT~@?C~rgBZvOA
zV1ynM7^mIc)z$-M8=FsP*o6Gsnd<ISBoG8?as5i49LsL|XI*-yx($0%s!-Wqu<Fl7
zeb6n}Je?y+Xx$%{cSkDo5P}qxUGP*=dO03bS|s_G%Ndlf^kCn&n=tX?9@j^=iAqvW
z{Yq2sK}{r9+&%SXEAs`ur7yS7Ma~7W(Yv{SMfraRCxm7<p|3P?Dr27<<{>*_wk-@c
zWmyTtLy(RbEv4UfCr#PIg*W>anvYvy&Z~Rc$=VSyNXY`WzQXK-fu3C#Kt{^Scp*hQ
z)_-#@S;WKcTc27~YEbFG9OZb<$V{V9co*$|Og3uZSeyRJe;v1+O*$oV8qsdKuQmjZ
zqhX@8nu_eQMTg6!H1+&r$A2T9*UmB)ern5Omt9%hEI-UMjW@+++&Js}6ta!1UGF+y
z?`sd0{bmu@=5^~nxo$^;{}*Uv#EH@OC9v({jiiw{#}D2Y|FC#BC%JcMb_+yuINEBZ
zZ5^Z=bZAMG8GHXDI4^5FTxfq7uKF-AExc^{U)kPs9E>9RcHrM_{-cOR?`G-;K9}^y
zCdVIS%3(mE2y%-}&EGjWf|gUhhJPodI`R+&cEgpf?tX8Bz`EP(GU0Bz#EE0xbu&hl
zU=>sucx+ZCmjYobaVgr=rp&Y{J-}eISG!s)_YJ?*drw?79j*k7s70aYx{quZk@_oA
ztMghuYR@@*uyF61Lu?ST4S!gPUY%b#Hi1uGNdcj}y;wRw+>u;${kYnjox!EQI~VoT
zecKHyG5o}X&Hc`>ix$;-<HS~<a(LsUdhf~Rn^jvaL;9QUn41OEe%NfW#&R2LVdkb#
zFW&qVMVXU;C&*(rO9i2^<e3(QKyex0Z&`Lr*GcxfWy;ph;PX;fINTjn&_Ut#aCPNR
zoA-4?mAJwRBO3dJ)3~Ky7}!@XFrHrd7{$3A__g{-msPI$;*s%?iNtj0sg(c^lvVnt
zO9sZiN#~XEN0-J*5(G_DxA1giP6nW`In*e<DL26QaWCPciNh2-Y)dwh!#FI2d0Bqa
z_@h-!^XSOm%I~40;8nwex2;knYpE8;QdW6?eytcbe@m8e2dVoz+zMu&+>{C?@)nZu
z6QY1^yk)&-fm{5nmUELzY^9uJi%ba$Rtx>?N`-gjwv^0)b_?T9cS#KlzY34vcO+0H
zQsbZpz~qhWcdOrFR<yJ>f4WicZD+CWMD>7oGkn8(h+6WHS|#K_iED_#wj4hoXoL3x
zb^UsJ6ciONsjsUN8!o86eCYUdbR$!P$Et4hw?PMeTt9z*{o5f)G+rCkHIpz(1>s7_
z)q~4@E=u-QmP!qCnM$e+A>0|=US0l6Cp$a>dB-4Xm3@^_=rXDt3}qIctg1m9GdIQ~
z5FC2Vs{feGf#i4Pa$xdaE=%mE8X8so@{=~s0fg!%kK}jj=CLnwv4uWB1dwOe@^$!j
zXfWGltG8)77oM#hPTI&x^WHzw{80`CMD_#sCr37<S}?Ae**swuXM<u~2W_XcYoq+*
zT&{C3wjUdInmO5W6(Ln46eEMm+b|3K`e`vT2i>LIO3Vk-z|pbLNQa(`Rv5V)-4i}K
z<GX*V#zC-&?e(@#S5Hd_#xo-ph8H9Is@>S@G4JRMw#$LH$u5${?8Nde=vJD(dJbdL
zQih@j>=FOojF(VO?x@^yOf<65zPk|lION~-`Rt>%Af7_Zzm%A_-3rd1{t5#+MMhlu
z)N?iui&ru#fi1kbYcUz062p<d7F+8CwYWLk9HK%XpM?SMo@uNEs9mqG4@3r&&?_Gh
z(P$?lz=Z!=6FSkb$o1znaINpVq+o|KPW;EsHcp@DedAmAVTO~Vpxw0h#w7Y=c+2GP
z_&dCML6(i|U&^<ihfA#;XXM{>y?GuB9l0R9^;XAEpfxT02eG8<;)0^6RZsR^{SzRF
zob6er2%>4MUje&3oJ;M2BS&`F!PZu-TdcUx8sxsIyX&`f_xSOD`_wO4sW}HYbNn|s
zPrmnf7gNO*&?K<;I3s)iY-Fsf(85H?%3NVLT<W+^#5zomdKMMdVkcKoC@z1w*k+Rt
z{`Jr&I@oG-95G7%@A;<QHK+BQCxnajc&m;JqN0zN1#fcMF|1vlC4-fW%%kVhHN6^1
z^|6`44(30+DUJt%c`x<C9+;U-ni7LzJi_;1Vl?yn8g7k$rwi_|^_~?R{hHt1JQ&(}
zZ_yF=v$iM)Wn|?+N83u*QnQb^eQ%U;i?83+#(1&=_%nq@H;hWcQcXSI3C4{^IlBSD
zrM&xn)qw@EHeZB^RJfRYKV?*NW2w9bRD6rp!BQnxK<$Y{`SG$3!CK>Omqv=|89OSw
zy&u{{F~b0A(PHh}igLOoxFT`2BC*FIBP0cx%@(Py{$C&25BTf>%WMbg(dq3btqbN>
z|3nOX7yd{zeiyjzJ}_8|H|&yLyT$HI`<qQ23QDwl=E!uKay+#E)nQs5d3~gn95Z23
ztyXIdFGP0eKf3d*IZA3Y;44i_g_6I^$<1)mub^qWt48ydyXmLSpF;7SB!nhX1}bm2
zcrz`B_tV{@3&W>x)T}()b?$QRs*J97w1OO|&VSdU{+J1KsLz4cgyFMtCB+Xe>n3ci
z+d+TXAr2-x;-^O)GvW=pC7OP@F3yjY?S(yY6_eO6$&%~@x~)#2yuCN=)m>igMsO2R
zek<n;#{M(lr3&GHMd!>)_z0yq&2l4#UbE@f_GD4-6z<zP8>_7L!9M^-CYH_d;*qwR
z)f|Iz4qC`b(}?=o1oKC=x8{I0%g&4+m;CNM@he;Pu%0=a8gc%bF$6<1i#(cX*<~ks
z?~l-HQBWxi!1|4O?D2kINC?b_-6!L)7puBI5Ug?OEgU(_;)gUke7ur9%}F<(NCHq2
z0KjA}mF1pzetLWXazERjant}}(qY%2@m}nBFw%QFdUGAJxs91qIj;_%O|>Az-TZjc
z(5aD3z)c{I`M@l42VR~QU&Pm6hbBT@ZKsH(#+0F&A6}m5R=J2)IprZ@k<Y&Jie-gl
z&xR_>!19v5o+4>)@mVKip=TnVahYNaFpB^xrH5)Uh|DRRbCjR&UOYVZc-{X17lV=@
zaaAau)%n@WV#YHh2nIwz^AJn^dbJY<6Hx~x5PrV$^~|Mlvy~OA(W&G#Ku25pcrihy
zs^rHELP6Tz?g{(9t;K4A`H~$ekNhb}%A>m@{C!?Y#q)t0>*TWUQQ?2oS#>q)GUtNM
zS6REB`K@tbU|Q10UTcLq(yhFeq&Pei_;)>AX5md1g>54@e2v}YHuf;B<9r;*B%t;v
zB#Yer;q(O{wtHP9LMNCa_)mac<juB1(7t9B15%Q5t<wMeS+!cW)*t~^>{$O6Y?V6-
zt2te^_)O`;nF}b6%N}RHB`b1c2hRhTN$o2eE8~bi*FcAxp?~OVVyciyrL*L&x?a>1
zl33J}ZDxVUXgc^erM_Bg#54tyqMxFcwVOaJ9M9ICyHIJC><@PI^KjYDY&L4yu9nP^
zDWUln?3;`-agGc_^>>n`M-@gLuK5iDp34V6<53YD-}2qmj`)vEE7TB(1<b+D4t-XC
z(AJ-LlG+Qw`5&~sJU;Sgk`lMq^p6kzg!mu+A)5E7le;s0h}ji5<@n>}<Gn&`)G__|
zkdj^2X!I_71#Salxzj&?gUf;Mt!Cx07^r@veWuQjZrwtjcWbz!n#~Aq^LuKbFyyXQ
zrIp<xKK3@d)AX%FR=~R};6_modc`hluSlT@^(Qi^b?`x|@q@2Tm(+@9*1Fjvf-Kh!
zSY_T^s72mp=lQpt<Tvco%g!AxUEDh3dHRXo;ev##ZO-0PV{`97ifydFL1>&T8{wt%
z@z&3);gVa+vIk;)S6)yK&h~MJ<rPOaf$Q>%LNoCGeOc!Oe43gFR|c@R&REy8Tu@u&
z*nxXkv0w#o`c|9uvTC`YH!P-b4aO~V$MxMR<<;UH$6!@f%C2R1Yu||VkkI_}F79uK
z)?S2E#aK9r)5@8Ns*Q<MBP)@OmpZ2W*ce~5HJ7!!NSQ5DHA(|a<Ndca!3G`cO=U^K
z5&Ra1P_n+W|1F9d!;o>5DZX)yoWD^(&+VR#@neVH(pIhUZXB^wCb6nvW(!rGAmV<=
z6W$dJB31w$bwZqeE#iLarn~B#tK&ox<R?U5>q4f<lDa8j`!;l^WvQ{OUW)V>7^>Q>
zv*BiX_bG{)BBWlP8ZpYI)Kaf$V$!3gT-dR6QJ=UV3c#kX3s^)WTbiw3iNCHR8$9f(
zz6(l3vwr@hAfQU_?URz}0>iAYG1;VAZ^{GP56?xzrAmr!Cx~~xZ_@WPzfV0QI1>xx
z@cPocx^wE+qx<@Ad-exYZA@`Z_t2vwKc7o4-0d(>W^SP~F0dWA+w?1!r>=1YJ`h&F
zJyHph(qzMN3VudaOII<ztr+`ysq2T%iaKE@cVrr0u@n?L=XHITV}^Bf4>Fy?X>Sst
zRAEL*u)iku#9)u;LSrq%PjD|`<uuE1&s?dB_+!i@R!LG~{aMxS9Qx3=04MXY>&lq*
z1U&Lb{Y_@|#L9~{MXfEeyTmwv^rpbXcP{R9)I7q?Q@`a{%@2h3w4MA%7n@(dXgFrq
z_JHt#ozoX<EigVZS59+IJ!dxSBEu(9>L!cH09dMp`Iy`XXYZ8HPiNZr0FQSL2I44g
zH-`F-lp$j%O*M+GSVq&UQyS&tBlG3f^fRfTzICJ8mr=Fle?M0IV4S(SirSXZ%UA~}
z^?6{!_d}NTW#e4=^6p*Lyxwq%6|}x2r69lPXeIjb>0j5nN2jKzZ$nRd1E(0%Fa8Fs
zs@leg`t}dq_bQlIooY!Bf1-rDvo9=uu%Db(eKio4-UgtG9>l!2QF~Z&-rkti|3gdM
zV@(C~%X0EH&i_)}eN=_)(M@}Y8FzENd9p0#dRn#~);$|Rf79EhUq2vg50!vKRHV#m
zCsgTQ4+kr&X7Nv)lxoNLXLCd$jl!sN-@L1rZlI2Y2K%c$U3~XE%su?w*4}3{FYosT
z0A}+%t}IRUZ;cdIfje-$;L_q|P(ui!@vh5UP}DXMfT&#THM@gwK+G>*WZVEq_VVq!
zbLcaFsk82K&3};oI9I)@5sj;!dDojkdGru-@#?!(fV3WI(4bcdJR8YEvQs-d;`LFL
zB6BRE#4JRV>szBKjOaqi{6jG1pEGa_YF_A?r(or~rR-7v8mO*<tx;*tm$$RYFF^gC
z)!Nx(J9QlVF|)sc)+^>g+~XepuIgPG;qHrvDi`*wCx2j3t&bZom+IV`Y@3&Js%Z~_
zK{v@_D7_)>bwmO=Yd*z%bkGv1XN~90i=Mf!H7Y6PdSj|kdjwk6y4&mstsR`3zD_X3
zZzuClPQg!>97{)Mq}y-%)}-?m!7A@#CT-#bIv2`8%Xgw3dq+kHX%De7Zt$2Ex95p=
z=RY~29wy@3{gda8THDWApS7iAf4uTu@j5Ib0f<CCBy!Fm$)Y7$jDHe(#z}j&pw8XJ
znZu(iK<lDUjRIGja7%wLTfhaeJFdZ-_m|y-1E94PwgQzuz;{z<tV**FU*Lt2XRY2m
z*FSn-uAl5-+^MBy(tEN5cPa&*@K^)SwLIX<-QrPK{A(TUhZb2y|CkZe82#>uJpKJ7
z>a!AlJv4?~`T6$#BS6FX`IyrSKj53UejKhqpWAVcCbzfgZ#k9hi_3PNu=6U{3Vd63
zidRurbyM4^d=umcA#)xrF%g;0LVa$3&K4)<*|f?PA-W&P?>N&+`&Zk-Ct_6^Ryhis
zO4vIYQoUry{U(KU@JF48jQ_q2q^phIm`{k!m)cy-e;J%K4qYC53qIx0BzWYMAtX|Z
zG$6hsAE@yS8v^qte13manse=HkP6xS+`C#HBNp{8u0Sb6wZqaG?C!W|XX>)M2bF?E
zXpav2yKO$SVBb0HkT3PiLXS*J>y5uuemybW_xMP5@WD#Hy8V?AvN+IeLuPXj8GW@J
z$)T(Gn_T4bciy<MQf5AXgt)zP-Rwwd*^7VqFp>rq%f`%EQ#Y?;n_E5DSx~Q?OaJ7%
z&M+Ht$s;%9pH=?oHre?QwDn-{`f6x?se6>&qWE%)Yr%I?MDl2h8XvS^_87<9Ca@o`
zdHHWI9M0mH^j-GjJp$2ZuYuvk4kDtC*9gSJl_>(u-*2x2R_ig0T|GV=1@JSc3Ypfl
z{!ffi+{u$%7?JPAn8Na;ixt;57Txj+GZ*Bm9a9gw0Z9+~MMoh_M|6hj*}U8<B*+7G
zi={jaL27d#_U-SujXlmNqo#FejeIMY%5z=v`=QGG20$x~iVhpncRjyt3YCKAGr++W
zieI@N2aWMw0y?vwEq}+sN#71ya<cnniC!ErH+O_Bmpje#bo}U-&R4Zf99fMQ0Y=1+
zT@)L2TC^W|;b|*qL*3!$OKl~%OS1tM(Qa|GrJ%B@Z5NIc{<S(o7z{t{#EtUdOxn!n
zJkYj$abG!HWlY1b=Zk)QkkdhtiO$;jTg!WFWtc)_%;_xbbD#Bg2RvJ4_1}-B-JAlQ
zMqZqXVr9>oxf4pC`b&IH^IN-ik+;oW=kYiwt{y-<ZTE1(KEw*?*3y;q>sGRp!Rx;^
z!_<Ke47>TM76Jc@L59eKl+%4-E)jp*7+1S)f~lgOPt{$@4$k9yziK#H^n4U#Jst-u
z8!p6kVDQ%h(0gJ29KV(sOs&TS*Na!x;&oYJ=WSp?gzr3vdwLicos7QY^LwcMt;+NU
z=P&r&p`+Ew2;#o$(0KyA1`lsd<P?wqnk8!39ikfZ<g<~!;x%44p>vS~TagRR!Gf{2
znYrAQvQ+J0Zqr;U%C8IKr@Oj6Ylfzz@4DNXmAjL>4Yr%`2^pIh`JXsw&3|wA9cR<=
z=vu1GAa=*J6Hzp1TYB8ym83TZInn_`8S>Jv`HLjFSCe37dxFRK1;IF}lg%`fzcK<h
zS`MhEmWh*|S!jh8D%<?ZH<iW=Vx081i!Xu)cdkR3Q5+#;cNu?u>`tpa50rjj?D4Xo
z*y9(oo&r5z^QeT_=N3*6l4L21x(?-#P=b8e7eEMNZL!OAt_nkS85bAWF3r5nzf1#b
zXcZH(u}p>?qIbZ<0_azyU2P7ZvF2tCEEcuD`*wS|^a`#DK0eiN;;D7^94tH)8?jK5
z#(dCenze#<m(dsDG9i*R^^1S&rp|7jmvR5jYnEllb@gf0$CxV$ectb`>A$GU{y2$-
zl_WZ=_%0@|bil~%eMOxF)*7LbsXc`<x=t~EQPWWU+|7j!-XHBx<S36D;cfN{C#vrC
z^%O5FTi^KGU|GlE6C-`bj^+OKK|@Q_jelVV#`(rg4AUhwr6Cn;y<v?=HZT>7vgj{!
zcN`gED=<}Obgl@3@7PZi%MZ3+X)6Br^C~E9y+86!)2mC<eaW-@P-?)nw~VCM&;IV`
zao(P2{bGC0U{@o%z-J3QbW=xCeOUCVRV;;5()#rCaDjfy_F9#hJ?}kBsT|SdVs9J9
zq50+i#EUK})X__8c~j&7bc3*)k3XKy$PVf0InTzb^hPIVgO7hOr+4-%;O8v(&DN_2
ziTJ^}1kbkT&ilpon<Iz5SE5G#v<yWxZ{+Q6arGdTp?(t&(*@`A2pjo!*#v*5ww<V6
zHDA^lcs9L_v)GXz7Ktc4G#rzB8MWV>w)bK`cJq&y3x6A!q?tu#oVC!hMG&|{%!b>F
z`991a^7RtG>+O{C&uha`zx=wN9RC*7R&&8EHrQ+lX;j`v?;<Def-n={Y%gsy^4CeO
z^6#G|UCm+F`#!L$A-tSwV30BvujivsuRytJ=i<J&UyB|@J3{;~1~4@#-b0h_n{H?F
z2o3^2f>b$?%f?{XoW!~gw+6}0?6Vx`_0dSY#cs23hQYHH`tMRtm$UV{jVDojr=MCF
zow?l+ph<5|X14OCFWt8j;YQ9{UEd$6O$jPYaQSXC95b%w`oasXr@s?=#6M`M?Ap(H
z1tE1G5MHtF-csga^$8fsM&V#7djL$0hNmU|G^*+P=F$Og33>7<mhnI>z4EaTj%r>z
zsjG7F^c#_*Uu@$l4t-%waO`8NCm{+^#bEo>jhwDO<QDYK5{KSj$IYl!?ND0r(Z|wG
z6=@r%HCUw*yPs-Dw4F;FHgL%@s@cb|*o&(@X~6KiPkd&Nc^BR`-JsyDc!$@FcF%>u
zSuCNW#}^f{XYn+3A%msx2q7^Kc`vKgP1xBNTuCM1Y#Qz2n7kImR>&xDyo@HbE7$p_
zoVwaZ<{lI}v>?B;&yn*4jtHr5yL-@9)_4D{Gq`6BPf)I_?<cHXcAT~r$YJl`lS!VW
zBzCip1zatOHql38H^_;284j|bzDalJ(bs1~DYa$?KH6r<H`m0yU0G@yU28OSvVmJP
zzQ0HV|3VjSW(1!-t5;;Oh08%IA1P^EW?xHq#9$VoKZXlHL)U3F*UTDj9=9fF@5Ua`
z!ppUhmH3r>H>8*fxf?SB7<14keFmTYMrloSfhEBAAb?`t_qx=L)x|x#5>~ZpZ^^>w
zCtV8?tT^IRuPl$WJ8hf>I`Y^TRIh95%4~jP75xIYzL{X<?H|0aV9N!~8vWw(@V%U!
zu$wLH+GTSIP!tfJ_iq2@gGQZXVq+=x=GGqte(bORXj1@2k%+vri2ZjEUqgCWN}%EM
zLxR_q&>08iEIs|`ZpY?G0SjN6(V?7YSr@iwa@MAl;bJUq_(m8BmC0Ja|G_&r5A>HB
z+Y?Ki^-_$YkyLR;=CrS^V-}gp%HTZ+3a&yzOR8=?n`(-f$Udl??e_{`JgsK^KtGOB
zj@#A7L~+%AirLN3I;!f&jI0cM%Q5pAgsAKZNI}!+2&By6Ft|0BT$r^PFK519IU$;T
zPe9iL#+<EGzXWrbRM=PlQ?}uXkAze!oy9oGKy5rO?a-!_^5wv!VmHJ$L>Sb$dYK^w
z-GB~PXAhB)s5yT%uDz+F(Ga3Hs(aO+x!#~cEKp}%9m)7;Pt!sk!9Yh_Yh>c2LP+<h
ziZ~_p-&z>o@cip;?Y3Hn;!%~;;~{1GNjz`&zssu8Wm+U{)ZW!1T=5I!t#xr)N}u5s
zgk+=X9W`z|nYQbW?kV<Y!`Gi5PqY&oF;v<Y_8CZD{UzBuZZkh&yL!*%xQpjc%&ORp
z+O*LSiquaVep55n^H^e2wUfqs**2X*UoqpJwC^Ae5A_k8B+H%EIL7C;k_sEQJuO#E
z{pn+ON)~X$ollv=Ve75-LLDpaOh<rh#1E9M)eHH{B}Q3SBsqd#(LY>9k&+jcC{g-{
zp6I?Dwo`z5k@}z;#C!9^fwLsh*$j8%qR)kt2VcaL^6F;>aFhS$^@j$xN)+{y-ixay
z-hVteuka~)ohNC}S$1u&PB)lECAuzqje4DtVf>nPWJf<Okfj2`(X&lBuP{oY1FHlR
z^Nr1BHSwm^cJXF6!18o%pQ9V_Dg}MSEZN><__jpj(6PB&ZJgu{8vyYWWndjO_NvDm
z?Ia2z;Y!!>jUROrC58C?p4l!JOObdBymsa$W1(T%gh*(25Tdc$#^b9$WNNZuV!xzi
zwPq;4V#k`czG_&KU*TehunmFlYlq?HMLzlFO}7hOdfR1MYfZ6JfQS=T{|$?{FG0Jz
z62qlI#G|#f8Qg^W)C&pS(&#?12qhx)_}eVZZ=F>oEA<b=#9~2-Jhrnnz+*BF6C4&Z
zrfFfd{14<e<*O?=`|%#AB{7E73K+=<4_WnFA8$~A8C#@k1@OH0?b79AMICi<Pf2d&
zbK0dz`41Y)K=7Ck9)D&bKseox-XOHoUB^AUb%TOmvL<qdEc_-|?@&$ggGHO>HPkN!
zvwaVChb8w@Kb*KDSdDz;{I$V%OgwF6I-H&3Mv-Sxxt8$_*>c34u_DK&RQh;29rU&<
ztX1|CM;iFLF`I?ag%#R;+v<alOK12nxp|dpafE}WKpW>3WP)IyA{R*`QbHl+5%Ad$
zqU=}RaDXW-2HU4rwo#_%ChEO;Ch}NMw?6tj;oX+Q=G#bG1+Iw;WVtUDK4ZxH$Jl(z
z!|{Mmsgi!3Q;k*!(UC>pzo0KaDT8y>_&qtV!B5wY)`*3FRt&Bh{Mo|x(5eE+s0(r8
z9HpUuWR^X#1a5a_a(msLwEgS^A`p`9F>91shqRpRkLEp9&O&`oTBe`@DL&5HI0X-&
zzK1(v7J5_#`;(y+bfmwEgmid+5>?uPoOp|D^90JMhY5;x*N9(yOI*Z)V`P9ny*UH?
zR(b2DcYvnNQwc^L{YT>dU7jQtv;=q(ZOh#U{{6>M<#1J^3Q%OW9(GxoR2l#E?1v<C
zt2~Kb&K?sRiR$beGoPFp$F{tE<$ql35`QiF@@Gurh_^uiucfj6BSG~hE<cAa81#VP
z;dOn#GFwyGHr%`U;!b+$noLwdr8J#`F0<iHA9HGJh_M;rE^vEx1oJ;|y(hunIA(b<
zf6Z~2xWB^5?th#N9-wM~M$_~CEF`~O^(+e&qFrf7I4%d4(SYwGIijT}W@M^5YxEdx
zmg9I1y_%Slj4M-tg4{fT3KC((!Xs$V(caoTz#r+}XMZDy+l*%TFx&m>tPzK{dQUat
zm^sDne;6=6X{OgZ(o%)v`+@;O<9sk>TJk|=9sQ^_=X|AmJIQWlw{=$jOCIh|H0Ihh
zblxwl!i!ZcCHemZ+YB6WZ*s%S@x4oAzx-{pJ3Wpy*f9>@{{d}4lE21c50Lv?^lO<*
zbGy}t-VG~G0&5>yn4Ik76YM&Uv-{Jx#HRf$nJ%6|(<%U@shG@ySL7y)Nw72$rsJ7>
z!TZD`B*liSTmQwYSEgq-?09;Gk+-LQP<UZq20)K5H@G@5C?#LC@S?>YpuL@3mblx)
z67EskAB#*M+*VP5ZUYh+WzocF=pqZ9LLXni@&y-b=)w%#{6z)y#6y?yW*joDo<}@o
z!vYM}9M6={5w2?RD9km_owYfxVuOMlSg&mn?UTK44u9dlNUvAJhR5a_^9Pu2Q(!Zz
z+xpUeqo<_tQM^6+Bf%31zSw7TK#PYpH1n{Y7q!gS($C}5Ym<x2+moAn$hGzjwckB@
zE%#LY(bEuH$0N(-_F7CM*&eS)U$b>B{$77t@}jA})YleHi<a%w$2@pzI$6KfK05b%
zc6)T$UgJyd(t5`s*iCW$*vjuy?YCsN<<q6PIaB{u<9Vjf*T5nVJPmGVx!rS$ExFRT
z%g2$<h$sDCeTi-l^9*`#JKT$1$5;1Tcxy6#EmbatYT3G<@p;QYeV$pz+0t*(GVbTr
zuCMQV{<97ETPsX}bYHp;Y~>LflzLp8c&8tGk7--NP;RLI_WSP^D&j%R?S;?^ua@1?
z&$e3aOxuFDMdnOQu(W`*W;@^0Uz0tH|N70FHw!G@t;>PyNI|Q$gFAwCM2DU7M{bJQ
zp5p9e`nP{yUvFLPy)EMSL8S!#UE}esH)j8hJ=yd2?a#yXynFxQ;g#PQyzLFrp;IR?
z^99PW8QOQEKH2I^-Vdifamb#1j`3vw_q1CJ2Gajh^Kj-tO1aA0muRDUs}H^L@dbfL
zXEO0Bt$Z@}NMz*HsXQXoo~CXmL!6nis^NvFdP*i)u<4Zv&ScT`iwR-Cv|{uZwg~a*
zLmbV*M8Tf?3LWQh+$*XycnzGb$OM)?CtD)Su8S3Rf?w0rpWH7nSqBn+h~_#rh%8cl
z;75M!=*&|+Dth%Ai#l4o(PEpTm8@#$uWgZsG9%dq0(L5z3zd%#c43}vi<cJ$>IEjh
zKp?jTHx{qL>d2O`ztq~+Q>MkTd&z^3|FjW2&V1gl)Ldft5=wneza^$i&igaIa>&*(
z%5~PcW;^d2vYj41%Y4r5J=ZJj&$#{=mAp@6Px#ti&9$X)jDP=OjJ*Zt9%artx4iAX
z^M$<~V%yl|ocpykuavd9632CI=~?RYYWHp?KGl9pb}yX|Yx;TLw<g;f(IYx7oNLo`
zE_qEmaV|MN=UmH|#{HUp-V=76kKT583!iIW9@(^wCClV<-bb>Dz#9<ia*H^#ud;uO
z?LfAU*0<OOZeinS{FmtQR8C&7(Hu~H33gBSPxiYz-@jU}9rR@X?ax0xe0cX^^{Jko
zey0EPzyI*?4}bj^nRIHKuhIU~|NAcw|M5Tmcc<b$HQ4LF_|>l$KK%Cgzw>XdE{Lf!
znmn99Jmtpp?%msm-~R5cEpohm`1P;8x0^nE3;LY$+A!-}or4@GzyINfbV0-oyL>t}
zTN-`SPf~B}M=-Bd7=KFJyWm~xGuxzn=7IAm=k3vajpF|t<2jm#(YHpvraT<UU*4bR
z?du3uPp3CNtzDeoL@J8_*|#3H=zC<P)2s2zpIGs40Hv(J;4X*X>41(^Dp#=D7FL{R
zbYEaWzO&d9cS}N_y{SWiS;~U1SL}f+8R}x+^tl0pZT@lW@V@)NGc2JKCaj&XC9BZl
zxKr#1-YmJCUFAp)+qC&T8x?<Y$Y{`2gvSlPmn>QphCe}5UWZn1qgq(eyg>&@)j+5(
znr=j*i%VL#;TxCag1+OL?nfIhd}N^<zmRvI?SnQ8i^S!>RfI8$3q@1@f{h>fsiTk>
ze`-9&Quog#Pb#S?(v#Hqk>}qYC+;PNJwCT&fBpThEC>6d`S;)b&bN~m-X7XMt9y_y
zlkI)k;4;+fX|q4m{;y|+{ZDJr8hOTl>-7p>j{Dv3e?J`mypMbH%^Q1j!uVAj^7-et
zZy)~nr$4IuMb$bETpH9Zw;6T|zy6i&AN<I*-*b+o?>~Hac=ztz!%si|Y;&e_ezPC?
zNi;@ZmU|M1%;=@_VUPYES>@x1j&yHX>m2YFHb*%b@owSUt!(IAe9d0Y$5x#+%%l9Z
z(<Jtj+24H5JF=N|MsjEKT6T5qbJp?dY@eU;_Gch(H+hi<B+UWU`_b1mwl4C3^f$l!
zZt`gzc`Q*c^1QQmjl6j{?|wHgXus7WPaeZ@)R%L2f3@ctljp2&MkyU1>>Ff1{`ix>
z`b@No>qg+5^zu)Ii>u{fAvllZkgAU`i~c7+k$&~<asq?%?)^KNTip1Xh*xej((uUQ
zEhq3J<6kM^Ip;C(&w2R&*?Y4k$$2F~FSE9+z0T3;)>w}&G*2?uV>Yj?JITgu?(_oc
zqnYlb|F3<ol7H@S(g^}2$RHykGXqtT*nEQn1OYezmLGtUUPwEoD-ArOoGpEg%fAHH
zai)AdZKT5!kqCo{E@%iB`e_#ZaT>N@Z-xk3bvOr^^c~N6*aAv_DktURb%nJ6?*-l`
zy67VL3ZiisN;;nWL0n`<EDh@)%QK?(HZxj^C7C@BiFCMLG|e>-vek8Yk{;iWP7xq~
z|HXGd`J_KghCVmUcst41C@<0!PVm6*%BHz&JmHj(a~5d+%g<%h1f&!Y()p=#6p0`c
zbXY77h?)i%A8$P7qJjx{`FqhoSp#3?#J4?152cOPhxk8b!Isk&*3?6TFO$91sdGk?
zQx}f!vONkLw0?fAe=yc%tr>a9r$1vMp8o8S+ypSS>Ft5=e*Eb1)i>EF*W&DZp>;_5
zd<+3xbvbuZs78g>=v{FQ=B}sjp#$2*S7{eCjK|V9_{|TwE`83b^I6`DsW#|W^kxZZ
zQ`|biHR!Z6X>mD;lf-Gu5B)h3k+|-BQnrDTrEKA}(+{~B<F*^JS}S3ohE}8l;Ev7j
z08uAVTguS_Gs)zZWUlG2*WhfS4zCj*_isV7g@*HOf!z|u=TKYHywEzG22S<dw3Ik+
z<9o`wrIDwlY+J(Ysh3>NN1wlZ(E(eZu?_pYvU~dD&>o!-li0l+6PZt+K3#qNCX*Ri
z0A%+srz@U6d!FgiBtXY}8_1j)d2Vi=T)lY4{fMUTT<ag@xEpkD<f=9X*G$MifB7m`
z<MrjraT0@FJ--%iYG*?(m+zW#e^&tyBu=~a`aF=7qAz(4TKLuI<vNE}N#8a67T#OI
ztd*e!4?UF1y_M`!bXW(TdM>7L_s%P|57N--Ik}RY$Cgh&e*;tLcXh>#*#P~t&I-bj
zd{th^Q$dqf6({AXmL*;3P3(Zxd?C^ql?X#aer{|M-@xD_F3D&HA$V9a>B+>C<dzH-
z2cLjA^?4r=v=cg(5&s%2FgQ^7CAQ{UmR!vy?<V<H5YE5h!wwllSEoFLP@=^`pc3e_
zHA=QV!|D11cDI8=iSp+Q<BOgo0!P9aT<`@$9kf&ye2I{b{Jk@Lg%beHO1zNU%Z)rv
zl&~2Bjkd$|G>0K*Lf?JLeb>IzGc57#i+SDLKFP-tf0VJQIXTtw>ZLB?=TJX?{k-~9
zPLX$wItRET@l%Zjma6YYzRCIG`Ey01FaG-NkEJ9A@GMZj*ON4S7wy5-U*5cyzSBG6
z0jNg!CMbp;?dK##`)xX{%rO^vl3n+mf!N}|%z0|ArA~6z(q)eN{&RR{DT}|Cyj<?r
z%HNuv>tGI#>ldS;q7480D)tX=UcVM&@Nh#Rc>SDJ-nr6#sPkv?QRUc%XbZGi{C<90
z=6Z}&+QZG$o2%zf?YjfNWFYs4oK&khZ9n1VjUHahX(xMIUtQTP8{_fXR`pwx>lEor
z+uETr9-H5x^49QM%E5X<tv{{Ny-(?RKTdqhH5k`rd~Vz7P2`Y1gqsLTE|<Fmlc%-t
z&2jdO^Im1PH1)0PHRaq|=F<CrYaSPR+h|~~HvZ+;kE^E->KoTu+9_AgJ(nxzkg)}R
z8;wj;`9n6!{PpIh2(*=q57?0t?@->VlD48#Ed7ZS*%9T()+y)I6<T9arCXIr*cX`y
zVJPa{NMXO_RMKaejK7t^>YrEdKYWngaRI(VoDR@Xwsc$zmoK)>(Rcm0KL>xx{N?p&
z($@4ms@4Mww}f6v-%IdoWoj(~_$3-kWQe?4@k-^6boL4_eJ=KrV+*>adOMV^eYkX*
z${u{3V=L`$?(YVd*orl+M3L6%Ej0ShRr~?2)4&Hta2%8{LUeIJB%*0{HJQeR9^buD
zr+EUy!V{=mVaA>MjF3~FUQ;G<qBu1U2^0091wZ+N7oFl6=`LAeCFZ_n1j`6G@L|HH
zG5{b6iG}X=*6*nZ9BEwT#`B}bA7GOZe7^@cCL-jgWFS=A?9_)o&%Dx;bZAgSemGzV
zJ3r{Cb7IISKPGj$lUdS{TXlha{D2ia7Pa(BH}O4;Kjl2P^UL$bDH4~woi*Tfvoo9B
z$$OJFMx3lDe;<Y2>cgjLchMFx@N&4n2sr?^Ty-44PZSj6i5Oa<qaQld(JyJADSdm~
zAjamj+_8|)sUq!1wQtKIEY8Uv@9#=<fU31Qr?_T(*7ey=MErAjW+{uy=H+rfh<cP@
z>AOkOt%ws>b8|}Gi$7()`$guUAEg{)iWq>nF8vhFPyNE1xBT2`vnZpEO~P|#&)|Hk
zI#vBf*)QSSqXF8B{J*%026U_)@=(<t&)8C$^y~hv3*vRbyHKT+QQK!%s&U{P%IWXe
zv2>S|`=Ws<BM&;LwbM9a;mSFObKmtB;XG}LzZSn{<biiC;}UrytSmX$G5ap@Ox{Q+
z4Ga8~o6eoWT%Ko3MTR#wk!htKuVHJ&SqisRLs`oAzQNx`JufL!=~)DS360y*W=dtf
zEwZ=DS*o`~%RP0~Osianz?I76^af=Uzy~(U*X(O{yz*WQKV3l7xa`%7JH7(UV0PE-
zrbY_?Id}I(3h-G>(RUwGC0NomElnCVS^g+|@|-tVBGa|(#8ue@4^&RJbmbCQ%4O*?
z3=sMB<co_CpWKcIyyYi!lodWk3n4-pPSX*KII!Zu6JOyH3?D1?(>P>3l9M=ET7lJ_
zcjCtrKfvJJx0wCqL@+pd@&_1jAy#n6V&e)5iipTYIqZj^$f`TsB*1QF@-&|4>1iB6
zhXrykc7ixJwpmOwAnZvyu(<@!Py_k%Q8O5v`;|mq4)@opdJ3QC_Z%EX7OumJZfb!k
z!qvj`f3fVb9ej%$%n||kJ)SwK1E?FMFp5XxQIsR{is@U0&!yhBD$iDQVtqT!Qkv)C
z`#UqH)xuoEXIG&OA>@sD#fh|%+q<sfZ6L@{I-OLKi!^`76Xt5er<Gl`<hQuXd$&bP
z7nr>@aNF9U&ZTx#N;C%NY#8fb+<qF^yZ_%$@(bm=qJh>&ycMpdmqYPubhk*?RSI^^
zx@!}b#SnfEY$1e2BAr&WVLPv~Z?yH^XjzDYhfR9Mw8trBic_iiOR?H91!g~GlCy>T
z74*IZy>=R2e-tVdtewCr$XwT}h@25pwXK|yjK>n?7_F#P-?c?vbkRy<N^i#~YvjL$
z%#q&Gu<P^K<z6c1A@F(Gbg@dUAbpO#^DIsleWw~UEP~a5|LWY={#7G_&?SP*4`BZ(
zK}inp7RQ64rURQ!myibkL^FucrD?vx*&l}qNfs0uooJF`Csrh&iK&bdMJK+9*U{?4
zP5~$;Vo4RGAgL_E_`6(7k<!)cwBWF`IgB#?Z4KhQKQIn?dn8O++z|**aO8hyObL=H
zv!y%xOhS`~Q$aY4mhq&J!?gNqWoV(lC+!yg5hmAt%sWQ-)clW8##8d(xn``c22t^O
z($C2a>z{s(pUG?2P$6%3$>m0(_k7?rr4zoP<X5(B#9YawZmLL<E83(dALZRVc`Rx5
z`lGJ>YOrte4IY0Nj0nvnc>g^F>viUB34Xae;F5};pAbIH{K90^HYIgEHD4W;cPfT1
z`e&*(|II0IDi$}c*&MQu*K50n=|l3q4OfTv>1=oyU;jq_GWwC?vhYJOls=b_KCf}K
zw-+y7Xy32~-*Ni)E-kmOL;9K4I&@mCK4T!X9NP<VZv*bP($nWF^^w+=z5hB^P4<#5
zmA%Kj-=V%8p0^&3Rlm^~yi_B{l5weAw*~%Ky~dJv{~mR^2hF8*%(Z0a;?6-I=c0=}
z<XEa--A0DtqIL|w7g82BWH9wqzAeM#gnKI)%(w%eK_I)u_!#W6NdO;UW=D<C79Q-b
z@dYgCL+gjo<^!M5W;fYZE`KTZB^q;Z9odyP<NT#5l*4I)d-LQ*cFZx^s=LphzYMS1
zlR>22lU&B(iaGR#<9I#0me@VW;1&meuJJ&xc2a3+6H+owe{x+nA)PT0Z3S(`Hj~HK
zh+m^oOLrb~`3`rR{8*-_=czJGEpoKdn4*8|ysc$ks?#ZbmUI#6m-4664#S?N-IyyP
zV92<Z_Ym(MzCFX~82WvDT(V^(KzS2NGWp%5VE{nHdkBIaCaC$IhF@@spK4zjC=mnH
zApA@oPgImdO0P_#@|d_w8fEdT(Of)y<OGl90gcNazi!P9SuveFj4;q*9AHH=*`Px{
z@VaOKf6^r5T8twWm2b|;q7>iLaR)Y{U_4<59+NW`G2GAt9uN%ZRThcChZ8jp4*|kj
z7G)&k24U_KJ<l^^4RQs;l?*AM#aKu<$WMh<xs9H1M@VttBO5+qu`AjQKnk$oAJ@x<
zBq^E7y(P^WEv>d%W=fCsJOjEody4-t^{`m*lOWhPH+Z&7j^}{ydvF?vF**yYjMvRz
zVi<h;hdBV>_$ZwkdV2G;GZvJ1<Y^z|x+r}XrNzF~#n>F8kL!0Wg@1IEgau0-_sWSC
zcKpY=g^z!~efuV3;BNv=s~9?62u|P68Rl4wE118@_?f4QQw8v1J_8?QXPj?3`K<j3
zaGYr5bPOM~kMk%f(!cSo5bx(mV>8~T&;IQtQr0_Y{gg33br+pq%SK;~%byQU{CHiX
zucz`olr^qzUd$;U>bytiB(*su^S-ZQoRW6uxcL5^GwS!{3;VuNO&+c&cOwP!pyzVk
z*p$oXSvf2`scYtI)V1wOx97&*v%lo@%Xhi*Oul<0-&dV?c6}elHhZSD<tiHVbZ#wr
z@GHYh^!$BlQj{RZX}s;0xYRU8DDjyh^O<}nr)K;ipE+mzt`BoqpOccGJLdmfWgZ>R
z%>&WBK6BC*N6(Qol2<QZL=+hxI34?Tm416jp1~4t`Zs?TAQJn(0nrrc`x#lo$lC<{
zfIN4ch~nyZ<lyIHlSuEGe7kq*WRkaK`2FS8Gd+pt&n0xUzmc{=AI!CT^Ue&otUpj^
z=%Ux9jTw=3s*jxKu^eJaJ6=xxf^ljMc>TJ63VC>@<9jxjVM@0A{g~rk;>Xj#lC{OD
zI?W|+t>Yy$a~_ArT}G}YI~R9KeVmI9_L9TvdkK9V>2pu-AAkBs3n4Rfy!e@9F%hUa
z1{x1!p#&K2>}5j8z+lcXSa~4>Dh6vzfJ%-*BohICOpHE#`7GbKVPZ9wwS>P#8fEFP
zJlKXWlj5lN=g*$c-Z9LC=PHcdiieNg002M$Nkl<ZcuCp)+GX9b+ZjX8jZ8k(Ae8C&
z;q}9-hfi#<%tZIY$MHSPYZ}%Po$pP5i7ZQDbFJs`rSyQW(Ws?6kGXt@JC?siy`Ec{
zj-|Uj`RuLJ!{{G|y`|o4k}s`T-cdL=@;Hy-1i|Fi#XTB6Xp+<|jF3&bIx;*+2Jw9J
zf&(PD&I8_E40{?rh8m1giFU{u%53tH7G?FgqNM_;*A<4vgx=s1$Xnkf^c)lktR!5m
zW=0ayq9fnaY?rph)Sx0rz8DgXE8uk3riDVOfe4&Xc6K_m2&UJ?N!T^3*SP)km@TG8
zxd=ngcS-~A<#7WP#-n_ZE2ojiml%ayFFB1v;(O3ML>|}WAvD(Do|_)llwoTe#*(E)
zPv^rzvpmJg5HZ#u2B2q8AL`=c`w#DR(a~}7S;SuQ9gVANF>Xj_!XrL+;^gzSENn(*
z^0Z=1VSW)8LL&(2Pada_&CA?^@o)TQDX={CQ!?}XGe2P7;~#=|9^kr8=#M`Nea7L_
z*F=r4!ut8(rx<#i>ft_P{oK$)IjDa&ec|%j?XGQ;cNGiPPp|5Rp4g|Cq($$XqJUq`
z4d#^0`$PdHV;pT`s(zr)eDP<RIAN5TgU{z#y^OS}$6AN6O_?#0EdugS=(Etzk9pU}
zkDof7_H_CXh9?YFY@jST=uubvnEx`jF6p6!JD0|i{t{O=0{8Dx=wUFJ)^Xf@%>3RA
zW;I!^uf=eYWzv3Z-5|?I(b9A4bU2yz`sJ%KVoJ@|I~WgQY8{po;A?%JpNoNpx+*#C
z2U440Lo3E%H=5`0@MW!|=2(YdeD&tlt72bG|4siwJAL>5!=$0HytYGYrY+;dm?r(!
z6QRkoN+0my;|JkC)<R9*OZwf?05xB3<#(%`bEVlz2N!+LY5H{eTh;NNCAbYT?pgl(
zMte<tKk`ndf3cg|{5AeaCT_0_0}6I4F&M^>z@YA#?9}r?4g=s<G6?%9gM%-!tEA>I
z^1Ocaax}nY_sCcI_@nNWm4y~)`{WZIFT^0i<cWb&Y1mcccX+yM<#0+pg`OsZcZ@#_
zkf$a<J+3XdefacI2Zl^47*IyX<N<mwbw?t*usKD<4rD)0-b4GKIXCtUDfaMNg{J6x
z`>c$gw-VuGbi8~g6JSTiQ}wzd+p{9=_3Qpw2`9pwgIrODCC_=p+fuK*W3J`+hR$UZ
z6AKf`o%OhuBQ0KWMXzEAVS=nnKk}ZlQ=sXIppa_V!G~UnZs6na__2SNC_Ma>PM;@S
zrGK{_^&*GxH#9+jgs@|xuJ7v<C(+1Dg5+s=;YJzMs3U$#^!GfpNFyI-dh6wZ9%+m%
zU<{lt#7a85u`%34hXEvh$BCWiZsY+j8U%zycj9McIMIUil+ECU1oERil%Ibp7O5hS
zc*x4T(a|NOCy0=hfBHQG@{>-W$hd@es!>-`vwt|1v}6a`EWr!FMFSgYfH5`WdB&1)
zA=}A6kp)r~`~11Kr^o&}3_d&3F=*r;aOb1$oK9e@$wGL^VQ67t-kwqGBY0M!rHh!d
zd!O;YJI2jiC6-^Q%wi`-p7?A+lO6Tn^+XHv7K{}pKk_iwdH4SPz}Q2dz?hg5EhS?U
zc#PMd%Od=g1Ll+PqR;e`{P846#Pxhn%j2HzzIJ6EY4Fk4bJ2ypy?O-d#f=C<n!{_`
z=s~C4xJC5P4|GoXa!`ll`e_H9Q;JSGG30F!pUsBUe>@2S9`C=te#+m90nQ!s%O&uR
zhff_cxAitdy>P<9zuQW?=Ol=?PxSjjj5y4fc~1;=YKG5Oa(c_l>AbjJj|F|+H5hk1
zpXUyv5I)b+ah<*IN<6*%p7&XvUrz1BI`q7;|NH8roK|D5+6HeF);vc(hJk`VQ|K~(
z4#?NBO=#y9Cnq{iK|T}1%JA8THhmavjc1(xv^U<EA(xFO7-I61kDOfkE`}cZFJJeF
z95MLah_T4>{5c#yasJG_efLh#9j>QDxnGH~E1n+0$V=V%>Gfyw406)P9qpg8=Fgz#
zU;MkTfpJT_y=9&&3(v^jZhsDYmtaUgf5#1TFR_aT?w$siSCqyO=Tg#c+eqHoSX>5`
za`ZVLY%mCo6F8YO=&2EQnD_t~|Ap*u;feqTxwv5mzyN8yi-}2I9Q;6=0q>jFvU`OE
ztK=92F=0IZVAmzs-sgxvKMZJG-O%!JRJ3ZRjqLPPKHy}bj0qAGWeh@g%E}i1sM~zk
z5)(b8Cw08k4eQsiOKD5!db>paBfN8$h_(r=mv0kdcbo|8!Ys(1yQ9tH5Y@$XUo&R@
z<j+hV2fdUwL-tl=FL}<F=yXZpd1-6SVT=(b^gAk-EHa^a2LHs>z>)GKqJX{8103i8
zXL%M;L<uIs@KP{|TR_43KjbSw@>3Xm3EFqwNCV#-Lk^>5X?pqw<Al)Zp&%!C@;^>y
zqWnrt^9fOI@hH0+XkaUJH5~|{M^bc}0v`eLt7U+3qY1EX%pslAGZ?RsiFBuzNDSmh
zFM0w<V3EP(N(-ryz=9!;^xr}go*I@lVjq;T_@=o05pLvj9UV>$ILJa+@-Qelo>*!W
zhCdG9C0qg*4IEJe+g*JpV@Vd6Wz5Y<HCxag7KBA-o*u3R!eGL~!&sbO?ijyf1c_Hy
z@j-WfoDBNQo7aOez2kFj7IF-o$wnHB-+poi<3PSK2^bcQ?VBC_bIS7#tc`)0S4f_8
z5mqL!D2~l}{(Rdnj<OKT0w@L$nYE~)%b@Kq^AZ;G(F=1O7Pwgu9T~EO7d;YJu-P?i
z*SgUyAA{46{^O577mamdDWhz)FjqQJ4>|8gV#u@i+wfhHU)xrro2OIWp=@>qpBq9P
z9>df74<9572J=aqQ@)mYU)SeKI#;f-c>aCRKE!yJ>W6nnUM3|N{tW6<gquEj5)eg_
zyiM6DBRc^F<Tugl&kw(JzUp%0(-!ny6UedA@#S-wUx#@u+01F*%PAl!iaOr$b_JPh
z(&W%*N#=`(IMa3;1q>7UDKIeAsG%qQdOF+^!jCyRhO9ZqtrPqx6UNpS_fh7vq9$?I
zqVLxOg9WMR`;WKWR5C6ly#Dy}Pm%pfH$lce09;&d@U_(8U*^mCoSWzOy%=!f$w2xJ
z^gzGhjy?*bU0h2ck3K|yA|vI>9Vdb)vl(SD_Ss21yN=M?h+$#Gf#<kCy;8ZOk;nSd
zuz6P6-G`5IZLze^f!q_PjdEGHj8lul(zQ{KweR?z#u)LJ^w;uI>BgsRRC_4ekUy(^
zs)br)p2E-RUE*yTXwkb;k_IsaOCQ4c{@9zdretc<KpRvGpB9)lZf#IJlePh-4gS7B
zEv<*8@Kbu+lGYTTL-S50r;qh=`S+3sx;THIS?ucS-)unS(ax{EpA>*yBMe4z$6eU@
z;E4g6?F5U^3`lJPL~0aDccsYD;@q)2D&OJ4q=boO?zk#}LCD>6GO3lp-cpzEsxo|O
zf--qUe|C?!H4z5qTq&T7eHkOWdpp)v29ToO*gUs{QA%_8t^8b$rF0L)U&`+SzfJ?X
zXq-@EoyG7v%G+Ym<_~H25tm6ojJu?zG|#(j{;(`(P@5ux^@}Epj620)1o$aGIT0xn
zV6_YAAE_TLfcz0a`dWMvBMq$+Bh*(9aZLlBWc}0Uk3uOw@wIFfPJ-?e6vL9z7GKlh
zLr}^<P{ZV<G%1svII^;687zyxK&hiVW=yfTib1xP)nUx%ukj&dI$=SYUuSShvS8ft
z>(St?C!quxI7-Zqfy&Qt%^Fl$c;e^fBF*1tz^hS*2S^+=Odxy~r_e<*UXdhl{3w%7
z{yi_jfsVlpnUK}XK{{y{chSJ9G!Rd^Z6yQad3hoV<ZOXazZ1ctwk$OFS4j=Qnf&y-
z*oS$H@$LHMVLCAkIa7|)=<e*sfLu;n2+t>9EM&To$MApaxHvB-Mu?K(ijD&sCC6et
z#w#|dVZ^~`;eC<A+lIGXR?i*!ZBRp;rLpmRjec@`Z5(v%Bolq3^wr<XIw<kaYm(Do
zr8Yr((vCS0VWC`&i6IRM*+7_j^g#~z7}?rg8@$&AC*5)bRb0$>zC06AN3;VB9=W59
z<;Fnt`$-I)@e%lxPKHE$>WTU9p)RkVoSt8$ikujw@@EQpF`njgcBfN&UOYedzlcHP
zm@a=tm+)q+!gX!A!vHgG7jo?zYa3agijk3P<Vp^MW?4QA${4smX}hAmal+<vW^8ks
z<qe~D1);;>TO0M(?m54;U-KBg^<yz*-!KyC6}c<f(8m4eBs|W^YV_}$kUW2hAD`76
zZ|?a5ehNR*Tk^sAh<6{uEqX^U(lu+_sOiY(LdQOV-ztAi?$-1cK2tDxZvo#a_!hY?
zMxL^k4%KU!YFp|ldACV;o=|(~4Z0=T|EwF%#=7WdOyj)KP8C2Hz|=nQSDHAmW27{&
zw~5qHR0fSqFiTGFrOgIOOs?2bb0=M?2c0Qsc~Hp9%&x1*(YYqBG!VxK=cl-)#2D6I
zPW^cCL*+OW|9B<6-<;xo;<3(R_^xS)h3EU_&Y-{|La%pO+RzAz^19StUn3oyip`%i
zbI^{HICz?AzSB|?p?3|zo_ZciKpkus2!N!C%*-A|)=qXHR1}wmJWJC6<?39b(F16X
zPkOvKN|5q0TS`tzDj&&pRvMQ-0!EVlL_YHQVc6j}c)e>Lr*c?aG!}@c;lN8LYC=T7
zm9o>+7(A4Z<-$zM0Bwag98HHm$dY4!2|9dAOL3J~!Nr)TU|?L_MFV$913Yx4@?ad{
z1VhY}#XwoyEIAg(YmNoceBl$=xO;qz+n?BAw=E|@a>t@{Y0jbFUCDqM8<!U0Yx9b!
zF}7W2WP0HmIpi~kfrxK%<X;)dgbC4O^eEr;TOY9=O#r^_5qn6&zF5pjh+*9ap&NpB
zRo)Pg1?^HD0L8aVa!38v+DNT_avC0Qf%m%bIg#VyyIx`%Vans}^PL#|`nKSZT<0~O
zyQ1IRXu^BX^zOLkb8;y^4K##UAEW(q0_Cr7Ug@clzi?tEoa8`nJil+>y)Vk<oLltk
zu<GkOc-$rjFdIe=;e?T<)Fw)87LjQaTt}NbI}zlqY?xc>AAu+txJ&&Tu(>gjG+f?x
zs9ziDSib-LMO$WVMtrW=s~PAWerq4i_}Fu6+P8^bod4b?xzAbF@>6%t5}%iB7yX@A
zeMD(4brZF5zr*FNB%%{e%EV*6uBNFSQ?)!M4#Qm_<(+;EglYpr2Fm>M9aYf2OV_u#
zp11Td=&(z<<fxZ?(!^koy5|}KTSzNm+H~(ecP41`Hkbd{@_P$ht23PgmYd?)r>@oY
zT3zj-kTqo};gr($koO>(v9VDJBDC)`mpO0mfYZKJd=F}`Ag4pbc<@_Jp*Tp7FT@7@
zablo&gmGx8#W_5%2}c+X4H{01bYtzZbmdAQ{{Cfg$)y5D-vby!{Dq*tu^bmg;2b#x
zRXPQ<1k2{HEJQ$8obFhZppiCN{j>}q?DQi3P>BDQI9SL_S^EVMQq56TcN7nNhjSbU
z@i=g*I}j7?9)?7tV<9s<CGkD!;$JjyXElHUqqHdhTNg0(=^2#0dzJ>N7GXU9Lr#a}
zi{-?bf8-0^U|~4&#Z}u_2G60h9rFIvdLGw2j!#v{L#~Ujxjl--QSB47&vF`dpABlm
z#r(7rUc_KtWE<GVveqf?pW_?rM*2@1y_h$kMfB;5!fd!(>!{O*xuS)bUF}!ajm`<B
zj+m1~hs>qxs3OcU!g``ZMCUXYZOI*{H|%;oPZ|2OuY4{GL}?DqZt{|WfcD9?LcipM
z>DEr=6B%oc@^VW==^3rzw4@(n=RCDUYtq{4tE@%50Zi#64&#wh|7eKxK|e<8Dj21g
z^fgFc+Mro`R$6z+Gv|LP#=TGSoi?6*cZC1=vwQ`f!|d(*7;$et=S)A2XU|#ArLy_+
zxD;;__~C11MDl>{VMOb`B%HK9?~9Tz<t%Do=5Ze2+^2RLxA6{?P}&g(KpX%J5G2B&
z&liefUfc;pdbfi&0!er{QOYu9a1*bM@m!yfd3>!OJ9Lf-FLhJLY9npv66ZJMv!sba
zrQO*iA(z7NOA`ok&?|0gaIv4n%KGwWXAK;?{@{0Po;$0snS5Sue;VCyCUio;ybLX(
zlyLI2sDqk8x`o&FFea(JckKP+*L%T>zL6`>Y%Neqp%x%YV<ob0^GHmS1Ww*7zjd)O
zYQV$E$D8VOVT@B4fSn({BC8zi!i1mtB7n*wKG%;Skjf&m<FpJqv3FV|3XJ^J7dSg{
zV@4~|EFJhHH-wOdB>97$!2czNF0NeTr)UX`bk(cElb8%B0>nG1YX=es4zY21O;uva
z6>55&8t-eTli+zD%YVZaWKbAiX!#{9xscPi%Bk`iJaPRezTwzaeU#xxwx|X@zjjUu
zW`A6)n4B+K#=J}TqJeYPfc6R9!`4e`&r2Wo&|F(Dodg2MPL(*7-j643hS#w5Fr~Q*
z&tYwTQCh5~dz~^;_1r8pzN*Q-wZhJOzE#q9P&sBfZh9`~ZJAoSg@HaD{d>>Ob!#{b
zL#2}?t@5ytTxt)sIS6!Xb4sgYm#|fW!{AEuD9XdPX8GO}Co{|-hFqndzvo|SqvY1x
zXqlcH;Cjl-LS?FUa|b7O?DtI#6#mYH1K+gy>zg-P_9f5BGU@h#_Ev-6GSn(d!Dh(u
zK3SfTsb|-|sW_yF{z{u_hfsZtu~Dhd0z$`BJ*b7ez=L!dX9Z)9XR+z?jkM(#a!sF=
z&yi~_#WA<>kHf`e-IJG(SGUp1wp3cm-2$_fRNlh2qPHcT;`6j6gF8s;oa^%rs{2x;
zl^VFs?R07XJRYeerTMP8T!(WV9|N6-*TbDawsQ%!nt1(Scu3tBVTP%F4BEM4ki>6l
zaFyef;<3{G<Ip9zTH4f3q(~clk4HWXI9Zrl;sAGAQ+QLlsZA{W1Jf<^r+9Cbx2GJr
z9`ksu3?-So3QKu4I+yhOrvYyt?%!q&v_3DjGDaOt^`kZXsdQwc|I_!x6ky0>@kaYh
z-Ys%m!`^v8N;+$C^w-z&y$H|O-L)7MBwO>qmmlx7YhC0r!-C|3=s?DqcN=NA!h6m#
z(%+5&u%46TQ_rqtv-b%y2#o$xSv0}%@YBv<ot+vTlb<Z|dVNqZ&6AzPJ{IW2k${rM
z7gX{YNB9C!M^o~kiOYAJqHh&d{&ISUcO*sZ3_lV!Om_FnZe}4020R`KUu-_8k#@vr
z0Rxq9!icni2Twl7p+1U)&vlYZr4Ty!lnL1%N}i?*oWhF#NYdaLo#&C}vc`PqUEJw4
z5bNpm(i|=U?R@M0GncA1M(6qPXKH{m)7Cj5^6R-6lD+Rs$QX3&Dk5sh9UmPxqXWCw
z%O`Z?ndYI+9AcC{xDHsp5zSC%1UI1WZ%r>9z`_9Ujbk{jzVr8E<aZ-__A8qF#6RDi
zv0gGlwuh^8N*d3EV~*nU5eKW$^87t~+8jE_Pe8$oIW7jD=g*&AaU~w#OPM;wrQen-
z2#Hax)E;Z)J+z(FS{cu!T}9XnsbzfeqPrFm-Wa!#_x*>DGS2Bwf%I6nJ5+l|@OXBo
zqGJ)Kn{w5Vr|r9ZYkj1y;i>t2`S~o@GN-ai22;j3T&>TKK9=ufVGynvmqr2}kWiNN
z0D=>!d+_r1at^1wOXC5TZ$d4awTEuzc%EB6y3!49=O)-io>>jFw(~hc&|!|b_qcja
z4)V7uQwy&am~)e-MdrH#v$TJ%8hUC2$nrp<S6^w&p3^E0kc7n400s_4+Qp@mUY=ux
zbdBs-%n~23Y<N!%EJ}LqA5ubn%!}X)S{`5Ok0~ytFofTlUZn1N^t3$g=`H>txGnTA
zFc%G+n+B%ZeE!_#@x9A}Z29}AnWq?Be*a3w7eP{k2c50Hh@c!wBWaipaAf;BU2jOo
zSfaTKqsAP7G%?XJ07KTW^Dva|uNs2DCB=BAh<QpjrBep(IRi2oF-$;B5qPtVn4Hq<
zJ2?-C9F)t%l#u5uML~C&Qyk9?K%T|v9q=&PsQQE=Fytc|#u6NV4OT)A#BdW{YNhv)
zpRQqz82WrTx|2uNdZG0IW03d{A4wS!PYgCVN@o02k-BiBup%c;^9V|Il!ZK_zzJue
z9psF8giDS^;T)#~cj`1wDjo8X;#12%$KUI44(_((A7?UZ&&qZ>`F+KrvVF6&r^*5(
zi`#MWm<45D@UC6k28FzjGOF_NdO3x##FhG-8b19Fn%SyN(MPy3yvB2jbQW*z`sU&H
zV*H9c3=(dHtDT@~m1EO+pHwos-nunTm*w}>2f5bj&Ffd4PI7pp!dS!4zRM)nV98>+
zJ8vVkF-bhJ<~UP+xvkqZsB`@Rvb5;WW#$CPhmRkw{_^^bJY#)J=j6vrF#x@L|Gtsp
zREdt8=VO!_*S?fvh*JMxy+B>7wH(uDTZ^8Pc`Ru;g?{(T<8N0_pWKuIB$bm!kHmoa
z?Ag<+&vG5%m#^O>byx~-A5|jvC}XO<99s>g=l`7K+e#mMwLg6t^<le<TH8wd=SgR)
z8iP*hIXXqDw<^<CvY(qgTb1iB(&<;RveLzY+#qZCjvr)MWEP-+)0$MKT^ew;QrrqI
zTkBa&wp*ZDj5%%Ofn14p_mGN8OIoegf1`XI{YLmXurRNf13h_uNuMVd>K+NOxBiD$
zF0NDT-ZHjIwe@qF^Udk!@f`jf^ulDFKJz;545u$-k&1<X#{E9g);?0<q*MaKwQRuY
zk2r<VKjdLFN((-uW1Vn9-IA4;-nHcMTua22c=A)ec9Mc?j4dx&L9RI&la8L~f}4_-
zKPrb9d%)pGS)t=`s7P2zYZwqSn!r_c<C&x<GL|Q}+1rKpB&T>-KvX`Gk3oTFN>J6X
zfFXv#wO@(Gl{w(C_=#afc_dg^y8*}FbEG+2Xi#2g;}l+EK4#Feq)h0NZ&#yG|5i@u
zn~dDCZ^w6#0|!r+Qxh8g+?HK}ciG6(P)!$*BWZvg<6|cutCz*{=x*dXF?Z!F$1K<S
z6^C@czlIT`_S}ze5#@cCQ>yuNJ6k#Jqqib9L#M1u)4!jfvVLy;_jwS(``j6E3fbO!
zXIa|OXK6b}F+9F|_g?z^d0jkyB-dcQdaXtu7t7m7xsCGfnQxnew#!Sa;93@^OqMc!
zB<V7f2JceYG0fiF44>CX<6Oh*n^4rslG2s+&9kvymL>cn?C-0OA3w>rq(0B~br^fO
z&JBYePjYk~%S+~Cr3ibh@Q*Y;I~KnSpVMi;+ute)@<zLg$OY-5fr|z%8aQ7KR8REy
zqRB|~La$0+*(dj*9yjYvvuFobLhUjMlmV7b$Oa@sWLS=naodAswb|lVB~Mq=ZhN_p
zC_m#$79(*rm0IpTe@kt*baZ<k?&n2yZ~ezB7w1Y|suytfma)b+m$wxz$7_w>#x`gR
zQWzA_jyS0-7W?~&cwc*f$i{Pf_<!4VW@;di3|*{~G#sXduGS?8E!nfJc^XcqZZyIF
z7?8z4cX>QN`WS4iM;9MpaVox@*g<|TfG#o#ct=igC0jZ$Fyl2uwjgE36{Q*eAYLaT
z<RuUIM6P4th7AlN8cyPY@JN*xh)PiP4jswY3A+RkKR*RS@!$td9Rno!N$+5VhL_Xo
z1vwKuY$ac1PP8!NV2DzBLWo`U4sX}Yk=N5-(iaHYy>nVeS*c5D8OSAik2JvbK%9E;
z??QUf`A!TUe3vI!es)rVJidp+f^*GT|1r$Pc>BlOKZl8D9FuP8Ajf-`{;Guz-`zoH
z*4OuJF|_GBOp^SHH_xR;mb&hJ+JHFhBTA(R@8<@3TY_-1<=y)aYKVOL^hQrT=!>Ml
z1f9N=1@gD4jXOw97tsAF?fXzS)L)%iTMHTD>$i$}n`@VOBPT=3M%xbL>;ie4Zx?XQ
z69Dy6Tc(GjJ-2h;4)V8N(Z}=sOxiMr-6@^WCt&QvIF!4*WTzY(<tMOnSzDiv9Pg0$
zDR>?tZw>dZl%rO@LAvzc7kAOXMFSTN+-42fN>iZ(rSIVYRvH+_uCUZEGRcZAk9(di
zsCaDbmuNFUVp6k(i{)5DZ))<nhTc{<r8<MQmA9OWUeC#GRK>d$$&Ea_F_<A-YP&7F
z`ew*(ntfg^YQAW`YkwZH4pFL?I`ni-i~byDOTN=mP3PuqkcE4$*g9v&BFy00oYeVv
z9~2j9m=tTe{K40ER0^nZg@Teu8rP($pC<K$b%Jlc=c!NcOfpXKHK5VrAEg^|l(yl3
zu5t-M6GixAIdK@5Lr$ykv0#N&S-JKtz=5wFoe8E{i6uA~i1ai}N88FkI&?T8f>DMA
zNKQvF8HbYpTzX%iWT60q4RG>b;T*<&^5IQ8bA1C(J5vKtl0`sLnN1fu)=<W#3zeN+
z`~t7z6Uift8d%sOOn;z1kzL!d#Bpkj($M8kt^xF(yHZ*XbBVWWz*=ED{HY-8JG_5i
zeUT5p=Z-6({97h^m0PkPkH6hB#MJI=n3U!%x2NUBcT854rYQZP_IUz5*FNCILT~Oc
zymi;Medo%i-mTGOtxTnb<w(DjU#jC?y7o|sb{qC@`EJSAtG9Bx=dD~>MtG`x&!0UL
zvVEy)b70#{&?dnYRL^M&O5WMCtz(Q|o9EiI+RAxMw@=xEMo(=jByaes>}nqEuY_w)
zU0r_0DkW!Qmvyp~=q*jdxcK4Y$E!czzP<YRd3fDmPV1>$f$3Lh=6SazUa5v%#Xkkl
zJqi@daY&0j1b4G?#ljSb&SV!b&O2$VwbFAc<yo|&b5h7fmQ@;9+m6=atrBu2Qq|C-
zyLJjh=U(tWF^H%}t5*m2LYD?f(4Oj8OS~B%UH7CT+oOj|@*Q1>0z=(JT}UU%^2S=P
z4dyFkc26ElxI4?UWN;=;z0=JSd1`eyM`JF(6K3PU)N`{jt0r<a{YB&!m_y(r?wT-*
z@cvDfHTXxt5hd#PS&}lD)celn;Q|u?);c}ieXj?N+n*$QY|RqxqCKtQOAwe2s9bzq
zJpd;C2CE73rWJf$gi?P<q61YH@-7wQO0o(Gpq1j9_p$<-F=Zf55{)!K{Aai+6Uiiq
zLBx$EE`{sKzqdoWV=*c5e9{S%=lS3ec7rUu{WKEo40zjBYE&%T(*y=Llv$IMF6lit
zJLzL5h3p$U5#O+S+!)Wp#x+m9TXC<WmQpQ&ZJ{>@b8LQd@*J9fTiVmq4)1UB?T6Bw
zz^@MDn<wV@rq9smFw~#)F>-n$cRb7aiJ=^3%lNU{SmYDKl!Z8LqosXJ>6CZJ_wR?D
zGC>xrmt0Dpg9t#ymBwM5s#)S{2Ax`Zoi^8w%6CiTsxot8^#5Ut<bTK_xn6_C%Pw=k
z$9gEWQBd+W>Xh*PmytdwIz8iFM%v!EZ8$Z(m+IOnx-LAviNBhtT?LuTK)GT)=lP}e
z5nz_K=bY}^__Y|%xnj`kgAKl%qWtsSyQ_B}-s|bfoHixQ+i4DS&hgb%zOYb{4`KEN
z#k7Uw&Jw)V)?3P1!_5oYLMz8<jTgG=0cy>7KLEN#ktG`mYHPs9DDILvZ*51%ko&^p
z_G(~lJ6el(Oifh{Jq%t;6EELkR}zvzj1L%MKD46K-LvVg`<OnxBZytSV{zl;!54fY
z&AD*elsf|3+w#rs#@eo~kuV{FyTe^J_sNdo0ItCY`mKp_b9|V-Ly_ZfoyL^QJL0)P
zOij;IJO+6|{Hd}IaPD9vs9{y<<_0!Pi5@Bs{N`wQe%|I<WnJ6$oNp_iDY&(=PtmO9
zP4UX}@_3DADILAu#b?o{>i(>#AwU%1#vU^O$fQjGI+$cM(1i*C@CgF2!0^2vHJET!
znIsx5$<#(BaR7P0iQjJ+pmqSj=T~tYJBhA~CwvtW`pV3F<%gUSMRtk%H;Ht2X_5sn
zFN?ECs7j77&U`eN%A)c@*XYATH)SMPPE7ep7)Ws1$DdQi2ed0c4dx|JrBRK*^O4~2
z&WAVfy2v7YoNhNBfPpw|s{#xk#x@;}q=8G_A*Y@wLT*iyQ%E8!b;A3a$q#V&oJ=FF
zQ@63{j7=#z^1M8wi~|$2d<$tz$ugB!W8*kU<u$NH>E13DAx!|gY(D;iKDt|@zX>*%
ztOXiifw5#vFayu^)#Jx@R}?T_Z#j9dzkI%7aI(SR(Zd|2689ET$!<uN=2Uc*<RNV8
zdq|vvkCMhPx`^?}ptOL)iJsam4t8zZnzJ80S7iAnrv^%T@kQhKoO@le5UuaVo|J}u
z{9u04Yo~vPvv*;_qhF8oo!r!X)?wU7me;RdUj60u8@-NAUv8c$J{uT6fBCE@d`9m~
zeG&G+CVwCI5GU87oL+~tdBPr#A;;a?_xET{w^Sd%So`Vgq{kbl)fOCWuG(Y9P+#_#
zdRVwtmKLg|{)%?SIIrfst;c-YO6i%5Fx!W(ub$pKz545$H`1nG$%Ys=T+)uION_hU
z^~Bn@>@**;NGZ1$sevN>LVP}IDBgd#AK{KAIea32#+=9E9m{))@RsG=l2(f@2ECQc
z0S56F0s<53J%)4iHMDaYOX5p(x#Zm#W!R!jOXQ5a4+&FG-nUHE*&#fRiTgj^e(2b}
z|6!Y>wx}D3e!rStAMO}$6hQ7TzlL@MC;cv!&Eu_Q{4Df7{_Ib(x8gaUCBE}{o8G1H
zBJA;;mdo<#S1(@<vm{<0kx2K(i#~U@ld1d|_Pd;hSmS|$9D6Se9xyu823`!1O1pg`
zo$g-#A^TM^Uu0T^oFSDartEI44R$m1LG+;vB3(B30f^xNf<zvP?rM51*BQIirXS?I
z1<3ezau7rQR&J`STk*KFbT+I9UWVQwE!7uz{xWrcj)t%+x=L~fyJg$!h%O97wfkUV
zibShUN;D#`Jbx-Lhn<U~*Hb==hv;akY~xmK1`#z>WD1mdfX<B7IHEymjr^CN_qj~|
zduL%YrVvi?(75eXN~KjR!-R)1rs|wFbR-OYC5d5zFoqGylEo>H1Mpim5@>oy7a*J>
zn^B0Ek?4&Yg2aLaKjbIB!yzb67eRy^iv*f4f8upcs0rf<sKEFYe?QfrVv7&9K|;y(
ziz9+yG?Cu~qiv`NfSNEJ&yXW&5(7m^8m>_Xjl7hjKpyTOdOS}Oe{LL78V>~Y)xkB-
zqStPT=~P|`p0qu1{PC$RJl15HOZV}Y_hW1HH^<|W->!ku_dOW9K7K4$xTVF@uX>ut
zp2HetHB9tRe+Twdz9BN#LRn~j@%;HLTt~DL@HU%g51$YEn;gR8KGzw-mtjk(Q*F7t
zfK#6DLEEo&TPusp$J=T=7i50IA~ObdhwR4{80Sih$~hf>hP5;A$#fyR`mc*>0K-jr
z{xgL07}u|H`T>XGpZ9yI`nU#*tK6ca&rh!sO=g24-<rxD22I+0?uPYwt!=d55RyZ;
z$4l}%wx3w_xcepf$ormc=yZ4+&v?vy+93MR-Hm*gYq_Hog_1~n@-bMArj1n!xPmZc
zuz{Yn!67bfwWc51c*)rO{fB)ct0v=?>AXW38_{-NJb%{dCmNu5milM$x13V*@1*sT
z#4%K#Iv<K-_MztHae?4D=w++p`19gEkU76RUZU1*7Aj>H(inb5=lto~JRI}SF`O0Q
z97S9$Q*3`NM5s4@OI_^ir|NwT?WMA;g)jBnEws+LOjGhxPu{mo)!7t{bISV<Z(erd
zyn19O;<u`s$5-PqO2=%60CGQl&~HYFEMBLwd0J~38I!$u*pD|_@tjZYzY8{vzyITH
zo~`jW@(ny3ya;<-6T=^u<@tkNzL32*I38&Q680g~kRU;fF}1<*%?%T&p<^&x+F{2%
z+rEh67!1nz0Bgzd5x~65<$AGi-^Ru$@N0v#d~oRxc8TZGG!eh(e9~6TtR&V3sKZ6W
z3UZbVEz@7WxDE(|y(ej>WJJNpb8c6+#TRFk^%Cwi&@w=)G|`^Of{^ZfUjQFLk;7l6
z?#KL6xc01WS*DzCi%xT#xp=J{5wBGz5$A4%+ckcOj+(fL5rn?m-cyF5{ZQ!WyLjrL
z+&Dm=YhjsS2IC27+P5coNCk4C3iDP?Tx0;x5k$VSbD{|rR3}F<5*~6{h9;_<&{xs;
zqXFYUkarJ;6i+tUQ<~B;*iJknYsMf!V&Mo}UGxfc49mZD17BFni5@vglelXfT-zW3
znm}d0ky0h@ux>nY-a4xngzsGM1{_WT8kV>&9(uX}Q!by_#6>ye;2J;L0lfKP$nh|E
zgs0rn$>kKS`<z!>c)ZU#nJv3NLK$FfeCc51ch8b1ia&=JKsC4ozILkQ`BT1cVts-e
z9W){P<L`soynuPj(;L46E<=_mbAy*JUMNgl_>7RX#q}lSjPy#UXy}_tC+(LQI_YzC
zQCIpWoc4E#aZOe;q9vf@(f^e;{8@Inan<O9^ZAQ>%SJor#6UcBiZQR`Zqzt8%;C0y
z(rF5yepWW9VL%z47LhrT%vVddL@2jH<BtY1%E(-`tO0AQQR<&+1nFKlk|p1z5?$mD
zx)_vvlS7_QO|n4$RAl#S4aYF;DTQKABBMh;T{i@j&6_f(A3HX4j15|pgEB<NGuL0y
z7;|_wYnnDVe|JYL<|no1*x?{c8!tJGgHw)qa>=-LqkYU+Ug){OI?}avP?n)oZr`8~
zX-tJ9JnvBYNt2_OG0IHek-7~fi+-YX#A;U$PRT)@_tAuV?!@+W$e;7e<0T2^Htpm%
zqTXBY_AUZ#D<2axPCtf=?XSi!FMn+NHJm(s$yjv_{hXi2r)cFmcbuBPX5F$z>o%q9
z*hy||sOd5{{sI^K%e~54lV>k}=SJ60#Z1ZdKi_|x${yrJBhMgXW#n}-;(-1ye|s~a
zvV>Qx=lrzNfM`QRR`L~Ua)kl|&6lt2<`_D5ybTQ|40;)O4vjPn2tIuLG(=9ZUvkGX
z_ZjDoL2f=cdnCK4@<Aa3<&T2Tarv>surZXyVVF>q22Y;9-e*u*9vIpVpZxVP#i?%a
zvcdflTJZUO_3YVG9iTtdT~tis+0pZ&sEf~9=drm>$$EG5Zkq;{wu(b*?%4bem3IyP
zT`ObWrq8W>#~3it4yMY@;=K-XrH^!d4b}oughc}0CBRD{^`%C+Up#<<!0;CY3F$kF
zFcWC~DZ1v7-#dc`zY{p&^)RK7(9rjy@lcYHY)?#(V+Q;Z&lJ>hh>Xfs2EV;CJTIre
zQ+2^bfPIn>7hrrc2Q7ZNQyxmzjA3>?ni*=KU&2xO2YF^p@*;v0@hnb#2%{te)v4(+
zODC)bY{?@`M%-LM5{{Ea{4nywQ%ROP{F_fM<HE}j@)EmfU|S8uD|sSi8)Fw9!USZC
zx#N8BhGYLobsT09zq*({0xy4m$2XSgNuA<&KPM!+&_8`DZdQ6ObC+i_$1sBp<+HIc
z*BHHd{Yp-3JhJXmAO4&!SICve_?AA_S56_1*G|sA5`)uYnK&XRMjyhDXAH`bHhqpR
z>gvj)^v40Ww17>&X4m4yKAuw=wbKU7cgoYp`iTZv;FCrFl7ZvJv*)9B!dxex$I)+p
zl)A+)jTN}sx@1_O4ZeQ)Qs$9Qb*}pI#q+Bd(hl@J5EYlgj|<oQ<#mMtvNY#|*U(@g
zMxC!?qYj3(lHrm~d4H9**gvDzM$re?Y*7ZBzk4;TDNFxw!lc}uhm`C&p{%c7mgkI=
z?}xONlB;dTd@gfRo=3{{OmzAB<*U>N{4Lv4v)ew(?G32XW>TJQ)+#uKrO%O`4t&BB
z-x209FXPG%VlhOiF;JdAHE<<59kRAX)*y!RvJR4OOr_6!E(SVeitdY?{3;oXfPqew
z)$tfxl?=h2pZkn$@^)(91nTAHJ@M=*?Mi41LT|9ZAUD(*^%=)?Iln#Qy~NA747Fay
zT}!)6?UGv40BtKD6LWR*DICwnl%q`mvS-_v5n)QaQQj1-T<0U?OM+@Z$4)Y~sp&E|
z{sPzcm-{I(muJs5xnC*PQ*JSq{I9QGtY>g0n6aMo(@IN4l#z>~%#QLte)_C~Br`A=
z8wT=!dHq_Br+<^dHj5pPiUV?tJPhn>j=>=VuxC&ED;0D=E`wPH6AVTe?B2*g&=-=7
zmK`U~KOm|TA~EnVu*;23E(0)cUYYDn2x`04Vh0};?{v&(SC|cg7%-P6P{1L}3z=B3
z+l)aWdh-cJq-``rb}2(`f(|VP`^c^OZ}=!FN3R*gF_1>5$Y2BK$lXfBarD~2+#t$-
zU0>zg!&B#8$cJ+nQy6P!40Bsk{3X#+#=Yydro&V5jq;p>3{jp#!+Cw5Tlr4C2(J2I
z)FuHx!P5Cbc?4~KpRPcYVWCg>nTQj%d=_a0R*gYIQzDjb^u0cZ?>E>h5<nmenD`x2
z0^+(njmRIQkIv&-D#Q0|6#_a+)MzPwqU~__BD2aZS=tX%21F7cRr#cUVB7&#GJ%Jc
zc;Z|h%VR+8yQA<KKtFh7OB?K}QX`6_Ka!I@;HyrQj#1(pJ!V)+i9%2Lz*L+YXk6b;
zuYdPSIRq4N%7K)s(j=$L57l_e$J!`}sYHn%6fUk^1F2OO5xk{nwG<(?Nt}{l8xF^(
z!r~-tR>#hj4Nr_OIlbbGoQ%cMHMopJYbWdb@%cA30yuuHl)%5z@%%Tv&T8(Y_!l`9
z;0xI~&3wF{=f^nnlJ_4!$innXdA9pSB=o`pJ>y~e4tw?~dEe3~6C2UoF-PHK3C6Gb
z82^dzO4nCOqtY?Hc9Kq|>;uQ<w!HtR?l4UB?)TLnZ{O;B#nb4M{C$}}FVNG|Fj6e$
zLvdl41?Pui$Y57K*X8l9qV)8OjQ!)LI-3hPbz3lI4uWAT(sY^dp^wg>3lEG%(XqkL
zblRV;qP%aiIi<D1zFv?IBM$}zQp?@$Z~d73s~1ss#vLAZEuY1a@BPP*R~XfD{i6fY
zFz#>-8?elcp2}R6HiV-LZ=`Ljkt5A>;lUMw<hkP%Q+Y9;jgZ1$>N#&$7>fAOj)k?*
zgKMt~(ra^9=Uet|;hy%+HDXVsZIm4IT9=9UAkUD!-+LLrqfRi~l{RfUUD`Ms4ae=}
z`sx`*t8lOfUE*RJBu-s0@1+kwm$^}ibe`$F9$BjZ&mHy1bEHa@M9RboG1da|IXARk
z%4W#u-hFs4eM4^yWqpG>q%ZAx3!cpPX}{m4{qT$u&dX2zVUSLvh>i{YAMMC*;pj5D
zz;)_VtzFHx1A!L<9Xi6WX#E?foquTG;AH8C==-5Z8EM1F?bj6o$NGWQb>AmA{xSSv
z>)V#&p0u}>^g?Y;1GLU9>XUifZ|NiGH~jgYqdk{jlD&pgOE1y5q}>k<c$-+;z8&AS
z#dlzy$rWKs=<#hCPOvP&;f$1<&vF@DmMMP?&J?}-ls9YWVG#2E!$%oJ{nSCOPnsA6
z*bb1fx)?OsfV^(xF)R!=pLH@22ZHd(2YSxM2O9YVg#3pxNMUE#oMRBhpoRfxZII**
z*!>bs28r<V7eEQd(qwSW6$CP`VSui?xEQ@>&UQB0Zmle95M=<)4_$ow)=x+WKq10p
z>w^p^U&`*Le6Sh7R4xXs`M@*6)WTDGs^Kh3zdYrO)HMgZq`sEG&hcH6e=5FF#&eJ%
z%CmKNP5F-fD9Tt{Cd*rwyiCv8FtDeKmN~i^X$0K6Y)z9m{eB0R>@ZO6(?BLg9E*PP
zUgI>6zAwO)6CrAhAZ$<qMjH3uk|GN%QgGHBm8V)I98mIIAck8A?nDyE*h3x%0wKCY
z&yyoq%Ict<hWNilFl5v+3jlKB6h!`9bNw3l@-JR-26;d+?htlgf_g#*Skl>*j-NYl
zIUy8x0)FSj5xNw<7=ir64Qj*ri6ZhaB=OJnK^sFLCU#aQ5(fql{)z8pbE6Mr<<$zH
z{1OwQ73Wa%kQD-VEO|-~Q+bD|f$OZ|wHkZzIYh=oaofhEvc4({$-bD#SQy75eOyp3
z$-?64BN;#Q>4rH&5BLtxydW-;mT0qZ%EBu+(RJhIm^c$gwzu!z>s42_@vE<G%9!uW
z<T+++p5H>?^Mx69dNRQ`#8K8C4cF1|K47lIMzATp(|6>~jXb3HB{L}AG0~_Iap7a%
zE0R8u4Ryb54gn1ogMIErzn7mlVcx;qY05Ewa~PdtWriN~%shkoVcyp|r~D+F4s)YX
zNe<?QpB{V`8SI2Y35NEB-kJZg;BD<ALYkrM=7|}2c&8vQa-&z589mqL!Qg!Q@;RMg
zD%GH9LgaY={=J?ubK_E!LiuEs%k%M3<e&`hG7wK5`Mkq8MKjgbKw_R)pDS{j%C7Xo
zh&8pD(*$4*!HlkTn)b~b>V!7tZ7>$V`w=y)Wo^pW1-RC3{4_)=g!{UnjpKOcUyG&I
z>s1pQ|6l-o`~H`<3p)u$ZHRG47Q#w)cIu5Xd@g+Z2A1)esENSUYn+05FtLUaanU>Q
z$Yp)DJ%bpi(6{T1=V1ISK#AwT)^(y0oSdZ&jkmVfrL-^b7Y(3+=Jqwy80p{jX{O^$
zt`B+2(4Wg6LR+kcQ%f(=sHHL2In`;N8THmgzNAlR;LtXa<FEM+Ozqn;6JmEp9$<1A
zrtmy(4UXfk@fjP;jg!{UJq6CBp{IXH*(D_dRvDc54i5%hx*#G_GoWR690Ld+ILE=%
zbip{`@>Pv4>|kQ27&~ip5+gM$q;)W?_u<N*+$J7KHg>;c0PH)0#<}AZ8}N+NTT>`E
z{B+^25eXcAC5Jpq<p0IOR)a%6bo^E(pX?UW`>NB6i^dFJCS)b!Sp)f_fb?zgDC{<4
za;|dwBIsrXy%4$;8rY&e#ux4u^yU;NYyC%eY-C_jlCIjv_#XD|n|_=;`d8l_j-cc}
z5Cezq217wOf%&Z#8;m^+bjaXcKv)5ik->~XmP+rO@+_dSlqg{^QwD_8g|w7np5!Zx
z=ZlZ<x>$rS`NShl5F>ek(V-7uWPyV3=p?2*B}H7E>j_dRd*G8EjW%u&a@e2pjT}GK
z13h<c<Z-%M91+TSbG4AuRhUGOi4Oxf+SL#F>1it=`dI4hhn7*~Ob8eSN>{VFaf|N=
z{rCa!yz}z7yz>~Rhq!C+iD=|Gy>_>eAg`mPV^WnOJ<N*RJ)P6?v0L38&p6-cm>LVu
z8F`%6XW9Ife+LA^i9D$e((JP+$s#jX7qNiJ{DQvS<wS1eVzHm`Z)wq*{3YXZgwvCT
zjZ!i;XEPS3MKCHMAI@d-xz7@sRHt3L?Y+)e7#8dKMTZ{vHY@sJ@tnB#$G{=y5?$2i
zSj?Ti*t}2cg<A|Nb4pp%w{P+cWsPzV!!}>VD1^a)Qwm<+TGz5*O&NT_`iqpe)HdyH
zXZ7Ip^+bu_TU(M-8uE_u`T~YJ(^E_W!<<U5X?5g^51&t;KdDUKMp>Bm`8CZR`FIAH
z|1Q(Tkg2r`PVd#*#Snv-NGOQl(k8f4l`@jv>VU`SN`0_~;5vgY?b}bAK$A8%)pl#f
zK5#IC!6Vj7e2&l`<>(w?Q3cPv%SU_BZ`hzBEwhyuzSVC-3I01V*y(9VY0vnK5xm_o
z#y~b`(9cY@i@AQZlU)M&^98Q!Ho|#0!qcYSz5ftowkzs<{N!Z^2V)lXw4~0RXYMa)
zc{%P}JXPl$*saPx<wE_TiaMk|@^Q_fvJ~-W8{@D9!&s4`CYInzX>|(Q<($gMLkVWp
zyInCalr9?BR0GS#NSm_WoY?h${xAQP{^tOfMnf|c)^&1iv}ah~n{19)$7AYVaCS-A
zz}ZvqZ4k;Jc#FYXq`XIXUHjQXlI1j~WS`1g&gpzH{A#B2mGHbI+oN6VL33-Z<vZ@T
zrr$zOPkYG3JT3^B5kY!Qc_opqr$WR4LI0@}W6dCsczSP%??!~4uhoKr3lBfhqZrCi
zaNPJJGi@gAY+w*TA#SIFI^zs8aUDvF@A*BWN~rKoO|lz;(}H&It~p^N^5PUcW&2_j
zd4$RPC3=Z)t`8Or(Jf-5r?`Nq2;onuzwov}Uc~naAx0he<B)|34teMZLlGai#>cs_
zhs6#=0uMP^OvOlt^YUsq!4O7n=f#2&JNMn_ME;x;Qj-&)x%@f$m;7D>OLfVD<85-o
zZ-Lw<$xkh7>L|1iJ+&-3e~c`@3MY<oxVsvsZg_IXZt3&fQ>fGIVGarHMSm~4R@T!l
zBJW!^d&`#Fk=OKe`L~d73mPu%v3z94P3z8J=Kq$)a`RZ{2eH2IIlr=8Br*-m)v3=d
z?xViubmKZ{g>(43>~B4PV0`~oHlpDE_y6>t)W7^`*T9!wa=78a<aw5k*llRXT$wj_
zn(F`lZ~v#3>Hqle|5sPGHVt&(r2y)gcRqBOhl~(yv?IpS@bo}$DYf&2_gzIp>1?rV
zn0Cgo+wc@t)Hl`kr}8S~Ze{ZLnw0&DaTlKHJ=#Ty<{ppLAv909`^a0Q+FK-|hKrPl
zVSyc*>XN5M9Z6GzfaI&uLtk0)M~nt){J{4-Xv#kW5qL2YXhiZg4Qqoqt`G6u+u%Zr
ztlxeIndDnsfXT;f;64OeCg{q`@e+<2i&PSm4N4ey#K7c64?{CNE3e511!O=*O_!)Z
zsm`o)Vran_M56pSv70Xdz*)<kV1UPIG~tk9eM$PhkEFv7(8w1&($u)*bVS|o^!Fih
z5k@9_;<^i0IC>t;t<EG5z~rJ~Cx$(L%4Zk-R6FmvOi{VywP@g?%N8x&M;Pl{yyV2r
zr>)ox35SfC=<=-JwZ8Uz%H|rM@`rU3A?M2B_2_<&Yc$8{J;Y%lqkkBUF*bEL_r4fo
z-IkPnbk>)Cp8jCi)FY<QJpKq}l&}Sj5z(C!ww3vier&6PZZRjE>wC76^+M-}8lb69
zRIK(7M-=BavYJreZ7j}3<`ZjR=8;?^U8|7?qPL+P&l!1Y%CS2FYPps4I1JzgI^|++
z3`d+i-6km!Vp%w+H-#M+Z>HcpFRdAQwCd04rHXihjFPNNba)Jr&rP0Vl=(jJXsK6+
zao_cl$34Bah-$n09jf<f|20}Y)G)LZeu!O<?M1iFar>P7I+!IK_+a0u9q)wrT8k3J
zu*xY3ZeS?NT^c<=1fzj6k+vuC6m*hG(+m<r0wZ1+XK(|sLB#UAzo`UxFb&3w4umXu
z(YbL4;?7s)jDUnwBAyz<5(~22U`04O2f>)*VKr(+pb&N#3^nC<utOM<Yrxy$Fjq?y
zs1Jj^M14B+;BEYSYRKXSsUCVdv~<idP3?HOuFK<b+=o+Nd;8G8P`!N`@c!alB(HSK
zop`ii|CRbf@EHfu*BrBKYo4t&IGrN0fcDe!BC3_SmY$J0n<(<u#<Nk#wK}|Q`iNz|
zgwIa{Yun)=C28Tg^_XjlXIxL2f;*(ld*Vj<Tlj2I#w}>H$eP2<@tu1Q=k&IUQwOE@
zy3?PN+u<(${nCK93;e(T<Lyxj|L#ZIEeI~#oKUxQOv34*b9Ji>J+UI@UZ4>+hND9A
z^g}z>oL=OWhnK*D8)ajGr?&Vpr<YugW5-#ra#xZy>Lq|W%EiBE;2;fnn~LqMM$lp0
zcep(6={3K0cXg@#*Jyb9Qr?a?=kG;#j_;QFYxL>-DL+?iozk(;Zw3nqVZGodE{2K6
zcLDJpPGk0mMJE6g9W!PaO()&cNxWbunZO!PNSU$aw_NFi1J^7h1j!SELI40j07*na
zR6HC4JyFdb|LF)MVoAmzD@=@x$dMjcD+~MtUs+p=1x`=JB*B39(>Z1|G6`I6+KB0b
zlG#;iuu!s6M1_-dc7r+&JXrL?$C-Y_Pw*u3QyM}G{w|ZIE1Z;te+tyo0M_Yfxq>Xb
z<B`oBVK@4?ES`6leXaX5C}60HRnny9vkLcuG_QfBhR=D5cN-%1Ev+T>zE1r@gZ3o+
zHS}BM$l;tX>8*StPEM~Br+rN=MjkT$DDk@uX-jqJWzu)6NA97&R&U2p744jN9lxpe
zehgXWczB<WKI7U#3D03md27;I>aYc73GXR-vHzKZTf*;{uu=XNzFU-W3mPr5<}kD4
z_gR3P;x=*WV2bsr^tDatE>!l?fVT_$H-CM;7n=dym-5Mw7U#0Hz@7ulJGmxpiMx^S
zDxRzB3_WWn&wErJH0S89mG{t!h;r!yRz;8qcPP(R{MtfHDLoGG=VZ0vW~&tY!Btd!
z5W^bHTKYlEHbg#m`8F)-W~BV)t(+k3f9@H`aZ2%1*2yV(=eo#sxlA?|_RW})(>+=o
zfm0#W9*_zU!+jb1(vhDJ&b6NvV8_+C7i19WVVK%$R!`dGfq0ijpo=&~^HsQB!eQoA
z40_-dXb)udeB-ZKso<Jp$wJ0<RvW4$rgS`)&i$SxWp@G&*S~>a(@6-Jz^cJVf*nyI
zW?7Lx0Klq8B*AncPqB286<!b$A5F640Mn&52f8vhV~hbqF~*rYrEDShCZbXJi#uNp
zgy;8sg*+e0`rd7wOr_`29}(*(JhqO03s#9k?8jSZw7_V2SO70EOaw~$XrVm^Gr@8W
z^=$L^F4Nxh&sMRSV5{^Xjl65is|aL5Y#1GhlO9OYr=Cp}VHRLRL3biGDu3Y9)mcb8
zG#L4*jt?a`fxn0R6YNfs74<nKz*=3j(y#G`?^@n9np>qGTDGm^I0rhX&_yfH=MUNp
zFo)_<E|Nc##989cZCJ>KUUGJRst%{}a-Mm-HkarOJ-4l@_LjFr+1$;qJB~ALx9()i
z&<>1VUp;<Oey|@u=2J{Vjrm8ZOHA{!w_F_%>E!q^Zz?>8H!Q&vrS`lU&02bm@|^57
zyi)p>v`X~1Oj|2!&FHeFY&BYYJ)f;@X06U@<srS5_Y@8Ca@q3u9DXT(iEO;69zPoI
zj5y|wMlUl4a;_}%Ys+F780M=wXO%ztZI@WkL`3Z$lCQXu@3hn+1pV~W6GTE&;Wf=7
znl5oW6(d9y08fblr2i!_bSfAE`EQur#nh+<4dBEP*Dz_8q<<xiWLvx=PlgT(X9Y#R
z?iM++<@Z3oK}4?VCOz<^^V9m2KvI*wM;|a6k+cU7jTd}%_a|wB6Hm`dH0iX5r^re#
z!9xc=uJf+`6ghpn0>T1Mo}Tz4Y)P8_5LrYnuODR6a1uPqL|%*YGPl6iU{Vn}{ng;^
zVmkG+rQ4!Dwv^|#)8e~W7uTi%`loz+O&Wa}j{a}Wc*P$P{G7fLKJZWe%s9VDc8_m!
zz=wW)>8|rA*V=HYjB9DN;w`EByiQtpq}Or_=qi{~=&V&x^yR-+oV9Rk^x*v{xO+TH
zYs&&&UY;oH)Y<~L4okSjQAf3Fw_EV{fXU0>!lU*1)Uk0!F!CcD;~D5tqqQ>C(o4cE
zfi2;$N!we0d(+R$yjF*6<}XuaTPsIip4$?i(#2k7I0esR)+@YPuY=o5$I$)j&CPD)
zr|NJjZ%!atttf(YGqO@emeNX8TH#8#Eon~Q{T7^7m>SKpy2?~@*ewRnHMw#L9>{UA
zn<qC{FJHX4divx^o~`jm4<B8<e)aO|>CMxtCy#;YC5BJVtIsWx8^PyLbND&D$EWbN
zTHMLY>bz^|o<HYL$v%gh=eN*mfywEvjZf)jFFB@ot}Tb-ZSj0UYpsmA&YX_>F6Sxy
zT(&$OW!gjD+KXzB#j04AL0AgLLYrPamI^sI6`5AyJdLP=B@2NW7J37NFWRXE0a$(*
zMv`(nd>2BAS;S;W2?7X?8y%FLHf9I`!!OpDrzLzNS@`R4p!<+O(u5zfK@(mUCT{d7
zgBL<|L2|+Y7Nd~jbtohfsktP(z~l>gpq1LjBsnq@fO;A5VE0RwSCkMKDHAH!#SF)x
zuqSe{wofF&N!hk)ehTis>+qx%c+Z*hUdtEr<Gb#WbT)J26>r<_AF5L5_^*{#iMOY@
z-%4u^4*ksOB&XprxbOLSJWr=?d@+ph_aDR4#yLLm(n77%UFx@%6TX*zY?*e^7vqzy
z^p>}UWzrl6y3{t2`xLV0Jo9+e$Qt8N#6O2{-p*>}js5>Q$UP-PUM7dFrF(w#m%v86
zrQuTBpQEwIsnj{1F~8PEA`IcRI`lbgPH$~|tsHClUGNUmK;FLcxKm$KuQgguv*yF=
z^`r7yb`8hV;n~W!70zkjW}l-T>Pi(^N^|(=x57o7C1I!U{u10d*ga^@@odR&m5b9l
zFQ31dCHYwHQRMeac5Hut`%V|(Kx*X^8LYvH_-n$fr7<W!)YVAT(huc#7w}J^n^W*y
zT31`h$ie|9194l>TUv%g;P;T1MU{M!{}371;JP8@*ENO?F;)nXI4JOSeK%y#F`41U
zAjxOfF^6M1*tD9b3k$j8Qr!EuZgL?6t{*<U`t9E=5;!%kNOMu2!YNCn3q+MEam^G@
zd@VpM1<1jHgOSGtc6u5|56}Px9Q|X+ibh#~X||r?ad`#emxN(Y@iBmjXR>sBSUl5I
zeGd2p5eAr__=42wB$Wly=to6KVt}KxgN~`S9?k$&uR_pb$O5rMPsCCi7qkTe82JAr
zU2qgu!Z>!9Y8tQrT-?btkea|O|5mtJLKlFG1~%3}X))jXy%tv9w%u>Ri7@oO5jTb-
zTny(hc^p_zb3fuAJM1#J@BH(4o^EwxP=p_(v04rn<6A7$=5^xqYksX9HF!_wJ=#^2
zC*CK2!}71KgJY_@)HY8gcg{DDyIyMPo?r6U;FRd>nRXjw+_U`0rd?|L$ChC!zqvM&
zcVBsYj^7;pIsCQxOXXb4=Ylt_fxQ0n_%!x(?v}K^{Q7bA^g(B?s4`x{a|vvf;VsFS
z*W4{B$bA*_mn=Gc`_}%^C|Qov1+d$uui8U-d12hPn!LyxHQ+{`n0Cx??YTOp4Clt{
zS29{47YlB^Q2}lB#s^)D8?4A#Z*(mNB6R|;7!q|bqh&@2u_Q=_+;E#Rb<st6rhGqL
z0}*u)r2qDKO2vSJwWC|HNeIdum@Z%lwZBey1TG24s&vQ{hY>1#eDGRi<=aXacD!6_
zjG}Dv!&DN<njRozsesuvZE{*h?r<agNEX$8V)2nocD9pdSD1ifj*wsL0Szqz0Wy-_
z&2UB<dJ;pT(8Zt#i0{@0MV$mZm{Nr1P~pJpNh<*$oMk&&)rUS5b3RpsOU7mz&~|03
zw?&#yDyO20z?Z~}2KLfG*CFl2pr?Dven`6G#iw(6$?tCHXmxTlM$77Ii}&LeG?qzy
zUeNcs{P(%+=hX_&T$&>`ls#DG-FF^8SMA>xInPyp_f4)ABhP))$&&Kk2}92z9q1vr
zQPEq`i8Ncn-+z3U9ns%gKnE~U-a~fLa~yVRr^HqhxhQWe$e}NcAF?nxcc*aVch|ym
z^>-J_TWdRc*;?UpyhG!A^VJ1(5t7C7-0jUf=QkH0OIXL)P4={U@25oEx8-?T+(#-F
z_%UGkLchk{5M%ZUUC?(}d}yi9GQS!}gbtmgxXx$*h$abb>?>{(AkrCs#2_JX@a&B|
zMyH2QaTCnw7(s|u90X$)atbdozWn~F=UX%m596O?RM|)qjmYUAHC|Cx;qB$aV4|m`
zC?CjVOW84Y>M0pX#(?r&e!yYmL3I8k&MFw_IKM_mNlBuf>`64Q<wO+@{KQHlgx$E~
z@Cbrz<RK5Z>L>YCP*KW5`Su7@d7d68`h;l%OPxu#6l)9*XUa464jyCemojHafip61
zRsN+VDfNL%X#jsn8NUDcDQ$lI9kMlYoEqm4J)BzhbLP+UafogY!9_28h?YK)E0gAE
z&E-$&IlT_2Kqm7<%+64B%CT%s2l2%AWI@g>^OU+o6<fR?N;L9%vP~Q%$aNKx(_cd6
z5Zn@<C8gS$cS)W@)0VW!C1pQUu2%eW*6}&Z)~eq<;B@Y?(NOIi+vFZPyU@I7;G%(P
z4gCEdZ?FF4Z*Qi73c0N(n{rgYkUvj9dYl){TE}9}TG!I~`i|i#9xZgH@Q<Ar<vrx8
zq}l=#1{B<(6S+CPFwx_;*4<|#4?3&W-&%ccg?C7u<}$>#p6A7QD_q3cBD_0a7Hsx@
z5E~TX1}8vked>}&xm9mVXlgP1g(5zS|5yvuSRi5ZjkD%Swi|C0mw19JE+Rl%>0<y<
zD~`cS{gA9Qm7fqHuJx8mfRV=yHR7u=LPg<Ef^i{yc}|5PQSud65jv=0An=J9iI+$t
zG(>Jr*$6#OlbRE}qzo&wEfQiF(xN8W)L;gti>l?cw%}=rjAHa@y(tChX-6!eb+8mI
z(Zttu=b-|6na$^Yr*Jy@A*}}n*JwBnampi$pHg0Jw-R9?lrVKb-l-f?o>L#*($Dmo
z-A-%Au7X3#a9fN#Ejn%0Lkq3#TEp=v)!t`oC85u0)oqE#A-K@v<fMNWV~*Bbeo4Nq
z(&Dh@Sp6V<f9`5JIQ5C`T+UH^ToeU4mM%x~?3|F-lb@#EifvAR3Ee|*=lB$w>;Pp?
z-UCL3NTmsmAa8q{Jf)6D%FGhZS+-2n<3`U&F|E(ZaTRpoc4`g$<I9&*i*PIXpYnNH
zW8^`}Yfm;64Lye{U=F{Oe@K5C2m5=pIG^zBLHn-K+}eOrYai78-t~W1>!_t%{;ay+
zg0lt89(6ibhPnJL%5@ALxt!<Hf5k=MnoPcUpO@i~c&+cK(e%kb3;1?I#UP36FCh|k
z=Qtk6H~3^MiW#zmXShg~3<Z&x;zAsT8nKfw!64yQjtotJ#i2x3WWqOwo<5@yHzp}D
z2qbv*h?B(7!3;W@uY9C0&K_KhJ6!AL;wY~E=?x^b1@aSR4Z{jrhG&bO<h!v@jYg7y
z-d*>=8oY5qH^j~OWWMW5dG&vK1Phrg4lHTT2cKk6oo`;|L}Ib0;w;9xI|ysbuXV9N
z4bI2?1*rE5lxJ`&w<WEuWZIHW?R~bEN9cOilDJoRt<Q2R+?4X{`p_wewmf9nLLRtn
zMW;>jI@GzyQI~L-bDN*DCA!XrWvYbDjnhNbHpiyk4NJ-hL*HQ9%=`errLwiM+KX;0
z{|nrCYGCR)afvT&$2D<({J%YY+FaOkYadJHSqk5*m^%(QGW4LxmJTJ?3kxvaB21Q<
zS5A$7gp1*fT+^VIW4`C+&Ost%;OE!goSTj=@|>3j{8@GXQ0e>`IK}fhSBAO#L(2~S
z6c5+UQa)?(mhw3U{5f>nTpz|YT{t%q5VpmBO(d#D1pZi^dM%_H7SD_v&`7WeRd;Ix
z1};Bl;TksBp0>%rwQ2TE7)3I8Wrq<d;e(`-03(@Ekr-u?Va87NAP}^T9+<NZdhk>{
z@zd!bGvb*1(C}~9XuiPe)dysVgIn)RZpiX3>co}u;9!NGEWJO`13gaUu-K-0Gf2gO
zgTh0j7fJI>bfZn*qo-IUYV-x1EX5PL1m)*>5I`2n#xH?&;jYX89^l})r$Yzl?ADfl
zDW1c-zokyjUB_FN^|sSG=elwu@){klDc2etmkr!Q@SV;fe(STG!?nuY3b&*T$db!5
zg+C?35<2kcXe~vY5>WCsCO5<F@N6H>L`t1*V&aD3tqd${b?xQb+Bn@R&yux=R+?Tt
zTW>IRx+Q<9Y$Y0{w7uw-c-_`CZ^H-6Kd0=Kj8>xUUYvs&>`35LI@_^GXGeOjxkrIz
z_8p_UgqO&J088PggwAR0U)m{Y_RX`Eq`)IDe8`#%vRgau*p9%hG$ULLM_PoZ@*MUY
zkNLL9?eg#CBLly)zfvGmz+HX_v!dSVs6PYe_<WvP@vfUzIehop&vZ><E8YceUIV}6
zS~WkZp+*2fiG~L#bcO(_xGpS%Bf^*qygtF#H1YM6N-*pw9Wfq6g9w%chc}+q`>BB;
zeJ2LI>xXiAw9?LdM@dv*@foaPlu?4|b<gSsadqyeXwefGrAMa2LmM)bSJITerKwO|
z=?w;0HIj%=@u5SReL>U3bv-Bui9iZ%XU}^})kivdQ*lJsiKfC!dUTF!{47T>HDGv2
z-;&bFB=js#d|{|GOdQSV<aD6O@Jz!l{+Wt*C&aqw`1VxeIoFl;sVh~vrd(@qTs&|O
z!FM{R@LT0>h1;_XTgtE{tv$=UH|@dU<-}}n-E3`OS*oMG%Qq*_ZCiVoE6*kW_!`)(
z4G*UZ<ASVYv=VL1{~o1Wbaq~q-LGJrmt1>EbC=uS$k4Nw5-)U31Eu?&&p|hH@4Iu5
z@gl=T0~ZbS8t~m~?)N$88S=hws;@bxm#rBQ&}DCUv331qBruDG<n~fKjv5jqnRu6`
zVKG`LMgCwwxW0PuFyRvynr=ksm(QHOaZ-bcobti-h6#x~nC1`>h8T79CvjFav;Zc5
z2VdIOsx<sh!Uc_=@`1Jnv<w~w;~HaBIuVsXR0kS{6hPvLyAelqlt^kA(n)b{41zDa
zLHV&O)@duP9=uBjz58|ifTjpA@O+oZ3~_D*GC2&cBdWOkN%(zGLKk<NHQ*=Gk6DJb
zJlEnKvuqb$7Y%H!fxf3+PWcer<=ie9Yc$|v#5K4V9Miz1onKnQMFXeRz-@X4Z<DNB
z)aqR?@3G5wY<^p`f%`_|B!-^*+%UGf-+AkXnwGN`k=?o;r|7MbZ>lg;yr%FjdA$Z&
zbl3vZQ=AU{N(;Z!P<JnPOC8p<a||Bd_f0*^$B?6@15alS|A87(TpIJe`MKdi4Kk9L
z@4i+WfS>^H#t3{1V>C)Q$yfV}8+bqxXRu-vN~cLY&3XO)_4DfY&z}mbDEtA{mFscw
z?TRx6H5f&ZxUmvn3_z343FFI8@q`hyy@1M3GgK~>P3S5sg`-SnFw?Za&*O$|8KikS
z4E3rtN%Qaa^m)L`m1>gbqLAVz21(N&iQ*y*<<PvOV~M7^G2Y}+$K4MAaegvMX^9|1
zSq$d#_x5NYRcrM``kp1Jor0g^xfXAZ?j`>`G|=^{2Z-bC`T!EUp}K7N9&TagueT}K
z8od_WoNjhFG|!%l*W{Q)n#tUfCbaxLIm5P^z4aB$K1QrE+&E))kITmVjzMloJKwTB
z9iwdL#$yZpEv2y({@iqSEO}f{_-qXG_<%oS{6Rt%Svk00@(gAi4~kBfr9$~svR1^|
zJ^8Me_t@n-Hov6}WX&_Nl*hTlj|@G>P|DJV;a3#7GPk<p*6H16|B(7#T90#jUP_PM
zUs3)Tp4%C=l-Hc>_c8xa9p=yVZPDSe`=3Mgoa;3{C!W{B;~3?ZHTRP{HAqO$!Wcgv
z=X>16RG1{O8zG3QjwRC<?0O?dZx~Qq<zb8*#>k~pL`1}*peKC<0UR|jXu7?=HO?Ow
zN(mr={nQ}k@CMxFWk`xS`qPt2BqbpU|67ej0N|2QspBEkLxsAL%3zIzUuCCoi7e<W
zWRu4aR`lmA0y!t;CXsjz2>LTU8NBOLMMb)V-1#X@^&_}y&=D9{qr*Wb%87hh7I1(U
zx$e6|wMo4vRRt$qmiAL#dl4^&4=GS-q4?1E-Z|*`5S?s=OC5dxZpQ94G|yU(-YC};
zEa7ZGV#Ix`lVm@o?r_MsRL;42tI?aoQ-QKShji^VIWiVqv_-z9G&|yj0JP@jNd>e)
z(xHZ;0kIQAX7{+9K&E4|T+^Of+UYUNa7tcl^q0%Kl*Ur{oZo$kyPoj5F45z8`D^Jp
zoX3~8|9mXTm=FVbp6>Jy@t2eZ{(H#KBLAuQw#arLVcvfKwp+3P?Zc<5zyIUyZuF-J
z^Q_jMiKWl+6rZ`g;Y#%w`v3XYf7AK%UV3_zvvj3I1<^4v;EicH-rBg+biYI=UQGnt
zp?BO)@uCB+4PIRjB^lSGP4QYoZwZd;!0Wq)3mh)O?IAadAF-Z|Yxo-lNXhx`uLzRo
z1xlL*U5A?yJOZJp=VQnp>m(Kr-YlLjaq`&|x#nsb;}~T+wEvkdL0fK1IcqSkb@G?E
zW0bo^rdqj{l%)k$uT)FSTxa01xKA?q-4$TcE`8D8)3C_!f`2HJ3=9L^6>Wn1OZq!+
z|IqT6D<@YlEU*yfMhnH2XGKL(K?Fv=&x2t?jT@3CD3Y&#l-IR9D7j(94HCAvp@am9
z1-BG-A&^JI8n8bUuILi337`;q8cLEWBM!b^7Uy&=C#H1pBOu;ayI#T3cP)4lMo*mP
zBMtXZaAC)Lh#&l*rKf%*|Ci+XI~<)TUBPwz8&!!5k>C26Bde!tzd}!F=(I4!p{+O&
z5=VWq;E8OCBgy0tZ*fatTj))}?D4KH@oO!633tmhJL$4cn%KX^@FDU^(;c;q$#G7F
zS**vF((T>GqC6FmyW}OX9g+m*@aR+K>S;<3EqaXd5T2r!!|oYJ4jf&Qw=v>grjnec
zG;nZJc?MyBTTYTiS{`=aY321sKF;O0*zm8@n1UH%49hizSK=SrA7S1B`ab97v!T;>
zose{@UQT%8{9Q)a^9biI%A4oa;{NWf<=^G>wMMS*-+x>^emq(_xyz~=l+D}eAaP-2
zsvR#RRRuo=k1F#@M$0og1>=21h5YWxd+_Huw7H7(u!Y8@-5}|yLN!QO6S5*l0`%GP
za_l#{38th$k2!^s5Oh#%t|IUW|2K;b$?0YiO?M2*kRWpy#&Xn`&nft}t2=1<91Iwj
zL+3;KF0qqFHixr!3%J1PoJM7}GB|xJv%&Uh#u2^vON27=u40bE+{s}nhRftSaQQ9I
z@F3;59nnr>#t8aOmlWcbqmpRY(=;M~c$X<Z(ZBO${6f3JA%n}?)9=dx7;^O-6ixp)
zJ@9HU4!uV^cG}1GO?AV%XqDmn)z7QPR}Zh=Jbp4ldH3Vn)$2!=4v(nwJU&7W3IF`^
z`PJ3St1FC2uZ7RoUq7y%J}AFGo#&Scm_T|q7-i0Zr}A>xL*pNQ{J47c=y8?I|MnmM
zS@ZtoZ*QtFOLU$SDQjDxq7rxzaWTzO4*%;k-JheF!>uWNq><ytL3J*R>mg5H6Nm2<
zuNqD8xH^VTI+mjbUrTSrEy1mkyC#RH&n<rN;^HN6441cw8qXt%Q)9m>op$4{4|&Sr
z{u13St<SkOg7kZo>2R%VPobrpb1>vnUx&KZIxq2I!8|S=@tn_9;#9d9>>b)a*<iir
zqU}`v_(0vCA#5bS<WHPTm_Dt4rk}=fJqWno-n}pT#ja*$BYio~rDTAiSA&t>0lHFr
z;vjM8Vc-DA*ciMrMAm-7AsI*^SE_lM64$gOoP|{w=LMnU0Owb9!e7g2G~r?5B%ZJo
z9sS1oqI5(V;%nIxy{^1L$NE_zA>rjy9Sdbh{FJhFWtRB$wEzk(yTs73<_VjQuE4Oc
z`29z!1Fa7b<%j-~#2qqnkS2Ld!mJ*xPQl?DL0VrDjs`oS);TME7aas(dn|Y8vv4dq
z*YcUt^IDo);~iSxlyPZ29?El02TSBx0vq|x;mzekE7!xJ?IouNk1gIg(1}j3zN}1!
zfpa<N?_%6y_SYA(eKFs0B8-Rgyqca!XA=U(?i{B?%VFFH#%bfzzw1JmBPZWsVzSdH
z<wZCTx57ER)Ax96Uwt@-a_IHG56N9#=g{-)X_M!0eX6CE;3G^P_RR|3K0Lofe~&iq
z{7&Ul?lq??k7?v7)y1jwwIyF~<6F|qX&r+{&a)Jcb=dNZIt=}H{%m?*QQ~t*T9k2$
z&K~@&jB9FZ4kzV@4P5=3(@M!)fp_S`M;R@7zv0ZNTi#ttpNQVPzJPy_@Ccg=bO&+(
zOJ2zLu*2!gJ?2Bv<L(=m{I87*3?|B_6nxN-BS}2sQ8o-Fy^CQ|=i()pp}w4UY*)R!
zPT%~YERNaxW4t~u4@G`?-ffGUc3fA>q{qlpb8h4zKc}9@YcyiI8<T2A9$=m46ilSK
zRk*eP%lZBN|9f}!FaOKyJb4tKGV+XoD+z0K?xd_Y^Vlu)DLj?5)*Yg}wLA~O)9;js
z(7d3c<E1m@Vm^@ml5!l<{1?+ad}@F2zS&urC2d02ICqvOl8^Nm(@z!7<<a(85r+0f
zK0j}RfI1D*+<9~MJkn#n!$p5CKf>0+ThOS{zN_g=o|jw)(a1yIq0f0g*vmP!$FIf8
zpoa8rG$@9cR7f>ANFyO?i&WxECjfl>v3TXq2dLJjjh2QL#^BtDM0&(k+A?`CgGVw{
zKseFYMY13p@^!Kc%rz4aDUbR{O_*|FNs%Y23%k52T>xN`;9xl6Ivl{k6Bp@gf-UIj
zQ!Isa8QrHmh`>+N#P1dXg(m4t2>EH5@#8Py0SGLjLk9zo>r*FI<jX&=I|Wu8(VHTg
z3`WC@JSo}gK<b(}^}-Jspq0BHVhp;^4vmbpGOrPPKlE^OG^pwpH%G^5`1{M@<NK$F
z;XUTDu|JhJhMrGpE#o|;<q)OvFQ?Lu;regKn1-CFrfI|lxpVKFC<Ed6gAMH^hq3;1
zS_sD}wKvrx4M%=C{gyhdwK>wYPn{O6%Y~QA>4{DopFY>)5zf;bC;C%)z(qO{FNRAx
zi}^8}!#aKUBP`)sdJdPz=gQ^qz~wwzah(Rfz2jm^KRzF(<_ms#ZhO3??bWE2+ENWR
zPtWU+{K9KHM>~0ZO}!o>bl$E?v|3<Fcy~R`_2Rylw^WBYOzFL68;tB(2c9<U&j4``
z!zZUlTJ9oD3<Kvfp5jWa{5i~J%gf>M(lZ@(;&9OOA87;nUT{f=-l!gQ9s^dv4G0zT
z7bGF~#qFbkyQ*0)*Qa0Kum1JlPjfZep*8le|Mo|bgqN+DSx9{U<NGM}y_`tGy-U~9
zK{L|zF!}sAapXAU5oYwN;h*2TJZR#;!LSP+Fdqd5pPxI=gBD?~X#3;qSI=(pBfS#B
zR6ZxRa3e#{6jF)FIi!WY`W(dBRR-P%>>~C%uANz&Y_dx!uSfDd4Sf-hdL!&Vry27-
z?Nokz5Bg3;J@Q)+h?ge2t{Jq9Ufdb{p#z(ER5v571F=Ksm+nK4^_=rNMLd@$kLNP|
zl(xx$iu!a|_hVh@Bvg<+4?5&e>2r(rx&@8A%z50O!#uq-*sH<jxTlEkA)DSLc8bCT
zp1*&|%TXGFwS|6RzX!4Yl{QK{#-PBY41KWZPnfn!K6!o(n;Qkd@%D}Hh6qn|!-2xN
zAt533xWZ2&_7fTj4&#H$V#25~Lgh|~&Q}GO66(%z^7!G@x03g94VC{SKAt%^NhFh`
z{uGhv#^CH-1ee5BlX<V93oR3!1oG)$ffKt4#uI(FiHwv}orJ+7-)R{QGYkRa`lT;X
z{iL$u5!atYw4Ri|Af-Ijm;AwZeIpkDE+>BMyO7bZiG%-LaIV|dvY;V;OC1Zp(v&+F
zf?Zk+@SbmRb9B7#?Tmei=x~%|iO9#V_AZL@mt?EqIqXthJ6_4>+L2ZrmGI;Hx0Jti
z8v5U|{K)5ZOL(pw0OM|n98-1b3wF?T{`5T_j_HIIKD`L1WN>_MZ#k^?iP|p?=L@Ae
zosu7!aIQ!CH{cncI-c{6V^+rJ(U4<QhJS(a_DOksY(IpuXAL+3S9rqwT&4&Q9Y3uS
z(<5x`uYqx0MtH)MOL}@u_OkJwK+cjeAAQ97cD<C!<#|fH^2cG^cNwC*p5}E?A=M>2
zEq&HJl5$Dohx07_$WJp@{wcM+R#M=umBZ^W^5X0<@9x{`cn-IAE&6SN*;=mKK+pAp
zA8Ur^GJpQ_x$Mu0aSZf~rnFb-cRbGX5c`@r9f9XAhCK~Gzfv>i?G^Sr0Gz(d=KY$d
z=f1!9JRP0qeFQXIUryJ3$8#Tg^nLL?AsvKgkNkSh2y`|9xd(g;7<vCr+~1uovheP_
zkFHkVw_O)ruICRPU%mXxvt5d_rpc6WN@)oEAOHEUMY+9f#mqut>BNll^k)Q`PWRtG
ze@=!R8&7iDh`G|2uRpH%rV#FxTwBM81Dr2kxO%O>#*Tchq5JjY=hdfopB3i2Jdgk5
z&-WUJH^!as-+pQwTzEXVd2;pFo15x4i(F>tyK|X6zQp$vTcfMn4k5ZkfOCQ6h3f;`
zW6I=ureZzZ%Aac!a!04ny`J{+Gpaf5Rz5ji&g&HMD7znLc0ZSgmu5VQOL=h6$jf%<
zz@}6e%1r(shy3wq)z4*e!2Gm>r=PQ5D+@BC2|k|AfF>WLk*{kT^^>H4oMa%Xj5qvK
zYZ2OBGCnof4D2M)Lu6wBUTWWQ162)X_rmhrienPR=^pdth0hgX(sq69uLcA_lOl8y
zCb+=yPWdhAYP_JaCp<`l?=+NA`U8IF5}?tPR6%fDVD#Fy1g}OO4GVGp%MWK$oIdVG
z6wP-a5L6i9jq(YDuW5;pQmg$zX5bXbd=*qy7{%u#ks6a?#U@f_H~~y_CH&}1JENG8
zLw9u8aKLcFWkDE<J100QuILvPdma$b!q*A50I~DNCyrV-QdIbIqnvRN8q)W`6JGj{
z`c6*yP=-`Yt9ywE2(km;ISH1YWzUCdfcE6G_8BLwDN`(K49CLKbW87p2vZC5KH0+6
z%2txgC**PLKs&3AM{@ZhUY<vMP1|5HQ_{ub$1^X~k^m=4V+@OK)|7ur<XYWM;gGjP
zj;T7$+XZy#B5MwK_8QR6YO?dpX#XihfT#V@uR#|)<&he%^pb;dQ+%&>VL_}tpDt%7
zix7V#a}9G=5c^BIEm#4Ym)q$?9~k($^SIOV^nPw5<t~h=^FE|Bw&>&+qNRLz<vKtP
z$8#M2CY<XUJo^8rXV)wE-mZa*E-z!u^RUCX?=nO=J#DHDIz4=+MYuHg_59oieyxl0
zuBCAZp4VZdVejm=jRFzxT*56^S)Yrn_e~C-9qWfBL`UHdQ&NPs$Iy@2iThqlYKWEz
z@ediG`ZG#<WeDnPRyrn2eS9Q8^{lPU>m}PhWY6K4u9th9Q%;FGa0u@ksSEk9{eJ$g
za9H<CeTC;mKdC+@9TcA9oA>X;sZU&~?3UC2-CEwBavz#%Up>@$g>IwFxd#2pvpoIK
zcuTv=d6aNs8snteGoRD?_m7{i{^`Zoh{d;j9?KZ?*~In6u9Lq#dpa7|{nN8&dIATz
z$jAND^D!UIkU#$%WQb*X^z`x7&5fB+7`y(5zr1o3H4MGVEG#!Kj92YjJkV}wj}a!y
z(=pEiK3do0)*T`+aaabi#HZX`ehqgm9`xwg+gyv9j1pz7xyX%hjMs3`;T-o79Vr)!
zVfjuD7W=#-=LG=WDcX^4uB$1$T3JYsG-G%Q$Mbv<&f~ccJ*UZwI=U<xvosuav9@d@
zqA6hudU3bUTv=*zEKQ%&_fmQeAN4^!#qto=cuaIe@#iKkEiZ=(Y2`AN7sS{-crJ3}
z1+B^DJ0hmwYq&X%$9(`!JmSV~JHj|#?nm4_Z*^RR7`3(7uoGbx+SCn%un2>hE?XW1
z!p!&FAIK)@^xy6jE_HIA4xZrxoCpa5!H!WRkn%L2KL#is2t=iXBzJrzq6{=U{TUQ2
zW=w<;l=mcjWbun~r~ygRpieriynKMA7Qn<+<q=;&$r6XG#1o3it$0=jXek>_10IU@
zZY{MT@}S3iR3Q-y8RWm>5{xn=VKhw`c^-VV4JfMoF)B(2d2WMi71Jl(HTb0Z$s*Ta
z(D|M&zF(*KFdoaxm#Th%#)Ledl(#Hu4~T?_k1O2|S<S(ck*68?#DK2zd?CZrk11Y0
zE?Of)jfT&eVtEX9$i-uM8;f{!+q60QH(L~>?HVTUF6IZMYoAJoE3Ru_DP84B4IDcD
z7&`yRPZ!EkLa9$B3m7$ch6kGRLC=5WJAKy^{b(FxAcw<17e{&B(HHqM=>1a6ces+D
z)AhLfflUcr)8GYvRELiYo`#OmL#`vIABlPzzK3HzVV|4jFi}p`d#cAApLo<Qapz@q
zVKJ|#m`}Lmq`F2frsdhhqr4;JkboTMFsBhfszKLP6(Gy-mbivj3r-76CeminoVUlj
zK45~A!%lz<>Zb+1L~AMDQuq?jJZ(#wk(P$#4{c1xi89nIwa@OIAkS~a*FNRpL+wZC
zTXf7R5cZr|pC934xc02syCBy<9|te?75ZN3vsZ;8JqUPT<v!(69VP0NgCF_cr{^>h
z=*R;PSp3}4JyH+2$S+?<10sPL9r}628vlv7@a5UE@(lGDpE+pN=#4qdr+8g&Ni*k~
z)6U~_@Oc1<Kz6_W*6}%ca(+3y>kzm+FUKvxwBVHJtW9%%?nCG6Z@Z!^r|G9t^7K98
z|MUIFCbIoV)4;j%%v~%KY0=0--rxUWSJYv|DH(Zk{LjCBXr2DOaJ3y{4X(Tc<{zKG
z=+$Z;k}O=W#?@^2=+Lipb2;7TB<5E+1%#u|{x{J9{R?dhS)h$`{I4IsX#D?u`$7G+
ze`1?$p>h46|I2^Pb<={k84S#q3SDBp6`d08W2a4(g*;ujkT+L8G3MnMCx>%%T<;|~
zPji1t2Mku`7U`$(Tk=}uXrU8fpow!mSnTrm7#d&@>67$Uehi{~@E39RY8&^VUDR}n
zeqFEbN1g8#W>B2>YkTpk(PhwX#sxDx_yob>F-EY+#?>wIZtr0w!r)(xDKs8YoEico
z2#zL6yO;lphmV5>VZQ>-<CZIbCDV8Ojbur<2Ac@hdNd3$@nlVBkU=2X^AY1tLxsU0
zQgf6F=Y@d|^bpMP5{!BdZy>5?!6wypW$2P(|1DkVB|!0vo$*xoq+iWS{HY@|=3qpA
z&`n}HJxjjI3X(X`+_}Mta;R+N^RS2`&#tH=2gfO(2e$YpF;CU3!XK#2!4%L=nzUBt
zwRk5j>^)|F*XnDkJX3jFHi0#KxJDvBja8x((_&Z`_2@Ij5sveMEz5Q&EtlDSr$-zY
z`$^^*`kL}g$<GZciX?v`Jr7TfIh|zW=d^r`q(&^>84b?q#rKcP0eTGQ<fif*wgtvv
zr+gMWYC4?4%k#F9C#Rdox8%conG?L{VkD<*xV9E%u&4O@|FidIO>!hjerE!exUX*Z
z3}?DmNX{qO=M~wn%{Njg!Z$vy?8<caaQY}z;YI?v`cw12ySj#%U*QoMsDc3*W~R5P
z9qw-G$CRCW_D^oxDP?C}zl8cZ_GR|)Q{J~)BXi06#kgS|<|+R?x8}wa*MhC}LC4rE
zbB+CXO*_^8Dcs(6ql<(uquy9{f(!ro^r?IoW-l`zK74rf@k0p?{Fi_Em%YR`<!te{
zc$7DVpBce7ad=%?PBl9-EZ4J)TAjcUuXC%VE%KtA8h;H2&;R-#|IJzd>;L@ky@=5h
zJw5b{g}oNZD2L5LoN&o$%D&i^*amp|_d<6UaS@*{=2~*#iMUvnJi(L;Ud#6g@^HYO
zN^uizif#&*HPE&t?VFPMt>I?eeA4D{Vnm{^*9LjuBT`;V>ccv=#xW0*%z4VYqu8^(
zQ?2+Gd@*JJYMhNzf&m^|u(9ev8aDhpM`*)8%4KL1sBj)~kR83`5Cnpbq`U-$Ot$GO
zK_ak)5k3z%SOJ~V3p^*MwIAe_zk$N*w8tKGc6lA@p1rG#0EO%8gFIBjX@LwK#HA7q
z;t!+j&*^anK^_@8&<u7O+LHUk+D2X+jQxf{l=}<bE*Z>mh<82Zu4<u?;c^Us2yL<n
zKk|+7LsJBbqEwyEK&X%n^1V}bx_hp%i9NVgaJ?<$iQvJ+O&Ye^yp?a0k;b&OotaN{
zhGk5gTl!x{M4}Uv(b-B<xE4<fe(wHi*>MhdtF5WaCr!~XpbrQH)XrmDdL)>Qu9%99
zF$_F%z@ZDQ2R=Zg#CaC$Aolo8VTOm6O)HYdOOY`iSI|u*$!pQTo!ir*1uowxkF*)a
zbq3csdXk9%MNJnrMqZVt&N-pc`DwI)vZ?XJIl<BdC4bytP`hy@N{Sw*9C5YwwQSJX
zXyG$F>yjPHMPD1!!YPk73wDZpe?7Uir?7fjrsgAU+pB%FYR%Z0ws4GX*@@1*>>q0^
zwNuKj>Dvpob%fRC9Wd`(%jeP_`!aJnPJ#FKrDI}E596CL!Eudp^ay3<6~{B*GqdA{
zUhN#`g{~KB^5<ZZ2VIm|<5PD<8*6lPc=4!xs&CD=&Z(t#F0~6gWQ%fmV|l5pEp&Ug
zXNwI>c}{8ToWFU(MNiepUhyawyp?DBbG^j}JnI&1O-2mz#HWp9CxIX(V6=_Dzjc8%
zf1>Bp@I()F60egZ@%lLIp^kjnK&~fvoX~!dFFdeTJHXYFYz8OZ2&?f#*ybmCFnUhW
z5}Iht9_72xxrg0Z2FE$y)anp>w~GVMlIxbe_0F|3;-l_Zjx@ivvY-Fh#3NV#<*_p5
zlQpzca+ct7pK&t8#rE8~f$YrNf<-$xsryz>Zd>!Q=7-qkW6Q_LKZdtuw{Ilb0Lek?
z7X2yUTsW|)L;L)x5bGS)Vn!}=PUG0`=rS0W7|bh@=osT{$7#N4lSvz8&^gYT_$l&%
zYs1N9c5<0C1RF}wsqi{aCkJ6pZ$Q2=gYjbR5eRW8xh1bcMZl!URzSl8LN37MGB{;$
zFIBRNpmNCtS&>%^K&Aza;3aiBDB?@n>FdFOE~F*`-+`k|@=0DhFiT0+$jq4IbNUf|
zk|#duB<N+6>X|n7A@abR#1S8xkBID=gs6igJoYo5zLk{yQL~%5ahG$qX46_2n<;zT
z<-S($w@7~%zS20<Jeu+wbF9o!Uy=L=KIR=ciFwa7EcNmW1U_oR6fH1#Vgj$Z3_bSc
zG{`UVJW$P|L+X@BkIVRCA91x;_mBkliv4utouGk0bgq6R6O2P$L_rP5G6!})k&<%p
z%z!L6Ge}cutC}$w79@6d1V||cxc8?BI?*~dIJh(2Sq9bO80SH*5E?J4E6;E?{rtFZ
zGWs?dap_N6=K*%|dtT`yhqhz`zX3!7JL7~T`9*fUPE+v%dZE*?2L5>5Kqj<^R{y{^
zT8GhXgx&@OSVEqTX}ujFjJD_=+l6fT2Qreksq<Tr?@*WR1nCDo$R@#y&Hm#$yZPln
z2>731(02(kWGQz)QS10!mz{pZwmGW*G1>MptnoQY5$bY{4ja|x==08mUwnqE@WoxE
zsN1{Dj5wz)QUBWZPw~cMW-Kq&fo#UKj$v~?iS$c#oRe<rbvx}M&9%Q!4eJDTUmJ`O
z8JR!!$((Ik&yQUjkaLWTT07M4seFoW3O6P59Joudjq4qbQR=&I{fQp+-Hly=U)xu4
ze+_9(TR+(=@*du+fA{o@`OI#e-Nj{V#y78>(|3kYUVEY^eJW_Vjd^jbaW0|Bd}|t*
zd8wmwN=6L_ES?y%(RXbh`Hl9+3y&;=I(HAVITkmgGM$lml0GUr83z^<Czv_;39G?n
zi`oiZv_JF3x;A3+C2XX@7M(~`%!s5%mt#A#zoKp(n4`i>A8~h4*1=jnK7|ePmvVH(
zvUJK8)%!d;wFjQ2kX13(=`;Kj=gG0!<?pRQo~`4k#aB-LhKuN__DP0L6f&)R2pviF
zQakUjY(1q`bX?a4bVKQBY~qe+^R;z~0y6|VZq50nZ_OCW7FhJa2IJZg(37(kmq3YG
z=Qtbl9sn5{!FopW4L>g%FEso|799<5N7(m!A#vWqUy71P&FvZuD7el3o~W}AsO#pg
zZlm!sYE|w4j*tb)jipZ}Pi1G}r|`P`PA_7yt)yZDz8drz-b_*Z?K`vu1qqI~^DD?*
z7dR`ESWCEjulT&cpn-=?8WMWs7w2<cgmrlOl5?~UuKw49Nyv)N3G0o>=tg!73>1m3
zw)z7-YvHJis+6Y#(#$zKqjslXz@YOP@kTwdEc~8%Jg(@NfP97y=R2}Zb^y<P050}r
zlqp-GA*mlFFXr4u$~mM`be-sgyQ_Q;JEE`7)sAk;_T#tKkFuroZ}5I7iYDd?Y{c;c
zUcO?b%Ob!LTkqk0Axh5Wfk?7NLH))SnC#nZ6ZC#Ff#o{4lGJjnDrfoSo8JW!$8)i}
z!0{n|6+g10T^=kqf*2@(#sl)kg<R4PR;EN1d6ID;HvpWbt23u`ZpiTpoaQYg;7cm&
zr{wozz=#Cv2>V9CA8{H-q9b@u%J5PzI5|i(_}MY`lYOZE0xzrcdLucmxS|_hkVZB*
zWcZOxd_(`Jo@J|j183<s{0pt@2QS-2>#{n5T_%*@L_uXNzu4UqsZaxtU{)E`VXVO<
zaLDnoU$tlcTz8fN{A?Xqd*W)4T1vUi8GmVYpC-%PHfcI@umsR5-7?A9M#eeVrW)(3
zcHtEB3{E!m>qdosPN!s4SbGq2vd|6ll69T_VeQOoe9t&j&>YWX=u+Z|lOM|B->(V6
zW0%%d(JRk$UvTm%T%=o5=DNMC>w}N7uHwVAem9heHlt&IYCh8sDrcRUKi1{PRzF1A
zsj_TE&N04?ZENYP$zQ_Pl6RNzOR>8(hwjohE!`jNdohphUmqS`D-3%CdC>RR^c|CU
zYUcU%gK}J7D2S(NUf4H$mZmjrZ3TJIwO5b_S)VPS#F68JT}Q?zlfcCUO$_ox9AyQc
zj_ZXib%**M6L?BS4F@bv*xEZeb;3;(vkdC_WU;oW!b{5f1aoBL$938d2Zow0(f)=A
z2a8U5jTT+;;1_A)-3}2>nUf~c)`Oq7v4}YMt>J301I9A~)b1(Yp35BBlJ(uS&(N<6
z`D|OP*X(SaoF41%nyzR|tK8x%zHed2(w0#jt~p~V=c#2|bA1;pIP*SOVcU{VW`iw&
z3OKlS7DTtL%dC?LETCTjltRy@N!hIOAW_Ft=nmUa*4Zwg)2*MPLY`75rsy)T%yEAI
zDHr6T$5!zOra+WmxLYLSp1#Y|33F}m!60ULFm2Z>ejEUS#CCbEdy6|c1!rFAk1L4R
zcW(N&*actod3>z`G<f@~*)2`t`)d2?UJm}^m2}WK$O;5}d{fi_(2}shG<rGntL__{
zMp$^%c5E+v1fzTqxRhl-Wz;OAZBb2~Uu2%zF?(o9rmoi`{Tkg-T9gstdzIsHAeQrb
zEISH(T6>fuO7RP)ddzF*keOfgyUJznUFqLT?~;C8qAU7CeJeiUlnYK?gVnyV6M14e
zpV!pnL7!~`R=XAej*J?8ghf3yA4dFm6CVxZjj1eGYpn9ZQGTObAW^s4qdf8z+A+6k
zQHEgRjrCX-{+K+<i#8xL`{WosOL(KcmjAdRB3`S;V^Op}%37lP5@=#yT!KA!X+v$^
zZLvq=XiwknDQioc;E@e+%&(Ot88trP<iAWq|J)-=ydg!dSVmN=b9x`^5kED*x8gQ5
zrDv)6DcYfSDDN37g(Wl<HU~wWg@*I|Ly3owmc(gzlyj(ins+Vs)biKb^PT>@9ym7W
zvew!?^u1gWu#tUhf;{56v|ItDM-Qegi|21ON(G;WTMlbfzcLt0dfC1i-ebMh1+@cL
zlqEe8wzSNJq3$$a%Ci^U)Ey|CG#zu|G-YG-m)<qbh7*5bf6qlw*ZWzgWPZneTOOF3
zQ?1sp(YZElU|Fl;GUyww0nFU`H75^<m<w=&Hufp^7tvNxkjGz_X3p}vkfFDK{)CCa
zpqDE>IK(2^1x{SHke#!ockq1)H2B4#@p{0fx1$6{xELw2&}TIIgI-q?#)F-z-(rI*
zC__N3JFNeqV999uqoe~GKhSl_AC>9!BH#W>Rt!_~V_xi+0LzU*cb7WO+U4ew+hi3l
z9PXda&eCs#Ej@2P!HYdE%grgYMc&ZAs?&Dwb9`q&?{9m}(wezqCLf*FT(nIcv$q(s
zR5zZy`FwMrV(5(nlmnd4HYAs(+PTo2j-|4{(@lFo=VH@TcZXd&4%X7QTK3>U{q+ve
zXB|uDQ%e_c&Ueq9_ON@4&DnojWc*rrVvq+P*b(d(;|kUyTBE;O!{U0?p90hfzr(&s
z4{W*CyQ_xoaUM8j`(3TiQ#RFbmpl*sT)$?t!PtlU<*c5KMt%o>_rOXI{Q32pl^oy6
zr#zr<@`xtam#jgPPxK(5wI?x!lR<Zxv+=U_Ko{G3ctY&^I*S4sHOFtUC;LxwZJW&Q
zP|n!HI`r{w!PsZ9e8_ouka$w0yJhL$T05XE9~r;AcI<7@+Hu8(rS6H>{8%ePZGev?
zIf5(a1kN{f1+aXI2b_;>>%UlL3@gZkM(Qy!Z-tk9>N-iITzD8SW1FR?cFFtI?zn<J
z%0e#^uv1AgBP`>ngO9Y7GptB2)ci(z>W8UVQ{bg6x%eJeNd`1*e0#7V*pF|p!{4#)
zcYzIUV-ppr<;QZN5R3v+QERzV8i4hw$a6eoj<bUJL3A<{L}3%2M4t#&7)l+xV=C<Q
zGvCxH3!-nl`@oBoR+ecF4i(pl+z#e;mVihhHO~aK6d4cXLolNBEB5(o8IWe;;Nxi%
z?uPE4PLZtc+cjPltTO?*6eWXq#>ob?Sx>QGNoTU<=BG{u1`a&-0qgGd2#&WO*2nVd
zIrymu8%;ON&Tl`r-D}bG1djX4p9Z3l7z%?FQ&#GX@Ss*{r2q1RaM%mDFj)DX*+KGl
zS?*m9+!a|p9{2ET*A_;p&-1SD{n*gS(t5#hqs({obmPV7e?X|wB13msx_Rz*`DHNH
z`?I?n)9Jq2eKVtmcoUjGnCR_lQk)N=UW?t3wL>$7kMxlip6dsbr<jqoFFRk?KRs}w
zP<z^dalWEUdg)8s*)uFGFzy(WK3=yNwongbu@6?TZ6b}Yr`sgp)J*1#&lZS%aNQ2P
z9Am;3U*|Da9}}5s+RA*r!g$khs>P`>*~0Gu-{PM+z3V)4-uW<Rd&Y0|&sM&yzjf}O
z>E`Nt9@DnU(skjKt#Q3#8P+-H6i*E|_w9}v&lcF)em=*X-qMa5&)T-tWB`kD_P(iQ
ziOv?y-tgGYy=C8}yxQr|x!PLO9ette$_P_=t&LOkQ#k26Hm~WZ!A_y)Sea}2)h^|0
z`M(AFW7~Etf6GU=ym$Bx(jWi)*Q-DL^Y5-+fBrQ2y0bel?I@MM|J|z*f3C;#;h(>J
z9?_~Yx(erRmw2)i1&Bd8xKj-j=~$EN1V=XQbpk6JXfFqa{OG&z=;b>-TyWL=mhBNu
z_POLqkJ<wq9^jmuax$!wn%0Ri^ieN$$)nuTZKz$6bnPLm)kZdS*r1F4Sfj7z=QaGt
zwDA~T*%<p*u&JCe%zM;hnv>D8bB?dJnF){Ibu~7@&50WmQMN9|=xby~IvBB!L{5M}
zn;zneeVk|Kpr;rlS%Z0{Bw4QL;Afx~3OOC8KwTbbRaOp3EZ6EqOqNeTX4vo?WRx{)
z@Ch5*F)?rC13i^8;ElE;mk&p{q0*Jcvd8kqxWdrE!a(5y8hdWd&Vj>9llycLIY)b7
z^a0?rJ%xaF)^krLlCjq(iwl$w^mc(2@a9PQ+Q+H22^Qd-_9$=#O)rar)t!(wcQX2*
z5DH;{pu;<h>34ac?9T&qUN)<{I;=|Ncwx47@{w`Nw{k21bL#;MzF*T*dl2N-Hb=qv
z!#IX2_M31NxByNngR=ktKmbWZK~&=}hPw@3l!!h@2NjHijY=tdy|bs_41>=ELyeUo
zCr;#UfE5ugr_T%rUHMKAvOM9Vj*dF7Sc`rFCUO3KJt`){^<#(YK^=We#md-YiTB<X
z?vjB9rFjI+R0_j=itH>n`Cv*JG_`~`A}-hAiFoRV>j=>O6R-HI-Rx1baewV&iRkn-
z&2r{qey!KuKFK~GTQgl->z`wEO<S@P)(l(|<caNB5-fpU*B;U0HGa%NVt84LNNeOc
zL-)h+E4C%#RMvVGU+tL3qV$i?FP7WIn#-|r&K7K{2fSqM@;tT&E_ACMEh%&O$TvqT
z{F3XdDZj*VC7H5gicd7HT(a;Ja)$H97RNk%uVWgvt(CRJ&uT&RTe$%wLPRy6Z~Y$V
zyyP*(TI)2Y$5Jk<rt!D7ZLL=9C8s672A|Uz=dSo^Hwj+O7d6i1j~b6NbU*Mxt97zv
zgXBeCm8Wo`Q66m+Y|U|CNibCH$>X-wHrcTzU(@C(xivX0JL7RwerVC6Q@EvT8S;N<
z>@~L9ARQ53gGKz7GBO#<+(;PrljR56$a!#gisSj05FEH~H|6B<u+|Q>OZi&<Ir_P|
zI>)~hUyf0AYTA~^WUl}D$DgnM`8O}B0_J@4*Y_U|zW?&$r;k@3-hFDyMA1K|z(iH~
z1TeQ@f^>xXy#;wlQ{S|S3<QK;K6u#Ay@;!oTXqSa15AP=E0$}r*TDY#;~Vq;^D4`W
zAL{RVv+a-}#Ys6pO9?|`+SBkTcWYS-0|%m0B^K9KV$x`XIn%JfDN$yj!%j97@p+Uh
z%9gBd&q?Au=Y%1)TQZbK;yAgd`aAM$Ek|8jWz6x>b}p0^^n$+D?v@<+d8@5!_&7Mk
zw1zkG9aG*iJ<erkOON1ex{_3vA5<sj<mWII3eg?n2M9Dv=mw5@X%887$pxC_PhG5Q
zj8Dm%!|fqsPDgB6Zr?KAWmI_{wzxPE)yZg!Cc>HC{AC^m2qJv+bW(>POG|_g#t4&)
zk`s_9bD~WXElg#L;<z0|a}?1Tkrdg!qN30Z?1CIJ1bSCK)nk(5#4b(^g_XZ74vrE`
zgU3=QURq*%1Nu--ry5zK)WMxMp+tvo)oj7~+@DhFn%)CTA&5?xW%w%734x8~&uYwk
zxsHJBi*v0bEj`)BycTe7ws(?Fa{lsby;`>JZO0OMvULgVFP(;KiY5NgwMX>F623?u
z%g);^a9ic%<u1z4<#_I#$D;J-^9#YVtXOC6!?%PMzG}cL3fx%6jyW459(t`?%A@K?
z8{sMozvTMLuLaY=Ogd}hN;-1R^`H1MF4tvWhF?N`i)~B%d<kva%Z7NZzn9E&%GuJ#
z=d>^DYR#(_9DdE+e14&}vp)L`k3pGu07br7h95e~3G97wE3OyN6>7(jHOGNgFf{sY
z%2wGE=>=PpFVWGG7q3g^<c#f{qpoXQx|VUA-ln0hcL}t%cQkJEAcZlS+oJIRoHApH
zKaTqk?8$%q#d$DDJWr|Bo~->`Hs*7pYzJfDFK^%X;5;T;L7tvrqVT&Xj}NfFdH85D
z$OA*R0oasZf4cbuNuO%`J#wNj#aDZs+?pT{YFqLJ&jBXEr+6ot(?qp`JWU2e)OJOy
z^cbxJjyf&pVia8UWAZs!86PiNw(6Bk<&wkVnlFTeo6c`7m{X*&z;W>K3!wYuPV^`<
z4i03h4mW>1dOrPh!N4P4<eVa@O~=UAm|7|W`q(cVe1OH1rnT*k@+hmWe2HyKXxAPb
zmdIRUTc%knw?$ur=l;pMsn_}kSv7ez{tSzJUqVn7I^hVWsvP-iu!wJ!xAGw`o-l2d
zvz0HmC!WY=T&(8}&@K9x@F;K*b?;rCpNHKDHRos#20!U<^u&#gW~hZD_XK@{Ba1(1
z{K*lgrBZ0KE|=3eO~Il9>wrF@%ke}q3!MutW>B<0hnNE0N#%A9#$b>U5%4KbwK%i=
z^zypEMiFMxyiOi%fy*`Z{xE`0lybxK-}JZ*!ggq2I=)Htm`-w_D}RUpN?lFr%;;gz
zNKWJf?s^P3^a;?Cfc0Y%!a{6FQ`RNq&7F{o*6mbfL$3AJk9iX}o@@}knq}yTHtXP}
zHVv3~vrCJzr7s@)#VcCnE%{mt@Fz)QB+F6#T)FI?(pPIY^h<cB<VG2-<7VVXUp_v7
z2bOJ#^;S8`K2<sDK9!E8a_hE~?@vY<b-G9~l(`d>@$oHe3b*Fw(zRn~bt&a*z$_=8
ze`Q|k(2GZTjSg7u6Y7E~uhC%xcs9M_u_N-pr!;3r-hdU~RPKIi2yWRiCGVWLz3l6c
z?dLQ;=~~+s`g$o(jZU!InvH3~Co}u%9A1R2c`ba*B~y9CO_kTuPswWGYEuNgHF%c6
z`GsuE^jfDf9D1FXATO4wOCIG@r!BhXcw&1+t30+xut;CSi!S2!Dvz)Arlm$Zc~!QO
z0vxIzljpYP`kZXhE1#o}_*&p^Xl0~4t~_^qK2G{H_bJ*lt?@YQ|Ni;y)qOjMXr-n7
zT%)VOT6%#0n|+%n^27j6^y?r0@_HoZ<@Li+U6m*cSgwn%;HI$ll+mlhCt#vYQI^Wn
z5=FHwKSuTz-u4Ko*slW)4$>cF*<ra}5w#dJCky=E3xhU`rTf@`MhOr}T6yw^&RTPl
z2{#vQ{e@fV?JXzzi#wq~xW@rZW2&Lz01W-$$%4-H1Ev{3<fp8<@)>qVuc2Kc51LcR
z<Y2v4Zi~JKuk}xEZ*8n1E6UTxTjs6RBVQhq;D9j)Bc0@ASejNMJIFKN0l@^IcEU74
zUPZ0Z&#QutP5JJeHTtDn>nim%7lGnGXYS!-+bq+@FZ;mVUhrTBYY~n}&VWwqT8$5w
zzov}AX1scs23UE*rU<31Z_+rt(@Y?wmkWPa?+tUAg$4tA8=-^90hZaEP$1`?-E^P;
zOOaD-D74OH4&Z5ruI!CA|8TRy&;Z9HhwGgFTBp-zg#I7-TtLwV5@?Z>c9~p*>-gyY
za>}6(pclW(aC(ztJ}dM{7?Nr(-@xeS4|?b6%5W&)Gx)XV2HPnn=M@BGU;T9&@RCGb
zYUAH2T%?qjBaE`vS2hQEdJ#I&vauWMf(xb`IkC)K)SuwKNo>yI+{c4TCYzkK#Q$s)
z|Hht3=BecPha5JL_=DbgO$Kt130~(g;Lu0iioH#;A?Dc3EA}!BnQWLyk5xUf)^x@0
zCXnKJ@fWJ$SM#+Wh#d+p$r-3isRai+g{#TQ^h?HxHj1LWEGOdDl%+o&H(Roy9j_fl
zli!q-lu9n(Ryna8UDcR~f}-vSUsI0$U|erbQBWu3-VcYbFIi(VA{igGbFDc$nUDCa
zV%8k%2*1=`-RQ;~K1VXQ9F<sel(H!5f+>eKmZ?V@B93yb2UW23DPNNg78Me$u{}ej
z0FNF0dW>W!?-4r`_LezVz1=uuj9)LjUDh$2xh+e4m3dnFMHlCmVELS$aorkkQNH*i
zthWAXUProCSu&Nk=rVjQedeF5xB4RM&GT_iZp3FDS$3qQJcrNpv7U8BT&=uC*BTyZ
z^_TTWIW@b5SDyJ*kL$Hy%k#GMxOQfFk*hX0GX2telnYO;&&d(J@;Umuz{g|bUFiRo
zbpFQ|&#pec`QRo`&5^k_)y6{&R+FRKx&EUCQzB0*fb;u*7=Lr7CKDL4Vw<Au%ri9?
zGwss)l)j}hT6BN<%Z~@B*jAA8YbM-c*L?|BE}^+D_&M{XHn1>2%YwRfQ?sT^3p5+W
zWg~$Nkujy<*)KY=;@TzNIpa2Vn<V0qL{75&PP}eY1R1(BkAs!pU2L05U-*84eN+LM
zPR@|)wgBtOy$zirY)1!4Z7a2Xe$@`OaY`z-jQr9WR@z)RWj{5xZFKPhY4MXy1@jSl
zxu^v`<|%U4{L)|Yjrtn-b7=|9oQx&-*uEw7nI_svxg`&r#!8lPY#p67*%8)nyai=n
z!+Tr<xG$=|dccJhqVHwemPK2<OY{7ypb=lqs1d<_l^BFua}l_o1q&rvr^LH#;Zh(X
zO3<B^oMfmy<Xk{&UJ*pl1vTX;i_KAoSRc5279aXUeL{V7>g4SHBYWU?Dk}))1kMXm
z!W`-}&I4Guq8z{=!(bi+0T%@m)ajq@F&2a=b@q~dvG4ozhssV5VJ7_Qf94DBdgG7R
zs`+0h3w5PUW<ykOz`>M)Lv^Rttx%B0zn2&8bW(E%ejt`$?(9P;#bNWO0|;L@f)Umt
zU$jbaN$9!Uq*haeN4G{|(SVPC&~OYCjJjk9rW_o{QhdZXWz6bNG`>k}&SK7LEhc9z
z@jvseUbP@iyjxr2pD_=;^O{inj!f_}3>^BHN6<@m%-zQBd<eTa?vn`Q=$we0r1b?F
z&X06ACi1LjoxW^hmjINGH~2v3EVt$pQVdjaj#19D3LRyyOMr9C|J068SYGTK^Q%o`
zXua0kZfM$GlUL^)px9BfPZa35;k?z_thG%za-Aa!hAjGyT>cWSm9KPXOUwvv?r=MF
zu1Px>1K3nXO(CzfDGE}%%UCJruJ~g}<@^)>BOiS#lPeJvB?*SSFPzgG4}T<RSXbOe
z0}b5eoNQ>zphvKxmoZhYS&)xdu8Zgk@UiXQujUiM!6`A0rpAQm$h{r3rE4o<kVKm(
z^M`EL!?>`rx7p}5exp2KF-5z8yX>xQl=T;RVtEc9X>&Q-0-kmPSG(y?{;<>YWPg+?
zd77vACMNk17~dKrNe}*`d~3&G%JeIBg6=o;ai`yz?5Adn<|OMQb5yDCk8bP=o_!Bo
zCP!OEA9J-;b7hJiKSW)THkPA&*#QiG+FMKDa^kxX<p`89L%qT_AitrKABSh`<@jv>
zN5_w)?XS5#^0vw{N$X6c6YkXcluqVHZR}*5bnYRz#;cEMZieaRZN{q}$Gu=W9v~}^
zYt^L(9(2WetzBAkpa-A-fFnb2>Ll#1$t&3LZ;h@7g9}~p`z(?a=_yCMV;e<J9(f|1
zGXB-N7jZgAgcrHc;YX6|?QVs~-%&4g*o|LE&f6UisB7J$jP;SX!0NwKZO9y=)!fPW
zHFe2a!}Cj^-J{*e*h9`%fBwGvCQh6qTV-wG(=F6X^=G=4@7{g=?DJkL4^IgF`<r)e
z;@dBuuYOLieA{9>Jm|p&<nf&y68TT|_r~<_U7~;e*MIXexW9k=Nal&BnEv+RV{ez>
z<R5NtdLE6DKU#q2&wu~<>gMg6UYB^XU+`;uDoS)Uf8md*bnpNCzy9w7J{CHZa&i7G
zB@T?(95?A~ecsd#e85T1T*7%A1UQjmG1rOX6fbl-@rq?+k>=D!p2AUn)mH53HxVPj
zk!Adw?+}dbKT^COVDX!LuY8SM?3wdlq>1G<dPE!fx7dpu@hJyJACojrz>zowV~lBh
zMK}usiy}P03vJ<p2i#XXG0~%*8mr~BuF&Z>>WFO=o?N=sXTYLelxat_OYI_P!$u(q
zm~81zudF^CJS9gw8s7@?$yUi$U9!oQSR3Me1zvO)FHok059tES`n~{H>?*Xh3*X3p
z;*~CPwViU|F5{EF5Ir`b3pt{3nXVs=WsNDD$|ln`UrF$gicMDN0J7URz0G+&oe}c%
zf$iXTw4DGK(!pr3*-@l|fZSwQXFG}?XR|sTGm&J6NYq}{mB;oYn>B{9B}<@_kHs^U
zXFauglv(=G8)e|qrY|=)Hc!f3)p2~%WX&zvpgF2Kc^r$ltVkWkdaXT@Bi<;FGVMUW
z@X)IaUdKI+Z(t<qY8~}rOOd5H47{E?UK61HB-hh6j79e^I_&ShJ%XgoDc7Wi8xB8!
zZzX>U!#3#e8I8s|Oz`<n-w1Res~g*fhXh!(i$*&Ox^Y~GhfWFE)Llmh6y9+@5RpEl
z@}!Rov$@!{*O;`PAs!j>y82Ce*vP%UzINNtfpql)JQ@ds#&55L55``*tTw1Gu*$ob
z+<Rm==7OuWgZWg_a2#eH=6<X5GEY19Mut7y%=5t3pFvZD_=3PS9~@&I^F(tBpG3cv
z`BG+BTn8yfTFJ<Lk!_9RA=A#aA?j<<wrH?J>w|cKkxE-wH=!8DV!@XAmD>Sd@y&Ss
zsjVwDUifP^0iR1Pd6BP1BY)FuoxesK#1nJL%yUlgoJ&WB$GYU>znZ@y?UFL`I3{bl
zvu@}Zy@Vf{Jbq>4QvNAAWRao|m)O`E%Qams+7>+ftR}P7m-3Z(B3&%AIm>wq*AqYZ
zak$Ju_{xe>B&g@&y3@-xfgHOvRzwSoq&_dT>0i1*d@agoe~U)GL!NA|jh7{|S~`G7
z{wQB?VE98eOv#$U$+ytt54%tCoeCG*nsu~rI`_{0m>U0?7QZ5gc`DyBPRp1Z))e^*
zIOB;i2ro(FG0K8o<4$A7={tDMMzs?<$&Wd@py8Sd*`fzdLS{@{XOXzpQhTT;GH*LD
z`GH&(u%^n-<FyP)Wlqa?$_oYwgRg+ezl>eVq$tw?_9Ok!$r9nnqwiul(!_F21~9cZ
z(w?havuRHElC~maO18$x(qnLz#c_b7b4RZ%{K>Kwvh`%2K8?RO(<3$(9zJ|{_3=Y_
zV&}j7%fB?GeE52E_4t0-Piqp*AU@rGxq5i7d`<+?V||g;y*C@5n2iKhp5A|u(Fb+4
z>Bi0quhUm%Br7rzsCm(N4P2z#QYP>dzv=VN+VZsXH#ok}Bb!wJfB&EV@9IDQ%YX7<
z&5P>?-TGFvt8Fz~2r#{PFs853wP642fBZK`{n!8b-#xcm9<Qo_o|b@b39dn^yk&bn
z5Vhz8CsOZsr3$by&vs!)yeP<c>fBwz>sq)Afj`|To=sq(yS7bqI|fZAPI!@#b%{xE
z<$^1ZzG%^iHs{<C!De2o4K;dp&+vT=I8jD7cOyj`D03It68X@Q=KNKt>c7#ZI>r7_
zd`P=%sbx>}QH}2$usCkk=*x38%W1U@I(k=kT}8fFUSrD^TI;T(FMhV@o#NSQQ_I&=
zx|ZNt<8}$}UNqbxHN?WKrChOa(jTshSch4U0qYe-(RTd9`okt2;|`ew*$51P>*aZ6
zcp4$Di<F^r)i%$y#%Wy>9=pJ~4kO0Qc#Lv7*U1mb<Qfqx2geZw?E1PDJebjua`$i_
z+;FbWXe*=3$%@FX9PtEWtlm8Zgq&7VkjAukB!e8cIm3~KY~Zjsx$4-~={6L7r2cZY
zuHDY=KShlFc6?R<?xskOD;Nk(A+K2CDG-9^wF^Sz<<wc2!M~{bnl`SX-Rh(*I`Z)*
z*0;#RcJV2nYY#e^Iy!bOp<g-<=VS>lTUlS0+9BR^=CU*D6f7U-1&_3sRF=Kb&RE9Q
zT#xo=oa&MrX}6Xc4_Lvm2Dz-a`9`p<vd_gQzwD(CpXmDx9B)R);ZWCF67{Pbb2j=E
z9Gcto6W^M)=o&EdOr00z>&q>Av%Wp*^664pzP6h2j9EX-FJ(M2-KF^R#LYaJc==EH
z9*UOA()t?5Vr^~A*9xuiw^VPfjj|=;j*%DT>{*V!LT<~(439qCYEPD#<#6p3_gJ7O
z(O((oi0<bc=Sp+7M|*1JXk#2JDjzGqrc<yi8=a(hycBIM&-|HY4xj1ws-H8^v$c&&
z`FL98pDN$BfnQ1M7T*fuHF8_hAkVI+(~&sa`lr9XzS0v+Sy!gb_*|c(pTlQ))VJ_R
z&M~}_L7oFW!0vNGcj`2yBI}=`U1BE(4Uz_+@)qyhMRB%=g($<24NSUQx~AaR5pAU$
z^-xDQIVYz5UHuiIM6#o-C?}FD2-2bzTzTZBJau<=rp<LOUf3*MnrO2Q0+iJz?&_u<
z6F4AhU&O_-`U~2zgh;iloNZXfHBBY^T2!?Y5Yc9P<S*fu<Xd>A<f=W|SdO%L9f&xU
z%jj3FeNvF^W}LD?CAh{Bl4`aL8ul%vVY;vIPMUl?C0u@(qFIZZlR1a)^-WB~C2i~F
zR0`OjzY5wx;QJsC5TXK>wCB&9(>l!9@xZ$wx#)lw8)c`~X!zOu@h4TNSkkydJo?0g
z{Ky9#YaDdYU=4V5ksNPGE;^6{PO{w|;OKyn+-M@+b#=nX#4q`Ur>xU#<hd+=Du;5>
z<-T$GrA|BVnIhNYIt^|C5fZ_i!@qc}LlD|!L05OhoqUred8I)H@9~4J3LJqeci`a%
z<PARXHgFs=xQ=H74YC4#ah|t}HoESv@n(U`UHY?Ry;~~l7&>`*jSP*|DV{ann4(>x
z3p?iMrgXG?kk1WsJ-5FlC(_9d<q=maqjODgCELh2!>(vsEZ1~qm~5@tvFA8mBMaKt
zwo)IrERKnI%%^>7Q*LXlBW)}(VoO;%BJC7U4X1WUIy73N1tXucIoc5Uq!U}SA0i!P
zY>6pVNB&?uNFI4U=Ztcu%2@|?0?Tk@$lpuk)%Y{toKG{o>e)W&igSeH$(U|9Zs&Q5
zT-j2~fz^0q&(>VWy3B8<9L>a2FcGcSvNajbgvi$iHMJVwj3fN1n=1vc@thkLeY;lo
zlKHWOX06^!qL=@a?_pa@R$jZc9>%#^TN^p2S5ir9h~VVHMcC4^Y>D)1<>j$*ig^-!
z6>S}_ork`NK8$=@%hCTTM}GmME|>!FHGT0oP{VJ5$tLm24%KUU=7}`X#yD1DInwW0
z&bFg-jo!?kZJWbq`YrV--)@mN$HS8%Q72D+tWEGQ^-V3vgFg86$CD|ftwA1mC98G~
z9c}$+uR<fp!;>@756B=vTZ21t;wSP@Rz2#TVD~3>BA@spPYo9Fv0S6A!6Hwq{Q0VU
zpF8dqw`h-sPdO|c?_AFm?Gi&;2aP4XHJX||Em%u8os@Z~N4Uz$u_f~WQ<5&p0p6l-
z!6hr^YiMKNt>Hr-ulHdi?w&ug8!c#*qhpLj`jo?^vhs$c9xNXw(LMgug*m;lZ;2hT
z&G&sna+-@f`S}<5GRG7y(=Mq`wI`0HmMnj*hdsji&>n02fjjAK{CxgwOPRgyi6ncK
zOO9kmxGwajY!d%n=X2vE+P9=V(GQV#O*zZR^(<=+-_uVtc1hcMIhDc^8$?4qfF}-Z
zi1on%E-bhUigCviBH(-u87-S!+#L&?yOX|_c!OC{8iZU<tfA>QSBh{Ow`>6M4(W1N
z6PtVlZv1t5#=~6o;Ev_CJK-(spH2a9a{N_i1=Vj7QuQ^K##rKD-(R@FQ%rLlPYqxf
z^rHj}{l+Vd>ndEPpvZ>JDLP%W+uJ2QBcZus%r4w5?vO=pL0Gx7%gLgRJ-pD=Ry6kk
z))i)AIud8P42WN8XQ#RN%rJJ0(uQkRdGB7K7M%XW151;+^CHqxGh*wGorh~0zEsyz
zx|;5#K8U<Eo*FFjs2qJM*fF`}>6nHr!D{1dshvw~s>xr%SEE@YE7O)`XWTNz1b*0>
z#?tCiH0~m8EN8A7F1H09Z6ej^B{Sn<opq`BWLTACtN3eqwk67hKG(Ahjk^q=s%M$%
zw+u(dk~WA>c}q9TzS>)}y9Sf2EIZa!mj8PmLkQQyRA}T&+AG#9TkySe`Al->Zm1L8
zsq^gbnm*a?V~h^&lHjU;tND`gvV>-7`={o}QW@Ve9ph%|$qw39^AYQF|C&l$h`4?@
z*Zf`M?>)y#OK&_@i6)-Q2$u7`{T1b@++VAia=~QBIo6CgcCvoiF_mxOnUeP%H+bOO
zzUB2`G39#1waU@vIJR2!bMQyub#aoXe2#Z19(klk_sXm3c>P;UtfRWgBepNn)XJh&
z4oxiAcxteh;dODXwl9$@8s$+}4HogOvh9>QYV5HV^##Cd4iqe&u*=jX2MTCwyppk3
zu2=13d9m|UR-S36>T^8u)0B*@I6Xd#A34#^wzOUiuqJwsZA(t#x3(eMyj4z(Z_4-R
z)nl?X-fggZS1(>Xw@u~t)Mg{|$De+(4dfCaS)*sGy_^hlGRt>3;+-VYy=1;}()H5X
z_Q2kEsl*2B+qdtoKH3M<)}(v_i+)Zx75QFb&l-Kl(4KO99%EN+Y}a6Gbi@<uHGF&1
zD)^Gu9iGf#ts{9L#yiIXI9-rN!5+X)X7q6lBoIU*c*7b<kVa3NXia3UC=$V#RH96v
ziA8pbR%07hz3LMgH~xL5VKjN$_(3Fr3yL1_kw(iA(q(X#Ac&U=95$QR{_Q6hfQ*jG
zm>q}BvuThykc!Wh6yv(U(gBSe<+=5qw}n8B1%zq0^O+pOb6V(0Jf&i<U_)OC*|;;P
zLBk(UC}E<*?8g;AUy5E>@dIZkZUt8%&Af~k%u&Cu7Px?6sjYLH9t%|6IZ5f&p??_8
z`Jz7d!P;ZRQeQ50gyhBk7i_7FUjkk1lT)^5Z#`KaeYbWVWf^4^JIyG_x;Q7)ldW+K
z#y-eA5mzhc*Pe-nxv1l%!+Z_rX-s?C$8Mtd!H%z-LqOBAMY8Y_HwGXAdd~f#om1Y6
zbd=%eoFT6h>PNTRjzd4#@n}nm`lk5To}cTOmSxt~yiC)oFSWJBe=ay1M5V*0diq?}
zbf%n7-7cCg!*k_2ng;2-SdCjdAH=sN*EM5{2OnvT(3-@&)8^#bCn1lSn=Koc!cTEr
zDK~HqBZ-pywA8LW>7sv6(cYS^YtJETGJeUh@BTUL-D8il!;CwWogshx%j>H@{KKmr
ztnbl887eF9i9<m)EC1I&{pIQpzyF7WYhmGGMIQB>M5A>y<LBz)&vb$-&-ioK_psyo
z(Zff!5oMn{Vl$-b&T}`buCF$D1j{GaNLYHHsPWb<*bZ&o@if(eQ#f5b#FLmQdiiO4
z3+kLvf6I^cw%@bOdJ9BU4C4t6C0_q?ZBP8HNo+FSd?}9;a#6)MG1eq5FxLb?ru>fH
zN)t!s)LvhA>_N?+mb%a8Ut2qU*Xpy8fuTZM^NDmoPZo7DTnqBRs5@AvggmqWPYm)v
z%f<DRCr?%fdUC({dJ)_;WzP>5M5{eEeq-yXZwya0;unMq?&@1rgZ0R7sK5!nHm~>)
zEK^BO(Xgg5en@N*G0F7}7|mymuNdTk9{&I%0OC1<%YV>)a=Bd%62|9k_Y{fd5@;?G
zZ*DZ#i&X^Qp+Ogku?md!Kz0{Ptd$P$+QDX(F-bRN)sY1+ID$gHnP$8t@CBx9m_;6!
z(<zi3BW$FO(6Q1+-X%q#L?)8xJjaQ^_;|7rt4%pjjwk-TA@j$reeAw`Uq``iJ6$w#
zJ=x?u9=tO}hQNme=Sbd`u1wv1LSaP;EnC4=NHcF7537ak%y3DzY;9WhuXX)FilC+E
zng>EI7~u0|C;N21x|6u3nAj&xq8$)>>)AtIma$|^=3~=R*Ptto51kWE>CE;`=|Z+-
zW;*8L_}pxeeE-34tx&Ti+KgP$u%1iiAT4Vi*CQQe$%;Cnf3f?kACCq((CQ|@$Tv_7
zCF#?RbpvKxYWZzUHD9Z(d68?%UfCMymzJmGE|sx{ZtW)f8rf@UkMY}DUFW3V+NWHj
zaz4GpFRgLkf}gtmxy`geYXrKd+7$OyYIZU=bxh5AGW{Gr)6dp-6u6_PS+my=jMlz}
zk!!~065VlZh@O2Qu088}XrlUc<Td~NI$GXt<)8ol)79^P`?5{21Ew+jukuqhQB{iv
z8m)WE{|^iD)Y6Y1KVJRz(W4$E`W|3~8Z6@H$}N7umCwn^_^I}1+9ma^c0}3N&!3eK
zD6u*I{=>(s@-Zn){hm(g0X%$j^TihN;o>Cfo9bL`rf?FYK{Z7m<?dOYI*I!H`ODR{
zU3k^)y1jb*__4o&E86qbcG>>^<BvUvAkWL^FZ%kD!Y6xg<<D>4IO*fZkFFj+dc1_K
zr+N4Oz0HI7eSNM_Z{Hl(fUof%cTl_S+JDd5{^;u2(<i3vPz!yxCZL7Ihf;=i99vo<
zTi?*gx^$7A;ZyZFeN*^H4<B4Td5m9&L}1{@*RO5U_xsv8cx6jTE84)zp8DwIdqAD6
z#F%0XFc%51(5*V*Ru8KljmYDMO(Z34{16m!aA*`V`4c&^{esVF3Y0mEjDf0@WQT=c
zq-vA{<(LuW7dfR(t_OWC+hiI;Ka4w(YI6x5Gup=~^7On$;_RJ_fdfF6AnXCLBmA2<
za7lP#_aK%>X^7Sg$WR=~>3tYB-=Hn8YBL)UVK}cFqDZt~{|~<FhuFcc;rq{b4}AL`
zsL#S}?b7+`Htt_1*OCor9Z$A;NXLyW{7c4%FQHq6pNMCP&Bv;aI_CTft!N_OlnqP9
zMdqK|IL`D->PziK)|zp!M(z?je<?I;d{}GqvHUfeOJQ64bZUQLsqVX?+uCOAUD9qH
zQ|I)|;n(_yaa2oN#uv5`%(%-m=bzSkOCQy4$er^80Xk&P>BT2pd;V~}gmT)0b2d$L
zf5%>`2X1Zu_|!?GFFwsPr&Bzl(+tmW;cIye^3?D@-F&*@m1&nYw*U7}Z?697<@1`!
zcM9ttci>e|FD&Tu{Ml18Xib-t6S+HW#T_bV6Oy~ET50r}fk1?Z%cfZXE-X&zTPR{N
z8Jp`5?3*b5=da9KHa2pQCwu6N-BF!WhVf*|2YV&g$4{T^)d*~?)f3xrpX^B>Xg_~0
zPxC<e*%m@z1b_Ms!iorR&_JuFfhuGd4;uUU_UeO0#cwKpK1D;ux|eCV<_Vo{Y_)<o
zQ4)Pz>qq+ToQVtPoOm)qcl<}(7QaOPODWgdd@kEUJmg#|M@K+ic;N#!*IFx}=wzhd
zb=L*s<C-mg<mp1kj|avW#|zg$u`apFu?2jjL5{{@?lZx~Pc9k-d4T0>uNcc<{()B;
zpeIf76pHQ_VZFdC#U&xl{=lhqMQ24T3|)fW{`@!*JrdLz#liZ4;oY&Nn008=VS8m?
zr#d&j1Gln64K2#^!pB}Kx@)lWT6KBazWb!i%vR6HKgwHl-{Hr3KtY1zq+A|%JXW6C
zx;#d6{nYx;AwQp&?4fgwpY}T5t<kwh8!t)5QXjUqB9`)BQk(a%C5{=+bvY-WN^|%r
zUU1XW2X3AB50JYrEwQ~dj#}_JnaADKdbd8S+1~PlELD(a&M%_hBcHQr4;{ZI8RzV~
zX}=CVdnq>k+v{?ZeY6`Z2TLPnds*MbsmZ-dUwrvE>_2aB-Ts^hAM)T;Zl7$koNr##
zoQMfHGU8%+e-lqHF}Fu`<x>`v8?M_Pn0nekulZ@okNQ<6j~~WngHGRl$X194`CINj
zfR86)p4ihjBtClo<LjTiy!J=7$)+6Carr4I0nOXDZ(S9y%lgerp6-#P!Gd@1-dj-P
z!vN??p9u2UAL-%4hZe}JM_S*Qj&lbKIum!x<kzvlTYueD<Ds!J=}fIIe_%T7l>t{P
z5#;21JG!^!&AYc(A1yF7mH5`iPj7zqr)jQ#uy1eJ1}pvc?)_okOU|6#b9l*#d3EeI
zUVXXNFHuh}FR^tAO_n=XN9NpGH6>$>O&Skb-WnOl@?rzmgC3-}oYxJq=5msb9b7QV
zAF|7XIw>L*@4ADRKPBYXUP}Rmju~v^<#ju5?omb#V$tt06yukkxWR)Q=F8Xa8#nMe
zkI{P@I@4SRfDZ043<Cn^H3Ha-UyxOvekp;Jd*-7;YxS<3F3Wy^Q+BaV<Qh8-u4?o%
zi~-$;V)TXC<aCC59X8d=>*Q1?s3kiH!txCV66*bRb!H>E60{EYZxu;qp@qifd4q}!
z9t0ZN#kUzco=zZ=$B+(T*%WS|+NZQeSIyoUY#+;(OIg!~meeIOpdrn*c@AHs{IY-L
zgPLE3Q~t~L#A#Z`8e7{)+FM(#|DulCoY1n_Er-6IVXvNo6;G5C%N&as8?`mlk1bpW
zbRXQ6eKD6TWE5HD_#`;%GKs@bv_<1lvY-)+GD-DFQ!vN>-ttY%@a@bcI^ywP^va_}
zf+>%<RvB83YdT*LU6w&TCi>{B-9&f&pllyRpFyWO<tQtb!L?G0F3N53&cUPXJpQJ5
zBi|I=R@^c2BX4)=wa5(SvF6AgzFnmB3`Ef%v`h7PJ{-*euf!{TKYzK`zl}t^!YdbC
zxz<17kI5zT7#T}=H|be9Q0B<7fq58hiDlZMddz{#=dq80uSt^G;?W#w;iup=+49Gf
z%{9I$JrM^lNqS0c{CkXiNs4-+ybOn4Z56Nb8{21&$B@V?nytBH!N1fOu^myb;L2<A
zgo}K8l##1M*{-~=ezF_iImfwq_vY%yPvy-ruWWylmKJ(@yNuC1uxoW}ee>qy)jz&^
zX64co&<xku<NnEQLtmsdR<}X0A`56q|MaT9#|k#n0{ffo`TYlb^ANb7Y%lNC1KM#&
z3!YS3pnR~c;ERSaP2n4O2T5w{04aL;rG2y=99}FX$&!YvtJmh|$F{-x$UYN-&u`xT
ze0BSezwx%b?An67H#eWJUf2^zB<Z6ZZ75O;w&;g~G_<qShhL~A1L)Y!fzHm<GO2BB
znc?;pw~;{uOF22v0vK~>IB^h6Jh?7-&X>|<zGLgrQOHtV`LXg_{7lxgKQtnZ`c+mg
z!rEq`{1|C#C$$kTdE@+^1xSVu-;Z(AwU%tbtDEI1&=WO4|HU@b4?kvlYj@et0GmK$
zzeiEalRk%Ioej$+{Lz+kmDP_?Uu{Dfy!;wz_bl%rKemU!oemb*$y}ICW(42C`PqU!
zOVga)bNHGb{|1s7+;2J?QEfcO_SWRAg<<QQ%(d;AT7P8UTAAWgzKXv=+yG0(WWJVH
zKH<faU0lQiJt#p|eELkXYwPQjQRonFrjvf)b8<xsoyu9B<lNf9InqQP%|FLuSeTE#
zYGSVibHMxQV<(6#&SM%CSPOlp_v=?fcCCib11;gzX$Ejv*S8Ars3q`Oih><bhtp?Y
z?Tx7Y{Os(8g*K2__sb0vMJR3f+NpKFKoe+tS<XHG(^*y}oWfNZjsVW-Y8_9GP`Nv+
z+Az2FO(~X)V`--Nr*N{}p9r-1KoD3k<x~8ZhLg=p{CbzRV;^IjIc+W<C7v9~o7^LU
zN`-clBrDeow)}z<U4+N7mb7G=+Z*ZDlx0)5L=OS!iaqF^vXkRNYy-5ucFS<%hlcf2
zZBQ<oA=U8(7=1xqB7~E@_wD*qZGav+aP${zwvyY@RV0Hd7rRjDriP9r=o`-6bYVPL
zU-*$npshSJ>q7hEiQ8%Th;to-X<kJ}0{IicAk|jN(D|{*a=(EBra*ARO+=v%59<d(
zAbiP&rziMJY4qzh*pJ(zyrU?`z)}H^tfIbbZuoH#7{Dmsj@$GHU_V}3q3{A3+u`ue
zetgFH19_n=zs_Vo)M$g+QTlSA5x&qvRpxbUn|KRNktz3yzo!?G<-nAnhgRc|bN*JK
zKs?9f=-QgL@<yL*)qRP4*&n%{ZH{=>o~gA=yyEHe&Y7x<u9Y)qT8a5`iXeZPErZr<
zL9P#!LutzHfofdhoQH3D)_Bpm)Gpbl+}E8>m0)Ad^R+oXx+&VndJqmllL}^KH@N5c
z8f*@3@kYB_G+FKxUBs;|OHY|=s+IM2s7;q{?54xk_nzU5Hcd&2bg>+H1Xmt$Do^1=
zvn7vmDYG`cw4lx#d%Ml6Km5n57xqS+ax5*B`X%rq-Ul!38?${ap6Lu0iFSZ!+#h11
zghOnUFfP|<Up^}P%iw5>Ql^b{l@IL@X6XL`QY1&7nyx~x-+_84jk)LxsKL9>txu@a
zmLdiEdsol=ZCI2C+e)qciZ<#Itj;m!ldQ-u82R<bPoJ+Ic{FeUvuQcEm<gYAqzER7
zxh{Clm(pdvQ`I>@#XG3?*0V=@kr59>*eJOXJ`9@Wt>7Ckhk=#CA7!^B#=Frn4MCf1
zxn3ibX{q}K8A?+E#Klo7#RHs9?lbKaf2KWWUHuq!>M4N;J9T-kuj#iX7eN}U^vQ4h
z6_JDG6X(4>5huX$<gQ1Yq+m;?8)YNUhu1L17%#j^@P~WD_qA$)o?8DMYwF%ITXW)A
zJGb&*f<4->C;>gY*zXokg6eE;Fb5e8zG-VX#t9pSoP7F9Wf{i}YH{6->l-}rDG~gI
zj4$2OLXyo63jvY)cA-R`=Na;&T>O`hS=dbSb<W4MS1|nQ)@&B&pzd5oV?v?9<}c7Q
z2vKCMW1|RvdC<ln*vXSWB<BI&)w_x0xPHy`*QU{m!UOUvPt_ce=r2xq@TJQ&arb!=
z$&Wb*?(D}vZwDZ3uw?dpwYzglkqtIV{7l=FeGb=>gw#&+67t<??&5f|wGj-vrD^Fg
zX{o$p>12BxCxR(IR>wK{za^VnV|;Cp2mj`F=X&geDV*q53~8~<=2@?nHsx7BrjPiH
zQ$5NOEb_%NZH)BbVm+3pHaK&C!N;Mci^m6*GhU4ftE;c^cOv?)jAoHmZ{lQ0;^Bdw
zIR8a9&J&X(Yp@cC0*%l7T5Cy1z!1bKSmR)Wk2hlrVGj#@2NR*Q139EfI_m5W)BCQF
zkA6qIV@Xao;6te^(zOG05F2l306!W69a#+~mva7a&UDOq!PIqkY(w<ZG5U}>b-X=R
zcC<N{qwHADbW7_~>yKpA^vIVrocNaJvSG@nOJzkm*%)DK%e5es+N?awoLlp@+O?H0
z+SjsKG`YQ64}{+<pR*&6g|&N=OYF`%=4?8)?HcD><V*#5w#fUIcnI3?w21m=2;vN_
z`1jXuuXwuV-`~8u`p4&dj%9t?V}8wP{p;&@ZYOlLcFgf-ynK@3r>e`($jAOy_C_?7
zz0m#a(Y7dWZTXrz<YS<RS0R1;c<eVX$`C0VrR=5Pe*TuGvdO9wFWA@1SgNZ{H+O()
zan=W07i+UddAn#6|J*J9(-JoeZ%xP+9D#KkrRtu2JLUT7>67sX=Y=Jok}bNag`ED3
zPw3!_uUrZ3oR%EmvdU@SK;89WpEEw|V{~3t)?TY?>=Zqv_XfIsbPV!xshiE+#ZhUy
z#;xa?sF|@+rzNkRKWhdMD=a-|YCQEsvxV4qY+~E}5IE3h1UXN!J@nTVvZ)U0v)w(x
z>;3x8X@mNPe77MCS1r&3{L<5sd3@^e6@A2sb-%F?5j&Tcwn@hTPJ&6=G{kRuHJA;V
zKIl`{1|LvPAAcB61cSJ{4SqJRh{3S-0V#orf+K*UyL`X+C>$ik3waOXF#pOR_Bo1n
zWNG}mGT03<9D7PzKAQos`mwshccNS&58QZLjgj%MAWz{zlfg|!Fkw_ykf-301W!!B
zhJ-J2dkeOsa}kmuc_A0Re=4ZpeC`%=i+yHyZ!-l;_<)}9c6PARdT_mqJK9q3$sD@=
zr)XZ!_nl+f134fHb=UpVjNh@}oCl`*bqaUM4b2-&HCw0JHieU{B{ZVFlsumwoa0>K
zQrbRc$JTS>a!z~RqtGeGT}!sk+oBIS_u6uO7-gKQJl7+wc4E(#u`s8v*3NhyT07o~
zXNzp{uE{S!&r%zf&`tStjbHYp_4~4h`x{-f+M;}oKE^TE635KLskyMmmKN=)<)e4b
zcX=%6y0m5U7I@aNMaH+lLw}Oyd|s#ye~;xql(;`cuTa|><k{b%x<uJo$F9F-+uEls
z{d4ea|GD+g*(dxpA8Y???~h~C{gb_8Zg|C&NXYM9J%48J_-nM{#@^Ta{=<7`cwmnx
zJ+^PU@cK8MgmZGvW+~s&Xm94*>nx0iZv+u!;Pp6V%ZtvoAb@RL@ro@zFkd!RHQkDx
zkL>A_hvCt$0<HhPwQss$$9>zVeMT^&(Jr_}yJoRRoK7O|+fEs_(WW2lyCTIWMd6Zm
zN$-=#PuxHC7O~N2zWJbUmE5~}`t*tE>UQZ&NbbAXXOhWA7e0CX=%M`)nH)O?c`s~}
zKBk(Sh;Nq(><q7B<AQ@b+*?Uwp@rWQK5li9jBg%V6#apXu`;%8e2$!20$2EhdGyJ~
zJ0myCq8-odX_hhXU_C)reJ_ZosUAPJv14;euZv<XefeVF&~m!EjzQY7C3{lA#w_FE
z`r5{?*%R%i&wTx`w(?D{;g0#Nw>-AZ`&ggaI7VhpjJG_0`Sud)QXX@e7rk%az02TH
z9sg;Ki0vrjj<w~ptu3EC=&f&3A%Om)-yc63);Qq42KDPr>3iTxnzyPSlSf&xe2h%-
zE|HhVuy{|AZ@Ea_V#idQ`nAU9H+O=P2ylo;cZV|lfwGPhl<_BWY)<^Zx@(EKYAByG
zT*#E5Oc_&LBo8=(3wD7X?iyvD>4~4B+XQqsm5zMmU~DLQq*o{HzNQ+Dr*6H%R^nFv
zN+T;fuMw2oT4Z{#2R6}q1(*R@)jCmDE<qmT_jbbA>BJ{xGYLL4D0KuqjC#b2PB}QE
z!LWlQSVz+-$A{RdU{f)=qxA-s<6~)y?4gZ=&1JnSJf)0sYl(qqZy)3`=pf<1zypBJ
z4-OD30c!CYw!L(((K$!Er5$rJGk#6~WZq-y=V;p&`!3b4wQZT|*D2guJxl4&HNU3X
z76T$Jx)ywCJ2PEucPwYRsk%0gBfntq>6%h9d>w8yq&B`IEb54GWJg@D%L)xuBALmD
zW9<y#CUni_O@dC$bes>TjcduSjT1Y3SNnNd!)0J;XPg&zVb$n7v@;;jqrL3X*v|4>
z{k@mo?{ZG>0p{T`=}7iT)m-0Y`5KFpj4P3+oXcBf=XS4b(#4V=+pc5z@4_)}zkI%W
z_QUXMw5>LLy!~?Z(P9QqY(LO;CAW%M%=f21|Mlt*|NJ|rNB7_U_U7vMzk9Wq|D1$x
z%*W4XeC+!r2P?|yt+Xe4K41OxGaJU!iA~sHqhm~aef`w}Js*6d_PGTD^3ur3IRPBL
z%klcBpNyHq^^oq_&eF4|PmP9wcsPxa$L6p$o!O8r8@q#?Ym?1}?2`2EUI~a`7kBSZ
zrT6#Q!2VLPYKp5|kg!S5#&*kZxm~<Qk9P5tfs(vIpok6sR^mi9Ci<)O+O<M=*yqom
zJwMoNk1}b;p^p#cM4qWKo8@X#|C*gChC9MBwiwsb{$!)q#^^(Wl@C2w^X82`1(L=(
zn~emEJP6ll4}l)W<@2WmS2Y3(F(=nQ+<)l(0NtCn<MGlExf+b#GAEfIV;dw4ng=ZS
z_D67pHGsc7zX`JZ_SN*_jF0VI7^C$5zR9wwa5&~hWVUHN{UjLt1lJCI&L(wSTOPY$
zf=gd4a7tiGJh>m@6Fh9D$2p6Ty2r+LFRevcvAmT3TcYEAvs172I7eI9B(q^*^K5)1
zi;uRm+_K5f(>V%!5Xgat1fJsya!4NJv79?g_1Hb%l40Fu{J^IGlJwjdj@QyK_9VN!
z(#>v)Fa$6U#B@*Zcu>j&#(<IJx)Jt+KKKZ}a90>l7*STx!i@Ia59k=gw`Qbe-c|q#
znuy+CX#$JM#1kC3@C$Z%5VfmAZy5WFsZ-+cSdKKYoM>ERkgEul$#c8y6>v6{OQIua
z6n`FwEA2L8oz`vZXbjUUWnQy>dBCiLds|G-tqFIW`JV@f)Ym=pM8VxLfnSQE^=s2!
zy4UKRqTkYvDS1;kjiM=<J#pu3+Zwws#jaEJht|<GK3PkvH9PX{F~4#^Bg$V|rhoBm
z4EikP-9i`J9m{KEY4bSp3x=FLRw7;2QybqA7I`BaIT0thnMUxXImeBf3?18Z+iG%a
z<4trmn}uuTew=S1e?jopv3ZNn&gGWcc!+CSoELX#Q`B7>-x1#`*V-Cs)lXaHM_J#?
z+da^lM_VOcdVgx%#rD-;5x=FpWG>g{NQT9_%1dO|Y!-hFch|5>w(YKcQ`5gkkO#Ss
z>~jT=?17APB?Wn;`%;5E*hi3OZd}&S08i^pdl~(e?XE6MZ{EGN&82I9(j#wrvKdMs
z!k<*Ap2`*-*F4c;t7O?UDKLUOFP=a5&EjYYPYUpq0GmbxOnf8Q3GTTK$obhe(#xqo
zPqcjYO_zsehwo5DVO{y7Ke6=L^p}lcHgtI<6HhBVvZoeeCxH9#(VjpuJJzSNNnW0c
z8Myhb$A=Fe{V50!e(|J_=`M+2*hAl>`|0O5w)vzTs)BuQPeXmazHwh@^LdbbSmH?}
z{EhFH&?CV&<~qIQLNgzRZR0Fm8NjnO(Z@=0<6N-I$jj4gImko1-@SiteR*Sj_~7cv
zlgFd+LvZ2g(_0JNF-C_3AHh7IOV&4W%t1vklR%I^L1$RTOxXy|PY{&=$n$5+gEz*x
zy&neaN{Ko7^2PHo=L(!Y@+azS{*<8WW80uEQGM0~UiHVuXAZ(*8|@&`-n@PQ$1ib%
z6_}dz^V>J3XWdwKe*0l==ioE8U?Wcn>NRbmC+F!y`i*{lVSa#)cE=RQD^C<1F%6>F
zW|_NkSPS)}Mw9RxN8}afE}DZ4tbaH5jWz`=utUDrYe1;SS*IY6Y-R&Tv=OKAk9;q8
zjgCPe*JsDL61+ioxf}S}f-}e^NoEY}0FQd9D<R+InLbpqX`pfIff|5}v)CvDInH=J
zbvUn$LC)u72G=p3y0i2`JEoE6fF-3xTjRHr(X*%yfWPkBLFOkeOm&R%;FJzI2FLf&
ztU_B<7FyP<a=tP;ZwXA^PULiY3-|ZmOZGrFS}xhdUzKgA>JP2gr*x*)_q}zr*6^iz
zw7G+Cb-Y>1d&zXC8Y|~%bKQC!$LrX!WqXe|j%(l3WqP32Gv`>-ro2(JX^MU+?o@J@
z>i<r68V~5Ew$n&IUh20r?rOGXSggnL@t(RY3BTmFE%o$wA1w9Mw@+7gHMX_mzCC%R
zD8r0|+wwZ9t53dR`|Qc#$DP=iCZM2KIq`ZcJ+09eUN(n^U`L^3^E?JS-rFYY?|$<e
z-$a%yHqxIw9$u&9ugkJP&f`aq{YeDTDd(<#f;w?SiC_xBB>sTIOJZZ_=bwMJ4KwWu
zIe768L8BbxDKg)9U_gQhJiR2yWgGiS1ORxFBHKzJORv~_VZNSDyef~^@Y$xHZO#)&
zTaw<`hI&c&u730CrEjiBK7wCAzCL^+3z{th5!<v0N1songCM{Zal)}le!X#3Byd4c
zBPW791OiEX=L)*q=LGI}!bO3_B4e0Kv^{?ZiNMJVd$Ni|aJ76diw(>B9{kC#H=~Z8
zGKPzW<9<t!atN#zY0T@NY)&WuNigJ-J%RS>#f#Cr#0MoHRD1(Bfsnud_@fIVu%h-A
z$>sO$`*+sXaWF-A<~Hf0>HO`>mjW-ySzqimdA_YZkn<E2^MiEro>w^9Yu(H)C0<pB
zExZo#wT*YiJ>xG9A=<<rn<yn<$qYDM@{JoF?>J-HGN+F>Z`{UH+R&73Ca{;!#(&v_
zp^u?1iHjEkE?D}--~?fGK|}C_#09wrR;;dzbLrF5IPpEp!etn;${JZlvD?wPNId0f
zNGOZl3JL)y2-J~V8_YHurDWw2J_06bB{(7yt<Th9bt)@FxM)3xrnsP}>eS&+lVE>d
z2xaL7FC67o$~2PMDH^dUXjObiz!n?ab`#^r6gA4>h1s9j@!}v;fh)kJErnLU+E`V6
zE|0CAqy3IQtp`ro7pIiHr&lh`PGZ@6>_OjEyN=bX&Ht@^y!CZ0HGNClLU8I>yKDMS
z32U{j1wY0X4mC^CDRfJ3d?G}!Uu(YS966^C=f+2F*BpJcJ=5yEHqtCDXIV@6zSCW>
z2NtejYC*i75iI9<yk`6<>r-QtYb~y&VuGGDh8Eb=@01?+=J3#bU%T<b_w?NZm+OJG
z>%!8O94mMAOZIL)*Ei%#kcR-xPxhq4I}g<4KnWYh7@?W~06+jqL_t)*al=HfV#|%?
zTRz%Q2FC?ozM~#EPphy29+RFph%m}*$WA#cZr-$HWSP`ipt~Ea9gQAd*On6-@t+TZ
z8TNFC1q_OmfrCwFf?qnQ444vBi1oQLo7H<Jp6JPTsjj@GFXQ>14FLfDN<ia~lam0_
zbnhMVMabGsxX8<<ARGEI;eqlEo&ib_E+@tUPn5J00iwfXEj2!#{3?Nu(thRxPnqb+
zn8?RGsGWLbe@C8HxqK&U2t11Jy^Yg2CxH{dDz9o2RmuJ7rZg|d6x(Wio2a?LToFkC
zCkK(?CrCA<5_lWxl<cWNdlen?vaDg_AP*Gu!{1-O_OV-@U;<jwoZmI3m(q)FOJ1{O
zyk51XPOX=1t}!OJb&Nf%Wu3Wx$Wb7*cD85*Z^`c08k>ho3}@-+gFH5tbVn=~G~hkJ
zW5E^qn7J%IN%OiYodT9Q;QI5|$61|YAAtw_h)iA;2MjtykTcHFL$0TLz+nryH=`?1
zE_GX*Ob~n!^*J#6LryQt=}ea`gZH`cvHvc_c$rT;X@(el;r}`r0Qm>J_{pqweF&i3
zJq>91Qg*=zx4%|`w)=y`N%_GOlF6D)z4!zgUpL@W>dR(fXP*anI-6*d+uzesEwmmy
zGf_^D7K5X0X)2a2rT#YPa{F`rT&<wnT5C?rrZ6gYH}jOT_cV5`oh@JPWfM!n9!X+O
zZLcpk!C>hPpk}dVTYhrB$=o4oGUj-e;+MAN6hR*Ol{tNiHn!T<f-m)1jZWtZHQc#i
zy6Hsv;`O*-d*!F-uf1$T4%dYmUwg_uwJgRa=Zmf5T)y62GnO;$l8yRPw>QiG7W*u>
z;ajciCayj0&*ODZ*;{03eEAJ|Of*ZtC^0|s905O<=6L4t$j@u{oX%6@i)T))<^1xu
zv}+5``CE2b8u8id?uP5jvTVII>VM@e34F)rF4Fxu!?IE3bi2;Q$J2+6Sldty8`*4_
z{+ynuDNjI*mpC@n#~ZQ5g4#wp!6jZ9H%_<JpaHbk;mMgXk1eA8hWu?csFdkd$NG3t
z%9BQHxaGvtJL8rX3W8tpj%@L<(aa`UPJ@3lrd}`f_5P}JOhb94QMQV@{-(GyABRr;
zD~HiT?zxiyta<kIi9PM|eDaT{q|EfyiQopW4r46iOaDin&UjiEOnS9hO>ygD9v!u(
zTgu!x3=Z*%zS>hq8=Bial{Drgs8O4Lw26Qh_Qu3q<Ege&q&4H8pjA!^+(dW|NP|0E
zS0abXJ{qhtZvWVOq8;))rswDf&T(n1b&ON3COfWIqB~Wd+q$<N`JmR$DP7{5q8Z_g
zL35n-8h+Zwni4Mp-~p$#hRv_qJkP5rBZ)R}S2Q`n1MVmzM~1Jv=rVj<gP|QpVJQJq
z@>jK)ME+?buGH+N0SvcPWON(_<a)4Mn^Wjy?)9=eI^YC<YR9}zqWfdegoN&oHPrY?
zv{PjV44O){|20WwhyMYi>>8+8%Zj+(?xMGlq2pe6Z1&(&xLePq;QNlBAP!|0k7!Rt
zuY1`U0=-BU<fxEc3&<~mq#G(tTROM(_+0N1CoZtIwbq=wvgyQ4z7z4^nlDdnFXL`&
zs<E=Ju1T4CDZe(0m9Np+qMchSGM{wU_~chV)@H^d>(VPRv0k%9bjl;WV0+~B<r4jS
z=+g0h4_R489+Me=*L9A;8eey9%as1wxdO{W>o_+h>zue#_5Y>!*<N#ZFFE`6d2KXq
z*MPKrC2qGxoN;_{twaKL%y_=EHma>t`CNPF@Kdr|xJwxGTiUgSr{#mCaE+;-Zpzzf
zB3}&_@t07>XRmC(d`bFzRe!ry*@<aOkdN9F`(QqPE&H;qai9PX-|^Wa{cL^C-_jJ<
zw52>%y0RHO-h}>auXf@USA0W+SAKCv{*d@6eGbyF(M5+Ib`Kj|{L88KdN`h*sci`J
zN~YR!ec!(Sp-o{n^z#!OS?5$ezJ^V7*isATl-K;++J;{AOX1>41mFCUkPFTadu1^8
zR2M^ohVrc%o`RbDtMz@9(<*=X_{pDC;|Y$}KmFuStBp4R*$}LIL48=q)-^KzcwLxM
z^kF^V!pGSUgBs#rp4Xp1@{R3A`>?e(U-B#FmfO&#Y?#nwAmEgf->IK~+Q5%sHcH~t
zg1o{&8{?5~X*u_Qu1gMZ;gk!PbNZ#MAsMq@ea?oZn(!U6$o;Qj``m#$>uViYGhMDv
z@s)$(@O>S4v0Z@@Jt+emIl2hY_y8vP&X+dr#1nnX*pA<2fJU!)i}DP7ko?<CW&{`h
zjN2%er*^Q9pbc^gO%dS%B={h(gdECW^_sfQCh2D_<5hL#X%}l_hj7~oei^;POcp5U
z(ChRdfg8ftL<pcVF!<ZB-hpw1OL0Q0D|h7hM~1+K8k_f>x~p5ek%9^&tkJi2eu4f{
zcI8g%Hg^2Qo?x%}q}Xgij2HhP4_J{ktkq+~N-Hvq`b2AL<aZA1hRB`U@oP0eYs;^-
z{af2zX7N_;yW=}{eTe+>fzB1;Skh*(^4MbW36|@EE7$8y*RB=P9r<LdWY_XJdulup
zCcT1D=NL??ofC>qdFB_qmB&75(FiVIbmj=WIme&0Ec;c*$<2Zrj9dF=D__HmN${!K
zKB;sgGS1nJBAku>k7I!AnwXaOKk`T3xpI^p%Ts*N@t5tK;#rE*Twh9`=~``I?jAF5
zx2~hvzSGp#tYXV7xa4_QyGd_ZtHx9{uWg-2W67thwZ`}9G=3X__fxiijckpnm)D2E
zk{WD{?6tIi`s?epER*zqv4=)TfByT=o<F!M`<wsvTAvqjuzh_-UMThaJB4L@OMd2C
zTYr52+MiDP%iCgi=HDy8Be^fE&vl#-?C*_FUx2@!x)T?I+9odWoOm)NZ)<$wv8?V%
zC7c?a?swEyo_uNP^NkrhBKU?&P+ogt0vSAM5Wl+t3qct4=973fr4_m7$(!7cpl+tf
zM)~1}J3+8oAV7~|)`CE^lczNJVE3h`*ss~vwHHH~e@UGuz~&CfY|NG5!<4mkZR#hU
z;_+c#C!2(>(OSMGF&t*256e6|RK>Wnc+^pexzU<WZNcT$c0X9KG$zI_fs*{>Q>0m2
zUV4|R<W|Q-r6cm1BL27z9Qz7`tTxxBGs0t8_?T;4M%daiC)v!6R+>PySgLS+L|FOw
zXfUv79nQy#f#N)+IM%Rbu)D3lMw9)gSq(hmB5!OvaMht#P)1g`ZT6ZqlIX~lI2j;d
z6y<~Cc!aN^<szCVSmGGx7|FOOfeeCkg$G_AFUIIJCY!qJ>>A{O5(q6ht;6}@SY|mz
z85zqy7{Lii4nda*Tp+`*(|2+aY;}7U$?#To11!Uj1h6{ez5O!I_0kTbv2%xe=FbxN
zu{t>W>)6~h$_1?Rk<P}!qQ2C|^d#R4-#swvfjoa_8NL&^2fl^*XDf_WxxUwGupYUx
zARp*_Vd+>B-x77)jFi7;u!OHgBU@W|;A`xgk`r+{zRtCUja)mVoIFJx_&8n*OerTY
z5c^sFR4(3fUAlimwo2T}r}nJLqi@&notqY)#J)U6f9`)?L%cR+8B6(N-^TJ3zv$w*
z=@k7MoaTDuS+nlSw#cvYF=K0P?U9bV$Se0w9%TkT^mFT@q-<xN>O-XNHScN`>ebja
z+;`Y_56pRB%~};@ZRLwHa{2eadzC>))&IfX(nR{_-@X{h`KMRQzBezXskSorvwZPw
z$$2{IcAp1!7ODTOeGfgRT#vZx-@f`ytRIyB`uD#Z40lbZ96ZU?c|qlFdu{4PS~iIJ
zfO<|G3~C#)dQX0)=Q}%(9^&;uGy3=#{D%+2hvMM_M&hKG56o{(EgQ?pbXzKcpju+n
z{qR|z8s)DS#+KfjckgV|vHobEJVVgox36BUIH_RV<Z)j{EWuI!{PjMJV(a+(-Da!w
z9g|lB6rI2_&kmQbxq$5wd&(0m!+a4>Ykc<j%I!<|g!#aR=w4~cFEt%sD^a=HQ$BH1
zUvlvC+`JX-Idi%5u$Q)3K_6XWx@)r-()b2D_#+sii%m9M7&8Ru<3+D%;_OtJySJc4
zmikaQ{LjV=!3A{qorb1EvI+JBmwf0+)VWaCone}5g2B&4qwZYvz)XJyiY2DkWcWNa
z-5hJMA&v=T^5?#QS4ylEq7&YOBfb2>gO0*dg!r5?9qtH6@<TFMpB38*p!@Chj-W%u
z8_~%C>hxXRaHUF}BL?j#r6Qov^>!Mt1cvOm+%;0_rN+-Y3`lz%22T<8{r9^E?#u&>
z<hvY)cdb7={?hEp_EE3(7i;!0n@nqh0yUdzu$rBsivfsJ=$4*K$-`l8)}+10VD#~v
z(#2y(2evtZEAe}^d8OJWdCi`0VVu_NZZ>iU#MB($LHxyw__eoxwu4SEAWuCHr%ay0
z;fqt+w)a@WR<6Nx-8kEq-wD2H4{RMjOZevI#kb<PC2ifaEqmzW>8OAD-OJghmDbpO
z`0liho9&P_8=Y%tQBe!p0Kfj>+W#moNoDID_cAnx#E0(7=Gyp41D^8XL-(Q|y|#_-
zjU0L6TkIL^7)oRN26%7E2t=h7#5y4S4mhy~?)9&Q`@x=^sRax6G+=HF))wK`d|G2y
zE#Tq&vvcBbY!CBtjRn)RP5T;bHx}sORgm#%BkrPM&d0#ub>$V3JQ27x)%s_vtjxE^
zfh5b^yMD^{?`>;M9w*E>kkjq=J2@LJKH@DWK;s+zmMahxW$J0oc=Amj!WZu-M27CJ
z1g;pRf;8Af;+V!<hX#GfQKaGCy+T7kC+gEqB0A6wRas-GQSOe%R`9kIF})r=BJ6=2
zqlbpnndCCv46`W)V_mM%_4Kl}SAwEWSf9IOek{pa%e~kcR*>Ug-7LW^hi;f;TW<CK
z)o>B-;*bpwH(M0>YNr&){#fs6jE|>>FmGv+WlQNJ9WGq6$ryQm>1D>LKPkSZGk#%C
zx$CAc)vqA+Qu<$5H_bNZTlULbK;oG7Z`N@aNT9ytSiLv^bGCMrW8U8C)aESruKK=a
zPy1RuTYY!+n%#R^Q<HyJeY9u0vaCJY6lKJ6)^*ACW9{bJB3`!~E3fd^M&W*5JgxM$
zom0M<!nGw%WBj<=eRp&Z`FG{F=~jNns&AF!kyzftLzmX~OWXJFZ{A(~<MYD@MfUK?
z5*bVFoRSrvj@m;Wx~k9Z@_Hz~sq^<Ae>4(Qogg(v5*tvwdTq^LT~z&p_5_mRhwI~O
z+ddri+BSS-mQ;-KeWwW*5>%_Fim*E2(NH#!_|t~sQh_bYTl|qemRt094JRn@>g9{8
z=g*#%SO2jo*roDas?S#+?3+;Km4Ct<mrn<J%xhK0?J!mvBBHujRx4ZMJ8zJR{(SlE
z)%*A3r}vnk&z?TD&1d?tydLq%qsJqY@c;Yt)Zd6;rqXcZ4A<qNJ(#e<ykpvb_hW`p
zQ34hk*Lq?`E{OGb(F2~%=5hN?T4)sWrvz(ktu<QKTqWSWuT6-{GG5rRNr8XKgSOz1
zV1m3d))#GcJ;>?|lpWGv^2lGlvD1ky>|K-r-$)Mj_!BVL!@r<|I;r@Qa;My(T~%O5
z;SkZNY&k_w!l|Be1$=rL;H3aLy&wY*YxzK9r5P0>==)=3<Yi8LYSE~=Qq_p}mtNuN
zwU&QEum95BaPs!WbHuHFB}jeDJUDeDx3+6HyN-0uT(D)%PL50cv1pEXCBD|QV|B$b
zxze1okuSMX>HOoa)*GjD$$7OMw&VQ%J+$G{A$Q$Jr|{2~HpO#uY>F~s`4sk^O6IY4
z=X2wu<(;abJIU4c%NB7{zL~;p(Up0Y%GZsAnV<UiwX)`wyN9W7m19vXXA95Kk6U!C
z;hAb*Z0}tLdFJ{d+PSp6wB2*^s!#Or+Ov4a`<#$X_(m_k`|$qY)oSzd>+WFl{mGNZ
z2Tx$|RL{Hj<)vu)jt_w)zV&ikvPRLaWKKRfW*yCgN!*mmOuN^M;cG7tQ&VTHiuAD@
z89BjfZP-M*T&}o2ldPy0X)m5XJ9xs3z#IXD5)6ROzMWwKM2)2tN~)x-@#21gqw2U~
zM^)fLU@O2B$7fA-%ig_*Z#y7bZy+9F4dAtQPi-BL;|T8Z8b9VduiT66!v}BPzV#=8
zYHgLKGCEYNPJ2hP0kL#WWto!Rf(t*FXZ~ED<Ilm2Ir<i!@yl45^lL}cl5q~W#+f$q
zc{)ZLNo+{zX&E+dG{y;bc;F?(4*cmBYb;M0C~-%e0tIYLQ<ZGVg%&><%KixE@B~o_
z>JUmP_LcAMz+3RWSgG<HndtB|m=S_Th7q2bU^L8!o(D#No|TPsd!|SrD-D4wXxuiV
zcMkiJ{lrwh5(MiIuCj9;_M_FQ9o8F!NJ|A*%Bll4cpumiHquHCdih=$H2gpVFJ=29
zmkp{~pWRyE4LYF<KmFOUdtY|n^9g-y*vaHtP1ur+zezy>iA-8`Oxi<6)_)3_S;vz4
zvA&5qj&0Mi{Lz*z<zw~7<Jq2mUE7u=?bvGHR=%}ufSzR}FCX0s+OJk?<wluH{FLQm
zPhOu|a)6WYS)|dq59=oI*v3c)@6sn-BoiKz+KIn1u276^5^O5ZwoTDS9LIZ(W!Vn{
zO_415%E8C-K!04R<&Wc}o{w97yQjuDuj6>!Q^$8%ck6+@=F(k`54A!0-Fh@=_w(dX
z-=TPyey*9Xb*CJ+*7ZPKb9*_aB(6I(oh>abcumF}tR`!(Yx_Cf;A{G8uqCp7jWlb0
zCHlFsx7L@)B}KWhT(IL`*%G01Jdu|2hp#tRkMEB|7C+s5x_WtCKE0CZ@3J1-8_RcP
zQ`nXoOke|4Nu8?tUU_R%x%M&md7WkJE7S2I_}Vvj$~Sp<{ml<oyo!tT@iSktKd!aI
zfx$(FABOaJSVv7|qbrbi#g>l_DchD0n`kwz0>c(A>P~UMjqPk(Fve`PW9;Fv%q=wn
zH#`BB(<ciaytja4+v}NQZN1)-;umcJjAgDFD`U=~*1wKu0C{Qd+^#l$%Z%L8`$>7I
z)>|G3Qn7fEr*qz4z5e-)Z3+zQ8aDG}pJd`cf<mu<{^{zItp)WS7!+iDu8T5miV4=r
zTQV|ytykfPM|@dc%O1hk(u*(WjNf@#vNCg@uTG!wPxFB%UkFl=v>`=}c|n_?=Y-k+
ztPGAjr5Iz-$GfHo`as9Aq9mIsR#qJ_f3DAfj<Y&Cu*d(50GoLI5ybv_IA}XO8&2T#
zM_oZDm*-0ChyNXq6j7Hv@IJSl$&9eJ@)Q!Z^rQ0v4(;*am&qWY1s)lcOM76YO+|*W
zI-b|6`5$mkNVhb4;>H0E?tgmv9%6+~hWX14?ohY>4Yx6@paiqr<#o0t*|ADn`g&MX
zI#sh78uKRVEYl1DGN+O}J4G`mV~V$hJB7@at`<DEJ=a@&E%>o*iu}jg(z0<19`#Sr
zt;L<YElb*QZkyKn40>BW50c*yV9*NcfVKvRGPlU!c%fHzN`6ZYaN(50qjMjY;%E=J
zNJo8X%Il`&l1pEfwN}?!dg;S9@hTUNT(%3R9GuF^fgPJ#{us&i<KzgUBj^~oW9{H6
z<A!YYZ3BHYm`&MlHDS4*|J`u|vIlehF3j6=4&8<Q=u{h&-<2I%Z@Kv?LyxOp+C0>{
z!_qjlhStV<r0?Y{a8IkBTkr)U4@mZubs&5Jpr*eDTOjIJNs@h48-LL;nLgr<DI>Qf
z8=GVoL725kyfwZY<bmeWf;_T6`X`oa?V5tU_PE4AA^8JjS?O>W+xTo7qvt-x$X-aL
zz@{sDhEwB06kn?bguNmym=!TSsAVTlCp~^d#BZP=m~?$@-vY8AkMw{1I6o*kL5+7^
z_saVM8(ifnfgXhB%E|O#O&t$E-xyh%;sGY|=;9*6ayi<PajiNdB!7$?Se6y*HGE6{
zvG6@~J+LP~YN<9>DqUP-T?C6y0Az;JpV6y@LK_2_eobAQuz9?e*POK_)$lHh$-86n
z&FT&E_@>g+$K}aBUQ7A@<A;9UV<&T#xkh04TS$k!KW}x5hilzE$L_o(a}+I0=lJ%+
zdtk=Qh~M0aKJuq}x`R?YCdhJf-6_QkP_PG>2TJU?$8iLlZw_^gT)68yJ?QKE#tJ-K
zLY?H~M&NUddZg{~fgdJP$L~cimtYi#BAdEXcDjc%fxz0~3SIuyIwYIGm-Pf4$S;q>
z6!rYtA26SL;HW#76+66F?a8~YolXO)gBM<y1f&?sE>N<%2KKZapVN0ZQ&iZyx@5^Y
zt{8dX3r)O}&(Rn5)3xNntU8Z5x$U`r9xLivF3Fa<h0CFr*p_qKq*Y<L*A6L>@k95n
z{X^_|ARL$MKDdiZ=kCun4&eNB$@SJ6nWv;>oQ;o@MO*gq)-k=6@04whvRl{8k>*m%
z^ie!NowM^)=O=lr<ocXl-|_PvI91=xYd#Of(s_7k>j~@q65H3<y>!3xJUz2q%bGSW
zm-?Nu<$*u`<@MF?|KXLRFUj|D{f|MI2ye9sc&6L4ZCO^XPuaB;m+fD|H)Y4G2ji!e
zBp)7<>ZR8Htv=DyCf^U^clm}Vha}nWvhVKXbq6$Ca)^j`PD>seyEiw70TW)|q*ra-
ze7Ug=xI-GrP(HN*FPfX1@=7?>$>E#g&A}~r2!zEatALT%EE{cTIn+A55e?(jRHsOI
zKkGc~k+J4E{FwT<+8(;<eP)4yN<e_ugYA`g(yTRwYhwkyUu+B?n>ca#;&;A}B^8}5
zkJiUF^J&;79?LstjUdlU`^HiU+TC9L^ybagM+^RbuwV~=AMN`}pKm_fyt=y}PmG6Y
z0pIGQCHzs}Ry}+3EwOuVolSW*)HT{`#}qbmuva+zsWKO3vCMIf`Vc%VErrK%M$f!M
zA8`cA{2ipuAZWlM?|#?}1g<3R2m>9(f>Qvdl#703s7|@(k<bobY=B}&Q6JEuF6I(I
z8IECC<(73qTxw~cSc+%dkiuuAJ&v*}$~#d`Bp7+rB{(opmiNYXu!1S_M|dGbCs~3o
z%eQeYlRZaum%8U{I|u&P-u`p=_}8jKtHLquM@GNatrVoO&GF@w>BQoXGJS2^x3(qg
z8IJMO=Ii`mn(Fk}GFInXzjaKv`0{n*UM=lU1u6oka24`j9OzO4kn^MMa$V27J=NT;
z$<=khCG;KFHFup_>&YqY%4>YowX|Pac1Mq*hcVD!hD1d(<=WVpVR7`HecEc{5<9-r
ztoOk4-#qD=vhOvn{&vK7rhDN#6=YKu_jlWin@YKDQ?zrqEGyUN_-pvTeHhzMJ6mn7
z@dMjp2Ruvoc$|i`-|n=FTk~^UeXSZBvwD(ZsxglrJu<p+w5xt`bZZ;B<IccJ0zG9z
zy?n=GN%t&S*gt;aJ3r&D=Bc6DE5M4k4A;FruGGsf0i$|*%4T4sr5pof5%)0pDB{{>
zvwz$KC*?vW9*CCQ&Cu2{##Gde2<{|kFlSV-u-(6bXkn3{Q7dr7lW2#DS2fKCDi8TG
zq&_Ek0<IqPV%*)gd33mp)LxnB+8RD(-k%~pefqSY=bWq7f~Tj@yHqCbh+})n;yopM
z34hc#CF`8HC3c_F9uA}4H-3Rrtkcj{OkGc}uuk`7ix*NZyroPX9`Vx-@ep4`h3Nft
zXpzuq4_nSUTw8zL+@7ZdRtDcD!Pw-Wj>3eBzt?E^0xQ9&0iI-mz^G1CFQ?Pd*D3g4
zCkgr}s^c#}L&SP71YVf*M}1>*`TTT>JnIaK<qhIHDF>V__yOSxK$IEFH4!5>)(C##
zu&MS>`ECmL3wPN!-@H@xT}`fF%C|U=jh`*4Tg7Z;tP7|^>zXpH>g3WRPT^F)%e-~n
zB%89Wxh<lb%d?)Pa&)dB`Pm-9$-%|Hi2GK{weyxe+A`<IJ#<{EjH$L`^U|?->h@iV
z9b4L|bM`H~b3A!1fnN7PUxLs2`ykI~%-oos^QrL1$Y~u%k$tQyB0qX+V?b?~%HiA6
z#=cLdgn!3fqz8CfCXSzIq~IUl9`1Wbx^tFgQy!-u-h9|>MwYc!&=USg6U)-|@$I4g
zYO83D&1-s>%H#{=q}s0Wqzt`!B@->Zn<?^@*Q!0T{h_gI_;C8o=g-wACs<w|K74TX
z^yw3OQs?ICb9jfHK0M8`7f&Ynrpi~NkDJMQG6GugwM3*`ui&ZSfzb(U9BA{;ZL&@G
z2Wdn8k-a|dleGyB|7f>08NMghmfA-7JCDW=T}<?yfDw9l&D6u|@y0kfzCHBr{rj@t
zW<DkOpr@9wHKye9i1IRbljJuy5a=Q}=1)L%zR?9~Jk<DRr6IU;i+}|OqU=Y%n}Ft7
zd+4k22skju@J3M~${_4+2IMkvrv_ipj*R|dXIh6ZZ!2wzPk=Ee=-*n3W1BX|aap4W
z1`qvKz5_H+p@*@b6R*(oC)cbI`i6*TY5^MB8|@M;IZwMhe*9=OSD+<e$P+L*mGL$9
z98Wy)y_=l;HCNa5Cud3k3pvA>&-?^wYIg&++QR2VVh|O&-@JU;J;5{9Lyd9z)qKhP
zDgmJ}O}3|1himMS7G<HE8WT&$Ok2YtChB7>Y)xz0xkkq++Pp?TYYpoe2T6`8j9bP|
zc2#_JhSoRtXGcq!SFvdn!LL_;`ub;e;+yIZXm9NGZOGERWnKDuuUj{amQAAT@FWk%
zgIiO?cVk?3XQYpPqbo;^vc4f#dbx;ntPAKYwDI51F9+F=$#apWgZbfj%awq!-hPD$
zyfx^a*$WAOfboO#@^mS{Lh8By6a^if(K}i3fp3ZhJpFKc>j0NoUfX6;oQbss++g29
z9l9>j7XOY=UoXch;hndr>g?|1_PRPsTkq^Uk>&#<b`tZ6egBKKgycR0M;Q`m><8WQ
zH|>I+Cve?YR_}B8V6i{C%l#evh)iJgt>}!@#}zP0fz$swKI+^Cnc~6M8prU*gdE1H
za^V=C*i>{|pCQbWUT-LTftKG4!zN()*d@4RdOOSp+6zwew3kggT|ci#w0XpMbFY<l
z7hLf_ZN=v_zFhDYKjAF(K`+<n5R0!oVTdK>D5c&G5s28uL(?HF!@u$kGANT^sreo>
z{8}TRRo_uQSIYe^z1XMzY0-%GE^;43W+Z)8rsFW@ut$Ea*$ywg0}bWUON^fa_ipaS
zLYd3M`oX+rPC`jeDs!ZaBOh~&5B>}%J<#FPj>UQ4FA~6xQ)?`!U1MssU<^ec?aT8o
z>Qp}~A4@1c@iok3X*^__T0P2(eb8!`Xz&-Vuj%r6T~iid@W=KA)IqUDTbr*3%oRX9
zZvZ`I2mMmUACzWEWh}@(`3B0lgnzT*R$tU)#aGs5e6&MronYwj`Ge8co_-Csy803t
z?_4*~Doh6RgCDDDKXvgUhGZbk653YgJ9Do?`V3<p!-;}{UV0s9Kk%aiIR4lC6+Y%j
zaupGH2j#o8VTlja|IsxyKKdv3N94)nn%qo79k~pfnxBxt*qPJIntttLugHU*@l9?b
zTt2!qUeM`d&2ki%@mbDwVCc{NK${S!M^KmeZsiv{*YtDXKrg%tH4W^s=-Rl}4*7^2
z@qRpUdj?6JiOqA1c+CM<l-WAg4AO`5!3a~CcKdfh4bZt)D9?Q^ntzJz1ts6oUY#d{
zDP>BfKjZqtw3mGK5tK18-|+=Maeb$~;(?~fDmtJycY4x#A-x^g{Y4pjYva&;7_RHY
zM;_%zxXQYwTtX`z^4zBpcT5>yVYJ&F#y#~VW91m#zgAwJ-uVyqd4iaJ|GV;Pw+P$2
z?0vjKsOE`S9z3q^n+nepKfkqyI%+A)s?kirmhg*)?+p=L`*{21>e0PnXFU5D`uDaO
z@Z|n<Xys3TwO8*|`iCckK7IUr_2lUTUvFMa1oYH4Y<vC3zg+#N|NIAm4)SSh4#?T?
zJ=|!9%)fVID;Nvy()86{6GdR+nca0C6TyU+FJ4@|egDpbIf!I&yl-$ep&#2OI7?<t
z7YE&8*)398y7{iklQa>)ekP#AQKORLLSI2U2z5Gm6q?(sx9{Fvy)ygD32A}iHv%yP
zl@z17X9sAWILP+$4UsSD35{Waw5K=hsiBA0hyC!(f>{KO2sRPO!dE&e&N|8yKy@Et
zF*0(n3K`FyJ-vGO?)}w$>pNsWus*?uqr?Ggm<!oKPU`LPz8Y~hD}Lc@Z97an5p`{T
zEl;%QV4Uf^y=H$-8dtZKSDz7><H@m{7)QJgknzV!*7Il2`jdc6TLKN}9dFiTx^ey1
z$11^_5*);Z=6f3}F<2=c#tZWEg*x-7{-jZv$L~#kIUGw8C$L|APCkG7YykfM?7c~o
z9?Oxg_W>x(I3o|)9Mny2^$WDoy(Tj))MmZbf7HK_nXEUmZfmn%=}IeEQ<vl=&*7Oc
z6{-NpANP1OJ)<IW_=c*&Ap)o`B0St*_wX1pGa?5|J+{r)3+og46><3e(ZlJ%1w8sk
z>s4sJXpZ0PA16FHJh*@6b!L1p8hxSL5^O3e@SUAnjxv7=(8dF=V9S9%7K@i7m43kA
ze6TTgacT3<t4V#c3^q<{FqZb{BggvGIPTZWEOGLY<wH*LRQF@nQ#Si}tUN1y;yHE9
zF@%j0EoS(6tNrBT!R5u<SP(X2m~*&LVF4$ON#PmIMU>@}aqOpE@18|(rha~bVaFjY
z0&$_h!iK+|jecxQI!;BO4jpbAfEL)8CNNiG2cwLDEWtOds9ecI^TiVq7Vy0Z^4|k_
zNXR3`!Z}46gW3S2emd46BY8mJrEd4_sa1X4dnjkMGi|(tw$KmWczh47$OVnO1qZ@H
z(%R&S?D$)FRcVL9NPDRHZILB1tngRRydG4XgoW^_)yA{-Q|#tq{EpAF!RO57M^6lA
znx}<PO+zEUuoPcG@IE(UMR}ep=sz@`(-MsOr24Xoh()`qt>C~@{2?Ct1)-+vSfujc
zir<=sO*IY9D|8uaWuv}COv$@C$-)T*%M<)KDKUFHq`WiM`9WF$nGm!rlGDWoMe+>*
zIb8vIHvBRHn{B5S#xZE4zo%%q-FiGziR;aIA1VtdHcTx{eJN{Nwri}wlhc|LLbl)y
zV|j|NIT&7a3>z2&f1u=Fe9G~M(6TE^;&*P$X1=*J`f~kzimU|G&WdXu$*a5)?`AW9
zuB)H_HKfFYm#e90;Lex>F(sxglnSb(>bRD_5<f~-^^H2UEfH0`YZR;B#2G_KWWOz8
z)#o$PiO;gE<~Pw?$Ix)H+=NncT2~cVJQ*{hOY<13Zqgkk!9T?wWpYaL(A2MUc~1AO
z$Xq||bk7GGKh$3BFHh79;iBh0DB4_H<Mw#xaMxmFKH&|0tl@;Aua&a!ht75E{ML-U
zA~uW-^I06Q7LNCUF!zdn+3oQah2b%gbKh_JvYfn6Fq@21;^3z#*>m`5JYQ=t6jE}H
zjZ28d>Rgc1#v#3PbuaK94dcB)x9VAA71oKo^a>l?%qRMqc&!}QSzH_T=H@lJj=@v=
zCotF6;useE?P1iLJlEv`HZ`6sY--He^3~Jv6?E{uczdyXdT(6Y?rrkSVb&OPylcz!
zy`j9x)01VB=TG1MxclW7pY&Lon>;XoX>0bjsM?i#`R>B}J^BYmZ2NR(lLwy0MvonS
zI0B`>qRT^Gm1a+fw4Nw{m4gzSGrY2i1t%7%^P+Hqn<ROvhsE;8Ub%5HU@`X7XP?+2
zkuAvB1jb^n7Lx)XHgB2-%ZY}_@GTDC_^77ac=DjCaFJh&U8|$s&tJaSu~<IZe8)Bf
zp6o7c!-ppZpyNv&%!R?i{wrHx*0Pi6#5y@rU20hlCG}?3E8FSiqUqs-Y3IS8YA+VZ
z+1SYezqI_t#sP8)w9U8Qe-~?RA|5HGU-+x%tXuHpNbgYT8(b_x+Ch$Q%xL53%ohL8
ztRF@H?d=@j*?nRgHPn~7<jo9l$k!WQ_YL>+&pvg2HrD7T&tJTB8}-^jxd4BC^6Z)0
z6t56$_cieozkK!D+YveH<Ne?Cv^IHYJ2o+MU_(()bB+BkbL@6JHOClZBZa4U(4hbx
zeJhX2=r7O8_<rl<-X@b1tM~R465qX<+9TYE;|aRDKXTIc<nd$w6F*N7UA}zln{C({
zPY<bd8sjVUGVhE}9)&&kpRs8*&Mi~PR?apsZJd5)8(BZvH+||2BqX+=wKVKQ+jl;y
zb!n5wTYLp*&C;s7coa+LQXam%@kBSAbSW$61Xd^9!aa4&T%!+S^Q29#Gj%$<Z}Unr
zB44P;zm&t9dyDFt{Q@5s0+(qFx|n{9&K!=83wz=R#oB~|j{F4+`zh$@AV{R~Ar}CC
zw^ibsOJWXf!5<fZXkej|ix~9FTsDBf_q`V5fnTy5#|YFr&oR*0Ai+bz@rroH>SP$~
zTM3JLg7n@ur>$@ZfgjnN9vG^G@eFJvFg94Y@q&#Tr$NeU$SO|^jLT_5N`(mCOwjoO
z+7~a_Z^pz`T1%@Oop%-Bk(svC<q*}n!&Vr>hJq1D=b*7=-F+N;#!!ZyvoaqsUiIgN
zY;J0h7r^^=g9c+8OvdP2OEdpz6Bsu2Mc;fh%(;b#Y<E%dK}YqhAYUw1YFKcZVa$Ci
zrxBsatvU7zYs6}vTK`tXwS0}g6>klHt+SRnwGHSeNKdvWr{Ssd&=iD2URWx-iCE`2
zS@*GQobyB0_o1a*^=i>t^;*l<l05}Y_O$w<o{Xp$!h3e;*tK@v?BU!7dhL%oe;aIE
zw&BCH0DnM$zk0b|BZvC$NI5*Uuh+KOvQ3MwArG(Jhl{Yyzt>++?7Pk3kWHRez49hc
zt9%XZ7#M5x%<;@^@*uai-_?BeCQogHzR6S5hm9UKRJe<s#b_=1O9_j`Pi+x)=`@nX
z<+H~D$<sJ%HhQQsPxY`s%TojO=7leASM!&iI$%?zDyr)@<iU@pRL0$~sIz^Dy>P9?
z5~%#6e2#fC;Rjm`<B1*aX2y5?V-dc!<$x`(?XfKu%JE6d7M&b%@%`iT7q$`h%s0uh
z-Z5<(VYcapZ+V*N>cYA$0G9T8_1Z4XtBoGo?v-sgupuzl?pnwj_4ZS9{%T3FxL-f*
z)#9U%Jb4s5!*A?We%k1HXd7u)pE$XSfF3p(2>9#O@D#~=+rY66JKG4VH+$4CX$$%~
zn{PB)z0m_E{^!Xa-%x?D5O@q5y*xctkB2PE1_b>erVW+EW*==kAL8|J53M))Q#4kO
zF~pc;L((^d%AG~jx%Gq-a%@Q9yZAOw#u`zTdSOreU>|;;3o>a<A7G9O+xRBe_=-7R
z?fAp<o%JK0cAM8%+v4+{wZS*vjvJX;#IJdCUW@66Y)-TmYFcwb*1I(g-h4|1;u^i;
zQ7l`Q^5~aOd$t_nN7=C@*1D)pgyXdWyv8kR!(t|zK$Nq>2{rmEH7JXmw#tUx$VOSn
zGGDL|f}G}aY@?K?d4xGUX=AL&)2DbnTRd5UL~NcU<w3Mx_zix*)003KW()F!*Z__N
zTC~T4ZpcNsvkm=&3vQQ@{f@zAjCNx}mJJ`tyBz`J_@WQ`Bx?oO>veYVs4M)~1%JFs
z5lRVO4TPfH5~-l3Nl3AHEstVk?R9XziNs<y6lZZIo*!vgV}Ity_q3zQhA#)U8%xyv
z>BbD%LouF|w-v8}*}^-AS8tTe@op(QmVHY)>p3znjODa&hC|_;fj%~BhO19iYhDCf
z>DFdW^v@bk7De2$b;|M6%H_@3Tr=j|Jf3qI={#P@*IteD@{y`z8NB<1rSUOp+6Vor
z{rjP8liFhQq5kVxmTk!C`YGUCR%6YT)%~QF-Rk)zJZve4(1|M9i2NZzICB_FHfGt|
zva#H%SBsWyS9&d<_>U3K`Iy%a>&bL<WnPX`Tl&DMjt8>g5Z~o`ZT0CPvRid+<)61r
zp4*E0;qd1^(Yg*lMoqF8v%VU8uB_(e($3!iQ5SUbt#h;v>x4YtKPIfVKC{KYH&L0}
zZr`c;!jJYkGG1HO3u|7F)|0&r^j^FE;p^{rzy4}`<zA1kSBL2>$s^Ypi?b~9^W;Jt
zlE*?RVX-BiMi6D>>7@d%S6b8AlsNZ0@_EXICni`J*J3){ELO2u!y-N#Jgrwl)u?s;
zwLPf-wgQXN)Qwu@iC5)IwOuObX^9_x{Lyrt`33;5y7D`){Q}x_TAf%-XTd&f;cnsx
zTmW*C)6>C5%y&Qh;EUV%jHh1+)Q8uKv3MV^{jw)0%wMti-A1c7c{F(WMhs;<&Ba~Y
zSx0=!<g&Cy_^*u~_2GA^eB_K2I-=i^msK^_G@oDf6@75q2g$NL$h=;VHsom%e*()k
zCs^2{?ZC%AZD{gJz1R2J`slB%J~p*Jax5aA+os>;5B4;RJ?X|qk&gu{<^rDA`tkHi
zJULUNGM3)#F7L*x>{|FWEfzJWRP?;DF-bpajY(vv6Mao%GTLa0)5;<D{g3qBM>Z}P
z6Xj_=!(_vgi`)@&m-;{pt)`9spq->+-^Exqhj>+FeyRrzbA7p9bgL~~%7f2I+p$B_
z?i1+QVs}<gdThNdDBF2!U*tT#x8Xg6`h=KG9=E&9<DRhr!I;qZX>>u$ys5hq2S7`x
zu8Jwki#vWaHY1|nhB2-V9o$%NF^j4!!1)mjLJ1r<e8WftH#RcDK+3=a2n2ZShk4D@
zlC{|DWje(}k@&Dr8%fZqx1LNAj$+Cqgbpmu<I?Me<E4;tp0cs|z*5fZb_YnYJ$q1T
z#bRv5@K_l9R{ZC>i##qeUj|FUw4~~<?*(D@mNC1PKa6ev7R$X)-Ij%`wxvPEUU`$4
zpg3~vU&+;_J$mNrEvSoU?g@%D_`>;Hh(G${c&*H4PRpNH=gg_jrSalVWSNNgIEJ<K
zWYk<vnjm={&oa#!?5E}xoj7A=`doSrFU$3?MHdb-sz;V5E#4L#n&j5_uf|x)3tN2I
z7M*isT24!5DXw)~<p*_rSjLNUfErQ9r8V-k?AH7i9r^rYtDc&_)jlm+E%zbjdp`F1
z+T=w{n=`+`oBhy|&E@h(P2*e_w0xdb^GLRiYrd8ZJ(*fod@YO~uVfS>yR>;O-s_HA
z^%`xMFCF=Q8mYM6XxDjdTem#XMr?83KH7(Dd&`@Lv||hJHqt({I?iqK;NPWh;H+tb
z+vt<L$+M;oYkAJ<55Bhq**xI8Gp7&nS|=8l$H25glN=s+%~lRPVWH5PyjqW-i+_F0
zh5)Ys%ij-KuG8Emk94zOPWz2**u&k{e%TjO+k5!q1(9q!M$A(pFYFa~`v%|an)Ic8
zk+c!1{?hgXn-quGE9=TKvTriIv~d?BuPk)>s^k24c<oPx32nc4>;tD6lgug6cV%DO
zr<CsF(|h*BQ2oT4UhBug^kl)<bl`I~mw6&f{+DI<jQ+;PuY96d<;RXsg)a|J@E4AD
z3&qE-`ha(8y%Qgp)|}S`r4O}Ykw1lFX*O2|USnWEh6`GF^+XOa7#z0f=bp5yQ#NvZ
zVZi{XA9p^oksv(ycr}}EE*K_@eaI`2jEy8D#wUtgftW0oJ;?iO)eK8vfD3Su1HuYL
zCRPwln@2;LB|skRwYu;`kILO%EBDyzg8iVfd18|Mw>i@{b4s4lVDduS<Z;3jq>zZh
zU=mnDKmlo0pr*q^FuHC%Mz?~Cm%OFf&_PFThp@W{c-gj19whT(g;ehwtER}G>JHd_
zZEMNNUd1h*7OltMqi2k{rz7B-!<Z|}I&OP<&8ySa)ZwS%k2Q7O!o#ewk8fHsJ|(pZ
zRvNUAXqL%oscg|GgPvwkv3uN;nJ`ruWY#=E)58{BaRoogws@HGTk~^^cLKH{)9NcN
z{7lQ~4XPGcPlH-3y{*S3Kd$S9r8atFzU9dlZ6oXl#ZG;7BUC=-?&Ak!@;GVTK(`im
zEg#O7PRS^yY$<HvZPjN9|HjL=RPe@)J<I_5!CW|{E6QcZUgr~YHh(O;>VZ?X<5g`p
zLpgRG>pAjfYPDW#{iF<%$M(s3Yp$Pj>HSV8w`lpUhNbCtt%-bRXnYimCpUPSrgu^g
zs$N5}R>OL}sFUt;nv>Ij5|6x5chcZfskr-bjoatUCF}9C)G5+xH)0m;>yxJCeszdd
zr}aboE;AO457!Gq-FLQo;&40AsWysu*$SQfh_h9Xs2|_MdOAFT20u?<@nlwiNg_~r
z#UZjeuoU0Ym3cmt^p$#CAxvwsP&{Ss<o4-vYxHUk=%^jYTlMSiWxamFd%GE=yRQYz
zCJJRdrKyDs(R@BKnYafrdL3m7)bD*+O;TOK8A3{ne)lYuJ7eT|s^rRAg~?MY+quJ=
z4NlpKI2>N?SXY0fjZH4({HX{wv3$}EN$Sb7ig#=SjOhtI_Ej+lV*`jVZf1?zyo}eS
znGk=>c`?6Hmq_~t(kK#qezQhwo<+Ga7=NgQ8J)5m?F7q%X|IZ#%&oS`6ucpq@ni$9
zLdZ`np{gh&5vomGnxQQ{J)(IJ-eZPS;U3e*IjrnM_}DaLbCg$nrr|IgH(FS9O+w}m
z%DUe<AId((>kw}bS1i2MV`JIA52e0rl55UZt=8DiE30|3_|UM{vun=#^E|zeU5p()
zr6wOhtIv}&wvMO7%yVn&1I8BKth1Na6D_^+9$j_lmCxP0C)s`Hj-JjIJ^L%qIclpc
z1I?PQgt1*ksbkfnrq_A+KD5vp<ElgLyPVJCA@lD$&N4Q5bzVA75$FDLDE=wz&V75W
zeJvUKXPj@XMH%_l%&{Z)G?i^x@-_a;^J?u+<e7&v=2Y8w9bD6u&pOtQZOI-Iw{%U;
zpA&dp?^LvyJVbU8yke+z(JrleB1YNV@v_E3h5()Zr;X0JzNr3HV_nBF$c>%(nd`GN
zVjpx--43_8GT3_l&fd6m_pCX;Jb!z!dv<SpgVmvR`N2MefKIN5zp%GF?Hj(e2Q0Gu
z5B8e2PiJ1!hMse~<MZ6s>*0f+AQ|e(82jvgtnYtkZz6hR-%x(zZvwF!aBa=t(X)rU
zm+yklh0BlH`>DO@>!s}pFqf;}Yp#3#!|PZ(V7<pMpMLgu_rNgU@alOB;D7tqcp%|t
zKYO-&c<0{k>ld$f_aEHby?SxJduP`rgl8tF8-_10-tOMo8=weJK7H(bPt3-P3wxW<
z`NfW#%L$j3M(2BbE7zxg_xa9kF+be3>y>wJE^Xa~x;@&x`14<#|DDxA;eYub|5wAg
zzkBj`zun#WPyhH&w%l^FLC!v<6?_}#lc&!H6+EGozvVF$REg`btx`Fg<)B1pZEi%#
z_53m*2)dX&wN|wQ^}h`V!DG~yr%pcq^wTR)e5dK9efQ_kAfIKN-bsf}I5O|nbW2|g
zW6j-~ExEP$ybA7%&pw|({P6roe;wR}*vQ<myZpcWd@9TLlKdM?4gT9qTH`N`r_4)V
z;L?O~mg(0~*VZEWP1um<HRjqlM#it^Ui?Ie4U_O0`QW#W1(zFR=f$TXMz#bIMtm@}
zI024AOBZF(L$f$XyHxHQDWPKErwn`0K%K+pkl>eq9sXpCrP*|X*Yifd%gK+MXFQhR
z=^PJ6<03t8lECO2JjlC@1KE!;@*n+lNCd^8w8iKi@r=(g1Y|NEN<Q8LS7;UqXQVVz
zU}QX{#aeVYR-C-~Q$fCAWG8$p7mGo~SGkW#!y&Q8@*x;&WvPQazE<WY`E-2a*u#cv
zZM*6tzMHi5r=(kcRh{a*=&iVQENJn~(YEND!|9cs3V*I2&Xu+7Y}p8nP>+M4uKNW;
z$0#KB#j5RN9^*JPx5&26!rR5wYqz<1JNqW5k5#j|jgw>PI3-T@NzcDM{rH{mQ@Ngu
z+RXoEf6a5AzO=Z%J8#HV#aNWnr?>Dp8t9jtpC7ZIUAG@JwvIhtP!CS^kLjB&t=97+
zbdGIv?KU@W%U@e;Y4P^#Yw;eNc9XgfsSD?gY~c6KN2u4S`rPzwREN|eWLdX9FK_7i
z?eBg!0i4?)Irb*cL~+R+(mxK7ZP{`x+E%+_y=3fZFKpB4=@B>A*4B%gZFN=qU;gc1
zT<o9!{a*+1?)*xHuPcmh+O82hnN;hLDJ<STdHi@_;ls{r7j6%+ZBV!(k?oyhr@{C;
z^de}jzMd^CQt`*Zuoo!lm-eLQ>D>-HFZMT6n*H7CyS+Qu4((4kxTQ~ypV*ZeG1JA{
zi|6pOjJ{`-Y4Nl`rcBeEwE4-IEdovh-@F=k$){z@xh?fuM%%`*)?c7zUiyNXDATuT
z-_Ctu-+N*Wy$%hMv*6pg`KQL&lIH-<Km4Et`B<EflSFjuz^a4&w2wv0L(|0qSlB!K
z8`3dPQU;IeOAIeTr>)A1*9+|jJ;6Eh;!9zL&aq6~<)#$p6Ac{41JnM%;K>#?X|x$b
znoXWi03ViwHh$HP@VgyG)Fuy{9>C_CXDqzi29aZkpA92VBN({A4_KZkIC=$J$r_gm
z21?KXV0^1Q6+6u;F<{A?g2Kk|fm4hRo1;$lM3;SkkKn%ZRR0I<<Aznp^ttCt(03gR
zR+TrJ9>mg9jmLQ}ESKu|#*+fQ)(1s=%ZPtF9Ch30n$MNNSNpooXRJD{J7R14ZOU({
z&#~;@+DBV9LTk~t*oZu~9Ew%j!PFkt7FXxS+}}uZ&X#Fg?5}zE#X(MUaE+O-O?&M|
zeS0An<#guX*Y1q*p{BD>dj8FI%f6cPdzRUf&iFaK1>=;w*<Yv7v$f9I9%Qq;$`oU7
zPG{az)a6*1u~ssz+o9Om4`7|z7aCiq7&E89JjB2Fvi5Hak88DFkiRI;$ABI`H{iCy
zjEWZft)#VUw$0Rzdzv7eH_rKuvbnZvg5E;n+;?n_`6N%T>T^Es$%uZ4tT2up_r^V&
zH@#5zv?n*P$%AcN)$-I+FLd8ma2p$<RSE3Old5>K{d@=DyYIhO(XEK-RQcoE*<K86
zbvjL$vB(!)Otf15x+<wf(3U>gwC@uf8B<u9_MxZW+GBR#d>fD49jh)&e&T?669>78
zU8!?V=tpt_uVvio*YJtGp2{|Lw#v?#o$HsXJtuOlW!b*1ijO__+_%S7xf^HpEG-#{
z?Rva6kq=L@0bvXI%+K%d#3qJcjMxuzSp<7U8M`D#!w%B$u#p;DBV{q)u=3_=JQ;(;
z=sj`-$}ZXMFrawN7_Nw?Snk@ByzlH4YF=L#H{OzGGbR8^RN&v5?d+b&N8rEHJ|;!Q
z244N<xu}kBxKNh}1<dSldBgNF!+lQz)fQ80tc8bE4xvB!(fF)3&hp+KL-x%T)xoHq
zbhYOHf)6(JHwDH)p#nC*<ct>QGI%UMRtACvc<WH)%hNd{j?sPd$LyzXdI%kZokojV
zM_(vRvr+zDU~TcNU{#2$s(0+WEwie%HqM<7WV||EykC<)=5K)q)b_15QW<9Uo)*3Z
zRrsJxz|TGn#oHV9w!PI}dxM=&u)aQE6Ws5}&7G_0&m_{VYrlO>&DdKokkcGo%dN4G
z>$x?2HMa8m+AaBg;t!>7)pg6h@|zO8BAHrCT+PqbJtO7xRg~LHiJLXXo4vMJx~0xF
zZ=IjhJJ$wta!X~$#syDdseV~Er<z-?)0ePi!)=SP+i$S!O`u0MIQzH<Enmj@{Zyhp
zR`oFQxwv09*o|^?AzvekZ>m|;{o}2DYJW{r$@!X3d3(7nW1)v}9dw?+$>G>-n6dx(
zcGSnn002M$Nkl<Z=Wlnv{>2xWbh7j*>v1xpHv;$i>6RySzJBp?_w%PuHflZ&JFkrE
z1ugsZ%0JRi=m9^FY{1VQF;2!}F%g_$4QeG`##$+|Hyf|gIy5lR^uk<OPi70fcelkB
zjE}tGfkS+`rjKv!3tRaQX~*jvb8GE9r2f?BHr8ov-Fy8>8zaJrYY?lcO|;Fq%!=Tl
zjEfc)s+g0t=<S<V0`nNJU-jIehtgE6bi~sZi`>YJ-UyZ$GfonY%E#hC$cltTY=6pT
zmH-I+OgCPAPiU%Z5EwDDoEIm8bh>iOSP^)RIRP#*13y58ndCrE*v(BIm<gFLD3gXL
zq^r17Ip(-2MS>B}l}$Of=1zya)sijzjG3XSwg{ep_mJb$3Cvjqd0nke8_nse=gn2d
zABhRObWWTh$2l%jZ+F_q!Vw0<z7TV5yw!7Su=J#-csYJvJ!bYbx9*tk9ozdTH&-&;
zTQFk+zL9@gZ2Q!A4i6o@a&&PM%xw#`4|?_=YWLDScqqQ)xsiu)pktEiQ`b}cYh&T`
z__=?nHJA4IZ-Tz>+*aFK`|{Ghr@&(Vl8<Vfx{W9Y8#~w93ZHP0&ta=QS@t^8H(_7h
zc1N#g<GeIq|JC}l54cQGo^ueTdp5Sp*Y=wpW=)@5I@N_Ykl%8QpEIxJyE*>3vSVVQ
zXG<HRyY5phTVLCEdH!M>3h(S&G+b%&nl}6TOpDdOXY&{PUW~$)`u$kGg@f#tdPvvW
znEI>@*?y%ZuUK{~KZl)tGslN6>8trf=gFTpKg1_DNNW{9<+a>(<_}+izfNoahII7?
zmgu^G)!p3cRBK&yot^I}yYVvy%O`c+XGCh&*}Qx_%k;5ul?T*sS20&iYx;PvFH}&T
zrGISUZ(*E*cC3DT3fsWE?e?v$cdtKb!O~CGt+L){t)mb#mtE<@#UUz-`66D!=Ke}&
zZ^#O-ZD#?i3iYf=t-J2O^~rqApJ9X7p28_AC$d+7B{}ks6s2rDO+X4=SnaBf!b8up
z+Dc|96~SU9Df4tD3<B-Mq74IRk#Au`n*ynQ5TZmrC0~aYV$Jinl%Mm}b^GBq*|N3n
z(;m51dL92<-#yBC-a0d`aQTRAIG{JjUjg`u5gM*oUBkIHo&#y=T!W>jSq5GkgyYkb
za~h-Um}9n82ed%%m4y}i|E{^dz&C?`8mji`=W}?dm@D7XJ`HnELaKc*XaAuFoDy5M
z@l-c{R{xOBx}ID6rEsL@YHWzd98^~o+!{*lOCF`f?^_3Z-9}VUv@Pu;nc6|iZB|QM
zZ~i9iyV7<ms=VFKsl?nl&XhA>5zNn}n4M*}^p_P?9$2!Dfp~+7V;w_|S+8SaZFMbE
z?wUS*Ve`|)E89r0H`mb1czUe2X?A&OHw(OeGnlfaes2u-I>H;~+m#{Pue9X-Z77Dr
z4H5+Wv=p*$=6KO1eKnux+;H^v&1)}?n-`-*<+a>(=DV+MyG|oM6kTf{+gwwhLmJ;N
zj0ynnoR1I$x_HK}^J{E7CJ$?za|g$@y7s9@_QP5|Exysm)6#uxnj&0(?ig)xh}CEI
z2y^ca%q82jAaIEOEPJf}kui^vzD~P(y5{<JPo`IQ+B>)1EcRCLp3)U`J#o`w)d`n2
z9P&x2HZzF5E@m_56r6AokP$W~Z{&6CAni2cL)L@Q2;A9AIoWtETD*1*eg%~?uafp)
zv1G|Z&*wN{uhmF07LSl)3fbX8vnb_6a+t$iUwL&K%oajd(A-CS%cs~SKV{!JJ}RAw
zL^*!>H*ezAZ{Qk({aLwV1*Sju;dU9Z=<rH8I|Td7&Qh9J(>cHDO2&;SF#vTayb|$Z
z2XWJF8_y`E-i9CjCCZn+EdnI{;r^M9-L>xh^1S-+LkvBCsfDhyPWKjGv2mTYtko60
zZX6P8<$h{>t=FGu9{$kksyTFDyVWP=<c~eR*65tGM`ic#Pv@~EJ6rR|(z&E`)p4&5
zHN9t3k6!D^`C}Xh^7&Q9K1CYe-mTi`6neH|ZuJ4@Hp6+uR%O?~UylQ+tm|3#P4Vp;
z_pW#M#k;L?jMdtg_wVnwB-Zkojz?SCM>zY=U$@oQb9VVSoMY&e)os^d+nY>V#9BqW
z1?Sp`bq<BC^ZO_n2=r4z{?tQIKJ4GEe)+I#yKH^*r?RHdVAr_?-{yIF|IO~#zy0r=
z>x+jE8{9$q@{yJCw@3KdqerWyWZP<i#Q3SzXH977fF^l;mq`AY`-YM9{px3*DS2>=
z9iRO6H;zNPWpAxy4NonzPyV<o4f|MdLK&R66j;Ps8*~tcUt+}jJm&0+XER3%;b+^R
zNuSE4TlJ_963UA&<ZYqpqW!2}*0)buc1bo@u`a9gl4-^Dg|B24*D^hN&38y%zC8rv
zx@EQYDZTQO`j6<Rh&gF$ZStV2Zr_&vH8#@ERVQKebgCTOEVD+BWS8P&VlC;a%j>qj
zQeG2Mer?~*eoNw$WPL(DhrML;1o*&A&klM}ECSrw-Q(l!);W14+3<>)VkCW5^P88G
zhhPQtIoci>{)S_Rl{pRJDx8&tfAp!?d$X~+O$B=>cfR0--|6V%A36;WdyrK-d08;W
zpaTP3LP@I*MsPXVOSN;_-GE+nP*&{pgk{*&;ZJrMob$S`&{<@PYTK*|E8+KmJta-a
zSdK1u^anS1a7FCp$lqhpFg9f1@-WC2e$j{6=>A<E2=J7Eoc*~PCl&oh<70>ly}PSa
z?(v%U^-KhoGrhSc)!+)qT-g=oijej7aC-C#s+k=<+AMbBR{YGbG%<gSS+%Xo`kD*>
z97CgQ4Q>|WRW8~<&h%rax7u}V{l~zYJI<e~e$oj(<3agD>%%;~WOE*WwZ1iW)Z<I~
zz*V}&?d8|-FY&bMLp^FcWdnGvdbVhiS&Elzl)P+0SIchkDF#>9V4^9m?S;2zM~_}(
z2zxFDOUGCEdf0OqOJ%3PT++F=yd2KgvfmoMHT`YQ*rhgE+o!V*?xJ48&HTqqORwU!
z*c!uY`N7vX%er#9#+qxZIXUU6@oqkU)oXM0SYt2aa7#E;-&$S0`Ff3fZoiyfBU|&}
z9{z@F$eQ5;&7ZOz=f>S_cUNyXWcuHI_H1#!J*9tZAIi3#&K@0{AJQx0YTTShN6jm}
znNP27%d~IL&vSN2*LXugBeC@e*RoBp_=CS5&z!%6u_a#I=#fIjTflFcS5BO>U29yj
zwdN(SKk*O?n>e2=XP!DOY+?>>zMxh!mAvqCv1s`&MWvSxmII#wQs=Ykjw~j|rO8qD
zXPfG@Y^>?z>#{oEk`YY{Fv?4oxMi2h#ZRn9`ZDcU>6Q)0!d}AJaw1XNDL)N&@^_SR
z2l`y7aa+9fLH_EtWFbp-wB$rvGtsv2>x~BRWG^w}0sW%Ya{~PPh9xZN1V5+u)tBX3
zb3)d!#^)KYPP3rRJFO25``VP{4z-~vZylBT#fF7(kk>dc<!|lcmM3TE&wN*a;JhX6
zn~s)#XMFT={(Q5-1V+q|2SdI@4w=BSa^XRv3}y?^r6YJlPBiWS)AxeZxf~n$grRhR
zmQm*Ap~rPvMcL57t~^)d{R*c8ZFccT|BF5t`LG$A%Ih(_4u<O;kxd;4;Nx=~TlhR4
zH*3PjQEp5KP5%bHHrLQVJT2tl6xb<`>gJm{rH+O$3FucqmcL??$Zj84_Ggw3<klVL
zL*WS%t+|FVYL<H#EsQ6M4mBm~la3m*N3Z!>dF<uS`*TT8>X7P|<>BdVj%0h3CgzXh
zgm;sYw`d4iuZYfHGOcoo>SMEX3S%keuHfsq^YJ!*b3?uv=3EgvaqBoQ)~$X=u5LSF
z<ybnwXFMpMdG<-u=a+2G<FBQo?vr)CmXSVim9BYOd2pBbGY@qm)EnZh`ec9B_M<bG
zJ7|ZdGUB;*U8+aMA&vfQcb3ogLNC>J2$d1nb%0kmvQ_+|J4pJeF;Y=C>3R}iN_J(o
z7J!NEvVW`m7Jm6^t&E=w*)ilp3+itz%UiU%eoO6K_vu=`mFEeDC0xm^u~EEhV(C@9
zMrI!#@HNh|uAHuMX)De}>TmBX^;PMq@oqXl-2apenX7+dyON|I@TcR!6^0IBtG;|L
zxTWv0mN2Jl4hvZYwdEYYc*VoCq_oOzbIdu~soQZ&{WV8!k@-+~zrPq?#WrX2*aN3r
zy?)iBPo?K`EuDY<@x|_sfBv>Bs3&r|45?rZZ|g}OJ>`?2e0;6Jlg#{VFmfq17vm@S
zs^xc;&0%aQTYOGc3KeewzfNA9)v~ci@7Y#k)cINtd?q+e=5=|_3yZkbxS6+?&KU5h
zexmo}mgs9HMs#ohs(DqHtv1N^I&RsTj+<AC*XqrW(_R7DNRZ7ETWf>(j~TD&d(wv=
zb!}rys!rx^^mlLH+Ml(pz5WM0O1Lnkt>I^pcP`AAh9o@T)@{i`R;ypt`tr$BF5{#*
z`=Dp4*Tam!4%w?1yqwPXYtmk?HH;Nc)|t~4+tISMj(zRMAYafnXCHzA&cvokhSGl6
zz}g3yDPeQKUo~hQhcx|%K>wl-$J0J#oY;dSF<$UTFw$S&*}eiXGI`ORO8e6>09x8Z
z>9}CuwI_)<t|0rq+>U`h(b$G?K7Q;k=%9ptf6B&kcTJWPl(5NCMHkD6B=6V-I(8F0
zZvyaYnGqVVkwebw$14eppUoE8KW_ZMZ8$K;e^_ll#Oi^bVl-U7S>vTtF62<+c7ttc
z`4QR>ybPxXMH@=k<^kPEjsz!rmNF8Ghs_+}3nZZ8x5!xL4$dqg;77UM+$8=hvD<lt
zdN6?+$ST0{rqhIOeTg7<c;=S<$X@SH4cl_vaxrpiHb)rGldffQ(wXr{_vqs5mFMG<
zcyD98@7<zxe>+9pWZOQvdhN2t&YNsI=Cw)ZTw`$J8P4I1rt6d17g<G4b3Vn<A&-p&
zeI;4rUSTjo=Fwc)8^2jz=XQBvRCS|XTkI6>b+v1*GrILEC(<fY46PMp>k?l}b`33#
zlcz!t3u^Rz#_^iEs9(+TX5D^nHOIW63|vC4Yt|K|`~QNaaV(Uqt3^A8e$5MWSaW5v
zE5|Lnb9pb_(t~}RO6`l;(*<2~&m7NE*@eA1<*r@dVxN484D;40Y+kbG&~iWTJ(A&q
zSu<`Au{)m&N@tDN>RUC>mV8T3i&kU8gDq>1y|w&J<#mpBQ~G}jdYK=%=G8I3SEtYJ
zuY1*5Prla=d-RNRv3t8aw@p30CM}os(w{$jw0ryX;qIGPuXjIty#94?!h7@n?e4*y
z>Fa!7y}R7KF*^?Ng~r;Nddrrjcpt2NWR~ox%NI9#Y60^3fYmC$30lutJ^ET3b+F&J
z&b<RjebTnpo|<n>{aQAxk*jr)ZykWZLdVu18|t`+)2c_0cZq&XtR-E&@|&_@Y(sh8
z+O=%($W}_U78(2_uXy%}9k+3cJKD4zb?Lw;#Fh^A+g!W4UFy8>a-8LvD;N<hP-c4O
z@1<Mz^muOzU3Ju4d0Xt-*N)t3x)aY&z^nuL6EX&TadEl3cXs9%-&qfQ$P@FA5GmU<
zGI?LHGO~E+AkruPjOXdZgRkL>IMU*mwn3(JrWP*pgxDy--U#9qej$4VO?GROhdII@
znz2}$2Evjon>+-+NVFfyV{Wm<MDvN$og;pHlcx~1$>VmG0ql!Y1~QZ>ZE=X1J|BB%
z5-rl9!D#+UH|4Izcc>I#7+M)PWTO%oaur~P;@&V2EDF!eFwYdJMx09!hPElFZ!a%H
zy4c4?VQyb-pu~JtSSy3bN)Kvi0`j7Zoqo*)yKG@j7GEH*|J_$^PDAVbt#qeBOm!Iu
zCr+HAo^vLjT3^e@bM|a4J7xROZ%@78w57h9Gq=cOUbO{n-V3?^=k(h1sf>H7z9m}}
zt3LDv(Xt&&=~LOcq<3w3ZQseysdNlmVPI`rOJohtsmAfCa9j1PX)WC?S}ix1&%Ok!
z?(ez0g>$HWFf}jCVa=7b>}t`L>Vk~)WM0|V!aS5#`+E&uJznpX?<O+LWd9l*@#MOS
zc1V0M9soxQ2itNaJAmo8%|0cE8P?Z$816^u>-WHxJ|1(~VXZ%>@3n1?JiL20)7O{%
zbANYlcfY!J=9*G#7%c=gY4}X;<Hvh#H(c4|LC0gO$K!36-7V7&@<kKy8i;dk+*3ce
z$`_z|TowJ42iHo~Nwax@qo+s*`+e)&8@AxCc_jsns*uy7DV}S;mJM@QV9w!5XO7Xa
zgqit!=`My9wCd5rS)w0`wN<`VuO3E=cd0&<uaPTjf>+?w&#x4$Ags|V82bEpw|zX;
z4(;O7o+vUP{^)xxOf4b`s}&<BS;gXG(Y~h926a0u<%L&ciOyWXqEU^b@}+oRxg}hc
z_w4S`CASo-4;_Nj>z_-yk6BJH8DD#5Ir^im`_nFdSFYvH?4zyzx;83nfBvd9=d;&%
zu}J{g%i?uvYE$@-^S%tf@i=}YT!^H<O3BMZUN#j-mkD7LVbB<_WBaSsg66zl7B<2I
zHh*pdW!6vOb9u{yHN2h;eu|(EGVwY%@yDCdv{@5;2suu{i#MJ@Wb+f1#&2bgVNVjO
z{;tOW{dH?7L_g&=Ke&k|BYDSgdg9aOj2mFGZnu?N0M36(L)^iZnZFP~2Hgdt)W--x
z*Aqe587xuczx^Mvs!#Me`z2r?&mZHCKjt09r1`M0|Jfa-+|{UZW+u%~X1UX2&Kfa=
zEepuQLnzS@Lqjig*#b>`q_>16Ir&eRbzXF0mC1H!^|=i^EgHN8VMOkDzf1Kzo~F|w
z<vL^^pHAaxkhST#27ReNw94!LmG!jv^L&zdm6i;#(kdfno)n$2o?l->xymw~wCKe8
zgP)+X$W4EBL&TK2f3BxX{f+XT&K@0WB%dv-b=CRIt90)iwU({($Y-AmPjSmWm9?-`
zM!Zy>d582;M;jR1zDvcPoU<Q>y=<(N9FWiUk<Qm52f&_0G~3!zuvFfX>CtXeyYy^q
z(XsnFY+s`T9?pL<Z$7We^tD#E+BoCnbjv?At%X<f%;(!0+thg^*`j!!`$VPt7+y>{
zK=kO9xec~p1DAEYxuRaHf&RA`=-a2L`XBz%e-&EK&yNn=<mutgt#qBD?mau8|K<71
zx#E+Qef{E<<5Jh_sP8`W<*?Di0qRtN%FkWswDeww!#D_UwX1%wG`7>xv3JWiV0_|R
z;d8YL7cVT%p#xdjt&5s^uwl=f55b))-={ywR)by)*qwbTTfk^-fN|hZo#9owrV9rH
z@&PB)DO2dt*OrTKDc<6nExgi8J@%24Jbh~4uthJA_~wf%f2`WZ!5zEnTqni!y%we~
zrZ<etwlhW~tC+q<^;o1AP1WVe45b^;P0~2#n#BhZv2UwATNTO&wOmc#*Nz&e)-#ts
zM0O5$OBua{JF2JC^Z1>e;W2d?tJswA)C9pX0I<K-x7dV@D2t3*Xw>WA^t4Kp1s-WO
zFC@zvO_ueiZ0O^~9-bibbsSe>{mS{>B)8FjAUe}X0w<F_Uk2I0T9kNH>~Jd~*#CS-
zx^LtlZ}h;mI+zaf7%RKof<#j}vHSy%Hin|2UT6&FcmK1zDRDd9H~{0-ajp|H(9Nch
z{*2}aRps5*;nWC~7GUB}0o>39J!Y6F<J(QA4rtGT<6%CY=CV3X^%d3}=M%fcJl?mK
z`^rgqZBLySA2DTFA9FW#F5TH#C7;hpG{+0W84DI=I{y&;Lx@$+55ZP!Jcf_e_KJIT
zuNNkIe9-YDV@-Xew$snktvu_T8)vO{Vd?@uV`Q5oL#%Wzr#_++*T*20<=5Jf7M)ms
z8CU7rt~IY=lWZLeE64I}&8s}N<yWDJPTZmkJM$^6I`{IT=eT>0YHhX@vd>$*Jp@0d
znLtlPuMhX|&@IfC9ZR%1+*(%0rrOsv-y!*W94=wu$9!I~#D81L=jyqpO;|&qJ$S9y
zmUdZ#eVsnSo;*io`&6$(?a7!+SlA+)GB36;pXKqngiGJ71MLny<7hlJ*lY1s`6V1b
zcV2}w&x9WRwd#uWk}dNZqLdp*-T&r0Vtf}qEc2*25Sn!5u{C!L5>{;k*kDkukkg_a
zJ61n9HtrTYY~b30bAkHl=zPtw1#=(X+=lzeA5(^2&}y5pcZ~fUTIx2lX8ii+Uo^{G
zx@yL(yDrOl_~v-$%7i0(6pLQRyw<H1vX19(FNTerY#Xn3Q~3F_CrU2H-@Q3s<XK{1
zW9h4>kDZM=zIt~th^+mV(jag^;)l!m?=}sq8NI=g9lgbljE4hy7dpMk6dCE(M7eel
zY_YRtSJs<#$`0_Mi;q0^@;wJ={MB-~p5Tm!7GtggykOu%{h`1Yu~AflV_IIm&T$;j
zvV|dpUC3cWf^vH)t3!iSvkdyd%s#rpI77&`3=64=-xoAX-rVAW`eysUP#N{l2b_#w
zV6ON8I(YIU`Ia1Xh1{;99V^at1GCO&Jxg|GUedXaby+Jvr@O^p)6}=T%}Pcz+Nr!<
zXb@}Z(<1u7R1&!taFJivO&dJYhnW1ulNqEbL#~!BDu#%17EQ$0vU6bGGa=jd6kjbS
z+f|oZaE%?psAae0V?LZp>^qN4QTEF*cPPZhr1ckDd1Ha*%>J~Y@BXnL>%GH*0W{5D
z-Y*SGXxhN`{fDA}GWh(TWiIWmY63XPjA2hB%E|@~JoXP}=Pin`2O@V4qs#5%Hz^*H
zmOysUH58Z}Fc!4AqzxJ9uA!6{F6ZYM6;I}bF~SI*i9k_mG*gaD&2!hkiQ}4#rEG8+
zV8(P?GAtFTCzIj<uVWTQ;e^+Kly~N0k|HTnnewGvRp2~8X^GbH-Rkl8x_oP%^O)9R
zYin#T*|t`9c6&=t4=2mDc<2c&Jm}a?oP)-z&YLJB4F-Rd>o~1NV8)I7FzBNBlX0~j
zj%Pw(a3MR$nLgQ&?OL0@$+>Q=jr;J)=ljSa(>gcb2a9$}&Dw<6s*ACvzEsO{e&FXc
zO-&<F{kYYqeGaETj_g-bwSR>tM(XW--(o`D&nSmh>nE==<N{}Ei*R);ghzogm4l@}
zI3hjrlA<1AN6>q9Mwj#i)`%k;gDy+Pxgui<LtM3t=rv!?OLNwX9DRZ^m7~|6Vz-#O
zw?f9HT()L*c>JFcBQ-?=+xtu+OP}yk(QFdEwM{JnaI=jq9?@FICY9xwezPXDxa#_^
z!TAt*>iH(uVNF|ejPv=)Zc3`#CHmQjnEkw^y+Nqwn-Miq89S%zzDHTcQXLfwH|Ohm
ziKi83-5N8Zfr&jCyM-Z|_=Ho}MSM%KO(R1P7VXa1=iGx~A{${$?1S8vgs_%ydSk35
zm)oqiMf{3&EW=-_Z#@^tMzF*u-Wp%|`}XQQ`U1Wq#&`HufAB$*9eHfUT#m=b`A|0K
z*e^YzgDVWt#G|-Jr#$x;_>`ZM5$>8;e%5C18krA`NB-Z(?tT2sSX4XJI@PbP^ktJ#
z-$kAPFZ~BB{dqy>2YziP{5Z4F|EKSsI|4l4U!3oLxR~D*kUH6}d>vD+4F;<4*VnOG
zqo-=1{pH2W9pClog|A=6(=dFyM|BY|wtiEdmZ6Nkd+{nZXI|Q-&t<`VS>3G7_j_6a
zI)Ad6albaywn;8;EaBfLckioc_i5u>N~Ng>U%@OIuhzp~rwGGFPp{JJT`@JUhd1S#
z&(*pa1WQ3D)k}PH<x&#|hLMvze*_}&iGv1%#^1zNwWJs=*3$`^cp3EigI#n#1VOOB
zd{>^oUI#K|7I<9{ynb*kv?!VWVpn~Tf&==*<|0O2mVzeM>>p%i%!o1>Kck?=h{=%?
zuQ=I}68xny_oMyb3r)cFfsHb?WmZppRi8}nrEB|oe9*CLOQ?Nc=QGyaNmb@knmT54
zT5T_0;yK-!ua_2nH2K(8vLVkG(6T=9#3|R+ex>%p2k??7tv~ukl#TsCcA|%aFmpa~
zl2tje^pmC>8Dc{Gdh*78RL`vj$ad9bt33R|*hf|}jKdnIH8#b6@|gZ$|JYd6<_vSD
zZ=9HN7V-(u@Ev^{+PJ}^`II>-<{9rp^c~|JHY`gXUH%V0@Mr(v*WGjQGj23*)62mj
zzzZ`QP;P_CQ;uE&k>_Euzar*k{$mVa4VyqkP7LCAvi)d7#?kmeeUTx`X>dys3Wm&(
zv`rO1rWqk0+b9Q^O`}YAEOa(P$V>~M=?@>Mj;e!i>SR#Yn=vYaS5U8Vqq3GR6$F!h
zpnyJwF(rz{i8oP?o@|ZnZ0FGwR+Z#V$~9p)KAZoP#x*Uwp<_ESbGrDmGnvczqgmkd
z@nJFMF;61HUyQLh4$*d;Prc*yp=G0$R|e`F%!=dqVF~G`Zz_-yex@m{7`&Q~GB9%(
zkux>YtupG>3Zs8orQ6O&P$7S`(XLW8Us2Mt@v78FY1e8j&#5iES&lS*lDuN*IbC0<
z*X86D>h=_!IOB-V=cvqN>CvL9AZ7F12w8Y)ck=u(mZ*!`Lu~<_<7`|z7()q+F_l3I
z`LKw7fIKw*rd|pF&gJ-{94!689cfY#NMQqifftCfReZ!8-|~4PG<ALHk`BrU(v;g$
zdV(W(u&WL=U(VO;$}>i2W97l*ynuYxp*EsC*GX*yjxfZl^8)Np8S<)I-Dk2s`09}R
zB4zTW-%w~8gj>tye6}Ogl?Fq9qRq!kh#{T1YCGk_c|AYPQrcAQ*sG!Djn+AK-M8{_
z4!rF59=_-;e3cRRu<|$)-x5~F-I~sJFox()8XKf*dvgC`T|xfN{SXxBjBU;@tXVu{
zo}i55PWd{n$1(OAY}bET9(}4yrq}801I;(^$X>-+u0_i-OKpZuVQ9X`*Q#%x7s1Uo
zich?h_X0Yolji0gA9VSOHjX(s=34au*+5wiqC-QMpHmhd%1FwunKxsQmLAgZ$p>>X
z@U9K`nLf*P8n4KfP1I&uqqT_p@kwu%zE1sFZ%)^KUSktuQSC;*pzrXfr^&!qJ4zQa
z>br_n2A2LrQ=IK#L*=uwnLz)l!=HbA;hQI1*APBAdtm$z2br2*`9FR4<Bkm&HaGCc
z3)@ubg-;EW0N<C79_@IInbrWH-Lo4Se|&vu8ym4fQ=9qKlXx0N7<m(lk$0E14=dEB
z$=Ra^zDaUnuWf%*Ug`e&<%R19gTVKD2-vTUoo{Rt3VzD)EoEBAsck}*O&jPuMMV6l
zd~1!eR)9x;DyRMm^5M5-GwkAg+>H8bd7A0M#v?bkf|Emyo6FXwhm9VLsulD=*6`&@
zws3^O#G`cP%QPJ@GA;8eo!@<2^VWIE<OxSSip4`*PwXn=0Iq|B>Wn<NLu_eIn0&K=
z&WTX@gO1&B>z`v&ZUGuR{#U+dK-$YJ=78d9AnhNx8fac%Kfw_(?D7-#v_g-bc_rEN
z3%v4&<Rs(yVy8};mUL1lY-dswPR8ftmB~{)zL;t<nt-(kDqGa2l$TGGC&nh}%JiJp
z2}-8*()EPi;}d;OcP=Ac*GCw0arWOF-?7TFo%n*b!Uh)8iB+fCu2x>W*rqXz55!Aa
zf3y*44z~1%EQ>sJ$$<rL4&=1ppXsf1josqu(NvFI-lFAl(yFKEhu9#D8;hr|zY#Oo
z5MboHDV8R9AGIX&gyyW;IL(Es0x=sf1j>SLPXaKnC~u9$+(wt9oW9D!8*NG+I#2ni
z9CLh5vjk5V!HYb+1lr!R!v+U+@jg$1r5STNX27paCvX*VIbwQ^o$E0}z!%Rjatr_r
zZBBvZb%Yon>45^-<%uccSjIPa-Z2Lh8^Bl4vCcQZC@(xJCw6-zQ$XMs4l!dK`w3xV
z*y}cl5pc8Bb>VPlzGnjbSv%@!pC~E6MX`)GYRqS&p<`+ycu;1|ydVYju@#Gp)!n`<
z**caS^qj7bgWO5ECeSy1JXsrfzfMB*)jaR!V>J47@uWV&IY7{?t>4+E4X0ZCGaTB0
z{-_&lpc7~|POgdX_!DGK_O@0ZcZz{TpNey_7_+7a4Ew|OeB6Lm2W*C(`QXvQoA@OQ
zHa^I0tupZv*LW(c{V96XoG~HFx>$bhao6*r+K#!eXJZd*D}AXds&kH&#!twT&p0{l
z^HecP=Y_d^t-b2AOZ@Ql+Aqs$Y|Gb_Nshk5pVLa;BTZYV?&KLC$csihl5b%$PNr=c
zUT_SP69%{h?864p!Paq-a{4xU^4NBIx;8<UKlr%T5=P((8L0^3SDqN1S2;(!S~6K)
zX<=l3kimByIKqN_m-&Q2tTxIt((-MdPosa*X86+d*q`h`h8BRXxy^lwb1by<%iplX
zvCvuLSd5J0PtPfWWBV_*@#o(hoDo-j=H{4IU4V=pID*TExliReRGfHm>l`TAn~2q?
z>UlQX!n`q8Cw!=<#<N^6o%>;rXANDp_juKYt#<0k^ypqkn=AGSG1kWMDZ4VJ`WN`D
zD=;VK0I%91`ZE;zDd+xS7h_fJnB_H}Gd5K|8P7%sZ74HF($F#=`Useok2!fvIK#q^
zsz!{C+_zL#L-go1pYqe$Ec8&;o3oe7giB1n!{?mK*|g?!TrNjP4$9KT^nG13WFAig
zW??EFZ4lQz)`yC-;#_g;Q!cpjXR9yQz^c!l**Ut7L+Gh>)xKGSy_QFMGk>i=m)%r)
zt4*{yeTP3iO}1ov-G}D*vprv>*OZm6HtCPQ_~fZ?>iqhv&m0mxjJ1|NJ>^2#FF*fe
z#|DfxSDxOx-;`<D$!pBszKKl?HaA{Bf9-$gFXCw**`ICr_WazfLf<E4<4WaUJbE}$
zbNTwhH<TbgF}ny49^D@{lE~|cA~wO;*kLp3{Q0^20^HYc-tJy}|H{hl@BZ-h_ulWJ
zv$@5e^VsGTfw~j^{QEz7Oj-V8Q;a~pm=d&U^W)`(ZzlcK>PFBW!TAg7y4$_^{s*V?
z^X22HBO7t&<}a{l5B-ss-{5bB=ikmh33TT_|KmTQ*8_<qz`}a#qEp&lV{<^|pv~a&
zgNHR82h**-t8wR?3NJxAYFYAhraXa)N3jE>pInbbJ~-LaCtwbyoZP4`qVE&yz^sW*
z7)v&Y?<QkxB+U8Y*!7^zb>dg|m$2qIzb(5t$&9BH;-rgpQ`u7c06TME#wxz2!^g#2
zffG7CQPC>D?KJvv-G`TKZuxjgrY>KD(c{TB_jqc3OZkim@3C-~^q#uBSC5)r=a=ki
z@k1jprds7*n0_*UB70J0ew?$$Va$@|gq95}znC#aI+0;K=L-ZM*Oc2`MJ%N2shZ&;
zC@FN_lCSNiUvP@5IYD;8E;(rE23xcOH$=i^E@FXQc`de6C(<0BxJZBZu0nc0-iZS!
zuV8y;H(=bgizY7K*?eowU4sED64cd9IB8J5mz0Q4@e==<@_n)BV>({U>1pAMrnr`0
zi=B_3CvKwJM9=r^Z!#Z|xJGZ?Z+kL5dNvmx!l6FKCSSOgQtLz9$;w>BJZ2eFa@WsG
z=mT1$VF8@+%=wBI=a>^*#w|0koTmU|!5n54&lGMV)du>&oXR1W2{wR&&&6U+LMa>L
z#J))HdW}X!UA*Y%7%DPz9Dk~t3zwe)={ao#K#Msk@R9Qy0D{YY$GS%Ca{Y=;g(vz+
z+|VE=I=(HMQV%v>{VAi=2F(2mF}nwav@}79TpW+`T#?7Ou;5iXV+)Tujdc-6P$esU
zSvQ=*bv~<{ZXVI`YeWmrjsBhH^RYx}VsN890x4*brrzP>$mhC45l-fT9(5%(O$TM$
zFFFWy{XK7*@B<6f#P)v(J=+F`nFJ~`mg>W`0pmESaTHxoh)>WtU;9WrtvJiI$`7TT
z8vhXcj%AzL{}5Y_DVyuHr7vU~G6r<|-MzCjAHRGj>HhuYn_|!%K{HEi90Xk&qO6`{
zG!~(wn{!n{%_o%5bJ*`%EbM=KIbPQzM_|rUz|IyPvW$0)Yf0pMJtqjat~<Ka_cSj+
z0}~A9T9rdrU}H{C20#-Ja@awP&R!j?coHNxA+E@rQf^=D6=qLZ&_huBMef7+m5&r3
z%WCl6l&3&``@7#A1UqNjT-iZN8^S$ZJ^Hy_b3eQ@UH5xw-@?hSNo%MXNIccU(>!ea
z{Aim`pOnp;tdBHL@zD3!a3TMLy*iF3ph!QrGUx>OlqOF(0a?A`4P9(teep2ldClDY
z-PL&>S>#%n!qZb*#?}n~@&EpQ_w#@Fm2YZ+_kaJ}zc}AN|NFo0{*P~ezx&T${Qd4M
z8wv2%&^w9RigO6RoCO9akNJ*^jGOr~U1^oG2oxubb`n|^qgg;=;fIABP1?wD9~>8i
zch9bRFZ)&p1ZcX$AnqUyu6SfC(%_+AH1Y&!enA(;Mc#@yiPeRXY}W;v4rqZePNL$w
z$Hf)cu3_6xPcCD%csOYK9VvDIT)Q|Q{M3~T#PE+r9$cUK#CdWR4Bszki@#d7h#h~>
zwmGcXJeb4GSS>#4p#{&(o735snZ9qDi$GzjZ_Ht}?5_D+JkpVQ<zM&p!pUvENEe1q
zP|=qIx;PnY%|&x<*s{Adf57L9*Otg9b=iik=~}lg+G{#`ec`#GpylHR+wAzpZ)JG2
zM4i^sgPpP&rWTJ2cYZYg)Z@Qil-P>fQMZZgZEcF_gQOXw%-PzcVBt^;2f4jS6R?>y
zGO8aq&^6w@{y4A*F|SD%WjqN)o?}_oAs+Qp?#`tB(4#pLP}-G#gRQb5$|DaB^N;kH
zX>mL;8VS`w%3~qX%538U>`NY(FfMAv@d-V4aN51MjSuH#0hMx$^E8q1_!P_Q17fi<
z^K#4r#zid7LA3XGqHdH?Kk7hGId#7@8;3<u%8bw3(sVk$<+17AyI54^^4QakM7tH`
zZnusJl|hA<O&$XE(EPSEuY!bsdja!~8Rq6d^`o9XnWA{fl$}*nRR0_9K@?D8R9Z@;
zR63+XLPE-)NSBm!cMM1?F$2;$G$P$Fbj^@MGjzkyF!aCxhyOZjo%?gMulB{h+3)we
z-sgEf&jo{eVMFDXeFEo!D;!Z1n@rymjSI`eDm1}j-N!U)S2O}gMZSrO^gXv$L?A^j
z2OWx|ysh3#S2ItjrzW))O^$UL=jq%f`)d>`Ce!grTwRQIw0u1F8HLR72cJ_w0=ayC
z5IIKU+F!bXzWw&w%(yE*Cpgkz6Q&aM`fcT~{1@}>Xd07qx)ehMi;a3_oaQ=wngeOl
zd$+d2k<XMiK)hk<Pm=JN4zkhVl5JMCmiE#Igsvrm-g;|e2d@4|dPU<m%zf0I{Js7U
z6@yE6FA(XiG=&1wvB3(Thz9SInnb?Ok8|Q=qHwpcVLn6oyz!3bBxWE_F8wqvUP(<-
zo)JMSweyA82sf=VySQAvq=N+@I&wb?%}65c6(jhsjY1_NndCa`mCK&k;t0XMq0&V|
zrJrDV?@?lXkRr-tSF<jooZTXEz%3w(UQb;Y{>w(rf9XUAHsX-F<Dn0qTQITBGT+dM
z2_&ZNOsNrQcn8esO}7!{x^X|~5GC03r<xv7v^RX<IaqLpf0_BiH7-J>RwbyI0XI8R
zP!?fb6dOr>Q-2%l6PjpKCU`zBl5glbGbBol{FHi_deY4&Idb}k0^qaTRW9uI5vWdR
za#4j%04BhVro-&%nr+JxqTK)meK|!Q5)r}?Li$_O6SEhb1evimGELVUS<$TAV?E5*
zd}gLrrJCFQS4FCWp1ms2Z-Lj(NebH7um51BHd}=D-TY*?cDVotZ6_#3lfwXXKiN^U
zA#=lqz9$x){t$z|p1mR;LtaES4W32?O`pbXU;X@;#qJnjslkGFf1dsU#j|`xrdWYk
zzTP6z;6gvvxYGs^%>&Z8?f*Q~dipXXPiQ%ubT9)y?|pEWYqLM;xSTVtD<sLbvg{_1
zl7;oB^F!v@zyo#t%*)3t&o$Iv{H+iI-5n<%dA8qSP<TIJZa;2bgrD3WH@7bX%_{=^
zup({3++utWda*RPKU~aXHR7peOy^^$Rse2*8(RO??#e%Vwr8jdS1lX2FZU0nEAJOw
z!=*~+?3QS+j2?m{Yee@8#M>*d^efr7<Szbi6xxmN*9G01INyFJx5)cVB}K|D*3nTQ
zP4(L8@!88qz47S*(|sO3&am)~Cm&ltPvEbAF^1<u$O*aqX`1|<59iEEPpWu-Z9-H_
zCOG+Bf-^%Ph8DpE>Cp6^|0-ehDi<1K!u5|+McCQ0&KupVgZLwq_3q??tCaofHvIn+
z{m$ObWvmVp?Ht0D;;shO_2YS}ctwFEC5^9;p>msJV<-Ku79|<#sIm}CKlYcxI}G1z
znwj=*nF%0-1yj5~Q)`+-zR}!PWf5(_CA@M7>FPe5iu;&y;8QU?!6vD?DE_!<tZaO#
zibLy4<Hqq9n$v;OXL}~S5(<(TNojl}sc-+O^bc)?Q=<s`tJ)cL`=-u^*y|RX&Ytg|
z+wU~#H*7O`Twg4o{He>Zi-IVX7#DZMm&?lIRyO6;Ca02+hH&*W^aF7ooKb8%TwDxO
zQ}GX)6ni^Gl&BoPXd(SXrL$&B1xubt;OTk@`m5DOm`sUm|7|Dyzds|&Aogw%#1qB|
zGQh^zp9!+U)abb)co!yaE(ea+T4b<pHcZ`zx|1|IgOn28xV5qCOvIgj%7<#wO5U&(
zX;GQxB%ciBRf~~I!`O-jY(=NKso0{GG5dkmb_l%>cZSLD)_ni8)7jp>4?)z?pv+bT
z4OymbBvL2fZ+8BZNQy3A<^Axd1&F24WqWJVCw$Jw$d|}Us4?x^`O<h;9Az?K%ei47
z2$U|XFD`1RXNH7KT1KFXC>!i;$3|&uuB~Mx*<gRpi?hnl2_xLr;y$=dzG5gH`!ect
zuu55fGh54wv>Edq6gymZcBrcP-&~N?x0?L=aic#@uv)A5&Dvx}`3&2v`I%0gj3m3&
zC%c1weuD}7lPkLP$ERpkNJZyY<l<>TO+01bRvv@s!|CFcu|BKj{`_;_P7D6OlUIs+
zq<u~H__3o;*JMmO5pd?&Uk>?+$u<L!j?$<3TUkkMln!rN_?ho3EZVh&FzMzt#(AYN
z&rQ^HvYghQ>r0i`7~(lRwLf`&H{T$we{aN7qQ+nmXGXS_T#weAm-8vyv7xqEo3d>2
z1W8nU@w2Qjm=OLp0-5KS9{4@02eqM-z{sbwV49`34QWR?47|64V{JChcZJRZF4>F+
z@mr}jXs7aE=J8hV2$@eeo$f54;}fVHc{RSa(i?(@>R~3;*`#=>hol87h0Y&4Oj><v
zFy(FJpU9u;PfYTvhZ>bE>Bp%>zVO=Yek7Exb-~CGc4}B2v=9C)_|!=4?2_ya>mCV6
zPC0+aC1^$k72QUIcT!2!X`kCi_06No#=?5To<n1Y>!BYpoeXFzUv*}U(Y&H#)K5uk
zm%$;rt1kl+pPu<gN4;*Ve6}nbBNjVQUpqIl_w8y-apki0_PL@#;uF<ZVfSLy-7U{M
zX4q>faeSpJ_C5%Anyc*fNZu4%@s!m%0ew?a+bG-1c?g#L(FM@A2r+$|qkv+wzW#{c
z5-+A6bEPM{2e}QAWd7z;9QrF2AzM!Wh6_U19S&XWD7{h3l&(emIXZvh|7TAD5=Nze
z*xuFOlKUUe_HGRwyPmEogIB;Hs*NN1EfG1rL5>Y6=uZl5XdZ_5FkoH8+}9!C+?A}u
z6CLdqX(XR2In$3;(oN_F$eb7xte;}YIoHuE5*w*bqjv4&vyA*}&E1+(5}23qZBa1M
zQ1n=1G1;{ODI8zN5zc2X2z<ytiq=v~zfosqTaBQz=M5M%e*?!yb*ayq`zPR4tW+ep
z+0V^@wz&5nnKE|8uXH?l3-V|ojRe2L_wJ)KCy@T4kUo)Q%tM^2KRz@F4p?X)Q*UTb
z3GF?+d8$r5=t6k;m0nsXS*#JPlICjKQS`cLiTVX(pci0tOJexQliaO!m}Na|Q_+Kk
z=%{DOdZ!by^4m9FfpSR9Lc9e+LAM=ObOdi2uSDBe1_W8tAGAuMJ#6%87!%eTy#J7>
zW!%GLYtFt4IFPk!)j{;bczPJ$=79?{;xuozGal(Z8SB9_5=7GpARIUix1JBbn@Kuf
z2W)C0|7@Erj3<ZX>j+{L#By%@KSt@&+G+-0Z1`8h*fmP{6sm-;zt>9Yc1u`Ebec*K
z(g;*&Mj<;BrM`<18qzvDs`uSK@;!@P3Ca7_xRk<fSbA2hWW_jWyjudz3z1UFHk4_N
za~<#dLR@mH(swdxkb7bolwo+(`Op$DFov`x!%XN(^l|S-Np9k@Quq<qgV{wgoLUr~
zXHG}H3jK?vYC=<S&oxuLW&RtW+2g`ZRE|zI7dVxfE?)oO@h@Z`dqUgzs4MAo3<Yx`
zHo4;9)N21i8(ee|>}4<d85ZlXmPV{Pxo$A7$G!d+lG2Z0rgOrJX9eJI+5#$j6z2x_
z*%7@>3S>Mk&#oru#Eias_QE!v1=EB%0ta%((q_VQz?io*fxs-D`bO_}Ul*&86{&e|
zTnu>wV8xDhy_e5)1-9f{v#8VE+UNgi*%4$6*6-p<tv;s@s$6y&EaXy}mG)^{2xeN+
zZ$k{~oVg7r_Q@hevNgyA5$|%^m6Q0<f7q@Xgkxk-pInUh*yr4jsT(_Atpkj&i*83e
z$&*&EvRf{_CiF*!6G}j3btLwO=S2OM#5;TaHf`povj@o^BSjt8<^1#4qx_#bUXenc
znEZk*NL-H?np2NQfJ5WL>yp>gWe}rJxBE>S1A-VF*<U?6!N$``F>8DbWefFG>Ra1>
zHyrrntM9v5Vsl254bd1^>!y2)n|JJ!wSNKap=;(N_zv66wI-sZ0*s0V)a_$cah}AF
z2m8Mi)dp7LF>VDAu)qDKLRDKQQ*DR2+#K}%re!@fY8;A67YUUvD6(c@G2E06<h@;r
z*=|jVq0Dl{Jn;7tl6GA1jXbVz1F^`(<@;_YwxK$%_o3X9b16B*Vo)u7E~nuCiSzs3
zth7PMycb#Xn$jcoa@$tD>=k7hMLsFwS3eLnH8*#Wf>!mAn`wfkAINwKi4XGMHJ`RV
z9_W|~!8zxc&H|qF`7q9jPOYElJ@rNqr`vt<m%Zw7Z}+$3an1|klDOTkl6H}oF4}*%
zVw%j-U3I|V4a*YCi+<ld4Xy{Rz@8QM2ccELpoojs@5TmjkL3LyA^U3tH(Mj95jyev
z<AFv2KZK7<UeN~{5`qZhJxjHBtw}&Gh`lBOkyaY-bgq)EpnJ;pGhU@M@9^rl*3>Sc
z53Wf;u*YGEbi^m}v}D$*Z9ZL3Q~)#`_e+G;1e@QlyW+<5RP&=l*gVNB2yCnoPSTjQ
zkcVykvPY6`N;#ZrBlu-}J)h5XfNtDrSnJI54V;*z#*P@gXOFpqANtQXB)*zqH~Ozl
zisP+Uc{uVHEK3nT;`d|zc(iwgiWec3E7l|NW_3jiPR#mW{=u@e*_Swna%7qJ&<GTv
zDt8)0dErg1LQh|Qqv~zq7Ud!~$xzAuqbcf=bQ*Tvqv4~1f9*hWt{lBTR6gvX?E`WX
zwUJE%4A=zP(fVGe*6q4{{jfSld~g{iqF0b1`X`K_T~{Da3y>N6jwf&1Z8YV`pHu@D
zlDId#Xluxut>|uDRZRbDd-YNHlMj{3Zn$i%eK~PaOdEXyRM&>zeXa;svQ5}uNo>CK
z*Zz_6cvCXu=M&@BEjZ<~f~kSe=tGdxp{@~0x5oA$2_d(gVEgNnF#4?@-EV`-itNAO
z;-?GAb4O`causH*ry1bWs&e+_MgP8+toX0SHLqrc47d3qIQFxphNc}aoLl;B^QEc|
zIzm1}<1^Enh{K$zo_EkhT_Yo~9*Xw+<YF+gr;78|aRK`*pn1kpbLh;E;XcS5%YEP>
zoVK~P)`wdbvqt^-{;ig?QT`Zw_@pl+Y(&Y<k(f#ZI<dk;kePi-HY^vJM{Za%!oO|0
zjwgC8a0|ZMa;$b;3Ffh2(I95?w;$?+Gg9Xou!nHra2|*;(d6As5OFyOo@uH}t{dhZ
zi_@oCklEZZr37E{?P<JeHeihi{`!*>Ww{!1ToTtJ`8`O>vO_Kp)FUJ)IV$ux@H`1q
zLZ#YMnJZaH^eTs4AwXl!ozh1{iL7JLo!v9Qn&>=ih~NCA6RwYKVEt}Fnd69Ut?B>S
z6q?b{`B;$UTl+rznC$c)OiiZYQH=BhA*4?u?yd5EkcMZD+^WMYd0SrF!INsK>jD@~
zrb1hv15V6@{WW>Zv++kZ`N7556jJ?3jdh;A=RcxxCoCU*{a^;ChT&hiLSl{b4y)*1
zNs)oEc$paGC||@6>6rr0;P7(y#uxJxv=&zNGDA(Og6J26LWwU*=%n9D{-!<Sgz(wX
zH*va$1*RCJI#p7&HWkIgP*;UXfap;WCvN5t)6dqr-Y3RdkE<aSIF3O+T$dr%noOY*
zQ66791~-{FBwe^!^y!!fM0y^Rkt<wHaN?FmQ_zc>XSlo;$Sjx{>5mM+gihdqZf6eR
zI2~$uZQa7riu<g65y$wn?JilLM6r?(F=<ud=UvDf4#dISLxusa!<AJw=NC}wvWv;%
zt~ljduy2QT({WRu;n*E=cziOx{j?q;u}CI48XKZad%XmC7PEx!^j%S!Wh6$%MfLp$
z{#OGY+%Hj0W6QVxlc6oBF2%RLon|s$Jih>{$jJsoexX;ej?R9L6F#PIe=nP*t+#P4
z=k-{<GC%q6Q0JuC*^JX{jJo3!embG?h?&8cH6##~NVA!jHRkA;!q?kW9YkZ`%~UB0
z_!M{?(>H6Z>P%rw*W%Jjj=jp7B{ldfuG<%{gA?+RI0}KzZi?;|Q$zq0TrFOb=PT>J
zuRvEjSCneqLl^&=IOxPEHkA)5nRDdSs@FzoDzNZ=GlnCeYl@zRRX$<%V9VM{xQjn@
zj&iX{+F|VMc&A=1<HL+^=FBT&?~P{Pcq{e05$U!{WVhr&wV`pE*Q;jot7tX4U@RP;
zM8@;4M0<HzdA#usJ}mwX0N;P*Bf({J11k15ZJKN|B8K&e_tWSS_zzm)`GKTpfLeuh
z2&9^zKB(JHqTL5GFpYC~$RJUj+^R>=ZdHS*k%*(N2u%NLDE&s7%3?wfn0%Wt>cAB*
zPz8S4+V+u9jHI#*U&QwvZw_2W$h$^fUBg)w2AEUW_*w8%j7^fkZj#<Q0pum}sRr}A
zGP6n5v$V_&6Dr*GRLF+#B_|i-f06hyUwC*EiwtGjXxIx^WTlo$BroKf{uAlBiedDW
za(fQ89&pR{chVu6GzxHangQ@panH!kX*6bsn2@|&eXo6=%rK_$0!hH){@tm9tJX!p
zN~OLzLLgDmJSi=SAuV0bN{4PZV?!Mg968Qf!p08k|FUC|vC$kmsc5PtuA<66usq#3
zWmP|ycAGY_!b=M0YawZZrNtv!M2!Y{c^L8#9}biJhfqwbQK%#f?$vxe6HNH$qL;gF
zXBCD3=jS;Y0nb1@->8}0Xx+WDU4$7=lenR8t<aAcj0nXlOnDPhKCM{QM7VA#Md3@%
zsAT$AyIk<=n6Y5PyMIG@?h+*SN1y6@2W`DJrtTN7uYMEASH9d{6Y~s3<YV&7Lay7k
z0{kmP7RRK8rAy8{ReH%VnvnJ{4+UwrVb&neu3cM=JfT6;)gmJd1Q02Q?i|lQ*}G{d
z2rf*0!W!Wb@+@hbfy(;L`%M>n#-NiWce}%!YmlM3e5UYiOGAWVP>UU7N{?6fljN&m
zDL`a=>7Rf&<-*E8-<H!yoxg6qyQA|alaFHi@7%I?o4Kzit=#;-|5{m1bpIa<pym5&
zx>#ts4bX;&6t2p07z@=wP(2hFv>x;68JDDbH_@=!6RoY7XJ?YTJ8_uwHMsP>76|Dq
zSiTa`fQ@Urw@-^cF?W*Hs3#`*Vtqm~Dw_F?q~DO>*T)Udmk!%*&tEa6=bimg4bS1y
zcSyd~zYOARXFE9kzROz+?<pdQH6)l)vEQfypZ<qZX?~<fdS@!3JhNs9QTC~N5naV4
zc*%mJ0n)aVz^kB%kZ?fYzt$X=uW+$=H=cJ#Z7|9|f}=myD;K?-8oPbhdiP81P#`Lf
zDHHEGxqM*EOZxUcCQSlsI<MVRAr#s7m%-9E&zEau>zR_H#OJ5pQ>kGGI4Cozm-(D>
zG#i^3qo`)r-gJ6r46;Ab!uKb2325O3iYfk7b38Wxo@X*8)>pE8tR@MsgzS&(qx2c>
zB)Ion`(<?kUk=a1NXiR7a&<=Sx}LQ1QMv{^`>wo`?-cVkZonh-A)t^<w?S)NEuOr5
z;a&o<Om9k6LcLF2`V79aO9a(kU)ys2p46A#a`|Rndt%^ha59?2yMcF3fAY$0*RD8^
z#CALIvE$n7O&}!U7B%UXONAf&lgr1+Rf-ze-oTdaR5;}TC_VG$l-#xrUAWZZe6e<8
z#b=_*8^g`eKZ#x``u3W}p3?C>D1EBSMm@>y7#I0+ayCi}M@QWtNi+_Vg8AIW%stE-
zecMUN;c7teC?yZ@AColyGD;Nx1*TWyI9>-|3@7HbN6Z^`uEfiYt!<y7i(@}`X9FM!
z6f5V;P1p=VSxhJNHZ^`hDrJD!Z{8Lmb(1j$lv^h}jFM@U$MjUPdml{n9N&E7FB?w&
zaPf>IP;N_OS$Y|W*;!u%&i7fHR`#DBU@*tuK)k<v__tJU3S%EVC~4?!jta6JVJB`c
z>A-7?4u@40ZI`W1z_ZaBlb)ihs1v~ir)s6Yk*fzl9wm{GU&ERwr;PU}2Z!a0_T4c8
zPCVSL0kcUmh4CxOZ~d~0XwDq6#qW0*?|<+8R%T6qiB!$c&6wm?i>=c$p!+ts^lgc!
zR#lY!rH)t+gTgT8d$~P8iy*4B4ldkmAH)^-;AU|3_v36YAM+yAxcH=QHAG4ODyXen
zR8>sa)%VV(Uo(r3|DJDxO6WJ5X~6spZ9eBgTY0k9q3<`D6yLkXQhFi}{X-63;~*vp
z_D;Ummva|E5a#vPY|1;37KL%Wx|@C!V{Nw%^eFGW@4H;}{wW$0^-Wqz?<gQ}`Fct`
zeKlN-rG$tI(k5zA#Vuc_x1lu^tu#`dO~SF47pEh-wu$)U-rE)mez)Z>#N>VbQ6o|`
zVWb{Ls}ygN^>5c9EAKykQTh~f8kE(L^jc!DzQUPysmou?V@_VTKJQ}{=4+2T^7`)v
zeXa`xirDWf?a^Ee?!S4~w-=b^ACuyyZREsFOQ@$_$qS6LSTw)u<2UOe$x%$O$*@!+
z%NKS<b4a1ju2`4@HYt$qCs$;UDa${_<`}$yM;`eP*r!;;t*pbd;cp)Eyw6ESu3~dq
zc^sm-!@OJbNnQi0i8ZF{-dp-azoPYB;H^o&j%D4brA!Z(GMBisD9|#ls3*>d4daoW
zT)HX3!=E~JT<!yh`WFRAb?E9P(d@r%dlcsS1e4L?t!y{+JGs7_W1IjS>>bKV7sJOA
zrD*i{Zh=Z)<hZf9?AR7U181rpXXe^?+E>gl!_w}PbNfrf$84*J@3EQEGvFB`xpOgH
z+*bX60jjIQD0=0n;wYl5p5LEz_)63Vj>t4WVRAP^fV>qd3PeqA*WaYQU7_sj@XM3W
zNetj^_;GusYYPl9q~Xh$`YJE^{sDb>-sE`wRa$hiUF9lbvZ1~Vaan2CmT^?zryDVj
z25(_0r}?lt<{M)5?sw!i!dj>%bB6=e0+r*y(p;ff;N5jy{Qoq>sZVlu?RV(CMW=xC
zso3`N$5PDBh)SnX4&+6mBIvY?>cX}b8#>J7ZD)z$9!b+#dnkrjFuv^6l(@MngZbXx
zxn2%4-a&7ModQ;W+$RU#P=W6;%l96C!x#4`Wzb(-yPxjs`Q@gvPt9wCKR6oxvs$cF
zNpX`(;V|fu{Djf>TO-ZKh(q-+xG?psZiHn>ZAss0R+Ejpj6vmYJWAw^;d3$;c$oSg
zC}SnkWMiPNoXjGbm;b10!plMCOzdpmuED>m0?XujbEb)@ZczHt;4vs=`qmLWDTv+*
z_;z}B5@HYT+(u7Eirhp;)Kc9k1VDaTMr@aRFa&%2vUh;Pu~`h2ZTUr!{_${Rl-to)
z6QQ*DJ^sNO=b@id#78#OpR{S^E@X(fh}&pAeb>(aa=yGV3b)*(5In?Q6s0RzsDx2&
zx6NvjvQ$T6LW|z{llA>l7Z%W+oILrFbNz^I)0DN{lt^$`g)fUJp5n@^-VKz_E7AFr
zMqiaA-}xyw4j}pGju1H=DDl~6)1tu+O?w3dL|5R4sS0QY6S{f*$U(OSpJ)-b?xjFs
zS{d8x>3^iLK?B~^oiUS{;}wzGLj+<JCBAR|<oT;8wb%Lniku^6=jzE&%gR$mq2_}p
zbjG^{&Gunhk&^!$UVyhZw7^_$?_9;E`#mq|74%{?$o&G7zH(fOLE-R#mD|pN$?(a<
z>#*UAe}lF;^<Dwz2=10jOBEa<gB=5xLkeilf1X?QKdH~-N3^7~gmN>Nb-X?g;`9Ec
z0@!-b9ar-0;_HljpOg5qLFJ16XgH%#)tSDS5G}GTgDX~<tY*Let+_LEVk0hG!Dvn@
zzg0oXVr5-R$d~X*>!04o`|XjUb5HSQ>fa&+$Af4*l7MSkhJ&%~uX4*m7(A^5UKa@s
zZ{B@|jsMU=KbWugT-0{X*#HGi+?t2yO(jc4Jbn`y>9>DLMa{h>y?%X&`lnI6KZBNX
zs0K&xBRt5s;1S=gGPJJ~X1IVi<4OMC*1JYBV1~7V>VS(*?R8sotCKnh598drryrJC
zASvwbLG+@=C-+}5uiR{-3w&9rN`vlSdnfJsM}MxxQlcot4o}_Byvq&th7Lkoxmxtu
zPz@(mv*v}%$xZ`9rYDOuZ3Nwzm8vej*d!q<?t`$Bi<ZIC{>k4?hlXCdHT@xKb&#uR
zQd7vCrnzLbaL9t1hLouDxm)34SL9u%W%AMN&C0*Cv6XgH!2UdB-BPvz>2fo@cQL)4
zXIXUIfq{cupEU&ZNCXie4%Bm*GJGF#LqWv~BYb~JySSI3<UVek8Fj&SSIFbiDdf6J
z*{nIuwt@GHm+fn~+kL-31n76ySWNTrCb>^;8UT@NPGh!`2B%y@ln$g`9QSnEC7kOV
zlF0D)&E@M=_J0mMjL+yQR{OEY%p^dC!BNh&W3vgMZ28@W^z}Zu@$IjsDLlrTRc$bE
z<NlrBiku|=<y?Nirx|AfqnD#>U68h0m&-vOsaeT0WbO3)W9976)qH2gz!C))t$n>6
z>%^&AqWQ`u6Nj1C_Y1<=45~Bcn(y1<yV(F0w#@#v74I}h<3IKU#Aj+9<Sg=@16Nx|
z=DDzmb6%dyftR+IXA_ETv)ma}Y?oL#vUy~G%C&_uqeIQiUIINQ;<(a>yXXxtsHT6Y
z>KQ@PkHP`#TKoRXs2Itr?y?_F(WJz}s^}v>zw?Fvd8dky^Fh5O$oyyi16jbk9!U?=
z6CZ)!>_Kg}hXoBq_BY+gWfx>-b^O)6@9p_vd*IpBA>#g6`gUZT&-5w<vmFFxxcdKo
zG%^5&TA>H`H@A$Q5bO!ZvY^OhT1!>;7EU@K(qN`yAJhTA&Pw5{9P8S8f?LXNWi`z1
zTKnI6w1BjI+HsnIY`vUKFV8TD{YS(Jm(l>?<nLG?H?hxo8G`Au_Bg|6Le1))HTJvm
zYswdbvP=7M6|-2g)Avp%kG-(ZbZQrV_<GFW-yPi-QJrRH)ifcPV*&5V>5rLamuk-C
zDkhLTCI@Z4CyhF-5c1muhpF!Ao;eP9=lh#e0FP(;)!9xgmE>vdE0)u`0kDk0IC_rl
zL~-Y*D{&@`xpLsvB0JC)J6HID;On9AiXVOht1}Xg3<S2_OeI4+)x^RXJFnswMaQmn
znIc6irR(tgcr4oKyR)*g>dy8mO&RGik1ZUGq*&UP@8N-qx3|*RAD09-*=w6oD513D
zacV4|&ytMAYS!Y8+=}08r{$J<a6ib4T_(^8cnCUg@=B(%xSjfR2L&8nttvANO5E*6
zEzS=%1fX7^#r?`&9v{Cz?3d@NG~Z?crlh>jzI)xB_+M_sUwuJE9hn>)N<-WIzDwl0
zHJW=#OZr?`!;QW?XwZk;2p&(!9Q_w6PsK^h9uXXi9c0f^JdD}#KABNxQp1ETk`(C2
zOQR(1DYGPg#k(Ea2FpkQ4vhZm+y0O@0~(U6nKdbhEzC=HS|VkHx3<(G%~OHk{ql$N
z&#N^T*XVmzI>Ww|6mywt?{VDLiwAxcH}_%N+98+qjQsP1A?uV!t!>RjVi(d?7i&|G
zcm}<`cN9rMbkld#KWv#EHsU3(copG(4-((n1;^PAIk#wb;P3ZSeVfeb5TV5a4z387
zQZcyp?qpZT<?VBQkF)m{$HzBS1*D-W>YRps{f>!}RXX`oUi-fp4cd8*Y7~T*<E~RP
z$@q=ieC`HrSiD+YT)K_C0f+Mu%JCCu4n2Q{<?ibGP-__}Z%5$uj+IN+1Epmx=G^Eo
z+b(P_kShyl)5SI&wxG>vAxbJ)>4}m}#(sis8-+v=-=+&y;8`xWNi73^xCOV?I=&6(
zhWufB%h*)XZg<6hs1;hNX&x$7ApX{%CAD6`YpWe`hUD3wL+wo2L0C;5;3ZVwdh)LC
z(Id-;0k=^+@=|Y&-44AUf*N6)(!GIK*M8C$Xcv_6^NN8}(F|)3!@)0iw_!1<H|vVI
ztlbesftXnVOJVr7St&n$5^Vm*F+eqY^Ab?}49aF=cdC6SYPi@b4R?#T^q5D)bPttO
zxv0*I-XFW2nFnVPa@?%RzXv_?oABgM61iOWT~+Zbn3{7ld=EYdEx>pY{dk>qk!(lm
zP)_&n9fV`xe?K!OE%d}*5%$kA*?5*MVD2juYlP6t0|Gz&t>2k@96YIXNJ~-3!~=?@
z{(Xh(8l~Fid*-b)PB4k}#%9IPpy<!*+{&#r9iDxEOwEp{l-2pO_1V_99BW+or9Z{r
zM`hGeCg3(}eCHPxT!GIeZyd<fKCKn&3BAI#+*QkhBH(`q-tk^msoWl=T>Pzg+FZXU
z(_#UH%W9lKyt>z;jz4Er7ipODKTk5da!vCumorijWjcLaa6cGk8)b5F?oNmf0z%BV
ze=c@8_1D5~|J+d`DCdgN4KvkY4Hq_<%fWR$gR-j=)x0D}f65KBCiFDG0Y&~)OsDT9
z8~X7y90xO!upV%vX}6Tvmik%wyyDIF-gzd2HgN2348*U6788ICEA@%5aXVUlQPbc3
zW5lQ#Z+FSu&;4fA9DL-gI9X%W+^irT=yDjf2Jlf`km>^>x9<D^2u}b+>MV*@$Yo{&
zvwK?9Hs*S9+~$-}`RneqBCCn>TJJgg2n>L>N(!``Ng-eECOVyY{)j9<2-vZDG#gKF
zcW5W{&0lh9V@Bz_%iV7vONO+NxA&O@z)Z-71&<-HMUFP0<>wHsLB0<0B0)Mu^_9l=
z_y{#7<R1C_TmB|^cuc@JX3SUjE%R$R(X0G1<(~>)viQB_FJ*7Wswkx#<XfOE?!biA
zb}G+K^zLF9&&+Rt#1r+a*-<lG$Z>DCt{^!PJBAK2Q@aYozssJY`gZOe3La+lSD>xx
z?0E7laDgUVgBr}4BEfrqE`(!IWM)j&(Mx$^n6TZo=VBv(##}}v6r0&w#)ktUUP{-h
zeYD0(`dZx1Ti>;kF-sIU-ON-Dcbxz;f8P&zK5`gftgw4Mn!MAURo~Xd23XId8kMA)
zWQY{tLFq5p{Bk1XVnEo9nrhL?f_Wx9q|RXepBOc14lo1PcVC~kUqx~x(whyLF>Mr_
zFVp>LatvH@9`YAeGkKGpE}dy&X~56jBH96&PpJ}wLo8R<HBjwy3qumZiESoA&33m<
zj<bqu@r;S=(+ci|n`aX|*XeYr8aqkGi3hZhM*{O}=^MXaO9!AgH3Na4-$VaIBa5Rg
zENF~&OU(OI9-2%|O@AzG{0D4p6*Jdd7d5XP-!=<y+_6VTZ9x;2`P|E!q<SiQ32Z~;
z_Ix<$_D9zi_0!!0#@CY_BJJ8#1Ccusbwkts(@e9JYL=~yj`JB6XjdB4M>y#@=K2-(
z!+KLW{M7<$)_zplrS5h8$}gQN`?MNO3FHemN4dUb$5mLW1pBp51f;4*lnBLVw=ccr
zQ3jG;;7xL9CS^1#BApW0E4@+Y)`Xq|88)gpcbYeSKfCvKF@;a(T4ibb#;5`vzphKv
zVC*E)!C$ZE=q?QPU=LMy672o%1vfe}Yt}t>_>asanf=bf|Ha2Q63*EIyjY~}w-l34
zn?(<D+hAYXKyG2j<_9V#bHC#``9ClU;y)N(3qPiEx5@9J|6!qP+D4>Pi(M`K*Y*3w
zdoyWQE5$o1>@oT{W;v(=`aWIGWMW;Yz{~U##auLipv_oUGa}1A3Iya*oz)+-caW}t
z1eL~#Zv{c?Z)SW*U!QLsy`l$WGfPi1$}ax0+A?hJg=lj%I7u}yLR}e#Br5uon#csv
zlNT9@Ph_|2K)%G&)eT(W)*@T*yf$*Lwg~BE(LFoS7?ND%emMJRbT;0qx#kAE500TS
zD6MV3JX)UB6!8Q(n+DeZ_4Q%Um}seKx&%4kv^J20iP!%g))e<*lsIAcT@m-ge)k%o
zY@O|EvP;MFvhTerKc5v4BIoVUUhom-4MdHY+^}_rIous5j1c)f=y0rzGqTJY!Ih~e
z;{%b@xoy_ivwmRUt1yyOxF;Eo`}GSOYmNP41`tg}7UX7v-=8;AT1bXaLWY^~MSG<e
zgC@dE3Z>1t6kcYTvrNg9YmYq^6PX#%Wam7-G0|JuO2I|kRv!=MiuxKSLt|0k4$Tj3
zypE3~BxTe1&4eP#L9SG8{;x;ZHOTeSVhEn>e*CG>`pz{Vk58Dn+a;Mj^mtj*4|OUz
zw>Xf)L7rmj79|#i4;HM}>qZx6U81;ITvHqVLo=Sf3A9&wZWk9EwOmKU?k_^V{FnPE
zQqElhbv4XL3;Y+dn_=G55u^c)dp#Wg-_;|Zvlr_qDktuRBQ)vBHZ5TILb1NTr;s2}
z;^1Wb(7GD~cIb}xbd{X8qNe@lR$ow<L;H%4IpTY3zxmW}S48shx*ToAcCYr9G5R$Q
zqYt84N#Ks0pHZ;?hnQ<9!9|?bwPwi>%|A%8C%tU;{qxC-?{BTm-!Hr8)&C^$yE6aJ
zFvv5D=t`An^6#2Pit_p6NAs@En4Te#^&Ml`rN%g5*^?LNii^NhG%|m){X<J?#Pua&
z#1~bsC#B{0TCQZL)lD1yFSE)@?Jv3}T1a2OLCu%cEe87%3{xnJu*<$o`z!uBP3K!Q
zBqh+co&N_)_M&3{^$&zi*au)XYeQ1F<7}K|_Rbvdgp-@^p>M5~SIME|lIV4nv#*%g
zV)v{`J3VSXsl{H?@bn?NVZt-g0wu2|9bQhYXQ^<V6WJRKm?h9?yy?jZaUijSoq8bG
z|L+t)YfEQELiVDpJdFxsQ6C*whJ3aH7Qf1-iDqCP2$LNotY?9lnr?@Za2|2rDv-3j
zwZ;63Nt9nhhQPF#Ac7pZ3Gb_*m-)vo##&Q&R9pT&cCb4C$VKyeF3OeXttA(9z<{s5
zJZ`K>nF#L{wTIDGWk`FB_WZ=P0@<JY`9Q2RDX(!*Et)iU8b(;EsDtr<sm>W3tRD86
z+@Dx+!&m%Wk%yB9)LhUtQiJ#wmhO%n^q0aYuAobAtvk&qCqJMKvzqtprH^dr(po=O
zjN1PeThQ+K8LN8e!D!l1LU+HxP9;Aw;{n~8U0dgRV6|gBK9)QDJFPP`y35)0A&C_V
zY?!&*TEzC^>7EB!5#2FWyZM&4X#X=SM}3Gi5Zv|%3@dcj&U*gZ<TBNl__G3<uH5?<
z^0;iZKjp!7dVfFbSHmXm%HnXud#UQ!s}F_x<;OX)2;AoI8G&Z8e77l0=~Z!CpUn$t
zfsn%H5-4T0(pKA-BfOq;_ypOnUv4KX{k!kjH6A%%;Xi&9^b_wF9Lm52ELjoyuTAR=
zsH*Qdgd_5OrpgWI`LEwnjRF9#NX(Rg@{9O6?;GyTP?@lU97D3SU76XRmmObYOG6na
zy-aqMN;q!IHAbjV6@hvS+bOKp%Xj{zUz+JRLQSdgK$-mgWtq;Uwq{-<VIT={(UKCM
z<t(M7apasN9)pMu1Ka>NL$}ap%5@S~{Y8C{z)?pJlMnG5aMlr()r-rg(rl}^bV|q6
z#+3rwYqn?w5EE#6&5LVw8x|MBJkU}p8sMM%6sR_2wUU>IL)Ok-<RL@S>|-wuEo);y
zp682%{bB}V9j2bbAKocQI9X$PcO4FW_w=^2sdM82u?>g_Cb-;}U1Qjo;bUo;LcBy+
z;QxJp<s*e>NCX%P9IvXUbD+f~^Y*c@-wlBDe%k&wtz~O0l*tBm!6||LWw(Ej5S&X_
z60*4)Nq;UxunGn$Z-XaDJN;g!^P5EknHOiv9f6DLi1RVP3iIdyZNE*bRo(hel?>u>
zm6t^mHD=P5olDgms;ZMj>UAdWbWF(L6KBZ*?m{b9)vSM+B|wUzLOKu*98k~pgCAt~
z7&OX3Xmr_!+h^61a5IN3t+cmrF@n*t9z#i*xL5F=wVyP3;<zI?ew{^JKJOTj?RsO9
z6V&hc^J!QT=Y#=!EtlAFxOM=Mm9OkPwM!MlyAm*NG>CRSnO#0N!(+W%+Js=R-)5cu
z>-2(l#wDK|(xl|DXyCYN-_Yt!V7UWlN=mRUK*yJ@q2tnt13-69yP4u>mtyS-rRo|4
zTN$XRgPkZ^mY?dZV(6^bZ{A)=Xg?H#R@Wb@%&+=ATpB{%=9$C{+e>zx&X?UBwq7X$
zTW_4pTlk=j2{6l;GdC3Zn-M4UNHab)PYdHxdu4))ts&v3x!C!~!1lmG7i)c(TWuih
zt7Y_u;}pmE)ybMZs8r~KyK?(x$aLr?0bSL`JO2P5qItk}i<c-pBrwhj^H186EH~Xp
zo;9=@2LSHd?{Dw*s}I}t2tZlK3{sUz1WHld<%q~LFoo0Xu`V=sp^1w3q?8u&e*N1W
z)HiV0KkH*Iar!^TjVt6oT4Ps{Wl>ps_OU$%he*M#7>ool_E<K8Jz;xc{pa#Y;4^^O
z#;vm<dv+=NK)qB%q01Z-nfS+z#WR)6g9e0&IrSc@>flJtP`<Su{_w%yT-pSJB{Q0r
z9yN4MN4auG*l@AeD>Gu&Ti}E>cS#^oL9Vu6^P$mio#N3Wv-JeYZf_TY+H~9m_9tbw
zE><;x@ez4*wH%7ge)VTxotfbrq>RQ=4dE14BAWH2q|X?nJaH0ROl2t}#t3{2Q~|?F
zm3BEIP6O%`l_ItqN@EQk<q?bA6bARMCWf|NJvpHMlk>&|IsC>XY_-6zFQN8$s2gck
zjind%j|o%u2~ynQ${{uDpKE)6zy5)e)68HNq3Q{#ArKW2$z}osoZ;?k^XoEJ1J5We
z@kWV0f{E7$xtR(&SE;{y2<NYmBs_)Vf&c+O>4|%IJPL(D_&Dh~qVcCVjbAgX1@M{*
z4CO?<r@wkSON5Aq$6N`IfSv*3U(NiKa8@cJmH)coR?20W!?PX{j|Z3H_Ypm}?JOBg
zua0~UAJwp$`W-}k=EvDB^7AV&CMzd*G)nU^RHZ+g)LZnN8m9$L+V}3{8)*jnN~h1u
zbl3lu%f)w}09Txi`E4FaAD6Xh$)um4Y5Ia0o1J@N?+%BXdoL@Nx$bhE3Yf^^i>)e#
ziuU^|-QFK6C@A#P&b{+=Zh6hG`}T|Fwx86R2^Gt~r<2M&6}`{dX<%2jQTEEZ^l%l{
zb=H8Mpd?3)(7Kb-DIZe5YSsJLQIEvET;YqsqxhK=n}fgmvlN3*NT5IsUgN%c3(-_*
z)aCf}3rvgv010w39md^|*Ne0IM;w>r-2zuei9MlW0+qaxf}hF<Ae?5GyE>E6ZfuR_
zk~75+hm?4b>tEWXn)(Q4RzGY#m2rRsKpXe8Tdw(1Tt_856JFL59p&m~%(0m3w5s1y
zvbMAf6y@gb!*0hw%_S!8bd>#u1dCd#=tpvdNp8QD`=QjrFt6$kyH(>Y>rd7VBX(N<
zMxOn%%$z^c*Wt0Boy#+#y61COofWd|{&tsuZNi+v$EyB=&ok_dn)o+L_OsPLh2=4D
z=q1Ca^O?D9jzDa8U$PH;3EUE<{?$-|%v+@?yoY|=`Vix_S=Q_bQ@HMSqljC)>UNnf
z1yaM^9sSB&e@IPW8Sh8#%e-nHqu+d5rIft_QH`!{;(8Jn<S5X!#4(C>l+<umm<gX=
z5=i>v*t-;gN!$=A+|eA^e#^bYq6kQdZ6pM1uaYF02-$&;Gfc{oRa_=yn)4g@90|rQ
z3eG5WiS!i)WF&h)9jyF0V(9MmF&^WnB`z<OR~7N=Gl>q&2mS)5ojL<OM6Z^eR;Zsb
z$Q!@F0RW`gO0r}4+kJf9QTC&p^I|YrHqW1ZgD#xmSXoDZ;ykq%N8bM4v<>RpZKfc#
zy*|eE-{)jIYf;rV9MU+x5^?PmmyZ><z~}oC5q-Zg`9KJ5`@jqKW0<*i<!^ISPBIM5
zRf@y~7QL73RXqxTR4|gI)z6<~v~^Rzbjc0MOM27j|Kf7=%Dm72uNt`Xe)M5-tb$D!
z*ZUWcRkj~pK5wUsiCA2<A(m=F{w3fSo0RyAxLmaR7i5ZAMP;H!+@sSkUiR(O+!!9{
z-<}-`(Bl>^cF!xGi!{8neS?*4{xnnP!m6k?JFNMEi@R<j=wNSUcCmv<dKo%1jkcv*
z^_1M-TGcml8A?4AU>m})jg((qe4_-JiIZg=725ajJkCnCWE*kOO>k;%DPHGfRQBkB
zFZ4#NC-{4g^6rFj2KcWNh^ls(6D;qT76Gs`aIj!ykp%vJA8u?^G<`ub8TB<RW_xwj
zfCp#a9}rUAp+blH6RCp};M^DQE#o1Gj(%THPJezQx3TtoJxwZkA}aPeyFJ@*w3+)Y
z?WO{c_~V_3zPzLIk=0#QIbo5{O?lAg$}ZlwoOC1h&R9};_|+?#NxbD#8R>KF@o_nZ
zc$0#|d`77(#(?n3c*0lLMrZnW<@*9<UIIyW%&)G$ohWJQ7?yN#&@MEX(fWvh{!Ukf
zyU5!}cj1ZWVpbRX*D8)cS^@NSY)vykx0AHUSh1DO^!kB{{f`rMYhs#KlYEvxs$cpw
z{@6$4(rs_`sacE`UU(RVs~EP;hceJfY4m88CwlXzNUg8CMM{**5m8$$CeP)Tt}}r<
zv$xTo(#<43AD|Up!3RHU##~YhydcoWW)a$L>S+(vIJ`hzm|jPOwm#imn&D~JFw-fl
z%d57e+Wc+FN8&*H%Ul=b@ZN%sRQ?v$#N~Q`CE?jS2OieSJq#BZnQIJ#Yv$hxiKjm>
zlcsb!O);@YeD?VsEfn?p!rTL4Nbj``v=Nk2zKt5Z>}ZXqLHS0t^kfz~Ho|=ZKr*$-
zcj=hOUm-~I`J@abmHQ@thNS|9aQ0<;zkFT<66R`@r8M7FCDIjz=L46ASf(`eDl$M%
zxiwB}WnnKRgkc$y$3D20waoLe6M7~*kds?fU_f%eRe>MB>7g3A00FWYxk4MSeAfmu
z2S2Ul$~vfeSW7rIYhHYk+P7bfD75K+gu?0~t}<e-Jl3Pd`LpLQH4u9RM=DJ!v+vi9
zB5#U^LP)b{qS*yhjgpOhU~}Im;~akb^khnF0UlT}^)1Fyw^c`6*J=Dbg;G7&G}16s
z;^rvCmw#zw#+Z=^AnyB@nND!5klJf;3mPrhm-{fl1Z+_AKb-NM5yFnr{&`_78BS0*
zp;nz&hk42`RTkDk53(I!m+~Uip4qI~!jZ2ejoWd{Y~}Gj-|Gh4y%i&_uElT76;WqB
zHVvsGg7;VLSSPW9*PU5n$CSE+IVN3dI>v|RvJ^TCaqv&P5F7d*5)>^h6lx#n^$F&i
zvEw81kG`j7#ik?)L4pm$9dCv<tMVC^Fru}?u=M45q=@#LU3>tk7&6QUE?bt)Nmp9<
z3oe+#zWdAdQ-GfuCZ+PAE_2I^dXrk1QdUF!I3iNb&R+w(1C#YOmsFU`GMw{tL~gk)
zipRv@RKLaZkdt)tAWXI^uKS~B6LLBrAAWosPZU!jT)Pr$X<(Q6rH~bXjBg_OTaFP5
zeR-UEStj*m=)C~~QgNWbQC-j<!ewCw=x&_*cUaG{A%9!=;+mn+Ay%e&2N&>}!8K_<
zFFgt@lm-`55|F{?wK?Mpy2eS;aV}mw2f%JA$gDqZ+PNg|A3`qEDuGQER0ZE^&`{@t
zy>E^V1Zkk=f<mriyt7R(6Re}<q5^gWZ4yd;X={*^1zcvj$0PQ2HekExkYnfoG2j$E
zFJJ>0C8^x7kp#vSi6p-iwVw_6*MFhpBp#)zHth;Iuzi_+B<4RVORYVU?57aOdyooR
z@y)}1b9w7VKin(X8+9J$jX8X=eDg6G+^{ASL(FD*nGl#0EvGgI;gF{y>&eGWUR7-t
zUkFT!UIoN|E$3KYT~!2bD3_Q;*yFC%3GgtTG$I*^>?>~GC%aGpOBxQf@{t!W4{~7(
zjymz~$<TeAF~4&^dMyQJv&)!z5$lE^P!V;T)2Ym|GS<0<PN&r;j~+M8@7DR))Ft55
z(dG0Tm%9NXUQVyBK)bcTNMfW`+)AbSHE5%BW}Fm+vQ-qnCLO82-U8S@N!ut?Gaf8H
z4OMoQjB+{V#Q^h+-X+AC02!XsvQ1kuSLT5imhET9^jubGIB({)+~XPDYbX_e9pard
zj(FDj{T3SG)2EUR&U<>&g8QYjwcE=UM?9m6>NP8v(dOq(nEN|$XwbtPKMq|K4a-kb
z^gio)BKc+Gojj6kOyGxShX3~B2!QiLV3g5^1<8L@mXm&Tpl1aPJ$}W%PF<vQDSd;0
zT3_q(lbR>Pj``bX>X~O5&zi??NlCM%aQEM}GpVN*%Q$aWSR9uqKkXdZRDjY?_2p&Y
zYz8sj;D3t&K>OGALO7B1BXS^^c_|))xXb&*qZi6bt=dy;SC&kHpBV>hx;<RfX+C|S
z`_QYWnV1CSqO04V86>LR!v&6MyC2^u!#mN$?x~$q7r$bllm*0tp6&*hx%1l6#KXmO
zxg%L_wbikBoaV4e-_yDe(-)!#ydc&NjGkOyJnqMcn*~)9!Y+vG@T&_w$Jdz@VF2er
zmSKP&mh*K?nTsnfgMF{3H%8`2#QOwy%{mC;OQ2y;$zc6!9)FzcBYyQyT>wMKbjIeg
z5~ELSkW{u7_p#f45b57@`I0IRisfr@wXwB-=g4)g%M0g`ABQ5^Lq4s75)0C9K>-}y
zZHL8_A89I#Hiq1u@Kvx59>ll2jX$I2sjaBPSxfKq!Ea?cdErHyYqD0GTnfJ|aX{ZZ
zI#P86)OyU#zFQvGT&#9GR#nWG^6Muovvv3DAzj-(mz0(6_qu$ah4P4zBch+&Xae9u
zf}{jbZj9GOKh5Guo5$%q^eOSsm;Y?5Tl-vHi&<dofu2e#`{=C%%V%LG0WL_A@i!)k
z#`pYNi%VtopB7V>C8D3QTwmy!w#e<xI@GG6Qj_+SZ=18Cy;62!P&9$pwFmNoF?%qZ
zi2r_YG9zRB7B0yhKz1aOM4aAWYZ)%NB6DhLbq~fCteKCYl?nnXvcKiKEEP{6!#P@|
zRW-yVS+J-?x>-PSC1pQWZ&ByWwItf;dp~(Iezwy1vP}p^mMcD@){FI6+e;|w)x1tn
z2*0UN(B?zl%G9?;jJT^`Zd@L38GOk2b6G=)sEWT$q?{jv^J(BTE+$sxeQiV3kyvL`
z2&qbRWi1uqRQ|_Pt*Qr$@xA2zl%IFFL%RTACi*#l%j$9_rq<?wN3|J8U_p1}AMWS#
zXL0WoTPom0oXi?Iv(pGkPaU_cMl@Z%I5&VXOIT?MplK)ueg8ELTWQ>=J0hD7I@)HO
z;!Va)JHrPb0ajvBgw#k5f9&b$oR}(r$6@8`sqv5`=ARfuNPp=zSk?P2V>Rwxa75st
zOQ*pD`2<YmVaimiakt8NU;LTLs{1Fq&xG-X{TmoOD)Y2x+?(R`StWzU<oW&Wr8L(a
zy!~GML(3jN8l~m76;o?2f{F>_(&O^oo#$9fIo^15DFHO+ho9*CcK!$U$0<Ydl`+>8
z?Ny-xb#-vayuLch^?btunsTul<EpU@0PJ~Zi8jC6pXruEME1QC&&7JIa}#l}>X*FR
z>a%2xo4=&Kg})t1xzk8Y9yAiW`n@!U+|-=4Zyh78SAo-e70z1p0lN$pDvpY{-Ii$_
zqU+C&<JjQCZg;D;F8^lj+ioa3F|b~DP<r7xU3BfJS^-xHRT|{^N2?eff`5))arpY<
zrTCUBuuEY?-6My<EA3F596S-RXRkVd+Vm__HVLefe82`dj%OyM9)y`UA`_-XaORoI
z!QS3AdD3K_wg6CMZ#v`B!!p2qKf!)|UQ~D-K<|8Zj@;Z#Pc~>s{ig*dx;yqof8+QK
zX5G8aXE1sCLfr2zLVD5&^hxRJw1^Qwl{|m6IrOxTBDa}cG0JKr3wgA^{Z}YkHDj+J
zJeh*A88(2J7h3&A4EF@h2OrLA0_Wv<o~r(JCZ`jB`_qjiZPiL=sLj|DkpOoD5s88u
z$fxNk3!sq_p60I!q%4xwhiuO&d{j41ImMD&sCiPa@zhT*F0&&dVxLlPp!vfza|t5;
z#ZRpM;yJCgzLgz1{WGOA#!xQ4G8J5F`XvK7?3QryXp|)1@==C+Lbo^dtZEY9m>eTs
zkx%R@eP#m5M|@wmdv%7E1&(P`eeI5_wOApIrco{WxjK^5aT#C46<zClN8EjRW9%XE
zh&!7Y<cP0-XD93Ph)FWo)kyg{=_}Cxq3Nvuntb1|PbjD;6H#)o51?QWQnIlCL7YK|
z2uh0s0bz7EDkZRqv<whXP?2sJA+eDg-Mzu+*y!@%`#i7bAGm(G@AJIR`#O&EedI)1
z?Nt7k5OeK!-ts5Cj%~D5QB(!EOmC^B1`)L$l830*c$(TPUDQO(dZ?Uu$}xaylg~@B
zro1wlKIz1hB+xYxdO2oY<W~`hmDI_zK4gej_J2$5rLtHs8oKmpKRMA2Kk4j8Q%;$+
z_H<qrcrcPM&<k?LiEj7}hakd;3V4&mT`B&K<BwfzX`ag%&zPCRVLG9adqM`egcZIp
zoL)dtw|W>bWZaAvB^4!_3vohit<OcRqS^MZKeq3PZI(ZmJ$jN99=3sLJ2q}l(L(}i
zv*AMgQ#BHD6S#vZx0<vpSABuR$dgqQR0(8#J9jB5lgO5aO;Z)h+BIq8)9PF{B|l<7
zn*D|DS#i+4=4FLZ3VXML?zMQ`!pP7@H>*r5s)lIfspA)Qf+`ibl<z2<FpW>?mY`vj
z2K~tADLZAH=Cgi+vsl}9<DW+aicSO`jn!Y>2|U>iEaQWG6Xu%7KR~=VR#tND=Qn>M
zI=MOAnkkKgF{O^=i(ip@U_>i(A_;QP=q(3n?>7Ed!ag3_1E{zN*C;GA(nM!1r$MDZ
zUpUZAL@hOXV$YDom$%%;4gc<K({i4FPeyU3X-UG;m*u*7jVv63Rp8cgYu+@oV6rMG
zd-Mrm1ETjrcT;Xx!4@3v=So(QjQv1@iK{D%<hi=;cTn-eS@mhey2;`$L@6R-JK$H_
ztVYrq?1z<Mf^cT~dGja2B7+a|pDx&oxow{PyLZ#>wnN&n1hh*`1?LEnorp}i7aGHJ
z^g`gP-}-y`q^TK~%q~8+{=F-DRgZ~>Dmz-q_c&K<npANsCcQ(i+~PEp_cWJjVmtT$
zwX7$Q%Ca2muimKlew{p&BN3T6g*U?4+YMbxb+iQ@Wc^b)dVvKWI(6lzDEEt!Vr)BA
zRV=tK4@NlJwVsBn->mF$NJNVRuCeIsDBKUw$CfMiaL6HFWJG>;WE6ghHuSF3eV_H#
zJaz(n5vEJt)drVGucyEQ1D@gJfgPcXD;KY*I}bWt2Wks0@J*R4ysAv)4U2SfY{z8D
zEMIsjzL&f;Ujqd<c1JX^hja4WJ^E%vPo-U!`JhRqSiS)TFKRr;cs3=!u@2jB=lG7h
z{^z?m0Jh~ZeYrxo;L+C7TdN7>(C{ol?~OuZvUZ5NKC9ekTUFlDEtO#%|NCzLU8!?!
z=8|XPJm<x1pC!Z)ePOWh(rw?)^tyeCIxbyv@EHhdbQ}`WN)@=GM&o@xJIvZ$w|f1}
zP{yDlk6`zuE{-yxj`!;;kKqRScRf%3`&xI|_epZRaDO~1xvpC&;>c8d{P`}jv4ieh
z&ydU%Hn$^Eb3Z7b%=q3nU*3Dw-uw}e5$xt>DfAPb{wvQJZKE#jD}^eHa&C-RlQq2s
ziC2DTc>kQSe?V(p^z5|v@56!PaWvs_hJ#q;R752oaj7%5FDA{alk5Sihf5KM^s9Zg
z%;%?GD39KUja^|FD~tK#R$NoHHYrrw0Z_H%Sh3N87)e{$$<<4<32tS&@i2wLz~J@&
zlItLTU~d;w{2)vCOIxnCILQ_A`fU_!JHN(wP1fh2g&4%xUHP;M^>Gq+y_!A%ddK|X
z5J;IHjbHZugcXagdtrV)$@BLl_`N)6Oz8)4zX&!+!|OMOZ9IgdQ0<R)Whq4%A0LYU
zL>7P3uA3$P@QdV~zPTfVug-l1XBVdWLmYvG(LBb_u}OKDAhqgC)Tx<xs08`1XhXVY
zJFl;<bL>J4Z@Rakc_b|Zj17?rbDmLy?s!*Vx{GhjhRffZxcVQ!NeF4GXnj!NopXz(
z`v=t&N2NQIJS^)fvs!y)`61EacdOkTmfTdPQJ$u<+-gXnmtGpZI;ceaO&^>~>E)Kd
ziLRm+BbDFy>HEVE_O1xck)u`OK=6*PRoAk7<WTO1n`ln0@8*gQ```-tQ$g=sH)?iS
zXM$8w?R*%|5$)?m%ihrCRwB)d>pJNSk4u)8E|$`ISLh%nMBC|41CnPG7R17iUjAc6
zX9`nv|F=HrbhY~!w|zHX$CXyPnSR^Nzq71kt9*zXmlqadD||&dM<J(R;<nGQ>}3(o
zVt5fRLx{#~^G%vmYAimBL$)@rrVJ4)ZZ4qUyL+#L^0fi6MgV%5kxu&;yI4BCwx-`j
zWzbm5|CA8-dREbt`2LsA1j5hgP`gax?d(V!CBZEqf8f)BDWb=+9(dy@4Hj6Wpx(9J
z6rykw4~b`3?fHZI-mdCZA*2<1>VK>Uo3zQ2*PrKUlQM*avIx^(BRoAeU_G7uocuT_
zkX^5FM%X8pa(?ku2He{-+p)o3DouD*iZgvpX7U~Hm4;)2Qt(kN2&dSJvgWMJM2?MM
z-Npq={a9NX<U@V0NC0X+G69i6o;{bX*O+M!@qxEkV*@;<hq5?x^O`$2QoJ%`L2}yE
zpDa-{v(N;E@+&OU`+reD5ufO$eWpv2*|$22d6SP?*{5YyWxmWLw=~y#sI)y2&y`|N
z@fL1m&r9!`&S;8ZO{;r*x9THCFK}I2#*vW>dtH2G;{nT2^Zv|bqJi5RRu#95M4$Vo
zE0o&jO!GzW9u;vuD?bZwy>Ip>OifG_h}asf4d8~fP6LjBpr<P8DxYF=Rz?LU`)@`r
zU*>VEK8>LM*eWliAo=?5`KAjq>O%MZ>8D9s8IJm0%$A!etT5~_*CT%-{ULvUfk=A@
zdpMY2#&!BANq6#vR7btFx!>+Jy*wySJ$DN5F0)E~9Esm}AXpNW<RKVd$wzi+vzrmc
zQsC8T=)CG;|EQu%t?0J3nJy)s9cuKZ-wJFfWc%&uLjT3=Xp4ZC5#F*GnA_Xh(`j%5
z@b*Z(&-Zj#7Pc&jlw&!@9`=N%bCc(O2u}j0^6fUq+SJp$QMG0H9o5=tKhu314^BUi
z31;*t5Tz^%GIJZ%o(3DE79w};{;C*2v_m(C3VsON2yX4B{02~?%EtC4jgVY@ii$(!
zldg6~P4u)B^&06eui8K2?%ESp{;tc;*PhO@aP2v{&CA~QvTr*m@fXfjS7B?7D1I~U
z*NUN|V7<h9+?%a!yBeelX`YU4x~4iGI^1j4^)zA1HXYy7F4z=3a%F!qRMwCDaKlk?
za(xczoqn4HGbxwJ$wJq{2s#iM&?Os)F!z+uK6sZq)Kw<nUQtWp9Wlny*WNsO^ayCN
zww>DtU{CAUiX<GNs%g~CD*{k3qts2=ap3OYeXJ*^7PcaUc=C#^BQUpQmhznmPm-@y
zrvp9Dk(kUmbP%*hSZ2_Rc>((2F)z8J-DB2D?om|vd47~LtqT5m9y#jWdN+s?4176B
zgw1NIM*kHz;~`KTMGHJlGwp5WyQ_q0-4FLM6(mE?w5G6rTWVQeLdsb6W^QQJii4vR
zq$eW;&TiDh?Hm(UzHr{o&GOLKFG1G&GY7hS>Zm@X9wX4YnrWrzPb<gCfc+gxQ8^0Y
zd|^{v`-aXR?LorQq)-6dp-AsTiO=@oUB9unzAs~^z79;>CbIvUJGY(+^Lp~hJ|YiU
z(pDu#Y>)SlzD(~D3Yt(WwQ~Z-oLxPfX#uf!Z?^pi;2q%4ijcv{b<@!zm_QWqRDrf-
z{dQLSYO(Qx5Onp$P2fYTYl+XTM2$pb1pM#`8;_>MMX6w=0L{7-%<UP$Pwmp&j%j|B
zaPVyR0TBb-uu6>Ly9szk8aC*c<IvuIcyBulIO~_&;PD@8mE!pAJG<c}d#VeKAGC(q
zZyHgvSRqHP>e8w~_Ca*KAl1n}q~i>n1jJc<_Q9)9HI;B$^K>uSO_!N1Sv)iLm{BJ2
zT~O{_`9tZnv(&R5j#&NPA<8Kq!2`R@`~Kq!^o|_I%m%vvSPWzs7ww{_o>#2O`t0or
zfB#GBMxv<~@ZV)3Q@4vhQ~L+uK$-JE+l}Fmnxo&p{P3?_6%NJrH<p}NWk0AqDZ;m4
z?lYIOc0b(6JZp0)!D2Lt-5<tsV;;hp9#|pw14Ki?o{ROWEA2>@0o5wT!b|d|KS!#!
zxls*y_d{DLUXl44jwT*g9-NrLlM^o=v~n}DXQk?R9G;NW8K#~Iv3-P!#;P3`I8F8G
zWD-1zgGi?>tn?IuMc$8zf!*HHu4%vhrXTvHs4;yd$RxPzvqNKOg0?n^??Zyw((e$*
zwg^{_yd$ky``9L}_LHl{CkE|3xlolF<NdgaW2RjT$I1T~UnM9g<srNIi1oZR)@iaE
zJQ=lK?wbBfx7^*jM!xIS*E&p#93WVKo9Odk{nhbEewDIhN5;Od#dj2JP8Q4&SzF0A
zDSJ$n%@)jxI1n?9LVh*-e}%h|u<3I73qyU-zD7(Z>eL49CX4Wpc~A|+4&5Qh3oRxh
zJ7SfL_NNPvA4T$lSoA#ph(U*R@+mlsvO!zcMES~Yk{1e-mhi=93q^Dd$s}L9_VVV=
z%Bz8Zj<vQ!+{R`&a7>YS0#Qh)c-Q*$r+KK@I%+A(M_7#1#B*qDHN9d}-)*df#-uL_
zJ;d1xRi~<ry<^lBWzcR+$#n&~7?@9YrZBWBBD%{(?+|>8;Sqyf={J55YrXw)HTAY3
zgAaw@Sqyd+spxeFz%-q;XtZ)fS5J1URr`xgjXs&3yyY+ZNzd^3KisTRLUk%~>3yeT
zc%{McmrJJ2M6ag=;JCPDXAeWLYeHaHh%Lh}=;UC<t;pZ$XS&DC$CCFIp^v8o{ja+n
zj_7i!3g*=W+HiocQ@bnsJ%s*7lLWGJOHl0N+^YsRT@;c5)2hrnrBL=tpaFc<_RSdd
z9r)|pmz+Oo=8~S%F!-AajM-Z!gw30kXT%ETl?99EN68q6vA#D<)6=R@uLnd@8oKRG
z3={Ww&FCiU>%Z4XKfioIsg?RQ(VU!=d(veNx1O9;fl50K7W!b?yaUEIZg&+qzC4h%
zCT^n4g%>{jZf^>9{ZEhysL(L*+$(}o$n579!SFBT^9K7$U*{Dp43RzeW~K7?myAF3
z7z@ZX=+QNi;-JXC7V1EWPgWx7rVumJ1zzPkiL)DW%1I`@n*cnM)qOL!pd=cJi@%tN
z<f(g}$~?>><<q@-Pmi&*7X*Htd^!%vXoRrU@aa6)F>}`DLsyYaWu52D|5&>iep5^D
zaG#HIZ-nVS@OPDR+8H)mx%FQ#n80)*EWE}OMq4~I8oGg^0;+x1&lk9Ve13+iSK*P=
z0wbb%GAPtJmz1=-z5F_<UMxBBj&*;+tVLqUnl4*=NBGiwgYepMF_YMHhT`aciZ|T_
z@)()MS22Fnsd$O()!}L*+{nX;-}l{5_Ke~jg?wIC7)6As^lq&<S81m75em4Cay<3&
z#ZaBXGB=HpYgftohT(<ZGs&*r#ah!!kR2<0if5*b)V^Dh$r>c#Ph;O#<D@KB0rQlq
zt@^jsqmu_{=^+j6-S)}MvEJD)T0vHnEh~}pug8J=0m5506C{HwKv&<Fk&-UhT1x6e
zg3DI5pINCTay)wyoz-ch*G!J?8}dF<N~C4DjlAuxwF}BG!rMn$K|A+lQ*HT9!K(wl
zK_M+>CfwdW&I`2RWcM9v-lz1enj5Z9r-(;a3rV%A{HWuWy%+ohn`5)^s8iYBwpeQf
z^C*R3WZPF0fOF`1-+?mkt0Ie{`D>LIW=00&;oLUS+!i`*XT}J{hF!xS@GM0dyT4W)
zN46?3zMaZL$~0sF0u%bFp?#8P;VJvs)oz8jdVO{-Udl#G=@`svTeY*;SA_d!khibf
z#X5VL4;kQ$$=QsoM$BZx`MjQ;geWlAKzlvVpss>0O7cLCqndNt`+7(0#H5TUgFUP=
zv4zV_+-BZytc4)huBH#<JD%Nh)j#Tvo}HOu**oevFgWS2D%7JJ%LO(auMPwWNv?Qh
zIMz-CXYI5|^ZV2q<DP>`eM3!Q#uoU-UGMgO#ax&6c3_!w+xz_1INbaCo*0LNd(l@S
z^@Pr|f8%X%ZsKwnz`QrY-CloE@4DJubs27|;s*u0C|GD8&@&NDpUpj(Zk=)3xO|7$
zc>X2E!*%j*M!;EY_=R+wJkyR#sendB)7*2bx0lUuJmBQ-yBBmwOyO3%?WRBJNk8Y$
zJ@F7oy<}z82uhvHf!Wj2i04&-{{e9K+cLSKg<OLF)i(dSd-*Q_G@ZysznPz^G-zxu
zU^5ex1|qTH^t)#YX9XWesWAB5bTCklrDq6)=-2du4THLBv7ajry?xiJA9c9?i<^M;
z2CFmqF4p&8)j#@)qtDx-wqn|EyWi+l>p+f6aVrO?nhdO;T(0pV)sViar+c5PeNQ)`
zh*9a8y3|i3Z*cxiC7HRE(CRqdVNGL51o}B|gwLG7YbBq@dNh;eFG$~KI6uqJp5kF*
zXg2ma;B<;gt!z(Pz6D#pd0lpLc06W#upJA2qE<^bEach`cf~d=FY_?a>PKw)omgj^
z{hkB@6Ff$gFw?OpnX5^96e==G*e&QJ9}5H*r7=Vi1R@<}7KJ>~pB4U)Br)nRK7QGE
z*2_nFN`1kEJ<mg4Jki&I^VJD(RNv;Ogh{=l>;Go~r0Mgop>ChuiylJfTu~*SPqEKB
z99bDAEqX4g1Y@@YVpmOW)NWRrYV2D-dg_~7?-AKKj9a_*OXDRm6^_~W>Ug_S1l7a|
zDw(${%w><Z$D!9!C}`ytdGs;31vQ>4tR8J=rc$A_IR3@Pc%dm@BCQ2$IHY@zqB@VT
zPa$TT?ajJSL*Q$US(07^rja;bz02-m@;Nm=;sk5cx=WUpg2%As2+|f@nL_JjaM~J$
zCG=bn(~8pRDW}UzcZk6U$=d|uwUW;utNAKx8Rbro&u=&X9V;dA%$<?;z<r+II&=82
zcs~dYn3ZO?L4^3j#{!zB3WAH4DR|)9TbqYWBzbI~t2%>AnRC5JoLiKd#f<Y*_R|Xj
zJaW8(|N2Snwf2^O*u@?LTG1DQyN=S_-qTLM?^+;_UJ$uzK?WUVr;Y-@B@V3zY8+Ys
zAMa^^L|ywrmV-}mv`SCKxp`G`0tCO@yEWRXOwV31AM0uA(QdLdF3tS`FyU~lGcyA?
z*8RF1U{t&V9RH#{fbMT|P{P!*G!S7<eZ_L@CLj<3a0d}k1$s*Pz3uJ)yq@9apAq4v
z^<vVdS3pM1IqoVFn{Bs!<wb_Ff7<l01zzqVX-16H>;NHX#SVfEo*+axjdAcLS`3-X
zp0@4|#2wI=XAmKEz*<6)U2A-l>ndI3pV}^CF4^O0_`KTG`O(Cz9Pon&MEKn#-MS*m
zx{vpJV6tXr2_+l+;(H(YviL|O;~^&>P_tj?UaBqcS?w3%6U;cRl{7I^pz^}N$2#WQ
zP+UWd@GD*-kyhas$NqYj*|610&U`*zu23%mtU7sF%Jy{q>&3)XiQw>WC?p%@Ok%oR
z1N+smWHn1wPw(Il`RyTWrmZZeYA~Oj5hoafr5XxiyE9j|WVTF+zjINh9kDZ&c+qEp
z`5GT27k1eAt*cL>AhVTO!Y!VLHph5G&vtg2XUo%*0eyHe0LeygR}|bzU3fGi-y#+3
z++5@ow^B>x!yCoPrT%d4ymc=kOdznpQLT2UC!nDS_o>y&+MtY~Mr1NbmR1r$a?C!;
z3b182q4`7)aQ?Jjq@LDw&?TnFbVz2LaTf+ew_%k2h5P4Pb2>$v5HzPo^Uk0H3R?ZN
zU*_5wv(QQzt_w%5(K(f79<F@pi^jG<`2oTlL$=TRV$!B8I&~8w$Pii|I8{;_ca_ak
ziv?y6#~r)~u7C$g^4T8zqEojw_4iQx-CSt0at!oFitP4Pt%RBgH)&#=CgDv==r0#&
z`sZCvc#?tK-{ajUnqtkh=RKC@spw+|q4H6ONM+&4Jni1KCe-G_WBt}WU)WaeR=pG?
zLFwAtW7&yV!OfVu9J>{G_fb(7-%#wX?4~O^`Ki)d-_vb(C2=vDEW+e$_Rqyb;z1Fl
zn&4)CfLIx!Af(D`iW`0t3z?5_G`*)<`7$}@`Q`beJip$ZeFN)_YVj9k!s{KaPo?q-
z5KgzaaNe#0f<~3HV@BSFZ0l^i0+;1DZ;*+5a)fYqBG(|>f?u<H^L-t&!5M{OoxY?q
z0W@f7C-H)D&12ASU+KW?$n{GpHw%dMdd2y@(Eh8(xFeHt2n;SjV8ETG;uyYryG>Am
zt>M<gUcT{;<33LnMZuk<&%@4YiG3(M$=4<%KhyP^(OAW=&n}44KPiQCu7w%;oGSt~
z2<oxM6y$Dn(@hS3a#)t@8%UsX{s(m5316@5nqS!9{RObM`6cA{q<r4ukbPoopD(gv
zd%<__Rl+0J@2Cwt3v{+fvEX_~+LVd--rvnxlE>>=t4#HFQ{=SV|BO}~h)z!Q9gxq;
z`dsk(+~1lD`0fhn0B}A?07Mb&3(KoL5RQc;@Xm4ln>l{G&%f@Z`LMHN$NB_9Iqv*$
zGzAZ=4hLb&NPhsZN}fqr?OnzL=#HC0TwlK~hg@K!IHi-~`#wUWVejjK`|+{j*a>8C
z>$FYmZRZ+Y6i-oPZaM|XP;DlAxNMp^90~$T*cyu&Z>?X&$tk#C-k}mBg4<H*t7#Um
z!!Kf-BNf0eJlrJ55<DcGUa%7jxz>c2{NBxYx_=+j6KXM-txosY8tZd5_{g+_1|a_=
zv*XRHk7pZP&%Ope4lI^H{5M)GCb;9)DOi{_0Z^iK@ZR)TdwpkWedL-@>}8xX5~!M8
z?C1%W&CnSwydlAdZ2XIrdci%?I+BCY=<Km@?@f2?dqL|PQWlkI&rrYRajtFi;^aV+
zJg3vhO%B!8x%+J7p@MX2Y`qx3*XsnuQRCH3JD+fq)v(1(l-19_G%iG|B{|ziEggIQ
z@X5g=<n%~8SK*-giBfN*Ly%n26E1HP_N_1f=Kcm`>}zm;f;HWe_u88g0aY>09nJs3
z2*2=yt~~Qde>`%Odet(X{&Z+)Hsi;C>&E_vfQBCW4{CzrtH(U#YfYp{CqQ^wIf09P
z;|ehV(=;~H1OMT3U&3m)$3EdN2YerV*24{clgthYU;@U21BP{H;!|%>#QyyneDS0I
z+H_&|L130^q0Rtl(>Zu`e1EvAT-5p94eLRH<<NlkJMS_Mix}g(x_k-oeFWr4ZrtA^
zS(6ZzrD!=9pfXjvjli#@5o_fr3a6!IT-@St?SK*UZ`wIjWn=|T^Ml#)UHcxG67sG*
z3^<rL&M5<wMJ4?UD=_0xD5txosms;->mO>(Hk2Pc63@&CCLaIB9+l!?4>58vl-+x?
zOnz(~;+=lkA|Hu<oTz{(G3yn0VarsbMzY=}WZhR!4yQs)c;&b$>;ykKHB}Y(W@%qB
zcE}CrnX}GaoL^8v-+}RADCoH|C)X|A`YItqgIp7m>s|>)YV9xZM0lNsS%;@*uaJhC
zZ5tiU5j9pyCp;OCnp?cS^GDL^-bL-~n#1lK@F}^FS^ap!ic|-+zS<q@*>J=VR)e~o
zH6%pZgOZB=TW9wBpgey()4@9@)W$;bD&4TXc!k;Qz@+geyFd;<vH}z8TBq(c7+keI
z?O6Iy2ry79t>Rw%I2@C<dB8_DC1x84H9Ife7N`xs0-ct3<V@8$cjV?{aA3A}$4x%b
zkKUwn17K!P-NNt`9&vin0sr=N2KzFNIjrYC+WQ+<lo&tVZ?F_Jk~8v_0^$iJsayrg
z=5<jON4R?dlHcWV(9h{&>3_4K#*Qy7>F`-JfIe>~*uG|G0|gpQ;+&)bJ91UCf}bKH
zcw|_=^VWcQj4CL=NQ0aR_n0cQn1_Wt@>k)TO6$PoYRhC<^u!=Hqk(zB@~o7yEvrc&
zs{-)H>^=J<T;Y$B!9aJm81nW~{57G#0|p?n{*ZF2-6`6(pIJcS2erh6gSHfn%F4=@
z>Z~7rw_3W3xL!2{e>_AnKbt;R&<K5aNCQnup6qq;?ml`m(t5pXS30TE-RF;k|3B5I
zV*~hEbp5?B&{txRNj|ZA^Oxm`RARL9F#}sK>Et4R4tla^n*8VOP}ko?b4gM_fE1CF
zlvdqznMht^WVOV53{78T`6TuT>gTZ4$@uiTV0L1}SozEa=Ztz2px^eB*eT;ky}V}>
zLAd76^>@z~H17<j-Gz|XK>9Ee@Y~9PE{b@6qa_h(HKZ}pCMMC;<65_<XWp30_!*Ez
zqt+hK6WJq`RMI8Q>XdKSAd<XP1}@yg-2wZ(QfR<Z@Rf<-C5>IAZ<(+oi&>$ZXe=7~
zB`bn<n)tA9qaa>%vtCF$xHhhB_18iRR9+P>hlcbZ5lrX3fgNmY7?H@8vs2NV;3mu&
z!m}+vh1_M)`R6&T@-Brj8bIKwp>3|g@a!w=#|wbcZsx6)XChy%-=1;HV%u+318ydQ
ze?C~CtwoG#LL9V1GisEpeqXn*@Zac{y|!wxzbe?iy!mGC-9HV`5GG)-ylV0$x5r(x
z4l?hu2D>+tA8;&GkGAM!DdOI!sLI0xVp^+m`Z)Z=+izkb;ahG085#EnV;Kl0Ot2Jj
zs?E7M|M$75;7OY@7&kN9O74+Tq!{GQ2r{;VRf=-QgC~nEbv@9p!s}$4W3Jy$Rj>_e
zRoH5!_VCOkv&|v!Y~suHP5kVEYZQM?*z6&T+%Miy%$R!r-)%x2TIiomRS)N^Dm7v#
zL9Dk@$u`VHenB+Qf-_JO5NB1sU<~$>tJB}@W(JE47d3+i$!_*J_=Rj~epc4fTSI#H
z7Xa0t$<Rwj?3vaU+{Bea7a`zKO!?r~<|1IDy=QZbCjw6iHfOiVt%6%Yqy*)<PK^T;
zpVTJyuB!!e8r8Q(=94pB5gbhL$F?RHleeZx9*6eP<}Zt6+Qqd6qx>Po<0QFN^<KE3
z*rc;sgH`^o15bxzOj{xTDBt!76(&tXEhQW4pzb-1FGldJcnROAh%Y+HAJ8VcS&g7d
zm#K5g>3xXIv`(G<q$b<4p4Q!>*|fxnA^7M@Ws}y3BCT2@C38}1?7jWahi%Qp|I&+u
zP&|YzZ=(FZ-m!%a5$AF!EDG!Ab9l#FOiV1AaK1DPF0}$w!m^!&x-Mmn^I?ysX{4c<
zkK1ob)9^Qa^_<;Ws`>g>VPUT5*PU={B6w&5rC$+}z`5)>oN0Kr;ub_#f@sUY^!cg?
z_U(iI?Z{~K|1ZKL!7)8SmW%@X*x!u>;fGQ7_+;GF1z#n%$Q=~r)j-2L<Zb>W<LoEL
zV^d*+4X-Gq;+GjAp85wo0CZx6o3KaPGf->_=(|*Mj0z#clzl-Jv-N(Q)r|?;QywD(
zx}c%E*LTMp0D7gwbNs3rk&yQ5(_hZlOFm;}Ve!i+Q@{LNqk626I?IS6pm=>I#YHpU
zTeHbC9Cm&lUp4bv#@u!8<l*k}Z)StBLc#ZS|5C;MRTeZXp2kM{%3yeMRO-y^LGSSY
z<&fC5wBJ3AII&X-Unpo{e<AVN==+K5M?t7}zO>zB!ffe))Tte2Pib#^te2C=gKbr1
z0)&C(A}d$lRd5nbii}xux~ed6)&`>BkrDA{JY{H0%_HmayeUxac{AsE@uxA&7)le9
zIK03;iWXz=$>opwNQh`yZ?MNX*^B@)s%%xAzWBtx@a5NGSgakYH;K}Bs=Y}|Ur%WZ
zXsYr^0X{Jc)kEtlooKoM+^+J{`2S@Vi-1!6<ABIaE~KOn=L-kWu7*#_geadZlP*-a
zy}5Q{<6xw#gT;zHtdD7uLr5dk?E8O$G!6`BdsEDt6@;wNrus4X;d@G-nJw+2A3kn%
zaSw*$`(8z0at=};iBa{C)6TC8pT|Uq_W9!Of2G~_@DYvJ3198u+Az9v3$hf&tHx-K
zcpS0FBGX_(d^y_k5gouEnlazT<@OEmWh!B|?xy$xq&j+eQTg7rooCt}3h`p=rRo<2
z4083RJO*X=NxPoyYj>}GUT(*Tr*Un#0*Q6`hb>OuXY>|7MVfY7n>Ncw&1sBQxov3n
z&{v~rKFF`{Z$V<FWU6B%2x6<so0Gmq0m21hHAKd-%=-GDw`)D{tMzR@KH*a7y_m`E
zt~tYjhYQSW16obL(OB2D67D^bdwI|_oN<`5oMStwO71jsv+cH@-aXB^KI=rsfyG}%
zoWllTh7SShi#W6RNxyVsFcESD8Nn|eZn)kYqYWn*l-cSZkTWK8?34P!wBu9!{zEin
zd)+2(O}SGW7_lDS&9N-mifBwDs;lY>iXY|2S$}$+oo1lYo0&p+7VV;RjdpDPZl+t@
zO#_LY%iH~^xci{>g6ahx$MEwBW4r^8hz36(4WH78A9H6UP1{SJBopj2MG$9%+y7mj
zNKOpx)GB=(4hr#1Wn7Uvn;q?jzq#{GBi+{4yr@nd5Gy%7d!5Zzxp93GU-ME)$`PHw
z--kg87Ru)QwYnQ^1qR_64Kdi-qRGF=@GXKhgh0GJ;wq6K2U>&-!Rd2e9c!Ag%Mvul
z+sCq7)BPzdD|{s_?r#Ocbk{r!rsA&;5%=fC#0Xv1$~mZ42r%gKxJ1HC_ef70tHX=6
zhBI0T@|MRuM+U5S2AHPWnfDB$f`#C`e(GV#p3F5~!y_MJ!@zgk(`+PeH%Nmv94c2<
zEF=J&se=9~VU9Z{=0miJOS|%u=Z|Tx68nbMJ9PDxTpD5ikd+hNlWS9um=hmA98N}{
z;;zSxXLQ!wh?%pfNoq&L`#z0WN;i|{H&C3Zn(`GpFlKeyDF+;%WJ#=Qmi`Y1OM1`x
zKEfLZU8(%6lLdbNN&eA(_w-%5GissXmYRPI4YYSIA4+kekK4o}$OZ*#jcp$smP%G^
z_&q#cOE3u+>dkUMTD?x34EJWJfA6zg+j~0VjsXAi5@}}h;%eG-(!Wio{_HLQGE0~Z
zF(Y|Y!<yQh4Jh)`(_RV;S4lya$5~-wR_5sgD6>yyGtMj`pqk&8M&Z{X!|x#dA&ZRD
zpFYaDnf>S50&c9|Q*}m(|G*Cb#r1)?_t=7i><BSNEoq-_2ga!HUyO2R+Ut7y)sq=x
z)#PB~lpt6bc6DlDrusgk1czWEDW)+mY&zVvpFUvS=*<ptbb|qhcAj%Kq9JFDnfw6+
zx+vMCy0tHo>`ZT{`J}xh!+3CVl;2z?Ct2|1ALQ}jyT4;IP|-Ym*z9Ra1R!3K09v+<
zT9ak?nX6=Dv|1eo^zBs^!qJxwc*wuY_a1#|C6emg{~LoM3Cb{dRS|+f^=T=et4w!S
z4s8X-#@lLv&=Cp{1oECjHj0zzY{`|QW6ho_GB2paGinePlTd9tqs18V+#dU+&HWWo
zK9uy&T#B@v<g3M}RGKCt`B5&aO`t>mAB`s;B_BCOsidJ2+E#t@N&1UJzpz#1f7FO?
zL@-+R!Bgj3m%Hxh#HSo-Qnqq_vbr%=2XxNjk$Eincf8*CVJ`MTKe7|=IYID%xaJLs
zUVhM&Z<MbMm{yIH0BpH>`FL2F&D#e@DgC40vg<^0(yn^t=eXF9?TCNc9;Oc!$=;(*
zYM3F{lwk^%W(`W`^RDgwSukZ6KFDm{#%{xh*vzUS;+A?soyzx;Br3QzNUq*q5sm{j
z40-ro0<h2V_KKGD7PX@N0vh;v$Lr-^Yk2Mk;^Ux?(5Z|?^l;ExEW-_!k?{2(5)XDr
zg5KgaUzQfWd)=$E<YHJ%fCG$v1!PCjlOr1B>ujDRsF<>Dj_T+Inp^<og$^EDZ<_?*
zIINJI<oemi-L<cPm$spz-NnANcYL22=u|pdSV(r`0rC}k8HzCeIy{Bw-Y0BGs2Yrn
z1xskC`q=4kc8gr%wC;)qg8V9|5tt&Mofok)Uzq|kQ~ztZIuafmnkwo?4oXD;OM<w?
z>zG718ZOi&tA%R<Az*Zp&WAK-_bM8@>tpja{tsCZprOma*u(y0iTRPaXc$c$wg2W8
zQrw&CjF;K<n-ZJn!1@<Y#Lv$)Yd45EuN)i3a#B@4OeA&5Co44UTEdba+6caLsFmD4
zvdwRPW@9BPmi9Cn`1-Q(>?Qz5QU}y8oO|P9d8^jtz&D7CocVQ6Dt-8$QC+(G_vpy-
zZyfk~2F2xD*z(S5qz@a^DymCv)S_y-uJ4V&C{Q&Osq9qB5w!U%7oZ_AlsvG%dB7un
z{CD4KADmyvd%U9AG_#|g)PQ1-FUZ?h(MVdJ)*JrpKJUq2;k%GoE1PCj+AEBz(JHk$
zxidX@E0DUm9S&ZS$^qp{uii941~|x9<>iUVGn^zpFTa26JG*kz^ejF_;<Pv(Mq0mv
z+4^Q3McK7QZ^;T{)1thHL`y@N3<_#1VbMuCZ!xLK8OSeN_gk|M8=~hPTTkY)+8)ha
zzSVnWcQ!mKJ!MDUNBa2cM87Up?}(~5s3&wRT%c!v$B>P}o;Jb8@RmF?3GM?^5i4`a
z^CA{U;+^VL%SqW3U+*Y$BTAa&p{wTZXg&Hzr=CK#f9Gds%2WTGs<XnhFT3#v-)$VD
ztj*{Iy8LnVqCw5y&2iTyl~Ape+jit~;eZKc*FavOTg&Z6iC7_hsX|t9+OY<*t|Gr|
z@r_VRB}2Y|Xvgn(U`3YHeERLYu?4m+tRgB2YcW!pqyx*AYO?6`g!Mwx)c3SN6fSEa
zPN}9vgW*isHFBW8f3EzJGlKpGAMdaj1!fTMt`#${<K3x9Q;e}N^qikrE#pm+I}y+4
zxH|qiMaj$ZB8P|LaxcD{i8lIWigoPk>4lo+(364=JH&dpQ}vIgN*=AcK+Q2u<13?a
zxvy!D&^@J2@Y+0&7+cCW=nzLNGE{%-YYy*QxK#BYvTftRVWC&v076@=Nx?8}b0M0Q
zX=AG1+uRJMsoq+!Eyqj({o`p9YPC8ryhHL;{rASdyV;8xv3FBuHljrEzuC+vKJ1Y4
zYy^3R49oqlGVuG4Jq490zkK*=ar^|4FbVOQy<VRRm{Y!6l&5hHn76Ys&^KGTxxhx`
zsPQZh1q4jmD0S`B0a;i9J|k@&{s!z(#x5u0!LzV0Z#SM3h)NyO$$xcSe`j@;yWs?O
zIlpTb*ZV`qOEP6*x4xdEEhu6>j`66(u5Ei2{*DkMdX>|2UEDzLc-aGDrB>JrK<f;!
zU69MV`nQWX)0NDdHQsS#Plf3jK{a<Iwzsk5&WF#Bz39Q{I;raokfwI~y4sLe8&S8~
zx!#KM{AThh;At5ONy>bwFg))Qlw=BBpoGI0ehL~G2oy3`O#m7wc<$Fa1HY8H*!}tZ
zF3vALMGHdR?2Rq@LA&*v`~*%LuTFL;Zu}T{{?R!cK!qYmaW-Q9F@D2i7uzv=p0Y$W
zQZ};(tPYh{REG`>IVW<KyL<<|7W}h`6te9sB2>4}*+|tHp56OuBz>$n6nrpSwVjwg
znJ{pIZ|%h_mCYzSNRbBKK2lE)MsYBEy^3%1d}fmlTsK&Y{HvPJiqTXrAKUxY?I-Jo
zZ%JXeC^MSx-8cKb-jWeXu789SCcd%^SLioAlh3{%c#<DLjXzpfNC>~YpR<wgp07xe
z_6R5Xo(Z5@C2K>Yd<-HEO<!LV?t3^PyLM#2w-{b?lpVbf$9aoNp#|@f6cV@?F0fu|
z<=pnhRx%(2<o1twVQKNs8M_l_)4R4hu*qe+Pkxuj+)_|gO-**9Pi+Pqv)j&Q)g4`Q
zf|pa1WH3X~C7fXdXp~E|vBZOu+p*iV@dvt|<oQRj6looTZ_#}MS?iD=o*3{6lcfr}
z5T^+8UZPhlVPHH4mVA_m(u8Dn8-uDH-E4~4b;{Ow_b$=o;1SK`0gmEFzOBMYRk~)B
z-z4i}vlpF8k|U%u{G{a5N>@+4L2d~ah#U|!YKIHE_y;<gV2euI7nR0zwPL+x9n8WY
z*sb-l2AFcbtx61LuSBL))k*zx$LJNo?fM?h--E*re;kbEd@HM5rF5sNcVCi2Z^5*i
z2K~!(cds;TU!yIH)?E%!DU!q(hc>Aq_-h$FR>68EHwE-~=1Na_OHW^t$cwG&X$&Dx
z0cE1=0zO<~4Ybd%nX)N6aoQlT^}g<RrhI@#p|-xZ6l04xE67J@iO1jXhb|h?qf)AE
zKmT=q%gzk-@sTTW=pc)5tMb1xUDmHWX*EuuU8_7?q}1pRCZV%CmQczBt*O|>qtm1I
z@nVgOn(GQO|7(P)7tl>th4x)-Xu|`UvYbMgU3>bcIay!4Xw{7-qMvMGTJON@)OFIq
z`Q`Pt*?5CPaI|!vWi1z(slm2!Dsc199~TUGHZSqKQ=WdGA;?z8rh?x7ZvC^x1hal7
z7nv%<Bt09Ec-~G98Fa7v>luZoakk@AZ}!Z8;L?}NjLZLy-Z4#2i{oS*3u+zNv6B;k
zx4TUMn2?T&ug?ak%0VdF2gC|V$?vur_hwV`R#(}955$s54dX=dbPpc)!Rw!wX}%+j
zQKX!!sD|cD3L3lcy<qOW!AMf2?XGn_9uZNW+>~+Qn_m1J#j$!$8JY`q{w4$UvRsb9
zpaclHR`fL3I_cMiCG+)fBaI_J<QOsDz{kxUIxWsBY7pH>#`VT>n~EQunCo=S_ZRD)
zHW49<Dkd+7AC~D`KmU>MJi|V-FgF)TX^<Qv34^~;Qj~k{vvRVtyJ-sdkVV2!gOfP(
zap9K_`G*ZZG@S{U?H&of^3r|tBIU=q<fxa!!HwkD3&W3&M`K)n6+eG_@NS7W<=g!_
zqI+5Mz-Gn@3yhs#W-lc6U~lnwQBLZH=Da$wU|725Gsv@hiVhg!ni)o=e)OxL+a2$9
zLXJX*i4?b6Q!>4Z-~FbRs<uzBlH;&BCGSVAGH@o=26ZfPvCEo52Vr7S>!N=@<o#JV
z{36D9@eo629gYZ9U%w(V&D3*`GXHTtGJ_&@f=ld6CQ<|_V*XLkDGT7TrSiJLU8xLP
z8@qf*nlEsZI2EY>j4GcH{yKm%&ei~3gN6N3upDo3V9!U#-G^n*IFDw`{W;yFe*LMy
zN@;cMu2tCnsmlu6wh8Bqp1Z810xmE(JH^_!OE=yV#%^xqNEN1X8Y>X|vz9OY_ggHV
zszTK%MPaya^j6?907Yk!48EhEC%X=SDx+HJW6>PrZK%WJvM;n3-|gC*{WWq_z?l0=
z1x;onu60(*E1SV#Y$ru+2a*zq`p-QqV{zWbrEn8!8D4l*d{fu)>DuH2sQqSVi!Q0C
zc>e5}1Sp}&f8hw&sN%6I^-oqU?Io5O+)bAc%uh2o-tx+rQ!48#B-US;^w8S0tj49y
zrpKzd2X@9VYn4iMM0qE$llH*(I*K(>b$sP?In2(fnOB8p1J1UElN@wR6&};o2}pD{
z<CJ$it@cUIN<S^p|MiNvq9-`HvNd_^l*wD@3Z9R&2dUOvcH+d^_7>}A5u5a{=BbVK
z9y~iq=$vxO9caL%yp0{YRsS*n?P}96SHm~TgrR}{*!#peJYR({^foTw5jb0-LlI4I
zl&k%O(YH~urKnpT3!pu`Z6i~Jx8yFrD|2=;OMh2(9GMfjzP+06=q_oJyVf9k9?b6J
z^C}qPsb+&oo3c6au=_R?nt*m1c`_1TayYBu>@wY2RTrbDUgI4!aq1R*up1U>HP-U$
zgDX|tu-0_$Zk4~Qo<yE&o(pv>#AAaAXuI?A@t(`o@#o@7(db;vjnyk$Lfj|hdT?9`
zGFJQR<!NHRkdETovW-&E1Oj>RkgSMAM--3}UERjYJVenQhV%zeM}(yIWSaC2#toN1
z6<DI;#m9EtU_$sz-%NWQmzmZogW7SwSA)sk1!2CU+Q1F=(bKh5qtC`tc_guR*5P~Y
zYQQi)XQm5pfFnUA%=}92sO^e+#U@JyH}LsCyr<r>r8vjA)(=Hqa=1r=hf*CF$~EfP
zUrHhzlLm#xvN|!X5Pkui;C|!RY#DK|;~C!vzPDAX{UoKWA{1$r3^Xq#q{(_!1Tm)o
zul_THPk8Ljw6x$7Mqw=Nx;AX8(P%7N&z^g9xSW%6OIW?^cV>eLJqn5=EUF9XhLtad
zF}TaRJ{Wt)=blU?d|<zcHrp<@23U2S=>K!y2N|6+F9~ap*lFHVpCiq!mt|EQb(gK|
z$nv>TOZR9S_bI-SUQe}~TfkfJq^Cm^rMg{`^**Q8zh8~|tKnNx)(h(G!C8h_f^_pq
z;Zub#Gmk}MrzHc=HCHdR+4clQlK%)h0vwY$ech&&H2RkB%(t$ORB$?yF=-f{fG09j
zzxnLKq>@t_EM5XRG?GzslUx3L+X27#`W1bCnxeM%pN-UeoD@}qaw}x(a7yCVvD=B;
zJC=;DL1Iu^?2V<_u^hIQiT=G_IbVx>K@a6Q({T8txqLus`OB$K%0C{y=Ad(8H52=!
z_~GN~UcPd~4b(xaU<WAn(5^Mb_?|)8iGWBV*3F>?tDFCawYDdL4|}j!Ic+ufZmn5#
zLVY&O=+b3ocps8;yiA2o^OCM%n7$1o6kH)07FB%0*IyfTUyk!2G;yhh&Qb4{YtXJ`
zlR(tBHk*d|V^dPm;<d~;n5xd?<cYR?!QBv*!4F5qLLz~w3pnO(ld6<r+x$AL#?Ugy
zy<bzSp0RSq#p9ta!srbi1@C}Ut+pjn#jk8{cJ6)zCGs_I*G|v<{mn~4O$nIl>7Ny=
zta*Q0dc=%yOiH7+o0d<#Qsry&OhF&{%YS(8zSf!A88KU-8x)u#4yxNWV*i6!?f|y)
z%5pZO^QL-woOAqsnrUAY?>;>c1bx_DBIY~u&D2G6FwtD)4yQ0ZD3Y-}=1MTHQ#Jy&
zhQg1XS_{WmXx-N^{1>hdtNh%rf$H$oWY!EmK)8j%7s0Uccy&~6>$EY4$1Lz%V0Q5{
zvQGMdV?J(C@Gdk{gAk0<Z9Uttc;)J_{{Cn7!Vh6t;oX<CAo<!PVUGb-(uDwePmxOJ
z=aonvWz1)>^!VQ=y7EVi_J5;Tvuh`}Q8~?vLr>Bg#o|AJ6^muyQNJGjTbnC+pQ{*B
zw*3>*+dA|z@d={F^)vgVB0EMwWrfCmsm8Ry+g-}bz)=N0Dz&ZgwVLUh2)fX|w{_#A
zaQW<>8YWndIiJ^yVkv&aB%(q;e=Cs@hd)b?yeQOCDkE`C!9>PBgdGYI-c#L1;vCO8
zUCPKBM~=eT;jvmPS8k&z+{!-Sb_;C~IBdTB%T?O}q5VzP?S2k-4M9VFL~)@IKbMb>
zv4X3O*p#yK=%)<*hxXqlSWY2K_w)CNyLwo({>b>9T>6&VhxN9l8KUK!XPqp6@}N6H
zJdoSFM?Al)WdmJURqUG^Poq12$sY*trBAF0?JNwo^i?>_?#XNs?Mn0EDbs?vcSaqT
zhf}o{-1ij)Zj;+&p_+%!YEjCW_M_t{yk157U+{1Clxl#nkFPRuL+|*AZyg6C+|C_e
z*Uj!oDv-UWVjze)pf%qb%l3Rq>Vbxd52Y--ayowRa+#(HJopNc%dzoFAjRx2c^^wJ
z#No|OfI{%ey_45p6Yx>X6Ra`P1gWu^WBi<cRl0S))o@?^_hA#svDiMt*eQT~nwy>9
zK%>T_#J>&fztdEtU!rl`&zYU|`-%^4-+9;!0QEGOuN0^Ld2YnnFRLvuVkSk!hudui
zjh%TlmGlRqfI6aD@fZ1(<tOBFOL2V7Ik;QG371{E*GYGfM>wBJv6ZOue^64q3K6pP
z-E{X(0*0@8#nO$7D@O=hy5@hbV|F>MSP4so&-?51Il^9_qKCeaihZ|YDtNK@B2#sp
zG0YvG?4_)hzCMmLW)Z#5J)xQ)v_MveU+#CM8%m?2y(-z(lH_9i?a@U@#+<SOpj>-m
zIXhb9USV&_z}!VAe>}H>nJ!6Hq+HOT&-E0>qqEVS?e$ugALm(Vl|t;LGH*_{ZpF`5
z)YlI*xLvv{5}29QC(R~795P23;{$R`MZU&m=bg|Qn{RZqc|B%VRG_bL0H}*o{~m>o
z=taMTT1sP=m0tZxMvV@)v5qgK9V}s|B&8d}Xj;7Lb=TZ}ij_&1`p)mwVlKpHH3yZ1
zU7I8H?6B8&riu*=^EJV1$oXF_JlQe)r_7uEz$>XQGkN#n4a`m6E%SE-9aK3>!TO&$
z?(R1@WL(@REXjIbcV7YMZSC8Nt&a9QXP(U*#xER8K(KSfxgS2~yacWv?hcFr0OZD1
zzTE3B_)7`su(hoGr%pKTH8>B{3(MkYh#_#MKhhb04Ikd`xWRo#ameaz6sI)(HLrX7
ztB7jXWP7ZwsW^S#;T-ZxGc9p0NjT;6Ox?uMw5h{vH&U7U3_A!~{#tqHk27S-U2wio
z?Jqh8qpIj7+eOD1vCy;fI&j2jS^ik8ms4`yh(b|7D$?fY75`7iQI4Us!&nb?X@qrF
zXOY(+>Nr_qR$eFD|IZdqPL7kA;K86{o~Ba+GOT5+=8Y{|XN~pt)z?tl-R+T;uhQYi
zuA92~-bQy`Yw^0LU|`?er=I2Hla;+qe(l*Mhj{PBI{)8y%tmZz`O#Ft7?lN=HYq&0
zi|#bfF*)TcRcNaHX9l1R;C7~}6xM38t}>LbW5ND>lrZ;$#GCcJ*r949>twm+L<LIj
zUcA!K+pRfB*)2hne~^#Z#~zph>}uu>GCmJRIU2_dl@Zy~JX9xW#k*dAXhT=>^)SOq
zh=&ly-e2X+ya4I<^3<Ife3X(uXV%f_;GEmtc=*g`i3%qsZ`ucB>`Dne_W!Y-$$|@T
za(h8c0#;PzWfk#D9Df(fOYg8~WZ-1`_>l&QEpN6JKAP5cn5Gq<)6wCVjdRZ7>5Fgi
z76>!L{RH&)bXU+!vg{l^on1Y$&$=sT8Vgy%Ll4!D5qD%R8EuvCnylNjYF#5sryjFP
zb@tWF2CrDH9BQvSBXxFpBrXKYm;q09v=3DjRIM1x%a^!@hPA_#7Wp3fq`NX+{VK~b
z#M?7};>6cVH=SnLz0Q5<8!GLK&zXfW=CM~i9S{*1<LmT%fCO1XLutxGsF_!2A5cU0
zW}g)-R?Q;{U|oOBclcvBQT~!8FQSA+_v${OizEaQQz%`+3tU~W4@F=<Nr}J(^yR#p
z{pp}6A|$EPUjx5gSh!q&Z2r&s@d=~;Uv5w2?1J-cIiF&w*;8)Oep0^lO3TK3bAz7M
z&0eU<xt@JBqn+#~dtN%|w|u+Fz$6486Y2xgr!kVo#xv{mJDRlai)Cr^Z)~!3AvfvX
zzl6tqRD*=s&Sr0Su?YJH{aWs?ws>RMx@#Hi7*`Z5ga_VI%hAed++Uv#==~@M>iM+u
zN~iQM*dFWL%VrijEX8dF!EPK@tn*IQRr=0Dyv#oNO?_tbQGVxeENFY`o-3XGbFig^
zG1`|JzfCtjN@?tJHt7~VY8>f(YIw0!OZc_(<z-l-m%=4mg!Hm2@p0gj>ENbzA7LNo
zYul*d2dSHXYfp3f#--nkI;-&gvJ+kkGjhL#vr!wTYZ}XK(dv)u|M_p8rXmEcTgp>M
z3wvukSHEfsJ{Vo()0lT@#*;-(idHE=&-HVA!1jN9_xAQ5{;SV51*;tG<@=$i&$Q$9
zE)o1;<QTJi(lQD`eI10p#IVQ`y-QMaLCj7Y-2mnqfnoI~XB+DSz5{M?T6{zn`!J$E
zMql8J8L0kF4X@W%dZz{ZWf{J?n@Z;t5lt^$emJ(>^$N^<e3x1nd9ISVKz1z8-fh9k
z^6(kb-Xij(@Q~)Sw+4XK`24CG{B2uzwlr(vfuGvKld0|CG+q#9gaw%J@x&x;AW<x>
zo@zn9QS-^|-xXTYDKL((zpY7Z*A-6dlMnzg1Ah^;f0}w6FU*B)p8QyOjRZ{b`TU;H
z!@usXFZ2yN(%X)%sp&|)zxVSlz<6X)tXB@>|Es708ZF_(mKu4^40x%<@Ecg>J(&Bd
z6bz`#uA=`DK@eGy{449OxT4Vb7W0kb`+oqjKu*6y(eVH;eO}xtL7M>heht9O3_3md
zzmmb_MN$gEU&gi~JFo#=hpLxr(K0s24t6x{EB{<^YzOa`IP7QjJGQ*$#WVo*ex<T-
zG@q(;aZWwzI>mb6I6Gtl)B{c;0EZFJ39c8<B+AKHi{~~DKI#H}$VM*ZGH@InK)t&j
z>D&1TIP@OxK~ui7Bs*zy&~kJH_VDbdT+nk3(Y33xwq6F2^Ra+Zy1qV^&X%4jJErnm
zGB<(i>Fwb+q4Oa!dUp9eDdzJYbpGae;cp+sJl_l4m)yr5%OD3jfcug5>vXFfk!Rhl
z{32+@KbAMf7Rxe5w@O}?N&36wKm!Nlvd($8iSlux<2av!u?eCn-iPG-9ojPZUYd7-
z#j?{!#&77Dz8AV*rv0_X;Oyo>KF^@|UTr-d)A!_$;i-SVN%uSL+Kdq2w<A7YuZW+D
z_v#A1DARKZu|D&A-;}Po`B|^;_sk#ZVm#A1J%`4$n+|)|&3fk8HupZXOu{++Q=}&z
z+k4JiB|WE1tH^^EnY>HH`)ySjo1lM;vaM>D->FvRi}W9>96dcfe6%sgnM=i?(d!pX
zkL)-f!{k2W>mFaU1zP%b&ymIZH`1YtG(g|Ty>s^dY;h#VQq~h?@Qu>)4l|6w;kTUF
zzKLtdb59oZEwF8bUckSJcGA){XB^Sudn}xvWtQ){a5{(o@o)c9P|?v7&K<csaAE=+
zhz~&G<)z7Equ~@N47sUCncw3nh>1VZ@VSF3+ZB0Hlhal`N3La>Cx_p^J=SFX-2n<T
z*6A?w86V|yqn&Nl;}QJf+05s(`JJ8{MtyU6Oh;#s`J<C3+r-V!WS3XsMq10v@{Erm
za&t#Qe&prl$LaVIpKY>SX!B$^xYkD-EjLaq7>B$VCk_1h$6pryxK_p7|NFa!lLxBz
zTBXgTM1+=(m`^;?#CW7@)yG^XPgzN<iAV+>J^4}B$B5hGJ*LYKupUh;OB_^DURFlL
z_Jk8QK1|ArkZ^QOPsd~9Uuy!beB*(++d*0DTF<eg#g}o^nO_K_%vhJ;W7=`oac<)r
zkz_rlwe02HZOr}fF&*Uhwavria?IoUcvK(DK3~_rbtey3?oz9#v=pcSP97XJe8;9t
z)@$&Ma|0*E?Kfsbe*BgdUUIr9561_S)q7fbbs@(Pyd24PC1=Xkes#nYj|xs4cJ+YQ
zBt8x#2KOu!*kZEkNnHWt(ForJ`oOn;w9<<k?Sm$7wsZvC&4og(GLr#|>yzU;`H#m2
zN$UTYduM|GT9Zc$%EG{16Y1#0$-_5?4D3XL9*51f92Rl`l`L$dd<M!>oE35)Jkc?0
ztb7Buwc~*DLW{G9W8zMal_&jMDsP#{ADlb~T4{$P=vwI=IW^FGruqU8oJ)Vv$v9Nw
z18jJLqo?HJOws}+#rJj8o1G;%g}5hpF`@4VY&Omgj)Y?Z78)E&(Bs5_9$bWL(NjOv
z9S$Di6y$WfPMRU-in7YjhI0qU&Qq1+$DiLXcD14pxj=FFV80B<@no<a=m97ncF=ZM
zy+@fCvjNmKKb#)WmvR()N_MFqZ24tNFI!Sn>I%aJuXw)EJQN3Wrtjf+C()a$dQ6&M
z*jNn45qALf|8%F;fyY`N*K!Qo0Zum2@f^V+6Z`LKb=8MGTt|-qnZV>Fld&ZZH=ZY7
zN(R6SW~nzNu?c5iy*{c}Ed-zrTF<S*-1*~jce2+)=eDwzkQJF)<(RT#Du2v`mcB9i
zn3wg&_7Le>agWd7VGQTpQme>_a_5vI+Z*|blW#fnS9bC^j}W-iHvLX5JntTV{^8By
zkKcSz==d|HlRS8$-dn~;jEvdTQhPq^G-&1ZaOlXpq@Zo-rjJyYKMI~Cf@qGD2Ri!x
zAM}m2Z?rr1V?ut1wlh4JeHAAU_Pbw>?GKr5v+*4$me+bbrjL13aXW|5^G$N{!0Y}#
z%gJTU`MtCkHMIC64!!Oxda^cwXT|uj#%?ZWoL|WDqkS=+>0+Jcyqxy6&G?6=U7ntt
zG5gIk_dYZ|Vfl7qM9hfC_CABWOERZSvA(0@kZaaExo4I8D$4H~IwK>?nNzkIVjo1p
zTtT*LRma|Am&5k;^z`tq7~Yoj_IKww#!ey@oJFvnh(mU4%k^ZFtFnv#<#5b5d;ql3
zmV64z*!>&%(Dy+<aM`9^Idu%Y&)t-F2@Pn5_hRS9_RTxR+_^~0CVZ`hGhgcaoPg()
z+z%Gy2LANy%H`gGs#_6bvi&|~V%MkBbHY2V%8dg?gfWPo1Du6@N6^3&<yj7s;#n?f
zJdMEAL@hK~u6fDp0fRsw^TxE}**?p(lgD(%JM3M6h>yH`8kKV7#3ljxQ3qaPAmR#Q
zz!~q*d1bqwu1Ut{^c#JQ8Q>bwiNt{uD~H}|QVYm@PV=%Jfx3#<Gv`C=fv9DezgN%B
z9}CB1jq#E9hp%7WA#t5YB(`NdTdYq`>UL6SS^2t2i1lH)c&q0CgRj^zXCCUSb!{@8
z(|`H-t)tyAUxS?P*`Dco3V0y&?rkQ?SoNG`u}#wOhc90^Iy=nrGGok-ap-!s`b7;I
zUt@>h09NJ2`>0~){f)BT4aY@gPHc5;WxF#i=@w1Y6M0(SBym1+{G)&2ni{~ba<+Kn
zwT!vp-62}#YL#)WyiY|(9c~(W*A+d!cEtZWar788j|ogT_rUFF>YDBh--p@Ljv99A
zFtK;%J3Jb=Z|NCdzFDIQBgVk&lrb>L4qW2=C<A0U7lezhgb#AcJkd@#cHpq9W_RG{
z7yz6+H$3A-!;YMSE4z2ZBWKTzoI_=o539oE_`ySRb$l#c*rgq)`c;rd^>sK=(6ObR
zL_6pl`oH4Tk&~w!FDE<s=>3fb^?U<}NtJSxfgBF(6!H;t;P{one{nS6(2<S80XS}`
z7l3+zukQGfLq(3CaukBbZ96(L!AIQ{JA`w+alFE*o}uO2JA<Q&YlcJT-J8?J<)tQA
zbcDkpyOG#AG<*w*s^<7Qp1~KQ0eH#sixy)u1Nk_W*eSwyZgAq^WZ_5!UO<6;(LZ~t
zJjvk2r+jm0kWuVW{Hb;%QF6X*!%mZP*?aivS=o8SU>^K|96&sCY)28UKRB=WVHdw=
zI?4gS(ZV-tBJjPWy<uF5lLtQRLB7AMb0jAowjc}L*cP2H^$KZHZaDgnqtdVq=ahkO
z_8=R3vEx|JiYUW$mgVEtz{h7B`Q-7u;)OEH@Jns16YBI(Wu!dt@9VlU#wX47K?d<S
zrecTT7|a>ZpcmV5u=ao)=Q=cPPJ^f1Ej&2Bv%{o72GbL^?B;kqp4}{PLq1^(;6+B1
zXSy+5q+wj^U0vYk$g}Jy$8oMF_L|1>oW^GQZqjGk^;bfb@%1k!7I7Pr=lU`KZDu3l
zwwklx<;R%8a!OjZ&<Ai-;<@@V=OKqS!j4D4_>vg@#Css(iQCCDcZ_1Y9cFo}rg5tF
zpr2DYdig9&@U`g4o?}0&nRDHtlV@<&je|G>{ne^KUw!e5o|yr9<;eO2%Y&n~938Db
zAuCwb7Vat98#`2S%2CI;o~^^O$#>j&;BvdAe+K9)zSmA=fWF!95Cx$>S(fpHy*h98
zZGMJ~$wMaju2<54d26cwH@?r?Hk{27z@y{M$@${VpT1lC<@NpdVB1KY_39g$(3zGr
z@8=l@z6H6AEgIu`@7c~rr>}3c$N7Xke583uJeFtEy5g7t8(r3DZ%?P&LA38sV^IUw
zEPkHBa^C9s;kaoX8XxV5@=do+I3)+&xjt6aGd5Q0$%=Mi0}Gi}fpzCTT-WG|@u(X-
zePo18>vj-n9EX-;);u0IFSF>;5kp@v)*o0NVO`|4tTkfGRHZ~#?^<9(riTl>ZE(Ir
z=f|M}J2iQq@R%K%4D{DMXG*4Z+B)szF`kfniu#UY#+bjUxNXSt;(eL(dg-zD(c>Az
z$7fNbb6<Y@HjA+o#ElMOd(6z}JO)pWn~sTzh;!Tpa++f{vRu-0R_oi2SmDIu%>fO%
ztd}(46%;gD4-BA*vY-REJO}=6eUNXk4hH!g{gHuumzg{&+-ZkCYma>6h#SE1<5dr6
z(SWZvaLuPvCe8JvoadUHA?<P!2P_LbaRzzhQGU`~4+iGsGpUg$1TC+8AXzk*OtL8C
zXPHb$ztYE=4dfZjA00#F(Lu9)(X#w<A9*b!hmpqeO%wBsb4Z?TrX%F0iS#jU9_rWl
zC2ws%traTJ@~wmj2RcWWk4+n>^L%ePk8tXNYkSO-)8JzV8}o_#H^%Qg%FI3G2WsXu
zox`4!gf?e8@->TT%HA3vA>VXUHt;}*_jVkVvl`edylpdK?ptDBjQhTdIO4XAc<X^c
zzRMr$0v!D!_3M0xQ6_Q?p8O#FTob$4ZXVMRMqHE?@uqurI}yC(gU%n<Bp%BEUk3NL
z;IocQOZs}~qgr|B@{j(7YZ_s9;zi#23AyGm9p$VVM^39AEPqVi9GWryhvY@4VNcE`
z?en!m*KU|J?&%)G&#`NaX3nCL{M^HyC^NS|8bNXFX!MISfn6&&L=5aQU?TpBoB`#C
zGYMlCkkEw>r_7G>IWp}?yH(gK&#oMHIc&-C!Wq%$+Csl`__W}$W=l?+8##JVc;mC6
z)i(phnWIUTa;~&+{>~4191HhEvbVHCs{Rgd9V~k2>x}BtLPL3qBgqAqjo82$(wCat
z#OboFN&o$)2P)Tc;R-h6ys0~XbQZH{s0(c27`9UAVxxB63~viK$|yTD&vlF&{ACA^
z@}KGWyJENQ@s~Op4~G#ujdo<;w&>YaME%mnnCu5Qz72S%^S7Uf?p*rm=SzJn^if~*
z1&nw20sGJ><cAD)8?jp%J{$_{JUNt;3q9a}d3{pya0am(2=y$M@eXIs0f!fP0IS5=
zjRSCe9DO&y+02Z8Y)3)r56&ZO<eA{{Gn1C+;JQ;6l*?tikPLLOat?d32Y$elorK1D
z{B8RR;qV6<CVsJv`a`f+SaMt~cx;N#R`W-B$e^5r`Y$p3t6=#~M}BZP)~S>HoW(Zs
zJ^nn>bC2?7I^_NC<?`s<mLBRTg2y%yN4aqp8jxdI4iN*L!8Fb{o)BEV)}})a&+y*3
z$8^0hGB5ws1@c)f8UbGO+g_)kW1`n(U{WK}M>+7ai>6-BD$|<aaU7l)Cmm%jTSCR}
z$X?SEI!^oHoxy!K-#{ae_Um^M>opJKRKM$y7lAb6eewZzOWQ7oB}0FZ$AmTY;<mxc
z^5;0kbsyU4xYQxX@E6PXAIov*`<St<<&j66Z@vM3zAR-fCteD{@F(VjBJDB}kNGhU
zZ3{TwF+!B*dL=*0A|3Oi49YvUf;iF=Z^;JFQH`y>hP<qs{wc~KogwDuJkl&I{5yRU
zFG8yv^f_67e4opDNM{^6z}W>7^BqUWxt{Z5VA@_7d2J)*q`mR{qR$}iHflNW5qCMs
zcib|$Hg31Zd8ZxWw_ntLn0uCf9eMsC%erF7J;Ae`Ft=;>1-xJ7e&4p4$M#sZ(>cuf
zISswbMBMGc@BXGY{-uuMees8H7k_wF7YlOFL&FanDHnC=`WEUQK<DyUK56&UX7E_T
z=~Y9XI0ox+nh>33+%d$#tphFDV{&@*u9F^bkFQ714&=xa+cJ3CQctI^cjR~bj5zCF
z%2__|oJSaWvK=um#v}h!e2kAg?DN=gN|y8Q%j6!vk2QTObQ`t}Pt#(i7LRf8tZFm7
z)@3=%Yhz3enK$xt1Rmyg?X)qa+W~24A|RLhhp{Ta{R<6?&S+cwNV^|3ze9K;faloe
zs59m}&KLz~$?27qxo!HS77sXVfBX8lnO~HJKey#lx#4Ze4CE!+qjCRfdd9$v-+SdS
zT?}oHaamVRn=Xg(zRY<}XIZ{h#^*j9KId6R4r6|N7DYPG_uh`?{D8?76@@vEmt)5f
z3Jt%ABhJqb7DWpUTg&o<C*MjjuO|?p2bd(}Y}g(|`=Dv*jl7i8J3_F{wq!Zxp`nw<
z0OH3}9S1S!;KIPqkHJ28|KPP9#KBYHuZMC_UignW-(GnfO~HydR_SCJ@bh6X=v_zP
z4AAqk%gY$~ycp8(Tu#TKhn5#IgXIzOFoK5$NSfbmV^k;N2|@E%#%*1Smq(TJ`6dp4
zGs7yKXI_WKbK}uLdm;k7C)!P8o>mxXtkZNck82+D$s<ij3AkxY&#{E8Aj{CZxoHoU
z7kP^8^s-HU{Xp`Pk_U)&M;oD_Zj6heCnK|7%UMdV50jX7yx6Bi<cV?UtTXb&>tlN8
zOdCV4FTexMAxFNnI=Mm+{!sQ4R-`w;I?#_S>w$cUWRm~@KmbWZK~$7HI~s^b8$Bq8
zjyP@EvcRnZ?myvSAi#@S2GgL!dK}WfkZ=3U>j4WgJ(w_$X&ib$=`>-M7xRcmS<vQm
zq_-?W`h!zl4<5LH^GYK9p<Z)7eY*Sgs1w%Sj-K3C$LkCqI`{}_2JT_&upNXgyWll#
zl*9J>2ze!9lpE=ckGM_7a~({{Zk26{=OJZ4-w((4S6ZWUY}L<{P2~6Nie>b3D9WSG
z+$a3``=1wo`tuJZ+c^05he0trP%^~l0ncuLiDq_qu-hg&cFIl^_4|i%+OSiG#&v>|
zL=*5hNcj09p~zvk&pEqC6kbR-yHVUCIkv6PaHQI~cFMAIN3yGPgHYo#rEk=iT%EiZ
z&1s3_+__PiKU{Krn^qJa=y)zY(?5V?$IpdAm7{z}ROP^VLVj`ZNMZe!ke-ZGIQ60*
zH0;LVC^sBAI9Vv`mYhF0q}YW6>}rC8vhxidjx(cNIJin&N4h}=f2kYMm$FC^A3i<f
z`!M2TXAeB2q32k)V{~gUgJX<iAW!80+tV%=c0;ks|5&?q*kQ$PAAlp?4z(NSgk$qm
zwo8qviUUVwVnX~(@^Gdc%OM1uD!<liwdqht>*3U4>>Jw`4hrzkq>DU$Ww(!PJ=OI>
zCUS_M4bQq0*~Si}a`c+?&_>}SpYfU7b#?9(b8a0*v^Rcd;$srOIEBQ=K%O)|j)vpI
zt?)B;q<nm{2p;qRgxF4qJjUHFm&X%?^qtTFLr~(3c{z{UKwr)n_GcZfzz;w!_lo6G
zPm~ehD8rvmYd&elP0087pXVWbwvBRvP*I9r=s0eVGV#pEM#_sm>XmT?W#hgyxL%#k
zeJ9V~I8o@ajp+&5MO0s^sx)--!%mzdOi<Vvf&JtG*6pw$hd*2^0Qv~g7V`MPpXsur
z2pN={`;{N#X~q$L-n!h#@;9Qm9<F=xkJVnlQ7$HadCv1L1Wz76qtD}B;@F4ng;qSo
z`B^{s_^ai5a{Deu9@o`&Dp>%09OIDoI|^}6{<E7Yg6Uj;F^~AJoUmM1fOjeA`O%(z
z{m28j_UPoEG97Yhf3^)gKu8@rULxTSe3bFWKmGOgJqH}u>QwzQKt4EZaea6kSqRb}
z>ezLUolGFxPLpy?0n0O<5M2P}<GZ2;WVo)4BXn7f%W3eIX_{8pqG`#37kthmeZ^vj
z0LX07n#8MJvn=Nm($;f5!IyE+wyulc=jcbX9;dO>Kjd)zste?9IZglFn|Gz3Z`B$6
z0JNqz9U;#e_v`fE;7;_tKma_y_nNq4++n_wIs7q>V><z-9n$xCd<4#Thw#KP84gX`
zNeF*TLr<E%n0F5D4f+~@dy_n`uIHWHeE{}ZKVfdW)<Io!ExAu3IG=k3e$ufFF^-G~
zrU%EE5?<WVM;aTsKcPI<0gnC-p#FK+h%)JWamE3D2l9ItxLh{(L*%Uk){QRPnB~Oh
zDtUZ2FV;zqFUuhv7eTl3jf*l}4^gM-atI%FT*fH69#cHfGglW+A&ESCoK8Y)<9@%$
zZ1f2R+B9w7dK{9U`HYKro)2EQ8hI^$ozUZfxxBt6wkz7PoEhxjM^TU4uw^li7$M4z
zacJN%U*vJz&p)Rf!b97S@?zX{aoz|X=63D0Y0=x3n1^G=c;pgCE(^1`pL3pR2s1=l
z@}izdTgS7RdO60(e6+{0BNCCsz3}wqQL)$jv0X>~v0U!E$#=Vnw2q^rwOfVffJ*{D
z>7!v0hwZS896BF+;(cm)E&85~mIaT;1mK7}O&)W1o?jqM9YKdqLO;_Ky@i%NQC}<n
zmN`Xk@17o03(sT%9$MeL#9K$%pv(LY!STWufe*(TeB{J!S}*k<boO@-EQ#0eXai0Y
z0G#u%*YxBOx0lm+JcEuy<5m8w;^T$Kd6CXI=v&}?j?o~_ie6UOQSOq~OFwBeo7Qqj
zD@)tbZ+&vkDk1rlp~ZKHP>D!GJYGXhITh{qC1ojh)6^}pJRL?H8GD+10j#6aMGH#y
zoj<pAs>^Lk)}=1pE_cXpY3^cc%KW6V47}+vADWy72T%@P{#p>RG0xys(^>aen54*%
zuXvOb<K{CihH>?s@utsVT-iBAn|bHbM>@wH(hhd?Vhs?|m*oB+>vh_zh)7#rj?;!B
zK<9g>$CK$}dYw4!CAYBH23vNN<Bx(e(>SEAJ`6=OE|A|w$MQ@RX=0rGskCJf((k0M
zJZ3lJmfz{lnmk7O`dfw_!p<jjUc_+^xA2iB=Et~ccK7bLE9X5lZ0VaTGq?9CzK7_V
zlHJNjUwpo|Xf}<DW#mKh2H%6!nfvrz?LzS}V6lBM5U0`jIc6Dz9OYv0qXBu_a&R!I
zUmn;H&<}{VI3$#;N2qZ;8Y{U(r-vpxG03mGd+GxMnMF={VCV^2(MT*dLHk6<)Hs9V
z+KQ7$23tQk@=I7+KXQ~t7(bNgSI2V59cryAFRF?xVvuvA7kbi?#jN}tsOf@#g0AWe
zqRclp>@@P&NzZefTxW7l?4px;t7XA?^Z`diQKh<55x6$&=qcY~!R`S{S+Gec(53#A
zhZD%agg@WQ0no%>WC7(HKq`;R2tRVzr3BP!V4Pc0N~=r*p1dlr>Mj)EjD$|Q-NtIZ
z>RP7YjJV?ZffU&8B2zr%<E$Y+6t~cpm&1XP9ZXEdLQnlxI}4fWx0V@x%11q*8=<#)
zg#&D}$GPCjuAm`YXig1E<ikRrj~$E`m-hkfi0@;l4!M+~1-KO0X`SHwoU8X$Jys4p
zwM$vN0}5I2@#CGKd{+rN%FK9K#+3eF@u4WiQABw>wuaW@d$HbuB6nyTWRM1?#v(k2
z=|`|wr7rb{tv6e0*UH;chz)+{xE<4q!#dDo0EeA1&+#6w^DV<+#*_B7AkOswsAtpe
z3_ITVHdJ*u)$35kFatfr)Qhe3wW-qTU)lpZk1Q7&AU>a-?An^j!(aEw1rxGqv+SyK
zJuW%!#(2;3jwkD;%^`}kad6SbD7kr%ZJh5_-w)99&R2G@DIVpJ7uyMWl=DtmAQJfg
zuBp3!;iHVl`}KRc<YE(}6^~in2B^=LL(dXzA4f69F+SzX;P-`iZA@c1r1#{MH@H5X
zUL}k&dwiRqp>K(BHx+Yg-0kLm&Jsht=Q^a7a}Uh{U(YxWiF_?7^jU8At$JtC1no8Q
zx=lyf#Ib#OSL)tp?m_p~li|G<8H{ZSX&35>^i4iZU8-GzvmMsu&|^B6&v<M<7Z(t9
zJ0CpHr(CDbk8Ky)7MRxUcbBnHg0gEm;aW-eQ}KR|bUbUU&ms9SPMPC%ih1<y)DhQ*
z=caA84ru9jpXeKl=3}>a1p2crE{wqTI*fENPW|3(A<NuWKef#5>lovs?cB9GN^HkX
zAD<g-%KKP9*Dd2d#a#^b&a;_%o5Mptdh`FccPC4lB*}f?qjFy=vzA`a-2lkL+~!2#
zbMVYFg%F|0yv01!JOXZ{P=q&L<}d`%3#x0&eJ|?I{QE~}TDpHrL{?Q}1e*DUnW?F&
zsV#1sx|!DtRe|tu`-$Gf&u(8g+}QZ8Z|@Vx9na-XP)~1`y%wfZ_Eo~WMz;?)*Kb}x
z*~io4<~A8U;)b?|yFc;AxSg;2!`q+7;No7)TZ6f9wW}_#`{%y!g6pC`==43Fe=Dxz
zon#JN_=4H*?$a-vYcf3>YxrIkYw=``a<0f;-o5Q7-P!H^;M{Iz+O+E#pYZ%lZ|NHP
z9^aqA#K#RCcS}cc>1&fGFL}owS@@S8e8ET;Z=ShjJWPt=`q15?NZo(Pk#Tq}<^WDx
z%+X=WyPLab;sE=(<t=pmjf8t;nd`vHx%Q?^r{SkV2OqwBRzG-eM?I;xS=XSaClB*m
ze2MoaS2_rEEQo<)^PYTo^Mu7Z`Q~3Nl<!w#<ihs!l5SkO#mXF>#LaH&N2;M+r1z*k
zRy?1W=(6IUU2pX}gsgo19`DLI9|e29J7WIYHw0#7rf~Aub$r_D?GXPY?mRfxebPB6
z*m)v9F0_24-h_Lc@ET}$Go7$W-wZQm4tM$<*Zr|%ciJt#H^qKV&|abg*EcTm_QoN-
zmN#iXr0Qki`OPyNe4dRDIe#3Ce9c2=Z-db*sS8`+bbS1A$=k|*<jk(YBm8sR#|SIe
z91Y!)x0U|bNZLK<7roGhl$lRB^%|<2CQNwCO<$DwS$6RI+*wBNJNNPZ*l}~2{MhMT
zMt)9%UWWIoJg&tGWR9TB@+`Se_^)<XIt7P5Z!FIBF0Qw`Y!I5DyNf$UBXkTHck7gi
zE9w|P#5wWvh40$zcBhAtqiDy;mWl_iEkpaRUMv~QgZIXb>IZ6)c+JN##|g)3iA6v7
z^dY;qny=Nu&oyccLUBupBC9^Z+x`|b-eTvXD}R1t<GCjPu1$-9XBa?A9aosy$tJ_D
z&`@pXdYthv);QrK@A5IeTeQ1MjGki=U3jNj{O%N;@<A_cKUvCmcODdeT>Dl$0rV-m
zahf9=)+JB*n5)?=wdin?hdg<IP*=Vw)n!9e7M9*18+V^pJXkgjm;5bNwHT|^x4v3@
z-yE_?Pg_7ge)Xr+R9s^b4`-|%r+FX?=sN1N!gLT%PBxDg?FQ*mE)^&K!R|^;^wPdC
zq?JsDvjV(DIu<EIpB=BbCH3f4Cz5Io8pugUu4i}1sYvO1c9iS$I?0u{&@5_PS?9_V
zxXf#%(&dsrUUsU)AJ)7Iop|!i8|5YaYUXJ2W)8mSzvv*1F=wu~Jn?&L9<Oig0lZ|Z
zquVQ+fP<+1to5C_k|Q*;YaG^(JgGTe#&qd<qtnYh#za@&JQ-`EYSR@rL$6bD40~&E
zYkG@yjb;m{Sx0KxIO7RWR)N&7jXGs&4Ip0o5~mEw8gH%c_~hwa_D+yH=R1T(n@~P}
zX$#>+pFX+s@>3O0)|qp>FgfaB2KA8kjV9?yi$A)`D}I}>nUjW(%rPEx1Y`jj@1GoA
zo{P1-c-F%+JkuX}eDLM{b8OZ-@F4AHgqtqO6PoT{o`;Jzx9i@J4M3ZEi0oa|xq}9u
z6StAwl0K(f^Mrma?iAcUIpWshYXmzbiPlQC*_2U#UQpVYo^e|e(k9mCKx=s*M}Mbn
zieJmomsKy>=uW!Y@tVA(A3Azi&-Qg~<~Z`C4|w>XXll#_c~-(7H+_Hlz@+V&dHr(P
zYY`WG__)R=`Fl}&1IyEKQJ-n|oqo%o_Un0{Ve7_?&_ypY{I=FHTWizslZQX`)bZkU
z?aINP!0Q?_^@pI-;m;m}@h*ngL0Eq>uHZonw81W8Alsk$gi?#z>~qm1jkw41rBCVf
z$NDdN>ti{VXC&17${{q`6&?E=KXJ259@<~tyD-3anq&48@=W2&`l9Q}-}?mN`P2&D
zV?n+yQ;v}@{fv*nZMFTx&u-f0jQ_F2y<C%)7f-feUt8^YnB~rW>6ZLs@R^eh+Fj?N
z{ti9Y`(&pto@wTM%<(<XJ?@xi1^sMj>21Hui=RF{&F&9Pca7`&@^Fa1%bNZ3Gq{z`
zzBFF#_O?Uwq{ct~gr%v^=;fJo)257tfv!XRvm08jxP9|vYpU}RGJ?8*Ex$oD1F|-n
zL28Tsr@-gPZqiy7dUXXrUE@K&OiU8L<)$3wt9sC<<CEtM?C{+me)xT`k1lCPF8y=u
z!9JPd!m~xUH39mfCV+DX*k#3^W7!ES6B7~TonKG88_L2X-to!#N}t4ATqeK0-V%?$
zA9EG%;Fw)6UwU=ua#GgfgOA&J;-BM^PF!@O?S-M-9@o&rJIkBHoxaC)f2TR*&qm52
zx*sE6x#Ss&{6m96=+<0EBbDtRzWQ!0`aPVZ<vITTUPKFQ)zcQfv2J?9n>Hvs*PU=5
z6bQ>TZ8NR7V?kYCg?}rZ9PN*6dGg`sH3-cd=ef?a-Jd~pz8f<0&0%!v^fIs}>r^?1
z&DO_BbS!!6SnzQY{*`Jjfjo2Yy{@;$zL{V6P8mOrfo#&gWIWk1IK{&V;A%71xxId|
zT<ziQ?pJlim_`4)H;a~NTruW)vKW7c5rYwC-lf|m@T9IEd;7euf*Hp&Dv!09eZ8Ki
z(YF+ivFA;F+|{?P-qn%uac$c$?)XMdZH$)OcNPDxo}YJhRmgIETV0^xYFZmcWvF@g
z^3~b(GD^L5Jy+4w>6*73cUO$K*6sGE^=%({&6|%kZnG>5V^Eca>t9^UCTvW{%OLUi
zyXD%Kx4s2X`CqQ5t1DQZ)|EA76l1`7{kpC}+g<bZ_190v8^(+AtDd}Fx!kUFZ<o=G
z?Fl^{$$MT`qm8S(7^4Q%RbDj@6!blr)Klti7)nYegOMM2Rq`%c&vMEJ-Urvw$cNJ>
z`Y_pkS@h!I$CY8O4y*j;4P}1U#gR@&MhVw6&~@lrICE9=BJ#Yw@5Ih7tIJhA#q8TY
zez_h2{mWN%J<~2z#+hezC9u4`t-OvhE?IAsXQeT8ETMR8rpqUsYkc&>^vksZ@@tMp
zPat3VD&Kv?3ph?ayxz5EywOd1<Yl~|lRTQMyD|Ztbje^WGN;R!fDgDjBK^F1uzZU1
z%QZLnrQzg{1(Yz5e*6pO@ZV$T^&`r8jA3*tvGWxWXtqY*IgR9b)^4Vo<1fM~tqTmL
z=+M_@Tm7GOa`nhOKYLZ}By*Q0{k&YG<C<)7iC^P|dylUXzpSg{Kn(69F#SQkrdua#
zB1c&WCr#@e{C$(Rmlrvq-I58PG}dUh^joR1k;GP85ZUD?P5AB??l9NqKHBi1Nq@4B
zek~9C=p+9acTJA;{DyujLSD7al9lLz>T+|1^cHRKTUY%gexKXxY^LkF%yHUg;@8T#
zZyMp<YG-qMp3_+i8`oa$jOKZvV2utg=|)$$zk0ZQgJ-KP>udb1d2DQg&OZ<ve~zWw
zU?1qTA%V5B?dy@P^n%CVWx(fpHu9B^a<I<z3twC`9n@ET9nar+_<MM!-DimGZL;g$
z(^!+;`MQ5j`<QS~_t@vK%k68XHC;RHss2Oq4#m&u{qy6eGwOr7J5^=`_HuMJ)LL4h
z?QZyzF89>FA~L$%%lP{hSC`>W9y)LOm5w{rKg;cV%sua}%sVv?y{zYFX-@YTy%}5d
zGxX}Z;k}H!S;m$QbhoD3f;Bl`PuJjg-u_&7eDaXBK#x1*Kele9k#xGuEuBJREbe^A
zTuGmDN!|DTzb-5O*-hG9ah<=%#eZ(-iJ(O6bF&E_Ww{n<S<vm{y=+|4+Tvf+zduY{
zd>!9q{Bq^ErTDH#_jft*uf?z7LLYZ6?Jw_Kyj>N#eXhot2D+9eTz>oHe2loBe$w<X
z{PO<#he|C@zNq}JL1dxn_0?(f@W7=EdVFNSDeHvec3S?v@v%iS^Yk!1{qja-kK6K}
zlFli#J-y5Fbs1OnUquI#Xyy9bUwyu+MsFY9g4>$>Mu$${1#aOx&6fYNyz>HmO@8$5
zacj6k@y8B?ku&$J@_1-Lh&Q^$jl5q>eeJvYw#ZBq+U}~;tH3d7Y{|dM+f!sFZIgt#
zJh#&PSNw4$kY_F%&c|_GjVfu5E0A_q-Y|mvqCUob@8R8H0LaOkq{Ik<CWa2DVE{u0
zqmOU>j5q3QgPdW-DIFm0niCy0x%bVTacx_<;QH>Q#|YG)Yc)=SA8He9oWd#pxJshh
z-8k7(-}xEe@ToU!0pm^#PH!0lZdD&!=?z2Aax7-(lq-o`!^SWZ8ZriL;eB7nZ@D^b
zDIr&;<@-HveRpLCF%G?{G@SqnI2i<9)OU@f>4WP|<D93zvp9_J?Udh<_}jh@qzsGi
zn(MU66Z#>n{c%bq2A_t0&--q7O3JAfa>&iKX(1E`Vweg3JvR@1A+wSCUVH@@om>$N
z)P1KDr!2>hb*n-8kkOa#(z&^7FN}^cj$xtWuWIgP;YyY|1m;5H8#mP^1B?~9zDoLl
zUO+Otq(*q)O1QLqM}pjM%8(P?);9j4mlwxu{$*2jNuj>SQ{P&kf9j|&w$<Heca`vx
z#-fdNNc7tGrrWvFAHm~f;HeNDufj)$?@G-ZJedP4$CUrRbW+|kjdrpIUHACouF=Ew
z`0fvVbQaDUD7ZCxbip(Ff#3i6`{4#Ju$8w#c!fVbjYDE)9fZTbRzF!|ACfEHGap%<
zK5zA}#dW$<{q=+omN`$y%85R4kL8ozWz$R=J<Vg@%*S`k#!cjZOgFk7o9AQcx{gT$
z&NJk<7$)9-IkNNcC#}vu({#EX7v7GGoVat`8otxSza~3zac7!YcKV*sZ@Jm;n&qvH
zbF&Ql5_7sU&06>}JZo}-n{^NVr*w~f4!cen7dq}^mra*_75}dvEGITo)(kQkudX5&
zZ)D0l<8ad2lAAILU#DBc_4wJpMj!bz?Y{6a&&xjkpDNB=b_V&?eq->S<F6VfKV=$s
zpG<YKRxkT_dfYK-bow=am$9bH+`o4EwdcC0acIw{%L?By?Q4zioWIsdhhS}=c^}VO
zx$M(nmbFGZ<NlvN{=esJV@1~qy8oPN=Ske1y3>_E%6fyr#^gL<oiiW)ExGXFnQ4<2
znmG<`hR~d{0Q1IwPbV@nVHWQ3@h2_u8T>=r<AUoh{~Gi(6QA@F?(Uq92PXBg#s`-l
zIWvEsh}1sQIo#P@3lc;&iI>GjmH>%e+&TS(x$}T0sp!QY`HQ3_4n9A4{w?SoN9dA%
z5A(yzLC-t6zpO=kc;do?gyX`c4ZEwce}K#8&>E~^ySz1?j_Yl%<JbJXhE7RiP1eW4
z$^Tr>WW)DOo2^q6hw{HxzU220S_+3QZpZOo2G(@z^1IAa>VsVLv#iK~SJzy9iJx|v
zc<F}MB&D}gar+>AGtCwrow%eOH~e$>`n~(-w4u4G`)vByDtU9dS54y-{+@RCpCaR5
zGVQ*$^0D%D%Gf`r`KN#Q(~KJ7S?|xw*a+v1D*d}}$+)^?+{b?B*4Ut4?qIL>Ruz4x
zPQifNb<LZTJa_9v%H0Qd&c3MYdwiqjPF*#|5b~-{0x^<2{QT}P23=!FsQ+m1-Y)~n
zZe!VLoKMPF^Fy7ou!C=i=_d)^*2$ggx9j9h{qQq3y{R{<xYzI89oMyq=f82kbl4dX
zuGNbc!{9^d`Y}2&!oUrGotAJq>2;kRQ2@RhaGjB*@-vhuBtz1znk&Anyu2-=)`Q9m
z;=6b23~(6}xK8EVeOUVK<ceoVxL4()EN;~mGWRN;8{L%OxDw8ZpfUu>^Y~^>h4IGd
z!Z@H#%wHIvT#3@Jyn*+_!)V}CN$ys@Q+Z+6@KbR2oH8o;u3}Q3WcfCLlb1|C<ld@!
zyH~&4W#Ab8-S>X*-z`5J;G%y~Wy;@wWS2@q57z-Y5j3t$tMui?_jBY=I&?E1yH&Es
zcrbKb<dgH=KYw3FvFqa`4g;iX?4(s9$9U@lzPHP0=B1i1?=J1qrY2r<dSixf9XLTW
z^e((>+mx#|#wa97^mmI6&#!7z!|ykrKREMp(l1J`m%spd=2Gaif6wPH>UxxnJ@O!Z
z`45P<P*X?d2-3^%UVRhm(`rA$;_`hTbqeG)HxDeB?Y}+EP-PyMu=z*m;wJ7gF1?S1
zo?hqU|J|>?_+W!)Z9cV@)-iO_?&*`G4&lt-`x~vHU6UKVXIbO{GCQD)oXn%*b~^rC
zh4H&zel|y)x39>W38FK6(o8zSb2|us@{{nIyOk!|@UNxe**O*-+`M?{jVvJDyycm6
zmYA#m?c>X-(&!g`;+`sFi~iQ0YZ}AT)9n7luetiIeUSDsuP2V-*+;XcU;4^@d_69@
zbzJxNG+c9==cv=fA6e=!?X2^M?lNxV#O*TR{E}vmJI1eG%>TFe*U~&juKkrmyjyA7
zmoWYw1AX1QFFhw?9@khaze(4Km6s=uNefNtc_0108&x{rmaLwS?vLJG#<Bh`=UAT0
z(CzEfBd_OUE&~P`&spTCv#mZRa%LIQNgV(8H4Ym;BNMIze!oLybSR$8tu(|P!{7Bj
zmIvJ_VDIf?gSl)V`HUO<mb<Uaws_#lPa1JMy*gE%zG0bXvX?Kvv^%czcR!lU!ID<^
z$ndL`ZD$$&yqG3{KXa!!&vP4&><<KPAke!31$18hTMrL{_9uA>->e6Gr)jJQy-8Pp
z9{<d=g`er@koxR#_`3}LSzkEi-s%3sZ~l)V#^uD<|IdH-@Bi-+sy{Px@H!EfPWaA>
z#qG`AQKbmI-L5$yBY1<1LCQx;0~W@Hq1RPmHbZ=936I6XliIYfs51b16VFlX`Ht<#
zCNF7m8EJ0p^patdzQl)?8z4t~K=^qryxrIA{&DI1$(s9^y)pNuKIWzjx@;ZFv%0nj
zP9q<S(w$H`DJyt#m7ToFmmgWuaFxrQx~@psJgYKzQRzSa=IPnH+9csu_J987Np05D
zLcL@JJV+^fX~ZvI!lV2u1L&mVIvV*DC#(+8BtMbMO*+pj4|voAIiF5%4CjiQ#a&AL
zQQb*P7wM>ab==Fl*UO#Sq)I)YnL)ddwz5<f^rDk8bA^a}z`v?ZCHEW(ryO*A$w%bA
ztjSP5?nS@o9~~@WQXVPa#6<>IhcoUOuC&Q8P~Y6O^Q5WmyMiTY$Y|2;vex8_?`78M
zyC2P+)+Tx!{m;|hMHv2a8sU)qL-9m*8h)^qpC$I<ugSMELr>2xE4rsHa4qqljEQV7
z$Iyt^Hav5oAD$gv|Ii`z6@9sDc?e#=lD3$>CiLWnF7n{ecN%Xf->-#vKwS8O<8~RL
zfsd@TF*p$V&<d+-?O8sR*V|p+g&y>DQ-({-@;9=QX3~h;Y58Y-_(P|E(+8qcmhLu9
zzv7vN$Mg8n)B2Xi`kTaL&HBJ~_~W;a&+gy7xF$hd>bleYl>WIae#*43l2%>(Ro83G
zHG36_3-8C5cvz=n>cCY}Hjp1bdVKcqvj;<H`XYf@@5j47Hhulex_<0l8Ds93j7N{2
zpSjVBam2@s{cOye_mJaCtY>v~*XuH%xKPXqnzuhJDKJdX_hlJV_!&01PS4!Cb#sg>
zub(_Eg9#&#tGjBTExF^=Pr27`+^$nZ%LdiBn$7nqYK$DGco+|A9CwnZ?)IuH)he#$
z9mAD=@ed0K9SSek_v&QQ>$-O9M&;{T{f4!m=x*P;efF|W+t7XJTsZ7N_*L;R&^WpA
zvhpe)aLR-%@T{(Q8+j{(#k<mn>^o%yvVeH;q~@N*<BB_S@0CGGzxKTFWIFwGyG|jA
z^O0vz$)3F^<I=4%0yqsKEY6j2cZ=WEYlRxd6sJ^t8|QW5%7!t(`0=936dyW}4=Lkh
zco|TU#&<M~FwkU^iH;v1h^wD7?*ilp=s65JwFxk)vLcr2+ZG-0OGBQCWNbk%4dAbz
z`KLJ$9d4ANQ8Yu$#c3SWOZAEKL*M63{N$L60*0j}PZ*c9;cX0NaXHGb;`Hac(i!@o
zGtV`r7U6rDp0Dq|dUW=-j6>#aZ|WhrTkTj5d=o+$*{m{VnWO#i$x^0tz@LvHzbqPM
zG{%mqJ12sqjh^mBKJ@8xGx~tc&6m(JCsr@!l9@Ag|3ddk&EIQo+wvHG8?c$rfBXHj
zvrlXMQ6EE}$qkQq`IK(vEaG~96WT5(xX51%lli8~olbtr<D0J^pS>-^o@*wvHb~v4
z{-x7F8kr~e^fF%GuKBz5o%M}$%)!k^6W@8_XB?uN@<kU=?#VlsPI*&SnFI3_`eZ#7
zeaQyGkr7<hINF6a6`sLRf0UKEeA1wYwMFtEP5r!Z@(^C_KlLshy@9?XIDJCohKGJ+
z^|fF&3R2dYYjVlTnlx=QvbZt4Sd(~u(D;$1eW~v?zBw;lW^eoCX1<*=RfgT)d2LoC
z&!G{IC;992-M>#Kbcv7qA@5wfonpCpBy;CQ#(sL-zAL&mKaS+&HDwh1J~upTG|5|c
zLq`woW*>iO)my(AU`-y{UIP3*>{+!&$1BYq*Zt&xUiS;;qRxb)gLY|7o^^D}S3H<y
zkM%-5U!H~3-$Hruca4`m%0ODPOq|vvo|!QY)zhopr!N$bKe`DAl8-$9(aC;={?M4C
z-RMX3FZzbj=T;fwmG1r890dH*HP_Cwt|#%#FVF3wxwY;1CRqi?{G=qMZzJd1XD`Re
z=P&9MCJws0`qE1h`kT}x+JJFOpRM1@8Yyk7muLE8a)9x~zSLKL__|)yxx2bX@T{9h
z9D*gCr5B@j$`)VucRAWabnboST3v-N<3ilv(`GwfxyYY<s?(%}C-aN7K9w%!e(D~N
zHTU<mm-j>M<-M)($2Zt~!{bR^=lrtfu(#^_k*>ZnpEI}d`|k0xn!DEep}w2_-M3H9
z?%YRH2D@5MJgSq@-+lAl*`x0tm;X_%F`m>~<w?zz>9u^n^XuZVKlb?9i?gS71+$lG
z%=4@_v>*9-TC%k_I%-2NOHcLA=%fvL@ycB1y?j^RJ-6mz+RECzO8LMW_tdERoFVlE
zpLV585B0NM&Q`k!jT-wH*SBq@vxQIkJ&$mno4IX=cFpbbo$g)c#|veboqpxh+EdDy
zoig{lS2dlzsI}A9o|Ab^=F_BG1KEeqypZ0WWAiZne0`b@%J@;e==ZSJ8QO(CF*KPw
z@f)kg+NN?px22Sabot4YhoqV3Oc;#0%x=D~&&*h$mA)vQjZ+;$lRj16^!1sOWepHl
zUv5uIf9P3A+mwDUb9lJ){rYL;?74bVi!<{8es1bff08jm|KL*x(nx&tjx8W{2+u!{
z`E%I`r|<joAAXPR>_7is|7EzN&i><n`a5a82Xo?I)f=F{ny%Lwl~N0K;4M9h;4@)$
z$LV}LYY?vdnbShDhFf9XME$uoQ@1{(w#l?hS3X=1q2T`X{gbm_*N2)%L03gRt&TY#
zR@*1%P@F4ebiS|Z9_kyl@N$3pTwXSuSdf*oH%C}4=Og4-DW~wpopnyS;x%Y*A%ITb
z{UdVv<H4NYs0`0o>{Y(+*M?5sJnQVC^H7aVhtk1Dhq8e4=B-8k*zl==^0&YE?7XZD
z-00i|sa(<PsFeGiDtB>*fDCOJM3*%lGPm9w*MU5HUT?APo}DgpO;%5*``2hMi=)%p
z6=s)B@3MS{^k0)}67rzh;ruMD(XPqYj!pWHrJrRm_8vMKJ@d@rw8a@Oyf1Ez9&TUz
zn|Y3%c+Y&wE6>s#cS`u{7uBcTxp*sPYn=F}Km7UZuYdc?y`o*wviA1wUZ#&2Y0@$;
z9WUFiKB8Jbs;GVW`xwv0kE;BM!v9dl!ry+e7<$5LUg1=W(+Yq1=F!=2zkFEVFj?kC
z7B}9i@@*I=UOufW#w^ZEa-H6IREuy^HYY-|={_bFg>`kC(>>f9WwdZg=4OqHugfs<
ziIY4EzF=JKR<l$mW5$ahPE?e9Mv_lz;yDaHHQDt|9^e$u+ZWG@eG%XqGlqiebyxhm
zT8O&#EwnZRoDBM8XM}$DqI7xi;OzCIM`t%5J}kOg;FTf9H*}N%Ij)`?R)_KfhK099
z=d{prEnFFxi=KY(EM|Atunl7Lb>hPHUJOO!TD!`s{EQPeJEP4ggTfElmG1reP&DwZ
z7biBH=y_enu<@OvTF^MTVv*|mEaC&>doA$wEhMz+z~UN|cPxZmf#thH=3N$776t&V
z^jrz&dp6P-3)%XUxry)BIL!rIANRWSl9!uBJNS!FIhZT^o(mcVm}k2=^eD&TWfZ$!
zb*@O|qf{GdX;-(4haQ1@)ka<vp7G2a#GJ~BAK&~T=Vq0m(>Ppu2b4=T(2XBJnVYAW
zclyo{KRwBDx(VcJHS<s}e=+pX&v-2F^7hs4K6-|^-`430GCr--SmuFfr6>J2rI%Ag
za7!(CI@bM#9vki1G%{|u?(^FkPn;B!F5b*px#sGg(^G*iWc{4*F?I0l+psa`J!P)Y
z^ZzmRA@`VhPU?8Bi#1)=a9#gn{FjwsPVdxR-%vj^&yPOxs2}amW8&sH?kO^}-iWNN
zdf%648xvXcMb5`?ljhdk@7R2<>9V%oT%((D(P_pf>^o&!R~UxoW4VWP>FMmF-51wo
zT*iMH*>ig~R&AXgBA*dIdk2?E@7R2i4bsnCCarHPFTQacAj230vWGAN?N3iPZGJ87
zH5@Tp_?E;6-a|Zx;<x&d4-#sK$eh=H!8g*Ayx-SJ^{af}cVA*x)s5`z-}Q7dKM8Iv
zoi$wO|Nr7%rXBrMdH+i+|37^F{n+138JIs=U-WT4>rZp!o(KLv?$)Z|?{3Cn{F$S>
zcG<jt-(FYJpbN-)f`6-iv;T(XvJibeOV-udJ7!pMg<kZE{5c#vH~WblcdCD%{C%?6
z7T6Ll%TM;qn=$o&{`dbnlK%Vu@!!r>*ezw3Fh*d|i+><?3;E&S{jo;2z{)1Kg+vCm
z`I`%!ulo&f8DKlU<LL38E}d{A?pPy_^gBpbK9atQh~K0)71MbR`K2-IugsK{hsk88
zKd_N!kq=^&n2kK*q8tCIaEN!e2l7E0ov-`X($7S#$Au?N%0}241git#&s((K00Mcy
z@79KzZy<n_RoY?7HJy-h4-NWWFWui|=~()QcEttocm}qBK_xnD(R6(B4A=dg=Ew5S
z>9N-9m;-b91V8fyf2@0o+(&=>c8Hyy1qO^6)EltC2&Ao@qI1`8-@unNxZ0?bV{2d_
zp8K6Ox?^zis1Ml!gEuhA2H7YZffN2HuT$Fbmgu?MXPM_^oz7uPP~m@UBhNqlujQ?X
zJt}+<HRtaGf`8$Nefc}~+4z@S1m+dlJZ9tp8K?PevIF1K>EMd4^yuhEJ5OO&fZ+j*
zW?!3zG47PNj5oDaTZ6IR*zqqqo7mu!ItAeb3?t7ll<Y>dPwGvfHSn#OVe}!mAV!|h
zE^_K*Nf{?fV3I4k8V7bfj533_HrQRi_6dVb#X0fA$O9O8-qjm}bg~i8;Njam!&OF=
zQGNyrKj{s_(6Fp5x-mSI;31F2#c8N<+GoLxzr!$A>8Jc!)bKv9Q#~DQs)m-4C*#2d
zO&P%$)*9@x{a!z-4l^HVq{Ay;0eQ$(aA_xrO&D!(!(RbuCSF@n_3|IX6gt3=IOv)H
zUg^rmxttb{&wR+{rA1oC4`nmPlcE1|;!GR^odp#m4<PDMZL%;@eOtybK!?alKfuT{
zUJj`FkvCSm3dSMz9uW6xH(v5*zU8GWVPkk7yNny+yh&(IhsL6fo?zsE_Y8X5?-U?w
zAE=Ktu(-V}WK%Hw32&tpy4kfM5V)8VUF2warb2O?t4&-;*yhQCf2LT=L+0Z%Z4ayC
zwRDfgby@MJTodNf&s=xBExKD<Bz`MTiRW5tbR$oX+vkrid2Y67=egpc_B_w2%mF<k
zHUVZ>E6dR64=xL{&b?hG%~{W+m9<;Kar;^%xU0H<{O-vR5uLW=U9a)=r-FUDc+mvp
z<MnH_*>9NRQhqj|w`vwXaa=DuBrM&lz`SN#GxB83M)$sVFz4kIJx<l%Ti~ZcH}c@6
zuXgOW)rWZgldrV-guBkc>7P0T-*JchohNI<&eP*NK7RekDPT+_Q+~YQx0WyR@*MIX
z5ASQM0aGEY-rX}!2h4M|CqHdodD91-mjrEQ>z6u@IL`^m_=Evmr-R%2hFZ$C&u=m}
zFpmRXwk5lNls;vXabcm{{wq&5i^I1~`Ii^d$Go&_`S7fKjQpst)ZMAM@2x8R(Wx@z
z!LRQ<ohqZ#dx^b+`BTPE?JaFCbL2d8c=&&_TdQ^%@kfqR*vY#!u6g{vx(F>@;*ZR&
zdFEy2Ns$>h`U+<cm_Oh7>biJldl+5DL4Gfd|6Gt~zzYq>K=$xD?VPu@F*Wqd$|#}c
zMu?6l9egUD+bcM<VV}dn&u#{|8Q;T)WX<u>vCHWG9zUnw<2q0G&*}BJHNP~rc>2eP
zXWr-%x5vlW0XN6b;m#L-AEeekFov!##+;<N3>?FCvkq7u1LN9~BJ}Zcx)~I7Z2Vo<
z$l5CN#B)1M_xCtP%PnILp73qubqg<zeR1Mv*rGpWP+L3AGp8f$kY?|glg~BUtKeq2
zJ&*CnefXO%R+Cz*{|pH80%KLc4w-a1&8%OK+wyCpu6tYq^5%XgbZhxYT->gs#kdLO
z)`TPjVLwF^Is4q6M(8qGO_;kz*W=?4-51kop2W|tE2Y9S$Nf~{zx_AA`>B%p(bD<3
z%kW3b*iRt%mtQO&P(74XzHNfevu#5@zz~s*{>4yHS18r>3!_g_IrV$`_{G_ix>C%K
zfoGiNDC16G7=0=*j4HLE&X91wUTzpr8|c@)Nm>RR#t;UcA53lsUHR)dvfJ+@Prw-w
zCx3jG2gDtHNo@u)D4B#Z61=S|Cdj<=pceSm*9@&H!gXoyLB+X#t-~<#)CrvPE8R-&
z*nl5Kq~#RS&AM6*ydzJ71w+;Hn{qfg1jb3G;bY{gD<A75Fz7~=UGz9v;`9!<zxy6i
z;%-a^xLkXd1^B(s);BU3Sk}^tTy9Qj^+nDuOJ$(BFg9G&Y#9c3hKZa6iag&N0#%LY
z^NYl(oBV>~W@F!}GIg1>J5Bd%2j)+VG)`6db)aXAO$?Q;cu+&~=^w*SpTEXH=479;
z`crMBaLq4MzS;E3;=ZS6yp*P2%Fs(WHaAPm()M=o_?)^oM@e~Fv}JQp{{4(u7Z(HL
zx;iYa^g}sWBYhZi7}H-qLY?`+R$5(R@(<TH1p1;s@(;D)E;s(nqY^&G?ThJS@~|ee
z=W8wgGPsmO=cS)^WUgf1*3&&^KGtQ0H}#je;1lN3McFz%sIEf89lsqSvbg5?^>wo@
z<^#VCzz&%M?%M?D^zrAJO*r}`yw8o!q3P{1xKrfe>1&b5nB6@8Gkz=l#qI@*t@z+~
z-Eo3Ho0}I{?`*6sD5mV!p1G7?|IWl(9fX(bJ5dSuOt0a3{64=;?ydETwf34$Tk<lV
z^!S<okgmqfPLp!#GOyy_>O;(LwxH`1np4|$__o~i3tRCY1D`Q5`H4HrKK41B`FdEp
z?<Ww=QS`6rWA(T6LFXkwo7wuM&Lb}Mk1yi|T&I)n);H8ruE{&v_|q2b!AMIQGr)J)
zY9^lJ{v!HJH;22-pWENpOljBJhVt2JBd6%BzUFc~g+H`w^^>@*@`aZNzf<;}A~$aK
z^tRFoo%ZH8@6q;h?Edi0uFq`*%;A%d;P$y$`=HUc?<==8IfwEXo|K7jp3TIqxt`g1
zjL<f=(%ZtX$qw$AvJGu_(<h|;_qa9xse5>{jPqGzDt0cIUZBYoH&#Rf`PT{wF7npm
z@os^4Ps1y{Udf$jmKz<>^tkxf;>G#7j$$8ltA6C6?|u`RLlAnE-OI7_WxuHFoA_8S
zg%cl{S@0{b_`?@BA1fBVRaeJ>@Q=WOKi$x@M%wEtb%?Gn0Cj?=bO^okCf`N|Bcyin
zV}kNc8B1p?H8hEfi@x&;ci0!#>Ehp_zwDsa^-LPay5Wtxj}Of;bY%2A#vk{p7Wi&_
zS<^{6$LdMf%RuDEjf^=Q{OqPI;K!9Xbrs>(bF|x<okhGCUQ4&9f!K5A%EK8v&z^5G
z;-cv?^%d!dy4>6Ky|MHSaYJ|g?#0`@UDllbzPOq8V~1-i{9}v!=<-g{@aHVckDl?5
zAkaO-$3WWb&DsreANuk=nfeZ4?UcEi>{%HxY&yf=zh7^8*2x!lWdq-!`K+#Cb6wl#
zU)~$1cbtIXA5c5o_v?E<1FC`;VO)h4xK@ijCt+$weO%F27J%|z-<^E9sL}W=et~P-
ze%R&jU7PdJ^Ie_(d+qwIVKk~TsOR^^qvf>154GX{g9Sm+<r=pioY?WkvjMq&PV&gl
zC(%>hi_g!McCMvkyt3*1vW#3Y<_x{+{^aqUqB1bOuB+~1;Q8_w_ec3TDHY=bdgaSV
z@^)t!Bh!^=u0ryAUFiY&0%%?J1}skQPR0$VRQy2ZNITYZV}tA4oCKnWZ~LVF$&)67
z$E`YLL=S%7)**u*4h*^~t8tfg^@I05y*1+Q)>Uk-Nh?w3dJKm8lf~4H8zraY$FM?=
zdu1?79it~lKBOxj<_9SgE}d^xI$ZoVE&Pll=4J8-@Vc^yKIHq6n?<GZ2aYi_`u(cj
z1iV#sZzIKKgr5bxF+{nU^D;&mPnOnL0t(z&E$r9FGn<>fYn8cqctXpigL$-5K7gUf
z8;SBNZVT4pf{S|!Z)ow(zr?M%%4Q7?=@6cbp9ynkUg5sU+vn)v>pDjUT-tbWaXWns
zOXAGcLo+<}2W_DHgX8ve_#?m5$g5x060@xEpou?j)@HrD_@$)|(DB2C1}+dhT;j!r
zCT%nHJuWnHJI(AT=Vm?UGwhq^lk3+v(Zau#<`zDD-OY2>acllVGE>JN=59^eX}Dw7
zg3*u5uy+ikZ##ze=cLIqaYzS7dob5Q5BGA4yx<b%!mZ(FnuND*xgbOOU5D^Q{#sgz
z3+<_{v27omB7aT4F1yRhrhku*KlRw@6BjpeapAjvrc3x^xohLcWn~}x99~xb7T;W6
z$y4xY8VN^E`u-U&KDfBy>2V#`{cAkYq0=4XU*n5>`<QcHkC8dkMBcIZ&}E(`92s##
zb5++Dk}d;R)nl#PkI_TANn71ytUp#>URv9iR^;!a5oi6dFWs)IXCWoc-{<Z<z3!jW
zm)2H%$19V}A3C1A9`61*EphwuHq-x9VR}XHxQEg{md@AIuFb>-pmbz&LZRh$OB9P@
z3QwG$iL=F722$~d<QB<!MVho(pl1@EhL0wC@TZXkm-Oc_+8M}pq}^2%;6>fv4)BVL
zEPA7Jcez2h$2&@|<AM|CzGi?2j_W?rgyB+F=u(dGK2DEL+1ZqV>%SP9IzJhBqQkS=
zJQ;?!V#GT(T6g*h%h!Xd;{ZdRjs<?vbtwb5gj2=|X9GF%xXQ;*UPr>aw06ZYj;neL
z6s-l<sNeEuQd8am;{Aa1G9U>DuPxBqFPm55*C0G=ad5MtP2QwA<HJ9feeiS@H_D*O
z0T^tuh^I&A9c5nWbbyP#aeMqJezNE^Lzmn230;rpPdIfRy6`4{!LPY!{UU3P2JTS&
zF$+0z+?fobzqZ{$*+i#~=TE-S90U7wUW<$TxYFo!sWUD-H@~@k=yyE>o&SBEdn)h8
z%%N==%i9yunZbrm-omlblhop(XOv!B)Q|@+-NeVesD<vEQ(xjdZ^uIN826CD`dGO?
zCtdP#RULYrT~*dq`Q>Ttt^Z@?cHXt}NuRYAmo$_|#(<2wiR<n~o2}=#1Ug|ij4wrW
zdv{{RX8N7l(6>Qu<KKtKrTwO^I>Td(eDvtW`1Xzs@F!1e@?3I0eNb<>7Quru4hj1<
zPvGG%J{<{OyIt2{Ik{6X3@&A$xpt#A`X4?RMxSf--5uZLxpDJO*=}kBzY^1zkCQ*H
zXM>|e0lpvVLzF}R!u?Q&muq**5Ouen)AIj7U&ff{_1&MGQYpGywc%frqhUO9T8NRx
zRc&y?_~XP*!A*4JGJaKlZdCa=)ibVttN39cD!xy4#+Dp!e_G$AQHFTllo81FaZVMz
zuKC&sS9(s}Fd8jt)rVc{+BO^huL}39%9a6b<h|$^hTfE{?|1V(<<AIVv6Kaq5`oVq
z-}uIlOcY3K%mwTFd)kTcIQ3J0SGut~EhoE5W~BpmMLhh{A|EZh-|aFw=}*ZKeqDyR
z<rFi@!qq?dQLc0%KL#v9yI-2V88q5k$ur*?JjEw%_)(ta3U`p2`4w#Xe5YrCanyX3
zuAU*|4`AR7;L9^XR)DNr{{|n~iI>MM@CLiQ=eb3fXA_q+D|(chIAsVJ;`YrM@CD}a
zzQ=`zn{>D0(TGcYT=>A8(+|sHq3jzX;q7kZsKU$_60feb8Q`xgi89v~m-SR+ga*C-
z(@Cm7eD&Sn8%D#zOC$O5VQ~KZd~lqM$SdzT?LL3X9bY#VX_t{L-2fVfMW^9<T4?ms
z;X$XpB+PwX4<@>_Piga`RR?JJr41kX<jnffC-KS|P4bd<DqbB%2Rxw*FI<o32YKd}
z(DIk=k=^n9;p2AMiJNi3b7z=o^&c}1PjGSJq~GfcuE%%3xHUM2W{vL@96k5-4X4QZ
zOQD?&tjaMzI;LFXMsIDZ(*@sM{a)Hw+94M%c=T&>dVFLm`%b$?L;gssT#&T7{%d?m
zBQAP!KzQSN?&BW>)1Syk;}>J`M0PF%y*&`L(hT1ky|_T;m~iTmJ6|IgdT#10{M_hz
z2-o9}@h9C*yOvguKgHjl!#OSC8Ao<cKk;#oO?xiiNA+!3?R?%ZATRx{7gEeE_4Qqc
zq>0zMvd7`|6Nj(oxAP?pICQ71Ps#S3l6?^QPC8`#?`tKzq1h)VaXpX0ac6w$CC^Tu
z?}RpPm(%?jTRU!_|FSP)(j(99EWhh8w~^#==GnsMc@I5#4DiZ7b(20TG~CQTJC5J{
zca2_LbdWc5z)mMF@o~*P;>RBdEjei?W1jWRasnvxs_>>P)86qUZcS$7&opS^k|saf
zLdsBn{^9@o$I<@t{6-f2BF}eqLE`Z7tJlyBg7ydA6?|*)9T$Jxh)DSNfBNP<8Sv6i
zo<bj&PU53u@c6RlrF=osK)VLw&zE_fsZVk{V`$(sm?Yqege)q=xqH&X?6`P;sM8S8
zbfvmn4HkG%EreB+t2%N=UD8~rmP}%KPsT1X_^kyKedy?;ZfL!QsNlcb-3cCkI-v75
zsKqKg9B`Cg9^nkoIW-_2P8#GV3<u<Yz7FYLEu7JlXE7dH^vZiaJ_@fq@OU`o^UnR-
zV*xHbKO3#ek*p!F{;b^{x#;=+HE8MrUiss-*+Hg-w{+pW^~qH}Da%*8jh#ZDcUs=$
zLFae3{NKN<{6Bs+%4|N$olaERIz4G6z48tSjdJN7u*f_dy?kA1;7DSCJYcgz+(jD8
z8?f-9St$4ar7JyB4ybMFSzLlqRyMzOAYbXwp->5<1KME*#s#zeH-?Q)K^C26IQCA+
zVIZtF8hyTbzP_Q}^Dw71ha)Q;^j7?gPdz-ZO+@`b!lCP7e(gQLr_H8wgpbbYgXojN
zD7Y>o@^j^3!f44C?muaO3oVv#{~H;h3vKF*8{PT+dix0dmP|S7eRNMRJaCZ{S6uSa
z{rwrq1k@z44^;DWh#y4WH?8iXkIBt)RJe~Ibp&{KA9?tPpe8H$yEm83#)P^5^iTiH
z8IW`YdLD(>-X!bvXtp5jWh?%t!q0j7snR_)9p3~A94cdZ$_o(TImPX~Yvq$U$y%KI
zy93$C&IU^G-PNXn?~vFayiunQ@{Jx>w*{V;QQ&2bpHCSbY9rdm=oxPq7Myaaq2lau
zUElWnd7TC-I#-`Ps|Dq=$1e+4Cok$0hj=H0&^)Ti^}8BFoxowBxn`U!Ba97khM#dN
zr~KlCo#MGxV>F}B^_r)=t?=8ougYLjlURnQGVu6j&$asQPj<}V-_=~`&C}|0OU5Uj
zwJJ2%y*XX<iZP{5^Sszipl{SSQ=G=5$D3y_imonZbyZts&S@Y9ph|Rk5M{W43>}11
zKJp#Nw|^LY7<OJc;Zt<J=X0y(IOICPLl0&3uFCUz>8E^Ny?S?cv(kR;>bSx?1*KkY
zR2kmA>l9GwSNNNCnpvHZ`>g7qL+Z{v4CEx$%iW0}xZ8Cq#;L5F{t?H>!U*)D>IgW2
z>mRoO01qumL_t*D6Og7k+KZw?qaB!2`8}(jZ_`LF#-{OIqtR=X@u-T8aprZ+#a=RU
z?dDhvJkKhvo{kT-0yyc^|Fah*-|u<-d`n6j0t`atgMQ{q`eS3Pu?<eYZ?2`B0Kc~G
z1RMH+E%~qC{4nwc=cm-nqscH18GrGLBa5q#@85wimf!=K-{H?VD<6^J84oSKmz593
zEcE2lo5A#X6<o&jMe@bpK3=6yI5eK=6XZpIwzVda0^Qd77jZM*;xu^?zXhR5pF@tH
zxqs#j;Z+7t>X}tuHnj9%>KbUX0c$nslcQa7hcU0(_RTWzTJPi=OXgAGG=~QXKi{36
z!#{)h)mj}We{&jb0dyL=M0V2s%6AhgKi_@(^z6IG&quyeU+TWEkLW>m;xlI{vLRR6
z>Rg&+`XxPOmAM`{=(v7;UM38zRl8v(S2Bvvrb%Se>17$N#mRT(rHPj=UGx)h^v_Q@
z^mHov9Kz2fJ7pOoO?1T&+TQ-*1JVbX<Mg##Xu8{F!>27Koy7mc|N5t4d_^P9#!EID
zLmL;(8t6;5;8l%bTM)U?rPC#@<Kh>W=Vc!q`hD?han`+S_@uq%Zt(=4JaY3Ri(i}5
zjskeJDRDXJ11Ajfy|nbV;?%Eag6r8%8|LE6b?|wv$>~6bX9jK3ft!K+kyg`JyAF@I
z84?$K@{%ib=XiA5dtWPsmunmW`cS-{1!D}k=`%c^;<{{p`f+{BZeH`kXI~0``mV^}
z&hyU1Rn+-fFgVY$Z?^RBVWFGLRXE>|HHPIGSFSTHevmYl(DFCW#P7dazMBi%`Qz6H
z(w}tx#K%osPfMI|$Azx*B|dIw4!Lvr{>|V1ZtMrGwX4hz9@Ko=KA~{eA!&KG`nnR$
zi<fW5GbT>|ukFKkerfjc%DiocU^;L7WcQnX!W$l8@<GRCjWZ(ZPaj|U+^q9;{_yF`
zGUjN{j84%l^^!S5c-~*}2fOY@_x0EhvF1LP0ncfk+s-d9U5DO>qDvk-&p!Oz_H0?X
z>7keZT06|X<Uju7A4gKD3+o2-Df4-351n*p-3jWrIWOUbPn@!V&v*>4KbiMf>6E@c
z^Qi3iepCHu<dN0oSMa=@dSPB18vPFcXSJ5i%Wcy5H~;o;M*CMj!jToXSkCfYT{vJa
z$S*I!C9mS6)8}=|LE_;-;+D`wxu8p#h|~A|zFwI6^<RDdf#IH>y}r9FI*IS~BA$GE
z%pIZ|`mR5k*L9l5=TXj2^msUR${D0Ft2l~sK^|mr7O6?VH%tuM5*Z(0ZFf(03uO4o
zqEw(y2-3f495N|=8zQ-*PMS$K{NzUeSm`rqL6Zj7(+ThFj)itJRESq5(r_%z#79~t
zsIwkj)*L6Cazy9Bwo#Eu(oC-*4?48<%S2!x+aGas3%~10=$Eka%Nvea+<aLLw}U*-
z&>MfQcXRDV())V15o_WT(7^p%w?U*jo+mhH$#6x;uj}rAbp5!WIE%IpLLDOwvgPew
zy^&1U4&u;9UvWMv>>he$10v7j(RBxN8lj=zO7V#P0O+AUlZK8+x&aI3Er>peCo5%p
zDE)4f2>+IwiQSgW=<%X9>wHfl;2E4@9!%!=9`2prR$K1$-5;Ic^aDNI>G*qGHbmfc
zj`@hGFi0OnAN>tT)5``QTx6yU5)Qv`76OUGw+{}{<+)j-^Smcs8v0xt8J$*+<SV$i
zU8gmic=*J{uO0J$TbqEn)`}rB{dMT^cYxOy+w}7$y>jYff<E5n^ztv9Ka4&kp)81m
zqfh*OQXL)s`mcW-*7JOvBX%-+xceic<Nnou^%58-Dt2Q2lFQir-Jnmz_v!oP?z>a^
zQ)9X-%A7>W#`rykj)Lnp-K!6ED#bH{?m=D4#-K0^7KP_(;fq^_fEvpgG|(;vjWXgC
z?z|tmSw<Wd09@rCP-!!KFx)W0ysLhku?M_;S>H>s0HWvym<tR8Ptl!!zsD(`il9SI
z^5nElu7G<_+)h6I$M<}MOQ&niC`$*%8K-)biQg!Pk%qY?15Dv7?{})4ZkKVV13k&O
z`|xhbsk|z?Mfr+Y#@+Ms-gQz8EsSNi>(mq@k9h{8in?;$*wgRLa|tdZPQHUw{z13<
zUXZwjg6{&d@d1<4WpoPY`;;G5Qh&L^4^439YD?Rx@~J*|No)5Pf8mCaXt!yfYwNO!
zpY*tJiiGiLsk=pooH9Z~)-XC!tU&ssUPfD7^rGj$?`U>T6@PL%SZmR8F&7kLE^3}P
zPI52eFAAsPe{t{5*%xIT^gW+!q!=g6le55iQ*}$n%n4&ulcF^I=tJFMl&CxpU3X>g
zy4Q@;cslfXBROEStHrzd*IdGtc707U6Gt9=_s?{~uCiMLWnwPLAZ8OUa-RM9yZ6Y0
zk1JBURv*eSeVKX&{rdvu-r4MNwVOE|Inp+N{&Uqus;;jqk`^~S^QOJDfO$x-1M?T4
z{LSGrPwzCIhn@!iuS!=<?(_P7+UIrJaco+Y3?PjdbNc)u?aDlXd||Sr&4s&H{Zd|7
zg9FMa{sr^?UwU_lOnK`k+|14X^zGyGH36RUZ+;e27VVKIU3HcX4f^IM4cBn2fw=s{
zrz~x*!DpQ*P8js|zhIWX;mw@5$8~??bX@n(@qhKVzd2vih_msMmoh@jRc0XTs1exx
zNn7r?_#-Rf?nZvcC4W6`Ueolr?(cHq-xt3Y=h<7slh^0>kso)9C-}6b(0A87BzrvY
zc*0ks!yrv~1a^P=v)}Q)Kc{c*FJ_%z5!uEE&qYogd{r-8?kndmU%GuSr-|%jANaG^
zvX+;bSbNv5)@YNL-tLTPgSP%~jW@twW$zn%k87Qy4(0ocdZznYY!KI<^i9T`7?kWs
z`0azPXYWfNJh**!fBNoJrzKN3FC&j@S$DZs1~ctG9TL~SYuNeY*SGhzbodf4oaZoc
zaXU?LKZi??^hN4;>sg@Nth4YKx5~}lWawsE__PW8e|>Emp19%t>QQ|Yv5?ltHg#tZ
z7u^#V-h{i`<K{6iadA5zf5r`YFb0}CbXxw*5zw3C|5&g!wjD~(bGinzrcN}^u=D~m
zc0D}H%e=GW(?`z6TQq<3H@_KzL!b4guwT!Aa2WJS86UWb3l7Z;XaIfEhYQ0iuZLeQ
z-?P=9xH38PanW|jes9)1=sIou!T0_n{lB<8D_!;>f0y0)mwxoZra?}po#iHO#tZkc
z!ZY#Ui+g2iscYC<+X{c&)Ws}Md}w-H#*2>APG(w09`MP3{7-*(-adI|!a+6tQQ^HI
z*~GQso(K0>IT5=?Z;AW3?t@oQYw<I0Xmo150)r2Icf|q6z{W;!8E;G$Ef{pN^3^LR
zvMnTR+Ox))9J#h2gXbb~c_?}r!6!YP-K$#2?AFplUuAl)-li?v+nG}r;$5*u2Mb*d
z{7JnH9IIr~;cBEW>y1<o8QpBantb3<78;SodIk<**D+*)N5;N#A`@*~{8@<e0}rDK
zAESdZBTqhEWtTS%k_I`C>WVp=K)<bxwYkkH3mxBF_u-dcapVNlTjYc;aotTG)^OsK
znRcAALboPw4cBGPe)I<8-Zp2N%Y?}V@;>V{htYYiwkB5Me+roUk7Lt3Rd;U!JgIL+
zWHEhgI=$Swt{p$i+85Vpx9GO;o%gE#kDbO<<!$l*u{K72^TmU!X7^J3TRL4z*k1_K
zPm!mj5m(l`KYcbFSnIErnD_aDyR6%FI97)18H~*~ON}!DhGPb+Wdn5i=w#lQ(bqh!
zlOl}czF%{{t||Ne+d92a-wFA&zRj~hoy@RNUrL#n=i~-s&)qU+Jg+ek-2bfn&t9M1
zy;tVi8jy^wPrrY0ZU}LO+tWJf!@%*bj5of^!_YHM^f2ytv$i&Q$4Qr0HJ0w)biQ8K
zunqc}t2xO77*jHDvvGf|PW-&9amooH?vOpc*Hea{d-biH=g-e>+`nIZHQqA-(4oo&
z9b?WXh5MoUTyW!l8RZM-dbL!z@^IyvYutugrv_rkV(f7m%A9JPOse^TQw6W<MANM@
zyf8pG_2Y^+GQ~ON^0H0^+2D9yCpd0bnK`lYWM@FTTLvO{puB+cy;j%Oz1!v8yv2M8
zkCRj0*!*J|AMWnzO&u^0ed>Ekm7nj)uJ*xr@|)VM%tG2}tuJd6`bDMV4PqPk@?zsh
zy)SdFn`6PR?p-aGf-&C#ebY!-I?^BWxeCi0Ft2)E?G3m}?d2|S;<a5y9=w3g@e9-I
zR<#p6_p7a144Ctp(`sv(qiEamg8$qYvD3#K&&CDgN;cxu+2cBGp&fqq>(9oiA2Q|D
z{4?bu4Zm(ElAp|}>1eKGek}{jB5X0{C=1Y^&GaqYJXhwY^i#J>*(}#_4d&e+Z`%Ft
zSD$y(2mZ+ZAYu=q_i@P9`k>4!`doTVq_%at>dG8?O%9&SX?r}q9}eTMlReA4GwT^?
zYBT1zcvJ35mW*u1XuDoELeuBh((HhTOl{t}1*j(LANt>~zQ~-@DdQXEBj0+$rh|5r
z@Uo6vHe93Y5Z3<C1!xD#M7q6?Zlp(h>G7=kqzBbkB%jr*ojt2|oW8*`5kr!8+}Ben
z_tZn`bgbQW4>ta^Hz!&28%s-G)XbXGGxxH}P`RXEQI6^p+^#ZyQ!@L8l=QWA&xG%e
zz1ls;^ZIiLIN}d^X)_^LN6O}7LCRuHUdnThg9p9b7R>TjyfWLD?=Jh$dIGJw)SsqL
z?CJd!{;u!GN>5w!?CdMY#4%WGeH&F-_i~A@ThDIvcWOE>)C9NXMke=|x+N#NZ^@kT
zNn^$f&$^oD>B9nBym?;riN-pv-5o$br-B%aKx8<5XD+61%)D|C?D-SkJtXf%t<}x3
zGL|vrat5LIl$ZjCp2LZJ7_FUF&Wo^f{8hv9sekuEqrRk<PyAfa+gjq_jVbA)GcNRa
zG#%zRVQuYM&GC$HbG>Ek?CE!T$M~fI$e7a~5gUI_q1_PiqoaRp9WNtayE<jAvXw^f
z8$zRvnR}V{UZzb%&c2FSqv?CKnbU?YVa6ebKXaNj2>l!vJpSw<n?ssSl6|YzC7F|i
zc9wB!na}(mBfO^D8ctg9Aa&UF7w1Kv7;m;@{LlaXUx$pp{~!PDd>r}T-z21aJ)!+m
P00000NkvXXu0mjf{GrY%

diff --git a/doc/image_user_profile.png b/doc/image_user_profile.png
new file mode 100644
index 0000000000000000000000000000000000000000..d8e515b59a4f6c80925606fcab65b0757266f764
GIT binary patch
literal 80079
zcmZU41yo$kk}wk7-QC^Y-8I48A-F?uXK<I`9)ddr_rW!|4esvx^X<OfeY^j?r%zXR
zSy%NveP*hw0IJF|NbvaZU|?WKa<Y=@U|^79U|`?^u+X0~1glY6p92#c2?<p>2?<hF
z7e`AQI}0!{+C=w+ZUqPv+~Mx@a8wvBNnTFgP;?!j#{6$=G+zC*>G&hTnuh6{JwZTo
zr0&m9-Iczy_h9_*)(&#;_>|<<-)#W;^3?}-zvl`c_QoFi(+|A8yaJKcOc)QK5_lLn
z_%IL{nVDdGr!IR?(8Lhwzb8sr@QG19-rs9=hxEL8WK5w~sQ8>PxE`!Z@HD9&?j1Y&
zt)7fpOQpVsP?nE3f{}?cI%>z<U^~~yDWH)!Ef5KVvOOLPbs7?4Fth+Z^OZ?t{*=gN
zdM=$|H*HDJO^}<0^D*;+-X!HDUo)hG{lVyDGf3r9KzX@!mMVl$tag0^_-^JVYMv?R
z;e?#qI=HHA!K`(mur}G7&ksoJm8#&>n_w!-6SEOL%gkFfZOc7q)18pnh8%<6{Fxc%
zI|u)&jhL`Oy9qDd+oE2cLB`hGcWc?TJ1gE?1S#xbgmGAWDeqJg`t-S`g|?ifk`fsG
zCkzV)9$^Fa<r4z`Z1|t;b9<7)!C*dn^v@=l5Ak1ENU?m#|AN5<{vj0mDj_HL*?%>2
zv9NG(wRUvt1NdruidwMwrtPM!q{wgPXwPh7?r3Vk>}But4-1%}7yl<{Z{cP_>Sb@|
z;L7hMME)NX{GaeY(=6nq|3TtrD@3lXq)IB`=wd<2&CJToN-hjfN=ho|Vs6Q=E-C#l
z_Ro<JxwV^{6F&=!r>7^gCkL~mixmqSA0HnJD?1B2JJTlxldHFbn~4{bgDb^<3i&TN
zk`}IJE;deXHjWOY|Hw5lb#!+VA}9Yx(SJYxxlRi&oBz?|;QBAKJ`H5~2gAa~%*yiL
zvOie`|C!}iwehmB(~-2X|8&o%4q*;9F2Vnx|Nk)mqw(K3wf~2cmxJTqIsXmwe>pW=
zEnFlV?LSqz3IET|{EPYDk^f>8WckPPfAhqDy7@n5Km9BWFUa!WdnOG31ol-L3``VE
zPEzcf7x-x>Tn3|z`vF0x=jPk7;n_=9U?><9ESbr@6s}zin^}A;5;t^*;HsI=zR*k7
z)XqUX-s;rN5ckkNI3zABG{N}ygn_svN>p)KS?f*D^|$-8lHV^mn<IjmYC1X^I_rNw
zUz+o?QD7tn4f(sXDPP@QkM}_i56K!YL|h=n7VN)wwL@sF`!Jh-36ZO>e1F^%g3I)v
zIeq!}&RWzsqh+Q}DRD;>_S$G7P<vm8wvSAQ6(6<5+Z|)ES29A~MfPA;imDHQHxQeU
z*sy%1#eSvG>8#EpPEbKc_7elX&!i?A>Z6x4j?bG%;Hdug%FY5dq|T}85c@!xuxd^<
zjOU5iZcA;5V?@DKUrM)#!csvCFZM}8&V)_SA}2*Gb$WtR%0F%2VRnAJ>BC_3+D3b+
zL7q8=d*8|ddt)P(x;#(Q;1~Ycah$WxBS&2K*EW2nI;x{HgJ;u?*jE5UZ$6tJz>bEG
zr4&-+U9F`@-5;axmdB>dv`qSA$%^dW<Bs`(VZY5lk&lVntI>fa`QF=RuT(=<j);1w
zVVtR;ep6a83KaYOPI$+^ZqsFEt;RjMPoH^Ad}3nCW?he8hjY2!1EbA0uXOU=f=?$>
zFFS6j3)Rp$W1-{ss1ApLZLagUf>}#z)jByggG@vEV0MJ9o#UuA&bH;NURVxFgR4Z9
zAi<gp^g-YFZQ~yzpI7HhqfTyu@Nk*>hr*KNfW)q>hYU}<n$Zx}J6)w&k!+!nvGMr=
zR(|0EyV1uXhb5BMkKAxt4Y*N+6meO^SXd&dA}7kf%hq0#QxgDgHr)M_*+=V(JWuIR
zO>AsY!g(GVHj*P_j0$>@#BMqvUbRHk(Cu%H=NTLeO4wr>$ufC-ob;l8X}l_NpWh27
zY2fzU_si7AuZ@BgQ`6hcql?F56VI|&;?IM7;zKV@d*VNicFO6kG>+1Q2Z=){rXvIC
z#ctSrk_hS)jB+YlcB_%N=OomTtdtu$Vmz+4ts3$NnY$l}88+{DAs%I@krzd2I4Q;M
z3M0G5zq>7ejg0CT?Q3%7Fg&wPTSYmAiQu2Y&fii$2Nbq*mVY5ptn+<FYoRQ8Uqymg
z2jXL;>Z}<4EJ-TzP0dd)cQ<5%SWh$Hsc-&vrrXZhaua>isekXUF!g=Z^K%mOY{kW-
zDZ%RpomS^=Q$okBL<W8Rkhm!O(<Ta?p^!~R<y1J5M3q)2BMktZvo^3utniolsGLbk
zR+}DURc;lmUYqAwnXIGlna-^d(1*pFiSk`M1H7E|?zz!HgV3SwR418A%%C^fCC^B9
z8E5<SQ8!#IQ%^=(d0GYbHms5fKsXqyCqmXWG&?J`d>2*;9`Mg1^#Fw@Ew_v~(9z?b
zTsjp@jwFBdmoxxYiI3cpGYW=CF<Qe4MU^E44e)YC4Di#*DuQ%;4OqzV8>eB2Exr>i
zEc@_o^=9_Q#zy$R6EDEsb5LXm-M=YRM?Lo#TiKUW!8{x)eA$;o`#l1`UVEAsgXCbx
zM~JI@uj;`N&c@Or071%nQNJHts|f4W-mEC<!`W*w5oX+m3bdFQ-tO$tpQpmMS<N<^
zFw^NlS0!Gue~a$vna&vh_Lrc!oea~}B^*J7Cr)$<l{S&fiz0y8R!CwQ4-TSgg)5y4
zC*F`^T7IDe<&0}~k|nXJL@2j5S9QTYJdEgj)%AtEDuek#u)ykiYb!a|X(|csp|2tE
z@a+ipn{+;@t!p9JNMw_X6OG#!dJ5G9E}^;OdQCiGnp21%be$XFMugYwMt%q=oeJV<
zI-HO|2C-<dnyG%5jR*@djvsj9QfEtjk8^0hf&<5Nu~M>2abR03?;xZO)S$rAo9Joz
zlvq${M|D4s>i9|6R`hv1in9NG_I-|mxI#b0+Wuk{V)#FYXU`8Q(hcyxa3Mn#{rqYy
zW!?IE-28&`mx*0}vK`C+UUM*yWo8==H~FnDI6WVgg%U*&VOv^S_JyDb5ju45oI8z<
z(1dd+i*T)ZyyRT_r%U@h+mwY)Wbs(YEj3s<ww|2(G(4o%$Hd2nHB1RgRPRh)WVC<0
z!}Ra`z%otxovgx7mhC#3394>Chih5;@gh(4cgOvUCUg+jMqCx*FX9S{T4l6^dvm*D
z=Ou2vim@=6r{)LuD8*Gf6cm*3f?nUr0A&XNM9X2L=<4Np+cRo@OoeZ;mhIfYqqPJw
ziqe#M-_`BS6p{>KS3_@J`Wp;#sH6YMA67OtV&;HJDx~+9dlJ~q=X-Y~2Lx|C5+b;w
zqJAA|WTmO_x2AG+O+t4(lARyDzM{yWN_{wkQr_4Y#4u2wj1(US2Ny@yUMFWEF|VOl
z?_B@M$Ig`8hIHSA$RDokmw8voWdcO401yMHE!_xxXn44Pcemg&@8eLVKL`@0-}@jJ
zBiCoSfkW!>`C=3I{hsLJTudH&*woaQ{B^*U1<5MUPa0RQX8XM-LKwg6Hq|!eDxZ_T
z`=h-ZjoU6Gf&Ev6$djWZg3rTQ@&)ch*pW`#ZyzkKHAr$}GxF$uiR(_+I#-G5nB5VA
z*};)%Q;E5@%u_7y!87d{N50K=o8fIYq|sdncA10NaW*(k4n35nKb{+G$_VueJZ$92
zc8ZNN$|lzF-(3n|aHnbtn{C9NjLJoC=VmX+rdBrP^9%<Ok&$F|_Rd&L=xXquJ&#oE
zSXciDC0rc937vT?d*#dHr-g*W4c58Ync}KlG<8?fL`{aqxLV(Q5$AeARt_tV;$@a0
zUKteoDI*CeI%gBLryZus-Xcz{+Xx5DD58m5dYlO?F_)jWM4@eY@K%uI)ECH4?<0+*
zj=&DZAS#<RIPZiz9hcKYAmn-b?~nJqMewW_UbvC2wm2e`GJ^c)CR!{T5t?QF%3R_X
z!~4@>M|kY{ukbOY`k;SH3vdWv-V^esa3$?`Ih!g9i27KG>W#NQ-CYC3FK=pm+`+U`
zGFv|0eEKflNF3qHa(*OL=L@#kdh*_96!o94`?!Tgwu;<4=FMmnZoZuvU0A4NPuM=>
z3V{ON&IJ0W-cD7*(mM{lm0q=z)|zb7X~l-a{R&?dRa9(YfLjhlZn8co#PI0`Kjgf&
zM+`tiB=5iIe7kvL<jX=4XQ7m$lLRS`@^A?d2=<6d@c5dPGdNLZOcW&I1^>>)Av}a-
zkg5HN6gmHb(`Qu`Zjv8p`V+A-Nk6U0j3Mm<s?n_2-PexSZiTfv#CbGVko^&E7E@Vy
zA`~*Bl)#)B?-h_e&gk#b>!k!&)AS2bHLnL-F*BjytgJs9mn1Q0`IoPZz3^$WFr<a`
zI_@>gwbXj1L7#8YwBs^wbuIijzua<lY_UEq!!ZwZ3Cv!E7-v8|b0K@T$}{?7Y7>m&
zH&EGaU;OJ?UiNIgR;k~0$g6$II@WFcil?lwJ-Ru@S)^i&vdW^HV1o|p%zeJftRWTN
zEg**LXpR=c9Es-`^!l37Yzt$vnAXD@q%0(2H#RgHvO^_1L~lt4p{;Fvcrfmkyb|uN
z@g+f-lY7y3M*u{y*}<Lt!V^HofC;F-J%8}0t))V6z1%W?JcxhS^E#~??q_-Lv+Zx+
zxB)kyit=V0nmkp)lAw@z&ZJuZV<RrWZi%1429g`L3>Q0&)a>xK=X4e!joo$`M<sIK
zMKf7lsQQ_*9!$nT7Vt6oARodgpJ5l>zMr)T7jYq|h*gNES)v}IJ|7=`o5hGZ*P>==
z_zoi3z~Z*wKtK1p1hG=>^@Bx+M~1(T&#7o+ww?R%+*YAW@~Z_N>OY){WB?=GirseI
z#_g17M;Alqlabi-MW3n{2~NJE6)9g3n5U*qR(A!kkJmzMm4bsqbH1{*nmR?(*$(4F
zRP`wQ^@u4}lfm7cHa!z&r6aM&%mS`8D-m$U1BtxHbvduhE(Zc_nmtru00@(Xmn#Eb
z66vpCLJ&!C#($GQ2CJ#(Wzn$WX4|Y%y8DG}@o@+&$LnY(an_NCkF#m(%Tq<J<2Nfw
z8K>R5QYtQ39LNEB15PXxz2yj^P?9Z|N82AW$@4Mt<|KYP*rHgs*sm^$Wpav^8*>>G
zycXve$JFX+LKK>a$-$>h0X+P_^5ew@OY^DTY2%4gXL_x&a{rb`J+DNrqVA8&ty+m6
zrHnh}<Y&!P-j?OlZI>1q{(@o_Rl*7+`p~9wnqz{n#c6=1`jvVK64uK?+F?Hl&P?s)
zi$%<Vo~B&oOagxyw=)YF+PXR{il1&^nLno4?Eq553(au$!x9~s$MGM-l4YU93%|Ek
zk^_50s!)oGicIWG&2aIQn)~Yv?Bq4E<M^I`)tU|8*IHe4UG_ktSYW|g#L1T+9i=HH
zNh|$%o{}qyc4)Fc;nqpSL-C<HecgbB3wn)X8)ISMHjfiz_sYx7g_j|AFQ#Wjd?rZ{
z6KV~Fw3NkP$|>wP<+Y0_wxt(*8mO*y{O~vM$6+UubX@&IMd))nBR@6u%MC1IgdRV<
z!)`NUX=F7Vi9nL>=fri9B#}S2{R*yX{m_?+Vd4OC!S}}^^g`Vbi1!GhW+6iF0nlvb
zc(@F)ALD7deXND$SlUFCfc&b^!v;%~*o|o3HNSkbTHiw;?I?ly$yyAjKkd(Pu*8cz
zzX=t(7u*QHmj@f%zMzZ4L+zzDGnplN=lbj0B@C5cuR4Fi!;76$ueqqw|CUNbl=h(|
zp)@<6uX>WMd453ZA_F1z0JDAh-vpA7FQ-k^hc2g}>(Rr{VBEIG)AZJg0m$CuIZc$r
zedtnxMauk3QF1Uc#LI;<oCNfoZrj|D@OkWpK-%*^n>ysb;c9r9`)#U3c0v-2b^p?O
zs0ox>tdV<3GeF4_0-YHut77O#nu9ef!++XgqV5JLKXN>Nr9jan%A63EY&XWkd~b9O
z?QKlqOGLBOuQIHINV;OaG;NQ%R)3<bNqWWo3P={iV=uddAgup_*YNJ6?w?iCS!M_H
zafZTCToM+l#O1j+OCvZwt=-LXDS1;xWV+afey{7z^8%H#Cs)f4aAq0=2h%MDGhC<Q
zxT#NZ=+<eshbW>l+#*@f$EA2DQF~F_JC0-cHv%e~jb(+825AGv2>j2ktI7CZVOO0K
znd}8hVkz@zr4XZ{ZGz{*ty_kA%a5e@6lbAG%t@eRPBKEt-sq#%{%~nr=T-{OHz5Zh
zxR>%CRWy#h_G^lGlp4$K6qNg3C*Oy?&Y&D>zWuf-B7{O5#kS+nVS+QVTl~2ao{|m)
zr8`^V!u0ArmGYG$DmP`sa)d->_hzf$R|geqI|Fa7==-(@A-1A+!>a5m6Wkc3=UB7!
zBh4Mn6Q}G~;X+OwpP^8US2?YA`=R2_)jA1GgT;kE=IVuZ-&OutHjYBX05hT}xTQFl
z?7x`V?TQf?C8prv`LJy;@W@g%<|u`r&u0o2EOh`f&&?P9a@tz5N9|{r#ld$!GjyZI
zVef&Q@lqT&y9~6F9lD;<E;k-NzAiCvoMhM+zP=_^lIOh4H%Q?ge4|C1wXJz`cubEE
z(6yQ5J2<_ZtMN}oCU$<DP4BKIT2L^0^@WOXBKa7w+S8v%PswGX<RpZ>n0mcc36Ey>
zE!Cvd)Zci^eTX=C8`uLW^3AJAS)zV4*>+(GU$f+WgWAfR^7>h+`f7*2DB8OJ(~&;!
zK=5E09|vi)+yC_|=tj$5&I&7!9OlF5?zdJXy0dmh($Eh@OFP*Q&u$;T$Jc`z#Q2L9
znvl7tOQVlRls-$fZ{zyUr_p^>Q-4IYf`sofEBY7=-^FbZlPm(@-@0F~KB}yOW4iug
zgzadEJk4GP9~clrm#F==l@TWNcN>fE7kQhDz_6&-Fm@L4J&3zhaFIT$<WQNL{9;lu
zw&BJ4zFQ_zHj3M8!&vRL9DN&qqqut|+y)EsS=^i+cq8Paxh>^+SN}<ph+gpa_Hj*9
zm~x6T<@0{y_L9Aq%ApeQ-u&?!X=Zif9>?T*x|q{a0_2^PH^$Pf_<*WHj6UdO8nRU<
z;(P9@6?QVm8!z(SdTHBz&%h3pm3$f^aD+!0p~7JCJBpr1YHNLtHbTmK+%2;R&jLr4
zm#c@6r5J%XZntN#tDEmo;7uVosnyX9FrD%TQBqeqjxWu${WR$bBhll%KTl`MB#$27
zuM?N%PoUUALDF;jLY>-lrqm%x7m*IUGc4$zCbh%15L(7gq85*&F2S&v&OfRjcx(U%
zKNjMsp{>WY9X^E31jK5E*+?`4$JwmVS%H{D<RJrcN7I>|F-TFY4f6v*1=&qI@qm>j
zdr~FBOH%~Ia4`r-X!)Nc=7+UR1e{E~Ce5SY(V$n}?B-G^YwJ14Tx5hzGr!`G*bdoN
zTqf~5gyZQVT2;91Yo0dCr)HNan#cW_u=DB8?wyrPe`Vd`wj1$*FsA?+uCK`Z{5?eW
zK>XIr#wNE>9uGtap{-qI-ZSBtkZV)1Vo(W{7};Q`@Tv0!O)B!~g;|^RCf7^BS*{i#
z&FlEZO?9842?H~m_=<yEkp*oq^2<a5NxY?qZEm#H>K@A>sR@chgp>KjGH4z~<pHm@
znVvQcoEKOvuSwL`F!#k4vfB_GvyMk(NVu5HS&D_Zc+ddHYXJuLJAe=jehOFp%4A?H
zo(Afv1yml!Y(NMbv^PDzre9)+CI++AkCfj`<T!xJ1fz_plVkOMZe3?-f44%VQy{@!
z$P7P=^Y1m;?ZedsIYfI(d*W*ghnSW|qfwBCuD>cbl5I+fsc-_!`gzVMok;?U=E><A
zH@slI_!Cir++=y9QYf~Wo_?zIqqb098l1sOZ=rsIB4$7itrjwXT!z>b5Ab5=WVfhN
zZ1>|Dmp$kZu0>f<QF`=<acZxt)4lVs5TS4C<MpKRHI~DEfkx!1{Bpibm@3<O=;&CH
zWSsYsBk#$Rf<X{Nhlw|>3z^7!H!+~zLg5cSSzpG42M*Baeir8t9&&Qg_l;ndZOS@C
z!9ui0h5zdDGNmjV<NfVMOCdyaar8FN|A1-RU;J?kMuc&6ysVF>>t@~QsDT-gCfmuk
z^)5}XKe^qr!s#6p@I2vkuVV#F`!n2U^l`TtW?S=O*m}5)uvglcY+hnXzUHxdcI-LE
z)KbCTzx70`Y0+$tu;~Br0^-#%=#53db7%!g=NJ@%@@Goa*O+;#&XWH|mI?3K5#i=J
zKim9hU!f7X!z(c%ov5<2J8QqG_PH%>jv(=6)pzRUNbx_I%rTiORRWb}IizV+70ugk
zZoK^{J%E)PzIeUYoBz<;e2$~lzChS~y>|x%TL}tfd#sd$oOELUvEf@{>s6_qTh{8=
zJeRaWpELKxX!b|YrTemr*W83i(imkILGI0Iw1(O)!=aWRh#?cA-hB*PZ7sJ5MC`w-
z)_eG)>E>o-v4cJzu^#^6AmE9ou+`TPYm{S(mnqx#S`H+z*``d--OyCdD`TVu(<1Wv
zwwc5sN5eoJ@Z8)Dzu66naM3)_Da6dFM(V2)>B_ObUBelr=NO!RT^IOhsXp)6j!G{c
zRZg(C=ZR>)Xk!tiU!`z-2E~FG>=(k|FV^_pov`ByViKpP_Pm(?MYh1WDbjO9xOm?r
zffebvf~jwEh(}~B>%TF248E*S?@)c)eAx&<BzXo7OC|cV3!ZEv>^8SGxW7<c_x)bu
z>AvZ7Ix4}b@GZ0sd3vb+=n_QH3TvdU$Y@#fZQx$w%pV7v1J=*>ZH0u_FUfu%c$GCM
zEEv!>o7^fbC>Z#9XuWlWSqZY-iewibh7MS1vd>Dca;G(Hw}>Xd%I?iXGIx_^+DS!3
zHIA=e7+27gmKuVCWWg+VK2kkGwO&ZU2}<ve?0_%Uos^7tWlwL?H`0n{EB<Q#H;xWd
z7nz{(wHybLu!=~6ddS=f8YLMH6Aij2C2ur71<IayiDSCf75-A1ZM~Uq`Fvzt5K;r$
z%dK^RZO4#%X{K=m#XXFd4N(?=kja2dthk@SS}Qi>uOmgxQVh6F&b50<AJ1-(!ti`d
z_IH}aiuktg)~vKW>@*2cV#ETUz&f9Kmc(P8%HJSk9#?CO;!v#SZF5NVYLnEo?@`5g
zAIrW~Cep(SWKd$jDjQ%e=#?a~r?r3|GL|(c>Qb>GAjdf=uD~!JB=LtSnJ_s;+5iU5
zBIMw$sqq$+ZJLr&4;Hw{K4tkT$q9%*7sz1S#9NCf(v_WwIoM<+@XSh^G@M1BX&`<^
zo{1Jbx68&Ki(Xl239^OMgkhs_F{vin=dDKx!?aeFK`OfWRi>&#az#q8OO#YIm^6h6
zY+2e;((c#glu4WbCE?3v?$p0Ooci)Fbqq|T{;p-=W=JGg-gOy3Zt62<%@uny0;w2i
zQbeGk$dfG+5sqp1g6r!QMx+18cWDrZZahO#bD~R0ABUvSjk4`Q5`nCsI4;S)*W%-N
ztL*Y8iaAN;Iw#9ee=VltPN8^Z1*Q*@%wdit-u@0I_m#u(qkAe*#0wZW<9IU8qQKo&
zek-mj_E%rbs4Z6hw6pIk4GER{F@cA9-zaIC)FB-SJM3mi67sdUZU$xr{fo1q-U>p!
zrZf*vo{D<vQ`YKq;HiG%^{`zck*uXP19VBZH7A72_kD#|d-Y*!>}NsR^ztK7I<tgf
zTV)jCN7cvo2S?b0+?Q`{uyzf_?ySDgb9r!DMsFof>+XdjKs=Z0JCTor&Cm<KiyvoW
zunL*`)OEwm1QxGxWC1VtXKF758`so+niAjS2m-9Xp)rOo;&ah1Da4w%NA%K$Gby+s
zs{l#w30(H~jw`xpAy<YqL|GCLa9XADZeksM^)M4dK<7Rruq_)u5j*ny0EiAH;;4&7
z?1vU5Q2kJlYm)aWAFp*zC-$D9J{LZsnwDL^{9jd_?eDFXHMxkB&TK8gP9;d5v|BH_
zUi!N~UQ+uwmQ5p)zz<<Zh58(In|(}L2e$l39uF=7D!%+{eh)(RXXm|_-gKDb6-Mr1
zyDZDcmwPM$8eS8h@&jb2?ziGiy+-M1PDDZea1jv3&?T{F<`44xK%8{)Q}n@F0?47M
z&Qo+GmwqH3IuVDIm&qnTlH{6Ip1Y+&+7xyPCKbACDy_j*z*Of0fZVm_Pv<36dd2MU
z7(%XR=61j6RuRcdab>GuA<s%3Qk`|a_u&l9t?8E=vr^aLg;@0A_sdI)_>6{)(Bah0
zmQC+?V`ilLp;(aARJZ<Ho9;`Tz`6NPdfy$r#p5V)i`&`sE^X&B#@<*gYQ%b%y;Q9T
z6oCHPWk{A}fgh#sRkV?s2)^5@ZLFZF^?;3&9v=Q_YeA7~N%D-&??#iwT5=~ukHoJW
z{Ir-ntH<Fe?ohdIim*0oWVII3N}~vf6A2!V1T|jMQ(!^TB!+*AS%S2kzgy$PB#O9$
zS>?Fs2!tyx-D7U4<Dq(k%8AT%g&%|1h^p|+N1Vz+O2(}t5>9?YdpE@48X`hx0M_em
z^aW00Q~=qmxaxe0RtNTAaw~X@WUkHbu)^UI+su%;vit0Y<%%4n#yufoaDm^foJu+e
ztXK2APJFuGmPjMq2Jad_0|BIJ7i|1niYINUbRp72P6))6x~dlwq>$sftKZSi=~=>)
z1(Q=jl#C&2qvES{1tqV&2C=PQ<1A?zpd()_jHUfy%|c941TpP*Be^M*aaO0bZF;w?
zOuQnHpFr_$`I34sz<Z1}8{0~ydGLZJJyvUz$apv!TTi@<pOxhhmR`jm`><@SlH^9l
zg=sBT<DtMs^<2MW>Z)5SOGFHxi+enV09_fXycCuSOI!zt=6irJ)Xy~g5MU>MZ!SAe
zTj?)a8B7dwWY1kfZ#oWOR_5=gqTI~BVs@4erF)w^rNl7D*?nqf(uZP-#F}p-j&<s7
zD%1^;ApoWo_L%!#8<(9wWmXg4EN}kE;b*Yw35DB?ymQBNE|1*Yw|sOG+RA!#=zA=e
z{16P4g2JYFKfl<(J5Ob!Z>&<}c~UGz*LD8B<B$rEp2H*?CTm4Q3H6m+I~zPRK_hUn
z^PCGri5JtbcRH8%*5NQkN+>?W<1~JX4HbC9OEKpzq!J<k56vJPKZ=K?h##xp)L}FQ
z2D-oZ?2rlN&1|Y&MqVqICD>a)+aBTfahAc$!~&(~Dg{ZC3A{h<^lQ#!EbE4=OH#GK
zR~*C=|Dd%?k$!!<_{je@r0Zq+;13c&RTkQ#ESwd4^yh;t4pEic?JQg~&gWA|LDejx
z#~U-I+D26t1n~|cx=exx@jy$#O1#R9*}IB#98vqRyyL!GFIN86_QK>GZ%u)Lr;#73
z8zwGt$Szpv6llW@+1+qo?0mMTxQp?(=G_LzF+`Nq4nAPv<`n<4Cp~)HfuS<%((^&q
zOC%vIOJP-~`pVlq@ST(c#;oXxG)1V;6HqAM|IvlwtKDy(+*=f$E*sj(iouuk_=wRw
z0tWrSPQh78*sB5)gdWH`!gyI*Jvp6d9;`FezZQzqSY^p&vsQthY;`SaW-Nv`U<_ml
zpzhars)ph3kL&o1z8c${;qHwF%1y<q3c__<%Q9`7FZECHTSbu(>+8ZGE6GrpEVHND
zb<0;{P$NV0BI4rRt{B(DvskjANEiO4Kg;pLa{%JHSUi=on06F%^r^R*zUO!`mvxt<
z^WMxVsOP^)D<)9K1c^Lxvpn-R=oy5hcql2ulonh^D9NLeg-BVNz^NF}2o9RcH_}-S
ziS+Us=}fj!a~mirJ_(yx>CVN>+@F_Ig}@@DA7o}a$;Ep&dM9x!pd9h|WP4gG+Bgov
zw^>N*&;*1=cA$lOr0_!y)v23a^{B`aAg*>IHY|nIgKDgK`?*N3^1*1&ZTz@9qHo9E
zHWC)4q$jD+aY-oxhE}iT@wB_asuvEv=W}K+7uZ@<b{vg0iZG<8Nwh|){A~_$nLl%<
zEC}^X<GzYtr5dBD|DzYD*HcC-QXDJHQAVXPFLf-HYh^x8?f2~-j&{R2?|Vh8*IoC>
z&BBF^-y=w-?`7km_i~@-<9Bw`Cf>&L0?!EF1%;c|^1A2d`rIyyHe3``5ax2b&099;
zar~efxhDA}dD&&~{<i{48|_-a2jrjlcnw^|SENFt_dTQWr%14$prpXU%(PStX4{?=
zqVeGzcu^WjDu5NZl2~B)Xph(rL_IxiIZ-&%p-B0l%v7ub`+A~aF<$3PXn3Y{G~-rY
za5Jovd^<YU1`UT0Gip0JN!FnYni?->uctA?Uo%4-Hu}73me#T!qu!*5qx&SU8;;-2
zdfMIDTPsZY$rn=<{MjPDkn{&Wgd|rvszt4`*z&@ez1CHrIUUmKfOxMSC71iwHILBA
z)yQXCGlW(fZ5sGj+mH=u^Ze?e0>d$&VvrZ`75O7`R+Ct#b$xKA20$C;p*U6|xL!%Z
zmbq@+8nfe7`U4mTRdgMEutN2~@h+}fgd{jP^J_(7?M!wv@9oT80cP+b-A$E#WX#HS
zc1bT79^2u0gWBTp1eHE&q9R`GSl+=L>v4f*oRHMcRWPU$ygh=p5FQ&dh#(IotCKN^
zVTUAYO#VqNRUAb%fE@kuh#r5D0zY!{nCv*P0)Il_UKoqdvLOU9pUs)?Y7#8-j42Ig
z?bjSFejQ_e#!yA*v`J;G+{X2$7nmkA0titH`di7_*X&=}P4sHdx_!7(L4S=`2v&HQ
z{$Tr(rztJ8R@cs22?&{rD+HV_7~=q!<p)LQG*d5I{?eY(KGvaQE~b+QtSBR|tcF5@
zbgs;$yA|OI?Lab{<57c@7}-GaybKSQWes#FtI0r<1efN>%(}vZOyKWbqf^=fW5+Yy
zziRTOU$g?*hW=<@F!~MB)Qe?Ftqy)SM~~Cz!q<<Xac4><sB5t$AxI+9nvE242O^%a
z@M0<Hi$wXqckzCyPNzW9GJJv(+VuOpHyt^zU4@WQR!qYg7D8E9gCp-VrAUEsY_<@m
z{k;0TUtw(vs3CL+h|7&P5%nv~M*GoEU`Vp@H0c@f=D9J~^Z9e*EhG2N82g(W1{XzP
zj3UYR_g|@#w4AaokQJw47C!vJxUF7bC?<8tqEp272>pyb$@MiqqB%R*u&=h-mf&=?
zA6=tYTrwDH-d&oJ;vV-8N+8oa3t%k9BpJ+(Bzs>XJUdk!E>{0aj^3>~zegdJTv@|s
zw2&PS;z6eIwnT~RaGv5!HTy~CI(HmrONVC{#F%LJ#TA(b_-C2`vKzRhzA}CkidZ=0
z(2k<&fi11b6v?_Q?W;+_752~E;Gu(8q`V-HnrH!hhZlTBbz<7*1Z=Dn6xB-MwDuPO
zo4R*p*|ajvAB`!Cv^UNILtWk{V0_w@qqQlW=UDN5x3luPc-E{=X3f|Zk#4NCs+q}F
zRSRtfCZu*pk`5rqi?bQ9@>REFW=k867;<O}hs1{{U7gH^AP(Gt@YextwLCsZ)o+K?
z7T2+V^f<YnG~6!sb8cs(v60+62$EAAU8>ke!D0G!uGXkwCXQDBu%juCKdxf;WXP0t
zBokBPE{mbH9Plf!m_h7j!i7U#H+CdBa6H~v`BzS!B{i>A$Au%WnNp4RT^CPGjWSvh
zBv=0G%+Y7&G%<#vjxmTDdJb!yDlsr-z=NmO>-4e7sKrAVQRN`!Xo?D{r$$oL{S2%p
zvE%hV=G(;>`%-Eb<!8vC<jytb1gp%Lj~dUIC0<-|?Kvb?(tjyar~QFtogN?B9K5a*
zleb*p9-i&oy%k(Grjvj1z==kbN^0pKZ9c`pz>+oBFQsp?s45z~IWQ3FHp+!ZN*Nsd
zVN;e}Y^Atycm5ETScl{?{RRc%a|tBUb@<VGHtFMrkiHjLk-3Zl6tl>hG1+R(keAyT
zju<k<s9^+eG#HV21}Ch_i?R-n0vgRY#YJdP_Qj83P2*;*N9}Bd6pV+b)Y8_UKIbqw
zSy~&Pw|{{9r1I|d=N5i(rmh%gH3roOedY>iKud@G;HE(~LhR?-sD!t9Lg>uot=YGP
zWFm0;)ea^K!_7Cpma#;djC==@A?D$xXIPHb=8ilywg!^;a~2f+s&s|gY5hta{|bs%
zJjO0zh7Z<}<VWDg6fl$4siy+O8EA$6<|Lg*a{y1eB~$P)Eat_<xR{jgn`;|+HDKpW
zxNXe3J05pB`>HK4#N>AcR<RV9Nr;E^PMrlrmS-*f7COLoj)`FiS8`Pq%qfp>w1@h~
zzDS~Uxxb6zPf0li3mW$dJK~Y}{uT$WLCkDo1X=<Q+?XUsGAS_6zZ>|f)tg+UeM^I@
z8NX8zNjL4)1GN9938Y3y8#G+tTA^mbP{T|}bXcWP`1(p2-~7YED&#k(8XHQ3p2)a>
z{i67_!(9YZ$u1X1U&6E04;vV@%c|HC#-8B6Y?iB;hf%47XDT+bb1__(6Xddt_-2^)
zy*`6RctwpVW4!Jw-UG-`?YoHt8(%ojaDLq|jYT(;87zYdtUzz2<O}gA(gNdtFwtQW
z&|S}7rJ&2=<3=wv3#a6H=mpSeYT#0`-j!iUuA5gE_+bn&)t!dhNUenXmz!pUB?sw#
zWTtoQhJoh4s=x(2mwgDmE;x2l5z^%ac)y(ls7)0X(R?OGH76d>^X^kDV-wnB93r;D
zFOLIgax(}?6^N{V@iq>?e6`QJh8rd!@PAtj;1WF1!Ko1M_q1x&-}?MZG^<KW_$@js
zz^$SV=^)3mS`$^EgO@R6Tw$-dag9=ZvK?D5tDb}9%Q7_j(81$Q9M&Ld`PrJjB@VCC
zWTf83ui}N)pRwV*BtXhFyhv9C{BV2cHN_zrZ8^f036sO!^FJpJZ~oEz+I_oJ02zau
zyz#saOP1Kiwg&xqry6<|R^E?}GRv06E9D<~Z>LsAr^pu;LzfZv;%&v`+}S999{a4!
zl^@2!P;k328YL!oGeDMn9N}v9F%#S|v~}e~pzPQ#nn4+_Si}y8&a=2v0;yq=fb6;j
z_8mk_=)DWIHuGTUY+fcqbFL^g0$GQ)UTZ7)CrM~15yShqj+Df8)uKDndrFK~fJykA
z;&E}!Yz+_Y7)9KI4UauSjCf#JCSyRkM8~x`k+n1f8qWH-W2bNx5yi)qorecrSnFJz
z4xUX`Vt-AoPUaPay~B0zf>@<j&+rP{A^lX30xvX8s0=p&B6&LhfW0mQADKoP354!a
zH~<j7lGLb$ZPhSN<$|nYd2S{_6g6cw<o{-Rz|FuwGP@G$7#$kiWKs!J0KO2jq%jk-
zK3>z8%wAWosHuoQM%ZS<STm`(<I>=k+@Y!GS!hX}p3aFsYZ=wKP{5m~ag7`c<fs65
zep9<o@)n>N@aZRWy_&OqpL08e7bWyO?!qQM)nv24Xk)C9v4b<CYE1@Fi*X=%MHMZq
zx}l58;Lzm)Bb$(7&we`Vt|e>8Qj4$w>m2}$DOB~2Jbwq@N=Jw%zO89l_ELc~0~#k7
zt({l8%Mjs`!{+82Kg2BUA-*h6t;}*RJ!&`T`NXZiSUz0{AO1$ZYAnma@xP|2nI<xe
z7k;+(8WwmLF~~Ot-<xoJZ?Xdw8c-7q<=VsU<mz)d(PzD>0l&%q1`Z2fWuBgnGZ<c7
zU#=bKlSf>~r$&P-BYvnc3-i95a_qIU8_nQBZ4+QpK}zii;$^r#5(^IlWxlqI$*}Wv
zyzh`~mO=su9G?>jSGuK=A3qmCVEBHW9_|{pC{=%Ym$!F(YBhp2#y!1<4x!hyOs*we
zgg1mFjP5`bj?AwPXvwlegn?eyXPCHKXYpB%ZulEs<A0Wgo=#YuBGP<TcO10bY_Grz
zaHXdc>1jgV+u0&tt(S%AXdZdWbDb0v(!bBPveE9+>sP|IlQP-cB5Z?S8y}&``NcD}
z{K6*lnV=56mt#vhCAjl#4<Sy4?|_jad37yr24!r!XTmzK`&i-Fv)I@Ae!}rb=j`=a
zA<z3L%P5RofV42{YDfCt3<Da4=$4XnZhhvK@weLwLuLkvvR~r2FWnCWAaObv4x`6e
zEi97jXM_(B7l@FqOzUGFiyam>Sz3_K(kE><o_#wTaZ`yMDw(y|<{LAcv3nQ_sL;uD
zQt)lw2_%IBJ(_2WW@S~{Z?(4|JPt>bjhaig80)y)A*YF+SA#>YD*f?Xr(#PuUg#_a
z5rr3CFEui;8$8LGvzJ6ED<Mz-4-vZ)4p1|&XsrJReWfzrQ(xw%cnrx7MshzyVLLFj
zHL9rOP{$!*7SCP?54fxc#|7c>CPF)e*Nee&7+;Mh%=E{Li4G}jB-+@%0$_wML&EJ@
z@XM&oV?}y7m?fMVxnr!>;=rNcizj#+px`Z_*3K$nh)~!=dL4+!+#1bNawfis7eT_v
zL562~GE6CvQ@G{Q&I;P2%C2IjyP~tdY4FLJM>ZY>%6c{QDLz~KSSwVX!eNb#41H5x
zDwN|NkAHkrr8);jm#}HRZdrt|odPjkibIUN4;aIq8u<CynAsV<!Q;>B{W7>KYGSu~
zHEqWr^>Rlb?{Ix}Q{irRi69LT7k=5MAbB{+M;q<dUhV&GC5fsVS))>td!bC`69*I7
z)mWvWQ|Js+-yR@6phtP8_WOe2jqQ9qM3IK1*^!|4f=mB?IRkEipLRdL8&qE^Nawkt
zYMxf5j(gpuv7$1Y@15RY0~C`(Pa%%qeKmWoDn#$|GJzfZ>6=v$N6)<-nCEbS5x~Sy
zzDRXi?`2OID4{Qc5iI;7#cWX))P&-DVml|j?2wZN$<Xi1GdU`|M~Z)HtCtciW;D?3
zh-md4{~2l-!nQ($R@i%$v?Uv^E5*B#myBW{4BRz~cV8&k$dDgzghX$JQTRF}`|Cqw
zGUv4)(~3RW-}?pVkKKu5@d|b+;aFn6spI&|l6ULQG+#=dZYsRlS6N>vjP{V0JIKa%
zJt&c5m?JV2-y#+QD@?q*!yGRrM<LT?C<1viZzT3`h}(+p=dt=|JN9;i423yu^Ik#<
z9KX)HD#rzaUwS3gN+;8p=35;%lUXA7OBjD88VL7R>A`ZCU!v5+4WFXYN1LXsj`AsN
z1os5KxLEFvR&=3uUEevL^k7=T4l@{CC0+I%h-jt8l$T4)g2x|W4Qm;=ySMJ8nM1*D
z%r&@z$_$^_H3+#!?tt7rVb?v!Gu1SCh%5m|2>sDUui{m-X$a6O`UkJgA|FP_19xNe
zmvliT1TE2d-iBUtSsTBEKQDX6Ob$#|`<@O&l%5h%gvXx&D*n`h&%3#oqYQ?>Nk~f6
zf08lJ-c)-!V;I!W-xsd|CHmd3<%YvPE?2Z7l2wG|2<Yh<R(HGJ&Mrb(0Nb@zo8X;*
zgO9gpkvXv+`p!!bKu`Css|nk^kdhjEYirrX^lq0}TAzrq7U#+D+8QQzppWMr5d;;%
zHifb6<!jJiEyJs%&&(T>H1kgN*Bz3NVt)aEw=rhwb4TQ`aglncM=ZPR$RN%y^|W0_
z;a6=<^No+*c;JNZQ8NeK+`$;vvOdCOtn+k%Xy?uP2d@7~RX06vcD{2%$z|uq;}j^i
z`xQa1Ke;sW)d}VmM)(%ZUZ1)9boSEK#d+wL+dOVF&OfH8;uZ7?o@BfBP09XgB}xc_
z-Fy5@AUe8Rqn#Z1i22MSb@aO-z;~j^ix;^ve=PcV=BCRt$BLBB3TJRwq7#6w@_i83
z+vLo-3gPeBN=61i19;ioHbVF0SLf;AsBbwtt`zw~lGb0Sx!xXaV94ibSMAVDY89*k
z$&`M~C127w9NMEsbzk*O_GE~`n#d7^&rd{|N|%`}+Efgqe_*y4E}UH3$D!+-xU;Jw
z)8%o#j2bG)1|%(wPpv@eJMTo_j>zYA6yMG>H1n^B;^|(N%{4cN{dKT@R4xCzm%#%m
z*~MTMh1Y?*o9~NC0-XtHdHZelEFnnms$P9rhVhw3mbRR(te<8WObRo2LySnucOb48
zig+ZdYF(~nced*Sl`OF)iy#OPNK<<(a=H*m$S&#6`w{A^w=tp+EN9Y=5tQD;IjVc&
zQKd1zC6!}l3h7(_9E~gcW@Z}lXTA1K>XlYtcVZOl-FTRf&GHXl_b5-H;>M*+C9Z5;
ztV<R!nr^DqRdcX;H?uI<)C@%D*$lA^!|Z9dldZBrLl<BoY2diTPHCe-rEv|*ay_g=
zo#MT#-9LhIZNhOo_(O87eb=m<pB0A~F&D3L=!#;oc!C-;#rZaw*3H{Uf}hvQVs)i?
zlxHp~T`e#gR8mRlu#35uOT!LB!ZE%;k*7(6t4%{P$Q`vShdcXQD}m=nlb>x07d2d3
zwL!xVmGwS{q1_(nUHx1W_N19<$5S~6w?UTkhJ0{#rvm}g1r}0Q+NM(d>6U4dD<%O3
zE9#hpn$PuM>s^2O`b>YUh?V?TK<zqK1hmy3MQTN&u`v+)p5{S)<cbk+44*Efz0un*
zJY{+}e2n7@lJWs04M0gN{@DLTeMKH0bLL^Un(|PPlv`<^29^TjrAq+sc#OWu@vW`D
z0A|SCUeqChb8q<hF0}LX`_C~NxD_crcG}It<U#21x?z&^#DzWfx&ly#vL$i(zX~yi
zZyRVeA%HQ<uFL_KFx?gHqjskgkB>b5JDz=?VX=V48!^6;#?D!-kBy7Bg7eV&F51M~
z-SAm`w4eWE<)=VX1)!`2#IMvm^?Kfw%TEI(g`G4PxkV(0Vm77KG+T-6yY)-PC#>C|
zormohTk#yC-4H8(JDpt@e-ppg)4OkaK1BJ+@0*!J@RccnV8IQQVE{J7f}U!T)~?6l
zs_93r>jEAK#oPk!W{@Sopqpi2;=7%hSwetlSG*ad{&mJHRR~1$JX+biG*fL!PI&q<
zLE}<y;!LepwWTB&9VI_Jf+<u5Aku%^i^8hO<C-HN5|EdQdOZO31!h+rzN;oV8}7S^
zHfc94ECCgxJRHxVGc+(IlGh;RNK?nV9O0tV=Yb*CP??m3h(lrn1OSN@AtPTFsqyHD
zWv-A{h`3pPUYgB816Z4B0%ZAxHGgu$b@`O^O7%v~nOjI@NGG2*IXWe~C6MPhCvgT6
z$(5S|`vZ65X+Yb8oMAZlsaj+Xg|V-|h#{1VNI8k0<j>TGp>mMAzQqXq{XC}a#hSJ~
zXRpiZ$k;!ACSn=|DhglHuPK<5CxXnS1sHtbvikDA0Q#kTmD0$#e_ZG9I%kLvnP#47
zSK}KYJSn~z8>OX{x;5@VVO*tu=#qFzZ9Jx%!={&T#nvx$8Q5l_Pgu<^&}fv3uiL=R
z6%bhycmu4WY-fdoY&SepR9ighOOu1Xk8VO>kL7Uxy#H258{H0c<y+0<JaZ<<2IlNa
zbC|XrpEam0q-ny9s&61<hryJZb>}-%y?K;0lbN))Zi>I_rkZ`tVyG)!o_&nIuDE6j
z0&)m8c>HiKR}xnkn2)6b<&;jD+Y2#>S?ur)$#GjM&<m)G9?|S6>R>Rl0eqg5+M2!w
z|Jmpk4dr0JVdqHpb#G#;-{;|Mqp>*oI+N0*hnW2YJ+ZNi!Rd`krWtguW;+~6WXxBF
zNjy=+mZd3=@5RQ>U1PpAyy>q7Z|6~ii~8Xa?xmOM;g21Tg_?%?)PL!&zs#jN(vzp-
z)P*%aqFP!hIvBo;DaPP28xADY1x2cl8!>#v>4LHK&W&<&*1>xFxdCN}8fW#CWv$Dr
z%84jXk5^z%S7YIS=`h$DBV4Kk*&M@C>h8k<k#ngeq57`mLg7I7`uDN9$Z<abNm`-f
zXHIAfvguDh&O{~uv)S1=oY%S9)DsDvpU=HYH*_LZ3Uz#`2yk=bC2LfJO>&c?<*V&!
zB^-O3^v3K6ERt1<cvG5<ogON4UB?#H@XVb(d`v@eZ^1(vhl}c!6_oB`oLDzn3;U(Z
z^zmtK3Z#6!TIb!kizT}TNz6<<c<oes>9h@0=BYWzKzJav4)|zduO=bkepf2qLq_di
zFJ?Re4!JP)pCVZ|(|*B!G=u-d7mA3-jKBjj*_P856_RZWFjduRf8ZtGxcsl$I_+rK
z+Lh32ij(I~+#Lo<TG0$&x1X!?MO#_zRepM10XDJwSS7p5>H4%V?=?UdulKp!-#B1|
zR5%QbqmhKpz`V|P%i5ETIct@gr>D99TeueySOsxa+3;l+l7hg_r>6uXdM~o*EB^17
zn&{O0Gnd+2$X5>p{7a|O5r>}@$XXm!E#QNTzTyQw6?<*7g{-7lG=`ad!qX-ShOWr#
zgJ{13Ma34D`UT4LC>cWC(Pi=xzH}IzKrqr65JH4`OuS4F_8t}JQJ}WiRzhcvoJEtU
zjlxbZF@s$Bwhs-i7dy}gi2JAf&yP3AuwBtJccqiNxNh9rehY3s4;Rd`<{2HVwH)1t
zR`Gv<hOaM0xX_qVt@^X}BUlSFD7ezQy}7WYu9SkM*_PSX_D`!8r|XSc`;Zg|Eo0?c
zzWS}Sx11N375d(HcRJ<?Z|&BxpPR@nPS%>9z^CU%{hU>=9P(rYNOp<%JOY_~E!dQ?
zpPZ}YpLI%|;>%BgqHUEk)oqn+p|{vU3-$Yc){coD4nX{Nn(>U;{x;Hvtj)f>D$E?s
zbA)&suCV!3(&~NJ25wYb<9mbmU#j|ihMcX=2%El24(0?2K@yFpy2W>%GWkRhf6eU|
zEUJhi&$Zw_%Z=n&`P5KQe^J|VEuZ_yHfQ;etUv-6nWB~Jn;n95hGI@=b@sV9<%jy~
zeH7_EuJvX4?eOBue%>ZZl6GJDh@-vzwDOtY6h1(QnyieL3m7VvN!wdsgVuS&+mZ;c
z;$Y3*z+|U+(6ri{|Jh+BOsywmb{M0H8Kl3Ui?Av`vSJ!&f?!|9ZOQif!o^tSc#@>Q
z{H;bqN007XCzC_0FP9F%!b!%H>Y>YhLDkpJTVbh8x-J$!_yt46*m}#I&8GAz|H(vp
zF~}pW-)P^NxxRwZzRi&h5m9*Dsk}YD<5L$Y^wUFTWLTLhY+5+SgO!`~6d;c>A{F_I
z7rVV?_~3I}8prI?<tZmQ-U0sx+dV}|z#%9oOIFWN$7V=T5$<k-3BA%fFB#Lu8&Hf~
zvUz&x(sPYmJNQfV4LjWTY>c2*b7^GO41PnjIIl7zOsOg18PXi;c7K#Ul0EFPGfXO6
z6u=Iy?`|3=svkhn%SCaJ0%w;jg7*kDtemN!V@R1Z-LXM*v`9S=8g9;FXIsh0YtxCj
zci@y;uVtk9%cdL=rnly*x{@(4aOD{_Yz>!L+~Ro@?agydQ!NSiUD0Tkn6gcFmvXZ!
z|186+Nh|lW&V&%(o5lg*qpA6KjKsOtxGP1|I$_HEjI{0s?rG|x(G^;-=|;iDx@O!^
zW3!LK3Ky-e*;=HW&x(gLQ@T-mGl#MS-u|r!sTiKFz{>9&1_Y8DBX6G~eeAifRm9j5
zOt%KYx#fP=FPv1smbAgJp$)>gSp`LOy(V5W9on>jZJ4Rs70VOa0V-seuud8q*?_%7
zgC|B4*Y`&)1(FUQTSf=DSnXuSMC>)Ts@c>CTcu<t?eUtwiy1I%T5tPigw=Vc0v_kE
z5=UQr4Z8GFmXPKq>8KP9JlK#%9H@Qz%q)*paf$L&q4_qmNgU2%?zb`9-bYkk$IZcq
z=SexAhkfQ-k7K8IVXU5ZP62cn$Fxal9T<op1pq)2QWQ%n1+I98YD1@$vty_Njq_cb
z_KUt8V6L$_1=I5}k-134Hq`%-)0Jbv<~m6}nb&lYH%2-)G=!5%YjB1?FS2s}yNSGz
z7QpODv|OL9o-s8XT#t;_6rzi>_0{CGoF)2o#L-lTvWU}*Wo{u0lP~=)E>LL{(>WhG
zZ9#&}{Bm}*1~|MicWhJhuKxu>qmJb(|B;XN;<iU)xue3|f^1Xd7oTY?vB2w-X2e7&
zTMT%KVqg7~{h>oI`=kQJ9`rp3!-c$+2B3(=Njxk#0YflLfU%VW+A@BfrFets(@ZL9
zvsC5`xF7wxrHWB%sE)sUIE`kz^D^)MW9uCQGh3Fn;YlVF+s?$cZ9AFpj%}M0+xEmZ
z?%1}CiEVSfoU`|S-t(O2`?Y?p)!ntas;;W8zG_-S2?|Qs4G*jc)_{*e+AMwr#JfjP
z7Wbz*Hk@u$Q@G8wkl=cOg^gU3`v6VeZwC<Z0K#!E*~u9*7D=rF4K!l+!7U{@4fJ02
zc9RtbO*<l5J^jJiLC-UrkF0D_A4cBhz50t<`FvhQ5n~@lIT_-SyiJ-Y<Fg2NRqP&e
ze+ud@PTX%@&A2T32if4^?+atlUv?77sBw*B-}=pK8`3lCLZ4FGz5#hwJ<Cl6_cY@w
zxgPT>6m2W&EmP9G7OL@*Ua>$`>?e*yr_15fX(K23ikzwFgaLTH(rXU2-Q4EHrJC;h
zWzyP`8Wd{=-p38;g4l~TrnVB~@_q4afjp;eZUf7Jw4dzns9z<7abR4omiQuT4A>*j
zOL_WZOe#uwLU9uq(b_(<L$r^>HaAA*M5!(MuJsO(W*U-rsKj+Y^<8k%cA^}LE3l0~
z^8Eb}8Ulr?WXQJlcfocTTw!eL#Eh(JLcsgxc32=BrCo3fkEO%vS%gU1V}*OG2iS5$
zbGK1aGFnGs^p`Dqw~n}YA17o&;AZw=jPF83FlrFWlb;`BaFkYX@};J)ilREUReJ~6
zdJ^4Q=9CBp4$0so_yh=v0?Ev?8{$x~QF+;;yYmo3jbx)CVag|&8dKSFnIA>>$;t_h
zpF?-)4p<)mTF~jjYCY7^T@nPZJ=${2!?yY5E?_6{LC+&!Na}axI3mYoPM8n-6x!Vt
zGG7tCVRik(1t6>=lTsJz6VD<7FKyO{OQ(-&t$}fE*-zF(OIHlGm1>+F(4QPAxi?#b
zFEPdim7+j04|UnB7V3_I-<oP_*c`}ZZwcE{XFHabw;x%KR~A*5pqx&l?mc<x7JnnM
zB6*Y$KiOwsxo6oDgq%RJ70D=*jB|`qHmK{S?v_v`7Tyjgu`{+crnHRdLTH;@9u%mp
zo=QUCzS6H1KmU6v70_q>J+gn#`YaF=4@V_@BOw86--j=bY0ktm_3p6kl+zo%1_ON%
z8r{`G-l)caQUY_JkS~FEySS}|CdM8&`Cu8&vxkI<jY-GuVQC?{+sj(YM9HX|IsOMu
z#)v+B;Uq;gi6@1cCm-*-_s^fh*e`ci60=85(s69+LRM2$4b!P_2A4rMI+6Sbwmt)H
z2H3N4GAsof9;?mheH7X+C?U4u7$Kz=K$eu5j2zzGm4)SOD_uzrOzS%D0j_<c--gWf
zVhFUsJZU(WwONma4EF1TUE>pPXXel4LYDd-_w&*1fg0Lhc6Zl|!D$kdVtdRpMyUJN
zvlDIE@{r4~)m+0U4vh6jcXa81s>|wcsx}WgUUbgx*uD`MY1Q#XI!mq*{62%nq34XU
zT7$!1q#{R?zQ|)E@tckR4iQ-`gS{YQM&A@!+@z(E?-kCe@FJ<pDxlq`jN89fU6Vw-
z)eYUK*)|^>jeArzMhmeNHGFGz_s7BPA!w*~(ol{iL&-Ef<StL|=Vju+C@L;u!5YH#
zlu6Fiv=+00v)Pbe+Ix%CHc9+;5=VMq{b0V5&Pv8kqskaqIyI_6gf+LC&>%;U4r_>l
zlNfaUtxnkC^gg%;sRfR63gy8&hIbxO+oulLv!biY4f?fjSHFr5W<gPuT;AEfZz_s-
zni))USaa&>k(2ne{Dmh?<HiaxK-lVV16;t0p0y1Y#tAWyKUl6UsTM?$+Se6g!od(a
z{2ASTu@3pjo}L;cA6R^Cejli-GYBgMP&~L$aR-FfIi(X!aeOR$SFUhx*Wf<WCgX+6
zsvelHDB~ZPOy#Ky1oZ0Iq;C+?w)QH!v-ATZOL>CJss`Rv;efbK!35@!YkGY=HYDNa
zy$%JoEjuM;U(n{1H$e}6NFYNipzTU2pe3qON1go<L1JPans5z5C`r%Z$MX{<RETX9
zz1Af4^IS82q?<Dih)!*b(sfJk80UA1I3MV7JPXOv*kbAwZN2~4gd#RMHpIr6N85+T
zE~JiRVP=P#Zf50jAL7SxGH)UEqRFnip)4wmEW!bD`1GRPF2YscOJyW=ebIYRNBsfc
zKLx(_U~{MxUgyWZ-#hmvLw3eXM{yKsS_@XmQ*Vc=bTPaZk2Q7~FIodRNWrGJNEa2H
zW?-92OElStI$v`vnym$O@-5tZ)X?W=d2eA+p;!vQ_+jYthggDz*DYFpybs2Ji0IV6
z&^kvQhs~uSsK)z6@AsOJop;0u-;s^~qMe1pU4ce=*nxKH*WUo_-Y~3>34rW^lluLf
z^$M^Bpu84ObR?%NSH3lZ1tDENq+D5LrqK-wdtlH0AW`3OmHvY)KIiyg&UhDTt7IZ6
z=t_n1X1Km{)2D4LoL%b~uaFd9kVQ<DFH7^1Z7RWwqSInXR*oIgLS8Te7-Y8Pj~qv(
z_tce^Kk3+x+F13I*=E$*EB)3zCmSpfz++qmQ@?jONfn%vlCgAPV!CJBY5++Npagq5
zQ1sU0n>R}4!%7{OG^F5i3c(RlrRu=HJQ4k#-00FG06)`r>{W?kCTiPoFvVSj;o(KS
zDG58mci*4=XYe|~XI4J|2j_0cb3n4{xo=>n!E~}^1wAC{d?GO{;<|waxprV<PfZWh
zH*&_%HXKFXo@Q24<I0<D%u~kj<4OzK@HuLi(vf;l);(^Ms$rn!VKXq!U1Unt!RER*
zz@xkEM%q4Fy(BZOt7Yhuv?m$gv&k`JP$mvTUDNBp^hvXomc~n>YZ0OrC+itF+VzZ)
zV|<saI;(VG%acea`)efc`TiwMBw<W}ackg7B12!}j=FQGq2I6_u-~ChChF6UgJ2fb
z#d{rd-i`tHfjK68jn~EWFRe(e8YV9Ci)I0G_9Be<%S4*b+5oDMU4m8OT7v3LXL24n
z5NbOlq&1MKSxjiq4pa}OI)+uoad-JR|3tYT`FR%dk=Joi$iE9%!jP?3;Jz;Hksau#
z$y^#KNO4t8_P0-LQsJ6<vsr&K6wEOm{$${Jeux<u;w;Y9?7WY1<p0i1>yCw!XbWAq
zO+zC;-uJjwXy#G~HgUFwW^tpR=T|FRw|!9)nDFO;71Z+H<De7VJBw`6*g6Zn-E{4*
zX!MsYqQX6S#~(f4TEj_siNs=Hg0FjtMHS=avJNEg_Y^#ykE@v8gY&gMM)K^;=hYy{
zvfsnspeUrBTaJ3@9BZTz6{^^R-#{VXiVrMSMD-bXUY?W_ZH-AGzm?FjE5l}cXe56J
z{wJf&-~2Yy)L`Jzu-n}MTL=hSp7I<o{R!C1sw?oW5{cPR8c<BkbPwt4m})GoldN%@
z<T`ibo5qqAY|ad@xQ}<^Beeg;_3$s+oI6}U!yNpF%N0D6d|K=No@5c`dWQ*@w9*m#
zSSqp4s8Gtv)IvEyUvGEPh2-M!f6?sxdr|`Oh^Yi?xbPPnbKT_~gR=sw*P=<k0+i3`
zIa8}6idf^lB1?$H$v;Ua{+oqIj}cPF3Wl{Mva&}+b6O%IrqagdR&hC1ty7=Q0WaBh
z1QwpO_R4<ZK-&qa7J&EPuXiIIF@k#S&f2)|a(;(le$a8@#x)H#AssI{ViIGgH4O|t
zdFE@KUE2;03GQ8#Me!paotmEhrK9}xbmwRH9bs+vjFbES5ChqPjUowbV`VI39+=*2
zSoY0?BS25Zu49`nnGR_+({aPHQ8U3&3$C$QLlXB1>o0W%=wqAQ@TO(0u>~P1h6_3D
zyN%DC|1Id#FMfM%)J$EA>md~Z>z?R;uR#^sPjE(5!Dc)%Ci=KdCyur?TstnRYb8Wj
z&empxJrJ2rR?P@yq^9Cc?@ZXxSwk*8GaV^Ng5{SPf=+L*@84fpqyja~a&qFh4J7W1
zuFUAoJx;qCPI&EowCP|!<90KvG+DXb|LcEEn?jDu$cG#YL$InS-{7JLnK@WcMZzPD
zy60*<Gg$eUgtF#ozOb%t-_5cE*!>sB5BXl`8cnDLy%b&K9%kUen(D1@1D!nj%D#Nw
z<BvZ*7J(Y=;x^)WJO1nYHG_zlcv$dV&*_IDj`G}W>vn7J`cr-*SDD3G_-wYA2>iG0
zromlL-G{y2F&CPKEe~gZZ|*x{(0oI~R=DVM8>^6v%rxkX{SoM?1}}AIE25U>3BKfT
zI%Z}y1S`MtrTC=0jXa5w8^z>~l#Yj``D~i^P&Cq`ZjzVT_KV`R?_2Pzl*~WJLF;HI
zqyFN9>RuK9?M&awkm);ky`90niS==iuM2zm=Zw8i*ZYX-oVdvf7XgF+KCuzO|8V+m
z)uu-}YG9PLVTSa=T{v1E=+H(joyeak=UkCO<jhJeud{SG?Ja_uX*FVt<+aQ2>(e)o
zY4~J(K|&h;XH;y#-@Wzni<vxk{rD<gZY`&|Lw)Zt=kUAv#M`X|8yktrbn~1ZhivE6
z@FVCa;x6Yfo)7XSPP?-qk)!T&d7KmtGPSaihgr(nO-#5vMmU)M52E51%g7w{qwWjh
zL624w+yz9*9~b|UeYb~zo<vg%o}Og}?n;_j+_pz+J=|r$y%CO<HQ5&Y5S#-+%jLuL
zFR-Z4US@JUM7)VuEn1*?pSqst@FZko=vdZ2yXgyP#Y_D;TpxOKwa%YiyYDIzn-85+
z!B{KU+6fuL6L!&e_i9?$ba4xWHYgj7k9y0U#)iLnj@3F{EuXFymMn5^%El%U=sTY%
zJ=RKsHhk@<-X^-dB$G0MALF;Z!PiI~ZrekTyvQFp2s)J7KNLEvJ!eGI$tH`Uqu@it
z_pN3j`2U>S&1oZGQ~PrLMd?+c2v+o@YHnmAD!)R4V%da8ccVnTwyhA~6it#~nPG&p
zj^Cna^L=%G5YC~RdpnZBX)jLvV`R9<+Jlyjy=S{;!p<rt1HCfreY;*cq<21(hSqvn
zCvt4-pEW~*HoQ)8oPu#;v=SaI0*f<|;U)R6YAT>*C+2%IpN0+!@?C><Jyu9(vEYKJ
zySyErPGh_T&kr-bWH|1ZiF<hw_z5wn&ftKXw-Rm&ZdzV)c{IF~qq77Zw_CltZI2=3
z54ey^f1a|h+F>T)w?4MBwh2DY%gaP~-^WpUZ9^UUsb=|oE&$tEIZx>qN^YyiL#Hks
zmzxgqec_6kbe{}H|28BQAYkz4Pyn%m-Y2mZYyjC3_<qxg2InC&)uA8wGd!ZH&uqr-
zQ#PVaYKz}90}X-F0>gBe0Sn=jb_`$MhppRK#$DR6XvIT~FWpL0EDa(f^(Td5taaEl
z;>Yga3T%S(bfvO13Z{M|3r04$`7}D#`{!w5DG)>3DOxUk0Lyf29iz)ry7@Jid=EM3
z)hX#{5`sVtneKpdKQ+t3)EKgGzh}tx^i@@{E>%zXc*l0_Aj*#xUd`vwWaC6^tM3E}
zn4CwVeH*~-gmz-N)g^n(KAL=vq2=b1iTXdw{0cVsWF?Q&_haKbo`MmZSPGHIWGfzE
z%ha;5Mwea;(gCY?&~8s3(6_zDUcU9CA%+hA%;hf&?$;R9P5a1$`w_DK#dYXC#(eNd
zs+r^v!+78@#^)J&<HCfq$oA=H#q1mK(Ti%`<FfB$JlTR7XK-k|Y;=s>rd@c8lfbD*
zgWSYtxyAk>cj&QX?Xxt5*RWr@j@MIQ3dbIOc=9WrYv85K@E50uf7_Q+Fu&v8hh~E&
zmG_Z0!u#xWX|jj#1~^lHc--{a^KW14Uiy%cxVw(0oe<nS2g-++{@Hh`l%TGBSL#po
zkKSZF`-z5qV|4Cie8=i?$w^u4?K`tqYJZ%P4kFPqA|^RGc}~}rdlf^hp#b|gps$-+
z8jD?yw;S8VzB^bR@J9wK7LTNy?ER4&%%V^)y^X*lHjZ9<^duN}hT<_`_{YPeui$xs
zAA*M=e8K{8FFNq3wD`~8wzonOY{wHCx;6R$gWV=10`9W0<?Zvzl4Sw1nvF~<#sE40
zi$%eQ(_fBT{S@+7ztsvyE;R#x`>vneZ5F}_RwjGb3;CT(^QX<Hi;NuZJX`T_o4C!v
z()l=ZyuZPn1oQqhctN`{JTz|)mbwbia|xaZfdm`x3T`i0<biY1{6uH#O`K=35vZY;
zt4E71ZsdLEkGy3Vgs51n;R-Vw{|Ic64%G0tg}|YtoG+-tLDWTMt^iC0?#StHH){#w
zcm8>^^pDp&ngj9rkEgjuFHPnlsj++hPI|A8HTIyK*0<x6K`5okQ>~9X<{u!dp1&2I
zmeo2jii%A8YD0iqMT5`TxBCuXRpaP*UJL9}DFlIAf6N2~nfex1LIN?$1ySm#`c*k6
z;alwg7RMH3jrtsQf_CIIb@1fSU%o@nnW*~#n7^07Q*p4YvXYE~iaY+E7i*`&mT~}F
z4Do{#=ZAdtWc1WY;>6H~+v__>AyfcUzqPGZLIki;7C#;8k@u!x7PrySl0$G71!vaW
zRMpV+f3(9I%%d1|<Qudl4%JlFUk*+{1?<B6N%CEfjbO}YmpX^o)Qili@NnJ`0f9d5
z@~!6E`FqZLQC;8U!U`sxoLjJ#`NZfXsRQ3IYPuo(R?(qvBE|c*Cx6rZ`Q2{Je0g(q
z*ZpRy$Q1bdX-Ze|wS8DGz=(QNRGyPF)4f-5TO}ojz~|I7hX6h$<Mbce#R__`QL-fi
zux5DAN2iaT^W|!@sZNw1ccEK{d&I(y!n+??zcBw+@lTkr?0OYj(R~)5vC<S=FUKWT
zJ>auKdKf%5$<Mn^s;1j@i=w+wZ2YRDQMsKVkii~Z4(b%B6@3*Q<1IRcXq}j`?J^3N
zPr>Cvof)+Iwc^T1dL6;W_q*45@Y)2mXB}^J#)-OyLBvM=UO+{leo}b7_|{WTR;MVP
z@YrAAw?YtXk6REY_yleXDm|n8^wD3_n>_T~f%!%w_&Qg=Wu5lLC3ZPt3ksX#vi<ge
z9oN|auxEmB2G=-at*<vUVtu^+_lcC=J#n{tPLzY}J=rrp0$}R_FLM?kG&Z?!7Y9n9
zHgLAdURusUyhEl2Bo!DI<*q+gZRsa`ZoK5EY0QnRo{2|S4QLwOhhlbMJ@FeUJxjOB
zzb^g<wg~I_Eb;0O9XTz#JR#-sapMH<^_4WPU_A?*lz63jlF0l!2O{}LdAF~02R&mM
z?iG)H^VjCm+wZoQ`3R(g(}|_Q0hwN}j_n2AzhU>M3XNx<V+@3J$py<(TnWxCm(0%P
zuGP>w+BY+Co2R93R}1s2pf7Ly1JnJ`Axn^1Sf)Qx6+=2JzFE4u{9gqBiyg9INxd~f
zIO!Kii(^KHkAZT^sl-avPfOV*;^x0q<v)?XULm5{yefK1xJVnNoyZt=c&92mh3BO|
zuAIZ)p6EY`JiC>M85p=r4}FuV@U@}R6UyZdVZsrSZmhYSazRx6iyrA!ursUw659Ww
z0L22mg~)~k$oZsux#-h5o)&&aB(*Oh;@OQe!ukh(mmz+;@$|v2u)o3ne<C8Mu<rIF
zTtE{+(UShuf&*#Ep{f6%@L|kkl<hL=gA5-c{Kj)5Lyf&@DuHZO!6}ox+`p07|C&Iw
z01D_=+^#e}L^j(s>}bC00+T;vDM8#vQVI*d*Vu3|n&fct;3v-D?ng$-dNL<J?S~X%
zDl~*sFfo@8#z`G(<7j~h6u0Z=!K$B1sDK9dG!@Rx2q|l;Y)2sEpQmWz{E&tt{H~N5
z%_KRi?V)~ai=Ek5U2eK*&&l^bBo1jgDW@!fgfX4#Q!Wr%&^u2o^y{*eECA&Q4m;nH
z1?&B1Stvj(&WoerUeq+}jadw)roY|L6RZ5%n!E2Q7-gSEVVB)EQWgFxuqqJ)M*Ze2
zeX6@#VVS{29g6zjld?g?jJ4q5rvWnD=u0xPz#7)~=^GB?RN6doR(}4F3O=w3Jj{56
z-d6LNF9ZvCX8GI8;;w5m4KcYL)z|D)JEieOg;Ftv0j8g%qWE(S^^U}LUWCargYDq~
z4*ACqShr_vX9mZ#HrXZ>xbC(ps`kyQEQ7X$^1EA9G#~Z13#*|sJuM8pNnX3whciPa
zG`k2PCed*NZ=NcXNS2DDHwbBi52+@kXO$|wNfJ>*$B-4oV8MN<<EDFXj8W@)C5ywA
z#Hl>x(cS4{zn0-?$m(H@PG@cqE>o9fx0GG)wSLE3>*!r&LyQ4tv+FOogB6h;Nvvx@
zQBy3f$n#n#c6NrMb44e^<$JMm@^E(!i;zGDrcP*6AAVJ7!a5T2s~OrDPw7;ABwcIM
z0X+u1;<KyN{FA9s@MHepZe=0*tO-jys`Cq<-1}rg8u!@)2ab2dIjh-r{GL$>#0wqS
z@lXu(=tyFu8_UZY;bgvkBzcZDvSP{PM9Uszh9?Md_Kg0iA0%xxMTu^MlW4x!s3+6z
z5tq{FCqW-;hyTyk79b!(l<fxM9Su1p!<ocK1eEU3bk4juIuq&bPtKjeOi1E(ZcvHL
zOl>Nr8Q>2%^_L5x7fIA-Z&|ZWFnr0!WOmeDW>z!9K=0)F!A?!TkK5`<n*%!P4(ptI
z)`Fb4d%*Tj5dG^CG1T;9ZNh)hDLE_;G|<x?eaq~<g$t?1;n4gc=zs)XmyCN5cw~Ax
zY&za|bdFUGu7P<Ah?&aMMBe?1ulK_Vz#(_88587qE_ST$M&myGgeiE@`bi|X4$uC*
zY)iqT?{U3iTvLV~mIvWqp>6qBzea^ah#@08Mk-$itH{g2iLrftXU^AFk*!z&-nTY*
z<rNERTm=Tp_Gv%Ta^cEEVsMKNY`%eZ6H`llBO`5i;TI_f3}SX;tgtbT2Xg$q616(i
zTb;@WB6K}d=jNq-rF$6#p!(#<%iV9LWQ(?>L7TZwIzSHLGuzqlfymv#DWME+@bdlS
z(NF0eg>RNcu3jktbHCOlh6SV{)cm7x#Pw*u*QB5FoR5TRVREpreZN;a=oSINUj2>(
zS^$N!iQ79~<!)`6+?v^Lt!@!meoYUbzN5hA7QRuo3ZN7Sw8b!0`4Gdnb@(={ps{jj
zuvbL0Z;v(ffiCYgN%7$7X{gQI?h!Pmlm(?J_pL-t9W%$d1<%IJ-;x#ua&Pcb;BZr#
znS(AMnYJ#Fa%#aOc)zWs08i`Fd0T$-ThKaGbW-?fk#MVT;tE;si82sUHv^SoY+?F2
z7wSg+>*p(G(TJgGP7Y9GNYvsz4LD7xIoXPY@hAFx%Jq@jlGLRQsn|)A@LA?ITOEkk
zUqdwZ%2=a6#E{H3TxLgkX~CajtfUSFf5|36sb(7V{WcbuCXeR4R{ld*5N(Iew~B4J
zX7F2^<*CtueN}&jh=V1&KzmMnl~i_c>F6`HP(cQ!>YSBwV&)uwZfoL{q_JrZq26fC
z)PSJm+xVUf55Pe|(p)o=n>Oo^Cl5L()U9v5v<}E^xRGxSD@f95X=nhjYPS5S>`E=S
zN-;Q9Xk+F&R|xPOvCGp&d<2ky4t$m%V*jAFJd-BEO*-d^U>R_tiPzo&*L9Ubght5n
zwpMD2A1;H);;Zkkf5oEQ+h?H|jy%^RiEBnOIk8`{rG32-2Z~<gX&t$6L2<Y864W4C
zs5ezGxP3os2o%jF=QD8$A(eA0$C1aH{*}5wohd(C$y>1qNsrfDFbd5)Q2!eEDE4u~
zvIrmcn$4mknUL1nJQd3QwY0VmN6l9ZNbV$NsQ)iHuz{(1W))`}E`y4D{}>QijZY&_
zU?a5w{xak~9bWVkz%e<8xd<%uW*7+O%C7W>=`vqY6<;$MMRJ;xAGhQ-v$?u%0zp-s
z+@pOOE$lJ8KmaRfa*>fR4PT@|{+Q)x)<6O?>k{<Ml{#}E^+*gTLzC^2oO@C7FJKke
z)@S%>G$9+IM0^R$gvrs5LBcS{MH>()gCqq&Hqpje&-c&Sm<o@tU?@C{3N3bFAgu8x
zSV^alf@Q$$9}l}>_E#d2f(d{DCvvo&M5-N-OJdNE*a7#2L5SfBp9TTvdX`FI_2>Fh
zwtunc6W^4!aa)#Idk)Ur#?EOfK<ZQ<2+7eY4OvyA^VK<<4KzrM+{9%xZ~`QahBJ|x
znek@@d<-60TJoQ=yWvI7A_wy!v(w&K8Lu;B_Big<TE$81+863S=Dubq5Wymx@ei>s
zaZt#s!Ev+=aW_OJCNUJc@np(H^E1BaAwLsCauQ0w^|>>zc6UNuSQZS9WzdF@*2aWf
zZs=i@2codC9%t-%q8_N$)6L_tyIAV5imN#f3Y{qj@w$c{45-GZvPJ(4bYeWOe*3Zk
zk7B^rsM)Df$(OM<HoH2~MHAVw;6MHVWE9=BZ(RkG?kJc_kp5xOXk-_uHTV{(c7}dS
zbL+eJ(UxeWO(`gBJ~a6SG(f12hS5Yln>k1YQ&q!E4->LqY(q3G%vAh($xo030v374
z+69)z$`S@5a{@s$c<m&Vobw2_o?~#mikj!xQd_nDm*IWr0s*)*7XuRenJoBY9hcrk
zybg09eIYBT?*q7>ztw$ksUstps&`gOt40Evfwni2M(zYwl1jF_#y}NbqGW6XT+JI3
zp+N!;7Z%a>afUjniGzs*YFaMivBI)S+@dFU8wWiKNd+uIw3Jt#1Y@qjxkSa)^oRSJ
zUg+bI<O(`>I`Seax!Ba+@R%N#=z1v@p_YuY*T0xRm6O@l=L@8wJDNP2`syNmsrf`7
z-Rw$%cZJYuYfn@~EtqP+%Yunnq_rsdVy~2E>Wf|=9r_-GJ2NNMEOq7dZ(E-O6LSWB
zaMe~>`W;gim7Kw~S!GVGaOGN>^9=@?8)lh*@mZg7dtv$4F3jJE+($E(V~b~oY;O+=
zUk*GHP04+T*&2_(_PkgvbIfDL%nZoyMkyIQEO&MA3L2XkiJ<;E?nc3<;A5meV-<AZ
z*cdjqPit4l702@p*EFuOeaGBDM}fOg#22}kU$!zMWeGP^g(m668-aX1pD0mEqLL3F
zRsyL2$MoYJz!QZw3{PDR!dO#%8Uv?>k;!LzBoNn&`d)K#+Mf9JJh2K<+!}lo5hJyx
z7e#(>+DFd}!i;#l{~S)uIt6Pjd`#+YK7VAfg^9}*zzIg7{lP)r4HQ1>P!2~f9WD*O
zY?&He8=X~U%~4U(9qf{7CYrD~HQ?TF8{b+F7G0FtgoU_`ux9{1g>8I5TKb6k3<6J>
z4>VY1y5<zC;Rt5KJ9Oy@!{%1(Y|OC)Y(z>qd+XLtcJ2lVKgmn_7vWLysAvG(cQG2u
zfOo_|j?4PQ;7TeJnddrJ+0^zD(D}RugMAA4sh{GeCcBFihG*#Fdfr4e=qK$-xZzN1
zOUvnO><Ma49igI4%LZly!}aNzB<@k0hJ}=ca`=nrH|R6$(@$gRV%-rP9@mo#yU@zK
z>zeru+fplf-i0=7fOE{D<^gKq_jLNtn|7IEMXzgOA;?Yws>NlPr%dJ3RTkJAH-wGO
zI+JH59<mcjU7RbR27^t5E1qM;3oprF7u%m$fjX)?z4Uf=vt)-zHYr;&k(rl5coE@<
zv(sj^pnU{zOmy?~3hQUnUosWnc{YNDt;yb3EVZ0m37*)ZV$d`hh6@~k=`Mm^yA37P
z%ux3#cpnG(7V>_6=5EU{dlExH835~Er^Ld5X-yeWv+z?Nq(iiwmo%eR7R9ifr!}im
zH1qmIp3zG<JDgrovkh8pZ}&2~n@UKJcHTXaps9Vvr3Se`44*<c^R|}Rzk#6jo7`Ii
zEoJ0Xg*TzsIy@=t`x27+J7mKVTMCU~KZbLqG<poP3)l;uU)rkHaxDIWdUQjwwQIy~
z<rsQ(Zg@%?M(b$m#A<ClTvSF>bL6j)&^fI=n~=a<;bVq@0Kq_O;+f0a1|hoWylJ76
zbH`F&FFwf5!Hl*2A4nEb6j1REMF4#T69fF%1Q%t89lyzLyX;tlR1WMhYMbj{ltuP!
z#5LH6>eSBJw5<(r^MLZT;I2j#t0304M*DePn*?TNL}!I-+SG%s>y;wyQ_MvuTt8!h
z@0a_@giK)uExUkhEq9t|fk#Ly9POQY5^b{5GElHJRpDp4{IkzC9Hn=IB=MN8dy4V9
zJ^ak&MX~)-5VZI}ren3aee=%{0j^CJ>Y;+h{uGCKcA8OL|AFn+@Fy|p7kO*NfpW4X
zD<JdwlaYx?=ds^TpP=a^$I8OW(`kjBhcIpcVbuXRL<bqWakonIbvM-qCb7A{YXz!i
z`)8u-1>2AH$6-BSc{pKS-!h~5yQrF9!Nl)Z0}~4mUZ0ym*@Pm2>j@?q?W*poVzn#f
z$MZb2xB2YY>s7rSm#7Si4r{ZYLw^O_>MWoQdoHvOd1;&hQjj8!K}knycr#B(&jLBr
z2I59`zDIFh(+u{BcQ1-VQ?N|>&E~R6=vY6F*tv|Rd%pXvr9AOXWlkS=3-v~i0J)!B
zf_>VAQ4y-DRE(v-u{^RWGjlSNdmzG5wvf|PUdck(M_m1$vc)YzBYI@mju>Nl6ExPb
zo|C6*)?r0k&-#KY-o_=|GLXpVHt%;sC^VTMm9Qnr&1cgb(zNl3t7AyFCAz3o($yq$
zc$}yEom*`_qfY;B&9d7Dh!OVo)GG-x-qeQN{gSaVh6FPE3_(28HQWTEOf+mRB1)Ft
zqoh@tyih#hjTkkrwC`!LVF=b^(9YSXPmR{GY{Ji@{SiPosOXQ=f1g(fA^k=M<NjJe
z-0P*@tdNXf1p~JDSwyQ+eJ)dhm^5?zODGW@9A-T81R5r#cbI$Yk|AMD!K7cF&uo~+
zO3W7QX0J`40+RKrf6FY=Jc3+pWl>b}6cs<nTnZI;7VaEZ?LZ1wIZ)Q!GSR%fAgSCf
zDZ3royIpFLi1XqM@(vBuvU1Y#10smH{8CBRxsJFJD0-zaleHq7C4W-aBsh(JRGc^R
zizsHNtqCqq%Y98j#N@n_m~K(NVbmBcT~&W$<1<t@QVI*^Ws0(m3jw5n`LR}m-Gi~3
zA1i`8{nm$dXZIXYhQL3!nWov*=q)Z(Ubvdh9q*5{i`X)3o5^SROdu{&gU)~b*x?%Y
z#NkP9m`o`wmerInQ+`>SQM(Qxw+;issZ47ou?z#@de2tqh*iv1gIxLl9b>5?gQQ)$
z{zs<d6c0qrVzCX(?E}y6KIoIu|FpcF7x+))_wVyvTF~nup;yYdibOi#u^LS>PKeFU
z8Y?f}!(<Wr7iY-Nj0D8b^))1QDzKAWGJyO|%TiB|pna*TeF~NA<kIJAlQI07@zkpX
zlLwUB!$#NJM1KtzuU)AeosJ<%CIuF-+<UaEH92>eIaP8tJZPA?Hy$h#@C`UJb;zki
z;ur*3sZdU|1FN!$5>p2#n53gb?=$ZZjCpLRq!*fMXKI{9Y#z$=w4lnq#3yioO8a$N
zr9un$Ie3a$fbabz`AWq8xtJOs#LzXrAlN!||HJ`iv*bj{?M9|kqMT8Xw(q!_(Gsna
zQwojSa|)uZba|5!7@9g*0nQd-_QDmhl@^67ym}OjRl`m(fHp*j9D8nC%CY9awRk#>
z7GTVRH=dt!boa-V5hm^^+V|fAP}qs^D4$IqKF54OqOgQm^r_+viSAw%?G8S?M)ADf
zKxA<Bd^9x$^9W1Zgl9*~qqlR3-lF`>nd$jK&i6J{J-~+6G4srNaT43GJ_CzCBru_N
z^(k29Xc#OI4n7sY+h3sA<(GW?@yKUd%#RMKhBb6&ZaLMn+|}{Xlo_<>4V95bY0=1W
zH&>I^3jd;=R8Cx-Svcf}bmQr0i{`@FATGPij5gkwQZS%6(c<^FkHl+sc2BxMR()j@
zjrs~q%iDUI)wUp7zC*a~Jn<L!9Pmv>Z}3*HV&RV-%b%=TPy(qvKC5JUX%u6o++9pB
zg${roC)m*t)nEv5r-fpb2Zg``<BWB!4l?KjFIvYiSfXIF73F*lK`ScbdbMNA0r7PW
z3saS<k$(L|ItzsBzC<u-p5!NXIRImsi;w7PP^T-7IPfmWiFMejr+W4jbVt`IQ?a+K
z;VY86H1#3YOX;(1n!=;{0;FA0;X-nii%&V%T=r9^@Xw3JIaLnnJ2U;_s1}TzGB7xJ
zgi}4U!k0^fJg$y%WCbP(Mpw_K9m9}$O%)3maR^}+$2%Vb@cnO*93mlEZsXk%?9S3s
zQsA|YWfkKv9q?tSdhe2A*+SK$GT3sR)+J(`NcKGxF@h^X#t=-#DAcd&D~?*Y4QPy^
za~2K8EC*_#t#^%>_<2mg*>2M4ftq&1R!L>3wQR1qyn}ejywLrEHHp(+x90BDTOm;y
zgU;mSGfKCC03ti;V4!Osx=#U)J6nCC+t9xYttHhmH|4fz2+p$^FNu6_sRvPt{TOPy
zp3OO%Fq5!l8EGyfSd$N!_zp~`n9}nAz=pt>Yl<5{nd@w+*xnM!^7?Aw(2ZJ<OX=(;
zOs8?4$+;l?6MgcK(4102uvIgSV}z{2z!i+FtLQ^{bj5}`<pcGOs+X=a)+5*>9;mzY
z7MD4*Ba=5x*7dT)Zd^beVi+<>2SjhHj*ON$U>}R80?sB6g8tM?8Jz7)5&2*Qyt|$H
z4^&LJYcjF3+Z)jV5DhpAi~6q)%*$RyETns*1&A?B@*Mz#A!bnE9j0FciFC_gxCo^n
zjyOsj8;Pt=GP<WwH^@|;72CJ7-#>>3dAMWg+g$mJVl1SxvQvVAbjm0$p&1Piy%i$a
zDMXWnzo9yZPnE7OM4xHNV|>cmU}U8gLCQ3n2B5MqAIo?k>!+QDu4xx_WNfd!3Ey@l
z>w*F=QHw$|qJzStvr=gS_|+fEk3~5jEjN>?wQ-DqvI{r}KcMOamHA=Bka(mW(>$rW
zh^&hrRo!Y{n8XL7p=KZ3JIMh{JDcu4bc#O}j1&Uz9Pik^R{P}b-q=VBflZ$ELp<^X
zixA5pD<+$+H$TMP4U!c_olm1~?b;bX5P4sB(8zED>#mG#&fO|hyA1fwH?mc3(B{%N
zve7%AHd*S498e;ff=(Cac7UeborNsCQo59v6pS0s%3H>)F<}Ud8$k=-@qD?GB-=P9
zrn2#f1Af6|wr#Ib5+u#4ey@8n+&obIzOa<1`#BP9xfe`Y1;y^Z%?7ca@mf<wV2Spr
zF;8nu2dBtqzI0P&rG=LQCIG<LG4N>6%z;<Cab%WXpy_Ho>ZYPUOV=56Q-N$No{kDO
z+nq`+Cv8Q4!^rQ6{*Li#<AtFh0;h;WfNR<4CGG9a24n}5(B*ng2`W87zl}cM2+m2m
z{4-A~&t_A)VA87~o0vTxyVV-$6*3!Vi>qek`}ENk&j;ue1wchAE|3Q$A&iv=YvSU4
z(`s5EbfLek#xCC<0aK)QAgqCR&Wk*S{n6@VG|6?EX9G0^vXWn7ZnO|06-tzt?!Hd)
zr5q~o>yNu!hVNtU=0Sy$Mx>P2-B`HD-~(^i=uMa<*M;_2l`{pCl>E{iElob?eK>{J
zz{j#FB7W&xp=XsSSIQ#_;n5=KbmM0hnlla3C^4K{W_B3cOR@ZmAQ3v!P5Za39Y)Pr
z%wnK2$CzcC>3G}lRxjFa*xWDu0w?Q6j><ZvgKrE`X*|1Ej9EcZ5*Za)X5r6aSo20f
z?&?kBP5EKRwe*2(lge>hN*y>w_Upzr#@8rCRJ7X>Ybs-Ibc>bBob<_ml!A1lolrth
z)z`SlHcHaf@sU$D-_qN=t>v)Y$$0H7F$mTUt&N-6ONU$fn}<H@YB}$FtvdFMDb=B~
z2ixZ6cACiVR8~Rrfu_Z7t@oGx7=Oyhl83gpwFJL=U3U!E|3{y!ucV)t_D>($vM@1m
zdP=L7k6ydN&zdm_06F4WnbmE~F6U<%*&VAk6td?42w=Sb3>8l5%i3Ut((0Vq*@!zX
z_zNr=CIwUIPjnOnf{JYhJ`Z^D(0XfA<Wv|Ct7mTb47-GU=F(|YV0{lv{x*yYHZ?X7
zET%mna!ZmQ7uzJWAA}u_T3-7I2_i~*&d=6>gF4ZOP0o2ow|YUDDol7TIT%H$FuB?J
z+em7u<v9865sdVLZ~_$utr9Hgdq&@l2=P++fRlw+3Kc=<qeX9NsIbzX1pfoqTnA0j
zA7#_1bEcQltY4TvKF;t!BPb6y3-dhX#^P*tBs}{rP@2IA%Y6QrQm!>)y1IBT7aSrR
z0g=ZHaLo3NjZJs^rN%Rdktn|F8QOTFx7Vw&L^j(P;gcpC3FcEdLw$v@2oLVKZGc9E
z2$(dhu>ZZW9=Z70X7CV?dt1`!Et^sdox@wm@;JN$G9!q7Ikyi%R2)B@Vbf4KnSQQ4
zEo^p$f)E+~1tVnQ?1bLk;#VAp@t|rHg`8G?X4;S3u}x-a4GX3zG+^SkZgPo>Q`AvP
zkPC1jR3KN=IV&hB`AwCnZT9#CTLPh`4AJhHioVLm{y4dlXYr&;8c&x+vn0EOj*@#n
zcUkS$#)b#bm}We3gsP^Ix@=v$ZKz{&i2ybY%0($q&;T8^FQMiyBQa6CW?Ft#x!$(F
z71o&%z7Ur33|Nf<@)}l}$W0xQ5Hd{{9+>YZ5TN3ED@sd|&uy(k7MJtvRLh%TAwlEm
zEBYY2@<4{v1&V&`oyFpOLQ9EqF#?cW6_r`TDL$!ma$`8c-#02cslSE(Xe`jMM~~s@
zs^GTf0w#HA*s2DCFTpX2wfVpl0-u&9oo-_K$kw*|3wg&lBwDl~F86iHM;mDf4pgo5
z6un06`rzJ&dy|^RH3Vvh_aSKRN>KP$9yr}cQok57erqPJ0;^5CcPI;7^i?AVB?Gt8
zHXJjq3yNwVKFLLGuN0}8T&I;UgX`N5`)txB3yHcWV;wOEo4SsgWB)Nu?*fo&1Ad)2
zQUV{-hMy?4!}wf@x&>IBXb2n9grudOssCy$X-<Bzpmv46w*hCeA#GOWY@{bOKg`;N
zhKpR!dOEg-ea*o4o2n1<OU_@%u^8S{*NsbOrRdMOuuswmD5M{_%`(anIjMVi;=7cl
zsk4{hpOdGKQ03^qsKxsja9v+b`<HLfUS{h0mv1oES8D?><~&T+(;4%SyKbeVwQ`(}
zhn)58eQ8hQXm7b@d0d`2JAaEbsg;lD%RYZXGY`K!??r#F&`I^gqF<lS3Qt~a0r#+9
zT2;FXDEV!&J(_K7pE9H3Y&k^#5OIF)W_da*>hoJA$+gW5<3tVJH7(<soqvp6qp=CS
zO_oB?_T2XY71*9kikl>Fpgp@qb>97};p<A%390PFC<CyLdQk2;6qW4TO5qv0yNY|v
z+-HKb_g@LldPE>c_FwJ&bZ(eBrF-VwTs_ctqqht_<6@py>W{jXA9b#yo-4-B?S6>i
zw3C(ev)6AU7gOh&{Q)FMBb_*);6f{Ce}K<sT!fCe2_~dIGia0t!5tSO9uO37Bemec
zE4y>AZ-x*^xV`@0BilPxP}eh$_Ubt`Qv2d8SDzv!xUqA)`UD8!SaLGe#s`8_*v~#r
zaC1!hR?vvy2<#*JXJH%j4I)r=lnh?0sXcQ!OM^=t{~TTaN2RDb52z#1m$MCYtvM<j
zG*LT_Rw`QCPrH2N@Vn8y_(OkSDW?Y1PJXF@gD-#*x~FP*QRV&m+MZJlQMZZ%ktH6S
zJz_KGKjW#tYiI5tK)-shGOr~;Fnudg9{mA@jgLL~AdI4?&Uez{K37zu-xn;my|MKf
zKmFIN{LqB;oNh&y>`z^qzd|6efZ^iFCdZ12D?~pLF^Pw7{5`;bKBjdc=S>sF^2;`1
z#ZOeAb^;Z=_HhB*$jpd?V&Z+drNYN;{Qot|zeJ$!0oJ?`1p{y4@L3q^=Z2q|2Gz*v
zl-N^qx-H(4{ln%y69IvM0~1^{B&~d@_p1NT)TaBFWtGfFpC$`6&HK%^*gz5ip*yN@
z^CB#}2+y8=$;|d;d!1}unVqNSfj!RJU1Xf0Z#WpRY<k^yn@HaY=M^rsHIvTIy}LLc
z+gvYJJRdjPM58VPIu0{AnYYH@rdAf-F4c;Zw01SM?R#tH+)DU3mkD@Zo69S0!z2_+
z)xFY5vv7s16vCSOO7BioE>3Wi;2O1JfIbJ<j{%<94%`p;(?s0Zpus;0^&m0-<1tMN
z#!~A?MM+ER+7Sq=xT&Ca`VxbCVQan8z7mDcELQ$Z3?{Y49-(NeOEhzDI@Pwk^~fl=
zA9iP-_!~Mzr9?v=Rj6G-5&{aDEDC$a=5-8bBjw|aiQP+qe`}DhmB0_$oepC%<AN@m
zf9}3YL0bFq(GVLBM7SG~0Sxx<hhIJ%h}OoBvd@zJd4e$ywP(Y6b>E#wrOMfKk>p=Y
zjChR;i!a>Gy{xRPZG%N(Y75*q$$i*~(kAlw=OfectblbqgpOtF<Xp-8kXG|uqU01y
z*qHYAyHK0u6iYAd7T@bPtc#+ELr=6m@O$SjkG-U-lKQ6q5=FOS)X&yF#$mQ?^5#B5
z6@zYbvl2(r2}S$>CPu`g!4j2tN;)g?cf0y|>pBf8PgwV4VK$BmUWUTDS!7oUNJT@#
z1j|#}wz^T@{rwKJoD?T^tX@zi0|j5#P0G1`Dzx*t4wo@H*I#7e$_8iTkp$aA?|Z#y
zLYqV|syYEtD%i|czcM)O6Fr?*xYWEEh=OXV#LpZaKZL6a8&!~1P@QplcnlJ1qyJlU
zxR52>0Iyx;Q9wVd+@+cXGCMDViV4-6w?C0`%sicRYd;zkh=o&S-EW(CfdaFxU!h)s
zQjS^FSgS3UKikezVw^m-JFPZUho(vh-#6}{h`oVbzAtc(izgE!gzdUCBx>2b;-OmG
zyFW<JFkUwu<sD2X)6W)d*SzQBeyJr_IhV6v-jSHOI(xA3X9eK>_w;tZfizw)<V>dv
zx5y{Q)_|c&+60)4m_?0gRbq<Npa$Ld-JTmj34wk#@j4?SfB^#$!VZKjWEyzR^XCtS
zJ3HLibGhKAFgn_nOUMm|r{Nh}vUN<!;amfRha1DIbgTiYQkC+*{7}r!WEvq^?^*TW
zzjTyK-#M5{{e8GpEnkQ;UmXLa!Yh-0#?A8_{z)~Ic`~5X<UQqsu}Q89F_t7{?VVrE
zuDDYKes_Kzwh(;o8K~FvD0im3G8ToozIBg;7im5=>CLoBv9a~DY%{(lk7A~3Lgj%f
zp#nXQQ{i=IxmrF5hsw|px2+tZRoma)QJSiQ^PKAI(qL;ancC9y(O?Yb8TkF>?gK%Y
zV`Md{S^kST1d?&)c-pn|olHa4|B=`v(HB3zt5{T!?Q|J?4_T~iOUR5up`VB+P9&0I
z3YwZ!$NS-=crGH)W{FN=eKND6SZp=6z}Q@wJ~zNltj$N4tK00XT4;Bd&j+|<MoMP7
zZJoz0-Q(z*<pu;j?*s}oH2Y<R%b%M$i1n^@uSy&;Nxbiz!-re0xmUw%kE65B7a_iG
zgDFw?T+`lfW%cWOGksbKu?cmv2FKL{@5hCt7d;rskSk_!1>cz07Y?&C=l>L^Y`O%z
zJlLB~W@nNxQnOh(ZoS+aN&<Ob+7I58*9(g$?)D%y@`FE%V)n!z`A#eKeC>BZA>w9-
zuAVH%<_!E_9;)mv0Sr@F%%<VwPR(FKgH4wkmp;!|V>ZXDl|Jx?nJfz%PBUI39~cw2
z$Ly*vDF^=U<ZjfO<u$#m2G%c&)%eb=AS1n_F8dmefo`nLT!g;Oa6*9}ubqeMc`6z0
zHn0EB+Wcj7T2cI_v*45xhy-TbPkaE}t{fGjrLU!mQJCB#x{0s^d~b__ePV8)futp3
zvlET_M3Rulbt8yC+a-847PmZf@|Y!U26p%CDX8Q?X?gycELxbGfx4enP&Bl7jr`J~
zV9CgP1EiL-G*XetZdhQa{mRwVcET@eE!V^$kk1t8r-M3=Q91m3i`HwTs_{BDtIONj
zBC*{d<wK)U-0OO!)i1s;-M1GXhv=Ru+%$A5Ob5PC%V;*4>N3d{D`#PIqi}0@i}@(z
z(daiyhoQ>;X4WezS>JgFY|d%~t6mf;+b&nw2mI%k04c%HSUl=$zem7#dA%1&9^=L)
z?JNs;1J5L}JcA0ZGnkDdf9ve%06I6PFx2>Wl2Q@8OS!*Koh3hN|DH_e9Y`qgeO?K;
zc$DuTyof5a*}S_p*hoJb(xO$xUw>P)-Z-B`ulqrCt&$%MdzJi)pXavdlzx${LhJVz
zBmm*9uv|LVqW8;*<3<zj!UchCmwb)(@tAxz+aw@UC*D-w;D0$IJxC{VP;1YnHbP?Y
zSHvtfjYEvFR`nRG<2ld~Np;UFgM6CAn#kJ@?!3lxa@*&M>SPl-b8~AXM`aHKzu;i-
z+Oh^$dKm8Gad$NuVHmlr3rYWnHfgP{Vc23&`xfmA)UW-}%#J$}gu($E6(jk${ridL
zhP*xQXA8!f8*M)qAQ=@`$ZTv7!O@hIg>}A%sJl*g$!J|Qoo`+a!V~P#aT_KRQ+^ep
zv4XJgi4Q!*?LQnNXc;fh<Z?-kWS^B&f(V6$qO6$`7}asJY{a3sa75^6?pNy_1(O^l
znJeTK`R1;%6Ktm}jPGq+S}@Po)7GW!yUp`RVBp;gXS)k;?<(C+QD`vna{WH#UXEff
zru*ElvT497Iyj{MJ%wDtgl`anu+L(F<Px&qv_U-p-~HLTm5K3x{OI318xLgS=3v)1
z@XE}E<6OcpIDJ5rj!qv}!A-utuh1CT4D}n0TOUCMXl1;8ZZQ~>2(ApFe{T3UkDHK$
z)SBr(J(oESo(RJiwM@ks)t{Wp+wT~j(h~$EwL*kUtZ{i6n0PzRp_P;vv%Q@DFgo;6
z-=GKUoLw8q1?CBTB!Vyr(01%%DGh@;*|LM|&f{e4wo>5CgoV&1Q~`z2%_HVQkUQ=_
zQXfCk&AgOF*$tz|COaTVQS?MUzQ<e05|fol7{nst5zyJ`q5yj$2CNa(S}s2>B!L!l
zlg(0XZ2!I1Zy@U{y-x%?Kl4t_w8y;v5QBvo#sxm&BE!d8U?hbE5OU+v*>55@m~L%e
z>72DsoR?QcCpk~;SMKOEs;<{Vqpfeu<5lE0(fsameN?{i^*M7OnC7=$hkOKXeN@J_
z{w#GW5V#QhWmhTe5f-shnbWzNKwrgb%)TF+=cZ&$BS58J2O?1f?MO(xAtA~2W#*}0
z#k38Fo@;@ZEIAWG$8D^w$b3UaoaP6W^L^<DML9~He=G&aBZt1xhbI3?&>iw?VE+)t
zx75a#L-gvK<B?joZMvtKu_-5ZsW+r#`t_sM_;{@$eewSVm_USo-9n0-u83W$W5*Ul
z0Mfa&4!I}0NEro;!cJkIN4ipeOmxOmM#GvbS}E6(h;eKHLZN>PN*c$e|7kxR*~1qd
z!Etb?p{r{RG=9z$p6`4};BOt9N~dKMQ|`dQWt8aG)Q@{^>8EujDco@=Mz*7-y%9gC
z#`-gBvuE#jy0t{2dhB5TY*<ZZ5Il@3UKi*WFL>GI^T!Eavm*Sz0j4sf2zc=5&Rc0z
zH5QrBQEBcy%QL6_EJY5|`8^s=ymGh4^|9ro;}(YHG;UZ1q30S_eCW@9P=rIQDgTAg
zD5=_6L}0xxGM?4?Ox!-rink8Ay7}^-qYHA~6!k*U_;EddV1odgM7muu7($SF)vgp)
zh7^J1%P;>%s0~OPPq-+>R_l(O%QzZGa>@8$&m8SNTIFI1Aut-1KGKX$b7u0^oR4dF
z>7*?BRMn~fK&2qU-)+HI4nn~7BN6tu`2;6qk2mG6U-SVR<ii<Ay_mnO4&7teRFQ5R
zPtPCQ0G-cc_>CX!kC_SWg+gsZ;X<&}yYXNASSGuWnQATO;C)}oR<xYO40cs22`=1G
zF7Er@3*-H92p<mSu~6PxeD12tdY%sR|KS2~a+Jp`PT|X@a*x{=p=HAzX%}lp4nxGo
z&e({EN8KhDqCrH)DjATv%~8g0ouk!h<UEdNGrv=?@$?g6WDAV6JBte}IKDnc@UmYO
zAJv#@i3xpOqm5_sDlv*C=KC~wr|Z6?v%$jI+?M)Mr_*%*VDWz-IVNcTLrLK4#}mi*
zFof`E<VR)nO!A#T78i-#fFB6g0&fLgrFU}dcylB22jeEsE-G}_GzF=ugjvi`Xm_R$
zdyF>eB=kQ;YJLt$U9%Ik_iyR8Ij4}rxk1zEc5Re@PY&sDt?54U_Z_RED91*tLR?}Q
zhNmowGu`wWGL7BE1vZuk+RYG{xBa=8qLVf;CH65ZQZM*{Deg;XXpcqZ=vN>pnTq3y
zt?jru@`0PjD)~Spw5i_GLr&cdD(Yc{7EkzNWK#fg1SK2LFq)I&vm>G#Bda_Ek=7S<
z`4|rg+6(HQ&Zi$JA0zuU6gi`J@Ebe31eVbMN7r9RHQmPl<G3OqAtA^B=|(`1mhKqc
z9a19FIU11`q(cM-NDGdX8r|K}-937+5#QbK^ZuOQ?fpK#KX=Zy*Unzo>v}$)kGh(O
z&n;=_d-{cdl<|W4P9Cc3A8r3zp(vHHUJR%67CE*Rvv#{FJSksp>?cugt}smzX+f0K
z@K^>Md!K|kRU;%7A!*lSeV=g*ho+>Z4J)mAtX%KQ;z>;Yh`aRL_2*-VD=&(vi)aD4
z2tjZ6OWxVt+{kcUH#_%fL~LPMhDv^4Hfx?+^4ejze`y(vVb>DRU};H9@_GXN)hLMh
zoQYzg2qNB_<#$$7Bd$n!q6>VasXepXa$ZMs(RiJ6>wLdaCmqhV$g<)Wo59hAMigPT
zE%W^lN^Q={SU4VV=RN}+uL}P7;c*Asx&8{`_lhlcqDPdqBlLD?PCC}t2bU;d^$K39
zV1AlP3z1+v)i@jv4J4Ux4!SEMKm??<+)0S<n%>td+@!Ufi<B52in|XtIDKF3{RYgD
z402$McItyf{dXtG!m5Gz4ya~}m@h7WPt3><&nOV+TP|#7%i}sKAvXF-reM%>vtI|e
zSrcpZ&Xw?@z}|N{+fZPkSr$|0a-+;^BAnHqU>f)Y5dK_nJc+p;T(R04Rwm6s9iJ*w
z)kd(?OopqM-(neNQ9Uk0WUrk~%ULrjpf@Z}O7v!*WiTvRX8^N=QT7Aiv%)j=BtrZw
z^(HFE=Z-X{a&!BUPv}kIkAd}LjIx+NiYE3>`<vJqyS;R6!qn3OCAFL*Hni`&`2Jgc
zo&<xLrCyyYu_!>TR;1t|(CkBjL3MvJ!sIC)@j$_I&uv7*rKI;&lpIUO8vw@}g8&{N
zq_PJXz5l6rYD(~$<~45I+<chlTb5unbh#JZ&Nl~Yaq8Qusc!>S{`)}YJr91*;)lLP
zel*EMG^M~nYK|B&RbaprVE#{OXe0C{rHki!jF!mBBwyhj4;nYN(dPfpPmi@kH^Mog
zIUvFo{=!|6Oa3oJ;D7LeHYiqbqF4IkD)nVW4KIho*#F$icUY7%SUQ5+X$`ZAW84=S
zj>Z2Or$5ofQw2ZXYOY#&c6{B1WsNteps4$Q9wrnLA*s%g<A**@;D^pHSj0o(ZZ=_(
zN{QwvliQ-?-QqZD1moKQ5+Zvy6a58rh6@;)2#yAp>}Toc+osYa=dFio^bzvk?F!q$
z!r&>U<7!p43c(MJ9kNL2-f=ZxI9g9U0fhYf`_}{c+Sm>iR2r7mdRm`JZ<<3r0#`Rq
z6A9)k?S5Kgs2?YF1#q>{^V_W&_xY^P)BKKE(9ENsvx*Ytle6_OJ4B9cZr)dxFdr#H
zd=l;K+H}S6yuICYID~$be-+;Zj8f?Gof@F~2IMYJx&3*4nS>`qz%<SrDg{da;f%Q0
zATed)5!Wp-4gGm0G-F%ro{wh}2x(t*-@{qZ5Y+bT$-1QivRdD+&33vn{`6R&qqKqq
z)^KLILlR4~rVn<>_6yC5&0*b0?1KtT9>4IC-DYo!=FtS<FP;jQ9RfGtgKxZlX8e4I
zbXd%$HJ1SC3fQfb*twh&OYTa=u#@-_n2L=nlW?p5e$T?y#qSmdv8B<<V)Ja{+DiDv
zxz63Z1k7LDYJUJ0e!r!B*JB6U>vf@A^YQ)FQC_w8<$<cNP-qevqc(6;$I6+$-+SP(
zWwg_&DVNPKGY8h4<us=Hqu<AUv|`<!Cq|zjfN;hETu@zVaw63PrWA||VzS@R8S*c_
z<ufG9+-#)Ld40_{MFl4Pp(qF}Pz4&MuaNyNs1a?D`Y6zRil<{I;daP~Z#sy-!%x{m
zV1^$c)xogCA^$suZcoU5H<s~MZnt<@Ge~{Ae}y<u$e4K0?LnF1S*E#<I`!Luum=js
zj7N;~H9i31J6HJ3E^#7pmNxaBEUVuyE#@*=guLy=Q)zI-u2TY@@tsR?D&c_4xa&vP
zgtFJoEnW2<Wt7I&T#PFzQ@P@f>0G@r>%B7FF(3veRwXa+&z%g$ftw_%uf*%$w0bPx
zHmUs<7arAOP)A8D!EBY)CRkl7YbKYMfy}p;%av}=u*WKSwyo?kBR1d(7-jvH!H-~_
zYu+q4attf2Pv&~Sw>;d_uC<8DkG;b-is*j6$7NP36VDX`Qqs+qw@OVk;1``EAhb2y
zZ_XKcRp6(hyg%3K0EEx(e<0*2dtG>b$Y95qpJ>0ToL8Ps3KEUpq&Vq2Bj!szixlOK
z-&J_X`a|^5J=R4%=m$~5)%#bms1#Js#Qc8R3bU=>kN(8ow%H}ACGOD)v2fP3P~Y9<
zU~b)OseQKW17&{7!xz#6ylhX-3tq5QL$69a&U_sQ!9g6-or}E|Tai&8U7h_y(<{D$
zaabv0Md72)Fh=#zLM3=m4#r%#fRmw#$$$RrX<3I44l2C~fwX?|9hwH-#s%F98$+!i
zqL1rs5tTq@QU@)tNiXiJv50Jcz57st7qa<1N%Rd4B)HGIipWT5N1Zg7cjy=r&LTg`
zzD$x;nv12R69#WG4t#*rk(@`WFhPhMZ@;~r8OPQ<JF0o!Tf9)r=sM#4LodogTlLtk
zRdY)=DMz)K9hjZ(j!<BILb2TH09zfe9~-l!9_|n)5J5wrQC-C@fXL3MBmJ3AgPf~l
zgdDO<9Gf`Y2~Lf?mPpXOnQsW^9fGRQ_Ml@s;J6X#eQO`8BsRAtb-kLT&(9UzjYr3Y
zXZ$8oKyMUgdT$+<*J7suSy?%EcusMdao8Wt7hCdu4+^is^#0pJkfAAHp#5%V)jM1w
zF`>#Rn%o(IRR)tLQ$f7<!^8v#wH^L57NhpbM_BPGjk%mZJgq<5CuZ{i!zB~Z1N{r@
z8?o3&fXRfU!#Fz!%ubxXr>;TUKn>k&HrLlgb>L~h%<L@V&dcQI`ZhEFXN!J^bs>t@
z%;twX7o}yIM50#RbyMan{i!}Ko=pUd%{fFyS~wwoQx+YNkuPFWZnl*OshIt(SKEj}
z-+=(gd-Hbj|5T}ey2rPAgk#<(u{{SH`Su^}7p3ULv1E(&eLimF#=cw-`#;6%4Lbf)
zw6Bg;mr743=qZY5`}6gvrc{kSfA2n<<nU{Z=P#8qS^sY}{GaO>5%#6{Xb8CW`xV=f
zyAx1~uKtA6eAvO!H~PM*bo62ck#4_uB6T;PVjTQ`su$`GFrb0aFWF#BR!Mg{v))V%
z#7caN>+3`_<v*RLE7f1}cP(9|2{v-w&DqF@SN*USj73Kq?<rUKu~R%cU}kgt<A!%p
zu4u$1GvnWII;t*GVR;j)>t>5XKsE)XrRAVYJUpJNDKD!O5|TjeK9kNYF`pV18RcmQ
zSCHt%kA+IOB!t;#DVxu?I77h68`>Dx+NiP!pSP`qZq}qjLh;0&7&cfyjq5x?8gAEW
z>RH_Fc;p&?ATn6CjsZ!lH7ky@wY~g&RLxPWm~hy}_Gsp;6aE~|wA0+y@?1T`!Vxul
zdAxzZ3ALERCsf?);E{yRfu7_1UMggPZF=|$rWA+An0sBtYYROXk?1zZJDllVLLt-O
z<B1E3e{}^eCJY)3bL)t-_YZD28qxV<fYt`_uItFD-vUsT-${W^-QC>~b=Yz&hbtrS
zgKB<1=Rq3@CvnG;YPi8{EZNjb7HEJ^|7Co}5fl!UE1eK@p^|W`K=y6fGDlzleXoW=
zqWy{hfXhT3C``m^nz7FJd_*mY6|lh)L$S6EelwEPKRvBynC$FMvQ^-8F%}Gw=0Jb?
zaepzdoWql#^}<WuQok*JsW|t^Vi_=RsblM4Y556%7~ilFXevJVSSAg8jQ@wf{!*Gb
ziYkNKHTggn&@4U}YR$)ASFb69S)cm^*%)DF;x-E;l)Xtk;EKOJWjaPpNv5shDaZak
zw(;F(V1HztdLNhUPUM(urI#9&$YS%!^hbRLvaYSNxuh+4!Ep2K)qIUH3JFGtV~;E}
zl>4>3P&{b+Yqmo5W*T(VADOt(be3jUe}?uJ*Wf(f*JnmzGiljt2Ub*SI-`U=Bnylw
z14KB=%1Wj*mCvX|Jv7)zg!(jXW%~)ETNeUv6pLQVa8%ro3+1;`NLW@Jm_Bz5MHh-n
z%9(tx^5Q_rs%qq?s>tFfwSO*eC)`mjO`KL2Bz9_41RUdah9t|~2*ZDIEp8%ouH3##
z0|H$}wF5J)*Pwgr+Do5(CYQdF4udwr3A`kl;OT0{uUDb^VE%t54V^lhofansse*#;
zW!<xtpu2Addi1!X4@o!As0bT?_t*D3i^uyike?tf_~oc+w!Gf?<JNjIeU(ZHU(k@h
zQU1*1nbB65^i1o|I^>M`ZU1ZY)?wli3FzUo2pgB_P3~n~_k-j8i*O&4H7!tQ*!kk&
zE33r?`??wisrZ&SQ=i(PR&&EfPDlJvF0&)OvK3CRKLT~OK|9!Xp>ZF8<R8w-8g%~+
zhpk>Ky!#Y-xoalvfBNh(PMS(ecLAiVX}CKeSrpsSJjZSB6IR_@KtQI<r$(9o(hc&i
zq*jl`Ej9@a5A_<gyQ;J}R}eC6I;aP170_JWpT%!m#Wk>P87>!PjbC4)i-PwT;D>23
z-h0O%l6s%H>u38_&1k=xu9R9IL9Z6oFy4t<Y}<e1bxj-5jfFef67t(=5VBtr_U5u8
z^W7_~u!F{Md3bb$nVVG?y+_CO^X2^-9l-o(<R#mS$dzvvPnx4i-RG+xG(;$UoNP1S
z_8;Upu85Na4jFvFpPpv<)9q>TYjw_p(T0g>8K=fky_a!a!!cxm?VRAXmjQi5G?TO9
z16v%a)f@-<FY!J}EC^d?sLnHBeCccGRRa;-8jP5T$6)?pZBKA<;OoNgk7sXn!6e7;
zhdli{@Q9@^Z`XC5+Dj&vDA)z9GZ_qGP~6By;~*;R<i<hMj3cDgKyhjanCEZ)WV5-{
z`NP>9v3;n6e!u%epN%cTn`V1RM7kBZm4y~GUwRq!jTTFI)b8EN2P>}Rvt^Ce>|HQz
zs!CZ}hSwp<!MpwRu<JKw6|%6xEWf*1XM;DRy1IJx<`Rf63#^XMw*?X#u%)^_Ki$5?
zPyhGE>(WHu_N7XGaqXv)7QV~nI)0i*04_6_)jkPBE01G9+4@vx&siq;8Z$%AD2n4V
zlnoV!nlnbigYN3>cq0rty-lw=;Q+CyQnAk*+>P~MlLyZ(UUqdbVh`|b6Wd53U2nFj
z(YmB5K_Y<6R-Z#%_(URqOaH_5Y5~4-Y7fA|wBdHhP&yUA{^2gquRtwB<ftp^IM0&F
zk}gJs!YQP_{#V*H_#Y%wU5DqAhRaGV4(LD0M<?|zFc*DXf(3x?V%zS&WmvuB7Z!6#
zw1n(x{(7&}<yPhW=RLerLUd#25jk>xcm{zXhU>h!eJ}#YJ)6;Y@K}K&fvvPbum+X}
z_E9jyX64k8n85}jiWA{|p7r(k$}fBPWm=8HogRGHT=ekc!F&c$fNlHJR6$gB-SmSA
z2On`)R*W2yj!yCk@=J<lY-=J0bY0aQyu$(!9vsdPZ~rqT=zZ4NM2p`=s!GkP1BQQ#
zVZ_pk(RiSz6N8IhhG!#DoY~@zGgN!|ppefJr?OE6%)&@Yk5}3C+xb^Lf;&(Vi*eoD
zsYR_%|DyX5qV;-haI<<?tkNZng@gt`Dz!-d+=Rjz#T(z=NZjv=!_)0Y&b#<e-*(BJ
zIMNWKlOtBL{mTh&aLnO3O@6K3*EaC}(vhoCJ+r5DuP=|Ukl@SEW&K2gNC9Ene!qTX
z53;Q~Joz>8;0%h(TvTz@Tx7J~LDA@&I}r!en_#k6h=w~9DU`^cp<D15_xARf5DS|y
zsgzJz!`&Wv6r>bRKoe*$w{lPYNXx<sqr**eg)o2^@9_`+6$lmN!%3Hyq$#L<(aVO4
z<Yf4wpeTEHMBMIY+@HP|RtY6yuA~*vB6IV-jgaC_dP05PQ0GHs)G8WSI9S9uS39vW
zz~~4Qm`;AUJS}B3%kSu2@juXyYc#6t$z+y#I9DTsEgfHBjnCFw#g!6&4)w(Jx&6I$
zOd8}n*Ieb{ct(L}tv{uQ;mf080h*&j6*K+L_|zfSg%A8u%9`VHg1C+?dqJC91$F?A
zGAB~STO3G)?&r~s^4I5jcz|{D>$dLYiIZB!%<Fk|8W)k!$Sh6(&APKelNGteT9Z;F
zn)N3)bll12Q4XRQL7b8gF9u}`kjmt|`7qVW)Vu6#vE8Vb*aQY`G6IV9@xYfEMoqub
zZgrR4d!m+c+AkT1MrHj%cS?3}8F^nqd<Y*~y3u~;n2Iv~E@YAK7dZZt@wD;yv$j8g
zzl`3z`7W@bZazx_^!<QqBk=G~!Ld3K7e--vqJ`w<NVeZ+7LP7Wi`R=x$3A}D7<dO7
zCbTuM)=j>4Xx~4`*s*?G(>ik;)%N?oZ||cIC{yd1)6c))Erj+xJJVzJ@K;|XdtkJF
zKiBvge@6J^I}4y+>xOEMQ?JT73cAYc+_X^(tirV%Z~G34+)*-_oA_<qZL+WqHob`L
ziu?+CvME-=$^{4HAi>6mzaq864##7KL|HDU5<~52ICJXLq`S;X&w>*C$;XT>NA6FC
zD2_{J^y_E3*FDc_Dy5l6y)tTHo(~r*aHLyLFDd5k7L~wX)#-ymjw8%h(0cQOGuekE
zuE}T{F>RQP_h$6FC_MV@qA9;}IWT|nsf}ldL$Q&O&f$Wb&G=oj2)*+U_2n}|iA&iw
z^#qPhC93q*iu<W4`BU82RgI7o=YmQlkGf);-Ie=&X6N>P{RfeCqM}ohK#R0%W@`QV
zcyH^8rxu-#QDIv&+2)6&eH=!vf|gSLDgGg`s}qF*F3sv?%m;NxS1b4SivuBCO-!fM
z3tfLc=79u%-~pammvLlPDpkB4Rk+@61xB(+i<Hr0rT&x{Soi+TQA-M$UG0VXDt{?f
zc`4R4W2Iss=Axk8-W|tXT1v6G`IwwH`gBIW4Bj~Huw3wGsY_EPX=lmvtTBdMnQGMG
z>jkb|$UFM@HJ6|gAtLt>hvaU*-}NL_vn$N~T=Bf`Rj(lNxwu)a3D@tA;~$|9xu)C>
zus}S5SnL*xa4WGf9LkC8$<)Kkf(N??y6eHj(%~m&;8)fmAoNmdMnep^=$oyjPh}R|
zTBZ6S@yz?~G-wz%Mhu2$>u3p~jA#CYW(h^1v%L2!Q6$85{MKEofy+FbT$?hdkCo@G
zPU%iDj@oU*hQh?@OyA|eXwJP|`#;i29NdbOBMv|OYhY`u41Giy=Bchg6lhU?shAlL
zY!X?_d7N;WQ_^6lsBY?3Jw>!vqFU8;Br)qcq{!TTlUliZ$N-4*55ssU9i0B+H}6ns
zl!7R;=g5o-YSV?7!}|EyAOu5&z;Le3PaN;)yc$8cxTYvP@yi*>x@sSwaFqVbLS#;_
zF7|;sbEzEjoq=0NL?Nf+ZfKcQ{Y4-d&1Vu?a|>?&m1<|45~Jjal)guivC?3vmH;PW
z`~INi%TShY!cJm8@<28hi-t8;1(FrgKi|BUCB0CRylkO<^-y52@AX9&A0)xk87u*@
z;$@D#nA<-NJoAJR2^y7Ypjea7#dp#~KKQq8BxlsSWQ}enP=0#lA*+ZOyG+JMkBK?_
zx`GNek`U6;lYtu(MQB1YeaY9ORSDdX)bhyRFVYl=|2-JS+Mo|aYuSTU&RJ7ae3HCP
zb<w3HE!IOTzLsk>2eta=E$x#A?+u<7e^gc;Mb}yJMFxoDGf--X@8r{mqtl_<qq(9n
zGnSQ(q+AXPY&PPKPzuV=6ca&#r=9ni1c14|w!o54p5|liPo)k~;-odx^hk=}c)NiQ
ziwp4%(O30GSY-WA3H!*gi3tI%EMJghd+mlz#J4oeR3|BCx00l;YW0V?xX)@vMRi}8
zeDyJ<b^2(n$B;z1DN~7`5=;!h=a8u%2)0Tv*ycl0VspD$U|V$992-{dnz#&LuTu+k
zLMx@hY2^)~%}DIA<`&z~+I3-NLfW?t9xPk;Q@yX{Y$JA(>ZwywKcVlxO&1}1cSFfl
z>1^=YFIvuCV?6*vgrw>@*nQ#!Y#l3=271N;>Xa*LWa;WYySF9nUVQUA<?qUdAf!Do
zR@P1mCFW}+;*dI{nPK}*!pT8wA8YTWnWX}+b2d82@FU9iLz9bc>r&GA4LVP7UhF4Q
zAoR5j`>Gz*Z80ut&%@1p*Q~6m77(S&d91Q>f3$J*v-Eet(pQ7_7g01fiRk@Sz>d2;
z5h7TX>S9LA>CTKFyKI^7h_YgGBE|Z*Wb}ls<4OJOgI_P}m+sOSrQ5X5%CqCpPTbk>
z*(z-DvFQvXHw3N&qMR)f?oh)6>C}G9WK|Ure1@jJ%h@BEVj+tNRBd3H01<kyra=<J
zu$6~dE4Nh>Rh%DAPT2#RNP#aDL3eHE&fTSTP-fRhvnPRK*Xzc|Kvyom2d47?ge!;r
zd0`x!k<}k+?r;BfCHoab#p?U|mckh#Em1UBmiP-nD%7nKoG^uEABV%4V2<p6a<7c0
zNKJlFtV|U7I8i#&W1IQk7z!mWRx+j^T8iQ=fprAK*03A(@{Wm$Fhpg#aky%g-O)}N
zT7FK&?Ug8j7bE9eY8$&G71)vRe08`ZHV2lQ7;aqM73SG-&K!ko!^0=dYpKdsB^!M{
zJ-BmwIH#Qm@qjB2;jeTla01teL}*kssS;&(>+WCPj`9QWILNzv3Av<Iz8EQ-%}^UA
z(@GK2PB1<k_lJoy+gQ^_Nz*l+{Sq+&_v_F_V*#naK2|mns`iXli@uf9gvK(MypdSv
zOc42{ma13poM=+J9;3Q?S4^~oChf>4seFArE0nbLZ5D=)&C9e5B4aVyhaL0syO`IW
z^uPI0Gvjx6ubUf9$F1&=iu41r#IiGGTrAhG>Nr1U$n8N5o!rA)CkCig<?(;9sljj2
z()E_=I}`BmHSIN;vdlnEEIN)^y(s#$)Ee2imC<^bco_ld=ttG6DosFDSBM9rfzRwA
zkwuc0NwhQ_zSYF8j@zTpTe^HsvgQ5rh0^o3q0Nqp319mV!p?LI<mr2aWyWzz%zWrn
z-viU-AVv{A&(ynLcSC^`?(`P2`K5iM>F(>veC=(sZ!p;)sKCfCL@^bOAUowpQUmtr
z>%r_`rP*UQAVy$9{!<pa&LrG4WLcoWt=n^25Af?_b5RDKkud2Av|;?BjHX2IPtCc6
zmf;<?K=NbPS9S7fwRLK@Vvhz>yQmci*|6+eI_i^%GMNWA89Cl_<?u2+?of6ve?C*Y
zm5AZ`OP&5X5$$}S$>NGQ=<Y`Rw_s@63h{&^L&Bha%az(n*I%gGTh8YvrB^1WfUI;M
z>l+{0NOj~h!?1mZWz59jL4y<nSC<<L?m+^n!|x=qVN9-sY>cMwq)Inesa+e^GhAC<
zKTQ4Esb3Ml|85|QUw_Jg*<e2+B6pSMv-C^V`hwn%)RmdT!O*`7ySS_*;0&XT*j8!4
zY16@wuNrld-r*eh9pQ@lA#Np$K?Whm^P$HUF`bbC2@NV~DrWk9)Uu3PpB@D$Bw|%$
z<94ijHOGHCl^z-B{vFeOBfJOyS*?DCdybXF)*2M!6yB)I!NQi7#XSg3cKw?2Ec&Uz
zT1T({?1v<s@pyCqYQ4NL{*!RjN3Hz5Jyv*kshI{>`DeEi_HIPIrZeI?;W*Tl!I~?7
zEivGic@<GRGFspH%>9ow22J27xp)BofJX1)>?61t@ClU@?N23dP`{f3P0`4=Rc)JA
zR%~{cU<OCwz1BHfMZuh~N9(?DJD>Y5j<qU<b!@CBobI@xk*||<rV%6Ym+BaU4A|Ul
z4yQj=10`rU_;$(ZMlW#a+?oWN&XbeyO~Qsj=?tx;T~=ZJq=kKA&gm+IY>vE|SNhVV
zu+Qt~bM~XkL^3Dsjs53Ldl+-rK4}%PW}tf$P_GOZqAcX%>#9~sKeK;uz+bcL3x*tS
zDLhp?h~UDtqJsu&x@F)+piyQ#V8At_3;k5i-7-68s)MgM$!p$aSJDiL{`}zlXhn28
z(qfjSBo;Nxa-ha4xMGE`bL*^i9Nl7pV&2e?fN%~z!$YYU1IhER4e#|C)X;f8MU%Yx
z-j4=HoZ&fb#@`28X<qfwIicA<!v$(79ta*7;JRdG&@zVNDH*=2X0^#1!b=Rb2<<Ab
zU?HFguF@astb!A~ik>UNtO}0?NFfBx{aTQ#1w}V_O6{NhE_@*MZR4p1^p?Vbh&Eht
zVuMK1Q+aKsyrp+h7#30ipB*A^q9vL)i5bTv^qwisqJX<|PO~mz)kK|{c!@}yV|8mK
zt2H&m`>KuS44<J>e6G64hE>d0x#GAtlT7><8|9qv#tzreoos)YP%O=GS+AHtY?g%8
zj}^GWYX*mjeqK5o-cfv_-p^F}q9d-u8`R!1T78VY)uFVKqI1Q7vTqCk;U4TcJa0WW
zgPJ1o&>au*!GfYuH+`&ES*(6X?+mSYC*IsV3bo3+m3%{I853cR)<wI-)b-6{l5K%W
zwJtkkwa2d$tVK8%cGe<^xS}e)1TZ;AC7vdMGsN-~M~8cwKAx`moE*0>r_ZAbAnY8L
zdd019#k?*gRi=13vMh_`2dISqsq!?&$i5R-6@c-%>-y6wG%@I=kPx~59+%j0mEQkH
z8wc-&5y0?akC1dFHY&*4ynn;jDBSvxnD0aVVI8aa@%r{Q94E2)QzW+Z_m>AY**Se8
z%JAnnJI_<4B7cGu!qeeFyYlA0K8$1vIX;mufQE<RTr~L}Fzbt6N;j)e!>;#>ZiO~?
zVxOAtRt*|!xvq!GS?>a5E02fi>o(0aU6xunFAd*DOmiv{MJp@f&&?W?Sz$+VRqSP4
z(*eG7>=R5+-PHK)%06W*(*o=v38sBWc?G&TUID+>HV7%%X0qklDES9rM=Si~yL*;(
z$5d+C;9vmNzhbw1jtgX+4wMDfS6p7&nALCpndV2#KEJy%hb552Hpew_8fwTWHg$Ov
z#4$DwT!_|y#YvmDv4kGkDq`tT9f}yRUF6m(jm>+|fwG&oMHMh5%HDq8CP(YKe{03W
zF#ovy9@Ywq%T_Oz5bgaThbdWR{d6mR8h7uARaNM1Dz4gHsKGlr`Vr<NB-JJ6#jgwB
z5^aY1a6j8UN;<h1Vs@=sTYPz?iO$M%&JSf;v#D+zIu4(3EYN9sKAqfW(OW;2j@dd=
zez*~l2)vNdLSk`SW<Es+S+0-S{F~}&tNbK+@Uv*~qT#$7L2MHl%_9GyxYQG+cNiL(
z!cJnW*gnlDVJ2dC3A!|nzOBhNZyKXvp!vl~+up(~5ih3Ln0WLn@7R#+y>V0J?lYK|
zj^JTP7;C<5T1+#86b+|A`eMb{^VZ(a+V{>Zb3H2%%WyQ0p4T7aUt%ltv$nJt<nsUv
z{WuUzB}U7MTs=Pt-85STQwHMC;%i(3hJ9H44gl@9aPf0yIiQcR$z;oI!r-P4u(8u2
zNo`8rxxU{~3z$mmQkI1N?AC*E>Oj%CP{`|Ih&Ms*NP@Tp26X|0{B9c<`Mssr3#8jS
zK!ca0-ibhC^gDhYmcQi--gR_@fw`HE$ENgB28t~gS<tGzS~nS5F7%2jx|i@DUn_*O
zcwSd&B~GDLojN6l%ZlH=6*o9hQ)9gLw%WpUe&s8;Zsq>})BXj6{$Y$>Lnh>oT}-oF
z%@~WBPVwBEct1C(J`0A>OMZ7<|HYY_kKSdNVDgBRcR-Qt6ou9u&m`U>me~FQt&q_a
zfO>nIxPpGGaD-|w@NIjbZU|%e{;N6SS58uXb(S{g|GL@+Y4906cFC^f3E?0m<z$N{
z>VmQ@34+|{zBrFj$2u;7U{|LSolh@Gqcs{1cih{4C8J{;M&<b-FLWLhpvz39f>5q6
z-6=K4GW|m=m0Al|I>=CLl#=}tTd2xk-L*!2xX1Qi^P^-PUEBK-CPRkmNoapGdVydi
zyG#!9`8)S17g*6TY;AU0Q45===`Z;69hSBHCl0q8rF*EzcIo*3{}FV2q8odHMXA)K
zWUG}`&Q9rwTZNwRbENBE9oqj>elY|~>Z0lm<SI)Ld%%_}*B_I?IIB<mm8L2YC!w>f
z<;s|Mw<p;~&ZQj%uqmvD6eAz&eOxl@uuDGoGusJqEngk;k1ENbRw{p34`aCFm&!z#
zlGFy86wNK1jb@z9*aYi~;GQN%3cZR;M2{EjlL@q5Ck3kr#;dX^@=2Qi00nhuc<c4^
z9BJSZ;q6NBJE0U@utcuB=9td7VGygJSZqx~)9S7k>h-}DO^cbPrt8%746r5vvsh7z
z<kgAIE@5t0sT%m&(p<5jMc-(AG#C&C>Ezbb=hnXM5MsagAo<F!^<(RzY*aoj*@ieE
z+sExRaQ*1=yOjWTZ^AFNLbr#OH@T7?4hRv-tL##mU(t;&O%3xHOSWQop1qd`&E67g
zWY?w^yyhg$q|SNpjQ|fFE{7yoLo9jI9dSHC08vwPQ-q!8X5oFyM1(lY$Q`U=k+DXH
z$yeQHDX(?TNJ2O5dkJD*EgNbEf^&Ce34VVbY7@gdN%kVT!U+`js~X`!y*9hCBN})0
zp$MEeVwg>AWdXSjQ0Q>`Rl2qIyPOh<)tl&?@$HM6`M&#LVNOwY(SA|T1Crt07H{G4
zeZVN+14_TqFum1mIfOO(?J^DWay@txfqznVxsp=RQoFMv96-<sV%wLSA5cB^6LXV(
zxY&Ad8&IW+i;c3nz57unB3HRq+;KjBfI4@_A?;V}{@{)x><iVUZqUbJ%(HPasr1r9
z{4yGg?5(XWNN&qA=~tA<!>6>NLGx-nN&uh=z+^cM@cL|QDvswFv16*YD2GuJ)D0~z
zW*_b(eJpl*;O_yN3iqR08dZ20H)A=twG`B9y;|a9c=29~?XK`%lo^3@z|BdOk;<~L
z`<26L3IW6*zv%sr83fY;`3FT=j1}w-&$`#1B^PdzAIVUyp%R4ZvQLV5u$9t&gY>kw
zy_Q63In|F%b7;@q6|=pOaGq?vJ>dz+#CP`uy?DLfMV#}aQDFRsw~fNBL+7lr_8&*a
zoz3@!Aq652PF<c0m*ZiQwkxHZ3rV8DVfM6<lR)oSHpt2_%sf&7XB_>F9R|!P(^A6&
zX|tS~bG&cZY9VP3DRZ}2@_ty14mywDD;Ut3PB26CdrF9>=G_>zmH-0>$H(6hzjOUy
zaZ1Z#N8z_`4hfCrv!wGR9u7&$;cYGnQoLl<y{^R?XIu82V7^_1vlL7u)S=xdX})&c
zXdPTPrhMR06q}o^VfFrG5Kbr&G~Ow;KbD^VUHzh+Y%Ch1m*$6?TERDwRl<xGF6WK+
z(r*i51CwMb)w#+&FZ%Ep1F2E~xBa~G7v!WrX(yX-{zwH-Da;CJ8+x~jjBISc@Ut2`
zJSOu7XH~7NMjhOY0q!(dfvUIF(Lqm2;q8K)i?|Q>h?OKp*8U{#>dwHdqdU~FuacXj
ze`}X^Bw9W4;{KK*!ygjWN<7azk^Ykcrf|?5w`OqWNIa8pgfQ^t6u!E8px@2PLl%#6
z1>BJ|v&5Ld4ul>?B7&aOuP|l|rzO{#i-$YPyy$R)-GxkqhQe8K%9CIay+BgR_TF*|
zBKIx9<W{S!JGdBC6$)0p@QZ-|to_}8$$1QJ4s7H(>wJ%mL-Voa3I0Pu0bVQR8keHD
zuU6~FlKI`Oh?P!nni8Y1EYJ6WVV7gBM-Jz!DX^j0M7>Nx$>i?xR~sxD9|FYQ|4jQp
zqj&o8Ni*=g`e6h6)t8U#KCjrebql3rLurNY7515;5*7=ySb%&p_*;>}Oh#Z0rg=N6
zW8}Fk#l<zxE4Ar0%Bb+?S@zIfHbZ@`xNxgS)ii3BUoh~#!nq(tv>2wv{<4S^HIWc)
z(!x1M)#^LEVLb?yyF|cRJ}9<%<GGJ*<0I|bR+wvwXx@p?SzY#l{{l)M8O;?$1uH;8
zHa3J)!OCYV`Nts}kssE{L+wLP(szYMnLG+O*x6HBgZ6?RUy3Pn9eFD?+=D}5ZoS`Q
z%b{OzdmYR-J)C;NM6*UN<DE<Mf*)~DU+IVsKKohcG~m|z<v+%)$1<ym3>p;A-gPgV
zw*RU05e_)MrKa#+W{iCP0ws3s#kBmtpJ*d|6bv3wI$By|V`F3V(*S<*PGIL^Cq1_{
z04&yO{~O@@U)e}3J-QKP`_fGmi&OzqLI0}9_mtG+`W9VX2ne=(V^?YPZ?xrqCnicM
zbfq7Rhdy81pq*4dg_89ennVX?H8|Uy$6MHMs(<pyd0G6cnEtQdu}w2vl3v8bR}EmY
zM*qsK*?N-cFU0Zx{UB0wdQkm@&LKgi6w^9>(L||!FXQg?s0oS=-!J-e^lzB=|9!KC
zLQecQS{S}QXT~nSpdd1M9Y^6Fo|+1X!ehIXpkh+|#`iH2dc@`I@h{Csn??4?+OEsl
zw_GPlT{iiBN!4cK-*&xgyNGZ5Az#&ORbL(rpjkFU5%CKZ0qWB}6)T@_vo$;Jyo)=i
zC4M#fNc8~r!gj%a9rZQRJsscvqmPNDLC1;3X6wBB1sP%Z#7+r)EGWR=_;Xnk^T=eH
z_T>mzv{9jNEPq^Psm8T)K`cw2d?Q%q(g}Va;?!9yR_O?x*M+&YM-eMPmDMu#5c;~l
zri0Uto)7R?2~R`VX<<%%@oW~ieWHlX{M(tLs<iCoYUh@XK>r7K!=jf&R05b8JYL^2
zxGnN2X3OJN5~@&sBtP2?{5rP{FS?zu?jotrfK1`e1NPTZO%f=Z6Yf8}i8WQ+(%&Yn
zw=K_4I3gTeILzjhQ=Lx;{!BOa^f1p%Lsh+k(>&(3#w`VCSQ|5*=|<XooQi*ZyqHv^
zG;sXkIrROx+6+hZ+9|_R=-Gr;oum41<V^@yBMC+A4sVy*pN5A9xgS9pNECIjuY&|B
z=e@6_z29B7T}rQJff_?vb8_tBQ2chlS@TZ4omB+X^v}~JB_D7|D3<9PI6A5u^*hpo
z)3o{w@$=Vh{;K=^`_J7%76Fu?b>|sqZ;xerX@u=>qb+4CmTVz)x_Nqimz>s%=Q*`{
zm00{tSGY=>>a35RGRcU*sa)#r&Jxt1{T8lk5g2C|rwq4P$ASBvuNJoD6*(qmoLZ*;
zG34D1Kjm^Yzh4%EAl}vMJ3W3flDFUT<mRB~b(!j7Eo~}HeolYYos1`EAmD}}t)cy{
zl#&ttaJR1sDuzh)sV6`!+C|~Z?rLY#^6R4lutj~se8)3F94d7wDg&Fpp3h(_oZwIB
z3HVS&4vt}nDi-0y*z7J)L;pRW!#fEWbDjH&6rBtK==(GNNE@~nCdIW!b1;G9*@MCu
z@@+Pz;7;rf`FfN)R8`MP+#mt-^5Atp^%p8v(Tw~+0cEsT?Fu+^FIg`SUVdM!T|yS9
zZZdz<c`4Ga^#@8X=F^os?>j?gz%Gx!2P1_Wl<&5hXXBvMAVrM>UwA+}3|!FpM)^3g
zpT;(3ghJcA7E|w>ql{Y}>rhr`m4gOn7>=L{*5d#eptPT}o%D{a*_yL?ASrGJn(!wQ
zzsJ}tRmp#RVq*K_K81hv;{X(*{)Qwogu*D@0GkE3#Q>}B7;<&@fo3h?G?rSR&G>-K
z+5OE!g<0)6+}!_rZzQtS_ku+I)7u@P)*9=t(8Cx*r{Pr7BTJtOe0%^Jzk{TCd(+&}
zqTd+z@d(y2s1x(*(tVk3Gpn8sp<i;Sn+iCXnw?tT-)i0@FD|9>^trk`yFa-PE}<Vd
zUkg~D`%1yB2nJ>t*Ey!$g5^%V=i~lD^53BQ0)jW$8gv#1T%u6gGWDx(9#Ob}p8gCB
za78_1u(W=^M$>{zf|1epDg)GlX$!ay$G}HhCoRSMN;EoY%z?&#%o5};MI_{)rtLXD
z6m|S0Epr=l;0H<R<KFkK1-{!~Q~rlj?FSl#bamgR6E7-oNhmAo)l4ycc#sH0*83R~
z$z8y-_ncae{b}MlIw*g9;)@J(+?#hio+95^8p6(7yeU#~&$DjA+B}ee`E~@Tne>x|
z<f9R;*RKR@wcf9i&gfQsYFX=OI6h1&^zE3St<+4BcBoaB{vtvXIM>|l6+PoVo%!k?
z#TJ;7G_^Vidj32rF8vq+eFuq8`Owrge|x}1eHDGz+PFtAC}r~O`Skq?bmtX3IC@9<
zs`b7q2Lu+qUrQHUdNRs&<$3YsP#on7701N-<=1WL1><$zJ<9COJ#)(B^NcC4^LwD9
z`?8k`K9cr&IEI4!+-B)VWicj@34^1AhnP5p!|N}}6GS3UoT>d@vjppkUBp2bGhgKe
zOkf~blU5@5jlGnr;?hK23yv>IhuleV0<m`&t;#vwTLUk2;koe4<B?4MIV;)bA~W;7
zz|*2MO4$n<aj&p=ktj!^x$4%Qv~x8bT|Gr$28);Z!(0ePudq+p)!+4N>=k-qS@1I#
ze|J^s(BztKh&D*ruh9A*dbg0XXnNcDs<QLxP42q@!zy9=M=wm?IyX|F%;edVf`Rw(
zpcIA?bWrK2iM!LhAe=?*<(8Ar3I={!8ItiT0s~HClRHYE`j&`;b8a)NyL2g9SueVU
z2tHqR*^*sA1pVxPGBLTx0VI1Yb%TVhY>+*b{UE7J13z1^_qW7AyLh-^A@H6Ejb#He
zKYHAO;etAFEYLzsv4_&p#CTJjiJKQ4F@s%#+&_XhwBEy4>LbhpjZM&=)^jyISPE)Z
zUDAy<HE3D7gHKb*FFuNrcm}zQPC^VgSS^Gc$yV5vXKvg$B(hr&Ch(>q*bYJS>TBv`
zQ!ex`w#A<dVAV6=onI8jU8+YkHY%<D*G%T>2b;n??>Th(TAuJ*k^}1y7E*Hzo2-nI
z^$Pb#5nLqE1MlLtR<Q6z>%=DRDd{4+81^zn_OpfV2wV#oLl&n^kxMIqJDlwob;rL@
zo?)9)<)ABS2|<tB?MnTnyHU=w>rYf<^o&N=cazSHgDLMUnGn4+OYG^VR|voOUvcK1
zwSJB_zOH)<2P>+J$iJbWPjoKVgo=r@jShB>3C{VGAtMyll}p4EC;<|8aLw(xZTV^R
z&Oer;2l|!52#7sb)9?A$zP3rmm4O)#%-I+vJ&BTFta|;1^)jKr_85|#k}&qJUuE6K
zU#7O4f}S4hT9-3mx4B0zyGp#6Xty-|rR`oAdujRx(*&$I93zKSIzVnIWWc$_ankQm
zG))l6dg;^(GX#xJmP13NkEsJ)qd<ltQdOH-qOT`ze#Vn*%NTPI)eo-Sd)<YdH-$wu
zI4u8G2ipx~MYn6T00FD%Rb^|;^_qY~wwR1iLUkPG`h2?j098tUozi;T?W|Soml%W!
zJeSQamLaG_gGTEt6BZd{ks`^tB8}_z^h_qmE%B|`$*nOsKlpAks?3=;Y?e(gonN+q
zklAt|zSqon8|9ZjM?ViHT|TqPmuub66uL(xTzR=Z9{hSX#DksM{0sW??zIVCTG{ep
zAanj+Zd$7MAmF+OXwEFA_x`U2a4Z0eacNCsj@8M?)h_XE^t5b@#;qC1L?#^3W%<!<
z&t}}GxVGJK&hW@|vEn*jQ-w66#=Oq64hdfo-EVcYU5kr<rBVRGAcx@()ZHW|8d{pg
z-w~C&R&csK4SnXn6#*AEspRP<<PJ|Da=r*7FgZjb_MUJR9+^K$FH!8;46PLYIE=O7
zwkx8N`-z?E_`_g#7@l$*>S#sTUleOCB7WV7_Sq=abo%^f_DON){XNW8DZKHL(@sk%
zO=wS39!K%5nZ%JCK96KLHDoL8{&Gp(+~>-HYnki;5&F)UP0r{#8axG-SnarQ85A*`
z@Q~>fCJMOfvV2}%&|%`OxB8WOak*(TDr-NA=8x%fE$oD0P>ZXWm*d}mk`KBl_Ix$h
z?`!+=pD3GbY)ijWP`N!*-Is$YBRcl50b<-VWtCw)b(cuPhTF2aW8(Gqq4vaj&qJM0
zNQbhIt(;ukLud&%X-G|$ljBz{3*Tc|s&vS?HkpQ@_ixC{EfPqb*bv2`Ki6n~c`?$H
z<+R<EAU=Y5K;TN-xW-obb}qE+0~YT2G;MpEhO3x%6pPZB{k9?Jrg9S1hr1EV^fV>K
z&Dn|<?N#II+i{Y7orvF2*0}fYW8=Pe`IpEKX>+!}EV#3j(@azr;o!2~vWf}uSJCaA
z4<CRApEPxx9`O57J<q@>(Zx0IeJRp{>!U#CUTR>b`nP@^UjohU*d-zl6@(R^9d|z&
z1p4wQ%*t95ff(4$2n3W^6UrbWQ+8kyUJcNzCT;(+XHj@9td-QxE!Er()L9A#K0(y)
z<?YS6o2>`FtpPzC>%vZr9|C<b8t79oaR>N-M|T7?P^~Mn#tGY!PT|vh!Ul&#$4b-h
zky5!-=Xi4EwCw37OC62hfHi?eKBQa|y~#lY^_P2@$gATNUit9A#K|4e!Fsx?hS0KL
zHegm1S{4Ckgs}Lph5K)gFZTG|4}j&)`hO2iOL3AYdoa9I-7fSto*v4Znw^|_W!r@)
zzEH2X8zwx!EvZaMTQSRsQ6HE4HNB%)h97NW(x+&8_=_^J@om~F+Z9zmTCcdA0aN7H
z;KgDn@e$Co83?#GW@NVheHajO0KmP!w7lyW&@PS_Qf2a>vo=e?PLoEw6IYCeG3-YS
z`Ak_oSNm!Ed%L&=bswKNd1C?5#!4rOc*Gp*Q*DJIX>aLy&Yl;^b%bc5wuo+KI|@Y{
z?a#ZX;t49jL*8;=N_jmJ6bklLCA1+l#xc~%=UKroob&M)8@be;w)x=qc^j0tnSXl?
z+Y)4Qy8Fe8h(<ItGc(rnTQFg{r}C-U$fegSw>D{58q8>&@pCf@a#mU$7$uFV7F(K@
zvH^f8EGHnlpe)3$UR*Wx+jkAMgfff6oeiezDjGo!z;V8%d?Xd!IjOADR!HIx-}@Ns
zSVlbcrrO2cPlWc#QIEpT2j4FkaF7osug1qq*;K?o^{fujD8|_QgnnahiC$H|#u=%S
z*oFDhIA8dpxL$UqJ@ofy%1$lS{7=y@)8};xQBe^}Zv|1Fz<N4y#DJE%3)u|+4meSz
zHs^AT0Guxx(PvjiFdC!kuFD)pok&R&MWxfh9pP@mnm$jsOn@T9E@gov5VC|eX{JRF
ze*bLL0fFlSHD=z=kjgFI78oLef0+PxdC>eXvk4ym$XnzqN$QN^kch)N1!ByAv{D}A
z#Z?A}KGTr$1<pb-EJ6mL)NxqYsi`^|NP{}XzRr@#-lD#ksq8myks)s?a+pi!P-zNH
z6ac&2Aa-9eiM?Md?g$PzXa%}p9J8$wdsUSyadk<zbQLY_vK{3VNVZ|F{3CP$<~IUu
zuar~}59$m9lJZmI&uuHpsW#eR;i9IJ-6?us)O4IE?~_&M?Tryd@cBO5=bQ@EZ^yf2
z$D(!y;9O6Iq^j5%SkS)D|EQP7US@q@9WuU0ERU!MY-4JQ!Zz)I{*h^x=XzB?EN-y+
z2W9DA6gVg6cMEeTpGtcv9;iH<nshEBj8VIAEEjor5l6wz_bx&gN9!%{B0hm&&_KZF
zBE?30NeD?%2`ADMGob=0`MPq*CKHg-gC5+Yo4-g5B-}-fA;tS6;ge6naDdQ61cNgW
zP|K9~4z1nN<8wXZPZza)XDlY`FCLq`EU{<D$%jSgub@}A4K2d?J9PacMIAjg2I1b*
z+M(CCb2he-V8Fv|rIwU8t-`ygl!R}Ud~U43w|YQPgAbZoOaHN-?uw#S9oQ{@RsLU{
zYBE2Q@gO4Psmb_i)R9y5>3i0CXaG_$uHTHC%TF2TUCz6suQ(CH6?o0l$~t{zU~6!g
zh(=D-UUwtmcY@J+%IN7bZkZsYD=^um<W|d(V|XiZd!{uv<(9~y6TJx3?m@^2$wnT2
zWMm^_@x!lN!QU$dW)1y7LRbKeDVEspHgt*J6zhhBU3Q<_deh7CV4@)mPpT_-G2GQ%
z2nIUR1Ki0NtRrWM11mnb49jQNJN+_fUAjZAG_Al(&pQV?n`W?r=SbZ5I@mPXqdhqE
z>Q|*EP}w<~dQng+<ine64O%r`YZ<*Bc|h}cC+2m4Bdv?f(~C>=XIfzktAh8pzUuhD
z#k@1=)5m8ABk$Fe@^7!^uB~rTPJu0L_sOSSp95UlNQqAV(NYs@zB6q%Jz2lu4fyV;
zx|$QP`}}~++7on1=ks27nT^+WfXoMwwCF&W9ZztzF;^?1G|}!J6woa{w@^|pGM@B@
z_)fl%v6j{qGb2ShrX7hizgK+OVik_(me%L37(!*3BJjmNgRsrDQ?$KF=Y(vwypCoR
z1Ph!XmZf5WcoXs}5|27p#lHlzoZGa%S|7(8{N^1rOR#JHu#Tyo`dbY0U4OtIf<^QH
zvH+@W4&!D2hO2_FuxgG>Vc{{9M<1;TCW#DJNCrty9IIM(3^<S(wS0mN6IFf2(a*+`
z!-!H{TR^>DLCB;!|7n`&?V<EZfp^L1a+~q{zcNftr*IekG;Wk;pV`D`qolwdpdo^M
z1`ivYr=N5&GzS$tQV?CjpXG83A>$>+l$G3UuZfVvg~QDFp*7Qlm}RU6A$;scwXX5=
zFL56cO$+^_dbiTg+|s7qz%Pmp>Ps~GTEQA$L1!$Dxn1BM<Et1@Rp0x0O1Y2B$ywDh
z1-m-G;G;amclFcd`<3>W63p}VhtE7SGK{O&pGjfw$OrV0tjCJGuqTNBEFf@sGMbn;
zHJKMzf#=H;)kCU_ol6Ue`szXf{KE-}c6|)hay<|EkTB9La5~iSeeMKriHMxZceV^5
z_%@Qyy-&c*1h-N;AqwCAc|KI-O?(77o+joP2jjCv9^fpf5cyZv*ajsJp4L@w*mK_*
z^!?e-KPVth2nGnH_EpyCi015nG%X)krl#N5K&dVDSq|D;3btCPY_Q<^t-t%SCH}{}
zK1byg!Lb=cyOIhLygAKFii?f>G~MsTXv!mz%NKQ^B@j#~rI%b0dU-~!V6fAJQRIvJ
zI0GzR_d&N<iHPaYy}s}d&|J|L(8V!EFcrxZoK4w*MrE(h3SR8a!HSYGNIQ3NI`uQj
z<BCbkeI~l~@HxzjP$SR{dv<L)ynr=hGla|Cb`*R~S+$q%w5T-0NLK&0kVpJKs|c*{
z2=o?9LwtgXD3oF>tt-)Dg8H$lTHCT@ZiUc6KXzSJ82fZY-QnW@;R!9VtglL=blFa-
zlo?sGo%eAmM_sk=Kt#$HJBfsA;TY$OC}^>n2@}pwMDUDorl!~2c$=7qQjoYd>fh3r
zf6V5+SZF9>xVMiouJ6s<w@c^NJ7Xa88@12yQva*kUsdK~C|&lZZK?XeyI3uS&2fXG
zzZ;z1M`&UVdICgKAzY8RMMhuy7CrTQ^XKpPgtwhWNvhYkztSqvu`{y#Z<@+KVRW%p
z=s5p>IZ;$V-Q1g=eokp0Ja%WI<WVzepUYXi6vdfLy9bijGgg8gf9yz1$)DWEeY|IA
zu0D280<zd(r57uI7PG5?>`(Yk=^obL+Z^_Gvgg5B2bnDVeo=5u-d_gRS{2@V4hb{w
z=L1!QzA@SFiV2}9SJ|N%nCl<p=!|al*-hmrUI~ABS)+)sztTl{mSpi709~@K2J(~1
zA8cC$U&TW1?o}pYtx6AdnT$*m4!Nh4zBo2mw#X%X^eN2Oh@1FY{ZxB?L{m#`sfK4f
zrxkHeeP`{<mG~)NK0`F`VN(tc-bLp2#@0l5UYO<tzRixE2YGN!sWxl0E=R*<gw$EP
z^pRs%rvkp&dDCl6dA^HWIvzvm^Ls1tvHM*MBt~ghOJYUW-*`;f3oTaC2i6@u18=5K
zk*dq(H@Gv}tq-Q%YT3pDVq$lqx)OIG#18^HyoD!sLjqUQC`0+1TR6CO?QO@o^u_z1
z3t5diLF&6YOAWE}f(j(yI5S&6JqxIVr;fR8>`w_}<72`8`?oiWK0Xch#c{Rj<PsLr
zRVbHD5=Uef%2{4&HfB>SBFd!*1gD%cNfu8Ajh^c`{z2JG*CZsBKDk{Q_f_#<mePC}
zHXvJqt?K2_A{2tr_H92BCz1i)a~xOyQPs{I>!k@eIr+vYeDXlp=W`LKImY&g``VaP
zH)ll_8S^AQW1OaIw|Hls74d{Mh|dOXTx<%HYX6OOnVG+BCX4WPPEwIdUvhbg)emMk
zamV79P!gtowS7IEkb^FDI#}>NL7eNKwPvfkCXww_e`mBejWW(TBWifT)H1Z$ApXh1
zsW5!j>_D79K2v`rJ|{hU2za~i;zLTXi`7B|vz%H`Z|>%boKLf5>t7&BE~!lMlt3#B
zAnrk(e|^I=Sg|-_kk+B%$(!O6tx!l}2B5Nt5N5vEZwS;Txx%m*M}qs%FY~%ZKnu8y
z9<Ki%Tkjl~_rL#-ujS=h_OfkUVY$|7*~YSURm--QTeiD$E!(zj`*-bp&i8jad!KWA
z{u?g5o)6veSSx07$*D<%Qa^vQJ$koE``i9R1Qtu@sIguZed|a(S7JW69betSs7(*+
zaIT(U>C@1cX_%d&&^7@I_&AiQ^F)fG$-o&YdQsYuY8E=Vu%3>RT!az@yRQ<;H4~On
z5=_ywBlbrA&jeAEgjbWQpSodNFI1^obobp14QoT}vrp`2nBoZQ<yKeVwy*fIX3xq=
z!~l;rLaNsUV76&{&jZQgOAR&B+pSBx`_#M?ROIl|_&}!yTa?iEJg1NEed|b){AUBS
zoTB!gnz$p!SeU81ZSu7D9(PXmrHtIapb}-sJwJ3mZT#ewqJy22wnQfjrx!vEtR;16
z(I_Z?CI4ye!C_2SrC6q+L!XE%&y0RXtkt`Q;kp0rh?5kbK)^1W?k!3&=Y2Qj$7`r}
zkERr%)VGG4M9R5j*W&(^hmag-3)x;x7MjzP*=i{J@K5zu)?z<YKWwCXb8Qm5c}9*`
zj>*Cil$-^%ZKdogNv1SRVozp;Iic{au0Shc?|Fh!kJjD^U@dx^E@3W?U3>*hEgrfi
ztICgM=l-a$zp95)wpp2;+TwiqM<@d`Ii#Sz`#ICyo_;8g9GL>AC~S6HbTR@<OMA#c
z52tysdrKS%`yWpl(E#VE|BBpyC^F-XrPB{7=2BVfQePV@^Ard81DKerFuO6?rLjV1
zRYJs=|5K>)Z?DdEZjp|-tTs<~nZs6-kP$Pd+*!Jq<lxoK3$McbSq@H40*R@*qJKtr
zaOe$w$0FRQwc)rb=-ZrMKT62$*grEp4SIn^mK?7zDu{ifrQ^LV_(t)M3in@VP*wpN
z_q;IHjEJI<2fET)N4~UzH72VPhqMq{V-y~qJEgLPyr1FYf5VseqFEKLWyWzL)m=UI
z58|{IM~BO8hcAquBo>-N25nx3Ijrxqya-jRoB!+c1>4Y}`TX3?FU4a%AoxDu1+YwJ
z;Uk#M0vb;ilAhMLw@fgo*YvDhjE7nGtwKI<jx=Yisj+>lhD|X~mPl{adM2WGvv9WV
zUVjA(l_3VO<f>VAz&Xsg;^#AqBxikGIGp2dqmk0bk6$vGCvS=*Juz%6>sMQA>?FQ+
zefj&HL~=nr#yht>xF}2Y-C;(BlS|qh#*C0T3>**q@(4$C%ZZzZFDtH2`eWmV53mnD
z`lf#R{Sa@Z{Ls&Of)9TMEXQpOTsVCN3rJUx5Oy0CL_dA1efygJ)O@yja)|tkPH$}$
z^Uox{7Junzi0i(w$0m)}&=;RIi?+;lZ`JQKA_0GLFtWQRo~3I!Bu@MfrbON;9<$_9
zIm@8gFaj>(86qm3oh%Dv@sCE6ITnzPmzGL2z=HKB0Kj|PMrcuaWp?$iXK;Q0^ZJj{
z30Pwa5!680#JF1rFv%$)ep`9O;!KLQc76KbG!S90an%5#L!;~c(;rx29Kv%2mJuwU
ze2N-tPD(5_2Mi%gV=?NUfracFZLf`gRN9z?)5ndP6cjxf>}i}12KEZ_HZT4JD_{+0
zZ6uvXLUWXkS3Z8U+Oy$<yJruKDxJ}}<cF<?V~)MX6by+*_YfKniyITEn$_dc47Q$D
zu#MO9B17{E%9&@c)mlxvX5hn-gsu}p0QqZIU&Z+XVzXEg&EI!~yBKZ1|ENUS9&Z~T
zdwW4nVo=B@ejy+<Haq<Z8T3cAowTye2yjSz>YylqZ=}1lc9Cb)+R$DrcE7XNZGYh9
zs?;UE%b1d<PljTtcnqYpdEikbFVyH?PDg1!?z5ZWanN#bC^=Hrp@p*XZQeWJ(vp!G
z<gi%;ZA9{Lg)Zs1nTIs^uXwFF1kKoEdSlZL4lO4E>>FL^N{cLP|6sR`NK1G3-}N2=
z0O!0^ev`tdOybt%mB+0o0k?>w{oxy*pS*UO@ky=cb2V%$>iB;Hf?dHvHQjHu{i}H+
zBbUbB>LCD?gv241(5E7+QkR_`incH8rJOF6c?-&+G6W~A1NCwit!32hJBD5=d9p7c
zeaI&_u14>tl2qPwAqbf<B!n-Ruiv(3tyd2+fL3m-w|ii@GRVj3S|tc<g!i2brc@lG
z+X)%F2@GyDURFPsX=Jm)!B&axr+|$TK%035D5jQv{=o-P=*3>1v^@frwS4-|!D*g`
zJe{pJJ%LtjM}HLk1HS?TyKJulaLQ2UDjxlti4KNSH5P5Y54DiWf}hyEGPbVIs?^~P
z6{sJ32jq=EwdD_wIb?aOq?JhCl-lnsIK<j1gFPI{rGA*%X63KGJWi+rHG>bINUJ|x
zN71Ej&a@X#CT*PtbC?HqKP%lz-yIxNj9KS&XN{y$UkE8ysQ2qLzDGgkuULrdyT4~5
z*4_1x4ktZH5i@*;qV<?R#P&=69~C)Sm{3TVnT(<4_Z0+)5707S_4wdy>iF*H__1<R
z0xmh{7LlahtNF(X>F>4q8Q{8uNQTBAz%6zYiEhjHnsD%XyYu1H)6)VS?DsK?^O)5R
zJz!1oj_K$vR-6VY5`8d=<nWjcvG+)8a={Wo*zT06oIRH|`x;Vo0B@BD=4R~d^zt#0
zms^p?c|Zd{L>Jy=g1pYo5p&at2ePYbeQE`t{3GUXa%+p|B%9aG4dgh+=%3FCc}L&@
zL7zLqGSDu@Lt^;`RkD*lq_yd6LCiW0wMq83>zc=@6taP$j>h3FQR#)X6TjhRzL(=!
z4H%VH2~hh9Hcl%E@c+!hfnR^()b6>-;c`{C7p5#X2(=Aa?L9q4<T}xZ2d0Mfr+s5e
zzm@aeOXr`_6_I4^p+G}{vIboWZ=QbS0`#66S0U2ykYOG5-2u)7T~;1M21&@QthXXt
zPcyxqB}n-gGCs3%N@6$^71`LfW2kmoo>9&Y4edPWGL*aKiR_XjqyDVZbZN{KJojB{
zamegjr^^Jf9{Bb*hor8Iu)@gO8w-|jo;a@EKZp_vvi1Cq{5y551PPhb-xes93I!Pn
zEUFGxUHjkNXeENACRpnT><qW(rC!?#l#GoJ)<6ygYRSrAlaMyWIC_$1nRrA@1Ki)V
zS=C^!LI!0NNGt#<7=JF^JGA-Lw=vQ+b?Fq?{c_X(*8B3@iQT%iY--FhV!%(#v*T}<
zfg^+!JSL11W}r^|B)ICbp;ZAhG_3l1gkrsP`hxshC<ofWMMtArWw7|=w<nnO?`y&N
zI|Xk9Y^wkgjXgn3;q%mreHZKa&c4BNWmait9=F_H6&}Jjo<XN|z*9Yz|5<hyDa8M)
zHoJGAjRF<pkGH2FZ(IC^vzskKE`#W&>zTc#=1eWHB-me1NZE!IYR@evHsoD)VY`|`
z<v9r`-)ac;YKlA_n6xj7(>Gn!#ZS|CbNjUNofLN+y72-0>#jNV`cKwT*jPe`qB?KS
zj|f24;FrtNc2aoM*Q;`>WsgyAg=Ltf_U3E9r|k&r^uxh!$45ER72W&UiWRU~LEhQM
zDZ*L%{8ljjHu!z(G#H!K5qAX(zpV6?bIVBXYcx2`2?7Gq{GJ|M;L`t6aDi=Luzb+f
zxlH1Gxnxc>m3JPmPZ<yfb;`9kNe;CamXP-6m=3-xNe&)c`h_N8iv<3bH3Nn1pkqxZ
zrEyNFc71>_*F$@I-kkAbK}OT}K@>iel-_x*6Ui^sWk?JySW`PNw3E!OG#W8kyZ-Gq
zC9u)StE{<}Ob$`24l-zs(sjR1$g@84!E>olE3J+Ru_ZvwYM{7)6iMX3I)w|t==T-p
z-M4%%h!X7dy$*yFs~!0|Yq^k88flUI;j>u@Dmzs5MZQsz$41!<5fi-@+-#}?dLof7
z>Jy2Z--pkY%E4c?Dp3hPlsXQ>lluJL)mC^@b4gy3CvWT_&eS}vL}Wp66j4hsLqiA@
zvufm@>(kNR6}jU)Nn+8ZsZt)7JG?;Fb91*=*k&APQ!x#8ii$DFRz!3)u?ek1tbRV1
zk?}m91dGHk{pFmN{!{A(CXelZ&dRP^AT%2s!Nh6h!l{xh+ouRy_?(O|hZKQSo?p)?
z<?sHzI6B+5JxTTjHR@L^&8yhi0a{65@2lAl!8_ZMMf$aK3=7fH$>eOvsZG_|R-%x&
zi3mNZbJYA1op!4zzL##P)y>}-98J-;bVvTa4aR5WP?y)osU^U-5KSHwH8Qorn*N~Y
z_1*wFEq+ZcEwm^64My0Que|8!7^SoeMhU1?2{Wl=*!<;vNJGc&eFDKb9|rKMr6gZT
zw4^$76H$YSRgIc8YrlVsWsZb@)TQDsEM00^iev1Tp*EyAaU<436_X?^uM#6g+gd2e
z6<_xUijN&P_MEwj^z#}1rwrf^_KVgLQKr7Usv#L39i9JEV=nC+N|B38&iCt~hPIo6
z9jl$lvmkA~i~&@4T$bm|6Na@09%Oelvj(0?6!zsw^R*q=V{~BC%jF9phk^dYXDF!A
zJ@g2+Y#m80(5%uVeHsT&=B4a=o~|;)q7wNrW}g`ev90gd8d2m{LL{VR<&{0`Scs*h
z;TJ{=7_dVaRg#2_zbCNAEE35LxX<E?VoUY1<ny$#P%#gU;>HJ6=u|$zVODW`4;Yv(
z|5D0By8xGnND~?+trXtO_}!z<wu(u~LZ}cJt{7Mks$mh=;u6K>rA~m_d~TeIO)QP3
z3*halKMb5tF~vHpa`?N9C@0Ng5nzr-hhOtKuEk7xRLint5nd#Nl3cRD?L!Mm&hn>_
zOxzDBn-JB8POo`K$s!`SErDqh`_7GmlaQgm&P%G2nlaQSIi49$@8pwO8yO82WzlaD
z!Blp3PLE5U(XW#B<LW*}Rdb__o1x-Zb*V<gxh~}S&<61lmgCq`9|Y}Y7DNz3S1g9+
zn$Om*;)$2$04~}@qVT==!+F0N&x*>WVkyvBlq4@<`lXsADS1PbMLNvl$SL+qb7L;)
z$H;-ZOu-1`)(&>0V2M9`e)>Uu!rx?j0!k<n)@VXSBT-*&tU6o5GLD<VM3U+*Osd&e
z4J_i~{Qg^VYAb!*iUqm3bW*?~haKJ@T`HIff^$YRR#~`1d8HTA3MQsl{j4Q1_i0*1
z>uxvoj=(bIum0K_fpEv*R52M^9Aps-*9{*%do+la#4U+>g~tPGcd<K?B);j-g@y24
zUttI#kW>;{0(P5L$taPRi{Awqd|R}fq-NR}MB)Q3elN9{>Werpx77EPOl!itwB!t?
z(<iMy6DszugPSCxOOIrkZg4pB*``R@5J~FaM&kRgYByNZz8`sRS}iV#A~?j{gF+@g
z&p7B8NhO5sg4O#6QiS<GV4Lu4rtLeX_SV^gwES=ZgZUCFMx%?Shs7%(i_iI?1}5cu
z5`+a1nxAJnVf9x^bsuT7lkqFR%~5e*2%}_>T8x{VajfNQNVJyTOMSddX8nwN0Fmex
z3_OlwxLP2)4n^u0$>^rD4a)rT?}mW|o4O)&Qd(o=3w{z5-BweP$BFp#R46T#^Bcbd
z(>w-iK_?7S+aUQ;bi;4@!;St*3y)rbTR%UI32pg&;)r29G-MHVRsD!@vUkk$4Y=so
z`bDNZ+Jk#>_#cTr2-nA6YBf>rC&jzGFR^6%O)Q!MgK@7qE3LOORDdGA+enzW6C#y}
zvz3=z-T7)Wa5C097Ig1PIZcbpf6Rp-Zn<V}-;-+v<&mP2uvU?NqbR#puio$R?PwDQ
z+w-n&xl~gjhuQ54=maZxOTk90f9F(ZEd+aHjz_cp{PEUS?J}Zlnn`Z0jTu|--m(Ua
zdj!Lf^o@HH@2XrYNlQMcudu4qUVFLbuEJH#M_;-vpLb4rf3B27{UOv_jDdx_D+ELp
zc@O7=Y1Be)DWrCv$>UKSh0-K)D?Ql_B?_wopCV)Ic$KgNi7uKGH1#EIpNp1agi{2j
zVGpZyYplkTJ^>2#PVgnaAuyfS)AwoIW7O7=oCiXUf!nERtWsi$3Y*botR-c#=LH_(
z;uExMez<1^gYMofWvMf^OsGw-TT{40XA7>uNLZ<tp2P?H41B}3r-3<R{!uj1+dpo>
zaiof6-k&EEy`>d>=)7_v<ULFbFTG6`&MWQz4ksDHKRE0;^#ztlj&6~Q$y2fZZOvOX
z;A=@Wqsi5{S%h8R?AfOZNKg?k9V0|(8#gbJ|5-f`9p<g5GKQH1^72$fa4RMMgA;jC
z`3Kw+_}knKl;5^NU&VIAmOwn+|DpeQ=0IQ#SmwJ=^~5r2N(O(}4o2e0l%t;~Wg~VU
zJIJ8lxD6ztN6L-yI=sP*y(ycaV%9q&<J?x-5N03pyZ=L-a#k(Sy1s5BB1LTQdi;Lo
z-1n(X?E2*p+)vOWAl$XT&10TYRmbdu?YXYWcO2%lqZ1P%5KU}K=l9#_KF3^7{n?lx
zzn#2RehBI1r({c}|Ngz*{ici8p47T&k2_3?5i+3Z4)?ANVb4vmhnpmBz4EI;m_7WB
zd?J!gMeOwS_c9+#U%JE;Kx=Kc{=~Ni;r`wBx0@%=57sjy0{x3`4-2A~uelg|@MJ;U
zy}a)G@1BXE-!$%~0B!3bWaK#d$Y904VI-0hE(&8jmlW-_ek?j)^^ZCV#RR5HwQ8E9
zx2F>Psaw2B)7y5VF{;vlf0S84#QzUYc>=q&d$6;-msnoL1LG)63$w7DD<esDbs%1r
z48&+F?^@Zj3NMS8il^tzfY*5|@<OVCC}Q(-;rbd{^$wN(#7ZOf5;A8Atd10aF>Vz!
zt%Z5BnyR86P1jn+upL=TG`63N^F=agNJULgGVMksB{r7VVb(KSN5SQ?zkfH8Wn*r^
zQ5+MC^r=z*(#S_Em`SRNRW6rZe!C?sz36^V5GsT2DQR4XNCThq0(iUkYEX~GidIH8
zsMD4dD&I!iNahJ5ac`lcdsdT@i5MbLDF)>>W<!HlI`yy>7Lyz5HZD<-2g?Bum}-gY
z*c&Lduf7hncno;>&cGv};XaADw21VfB2N8d%;OH{Ga8*`OWi2>b<h&6o^~A=L=JuL
zqek*!;8x|wW4JOY|C6t4MtA~x0KKb}ZL|QsnitQqTK62+CB_&q@larCQ$U&QN?lTq
zbq4O{cY%FGa#imIL=ZjygbTdDCzDV)u`QKkS;Ik<z#G^h0ze~qnreB|vZ)T(LGLAM
z_O7T@v!;O4Xq7OcV4TIQ{U3uk?j}^T7-f4Y@hkBsJLeE6G}EiRv*j6`k^}-aI8?t}
z-=6M&3dy(^QBid>(?pAH%x_oqUwQ}`lXBnd+s_Q!W^Th$_EEm?-d+Yf27t@C9(Wdn
za%?aiypNJFwWeC(KTQY0Zi-QG>pTfEgFPi7u36V4al>WDC2zwZg1fvtbX8?_g<TA3
z|2RV4tYU4TqU+k|-@`LRpRR~^q@<1ga&6_d;VY8k8(zK=Qr_LO%NOG^5(8$IIDhwV
zfkMtux2N+bFeqP6JX4BHO1hy4<7Mq%hH}b`SZ5|!<$ZtCo4I%WBN_bbZxGNx^`DEo
zAMfspa|zx0e+H@idn_WHVrfhU_POSg;3Qil(@#A~oqeqKr5DaEX8Sv#{%hK~ZcF-A
z15e$VM3kb_t)g1c3+|ZEKbo+-n%pj9UbBA<PB56^Qay$y>`Vb;S|($6@oSwJ!Hu+m
z#|k;V%N2ZLiLBwQy@0br19Z_*2#e9A6fSxnnU3mFfD_wH$kD}M4X}IZ8^bsC{M{BN
z$@|gnCHshvtdwh6setPw3MCobt~jeu@*5C2LzHTckQebi#nznf65Jf{u~Fem4#rw?
z_ulf(N+OT~e#i2po3fIcI#wX+89sB;)RD+MKCCdaP}3U)Gq1WqIXKgxbbb=NcH-r}
zwOl9YW$#epXwsB>RqNyu*RYTeY-!W;l-(l6Jabb2`zhDJbgZb7UNG`BL$px0uhjVU
zCORpQB}P2srFPS));ixX+i7BAVwvq`g3vhB91PZ0yXyX*o$=yLjHnRU$QbSVkCE~9
z*zJd7(WJ5D-*;?F^UWGWk!1X5Us|<i@=lr^fW>TUzivkq0``9W|Gvd@DZdWKW)q*}
zd~V3VDSXJ7sVDH!UDx>8AO5vp0%ZE7nwc9c;VRTPH~)hyeU7nCl*V77uM_J>#n1gz
zeD~GRk6c{RSa9~KO>y&I4Swm;^2YNey=&3$(Wa;6_9ooN$)x1h-&d!n09{W3mOx8h
zwggk!Fhh(AjP_BEIoTVwgr1x8Hj;~`iz8VlU{}|3%%elfDoX|F=R!FzMaR#F9YAyu
z&GN#dn~%bd=v+JdZ|{bTYwavH1F5HZEn$Xo61{4ZRIC*X6F_9{jk@H+p>5YFP6Si3
za~erRn|CxvUld;4k0o7NzC1Vms&72qnM^onLnm<zx^r{RI`i2}7X4k%Q@rcMSI%^S
z^}uo&YTF)tF=4J{TJG~v1CZ1va%y_W_f5T!t%W!v$cZ&wqTTqdy_GGh{)8Igx$Wxu
z6rU9X4j$fzIe^8V#1G<QC&brow~RUaeZcgL%*o7RtEJ}2%<cQ6+L})*w0eGU!Gk2I
z(F!CuaI8>pqUf<)D^I7cr>!q9ZHsFQYYV1Pdue>FO06#PGv$x@kHyZ)o*XAR@hf5Y
ziF=jz%}F)4@u~637|Hy5^}LfV8G$NVn1FDK)xJ}{lt4(6V5UtgyVam19yJTRvxecg
ziPxjtpYO);Ca5rZav~@X{OzupX2)EoJE(-zTl!GuN4yhapK}>v4|}GrHg=Ufu9SjI
zcLFM2m{MfZ^A>9E7)F4ZoT2B1`|i9rFsnP6o<G+Zcef<aZfiY33|3JxI*!VPab=G}
zHU_42>I?A<_J&?zT0PF(fOp^N#@jjyFdKp!l=ojoF%oR2SwBwH4rc5n1aWXXnpBE0
zC#vKw0kw!SuX8~BItva&Nv2x(DT8Kb;y_aex|E+kzfP9I$idL%4GrR5F{lp=EY5#r
z+Q;cj5&LL1lDA$%Z4~(>T$#YJLssw}M^KpRGD~u|KrJx?Pn70s)T?Y?z9#e4;|L&C
z_Sd<<8?PV?f^Hl*$hJ`0yDw(6^^IJy;3D=GX#R|1<Gw-)+@l3WL_DG(;g4{>##>?*
zZXTKM+-yx<>JO=YTXcKY&j@qH?%8oud;Qt5Yj@JRhgmDEPh>JZmOfe!Z0g95T>M6i
z)h^Jt(phOw_e)UH4X=#S*!;Sg+dNfM`HqGo(;7G&l!)SDUH~D(wl?Vzor|9ng(LGs
zf)r+HMd&%>sw;3rJXBkdCiR{X=4;HCA9(T3-}p;Jcj@Gh_%?KNxctC1p+XaHQ!ba2
zAc_}ew~BP}zUw_LNO+Rpir89YD4a{>$1tjIpf0>GvYqK^6H^L*b+nc<P(=#+qRDca
zPFRkWMH;6BwT-Nr`~Lo%`k_9%4j=r5OsoD{ho<qpY0B<2Oz3Q8B|zEg5<3^OBPyWP
z^4I*RDrQ(QNtsu}yr4!}S-B^_AbX7?F{_lCDY7@{Q~u6fe<&b`;y6>FVbfN^L@d`B
zn@)PtsaSqctKOz4&NUr^T!$dSl#AHcAibD!5T!1FL0|G6k(&P!_b(eSOlzWEfQ5?v
zupANa)wTssLl7NC;)y%NNQU&m56aPsBdD^t%S==#E7@H6Me@_q5RMvcmE)swLmIph
z#@^Kgl$R2-)qTB>SX=%1ufcD)XS{B}CX~5>oi&dVLnrR8e0gbTYcj)(waR-p_ooM~
zJlJ%tORITCgzweKy1nI$t%wy?EqhyCatHA<al&RZ7t2ULo%k76T(|D7-QMe&!yI{n
zzpH|XPyDn^8Iiasl)A4An-?RXbGu(3JQf9)<6o;Sq0=zdM@EovgTg6f@T-4QU<5Ah
zCB{w3lt`O6>+>G;-^mp0nZ^<NwC`=3FW_-!gv$raRN`_?J>M>^WiCJVeA@`MwBbmE
zcYGXjNu=nxFZB8E1DWe1pZR<U5o^M%P!9GY9Y7QAas3%<QE`F2GmJvVe^*{gk)#96
zMrYJGxU^e+*YeOrMa88%cgZKbHMh#IhWtroTIhcrQKab=X#tj%!`LIi@=yT|o+2zv
z(~q`%)vdp3>Jlsm?WgL~YmY=D&=es5*J=I*p<-)^Kq#0*n!GeIgjPL7{QMn-IWP&u
zTdWPEJJN;2H<FPk!NRL&|G;tnA!Ar}EYLId>T*VT!DjASLooR>HV5keILxIm#|Dsy
zcu!c-+A#u*OI!3S^=)OAHk4Oe4CVbgNnK(?d!r#~8_9eAKrjcT>eqL;k&#hxGP8N~
z@caMuLSe-ro6u`pTIhG>oyX)%OX3_k`3$t>C@qeT5*myNlG8p5=fM2;8O2AMcjkVf
zjMj=dY3Q;Vw6aZatRx(*Z<CW5<*UZ?ui^Hz1=JIML4k0C?AJ@|X6AK!RH|ka|C;bL
zL+l}Je*i5}B4nWPzdrftj(&D#$}kSMIdi)tVSuxA*+BrDs;-y9<r3iPWS~#*DsLOp
z#$De2ZlK1^EzoQCiO}7bMqUc5;-2wG<1<*=rMsi+yiv97gH3$qYw27cyCnEz%tN-r
z`0*^C==q0^3nTA9?c|r;KEQM5W9!GR$4!9(=RDbnSL?^=5ey0JCVEaIr#N>!?SH(w
zj*W?qZ6grTk_<nWG4XZW8N?thnzcQ>*5&R79Cq1<1UzB4?WVmxt*Z`gw!rCeDTy12
zU6a@&;>WD*L0GML`e8Y$Jr%CliVPiT!OJsU$y0$=Eh-b;M?XfMn3j8Wd0EDGY8(;M
zMAbOm=ao6~lB0za3{Oc|Y6B7LYf)m97>Kuvo(CiKM3}Rdr`-Lm(OI`FT7jzz6nW~Q
zBB=+cI#9dU5Ns@0+mi|saed}dxx0}46-LETxzehK3C{y~6^nEm#^Unl&lM=`m#ia?
z?W?a3+Zo?vg5;1CR{y?QqGb^Eog>2x*Ooz=>hN|(!@6l-;}zIqff|`;>;eMXp#9C5
z;@{2Yhyq7aa}-Q%4<Q53$F=3NX&M?*8P~ydabZzRlxq)Z__2W3X3bddhk4WD?jxC}
zbt=Im;NfU<obKCzja($-@NFLPb})ib!j`Gd<?*&KcBZ7D>(k!#@$B5nxn$#t4eP_C
zsys#e4YtjfafNR|=I8CNhncZ97=-QXxKS3Noamqef!jqRQEMvTo0UtrKhfva>x?K0
z;`{r(SlzcGY>Jvo{W4*KrjGvZ?6bEglrwyP+3pa$;J*&eSI8rLLKfe7h(t6TCY+@*
z8-!)oOt(1F#;?TF$HS$-7$?PwK~9idV^DJ$o8f-OH`BN^uN?qlYu=$7mjLg>XoW|$
zWf=aHm5a^O`U&qofE5t;3?8~jHV6pZ^-9!n`u5K)KIWJIB%mY<#>^sqNGzb3<3Hcb
zn8B;Re7O(HpwWE(LLLR|-x+AOd2w9Ik@=Foli$(mU@7(NV*@NM*VC^(1~Ei-NmqTM
zb8q4H!x6?N<cqER>G(xu4$BDoQYs2W>vp3FpjI8QL84RXj{58U1>>0&LQvk@Z7m>!
zj7&@3_6Os`qmp)bPspkT%1>=2GtnHK&1q6nQxk;OT3EZ@@j5u4GYBQ6aThOb&x>C#
zRRFJqdQs^Fn7n4=t_M%YfEO9tR>b$8?}mpi2i?Iwea|A7*L~WlEl0mH6;obNO2qqr
z_;eM+nI>}MPtM8>-ByOl(^Fy(b|f>Hj_cc6M%HlS7?oRojO>z?Fj~rR8LAGj(Htch
z8Ic{(O@z@l@;J}K-k+_18cmzRB#g4y*a~wPs8iJ{j?#TSG!OeC_*ldj#}+#lXVAa=
z`cUDx#n8doFtq&%VRSGJ3)BoV4(kT>R*mq`HvPqZ&eIgV9u2E;AjHmxFW;TaW*=!`
zSb=fvt|c`Vm!*Y@iqf$Tnry#zToim&#Z;5DZRyoN-ai>FkvWErS_yAKU=vPVzQ*oD
zXNcK~Hw=1T{vwJkl}uUw6u8ahIdTIpKG8)4EGidqZ-Fi{WV7ZpFgIpch{e7F-}hwt
z(Xryv{v$#rwg(<(BgRJgYr8v7(;LVhqQG-d%Sq`qVO&{cT7>ABZp&V7he^}GR`=fD
z*YHXka;r#x`-&$cJPhLFDD=h26m5ZcG*Q0GSFiJd2l!J8oLTKpFi_akhz5uPSMVrj
z843^k1#R;jKV<Xg+P1m|ELLB$iGCSrz|4$!9~#E)tz0KVK?Y(%Am-Gt25+YgdEi;6
z8m_x-{i^r^zJU0#TdaCJrQ@YIj6k*9k+W90u_5O1j0E5NI$}>B9HwfOB;HZ1DuXK-
zz<mTOPDa)Kc9PDk1_wc#c>NF0pz-u%xDQ?{e1frBj?ck(j_WM0Z&zGxR1bUJFGkpN
z?`x{9(7GU8YAi9H?+JCk%!(l+*1X+&c>gLP_Bxlx72McRrCfVLl2?GVkW-Zan=fsr
zA!@cTlc$2VBzEJ@-g;AG`vk5(aQwi@HLPtpBFmJ1eJmp$SLXP2Ran9#-}clfFU5#W
zwm;mc%mF89h_FfgveF!D%73aq;L^`2u9*jB3LxZ0wx~_WNsN2n#u3G|>|10;4~Jw6
z6Aj#9Z!cqD+W~iFbUB>i_nN&9HQPBGwHTXluwTk-EGa9z=Oi!-5T-w!bIbvhZhoti
z9HKjI<V)yInL-K3PquB)M5*{eUwU8My?xT$FP9{GV{SAO**4wTchSA7-|osWT&)Ag
zs%(u+?zQ}fn5L@)k1^_1a6H5<FTH%p-45cdqI5fOtXB?d5%oTAav{%5g!kG1JM)gJ
zBF*=GW$@OD)+3f#NwkZRzuQ0*BnVnBwsJrLZGJ~8E>iM0`T=fZaUO{QPqC*yH@rkq
z(F{_%ZT7nLhx&G=#y?I}jf1@>*oZTMC5UNMMiiS3Z({)Rw+)w9X`~M-%wEp{W3txK
zWpZ)^4S6sA9lExTf^G?x0luR)&kw0a>%^|Wr4jkBtR>ZXr+}BvBclXEF;#Z87(LA3
zlOOHZmv040oTz=73-RDcNEMsZc38#=BAUCeXugPPYK<uKDkay0(<Utu5j(xUl)h^j
z+bVDU^sqzxn7<t|r(NXtmQWYbHWL*DxUb$WXn#?UjUod-!?)Snd#`)Cls4{neU1r;
zT|P4fynXLpP20B?2I9(_rB%5=M(e6xZ*_x53tapnFt_rDu|HPF2Ho=Lx|#I@PO)`v
z54SN52rAK?rVPua%G;hdw?z%KTEJ`28k<@B<Q(5_<X5AglK@_Rc&4aZIWUL~ZY=yP
z<k3&MJLrgv^y~cbR{H*@QLXE^#Z2LoNv}7Wv1`DdH7fb|%VxFk;QG_=EJ7fTbI%4(
z&c>~wJLd*CMdD#Iwx1O*dni=il)c3XX9$@OP#o-Q4>8`VkZ23Xn@IK3Fa2Bq?VZ;t
zt~S00^#K$Zau|Z0uHu5my-@w0o2$~t=SAXH!H1o<mqmYNN|kl5ZOsg8y@~&d?L`p(
zDkfuQBGhpmhu&$!Y$Fc3EY-0DCc|^}S=l;8LLTehZA;M3pRm@9(!$qVGyt!s22_o)
z(ia|Vk|6kC6cY^T<`8<4-)Eetm-bu25i%;?&zrUt7A^~#Kx!dYdF<r#o-ro|Gc_rP
zE7vnEd+N^V@z81(c*%#qS3y6$A8du!e0#m{Y;dz*VP@B*qSLEqIHX;0516>H0KC5E
zU}er4uA&tY4Ax8(zSiuy2Rk_R1p$gw!)p?b8s**MeUct=)WP`>UCukqQp(6F2s1ot
z4NZk<9J<5`wAT-fWO-wluzy0_q<tf=-jQ@-yB~7miX*rW*H_%X*Jhvjyi#&*iTjja
zpj}Zy)or&$$b9x(iG%%Kp2s2kePo%>{<}LLdEgf$dX*403PhM9N{(0k9$3G`87!4)
zLHa=K$P=q5DJE%VgIq)E6vI?9@p-yDf(6=so-)mD)Lg;V^@FIHN5>>f?=5tc>+A<K
z3L$*LpQ1L?HQ>UMBHKX)iDXG7NZ-YgRMUw)ZC?WE^M9OWVs>VMDVO2`<$BE26B~r9
zG>8up!D>b8e8TwE(57X(hDBu{%y))}cKJDjRp}$kyDBf)*RVtNwDnyZg5GnUfyJhb
zZPX3<cJ)CHB58H!2ihI%%vAOZL3-*F7jq_TYV8*asz=4|g7(^hHg6vi3X&qPJe73s
zOWbF7!h35LyVsizdG}m`6?zBX?d7S}>E#7tCTGTnp=CVx2lXcwxfBoYpA8QA`q)aV
zL{sa+9H#AUOuqY#5}y#v>vDpb1Q+$L$7yPPgE`O^M;)6;6E!1spbiDO7hS*E%!lnu
z`)!|hK_K-~^C5w;qk)}p=lIC=#~VG7hGhlQjlK^bkk{&+Rxk2~NgBU}(<364pnT8*
z%Xvm3?1}bU-)_}8df%DOgj%_sgAQ0!hZ5WPntC-!k`kCrV~$nYbRS;k4vLB9S{z>Y
z6nS9%o6nA`@x*Y%`x|IV1mfd%U~%euS6&~Qn|P^4>x&JNe|f72{|ftO=~Dnv2WC?8
zl<X1yHj_al#T^U4Z>sju=Bpmb^1eZN<9sMh6=v$<rCMqRr6ct5gz^UI8;{~Mc9-SS
zNg;m1jaT6bXIpx#ZJOmXGaX9dsUM8^UDaer>}_QaiJRB8{yZ!1t<Z2+<r!*afH<vQ
z=eEH{ob<c{Xw$u7&OD4bF=u*uI8(5zGQ~OHVnj8BZ25YMH*@EwU=XQHRYHi7BjX^k
zyW7ppLkD@SH*bH!0SVNSKvZkfpPsIo^o#jn`1DD0!q&DXfsrlt3y77wAyHdxA|`=P
zC2A2zc-c~Nr9<ttydnu7zDI6=Q+&`w#PK@0gqZxsRt4#+-0(}~qRD8|mhbxf3FY##
zsuJ?HUya1Mfd#=5Vq>@u0uQr$X-l}*^{;<)c#9S!+7gTKem3|s^#xNnem;T?-5e}l
z;na|(Y?j<p-}dHbP0P<rf*o5%^)hkqY4;kS2SK&dRX-I?g_N?ACd*_xgbZ!RzQ4m3
zXnEUm*>^5Jdwj&7ab4FFOcEk#zStA-0X#0b+{D}_QD0N!qv^nUKR>z*j5xiN=+<iw
znPquzS=q^Xj-82SJ_9cwUA%Hp>deGKs?W>IpyU$}&_DqZ%A9PwSrP&*(t?v=q<%5#
zj{4BHvY$TkFwIaIA#CWrOg%kyFyq<bfOC@HnCe!Lavs35i=kTJvZQHeLU464h=A<x
z+?oS}XHLT5I}ewH6pyc84~5r#&w`cKr#BwrvDsB2yL$N=rKuQY69}hSH@xsc@s4N3
zy-$U?-c*@K<zw=I>8+=Jy$kf9&Bj(sR9Tj*;cuI*ni^K+=fzy@tE{lNOpgJFSk0Z`
zU|fzg6cH}O*ac+l<z#F5tw$@BiGJ9?=z*oc0xnx_eC7ItXnVc9a36kEO+)sM_Of5V
zevA^~WjNJdl%|61=SUu1l{DB7OZ)V7i+gEE#1bzBcb@!iH%o%^ki%9k?f0!~Sukhc
zwtG8VY&naDHxD`yQHDw^nqE*n!4o65O4jQ>fn#t9)$we~gWv5&&2id7q1-_xF7_WW
z>recew#f2u_iE@>!2z^C235#=9X9AJe81@n8(ly|4%{Moh=$!KnJK{7?Jdo8-C$$?
z33)%fxdqx8B7oe4Eig628GkyPx$l8QU>e{GKE~UWVN{zwdxu+A(gW9Nw;Le(w0les
zH>NcczO}>&#K`Zj)6QeYlaI$=j2vn(1Qm9a4^ko-CMM7j1l{n6{yu!4M8Ez8F49w1
zl59KZL?w;puqk%11Hv2s+EiSnL!`l33W7P!xRVs0YswAQ8<CvAw1wL{FcrLhUIy}7
z*o&HsvAV;KXuy_T#B2UjUqgy5m^ozGHrz<xS^qnxjMsDRyuZZpX+6mLjljr*z-?1S
zkRXt94T*KG=20mJ&}QDl<z<`J+4P*V>Y#?BYw6MD_G<xAGMuNKLIGk+{;xqn$IDkp
zHtW4Pl7&pr^`wlqlXI2pk5l+^hIF}JMqqH%B=edLjXe(F?PP2<zU6Y{?1TuN$=3qC
z?{Ki)d&62A$-vkrO3mKH(9Y`%p!*@b4?zFmvj<RC`S693I3_CKP}o7e$hB09a3m{J
z^ah82to1@c+HVMEUZ>j_CR`Pl9UwERQ@Pop9_Rh`TRXHX>+v+sberT^6oXM}y%eSd
zEh>?XY>k?D4C<A5^~t6QY2>o=mPVxv=c5BI1oPT+ZC@xU`<wuZuYQjT>nH^&*#iX8
z{N0*!kdBGo^&s$m`8+siA4z}t^j5tQJ_0BpJURlbSr+PTj9*I%LsJ_usPjQ7zFbMA
z!IgB_;THb{X~pPr?=?ZlwQ4toFOI;~vjE|#@>?xHPi&WnMVi~8yHB3wTy2CT9q9FT
zWuy4E(%>a6*6ZpPsud6cVL#%>D3v%04~R)Qs%k%Tlb@)O#rcS1HWq&dU1iWmd`-<(
z->A=?#?ugMTgL8Kj%jEmRsX7R>rT-CG~5?x@DjhWcQ@zQphpRu%rijs8+!PO?Rm15
zFpJjc7E%3@8_(bbVaRgEEabJHnvrimf8s(Eh$-B5(e)k8$>;ek6(^AUIvCB5?4;Gc
zj?*-z<(FWUG(7!-+kHYTF%7JBz%OL)21`Su8H|T$Or2>2z4lK|HxG(?M2-W{*+f$O
zH}R}tjNeKA<E#UxIz|V^D6j?AeR#9m$?7{mvY=D|Bh2{aV(~Hehqw*dM~uDnC^F#X
zDYNif7vN^=1E=%L&D<G5`}2pWZ#aT4JA82$7B;6c9zEW#F&W&Wk1Zl?-jln-@+dx`
zurMI4n_&sS$rJOBnwwC%;Dt7$5)#DE3)R<FW9~V~;Vh`kyg2OkSbilzEXeoChcX#t
zbH+niQv~93nQa>@?oKRrrCPu{2b*bOK}hv|Ta)dGvo_@l?|O@3ALFcW;of36+i~|2
zEvgr2mT|9Z=#3@D?uD7Tu~wF+e#(0-qH~+f2m4Dt?}w-WNGCm8#K?9P!B0;usrQM)
zEG$I48=O2bQBn4<WoelgOLHCi%G(~>seE(&T<(-&UoL-3W2qXqYqa@vV|A`Q1)NEN
zK6f&i=qbI<+q8&m;6}Hf<3`Ec|6uTQG1L31=l_zjA719X%MIx2d0z^Zji9GK<646g
zX%lbrQGwL1l=kJE3sBbl2+_Xw9M+EAbxbru)r&-|To;w~Av^2s4k&vAr%g~o4!HeY
zVBx0linCb}ivp_7=@UZ+0T;gKK7VcSIx5gDg@Lt;r-2z5+UEPhjYwb??d<^Gfr4mR
zzG(stpRvolyXtvV5C?c@)})vABq5?Kk;LI&-tOZ#>5mG3yMA4Y<yy8c8H6jt9^yC!
z`eSb4_#K{6w8tOF_n&xkUtTs>c%|aS<Dq%~osLE^kSx6qPX?UiLBfJq?J1U4)`t>`
zBYXG-1;#$R_XeO^gyd%zc~623`8B-J#I?%_+hu(t@o{TBz(aLftnda}N<?Nt4JMGw
zETM*v)r`R&hf9m3qn@VResw<nMD<`D-!(kMa{s4--VX~u3xkJQWK&=yysmYL=F<GQ
zk9}ZFK-K^)g7|y37aa5R<m6$*vXe4fTrrGQ+?`j{2fwl9<9IxF#An<Nj|Bm&-u~(C
z-12-qTL$>&e9fwJNVh1{hFNt2NCy?`afTAclANfVqn3@J4O>~26Hb3(Iwc%9H&{Xy
z*Gf~+zK?MGnBSqUO4T^PeEg!fU;;7{6K@otUs#><*-e&r<%)DsL>+pRPZw5_wwz{X
z?|{FpKxTf9YE;!uhq?kgS05VPkC@C{Y%sLt-d!m%iNF1x;IGlSZ8G~;Q>e0va&k6@
zjX)Edb2TDQtr)a;ISYJonK$c&+w04@7AsX1ufflq>s2;m@ttb?H9MB@S96q_c!H2t
zFeOgNUFKGP?W-l#Vdq&PEWa~(TrI;O5?$n6_w1h89t<^8=q1_h5ntyTxOTS6dLH`m
z`HVla2r;vclFzIa`K<jc<-Abc89!Pau*aEsYuaxQM7H&8O1hk<ANSPBC~h%KQ=W;h
z-<i)96vlx332>M$%W*e4_(!V;0+?lynGSkXFjyoVWrvffy!YD3tN7l>BJy|3@&DQn
z#B85|D$AgRn3xxxfzOk~HV3OVFgq+-^A(PueVp)r)XD!h5}p5sRBCf~)<lb;M@?23
zZWjK}Zu8$iCWAB_CC_1k9g)>r9Qxyb0SkM6Fga+Bg7F5!qJO2wf1@piNVQEkJ|%}a
z|E$;l?;mgv<C$4PaXe%l>d^RF_&(@Jud25^JY2T|SCMc-&Cr8}j>u3fY>DdODCH$W
zlAgCOAhrgr-!>M|r+@GB3&b@>$OibMxEd)4%3<$v4fuAtyZi4X+6&s}M&{|Z6$TK0
zj}$+WFO=4T%4pT}&k|lt23`#;N22avDgh!?pM}VwuLz)$-+XgxClUJkwoS0xz^1cn
z>7PjZk%=FuS?v53S-a8KM#B9m{ke!APMUt;xx$NnUP6D5@p}%ax3P7kypCU}rXlYx
z^x@rsjl4T;ZqyM(BJ9|&4~$WB+Y3gFzIE(ZVS1N#R{EX9P8pZ1m;z)}qgRZfLt7gE
zf>BYn9L_Wya2!yL>h0-<ATo`zv@dV65uKx5OPB%?GI>)DCnIJn+rSXruV7X*|99`$
z>!dylE9LUhGc0~zJXRQ-Q=#07uo)eU*{g1XY`IA{W)`<WUE8MZqVW7Ch_%=)CofNR
zWMT2aeMec<Nr991=(2U_5+Uz}%c)6{?1Sr~ApfUJ$T;|4;XV^VqDEg?gXv!BzvB=g
zN!Lh6cIKAmxW49J2v59>tIlwFjp|x)4dVv-a6iq{37(<u7*%9F(hyW?Cm~bc?=uYG
z2=HA-P>Uh{iulMbN)(4Od&{?x$cji9+&Pp*!Qt$9ougIEm-Epd(HQ2dG%`*07+{#k
z0!5E-^O1rsl*Ym^RBx;SMD*(tIHWZ^mSO(G#d<8AgU)00{#g&S^g;~-_Sr4)xd@yS
zKg;wV;hw8w(27}^=FoxvE?tR4P*dXv$|1#a%;L-#9Yz#sKs05h87fR)yFldct0TYV
z%#8PNq`;HZ3{Pia;f5kFjg-vFrC`x!Ejc8VFU0keQVoFH7v-(#L(E*1c=ryj{?(Or
zC4u-bWOSJ}yn`%_SH+*aQq1auc!W5lfuoNyXXiW%r(QBbJ}R9q>6%rsWLt#P0<QUu
z!kG2-0Y3ghpPdH|@&hx+%yBP4Y7t?q_Gf?)-hP{g2lbTK`Eu8BdQxfW34n=BHI625
z<iWWehJx}e4n};TzMGXe2eoEHg$XUy^E9H>o1f+Ux?ENQ=3A1rt;fz@?~6x|v+6_P
z#(HSLV-ZtZcagpjDaa8Ko&x|Ssj3VGr^_dax(%Qc0i9y;)M+4Qe3VlC`8=<yvB>|S
z1y7wuC#wi6{`b-MSj&WOkJ1w4?-SpmV*o~qJi=ep3n9MnlAJR?Uh)rnEYK9<oJW>7
z!T{CT`*`Q0t6&I9FDE!R^+FtVq}T=@1*J89eYKD4JlDYuftBO>HC{0>AJ($dm)+ak
zZ=-lB=Gu%tr^GXY-Q+Ab%&0(^>XcdCkwvG48?b`JF>^(txD6X<%*(pybO6(!`f5FV
ze3o!*47AkKLvsC77Va;?l`HyVepGR(<k-%HYI^I&=@9DVo+z{H&?Cu(9(W+yO5$qs
z1(JXOA=45o22?0G>fOTM5)+zwu`VcKFFMXka$P6kINYag*?dGoU!AW`HZMEl6~_#C
z?eh6h*8DSq+S`M|99@&0L|VDcPUiGy_o$6**zR$m+pi+Ty>dPQ?51nmH2F00GV^PT
z&5oTk3Z|aP(){{O-1#J%SW;jH<k3cTw&kcq!;-PP51hI2Yd_|meg6pFj0S4iTA$<e
z`2Ia@!~{QA-GNunP-$m*yF2&tz|tu;0x3HaIRoe1*9L@aOy0RNZh<B@V#PRSlm(~4
zmhU$Z-NP^Y$|IEUcH1esH1J!kezY<R(Us-?sAp#6D{tk<6v>!ho2;A>`sF~PWglK@
zG$tz1hvmYf$T7a-<Avw<XpcGkr2)-NGD87u#H7do96mJ+c7lD%7%K;EJuQVm?g`jP
zRYwDREft4rDcjF7zWJQt&-=7U*hJC8HK^a*{MXf_*MLHbLP(2IcnYyu?8=oN!`UQo
zG;s{Imp{sU+V77^2Z&)b`l2Om7G%u^;s9=-v3dM-e2M7OT}P5Wo2dbJGT)oMB7%Bt
zAT>pcmF@A{tn|=w7=U!07C-(FNN`ht8m|lMhtI2cim>T<OBSNg-xu3DhUOZ=7hmYW
zHi&$a%m>DFm*uDRSN8X|XlqT^pVD$bO@e(_7y7@HgXCXmf76As{oLI%$4+?1e0~lL
z?ppA8N927gyI80#2zXrxx&ofOl=u#8)w<`unFoADJl@$tfA5f*I-^<EV2{m|rlldp
zsFuccolAB$NACIZj!}MTXvUx5KK%=2dx2>!Fs(p6-fJ#BQyG=RcXxM}A1Lo?Wu=`V
zp+GNSbfd2Y!%|aIGcaxB%zkh@saQ=DfCa;<vh)6QdfJgyKd6w9XOWMPTAk!(Zf<TT
z;7e>z51BpLfWO``dBIp(Lf5qBKxDMa;d@jV%}6~wqiFVhOQwX%aBDz;35ysz$CvS|
zmdi^1R5mp~QfTe@n$DRAvGdp?BK1iY2MQYLUXv_YPEjGoRACdL5=e&-zXoI_iqai3
zI0``!J5zi?5`5eC+Temo8#(sCwh@Q%Hc<c>6L+S|RM~^vJx*zXgc={NehlD9-(AdS
zTXln+Aa{WKGHsCoujm@ux+Kn7w-DYBcFP|iOd_MMt#^c9Y1`RPQUL+u>l`Bzt%(MK
z<oIsgSkCDqe|DM$Y*u)*G+)+WMK|qtS8UJJ_e^vS#-wAk*(4nr4wN}rHF7C0qT+s3
z+Ogd+_zpg>#*zfgZrGRD2A=AB8_rh?SLx9M*+oUACpW-3l^v7p99h>acFTv$T7(W~
zisE1w1uZm(?ew2%=!R`~axmVW>7$%<+UG^MY;5*5ieH%0Ha^@m2g}%{*$+<)h`!%?
zBIzh2+*;XLD8O&4*uzS?3)7>_Bi>~?8;7N<UVd(*zzBIdh|Vj`IPjoy?@Z`Kq-&mG
z*n-HUP5DtGV?*}MM((8N8VH;hWdW8IITP{wWELnZts!Ph*YhK?o2L~8&6eZI;y)l3
zMA$qi;3|HmlOP&8L_QDJhQ?5Yed1{3uaA`QH*Gt+Wfs%V43aY68n%)d+s$NHeG+_$
z#pnYnPOkA@lg|m_ScRh_%zfH<N4afD*NZI34(FjEi1;11VyGOXL1U4Dwd<SL_W^Op
zvzDxr8CPZ>F>l|&Abi)(!h(*+;a1k(@xqKXooEpWQx1<;e}j&Oz|=G!+M4$OZnYkV
z6-S};a$V_mByyB{DtbyxuLDgUHUO;~p#Z57x&QaNdw6Ks$4757=1bOV(S$xkKTUCt
zeTk(gMoF~HYVNmc4;;t!%;s*3_iky-{Uhue(+j&&?|*4GeRM|=2to0i&^!0%C1n=r
zNYd<OkmvD$nLCV68t8+ROM#_w5>xRNI>bXrjDO7d!dNt8D4#({d0}jhTuNnjzFIKs
zR=|La^uhdPYnJPM@waR!SsWec>f9`{v-b=W2(#pl<T2bL+RxWKatSUR0hoR!1iSo^
za@34boj&B=uA9AmlL_Qm{=%wQw#HwvTw&&;{ZMyZ5WA#U-rE_|2%|`7t`DgkDbXU%
z-<HdUc8al|NH#bqfMj7V3AsB71$!9j{5G#w!bh$sqqIx`Lmh(3yjGWij<g%;mXvZC
z)@-E{LY+&41K+mM;TE6kZfxaB^As30*h*Wr_3aJ7Zc>Esx$B_N`YEpU%%2(3Si3dq
z3a%A4{vq}Cyh81}K*CyXpoMNYC-t;j0&7BhVbrG;TYdpzM!%Gg=_8T=i<|7J12gT6
zkD_R4k(S_n2o3aG{u>dbkM@fJ94V#8E`}ORMNzV$?y3n5#q9Ei<F>dj)votSd36b0
zho5mX_IA2pbc1Yc!7=aS440pk2!8T-`+Cr8qBD-^8N^N6>~sc)T{mDm3}7+FkhzOp
z_03g52vZhKXA1PGiI)gX3^Px8;^zFQgid>0PF)JD9bBE<Q~j1W#~DjjvtZ3&t(|K7
zE0=L25JG+_C^f6*!*ooWIj8zKCTT_krJrS@1o56Z*2Uww$7+sKWnblyWLG;m5&;qk
zAsLe3+9CDNo$AqUHazy@gKsot2DelAapIu3wAdMfUZvqpLTxQ2X1t-AA(>E_ttQuP
zNgP+JC<@c0L?>vP^OzJ^lag<yEyX6>i1iWj6zq#u=f!#vCS&VNJKws8`fwSqJzG%r
zrizV7N2b`-tTxkrJg5`-UA1*?AE?v(e*wQKK-Tri2(shhX*CHOzdG0nzhqPZ+>}lM
zYGIVmr`Ls#FGNfu4aa=LGaoS0sQ>ud0uaeNhevw)C!RXCmUyRzfV_ODf1B-s;s;vL
z3`PR8)q$r`#DEYgT302k{6Lfm7e{<d24K)&asxhysHTO{OP6oMt#xTTq5OmF#51T0
zc+#mNr~C`@1tZEjR?~yA4_gF<Sjdp@1wOcK8Tio-j*BjcLuT?vmW*D24DlAu`|z}~
zFh38oo6qp{d}&KND;;SYC$uev_sZ(37(pAV1`pdruD0+^Xf}oE;Vp%9WMM(;`MTb;
zXC)cR&S-+eD$b^Ck)Aw0spT8!$cm2`MXPJQgN2#t8Ldmk3)<&a!T`hKjCP&hr0vMV
zP&IecVc0CoBq%FmHo16C22f0lQb&#*T~vdSK9?O8ysPHyy|Cwce@x2=y{(J>dDFea
zh@}iMkf^8T{5xcEJ$_op+m=TUjfM}JXW>^Xlh7NLJo!4|_3q{HVV|8mzE3L(BX=3C
zJbV+{216TdI<YvTCoIM7US%)JkI$#cW%78RK7gK~3dX~?4Oh`c?xxPunMoHpz@6xX
z)sr$h`Bf8-TZYp#-qZHsruaejumS{*tps5JnVgnCj=b2pAkT}ixyL{CIU4LSf-I2+
zX|rUehf2VJfWZY8ycrf5t`Eu9BN@;DZ_DYv;q;JA!n_>1Ql$-M`AZUJsTjt{taEDc
zSj7+ZC{$u!QDoXpL|_mld1J6c-drQ!M7{{qKPZek*?$fAj?zox<3pzuQvG|O4X=CE
zbVWzrD|?^7P{TufJv}9`yr=ptA>z{ak9SMM!bop?({~HP*02Yf5+TM_y0s1%OJ>$X
zaIGUy9+DP}kN~7|3z!)uW<ZHx{FgtvuMmchIb+d%PJw{OA$JJFKZt}lg~Sx&aQ%=c
zbE-3akkXWPFjax(RcCP}-#S1RpNXf?{s(N@%`kv5YT^rR!|SbAlH}S~(pb5{@9$TT
zMHr%UNKyuZ3%@BDup@IguuMkAl7)pir3LfQ{<9sEH2zAQByDbOsxgQR!3bg%Dc~qC
z**eH7Sy(lKP8mJP3r4flo8o@2H$PAc@`ZPaXLJ2>r?ly#1c)6Gz6=ZxJ-}ZlJOpw}
z-YKu|46Z+9?^u=s{+9kx_S6e@iUec1x6Do6p1LE1EI!bbGGc-i{YYWv%ZPelD{hp9
zrb}k2M_%WY32C%H=zu)-I&jlU1E!#(1L$Ma+0GYFGs8#EGUqM$p$ugUz4G@ES$Mxf
zS?=~rg8LU=+M1~I!&Bz87<I112nrqCnX_lp#*4;zTpK2~ttnyop<iNa+VZ&-X@IiX
z8X92W=I7~*Xy^hgcx+K)wz0V3sTTK~@X0#qa#%XiGiR4nW>$od)0~|NVsH1noQ1Na
zrR8&SdBT+27O|t3C@im>>udasR`jC=4ub})j*miC-r-jC08ZckX8Z-gA3!QWvNok8
zxiq>+_M*5_0zP42L{H@4h8JbVpRIs6^u2uUTzBvq;&!AhUy&7)u;py=$bt@z6#*Ri
z!=ODQ=RDbBH9srut>}#S&=0VRGM3EAxzM?}S*=%Au1vzw-gADFf$*`TN4o=Bwq&hJ
zAHqrjX*jk_om*K=j)}|`2kjHbj%xa{tk`fyw8}A%hA({=Mhx5JY!$-Jtun!u89{k!
z;i8`f=~r}ME}6l2BTwik=!prqq$fX6VRr+L)fO}2dA9zFFlV@qEX?a59=$M_M`^ku
zCAtclrLUe)IAn%*P^{=&@;_+b3;!ZV|B4}MZb|A0t81(l`SZJs5!gD25k;Dn5q`If
z1cq&Q^vyN_fv#tyU&a7BedbKakEGIfvs(pSl7`M$&83do;cw2nEiPpJKL#wu^1Rfw
zpp-CvSz2C}aZa}C$BG|g7b+@-62=f=R{iqZD=pvAj~a*?ICvTey+gHJy?b4HU}bhq
z4V%!4E*@Fb^Een7r%s<qPM$uk%h)t(q2tL&(fHCK(nzmJ;}7g<F-n+8p~G8{hI|l5
z!(Tk>m8TYDr$aj>4JR^(Mo(0dYK2ai2Gin^i3S<tX>|=FEp%u!w$U}$K#`HEhx?k7
zh>&}v0pCEl2~&Mfc~^DTOWuk6KbsMiiC7(ksJPNhXgqj2>n>^O06M&L_U_T1;5>K2
z5Mbpeto|5}E?7E~&pR4c92i_OyDj|EKWIo$CvKQ|hLjGrQkD{<2gy_MREp9c5eayu
z4J{43{H8g87K5*pMJc(%m#`yk;aQpOXL1Xjv%)f;tjZZ1wn&j<;^t+mMVQG?blY(l
z=u@6@p@a-NPYVso$+x#FVyuM0c%sISwXqmWr=+joA_}%TGr)!p_^0LUsI}u5X7jR=
zz@VBw0jD$j66Ru^tdAfMOWOBvD`*p!eW<qg=2C7k<(yE}?f{IZMKR3K$2l?5=qJe&
zjyzgEsZ;31@<SeR48()c<N7O2;b%<yRQg-CeUmTBhl)2VgZ6C2V=OtUhRKxbk^b+j
zz`{a}P3eCJsjK8kjREA8^01vE=o{lA<RRn(X~|#Ed8iA-;m9-cn;y|cw~U0&DwCvA
zKg_Y(#Mpr??Y!GKR*A?9F4UPbs&6J-kI5xS_EBKUi}ZjdmtaWwU>Gflu?1ZmhOW^H
z&%w~NGD0?tAL{z-Y-XG(@*(}H>W{0^2F-3+eWTqx=x#UkQI%)Y$X@fUlIb9(HR*Ae
z+)Z+2dEkswX1SC>Nj$SLG!{5IGdd+&Uz-68ouL^EGP9?K1D)57!UM}nk#$7SV^(YT
zov2D=?O!H`v&bZ!t@a}FOCFf5!th&>%OFArMc$chYNRHJMpvML{GQ(ekQlWO8n7V?
zzVDBC7xYgDpu9u9e>!F}<Wz1!3C?gafNa!4ilno}6bC(-y5*&lt!B(-8?q*cAH_}O
zFk>pk)o&3AE#0Q3G@z5xQy8qDKvsFBWCbexOW9c`+>>`;({wh$aVrnB^&d1{Ivo~l
z6-L6=fte5J;dWsgl;wf;drc;__`|@#QBSOIKm(zF2pZ6p3PYAOJzpyAF~$q9w5xM7
zVK56lR$1uWLV7T4Se?gFj%>*aE}Z4jnR)4tFnTbmfMI6>G|cf#-xK;Qi!JvV&|94e
zyrtwtP6p4JfmDQV{a!E<)<rJzNS=a`z{Nf@VfBU-(L-jeW%SDWD>Xu7kdLD-LK_1R
zU6J>&-Awu`S#dLCg>r%(ItbeZMTV9Gc?^UjvYNsQ6?`!sFbpz%`f^e6(w|TkK8CQ$
zL%*99k;-Rfgdv!fGln7b?UITet!v@phD7yhTi?-lFs#B>yYQy64KmxthSpdu+z_4N
zTRjP@eEMd~N)+P~ecKUNT-Xu9YMa(EFL(erbiY;oaaG!&84a54$EMvp=q|61{PUpl
zml-*xkx}u<%V%{yAAi`MLmOBwHGTJDdTByAl=LxT0yP@lvqWu<2GtymleK2dg5em?
zSf@xMpEZ^&J%=rGWWeAR=>Yt&ZNoZ0@)R0Z@*~ocsp$4i1FpYpEkjvj)0)sBMrfix
z>yOx0LUrd^--m*#PXoih)K2LR3z&@t96bAIDyu^F7XTH4(T!@<45NDWrnx{k{>y+6
zx@?8v-PQM3-(duh|JwFwP)T=?1G)4saFbw<@kgw3ksqX#7C{D@sr(XFUP?RQ^Xuyu
zI5_je)A{pjZ6Z1H6H9v5E6t{^5~K!_Z7B;$wE5FI5vKJ@BIzfrZG~Q_RZ_1+=jN<0
zv1jwm2xFkn>QUWRH^Zy*>96cIGFBGo`)n(mt$q~I;JvZ!U*tt<O&81t_2jkeO7bpn
zNkHpun#Ly|7UpHAg-j%32!@>tfgdYq!BJm%$LyU9?;CPvt41kkl}CLSI+h0vEYn97
z+HR#yh>;R<L><h=BIw(mk&&{?JAHq}D8i8^tA|zxiMJIhjLze-1#>Rl!pV(zR_A~_
z-<)s9b~A6;Td|v54nFPOGIHv2ZtqSVoKHW9W#-#gFY6%TM$0jNM$Z}!V`lqtX*9IK
zl}6SY71F%(xUCvuPh^+ERkG5c`aQ$DBtj%xqJd!smHXA$lw{>_MVsQyPo?Kwv3H!Y
z9?3l0AcVFc%uIVk0Mlohg*2$E18NZEU==``I=K-}Q^4)ECjl0ogdnFpW$>v^BE5XZ
z%hCc6gK)l65ez<VL&quQKt4neXL6?i-Ae-n=P=A>b~7?$zz>vsCzS74*JRc_d;~je
zDeJ7t4Rlf`Feo_7V~(;nHzz9;tWw~-`a{2BP?>M-waJl}A%komV%8sn4N2Iwz)BEX
ztNessA62c3L2}>E0cQhi!e#j(t;)fBRdy(_rH;I^s>CNFFw9uh2osX&B?3xFw#CWW
z8*`a{kkM<{Ac}pqKL+EM6&B9nurjouBt~U;7IQ9+ts0A%-pQ-345!V3Q6k-%5cPTI
zQ&r&c$ZF^2Ot>I}vO?dSSL3@;VYM+cyo~Po;O|vG*}UeR4_ZAx@^BE#ENjn(*Ks2w
z$Qp_&cGduSm@4I2W8~pS6?Sdgs7OXS`>28Qtvp$kn6#yrJ<w~_dsxKY!}m5ep(Gf1
z^1gdGo|%Eo+5u97v1IRh7?=nRMGV-i;OMFjLl6U0fLy6-8Zd$`gOdOrZV1ynKn`OC
zUnjsrSnU@`iNlz&K{L-Cv3!_}(PZ$%5$=kiF+58?_`Z=6V^A9mCz4qyy1y7gO{*+3
z=5}zOEoCdKvI4^4b}@GSYD%hd)$*DQGK;<!Q)zgl<uhu}&i-)2Sup0pq4+93J}*^p
z`MOmETSklR=j_D5U<@lh;bX5y7bkR!TJQEcxg<a?`>}1^o=-2FPpJxrZDANeIJPnF
zhY?gIVV0H|J6T-Q(2kxz!XUTpi?o3uJg#TXd>hh~YJ>|KpnWZ?-%&^98qhwfJh^d7
zth@cH>NA3}K}-frLu!lcbcCI5OP&o3fT6?M-sb8~21xRc0Y(=#3_^@5oup5*c>J>X
zIvB&_C9|-{(XiAeOY;>%XM06fi}SO;RkHTccM-<NMGYM+qgfPR6o&vl-iPyPO@z(k
zGf>ROHKAiLg}7QS266a({l=(!FDys-yxX(+23O^KSPPqPcty{?zWa1D8?r_?gi?Dz
zjRt++gm@30y!%Fz%ieU06R7eKE`JMcg!YKiZfM77L{qB96S|sCn<4?-LMH*}6K19!
z`U-<jKxjf{N76(vwOUj@1qxf+Qb0}CA4LpWQW6Rq+AWfF^&t=^BV9mNNLbTQc?-%l
z0QUZRUG=8y-cwWgOg8hqL{)KFS{N|T&9W^iY@eFxZiUOH=lzyCc=&HPc%Fq}(R0R+
zBG%klsbna?g_6nQ(vTIEPlwKqe7_1qh*cmAA>HC9D+I>F6S6q!2LGg1s<CYG!0L6;
zDN;{g5=7P%ea}*Xbh03Yr&bv)w$g3Omhv-aH*_l@XKzhLwwJRlE;Ddg<>ce%%`SxK
zvMUG}fAaxhmBgM+kHnLIi|@^cE2H$)0bX6IJA2IgMvr*EBSYVR$DDQmN0dJcq><g9
zpilRzYl@bhAR<48yPkcMl{bucF8^R#CKoH{n1)%9)R}DNTYTQHOV7g(`t14m@SyYz
zuZR{}$QPx-SNW#BBAQirqf9g2`1yKdX+``j&HN&mVd6$b*VFNB0QK8H=VdMWsAwK$
z&}UEEhuhKvmkq}p(jhN=AeHX};fm<`I3K?2&d(}NuMs3Q+6FF~n*MMglpAPpld|X7
zX7j@32H`b87{g^IP`?H(jS|;&+H9WfC8ojA@BQRkqLiP}Ngw+>3b+V{IR47gm5|FL
z6~Y|*xx);zP9y8q4bq?`|FwH-IumMmMKT({LHwe85r@n3%EKC*hb@B1(^29jKY4t>
zisDMp`=ULr4J-j_BU2yILl8l}aXz1>3-g;~!(cr<i!*$CHkdqL|7<4#v$sSV8Dy0R
zl{knkhK67fPJPWscxeKh5|sgp5;C9s%3!6F5|MCs(3E}@b<9tc0IUNqnm_oZofly-
zgp^oU2xUZy3_uZ$b^-S38?SMaD1;c#1nBUqNG@Fyc}~R)=?>Q7C5l+`PLzXC6n?67
z-UZslfXP?KLi3R7HxKLUsHIt)`NXrOj1>zsp|j{Rh~%1T55uYk1NE@tA_L5ot|_Uc
zxs(PvT2LC}enijEGM#30C6`tv0<Lw+JsSo89;CHGlZ+019tH-E?O^m-mZp*EkHtX`
zI96!b=I8r8J5w8A!j<2k`)zXcrA#C;kni0x7xGEJ%gtx325rffLX0VHH)Zve0lLXY
z-)nN?`*U2#%0RyEf(M^J!j?yPnd5w~#rNhTJLhHnz2O)<bpm<KW$_`#%9JP4e0tcH
z7|Mc63@i=5dnQZUjeguf9ePSRJf^=|2P9s%zsWhEGfQVxcj$9@YdJY4SFI~d4_4qo
zn^?7-tXtOvS6Sz1jOx+&K+lJVrC&yCn7owH99GsQ_{Q6OaHu~#TUZ0RAA>hPmpt@i
z9bJ>J2_t(ifTis)$2i?TTfFz{(mib3u&3?AP3V~M#z#_j>~=@V+m>8_;As)2eK{-d
z?lIlg+^YlQb=ehHYmi>2>u9)(UQ-%^?hQB6vC&}DSo1vu9EHP;YQYf>H?L3-x`$a{
z=^Vm7Yv~|m>qFSXOyy&*cxH=*gQHDeukbOml&N+czTkblKU{0u2rRwpOa15SO|Bs2
zMjYFFOec&H*Tw+q(u{$DHBi;5G$E>k^(KH3rj-Wr<)ZJLygT;p;oE>LPuLE;>xWbY
zhs7E}Os~=wJh4c^M^?qhu<1J513_Sr$S@-9tol2J6fR8%&S19FLqF6DX6<#jt%J_M
z#8C7}SmE$AUVhT35O1QT%@kYeSk>Tfb5m~Ol0U0|7+L&nb__{L9H5_U@Q6@hpc^tq
zS=s=eG}^|Z{vp7m$^cLt=fk+_8pq5!1AIF(rjhbPyUIBi9EOly?JOM}Z8Qh}?Xv}z
z?g!yvfSuzh<+LR<Gd|%8Z~7hI$0$Sn38hN;VJt!42D(i2+2DS?d%d>6Xdf9Ugs{nw
z^?T%nK7_sv-gad(+x`4X1_OF?e3xccaQL)@cU%O3&eIanZ|H={N6w>RFonUJ(W1Yy
zNlJ`Q4)&9Gw)JsFh!wdiCq>?4gkg{|E|?NL&c@(Ef1jQ0>S#WKE{!MWhP*M?TSvIW
zAiNLMG4ifur1gs}rc87MdaQ6^3?o0V7=LyvEn@}n&@=iDD{edbeW~#Fa&uv3D3R{#
zkm-T4A2$OJ1&Fh83%U=NF@$eXSsP}!_&tLtpO6b%P02;a$Ebm*ffgFDwxJ4z7Sjx%
zDmc={uD|~JoogQT*e)Cl_()y<N5Qg#k+c(`iIRpv8bD^GeB%x=8lMGe2*Ve>L)XdR
zgHA6rV(iBsPfi>^-fiMd2xgi1Fj|p=#)LFxCr%ufMvrDCYg&+<gSW^7Z8}q~_Vh5&
z#GpGXGd*--CypP}bUF|6Kx1m~TnEaB)NCLH?tVJtpd-jT%A!9uK<R)_a4Zi+_(rdc
zzNP2=B6#DMhpl^-tV!Pbp^A9)L+wsXm0sI$98@Q5cz9S})_MKeGlLNA{L{K&^}^CU
zy`nIA0>^k2!PLdkQ1Q;1Posj))hT_R!vHYI6`qA@;7zK^kk*|pvQUt^6fjeA8D)!*
zHi9ifMgvZn^P~hs>n7>MXInmnLyE*UBxEr?W|Xp!>DK%#r8mT+({4B$j2mqR$P7Te
zp2#72yzk_x<iZmt`toCKCtHLVD4OG{X4*>cx`*f`l8>q~IacIYvEaKu-oRiWydo<{
z&@JmHko454Q#xq&dWTN<xN0^$&GWZ1V6g5#&WYhe-$QxwogxO=eCG~>4I_eQ?~sA-
z_1KCF1_w05>Wo%Hp(O^AY*pj>@KA${!kEHhI0P527^nZ^T55~q+d1dtquor>p;PLd
z)mscc`fIz<j(ChC^vUjxOfNfX9Pv@`d|#b13acPerbYDuePIB=M^~eSuXpsD@HAZD
z=!Z`nJKhDuI6EWXd%`h3K%beRB;yXb;33>}Cq{_W%{f;$DHm=5wD%01g>O{Rh6Fvc
zvUpB-(w9RQ<0%-Gy>B+(9HJgyAY%{DpKmHrR^VWKLDz`uija<Tg7JnvaG8GfabJI|
zP9ua54=Yrj7Ilm^(ELH@9~AxB_R+LB_1y|3ORl8eSOQP!qxs!@OFZ?Ut0$|F;#&|>
zbq<h>cRw_uupPFE(-q(J<IpEAzEDw+Z-Pl889(<qkstmu(sr)1l4<~gV|gQLHl3&-
zyq2y8gTu51x5hSgY+;*haD}y}2M%s!g^O!sf<xX>XcW;xo(*XakT;)x$0u%6Ixl<q
ze^+0PpyQWcp$~`Ur(4-_0$aTKx#T^4+zlCuXcC9ZH^zBbgY&ROFnKykyyPd34_HxL
z33^|29+zia8=3lu9)gIXe3~xIZ;}mz_4F*x@a@@PEZ+O3Z?qq7d6R&(ZxU(g;B@Ip
zqeF37L(WQ0HoyQfzyhRj0;t7Hj}j6y<PafzfRV;o1~g&W2#M%y6hMFUQU#3&jdh3=
z7&Y=jV;|@+!|ggMk(U)3HE1Nve@VvE!5#*p@zH^o$RIIfUPa+zGrzV2Cvd~Wj7rh>
z<8|(J*atX3gq1@Ps->A<1e1?z3k&|J^OTpV=ovY2o_A4P9@p~M1U8R**q_l@`qmw~
ztXR2~39iN=`LlMV3~+#Wt!dUMr1<culc_F!08t82JRct}1a1*IN?OKtH)tB|I^THX
zo3nD{xvcJZyfPT0H;8Za%WzHJI^LmQlO0#GbryH!RadFcPu+I7y7Vrg0Ro=34>u(R
z4rhAU4%d+dib<J}V(Z(wOt8(!q#VxZ$~t~BnXb`VoTY#I){iLcEg$`$`nnIzl&dj^
z;3w4GtOA?JmJjnSe!u+PkfVrZn>hcn=kfB-P4t$>$v-!t;bA<!4||yNEnJpQ;>*&?
zbYSp#-@_YS-k@w%BS4VmjTa09bG&yVQ5fx8WcNAD1d$F3Mnn!!<2mkQh5{2(`Yy|l
z8XeG5e#}aFd{$YQ{T4$+d^JXlzszif@s40fnr&B6gC{jp)R3|?eE1=w<tL@3`BQ;X
zI)QM;Gfe~6t^l_VN%bP}LbB3Xxl{rTkXS_*BaSkyg2}VoE!aGtVV{xHw`rbURUUkr
z&S*PCia9NM7u!5IEPiuNgF)8S(ZQ(h9hNV<KCYtt-OOJRZ_2ys%`d{~lY2xFo|g{~
zLw;0rybQZR)6=gD!=Jxy`yC3gvW+yE&9CFpHoYm`Iym~LG&?E+MuIt3ML0uK<^1IV
zRd9o(dD=eQgch=Q#0Z*A7LqLtBL336SRO5Zv;u1~HkcX<=Ov4HefT|Z(Q$wpK~@o(
z2y!rBT;&ykD0Fj{7xOKCuYD@&sAVXkX>mooe4N3d2Osal`LsHnd72(JPt&JWg-1=V
ziq}C*^Lq2T@bOi08;<#gXFj-=rs0&BPbX~8cCJOi(vbyQZHNy>3IydZ9WjQ6*YN3N
zQU-YPk}4>H^o8Ym?C?MYD=b2nRUdI(RPa(lYWzroJu9I81OrG76GhB;XDdZ1p0HE{
zO;>sZPM$eXO{0kvDSB#@r5rR;QUO3t@i`lX^duo3h8@YKI|=qVulp*vKE#F*Wi&B>
zEY18TB+FoJrZG>W3_cGx%I9tkO6wrBt$x(WL%NlThjpX|5_MSofta>H(rVlS{vd%`
zd6mgzTeN(6mch1-=Uvb1PU{eDM@^yErMH!(2;S?)hl^m#;%M9R=RSem4f&zTCkB|X
z%ImdEzEQ*%=>}OWzACSlX8tJT@ieOFA|I{{rmXC2l>}UFQS+Aw5a*p=HSzJI2z%N-
z+>|!5PfUiDAo}Z~m7t>hd0vFgJ^rcB(O?*HSgiO#%`Req;3#M`Wn1*}c>X@zlxBn*
ztp=KuK@*tKN_GFGme;j^ChVV8OA=uXZ<a<g6rRlY8tuaIf~GHNew+Z!@CPOa5P*hh
zK56bP&l4fp94MCFMG17${v##TO;k)emu`AWi-A6(5O05)l%i^K3e!t1C2)W|#H@2^
zRw}`a&VkCW2~{b1bz;8k2O;Tk^7N{?m#2D~21Q&^yrmx`e?R2h&Ea&u-dq06KS;VC
zRF*pZ`Mmn@!2IX~21Sl|JsQ|+O>$B7ZoBj@*e;u*v4-3++D*3raCuxy%lm2sQJJkm
zch3OiAzM6;f<_B=dkSiL+Eb<tn3c&@7T<v6$Uyv{piCYQV~opSE&U+*qmb2Tm_Mrg
zLC@1<wp5*XKAt$6ZR*CS0kbV^wM9vgMh_;wHo%1CPib@sNF2_?gcyMpU~uF|apbux
zngvKzn1JMhA%b7!6;jk#bI*Z{L>N6y3phkUTa#qVm^P)BF3$w@mO5E3vfJvEizY~#
ztYjHk#HeycwhE>wZBV=-T2*o6qb@I&?#rW!=BU$(bWs%7mamnq;ZetpuKA={*!vdt
z>D~9s2oC6*zCF6iCQLnFpMH?SMe+|KU(>uQTwwz-2-kj`GC5dH9>#gPgzaw_nuFr%
zJX8g;FH`F)D!^-#>iE?X7!d$E*NLwjJzqAtq{7Fx@P$MMjBKXM(Mh=CZnrElL<Y=8
z58B#x??n)u^kwt8*eJMpm8%_kE_mpn+WQc;!@0OoeEobBG@MASgT5W~;nsu-5JpX}
zfW<&jWpb6p55#0ZB>!@j$pegUU}04TJD?c5iZ%*ajRwAd$%hr(=o%fiJJ5l_V@Ixz
zp5W`Psk3ceJnFWHEJ<#iau0@p-;GWK-Dj@56)y6OaiSbyD$mTm=-9#1^t_L3F_Q$)
zx%QU9sp<g|8VQ!hi0_?H7FeA6zzSS<Yn|kEjGGKJnyo8rUz7Yzj_Z}y`T<wrs?;SR
z&=LUfI7=h6k>AcTg?O&jl`D5sozx8o5UEJ=AXND;c*D9TG|GzB^KJs;;YS@V%7?4E
zBV%V}3PI1)!g)Ur=kbRTsz~3XqCJdUO>v9lZ;I>TOy54PDBKjT2)9Wb^!+<*{H$qd
zLZb-A=#b{aMX>uF=gYL)9`3hLBjAtTecp^@+wy1&Z+!O7ADx8$3hIgU5`t0ZV}3mQ
z#u2oq<%|!`=VH!%Vn-9@hf5JA)2;OuSLUk`RF>%DdHFqKpSGXjwsN(FZ<Dv(q}@i2
z`zvhJf4h>`0ptLqNc4=uXl3xvz|77G@!VtZ1W*|~NWA<DrdMVFL4%=%8WGaKU~m|H
z455IL#t5+DW5pOyHxMHH86a)20t1Ct9z2V+vn@1O$(fi8jjE<;o}fcZhfE=4hESL(
zlr;DQA?Y27<dK*V8$`3P0>+<R(*UM#YTh>DVWsPVF=pGiA{uS*_Y?Lo$ex!m|Gc02
zE8>07<tvhT(73~rSLDQvx*ayUfqMMiYYk}sFHBkKX>7dI2-&#xA9ZD_k*jfXTe;f8
zHx@(ycawG-IrbYi-%v0>-U~Ns{qK#`d!d<P9!1Ol#gM(RZ1?HvIt_FU9C;4LMv7yA
zqRe2x@SXq^(k<C@;2?h(m<B^6yjeg70}EKr4~3PUUJBuMI7!3E!Jwk<@gJGhS_$yF
zl}!{7K5{Bes=#LYNQHx=-~o$MS|OyrX8c%QV1?_ID8#k-&E#l<6=9obpr~xJ(r!Xw
z`0{PWr5K8j@ebpP@?S)6uiES)-c4mJ!rhNJd+&$ky`MTa+3RHRM;(UU_BkxR$?ScT
zpE|S~^=LO}dii6xR|BWzJ8?yBBmEF}Q5tbfrJhZ{D^v#a_Ip3CNsW!5GD0!_@N1y$
zAfPVC`G%tCwwXE_2O&M5kAoo7C}k*G{%=Cx$`gDGX9hx!0+q{GFb$^>;@KGouxv2F
zmCoC;O|Oe*2^cmKw(FkxfMve}590}=C%qL7Yek54;hiz$jVbbDS8Qj@1vQ|{c8T<w
z=~R5=gpd4Jr))xSsNX@$WhkbH6xj^aVZ=X7o#50|Sh*Y~zfExMy>0@#+c3O-_Xl?n
zbY3Te-;O!#cDP~ji|Rz3{6#qX5tnuu`@#B9@jNacJofJFnI(Okn#kJ0B3-hF<Y>)R
z;5T1%rP4Y~?Z4N>p$QiiP?hj);j3shO}oEzGEMa{2wYvc)up%PIZAqUv`Du<<h0o~
z&imM^g|4#rOi-!hkNB_x!?vYR)6{54E1h1CN-G4MdKfP9=Pd!XE;1?gYHfK04p+1y
zqyh^;Gip=@B_Tif(3HPy5Y25eM5z=aER8H-;)D;dCb+_+@uW9sMU}NuYxgrKUOO%a
z7rHk84flOl<<RPI70f|OL*81w93+8<l%+1Ob?HO0@1mFSDU0tSo=1Vh_O2XM)rX$t
zph|ae<jSuE9X+<F!g&LFW4nKKyv;=o4Z823ko(h%KFTW4wz9W{AEm63*5DebE4RAz
zwwy;vuZ|Y!_J^Fzwy7bMBAQXu9hn)yT%)1e=+ZP6aY)9lQ1eIpu(HDHiu}ncO9}zY
z;4nVGPmK_Tk!Lk@G)@x*Kq-rXlQxdFR09=~QbAKP!4P5#9!ZcOJ>`#4q$@~4P)U+B
zyP-yXq*tVrD){{_K=jAg#SDs9NB6L#*X6%1y-7a3z1;+UzhRJf6#3rod^F`>m)E-V
zrhMw)%J^72Qip#&q#w3-<$Tb2%tNLD>ubH?GOZJq84RH|7d7bFqV=`4UJCIkY9MML
zYG98v&_9o}3yeJzA2wDJR}^i@_<q7^%iY{rhv#X;aMVE5!1<>E*BqRGnTt7!8rUZd
z<hR4=9Hz)(K4NuthU<a7LA6iWjBKL@q6VS{q6VS{q6VS{q6VS{q6VS{f(BM|-w7&O
zjn&U^452Vx3A*yCtA-%4_o$uDW->Y1dw)1mtgo&lGjsD@fA1yBIF`wnL9`DQ(IgI+
z*GC@K;5=**OrDMsFZs#i17>AVf>Ifsr)M&^k*Siw0ZlB*r|H7{CZZdxr)P17Z_frp
z{MP1XGBFk2C&Ii#qjK!&WlPzdTt$6uqmzuu7PYy#nPf}q?fUjuGA-Zg;bUj`>zmth
zTbu0WmD||n=ckkPbL;ZmAo&)L;BKv~>9>ByWI1MieJh!qpB1?~BKKsnEjPt%Z)_&x
zGt<fTn!8O#wyt$H*W^1$Q^}aL7t=@Q^}+>ip4(nslP!JHq_VwD_IQlT($2997L(O;
zs|h?d<by@ybF-obp|RdcCa0yTl=dZ&gPPmI_xQq0a&l!e*^%(Vg-4R<j^s`L%~t5n
z>wMI8y6U~NG%NciW!n~Z)b*x`jKG{)IhQO>PxrKcrspahd!0b13YPTf4u%Hm>Lulh
z8)X~jb8dBuIIHt_-*aDb=_MC>u%2+8%m&QtcF`cu((8QB$A|0i>e5F^%XsYuO`~1s
z8*hB;`*{C(#+9`N9<L0hj(1hM(JzCkldlZ#Fvri0_g;i;D?mN($&mX|<4RCL4apzI
zDgcR5Q3Fu}P6KB*H@Z)}KRlZOM1#fGrM0b+hZQk~>?_asT#jJRlVb2p&q#+Px5I6%
zZzP?yO)+G)#E2PBro<581E&}+TS5Z^q_Zl9k{Ce5?T9giA%q)CZ;!zM+S(Et>nq9n
z$umi3c_rC6doFxD6$41zPHI4{3m*(CJ8QJLvMzii?%cU#{nVLca&|h|5+iFz3?9Xk
zQ^M27=9A@nmt;wPTMVb+mgSXb{As0ujpe=D%6)(77nW+SmTgr)_XFVRBTSYI?|KGy
zHRSF(6>iC@0ULSp8Q0E_oOmP{HIUOlleVZMeRSS&MRrryYi2dYI?v(q`w-9b?8fXx
z4GgV;qf^27fRKNG#5B<r5@TSqUFKW7`S&cHlVBJQOo-7jJ~5e0PfsLkr`8hY>5fi|
zQFDAc*<9TayGM)=S>4%^6{Kx3n79Z5<7Z4(24q@Kd|BP<h|wfQ5DsH#OjdlR7LN)G
zx$r<{{>DUZg_qT*BLcrZHk)jSp`$Z{GQj3Chlzy+HMq9cHj=3YS%DJ64aW_38)sHT
zh6SZv^yO*=j6b`kfeSB?vfLx*>-8S7+!r>BHCM~-i%<uc>84E7B%@nAZ<N!^PJ~;s
z8Uc?G=$LEgM@Tl3I;0vHcJ;C@+fCY{h1vdHVRjR39evXOKNL}d4&pJlu>b%707*qo
IM6N<$f*8B%#{d8T

literal 0
HcmV?d00001


From c1aaa03d89624f1533aa85a716f0b34f998b8fd0 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 24 Mar 2025 14:32:50 +0000
Subject: [PATCH 10/33] add fmt lib

---
 CMakeLists.txt    | 19 +++++++++++++++++++
 install.cmake.bat | 13 +++++++++++++
 install.cmake.sh  | 11 +++++++++++
 3 files changed, 43 insertions(+)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 3c27166..0109195 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -76,11 +76,27 @@ include_directories(
     subprojects/glfw/include
     subprojects/libdeflate
     subprojects/glm
+    subprojects/fmt/include
 )
 
 add_subdirectory(subprojects/glfw)
 add_subdirectory(subprojects/glm)
 
+find_library(
+    lib_fmt
+    NAMES fmt libfmt
+    PATHS 
+        subprojects/fmt/build 
+        subprojects/fmt/build/debug 
+        subprojects/fmt/build/Release
+        subprojects/fmt/build/release)
+if(NOT lib_fmt)
+    message(FATAL_ERROR "lib_fmt not found")
+else()
+    message(STATUS "lib_fmt found: ${lib_fmt}")
+endif()
+
+
 find_library(
     lib_deflate_static 
     NAMES deflate deflatestatic
@@ -96,6 +112,8 @@ else()
 endif()
 
 
+
+
 # set(CMAKE_PREFIX_PATH "subprojects/libtorch/share/cmake/Torch")
 if (WITH_TORCH)
     message(STATUS "CMAKE_PREFIX_PATH: ${CMAKE_PREFIX_PATH}")
@@ -151,6 +169,7 @@ endif()
 target_link_libraries(
     ${target_name} PRIVATE 
     ${lib_deflate_static} 
+    ${lib_fmt}
     sortlib 
     glm 
     glad 
diff --git a/install.cmake.bat b/install.cmake.bat
index 63c7ea1..72f4906 100644
--- a/install.cmake.bat
+++ b/install.cmake.bat
@@ -29,6 +29,19 @@ REM ========= pull git repo =========
 git submodule update --init --recursive
 
 
+REM ========= fmt =========
+cd subprojects\fmt
+cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_OSX_ARCHITECTURES=%ARCH% -S . -B build
+cmake --build build --config Release --target fmt 
+if errorlevel 1 (
+    echo "CMake fmt failed."
+    goto :error
+) else (
+    echo "CMake fmt completed successfully."
+)
+cd ..\..\
+
+
 REM ========= deflate =========
 cd subprojects\libdeflate
 cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_OSX_ARCHITECTURES=%ARCH% -S . -B build
diff --git a/install.cmake.sh b/install.cmake.sh
index 14c7b9a..fedc78c 100755
--- a/install.cmake.sh
+++ b/install.cmake.sh
@@ -49,6 +49,17 @@ git submodule update --init --recursive
 # cd ico_design && cd ico_design &&  iconutil -c icns icon_v2.iconset && cd ..
 
 
+# ========= fmt =========
+cd subprojects/fmt
+cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_OSX_ARCHITECTURES=${ARCH} -S . -B build
+
+cmake --build build --target fmt --config Release || {
+    echo "fmt: compile failed!"
+    exit 1
+}
+cd ../../
+
+
 # ========= libdeflate =========
 cd subprojects/libdeflate
 cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_OSX_ARCHITECTURES=${ARCH} -S . -B build

From f134adcb08fea21dd03e78966ac69ea2a361de3f Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 24 Mar 2025 14:34:15 +0000
Subject: [PATCH 11/33] auto cut run within selected area and local sort with
 double ends fixed

---
 PretextView.cpp            | 501 ++++++++++++++++++++++++++-----------
 src/auto_curation_state.h  | 249 +++++++++++++++++-
 src/copy_texture.cpp       |  29 ++-
 src/copy_texture.h         |  71 ++++--
 src/frag_cut_calculation.h |  75 +++---
 src/frag_sort.cpp          | 378 ++++++++++++++++++++--------
 src/frag_sort.h            |   2 +
 src/genomeData.h           |  18 +-
 src/utilsPretextView.h     |   1 +
 9 files changed, 993 insertions(+), 331 deletions(-)

diff --git a/PretextView.cpp b/PretextView.cpp
index b9f8980..978f593 100644
--- a/PretextView.cpp
+++ b/PretextView.cpp
@@ -659,11 +659,11 @@ Loading = 0;
 
 global_variable
 u32
-Yahs_sorting = 0;
+auto_sort_state = 0;
 
 global_variable
 u32 
-Auto_cutting = 0;
+auto_cut_state = 0;
 
 global_variable s32 auto_cut_button = 0; 
 global_variable s32 auto_sort_button = 0;
@@ -1635,7 +1635,7 @@ global_function
 void
 MouseMove(GLFWwindow* window, f64 x, f64 y)
 {
-    if (Loading || Yahs_sorting)
+    if (Loading || auto_sort_state || auto_cut_state)
     {
         return;
     }
@@ -1647,7 +1647,7 @@ MouseMove(GLFWwindow* window, f64 x, f64 y)
     if (UI_On)
     {
     }
-    else
+    else // ui_on == false
     {
         if (Edit_Mode)
         {
@@ -1672,24 +1672,51 @@ MouseMove(GLFWwindow* window, f64 x, f64 y)
 
             u32 pixel1 = (u32)((f64)nPixels * (0.5 + (f64)wx));
             u32 pixel2 = (u32)((f64)nPixels * (0.5 + (f64)wy));
+            
+            // 防止在尾部溢出
+            pixel1 = my_Max(0, my_Min(nPixels - 1, pixel1));
+            pixel2 = my_Max(0, my_Min(nPixels - 1, pixel2));
 
-            u32 contig = Map_State->contigIds[pixel1];
+            u32 contig = Map_State->contigIds[pixel1]; 
 
             if (!Edit_Pixels.editing && !Edit_Pixels.selecting && Map_State->contigIds[pixel2] != contig)
             {
                 u32 testPixel = pixel1;
                 u32 testContig = contig;
-                while (testContig == contig)
-                {
+                while (testContig == contig) // move testContig to the 1 before the first or 1 after the last pixel of the contig
+                {   
+
+                    if (pixel1 < pixel2) 
+                    {
+                        testPixel ++ ;
+                    }
+                    else
+                    {
+                        testPixel -- ;
+                    }
+
+                    testContig = Map_State->contigIds[testPixel];
+
+                    if (testPixel == 0 || testPixel >= (nPixels - 1)) break;
+
                     testContig = pixel1 < pixel2 ? Map_State->contigIds[++testPixel] : Map_State->contigIds[--testPixel];
                     if (testPixel == 0 || testPixel >= (nPixels - 1)) break;
+                }   
+                if (Map_State->contigIds[testPixel] == contig)
+                {
+                    pixel2 = testPixel;
+                }
+                else
+                {
+                    pixel2 = pixel1 < pixel2 ? testPixel - 1 : testPixel + 1;
                 }
-                pixel2 = pixel1 < pixel2 ? testPixel - 1 : testPixel + 1;
             }
+
             
-            if (Edit_Pixels.selecting)
+            if (Edit_Pixels.selecting) // select the contig
             {
-                Edit_Pixels.selectPixels.x = my_Max(pixel1, my_Max(Edit_Pixels.selectPixels.x, Edit_Pixels.selectPixels.y));
+                Edit_Pixels.selectPixels.x = my_Max(pixel1, my_Max(Edit_Pixels.selectPixels.x, Edit_Pixels.selectPixels.y)); // used to select multiple contigs with pressing space and dragging the mouse
+                Edit_Pixels.selectPixels.x = pixel1;
                 while(  Edit_Pixels.selectPixels.x < (Number_of_Pixels_1D - 1) &&
                         ((Edit_Pixels.scaffSelecting && Map_State->scaffIds[Edit_Pixels.selectPixels.x] && Map_State->scaffIds[Edit_Pixels.selectPixels.x] == Map_State->scaffIds[1 + Edit_Pixels.selectPixels.x] ) || 
                           Map_State->contigIds[Edit_Pixels.selectPixels.x] == Map_State->contigIds[1 + Edit_Pixels.selectPixels.x]))
@@ -1698,6 +1725,7 @@ MouseMove(GLFWwindow* window, f64 x, f64 y)
                 }
 
                 Edit_Pixels.selectPixels.y = my_Min(pixel1, my_Min(Edit_Pixels.selectPixels.x, Edit_Pixels.selectPixels.y));
+                // Edit_Pixels.selectPixels.y = pixel1;
                 while(  Edit_Pixels.selectPixels.y > 0 &&
                         ((Edit_Pixels.scaffSelecting && Map_State->scaffIds[Edit_Pixels.selectPixels.y] && Map_State->scaffIds[Edit_Pixels.selectPixels.y] == Map_State->scaffIds[Edit_Pixels.selectPixels.y - 1]) || Map_State->contigIds[Edit_Pixels.selectPixels.y] == Map_State->contigIds[Edit_Pixels.selectPixels.y - 1]))
                 {
@@ -1741,7 +1769,7 @@ MouseMove(GLFWwindow* window, f64 x, f64 y)
                 Edit_Pixels.worldCoords.x = (f32)(((f64)((2 * Edit_Pixels.pixels.x) + 1)) / ((f64)(2 * nPixels))) - 0.5f;
                 Edit_Pixels.worldCoords.y = (f32)(((f64)((2 * Edit_Pixels.pixels.y) + 1)) / ((f64)(2 * nPixels))) - 0.5f;
             }
-            else
+            else  // edit_pixels.editing == 0
             {
                 if (netDelta || Global_Edit_Invert_Flag)
                 {
@@ -1957,7 +1985,7 @@ Mouse(GLFWwindow* window, s32 button, s32 action, s32 mods)
     s32 primaryMouse = Mouse_Invert ? GLFW_MOUSE_BUTTON_RIGHT : GLFW_MOUSE_BUTTON_LEFT;
     s32 secondaryMouse = Mouse_Invert ? GLFW_MOUSE_BUTTON_LEFT : GLFW_MOUSE_BUTTON_RIGHT;    
     
-    if (Loading || Yahs_sorting)
+    if (Loading || auto_sort_state)
     {
         return;
     }
@@ -2087,7 +2115,7 @@ global_function
 void
 Scroll(GLFWwindow* window, f64 x, f64 y)
 {
-    if (Loading || Yahs_sorting)
+    if (Loading || auto_sort_state)
     {
         return;
     }
@@ -2741,7 +2769,7 @@ Render() {
     }
 
     // Text / UI Rendering
-    if (Contig_Name_Labels->on || Scale_Bars->on || Tool_Tip->on || UI_On || Loading || Yahs_sorting || Edit_Mode || Waypoint_Edit_Mode || Waypoints_Always_Visible || Scaff_Edit_Mode || Scaffs_Always_Visible || MetaData_Edit_Mode || MetaData_Always_Visible || graphNamesOn)
+    if (Contig_Name_Labels->on || Scale_Bars->on || Tool_Tip->on || UI_On || Loading || auto_sort_state || Edit_Mode || Waypoint_Edit_Mode || Waypoints_Always_Visible || Scaff_Edit_Mode || Scaffs_Always_Visible || MetaData_Edit_Mode || MetaData_Always_Visible || graphNamesOn)
     {
         f32 textNormalMat[16];
         f32 textRotMat[16];
@@ -2847,7 +2875,7 @@ Render() {
 
                 if (rightPixel > 0.0f && leftPixel < width)
                 {
-                    const char *name = (const char *)(Original_Contigs + cont->originalContigId)->name;
+                    const char *name = (const char *)(Original_Contigs + cont->get_original_contig_id())->name;
                     f32 x = (rightPixel + leftPixel) * 0.5f;
                     f32 y = my_Max(wy0, 0.0f) + 10.0f;
                     f32 textWidth = fonsTextBounds(FontStash_Context, x, y, name, 0, NULL);
@@ -2896,7 +2924,7 @@ Render() {
 
                 if (topPixel < height && bottomPixel > 0.0f)
                 {
-                    const char *name = (const char *)(Original_Contigs + cont->originalContigId)->name;
+                    const char *name = (const char *)(Original_Contigs + cont->get_original_contig_id())->name;
                     f32 y = (topPixel + bottomPixel) * 0.5f;
                     f32 x = my_Max(wx0, 0.0f) + 10.0f;
                     f32 textWidth = fonsTextBounds(FontStash_Context, x, y, name, 0, NULL);
@@ -3504,13 +3532,11 @@ Render() {
             f32 end_fraction   = (f32)auto_curation_state.get_end()  /(f32)Number_of_Pixels_1D;
 
             // selected frags into a string
-            std::vector<u32> selected_frag_ids;
-            auto_curation_state.get_selected_fragments(selected_frag_ids, Map_State, Number_of_Pixels_1D);
-            std::stringstream ss;
-            if (selected_frag_ids.size() > 0) ss << selected_frag_ids[0];
-            for (u32 i = 1; i < selected_frag_ids.size(); i++) ss <<", " << selected_frag_ids[i] ;
-            std::string selectedFragmentsStr = "Selected fragments number: " + std::to_string(selected_frag_ids.size());
-            std::string start_and_end_pixel_str = "Select range: " + std::to_string(auto_curation_state.get_start()) + " - " + std::to_string(auto_curation_state.get_end());
+            SelectArea select_area;
+            auto_curation_state.get_selected_fragments(select_area, Map_State, Number_of_Pixels_1D, Contigs);
+            std::string selectedFragmentsStr = fmt::format(
+                "Selected fragments number: {}", select_area.selected_frag_ids.size());
+            std::string start_and_end_pixel_str = fmt::format("Select range: {} - {}", auto_curation_state.get_start(), auto_curation_state.get_end());
 
             { // draw the help text in the bottom right corner
                 fonsSetFont(FontStash_Context, Font_Bold);
@@ -3574,7 +3600,7 @@ Render() {
                     // draw the start & end point
                     {   
                         f32 line_width = 0.002f / Camera_Position.z;
-                        f32 mask_color[4] = {1.0f, 0.f, 0.f, 0.8f};
+                        f32 mask_color[4] = {1.0f, 0.f, 0.f, 0.8f}; // red
                         glUseProgram(Flat_Shader->shaderProgram);
                         glUniform4fv(Flat_Shader->colorLocation, 1, (GLfloat *)&mask_color);
                         for (auto loc_fraction : {start_fraction, end_fraction})
@@ -3593,7 +3619,7 @@ Render() {
 
                     // draw the cover on selected area
                     {   
-                        f32 mask_color[4] = {0.906f, 0.03921f, 0.949f, 0.5f}; 
+                        // f32 mask_color[4] = {0.906f, 0.03921f, 0.949f, 0.5f}; 
                         vert[0].x = ModelXToScreen(start_fraction - 0.5f);     vert[0].y = ModelYToScreen(0.5f - start_fraction);
                         vert[1].x = ModelXToScreen(start_fraction - 0.5f);     vert[1].y = ModelYToScreen(0.5f - end_fraction);
                         vert[2].x = ModelXToScreen(end_fraction - 0.5f);       vert[2].y = ModelYToScreen(0.5f - end_fraction);
@@ -3603,7 +3629,7 @@ Render() {
                         u32 colour = FourFloatColorToU32(*((nk_colorf *)font_color));
 
                         glUseProgram(Flat_Shader->shaderProgram);
-                        glUniform4fv(Flat_Shader->colorLocation, 1, (GLfloat *)&mask_color);
+                        glUniform4fv(Flat_Shader->colorLocation, 1, (GLfloat *)&auto_curation_state.mask_color);
 
                         glBindBuffer(GL_ARRAY_BUFFER, Scaff_Bar_Data->vbos[ptr]);
                         glBufferSubData(GL_ARRAY_BUFFER, 0, 4 * sizeof(vertex), vert);
@@ -3612,15 +3638,19 @@ Render() {
 
                         glUseProgram(UI_Shader->shaderProgram);
                         fonsSetColor(FontStash_Context, colour);
+                        
+                        std::string buff = fmt::format(
+                            "{}: ({}) pixels, ({}) fragments", 
+                            auto_curation_state.selected_or_exclude==0?"Selected" :"Excluded", 
+                            std::abs(auto_curation_state.get_end() - auto_curation_state.get_start()), 
+                            select_area.selected_frag_ids.size());
 
-                        char buff[256];
                         f32 lh = 0.0f;
-                        stbsp_snprintf(buff, sizeof(buff), "%s (%d) Pixs: %s", auto_curation_state.selected_or_exclude==0?"Select" :"Exclude", std::abs(auto_curation_state.get_end() - auto_curation_state.get_start()), std::to_string(selected_frag_ids.size()).c_str());
-                        f32 textWidth = fonsTextBounds(FontStash_Context, 0, 0, buff, 0, NULL);
+                        f32 textWidth = fonsTextBounds(FontStash_Context, 0, 0, buff.c_str(), 0, NULL);
                         fonsDrawText(
                             FontStash_Context, 
                             ModelXToScreen( 0.5f * (end_fraction + start_fraction ) - 0.5f) - (0.5f * textWidth), 
-                            ModelYToScreen(0.5f - 0.5f*(start_fraction + end_fraction)) - lh * 0.5f, buff, 0);
+                            ModelYToScreen(0.5f - 0.5f*(start_fraction + end_fraction)) - lh * 0.5f, buff.c_str(), 0);
                     }
                 }
             }
@@ -4326,7 +4356,7 @@ Render() {
         }
 
 
-        if (Yahs_sorting)
+        if (auto_sort_state)
         {
             u32 colour = glfonsRGBA(Theme_Colour.r, Theme_Colour.g, Theme_Colour.b, Theme_Colour.a);
 
@@ -4344,7 +4374,7 @@ Render() {
             ChangeSize((s32)width, (s32)height);
         }
 
-        if (Auto_cutting)
+        if (auto_cut_state)
         {
             u32 colour = glfonsRGBA(Theme_Colour.r, Theme_Colour.g, Theme_Colour.b, Theme_Colour.a);
 
@@ -6059,20 +6089,26 @@ RearrangeMap(       // NOTE: VERY IMPORTANT
     bool update_contigs_flag // if true, update the contigs from the map state
     )       
 {
-    if (pixelFrom > pixelTo)  // Swap, make sure pixelFrom is less than pixelTo
-    {
-        u32 tmp = pixelFrom;
-        pixelFrom = pixelTo;
-        pixelTo = tmp;
-    }
 
     u32 nPixels = Number_of_Pixels_1D;
 
-    Assert((delta > 0 ? (u32)delta : (u32)(-delta)) < nPixels);
-
-    Assert(pixelFrom < nPixels);
-    Assert(pixelTo < nPixels);
+    if (std::abs(delta) >= nPixels ||
+        pixelFrom >= nPixels ||
+        pixelTo >= nPixels)
+    {
+        fmt::print(
+            stderr,
+            "RearrangeMap: Invalid parameters: delta = {}, pixelFrom = {}, pixelTo = {},  nPixels = {}, file: {}, line: {}\n",
+            delta, pixelFrom, pixelTo, nPixels, 
+            __FILE__, __LINE__
+        );
+        assert(0);
+    }
 
+    if (pixelFrom > pixelTo)  // Swap, make sure pixelFrom is less than pixelTo
+    {
+        std::swap(pixelFrom, pixelTo);
+    }
     u32 nPixelsInRange = pixelTo - pixelFrom + 1;
 
     pixelFrom += nPixels;
@@ -6269,42 +6305,55 @@ BreakMap(
 
     s32 ptr_left = (s32)loc, ptr_right = (s32)loc;
     u08 inversed = IsContigInverted(contig_id);
-    while (
+    while ( // 从loc向左遍历，找到第一个不满足条件的像素点索引
         --ptr_left >= 0 && 
         Map_State->contigIds[ptr_left] == contig_id && 
         (Map_State->contigRelCoords[ptr_left] ==  Map_State->contigRelCoords[ptr_left+1]+ (inversed ? +1 : -1)) ) {};
-    while (
+    while ( // 从loc向右遍历，找到第一个不满足条件的像素点索引
         ++ptr_right < Number_of_Pixels_1D && 
         Map_State->contigIds[ptr_right] == contig_id && 
         (Map_State->contigRelCoords[ptr_right - 1] ==  Map_State->contigRelCoords[ptr_right]+ (inversed ? +1 : -1)) ) {};
     
     if ((loc - ptr_left) < ignore_len || (ptr_right - loc) < ignore_len )  
     {   
-        char buff[256];
-        snprintf(buff, sizeof(buff), "Error: original_contig_id %d current_contig_id %d, pixel range: [%d, cut(%d), %d], left(%d), right(%d), smaller than ignore_len (%d) \n", original_contig_id>Max_Number_of_Contigs? original_contig_id%Max_Number_of_Contigs:original_contig_id, contig_id, ptr_left, loc, ptr_right, loc - ptr_left, ptr_right - loc, ignore_len);
-        std::cout << buff << std::endl;
+        fmt::print(
+            "Error: original_contig_id {} current_contig_id {}, pixel range: [{}, cut({}), {}], left({}), right({}), smaller than ignore_len ({})\n", 
+            original_contig_id%Max_Number_of_Contigs, 
+            contig_id, 
+            ptr_left, 
+            loc, 
+            ptr_right, 
+            loc - ptr_left, 
+            ptr_right - loc, 
+            ignore_len);
         return;
     } 
 
     ptr_left++; ptr_right--;
-    // if ((inversed && Map_State->contigRelCoords[ptr_right] != 0 ) || 
-    //     (!inversed && Map_State->contigRelCoords[ptr_left] != 0))
-    // {
-    //     char buff[256];
-    //     snprintf(buff, sizeof(buff), "Error: contig %d, pixel range: [%d, %d] %s inversed, but the start real coor is not 0 (%d instead)\n", contig_id, ptr_left, ptr_right, inversed ? "(is)" : "(isn\'t)", inversed ? Map_State->contigRelCoords[ptr_right] : Map_State->contigRelCoords[ptr_left]);
-    //     MY_CHECK(buff);
-    //     assert(0);
-    // }
 
-    // cut the contig by amending the relCoords..
+    // cut the contig by amending the original Contig Ids.
     for (u32 tmp = loc; tmp <= ptr_right; tmp++) // left side
-    {
-        Map_State->originalContigIds[tmp] += Max_Number_of_Contigs;
+    {   
+        if (Map_State->originalContigIds[tmp] > (std::numeric_limits<u32>::max() - Max_Number_of_Contigs))
+        {
+            fmt::print("Error: originalContigIds[{}] + {} = {} is greater than the max number of contigs ({}), file: {}, line: {}\n", tmp, 
+            Max_Number_of_Contigs, 
+            (u64)Map_State->originalContigIds[tmp] + Max_Number_of_Contigs, 
+            std::numeric_limits<u32>::max(), 
+            __FILE__, __LINE__);
+            assert(0);
+        }
+        Map_State->originalContigIds[tmp] += Max_Number_of_Contigs;  // NOTE: the number can not exceed the max number 2**32 - 1 = 4294967295. However, it is not easy to exceed the max number of contigs, as 4294967295 / 4096 = 1048575.9997
     }
 
-    char buff[256];
-    snprintf(buff, sizeof(buff), "Original contig_id (%d), current_id (%d), pixel range: [%d, %d] %s inversed, cut at %d\n", original_contig_id, contig_id, ptr_left, ptr_right, inversed ? "(is)" : "(isn\'t)", loc);
-    std::cout << buff;
+    fmt::print(
+        "Original contig_id ({}), current_id ({}), pixel range: [{}, {}] {} inversed, cut at {}\n", 
+        original_contig_id%Max_Number_of_Contigs, 
+        contig_id, 
+        ptr_left, 
+        ptr_right, 
+        inversed ? "(is)" : "(isn\'t)", 
+        loc );
 
     return ; 
 
@@ -6385,10 +6434,25 @@ void AutoCurationFromFragsOrder(
 {   
     u08 using_select_area = (select_area && select_area->select_flag)? 1 : 0;
     u32 num_frags = contigs_->numberOfContigs;
-    if (!using_select_area && num_frags != frags_order_->get_num_frags()) 
+    if (!using_select_area ) 
+    {   
+        if ( num_frags != frags_order_->get_num_frags())
+        {
+            fprintf(stderr, "Number of contigs(%d) and fragsOrder.num_frags(%d) do not match.\n", num_frags, frags_order_->get_num_frags());
+            assert(0);
+        }
+    }
+    else
     {
-        fprintf(stderr, "Number of contigs(%d) and fragsOrder.num_frags(%d) do not match.\n", num_frags, frags_order_->get_num_frags());
-        return;
+        if (select_area->get_to_sort_frags_num() != frags_order_->get_num_frags())
+        {
+            fmt::print(
+                stderr, 
+                "num_to_sort_contigs({}) != fragsOrder.num_frags({}), file:{}, line:{}\n", select_area->get_to_sort_frags_num(), 
+                frags_order_->get_num_frags(), 
+                __FILE__, __LINE__);
+            assert(0);
+        }
     }
 
     u32 num_autoCurated_edits=0;
@@ -6401,8 +6465,9 @@ void AutoCurationFromFragsOrder(
     {
         std::vector<s32> full_predicted_order(num_frags);
         std::iota(full_predicted_order.begin(), full_predicted_order.end(), 1);
-        for (u32 i=0; i< select_area->selected_frag_ids.size(); i++) 
-            full_predicted_order[select_area->selected_frag_ids[i]] = (predicted_order[i]>0?1:-1) * (select_area->selected_frag_ids[0] + std::abs(predicted_order[i]));
+        auto frags_id_to_sort = select_area->get_to_sort_frags_id(Contigs);
+        for (u32 i=0; i< select_area->get_to_sort_frags_num(); i++) 
+            full_predicted_order[frags_id_to_sort[i]] = (predicted_order[i]>0?1:-1) * (select_area->get_first_frag_id() + std::abs(predicted_order[i]));
         predicted_order = full_predicted_order;
     }
     for (s32 i = 0; i < num_frags; ++i) current_order[i] = {i+1, contigs_->contigs_arr[i].length}; // start from 1
@@ -6505,6 +6570,7 @@ void AutoCurationFromFragsOrder(
     return ;
 }
 
+
 global_function
 void
 Auto_cutting_func(
@@ -6513,7 +6579,20 @@ Auto_cutting_func(
     std::string file_save_name = std::string("/auto_curation_tmp")
 )
 {   
-    if (!currFileName || !Auto_cutting) return;
+    if (!currFileName || !auto_cut_state) return;
+
+    if (auto_cut_state == 2) // restore the cutted frags within the selected area
+    {   
+        u32 start = auto_curation_state.get_start();
+        u32 end = auto_curation_state.get_end();
+        if (start < 0 || end < 0) return ;
+        fprintf(stdout, "========================\n");
+        fmt::print(stdout, "[Auto cutting] restoring the cutted frags within the selected area: [{}, {}]\n", start, end);
+        Map_State->restore_cutted_contigs( start, end );
+        UpdateContigsFromMapState();
+        Redisplay = 1;
+        return ;
+    }
 
     fprintf(stdout, "========================\n");
     // copy textures to cpu
@@ -6526,9 +6605,21 @@ Auto_cutting_func(
     restore_settings_after_copy(original_control_points);
     
     // classic cut 
-    if (1)
-    {
-        FragCutCal frag_cut_cal(&texture_array_4_ai, Contigs, Number_of_Pixels_1D); 
+    {   
+        SelectArea selected_area;
+        auto_curation_state.get_selected_fragments(
+            selected_area, 
+            Map_State, 
+            Number_of_Pixels_1D, 
+            Contigs);
+        FragCutCal frag_cut_cal(
+            &texture_array_4_ai, 
+            Contigs, 
+            Number_of_Pixels_1D, 
+            &selected_area, 
+            auto_curation_state.auto_cut_diag_window_for_pixel_mean,
+            auto_curation_state.auto_cut_threshold, 
+            auto_curation_state.auto_cut_smallest_frag_size_in_pixel); 
         std::vector<u32> cut_locs_pixel = frag_cut_cal.get_cut_locs_pixel();
         #ifdef DEBUG
             std::cout << "Cut locs in pixel: [";
@@ -6537,19 +6628,21 @@ Auto_cutting_func(
         #endif
         
         // add the cut locs to the extensions
-        if (arena)
-        {   
-            u32 added_num = Extensions.get_num_extensions();
-            std::string graph_name = "hic_pixel_density";
-            u32* graph_data = new u32[Number_of_Pixels_1D];
-            for (u32 i=0; i < Number_of_Pixels_1D; i ++)
-            {
-                graph_data[i] = (u32)(frag_cut_cal.hic_density[i] * 100.f);
+        #ifdef DEBUG    
+            if (arena)
+            {   
+                u32 added_num = Extensions.get_num_extensions();
+                std::string graph_name = "hic_pixel_density";
+                u32* graph_data = new u32[Number_of_Pixels_1D];
+                for (u32 i=0; i < Number_of_Pixels_1D; i ++)
+                {
+                    graph_data[i] = (u32)(frag_cut_cal.hic_density[i] * 100.f);
+                }
+                add_graph_to_extensions(arena, (u32*)graph_name.c_str(), graph_data);
+                
+                push_extensions_to_opengl(arena, added_num);
             }
-            add_graph_to_extensions(arena, (u32*)graph_name.c_str(), graph_data);
-            
-            push_extensions_to_opengl(arena, added_num);
-        }
+        #endif // DEBUG
         
         // cut the contigs
         cut_frags(cut_locs_pixel);
@@ -6584,9 +6677,9 @@ Auto_cutting_func(
 
 global_function
 void
-Sort_auto_sort(char* currFileName)
+Auto_sort_func(char* currFileName)
 {   
-    if (!currFileName || !Yahs_sorting) return;
+    if (!currFileName || !auto_sort_state) return;
     
     fprintf(stdout, "========================\n");
     fprintf(stdout, "[Auto Sort] start...\n");
@@ -6604,7 +6697,12 @@ Sort_auto_sort(char* currFileName)
 
     // check if select the area for sorting
     SelectArea selected_area;
-    auto_curation_state.get_selected_fragments(selected_area.selected_frag_ids, Map_State, Number_of_Pixels_1D);
+    auto_curation_state.get_selected_fragments(
+        selected_area, 
+        Map_State, 
+        Number_of_Pixels_1D, 
+        Contigs
+    );
     if (auto_curation_state.get_start() >=0 && auto_curation_state.get_end() >= 0 && selected_area.selected_frag_ids.size() > 0)
     {   
         fprintf(stdout, "Using selected area for sorting:\n");
@@ -6630,8 +6728,8 @@ Sort_auto_sort(char* currFileName)
     std::vector<std::string> exclude_tags = {"haplotig", "unloc"};
     std::vector<s32> exclude_meta_tag_idx = get_exclude_metaData_idx(exclude_tags);
     LikelihoodTable likelihood_table(
-        texture_array_4_ai.get_frags(), 
-        texture_array_4_ai.get_compressed_hic(), 
+        texture_array_4_ai.get_frags(),
+        texture_array_4_ai.get_compressed_hic(),
         (f32)auto_curation_state.smallest_frag_size_in_pixel / ((f32)Number_of_Pixels_1D + 1.f), 
         exclude_meta_tag_idx, 
         Number_of_Pixels_1D);
@@ -6641,6 +6739,7 @@ Sort_auto_sort(char* currFileName)
         frag_sort_method->sort_according_likelihood_unionFind( 
             likelihood_table, 
             frags_order, 
+            selected_area,
             auto_curation_state.link_score_threshold, 
             texture_array_4_ai.get_frags());
     }
@@ -6649,6 +6748,7 @@ Sort_auto_sort(char* currFileName)
         frag_sort_method->sort_according_likelihood_unionFind_doFuse( 
             likelihood_table, 
             frags_order, 
+            selected_area,
             auto_curation_state.link_score_threshold, 
             texture_array_4_ai.get_frags(), true, true);
     }
@@ -6657,6 +6757,7 @@ Sort_auto_sort(char* currFileName)
         frag_sort_method->sort_according_likelihood_unionFind_doFuse( 
             likelihood_table, 
             frags_order, 
+            selected_area,
             auto_curation_state.link_score_threshold, 
             texture_array_4_ai.get_frags(), false, true);
     }
@@ -6671,7 +6772,7 @@ Sort_auto_sort(char* currFileName)
         Map_State, 
         &selected_area);
     std::cout << std::endl;
-    Yahs_sorting = 0;
+    auto_sort_state = 0;
 
     return;
 }
@@ -7101,7 +7202,7 @@ global_function
 void
 KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
 {
-    if (!Loading && !Yahs_sorting && !Auto_cutting && (action != GLFW_RELEASE || key == GLFW_KEY_SPACE || key == GLFW_KEY_A || key == GLFW_KEY_LEFT_SHIFT))
+    if (!Loading && !auto_sort_state && !auto_cut_state && (action != GLFW_RELEASE || key == GLFW_KEY_SPACE || key == GLFW_KEY_A || key == GLFW_KEY_LEFT_SHIFT))
     {
         if (UI_On)
         {
@@ -7234,23 +7335,31 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                     break;
 
                 case GLFW_KEY_C:
-                    if (!Extensions.head) break;
-                    TraverseLinkedList(Extensions.head, extension_node)
+                    if (Select_Sort_Area_Mode && auto_curation_state.get_selected_fragments_num(Map_State, Number_of_Pixels_1D) > 0) // local area break
                     {
-                        switch (node->type)
-                        {
-                        case extension_graph:
+                        auto_cut_state = 1;
+                    }
+                    else // open "coverage" extension
+                    {
+                        if (!Extensions.head) break;
+                        TraverseLinkedList(Extensions.head, extension_node)
                         {
-
-                            graph *gph = (graph *)node->extension;
-                            if (strcmp((char *)gph->name, "coverage") == 0)
+                            switch (node->type)
                             {
-                                gph->on = !gph->on;
-                                break;
+                            case extension_graph:
+                            {
+
+                                graph *gph = (graph *)node->extension;
+                                if (strcmp((char *)gph->name, "coverage") == 0)
+                                {
+                                    gph->on = !gph->on;
+                                    break;
+                                }
+                            }
                             }
-                        }
                         }
                     }
+                    
                     break;
 
                 case GLFW_KEY_D:
@@ -7268,6 +7377,7 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                     break;
 
                 case GLFW_KEY_F:
+                    keyPressed = ToggleSelectSortAreaMode(window);
                     break;
 
                 case GLFW_KEY_G:
@@ -7302,7 +7412,6 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                     break;
 
                 case GLFW_KEY_K:
-                    keyPressed = ToggleSelectSortAreaMode(window);
                     break;
 
                 case GLFW_KEY_L:
@@ -7382,7 +7491,7 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                     }
                     else if (Select_Sort_Area_Mode)
                     {
-                        auto_curation_state.clear();
+                        auto_curation_state.clear(); // clear the selected area
                     }
                     else
                     {
@@ -7424,6 +7533,10 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                     break;
 
                 case GLFW_KEY_V:
+                    if (Select_Sort_Area_Mode)
+                    {
+                        auto_cut_state = 2; // restore the cutted frags within the selected area
+                    }
                     break;
 
                 case GLFW_KEY_W:
@@ -7493,11 +7606,10 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                     }
                     else if (Select_Sort_Area_Mode && action == GLFW_PRESS)
                     {   
-                        std::vector<u32> frags_id;
-                        auto_curation_state.get_selected_fragments(frags_id, Map_State, Number_of_Pixels_1D);
-                        if (frags_id.size() >= 2)
+                        u32 selected_frag_num = auto_curation_state.get_selected_fragments_num(Map_State, Number_of_Pixels_1D);
+                        if (selected_frag_num >= 2)
                         {
-                            Yahs_sorting = 1;
+                            auto_sort_state = 1;
                         }
                     }
                     else
@@ -8608,6 +8720,28 @@ global_function
             nk_tree_pop(NK_Context);
         }
 
+
+        // select the color for the selected area
+        if (nk_tree_push(NK_Context, NK_TREE_TAB, "Colour Selected Area", NK_MINIMIZED))
+        {
+            struct nk_colorf colour_bg = {
+                auto_curation_state.mask_color[0], 
+                auto_curation_state.mask_color[1], 
+                auto_curation_state.mask_color[2], 
+                auto_curation_state.mask_color[3]};
+            nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 200.0f, 4);
+            colour_bg = nk_color_picker(NK_Context, colour_bg, NK_RGBA);
+            auto_curation_state.mask_color[0] = colour_bg.r;
+            auto_curation_state.mask_color[1] = colour_bg.g;
+            auto_curation_state.mask_color[2] = colour_bg.b;
+            auto_curation_state.mask_color[3] = colour_bg.a;
+            nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30, 4);
+            if (nk_button_label(NK_Context, "Default")) 
+            {
+                auto_curation_state.set_mask_color(auto_curation_state.mask_color_default);
+            }
+            nk_tree_pop(NK_Context);
+        }
         // nk_tree_pop(NK_Context); // used to have this here, but it was causing a crash
 
         nk_style_set_font(ctx, &NK_Font->handle);
@@ -8808,9 +8942,19 @@ global_variable
 u08
 SaveState_Magic_Tail_Manual = 3;
 
+/* 
+保存当前状态，
+    headerHash: 用于生成文件名的哈希值
+    path: 保存路径
+    overwrite: 是否覆盖
+    version: 版本号，用于控制生成文件的版本，兼容上一个版本
+*/
 global_function
 u08
-SaveState(u64 headerHash, char *path = 0, u08 overwrite = 0)
+SaveState(
+    u64 headerHash, char *path = 0, u08 overwrite = 0, 
+    u32 version = 0
+)
 {
     if (!path)
     {
@@ -8851,7 +8995,8 @@ SaveState(u64 headerHash, char *path = 0, u08 overwrite = 0)
                     break;
             }
         }
-
+        
+        // number of scaffs and metaFlags
         u32 nScaffs = 0;
         u32 nMetaFlags = 0;
         ForLoop(Contigs->numberOfContigs)
@@ -8859,7 +9004,7 @@ SaveState(u64 headerHash, char *path = 0, u08 overwrite = 0)
             if ((Contigs->contigs_arr + index)->scaffId) ++nScaffs;
             if (*(Contigs->contigs_arr + index)->metaDataFlags) ++nMetaFlags;
         }
-        
+        // number of meta tags
         u08 nMetaTags = 0;
         u32 totalMetaTagSpace = 0;
         ForLoop(ArrayCount(Meta_Data->tags))
@@ -9213,8 +9358,10 @@ SaveState(u64 headerHash, char *path = 0, u08 overwrite = 0)
 
         if (!path) // save the name SaveState_Name to file named as ptlsn 
         {
-            u08 nameCache[17];
-            CopyNullTerminatedString((u08 *)SaveState_Name, (u08 *)nameCache);
+            // u08 nameCache[17];
+            // CopyNullTerminatedString((u08 *)SaveState_Name, (u08 *)nameCache);
+
+            std::string nameCache_string = std::string((char *)SaveState_Name);
 
             *(SaveState_Name + 0) = 'p';
             *(SaveState_Name + 1) = 't';
@@ -9224,7 +9371,8 @@ SaveState(u64 headerHash, char *path = 0, u08 overwrite = 0)
             *(SaveState_Name + 5) = '\0';
 
             file = fopen((const char *)SaveState_Path, "wb");
-            fwrite((u08 *)nameCache, 1, sizeof(nameCache), file);
+            // fwrite((u08 *)nameCache, 1, sizeof(nameCache), file);
+            fwrite((u08 *)nameCache_string.c_str(), 1, nameCache_string.size(), file);
             fclose(file);
         }
     }
@@ -9884,7 +10032,7 @@ restore_initial_state()
     
     // restore all the splited contigs
     {
-        Map_State->link_splited_contigs(Number_of_Pixels_1D);
+        Map_State->restore_cutted_contigs_all(Number_of_Pixels_1D);
     }
 
 
@@ -10138,6 +10286,9 @@ UserSaveState(const char *headerHash = "userprofile", u08 overwrite = 1, char *p
     fwrite(&userColourMapOrder.nMaps, sizeof(userColourMapOrder.nMaps), 1, file);
     fwrite(userColourMapOrder.order, sizeof(u32), userColourMapOrder.nMaps, file);
 
+    // save selected area color 
+    fwrite(&auto_curation_state.mask_color, sizeof(auto_curation_state.mask_color), 1, file);
+
     fclose(file);
 
     printf("[UserProfile]: Saved to: %s\n", path);
@@ -10249,6 +10400,17 @@ global_function
     fread(&userColourMapOrder.nMaps, sizeof(userColourMapOrder.nMaps), 1, file);
     fread(userColourMapOrder.order, sizeof(u32), userColourMapOrder.nMaps, file);
 
+    // load selected area color 
+    f32 tmp_mask_color[4];
+    if (fread(tmp_mask_color, sizeof(tmp_mask_color), 1, file) == 1)
+    {
+        auto_curation_state.set_mask_color(tmp_mask_color);
+    }
+    else 
+    {
+        auto_curation_state.set_mask_color(auto_curation_state.mask_color_default);
+    }
+
     fclose(file);
     return 0; // Success
 }
@@ -10373,7 +10535,7 @@ GenerateAGP(char *path, u08 overwrite, u08 formatSingletons, u08 preserveOrder)
                 u64 contRealEndCoord = (u64)((f64)(startCoord + cont->length) / (f64)Number_of_Pixels_1D * (f64)Total_Genome_Length);
                 u64 contRealSize = contRealEndCoord - contRealStartCoord + 1;
 
-                char *contName = (char *)((Original_Contigs + cont->originalContigId)->name);
+                char *contName = (char *)((Original_Contigs + cont->get_original_contig_id())->name);
 
                 if (cont->scaffId && !type)
                 {
@@ -10705,19 +10867,19 @@ MainArgs {
             Redisplay = 1;
         }
 
-        if (Yahs_sorting)
+        if (auto_sort_state)
         {
             if (currFileName) SaveState(headerHash);
-            Sort_auto_sort((char*)currFileName);
-            Yahs_sorting = 0;
+            Auto_sort_func((char*)currFileName);
+            auto_sort_state = 0;
             Redisplay = 1;
         }
 
-        if (Auto_cutting)
+        if (auto_cut_state)
         {   
             if (currFileName) SaveState(headerHash);
             Auto_cutting_func((char*)currFileName, Loading_Arena);
-            Auto_cutting = 0;
+            auto_cut_state = 0;
             Redisplay = 1;
         }
 
@@ -10829,9 +10991,6 @@ MainArgs {
                     struct nk_color sort_button_active = nk_rgb(128, 0, 128);  
                     push_nk_style(NK_Context, sort_button_normal, sort_button_hover, sort_button_active);
                     nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 3);
-                    auto_cut_button = nk_button_label(NK_Context, "Auto cut");
-                    bounds = nk_widget_bounds(NK_Context);
-                    auto_sort_button = nk_button_label(NK_Context, "Auto Sort");
 
                     // AI sort button
                     /*
@@ -10877,24 +11036,84 @@ MainArgs {
                     */
                     // auto cut
                     {
+                        bounds = nk_widget_bounds(NK_Context);
+                        auto_cut_button = nk_button_label(NK_Context, "Auto cut");
+
                         if (auto_cut_button && currFileName)
                         {
-                            Auto_cutting = 1;
+                            auto_cut_state = 1;
+                            auto_curation_state.clear(); // if click the button, the sort/cut will be applied globally
+                        }
+                        // window to set the parameters for cut
+                        if (nk_contextual_begin(NK_Context, 0, nk_vec2(Screen_Scale.x * 480, Screen_Scale.y * 800), bounds))
+                        {   
+                            nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 1);
+
+                            nk_label(NK_Context, "Cut threshold (default: 0.05)", NK_TEXT_LEFT);
+                            nk_edit_string_zero_terminated(
+                                NK_Context, 
+                                NK_EDIT_FIELD, 
+                                (char*)auto_curation_state.auto_cut_threshold_buf, 
+                                sizeof(auto_curation_state.auto_cut_threshold_buf), 
+                                nk_filter_float);
+                            
+                            nk_label(NK_Context, "Pixel_mean window size (default: 8)", NK_TEXT_LEFT); //  见鬼了，见鬼了，这个地方，如果我使用 "Pixel_mean windows (default: 8)" 点击弹出的窗口外面程序就会卡死，但是使用"Pixel_mean window size (default: 8)" 就不会卡死
+                            nk_edit_string_zero_terminated(
+                                NK_Context, 
+                                NK_EDIT_FIELD, 
+                                (char*)auto_curation_state.auto_cut_diag_window_for_pixel_mean_buf, 
+                                sizeof(auto_curation_state.auto_cut_diag_window_for_pixel_mean_buf), 
+                                nk_filter_decimal);
+
+                            nk_label(NK_Context, "Smallest frag size (default: 8)", NK_TEXT_LEFT);
+                            nk_edit_string_zero_terminated(
+                                NK_Context, 
+                                NK_EDIT_FIELD, 
+                                (char*)auto_curation_state.auto_cut_smallest_frag_size_in_pixel_buf, 
+                                sizeof(auto_curation_state.auto_cut_smallest_frag_size_in_pixel_buf), 
+                                nk_filter_decimal);
+
+                            nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 2);
+                            if (nk_button_label(NK_Context, "Cancel")) 
+                            {   
+                                printf("[Auto Cut] Cancel button clicked\n");
+                                auto_curation_state.set_buf();
+                                nk_contextual_close(NK_Context);
+                            }
+                            if (nk_button_label(NK_Context, "Apply")) 
+                            {
+                                // Apply changes
+                                // Convert text to integer and float
+                                auto_curation_state.update_value_from_buf();
+                                
+                                nk_contextual_close(NK_Context);
+
+                                auto_curation_state.set_buf();
+                                fmt::print("[Auto Cut] cut_threshold:               {:.3f}\n", auto_curation_state.auto_cut_threshold);
+                                fmt::print("[Auto Cut] pixel mean window size:      {}\n", auto_curation_state.auto_cut_diag_window_for_pixel_mean);
+                                fmt::print("[Auto Cut] smallest_frag_size_in_pixel: {}\n", auto_curation_state.auto_cut_smallest_frag_size_in_pixel);
+                            }
+
+                            nk_contextual_end(NK_Context);
+
+                            
                         }
                     }
 
                     // Auto sort button
+                    bounds = nk_widget_bounds(NK_Context);
+                    auto_sort_button = nk_button_label(NK_Context, "Auto Sort");
                     {   
-                        if (auto_sort_button) 
+                        if (auto_sort_button && currFileName) 
                         {
-                            Yahs_sorting = 1;
-                            auto_curation_state.clear();
+                            auto_sort_state = 1;
+                            auto_curation_state.clear(); // if click the button, the sort/cut will be applied globally
                         }
                         // window to set the parameters for sort
                         if (nk_contextual_begin(NK_Context, 0, nk_vec2(Screen_Scale.x * 480, Screen_Scale.y * 400), bounds))
                         {
                             nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 1);
-                            nk_label(NK_Context, "Smallest Frag Size (pixels):", NK_TEXT_LEFT);
+                            nk_label(NK_Context, "Smallest Frag Size (default: 2)", NK_TEXT_LEFT);
                             nk_edit_string_zero_terminated(
                                 NK_Context, 
                                 NK_EDIT_FIELD, 
@@ -10902,7 +11121,7 @@ MainArgs {
                                 sizeof(auto_curation_state.frag_size_buf), 
                                 nk_filter_decimal);
 
-                            nk_label(NK_Context, "Link Score Threshold:", NK_TEXT_LEFT);
+                            nk_label(NK_Context, "Link score threshold (default: 0.4)", NK_TEXT_LEFT);
                             nk_edit_string_zero_terminated(
                                 NK_Context, 
                                 NK_EDIT_FIELD, 
@@ -10930,24 +11149,21 @@ MainArgs {
                             if (nk_button_label(NK_Context, "Cancel")) 
                             {   
                                 printf("[Auto Sort] Cancel button clicked\n");
+                                auto_curation_state.set_buf();
                                 nk_contextual_close(NK_Context);
                             }
                             if (nk_button_label(NK_Context, "Apply")) 
                             {
                                 // Apply changes
                                 // Convert text to integer and float
-                                auto_curation_state.smallest_frag_size_in_pixel = (u32)atoi((char*)auto_curation_state.frag_size_buf);
-                                auto_curation_state.link_score_threshold = (float)atof((char*)auto_curation_state.score_threshold_buf);
+                                auto_curation_state.update_value_from_buf();
+
                                 nk_contextual_close(NK_Context);
-                                if (auto_curation_state.link_score_threshold > 1.0f || auto_curation_state.link_score_threshold < 0.0f) 
-                                {   
-                                    printf("[Auto Sort] Warning: link Score Threshold should be in the range of [0, 1]\n");
-                                    auto_curation_state.link_score_threshold = std::max(0.0f, std::min(1.0f, auto_curation_state.link_score_threshold));
-                                }
+
                                 auto_curation_state.set_buf();
-                                printf("[Auto Sort] smallest_frag_size_in_pixel: %d\n", auto_curation_state.smallest_frag_size_in_pixel);
-                                printf("[Auto Sort] link_score_threshold:        %.3f\n", auto_curation_state.link_score_threshold);
-                                printf("[Auto Sort] Sort mode:                   %s\n", auto_curation_state.get_sort_mode_name().c_str());
+                                fmt::print("[Auto Sort] smallest_frag_size_in_pixel: {}\n", auto_curation_state.smallest_frag_size_in_pixel);
+                                fmt::print("[Auto Sort] link_score_threshold:        {:.3f}\n", auto_curation_state.link_score_threshold);
+                                fmt::print("[Auto Sort] Sort mode:                   {}\n", auto_curation_state.get_sort_mode_name());
                             }
                             // redo all the changes
                             if (!auto_curation_state.show_autoSort_redo_confirm_popup)
@@ -11000,7 +11216,7 @@ MainArgs {
                             nk_contextual_end(NK_Context);
                         }
                     }
-                    pop_nk_style(NK_Context, 3);
+                    pop_nk_style(NK_Context, 3); // pop the style for auto cut and auto sort button
 
                     // waypoint edit mode
                     nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 3);
@@ -11600,9 +11816,10 @@ MainArgs {
                         }
 
                         // Input Sequences
-                        {
+                        {   
+                            std::string input_sequence_name = fmt::format("Input Sequences ({})", Number_of_Original_Contigs);
                             nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 1);
-                            if (nk_tree_push(NK_Context, NK_TREE_TAB, "Input Sequences", NK_MINIMIZED))
+                            if (nk_tree_push(NK_Context, NK_TREE_TAB, input_sequence_name.c_str(), NK_MINIMIZED))
                             {   
                                 nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 3);
                                 nk_label(NK_Context, "Search: ", NK_TEXT_LEFT);
@@ -11697,7 +11914,7 @@ MainArgs {
                                                     u32 startCoord = cont->startCoord;
                                                     u32 endCoord = IsContigInverted(index2) ? (startCoord - cont->length + 1) : (startCoord + cont->length);
 
-                                                    stbsp_snprintf((char *)buff, sizeof(buff), "%s [%$" PRIu64 "bp to %$" PRIu64 "bp]", (char *)((Original_Contigs + cont->originalContigId)->name), (u64)((f64)startCoord / (f64)Number_of_Pixels_1D * (f64)Total_Genome_Length), (u64)((f64)(endCoord) / (f64)Number_of_Pixels_1D * (f64)Total_Genome_Length));
+                                                    stbsp_snprintf((char *)buff, sizeof(buff), "%s [%$" PRIu64 "bp to %$" PRIu64 "bp]", (char *)((Original_Contigs + cont->get_original_contig_id())->name), (u64)((f64)startCoord / (f64)Number_of_Pixels_1D * (f64)Total_Genome_Length), (u64)((f64)(endCoord) / (f64)Number_of_Pixels_1D * (f64)Total_Genome_Length));
                                                     if (nk_button_label(NK_Context, (char *)buff))
                                                     {
                                                         f32 p = pos + (0.5f * contLen);
diff --git a/src/auto_curation_state.h b/src/auto_curation_state.h
index 074aca3..f5162f7 100644
--- a/src/auto_curation_state.h
+++ b/src/auto_curation_state.h
@@ -16,7 +16,78 @@ struct SelectArea
     u08 select_flag = 0; // no selected area set
     u32 start_pixel = 0;
     u32 end_pixel = 0;
+    s32 source_frag_id = -1;
+    s32 sink_frag_id = -1;
     std::vector<u32> selected_frag_ids;
+
+
+    void clearup()
+    {
+        this->select_flag = 0;
+        this->start_pixel = 0;
+        this->end_pixel = 0;
+        this->source_frag_id = -1;
+        this->sink_frag_id = -1;
+        this->selected_frag_ids.clear();
+    }
+
+    /*
+    first frag id, including the source fragment
+    */
+    u32 get_first_frag_id()
+    {
+        return this->source_frag_id >=0? this->source_frag_id : this->selected_frag_ids[0];
+    }
+
+    /*
+    len in pixel, including the source and sink fragments
+    */
+    u32 get_selected_len(const contigs* Contigs) const
+    {
+        u32 len = 0;
+        if (this->source_frag_id>=0)
+        {
+            len += Contigs->contigs_arr[this->source_frag_id].length;
+        }
+        for (const auto& frag_id : this->selected_frag_ids)
+        {
+            len += Contigs->contigs_arr[frag_id].length;
+        }
+        if (this->sink_frag_id>=0)
+        {
+            len += Contigs->contigs_arr[this->sink_frag_id].length;
+        }
+        return len;
+    }
+
+    /*
+    number of fragments to be sorted, including the source and sink fragments
+    */
+    u32 get_to_sort_frags_num()
+    {
+        return this->selected_frag_ids.size() + (this->source_frag_id>=0) + (this->sink_frag_id>=0);
+    }
+
+    /*
+    id of fragments to be sorted, including the source and sink fragments
+    */
+    std::vector<u32> get_to_sort_frags_id(const contigs* Contigs)
+    {
+        std::vector<u32> to_sort_frags;
+        if (this->source_frag_id>=0)
+        {
+            to_sort_frags.push_back(this->source_frag_id);
+        }
+        for (const auto& frag_id : this->selected_frag_ids)
+        {
+            to_sort_frags.push_back(frag_id);
+        }
+        if (this->sink_frag_id>=0)
+        {
+            to_sort_frags.push_back(this->sink_frag_id);
+        }
+        return to_sort_frags;
+    }
 };
 
 
@@ -28,32 +99,89 @@ struct AutoCurationState
     s32 end_pixel = -1;   // select contigs for sort
 
 public:
+    f32 mask_color[4] = {0.906f, 0.03921f, 0.949f, 0.2f}; // used to show the selected area.
+    f32 mask_color_default[4] = {0.906f, 0.03921f, 0.949f, 0.2f}; // default color
     u08 selected_or_exclude = 0; // 0 for select, 1 for exclude
     u08 select_mode = 0; // 0: deactive, 1: set mode, 2: delete mode from start, 3: delete mode from end
     // Variables to store the current values
     u32 smallest_frag_size_in_pixel = 2;
     f32 link_score_threshold = 0.4f;
 
+    f32 auto_cut_threshold = 0.05f;
+    u32 auto_cut_diag_window_for_pixel_mean= 8;
+    u32 auto_cut_smallest_frag_size_in_pixel = 8;
+
     // Variables for the editing UI state
     bool show_autoSort_erase_confirm_popup = false;
     bool show_autoSort_redo_confirm_popup = false;
     //sorting mode
     u32 sort_mode = 0; // 0: union find, 1: fuse union find, 2 deep fuse
     std::vector<std::string> sort_mode_names = {"Union Find", "Fuse Union Find", "Deep Fuse"};
-    u08 frag_size_buf[16];
-    u08 score_threshold_buf[16];
+    
+    // auto sort
+    char frag_size_buf[16];
+    char score_threshold_buf[16];
+
+    // auto cut
+    char auto_cut_threshold_buf[16];
+    char auto_cut_diag_window_for_pixel_mean_buf[16];
+    char auto_cut_smallest_frag_size_in_pixel_buf[16];
+
     AutoCurationState()
     {
         set_buf();
     }
-
+    
     void set_buf()
-    {
-        memset(frag_size_buf, 0, 16);
-        memset(score_threshold_buf, 0, 16);
-        snprintf((char*)frag_size_buf, sizeof(frag_size_buf), "%u", smallest_frag_size_in_pixel);
+    {   
+        // sort: smallest_frag_size
+        memset(this->frag_size_buf, 0, sizeof(frag_size_buf));
+        snprintf((char*)frag_size_buf, sizeof(frag_size_buf), "%u", this->smallest_frag_size_in_pixel);
+        // sort: link_score_threshold
         memset(score_threshold_buf, 0, sizeof(score_threshold_buf));
-        snprintf((char*)score_threshold_buf, sizeof(score_threshold_buf), "%.3f", link_score_threshold);
+        snprintf((char*)this->score_threshold_buf, sizeof(this->score_threshold_buf), "%.3f", this->link_score_threshold);
+
+        // // cut: auto_cut_threshold
+        // memset(this->auto_cut_threshold_buf, 0, sizeof(this->auto_cut_threshold_buf));
+        // snprintf((char*)this->auto_cut_threshold_buf, sizeof(this->auto_cut_threshold_buf), "%.3f", this->auto_cut_threshold);
+        // // cut: auto_cut_diag_window_for_pixel_mean
+        // memset(this->auto_cut_diag_window_for_pixel_mean_buf, 0, sizeof(this->auto_cut_diag_window_for_pixel_mean_buf));
+        // snprintf((char*)this->auto_cut_diag_window_for_pixel_mean_buf, sizeof(this->auto_cut_diag_window_for_pixel_mean_buf), "%u", this->auto_cut_diag_window_for_pixel_mean);
+        // // cut: auto_cut_smallest_frag_size_in_pixel
+        // memset(this->auto_cut_smallest_frag_size_in_pixel_buf, 0, sizeof(this->auto_cut_smallest_frag_size_in_pixel_buf));
+        // snprintf((char*)this->auto_cut_smallest_frag_size_in_pixel_buf, sizeof(this->auto_cut_smallest_frag_size_in_pixel_buf), "%u", this->auto_cut_smallest_frag_size_in_pixel);
+
+
+        fmt::format_to((char*)auto_cut_threshold_buf, "{:.3f}", this->auto_cut_threshold);
+        fmt::format_to((char*)auto_cut_diag_window_for_pixel_mean_buf, "{}", this->auto_cut_diag_window_for_pixel_mean);
+        fmt::format_to((char*)auto_cut_smallest_frag_size_in_pixel_buf, "{}", this->auto_cut_smallest_frag_size_in_pixel);
+    }
+
+    void set_mask_color(f32* color)
+    {
+        for (u32 i = 0; i < 4; i++)
+        {
+            this->mask_color[i] = color[i];
+        }
+    }
+
+    void update_value_from_buf()
+    {   
+        // sort 
+        this->smallest_frag_size_in_pixel = (u32)atoi((char*)this->frag_size_buf);
+        this->link_score_threshold = (float)atof((char*)this->score_threshold_buf);
+        if (this->link_score_threshold > 1.0f || this->link_score_threshold < 0.0f) 
+        {   
+            printf("[Auto Sort] Warning: link Score Threshold should be in the range of [0, 1]\n");
+            this->link_score_threshold = std::max(0.0f, std::min(1.0f, this->link_score_threshold));
+        }
+
+        // cut
+        this->auto_cut_threshold = (f32)atof((char*)this->auto_cut_threshold_buf);
+        this->auto_cut_diag_window_for_pixel_mean = (u32)atoi((char*)this->auto_cut_diag_window_for_pixel_mean_buf);
+        this->auto_cut_smallest_frag_size_in_pixel = (u32)atoi((char*)this->auto_cut_smallest_frag_size_in_pixel_buf);
+
+        this->set_buf();
     }
 
     std::string get_sort_mode_name() const
@@ -89,7 +217,67 @@ struct AutoCurationState
         return this->end_pixel;
     }
 
-    void get_selected_fragments(std::vector<u32>& frag_ids, map_state* Map_State, u32 number_of_pixels_1D, u08 exclude_flag=false)
+    u32 get_selected_fragments_num(
+        map_state* Map_State,
+        u32 number_of_pixels_1D
+    ) 
+    {
+        if (this->start_pixel > number_of_pixels_1D || this->end_pixel > number_of_pixels_1D || this->start_pixel < 0 || this->end_pixel < 0)
+        {
+            return 0;
+        }
+        if (this->start_pixel > this->end_pixel)
+        {
+            std::swap(this->start_pixel, this->end_pixel);
+        }
+
+        u32 num_selected_frags = 1;
+        for (u32 i = this->start_pixel+1; i <= this->end_pixel; i++)
+        {
+            if (Map_State->contigIds[i] != Map_State->contigIds[i-1])
+            {
+                num_selected_frags ++;
+            }
+        }
+        return num_selected_frags;
+    }
+
+    std::vector<u32> get_selected_fragments_id(
+        map_state* Map_State,
+        u32 number_of_pixels_1D
+    )
+    {
+        std::vector<u32> selected_frag_ids={};
+        if (this->start_pixel > number_of_pixels_1D || this->end_pixel > number_of_pixels_1D || this->start_pixel < 0 || this->end_pixel < 0)
+        {
+            return selected_frag_ids;
+        }
+        if (this->start_pixel > this->end_pixel)
+        {
+            std::swap(this->start_pixel, this->end_pixel);
+        }
+
+        u32 last_contig_id = Map_State->contigIds[this->start_pixel];
+        selected_frag_ids.push_back(last_contig_id);
+        for (u32 i = this->start_pixel+1; i <= this->end_pixel; i++)
+        {
+            if (last_contig_id == Map_State->contigIds[i]) continue;
+            else
+            {
+                last_contig_id = Map_State->contigIds[i];
+                selected_frag_ids.push_back(last_contig_id);
+            }
+        }
+        return selected_frag_ids;
+    }
+
+
+    void get_selected_fragments(
+        SelectArea& select_area, 
+        map_state* Map_State, 
+        u32 number_of_pixels_1D, 
+        contigs* Contigs,
+        u08 exclude_flag=false)
     {   
         if (this->start_pixel > number_of_pixels_1D || this->end_pixel > number_of_pixels_1D || this->start_pixel < 0 || this->end_pixel < 0)
         {   
@@ -103,22 +291,57 @@ struct AutoCurationState
             std::swap(this->start_pixel, this->end_pixel);
         }
 
-        if (!frag_ids.empty() )
+        if (!select_area.selected_frag_ids.empty() )
         {
-            frag_ids.clear();
+            select_area.clearup();
         }
         
+        // define source frag
+        if (this->start_pixel>0)
+        {   
+            u32 frag_id = Map_State->contigIds[this->start_pixel-1];
+            if (Contigs->contigs_arr[frag_id].length >= this->smallest_frag_size_in_pixel)
+            {
+                select_area.source_frag_id = frag_id;
+            }
+            else
+            {   
+                select_area.source_frag_id = -1;
+                fmt::print("The source_frag_len ({}) < smallest_length ({}), not set the source frag.\n", Contigs->contigs_arr[frag_id].length, this->smallest_frag_size_in_pixel);
+            }
+        }
+        
+        // push the selected frag id into select_area
+        select_area.start_pixel = this->start_pixel;
+        select_area.end_pixel = this->end_pixel;
         u32 last_contig_id = Map_State->contigIds[this->start_pixel];
-        frag_ids.push_back(last_contig_id);
+        select_area.selected_frag_ids.push_back(last_contig_id);
         for (u32 i = start_pixel+1; i <= this->end_pixel; i ++ )
         {
             if (last_contig_id == Map_State->contigIds[i]) continue;
             else
             {   
                 last_contig_id = Map_State->contigIds[i];
-                frag_ids.push_back(last_contig_id);
+                select_area.selected_frag_ids.push_back(last_contig_id);
+            }
+        }
+
+        // define sink frag
+        if (this->end_pixel + 1 <= number_of_pixels_1D - 1)
+        {   
+            u32 frag_id = Map_State->contigIds[this->end_pixel+1];
+            if (Contigs->contigs_arr[frag_id].length >= this->smallest_frag_size_in_pixel)
+            {
+                select_area.sink_frag_id = frag_id;
+            }
+            else
+            {   
+                select_area.sink_frag_id = -1;
+                fmt::print("The tail_frag_len ({}) < smallest_length ({}), not set the tail frag.\n", Contigs->contigs_arr[frag_id].length, this->smallest_frag_size_in_pixel);
             }
         }
+        // set the select_area to valid
+        if (!select_area.selected_frag_ids.empty()) select_area.select_flag = 1;
     }
 
     void update_sort_area(
diff --git a/src/copy_texture.cpp b/src/copy_texture.cpp
index a302bb4..c41e787 100644
--- a/src/copy_texture.cpp
+++ b/src/copy_texture.cpp
@@ -23,7 +23,7 @@ SOFTWARE.
 
 #include "copy_texture.h"
 #include "shaderSource.h"
-
+#include "utilsPretextView.h"
 
 
 Show_State::Show_State()
@@ -263,8 +263,17 @@ TexturesArray4AI::~TexturesArray4AI()
 u08 TexturesArray4AI::operator()(u32 row, u32 column) const
 {   
     if (row>=num_pixels_1d || column>=num_pixels_1d) 
-    {
-        printf("Index [%d, %d] is out of range [%d, %d]\n", row, column, num_pixels_1d - 1, num_pixels_1d - 1);
+    {   
+        fmt::print(
+            stderr,
+            "Index [{}, {}] is out of range [{}, {}], File:{}, line:{}\n", 
+            row, 
+            column, 
+            num_pixels_1d - 1, 
+            num_pixels_1d - 1,
+            __FILE__,
+            __LINE__
+        );
         assert(0);
     }
     if (!this->textures) 
@@ -805,9 +814,17 @@ void TexturesArray4AI::cal_compressed_hic(
     u08 using_select_area = (select_area != nullptr && select_area->select_flag)?1:0;
     // clean the memory of compressed_hic_mx
     frags->re_allocate_mem(Contigs, select_area);
-    if (frags->total_length != (using_select_area? (select_area->end_pixel - select_area->start_pixel + 1): num_pixels_1d))
-    {
-        fprintf(stderr, "\n[Compress Hic] warning: frags->total_length(%d) != num_pixels_1d (%d) (%s).\n\n", frags->total_length, (using_select_area? (select_area->end_pixel - select_area->start_pixel + 1): num_pixels_1d), (using_select_area?"selected area":"full area"));
+    if (frags->total_length != (using_select_area ? select_area->get_selected_len(Contigs) : num_pixels_1d))
+    {   
+        fmt::print(
+            "\n[Compress Hic] warning: frags->total_length({}) != num_pixels_1d ({}) ({}). file:{}, line:{}\n\n",
+            frags->total_length,
+            (using_select_area? (select_area->end_pixel - select_area->start_pixel + 1): num_pixels_1d),
+            (using_select_area?"selected area":"full area"),
+            __FILE__,
+            __LINE__    
+        );
+        assert(0);
     }
     compressed_hic->re_allocate_mem(
         frags->num, 
diff --git a/src/copy_texture.h b/src/copy_texture.h
index 7fe1ae9..5fa79ac 100644
--- a/src/copy_texture.h
+++ b/src/copy_texture.h
@@ -53,47 +53,78 @@ struct Frag4compress {
 
     void re_allocate_mem(
         const contigs* Contigs, 
-        const SelectArea* select_area=nullptr)
+        const SelectArea* select_area=nullptr, 
+        bool use_for_cut_flag=false)
     {
         cleanup();
         u08 using_select_area = (select_area != nullptr && select_area->select_flag)?1:0;
-        num = Contigs->numberOfContigs;
-        if (using_select_area)
-        {
-            num = select_area->selected_frag_ids.size();
-            if (num < 2)
+        this->num = Contigs->numberOfContigs;
+        std::vector<u32> selected_frag_ids_tmp; 
+        if (using_select_area )
+        {   
+            selected_frag_ids_tmp = select_area->selected_frag_ids;
+            this->num = select_area->selected_frag_ids.size();
+            if (this->num < 2 && !use_for_cut_flag)
+            {
+                fmt::print(
+                    stderr, 
+                    "The number_of_select_fragments_for_sorting({}) should not be less than 2, file:{}, line:{}\n", this->num, __FILE__, __LINE__);
+                assert(0);
+            }
+            if (select_area->source_frag_id >=0 && !use_for_cut_flag) 
             {
-                fprintf(stderr, "The number_of_select_fragments_for_sorting(%d) should not be less than 2, file:%s (line:%d)\n", num, __FILE__, __LINE__);
-                std::abort();
+                this->num ++;
+                selected_frag_ids_tmp.insert(selected_frag_ids_tmp.begin(), select_area->source_frag_id);
             }
+            if (select_area->sink_frag_id >=0 && !use_for_cut_flag) 
+            {
+                this->num ++;
+                selected_frag_ids_tmp.push_back(select_area->sink_frag_id);
+            }
+        }
+        else // global area
+        {
+            selected_frag_ids_tmp.resize(this->num);
+            std::iota(selected_frag_ids_tmp.begin(), selected_frag_ids_tmp.end(), 0);
         }
+
         frag_id = new u32[num];
-        startCoord = new u32[num]; // TODO (shaoheng): it is wired that if I use std::vector<u32> here, the debugger will be stuck where I call the constructor. Sometimes it can work with std::vector, but sometimes it cannot.
+        startCoord = new u32[num]; 
         length = new u32[num];
         inversed = new bool[num];
         metaDataFlags = new u64[num];
         total_length = 0;
 
         // Initialize the startCoord, length, and inversed
+        frag_id[0] = selected_frag_ids_tmp[0];
         inversed[0] = false;
+        startCoord[0] = 0;
         if (using_select_area)
-        {   
-            frag_id[0] = select_area->selected_frag_ids[0];
-            startCoord[0] =select_area->start_pixel;
-            length[0] = Contigs->contigs_arr[select_area->selected_frag_ids[0]].length;
-        }
-        else
         {
-            frag_id[0] = 0;
-            startCoord[0] = 0;
-            length[0] = Contigs->contigs_arr[0].length;
+            s32 tmp = 0;
+            while (tmp < frag_id[0])
+            {
+                startCoord[0] += Contigs->contigs_arr[tmp].length;
+                tmp ++;
+            }
+            if (tmp>=Contigs->numberOfContigs)
+            {
+                fmt::print(
+                    stderr, 
+                    "The frag_id[0]({}) should be less than the number_of_contigs({}), file:{}, line:{}\n", 
+                    frag_id[0], Contigs->numberOfContigs, __FILE__, __LINE__);
+                assert(0);
+            }
         }
+        length[0] = Contigs->contigs_arr[frag_id[0]].length;
+        metaDataFlags[0] = (Contigs->contigs_arr[frag_id[0]].metaDataFlags == nullptr)?0:*(Contigs->contigs_arr[frag_id[0]].metaDataFlags);
         total_length = length[0];
+
         for (u32 i = 1; i < num; i++)
         {   
-            u32 contig_id = using_select_area?select_area->selected_frag_ids[i]:i;
+            u32 contig_id = selected_frag_ids_tmp[i];
             frag_id[i] = contig_id;
-            inversed[i] = false; // todo (shaoheng) add the inversed information
+            inversed[i] = false; // currently, this is not used
             startCoord[i] = startCoord[i-1] + length[i-1];
             length[i] = Contigs->contigs_arr[contig_id].length;
             metaDataFlags[i] = (Contigs->contigs_arr[contig_id].metaDataFlags == nullptr)?0:*(Contigs->contigs_arr[contig_id].metaDataFlags);
diff --git a/src/frag_cut_calculation.h b/src/frag_cut_calculation.h
index 4a30f7b..dcd0f09 100644
--- a/src/frag_cut_calculation.h
+++ b/src/frag_cut_calculation.h
@@ -34,10 +34,15 @@ class FragCutCal
 {
 public: 
     TexturesArray4AI *textures_array = nullptr;
-    const Frag4compress* frags = nullptr;
+    Frag4compress* frags = nullptr;
     contigs* Contigs = nullptr;
     std::string file_save_dir = std::filesystem::current_path().string() + "/hic_density_curve";
     std::vector<f32> hic_density={};
+    const SelectArea* select_area = nullptr;
+
+    const u32 pixel_mean_window_size;
+    const f32 cut_threshold;
+    const u32 smallest_frag_size;
 
     u32 D = 0;
     std::vector<f32> norm_diag_mean={};
@@ -46,8 +51,15 @@ class FragCutCal
     u32 maximum_D=5000;
     f32 min_hic_density=30.f;
 
-    FragCutCal(TexturesArray4AI* textures_array_, contigs* Contigs_, const u32& num_pixels_1d)
-        : textures_array(textures_array_), Contigs(Contigs_)
+    FragCutCal(
+        TexturesArray4AI* textures_array_, 
+        contigs* Contigs_, 
+        const u32& num_pixels_1d, 
+        const SelectArea* select_area_, 
+        u32 pixel_mean_window_size_, 
+        f32 cut_threshold_, 
+        u32 smallest_frag_size_)
+        : textures_array(textures_array_), Contigs(Contigs_), select_area(select_area_), pixel_mean_window_size(pixel_mean_window_size_), cut_threshold(cut_threshold_), smallest_frag_size(smallest_frag_size_)
     {   
         if (textures_array == nullptr)
         {
@@ -64,15 +76,17 @@ class FragCutCal
         #endif
 
         this->frags = textures_array->get_frags();
-        textures_array->cal_maximum_number_of_shift(
+        this->frags->re_allocate_mem(this->Contigs, this->select_area, true);
+        textures_array->cal_maximum_number_of_shift( // and get the norm_diag_mean
             this->D, 
             this->norm_diag_mean, 
             this->Contigs,
             this->D_hic_ratio,
             this->maximum_D,
             this->min_hic_density);
-        this->hic_density.resize(num_pixels_1d, 0.f);
         
+        // cal the hic_density
+        this->hic_density.resize(num_pixels_1d, 0.f);
         for (u32 frag_id = 0; frag_id < this->frags->num; frag_id++)
         {
             this->get_single_fragment_hic_density(
@@ -80,7 +94,6 @@ class FragCutCal
                 this->frags->length[frag_id],
                 this->hic_density.data() + this->frags->startCoord[frag_id]);
         }
-
     }
 
     std::vector<u32> get_cut_locs_pixel()
@@ -97,19 +110,18 @@ class FragCutCal
     void find_break_points(
         const u32 start_pixel,
         const std::vector<f32>& arr,
-        const std::vector<f32>& order1_derivative,
-        const std::vector<f32>& order2_derivative,
-        std::vector<u32>& break_points,
-        const u32 windows_size=8,  
-        const f32 threshold=0.05)
-    {
+        std::vector<u32>& break_points)
+    {   
+        const auto& windows_size = this->pixel_mean_window_size;
+        const auto& threshold = this->cut_threshold;
+
         int n = arr.size();
         if (n < 2 * windows_size + 1) 
         {
             return ; 
         }
 
-        if (1)  // detect the break points according to the value
+        // detect the break points according to the value
         {
             std::vector<f32> prefix_sum(n + 1, 0.f);
             cal_prefix_sum(arr, prefix_sum);
@@ -148,28 +160,6 @@ class FragCutCal
             }
         }
 
-        if (0) // detect break points according to derivatives
-        {   
-            u08 increase_flag = 1;
-            u32 len_count = 0;
-            for (u32 i = 1; i < n; i ++ )
-            {
-                if ((order1_derivative[i] > 0) == increase_flag)
-                {
-                    len_count++;
-                }
-                else
-                {
-                    if (len_count > windows_size * 2)
-                    {
-                        break_points.push_back(i + start_pixel);
-                    }
-                    len_count = 0;
-                    increase_flag = !increase_flag;
-                }
-            }
-        }
-
         return ;
     }
 
@@ -181,10 +171,11 @@ class FragCutCal
             this->hic_density.begin() + this->frags->startCoord[frag_id] + this->frags->length[frag_id]);
 
         u32 n = hic_density_tmp.size();
-        std::vector<f32> order1_derivative(n, 0.f);
-        std::vector<f32> order2_derivative(n, 0.f);
-        get_smoothed_1st_order_derivative(hic_density_tmp, order1_derivative);
-        get_smoothed_1st_order_derivative(order1_derivative, order2_derivative);
+
+        // std::vector<f32> order1_derivative(n, 0.f);
+        // std::vector<f32> order2_derivative(n, 0.f);
+        // get_smoothed_1st_order_derivative(hic_density_tmp, order1_derivative);
+        // get_smoothed_1st_order_derivative(order1_derivative, order2_derivative);
         
         #ifdef DEBUG // output the hic_density to text file
             std::string filename = this->file_save_dir +  "/current_id_" + std::to_string(frag_id) + ".txt";
@@ -192,7 +183,7 @@ class FragCutCal
             std::ofstream out(filename);
             for (u32 i = 0; i < hic_density_tmp.size(); i++)
             {
-                out << hic_density_tmp[i] << ", " << order1_derivative[i] << ", " << order2_derivative[i] << std::endl;
+                out << hic_density_tmp[i] << std::endl;
             }
             out.close();
         #endif
@@ -201,8 +192,6 @@ class FragCutCal
         this->find_break_points(
             this->frags->startCoord[frag_id],
             hic_density_tmp,
-            order1_derivative,
-            order2_derivative,
             break_points);
         if (break_points.size() > 0)
         {
@@ -246,7 +235,7 @@ class FragCutCal
         u32 num = 0, pixel_index=frag_start_pixel+relative_index;
         f32 sum = 0.f;
 
-        for (u32 i = 1; i < 8 ; i++) 
+        for (u32 i = 1; i < this->pixel_mean_window_size ; i++) 
         {   
             if (diag_flag)
             {   
diff --git a/src/frag_sort.cpp b/src/frag_sort.cpp
index 1328415..6a06937 100644
--- a/src/frag_sort.cpp
+++ b/src/frag_sort.cpp
@@ -218,52 +218,80 @@ void initilise_chromosomes(
     }
 }
 
+/*
+source and sink frags should not be sorted
+    return: 
+            0: not attachable, 
+            1: attachable
+*/
+bool check_source_sink_attachable(
+    const Link& lnk, 
+    u32 source_frag_id, u32 sink_frag_id, 
+    const u32& source_chain_id, const u32& sink_chain_id,
+    const u32& ca, const u32& cb
+)
+{   
+    if ((source_chain_id == ca && sink_chain_id == cb) || 
+        (sink_chain_id == ca && source_chain_id == cb)) return 0;
+    source_frag_id--; sink_frag_id -- ; // 在link中，从0开始编号，但是source_frag_id从1开始编号
 
-FragSortTool::FragSortTool() {}
-
-FragSortTool::~FragSortTool() {}
+    if ((lnk.frag_a == source_frag_id && lnk.frag_b==sink_frag_id)||
+        (lnk.frag_b == source_frag_id && lnk.frag_a==sink_frag_id))
+        return 0;
+    // source can not be linked to the head
+    if ( source_frag_id >= 0 && 
+        ((lnk.frag_a == source_frag_id && lnk.is_a_head_linked) || 
+        (lnk.frag_b == source_frag_id && lnk.is_b_head_linked))
+    ) 
+        return 0;
+    // sink can not be linked to the end
+    if ( sink_frag_id >= 0 && 
+        ((lnk.frag_a == sink_frag_id && !lnk.is_a_head_linked) || 
+        (lnk.frag_b == sink_frag_id && !lnk.is_b_head_linked))
+    ) 
+        return 0;
+    return 1;
+}
 
-void FragSortTool::sort_according_likelihood_unionFind(
-    const LikelihoodTable& likelihood_table, 
-    FragsOrder& frags_order, 
-    const f32 threshold, 
-    const Frag4compress* frags, 
-    bool sort_according_len_flag) const
+/*
+reverse the chain and return the new chain_id
+*/
+void reverse_chain(
+    std::deque<s32>& cha
+)
 {
-    u32 n = likelihood_table.get_num_frags();
-    std::vector<Link> all_links;
+    std::reverse(cha.begin(), cha.end());
+    for (auto &val: cha) val = -val; // flip orientation
+}
+void reverse_chain(
+    std::deque<s32>& cha,
+    u32& i
+)
+{
+    std::reverse(cha.begin(), cha.end());
+    for (auto &val: cha) val = -val; // flip orientation
+    i = cha.size() - 1 - i;
+}
 
-    std::vector<std::pair<bool, bool>> link_kinds = {
-        {true,  true}, {true,  false}, 
-        {false, true}, {false, false}};
-    for (u32 i = 0; i < n; i++) 
-    {
-        for (u32 j = i+1; j < n; j++)
-        {
-            for (u32 k = 0; k < 4; k++)
-            {   
-                f32 val = likelihood_table(i, j, k);
-                if (val >= threshold) {
-                    all_links.push_back(
-                        Link(val, i, j, link_kinds[k].first, link_kinds[k].second));
-                }
-            }
-        }
-    }
 
-    // Sort all links by descending score
-    std::sort(all_links.begin(), all_links.end(), [](const Link& a, const Link& b) {
-        return a.score > b.score;
-    });
-    // initialize the chromosomes and union-find
-    // every fragment is initilised as a not-inverted chain
-    std::vector<std::deque<s32>> chromosomes;
-    std::vector<s32> chain_id(n, -1);
-    std::vector<std::deque<s32>> chromosomes_excluded;
-    initilise_chromosomes(chromosomes, chromosomes_excluded, chain_id, n, likelihood_table);
 
+/* 
+union find operation
+*/
+void union_find_process(
+    const std::vector<Link>& all_links, 
+    std::vector<std::deque<s32>>& chromosomes, 
+    SelectArea& select_area, 
+    std::vector<s32> & chain_id, 
+    const s32& source_frag_id, 
+    const s32& sink_frag_id,
+    s32& source_chain_id,
+    s32& sink_chain_id
+)
+{
+    
     // Process links in order of decreasing score
-    for (auto &lnk : all_links)
+    for (const Link &lnk : all_links)
     {
         u32 a = lnk.frag_a;
         u32 b = lnk.frag_b;
@@ -277,15 +305,18 @@ void FragSortTool::sort_according_likelihood_unionFind(
 
         if (ca == cb) continue;
 
+
         // Decide how to connect them:
         auto &cha = chromosomes[ca];
         auto &chb = chromosomes[cb];
-        
-        // if the fragments are not the head or end of the chain, 
-        // thus they can not be linked
-        if (!is_head_or_end(cha, a) || !is_head_or_end(chb, b)) continue;
 
-        // find the index i/j of a/b on cha/b
+        // if a/b is the source, it can only be linked to the tail
+        // if a/b is the sink, it can only be linked to the head
+        // if a and b are source and sink or sink and source, they can not be linked
+        // source and link can not be linked into one chromosome
+        if (!check_source_sink_attachable(lnk, source_frag_id, sink_frag_id, source_chain_id, sink_chain_id, ca, cb)) continue;
+        
+        // find the index i/j of a/b on cha/chb
         u32 i = 0, j = 0;
         for (; i < cha.size(); i++)
             if (std::abs(cha[i]) == a+1)
@@ -302,15 +333,11 @@ void FragSortTool::sort_according_likelihood_unionFind(
         // Reverse a and/or b if necessary to make it always a->b
         if (lnk.is_a_head_linked != (cha[i] < 0))
         {
-            std::reverse(cha.begin(), cha.end());
-            for (auto &val: cha) val = -val; // flip orientation
-            i = cha.size() - 1 - i;
+            reverse_chain(cha, i);
         }
         if (lnk.is_b_head_linked != (chb[j] > 0))
         {
-            std::reverse(chb.begin(), chb.end());
-            for (auto &val: chb) val = -val; // flip orientation
-            j = chb.size() - 1 - j;
+            reverse_chain(chb, j);
         }
         
         // join cha and chb if a is the last and b is the first
@@ -319,10 +346,142 @@ void FragSortTool::sort_according_likelihood_unionFind(
             std::copy(chb.begin(), chb.end(), std::back_inserter(cha));
             for (auto &val: chb)
                 chain_id[std::abs(val)-1] = ca;
+            if (sink_chain_id == cb) sink_chain_id = ca;
+            if (source_chain_id == cb) source_chain_id = ca;
             chb.clear();
         }
     }
 
+}
+
+
+void fix_source_sink_frag(
+    std::vector<std::deque<s32>>& chromosomes, 
+    SelectArea& select_area, 
+    const s32& source_frag_id,
+    const s32 sink_frag_id
+)
+{   
+    /*
+    TODO (SHAOHENG) 这里还是有问题，当只有一个片段的时候有时候会把第一个片段反向，检查这个地方
+    */
+
+    if (sink_frag_id < 0 && source_frag_id < 0) return ;
+
+    if (select_area.selected_frag_ids.size() > 0)
+    {   
+        for (u32 i = 0; i < chromosomes.size() ; i ++ )
+        {   
+            std::deque<s32>& tmp = chromosomes[i];
+            if (tmp.empty()) continue;
+            
+            if (std::abs(tmp.back()) == source_frag_id && tmp.size() > 1 ) reverse_chain(tmp);
+            if (std::abs(tmp.front()) == source_frag_id && i != 0)
+            {   
+                if (tmp.front() < 0)
+                {
+                    fmt::print(
+                        stderr,
+                        "Error: the source_frag_id({}) should not be inversed, chromosome[{}].font() = {} \n", 
+                        source_frag_id, 
+                        i, 
+                        tmp.front());
+                }
+                std::swap(tmp, chromosomes.front());
+                continue;
+            }
+
+            if (std::abs(tmp.front()) == sink_frag_id && tmp.size() > 1) reverse_chain(tmp);
+            if (std::abs(tmp.back()) == sink_frag_id)
+            {   
+                if (tmp.back() < 0)
+                {
+                    fmt::print(
+                        stderr,
+                        "Error: the sink_frag_id({}) should not be inversed, chromosome[{}].back() = {} \n", 
+                        sink_frag_id, 
+                        i, 
+                        tmp.back());
+                }
+                std::swap(tmp, chromosomes.back());
+                continue;
+            }
+        }
+    }
+}
+
+
+
+FragSortTool::FragSortTool() {}
+
+FragSortTool::~FragSortTool() {}
+
+void FragSortTool::sort_according_likelihood_unionFind(
+    const LikelihoodTable& likelihood_table, 
+    FragsOrder& frags_order, 
+    SelectArea& select_area,
+    const f32 threshold, 
+    const Frag4compress* frags, 
+    bool sort_according_len_flag) const
+{   
+    // this is used to fix the source and sink during local sort
+    s32 source_frag_id = -1, source_chain_id = -1, 
+        sink_frag_id = -1,   sink_chain_id = -1;
+    if (select_area.source_frag_id >=0)
+    {
+        source_frag_id = 1;
+        source_chain_id = 0;
+    }
+    if (select_area.sink_frag_id >=0)
+    {
+        sink_frag_id = select_area.get_to_sort_frags_num();
+        sink_chain_id = sink_frag_id - 1;
+    }
+
+    u32 n = likelihood_table.get_num_frags();
+    std::vector<Link> all_links;
+
+    std::vector<std::pair<bool, bool>> link_kinds = {
+        {true,  true}, {true,  false}, 
+        {false, true}, {false, false}};
+    for (u32 i = 0; i < n; i++) 
+    {
+        for (u32 j = i+1; j < n; j++)
+        {
+            for (u32 k = 0; k < 4; k++)
+            {   
+                f32 val = likelihood_table(i, j, k);
+                if (val >= threshold) {
+                    all_links.push_back(
+                        Link(val, i, j, link_kinds[k].first, link_kinds[k].second));
+                }
+            }
+        }
+    }
+
+    // Sort all links by descending score
+    std::sort(all_links.begin(), all_links.end(), [](const Link& a, const Link& b) {
+        return a.score > b.score;
+    });
+    // initialize the chromosomes and union-find
+    // every fragment is initilised as a not-inverted chain
+    std::vector<std::deque<s32>> chromosomes;
+    std::vector<s32> chain_id(n, -1);
+    std::vector<std::deque<s32>> chromosomes_excluded;
+    initilise_chromosomes(chromosomes, chromosomes_excluded, chain_id, n, likelihood_table);
+
+    // run union find sort
+    union_find_process(
+        all_links, 
+        chromosomes, 
+        select_area,
+        chain_id,
+        source_frag_id, 
+        sink_frag_id, 
+        source_chain_id, 
+        sink_chain_id
+    );
+
     // sort the chromosomes according to the length
     if (frags && sort_according_len_flag)
     {
@@ -342,6 +501,15 @@ void FragSortTool::sort_according_likelihood_unionFind(
                 return len_a > len_b;
             });
     }
+
+    // fix source and sink to the head and end
+    fix_source_sink_frag(
+        chromosomes, 
+        select_area, 
+        source_frag_id, 
+        sink_frag_id
+    );
+
     // merge two vectors
     chromosomes.insert(chromosomes.end(), chromosomes_excluded.begin(), chromosomes_excluded.end());
 
@@ -360,7 +528,9 @@ f32 link_score(
     u64 k = (u64)(std::abs(source)<<1|(source<0?1:0))<<32|(std::abs(sink)<<1|(sink<0?1:0));
     auto it = map_links.find(k);
     if (it != map_links.end())
+    {
         return all_links[it->second].score;
+    }
     return -std::numeric_limits<f32>::infinity();
 }
 
@@ -564,12 +734,29 @@ std::deque<s32> fuse_chromosomes_shguan(
 void FragSortTool::sort_according_likelihood_unionFind_doFuse(
     const LikelihoodTable& likelihood_table, 
     FragsOrder& frags_order, 
+    SelectArea& select_area,
     const f32 threshold, 
     const Frag4compress* frags,
     const bool doStageOne,
     const bool doStageTwo, 
     bool sort_according_len_flag) const 
 {   
+
+    // this is used to fix the source and sink during local sort
+    s32 source_frag_id = -1, source_chain_id = -1, 
+        sink_frag_id = -1,   sink_chain_id = -1;
+    if (select_area.source_frag_id >=0)
+    {
+        source_frag_id = 1;
+        source_chain_id = 0;
+    }
+    if (select_area.sink_frag_id >=0)
+    {
+        sink_frag_id = select_area.get_to_sort_frags_num();
+        sink_chain_id = sink_frag_id - 1;
+    }
+
+    // *** 
     u32 n = likelihood_table.get_num_frags();
     std::vector<Link> all_links;
     std::map<u64, u32> map_links;
@@ -596,7 +783,7 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
     std::sort(all_links.begin(), all_links.end(), [](const Link& a, const Link& b) {
         return a.score > b.score;
     });
-    if (doStageTwo) 
+    if (doStageTwo)
     {
         // Map sort links to frag pairs  (frga<<1|oria)<<32|(frgb<<1|orib)
         for (u32 i = 0; i < all_links.size(); i++)
@@ -616,51 +803,16 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
     // Process links in order of decreasing score
     if (doStageOne)
     {
-        // Stage I end to end join
-        for (auto &lnk : all_links)
-        {
-            u32 a = lnk.frag_a;
-            u32 b = lnk.frag_b;
-            u32 ca = chain_id[a], cb = chain_id[b];
-
-            if (ca == cb) continue;
-
-            // Decide how to connect them:
-            auto &cha = chromosomes[ca];
-            auto &chb = chromosomes[cb];
-            
-            // Find the index i/j of a/b on cha/b
-            u32 i = 0, j = 0;
-            for (; i < cha.size(); i++)
-                if (std::abs(cha[i]) == a+1)
-                    break;
-            for (; j < chb.size(); j++)
-                if (std::abs(chb[j]) == b+1)
-                    break;
-            
-            // Reverse a and/or b if necessary to make it always a->b
-            if (lnk.is_a_head_linked != (cha[i] < 0))
-            {
-                std::reverse(cha.begin(), cha.end());
-                for (auto &val: cha) val = -val; // flip orientation
-                i = cha.size() - 1 - i;
-            }
-            if (lnk.is_b_head_linked != (chb[j] > 0))
-            {
-                std::reverse(chb.begin(), chb.end());
-                for (auto &val: chb) val = -val; // flip orientation
-                j = chb.size() - 1 - j;
-            }
-            
-            // join cha and chb if a is the last and b is the first
-            if (i == cha.size() - 1 && j == 0) 
-            {
-                std::copy(chb.begin(), chb.end(), std::back_inserter(cha));
-                for (auto &val: chb)
-                    chain_id[std::abs(val)-1] = ca;
-                chb.clear();
-            }
-        }
+        union_find_process(
+            all_links, 
+            chromosomes,
+            select_area,
+            chain_id, 
+            source_frag_id, 
+            sink_frag_id, 
+            source_chain_id,
+            sink_chain_id
+        );
     }
 
     auto start_time = std::chrono::high_resolution_clock::now();
@@ -674,6 +826,7 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
             u32 ca = chain_id[a], cb = chain_id[b];
 
             if (ca == cb) continue;
+            if (!check_source_sink_attachable(lnk, source_frag_id, sink_frag_id, source_chain_id, sink_chain_id, ca, cb)) continue;
 
             // Decide how to connect them:
             auto &cha = chromosomes[ca];
@@ -691,24 +844,24 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
             // Reverse a and/or b if necessary to make it always a->b
             if (lnk.is_a_head_linked != (cha[i] < 0))
             {
-                std::reverse(cha.begin(), cha.end());
-                for (auto &val: cha) val = -val; // flip orientation
-                i = cha.size() - 1 - i;
+                reverse_chain(cha, i);
             }
             if (lnk.is_b_head_linked != (chb[j] > 0))
             {
-                std::reverse(chb.begin(), chb.end());
-                for (auto &val: chb) val = -val; // flip orientation
-                j = chb.size() - 1 - j;
+                reverse_chain(chb, j);
             }
+            
+            // should not fuse the source or sink inside of the chain   
+            if (std::abs(cha.back()) ==  source_frag_id || std::abs(cha.back()) ==  sink_frag_id || 
+                std::abs(chb.front()) == source_frag_id || std::abs(chb.front()) == sink_frag_id ) continue;
 
             // Now try to fuse cha and chb at position i and j
             // the fragments (i...end] on cha
             // and the fragments [beg...j) on chb
             // the sequence order in each chromosome keep unchanged
             // are fused to maximum the overall score (measured by link strength)
-            std::deque<s32> atail(0); // TODO (shaoheng): 检查这里是不是应该填 a 的上一个片段
-            if (cha.size() > i+1) // TODO (shaoheng): 检查这里到底是 i + 1 还是 i
+            std::deque<s32> atail(0); 
+            if (cha.size() > i+1) 
                 std::copy(cha.begin()+i+1, cha.end(), std::back_inserter(atail));
             std::deque<s32> bhead(chb.begin(), chb.begin()+j+1);
 
@@ -729,13 +882,19 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
                     std::copy(chb.begin()+j+1, chb.end(), std::back_inserter(cha)); // append the rest of chb to cha
                 for (auto &val: chb)                                                // update chain_id of fragments in chb 
                     chain_id[std::abs(val)-1] = ca;
+                if (sink_chain_id == cb) sink_chain_id = ca;
+                if (source_chain_id == cb) source_chain_id = ca;
                 chb.clear();
             }
         }
     }
     auto end_time = std::chrono::high_resolution_clock::now();
     std::chrono::duration<f64> elapsed_seconds = end_time-start_time;
-    fprintf(stdout, "[Fusing]: consumed time %.2fs\n", elapsed_seconds.count());
+    fmt::print(
+        stdout, 
+        "[Fusing]: consumed time {:.2f}s\n", 
+        elapsed_seconds.count()
+    );
 
     // sort the chromosomes according to the length
     if (frags && sort_according_len_flag)
@@ -756,6 +915,15 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
                 return len_a > len_b;
             });
     }
+
+    // fix source and sink to the head and end
+    fix_source_sink_frag(
+        chromosomes, 
+        select_area, 
+        source_frag_id, 
+        sink_frag_id
+    );
+
     // merge two vectors
     chromosomes.insert(chromosomes.end(), chromosomes_excluded.begin(), chromosomes_excluded.end());
 
diff --git a/src/frag_sort.h b/src/frag_sort.h
index 50bbc18..4be194a 100644
--- a/src/frag_sort.h
+++ b/src/frag_sort.h
@@ -265,6 +265,7 @@ class FragSortTool
     void sort_according_likelihood_unionFind(
         const LikelihoodTable& likelihood_table, 
         FragsOrder& frags_order,
+        SelectArea& select_area,
         const f32 threshold=-0.001, 
         const Frag4compress* frags=nullptr, 
         bool sort_according_len_flag=true) const;
@@ -272,6 +273,7 @@ class FragSortTool
     void sort_according_likelihood_unionFind_doFuse(
         const LikelihoodTable& likelihood_table, 
         FragsOrder& frags_order,
+        SelectArea& select_area,
         const f32 threshold=-0.001, 
         const Frag4compress* frags=nullptr,
         const bool doStageOne=true,
diff --git a/src/genomeData.h b/src/genomeData.h
index 0942675..a503b76 100644
--- a/src/genomeData.h
+++ b/src/genomeData.h
@@ -60,7 +60,7 @@ struct original_contig
 {
     u32 name[16];         // contig name
     u32 *contigMapPixels = nullptr; // global center coordinate of the fragments
-    u32 nContigs;         // num of contigs
+    u32 nContigs;         // num of contigs, 原contig被分成了多少个fragments
     u32 pad;
 };
 
@@ -73,6 +73,11 @@ struct contig
     u32 startCoord;       // local coordinate on the original contig
     u32 scaffId;
     u32 pad;
+
+    u32 get_original_contig_id() const 
+    {
+        return originalContigId % Max_Number_of_Contigs;
+    }
 };
 
 
@@ -104,13 +109,22 @@ struct map_state
         return originalContigIds[pixel] % Max_Number_of_Contigs;
     }
 
-    void link_splited_contigs(const u32 num_pixel_1d)
+    void restore_cutted_contigs_all(const u32& num_pixel_1d)
     {
         for (u32 i = 0; i < num_pixel_1d; i++)
         {
             originalContigIds[i] = originalContigIds[i] % Max_Number_of_Contigs;
         }
     }
+
+    void restore_cutted_contigs(const s32 start_pixel, const s32 end_pixel)
+    {   
+        if (start_pixel < 0 || end_pixel < 0) return;
+        for (u32 i = start_pixel; i <= end_pixel; i++)
+        {
+            originalContigIds[i] = originalContigIds[i] % Max_Number_of_Contigs;
+        }
+    }
 };
 
 
diff --git a/src/utilsPretextView.h b/src/utilsPretextView.h
index 7e57ba1..29ec32c 100644
--- a/src/utilsPretextView.h
+++ b/src/utilsPretextView.h
@@ -34,6 +34,7 @@ SOFTWARE.
 #include <chrono>
 #include <unordered_set>
 #include <queue>
+#include <fmt/core.h>
 
 
 #include <glm/glm.hpp>

From 05de40d558a8da2a88948135a96b7b9fef776374 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 24 Mar 2025 14:35:24 +0000
Subject: [PATCH 12/33] remove track

---
 .gitignore | 1 +
 1 file changed, 1 insertion(+)

diff --git a/.gitignore b/.gitignore
index a3fcf7c..3289690 100644
--- a/.gitignore
+++ b/.gitignore
@@ -31,6 +31,7 @@ python/compressed_data_test/
 python/autoCut/models/cfgs/error_model.pth
 python/autoCut/models/cfgs/chr_model.pth
 python/autoCut/models/swin/mmdet.egg-info/*
+python/classic_cut/show_hic_density.py
 
 subprojects/eigen
 subprojects/glfw

From 94079ae2077c651ebe8b3e39e9429048b7e9d919 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 24 Mar 2025 14:42:00 +0000
Subject: [PATCH 13/33] force msvc to use utf-8

---
 CMakeLists.txt | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 0109195..2ec5938 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -55,6 +55,9 @@ else()
     endif()
 endif()
 
+if (MSVC)
+    add_compile_options(/utf-8)  # force MSVC to use UTF-8 
+endif()
 
 message(STATUS "Install path: ${CMAKE_INSTALL_PREFIX}")
 

From 943854cce5891d95db1a5d35ba1c04e229f65055 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 24 Mar 2025 14:47:16 +0000
Subject: [PATCH 14/33] update version

---
 CMakeLists.txt | 2 +-
 Info.plist     | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 2ec5938..0a9d104 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -1,5 +1,5 @@
 cmake_minimum_required(VERSION 3.17)
-project(PretextView VERSION "1.0.1")
+project(PretextView VERSION "1.0.2")
 
 
 option(FORCE_MAC_X86 "Force macOS x86_64 build" OFF)  
diff --git a/Info.plist b/Info.plist
index 1d41b42..608a4e5 100644
--- a/Info.plist
+++ b/Info.plist
@@ -13,7 +13,7 @@
   <key>CFBundleIconFile</key>
   <string>icon.icns</string>
   <key>CFBundleShortVersionString</key>
-  <string>1.0.1</string>
+  <string>1.0.2</string>
   <key>CFBundleInfoDictionaryVersion</key>
   <string>6.0</string>
   <key>CFBundlePackageType</key>

From 00dc83b87fd70be6d7d90b879163f132e0a274d9 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:02:53 +0100
Subject: [PATCH 15/33] color blind color palette

---
 src/ColorMapData.cpp | 7349 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 7349 insertions(+)
 create mode 100644 src/ColorMapData.cpp

diff --git a/src/ColorMapData.cpp b/src/ColorMapData.cpp
new file mode 100644
index 0000000..3782ad1
--- /dev/null
+++ b/src/ColorMapData.cpp
@@ -0,0 +1,7349 @@
+/*
+Copyright (c) 2021 Ed Harry, Wellcome Sanger Institute
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+*/
+
+#ifndef HEADER_H  // 防止重定义错误
+#include "Header.h"
+#endif
+
+#define Number_of_Color_Maps (sizeof(Color_Map_Data) / sizeof(Color_Map_Data[0]))
+
+
+
+// global_variable
+f32 Color_Map_Data[][768]= 
+{   
+
+    // lightsunset
+    { 
+        0.98000, 0.98000, 0.95000, 0.97643, 0.97835, 0.95137, 0.97286, 0.97671, 0.95275, 0.96929, 0.97506, 0.95412, 0.96573, 0.97341, 0.95549, 0.96216, 0.97176, 0.95686, 0.95859, 0.97012, 0.95824, 0.95502, 0.96847, 0.95961, 0.95145, 0.96682, 0.96098, 0.94788, 0.96518, 0.96235, 0.94431, 0.96353, 0.96373, 0.94075, 0.96188, 0.96510, 0.93718, 0.96024, 0.96647, 0.93361, 0.95859, 0.96784, 0.93004, 0.95694, 0.96922, 0.92647, 0.95529, 0.97059, 0.92290, 0.95365, 0.97196, 0.91933, 0.95200, 0.97333, 0.91576, 0.95035, 0.97471, 0.91220, 0.94871, 0.97608, 0.90863, 0.94706, 0.97745, 0.90506, 0.94541, 0.97882, 0.90149, 0.94376, 0.98020, 0.89792, 0.94212, 0.98157, 0.89435, 0.94047, 0.98294, 0.89078, 0.93882, 0.98431, 0.88722, 0.93718, 0.98569, 0.88365, 0.93553, 0.98706, 0.88008, 0.93388, 0.98843, 0.87651, 0.93224, 0.98980, 0.87294, 0.93059, 0.99118, 0.86937, 0.92894, 0.99255, 0.86580, 0.92729, 0.99392, 0.86224, 0.92565, 0.99529, 0.85867, 0.92400, 0.99667, 0.85510, 0.92235, 0.99804, 0.85153, 0.92071, 0.99941, 0.84529, 0.91765, 0.99984, 0.83706, 0.91353, 0.99957, 0.82882, 0.90941, 0.99929, 0.82059, 0.90529, 0.99902, 0.81235, 0.90118, 0.99875, 0.80412, 0.89706, 0.99847, 0.79588, 0.89294, 0.99820, 0.78765, 0.88882, 0.99792, 0.77941, 0.88471, 0.99765, 0.77118, 0.88059, 0.99737, 0.76294, 0.87647, 0.99710, 0.75471, 0.87235, 0.99682, 0.74647, 0.86824, 0.99655, 0.73824, 0.86412, 0.99627, 0.73000, 0.86000, 0.99600, 0.72176, 0.85588, 0.99573, 0.71353, 0.85176, 0.99545, 0.70529, 0.84765, 0.99518, 0.69706, 0.84353, 0.99490, 0.68882, 0.83941, 0.99463, 0.68059, 0.83529, 0.99435, 0.67235, 0.83118, 0.99408, 0.66412, 0.82706, 0.99380, 0.65588, 0.82294, 0.99353, 0.64765, 0.81882, 0.99325, 0.63941, 0.81471, 0.99298, 0.63118, 0.81059, 0.99271, 0.62294, 0.80647, 0.99243, 0.61471, 0.80235, 0.99216, 0.60647, 0.79824, 0.99188, 0.59824, 0.79412, 0.99161, 0.59000, 0.79000, 0.99133, 0.58176, 0.78588, 0.99106, 0.57353, 0.78176, 0.99078, 0.56529, 0.77765, 0.99051, 0.55706, 0.77353, 0.99024, 0.54863, 0.76914, 0.98969, 0.53902, 0.76310, 0.98749, 0.52941, 0.75706, 0.98529, 0.51980, 0.75102, 0.98310, 0.51020, 0.74498, 0.98090, 0.50059, 0.73894, 0.97871, 0.49098, 0.73290, 0.97651, 0.48137, 0.72686, 0.97431, 0.47176, 0.72082, 0.97212, 0.46216, 0.71478, 0.96992, 0.45255, 0.70875, 0.96773, 0.44294, 0.70271, 0.96553, 0.43333, 0.69667, 0.96333, 0.42373, 0.69063, 0.96114, 0.41412, 0.68459, 0.95894, 0.40451, 0.67855, 0.95675, 0.39490, 0.67251, 0.95455, 0.38529, 0.66647, 0.95235, 0.37569, 0.66043, 0.95016, 0.36608, 0.65439, 0.94796, 0.35647, 0.64835, 0.94576, 0.34686, 0.64231, 0.94357, 0.33725, 0.63627, 0.94137, 0.32765, 0.63024, 0.93918, 0.31804, 0.62420, 0.93698, 0.30843, 0.61816, 0.93478, 0.29882, 0.61212, 0.93259, 0.28922, 0.60608, 0.93039, 0.27961, 0.60004, 0.92820, 0.27000, 0.59400, 0.92600, 0.26039, 0.58796, 0.92380, 0.25078, 0.58192, 0.92161, 0.24118, 0.57588, 0.91941, 0.23157, 0.56984, 0.91722, 0.22196, 0.56380, 0.91502, 0.21235, 0.55776, 0.91282, 0.20275, 0.55173, 0.91063, 0.19804, 0.54608, 0.90588, 0.19529, 0.54059, 0.90012, 0.19255, 0.53510, 0.89435, 0.18980, 0.52961, 0.88859, 0.18706, 0.52412, 0.88282, 0.18431, 0.51863, 0.87706, 0.18157, 0.51314, 0.87129, 0.17882, 0.50765, 0.86553, 0.17608, 0.50216, 0.85976, 0.17333, 0.49667, 0.85400, 0.17059, 0.49118, 0.84824, 0.16784, 0.48569, 0.84247, 0.16510, 0.48020, 0.83671, 0.16235, 0.47471, 0.83094, 0.15961, 0.46922, 0.82518, 0.15686, 0.46373, 0.81941, 0.15412, 0.45824, 0.81365, 0.15137, 0.45275, 0.80788, 0.14863, 0.44725, 0.80212, 0.14588, 0.44176, 0.79635, 0.14314, 0.43627, 0.79059, 0.14039, 0.43078, 0.78482, 0.13765, 0.42529, 0.77906, 0.13490, 0.41980, 0.77329, 0.13216, 0.41431, 0.76753, 0.12941, 0.40882, 0.76176, 0.12667, 0.40333, 0.75600, 0.12392, 0.39784, 0.75024, 0.12118, 0.39235, 0.74447, 0.11843, 0.38686, 0.73871, 0.11569, 0.38137, 0.73294, 0.11294, 0.37588, 0.72718, 0.11020, 0.37039, 0.72141, 0.10745, 0.36490, 0.71565, 0.10471, 0.35941, 0.70988, 0.10196, 0.35392, 0.70412, 0.10157, 0.34882, 0.69961, 0.10706, 0.34471, 0.69824, 0.11255, 0.34059, 0.69686, 0.11804, 0.33647, 0.69549, 0.12353, 0.33235, 0.69412, 0.12902, 0.32824, 0.69275, 0.13451, 0.32412, 0.69137, 0.14000, 0.32000, 0.69000, 0.14549, 0.31588, 0.68863, 0.15098, 0.31176, 0.68725, 0.15647, 0.30765, 0.68588, 0.16196, 0.30353, 0.68451, 0.16745, 0.29941, 0.68314, 0.17294, 0.29529, 0.68176, 0.17843, 0.29118, 0.68039, 0.18392, 0.28706, 0.67902, 0.18941, 0.28294, 0.67765, 0.19490, 0.27882, 0.67627, 0.20039, 0.27471, 0.67490, 0.20588, 0.27059, 0.67353, 0.21137, 0.26647, 0.67216, 0.21686, 0.26235, 0.67078, 0.22235, 0.25824, 0.66941, 0.22784, 0.25412, 0.66804, 0.23333, 0.25000, 0.66667, 0.23882, 0.24588, 0.66529, 0.24431, 0.24176, 0.66392, 0.24980, 0.23765, 0.66255, 0.25529, 0.23353, 0.66118, 0.26078, 0.22941, 0.65980, 0.26627, 0.22529, 0.65843, 0.27176, 0.22118, 0.65706, 0.27725, 0.21706, 0.65569, 0.28275, 0.21294, 0.65431, 0.28824, 0.20882, 0.65294, 0.29373, 0.20471, 0.65157, 0.29922, 0.20059, 0.65020, 0.30706, 0.20000, 0.64882, 0.31529, 0.20000, 0.64745, 0.32353, 0.20000, 0.64608, 0.33176, 0.20000, 0.64471, 0.34000, 0.20000, 0.64333, 0.34824, 0.20000, 0.64196, 0.35647, 0.20000, 0.64059, 0.36471, 0.20000, 0.63922, 0.37294, 0.20000, 0.63784, 0.38118, 0.20000, 0.63647, 0.38941, 0.20000, 0.63510, 0.39765, 0.20000, 0.63373, 0.40588, 0.20000, 0.63235, 0.41412, 0.20000, 0.63098, 0.42235, 0.20000, 0.62961, 0.43059, 0.20000, 0.62824, 0.43882, 0.20000, 0.62686, 0.44706, 0.20000, 0.62549, 0.45529, 0.20000, 0.62412, 0.46353, 0.20000, 0.62275, 0.47176, 0.20000, 0.62137, 0.48000, 0.20000, 0.62000, 0.48824, 0.20000, 0.61863, 0.49647, 0.20000, 0.61725, 0.50471, 0.20000, 0.61588, 0.51294, 0.20000, 0.61451, 0.52118, 0.20000, 0.61314, 0.52941, 0.20000, 0.61176, 0.53765, 0.20000, 0.61039, 0.54588, 0.20000, 0.60902, 0.55412, 0.20000, 0.60765, 0.56235, 0.20000, 0.60627, 0.57059, 0.20000, 0.60490, 0.57882, 0.20000, 0.60353, 0.58706, 0.20000, 0.60216, 0.59529, 0.20000, 0.60078, 0.60353, 0.20471, 0.59765, 0.61176, 0.21569, 0.59216, 0.62000, 0.22667, 0.58667, 0.62824, 0.23765, 0.58118, 0.63647, 0.24863, 0.57569, 0.64471, 0.25961, 0.57020, 0.65294, 0.27059, 0.56471, 0.66118, 0.28157, 0.55922, 0.66941, 0.29255, 0.55373, 0.67765, 0.30353, 0.54824, 0.68588, 0.31451, 0.54275, 0.69412, 0.32549, 0.53725, 0.70235, 0.33647, 0.53176, 0.71059, 0.34745, 0.52627, 0.71882, 0.35843, 0.52078, 0.72706, 0.36941, 0.51529, 0.73529, 0.38039, 0.50980, 0.74353, 0.39137, 0.50431, 0.75176, 0.40235, 0.49882, 0.76000, 0.41333, 0.49333, 0.76824, 0.42431, 0.48784, 0.77647, 0.43529, 0.48235, 0.78471, 0.44627, 0.47686, 0.79294, 0.45725, 0.47137, 0.80118, 0.46824, 0.46588, 0.80941, 0.47922, 0.46039, 0.81765, 0.49020, 0.45490, 0.82588, 0.50118, 0.44941, 0.83412, 0.51216, 0.44392, 0.84235, 0.52314, 0.43843, 0.85059, 0.53412, 0.43294, 0.85882, 0.54510, 0.42745, 0.86706, 0.55608, 0.42196, 0.87529, 0.56706, 0.41647, 0.88353, 0.57804, 0.41098, 0.89176, 0.58902, 0.40549, 0.90000, 0.60000, 0.40000 
+    },
+
+    // viridis color blind
+    {
+        0.267004, 0.004874, 0.329415, 0.26851, 0.009605, 0.335427, 0.269944, 0.014625, 0.341379, 0.271305, 0.019942, 0.347269, 0.272594, 0.025563, 0.353093, 0.273809, 0.031497, 0.358853, 0.274952, 0.037752, 0.364543, 0.276022, 0.044167, 0.370164, 0.277018, 0.050344, 0.375715, 0.277941, 0.056324, 0.381191, 0.278791, 0.062145, 0.386592, 0.279566, 0.067836, 0.391917, 0.280267, 0.073417, 0.397163, 0.280894, 0.078907, 0.402329, 0.281446, 0.08432, 0.407414, 0.281924, 0.089666, 0.412415, 0.282327, 0.094955, 0.417331, 0.282656, 0.100196, 0.42216, 0.28291, 0.105393, 0.426902, 0.283091, 0.110553, 0.431554, 0.283197, 0.11568, 0.436115, 0.283229, 0.120777, 0.440584, 0.283187, 0.125848, 0.44496, 0.283072, 0.130895, 0.449241, 0.282884, 0.13592, 0.453427, 0.282623, 0.140926, 0.457517, 0.28229, 0.145912, 0.46151, 0.281887, 0.150881, 0.465405, 0.281412, 0.155834, 0.469201, 0.280868, 0.160771, 0.472899, 0.280255, 0.165693, 0.476498, 0.279574, 0.170599, 0.479997, 0.278826, 0.17549, 0.483397, 0.278012, 0.180367, 0.486697, 0.277134, 0.185228, 0.489898, 0.276194, 0.190074, 0.493001, 0.275191, 0.194905, 0.496005, 0.274128, 0.199721, 0.498911, 0.273006, 0.20452, 0.501721, 0.271828, 0.209303, 0.504434, 0.270595, 0.214069, 0.507052, 0.269308, 0.218818, 0.509577, 0.267968, 0.223549, 0.512008, 0.26658, 0.228262, 0.514349, 0.265145, 0.232956, 0.516599, 0.263663, 0.237631, 0.518762, 0.262138, 0.242286, 0.520837, 0.260571, 0.246922, 0.522828, 0.258965, 0.251537, 0.524736, 0.257322, 0.25613, 0.526563, 0.255645, 0.260703, 0.528312, 0.253935, 0.265254, 0.529983, 0.252194, 0.269783, 0.531579, 0.250425, 0.27429, 0.533103, 0.248629, 0.278775, 0.534556, 0.246811, 0.283237, 0.535941, 0.244972, 0.287675, 0.53726, 0.243113, 0.292092, 0.538516, 0.241237, 0.296485, 0.539709, 0.239346, 0.300855, 0.540844, 0.237441, 0.305202, 0.541921, 0.235526, 0.309527, 0.542944, 0.233603, 0.313828, 0.543914, 0.231674, 0.318106, 0.544834, 0.229739, 0.322361, 0.545706, 0.227802, 0.326594, 0.546532, 0.225863, 0.330805, 0.547314, 0.223925, 0.334994, 0.548053, 0.221989, 0.339161, 0.548752, 0.220057, 0.343307, 0.549413, 0.21813, 0.347432, 0.550038, 0.21621, 0.351535, 0.550627, 0.214298, 0.355619, 0.551184, 0.212395, 0.359683, 0.55171, 0.210503, 0.363727, 0.552206, 0.208623, 0.367752, 0.552675, 0.206756, 0.371758, 0.553117, 0.204903, 0.375746, 0.553533, 0.203063, 0.379716, 0.553925, 0.201239, 0.38367, 0.554294, 0.19943, 0.387607, 0.554642, 0.197636, 0.391528, 0.554969, 0.19586, 0.395433, 0.555276, 0.1941, 0.399323, 0.555565, 0.192357, 0.403199, 0.555836, 0.190631, 0.407061, 0.556089, 0.188923, 0.41091, 0.556326, 0.187231, 0.414746, 0.556547, 0.185556, 0.41857, 0.556753, 0.183898, 0.422383, 0.556944, 0.182256, 0.426184, 0.55712, 0.180629, 0.429975, 0.557282, 0.179019, 0.433756, 0.55743, 0.177423, 0.437527, 0.557565, 0.175841, 0.44129, 0.557685, 0.174274, 0.445044, 0.557792, 0.172719, 0.448791, 0.557885, 0.171176, 0.45253, 0.557965, 0.169646, 0.456262, 0.55803, 0.168126, 0.459988, 0.558082, 0.166617, 0.463708, 0.558119, 0.165117, 0.467423, 0.558141, 0.163625, 0.471133, 0.558148, 0.162142, 0.474838, 0.55814, 0.160665, 0.47854, 0.558115, 0.159194, 0.482237, 0.558073, 0.157729, 0.485932, 0.558013, 0.15627, 0.489624, 0.557936, 0.154815, 0.493313, 0.55784, 0.153364, 0.497, 0.557724, 0.151918, 0.500685, 0.557587, 0.150476, 0.504369, 0.55743, 0.149039, 0.508051, 0.55725, 0.147607, 0.511733, 0.557049, 0.14618, 0.515413, 0.556823, 0.144759, 0.519093, 0.556572, 0.143343, 0.522773, 0.556295, 0.141935, 0.526453, 0.555991, 0.140536, 0.530132, 0.555659, 0.139147, 0.533812, 0.555298, 0.13777, 0.537492, 0.554906, 0.136408, 0.541173, 0.554483, 0.135066, 0.544853, 0.554029, 0.133743, 0.548535, 0.553541, 0.132444, 0.552216, 0.553018, 0.131172, 0.555899, 0.552459, 0.129933, 0.559582, 0.551864, 0.128729, 0.563265, 0.551229, 0.127568, 0.566949, 0.550556, 0.126453, 0.570633, 0.549841, 0.125394, 0.574318, 0.549086, 0.124395, 0.578002, 0.548287, 0.123463, 0.581687, 0.547445, 0.122606, 0.585371, 0.546557, 0.121831, 0.589055, 0.545623, 0.121148, 0.592739, 0.544641, 0.120565, 0.596422, 0.543611, 0.120092, 0.600104, 0.54253, 0.119738, 0.603785, 0.5414, 0.119512, 0.607464, 0.540218, 0.119423, 0.611141, 0.538982, 0.119483, 0.614817, 0.537692, 0.119699, 0.61849, 0.536347, 0.120081, 0.622161, 0.534946, 0.120638, 0.625828, 0.533488, 0.12138, 0.629492, 0.531973, 0.122312, 0.633153, 0.530398, 0.123444, 0.636809, 0.528763, 0.12478, 0.640461, 0.527068, 0.126326, 0.644107, 0.525311, 0.128087, 0.647749, 0.523491, 0.130067, 0.651384, 0.521608, 0.132268, 0.655014, 0.519661, 0.134692, 0.658636, 0.517649, 0.137339, 0.662252, 0.515571, 0.14021, 0.665859, 0.513427, 0.143303, 0.669459, 0.511215, 0.146616, 0.67305, 0.508936, 0.150148, 0.676631, 0.506589, 0.153894, 0.680203, 0.504172, 0.157851, 0.683765, 0.501686, 0.162016, 0.687316, 0.499129, 0.166383, 0.690856, 0.496502, 0.170948, 0.694384, 0.493803, 0.175707, 0.6979, 0.491033, 0.180653, 0.701402, 0.488189, 0.185783, 0.704891, 0.485273, 0.19109, 0.708366, 0.482284, 0.196571, 0.711827, 0.479221, 0.202219, 0.715272, 0.476084, 0.20803, 0.718701, 0.472873, 0.214, 0.722114, 0.469588, 0.220124, 0.725509, 0.466226, 0.226397, 0.728888, 0.462789, 0.232815, 0.732247, 0.459277, 0.239374, 0.735588, 0.455688, 0.24607, 0.73891, 0.452024, 0.252899, 0.742211, 0.448284, 0.259857, 0.745492, 0.444467, 0.266941, 0.748751, 0.440573, 0.274149, 0.751988, 0.436601, 0.281477, 0.755203, 0.432552, 0.288921, 0.758394, 0.428426, 0.296479, 0.761561, 0.424223, 0.304148, 0.764704, 0.419943, 0.311925, 0.767822, 0.415586, 0.319809, 0.770914, 0.411152, 0.327796, 0.77398, 0.40664, 0.335885, 0.777018, 0.402049, 0.344074, 0.780029, 0.397381, 0.35236, 0.783011, 0.392636, 0.360741, 0.785964, 0.387814, 0.369214, 0.788888, 0.382914, 0.377779, 0.791781, 0.377939, 0.386433, 0.794644, 0.372886, 0.395174, 0.797475, 0.367757, 0.404001, 0.800275, 0.362552, 0.412913, 0.803041, 0.357269, 0.421908, 0.805774, 0.35191, 0.430983, 0.808473, 0.346476, 0.440137, 0.811138, 0.340967, 0.449368, 0.813768, 0.335384, 0.458674, 0.816363, 0.329727, 0.468053, 0.818921, 0.323998, 0.477504, 0.821444, 0.318195, 0.487026, 0.823929, 0.312321, 0.496615, 0.826376, 0.306377, 0.506271, 0.828786, 0.300362, 0.515992, 0.831158, 0.294279, 0.525776, 0.833491, 0.288127, 0.535621, 0.835785, 0.281908, 0.545524, 0.838039, 0.275626, 0.555484, 0.840254, 0.269281, 0.565498, 0.84243, 0.262877, 0.575563, 0.844566, 0.256415, 0.585678, 0.846661, 0.249897, 0.595839, 0.848717, 0.243329, 0.606045, 0.850733, 0.236712, 0.616293, 0.852709, 0.230052, 0.626579, 0.854645, 0.223353, 0.636902, 0.856542, 0.21662, 0.647257, 0.8584, 0.209861, 0.657642, 0.860219, 0.203082, 0.668054, 0.861999, 0.196293, 0.678489, 0.863742, 0.189503, 0.688944, 0.865448, 0.182725, 0.699415, 0.867117, 0.175971, 0.709898, 0.868751, 0.169257, 0.720391, 0.87035, 0.162603, 0.730889, 0.871916, 0.156029, 0.741388, 0.873449, 0.149561, 0.751884, 0.874951, 0.143228, 0.762373, 0.876424, 0.137064, 0.772852, 0.877868, 0.131109, 0.783315, 0.879285, 0.125405, 0.79376, 0.880678, 0.120005, 0.804182, 0.882046, 0.114965, 0.814576, 0.883393, 0.110347, 0.82494, 0.88472, 0.106217, 0.83527, 0.886029, 0.102646, 0.845561, 0.887322, 0.099702, 0.85581, 0.888601, 0.097452, 0.866013, 0.889868, 0.095953, 0.876168, 0.891125, 0.09525, 0.886271, 0.892374, 0.095374, 0.89632, 0.893616, 0.096335, 0.906311, 0.894855, 0.098125, 0.916242, 0.896091, 0.100717, 0.926106, 0.89733, 0.104071, 0.935904, 0.89857, 0.108131, 0.945636, 0.899815, 0.112838, 0.9553, 0.901065, 0.118128, 0.964894, 0.902323, 0.123941, 0.974417, 0.90359, 0.130215, 0.983868, 0.904867, 0.136897, 0.993248, 0.906157, 0.143936
+    }, 
+
+    // cividis
+    { 
+        0.0, 0.135112, 0.304751, 0.0, 0.138068, 0.311105, 0.0, 0.141013, 0.317579, 0.0, 0.143951, 0.323982, 0.0, 0.146877, 0.330479, 0.0, 0.149791, 0.337065, 0.0, 0.152673, 0.343704, 0.0, 0.155377, 0.3505, 0.0, 0.157932, 0.357521, 0.0, 0.160495, 0.364534, 0.0, 0.163058, 0.371608, 0.0, 0.165621, 0.378769, 0.0, 0.168204, 0.385902, 0.0, 0.1708, 0.3931, 0.0, 0.17342, 0.400353, 0.0, 0.176082, 0.407577, 0.0, 0.178802, 0.414764, 0.0, 0.18161, 0.421859, 0.0, 0.18455, 0.428802, 0.0, 0.186915, 0.435532, 0.0, 0.188769, 0.439563, 0.0, 0.19095, 0.441085, 0.0, 0.193366, 0.441561, 0.003602, 0.195911, 0.441564, 0.017852, 0.198528, 0.441248, 0.03211, 0.201199, 0.440785, 0.046205, 0.203903, 0.440196, 0.058378, 0.206629, 0.439531, 0.068968, 0.209372, 0.438863, 0.078624, 0.212122, 0.438105, 0.087465, 0.214879, 0.437342, 0.095645, 0.217643, 0.436593, 0.103401, 0.220406, 0.43579, 0.110658, 0.22317, 0.435067, 0.117612, 0.225935, 0.434308, 0.124291, 0.228697, 0.433547, 0.130669, 0.231458, 0.43284, 0.13683, 0.234216, 0.432148, 0.142852, 0.236972, 0.431404, 0.148638, 0.239724, 0.430752, 0.154261, 0.242475, 0.43012, 0.159733, 0.245221, 0.429528, 0.165113, 0.247965, 0.428908, 0.170362, 0.250707, 0.428325, 0.17549, 0.253444, 0.42779, 0.180503, 0.25618, 0.427299, 0.185453, 0.258914, 0.426788, 0.190303, 0.261644, 0.426329, 0.195057, 0.264372, 0.425924, 0.199764, 0.267099, 0.425497, 0.204385, 0.269823, 0.425126, 0.208926, 0.272546, 0.424809, 0.213431, 0.275266, 0.42448, 0.217863, 0.277985, 0.424206, 0.222264, 0.280702, 0.423914, 0.226598, 0.283419, 0.423678, 0.230871, 0.286134, 0.423498, 0.23512, 0.288848, 0.423304, 0.239312, 0.291562, 0.423167, 0.243485, 0.294274, 0.423014, 0.247605, 0.296986, 0.422917, 0.251675, 0.299698, 0.422873, 0.255731, 0.302409, 0.422814, 0.25974, 0.30512, 0.42281, 0.263738, 0.307831, 0.422789, 0.267693, 0.310542, 0.422821, 0.271639, 0.313253, 0.422837, 0.275513, 0.315965, 0.422979, 0.279411, 0.318677, 0.423031, 0.28324, 0.32139, 0.423211, 0.287065, 0.324103, 0.423373, 0.290884, 0.326816, 0.423517, 0.294669, 0.329531, 0.423716, 0.298421, 0.332247, 0.423973, 0.302169, 0.334963, 0.424213, 0.305886, 0.337681, 0.424512, 0.309601, 0.340399, 0.42479, 0.313287, 0.34312, 0.42512, 0.316941, 0.345842, 0.425512, 0.320595, 0.348565, 0.425889, 0.32425, 0.351289, 0.42625, 0.327875, 0.354016, 0.42667, 0.331474, 0.356744, 0.427144, 0.335073, 0.359474, 0.427605, 0.338673, 0.362206, 0.428053, 0.342246, 0.364939, 0.428559, 0.345793, 0.367676, 0.429127, 0.349341, 0.370414, 0.429685, 0.352892, 0.373153, 0.430226, 0.356418, 0.375896, 0.430823, 0.359916, 0.378641, 0.431501, 0.363446, 0.381388, 0.432075, 0.366923, 0.384139, 0.432796, 0.37043, 0.38689, 0.433428, 0.373884, 0.389646, 0.434209, 0.377371, 0.392404, 0.43489, 0.38083, 0.395164, 0.435653, 0.384268, 0.397928, 0.436475, 0.387705, 0.400694, 0.437305, 0.391151, 0.403464, 0.438096, 0.394568, 0.406236, 0.438986, 0.397991, 0.409011, 0.439848, 0.401418, 0.41179, 0.440708, 0.40482, 0.414572, 0.441642, 0.408226, 0.417357, 0.44257, 0.411607, 0.420145, 0.443577, 0.414992, 0.422937, 0.444578, 0.418383, 0.425733, 0.44556, 0.421748, 0.428531, 0.44664, 0.42512, 0.431334, 0.447692, 0.428462, 0.43414, 0.448864, 0.431817, 0.43695, 0.449982, 0.435168, 0.439763, 0.451134, 0.438504, 0.44258, 0.452341, 0.44181, 0.445402, 0.453659, 0.445148, 0.448226, 0.454885, 0.448447, 0.451053, 0.456264, 0.451759, 0.453887, 0.457582, 0.455072, 0.456718, 0.458976, 0.458366, 0.459552, 0.460457, 0.461616, 0.462405, 0.461969, 0.464947, 0.465241, 0.463395, 0.468254, 0.468083, 0.464908, 0.471501, 0.47096, 0.466357, 0.474812, 0.473832, 0.467681, 0.478186, 0.476699, 0.468845, 0.481622, 0.479573, 0.469767, 0.485141, 0.482451, 0.470384, 0.488697, 0.485318, 0.471008, 0.492278, 0.488198, 0.471453, 0.495913, 0.491076, 0.471751, 0.499552, 0.49396, 0.472032, 0.503185, 0.496851, 0.472305, 0.506866, 0.499743, 0.472432, 0.51054, 0.502643, 0.47255, 0.514226, 0.505546, 0.47264, 0.51792, 0.508454, 0.472707, 0.521643, 0.511367, 0.472639, 0.525348, 0.514285, 0.47266, 0.529086, 0.517207, 0.472543, 0.532829, 0.520135, 0.472401, 0.536553, 0.523067, 0.472352, 0.540307, 0.526005, 0.472163, 0.544069, 0.528948, 0.471947, 0.54784, 0.531895, 0.471704, 0.551612, 0.534849, 0.471439, 0.555393, 0.537807, 0.471147, 0.559181, 0.540771, 0.470829, 0.562972, 0.543741, 0.470488, 0.566802, 0.546715, 0.469988, 0.570607, 0.549695, 0.469593, 0.574417, 0.552682, 0.469172, 0.578236, 0.555673, 0.468724, 0.582087, 0.55867, 0.468118, 0.585916, 0.561674, 0.467618, 0.589753, 0.564682, 0.46709, 0.593622, 0.567697, 0.466401, 0.597469, 0.570718, 0.465821, 0.601354, 0.573743, 0.465074, 0.605211, 0.576777, 0.464441, 0.609105, 0.579816, 0.463638, 0.612977, 0.582861, 0.46295, 0.616852, 0.585913, 0.462237, 0.620765, 0.58897, 0.461351, 0.624654, 0.592034, 0.460583, 0.628576, 0.595104, 0.459641, 0.632506, 0.59818, 0.458668, 0.636412, 0.601264, 0.457818, 0.640352, 0.604354, 0.456791, 0.64427, 0.60745, 0.455886, 0.648222, 0.610553, 0.454801, 0.652178, 0.613664, 0.453689, 0.656114, 0.61678, 0.452702, 0.660082, 0.619904, 0.451534, 0.664055, 0.623034, 0.450338, 0.668008, 0.626171, 0.44927, 0.671991, 0.629316, 0.448018, 0.675981, 0.632468, 0.446736, 0.679979, 0.635626, 0.445424, 0.68395, 0.638793, 0.444251, 0.687957, 0.641966, 0.442886, 0.691971, 0.645145, 0.441491, 0.695985, 0.648334, 0.440072, 0.700008, 0.651529, 0.438624, 0.704037, 0.654731, 0.437147, 0.708067, 0.657942, 0.435647, 0.712105, 0.66116, 0.434117, 0.716177, 0.664384, 0.432386, 0.720222, 0.667618, 0.430805, 0.724274, 0.670859, 0.429194, 0.728334, 0.674107, 0.427554, 0.732422, 0.677364, 0.425717, 0.736488, 0.680629, 0.424028, 0.740589, 0.6839, 0.422131, 0.744664, 0.687181, 0.420393, 0.748772, 0.69047, 0.418448, 0.752886, 0.693766, 0.416472, 0.756975, 0.697071, 0.414659, 0.761096, 0.700384, 0.412638, 0.765223, 0.703705, 0.410587, 0.769353, 0.707035, 0.408516, 0.773486, 0.710373, 0.406422, 0.777651, 0.713719, 0.404112, 0.781795, 0.717074, 0.401966, 0.785965, 0.720438, 0.399613, 0.790116, 0.72381, 0.397423, 0.794298, 0.72719, 0.395016, 0.79848, 0.73058, 0.392597, 0.802667, 0.733978, 0.390153, 0.806859, 0.737385, 0.387684, 0.811054, 0.740801, 0.385198, 0.815274, 0.744226, 0.382504, 0.819499, 0.747659, 0.379785, 0.823729, 0.751101, 0.377043, 0.827959, 0.754553, 0.374292, 0.832192, 0.758014, 0.371529, 0.836429, 0.761483, 0.368747, 0.840693, 0.764962, 0.365746, 0.844957, 0.76845, 0.362741, 0.849223, 0.771947, 0.359729, 0.853515, 0.775454, 0.3565, 0.857809, 0.778969, 0.353259, 0.862105, 0.782494, 0.350011, 0.866421, 0.786028, 0.346571, 0.870717, 0.789572, 0.343333, 0.875057, 0.793125, 0.339685, 0.879378, 0.796687, 0.336241, 0.88372, 0.800258, 0.332599, 0.888081, 0.803839, 0.32877, 0.89244, 0.80743, 0.324968, 0.896818, 0.81103, 0.320982, 0.901195, 0.814639, 0.317021, 0.905589, 0.818257, 0.312889, 0.91, 0.821885, 0.308594, 0.914407, 0.825522, 0.304348, 0.918828, 0.829168, 0.29996, 0.923279, 0.832822, 0.295244, 0.927724, 0.836486, 0.290611, 0.93218, 0.840159, 0.28588, 0.93666, 0.843841, 0.280876, 0.941147, 0.84753, 0.275815, 0.945654, 0.851228, 0.270532, 0.950178, 0.854933, 0.265085, 0.954725, 0.858646, 0.259365, 0.959284, 0.862365, 0.253563, 0.963872, 0.866089, 0.247445, 0.968469, 0.869819, 0.24131, 0.973114, 0.87355, 0.234677, 0.97778, 0.877281, 0.227954, 0.982497, 0.881008, 0.220878, 0.987293, 0.884718, 0.213336, 0.992218, 0.888385, 0.205468, 0.994847, 0.892954, 0.203445, 0.995249, 0.898384, 0.207561, 0.995503, 0.903866, 0.21237, 0.995737, 0.909344, 0.217772 },
+
+    {
+        0.0028108909432653585f,0.00023971464598377364f,0.0139849755645124f,0.004019847439892661f,0.0009782136527954485f,0.018722506642108738f,0.00553592753551016f,0.0018785803659367136f,0.024433316141969492f,0.007352275238375931f,0.0029285938665143995f,0.031151041807332518f,0.009463157808182366f,0.004122022987412015f,0.03885690760774683f,0.01187516425529045f,0.005443920038504773f,0.04716007356275464f,0.01459882064653067f,0.006882989229735782f,0.055492610304536484f,0.017636358974306172f,0.008432150815965444f,0.06383358056387142f,0.02101796809093932f,0.010065826612517249f,0.07226064067376634f,0.02474393539914235f,0.011784158597413285f,0.08070837752145013f,0.028837804117072482f,0.013566889108044652f,0.0892198151329586f,0.033326247190342086f,0.015396259291835603f,0.09780925050843237f,0.03821642380815322f,0.017268324012382f,0.10644014290558228f,0.043430971060037175f,0.01915497115064957f,0.11516213940852653f,0.04869376377787085f,0.02104088811842354f,0.12396774807241907f,0.05400341174839573f,0.02291932904231329f,0.13283301770235545f,0.05937446015473189f,0.0247676534264442f,0.14177532141983173f,0.06482608437087267f,0.02654913857103515f,0.1508303085225832f,0.07034982700960057f,0.02826358681647732f,0.15995532205231544f,0.07595041067967645f,0.02988842186530461f,0.16914872242514994f,0.08163782650304266f,0.03140401564457005f,0.17841319040234854f,0.08742545856288386f,0.03277327931516748f,0.18776813322972746f,0.09331557740511903f,0.033979154409542016f,0.19719580690248006f,0.09930765162558787f,0.0350072967471369f,0.2066795437361335f,0.105402262424314f,0.0358364665010164f,0.21620646706666896f,0.11160571005993522f,0.03644452450798942f,0.22577039634735835f,0.11791676181366892f,0.03681366238931068f,0.2353533948315683f,0.12433690815307286f,0.03692755679726379f,0.2449402569352536f,0.130861794027333f,0.0367738251284066f,0.25450501190842256f,0.13748923050603104f,0.036346007036179566f,0.2640284014268764f,0.14420943125208832f,0.03564202294027663f,0.2734764666562528f,0.1510154327209675f,0.034667061792608964f,0.282818469890769f,0.1578973488336295f,0.03343446973536865f,0.29202361159436785f,0.16483784305323396f,0.031964803318426126f,0.3010500119503021f,0.1718271766829396f,0.03028588882044497f,0.3098670941619176f,0.17884663206387366f,0.028428459100255845f,0.31843622115540393f,0.1858839310036252f,0.026431189198060315f,0.3267275455559629f,0.19291386085475284f,0.02434997053816215f,0.33470171605932664f,0.19992220593124488f,0.02223366886218197f,0.3423326727997547f,0.20689679435120226f,0.020130048116207458f,0.34960418789627484f,0.21382118808651815f,0.018085948264879845f,0.3564954212152869f,0.22068623935476184f,0.016147650423161998f,0.3630028398603805f,0.22748122267063475f,0.014355534805442475f,0.3691191047604013f,0.2342036115077947f,0.012742013345224536f,0.3748502927327437f,0.24084627534795414f,0.011335284339993059f,0.3802005626251185f,0.2474096049888127f,0.01015735838738674f,0.3851833827603531f,0.25389243988976723f,0.009226759851114728f,0.3898140016153026f,0.2602964252633522f,0.00854893129141548f,0.3941060822693162f,0.2666282702790725f,0.00812889690971778f,0.3980799156919193f,0.2728881836012367f,0.00796777080577268f,0.4017525543519634f,0.2790829843835597f,0.008062521199618247f,0.4051419276648617f,0.2852151444728726f,0.008411444340647998f,0.4082654255391529f,0.2912905872823367f,0.00900696229306688f,0.41114096865607663f,0.2973134150580773f,0.009846182094486277f,0.4137873198524331f,0.30328762487316085f,0.010914425884934178f,0.41621716948105103f,0.30922074689328993f,0.012204295605290823f,0.4184468468595007f,0.3151147196399223f,0.013704677218996394f,0.4204882743475308f,0.32097639646999004f,0.015407335075544953f,0.422354544204609f,0.3268065495346441f,0.017302460620371732f,0.4240567802998516f,0.33261022038705f,0.01938047884522835f,0.4256065016808433f,0.33839282001375265f,0.021633450732188426f,0.4270121205544362f,0.34415378701257915f,0.024052574290460142f,0.4282829405343027f,0.3498988918271099f,0.02662930883000891f,0.4294266653701039f,0.35562785044944545f,0.029355256148443618f,0.43044874030607067f,0.3613465905418141f,0.032224412611494564f,0.4313587270615997f,0.3670534821768672f,0.03523065438609198f,0.4321591931513488f,0.3727525670853749f,0.03836576069038865f,0.43285640046483465f,0.37844680040449136f,0.04159569733823424f,0.4334555349565255f,0.3841355329782064f,0.0447886211961958f,0.43396029607797f,0.3898229228233572f,0.04794477857573524f,0.43437397708144543f,0.3955064846961563f,0.0510632756323768f,0.4347003809417806f,0.40118951085798776f,0.05414611240426622f,0.43494369314594333f,0.4068728001787072f,0.05719222228379529f,0.4351048143031719f,0.4125561659556076f,0.060199930687386674f,0.43518581888017777f,0.4182439569546356f,0.06317128301685854f,0.4351887312325765f,0.4239333136155822f,0.06610722021585971f,0.4351145439646258f,0.4296274038777328f,0.06900863885365052f,0.43496624880705104f,0.4353241735726936f,0.07187617069246008f,0.4347439247528073f,0.4410274908705713f,0.07471304329299233f,0.4344484704151144f,0.4467336685929623f,0.0775180640738074f,0.4340820535925406f,0.4524463832839706f,0.08029430493651157f,0.4336434847073313f,0.4581629519603063f,0.08304144488312268f,0.43313601580946376f,0.4638857742577249f,0.08576376530326196f,0.432560538831606f,0.4696151520071452f,0.08846085367401663f,0.4319148958388981f,0.4753481825505309f,0.09113464993118832f,0.43120041645865226f,0.4810895258391375f,0.09378748211627656f,0.4304157480526235f,0.4868354565595182f,0.09641864471026834f,0.4295633056730111f,0.4925879104627671f,0.09903140399175724f,0.4286426514306487f,0.4983447389473999f,0.1016274927015326f,0.42765327068757697f,0.5041074631513899f,0.10420806397082304f,0.4265959365704464f,0.5098765596563365f,0.10677546051510424f,0.42547036940392857f,0.5156490293025526f,0.1093311312138189f,0.4242771408904675f,0.5214267691404487f,0.11187572305489003f,0.4230146529372112f,0.5272086759775392f,0.11441057366051095f,0.42168356166547744f,0.532994806491818f,0.1169395877882898f,0.4202831796072165f,0.5387841442602808f,0.1194627941731479f,0.4188152468547687f,0.5445752162374711f,0.12198117471380225f,0.4172784035251669f,0.550371411794682f,0.1244988719881559f,0.4156732538227701f,0.5561677070769875f,0.12701492382448096f,0.4139986102201316f,0.5619661289256851f,0.12953294280559788f,0.4122546278213512f,0.5677646289876168f,0.1320539831226978f,0.4104422052829828f,0.5735651914497292f,0.1345794432111999f,0.4085604229551985f,0.579363842280652f,0.13711144234826714f,0.40661024958520653f,0.5851601782714847f,0.13965174564712465f,0.4045922065996268f,0.5909574913960268f,0.14220219189728348f,0.4025047796622188f,0.5967498429288651f,0.14476378100450754f,0.4003480187832498f,0.6025402031913455f,0.14733996311413705f,0.3981218371952534f,0.6083246722235879f,0.14993155878890302f,0.3958283734200633f,0.614106176674807f,0.15253988593803774f,0.39346646772201055f,0.6198788025701961f,0.1551679529737472f,0.3910373196621276f,0.6256450195747313f,0.15781633943679754f,0.3885413427073905f,0.6314052760627059f,0.1604878602151773f,0.3859778883860768f,0.637153671587794f,0.16318436257638144f,0.3833482435611563f,0.6428950257611473f,0.16590822465679886f,0.3806510890190323f,0.6486225037023093f,0.16865910560313505f,0.377887791334078f,0.6543400237812237f,0.17144146614273248f,0.37505894495304937f,0.6600427220068441f,0.17425592860145306f,0.37216699339894543f,0.6657309854853277f,0.17710424966773933f,0.3692102159546668f,0.671404337541839f,0.1799902187518531f,0.3661898345073855f,0.6770589159382479f,0.182912273416154f,0.3631083856894559f,0.6826975017706738f,0.1858760680408481f,0.3599642927599365f,0.6883152752699135f,0.18888000215377848f,0.3567601555176547f,0.6939130936485612f,0.1919297404645243f,0.3534953157761731f,0.6994880189268522f,0.19502263596715352f,0.3501704524961132f,0.7050385790436794f,0.19816353086298186f,0.3467877021993576f,0.7105661060122954f,0.20135422039004186f,0.3433461843544269f,0.7160668362113027f,0.20459503779649424f,0.3398506288928686f,0.7215395551311659f,0.20788868051000065f,0.33629822944329f,0.7269824482031878f,0.21123646007205965f,0.3326937750819364f,0.7323962498882949f,0.21464109358472336f,0.32903540290632194f,0.7377762325604813f,0.21810179955417053f,0.3253269506392982f,0.7431240483143292f,0.22162240347649032f,0.3215664902244945f,0.7484350306113071f,0.2252020416267312f,0.31775890499434706f,0.7537094301237001f,0.2288447108656564f,0.31390318773099685f,0.7589465700904082f,0.2325513179438209f,0.3099993007197641f,0.7641427572948342f,0.23632095843046075f,0.3060511850445163f,0.7692985739778287f,0.24015863799868106f,0.30205775317511674f,0.7744103608994938f,0.24406133999228075f,0.2980219987542918f,0.7794786898294086f,0.2480331020679257f,0.2939437813724841f,0.784498961427673f,0.25207289596277377f,0.2898270979003709f,0.7894723227633637f,0.2561838131658038f,0.2856708304197678f,0.7943970958045569f,0.2603648353176219f,0.2814768234287905f,0.7992696028577546f,0.2646179852095233f,0.27724708700681994f,0.8040913179125936f,0.26894337006009084f,0.2729813802962653f,0.8088576856149733f,0.2733398560423693f,0.2686817498061221f,0.8135691155681037f,0.2778117319366859f,0.2643498423217051f,0.8182230524153231f,0.2823557514806422f,0.259985751711718f,0.8228187030680942f,0.2869730612436132f,0.25559221349785954f,0.8273560901963927f,0.2916649238384595f,0.2511669292787484f,0.8318308105503314f,0.2964310422772357f,0.24671393625869514f,0.8362441082220449f,0.3012708228202772f,0.2422317977788539f,0.8405915992051021f,0.3061839140531289f,0.2377225424355419f,0.8448754218830034f,0.3111718447857894f,0.2331856496191531f,0.8490922820892888f,0.3162321293857926f,0.228623159368174f,0.8532412230472858f,0.3213650843479945f,0.22403563620206654f,0.8573221666233893f,0.3265710944674797f,0.21942058959604985f,0.8613319022896901f,0.33184860342904376f,0.2147810099094718f,0.8652713962071545f,0.33719833283248074f,0.2101141748022448f,0.8691385576642853f,0.3426165911885936f,0.20542406206466107f,0.8729332123141957f,0.34810624963200004f,0.20070581716761515f,0.876653351754305f,0.3536635338530963f,0.195963342516276f,0.8802978200638566f,0.3592889341611301f,0.19119290796426572f,0.8838684350097556f,0.3649829259277349f,0.18639467968336265f,0.8873602825148426f,0.3707406901646968f,0.18156845817161654f,0.890776984152265f,0.37656622385933053f,0.17671201671816958f,0.894114774529188f,0.3824525761944193f,0.1718270686976245f,0.8973741852518462f,0.388403855880936f,0.16690816145867332f,0.9005544630688599f,0.3944150503711016f,0.1619568111176787f,0.9036543550377922f,0.4004867447796602f,0.15697061628964454f,0.906674539526174f,0.4066185337569268f,0.1519460056654663f,0.9096133361541572f,0.4128073699691127f,0.14688407634155784f,0.9124721600806108f,0.419054437663371f,0.14177960453650024f,0.9152474734642112f,0.4253546069506915f,0.1366313113758783f,0.9179415758304628f,0.43171113773754055f,0.13143638978707065f,0.920553002660083f,0.4381188215811716f,0.12619301932742935f,0.9230823502164308f,0.44457832435590056f,0.12089804008534748f,0.9255281583142996f,0.4510903508911226f,0.115545643733264f,0.9278900885067396f,0.4576496172590345f,0.11013577030361796f,0.9301682735211648f,0.4642584989903114f,0.1046620258358874f,0.9323634323492204f,0.470912661585951f,0.09912286212434912f,0.9344736508745052f,0.4776135317653246f,0.09351101566047268f,0.9364987505473671f,0.4843577227601096f,0.0878266775099058f,0.9384406797563624f,0.4911467021518643f,0.08206088279212748f,0.94029638756102f,0.4979770365588466f,0.07621335449370882f,0.9420673342348812f,0.5048474657057321f,0.07027747421356159f,0.943752835993572f,0.5117607949728211f,0.06424712981979414f,0.9453530086488392f,0.5187105010375903f,0.05812067587313205f,0.9468675926635921f,0.5257001589178919f,0.05189145509095276f,0.9482957691429188f,0.5327252159488806f,0.04556084013205488f,0.9496381978647562f,0.5397882797332773f,0.039112475813857225f,0.9508941688983458f,0.5468847469930004f,0.0329036576052177f,0.9520630153802706f,0.5540144303588769f,0.02723224588536035f,0.953145915117146f,0.561180174976782f,0.02211629990383893f,0.9541413109174116f,0.5683763342018563f,0.017577226490710227f,0.9550496977801548f,0.5756055764119579f,0.013636710338596076f,0.9558695965052876f,0.5828632209633262f,0.010318691317706429f,0.9566033661162672f,0.5901529783933634f,0.007647565236664239f,0.9572476931570264f,0.5974691178771274f,0.005653645257891558f,0.9578046847922824f,0.6048145165947891f,0.004365191753448543f,0.9582725531896952f,0.6121880260521443f,0.003815358887601479f,0.9586519941780332f,0.6195869032693965f,0.004038403733227168f,0.9589432359293408f,0.6270139069375642f,0.0050725965725081324f,0.9591452081062959f,0.6344632304037671f,0.006957730473759643f,0.9592570195286264f,0.6419386757168154f,0.009736863918784179f,0.959279647369708f,0.6494364125427067f,0.013456640602907992f,0.9592132245352332f,0.656956325469712f,0.01816672798060043f,0.9590557212865652f,0.6645013350837338f,0.023921375872797367f,0.9588093894625144f,0.6720645507841265f,0.03077795215314844f,0.9584721753486866f,0.6796498174204002f,0.03879846557918752f,0.9580453442160642f,0.6872532448327471f,0.04754586688050162f,0.9575278692108288f,0.6948766727751612f,0.05639000971583572f,0.9569201366476526f,0.7025161872712079f,0.0653272271581101f,0.95622222088975f,0.7101726957961991f,0.07436297745193415f,0.9554351276542836f,0.7178471084562286f,0.08350080196415507f,0.9545585022368179f,0.7255344740989906f,0.09274251876146856f,0.95359012489534f,0.7332386758872225f,0.10209834873587548f,0.9525337955331732f,0.7409547279741554f,0.11157046121752068f,0.9513894499647296f,0.7486824818249713f,0.1211710906915065f,0.9501579394841028f,0.7564199576460588f,0.13090499817983486f,0.9488415317142022f,0.7641650350600445f,0.14078130910755238f,0.9474413345042012f,0.771919616464573f,0.15081436111263724f,0.94595966334309f,0.7796766610808825f,0.16100576302194788f,0.9443984072166398f,0.7874390279521042f,0.17137304382242416f,0.9427621029656729f,0.7952006533320073f,0.18192158708279615f,0.941049699471568f,0.8029643911682793f,0.19266999933666f,0.9392640761941338f,0.8107231422033202f,0.20362775766846952f,0.9374158407051084f,0.8184757458260994f,0.21480475650403494f,0.9355108618569462f,0.8262179886623421f,0.22621716723273394f,0.9335594549301716f,0.8339427520154467f,0.23787212746051176f,0.931563069969426f,0.8416528004815411f,0.2497953642795853f,0.9295295956536241f,0.8493389652326568f,0.2620016582307053f,0.927483113892923f,0.8569948963180799f,0.2745012512443162f,0.9254408066615336f,0.8646104422830595f,0.28729978853917554f,0.9234033328223992f,0.8721861539859546f,0.3004428353086889f,0.9214094735748476f,0.8797058578924303f,0.3139219503891981f,0.9194909179295936f,0.8871561145803522f,0.3277441860383484f,0.9176568416750964f,0.8945363755302687f,0.34196084494650136f,0.9159695644129056f,0.9018205155640128f,0.3565351604857609f,0.9144604035423052f,0.9089997667171816f,0.3714975395911552f,0.91318592603147f,0.9160520672097188f,0.3868282485461301f,0.9122023005193108f,0.922961587897254f,0.40251902826986985f,0.9115712707791952f,0.929705490864802f,0.4185354153286997f,0.9113589610617012f,0.9362663045373388f,0.4348346331778865f,0.911626841657708f,0.9426265331991612f,0.4513649463134185f,0.9124352877263222f,0.9487690293791992f,0.4680385459173532f,0.9138280045063198f,0.9546871530908342f,0.484787690701253f,0.9158401268012052f,0.9603724114950956f,0.5015022040367715f,0.9184812096938036f,0.9658321994102153f,0.5181206698543415f,0.9217499769357954f,0.971070575662174f,0.5345393044762992f,0.925624725190333f,0.9761026354466212f,0.5507027927528236f,0.9300724918488174f,0.9809461333524748f,0.5665954408447937f,0.93505290022375f,0.9856188564238579f,0.5821026104689219f,0.9405209651098092f,0.9901403564326076f,0.5972696263274475f,0.9464291624522388f,0.9945258421274078f,0.6120801460623929f,0.9527369951544834f,0.9987960656048062f,0.6265350434944937f,0.9593995073432898f,1.0029632847639067f,0.6406264784793351f}, {0.0f,0.0f,0.0f,0.0124985675549f,0.00376710208302f,0.00210713336054f,0.0249971351098f,0.00753420416604f,0.00421426672107f,0.037495702664699995f,0.0113013062491f,0.00632140008161f,0.0492581088139f,0.0150684083321f,0.008428533442149999f,0.0594165938721f,0.0188355104151f,0.010535666802700001f,0.0684471877541f,0.0226026124981f,0.012642800163200001f,0.0766363089899f,0.0263697145811f,0.014749933523799998f,0.0841678787653f,0.0301368166641f,0.016857066884299998f,0.09116808551160001f,0.0339039187472f,0.0189642002448f,0.09772782947889999f,0.0376710208302f,0.021071333605400002f,0.103915087694f,0.0414146679499f,0.023178466965900004f,0.109782226099f,0.044974293467f,0.025285600326400003f,0.11537057005200001f,0.048364772838199996f,0.027392733687f,0.120713391496f,0.0516062923368f,0.0294998670475f,0.125837934386f,0.054715381023000005f,0.031607000408099994f,0.130854584645f,0.0576505684502f,0.0337193392032f,0.13604190193f,0.06025908266f,0.0358534453625f,0.14140417663f,0.0625542306901f,0.0380109209821f,0.146927404384f,0.0645557290717f,0.0401922566139f,0.152598730721f,0.0662787359665f,0.0423389683918f,0.158389635306f,0.0677778384763f,0.0444339133373f,0.164222599234f,0.0692401824912f,0.046456344033f,0.170089016456f,0.07068488518239999f,0.04840869742790001f,0.17598792054499998f,0.0721120614255f,0.050295820716699996f,0.18191842184300003f,0.0735218171731f,0.0521219928396f,0.187879698899f,0.07491424994469999f,0.05389101229169999f,0.19387099103f,0.0762894492712f,0.0556062682658f,0.19989159185400002f,0.07764749709919999f,0.057270798828499996f,0.20594084362799997f,0.0789884681608f,0.0588873389018f,0.212018132302f,0.0803124303088f,0.0604583601449f,0.218122883178f,0.0816194448242f,0.0619861043454f,0.224254557086f,0.0829095666958f,0.0634726115611f,0.23041264701699998f,0.0841828448758f,0.06491974398580001f,0.236596675144f,0.0854393225134f,0.0663292063043f,0.242806190184f,0.0866790371676f,0.0677025631472f,0.249040765054f,0.08790202100120001f,0.0690412541328f,0.255299994784f,0.0891083009582f,0.0703466068909f,0.26158349465999997f,0.09029789892519999f,0.07161984838819999f,0.267890898557f,0.0914708318774f,0.07286211481869999f,0.274221857454f,0.0926271120122f,0.0740744602728f,0.280576038088f,0.0937667468691f,0.0752578643655f,0.286953121761f,0.09488973943880001f,0.07641323897f,0.293352803248f,0.0959960882604f,0.077541434183f,0.299774789817f,0.0970857875089f,0.0786432436236f,0.30621880034800003f,0.0981588270728f,0.0797194091565f,0.312684564525f,0.0992151926224f,0.0807706251109f,0.319171134912f,0.10025502620999999f,0.0818055914811f,0.32567691454500003f,0.10127863170399999f,0.0828412386963f,0.332201791044f,0.102285972873f,0.0838775716485f,0.33874565083999997f,0.103277010196f,0.08491459706849999f,0.34530837961099997f,0.104251700818f,0.0859523214123f,0.351889862505f,0.105209998536f,0.0869907508714f,0.358489984344f,0.10615185378900001f,0.08802989138360001f,0.36510862979900005f,0.107077213638f,0.0890697486419f,0.37174568354500004f,0.10798602174699999f,0.09011032810450001f,0.378401030403f,0.10887821835599999f,0.0911516350032f,0.385074555462f,0.109753740249f,0.0921936743521f,0.391766144185f,0.110612520723f,0.09323645095560001f,0.398475682511f,0.111454489549f,0.09427996941639999f,0.40520305693400005f,0.112279572927f,0.09532423414249999f,0.411948154577f,0.113087693439f,0.0963692493548f,0.418710863264f,0.113878770002f,0.0974150190938f,0.425491071569f,0.114652717803f,0.09846154722599999f,0.43228866887100004f,0.115409448248f,0.09950883745030001f,0.439103545394f,0.116148868889f,0.100556893304f,0.44593559224600005f,0.116870883353f,0.101605718168f,0.452784701449f,0.117575391273f,0.10265531527499999f,0.45965076597099996f,0.118262288197f,0.10370568771f,0.46653367974f,0.118931465509f,0.10475683842f,0.473433337672f,0.11958281033f,0.105808770216f,0.480349635682f,0.120216205427f,0.10686148578f,0.48728247069700004f,0.12083152909900001f,0.10791498766799999f,0.49423174066500003f,0.121428655073f,0.108969278312f,0.5011973445669999f,0.12200745238199999f,0.110024360032f,0.5081791824140001f,0.122567785241f,0.11108023502999999f,0.515177155256f,0.123109512914f,0.112136905403f,0.522191165184f,0.12363248957299999f,0.113194373138f,0.529221115323f,0.12413656414999999f,0.114252640123f,0.536266909839f,0.12462158018f,0.115311708147f,0.543328453928f,0.12508737563300001f,0.116371578905f,0.550405653818f,0.125533782739f,0.11743225399700001f,0.557498416761f,0.125960627799f,0.118493734936f,0.564606651025f,0.126367730992f,0.11955602315f,0.571730265893f,0.126754906165f,0.12061911998199999f,0.57886917165f,0.12712196061f,0.121683026697f,0.586023279576f,0.127468694834f,0.12274774447899998f,0.59319250194f,0.127794902308f,0.123813274439f,0.600376751988f,0.12810036920999998f,0.124879617617f,0.607575943936f,0.128384874145f,0.125946774979f,0.614789992957f,0.128648187853f,0.127014747427f,0.6220188151739999f,0.128890072894f,0.128083535793f,0.6292623276470001f,0.129110283326f,0.12915314085000001f,0.636520448365f,0.129308564349f,0.13022356330499998f,0.643793096232f,0.129484651934f,0.131294803809f,0.65108019106f,0.129638272432f,0.132366862954f,0.658381653554f,0.12976914215500002f,0.133439741276f,0.665697405303f,0.129876966926f,0.13451343925599998f,0.67302736877f,0.129961441609f,0.135587957327f,0.680371467279f,0.13002224960699998f,0.136663295866f,0.686819633179f,0.132061139735f,0.13726053577199998f,0.690577321338f,0.139841365035f,0.136421938331f,0.694333453232f,0.147366583596f,0.13555415609899998f,0.69808804669f,0.154670100683f,0.134656431336f,0.701841119323f,0.161779110231f,0.133727968564f,0.705592688522f,0.16871613022f,0.132767931803f,0.709342771462f,0.17550003480300003f,0.131775441528f,0.713091385109f,0.18214681303599997f,0.130749571337f,0.716838546221f,0.188670137478f,0.12968934428099999f,0.720584271352f,0.195081797593f,0.12859372881699999f,0.724328576854f,0.20139203510899997f,0.127461634333f,0.7280714788830001f,0.207609807026f,0.126291906198f,0.7318129934000001f,0.21374299438299998f,0.125083320255f,0.7355531361740001f,0.21979856979499998f,0.123834576701f,0.7392919227879999f,0.225782733244f,0.122544293247f,0.743029368637f,0.23170102312899998f,0.12121099746899999f,0.746765488936f,0.237558407857f,0.11983311822f,0.7505002987199999f,0.24335936194200003f,0.118408975956f,0.754233812848f,0.249107929694f,0.116936771817f,0.7579660460050001f,0.254807778863f,0.115414575242f,0.761697012704f,0.26046224609499996f,0.11384030988800001f,0.765426727292f,0.266074375673f,0.112211737544f,0.7691552039480001f,0.27164695270100003f,0.11052643969699999f,0.77288245669f,0.277182531676f,0.10878179629899999f,0.7766084993729999f,0.28268346119000004f,0.106974961217f,0.7803333456970001f,0.288151905399f,0.105102833683f,0.784057009203f,0.293589862731f,0.103162024955f,0.787779503281f,0.29899918227600003f,0.101148819133f,0.791500841169f,0.304381578185f,0.0990591268518f,0.795221035957f,0.309738642359f,0.0968884301939f,0.7989401005869999f,0.315071855678f,0.0946317167004f,0.8026580478569999f,0.32038259796199997f,0.0922833997264f,0.8063748904260001f,0.325672156837f,0.08983722152339998f,0.810090640808f,0.33094173563900003f,0.08728613424630001f,0.813805311384f,0.336192460496f,0.08462215241789998f,0.817518914394f,0.341425386671f,0.0818361680288f,0.821231461948f,0.346641504267f,0.0789177160322f,0.824942966022f,0.351841743372f,0.0758546729705f,0.828653438464f,0.35702697870300004f,0.0726328638943f,0.8323628909909999f,0.362198033804f,0.06923554106180001f,0.8360713351960001f,0.36735568486500003f,0.065642679414f,0.839778782546f,0.372500664176f,0.061830003621300006f,0.843485244386f,0.377633663287f,0.05776761042780001f,0.847190731941f,0.382755335881f,0.0534179600786f,0.8508952563140001f,0.387866300395f,0.048732844522199996f,0.854598828494f,0.39296714243f,0.0436486154689f,0.858301459351f,0.398058416948f,0.0380796196638f,0.8620031596430001f,0.403140650294f,0.032290438683500004f,0.865703940014f,0.408214342055f,0.026411833944599995f,0.869403810997f,0.413279966769f,0.020443258979599997f,0.870620849202f,0.42025968812700004f,0.0196207043343f,0.871622154301f,0.42734161938799997f,0.0192250591085f,0.87260643877f,0.434372519047f,0.018839607349200004f,0.87357362974f,0.441355114094f,0.0184645577569f,0.8745236535169999f,0.448291940683f,0.0181001190322f,0.875456435563f,0.45518536158999995f,0.0177464998757f,0.876371900482f,0.462037581704f,0.017403908988f,0.8772699720039999f,0.468850661816f,0.0170725550697f,0.878150572968f,0.47562653091000007f,0.0167526468212f,0.8790136253059999f,0.482366997164f,0.0164443929432f,0.879859050028f,0.489073757822f,0.0161480021363f,0.880686767202f,0.49574840805799997f,0.0158636831011f,0.8814966959389999f,0.502392448976f,0.015591644538f,0.8822887543729999f,0.509007294828f,0.0153320951477f,0.883062859646f,0.51559427955f,0.0150852436308f,0.883818927886f,0.522154662692f,0.0148512986878f,0.884556874192f,0.528689634804f,0.0146304690194f,0.885276612612f,0.5352003223390001f,0.014422963326f,0.885978056125f,0.541687792133f,0.0142289903082f,0.886661116621f,0.548153055488f,0.0140487586667f,0.8873257048790001f,0.554597071914f,0.013882477102000001f,0.8879717305500001f,0.561020752565f,0.013730354314700002f,0.888599102131f,0.56742496338f,0.0135925990053f,0.889207726947f,0.573810527989f,0.013469419874499998f,0.889797511128f,0.580178230371f,0.013361025622700001f,0.890368359587f,0.586528817319f,0.0132676249506f,0.890920175994f,0.592863000709f,0.0131894265588f,0.891452862757f,0.5991814596f,0.0131266391478f,0.891966320994f,0.605484842178f,0.0130794714182f,0.892460450511f,0.611773767557f,0.013048132070599999f,0.8929351497750001f,0.6180488274539999f,0.0130328298055f,0.893390315889f,0.624310587737f,0.0130337733235f,0.893825844568f,0.6305595898749999f,0.0130511713252f,0.894241630107f,0.636796352279f,0.013085232511200001f,0.8946375653569999f,0.643021371558f,0.013136165582f,0.895013541696f,0.649235123688f,0.0132041792383f,0.8953694490009999f,0.6554380651039999f,0.0132894821805f,0.895705175617f,0.6616306337210001f,0.0133922831093f,0.896020608327f,0.6678132498899999f,0.013512790725200001f,0.8963156323200001f,0.673986317291f,0.0136512137289f,0.896590131162f,0.6801502237729999f,0.013807760820799998f,0.8968439867620001f,0.686305342141f,0.013982640701600001f,0.897077079337f,0.692452030892f,0.0141760620719f,0.8972892873780001f,0.698590634912f,0.014388233632100001f,0.8974804876200001f,0.704721486124f,0.014619364082899999f,0.897650554997f,0.7108449041079999f,0.0148696621249f,0.897799362612f,0.716961196674f,0.0151393364586f,0.897926781699f,0.7230706604100001f,0.0154285957846f,0.8980326815799999f,0.729173581195f,0.0157376488035f,0.898116929626f,0.735270234683f,0.0160667042159f,0.8981793912200001f,0.741360886763f,0.0164159707222f,0.89821992971f,0.7474457939910001f,0.0167856570232f,0.8982384063700001f,0.7535252040000001f,0.0171759718194f,0.898234680352f,0.7595993558840001f,0.0175871238113f,0.898208608643f,0.7656684805699999f,0.0180193216995f,0.898160046015f,0.771732801162f,0.0184727741846f,0.898088844982f,0.777792533271f,0.0189476899672f,0.8979948557459999f,0.783847885327f,0.019444277747900002f,0.8978779261480001f,0.7898990588770001f,0.019962746227200002f,0.8977379016150001f,0.795946248863f,0.020503304105599997f,0.897574625107f,0.801989643893f,0.0210661600838f,0.897387937062f,0.808029426494f,0.0216515228624f,0.897177675337f,0.8140657733519999f,0.022259601141900003f,0.896943675152f,0.820098855544f,0.0228906036229f,0.8966857690279999f,0.8261288387540001f,0.023544739005999998f,0.8985530078489999f,0.831233249825f,0.06826100781569999f,0.9028903792290001f,0.835257282572f,0.127173791555f,0.907153202363f,0.8392904152249999f,0.16936596405f,0.9113400634639999f,0.8433327452760001f,0.20455362557700002f,0.9154495631820001f,0.847384366892f,0.235764224808f,0.9194803145479999f,0.8514453709909999f,0.264391329421f,0.923430940957f,0.85551584532f,0.291204862021f,0.927300074214f,0.859595874528f,0.316679372942f,0.931086352617f,0.8636855402410001f,0.341129701928f,0.934788419075f,0.867784921127f,0.364776124263f,0.938404919262f,0.8718940929729999f,0.38777907412100004f,0.941934499804f,0.8760131287430001f,0.41025913859200003f,0.945375806484f,0.880142098652f,0.43230927905f,0.948727482467f,0.884281070227f,0.454002666231f,0.951988166539f,0.8884301083689999f,0.47539790299999996f,0.955156491362f,0.892589275414f,0.49654262079899997f,0.9582310817219999f,0.896758631197f,0.51747602533f,0.96121055279f,0.9009382331069999f,0.538230741549f,0.9640935083679999f,0.9051281361460001f,0.55883417852f,0.9668785391320001f,0.909328392984f,0.579309557384f,0.969564220862f,0.913539054016f,0.599676697971f,0.9721491126430001f,0.917760167413f,0.619952629264f,0.9746317550560001f,0.9219917791760001f,0.640152069209f,0.9770106683249999f,0.9262339331899999f,0.660287806118f,0.979284350439f,0.930486671268f,0.6803710050120001f,0.981451275229f,0.934750033206f,0.700411455972f,0.9835098904f,0.9390240568279999f,0.720417777213f,0.9854586155060001f,0.943308778035f,0.7403975824410001f,0.987295839871f,0.9476042308499999f,0.760357619768f,0.9890199204369999f,0.951910447464f,0.780303887791f,0.990629179549f,0.956227458278f,0.8002417331899999f,0.9921219026450001f,0.960555291948f,0.8201759332430001f,0.993496335869f,0.9648939754279999f,0.840110765987f,0.9947506835750001f,0.9692435340090001f,0.8600500701389999f,0.995883105733f,0.973603991364f,0.879997296537f,0.99689171521f,0.9779753695799999f,0.899955552467f,0.997774574927f,0.982357689206f,0.9199276400409999f,0.9985296948710001f,0.986750969284f,0.939916089534f,0.9991550289650001f,0.9911552273890001f,0.9599231884430001f,0.999648471753f,0.995570479665f,0.9799510069290001f,1.0f,1.0f,1.0f
+    },
+    {
+        0.0f,0.0f,0.0f,0.017991146422900003f,0.0009200649235139999f,0.0156157177605f,0.0359573389421f,0.0017855321075899999f,0.0319686152223f,0.052717504614f,0.00252887372289f,0.0485944475368f,0.06645936690569999f,0.00322312300316f,0.0629875526327f,0.07837890807539999f,0.00380156463246f,0.0758728198312f,0.08901944585010002f,0.00426758967588f,0.0877407961252f,0.0985268713734f,0.00475880811033f,0.098703029916f,0.107066588617f,0.00523157681353f,0.109516758625f,0.11449012489f,0.00579292222525f,0.120833508104f,0.120982700612f,0.00638735196241f,0.132756772803f,0.126759943415f,0.0069730244716199995f,0.14502000078900001f,0.132114155782f,0.007512258382419999f,0.157034897655f,0.137058036252f,0.00806004196545f,0.168740910277f,0.141649687793f,0.0086020586774f,0.18018158636599999f,0.145897662326f,0.009187402070350001f,0.191306061277f,0.149911730845f,0.009699306680750001f,0.20228018041999998f,0.153699794586f,0.0101651632923f,0.21306910708000001f,0.157229049852f,0.0106966648307f,0.223525579263f,0.160604333941f,0.0111474503315f,0.23383850216999996f,0.163796762742f,0.0116231373033f,0.243863642246f,0.166848448357f,0.0121134415658f,0.253685702911f,0.169819141386f,0.0126114478884f,0.26363818667f,0.172738904452f,0.013045630466799999f,0.273838903033f,0.17553987372299998f,0.0136060166319f,0.284035834989f,0.178292710791f,0.014093826366700002f,0.294472726772f,0.180967301212f,0.0146031444225f,0.305018418317f,0.183568732274f,0.0151304937855f,0.315665343158f,0.18612659789500002f,0.0155638077972f,0.326540250824f,0.188597679232f,0.016112598419f,0.337367139559f,0.19101527950999997f,0.0166661392128f,0.348269803401f,0.19338747153000002f,0.0172191510076f,0.35923893975599996f,0.195737845385f,0.0176485273411f,0.370407572449f,0.198044113384f,0.018180747506500002f,0.381482230849f,0.200323209309f,0.018816828319f,0.392447152667f,0.20260874331f,0.0193080320772f,0.40358286922799996f,0.204899513261f,0.019767508720099997f,0.414734365044f,0.207203487335f,0.020318509756f,0.42573979336900003f,0.20954069120299998f,0.020829322049200003f,0.43673700905699997f,0.211923423866f,0.0214286969089f,0.447561641367f,0.214365276602f,0.0218414047174f,0.458511304496f,0.216884168073f,0.022476171082099997f,0.469109681584f,0.219485130374f,0.022911157438400002f,0.479815538029f,0.222188067595f,0.023425825940400002f,0.49030580992600004f,0.224998688279f,0.0238728708752f,0.500730803846f,0.227935385801f,0.0244003542471f,0.51092309047f,0.230997962124f,0.024855742978500003f,0.521037675039f,0.234203821386f,0.0253947921946f,0.530907891495f,0.237546372868f,0.0258598394405f,0.540693759978f,0.241027046685f,0.026247579899400002f,0.55039511939f,0.244584484796f,0.0267693440447f,0.55991166928f,0.2481175218f,0.0271274399767f,0.569636574265f,0.24989475832f,0.0278631362956f,0.581398671653f,0.251525497218f,0.0284219720302f,0.593483256725f,0.252639757507f,0.029041628899500002f,0.606152585473f,0.252696088643f,0.0296094624578f,0.620118665558f,0.251496133219f,0.0303909420016f,0.635313719985f,0.248691606409f,0.0311547780908f,0.652158617486f,0.243878137586f,0.0320817761944f,0.670731330201f,0.23639962754000002f,0.0330706616692f,0.691424807068f,0.225267244623f,0.0341252602766f,0.7146108666329999f,0.20888985034200003f,0.0353832411736f,0.7406363862800001f,0.18419045863800002f,0.0368327602548f,0.7701406520819999f,0.14386341454600002f,0.038418444306699995f,0.804039880295f,0.0746219638915f,0.0463798951671f,0.836091603811f,0.051929680457699995f,0.0798472986086f,0.829374357579f,0.0394515282547f,0.107189845036f,0.816977772194f,0.0384497825314f,0.130572940256f,0.8006196394949999f,0.0376049080062f,0.150475345512f,0.783946103531f,0.0366209160453f,0.167937309627f,0.767178191635f,0.036209539045199995f,0.183593242522f,0.750243007035f,0.035009786269f,0.197766483702f,0.733612552706f,0.0341816624553f,0.21076933108100002f,0.717059181752f,0.0338207279497f,0.222790052386f,0.700632749393f,0.0330602243752f,0.233954587651f,0.684623431271f,0.0320054471262f,0.244386794417f,0.669049419546f,0.0311985935446f,0.25418969331700003f,0.653810060935f,0.0306931539107f,0.263438905191f,0.6389273191579999f,0.0300923241428f,0.272196250421f,0.624527084061f,0.0294532825196f,0.280518738048f,0.610613760314f,0.0288266759948f,0.288454796768f,0.597189242503f,0.028258218845900004f,0.296045951278f,0.5842532441480001f,0.027358108648900002f,0.303331720628f,0.571895679741f,0.0270296599782f,0.31033700714699997f,0.559924872093f,0.0264448001694f,0.317097231559f,0.548516147786f,0.0256345140463f,0.323638967672f,0.537652138675f,0.025492765819700003f,0.329971350849f,0.527160032449f,0.024740170013099996f,0.3361344932f,0.517263775647f,0.0242644107155f,0.342132084344f,0.507792879869f,0.0240860234118f,0.347978814433f,0.49873824455f,0.023770524347099996f,0.35369797069f,0.49015319357500003f,0.0233150400472f,0.35930342735799997f,0.48201715982199994f,0.022709995372599996f,0.364807798254f,0.474309413085f,0.022415522700799998f,0.37021053997000003f,0.46695688834f,0.021951068805999997f,0.375533039074f,0.459997929716f,0.021785245042699998f,0.380771775449f,0.453366342385f,0.0214109796297f,0.38594788015500003f,0.447093322224f,0.021308362673299996f,0.391055299678f,0.44111961202399996f,0.0209409639731f,0.396115195541f,0.43546940626900005f,0.0208032233745f,0.40111931542199997f,0.43009041141800003f,0.020326797125099997f,0.406089062357f,0.424999913907f,0.0200232189201f,0.41101424043799994f,0.420152583348f,0.0198728054352f,0.415899382496f,0.415537874164f,0.0201730386418f,0.42075220537299995f,0.410977892936f,0.0204269128371f,0.425601519313f,0.40633093586999997f,0.020648268745f,0.430446943879f,0.40159664050300004f,0.0208515375655f,0.435288096472f,0.3967746856f,0.0210517249569f,0.4401245929f,0.391864797966f,0.0212643917258f,0.44495604798f,0.386866760343f,0.021505630995400003f,0.449782076175f,0.381780420593f,0.021792041613500002f,0.45460229226000004f,0.376605702372f,0.0221406975556f,0.459416312029f,0.37134261756399994f,0.0225691130702f,0.46422375303499996f,0.365991280754f,0.0224259562301f,0.469044405085f,0.36052987785499996f,0.0230580790144f,0.47383779193099995f,0.354998747537f,0.0231350670787f,0.47864410868699997f,0.349350101263f,0.0233439830722f,0.48344280665899997f,0.343606556264f,0.0237028430224f,0.488233513041f,0.337768885308f,0.0235081053068f,0.493037171153f,0.331794869685f,0.024209893148099997f,0.497811021508f,0.325767842034f,0.0243709990305f,0.502597537708f,0.319598395605f,0.024730286079f,0.507375060891f,0.313331518429f,0.0245356175547f,0.512165384947f,0.306908018078f,0.0253294179759f,0.516924115438f,0.300448540314f,0.025576676730900003f,0.521695388489f,0.293828618023f,0.0252539941421f,0.5264794575689999f,0.287037203286f,0.0259778918909f,0.53123082017f,0.280225448366f,0.0261331721134f,0.535994757369f,0.27324116767400003f,0.0265399835536f,0.540748306125f,0.26616608007f,0.0263475128744f,0.54551458806f,0.25890964136100003f,0.026400511586099996f,0.5502703191f,0.251565875201f,0.026708122111499998f,0.555015282958f,0.244143954883f,0.0272784546489f,0.5597492873650001f,0.23665484777600002f,0.0271978964134f,0.564496121687f,0.22898932204f,0.027362887814599997f,0.5692319597209999f,0.22127137955499998f,0.027777471233f,0.573956676974f,0.213518485566f,0.0284444308696f,0.578670178335f,0.205751413881f,0.028384328963700002f,0.583396856402f,0.19784223614f,0.0285476524053f,0.588112448966f,0.18995533481199997f,0.0289331144526f,0.5928169448489999f,0.182123550818f,0.0285115878732f,0.597535139873f,0.17420506477000003f,0.029317533151700003f,0.602217649212f,0.166595804546f,0.029277909690400002f,0.606914200514f,0.15897569628900002f,0.0294140830833f,0.611600149328f,0.151584202296f,0.029720091464999997f,0.616275666466f,0.144488804029f,0.0301894614322f,0.620940969293f,0.137765522615f,0.030815365661999997f,0.62559632569f,0.131498161336f,0.0304596782636f,0.630267391981f,0.12551028490900001f,0.0313631945057f,0.6349039514029999f,0.12041327253699999f,0.031242877616f,0.639557176769f,0.115752028275f,0.0312235432034f,0.644202264115f,0.11187543455f,0.031299156354f,0.648839821508f,0.108851437083f,0.0314184113993f,0.653477231267f,0.106176991919f,0.031535472753400004f,0.658121193765f,0.103322272239f,0.0325849440421f,0.662732380265f,0.10250629337000002f,0.0324964623759f,0.667373624657f,0.10131367318199999f,0.0322681389383f,0.672026774497f,0.09970463821270001f,0.0330969943958f,0.676670839949f,0.0976295890318f,0.0326757224886f,0.6813534245209999f,0.09429961297700001f,0.0337229991078f,0.686019956215f,0.09035971130699999f,0.033906420328900004f,0.6907178321999999f,0.0849030421592f,0.0334148100448f,0.695443242051f,0.07765277688510001f,0.0337425640662f,0.700167125211f,0.0686502617223f,0.0338207022037f,0.704910801625f,0.056740271177f,0.0338743964261f,0.709670103135f,0.0403918706963f,0.058733599314900006f,0.713922150001f,0.034390706820699994f,0.08742958138599999f,0.7178914983329999f,0.0346787868923f,0.111915239145f,0.7217992682150001f,0.0348193031422f,0.134123009678f,0.7256434216609999f,0.034803745147699995f,0.155347319743f,0.729397564257f,0.035371253705500004f,0.175139323574f,0.733108825242f,0.03502508946419999f,0.194611949127f,0.736726660802f,0.0352592690818f,0.213537378156f,0.740273862291f,0.0353178375278f,0.23233260534699998f,0.743724886122f,0.035968753389f,0.250513531308f,0.747126350622f,0.0356584622917f,0.268691545657f,0.750429081044f,0.035938694373699995f,0.286648031957f,0.753655779162f,0.0360251867582f,0.30462964428100003f,0.75678196096f,0.0367156189835f,0.32225527245200003f,0.75985331015f,0.0363983061994f,0.339942737704f,0.762822451337f,0.036684313670300006f,0.35751599238799997f,0.765711772711f,0.036760521877800004f,0.37499478500100003f,0.768520445955f,0.03661958726469999f,0.39253668911399997f,0.7712256456290001f,0.0370972727808f,0.40999430756399996f,0.773849358368f,0.0373587187295f,0.42737966438600006f,0.776391028603f,0.0373970537272f,0.44470245939499997f,0.778850167358f,0.0372056550069f,0.46207959141000005f,0.7812056023f,0.037651709852999996f,0.479394440292f,0.783478485675f,0.0378707411309f,0.496653275128f,0.785668578238f,0.037856699162400004f,0.513861188457f,0.787775695975f,0.0376038180367f,0.531105711117f,0.7897806400059999f,0.0380113165991f,0.548295506065f,0.7917033548639999f,0.038184710776999996f,0.565434007525f,0.793543877596f,0.0381188749731f,0.5825911959669999f,0.795284740401f,0.038737631302199996f,0.599630010222f,0.796961716746f,0.0381868400101f,0.6166820495040001f,0.798540479427f,0.038327533972699994f,0.6336821068330001f,0.800038845211f,0.038226918714199996f,0.650680153913f,0.801441969801f,0.038841776359800004f,0.667576491456f,0.802781137769f,0.0382551020787f,0.684464833385f,0.804027141476f,0.0383929993421f,0.7013338115350001f,0.80518254406f,0.0392768048004f,0.718108996552f,0.80627465814f,0.038941355259f,0.734858978841f,0.807278722198f,0.0393626753188f,0.7515248991399999f,0.808219025281f,0.0385431989517f,0.7681800140530001f,0.8090642018549999f,0.0395103035394f,0.784981358809f,0.8097796552509999f,0.0389863784715f,0.8022058529970001f,0.8102699963259999f,0.0386723126896f,0.8198700490620001f,0.810514774272f,0.039587106598699996f,0.837964828179f,0.81050820531f,0.0407349043605f,0.8564957604919999f,0.810239070669f,0.0411533487121f,0.875477677159f,0.809684424733f,0.041848599778699994f,0.8949159257420001f,0.808826947298f,0.0428263881801f,0.914815073355f,0.807648343923f,0.0440862680746f,0.935192099963f,0.8061279113519999f,0.0447502798991f,0.9555538065289999f,0.8043291392380001f,0.06738231609709999f,0.9643886421690001f,0.8047460502860001f,0.252819656976f,0.969264675898f,0.806479603483f,0.355213883146f,0.972679058372f,0.808946322667f,0.427168888808f,0.975344341181f,0.8119109170919999f,0.482670299842f,0.977488081448f,0.8152639825329999f,0.527978708904f,0.9793252066920001f,0.81891153223f,0.5660448050119999f,0.980856323558f,0.822821418552f,0.599002327948f,0.982248942f,0.8269254258490001f,0.627796037997f,0.9834291778550001f,0.8312222396439999f,0.653523321339f,0.984581461736f,0.835645803538f,0.676478782658f,0.9855303208130001f,0.8402290853319999f,0.6974790648420001f,0.986439873531f,0.844915604813f,0.716603137535f,0.987284696794f,0.8497017593199999f,0.734192410112f,0.9880974797740001f,0.8545693999539999f,0.750435011311f,0.988784936521f,0.859536817855f,0.765642621436f,0.989434072415f,0.8645728128650001f,0.779848654014f,0.990132548978f,0.8696463974869999f,0.793083975622f,0.99072699097f,0.874795481798f,0.8056380470569999f,0.9912757574009999f,0.879998726453f,0.817531592142f,0.991837792113f,0.885234998583f,0.8287840677949999f,0.992337137685f,0.890521929857f,0.839541460436f,0.992819103751f,0.895843059048f,0.8498184551350001f,0.993184915734f,0.901223132537f,0.859755146555f,0.9937648480589999f,0.906565576495f,0.869111246785f,0.99415826293f,0.911981196241f,0.8782758180429999f,0.994542722176f,0.9174171683239999f,0.887132207634f,0.994951529848f,0.922861877404f,0.895689414959f,0.995417956347f,0.928304043215f,0.9039557832330001f,0.995805993584f,0.9337798170880001f,0.912062161581f,0.996137505124f,0.939281131187f,0.920014523899f,0.996434662824f,0.944800045145f,0.927818426777f,0.9968989074510001f,0.950279714893f,0.935356647591f,0.997197218778f,0.955810300499f,0.9428787189710001f,0.997527495307f,0.961336154068f,0.950266270209f,0.9979112170119999f,0.9668504867319999f,0.9575230453650001f,0.998179199116f,0.9723977961929999f,0.964774153088f,0.9985368445209999f,0.977921894815f,0.9719005570520001f,0.998809261864f,0.983468565536f,0.979026563766f,0.999205022643f,0.9889818427570001f,0.986032578677f,0.999545232406f,0.994508167954f,0.993042540059f,1.0f,1.0f,1.0f
+
+    },
+    {
+        0.0f,0.0f,0.0f,0.018058465396400002f,0.000875333094297f,0.0158976099199f,0.0358446051389f,0.00174980699631f,0.0328047687723f,0.0522636499211f,0.0025563525718699997f,0.050049172046f,0.06595973062999999f,0.00316349374618f,0.0653730066735f,0.0775624917691f,0.00377981404209f,0.0790449485008f,0.08770618802069999f,0.00440417464933f,0.0916119341295f,0.0968151684079f,0.0049730454100600004f,0.103540340406f,0.104722118803f,0.00554119297516f,0.11599716230999998f,0.111370425372f,0.00620204611365f,0.129261639264f,0.116946513903f,0.006848121858689999f,0.14355298095799998f,0.12153532394899999f,0.007593629006810001f,0.15827138042199998f,0.12560895274599998f,0.00824751570213f,0.17294829718899998f,0.129116053052f,0.00897961657386f,0.187269898157f,0.132214226742f,0.0096321961349f,0.20148045969300002f,0.134903090655f,0.0103013238918f,0.215382661822f,0.137265159112f,0.0109393452309f,0.229013577036f,0.139353106158f,0.011562120307200002f,0.242296207746f,0.141228114177f,0.012182397830399998f,0.25515394106899997f,0.14295845935100002f,0.012810606763500001f,0.267511245419f,0.144632880203f,0.013374506779399999f,0.279430555591f,0.146311489693f,0.013856517071799999f,0.290986319638f,0.147985088662f,0.014435025153600001f,0.30226197245700004f,0.14970770896299998f,0.0149622508005f,0.313501070947f,0.15150937992f,0.015514977713799998f,0.324530460502f,0.15343108664400001f,0.0159967974703f,0.335465361707f,0.155506784775f,0.0164900613996f,0.346132556517f,0.15777230058f,0.0169915776714f,0.356506192632f,0.16025809158399998f,0.0174996875948f,0.366565290795f,0.162987579688f,0.0180145188299f,0.376294691501f,0.165965739373f,0.0184349864779f,0.38584077770400005f,0.16918579305999998f,0.018871069297799997f,0.39506689055f,0.172195422223f,0.0192910917595f,0.40464415821000005f,0.173403705169f,0.0198899561305f,0.41645113278800006f,0.173861044438f,0.0205279465565f,0.42913879139600003f,0.172365040039f,0.0212232568816f,0.44407754921099996f,0.168164143733f,0.022081702120099998f,0.461671910382f,0.15995214689200002f,0.0230256208157f,0.482744336407f,0.145360278175f,0.024227775932999997f,0.508035519531f,0.118833057349f,0.0257460880057f,0.538929643256f,0.059215889274900006f,0.0275724315861f,0.577825147355f,0.027433370317599997f,0.055696941466199995f,0.573236714621f,0.026610656155599998f,0.0822574864661f,0.552306416405f,0.025323468955f,0.10275920217599999f,0.531339176589f,0.024581636462599997f,0.11971852687699999f,0.510265658092f,0.023508081587900003f,0.134168958766f,0.489764126426f,0.022623365942599998f,0.14677633269200002f,0.46981996249799995f,0.0215547333709f,0.157937576357f,0.450747864098f,0.020699329830499998f,0.16795289040999997f,0.43248433008299997f,0.0198967605312f,0.177031635917f,0.41516848254f,0.0192303955284f,0.18533855557f,0.398809925251f,0.0182867657147f,0.193007654432f,0.383581078901f,0.017612817051200002f,0.20014041547699998f,0.36927508329000003f,0.017014848798499998f,0.206823590156f,0.355950858235f,0.01653121034f,0.213126962847f,0.343569825475f,0.0159456483298f,0.219116312109f,0.332154924132f,0.015511386092299998f,0.224833901477f,0.321585900677f,0.014978811725399999f,0.230330436099f,0.311871103985f,0.0145817257042f,0.23563231808600002f,0.302898649355f,0.0140334508886f,0.24078128980900002f,0.294664914215f,0.0138286995253f,0.24578010019000002f,0.287029297837f,0.013384075107200001f,0.250676352423f,0.280026567675f,0.0132072369246f,0.255463537867f,0.27353693590100003f,0.012963478530099999f,0.260172611311f,0.267551533796f,0.01275628057f,0.264812412048f,0.261996025246f,0.0130214917563f,0.269399802958f,0.2564471131f,0.013275541578999999f,0.27397345914099996f,0.250784410116f,0.0135425133387f,0.278532552634f,0.245007619091f,0.0135126185665f,0.283093979942f,0.23911193732900002f,0.0138754745934f,0.287621824293f,0.23310130526f,0.013980997516000002f,0.29215106711800004f,0.22696086059599999f,0.014190400945600002f,0.296663130996f,0.22069629667100002f,0.0145304863521f,0.301157183524f,0.21430915847800003f,0.0146578033742f,0.305651900786f,0.207766742968f,0.014953406351900001f,0.310127653915f,0.20109459745799999f,0.0150555796267f,0.314603774411f,0.194248861492f,0.0153591079009f,0.319060012372f,0.187269706758f,0.015481614838900001f,0.323516334236f,0.18010245883200002f,0.0158320447878f,0.327951947526f,0.17280379686f,0.0160049395278f,0.332387434875f,0.16530897460400001f,0.0164236283821f,0.336801542341f,0.157694092826f,0.016657244420399998f,0.341215451176f,0.149885785417f,0.0166866212706f,0.34562941453199997f,0.141882953872f,0.0169565382695f,0.350021626219f,0.133799378962f,0.016994284899799997f,0.354414106962f,0.125553279371f,0.0172595614171f,0.358784804599f,0.117294366612f,0.0177521102088f,0.36313367696500004f,0.109085926267f,0.017956123680099998f,0.36748358135799997f,0.10085354282400001f,0.0178318333763f,0.371835170462f,0.09265963234850001f,0.0184069150757f,0.376143131735f,0.084959044806f,0.0186127396493f,0.380453888309f,0.07755167092710001f,0.0184053173473f,0.384768452568f,0.0705745923215f,0.0188750949188f,0.389041323406f,0.0646649213733f,0.018887726719f,0.39332016766000005f,0.059585575389f,0.018989240106099998f,0.39758314905000003f,0.0557613443634f,0.019155370293599996f,0.401836286524f,0.0530002144934f,0.0195726975527f,0.406079284257f,0.0509259961276f,0.0200605678256f,0.41031532947899996f,0.0499257009659f,0.019933992650500002f,0.414589372335f,0.047897334140099994f,0.0201198564949f,0.41887075826700004f,0.04469494276930001f,0.0202744574697f,0.423173801671f,0.039760474672f,0.020678514617299998f,0.427489232063f,0.032996469891599996f,0.0209912092412f,0.43183050815499996f,0.024250178731400004f,0.0349723445527f,0.43571010284100004f,0.021003309972200002f,0.0568879927503f,0.439264307473f,0.0211778526846f,0.0767479553296f,0.442716711831f,0.0212184127479f,0.0957871138849f,0.446040085637f,0.021486782986400003f,0.11404484267100001f,0.44925275321f,0.021609207331500003f,0.1318104468f,0.45235073162199996f,0.0215750256604f,0.149470028439f,0.455308584136f,0.021761699914500002f,0.167053697064f,0.45812383476999996f,0.0221744111686f,0.184236193933f,0.460837074209f,0.0220219560819f,0.201440136403f,0.46340308318499995f,0.022086341869099996f,0.21865425261799998f,0.465820835436f,0.022373913789f,0.23574290226f,0.468109977517f,0.022477547036000002f,0.252842941703f,0.47024955625f,0.0228083262913f,0.269736892289f,0.472278349938f,0.0225244412619f,0.286742684767f,0.474138353353f,0.0228931346916f,0.303649814199f,0.475868629906f,0.0230684253847f,0.320467842222f,0.47746917857f,0.0230437190067f,0.33720349255099996f,0.47894019618600003f,0.0228130802442f,0.35397751849200004f,0.48025060826399996f,0.0232755329729f,0.37059809134000005f,0.481450683186f,0.0230858051611f,0.38717825679799994f,0.482510801383f,0.0231497853431f,0.40370139153499995f,0.483434884226f,0.023480468515f,0.42008733990599995f,0.48425111724799996f,0.023141666200700003f,0.436435341245f,0.484925361047f,0.023555276293700002f,0.452650469281f,0.485494834886f,0.0232942153785f,0.468883184798f,0.485908993767f,0.0236001754074f,0.48565148479099995f,0.48601265707900004f,0.023419886257299998f,0.503100421156f,0.485724148423f,0.024240176158599996f,0.521220887411f,0.48502097767499996f,0.0251173152935f,0.540032006724f,0.483860737672f,0.0260755646195f,0.559561329233f,0.482196106284f,0.0266793371253f,0.579802776807f,0.47997623333000006f,0.0278581744233f,0.600792297647f,0.477141258357f,0.028716149772399998f,0.622520503646f,0.47363135725500005f,0.0297062017882f,0.644983693461f,0.46938021635899996f,0.0308169231996f,0.668171263815f,0.464315602087f,0.0320261521594f,0.692064697472f,0.45835888974399996f,0.0332998807226f,0.716681934384f,0.45140229749899996f,0.0341928395725f,0.7418988995450001f,0.443392751729f,0.0354509848299f,0.767707726612f,0.434210395185f,0.0365931233555f,0.7939891186429999f,0.42378120354099996f,0.037902022134800005f,0.820656977454f,0.41200156363000007f,0.039248180042500004f,0.8476125158560001f,0.398759158615f,0.0404892124241f,0.8746742422210001f,0.38398939181999997f,0.04178030419f,0.9017077994730001f,0.3675822286f,0.042949926405800004f,0.928415174219f,0.349573677887f,0.044424104252f,0.954545481796f,0.329991381325f,0.0462442266644f,0.9573013443709999f,0.33609564364f,0.10421581297900001f,0.9591174926260001f,0.34322000200099995f,0.142599846881f,0.960545553503f,0.350720062574f,0.17223852748f,0.9616946896969999f,0.358466816251f,0.196508613818f,0.962634746497f,0.36637856625500004f,0.21695690863000003f,0.963502228203f,0.374312232539f,0.23436999222399998f,0.964204505811f,0.38235618254799997f,0.24948301355900002f,0.964827517636f,0.39042044189400005f,0.262649173343f,0.965339521585f,0.39852906948100003f,0.274255453153f,0.965913832211f,0.40649656131399997f,0.28484251263300003f,0.96660239999f,0.414004257475f,0.299477112022f,0.967424045438f,0.421156664498f,0.31595297826499996f,0.96824448919f,0.428060999844f,0.334462757437f,0.96925265516f,0.434535360032f,0.355085184136f,0.970255192862f,0.440768027519f,0.377386202244f,0.971400042438f,0.44666745096999994f,0.40071500981299996f,0.972494124682f,0.4524136941f,0.42479050884999997f,0.973682286844f,0.457914388253f,0.449200759171f,0.974414038449f,0.46361271767800005f,0.47401216261600004f,0.9747062520569999f,0.469501541549f,0.498966836373f,0.974924322329f,0.47531196257200004f,0.5237342872290001f,0.975203228437f,0.480956638403f,0.54818506882f,0.975442392691f,0.486524219209f,0.5723036278320001f,0.975773849859f,0.49192905913699997f,0.596023932317f,0.9759748034810001f,0.497346258966f,0.619361874151f,0.976292748071f,0.50260450503f,0.642262831649f,0.976501574349f,0.507878764176f,0.664729824313f,0.9767289113650001f,0.513085162851f,0.6867407604160001f,0.97698331835f,0.518226546562f,0.7082899559579999f,0.9772725006299999f,0.523305979649f,0.7293745397600001f,0.977484712652f,0.528412817876f,0.749962011074f,0.977744873326f,0.533464543877f,0.770070436051f,0.9779400111409999f,0.538551100225f,0.7896558164390001f,0.978193692765f,0.543590248049f,0.8087531397859999f,0.978391929879f,0.548672499326f,0.8273059777480001f,0.9786568878949999f,0.55371563806f,0.845365210221f,0.978873730796f,0.558810727197f,0.862861541059f,0.9791631966959999f,0.5638753437590001f,0.8798610494939999f,0.979409739754f,0.569001080649f,0.8962813946610001f,0.9796153949680001f,0.574192828239f,0.9121082288019999f,0.979898784777f,0.5793675187899999f,0.9274295732439999f,0.980143480993f,0.584617488548f,0.942143565762f,0.980428066947f,0.589851275613f,0.956547783518f,0.9806167300409999f,0.595083361726f,0.9710772006f,0.976468972257f,0.603553543446f,0.981004612526f,0.9648058479520001f,0.617572961021f,0.981714065719f,0.954176640447f,0.63062291544f,0.9823004555700001f,0.944661555225f,0.642731489199f,0.982953296597f,0.93603016065f,0.654130441515f,0.98346643582f,0.928360756143f,0.6648172405899999f,0.984025523979f,0.9214349788770001f,0.674983455547f,0.984397251418f,0.915422862482f,0.684546158619f,0.984911837263f,0.910111827943f,0.693674997183f,0.985320520096f,0.905571512771f,0.702343324126f,0.985812122819f,0.901601922737f,0.710699556338f,0.986111780451f,0.8983568718520001f,0.718653125358f,0.986567164996f,0.8956423445839999f,0.7263409954760001f,0.986888961516f,0.893521424939f,0.733727097324f,0.98726874984f,0.891890609952f,0.740883482457f,0.9875661752700001f,0.8908078534280001f,0.747773239593f,0.987979040431f,0.890095471452f,0.7545144814040001f,0.988182442835f,0.889886635315f,0.7610171457439999f,0.988552456896f,0.890008347904f,0.767394121872f,0.988753387001f,0.890588245246f,0.773556490152f,0.989172395609f,0.8914579467329999f,0.779612719008f,0.989464223106f,0.892598870881f,0.785570447984f,0.989646351742f,0.8940631843790001f,0.791389242524f,0.989928948566f,0.895831034216f,0.7970767481309999f,0.990336671576f,0.897743294753f,0.802733631489f,0.990505508391f,0.899923597083f,0.808271006036f,0.990843972112f,0.9022150723129999f,0.8137872556219999f,0.9909846661f,0.904742107148f,0.8191935117679999f,0.991342528157f,0.907348759393f,0.82458640736f,0.991548838469f,0.910053474568f,0.829929090757f,0.9918886243799999f,0.91215063232f,0.835494723044f,0.9921146238449999f,0.914160237339f,0.841082517336f,0.9923880661700001f,0.915831974426f,0.8467923602140001f,0.992574850985f,0.917215934105f,0.85258644636f,0.9928525746119999f,0.918343892019f,0.858451030257f,0.993234361466f,0.919146606622f,0.864444157258f,0.9933642488940001f,0.919795086974f,0.870464877501f,0.993647603398f,0.920331532515f,0.8764968618369999f,0.9941021320110001f,0.9207174933589999f,0.882585544639f,0.9943942612459999f,0.9210197452869999f,0.888704026867f,0.99456239769f,0.921345666753f,0.8947817520200001f,0.994968207377f,0.921656719823f,0.900877705629f,0.995139198777f,0.9220365833640001f,0.9069316285519999f,0.9954233926039999f,0.9224997720460001f,0.912939395805f,0.99582750669f,0.923064404149f,0.9189223523360001f,0.996076872691f,0.923761725923f,0.92485561751f,0.996340836063f,0.9246478367640001f,0.930744885063f,0.996384013549f,0.9257491717f,0.9365672216680001f,0.996378230633f,0.9270589345389999f,0.942312986593f,0.9964670712720001f,0.9286197534080001f,0.947979495601f,0.99655468174f,0.9304505616230001f,0.95355995137f,0.9966691854180001f,0.932620027359f,0.959041009932f,0.996745582305f,0.935159261824f,0.9644107087809999f,0.996829850665f,0.938030872519f,0.9696706896549999f,0.9970449960400001f,0.941454329103f,0.9747737451080001f,0.9972004007619999f,0.945484787756f,0.979697280271f,0.997379626347f,0.9502538230959999f,0.9844008257060001f,0.997607280329f,0.955994721365f,0.988815107679f,0.9978827357819999f,0.963064049972f,0.992831571263f,0.998213482077f,0.971958790258f,0.996291841212f,0.998647995176f,0.983626075978f,0.9988935025700001f,0.999215372248f,1.0f,1.0f,1.0f
+    },
+
+    {
+        0.301961f, 0.047059f, 0.090196f,
+        0.3244797567486321f, 0.04518239041438145f, 0.0826897493142621f,
+        0.3469985134972642f, 0.04330578082876291f, 0.07518349862852419f,
+        0.3695172702458963f, 0.04142917124314436f, 0.06767724794278629f,
+        0.39203602699452844f, 0.03955256165752581f, 0.060170997257048386f,
+        0.4153246758645797f, 0.04229508539323472f, 0.05420443073367733f,
+        0.43878176790024087f, 0.0460482201189405f, 0.048574728645118664f,
+        0.46223885993590197f, 0.049801354844646276f, 0.04294502655656f,
+        0.4856959519715631f, 0.053554489570352054f, 0.03731532446800133f,
+        0.5085516281292881f, 0.06392319895335406f, 0.03348987001325075f,
+        0.5310704364835226f, 0.07799745417475062f, 0.030675018968971443f,
+        0.5535892448377571f, 0.0920717093961472f, 0.027860167924692127f,
+        0.5761080531919917f, 0.10614596461754378f, 0.025045316880412817f,
+        0.5964619516652722f, 0.11978723786274958f, 0.022663447812324287f,
+        0.6142893416123747f, 0.13292320940271982f, 0.020786880449471398f,
+        0.6321167315594771f, 0.14605918094269005f, 0.01891031308661851f,
+        0.6499441215065795f, 0.15919515248266028f, 0.01703374572376562f,
+        0.6680360595281275f, 0.17259567209707607f, 0.014892630286467139f,
+        0.6868017331566564f, 0.1866699273184727f, 0.012077779242187815f,
+        0.7055674067851853f, 0.2007441825398693f, 0.009262928197908492f,
+        0.724333080413714f, 0.2148184377612659f, 0.006448077153629168f,
+        0.7423298406606399f, 0.2289888071553629f, 0.003825454454750602f,
+        0.7535892448377571f, 0.244001346058186f, 0.002887170773324161f,
+        0.7648486490148745f, 0.2590138849610091f, 0.0019488870918977193f,
+        0.7761080531919917f, 0.27402642386383214f, 0.0010106034104712786f,
+        0.787367457369109f, 0.2890389627666552f, 7.231972904483746e-05f,
+        0.8046886092128975f, 0.3101132493361495f, 0.0025978918571448154f,
+        0.82251599916f, 0.33169377400895766f, 0.0054127429014241436f,
+        0.8403433891071024f, 0.35327429868176585f, 0.008227593945703472f,
+        0.8581707790542048f, 0.374854823354574f, 0.0110424449899828f,
+        0.8697203985455797f, 0.39434275787547296f, 0.014554826084898522f,
+        0.8791032353598441f, 0.41310843150400184f, 0.01830796081060429f,
+        0.8884860721741085f, 0.43187410513253066f, 0.022061095536310055f,
+        0.8978689089883728f, 0.4506397787610594f, 0.0258142302620158f,
+        0.9040771689092902f, 0.4672890677940236f, 0.030096461136612775f,
+        0.907830303634996f, 0.4823016066968467f, 0.03478787954374499f,
+        0.9115834383607018f, 0.4973141455996698f, 0.03947929795087719f,
+        0.9153365730864075f, 0.5123266845024927f, 0.04417071635800937f,
+        0.9183683833178213f, 0.52697856115817f, 0.04994412150657952f,
+        0.9202449506806742f, 0.5410528163795666f, 0.05745039095799106f,
+        0.9221215180435272f, 0.5551270716009632f, 0.0649566604094026f,
+        0.9239980854063801f, 0.5692013268223598f, 0.07246292986081407f,
+        0.9260668811146336f, 0.5834678103891573f, 0.08054588434842794f,
+        0.928881732158913f, 0.5984803492919804f, 0.09086700484411882f,
+        0.9316965832031924f, 0.6134928881948034f, 0.1011881253398097f,
+        0.9345114342474716f, 0.6285054270976265f, 0.1115092458355005f,
+        0.9373024908480955f, 0.6435417604441052f, 0.12194933854946921f,
+        0.9391790582109483f, 0.6594925830283547f, 0.13696187745229227f,
+        0.9410556255738012f, 0.6754434056126042f, 0.1519744163551153f,
+        0.942932192936654f, 0.6913942281968535f, 0.16698695525793825f,
+        0.9448087602995069f, 0.7073450507811031f, 0.1819994941607614f,
+        0.9466853276623598f, 0.7217085849186791f, 0.19701203306358447f,
+        0.9485618950252127f, 0.7357828401400757f, 0.21202457196640753f,
+        0.9504384623880655f, 0.7498570953614723f, 0.2270371108692305f,
+        0.9523150297509184f, 0.763931350582869f, 0.24204964977205368f,
+        0.9548168074353629f, 0.7780056058042656f, 0.259563029961243f,
+        0.9576316584796423f, 0.7920798610256622f, 0.27832870358977185f,
+        0.9604465095239215f, 0.8061541162470587f, 0.29709437721830056f,
+        0.9632613605682009f, 0.8202283714684554f, 0.3158600508468295f,
+        0.9651626587727875f, 0.8333890738501594f, 0.33690960657459007f,
+        0.966100942454214f, 0.8455867617087031f, 0.3603666986102511f,
+        0.9670392261356404f, 0.8577844495672468f, 0.383823790645912f,
+        0.9679775098170669f, 0.8699821374257906f, 0.40728088268157325f,
+        0.9689157934984933f, 0.881314797730031f, 0.430161289681032f,
+        0.9698540771799198f, 0.8906976345442954f, 0.45174181435384014f,
+        0.9707923608613461f, 0.9000804713585597f, 0.47332233902664833f,
+        0.9717306445427726f, 0.9094633081728241f, 0.49490286369945613f,
+        0.972788836840555f, 0.9186063277543768f, 0.5162435711395528f,
+        0.9746654042034079f, 0.9261125972057883f, 0.5359475284495081f,
+        0.9765419715662608f, 0.9336188666571998f, 0.5556514857594633f,
+        0.9784185389291136f, 0.9411251361086114f, 0.5753554430694187f,
+        0.9802951062919665f, 0.9486314055600229f, 0.5950594003793739f,
+        0.9803921568627451f, 0.9543581582193601f, 0.6129838408972549f,
+        0.9803921568627451f, 0.9599878603079188f, 0.6308112308443573f,
+        0.9803921568627451f, 0.9656175623964773f, 0.6486386207914594f,
+        0.9803921568627451f, 0.9712472644850361f, 0.666466010738562f,
+        0.981834805851329f, 0.9754343175850109f, 0.6864573741685402f,
+        0.9837113732141819f, 0.9791874523107167f, 0.707099615159922f,
+        0.9855879405770348f, 0.9829405870364225f, 0.7277418561513037f,
+        0.9874645079398877f, 0.9866937217621282f, 0.7483840971426855f,
+        0.9876593664170776f, 0.9882352941176471f, 0.7712609376122769f,
+        0.9866819875822583f, 0.9882352941176471f, 0.7956954084827571f,
+        0.9857046087474391f, 0.9882352941176471f, 0.820129879353237f,
+        0.9847272299126199f, 0.9882352941176471f, 0.8445643502237176f,
+        0.9470546069594605f, 0.9470546069594605f, 0.886023613981023f,
+        0.932598204007831f, 0.932598204007831f, 0.8661460599225324f,
+        0.9181418010562015f, 0.9181418010562015f, 0.8462685058640418f,
+        0.9036853981045719f, 0.9036853981045719f, 0.8263909518055512f,
+        0.8892289951529424f, 0.8891685524480406f, 0.8067551685666682f,
+        0.8747725922013129f, 0.8738086243119342f, 0.790491715246085f,
+        0.8603161892496836f, 0.858448696175828f, 0.774228261925502f,
+        0.8458597862980538f, 0.8430887680397214f, 0.7579648086049184f,
+        0.8314033833464243f, 0.827728839903615f, 0.7417013552843352f,
+        0.8189164803943001f, 0.8136819117671789f, 0.7274074019632574f,
+        0.8071706529961011f, 0.8001290340000261f, 0.7138545241961046f,
+        0.7954248255979021f, 0.7865761562328735f, 0.700301646428952f,
+        0.7836789981997031f, 0.7730232784657207f, 0.6867487686617992f,
+        0.7719331708015041f, 0.7601684649191511f, 0.6738939551152296f,
+        0.7601873434033053f, 0.7484226375209523f, 0.6621481277170308f,
+        0.7484415160051061f, 0.7366768101227531f, 0.6504023003188316f,
+        0.736695688606907f, 0.7249309827245541f, 0.6386564729206325f,
+        0.7249914254294559f, 0.7131851553263551f, 0.6269522097431814f,
+        0.7141491232157339f, 0.7014393279281561f, 0.6161099075294594f,
+        0.7033068210020117f, 0.6896935005299571f, 0.6052676053157372f,
+        0.6924645187882896f, 0.6779476731317582f, 0.5944253031020151f,
+        0.6816222165745675f, 0.6662018457335592f, 0.583583000888293f,
+        0.6707799143608455f, 0.6538183968196791f, 0.572740698674571f,
+        0.6599376121471232f, 0.6411690442370032f, 0.5618983964608487f,
+        0.649095309933401f, 0.6285196916543273f, 0.5510560942471265f,
+        0.6382530077196789f, 0.6158703390716513f, 0.5402137920334044f,
+        0.627080551879819f, 0.6035511401151131f, 0.5293714898196822f,
+        0.6153347244816201f, 0.5918053127169142f, 0.51852918760596f,
+        0.603588897083421f, 0.5800594853187151f, 0.5076868853922378f,
+        0.591843069685222f, 0.5683136579205161f, 0.49684458317851565f,
+        0.5801426137602113f, 0.5565678305223173f, 0.4860249667013876f,
+        0.5702038367309658f, 0.5448220031241182f, 0.4760861896721422f,
+        0.5602650597017206f, 0.5330761757259193f, 0.466147412642897f,
+        0.5503262826724753f, 0.5213303483277203f, 0.4562086356136517f,
+        0.54038750564323f, 0.5095845209295213f, 0.4462698585844065f,
+        0.5310674716455119f, 0.49845743656284963f, 0.43756856761821566f,
+        0.5220322198007434f, 0.48761513434912745f, 0.429436840957924f,
+        0.512996967955975f, 0.4767728321354053f, 0.4213051142976324f,
+        0.5039617161112067f, 0.46593052992168327f, 0.4131733876373409f,
+        0.4952377394084217f, 0.4557107779919283f, 0.40504166097704913f,
+        0.4871060127481301f, 0.4466755261471598f, 0.39690993431675753f,
+        0.4789742860878385f, 0.43764027430239133f, 0.3887782076564659f,
+        0.4708425594275469f, 0.42860502245762283f, 0.3806464809961743f,
+        0.4627146400196953f, 0.4195773851177345f, 0.3725223688407628f,
+        0.4554864385438805f, 0.4123491836419197f, 0.36619769254942486f,
+        0.4482582370680657f, 0.4051209821661049f, 0.35987301625808693f,
+        0.44103003559225096f, 0.39789278069029016f, 0.353548339966749f,
+        0.4338018341164362f, 0.3906645792144754f, 0.3472236636754111f,
+        0.4271734971879949f, 0.38343637773866085f, 0.3408989873840733f,
+        0.4208488208966568f, 0.37620817626284586f, 0.3345743110927352f,
+        0.4145241446053189f, 0.36897997478703115f, 0.3282496348013973f,
+        0.408199468313981f, 0.3617517733112164f, 0.3219249585100594f,
+        0.4015823953648134f, 0.3542311751775719f, 0.3158926788765512f,
+        0.39435419388899856f, 0.34609944851728025f, 0.31047152776969006f,
+        0.38712599241318374f, 0.3379677218569886f, 0.30505037666282897f,
+        0.379897790937369f, 0.32983599519669693f, 0.2996292255559679f,
+        0.37266958946155415f, 0.3217042685364053f, 0.2942080744491068f,
+        0.3663298419385024f, 0.31534944978163965f, 0.28789846938948277f,
+        0.36000516564716445f, 0.3090247734903017f, 0.28157379309814484f,
+        0.3536804893558265f, 0.3027000971989638f, 0.2752491168068069f,
+        0.34735581306448865f, 0.2963754209076259f, 0.26892444051546904f,
+        0.3410311367731507f, 0.2894697585530686f, 0.2625997642241311f,
+        0.3347064604818128f, 0.2822415570772539f, 0.2562750879327932f,
+        0.3283817841904749f, 0.27501335560143914f, 0.24995041164145532f,
+        0.32205710789913716f, 0.2677851541256246f, 0.24362573535011758f,
+        0.3154589134341233f, 0.2605569526498096f, 0.2373010590587795f,
+        0.3082307119583085f, 0.2533287511739948f, 0.23097638276744153f,
+        0.3010025104824937f, 0.24610054969818002f, 0.22465170647610358f,
+        0.2937743090066789f, 0.2388723482223652f, 0.21832703018476562f,
+        0.2865461075308641f, 0.2316441467465504f, 0.21200235389342767f,
+        0.27931790605504936f, 0.22441594527073566f, 0.20567767760208977f,
+        0.2720897045792346f, 0.2171877437949209f, 0.19935300131075184f,
+        0.2648615031034198f, 0.20995954231910616f, 0.19302832501941394f,
+        0.2576333016276051f, 0.2027313408432914f, 0.18670364872807604f,
+        0.2515293153099212f, 0.19606524694654207f, 0.18150318759486897f,
+        0.24610816420306011f, 0.18974057065520414f, 0.17698556167248475f,
+        0.24068701309619905f, 0.18341589436386624f, 0.17246793575010053f,
+        0.23526586198933802f, 0.1770912180725283f, 0.1679503098277163f,
+        0.2293354315034333f, 0.1707665417811904f, 0.16292340452628845f,
+        0.2221072300276185f, 0.16444186548985246f, 0.1565987282349505f,
+        0.2148790285518039f, 0.1581171891985147f, 0.15027405194361274f,
+        0.2076508270759889f, 0.15179251290717655f, 0.1439493756522746f,
+        0.2004226256001741f, 0.14546783661583862f, 0.13762469936093666f,
+        0.18894751779574606f, 0.15615051291796544f, 0.16786193042772124f,
+        0.17720867752123598f, 0.16788934416252893f, 0.20036965238654675f,
+        0.16546983724672593f, 0.1796281754070924f, 0.2328773743453723f,
+        0.15373099697221587f, 0.19136700665165585f, 0.2653850963041978f,
+        0.15262772233867725f, 0.20024081909319424f, 0.28314885340588275f,
+        0.15865382088226443f, 0.20719411419736025f, 0.29102924077196585f,
+        0.1646799194258516f, 0.21414740930152623f, 0.29890962813804894f,
+        0.1707060179694388f, 0.22110070440569224f, 0.30679001550413204f,
+        0.17673211651302598f, 0.22805399950985822f, 0.3146704028702152f,
+        0.18275821505661316f, 0.2350072946140242f, 0.3225507902362983f,
+        0.18878431360020034f, 0.2419605897181902f, 0.3304311776023814f,
+        0.19481041214378753f, 0.2489138848223562f, 0.3383115649684645f,
+        0.20077216965118935f, 0.25593152031923283f, 0.3462562936924703f,
+        0.20633476494215652f, 0.2633483140405224f, 0.35460018662892107f,
+        0.21189736023312403f, 0.2707651077618124f, 0.3629440795653723f,
+        0.21745995552409134f, 0.2781819014831022f, 0.37128797250182327f,
+        0.22302255081505867f, 0.2855986952043919f, 0.37963186543827426f,
+        0.228585146106026f, 0.2930154889256817f, 0.38797575837472525f,
+        0.23414774139699335f, 0.30043228264697147f, 0.39631965131117625f,
+        0.23971033668796066f, 0.30784907636826125f, 0.4046635442476273f,
+        0.245272931978928f, 0.31526587008955104f, 0.4130074371840783f,
+        0.2517797032050162f, 0.32331211443425467f, 0.4219807807439432f,
+        0.2587329473187253f, 0.33165600737070566f, 0.43125177289555533f,
+        0.2656861914324345f, 0.33999990030715666f, 0.44052276504716753f,
+        0.2726394355461436f, 0.3483437932436076f, 0.44979375719877973f,
+        0.27959267965985274f, 0.3566876861800586f, 0.4590647493503919f,
+        0.2865459237735619f, 0.3650315791165095f, 0.4683357415020041f,
+        0.29349916788727104f, 0.3733754720529605f, 0.4776067336536163f,
+        0.30045241200098016f, 0.3817193649894115f, 0.4868777258052285f,
+        0.3076087632044908f, 0.3900632579258622f, 0.4960471644119395f,
+        0.3154891065333614f, 0.39840715086231343f, 0.5048546069559714f,
+        0.32336944986223176f, 0.4067510437987644f, 0.5136620495000029f,
+        0.33124979319110215f, 0.4150949367352154f, 0.5224694920440346f,
+        0.33913013651997254f, 0.4234388296716664f, 0.5312769345880661f,
+        0.3470104798488429f, 0.4317827226081174f, 0.5400843771320978f,
+        0.3548908231777133f, 0.4401266155445684f, 0.5488918196761294f,
+        0.3627711665065837f, 0.4484705084810194f, 0.557699262220161f,
+        0.3706515098354541f, 0.4568144014174704f, 0.5665067047641925f,
+        0.3788837845499998f, 0.4655102257395967f, 0.5753141473082242f,
+        0.38722767748645076f, 0.4743176682836283f, 0.5841215898522557f,
+        0.3955715704229017f, 0.48312511082765985f, 0.5929290323962874f,
+        0.4039154633593527f, 0.49193255337169145f, 0.6017364749403189f,
+        0.4122593562958036f, 0.500739995915723f, 0.6105439174843506f,
+        0.4206032492322546f, 0.5095474384597546f, 0.6193513600283821f,
+        0.42894714216870555f, 0.5183548810037861f, 0.6281588025724136f,
+        0.4372910351051563f, 0.5271623235478174f, 0.636966245116445f,
+        0.44577368766047676f, 0.5362472853295879f, 0.6460512068982154f,
+        0.4545811302045084f, 0.5459818270887807f, 0.6557857486574082f,
+        0.46338857274854006f, 0.5557163688479736f, 0.6655202904166011f,
+        0.4721960152925717f, 0.5654509106071665f, 0.675254832175794f,
+        0.48100345783660337f, 0.5751854523663593f, 0.6849893739349868f,
+        0.489810900380635f, 0.5849199941255522f, 0.6947239156941796f,
+        0.4986183429246666f, 0.594654535884745f, 0.7044584574533724f,
+        0.5074257854686983f, 0.6043890776439379f, 0.7141929992125653f,
+        0.51623322801273f, 0.6141236194031308f, 0.7239275409717582f,
+        0.525818945476049f, 0.6238581611623235f, 0.7328838078116634f,
+        0.5355534872352418f, 0.6335927029215163f, 0.7416912503556949f,
+        0.5452880289944345f, 0.643327244680709f, 0.7504986928997265f,
+        0.5550225707536273f, 0.6530617864399018f, 0.759306135443758f,
+        0.56475711251282f, 0.6627963281990945f, 0.7681135779877896f,
+        0.5744916542720128f, 0.6725308699582873f, 0.7769210205318211f,
+        0.5842261960312053f, 0.6822654117174798f, 0.7857284630758524f,
+        0.5939607377903984f, 0.6919999534766728f, 0.7945359056198842f,
+        0.6038712452424287f, 0.7019104609287032f, 0.8031673824710781f,
+        0.6140693366092023f, 0.7121085522954768f, 0.8115112754075292f,
+        0.6242674279759757f, 0.7223066436622502f, 0.8198551683439802f,
+        0.6344655193427492f, 0.7325047350290237f, 0.8281990612804312f,
+        0.6446636107095227f, 0.7427028263957972f, 0.8365429542168822f,
+        0.6548617020762961f, 0.7529009177625706f, 0.8448868471533333f,
+        0.6650597934430696f, 0.7630990091293441f, 0.8532307400897843f,
+        0.6752578848098431f, 0.7732971004961176f, 0.8615746330262353f,
+        0.6854559761766166f, 0.7834951918628911f, 0.8699185259626863f,
+        0.6943750369425534f, 0.7924142526288279f, 0.8769833882983007f,
+        0.703182479486585f, 0.8012216951728595f, 0.8839366324120098f,
+        0.7119899220306165f, 0.810029137716891f, 0.8908898765257189f,
+        0.720797364574648f, 0.8188365802609225f, 0.8978431206394281f,
+        0.7296048071186797f, 0.8276440228049542f, 0.9047963647531372f,
+        0.738412249662711f, 0.8364514653489855f, 0.9117496088668462f,
+        0.7472196922067428f, 0.8452589078930173f, 0.9187028529805555f,
+        0.7560271347507743f, 0.8540663504370488f, 0.9256560970942647f,
+        0.7659004361288364f, 0.8635133082814987f, 0.9328225129747799f,
+        0.777025626710771f, 0.8737113996482722f, 0.9402393066960696f,
+        0.7881508172927056f, 0.8839094910150456f, 0.9476561004173594f,
+        0.7992760078746403f, 0.894107582381819f, 0.9550728941386492f,
+        0.8123197443578297f, 0.9049451890490107f, 0.9620633443263267f,
+        0.8317888278762154f, 0.9179245780612678f, 0.967625939617294f,
+        0.851257911394601f, 0.930903967073525f, 0.9731885349082614f,
+        0.8707269949129868f, 0.9438833560857821f, 0.9787511301992287f,
+        0.8901960784313725f, 0.9568627450980393f, 0.984313725490196f
+    }, 
+    {
+        0.831373f, 0.909804f, 0.980392f,
+        0.8055367647058824f, 0.8950403921568627f, 0.9742404705882353f,
+        0.7797005294117647f, 0.8802767843137255f, 0.9680889411764706f,
+        0.7538642941176471f, 0.8655131764705882f, 0.9619374117647058f,
+        0.7350253725490197f, 0.8537484901960785f, 0.9557858823529412f,
+        0.7178012156862745f, 0.8426758627450981f, 0.9496343529411765f,
+        0.7005770588235294f, 0.8316032352941177f, 0.9434828235294118f,
+        0.6818149607843137f, 0.8209150196078432f, 0.9392537450980393f,
+        0.6621300980392157f, 0.8104574509803922f, 0.936178137254902f,
+        0.6424452352941177f, 0.7999998823529412f, 0.9331025294117647f,
+        0.6227603725490196f, 0.7895423137254902f, 0.9300269215686274f,
+        0.6030755098039216f, 0.7790847450980393f, 0.9269513137254902f,
+        0.5833906470588235f, 0.7686271764705883f, 0.9238757058823529f,
+        0.5640133921568627f, 0.7586309803921569f, 0.9208000588235293f,
+        0.5455589607843138f, 0.7500189019607844f, 0.917724294117647f,
+        0.5271045294117647f, 0.7414068235294118f, 0.9146485294117647f,
+        0.5086500980392157f, 0.7327947450980392f, 0.9115727647058823f,
+        0.4901956666666667f, 0.7241826666666666f, 0.908497f,
+        0.47174123529411766f, 0.7155705882352942f, 0.9054212352941177f,
+        0.4532868039215686f, 0.7069585098039216f, 0.9023454705882353f,
+        0.4364471764705882f, 0.6972699411764706f, 0.8960402352941177f,
+        0.4198382352941176f, 0.6874275882352942f, 0.8892736470588235f,
+        0.4032292941176471f, 0.6775852352941176f, 0.8825070588235294f,
+        0.38662035294117647f, 0.6677428823529412f, 0.8757404705882353f,
+        0.37001141176470587f, 0.6579005294117647f, 0.8689738823529412f,
+        0.3534024705882353f, 0.6480581764705882f, 0.8622072941176471f,
+        0.3374086666666667f, 0.6376006862745097f, 0.8557482352941177f,
+        0.32203000000000004f, 0.6265280588235294f, 0.8495967058823529f,
+        0.30665133333333333f, 0.615455431372549f, 0.8434451764705883f,
+        0.2912726666666667f, 0.6043828039215686f, 0.8372936470588236f,
+        0.275894f, 0.5933101764705881f, 0.8311421176470588f,
+        0.2605153333333333f, 0.5822375490196078f, 0.8249905882352941f,
+        0.24575180392156862f, 0.5715493921568627f, 0.8188390588235294f,
+        0.23529423529411772f, 0.5635525294117647f, 0.8126875294117647f,
+        0.22483666666666666f, 0.5555556666666667f, 0.806536f,
+        0.21437909803921568f, 0.5475588039215686f, 0.8003844705882353f,
+        0.2039215294117647f, 0.5395619411764705f, 0.7942329411764706f,
+        0.1934639607843138f, 0.5315650784313726f, 0.7880814117647059f,
+        0.18300639215686273f, 0.5235682156862744f, 0.7819298823529411f,
+        0.1711649411764706f, 0.5123416470588236f, 0.7711649411764706f,
+        0.1588621960784314f, 0.5000385098039216f, 0.7588621960784314f,
+        0.14825105882352946f, 0.4885043529411766f, 0.747482117647059f,
+        0.14440635294117649f, 0.48004611764705885f, 0.739792705882353f,
+        0.14056164705882354f, 0.4715878823529412f, 0.7321032941176471f,
+        0.1367169411764706f, 0.46312964705882353f, 0.7244138823529411f,
+        0.13287223529411768f, 0.45467141176470593f, 0.7167244705882353f,
+        0.1288737450980392f, 0.44667452941176466f, 0.7094195294117647f,
+        0.12349119607843137f, 0.44282982352941175f, 0.7055748235294118f,
+        0.11810864705882353f, 0.43898511764705883f, 0.7017301176470588f,
+        0.11272609803921571f, 0.4351404117647059f, 0.697885411764706f,
+        0.10734354901960784f, 0.43129570588235294f, 0.694040705882353f,
+        0.101961f, 0.427451f, 0.690196f,
+        0.09888531372549018f, 0.4182239411764706f, 0.6778932549019607f,
+        0.0958096274509804f, 0.40899688235294124f, 0.6655905098039216f,
+        0.09342588235294116f, 0.4018458235294117f, 0.6553637647058823f,
+        0.0918878431372549f, 0.39723209803921566f, 0.6476743529411765f,
+        0.09034980392156863f, 0.3926183725490196f, 0.6399849411764706f,
+        0.0874278823529412f, 0.38385229411764715f, 0.6322955294117648f,
+        0.08435219607843138f, 0.3746248431372549f, 0.6246061176470589f,
+        0.08073819607843138f, 0.36647401960784315f, 0.616916705882353f,
+        0.0761244705882353f, 0.36032264705882355f, 0.6092272941176471f,
+        0.07151074509803923f, 0.35417127450980396f, 0.6015378823529411f,
+        0.06935788235294117f, 0.34555935294117646f, 0.5901574901960783f,
+        0.06782023529411765f, 0.33633229411764703f, 0.5778543529411765f,
+        0.06743592156862746f, 0.3278740588235294f, 0.5659357254901961f,
+        0.07051160784313726f, 0.321722294117647f, 0.5551706274509803f,
+        0.07358729411764703f, 0.3155705294117648f, 0.5444055294117648f,
+        0.07773960784313726f, 0.31157201960784314f, 0.5325640784313725f,
+        0.08235333333333332f, 0.3084963333333333f, 0.5202613333333335f,
+        0.08673635294117645f, 0.3047286470588236f, 0.5079585294117649f,
+        0.08981203921568626f, 0.29703962745098045f, 0.49565539215686283f,
+        0.09288772549019605f, 0.2893506078431373f, 0.4833522549019609f,
+        0.09596341176470587f, 0.2844293529411765f, 0.4719719411764706f,
+        0.09903909803921569f, 0.2813532745098039f, 0.46120684313725485f,
+        0.10211478431372548f, 0.2782003333333333f, 0.4502110784313726f,
+        0.10519047058823532f, 0.27358699999999997f, 0.4348326470588234f,
+        0.1082661568627451f, 0.26897366666666667f, 0.41945421568627445f,
+        0.11057282352941177f, 0.2651291568627451f, 0.405613431372549f,
+        0.1121104705882353f, 0.2620534705882353f, 0.39331029411764706f,
+        0.11364811764705883f, 0.2589777843137255f, 0.38100715686274506f,
+        0.113725f, 0.2559017254901961f, 0.3687043921568628f,
+        0.113725f, 0.25282564705882354f, 0.3564016470588236f,
+        0.11310994117647062f, 0.2534407058823528f, 0.344713960784314f,
+        0.11157229411764708f, 0.25959247058823526f, 0.3339488627450981f,
+        0.11003464705882354f, 0.2657442352941176f, 0.3231837647058824f,
+        0.10849666666666667f, 0.2784316666666666f, 0.3019610000000001f,
+        0.1069586274509804f, 0.29227245098039206f, 0.2788927647058824f,
+        0.10449800000000001f, 0.3028837058823529f, 0.26043817647058826f,
+        0.09988466666666668f, 0.30595939215686274f, 0.25274876470588237f,
+        0.09527133333333333f, 0.30903507843137257f, 0.24505935294117648f,
+        0.090657705882353f, 0.313264294117647f, 0.23852347058823536f,
+        0.08604398039215685f, 0.3178780196078432f, 0.23237209803921566f,
+        0.08173786274509802f, 0.32310680392156865f, 0.226835862745098f,
+        0.07866217647058822f, 0.3307958235294118f, 0.2237601764705882f,
+        0.07558649019607842f, 0.3384848431372549f, 0.22068449019607841f,
+        0.07351027450980392f, 0.34617411764705885f, 0.21760880392156862f,
+        0.07197223529411766f, 0.3538635294117647f, 0.21453311764705885f,
+        0.07043423529411766f, 0.3615529411764705f, 0.21191876470588236f,
+        0.06889658823529415f, 0.36924235294117635f, 0.21345641176470587f,
+        0.0673589411764706f, 0.37693176470588236f, 0.2149940588235294f,
+        0.06582107843137255f, 0.38462117647058824f, 0.20722786274509805f,
+        0.06428303921568627f, 0.39231058823529413f, 0.19184943137254903f,
+        0.062745f, 0.4f, 0.176471f,
+        0.06459049019607843f, 0.40307576470588236f, 0.17154974509803922f,
+        0.06643598039215687f, 0.4061515294117647f, 0.16662849019607842f,
+        0.06828147058823529f, 0.4092272941176471f, 0.16170723529411765f,
+        0.0701269607843137f, 0.41230305882352936f, 0.15678598039215694f,
+        0.07197245098039216f, 0.41537882352941174f, 0.15186472549019608f,
+        0.0738179411764706f, 0.4184545882352941f, 0.1469434705882353f,
+        0.07566343137254902f, 0.42153035294117647f, 0.1424067254901961f,
+        0.07750892156862746f, 0.4246061176470588f, 0.1381006862745098f,
+        0.07935441176470588f, 0.4276818823529412f, 0.13379464705882355f,
+        0.08119990196078432f, 0.4307576470588235f, 0.12948860784313726f,
+        0.08304539215686275f, 0.43383341176470586f, 0.12518256862745097f,
+        0.08489088235294115f, 0.4369091764705882f, 0.12087652941176477f,
+        0.08765907843137255f, 0.4404462745098039f, 0.117647f,
+        0.09319539215686275f, 0.4453673725490196f, 0.117647f,
+        0.09873170588235294f, 0.4502884705882353f, 0.117647f,
+        0.10426801960784314f, 0.455209568627451f, 0.117647f,
+        0.10980433333333334f, 0.4601306666666667f, 0.117647f,
+        0.11534064705882352f, 0.46505176470588233f, 0.117647f,
+        0.12087696078431373f, 0.46997286274509803f, 0.117647f,
+        0.13018105882352926f, 0.4748940980392156f, 0.12195309803921561f,
+        0.1400234117647059f, 0.4798153529411765f, 0.12687435294117647f,
+        0.14986576470588236f, 0.48473660784313727f, 0.13179560784313726f,
+        0.15970811764705883f, 0.48965786274509804f, 0.13671686274509803f,
+        0.1695504705882353f, 0.4945791176470588f, 0.14163811764705883f,
+        0.17939282352941177f, 0.4995003725490196f, 0.14655937254901963f,
+        0.18985031372549022f, 0.5050366862745098f, 0.1520956862745098f,
+        0.20092294117647058f, 0.5111880588235294f, 0.15824705882352944f,
+        0.211995568627451f, 0.517339431372549f, 0.16439843137254903f,
+        0.2230681960784314f, 0.5234908039215687f, 0.17054980392156865f,
+        0.23414082352941146f, 0.5296421764705881f, 0.17670117647058808f,
+        0.24521345098039218f, 0.5357935490196079f, 0.18285254901960785f,
+        0.2562860784313726f, 0.5419449411764706f, 0.18908084313725493f,
+        0.26735870588235294f, 0.5480964705882353f, 0.19584758823529413f,
+        0.27843133333333336f, 0.554248f, 0.20261433333333334f,
+        0.28950396078431373f, 0.5603995294117647f, 0.20938107843137255f,
+        0.30057658823529415f, 0.5665510588235294f, 0.21614782352941175f,
+        0.3116492156862745f, 0.5727025882352941f, 0.222914568627451f,
+        0.32272184313725494f, 0.5788541176470589f, 0.2296813137254902f,
+        0.33471729411764706f, 0.5850056470588235f, 0.23690923529411764f,
+        0.3470203529411765f, 0.5911571764705883f, 0.24429088235294116f,
+        0.3593234117647059f, 0.597308705882353f, 0.2516725294117647f,
+        0.3716264705882353f, 0.6034602352941176f, 0.2590541764705882f,
+        0.3839295294117647f, 0.6096117647058824f, 0.26643582352941175f,
+        0.3962325882352941f, 0.6157632941176471f, 0.2738174705882353f,
+        0.409689f, 0.6226068235294118f, 0.28166052941176467f,
+        0.42506766666666623f, 0.6306036862745096f, 0.290272607843137f,
+        0.44044633333333333f, 0.6386005490196078f, 0.2988846862745098f,
+        0.455825f, 0.6465974117647059f, 0.30749676470588233f,
+        0.4712036666666667f, 0.6545942745098039f, 0.3161088431372549f,
+        0.48658233333333334f, 0.6625911372549019f, 0.3247209215686274f,
+        0.501961f, 0.670588f, 0.333333f,
+        0.5137513267973856f, 0.6782774117647058f, 0.3420476405228758f,
+        0.5255416535947712f, 0.6859668235294117f, 0.35076228104575163f,
+        0.5373319803921568f, 0.6936562352941176f, 0.3594769215686274f,
+        0.5491223071895425f, 0.7013456470588234f, 0.3681915620915033f,
+        0.5609126339869281f, 0.7090350588235294f, 0.37690620261437907f,
+        0.5727029607843137f, 0.7167244705882352f, 0.38562084313725487f,
+        0.5844932875816993f, 0.7244138823529411f, 0.3943354836601307f,
+        0.602383794117647f, 0.7334488725490195f, 0.4061899215686274f,
+        0.6316032058823529f, 0.744982794117647f, 0.42387541176470583f,
+        0.6608226176470587f, 0.7565167156862744f, 0.4415609019607843f,
+        0.6900420294117637f, 0.7680506372549016f, 0.45924639215686225f,
+        0.7192614411764705f, 0.7795845588235294f, 0.47693188235294115f,
+        0.7469430588235293f, 0.7916952450980392f, 0.49577080392156864f,
+        0.7700112941176469f, 0.805536225490196f, 0.518070019607843f,
+        0.7930795294117645f, 0.8193772058823529f, 0.5403692352941174f,
+        0.816147764705882f, 0.8332181862745096f, 0.5626684509803919f,
+        0.8392159999999995f, 0.8470591666666665f, 0.5849676666666662f,
+        0.8600542941176469f, 0.8583626274509805f, 0.6034606274509804f,
+        0.8682562549019607f, 0.8552868104575164f, 0.6003848104575163f,
+        0.8764582156862744f, 0.8522109934640524f, 0.5973089934640523f,
+        0.8846601764705881f, 0.8491351764705883f, 0.5942331764705883f,
+        0.8928621372549018f, 0.8460593594771242f, 0.5911573594771242f,
+        0.9010640980392156f, 0.8429835424836601f, 0.5880815424836602f,
+        0.9092660588235294f, 0.8399077254901961f, 0.5850057254901961f,
+        0.9174680196078431f, 0.8368319084967321f, 0.5819299084967321f,
+        0.9208000588235293f, 0.8311417647058823f, 0.577008705882353f,
+        0.9192621764705883f, 0.8228372941176473f, 0.570242117647059f,
+        0.917724294117647f, 0.8145328235294117f, 0.5634755294117647f,
+        0.9161864117647058f, 0.8062283529411765f, 0.5567089411764706f,
+        0.9146485294117647f, 0.7979238823529411f, 0.5499423529411764f,
+        0.9131106470588235f, 0.7896194117647058f, 0.5431757647058824f,
+        0.9115727647058823f, 0.7813149411764706f, 0.5364091764705883f,
+        0.9100348823529412f, 0.7730104705882352f, 0.5296425882352941f,
+        0.908497f, 0.764706f, 0.522876f,
+        0.9069591176470588f, 0.7564015294117646f, 0.5161094117647059f,
+        0.9054212352941177f, 0.7480970588235294f, 0.5093428235294117f,
+        0.9038833529411765f, 0.739792588235294f, 0.5025762352941177f,
+        0.9023454705882353f, 0.7314881176470588f, 0.4958096470588235f,
+        0.8994235294117647f, 0.7208768235294117f, 0.4878895882352941f,
+        0.8960402352941177f, 0.709496588235294f, 0.4795850392156863f,
+        0.8926569411764707f, 0.6981163529411765f, 0.4712804901960784f,
+        0.8892736470588235f, 0.6867361176470588f, 0.46297594117647056f,
+        0.8858903529411766f, 0.6753558823529414f, 0.454671392156863f,
+        0.8825070588235294f, 0.6639756470588235f, 0.4463668431372549f,
+        0.8791237647058824f, 0.6525954117647058f, 0.4380622941176471f,
+        0.8757404705882353f, 0.6412151764705882f, 0.4297577450980392f,
+        0.8723571764705883f, 0.6298349411764705f, 0.42145319607843135f,
+        0.8689738823529412f, 0.618454705882353f, 0.41314864705882354f,
+        0.8655905882352941f, 0.6070744705882353f, 0.40484409803921567f,
+        0.8622072941176471f, 0.5956942352941177f, 0.39653954901960786f,
+        0.858824f, 0.584314f, 0.388235f,
+        0.8542103529411765f, 0.5729337647058823f, 0.38300629411764703f,
+        0.8495967058823529f, 0.5615535294117647f, 0.3777775882352941f,
+        0.8449830588235294f, 0.550173294117647f, 0.37254888235294115f,
+        0.8403694117647059f, 0.5387930588235295f, 0.36732017647058823f,
+        0.8357557647058824f, 0.5274128235294118f, 0.36209147058823526f,
+        0.8311421176470588f, 0.5160325882352941f, 0.35686276470588235f,
+        0.8265284705882353f, 0.5046523529411765f, 0.3516340588235294f,
+        0.8219148235294119f, 0.49327211764705914f, 0.34640535294117664f,
+        0.8173011764705883f, 0.48189188235294117f, 0.3411766470588235f,
+        0.8126875294117647f, 0.47051164705882353f, 0.3359479411764706f,
+        0.8080738823529412f, 0.45913141176470584f, 0.3307192352941176f,
+        0.8034602352941177f, 0.44775117647058826f, 0.3254905294117647f,
+        0.7976162941176471f, 0.43644782352941175f, 0.31926219607843137f,
+        0.7880814705882353f, 0.4253751176470588f, 0.3100349803921569f,
+        0.7785466470588236f, 0.4143024117647059f, 0.30080776470588233f,
+        0.7690118235294118f, 0.40322970588235296f, 0.29158054901960784f,
+        0.7594770000000001f, 0.392157f, 0.28235333333333335f,
+        0.7499421764705883f, 0.38108429411764705f, 0.2731261176470588f,
+        0.7404073529411765f, 0.3700115882352941f, 0.2638989019607843f,
+        0.7308725294117647f, 0.3589388823529412f, 0.2546716862745098f,
+        0.7213377058823529f, 0.34786617647058826f, 0.24544447058823526f,
+        0.7118028823529412f, 0.3367934705882353f, 0.23621725490196077f,
+        0.7022680588235294f, 0.32572076470588235f, 0.22699003921568628f,
+        0.6927332352941179f, 0.31464805882352975f, 0.217762823529412f,
+        0.6831984117647059f, 0.30357535294117644f, 0.20853560784313724f,
+        0.6733560588235294f, 0.29296401960784313f, 0.20023109803921568f,
+        0.6632061764705882f, 0.2828140588235294f, 0.19284929411764706f,
+        0.6530562941176471f, 0.2726640980392157f, 0.18546749019607842f,
+        0.6429064117647059f, 0.26251413725490197f, 0.1780856862745098f,
+        0.6327565294117647f, 0.25236417647058823f, 0.17070388235294118f,
+        0.6226066470588235f, 0.2422142156862745f, 0.16332207843137253f,
+        0.6124567647058823f, 0.23206425490196078f, 0.1559402745098039f,
+        0.6023068823529412f, 0.22191429411764707f, 0.1485584705882353f,
+        0.5921569999999999f, 0.21176433333333333f, 0.14117666666666667f,
+        0.5820071176470588f, 0.2016143725490196f, 0.13379486274509805f,
+        0.5718572352941176f, 0.19146441176470588f, 0.12641305882352943f,
+        0.5617073529411765f, 0.18131445098039217f, 0.11903125490196079f,
+        0.5515574705882352f, 0.17116449019607843f, 0.11164945098039215f,
+        0.5407154117647058f, 0.15847711764705882f, 0.10242223529411765f,
+        0.5296426274509807f, 0.14494394117647097f, 0.09257988235294146f,
+        0.5185698431372548f, 0.13141076470588237f, 0.08273752941176471f,
+        0.5074970588235294f, 0.11787758823529412f, 0.07289517647058824f,
+        0.4964242745098039f, 0.10434441176470588f, 0.06305282352941177f,
+        0.48535149019607843f, 0.09081123529411764f, 0.0532104705882353f,
+        0.47681623529411765f, 0.0779700588235294f, 0.044290823529411764f,
+        0.47251019607843137f, 0.0662822156862745f, 0.03690901960784314f,
+        0.4682041568627451f, 0.05459437254901961f, 0.02952721568627451f,
+        0.4638981176470588f, 0.042906529411764704f, 0.02214541176470588f,
+        0.45959207843137256f, 0.031218686274509806f, 0.014763607843137252f,
+        0.4552860392156863f, 0.019530843137254908f, 0.007381803921568626f,
+        0.45098f, 0.007843f, 0.0f
+    }, 
+    {
+        0.301961f, 0.266667f, 0.239216f,
+        0.31610937062668204f, 0.2789699081122645f, 0.24844317108804306f,
+        0.3302577412533641f, 0.29127281622452905f, 0.25767034217608614f,
+        0.3444061118800461f, 0.30357572433679353f, 0.2668975132641292f,
+        0.35855448250672817f, 0.31587863244905806f, 0.27612468435217224f,
+        0.37270285313341023f, 0.3281815405613226f, 0.2853518554402153f,
+        0.3868512237600923f, 0.34048444867358707f, 0.29457902652825835f,
+        0.40099961553248753f, 0.3531718569780854f, 0.3022683583237217f,
+        0.41514801999231066f, 0.3660899653979239f, 0.3090349865436371f,
+        0.42929642445213384f, 0.3790080738177624f, 0.3158016147635525f,
+        0.44344482891195697f, 0.39192618223760095f, 0.3225682429834679f,
+        0.45759323337178015f, 0.4048442906574395f, 0.32933487120338334f,
+        0.4717416378316033f, 0.41776239907727797f, 0.33610149942329876f,
+        0.48558246828143026f, 0.43052672049211843f, 0.342560553633218f,
+        0.4985005767012688f, 0.4428296808919646f, 0.3480968858131488f,
+        0.5114186851211073f, 0.45513264129181086f, 0.3536332179930796f,
+        0.5243367935409458f, 0.46743560169165704f, 0.3591695501730104f,
+        0.5372549019607843f, 0.47973856209150323f, 0.36470588235294116f,
+        0.5501730103806228f, 0.4920415224913494f, 0.370242214532872f,
+        0.5630911188004613f, 0.5043444828911956f, 0.37577854671280275f,
+        0.5733179546328335f, 0.5139561707035755f, 0.3780853517877739f,
+        0.5831603229527105f, 0.5231833910034601f, 0.37993079584775086f,
+        0.5930026912725874f, 0.5324106113033449f, 0.3817762399077278f,
+        0.6028450595924644f, 0.5416378316032295f, 0.38362168396770474f,
+        0.6126874279123414f, 0.5508650519031142f, 0.3854671280276817f,
+        0.6225297962322184f, 0.5600922722029988f, 0.3873125720876586f,
+        0.6314494425221069f, 0.568396770472895f, 0.38762014609765477f,
+        0.639446366782007f, 0.5757785467128028f, 0.3863898500576701f,
+        0.647443291041907f, 0.5831603229527105f, 0.3851595540176855f,
+        0.655440215301807f, 0.5905420991926182f, 0.3839292579777009f,
+        0.663437139561707f, 0.597923875432526f, 0.38269896193771624f,
+        0.6714340638216071f, 0.6053056516724338f, 0.38146866589773165f,
+        0.6795847750865053f, 0.6129181084198385f, 0.3803921568627451f,
+        0.6888119953863898f, 0.6221453287197232f, 0.3803921568627451f,
+        0.6980392156862746f, 0.6313725490196078f, 0.3803921568627451f,
+        0.7072664359861592f, 0.6405997693194925f, 0.3803921568627451f,
+        0.7164936562860439f, 0.6498269896193771f, 0.3803921568627451f,
+        0.7257208765859284f, 0.6590542099192618f, 0.3803921568627451f,
+        0.7349480968858132f, 0.6682814302191464f, 0.3803921568627451f,
+        0.7455594002306806f, 0.6793540945790081f, 0.3845444059976932f,
+        0.7566320645905421f, 0.6910419069588619f, 0.390080738177624f,
+        0.7677047289504036f, 0.7027297193387158f, 0.3956170703575547f,
+        0.7787773933102653f, 0.7144175317185698f, 0.4011534025374856f,
+        0.789850057670127f, 0.7261053440984236f, 0.4066897347174164f,
+        0.8009227220299885f, 0.7377931564782776f, 0.4122260668973472f,
+        0.8117647058823528f, 0.7499423298731256f, 0.4200692041522491f,
+        0.8222222222222222f, 0.7628604382929642f, 0.4317570165321031f,
+        0.8326797385620915f, 0.7757785467128028f, 0.44344482891195697f,
+        0.8431372549019608f, 0.7886966551326413f, 0.45513264129181086f,
+        0.85359477124183f, 0.8016147635524797f, 0.46682045367166464f,
+        0.8640522875816994f, 0.8145328719723183f, 0.47850826605151864f,
+        0.8745098039215686f, 0.8274509803921568f, 0.49019607843137253f,
+        0.8837370242214533f, 0.8391387927720108f, 0.5037293348712033f,
+        0.8929642445213379f, 0.8508266051518646f, 0.5172625913110341f,
+        0.9021914648212226f, 0.8625144175317185f, 0.530795847750865f,
+        0.9114186851211072f, 0.8742022299115725f, 0.5443291041906958f,
+        0.9206459054209919f, 0.8858900422914264f, 0.5578623606305266f,
+        0.9298731257208765f, 0.8975778546712803f, 0.5713956170703574f,
+        0.9348712033833141f, 0.9050365244136871f, 0.5864667435601691f,
+        0.9373317954632834f, 0.9099577085736256f, 0.6024605920799692f,
+        0.9397923875432526f, 0.9148788927335639f, 0.6184544405997693f,
+        0.9427912341407151f, 0.9203383314109957f, 0.6355247981545558f,
+        0.946482122260669f, 0.9264898116109188f, 0.6539792387543253f,
+        0.9501730103806227f, 0.932641291810842f, 0.6724336793540946f,
+        0.9803921568627451f, 0.9598796012718599f, 0.6248876284402043f,
+        0.9803921568627451f, 0.9522614779722316f, 0.6033029457579241f,
+        0.9803921568627451f, 0.9446433546726033f, 0.5817182630756441f,
+        0.9802819149805966f, 0.9369609236083882f, 0.5600861151385496f,
+        0.976625215796775f, 0.9272097257848639f, 0.5369270203076794f,
+        0.9729685166129534f, 0.9174585279613396f, 0.513767925476809f,
+        0.9693118174291318f, 0.9077073301378153f, 0.4906088306459388f,
+        0.9686274509803922f, 0.8969653547359303f, 0.4654681806583472f,
+        0.9686274509803922f, 0.8859952571844655f, 0.4398712863715959f,
+        0.9686274509803922f, 0.8750251596330006f, 0.41427439208484457f,
+        0.9679015428456609f, 0.8618773376773421f, 0.3886774977980937f,
+        0.9666826431177203f, 0.8472505409420554f, 0.3630806035113421f,
+        0.9654637433897798f, 0.832623744206769f, 0.3374837092245909f,
+        0.9637838049707373f, 0.8175359087803806f, 0.3137309697022474f,
+        0.9613460055148563f, 0.8016902123171536f, 0.2930096743272582f,
+        0.9589082060589752f, 0.7858445158539267f, 0.27228837895226904f,
+        0.9562742373556216f, 0.7699988193906997f, 0.25195942207222505f,
+        0.9526175381717998f, 0.7541531229274726f, 0.23367592615311694f,
+        0.9489608389879783f, 0.7383074264642459f, 0.2153924302340091f,
+        0.9453041398041566f, 0.7224617300010185f, 0.19710893431490079f,
+        0.9439478396839025f, 0.7054658340060078f, 0.17997563792757645f,
+        0.942728939955962f, 0.6884012378148402f, 0.16291104173640888f,
+        0.9415100402280214f, 0.6713366416236727f, 0.14584644554524134f,
+        0.9385204803237721f, 0.6533867153443508f, 0.13055250953038264f,
+        0.9348637811399505f, 0.6351032194252428f, 0.11592571279509623f,
+        0.9312070819561289f, 0.6168197235061348f, 0.10129891605980981f,
+        0.9281708434168324f, 0.599777148876077f, 0.08915396190262356f,
+        0.9257330439609512f, 0.5839314524128498f, 0.07940276407909912f,
+        0.9232952445050702f, 0.5680857559496227f, 0.0696515662555748f,
+        0.9208574450491891f, 0.5522400594863958f, 0.060967142034737444f,
+        0.918419645593308f, 0.5363943630231689f, 0.054872643395034774f,
+        0.915981846137427f, 0.5205486665599419f, 0.048778144755332105f,
+        0.9134533249242796f, 0.5046122483394486f, 0.04277436787289571f,
+        0.909796625740458f, 0.4875476521482812f, 0.03789876896113357f,
+        0.9061399265566364f, 0.470483055957114f, 0.033023170049371504f,
+        0.9024832273728148f, 0.4534184597659463f, 0.028147571137609302f,
+        0.8904685118563737f, 0.4311301033668914f, 0.023271972225847107f,
+        0.8770606148490276f, 0.407971008536021f, 0.01839637331408491f,
+        0.8636527178416815f, 0.3848119137051505f, 0.013520774402322707f,
+        0.845565525246647f, 0.36009305367838407f, 0.009425058088508648f,
+        0.824844229871658f, 0.33449615939163285f, 0.005768358904687046f,
+        0.8041229344966688f, 0.30889926510488164f, 0.002111659720865444f,
+        0.7880367575105484f, 0.2869074628983617f, 0.0f,
+        0.778285559687024f, 0.269842866707194f, 0.0f,
+        0.7685343618634997f, 0.2527782705160265f, 0.0f,
+        0.756281726933082f, 0.23621396174623774f, 0.0007504311320680208f,
+        0.7343415318301524f, 0.22158716501095133f, 0.004407130315889622f,
+        0.7124013367272228f, 0.2069603682756649f, 0.008063829499711224f,
+        0.6904611416242933f, 0.1923335715403785f, 0.011720528683532825f,
+        0.6709292945113645f, 0.17650260081009153f, 0.014173053872353942f,
+        0.6514268988643162f, 0.16065690434686475f, 0.016610853328234986f,
+        0.6319245032172672f, 0.1448112078836375f, 0.019048652784116098f,
+        0.6086648893647341f, 0.1289655114204105f, 0.021486452239997178f,
+        0.5842868948059233f, 0.11311981495718348f, 0.02392425169587826f,
+        0.5599089002471126f, 0.09727411849395648f, 0.026362051151759335f,
+        0.5348564705805601f, 0.08345172735395483f, 0.03014872082312395f,
+        0.5092595762938088f, 0.07126273007454949f, 0.03502431973488609f,
+        0.48366268200705764f, 0.05907373279514415f, 0.03989991864664823f,
+        0.4580657877203067f, 0.050161260828637984f, 0.045594648886635065f,
+        0.4324688934335549f, 0.04772346137275686f, 0.05290804725427843f,
+        0.4068719991468035f, 0.045285661916875775f, 0.06022144562192171f,
+        0.37216887030186757f, 0.04559514544574167f, 0.06869082317427866f,
+        0.26986239238895643f, 0.06630005470405285f, 0.08574202056008974f,
+        0.16755591447604534f, 0.08700496396236404f, 0.10279321794590082f,
+        0.0652494365631342f, 0.10770987322067523f, 0.11984441533171192f,
+        0.060405735428046226f, 0.12741578495679112f, 0.13807992559587734f,
+        0.06652876344946963f, 0.14700928726218748f, 0.15644869738821665f,
+        0.07265179147089304f, 0.16660278956758384f, 0.17481746918055596f,
+        0.07706893519620056f, 0.18534345847694872f, 0.18892163898463246f,
+        0.08074281446344087f, 0.2037125425412193f, 0.20116769502747953f,
+        0.08441669373068106f, 0.22208162660548933f, 0.21341375107032623f,
+        0.08748533055866523f, 0.23863498335199132f, 0.22444932223466088f,
+        0.08993447931284838f, 0.25332987587709027f, 0.23424591725139352f,
+        0.09238362806703153f, 0.2680247684021892f, 0.24404251226812612f,
+        0.0955475536424294f, 0.28486417366221145f, 0.2534817644420712f,
+        0.10044585115079575f, 0.3069071369937219f, 0.26205394121767783f,
+        0.10534414865916211f, 0.32895010032523236f, 0.2706261179932844f,
+        0.10991359756616595f, 0.3504449640201837f, 0.278431f,
+        0.11113801580729209f, 0.36636458705833985f, 0.278431f,
+        0.11236243404841822f, 0.382284210096496f, 0.278431f,
+        0.11358685228954435f, 0.39820383313465213f, 0.278431f,
+        0.11481154756982646f, 0.414123456172808f, 0.2762581818995031f,
+        0.11603627808288348f, 0.4300430792109638f, 0.27380903314531996f,
+        0.11726100859594052f, 0.4459627022491196f, 0.2713598843911368f,
+        0.117647f, 0.4593665356701801f, 0.26471746780186367f,
+        0.117647f, 0.47161259171302716f, 0.25614529102625694f,
+        0.117647f, 0.48385864775587384f, 0.24757311425065043f,
+        0.11705606935183134f, 0.49492299317312477f, 0.23722844687202088f,
+        0.11583133883877429f, 0.5047199004617884f, 0.224982703101105f,
+        0.11460660832571724f, 0.514516807750452f, 0.2127369593301891f,
+        0.11338196529919445f, 0.5239705928517758f, 0.19946184899725378f,
+        0.11215754705806832f, 0.5325427696273825f, 0.18354191368716674f,
+        0.11093312881694219f, 0.541114946402989f, 0.1676219783770797f,
+        0.10913687682140437f, 0.5499730157534745f, 0.15122554734141894f,
+        0.10056470004579793f, 0.56221875952439f, 0.12918289628183982f,
+        0.0919925232701915f, 0.5744645032953057f, 0.10714024522226068f,
+        0.08342034649458506f, 0.5867102470662214f, 0.08509759416268153f,
+        0.0898578302810215f, 0.6000285000059271f, 0.07377593394023943f,
+        0.09843000705662808f, 0.613499286561831f, 0.06397902665157579f,
+        0.10700218383223467f, 0.6269700731177348f, 0.05418211936291216f,
+        0.1312367980013871f, 0.6412648665798616f, 0.05757478792575119f,
+        0.16307635634963116f, 0.6559597591049612f, 0.0673716952144151f,
+        0.19491591469787434f, 0.6706546516300602f, 0.07716860250307873f,
+        0.22733194488156255f, 0.6847732193413503f, 0.08696536277010616f,
+        0.26039592147093155f, 0.698244005897254f, 0.09676195778683878f,
+        0.2934598980603006f, 0.7117147924531577f, 0.10655855280357142f,
+        0.3271814954421744f, 0.72485668477538f, 0.11767030556788498f,
+        0.3626949330576572f, 0.7371024285462957f, 0.1323655103649147f,
+        0.39820837067313997f, 0.7493481723172114f, 0.14706071516194444f,
+        0.43396477278845f, 0.7615939367413488f, 0.16240387348269752f,
+        0.47315177739924275f, 0.7738399927841956f, 0.18689598556839115f,
+        0.5123387820100355f, 0.7860860488270425f, 0.21138809765408478f,
+        0.5515257866208281f, 0.7983321048698893f, 0.2358802097397784f,
+        0.5864816347629046f, 0.810578160912736f, 0.2540253173814523f,
+        0.620770341865331f, 0.8228242169555828f, 0.2711696709326655f,
+        0.6550590489677572f, 0.8350702729984296f, 0.2883140244838786f,
+        0.69096778056709f, 0.8473161224845379f, 0.3151785250165367f,
+        0.7277056364237002f, 0.859561866255454f, 0.3470177710928496f,
+        0.7644434922803094f, 0.8718076100263699f, 0.37885701716916154f,
+        0.7989324066180044f, 0.8846158042358104f, 0.41575659672182685f,
+        0.8307719649662473f, 0.8980865907917142f, 0.4586174805998594f,
+        0.8626115233144902f, 0.911557377347618f, 0.501478364477892f,
+        0.8922497518032245f, 0.9237702496711849f, 0.5443392483559245f,
+        0.91551682110393f, 0.9323424264467914f, 0.5872001322339571f,
+        0.9387838904046354f, 0.940914603222398f, 0.6300610161119896f,
+        0.9614508334268967f, 0.9496201432825605f, 0.6745222934129278f,
+        0.9736968894697435f, 0.9606414688123502f, 0.7467735868850207f,
+        0.9859429455125903f, 0.9716627943421401f, 0.8190248803571136f,
+        0.9981890015554372f, 0.9826841198719298f, 0.8912761738292065f,
+        0.8400593006492788f, 0.890828975486959f, 0.7926814714742926f,
+        0.8225883764452582f, 0.8777257823339436f, 0.7702187864505137f,
+        0.8051174522412377f, 0.8646225891809282f, 0.7477561014267349f,
+        0.7876465280372176f, 0.8515193960279132f, 0.7252934164029566f,
+        0.7701756038331965f, 0.8384162028748974f, 0.7028307313791772f,
+        0.7527046796291759f, 0.8253130097218819f, 0.6803680463553983f,
+        0.735938331489784f, 0.8125620596836707f, 0.659666936245722f,
+        0.7197153077131807f, 0.8000827371893193f, 0.6403242409556197f,
+        0.7034922839365773f, 0.787603414694968f, 0.6209815456655174f,
+        0.6872692601599739f, 0.7751240922006166f, 0.6016388503754151f,
+        0.6710462363833705f, 0.7626447697062653f, 0.5822961550853128f,
+        0.6548232126067671f, 0.7501654472119139f, 0.5629534597952105f,
+        0.6389492119548505f, 0.737511702146327f, 0.5439596986386873f,
+        0.6239742477157536f, 0.7244086681034004f, 0.5258647446659137f,
+        0.6089992834766567f, 0.711305634060474f, 0.50776979069314f,
+        0.5940243192375596f, 0.6982026000175474f, 0.48967483672036627f,
+        0.5790493549984627f, 0.685099565974621f, 0.47157988274759255f,
+        0.5640743907593657f, 0.6719965319316944f, 0.4534849287748189f,
+        0.5490994265202687f, 0.6588934978887679f, 0.43538997480204517f,
+        0.5347450239586111f, 0.6464108672571044f, 0.4197777423375569f,
+        0.5203939303781779f, 0.6339315447627528f, 0.40417874832970635f,
+        0.5060428367977452f, 0.6214522222684016f, 0.38857975432185626f,
+        0.49169174321731246f, 0.6089728997740504f, 0.3729807603140061f,
+        0.47734064963687967f, 0.5964935772796992f, 0.35738176630615603f,
+        0.4629895560564469f, 0.584014254785348f, 0.34178277229830595f,
+        0.4495241466615725f, 0.5715350452177066f, 0.32751224810543833f,
+        0.436420953508557f, 0.5590558818334442f, 0.31378502518366974f,
+        0.4233177603555415f, 0.5465767184491819f, 0.3000578022619011f,
+        0.41021456720252597f, 0.5340975550649195f, 0.28633057934013245f,
+        0.39711137404951047f, 0.5216183916806572f, 0.27260335641836386f,
+        0.384008180896495f, 0.5091392282963948f, 0.2588761334965952f,
+        0.371169997324889f, 0.4963949877434795f, 0.2459441420807853f,
+        0.35869067483053774f, 0.483291794590464f, 0.23408900846527608f,
+        0.34621135233618644f, 0.4701886014374485f, 0.22223387484976687f,
+        0.33373202984183514f, 0.457085408284433f, 0.21037874123425765f,
+        0.32125270734748423f, 0.4439822151314179f, 0.19852360761874877f,
+        0.3087733848531326f, 0.430879021978402f, 0.1866684740032392f,
+        0.29646857392112475f, 0.4178630623587813f, 0.17507504098791427f,
+        0.28523731096427973f, 0.40538373986443f, 0.16509151934839766f,
+        0.2740060480074347f, 0.39290441737007875f, 0.1551079977088811f,
+        0.2627747850505897f, 0.38042509487572745f, 0.1451244760693645f,
+        0.25154352209374464f, 0.36794577238137616f, 0.13514095442984791f,
+        0.24031225913689966f, 0.3554664498870249f, 0.1251574327903313f,
+        0.22908099618005462f, 0.3429871273926736f, 0.11517391115081471f,
+        0.21891652552969754f, 0.32997461277267753f, 0.10839103846756061f,
+        0.20893316300026993f, 0.316871578729751f, 0.10215153633047391f,
+        0.1989498004708423f, 0.3037685446868245f, 0.09591203419338723f,
+        0.18896643794141466f, 0.290665510643898f, 0.08967253205630053f,
+        0.17898307541198705f, 0.2775624766009715f, 0.08343302991921385f,
+        0.1689997128825594f, 0.264459442558045f, 0.07719352778212715f,
+        0.15937190244916125f, 0.251711869932215f, 0.07059838287007802f,
+        0.15001241057839806f, 0.2392325474378641f, 0.06373485096423837f,
+        0.14065291870763433f, 0.22675322494351247f, 0.05687131905839832f,
+        0.1312934268368709f, 0.21427390244916117f, 0.050007787152558464f,
+        0.12193393496610741f, 0.20179457995480987f, 0.04314425524671861f,
+        0.11257444309534397f, 0.18931525746045863f, 0.036280723340878764f,
+        0.10357059399954295f, 0.17683598030557382f, 0.030483983741526836f,
+        0.09545916166628579f, 0.16435681692131152f, 0.02736415311793903f,
+        0.08734772933302863f, 0.1518776535370492f, 0.024244322494351222f,
+        0.07923629699977147f, 0.1393984901527869f, 0.021124491870763415f,
+        0.07112486466651433f, 0.1269193267685246f, 0.01800466124717561f,
+        0.06301343233325715f, 0.1144401633842623f, 0.014884830623587805f,
+        0.054902f, 0.101961f, 0.011765f
+    }, 
+    {
+        0.984313725490196f, 0.9647058823529412f, 0.7372549019607844f,
+        0.9827921757263818f, 0.9612823953843591f, 0.7273648284959916f,
+        0.9812706259625675f, 0.957858908415777f, 0.717474755031199f,
+        0.9797490761987533f, 0.954435421447195f, 0.7075846815664062f,
+        0.978227526434939f, 0.9510119344786129f, 0.6976946081016135f,
+        0.9767059766711247f, 0.9475884475100308f, 0.6878045346368208f,
+        0.9751844269073104f, 0.9441649605414487f, 0.6779144611720281f,
+        0.9736628771434962f, 0.9407414735728665f, 0.6680243877072354f,
+        0.972141327379682f, 0.9373179866042844f, 0.6581343142424427f,
+        0.9706197776158677f, 0.9338944996357024f, 0.64824424077765f,
+        0.9690982278520535f, 0.9304710126671203f, 0.6383541673128572f,
+        0.9670512916421627f, 0.925734059583347f, 0.6237356158333762f,
+        0.9647689669964413f, 0.9204086354099971f, 0.6069985684314194f,
+        0.9624866423507199f, 0.9150832112366472f, 0.5902615210294625f,
+        0.9602043177049986f, 0.9097577870632972f, 0.5735244736275057f,
+        0.9579219930592772f, 0.9044323628899473f, 0.5567874262255488f,
+        0.9556396684135557f, 0.8982915542602752f, 0.540458071051753f,
+        0.9533573437678344f, 0.8914445803231109f, 0.5244817985317033f,
+        0.951075019122113f, 0.8845976063859468f, 0.5085055260116536f,
+        0.9487926944763916f, 0.8777506324487826f, 0.4925292534916038f,
+        0.9465103698306702f, 0.8709036585116184f, 0.4765529809715541f,
+        0.9436480491644571f, 0.8640566845744543f, 0.4620266985027334f,
+        0.9398441747549214f, 0.8572097106372901f, 0.44985430039221935f,
+        0.9360403003453859f, 0.8503627367001259f, 0.4376819022817052f,
+        0.9322364259358502f, 0.8435157627629617f, 0.42550950417119116f,
+        0.9284325515263145f, 0.8366687888257975f, 0.4133371060606771f,
+        0.9248009809091094f, 0.8299941186809638f, 0.4023708344964765f,
+        0.9217578813814808f, 0.8239079196257069f, 0.3955238605593123f,
+        0.9187147818538524f, 0.8178217205704498f, 0.3886768866221481f,
+        0.9156716823262238f, 0.8117355215151927f, 0.38182991268498395f,
+        0.9126285827985953f, 0.8056493224599357f, 0.3749829387478198f,
+        0.9095854832709668f, 0.7995631234046786f, 0.3681905743850708f,
+        0.9065423837433382f, 0.7934769243494216f, 0.3621043753298137f,
+        0.9034992842157097f, 0.7873907252941645f, 0.35601817627455673f,
+        0.9004561846880812f, 0.7813045262389074f, 0.34993197721929964f,
+        0.8974130851604527f, 0.7752183271836505f, 0.3438457781640426f,
+        0.8943699856328242f, 0.7691321281283934f, 0.33775957910878557f,
+        0.8913268861051956f, 0.7623482388347295f, 0.3323710702919355f,
+        0.8882837865775671f, 0.7555012648975652f, 0.32704564611858555f,
+        0.8852406870499385f, 0.7486542909604009f, 0.32172022194523564f,
+        0.88219758752231f, 0.7418073170232368f, 0.3163947977718857f,
+        0.8791544879946815f, 0.7349603430860726f, 0.3110693735985358f,
+        0.8766913844875447f, 0.7292733611898917f, 0.30632394544567754f,
+        0.8744090598418233f, 0.7239479370165418f, 0.30175929615423475f,
+        0.8721267351961018f, 0.7186225128431919f, 0.29719464686279196f,
+        0.8698444105503805f, 0.7132970886698421f, 0.2926299975713492f,
+        0.8675620859046591f, 0.7079716644964922f, 0.2880653482799064f,
+        0.8651256606580789f, 0.702800340924001f, 0.2839630007910399f,
+        0.8625897443850551f, 0.6977285083779534f, 0.28015912638150425f,
+        0.8600538281120313f, 0.6926566758319059f, 0.27635525197196864f,
+        0.8588235294117647f, 0.687258438892669f, 0.2735305907420008f,
+        0.8588235294117647f, 0.6815526272783656f, 0.271628653537233f,
+        0.47016490869624633f, 0.24019165648847035f, 0.17072639619813998f,
+        0.47793154493285583f, 0.2510649472197236f, 0.17538637794010564f,
+        0.4856981811694654f, 0.26193823795097704f, 0.1800463596820714f,
+        0.4949028789265056f, 0.2728115286822304f, 0.18758246446489846f,
+        0.504222842410437f, 0.2836848194134837f, 0.195349100701508f,
+        0.5135428058943684f, 0.2951812575323605f, 0.20311573693811746f,
+        0.5228627693782999f, 0.3076078755109359f, 0.21088237317472708f,
+        0.5321827328622314f, 0.3200344934895112f, 0.2186490094113366f,
+        0.5428642568483011f, 0.3351842324723629f, 0.22913876665222266f,
+        0.5537375475795543f, 0.35071750494558185f, 0.24001205738347592f,
+        0.5646108383108077f, 0.3657041310494699f, 0.24979205537606716f,
+        0.5754841290420611f, 0.37968407627536704f, 0.2575586916126767f,
+        0.5863574197733145f, 0.3936640215012642f, 0.2653253278492862f,
+        0.595945651020722f, 0.4063589072433155f, 0.2730919640858957f,
+        0.6052656145046532f, 0.41878552522189055f, 0.2808586003225051f,
+        0.6150557233396234f, 0.4316822885515046f, 0.28862523655911476f,
+        0.6259290140708768f, 0.4456622337774018f, 0.2963918727957243f,
+        0.6368023048021301f, 0.459642179003299f, 0.3041585090323339f,
+        0.6476755955333834f, 0.47362212422919614f, 0.3143422622000501f,
+        0.6585488862646367f, 0.4876020694550932f, 0.32521555293130344f,
+        0.6698158213286363f, 0.5023693033464828f, 0.33569519932981057f,
+        0.6822424393072116f, 0.5194559030670238f, 0.34501516281374206f,
+        0.6946690572857868f, 0.5365425027875648f, 0.3543351262976735f,
+        0.7082277327116232f, 0.5536291025081059f, 0.364787147228866f,
+        0.7222076779375203f, 0.5707157022286469f, 0.37566043796011933f,
+        0.7361876231634175f, 0.5878023019491878f, 0.38653372869137265f,
+        0.7501675683893146f, 0.6048889016697289f, 0.397407019422626f,
+        0.7641475136152119f, 0.6219755013902698f, 0.40828031015387933f,
+        0.7770719024121405f, 0.6390621011108107f, 0.4191536008851327f,
+        0.7894985203907157f, 0.6561487008313517f, 0.430026891616386f,
+        0.8016844960731294f, 0.672994658255731f, 0.44090018234763917f,
+        0.8125577868043828f, 0.6885279307289504f, 0.4517734730788927f,
+        0.8234310775356363f, 0.7040612032021695f, 0.4626467638101461f,
+        0.8343043682668896f, 0.7205735310860647f, 0.47449910995207556f,
+        0.845177658998143f, 0.7376601308066056f, 0.48692572793065086f,
+        0.8560509497293963f, 0.7547467305271467f, 0.49984476974283293f,
+        0.8669242404606496f, 0.7718333302476875f, 0.5169313694633738f,
+        0.877797531191903f, 0.7889199299682285f, 0.5340179691839148f,
+        0.8886708219231562f, 0.8069090840811528f, 0.5529096776892228f,
+        0.8995441126544097f, 0.825549011049016f, 0.573102931904408f,
+        0.9102421228665101f, 0.8441012977573025f, 0.5937343874174753f,
+        0.9180087591031196f, 0.8611878974778435f, 0.6216942778692697f,
+        0.9257753953397292f, 0.8782744971983845f, 0.6496541683210642f,
+        0.9335420315763387f, 0.8945350435448343f, 0.6833964323914973f,
+        0.9413086678129482f, 0.9100683160180534f, 0.7222296135745451f,
+        0.9490864432908418f, 0.9256015884912725f, 0.7610627947575929f,
+        0.9584064067747732f, 0.9411348609644913f, 0.7998959759406398f,
+        0.9677263702587047f, 0.9566681334377105f, 0.8387291571236879f,
+        0.9777958860984349f, 0.9714518535551306f, 0.8760633085503734f,
+        0.9886691768296884f, 0.9854317987810279f, 0.9117899905715021f,
+        0.9995424675609417f, 0.999411744006925f, 0.9475166725926307f,
+        0.9305889327343205f, 0.958497068017665f, 0.9016780969008078f,
+        0.9030875168212531f, 0.9355793512916724f, 0.8726486513904852f,
+        0.8732106722226683f, 0.9090985672506915f, 0.8424315672506916f,
+        0.8395975275252713f, 0.8770134521505495f, 0.8103464521505496f,
+        0.8059843828278732f, 0.8449283370504066f, 0.7782613370504068f,
+        0.7811621641470541f, 0.8228898115459841f, 0.7549671479668429f,
+        0.7582440578163706f, 0.8030273748051202f, 0.7335770712334146f,
+        0.7364960805285846f, 0.784335067107154f, 0.7133570240928547f,
+        0.718162062989667f, 0.7690567191580558f, 0.6965506465465015f,
+        0.6998280454507493f, 0.7537783712089577f, 0.6797442690001482f,
+        0.6857017503437861f, 0.743759877497879f, 0.6681977456918142f,
+        0.673478682379817f, 0.7361205087209846f, 0.6590303473176646f,
+        0.6616279964169979f, 0.72848113994409f, 0.6500491874175731f,
+        0.652460987647539f, 0.7208417711671955f, 0.6424102082453693f,
+        0.64329397887808f, 0.713202402390301f, 0.6347712290731656f,
+        0.6324304105239394f, 0.7055630336134066f, 0.6262837538210657f,
+        0.62020734255997f, 0.697923664836512f, 0.617116355446916f,
+        0.6079842745960007f, 0.6902842960596176f, 0.6079489570727663f,
+        0.6002919272827231f, 0.6826449272827231f, 0.6033122793493083f,
+        0.5926525585058288f, 0.6750055585058288f, 0.5987285801622335f,
+        0.5843689262613266f, 0.6673661897289341f, 0.5935006175075511f,
+        0.5752019174918677f, 0.6597268209520397f, 0.5873892783279118f,
+        0.5660349087224088f, 0.6520874521751452f, 0.5812779391482725f,
+        0.5594805999686333f, 0.6444484165233324f, 0.5764727834139342f,
+        0.553369260788994f, 0.6368094373511287f, 0.5718890842268592f,
+        0.5463767700795689f, 0.6291703458444619f, 0.5668649216093548f,
+        0.5372093717054192f, 0.6215309770675673f, 0.5607535824297155f,
+        0.5280419733312695f, 0.6138916082906728f, 0.5546422432500763f,
+        0.5199772395137784f, 0.6051495749571197f, 0.5496332874785599f,
+        0.5123378707368841f, 0.5959821765829703f, 0.5450495882914851f,
+        0.5046985623179064f, 0.5870515019599892f, 0.5402292257111583f,
+        0.4970595831457028f, 0.5794121331830948f, 0.534117886531519f,
+        0.4894206039734991f, 0.5717727644062004f, 0.5280065473518798f,
+        0.48178139562930594f, 0.5641336248012955f, 0.5227937915431856f,
+        0.47414202685241147f, 0.5564946456290918f, 0.5182100923561108f,
+        0.466502658075517f, 0.548855658075517f, 0.5136264015504071f,
+        0.4588632892986226f, 0.5412162892986225f, 0.509043091968023f,
+        0.4512239205217281f, 0.5335769205217281f, 0.5044597823856389f,
+        0.44358455174483363f, 0.5259375517448337f, 0.4998762956078115f,
+        0.4359451829679392f, 0.5182981829679393f, 0.4952925964207367f,
+        0.4283058141910448f, 0.5106588141910448f, 0.49070889723366184f,
+        0.42066644541415016f, 0.5030194454141502f, 0.48476849473512645f,
+        0.4130270766372556f, 0.4953800766372556f, 0.47865715555548705f,
+        0.40440561604828484f, 0.4862676327516955f, 0.47205470786036097f,
+        0.39371057768157075f, 0.47404495439241695f, 0.4644153390834665f,
+        0.383015539314857f, 0.4618222760331388f, 0.45677597030657224f,
+        0.373473698255395f, 0.4507525009481426f, 0.44798369825539497f,
+        0.3643066894859361f, 0.4400574625814285f, 0.4388166894859361f,
+        0.35485242421471447f, 0.4290751677129517f, 0.4290751677129517f,
+        0.3441573858480004f, 0.4168520997489824f, 0.4168520997489824f,
+        0.3334623474812863f, 0.40462903178501314f, 0.40462903178501314f,
+        0.2611752601292498f, 0.3327286183995115f, 0.43173295706193f,
+        0.27131413479749517f, 0.3444273199397946f, 0.4449914854742508f,
+        0.28145300946574053f, 0.3561260214800777f, 0.45825001388657166f,
+        0.2915918841339859f, 0.3678247230203608f, 0.4715085422988925f,
+        0.30173075880223127f, 0.3795234245606439f, 0.4847670707112134f,
+        0.312330193511192f, 0.391222126100927f, 0.49756503908281885f,
+        0.32324898161545623f, 0.40292082764121007f, 0.5100436540591208f,
+        0.33416776971972045f, 0.4146195291814932f, 0.5225222690354228f,
+        0.3450865578239847f, 0.42631823072177627f, 0.5350008840117247f,
+        0.3560053459282489f, 0.4380169322620594f, 0.5474794989880267f,
+        0.3669241340325128f, 0.45015419239570087f, 0.5599581139643284f,
+        0.37784292213677734f, 0.46263280737200324f, 0.5724367289406307f,
+        0.38876171024104156f, 0.4751114223483052f, 0.5849153439169327f,
+        0.39968049834530583f, 0.48759003732460715f, 0.5973939588932347f,
+        0.41059928644957006f, 0.5000686523009091f, 0.6098725738695366f,
+        0.4219346316998364f, 0.5125472672772111f, 0.6223511888458386f,
+        0.43363333324011943f, 0.525025882253513f, 0.6348298038221405f,
+        0.44533203478040245f, 0.5375044972298149f, 0.6473084187984424f,
+        0.45703073632068547f, 0.5499831122061168f, 0.6597870337747443f,
+        0.46872943786096855f, 0.5624617271824187f, 0.6722656487510462f,
+        0.48082269509989706f, 0.5753348978573661f, 0.6851388194259936f,
+        0.493301310076199f, 0.5885934262696869f, 0.6983973478383144f,
+        0.5057799250525009f, 0.6018519546820077f, 0.7116558762506352f,
+        0.5182585400288028f, 0.6151104830943285f, 0.724914404662956f,
+        0.5307371550051047f, 0.6283690115066493f, 0.7381729330752768f,
+        0.5439608784839844f, 0.6420000941702589f, 0.7506863529850197f,
+        0.5579993203323239f, 0.6560385360185984f, 0.7623850545253026f,
+        0.572037762180664f, 0.6700769778669387f, 0.7740837560655861f,
+        0.5860762040290038f, 0.6841154197152784f, 0.7857824576058693f,
+        0.6001146458773436f, 0.6981538615636182f, 0.7974811591461525f,
+        0.6141530877256833f, 0.7121923034119579f, 0.8091798606864355f,
+        0.628191529574023f, 0.7262307452602976f, 0.8208785622267186f,
+        0.6422299714223627f, 0.7402691871086372f, 0.8325772637670017f,
+        0.6562684132707024f, 0.7543076289569769f, 0.8442759653072847f,
+        0.670306855119042f, 0.7683460708053165f, 0.8559746668475677f,
+        0.6836881942542303f, 0.7817274099405048f, 0.8670162656746994f,
+        0.6961668092305322f, 0.7942060249168067f, 0.8771551403429447f,
+        0.7086454242068342f, 0.8066846398931087f, 0.88729401501119f,
+        0.7211240391831361f, 0.8191632548694106f, 0.8974328896794352f,
+        0.7336026541594379f, 0.8316418698457124f, 0.9075717643476805f,
+        0.747000918863397f, 0.8444270347312335f, 0.9174040891067069f,
+        0.7618192741477557f, 0.8576855631435544f, 0.9267630503389334f,
+        0.7766376294321139f, 0.8709440915558749f, 0.9361220115711597f,
+        0.791455984716473f, 0.8842026199681962f, 0.9454809728033865f,
+        0.8062743400008316f, 0.8974611483805172f, 0.9548399340356131f,
+        0.825360922213127f, 0.9121424191021502f, 0.9641988952678395f,
+        0.8518779790377685f, 0.9293005146945653f, 0.9735578565000659f,
+        0.8783950358624101f, 0.9464586102869804f, 0.9829168177322923f,
+        0.9088930132634999f, 0.96228959744134f, 0.9896217611397445f,
+        0.940089425918105f, 0.977887749174702f, 0.9958610686278955f,
+        0.8706129528586563f, 0.9752876514848623f, 0.9504825480858896f,
+        0.8354721694692707f, 0.9600091092988845f, 0.9313843216528329f,
+        0.8003313860798852f, 0.9447305671129068f, 0.9122860952197762f,
+        0.7651906026904998f, 0.929452024926929f, 0.8931878687867195f,
+        0.7300498193011142f, 0.9141734827409513f, 0.8740896423536628f,
+        0.6964420460880201f, 0.8979750953567552f, 0.8551446973920135f,
+        0.6689404363904545f, 0.8781128541533475f, 0.8368102909269698f,
+        0.641438826692889f, 0.8582506129499399f, 0.818475884461926f,
+        0.6139372169953242f, 0.8383883717465328f, 0.8001414779968828f,
+        0.5864356072977578f, 0.8185261305431246f, 0.7818070715318386f,
+        0.5595506781709293f, 0.7993319512219257f, 0.7642435026760817f,
+        0.5412164665082237f, 0.789400928021391f, 0.7573681489522748f,
+        0.5228822548455181f, 0.7794699048208561f, 0.7504927952284679f,
+        0.5045480431828124f, 0.7695388816203214f, 0.743617441504661f,
+        0.48621383152010683f, 0.7596078584197865f, 0.7367420877808541f,
+        0.46787961985740123f, 0.7496768352192518f, 0.7298667340570473f,
+        0.4459781749242958f, 0.7311847432388193f, 0.7151437945910554f,
+        0.42382408421217305f, 0.7120863220034245f, 0.6998650576027395f,
+        0.4016699935000503f, 0.6929879007680295f, 0.6845863206144236f,
+        0.37951590278792746f, 0.6738894795326347f, 0.6693075836261078f,
+        0.35736181207580464f, 0.6547910582972398f, 0.6540288466377918f,
+        0.339488307730249f, 0.634469835142618f, 0.6381387086898624f,
+        0.3226819308059074f, 0.6138437739711975f, 0.6220961517335338f,
+        0.30587555388156584f, 0.5932177127997769f, 0.6060535947772051f,
+        0.2890691769572247f, 0.5725916516283569f, 0.5900110378208768f,
+        0.27226280003288267f, 0.5519655904569358f, 0.5739684808645477f,
+        0.257494556248798f, 0.5333777923673176f, 0.5584355546494423f,
+        0.24374365399884607f, 0.5158074006726249f, 0.5431570124634645f,
+        0.2299927517488941f, 0.49823700897793216f, 0.5278784702774868f,
+        0.21624184949894218f, 0.4806666172832394f, 0.5125999280915091f,
+        0.20249094724899022f, 0.4630962255885467f, 0.4973213859055314f,
+        0.18955545418499564f, 0.4467488437195536f, 0.48285814895227464f,
+        0.17733258147574582f, 0.4314703015335759f, 0.4691074415046609f,
+        0.165109708766496f, 0.41619175934759817f, 0.4553567340570471f,
+        0.15288683605724618f, 0.4009132171616204f, 0.4416060266094333f,
+        0.14066396334799636f, 0.38563467497564274f, 0.4278553191618195f,
+        0.13027560537729616f, 0.3728020484878281f, 0.41655052741236887f,
+        0.12263643168547643f, 0.36363484525530626f, 0.40891115891821095f,
+        0.11499725799365669f, 0.3544676420227844f, 0.40127179042405303f,
+        0.10735808430183696f, 0.3453004387902625f, 0.39363242192989517f,
+        0.09971891061001745f, 0.33613323555774094f, 0.3859930534357375f,
+        0.09167203232521882f, 0.3259468747960085f, 0.3773345793889872f,
+        0.08250482909269696f, 0.3129599873164077f, 0.3658757214500885f,
+        0.07333762586017512f, 0.29997309983680687f, 0.3544168635111898f,
+        0.06417042262765325f, 0.28698621235720606f, 0.3429580055722911f,
+        0.0550032193951314f, 0.27399932487760525f, 0.33149914763339244f,
+        0.04583601616260954f, 0.2608086110898242f, 0.31973451124975977f,
+        0.03666881293008763f, 0.24629388887185935f, 0.3059836089998078f,
+        0.027501609697565724f, 0.23177916665389453f, 0.29223270674985585f,
+        0.018334406465043813f, 0.21726444443592968f, 0.2784818044999039f,
+        0.00916720323252191f, 0.20274972221796486f, 0.26473090224995194f,
+        0.0f, 0.188235f, 0.25098f
+    }, 
+    {
+        0.011765f, 0.015686f, 0.101961f,
+        0.013918019607843137f, 0.018146588235294116f, 0.10580566666666666f,
+        0.016071039215686273f, 0.020607176470588233f, 0.10965033333333334f,
+        0.018224058823529413f, 0.02306776470588235f, 0.113495f,
+        0.02037707843137255f, 0.025528352941176467f, 0.11733966666666666f,
+        0.022530098039215685f, 0.027988941176470584f, 0.12118433333333334f,
+        0.024683117647058824f, 0.0304495294117647f, 0.125029f,
+        0.026836137254901964f, 0.032910117647058826f, 0.12887366666666666f,
+        0.0289891568627451f, 0.03537070588235294f, 0.13271833333333333f,
+        0.031142176470588236f, 0.03783129411764706f, 0.136563f,
+        0.03329519607843137f, 0.04029188235294118f, 0.14040766666666665f,
+        0.035448215686274515f, 0.042752470588235295f, 0.14425233333333334f,
+        0.03760123529411765f, 0.04521305882352941f, 0.148097f,
+        0.03975425490196079f, 0.04767364705882352f, 0.15194166666666667f,
+        0.04190727450980393f, 0.050134235294117646f, 0.15578633333333333f,
+        0.04406029411764706f, 0.05259482352941176f, 0.15963100000000002f,
+        0.046213313725490195f, 0.05505541176470588f, 0.16347566666666669f,
+        0.04836633333333333f, 0.05751599999999999f, 0.16732033333333335f,
+        0.050519352941176474f, 0.059976588235294115f, 0.171165f,
+        0.05267237254901961f, 0.06243717647058823f, 0.17500966666666667f,
+        0.054825392156862746f, 0.06489776470588235f, 0.17885433333333334f,
+        0.05697841176470588f, 0.06735835294117645f, 0.182699f,
+        0.059131431372549025f, 0.06981894117647058f, 0.18654366666666666f,
+        0.06128445098039216f, 0.0722795294117647f, 0.19038833333333333f,
+        0.0634374705882353f, 0.07474011764705882f, 0.194233f,
+        0.06559049019607843f, 0.07720070588235294f, 0.19807766666666668f,
+        0.06766660784313726f, 0.08019954901960784f, 0.20238370588235297f,
+        0.06966582352941177f, 0.08373664705882353f, 0.20715111764705885f,
+        0.07166503921568627f, 0.08727374509803922f, 0.21191852941176473f,
+        0.07366425490196078f, 0.0908108431372549f, 0.2166859411764706f,
+        0.0756634705882353f, 0.09434794117647059f, 0.2214533529411765f,
+        0.07766268627450981f, 0.09788503921568628f, 0.22622076470588237f,
+        0.07966190196078432f, 0.10142213725490196f, 0.23098817647058822f,
+        0.08166111764705881f, 0.10495923529411763f, 0.23575558823529408f,
+        0.08366033333333334f, 0.10849633333333333f, 0.240523f,
+        0.08565954901960784f, 0.11203343137254902f, 0.2452904117647059f,
+        0.08765876470588235f, 0.11557052941176471f, 0.25005782352941175f,
+        0.08965798039215685f, 0.11910762745098036f, 0.2548252352941176f,
+        0.09165719607843137f, 0.12264472549019607f, 0.2595926470588235f,
+        0.09365641176470588f, 0.12618182352941176f, 0.2643600588235294f,
+        0.0956556274509804f, 0.12971892156862747f, 0.2691274705882353f,
+        0.0976548431372549f, 0.1332560196078431f, 0.2738948823529411f,
+        0.09965405882352942f, 0.13679311764705882f, 0.2786622941176471f,
+        0.10165327450980392f, 0.14033021568627452f, 0.28342970588235294f,
+        0.10365249019607843f, 0.1438673137254902f, 0.2881971176470588f,
+        0.10565170588235293f, 0.14740441176470587f, 0.29296452941176465f,
+        0.10765092156862746f, 0.15094150980392157f, 0.2977319411764706f,
+        0.10965013725490197f, 0.15447860784313727f, 0.30249935294117647f,
+        0.11164935294117648f, 0.15801570588235292f, 0.3072667647058823f,
+        0.11364856862745096f, 0.16155280392156862f, 0.3120341764705882f,
+        0.1156477843137255f, 0.16508990196078432f, 0.31680158823529414f,
+        0.117647f, 0.168627f, 0.321569f,
+        0.11949245098039216f, 0.17124139215686274f, 0.3246447254901961f,
+        0.1213379019607843f, 0.17385578431372548f, 0.32772045098039215f,
+        0.12318335294117647f, 0.17647017647058824f, 0.33079617647058823f,
+        0.12502880392156862f, 0.17908456862745098f, 0.3338719019607843f,
+        0.12687425490196078f, 0.18169896078431372f, 0.3369476274509804f,
+        0.12871970588235293f, 0.18431335294117646f, 0.34002335294117647f,
+        0.13056515686274509f, 0.18692774509803922f, 0.34309907843137255f,
+        0.13241060784313724f, 0.18954213725490196f, 0.34617480392156863f,
+        0.13425605882352942f, 0.1921565294117647f, 0.3492505294117647f,
+        0.13610150980392155f, 0.19477092156862744f, 0.3523262549019608f,
+        0.13794696078431373f, 0.1973853137254902f, 0.35540198039215687f,
+        0.1397924117647059f, 0.19999970588235294f, 0.35847770588235295f,
+        0.14163786274509804f, 0.20261409803921568f, 0.36155343137254903f,
+        0.1434833137254902f, 0.20522849019607844f, 0.3646291568627451f,
+        0.14532876470588232f, 0.20784288235294113f, 0.36770488235294113f,
+        0.1471742156862745f, 0.21045727450980392f, 0.37078060784313727f,
+        0.14901966666666666f, 0.21307166666666666f, 0.37385633333333335f,
+        0.15086511764705882f, 0.21568605882352943f, 0.3769320588235294f,
+        0.15271056862745097f, 0.21830045098039216f, 0.3800077843137255f,
+        0.15455601960784313f, 0.2209148431372549f, 0.3830835098039216f,
+        0.15640147058823528f, 0.22352923529411764f, 0.38615923529411766f,
+        0.15824692156862744f, 0.2261436274509804f, 0.38923496078431374f,
+        0.1600923725490196f, 0.2287580196078431f, 0.39231068627450977f,
+        0.16193782352941177f, 0.23137241176470588f, 0.3953864117647059f,
+        0.1637832745098039f, 0.23398680392156862f, 0.398462137254902f,
+        0.16609007843137255f, 0.23721633333333333f, 0.4019992352941177f,
+        0.16885823529411764f, 0.241061f, 0.40599770588235295f,
+        0.17162639215686273f, 0.24490566666666666f, 0.4099961764705883f,
+        0.17439454901960783f, 0.24875033333333332f, 0.41399464705882355f,
+        0.17716270588235294f, 0.252595f, 0.4179931176470588f,
+        0.179930862745098f, 0.2564396666666666f, 0.4219915882352941f,
+        0.18269901960784313f, 0.26028433333333334f, 0.4259900588235294f,
+        0.18546717647058822f, 0.264129f, 0.42998852941176474f,
+        0.18823533333333334f, 0.26797366666666667f, 0.433987f,
+        0.19100349019607843f, 0.27181833333333333f, 0.4379854705882353f,
+        0.19377164705882352f, 0.275663f, 0.4419839411764706f,
+        0.1965398039215686f, 0.27950766666666665f, 0.4459824117647059f,
+        0.19930796078431373f, 0.2833523333333333f, 0.4499808823529412f,
+        0.2020761176470588f, 0.2871969999999999f, 0.4539793529411764f,
+        0.20484427450980391f, 0.29104166666666664f, 0.45797782352941174f,
+        0.207612431372549f, 0.2948863333333333f, 0.46197629411764707f,
+        0.21038058823529412f, 0.29873099999999997f, 0.46597476470588234f,
+        0.21314874509803922f, 0.3025756666666667f, 0.46997323529411766f,
+        0.2159169019607843f, 0.3064203333333333f, 0.47397170588235293f,
+        0.2186850588235294f, 0.310265f, 0.47797017647058826f,
+        0.22145321568627452f, 0.3141096666666667f, 0.48196864705882353f,
+        0.22422137254901958f, 0.3179543333333333f, 0.48596711764705874f,
+        0.2269895294117647f, 0.321799f, 0.4899655882352941f,
+        0.22975768627450982f, 0.32564366666666666f, 0.4939640588235294f,
+        0.2325258431372549f, 0.3294883333333333f, 0.4979625294117647f,
+        0.235294f, 0.333333f, 0.501961f,
+        0.23744701960784315f, 0.3365625294117647f, 0.5050367254901961f,
+        0.2396000392156863f, 0.3397920588235294f, 0.5081124509803921f,
+        0.2417530588235294f, 0.3430215882352941f, 0.5111881764705882f,
+        0.24390607843137252f, 0.3462511176470588f, 0.5142639019607843f,
+        0.2460590980392157f, 0.34948064705882353f, 0.5173396274509804f,
+        0.24821211764705883f, 0.3527101764705882f, 0.5204153529411765f,
+        0.25036513725490195f, 0.3559397058823529f, 0.5234910784313725f,
+        0.2525181568627451f, 0.35916923529411765f, 0.5265668039215686f,
+        0.25467117647058823f, 0.36239876470588234f, 0.5296425294117647f,
+        0.2568241960784314f, 0.3656282941176471f, 0.5327182549019608f,
+        0.2589772156862745f, 0.36885782352941177f, 0.5357939803921569f,
+        0.2611302352941176f, 0.3720873529411764f, 0.538869705882353f,
+        0.2632832549019608f, 0.37531688235294114f, 0.541945431372549f,
+        0.26543627450980395f, 0.3785464117647059f, 0.5450211568627451f,
+        0.2675892941176471f, 0.38177594117647057f, 0.5480968823529412f,
+        0.2697423137254902f, 0.3850054705882353f, 0.5511726078431373f,
+        0.2718953333333333f, 0.388235f, 0.5542483333333333f,
+        0.27404835294117647f, 0.3914645294117647f, 0.5573240588235294f,
+        0.2762013725490196f, 0.3946940588235294f, 0.5603997843137255f,
+        0.27835439215686275f, 0.39792358823529406f, 0.5634755098039216f,
+        0.2805074117647059f, 0.4011531176470588f, 0.5665512352941177f,
+        0.28266043137254904f, 0.40438264705882354f, 0.5696269607843137f,
+        0.2848134509803922f, 0.40761217647058823f, 0.5727026862745098f,
+        0.2869664705882353f, 0.4108417058823529f, 0.5757784117647059f,
+        0.2891194901960784f, 0.41407123529411766f, 0.578854137254902f,
+        0.2915031960784314f, 0.41737766666666665f, 0.5817760784313726f,
+        0.2941175882352941f, 0.420761f, 0.5845442352941177f,
+        0.29673198039215687f, 0.42414433333333335f, 0.5873123921568628f,
+        0.29934637254901963f, 0.42752766666666664f, 0.5900805490196078f,
+        0.3019607647058823f, 0.4309109999999999f, 0.5928487058823528f,
+        0.3045751568627451f, 0.43429433333333334f, 0.5956168627450981f,
+        0.3071895490196078f, 0.4376776666666667f, 0.5983850196078432f,
+        0.3098039411764706f, 0.441061f, 0.6011531764705883f,
+        0.31241833333333335f, 0.44444433333333333f, 0.6039213333333333f,
+        0.31503272549019606f, 0.4478276666666667f, 0.6066894901960784f,
+        0.31764711764705883f, 0.451211f, 0.6094576470588235f,
+        0.32026150980392154f, 0.4545943333333333f, 0.6122258039215687f,
+        0.3228759019607843f, 0.45797766666666667f, 0.6149939607843138f,
+        0.32549029411764707f, 0.461361f, 0.6177621176470588f,
+        0.3281046862745098f, 0.4647443333333333f, 0.6205302745098039f,
+        0.33071907843137255f, 0.46812766666666666f, 0.623298431372549f,
+        0.3333334705882353f, 0.471511f, 0.6260665882352942f,
+        0.335947862745098f, 0.4748943333333333f, 0.6288347450980393f,
+        0.3385622549019608f, 0.47827766666666666f, 0.6316029019607843f,
+        0.3411766470588235f, 0.481661f, 0.6343710588235294f,
+        0.3437910392156862f, 0.48504433333333324f, 0.6371392156862744f,
+        0.34640543137254903f, 0.48842766666666665f, 0.6399073725490196f,
+        0.34901982352941174f, 0.491811f, 0.6426755294117648f,
+        0.3516342156862745f, 0.4951943333333333f, 0.6454436862745098f,
+        0.3542486078431373f, 0.49857766666666664f, 0.6482118431372549f,
+        0.356863f, 0.501961f, 0.65098f,
+        0.3600925294117647f, 0.5062670392156863f, 0.6540557647058823f,
+        0.3633220588235294f, 0.5105730784313726f, 0.6571315294117647f,
+        0.3665515882352941f, 0.5148791176470588f, 0.660207294117647f,
+        0.3697811176470588f, 0.5191851568627451f, 0.6632830588235294f,
+        0.37301064705882353f, 0.5234911960784314f, 0.6663588235294118f,
+        0.3762401764705882f, 0.5277972352941176f, 0.6694345882352941f,
+        0.37946970588235296f, 0.5321032745098039f, 0.6725103529411764f,
+        0.38269923529411765f, 0.5364093137254902f, 0.6755861176470588f,
+        0.38592876470588233f, 0.5407153529411765f, 0.6786618823529411f,
+        0.389158294117647f, 0.5450213921568627f, 0.6817376470588236f,
+        0.39238782352941165f, 0.5493274313725489f, 0.6848134117647058f,
+        0.39561735294117645f, 0.5536334705882353f, 0.6878891764705882f,
+        0.3988468823529412f, 0.5579395098039216f, 0.6909649411764706f,
+        0.4020764117647059f, 0.5622455490196079f, 0.6940407058823529f,
+        0.40530594117647056f, 0.5665515882352941f, 0.6971164705882352f,
+        0.4085354705882353f, 0.5708576274509803f, 0.7001922352941177f,
+        0.411765f, 0.5751636666666666f, 0.703268f,
+        0.4149945294117647f, 0.5794697058823529f, 0.7063437647058823f,
+        0.4182240588235294f, 0.5837757450980392f, 0.7094195294117647f,
+        0.4214535882352941f, 0.5880817843137255f, 0.712495294117647f,
+        0.4246831176470588f, 0.5923878235294118f, 0.7155710588235293f,
+        0.42791264705882354f, 0.5966938627450981f, 0.7186468235294117f,
+        0.4311421764705882f, 0.6009999019607843f, 0.7217225882352941f,
+        0.43437170588235297f, 0.6053059411764706f, 0.7247983529411764f,
+        0.43760123529411765f, 0.6096119803921569f, 0.7278741176470588f,
+        0.44083076470588234f, 0.6138411176470588f, 0.7307960784313725f,
+        0.44406029411764697f, 0.6179933529411764f, 0.7335642352941175f,
+        0.44728982352941177f, 0.6221455882352941f, 0.7363323921568627f,
+        0.45051935294117645f, 0.6262978235294118f, 0.7391005490196078f,
+        0.45374888235294114f, 0.6304500588235294f, 0.7418687058823529f,
+        0.4569784117647059f, 0.634602294117647f, 0.744636862745098f,
+        0.46020794117647057f, 0.6387545294117647f, 0.7474050196078431f,
+        0.46343747058823526f, 0.6429067647058824f, 0.7501731764705882f,
+        0.466667f, 0.647059f, 0.7529413333333334f,
+        0.4698965294117647f, 0.6512112352941176f, 0.7557094901960784f,
+        0.47312605882352937f, 0.6553634705882353f, 0.7584776470588235f,
+        0.4763555882352941f, 0.659515705882353f, 0.7612458039215686f,
+        0.4795851176470588f, 0.6636679411764707f, 0.7640139607843137f,
+        0.4828146470588235f, 0.6678201764705882f, 0.7667821176470588f,
+        0.4860441764705882f, 0.6719724117647059f, 0.769550274509804f,
+        0.4892737058823529f, 0.6761246470588236f, 0.7723184313725491f,
+        0.4925032352941176f, 0.6802768823529413f, 0.7750865882352941f,
+        0.49573276470588223f, 0.6844291176470587f, 0.7778547450980392f,
+        0.49896229411764703f, 0.6885813529411765f, 0.7806229019607843f,
+        0.5021918235294117f, 0.6927335882352942f, 0.7833910588235294f,
+        0.5054213529411764f, 0.6968858235294118f, 0.7861592156862746f,
+        0.5086508823529411f, 0.7010380588235294f, 0.7889273725490197f,
+        0.5118804117647058f, 0.7051902941176471f, 0.7916955294117647f,
+        0.5151099411764706f, 0.7093425294117648f, 0.7944636862745098f,
+        0.5183394705882353f, 0.7134947647058824f, 0.7972318431372549f,
+        0.521569f, 0.717647f, 0.8f,
+        0.5263363725490195f, 0.7221068431372549f, 0.8030757254901961f,
+        0.5311037450980391f, 0.7265666862745098f, 0.8061514509803922f,
+        0.5358711176470587f, 0.7310265294117647f, 0.8092271764705883f,
+        0.5406384901960783f, 0.7354863725490196f, 0.8123029019607844f,
+        0.545405862745098f, 0.7399462156862745f, 0.8153786274509804f,
+        0.5501732352941175f, 0.7444060588235294f, 0.8184543529411765f,
+        0.5549406078431373f, 0.7488659019607844f, 0.8215300784313726f,
+        0.5597079803921566f, 0.7533257450980391f, 0.8246058039215686f,
+        0.5644753529411765f, 0.7577855882352942f, 0.8276815294117648f,
+        0.569242725490196f, 0.7622454313725491f, 0.8307572549019608f,
+        0.5740100980392157f, 0.766705274509804f, 0.8338329803921569f,
+        0.5787774705882353f, 0.7711651176470589f, 0.836908705882353f,
+        0.5835448431372549f, 0.7756249607843138f, 0.839984431372549f,
+        0.5883122156862745f, 0.7800848039215686f, 0.8430601568627452f,
+        0.593079588235294f, 0.7845446470588235f, 0.8461358823529411f,
+        0.5978469607843137f, 0.7890044901960784f, 0.8492116078431372f,
+        0.6026143333333333f, 0.7934643333333333f, 0.8522873333333333f,
+        0.607381705882353f, 0.7979241764705882f, 0.8553630588235294f,
+        0.6121490784313726f, 0.8023840196078431f, 0.8584387843137254f,
+        0.6169164509803922f, 0.806843862745098f, 0.8615145098039215f,
+        0.6216838235294118f, 0.811303705882353f, 0.8645902352941176f,
+        0.6264511960784314f, 0.8157635490196079f, 0.8676659607843137f,
+        0.631218568627451f, 0.8202233921568628f, 0.8707416862745098f,
+        0.6359859411764704f, 0.8246832352941176f, 0.8738174117647057f,
+        0.6407533137254902f, 0.8291430784313726f, 0.8768931372549019f,
+        0.6485195490196078f, 0.835063862745098f, 0.8811991960784313f,
+        0.6592846470588235f, 0.8424455882352941f, 0.8867355882352941f,
+        0.6700497450980392f, 0.8498273137254903f, 0.8922719803921568f,
+        0.6808148431372548f, 0.8572090392156864f, 0.8978083725490196f,
+        0.6915799411764706f, 0.8645907647058824f, 0.9033447647058823f,
+        0.7023450392156863f, 0.8719724901960785f, 0.9088811568627451f,
+        0.7131101372549019f, 0.8793542156862746f, 0.9144175490196078f,
+        0.7238752352941176f, 0.8867359411764706f, 0.9199539411764706f,
+        0.7346403333333333f, 0.8941176666666667f, 0.9254903333333333f,
+        0.745405431372549f, 0.9014993921568628f, 0.931026725490196f,
+        0.7561705294117647f, 0.9088811176470588f, 0.9365631176470588f,
+        0.7669356274509804f, 0.9162628431372549f, 0.9420995098039215f,
+        0.7777007254901961f, 0.923644568627451f, 0.9476359019607843f,
+        0.7891578823529412f, 0.9298729411764706f, 0.9520188235294117f,
+        0.8008457254901957f, 0.9357168627450979f, 0.9560172549019607f,
+        0.812533568627451f, 0.9415607843137255f, 0.9600156862745098f,
+        0.8242214117647059f, 0.9474047058823529f, 0.9640141176470588f,
+        0.8359092549019608f, 0.9532486274509804f, 0.9680125490196079f,
+        0.8475970980392157f, 0.9590925490196078f, 0.9720109803921568f,
+        0.8592849411764706f, 0.9649364705882353f, 0.9760094117647059f,
+        0.8709727843137255f, 0.9707803921568627f, 0.9800078431372549f,
+        0.8826606274509804f, 0.9766243137254902f, 0.9840062745098039f,
+        0.8943484705882353f, 0.9824682352941176f, 0.9880047058823529f,
+        0.9060363137254902f, 0.9883121568627451f, 0.992003137254902f,
+        0.9177241568627451f, 0.9941560784313725f, 0.9960015686274509f,
+        0.929412f, 1.0f, 1.0f
+    }
+, 
+{
+0.8f, 1.0f, 0.984314f,
+0.7684737254901961f, 0.9938484313725491f, 0.9797002745098039f,
+0.7369474509803922f, 0.987696862745098f, 0.9750865490196079f,
+0.7054211764705882f, 0.981545294117647f, 0.9704728235294118f,
+0.6738949019607844f, 0.9753937254901961f, 0.9658590980392157f,
+0.6423686274509804f, 0.9692421568627452f, 0.9612453725490195f,
+0.629527411764706f, 0.9626293529411765f, 0.9570931764705882f,
+0.6187623137254902f, 0.9559653006535948f, 0.9529922614379085f,
+0.6079972156862745f, 0.949301248366013f, 0.9488913464052288f,
+0.5972321176470589f, 0.9426371960784313f, 0.944790431372549f,
+0.5864670196078432f, 0.9359731437908497f, 0.9406895163398693f,
+0.5757019215686274f, 0.929309091503268f, 0.9365886013071896f,
+0.5649368235294118f, 0.9226450392156863f, 0.9324876862745098f,
+0.5553251176470588f, 0.9163398039215687f, 0.9288737450980392f,
+0.5491735882352942f, 0.9111110196078431f, 0.9267207254901961f,
+0.5430220588235294f, 0.9058822352941176f, 0.9245677058823529f,
+0.5368705294117647f, 0.9006534509803922f, 0.9224146862745098f,
+0.5307189999999999f, 0.8954246666666666f, 0.9202616666666666f,
+0.5245674705882353f, 0.8901958823529411f, 0.9181086470588236f,
+0.5184159411764706f, 0.8849670980392157f, 0.9159556274509804f,
+0.5122644117647058f, 0.8797383137254902f, 0.9138026078431373f,
+0.5061128823529412f, 0.8745095294117646f, 0.9116495882352942f,
+0.4999613529411765f, 0.8692807450980392f, 0.909496568627451f,
+0.4938098235294118f, 0.8640519607843137f, 0.9073435490196079f,
+0.48765829411764705f, 0.8588231764705883f, 0.9051905294117647f,
+0.48150676470588233f, 0.8535943921568627f, 0.9030375098039216f,
+0.4753552745098039f, 0.8480580784313725f, 0.9007306862745098f,
+0.46920382352941176f, 0.8422142352941177f, 0.8982700588235294f,
+0.4630523725490196f, 0.8363703921568627f, 0.895809431372549f,
+0.45690092156862744f, 0.8305265490196079f, 0.8933488039215687f,
+0.4507494705882353f, 0.8246827058823529f, 0.8908881764705883f,
+0.4445980196078431f, 0.818838862745098f, 0.8884275490196079f,
+0.438446568627451f, 0.8129950196078432f, 0.8859669215686274f,
+0.43229511764705886f, 0.8071511764705883f, 0.8835062941176471f,
+0.4261436666666667f, 0.8013073333333334f, 0.8810456666666667f,
+0.41999221568627454f, 0.7954634901960784f, 0.8785850392156863f,
+0.4138407647058824f, 0.7896196470588236f, 0.8761244117647059f,
+0.4076893137254903f, 0.7837758039215686f, 0.8736637843137255f,
+0.40153786274509806f, 0.7779319607843137f, 0.8712031568627451f,
+0.39653976470588237f, 0.772318705882353f, 0.8694345882352942f,
+0.3919261176470588f, 0.7667823137254902f, 0.867896705882353f,
+0.3873124705882353f, 0.7612459215686275f, 0.8663588235294117f,
+0.38269882352941176f, 0.7557095294117647f, 0.8648209411764706f,
+0.37808517647058826f, 0.750173137254902f, 0.8632830588235294f,
+0.3734715294117647f, 0.7446367450980392f, 0.8617451764705882f,
+0.3688578823529412f, 0.7391003529411765f, 0.8602072941176471f,
+0.36424423529411765f, 0.7335639607843137f, 0.8586694117647059f,
+0.3596305882352941f, 0.728027568627451f, 0.8571315294117647f,
+0.3550169411764706f, 0.7224911764705882f, 0.8555936470588235f,
+0.3504032941176471f, 0.7169547843137255f, 0.8540557647058823f,
+0.34578964705882353f, 0.7114183921568628f, 0.8525178823529411f,
+0.341176f, 0.705882f, 0.85098f,
+0.336562431372549f, 0.7006532941176471f, 0.8494421960784313f,
+0.33194886274509805f, 0.6954245882352942f, 0.8479043921568628f,
+0.32733529411764706f, 0.6901958823529412f, 0.8463665882352941f,
+0.32272172549019607f, 0.6849671764705882f, 0.8448287843137254f,
+0.3181081568627451f, 0.6797384705882353f, 0.8432909803921569f,
+0.31349458823529414f, 0.6745097647058824f, 0.8417531764705882f,
+0.30888101960784314f, 0.6692810588235294f, 0.8402153725490196f,
+0.30426745098039215f, 0.6640523529411765f, 0.838677568627451f,
+0.29965388235294116f, 0.6588236470588236f, 0.8371397647058824f,
+0.2950403137254902f, 0.6535949411764707f, 0.8356019607843137f,
+0.29042674509803923f, 0.6483662352941176f, 0.8340641568627452f,
+0.28581317647058824f, 0.6431375294117647f, 0.8325263529411765f,
+0.2812764901960785f, 0.6379857058823529f, 0.8309885294117647f,
+0.2769704509803922f, 0.6330645294117647f, 0.8294506470588235f,
+0.27266441176470596f, 0.6281433529411765f, 0.8279127647058824f,
+0.2683583725490196f, 0.6232221764705882f, 0.8263748823529412f,
+0.26405233333333333f, 0.618301f, 0.824837f,
+0.25974629411764705f, 0.6133798235294118f, 0.8232991176470589f,
+0.25544025490196076f, 0.6084586470588235f, 0.8217612352941176f,
+0.25113421568627453f, 0.6035374705882353f, 0.8202233529411764f,
+0.24682817647058825f, 0.598616294117647f, 0.8186854705882353f,
+0.24252213725490196f, 0.5936951176470588f, 0.8171475882352941f,
+0.23821609803921573f, 0.5887739411764706f, 0.8156097058823529f,
+0.2339100588235294f, 0.5838527647058823f, 0.8140718235294118f,
+0.22960401960784313f, 0.5789315882352941f, 0.8125339411764706f,
+0.22529798039215684f, 0.5735490392156862f, 0.8108422549019607f,
+0.22099194117647059f, 0.5677051176470588f, 0.8089967647058823f,
+0.2166859019607843f, 0.5618611960784313f, 0.8071512745098038f,
+0.21237986274509804f, 0.5560172745098039f, 0.8053057843137255f,
+0.20807382352941176f, 0.5501733529411764f, 0.8034602941176471f,
+0.20376778431372555f, 0.5443294313725491f, 0.8016148039215686f,
+0.1994617450980392f, 0.5384855098039215f, 0.7997693137254902f,
+0.19515570588235293f, 0.5326415882352941f, 0.7979238235294117f,
+0.19084966666666667f, 0.5267976666666666f, 0.7960783333333333f,
+0.1865436274509804f, 0.5209537450980392f, 0.794232843137255f,
+0.18223758823529412f, 0.5151098235294117f, 0.7923873529411765f,
+0.17793154901960784f, 0.5092659019607844f, 0.7905418627450981f,
+0.17362550980392158f, 0.5034219803921569f, 0.7886963725490196f,
+0.1695501764705883f, 0.49803941176470595f, 0.7870815882352942f,
+0.16555174509803922f, 0.49281062745098037f, 0.785543705882353f,
+0.1615533137254902f, 0.4875818431372549f, 0.7840058235294117f,
+0.15755488235294118f, 0.4823530588235294f, 0.7824679411764706f,
+0.15355645098039217f, 0.47712427450980394f, 0.7809300588235294f,
+0.14955801960784315f, 0.47189549019607846f, 0.7793921764705882f,
+0.1455595882352941f, 0.46666670588235293f, 0.7778542941176471f,
+0.1415611568627451f, 0.46143792156862745f, 0.7763164117647059f,
+0.13756272549019613f, 0.45620913725490203f, 0.7747785294117647f,
+0.13356429411764706f, 0.4509803529411765f, 0.7732406470588234f,
+0.12956586274509804f, 0.44575156862745097f, 0.7717027647058823f,
+0.12556743137254903f, 0.4405227843137255f, 0.7701648823529411f,
+0.121569f, 0.435294f, 0.768627f,
+0.1191083725490196f, 0.42975768627450983f, 0.7670891960784313f,
+0.11664774509803921f, 0.4242213725490196f, 0.7655513921568627f,
+0.11418711764705883f, 0.4186850588235294f, 0.7640135882352941f,
+0.11172649019607847f, 0.4131487450980393f, 0.7624757843137255f,
+0.10926586274509803f, 0.40761243137254904f, 0.7609379803921569f,
+0.10680523529411765f, 0.4020761176470588f, 0.7594001764705882f,
+0.10434460784313725f, 0.3965398039215686f, 0.7578623725490196f,
+0.10188398039215686f, 0.39100349019607844f, 0.756324568627451f,
+0.09942335294117646f, 0.3854671764705882f, 0.7547867647058824f,
+0.09696272549019608f, 0.379930862745098f, 0.7532489607843137f,
+0.09450209803921568f, 0.37439454901960784f, 0.7517111568627451f,
+0.09204147058823534f, 0.3688582352941177f, 0.7501733529411765f,
+0.0901703725490196f, 0.36437279084967317f, 0.7488918431372549f,
+0.09006786274509804f, 0.363039954248366f, 0.7483792156862745f,
+0.08996535294117647f, 0.3617071176470588f, 0.7478665882352942f,
+0.0898628431372549f, 0.3603742810457516f, 0.7473539607843137f,
+0.08976033333333333f, 0.3590414444444444f, 0.7468413333333334f,
+0.08965782352941176f, 0.35770860784313724f, 0.746328705882353f,
+0.0895553137254902f, 0.35637577124183006f, 0.7458160784313725f,
+0.08945280392156862f, 0.3550429346405229f, 0.7453034509803922f,
+0.08935029411764706f, 0.35371009803921566f, 0.7447908235294117f,
+0.08924778431372549f, 0.3523772614379085f, 0.7442781960784314f,
+0.08914527450980392f, 0.35104442483660125f, 0.743765568627451f,
+0.08904276470588235f, 0.3497115882352941f, 0.7432529411764706f,
+0.08894025490196078f, 0.3483787516339869f, 0.7427403137254902f,
+0.08883774509803921f, 0.34704591503267973f, 0.7422276862745097f,
+0.08873523529411764f, 0.3457130784313725f, 0.7417150588235294f,
+0.08863272549019607f, 0.3443802418300653f, 0.741202431372549f,
+0.0885302156862745f, 0.34304740522875815f, 0.7406898039215686f,
+0.08842770588235295f, 0.341714568627451f, 0.7401771764705882f,
+0.08832519607843137f, 0.34038173202614375f, 0.7396645490196079f,
+0.0882226862745098f, 0.33904889542483657f, 0.7391519215686274f,
+0.08812017647058824f, 0.3377160588235294f, 0.7386392941176471f,
+0.08801766666666667f, 0.3363832222222222f, 0.7381266666666666f,
+0.0879151568627451f, 0.335050385620915f, 0.7376140392156862f,
+0.08781264705882354f, 0.3337175490196078f, 0.7371014117647059f,
+0.08771013725490197f, 0.33238471241830064f, 0.7365887843137254f,
+0.0876076274509804f, 0.33105187581699347f, 0.7360761568627451f,
+0.08750511764705883f, 0.32971903921568624f, 0.7355635294117647f,
+0.08740260784313726f, 0.32838620261437906f, 0.7350509019607843f,
+0.08730009803921569f, 0.3270533660130719f, 0.7345382745098039f,
+0.08719758823529412f, 0.32572052941176466f, 0.7340256470588234f,
+0.08709507843137256f, 0.3243876928104575f, 0.7335130196078431f,
+0.08699256862745099f, 0.3230548562091503f, 0.7330003921568627f,
+0.08689005882352942f, 0.32172201960784313f, 0.7324877647058823f,
+0.08678754901960785f, 0.32038918300653596f, 0.7319751372549019f,
+0.08668503921568628f, 0.31905634640522873f, 0.7314625098039216f,
+0.08658252941176471f, 0.31772350980392156f, 0.7309498823529411f,
+0.08648001960784314f, 0.3163906732026144f, 0.7304372549019608f,
+0.08637750980392157f, 0.31505783660130715f, 0.7299246274509803f,
+0.086275f, 0.313725f, 0.729412f,
+0.08565978431372549f, 0.30941903921568625f, 0.7278741176470588f,
+0.08504456862745098f, 0.30511307843137253f, 0.7263362352941176f,
+0.08442935294117647f, 0.3008071176470588f, 0.7247983529411764f,
+0.08381413725490196f, 0.2965011568627451f, 0.7232604705882353f,
+0.08319892156862746f, 0.29219519607843136f, 0.7217225882352941f,
+0.08258370588235295f, 0.28788923529411764f, 0.720184705882353f,
+0.08196849019607844f, 0.2835832745098039f, 0.7186468235294118f,
+0.08135327450980392f, 0.2792773137254902f, 0.7171089411764706f,
+0.08073805882352941f, 0.2749713529411765f, 0.7155710588235293f,
+0.0801228431372549f, 0.27066539215686275f, 0.7140331764705882f,
+0.07950762745098042f, 0.26635943137254914f, 0.7124952941176471f,
+0.07889241176470588f, 0.2620534705882353f, 0.7109574117647058f,
+0.07850790196078432f, 0.25790127450980393f, 0.7091888431372549f,
+0.07881550980392157f, 0.2542103725490196f, 0.7067282156862745f,
+0.07912311764705883f, 0.2505194705882353f, 0.7042675882352941f,
+0.07943072549019607f, 0.24682856862745098f, 0.7018069607843137f,
+0.07973833333333333f, 0.24313766666666667f, 0.6993463333333333f,
+0.08004594117647058f, 0.23944676470588236f, 0.696885705882353f,
+0.08035354901960784f, 0.23575586274509805f, 0.6944250784313726f,
+0.0806611568627451f, 0.23206496078431374f, 0.6919644509803922f,
+0.08096876470588235f, 0.2283740588235294f, 0.6895038235294118f,
+0.08127637254901961f, 0.2246831568627451f, 0.6870431960784313f,
+0.08158398039215686f, 0.22099225490196078f, 0.6845825686274509f,
+0.08189158823529412f, 0.21730135294117647f, 0.6821219411764706f,
+0.08219919607843137f, 0.21361045098039216f, 0.6796613137254902f,
+0.08266056862745097f, 0.20991954901960785f, 0.6768931568627451f,
+0.08327570588235292f, 0.20622864705882366f, 0.6738174705882354f,
+0.0838908431372549f, 0.20253774509803923f, 0.6707417843137254f,
+0.08450598039215686f, 0.1988468431372549f, 0.6676660980392157f,
+0.08512111764705882f, 0.1951559411764706f, 0.6645904117647059f,
+0.08573625490196078f, 0.19146503921568628f, 0.6615147254901961f,
+0.08635139215686274f, 0.18777413725490197f, 0.6584390392156863f,
+0.0869665294117647f, 0.18408323529411763f, 0.6553633529411764f,
+0.08758166666666667f, 0.18039233333333332f, 0.6522876666666667f,
+0.08819680392156863f, 0.176701431372549f, 0.6492119803921569f,
+0.08881194117647058f, 0.1730105294117647f, 0.6461362941176471f,
+0.08942707843137254f, 0.1693196274509804f, 0.6430606078431372f,
+0.0900422156862745f, 0.16562872549019608f, 0.6399849215686274f,
+0.0908880588235294f, 0.1628605294117647f, 0.6369091764705882f,
+0.09181080392156862f, 0.1603999019607843f, 0.6338334117647059f,
+0.09273354901960784f, 0.1579392745098039f, 0.6307576470588235f,
+0.09365629411764706f, 0.15547864705882353f, 0.6276818823529412f,
+0.09457903921568625f, 0.1530180196078432f, 0.6246061176470589f,
+0.09550178431372548f, 0.15055739215686276f, 0.6215303529411764f,
+0.0964245294117647f, 0.14809676470588234f, 0.6184545882352941f,
+0.09734727450980392f, 0.14563613725490196f, 0.6153788235294118f,
+0.09827001960784314f, 0.14317550980392157f, 0.6123030588235294f,
+0.09919276470588234f, 0.1407148823529412f, 0.6092272941176471f,
+0.10011550980392156f, 0.13825425490196078f, 0.6061515294117646f,
+0.10103825490196078f, 0.1357936274509804f, 0.6030757647058823f,
+0.101961f, 0.133333f, 0.6f,
+0.10257613725490196f, 0.13148758823529413f, 0.5960015686274509f,
+0.10319127450980392f, 0.12964217647058823f, 0.592003137254902f,
+0.10380641176470588f, 0.12779676470588236f, 0.5880047058823529f,
+0.10442154901960785f, 0.1259513529411765f, 0.5840062745098039f,
+0.1050366862745098f, 0.12410594117647059f, 0.5800078431372548f,
+0.10565182352941176f, 0.12226052941176471f, 0.5760094117647059f,
+0.10626696078431372f, 0.12041511764705883f, 0.5720109803921568f,
+0.10688209803921567f, 0.118569705882353f, 0.568012549019608f,
+0.10749723529411764f, 0.11672429411764706f, 0.5640141176470588f,
+0.1081123725490196f, 0.11487888235294118f, 0.5600156862745098f,
+0.10872750980392157f, 0.11303347058823529f, 0.5560172549019607f,
+0.10934264705882353f, 0.11118805882352942f, 0.5520188235294117f,
+0.11064982352941176f, 0.10988088235294118f, 0.5480972745098038f,
+0.11403311764705883f, 0.11018841176470588f, 0.5444063725490196f,
+0.11741641176470588f, 0.11049594117647059f, 0.5407154705882352f,
+0.12079970588235293f, 0.1108034705882353f, 0.537024568627451f,
+0.124183f, 0.111111f, 0.5333336666666666f,
+0.12756629411764706f, 0.11141852941176471f, 0.5296427647058823f,
+0.13094958823529412f, 0.11172605882352941f, 0.525951862745098f,
+0.13433288235294116f, 0.11203358823529412f, 0.5222609607843137f,
+0.13771617647058823f, 0.11234111764705883f, 0.5185700588235294f,
+0.1410994705882353f, 0.11264864705882353f, 0.5148791568627451f,
+0.14448276470588234f, 0.11295617647058824f, 0.5111882549019607f,
+0.14786605882352932f, 0.11326370588235293f, 0.5074973529411766f,
+0.15124935294117647f, 0.11357123529411765f, 0.5038064509803921f,
+0.1534023725490196f, 0.11357123529411765f, 0.4987314705882353f,
+0.1543251176470588f, 0.11326370588235295f, 0.4922724117647059f,
+0.15524786274509803f, 0.11295617647058824f, 0.48581335294117645f,
+0.15617060784313724f, 0.11264864705882353f, 0.4793542941176471f,
+0.15709335294117646f, 0.11234111764705883f, 0.47289523529411764f,
+0.15801609803921568f, 0.11203358823529412f, 0.4664361764705882f,
+0.1589388431372549f, 0.11172605882352941f, 0.45997711764705884f,
+0.15986158823529412f, 0.11141852941176471f, 0.4535180588235294f,
+0.16078433333333333f, 0.111111f, 0.447059f,
+0.16170707843137255f, 0.1108034705882353f, 0.44059994117647056f,
+0.16262982352941174f, 0.11049594117647059f, 0.4341408823529412f,
+0.16355256862745096f, 0.11018841176470588f, 0.42768182352941175f,
+0.16447531372549018f, 0.10988088235294118f, 0.4212227647058823f,
+0.16262988235294118f, 0.10680517647058824f, 0.4126875882352941f,
+0.15986172549019614f, 0.10280674509803933f, 0.40346037254901984f,
+0.15709356862745097f, 0.09880831372549019f, 0.39423315686274507f,
+0.15432541176470588f, 0.09480988235294117f, 0.3850059411764706f,
+0.1515572549019608f, 0.09081145098039216f, 0.3757787254901961f,
+0.1487890980392157f, 0.08681301960784313f, 0.36655150980392154f,
+0.14602094117647058f, 0.08281458823529411f, 0.35732429411764705f,
+0.14325278431372548f, 0.0788161568627451f, 0.34809707843137255f,
+0.1404846274509804f, 0.07481772549019608f, 0.338869862745098f,
+0.1377164705882353f, 0.07081929411764706f, 0.3296426470588235f,
+0.13494831372549018f, 0.06682086274509805f, 0.320415431372549f,
+0.1321801568627451f, 0.06282243137254903f, 0.3111882156862745f,
+0.129412f, 0.058824f, 0.301961f
+}
+, 
+{
+0.94902f, 1.0f, 0.980392f,
+0.9471745490196078f, 1.0f, 0.9791617254901961f,
+0.9453290980392156f, 1.0f, 0.9779314509803922f,
+0.9434836470588235f, 1.0f, 0.9767011764705883f,
+0.9416381960784314f, 1.0f, 0.9754709019607843f,
+0.9397927450980392f, 1.0f, 0.9742406274509804f,
+0.937947294117647f, 1.0f, 0.9730103529411765f,
+0.9361018431372549f, 1.0f, 0.9717800784313726f,
+0.9342563921568627f, 1.0f, 0.9705498039215686f,
+0.9324109411764706f, 1.0f, 0.9693195294117647f,
+0.9305654901960784f, 1.0f, 0.9680892549019608f,
+0.9287200392156862f, 1.0f, 0.9668589803921569f,
+0.9268745882352941f, 1.0f, 0.965628705882353f,
+0.925029137254902f, 1.0f, 0.964398431372549f,
+0.9231836862745098f, 1.0f, 0.9631681568627452f,
+0.9213382352941176f, 1.0f, 0.9619378823529412f,
+0.9194927843137255f, 1.0f, 0.9607076078431372f,
+0.9176473333333334f, 1.0f, 0.9594773333333333f,
+0.9158018823529411f, 1.0f, 0.9582470588235295f,
+0.913956431372549f, 1.0f, 0.9570167843137255f,
+0.9121109803921569f, 1.0f, 0.9557865098039215f,
+0.9102655294117648f, 1.0f, 0.9545562352941176f,
+0.9084200784313725f, 1.0f, 0.9533259607843138f,
+0.9065746274509804f, 1.0f, 0.9520956862745098f,
+0.9047291764705883f, 1.0f, 0.9508654117647058f,
+0.9028837254901961f, 1.0f, 0.9496351372549019f,
+0.8991928431372549f, 0.9992310588235294f, 0.9477896862745098f,
+0.8936565294117648f, 0.9976931764705882f, 0.9453290588235294f,
+0.8881202156862745f, 0.996155294117647f, 0.9428684313725489f,
+0.8825839019607843f, 0.9946174117647059f, 0.9404078039215686f,
+0.8770475882352942f, 0.9930795294117647f, 0.9379471764705882f,
+0.8715112745098039f, 0.9915416470588235f, 0.9354865490196078f,
+0.8659749607843138f, 0.9900037647058824f, 0.9330259215686274f,
+0.8604386470588236f, 0.9884658823529412f, 0.9305652941176471f,
+0.8549023333333333f, 0.986928f, 0.9281046666666667f,
+0.8493660196078432f, 0.9853901176470589f, 0.9256440392156863f,
+0.843829705882353f, 0.9838522352941177f, 0.9231834117647059f,
+0.8382933921568628f, 0.9823143529411765f, 0.9207227843137255f,
+0.8327570784313726f, 0.9807764705882354f, 0.918262156862745f,
+0.8276820588235294f, 0.9792385882352942f, 0.9162629411764706f,
+0.8227608039215687f, 0.977700705882353f, 0.9144175294117647f,
+0.8178395490196079f, 0.9761628235294117f, 0.9125721176470588f,
+0.812918294117647f, 0.9746249411764706f, 0.9107267058823529f,
+0.8079970392156863f, 0.9730870588235294f, 0.9088812941176471f,
+0.8030757843137255f, 0.9715491764705882f, 0.9070358823529412f,
+0.7981545294117647f, 0.9700112941176471f, 0.9051904705882353f,
+0.7932332745098039f, 0.9684734117647059f, 0.9033450588235294f,
+0.7883120196078431f, 0.9669355294117647f, 0.9014996470588235f,
+0.7833907647058823f, 0.9653976470588235f, 0.8996542352941176f,
+0.7784695098039216f, 0.9638597647058823f, 0.8978088235294117f,
+0.7735482549019608f, 0.9623218823529411f, 0.8959634117647058f,
+0.768627f, 0.960784f, 0.894118f,
+0.7630906862745097f, 0.9583234509803922f, 0.8916573725490196f,
+0.7575543725490196f, 0.9558629019607843f, 0.8891967450980393f,
+0.7520180588235293f, 0.9534023529411765f, 0.8867361176470588f,
+0.7464817450980392f, 0.9509418039215686f, 0.8842754901960784f,
+0.740945431372549f, 0.9484812549019608f, 0.881814862745098f,
+0.7354091176470589f, 0.946020705882353f, 0.8793542352941176f,
+0.7298728039215686f, 0.9435601568627451f, 0.8768936078431372f,
+0.7243364901960784f, 0.9410996078431373f, 0.8744329803921569f,
+0.7188001764705882f, 0.9386390588235294f, 0.8719723529411765f,
+0.713263862745098f, 0.9361785098039216f, 0.8695117254901961f,
+0.7077275490196078f, 0.9337179607843137f, 0.8670510980392157f,
+0.7021912352941176f, 0.9312574117647059f, 0.8645904705882352f,
+0.6968856078431372f, 0.9288737450980392f, 0.8623605294117647f,
+0.6922720392156863f, 0.9267207254901961f, 0.8608226470588235f,
+0.6876584705882354f, 0.924567705882353f, 0.8592847647058823f,
+0.6830449019607843f, 0.9224146862745098f, 0.8577468823529412f,
+0.6784313333333333f, 0.9202616666666666f, 0.856209f,
+0.6738177647058823f, 0.9181086470588236f, 0.8546711176470588f,
+0.6692041960784314f, 0.9159556274509804f, 0.8531332352941177f,
+0.6645906274509804f, 0.9138026078431373f, 0.8515953529411765f,
+0.6599770588235294f, 0.9116495882352942f, 0.8500574705882353f,
+0.6553634901960784f, 0.909496568627451f, 0.8485195882352942f,
+0.6507499215686275f, 0.9073435490196079f, 0.846981705882353f,
+0.6461363529411764f, 0.9051905294117647f, 0.8454438235294118f,
+0.6415227843137254f, 0.9030375098039216f, 0.8439059411764707f,
+0.6372167450980393f, 0.9007306862745098f, 0.8425218627450981f,
+0.6332182352941177f, 0.8982700588235294f, 0.8412915882352942f,
+0.6292197254901961f, 0.895809431372549f, 0.8400613137254902f,
+0.6252212156862745f, 0.8933488039215687f, 0.8388310392156864f,
+0.621222705882353f, 0.8908881764705883f, 0.8376007647058824f,
+0.6172241960784314f, 0.8884275490196079f, 0.8363704901960785f,
+0.6132256862745098f, 0.8859669215686274f, 0.8351402156862745f,
+0.6092271764705882f, 0.883506294117647f, 0.8339099411764707f,
+0.6052286666666666f, 0.8810456666666667f, 0.8326796666666667f,
+0.6012301568627451f, 0.8785850392156863f, 0.8314493921568628f,
+0.5972316470588235f, 0.8761244117647059f, 0.8302191176470589f,
+0.5932331372549019f, 0.8736637843137255f, 0.828988843137255f,
+0.5892346274509803f, 0.8712031568627451f, 0.827758568627451f,
+0.5854668235294118f, 0.8682812352941177f, 0.8260669411764706f,
+0.5817759215686275f, 0.8652055490196079f, 0.8242215294117647f,
+0.5780850196078431f, 0.8621298627450981f, 0.8223761176470589f,
+0.5743941176470588f, 0.8590541764705882f, 0.820530705882353f,
+0.5707032156862745f, 0.8559784901960784f, 0.8186852941176471f,
+0.5670123137254902f, 0.8529028039215687f, 0.8168398823529412f,
+0.5633214117647058f, 0.8498271176470589f, 0.8149944705882354f,
+0.5596305098039216f, 0.846751431372549f, 0.8131490588235295f,
+0.5559396078431373f, 0.8436757450980393f, 0.8113036470588236f,
+0.552248705882353f, 0.8406000588235294f, 0.8094582352941176f,
+0.5485578039215686f, 0.8375243725490197f, 0.8076128235294118f,
+0.5448669019607844f, 0.8344486862745099f, 0.8057674117647059f,
+0.541176f, 0.831373f, 0.803922f,
+0.5381003137254902f, 0.8289123725490196f, 0.8033067843137255f,
+0.5350246274509803f, 0.8264517450980392f, 0.802691568627451f,
+0.5319489411764706f, 0.8239911176470589f, 0.8020763529411765f,
+0.5288732549019608f, 0.8215304901960785f, 0.801461137254902f,
+0.525797568627451f, 0.8190698627450981f, 0.8008459215686274f,
+0.5227218823529411f, 0.8166092352941177f, 0.8002307058823529f,
+0.5196461960784313f, 0.8141486078431373f, 0.7996154901960785f,
+0.5165705098039216f, 0.811687980392157f, 0.7990002745098039f,
+0.5134948235294118f, 0.8092273529411765f, 0.7983850588235294f,
+0.510419137254902f, 0.8067667254901961f, 0.7977698431372549f,
+0.5073434509803921f, 0.8043060980392157f, 0.7971546274509803f,
+0.5042677647058824f, 0.8018454705882354f, 0.7965394117647059f,
+0.5010382745098039f, 0.7989234901960784f, 0.7953859607843137f,
+0.4973473725490196f, 0.7946174509803922f, 0.7926178039215686f,
+0.4936564705882353f, 0.7903114117647059f, 0.7898496470588235f,
+0.48996556862745094f, 0.7860053725490197f, 0.7870814901960784f,
+0.48627466666666663f, 0.7816993333333334f, 0.7843133333333333f,
+0.4825837647058823f, 0.7773932941176471f, 0.7815451764705882f,
+0.478892862745098f, 0.7730872549019608f, 0.7787770196078431f,
+0.47520196078431376f, 0.7687812156862747f, 0.7760088627450981f,
+0.4715110588235294f, 0.7644751764705883f, 0.7732407058823529f,
+0.4678201568627451f, 0.760169137254902f, 0.7704725490196078f,
+0.4641292549019608f, 0.7558630980392157f, 0.7677043921568627f,
+0.46043835294117647f, 0.7515570588235294f, 0.7649362352941177f,
+0.45674745098039216f, 0.7472510196078431f, 0.7621680784313726f,
+0.4535179215686274f, 0.7430987843137256f, 0.7595537254901961f,
+0.45074976470588235f, 0.7391003529411765f, 0.7570931764705883f,
+0.44798160784313723f, 0.7351019215686275f, 0.7546326274509804f,
+0.4452134509803921f, 0.7311034901960785f, 0.7521720784313726f,
+0.4424452941176471f, 0.7271050588235296f, 0.7497115294117648f,
+0.43967713725490193f, 0.7231066274509804f, 0.7472509803921569f,
+0.43690898039215686f, 0.7191081960784313f, 0.744790431372549f,
+0.43414082352941175f, 0.7151097647058824f, 0.7423298823529412f,
+0.4313726666666666f, 0.7111113333333333f, 0.7398693333333333f,
+0.42860450980392156f, 0.7071129019607844f, 0.7374087843137255f,
+0.42583635294117644f, 0.7031144705882353f, 0.7349482352941176f,
+0.4230681960784314f, 0.6991160392156863f, 0.7324876862745098f,
+0.42030003921568626f, 0.6951176078431373f, 0.7300271372549019f,
+0.41730117647058823f, 0.6904271176470589f, 0.7271051764705881f,
+0.41422541176470584f, 0.6855059411764706f, 0.7240294117647058f,
+0.4111496470588235f, 0.6805847647058824f, 0.7209536470588235f,
+0.4080738823529412f, 0.6756635882352942f, 0.7178778823529411f,
+0.40499811764705884f, 0.6707424117647058f, 0.7148021176470588f,
+0.40192235294117645f, 0.6658212352941176f, 0.7117263529411765f,
+0.3988465882352941f, 0.6609000588235294f, 0.7086505882352941f,
+0.39577082352941184f, 0.6559788823529413f, 0.7055748235294118f,
+0.3926950588235294f, 0.6510577058823529f, 0.7024990588235294f,
+0.38961929411764706f, 0.6461365294117647f, 0.699423294117647f,
+0.38654352941176473f, 0.6412153529411765f, 0.6963475294117647f,
+0.38346776470588234f, 0.6362941764705882f, 0.6932717647058824f,
+0.380392f, 0.631373f, 0.690196f,
+0.3776238431372549f, 0.6273744901960784f, 0.6877354509803922f,
+0.3748556862745098f, 0.6233759803921568f, 0.6852749019607843f,
+0.3720875294117647f, 0.6193774705882352f, 0.6828143529411765f,
+0.36931937254901964f, 0.6153789607843138f, 0.6803538039215686f,
+0.3665512156862745f, 0.6113804509803922f, 0.6778932549019608f,
+0.3637830588235294f, 0.6073819411764706f, 0.6754327058823529f,
+0.36101490196078434f, 0.603383431372549f, 0.6729721568627451f,
+0.3582467450980392f, 0.5993849215686274f, 0.6705116078431372f,
+0.3554785882352941f, 0.5953864117647059f, 0.6680510588235294f,
+0.35271043137254904f, 0.5913879019607843f, 0.6655905098039215f,
+0.34994227450980403f, 0.5873893921568629f, 0.6631299607843137f,
+0.34717411764705886f, 0.5833908823529412f, 0.6606694117647058f,
+0.3445597450980392f, 0.5793154901960784f, 0.6580550588235293f,
+0.3424067254901961f, 0.5750094509803921f, 0.654979294117647f,
+0.34025370588235293f, 0.5707034117647058f, 0.6519035294117647f,
+0.33810068627450984f, 0.5663973725490197f, 0.6488277647058823f,
+0.3359476666666667f, 0.5620913333333334f, 0.645752f,
+0.33379464705882356f, 0.5577852941176471f, 0.6426762352941177f,
+0.3316416274509804f, 0.5534792549019608f, 0.6396004705882353f,
+0.32948860784313727f, 0.5491732156862745f, 0.636524705882353f,
+0.32733558823529413f, 0.5448671764705882f, 0.6334489411764707f,
+0.325182568627451f, 0.540561137254902f, 0.6303731764705882f,
+0.32302954901960784f, 0.5362550980392157f, 0.6272974117647059f,
+0.3208765294117647f, 0.5319490588235294f, 0.6242216470588235f,
+0.31872350980392156f, 0.5276430196078431f, 0.6211458823529412f,
+0.3171856274509804f, 0.5233369803921569f, 0.6180701176470589f,
+0.3162628823529412f, 0.5190309411764708f, 0.6149943529411765f,
+0.31534013725490195f, 0.5147249019607844f, 0.6119185882352942f,
+0.31441739215686276f, 0.5104188627450981f, 0.6088428235294118f,
+0.3134946470588235f, 0.5061128235294118f, 0.6057670588235294f,
+0.3125719019607843f, 0.5018067843137255f, 0.6026912941176471f,
+0.3116491568627451f, 0.4975007450980392f, 0.5996155294117648f,
+0.3107264117647059f, 0.4931947058823529f, 0.5965397647058824f,
+0.30980366666666664f, 0.4888886666666667f, 0.593464f,
+0.30888092156862745f, 0.4845826274509804f, 0.5903882352941177f,
+0.3079581764705882f, 0.4802765882352941f, 0.5873124705882353f,
+0.307035431372549f, 0.47597054901960784f, 0.584236705882353f,
+0.30611268627450977f, 0.4716645098039216f, 0.5811609411764707f,
+0.30495929411764705f, 0.4673584705882353f, 0.5773931764705883f,
+0.30372901960784315f, 0.46305243137254903f, 0.5733947450980392f,
+0.3024987450980392f, 0.45874639215686275f, 0.5693963137254902f,
+0.3012684705882353f, 0.45444035294117646f, 0.5653978823529412f,
+0.3000381960784314f, 0.45013431372549034f, 0.5613994509803922f,
+0.29880792156862745f, 0.44582827450980395f, 0.5574010196078432f,
+0.2975776470588235f, 0.44152223529411766f, 0.5534025882352941f,
+0.2963473725490196f, 0.4372161960784314f, 0.5494041568627451f,
+0.2951170980392157f, 0.4329101568627451f, 0.5454057254901961f,
+0.29388682352941176f, 0.42860411764705886f, 0.5414072941176471f,
+0.29265654901960786f, 0.42429807843137257f, 0.537408862745098f,
+0.2914262745098039f, 0.4199920392156863f, 0.5334104313725491f,
+0.290196f, 0.415686f, 0.529412f,
+0.2892732549019608f, 0.41076482352941174f, 0.5247983529411765f,
+0.2883505098039216f, 0.40584364705882353f, 0.5201847058823529f,
+0.2874277647058824f, 0.40092247058823527f, 0.5155710588235294f,
+0.28650501960784314f, 0.39600129411764706f, 0.5109574117647059f,
+0.28558227450980395f, 0.3910801176470588f, 0.5063437647058824f,
+0.2846595294117647f, 0.3861589411764706f, 0.5017301176470588f,
+0.2837367843137255f, 0.38123776470588233f, 0.49711647058823527f,
+0.2828140392156863f, 0.37631658823529424f, 0.4925028235294119f,
+0.2818912941176471f, 0.37139541176470586f, 0.4878891764705882f,
+0.2809685490196078f, 0.36647423529411766f, 0.4832755294117647f,
+0.28004580392156864f, 0.3615530588235294f, 0.47866188235294116f,
+0.2791230588235294f, 0.3566318823529412f, 0.47404823529411766f,
+0.27804654901960785f, 0.35186449019607846f, 0.46920392156862745f,
+0.2765087450980392f, 0.3475584509803922f, 0.46366760784313726f,
+0.2749709411764706f, 0.3432524117647059f, 0.4581312941176471f,
+0.2734331372549019f, 0.3389463725490196f, 0.4525949803921569f,
+0.2718953333333333f, 0.3346403333333333f, 0.44705866666666666f,
+0.2703575294117647f, 0.3303342941176471f, 0.4415223529411765f,
+0.26881972549019606f, 0.3260282549019608f, 0.4359860392156863f,
+0.26728192156862746f, 0.3217222156862745f, 0.4304497254901961f,
+0.2657441176470588f, 0.31741617647058823f, 0.4249134117647059f,
+0.2642063137254902f, 0.31311013725490194f, 0.4193770980392157f,
+0.26266850980392153f, 0.3088040980392157f, 0.4138407843137255f,
+0.261130705882353f, 0.30449805882352954f, 0.4083044705882355f,
+0.25959290196078433f, 0.30019201960784314f, 0.40276815686274514f,
+0.2579012549019608f, 0.29603978431372546f, 0.3972318431372549f,
+0.25605576470588237f, 0.29204135294117645f, 0.3916955294117647f,
+0.25421027450980394f, 0.28804292156862743f, 0.38615921568627454f,
+0.2523647843137255f, 0.2840444901960784f, 0.3806229019607843f,
+0.25051929411764706f, 0.2800460588235294f, 0.3750865882352941f,
+0.24867380392156863f, 0.2760476274509804f, 0.36955027450980393f,
+0.2468283137254902f, 0.27204919607843137f, 0.36401396078431375f,
+0.24498282352941175f, 0.26805076470588235f, 0.3584776470588235f,
+0.24313733333333334f, 0.26405233333333333f, 0.35294133333333333f,
+0.2412918431372549f, 0.2600539019607843f, 0.34740501960784315f,
+0.23944635294117647f, 0.2560554705882353f, 0.34186870588235296f,
+0.23760086274509803f, 0.2520570392156863f, 0.3363323921568627f,
+0.23575537254901963f, 0.24805860784313727f, 0.33079607843137254f,
+0.2320644705882353f, 0.24290676470588235f, 0.32364499999999996f,
+0.22775843137254914f, 0.2373704509803923f, 0.31595566666666686f,
+0.22345239215686274f, 0.23183413725490196f, 0.3082663333333333f,
+0.21914635294117646f, 0.22629782352941177f, 0.300577f,
+0.2148403137254902f, 0.22076150980392156f, 0.29288766666666666f,
+0.21053427450980391f, 0.21522519607843138f, 0.28519833333333333f,
+0.20622823529411766f, 0.20968888235294117f, 0.277509f,
+0.20192219607843137f, 0.204152568627451f, 0.2698196666666667f,
+0.19761615686274508f, 0.19861625490196078f, 0.26213033333333335f,
+0.19331011764705883f, 0.1930799411764706f, 0.25444100000000003f,
+0.18900407843137254f, 0.18754362745098038f, 0.24675166666666665f,
+0.18469803921568628f, 0.18200731372549017f, 0.23906233333333332f,
+0.180392f, 0.176471f, 0.231373f
+}
+, 
+{
+0.960784f, 0.988235f, 1.0f,
+0.9530946666666666f, 0.9845440980392156f, 0.9984621176470588f,
+0.9454053333333333f, 0.9808531960784314f, 0.9969242352941177f,
+0.937716f, 0.977162294117647f, 0.9953863529411765f,
+0.9300266666666667f, 0.9734713921568627f, 0.9938484705882353f,
+0.9223373333333333f, 0.9697804901960784f, 0.9923105882352942f,
+0.914648f, 0.9660895882352941f, 0.990772705882353f,
+0.9069586666666667f, 0.9623986862745098f, 0.9892348235294118f,
+0.8992693333333334f, 0.9587077843137255f, 0.9876969411764707f,
+0.8915799999999999f, 0.9550168823529411f, 0.9861590588235294f,
+0.8838906666666666f, 0.9513259803921569f, 0.9846211764705882f,
+0.8762013333333333f, 0.9476350784313725f, 0.983083294117647f,
+0.868512f, 0.9439441764705883f, 0.9815454117647059f,
+0.8611302352941176f, 0.9400226078431373f, 0.9797768627450981f,
+0.8546711764705882f, 0.9354090392156863f, 0.9773163137254902f,
+0.8482121176470588f, 0.9307954705882353f, 0.9748557647058824f,
+0.8417530588235294f, 0.9261819019607843f, 0.9723952156862745f,
+0.835294f, 0.9215683333333333f, 0.9699346666666667f,
+0.8288349411764706f, 0.9169547647058823f, 0.9674741176470588f,
+0.8223758823529411f, 0.9123411960784314f, 0.965013568627451f,
+0.8159168235294117f, 0.9077276274509805f, 0.9625530196078431f,
+0.8094577647058824f, 0.9031140588235295f, 0.9600924705882353f,
+0.8029987058823529f, 0.8985004901960785f, 0.9576319215686274f,
+0.7965396470588235f, 0.8938869215686275f, 0.9551713725490196f,
+0.7900805882352941f, 0.8892733529411765f, 0.9527108235294117f,
+0.7836215294117647f, 0.8846597843137255f, 0.9502502745098039f,
+0.7790848039215686f, 0.8808920196078431f, 0.9478665882352941f,
+0.7764704117647059f, 0.8779700588235294f, 0.9455597647058823f,
+0.7738560196078431f, 0.8750480980392157f, 0.9432529411764705f,
+0.7712416274509803f, 0.872126137254902f, 0.9409461176470588f,
+0.7686272352941176f, 0.8692041764705882f, 0.938639294117647f,
+0.7660128431372549f, 0.8662822156862745f, 0.9363324705882353f,
+0.7633984509803922f, 0.8633602549019608f, 0.9340256470588235f,
+0.7607840588235294f, 0.8604382941176472f, 0.9317188235294117f,
+0.7581696666666666f, 0.8575163333333333f, 0.929412f,
+0.755555274509804f, 0.8545943725490196f, 0.9271051764705882f,
+0.7529408823529412f, 0.8516724117647059f, 0.9247983529411764f,
+0.7503264901960784f, 0.8487504509803921f, 0.9224915294117647f,
+0.7477120980392157f, 0.8458284901960784f, 0.9201847058823529f,
+0.7450977058823529f, 0.8429065294117647f, 0.9178778823529412f,
+0.7424833137254903f, 0.839984568627451f, 0.9155710588235294f,
+0.7398689215686275f, 0.8370626078431372f, 0.9132642352941176f,
+0.7372545294117647f, 0.8341406470588235f, 0.9109574117647059f,
+0.734640137254902f, 0.8312186862745098f, 0.9086505882352941f,
+0.7320257450980392f, 0.8282967254901961f, 0.9063437647058823f,
+0.7294113529411765f, 0.8253747647058823f, 0.9040369411764706f,
+0.7267969607843138f, 0.8224528039215686f, 0.9017301176470588f,
+0.724182568627451f, 0.8195308431372549f, 0.899423294117647f,
+0.7215681764705882f, 0.8166088823529412f, 0.8971164705882353f,
+0.7189537843137256f, 0.8136869215686274f, 0.8948096470588235f,
+0.7163393921568628f, 0.8107649607843137f, 0.8925028235294118f,
+0.713725f, 0.807843f, 0.890196f,
+0.7101879019607844f, 0.8044597058823529f, 0.8874278431372549f,
+0.7066508039215686f, 0.8010764117647059f, 0.8846596862745099f,
+0.703113705882353f, 0.7976931176470589f, 0.8818915294117647f,
+0.6995766078431374f, 0.7943098235294117f, 0.8791233725490196f,
+0.6960395098039216f, 0.7909265294117647f, 0.8763552156862745f,
+0.692502411764706f, 0.7875432352941176f, 0.8735870588235294f,
+0.6889653137254902f, 0.7841599411764706f, 0.8708189019607843f,
+0.6854282156862745f, 0.7807766470588235f, 0.8680507450980393f,
+0.6818911176470589f, 0.7773933529411765f, 0.8652825882352941f,
+0.6783540196078433f, 0.7740100588235295f, 0.862514431372549f,
+0.6748169215686275f, 0.7706267647058823f, 0.8597462745098039f,
+0.6712798235294118f, 0.7672434705882353f, 0.8569781176470588f,
+0.6677427254901961f, 0.7638601764705882f, 0.8542099607843138f,
+0.6642056274509804f, 0.7604768823529412f, 0.8514418039215687f,
+0.6606685294117648f, 0.7570935882352942f, 0.8486736470588235f,
+0.657131431372549f, 0.753710294117647f, 0.8459054901960784f,
+0.6535943333333334f, 0.750327f, 0.8431373333333333f,
+0.6500572352941176f, 0.7469437058823529f, 0.8403691764705883f,
+0.646520137254902f, 0.7435604117647059f, 0.8376010196078432f,
+0.6429830392156863f, 0.7401771176470588f, 0.834832862745098f,
+0.6394459411764706f, 0.7367938235294118f, 0.8320647058823529f,
+0.6359088431372549f, 0.7334105294117648f, 0.8292965490196078f,
+0.6323717450980393f, 0.7300272352941177f, 0.8265283921568628f,
+0.6288346470588235f, 0.7266439411764706f, 0.8237602352941177f,
+0.6252975490196079f, 0.7232606470588235f, 0.8209920784313726f,
+0.6217604509803921f, 0.7198004509803921f, 0.8180701176470588f,
+0.6182233529411765f, 0.7162633529411765f, 0.8149943529411765f,
+0.6146862549019608f, 0.7127262549019608f, 0.8119185882352942f,
+0.6111491568627451f, 0.7091891568627451f, 0.8088428235294117f,
+0.6076120588235294f, 0.7056520588235294f, 0.8057670588235294f,
+0.6040749607843138f, 0.7021149607843138f, 0.8026912941176471f,
+0.600537862745098f, 0.698577862745098f, 0.7996155294117647f,
+0.5970007647058824f, 0.6950407647058824f, 0.7965397647058824f,
+0.5934636666666666f, 0.6915036666666666f, 0.793464f,
+0.589926568627451f, 0.687966568627451f, 0.7903882352941176f,
+0.5863894705882353f, 0.6844294705882353f, 0.7873124705882353f,
+0.5828523725490196f, 0.6808923725490196f, 0.784236705882353f,
+0.5793152745098039f, 0.6773552745098039f, 0.7811609411764706f,
+0.5757781764705883f, 0.6738181764705883f, 0.7780851764705883f,
+0.5722410784313725f, 0.6702810784313725f, 0.7750094117647058f,
+0.5687039803921569f, 0.6667439803921569f, 0.7719336470588235f,
+0.5651668823529411f, 0.6632068823529411f, 0.7688578823529412f,
+0.5616297843137255f, 0.6596697843137255f, 0.7657821176470588f,
+0.5580926862745098f, 0.6561326862745098f, 0.7627063529411764f,
+0.554555588235294f, 0.6525955882352941f, 0.7596305882352941f,
+0.5510184901960784f, 0.6490584901960784f, 0.7565548235294117f,
+0.5474813921568628f, 0.6455213921568628f, 0.7534790588235294f,
+0.543944294117647f, 0.641984294117647f, 0.7504032941176471f,
+0.5404071960784314f, 0.6384471960784314f, 0.7473275294117646f,
+0.5368700980392156f, 0.6349100980392156f, 0.7442517647058823f,
+0.533333f, 0.631373f, 0.741176f,
+0.5302572745098039f, 0.6281434705882353f, 0.7379465098039215f,
+0.5271815490196078f, 0.6249139411764706f, 0.7347170196078431f,
+0.5241058235294117f, 0.6216844117647058f, 0.7314875294117646f,
+0.5210300980392156f, 0.6184548823529412f, 0.7282580392156862f,
+0.5179543725490195f, 0.6152253529411764f, 0.7250285490196078f,
+0.5148786470588235f, 0.6119958235294117f, 0.7217990588235293f,
+0.5118029215686274f, 0.608766294117647f, 0.7185695686274509f,
+0.5087271960784313f, 0.6055367647058824f, 0.7153400784313725f,
+0.5056514705882352f, 0.6023072352941176f, 0.712110588235294f,
+0.5025757450980392f, 0.5990777058823529f, 0.7088810980392156f,
+0.49950001960784307f, 0.5958481764705882f, 0.7056516078431372f,
+0.49642429411764705f, 0.5926186470588235f, 0.7024221176470589f,
+0.4933485686274509f, 0.5893891176470588f, 0.6991926274509803f,
+0.49027284313725483f, 0.5861595882352941f, 0.6959631372549019f,
+0.48719711764705875f, 0.5829300588235293f, 0.6927336470588235f,
+0.48412139215686273f, 0.5797005294117646f, 0.689504156862745f,
+0.48104566666666665f, 0.576471f, 0.6862746666666666f,
+0.47796994117647057f, 0.5732414705882353f, 0.6830451764705882f,
+0.4748942156862745f, 0.5700119411764706f, 0.6798156862745097f,
+0.47181849019607847f, 0.5667824117647059f, 0.6765861960784314f,
+0.46874276470588233f, 0.5635528823529411f, 0.6733567058823529f,
+0.46566703921568625f, 0.5603233529411764f, 0.6701272156862744f,
+0.46259131372549017f, 0.5570938235294117f, 0.666897725490196f,
+0.4595155882352941f, 0.553864294117647f, 0.6636682352941177f,
+0.456439862745098f, 0.5506347647058824f, 0.6604387450980391f,
+0.4535179215686274f, 0.5474821176470588f, 0.6572861176470588f,
+0.45074976470588235f, 0.5444063529411765f, 0.6542103529411765f,
+0.44798160784313723f, 0.5413305882352941f, 0.6511345882352941f,
+0.4452134509803921f, 0.5382548235294117f, 0.6480588235294117f,
+0.4424452941176471f, 0.5351790588235295f, 0.6449830588235295f,
+0.43967713725490193f, 0.532103294117647f, 0.641907294117647f,
+0.43690898039215686f, 0.5290275294117647f, 0.6388315294117647f,
+0.43414082352941175f, 0.5259517647058823f, 0.6357557647058824f,
+0.4313726666666666f, 0.522876f, 0.63268f,
+0.42860450980392156f, 0.5198002352941176f, 0.6296042352941176f,
+0.42583635294117644f, 0.5167244705882352f, 0.6265284705882352f,
+0.4230681960784314f, 0.5136487058823529f, 0.6234527058823529f,
+0.42030003921568626f, 0.5105729411764706f, 0.6203769411764706f,
+0.41753188235294114f, 0.5074971764705882f, 0.6173011764705882f,
+0.4147637254901961f, 0.5044214117647059f, 0.6142254117647059f,
+0.41199556862745096f, 0.5013456470588236f, 0.6111496470588236f,
+0.40922741176470584f, 0.4982698823529412f, 0.6080738823529411f,
+0.4064592549019608f, 0.4951941176470588f, 0.6049981176470588f,
+0.40369109803921566f, 0.49211835294117645f, 0.6019223529411765f,
+0.4009229411764706f, 0.4890425882352941f, 0.5988465882352941f,
+0.39815478431372553f, 0.48596682352941184f, 0.5957708235294119f,
+0.39538662745098035f, 0.4828910588235294f, 0.5926950588235295f,
+0.3926184705882353f, 0.47981529411764706f, 0.589619294117647f,
+0.38985031372549017f, 0.47673952941176473f, 0.5865435294117647f,
+0.38708215686274505f, 0.47366376470588234f, 0.5834677647058824f,
+0.384314f, 0.470588f, 0.580392f,
+0.3815458039215686f, 0.4676660392156863f, 0.5773162745098039f,
+0.37877760784313724f, 0.46474407843137255f, 0.5742405490196079f,
+0.37600941176470587f, 0.4618221176470588f, 0.5711648235294118f,
+0.3732412156862745f, 0.4589001568627451f, 0.5680890980392157f,
+0.3704730196078431f, 0.4559781960784314f, 0.5650133725490196f,
+0.36770482352941175f, 0.45305623529411765f, 0.5619376470588235f,
+0.3649366274509804f, 0.4501342745098039f, 0.5588619215686275f,
+0.362168431372549f, 0.4472123137254902f, 0.5557861960784314f,
+0.35940023529411763f, 0.44429035294117647f, 0.5527104705882353f,
+0.35663203921568626f, 0.44136839215686274f, 0.5496347450980392f,
+0.35386384313725494f, 0.43844643137254913f, 0.5465590196078433f,
+0.3510956470588235f, 0.4355244705882353f, 0.5434832941176471f,
+0.34832745098039214f, 0.43260250980392156f, 0.540407568627451f,
+0.34555925490196077f, 0.42968054901960784f, 0.5373318431372549f,
+0.3427910588235294f, 0.4267585882352941f, 0.5342561176470588f,
+0.340022862745098f, 0.4238366274509804f, 0.5311803921568627f,
+0.33725466666666665f, 0.42091466666666666f, 0.5281046666666667f,
+0.3344864705882353f, 0.41799270588235293f, 0.5250289411764706f,
+0.3317182745098039f, 0.4150707450980392f, 0.5219532156862745f,
+0.32895007843137253f, 0.4121487843137255f, 0.5188774901960784f,
+0.32618188235294115f, 0.40922682352941175f, 0.5158017647058823f,
+0.3234136862745098f, 0.406304862745098f, 0.5127260392156863f,
+0.3206454901960784f, 0.4033829019607843f, 0.5096503137254902f,
+0.31787729411764704f, 0.4004609411764706f, 0.5065745882352941f,
+0.31510909803921566f, 0.39753898039215685f, 0.503498862745098f,
+0.3124947058823529f, 0.39461703921568625f, 0.5003462352941176f,
+0.31003411764705885f, 0.3916951176470589f, 0.497116705882353f,
+0.3075735294117647f, 0.38877319607843136f, 0.4938871764705882f,
+0.3051129411764706f, 0.3858512745098039f, 0.49065764705882353f,
+0.30265235294117643f, 0.38292935294117647f, 0.4874281176470588f,
+0.3001917647058823f, 0.380007431372549f, 0.4841985882352941f,
+0.2977311764705882f, 0.3770855098039216f, 0.4809690588235294f,
+0.2952705882352941f, 0.3741635882352941f, 0.4777395294117647f,
+0.29280999999999996f, 0.37124166666666664f, 0.47451f,
+0.29034941176470586f, 0.3683197450980392f, 0.4712804705882353f,
+0.28788882352941175f, 0.36539782352941175f, 0.46805094117647056f,
+0.28542823529411765f, 0.3624759019607843f, 0.46482141176470587f,
+0.2829676470588235f, 0.35955398039215686f, 0.4615918823529412f,
+0.2805070588235294f, 0.3566320588235294f, 0.45836235294117644f,
+0.2780464705882353f, 0.35371013725490197f, 0.45513282352941176f,
+0.2755858823529412f, 0.3507882156862745f, 0.45190329411764707f,
+0.273125294117647f, 0.347866294117647f, 0.4486737647058823f,
+0.270664705882353f, 0.34494437254901966f, 0.4454442352941177f,
+0.2682041176470588f, 0.34202245098039213f, 0.4422147058823529f,
+0.2657435294117647f, 0.3391005294117647f, 0.4389851764705882f,
+0.26328294117647055f, 0.33617860784313724f, 0.4357556470588235f,
+0.26082235294117645f, 0.3332566862745098f, 0.43252611764705884f,
+0.25836176470588235f, 0.33033476470588236f, 0.4292965882352941f,
+0.25590117647058824f, 0.3274128431372549f, 0.4260670588235294f,
+0.2534405882352941f, 0.32449092156862747f, 0.42283752941176467f,
+0.25098f, 0.321569f, 0.419608f,
+0.24898078431372547f, 0.3191084117647059f, 0.41683984313725486f,
+0.24698156862745096f, 0.31664782352941173f, 0.4140716862745098f,
+0.24498235294117646f, 0.31418723529411763f, 0.4113035294117647f,
+0.24298313725490195f, 0.3117266470588235f, 0.4085353725490196f,
+0.24098392156862744f, 0.3092660588235294f, 0.4057672156862745f,
+0.23898470588235293f, 0.30680547058823526f, 0.4029990588235294f,
+0.23698549019607842f, 0.30434488235294116f, 0.4002309019607843f,
+0.23498627450980397f, 0.3018842941176471f, 0.3974627450980393f,
+0.2329870588235294f, 0.29942370588235295f, 0.39469458823529413f,
+0.2309878431372549f, 0.2969631176470588f, 0.391926431372549f,
+0.2289886274509804f, 0.2945025294117647f, 0.3891582745098039f,
+0.22698941176470588f, 0.2920419411764706f, 0.38639011764705883f,
+0.22499019607843138f, 0.2895813529411765f, 0.3836219607843137f,
+0.22299098039215687f, 0.2871207647058823f, 0.3808538039215686f,
+0.22099176470588236f, 0.2846601764705882f, 0.3780856470588235f,
+0.21899254901960785f, 0.2821995882352941f, 0.3753174901960784f,
+0.21699333333333334f, 0.279739f, 0.37254933333333334f,
+0.21499411764705884f, 0.27727841176470586f, 0.3697811764705882f,
+0.2129949019607843f, 0.27481782352941175f, 0.3670130196078431f,
+0.21099568627450982f, 0.27235723529411765f, 0.36424486274509804f,
+0.2089964705882353f, 0.26989664705882355f, 0.3614767058823529f,
+0.2069972549019608f, 0.2674360588235294f, 0.35870854901960786f,
+0.20499803921568627f, 0.2649754705882353f, 0.35594039215686274f,
+0.20299882352941184f, 0.26251488235294124f, 0.35317223529411773f,
+0.20099960784313725f, 0.2600542941176471f, 0.35040407843137256f,
+0.19900039215686274f, 0.2575936862745098f, 0.3476359019607843f,
+0.19700117647058824f, 0.25513305882352943f, 0.34486770588235294f,
+0.19500196078431373f, 0.252672431372549f, 0.34209950980392156f,
+0.19300274509803922f, 0.2502118039215686f, 0.3393313137254902f,
+0.1910035294117647f, 0.24775117647058822f, 0.3365631176470588f,
+0.1890043137254902f, 0.24529054901960784f, 0.33379492156862745f,
+0.1870050980392157f, 0.24282992156862745f, 0.33102672549019607f,
+0.1850058823529412f, 0.24036929411764707f, 0.3282585294117647f,
+0.18300666666666668f, 0.23790866666666666f, 0.3254903333333333f,
+0.18100745098039217f, 0.23544803921568627f, 0.32272213725490195f,
+0.17900823529411766f, 0.2329874117647059f, 0.3199539411764706f,
+0.17700901960784315f, 0.23052678431372547f, 0.3171857450980392f,
+0.17500980392156865f, 0.2280661568627451f, 0.31441754901960783f,
+0.17301058823529414f, 0.2256055294117647f, 0.31164935294117646f,
+0.17101137254901969f, 0.22314490196078438f, 0.3088811568627452f,
+0.16901215686274512f, 0.22068427450980393f, 0.3061129607843137f,
+0.1670129411764706f, 0.21822364705882352f, 0.30334476470588234f,
+0.1650137254901961f, 0.21576301960784314f, 0.30057656862745097f,
+0.16301450980392157f, 0.21330239215686275f, 0.2978083725490196f,
+0.1610152941176471f, 0.21084176470588237f, 0.2950401764705882f,
+0.15901607843137255f, 0.20838113725490195f, 0.29227198039215685f,
+0.15701686274509805f, 0.20592050980392157f, 0.2895037843137255f,
+0.15501764705882354f, 0.20345988235294118f, 0.2867355882352941f,
+0.15301843137254903f, 0.2009992549019608f, 0.28396739215686273f,
+0.15101921568627452f, 0.1985386274509804f, 0.28119919607843136f,
+0.14902f, 0.196078f, 0.278431f
+}
+, 
+{
+0.0862745098039216f, 0.00392156862745098f, 0.298039215686275f,
+0.08982329931935286f, 0.006456462026374212f, 0.317811345363286f,
+0.09337208883478411f, 0.008991355425297444f, 0.33758347504029695f,
+0.09692087835021537f, 0.011526248824220677f, 0.35735560471730793f,
+0.10046966786564662f, 0.014061142223143908f, 0.3771277343943189f,
+0.10401845738107789f, 0.016596035622067143f, 0.3968998640713299f,
+0.10756724689650915f, 0.019130929020990374f, 0.41667199374834085f,
+0.1111160364119404f, 0.021665822419913605f, 0.4364441234253518f,
+0.11340168241333767f, 0.024661032165175043f, 0.45340494373553814f,
+0.11218083286206953f, 0.02893408342513443f, 0.4625615488616119f,
+0.1109599833108014f, 0.03320713468509382f, 0.4717181539876857f,
+0.10973913375953327f, 0.03748018594505321f, 0.48087475911375943f,
+0.10851828420826513f, 0.0417532372050126f, 0.49003136423983323f,
+0.10729743465699701f, 0.04602628846497199f, 0.499187969365907f,
+0.10607658510572887f, 0.05029933972493138f, 0.5083445744919808f,
+0.09331021241785199f, 0.04876183815739442f, 0.5194176988394122f,
+0.07835474816590786f, 0.046122625387451474f, 0.5308542052536591f,
+0.06339928391396372f, 0.04348341261750853f, 0.5422907116679061f,
+0.04844381966201959f, 0.04084419984756559f, 0.5537272180821531f,
+0.038521432139711945f, 0.04442266469293837f, 0.5642553393014402f,
+0.03670853298500114f, 0.05801929278187287f, 0.5733198350749943f,
+0.03489563383029033f, 0.07161592087080738f, 0.5823843308485482f,
+0.033082734675579525f, 0.0852125489597419f, 0.5914488266221023f,
+0.03153129829389241f, 0.09893905159962488f, 0.6005293276463081f,
+0.0343354355690541f, 0.11482906748563648f, 0.609876451896847f,
+0.03713957284421579f, 0.1307190833716481f, 0.619223576147386f,
+0.03994371011937749f, 0.1466090992576597f, 0.628570700397925f,
+0.04274784739453918f, 0.1624991151436713f, 0.6379178246484639f,
+0.04481715070521909f, 0.1759049571056668f, 0.6448154071591428f,
+0.046767936262570996f, 0.1889101388823944f, 0.6513179151390789f,
+0.0487187218199229f, 0.20191532065912202f, 0.657820423119015f,
+0.050669507377274804f, 0.2149205024358496f, 0.6643229310989512f,
+0.05262029293462671f, 0.2279256842125772f, 0.6708254390788874f,
+0.0545710784919786f, 0.24093086598930472f, 0.6773279470588235f,
+0.054902f, 0.2508731667144937f, 0.6816883716056075f,
+0.054902f, 0.26018975245027703f, 0.6856111906136431f,
+0.054902f, 0.2695063381860604f, 0.6895340096216785f,
+0.054902f, 0.27882292392184366f, 0.6934568286297141f,
+0.054902f, 0.288139509657627f, 0.6973796476377497f,
+0.054902f, 0.2974560953934104f, 0.7013024666457853f,
+0.054902f, 0.3067726811291937f, 0.7052252856538209f,
+0.054902f, 0.31608926686497696f, 0.7091481046618564f,
+0.054569855699292166f, 0.3242898013180376f, 0.7124610866487674f,
+0.054171027288259704f, 0.33226626783825564f, 0.7156516325766821f,
+0.05377219887722724f, 0.3402427343584737f, 0.7188421785045969f,
+0.05337337046619478f, 0.3482192008786917f, 0.7220327244325115f,
+0.05297454205516231f, 0.3561956673989098f, 0.7252232703604263f,
+0.05257571364412985f, 0.36417213391912784f, 0.728413816288341f,
+0.05217688523309739f, 0.37214860043934583f, 0.7316043622162558f,
+0.05177805682206493f, 0.38012506695956383f, 0.7347949081441705f,
+0.051379228411032464f, 0.38810153347978193f, 0.7379854540720853f,
+0.0509804f, 0.396078f, 0.741176f,
+0.0491349568627451f, 0.4071507843137255f, 0.7454820392156862f,
+0.04728951372549021f, 0.4182235686274509f, 0.7497880784313725f,
+0.0454440705882353f, 0.42929635294117646f, 0.7540941176470588f,
+0.04359862745098039f, 0.44036913725490195f, 0.7584001568627451f,
+0.04175318431372549f, 0.45144192156862745f, 0.7627061960784314f,
+0.0399077411764706f, 0.4625147058823529f, 0.7670122352941176f,
+0.03844675882352941f, 0.47358739215686274f, 0.7705493529411764f,
+0.03721645294117647f, 0.48466001960784316f, 0.7736251176470588f,
+0.035986147058823525f, 0.49573264705882353f, 0.7767008823529411f,
+0.0347558411764706f, 0.5068052745098038f, 0.7797766470588235f,
+0.03352553529411764f, 0.5178779019607843f, 0.7828524117647059f,
+0.0322952294117647f, 0.5289505294117647f, 0.7859281764705882f,
+0.0313725f, 0.5402018875658186f, 0.789119092570842f,
+0.0313725f, 0.5519894378290925f, 0.7926554628542101f,
+0.0313725f, 0.5637769880923663f, 0.796191833137578f,
+0.0313725f, 0.5755645383556405f, 0.7997282034209461f,
+0.0313725f, 0.5873520886189144f, 0.803264573704314f,
+0.0313725f, 0.5991396388821885f, 0.8068009439876821f,
+0.0313725f, 0.6109271891454624f, 0.81033731427105f,
+0.030686504498252627f, 0.6239180159928419f, 0.8134800106205788f,
+0.02953605546507661f, 0.637723521737313f, 0.8163561698742561f,
+0.028385606431900596f, 0.6515290274817839f, 0.8192323291279332f,
+0.027235157398724598f, 0.6653345332262547f, 0.8221084883816103f,
+0.026084708365548565f, 0.6791400389707258f, 0.8249846476352876f,
+0.024934259332372546f, 0.6929455447151969f, 0.8278608068889647f,
+0.02378381029919653f, 0.7067510504596678f, 0.8307369661426419f,
+0.026665552705750366f, 0.7201084854266825f, 0.8336130111508377f,
+0.030692146324308684f, 0.7333387006472386f, 0.8364890237215664f,
+0.034718739942867005f, 0.7465689158677947f, 0.8393650362922951f,
+0.03874533356142532f, 0.7597991310883508f, 0.8422410488630238f,
+0.042771927179983585f, 0.7730293463089067f, 0.8451170614337524f,
+0.04679852079854196f, 0.7862595615294631f, 0.8479930740044811f,
+0.050825114417100276f, 0.7994897767500192f, 0.8508690865752098f,
+0.0543243452656655f, 0.8093630740305169f, 0.8543239793739822f,
+0.057802422583076996f, 0.8191017189004424f, 0.8578020921678604f,
+0.06128049990048849f, 0.828840363770368f, 0.8612802049617386f,
+0.06475857721789999f, 0.8385790086402934f, 0.8647583177556168f,
+0.06823665453531148f, 0.848317653510219f, 0.868236430549495f,
+0.08245599373837956f, 0.8578084103145681f, 0.8701036085744033f,
+0.11909686111324025f, 0.8667817205372278f, 0.8686080886509812f,
+0.1557377284881004f, 0.8757550307598874f, 0.8671125687275592f,
+0.19237859586296055f, 0.8847283409825469f, 0.865617048804137f,
+0.2290194632378207f, 0.8937016512052064f, 0.8641215288807149f,
+0.26553302715502125f, 0.9026409960498816f, 0.8629490074850659f,
+0.30057390134620043f, 0.911187417841895f, 0.8655130429914547f,
+0.3356147755373796f, 0.9197338396339085f, 0.8680770784978435f,
+0.3706556497285583f, 0.9282802614259218f, 0.8706411140042323f,
+0.405696523919738f, 0.9368266832179354f, 0.8732051495106212f,
+0.43386584145924617f, 0.9428081693701252f, 0.8717897639601007f,
+0.4549143991909399f, 0.9461317001943552f, 0.8662506400840699f,
+0.4759629569226337f, 0.9494552310185852f, 0.8607115162080392f,
+0.4970115146543274f, 0.9527787618428152f, 0.8551723923320084f,
+0.5180600723860211f, 0.9561022926670452f, 0.8496332684559776f,
+0.5391086301177149f, 0.9594258234912751f, 0.8440941445799469f,
+0.5601571878494084f, 0.9627493543155051f, 0.8385550207039163f,
+0.5800668584292327f, 0.9660728516641083f, 0.8359774258320541f,
+0.5983462540385759f, 0.9693963010935857f, 0.8376391505467928f,
+0.616625649647919f, 0.9727197505230629f, 0.8393008752615315f,
+0.6349050452572622f, 0.9760431999525403f, 0.8409625999762702f,
+0.6531844408666054f, 0.9793666493820177f, 0.8426243246910089f,
+0.6735201542774042f, 0.980392f, 0.8464029045932956f,
+0.6947733420448079f, 0.980392f, 0.8511259690250876f,
+0.7160265298122113f, 0.980392f, 0.8558490334568797f,
+0.7372797175796152f, 0.980392f, 0.8605720978886717f,
+0.7585329053470189f, 0.980392f, 0.8652951623204639f,
+0.7782885274576924f, 0.980392f, 0.8709115804190044f,
+0.7974311152536634f, 0.980392f, 0.8768936963087709f,
+0.8165737030496344f, 0.980392f, 0.8828758121985375f,
+0.8353990946608948f, 0.9813855320133272f, 0.891242498620913f,
+0.853807049407915f, 0.983686573095667f, 0.902747330120914f,
+0.8722150041549349f, 0.9859876141780067f, 0.9142521616209148f,
+0.8904041643798827f, 0.9882352915165185f, 0.9255942420873363f,
+0.8993772540094933f, 0.9882351793505613f, 0.930080917762425f,
+0.908350343639104f, 0.9882350671846042f, 0.9345675934375137f,
+0.9292490417620138f, 0.9903519556804415f, 0.9436056906327613f,
+0.9679599494660566f, 0.9956308177973706f, 0.9594418370893697f,
+0.9990904468949905f, 0.9990904468949905f, 0.9673948853461224f,
+0.9938123206509195f, 0.9938123206509195f, 0.9374855032963872f,
+0.9885341944068486f, 0.9885341944068486f, 0.9075761212466521f,
+0.9891171203205709f, 0.9847079893059514f, 0.882954871665155f,
+0.9900518820820471f, 0.9809689422600464f, 0.8586510658667726f,
+0.9909866438435234f, 0.9772298952141412f, 0.8343472600683889f,
+0.9919214056049996f, 0.9734908481682361f, 0.8100434542700058f,
+0.9921568928672992f, 0.9708277767807161f, 0.784126871617359f,
+0.9921569331317006f, 0.9685269883577208f, 0.7576672640595238f,
+0.992156973396102f, 0.9662261999347256f, 0.7312076565016885f,
+0.9918070264488894f, 0.9640061421312186f, 0.7074261899475284f,
+0.9907754776791468f, 0.9619433076077379f, 0.6888604162201971f,
+0.9897439289094041f, 0.9598804730842572f, 0.6702946424928657f,
+0.9887123801396615f, 0.9578176385607764f, 0.6517288687655345f,
+0.9874876143566166f, 0.9531259764897456f, 0.630431253475807f,
+0.986096404715532f, 0.9461697509019893f, 0.6067803348127022f,
+0.9847051950744474f, 0.9392135253142329f, 0.5831294161495973f,
+0.9833139854333628f, 0.9322572997264765f, 0.5594784974864926f,
+0.9819227757922783f, 0.9253010741387202f, 0.5358275788233877f,
+0.9805315661511936f, 0.9183448485509638f, 0.512176660160283f,
+0.9785571151317216f, 0.9109192967647322f, 0.4939020283846548f,
+0.9765176290163489f, 0.9034414121443467f, 0.47622688653939166f,
+0.9744781429009761f, 0.8959635275239616f, 0.458551744694129f,
+0.9724386567856035f, 0.8884856429035763f, 0.4408766028488664f,
+0.9703991706702307f, 0.8810077582831911f, 0.4232014610036037f,
+0.9681369441769997f, 0.8732357848772184f, 0.4056866743982491f,
+0.9643980579101176f, 0.8635141466474067f, 0.3892349646114899f,
+0.9606591716432354f, 0.853792508417595f, 0.3727832548247306f,
+0.9569202853763533f, 0.8440708701877834f, 0.35633154503797143f,
+0.9531813991094712f, 0.8343492319579716f, 0.3398798352512122f,
+0.9494425128425891f, 0.82462759372816f, 0.323428125464453f,
+0.9448745737626966f, 0.813580230728528f, 0.3124489777119139f,
+0.9402010116374271f, 0.8023639676475951f, 0.3021670456964163f,
+0.9355274495121576f, 0.7911477045666622f, 0.29188511368091874f,
+0.9308538873868881f, 0.7799314414857293f, 0.28160318166542114f,
+0.9272115884592539f, 0.7698697653777618f, 0.27359016194241303f,
+0.924029411764706f, 0.7603232352941179f, 0.26658947058823546f,
+0.9208472350701579f, 0.7507767052104736f, 0.2595887792340575f,
+0.9176650583756099f, 0.7412301751268295f, 0.2525880878798797f,
+0.9144828816810617f, 0.7316836450431854f, 0.2455873965257019f,
+0.9113007049865137f, 0.7221371149595414f, 0.23858670517152414f,
+0.9079179326501371f, 0.7119888941390766f, 0.2315218653002742f,
+0.904357030656664f, 0.701306369763206f, 0.2244000613133279f,
+0.9007961286631908f, 0.6906238453873353f, 0.2172782573263816f,
+0.8972352266697177f, 0.6799413210114648f, 0.2101564533394353f,
+0.8936743246762445f, 0.6692587966355942f, 0.20303464935248902f,
+0.890135681872749f, 0.6586229429937721f, 0.195937261121023f,
+0.8875346414565826f, 0.6499529556618228f, 0.1898683142052401f,
+0.8849336010404162f, 0.6412829683298734f, 0.18379936728945723f,
+0.8823325606242497f, 0.6326129809979241f, 0.17773042037367434f,
+0.8797315202080832f, 0.6239429936659747f, 0.17166147345789146f,
+0.877815862745098f, 0.6140716862745098f, 0.16770433333333334f,
+0.8765855882352941f, 0.6029990588235298f, 0.16585900000000006f,
+0.8753553137254901f, 0.5919264313725491f, 0.16401366666666667f,
+0.8741250392156863f, 0.5808538039215686f, 0.16216833333333333f,
+0.8728947647058823f, 0.5697811764705882f, 0.160323f,
+0.8716644901960785f, 0.5587085490196078f, 0.15847766666666666f,
+0.8702035294117647f, 0.5475590196078431f, 0.1566323137254902f,
+0.8671277647058824f, 0.5358711764705882f, 0.15478682352941175f,
+0.864052f, 0.5241833333333333f, 0.15294133333333335f,
+0.8609762352941176f, 0.5124954901960784f, 0.1510958431372549f,
+0.8579004705882353f, 0.5008076470588235f, 0.14925035294117647f,
+0.8548247058823529f, 0.4891198039215686f, 0.14740486274509804f,
+0.8517489411764706f, 0.4774319607843137f, 0.1455593725490196f,
+0.848673294117647f, 0.4671282352941176f, 0.1437138823529412f,
+0.8455976862745098f, 0.45728588235294115f, 0.14186839215686275f,
+0.8425220784313725f, 0.4474435294117647f, 0.14002290196078432f,
+0.8394464705882353f, 0.4376011764705882f, 0.13817741176470588f,
+0.8363708627450981f, 0.427758823529412f, 0.1363319215686275f,
+0.8332952549019608f, 0.4179164705882353f, 0.13448643137254904f,
+0.8302195882352942f, 0.4078434117647059f, 0.13217958823529413f,
+0.8271438235294117f, 0.3973858431372549f, 0.12910382352941177f,
+0.8240680588235294f, 0.38692827450980394f, 0.1260280588235294f,
+0.8209922941176471f, 0.37647070588235293f, 0.12295229411764706f,
+0.8179165294117647f, 0.366013137254902f, 0.11987652941176472f,
+0.8148407647058823f, 0.35555556862745097f, 0.11680076470588235f,
+0.811765f, 0.345098f, 0.113725f,
+0.8080740196078431f, 0.3327950980392157f, 0.11064932941176471f,
+0.8043830392156862f, 0.3204921960784314f, 0.10757365882352941f,
+0.8006920588235295f, 0.30818929411764706f, 0.10449798823529412f,
+0.7970010784313726f, 0.29588639215686274f, 0.10142231764705882f,
+0.7933100980392157f, 0.2835834901960784f, 0.09834664705882352f,
+0.7896191176470588f, 0.2712805882352941f, 0.09527097647058823f,
+0.7836213725490196f, 0.2585931764705882f, 0.09219526666666666f,
+0.7762395686274511f, 0.24567505882352977f, 0.08911953333333342f,
+0.7688577647058823f, 0.23275694117647058f, 0.0860438f,
+0.7614759607843137f, 0.21983882352941175f, 0.08296806666666667f,
+0.7540941568627451f, 0.20692070588235295f, 0.07989233333333333f,
+0.7467123529411764f, 0.19400258823529412f, 0.0768166f,
+0.7391767843137255f, 0.18200717647058823f, 0.07404843921568628f,
+0.7311799215686274f, 0.17277988235294117f, 0.07220299607843138f,
+0.7231830588235294f, 0.16355258823529412f, 0.07035755294117647f,
+0.7151861960784314f, 0.15432529411764706f, 0.06851210980392157f,
+0.7071893333333333f, 0.145098f, 0.06666666666666667f,
+0.6991924705882353f, 0.13587070588235295f, 0.06482122352941176f,
+0.6911956078431373f, 0.12664341176470587f, 0.06297578039215686f,
+0.6805074509803922f, 0.11687795490196078f, 0.060053833333333334f,
+0.6694348235294117f, 0.10703561764705882f, 0.0569781f,
+0.6583621960784314f, 0.09719328039215686f, 0.053902366666666666f,
+0.647289568627451f, 0.0873509431372549f, 0.050826633333333336f,
+0.636216941176471f, 0.07750860588235323f, 0.04775090000000009f,
+0.6251443137254903f, 0.06766626862745098f, 0.04467516666666667f,
+0.6136913649751099f, 0.059786774105627376f, 0.04543821455211589f,
+0.6018580949253296f, 0.053870122316882124f, 0.05004004365634768f,
+0.5900248248755493f, 0.04795347052813688f, 0.05464187276057946f,
+0.578191554825769f, 0.04203681873939163f, 0.059243701864811246f,
+0.5663582847759888f, 0.036120166950646376f, 0.06384553096904302f,
+0.5545250147262085f, 0.03020351516190113f, 0.06844736007327482f,
+0.5410219798296445f, 0.026251307171840824f, 0.07258769484519471f,
+0.5260663992311696f, 0.0240079891502823f, 0.07632657092660074f,
+0.5111108186326947f, 0.021764671128723774f, 0.08006544700800676f,
+0.49615523803421974f, 0.019521353107165247f, 0.08380432308941278f,
+0.4811996574357448f, 0.017278035085606727f, 0.08754319917081879f,
+0.4671127882020719f, 0.015107094915631003f, 0.09077531345386532f,
+0.455148071471673f, 0.013112964725519002f, 0.09276947245909949f,
+0.44318335474127407f, 0.011118834535407001f, 0.09476363146433366f,
+0.43121863801087557f, 0.009124704345295059f, 0.09675779046956778f,
+0.4192539212804763f, 0.007130574155183001f, 0.09875194947480201f,
+0.4072892045500774f, 0.0051364439650710015f, 0.10074610848003619f,
+0.39013769849135815f, 0.0037389347229546776f, 0.10049990825007929f,
+0.3648998979446442f, 0.003271567882585343f, 0.09676094913058408f,
+0.33966209739793035f, 0.0028042010422160086f, 0.09302199001108886f,
+0.3144242968512165f, 0.0023368342018466734f, 0.08928303089159366f,
+0.28918649630450255f, 0.001869467361477339f, 0.08554407177209844f,
+0.2639486957577887f, 0.0014021005211080043f, 0.08180511265260323f,
+0.23871089521107478f, 0.0009347336807386691f, 0.07806615353310803f,
+0.2134730946643609f, 0.00046736684036933476f, 0.07432719441361282f,
+0.188235294117647f, 0.0f, 0.0705882352941176f
+}
+, 
+{
+1.0f, 0.984314f, 0.8f,
+0.9992310588235294f, 0.9812382352941177f, 0.789388705882353f,
+0.9984621176470588f, 0.9781624705882354f, 0.7787774117647059f,
+0.9976931764705882f, 0.9750867058823529f, 0.7681661176470589f,
+0.9969242352941177f, 0.9720109411764706f, 0.7575548235294118f,
+0.996155294117647f, 0.9689351764705882f, 0.7469435294117648f,
+0.9953863529411765f, 0.9658594117647059f, 0.7363322352941177f,
+0.9946174117647059f, 0.9627836470588236f, 0.7257209411764706f,
+0.9938484705882353f, 0.9597078823529411f, 0.7151096470588235f,
+0.9930795294117647f, 0.9566321176470588f, 0.7044983529411765f,
+0.9923105882352942f, 0.9535563529411765f, 0.6938870588235294f,
+0.9915416470588235f, 0.9504805882352941f, 0.6832757647058824f,
+0.990772705882353f, 0.9474048235294118f, 0.6726644705882353f,
+0.9900037647058824f, 0.9443290588235294f, 0.6620531764705883f,
+0.9892348235294118f, 0.941253294117647f, 0.6514418823529412f,
+0.9884658823529412f, 0.9381775294117647f, 0.6408305882352942f,
+0.9876969411764707f, 0.9351017647058824f, 0.6302192941176471f,
+0.986928f, 0.932026f, 0.619608f,
+0.9861590588235294f, 0.9289502352941176f, 0.6089967058823529f,
+0.9853901176470589f, 0.9258744705882352f, 0.5983854117647058f,
+0.9846211764705882f, 0.9227987058823529f, 0.5877741176470588f,
+0.9838522352941177f, 0.9197229411764706f, 0.5771628235294117f,
+0.983083294117647f, 0.9166471764705882f, 0.5665515294117647f,
+0.9823143529411765f, 0.9135714117647058f, 0.5559402352941176f,
+0.9815454117647059f, 0.9104956470588235f, 0.5453289411764706f,
+0.9807764705882354f, 0.9074198823529411f, 0.5347176470588235f,
+0.9800075294117647f, 0.9041134509803921f, 0.5265669411764706f,
+0.9792385882352942f, 0.9005763529411764f, 0.5208768235294118f,
+0.9784696470588236f, 0.8970392549019608f, 0.5151867058823529f,
+0.977700705882353f, 0.893502156862745f, 0.5094965882352941f,
+0.9769317647058824f, 0.8899650588235294f, 0.5038064705882352f,
+0.9761628235294117f, 0.8864279607843137f, 0.49811635294117645f,
+0.9753938823529412f, 0.882890862745098f, 0.49242623529411766f,
+0.9746249411764706f, 0.8793537647058823f, 0.4867361176470589f,
+0.973856f, 0.8758166666666667f, 0.481046f,
+0.9730870588235294f, 0.8722795686274509f, 0.4753558823529412f,
+0.9723181176470589f, 0.8687424705882353f, 0.46966576470588234f,
+0.9715491764705882f, 0.8652053725490196f, 0.46397564705882355f,
+0.9707802352941176f, 0.8616682745098039f, 0.4582855294117647f,
+0.9700112941176471f, 0.8581311764705882f, 0.45259541176470586f,
+0.9692423529411764f, 0.8545940784313726f, 0.44690529411764707f,
+0.9684734117647059f, 0.8510569803921569f, 0.4412151764705883f,
+0.9677044705882353f, 0.8475198823529412f, 0.43552505882352943f,
+0.9669355294117647f, 0.8439827843137255f, 0.4298349411764706f,
+0.9661665882352941f, 0.8404456862745098f, 0.42414482352941174f,
+0.9653976470588235f, 0.8369085882352941f, 0.41845470588235295f,
+0.9646287058823529f, 0.8333714901960785f, 0.4127645882352941f,
+0.9638597647058823f, 0.8298343921568627f, 0.4070744705882353f,
+0.9630908235294118f, 0.8262972941176471f, 0.40138435294117647f,
+0.9623218823529411f, 0.8227601960784314f, 0.3956942352941177f,
+0.9615529411764706f, 0.8192230980392157f, 0.3900041176470588f,
+0.960784f, 0.815686f, 0.384314f,
+0.9560166274509804f, 0.8095345490196079f, 0.3807769019607843f,
+0.9512492549019608f, 0.8033830980392157f, 0.37723980392156864f,
+0.9464818823529412f, 0.7972316470588235f, 0.37370270588235294f,
+0.9417145098039216f, 0.7910801960784314f, 0.37016560784313723f,
+0.936947137254902f, 0.7849287450980392f, 0.3666285098039216f,
+0.9321797647058824f, 0.7787772941176471f, 0.3630914117647059f,
+0.9274123921568627f, 0.7726258431372549f, 0.3595543137254902f,
+0.9226450196078431f, 0.7664743921568627f, 0.3560172156862745f,
+0.9178776470588235f, 0.7603229411764706f, 0.35248011764705883f,
+0.9131102745098039f, 0.7541714901960784f, 0.34894301960784313f,
+0.9083429019607843f, 0.7480200392156863f, 0.3454059215686274f,
+0.9035755294117647f, 0.7418685882352941f, 0.3418688235294117f,
+0.8988081568627451f, 0.735717137254902f, 0.3383317254901961f,
+0.8940407843137255f, 0.7295656862745098f, 0.3347946274509804f,
+0.8892734117647059f, 0.7234142352941177f, 0.33125752941176473f,
+0.8845060392156863f, 0.7172627843137255f, 0.327720431372549f,
+0.8797386666666667f, 0.7111113333333333f, 0.3241833333333333f,
+0.8749712941176471f, 0.7049598823529412f, 0.3206462352941176f,
+0.8702039215686275f, 0.698808431372549f, 0.317109137254902f,
+0.8654365490196079f, 0.6926569803921568f, 0.31357203921568627f,
+0.8606691764705882f, 0.6865055294117647f, 0.31003494117647057f,
+0.8559018039215686f, 0.6803540784313725f, 0.3064978431372549f,
+0.8511344313725491f, 0.6742026274509805f, 0.3029607450980393f,
+0.8463670588235294f, 0.6680511764705882f, 0.2994236470588235f,
+0.8415996862745098f, 0.661899725490196f, 0.29588654901960787f,
+0.8369091960784313f, 0.6559020392156862f, 0.2925801176470588f,
+0.832295588235294f, 0.6500581176470588f, 0.28950435294117643f,
+0.8276819803921568f, 0.6442141960784313f, 0.2864285882352941f,
+0.8230683725490195f, 0.6383702745098039f, 0.28335282352941177f,
+0.8184547647058823f, 0.6325263529411764f, 0.2802770588235294f,
+0.8138411568627452f, 0.6266824313725491f, 0.2772012941176471f,
+0.8092275490196078f, 0.6208385098039215f, 0.2741255294117647f,
+0.8046139411764706f, 0.6149945882352941f, 0.2710497647058823f,
+0.8000003333333333f, 0.6091506666666666f, 0.267974f,
+0.7953867254901961f, 0.6033067450980392f, 0.26489823529411766f,
+0.7907731176470588f, 0.5974628235294117f, 0.26182247058823527f,
+0.7861595098039216f, 0.5916189019607843f, 0.25874670588235293f,
+0.7815459019607843f, 0.5857749803921568f, 0.2556709411764706f,
+0.7769322941176471f, 0.5799310588235295f, 0.25259517647058827f,
+0.7723186862745098f, 0.574087137254902f, 0.24951941176470588f,
+0.7677050784313726f, 0.5682432156862746f, 0.24644364705882352f,
+0.7630914705882352f, 0.5623992941176471f, 0.24336788235294116f,
+0.7584778627450981f, 0.5565553725490197f, 0.24029211764705882f,
+0.7538642549019607f, 0.5507114509803922f, 0.23721635294117646f,
+0.7492506470588235f, 0.5448675294117648f, 0.2341405882352941f,
+0.7446370392156862f, 0.5390236078431373f, 0.23106482352941177f,
+0.7400234313725491f, 0.53317968627451f, 0.22798905882352943f,
+0.7354098235294118f, 0.5273357647058824f, 0.22491329411764704f,
+0.7307962156862745f, 0.5214918431372549f, 0.2218375294117647f,
+0.7261826078431373f, 0.5156479215686275f, 0.21876176470588232f,
+0.721569f, 0.509804f, 0.215686f,
+0.7188008039215686f, 0.5061130980392157f, 0.21430192156862743f,
+0.7160326078431373f, 0.5024221960784314f, 0.2129178431372549f,
+0.7132644117647059f, 0.4987312941176471f, 0.21153376470588234f,
+0.7104962156862745f, 0.49504039215686285f, 0.2101496862745098f,
+0.7077280196078432f, 0.49134949019607843f, 0.20876560784313725f,
+0.7049598235294118f, 0.4876585882352941f, 0.2073815294117647f,
+0.7021916274509804f, 0.4839676862745098f, 0.20599745098039215f,
+0.699423431372549f, 0.4802767843137255f, 0.2046133725490196f,
+0.6966552352941177f, 0.4765858823529412f, 0.20322929411764706f,
+0.6938870392156863f, 0.4728949803921569f, 0.2018452156862745f,
+0.6911188431372549f, 0.4692040784313726f, 0.20046113725490194f,
+0.6883506470588235f, 0.4655131764705883f, 0.1990770588235294f,
+0.6855824509803922f, 0.46182227450980395f, 0.19769298039215685f,
+0.6828142549019608f, 0.4581313725490196f, 0.1963089019607843f,
+0.6800460588235294f, 0.45444047058823533f, 0.19492482352941176f,
+0.677277862745098f, 0.45074956862745097f, 0.1935407450980392f,
+0.6745096666666667f, 0.4470586666666667f, 0.19215666666666667f,
+0.6717414705882353f, 0.44336776470588235f, 0.1907725882352941f,
+0.6689732745098039f, 0.43967686274509804f, 0.18938850980392155f,
+0.6662050784313726f, 0.4359859607843138f, 0.18800443137254905f,
+0.6634368823529412f, 0.4322950588235294f, 0.18662035294117646f,
+0.6606686862745098f, 0.4286041568627451f, 0.18523627450980393f,
+0.6579004901960784f, 0.4249132549019608f, 0.18385219607843137f,
+0.6551322941176471f, 0.4212223529411765f, 0.18246811764705884f,
+0.6523640980392157f, 0.4175314509803921f, 0.18108403921568628f,
+0.6489807843137255f, 0.41360988235294116f, 0.17954617647058824f,
+0.6449823529411765f, 0.40945764705882354f, 0.1778545294117647f,
+0.6409839215686275f, 0.40530541176470586f, 0.17616288235294117f,
+0.6369854901960784f, 0.40115317647058824f, 0.17447123529411765f,
+0.6329870588235296f, 0.3970009411764707f, 0.17277958823529416f,
+0.6289886274509804f, 0.39284870588235293f, 0.17108794117647058f,
+0.6249901960784313f, 0.3886964705882353f, 0.16939629411764706f,
+0.6209917647058824f, 0.38454423529411763f, 0.1677046470588235f,
+0.6169933333333333f, 0.380392f, 0.166013f,
+0.6129949019607843f, 0.3762397647058824f, 0.16432135294117647f,
+0.6089964705882353f, 0.3720875294117647f, 0.16262970588235293f,
+0.6049980392156863f, 0.3679352941176471f, 0.1609380588235294f,
+0.6009996078431372f, 0.3637830588235294f, 0.1592464117647059f,
+0.5970011764705883f, 0.3596308235294118f, 0.15755476470588234f,
+0.5930027450980392f, 0.35547858823529416f, 0.15586311764705882f,
+0.5890043137254902f, 0.3513263529411765f, 0.1541714705882353f,
+0.5850058823529412f, 0.34717411764705886f, 0.15247982352941175f,
+0.5810074509803921f, 0.3430218823529412f, 0.15078817647058823f,
+0.5770090196078431f, 0.33886964705882355f, 0.1490965294117647f,
+0.573010588235294f, 0.33471741176470593f, 0.14740488235294116f,
+0.5690121568627452f, 0.33056517647058836f, 0.14571323529411767f,
+0.565013725490196f, 0.32641294117647063f, 0.1440215882352941f,
+0.5610152941176471f, 0.32226070588235295f, 0.14232994117647058f,
+0.557016862745098f, 0.3181084705882353f, 0.14063829411764706f,
+0.553018431372549f, 0.31395623529411765f, 0.1389466470588235f,
+0.54902f, 0.309804f, 0.137255f,
+0.545175294117647f, 0.3061130980392157f, 0.13571711764705882f,
+0.5413305882352941f, 0.3024221960784314f, 0.13417923529411763f,
+0.5374858823529411f, 0.2987312941176471f, 0.13264135294117646f,
+0.5336411764705882f, 0.2950403921568628f, 0.1311034705882353f,
+0.5297964705882352f, 0.2913494901960785f, 0.1295655882352941f,
+0.5259517647058823f, 0.2876585882352941f, 0.12802770588235293f,
+0.5221070588235294f, 0.2839676862745098f, 0.12648982352941177f,
+0.5182623529411764f, 0.2802767843137255f, 0.12495194117647059f,
+0.5144176470588235f, 0.2765858823529412f, 0.1234140588235294f,
+0.5105729411764706f, 0.27289498039215687f, 0.12187617647058822f,
+0.5067282352941177f, 0.2692040784313727f, 0.1203382941176471f,
+0.5028835294117647f, 0.26551317647058825f, 0.11880041176470588f,
+0.49903882352941176f, 0.26182227450980394f, 0.11726252941176471f,
+0.4951941176470588f, 0.25813137254901963f, 0.11572464705882352f,
+0.49134941176470587f, 0.25444047058823527f, 0.11418676470588235f,
+0.4875047058823529f, 0.250749568627451f, 0.11264888235294117f,
+0.48366f, 0.24705866666666668f, 0.111111f,
+0.479815294117647f, 0.24336776470588234f, 0.10957311764705882f,
+0.4759705882352941f, 0.23967686274509803f, 0.10803523529411764f,
+0.4721258823529412f, 0.23598596078431372f, 0.10649735294117647f,
+0.4682811764705882f, 0.2322950588235294f, 0.10495947058823529f,
+0.4644364705882353f, 0.2286041568627451f, 0.10342158823529411f,
+0.46059176470588237f, 0.2249132549019608f, 0.10188370588235293f,
+0.4567470588235294f, 0.22122235294117648f, 0.10034582352941177f,
+0.4529023529411765f, 0.21753145098039214f, 0.0988079411764706f,
+0.44898078431372546f, 0.21376366666666666f, 0.09696249019607843f,
+0.44498235294117655f, 0.2099190000000001f, 0.09480947058823536f,
+0.4409839215686274f, 0.20607433333333333f, 0.09265645098039216f,
+0.4369854901960784f, 0.20222966666666667f, 0.09050343137254901f,
+0.4329870588235294f, 0.19838499999999998f, 0.08835041176470589f,
+0.4289886274509804f, 0.19454033333333332f, 0.08619739215686274f,
+0.42499019607843136f, 0.19069566666666665f, 0.08404437254901961f,
+0.42099176470588234f, 0.186851f, 0.08189135294117647f,
+0.4169933333333333f, 0.18300633333333333f, 0.07973833333333333f,
+0.4129949019607843f, 0.17916166666666666f, 0.0775853137254902f,
+0.4089964705882353f, 0.175317f, 0.07543229411764707f,
+0.4049980392156863f, 0.17147233333333334f, 0.07327927450980393f,
+0.40099960784313726f, 0.16762766666666667f, 0.07112625490196078f,
+0.39700117647058825f, 0.163783f, 0.06897323529411764f,
+0.39300274509803923f, 0.15993833333333332f, 0.06682021568627451f,
+0.3890043137254902f, 0.15609366666666666f, 0.06466719607843138f,
+0.3850058823529412f, 0.152249f, 0.06251417647058824f,
+0.38100745098039224f, 0.14840433333333344f, 0.06036115686274516f,
+0.37700901960784317f, 0.14455966666666667f, 0.05820813725490196f,
+0.3730105882352941f, 0.140715f, 0.056055117647058825f,
+0.3690121568627451f, 0.13687033333333332f, 0.05390209803921569f,
+0.36501372549019606f, 0.13302566666666665f, 0.05174907843137255f,
+0.36101529411764705f, 0.129181f, 0.04959605882352942f,
+0.35701686274509803f, 0.12533633333333333f, 0.04744303921568628f,
+0.353018431372549f, 0.12149166666666666f, 0.04529001960784314f,
+0.34902f, 0.117647f, 0.043137f,
+0.3451752941176471f, 0.1153402156862745f, 0.04221429411764706f,
+0.3413305882352941f, 0.11303343137254902f, 0.04129158823529412f,
+0.3374858823529412f, 0.11072664705882353f, 0.04036888235294118f,
+0.3336411764705882f, 0.10841986274509804f, 0.039446176470588235f,
+0.3297964705882353f, 0.10611307843137255f, 0.0385234705882353f,
+0.32595176470588233f, 0.10380629411764707f, 0.037600764705882354f,
+0.3221070588235294f, 0.10149950980392157f, 0.03667805882352941f,
+0.31826235294117655f, 0.09919272549019614f, 0.0357553529411765f,
+0.3144176470588235f, 0.09688594117647059f, 0.03483264705882353f,
+0.3105729411764706f, 0.0945791568627451f, 0.03390994117647059f,
+0.30672823529411763f, 0.09227237254901961f, 0.03298723529411765f,
+0.3028835294117647f, 0.08996558823529412f, 0.03206452941176471f,
+0.29903882352941175f, 0.08765880392156863f, 0.031141823529411767f,
+0.29519411764705883f, 0.08535201960784314f, 0.030219117647058823f,
+0.29134941176470586f, 0.08304523529411764f, 0.029296411764705883f,
+0.28750470588235294f, 0.08073845098039216f, 0.02837370588235294f,
+0.28366f, 0.07843166666666668f, 0.027451000000000003f,
+0.27981529411764705f, 0.07612488235294118f, 0.02652829411764706f,
+0.2759705882352941f, 0.07381809803921568f, 0.02560558823529412f,
+0.27212588235294116f, 0.07151131372549019f, 0.024682882352941176f,
+0.26828117647058825f, 0.0692045294117647f, 0.023760176470588236f,
+0.2644364705882353f, 0.06689774509803922f, 0.022837470588235292f,
+0.2605917647058823f, 0.06459096078431373f, 0.021914764705882352f,
+0.2567470588235295f, 0.0622841764705883f, 0.02099205882352944f,
+0.25290235294117647f, 0.05997739215686275f, 0.020069352941176472f,
+0.24898078431372547f, 0.05774749019607843f, 0.019223529411764705f,
+0.24498235294117646f, 0.055594470588235294f, 0.01845458823529412f,
+0.24098392156862744f, 0.05344145098039216f, 0.017685647058823528f,
+0.23698549019607842f, 0.05128843137254902f, 0.01691670588235294f,
+0.2329870588235294f, 0.049135411764705886f, 0.016147764705882355f,
+0.22898862745098036f, 0.04698239215686274f, 0.015378823529411764f,
+0.22499019607843135f, 0.044829372549019614f, 0.014609882352941177f,
+0.22099176470588233f, 0.04267635294117647f, 0.013840941176470587f,
+0.21699333333333332f, 0.040523333333333335f, 0.013072f,
+0.2129949019607843f, 0.0383703137254902f, 0.012303058823529412f,
+0.20899647058823528f, 0.03621729411764706f, 0.011534117647058825f,
+0.20499803921568627f, 0.03406427450980393f, 0.010765176470588235f,
+0.20099960784313725f, 0.031911254901960784f, 0.009996235294117648f,
+0.19700117647058824f, 0.029758235294117648f, 0.00922729411764706f,
+0.19300274509803933f, 0.027605215686274574f, 0.008458352941176493f,
+0.1890043137254902f, 0.025452196078431376f, 0.007689411764705884f,
+0.1850058823529412f, 0.023299176470588233f, 0.0069204705882352936f,
+0.18100745098039217f, 0.021146156862745097f, 0.006151529411764707f,
+0.17700901960784315f, 0.01899313725490196f, 0.005382588235294118f,
+0.17301058823529414f, 0.016840117647058825f, 0.004613647058823531f,
+0.16901215686274512f, 0.014687098039215689f, 0.003844705882352941f,
+0.1650137254901961f, 0.012534078431372553f, 0.0030757647058823542f,
+0.1610152941176471f, 0.010381058823529417f, 0.002306823529411764f,
+0.15701686274509807f, 0.00822803921568628f, 0.0015378823529411771f,
+0.15301843137254903f, 0.006075019607843138f, 0.0007689411764705868f,
+0.14902f, 0.003922f, 0.0f
+}
+, 
+{
+0.941176f, 0.941176f, 0.894118f,
+0.9362548235294118f, 0.9362548235294118f, 0.8873513333333333f,
+0.9313336470588235f, 0.9313336470588235f, 0.8805846666666667f,
+0.9264124705882353f, 0.9264124705882353f, 0.873818f,
+0.9214912941176471f, 0.9214912941176471f, 0.8670513333333333f,
+0.9165701176470589f, 0.9165701176470589f, 0.8602846666666666f,
+0.9116489411764706f, 0.9116489411764706f, 0.853518f,
+0.9067277647058823f, 0.9067277647058823f, 0.8467513333333333f,
+0.9018065882352941f, 0.9018065882352941f, 0.8399846666666666f,
+0.8968854117647058f, 0.8968854117647058f, 0.833218f,
+0.8919642352941176f, 0.8919642352941176f, 0.8264513333333333f,
+0.8870430588235294f, 0.8870430588235294f, 0.8196846666666666f,
+0.8821218823529411f, 0.8821218823529411f, 0.812918f,
+0.8772776078431372f, 0.8772776078431372f, 0.8064589215686274f,
+0.8726640392156862f, 0.8726640392156862f, 0.8009226078431372f,
+0.8680504705882353f, 0.8680504705882353f, 0.795386294117647f,
+0.8634369019607843f, 0.8634369019607843f, 0.7898499803921568f,
+0.8588233333333333f, 0.8588233333333333f, 0.7843136666666667f,
+0.8542097647058823f, 0.8542097647058823f, 0.7787773529411764f,
+0.8495961960784314f, 0.8495961960784314f, 0.7732410392156862f,
+0.8449826274509804f, 0.8449826274509804f, 0.7677047254901961f,
+0.8403690588235294f, 0.8403690588235294f, 0.7621684117647058f,
+0.8357554901960784f, 0.8357554901960784f, 0.7566320980392157f,
+0.8311419215686274f, 0.8311419215686274f, 0.7510957843137255f,
+0.8265283529411764f, 0.8265283529411764f, 0.7455594705882352f,
+0.8219147843137254f, 0.8219147843137254f, 0.7400231568627451f,
+0.8183776862745098f, 0.8182239215686274f, 0.7358709215686274f,
+0.8159170588235294f, 0.8154557647058823f, 0.7331027647058823f,
+0.813456431372549f, 0.8126876078431372f, 0.7303346078431372f,
+0.8109958039215687f, 0.8099194509803922f, 0.7275664509803922f,
+0.8085351764705883f, 0.8071512941176471f, 0.7247982941176471f,
+0.8060745490196078f, 0.804383137254902f, 0.722030137254902f,
+0.8036139215686274f, 0.8016149803921568f, 0.7192619803921568f,
+0.8011532941176471f, 0.7988468235294117f, 0.7164938235294117f,
+0.7986926666666667f, 0.7960786666666666f, 0.7137256666666666f,
+0.7962320392156863f, 0.7933105098039215f, 0.7109575098039216f,
+0.7937714117647059f, 0.7905423529411765f, 0.7081893529411765f,
+0.7913107843137255f, 0.7877741960784314f, 0.7054211960784313f,
+0.7888501568627451f, 0.7850060392156862f, 0.7026530392156862f,
+0.7859282352941177f, 0.7817764705882353f, 0.6994235294117647f,
+0.7828525490196079f, 0.7783930980392156f, 0.6960402352941176f,
+0.779776862745098f, 0.7750097254901961f, 0.6926569411764706f,
+0.7767011764705882f, 0.7716263529411764f, 0.6892736470588234f,
+0.7736254901960784f, 0.7682429803921568f, 0.6858903529411764f,
+0.7705498039215687f, 0.7648596078431372f, 0.6825070588235294f,
+0.7674741176470589f, 0.7614762352941176f, 0.6791237647058823f,
+0.764398431372549f, 0.758092862745098f, 0.6757404705882353f,
+0.7613227450980392f, 0.7547094901960784f, 0.6723571764705882f,
+0.7582470588235294f, 0.7513261176470588f, 0.6689738823529412f,
+0.7551713725490197f, 0.7479427450980392f, 0.6655905882352942f,
+0.7520956862745098f, 0.7445593725490196f, 0.662207294117647f,
+0.74902f, 0.741176f, 0.658824f,
+0.7453290980392157f, 0.737177568627451f, 0.6548254901960784f,
+0.7416381960784314f, 0.7331791372549019f, 0.6508269803921569f,
+0.737947294117647f, 0.729180705882353f, 0.6468284705882352f,
+0.7342563921568628f, 0.7251822745098039f, 0.6428299607843138f,
+0.7305654901960784f, 0.7211838431372549f, 0.6388314509803922f,
+0.7268745882352942f, 0.7171854117647058f, 0.6348329411764706f,
+0.7231836862745098f, 0.7131869803921569f, 0.630834431372549f,
+0.7194927843137254f, 0.7091885490196078f, 0.6268359215686274f,
+0.7158018823529412f, 0.7051901176470589f, 0.6228374117647059f,
+0.7121109803921568f, 0.7011916862745099f, 0.6188389019607844f,
+0.7084200784313726f, 0.6971932549019608f, 0.6148403921568628f,
+0.7047291764705882f, 0.6931948235294118f, 0.6108418823529412f,
+0.7009613725490196f, 0.6891963921568628f, 0.6069971764705883f,
+0.696962862745098f, 0.6851979607843137f, 0.6036138823529412f,
+0.6929643529411765f, 0.6811995294117648f, 0.6002305882352942f,
+0.6889658431372548f, 0.6772010980392157f, 0.596847294117647f,
+0.6849673333333333f, 0.6732026666666667f, 0.593464f,
+0.6809688235294117f, 0.6692042352941177f, 0.590080705882353f,
+0.6769703137254902f, 0.6652058039215687f, 0.5866974117647059f,
+0.6729718039215686f, 0.6612073725490196f, 0.5833141176470589f,
+0.668973294117647f, 0.6572089411764706f, 0.5799308235294118f,
+0.6649747843137255f, 0.6532105098039216f, 0.5765475294117648f,
+0.660976274509804f, 0.6492120784313726f, 0.5731642352941178f,
+0.6569777647058823f, 0.6452136470588236f, 0.5697809411764706f,
+0.6529792549019607f, 0.6412152156862745f, 0.5663976470588236f,
+0.6494421568627451f, 0.6373705490196079f, 0.5631681176470589f,
+0.6463664705882353f, 0.6336796470588235f, 0.5600923529411765f,
+0.6432907843137254f, 0.6299887450980393f, 0.5570165882352942f,
+0.6402150980392157f, 0.6262978431372549f, 0.5539408235294118f,
+0.6371394117647059f, 0.6226069411764706f, 0.5508650588235294f,
+0.6340637254901961f, 0.6189160392156864f, 0.5477892941176471f,
+0.6309880392156862f, 0.615225137254902f, 0.5447135294117648f,
+0.6279123529411764f, 0.6115342352941177f, 0.5416377647058824f,
+0.6248366666666667f, 0.6078433333333334f, 0.538562f,
+0.6217609803921569f, 0.604152431372549f, 0.5354862352941177f,
+0.618685294117647f, 0.6004615294117648f, 0.5324104705882353f,
+0.6156096078431372f, 0.5967706274509804f, 0.529334705882353f,
+0.6125339215686274f, 0.5930797254901962f, 0.5262589411764707f,
+0.6099195294117647f, 0.5900808823529412f, 0.5236445882352941f,
+0.6074589019607843f, 0.5873127254901961f, 0.5211840392156862f,
+0.6049982745098039f, 0.584544568627451f, 0.5187234901960784f,
+0.6025376470588235f, 0.581776411764706f, 0.5162629411764706f,
+0.6000770196078431f, 0.5790082549019608f, 0.5138023921568627f,
+0.5976163921568628f, 0.5762400980392157f, 0.511341843137255f,
+0.5951557647058824f, 0.5734719411764706f, 0.508881294117647f,
+0.592695137254902f, 0.5707037843137255f, 0.5064207450980392f,
+0.5902345098039217f, 0.5679356274509805f, 0.5039601960784315f,
+0.5877738823529411f, 0.5651674705882354f, 0.5014996470588235f,
+0.5853132549019608f, 0.5623993137254902f, 0.4990390980392157f,
+0.5828526274509804f, 0.5596311568627451f, 0.49657854901960785f,
+0.580392f, 0.556863f, 0.494118f,
+0.5773162352941177f, 0.5534796274509803f, 0.49104223529411767f,
+0.5742404705882354f, 0.5500962549019608f, 0.4879664705882353f,
+0.5711647058823529f, 0.5467128823529411f, 0.48489070588235295f,
+0.5680889411764707f, 0.5433295098039216f, 0.4818149411764706f,
+0.5650131764705882f, 0.5399461372549019f, 0.4787391764705882f,
+0.5619374117647059f, 0.5365627647058824f, 0.4756634117647059f,
+0.5588616470588236f, 0.5331793921568627f, 0.4725876470588235f,
+0.5557858823529411f, 0.5297960196078432f, 0.46951188235294117f,
+0.5527101176470588f, 0.5264126470588235f, 0.4664361176470588f,
+0.5496343529411765f, 0.523029274509804f, 0.46336035294117645f,
+0.5465585882352941f, 0.5196459019607843f, 0.4602845882352941f,
+0.5434828235294118f, 0.5162625294117648f, 0.4572088235294118f,
+0.5405608627450981f, 0.5130329607843137f, 0.4543637450980392f,
+0.5381003137254902f, 0.5102648039215686f, 0.45221072549019603f,
+0.5356397647058824f, 0.5074966470588235f, 0.45005770588235294f,
+0.5331792156862745f, 0.5047284901960785f, 0.4479046862745098f,
+0.5307186666666667f, 0.5019603333333333f, 0.44575166666666666f,
+0.5282581176470589f, 0.4991921764705882f, 0.4435986470588235f,
+0.525797568627451f, 0.4964240196078431f, 0.44144562745098037f,
+0.5233370196078432f, 0.4936558627450981f, 0.4392926078431373f,
+0.5208764705882353f, 0.4908877058823529f, 0.43713958823529414f,
+0.5184159215686275f, 0.48811954901960786f, 0.434986568627451f,
+0.5159553725490197f, 0.48535139215686274f, 0.43283354901960785f,
+0.5134948235294118f, 0.4825832352941176f, 0.4306805294117647f,
+0.511034274509804f, 0.47981507843137255f, 0.42852750980392157f,
+0.5090350588235294f, 0.47750829411764706f, 0.42668205882352944f,
+0.5074971764705882f, 0.4756628823529412f, 0.4251441764705883f,
+0.5059592941176471f, 0.4738174705882353f, 0.4236062941176471f,
+0.5044214117647059f, 0.4719720588235294f, 0.4220684117647059f,
+0.5028835294117647f, 0.47012664705882357f, 0.4205305294117648f,
+0.5013456470588236f, 0.46828123529411764f, 0.41899264705882355f,
+0.4998077647058824f, 0.46643582352941176f, 0.4174547647058824f,
+0.49826988235294123f, 0.4645904117647059f, 0.4159168823529412f,
+0.496732f, 0.46274499999999996f, 0.414379f,
+0.49519411764705884f, 0.4608995882352941f, 0.41284111764705883f,
+0.4936562352941177f, 0.4590541764705882f, 0.41130323529411766f,
+0.4921183529411765f, 0.45720876470588234f, 0.4097653529411765f,
+0.49058047058823534f, 0.45536335294117647f, 0.40822747058823533f,
+0.4890425882352941f, 0.45351794117647054f, 0.4066895882352941f,
+0.48750470588235295f, 0.45167252941176467f, 0.40515170588235294f,
+0.4859668235294118f, 0.4498271176470588f, 0.4036138235294118f,
+0.4844289411764706f, 0.4479817058823529f, 0.4020759411764706f,
+0.48289105882352945f, 0.44613629411764705f, 0.40053805882352944f,
+0.48135317647058823f, 0.4442908823529412f, 0.3990001764705882f,
+0.47981529411764706f, 0.4424454705882353f, 0.39746229411764705f,
+0.47827741176470595f, 0.4406000588235295f, 0.39592441176470594f,
+0.47673952941176473f, 0.4387546470588235f, 0.3943865294117647f,
+0.47520164705882356f, 0.4369092352941176f, 0.39284864705882355f,
+0.47366376470588234f, 0.43506382352941175f, 0.39131076470588233f,
+0.4721258823529412f, 0.4332184117647059f, 0.38977288235294116f,
+0.470588f, 0.431373f, 0.388235f,
+0.46905011764705884f, 0.42952750980392157f, 0.38669711764705883f,
+0.4675122352941177f, 0.42768201960784313f, 0.38515923529411766f,
+0.46597435294117645f, 0.4258365294117647f, 0.38362135294117644f,
+0.4644364705882353f, 0.42399103921568626f, 0.3820834705882353f,
+0.4628985882352941f, 0.4221455490196079f, 0.3805455882352941f,
+0.46136070588235295f, 0.42030005882352944f, 0.37900770588235294f,
+0.4598228235294118f, 0.418454568627451f, 0.3774698235294118f,
+0.45828494117647056f, 0.41660907843137257f, 0.37593194117647055f,
+0.4567470588235294f, 0.41476358823529413f, 0.3743940588235294f,
+0.45520917647058823f, 0.4129180980392157f, 0.3728561764705882f,
+0.4536712941176471f, 0.4110726078431373f, 0.3713182941176471f,
+0.4521334117647059f, 0.4092271176470588f, 0.3697804117647059f,
+0.450364862745098f, 0.40722786274509803f, 0.3681656470588235f,
+0.44790431372549017f, 0.4047673137254902f, 0.3663202352941176f,
+0.4454437647058823f, 0.40230676470588234f, 0.36447482352941174f,
+0.4429832156862745f, 0.39984621568627454f, 0.36262941176470587f,
+0.4405226666666667f, 0.3973856666666667f, 0.360784f,
+0.4380621176470588f, 0.39492511764705884f, 0.3589385882352941f,
+0.435601568627451f, 0.392464568627451f, 0.35709317647058825f,
+0.4331410196078431f, 0.39000401960784314f, 0.3552477647058824f,
+0.4306804705882353f, 0.3875434705882353f, 0.3534023529411765f,
+0.4282199215686274f, 0.38508292156862745f, 0.35155694117647057f,
+0.4257593725490196f, 0.3826223725490196f, 0.3497115294117647f,
+0.4232988235294117f, 0.38016182352941175f, 0.3478661176470588f,
+0.4208382745098039f, 0.3777012745098039f, 0.34602070588235295f,
+0.4183776862745098f, 0.3750868823529412f, 0.34402149019607847f,
+0.4159170588235295f, 0.3723186470588236f, 0.3418684705882354f,
+0.413456431372549f, 0.3695504117647059f, 0.3397154509803922f,
+0.4109958039215686f, 0.36678217647058825f, 0.33756243137254904f,
+0.40853517647058823f, 0.36401394117647057f, 0.3354094117647059f,
+0.4060745490196078f, 0.36124570588235294f, 0.33325639215686276f,
+0.40361392156862747f, 0.3584774705882353f, 0.3311033725490196f,
+0.40115329411764705f, 0.35570923529411763f, 0.32895035294117647f,
+0.39869266666666664f, 0.352941f, 0.32679733333333333f,
+0.3962320392156863f, 0.3501727647058823f, 0.3246443137254902f,
+0.39377141176470587f, 0.3474045294117647f, 0.3224912941176471f,
+0.39131078431372546f, 0.34463629411764707f, 0.32033827450980396f,
+0.3888501568627451f, 0.3418680588235294f, 0.3181852549019608f,
+0.38662023529411765f, 0.33956123529411764f, 0.31603223529411767f,
+0.3844672156862745f, 0.3374082156862745f, 0.3138792156862745f,
+0.38231419607843137f, 0.33525519607843135f, 0.3117261960784314f,
+0.3801611764705882f, 0.3331021764705882f, 0.30957317647058824f,
+0.37800815686274514f, 0.3309491568627451f, 0.30742015686274515f,
+0.37585513725490194f, 0.3287961372549019f, 0.30526713725490195f,
+0.3737021176470588f, 0.3266431176470588f, 0.3031141176470588f,
+0.3715490980392157f, 0.3244900980392157f, 0.3009610980392157f,
+0.36939607843137257f, 0.32233707843137255f, 0.2988080784313726f,
+0.3672430588235294f, 0.3201840588235294f, 0.29665505882352944f,
+0.3650900392156863f, 0.31803103921568626f, 0.2945020392156863f,
+0.36293701960784314f, 0.3158780196078431f, 0.29234901960784315f,
+0.360784f, 0.313725f, 0.290196f,
+0.35832345098039214f, 0.3112644509803921f, 0.28835058823529414f,
+0.3558629019607843f, 0.3088039019607843f, 0.28650517647058826f,
+0.35340235294117645f, 0.3063433529411765f, 0.28465976470588233f,
+0.3509418039215686f, 0.30388280392156863f, 0.28281435294117646f,
+0.3484812549019608f, 0.3014222549019608f, 0.2809689411764706f,
+0.34602070588235295f, 0.29896170588235293f, 0.2791235294117647f,
+0.3435601568627451f, 0.2965011568627451f, 0.27727811764705884f,
+0.3410996078431373f, 0.29404060784313735f, 0.27543270588235297f,
+0.3386390588235294f, 0.29158005882352944f, 0.27358729411764704f,
+0.33617850980392155f, 0.2891195098039216f, 0.27174188235294117f,
+0.3337179607843137f, 0.28665896078431374f, 0.2698964705882353f,
+0.33125741176470586f, 0.2841984117647059f, 0.2680510588235294f,
+0.3288737450980392f, 0.2817378431372549f, 0.2661287450980392f,
+0.32672072549019604f, 0.2792772156862745f, 0.26397572549019604f,
+0.3245677058823529f, 0.27681658823529415f, 0.26182270588235296f,
+0.3224146862745098f, 0.27435596078431373f, 0.2596696862745098f,
+0.32026166666666667f, 0.2718953333333333f, 0.25751666666666667f,
+0.3181086470588235f, 0.26943470588235297f, 0.25536364705882353f,
+0.3159556274509804f, 0.26697407843137255f, 0.2532106274509804f,
+0.31380260784313724f, 0.26451345098039214f, 0.25105760784313724f,
+0.3116495882352941f, 0.2620528235294118f, 0.24890458823529413f,
+0.30949656862745095f, 0.25959219607843137f, 0.24675156862745098f,
+0.3073435490196078f, 0.25713156862745096f, 0.24459854901960784f,
+0.3051905294117647f, 0.25467094117647066f, 0.24244552941176478f,
+0.3030375098039215f, 0.2522103137254902f, 0.24029250980392158f,
+0.2999617450980392f, 0.2491345882352941f, 0.23767811764705885f,
+0.2959632352941176f, 0.24544376470588233f, 0.23460235294117648f,
+0.29196472549019603f, 0.24175294117647056f, 0.23152658823529412f,
+0.2879662156862745f, 0.2380621176470588f, 0.22845082352941176f,
+0.2839677058823529f, 0.23437129411764704f, 0.22537505882352943f,
+0.27996919607843135f, 0.23068047058823526f, 0.22229929411764707f,
+0.27597068627450977f, 0.22698964705882352f, 0.2192235294117647f,
+0.2719721764705882f, 0.22329882352941174f, 0.21614776470588237f,
+0.26797366666666667f, 0.219608f, 0.213072f,
+0.2639751568627451f, 0.21591717647058822f, 0.20999623529411765f,
+0.2599766470588235f, 0.21222635294117648f, 0.20692047058823532f,
+0.2559781372549019f, 0.2085355294117647f, 0.20384470588235296f,
+0.25197962745098035f, 0.20484470588235293f, 0.2007689411764706f,
+0.2491345882352941f, 0.20253788235294118f, 0.1986159411764706f,
+0.2466740392156863f, 0.2006923921568628f, 0.19677052941176476f,
+0.2442134901960784f, 0.1988469019607843f, 0.19492511764705883f,
+0.24175294117647056f, 0.19700141176470587f, 0.19307970588235296f,
+0.23929239215686274f, 0.19515592156862743f, 0.19123429411764706f,
+0.2368318431372549f, 0.19331043137254902f, 0.18938888235294118f,
+0.23437129411764704f, 0.1914649411764706f, 0.18754347058823528f,
+0.23191074509803922f, 0.18961945098039215f, 0.1856980588235294f,
+0.22945019607843137f, 0.18777396078431371f, 0.18385264705882354f,
+0.22698964705882352f, 0.1859284705882353f, 0.18200723529411764f,
+0.22452909803921567f, 0.18408298039215687f, 0.18016182352941176f,
+0.22206854901960785f, 0.18223749019607843f, 0.17831641176470586f,
+0.219608f, 0.180392f, 0.176471f
+}
+, 
+{
+1.0f, 0.933333f, 0.8f,
+0.996078431372549f, 0.9291807647058823f, 0.7936947254901962f,
+0.9921568627450981f, 0.9250285294117647f, 0.7873894509803921f,
+0.9882352941176471f, 0.920876294117647f, 0.7810841764705883f,
+0.9843137254901961f, 0.9167240588235294f, 0.7747789019607844f,
+0.9803921568627451f, 0.9125718235294117f, 0.7684736274509805f,
+0.9764705882352941f, 0.9084195882352941f, 0.7621683529411765f,
+0.9725490196078431f, 0.9042673529411764f, 0.7558630784313726f,
+0.9686274509803922f, 0.9001151176470588f, 0.7495578039215687f,
+0.9647058823529412f, 0.8959628823529412f, 0.7432525294117648f,
+0.9607843137254902f, 0.8918106470588235f, 0.7369472549019608f,
+0.9568627450980393f, 0.8876584117647058f, 0.7306419803921569f,
+0.9529411764705883f, 0.8835061764705883f, 0.724336705882353f,
+0.9490196078431372f, 0.8793539411764706f, 0.7180314313725491f,
+0.9450980392156862f, 0.8752017058823529f, 0.7117261568627451f,
+0.9411764705882353f, 0.8710494705882352f, 0.7054208823529412f,
+0.9372549019607843f, 0.8668972352941177f, 0.6991156078431373f,
+0.9333333333333333f, 0.862745f, 0.6928103333333333f,
+0.9294117647058824f, 0.8585927647058823f, 0.6865050588235294f,
+0.9254901960784314f, 0.8544405294117647f, 0.6801997843137255f,
+0.9215686274509804f, 0.8502882941176471f, 0.6738945098039216f,
+0.9176470588235295f, 0.8461360588235294f, 0.6675892352941177f,
+0.9137254901960784f, 0.8419838235294117f, 0.6612839607843137f,
+0.9098039215686274f, 0.8378315882352941f, 0.6549786862745098f,
+0.9058823529411765f, 0.8336793529411765f, 0.6486734117647059f,
+0.9019607843137255f, 0.8295271176470588f, 0.6423681372549019f,
+0.8980392156862745f, 0.8253748823529412f, 0.636062862745098f,
+0.8941176470588236f, 0.8212226470588235f, 0.6297575882352942f,
+0.8901960784313726f, 0.8170704117647059f, 0.6234523137254901f,
+0.8862745098039216f, 0.8129181764705883f, 0.6171470392156864f,
+0.8823529411764706f, 0.8087659411764706f, 0.6108417647058824f,
+0.8784313725490196f, 0.8046137058823529f, 0.6045364901960785f,
+0.8745098039215686f, 0.8004614705882354f, 0.5982312156862746f,
+0.8705882352941177f, 0.7963092352941177f, 0.5919259411764707f,
+0.8666666666666667f, 0.792157f, 0.5856206666666667f,
+0.8627450980392157f, 0.7880047647058823f, 0.5793153921568628f,
+0.8588235294117648f, 0.7838525294117646f, 0.5730101176470588f,
+0.8549019607843138f, 0.7797002941176471f, 0.566704843137255f,
+0.8509803921568628f, 0.7755480588235294f, 0.560399568627451f,
+0.8470588235294119f, 0.7713958235294118f, 0.5540942941176471f,
+0.8431372549019608f, 0.7672435882352941f, 0.5477890196078432f,
+0.8392156862745099f, 0.7630913529411765f, 0.5414837450980392f,
+0.8352941176470589f, 0.7589391176470588f, 0.5351784705882353f,
+0.8313725490196079f, 0.7547868823529412f, 0.5288731960784314f,
+0.8274509803921569f, 0.7506346470588235f, 0.5225679215686274f,
+0.823529411764706f, 0.746482411764706f, 0.5162626470588236f,
+0.8196078431372549f, 0.7423301764705883f, 0.5099573725490196f,
+0.8156862745098039f, 0.7381779411764706f, 0.5036520980392156f,
+0.8117647058823529f, 0.7340257058823529f, 0.4973468235294118f,
+0.807843137254902f, 0.7298734705882354f, 0.4910415490196079f,
+0.803921568627451f, 0.7257212352941177f, 0.48473627450980394f,
+0.8f, 0.721569f, 0.478431f,
+0.7960784313725491f, 0.7181856862745098f, 0.4749708039215686f,
+0.7921568627450981f, 0.7148023725490197f, 0.47151060784313725f,
+0.7882352941176471f, 0.7114190588235294f, 0.46805041176470585f,
+0.7843137254901961f, 0.7080357450980392f, 0.4645902156862745f,
+0.7803921568627451f, 0.7046524313725491f, 0.4611300196078431f,
+0.7764705882352941f, 0.7012691176470589f, 0.45766982352941177f,
+0.7725490196078432f, 0.6978858039215686f, 0.45420962745098037f,
+0.7686274509803922f, 0.6945024901960785f, 0.450749431372549f,
+0.7647058823529412f, 0.6911191764705882f, 0.4472892352941176f,
+0.7607843137254903f, 0.687735862745098f, 0.4438290392156863f,
+0.7568627450980392f, 0.6843525490196078f, 0.4403688431372549f,
+0.7529411764705882f, 0.6809692352941177f, 0.43690864705882354f,
+0.7490196078431373f, 0.6775859215686274f, 0.43344845098039214f,
+0.7450980392156863f, 0.6742026078431372f, 0.4299882549019608f,
+0.7411764705882353f, 0.6708192941176471f, 0.42652805882352945f,
+0.7372549019607844f, 0.6674359803921569f, 0.423067862745098f,
+0.7333333333333334f, 0.6640526666666666f, 0.41960766666666666f,
+0.7294117647058824f, 0.6606693529411765f, 0.41614747058823526f,
+0.7254901960784313f, 0.6572860392156863f, 0.4126872745098039f,
+0.7215686274509804f, 0.6539027254901961f, 0.4092270784313725f,
+0.7176470588235294f, 0.6505194117647058f, 0.40576688235294117f,
+0.7137254901960784f, 0.6471360980392157f, 0.4023066862745098f,
+0.7098039215686275f, 0.6437527843137255f, 0.3988464901960785f,
+0.7058823529411765f, 0.6403694705882352f, 0.39538629411764703f,
+0.7019607843137255f, 0.6369861568627451f, 0.3919260980392157f,
+0.6980392156862745f, 0.6336028431372549f, 0.3884659019607843f,
+0.6941176470588235f, 0.6302195294117647f, 0.38500570588235294f,
+0.6901960784313725f, 0.6268362156862745f, 0.38154550980392155f,
+0.6862745098039216f, 0.6234529019607843f, 0.3780853137254902f,
+0.6823529411764706f, 0.6200695882352941f, 0.3746251176470588f,
+0.6784313725490196f, 0.616686274509804f, 0.37116492156862746f,
+0.6745098039215687f, 0.6133029607843137f, 0.36770472549019606f,
+0.6705882352941177f, 0.6099196470588235f, 0.36424452941176466f,
+0.6666666666666667f, 0.6065363333333333f, 0.3607843333333333f,
+0.6627450980392157f, 0.6031530196078431f, 0.357324137254902f,
+0.6588235294117647f, 0.5997697058823529f, 0.3538639411764706f,
+0.6549019607843137f, 0.5963863921568627f, 0.3504037450980392f,
+0.6509803921568628f, 0.5930030784313725f, 0.34694354901960783f,
+0.6470588235294118f, 0.5896197647058824f, 0.3434833529411765f,
+0.6431372549019607f, 0.5862364509803921f, 0.3400231568627451f,
+0.6392156862745098f, 0.582853137254902f, 0.3365629607843137f,
+0.6352941176470588f, 0.5794698235294118f, 0.33310276470588235f,
+0.6313725490196078f, 0.5760865098039215f, 0.32964256862745095f,
+0.6274509803921569f, 0.5727031960784313f, 0.3261823725490196f,
+0.6235294117647059f, 0.5693198823529412f, 0.3227221764705882f,
+0.6196078431372549f, 0.5659365686274509f, 0.3192619803921568f,
+0.615686274509804f, 0.5625532549019607f, 0.3158017843137255f,
+0.611764705882353f, 0.5591699411764706f, 0.3123415882352941f,
+0.6078431372549019f, 0.5557866274509804f, 0.3088813921568627f,
+0.6039215686274509f, 0.5524033137254901f, 0.3054211960784313f,
+0.6f, 0.54902f, 0.301961f,
+0.596078431372549f, 0.5457135686274509f, 0.29919282352941173f,
+0.592156862745098f, 0.5424071372549019f, 0.2964246470588235f,
+0.5882352941176471f, 0.5391007058823529f, 0.2936564705882353f,
+0.5843137254901961f, 0.5357942745098039f, 0.2908882941176471f,
+0.580392156862745f, 0.5324878431372548f, 0.2881201176470588f,
+0.5764705882352941f, 0.5291814117647058f, 0.28535194117647056f,
+0.5725490196078431f, 0.5258749803921569f, 0.2825837647058823f,
+0.5686274509803921f, 0.5225685490196078f, 0.2798155882352941f,
+0.5647058823529412f, 0.5192621176470588f, 0.2770474117647059f,
+0.5607843137254902f, 0.5159556862745097f, 0.27427923529411763f,
+0.5568627450980392f, 0.5126492549019608f, 0.2715110588235294f,
+0.5529411764705883f, 0.5093428235294117f, 0.2687428823529412f,
+0.5490196078431373f, 0.5060363921568627f, 0.26597470588235295f,
+0.5450980392156862f, 0.5027299607843136f, 0.2632065294117647f,
+0.5411764705882353f, 0.49942352941176466f, 0.26043835294117645f,
+0.5372549019607843f, 0.49611709803921566f, 0.2576701764705882f,
+0.5333333333333333f, 0.4928106666666666f, 0.25490199999999996f,
+0.5294117647058824f, 0.48950423529411763f, 0.25213382352941177f,
+0.5254901960784314f, 0.4861978039215686f, 0.24936564705882353f,
+0.5215686274509804f, 0.48289137254901965f, 0.24659747058823533f,
+0.5176470588235295f, 0.47958494117647055f, 0.24382929411764706f,
+0.5137254901960784f, 0.47627850980392156f, 0.24106111764705881f,
+0.5098039215686274f, 0.4729720784313725f, 0.2382929411764706f,
+0.5058823529411764f, 0.4696656470588235f, 0.23552476470588235f,
+0.5019607843137255f, 0.4663592156862745f, 0.2327565882352941f,
+0.4980392156862745f, 0.4630527843137255f, 0.22998841176470589f,
+0.49411764705882355f, 0.45974635294117644f, 0.22722023529411764f,
+0.49019607843137253f, 0.45643992156862745f, 0.22445205882352942f,
+0.48627450980392156f, 0.4531334901960784f, 0.22168388235294118f,
+0.4823529411764707f, 0.44982705882352947f, 0.218915705882353f,
+0.47843137254901963f, 0.44652062745098037f, 0.2161475294117647f,
+0.4745098039215686f, 0.4432141960784314f, 0.21337935294117646f,
+0.47058823529411764f, 0.43990776470588233f, 0.21061117647058825f,
+0.4666666666666667f, 0.43660133333333334f, 0.207843f,
+0.4627450980392157f, 0.4332949019607843f, 0.20507482352941175f,
+0.45882352941176474f, 0.4299884705882353f, 0.2023066470588235f,
+0.4549019607843138f, 0.42668203921568626f, 0.19953847058823532f,
+0.45098039215686275f, 0.4233756078431372f, 0.19677029411764707f,
+0.4470588235294118f, 0.4200691764705883f, 0.19400211764705882f,
+0.44313725490196076f, 0.41676274509803923f, 0.19123394117647058f,
+0.4392156862745098f, 0.4134563137254902f, 0.18846576470588233f,
+0.43529411764705883f, 0.41014988235294114f, 0.18569758823529414f,
+0.43137254901960786f, 0.40684345098039215f, 0.1829294117647059f,
+0.4274509803921569f, 0.40353701960784316f, 0.18016123529411765f,
+0.42352941176470593f, 0.4002305882352941f, 0.1773930588235294f,
+0.419607843137255f, 0.3969241568627452f, 0.17462488235294127f,
+0.41568627450980394f, 0.3936177254901961f, 0.17185670588235297f,
+0.411764705882353f, 0.3903112941176471f, 0.16908852941176472f,
+0.40784313725490196f, 0.38700486274509804f, 0.16632035294117647f,
+0.403921568627451f, 0.383698431372549f, 0.16355217647058826f,
+0.4f, 0.380392f, 0.160784f,
+0.39807766666666666f, 0.37885413725490197f, 0.1587847843137255f,
+0.39615533333333336f, 0.37731627450980393f, 0.156785568627451f,
+0.394233f, 0.3757784117647059f, 0.15478635294117649f,
+0.3923106666666667f, 0.37424054901960785f, 0.15278713725490198f,
+0.39038833333333334f, 0.3727026862745098f, 0.15078792156862747f,
+0.38846600000000003f, 0.37116482352941177f, 0.14878870588235296f,
+0.3865436666666667f, 0.36962696078431373f, 0.14678949019607845f,
+0.38462133333333337f, 0.3680890980392157f, 0.14479027450980392f,
+0.382699f, 0.36655123529411765f, 0.1427910588235294f,
+0.38077666666666665f, 0.3650133725490196f, 0.1407918431372549f,
+0.3788543333333334f, 0.3634755098039216f, 0.13879262745098045f,
+0.376932f, 0.36193764705882353f, 0.13679341176470589f,
+0.3750096666666667f, 0.3603997843137255f, 0.13479419607843138f,
+0.3730873333333333f, 0.35886192156862745f, 0.13279498039215687f,
+0.371165f, 0.3573240588235294f, 0.13079576470588236f,
+0.36924266666666666f, 0.35578619607843137f, 0.12879654901960785f,
+0.36732033333333336f, 0.35424833333333333f, 0.12679733333333335f,
+0.365398f, 0.3527104705882353f, 0.12479811764705884f,
+0.36347566666666664f, 0.35117260784313725f, 0.12279890196078433f,
+0.36155333333333334f, 0.3496347450980392f, 0.12079968627450982f,
+0.35963100000000003f, 0.3480968823529412f, 0.11880047058823531f,
+0.3577086666666667f, 0.34655901960784313f, 0.11680125490196079f,
+0.3557863333333333f, 0.3450211568627451f, 0.11480203921568628f,
+0.353864f, 0.34348329411764705f, 0.11280282352941177f,
+0.35194166666666665f, 0.341945431372549f, 0.11080360784313725f,
+0.35001933333333335f, 0.340407568627451f, 0.10880439215686274f,
+0.34809700000000005f, 0.338869705882353f, 0.10680517647058829f,
+0.34617466666666663f, 0.3373318431372549f, 0.10480596078431373f,
+0.3442523333333333f, 0.33579398039215685f, 0.10280674509803922f,
+0.34233f, 0.3342561176470588f, 0.10080752941176471f,
+0.34040766666666666f, 0.3327182549019608f, 0.0988083137254902f,
+0.3384853333333333f, 0.33118039215686274f, 0.0968090980392157f,
+0.336563f, 0.3296425294117647f, 0.09480988235294117f,
+0.3346406666666667f, 0.32810466666666666f, 0.09281066666666668f,
+0.33271833333333334f, 0.3265668039215686f, 0.09081145098039216f,
+0.330796f, 0.3250289411764706f, 0.08881223529411765f,
+0.3288736666666667f, 0.32349107843137254f, 0.08681301960784314f,
+0.3269513333333333f, 0.3219532156862745f, 0.08481380392156863f,
+0.325029f, 0.32041535294117646f, 0.08281458823529413f,
+0.32310666666666665f, 0.3188774901960784f, 0.08081537254901962f,
+0.3211843333333333f, 0.3173396274509804f, 0.0788161568627451f,
+0.319262f, 0.31580176470588234f, 0.0768169411764706f,
+0.3173396666666667f, 0.31426390196078435f, 0.07481772549019614f,
+0.3154173333333333f, 0.31272603921568626f, 0.07281850980392157f,
+0.31349499999999997f, 0.3111881764705882f, 0.07081929411764706f,
+0.31157266666666666f, 0.3096503137254902f, 0.06882007843137256f,
+0.3096503333333333f, 0.30811245098039214f, 0.06682086274509805f,
+0.307728f, 0.3065745882352941f, 0.06482164705882354f,
+0.30580566666666664f, 0.30503672549019606f, 0.06282243137254902f,
+0.3038833333333333f, 0.303498862745098f, 0.06082321568627451f,
+0.301961f, 0.301961f, 0.058824f,
+0.29965419607843136f, 0.2999617647058823f, 0.05767058823529412f,
+0.29734739215686273f, 0.2979625294117647f, 0.05651717647058824f,
+0.2950405882352941f, 0.29596329411764705f, 0.055363764705882355f,
+0.2927337843137255f, 0.2939640588235294f, 0.05421035294117647f,
+0.29042698039215686f, 0.2919648235294117f, 0.05305694117647059f,
+0.28812017647058824f, 0.2899655882352941f, 0.05190352941176471f,
+0.2858133725490196f, 0.28796635294117645f, 0.05075011764705882f,
+0.28350656862745105f, 0.28596711764705884f, 0.04959670588235297f,
+0.2811997647058823f, 0.2839678823529412f, 0.04844329411764706f,
+0.2788929607843137f, 0.2819686470588235f, 0.04728988235294118f,
+0.27658615686274507f, 0.27996941176470586f, 0.04613647058823529f,
+0.27427935294117645f, 0.27797017647058825f, 0.04498305882352941f,
+0.2719725490196078f, 0.2759709411764706f, 0.04382964705882353f,
+0.2696657450980392f, 0.2739717058823529f, 0.04267623529411765f,
+0.2673589411764706f, 0.27197247058823526f, 0.041522823529411765f,
+0.26505213725490195f, 0.26997323529411765f, 0.04036941176470588f,
+0.26274533333333333f, 0.267974f, 0.039216f,
+0.2604385294117647f, 0.2659747647058823f, 0.03806258823529411f,
+0.2581317254901961f, 0.2639755294117647f, 0.03690917647058824f,
+0.25582492156862746f, 0.26197629411764706f, 0.035755764705882355f,
+0.25351811764705884f, 0.2599770588235294f, 0.03460235294117647f,
+0.2512113137254902f, 0.25797782352941173f, 0.033448941176470584f,
+0.24890450980392156f, 0.2559785882352941f, 0.03229552941176471f,
+0.246597705882353f, 0.2539793529411765f, 0.03114211764705886f,
+0.24429090196078432f, 0.2519801176470588f, 0.02998870588235294f,
+0.24198409803921567f, 0.2499808823529412f, 0.028835294117647063f,
+0.23967729411764704f, 0.24798164705882353f, 0.027681882352941178f,
+0.23737049019607842f, 0.24598241176470587f, 0.026528470588235292f,
+0.2350636862745098f, 0.24398317647058823f, 0.02537505882352941f,
+0.23275688235294117f, 0.2419839411764706f, 0.02422164705882353f,
+0.23045007843137255f, 0.23998470588235293f, 0.023068235294117653f,
+0.22814327450980393f, 0.2379854705882353f, 0.021914823529411764f,
+0.2258364705882353f, 0.23598623529411766f, 0.020761411764705882f,
+0.22352966666666668f, 0.233987f, 0.019608f,
+0.22122286274509803f, 0.23198776470588234f, 0.01845458823529412f,
+0.2189160588235294f, 0.2299885294117647f, 0.01730117647058823f,
+0.21660925490196078f, 0.22798929411764707f, 0.016147764705882355f,
+0.21430245098039213f, 0.2259900588235294f, 0.014994352941176473f,
+0.21199564705882354f, 0.22399082352941177f, 0.01384094117647059f,
+0.20968884313725497f, 0.2219915882352942f, 0.012687529411764736f,
+0.20738203921568626f, 0.21999235294117647f, 0.01153411764705882f,
+0.20507523529411764f, 0.21799311764705884f, 0.010380705882352945f,
+0.20276843137254902f, 0.2159938823529412f, 0.009227294117647056f,
+0.2004616274509804f, 0.21399464705882354f, 0.008073882352941174f,
+0.19815482352941177f, 0.21199541176470588f, 0.006920470588235299f,
+0.19584801960784315f, 0.20999617647058824f, 0.00576705882352941f,
+0.19354121568627453f, 0.2079969411764706f, 0.004613647058823535f,
+0.1912344117647059f, 0.20599770588235294f, 0.003460235294117646f,
+0.18892760784313725f, 0.20399847058823528f, 0.002306823529411764f,
+0.18662080392156866f, 0.20199923529411767f, 0.001153411764705882f,
+0.184314f, 0.2f, 0.0f
+}
+, 
+{
+0.054902f, 0.109804f, 0.121569f,
+0.05643988235294117f, 0.11472517647058823f, 0.12618256862745098f,
+0.057977764705882354f, 0.11964635294117647f, 0.13079613725490197f,
+0.059515647058823534f, 0.1245675294117647f, 0.13540970588235293f,
+0.06105352941176471f, 0.12948870588235295f, 0.14002327450980392f,
+0.06259141176470588f, 0.13440988235294118f, 0.1446368431372549f,
+0.06412929411764706f, 0.13933105882352942f, 0.14925041176470588f,
+0.06566717647058824f, 0.14425223529411765f, 0.15386398039215687f,
+0.06720505882352942f, 0.1491734117647059f, 0.15847754901960784f,
+0.06874294117647059f, 0.15409458823529412f, 0.16309111764705883f,
+0.07028082352941177f, 0.15901576470588236f, 0.1677046862745098f,
+0.07181870588235295f, 0.1639369411764706f, 0.1723182549019608f,
+0.07335658823529412f, 0.16885811764705883f, 0.17693182352941175f,
+0.0747406862745098f, 0.17370241176470588f, 0.18116094117647058f,
+0.07566343137254902f, 0.1783160588235294f, 0.18423670588235294f,
+0.07658617647058824f, 0.18292970588235294f, 0.1873124705882353f,
+0.07750892156862746f, 0.18754335294117647f, 0.19038823529411764f,
+0.07843166666666668f, 0.192157f, 0.193464f,
+0.07935441176470588f, 0.19677064705882352f, 0.19653976470588236f,
+0.0802771568627451f, 0.20138429411764705f, 0.1996155294117647f,
+0.08119990196078432f, 0.20599794117647058f, 0.20269129411764705f,
+0.08212264705882354f, 0.21061158823529413f, 0.20576705882352941f,
+0.08304539215686275f, 0.21522523529411763f, 0.20884282352941175f,
+0.08396813725490196f, 0.2198388823529412f, 0.2119185882352941f,
+0.08489088235294118f, 0.2244525294117647f, 0.21499435294117647f,
+0.0858136274509804f, 0.22906617647058825f, 0.21807011764705883f,
+0.08658256862745098f, 0.23321841176470587f, 0.22083827450980392f,
+0.08719770588235294f, 0.23690923529411764f, 0.22329882352941177f,
+0.0878128431372549f, 0.24060005882352942f, 0.2257593725490196f,
+0.08842798039215687f, 0.24429088235294116f, 0.22821992156862744f,
+0.08904311764705883f, 0.24798170588235294f, 0.2306804705882353f,
+0.08965825490196079f, 0.2516725294117647f, 0.23314101960784311f,
+0.09027339215686274f, 0.25536335294117646f, 0.23560156862745096f,
+0.0908885294117647f, 0.2590541764705882f, 0.23806211764705879f,
+0.09150366666666666f, 0.262745f, 0.24052266666666666f,
+0.09211880392156863f, 0.26643582352941175f, 0.24298321568627448f,
+0.09273394117647059f, 0.2701266470588235f, 0.24544376470588233f,
+0.09334907843137254f, 0.27381747058823525f, 0.24790431372549016f,
+0.0939642156862745f, 0.27750829411764705f, 0.250364862745098f,
+0.09504070588235293f, 0.28258329411764704f, 0.2525947647058823f,
+0.09627098039215685f, 0.2881196862745098f, 0.25474778431372547f,
+0.09750125490196077f, 0.2936560784313725f, 0.2569008039215686f,
+0.0987315294117647f, 0.2991924705882353f, 0.25905382352941175f,
+0.09996180392156863f, 0.30472886274509803f, 0.2612068431372549f,
+0.10119207843137254f, 0.3102652549019608f, 0.26335986274509804f,
+0.10242235294117646f, 0.31580164705882346f, 0.2655128823529411f,
+0.10365262745098039f, 0.32133803921568627f, 0.26766590196078427f,
+0.10488290196078431f, 0.326874431372549f, 0.2698189215686274f,
+0.10611317647058824f, 0.33241082352941176f, 0.27197194117647056f,
+0.10734345098039215f, 0.33794721568627445f, 0.2741249607843137f,
+0.10857372549019607f, 0.34348360784313725f, 0.27627798039215684f,
+0.109804f, 0.34902f, 0.278431f,
+0.1101115294117647f, 0.353018431372549f, 0.278431f,
+0.11041905882352941f, 0.35701686274509803f, 0.278431f,
+0.11072658823529412f, 0.36101529411764705f, 0.278431f,
+0.11103411764705882f, 0.36501372549019606f, 0.278431f,
+0.11134164705882353f, 0.36901215686274513f, 0.278431f,
+0.11164917647058824f, 0.3730105882352941f, 0.278431f,
+0.11195670588235294f, 0.37700901960784317f, 0.278431f,
+0.11226423529411765f, 0.3810074509803922f, 0.278431f,
+0.11257176470588236f, 0.3850058823529412f, 0.278431f,
+0.11287929411764706f, 0.38900431372549016f, 0.278431f,
+0.11318682352941177f, 0.39300274509803923f, 0.278431f,
+0.11349435294117648f, 0.39700117647058825f, 0.278431f,
+0.11380190196078432f, 0.40099960784313726f, 0.2782772156862745f,
+0.11410950980392158f, 0.4049980392156863f, 0.2776620784313725f,
+0.11441711764705882f, 0.40899647058823524f, 0.2770469411764706f,
+0.11472472549019608f, 0.4129949019607843f, 0.27643180392156863f,
+0.11503233333333333f, 0.4169933333333333f, 0.27581666666666665f,
+0.11533994117647059f, 0.42099176470588234f, 0.2752015294117647f,
+0.11564754901960785f, 0.42499019607843136f, 0.27458639215686276f,
+0.1159551568627451f, 0.4289886274509804f, 0.2739712549019608f,
+0.11626276470588236f, 0.4329870588235294f, 0.2733561176470588f,
+0.11657037254901961f, 0.4369854901960784f, 0.27274098039215683f,
+0.11687798039215685f, 0.4409839215686274f, 0.2721258431372549f,
+0.11718558823529412f, 0.4449823529411765f, 0.27151070588235293f,
+0.11749319607843138f, 0.44898078431372546f, 0.27089556862745096f,
+0.117647f, 0.45251788235294116f, 0.26951149019607845f,
+0.117647f, 0.45559364705882355f, 0.2673584705882353f,
+0.117647f, 0.4586694117647059f, 0.26520545098039217f,
+0.117647f, 0.4617451764705882f, 0.263052431372549f,
+0.117647f, 0.4648209411764706f, 0.2608994117647059f,
+0.117647f, 0.4678967058823529f, 0.2587463921568628f,
+0.117647f, 0.4709724705882353f, 0.2565933725490196f,
+0.117647f, 0.47404823529411766f, 0.25444035294117645f,
+0.117647f, 0.477124f, 0.2522873333333333f,
+0.117647f, 0.4801997647058824f, 0.25013431372549016f,
+0.117647f, 0.4832755294117647f, 0.24798129411764705f,
+0.117647f, 0.4863512941176471f, 0.2458282745098039f,
+0.117647f, 0.48942705882352944f, 0.2436752549019608f,
+0.11741629411764706f, 0.4920414705882353f, 0.24083023529411768f,
+0.11710868627450981f, 0.4945020980392157f, 0.23775454901960782f,
+0.11680107843137255f, 0.4969627254901961f, 0.23467886274509803f,
+0.1164934705882353f, 0.49942335294117646f, 0.23160317647058823f,
+0.11618586274509804f, 0.5018839803921569f, 0.22852749019607843f,
+0.11587825490196078f, 0.5043446078431373f, 0.2254518039215686f,
+0.11557064705882353f, 0.5068052352941176f, 0.2223761176470588f,
+0.11526303921568629f, 0.509265862745098f, 0.219300431372549f,
+0.11495543137254903f, 0.5117264901960784f, 0.21622474509803924f,
+0.11464782352941177f, 0.5141871176470588f, 0.2131490588235294f,
+0.11434021568627452f, 0.5166477450980391f, 0.2100733725490196f,
+0.11403260784313726f, 0.5191083725490195f, 0.2069976862745098f,
+0.113725f, 0.521569f, 0.203922f,
+0.1134174705882353f, 0.523722019607843f, 0.1999234901960784f,
+0.1131099411764706f, 0.5258750392156862f, 0.19592498039215686f,
+0.11280241176470589f, 0.5280280588235293f, 0.19192647058823528f,
+0.11249488235294118f, 0.5301810784313725f, 0.18792796078431379f,
+0.11218735294117647f, 0.5323340980392156f, 0.18392945098039215f,
+0.11187982352941177f, 0.5344871176470588f, 0.17993094117647057f,
+0.11157229411764706f, 0.5366401372549019f, 0.17593243137254902f,
+0.11126476470588235f, 0.5387931568627451f, 0.17193392156862744f,
+0.11095723529411765f, 0.5409461764705882f, 0.1679354117647059f,
+0.11064970588235294f, 0.5430991960784313f, 0.1639369019607843f,
+0.11034217647058824f, 0.5452522156862745f, 0.15993839215686273f,
+0.11003464705882353f, 0.5474052352941176f, 0.15593988235294123f,
+0.10926574509803921f, 0.5497889215686274f, 0.15155692156862743f,
+0.10711272549019607f, 0.5528646078431372f, 0.14602060784313725f,
+0.10495970588235294f, 0.555940294117647f, 0.14048429411764704f,
+0.1028066862745098f, 0.5590159803921568f, 0.13494798039215686f,
+0.10065366666666667f, 0.5620916666666667f, 0.12941166666666665f,
+0.09850064705882353f, 0.5651673529411764f, 0.12387535294117646f,
+0.09634762745098038f, 0.5682430392156862f, 0.11833903921568627f,
+0.09419460784313728f, 0.571318725490196f, 0.11280272549019615f,
+0.09204158823529411f, 0.5743944117647058f, 0.10726641176470587f,
+0.08988856862745098f, 0.5774700980392157f, 0.10173009803921568f,
+0.08773554901960784f, 0.5805457843137254f, 0.09619378431372548f,
+0.0855825294117647f, 0.5836214705882352f, 0.09065747058823528f,
+0.08342950980392157f, 0.586697156862745f, 0.08512115686274509f,
+0.08342950980392157f, 0.5899266862745097f, 0.0811226862745098f,
+0.0855825294117647f, 0.5933100588235294f, 0.0786620588235294f,
+0.08773554901960784f, 0.5966934313725489f, 0.07620143137254902f,
+0.08988856862745098f, 0.6000768039215686f, 0.07374080392156862f,
+0.09204158823529406f, 0.6034601764705881f, 0.0712801764705883f,
+0.09419460784313725f, 0.6068435490196078f, 0.06881954901960784f,
+0.09634762745098038f, 0.6102269215686275f, 0.06635892156862745f,
+0.09850064705882353f, 0.613610294117647f, 0.06389829411764705f,
+0.10065366666666667f, 0.6169936666666667f, 0.06143766666666667f,
+0.1028066862745098f, 0.6203770392156862f, 0.05897703921568627f,
+0.10495970588235294f, 0.6237604117647059f, 0.056516411764705884f,
+0.10711272549019607f, 0.6271437843137254f, 0.054055784313725486f,
+0.10926574509803921f, 0.6305271568627451f, 0.0515951568627451f,
+0.11580170588235295f, 0.6341411176470588f, 0.05282547058823529f,
+0.12379864705882353f, 0.6378319411764706f, 0.055286098039215685f,
+0.1317955882352941f, 0.6415227647058823f, 0.05774672549019608f,
+0.1397925294117647f, 0.6452135882352941f, 0.06020735294117647f,
+0.1477894705882353f, 0.6489044117647058f, 0.06266798039215686f,
+0.1557864117647059f, 0.6525952352941177f, 0.06512860784313726f,
+0.16378335294117646f, 0.6562860588235294f, 0.06758923529411764f,
+0.17178029411764684f, 0.659976882352941f, 0.07004986274509796f,
+0.17977723529411765f, 0.6636677058823529f, 0.07251049019607843f,
+0.18777417647058825f, 0.6673585294117647f, 0.07497111764705881f,
+0.19577111764705885f, 0.6710493529411765f, 0.07743174509803921f,
+0.2037680588235294f, 0.6747401764705883f, 0.0798923725490196f,
+0.211765f, 0.678431f, 0.082353f,
+0.2200694705882353f, 0.6818143725490197f, 0.08481354901960785f,
+0.22837394117647059f, 0.6851977450980392f, 0.08727409803921568f,
+0.2366784117647059f, 0.6885811176470589f, 0.08973464705882353f,
+0.2449828823529412f, 0.6919644901960784f, 0.09219519607843137f,
+0.2532873529411765f, 0.6953478627450981f, 0.09465574509803922f,
+0.2615918235294118f, 0.6987312352941176f, 0.09711629411764706f,
+0.2698962941176471f, 0.7021146078431373f, 0.0995768431372549f,
+0.2782007647058824f, 0.7054979803921568f, 0.10203739215686275f,
+0.28650523529411764f, 0.7088813529411765f, 0.1044979411764706f,
+0.29480970588235295f, 0.7122647254901961f, 0.10695849019607843f,
+0.303114176470588f, 0.7156480980392156f, 0.10941903921568621f,
+0.31141864705882355f, 0.7190314705882354f, 0.11187958823529412f,
+0.31987692156862746f, 0.7223379215686274f, 0.11464772549019608f,
+0.32879660784313725f, 0.7254136078431372f, 0.1183386274509804f,
+0.3377162941176471f, 0.7284892941176471f, 0.12202952941176472f,
+0.3466359803921569f, 0.7315649803921569f, 0.125720431372549f,
+0.35555566666666666f, 0.7346406666666667f, 0.12941133333333335f,
+0.3644753529411765f, 0.7377163529411764f, 0.13310223529411766f,
+0.3733950392156863f, 0.7407920392156863f, 0.13679313725490197f,
+0.3823147254901961f, 0.7438677254901961f, 0.14048403921568628f,
+0.3912344117647059f, 0.7469434117647059f, 0.1441749411764706f,
+0.4001540980392157f, 0.7500190980392157f, 0.14786584313725493f,
+0.4090737843137255f, 0.7530947843137255f, 0.15155674509803924f,
+0.4179934705882353f, 0.7561704705882353f, 0.15524764705882355f,
+0.4269131568627451f, 0.7592461568627451f, 0.15893854901960786f,
+0.43629417647058827f, 0.7623218823529412f, 0.16385976470588237f,
+0.4461365294117644f, 0.7653976470588235f, 0.1700112941176469f,
+0.4559788823529412f, 0.7684734117647058f, 0.17616282352941176f,
+0.4658212352941177f, 0.7715491764705883f, 0.18231435294117648f,
+0.47566358823529414f, 0.7746249411764706f, 0.18846588235294118f,
+0.4855059411764706f, 0.777700705882353f, 0.1946174117647059f,
+0.495348294117647f, 0.7807764705882353f, 0.2007689411764706f,
+0.5051906470588235f, 0.7838522352941176f, 0.20692047058823532f,
+0.515033f, 0.7869280000000001f, 0.213072f,
+0.5248753529411765f, 0.7900037647058824f, 0.2192235294117647f,
+0.5347177058823529f, 0.7930795294117647f, 0.22537505882352943f,
+0.5445600588235294f, 0.7961552941176471f, 0.23152658823529415f,
+0.5544024117647058f, 0.7992310588235294f, 0.23767811764705882f,
+0.5633220588235294f, 0.8023068235294118f, 0.24244552941176473f,
+0.5719341372549019f, 0.8053825882352942f, 0.24675156862745098f,
+0.5805462156862745f, 0.8084583529411765f, 0.25105760784313724f,
+0.5891582941176471f, 0.8115341176470589f, 0.25536364705882353f,
+0.5977703725490193f, 0.8146098823529411f, 0.2596696862745097f,
+0.6063824509803921f, 0.8176856470588235f, 0.2639757254901961f,
+0.6149945294117647f, 0.8207614117647059f, 0.26828176470588233f,
+0.6236066078431373f, 0.8238371764705882f, 0.2725878039215686f,
+0.6322186862745098f, 0.8269129411764706f, 0.2768938431372549f,
+0.6408307647058824f, 0.829988705882353f, 0.2811998823529412f,
+0.6494428431372549f, 0.8330644705882353f, 0.2855059215686274f,
+0.6580549215686274f, 0.8361402352941176f, 0.2898119607843137f,
+0.666667f, 0.839216f, 0.294118f,
+0.6758942156862745f, 0.8422916862745098f, 0.302114862745098f,
+0.685121431372549f, 0.8453673725490196f, 0.31011172549019606f,
+0.6943486470588235f, 0.8484430588235293f, 0.3181085882352941f,
+0.703575862745098f, 0.8515187450980392f, 0.3261054509803921f,
+0.7128030784313726f, 0.854594431372549f, 0.3341023137254902f,
+0.7220302941176471f, 0.8576701176470588f, 0.34209917647058824f,
+0.7312575098039216f, 0.8607458039215686f, 0.35009603921568627f,
+0.7404847254901958f, 0.8638214901960783f, 0.3580929019607841f,
+0.7497119411764706f, 0.8668971764705882f, 0.36608976470588234f,
+0.758939156862745f, 0.869972862745098f, 0.37408662745098037f,
+0.7681663725490195f, 0.8730485490196078f, 0.3820834901960784f,
+0.777393588235294f, 0.8761242352941176f, 0.3900803529411765f,
+0.7863132352941176f, 0.8792768431372548f, 0.39876927450980393f,
+0.7943101764705882f, 0.8826602156862745f, 0.4095343725490196f,
+0.8023071176470588f, 0.886043588235294f, 0.42029947058823525f,
+0.8103040588235294f, 0.8894269607843137f, 0.43106456862745096f,
+0.818301f, 0.8928103333333333f, 0.4418296666666666f,
+0.8262979411764706f, 0.8961937058823529f, 0.45259476470588234f,
+0.8342948823529411f, 0.8995770784313725f, 0.463359862745098f,
+0.8422918235294118f, 0.9029604509803921f, 0.47412496078431365f,
+0.8502887647058823f, 0.9063438235294118f, 0.48489005882352937f,
+0.858285705882353f, 0.9097271960784313f, 0.4956551568627451f,
+0.8662826470588235f, 0.913110568627451f, 0.5064202549019607f,
+0.874279588235294f, 0.9164939411764704f, 0.5171853529411761f,
+0.8822765294117647f, 0.9198773137254902f, 0.5279504509803921f,
+0.8891969215686275f, 0.9226455098039216f, 0.5387155490196078f,
+0.8950407647058823f, 0.9247985294117647f, 0.5494806470588235f,
+0.9008846078431373f, 0.9269515490196079f, 0.5602457450980391f,
+0.9067284509803921f, 0.9291045686274509f, 0.5710108431372548f,
+0.9125722941176471f, 0.9312575882352941f, 0.5817759411764706f,
+0.918416137254902f, 0.9334106078431372f, 0.5925410392156862f,
+0.9242599803921568f, 0.9355636274509803f, 0.6033061372549019f,
+0.9301038235294118f, 0.9377166470588235f, 0.6140712352941176f,
+0.9359476666666666f, 0.9398696666666666f, 0.6248363333333333f,
+0.9417915098039216f, 0.9420226862745098f, 0.635601431372549f,
+0.9476353529411764f, 0.9441757058823529f, 0.6463665294117646f,
+0.9534791960784313f, 0.9463287254901961f, 0.6571316274509804f,
+0.9593230392156863f, 0.9484817450980392f, 0.6678967254901961f,
+0.9630908235294118f, 0.9510961176470588f, 0.6841981764705882f,
+0.966166588235294f, 0.9538642745098038f, 0.702345078431372f,
+0.9692423529411764f, 0.956632431372549f, 0.7204919803921568f,
+0.9723181176470588f, 0.9594005882352941f, 0.7386388823529412f,
+0.9753938823529411f, 0.9621687450980392f, 0.7567857843137255f,
+0.9784696470588236f, 0.9649369019607843f, 0.7749326862745098f,
+0.9815454117647059f, 0.9677050588235294f, 0.7930795882352941f,
+0.9846211764705882f, 0.9704732156862745f, 0.8112264901960784f,
+0.9876969411764706f, 0.9732413725490197f, 0.8293733921568627f,
+0.9907727058823529f, 0.9760095294117647f, 0.8475202941176471f,
+0.9938484705882353f, 0.9787776862745098f, 0.8656671960784313f,
+0.9969242352941177f, 0.9815458431372549f, 0.8838140980392157f,
+1.0f, 0.984314f, 0.901961f
+}
+, 
+{
+0.909804f, 0.980392f, 0.831373f,
+0.9051903529411764f, 0.9779314509803922f, 0.8242987254901961f,
+0.9005767058823528f, 0.9754709019607843f, 0.8172244509803922f,
+0.8959630588235293f, 0.9730103529411765f, 0.8101501764705883f,
+0.8913494117647058f, 0.9705498039215686f, 0.8030759019607843f,
+0.8867357647058823f, 0.9680892549019608f, 0.7960016274509804f,
+0.8821221176470587f, 0.965628705882353f, 0.7889273529411764f,
+0.8775084705882352f, 0.9631681568627452f, 0.7818530784313725f,
+0.8728948235294117f, 0.9607076078431372f, 0.7747788039215686f,
+0.8682811764705882f, 0.9582470588235295f, 0.7677045294117647f,
+0.8636675294117646f, 0.9557865098039215f, 0.7606302549019608f,
+0.8590538823529411f, 0.9533259607843138f, 0.7535559803921568f,
+0.8544402352941176f, 0.9508654117647058f, 0.7464817058823529f,
+0.8495959215686274f, 0.9480972745098039f, 0.739023f,
+0.8440596078431372f, 0.9444063725490196f, 0.7304109999999999f,
+0.838523294117647f, 0.9407154705882352f, 0.721799f,
+0.8329869803921568f, 0.937024568627451f, 0.7131869999999999f,
+0.8274506666666667f, 0.9333336666666666f, 0.704575f,
+0.8219143529411764f, 0.9296427647058824f, 0.695963f,
+0.8163780392156862f, 0.925951862745098f, 0.6873509999999999f,
+0.8108417254901961f, 0.9222609607843137f, 0.678739f,
+0.8053054117647058f, 0.9185700588235294f, 0.6701269999999999f,
+0.7997690980392157f, 0.9148791568627451f, 0.661515f,
+0.7942327843137255f, 0.9111882549019608f, 0.652903f,
+0.7886964705882352f, 0.9074973529411765f, 0.644291f,
+0.7831601568627451f, 0.9038064509803921f, 0.635679f,
+0.7783927647058824f, 0.9007307058823529f, 0.6282203529411764f,
+0.774394294117647f, 0.8982701176470589f, 0.6219150588235294f,
+0.7703958235294117f, 0.8958095294117647f, 0.6156097647058824f,
+0.7663973529411765f, 0.8933489411764706f, 0.6093044705882352f,
+0.7623988823529412f, 0.8908883529411765f, 0.6029991764705882f,
+0.7584004117647059f, 0.8884277647058824f, 0.5966938823529412f,
+0.7544019411764705f, 0.8859671764705882f, 0.590388588235294f,
+0.7504034705882353f, 0.8835065882352942f, 0.584083294117647f,
+0.746405f, 0.881046f, 0.577778f,
+0.7424065294117647f, 0.8785854117647058f, 0.5714727058823529f,
+0.7384080588235294f, 0.8761248235294118f, 0.5651674117647059f,
+0.7344095882352941f, 0.8736642352941176f, 0.5588621176470588f,
+0.7304111176470588f, 0.8712036470588235f, 0.5525568235294117f,
+0.7264126470588235f, 0.8687430588235294f, 0.5462515294117647f,
+0.7224141764705883f, 0.8662824705882353f, 0.5399462352941177f,
+0.7184157058823529f, 0.8638218823529412f, 0.5336409411764707f,
+0.7144172352941176f, 0.8613612941176471f, 0.5273356470588235f,
+0.7104187647058824f, 0.8589007058823529f, 0.5210303529411765f,
+0.7064202941176471f, 0.8564401176470589f, 0.5147250588235294f,
+0.7024218235294117f, 0.8539795294117647f, 0.5084197647058823f,
+0.6984233529411765f, 0.8515189411764705f, 0.5021144705882352f,
+0.6944248823529412f, 0.8490583529411765f, 0.4958091764705882f,
+0.6904264117647059f, 0.8465977647058823f, 0.4895038823529412f,
+0.6864279411764707f, 0.8441371764705882f, 0.48319858823529416f,
+0.6824294705882353f, 0.8416765882352941f, 0.476893294117647f,
+0.678431f, 0.839216f, 0.470588f,
+0.6755090784313725f, 0.8364478039215686f, 0.46658952941176474f,
+0.6725871568627452f, 0.8336796078431372f, 0.46259105882352947f,
+0.6696652352941177f, 0.8309114117647058f, 0.45859258823529414f,
+0.6667433137254902f, 0.8281432156862745f, 0.4545941176470588f,
+0.6638213921568628f, 0.8253750196078431f, 0.45059564705882355f,
+0.6608994705882353f, 0.8226068235294117f, 0.4465971764705883f,
+0.6579775490196078f, 0.8198386274509804f, 0.44259870588235295f,
+0.6550556274509803f, 0.817070431372549f, 0.4386002352941176f,
+0.652133705882353f, 0.8143022352941176f, 0.43460176470588235f,
+0.6492117843137255f, 0.8115340392156862f, 0.4306032941176471f,
+0.646289862745098f, 0.8087658431372549f, 0.42660482352941176f,
+0.6433679411764706f, 0.8059976470588235f, 0.4226063529411765f,
+0.6404460196078431f, 0.8032294509803921f, 0.41860788235294116f,
+0.6375240980392156f, 0.8004612549019607f, 0.4146094117647059f,
+0.6346021764705883f, 0.7976930588235294f, 0.4106109411764706f,
+0.6316802549019608f, 0.794924862745098f, 0.4066124705882353f,
+0.6287583333333333f, 0.7921566666666666f, 0.402614f,
+0.6258364117647058f, 0.7893884705882352f, 0.3986155294117647f,
+0.6229144901960784f, 0.7866202745098039f, 0.3946170588235294f,
+0.619992568627451f, 0.7838520784313725f, 0.3906185882352941f,
+0.6170706470588235f, 0.7810838823529411f, 0.38662011764705884f,
+0.6141487254901961f, 0.7783156862745098f, 0.3826216470588235f,
+0.6112268039215686f, 0.7755474901960784f, 0.3786231764705883f,
+0.6083048823529411f, 0.772779294117647f, 0.3746247058823529f,
+0.6053829607843138f, 0.7700110980392156f, 0.37062623529411765f,
+0.6028454901960784f, 0.767704294117647f, 0.36701225490196077f,
+0.6006924705882353f, 0.7658588823529411f, 0.36378276470588233f,
+0.5985394509803921f, 0.7640134705882352f, 0.3605532745098039f,
+0.596386431372549f, 0.7621680588235293f, 0.35732378431372547f,
+0.5942334117647058f, 0.7603226470588235f, 0.35409429411764703f,
+0.5920803921568627f, 0.7584772352941176f, 0.35086480392156866f,
+0.5899273725490196f, 0.7566318235294117f, 0.34763531372549017f,
+0.5877743529411764f, 0.7547864117647058f, 0.34440582352941174f,
+0.5856213333333333f, 0.752941f, 0.3411763333333333f,
+0.5834683137254901f, 0.7510955882352941f, 0.33794684313725487f,
+0.581315294117647f, 0.7492501764705882f, 0.33471735294117644f,
+0.5791622745098038f, 0.7474047647058824f, 0.331487862745098f,
+0.5770092549019608f, 0.7455593529411765f, 0.3282583725490196f,
+0.5748562352941177f, 0.7437139411764706f, 0.3250288823529412f,
+0.5727032156862745f, 0.7418685294117647f, 0.32179939215686276f,
+0.5705501960784314f, 0.7400231176470589f, 0.31856990196078433f,
+0.5683971764705882f, 0.7381777058823529f, 0.3153404117647059f,
+0.5662441568627451f, 0.736332294117647f, 0.31211092156862746f,
+0.5640911372549019f, 0.7344868823529411f, 0.30888143137254903f,
+0.5619381176470588f, 0.7326414705882353f, 0.3056519411764706f,
+0.5597850980392156f, 0.7307960588235294f, 0.30242245098039217f,
+0.5576320784313725f, 0.7289506470588236f, 0.2991929607843138f,
+0.5554790588235293f, 0.7271052352941176f, 0.2959634705882353f,
+0.5533260392156862f, 0.7252598235294118f, 0.29273398039215687f,
+0.551173019607843f, 0.7234144117647059f, 0.28950449019607843f,
+0.54902f, 0.721569f, 0.286275f,
+0.5473283137254902f, 0.7195697450980393f, 0.2845833137254902f,
+0.5456366274509804f, 0.7175704901960784f, 0.2828916274509804f,
+0.5439449411764705f, 0.7155712352941177f, 0.28119994117647057f,
+0.5422532549019607f, 0.7135719803921569f, 0.2795082549019608f,
+0.540561568627451f, 0.7115727254901961f, 0.27781656862745097f,
+0.5388698823529412f, 0.7095734705882353f, 0.2761248823529412f,
+0.5371781960784313f, 0.7075742156862745f, 0.27443319607843136f,
+0.5354865098039215f, 0.7055749607843137f, 0.2727415098039216f,
+0.5337948235294118f, 0.703575705882353f, 0.27104982352941176f,
+0.532103137254902f, 0.7015764509803921f, 0.269358137254902f,
+0.5304114509803921f, 0.6995771960784314f, 0.26766645098039216f,
+0.5287197647058823f, 0.6975779411764707f, 0.2659747647058824f,
+0.5270280784313726f, 0.6955786862745098f, 0.26428307843137255f,
+0.5253363921568628f, 0.693579431372549f, 0.2625913921568627f,
+0.5236447058823529f, 0.6915801764705882f, 0.26089970588235295f,
+0.5219530196078431f, 0.6895809215686275f, 0.2592080196078431f,
+0.5202613333333334f, 0.6875816666666666f, 0.25751633333333335f,
+0.5185696470588236f, 0.6855824117647058f, 0.2558246470588235f,
+0.5168779607843137f, 0.6835831568627451f, 0.2541329607843137f,
+0.5151862745098039f, 0.6815839019607843f, 0.2524412745098039f,
+0.5134945882352941f, 0.6795846470588235f, 0.25074958823529414f,
+0.5118029019607844f, 0.6775853921568628f, 0.2490579019607843f,
+0.5101112156862746f, 0.6755861372549019f, 0.2473662156862745f,
+0.5084195294117647f, 0.6735868823529412f, 0.2456745294117647f,
+0.5067278431372549f, 0.6715876274509803f, 0.24398284313725488f,
+0.5048054901960785f, 0.6692039215686274f, 0.24213739215686272f,
+0.5026524705882354f, 0.6664357647058823f, 0.24013817647058822f,
+0.5004994509803922f, 0.6636676078431372f, 0.2381389607843137f,
+0.4983464313725491f, 0.6608994509803922f, 0.2361397450980392f,
+0.496193411764706f, 0.6581312941176471f, 0.23414052941176475f,
+0.4940403921568628f, 0.6553631372549019f, 0.23214131372549018f,
+0.49188737254901965f, 0.6525949803921568f, 0.23014209803921568f,
+0.4897343529411765f, 0.6498268235294117f, 0.22814288235294117f,
+0.48758133333333337f, 0.6470586666666667f, 0.22614366666666666f,
+0.4854283137254902f, 0.6442905098039216f, 0.22414445098039215f,
+0.4832752941176471f, 0.6415223529411764f, 0.22214523529411764f,
+0.48112227450980394f, 0.6387541960784313f, 0.22014601960784314f,
+0.4789692549019608f, 0.6359860392156862f, 0.21814680392156863f,
+0.47681623529411765f, 0.6332178823529412f, 0.21614758823529412f,
+0.4746632156862745f, 0.6304497254901961f, 0.2141483725490196f,
+0.47251019607843137f, 0.627681568627451f, 0.2121491568627451f,
+0.4703571764705883f, 0.6249134117647058f, 0.21014994117647057f,
+0.4682041568627451f, 0.6221452549019607f, 0.2081507254901961f,
+0.466051137254902f, 0.6193770980392157f, 0.20615150980392155f,
+0.46389811764705885f, 0.6166089411764706f, 0.20415229411764707f,
+0.46174509803921576f, 0.6138407843137256f, 0.2021530784313726f,
+0.45959207843137256f, 0.6110726274509803f, 0.20015386274509803f,
+0.4574390588235294f, 0.6083044705882352f, 0.19815464705882352f,
+0.4552860392156863f, 0.6055363137254902f, 0.196155431372549f,
+0.45313301960784313f, 0.6027681568627451f, 0.1941562156862745f,
+0.45098f, 0.6f, 0.192157f,
+0.44882698039215685f, 0.5972318431372549f, 0.19046535294117647f,
+0.4466739607843137f, 0.5944636862745097f, 0.18877370588235293f,
+0.44452094117647056f, 0.5916955294117647f, 0.1870820588235294f,
+0.4423679215686274f, 0.5889273725490196f, 0.1853904117647059f,
+0.4402149019607843f, 0.5861592156862745f, 0.18369876470588234f,
+0.4380618823529412f, 0.5833910588235294f, 0.18200711764705882f,
+0.43590886274509805f, 0.5806229019607843f, 0.1803154705882353f,
+0.4337558431372549f, 0.5778547450980392f, 0.17862382352941175f,
+0.43160282352941176f, 0.5750865882352941f, 0.17693217647058823f,
+0.4294498039215686f, 0.572318431372549f, 0.1752405294117647f,
+0.42729678431372553f, 0.569550274509804f, 0.17354888235294122f,
+0.42514376470588233f, 0.5667821176470588f, 0.17185723529411764f,
+0.4229907450980392f, 0.5640139607843138f, 0.17016558823529412f,
+0.42083772549019605f, 0.5612458039215686f, 0.16847394117647058f,
+0.41868470588235296f, 0.5584776470588235f, 0.16678229411764706f,
+0.4165316862745098f, 0.5557094901960784f, 0.16509064705882354f,
+0.4143786666666667f, 0.5529413333333333f, 0.16339900000000002f,
+0.41222564705882353f, 0.5501731764705882f, 0.16170735294117647f,
+0.4100726274509804f, 0.5474050196078432f, 0.16001570588235295f,
+0.40791960784313724f, 0.544636862745098f, 0.15832405882352943f,
+0.4057665882352941f, 0.5418687058823529f, 0.15663241176470588f,
+0.40361356862745096f, 0.5391005490196078f, 0.15494076470588236f,
+0.4014605490196078f, 0.5363323921568628f, 0.15324911764705884f,
+0.3993075294117647f, 0.5335642352941177f, 0.15155747058823532f,
+0.39715450980392153f, 0.5307960784313726f, 0.14986582352941177f,
+0.39500149019607844f, 0.5278741176470588f, 0.14825105882352943f,
+0.39284847058823535f, 0.5247983529411766f, 0.1467131764705883f,
+0.39069545098039216f, 0.5217225882352942f, 0.14517529411764707f,
+0.388542431372549f, 0.5186468235294117f, 0.1436374117647059f,
+0.38638941176470587f, 0.5155710588235294f, 0.1420995294117647f,
+0.3842363921568627f, 0.512495294117647f, 0.14056164705882354f,
+0.3820833725490196f, 0.5094195294117647f, 0.13902376470588237f,
+0.37993035294117644f, 0.5063437647058824f, 0.13748588235294118f,
+0.3777773333333333f, 0.503268f, 0.135948f,
+0.37562431372549016f, 0.5001922352941176f, 0.13441011764705882f,
+0.37347129411764707f, 0.49711647058823527f, 0.13287223529411765f,
+0.3713182745098039f, 0.49404070588235294f, 0.13133435294117649f,
+0.3691652549019608f, 0.4909649411764706f, 0.1297964705882353f,
+0.36701223529411764f, 0.4878891764705882f, 0.12825858823529412f,
+0.3648592156862745f, 0.4848134117647059f, 0.12672070588235296f,
+0.36270619607843135f, 0.48173764705882355f, 0.12518282352941176f,
+0.3605531764705882f, 0.47866188235294116f, 0.1236449411764706f,
+0.3584001568627451f, 0.4755861176470589f, 0.12210705882352946f,
+0.3562471372549019f, 0.4725103529411765f, 0.12056917647058824f,
+0.3540941176470588f, 0.4694345882352941f, 0.11903129411764707f,
+0.3519410980392157f, 0.46635882352941177f, 0.11749341176470587f,
+0.34978807843137255f, 0.4632830588235294f, 0.11595552941176471f,
+0.3476350588235294f, 0.46020729411764705f, 0.11441764705882354f,
+0.34548203921568627f, 0.4571315294117647f, 0.11287976470588235f,
+0.3433290196078431f, 0.4540557647058823f, 0.11134188235294118f,
+0.341176f, 0.45098f, 0.109804f,
+0.3390230196078431f, 0.4477504705882353f, 0.10841992156862745f,
+0.33687003921568626f, 0.44452094117647056f, 0.10703584313725491f,
+0.33471705882352937f, 0.4412914117647059f, 0.10565176470588235f,
+0.33256407843137253f, 0.4380618823529412f, 0.1042676862745098f,
+0.33041109803921564f, 0.43483235294117645f, 0.10288360784313726f,
+0.3282581176470588f, 0.43160282352941176f, 0.10149952941176471f,
+0.3261051372549019f, 0.428373294117647f, 0.10011545098039215f,
+0.32395215686274514f, 0.42514376470588244f, 0.09873137254901965f,
+0.3217991764705882f, 0.42191423529411765f, 0.09734729411764706f,
+0.31964619607843137f, 0.41868470588235296f, 0.09596321568627451f,
+0.3174932156862745f, 0.4154551764705882f, 0.09457913725490197f,
+0.31534023529411764f, 0.41222564705882353f, 0.09319505882352941f,
+0.31318725490196075f, 0.4089961176470588f, 0.09181098039215686f,
+0.3110342745098039f, 0.4057665882352941f, 0.09042690196078432f,
+0.30888129411764703f, 0.4025370588235294f, 0.08904282352941177f,
+0.3067283137254902f, 0.3993075294117647f, 0.08765874509803923f,
+0.3045753333333333f, 0.396078f, 0.08627466666666667f,
+0.3024223529411765f, 0.3928484705882353f, 0.08489058823529412f,
+0.3002693725490196f, 0.38961894117647056f, 0.08350650980392157f,
+0.29811639215686275f, 0.38638941176470587f, 0.08212243137254903f,
+0.29596341176470586f, 0.3831598823529412f, 0.08073835294117648f,
+0.29381043137254903f, 0.37993035294117644f, 0.07935427450980392f,
+0.29165745098039214f, 0.37670082352941175f, 0.07797019607843138f,
+0.28950447058823536f, 0.3734712941176471f, 0.07658611764705886f,
+0.2873514901960784f, 0.3702417647058823f, 0.07520203921568627f,
+0.2849678039215686f, 0.36685845098039216f, 0.0738948431372549f,
+0.2823534117647059f, 0.36332135294117646f, 0.07266452941176471f,
+0.27973901960784314f, 0.35978425490196075f, 0.07143421568627452f,
+0.2771246274509804f, 0.3562471568627451f, 0.07020390196078433f,
+0.27451023529411767f, 0.3527100588235294f, 0.06897358823529412f,
+0.2718958431372549f, 0.3491729607843137f, 0.06774327450980393f,
+0.26928145098039213f, 0.345635862745098f, 0.06651296078431373f,
+0.2666670588235294f, 0.34209876470588235f, 0.06528264705882353f,
+0.26405266666666666f, 0.33856166666666665f, 0.06405233333333334f,
+0.26143827450980395f, 0.33502456862745095f, 0.06282201960784314f,
+0.2588238823529412f, 0.3314874705882353f, 0.061591705882352944f,
+0.2562094901960784f, 0.3279503725490196f, 0.06036139215686275f,
+0.2535950980392157f, 0.3244132745098039f, 0.05913107843137255f,
+0.25098070588235294f, 0.3208761764705882f, 0.05790076470588236f,
+0.24836631372549028f, 0.31733907843137266f, 0.056670450980392195f,
+0.24575192156862746f, 0.31380198039215684f, 0.05544013725490196f,
+0.2431375294117647f, 0.31026488235294114f, 0.05420982352941177f,
+0.24052313725490196f, 0.3067277843137255f, 0.052979509803921576f,
+0.23790874509803922f, 0.3031906862745098f, 0.05174919607843138f,
+0.23529435294117648f, 0.2996535882352941f, 0.05051888235294118f,
+0.23267996078431372f, 0.29611649019607844f, 0.049288568627450985f,
+0.23006556862745098f, 0.29257939215686274f, 0.048058254901960785f,
+0.22745117647058824f, 0.28904229411764704f, 0.04682794117647059f,
+0.2248367843137255f, 0.2855051960784314f, 0.045597627450980394f,
+0.22222239215686274f, 0.2819680980392157f, 0.0443673137254902f,
+0.219608f, 0.278431f, 0.043137f
+}
+, 
+{
+0.964706f, 0.988235f, 0.941176f,
+0.9619378431372548f, 0.9871584901960784f, 0.9373313333333333f,
+0.9591696862745097f, 0.9860819803921569f, 0.9334866666666667f,
+0.9564015294117647f, 0.9850054705882353f, 0.929642f,
+0.9536333725490196f, 0.9839289607843137f, 0.9257973333333334f,
+0.9508652156862745f, 0.9828524509803921f, 0.9219526666666666f,
+0.9480970588235293f, 0.9817759411764706f, 0.918108f,
+0.9453289019607842f, 0.980699431372549f, 0.9142633333333333f,
+0.9425607450980392f, 0.9796229215686274f, 0.9104186666666667f,
+0.9397925882352941f, 0.9785464117647058f, 0.906574f,
+0.937024431372549f, 0.9774699019607843f, 0.9027293333333334f,
+0.9342562745098039f, 0.9763933921568627f, 0.8988846666666667f,
+0.9314881176470587f, 0.9753168823529411f, 0.8950400000000001f,
+0.9287199607843137f, 0.9742403725490196f, 0.8911953333333333f,
+0.9259518039215686f, 0.973163862745098f, 0.8873506666666667f,
+0.9231836470588235f, 0.9720873529411764f, 0.883506f,
+0.9204154901960784f, 0.9710108431372548f, 0.8796613333333334f,
+0.9176473333333333f, 0.9699343333333333f, 0.8758166666666667f,
+0.9148791764705881f, 0.9688578235294117f, 0.871972f,
+0.9121110196078431f, 0.9677813137254901f, 0.8681273333333334f,
+0.909342862745098f, 0.9667048039215687f, 0.8642826666666666f,
+0.9065747058823529f, 0.9656282941176471f, 0.860438f,
+0.9038065490196078f, 0.9645517843137255f, 0.8565933333333333f,
+0.9010383921568628f, 0.9634752745098039f, 0.8527486666666667f,
+0.8982702352941176f, 0.9623987647058824f, 0.848904f,
+0.8955020784313725f, 0.9613222549019608f, 0.8450593333333334f,
+0.8921956470588235f, 0.9600150784313726f, 0.8413684509803921f,
+0.8883509411764705f, 0.9584772352941177f, 0.8378313529411765f,
+0.8845062352941176f, 0.9569393921568627f, 0.8342942549019609f,
+0.8806615294117647f, 0.9554015490196078f, 0.8307571568627451f,
+0.8768168235294117f, 0.9538637058823529f, 0.8272200588235294f,
+0.8729721176470588f, 0.952325862745098f, 0.8236829607843137f,
+0.8691274117647059f, 0.9507880196078431f, 0.820145862745098f,
+0.8652827058823529f, 0.9492501764705882f, 0.8166087647058824f,
+0.8614379999999999f, 0.9477123333333333f, 0.8130716666666666f,
+0.8575932941176471f, 0.9461744901960784f, 0.809534568627451f,
+0.8537485882352941f, 0.9446366470588236f, 0.8059974705882353f,
+0.8499038823529411f, 0.9430988039215686f, 0.8024603725490196f,
+0.8460591764705881f, 0.9415609607843137f, 0.7989232745098039f,
+0.8422144705882353f, 0.9400231176470588f, 0.7953861764705882f,
+0.8383697647058823f, 0.9384852745098039f, 0.7918490784313725f,
+0.8345250588235293f, 0.936947431372549f, 0.7883119803921569f,
+0.8306803529411764f, 0.9354095882352941f, 0.7847748823529411f,
+0.8268356470588235f, 0.9338717450980392f, 0.7812377843137255f,
+0.8229909411764705f, 0.9323339019607843f, 0.7777006862745098f,
+0.8191462352941177f, 0.9307960588235293f, 0.7741635882352942f,
+0.8153015294117647f, 0.9292582156862745f, 0.7706264901960784f,
+0.8114568235294117f, 0.9277203725490196f, 0.7670893921568627f,
+0.8076121176470588f, 0.9261825294117647f, 0.763552294117647f,
+0.8037674117647059f, 0.9246446862745098f, 0.7600151960784314f,
+0.7999227058823529f, 0.9231068431372549f, 0.7564780980392156f,
+0.796078f, 0.921569f, 0.752941f,
+0.7942325882352941f, 0.9203386862745098f, 0.751710725490196f,
+0.7923871764705882f, 0.9191083725490196f, 0.7504804509803922f,
+0.7905417647058823f, 0.9178780588235294f, 0.7492501764705882f,
+0.7886963529411765f, 0.9166477450980391f, 0.7480199019607843f,
+0.7868509411764706f, 0.9154174313725489f, 0.7467896274509803f,
+0.7850055294117647f, 0.9141871176470588f, 0.7455593529411765f,
+0.7831601176470588f, 0.9129568039215686f, 0.7443290784313725f,
+0.781314705882353f, 0.9117264901960784f, 0.7430988039215686f,
+0.779469294117647f, 0.9104961764705882f, 0.7418685294117647f,
+0.7776238823529412f, 0.909265862745098f, 0.7406382549019608f,
+0.7757784705882352f, 0.9080355490196078f, 0.7394079803921568f,
+0.7739330588235294f, 0.9068052352941176f, 0.738177705882353f,
+0.7720876470588235f, 0.9055749215686274f, 0.736947431372549f,
+0.7702422352941176f, 0.9043446078431372f, 0.7357171568627451f,
+0.7683968235294117f, 0.9031142941176471f, 0.7344868823529412f,
+0.7665514117647059f, 0.9018839803921569f, 0.7332566078431373f,
+0.764706f, 0.9006536666666667f, 0.7320263333333333f,
+0.7628605882352941f, 0.8994233529411765f, 0.7307960588235294f,
+0.7610151764705883f, 0.8981930392156863f, 0.7295657843137255f,
+0.7591697647058824f, 0.8969627254901961f, 0.7283355098039216f,
+0.7573243529411765f, 0.8957324117647059f, 0.7271052352941176f,
+0.7554789411764706f, 0.8945020980392157f, 0.7258749607843137f,
+0.7536335294117648f, 0.8932717843137254f, 0.7246446862745098f,
+0.7517881176470589f, 0.8920414705882352f, 0.7234144117647059f,
+0.749942705882353f, 0.890811156862745f, 0.7221841372549019f,
+0.7479434901960784f, 0.8894270588235293f, 0.7207231568627451f,
+0.7457904705882353f, 0.8878891764705882f, 0.7190314705882354f,
+0.7436374509803921f, 0.886351294117647f, 0.7173397843137255f,
+0.741484431372549f, 0.8848134117647058f, 0.7156480980392157f,
+0.7393314117647058f, 0.8832755294117647f, 0.7139564117647059f,
+0.7371783921568628f, 0.8817376470588235f, 0.7122647254901961f,
+0.7350253725490197f, 0.8801997647058823f, 0.7105730392156863f,
+0.7328723529411765f, 0.8786618823529412f, 0.7088813529411765f,
+0.7307193333333334f, 0.877124f, 0.7071896666666667f,
+0.7285663137254902f, 0.8755861176470588f, 0.7054979803921568f,
+0.7264132941176471f, 0.8740482352941176f, 0.7038062941176471f,
+0.7242602745098039f, 0.8725103529411764f, 0.7021146078431373f,
+0.7221072549019608f, 0.8709724705882352f, 0.7004229215686275f,
+0.7199542352941177f, 0.8694345882352941f, 0.6987312352941177f,
+0.7178012156862745f, 0.8678967058823529f, 0.6970395490196079f,
+0.7156481960784314f, 0.8663588235294117f, 0.6953478627450981f,
+0.7134951764705882f, 0.8648209411764706f, 0.6936561764705882f,
+0.7113421568627452f, 0.8632830588235294f, 0.6919644901960784f,
+0.709189137254902f, 0.8617451764705882f, 0.6902728039215686f,
+0.7070361176470589f, 0.8602072941176471f, 0.6885811176470589f,
+0.7048830980392157f, 0.8586694117647058f, 0.686889431372549f,
+0.7027300784313726f, 0.8571315294117647f, 0.6851977450980392f,
+0.7005770588235294f, 0.8555936470588235f, 0.6835060588235294f,
+0.6984240392156863f, 0.8540557647058823f, 0.6818143725490197f,
+0.6962710196078431f, 0.8525178823529411f, 0.6801226862745098f,
+0.694118f, 0.85098f, 0.678431f,
+0.6922725490196079f, 0.849749725490196f, 0.677200725490196f,
+0.6904270980392156f, 0.8485194509803922f, 0.6759704509803922f,
+0.6885816470588235f, 0.8472891764705882f, 0.6747401764705883f,
+0.6867361960784314f, 0.8460589019607843f, 0.6735099019607843f,
+0.6848907450980393f, 0.8448286274509803f, 0.6722796274509804f,
+0.683045294117647f, 0.8435983529411765f, 0.6710493529411765f,
+0.6811998431372549f, 0.8423680784313725f, 0.6698190784313726f,
+0.6793543921568628f, 0.8411378039215686f, 0.6685888039215686f,
+0.6775089411764706f, 0.8399075294117647f, 0.6673585294117648f,
+0.6756634901960784f, 0.8386772549019608f, 0.6661282549019608f,
+0.6738180392156863f, 0.8374469803921568f, 0.6648979803921569f,
+0.6719725882352942f, 0.836216705882353f, 0.663667705882353f,
+0.670127137254902f, 0.834986431372549f, 0.6624374313725491f,
+0.6682816862745098f, 0.8337561568627451f, 0.6612071568627451f,
+0.6664362352941177f, 0.8325258823529411f, 0.6599768823529412f,
+0.6645907843137255f, 0.8312956078431373f, 0.6587466078431373f,
+0.6627453333333334f, 0.8300653333333333f, 0.6575163333333334f,
+0.6608998823529412f, 0.8288350588235294f, 0.6562860588235294f,
+0.659054431372549f, 0.8276047843137255f, 0.6550557843137256f,
+0.6572089803921569f, 0.8263745098039216f, 0.6538255098039216f,
+0.6553635294117648f, 0.8251442352941176f, 0.6525952352941177f,
+0.6535180784313726f, 0.8239139607843137f, 0.6513649607843137f,
+0.6516726274509804f, 0.8226836862745098f, 0.6501346862745099f,
+0.6498271764705883f, 0.8214534117647059f, 0.6489044117647059f,
+0.6479817254901962f, 0.8202231372549019f, 0.647674137254902f,
+0.6462900588235294f, 0.8188390588235294f, 0.6465976274509805f,
+0.6447521764705882f, 0.8173011764705882f, 0.6456748823529412f,
+0.6432142941176471f, 0.815763294117647f, 0.644752137254902f,
+0.6416764117647059f, 0.8142254117647059f, 0.6438293921568627f,
+0.6401385294117647f, 0.8126875294117647f, 0.6429066470588236f,
+0.6386006470588236f, 0.8111496470588235f, 0.6419839019607844f,
+0.6370627647058824f, 0.8096117647058824f, 0.6410611568627451f,
+0.6355248823529412f, 0.8080738823529412f, 0.6401384117647059f,
+0.6339870000000001f, 0.806536f, 0.6392156666666667f,
+0.6324491176470589f, 0.8049981176470589f, 0.6382929215686275f,
+0.6309112352941176f, 0.8034602352941176f, 0.6373701764705882f,
+0.6293733529411765f, 0.8019223529411764f, 0.636447431372549f,
+0.6278354705882353f, 0.8003844705882353f, 0.6355246862745099f,
+0.6262975882352941f, 0.7988465882352941f, 0.6346019411764706f,
+0.624759705882353f, 0.7973087058823529f, 0.6336791960784314f,
+0.6232218235294118f, 0.7957708235294118f, 0.6327564509803921f,
+0.6216839411764706f, 0.7942329411764706f, 0.631833705882353f,
+0.6201460588235295f, 0.7926950588235294f, 0.6309109607843137f,
+0.6186081764705883f, 0.7911571764705883f, 0.6299882156862745f,
+0.6170702941176471f, 0.7896192941176471f, 0.6290654705882353f,
+0.615532411764706f, 0.7880814117647059f, 0.6281427254901961f,
+0.6139945294117647f, 0.7865435294117646f, 0.6272199803921569f,
+0.6124566470588235f, 0.7850056470588235f, 0.6262972352941176f,
+0.6109187647058824f, 0.7834677647058823f, 0.6253744901960785f,
+0.6093808823529412f, 0.7819298823529411f, 0.6244517450980392f,
+0.607843f, 0.780392f, 0.623529f,
+0.6059975490196079f, 0.7788541176470588f, 0.622298725490196f,
+0.6041520980392157f, 0.7773162352941176f, 0.6210684509803922f,
+0.6023066470588235f, 0.7757783529411765f, 0.6198381764705883f,
+0.6004611960784314f, 0.7742404705882353f, 0.6186079019607843f,
+0.5986157450980393f, 0.7727025882352941f, 0.6173776274509803f,
+0.596770294117647f, 0.7711647058823529f, 0.6161473529411765f,
+0.5949248431372549f, 0.7696268235294117f, 0.6149170784313726f,
+0.5930793921568628f, 0.7680889411764705f, 0.6136868039215686f,
+0.5912339411764705f, 0.7665510588235294f, 0.6124565294117648f,
+0.5893884901960784f, 0.7650131764705882f, 0.6112262549019608f,
+0.5875430392156863f, 0.763475294117647f, 0.609995980392157f,
+0.585697588235294f, 0.7619374117647059f, 0.608765705882353f,
+0.5838521372549019f, 0.7603995294117647f, 0.6075354313725491f,
+0.5820066862745098f, 0.7588616470588235f, 0.6063051568627451f,
+0.5801612352941177f, 0.7573237647058824f, 0.6050748823529412f,
+0.5783157843137254f, 0.7557858823529411f, 0.6038446078431373f,
+0.5764703333333333f, 0.7542479999999999f, 0.6026143333333334f,
+0.5746248823529412f, 0.7527101176470588f, 0.6013840588235294f,
+0.5727794313725489f, 0.7511722352941176f, 0.6001537843137256f,
+0.5709339803921568f, 0.7496343529411764f, 0.5989235098039216f,
+0.5690885294117647f, 0.7480964705882353f, 0.5976932352941177f,
+0.5672430784313724f, 0.7465585882352941f, 0.5964629607843137f,
+0.5653976274509803f, 0.7450207058823529f, 0.5952326862745099f,
+0.5635521764705882f, 0.7434828235294118f, 0.5940024117647059f,
+0.5617067254901961f, 0.7419449411764706f, 0.592772137254902f,
+0.5599381764705882f, 0.7404070784313725f, 0.5916187450980392f,
+0.5582465294117647f, 0.7388692352941176f, 0.5905422352941178f,
+0.5565548823529411f, 0.7373313921568627f, 0.5894657254901962f,
+0.5548632352941176f, 0.7357935490196078f, 0.5883892156862746f,
+0.5531715882352941f, 0.7342557058823529f, 0.587312705882353f,
+0.5514799411764706f, 0.732717862745098f, 0.5862361960784315f,
+0.5497882941176471f, 0.7311800196078431f, 0.5851596862745099f,
+0.5480966470588234f, 0.7296421764705882f, 0.5840831764705883f,
+0.5464049999999999f, 0.7281043333333332f, 0.5830066666666667f,
+0.5447133529411764f, 0.7265664901960783f, 0.5819301568627452f,
+0.5430217058823529f, 0.7250286470588235f, 0.5808536470588236f,
+0.5413300588235294f, 0.7234908039215686f, 0.579777137254902f,
+0.5396384117647058f, 0.7219529607843137f, 0.5787006274509804f,
+0.5379467647058823f, 0.7204151176470588f, 0.5776241176470589f,
+0.5362551176470588f, 0.7188772745098039f, 0.5765476078431373f,
+0.5345634705882353f, 0.717339431372549f, 0.5754710980392157f,
+0.5328718235294118f, 0.715801588235294f, 0.5743945882352942f,
+0.5311801764705882f, 0.7142637450980392f, 0.5733180784313726f,
+0.5294885294117647f, 0.7127259019607842f, 0.572241568627451f,
+0.5277968823529411f, 0.7111880588235293f, 0.5711650588235294f,
+0.5261052352941176f, 0.7096502156862745f, 0.5700885490196079f,
+0.524413588235294f, 0.7081123725490196f, 0.5690120392156863f,
+0.5227219411764705f, 0.7065745294117647f, 0.5679355294117647f,
+0.521030294117647f, 0.7050366862745098f, 0.5668590196078431f,
+0.5193386470588235f, 0.7034988431372549f, 0.5657825098039216f,
+0.517647f, 0.701961f, 0.564706f,
+0.5156477843137255f, 0.7002693529411764f, 0.5634756862745098f,
+0.5136485686274509f, 0.6985777058823529f, 0.5622453725490196f,
+0.5116493529411764f, 0.6968860588235294f, 0.5610150588235294f,
+0.5096501372549019f, 0.6951944117647059f, 0.5597847450980392f,
+0.5076509215686275f, 0.6935027647058823f, 0.558554431372549f,
+0.5056517058823529f, 0.6918111176470588f, 0.5573241176470588f,
+0.5036524901960784f, 0.6901194705882352f, 0.5560938039215686f,
+0.5016532745098039f, 0.6884278235294118f, 0.5548634901960785f,
+0.49965405882352937f, 0.6867361764705882f, 0.5536331764705883f,
+0.4976548431372549f, 0.6850445294117646f, 0.5524028627450981f,
+0.49565562745098035f, 0.6833528823529411f, 0.5511725490196079f,
+0.49365641176470587f, 0.6816612352941176f, 0.5499422352941177f,
+0.49165719607843134f, 0.6799695882352941f, 0.5487119215686275f,
+0.48965798039215686f, 0.6782779411764706f, 0.5474816078431373f,
+0.4876587647058823f, 0.676586294117647f, 0.5462512941176471f,
+0.48565954901960784f, 0.6748946470588235f, 0.5450209803921569f,
+0.4836603333333333f, 0.673203f, 0.5437906666666666f,
+0.4816611176470588f, 0.6715113529411765f, 0.5425603529411764f,
+0.4796619019607843f, 0.669819705882353f, 0.5413300392156862f,
+0.4776626862745098f, 0.6681280588235293f, 0.540099725490196f,
+0.47566347058823527f, 0.6664364117647058f, 0.5388694117647058f,
+0.4736642549019608f, 0.6647447647058823f, 0.5376390980392156f,
+0.47166503921568625f, 0.6630531176470588f, 0.5364087843137254f,
+0.46966582352941183f, 0.6613614705882354f, 0.5351784705882353f,
+0.46766660784313724f, 0.6596698235294117f, 0.533948156862745f,
+0.4655135882352941f, 0.6574399019607843f, 0.5323333921568627f,
+0.46320676470588235f, 0.6546717058823529f, 0.5303341764705882f,
+0.4608999411764706f, 0.6519035098039215f, 0.5283349607843136f,
+0.45859311764705885f, 0.6491353137254902f, 0.5263357450980392f,
+0.45628629411764704f, 0.6463671176470588f, 0.5243365294117647f,
+0.4539794705882353f, 0.6435989215686274f, 0.5223373137254902f,
+0.45167264705882354f, 0.640830725490196f, 0.5203380980392156f,
+0.4493658235294118f, 0.6380625294117647f, 0.5183388823529411f,
+0.447059f, 0.6352943333333333f, 0.5163396666666666f,
+0.44475217647058823f, 0.6325261372549019f, 0.5143404509803922f,
+0.4424453529411765f, 0.6297579411764705f, 0.5123412352941176f,
+0.44013852941176473f, 0.6269897450980392f, 0.5103420196078431f,
+0.43783170588235293f, 0.6242215490196078f, 0.5083428039215686f,
+0.4355248823529412f, 0.6214533529411764f, 0.5063435882352941f,
+0.4332180588235295f, 0.6186851568627452f, 0.5043443725490196f,
+0.4309112352941177f, 0.6159169607843137f, 0.5023451568627451f,
+0.4286044117647059f, 0.6131487647058823f, 0.5003459411764706f,
+0.4262975882352941f, 0.6103805686274509f, 0.49834672549019604f,
+0.4239907647058824f, 0.6076123725490196f, 0.49634750980392156f,
+0.4216839411764706f, 0.6048441764705882f, 0.494348294117647f,
+0.4193771176470588f, 0.6020759803921568f, 0.49234907843137254f,
+0.41707029411764707f, 0.5993077843137254f, 0.490349862745098f,
+0.4147634705882353f, 0.5965395882352941f, 0.4883506470588235f,
+0.41245664705882357f, 0.5937713921568627f, 0.486351431372549f,
+0.41014982352941176f, 0.5910031960784313f, 0.4843522156862745f,
+0.407843f, 0.588235f, 0.482353f
+}
+, 
+{
+0.976471f, 1.0f, 0.94902f,
+0.9740103725490196f, 0.9984621176470588f, 0.9459442352941176f,
+0.9715497450980392f, 0.9969242352941177f, 0.9428684705882353f,
+0.9690891176470589f, 0.9953863529411765f, 0.9397927058823529f,
+0.9666284901960784f, 0.9938484705882353f, 0.9367169411764705f,
+0.964167862745098f, 0.9923105882352942f, 0.9336411764705882f,
+0.9617072352941176f, 0.990772705882353f, 0.9305654117647059f,
+0.9592466078431372f, 0.9892348235294118f, 0.9274896470588235f,
+0.9567859803921569f, 0.9876969411764707f, 0.9244138823529411f,
+0.9543253529411765f, 0.9861590588235294f, 0.9213381176470588f,
+0.9518647254901961f, 0.9846211764705882f, 0.9182623529411764f,
+0.9494040980392157f, 0.983083294117647f, 0.9151865882352941f,
+0.9469434705882352f, 0.9815454117647059f, 0.9121108235294118f,
+0.9443290588235294f, 0.9797768627450981f, 0.9091119607843137f,
+0.941253294117647f, 0.9773163137254902f, 0.9063438039215685f,
+0.9381775294117647f, 0.9748557647058824f, 0.9035756470588235f,
+0.9351017647058824f, 0.9723952156862745f, 0.9008074901960784f,
+0.932026f, 0.9699346666666667f, 0.8980393333333333f,
+0.9289502352941176f, 0.9674741176470588f, 0.8952711764705882f,
+0.9258744705882352f, 0.965013568627451f, 0.8925030196078431f,
+0.9227987058823529f, 0.9625530196078431f, 0.889734862745098f,
+0.9197229411764706f, 0.9600924705882353f, 0.8869667058823529f,
+0.9166471764705882f, 0.9576319215686274f, 0.8841985490196078f,
+0.9135714117647058f, 0.9551713725490196f, 0.8814303921568627f,
+0.9104956470588235f, 0.9527108235294117f, 0.8786622352941177f,
+0.9074198823529411f, 0.9502502745098039f, 0.8758940784313726f,
+0.904344156862745f, 0.9479434901960784f, 0.8729721176470588f,
+0.9012684705882352f, 0.9457904705882353f, 0.8698963529411765f,
+0.8981927843137254f, 0.9436374509803921f, 0.8668205882352942f,
+0.8951170980392157f, 0.941484431372549f, 0.8637448235294117f,
+0.8920414117647059f, 0.9393314117647058f, 0.8606690588235294f,
+0.888965725490196f, 0.9371783921568627f, 0.8575932941176471f,
+0.8858900392156862f, 0.9350253725490196f, 0.8545175294117647f,
+0.8828143529411765f, 0.9328723529411764f, 0.8514417647058824f,
+0.8797386666666667f, 0.9307193333333333f, 0.848366f,
+0.8766629803921568f, 0.9285663137254901f, 0.8452902352941176f,
+0.873587294117647f, 0.9264132941176471f, 0.8422144705882353f,
+0.8705116078431372f, 0.9242602745098039f, 0.839138705882353f,
+0.8674359215686274f, 0.9221072549019608f, 0.8360629411764706f,
+0.8641294705882353f, 0.9197235294117647f, 0.8332178823529411f,
+0.8607460980392156f, 0.9172629019607843f, 0.830449725490196f,
+0.8573627254901961f, 0.9148022745098039f, 0.827681568627451f,
+0.8539793529411764f, 0.9123416470588235f, 0.8249134117647059f,
+0.8505959803921568f, 0.9098810196078431f, 0.8221452549019608f,
+0.8472126078431372f, 0.9074203921568628f, 0.8193770980392157f,
+0.8438292352941177f, 0.9049597647058824f, 0.8166089411764706f,
+0.840445862745098f, 0.9024991372549019f, 0.8138407843137255f,
+0.8370624901960784f, 0.9000385098039215f, 0.8110726274509804f,
+0.8336791176470588f, 0.8975778823529411f, 0.8083044705882353f,
+0.8302957450980392f, 0.8951172549019608f, 0.8055363137254903f,
+0.8269123725490196f, 0.8926566274509804f, 0.8027681568627452f,
+0.823529f, 0.890196f, 0.8f,
+0.819530568627451f, 0.8871202352941177f, 0.796616705882353f,
+0.8155321372549019f, 0.8840444705882353f, 0.793233411764706f,
+0.811533705882353f, 0.8809687058823529f, 0.7898501176470588f,
+0.8075352745098039f, 0.8778929411764705f, 0.7864668235294118f,
+0.8035368431372549f, 0.8748171764705882f, 0.7830835294117647f,
+0.7995384117647059f, 0.8717414117647059f, 0.7797002352941177f,
+0.7955399803921569f, 0.8686656470588235f, 0.7763169411764705f,
+0.7915415490196078f, 0.8655898823529411f, 0.7729336470588235f,
+0.7875431176470589f, 0.8625141176470588f, 0.7695503529411765f,
+0.7835446862745099f, 0.8594383529411764f, 0.7661670588235294f,
+0.7795462549019608f, 0.8563625882352941f, 0.7627837647058823f,
+0.7755478235294118f, 0.8532868235294118f, 0.7594004705882352f,
+0.7718569607843138f, 0.850364862745098f, 0.7563247450980392f,
+0.7690888039215686f, 0.8479043137254901f, 0.7541717254901961f,
+0.7663206470588236f, 0.8454437647058823f, 0.752018705882353f,
+0.7635524901960785f, 0.8429832156862744f, 0.7498656862745098f,
+0.7607843333333334f, 0.8405226666666666f, 0.7477126666666666f,
+0.7580161764705883f, 0.8380621176470588f, 0.7455596470588235f,
+0.7552480196078432f, 0.8356015686274509f, 0.7434066274509803f,
+0.752479862745098f, 0.8331410196078431f, 0.7412536078431372f,
+0.7497117058823529f, 0.8306804705882352f, 0.7391005882352941f,
+0.7469435490196078f, 0.8282199215686274f, 0.7369475686274509f,
+0.7441753921568628f, 0.8257593725490197f, 0.7347945490196078f,
+0.7414072352941177f, 0.8232988235294117f, 0.7326415294117646f,
+0.7386390784313726f, 0.820838274509804f, 0.7304885098039215f,
+0.7352557843137255f, 0.8180701176470588f, 0.7277203529411764f,
+0.7312573529411764f, 0.8149943529411765f, 0.7243370588235294f,
+0.7272589215686275f, 0.8119185882352942f, 0.7209537647058823f,
+0.7232604901960784f, 0.8088428235294117f, 0.7175704705882352f,
+0.7192620588235294f, 0.8057670588235294f, 0.7141871764705882f,
+0.7152636274509804f, 0.8026912941176471f, 0.7108038823529412f,
+0.7112651960784313f, 0.7996155294117647f, 0.7074205882352941f,
+0.7072667647058823f, 0.7965397647058824f, 0.704037294117647f,
+0.7032683333333333f, 0.793464f, 0.700654f,
+0.6992699019607843f, 0.7903882352941176f, 0.6972707058823528f,
+0.6952714705882352f, 0.7873124705882353f, 0.6938874117647058f,
+0.6912730392156863f, 0.784236705882353f, 0.6905041176470588f,
+0.6872746078431372f, 0.7811609411764706f, 0.6871208235294117f,
+0.6841988235294117f, 0.7780851764705883f, 0.6846602352941177f,
+0.6814305882352941f, 0.7750094117647058f, 0.6825072156862745f,
+0.6786623529411765f, 0.7719336470588235f, 0.6803541960784314f,
+0.6758941176470588f, 0.7688578823529412f, 0.6782011764705882f,
+0.6731258823529411f, 0.7657821176470588f, 0.6760481568627451f,
+0.6703576470588235f, 0.7627063529411764f, 0.6738951372549019f,
+0.6675894117647059f, 0.7596305882352941f, 0.6717421176470588f,
+0.6648211764705882f, 0.7565548235294117f, 0.6695890980392156f,
+0.6620529411764706f, 0.7534790588235294f, 0.6674360784313725f,
+0.659284705882353f, 0.7504032941176471f, 0.6652830588235293f,
+0.6565164705882353f, 0.7473275294117646f, 0.6631300392156863f,
+0.6537482352941176f, 0.7442517647058823f, 0.660977019607843f,
+0.65098f, 0.741176f, 0.658824f,
+0.647596705882353f, 0.7381003137254901f, 0.6560557647058823f,
+0.6442134117647059f, 0.7350246274509803f, 0.6532875294117647f,
+0.6408301176470589f, 0.7319489411764706f, 0.6505192941176471f,
+0.6374468235294118f, 0.7288732549019608f, 0.6477510588235295f,
+0.6340635294117647f, 0.7257975686274509f, 0.6449828235294117f,
+0.6306802352941177f, 0.7227218823529411f, 0.6422145882352941f,
+0.6272969411764706f, 0.7196461960784313f, 0.6394463529411765f,
+0.6239136470588236f, 0.7165705098039216f, 0.6366781176470588f,
+0.6205303529411765f, 0.7134948235294117f, 0.6339098823529412f,
+0.6171470588235294f, 0.7104191372549019f, 0.6311416470588235f,
+0.6137637647058823f, 0.7073434509803921f, 0.6283734117647058f,
+0.6103804705882354f, 0.7042677647058824f, 0.6256051764705883f,
+0.6071509607843137f, 0.7011151764705882f, 0.6229907450980392f,
+0.6043828039215686f, 0.6977318823529411f, 0.6208377254901961f,
+0.6016146470588235f, 0.6943485882352941f, 0.6186847058823529f,
+0.5988464901960785f, 0.690965294117647f, 0.6165316862745098f,
+0.5960783333333334f, 0.6875819999999999f, 0.6143786666666666f,
+0.5933101764705883f, 0.6841987058823529f, 0.6122256470588235f,
+0.5905420196078431f, 0.6808154117647058f, 0.6100726274509803f,
+0.5877738627450981f, 0.6774321176470588f, 0.6079196078431373f,
+0.5850057058823529f, 0.6740488235294118f, 0.6057665882352942f,
+0.5822375490196078f, 0.6706655294117647f, 0.603613568627451f,
+0.5794693921568628f, 0.6672822352941176f, 0.6014605490196079f,
+0.5767012352941177f, 0.6638989411764705f, 0.5993075294117647f,
+0.5739330784313725f, 0.6605156470588235f, 0.5971545098039216f,
+0.5707035490196078f, 0.6569785490196078f, 0.5943863529411765f,
+0.5670126470588235f, 0.6532876470588235f, 0.5910030588235294f,
+0.5633217450980392f, 0.6495967450980392f, 0.5876197647058824f,
+0.5596308431372549f, 0.6459058431372549f, 0.5842364705882352f,
+0.5559399411764707f, 0.6422149411764707f, 0.5808531764705883f,
+0.5522490392156862f, 0.6385240392156862f, 0.5774698823529412f,
+0.548558137254902f, 0.634833137254902f, 0.5740865882352941f,
+0.5448672352941176f, 0.6311422352941176f, 0.5707032941176471f,
+0.5411763333333334f, 0.6274513333333334f, 0.56732f,
+0.537485431372549f, 0.623760431372549f, 0.563936705882353f,
+0.5337945294117648f, 0.6200695294117647f, 0.5605534117647059f,
+0.5301036274509804f, 0.6163786274509804f, 0.5571701176470588f,
+0.5264127254901961f, 0.6126877254901961f, 0.5537868235294118f,
+0.5238752352941176f, 0.6099195294117647f, 0.5517875882352942f,
+0.5217222156862745f, 0.6074589019607843f, 0.550249705882353f,
+0.5195691960784314f, 0.6049982745098039f, 0.5487118235294117f,
+0.5174161764705882f, 0.6025376470588235f, 0.5471739411764706f,
+0.5152631568627452f, 0.6000770196078431f, 0.5456360588235294f,
+0.513110137254902f, 0.5976163921568628f, 0.5440981764705882f,
+0.5109571176470589f, 0.5951557647058824f, 0.5425602941176471f,
+0.5088040980392158f, 0.5926951372549021f, 0.5410224117647059f,
+0.5066510784313726f, 0.5902345098039216f, 0.5394845294117647f,
+0.5044980588235294f, 0.5877738823529411f, 0.5379466470588234f,
+0.5023450392156863f, 0.5853132549019608f, 0.5364087647058823f,
+0.5001920196078431f, 0.5828526274509804f, 0.5348708823529411f,
+0.498039f, 0.580392f, 0.533333f,
+0.49557845098039216f, 0.5779314509803922f, 0.5314875882352941f,
+0.4931179019607843f, 0.5754709019607843f, 0.5296421764705882f,
+0.49065735294117646f, 0.5730103529411765f, 0.5277967647058823f,
+0.4881968039215686f, 0.5705498039215686f, 0.5259513529411765f,
+0.4857362549019608f, 0.5680892549019608f, 0.5241059411764706f,
+0.48327570588235297f, 0.5656287058823529f, 0.5222605294117647f,
+0.4808151568627451f, 0.5631681568627451f, 0.5204151176470588f,
+0.47835460784313727f, 0.5607076078431372f, 0.518569705882353f,
+0.4758940588235294f, 0.5582470588235294f, 0.5167242941176471f,
+0.47343350980392157f, 0.5557865098039215f, 0.5148788823529412f,
+0.4709729607843138f, 0.5533259607843137f, 0.5130334705882353f,
+0.46851241176470587f, 0.5508654117647058f, 0.5111880588235295f,
+0.4662056274509804f, 0.5484817450980392f, 0.5093426470588236f,
+0.46436013725490194f, 0.5463287254901961f, 0.5074972352941177f,
+0.4625146470588235f, 0.5441757058823529f, 0.5056518235294118f,
+0.4606691568627451f, 0.5420226862745098f, 0.5038064117647059f,
+0.4588236666666667f, 0.5398696666666666f, 0.501961f,
+0.45697817647058825f, 0.5377166470588235f, 0.5001155882352941f,
+0.4551326862745098f, 0.5355636274509803f, 0.49827017647058824f,
+0.4532871960784314f, 0.5334106078431372f, 0.49642476470588237f,
+0.45144170588235294f, 0.5312575882352941f, 0.4945793529411765f,
+0.4495962156862745f, 0.5291045686274509f, 0.4927339411764706f,
+0.44775072549019607f, 0.5269515490196078f, 0.49088852941176475f,
+0.44590523529411763f, 0.5247985294117646f, 0.4890431176470588f,
+0.4440597450980392f, 0.5226455098039215f, 0.48719770588235295f,
+0.4423680588235294f, 0.5208000588235293f, 0.4858136274509804f,
+0.44083017647058825f, 0.5192621764705883f, 0.4848908823529412f,
+0.4392922941176471f, 0.517724294117647f, 0.48396813725490195f,
+0.43775441176470586f, 0.5161864117647058f, 0.48304539215686276f,
+0.4362165294117647f, 0.5146485294117646f, 0.4821226470588235f,
+0.43467864705882353f, 0.5131106470588235f, 0.4811999019607843f,
+0.43314076470588236f, 0.5115727647058823f, 0.4802771568627451f,
+0.4316028823529412f, 0.5100348823529411f, 0.4793544117647059f,
+0.430065f, 0.508497f, 0.47843166666666664f,
+0.4285271176470588f, 0.5069591176470588f, 0.47750892156862745f,
+0.42698923529411764f, 0.5054212352941176f, 0.4765861764705882f,
+0.4254513529411765f, 0.5038833529411765f, 0.475663431372549f,
+0.4239134705882353f, 0.5023454705882353f, 0.47474068627450977f,
+0.42283699999999996f, 0.5005768823529412f, 0.4738179411764706f,
+0.42191433333333334f, 0.49873139215686274f, 0.4728951960784314f,
+0.42099166666666665f, 0.4968859019607843f, 0.47197245098039214f,
+0.42006899999999997f, 0.49504041176470587f, 0.47104970588235295f,
+0.41914633333333334f, 0.4931949215686275f, 0.47012696078431376f,
+0.41822366666666666f, 0.491349431372549f, 0.4692042156862745f,
+0.417301f, 0.48950394117647056f, 0.46828147058823527f,
+0.41637833333333335f, 0.4876584509803922f, 0.4673587254901961f,
+0.41545566666666667f, 0.48581296078431374f, 0.4664359803921569f,
+0.414533f, 0.4839674705882353f, 0.46551323529411764f,
+0.4136103333333333f, 0.48212198039215687f, 0.46459049019607845f,
+0.4126876666666667f, 0.48027649019607843f, 0.4636677450980392f,
+0.411765f, 0.478431f, 0.462745f,
+0.411149862745098f, 0.4768931960784314f, 0.4618222549019608f,
+0.4105347254901961f, 0.47535539215686273f, 0.4608995098039216f,
+0.4099195882352941f, 0.47381758823529413f, 0.4599767647058824f,
+0.40930445098039214f, 0.47227978431372547f, 0.45905401960784314f,
+0.40868931372549017f, 0.47074198039215687f, 0.45813127450980395f,
+0.40807417647058825f, 0.46920417647058826f, 0.4572085294117647f,
+0.40745903921568627f, 0.4676663725490196f, 0.4562857843137255f,
+0.40684390196078435f, 0.466128568627451f, 0.4553630392156863f,
+0.4062287647058824f, 0.46459076470588234f, 0.4544402941176471f,
+0.4056136274509804f, 0.46305296078431374f, 0.45351754901960784f,
+0.4049984901960784f, 0.4615151568627451f, 0.45259480392156864f,
+0.4043833529411765f, 0.4599773529411765f, 0.4516720588235294f,
+0.4037681960784314f, 0.4584395294117647f, 0.45067243137254903f,
+0.40315298039215686f, 0.4569016470588235f, 0.4494421568627451f,
+0.4025377647058824f, 0.45536376470588236f, 0.4482118823529412f,
+0.40192254901960783f, 0.4538258823529412f, 0.44698160784313723f,
+0.40130733333333335f, 0.452288f, 0.44575133333333333f,
+0.4006921176470588f, 0.4507501176470588f, 0.44452105882352944f,
+0.4000769019607843f, 0.44921223529411763f, 0.4432907843137255f,
+0.3994616862745098f, 0.44767435294117647f, 0.4420605098039216f,
+0.3988464705882353f, 0.4461364705882353f, 0.44083023529411763f,
+0.39823125490196076f, 0.44459858823529413f, 0.43959996078431374f,
+0.3976160392156863f, 0.4430607058823529f, 0.43836968627450984f,
+0.3970008235294118f, 0.4415228235294118f, 0.43713941176470594f,
+0.39638560784313726f, 0.4399849411764706f, 0.435909137254902f,
+0.395770431372549f, 0.4384470588235294f, 0.43467886274509804f,
+0.39515529411764705f, 0.43690917647058825f, 0.43344858823529414f,
+0.3945401568627451f, 0.4353712941176471f, 0.4322183137254902f,
+0.39392501960784315f, 0.43383341176470586f, 0.4309880392156863f,
+0.3933098823529412f, 0.4322955294117647f, 0.42975776470588234f,
+0.3926947450980392f, 0.4307576470588235f, 0.42852749019607844f,
+0.3920796078431372f, 0.42921976470588236f, 0.4272972156862745f,
+0.3914644705882353f, 0.4276818823529412f, 0.4260669411764706f,
+0.3908493333333333f, 0.42614399999999997f, 0.42483666666666664f,
+0.39023419607843135f, 0.4246061176470588f, 0.42360639215686274f,
+0.38961905882352943f, 0.42306823529411763f, 0.4223761176470588f,
+0.38900392156862745f, 0.42153035294117647f, 0.4211458431372549f,
+0.3883887843137255f, 0.4199924705882353f, 0.41991556862745094f,
+0.3877736470588235f, 0.41822388235294117f, 0.4184545882352941f,
+0.3871585098039216f, 0.4163783921568628f, 0.41691670588235297f,
+0.3865433725490196f, 0.4145329019607843f, 0.41537882352941174f,
+0.3859282352941176f, 0.41268741176470586f, 0.4138409411764706f,
+0.3853130980392157f, 0.4108419215686274f, 0.4123030588235294f,
+0.38469796078431373f, 0.408996431372549f, 0.41076517647058824f,
+0.38408282352941175f, 0.40715094117647055f, 0.4092272941176471f,
+0.38346768627450983f, 0.40530545098039217f, 0.4076894117647059f,
+0.38285254901960786f, 0.40345996078431373f, 0.4061515294117647f,
+0.3822374117647059f, 0.4016144705882353f, 0.4046136470588235f,
+0.3816222745098039f, 0.39976898039215686f, 0.40307576470588236f,
+0.381007137254902f, 0.3979234901960784f, 0.4015378823529412f,
+0.380392f, 0.396078f, 0.4f
+}
+, 
+{
+1.0f, 0.988235f, 0.968627f,
+1.0f, 0.9839289607843137f, 0.9532483333333334f,
+1.0f, 0.9796229215686274f, 0.9378696666666667f,
+1.0f, 0.9753168823529411f, 0.9224910000000001f,
+1.0f, 0.9710108431372548f, 0.9071123333333334f,
+1.0f, 0.9667048039215687f, 0.8917336666666666f,
+1.0f, 0.9623987647058824f, 0.876355f,
+0.9969242156862745f, 0.9569393921568627f, 0.8602074117647059f,
+0.9920029607843137f, 0.9507880196078431f, 0.8435984705882353f,
+0.987081705882353f, 0.9446366470588236f, 0.8269895294117647f,
+0.9821604509803922f, 0.9384852745098039f, 0.8103805882352941f,
+0.9772391960784313f, 0.9323339019607843f, 0.7937716470588235f,
+0.9723179411764706f, 0.9261825294117647f, 0.7771627058823529f,
+0.968319431372549f, 0.9212613921568628f, 0.7610150980392156f,
+0.9670891568627451f, 0.9200309607843137f, 0.7462514901960784f,
+0.9658588823529412f, 0.9188005294117647f, 0.7314878823529412f,
+0.9646286078431372f, 0.9175700980392156f, 0.7167242745098039f,
+0.9633983333333334f, 0.9163396666666667f, 0.7019606666666667f,
+0.9621680588235294f, 0.9151092352941177f, 0.6871970588235294f,
+0.9609377843137255f, 0.9138788039215686f, 0.6724334509803921f,
+0.958092705882353f, 0.9126485098039215f, 0.6608994509803922f,
+0.9550169411764706f, 0.9114182352941177f, 0.6498268235294117f,
+0.9519411764705882f, 0.9101879607843137f, 0.6387541960784313f,
+0.9488654117647058f, 0.9089576862745098f, 0.627681568627451f,
+0.9457896470588235f, 0.9077274117647058f, 0.6166089411764706f,
+0.9427138823529412f, 0.9064971372549019f, 0.6055363137254902f,
+0.9402533333333334f, 0.9068047450980392f, 0.5953863529411765f,
+0.938408f, 0.9086502352941176f, 0.5861590588235294f,
+0.9365626666666667f, 0.9104957254901961f, 0.5769317647058824f,
+0.9347173333333334f, 0.9123412156862745f, 0.5677044705882353f,
+0.932872f, 0.914186705882353f, 0.5584771764705883f,
+0.9310266666666667f, 0.9160321960784313f, 0.5492498823529411f,
+0.928796843137255f, 0.9174932156862745f, 0.5404070784313726f,
+0.9238755882352941f, 0.9162629411764706f, 0.5342557058823529f,
+0.9189543333333333f, 0.9150326666666666f, 0.5281043333333333f,
+0.9140330784313726f, 0.9138023921568628f, 0.5219529607843137f,
+0.9091118235294118f, 0.9125721176470588f, 0.5158015882352941f,
+0.9041905686274511f, 0.9113418431372549f, 0.5096502156862746f,
+0.8992693137254902f, 0.9101115686274509f, 0.5034988431372549f,
+0.8920413529411765f, 0.9074971764705881f, 0.4973473529411765f,
+0.8840444901960784f, 0.9044214117647058f, 0.49119582352941177f,
+0.8760476274509804f, 0.9013456470588235f, 0.4850442941176471f,
+0.8680507647058824f, 0.8982698823529411f, 0.4788927647058824f,
+0.8600539019607844f, 0.8951941176470588f, 0.47274123529411766f,
+0.8520570392156863f, 0.8921183529411765f, 0.46658970588235293f,
+0.8426760588235295f, 0.8883505882352941f, 0.46089958823529414f,
+0.8309882156862746f, 0.8834294901960784f, 0.45597849019607845f,
+0.8193003725490197f, 0.8785083921568627f, 0.45105739215686275f,
+0.8076125294117648f, 0.873587294117647f, 0.44613629411764705f,
+0.79592468627451f, 0.8686661960784314f, 0.4412151960784314f,
+0.784236843137255f, 0.8637450980392157f, 0.4362940980392157f,
+0.772549f, 0.858824f, 0.431373f,
+0.7590158235294118f, 0.8545179607843137f, 0.42645174509803924f,
+0.7454826470588236f, 0.8502119215686275f, 0.42153049019607847f,
+0.7319494705882353f, 0.8459058823529412f, 0.41660923529411764f,
+0.7184162941176471f, 0.8415998431372549f, 0.41168798039215687f,
+0.7048831176470588f, 0.8372938039215687f, 0.4067667254901961f,
+0.6913499411764706f, 0.8329877647058824f, 0.40184547058823533f,
+0.6782011764705882f, 0.8282972156862746f, 0.3969242156862745f,
+0.6652830588235293f, 0.8233759607843137f, 0.39200296078431374f,
+0.6523649411764706f, 0.818454705882353f, 0.38708170588235297f,
+0.6394468235294118f, 0.8135334509803922f, 0.3821604509803922f,
+0.6265287058823529f, 0.8086121960784314f, 0.37723919607843137f,
+0.613610588235294f, 0.8036909411764707f, 0.3723179411764706f,
+0.6005386666666667f, 0.7987696862745098f, 0.3673967254901961f,
+0.5870053333333333f, 0.793848431372549f, 0.3624756274509804f,
+0.5734720000000002f, 0.7889271764705883f, 0.3575545294117648f,
+0.5599386666666666f, 0.7840059215686275f, 0.35263343137254904f,
+0.5464053333333333f, 0.7790846666666666f, 0.34771233333333335f,
+0.532872f, 0.7741634117647058f, 0.34279123529411765f,
+0.5193386666666666f, 0.769242156862745f, 0.337870137254902f,
+0.5036523529411765f, 0.7621680588235293f, 0.3324107254901961f,
+0.48765847058823525f, 0.7547864117647058f, 0.3268744117647059f,
+0.4716645882352941f, 0.7474047647058824f, 0.3213380980392157f,
+0.45567070588235314f, 0.740023117647059f, 0.3158017843137256f,
+0.43967682352941173f, 0.7326414705882353f, 0.3102654705882353f,
+0.4236829411764706f, 0.7252598235294118f, 0.3047291568627451f,
+0.40861180392156865f, 0.7181856274509804f, 0.29950037254901957f,
+0.3944634117647059f, 0.7114188823529412f, 0.2945791176470588f,
+0.38031501960784314f, 0.704652137254902f, 0.289657862745098f,
+0.3661666274509804f, 0.6978853921568627f, 0.28473660784313726f,
+0.35201823529411763f, 0.6911186470588235f, 0.27981535294117643f,
+0.3378698431372551f, 0.6843519019607844f, 0.2748940980392158f,
+0.32395213725490196f, 0.6778927450980392f, 0.2701266470588235f,
+0.31164923529411764f, 0.6735867058823529f, 0.26643582352941175f,
+0.2993463333333333f, 0.6692806666666666f, 0.262745f,
+0.287043431372549f, 0.6649746274509803f, 0.25905417647058826f,
+0.2747405294117647f, 0.6606685882352942f, 0.25536335294117646f,
+0.26243762745098037f, 0.6563625490196079f, 0.2516725294117647f,
+0.2501347254901961f, 0.6520565098039216f, 0.24798170588235294f,
+0.24382947058823534f, 0.6463664705882354f, 0.24936582352941172f,
+0.23952343137254903f, 0.6402150980392157f, 0.2524415882352941f,
+0.23521739215686274f, 0.6340637254901961f, 0.25551735294117645f,
+0.23091135294117648f, 0.6279123529411764f, 0.25859311764705883f,
+0.2266053137254902f, 0.6217609803921569f, 0.26166888235294117f,
+0.2222992745098039f, 0.6156096078431372f, 0.2647446470588235f,
+0.21799323529411765f, 0.6099195294117647f, 0.2675897058823529f,
+0.21368719607843137f, 0.6049982745098039f, 0.2700502549019608f,
+0.20938115686274517f, 0.6000770196078432f, 0.27251080392156857f,
+0.20507511764705882f, 0.5951557647058824f, 0.2749713529411765f,
+0.20076907843137254f, 0.5902345098039216f, 0.2774319019607843f,
+0.19646303921568628f, 0.5853132549019608f, 0.27989245098039217f,
+0.192157f, 0.580392f, 0.282353f,
+0.1866206862745098f, 0.5773162352941177f, 0.2835832745098039f,
+0.1810843725490196f, 0.5742404705882352f, 0.28481354901960787f,
+0.17554805882352942f, 0.5711647058823529f, 0.28604382352941177f,
+0.1700117450980393f, 0.5680889411764706f, 0.28727409803921566f,
+0.16447543137254902f, 0.5650131764705882f, 0.2885043725490196f,
+0.15893911764705881f, 0.5619374117647058f, 0.2897346470588235f,
+0.15301829411764706f, 0.5588616470588235f, 0.2905805098039216f,
+0.14686676470588236f, 0.5557858823529411f, 0.29119572549019607f,
+0.14071523529411764f, 0.5527101176470588f, 0.2918109411764706f,
+0.13456370588235295f, 0.5496343529411765f, 0.2924261568627451f,
+0.12841217647058822f, 0.5465585882352941f, 0.29304137254901963f,
+0.12226064705882361f, 0.5434828235294118f, 0.2936565882352941f,
+0.11657049019607843f, 0.5404070980392157f, 0.2945793333333333f,
+0.11226445098039216f, 0.5373314901960784f, 0.29642466666666667f,
+0.10795841176470589f, 0.5342558823529412f, 0.29827f,
+0.10365237254901961f, 0.5311802745098039f, 0.30011533333333335f,
+0.09934633333333334f, 0.5281046666666667f, 0.30196066666666666f,
+0.09504029411764706f, 0.5250290588235293f, 0.30380599999999996f,
+0.09073425490196078f, 0.5219534509803921f, 0.30565133333333333f,
+0.08750470588235298f, 0.5199541960784314f, 0.308573294117647f,
+0.08442894117647058f, 0.5181087058823529f, 0.3116490588235294f,
+0.08135317647058823f, 0.5162632156862745f, 0.3147248235294118f,
+0.07827741176470589f, 0.514417725490196f, 0.3178005882352941f,
+0.07520164705882353f, 0.5125722352941177f, 0.32087635294117645f,
+0.07212588235294118f, 0.5107267450980393f, 0.32395211764705883f,
+0.07151074509803922f, 0.509804f, 0.32887329411764704f,
+0.07335623529411764f, 0.509804f, 0.3356398823529412f,
+0.07520172549019608f, 0.509804f, 0.3424064705882353f,
+0.07704721568627451f, 0.509804f, 0.3491730588235294f,
+0.07889270588235288f, 0.509804f, 0.3559396470588233f,
+0.08073819607843137f, 0.509804f, 0.3627062352941176f,
+0.08250678431372549f, 0.5100346862745099f, 0.36947284313725487f,
+0.08373705882352941f, 0.5118801764705883f, 0.3762395882352941f,
+0.08496733333333333f, 0.5137256666666666f, 0.38300633333333334f,
+0.08619760784313725f, 0.5155711568627451f, 0.38977307843137254f,
+0.08742788235294117f, 0.5174166470588235f, 0.39653982352941175f,
+0.0886581568627451f, 0.5192621372549019f, 0.40330656862745096f,
+0.08988843137254902f, 0.5211076274509804f, 0.41007331372549016f,
+0.09065741176470588f, 0.5224917058823529f, 0.4173013529411765f,
+0.09127262745098039f, 0.5237219803921568f, 0.4246831568627451f,
+0.0918878431372549f, 0.5249522549019607f, 0.4320649607843137f,
+0.09250305882352941f, 0.5261825294117647f, 0.43944676470588234f,
+0.09311827450980392f, 0.5274128039215686f, 0.446828568627451f,
+0.09373349019607842f, 0.5286430784313726f, 0.45421037254901964f,
+0.09457935294117646f, 0.530104f, 0.4613614705882353f,
+0.09580962745098034f, 0.5319493333333333f, 0.4681280588235292f,
+0.09703990196078431f, 0.5337946666666666f, 0.47489464705882356f,
+0.09827017647058824f, 0.53564f, 0.48166123529411764f,
+0.09950045098039215f, 0.5374853333333334f, 0.4884278235294118f,
+0.10073072549019607f, 0.5393306666666666f, 0.49519441176470586f,
+0.101961f, 0.541176f, 0.501961f,
+0.10380633333333333f, 0.542406431372549f, 0.5093428039215686f,
+0.10565166666666667f, 0.543636862745098f, 0.5167246078431372f,
+0.107497f, 0.544867294117647f, 0.5241064117647058f,
+0.10934233333333333f, 0.5460977254901961f, 0.5314882156862745f,
+0.11118766666666667f, 0.547328156862745f, 0.5388700196078431f,
+0.11303300000000001f, 0.548558588235294f, 0.5462518235294117f,
+0.1152629411764706f, 0.5470976470588235f, 0.5501733333333333f,
+0.11772364705882353f, 0.5440218823529411f, 0.5520186666666667f,
+0.12018435294117648f, 0.5409461176470588f, 0.5538639999999999f,
+0.12264505882352941f, 0.5378703529411765f, 0.5557093333333333f,
+0.12510576470588228f, 0.5347945882352942f, 0.5575546666666665f,
+0.12756647058823528f, 0.5317188235294118f, 0.5593999999999999f,
+0.13002713725490195f, 0.529104431372549f, 0.5615529411764706f,
+0.1324876862745098f, 0.5278741568627451f, 0.5646287058823529f,
+0.13494823529411765f, 0.5266438823529411f, 0.5677044705882353f,
+0.1374087843137255f, 0.5254136078431372f, 0.5707802352941176f,
+0.13986933333333335f, 0.5241833333333333f, 0.573856f,
+0.14232988235294117f, 0.5229530588235294f, 0.5769317647058824f,
+0.14479043137254902f, 0.5217227843137254f, 0.5800075294117647f,
+0.14671280392156863f, 0.5183393921568628f, 0.583083294117647f,
+0.14855829411764707f, 0.5146484117647059f, 0.5861590588235294f,
+0.1504037843137255f, 0.510957431372549f, 0.5892348235294118f,
+0.1522492745098039f, 0.5072664509803921f, 0.5923105882352941f,
+0.15409476470588235f, 0.5035754705882353f, 0.5953863529411765f,
+0.15594025490196078f, 0.49988449019607845f, 0.5984621176470588f,
+0.15747813725490195f, 0.494040568627451f, 0.6009227450980392f,
+0.15870841176470585f, 0.4860437058823532f, 0.6027682352941176f,
+0.1599386862745098f, 0.4780468431372549f, 0.6046137254901961f,
+0.16116896078431373f, 0.47004998039215684f, 0.6064592156862745f,
+0.16239923529411765f, 0.4620531176470588f, 0.608304705882353f,
+0.16362950980392157f, 0.4540562549019608f, 0.6101501960784314f,
+0.16485978431372547f, 0.4460593725490196f, 0.6119187843137255f,
+0.1660900588235294f, 0.43806235294117646f, 0.6131490588235294f,
+0.16732033333333332f, 0.4300653333333333f, 0.6143793333333334f,
+0.16855060784313725f, 0.4220683137254902f, 0.6156096078431372f,
+0.16978088235294117f, 0.41407129411764704f, 0.6168398823529412f,
+0.1710111568627451f, 0.40607427450980393f, 0.6180701568627451f,
+0.17224143137254902f, 0.3980772549019608f, 0.6193004313725491f,
+0.17301041176470588f, 0.38961894117647056f, 0.6209921176470589f,
+0.1736256274509804f, 0.38100686274509804f, 0.6228376078431372f,
+0.1742408431372549f, 0.37239478431372547f, 0.6246830980392157f,
+0.17485605882352942f, 0.36378270588235295f, 0.6265285882352941f,
+0.1754712745098039f, 0.3551706274509806f, 0.6283740784313725f,
+0.17608649019607842f, 0.3465585490196078f, 0.630219568627451f,
+0.17693235294117646f, 0.33956123529411764f, 0.6318343529411764f,
+0.17816262745098038f, 0.33525519607843135f, 0.6330646274509804f,
+0.1793929019607843f, 0.33094915686274506f, 0.6342949019607843f,
+0.18062317647058823f, 0.3266431176470588f, 0.6355251764705883f,
+0.18185345098039216f, 0.32233707843137255f, 0.6367554509803921f,
+0.18308372549019608f, 0.31803103921568626f, 0.637985725490196f,
+0.184314f, 0.313725f, 0.639216f,
+0.184314f, 0.3075736274509804f, 0.6361402352941177f,
+0.184314f, 0.3014222549019608f, 0.6330644705882353f,
+0.184314f, 0.29527088235294113f, 0.629988705882353f,
+0.184314f, 0.28911950980392154f, 0.6269129411764707f,
+0.184314f, 0.28296813725490194f, 0.6238371764705882f,
+0.184314f, 0.27681676470588235f, 0.6207614117647059f,
+0.184314f, 0.27066529411764706f, 0.6176856470588236f,
+0.184314f, 0.2645137647058825f, 0.6146098823529413f,
+0.184314f, 0.2583622352941176f, 0.6115341176470589f,
+0.184314f, 0.25221070588235295f, 0.6084583529411765f,
+0.184314f, 0.24605917647058823f, 0.6053825882352941f,
+0.184314f, 0.23990764705882353f, 0.6023068235294118f,
+0.184314f, 0.23390992156862744f, 0.5992310588235293f,
+0.184314f, 0.22837360784313726f, 0.596155294117647f,
+0.184314f, 0.22283729411764708f, 0.5930795294117647f,
+0.184314f, 0.21730098039215687f, 0.5900037647058823f,
+0.184314f, 0.21176466666666668f, 0.586928f,
+0.184314f, 0.20622835294117647f, 0.5838522352941177f,
+0.184314f, 0.2006920392156863f, 0.5807764705882353f,
+0.18861996078431373f, 0.1994616862745098f, 0.577700705882353f,
+0.1935410588235294f, 0.1988464705882353f, 0.5746249411764706f,
+0.1984621568627451f, 0.19823125490196078f, 0.5715491764705882f,
+0.20338325490196077f, 0.19761603921568627f, 0.5684734117647059f,
+0.20830435294117633f, 0.19700082352941178f, 0.5653976470588236f,
+0.21322545098039214f, 0.19638560784313724f, 0.5623218823529411f,
+0.21722388235294116f, 0.1957704705882353f, 0.556785568627451f,
+0.22029964705882352f, 0.19515541176470588f, 0.548788705882353f,
+0.22337541176470588f, 0.19454035294117647f, 0.5407918431372549f,
+0.22645117647058824f, 0.19392529411764706f, 0.5327949803921569f,
+0.2295269411764706f, 0.19331023529411764f, 0.5247981176470589f,
+0.23260270588235293f, 0.19269517647058823f, 0.5168012549019608f,
+0.23583225490196078f, 0.19231078431372547f, 0.509188843137255f,
+0.24013829411764706f, 0.1935410588235294f, 0.5042675882352942f,
+0.24444433333333335f, 0.19477133333333332f, 0.49934633333333334f,
+0.2487503725490196f, 0.19600160784313725f, 0.49442507843137257f,
+0.2530564117647059f, 0.19723188235294117f, 0.4895038235294118f,
+0.2573624509803922f, 0.1984621568627451f, 0.48458256862745097f,
+0.26166849019607846f, 0.19969243137254902f, 0.4796613137254902f,
+0.26551323529411763f, 0.20230682352941176f, 0.4738174705882353f,
+0.2692042156862744f, 0.20538258823529404f, 0.4676660980392159f,
+0.2728951960784314f, 0.20845835294117648f, 0.46151472549019606f,
+0.27658617647058825f, 0.21153411764705882f, 0.45536335294117647f,
+0.2802771568627451f, 0.21460988235294118f, 0.4492119803921569f,
+0.28396813725490194f, 0.21768564705882354f, 0.4430606078431373f,
+0.28719770588235294f, 0.22076141176470587f, 0.43690917647058825f,
+0.2896582549019608f, 0.22383717647058823f, 0.4307576470588235f,
+0.29211880392156864f, 0.2269129411764706f, 0.4246061176470588f,
+0.2945793529411765f, 0.22998870588235296f, 0.41845458823529413f,
+0.2970399019607843f, 0.2330644705882353f, 0.4123030588235294f,
+0.29950045098039213f, 0.23614023529411765f, 0.40615152941176474f,
+0.301961f, 0.239216f, 0.4f
+}
+, 
+{
+1.0f, 1.0f, 0.988235f,
+0.9993720473856209f, 0.9995290155228758f, 0.9824263382352941f,
+0.9980905114379085f, 0.9985678227124183f, 0.9705719264705882f,
+0.9968089754901961f, 0.9976066299019608f, 0.9587175147058823f,
+0.9955274395424837f, 0.9966454370915032f, 0.9468631029411765f,
+0.9942459035947713f, 0.9956842442810457f, 0.9350086911764706f,
+0.9929643676470589f, 0.9947230514705883f, 0.9231542794117646f,
+0.9916828316993465f, 0.9937618586601307f, 0.9112998676470588f,
+0.990401295751634f, 0.9928006658496732f, 0.8994454558823529f,
+0.9891197598039216f, 0.9918394730392157f, 0.8875910441176471f,
+0.9878382238562091f, 0.9908782802287581f, 0.8757366323529412f,
+0.9865566879084967f, 0.9899170874183006f, 0.8638822205882353f,
+0.9852751519607843f, 0.9889558946078432f, 0.8520278088235295f,
+0.9843908823529411f, 0.988235f, 0.8411377647058824f,
+0.9846984117647058f, 0.988235f, 0.8331408235294118f,
+0.9850059411764706f, 0.988235f, 0.8251438823529412f,
+0.9853134705882353f, 0.988235f, 0.8171469411764706f,
+0.985621f, 0.988235f, 0.80915f,
+0.9859285294117647f, 0.988235f, 0.8011530588235294f,
+0.9862360588235294f, 0.988235f, 0.7931561176470588f,
+0.9865435882352941f, 0.988235f, 0.7851591764705882f,
+0.9868511176470588f, 0.988235f, 0.7771622352941177f,
+0.9871586470588235f, 0.988235f, 0.769165294117647f,
+0.9874661764705882f, 0.988235f, 0.7611683529411765f,
+0.9877737058823529f, 0.988235f, 0.7531714117647058f,
+0.9880812352941176f, 0.988235f, 0.7451744705882353f,
+0.987927431372549f, 0.9874660588235293f, 0.7377927058823529f,
+0.9873122941176471f, 0.9859281764705882f, 0.7310261176470588f,
+0.9866971568627451f, 0.984390294117647f, 0.7242595294117646f,
+0.9860820196078431f, 0.9828524117647058f, 0.7174929411764706f,
+0.9854668823529412f, 0.9813145294117647f, 0.7107263529411765f,
+0.9848517450980392f, 0.9797766470588235f, 0.7039597647058823f,
+0.9842366078431373f, 0.9782387647058823f, 0.6971931764705882f,
+0.9836214705882353f, 0.9767008823529412f, 0.6904265882352941f,
+0.9830063333333333f, 0.975163f, 0.6836599999999999f,
+0.9823911960784314f, 0.9736251176470588f, 0.6768934117647059f,
+0.9817760588235295f, 0.9720872352941177f, 0.6701268235294118f,
+0.9811609215686274f, 0.9705493529411765f, 0.6633602352941177f,
+0.9805457843137255f, 0.9690114705882353f, 0.6565936470588235f,
+0.980392f, 0.9672429411764706f, 0.6502884117647059f,
+0.980392f, 0.9653975294117647f, 0.6441369607843137f,
+0.980392f, 0.9635521176470588f, 0.6379855098039215f,
+0.980392f, 0.961706705882353f, 0.6318340588235294f,
+0.980392f, 0.9598612941176471f, 0.6256826078431372f,
+0.980392f, 0.9580158823529412f, 0.6195311568627451f,
+0.980392f, 0.9561704705882353f, 0.6133797058823529f,
+0.980392f, 0.9543250588235295f, 0.6072282549019608f,
+0.980392f, 0.9524796470588235f, 0.6010768039215686f,
+0.980392f, 0.9506342352941176f, 0.5949253529411764f,
+0.980392f, 0.9487888235294117f, 0.5887739019607843f,
+0.980392f, 0.9469434117647059f, 0.5826224509803921f,
+0.980392f, 0.945098f, 0.576471f,
+0.9794692549019608f, 0.9420222352941177f, 0.5693968039215686f,
+0.9785465098039217f, 0.9389464705882353f, 0.5623226078431373f,
+0.9776237647058824f, 0.9358707058823529f, 0.5552484117647059f,
+0.9767010196078432f, 0.9327949411764705f, 0.5481742156862744f,
+0.9757782745098039f, 0.9297191764705882f, 0.5411000196078432f,
+0.9748555294117648f, 0.9266434117647059f, 0.5340258235294117f,
+0.9739327843137255f, 0.9235676470588235f, 0.5269516274509803f,
+0.9730100392156863f, 0.9204918823529411f, 0.519877431372549f,
+0.972087294117647f, 0.9174161176470588f, 0.5128032352941176f,
+0.9711645490196079f, 0.9143403529411764f, 0.5057290392156863f,
+0.9702418039215687f, 0.9112645882352941f, 0.4986548431372549f,
+0.9693190588235294f, 0.9081888235294118f, 0.49158064705882354f,
+0.968627f, 0.9050361764705882f, 0.4843526470588235f,
+0.968627f, 0.9016528823529412f, 0.47666323529411764f,
+0.968627f, 0.8982695882352941f, 0.46897382352941186f,
+0.968627f, 0.894886294117647f, 0.4612844117647059f,
+0.968627f, 0.8915029999999999f, 0.453595f,
+0.968627f, 0.8881197058823529f, 0.44590558823529414f,
+0.968627f, 0.8847364117647059f, 0.43821617647058825f,
+0.968627f, 0.8813531176470588f, 0.43052676470588236f,
+0.968627f, 0.8779698235294118f, 0.42283735294117647f,
+0.968627f, 0.8745865294117647f, 0.4151479411764706f,
+0.968627f, 0.8712032352941177f, 0.4074585294117648f,
+0.968627f, 0.8678199411764705f, 0.3997691176470588f,
+0.968627f, 0.8644366470588235f, 0.39207970588235297f,
+0.968319431372549f, 0.8604382156862745f, 0.384236568627451f,
+0.967704294117647f, 0.8558246470588236f, 0.37623970588235295f,
+0.9670891568627451f, 0.8512110784313726f, 0.3682428431372549f,
+0.9664740196078432f, 0.8465975098039216f, 0.3602459803921569f,
+0.9658588823529412f, 0.8419839411764706f, 0.3522491176470588f,
+0.9652437450980392f, 0.8373703725490197f, 0.3442522549019609f,
+0.9646286078431372f, 0.8327568039215686f, 0.33625539215686273f,
+0.9640134705882353f, 0.8281432352941176f, 0.3282585294117647f,
+0.9633983333333334f, 0.8235296666666667f, 0.32026166666666667f,
+0.9627831960784313f, 0.8189160980392157f, 0.31226480392156863f,
+0.9621680588235294f, 0.8143025294117647f, 0.3042679411764706f,
+0.9615529215686274f, 0.8096889607843137f, 0.29627107843137257f,
+0.9609377843137255f, 0.8050753921568627f, 0.28827421568627454f,
+0.960092f, 0.8002310588235295f, 0.28235341176470596f,
+0.9591693333333333f, 0.7953098039215687f, 0.2771246274509804f,
+0.9582466666666667f, 0.7903885490196079f, 0.2718958431372549f,
+0.957324f, 0.785467294117647f, 0.2666670588235294f,
+0.9564013333333333f, 0.7805460392156862f, 0.26143827450980395f,
+0.9554786666666666f, 0.7756247843137255f, 0.2562094901960784f,
+0.954556f, 0.7707035294117647f, 0.25098070588235294f,
+0.9536333333333333f, 0.7657822745098039f, 0.24575192156862746f,
+0.9527106666666667f, 0.7608610196078431f, 0.24052313725490204f,
+0.951788f, 0.7559397647058823f, 0.23529435294117648f,
+0.9508653333333333f, 0.7510185098039215f, 0.23006556862745098f,
+0.9499426666666666f, 0.7460972549019608f, 0.2248367843137255f,
+0.94902f, 0.741176f, 0.219608f,
+0.9484047843137254f, 0.7362548235294117f, 0.21407168627450981f,
+0.947789568627451f, 0.7313336470588235f, 0.2085353725490196f,
+0.9471743529411765f, 0.7264124705882352f, 0.20299905882352942f,
+0.9465591372549019f, 0.7214912941176471f, 0.1974627450980393f,
+0.9459439215686274f, 0.7165701176470588f, 0.19192643137254903f,
+0.945328705882353f, 0.7116489411764706f, 0.18639011764705882f,
+0.9447134901960784f, 0.7067277647058823f, 0.18085380392156863f,
+0.9440982745098039f, 0.7018065882352941f, 0.17531749019607845f,
+0.9434830588235295f, 0.6968854117647059f, 0.16978117647058824f,
+0.9428678431372549f, 0.6919642352941177f, 0.16424486274509806f,
+0.9422526274509804f, 0.6870430588235295f, 0.15870854901960785f,
+0.941637411764706f, 0.6821218823529412f, 0.15317223529411772f,
+0.9409453333333333f, 0.6770469215686274f, 0.14794349019607844f,
+0.9400226666666667f, 0.6715106078431372f, 0.14363745098039216f,
+0.9391f, 0.6659742941176471f, 0.1393314117647059f,
+0.9381773333333333f, 0.6604379803921568f, 0.1350253725490196f,
+0.9372546666666667f, 0.6549016666666667f, 0.13071933333333335f,
+0.936332f, 0.6493653529411765f, 0.12641329411764707f,
+0.9354093333333333f, 0.6438290392156862f, 0.12210725490196078f,
+0.9344866666666667f, 0.6382927254901962f, 0.11780121568627458f,
+0.9335640000000001f, 0.6327564117647059f, 0.11349517647058824f,
+0.9326413333333333f, 0.6272200980392157f, 0.10918913725490195f,
+0.9317186666666667f, 0.6216837843137255f, 0.10488309803921568f,
+0.9307960000000001f, 0.6161474705882353f, 0.10057705882352941f,
+0.9298733333333333f, 0.6106111568627451f, 0.09627101960784312f,
+0.9291044313725491f, 0.6053824117647059f, 0.09258011764705881f,
+0.928489294117647f, 0.6004612352941177f, 0.08950435294117647f,
+0.9278741568627451f, 0.5955400588235295f, 0.08642858823529412f,
+0.9272590196078432f, 0.5906188823529411f, 0.08335282352941176f,
+0.9266438823529412f, 0.585697705882353f, 0.08027705882352949f,
+0.9260287450980392f, 0.5807765294117647f, 0.07720129411764706f,
+0.9254136078431372f, 0.5758553529411765f, 0.0741255294117647f,
+0.9247984705882353f, 0.5709341764705882f, 0.07104976470588235f,
+0.9241833333333334f, 0.566013f, 0.067974f,
+0.9235681960784313f, 0.5610918235294118f, 0.06489823529411765f,
+0.9229530588235294f, 0.5561706470588235f, 0.06182247058823529f,
+0.9223379215686274f, 0.5512494705882353f, 0.05874670588235294f,
+0.9217227843137255f, 0.546328294117647f, 0.05567094117647059f,
+0.9208769411764706f, 0.5416378235294117f, 0.05374858823529412f,
+0.9199541960784313f, 0.5370242549019607f, 0.05221070588235294f,
+0.9190314509803921f, 0.5324106862745098f, 0.05067282352941176f,
+0.9181087058823529f, 0.5277971176470588f, 0.04913494117647059f,
+0.9171859607843137f, 0.5231835490196078f, 0.04759705882352941f,
+0.9162632156862744f, 0.5185699803921568f, 0.046059176470588235f,
+0.9153404705882352f, 0.5139564117647059f, 0.04452129411764706f,
+0.9144177254901961f, 0.5093428431372551f, 0.04298341176470592f,
+0.9134949803921568f, 0.5047292745098039f, 0.04144552941176471f,
+0.9125722352941176f, 0.500115705882353f, 0.039907647058823534f,
+0.9116494901960783f, 0.495502137254902f, 0.03836976470588235f,
+0.9107267450980392f, 0.490888568627451f, 0.03683188235294117f,
+0.909804f, 0.486275f, 0.035294f,
+0.9082661176470588f, 0.48043107843137256f, 0.034063725490196074f,
+0.9067282352941176f, 0.4745871568627451f, 0.032833450980392156f,
+0.9051903529411764f, 0.46874323529411765f, 0.03160317647058823f,
+0.9036524705882353f, 0.4628993137254902f, 0.030372901960784314f,
+0.9021145882352941f, 0.45705539215686275f, 0.029142627450980393f,
+0.900576705882353f, 0.4512114705882353f, 0.02791235294117647f,
+0.8990388235294118f, 0.44536754901960784f, 0.026682078431372547f,
+0.8975009411764706f, 0.4395236274509804f, 0.02545180392156863f,
+0.8959630588235293f, 0.43367970588235294f, 0.024221529411764704f,
+0.8944251764705882f, 0.4278357843137255f, 0.022991254901960786f,
+0.8928872941176471f, 0.4219918627450982f, 0.021760980392156896f,
+0.8913494117647058f, 0.4161479411764706f, 0.020530705882352944f,
+0.8884274509803921f, 0.409688862745098f, 0.019223529411764705f,
+0.8813532549019608f, 0.4013843137254902f, 0.017685647058823528f,
+0.8742790588235294f, 0.39307976470588235f, 0.016147764705882355f,
+0.867204862745098f, 0.3847752156862745f, 0.014609882352941177f,
+0.8601306666666667f, 0.3764706666666667f, 0.013072f,
+0.8530564705882353f, 0.3681661176470588f, 0.011534117647058825f,
+0.8459822745098039f, 0.359861568627451f, 0.009996235294117648f,
+0.8389080784313726f, 0.35155701960784314f, 0.00845835294117647f,
+0.8318338823529412f, 0.34325247058823527f, 0.0069204705882352936f,
+0.8247596862745098f, 0.33494792156862746f, 0.005382588235294118f,
+0.8176854901960785f, 0.3266433725490196f, 0.003844705882352941f,
+0.8106112941176471f, 0.3183388235294118f, 0.002306823529411764f,
+0.8035370980392157f, 0.3100342745098039f, 0.0007689411764705868f,
+0.7984621176470589f, 0.30326764705882353f, 0.0f,
+0.7953863529411765f, 0.29803894117647073f, 0.0f,
+0.7923105882352942f, 0.29281023529411765f, 0.0f,
+0.7892348235294118f, 0.2875815294117647f, 0.0f,
+0.7861590588235294f, 0.28235282352941177f, 0.0f,
+0.7830832941176471f, 0.2771241176470588f, 0.0f,
+0.7800075294117648f, 0.2718954117647059f, 0.0f,
+0.7769317647058824f, 0.266666705882353f, 0.0f,
+0.773856f, 0.261438f, 0.0f,
+0.7707802352941177f, 0.25620929411764704f, 0.0f,
+0.7677044705882353f, 0.2509805882352941f, 0.0f,
+0.764628705882353f, 0.24575188235294118f, 0.0f,
+0.7615529411764707f, 0.24052317647058824f, 0.0f,
+0.7559397058823529f, 0.23575576470588236f, 0.0006920588235294117f,
+0.7494806470588236f, 0.23114211764705883f, 0.0016148039215686276f,
+0.7430215882352942f, 0.2265284705882353f, 0.002537549019607843f,
+0.7365625294117647f, 0.22191482352941178f, 0.003460294117647059f,
+0.7301034705882354f, 0.2173011764705884f, 0.004383039215686248f,
+0.7236444117647058f, 0.21268752941176472f, 0.00530578431372549f,
+0.7171853529411765f, 0.2080738823529412f, 0.006228529411764706f,
+0.7107262941176471f, 0.20346023529411766f, 0.00715127450980392f,
+0.7042672352941176f, 0.1988465882352941f, 0.008074019607843137f,
+0.6978081764705882f, 0.1942329411764706f, 0.008996764705882352f,
+0.6913491176470589f, 0.18961929411764705f, 0.009919509803921568f,
+0.6848900588235294f, 0.18500564705882352f, 0.010842254901960783f,
+0.678431f, 0.180392f, 0.011765f,
+0.6722795490196078f, 0.175778431372549f, 0.012687666666666667f,
+0.6661280980392157f, 0.17116486274509804f, 0.013610333333333334f,
+0.6599766470588235f, 0.16655129411764705f, 0.014533f,
+0.6538251960784314f, 0.16193772549019608f, 0.015455666666666666f,
+0.6476737450980392f, 0.1573241568627451f, 0.016378333333333335f,
+0.641522294117647f, 0.15271058823529413f, 0.017301f,
+0.6353708431372549f, 0.14809701960784313f, 0.018223666666666666f,
+0.629219392156863f, 0.14348345098039228f, 0.019146333333333307f,
+0.6230679411764706f, 0.13886988235294118f, 0.020069000000000004f,
+0.6169164901960784f, 0.13425631372549018f, 0.020991666666666665f,
+0.6107650392156863f, 0.12964274509803922f, 0.021914333333333334f,
+0.6046135882352941f, 0.12502917647058823f, 0.022837000000000003f,
+0.5980776666666666f, 0.12026180392156863f, 0.02368280392156863f,
+0.5903883333333333f, 0.11503301960784314f, 0.024298019607843137f,
+0.582699f, 0.10980423529411765f, 0.02491323529411765f,
+0.5750096666666666f, 0.10457545098039216f, 0.025528450980392157f,
+0.5673203333333333f, 0.09934666666666667f, 0.026143666666666666f,
+0.559631f, 0.09411788235294118f, 0.026758882352941177f,
+0.5519416666666667f, 0.08888909803921569f, 0.027374098039215686f,
+0.5442523333333333f, 0.08366031372549018f, 0.027989313725490197f,
+0.536563f, 0.0784315294117647f, 0.028604529411764706f,
+0.5288736666666667f, 0.0732027450980392f, 0.029219745098039214f,
+0.5211843333333334f, 0.06797396078431373f, 0.029834960784313726f,
+0.5134950000000003f, 0.06274517647058839f, 0.030450176470588217f,
+0.5058056666666666f, 0.05751639215686273f, 0.031065392156862742f,
+0.4979625294117647f, 0.05428686274509804f, 0.03244950980392157f,
+0.4899655882352941f, 0.05305658823529412f, 0.034602529411764706f,
+0.48196864705882353f, 0.051826313725490195f, 0.03675554901960784f,
+0.47397170588235293f, 0.05059603921568628f, 0.03890856862745098f,
+0.46597476470588234f, 0.04936576470588235f, 0.041061588235294114f,
+0.45797782352941174f, 0.048135490196078434f, 0.04321460784313726f,
+0.4499808823529412f, 0.04690521568627451f, 0.04536762745098039f,
+0.4419839411764706f, 0.045674941176470585f, 0.04752064705882353f,
+0.433987f, 0.04444466666666667f, 0.049673666666666665f,
+0.4259900588235294f, 0.04321439215686275f, 0.0518266862745098f,
+0.4179931176470588f, 0.041984117647058825f, 0.053979705882352944f,
+0.4099961764705883f, 0.0407538431372549f, 0.05613272549019608f,
+0.4019992352941177f, 0.03952356862745098f, 0.058285745098039216f,
+0.394233f, 0.03967735294117647f, 0.06066941176470588f,
+0.3865436666666669f, 0.04029249019607842f, 0.06312996078431365f,
+0.37885433333333335f, 0.040907627450980394f, 0.06559050980392157f,
+0.371165f, 0.041522764705882356f, 0.06805105882352941f,
+0.36347566666666664f, 0.04213790196078431f, 0.07051160784313726f,
+0.3557863333333333f, 0.042753039215686274f, 0.0729721568627451f,
+0.348097f, 0.043368176470588236f, 0.07543270588235294f,
+0.34040766666666666f, 0.04398331372549019f, 0.07789325490196078f,
+0.33271833333333334f, 0.044598450980392154f, 0.08035380392156863f,
+0.325029f, 0.045213588235294117f, 0.08281435294117646f,
+0.3173396666666667f, 0.04582872549019608f, 0.08527490196078431f,
+0.3096503333333333f, 0.046443862745098034f, 0.08773545098039215f,
+0.301961f, 0.047059f, 0.090196f
+}
+, 
+{
+0.941176f, 0.8f, 0.552941f,
+0.9396381960784314f, 0.7956939607843138f, 0.5470970784313726f,
+0.9381003921568628f, 0.7913879215686275f, 0.5412531568627451f,
+0.9365625882352941f, 0.7870818823529412f, 0.5354092352941177f,
+0.9350247843137255f, 0.7827758431372549f, 0.5295653137254902f,
+0.9334869803921568f, 0.7784698039215687f, 0.5237213921568628f,
+0.9319491764705883f, 0.7741637647058824f, 0.5178774705882353f,
+0.9304113725490196f, 0.7698577254901962f, 0.5120335490196078f,
+0.928873568627451f, 0.7655516862745099f, 0.5061896274509804f,
+0.9273357647058823f, 0.7612456470588236f, 0.5003457058823529f,
+0.9257979607843138f, 0.7569396078431373f, 0.4945017843137255f,
+0.9242601568627451f, 0.752633568627451f, 0.48865786274509804f,
+0.9227223529411764f, 0.7483275294117647f, 0.4828139411764706f,
+0.9211845294117647f, 0.743867705882353f, 0.47681623529411765f,
+0.9196466470588235f, 0.7389465294117648f, 0.4703571764705882f,
+0.9181087647058823f, 0.7340253529411765f, 0.4638981176470588f,
+0.9165708823529412f, 0.7291041764705882f, 0.4574390588235294f,
+0.915033f, 0.724183f, 0.45098f,
+0.9134951176470588f, 0.7192618235294118f, 0.44452094117647056f,
+0.9119572352941177f, 0.7143406470588235f, 0.4380618823529412f,
+0.9104193529411765f, 0.7094194705882353f, 0.43160282352941176f,
+0.9088814705882353f, 0.7044982941176471f, 0.42514376470588233f,
+0.9073435882352942f, 0.6995771176470589f, 0.41868470588235296f,
+0.905805705882353f, 0.6946559411764706f, 0.41222564705882353f,
+0.9042678235294118f, 0.6897347647058824f, 0.4057665882352941f,
+0.9027299411764707f, 0.6848135882352941f, 0.3993075294117647f,
+0.9011151764705883f, 0.6807382352941176f, 0.39454013725490195f,
+0.8994235294117647f, 0.6775087058823529f, 0.39146441176470587f,
+0.8977318823529412f, 0.6742791764705882f, 0.3883886862745098f,
+0.8960402352941177f, 0.6710496470588235f, 0.3853129607843137f,
+0.8943485882352942f, 0.6678201176470588f, 0.3822372352941176f,
+0.8926569411764707f, 0.6645905882352942f, 0.37916150980392155f,
+0.8909652941176471f, 0.6613610588235294f, 0.37608578431372547f,
+0.8892736470588236f, 0.6581315294117647f, 0.37301005882352944f,
+0.887582f, 0.654902f, 0.3699343333333333f,
+0.8858903529411765f, 0.6516724705882353f, 0.36685860784313723f,
+0.8841987058823529f, 0.6484429411764706f, 0.3637828823529412f,
+0.8825070588235294f, 0.6452134117647059f, 0.3607071568627451f,
+0.8808154117647059f, 0.6419838823529411f, 0.35763143137254905f,
+0.8791237647058824f, 0.6387543529411764f, 0.35455570588235297f,
+0.8774321176470589f, 0.6355248235294118f, 0.3514799803921569f,
+0.8757404705882353f, 0.6322952941176471f, 0.3484042549019608f,
+0.8740488235294118f, 0.6290657647058824f, 0.34532852941176473f,
+0.8723571764705883f, 0.6258362352941176f, 0.34225280392156865f,
+0.8706655294117648f, 0.6226067058823529f, 0.33917707843137257f,
+0.8689738823529413f, 0.6193771764705882f, 0.3361013529411765f,
+0.8672822352941176f, 0.6161476470588235f, 0.3330256274509804f,
+0.8655905882352941f, 0.6129181176470588f, 0.32994990196078433f,
+0.8638989411764706f, 0.609688588235294f, 0.32687417647058825f,
+0.8622072941176471f, 0.6064590588235295f, 0.32379845098039217f,
+0.8605156470588236f, 0.6032295294117647f, 0.3207227254901961f,
+0.858824f, 0.6f, 0.317647f,
+0.8572861176470589f, 0.5972318431372549f, 0.31610911764705885f,
+0.8557482352941177f, 0.5944636862745099f, 0.3145712352941177f,
+0.8542103529411765f, 0.5916955294117647f, 0.31303335294117646f,
+0.8526724705882354f, 0.5889273725490196f, 0.3114954705882353f,
+0.8511345882352942f, 0.5861592156862745f, 0.3099575882352941f,
+0.849596705882353f, 0.5833910588235294f, 0.30841970588235296f,
+0.8480588235294118f, 0.5806229019607843f, 0.3068818235294118f,
+0.8465209411764706f, 0.5778547450980392f, 0.30534394117647057f,
+0.8449830588235294f, 0.5750865882352941f, 0.3038060588235294f,
+0.8434451764705883f, 0.572318431372549f, 0.30226817647058823f,
+0.8419072941176471f, 0.5695502745098039f, 0.30073029411764707f,
+0.8403694117647059f, 0.5667821176470588f, 0.2991924117647059f,
+0.8388315294117648f, 0.5640139607843138f, 0.29765452941176473f,
+0.8372936470588236f, 0.5612458039215686f, 0.2961166470588235f,
+0.8357557647058824f, 0.5584776470588235f, 0.29457876470588235f,
+0.8342178823529411f, 0.5557094901960784f, 0.2930408823529412f,
+0.83268f, 0.5529413333333333f, 0.291503f,
+0.8311421176470588f, 0.5501731764705882f, 0.2899651176470588f,
+0.8296042352941176f, 0.5474050196078432f, 0.2884272352941176f,
+0.8280663529411765f, 0.544636862745098f, 0.28688935294117646f,
+0.8265284705882353f, 0.5418687058823529f, 0.2853514705882353f,
+0.8249905882352941f, 0.5391005490196078f, 0.2838135882352941f,
+0.823452705882353f, 0.5363323921568628f, 0.28227570588235296f,
+0.8219148235294118f, 0.5335642352941177f, 0.28073782352941173f,
+0.8203769411764706f, 0.5307960784313726f, 0.27919994117647057f,
+0.8192235294117647f, 0.5282585882352941f, 0.27750829411764705f,
+0.8184545882352942f, 0.5259517647058823f, 0.2756628823529412f,
+0.8176856470588235f, 0.5236449411764705f, 0.2738174705882353f,
+0.816916705882353f, 0.5213381176470588f, 0.2719720588235294f,
+0.8161477647058824f, 0.519031294117647f, 0.2701266470588235f,
+0.8153788235294118f, 0.5167244705882353f, 0.2682812352941177f,
+0.8146098823529412f, 0.5144176470588235f, 0.26643582352941175f,
+0.8138409411764707f, 0.5121108235294117f, 0.2645904117647059f,
+0.813072f, 0.509804f, 0.262745f,
+0.8123030588235294f, 0.5074971764705882f, 0.26089958823529413f,
+0.8115341176470589f, 0.5051903529411764f, 0.2590541764705882f,
+0.8107651764705882f, 0.5028835294117647f, 0.25720876470588233f,
+0.8099962352941177f, 0.5005767058823529f, 0.25536335294117646f,
+0.809227294117647f, 0.49826988235294123f, 0.2535179411764706f,
+0.8084583529411765f, 0.4959630588235294f, 0.2516725294117647f,
+0.8076894117647059f, 0.4936562352941177f, 0.2498271176470588f,
+0.8069204705882354f, 0.49134941176470587f, 0.24798170588235294f,
+0.8061515294117647f, 0.4890425882352941f, 0.24613629411764704f,
+0.8053825882352942f, 0.48673576470588237f, 0.24429088235294116f,
+0.8046136470588235f, 0.4844289411764706f, 0.2424454705882353f,
+0.803844705882353f, 0.4821221176470588f, 0.24060005882352942f,
+0.8030757647058824f, 0.4798152941176471f, 0.23875464705882354f,
+0.8023068235294118f, 0.4775084705882353f, 0.23690923529411764f,
+0.8015378823529412f, 0.47520164705882356f, 0.23506382352941177f,
+0.8007689411764707f, 0.47289482352941176f, 0.23321841176470587f,
+0.8f, 0.470588f, 0.231373f,
+0.7992310588235294f, 0.46843498039215686f, 0.23029649019607842f,
+0.7984621176470589f, 0.4662819607843137f, 0.22921998039215685f,
+0.7976931764705882f, 0.4641289411764706f, 0.22814347058823528f,
+0.7969242352941177f, 0.4619759215686275f, 0.22706696078431374f,
+0.7961552941176471f, 0.45982290196078435f, 0.22599045098039217f,
+0.7953863529411765f, 0.4576698823529412f, 0.2249139411764706f,
+0.7946174117647059f, 0.45551686274509806f, 0.22383743137254902f,
+0.7938484705882353f, 0.4533638431372549f, 0.22276092156862745f,
+0.7930795294117647f, 0.4512108235294118f, 0.22168441176470588f,
+0.7923105882352941f, 0.44905780392156863f, 0.2206079019607843f,
+0.7915416470588236f, 0.4469047843137255f, 0.21953139215686274f,
+0.7907727058823529f, 0.4447517647058824f, 0.2184548823529412f,
+0.7900037647058824f, 0.4425987450980392f, 0.2173783725490196f,
+0.7892348235294118f, 0.44044572549019606f, 0.21630186274509802f,
+0.7884658823529411f, 0.4382927058823529f, 0.21522535294117645f,
+0.7876969411764706f, 0.43613968627450983f, 0.2141488431372549f,
+0.786928f, 0.4339866666666667f, 0.21307233333333334f,
+0.7861590588235294f, 0.43183364705882354f, 0.21199582352941176f,
+0.7853901176470588f, 0.4296806274509804f, 0.2109193137254902f,
+0.7846211764705883f, 0.42752760784313726f, 0.20984280392156862f,
+0.7838522352941176f, 0.4253745882352941f, 0.20876629411764705f,
+0.783083294117647f, 0.423221568627451f, 0.20768978431372548f,
+0.7823143529411765f, 0.42106854901960783f, 0.2066132745098039f,
+0.7815454117647058f, 0.4189155294117647f, 0.20553676470588234f,
+0.7807764705882353f, 0.4167625098039216f, 0.2044602549019608f,
+0.7800075294117647f, 0.41460949019607846f, 0.20338374509803922f,
+0.7792385882352941f, 0.4124564705882353f, 0.20230723529411765f,
+0.7784696470588235f, 0.41030345098039217f, 0.20123072549019608f,
+0.777700705882353f, 0.408150431372549f, 0.2001542156862745f,
+0.7769317647058823f, 0.40599741176470594f, 0.19907770588235296f,
+0.7761628235294118f, 0.40384439215686274f, 0.19800119607843136f,
+0.7753938823529412f, 0.4016913725490196f, 0.1969246862745098f,
+0.7746249411764706f, 0.39953835294117646f, 0.19584817647058822f,
+0.773856f, 0.3973853333333333f, 0.19477166666666665f,
+0.7730870588235293f, 0.39523231372549017f, 0.19369515686274508f,
+0.7723181176470588f, 0.3930792941176471f, 0.19261864705882353f,
+0.7715491764705882f, 0.39092627450980394f, 0.19154213725490196f,
+0.7707802352941177f, 0.3887732549019608f, 0.1904656274509804f,
+0.770011294117647f, 0.38662023529411765f, 0.18938911764705882f,
+0.7692423529411765f, 0.3844672156862745f, 0.18831260784313725f,
+0.7684734117647058f, 0.38231419607843137f, 0.18723609803921568f,
+0.7677044705882353f, 0.3801611764705882f, 0.1861595882352941f,
+0.7669355294117647f, 0.3780081568627451f, 0.18508307843137253f,
+0.7661665882352942f, 0.37585513725490194f, 0.18400656862745096f,
+0.7653976470588235f, 0.37370211764705885f, 0.18293005882352942f,
+0.764628705882353f, 0.37154909803921576f, 0.18185354901960787f,
+0.7638597647058823f, 0.36939607843137257f, 0.18077703921568627f,
+0.7630908235294118f, 0.3672430588235294f, 0.1797005294117647f,
+0.7623218823529412f, 0.3650900392156863f, 0.17862401960784313f,
+0.7615529411764707f, 0.36293701960784314f, 0.17754750980392156f,
+0.760784f, 0.360784f, 0.176471f,
+0.7600150588235294f, 0.3592461568627451f, 0.1757020588235294f,
+0.7592461176470588f, 0.35770831372549017f, 0.17493311764705882f,
+0.7584771764705882f, 0.3561704705882353f, 0.17416417647058824f,
+0.7577082352941177f, 0.3546326274509804f, 0.17339523529411763f,
+0.756939294117647f, 0.3530947843137255f, 0.17262629411764704f,
+0.7561703529411765f, 0.35155694117647057f, 0.17185735294117646f,
+0.7554014117647059f, 0.35001909803921566f, 0.17108841176470588f,
+0.7546324705882352f, 0.3484812549019608f, 0.1703194705882353f,
+0.7538635294117647f, 0.3469434117647059f, 0.1695505294117647f,
+0.7530945882352941f, 0.345405568627451f, 0.1687815882352941f,
+0.7523256470588235f, 0.3438677254901961f, 0.16801264705882354f,
+0.7515567058823529f, 0.34232988235294115f, 0.16724370588235293f,
+0.7507877647058824f, 0.3407920392156863f, 0.16647476470588235f,
+0.7500188235294117f, 0.3392541960784314f, 0.16570582352941177f,
+0.7492498823529411f, 0.33771635294117647f, 0.16493688235294118f,
+0.7484809411764706f, 0.33617850980392155f, 0.1641679411764706f,
+0.7477119999999999f, 0.33464066666666664f, 0.163399f,
+0.7469430588235294f, 0.3331028235294118f, 0.1626300588235294f,
+0.7461741176470588f, 0.33156498039215687f, 0.16186111764705882f,
+0.7454051764705882f, 0.33002713725490196f, 0.16109217647058824f,
+0.7446362352941176f, 0.32848929411764705f, 0.16032323529411766f,
+0.743867294117647f, 0.32695145098039213f, 0.15955429411764704f,
+0.7430983529411764f, 0.3254136078431372f, 0.15878535294117646f,
+0.7423294117647058f, 0.32387576470588236f, 0.15801641176470588f,
+0.7415604705882353f, 0.32233792156862745f, 0.1572474705882353f,
+0.7407915490196078f, 0.3208000588235294f, 0.1564785294117647f,
+0.7400226470588235f, 0.3192621764705883f, 0.15570958823529413f,
+0.7392537450980392f, 0.3177242941176471f, 0.15494064705882354f,
+0.7384848431372548f, 0.31618641176470585f, 0.15417170588235293f,
+0.7377159411764705f, 0.3146485294117647f, 0.15340276470588235f,
+0.7369470392156863f, 0.3131106470588235f, 0.15263382352941177f,
+0.7361781372549019f, 0.31157276470588235f, 0.15186488235294118f,
+0.7354092352941176f, 0.3100348823529412f, 0.1510959411764706f,
+0.7346403333333333f, 0.308497f, 0.150327f,
+0.733871431372549f, 0.30695911764705885f, 0.1495580588235294f,
+0.7331025294117647f, 0.30542123529411763f, 0.14878911764705882f,
+0.7323336274509804f, 0.30388335294117647f, 0.14802017647058824f,
+0.731564725490196f, 0.3023454705882353f, 0.14725123529411765f,
+0.7307958235294117f, 0.30080758823529413f, 0.14648229411764704f,
+0.7300269215686275f, 0.29926970588235297f, 0.14571335294117646f,
+0.7292580196078431f, 0.2977318235294118f, 0.14494441176470588f,
+0.7284891176470588f, 0.2961939411764706f, 0.1441754705882353f,
+0.7277202156862745f, 0.29465605882352947f, 0.1434065294117647f,
+0.7269513137254902f, 0.29311817647058824f, 0.1426375882352941f,
+0.7261824117647059f, 0.2915802941176471f, 0.14186864705882352f,
+0.7254135098039216f, 0.2900424117647059f, 0.14109970588235293f,
+0.7246446078431372f, 0.28850452941176474f, 0.14033076470588235f,
+0.723875705882353f, 0.2869666470588236f, 0.13956182352941177f,
+0.7231068039215687f, 0.28542876470588235f, 0.13879288235294115f,
+0.7223379019607843f, 0.2838908823529412f, 0.13802394117647057f,
+0.721569f, 0.282353f, 0.137255f,
+0.7208000588235294f, 0.2811226862745098f, 0.1367936274509804f,
+0.7200311176470588f, 0.2798923725490196f, 0.13633225490196077f,
+0.7192621764705882f, 0.2786620588235294f, 0.13587088235294117f,
+0.7184932352941177f, 0.27743174509803925f, 0.13540950980392155f,
+0.717724294117647f, 0.27620143137254904f, 0.13494813725490196f,
+0.7169553529411765f, 0.27497111764705884f, 0.13448676470588233f,
+0.7161864117647059f, 0.27374080392156863f, 0.13402539215686274f,
+0.7154174705882353f, 0.2725104901960785f, 0.13356401960784314f,
+0.7146485294117647f, 0.2712801764705882f, 0.13310264705882352f,
+0.7138795882352941f, 0.27004986274509807f, 0.1326412745098039f,
+0.7131106470588235f, 0.26881954901960786f, 0.1321799019607843f,
+0.7123417058823529f, 0.26758923529411766f, 0.1317185294117647f,
+0.7115727647058824f, 0.26635892156862745f, 0.13125715686274508f,
+0.7108038235294117f, 0.26512860784313724f, 0.1307957843137255f,
+0.7100348823529411f, 0.26389829411764704f, 0.13033441176470587f,
+0.7092659411764706f, 0.2626679803921569f, 0.12987303921568627f,
+0.7084969999999999f, 0.2614376666666667f, 0.12941166666666665f,
+0.7077280588235294f, 0.2602073529411765f, 0.12895029411764705f,
+0.7069591176470588f, 0.25897703921568627f, 0.12848892156862743f,
+0.7061901764705882f, 0.25774672549019606f, 0.12802754901960783f,
+0.7054212352941176f, 0.25651641176470585f, 0.12756617647058824f,
+0.704652294117647f, 0.25528609803921565f, 0.12710480392156862f,
+0.7038833529411764f, 0.2540557843137255f, 0.12664343137254902f,
+0.7031144117647059f, 0.2528254705882353f, 0.12618205882352942f,
+0.7023454705882353f, 0.2515951568627451f, 0.1257206862745098f,
+0.7010382549019607f, 0.25005729411764704f, 0.1250286274509804f,
+0.6991927647058823f, 0.24821188235294117f, 0.12410588235294116f,
+0.6973472745098038f, 0.24636647058823527f, 0.12318313725490196f,
+0.6955017843137254f, 0.2445210588235294f, 0.12226039215686274f,
+0.6936562941176471f, 0.24267564705882352f, 0.12133764705882352f,
+0.6918108039215686f, 0.24083023529411762f, 0.1204149019607843f,
+0.6899653137254902f, 0.23898482352941175f, 0.1194921568627451f,
+0.6881198235294117f, 0.23713941176470588f, 0.11856941176470588f,
+0.6862743333333333f, 0.23529399999999998f, 0.11764666666666666f,
+0.6844288431372549f, 0.2334485882352941f, 0.11672392156862746f,
+0.6825833529411764f, 0.23160317647058823f, 0.11580117647058824f,
+0.6807378627450981f, 0.22975776470588233f, 0.11487843137254902f,
+0.6788923725490196f, 0.22791235294117645f, 0.1139556862745098f,
+0.6768162352941176f, 0.2256055294117647f, 0.11326364705882354f,
+0.6746632156862746f, 0.22314490196078438f, 0.11264850980392159f,
+0.6725101960784314f, 0.2206842745098039f, 0.11203337254901961f,
+0.6703571764705882f, 0.21822364705882352f, 0.11141823529411765f,
+0.6682041568627451f, 0.21576301960784314f, 0.11080309803921569f,
+0.666051137254902f, 0.21330239215686275f, 0.11018796078431373f,
+0.6638981176470589f, 0.21084176470588234f, 0.10957282352941176f,
+0.6617450980392157f, 0.20838113725490195f, 0.10895768627450982f,
+0.6595920784313726f, 0.20592050980392157f, 0.10834254901960785f,
+0.6574390588235294f, 0.20345988235294118f, 0.10772741176470589f,
+0.6552860392156863f, 0.20099925490196077f, 0.10711227450980393f,
+0.6531330196078431f, 0.1985386274509804f, 0.10649713725490197f,
+0.65098f, 0.196078f, 0.105882f
+}
+, 
+{
+0.301961f, 0.047059f, 0.090196f,
+0.33271833333333334f, 0.044598450980392154f, 0.08035380392156863f,
+0.36347566666666664f, 0.04213790196078431f, 0.07051160784313726f,
+0.394233f, 0.03967735294117647f, 0.06066941176470589f,
+0.4259900588235294f, 0.04321439215686274f, 0.05182668627450981f,
+0.45797782352941174f, 0.048135490196078434f, 0.04321460784313726f,
+0.4899655882352941f, 0.05305658823529412f, 0.034602529411764706f,
+0.5211843333333334f, 0.06797396078431373f, 0.029834960784313726f,
+0.5519416666666667f, 0.08888909803921569f, 0.027374098039215686f,
+0.582699f, 0.10980423529411765f, 0.02491323529411765f,
+0.6107650392156863f, 0.12964274509803922f, 0.021914333333333334f,
+0.6353708431372549f, 0.14809701960784313f, 0.018223666666666666f,
+0.6599766470588235f, 0.16655129411764705f, 0.014533f,
+0.6848900588235294f, 0.18500564705882352f, 0.010842254901960783f,
+0.7107262941176471f, 0.20346023529411766f, 0.007151274509803921f,
+0.7365625294117647f, 0.22191482352941178f, 0.003460294117647058f,
+0.7615529411764707f, 0.24052317647058824f, 0.0f,
+0.773856f, 0.261438f, 0.0f,
+0.7861590588235294f, 0.28235282352941177f, 0.0f,
+0.7984621176470589f, 0.30326764705882353f, 0.0f,
+0.8247596862745098f, 0.33494792156862746f, 0.005382588235294118f,
+0.8530564705882353f, 0.3681661176470588f, 0.011534117647058825f,
+0.8813532549019608f, 0.4013843137254902f, 0.01768564705882353f,
+0.8944251764705882f, 0.4278357843137255f, 0.022991254901960786f,
+0.900576705882353f, 0.4512114705882353f, 0.02791235294117647f,
+0.9067282352941176f, 0.4745871568627451f, 0.032833450980392156f,
+0.9116494901960783f, 0.49550213725490194f, 0.03836976470588235f,
+0.9153404705882352f, 0.5139564117647059f, 0.04452129411764706f,
+0.9190314509803921f, 0.5324106862745098f, 0.05067282352941176f,
+0.9223379215686274f, 0.5512494705882353f, 0.05874670588235294f,
+0.9247984705882353f, 0.5709341764705882f, 0.07104976470588235f,
+0.9272590196078432f, 0.5906188823529411f, 0.08335282352941176f,
+0.9298733333333333f, 0.6106111568627451f, 0.09627101960784314f,
+0.933564f, 0.6327564117647058f, 0.11349517647058811f,
+0.9372546666666667f, 0.6549016666666667f, 0.13071933333333335f,
+0.9409453333333333f, 0.6770469215686274f, 0.14794349019607844f,
+0.9434830588235295f, 0.6968854117647059f, 0.16978117647058824f,
+0.9459439215686274f, 0.7165701176470587f, 0.19192643137254886f,
+0.9484047843137254f, 0.7362548235294117f, 0.2140716862745098f,
+0.951788f, 0.7559397647058823f, 0.23529435294117645f,
+0.9554786666666666f, 0.7756247843137255f, 0.2562094901960784f,
+0.9591693333333333f, 0.7953098039215685f, 0.27712462745098027f,
+0.9621680588235294f, 0.8143025294117647f, 0.3042679411764706f,
+0.9646286078431372f, 0.8327568039215686f, 0.33625539215686273f,
+0.9670891568627451f, 0.8512110784313726f, 0.3682428431372549f,
+0.968627f, 0.8678199411764704f, 0.3997691176470586f,
+0.968627f, 0.8813531176470588f, 0.43052676470588236f,
+0.968627f, 0.894886294117647f, 0.4612844117647059f,
+0.9693190588235294f, 0.9081888235294118f, 0.49158064705882354f,
+0.9730100392156863f, 0.9204918823529411f, 0.5198774313725488f,
+0.9767010196078432f, 0.9327949411764705f, 0.5481742156862744f,
+0.980392f, 0.945098f, 0.576471f,
+0.980392f, 0.9524796470588235f, 0.6010768039215686f,
+0.980392f, 0.959861294117647f, 0.625682607843137f,
+0.980392f, 0.9672429411764706f, 0.6502884117647059f,
+0.9823911960784314f, 0.9736251176470588f, 0.6768934117647059f,
+0.9848517450980392f, 0.9797766470588235f, 0.7039597647058823f,
+0.987312294117647f, 0.9859281764705882f, 0.7310261176470586f,
+0.9874661764705882f, 0.988235f, 0.7611683529411765f,
+0.9862360588235294f, 0.988235f, 0.7931561176470588f,
+0.9850059411764706f, 0.988235f, 0.8251438823529412f,
+0.9865566879084967f, 0.9899170874183006f, 0.8638822205882349f,
+0.9916828316993465f, 0.9937618586601307f, 0.9112998676470587f,
+0.9968089754901961f, 0.9976066299019608f, 0.9587175147058823f,
+0.9052372294703552f, 0.9779564511584385f, 0.8243706029943159f,
+0.8867819030220032f, 0.9681138613764159f, 0.7960723730259015f,
+0.8683265765736515f, 0.9582712715943934f, 0.7677741430574876f,
+0.8496495152292368f, 0.9481330038699588f, 0.7391063674947f,
+0.827503374481646f, 0.9333688054588851f, 0.704656989519581f,
+0.8053572337340553f, 0.9186046070478114f, 0.670207611544462f,
+0.7832110929864646f, 0.9038404086367376f, 0.6357582335693429f,
+0.766433500561199f, 0.893371185783902f, 0.6093614727271444f,
+0.7504389784273724f, 0.8835284391328594f, 0.5841392873691418f,
+0.7344444562935458f, 0.8736856924818169f, 0.5589171020111392f,
+0.7184499341597195f, 0.8638429458307745f, 0.533694916653137f,
+0.7024554120258928f, 0.8540001991797318f, 0.508472731295134f,
+0.6864608898920662f, 0.8441574525286893f, 0.48325054593713146f,
+0.6726107669334224f, 0.8337019757621676f, 0.46262336775823965f,
+0.6609226131327606f, 0.8226287485193524f, 0.44662884562441313f,
+0.6492344593320987f, 0.8115555212765373f, 0.4306343234905866f,
+0.6375463055314369f, 0.8004822940337221f, 0.4146398013567601f,
+0.6258581517307751f, 0.7894090667909068f, 0.3986452792229336f,
+0.6141699979301134f, 0.7783358395480918f, 0.3826507570891073f,
+0.6028608203088908f, 0.767717433975006f, 0.3670352497919524f,
+0.5942483973806011f, 0.760335491638489f, 0.3541167722685378f,
+0.5856359744523114f, 0.7529535493019721f, 0.3411982947451232f,
+0.5770235515240217f, 0.7455716069654551f, 0.32827981722170857f,
+0.5684111285957321f, 0.7381896646289382f, 0.31536133969829394f,
+0.5597987056674425f, 0.7308077222924212f, 0.3024428621748793f,
+0.5511862827391528f, 0.7234257799559043f, 0.2895243846514647f,
+0.5439550917001388f, 0.71558323130337f, 0.2812100917001388f,
+0.5371880759214683f, 0.7075858918019468f, 0.2744430759214683f,
+0.530421060142798f, 0.6995885523005235f, 0.26767606014279793f,
+0.5236540443641275f, 0.6915912127991003f, 0.2609090443641276f,
+0.5168870285854572f, 0.6835938732976771f, 0.2541420285854572f,
+0.5101200128067869f, 0.6755965337962538f, 0.24737501280678684f,
+0.5026633222411251f, 0.6664497167745533f, 0.2401482529207051f,
+0.49405089931283536f, 0.655376646400758f, 0.2321510702883018f,
+0.4854384763845458f, 0.6443035760269629f, 0.22415388765589855f,
+0.476826053456256f, 0.6332305056531674f, 0.21615670502349513f,
+0.46821363052796633f, 0.6221574352793721f, 0.2081595223910918f,
+0.4596012075996766f, 0.6110843649055767f, 0.20016233975868847f,
+0.4509887846713869f, 0.6000112945317814f, 0.19216515712628512f,
+0.44237636174309725f, 0.588938224157986f, 0.1853970432864374f,
+0.43376393881480757f, 0.5778651537841907f, 0.1786301843767869f,
+0.42515151588651784f, 0.5667920834103953f, 0.17186332546713637f,
+0.4165390929582283f, 0.5557190130366001f, 0.16509646655748597f,
+0.4079266700299385f, 0.5446459426628045f, 0.15832960764783535f,
+0.39931424710164876f, 0.5335728722890092f, 0.15156274873818484f,
+0.3907018241733591f, 0.5217316928630087f, 0.14517984643150433f,
+0.3820894012450694f, 0.5094281418974407f, 0.13902807094872033f,
+0.37347697831677973f, 0.4971245909318726f, 0.13287629546593632f,
+0.36486455538849005f, 0.48482103996630455f, 0.1267245199831523f,
+0.3562521324602004f, 0.47251748900073653f, 0.12057274450036828f,
+0.3476397095319108f, 0.4602139380351687f, 0.11442096901758436f,
+0.3390273257408728f, 0.4477569299054316f, 0.10842268983622093f,
+0.33041505968160295f, 0.434838295512997f, 0.10288615464932324f,
+0.3218027936223331f, 0.4219196611205625f, 0.09734961946242557f,
+0.3131905275630633f, 0.40900102672812794f, 0.0918130842755279f,
+0.30457826150379347f, 0.3960823923356934f, 0.08627654908863021f,
+0.2959659954445236f, 0.38316375794325885f, 0.08074001390173252f,
+0.28735372938525383f, 0.37024512355082434f, 0.07520347871483485f,
+0.2771269282081088f, 0.35625026963353057f, 0.07020498468017158f,
+0.26666894126117985f, 0.342101311518343f, 0.06528353292014033f,
+0.25621095431425095f, 0.3279523534031557f, 0.06036208116010914f,
+0.24575296736732213f, 0.31380339528796836f, 0.05544062940007796f,
+0.23529498042039326f, 0.299654437172781f, 0.05051917764004678f,
+0.22483699347346442f, 0.28550547905759366f, 0.045597725880015594f,
+0.3816098233847023f, 0.39398373507705337f, 0.4030446360993551f,
+0.3840705692600429f, 0.39767485389006435f, 0.4091966576630016f,
+0.3865313151353836f, 0.4013659727030754f, 0.41534867922664814f,
+0.38899206101072425f, 0.4050570915160864f, 0.4215007007902947f,
+0.3914528068860649f, 0.4087482103290973f, 0.4276527223539411f,
+0.3939135527614056f, 0.4124393291421084f, 0.4338047439175878f,
+0.3963743175261149f, 0.4162045320589903f, 0.4399567465918657f,
+0.39883522027675045f, 0.42051091577848393f, 0.44610861128021734f,
+0.401296123027386f, 0.42481729949797753f, 0.452260475968569f,
+0.4037570257780215f, 0.42912368321747113f, 0.4584123406569207f,
+0.40621792852865707f, 0.43343006693696473f, 0.4645642053452724f,
+0.4086788312792926f, 0.43773645065645833f, 0.47071607003362403f,
+0.41113973402992815f, 0.44204283437595193f, 0.4768679347219757f,
+0.4145182796461964f, 0.4468081565446947f, 0.4834788548760093f,
+0.41820939845920746f, 0.4517298051706709f, 0.4902461413461384f,
+0.4219005172722185f, 0.4566514537966471f, 0.4970134278162676f,
+0.4255916360852295f, 0.46157310242262334f, 0.5037807142863968f,
+0.4292827548982405f, 0.4664947510485996f, 0.510548000756526f,
+0.43297387371125157f, 0.4714163996745758f, 0.5173152872266551f,
+0.43803604331672713f, 0.4774805129366699f, 0.5249965104447574f,
+0.4454184378180438f, 0.4854780154691743f, 0.534224385914932f,
+0.4528008323193609f, 0.49347551800167916f, 0.5434522613851074f,
+0.4601832268206778f, 0.5014730205341839f, 0.5526801368552824f,
+0.46756562132199475f, 0.5094705230666885f, 0.5619080123254574f,
+0.47494801582331164f, 0.5174680255991931f, 0.5711358877956323f,
+0.48233041032462853f, 0.5254655281316978f, 0.5803637632658073f,
+0.49155263873598226f, 0.5371430112748081f, 0.5926583935083312f,
+0.5007805142061572f, 0.5488317894956187f, 0.6049624366356243f,
+0.5100083896763322f, 0.5605205677164292f, 0.6172664797629175f,
+0.5192362651465072f, 0.5722093459372397f, 0.6295705228902105f,
+0.5284641406166822f, 0.5838981241580503f, 0.6418745660175037f,
+0.5376920160868572f, 0.5955869023788608f, 0.6541786091447969f,
+0.5469199892040675f, 0.6068927089272528f, 0.666865428650438f,
+0.5561480215495374f, 0.6179662222415807f, 0.679784422933624f,
+0.5653760538950072f, 0.6290397355559086f, 0.6927034172168101f,
+0.5746040862404771f, 0.6401132488702366f, 0.705622411499996f,
+0.5838321185859466f, 0.6511867621845642f, 0.7185414057831816f,
+0.5930601509314167f, 0.6622602754988924f, 0.731460400066368f,
+0.6027457888132203f, 0.6737914332483586f, 0.7442269110358488f,
+0.6138193021275482f, 0.6867105844068394f, 0.7565309541631419f,
+0.624892815441876f, 0.6996297355653202f, 0.768834997290435f,
+0.635966328756204f, 0.712548886723801f, 0.7811390404177281f,
+0.6470398420705319f, 0.7254680378822819f, 0.7934430835450211f,
+0.6581133553848597f, 0.7383871890407626f, 0.8057471266723143f,
+0.6691868686991876f, 0.7513063401992435f, 0.8180511697996073f,
+0.6797231035826917f, 0.7626133820135156f, 0.8292803820135156f,
+0.6901815088658318f, 0.7736868953278435f, 0.8403538953278435f,
+0.7006399141489721f, 0.7847604086421713f, 0.8514274086421714f,
+0.7110983194321123f, 0.7958339219564993f, 0.8625009219564993f,
+0.7215567247152525f, 0.8069074352708272f, 0.8735744352708272f,
+0.7320151299983927f, 0.8179809485851552f, 0.8846479485851553f,
+0.7433943934130215f, 0.829054540177121f, 0.8948006037679945f,
+0.7556982796650195f, 0.8401282103667435f, 0.9040286361134641f,
+0.7680021659170181f, 0.8512018805563666f, 0.9132566684589342f,
+0.7803060521690164f, 0.8622755507459894f, 0.922484700804404f,
+0.7926099384210146f, 0.8733492209356122f, 0.9317127331498739f,
+0.8049138246730129f, 0.884422891125235f, 0.9409407654953438f,
+0.8182899339607481f, 0.8958028800645715f, 0.9502453531164833f,
+0.8392067445270285f, 0.9093371392542401f, 0.960088336617846f,
+0.8601235550933087f, 0.9228713984439086f, 0.9699313201192087f,
+0.8810403656595891f, 0.9364056576335772f, 0.9797743036205714f,
+0.9169484148295987f, 0.955705868305041f, 0.9861579514817953f,
+0.9538602304608881f, 0.9753921490583561f, 0.9923099730454419f,
+0.9907720460921776f, 0.9950784298116713f, 0.9984619946090884f,
+0.9744873092420462f, 0.9744873092420462f, 0.9270234143318583f,
+0.9560315583016963f, 0.9560315583016963f, 0.8974938363265909f,
+0.9375758073613464f, 0.9375758073613464f, 0.8679642583213234f,
+0.9170615977193497f, 0.9170615977193497f, 0.8404931390177027f,
+0.8949145710906945f, 0.8949145710906945f, 0.8146548367007419f,
+0.8727675444620381f, 0.8727675444620381f, 0.7888165343837795f,
+0.8519876873835621f, 0.8510762022779806f, 0.7648012022779807f,
+0.8335316226926223f, 0.8301593917117002f, 0.7438843917117002f,
+0.8150755580016825f, 0.8092425811454197f, 0.7229675811454197f,
+0.7970702735105003f, 0.7890018834106215f, 0.7027268834106215f,
+0.7810752684454911f, 0.7717763485326472f, 0.6855013485326472f,
+0.7650802633804821f, 0.7545508136546729f, 0.6682758136546729f,
+0.749085258315473f, 0.7373252787766986f, 0.6510502787766986f,
+0.7343154501316902f, 0.7213252532504636f, 0.6362757626022031f,
+0.7195506611290563f, 0.7053302481854542f, 0.6215112873501589f,
+0.7047858721264224f, 0.6893352431204448f, 0.6067468120981148f,
+0.6890259843331534f, 0.6723451392648004f, 0.5909869843331534f,
+0.6730306655175545f, 0.655119604386826f, 0.5749916655175545f,
+0.6570353467019555f, 0.6378940695088515f, 0.5589963467019555f,
+0.641040222860408f, 0.6206685346308771f, 0.5437657161158874f,
+0.6250452177953992f, 0.6034429997529032f, 0.5290009271132541f,
+0.6090502127303895f, 0.5862174648749283f, 0.51423613811062f,
+0.5930552076653802f, 0.5689919299969538f, 0.4989372076653803f,
+0.5770602026003708f, 0.5517663951189792f, 0.4829422026003709f,
+0.5610651975353614f, 0.5345408602410047f, 0.4669471975353615f,
+0.5459816388371387f, 0.5179229821000844f, 0.45216750594446076f,
+0.5336775957098456f, 0.5031581930974507f, 0.4410939926301329f,
+0.5213735525825526f, 0.4883934040948169f, 0.430020479315805f,
+0.5090695094552595f, 0.47370207163840083f, 0.4188735094552594f,
+0.49676546632796625f, 0.46016781244873217f, 0.40656946632796626f,
+0.4844614232006731f, 0.4466335532590635f, 0.39426542320067304f,
+0.4721573800733799f, 0.4330992940693948f, 0.3819613800733799f,
+0.46200048854224385f, 0.4217116390124473f, 0.3728777906086043f,
+0.45215750504088115f, 0.41063781194752963f, 0.3642650231696171f,
+0.44231452153951845f, 0.399563984882612f, 0.3556522557306299f,
+0.43331448829164265f, 0.38933353803815574f, 0.34703948829164266f,
+0.42470172085265556f, 0.3794905545367932f, 0.33842672085265557f,
+0.41608895341366803f, 0.36964757103543017f, 0.32981395341366804f,
+0.40686343129881103f, 0.35919183285819767f, 0.32120118597468084f,
+0.3970201340468586f, 0.3481183195438698f, 0.3125884185356936f,
+0.38717683679490617f, 0.3370448062295419f, 0.3039756510967064f,
+0.37733363702997946f, 0.32635353954295376f, 0.2953628836577192f,
+0.36749065352861676f, 0.31651024229100133f, 0.286750116218732f,
+0.3576476700272541f, 0.3066669450390489f, 0.2781373487797447f,
+0.3479565813407575f, 0.2968236865258914f, 0.2695245813407575f,
+0.3393438139017703f, 0.2869807030245287f, 0.2609118139017703f,
+0.3307310464627831f, 0.27713771952316607f, 0.2522990464627831f,
+0.32211827902379586f, 0.26729473602180337f, 0.2436862790237959f,
+0.3123534587792868f, 0.25745145877928677f, 0.2350735115848086f,
+0.30251016152733423f, 0.24760816152733423f, 0.22646074414582132f,
+0.29266686427538174f, 0.23776486427538174f, 0.21784797670683403f,
+0.28282380201635243f, 0.2279218020163524f, 0.2092352092678468f,
+0.27298081851499f, 0.21807881851499f, 0.20062244182885983f,
+0.26313783501362703f, 0.20823583501362705f, 0.19200967438987238f,
+0.25398590695088513f, 0.19839267526757198f, 0.18408813863419826f,
+0.24537313951189782f, 0.1885493780156195f, 0.17670590100817618f,
+0.23676037207291054f, 0.17870608076366698f, 0.1693236633821541f,
+0.2276867835117145f, 0.1693236046339233f, 0.16148060463392327f,
+0.2178434862597621f, 0.16071083719493606f, 0.15286783719493605f,
+0.20800018900780967f, 0.15209806975594886f, 0.14425506975594884f,
+0.19884806469093963f, 0.14394606469093962f, 0.136333475252044f,
+0.1926960431272931f, 0.1377940431272931f, 0.13141198350136266f,
+0.18654402156364655f, 0.13164202156364654f, 0.12649049175068133f,
+0.180392f, 0.12549f, 0.121569f
+}
+, 
+{
+0.960784f, 0.964706f, 1.0f,
+0.9524795294117647f, 0.9567090588235294f, 0.9953863529411765f,
+0.9441750588235294f, 0.9487121176470588f, 0.9907727058823529f,
+0.9358705882352941f, 0.9407151764705882f, 0.9861590588235294f,
+0.9275661176470589f, 0.9327182352941176f, 0.9815454117647059f,
+0.9192616470588235f, 0.924721294117647f, 0.9769317647058824f,
+0.9109571764705883f, 0.9167243529411765f, 0.9723181176470588f,
+0.9026527058823529f, 0.9087274117647058f, 0.9677044705882353f,
+0.8943482352941177f, 0.9007304705882353f, 0.9630908235294118f,
+0.8860437647058824f, 0.8927335294117646f, 0.9584771764705883f,
+0.8777392941176471f, 0.8847365882352941f, 0.9538635294117647f,
+0.8694348235294118f, 0.8767396470588235f, 0.9492498823529412f,
+0.8611303529411765f, 0.8687427058823529f, 0.9446362352941177f,
+0.853671705882353f, 0.8613609215686274f, 0.9401763921568628f,
+0.8487505294117648f, 0.8558246078431372f, 0.9361779607843137f,
+0.8438293529411766f, 0.8502882941176471f, 0.9321795294117647f,
+0.8389081764705882f, 0.8447519803921568f, 0.9281810980392157f,
+0.833987f, 0.8392156666666667f, 0.9241826666666667f,
+0.8290658235294118f, 0.8336793529411765f, 0.9201842352941176f,
+0.8241446470588235f, 0.8281430392156862f, 0.9161858039215687f,
+0.8192234705882353f, 0.8226067254901961f, 0.9121873725490196f,
+0.8143022941176471f, 0.8170704117647059f, 0.9081889411764705f,
+0.8093811176470589f, 0.8115340980392157f, 0.9041905098039216f,
+0.8044599411764706f, 0.8059977843137255f, 0.9001920784313725f,
+0.7995387647058824f, 0.8004614705882352f, 0.8961936470588235f,
+0.7946175882352942f, 0.7949251568627451f, 0.8921952156862745f,
+0.7912342745098039f, 0.7906191176470588f, 0.8888119215686274f,
+0.7893888235294118f, 0.7875433529411765f, 0.8860437647058823f,
+0.7875433725490196f, 0.7844675882352942f, 0.8832756078431372f,
+0.7856979215686275f, 0.7813918235294117f, 0.8805074509803922f,
+0.7838524705882353f, 0.7783160588235294f, 0.8777392941176471f,
+0.7820070196078431f, 0.775240294117647f, 0.8749711372549019f,
+0.780161568627451f, 0.7721645294117647f, 0.8722029803921568f,
+0.7783161176470589f, 0.7690887647058824f, 0.8694348235294118f,
+0.7764706666666666f, 0.766013f, 0.8666666666666667f,
+0.7746252156862745f, 0.7629372352941177f, 0.8638985098039216f,
+0.7727797647058824f, 0.7598614705882353f, 0.8611303529411765f,
+0.7709343137254903f, 0.7567857058823529f, 0.8583621960784313f,
+0.769088862745098f, 0.7537099411764706f, 0.8555940392156862f,
+0.7672434117647059f, 0.7506341764705883f, 0.8528258823529412f,
+0.7653979607843138f, 0.7475584117647059f, 0.8500577254901961f,
+0.7635525098039216f, 0.7444826470588236f, 0.847289568627451f,
+0.7617070588235294f, 0.7414068823529412f, 0.8445214117647059f,
+0.7598616078431373f, 0.7383311176470588f, 0.8417532549019607f,
+0.7580161568627451f, 0.7352553529411765f, 0.8389850980392157f,
+0.756170705882353f, 0.7321795882352942f, 0.8362169411764706f,
+0.7543252549019608f, 0.7291038235294118f, 0.8334487843137255f,
+0.7524798039215687f, 0.7260280588235295f, 0.8306806274509804f,
+0.7506343529411765f, 0.7229522941176472f, 0.8279124705882352f,
+0.7487889019607844f, 0.7198765294117648f, 0.8251443137254902f,
+0.7469434509803922f, 0.7168007647058824f, 0.8223761568627451f,
+0.745098f, 0.713725f, 0.819608f,
+0.7410995294117647f, 0.710034137254902f, 0.8157633333333333f,
+0.7371010588235295f, 0.706343274509804f, 0.8119186666666667f,
+0.7331025882352942f, 0.7026524117647059f, 0.808074f,
+0.7291041176470588f, 0.6989615490196079f, 0.8042293333333334f,
+0.7251056470588235f, 0.6952706862745098f, 0.8003846666666666f,
+0.7211071764705883f, 0.6915798235294118f, 0.79654f,
+0.717108705882353f, 0.6878889607843137f, 0.7926953333333333f,
+0.7131102352941177f, 0.6841980980392157f, 0.7888506666666667f,
+0.7091117647058823f, 0.6805072352941177f, 0.785006f,
+0.7051132941176471f, 0.6768163725490197f, 0.7811613333333334f,
+0.7011148235294118f, 0.6731255098039216f, 0.7773166666666667f,
+0.6971163529411765f, 0.6694346470588236f, 0.773472f,
+0.6931178823529411f, 0.6657437843137255f, 0.7696273333333333f,
+0.6891194117647059f, 0.6620529215686275f, 0.7657826666666667f,
+0.6851209411764706f, 0.6583620588235295f, 0.7619380000000001f,
+0.6811224705882353f, 0.6546711960784314f, 0.7580933333333334f,
+0.677124f, 0.6509803333333334f, 0.7542486666666667f,
+0.6731255294117646f, 0.6472894705882354f, 0.750404f,
+0.6691270588235294f, 0.6435986078431373f, 0.7465593333333334f,
+0.6651285882352941f, 0.6399077450980393f, 0.7427146666666666f,
+0.6611301176470588f, 0.6362168823529413f, 0.73887f,
+0.6571316470588235f, 0.6325260196078432f, 0.7350253333333333f,
+0.6531331764705882f, 0.6288351568627452f, 0.7311806666666667f,
+0.6491347058823529f, 0.625144294117647f, 0.727336f,
+0.6451362352941176f, 0.621453431372549f, 0.7234913333333334f,
+0.6414453529411764f, 0.6177625490196079f, 0.7195697647058824f,
+0.6380620588235294f, 0.6140716470588236f, 0.7155712941176471f,
+0.6346787647058824f, 0.6103807450980393f, 0.7115728235294118f,
+0.6312954705882352f, 0.6066898431372549f, 0.7075743529411765f,
+0.6279121764705882f, 0.6029989411764707f, 0.7035758823529412f,
+0.6245288823529412f, 0.5993080392156863f, 0.699577411764706f,
+0.6211455882352941f, 0.595617137254902f, 0.6955789411764706f,
+0.617762294117647f, 0.5919262352941177f, 0.6915804705882354f,
+0.614379f, 0.5882353333333333f, 0.687582f,
+0.610995705882353f, 0.5845444313725491f, 0.6835835294117647f,
+0.6076124117647058f, 0.5808535294117647f, 0.6795850588235295f,
+0.6042291176470588f, 0.5771626274509805f, 0.6755865882352942f,
+0.6008458235294117f, 0.5734717254901961f, 0.6715881176470588f,
+0.5974625294117647f, 0.5697808235294118f, 0.6675896470588236f,
+0.5940792352941177f, 0.5660899215686275f, 0.6635911764705883f,
+0.5906959411764706f, 0.5623990196078432f, 0.659592705882353f,
+0.5873126470588235f, 0.5587081176470589f, 0.6555942352941176f,
+0.5839293529411764f, 0.5550172156862745f, 0.6515957647058824f,
+0.5805460588235294f, 0.5513263137254902f, 0.6475972941176471f,
+0.5771627647058823f, 0.5476354117647059f, 0.6435988235294118f,
+0.5737794705882353f, 0.5439445098039216f, 0.6396003529411765f,
+0.5703961764705883f, 0.5402536078431374f, 0.6356018823529412f,
+0.5670128823529412f, 0.536562705882353f, 0.6316034117647059f,
+0.5636295882352941f, 0.5328718039215686f, 0.6276049411764706f,
+0.560246294117647f, 0.5291809019607843f, 0.6236064705882354f,
+0.556863f, 0.52549f, 0.619608f,
+0.5548637450980393f, 0.5233369803921569f, 0.6173011764705882f,
+0.5528644901960784f, 0.5211839607843137f, 0.6149943529411765f,
+0.5508652352941177f, 0.5190309411764706f, 0.6126875294117647f,
+0.548865980392157f, 0.5168779215686274f, 0.610380705882353f,
+0.5468667254901961f, 0.5147249019607844f, 0.6080738823529412f,
+0.5448674705882353f, 0.5125718823529412f, 0.6057670588235294f,
+0.5428682156862745f, 0.5104188627450981f, 0.6034602352941176f,
+0.5408689607843138f, 0.5082658431372549f, 0.6011534117647059f,
+0.538869705882353f, 0.5061128235294118f, 0.5988465882352941f,
+0.5368704509803922f, 0.5039598039215687f, 0.5965397647058823f,
+0.5348711960784314f, 0.5018067843137255f, 0.5942329411764706f,
+0.5328719411764706f, 0.4996537647058824f, 0.5919261176470588f,
+0.5308726862745098f, 0.4975007450980392f, 0.589619294117647f,
+0.528873431372549f, 0.49534772549019607f, 0.5873124705882353f,
+0.5268741764705883f, 0.4931947058823529f, 0.5850056470588235f,
+0.5248749215686275f, 0.49104168627450984f, 0.5826988235294117f,
+0.5228756666666667f, 0.4888886666666667f, 0.580392f,
+0.5208764117647059f, 0.48673564705882355f, 0.5780851764705882f,
+0.5188771568627452f, 0.4845826274509804f, 0.5757783529411764f,
+0.5168779019607844f, 0.48242960784313726f, 0.5734715294117647f,
+0.5148786470588236f, 0.4802765882352941f, 0.5711647058823529f,
+0.5128793921568628f, 0.478123568627451f, 0.5688578823529411f,
+0.510880137254902f, 0.47597054901960784f, 0.5665510588235294f,
+0.5088808823529413f, 0.4738175294117647f, 0.5642442352941176f,
+0.5068816274509804f, 0.4716645098039216f, 0.5619374117647058f,
+0.5049592941176471f, 0.46951149019607846f, 0.5596306078431372f,
+0.5031138823529412f, 0.4673584705882353f, 0.5573238235294117f,
+0.5012684705882353f, 0.4652054509803922f, 0.5550170392156862f,
+0.49942305882352944f, 0.46305243137254903f, 0.5527102549019607f,
+0.4975776470588236f, 0.46089941176470595f, 0.5504034705882354f,
+0.4957322352941177f, 0.45874639215686275f, 0.5480966862745098f,
+0.4938868235294118f, 0.4565933725490196f, 0.5457899019607843f,
+0.49204141176470595f, 0.45444035294117646f, 0.5434831176470588f,
+0.490196f, 0.4522873333333333f, 0.5411763333333333f,
+0.48835058823529415f, 0.4501343137254902f, 0.5388695490196078f,
+0.4865051764705883f, 0.4479812941176471f, 0.5365627647058823f,
+0.4846597647058824f, 0.44582827450980395f, 0.5342559803921568f,
+0.48281435294117653f, 0.4436752549019608f, 0.5319491960784314f,
+0.4809689411764706f, 0.44152223529411766f, 0.5296424117647058f,
+0.4791235294117647f, 0.4393692156862745f, 0.5273356274509804f,
+0.47727811764705885f, 0.4372161960784314f, 0.5250288431372548f,
+0.475432705882353f, 0.43506317647058823f, 0.5227220588235294f,
+0.4735872941176471f, 0.4329101568627451f, 0.5204152745098038f,
+0.4717418823529412f, 0.43075713725490195f, 0.5181084901960784f,
+0.4698964705882353f, 0.42860411764705886f, 0.515801705882353f,
+0.4680510588235295f, 0.42645109803921577f, 0.5134949215686275f,
+0.46620564705882356f, 0.42429807843137257f, 0.511188137254902f,
+0.4643602352941177f, 0.42214505882352943f, 0.5088813529411764f,
+0.46251482352941176f, 0.4199920392156863f, 0.506574568627451f,
+0.4606694117647059f, 0.41783901960784314f, 0.5042677843137254f,
+0.458824f, 0.415686f, 0.501961f,
+0.4571323137254902f, 0.4138405490196078f, 0.4999617450980392f,
+0.4554406274509804f, 0.4119950980392157f, 0.4979624901960784f,
+0.4537489411764706f, 0.4101496470588235f, 0.4959632352941176f,
+0.4520572549019608f, 0.4083041960784314f, 0.49396398039215683f,
+0.450365568627451f, 0.4064587450980392f, 0.4919647254901961f,
+0.4486738823529412f, 0.40461329411764707f, 0.4899654705882353f,
+0.44698219607843137f, 0.4027678431372549f, 0.4879662156862745f,
+0.4452905098039216f, 0.40092239215686276f, 0.48596696078431373f,
+0.44359882352941177f, 0.3990769411764706f, 0.48396770588235294f,
+0.44190713725490194f, 0.3972314901960784f, 0.48196845098039215f,
+0.4402154509803922f, 0.3953860392156863f, 0.4799691960784314f,
+0.43852376470588234f, 0.3935405882352941f, 0.47796994117647057f,
+0.43683207843137256f, 0.39169513725490196f, 0.4759706862745098f,
+0.43514039215686273f, 0.3898496862745098f, 0.473971431372549f,
+0.43344870588235296f, 0.38800423529411765f, 0.4719721764705882f,
+0.43175701960784313f, 0.38615878431372547f, 0.46997292156862747f,
+0.43006533333333336f, 0.38431333333333334f, 0.4679736666666667f,
+0.4283736470588235f, 0.38246788235294116f, 0.4659744117647059f,
+0.42668196078431375f, 0.38062243137254903f, 0.4639751568627451f,
+0.4249902745098039f, 0.37877698039215685f, 0.4619759019607843f,
+0.4232985882352941f, 0.37693152941176467f, 0.4599766470588235f,
+0.4216069019607843f, 0.37508607843137254f, 0.45797739215686273f,
+0.4199152156862745f, 0.37324062745098036f, 0.45597813725490194f,
+0.4182235294117647f, 0.37139517647058823f, 0.45397888235294115f,
+0.4165318431372549f, 0.36954972549019605f, 0.4519796274509804f,
+0.41491707843137254f, 0.3677811764705882f, 0.44998039215686275f,
+0.4133792352941177f, 0.36608952941176476f, 0.44798117647058827f,
+0.4118413921568627f, 0.3643978823529412f, 0.44598196078431374f,
+0.41030354901960786f, 0.3627062352941176f, 0.4439827450980392f,
+0.40876570588235295f, 0.3610145882352941f, 0.4419835294117647f,
+0.40722786274509803f, 0.35932294117647057f, 0.4399843137254902f,
+0.4056900196078431f, 0.35763129411764705f, 0.4379850980392157f,
+0.4041521764705882f, 0.3559396470588235f, 0.43598588235294117f,
+0.40261433333333335f, 0.354248f, 0.4339866666666667f,
+0.40107649019607844f, 0.3525563529411765f, 0.43198745098039215f,
+0.3995386470588235f, 0.3508647058823529f, 0.42998823529411767f,
+0.3980008039215686f, 0.3491730588235294f, 0.42798901960784314f,
+0.3964629607843137f, 0.34748141176470587f, 0.42598980392156865f,
+0.39492511764705884f, 0.34578976470588235f, 0.4239905882352941f,
+0.39338727450980393f, 0.34409811764705883f, 0.42199137254901964f,
+0.391849431372549f, 0.3424064705882353f, 0.4199921568627451f,
+0.3903115882352941f, 0.34071482352941174f, 0.41799294117647057f,
+0.38877374509803925f, 0.33902317647058827f, 0.41599372549019614f,
+0.38723590196078433f, 0.3373315294117647f, 0.4139945098039216f,
+0.3856980588235294f, 0.3356398823529412f, 0.41199529411764707f,
+0.3841602156862745f, 0.33394823529411766f, 0.40999607843137253f,
+0.3826223725490196f, 0.33225658823529414f, 0.40799686274509805f,
+0.3810845294117647f, 0.3305649411764706f, 0.4059976470588236f,
+0.3795466862745098f, 0.32887329411764704f, 0.40399843137254904f,
+0.3780088431372549f, 0.3271816470588235f, 0.4019992156862745f,
+0.376471f, 0.32549f, 0.4f,
+0.37477931372549017f, 0.3236445490196078f, 0.3980007843137255f,
+0.3730876274509804f, 0.3217990980392157f, 0.396001568627451f,
+0.37139594117647057f, 0.3199536470588235f, 0.39400235294117647f,
+0.3697042549019608f, 0.3181081960784314f, 0.392003137254902f,
+0.36801256862745096f, 0.3162627450980392f, 0.39000392156862745f,
+0.3663208823529412f, 0.3144172941176471f, 0.388004705882353f,
+0.36462919607843136f, 0.3125718431372549f, 0.38600549019607844f,
+0.3629375098039216f, 0.3107263921568628f, 0.384006274509804f,
+0.36124582352941176f, 0.3088809411764706f, 0.3820070588235294f,
+0.35955413725490193f, 0.30703549019607845f, 0.3800078431372549f,
+0.35786245098039215f, 0.30519003921568627f, 0.3780086274509804f,
+0.3561707647058823f, 0.3033445882352941f, 0.37600941176470587f,
+0.35447907843137255f, 0.30149913725490196f, 0.3740101960784314f,
+0.3527873921568627f, 0.2996536862745098f, 0.37201098039215685f,
+0.35109570588235295f, 0.29780823529411765f, 0.3700117647058824f,
+0.3494040196078431f, 0.29596278431372547f, 0.36801254901960784f,
+0.34771233333333335f, 0.29411733333333334f, 0.36601333333333336f,
+0.3460206470588235f, 0.29227188235294116f, 0.3640141176470588f,
+0.34432896078431374f, 0.290426431372549f, 0.36201490196078434f,
+0.3426372745098039f, 0.28858098039215685f, 0.3600156862745098f,
+0.3409455882352941f, 0.2867355294117647f, 0.35801647058823527f,
+0.3392539019607843f, 0.28489007843137254f, 0.3560172549019608f,
+0.3375622156862745f, 0.28304462745098036f, 0.35401803921568625f,
+0.33587052941176476f, 0.2811991764705883f, 0.35201882352941183f,
+0.3341788431372549f, 0.27935372549019605f, 0.35001960784313724f,
+0.33192329411764704f, 0.27689315686274507f, 0.34748212418300656f,
+0.3291038823529412f, 0.2738174705882353f, 0.3444063725490196f,
+0.3262844705882353f, 0.2707417843137255f, 0.3413306209150327f,
+0.3234650588235294f, 0.26766609803921565f, 0.33825486928104576f,
+0.32064564705882354f, 0.2645904117647059f, 0.3351791176470588f,
+0.31782623529411763f, 0.26151472549019605f, 0.3321033660130719f,
+0.31500682352941173f, 0.25843903921568623f, 0.32902761437908495f,
+0.3121874117647059f, 0.25536335294117646f, 0.325951862745098f,
+0.309368f, 0.25228766666666663f, 0.3228761111111111f,
+0.3065485882352941f, 0.2492119803921568f, 0.31980035947712415f,
+0.3037291764705882f, 0.246136294117647f, 0.3167246078431372f,
+0.3009097647058823f, 0.2430606078431372f, 0.3136488562091503f,
+0.2980903529411764f, 0.2399849215686274f, 0.31057310457516335f,
+0.29527094117647057f, 0.2369092352941176f, 0.3074973529411764f,
+0.2924515294117648f, 0.23383354901960787f, 0.30442160130718954f,
+0.28927327450980383f, 0.23052715686274503f, 0.3009613725490195f,
+0.28465964705882346f, 0.22629794117647054f, 0.29596323529411755f,
+0.2800460196078431f, 0.22206872549019602f, 0.2909650980392156f,
+0.2754323921568627f, 0.21783950980392153f, 0.28596696078431366f,
+0.2708187647058823f, 0.213610294117647f, 0.2809688235294117f,
+0.2662051372549019f, 0.20938107843137252f, 0.2759706862745097f,
+0.26159150980392154f, 0.20515186274509803f, 0.27097254901960777f,
+0.25697788235294117f, 0.2009226470588235f, 0.2659744117647058f,
+0.25236425490196074f, 0.19669343137254902f, 0.2609762745098039f,
+0.24775062745098037f, 0.19246421568627453f, 0.2559781372549019f,
+0.243137f, 0.188235f, 0.25098f
+}
+, 
+{
+0.25098f, 0.015686f, 0.168627f,
+0.2474429019607843f, 0.017070078431372548f, 0.1706262549019608f,
+0.2439058039215686f, 0.018454156862745097f, 0.17262550980392158f,
+0.24036870588235293f, 0.019838235294117646f, 0.17462476470588234f,
+0.23683160784313723f, 0.021222313725490195f, 0.17662401960784313f,
+0.23329450980392155f, 0.022606392156862744f, 0.17862327450980392f,
+0.22975741176470588f, 0.023990470588235294f, 0.1806225294117647f,
+0.22622031372549017f, 0.025374549019607843f, 0.1826217843137255f,
+0.2226832156862745f, 0.026758627450980392f, 0.18462103921568626f,
+0.21914611764705882f, 0.028142705882352938f, 0.18662029411764705f,
+0.21560901960784312f, 0.02952678431372549f, 0.18861954901960784f,
+0.21207192156862745f, 0.030910862745098036f, 0.19061880392156863f,
+0.20853482352941177f, 0.03229494117647058f, 0.19261805882352942f,
+0.20499772549019607f, 0.033679019607843134f, 0.19461731372549018f,
+0.2014606274509804f, 0.03506309803921569f, 0.19661656862745097f,
+0.1979235294117647f, 0.03644717647058823f, 0.19861582352941176f,
+0.19438643137254902f, 0.03783125490196078f, 0.20061507843137255f,
+0.19084933333333334f, 0.03921533333333333f, 0.20261433333333334f,
+0.18731223529411764f, 0.040599411764705884f, 0.20461358823529413f,
+0.18377513725490197f, 0.04198349019607843f, 0.2066128431372549f,
+0.18023803921568626f, 0.043367568627450975f, 0.20861209803921568f,
+0.17670094117647062f, 0.04475164705882352f, 0.21061135294117647f,
+0.1731638431372549f, 0.04613572549019608f, 0.21261060784313726f,
+0.1696267450980392f, 0.047519803921568626f, 0.21460986274509802f,
+0.16608964705882354f, 0.04890388235294117f, 0.2166091176470588f,
+0.16255254901960786f, 0.050287960784313725f, 0.2186083725490196f,
+0.16132225490196078f, 0.05205650980392157f, 0.22122276470588234f,
+0.16239876470588235f, 0.054209529411764705f, 0.22445229411764706f,
+0.16347527450980393f, 0.05636254901960784f, 0.22768182352941177f,
+0.1645517843137255f, 0.05851556862745098f, 0.23091135294117646f,
+0.16562829411764707f, 0.06066858823529411f, 0.23414088235294117f,
+0.16670480392156864f, 0.06282160784313726f, 0.23737041176470589f,
+0.1677813137254902f, 0.06497462745098039f, 0.24059994117647057f,
+0.16885782352941175f, 0.06712764705882351f, 0.24382947058823526f,
+0.16993433333333335f, 0.06928066666666666f, 0.247059f,
+0.17101084313725493f, 0.0714336862745098f, 0.2502885294117647f,
+0.17208735294117647f, 0.07358670588235294f, 0.2535180588235294f,
+0.17316386274509804f, 0.07573972549019606f, 0.2567475882352941f,
+0.1742403725490196f, 0.07789274509803921f, 0.25997711764705883f,
+0.17531688235294118f, 0.08004576470588236f, 0.2632066470588235f,
+0.17639339215686275f, 0.08219878431372549f, 0.2664361764705882f,
+0.17746990196078433f, 0.08435180392156862f, 0.2696657058823529f,
+0.1785464117647059f, 0.08650482352941177f, 0.27289523529411763f,
+0.17962292156862747f, 0.0886578431372549f, 0.2761247647058823f,
+0.18069943137254904f, 0.09081086274509803f, 0.27935429411764706f,
+0.18177594117647058f, 0.09296388235294117f, 0.2825838235294117f,
+0.18285245098039216f, 0.09511690196078432f, 0.2858133529411765f,
+0.18392896078431373f, 0.09726992156862746f, 0.2890428823529412f,
+0.1850054705882353f, 0.09942294117647059f, 0.29227241176470586f,
+0.18608198039215687f, 0.10157596078431372f, 0.29550194117647055f,
+0.18715849019607844f, 0.10372898039215686f, 0.2987314705882353f,
+0.188235f, 0.105882f, 0.301961f,
+0.18931150980392158f, 0.10803501960784315f, 0.305344294117647f,
+0.19038801960784313f, 0.11018803921568626f, 0.30872758823529406f,
+0.19146452941176473f, 0.11234105882352942f, 0.31211088235294115f,
+0.19254103921568627f, 0.11449407843137255f, 0.3154941764705882f,
+0.19361754901960784f, 0.11664709803921569f, 0.3188774705882353f,
+0.19469405882352941f, 0.1188001176470588f, 0.32226076470588233f,
+0.19577056862745099f, 0.12095313725490196f, 0.3256440588235294f,
+0.19684707843137256f, 0.1231061568627451f, 0.32902735294117647f,
+0.19792358823529413f, 0.12525917647058823f, 0.3324106470588235f,
+0.19900009803921567f, 0.12741219607843135f, 0.33579394117647055f,
+0.20007660784313724f, 0.12956521568627452f, 0.33917723529411764f,
+0.20115311764705882f, 0.13171823529411766f, 0.3425605294117647f,
+0.2022296274509804f, 0.13387125490196078f, 0.3459438235294118f,
+0.20330613725490196f, 0.13602427450980392f, 0.3493271176470588f,
+0.2043826470588235f, 0.13817729411764704f, 0.35271041176470586f,
+0.2054591568627451f, 0.1403303137254902f, 0.35609370588235295f,
+0.20653566666666667f, 0.14248333333333335f, 0.359477f,
+0.20761217647058824f, 0.1446363529411765f, 0.36286029411764703f,
+0.2086886862745098f, 0.1467893725490196f, 0.3662435882352941f,
+0.20976519607843136f, 0.14894239215686275f, 0.36962688235294117f,
+0.21084170588235293f, 0.1510954117647059f, 0.3730101764705882f,
+0.2119182156862745f, 0.15324843137254904f, 0.3763934705882353f,
+0.21299472549019605f, 0.15540145098039215f, 0.3797767647058823f,
+0.21407123529411765f, 0.1575544705882353f, 0.38316005882352944f,
+0.2151477450980392f, 0.15970749019607844f, 0.3865433529411765f,
+0.21691629411764704f, 0.1620911960784314f, 0.3913876470588235f,
+0.21937688235294117f, 0.16470558823529413f, 0.3976929411764706f,
+0.22183747058823527f, 0.16731998039215687f, 0.40399823529411766f,
+0.2242980588235294f, 0.1699343725490196f, 0.4103035294117647f,
+0.2267586470588235f, 0.17254876470588237f, 0.41660882352941175f,
+0.2292192352941176f, 0.17516315686274506f, 0.4229141176470587f,
+0.23167982352941174f, 0.17777754901960785f, 0.42921941176470585f,
+0.23414041176470587f, 0.1803919411764706f, 0.4355247058823529f,
+0.23660099999999998f, 0.18300633333333333f, 0.44183f,
+0.2390615882352941f, 0.1856207254901961f, 0.448135294117647f,
+0.2415221764705882f, 0.18823511764705883f, 0.4544405882352941f,
+0.24398276470588234f, 0.19084950980392157f, 0.46074588235294117f,
+0.24644335294117645f, 0.1934639019607843f, 0.4670511764705882f,
+0.24890394117647055f, 0.19607829411764702f, 0.47335647058823516f,
+0.2513645294117647f, 0.1986926862745098f, 0.4796617647058823f,
+0.2538251176470588f, 0.20130707843137255f, 0.48596705882352936f,
+0.25628570588235294f, 0.2039214705882353f, 0.49227235294117644f,
+0.25874629411764705f, 0.20653586274509803f, 0.4985776470588235f,
+0.26120688235294115f, 0.20915025490196076f, 0.5048829411764706f,
+0.26366747058823525f, 0.21176464705882353f, 0.5111882352941176f,
+0.2661280588235294f, 0.21437903921568627f, 0.5174935294117646f,
+0.26858864705882346f, 0.21699343137254898f, 0.5237988235294117f,
+0.2710492352941176f, 0.21960782352941174f, 0.5301041176470588f,
+0.2735098235294117f, 0.22222221568627448f, 0.5364094117647058f,
+0.2759704117647059f, 0.22483660784313725f, 0.5427147058823529f,
+0.278431f, 0.227451f, 0.54902f,
+0.2801226862745098f, 0.22975782352941176f, 0.5550176862745098f,
+0.2818143725490196f, 0.23206464705882351f, 0.5610153725490196f,
+0.2835060588235294f, 0.2343714705882353f, 0.5670130588235294f,
+0.2851977450980392f, 0.23667829411764701f, 0.5730107450980391f,
+0.286889431372549f, 0.23898511764705882f, 0.579008431372549f,
+0.2885811176470588f, 0.24129194117647057f, 0.5850061176470588f,
+0.2902728039215686f, 0.24359876470588235f, 0.5910038039215686f,
+0.2919644901960784f, 0.2459055882352941f, 0.5970014901960784f,
+0.2936561764705882f, 0.24821241176470588f, 0.6029991764705882f,
+0.29534786274509806f, 0.2505192352941176f, 0.608996862745098f,
+0.29703954901960783f, 0.25282605882352943f, 0.6149945490196078f,
+0.2987312352941176f, 0.2551328823529411f, 0.6209922352941175f,
+0.30042292156862743f, 0.25743970588235293f, 0.6269899215686274f,
+0.30211460784313726f, 0.2597465294117647f, 0.6329876078431372f,
+0.30380629411764704f, 0.2620533529411765f, 0.638985294117647f,
+0.30549798039215686f, 0.26436017647058824f, 0.6449829803921568f,
+0.30718966666666664f, 0.266667f, 0.6509806666666667f,
+0.30888135294117647f, 0.2689738235294118f, 0.6569783529411765f,
+0.31057303921568624f, 0.27128064705882354f, 0.6629760392156863f,
+0.312264725490196f, 0.27358747058823524f, 0.668973725490196f,
+0.3139564117647059f, 0.27589429411764704f, 0.6749714117647059f,
+0.3156480980392157f, 0.27820111764705885f, 0.6809690980392157f,
+0.3173397843137255f, 0.2805079411764706f, 0.6869667843137255f,
+0.3190314705882353f, 0.28281476470588235f, 0.6929644705882352f,
+0.3207231568627451f, 0.2851215882352941f, 0.6989621568627451f,
+0.32218413725490197f, 0.28758217647058826f, 0.7038833333333333f,
+0.32341441176470587f, 0.2901965294117647f, 0.7077279999999999f,
+0.3246446862745098f, 0.29281088235294117f, 0.7115726666666666f,
+0.3258749607843137f, 0.2954252352941176f, 0.7154173333333333f,
+0.3271052352941176f, 0.29803958823529403f, 0.7192619999999998f,
+0.32833550980392157f, 0.3006539411764706f, 0.7231066666666667f,
+0.32956578431372546f, 0.30326829411764705f, 0.7269513333333333f,
+0.3307960588235294f, 0.3058826470588235f, 0.730796f,
+0.3320263333333333f, 0.308497f, 0.7346406666666666f,
+0.33325660784313726f, 0.3111113529411765f, 0.7384853333333333f,
+0.33448688235294116f, 0.31372570588235293f, 0.7423299999999999f,
+0.3357171568627451f, 0.3163400588235294f, 0.7461746666666667f,
+0.336947431372549f, 0.3189544117647059f, 0.7500193333333334f,
+0.33817770588235296f, 0.32156876470588236f, 0.753864f,
+0.33940798039215686f, 0.3241831176470588f, 0.7577086666666667f,
+0.34063825490196076f, 0.3267974705882353f, 0.7615533333333333f,
+0.3418685294117647f, 0.3294118235294118f, 0.765398f,
+0.3430988039215686f, 0.33202617647058824f, 0.7692426666666667f,
+0.34432907843137256f, 0.3346405294117647f, 0.7730873333333333f,
+0.34555935294117646f, 0.33725488235294115f, 0.776932f,
+0.34678962745098035f, 0.33986923529411756f, 0.7807766666666666f,
+0.3480199019607843f, 0.3424835882352941f, 0.7846213333333334f,
+0.34925017647058826f, 0.3450979411764706f, 0.788466f,
+0.35048045098039216f, 0.34771229411764704f, 0.7923106666666667f,
+0.3517107254901961f, 0.35032664705882355f, 0.7961553333333333f,
+0.352941f, 0.352941f, 0.8f,
+0.35524782352941175f, 0.35709327450980394f, 0.8030757254901961f,
+0.35755464705882356f, 0.36124554901960787f, 0.8061514509803922f,
+0.3598614705882353f, 0.36539782352941175f, 0.8092271764705883f,
+0.36216829411764706f, 0.3695500980392157f, 0.8123029019607844f,
+0.3644751176470588f, 0.3737023725490196f, 0.8153786274509804f,
+0.3667819411764706f, 0.37785464705882355f, 0.8184543529411765f,
+0.36908876470588237f, 0.3820069215686275f, 0.8215300784313726f,
+0.3713955882352941f, 0.38615919607843135f, 0.8246058039215687f,
+0.37370241176470587f, 0.3903114705882353f, 0.8276815294117648f,
+0.37600923529411767f, 0.3944637450980392f, 0.8307572549019608f,
+0.37831605882352937f, 0.39861601960784304f, 0.8338329803921568f,
+0.38062288235294117f, 0.4027682941176471f, 0.836908705882353f,
+0.3829297058823529f, 0.40692056862745096f, 0.839984431372549f,
+0.3852365294117647f, 0.4110728431372549f, 0.8430601568627452f,
+0.3875433529411765f, 0.41522511764705883f, 0.8461358823529411f,
+0.38985017647058823f, 0.41937739215686276f, 0.8492116078431372f,
+0.392157f, 0.4235296666666667f, 0.8522873333333333f,
+0.3944638235294118f, 0.4276819411764706f, 0.8553630588235294f,
+0.39677064705882353f, 0.4318342156862745f, 0.8584387843137254f,
+0.3990774705882353f, 0.43598649019607844f, 0.8615145098039215f,
+0.40138429411764703f, 0.44013876470588237f, 0.8645902352941176f,
+0.40369111764705884f, 0.4442910392156863f, 0.8676659607843137f,
+0.4059979411764706f, 0.4484433137254902f, 0.8707416862745098f,
+0.40830476470588234f, 0.4525955882352941f, 0.8738174117647058f,
+0.4106115882352941f, 0.45674786274509804f, 0.8768931372549019f,
+0.413610431372549f, 0.4612845882352941f, 0.8788923725490195f,
+0.41730129411764694f, 0.4662057647058822f, 0.8798151176470588f,
+0.4209921568627451f, 0.4711269411764706f, 0.880737862745098f,
+0.42468301960784316f, 0.47604811764705884f, 0.8816606078431373f,
+0.42837388235294116f, 0.48096929411764705f, 0.8825833529411764f,
+0.4320647450980392f, 0.4858904705882353f, 0.8835060980392156f,
+0.43575560784313727f, 0.4908116470588235f, 0.8844288431372549f,
+0.4394464705882353f, 0.4957328235294118f, 0.8853515882352941f,
+0.4431373333333333f, 0.500654f, 0.8862743333333333f,
+0.4468281960784314f, 0.5055751764705883f, 0.8871970784313725f,
+0.45051905882352944f, 0.5104963529411765f, 0.8881198235294118f,
+0.4542099215686275f, 0.5154175294117647f, 0.889042568627451f,
+0.4579007843137255f, 0.520338705882353f, 0.8899653137254901f,
+0.46159164705882355f, 0.5252598823529412f, 0.8908880588235294f,
+0.4652825098039216f, 0.5301810588235294f, 0.8918108039215686f,
+0.4689733725490196f, 0.5351022352941176f, 0.8927335490196079f,
+0.47266423529411766f, 0.5400234117647059f, 0.893656294117647f,
+0.4763550980392156f, 0.544944588235294f, 0.8945790392156863f,
+0.4800459607843138f, 0.5498657647058823f, 0.8955017843137255f,
+0.48373682352941183f, 0.5547869411764705f, 0.8964245294117648f,
+0.48742768627450983f, 0.5597081176470589f, 0.8973472745098039f,
+0.4911185490196079f, 0.5646292941176471f, 0.8982700196078431f,
+0.49480941176470594f, 0.5695504705882353f, 0.8991927647058824f,
+0.49850027450980394f, 0.5744716470588235f, 0.9001155098039215f,
+0.502191137254902f, 0.5793928235294118f, 0.9010382549019608f,
+0.505882f, 0.584314f, 0.901961f,
+0.5083425882352942f, 0.5883124705882353f, 0.9027299411764707f,
+0.5108031764705883f, 0.5923109411764705f, 0.9034988823529412f,
+0.5132637647058824f, 0.5963094117647059f, 0.9042678235294118f,
+0.5157243529411765f, 0.6003078823529412f, 0.9050367647058823f,
+0.5181849411764706f, 0.6043063529411765f, 0.905805705882353f,
+0.5206455294117648f, 0.6083048235294117f, 0.9065746470588235f,
+0.5231061176470588f, 0.6123032941176471f, 0.9073435882352942f,
+0.5255667058823529f, 0.6163017647058823f, 0.9081125294117647f,
+0.528027294117647f, 0.6203002352941176f, 0.9088814705882353f,
+0.5304878823529412f, 0.6242987058823529f, 0.9096504117647058f,
+0.5329484705882354f, 0.6282971764705882f, 0.9104193529411765f,
+0.5354090588235294f, 0.6322956470588235f, 0.911188294117647f,
+0.5378696470588236f, 0.6362941176470588f, 0.9119572352941177f,
+0.5403302352941177f, 0.6402925882352941f, 0.9127261764705882f,
+0.5427908235294118f, 0.6442910588235294f, 0.9134951176470588f,
+0.5452514117647059f, 0.6482895294117647f, 0.9142640588235293f,
+0.547712f, 0.652288f, 0.915033f,
+0.5501725882352941f, 0.6562864705882353f, 0.9158019411764706f,
+0.5526331764705883f, 0.6602849411764705f, 0.9165708823529412f,
+0.5550937647058823f, 0.6642834117647058f, 0.9173398235294118f,
+0.5575543529411765f, 0.6682818823529412f, 0.9181087647058823f,
+0.5600149411764705f, 0.6722803529411765f, 0.918877705882353f,
+0.5624755294117647f, 0.6762788235294117f, 0.9196466470588235f,
+0.5649361176470588f, 0.6802772941176469f, 0.920415588235294f,
+0.5673967058823529f, 0.6842757647058824f, 0.9211845294117647f,
+0.5717796470588236f, 0.6888893725490196f, 0.922491705882353f,
+0.5780849411764706f, 0.6941181176470588f, 0.9243371176470588f,
+0.5843902352941176f, 0.699346862745098f, 0.9261825294117647f,
+0.5906955294117647f, 0.7045756078431372f, 0.9280279411764706f,
+0.5970008235294118f, 0.7098043529411765f, 0.9298733529411765f,
+0.6033061176470588f, 0.7150330980392157f, 0.9317187647058823f,
+0.6096114117647059f, 0.7202618431372548f, 0.9335641764705882f,
+0.6159167058823529f, 0.7254905882352941f, 0.9354095882352941f,
+0.6222219999999999f, 0.7307193333333333f, 0.937255f,
+0.6285272941176471f, 0.7359480784313726f, 0.9391004117647058f,
+0.6348325882352941f, 0.7411768235294117f, 0.9409458235294118f,
+0.6411378823529411f, 0.746405568627451f, 0.9427912352941177f,
+0.6474431764705882f, 0.7516343137254902f, 0.9446366470588236f,
+0.6537484705882353f, 0.7568630588235294f, 0.9464820588235294f,
+0.6600537647058822f, 0.7620918039215685f, 0.9483274705882352f,
+0.6663590588235294f, 0.7673205490196078f, 0.9501728823529412f,
+0.6726643529411764f, 0.7725492941176471f, 0.952018294117647f,
+0.6789696470588235f, 0.7777780392156862f, 0.9538637058823529f,
+0.6852749411764706f, 0.7830067843137255f, 0.9557091176470588f,
+0.6915802352941176f, 0.7882355294117647f, 0.9575545294117647f,
+0.6978855294117647f, 0.793464274509804f, 0.9593999411764706f,
+0.7041908235294118f, 0.7986930196078431f, 0.9612453529411765f,
+0.7104961176470588f, 0.8039217647058824f, 0.9630907647058824f,
+0.7168014117647058f, 0.8091505098039216f, 0.9649361764705883f,
+0.7231067058823529f, 0.8143792549019608f, 0.9667815882352941f,
+0.729412f, 0.819608f, 0.968627f
+}
+, 
+{
+0.988235f, 0.976471f, 0.909804f,
+0.9866971176470588f, 0.9724724901960784f, 0.8968858823529411f,
+0.9851592352941176f, 0.9684739803921568f, 0.8839677647058823f,
+0.9836213529411765f, 0.9644754705882352f, 0.8710496470588235f,
+0.9820834705882353f, 0.9604769607843138f, 0.8581315294117646f,
+0.9805455882352941f, 0.9564784509803922f, 0.8452134117647059f,
+0.979007705882353f, 0.9524799411764706f, 0.832295294117647f,
+0.9774698235294118f, 0.948481431372549f, 0.8193771764705882f,
+0.9759319411764706f, 0.9444829215686275f, 0.8064590588235294f,
+0.9743940588235294f, 0.9404844117647059f, 0.7935409411764706f,
+0.9728561764705882f, 0.9364859019607843f, 0.7806228235294118f,
+0.971318294117647f, 0.9324873921568628f, 0.767704705882353f,
+0.9697804117647059f, 0.9284888823529412f, 0.7547865882352942f,
+0.9682425490196078f, 0.9238752352941176f, 0.7420222549019608f,
+0.9667047450980393f, 0.9174161764705883f, 0.729719274509804f,
+0.9651669411764706f, 0.9109571176470589f, 0.7174162941176471f,
+0.963629137254902f, 0.9044980588235294f, 0.7051133137254902f,
+0.9620913333333333f, 0.898039f, 0.6928103333333333f,
+0.9605535294117648f, 0.8915799411764707f, 0.6805073529411765f,
+0.9590157254901961f, 0.8851208823529412f, 0.6682043725490197f,
+0.9574779215686274f, 0.8786618235294118f, 0.6559013921568627f,
+0.9559401176470588f, 0.8722027647058824f, 0.6435984117647059f,
+0.9544023137254902f, 0.8657437058823529f, 0.631295431372549f,
+0.9528645098039216f, 0.8592846470588236f, 0.6189924509803921f,
+0.9513267058823529f, 0.8528255882352942f, 0.6066894705882353f,
+0.9497889019607842f, 0.8463665294117647f, 0.5943864901960784f,
+0.9487123921568628f, 0.8403688431372549f, 0.5854668431372548f,
+0.9480971764705882f, 0.8348325294117648f, 0.5799305294117647f,
+0.9474819607843137f, 0.8292962156862745f, 0.5743942156862745f,
+0.9468667450980393f, 0.8237599019607843f, 0.5688579019607842f,
+0.9462515294117647f, 0.8182235882352942f, 0.5633215882352941f,
+0.9456363137254902f, 0.8126872745098039f, 0.5577852745098039f,
+0.9450210980392156f, 0.8071509607843138f, 0.5522489607843137f,
+0.9444058823529412f, 0.8016146470588236f, 0.5467126470588235f,
+0.9437906666666667f, 0.7960783333333333f, 0.5411763333333333f,
+0.9431754509803921f, 0.7905420196078432f, 0.5356400196078431f,
+0.9425602352941177f, 0.785005705882353f, 0.5301037058823529f,
+0.9419450196078432f, 0.7794693921568628f, 0.5245673921568628f,
+0.9413298039215686f, 0.7739330784313726f, 0.5190310784313725f,
+0.940484f, 0.7674740588235295f, 0.5134947647058823f,
+0.9395613333333334f, 0.7607074705882353f, 0.5079584509803922f,
+0.9386386666666667f, 0.7539408823529412f, 0.5024221372549019f,
+0.937716f, 0.7471742941176471f, 0.49688582352941174f,
+0.9367933333333334f, 0.7404077058823529f, 0.49134950980392156f,
+0.9358706666666666f, 0.7336411176470589f, 0.4858131960784313f,
+0.934948f, 0.7268745294117648f, 0.4802768823529412f,
+0.9340253333333334f, 0.7201079411764706f, 0.47474056862745095f,
+0.9331026666666666f, 0.7133413529411765f, 0.46920425490196077f,
+0.93218f, 0.7065747647058823f, 0.4636679411764706f,
+0.9312573333333334f, 0.6998081764705882f, 0.4581316274509804f,
+0.9303346666666666f, 0.693041588235294f, 0.45259531372549017f,
+0.929412f, 0.686275f, 0.447059f,
+0.9278741176470588f, 0.6807386078431372f, 0.4427529607843137f,
+0.9263362352941177f, 0.6752022156862745f, 0.43844692156862747f,
+0.9247983529411765f, 0.6696658235294117f, 0.4341408823529412f,
+0.9232604705882352f, 0.664129431372549f, 0.4298348431372549f,
+0.9217225882352941f, 0.6585930392156862f, 0.4255288039215686f,
+0.9201847058823529f, 0.6530566470588235f, 0.4212227647058824f,
+0.9186468235294117f, 0.6475202549019607f, 0.41691672549019604f,
+0.9171089411764706f, 0.641983862745098f, 0.4126106862745098f,
+0.9155710588235294f, 0.6364474705882353f, 0.4083046470588235f,
+0.9140331764705882f, 0.6309110784313725f, 0.40399860784313724f,
+0.912495294117647f, 0.6253746862745098f, 0.39969256862745095f,
+0.9109574117647058f, 0.619838294117647f, 0.39538652941176466f,
+0.9096502156862745f, 0.6145326078431372f, 0.3915418431372549f,
+0.9090350784313725f, 0.6099190392156862f, 0.38908121568627446f,
+0.9084199411764705f, 0.6053054705882354f, 0.38662058823529416f,
+0.9078048039215686f, 0.6006919019607843f, 0.3841599607843137f,
+0.9071896666666667f, 0.5960783333333333f, 0.38169933333333333f,
+0.9065745294117646f, 0.5914647647058823f, 0.3792387058823529f,
+0.9059593921568627f, 0.5868511960784314f, 0.3767780784313725f,
+0.9053442549019608f, 0.5822376274509804f, 0.37431745098039215f,
+0.9047291176470588f, 0.5776240588235294f, 0.37185682352941174f,
+0.9041139803921568f, 0.5730104901960784f, 0.3693961960784314f,
+0.9034988431372549f, 0.5683969215686275f, 0.366935568627451f,
+0.9028837058823529f, 0.5637833529411764f, 0.36447494117647056f,
+0.902268568627451f, 0.5591697843137254f, 0.3620143137254902f,
+0.9014996274509804f, 0.5536334705882353f, 0.35986129411764706f,
+0.9005768823529412f, 0.5471744117647058f, 0.3580158823529412f,
+0.899654137254902f, 0.5407153529411765f, 0.3561704705882353f,
+0.8987313921568627f, 0.5342562941176471f, 0.35432505882352944f,
+0.8978086470588236f, 0.5277972352941176f, 0.3524796470588235f,
+0.8968859019607843f, 0.5213381764705883f, 0.3506342352941177f,
+0.8959631568627451f, 0.5148791176470588f, 0.34878882352941176f,
+0.8950404117647058f, 0.5084200588235294f, 0.3469434117647059f,
+0.8941176666666667f, 0.501961f, 0.345098f,
+0.8931949215686275f, 0.49550194117647056f, 0.34325258823529414f,
+0.8922721764705882f, 0.4890428823529412f, 0.34140717647058827f,
+0.891349431372549f, 0.48258382352941176f, 0.3395617647058824f,
+0.8904266862745098f, 0.47612476470588233f, 0.3377163529411765f,
+0.8895039411764706f, 0.4703577647058824f, 0.3356402352941177f,
+0.8885811960784313f, 0.4648214509803921f, 0.33348721568627454f,
+0.8876584509803921f, 0.45928513725490194f, 0.3313341960784314f,
+0.886735705882353f, 0.45374882352941176f, 0.32918117647058825f,
+0.8858129607843137f, 0.4482125098039216f, 0.3270281568627451f,
+0.8848902156862745f, 0.44267619607843134f, 0.32487513725490197f,
+0.8839674705882352f, 0.43713988235294116f, 0.3227221176470588f,
+0.8830447254901961f, 0.431603568627451f, 0.32056909803921574f,
+0.8821219803921568f, 0.42606725490196085f, 0.3184160784313726f,
+0.8811992352941176f, 0.4205309411764706f, 0.31626305882352945f,
+0.8802764901960783f, 0.41499462745098037f, 0.3141100392156863f,
+0.8793537450980392f, 0.4094583137254902f, 0.31195701960784317f,
+0.878431f, 0.403922f, 0.309804f,
+0.8753553137254901f, 0.39807807843137255f, 0.305805568627451f,
+0.8722796274509803f, 0.3922341568627451f, 0.301807137254902f,
+0.8692039411764706f, 0.38639023529411765f, 0.297808705882353f,
+0.8661282549019608f, 0.3805463137254903f, 0.293810274509804f,
+0.8630525686274509f, 0.37470239215686274f, 0.28981184313725494f,
+0.8599768823529411f, 0.3688584705882353f, 0.28581341176470587f,
+0.8569011960784313f, 0.36301454901960784f, 0.28181498039215686f,
+0.8538255098039216f, 0.3571706274509804f, 0.27781654901960784f,
+0.8507498235294118f, 0.35132670588235293f, 0.2738181176470588f,
+0.8476741372549019f, 0.3454827843137255f, 0.2698196862745098f,
+0.8445984509803921f, 0.339638862745098f, 0.2658212549019608f,
+0.8415227647058824f, 0.3337949411764707f, 0.26182282352941183f,
+0.8384470588235293f, 0.3284123725490196f, 0.2584395294117647f,
+0.835371294117647f, 0.32441386274509804f, 0.2569016470588235f,
+0.8322955294117647f, 0.32041535294117646f, 0.25536376470588235f,
+0.8292197647058823f, 0.3164168431372549f, 0.2538258823529412f,
+0.826144f, 0.3124183333333333f, 0.252288f,
+0.8230682352941177f, 0.3084198235294118f, 0.25075011764705885f,
+0.8199924705882353f, 0.3044213137254902f, 0.24921223529411765f,
+0.816916705882353f, 0.30042280392156867f, 0.2476743529411765f,
+0.8138409411764707f, 0.29642429411764704f, 0.2461364705882353f,
+0.8107651764705882f, 0.29242578431372546f, 0.24459858823529412f,
+0.8076894117647059f, 0.2884272745098039f, 0.24306070588235296f,
+0.8046136470588235f, 0.28442876470588235f, 0.2415228235294118f,
+0.8015378823529412f, 0.2804302549019608f, 0.2399849411764706f,
+0.7980007843137256f, 0.2758166470588235f, 0.23813949019607844f,
+0.7940023529411765f, 0.27058794117647056f, 0.2359864705882353f,
+0.7900039215686275f, 0.26535923529411765f, 0.23383345098039215f,
+0.7860054901960785f, 0.2601305294117647f, 0.23168043137254904f,
+0.7820070588235296f, 0.25490182352941193f, 0.22952741176470595f,
+0.7780086274509804f, 0.24967311764705882f, 0.22737439215686275f,
+0.7740101960784314f, 0.24444441176470588f, 0.2252213725490196f,
+0.7700117647058824f, 0.23921570588235294f, 0.2230683529411765f,
+0.7660133333333333f, 0.233987f, 0.22091533333333335f,
+0.7620149019607844f, 0.22875829411764706f, 0.2187623137254902f,
+0.7580164705882353f, 0.22352958823529412f, 0.21660929411764707f,
+0.7540180392156863f, 0.21830088235294118f, 0.21445627450980392f,
+0.7500196078431373f, 0.21307217647058824f, 0.21230325490196078f,
+0.7462518235294118f, 0.20991952941176473f, 0.21153429411764707f,
+0.7425609215686274f, 0.20745890196078431f, 0.2112266862745098f,
+0.7388700196078432f, 0.20499827450980393f, 0.21091907843137256f,
+0.7351791176470588f, 0.20253764705882354f, 0.2106114705882353f,
+0.7314882156862745f, 0.20007701960784313f, 0.21030386274509805f,
+0.7277973137254902f, 0.19761639215686275f, 0.20999625490196078f,
+0.7241064117647059f, 0.19515576470588236f, 0.20968864705882354f,
+0.7204155098039217f, 0.19269513725490203f, 0.2093810392156863f,
+0.7167246078431372f, 0.19023450980392156f, 0.20907343137254902f,
+0.7130337058823529f, 0.18777388235294118f, 0.20876582352941175f,
+0.7093428039215686f, 0.1853132549019608f, 0.2084582156862745f,
+0.7056519019607843f, 0.18285262745098038f, 0.20815060784313727f,
+0.701961f, 0.180392f, 0.207843f,
+0.6985777058823529f, 0.17916172549019607f, 0.207843f,
+0.6951944117647059f, 0.17793145098039215f, 0.207843f,
+0.6918111176470588f, 0.17670117647058822f, 0.207843f,
+0.6884278235294117f, 0.1754709019607843f, 0.207843f,
+0.6850445294117646f, 0.1742406274509804f, 0.207843f,
+0.6816612352941176f, 0.17301035294117648f, 0.207843f,
+0.6782779411764706f, 0.17178007843137255f, 0.207843f,
+0.6748946470588235f, 0.17054980392156863f, 0.207843f,
+0.6715113529411765f, 0.1693195294117647f, 0.207843f,
+0.6681280588235293f, 0.16808925490196078f, 0.207843f,
+0.6647447647058824f, 0.16685898039215688f, 0.207843f,
+0.6613614705882352f, 0.16562870588235293f, 0.207843f,
+0.6576705882352941f, 0.1640908431372549f, 0.2074585294117647f,
+0.6530569411764705f, 0.1616302156862745f, 0.20592064705882354f,
+0.648443294117647f, 0.15916958823529412f, 0.20438276470588235f,
+0.6438296470588235f, 0.15670896078431373f, 0.20284488235294118f,
+0.639216f, 0.15424833333333332f, 0.201307f,
+0.6346023529411764f, 0.15178770588235294f, 0.19976911764705882f,
+0.6299887058823529f, 0.14932707843137255f, 0.19823123529411765f,
+0.6253750588235294f, 0.14686645098039217f, 0.19669335294117649f,
+0.6207614117647059f, 0.14440582352941175f, 0.1951554705882353f,
+0.6161477647058823f, 0.14194519607843137f, 0.19361758823529412f,
+0.6115341176470588f, 0.13948456862745098f, 0.19207970588235296f,
+0.6069204705882353f, 0.1370239411764706f, 0.1905418235294118f,
+0.6023068235294118f, 0.1345633137254902f, 0.1890039411764706f,
+0.5976931764705882f, 0.13256405882352942f, 0.18808123529411766f,
+0.5930795294117648f, 0.13102617647058828f, 0.18777370588235295f,
+0.5884658823529412f, 0.12948829411764706f, 0.18746617647058825f,
+0.5838522352941177f, 0.1279504117647059f, 0.18715864705882354f,
+0.5792385882352941f, 0.1264125294117647f, 0.18685111764705883f,
+0.5746249411764706f, 0.12487464705882353f, 0.18654358823529413f,
+0.5700112941176471f, 0.12333676470588237f, 0.18623605882352942f,
+0.5653976470588236f, 0.12179888235294119f, 0.18592852941176471f,
+0.560784f, 0.120261f, 0.185621f,
+0.5561703529411764f, 0.11872311764705883f, 0.1853134705882353f,
+0.551556705882353f, 0.11718523529411766f, 0.1850059411764706f,
+0.5469430588235294f, 0.11564735294117648f, 0.1846984117647059f,
+0.5423294117647058f, 0.1141094705882353f, 0.18439088235294118f,
+0.5388692352941177f, 0.11280229411764707f, 0.18408329411764707f,
+0.5357935490196079f, 0.11157201960784315f, 0.1837756862745098f,
+0.532717862745098f, 0.11034174509803922f, 0.18346807843137256f,
+0.5296421764705882f, 0.1091114705882353f, 0.18316047058823529f,
+0.5265664901960785f, 0.10788119607843141f, 0.18285286274509804f,
+0.5234908039215687f, 0.10665092156862746f, 0.18254525490196077f,
+0.5204151176470588f, 0.10542064705882354f, 0.18223764705882353f,
+0.517339431372549f, 0.10419037254901961f, 0.18193003921568626f,
+0.5142637450980392f, 0.10296009803921569f, 0.18162243137254902f,
+0.5111880588235294f, 0.10172982352941176f, 0.18131482352941175f,
+0.5081123725490196f, 0.10049954901960785f, 0.1810072156862745f,
+0.5050366862745098f, 0.09926927450980393f, 0.18069960784313727f,
+0.501961f, 0.098039f, 0.180392f,
+0.49703982352941173f, 0.09650111764705882f, 0.17916172549019607f,
+0.4921186470588235f, 0.09496323529411765f, 0.17793145098039215f,
+0.48719747058823526f, 0.09342535294117647f, 0.17670117647058822f,
+0.48227629411764705f, 0.09188747058823529f, 0.1754709019607843f,
+0.4773551176470588f, 0.09034958823529413f, 0.1742406274509804f,
+0.4724339411764706f, 0.08881170588235295f, 0.17301035294117648f,
+0.4675127647058823f, 0.08727382352941176f, 0.17178007843137255f,
+0.4625915882352942f, 0.08573594117647063f, 0.17054980392156865f,
+0.45767041176470585f, 0.08419805882352942f, 0.1693195294117647f,
+0.45274923529411765f, 0.08266017647058824f, 0.16808925490196078f,
+0.4478280588235294f, 0.08112229411764706f, 0.16685898039215685f,
+0.4429068823529412f, 0.07958441176470588f, 0.16562870588235293f,
+0.4388315294117647f, 0.07820033333333333f, 0.1646290980392157f,
+0.4372936470588235f, 0.07727766666666666f, 0.16432149019607842f,
+0.43575576470588234f, 0.076355f, 0.16401388235294118f,
+0.4342178823529412f, 0.07543233333333334f, 0.1637062745098039f,
+0.43268f, 0.07450966666666667f, 0.16339866666666666f,
+0.4311421176470588f, 0.073587f, 0.16309105882352942f,
+0.4296042352941176f, 0.07266433333333333f, 0.16278345098039215f,
+0.42806635294117645f, 0.07174166666666668f, 0.1624758431372549f,
+0.4265284705882353f, 0.07081900000000001f, 0.16216823529411764f,
+0.4249905882352941f, 0.06989633333333334f, 0.1618606274509804f,
+0.42345270588235295f, 0.06897366666666667f, 0.16155301960784316f,
+0.4219148235294118f, 0.06805100000000003f, 0.1612454117647059f,
+0.42037694117647056f, 0.06712833333333333f, 0.16093780392156865f,
+0.41653223529411765f, 0.06605182352941177f, 0.15986129411764707f,
+0.4103807058823529f, 0.0648214705882353f, 0.15801588235294117f,
+0.4042291764705882f, 0.06359111764705883f, 0.1561704705882353f,
+0.39807764705882354f, 0.06236076470588235f, 0.15432505882352943f,
+0.3919261176470588f, 0.061130411764705885f, 0.15247964705882353f,
+0.3857745882352941f, 0.05990005882352941f, 0.15063423529411765f,
+0.3796230588235294f, 0.05866970588235294f, 0.14878882352941175f,
+0.3734715294117647f, 0.05743935294117647f, 0.14694341176470588f,
+0.36732f, 0.056209f, 0.145098f,
+0.36116847058823526f, 0.05497864705882353f, 0.1432525882352941f,
+0.3550169411764706f, 0.05374829411764706f, 0.14140717647058823f,
+0.34886541176470587f, 0.05251794117647059f, 0.13956176470588233f,
+0.34271388235294115f, 0.05128758823529411f, 0.13771635294117646f,
+0.3374851176470588f, 0.050288f, 0.1361015882352941f,
+0.3325639411764707f, 0.04936533333333336f, 0.13456370588235297f,
+0.32764276470588233f, 0.04844266666666666f, 0.13302582352941175f,
+0.3227215882352941f, 0.04752f, 0.13148794117647059f,
+0.31780041176470586f, 0.04659733333333333f, 0.1299500588235294f,
+0.31287923529411765f, 0.04567466666666667f, 0.12841217647058822f,
+0.3079580588235294f, 0.044752f, 0.12687429411764706f,
+0.3030368823529412f, 0.04382933333333333f, 0.1253364117647059f,
+0.2981157058823529f, 0.04290666666666666f, 0.1237985294117647f,
+0.2931945294117647f, 0.041984f, 0.12226064705882353f,
+0.28827335294117645f, 0.04106133333333333f, 0.12072276470588235f,
+0.28335217647058825f, 0.04013866666666667f, 0.11918488235294118f,
+0.278431f, 0.039216f, 0.117647f
+}
+, 
+{
+0.988235f, 0.882353f, 0.843137f,
+0.9863126470588235f, 0.8738946666666667f, 0.8327564117647059f,
+0.984390294117647f, 0.8654363333333334f, 0.8223758235294117f,
+0.9824679411764706f, 0.856978f, 0.8119952352941177f,
+0.9805455882352941f, 0.8485196666666667f, 0.8016146470588236f,
+0.9786232352941177f, 0.8400613333333333f, 0.7912340588235294f,
+0.9767008823529412f, 0.831603f, 0.7808534705882353f,
+0.9747785294117647f, 0.8231446666666666f, 0.7704728823529412f,
+0.9728561764705882f, 0.8146863333333333f, 0.7600922941176471f,
+0.9709338235294118f, 0.806228f, 0.7497117058823529f,
+0.9690114705882353f, 0.7977696666666666f, 0.7393311176470588f,
+0.968627f, 0.7893114117647059f, 0.7280277843137255f,
+0.968627f, 0.7808531764705882f, 0.7164937647058823f,
+0.968627f, 0.7723949411764706f, 0.7049597450980393f,
+0.968627f, 0.763936705882353f, 0.6934257254901961f,
+0.968627f, 0.7554784705882353f, 0.681891705882353f,
+0.968627f, 0.7470202352941177f, 0.6703576862745099f,
+0.968627f, 0.7385619999999999f, 0.6588236666666667f,
+0.968627f, 0.7301037647058823f, 0.6472896470588236f,
+0.968627f, 0.7216455294117647f, 0.6357556274509805f,
+0.968627f, 0.713187294117647f, 0.6242216078431373f,
+0.9683963235294117f, 0.7065744705882353f, 0.6160324411764706f,
+0.9680118627450981f, 0.7011919215686274f, 0.6100731764705883f,
+0.9676274019607843f, 0.6958093725490196f, 0.604113911764706f,
+0.9672429411764706f, 0.6904268235294118f, 0.5981546470588236f,
+0.9668584803921568f, 0.6850442745098039f, 0.5921953823529412f,
+0.9664740196078432f, 0.6796617254901961f, 0.5862361176470589f,
+0.9660895588235294f, 0.6742791764705882f, 0.5802768529411765f,
+0.9657050980392157f, 0.6688966274509804f, 0.5743175882352942f,
+0.9653206372549019f, 0.6635140784313726f, 0.5683583235294118f,
+0.9649361764705883f, 0.6581315294117647f, 0.5623990588235295f,
+0.9645517156862745f, 0.6527489803921569f, 0.5564397941176471f,
+0.9641672549019608f, 0.647366431372549f, 0.5504805294117647f,
+0.963782794117647f, 0.6419838823529412f, 0.5445212647058825f,
+0.9633983333333334f, 0.6366013333333334f, 0.538562f,
+0.9630138725490196f, 0.6312187843137255f, 0.5326027352941177f,
+0.9626294117647058f, 0.6258362352941177f, 0.5266434705882354f,
+0.9622449509803921f, 0.6204536862745098f, 0.520684205882353f,
+0.9618604901960784f, 0.6150711372549019f, 0.5147249411764706f,
+0.9614760294117647f, 0.6096885882352941f, 0.5087656764705882f,
+0.9610915686274509f, 0.6043060392156863f, 0.502806411764706f,
+0.9606302222222222f, 0.5986159215686275f, 0.49650113071895435f,
+0.9598613333333333f, 0.5916955294117647f, 0.4888117843137255f,
+0.9590924444444444f, 0.5847751372549019f, 0.4811224379084968f,
+0.9583235555555555f, 0.5778547450980392f, 0.473433091503268f,
+0.9575546666666667f, 0.5709343529411766f, 0.4657437450980393f,
+0.9567857777777777f, 0.5640139607843138f, 0.4580543986928105f,
+0.9560168888888889f, 0.557093568627451f, 0.4503650522875817f,
+0.955248f, 0.5501731764705883f, 0.44267570588235294f,
+0.9544791111111111f, 0.5432527843137256f, 0.4349863594771243f,
+0.9537102222222222f, 0.5363323921568628f, 0.42729701307189544f,
+0.9529413333333333f, 0.529412f, 0.4196076666666667f,
+0.9521724444444444f, 0.5224916078431373f, 0.41191832026143793f,
+0.9514035555555556f, 0.5155712156862746f, 0.40422897385620926f,
+0.9506346666666666f, 0.5086508235294118f, 0.3965396274509804f,
+0.9498657777777777f, 0.501730431372549f, 0.38885028104575164f,
+0.9490968888888889f, 0.4948100392156863f, 0.3811609346405229f,
+0.9474052352941177f, 0.48904300000000006f, 0.37531705882352945f,
+0.9456110522875817f, 0.48340411111111115f, 0.36967823529411764f,
+0.9438168692810457f, 0.47776522222222223f, 0.3640394117647059f,
+0.9420226862745098f, 0.47212633333333337f, 0.35840058823529414f,
+0.9402285032679738f, 0.4664874444444445f, 0.3527617647058824f,
+0.9384343202614379f, 0.4608485555555556f, 0.3471229411764706f,
+0.9366401372549019f, 0.4552096666666667f, 0.3414841176470588f,
+0.934845954248366f, 0.4495707777777778f, 0.3358452941176471f,
+0.93305177124183f, 0.4439318888888889f, 0.3302064705882353f,
+0.9312575882352941f, 0.4382930000000001f, 0.3245676470588236f,
+0.9294634052287581f, 0.43265411111111113f, 0.31892882352941176f,
+0.9276692222222223f, 0.42701522222222227f, 0.31329f,
+0.9258750392156863f, 0.42137633333333335f, 0.30765117647058826f,
+0.9240808562091503f, 0.4157374444444445f, 0.3020123529411765f,
+0.9222866732026144f, 0.4100985555555556f, 0.29637352941176476f,
+0.9198773137254902f, 0.404767274509804f, 0.2911960392156863f,
+0.9170578366013071f, 0.3996410653594772f, 0.28632610457516344f,
+0.9142383594771242f, 0.39451485620915044f, 0.2814561699346406f,
+0.9114188823529411f, 0.3893886470588236f, 0.2765862352941177f,
+0.9085994052287582f, 0.38426243790849673f, 0.2717163006535948f,
+0.9057799281045752f, 0.37913622875816994f, 0.26684636601307193f,
+0.9029604509803921f, 0.3740100196078432f, 0.26197643137254906f,
+0.9001409738562092f, 0.36888381045751634f, 0.2571064967320262f,
+0.8973214967320261f, 0.36375760130718954f, 0.2522365620915033f,
+0.8945020196078431f, 0.35863139215686274f, 0.2473666274509804f,
+0.8916825424836602f, 0.35350518300653605f, 0.2424966928104576f,
+0.8888630653594771f, 0.34837897385620914f, 0.23762675816993467f,
+0.886043588235294f, 0.34325276470588234f, 0.2327568235294118f,
+0.8832241111111111f, 0.33812655555555554f, 0.22788688888888892f,
+0.880404633986928f, 0.33300034640522874f, 0.22301695424836604f,
+0.8776620784313726f, 0.3284892745098039f, 0.21891596078431372f,
+0.8750990065359476f, 0.325413522875817f, 0.21660916339869282f,
+0.8725359346405228f, 0.32233777124183005f, 0.2143023660130719f,
+0.869972862745098f, 0.3192620196078432f, 0.21199556862745103f,
+0.8674097908496732f, 0.3161862679738562f, 0.20968877124183008f,
+0.8648467189542484f, 0.3131105163398693f, 0.20738197385620916f,
+0.8622836470588235f, 0.31003476470588237f, 0.20507517647058823f,
+0.8597205751633986f, 0.30695901307189544f, 0.2027683790849673f,
+0.8571575032679738f, 0.3038832614379085f, 0.20046158169934641f,
+0.854594431372549f, 0.30080750980392157f, 0.1981547843137255f,
+0.8520313594771242f, 0.29773175816993463f, 0.19584798692810457f,
+0.8494682875816993f, 0.29465600653594776f, 0.1935411895424837f,
+0.8469052156862745f, 0.2915802549019608f, 0.19123439215686275f,
+0.8443421437908496f, 0.2885045032679739f, 0.18892759477124182f,
+0.8417790718954248f, 0.28542875163398695f, 0.18662079738562093f,
+0.839216f, 0.282353f, 0.184314f,
+0.836716931372549f, 0.28004618627450983f, 0.1825839019607843f,
+0.834217862745098f, 0.27773937254901965f, 0.18085380392156863f,
+0.831718794117647f, 0.2754325588235294f, 0.17912370588235293f,
+0.8292197254901961f, 0.2731257450980393f, 0.1773936078431373f,
+0.8267206568627451f, 0.27081893137254903f, 0.17566350980392156f,
+0.8242215882352941f, 0.26851211764705885f, 0.1739334117647059f,
+0.8217225196078431f, 0.26620530392156866f, 0.1722033137254902f,
+0.8192234509803922f, 0.2638984901960784f, 0.17047321568627452f,
+0.8167243823529412f, 0.26159167647058823f, 0.16874311764705882f,
+0.8142253137254902f, 0.25928486274509804f, 0.16701301960784315f,
+0.8117262450980391f, 0.25697804901960786f, 0.16528292156862745f,
+0.8092271764705883f, 0.25467123529411767f, 0.16355282352941178f,
+0.8067281078431372f, 0.25236442156862743f, 0.16182272549019608f,
+0.8042290392156862f, 0.25005760784313724f, 0.16009262745098038f,
+0.8017299705882353f, 0.24775079411764706f, 0.1583625294117647f,
+0.7992309019607843f, 0.24544398039215687f, 0.156632431372549f,
+0.7967318333333333f, 0.24313716666666665f, 0.15490233333333334f,
+0.7942327647058823f, 0.24083035294117644f, 0.15317223529411764f,
+0.7917336960784314f, 0.23852353921568625f, 0.15144213725490197f,
+0.7892346274509804f, 0.2362167254901961f, 0.1497120392156863f,
+0.7863895490196078f, 0.23314098039215683f, 0.1477896862745098f,
+0.7833137973856209f, 0.22955261437908495f, 0.1457391633986928f,
+0.780238045751634f, 0.22596424836601306f, 0.14368864052287583f,
+0.777162294117647f, 0.22237588235294117f, 0.14163811764705883f,
+0.7740865424836602f, 0.2187875163398693f, 0.13958759477124183f,
+0.7710107908496732f, 0.2151991503267974f, 0.13753707189542486f,
+0.7679350392156863f, 0.2116107843137255f, 0.13548654901960785f,
+0.7648592875816993f, 0.2080224183006536f, 0.13343602614379085f,
+0.7617835359477124f, 0.2044340522875817f, 0.13138550326797388f,
+0.7587077843137255f, 0.20084568627450994f, 0.12933498039215693f,
+0.7556320326797386f, 0.19725732026143794f, 0.12728445751633988f,
+0.7525562810457517f, 0.19366895424836605f, 0.1252339346405229f,
+0.7494805294117647f, 0.19008058823529417f, 0.12318341176470592f,
+0.7464047777777778f, 0.18649222222222228f, 0.12113288888888893f,
+0.7433290261437908f, 0.1829038562091504f, 0.11908236601307193f,
+0.7404070784313725f, 0.1795461764705883f, 0.11726252941176472f,
+0.7378440065359477f, 0.1767267647058824f, 0.11598096078431375f,
+0.7352809346405229f, 0.1739073529411765f, 0.11469939215686277f,
+0.732717862745098f, 0.1710879411764706f, 0.11341782352941178f,
+0.7301547908496732f, 0.16826852941176473f, 0.1121362549019608f,
+0.7275917189542483f, 0.16544911764705886f, 0.11085468627450983f,
+0.7250286470588235f, 0.16262970588235295f, 0.10957311764705883f,
+0.7224655751633987f, 0.15981029411764708f, 0.10829154901960786f,
+0.7199025032679738f, 0.15699088235294117f, 0.10700998039215687f,
+0.717339431372549f, 0.1541714705882353f, 0.10572841176470589f,
+0.7147763594771243f, 0.1513520588235295f, 0.10444684313725494f,
+0.7122132875816993f, 0.14853264705882352f, 0.10316527450980392f,
+0.7096502156862745f, 0.14571323529411764f, 0.10188370588235295f,
+0.7070871437908496f, 0.14289382352941177f, 0.10060213725490197f,
+0.7045240718954248f, 0.14007441176470586f, 0.09932056862745098f,
+0.701961f, 0.137255f, 0.098039f,
+0.699141588235294f, 0.13469186274509803f, 0.09675743137254902f,
+0.6963221764705881f, 0.13212872549019608f, 0.09547586274509805f,
+0.6935027647058823f, 0.12956558823529413f, 0.09419429411764706f,
+0.6906833529411764f, 0.12700245098039215f, 0.09291272549019608f,
+0.6878639411764705f, 0.12443931372549019f, 0.0916311568627451f,
+0.6850445294117646f, 0.12187617647058824f, 0.09034958823529413f,
+0.6822251176470588f, 0.11931303921568628f, 0.08906801960784315f,
+0.6794057058823529f, 0.11674990196078433f, 0.08778645098039217f,
+0.676586294117647f, 0.11418676470588236f, 0.08650488235294118f,
+0.6737668823529411f, 0.11162362745098041f, 0.0852233137254902f,
+0.6709474705882353f, 0.10906049019607852f, 0.08394174509803927f,
+0.6681280588235294f, 0.10649735294117649f, 0.08266017647058825f,
+0.6653086470588235f, 0.10393421568627453f, 0.08137860784313727f,
+0.6624892352941176f, 0.10137107843137258f, 0.08009703921568628f,
+0.6596698235294117f, 0.09880794117647063f, 0.0788154705882353f,
+0.6570298039215686f, 0.09750077777777778f, 0.07878987581699345f,
+0.6544666666666666f, 0.0967318888888889f, 0.07930255555555556f,
+0.6519035294117647f, 0.095963f, 0.07981523529411765f,
+0.6493403921568628f, 0.09519411111111112f, 0.08032791503267973f,
+0.6467772549019608f, 0.09442522222222223f, 0.08084059477124182f,
+0.6442141176470588f, 0.09365633333333334f, 0.08135327450980392f,
+0.6416509803921568f, 0.09288744444444445f, 0.08186595424836601f,
+0.6390878431372549f, 0.09211855555555556f, 0.0823786339869281f,
+0.636524705882353f, 0.09134966666666668f, 0.0828913137254902f,
+0.633961568627451f, 0.09058077777777779f, 0.08340399346405229f,
+0.631398431372549f, 0.08981188888888889f, 0.08391667320261438f,
+0.6288352941176472f, 0.08904300000000002f, 0.08442935294117646f,
+0.6262721568627452f, 0.08827411111111111f, 0.08494203267973857f,
+0.6237090196078432f, 0.08750522222222222f, 0.08545471241830066f,
+0.6211458823529412f, 0.08673633333333333f, 0.08596739215686275f,
+0.6185827450980392f, 0.086480045751634f, 0.08689013725490197f,
+0.6160196078431373f, 0.08699266013071896f, 0.08842798039215688f,
+0.6134564705882354f, 0.08750527450980393f, 0.08996582352941178f,
+0.6108933333333334f, 0.0880178888888889f, 0.09150366666666668f,
+0.6083301960784314f, 0.08853050326797386f, 0.09304150980392158f,
+0.6057670588235294f, 0.08904311764705883f, 0.09457935294117648f,
+0.6032039215686275f, 0.0895557320261438f, 0.09611719607843137f,
+0.6006407843137255f, 0.09006834640522876f, 0.09765503921568627f,
+0.5980776470588236f, 0.09058096078431373f, 0.09919288235294117f,
+0.5955145098039216f, 0.0910935751633987f, 0.10073072549019607f,
+0.5929513725490196f, 0.09160618954248366f, 0.10226856862745098f,
+0.5903882352941177f, 0.09211880392156863f, 0.10380641176470587f,
+0.5878250980392158f, 0.09263141830065358f, 0.10534425490196073f,
+0.5852619607843138f, 0.09314403267973856f, 0.10688209803921568f,
+0.5826988235294118f, 0.09365664705882353f, 0.10841994117647058f,
+0.5801356862745098f, 0.09414362745098039f, 0.10995778431372548f,
+0.5775725490196079f, 0.09439990196078431f, 0.11149562745098038f,
+0.5750094117647059f, 0.09465617647058823f, 0.11303347058823529f,
+0.5724462745098039f, 0.09491245098039215f, 0.11457131372549019f,
+0.569883137254902f, 0.09516872549019607f, 0.1161091568627451f,
+0.56732f, 0.095425f, 0.117647f,
+0.5647568627450981f, 0.09568127450980392f, 0.1191848431372549f,
+0.5621937254901961f, 0.09593754901960784f, 0.12072268627450981f,
+0.5596305882352941f, 0.09619382352941176f, 0.12226052941176471f,
+0.5570674509803921f, 0.09645009803921568f, 0.12379837254901963f,
+0.5545043137254901f, 0.0967063725490196f, 0.1253362156862745f,
+0.5519411764705883f, 0.09696264705882353f, 0.12687405882352942f,
+0.5493780392156863f, 0.09721892156862745f, 0.12841190196078434f,
+0.5468149019607844f, 0.09747519607843137f, 0.1299497450980392f,
+0.5442517647058823f, 0.0977314705882353f, 0.13148758823529413f,
+0.5416886274509803f, 0.09798774509803922f, 0.13302543137254905f,
+0.5387154509803921f, 0.098039f, 0.13456330588235296f,
+0.5356397647058824f, 0.098039f, 0.13610118823529413f,
+0.5325640784313725f, 0.098039f, 0.1376390705882353f,
+0.5294883921568627f, 0.098039f, 0.1391769529411765f,
+0.5264127058823529f, 0.098039f, 0.14071483529411766f,
+0.5233370196078432f, 0.098039f, 0.14225271764705882f,
+0.5202613333333334f, 0.098039f, 0.1437906f,
+0.5171856470588235f, 0.098039f, 0.14532848235294116f,
+0.5141099607843137f, 0.098039f, 0.14686636470588232f,
+0.511034274509804f, 0.098039f, 0.14840424705882352f,
+0.5079585882352942f, 0.098039f, 0.14994212941176469f,
+0.5048829019607843f, 0.098039f, 0.15148001176470585f,
+0.5017918352941177f, 0.09805438039215686f, 0.15297175686274508f,
+0.4984085411764707f, 0.0983619882352941f, 0.15358689411764703f,
+0.49502524705882356f, 0.09866959607843137f, 0.154202031372549f,
+0.4916419529411765f, 0.09897720392156863f, 0.15481716862745099f,
+0.4882586588235294f, 0.09928481176470588f, 0.15543230588235293f,
+0.4848753647058824f, 0.09959241960784314f, 0.1560474431372549f,
+0.4814920705882353f, 0.0999000274509804f, 0.15666258039215686f,
+0.47810877647058825f, 0.10020763529411765f, 0.15727771764705883f,
+0.47472548235294115f, 0.1005152431372549f, 0.15789285490196078f,
+0.4713421882352941f, 0.10082285098039216f, 0.15850799215686276f,
+0.467958894117647f, 0.1011304588235294f, 0.1591231294117647f,
+0.4645756f, 0.10143806666666666f, 0.15973826666666668f,
+0.46119230588235294f, 0.10174567450980392f, 0.16035340392156863f,
+0.45805505882352937f, 0.10211478431372549f, 0.16109158823529413f,
+0.45549192156862744f, 0.10262739869281046f, 0.1621168823529412f,
+0.45292878431372546f, 0.10314001307189542f, 0.16314217647058826f,
+0.45036564705882354f, 0.10365262745098039f, 0.1641674705882353f,
+0.4478025098039216f, 0.10416524183006534f, 0.16519276470588234f,
+0.4452393725490196f, 0.10467785620915032f, 0.1662180588235294f,
+0.44267623529411765f, 0.10519047058823529f, 0.16724335294117648f,
+0.44011309803921567f, 0.10570308496732027f, 0.16826864705882352f,
+0.4375499607843137f, 0.10621569934640523f, 0.1692939411764706f,
+0.43498682352941176f, 0.1067283137254902f, 0.17031923529411766f,
+0.4324236862745098f, 0.10724092810457517f, 0.1713445294117647f,
+0.42986054901960785f, 0.10775354248366013f, 0.17236982352941177f,
+0.42729741176470587f, 0.1082661568627451f, 0.1733951176470588f,
+0.4247342745098039f, 0.10877877124183007f, 0.17442041176470588f,
+0.42217113725490196f, 0.10929138562091503f, 0.17544570588235292f,
+0.419608f, 0.109804f, 0.176471f
+}
+, 
+{
+1.0f, 0.972549f, 0.94902f,
+0.9953863529411765f, 0.9642445294117648f, 0.9394851764705883f,
+0.9907727058823529f, 0.9559400588235294f, 0.9299503529411765f,
+0.9861590588235294f, 0.9476355882352941f, 0.9204155294117647f,
+0.9815454117647059f, 0.9393311176470588f, 0.9108807058823529f,
+0.9769317647058824f, 0.9310266470588235f, 0.9013458823529412f,
+0.9723181176470588f, 0.9227221764705882f, 0.8918110588235294f,
+0.9677044705882353f, 0.9144177058823529f, 0.8822762352941177f,
+0.9630908235294118f, 0.9061132352941176f, 0.8727414117647059f,
+0.9584771764705883f, 0.8978087647058823f, 0.8632065882352942f,
+0.9538635294117647f, 0.889504294117647f, 0.8536717647058824f,
+0.9492498823529412f, 0.8811998235294117f, 0.8441369411764706f,
+0.9446362352941177f, 0.8728953529411765f, 0.8346021176470588f,
+0.9399457058823529f, 0.864590862745098f, 0.8250672941176471f,
+0.9350245294117647f, 0.8562863137254901f, 0.8155324705882353f,
+0.9301033529411765f, 0.8479817647058824f, 0.8059976470588236f,
+0.9251821764705882f, 0.8396772156862745f, 0.7964628235294118f,
+0.920261f, 0.8313726666666666f, 0.7869280000000001f,
+0.9153398235294118f, 0.8230681176470588f, 0.7773931764705883f,
+0.9104186470588235f, 0.814763568627451f, 0.7678583529411765f,
+0.9054974705882353f, 0.8064590196078432f, 0.7583235294117647f,
+0.9005762941176471f, 0.7981544705882353f, 0.7487887058823529f,
+0.8956551176470589f, 0.7898499215686274f, 0.7392538823529412f,
+0.8907339411764705f, 0.7815453725490196f, 0.7297190588235294f,
+0.8858127647058823f, 0.7732408235294118f, 0.7201842352941177f,
+0.8808915882352941f, 0.764936274509804f, 0.7106494117647059f,
+0.8761242156862745f, 0.757400705882353f, 0.7026524705882353f,
+0.8715106470588235f, 0.7506341176470589f, 0.6961934117647058f,
+0.8668970784313725f, 0.7438675294117647f, 0.6897343529411765f,
+0.8622835098039215f, 0.7371009411764706f, 0.6832752941176471f,
+0.8576699411764706f, 0.7303343529411765f, 0.6768162352941176f,
+0.8530563725490196f, 0.7235677647058824f, 0.6703571764705882f,
+0.8484428039215686f, 0.7168011764705883f, 0.6638981176470589f,
+0.8438292352941177f, 0.7100345882352942f, 0.6574390588235295f,
+0.8392156666666667f, 0.703268f, 0.65098f,
+0.8346020980392157f, 0.6965014117647059f, 0.6445209411764706f,
+0.8299885294117647f, 0.6897348235294118f, 0.6380618823529411f,
+0.8253749607843137f, 0.6829682352941178f, 0.6316028235294118f,
+0.8207613921568627f, 0.6762016470588236f, 0.6251437647058824f,
+0.8173011764705882f, 0.670588411764706f, 0.6198381176470589f,
+0.8142254117647059f, 0.6653596274509804f, 0.6149169411764706f,
+0.8111496470588235f, 0.660130843137255f, 0.6099957647058823f,
+0.8080738823529412f, 0.6549020588235295f, 0.6050745882352941f,
+0.8049981176470589f, 0.6496732745098039f, 0.6001534117647058f,
+0.8019223529411764f, 0.6444444901960785f, 0.5952322352941176f,
+0.7988465882352941f, 0.639215705882353f, 0.5903110588235294f,
+0.7957708235294118f, 0.6339869215686275f, 0.5853898823529411f,
+0.7926950588235294f, 0.628758137254902f, 0.5804687058823529f,
+0.7896192941176471f, 0.6235293529411765f, 0.5755475294117647f,
+0.7865435294117648f, 0.6183005686274511f, 0.5706263529411765f,
+0.7834677647058823f, 0.6130717843137256f, 0.5657051764705882f,
+0.780392f, 0.607843f, 0.560784f,
+0.7773162352941176f, 0.6029218235294118f, 0.556170431372549f,
+0.7742404705882353f, 0.5980006470588236f, 0.5515568627450981f,
+0.7711647058823529f, 0.5930794705882353f, 0.546943294117647f,
+0.7680889411764705f, 0.5881582941176471f, 0.5423297254901961f,
+0.7650131764705882f, 0.5832371176470589f, 0.5377161568627451f,
+0.7619374117647059f, 0.5783159411764707f, 0.5331025882352941f,
+0.7588616470588235f, 0.5733947647058824f, 0.5284890196078431f,
+0.7557858823529411f, 0.5684735882352941f, 0.5238754509803921f,
+0.7527101176470588f, 0.5635524117647058f, 0.5192618823529411f,
+0.7496343529411764f, 0.5586312352941176f, 0.5146483137254902f,
+0.7465585882352941f, 0.5537100588235294f, 0.5100347450980391f,
+0.7434828235294118f, 0.5487888823529412f, 0.5054211764705883f,
+0.740560862745098f, 0.5440983921568627f, 0.5010382745098039f,
+0.7381003137254901f, 0.5400999607843137f, 0.4973473725490196f,
+0.7356397647058823f, 0.5361015294117647f, 0.49365647058823536f,
+0.7331792156862744f, 0.5321030980392156f, 0.48996556862745094f,
+0.7307186666666666f, 0.5281046666666667f, 0.48627466666666663f,
+0.7282581176470588f, 0.5241062352941176f, 0.4825837647058823f,
+0.7257975686274509f, 0.5201078039215686f, 0.478892862745098f,
+0.7233370196078431f, 0.5161093725490196f, 0.4752019607843137f,
+0.7208764705882352f, 0.5121109411764706f, 0.4715110588235294f,
+0.7184159215686274f, 0.5081125098039215f, 0.4678201568627451f,
+0.7159553725490196f, 0.5041140784313726f, 0.46412925490196083f,
+0.7134948235294117f, 0.5001156470588235f, 0.46043835294117647f,
+0.7110342745098039f, 0.49611721568627454f, 0.45674745098039216f,
+0.7082661176470588f, 0.49196498039215686f, 0.4530565490196078f,
+0.7051903529411765f, 0.4876589411764706f, 0.4493656470588235f,
+0.7021145882352942f, 0.4833529019607843f, 0.44567474509803917f,
+0.6990388235294117f, 0.47904686274509806f, 0.44198384313725486f,
+0.6959630588235294f, 0.47474082352941177f, 0.43829294117647055f,
+0.692887294117647f, 0.47043478431372554f, 0.4346020392156863f,
+0.6898115294117647f, 0.4661287450980392f, 0.43091113725490193f,
+0.6867357647058824f, 0.46182270588235297f, 0.4272202352941176f,
+0.6836599999999999f, 0.4575166666666667f, 0.4235293333333333f,
+0.6805842352941176f, 0.4532106274509804f, 0.419838431372549f,
+0.6775084705882353f, 0.4489045882352941f, 0.4161475294117647f,
+0.6744327058823529f, 0.4445985490196078f, 0.4124566274509804f,
+0.6713569411764706f, 0.44029250980392154f, 0.4087657254901961f,
+0.6687425882352941f, 0.4369091764705883f, 0.40599758823529414f,
+0.6662820392156863f, 0.43383341176470586f, 0.4035370392156863f,
+0.6638214901960784f, 0.4307576470588235f, 0.40107649019607844f,
+0.6613609411764706f, 0.4276818823529412f, 0.3986159411764706f,
+0.6589003921568627f, 0.4246061176470588f, 0.39615539215686274f,
+0.6564398431372549f, 0.42153035294117647f, 0.3936948431372549f,
+0.653979294117647f, 0.41845458823529413f, 0.39123429411764704f,
+0.6515187450980392f, 0.4153788235294118f, 0.38877374509803925f,
+0.6490581960784314f, 0.41230305882352947f, 0.3863131960784314f,
+0.6465976470588235f, 0.4092272941176471f, 0.38385264705882355f,
+0.6441370980392157f, 0.40615152941176474f, 0.3813920980392157f,
+0.6416765490196078f, 0.40307576470588236f, 0.37893154901960785f,
+0.639216f, 0.4f, 0.376471f,
+0.6370629803921569f, 0.3972318431372549f, 0.37462550980392156f,
+0.6349099607843137f, 0.39446368627450984f, 0.3727800196078431f,
+0.6327569411764706f, 0.3916955294117647f, 0.3709345294117647f,
+0.6306039215686274f, 0.38892737254901966f, 0.3690890392156863f,
+0.6284509019607843f, 0.38615921568627454f, 0.3672435490196079f,
+0.6262978823529411f, 0.3833910588235294f, 0.36539805882352944f,
+0.6241448627450981f, 0.3806229019607843f, 0.363552568627451f,
+0.6219918431372549f, 0.3778547450980392f, 0.36170707843137256f,
+0.6198388235294118f, 0.3750865882352941f, 0.3598615882352941f,
+0.6176858039215687f, 0.372318431372549f, 0.3580160980392157f,
+0.6155327843137255f, 0.3695502745098039f, 0.35617060784313725f,
+0.6133797647058824f, 0.3667821176470588f, 0.3543251176470589f,
+0.611149843137255f, 0.36409084313725487f, 0.3524796470588235f,
+0.6086892156862745f, 0.36163021568627446f, 0.35063423529411764f,
+0.6062285882352941f, 0.3591695882352941f, 0.34878882352941176f,
+0.6037679607843137f, 0.3567089607843137f, 0.3469434117647059f,
+0.6013073333333333f, 0.35424833333333333f, 0.345098f,
+0.598846705882353f, 0.3517877058823529f, 0.3432525882352941f,
+0.5963860784313726f, 0.3493270784313725f, 0.3414071764705882f,
+0.5939254509803922f, 0.34686645098039215f, 0.3395617647058824f,
+0.5914648235294118f, 0.34440582352941174f, 0.33771635294117647f,
+0.5890041960784314f, 0.3419451960784314f, 0.3358709411764706f,
+0.5865435686274509f, 0.33948456862745097f, 0.33402552941176467f,
+0.5840829411764706f, 0.33702394117647055f, 0.3321801176470588f,
+0.5816223137254902f, 0.3345633137254902f, 0.3303347058823529f,
+0.5791617254901961f, 0.3321027254901961f, 0.3286430588235294f,
+0.5767011764705883f, 0.32964217647058824f, 0.32710517647058823f,
+0.5742406274509804f, 0.3271816274509804f, 0.32556729411764707f,
+0.5717800784313726f, 0.32472107843137255f, 0.3240294117647059f,
+0.5693195294117648f, 0.32226052941176475f, 0.32249152941176473f,
+0.5668589803921569f, 0.31979998039215685f, 0.3209536470588235f,
+0.564398431372549f, 0.317339431372549f, 0.31941576470588234f,
+0.5619378823529412f, 0.31487888235294115f, 0.3178778823529412f,
+0.5594773333333333f, 0.3124183333333333f, 0.31634f,
+0.5570167843137255f, 0.30995778431372545f, 0.31480211764705884f,
+0.5545562352941176f, 0.30749723529411765f, 0.3132642352941177f,
+0.5520956862745098f, 0.3050366862745098f, 0.3117263529411765f,
+0.5496351372549019f, 0.30257613725490196f, 0.31018847058823534f,
+0.5474052352941176f, 0.30034623529411764f, 0.3086505882352941f,
+0.5452522156862745f, 0.2981932156862745f, 0.30711270588235295f,
+0.5430991960784314f, 0.29604019607843135f, 0.3055748235294118f,
+0.5409461764705882f, 0.2938871764705882f, 0.3040369411764706f,
+0.5387931568627451f, 0.29173415686274506f, 0.30249905882352945f,
+0.5366401372549019f, 0.2895811372549019f, 0.30096117647058823f,
+0.5344871176470588f, 0.2874281176470588f, 0.29942329411764707f,
+0.5323340980392157f, 0.28527509803921575f, 0.29788541176470595f,
+0.5301810784313725f, 0.28312207843137255f, 0.29634752941176473f,
+0.5280280588235293f, 0.2809690588235294f, 0.29480964705882357f,
+0.5258750392156862f, 0.27881603921568626f, 0.29327176470588234f,
+0.523722019607843f, 0.2766630196078431f, 0.2917338823529412f,
+0.521569f, 0.27451f, 0.290196f,
+0.5191083725490195f, 0.27266450980392154f, 0.2889657254901961f,
+0.5166477450980391f, 0.2708190196078431f, 0.28773545098039216f,
+0.5141871176470588f, 0.26897352941176467f, 0.28650517647058826f,
+0.5117264901960784f, 0.26712803921568623f, 0.2852749019607843f,
+0.509265862745098f, 0.2652825490196078f, 0.2840446274509804f,
+0.5068052352941176f, 0.2634370588235294f, 0.28281435294117646f,
+0.5043446078431373f, 0.261591568627451f, 0.28158407843137256f,
+0.5018839803921569f, 0.25974607843137254f, 0.2803538039215686f,
+0.49942335294117646f, 0.2579005882352941f, 0.2791235294117647f,
+0.4969627254901961f, 0.25605509803921567f, 0.27789325490196076f,
+0.49450209803921574f, 0.2542096078431373f, 0.27666298039215687f,
+0.49204147058823533f, 0.2523641176470588f, 0.2754327058823529f,
+0.48958086274509804f, 0.2504417450980392f, 0.27412552941176466f,
+0.4871203137254902f, 0.24828872549019607f, 0.2725876470588235f,
+0.48465976470588235f, 0.24613570588235292f, 0.2710497647058823f,
+0.48219921568627455f, 0.24398268627450978f, 0.26951188235294116f,
+0.4797386666666667f, 0.24182966666666666f, 0.267974f,
+0.47727811764705885f, 0.23967664705882352f, 0.2664361176470588f,
+0.474817568627451f, 0.23752362745098038f, 0.26489823529411766f,
+0.47235701960784315f, 0.23537060784313724f, 0.2633603529411765f,
+0.4698964705882353f, 0.23321758823529412f, 0.2618224705882353f,
+0.46743592156862745f, 0.23106456862745098f, 0.2602845882352941f,
+0.4649753725490196f, 0.22891154901960784f, 0.25874670588235293f,
+0.46251482352941176f, 0.22675852941176472f, 0.25720882352941177f,
+0.4600542745098039f, 0.22460550980392158f, 0.2556709411764706f,
+0.45774749019607847f, 0.2227600980392157f, 0.25428686274509804f,
+0.4555944705882354f, 0.2212222941176471f, 0.25305658823529414f,
+0.4534414509803922f, 0.21968449019607844f, 0.2518263137254902f,
+0.45128843137254904f, 0.2181466862745098f, 0.2505960392156863f,
+0.4491354117647059f, 0.21660888235294118f, 0.24936576470588237f,
+0.44698239215686275f, 0.21507107843137255f, 0.24813549019607845f,
+0.4448293725490196f, 0.21353327450980392f, 0.24690521568627452f,
+0.44267635294117647f, 0.21199547058823529f, 0.2456749411764706f,
+0.4405233333333333f, 0.21045766666666665f, 0.24444466666666667f,
+0.4383703137254902f, 0.20891986274509802f, 0.24321439215686275f,
+0.4362172941176471f, 0.2073820588235294f, 0.24198411764705885f,
+0.43406427450980395f, 0.2058442549019608f, 0.24075384313725492f,
+0.4319112549019608f, 0.20430645098039216f, 0.239523568627451f,
+0.4295275294117647f, 0.20253788235294118f, 0.23829323529411767f,
+0.42706690196078434f, 0.20069239215686274f, 0.23706288235294118f,
+0.4246062745098039f, 0.1988469019607843f, 0.23583252941176472f,
+0.4221456470588235f, 0.19700141176470587f, 0.23460217647058826f,
+0.4196850196078432f, 0.1951559215686275f, 0.2333718235294118f,
+0.41722439215686274f, 0.19331043137254902f, 0.2321414705882353f,
+0.4147637647058824f, 0.1914649411764706f, 0.23091111764705882f,
+0.412303137254902f, 0.18961945098039215f, 0.22968076470588236f,
+0.40984250980392156f, 0.18777396078431371f, 0.2284504117647059f,
+0.4073818823529412f, 0.1859284705882353f, 0.2272200588235294f,
+0.4049212549019608f, 0.18408298039215687f, 0.22598970588235295f,
+0.40246062745098043f, 0.18223749019607843f, 0.22475935294117647f,
+0.4f, 0.180392f, 0.223529f,
+0.3975393725490196f, 0.17854658823529412f, 0.22199119607843137f,
+0.39507874509803925f, 0.17670117647058822f, 0.22045339215686274f,
+0.39261811764705884f, 0.17485576470588235f, 0.2189155882352941f,
+0.3901574901960784f, 0.17301035294117648f, 0.21737778431372548f,
+0.38769686274509807f, 0.17116494117647058f, 0.21583998039215685f,
+0.38523623529411766f, 0.1693195294117647f, 0.21430217647058825f,
+0.38277560784313724f, 0.16747411764705883f, 0.21276437254901961f,
+0.38031498039215694f, 0.16562870588235298f, 0.211226568627451f,
+0.3778543529411765f, 0.16378329411764705f, 0.20968876470588235f,
+0.37539372549019606f, 0.16193788235294118f, 0.20815096078431372f,
+0.37293309803921565f, 0.16009247058823528f, 0.2066131568627451f,
+0.3704724705882353f, 0.1582470588235294f, 0.20507535294117646f,
+0.3678580784313725f, 0.1564016274509804f, 0.2035375294117647f,
+0.3647823921568627f, 0.15455613725490197f, 0.20199964705882353f,
+0.36170670588235293f, 0.15271064705882353f, 0.20046176470588234f,
+0.3586310196078431f, 0.1508651568627451f, 0.19892388235294117f,
+0.35555533333333333f, 0.14901966666666666f, 0.197386f,
+0.3524796470588235f, 0.14717417647058825f, 0.1958481176470588f,
+0.3494039607843137f, 0.14532868627450982f, 0.19431023529411764f,
+0.3463282745098039f, 0.14348319607843138f, 0.19277235294117648f,
+0.3432525882352941f, 0.14163770588235294f, 0.19123447058823528f,
+0.3401769019607843f, 0.1397922156862745f, 0.18969658823529412f,
+0.3371012156862745f, 0.1379467254901961f, 0.18815870588235295f,
+0.3340255294117648f, 0.13610123529411772f, 0.1866208235294118f,
+0.3309498431372549f, 0.13425574509803923f, 0.1850829411764706f,
+0.3278741176470588f, 0.13225649019607844f, 0.1833912549019608f,
+0.3247983529411765f, 0.1301034705882353f, 0.18154576470588235f,
+0.3217225882352941f, 0.12795045098039215f, 0.17970027450980391f,
+0.31864682352941176f, 0.12579743137254903f, 0.1778547843137255f,
+0.3155710588235294f, 0.12364441176470589f, 0.17600929411764707f,
+0.31249529411764704f, 0.12149139215686275f, 0.17416380392156863f,
+0.3094195294117647f, 0.11933837254901962f, 0.1723183137254902f,
+0.30634376470588237f, 0.11718535294117648f, 0.17047282352941176f,
+0.303268f, 0.11503233333333335f, 0.16862733333333335f,
+0.30019223529411765f, 0.1128793137254902f, 0.16678184313725491f,
+0.2971164705882353f, 0.11072629411764706f, 0.16493635294117648f,
+0.2940407058823529f, 0.10857327450980392f, 0.16309086274509804f,
+0.2909649411764706f, 0.10642025490196079f, 0.1612453725490196f,
+0.28788917647058826f, 0.10426723529411765f, 0.15778517647058823f,
+0.284813411764706f, 0.10211421568627457f, 0.15378674509803933f,
+0.28173764705882354f, 0.09996119607843137f, 0.1497883137254902f,
+0.2786618823529412f, 0.09780817647058823f, 0.14578988235294119f,
+0.2755861176470588f, 0.0956551568627451f, 0.14179145098039217f,
+0.2725103529411765f, 0.09350213725490196f, 0.13779301960784315f,
+0.2694345882352941f, 0.09134911764705883f, 0.1337945882352941f,
+0.26635882352941176f, 0.08919609803921569f, 0.1297961568627451f,
+0.2632830588235294f, 0.08704307843137254f, 0.12579772549019608f,
+0.26020729411764704f, 0.08489005882352942f, 0.12179929411764706f,
+0.2571315294117647f, 0.08273703921568627f, 0.11780086274509805f,
+0.2540557647058823f, 0.08058401960784314f, 0.11380243137254903f,
+0.25098f, 0.078431f, 0.109804f
+}
+, 
+{
+1.0f, 0.996078f, 0.890196f,
+0.9995386274509804f, 0.9950014901960784f, 0.8832755686274509f,
+0.9990772549019608f, 0.9939249803921569f, 0.876355137254902f,
+0.9986158823529412f, 0.9928484705882353f, 0.8694347058823529f,
+0.9981545098039216f, 0.9917719607843137f, 0.8625142745098039f,
+0.9976931372549019f, 0.9906954509803921f, 0.855593843137255f,
+0.9972317647058824f, 0.9896189411764706f, 0.8486734117647059f,
+0.9967703921568627f, 0.988542431372549f, 0.8417529803921568f,
+0.9963090196078431f, 0.9874659215686274f, 0.8348325490196078f,
+0.9958476470588236f, 0.9863894117647058f, 0.8279121176470589f,
+0.9953862745098039f, 0.9853129019607844f, 0.8209916862745098f,
+0.9949249019607843f, 0.9842363921568628f, 0.8140712549019608f,
+0.9944635294117647f, 0.9831598823529412f, 0.8071508235294118f,
+0.9940021568627451f, 0.9820833725490197f, 0.8002303921568628f,
+0.9935407843137255f, 0.9810068627450981f, 0.7933099607843137f,
+0.9930794117647058f, 0.9799303529411765f, 0.7863895294117648f,
+0.9926180392156863f, 0.9788538431372549f, 0.7794690980392157f,
+0.9921566666666667f, 0.9777773333333334f, 0.7725486666666667f,
+0.991695294117647f, 0.9767008235294118f, 0.7656282352941177f,
+0.9912339215686274f, 0.9756243137254902f, 0.7587078039215687f,
+0.9907725490196079f, 0.9745478039215687f, 0.7517873725490196f,
+0.9903111764705882f, 0.9734712941176471f, 0.7448669411764707f,
+0.9898498039215686f, 0.9723947843137255f, 0.7379465098039216f,
+0.989388431372549f, 0.9713182745098039f, 0.7310260784313726f,
+0.9889270588235294f, 0.9702417647058824f, 0.7241056470588236f,
+0.9884656862745098f, 0.9691652549019608f, 0.7171852156862746f,
+0.9880812156862745f, 0.9679349607843137f, 0.7112644117647059f,
+0.9877736470588235f, 0.9665508823529412f, 0.7063432352941177f,
+0.9874660784313726f, 0.9651668039215686f, 0.7014220588235295f,
+0.9871585098039215f, 0.9637827254901961f, 0.6965008823529412f,
+0.9868509411764705f, 0.9623986470588235f, 0.691579705882353f,
+0.9865433725490196f, 0.9610145686274509f, 0.6866585294117648f,
+0.9862358039215686f, 0.9596304901960785f, 0.6817373529411765f,
+0.9859282352941177f, 0.9582464117647059f, 0.6768161764705883f,
+0.9856206666666667f, 0.9568623333333334f, 0.671895f,
+0.9853130980392157f, 0.9554782549019608f, 0.6669738235294118f,
+0.9850055294117647f, 0.9540941764705883f, 0.6620526470588235f,
+0.9846979607843137f, 0.9527100980392157f, 0.6571314705882353f,
+0.9843903921568627f, 0.9513260196078431f, 0.6522102941176471f,
+0.9840828235294118f, 0.9499419411764706f, 0.6472891176470588f,
+0.9837752549019608f, 0.948557862745098f, 0.6423679411764706f,
+0.9834676862745099f, 0.9471737843137255f, 0.6374467647058824f,
+0.9831601176470588f, 0.9457897058823529f, 0.6325255882352941f,
+0.9828525490196078f, 0.9444056274509803f, 0.6276044117647058f,
+0.9825449803921569f, 0.9430215490196078f, 0.6226832352941176f,
+0.9822374117647059f, 0.9416374705882353f, 0.6177620588235294f,
+0.981929843137255f, 0.9402533921568628f, 0.6128408823529411f,
+0.981622274509804f, 0.9388693137254902f, 0.6079197058823529f,
+0.9813147058823529f, 0.9374852352941176f, 0.6029985294117647f,
+0.981007137254902f, 0.9361011568627451f, 0.5980773529411765f,
+0.980699568627451f, 0.9347170784313725f, 0.5931561764705882f,
+0.980392f, 0.933333f, 0.588235f,
+0.9799306274509805f, 0.932102725490196f, 0.5850054705882353f,
+0.9794692549019608f, 0.9308724509803922f, 0.5817759411764706f,
+0.9790078823529412f, 0.9296421764705882f, 0.5785464117647059f,
+0.9785465098039217f, 0.9284119019607843f, 0.5753168823529411f,
+0.978085137254902f, 0.9271816274509803f, 0.5720873529411764f,
+0.9776237647058824f, 0.9259513529411765f, 0.5688578235294117f,
+0.9771623921568627f, 0.9247210784313725f, 0.565628294117647f,
+0.9767010196078432f, 0.9234908039215686f, 0.5623987647058823f,
+0.9762396470588236f, 0.9222605294117647f, 0.5591692352941177f,
+0.9757782745098039f, 0.9210302549019608f, 0.555939705882353f,
+0.9753169019607844f, 0.9197999803921568f, 0.5527101764705882f,
+0.9748555294117648f, 0.918569705882353f, 0.5494806470588235f,
+0.9743941568627451f, 0.917339431372549f, 0.5462511176470588f,
+0.9739327843137255f, 0.9161091568627451f, 0.5430215882352941f,
+0.9734714117647059f, 0.9148788823529412f, 0.5397920588235294f,
+0.9730100392156863f, 0.9136486078431373f, 0.5365625294117647f,
+0.9725486666666667f, 0.9124183333333333f, 0.5333330000000001f,
+0.972087294117647f, 0.9111880588235294f, 0.5301034705882353f,
+0.9716259215686275f, 0.9099577843137255f, 0.5268739411764706f,
+0.9711645490196079f, 0.9087275098039216f, 0.5236444117647059f,
+0.9707031764705882f, 0.9074972352941176f, 0.5204148823529412f,
+0.9702418039215687f, 0.9062669607843137f, 0.5171853529411765f,
+0.969780431372549f, 0.9050366862745098f, 0.5139558235294118f,
+0.9693190588235294f, 0.9038064117647059f, 0.5107262941176471f,
+0.9688576862745099f, 0.9025761372549019f, 0.5074967647058825f,
+0.9684732156862745f, 0.9011920588235294f, 0.5042672352941177f,
+0.9681656470588236f, 0.8996541764705882f, 0.501037705882353f,
+0.9678580784313726f, 0.898116294117647f, 0.4978081764705883f,
+0.9675505098039215f, 0.8965784117647059f, 0.4945786470588236f,
+0.9672429411764706f, 0.8950405294117647f, 0.49134911764705885f,
+0.9669353725490196f, 0.8935026470588235f, 0.4881195882352942f,
+0.9666278039215687f, 0.8919647647058824f, 0.4848900588235294f,
+0.9663202352941177f, 0.8904268823529412f, 0.48166052941176474f,
+0.9660126666666666f, 0.888889f, 0.47843100000000005f,
+0.9657050980392157f, 0.8873511176470589f, 0.4752014705882353f,
+0.9653975294117647f, 0.8858132352941176f, 0.4719719411764706f,
+0.9650899607843137f, 0.8842753529411764f, 0.4687424117647059f,
+0.9647823921568628f, 0.8827374705882353f, 0.4655128823529412f,
+0.9644748235294117f, 0.8811995882352941f, 0.4622833529411765f,
+0.9641672549019608f, 0.8796617058823529f, 0.45905382352941176f,
+0.9638596862745098f, 0.8781238235294118f, 0.4558242941176471f,
+0.9635521176470588f, 0.8765859411764706f, 0.45259476470588234f,
+0.9632445490196079f, 0.8750480588235294f, 0.44936523529411765f,
+0.9629369803921568f, 0.8735101764705883f, 0.44613570588235296f,
+0.9626294117647058f, 0.8719722941176471f, 0.4429061764705883f,
+0.9623218431372549f, 0.8704344117647058f, 0.43967664705882353f,
+0.9620142745098039f, 0.8688965294117648f, 0.4364471176470589f,
+0.961706705882353f, 0.8673586470588235f, 0.4332175882352941f,
+0.9613991372549019f, 0.8658207647058823f, 0.4299880588235294f,
+0.9610915686274509f, 0.8642828823529412f, 0.4267585294117647f,
+0.960784f, 0.862745f, 0.423529f,
+0.9603226666666667f, 0.861514725490196f, 0.4210684117647059f,
+0.9598613333333333f, 0.8602844509803922f, 0.4186078235294117f,
+0.9593999999999999f, 0.8590541764705882f, 0.4161472352941176f,
+0.9589386666666666f, 0.8578239019607843f, 0.4136866470588236f,
+0.9584773333333333f, 0.8565936274509803f, 0.4112260588235294f,
+0.958016f, 0.8553633529411765f, 0.40876547058823526f,
+0.9575546666666667f, 0.8541330784313725f, 0.40630488235294115f,
+0.9570933333333334f, 0.8529028039215686f, 0.40384429411764705f,
+0.9566319999999999f, 0.8516725294117647f, 0.40138370588235295f,
+0.9561706666666666f, 0.8504422549019608f, 0.3989231176470588f,
+0.9557093333333333f, 0.8492119803921568f, 0.3964625294117647f,
+0.955248f, 0.847981705882353f, 0.39400194117647064f,
+0.9547866666666667f, 0.846751431372549f, 0.3915413529411765f,
+0.9543253333333334f, 0.8455211568627451f, 0.3890807647058823f,
+0.9538639999999999f, 0.8442908823529411f, 0.3866201764705882f,
+0.9534026666666666f, 0.8430606078431373f, 0.3841595882352941f,
+0.9529413333333333f, 0.8418303333333333f, 0.381699f,
+0.95248f, 0.8406000588235294f, 0.37923841176470585f,
+0.9520186666666667f, 0.8393697843137256f, 0.37677782352941175f,
+0.9515573333333334f, 0.8381395098039216f, 0.37431723529411765f,
+0.9510959999999999f, 0.8369092352941176f, 0.37185664705882354f,
+0.9506346666666666f, 0.8356789607843137f, 0.3693960588235294f,
+0.9501733333333333f, 0.8344486862745099f, 0.3669354705882353f,
+0.949712f, 0.8332184117647059f, 0.3644748823529412f,
+0.9492506666666667f, 0.831988137254902f, 0.3620142941176471f,
+0.9488661960784314f, 0.830757843137255f, 0.3596306078431373f,
+0.948558588235294f, 0.8295275294117648f, 0.3573238235294118f,
+0.9482509803921568f, 0.8282972156862746f, 0.3550170392156863f,
+0.9479433725490196f, 0.8270669019607844f, 0.3527102549019608f,
+0.9476357647058823f, 0.8258365882352942f, 0.35040347058823534f,
+0.9473281568627451f, 0.824606274509804f, 0.3480966862745098f,
+0.9470205490196079f, 0.8233759607843137f, 0.3457899019607843f,
+0.9467129411764705f, 0.8221456470588235f, 0.3434831176470588f,
+0.9464053333333333f, 0.8209153333333333f, 0.3411763333333333f,
+0.9460977254901961f, 0.8196850196078431f, 0.3388695490196078f,
+0.9457901176470588f, 0.818454705882353f, 0.33656276470588237f,
+0.9454825098039216f, 0.8172243921568628f, 0.33425598039215687f,
+0.9451749019607844f, 0.8159940784313726f, 0.3319491960784314f,
+0.944867294117647f, 0.8147637647058824f, 0.3296424117647059f,
+0.9445596862745098f, 0.8135334509803922f, 0.3273356274509804f,
+0.9442520784313726f, 0.812303137254902f, 0.3250288431372549f,
+0.9439444705882353f, 0.8110728235294118f, 0.3227220588235294f,
+0.9436368627450981f, 0.8098425098039216f, 0.3204152745098039f,
+0.9433292549019607f, 0.8086121960784314f, 0.3181084901960784f,
+0.9430216470588235f, 0.8073818823529412f, 0.3158017058823529f,
+0.9427140392156863f, 0.8061515686274511f, 0.3134949215686275f,
+0.942406431372549f, 0.8049212549019609f, 0.31118813725490196f,
+0.9420988235294118f, 0.8036909411764707f, 0.30888135294117647f,
+0.9417912156862746f, 0.8024606274509805f, 0.30657456862745097f,
+0.9414836078431372f, 0.8012303137254903f, 0.3042677843137255f,
+0.941176f, 0.8f, 0.301961f,
+0.9407146666666667f, 0.7987696862745098f, 0.29965417647058823f,
+0.9402533333333334f, 0.7975393725490196f, 0.2973473529411764f,
+0.939792f, 0.7963090588235294f, 0.2950405294117647f,
+0.9393306666666666f, 0.7950787450980392f, 0.2927337058823529f,
+0.9388693333333333f, 0.793848431372549f, 0.2904268823529412f,
+0.938408f, 0.7926181176470588f, 0.28812005882352937f,
+0.9379466666666667f, 0.7913878039215686f, 0.2858132352941176f,
+0.9374853333333334f, 0.7901574901960784f, 0.28350641176470587f,
+0.937024f, 0.7889271764705883f, 0.2811995882352941f,
+0.9365626666666667f, 0.7876968627450981f, 0.2788927647058823f,
+0.9361013333333333f, 0.7864665490196079f, 0.2765859411764706f,
+0.93564f, 0.7852362352941177f, 0.2742791176470588f,
+0.9351786666666667f, 0.7840059215686275f, 0.27197229411764706f,
+0.9347173333333334f, 0.7827756078431373f, 0.26966547058823526f,
+0.934256f, 0.7815452941176471f, 0.2673586470588235f,
+0.9337946666666667f, 0.7803149803921569f, 0.26505182352941176f,
+0.9333333333333333f, 0.7790846666666666f, 0.262745f,
+0.932872f, 0.7778543529411764f, 0.2604381764705882f,
+0.9324106666666667f, 0.7766240392156862f, 0.25813135294117645f,
+0.9319493333333333f, 0.775393725490196f, 0.2558245294117647f,
+0.931488f, 0.7741634117647058f, 0.25351770588235295f,
+0.9310266666666667f, 0.7729330980392156f, 0.25121088235294114f,
+0.9305653333333334f, 0.7717027843137254f, 0.2489040588235294f,
+0.930104f, 0.7704724705882352f, 0.24659723529411764f,
+0.9296426666666667f, 0.769242156862745f, 0.24429041176470587f,
+0.9292582156862745f, 0.7679349607843137f, 0.24159913725490195f,
+0.9289506470588236f, 0.7665508823529412f, 0.23852341176470596f,
+0.9286430784313726f, 0.7651668039215686f, 0.2354476862745098f,
+0.9283355098039215f, 0.763782725490196f, 0.2323719607843137f,
+0.9280279411764706f, 0.7623986470588235f, 0.22929623529411763f,
+0.9277203725490196f, 0.761014568627451f, 0.22622050980392155f,
+0.9274128039215687f, 0.7596304901960784f, 0.22314478431372547f,
+0.9271052352941177f, 0.7582464117647059f, 0.2200690588235294f,
+0.9267976666666666f, 0.7568623333333333f, 0.21699333333333332f,
+0.9264900980392157f, 0.7554782549019607f, 0.21391760784313724f,
+0.9261825294117647f, 0.7540941764705882f, 0.21084188235294116f,
+0.9258749607843137f, 0.7527100980392156f, 0.20776615686274508f,
+0.9255673921568628f, 0.7513260196078431f, 0.204690431372549f,
+0.9252598235294117f, 0.7499419411764706f, 0.20161470588235292f,
+0.9249522549019608f, 0.7485578627450981f, 0.19853898039215687f,
+0.9246446862745098f, 0.7471737843137255f, 0.1954632549019608f,
+0.9243371176470588f, 0.7457897058823529f, 0.19238752941176468f,
+0.9240295490196079f, 0.7444056274509804f, 0.1893118039215687f,
+0.9237219803921568f, 0.7430215490196078f, 0.18623607843137255f,
+0.9234144117647058f, 0.7416374705882353f, 0.18316035294117647f,
+0.9231068431372549f, 0.7402533921568627f, 0.18008462745098036f,
+0.9227992745098039f, 0.7388693137254903f, 0.1770089019607843f,
+0.922491705882353f, 0.7374852352941177f, 0.17393317647058823f,
+0.9221841372549019f, 0.7361011568627451f, 0.17085745098039215f,
+0.9218765686274509f, 0.7347170784313726f, 0.16778172549019607f,
+0.921569f, 0.733333f, 0.164706f,
+0.9211076274509804f, 0.7316413529411765f, 0.16178403921568627f,
+0.9206462549019607f, 0.729949705882353f, 0.15886207843137254f,
+0.9201848823529412f, 0.7282580588235295f, 0.1559401176470588f,
+0.9197235098039216f, 0.7265664117647059f, 0.15301815686274509f,
+0.9192621372549019f, 0.7248747647058824f, 0.15009619607843136f,
+0.9188007647058823f, 0.7231831176470589f, 0.14717423529411763f,
+0.9183393921568627f, 0.7214914705882353f, 0.1442522745098039f,
+0.9178780196078431f, 0.7197998235294119f, 0.14133031372549026f,
+0.9174166470588235f, 0.7181081764705882f, 0.13840835294117645f,
+0.9169552745098039f, 0.7164165294117647f, 0.13548639215686273f,
+0.9164939019607843f, 0.7147248823529412f, 0.132564431372549f,
+0.9160325294117647f, 0.7130332352941177f, 0.1296424705882353f,
+0.915571156862745f, 0.7113415882352941f, 0.12672050980392158f,
+0.9151097843137255f, 0.7096499411764706f, 0.12379854901960784f,
+0.9146484117647058f, 0.7079582941176471f, 0.12087658823529411f,
+0.9141870392156862f, 0.7062666470588236f, 0.1179546274509804f,
+0.9137256666666667f, 0.7045750000000001f, 0.11503266666666667f,
+0.913264294117647f, 0.7028833529411765f, 0.11211070588235293f,
+0.9128029215686274f, 0.7011917058823529f, 0.10918874509803922f,
+0.9123415490196078f, 0.6995000588235294f, 0.10626678431372549f,
+0.9118801764705882f, 0.6978084117647059f, 0.10334482352941177f,
+0.9114188039215686f, 0.6961167647058824f, 0.10042286274509803f,
+0.9109574313725489f, 0.6944251176470588f, 0.09750090196078431f,
+0.9104960588235294f, 0.6927334705882354f, 0.09457894117647067f,
+0.9100346862745098f, 0.6910418235294118f, 0.09165698039215686f,
+0.9095476928104574f, 0.6891707712418301f, 0.08865812418300653f,
+0.9090350784313725f, 0.6871203137254902f, 0.0855823725490196f,
+0.9085224640522875f, 0.6850698562091504f, 0.08250662091503266f,
+0.9080098496732025f, 0.6830193986928105f, 0.07943086928104574f,
+0.9074972352941176f, 0.6809689411764706f, 0.07635511764705881f,
+0.9069846209150326f, 0.6789184836601307f, 0.07327936601307188f,
+0.9064720065359477f, 0.6768680261437908f, 0.07020361437908494f,
+0.9059593921568627f, 0.674817568627451f, 0.06712786274509802f,
+0.9054467777777777f, 0.6727671111111111f, 0.06405211111111109f,
+0.9049341633986928f, 0.6707166535947712f, 0.060976359477124153f,
+0.9044215490196078f, 0.6686661960784314f, 0.05790060784313722f,
+0.9039089346405229f, 0.6666157385620914f, 0.054824856209150286f,
+0.9033963202614379f, 0.6645652810457516f, 0.05174910457516336f,
+0.9028837058823529f, 0.6625148235294117f, 0.048673352941176425f,
+0.902371091503268f, 0.6604643660130719f, 0.04559760130718958f,
+0.9017303137254902f, 0.6584395294117646f, 0.042291176470588165f,
+0.9005768823529412f, 0.6565171764705882f, 0.03806205882352935f,
+0.8994234509803921f, 0.6545948235294117f, 0.03383294117647053f,
+0.8982700196078431f, 0.6526724705882353f, 0.029603823529411717f,
+0.897116588235294f, 0.6507501176470588f, 0.0253747058823529f,
+0.8959631568627451f, 0.6488277647058823f, 0.021145588235294083f,
+0.8948097254901961f, 0.6469054117647058f, 0.01691647058823527f,
+0.893656294117647f, 0.6449830588235294f, 0.012687352941176448f,
+0.892502862745098f, 0.643060705882353f, 0.008458235294117634f,
+0.891349431372549f, 0.6411383529411765f, 0.004229117647058821f,
+0.890196f, 0.639216f, 0.0f
+    }
+    , 
+    {
+        1.0f, 0.996078f, 0.941176f,
+        0.996155294117647f, 0.9922334901960784f, 0.9257973725490196f,
+        0.9923105882352942f, 0.9883889803921568f, 0.9104187450980392f,
+        0.9884658823529412f, 0.9845444705882352f, 0.8950401176470588f,
+        0.9846211764705882f, 0.9806999607843138f, 0.8796614901960784f,
+        0.9807764705882354f, 0.9768554509803922f, 0.864282862745098f,
+        0.9790078823529412f, 0.9741641764705882f, 0.8549018235294118f,
+        0.9774699738562092f, 0.9716010392156863f, 0.846187183006536f,
+        0.9759320653594772f, 0.9690379019607843f, 0.8374725424836601f,
+        0.9743941568627451f, 0.9664747647058823f, 0.8287579019607842f,
+        0.9728562483660131f, 0.9639116274509804f, 0.8200432614379085f,
+        0.9713183398692811f, 0.9613484901960785f, 0.8113286209150327f,
+        0.969780431372549f, 0.9587853529411765f, 0.8026139803921568f,
+        0.9680887450980392f, 0.9558633725490197f, 0.7937711960784313f,
+        0.9659357254901961f, 0.9518648627450981f, 0.7845439803921568f,
+        0.9637827058823529f, 0.9478663529411765f, 0.7753167647058823f,
+        0.9616296862745098f, 0.9438678431372549f, 0.7660895490196078f,
+        0.9594766666666666f, 0.9398693333333333f, 0.7568623333333333f,
+        0.9573236470588236f, 0.9358708235294118f, 0.7476351176470588f,
+        0.9551706274509804f, 0.9318723137254902f, 0.7384079019607843f,
+        0.9530176078431373f, 0.9278738039215686f, 0.7291806862745098f,
+        0.9508645882352942f, 0.923875294117647f, 0.7199534705882353f,
+        0.948711568627451f, 0.9198767843137254f, 0.7107262549019607f,
+        0.9465585490196079f, 0.9158782745098039f, 0.7014990392156862f,
+        0.9444055294117647f, 0.9118797647058823f, 0.6922718235294117f,
+        0.9422525098039216f, 0.9078812549019607f, 0.6830446078431373f,
+        0.9404070784313726f, 0.9043441372549019f, 0.6757397254901961f,
+        0.9388692352941177f, 0.9012684117647058f, 0.6703571764705882f,
+        0.9373313921568628f, 0.8981926862745098f, 0.6649746274509803f,
+        0.9357935490196079f, 0.8951169607843137f, 0.6595920784313726f,
+        0.934255705882353f, 0.8920412352941176f, 0.6542095294117647f,
+        0.932717862745098f, 0.8889655098039215f, 0.6488269803921569f,
+        0.9311800196078431f, 0.8858897843137254f, 0.643444431372549f,
+        0.9296421764705882f, 0.8828140588235294f, 0.6380618823529413f,
+        0.9281043333333333f, 0.8797383333333333f, 0.6326793333333334f,
+        0.9265664901960784f, 0.8766626078431372f, 0.6272967843137255f,
+        0.9250286470588236f, 0.8735868823529411f, 0.6219142352941176f,
+        0.9234908039215687f, 0.8705111568627452f, 0.6165316862745098f,
+        0.9219529607843138f, 0.867435431372549f, 0.6111491372549019f,
+        0.9204151176470589f, 0.8643597058823529f, 0.6057665882352941f,
+        0.9188772745098039f, 0.8612839803921569f, 0.6003840392156863f,
+        0.917339431372549f, 0.8582082549019608f, 0.5950014901960785f,
+        0.9158015882352941f, 0.8551325294117648f, 0.5896189411764706f,
+        0.9142637450980392f, 0.8520568039215687f, 0.5842363921568627f,
+        0.9127259019607843f, 0.8489810784313726f, 0.5788538431372549f,
+        0.9111880588235295f, 0.8459053529411765f, 0.5734712941176471f,
+        0.9096502156862745f, 0.8428296274509804f, 0.5680887450980392f,
+        0.9081123725490197f, 0.8397539019607844f, 0.5627061960784314f,
+        0.9065745294117648f, 0.8366781764705883f, 0.5573236470588235f,
+        0.9050366862745098f, 0.8336024509803922f, 0.5519410980392156f,
+        0.9034988431372549f, 0.8305267254901961f, 0.5465585490196079f,
+        0.901961f, 0.827451f, 0.541176f,
+        0.8999617450980393f, 0.8237600980392157f, 0.5379465098039216f,
+        0.8979624901960784f, 0.8200691960784314f, 0.5347170196078431f,
+        0.8959632352941177f, 0.8163782941176471f, 0.5314875294117647f,
+        0.8939639803921569f, 0.8126873921568628f, 0.5282580392156863f,
+        0.8919647254901961f, 0.8089964901960784f, 0.5250285490196078f,
+        0.8899654705882353f, 0.8053055882352942f, 0.5217990588235294f,
+        0.8879662156862745f, 0.8016146862745098f, 0.518569568627451f,
+        0.8859669607843137f, 0.7979237843137256f, 0.5153400784313725f,
+        0.883967705882353f, 0.7942328823529412f, 0.5121105882352941f,
+        0.8819684509803921f, 0.790541980392157f, 0.5088810980392157f,
+        0.8799691960784314f, 0.7868510784313726f, 0.5056516078431372f,
+        0.8779699411764705f, 0.7831601764705882f, 0.5024221176470588f,
+        0.8759706862745098f, 0.779469274509804f, 0.49919262745098036f,
+        0.873971431372549f, 0.7757783725490196f, 0.49596313725490193f,
+        0.8719721764705882f, 0.7720874705882353f, 0.49273364705882355f,
+        0.8699729215686274f, 0.768396568627451f, 0.4895041568627451f,
+        0.8679736666666666f, 0.7647056666666667f, 0.4862746666666667f,
+        0.8659744117647059f, 0.7610147647058824f, 0.48304517647058826f,
+        0.8639751568627451f, 0.757323862745098f, 0.4798156862745098f,
+        0.8619759019607843f, 0.7536329607843137f, 0.4765861960784314f,
+        0.8599766470588235f, 0.7499420588235295f, 0.47335670588235296f,
+        0.8579773921568628f, 0.7462511568627451f, 0.4701272156862745f,
+        0.8559781372549019f, 0.7425602549019609f, 0.46689772549019615f,
+        0.8539788823529412f, 0.7388693529411765f, 0.46366823529411766f,
+        0.8519796274509803f, 0.7351784509803921f, 0.4604387450980392f,
+        0.8499803921568627f, 0.7316413529411765f, 0.4576705882352941f,
+        0.8479811764705882f, 0.7282580588235295f, 0.45536376470588236f,
+        0.8459819607843136f, 0.7248747647058824f, 0.4530569411764706f,
+        0.8439827450980392f, 0.7214914705882353f, 0.45075011764705886f,
+        0.8419835294117647f, 0.7181081764705882f, 0.44844329411764705f,
+        0.8399843137254902f, 0.7147248823529413f, 0.44613647058823536f,
+        0.8379850980392156f, 0.7113415882352941f, 0.44382964705882355f,
+        0.8359858823529411f, 0.7079582941176471f, 0.4415228235294118f,
+        0.8339866666666667f, 0.7045750000000001f, 0.439216f,
+        0.8319874509803922f, 0.7011917058823529f, 0.43690917647058825f,
+        0.8299882352941177f, 0.6978084117647059f, 0.4346023529411765f,
+        0.8279890196078431f, 0.6944251176470588f, 0.43229552941176475f,
+        0.8259898039215686f, 0.6910418235294118f, 0.42998870588235294f,
+        0.8239905882352941f, 0.6876585294117648f, 0.42768188235294124f,
+        0.8219913725490197f, 0.6842752352941177f, 0.42537505882352944f,
+        0.8199921568627451f, 0.6808919411764707f, 0.4230682352941177f,
+        0.8179929411764706f, 0.6775086470588235f, 0.4207614117647059f,
+        0.8159937254901961f, 0.6741253529411765f, 0.41845458823529413f,
+        0.8139945098039216f, 0.6707420588235294f, 0.4161477647058824f,
+        0.811995294117647f, 0.6673587647058824f, 0.41384094117647063f,
+        0.8099960784313726f, 0.6639754705882354f, 0.41153411764705883f,
+        0.8079968627450981f, 0.6605921764705883f, 0.40922729411764713f,
+        0.8059976470588236f, 0.6572088823529412f, 0.40692047058823533f,
+        0.803998431372549f, 0.6538255882352941f, 0.4046136470588236f,
+        0.8019992156862745f, 0.6504422941176471f, 0.4023068235294118f,
+        0.8f, 0.647059f, 0.4f,
+        0.7984621176470589f, 0.6436756666666668f, 0.3980007843137255f,
+        0.7969242352941177f, 0.6402923333333334f, 0.396001568627451f,
+        0.7953863529411765f, 0.6369090000000001f, 0.39400235294117647f,
+        0.7938484705882354f, 0.6335256666666668f, 0.392003137254902f,
+        0.7923105882352942f, 0.6301423333333334f, 0.39000392156862745f,
+        0.7907727058823529f, 0.6267590000000001f, 0.388004705882353f,
+        0.7892348235294118f, 0.6233756666666667f, 0.38600549019607844f,
+        0.7876969411764706f, 0.6199923333333334f, 0.38400627450980396f,
+        0.7861590588235294f, 0.616609f, 0.3820070588235294f,
+        0.7846211764705883f, 0.6132256666666667f, 0.3800078431372549f,
+        0.7830832941176471f, 0.6098423333333334f, 0.3780086274509804f,
+        0.7815454117647059f, 0.6064590000000001f, 0.3760094117647059f,
+        0.7800075294117648f, 0.6030756666666667f, 0.3740101960784314f,
+        0.7784696470588236f, 0.5996923333333333f, 0.37201098039215685f,
+        0.7769317647058824f, 0.596309f, 0.3700117647058824f,
+        0.7753938823529412f, 0.5929256666666667f, 0.36801254901960784f,
+        0.773856f, 0.5895423333333333f, 0.36601333333333336f,
+        0.7723181176470588f, 0.586159f, 0.3640141176470588f,
+        0.7707802352941177f, 0.5827756666666666f, 0.36201490196078434f,
+        0.7692423529411765f, 0.5793923333333334f, 0.36001568627450986f,
+        0.7677044705882353f, 0.576009f, 0.35801647058823527f,
+        0.7661665882352942f, 0.5726256666666666f, 0.3560172549019608f,
+        0.764628705882353f, 0.5692423333333333f, 0.35401803921568625f,
+        0.7630908235294118f, 0.565859f, 0.3520188235294118f,
+        0.7615529411764707f, 0.5624756666666666f, 0.35001960784313724f,
+        0.7592461568627451f, 0.5587847843137255f, 0.34748211764705883f,
+        0.7561704705882353f, 0.5547863529411764f, 0.3444063529411765f,
+        0.7530947843137255f, 0.5507879215686274f, 0.3413305882352941f,
+        0.7500190980392157f, 0.5467894901960784f, 0.3382548235294118f,
+        0.746943411764706f, 0.5427910588235295f, 0.3351790588235295f,
+        0.7438677254901961f, 0.5387926274509803f, 0.33210329411764705f,
+        0.7407920392156863f, 0.5347941960784314f, 0.3290275294117647f,
+        0.7377163529411764f, 0.5307957647058823f, 0.3259517647058824f,
+        0.7346406666666667f, 0.5267973333333333f, 0.322876f,
+        0.7315649803921569f, 0.5227989019607844f, 0.31980023529411766f,
+        0.7284892941176471f, 0.5188004705882353f, 0.31672447058823533f,
+        0.7254136078431372f, 0.5148020392156863f, 0.31364870588235294f,
+        0.7223379215686274f, 0.5108036078431373f, 0.3105729411764706f,
+        0.7197235294117648f, 0.5070358235294118f, 0.3077278823529412f,
+        0.7172629019607843f, 0.5033449215686275f, 0.3049597254901961f,
+        0.7148022745098039f, 0.49965401960784317f, 0.302191568627451f,
+        0.7123416470588235f, 0.49596311764705886f, 0.2994234117647059f,
+        0.7098810196078431f, 0.49227221568627455f, 0.29665525490196076f,
+        0.7074203921568628f, 0.48858131372549024f, 0.2938870980392157f,
+        0.7049597647058824f, 0.48489041176470593f, 0.2911189411764706f,
+        0.7024991372549021f, 0.4811995098039217f, 0.28835078431372557f,
+        0.7000385098039216f, 0.47750860784313726f, 0.2855826274509804f,
+        0.6975778823529412f, 0.47381770588235295f, 0.2828144705882353f,
+        0.6951172549019609f, 0.47012680392156864f, 0.28004631372549016f,
+        0.6926566274509804f, 0.4664359019607843f, 0.2772781568627451f,
+        0.690196f, 0.462745f, 0.27451f,
+        0.6877354509803922f, 0.4599768431372549f, 0.27512513725490195f,
+        0.6852749019607843f, 0.45720868627450983f, 0.27574027450980393f,
+        0.6828143529411765f, 0.4544405294117647f, 0.27635541176470585f,
+        0.6803538039215686f, 0.45167237254901965f, 0.2769705490196078f,
+        0.6778932549019608f, 0.44890421568627453f, 0.2775856862745098f,
+        0.6754327058823529f, 0.4461360588235294f, 0.2782008235294118f,
+        0.6729721568627451f, 0.44336790196078435f, 0.2788159607843137f,
+        0.6705116078431372f, 0.44059974509803923f, 0.2794310980392157f,
+        0.6680510588235294f, 0.4378315882352941f, 0.28004623529411765f,
+        0.6655905098039215f, 0.43506343137254905f, 0.2806613725490196f,
+        0.6631299607843137f, 0.43229527450980404f, 0.28127650980392155f,
+        0.6606694117647058f, 0.42952711764705886f, 0.2818916470588235f,
+        0.6582857450980392f, 0.42698964705882353f, 0.2825067843137255f,
+        0.6561327254901961f, 0.42514423529411766f, 0.2831219215686275f,
+        0.6539797058823529f, 0.4232988235294118f, 0.28373705882352945f,
+        0.6518266862745098f, 0.4214534117647059f, 0.2843521960784314f,
+        0.6496736666666666f, 0.41960800000000004f, 0.28496733333333335f,
+        0.6475206470588235f, 0.4177625882352941f, 0.2855824705882353f,
+        0.6453676274509803f, 0.41591717647058823f, 0.28619760784313725f,
+        0.6432146078431372f, 0.41407176470588236f, 0.2868127450980392f,
+        0.6410615882352941f, 0.4122263529411765f, 0.2874278823529412f,
+        0.6389085686274509f, 0.4103809411764706f, 0.2880430196078432f,
+        0.6367555490196078f, 0.40853552941176474f, 0.2886581568627451f,
+        0.6346025294117646f, 0.4066901176470588f, 0.2892732941176471f,
+        0.6324495098039216f, 0.40484470588235294f, 0.28988843137254905f,
+        0.6301426862745098f, 0.4029992549019608f, 0.2903498039215686f,
+        0.6276820588235295f, 0.40115376470588243f, 0.2906574117647059f,
+        0.6252214313725489f, 0.39930827450980394f, 0.29096501960784316f,
+        0.6227608039215686f, 0.3974627843137255f, 0.29127262745098037f,
+        0.6203001764705882f, 0.39561729411764707f, 0.29158023529411764f,
+        0.6178395490196078f, 0.39377180392156863f, 0.2918878431372549f,
+        0.6153789215686274f, 0.3919263137254902f, 0.2921954509803922f,
+        0.612918294117647f, 0.39008082352941176f, 0.2925030588235294f,
+        0.6104576666666667f, 0.3882353333333333f, 0.29281066666666666f,
+        0.6079970392156863f, 0.3863898431372549f, 0.29311827450980393f,
+        0.6055364117647059f, 0.3845443529411765f, 0.29342588235294115f,
+        0.6030757843137254f, 0.38269886274509807f, 0.2937334901960784f,
+        0.600615156862745f, 0.38085337254901963f, 0.2940410980392157f,
+        0.5970011764705883f, 0.37877723529411766f, 0.2936566470588235f,
+        0.5930027450980392f, 0.3766242156862745f, 0.2930415098039216f,
+        0.5890043137254901f, 0.3744711960784314f, 0.2924263725490196f,
+        0.5850058823529412f, 0.37231817647058824f, 0.2918112352941176f,
+        0.5810074509803922f, 0.37016515686274515f, 0.2911960980392157f,
+        0.5770090196078431f, 0.36801213725490195f, 0.2905809607843137f,
+        0.573010588235294f, 0.3658591176470588f, 0.28996582352941175f,
+        0.5690121568627451f, 0.3637060980392157f, 0.2893506862745098f,
+        0.565013725490196f, 0.3615530784313726f, 0.28873554901960785f,
+        0.5610152941176471f, 0.35940005882352943f, 0.2881204117647059f,
+        0.557016862745098f, 0.3572470392156863f, 0.2875052745098039f,
+        0.553018431372549f, 0.35509401960784315f, 0.286890137254902f,
+        0.54902f, 0.352941f, 0.286275f,
+        0.5459442352941176f, 0.35078798039215686f, 0.28596739215686273f,
+        0.5428684705882353f, 0.3486349607843137f, 0.2856597843137255f,
+        0.539792705882353f, 0.3464819411764706f, 0.28535217647058825f,
+        0.5367169411764706f, 0.34432892156862743f, 0.285044568627451f,
+        0.5336411764705882f, 0.3421759019607843f, 0.2847369607843137f,
+        0.5305654117647058f, 0.3400228823529412f, 0.2844293529411765f,
+        0.5274896470588235f, 0.33786986274509806f, 0.2841217450980392f,
+        0.5244138823529413f, 0.33571684313725497f, 0.283814137254902f,
+        0.5213381176470588f, 0.3335638235294118f, 0.28350652941176474f,
+        0.5182623529411765f, 0.33141080392156863f, 0.28319892156862747f,
+        0.5151865882352942f, 0.3292577843137255f, 0.2828913137254902f,
+        0.5121108235294118f, 0.32710476470588234f, 0.282583705882353f,
+        0.5090350588235294f, 0.3250286470588235f, 0.28219921568627454f,
+        0.5059592941176471f, 0.32318323529411763f, 0.28158407843137256f,
+        0.5028835294117647f, 0.32133782352941176f, 0.2809689411764706f,
+        0.4998077647058824f, 0.3194924117647059f, 0.2803538039215686f,
+        0.496732f, 0.317647f, 0.2797386666666667f,
+        0.4936562352941177f, 0.3158015882352941f, 0.2791235294117647f,
+        0.49058047058823534f, 0.3139561764705882f, 0.27850839215686274f,
+        0.48750470588235295f, 0.31211076470588234f, 0.27789325490196076f,
+        0.4844289411764706f, 0.31026535294117646f, 0.27727811764705884f,
+        0.48135317647058823f, 0.3084199411764706f, 0.27666298039215687f,
+        0.4782774117647059f, 0.30657452941176466f, 0.2760478431372549f,
+        0.4752016470588236f, 0.30472911764705884f, 0.2754327058823529f,
+        0.4721258823529412f, 0.3028837058823529f, 0.27481756862745094f,
+        0.4690501568627451f, 0.30088449019607844f, 0.27404862745098035f,
+        0.4659744705882353f, 0.2987314705882353f, 0.27312588235294116f,
+        0.4628987843137255f, 0.29657845098039215f, 0.2722031372549019f,
+        0.45982309803921567f, 0.294425431372549f, 0.2712803921568627f,
+        0.4567474117647059f, 0.29227241176470586f, 0.27035764705882354f,
+        0.4536717254901961f, 0.2901193921568627f, 0.2694349019607843f,
+        0.4505960392156863f, 0.2879663725490196f, 0.2685121568627451f,
+        0.4475203529411765f, 0.28581335294117644f, 0.26758941176470585f,
+        0.44444466666666665f, 0.2836603333333333f, 0.26666666666666666f,
+        0.4413689803921569f, 0.28150731372549015f, 0.2657439215686275f,
+        0.43829329411764706f, 0.27935429411764706f, 0.2648211764705882f,
+        0.43521760784313723f, 0.2772012745098039f, 0.26389843137254904f,
+        0.43214192156862746f, 0.2750482549019608f, 0.2629756862745098f,
+        0.42837411764705885f, 0.2733565882352941f, 0.26182229411764707f,
+        0.4243756078431374f, 0.27181870588235296f, 0.2605920196078432f,
+        0.4203770980392157f, 0.27028082352941174f, 0.2593617450980392f,
+        0.4163785882352941f, 0.2687429411764706f, 0.2581314705882353f,
+        0.41238007843137253f, 0.2672050588235294f, 0.25690119607843137f,
+        0.408381568627451f, 0.26566717647058824f, 0.2556709215686275f,
+        0.40438305882352943f, 0.2641292941176471f, 0.2544406470588235f,
+        0.40038454901960785f, 0.2625914117647059f, 0.2532103725490196f,
+        0.39638603921568627f, 0.2610535294117647f, 0.25198009803921567f,
+        0.3923875294117647f, 0.2595156470588235f, 0.2507498235294118f,
+        0.38838901960784317f, 0.25797776470588235f, 0.24951954901960785f,
+        0.3843905098039216f, 0.2564398823529412f, 0.24828927450980393f,
+        0.380392f, 0.254902f, 0.247059f,
+    },
+
+    // one channel red for 
+    {
+        0.0f, 0.f, 0.f, 
+        0.00392156862745098f, 0.f, 0.f, 
+        0.00784313725490196f, 0.f, 0.f, 
+        0.011764705882352941f, 0.f, 0.f, 
+        0.01568627450980392f, 0.f, 0.f, 
+        0.0196078431372549f, 0.f, 0.f, 
+        0.023529411764705882f, 0.f, 0.f, 
+        0.027450980392156862f, 0.f, 0.f, 
+        0.03137254901960784f, 0.f, 0.f, 
+        0.03529411764705882f, 0.f, 0.f, 
+        0.0392156862745098f, 0.f, 0.f, 
+        0.043137254901960784f, 0.f, 0.f, 
+        0.047058823529411764f, 0.f, 0.f, 
+        0.050980392156862744f, 0.f, 0.f, 
+        0.054901960784313725f, 0.f, 0.f, 
+        0.058823529411764705f, 0.f, 0.f, 
+        0.06274509803921569f, 0.f, 0.f, 
+        0.06666666666666667f, 0.f, 0.f, 
+        0.07058823529411765f, 0.f, 0.f, 
+        0.07450980392156863f, 0.f, 0.f, 
+        0.0784313725490196f, 0.f, 0.f, 
+        0.08235294117647059f, 0.f, 0.f, 
+        0.08627450980392157f, 0.f, 0.f, 
+        0.09019607843137255f, 0.f, 0.f, 
+        0.09411764705882353f, 0.f, 0.f, 
+        0.09803921568627451f, 0.f, 0.f, 
+        0.10196078431372549f, 0.f, 0.f, 
+        0.10588235294117647f, 0.f, 0.f, 
+        0.10980392156862745f, 0.f, 0.f, 
+        0.11372549019607843f, 0.f, 0.f, 
+        0.11764705882352941f, 0.f, 0.f, 
+        0.12156862745098039f, 0.f, 0.f, 
+        0.12549019607843137f, 0.f, 0.f, 
+        0.12941176470588234f, 0.f, 0.f, 
+        0.13333333333333333f, 0.f, 0.f, 
+        0.13725490196078433f, 0.f, 0.f, 
+        0.1411764705882353f, 0.f, 0.f, 
+        0.14509803921568626f, 0.f, 0.f, 
+        0.14901960784313725f, 0.f, 0.f, 
+        0.15294117647058825f, 0.f, 0.f, 
+        0.1568627450980392f, 0.f, 0.f, 
+        0.16078431372549018f, 0.f, 0.f, 
+        0.16470588235294117f, 0.f, 0.f, 
+        0.16862745098039217f, 0.f, 0.f, 
+        0.17254901960784313f, 0.f, 0.f, 
+        0.1764705882352941f, 0.f, 0.f, 
+        0.1803921568627451f, 0.f, 0.f, 
+        0.1843137254901961f, 0.f, 0.f, 
+        0.18823529411764706f, 0.f, 0.f, 
+        0.19215686274509802f, 0.f, 0.f, 
+        0.19607843137254902f, 0.f, 0.f, 
+        0.2f, 0.f, 0.f, 
+        0.20392156862745098f, 0.f, 0.f, 
+        0.20784313725490194f, 0.f, 0.f, 
+        0.21176470588235294f, 0.f, 0.f, 
+        0.21568627450980393f, 0.f, 0.f, 
+        0.2196078431372549f, 0.f, 0.f, 
+        0.22352941176470587f, 0.f, 0.f, 
+        0.22745098039215686f, 0.f, 0.f, 
+        0.23137254901960785f, 0.f, 0.f, 
+        0.23529411764705882f, 0.f, 0.f, 
+        0.2392156862745098f, 0.f, 0.f, 
+        0.24313725490196078f, 0.f, 0.f, 
+        0.24705882352941178f, 0.f, 0.f, 
+        0.25098039215686274f, 0.f, 0.f, 
+        0.2549019607843137f, 0.f, 0.f, 
+        0.2588235294117647f, 0.f, 0.f, 
+        0.2627450980392157f, 0.f, 0.f, 
+        0.26666666666666666f, 0.f, 0.f, 
+        0.27058823529411763f, 0.f, 0.f, 
+        0.27450980392156865f, 0.f, 0.f, 
+        0.2784313725490196f, 0.f, 0.f, 
+        0.2823529411764706f, 0.f, 0.f, 
+        0.28627450980392155f, 0.f, 0.f, 
+        0.2901960784313725f, 0.f, 0.f, 
+        0.29411764705882354f, 0.f, 0.f, 
+        0.2980392156862745f, 0.f, 0.f, 
+        0.30196078431372547f, 0.f, 0.f, 
+        0.3058823529411765f, 0.f, 0.f, 
+        0.30980392156862746f, 0.f, 0.f, 
+        0.3137254901960784f, 0.f, 0.f, 
+        0.3176470588235294f, 0.f, 0.f, 
+        0.32156862745098036f, 0.f, 0.f, 
+        0.3254901960784314f, 0.f, 0.f, 
+        0.32941176470588235f, 0.f, 0.f, 
+        0.3333333333333333f, 0.f, 0.f, 
+        0.33725490196078434f, 0.f, 0.f, 
+        0.3411764705882353f, 0.f, 0.f, 
+        0.34509803921568627f, 0.f, 0.f, 
+        0.34901960784313724f, 0.f, 0.f, 
+        0.3529411764705882f, 0.f, 0.f, 
+        0.3568627450980392f, 0.f, 0.f, 
+        0.3607843137254902f, 0.f, 0.f, 
+        0.36470588235294116f, 0.f, 0.f, 
+        0.3686274509803922f, 0.f, 0.f, 
+        0.37254901960784315f, 0.f, 0.f, 
+        0.3764705882352941f, 0.f, 0.f, 
+        0.3803921568627451f, 0.f, 0.f, 
+        0.38431372549019605f, 0.f, 0.f, 
+        0.38823529411764707f, 0.f, 0.f, 
+        0.39215686274509803f, 0.f, 0.f, 
+        0.396078431372549f, 0.f, 0.f, 
+        0.4f, 0.f, 0.f, 
+        0.403921568627451f, 0.f, 0.f, 
+        0.40784313725490196f, 0.f, 0.f, 
+        0.4117647058823529f, 0.f, 0.f, 
+        0.4156862745098039f, 0.f, 0.f, 
+        0.4196078431372549f, 0.f, 0.f, 
+        0.4235294117647059f, 0.f, 0.f, 
+        0.42745098039215684f, 0.f, 0.f, 
+        0.43137254901960786f, 0.f, 0.f, 
+        0.43529411764705883f, 0.f, 0.f, 
+        0.4392156862745098f, 0.f, 0.f, 
+        0.44313725490196076f, 0.f, 0.f, 
+        0.44705882352941173f, 0.f, 0.f, 
+        0.45098039215686275f, 0.f, 0.f, 
+        0.4549019607843137f, 0.f, 0.f, 
+        0.4588235294117647f, 0.f, 0.f, 
+        0.4627450980392157f, 0.f, 0.f, 
+        0.4666666666666667f, 0.f, 0.f, 
+        0.47058823529411764f, 0.f, 0.f, 
+        0.4745098039215686f, 0.f, 0.f, 
+        0.4784313725490196f, 0.f, 0.f, 
+        0.4823529411764706f, 0.f, 0.f, 
+        0.48627450980392156f, 0.f, 0.f, 
+        0.49019607843137253f, 0.f, 0.f, 
+        0.49411764705882355f, 0.f, 0.f, 
+        0.4980392156862745f, 0.f, 0.f, 
+        0.5019607843137255f, 0.f, 0.f, 
+        0.5058823529411764f, 0.f, 0.f, 
+        0.5098039215686274f, 0.f, 0.f, 
+        0.5137254901960784f, 0.f, 0.f, 
+        0.5176470588235293f, 0.f, 0.f, 
+        0.5215686274509804f, 0.f, 0.f, 
+        0.5254901960784314f, 0.f, 0.f, 
+        0.5294117647058824f, 0.f, 0.f, 
+        0.5333333333333333f, 0.f, 0.f, 
+        0.5372549019607843f, 0.f, 0.f, 
+        0.5411764705882353f, 0.f, 0.f, 
+        0.5450980392156862f, 0.f, 0.f, 
+        0.5490196078431373f, 0.f, 0.f, 
+        0.5529411764705883f, 0.f, 0.f, 
+        0.5568627450980392f, 0.f, 0.f, 
+        0.5607843137254902f, 0.f, 0.f, 
+        0.5647058823529412f, 0.f, 0.f, 
+        0.5686274509803921f, 0.f, 0.f, 
+        0.5725490196078431f, 0.f, 0.f, 
+        0.5764705882352941f, 0.f, 0.f, 
+        0.580392156862745f, 0.f, 0.f, 
+        0.5843137254901961f, 0.f, 0.f, 
+        0.5882352941176471f, 0.f, 0.f, 
+        0.592156862745098f, 0.f, 0.f, 
+        0.596078431372549f, 0.f, 0.f, 
+        0.6f, 0.f, 0.f, 
+        0.6039215686274509f, 0.f, 0.f, 
+        0.6078431372549019f, 0.f, 0.f, 
+        0.611764705882353f, 0.f, 0.f, 
+        0.615686274509804f, 0.f, 0.f, 
+        0.6196078431372549f, 0.f, 0.f, 
+        0.6235294117647059f, 0.f, 0.f, 
+        0.6274509803921569f, 0.f, 0.f, 
+        0.6313725490196078f, 0.f, 0.f, 
+        0.6352941176470588f, 0.f, 0.f, 
+        0.6392156862745098f, 0.f, 0.f, 
+        0.6431372549019607f, 0.f, 0.f, 
+        0.6470588235294118f, 0.f, 0.f, 
+        0.6509803921568628f, 0.f, 0.f, 
+        0.6549019607843137f, 0.f, 0.f, 
+        0.6588235294117647f, 0.f, 0.f, 
+        0.6627450980392157f, 0.f, 0.f, 
+        0.6666666666666666f, 0.f, 0.f, 
+        0.6705882352941176f, 0.f, 0.f, 
+        0.6745098039215687f, 0.f, 0.f, 
+        0.6784313725490196f, 0.f, 0.f, 
+        0.6823529411764706f, 0.f, 0.f, 
+        0.6862745098039216f, 0.f, 0.f, 
+        0.6901960784313725f, 0.f, 0.f, 
+        0.6941176470588235f, 0.f, 0.f, 
+        0.6980392156862745f, 0.f, 0.f, 
+        0.7019607843137254f, 0.f, 0.f, 
+        0.7058823529411764f, 0.f, 0.f, 
+        0.7098039215686275f, 0.f, 0.f, 
+        0.7137254901960784f, 0.f, 0.f, 
+        0.7176470588235294f, 0.f, 0.f, 
+        0.7215686274509804f, 0.f, 0.f, 
+        0.7254901960784313f, 0.f, 0.f, 
+        0.7294117647058823f, 0.f, 0.f, 
+        0.7333333333333333f, 0.f, 0.f, 
+        0.7372549019607844f, 0.f, 0.f, 
+        0.7411764705882353f, 0.f, 0.f, 
+        0.7450980392156863f, 0.f, 0.f, 
+        0.7490196078431373f, 0.f, 0.f, 
+        0.7529411764705882f, 0.f, 0.f, 
+        0.7568627450980392f, 0.f, 0.f, 
+        0.7607843137254902f, 0.f, 0.f, 
+        0.7647058823529411f, 0.f, 0.f, 
+        0.7686274509803921f, 0.f, 0.f, 
+        0.7725490196078432f, 0.f, 0.f, 
+        0.7764705882352941f, 0.f, 0.f, 
+        0.7803921568627451f, 0.f, 0.f, 
+        0.7843137254901961f, 0.f, 0.f, 
+        0.788235294117647f, 0.f, 0.f, 
+        0.792156862745098f, 0.f, 0.f, 
+        0.796078431372549f, 0.f, 0.f, 
+        0.8f, 0.f, 0.f, 
+        0.803921568627451f, 0.f, 0.f, 
+        0.807843137254902f, 0.f, 0.f, 
+        0.8117647058823529f, 0.f, 0.f, 
+        0.8156862745098039f, 0.f, 0.f, 
+        0.8196078431372549f, 0.f, 0.f, 
+        0.8235294117647058f, 0.f, 0.f, 
+        0.8274509803921568f, 0.f, 0.f, 
+        0.8313725490196078f, 0.f, 0.f, 
+        0.8352941176470589f, 0.f, 0.f, 
+        0.8392156862745098f, 0.f, 0.f, 
+        0.8431372549019608f, 0.f, 0.f, 
+        0.8470588235294118f, 0.f, 0.f, 
+        0.8509803921568627f, 0.f, 0.f, 
+        0.8549019607843137f, 0.f, 0.f, 
+        0.8588235294117647f, 0.f, 0.f, 
+        0.8627450980392157f, 0.f, 0.f, 
+        0.8666666666666667f, 0.f, 0.f, 
+        0.8705882352941177f, 0.f, 0.f, 
+        0.8745098039215686f, 0.f, 0.f, 
+        0.8784313725490196f, 0.f, 0.f, 
+        0.8823529411764706f, 0.f, 0.f, 
+        0.8862745098039215f, 0.f, 0.f, 
+        0.8901960784313725f, 0.f, 0.f, 
+        0.8941176470588235f, 0.f, 0.f, 
+        0.8980392156862745f, 0.f, 0.f, 
+        0.9019607843137255f, 0.f, 0.f, 
+        0.9058823529411765f, 0.f, 0.f, 
+        0.9098039215686274f, 0.f, 0.f, 
+        0.9137254901960784f, 0.f, 0.f, 
+        0.9176470588235294f, 0.f, 0.f, 
+        0.9215686274509803f, 0.f, 0.f, 
+        0.9254901960784314f, 0.f, 0.f, 
+        0.9294117647058824f, 0.f, 0.f, 
+        0.9333333333333333f, 0.f, 0.f, 
+        0.9372549019607843f, 0.f, 0.f, 
+        0.9411764705882353f, 0.f, 0.f, 
+        0.9450980392156862f, 0.f, 0.f, 
+        0.9490196078431372f, 0.f, 0.f, 
+        0.9529411764705882f, 0.f, 0.f, 
+        0.9568627450980391f, 0.f, 0.f, 
+        0.9607843137254902f, 0.f, 0.f, 
+        0.9647058823529412f, 0.f, 0.f, 
+        0.9686274509803922f, 0.f, 0.f, 
+        0.9725490196078431f, 0.f, 0.f, 
+        0.9764705882352941f, 0.f, 0.f, 
+        0.9803921568627451f, 0.f, 0.f, 
+        0.984313725490196f, 0.f, 0.f, 
+        0.9882352941176471f, 0.f, 0.f, 
+        0.9921568627450981f, 0.f, 0.f, 
+        0.996078431372549f, 0.f, 0.f, 
+        1.0f, 0.f, 0.f, 
+    }
+};
+
+
+// global_variable
+const char* Color_Map_Names[Number_of_Color_Maps] = 
+{   
+    "light_sunset (colorblind)",
+    "viridis (colorblind)",
+    "cividis (colorblind)", 
+    "Inferno",
+    "Black Body",
+    "Kindlmann",
+    "Extended Kindlmann",
+    "Three Wave Yellow-Grey-Blue",
+    "Three Wave Blue-Green-Yellow",
+    "Four Wave Grey-Red-Green",
+    "Five Wave Yellow-Brown-Blue",
+    "Blue 1",
+    "Blue 2",
+    "Blue 3",
+    "Blue 4",
+    "Blue-Orange Divergent",
+    "Brown 1",
+    "Brown 2",
+    "Brown 3",
+    "Green 1",
+    "Green 2",
+    "Green 3",
+    "Green 4",
+    "Mellow Rainbow",
+    "Orange 1",
+    "Orange 2",
+    "Orange-Green-Blue-Gray",
+    "Purple 1",
+    "Purple 2",
+    "Red 1",
+    "Red 2",
+    "Red 3",
+    "Yellow 1",
+    "Yellow 2", 
+    "Copy_texture(always at end)"
+};
\ No newline at end of file

From 35d58047fcffd00fea4104232061e3ed5a212f99 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:08:04 +0100
Subject: [PATCH 16/33] update copy_texture_to_cpu and weight for link score
 calculate

---
 src/copy_texture.cpp | 212 ++++++++++++++++++++++++++++++-------------
 src/copy_texture.h   |  39 +++++---
 2 files changed, 176 insertions(+), 75 deletions(-)

diff --git a/src/copy_texture.cpp b/src/copy_texture.cpp
index c41e787..81ae13e 100644
--- a/src/copy_texture.cpp
+++ b/src/copy_texture.cpp
@@ -24,6 +24,7 @@ SOFTWARE.
 #include "copy_texture.h"
 #include "shaderSource.h"
 #include "utilsPretextView.h"
+#include "cmath"
 
 
 Show_State::Show_State()
@@ -39,10 +40,15 @@ Show_State::~Show_State() {}
 
 
 void get_linear_mask(f32* linear_array, u32 length)
-{
+{   
+    if (length < 2)
+    {
+        fmt::print(stderr, "Error: link score calculation: length({}) < 2\n", length);
+        assert(0);
+    }
     for (u32 i = 0; i < length; i++)
     {
-        linear_array[i] = 1.0f - (f32)i / (f32)length;
+        linear_array[i] = std::sqrt(1.0f / (f32)(i+1)) ;
     }
     f32 sum = std::accumulate(linear_array, linear_array+length, 0.0f);
     for (int i  = 0 ; i < length; i++)
@@ -121,15 +127,16 @@ void cursor_position_callback(GLFWwindow* window, double xpos, double ypos)
 
 
 TexturesArray4AI::TexturesArray4AI(
-    u32 num_textures_1d_, u32 texture_resolution_, char* fileName, const contigs* Contigs)
+    u32 num_textures_1d_, 
+    u32 texture_resolution_, 
+    char* fileName, 
+    const contigs* Contigs)
     :num_textures_1d(num_textures_1d_), texture_resolution(texture_resolution_),
     is_copied_from_buffer(false), is_compressed(false), hic_shader_initilised(false)
 {   
     
     this->frags = new Frag4compress(Contigs);
 
-    this->compressed_hic = new Matrix3D<f32>(1, 1, 5);  // initialize the compressed_hic with 1x1x5
-
     this->compressed_extensions = new CompressedExtensions(1); // initialize the compressed_extensions with 1
 
     char* tmp = fileName;
@@ -147,30 +154,10 @@ TexturesArray4AI::TexturesArray4AI(
     }
     this->texture_resolution_exp = (u32)std::log2(this->texture_resolution);
     this->total_num_textures = ((1 + this->num_textures_1d) * this->num_textures_1d ) >> 1;
-    this->textures = new u08*[this->total_num_textures];
-    try
-    {
-        for (u32 i = 0; i < this->total_num_textures; i++)
-        {
-            this->textures[i] = new u08[this->texture_resolution * this->texture_resolution];
-        }
-    }
-    catch(const std::exception& e)
-    {   
-        for (u32 i = 0; i < this->total_num_textures; i++)
-        {
-            if (this->textures[i])
-            {
-                delete[] this->textures[i];
-                this->textures[i] = nullptr;
-            }
-        }
-        delete[] this->textures;
-        this->textures = nullptr;
-        std::cerr << e.what() << " Allocate mem error\n";
-        assert(0);
-    }
     
+    // 此处不再为copy textures申请空间，因为耗费太多内存
+    // this->initialise_textures();
+
     // Shader setup and texture binding
     this->shaderProgram = CreateShader(FragmentSource_copyTexture.c_str(), VertexSource_copyTexture.c_str());
 
@@ -209,21 +196,56 @@ TexturesArray4AI::TexturesArray4AI(
 }
 
 
-TexturesArray4AI::~TexturesArray4AI()
+void TexturesArray4AI::initialise_textures()
 {   
-    if (textures)
+    if (this->textures) this->clear_textures();
+    this->textures = new u08*[this->total_num_textures];
+    try
     {
-        for (u32 i = 0; i < total_num_textures; i++)
+        for (u32 i = 0; i < this->total_num_textures; i++)
         {
-            if (textures[i]) 
-            {
-                delete[] textures[i];
-                textures[i] = nullptr;
-            }
+            this->textures[i] = new u08[this->texture_resolution * this->texture_resolution];
+        }
+    }
+    catch(const std::exception& e)
+    {   
+        this->clear_textures();
+        std::cerr << e.what() << " Allocate mem error\n";
+        assert(0);
+    }
+}
+
+void TexturesArray4AI::clear_textures()
+{
+    for (u32 i = 0; i < this->total_num_textures; i++)
+    {
+        if (this->textures && this->textures[i])
+        {
+            delete[] this->textures[i];
+            this->textures[i] = nullptr;
         }
-        delete[] textures;
-        textures = nullptr;
     }
+    delete[] this->textures;
+    this->textures = nullptr;
+    is_copied_from_buffer = false;
+}
+
+
+void TexturesArray4AI::clear_compressed_hic()
+{
+    if (this->compressed_hic)
+    {
+        delete this->compressed_hic;
+        this->compressed_hic = nullptr;
+    }
+    return;
+}
+
+
+TexturesArray4AI::~TexturesArray4AI()
+{   
+    this->clear_textures();
+    this->clear_compressed_hic();
 
     if (frags)
     {
@@ -231,12 +253,6 @@ TexturesArray4AI::~TexturesArray4AI()
         frags = nullptr;
     }
     
-    if (compressed_hic)
-    {
-        delete compressed_hic;
-        compressed_hic = nullptr;
-    }
-
     if (compressed_extensions)
     {
         delete compressed_extensions;
@@ -321,11 +337,17 @@ void TexturesArray4AI::check_copied_from_buffer() const
     return ;
 }
 
+/* 
+复制没有经过re-order的buffer到 this->textures 中
+其实就是存储在.pretext 文件中的textures
+*/
 void TexturesArray4AI::copy_buffer_to_textures(
     const contact_matrix *contact_matrix_, 
     bool show_flag) 
 {   
 
+    this->initialise_textures(); // 申请空间 textures
+
     const GLuint &contact_matrix_textures = contact_matrix_->textures;
     if (is_copied_from_buffer) return;
     
@@ -350,8 +372,6 @@ void TexturesArray4AI::copy_buffer_to_textures(
         assert(0);
     }
 
-    
-
     s32 orignal_viewport[4];
     glGetIntegerv(GL_VIEWPORT, orignal_viewport);
     GLcall(glActiveTexture(GL_TEXTURE0));
@@ -402,12 +422,16 @@ void TexturesArray4AI::copy_buffer_to_textures(
 }
 
 
-
+/*
+这个会把操作修改后的像素拷贝到 this->textures 中
+*/
 void TexturesArray4AI::copy_buffer_to_textures_dynamic(
     const contact_matrix *contact_matrix_, 
     bool show_flag) 
 {
 
+    this->initialise_textures(); // 申请空间textures
+
     GLuint temptexture;
     glGenTextures(1, &temptexture);
     glBindTexture(GL_TEXTURE_2D, temptexture);
@@ -443,8 +467,8 @@ void TexturesArray4AI::copy_buffer_to_textures_dynamic(
                 ptr_cnt++;
                 continue;
             }
-            // GLcall(glBindFramebuffer(GL_DRAW_FRAMEBUFFER, framebuffer));
-            GLcall(glBindFramebuffer(GL_DRAW_FRAMEBUFFER, 0));
+            GLcall(glBindFramebuffer(GL_DRAW_FRAMEBUFFER, framebuffer));
+            // GLcall(glBindFramebuffer(GL_DRAW_FRAMEBUFFER, 0));
             GLcall(glViewport(0, 0, texture_resolution, texture_resolution));
             GLcall(glClearColor(0.3f, 0.3f, 0.3f, 1.0f));
             GLcall(glClear(GL_COLOR_BUFFER_BIT));
@@ -456,12 +480,11 @@ void TexturesArray4AI::copy_buffer_to_textures_dynamic(
             // GLcall(glBindVertexArray(this->vao));
             // GLcall(glDrawElements(GL_TRIANGLES, 6, GL_UNSIGNED_INT, 0));
 
-            glBindTexture(GL_TEXTURE_2D_ARRAY, contact_matrix_->textures);
             GLcall(glBindVertexArray(contact_matrix_->vaos[ptr_cnt])); 
             GLcall(glDrawRangeElements(GL_TRIANGLES, 0, 3, 6, GL_UNSIGNED_SHORT, NULL));
 
             // GLcall( glBindFramebuffer(GL_READ_FRAMEBUFFER, framebuffer));
-            GLcall( glBindFramebuffer(GL_READ_FRAMEBUFFER, 0));
+            // GLcall( glBindFramebuffer(GL_READ_FRAMEBUFFER, 0));
             GLcall(glReadPixels(0, 0, texture_resolution, texture_resolution, GL_RED, GL_UNSIGNED_BYTE, textures[layer_cnt]));
             if ((layer_cnt + 1) % (total_num_textures / 100) == 0)
             {
@@ -498,6 +521,55 @@ void TexturesArray4AI::copy_buffer_to_textures_dynamic(
     return ;
 }
 
+/*
+存储一个matrix然后 rearrange
+
+目前没有这样做，因为需要存两个textures... 目前还是从opengl的buffer往外copy
+*/
+void TexturesArray4AI::rearrange_textures(const u32* rearrange_index)
+{
+    if (!is_copied_from_buffer || this->textures == nullptr) 
+    {
+        std::cerr << "TexturesArray4AI is not initialized, "
+            "please first copy the buffer textures to this->textures[], "
+            << std::endl;
+        assert(0);
+    }
+    u08** new_textures = new u08*[total_num_textures];
+    for (u32 i = 0; i < total_num_textures; i++)
+    {
+        new_textures[i] = new u08[texture_resolution * texture_resolution];
+    }
+    u32 texture_id = 0, row_in_texture = 0, column_in_texture = 0, 
+        row_re = 0, column_re =0,
+        texture_id_re = 0, row_in_texture_re = 0, column_in_texture_re = 0;
+
+    for (u32 row = 0; row < num_pixels_1d; row++)
+    {
+        for (u32 column = 0; column < num_pixels_1d; column++)
+        {   
+            /* 
+            id= texture_id_cal(row>>texture_resolution_exp, column>>texture_resolution_exp, num_textures_1d)
+            row_in_texture = row & ((1<<texture_resolution_exp) - 1)
+            colomn_in_texture = column & ((1<<texture_resolution_exp) - 1)
+            */
+            texture_id = texture_id_cal(row>>texture_resolution_exp, column>>texture_resolution_exp, num_textures_1d);
+            row_in_texture = row & ((1<<texture_resolution_exp) - 1);
+            column_in_texture = column & ((1<<texture_resolution_exp) - 1);
+
+            row_re = rearrange_index[row];
+            column_re = rearrange_index[column];
+            texture_id_re = texture_id_cal(row_re>>texture_resolution_exp, column_re>>texture_resolution_exp, num_textures_1d);
+            row_in_texture_re = row_re & ((1<<texture_resolution_exp) - 1);
+            column_in_texture_re = column_re & ((1<<texture_resolution_exp) - 1);
+            new_textures[texture_id][row_in_texture * texture_resolution + column_in_texture] = textures[texture_id_re][row_in_texture_re * texture_resolution + column_in_texture_re];
+        }
+    }
+    clear_textures();
+    is_copied_from_buffer = true;
+    textures = new_textures;
+    return ;
+}
 
 
 void TexturesArray4AI::prepare_tmp_texture2d_array(GLuint& tmp_texture2d_array)
@@ -632,7 +704,7 @@ void TexturesArray4AI::show_collected_texture()
 void TexturesArray4AI::show_collected_textures()
 {
     GLuint tmp_texture2d_array;
-    prepare_tmp_texture2d_array(tmp_texture2d_array);
+    this->prepare_tmp_texture2d_array(tmp_texture2d_array);
     
     GLFWwindow* window = glfwGetCurrentContext();
     if(!window) assert(0);
@@ -674,13 +746,13 @@ void TexturesArray4AI::show_collected_textures()
     glfwSetCursorPosCallback(window, cursor_position_callback);
     glUseProgram(shaderProgram);
     GLcall(glUniform1i(glGetUniformLocation(shaderProgram, "texArray"), 0));
-    glm::mat4 model_not_flipped = glm::scale(
+    glm::mat4 model_not_flipped = glm::scale( // 控制显示大小
         glm::mat4(1.0f), 
         glm::vec3(
             1.0f/(f32)num_textures_1d, 
             -1.0f/(f32)num_textures_1d, // flipped on y axis 
             1.0f));
-    glm::mat4 model_flipped = glm::scale(
+    glm::mat4 model_flipped = glm::scale( // 控制显示大小
         glm::rotate(model_not_flipped, glm::radians(90.0f), glm::vec3(0.0f, 0.0f, 1.0f)), 
         glm::vec3(1.0f, -1.0f, 1.0f));
     while (!glfwWindowShouldClose(window))
@@ -696,20 +768,20 @@ void TexturesArray4AI::show_collected_textures()
 
         glBindVertexArray(vao);
         glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, ebo);
-        glm::mat4 projection = glm::ortho(
+        glm::mat4 projection = glm::ortho( // 控制镜头远近
             -1.0f * ratio_w_h * show_state.zoomlevel, 1.0f * ratio_w_h * show_state.zoomlevel, 
             -1.0f * show_state.zoomlevel, 1.0f * show_state.zoomlevel, 
             -1.0f, 1.0f);
         
         glm::mat4 model_tmp;
-        for (int row = 0; row < num_textures_1d; row++ )
+        for (int row = 0; row < num_textures_1d; row++ )// 显示所有的 tile
         {
             for (int column = 0; column < num_textures_1d; column ++ )
             {   
                 // if (row > column) continue;
                 u32 layer = texture_id_cal((u32)row, (u32)column, num_textures_1d);
                 GLcall(glUniform1i(glGetUniformLocation(shaderProgram, "layer"), (int)layer));
-                model_tmp =  glm::translate(
+                model_tmp =  glm::translate( // 控制显示位置
                     glm::mat4(1.0f), 
                     glm::vec3( 
                         -1.0f + ((f32)column + 0.5f) * 2.0f / (f32)num_textures_1d + show_state.translationOffset.x,
@@ -826,10 +898,20 @@ void TexturesArray4AI::cal_compressed_hic(
         );
         assert(0);
     }
-    compressed_hic->re_allocate_mem(
-        frags->num, 
-        frags->num, 
-        5);
+    if (compressed_hic)
+    {
+        compressed_hic->re_allocate_mem(
+            frags->num, 
+            frags->num, 
+            5);
+    }
+    else 
+    {
+        compressed_hic = new Matrix3D<f32>(
+            frags->num, 
+            frags->num, 
+            5);
+    }
     if (mass_centres) 
     {
         delete mass_centres;
@@ -973,8 +1055,10 @@ void TexturesArray4AI::get_interaction_score(
         };
 
     u32 number_of_scores = std::min(D-1, std::max(num_row, num_column));
-    f32* linear_weight = new f32[number_of_scores]; 
-    get_linear_mask(linear_weight, number_of_scores);  // assign the linear weight mask
+    // f32 linear_weight[number_of_scores]; 
+    // get_linear_mask(linear_weight, number_of_scores);  // assign the linear weight mask
+    f32 linear_weight[std::min(D-1, (u32)(number_of_scores +5) ) ];  // use this to consider the score caused by repeat 
+    get_linear_mask(linear_weight, std::min(D-1, (u32)(number_of_scores +5) ) );  // 往外多算几个像素点，因此小的片段的link score分就会相对下降, 特别是很小的片段
 
     buffer_values_on_channel.initilize();
     for (int channel = 0; channel < 4; channel++)
@@ -988,7 +1072,7 @@ void TexturesArray4AI::get_interaction_score(
             buffer_values_on_channel.c[channel] += linear_weight[shift-1] * score_cal(tmp_mean, shift);
         }
     }
-    delete[] linear_weight;
+    // delete[] linear_weight;
     return ;
 }
 
diff --git a/src/copy_texture.h b/src/copy_texture.h
index 5fa79ac..a13f6e8 100644
--- a/src/copy_texture.h
+++ b/src/copy_texture.h
@@ -26,10 +26,11 @@ SOFTWARE.
 #include <utility>
 #include <fstream>
 #include <random>
-
+#include <cstdio>
 #include "utilsPretextView.h"
 #include "genomeData.h"
 #include "auto_curation_state.h"
+#include <fmt/core.h>
 
 struct Frag4compress {
     u32 num;
@@ -293,10 +294,8 @@ class Matrix3D
     void cleanup()
     {   
 
-        MY_CHECK(0);
         if (data)
         {   
-            MY_CHECK(0);
             delete[] data;
             data = nullptr;
         }
@@ -304,7 +303,6 @@ class Matrix3D
 
     void re_allocate_mem(u32 row_num_, u32 col_num_, u32 layer_)
     {   
-        MY_CHECK(0);
         cleanup();
         row_num = row_num_;
         col_num = col_num_;
@@ -314,12 +312,9 @@ class Matrix3D
         shape[2] = layer_;
         length = row_num * col_num * layer_num;
 
-        MY_CHECK(0);
         data = new T[length];
 
-        MY_CHECK(0);
         memset(data, 0, length * sizeof(T));
-        MY_CHECK(0);
     }
 
     void check_indexing(const u32& row, const u32& col, const u32& layer) const
@@ -348,6 +343,23 @@ class Matrix3D
         check_indexing(row, col, layer);
         data[row * col_num * layer_num + col* layer_num + layer] = value;
     }
+
+    void output_to_file(FILE* fp) const
+    {
+        fmt::print(fp, "# Matrix shape: {} {} {}\n", row_num, col_num, layer_num);
+        for (u32 l = 0; l < layer_num; l++)
+        {   
+            fmt::print(fp, "# layer: {}\n", l);
+            for (u32 i = 0; i < row_num; i++)
+            {
+                for (u32 j = 0; j < col_num; j++)
+                {
+                    fmt::print(fp, "{:.4f} ", data[i * col_num * layer_num + j * layer_num + l]);
+                }
+                fmt::print(fp, "\n");
+            }
+        }
+    }
 };
 
 
@@ -469,10 +481,11 @@ struct Values_on_Channel
     }
     void initilize()
     {
-        c[0] = 0.f;
-        c[1] = 0.f;
-        c[2] = 0.f;
-        c[3] = 0.f;
+        // c[0] = 0.f;
+        // c[1] = 0.f;
+        // c[2] = 0.f;
+        // c[3] = 0.f;
+        memset(c, 0, 4 * sizeof(f32));
     }
 }; 
 
@@ -755,6 +768,10 @@ class TexturesArray4AI
     void output_mass_centres_to_file() const;
     void output_frags4AI_to_file() const;
     void output_compressed_hic_massCetre_extension_for_python() const;
+    void initialise_textures();
+    void clear_textures();
+    void clear_compressed_hic();
+    void rearrange_textures(const u32* rearrange_index);
 };
 
 

From 509749cb5bfc6f548d684b12c6b0280b915ec2a3 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:08:47 +0100
Subject: [PATCH 17/33] change parameters for pixel cut

---
 src/auto_curation_state.h  |  27 +++++-
 src/frag_cut_calculation.h | 181 ++++++++++++++++++++-----------------
 2 files changed, 122 insertions(+), 86 deletions(-)

diff --git a/src/auto_curation_state.h b/src/auto_curation_state.h
index f5162f7..e1b8454 100644
--- a/src/auto_curation_state.h
+++ b/src/auto_curation_state.h
@@ -1,6 +1,3 @@
-
-
-
 #ifndef AUTO_CURATION_STATE_H 
 #define AUTO_CURATION_STATE_H
 
@@ -115,8 +112,8 @@ struct AutoCurationState
     bool show_autoSort_erase_confirm_popup = false;
     bool show_autoSort_redo_confirm_popup = false;
     //sorting mode
-    u32 sort_mode = 0; // 0: union find, 1: fuse union find, 2 deep fuse
-    std::vector<std::string> sort_mode_names = {"Union Find", "Fuse Union Find", "Deep Fuse"};
+    u32 sort_mode = 1; // 0: union find, 1: fuse union find, 2 deep fuse, 3 yahs
+    std::vector<std::string> sort_mode_names = {"Union Find", "Fuse", "Deep Fuse"};
     
     // auto sort
     char frag_size_buf[16];
@@ -165,6 +162,14 @@ struct AutoCurationState
         }
     }
 
+    void set_default_mask_color()
+    {
+        for (u32 i = 0; i < 4; i++)
+        {
+            this->mask_color[i] = this->mask_color_default[i];
+        }
+    }
+
     void update_value_from_buf()
     {   
         // sort 
@@ -416,6 +421,18 @@ struct AutoCurationState
         return;
     }
 
+    void adjust_cut_threshold(s32 delta)
+    {
+        this->auto_cut_threshold += delta * 0.002f;
+        this->auto_cut_threshold = std::max(0.0f, std::min(1.0f, this->auto_cut_threshold));
+        fmt::format_to((char*)auto_cut_threshold_buf, "{:.4f}", this->auto_cut_threshold);
+    }
+
+    void change_sort_mode(s32 delta)
+    {
+        this->sort_mode = (this->sort_mode + this->sort_mode_names.size() + (delta>0?1:-1)) % this->sort_mode_names.size();
+    }
+
 };
 
 #endif // AUTO_CURATION_STATE_H 
\ No newline at end of file
diff --git a/src/frag_cut_calculation.h b/src/frag_cut_calculation.h
index dcd0f09..80023b6 100644
--- a/src/frag_cut_calculation.h
+++ b/src/frag_cut_calculation.h
@@ -13,6 +13,7 @@
 #include "genomeData.h"
 #include "utilsPretextView.h"
 #include "copy_texture.h"
+#include "auto_curation_state.h"
 
 
 
@@ -35,31 +36,22 @@ class FragCutCal
 public: 
     TexturesArray4AI *textures_array = nullptr;
     Frag4compress* frags = nullptr;
-    contigs* Contigs = nullptr;
-    std::string file_save_dir = std::filesystem::current_path().string() + "/hic_density_curve";
-    std::vector<f32> hic_density={};
+    std::string file_save_dir = std::filesystem::current_path().string() + "/hic_pixel_density_curve";
+    std::vector<f32> hic_pixel_density_origin={};
+    std::vector<f32> hic_pixel_density={};
     const SelectArea* select_area = nullptr;
 
     const u32 pixel_mean_window_size;
-    const f32 cut_threshold;
-    const u32 smallest_frag_size;
-
-    u32 D = 0;
-    std::vector<f32> norm_diag_mean={};
-
-    f32 D_hic_ratio=0.05f;
-    u32 maximum_D=5000;
-    f32 min_hic_density=30.f;
+    f32 cut_threshold=0.05f;
+    u32 smallest_frag_size_in_pixel = 8;
 
     FragCutCal(
         TexturesArray4AI* textures_array_, 
-        contigs* Contigs_, 
         const u32& num_pixels_1d, 
-        const SelectArea* select_area_, 
-        u32 pixel_mean_window_size_, 
-        f32 cut_threshold_, 
-        u32 smallest_frag_size_)
-        : textures_array(textures_array_), Contigs(Contigs_), select_area(select_area_), pixel_mean_window_size(pixel_mean_window_size_), cut_threshold(cut_threshold_), smallest_frag_size(smallest_frag_size_)
+        const u32 pixel_mean_window_size_)
+
+        : textures_array(textures_array_), 
+        pixel_mean_window_size(pixel_mean_window_size_)
     {   
         if (textures_array == nullptr)
         {
@@ -67,41 +59,80 @@ class FragCutCal
             assert(0);
         }
 
-        #ifdef DEBUG // create folder to save the hic_density_curve
+        #ifdef DEBUG // create folder to save the hic_pixel_density_curve
             if (std::filesystem::exists(this->file_save_dir))
             {
                 std::filesystem::remove_all(this->file_save_dir);
             }
             std::filesystem::create_directory(this->file_save_dir);
         #endif
-
-        this->frags = textures_array->get_frags();
-        this->frags->re_allocate_mem(this->Contigs, this->select_area, true);
-        textures_array->cal_maximum_number_of_shift( // and get the norm_diag_mean
-            this->D, 
-            this->norm_diag_mean, 
-            this->Contigs,
-            this->D_hic_ratio,
-            this->maximum_D,
-            this->min_hic_density);
         
-        // cal the hic_density
-        this->hic_density.resize(num_pixels_1d, 0.f);
+        this->frags = textures_array->get_frags(); // 需要得到最初没有进行排序的contig 信息
+
+        // cal the hic_pixel_density
+        this->cal_pixel_density(num_pixels_1d);
+    }
+
+    void cal_pixel_density(u32 num_pixels_1d)
+    {   
+        this->hic_pixel_density_origin.resize(num_pixels_1d, 0.f);
         for (u32 frag_id = 0; frag_id < this->frags->num; frag_id++)
         {
-            this->get_single_fragment_hic_density(
+            this->get_single_fragment_hic_pixel_density(
                 this->frags->startCoord[frag_id],
                 this->frags->length[frag_id],
-                this->hic_density.data() + this->frags->startCoord[frag_id]);
+                this->hic_pixel_density_origin.data() + this->frags->startCoord[frag_id]);
         }
+
+        #ifdef DEBUG // output the hic_pixel_density to text file
+            std::string filename = fmt::format("{}/hic_pixel_density_origin.txt", this->file_save_dir);
+            std::ofstream out(filename);
+            if (!out)
+            {
+                fmt::print(stderr, "Cannot open the file {}\n", filename);
+                assert(0);
+            }
+            for (u32 i = 0; i < this->hic_pixel_density_origin.size(); i++)
+            {
+                out << this->hic_pixel_density_origin[i] << std::endl;
+            }
+            out.close();
+            fmt::print("[Pixel Cut]: output hic_pixel_density to {}\n", filename);
+        #endif
     }
 
-    std::vector<u32> get_cut_locs_pixel()
-    {
+    void rearrange_hic_pixel_density(const u32* pixel_rearrange_index)
+    {   
+        this->hic_pixel_density.resize(this->hic_pixel_density_origin.size(), 0.f);
+        for (u32 i = 0; i < this->hic_pixel_density.size(); i++)
+        {   
+            if (pixel_rearrange_index[i]<0 || pixel_rearrange_index[i] >= this->hic_pixel_density.size())
+            {
+                fmt::print("The pixel_rearrange_index ({}) is out of range[{}, {}]", pixel_rearrange_index[i], 0, this->hic_pixel_density.size()-1);
+            }
+            this->hic_pixel_density[i] = this->hic_pixel_density_origin[pixel_rearrange_index[i]];
+        }
+    }
+
+    std::vector<u32> get_cut_locs_pixel(
+        const AutoCurationState& auto_curation_state,
+        const u32* pixel_rearrange_index_, 
+        const contigs* Contigs,
+        const SelectArea* select_area=nullptr)
+    {   
+        // 设定cut的参数
+        this->smallest_frag_size_in_pixel = auto_curation_state.auto_cut_smallest_frag_size_in_pixel;
+        this->cut_threshold = auto_curation_state.auto_cut_threshold;
+
+        // rearrange index 
+        this->rearrange_hic_pixel_density(pixel_rearrange_index_);
+
+        // 在 select_area 中进行cut 
+        this->frags->re_allocate_mem(Contigs, select_area, 1);
         std::vector<u32> cut_locs_pixel;
         for (u32 i = 0; i < this->frags->num; i++)
         {
-            this->get_single_fragment_cut_locs(i, cut_locs_pixel);
+            this->get_single_fragment_cut_locs(i, cut_locs_pixel); 
         }
         return cut_locs_pixel;
     }
@@ -112,7 +143,7 @@ class FragCutCal
         const std::vector<f32>& arr,
         std::vector<u32>& break_points)
     {   
-        const auto& windows_size = this->pixel_mean_window_size;
+        const auto& windows_size = this->smallest_frag_size_in_pixel;
         const auto& threshold = this->cut_threshold;
 
         int n = arr.size();
@@ -164,34 +195,40 @@ class FragCutCal
     }
 
     void get_single_fragment_cut_locs(
-        const u32& frag_id, std::vector<u32>& cut_locs)
+        const u32& frag_id, 
+        std::vector<u32>& cut_locs)
     {   
         // copy orginal to local
-        std::vector<f32> hic_density_tmp(this->hic_density.begin() + this->frags->startCoord[frag_id], 
-            this->hic_density.begin() + this->frags->startCoord[frag_id] + this->frags->length[frag_id]);
+        std::vector<f32> hic_pixel_density_tmp(this->hic_pixel_density.begin() + this->frags->startCoord[frag_id], 
+            this->hic_pixel_density.begin() + this->frags->startCoord[frag_id] + this->frags->length[frag_id]);
 
-        u32 n = hic_density_tmp.size();
+        u32 n = hic_pixel_density_tmp.size();
 
         // std::vector<f32> order1_derivative(n, 0.f);
         // std::vector<f32> order2_derivative(n, 0.f);
-        // get_smoothed_1st_order_derivative(hic_density_tmp, order1_derivative);
+        // get_smoothed_1st_order_derivative(hic_pixel_density_tmp, order1_derivative);
         // get_smoothed_1st_order_derivative(order1_derivative, order2_derivative);
         
-        #ifdef DEBUG // output the hic_density to text file
-            std::string filename = this->file_save_dir +  "/current_id_" + std::to_string(frag_id) + ".txt";
-            std::cout << "[Auto Cut]: (classic) output hic_density to " << filename << std::endl;
+        #ifdef DEBUG_OUTPUT_PIXEL_CUT_FILE // output the hic_pixel_density to text file
+            std::string filename = fmt::format("{}/current_id_{}.txt", this->file_save_dir, frag_id);
+            fmt::print("[Pixel Cut]: output hic_pixel_density to {}\n", filename);
             std::ofstream out(filename);
-            for (u32 i = 0; i < hic_density_tmp.size(); i++)
+            if (!out)
+            {
+                fmt::print(stderr, "Cannot open the file {}\n", filename);
+                assert(0);
+            }
+            for (u32 i = 0; i < hic_pixel_density_tmp.size(); i++)
             {
-                out << hic_density_tmp[i] << std::endl;
+                out << hic_pixel_density_tmp[i] << std::endl;
             }
             out.close();
-        #endif
+        #endif // DEBUG_OUTPUT_PIXEL_CUT_FILE
 
         std::vector<u32> break_points(0);
         this->find_break_points(
             this->frags->startCoord[frag_id],
-            hic_density_tmp,
+            hic_pixel_density_tmp,
             break_points);
         if (break_points.size() > 0)
         {
@@ -201,68 +238,50 @@ class FragCutCal
     }
 
     
-    void get_single_fragment_hic_density(
+    void get_single_fragment_hic_pixel_density(
         const u32& start_pixel,
         const u32& len_pixel,
-        f32* hic_density
+        f32* hic_pixel_density
     )
     {   
         for (u32 i=0; i < len_pixel; i++)
         {
-            hic_density[i] = this->pixel_mean(start_pixel, i, len_pixel);
+            hic_pixel_density[i] = this->pixel_mean(start_pixel, i, len_pixel);
         }
         if (len_pixel > 4)
         {
-            hic_density[0] = hic_density[1] = hic_density[2];
-            hic_density[len_pixel-1] = hic_density[len_pixel-2] = hic_density[len_pixel-3];
+            hic_pixel_density[0] = hic_pixel_density[1] = hic_pixel_density[2];
+            hic_pixel_density[len_pixel-1] = hic_pixel_density[len_pixel-2] = hic_pixel_density[len_pixel-3];
         }
         else 
         {
             for (u32 i = 0; i < len_pixel; i++)
             {
-                hic_density[i] = 1.0f;
+                hic_pixel_density[i] = 1.0f;
             }
         }
         return ;
     }
 
+    // 计算对角线上的距离为 i < this->pixel_mean_window_size 的像素的和
     f32 pixel_mean(
         const u32& frag_start_pixel,
         const u32& relative_index, 
-        const u32& len_pixel, 
-        const u08 diag_flag=1) 
+        const u32& len_pixel) 
     {   
         u32 num = 0, pixel_index=frag_start_pixel+relative_index;
         f32 sum = 0.f;
 
         for (u32 i = 1; i < this->pixel_mean_window_size ; i++) 
         {   
-            if (diag_flag)
-            {   
-                if (i < relative_index && i + relative_index < len_pixel-1)
-                {
-                    sum += ( (f32)((*(this->textures_array))(
-                        pixel_index - i,
-                        pixel_index + i)) ); // / this->norm_diag_mean[i]
-                    num++;
-                }
-            } else 
+            if (i < relative_index && i + relative_index < len_pixel-1)
             {
-                if ( i+relative_index < len_pixel)
-                {
-                    sum += ((*(this->textures_array))(
-                        pixel_index, 
-                        pixel_index + i) / this->norm_diag_mean[i]);
-                    num++;
-                }
-                if (relative_index >= i)
-                {
-                    sum += ((*(this->textures_array))(
-                        pixel_index, 
-                        pixel_index - i) / this->norm_diag_mean[i]);
-                    num++;
-                }
+                sum += ( (f32)((*(this->textures_array))(
+                    pixel_index - i,
+                    pixel_index + i)) ); // / this->norm_diag_mean[i]
+                num++;
             }
+            else break;
         }
         return num==0 ? 0.f : (f32)sum / (f32)num;
     }

From dc9fd4f9d55cb5a77da46fc105dcc5170e393768 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:10:03 +0100
Subject: [PATCH 18/33] fix the end while sorting locally

---
 src/frag_sort.cpp | 297 +++++++++++++++++++---------------------------
 src/frag_sort.h   | 211 ++------------------------------
 src/frags_order.h | 121 +++++++++++++++++++
 3 files changed, 256 insertions(+), 373 deletions(-)
 create mode 100644 src/frags_order.h

diff --git a/src/frag_sort.cpp b/src/frag_sort.cpp
index 6a06937..b158388 100644
--- a/src/frag_sort.cpp
+++ b/src/frag_sort.cpp
@@ -1,4 +1,3 @@
-
 /*
 Copyright (c) 2024 Shaoheng Guan, Wellcome Sanger Institute
 
@@ -22,8 +21,7 @@ SOFTWARE.
 */
 
 
-#include <frag_sort.h>
-
+#include "frag_sort.h"
 
 
 void dfs(
@@ -219,25 +217,38 @@ void initilise_chromosomes(
 }
 
 /*
-source and sink frags should not be sorted
+条件：
+1. source frag 的头不能连
+2. sink frag   的尾不能连
+3. source 和 sink不能连到一起
+
+
+
     return: 
             0: not attachable, 
             1: attachable
 */
 bool check_source_sink_attachable(
     const Link& lnk, 
-    u32 source_frag_id, u32 sink_frag_id, 
-    const u32& source_chain_id, const u32& sink_chain_id,
-    const u32& ca, const u32& cb
+    u32 source_frag_id, 
+    u32 sink_frag_id, 
+    const u32& source_chain_id, 
+    const u32& sink_chain_id,
+    const u32& ca, 
+    const u32& cb
 )
 {   
+    // source and sink can not be linked to the same chain
     if ((source_chain_id == ca && sink_chain_id == cb) || 
         (sink_chain_id == ca && source_chain_id == cb)) return 0;
+
     source_frag_id--; sink_frag_id -- ; // 在link中，从0开始编号，但是source_frag_id从1开始编号
 
+    // source and sink can not be linked to each other
     if ((lnk.frag_a == source_frag_id && lnk.frag_b==sink_frag_id)||
         (lnk.frag_b == source_frag_id && lnk.frag_a==sink_frag_id))
         return 0;
+
     // source can not be linked to the head
     if ( source_frag_id >= 0 && 
         ((lnk.frag_a == source_frag_id && lnk.is_a_head_linked) || 
@@ -257,15 +268,15 @@ bool check_source_sink_attachable(
 reverse the chain and return the new chain_id
 */
 void reverse_chain(
-    std::deque<s32>& cha
-)
+    std::deque<s32>& cha)
 {
     std::reverse(cha.begin(), cha.end());
     for (auto &val: cha) val = -val; // flip orientation
 }
+
 void reverse_chain(
     std::deque<s32>& cha,
-    u32& i
+    u32& i // frag_id in the chain
 )
 {
     std::reverse(cha.begin(), cha.end());
@@ -289,7 +300,6 @@ void union_find_process(
     s32& sink_chain_id
 )
 {
-    
     // Process links in order of decreasing score
     for (const Link &lnk : all_links)
     {
@@ -362,50 +372,52 @@ void fix_source_sink_frag(
     const s32 sink_frag_id
 )
 {   
-    /*
-    TODO (SHAOHENG) 这里还是有问题，当只有一个片段的时候有时候会把第一个片段反向，检查这个地方
-    */
 
-    if (sink_frag_id < 0 && source_frag_id < 0) return ;
+    if (select_area.selected_frag_ids.empty() || (sink_frag_id < 0 && source_frag_id < 0 ) ) return ;
 
-    if (select_area.selected_frag_ids.size() > 0)
+    for (u32 i = 0; i < chromosomes.size() ; i ++ )
     {   
-        for (u32 i = 0; i < chromosomes.size() ; i ++ )
-        {   
-            std::deque<s32>& tmp = chromosomes[i];
-            if (tmp.empty()) continue;
-            
-            if (std::abs(tmp.back()) == source_frag_id && tmp.size() > 1 ) reverse_chain(tmp);
-            if (std::abs(tmp.front()) == source_frag_id && i != 0)
-            {   
-                if (tmp.front() < 0)
-                {
-                    fmt::print(
-                        stderr,
-                        "Error: the source_frag_id({}) should not be inversed, chromosome[{}].font() = {} \n", 
-                        source_frag_id, 
-                        i, 
-                        tmp.front());
-                }
-                std::swap(tmp, chromosomes.front());
-                continue;
-            }
+        std::deque<s32>& tmp = chromosomes[i];
+        if (tmp.empty()) continue;
+        
+        if (std::abs(tmp.back()) == source_frag_id && tmp.size() > 1 ) 
+        {
+            if (tmp.back() > 0) 
+            {
+                fmt::print(
+                    stderr,
+                    "Error: source_frag({})'s head should not be linked, chromosome[{}].back() = {} \n", 
+                    source_frag_id, 
+                    i, 
+                    tmp.back());
+                assert(0);
+            } 
+            reverse_chain(tmp);
+        }
+        if (std::abs(tmp.front()) == source_frag_id && i != 0)
+        {
+            std::swap(tmp, chromosomes.front());
+            continue;
+        }
 
-            if (std::abs(tmp.front()) == sink_frag_id && tmp.size() > 1) reverse_chain(tmp);
-            if (std::abs(tmp.back()) == sink_frag_id)
-            {   
-                if (tmp.back() < 0)
-                {
-                    fmt::print(
-                        stderr,
-                        "Error: the sink_frag_id({}) should not be inversed, chromosome[{}].back() = {} \n", 
-                        sink_frag_id, 
-                        i, 
-                        tmp.back());
-                }
-                std::swap(tmp, chromosomes.back());
-                continue;
+        if (std::abs(tmp.front()) == sink_frag_id && tmp.size() > 1) 
+        {
+            if (tmp.front() > 0 )
+            {
+                fmt::print(
+                    stderr,
+                    "Error: the sink_frag_id({})'s end should not be linked, chromosome[{}].front() = {} \n", 
+                    sink_frag_id, 
+                    i, 
+                    tmp.front());
+                assert(0);
             }
+            reverse_chain(tmp);
+        }
+        if (std::abs(tmp.back()) == sink_frag_id && i != chromosomes.size()-1)
+        {   
+            std::swap(tmp, chromosomes.back());
+            continue;
         }
     }
 }
@@ -416,6 +428,8 @@ FragSortTool::FragSortTool() {}
 
 FragSortTool::~FragSortTool() {}
 
+
+
 void FragSortTool::sort_according_likelihood_unionFind(
     const LikelihoodTable& likelihood_table, 
     FragsOrder& frags_order, 
@@ -441,6 +455,12 @@ void FragSortTool::sort_according_likelihood_unionFind(
     u32 n = likelihood_table.get_num_frags();
     std::vector<Link> all_links;
 
+    // 计算的分位数
+    f32 threshold_val = percentile_cal(
+        likelihood_table.data, likelihood_table.size, this->threshold_ratio);
+    f32 threshold_using = std::min(threshold_val, threshold);
+    fmt::print(stderr, "[Pixel Sort] threshold at {}: {}, fixed_threshold: {}, using: {}\n", this->threshold_ratio, threshold_val, threshold, threshold_using);
+
     std::vector<std::pair<bool, bool>> link_kinds = {
         {true,  true}, {true,  false}, 
         {false, true}, {false, false}};
@@ -451,7 +471,7 @@ void FragSortTool::sort_according_likelihood_unionFind(
             for (u32 k = 0; k < 4; k++)
             {   
                 f32 val = likelihood_table(i, j, k);
-                if (val >= threshold) {
+                if (val >= threshold_using) {
                     all_links.push_back(
                         Link(val, i, j, link_kinds[k].first, link_kinds[k].second));
                 }
@@ -517,7 +537,6 @@ void FragSortTool::sort_according_likelihood_unionFind(
 }
 
 
-// Chenxi Zhou
 f32 link_score(
     const std::vector<Link>& all_links,
     const std::map<u64,u32>& map_links,
@@ -535,113 +554,7 @@ f32 link_score(
 }
 
 
-// Chenxi Zhou
 std::deque<s32> fuse_chromosomes(
-    const std::vector<Link>& all_links,
-    const std::map<u64,u32>& map_links,
-    std::deque<s32>& cha,
-    std::deque<s32>& chb,
-    s32 source, s32 sink,
-    const f32 threshold)
-    /*
-    cha:    tail of a
-    chb:    head of b
-    source: cha[i]
-    sink:   chb.size() > (j+1) ? chb[j+1] : 0
-    */
-{
-    int n = cha.size();
-    int m = chb.size();
-    
-    // push source
-    cha.push_front(source);
-    chb.push_front(source);
-
-    // need two DP matrices
-    // one for DP[i,j] ending with i-th element in cha
-    // one for DP[i,j] ending with j-th element in chb
-    // fill matrices row by row
-    std::vector<double> pre_row_a(m+1, -std::numeric_limits<double>::infinity());   // i - 1
-    std::vector<double> curr_row_a(m+1, -std::numeric_limits<double>::infinity());  // i
-    std::vector<double> pre_row_b(m+1, -std::numeric_limits<double>::infinity());   // i - 1
-    std::vector<double> curr_row_b(m+1, -std::numeric_limits<double>::infinity());  // i
-
-    // backtracking matrix could be big - n*m*2
-    // use one bit per position
-    // tracking if need to switch list
-    std::vector<std::vector<uint8_t>> bt(2, std::vector<uint8_t>(((int64_t)n*m-1)/8+1, 0));
-
-    pre_row_a[0] = 0;
-    pre_row_b[0] = 0;
-    for (int j = 1; j <= m; ++j)  // initialise the score of the original chb
-        pre_row_b[j] = pre_row_b[j-1] + link_score(all_links, map_links, chb[j-1], chb[j]);
-    
-    // fill DP and backtracking matrices
-    double link1, link2;
-    int64_t b;
-    for (int i = 1; i <= n; ++i)  // check all the elements in cha
-    {
-        curr_row_a[0] = pre_row_a[0] + link_score(all_links, map_links, cha[i-1], cha[i]); // i = i, j = 0
-        curr_row_b[0] = -std::numeric_limits<double>::infinity();                          // i = i, j = 0  // original code
-        // curr_row_b[0] = pre_row_a[0] ; // i = i, j = 0  // changed code
-        for (int j = 1; j <= m; ++j) {
-            b = (i-1)*m + (j-1);
-            link1 = pre_row_a[j] + link_score(all_links, map_links, cha[i-1], cha[i]);    // last is cha[i-1]
-            link2 = pre_row_b[j] + link_score(all_links, map_links, chb[j], cha[i]);      // last is chb[j]
-            curr_row_a[j] = std::max(link1, link2);
-            bt[0][b/8] |= (uint8_t)(link1<link2?1:0)<<(b%8);
-            link1 = curr_row_a[j-1] + link_score(all_links, map_links, cha[i], chb[j]);   // last is cha[i]
-            link2 = curr_row_b[j-1] + link_score(all_links, map_links, chb[j-1], chb[j]); // last is chb[j-1]
-            curr_row_b[j] = std::max(link1, link2);
-            bt[1][b/8] |= (uint8_t)(link1<link2?0:1)<<(b%8);
-        }
-        std::swap(curr_row_a, pre_row_a);
-        std::swap(curr_row_b, pre_row_b);
-    }
-
-    double link, maxLink;
-    link = 0;
-    for (int i = 1; i < cha.size(); ++i)
-        link += link_score(all_links, map_links, cha[i-1], cha[i]);
-    for (int j = 2; j < chb.size(); ++j)
-        link += link_score(all_links, map_links, chb[j-1], chb[j]);
-    link += link_score(all_links, map_links, chb.back(), sink);
-        
-    link1 = pre_row_a[m] + link_score(all_links, map_links, cha[n], sink);
-    link2 = pre_row_b[m] + link_score(all_links, map_links, chb[m], sink);
-    maxLink = std::max(link1, link2);
-
-    // link is not strong enough
-    if (maxLink < link + threshold)
-        return std::deque<s32>();
-
-    // pop source
-    cha.pop_front();
-    chb.pop_front();
-
-    // do backtracking
-    std::deque<s32> fused(0);
-    std::array<std::reference_wrapper<std::deque<s32>>, 2> 
-        lists = {std::ref(cha), std::ref(chb)};
-    std::array<int, 2> sizes = {n-1, m-1};
-    uint8_t k = link1<link2? 1 : 0;
-    while(sizes[0] >= 0 && sizes[1] >= 0) // push cha and chb to fused according to backtracking matrix
-    {
-        b = sizes[0]*m + sizes[1];
-        fused.push_front(lists[k].get()[sizes[k]--]);
-        k ^= !!(bt[k][b/8]&((uint8_t)1<<(b%8)));
-    }
-    while(sizes[0] >= 0)
-        fused.push_front(lists[0].get()[sizes[0]--]);
-    while(sizes[1] >= 0)
-        fused.push_front(lists[1].get()[sizes[1]--]);
-    
-    return fused;
-}
-
-
-// from Chenxi's code
-std::deque<s32> fuse_chromosomes_shguan(
     const std::vector<Link>& all_links,
     const std::map<u64, u32>& map_links,
     std::deque<s32>& cha,
@@ -730,7 +643,6 @@ std::deque<s32> fuse_chromosomes_shguan(
 }
 
 
-// Chenxi Zhou
 void FragSortTool::sort_according_likelihood_unionFind_doFuse(
     const LikelihoodTable& likelihood_table, 
     FragsOrder& frags_order, 
@@ -743,8 +655,10 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
 {   
 
     // this is used to fix the source and sink during local sort
-    s32 source_frag_id = -1, source_chain_id = -1, 
-        sink_frag_id = -1,   sink_chain_id = -1;
+    s32 source_frag_id = -1,  // 第一个片段的id
+        sink_frag_id = -1,    // 最后一个片段的id
+        source_chain_id = -1, // 第一个片段所在的chain的id
+        sink_chain_id = -1;   // 最后一个片段所在的chain的id 
     if (select_area.source_frag_id >=0)
     {
         source_frag_id = 1;
@@ -756,7 +670,12 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
         sink_chain_id = sink_frag_id - 1;
     }
 
-    // *** 
+    // 计算的分位数
+    f32 threshold_val = percentile_cal(likelihood_table.data, likelihood_table.size, this->threshold_ratio);
+    f32 threshold_using = std::min(threshold_val, threshold);
+    fmt::print(stderr, "[Pixel Sort]: link score threshold at {}: {}, fixed_threshold: {}, using: {}\n", this->threshold_ratio, threshold_val, threshold, threshold_using);
+
+    // 建立所有的links 
     u32 n = likelihood_table.get_num_frags();
     std::vector<Link> all_links;
     std::map<u64, u32> map_links;
@@ -771,7 +690,7 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
             for (u32 k = 0; k < 4; k++)
             {   
                 f32 val = likelihood_table(i, j, k);
-                if (val >= threshold) {
+                if (val >= threshold_using) {
                     all_links.push_back(
                         Link(val, i, j, link_kinds[k].first, link_kinds[k].second));
                 }
@@ -808,10 +727,10 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
             chromosomes,
             select_area,
             chain_id, 
-            source_frag_id, 
-            sink_frag_id, 
-            source_chain_id,
-            sink_chain_id
+            source_frag_id,   
+            sink_frag_id,     
+            source_chain_id,  
+            sink_chain_id     
         );
     }
 
@@ -851,9 +770,10 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
                 reverse_chain(chb, j);
             }
             
-            // should not fuse the source or sink inside of the chain   
+            // should not fuse the source or sink inside of the chain 
             if (std::abs(cha.back()) ==  source_frag_id || std::abs(cha.back()) ==  sink_frag_id || 
-                std::abs(chb.front()) == source_frag_id || std::abs(chb.front()) == sink_frag_id ) continue;
+                std::abs(chb.front()) == source_frag_id || std::abs(chb.front()) == sink_frag_id 
+            ) continue;
 
             // Now try to fuse cha and chb at position i and j
             // the fragments (i...end] on cha
@@ -865,14 +785,15 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
                 std::copy(cha.begin()+i+1, cha.end(), std::back_inserter(atail));
             std::deque<s32> bhead(chb.begin(), chb.begin()+j+1);
 
-            std::deque<s32> fused = fuse_chromosomes_shguan(
+            // todo （3, Apr. 2025） 检查为什么在deep fuse的时候还是会弄乱
+            std::deque<s32> fused = fuse_chromosomes(
                 all_links, 
                 map_links, 
                 atail,                       // cha 
                 bhead,                       // chb
-                cha[i],                      // source
-                chb.size()>(j+1)?chb[j+1]:0, // sink 
-                threshold);
+                cha[i],                      // source frag id
+                chb.size()>(j+1)?chb[j+1]:0, // sink frag id
+                threshold_using);
             if (!fused.empty()) 
             {
                 if (cha.size() > i+1) 
@@ -892,7 +813,7 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
     std::chrono::duration<f64> elapsed_seconds = end_time-start_time;
     fmt::print(
         stdout, 
-        "[Fusing]: consumed time {:.2f}s\n", 
+        "[Pixel Sort]: fusing consumed time {:.2f}s\n", 
         elapsed_seconds.count()
     );
 
@@ -928,4 +849,32 @@ void FragSortTool::sort_according_likelihood_unionFind_doFuse(
     chromosomes.insert(chromosomes.end(), chromosomes_excluded.begin(), chromosomes_excluded.end());
 
     frags_order.set_order(chromosomes);
+}
+
+void FragSortTool::sort_according_yahs(
+    const LikelihoodTable& likelihood_table,
+    FragsOrder& frags_order,
+    SelectArea& select_area,
+    const f32 threshold,
+    const Frag4compress* frags) const
+{
+    // Create YaHS instance
+    YaHS yahs;
+    
+    // Run YaHS sorting with the given parameters
+    auto chromosomes = yahs.yahs_sorting(likelihood_table, threshold, 0.1);
+    
+    // Convert chromosomes to frags_order format
+    std::vector<std::deque<s32>> formatted_chromosomes;
+    for (const auto& chromosome : chromosomes) {
+        std::deque<s32> formatted_chromosome;
+        for (uint32_t frag_id : chromosome) {
+            formatted_chromosome.push_back(frag_id + 1); // Convert to 1-based indexing
+        }
+        formatted_chromosomes.push_back(std::move(formatted_chromosome));
+    }
+    
+    // Set the order in frags_order
+    frags_order.set_order(formatted_chromosomes);
+    
 }
\ No newline at end of file
diff --git a/src/frag_sort.h b/src/frag_sort.h
index 4be194a..9400d79 100644
--- a/src/frag_sort.h
+++ b/src/frag_sort.h
@@ -33,7 +33,9 @@ SOFTWARE.
 #include <functional>
 #include "copy_texture.h"
 #include "auto_curation_state.h"
-
+#include "yahs_sort.h"
+#include "likelihood_table.h"
+#include "frags_order.h"
 
 struct Link {
     f32 score;
@@ -46,208 +48,12 @@ struct Link {
 };
 
 
-struct LikelihoodTable
+class FragSortTool
 {
-private:
-    u32 num_frags;   // all the frags including the filtered out ones
-    f32* data = nullptr;       // with shape [num_frags, num_frags, 4]
 public:
-    std::unordered_set<u32> excluded_fragment_idx;
-    Frag4compress* frags = nullptr;
-
-    LikelihoodTable(
-        Frag4compress* frags, 
-        const Matrix3D<f32>* compressed_hic, 
-        const f32 threshold=10.f/32769.f, 
-        const std::vector<s32>& exclude_tag_idx=std::vector<s32>(),
-        const u32 num_pixels_1d=32768) 
-        :num_frags(frags->num), frags(frags), data(nullptr)
-    {   
-        auto is_exclude_tag = [&](u32 idx) -> bool
-        {   
-            if ((f32)frags->length[idx]/(f32)num_pixels_1d <= threshold) 
-            {
-                this->excluded_fragment_idx.insert(idx);
-                return true;
-            }
-            if (exclude_tag_idx.size()==0) return false;
-            for (auto i : exclude_tag_idx)
-            {   
-                if (i < 0) continue;
-                if (frags->metaDataFlags[idx] & (1<<i)) 
-                {
-                    this->excluded_fragment_idx.insert(idx);
-                    return true;
-                }
-            }
-            return false;
-        };
-
-        data = new f32[num_frags * num_frags * 4];
-        for (u32 i = 0; i<num_frags * num_frags * 4; i++) data[i] = -1.f;
-        for (u32 i = 0; i<num_frags; i++)
-        {   
-            if ( is_exclude_tag(i) ) continue;
-            for (u32 j = 0; j < num_frags; j++)
-            {   
-                if ( is_exclude_tag(j)) continue;
-                (*this)(i, j, 0) = (*compressed_hic)(i, j, 0);
-                (*this)(i, j, 1) = (*compressed_hic)(i, j, 1);
-                (*this)(i, j, 2) = (*compressed_hic)(i, j, 3);
-                (*this)(i, j, 3) = (*compressed_hic)(i, j, 2);
-            }
-        }
-    }
-
-    ~LikelihoodTable()
-    {
-        if (data) 
-        {
-            delete[] data;
-            data = nullptr;
-        }
-    }
-
-    u32 get_num_frags() const {return num_frags;}
-
-    f32& operator() (const u32 i, const u32 j, const u32 k) 
-    {   
-        if (i >= num_frags || j >= num_frags || k >= 4) 
-        {
-            std::cerr << "Error: index out of range" << std::endl;
-            assert(0);
-        }
-        return data[ ((i * num_frags) << 2) + (j << 2) + k];
-    }
-
-    const f32& operator() (const u32 i, const u32 j, const u32 k) const
-    {
-        if (i >= num_frags || j >= num_frags || k >= 4) 
-        {
-            std::cerr << "Error: index out of range" << std::endl;
-            assert(0);
-        }
-        return data[ ((i * num_frags) << 2) + (j << 2) + k];
-    }
 
-};
-
-
-struct FragsOrder
-{
-private:
-    u32 num_frags;                  // number of all the fragments
-    u32 num_chromosomes;
-    u32* num_frags_in_chromosomes = nullptr;  // [num_chromosomes]
-    s32** order = nullptr;                    // [num_chromosomes, num_frags_in_chromosomes], + - represents the direction of the fragment, start from 1
-    void cleanup()
-    {   
-        if (order)
-        {   
-            for (u32 i = 0; i < num_chromosomes; i++)
-            {
-                if (order[i])
-                {
-                    delete[] order[i];
-                    order[i] = nullptr;
-                }
-            }
-            delete[] order;
-            order = nullptr;
-        }
-
-        if (num_frags_in_chromosomes)
-        {
-            delete[] num_frags_in_chromosomes;
-            num_frags_in_chromosomes = nullptr;
-        }
-        num_chromosomes = 0;
-    }
-
-public:
-    FragsOrder(u32 num_frags_)
-        :num_frags(num_frags_), num_chromosomes(0), num_frags_in_chromosomes(nullptr), order(nullptr)
-    {}
+    f32 threshold_ratio = 0.85;
 
-    ~FragsOrder()
-    {   
-        cleanup();
-    }
-    
-    u32 get_num_frags() const {return num_frags;}
-
-    void set_order(const std::vector<std::deque<s32>> chromosomes)
-    {   
-        cleanup(); // make sure the memory is released
-        u32 tmp = 0;
-        for (auto chromosome : chromosomes) 
-        {
-            if (chromosome.size()!=0) tmp += 1;
-        }
-        num_chromosomes = tmp;
-        num_frags_in_chromosomes = new u32[num_chromosomes];
-        order = new s32*[num_chromosomes];
-        u32 cnt= 0;
-        for (u32 i = 0; i < chromosomes.size(); i++)
-        {   
-            if (chromosomes[i].size() == 0) continue;
-            num_frags_in_chromosomes[cnt] = chromosomes[i].size();
-            order[cnt] = new s32[num_frags_in_chromosomes[cnt]];
-            for (u32 j = 0; j < num_frags_in_chromosomes[cnt]; j++)
-            {
-                order[cnt][j] = chromosomes[i][j];
-                if (order[cnt][j] == 0)
-                {
-                    std::cerr << "Error: the order should not contain ("<< order[cnt][j] << ")." << std::endl;
-                    assert(0);
-                }
-            }
-            cnt++ ;
-        }
-        if (cnt != num_chromosomes)
-        {
-            fprintf(stderr, "The cnt(%d) != num_chromosomes(%d)\n", cnt, num_chromosomes);
-            assert(0);
-        }
-    }
-
-    std::vector<s32> get_order_without_chromosomeInfor() const
-    {
-        if (num_chromosomes == 0)
-        {
-            std::cerr << "Error: the order is not set yet" << std::endl;
-            assert(0);
-        }
-        std::vector<s32> output;
-        for (u32 i = 0; i < num_chromosomes; i++)
-        {
-            for (u32 j = 0; j < num_frags_in_chromosomes[i]; j++)
-            {
-                output.push_back(order[i][j]);
-            }
-        }
-
-        #ifdef DEBUG // echo the results
-            std::cout << "\n\nSorting results \n";
-            for (u32 i = 0; i < num_chromosomes; i++)
-            {   
-                std::cout << "Chr [" << i+1 << "]: ";
-                for (u32 j = 0; j < num_frags_in_chromosomes[i]; j++)
-                {   
-                    std::cout << order[i][j] << " ";
-                }
-                std::cout << std::endl;
-            }
-        #endif //DEBUG
-
-        return output;
-    }
-};
-
-
-class FragSortTool
-{
-public:
     FragSortTool();
 
     ~FragSortTool();
@@ -279,6 +85,13 @@ class FragSortTool
         const bool doStageOne=true,
         const bool doStageTwo=false, 
         bool sort_according_len_flag=true) const;
+
+    void sort_according_yahs(
+        const LikelihoodTable& likelihood_table,
+        FragsOrder& frags_order,
+        SelectArea& select_area,
+        const f32 threshold=-0.001,
+        const Frag4compress* frags=nullptr) const;
 };
 
 
diff --git a/src/frags_order.h b/src/frags_order.h
new file mode 100644
index 0000000..ceaf93e
--- /dev/null
+++ b/src/frags_order.h
@@ -0,0 +1,121 @@
+#ifndef FRAGS_ORDER_H
+#define FRAGS_ORDER_H
+
+#include "utilsPretextView.h"
+
+
+struct FragsOrder
+{
+private:
+    u32 num_frags;                  // number of all the fragments
+    u32 num_chromosomes;
+    u32* num_frags_in_chromosomes = nullptr;  // [num_chromosomes]
+    s32** order = nullptr;                    // [num_chromosomes, num_frags_in_chromosomes], + - represents the direction of the fragment, start from 1
+    void cleanup()
+    {   
+        if (order)
+        {   
+            for (u32 i = 0; i < num_chromosomes; i++)
+            {
+                if (order[i])
+                {
+                    delete[] order[i];
+                    order[i] = nullptr;
+                }
+            }
+            delete[] order;
+            order = nullptr;
+        }
+
+        if (num_frags_in_chromosomes)
+        {
+            delete[] num_frags_in_chromosomes;
+            num_frags_in_chromosomes = nullptr;
+        }
+        num_chromosomes = 0;
+    }
+
+public:
+    FragsOrder(u32 num_frags_)
+        :num_frags(num_frags_), num_chromosomes(0), num_frags_in_chromosomes(nullptr), order(nullptr)
+    {}
+
+    ~FragsOrder()
+    {   
+        cleanup();
+    }
+    
+    u32 get_num_frags() const {return num_frags;}
+
+    void set_order(const std::vector<std::deque<s32>> chromosomes)
+    {   
+        cleanup(); // make sure the memory is released
+        u32 tmp = 0;
+        for (auto chromosome : chromosomes) 
+        {
+            if (chromosome.size()!=0) tmp += 1;
+        }
+        num_chromosomes = tmp;
+        num_frags_in_chromosomes = new u32[num_chromosomes];
+        order = new s32*[num_chromosomes];
+        u32 cnt= 0;
+        for (u32 i = 0; i < chromosomes.size(); i++)
+        {   
+            if (chromosomes[i].size() == 0) continue;
+            num_frags_in_chromosomes[cnt] = chromosomes[i].size();
+            order[cnt] = new s32[num_frags_in_chromosomes[cnt]];
+            for (u32 j = 0; j < num_frags_in_chromosomes[cnt]; j++)
+            {
+                order[cnt][j] = chromosomes[i][j];
+                if (order[cnt][j] == 0)
+                {
+                    std::cerr << "Error: the order should not contain ("<< order[cnt][j] << ")." << std::endl;
+                    assert(0);
+                }
+            }
+            cnt++ ;
+        }
+        if (cnt != num_chromosomes)
+        {
+            fprintf(stderr, "The cnt(%d) != num_chromosomes(%d)\n", cnt, num_chromosomes);
+            assert(0);
+        }
+    }
+
+    std::vector<s32> get_order_without_chromosomeInfor() const
+    {
+        if (num_chromosomes == 0)
+        {
+            std::cerr << "Error: the order is not set yet" << std::endl;
+            assert(0);
+        }
+        std::vector<s32> output;
+        for (u32 i = 0; i < num_chromosomes; i++)
+        {
+            for (u32 j = 0; j < num_frags_in_chromosomes[i]; j++)
+            {
+                output.push_back(order[i][j]);
+            }
+        }
+
+        #ifdef DEBUG // echo the results
+            std::cout << "\n\nSorting results \n";
+            for (u32 i = 0; i < num_chromosomes; i++)
+            {   
+                std::cout << "Chr [" << i+1 << "]: ";
+                for (u32 j = 0; j < num_frags_in_chromosomes[i]; j++)
+                {   
+                    std::cout << order[i][j] << " ";
+                }
+                std::cout << std::endl;
+            }
+        #endif //DEBUG
+
+        return output;
+    }
+};
+
+
+
+
+#endif // FRAGS_ORDER_H

From cf3128fffcd7cb9689ab394241453cbf1539d9d7 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:10:33 +0100
Subject: [PATCH 19/33] add pixel density extension for pixel cut

---
 src/genomeData.h | 49 ++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 49 insertions(+)

diff --git a/src/genomeData.h b/src/genomeData.h
index a503b76..7453bd4 100644
--- a/src/genomeData.h
+++ b/src/genomeData.h
@@ -93,6 +93,11 @@ struct contigs
 struct meta_data
 {
     u32 tags[64][16];
+
+    meta_data()
+    {
+        memset(tags, 0, sizeof(tags));
+    }
 };
 
 
@@ -128,6 +133,30 @@ struct map_state
 };
 
 
+/*
+用于存储计算得到的pixel_density值
+*/
+struct Extension_Graph_Data
+{   
+    bool is_prepared = false;
+    f32* data=nullptr;
+    Extension_Graph_Data(u32 num_pixel_1d) 
+    {
+        data = new f32[num_pixel_1d];
+        memset(data, 0, num_pixel_1d * sizeof(f32));
+    }
+
+    ~Extension_Graph_Data()
+    {
+        if (data)
+        {
+            delete[] data;
+            data = nullptr;
+        }
+    }
+    
+};
+
 
 // extension structures
 struct
@@ -180,6 +209,26 @@ extension_sentinel
         }
         return added_index;
     }
+
+    bool is_empty()
+    {
+        return this->head == nullptr;
+    }
+
+    bool is_graph_name_exist(const std::string& name)
+    {
+        TraverseLinkedList(this->head, extension_node)
+        {
+            if (node->type == extension_graph)
+            {
+                if (strcmp((char*)((graph*)node->extension)->name, name.c_str()) == 0)
+                {
+                    return true;
+                }
+            }
+        }
+        return false;
+    }
 };
 
 

From d1e6cf59906273b88ed86e3700a0468fa7d7a62c Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:11:14 +0100
Subject: [PATCH 20/33] seperate this part from frag sort file

---
 src/likelihood_table.h | 128 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 128 insertions(+)
 create mode 100644 src/likelihood_table.h

diff --git a/src/likelihood_table.h b/src/likelihood_table.h
new file mode 100644
index 0000000..20d1713
--- /dev/null
+++ b/src/likelihood_table.h
@@ -0,0 +1,128 @@
+#ifndef LIKELIHOOD_TABLE_H
+#define LIKELIHOOD_TABLE_H
+
+#include "utilsPretextView.h"
+#include "copy_texture.h"
+#include <fmt/core.h>
+#include <cstdio>
+struct LikelihoodTable
+{
+public:
+    f32* data = nullptr;       // with shape [num_frags, num_frags, 4]
+    u32 num_frags;   // all the frags including the filtered out ones
+    u32 size ;
+    std::unordered_set<u32> excluded_fragment_idx;
+    Frag4compress* frags = nullptr;
+
+    LikelihoodTable(
+        Frag4compress* frags, 
+        const Matrix3D<f32>* compressed_hic, 
+        const f32 threshold=10.f/32769.f, 
+        const std::vector<s32>& exclude_tag_idx=std::vector<s32>(),
+        const u32 num_pixels_1d=32768) 
+        :num_frags(frags->num), frags(frags), data(nullptr)
+    {   
+        auto is_exclude_tag = [&](u32 idx) -> bool
+        {   
+            if ((f32)frags->length[idx]/(f32)num_pixels_1d <= threshold) 
+            {
+                this->excluded_fragment_idx.insert(idx);
+                return true;
+            }
+            if (exclude_tag_idx.size()==0) return false;
+            for (auto i : exclude_tag_idx)
+            {   
+                if (i < 0) continue;
+                if (frags->metaDataFlags[idx] & (1<<i)) 
+                {
+                    this->excluded_fragment_idx.insert(idx);
+                    return true;
+                }
+            }
+            return false;
+        };
+        size = num_frags * num_frags * 4;
+        data = new f32[size];
+        for (u32 i = 0; i<num_frags * num_frags * 4; i++) data[i] = -1.f;
+        for (u32 i = 0; i<num_frags; i++)
+        {   
+            if ( is_exclude_tag(i) ) continue;
+            for (u32 j = 0; j < num_frags; j++)
+            {   
+                if ( is_exclude_tag(j)) continue;
+                (*this)(i, j, 0) = (*compressed_hic)(i, j, 0);
+                (*this)(i, j, 1) = (*compressed_hic)(i, j, 1);
+                (*this)(i, j, 2) = (*compressed_hic)(i, j, 3);
+                (*this)(i, j, 3) = (*compressed_hic)(i, j, 2);
+            }
+        }
+    }
+
+    ~LikelihoodTable()
+    {
+        if (data) 
+        {
+            delete[] data;
+            data = nullptr;
+        }
+    }
+
+    u32 get_num_frags() const {return num_frags;}
+
+    f32& operator() (const u32 i, const u32 j, const u32 k) 
+    {   
+        if (i >= num_frags || j >= num_frags || k >= 4) 
+        {
+            std::cerr << "Error: index out of range" << std::endl;
+            assert(0);
+        }
+        return data[ ((i * num_frags) << 2) + (j << 2) + k];
+    }
+
+    const f32& operator() (const u32 i, const u32 j, const u32 k) const
+    {
+        if (i >= num_frags || j >= num_frags || k >= 4) 
+        {
+            std::cerr << "Error: index out of range" << std::endl;
+            assert(0);
+        }
+        return data[ ((i * num_frags) << 2) + (j << 2) + k];
+    }
+
+    const void output_fragsInfo_likelihoodTable(const std::string& file_name, const Matrix3D<f32>* compressed_hic) const
+    {       
+        FILE* ofs = fopen(file_name.c_str(), "w");
+        if (!ofs)
+        {
+            fmt::print("Error: failed to open: {}", file_name) ;
+            assert(0);
+        }
+
+        fmt::print( ofs, "# frags num: {} \n", num_frags);
+        fmt::print( ofs, "# frags_index\tlength\trepeat\tcoverage\n");
+        for (u32 i = 0; i < num_frags; i++)
+        {
+            fmt::print(
+                ofs, 
+                "{}\t{}\t{}\t{}\t{}\n", 
+                i, 
+                frags->length[i], 
+                0, 
+                0, 
+                0);
+        }
+        fmt::print(ofs, "\n");
+
+        // likelihood table
+        fmt::print(ofs, "# likelihood table\n");
+        compressed_hic->output_to_file(ofs);
+
+        fclose(ofs);
+
+        fmt::print("[DEBUG_OUTPUT_LIKELIHOOD_TABLE]: Likelihood table saved to: {}\n", file_name);
+    }
+};
+
+
+
+#endif // LIKELIHOOD_TABLE_H
\ No newline at end of file

From 69388ac291f317860c84421eabb69fba52efc6c1 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:11:57 +0100
Subject: [PATCH 21/33] re-organise files

---
 ColorMapData.cpp                              | 7330 -----------------
 Resources.cpp => src/Resources.cpp            |    0
 .../TextureLoadQueue.cpp                      |    0
 glad.c => src/glad.c                          |    0
 4 files changed, 7330 deletions(-)
 delete mode 100644 ColorMapData.cpp
 rename Resources.cpp => src/Resources.cpp (100%)
 rename TextureLoadQueue.cpp => src/TextureLoadQueue.cpp (100%)
 rename glad.c => src/glad.c (100%)

diff --git a/ColorMapData.cpp b/ColorMapData.cpp
deleted file mode 100644
index 6380ed5..0000000
--- a/ColorMapData.cpp
+++ /dev/null
@@ -1,7330 +0,0 @@
-/*
-Copyright (c) 2021 Ed Harry, Wellcome Sanger Institute
-
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
-*/
-
-#ifndef HEADER_H  // 防止重定义错误
-#include "Header.h"
-#endif
-
-#define Number_of_Color_Maps (sizeof(Color_Map_Data) / sizeof(Color_Map_Data[0]))
-
-
-// global_variable
-f32 Color_Map_Data[][768]= 
-{
-    {
-        0.0028108909432653585f,0.00023971464598377364f,0.0139849755645124f,0.004019847439892661f,0.0009782136527954485f,0.018722506642108738f,0.00553592753551016f,0.0018785803659367136f,0.024433316141969492f,0.007352275238375931f,0.0029285938665143995f,0.031151041807332518f,0.009463157808182366f,0.004122022987412015f,0.03885690760774683f,0.01187516425529045f,0.005443920038504773f,0.04716007356275464f,0.01459882064653067f,0.006882989229735782f,0.055492610304536484f,0.017636358974306172f,0.008432150815965444f,0.06383358056387142f,0.02101796809093932f,0.010065826612517249f,0.07226064067376634f,0.02474393539914235f,0.011784158597413285f,0.08070837752145013f,0.028837804117072482f,0.013566889108044652f,0.0892198151329586f,0.033326247190342086f,0.015396259291835603f,0.09780925050843237f,0.03821642380815322f,0.017268324012382f,0.10644014290558228f,0.043430971060037175f,0.01915497115064957f,0.11516213940852653f,0.04869376377787085f,0.02104088811842354f,0.12396774807241907f,0.05400341174839573f,0.02291932904231329f,0.13283301770235545f,0.05937446015473189f,0.0247676534264442f,0.14177532141983173f,0.06482608437087267f,0.02654913857103515f,0.1508303085225832f,0.07034982700960057f,0.02826358681647732f,0.15995532205231544f,0.07595041067967645f,0.02988842186530461f,0.16914872242514994f,0.08163782650304266f,0.03140401564457005f,0.17841319040234854f,0.08742545856288386f,0.03277327931516748f,0.18776813322972746f,0.09331557740511903f,0.033979154409542016f,0.19719580690248006f,0.09930765162558787f,0.0350072967471369f,0.2066795437361335f,0.105402262424314f,0.0358364665010164f,0.21620646706666896f,0.11160571005993522f,0.03644452450798942f,0.22577039634735835f,0.11791676181366892f,0.03681366238931068f,0.2353533948315683f,0.12433690815307286f,0.03692755679726379f,0.2449402569352536f,0.130861794027333f,0.0367738251284066f,0.25450501190842256f,0.13748923050603104f,0.036346007036179566f,0.2640284014268764f,0.14420943125208832f,0.03564202294027663f,0.2734764666562528f,0.1510154327209675f,0.034667061792608964f,0.282818469890769f,0.1578973488336295f,0.03343446973536865f,0.29202361159436785f,0.16483784305323396f,0.031964803318426126f,0.3010500119503021f,0.1718271766829396f,0.03028588882044497f,0.3098670941619176f,0.17884663206387366f,0.028428459100255845f,0.31843622115540393f,0.1858839310036252f,0.026431189198060315f,0.3267275455559629f,0.19291386085475284f,0.02434997053816215f,0.33470171605932664f,0.19992220593124488f,0.02223366886218197f,0.3423326727997547f,0.20689679435120226f,0.020130048116207458f,0.34960418789627484f,0.21382118808651815f,0.018085948264879845f,0.3564954212152869f,0.22068623935476184f,0.016147650423161998f,0.3630028398603805f,0.22748122267063475f,0.014355534805442475f,0.3691191047604013f,0.2342036115077947f,0.012742013345224536f,0.3748502927327437f,0.24084627534795414f,0.011335284339993059f,0.3802005626251185f,0.2474096049888127f,0.01015735838738674f,0.3851833827603531f,0.25389243988976723f,0.009226759851114728f,0.3898140016153026f,0.2602964252633522f,0.00854893129141548f,0.3941060822693162f,0.2666282702790725f,0.00812889690971778f,0.3980799156919193f,0.2728881836012367f,0.00796777080577268f,0.4017525543519634f,0.2790829843835597f,0.008062521199618247f,0.4051419276648617f,0.2852151444728726f,0.008411444340647998f,0.4082654255391529f,0.2912905872823367f,0.00900696229306688f,0.41114096865607663f,0.2973134150580773f,0.009846182094486277f,0.4137873198524331f,0.30328762487316085f,0.010914425884934178f,0.41621716948105103f,0.30922074689328993f,0.012204295605290823f,0.4184468468595007f,0.3151147196399223f,0.013704677218996394f,0.4204882743475308f,0.32097639646999004f,0.015407335075544953f,0.422354544204609f,0.3268065495346441f,0.017302460620371732f,0.4240567802998516f,0.33261022038705f,0.01938047884522835f,0.4256065016808433f,0.33839282001375265f,0.021633450732188426f,0.4270121205544362f,0.34415378701257915f,0.024052574290460142f,0.4282829405343027f,0.3498988918271099f,0.02662930883000891f,0.4294266653701039f,0.35562785044944545f,0.029355256148443618f,0.43044874030607067f,0.3613465905418141f,0.032224412611494564f,0.4313587270615997f,0.3670534821768672f,0.03523065438609198f,0.4321591931513488f,0.3727525670853749f,0.03836576069038865f,0.43285640046483465f,0.37844680040449136f,0.04159569733823424f,0.4334555349565255f,0.3841355329782064f,0.0447886211961958f,0.43396029607797f,0.3898229228233572f,0.04794477857573524f,0.43437397708144543f,0.3955064846961563f,0.0510632756323768f,0.4347003809417806f,0.40118951085798776f,0.05414611240426622f,0.43494369314594333f,0.4068728001787072f,0.05719222228379529f,0.4351048143031719f,0.4125561659556076f,0.060199930687386674f,0.43518581888017777f,0.4182439569546356f,0.06317128301685854f,0.4351887312325765f,0.4239333136155822f,0.06610722021585971f,0.4351145439646258f,0.4296274038777328f,0.06900863885365052f,0.43496624880705104f,0.4353241735726936f,0.07187617069246008f,0.4347439247528073f,0.4410274908705713f,0.07471304329299233f,0.4344484704151144f,0.4467336685929623f,0.0775180640738074f,0.4340820535925406f,0.4524463832839706f,0.08029430493651157f,0.4336434847073313f,0.4581629519603063f,0.08304144488312268f,0.43313601580946376f,0.4638857742577249f,0.08576376530326196f,0.432560538831606f,0.4696151520071452f,0.08846085367401663f,0.4319148958388981f,0.4753481825505309f,0.09113464993118832f,0.43120041645865226f,0.4810895258391375f,0.09378748211627656f,0.4304157480526235f,0.4868354565595182f,0.09641864471026834f,0.4295633056730111f,0.4925879104627671f,0.09903140399175724f,0.4286426514306487f,0.4983447389473999f,0.1016274927015326f,0.42765327068757697f,0.5041074631513899f,0.10420806397082304f,0.4265959365704464f,0.5098765596563365f,0.10677546051510424f,0.42547036940392857f,0.5156490293025526f,0.1093311312138189f,0.4242771408904675f,0.5214267691404487f,0.11187572305489003f,0.4230146529372112f,0.5272086759775392f,0.11441057366051095f,0.42168356166547744f,0.532994806491818f,0.1169395877882898f,0.4202831796072165f,0.5387841442602808f,0.1194627941731479f,0.4188152468547687f,0.5445752162374711f,0.12198117471380225f,0.4172784035251669f,0.550371411794682f,0.1244988719881559f,0.4156732538227701f,0.5561677070769875f,0.12701492382448096f,0.4139986102201316f,0.5619661289256851f,0.12953294280559788f,0.4122546278213512f,0.5677646289876168f,0.1320539831226978f,0.4104422052829828f,0.5735651914497292f,0.1345794432111999f,0.4085604229551985f,0.579363842280652f,0.13711144234826714f,0.40661024958520653f,0.5851601782714847f,0.13965174564712465f,0.4045922065996268f,0.5909574913960268f,0.14220219189728348f,0.4025047796622188f,0.5967498429288651f,0.14476378100450754f,0.4003480187832498f,0.6025402031913455f,0.14733996311413705f,0.3981218371952534f,0.6083246722235879f,0.14993155878890302f,0.3958283734200633f,0.614106176674807f,0.15253988593803774f,0.39346646772201055f,0.6198788025701961f,0.1551679529737472f,0.3910373196621276f,0.6256450195747313f,0.15781633943679754f,0.3885413427073905f,0.6314052760627059f,0.1604878602151773f,0.3859778883860768f,0.637153671587794f,0.16318436257638144f,0.3833482435611563f,0.6428950257611473f,0.16590822465679886f,0.3806510890190323f,0.6486225037023093f,0.16865910560313505f,0.377887791334078f,0.6543400237812237f,0.17144146614273248f,0.37505894495304937f,0.6600427220068441f,0.17425592860145306f,0.37216699339894543f,0.6657309854853277f,0.17710424966773933f,0.3692102159546668f,0.671404337541839f,0.1799902187518531f,0.3661898345073855f,0.6770589159382479f,0.182912273416154f,0.3631083856894559f,0.6826975017706738f,0.1858760680408481f,0.3599642927599365f,0.6883152752699135f,0.18888000215377848f,0.3567601555176547f,0.6939130936485612f,0.1919297404645243f,0.3534953157761731f,0.6994880189268522f,0.19502263596715352f,0.3501704524961132f,0.7050385790436794f,0.19816353086298186f,0.3467877021993576f,0.7105661060122954f,0.20135422039004186f,0.3433461843544269f,0.7160668362113027f,0.20459503779649424f,0.3398506288928686f,0.7215395551311659f,0.20788868051000065f,0.33629822944329f,0.7269824482031878f,0.21123646007205965f,0.3326937750819364f,0.7323962498882949f,0.21464109358472336f,0.32903540290632194f,0.7377762325604813f,0.21810179955417053f,0.3253269506392982f,0.7431240483143292f,0.22162240347649032f,0.3215664902244945f,0.7484350306113071f,0.2252020416267312f,0.31775890499434706f,0.7537094301237001f,0.2288447108656564f,0.31390318773099685f,0.7589465700904082f,0.2325513179438209f,0.3099993007197641f,0.7641427572948342f,0.23632095843046075f,0.3060511850445163f,0.7692985739778287f,0.24015863799868106f,0.30205775317511674f,0.7744103608994938f,0.24406133999228075f,0.2980219987542918f,0.7794786898294086f,0.2480331020679257f,0.2939437813724841f,0.784498961427673f,0.25207289596277377f,0.2898270979003709f,0.7894723227633637f,0.2561838131658038f,0.2856708304197678f,0.7943970958045569f,0.2603648353176219f,0.2814768234287905f,0.7992696028577546f,0.2646179852095233f,0.27724708700681994f,0.8040913179125936f,0.26894337006009084f,0.2729813802962653f,0.8088576856149733f,0.2733398560423693f,0.2686817498061221f,0.8135691155681037f,0.2778117319366859f,0.2643498423217051f,0.8182230524153231f,0.2823557514806422f,0.259985751711718f,0.8228187030680942f,0.2869730612436132f,0.25559221349785954f,0.8273560901963927f,0.2916649238384595f,0.2511669292787484f,0.8318308105503314f,0.2964310422772357f,0.24671393625869514f,0.8362441082220449f,0.3012708228202772f,0.2422317977788539f,0.8405915992051021f,0.3061839140531289f,0.2377225424355419f,0.8448754218830034f,0.3111718447857894f,0.2331856496191531f,0.8490922820892888f,0.3162321293857926f,0.228623159368174f,0.8532412230472858f,0.3213650843479945f,0.22403563620206654f,0.8573221666233893f,0.3265710944674797f,0.21942058959604985f,0.8613319022896901f,0.33184860342904376f,0.2147810099094718f,0.8652713962071545f,0.33719833283248074f,0.2101141748022448f,0.8691385576642853f,0.3426165911885936f,0.20542406206466107f,0.8729332123141957f,0.34810624963200004f,0.20070581716761515f,0.876653351754305f,0.3536635338530963f,0.195963342516276f,0.8802978200638566f,0.3592889341611301f,0.19119290796426572f,0.8838684350097556f,0.3649829259277349f,0.18639467968336265f,0.8873602825148426f,0.3707406901646968f,0.18156845817161654f,0.890776984152265f,0.37656622385933053f,0.17671201671816958f,0.894114774529188f,0.3824525761944193f,0.1718270686976245f,0.8973741852518462f,0.388403855880936f,0.16690816145867332f,0.9005544630688599f,0.3944150503711016f,0.1619568111176787f,0.9036543550377922f,0.4004867447796602f,0.15697061628964454f,0.906674539526174f,0.4066185337569268f,0.1519460056654663f,0.9096133361541572f,0.4128073699691127f,0.14688407634155784f,0.9124721600806108f,0.419054437663371f,0.14177960453650024f,0.9152474734642112f,0.4253546069506915f,0.1366313113758783f,0.9179415758304628f,0.43171113773754055f,0.13143638978707065f,0.920553002660083f,0.4381188215811716f,0.12619301932742935f,0.9230823502164308f,0.44457832435590056f,0.12089804008534748f,0.9255281583142996f,0.4510903508911226f,0.115545643733264f,0.9278900885067396f,0.4576496172590345f,0.11013577030361796f,0.9301682735211648f,0.4642584989903114f,0.1046620258358874f,0.9323634323492204f,0.470912661585951f,0.09912286212434912f,0.9344736508745052f,0.4776135317653246f,0.09351101566047268f,0.9364987505473671f,0.4843577227601096f,0.0878266775099058f,0.9384406797563624f,0.4911467021518643f,0.08206088279212748f,0.94029638756102f,0.4979770365588466f,0.07621335449370882f,0.9420673342348812f,0.5048474657057321f,0.07027747421356159f,0.943752835993572f,0.5117607949728211f,0.06424712981979414f,0.9453530086488392f,0.5187105010375903f,0.05812067587313205f,0.9468675926635921f,0.5257001589178919f,0.05189145509095276f,0.9482957691429188f,0.5327252159488806f,0.04556084013205488f,0.9496381978647562f,0.5397882797332773f,0.039112475813857225f,0.9508941688983458f,0.5468847469930004f,0.0329036576052177f,0.9520630153802706f,0.5540144303588769f,0.02723224588536035f,0.953145915117146f,0.561180174976782f,0.02211629990383893f,0.9541413109174116f,0.5683763342018563f,0.017577226490710227f,0.9550496977801548f,0.5756055764119579f,0.013636710338596076f,0.9558695965052876f,0.5828632209633262f,0.010318691317706429f,0.9566033661162672f,0.5901529783933634f,0.007647565236664239f,0.9572476931570264f,0.5974691178771274f,0.005653645257891558f,0.9578046847922824f,0.6048145165947891f,0.004365191753448543f,0.9582725531896952f,0.6121880260521443f,0.003815358887601479f,0.9586519941780332f,0.6195869032693965f,0.004038403733227168f,0.9589432359293408f,0.6270139069375642f,0.0050725965725081324f,0.9591452081062959f,0.6344632304037671f,0.006957730473759643f,0.9592570195286264f,0.6419386757168154f,0.009736863918784179f,0.959279647369708f,0.6494364125427067f,0.013456640602907992f,0.9592132245352332f,0.656956325469712f,0.01816672798060043f,0.9590557212865652f,0.6645013350837338f,0.023921375872797367f,0.9588093894625144f,0.6720645507841265f,0.03077795215314844f,0.9584721753486866f,0.6796498174204002f,0.03879846557918752f,0.9580453442160642f,0.6872532448327471f,0.04754586688050162f,0.9575278692108288f,0.6948766727751612f,0.05639000971583572f,0.9569201366476526f,0.7025161872712079f,0.0653272271581101f,0.95622222088975f,0.7101726957961991f,0.07436297745193415f,0.9554351276542836f,0.7178471084562286f,0.08350080196415507f,0.9545585022368179f,0.7255344740989906f,0.09274251876146856f,0.95359012489534f,0.7332386758872225f,0.10209834873587548f,0.9525337955331732f,0.7409547279741554f,0.11157046121752068f,0.9513894499647296f,0.7486824818249713f,0.1211710906915065f,0.9501579394841028f,0.7564199576460588f,0.13090499817983486f,0.9488415317142022f,0.7641650350600445f,0.14078130910755238f,0.9474413345042012f,0.771919616464573f,0.15081436111263724f,0.94595966334309f,0.7796766610808825f,0.16100576302194788f,0.9443984072166398f,0.7874390279521042f,0.17137304382242416f,0.9427621029656729f,0.7952006533320073f,0.18192158708279615f,0.941049699471568f,0.8029643911682793f,0.19266999933666f,0.9392640761941338f,0.8107231422033202f,0.20362775766846952f,0.9374158407051084f,0.8184757458260994f,0.21480475650403494f,0.9355108618569462f,0.8262179886623421f,0.22621716723273394f,0.9335594549301716f,0.8339427520154467f,0.23787212746051176f,0.931563069969426f,0.8416528004815411f,0.2497953642795853f,0.9295295956536241f,0.8493389652326568f,0.2620016582307053f,0.927483113892923f,0.8569948963180799f,0.2745012512443162f,0.9254408066615336f,0.8646104422830595f,0.28729978853917554f,0.9234033328223992f,0.8721861539859546f,0.3004428353086889f,0.9214094735748476f,0.8797058578924303f,0.3139219503891981f,0.9194909179295936f,0.8871561145803522f,0.3277441860383484f,0.9176568416750964f,0.8945363755302687f,0.34196084494650136f,0.9159695644129056f,0.9018205155640128f,0.3565351604857609f,0.9144604035423052f,0.9089997667171816f,0.3714975395911552f,0.91318592603147f,0.9160520672097188f,0.3868282485461301f,0.9122023005193108f,0.922961587897254f,0.40251902826986985f,0.9115712707791952f,0.929705490864802f,0.4185354153286997f,0.9113589610617012f,0.9362663045373388f,0.4348346331778865f,0.911626841657708f,0.9426265331991612f,0.4513649463134185f,0.9124352877263222f,0.9487690293791992f,0.4680385459173532f,0.9138280045063198f,0.9546871530908342f,0.484787690701253f,0.9158401268012052f,0.9603724114950956f,0.5015022040367715f,0.9184812096938036f,0.9658321994102153f,0.5181206698543415f,0.9217499769357954f,0.971070575662174f,0.5345393044762992f,0.925624725190333f,0.9761026354466212f,0.5507027927528236f,0.9300724918488174f,0.9809461333524748f,0.5665954408447937f,0.93505290022375f,0.9856188564238579f,0.5821026104689219f,0.9405209651098092f,0.9901403564326076f,0.5972696263274475f,0.9464291624522388f,0.9945258421274078f,0.6120801460623929f,0.9527369951544834f,0.9987960656048062f,0.6265350434944937f,0.9593995073432898f,1.0029632847639067f,0.6406264784793351f}, {0.0f,0.0f,0.0f,0.0124985675549f,0.00376710208302f,0.00210713336054f,0.0249971351098f,0.00753420416604f,0.00421426672107f,0.037495702664699995f,0.0113013062491f,0.00632140008161f,0.0492581088139f,0.0150684083321f,0.008428533442149999f,0.0594165938721f,0.0188355104151f,0.010535666802700001f,0.0684471877541f,0.0226026124981f,0.012642800163200001f,0.0766363089899f,0.0263697145811f,0.014749933523799998f,0.0841678787653f,0.0301368166641f,0.016857066884299998f,0.09116808551160001f,0.0339039187472f,0.0189642002448f,0.09772782947889999f,0.0376710208302f,0.021071333605400002f,0.103915087694f,0.0414146679499f,0.023178466965900004f,0.109782226099f,0.044974293467f,0.025285600326400003f,0.11537057005200001f,0.048364772838199996f,0.027392733687f,0.120713391496f,0.0516062923368f,0.0294998670475f,0.125837934386f,0.054715381023000005f,0.031607000408099994f,0.130854584645f,0.0576505684502f,0.0337193392032f,0.13604190193f,0.06025908266f,0.0358534453625f,0.14140417663f,0.0625542306901f,0.0380109209821f,0.146927404384f,0.0645557290717f,0.0401922566139f,0.152598730721f,0.0662787359665f,0.0423389683918f,0.158389635306f,0.0677778384763f,0.0444339133373f,0.164222599234f,0.0692401824912f,0.046456344033f,0.170089016456f,0.07068488518239999f,0.04840869742790001f,0.17598792054499998f,0.0721120614255f,0.050295820716699996f,0.18191842184300003f,0.0735218171731f,0.0521219928396f,0.187879698899f,0.07491424994469999f,0.05389101229169999f,0.19387099103f,0.0762894492712f,0.0556062682658f,0.19989159185400002f,0.07764749709919999f,0.057270798828499996f,0.20594084362799997f,0.0789884681608f,0.0588873389018f,0.212018132302f,0.0803124303088f,0.0604583601449f,0.218122883178f,0.0816194448242f,0.0619861043454f,0.224254557086f,0.0829095666958f,0.0634726115611f,0.23041264701699998f,0.0841828448758f,0.06491974398580001f,0.236596675144f,0.0854393225134f,0.0663292063043f,0.242806190184f,0.0866790371676f,0.0677025631472f,0.249040765054f,0.08790202100120001f,0.0690412541328f,0.255299994784f,0.0891083009582f,0.0703466068909f,0.26158349465999997f,0.09029789892519999f,0.07161984838819999f,0.267890898557f,0.0914708318774f,0.07286211481869999f,0.274221857454f,0.0926271120122f,0.0740744602728f,0.280576038088f,0.0937667468691f,0.0752578643655f,0.286953121761f,0.09488973943880001f,0.07641323897f,0.293352803248f,0.0959960882604f,0.077541434183f,0.299774789817f,0.0970857875089f,0.0786432436236f,0.30621880034800003f,0.0981588270728f,0.0797194091565f,0.312684564525f,0.0992151926224f,0.0807706251109f,0.319171134912f,0.10025502620999999f,0.0818055914811f,0.32567691454500003f,0.10127863170399999f,0.0828412386963f,0.332201791044f,0.102285972873f,0.0838775716485f,0.33874565083999997f,0.103277010196f,0.08491459706849999f,0.34530837961099997f,0.104251700818f,0.0859523214123f,0.351889862505f,0.105209998536f,0.0869907508714f,0.358489984344f,0.10615185378900001f,0.08802989138360001f,0.36510862979900005f,0.107077213638f,0.0890697486419f,0.37174568354500004f,0.10798602174699999f,0.09011032810450001f,0.378401030403f,0.10887821835599999f,0.0911516350032f,0.385074555462f,0.109753740249f,0.0921936743521f,0.391766144185f,0.110612520723f,0.09323645095560001f,0.398475682511f,0.111454489549f,0.09427996941639999f,0.40520305693400005f,0.112279572927f,0.09532423414249999f,0.411948154577f,0.113087693439f,0.0963692493548f,0.418710863264f,0.113878770002f,0.0974150190938f,0.425491071569f,0.114652717803f,0.09846154722599999f,0.43228866887100004f,0.115409448248f,0.09950883745030001f,0.439103545394f,0.116148868889f,0.100556893304f,0.44593559224600005f,0.116870883353f,0.101605718168f,0.452784701449f,0.117575391273f,0.10265531527499999f,0.45965076597099996f,0.118262288197f,0.10370568771f,0.46653367974f,0.118931465509f,0.10475683842f,0.473433337672f,0.11958281033f,0.105808770216f,0.480349635682f,0.120216205427f,0.10686148578f,0.48728247069700004f,0.12083152909900001f,0.10791498766799999f,0.49423174066500003f,0.121428655073f,0.108969278312f,0.5011973445669999f,0.12200745238199999f,0.110024360032f,0.5081791824140001f,0.122567785241f,0.11108023502999999f,0.515177155256f,0.123109512914f,0.112136905403f,0.522191165184f,0.12363248957299999f,0.113194373138f,0.529221115323f,0.12413656414999999f,0.114252640123f,0.536266909839f,0.12462158018f,0.115311708147f,0.543328453928f,0.12508737563300001f,0.116371578905f,0.550405653818f,0.125533782739f,0.11743225399700001f,0.557498416761f,0.125960627799f,0.118493734936f,0.564606651025f,0.126367730992f,0.11955602315f,0.571730265893f,0.126754906165f,0.12061911998199999f,0.57886917165f,0.12712196061f,0.121683026697f,0.586023279576f,0.127468694834f,0.12274774447899998f,0.59319250194f,0.127794902308f,0.123813274439f,0.600376751988f,0.12810036920999998f,0.124879617617f,0.607575943936f,0.128384874145f,0.125946774979f,0.614789992957f,0.128648187853f,0.127014747427f,0.6220188151739999f,0.128890072894f,0.128083535793f,0.6292623276470001f,0.129110283326f,0.12915314085000001f,0.636520448365f,0.129308564349f,0.13022356330499998f,0.643793096232f,0.129484651934f,0.131294803809f,0.65108019106f,0.129638272432f,0.132366862954f,0.658381653554f,0.12976914215500002f,0.133439741276f,0.665697405303f,0.129876966926f,0.13451343925599998f,0.67302736877f,0.129961441609f,0.135587957327f,0.680371467279f,0.13002224960699998f,0.136663295866f,0.686819633179f,0.132061139735f,0.13726053577199998f,0.690577321338f,0.139841365035f,0.136421938331f,0.694333453232f,0.147366583596f,0.13555415609899998f,0.69808804669f,0.154670100683f,0.134656431336f,0.701841119323f,0.161779110231f,0.133727968564f,0.705592688522f,0.16871613022f,0.132767931803f,0.709342771462f,0.17550003480300003f,0.131775441528f,0.713091385109f,0.18214681303599997f,0.130749571337f,0.716838546221f,0.188670137478f,0.12968934428099999f,0.720584271352f,0.195081797593f,0.12859372881699999f,0.724328576854f,0.20139203510899997f,0.127461634333f,0.7280714788830001f,0.207609807026f,0.126291906198f,0.7318129934000001f,0.21374299438299998f,0.125083320255f,0.7355531361740001f,0.21979856979499998f,0.123834576701f,0.7392919227879999f,0.225782733244f,0.122544293247f,0.743029368637f,0.23170102312899998f,0.12121099746899999f,0.746765488936f,0.237558407857f,0.11983311822f,0.7505002987199999f,0.24335936194200003f,0.118408975956f,0.754233812848f,0.249107929694f,0.116936771817f,0.7579660460050001f,0.254807778863f,0.115414575242f,0.761697012704f,0.26046224609499996f,0.11384030988800001f,0.765426727292f,0.266074375673f,0.112211737544f,0.7691552039480001f,0.27164695270100003f,0.11052643969699999f,0.77288245669f,0.277182531676f,0.10878179629899999f,0.7766084993729999f,0.28268346119000004f,0.106974961217f,0.7803333456970001f,0.288151905399f,0.105102833683f,0.784057009203f,0.293589862731f,0.103162024955f,0.787779503281f,0.29899918227600003f,0.101148819133f,0.791500841169f,0.304381578185f,0.0990591268518f,0.795221035957f,0.309738642359f,0.0968884301939f,0.7989401005869999f,0.315071855678f,0.0946317167004f,0.8026580478569999f,0.32038259796199997f,0.0922833997264f,0.8063748904260001f,0.325672156837f,0.08983722152339998f,0.810090640808f,0.33094173563900003f,0.08728613424630001f,0.813805311384f,0.336192460496f,0.08462215241789998f,0.817518914394f,0.341425386671f,0.0818361680288f,0.821231461948f,0.346641504267f,0.0789177160322f,0.824942966022f,0.351841743372f,0.0758546729705f,0.828653438464f,0.35702697870300004f,0.0726328638943f,0.8323628909909999f,0.362198033804f,0.06923554106180001f,0.8360713351960001f,0.36735568486500003f,0.065642679414f,0.839778782546f,0.372500664176f,0.061830003621300006f,0.843485244386f,0.377633663287f,0.05776761042780001f,0.847190731941f,0.382755335881f,0.0534179600786f,0.8508952563140001f,0.387866300395f,0.048732844522199996f,0.854598828494f,0.39296714243f,0.0436486154689f,0.858301459351f,0.398058416948f,0.0380796196638f,0.8620031596430001f,0.403140650294f,0.032290438683500004f,0.865703940014f,0.408214342055f,0.026411833944599995f,0.869403810997f,0.413279966769f,0.020443258979599997f,0.870620849202f,0.42025968812700004f,0.0196207043343f,0.871622154301f,0.42734161938799997f,0.0192250591085f,0.87260643877f,0.434372519047f,0.018839607349200004f,0.87357362974f,0.441355114094f,0.0184645577569f,0.8745236535169999f,0.448291940683f,0.0181001190322f,0.875456435563f,0.45518536158999995f,0.0177464998757f,0.876371900482f,0.462037581704f,0.017403908988f,0.8772699720039999f,0.468850661816f,0.0170725550697f,0.878150572968f,0.47562653091000007f,0.0167526468212f,0.8790136253059999f,0.482366997164f,0.0164443929432f,0.879859050028f,0.489073757822f,0.0161480021363f,0.880686767202f,0.49574840805799997f,0.0158636831011f,0.8814966959389999f,0.502392448976f,0.015591644538f,0.8822887543729999f,0.509007294828f,0.0153320951477f,0.883062859646f,0.51559427955f,0.0150852436308f,0.883818927886f,0.522154662692f,0.0148512986878f,0.884556874192f,0.528689634804f,0.0146304690194f,0.885276612612f,0.5352003223390001f,0.014422963326f,0.885978056125f,0.541687792133f,0.0142289903082f,0.886661116621f,0.548153055488f,0.0140487586667f,0.8873257048790001f,0.554597071914f,0.013882477102000001f,0.8879717305500001f,0.561020752565f,0.013730354314700002f,0.888599102131f,0.56742496338f,0.0135925990053f,0.889207726947f,0.573810527989f,0.013469419874499998f,0.889797511128f,0.580178230371f,0.013361025622700001f,0.890368359587f,0.586528817319f,0.0132676249506f,0.890920175994f,0.592863000709f,0.0131894265588f,0.891452862757f,0.5991814596f,0.0131266391478f,0.891966320994f,0.605484842178f,0.0130794714182f,0.892460450511f,0.611773767557f,0.013048132070599999f,0.8929351497750001f,0.6180488274539999f,0.0130328298055f,0.893390315889f,0.624310587737f,0.0130337733235f,0.893825844568f,0.6305595898749999f,0.0130511713252f,0.894241630107f,0.636796352279f,0.013085232511200001f,0.8946375653569999f,0.643021371558f,0.013136165582f,0.895013541696f,0.649235123688f,0.0132041792383f,0.8953694490009999f,0.6554380651039999f,0.0132894821805f,0.895705175617f,0.6616306337210001f,0.0133922831093f,0.896020608327f,0.6678132498899999f,0.013512790725200001f,0.8963156323200001f,0.673986317291f,0.0136512137289f,0.896590131162f,0.6801502237729999f,0.013807760820799998f,0.8968439867620001f,0.686305342141f,0.013982640701600001f,0.897077079337f,0.692452030892f,0.0141760620719f,0.8972892873780001f,0.698590634912f,0.014388233632100001f,0.8974804876200001f,0.704721486124f,0.014619364082899999f,0.897650554997f,0.7108449041079999f,0.0148696621249f,0.897799362612f,0.716961196674f,0.0151393364586f,0.897926781699f,0.7230706604100001f,0.0154285957846f,0.8980326815799999f,0.729173581195f,0.0157376488035f,0.898116929626f,0.735270234683f,0.0160667042159f,0.8981793912200001f,0.741360886763f,0.0164159707222f,0.89821992971f,0.7474457939910001f,0.0167856570232f,0.8982384063700001f,0.7535252040000001f,0.0171759718194f,0.898234680352f,0.7595993558840001f,0.0175871238113f,0.898208608643f,0.7656684805699999f,0.0180193216995f,0.898160046015f,0.771732801162f,0.0184727741846f,0.898088844982f,0.777792533271f,0.0189476899672f,0.8979948557459999f,0.783847885327f,0.019444277747900002f,0.8978779261480001f,0.7898990588770001f,0.019962746227200002f,0.8977379016150001f,0.795946248863f,0.020503304105599997f,0.897574625107f,0.801989643893f,0.0210661600838f,0.897387937062f,0.808029426494f,0.0216515228624f,0.897177675337f,0.8140657733519999f,0.022259601141900003f,0.896943675152f,0.820098855544f,0.0228906036229f,0.8966857690279999f,0.8261288387540001f,0.023544739005999998f,0.8985530078489999f,0.831233249825f,0.06826100781569999f,0.9028903792290001f,0.835257282572f,0.127173791555f,0.907153202363f,0.8392904152249999f,0.16936596405f,0.9113400634639999f,0.8433327452760001f,0.20455362557700002f,0.9154495631820001f,0.847384366892f,0.235764224808f,0.9194803145479999f,0.8514453709909999f,0.264391329421f,0.923430940957f,0.85551584532f,0.291204862021f,0.927300074214f,0.859595874528f,0.316679372942f,0.931086352617f,0.8636855402410001f,0.341129701928f,0.934788419075f,0.867784921127f,0.364776124263f,0.938404919262f,0.8718940929729999f,0.38777907412100004f,0.941934499804f,0.8760131287430001f,0.41025913859200003f,0.945375806484f,0.880142098652f,0.43230927905f,0.948727482467f,0.884281070227f,0.454002666231f,0.951988166539f,0.8884301083689999f,0.47539790299999996f,0.955156491362f,0.892589275414f,0.49654262079899997f,0.9582310817219999f,0.896758631197f,0.51747602533f,0.96121055279f,0.9009382331069999f,0.538230741549f,0.9640935083679999f,0.9051281361460001f,0.55883417852f,0.9668785391320001f,0.909328392984f,0.579309557384f,0.969564220862f,0.913539054016f,0.599676697971f,0.9721491126430001f,0.917760167413f,0.619952629264f,0.9746317550560001f,0.9219917791760001f,0.640152069209f,0.9770106683249999f,0.9262339331899999f,0.660287806118f,0.979284350439f,0.930486671268f,0.6803710050120001f,0.981451275229f,0.934750033206f,0.700411455972f,0.9835098904f,0.9390240568279999f,0.720417777213f,0.9854586155060001f,0.943308778035f,0.7403975824410001f,0.987295839871f,0.9476042308499999f,0.760357619768f,0.9890199204369999f,0.951910447464f,0.780303887791f,0.990629179549f,0.956227458278f,0.8002417331899999f,0.9921219026450001f,0.960555291948f,0.8201759332430001f,0.993496335869f,0.9648939754279999f,0.840110765987f,0.9947506835750001f,0.9692435340090001f,0.8600500701389999f,0.995883105733f,0.973603991364f,0.879997296537f,0.99689171521f,0.9779753695799999f,0.899955552467f,0.997774574927f,0.982357689206f,0.9199276400409999f,0.9985296948710001f,0.986750969284f,0.939916089534f,0.9991550289650001f,0.9911552273890001f,0.9599231884430001f,0.999648471753f,0.995570479665f,0.9799510069290001f,1.0f,1.0f,1.0f
-    },
-    {
-        0.0f,0.0f,0.0f,0.017991146422900003f,0.0009200649235139999f,0.0156157177605f,0.0359573389421f,0.0017855321075899999f,0.0319686152223f,0.052717504614f,0.00252887372289f,0.0485944475368f,0.06645936690569999f,0.00322312300316f,0.0629875526327f,0.07837890807539999f,0.00380156463246f,0.0758728198312f,0.08901944585010002f,0.00426758967588f,0.0877407961252f,0.0985268713734f,0.00475880811033f,0.098703029916f,0.107066588617f,0.00523157681353f,0.109516758625f,0.11449012489f,0.00579292222525f,0.120833508104f,0.120982700612f,0.00638735196241f,0.132756772803f,0.126759943415f,0.0069730244716199995f,0.14502000078900001f,0.132114155782f,0.007512258382419999f,0.157034897655f,0.137058036252f,0.00806004196545f,0.168740910277f,0.141649687793f,0.0086020586774f,0.18018158636599999f,0.145897662326f,0.009187402070350001f,0.191306061277f,0.149911730845f,0.009699306680750001f,0.20228018041999998f,0.153699794586f,0.0101651632923f,0.21306910708000001f,0.157229049852f,0.0106966648307f,0.223525579263f,0.160604333941f,0.0111474503315f,0.23383850216999996f,0.163796762742f,0.0116231373033f,0.243863642246f,0.166848448357f,0.0121134415658f,0.253685702911f,0.169819141386f,0.0126114478884f,0.26363818667f,0.172738904452f,0.013045630466799999f,0.273838903033f,0.17553987372299998f,0.0136060166319f,0.284035834989f,0.178292710791f,0.014093826366700002f,0.294472726772f,0.180967301212f,0.0146031444225f,0.305018418317f,0.183568732274f,0.0151304937855f,0.315665343158f,0.18612659789500002f,0.0155638077972f,0.326540250824f,0.188597679232f,0.016112598419f,0.337367139559f,0.19101527950999997f,0.0166661392128f,0.348269803401f,0.19338747153000002f,0.0172191510076f,0.35923893975599996f,0.195737845385f,0.0176485273411f,0.370407572449f,0.198044113384f,0.018180747506500002f,0.381482230849f,0.200323209309f,0.018816828319f,0.392447152667f,0.20260874331f,0.0193080320772f,0.40358286922799996f,0.204899513261f,0.019767508720099997f,0.414734365044f,0.207203487335f,0.020318509756f,0.42573979336900003f,0.20954069120299998f,0.020829322049200003f,0.43673700905699997f,0.211923423866f,0.0214286969089f,0.447561641367f,0.214365276602f,0.0218414047174f,0.458511304496f,0.216884168073f,0.022476171082099997f,0.469109681584f,0.219485130374f,0.022911157438400002f,0.479815538029f,0.222188067595f,0.023425825940400002f,0.49030580992600004f,0.224998688279f,0.0238728708752f,0.500730803846f,0.227935385801f,0.0244003542471f,0.51092309047f,0.230997962124f,0.024855742978500003f,0.521037675039f,0.234203821386f,0.0253947921946f,0.530907891495f,0.237546372868f,0.0258598394405f,0.540693759978f,0.241027046685f,0.026247579899400002f,0.55039511939f,0.244584484796f,0.0267693440447f,0.55991166928f,0.2481175218f,0.0271274399767f,0.569636574265f,0.24989475832f,0.0278631362956f,0.581398671653f,0.251525497218f,0.0284219720302f,0.593483256725f,0.252639757507f,0.029041628899500002f,0.606152585473f,0.252696088643f,0.0296094624578f,0.620118665558f,0.251496133219f,0.0303909420016f,0.635313719985f,0.248691606409f,0.0311547780908f,0.652158617486f,0.243878137586f,0.0320817761944f,0.670731330201f,0.23639962754000002f,0.0330706616692f,0.691424807068f,0.225267244623f,0.0341252602766f,0.7146108666329999f,0.20888985034200003f,0.0353832411736f,0.7406363862800001f,0.18419045863800002f,0.0368327602548f,0.7701406520819999f,0.14386341454600002f,0.038418444306699995f,0.804039880295f,0.0746219638915f,0.0463798951671f,0.836091603811f,0.051929680457699995f,0.0798472986086f,0.829374357579f,0.0394515282547f,0.107189845036f,0.816977772194f,0.0384497825314f,0.130572940256f,0.8006196394949999f,0.0376049080062f,0.150475345512f,0.783946103531f,0.0366209160453f,0.167937309627f,0.767178191635f,0.036209539045199995f,0.183593242522f,0.750243007035f,0.035009786269f,0.197766483702f,0.733612552706f,0.0341816624553f,0.21076933108100002f,0.717059181752f,0.0338207279497f,0.222790052386f,0.700632749393f,0.0330602243752f,0.233954587651f,0.684623431271f,0.0320054471262f,0.244386794417f,0.669049419546f,0.0311985935446f,0.25418969331700003f,0.653810060935f,0.0306931539107f,0.263438905191f,0.6389273191579999f,0.0300923241428f,0.272196250421f,0.624527084061f,0.0294532825196f,0.280518738048f,0.610613760314f,0.0288266759948f,0.288454796768f,0.597189242503f,0.028258218845900004f,0.296045951278f,0.5842532441480001f,0.027358108648900002f,0.303331720628f,0.571895679741f,0.0270296599782f,0.31033700714699997f,0.559924872093f,0.0264448001694f,0.317097231559f,0.548516147786f,0.0256345140463f,0.323638967672f,0.537652138675f,0.025492765819700003f,0.329971350849f,0.527160032449f,0.024740170013099996f,0.3361344932f,0.517263775647f,0.0242644107155f,0.342132084344f,0.507792879869f,0.0240860234118f,0.347978814433f,0.49873824455f,0.023770524347099996f,0.35369797069f,0.49015319357500003f,0.0233150400472f,0.35930342735799997f,0.48201715982199994f,0.022709995372599996f,0.364807798254f,0.474309413085f,0.022415522700799998f,0.37021053997000003f,0.46695688834f,0.021951068805999997f,0.375533039074f,0.459997929716f,0.021785245042699998f,0.380771775449f,0.453366342385f,0.0214109796297f,0.38594788015500003f,0.447093322224f,0.021308362673299996f,0.391055299678f,0.44111961202399996f,0.0209409639731f,0.396115195541f,0.43546940626900005f,0.0208032233745f,0.40111931542199997f,0.43009041141800003f,0.020326797125099997f,0.406089062357f,0.424999913907f,0.0200232189201f,0.41101424043799994f,0.420152583348f,0.0198728054352f,0.415899382496f,0.415537874164f,0.0201730386418f,0.42075220537299995f,0.410977892936f,0.0204269128371f,0.425601519313f,0.40633093586999997f,0.020648268745f,0.430446943879f,0.40159664050300004f,0.0208515375655f,0.435288096472f,0.3967746856f,0.0210517249569f,0.4401245929f,0.391864797966f,0.0212643917258f,0.44495604798f,0.386866760343f,0.021505630995400003f,0.449782076175f,0.381780420593f,0.021792041613500002f,0.45460229226000004f,0.376605702372f,0.0221406975556f,0.459416312029f,0.37134261756399994f,0.0225691130702f,0.46422375303499996f,0.365991280754f,0.0224259562301f,0.469044405085f,0.36052987785499996f,0.0230580790144f,0.47383779193099995f,0.354998747537f,0.0231350670787f,0.47864410868699997f,0.349350101263f,0.0233439830722f,0.48344280665899997f,0.343606556264f,0.0237028430224f,0.488233513041f,0.337768885308f,0.0235081053068f,0.493037171153f,0.331794869685f,0.024209893148099997f,0.497811021508f,0.325767842034f,0.0243709990305f,0.502597537708f,0.319598395605f,0.024730286079f,0.507375060891f,0.313331518429f,0.0245356175547f,0.512165384947f,0.306908018078f,0.0253294179759f,0.516924115438f,0.300448540314f,0.025576676730900003f,0.521695388489f,0.293828618023f,0.0252539941421f,0.5264794575689999f,0.287037203286f,0.0259778918909f,0.53123082017f,0.280225448366f,0.0261331721134f,0.535994757369f,0.27324116767400003f,0.0265399835536f,0.540748306125f,0.26616608007f,0.0263475128744f,0.54551458806f,0.25890964136100003f,0.026400511586099996f,0.5502703191f,0.251565875201f,0.026708122111499998f,0.555015282958f,0.244143954883f,0.0272784546489f,0.5597492873650001f,0.23665484777600002f,0.0271978964134f,0.564496121687f,0.22898932204f,0.027362887814599997f,0.5692319597209999f,0.22127137955499998f,0.027777471233f,0.573956676974f,0.213518485566f,0.0284444308696f,0.578670178335f,0.205751413881f,0.028384328963700002f,0.583396856402f,0.19784223614f,0.0285476524053f,0.588112448966f,0.18995533481199997f,0.0289331144526f,0.5928169448489999f,0.182123550818f,0.0285115878732f,0.597535139873f,0.17420506477000003f,0.029317533151700003f,0.602217649212f,0.166595804546f,0.029277909690400002f,0.606914200514f,0.15897569628900002f,0.0294140830833f,0.611600149328f,0.151584202296f,0.029720091464999997f,0.616275666466f,0.144488804029f,0.0301894614322f,0.620940969293f,0.137765522615f,0.030815365661999997f,0.62559632569f,0.131498161336f,0.0304596782636f,0.630267391981f,0.12551028490900001f,0.0313631945057f,0.6349039514029999f,0.12041327253699999f,0.031242877616f,0.639557176769f,0.115752028275f,0.0312235432034f,0.644202264115f,0.11187543455f,0.031299156354f,0.648839821508f,0.108851437083f,0.0314184113993f,0.653477231267f,0.106176991919f,0.031535472753400004f,0.658121193765f,0.103322272239f,0.0325849440421f,0.662732380265f,0.10250629337000002f,0.0324964623759f,0.667373624657f,0.10131367318199999f,0.0322681389383f,0.672026774497f,0.09970463821270001f,0.0330969943958f,0.676670839949f,0.0976295890318f,0.0326757224886f,0.6813534245209999f,0.09429961297700001f,0.0337229991078f,0.686019956215f,0.09035971130699999f,0.033906420328900004f,0.6907178321999999f,0.0849030421592f,0.0334148100448f,0.695443242051f,0.07765277688510001f,0.0337425640662f,0.700167125211f,0.0686502617223f,0.0338207022037f,0.704910801625f,0.056740271177f,0.0338743964261f,0.709670103135f,0.0403918706963f,0.058733599314900006f,0.713922150001f,0.034390706820699994f,0.08742958138599999f,0.7178914983329999f,0.0346787868923f,0.111915239145f,0.7217992682150001f,0.0348193031422f,0.134123009678f,0.7256434216609999f,0.034803745147699995f,0.155347319743f,0.729397564257f,0.035371253705500004f,0.175139323574f,0.733108825242f,0.03502508946419999f,0.194611949127f,0.736726660802f,0.0352592690818f,0.213537378156f,0.740273862291f,0.0353178375278f,0.23233260534699998f,0.743724886122f,0.035968753389f,0.250513531308f,0.747126350622f,0.0356584622917f,0.268691545657f,0.750429081044f,0.035938694373699995f,0.286648031957f,0.753655779162f,0.0360251867582f,0.30462964428100003f,0.75678196096f,0.0367156189835f,0.32225527245200003f,0.75985331015f,0.0363983061994f,0.339942737704f,0.762822451337f,0.036684313670300006f,0.35751599238799997f,0.765711772711f,0.036760521877800004f,0.37499478500100003f,0.768520445955f,0.03661958726469999f,0.39253668911399997f,0.7712256456290001f,0.0370972727808f,0.40999430756399996f,0.773849358368f,0.0373587187295f,0.42737966438600006f,0.776391028603f,0.0373970537272f,0.44470245939499997f,0.778850167358f,0.0372056550069f,0.46207959141000005f,0.7812056023f,0.037651709852999996f,0.479394440292f,0.783478485675f,0.0378707411309f,0.496653275128f,0.785668578238f,0.037856699162400004f,0.513861188457f,0.787775695975f,0.0376038180367f,0.531105711117f,0.7897806400059999f,0.0380113165991f,0.548295506065f,0.7917033548639999f,0.038184710776999996f,0.565434007525f,0.793543877596f,0.0381188749731f,0.5825911959669999f,0.795284740401f,0.038737631302199996f,0.599630010222f,0.796961716746f,0.0381868400101f,0.6166820495040001f,0.798540479427f,0.038327533972699994f,0.6336821068330001f,0.800038845211f,0.038226918714199996f,0.650680153913f,0.801441969801f,0.038841776359800004f,0.667576491456f,0.802781137769f,0.0382551020787f,0.684464833385f,0.804027141476f,0.0383929993421f,0.7013338115350001f,0.80518254406f,0.0392768048004f,0.718108996552f,0.80627465814f,0.038941355259f,0.734858978841f,0.807278722198f,0.0393626753188f,0.7515248991399999f,0.808219025281f,0.0385431989517f,0.7681800140530001f,0.8090642018549999f,0.0395103035394f,0.784981358809f,0.8097796552509999f,0.0389863784715f,0.8022058529970001f,0.8102699963259999f,0.0386723126896f,0.8198700490620001f,0.810514774272f,0.039587106598699996f,0.837964828179f,0.81050820531f,0.0407349043605f,0.8564957604919999f,0.810239070669f,0.0411533487121f,0.875477677159f,0.809684424733f,0.041848599778699994f,0.8949159257420001f,0.808826947298f,0.0428263881801f,0.914815073355f,0.807648343923f,0.0440862680746f,0.935192099963f,0.8061279113519999f,0.0447502798991f,0.9555538065289999f,0.8043291392380001f,0.06738231609709999f,0.9643886421690001f,0.8047460502860001f,0.252819656976f,0.969264675898f,0.806479603483f,0.355213883146f,0.972679058372f,0.808946322667f,0.427168888808f,0.975344341181f,0.8119109170919999f,0.482670299842f,0.977488081448f,0.8152639825329999f,0.527978708904f,0.9793252066920001f,0.81891153223f,0.5660448050119999f,0.980856323558f,0.822821418552f,0.599002327948f,0.982248942f,0.8269254258490001f,0.627796037997f,0.9834291778550001f,0.8312222396439999f,0.653523321339f,0.984581461736f,0.835645803538f,0.676478782658f,0.9855303208130001f,0.8402290853319999f,0.6974790648420001f,0.986439873531f,0.844915604813f,0.716603137535f,0.987284696794f,0.8497017593199999f,0.734192410112f,0.9880974797740001f,0.8545693999539999f,0.750435011311f,0.988784936521f,0.859536817855f,0.765642621436f,0.989434072415f,0.8645728128650001f,0.779848654014f,0.990132548978f,0.8696463974869999f,0.793083975622f,0.99072699097f,0.874795481798f,0.8056380470569999f,0.9912757574009999f,0.879998726453f,0.817531592142f,0.991837792113f,0.885234998583f,0.8287840677949999f,0.992337137685f,0.890521929857f,0.839541460436f,0.992819103751f,0.895843059048f,0.8498184551350001f,0.993184915734f,0.901223132537f,0.859755146555f,0.9937648480589999f,0.906565576495f,0.869111246785f,0.99415826293f,0.911981196241f,0.8782758180429999f,0.994542722176f,0.9174171683239999f,0.887132207634f,0.994951529848f,0.922861877404f,0.895689414959f,0.995417956347f,0.928304043215f,0.9039557832330001f,0.995805993584f,0.9337798170880001f,0.912062161581f,0.996137505124f,0.939281131187f,0.920014523899f,0.996434662824f,0.944800045145f,0.927818426777f,0.9968989074510001f,0.950279714893f,0.935356647591f,0.997197218778f,0.955810300499f,0.9428787189710001f,0.997527495307f,0.961336154068f,0.950266270209f,0.9979112170119999f,0.9668504867319999f,0.9575230453650001f,0.998179199116f,0.9723977961929999f,0.964774153088f,0.9985368445209999f,0.977921894815f,0.9719005570520001f,0.998809261864f,0.983468565536f,0.979026563766f,0.999205022643f,0.9889818427570001f,0.986032578677f,0.999545232406f,0.994508167954f,0.993042540059f,1.0f,1.0f,1.0f
-
-    },
-    {
-        0.0f,0.0f,0.0f,0.018058465396400002f,0.000875333094297f,0.0158976099199f,0.0358446051389f,0.00174980699631f,0.0328047687723f,0.0522636499211f,0.0025563525718699997f,0.050049172046f,0.06595973062999999f,0.00316349374618f,0.0653730066735f,0.0775624917691f,0.00377981404209f,0.0790449485008f,0.08770618802069999f,0.00440417464933f,0.0916119341295f,0.0968151684079f,0.0049730454100600004f,0.103540340406f,0.104722118803f,0.00554119297516f,0.11599716230999998f,0.111370425372f,0.00620204611365f,0.129261639264f,0.116946513903f,0.006848121858689999f,0.14355298095799998f,0.12153532394899999f,0.007593629006810001f,0.15827138042199998f,0.12560895274599998f,0.00824751570213f,0.17294829718899998f,0.129116053052f,0.00897961657386f,0.187269898157f,0.132214226742f,0.0096321961349f,0.20148045969300002f,0.134903090655f,0.0103013238918f,0.215382661822f,0.137265159112f,0.0109393452309f,0.229013577036f,0.139353106158f,0.011562120307200002f,0.242296207746f,0.141228114177f,0.012182397830399998f,0.25515394106899997f,0.14295845935100002f,0.012810606763500001f,0.267511245419f,0.144632880203f,0.013374506779399999f,0.279430555591f,0.146311489693f,0.013856517071799999f,0.290986319638f,0.147985088662f,0.014435025153600001f,0.30226197245700004f,0.14970770896299998f,0.0149622508005f,0.313501070947f,0.15150937992f,0.015514977713799998f,0.324530460502f,0.15343108664400001f,0.0159967974703f,0.335465361707f,0.155506784775f,0.0164900613996f,0.346132556517f,0.15777230058f,0.0169915776714f,0.356506192632f,0.16025809158399998f,0.0174996875948f,0.366565290795f,0.162987579688f,0.0180145188299f,0.376294691501f,0.165965739373f,0.0184349864779f,0.38584077770400005f,0.16918579305999998f,0.018871069297799997f,0.39506689055f,0.172195422223f,0.0192910917595f,0.40464415821000005f,0.173403705169f,0.0198899561305f,0.41645113278800006f,0.173861044438f,0.0205279465565f,0.42913879139600003f,0.172365040039f,0.0212232568816f,0.44407754921099996f,0.168164143733f,0.022081702120099998f,0.461671910382f,0.15995214689200002f,0.0230256208157f,0.482744336407f,0.145360278175f,0.024227775932999997f,0.508035519531f,0.118833057349f,0.0257460880057f,0.538929643256f,0.059215889274900006f,0.0275724315861f,0.577825147355f,0.027433370317599997f,0.055696941466199995f,0.573236714621f,0.026610656155599998f,0.0822574864661f,0.552306416405f,0.025323468955f,0.10275920217599999f,0.531339176589f,0.024581636462599997f,0.11971852687699999f,0.510265658092f,0.023508081587900003f,0.134168958766f,0.489764126426f,0.022623365942599998f,0.14677633269200002f,0.46981996249799995f,0.0215547333709f,0.157937576357f,0.450747864098f,0.020699329830499998f,0.16795289040999997f,0.43248433008299997f,0.0198967605312f,0.177031635917f,0.41516848254f,0.0192303955284f,0.18533855557f,0.398809925251f,0.0182867657147f,0.193007654432f,0.383581078901f,0.017612817051200002f,0.20014041547699998f,0.36927508329000003f,0.017014848798499998f,0.206823590156f,0.355950858235f,0.01653121034f,0.213126962847f,0.343569825475f,0.0159456483298f,0.219116312109f,0.332154924132f,0.015511386092299998f,0.224833901477f,0.321585900677f,0.014978811725399999f,0.230330436099f,0.311871103985f,0.0145817257042f,0.23563231808600002f,0.302898649355f,0.0140334508886f,0.24078128980900002f,0.294664914215f,0.0138286995253f,0.24578010019000002f,0.287029297837f,0.013384075107200001f,0.250676352423f,0.280026567675f,0.0132072369246f,0.255463537867f,0.27353693590100003f,0.012963478530099999f,0.260172611311f,0.267551533796f,0.01275628057f,0.264812412048f,0.261996025246f,0.0130214917563f,0.269399802958f,0.2564471131f,0.013275541578999999f,0.27397345914099996f,0.250784410116f,0.0135425133387f,0.278532552634f,0.245007619091f,0.0135126185665f,0.283093979942f,0.23911193732900002f,0.0138754745934f,0.287621824293f,0.23310130526f,0.013980997516000002f,0.29215106711800004f,0.22696086059599999f,0.014190400945600002f,0.296663130996f,0.22069629667100002f,0.0145304863521f,0.301157183524f,0.21430915847800003f,0.0146578033742f,0.305651900786f,0.207766742968f,0.014953406351900001f,0.310127653915f,0.20109459745799999f,0.0150555796267f,0.314603774411f,0.194248861492f,0.0153591079009f,0.319060012372f,0.187269706758f,0.015481614838900001f,0.323516334236f,0.18010245883200002f,0.0158320447878f,0.327951947526f,0.17280379686f,0.0160049395278f,0.332387434875f,0.16530897460400001f,0.0164236283821f,0.336801542341f,0.157694092826f,0.016657244420399998f,0.341215451176f,0.149885785417f,0.0166866212706f,0.34562941453199997f,0.141882953872f,0.0169565382695f,0.350021626219f,0.133799378962f,0.016994284899799997f,0.354414106962f,0.125553279371f,0.0172595614171f,0.358784804599f,0.117294366612f,0.0177521102088f,0.36313367696500004f,0.109085926267f,0.017956123680099998f,0.36748358135799997f,0.10085354282400001f,0.0178318333763f,0.371835170462f,0.09265963234850001f,0.0184069150757f,0.376143131735f,0.084959044806f,0.0186127396493f,0.380453888309f,0.07755167092710001f,0.0184053173473f,0.384768452568f,0.0705745923215f,0.0188750949188f,0.389041323406f,0.0646649213733f,0.018887726719f,0.39332016766000005f,0.059585575389f,0.018989240106099998f,0.39758314905000003f,0.0557613443634f,0.019155370293599996f,0.401836286524f,0.0530002144934f,0.0195726975527f,0.406079284257f,0.0509259961276f,0.0200605678256f,0.41031532947899996f,0.0499257009659f,0.019933992650500002f,0.414589372335f,0.047897334140099994f,0.0201198564949f,0.41887075826700004f,0.04469494276930001f,0.0202744574697f,0.423173801671f,0.039760474672f,0.020678514617299998f,0.427489232063f,0.032996469891599996f,0.0209912092412f,0.43183050815499996f,0.024250178731400004f,0.0349723445527f,0.43571010284100004f,0.021003309972200002f,0.0568879927503f,0.439264307473f,0.0211778526846f,0.0767479553296f,0.442716711831f,0.0212184127479f,0.0957871138849f,0.446040085637f,0.021486782986400003f,0.11404484267100001f,0.44925275321f,0.021609207331500003f,0.1318104468f,0.45235073162199996f,0.0215750256604f,0.149470028439f,0.455308584136f,0.021761699914500002f,0.167053697064f,0.45812383476999996f,0.0221744111686f,0.184236193933f,0.460837074209f,0.0220219560819f,0.201440136403f,0.46340308318499995f,0.022086341869099996f,0.21865425261799998f,0.465820835436f,0.022373913789f,0.23574290226f,0.468109977517f,0.022477547036000002f,0.252842941703f,0.47024955625f,0.0228083262913f,0.269736892289f,0.472278349938f,0.0225244412619f,0.286742684767f,0.474138353353f,0.0228931346916f,0.303649814199f,0.475868629906f,0.0230684253847f,0.320467842222f,0.47746917857f,0.0230437190067f,0.33720349255099996f,0.47894019618600003f,0.0228130802442f,0.35397751849200004f,0.48025060826399996f,0.0232755329729f,0.37059809134000005f,0.481450683186f,0.0230858051611f,0.38717825679799994f,0.482510801383f,0.0231497853431f,0.40370139153499995f,0.483434884226f,0.023480468515f,0.42008733990599995f,0.48425111724799996f,0.023141666200700003f,0.436435341245f,0.484925361047f,0.023555276293700002f,0.452650469281f,0.485494834886f,0.0232942153785f,0.468883184798f,0.485908993767f,0.0236001754074f,0.48565148479099995f,0.48601265707900004f,0.023419886257299998f,0.503100421156f,0.485724148423f,0.024240176158599996f,0.521220887411f,0.48502097767499996f,0.0251173152935f,0.540032006724f,0.483860737672f,0.0260755646195f,0.559561329233f,0.482196106284f,0.0266793371253f,0.579802776807f,0.47997623333000006f,0.0278581744233f,0.600792297647f,0.477141258357f,0.028716149772399998f,0.622520503646f,0.47363135725500005f,0.0297062017882f,0.644983693461f,0.46938021635899996f,0.0308169231996f,0.668171263815f,0.464315602087f,0.0320261521594f,0.692064697472f,0.45835888974399996f,0.0332998807226f,0.716681934384f,0.45140229749899996f,0.0341928395725f,0.7418988995450001f,0.443392751729f,0.0354509848299f,0.767707726612f,0.434210395185f,0.0365931233555f,0.7939891186429999f,0.42378120354099996f,0.037902022134800005f,0.820656977454f,0.41200156363000007f,0.039248180042500004f,0.8476125158560001f,0.398759158615f,0.0404892124241f,0.8746742422210001f,0.38398939181999997f,0.04178030419f,0.9017077994730001f,0.3675822286f,0.042949926405800004f,0.928415174219f,0.349573677887f,0.044424104252f,0.954545481796f,0.329991381325f,0.0462442266644f,0.9573013443709999f,0.33609564364f,0.10421581297900001f,0.9591174926260001f,0.34322000200099995f,0.142599846881f,0.960545553503f,0.350720062574f,0.17223852748f,0.9616946896969999f,0.358466816251f,0.196508613818f,0.962634746497f,0.36637856625500004f,0.21695690863000003f,0.963502228203f,0.374312232539f,0.23436999222399998f,0.964204505811f,0.38235618254799997f,0.24948301355900002f,0.964827517636f,0.39042044189400005f,0.262649173343f,0.965339521585f,0.39852906948100003f,0.274255453153f,0.965913832211f,0.40649656131399997f,0.28484251263300003f,0.96660239999f,0.414004257475f,0.299477112022f,0.967424045438f,0.421156664498f,0.31595297826499996f,0.96824448919f,0.428060999844f,0.334462757437f,0.96925265516f,0.434535360032f,0.355085184136f,0.970255192862f,0.440768027519f,0.377386202244f,0.971400042438f,0.44666745096999994f,0.40071500981299996f,0.972494124682f,0.4524136941f,0.42479050884999997f,0.973682286844f,0.457914388253f,0.449200759171f,0.974414038449f,0.46361271767800005f,0.47401216261600004f,0.9747062520569999f,0.469501541549f,0.498966836373f,0.974924322329f,0.47531196257200004f,0.5237342872290001f,0.975203228437f,0.480956638403f,0.54818506882f,0.975442392691f,0.486524219209f,0.5723036278320001f,0.975773849859f,0.49192905913699997f,0.596023932317f,0.9759748034810001f,0.497346258966f,0.619361874151f,0.976292748071f,0.50260450503f,0.642262831649f,0.976501574349f,0.507878764176f,0.664729824313f,0.9767289113650001f,0.513085162851f,0.6867407604160001f,0.97698331835f,0.518226546562f,0.7082899559579999f,0.9772725006299999f,0.523305979649f,0.7293745397600001f,0.977484712652f,0.528412817876f,0.749962011074f,0.977744873326f,0.533464543877f,0.770070436051f,0.9779400111409999f,0.538551100225f,0.7896558164390001f,0.978193692765f,0.543590248049f,0.8087531397859999f,0.978391929879f,0.548672499326f,0.8273059777480001f,0.9786568878949999f,0.55371563806f,0.845365210221f,0.978873730796f,0.558810727197f,0.862861541059f,0.9791631966959999f,0.5638753437590001f,0.8798610494939999f,0.979409739754f,0.569001080649f,0.8962813946610001f,0.9796153949680001f,0.574192828239f,0.9121082288019999f,0.979898784777f,0.5793675187899999f,0.9274295732439999f,0.980143480993f,0.584617488548f,0.942143565762f,0.980428066947f,0.589851275613f,0.956547783518f,0.9806167300409999f,0.595083361726f,0.9710772006f,0.976468972257f,0.603553543446f,0.981004612526f,0.9648058479520001f,0.617572961021f,0.981714065719f,0.954176640447f,0.63062291544f,0.9823004555700001f,0.944661555225f,0.642731489199f,0.982953296597f,0.93603016065f,0.654130441515f,0.98346643582f,0.928360756143f,0.6648172405899999f,0.984025523979f,0.9214349788770001f,0.674983455547f,0.984397251418f,0.915422862482f,0.684546158619f,0.984911837263f,0.910111827943f,0.693674997183f,0.985320520096f,0.905571512771f,0.702343324126f,0.985812122819f,0.901601922737f,0.710699556338f,0.986111780451f,0.8983568718520001f,0.718653125358f,0.986567164996f,0.8956423445839999f,0.7263409954760001f,0.986888961516f,0.893521424939f,0.733727097324f,0.98726874984f,0.891890609952f,0.740883482457f,0.9875661752700001f,0.8908078534280001f,0.747773239593f,0.987979040431f,0.890095471452f,0.7545144814040001f,0.988182442835f,0.889886635315f,0.7610171457439999f,0.988552456896f,0.890008347904f,0.767394121872f,0.988753387001f,0.890588245246f,0.773556490152f,0.989172395609f,0.8914579467329999f,0.779612719008f,0.989464223106f,0.892598870881f,0.785570447984f,0.989646351742f,0.8940631843790001f,0.791389242524f,0.989928948566f,0.895831034216f,0.7970767481309999f,0.990336671576f,0.897743294753f,0.802733631489f,0.990505508391f,0.899923597083f,0.808271006036f,0.990843972112f,0.9022150723129999f,0.8137872556219999f,0.9909846661f,0.904742107148f,0.8191935117679999f,0.991342528157f,0.907348759393f,0.82458640736f,0.991548838469f,0.910053474568f,0.829929090757f,0.9918886243799999f,0.91215063232f,0.835494723044f,0.9921146238449999f,0.914160237339f,0.841082517336f,0.9923880661700001f,0.915831974426f,0.8467923602140001f,0.992574850985f,0.917215934105f,0.85258644636f,0.9928525746119999f,0.918343892019f,0.858451030257f,0.993234361466f,0.919146606622f,0.864444157258f,0.9933642488940001f,0.919795086974f,0.870464877501f,0.993647603398f,0.920331532515f,0.8764968618369999f,0.9941021320110001f,0.9207174933589999f,0.882585544639f,0.9943942612459999f,0.9210197452869999f,0.888704026867f,0.99456239769f,0.921345666753f,0.8947817520200001f,0.994968207377f,0.921656719823f,0.900877705629f,0.995139198777f,0.9220365833640001f,0.9069316285519999f,0.9954233926039999f,0.9224997720460001f,0.912939395805f,0.99582750669f,0.923064404149f,0.9189223523360001f,0.996076872691f,0.923761725923f,0.92485561751f,0.996340836063f,0.9246478367640001f,0.930744885063f,0.996384013549f,0.9257491717f,0.9365672216680001f,0.996378230633f,0.9270589345389999f,0.942312986593f,0.9964670712720001f,0.9286197534080001f,0.947979495601f,0.99655468174f,0.9304505616230001f,0.95355995137f,0.9966691854180001f,0.932620027359f,0.959041009932f,0.996745582305f,0.935159261824f,0.9644107087809999f,0.996829850665f,0.938030872519f,0.9696706896549999f,0.9970449960400001f,0.941454329103f,0.9747737451080001f,0.9972004007619999f,0.945484787756f,0.979697280271f,0.997379626347f,0.9502538230959999f,0.9844008257060001f,0.997607280329f,0.955994721365f,0.988815107679f,0.9978827357819999f,0.963064049972f,0.992831571263f,0.998213482077f,0.971958790258f,0.996291841212f,0.998647995176f,0.983626075978f,0.9988935025700001f,0.999215372248f,1.0f,1.0f,1.0f
-    },
-
-    {
-        0.301961f, 0.047059f, 0.090196f,
-        0.3244797567486321f, 0.04518239041438145f, 0.0826897493142621f,
-        0.3469985134972642f, 0.04330578082876291f, 0.07518349862852419f,
-        0.3695172702458963f, 0.04142917124314436f, 0.06767724794278629f,
-        0.39203602699452844f, 0.03955256165752581f, 0.060170997257048386f,
-        0.4153246758645797f, 0.04229508539323472f, 0.05420443073367733f,
-        0.43878176790024087f, 0.0460482201189405f, 0.048574728645118664f,
-        0.46223885993590197f, 0.049801354844646276f, 0.04294502655656f,
-        0.4856959519715631f, 0.053554489570352054f, 0.03731532446800133f,
-        0.5085516281292881f, 0.06392319895335406f, 0.03348987001325075f,
-        0.5310704364835226f, 0.07799745417475062f, 0.030675018968971443f,
-        0.5535892448377571f, 0.0920717093961472f, 0.027860167924692127f,
-        0.5761080531919917f, 0.10614596461754378f, 0.025045316880412817f,
-        0.5964619516652722f, 0.11978723786274958f, 0.022663447812324287f,
-        0.6142893416123747f, 0.13292320940271982f, 0.020786880449471398f,
-        0.6321167315594771f, 0.14605918094269005f, 0.01891031308661851f,
-        0.6499441215065795f, 0.15919515248266028f, 0.01703374572376562f,
-        0.6680360595281275f, 0.17259567209707607f, 0.014892630286467139f,
-        0.6868017331566564f, 0.1866699273184727f, 0.012077779242187815f,
-        0.7055674067851853f, 0.2007441825398693f, 0.009262928197908492f,
-        0.724333080413714f, 0.2148184377612659f, 0.006448077153629168f,
-        0.7423298406606399f, 0.2289888071553629f, 0.003825454454750602f,
-        0.7535892448377571f, 0.244001346058186f, 0.002887170773324161f,
-        0.7648486490148745f, 0.2590138849610091f, 0.0019488870918977193f,
-        0.7761080531919917f, 0.27402642386383214f, 0.0010106034104712786f,
-        0.787367457369109f, 0.2890389627666552f, 7.231972904483746e-05f,
-        0.8046886092128975f, 0.3101132493361495f, 0.0025978918571448154f,
-        0.82251599916f, 0.33169377400895766f, 0.0054127429014241436f,
-        0.8403433891071024f, 0.35327429868176585f, 0.008227593945703472f,
-        0.8581707790542048f, 0.374854823354574f, 0.0110424449899828f,
-        0.8697203985455797f, 0.39434275787547296f, 0.014554826084898522f,
-        0.8791032353598441f, 0.41310843150400184f, 0.01830796081060429f,
-        0.8884860721741085f, 0.43187410513253066f, 0.022061095536310055f,
-        0.8978689089883728f, 0.4506397787610594f, 0.0258142302620158f,
-        0.9040771689092902f, 0.4672890677940236f, 0.030096461136612775f,
-        0.907830303634996f, 0.4823016066968467f, 0.03478787954374499f,
-        0.9115834383607018f, 0.4973141455996698f, 0.03947929795087719f,
-        0.9153365730864075f, 0.5123266845024927f, 0.04417071635800937f,
-        0.9183683833178213f, 0.52697856115817f, 0.04994412150657952f,
-        0.9202449506806742f, 0.5410528163795666f, 0.05745039095799106f,
-        0.9221215180435272f, 0.5551270716009632f, 0.0649566604094026f,
-        0.9239980854063801f, 0.5692013268223598f, 0.07246292986081407f,
-        0.9260668811146336f, 0.5834678103891573f, 0.08054588434842794f,
-        0.928881732158913f, 0.5984803492919804f, 0.09086700484411882f,
-        0.9316965832031924f, 0.6134928881948034f, 0.1011881253398097f,
-        0.9345114342474716f, 0.6285054270976265f, 0.1115092458355005f,
-        0.9373024908480955f, 0.6435417604441052f, 0.12194933854946921f,
-        0.9391790582109483f, 0.6594925830283547f, 0.13696187745229227f,
-        0.9410556255738012f, 0.6754434056126042f, 0.1519744163551153f,
-        0.942932192936654f, 0.6913942281968535f, 0.16698695525793825f,
-        0.9448087602995069f, 0.7073450507811031f, 0.1819994941607614f,
-        0.9466853276623598f, 0.7217085849186791f, 0.19701203306358447f,
-        0.9485618950252127f, 0.7357828401400757f, 0.21202457196640753f,
-        0.9504384623880655f, 0.7498570953614723f, 0.2270371108692305f,
-        0.9523150297509184f, 0.763931350582869f, 0.24204964977205368f,
-        0.9548168074353629f, 0.7780056058042656f, 0.259563029961243f,
-        0.9576316584796423f, 0.7920798610256622f, 0.27832870358977185f,
-        0.9604465095239215f, 0.8061541162470587f, 0.29709437721830056f,
-        0.9632613605682009f, 0.8202283714684554f, 0.3158600508468295f,
-        0.9651626587727875f, 0.8333890738501594f, 0.33690960657459007f,
-        0.966100942454214f, 0.8455867617087031f, 0.3603666986102511f,
-        0.9670392261356404f, 0.8577844495672468f, 0.383823790645912f,
-        0.9679775098170669f, 0.8699821374257906f, 0.40728088268157325f,
-        0.9689157934984933f, 0.881314797730031f, 0.430161289681032f,
-        0.9698540771799198f, 0.8906976345442954f, 0.45174181435384014f,
-        0.9707923608613461f, 0.9000804713585597f, 0.47332233902664833f,
-        0.9717306445427726f, 0.9094633081728241f, 0.49490286369945613f,
-        0.972788836840555f, 0.9186063277543768f, 0.5162435711395528f,
-        0.9746654042034079f, 0.9261125972057883f, 0.5359475284495081f,
-        0.9765419715662608f, 0.9336188666571998f, 0.5556514857594633f,
-        0.9784185389291136f, 0.9411251361086114f, 0.5753554430694187f,
-        0.9802951062919665f, 0.9486314055600229f, 0.5950594003793739f,
-        0.9803921568627451f, 0.9543581582193601f, 0.6129838408972549f,
-        0.9803921568627451f, 0.9599878603079188f, 0.6308112308443573f,
-        0.9803921568627451f, 0.9656175623964773f, 0.6486386207914594f,
-        0.9803921568627451f, 0.9712472644850361f, 0.666466010738562f,
-        0.981834805851329f, 0.9754343175850109f, 0.6864573741685402f,
-        0.9837113732141819f, 0.9791874523107167f, 0.707099615159922f,
-        0.9855879405770348f, 0.9829405870364225f, 0.7277418561513037f,
-        0.9874645079398877f, 0.9866937217621282f, 0.7483840971426855f,
-        0.9876593664170776f, 0.9882352941176471f, 0.7712609376122769f,
-        0.9866819875822583f, 0.9882352941176471f, 0.7956954084827571f,
-        0.9857046087474391f, 0.9882352941176471f, 0.820129879353237f,
-        0.9847272299126199f, 0.9882352941176471f, 0.8445643502237176f,
-        0.9470546069594605f, 0.9470546069594605f, 0.886023613981023f,
-        0.932598204007831f, 0.932598204007831f, 0.8661460599225324f,
-        0.9181418010562015f, 0.9181418010562015f, 0.8462685058640418f,
-        0.9036853981045719f, 0.9036853981045719f, 0.8263909518055512f,
-        0.8892289951529424f, 0.8891685524480406f, 0.8067551685666682f,
-        0.8747725922013129f, 0.8738086243119342f, 0.790491715246085f,
-        0.8603161892496836f, 0.858448696175828f, 0.774228261925502f,
-        0.8458597862980538f, 0.8430887680397214f, 0.7579648086049184f,
-        0.8314033833464243f, 0.827728839903615f, 0.7417013552843352f,
-        0.8189164803943001f, 0.8136819117671789f, 0.7274074019632574f,
-        0.8071706529961011f, 0.8001290340000261f, 0.7138545241961046f,
-        0.7954248255979021f, 0.7865761562328735f, 0.700301646428952f,
-        0.7836789981997031f, 0.7730232784657207f, 0.6867487686617992f,
-        0.7719331708015041f, 0.7601684649191511f, 0.6738939551152296f,
-        0.7601873434033053f, 0.7484226375209523f, 0.6621481277170308f,
-        0.7484415160051061f, 0.7366768101227531f, 0.6504023003188316f,
-        0.736695688606907f, 0.7249309827245541f, 0.6386564729206325f,
-        0.7249914254294559f, 0.7131851553263551f, 0.6269522097431814f,
-        0.7141491232157339f, 0.7014393279281561f, 0.6161099075294594f,
-        0.7033068210020117f, 0.6896935005299571f, 0.6052676053157372f,
-        0.6924645187882896f, 0.6779476731317582f, 0.5944253031020151f,
-        0.6816222165745675f, 0.6662018457335592f, 0.583583000888293f,
-        0.6707799143608455f, 0.6538183968196791f, 0.572740698674571f,
-        0.6599376121471232f, 0.6411690442370032f, 0.5618983964608487f,
-        0.649095309933401f, 0.6285196916543273f, 0.5510560942471265f,
-        0.6382530077196789f, 0.6158703390716513f, 0.5402137920334044f,
-        0.627080551879819f, 0.6035511401151131f, 0.5293714898196822f,
-        0.6153347244816201f, 0.5918053127169142f, 0.51852918760596f,
-        0.603588897083421f, 0.5800594853187151f, 0.5076868853922378f,
-        0.591843069685222f, 0.5683136579205161f, 0.49684458317851565f,
-        0.5801426137602113f, 0.5565678305223173f, 0.4860249667013876f,
-        0.5702038367309658f, 0.5448220031241182f, 0.4760861896721422f,
-        0.5602650597017206f, 0.5330761757259193f, 0.466147412642897f,
-        0.5503262826724753f, 0.5213303483277203f, 0.4562086356136517f,
-        0.54038750564323f, 0.5095845209295213f, 0.4462698585844065f,
-        0.5310674716455119f, 0.49845743656284963f, 0.43756856761821566f,
-        0.5220322198007434f, 0.48761513434912745f, 0.429436840957924f,
-        0.512996967955975f, 0.4767728321354053f, 0.4213051142976324f,
-        0.5039617161112067f, 0.46593052992168327f, 0.4131733876373409f,
-        0.4952377394084217f, 0.4557107779919283f, 0.40504166097704913f,
-        0.4871060127481301f, 0.4466755261471598f, 0.39690993431675753f,
-        0.4789742860878385f, 0.43764027430239133f, 0.3887782076564659f,
-        0.4708425594275469f, 0.42860502245762283f, 0.3806464809961743f,
-        0.4627146400196953f, 0.4195773851177345f, 0.3725223688407628f,
-        0.4554864385438805f, 0.4123491836419197f, 0.36619769254942486f,
-        0.4482582370680657f, 0.4051209821661049f, 0.35987301625808693f,
-        0.44103003559225096f, 0.39789278069029016f, 0.353548339966749f,
-        0.4338018341164362f, 0.3906645792144754f, 0.3472236636754111f,
-        0.4271734971879949f, 0.38343637773866085f, 0.3408989873840733f,
-        0.4208488208966568f, 0.37620817626284586f, 0.3345743110927352f,
-        0.4145241446053189f, 0.36897997478703115f, 0.3282496348013973f,
-        0.408199468313981f, 0.3617517733112164f, 0.3219249585100594f,
-        0.4015823953648134f, 0.3542311751775719f, 0.3158926788765512f,
-        0.39435419388899856f, 0.34609944851728025f, 0.31047152776969006f,
-        0.38712599241318374f, 0.3379677218569886f, 0.30505037666282897f,
-        0.379897790937369f, 0.32983599519669693f, 0.2996292255559679f,
-        0.37266958946155415f, 0.3217042685364053f, 0.2942080744491068f,
-        0.3663298419385024f, 0.31534944978163965f, 0.28789846938948277f,
-        0.36000516564716445f, 0.3090247734903017f, 0.28157379309814484f,
-        0.3536804893558265f, 0.3027000971989638f, 0.2752491168068069f,
-        0.34735581306448865f, 0.2963754209076259f, 0.26892444051546904f,
-        0.3410311367731507f, 0.2894697585530686f, 0.2625997642241311f,
-        0.3347064604818128f, 0.2822415570772539f, 0.2562750879327932f,
-        0.3283817841904749f, 0.27501335560143914f, 0.24995041164145532f,
-        0.32205710789913716f, 0.2677851541256246f, 0.24362573535011758f,
-        0.3154589134341233f, 0.2605569526498096f, 0.2373010590587795f,
-        0.3082307119583085f, 0.2533287511739948f, 0.23097638276744153f,
-        0.3010025104824937f, 0.24610054969818002f, 0.22465170647610358f,
-        0.2937743090066789f, 0.2388723482223652f, 0.21832703018476562f,
-        0.2865461075308641f, 0.2316441467465504f, 0.21200235389342767f,
-        0.27931790605504936f, 0.22441594527073566f, 0.20567767760208977f,
-        0.2720897045792346f, 0.2171877437949209f, 0.19935300131075184f,
-        0.2648615031034198f, 0.20995954231910616f, 0.19302832501941394f,
-        0.2576333016276051f, 0.2027313408432914f, 0.18670364872807604f,
-        0.2515293153099212f, 0.19606524694654207f, 0.18150318759486897f,
-        0.24610816420306011f, 0.18974057065520414f, 0.17698556167248475f,
-        0.24068701309619905f, 0.18341589436386624f, 0.17246793575010053f,
-        0.23526586198933802f, 0.1770912180725283f, 0.1679503098277163f,
-        0.2293354315034333f, 0.1707665417811904f, 0.16292340452628845f,
-        0.2221072300276185f, 0.16444186548985246f, 0.1565987282349505f,
-        0.2148790285518039f, 0.1581171891985147f, 0.15027405194361274f,
-        0.2076508270759889f, 0.15179251290717655f, 0.1439493756522746f,
-        0.2004226256001741f, 0.14546783661583862f, 0.13762469936093666f,
-        0.18894751779574606f, 0.15615051291796544f, 0.16786193042772124f,
-        0.17720867752123598f, 0.16788934416252893f, 0.20036965238654675f,
-        0.16546983724672593f, 0.1796281754070924f, 0.2328773743453723f,
-        0.15373099697221587f, 0.19136700665165585f, 0.2653850963041978f,
-        0.15262772233867725f, 0.20024081909319424f, 0.28314885340588275f,
-        0.15865382088226443f, 0.20719411419736025f, 0.29102924077196585f,
-        0.1646799194258516f, 0.21414740930152623f, 0.29890962813804894f,
-        0.1707060179694388f, 0.22110070440569224f, 0.30679001550413204f,
-        0.17673211651302598f, 0.22805399950985822f, 0.3146704028702152f,
-        0.18275821505661316f, 0.2350072946140242f, 0.3225507902362983f,
-        0.18878431360020034f, 0.2419605897181902f, 0.3304311776023814f,
-        0.19481041214378753f, 0.2489138848223562f, 0.3383115649684645f,
-        0.20077216965118935f, 0.25593152031923283f, 0.3462562936924703f,
-        0.20633476494215652f, 0.2633483140405224f, 0.35460018662892107f,
-        0.21189736023312403f, 0.2707651077618124f, 0.3629440795653723f,
-        0.21745995552409134f, 0.2781819014831022f, 0.37128797250182327f,
-        0.22302255081505867f, 0.2855986952043919f, 0.37963186543827426f,
-        0.228585146106026f, 0.2930154889256817f, 0.38797575837472525f,
-        0.23414774139699335f, 0.30043228264697147f, 0.39631965131117625f,
-        0.23971033668796066f, 0.30784907636826125f, 0.4046635442476273f,
-        0.245272931978928f, 0.31526587008955104f, 0.4130074371840783f,
-        0.2517797032050162f, 0.32331211443425467f, 0.4219807807439432f,
-        0.2587329473187253f, 0.33165600737070566f, 0.43125177289555533f,
-        0.2656861914324345f, 0.33999990030715666f, 0.44052276504716753f,
-        0.2726394355461436f, 0.3483437932436076f, 0.44979375719877973f,
-        0.27959267965985274f, 0.3566876861800586f, 0.4590647493503919f,
-        0.2865459237735619f, 0.3650315791165095f, 0.4683357415020041f,
-        0.29349916788727104f, 0.3733754720529605f, 0.4776067336536163f,
-        0.30045241200098016f, 0.3817193649894115f, 0.4868777258052285f,
-        0.3076087632044908f, 0.3900632579258622f, 0.4960471644119395f,
-        0.3154891065333614f, 0.39840715086231343f, 0.5048546069559714f,
-        0.32336944986223176f, 0.4067510437987644f, 0.5136620495000029f,
-        0.33124979319110215f, 0.4150949367352154f, 0.5224694920440346f,
-        0.33913013651997254f, 0.4234388296716664f, 0.5312769345880661f,
-        0.3470104798488429f, 0.4317827226081174f, 0.5400843771320978f,
-        0.3548908231777133f, 0.4401266155445684f, 0.5488918196761294f,
-        0.3627711665065837f, 0.4484705084810194f, 0.557699262220161f,
-        0.3706515098354541f, 0.4568144014174704f, 0.5665067047641925f,
-        0.3788837845499998f, 0.4655102257395967f, 0.5753141473082242f,
-        0.38722767748645076f, 0.4743176682836283f, 0.5841215898522557f,
-        0.3955715704229017f, 0.48312511082765985f, 0.5929290323962874f,
-        0.4039154633593527f, 0.49193255337169145f, 0.6017364749403189f,
-        0.4122593562958036f, 0.500739995915723f, 0.6105439174843506f,
-        0.4206032492322546f, 0.5095474384597546f, 0.6193513600283821f,
-        0.42894714216870555f, 0.5183548810037861f, 0.6281588025724136f,
-        0.4372910351051563f, 0.5271623235478174f, 0.636966245116445f,
-        0.44577368766047676f, 0.5362472853295879f, 0.6460512068982154f,
-        0.4545811302045084f, 0.5459818270887807f, 0.6557857486574082f,
-        0.46338857274854006f, 0.5557163688479736f, 0.6655202904166011f,
-        0.4721960152925717f, 0.5654509106071665f, 0.675254832175794f,
-        0.48100345783660337f, 0.5751854523663593f, 0.6849893739349868f,
-        0.489810900380635f, 0.5849199941255522f, 0.6947239156941796f,
-        0.4986183429246666f, 0.594654535884745f, 0.7044584574533724f,
-        0.5074257854686983f, 0.6043890776439379f, 0.7141929992125653f,
-        0.51623322801273f, 0.6141236194031308f, 0.7239275409717582f,
-        0.525818945476049f, 0.6238581611623235f, 0.7328838078116634f,
-        0.5355534872352418f, 0.6335927029215163f, 0.7416912503556949f,
-        0.5452880289944345f, 0.643327244680709f, 0.7504986928997265f,
-        0.5550225707536273f, 0.6530617864399018f, 0.759306135443758f,
-        0.56475711251282f, 0.6627963281990945f, 0.7681135779877896f,
-        0.5744916542720128f, 0.6725308699582873f, 0.7769210205318211f,
-        0.5842261960312053f, 0.6822654117174798f, 0.7857284630758524f,
-        0.5939607377903984f, 0.6919999534766728f, 0.7945359056198842f,
-        0.6038712452424287f, 0.7019104609287032f, 0.8031673824710781f,
-        0.6140693366092023f, 0.7121085522954768f, 0.8115112754075292f,
-        0.6242674279759757f, 0.7223066436622502f, 0.8198551683439802f,
-        0.6344655193427492f, 0.7325047350290237f, 0.8281990612804312f,
-        0.6446636107095227f, 0.7427028263957972f, 0.8365429542168822f,
-        0.6548617020762961f, 0.7529009177625706f, 0.8448868471533333f,
-        0.6650597934430696f, 0.7630990091293441f, 0.8532307400897843f,
-        0.6752578848098431f, 0.7732971004961176f, 0.8615746330262353f,
-        0.6854559761766166f, 0.7834951918628911f, 0.8699185259626863f,
-        0.6943750369425534f, 0.7924142526288279f, 0.8769833882983007f,
-        0.703182479486585f, 0.8012216951728595f, 0.8839366324120098f,
-        0.7119899220306165f, 0.810029137716891f, 0.8908898765257189f,
-        0.720797364574648f, 0.8188365802609225f, 0.8978431206394281f,
-        0.7296048071186797f, 0.8276440228049542f, 0.9047963647531372f,
-        0.738412249662711f, 0.8364514653489855f, 0.9117496088668462f,
-        0.7472196922067428f, 0.8452589078930173f, 0.9187028529805555f,
-        0.7560271347507743f, 0.8540663504370488f, 0.9256560970942647f,
-        0.7659004361288364f, 0.8635133082814987f, 0.9328225129747799f,
-        0.777025626710771f, 0.8737113996482722f, 0.9402393066960696f,
-        0.7881508172927056f, 0.8839094910150456f, 0.9476561004173594f,
-        0.7992760078746403f, 0.894107582381819f, 0.9550728941386492f,
-        0.8123197443578297f, 0.9049451890490107f, 0.9620633443263267f,
-        0.8317888278762154f, 0.9179245780612678f, 0.967625939617294f,
-        0.851257911394601f, 0.930903967073525f, 0.9731885349082614f,
-        0.8707269949129868f, 0.9438833560857821f, 0.9787511301992287f,
-        0.8901960784313725f, 0.9568627450980393f, 0.984313725490196f
-    }, 
-    {
-        0.831373f, 0.909804f, 0.980392f,
-        0.8055367647058824f, 0.8950403921568627f, 0.9742404705882353f,
-        0.7797005294117647f, 0.8802767843137255f, 0.9680889411764706f,
-        0.7538642941176471f, 0.8655131764705882f, 0.9619374117647058f,
-        0.7350253725490197f, 0.8537484901960785f, 0.9557858823529412f,
-        0.7178012156862745f, 0.8426758627450981f, 0.9496343529411765f,
-        0.7005770588235294f, 0.8316032352941177f, 0.9434828235294118f,
-        0.6818149607843137f, 0.8209150196078432f, 0.9392537450980393f,
-        0.6621300980392157f, 0.8104574509803922f, 0.936178137254902f,
-        0.6424452352941177f, 0.7999998823529412f, 0.9331025294117647f,
-        0.6227603725490196f, 0.7895423137254902f, 0.9300269215686274f,
-        0.6030755098039216f, 0.7790847450980393f, 0.9269513137254902f,
-        0.5833906470588235f, 0.7686271764705883f, 0.9238757058823529f,
-        0.5640133921568627f, 0.7586309803921569f, 0.9208000588235293f,
-        0.5455589607843138f, 0.7500189019607844f, 0.917724294117647f,
-        0.5271045294117647f, 0.7414068235294118f, 0.9146485294117647f,
-        0.5086500980392157f, 0.7327947450980392f, 0.9115727647058823f,
-        0.4901956666666667f, 0.7241826666666666f, 0.908497f,
-        0.47174123529411766f, 0.7155705882352942f, 0.9054212352941177f,
-        0.4532868039215686f, 0.7069585098039216f, 0.9023454705882353f,
-        0.4364471764705882f, 0.6972699411764706f, 0.8960402352941177f,
-        0.4198382352941176f, 0.6874275882352942f, 0.8892736470588235f,
-        0.4032292941176471f, 0.6775852352941176f, 0.8825070588235294f,
-        0.38662035294117647f, 0.6677428823529412f, 0.8757404705882353f,
-        0.37001141176470587f, 0.6579005294117647f, 0.8689738823529412f,
-        0.3534024705882353f, 0.6480581764705882f, 0.8622072941176471f,
-        0.3374086666666667f, 0.6376006862745097f, 0.8557482352941177f,
-        0.32203000000000004f, 0.6265280588235294f, 0.8495967058823529f,
-        0.30665133333333333f, 0.615455431372549f, 0.8434451764705883f,
-        0.2912726666666667f, 0.6043828039215686f, 0.8372936470588236f,
-        0.275894f, 0.5933101764705881f, 0.8311421176470588f,
-        0.2605153333333333f, 0.5822375490196078f, 0.8249905882352941f,
-        0.24575180392156862f, 0.5715493921568627f, 0.8188390588235294f,
-        0.23529423529411772f, 0.5635525294117647f, 0.8126875294117647f,
-        0.22483666666666666f, 0.5555556666666667f, 0.806536f,
-        0.21437909803921568f, 0.5475588039215686f, 0.8003844705882353f,
-        0.2039215294117647f, 0.5395619411764705f, 0.7942329411764706f,
-        0.1934639607843138f, 0.5315650784313726f, 0.7880814117647059f,
-        0.18300639215686273f, 0.5235682156862744f, 0.7819298823529411f,
-        0.1711649411764706f, 0.5123416470588236f, 0.7711649411764706f,
-        0.1588621960784314f, 0.5000385098039216f, 0.7588621960784314f,
-        0.14825105882352946f, 0.4885043529411766f, 0.747482117647059f,
-        0.14440635294117649f, 0.48004611764705885f, 0.739792705882353f,
-        0.14056164705882354f, 0.4715878823529412f, 0.7321032941176471f,
-        0.1367169411764706f, 0.46312964705882353f, 0.7244138823529411f,
-        0.13287223529411768f, 0.45467141176470593f, 0.7167244705882353f,
-        0.1288737450980392f, 0.44667452941176466f, 0.7094195294117647f,
-        0.12349119607843137f, 0.44282982352941175f, 0.7055748235294118f,
-        0.11810864705882353f, 0.43898511764705883f, 0.7017301176470588f,
-        0.11272609803921571f, 0.4351404117647059f, 0.697885411764706f,
-        0.10734354901960784f, 0.43129570588235294f, 0.694040705882353f,
-        0.101961f, 0.427451f, 0.690196f,
-        0.09888531372549018f, 0.4182239411764706f, 0.6778932549019607f,
-        0.0958096274509804f, 0.40899688235294124f, 0.6655905098039216f,
-        0.09342588235294116f, 0.4018458235294117f, 0.6553637647058823f,
-        0.0918878431372549f, 0.39723209803921566f, 0.6476743529411765f,
-        0.09034980392156863f, 0.3926183725490196f, 0.6399849411764706f,
-        0.0874278823529412f, 0.38385229411764715f, 0.6322955294117648f,
-        0.08435219607843138f, 0.3746248431372549f, 0.6246061176470589f,
-        0.08073819607843138f, 0.36647401960784315f, 0.616916705882353f,
-        0.0761244705882353f, 0.36032264705882355f, 0.6092272941176471f,
-        0.07151074509803923f, 0.35417127450980396f, 0.6015378823529411f,
-        0.06935788235294117f, 0.34555935294117646f, 0.5901574901960783f,
-        0.06782023529411765f, 0.33633229411764703f, 0.5778543529411765f,
-        0.06743592156862746f, 0.3278740588235294f, 0.5659357254901961f,
-        0.07051160784313726f, 0.321722294117647f, 0.5551706274509803f,
-        0.07358729411764703f, 0.3155705294117648f, 0.5444055294117648f,
-        0.07773960784313726f, 0.31157201960784314f, 0.5325640784313725f,
-        0.08235333333333332f, 0.3084963333333333f, 0.5202613333333335f,
-        0.08673635294117645f, 0.3047286470588236f, 0.5079585294117649f,
-        0.08981203921568626f, 0.29703962745098045f, 0.49565539215686283f,
-        0.09288772549019605f, 0.2893506078431373f, 0.4833522549019609f,
-        0.09596341176470587f, 0.2844293529411765f, 0.4719719411764706f,
-        0.09903909803921569f, 0.2813532745098039f, 0.46120684313725485f,
-        0.10211478431372548f, 0.2782003333333333f, 0.4502110784313726f,
-        0.10519047058823532f, 0.27358699999999997f, 0.4348326470588234f,
-        0.1082661568627451f, 0.26897366666666667f, 0.41945421568627445f,
-        0.11057282352941177f, 0.2651291568627451f, 0.405613431372549f,
-        0.1121104705882353f, 0.2620534705882353f, 0.39331029411764706f,
-        0.11364811764705883f, 0.2589777843137255f, 0.38100715686274506f,
-        0.113725f, 0.2559017254901961f, 0.3687043921568628f,
-        0.113725f, 0.25282564705882354f, 0.3564016470588236f,
-        0.11310994117647062f, 0.2534407058823528f, 0.344713960784314f,
-        0.11157229411764708f, 0.25959247058823526f, 0.3339488627450981f,
-        0.11003464705882354f, 0.2657442352941176f, 0.3231837647058824f,
-        0.10849666666666667f, 0.2784316666666666f, 0.3019610000000001f,
-        0.1069586274509804f, 0.29227245098039206f, 0.2788927647058824f,
-        0.10449800000000001f, 0.3028837058823529f, 0.26043817647058826f,
-        0.09988466666666668f, 0.30595939215686274f, 0.25274876470588237f,
-        0.09527133333333333f, 0.30903507843137257f, 0.24505935294117648f,
-        0.090657705882353f, 0.313264294117647f, 0.23852347058823536f,
-        0.08604398039215685f, 0.3178780196078432f, 0.23237209803921566f,
-        0.08173786274509802f, 0.32310680392156865f, 0.226835862745098f,
-        0.07866217647058822f, 0.3307958235294118f, 0.2237601764705882f,
-        0.07558649019607842f, 0.3384848431372549f, 0.22068449019607841f,
-        0.07351027450980392f, 0.34617411764705885f, 0.21760880392156862f,
-        0.07197223529411766f, 0.3538635294117647f, 0.21453311764705885f,
-        0.07043423529411766f, 0.3615529411764705f, 0.21191876470588236f,
-        0.06889658823529415f, 0.36924235294117635f, 0.21345641176470587f,
-        0.0673589411764706f, 0.37693176470588236f, 0.2149940588235294f,
-        0.06582107843137255f, 0.38462117647058824f, 0.20722786274509805f,
-        0.06428303921568627f, 0.39231058823529413f, 0.19184943137254903f,
-        0.062745f, 0.4f, 0.176471f,
-        0.06459049019607843f, 0.40307576470588236f, 0.17154974509803922f,
-        0.06643598039215687f, 0.4061515294117647f, 0.16662849019607842f,
-        0.06828147058823529f, 0.4092272941176471f, 0.16170723529411765f,
-        0.0701269607843137f, 0.41230305882352936f, 0.15678598039215694f,
-        0.07197245098039216f, 0.41537882352941174f, 0.15186472549019608f,
-        0.0738179411764706f, 0.4184545882352941f, 0.1469434705882353f,
-        0.07566343137254902f, 0.42153035294117647f, 0.1424067254901961f,
-        0.07750892156862746f, 0.4246061176470588f, 0.1381006862745098f,
-        0.07935441176470588f, 0.4276818823529412f, 0.13379464705882355f,
-        0.08119990196078432f, 0.4307576470588235f, 0.12948860784313726f,
-        0.08304539215686275f, 0.43383341176470586f, 0.12518256862745097f,
-        0.08489088235294115f, 0.4369091764705882f, 0.12087652941176477f,
-        0.08765907843137255f, 0.4404462745098039f, 0.117647f,
-        0.09319539215686275f, 0.4453673725490196f, 0.117647f,
-        0.09873170588235294f, 0.4502884705882353f, 0.117647f,
-        0.10426801960784314f, 0.455209568627451f, 0.117647f,
-        0.10980433333333334f, 0.4601306666666667f, 0.117647f,
-        0.11534064705882352f, 0.46505176470588233f, 0.117647f,
-        0.12087696078431373f, 0.46997286274509803f, 0.117647f,
-        0.13018105882352926f, 0.4748940980392156f, 0.12195309803921561f,
-        0.1400234117647059f, 0.4798153529411765f, 0.12687435294117647f,
-        0.14986576470588236f, 0.48473660784313727f, 0.13179560784313726f,
-        0.15970811764705883f, 0.48965786274509804f, 0.13671686274509803f,
-        0.1695504705882353f, 0.4945791176470588f, 0.14163811764705883f,
-        0.17939282352941177f, 0.4995003725490196f, 0.14655937254901963f,
-        0.18985031372549022f, 0.5050366862745098f, 0.1520956862745098f,
-        0.20092294117647058f, 0.5111880588235294f, 0.15824705882352944f,
-        0.211995568627451f, 0.517339431372549f, 0.16439843137254903f,
-        0.2230681960784314f, 0.5234908039215687f, 0.17054980392156865f,
-        0.23414082352941146f, 0.5296421764705881f, 0.17670117647058808f,
-        0.24521345098039218f, 0.5357935490196079f, 0.18285254901960785f,
-        0.2562860784313726f, 0.5419449411764706f, 0.18908084313725493f,
-        0.26735870588235294f, 0.5480964705882353f, 0.19584758823529413f,
-        0.27843133333333336f, 0.554248f, 0.20261433333333334f,
-        0.28950396078431373f, 0.5603995294117647f, 0.20938107843137255f,
-        0.30057658823529415f, 0.5665510588235294f, 0.21614782352941175f,
-        0.3116492156862745f, 0.5727025882352941f, 0.222914568627451f,
-        0.32272184313725494f, 0.5788541176470589f, 0.2296813137254902f,
-        0.33471729411764706f, 0.5850056470588235f, 0.23690923529411764f,
-        0.3470203529411765f, 0.5911571764705883f, 0.24429088235294116f,
-        0.3593234117647059f, 0.597308705882353f, 0.2516725294117647f,
-        0.3716264705882353f, 0.6034602352941176f, 0.2590541764705882f,
-        0.3839295294117647f, 0.6096117647058824f, 0.26643582352941175f,
-        0.3962325882352941f, 0.6157632941176471f, 0.2738174705882353f,
-        0.409689f, 0.6226068235294118f, 0.28166052941176467f,
-        0.42506766666666623f, 0.6306036862745096f, 0.290272607843137f,
-        0.44044633333333333f, 0.6386005490196078f, 0.2988846862745098f,
-        0.455825f, 0.6465974117647059f, 0.30749676470588233f,
-        0.4712036666666667f, 0.6545942745098039f, 0.3161088431372549f,
-        0.48658233333333334f, 0.6625911372549019f, 0.3247209215686274f,
-        0.501961f, 0.670588f, 0.333333f,
-        0.5137513267973856f, 0.6782774117647058f, 0.3420476405228758f,
-        0.5255416535947712f, 0.6859668235294117f, 0.35076228104575163f,
-        0.5373319803921568f, 0.6936562352941176f, 0.3594769215686274f,
-        0.5491223071895425f, 0.7013456470588234f, 0.3681915620915033f,
-        0.5609126339869281f, 0.7090350588235294f, 0.37690620261437907f,
-        0.5727029607843137f, 0.7167244705882352f, 0.38562084313725487f,
-        0.5844932875816993f, 0.7244138823529411f, 0.3943354836601307f,
-        0.602383794117647f, 0.7334488725490195f, 0.4061899215686274f,
-        0.6316032058823529f, 0.744982794117647f, 0.42387541176470583f,
-        0.6608226176470587f, 0.7565167156862744f, 0.4415609019607843f,
-        0.6900420294117637f, 0.7680506372549016f, 0.45924639215686225f,
-        0.7192614411764705f, 0.7795845588235294f, 0.47693188235294115f,
-        0.7469430588235293f, 0.7916952450980392f, 0.49577080392156864f,
-        0.7700112941176469f, 0.805536225490196f, 0.518070019607843f,
-        0.7930795294117645f, 0.8193772058823529f, 0.5403692352941174f,
-        0.816147764705882f, 0.8332181862745096f, 0.5626684509803919f,
-        0.8392159999999995f, 0.8470591666666665f, 0.5849676666666662f,
-        0.8600542941176469f, 0.8583626274509805f, 0.6034606274509804f,
-        0.8682562549019607f, 0.8552868104575164f, 0.6003848104575163f,
-        0.8764582156862744f, 0.8522109934640524f, 0.5973089934640523f,
-        0.8846601764705881f, 0.8491351764705883f, 0.5942331764705883f,
-        0.8928621372549018f, 0.8460593594771242f, 0.5911573594771242f,
-        0.9010640980392156f, 0.8429835424836601f, 0.5880815424836602f,
-        0.9092660588235294f, 0.8399077254901961f, 0.5850057254901961f,
-        0.9174680196078431f, 0.8368319084967321f, 0.5819299084967321f,
-        0.9208000588235293f, 0.8311417647058823f, 0.577008705882353f,
-        0.9192621764705883f, 0.8228372941176473f, 0.570242117647059f,
-        0.917724294117647f, 0.8145328235294117f, 0.5634755294117647f,
-        0.9161864117647058f, 0.8062283529411765f, 0.5567089411764706f,
-        0.9146485294117647f, 0.7979238823529411f, 0.5499423529411764f,
-        0.9131106470588235f, 0.7896194117647058f, 0.5431757647058824f,
-        0.9115727647058823f, 0.7813149411764706f, 0.5364091764705883f,
-        0.9100348823529412f, 0.7730104705882352f, 0.5296425882352941f,
-        0.908497f, 0.764706f, 0.522876f,
-        0.9069591176470588f, 0.7564015294117646f, 0.5161094117647059f,
-        0.9054212352941177f, 0.7480970588235294f, 0.5093428235294117f,
-        0.9038833529411765f, 0.739792588235294f, 0.5025762352941177f,
-        0.9023454705882353f, 0.7314881176470588f, 0.4958096470588235f,
-        0.8994235294117647f, 0.7208768235294117f, 0.4878895882352941f,
-        0.8960402352941177f, 0.709496588235294f, 0.4795850392156863f,
-        0.8926569411764707f, 0.6981163529411765f, 0.4712804901960784f,
-        0.8892736470588235f, 0.6867361176470588f, 0.46297594117647056f,
-        0.8858903529411766f, 0.6753558823529414f, 0.454671392156863f,
-        0.8825070588235294f, 0.6639756470588235f, 0.4463668431372549f,
-        0.8791237647058824f, 0.6525954117647058f, 0.4380622941176471f,
-        0.8757404705882353f, 0.6412151764705882f, 0.4297577450980392f,
-        0.8723571764705883f, 0.6298349411764705f, 0.42145319607843135f,
-        0.8689738823529412f, 0.618454705882353f, 0.41314864705882354f,
-        0.8655905882352941f, 0.6070744705882353f, 0.40484409803921567f,
-        0.8622072941176471f, 0.5956942352941177f, 0.39653954901960786f,
-        0.858824f, 0.584314f, 0.388235f,
-        0.8542103529411765f, 0.5729337647058823f, 0.38300629411764703f,
-        0.8495967058823529f, 0.5615535294117647f, 0.3777775882352941f,
-        0.8449830588235294f, 0.550173294117647f, 0.37254888235294115f,
-        0.8403694117647059f, 0.5387930588235295f, 0.36732017647058823f,
-        0.8357557647058824f, 0.5274128235294118f, 0.36209147058823526f,
-        0.8311421176470588f, 0.5160325882352941f, 0.35686276470588235f,
-        0.8265284705882353f, 0.5046523529411765f, 0.3516340588235294f,
-        0.8219148235294119f, 0.49327211764705914f, 0.34640535294117664f,
-        0.8173011764705883f, 0.48189188235294117f, 0.3411766470588235f,
-        0.8126875294117647f, 0.47051164705882353f, 0.3359479411764706f,
-        0.8080738823529412f, 0.45913141176470584f, 0.3307192352941176f,
-        0.8034602352941177f, 0.44775117647058826f, 0.3254905294117647f,
-        0.7976162941176471f, 0.43644782352941175f, 0.31926219607843137f,
-        0.7880814705882353f, 0.4253751176470588f, 0.3100349803921569f,
-        0.7785466470588236f, 0.4143024117647059f, 0.30080776470588233f,
-        0.7690118235294118f, 0.40322970588235296f, 0.29158054901960784f,
-        0.7594770000000001f, 0.392157f, 0.28235333333333335f,
-        0.7499421764705883f, 0.38108429411764705f, 0.2731261176470588f,
-        0.7404073529411765f, 0.3700115882352941f, 0.2638989019607843f,
-        0.7308725294117647f, 0.3589388823529412f, 0.2546716862745098f,
-        0.7213377058823529f, 0.34786617647058826f, 0.24544447058823526f,
-        0.7118028823529412f, 0.3367934705882353f, 0.23621725490196077f,
-        0.7022680588235294f, 0.32572076470588235f, 0.22699003921568628f,
-        0.6927332352941179f, 0.31464805882352975f, 0.217762823529412f,
-        0.6831984117647059f, 0.30357535294117644f, 0.20853560784313724f,
-        0.6733560588235294f, 0.29296401960784313f, 0.20023109803921568f,
-        0.6632061764705882f, 0.2828140588235294f, 0.19284929411764706f,
-        0.6530562941176471f, 0.2726640980392157f, 0.18546749019607842f,
-        0.6429064117647059f, 0.26251413725490197f, 0.1780856862745098f,
-        0.6327565294117647f, 0.25236417647058823f, 0.17070388235294118f,
-        0.6226066470588235f, 0.2422142156862745f, 0.16332207843137253f,
-        0.6124567647058823f, 0.23206425490196078f, 0.1559402745098039f,
-        0.6023068823529412f, 0.22191429411764707f, 0.1485584705882353f,
-        0.5921569999999999f, 0.21176433333333333f, 0.14117666666666667f,
-        0.5820071176470588f, 0.2016143725490196f, 0.13379486274509805f,
-        0.5718572352941176f, 0.19146441176470588f, 0.12641305882352943f,
-        0.5617073529411765f, 0.18131445098039217f, 0.11903125490196079f,
-        0.5515574705882352f, 0.17116449019607843f, 0.11164945098039215f,
-        0.5407154117647058f, 0.15847711764705882f, 0.10242223529411765f,
-        0.5296426274509807f, 0.14494394117647097f, 0.09257988235294146f,
-        0.5185698431372548f, 0.13141076470588237f, 0.08273752941176471f,
-        0.5074970588235294f, 0.11787758823529412f, 0.07289517647058824f,
-        0.4964242745098039f, 0.10434441176470588f, 0.06305282352941177f,
-        0.48535149019607843f, 0.09081123529411764f, 0.0532104705882353f,
-        0.47681623529411765f, 0.0779700588235294f, 0.044290823529411764f,
-        0.47251019607843137f, 0.0662822156862745f, 0.03690901960784314f,
-        0.4682041568627451f, 0.05459437254901961f, 0.02952721568627451f,
-        0.4638981176470588f, 0.042906529411764704f, 0.02214541176470588f,
-        0.45959207843137256f, 0.031218686274509806f, 0.014763607843137252f,
-        0.4552860392156863f, 0.019530843137254908f, 0.007381803921568626f,
-        0.45098f, 0.007843f, 0.0f
-    }, 
-    {
-        0.301961f, 0.266667f, 0.239216f,
-        0.31610937062668204f, 0.2789699081122645f, 0.24844317108804306f,
-        0.3302577412533641f, 0.29127281622452905f, 0.25767034217608614f,
-        0.3444061118800461f, 0.30357572433679353f, 0.2668975132641292f,
-        0.35855448250672817f, 0.31587863244905806f, 0.27612468435217224f,
-        0.37270285313341023f, 0.3281815405613226f, 0.2853518554402153f,
-        0.3868512237600923f, 0.34048444867358707f, 0.29457902652825835f,
-        0.40099961553248753f, 0.3531718569780854f, 0.3022683583237217f,
-        0.41514801999231066f, 0.3660899653979239f, 0.3090349865436371f,
-        0.42929642445213384f, 0.3790080738177624f, 0.3158016147635525f,
-        0.44344482891195697f, 0.39192618223760095f, 0.3225682429834679f,
-        0.45759323337178015f, 0.4048442906574395f, 0.32933487120338334f,
-        0.4717416378316033f, 0.41776239907727797f, 0.33610149942329876f,
-        0.48558246828143026f, 0.43052672049211843f, 0.342560553633218f,
-        0.4985005767012688f, 0.4428296808919646f, 0.3480968858131488f,
-        0.5114186851211073f, 0.45513264129181086f, 0.3536332179930796f,
-        0.5243367935409458f, 0.46743560169165704f, 0.3591695501730104f,
-        0.5372549019607843f, 0.47973856209150323f, 0.36470588235294116f,
-        0.5501730103806228f, 0.4920415224913494f, 0.370242214532872f,
-        0.5630911188004613f, 0.5043444828911956f, 0.37577854671280275f,
-        0.5733179546328335f, 0.5139561707035755f, 0.3780853517877739f,
-        0.5831603229527105f, 0.5231833910034601f, 0.37993079584775086f,
-        0.5930026912725874f, 0.5324106113033449f, 0.3817762399077278f,
-        0.6028450595924644f, 0.5416378316032295f, 0.38362168396770474f,
-        0.6126874279123414f, 0.5508650519031142f, 0.3854671280276817f,
-        0.6225297962322184f, 0.5600922722029988f, 0.3873125720876586f,
-        0.6314494425221069f, 0.568396770472895f, 0.38762014609765477f,
-        0.639446366782007f, 0.5757785467128028f, 0.3863898500576701f,
-        0.647443291041907f, 0.5831603229527105f, 0.3851595540176855f,
-        0.655440215301807f, 0.5905420991926182f, 0.3839292579777009f,
-        0.663437139561707f, 0.597923875432526f, 0.38269896193771624f,
-        0.6714340638216071f, 0.6053056516724338f, 0.38146866589773165f,
-        0.6795847750865053f, 0.6129181084198385f, 0.3803921568627451f,
-        0.6888119953863898f, 0.6221453287197232f, 0.3803921568627451f,
-        0.6980392156862746f, 0.6313725490196078f, 0.3803921568627451f,
-        0.7072664359861592f, 0.6405997693194925f, 0.3803921568627451f,
-        0.7164936562860439f, 0.6498269896193771f, 0.3803921568627451f,
-        0.7257208765859284f, 0.6590542099192618f, 0.3803921568627451f,
-        0.7349480968858132f, 0.6682814302191464f, 0.3803921568627451f,
-        0.7455594002306806f, 0.6793540945790081f, 0.3845444059976932f,
-        0.7566320645905421f, 0.6910419069588619f, 0.390080738177624f,
-        0.7677047289504036f, 0.7027297193387158f, 0.3956170703575547f,
-        0.7787773933102653f, 0.7144175317185698f, 0.4011534025374856f,
-        0.789850057670127f, 0.7261053440984236f, 0.4066897347174164f,
-        0.8009227220299885f, 0.7377931564782776f, 0.4122260668973472f,
-        0.8117647058823528f, 0.7499423298731256f, 0.4200692041522491f,
-        0.8222222222222222f, 0.7628604382929642f, 0.4317570165321031f,
-        0.8326797385620915f, 0.7757785467128028f, 0.44344482891195697f,
-        0.8431372549019608f, 0.7886966551326413f, 0.45513264129181086f,
-        0.85359477124183f, 0.8016147635524797f, 0.46682045367166464f,
-        0.8640522875816994f, 0.8145328719723183f, 0.47850826605151864f,
-        0.8745098039215686f, 0.8274509803921568f, 0.49019607843137253f,
-        0.8837370242214533f, 0.8391387927720108f, 0.5037293348712033f,
-        0.8929642445213379f, 0.8508266051518646f, 0.5172625913110341f,
-        0.9021914648212226f, 0.8625144175317185f, 0.530795847750865f,
-        0.9114186851211072f, 0.8742022299115725f, 0.5443291041906958f,
-        0.9206459054209919f, 0.8858900422914264f, 0.5578623606305266f,
-        0.9298731257208765f, 0.8975778546712803f, 0.5713956170703574f,
-        0.9348712033833141f, 0.9050365244136871f, 0.5864667435601691f,
-        0.9373317954632834f, 0.9099577085736256f, 0.6024605920799692f,
-        0.9397923875432526f, 0.9148788927335639f, 0.6184544405997693f,
-        0.9427912341407151f, 0.9203383314109957f, 0.6355247981545558f,
-        0.946482122260669f, 0.9264898116109188f, 0.6539792387543253f,
-        0.9501730103806227f, 0.932641291810842f, 0.6724336793540946f,
-        0.9803921568627451f, 0.9598796012718599f, 0.6248876284402043f,
-        0.9803921568627451f, 0.9522614779722316f, 0.6033029457579241f,
-        0.9803921568627451f, 0.9446433546726033f, 0.5817182630756441f,
-        0.9802819149805966f, 0.9369609236083882f, 0.5600861151385496f,
-        0.976625215796775f, 0.9272097257848639f, 0.5369270203076794f,
-        0.9729685166129534f, 0.9174585279613396f, 0.513767925476809f,
-        0.9693118174291318f, 0.9077073301378153f, 0.4906088306459388f,
-        0.9686274509803922f, 0.8969653547359303f, 0.4654681806583472f,
-        0.9686274509803922f, 0.8859952571844655f, 0.4398712863715959f,
-        0.9686274509803922f, 0.8750251596330006f, 0.41427439208484457f,
-        0.9679015428456609f, 0.8618773376773421f, 0.3886774977980937f,
-        0.9666826431177203f, 0.8472505409420554f, 0.3630806035113421f,
-        0.9654637433897798f, 0.832623744206769f, 0.3374837092245909f,
-        0.9637838049707373f, 0.8175359087803806f, 0.3137309697022474f,
-        0.9613460055148563f, 0.8016902123171536f, 0.2930096743272582f,
-        0.9589082060589752f, 0.7858445158539267f, 0.27228837895226904f,
-        0.9562742373556216f, 0.7699988193906997f, 0.25195942207222505f,
-        0.9526175381717998f, 0.7541531229274726f, 0.23367592615311694f,
-        0.9489608389879783f, 0.7383074264642459f, 0.2153924302340091f,
-        0.9453041398041566f, 0.7224617300010185f, 0.19710893431490079f,
-        0.9439478396839025f, 0.7054658340060078f, 0.17997563792757645f,
-        0.942728939955962f, 0.6884012378148402f, 0.16291104173640888f,
-        0.9415100402280214f, 0.6713366416236727f, 0.14584644554524134f,
-        0.9385204803237721f, 0.6533867153443508f, 0.13055250953038264f,
-        0.9348637811399505f, 0.6351032194252428f, 0.11592571279509623f,
-        0.9312070819561289f, 0.6168197235061348f, 0.10129891605980981f,
-        0.9281708434168324f, 0.599777148876077f, 0.08915396190262356f,
-        0.9257330439609512f, 0.5839314524128498f, 0.07940276407909912f,
-        0.9232952445050702f, 0.5680857559496227f, 0.0696515662555748f,
-        0.9208574450491891f, 0.5522400594863958f, 0.060967142034737444f,
-        0.918419645593308f, 0.5363943630231689f, 0.054872643395034774f,
-        0.915981846137427f, 0.5205486665599419f, 0.048778144755332105f,
-        0.9134533249242796f, 0.5046122483394486f, 0.04277436787289571f,
-        0.909796625740458f, 0.4875476521482812f, 0.03789876896113357f,
-        0.9061399265566364f, 0.470483055957114f, 0.033023170049371504f,
-        0.9024832273728148f, 0.4534184597659463f, 0.028147571137609302f,
-        0.8904685118563737f, 0.4311301033668914f, 0.023271972225847107f,
-        0.8770606148490276f, 0.407971008536021f, 0.01839637331408491f,
-        0.8636527178416815f, 0.3848119137051505f, 0.013520774402322707f,
-        0.845565525246647f, 0.36009305367838407f, 0.009425058088508648f,
-        0.824844229871658f, 0.33449615939163285f, 0.005768358904687046f,
-        0.8041229344966688f, 0.30889926510488164f, 0.002111659720865444f,
-        0.7880367575105484f, 0.2869074628983617f, 0.0f,
-        0.778285559687024f, 0.269842866707194f, 0.0f,
-        0.7685343618634997f, 0.2527782705160265f, 0.0f,
-        0.756281726933082f, 0.23621396174623774f, 0.0007504311320680208f,
-        0.7343415318301524f, 0.22158716501095133f, 0.004407130315889622f,
-        0.7124013367272228f, 0.2069603682756649f, 0.008063829499711224f,
-        0.6904611416242933f, 0.1923335715403785f, 0.011720528683532825f,
-        0.6709292945113645f, 0.17650260081009153f, 0.014173053872353942f,
-        0.6514268988643162f, 0.16065690434686475f, 0.016610853328234986f,
-        0.6319245032172672f, 0.1448112078836375f, 0.019048652784116098f,
-        0.6086648893647341f, 0.1289655114204105f, 0.021486452239997178f,
-        0.5842868948059233f, 0.11311981495718348f, 0.02392425169587826f,
-        0.5599089002471126f, 0.09727411849395648f, 0.026362051151759335f,
-        0.5348564705805601f, 0.08345172735395483f, 0.03014872082312395f,
-        0.5092595762938088f, 0.07126273007454949f, 0.03502431973488609f,
-        0.48366268200705764f, 0.05907373279514415f, 0.03989991864664823f,
-        0.4580657877203067f, 0.050161260828637984f, 0.045594648886635065f,
-        0.4324688934335549f, 0.04772346137275686f, 0.05290804725427843f,
-        0.4068719991468035f, 0.045285661916875775f, 0.06022144562192171f,
-        0.37216887030186757f, 0.04559514544574167f, 0.06869082317427866f,
-        0.26986239238895643f, 0.06630005470405285f, 0.08574202056008974f,
-        0.16755591447604534f, 0.08700496396236404f, 0.10279321794590082f,
-        0.0652494365631342f, 0.10770987322067523f, 0.11984441533171192f,
-        0.060405735428046226f, 0.12741578495679112f, 0.13807992559587734f,
-        0.06652876344946963f, 0.14700928726218748f, 0.15644869738821665f,
-        0.07265179147089304f, 0.16660278956758384f, 0.17481746918055596f,
-        0.07706893519620056f, 0.18534345847694872f, 0.18892163898463246f,
-        0.08074281446344087f, 0.2037125425412193f, 0.20116769502747953f,
-        0.08441669373068106f, 0.22208162660548933f, 0.21341375107032623f,
-        0.08748533055866523f, 0.23863498335199132f, 0.22444932223466088f,
-        0.08993447931284838f, 0.25332987587709027f, 0.23424591725139352f,
-        0.09238362806703153f, 0.2680247684021892f, 0.24404251226812612f,
-        0.0955475536424294f, 0.28486417366221145f, 0.2534817644420712f,
-        0.10044585115079575f, 0.3069071369937219f, 0.26205394121767783f,
-        0.10534414865916211f, 0.32895010032523236f, 0.2706261179932844f,
-        0.10991359756616595f, 0.3504449640201837f, 0.278431f,
-        0.11113801580729209f, 0.36636458705833985f, 0.278431f,
-        0.11236243404841822f, 0.382284210096496f, 0.278431f,
-        0.11358685228954435f, 0.39820383313465213f, 0.278431f,
-        0.11481154756982646f, 0.414123456172808f, 0.2762581818995031f,
-        0.11603627808288348f, 0.4300430792109638f, 0.27380903314531996f,
-        0.11726100859594052f, 0.4459627022491196f, 0.2713598843911368f,
-        0.117647f, 0.4593665356701801f, 0.26471746780186367f,
-        0.117647f, 0.47161259171302716f, 0.25614529102625694f,
-        0.117647f, 0.48385864775587384f, 0.24757311425065043f,
-        0.11705606935183134f, 0.49492299317312477f, 0.23722844687202088f,
-        0.11583133883877429f, 0.5047199004617884f, 0.224982703101105f,
-        0.11460660832571724f, 0.514516807750452f, 0.2127369593301891f,
-        0.11338196529919445f, 0.5239705928517758f, 0.19946184899725378f,
-        0.11215754705806832f, 0.5325427696273825f, 0.18354191368716674f,
-        0.11093312881694219f, 0.541114946402989f, 0.1676219783770797f,
-        0.10913687682140437f, 0.5499730157534745f, 0.15122554734141894f,
-        0.10056470004579793f, 0.56221875952439f, 0.12918289628183982f,
-        0.0919925232701915f, 0.5744645032953057f, 0.10714024522226068f,
-        0.08342034649458506f, 0.5867102470662214f, 0.08509759416268153f,
-        0.0898578302810215f, 0.6000285000059271f, 0.07377593394023943f,
-        0.09843000705662808f, 0.613499286561831f, 0.06397902665157579f,
-        0.10700218383223467f, 0.6269700731177348f, 0.05418211936291216f,
-        0.1312367980013871f, 0.6412648665798616f, 0.05757478792575119f,
-        0.16307635634963116f, 0.6559597591049612f, 0.0673716952144151f,
-        0.19491591469787434f, 0.6706546516300602f, 0.07716860250307873f,
-        0.22733194488156255f, 0.6847732193413503f, 0.08696536277010616f,
-        0.26039592147093155f, 0.698244005897254f, 0.09676195778683878f,
-        0.2934598980603006f, 0.7117147924531577f, 0.10655855280357142f,
-        0.3271814954421744f, 0.72485668477538f, 0.11767030556788498f,
-        0.3626949330576572f, 0.7371024285462957f, 0.1323655103649147f,
-        0.39820837067313997f, 0.7493481723172114f, 0.14706071516194444f,
-        0.43396477278845f, 0.7615939367413488f, 0.16240387348269752f,
-        0.47315177739924275f, 0.7738399927841956f, 0.18689598556839115f,
-        0.5123387820100355f, 0.7860860488270425f, 0.21138809765408478f,
-        0.5515257866208281f, 0.7983321048698893f, 0.2358802097397784f,
-        0.5864816347629046f, 0.810578160912736f, 0.2540253173814523f,
-        0.620770341865331f, 0.8228242169555828f, 0.2711696709326655f,
-        0.6550590489677572f, 0.8350702729984296f, 0.2883140244838786f,
-        0.69096778056709f, 0.8473161224845379f, 0.3151785250165367f,
-        0.7277056364237002f, 0.859561866255454f, 0.3470177710928496f,
-        0.7644434922803094f, 0.8718076100263699f, 0.37885701716916154f,
-        0.7989324066180044f, 0.8846158042358104f, 0.41575659672182685f,
-        0.8307719649662473f, 0.8980865907917142f, 0.4586174805998594f,
-        0.8626115233144902f, 0.911557377347618f, 0.501478364477892f,
-        0.8922497518032245f, 0.9237702496711849f, 0.5443392483559245f,
-        0.91551682110393f, 0.9323424264467914f, 0.5872001322339571f,
-        0.9387838904046354f, 0.940914603222398f, 0.6300610161119896f,
-        0.9614508334268967f, 0.9496201432825605f, 0.6745222934129278f,
-        0.9736968894697435f, 0.9606414688123502f, 0.7467735868850207f,
-        0.9859429455125903f, 0.9716627943421401f, 0.8190248803571136f,
-        0.9981890015554372f, 0.9826841198719298f, 0.8912761738292065f,
-        0.8400593006492788f, 0.890828975486959f, 0.7926814714742926f,
-        0.8225883764452582f, 0.8777257823339436f, 0.7702187864505137f,
-        0.8051174522412377f, 0.8646225891809282f, 0.7477561014267349f,
-        0.7876465280372176f, 0.8515193960279132f, 0.7252934164029566f,
-        0.7701756038331965f, 0.8384162028748974f, 0.7028307313791772f,
-        0.7527046796291759f, 0.8253130097218819f, 0.6803680463553983f,
-        0.735938331489784f, 0.8125620596836707f, 0.659666936245722f,
-        0.7197153077131807f, 0.8000827371893193f, 0.6403242409556197f,
-        0.7034922839365773f, 0.787603414694968f, 0.6209815456655174f,
-        0.6872692601599739f, 0.7751240922006166f, 0.6016388503754151f,
-        0.6710462363833705f, 0.7626447697062653f, 0.5822961550853128f,
-        0.6548232126067671f, 0.7501654472119139f, 0.5629534597952105f,
-        0.6389492119548505f, 0.737511702146327f, 0.5439596986386873f,
-        0.6239742477157536f, 0.7244086681034004f, 0.5258647446659137f,
-        0.6089992834766567f, 0.711305634060474f, 0.50776979069314f,
-        0.5940243192375596f, 0.6982026000175474f, 0.48967483672036627f,
-        0.5790493549984627f, 0.685099565974621f, 0.47157988274759255f,
-        0.5640743907593657f, 0.6719965319316944f, 0.4534849287748189f,
-        0.5490994265202687f, 0.6588934978887679f, 0.43538997480204517f,
-        0.5347450239586111f, 0.6464108672571044f, 0.4197777423375569f,
-        0.5203939303781779f, 0.6339315447627528f, 0.40417874832970635f,
-        0.5060428367977452f, 0.6214522222684016f, 0.38857975432185626f,
-        0.49169174321731246f, 0.6089728997740504f, 0.3729807603140061f,
-        0.47734064963687967f, 0.5964935772796992f, 0.35738176630615603f,
-        0.4629895560564469f, 0.584014254785348f, 0.34178277229830595f,
-        0.4495241466615725f, 0.5715350452177066f, 0.32751224810543833f,
-        0.436420953508557f, 0.5590558818334442f, 0.31378502518366974f,
-        0.4233177603555415f, 0.5465767184491819f, 0.3000578022619011f,
-        0.41021456720252597f, 0.5340975550649195f, 0.28633057934013245f,
-        0.39711137404951047f, 0.5216183916806572f, 0.27260335641836386f,
-        0.384008180896495f, 0.5091392282963948f, 0.2588761334965952f,
-        0.371169997324889f, 0.4963949877434795f, 0.2459441420807853f,
-        0.35869067483053774f, 0.483291794590464f, 0.23408900846527608f,
-        0.34621135233618644f, 0.4701886014374485f, 0.22223387484976687f,
-        0.33373202984183514f, 0.457085408284433f, 0.21037874123425765f,
-        0.32125270734748423f, 0.4439822151314179f, 0.19852360761874877f,
-        0.3087733848531326f, 0.430879021978402f, 0.1866684740032392f,
-        0.29646857392112475f, 0.4178630623587813f, 0.17507504098791427f,
-        0.28523731096427973f, 0.40538373986443f, 0.16509151934839766f,
-        0.2740060480074347f, 0.39290441737007875f, 0.1551079977088811f,
-        0.2627747850505897f, 0.38042509487572745f, 0.1451244760693645f,
-        0.25154352209374464f, 0.36794577238137616f, 0.13514095442984791f,
-        0.24031225913689966f, 0.3554664498870249f, 0.1251574327903313f,
-        0.22908099618005462f, 0.3429871273926736f, 0.11517391115081471f,
-        0.21891652552969754f, 0.32997461277267753f, 0.10839103846756061f,
-        0.20893316300026993f, 0.316871578729751f, 0.10215153633047391f,
-        0.1989498004708423f, 0.3037685446868245f, 0.09591203419338723f,
-        0.18896643794141466f, 0.290665510643898f, 0.08967253205630053f,
-        0.17898307541198705f, 0.2775624766009715f, 0.08343302991921385f,
-        0.1689997128825594f, 0.264459442558045f, 0.07719352778212715f,
-        0.15937190244916125f, 0.251711869932215f, 0.07059838287007802f,
-        0.15001241057839806f, 0.2392325474378641f, 0.06373485096423837f,
-        0.14065291870763433f, 0.22675322494351247f, 0.05687131905839832f,
-        0.1312934268368709f, 0.21427390244916117f, 0.050007787152558464f,
-        0.12193393496610741f, 0.20179457995480987f, 0.04314425524671861f,
-        0.11257444309534397f, 0.18931525746045863f, 0.036280723340878764f,
-        0.10357059399954295f, 0.17683598030557382f, 0.030483983741526836f,
-        0.09545916166628579f, 0.16435681692131152f, 0.02736415311793903f,
-        0.08734772933302863f, 0.1518776535370492f, 0.024244322494351222f,
-        0.07923629699977147f, 0.1393984901527869f, 0.021124491870763415f,
-        0.07112486466651433f, 0.1269193267685246f, 0.01800466124717561f,
-        0.06301343233325715f, 0.1144401633842623f, 0.014884830623587805f,
-        0.054902f, 0.101961f, 0.011765f
-    }, 
-    {
-        0.984313725490196f, 0.9647058823529412f, 0.7372549019607844f,
-        0.9827921757263818f, 0.9612823953843591f, 0.7273648284959916f,
-        0.9812706259625675f, 0.957858908415777f, 0.717474755031199f,
-        0.9797490761987533f, 0.954435421447195f, 0.7075846815664062f,
-        0.978227526434939f, 0.9510119344786129f, 0.6976946081016135f,
-        0.9767059766711247f, 0.9475884475100308f, 0.6878045346368208f,
-        0.9751844269073104f, 0.9441649605414487f, 0.6779144611720281f,
-        0.9736628771434962f, 0.9407414735728665f, 0.6680243877072354f,
-        0.972141327379682f, 0.9373179866042844f, 0.6581343142424427f,
-        0.9706197776158677f, 0.9338944996357024f, 0.64824424077765f,
-        0.9690982278520535f, 0.9304710126671203f, 0.6383541673128572f,
-        0.9670512916421627f, 0.925734059583347f, 0.6237356158333762f,
-        0.9647689669964413f, 0.9204086354099971f, 0.6069985684314194f,
-        0.9624866423507199f, 0.9150832112366472f, 0.5902615210294625f,
-        0.9602043177049986f, 0.9097577870632972f, 0.5735244736275057f,
-        0.9579219930592772f, 0.9044323628899473f, 0.5567874262255488f,
-        0.9556396684135557f, 0.8982915542602752f, 0.540458071051753f,
-        0.9533573437678344f, 0.8914445803231109f, 0.5244817985317033f,
-        0.951075019122113f, 0.8845976063859468f, 0.5085055260116536f,
-        0.9487926944763916f, 0.8777506324487826f, 0.4925292534916038f,
-        0.9465103698306702f, 0.8709036585116184f, 0.4765529809715541f,
-        0.9436480491644571f, 0.8640566845744543f, 0.4620266985027334f,
-        0.9398441747549214f, 0.8572097106372901f, 0.44985430039221935f,
-        0.9360403003453859f, 0.8503627367001259f, 0.4376819022817052f,
-        0.9322364259358502f, 0.8435157627629617f, 0.42550950417119116f,
-        0.9284325515263145f, 0.8366687888257975f, 0.4133371060606771f,
-        0.9248009809091094f, 0.8299941186809638f, 0.4023708344964765f,
-        0.9217578813814808f, 0.8239079196257069f, 0.3955238605593123f,
-        0.9187147818538524f, 0.8178217205704498f, 0.3886768866221481f,
-        0.9156716823262238f, 0.8117355215151927f, 0.38182991268498395f,
-        0.9126285827985953f, 0.8056493224599357f, 0.3749829387478198f,
-        0.9095854832709668f, 0.7995631234046786f, 0.3681905743850708f,
-        0.9065423837433382f, 0.7934769243494216f, 0.3621043753298137f,
-        0.9034992842157097f, 0.7873907252941645f, 0.35601817627455673f,
-        0.9004561846880812f, 0.7813045262389074f, 0.34993197721929964f,
-        0.8974130851604527f, 0.7752183271836505f, 0.3438457781640426f,
-        0.8943699856328242f, 0.7691321281283934f, 0.33775957910878557f,
-        0.8913268861051956f, 0.7623482388347295f, 0.3323710702919355f,
-        0.8882837865775671f, 0.7555012648975652f, 0.32704564611858555f,
-        0.8852406870499385f, 0.7486542909604009f, 0.32172022194523564f,
-        0.88219758752231f, 0.7418073170232368f, 0.3163947977718857f,
-        0.8791544879946815f, 0.7349603430860726f, 0.3110693735985358f,
-        0.8766913844875447f, 0.7292733611898917f, 0.30632394544567754f,
-        0.8744090598418233f, 0.7239479370165418f, 0.30175929615423475f,
-        0.8721267351961018f, 0.7186225128431919f, 0.29719464686279196f,
-        0.8698444105503805f, 0.7132970886698421f, 0.2926299975713492f,
-        0.8675620859046591f, 0.7079716644964922f, 0.2880653482799064f,
-        0.8651256606580789f, 0.702800340924001f, 0.2839630007910399f,
-        0.8625897443850551f, 0.6977285083779534f, 0.28015912638150425f,
-        0.8600538281120313f, 0.6926566758319059f, 0.27635525197196864f,
-        0.8588235294117647f, 0.687258438892669f, 0.2735305907420008f,
-        0.8588235294117647f, 0.6815526272783656f, 0.271628653537233f,
-        0.47016490869624633f, 0.24019165648847035f, 0.17072639619813998f,
-        0.47793154493285583f, 0.2510649472197236f, 0.17538637794010564f,
-        0.4856981811694654f, 0.26193823795097704f, 0.1800463596820714f,
-        0.4949028789265056f, 0.2728115286822304f, 0.18758246446489846f,
-        0.504222842410437f, 0.2836848194134837f, 0.195349100701508f,
-        0.5135428058943684f, 0.2951812575323605f, 0.20311573693811746f,
-        0.5228627693782999f, 0.3076078755109359f, 0.21088237317472708f,
-        0.5321827328622314f, 0.3200344934895112f, 0.2186490094113366f,
-        0.5428642568483011f, 0.3351842324723629f, 0.22913876665222266f,
-        0.5537375475795543f, 0.35071750494558185f, 0.24001205738347592f,
-        0.5646108383108077f, 0.3657041310494699f, 0.24979205537606716f,
-        0.5754841290420611f, 0.37968407627536704f, 0.2575586916126767f,
-        0.5863574197733145f, 0.3936640215012642f, 0.2653253278492862f,
-        0.595945651020722f, 0.4063589072433155f, 0.2730919640858957f,
-        0.6052656145046532f, 0.41878552522189055f, 0.2808586003225051f,
-        0.6150557233396234f, 0.4316822885515046f, 0.28862523655911476f,
-        0.6259290140708768f, 0.4456622337774018f, 0.2963918727957243f,
-        0.6368023048021301f, 0.459642179003299f, 0.3041585090323339f,
-        0.6476755955333834f, 0.47362212422919614f, 0.3143422622000501f,
-        0.6585488862646367f, 0.4876020694550932f, 0.32521555293130344f,
-        0.6698158213286363f, 0.5023693033464828f, 0.33569519932981057f,
-        0.6822424393072116f, 0.5194559030670238f, 0.34501516281374206f,
-        0.6946690572857868f, 0.5365425027875648f, 0.3543351262976735f,
-        0.7082277327116232f, 0.5536291025081059f, 0.364787147228866f,
-        0.7222076779375203f, 0.5707157022286469f, 0.37566043796011933f,
-        0.7361876231634175f, 0.5878023019491878f, 0.38653372869137265f,
-        0.7501675683893146f, 0.6048889016697289f, 0.397407019422626f,
-        0.7641475136152119f, 0.6219755013902698f, 0.40828031015387933f,
-        0.7770719024121405f, 0.6390621011108107f, 0.4191536008851327f,
-        0.7894985203907157f, 0.6561487008313517f, 0.430026891616386f,
-        0.8016844960731294f, 0.672994658255731f, 0.44090018234763917f,
-        0.8125577868043828f, 0.6885279307289504f, 0.4517734730788927f,
-        0.8234310775356363f, 0.7040612032021695f, 0.4626467638101461f,
-        0.8343043682668896f, 0.7205735310860647f, 0.47449910995207556f,
-        0.845177658998143f, 0.7376601308066056f, 0.48692572793065086f,
-        0.8560509497293963f, 0.7547467305271467f, 0.49984476974283293f,
-        0.8669242404606496f, 0.7718333302476875f, 0.5169313694633738f,
-        0.877797531191903f, 0.7889199299682285f, 0.5340179691839148f,
-        0.8886708219231562f, 0.8069090840811528f, 0.5529096776892228f,
-        0.8995441126544097f, 0.825549011049016f, 0.573102931904408f,
-        0.9102421228665101f, 0.8441012977573025f, 0.5937343874174753f,
-        0.9180087591031196f, 0.8611878974778435f, 0.6216942778692697f,
-        0.9257753953397292f, 0.8782744971983845f, 0.6496541683210642f,
-        0.9335420315763387f, 0.8945350435448343f, 0.6833964323914973f,
-        0.9413086678129482f, 0.9100683160180534f, 0.7222296135745451f,
-        0.9490864432908418f, 0.9256015884912725f, 0.7610627947575929f,
-        0.9584064067747732f, 0.9411348609644913f, 0.7998959759406398f,
-        0.9677263702587047f, 0.9566681334377105f, 0.8387291571236879f,
-        0.9777958860984349f, 0.9714518535551306f, 0.8760633085503734f,
-        0.9886691768296884f, 0.9854317987810279f, 0.9117899905715021f,
-        0.9995424675609417f, 0.999411744006925f, 0.9475166725926307f,
-        0.9305889327343205f, 0.958497068017665f, 0.9016780969008078f,
-        0.9030875168212531f, 0.9355793512916724f, 0.8726486513904852f,
-        0.8732106722226683f, 0.9090985672506915f, 0.8424315672506916f,
-        0.8395975275252713f, 0.8770134521505495f, 0.8103464521505496f,
-        0.8059843828278732f, 0.8449283370504066f, 0.7782613370504068f,
-        0.7811621641470541f, 0.8228898115459841f, 0.7549671479668429f,
-        0.7582440578163706f, 0.8030273748051202f, 0.7335770712334146f,
-        0.7364960805285846f, 0.784335067107154f, 0.7133570240928547f,
-        0.718162062989667f, 0.7690567191580558f, 0.6965506465465015f,
-        0.6998280454507493f, 0.7537783712089577f, 0.6797442690001482f,
-        0.6857017503437861f, 0.743759877497879f, 0.6681977456918142f,
-        0.673478682379817f, 0.7361205087209846f, 0.6590303473176646f,
-        0.6616279964169979f, 0.72848113994409f, 0.6500491874175731f,
-        0.652460987647539f, 0.7208417711671955f, 0.6424102082453693f,
-        0.64329397887808f, 0.713202402390301f, 0.6347712290731656f,
-        0.6324304105239394f, 0.7055630336134066f, 0.6262837538210657f,
-        0.62020734255997f, 0.697923664836512f, 0.617116355446916f,
-        0.6079842745960007f, 0.6902842960596176f, 0.6079489570727663f,
-        0.6002919272827231f, 0.6826449272827231f, 0.6033122793493083f,
-        0.5926525585058288f, 0.6750055585058288f, 0.5987285801622335f,
-        0.5843689262613266f, 0.6673661897289341f, 0.5935006175075511f,
-        0.5752019174918677f, 0.6597268209520397f, 0.5873892783279118f,
-        0.5660349087224088f, 0.6520874521751452f, 0.5812779391482725f,
-        0.5594805999686333f, 0.6444484165233324f, 0.5764727834139342f,
-        0.553369260788994f, 0.6368094373511287f, 0.5718890842268592f,
-        0.5463767700795689f, 0.6291703458444619f, 0.5668649216093548f,
-        0.5372093717054192f, 0.6215309770675673f, 0.5607535824297155f,
-        0.5280419733312695f, 0.6138916082906728f, 0.5546422432500763f,
-        0.5199772395137784f, 0.6051495749571197f, 0.5496332874785599f,
-        0.5123378707368841f, 0.5959821765829703f, 0.5450495882914851f,
-        0.5046985623179064f, 0.5870515019599892f, 0.5402292257111583f,
-        0.4970595831457028f, 0.5794121331830948f, 0.534117886531519f,
-        0.4894206039734991f, 0.5717727644062004f, 0.5280065473518798f,
-        0.48178139562930594f, 0.5641336248012955f, 0.5227937915431856f,
-        0.47414202685241147f, 0.5564946456290918f, 0.5182100923561108f,
-        0.466502658075517f, 0.548855658075517f, 0.5136264015504071f,
-        0.4588632892986226f, 0.5412162892986225f, 0.509043091968023f,
-        0.4512239205217281f, 0.5335769205217281f, 0.5044597823856389f,
-        0.44358455174483363f, 0.5259375517448337f, 0.4998762956078115f,
-        0.4359451829679392f, 0.5182981829679393f, 0.4952925964207367f,
-        0.4283058141910448f, 0.5106588141910448f, 0.49070889723366184f,
-        0.42066644541415016f, 0.5030194454141502f, 0.48476849473512645f,
-        0.4130270766372556f, 0.4953800766372556f, 0.47865715555548705f,
-        0.40440561604828484f, 0.4862676327516955f, 0.47205470786036097f,
-        0.39371057768157075f, 0.47404495439241695f, 0.4644153390834665f,
-        0.383015539314857f, 0.4618222760331388f, 0.45677597030657224f,
-        0.373473698255395f, 0.4507525009481426f, 0.44798369825539497f,
-        0.3643066894859361f, 0.4400574625814285f, 0.4388166894859361f,
-        0.35485242421471447f, 0.4290751677129517f, 0.4290751677129517f,
-        0.3441573858480004f, 0.4168520997489824f, 0.4168520997489824f,
-        0.3334623474812863f, 0.40462903178501314f, 0.40462903178501314f,
-        0.2611752601292498f, 0.3327286183995115f, 0.43173295706193f,
-        0.27131413479749517f, 0.3444273199397946f, 0.4449914854742508f,
-        0.28145300946574053f, 0.3561260214800777f, 0.45825001388657166f,
-        0.2915918841339859f, 0.3678247230203608f, 0.4715085422988925f,
-        0.30173075880223127f, 0.3795234245606439f, 0.4847670707112134f,
-        0.312330193511192f, 0.391222126100927f, 0.49756503908281885f,
-        0.32324898161545623f, 0.40292082764121007f, 0.5100436540591208f,
-        0.33416776971972045f, 0.4146195291814932f, 0.5225222690354228f,
-        0.3450865578239847f, 0.42631823072177627f, 0.5350008840117247f,
-        0.3560053459282489f, 0.4380169322620594f, 0.5474794989880267f,
-        0.3669241340325128f, 0.45015419239570087f, 0.5599581139643284f,
-        0.37784292213677734f, 0.46263280737200324f, 0.5724367289406307f,
-        0.38876171024104156f, 0.4751114223483052f, 0.5849153439169327f,
-        0.39968049834530583f, 0.48759003732460715f, 0.5973939588932347f,
-        0.41059928644957006f, 0.5000686523009091f, 0.6098725738695366f,
-        0.4219346316998364f, 0.5125472672772111f, 0.6223511888458386f,
-        0.43363333324011943f, 0.525025882253513f, 0.6348298038221405f,
-        0.44533203478040245f, 0.5375044972298149f, 0.6473084187984424f,
-        0.45703073632068547f, 0.5499831122061168f, 0.6597870337747443f,
-        0.46872943786096855f, 0.5624617271824187f, 0.6722656487510462f,
-        0.48082269509989706f, 0.5753348978573661f, 0.6851388194259936f,
-        0.493301310076199f, 0.5885934262696869f, 0.6983973478383144f,
-        0.5057799250525009f, 0.6018519546820077f, 0.7116558762506352f,
-        0.5182585400288028f, 0.6151104830943285f, 0.724914404662956f,
-        0.5307371550051047f, 0.6283690115066493f, 0.7381729330752768f,
-        0.5439608784839844f, 0.6420000941702589f, 0.7506863529850197f,
-        0.5579993203323239f, 0.6560385360185984f, 0.7623850545253026f,
-        0.572037762180664f, 0.6700769778669387f, 0.7740837560655861f,
-        0.5860762040290038f, 0.6841154197152784f, 0.7857824576058693f,
-        0.6001146458773436f, 0.6981538615636182f, 0.7974811591461525f,
-        0.6141530877256833f, 0.7121923034119579f, 0.8091798606864355f,
-        0.628191529574023f, 0.7262307452602976f, 0.8208785622267186f,
-        0.6422299714223627f, 0.7402691871086372f, 0.8325772637670017f,
-        0.6562684132707024f, 0.7543076289569769f, 0.8442759653072847f,
-        0.670306855119042f, 0.7683460708053165f, 0.8559746668475677f,
-        0.6836881942542303f, 0.7817274099405048f, 0.8670162656746994f,
-        0.6961668092305322f, 0.7942060249168067f, 0.8771551403429447f,
-        0.7086454242068342f, 0.8066846398931087f, 0.88729401501119f,
-        0.7211240391831361f, 0.8191632548694106f, 0.8974328896794352f,
-        0.7336026541594379f, 0.8316418698457124f, 0.9075717643476805f,
-        0.747000918863397f, 0.8444270347312335f, 0.9174040891067069f,
-        0.7618192741477557f, 0.8576855631435544f, 0.9267630503389334f,
-        0.7766376294321139f, 0.8709440915558749f, 0.9361220115711597f,
-        0.791455984716473f, 0.8842026199681962f, 0.9454809728033865f,
-        0.8062743400008316f, 0.8974611483805172f, 0.9548399340356131f,
-        0.825360922213127f, 0.9121424191021502f, 0.9641988952678395f,
-        0.8518779790377685f, 0.9293005146945653f, 0.9735578565000659f,
-        0.8783950358624101f, 0.9464586102869804f, 0.9829168177322923f,
-        0.9088930132634999f, 0.96228959744134f, 0.9896217611397445f,
-        0.940089425918105f, 0.977887749174702f, 0.9958610686278955f,
-        0.8706129528586563f, 0.9752876514848623f, 0.9504825480858896f,
-        0.8354721694692707f, 0.9600091092988845f, 0.9313843216528329f,
-        0.8003313860798852f, 0.9447305671129068f, 0.9122860952197762f,
-        0.7651906026904998f, 0.929452024926929f, 0.8931878687867195f,
-        0.7300498193011142f, 0.9141734827409513f, 0.8740896423536628f,
-        0.6964420460880201f, 0.8979750953567552f, 0.8551446973920135f,
-        0.6689404363904545f, 0.8781128541533475f, 0.8368102909269698f,
-        0.641438826692889f, 0.8582506129499399f, 0.818475884461926f,
-        0.6139372169953242f, 0.8383883717465328f, 0.8001414779968828f,
-        0.5864356072977578f, 0.8185261305431246f, 0.7818070715318386f,
-        0.5595506781709293f, 0.7993319512219257f, 0.7642435026760817f,
-        0.5412164665082237f, 0.789400928021391f, 0.7573681489522748f,
-        0.5228822548455181f, 0.7794699048208561f, 0.7504927952284679f,
-        0.5045480431828124f, 0.7695388816203214f, 0.743617441504661f,
-        0.48621383152010683f, 0.7596078584197865f, 0.7367420877808541f,
-        0.46787961985740123f, 0.7496768352192518f, 0.7298667340570473f,
-        0.4459781749242958f, 0.7311847432388193f, 0.7151437945910554f,
-        0.42382408421217305f, 0.7120863220034245f, 0.6998650576027395f,
-        0.4016699935000503f, 0.6929879007680295f, 0.6845863206144236f,
-        0.37951590278792746f, 0.6738894795326347f, 0.6693075836261078f,
-        0.35736181207580464f, 0.6547910582972398f, 0.6540288466377918f,
-        0.339488307730249f, 0.634469835142618f, 0.6381387086898624f,
-        0.3226819308059074f, 0.6138437739711975f, 0.6220961517335338f,
-        0.30587555388156584f, 0.5932177127997769f, 0.6060535947772051f,
-        0.2890691769572247f, 0.5725916516283569f, 0.5900110378208768f,
-        0.27226280003288267f, 0.5519655904569358f, 0.5739684808645477f,
-        0.257494556248798f, 0.5333777923673176f, 0.5584355546494423f,
-        0.24374365399884607f, 0.5158074006726249f, 0.5431570124634645f,
-        0.2299927517488941f, 0.49823700897793216f, 0.5278784702774868f,
-        0.21624184949894218f, 0.4806666172832394f, 0.5125999280915091f,
-        0.20249094724899022f, 0.4630962255885467f, 0.4973213859055314f,
-        0.18955545418499564f, 0.4467488437195536f, 0.48285814895227464f,
-        0.17733258147574582f, 0.4314703015335759f, 0.4691074415046609f,
-        0.165109708766496f, 0.41619175934759817f, 0.4553567340570471f,
-        0.15288683605724618f, 0.4009132171616204f, 0.4416060266094333f,
-        0.14066396334799636f, 0.38563467497564274f, 0.4278553191618195f,
-        0.13027560537729616f, 0.3728020484878281f, 0.41655052741236887f,
-        0.12263643168547643f, 0.36363484525530626f, 0.40891115891821095f,
-        0.11499725799365669f, 0.3544676420227844f, 0.40127179042405303f,
-        0.10735808430183696f, 0.3453004387902625f, 0.39363242192989517f,
-        0.09971891061001745f, 0.33613323555774094f, 0.3859930534357375f,
-        0.09167203232521882f, 0.3259468747960085f, 0.3773345793889872f,
-        0.08250482909269696f, 0.3129599873164077f, 0.3658757214500885f,
-        0.07333762586017512f, 0.29997309983680687f, 0.3544168635111898f,
-        0.06417042262765325f, 0.28698621235720606f, 0.3429580055722911f,
-        0.0550032193951314f, 0.27399932487760525f, 0.33149914763339244f,
-        0.04583601616260954f, 0.2608086110898242f, 0.31973451124975977f,
-        0.03666881293008763f, 0.24629388887185935f, 0.3059836089998078f,
-        0.027501609697565724f, 0.23177916665389453f, 0.29223270674985585f,
-        0.018334406465043813f, 0.21726444443592968f, 0.2784818044999039f,
-        0.00916720323252191f, 0.20274972221796486f, 0.26473090224995194f,
-        0.0f, 0.188235f, 0.25098f
-    }, 
-    {
-        0.011765f, 0.015686f, 0.101961f,
-        0.013918019607843137f, 0.018146588235294116f, 0.10580566666666666f,
-        0.016071039215686273f, 0.020607176470588233f, 0.10965033333333334f,
-        0.018224058823529413f, 0.02306776470588235f, 0.113495f,
-        0.02037707843137255f, 0.025528352941176467f, 0.11733966666666666f,
-        0.022530098039215685f, 0.027988941176470584f, 0.12118433333333334f,
-        0.024683117647058824f, 0.0304495294117647f, 0.125029f,
-        0.026836137254901964f, 0.032910117647058826f, 0.12887366666666666f,
-        0.0289891568627451f, 0.03537070588235294f, 0.13271833333333333f,
-        0.031142176470588236f, 0.03783129411764706f, 0.136563f,
-        0.03329519607843137f, 0.04029188235294118f, 0.14040766666666665f,
-        0.035448215686274515f, 0.042752470588235295f, 0.14425233333333334f,
-        0.03760123529411765f, 0.04521305882352941f, 0.148097f,
-        0.03975425490196079f, 0.04767364705882352f, 0.15194166666666667f,
-        0.04190727450980393f, 0.050134235294117646f, 0.15578633333333333f,
-        0.04406029411764706f, 0.05259482352941176f, 0.15963100000000002f,
-        0.046213313725490195f, 0.05505541176470588f, 0.16347566666666669f,
-        0.04836633333333333f, 0.05751599999999999f, 0.16732033333333335f,
-        0.050519352941176474f, 0.059976588235294115f, 0.171165f,
-        0.05267237254901961f, 0.06243717647058823f, 0.17500966666666667f,
-        0.054825392156862746f, 0.06489776470588235f, 0.17885433333333334f,
-        0.05697841176470588f, 0.06735835294117645f, 0.182699f,
-        0.059131431372549025f, 0.06981894117647058f, 0.18654366666666666f,
-        0.06128445098039216f, 0.0722795294117647f, 0.19038833333333333f,
-        0.0634374705882353f, 0.07474011764705882f, 0.194233f,
-        0.06559049019607843f, 0.07720070588235294f, 0.19807766666666668f,
-        0.06766660784313726f, 0.08019954901960784f, 0.20238370588235297f,
-        0.06966582352941177f, 0.08373664705882353f, 0.20715111764705885f,
-        0.07166503921568627f, 0.08727374509803922f, 0.21191852941176473f,
-        0.07366425490196078f, 0.0908108431372549f, 0.2166859411764706f,
-        0.0756634705882353f, 0.09434794117647059f, 0.2214533529411765f,
-        0.07766268627450981f, 0.09788503921568628f, 0.22622076470588237f,
-        0.07966190196078432f, 0.10142213725490196f, 0.23098817647058822f,
-        0.08166111764705881f, 0.10495923529411763f, 0.23575558823529408f,
-        0.08366033333333334f, 0.10849633333333333f, 0.240523f,
-        0.08565954901960784f, 0.11203343137254902f, 0.2452904117647059f,
-        0.08765876470588235f, 0.11557052941176471f, 0.25005782352941175f,
-        0.08965798039215685f, 0.11910762745098036f, 0.2548252352941176f,
-        0.09165719607843137f, 0.12264472549019607f, 0.2595926470588235f,
-        0.09365641176470588f, 0.12618182352941176f, 0.2643600588235294f,
-        0.0956556274509804f, 0.12971892156862747f, 0.2691274705882353f,
-        0.0976548431372549f, 0.1332560196078431f, 0.2738948823529411f,
-        0.09965405882352942f, 0.13679311764705882f, 0.2786622941176471f,
-        0.10165327450980392f, 0.14033021568627452f, 0.28342970588235294f,
-        0.10365249019607843f, 0.1438673137254902f, 0.2881971176470588f,
-        0.10565170588235293f, 0.14740441176470587f, 0.29296452941176465f,
-        0.10765092156862746f, 0.15094150980392157f, 0.2977319411764706f,
-        0.10965013725490197f, 0.15447860784313727f, 0.30249935294117647f,
-        0.11164935294117648f, 0.15801570588235292f, 0.3072667647058823f,
-        0.11364856862745096f, 0.16155280392156862f, 0.3120341764705882f,
-        0.1156477843137255f, 0.16508990196078432f, 0.31680158823529414f,
-        0.117647f, 0.168627f, 0.321569f,
-        0.11949245098039216f, 0.17124139215686274f, 0.3246447254901961f,
-        0.1213379019607843f, 0.17385578431372548f, 0.32772045098039215f,
-        0.12318335294117647f, 0.17647017647058824f, 0.33079617647058823f,
-        0.12502880392156862f, 0.17908456862745098f, 0.3338719019607843f,
-        0.12687425490196078f, 0.18169896078431372f, 0.3369476274509804f,
-        0.12871970588235293f, 0.18431335294117646f, 0.34002335294117647f,
-        0.13056515686274509f, 0.18692774509803922f, 0.34309907843137255f,
-        0.13241060784313724f, 0.18954213725490196f, 0.34617480392156863f,
-        0.13425605882352942f, 0.1921565294117647f, 0.3492505294117647f,
-        0.13610150980392155f, 0.19477092156862744f, 0.3523262549019608f,
-        0.13794696078431373f, 0.1973853137254902f, 0.35540198039215687f,
-        0.1397924117647059f, 0.19999970588235294f, 0.35847770588235295f,
-        0.14163786274509804f, 0.20261409803921568f, 0.36155343137254903f,
-        0.1434833137254902f, 0.20522849019607844f, 0.3646291568627451f,
-        0.14532876470588232f, 0.20784288235294113f, 0.36770488235294113f,
-        0.1471742156862745f, 0.21045727450980392f, 0.37078060784313727f,
-        0.14901966666666666f, 0.21307166666666666f, 0.37385633333333335f,
-        0.15086511764705882f, 0.21568605882352943f, 0.3769320588235294f,
-        0.15271056862745097f, 0.21830045098039216f, 0.3800077843137255f,
-        0.15455601960784313f, 0.2209148431372549f, 0.3830835098039216f,
-        0.15640147058823528f, 0.22352923529411764f, 0.38615923529411766f,
-        0.15824692156862744f, 0.2261436274509804f, 0.38923496078431374f,
-        0.1600923725490196f, 0.2287580196078431f, 0.39231068627450977f,
-        0.16193782352941177f, 0.23137241176470588f, 0.3953864117647059f,
-        0.1637832745098039f, 0.23398680392156862f, 0.398462137254902f,
-        0.16609007843137255f, 0.23721633333333333f, 0.4019992352941177f,
-        0.16885823529411764f, 0.241061f, 0.40599770588235295f,
-        0.17162639215686273f, 0.24490566666666666f, 0.4099961764705883f,
-        0.17439454901960783f, 0.24875033333333332f, 0.41399464705882355f,
-        0.17716270588235294f, 0.252595f, 0.4179931176470588f,
-        0.179930862745098f, 0.2564396666666666f, 0.4219915882352941f,
-        0.18269901960784313f, 0.26028433333333334f, 0.4259900588235294f,
-        0.18546717647058822f, 0.264129f, 0.42998852941176474f,
-        0.18823533333333334f, 0.26797366666666667f, 0.433987f,
-        0.19100349019607843f, 0.27181833333333333f, 0.4379854705882353f,
-        0.19377164705882352f, 0.275663f, 0.4419839411764706f,
-        0.1965398039215686f, 0.27950766666666665f, 0.4459824117647059f,
-        0.19930796078431373f, 0.2833523333333333f, 0.4499808823529412f,
-        0.2020761176470588f, 0.2871969999999999f, 0.4539793529411764f,
-        0.20484427450980391f, 0.29104166666666664f, 0.45797782352941174f,
-        0.207612431372549f, 0.2948863333333333f, 0.46197629411764707f,
-        0.21038058823529412f, 0.29873099999999997f, 0.46597476470588234f,
-        0.21314874509803922f, 0.3025756666666667f, 0.46997323529411766f,
-        0.2159169019607843f, 0.3064203333333333f, 0.47397170588235293f,
-        0.2186850588235294f, 0.310265f, 0.47797017647058826f,
-        0.22145321568627452f, 0.3141096666666667f, 0.48196864705882353f,
-        0.22422137254901958f, 0.3179543333333333f, 0.48596711764705874f,
-        0.2269895294117647f, 0.321799f, 0.4899655882352941f,
-        0.22975768627450982f, 0.32564366666666666f, 0.4939640588235294f,
-        0.2325258431372549f, 0.3294883333333333f, 0.4979625294117647f,
-        0.235294f, 0.333333f, 0.501961f,
-        0.23744701960784315f, 0.3365625294117647f, 0.5050367254901961f,
-        0.2396000392156863f, 0.3397920588235294f, 0.5081124509803921f,
-        0.2417530588235294f, 0.3430215882352941f, 0.5111881764705882f,
-        0.24390607843137252f, 0.3462511176470588f, 0.5142639019607843f,
-        0.2460590980392157f, 0.34948064705882353f, 0.5173396274509804f,
-        0.24821211764705883f, 0.3527101764705882f, 0.5204153529411765f,
-        0.25036513725490195f, 0.3559397058823529f, 0.5234910784313725f,
-        0.2525181568627451f, 0.35916923529411765f, 0.5265668039215686f,
-        0.25467117647058823f, 0.36239876470588234f, 0.5296425294117647f,
-        0.2568241960784314f, 0.3656282941176471f, 0.5327182549019608f,
-        0.2589772156862745f, 0.36885782352941177f, 0.5357939803921569f,
-        0.2611302352941176f, 0.3720873529411764f, 0.538869705882353f,
-        0.2632832549019608f, 0.37531688235294114f, 0.541945431372549f,
-        0.26543627450980395f, 0.3785464117647059f, 0.5450211568627451f,
-        0.2675892941176471f, 0.38177594117647057f, 0.5480968823529412f,
-        0.2697423137254902f, 0.3850054705882353f, 0.5511726078431373f,
-        0.2718953333333333f, 0.388235f, 0.5542483333333333f,
-        0.27404835294117647f, 0.3914645294117647f, 0.5573240588235294f,
-        0.2762013725490196f, 0.3946940588235294f, 0.5603997843137255f,
-        0.27835439215686275f, 0.39792358823529406f, 0.5634755098039216f,
-        0.2805074117647059f, 0.4011531176470588f, 0.5665512352941177f,
-        0.28266043137254904f, 0.40438264705882354f, 0.5696269607843137f,
-        0.2848134509803922f, 0.40761217647058823f, 0.5727026862745098f,
-        0.2869664705882353f, 0.4108417058823529f, 0.5757784117647059f,
-        0.2891194901960784f, 0.41407123529411766f, 0.578854137254902f,
-        0.2915031960784314f, 0.41737766666666665f, 0.5817760784313726f,
-        0.2941175882352941f, 0.420761f, 0.5845442352941177f,
-        0.29673198039215687f, 0.42414433333333335f, 0.5873123921568628f,
-        0.29934637254901963f, 0.42752766666666664f, 0.5900805490196078f,
-        0.3019607647058823f, 0.4309109999999999f, 0.5928487058823528f,
-        0.3045751568627451f, 0.43429433333333334f, 0.5956168627450981f,
-        0.3071895490196078f, 0.4376776666666667f, 0.5983850196078432f,
-        0.3098039411764706f, 0.441061f, 0.6011531764705883f,
-        0.31241833333333335f, 0.44444433333333333f, 0.6039213333333333f,
-        0.31503272549019606f, 0.4478276666666667f, 0.6066894901960784f,
-        0.31764711764705883f, 0.451211f, 0.6094576470588235f,
-        0.32026150980392154f, 0.4545943333333333f, 0.6122258039215687f,
-        0.3228759019607843f, 0.45797766666666667f, 0.6149939607843138f,
-        0.32549029411764707f, 0.461361f, 0.6177621176470588f,
-        0.3281046862745098f, 0.4647443333333333f, 0.6205302745098039f,
-        0.33071907843137255f, 0.46812766666666666f, 0.623298431372549f,
-        0.3333334705882353f, 0.471511f, 0.6260665882352942f,
-        0.335947862745098f, 0.4748943333333333f, 0.6288347450980393f,
-        0.3385622549019608f, 0.47827766666666666f, 0.6316029019607843f,
-        0.3411766470588235f, 0.481661f, 0.6343710588235294f,
-        0.3437910392156862f, 0.48504433333333324f, 0.6371392156862744f,
-        0.34640543137254903f, 0.48842766666666665f, 0.6399073725490196f,
-        0.34901982352941174f, 0.491811f, 0.6426755294117648f,
-        0.3516342156862745f, 0.4951943333333333f, 0.6454436862745098f,
-        0.3542486078431373f, 0.49857766666666664f, 0.6482118431372549f,
-        0.356863f, 0.501961f, 0.65098f,
-        0.3600925294117647f, 0.5062670392156863f, 0.6540557647058823f,
-        0.3633220588235294f, 0.5105730784313726f, 0.6571315294117647f,
-        0.3665515882352941f, 0.5148791176470588f, 0.660207294117647f,
-        0.3697811176470588f, 0.5191851568627451f, 0.6632830588235294f,
-        0.37301064705882353f, 0.5234911960784314f, 0.6663588235294118f,
-        0.3762401764705882f, 0.5277972352941176f, 0.6694345882352941f,
-        0.37946970588235296f, 0.5321032745098039f, 0.6725103529411764f,
-        0.38269923529411765f, 0.5364093137254902f, 0.6755861176470588f,
-        0.38592876470588233f, 0.5407153529411765f, 0.6786618823529411f,
-        0.389158294117647f, 0.5450213921568627f, 0.6817376470588236f,
-        0.39238782352941165f, 0.5493274313725489f, 0.6848134117647058f,
-        0.39561735294117645f, 0.5536334705882353f, 0.6878891764705882f,
-        0.3988468823529412f, 0.5579395098039216f, 0.6909649411764706f,
-        0.4020764117647059f, 0.5622455490196079f, 0.6940407058823529f,
-        0.40530594117647056f, 0.5665515882352941f, 0.6971164705882352f,
-        0.4085354705882353f, 0.5708576274509803f, 0.7001922352941177f,
-        0.411765f, 0.5751636666666666f, 0.703268f,
-        0.4149945294117647f, 0.5794697058823529f, 0.7063437647058823f,
-        0.4182240588235294f, 0.5837757450980392f, 0.7094195294117647f,
-        0.4214535882352941f, 0.5880817843137255f, 0.712495294117647f,
-        0.4246831176470588f, 0.5923878235294118f, 0.7155710588235293f,
-        0.42791264705882354f, 0.5966938627450981f, 0.7186468235294117f,
-        0.4311421764705882f, 0.6009999019607843f, 0.7217225882352941f,
-        0.43437170588235297f, 0.6053059411764706f, 0.7247983529411764f,
-        0.43760123529411765f, 0.6096119803921569f, 0.7278741176470588f,
-        0.44083076470588234f, 0.6138411176470588f, 0.7307960784313725f,
-        0.44406029411764697f, 0.6179933529411764f, 0.7335642352941175f,
-        0.44728982352941177f, 0.6221455882352941f, 0.7363323921568627f,
-        0.45051935294117645f, 0.6262978235294118f, 0.7391005490196078f,
-        0.45374888235294114f, 0.6304500588235294f, 0.7418687058823529f,
-        0.4569784117647059f, 0.634602294117647f, 0.744636862745098f,
-        0.46020794117647057f, 0.6387545294117647f, 0.7474050196078431f,
-        0.46343747058823526f, 0.6429067647058824f, 0.7501731764705882f,
-        0.466667f, 0.647059f, 0.7529413333333334f,
-        0.4698965294117647f, 0.6512112352941176f, 0.7557094901960784f,
-        0.47312605882352937f, 0.6553634705882353f, 0.7584776470588235f,
-        0.4763555882352941f, 0.659515705882353f, 0.7612458039215686f,
-        0.4795851176470588f, 0.6636679411764707f, 0.7640139607843137f,
-        0.4828146470588235f, 0.6678201764705882f, 0.7667821176470588f,
-        0.4860441764705882f, 0.6719724117647059f, 0.769550274509804f,
-        0.4892737058823529f, 0.6761246470588236f, 0.7723184313725491f,
-        0.4925032352941176f, 0.6802768823529413f, 0.7750865882352941f,
-        0.49573276470588223f, 0.6844291176470587f, 0.7778547450980392f,
-        0.49896229411764703f, 0.6885813529411765f, 0.7806229019607843f,
-        0.5021918235294117f, 0.6927335882352942f, 0.7833910588235294f,
-        0.5054213529411764f, 0.6968858235294118f, 0.7861592156862746f,
-        0.5086508823529411f, 0.7010380588235294f, 0.7889273725490197f,
-        0.5118804117647058f, 0.7051902941176471f, 0.7916955294117647f,
-        0.5151099411764706f, 0.7093425294117648f, 0.7944636862745098f,
-        0.5183394705882353f, 0.7134947647058824f, 0.7972318431372549f,
-        0.521569f, 0.717647f, 0.8f,
-        0.5263363725490195f, 0.7221068431372549f, 0.8030757254901961f,
-        0.5311037450980391f, 0.7265666862745098f, 0.8061514509803922f,
-        0.5358711176470587f, 0.7310265294117647f, 0.8092271764705883f,
-        0.5406384901960783f, 0.7354863725490196f, 0.8123029019607844f,
-        0.545405862745098f, 0.7399462156862745f, 0.8153786274509804f,
-        0.5501732352941175f, 0.7444060588235294f, 0.8184543529411765f,
-        0.5549406078431373f, 0.7488659019607844f, 0.8215300784313726f,
-        0.5597079803921566f, 0.7533257450980391f, 0.8246058039215686f,
-        0.5644753529411765f, 0.7577855882352942f, 0.8276815294117648f,
-        0.569242725490196f, 0.7622454313725491f, 0.8307572549019608f,
-        0.5740100980392157f, 0.766705274509804f, 0.8338329803921569f,
-        0.5787774705882353f, 0.7711651176470589f, 0.836908705882353f,
-        0.5835448431372549f, 0.7756249607843138f, 0.839984431372549f,
-        0.5883122156862745f, 0.7800848039215686f, 0.8430601568627452f,
-        0.593079588235294f, 0.7845446470588235f, 0.8461358823529411f,
-        0.5978469607843137f, 0.7890044901960784f, 0.8492116078431372f,
-        0.6026143333333333f, 0.7934643333333333f, 0.8522873333333333f,
-        0.607381705882353f, 0.7979241764705882f, 0.8553630588235294f,
-        0.6121490784313726f, 0.8023840196078431f, 0.8584387843137254f,
-        0.6169164509803922f, 0.806843862745098f, 0.8615145098039215f,
-        0.6216838235294118f, 0.811303705882353f, 0.8645902352941176f,
-        0.6264511960784314f, 0.8157635490196079f, 0.8676659607843137f,
-        0.631218568627451f, 0.8202233921568628f, 0.8707416862745098f,
-        0.6359859411764704f, 0.8246832352941176f, 0.8738174117647057f,
-        0.6407533137254902f, 0.8291430784313726f, 0.8768931372549019f,
-        0.6485195490196078f, 0.835063862745098f, 0.8811991960784313f,
-        0.6592846470588235f, 0.8424455882352941f, 0.8867355882352941f,
-        0.6700497450980392f, 0.8498273137254903f, 0.8922719803921568f,
-        0.6808148431372548f, 0.8572090392156864f, 0.8978083725490196f,
-        0.6915799411764706f, 0.8645907647058824f, 0.9033447647058823f,
-        0.7023450392156863f, 0.8719724901960785f, 0.9088811568627451f,
-        0.7131101372549019f, 0.8793542156862746f, 0.9144175490196078f,
-        0.7238752352941176f, 0.8867359411764706f, 0.9199539411764706f,
-        0.7346403333333333f, 0.8941176666666667f, 0.9254903333333333f,
-        0.745405431372549f, 0.9014993921568628f, 0.931026725490196f,
-        0.7561705294117647f, 0.9088811176470588f, 0.9365631176470588f,
-        0.7669356274509804f, 0.9162628431372549f, 0.9420995098039215f,
-        0.7777007254901961f, 0.923644568627451f, 0.9476359019607843f,
-        0.7891578823529412f, 0.9298729411764706f, 0.9520188235294117f,
-        0.8008457254901957f, 0.9357168627450979f, 0.9560172549019607f,
-        0.812533568627451f, 0.9415607843137255f, 0.9600156862745098f,
-        0.8242214117647059f, 0.9474047058823529f, 0.9640141176470588f,
-        0.8359092549019608f, 0.9532486274509804f, 0.9680125490196079f,
-        0.8475970980392157f, 0.9590925490196078f, 0.9720109803921568f,
-        0.8592849411764706f, 0.9649364705882353f, 0.9760094117647059f,
-        0.8709727843137255f, 0.9707803921568627f, 0.9800078431372549f,
-        0.8826606274509804f, 0.9766243137254902f, 0.9840062745098039f,
-        0.8943484705882353f, 0.9824682352941176f, 0.9880047058823529f,
-        0.9060363137254902f, 0.9883121568627451f, 0.992003137254902f,
-        0.9177241568627451f, 0.9941560784313725f, 0.9960015686274509f,
-        0.929412f, 1.0f, 1.0f
-    }
-, 
-{
-0.8f, 1.0f, 0.984314f,
-0.7684737254901961f, 0.9938484313725491f, 0.9797002745098039f,
-0.7369474509803922f, 0.987696862745098f, 0.9750865490196079f,
-0.7054211764705882f, 0.981545294117647f, 0.9704728235294118f,
-0.6738949019607844f, 0.9753937254901961f, 0.9658590980392157f,
-0.6423686274509804f, 0.9692421568627452f, 0.9612453725490195f,
-0.629527411764706f, 0.9626293529411765f, 0.9570931764705882f,
-0.6187623137254902f, 0.9559653006535948f, 0.9529922614379085f,
-0.6079972156862745f, 0.949301248366013f, 0.9488913464052288f,
-0.5972321176470589f, 0.9426371960784313f, 0.944790431372549f,
-0.5864670196078432f, 0.9359731437908497f, 0.9406895163398693f,
-0.5757019215686274f, 0.929309091503268f, 0.9365886013071896f,
-0.5649368235294118f, 0.9226450392156863f, 0.9324876862745098f,
-0.5553251176470588f, 0.9163398039215687f, 0.9288737450980392f,
-0.5491735882352942f, 0.9111110196078431f, 0.9267207254901961f,
-0.5430220588235294f, 0.9058822352941176f, 0.9245677058823529f,
-0.5368705294117647f, 0.9006534509803922f, 0.9224146862745098f,
-0.5307189999999999f, 0.8954246666666666f, 0.9202616666666666f,
-0.5245674705882353f, 0.8901958823529411f, 0.9181086470588236f,
-0.5184159411764706f, 0.8849670980392157f, 0.9159556274509804f,
-0.5122644117647058f, 0.8797383137254902f, 0.9138026078431373f,
-0.5061128823529412f, 0.8745095294117646f, 0.9116495882352942f,
-0.4999613529411765f, 0.8692807450980392f, 0.909496568627451f,
-0.4938098235294118f, 0.8640519607843137f, 0.9073435490196079f,
-0.48765829411764705f, 0.8588231764705883f, 0.9051905294117647f,
-0.48150676470588233f, 0.8535943921568627f, 0.9030375098039216f,
-0.4753552745098039f, 0.8480580784313725f, 0.9007306862745098f,
-0.46920382352941176f, 0.8422142352941177f, 0.8982700588235294f,
-0.4630523725490196f, 0.8363703921568627f, 0.895809431372549f,
-0.45690092156862744f, 0.8305265490196079f, 0.8933488039215687f,
-0.4507494705882353f, 0.8246827058823529f, 0.8908881764705883f,
-0.4445980196078431f, 0.818838862745098f, 0.8884275490196079f,
-0.438446568627451f, 0.8129950196078432f, 0.8859669215686274f,
-0.43229511764705886f, 0.8071511764705883f, 0.8835062941176471f,
-0.4261436666666667f, 0.8013073333333334f, 0.8810456666666667f,
-0.41999221568627454f, 0.7954634901960784f, 0.8785850392156863f,
-0.4138407647058824f, 0.7896196470588236f, 0.8761244117647059f,
-0.4076893137254903f, 0.7837758039215686f, 0.8736637843137255f,
-0.40153786274509806f, 0.7779319607843137f, 0.8712031568627451f,
-0.39653976470588237f, 0.772318705882353f, 0.8694345882352942f,
-0.3919261176470588f, 0.7667823137254902f, 0.867896705882353f,
-0.3873124705882353f, 0.7612459215686275f, 0.8663588235294117f,
-0.38269882352941176f, 0.7557095294117647f, 0.8648209411764706f,
-0.37808517647058826f, 0.750173137254902f, 0.8632830588235294f,
-0.3734715294117647f, 0.7446367450980392f, 0.8617451764705882f,
-0.3688578823529412f, 0.7391003529411765f, 0.8602072941176471f,
-0.36424423529411765f, 0.7335639607843137f, 0.8586694117647059f,
-0.3596305882352941f, 0.728027568627451f, 0.8571315294117647f,
-0.3550169411764706f, 0.7224911764705882f, 0.8555936470588235f,
-0.3504032941176471f, 0.7169547843137255f, 0.8540557647058823f,
-0.34578964705882353f, 0.7114183921568628f, 0.8525178823529411f,
-0.341176f, 0.705882f, 0.85098f,
-0.336562431372549f, 0.7006532941176471f, 0.8494421960784313f,
-0.33194886274509805f, 0.6954245882352942f, 0.8479043921568628f,
-0.32733529411764706f, 0.6901958823529412f, 0.8463665882352941f,
-0.32272172549019607f, 0.6849671764705882f, 0.8448287843137254f,
-0.3181081568627451f, 0.6797384705882353f, 0.8432909803921569f,
-0.31349458823529414f, 0.6745097647058824f, 0.8417531764705882f,
-0.30888101960784314f, 0.6692810588235294f, 0.8402153725490196f,
-0.30426745098039215f, 0.6640523529411765f, 0.838677568627451f,
-0.29965388235294116f, 0.6588236470588236f, 0.8371397647058824f,
-0.2950403137254902f, 0.6535949411764707f, 0.8356019607843137f,
-0.29042674509803923f, 0.6483662352941176f, 0.8340641568627452f,
-0.28581317647058824f, 0.6431375294117647f, 0.8325263529411765f,
-0.2812764901960785f, 0.6379857058823529f, 0.8309885294117647f,
-0.2769704509803922f, 0.6330645294117647f, 0.8294506470588235f,
-0.27266441176470596f, 0.6281433529411765f, 0.8279127647058824f,
-0.2683583725490196f, 0.6232221764705882f, 0.8263748823529412f,
-0.26405233333333333f, 0.618301f, 0.824837f,
-0.25974629411764705f, 0.6133798235294118f, 0.8232991176470589f,
-0.25544025490196076f, 0.6084586470588235f, 0.8217612352941176f,
-0.25113421568627453f, 0.6035374705882353f, 0.8202233529411764f,
-0.24682817647058825f, 0.598616294117647f, 0.8186854705882353f,
-0.24252213725490196f, 0.5936951176470588f, 0.8171475882352941f,
-0.23821609803921573f, 0.5887739411764706f, 0.8156097058823529f,
-0.2339100588235294f, 0.5838527647058823f, 0.8140718235294118f,
-0.22960401960784313f, 0.5789315882352941f, 0.8125339411764706f,
-0.22529798039215684f, 0.5735490392156862f, 0.8108422549019607f,
-0.22099194117647059f, 0.5677051176470588f, 0.8089967647058823f,
-0.2166859019607843f, 0.5618611960784313f, 0.8071512745098038f,
-0.21237986274509804f, 0.5560172745098039f, 0.8053057843137255f,
-0.20807382352941176f, 0.5501733529411764f, 0.8034602941176471f,
-0.20376778431372555f, 0.5443294313725491f, 0.8016148039215686f,
-0.1994617450980392f, 0.5384855098039215f, 0.7997693137254902f,
-0.19515570588235293f, 0.5326415882352941f, 0.7979238235294117f,
-0.19084966666666667f, 0.5267976666666666f, 0.7960783333333333f,
-0.1865436274509804f, 0.5209537450980392f, 0.794232843137255f,
-0.18223758823529412f, 0.5151098235294117f, 0.7923873529411765f,
-0.17793154901960784f, 0.5092659019607844f, 0.7905418627450981f,
-0.17362550980392158f, 0.5034219803921569f, 0.7886963725490196f,
-0.1695501764705883f, 0.49803941176470595f, 0.7870815882352942f,
-0.16555174509803922f, 0.49281062745098037f, 0.785543705882353f,
-0.1615533137254902f, 0.4875818431372549f, 0.7840058235294117f,
-0.15755488235294118f, 0.4823530588235294f, 0.7824679411764706f,
-0.15355645098039217f, 0.47712427450980394f, 0.7809300588235294f,
-0.14955801960784315f, 0.47189549019607846f, 0.7793921764705882f,
-0.1455595882352941f, 0.46666670588235293f, 0.7778542941176471f,
-0.1415611568627451f, 0.46143792156862745f, 0.7763164117647059f,
-0.13756272549019613f, 0.45620913725490203f, 0.7747785294117647f,
-0.13356429411764706f, 0.4509803529411765f, 0.7732406470588234f,
-0.12956586274509804f, 0.44575156862745097f, 0.7717027647058823f,
-0.12556743137254903f, 0.4405227843137255f, 0.7701648823529411f,
-0.121569f, 0.435294f, 0.768627f,
-0.1191083725490196f, 0.42975768627450983f, 0.7670891960784313f,
-0.11664774509803921f, 0.4242213725490196f, 0.7655513921568627f,
-0.11418711764705883f, 0.4186850588235294f, 0.7640135882352941f,
-0.11172649019607847f, 0.4131487450980393f, 0.7624757843137255f,
-0.10926586274509803f, 0.40761243137254904f, 0.7609379803921569f,
-0.10680523529411765f, 0.4020761176470588f, 0.7594001764705882f,
-0.10434460784313725f, 0.3965398039215686f, 0.7578623725490196f,
-0.10188398039215686f, 0.39100349019607844f, 0.756324568627451f,
-0.09942335294117646f, 0.3854671764705882f, 0.7547867647058824f,
-0.09696272549019608f, 0.379930862745098f, 0.7532489607843137f,
-0.09450209803921568f, 0.37439454901960784f, 0.7517111568627451f,
-0.09204147058823534f, 0.3688582352941177f, 0.7501733529411765f,
-0.0901703725490196f, 0.36437279084967317f, 0.7488918431372549f,
-0.09006786274509804f, 0.363039954248366f, 0.7483792156862745f,
-0.08996535294117647f, 0.3617071176470588f, 0.7478665882352942f,
-0.0898628431372549f, 0.3603742810457516f, 0.7473539607843137f,
-0.08976033333333333f, 0.3590414444444444f, 0.7468413333333334f,
-0.08965782352941176f, 0.35770860784313724f, 0.746328705882353f,
-0.0895553137254902f, 0.35637577124183006f, 0.7458160784313725f,
-0.08945280392156862f, 0.3550429346405229f, 0.7453034509803922f,
-0.08935029411764706f, 0.35371009803921566f, 0.7447908235294117f,
-0.08924778431372549f, 0.3523772614379085f, 0.7442781960784314f,
-0.08914527450980392f, 0.35104442483660125f, 0.743765568627451f,
-0.08904276470588235f, 0.3497115882352941f, 0.7432529411764706f,
-0.08894025490196078f, 0.3483787516339869f, 0.7427403137254902f,
-0.08883774509803921f, 0.34704591503267973f, 0.7422276862745097f,
-0.08873523529411764f, 0.3457130784313725f, 0.7417150588235294f,
-0.08863272549019607f, 0.3443802418300653f, 0.741202431372549f,
-0.0885302156862745f, 0.34304740522875815f, 0.7406898039215686f,
-0.08842770588235295f, 0.341714568627451f, 0.7401771764705882f,
-0.08832519607843137f, 0.34038173202614375f, 0.7396645490196079f,
-0.0882226862745098f, 0.33904889542483657f, 0.7391519215686274f,
-0.08812017647058824f, 0.3377160588235294f, 0.7386392941176471f,
-0.08801766666666667f, 0.3363832222222222f, 0.7381266666666666f,
-0.0879151568627451f, 0.335050385620915f, 0.7376140392156862f,
-0.08781264705882354f, 0.3337175490196078f, 0.7371014117647059f,
-0.08771013725490197f, 0.33238471241830064f, 0.7365887843137254f,
-0.0876076274509804f, 0.33105187581699347f, 0.7360761568627451f,
-0.08750511764705883f, 0.32971903921568624f, 0.7355635294117647f,
-0.08740260784313726f, 0.32838620261437906f, 0.7350509019607843f,
-0.08730009803921569f, 0.3270533660130719f, 0.7345382745098039f,
-0.08719758823529412f, 0.32572052941176466f, 0.7340256470588234f,
-0.08709507843137256f, 0.3243876928104575f, 0.7335130196078431f,
-0.08699256862745099f, 0.3230548562091503f, 0.7330003921568627f,
-0.08689005882352942f, 0.32172201960784313f, 0.7324877647058823f,
-0.08678754901960785f, 0.32038918300653596f, 0.7319751372549019f,
-0.08668503921568628f, 0.31905634640522873f, 0.7314625098039216f,
-0.08658252941176471f, 0.31772350980392156f, 0.7309498823529411f,
-0.08648001960784314f, 0.3163906732026144f, 0.7304372549019608f,
-0.08637750980392157f, 0.31505783660130715f, 0.7299246274509803f,
-0.086275f, 0.313725f, 0.729412f,
-0.08565978431372549f, 0.30941903921568625f, 0.7278741176470588f,
-0.08504456862745098f, 0.30511307843137253f, 0.7263362352941176f,
-0.08442935294117647f, 0.3008071176470588f, 0.7247983529411764f,
-0.08381413725490196f, 0.2965011568627451f, 0.7232604705882353f,
-0.08319892156862746f, 0.29219519607843136f, 0.7217225882352941f,
-0.08258370588235295f, 0.28788923529411764f, 0.720184705882353f,
-0.08196849019607844f, 0.2835832745098039f, 0.7186468235294118f,
-0.08135327450980392f, 0.2792773137254902f, 0.7171089411764706f,
-0.08073805882352941f, 0.2749713529411765f, 0.7155710588235293f,
-0.0801228431372549f, 0.27066539215686275f, 0.7140331764705882f,
-0.07950762745098042f, 0.26635943137254914f, 0.7124952941176471f,
-0.07889241176470588f, 0.2620534705882353f, 0.7109574117647058f,
-0.07850790196078432f, 0.25790127450980393f, 0.7091888431372549f,
-0.07881550980392157f, 0.2542103725490196f, 0.7067282156862745f,
-0.07912311764705883f, 0.2505194705882353f, 0.7042675882352941f,
-0.07943072549019607f, 0.24682856862745098f, 0.7018069607843137f,
-0.07973833333333333f, 0.24313766666666667f, 0.6993463333333333f,
-0.08004594117647058f, 0.23944676470588236f, 0.696885705882353f,
-0.08035354901960784f, 0.23575586274509805f, 0.6944250784313726f,
-0.0806611568627451f, 0.23206496078431374f, 0.6919644509803922f,
-0.08096876470588235f, 0.2283740588235294f, 0.6895038235294118f,
-0.08127637254901961f, 0.2246831568627451f, 0.6870431960784313f,
-0.08158398039215686f, 0.22099225490196078f, 0.6845825686274509f,
-0.08189158823529412f, 0.21730135294117647f, 0.6821219411764706f,
-0.08219919607843137f, 0.21361045098039216f, 0.6796613137254902f,
-0.08266056862745097f, 0.20991954901960785f, 0.6768931568627451f,
-0.08327570588235292f, 0.20622864705882366f, 0.6738174705882354f,
-0.0838908431372549f, 0.20253774509803923f, 0.6707417843137254f,
-0.08450598039215686f, 0.1988468431372549f, 0.6676660980392157f,
-0.08512111764705882f, 0.1951559411764706f, 0.6645904117647059f,
-0.08573625490196078f, 0.19146503921568628f, 0.6615147254901961f,
-0.08635139215686274f, 0.18777413725490197f, 0.6584390392156863f,
-0.0869665294117647f, 0.18408323529411763f, 0.6553633529411764f,
-0.08758166666666667f, 0.18039233333333332f, 0.6522876666666667f,
-0.08819680392156863f, 0.176701431372549f, 0.6492119803921569f,
-0.08881194117647058f, 0.1730105294117647f, 0.6461362941176471f,
-0.08942707843137254f, 0.1693196274509804f, 0.6430606078431372f,
-0.0900422156862745f, 0.16562872549019608f, 0.6399849215686274f,
-0.0908880588235294f, 0.1628605294117647f, 0.6369091764705882f,
-0.09181080392156862f, 0.1603999019607843f, 0.6338334117647059f,
-0.09273354901960784f, 0.1579392745098039f, 0.6307576470588235f,
-0.09365629411764706f, 0.15547864705882353f, 0.6276818823529412f,
-0.09457903921568625f, 0.1530180196078432f, 0.6246061176470589f,
-0.09550178431372548f, 0.15055739215686276f, 0.6215303529411764f,
-0.0964245294117647f, 0.14809676470588234f, 0.6184545882352941f,
-0.09734727450980392f, 0.14563613725490196f, 0.6153788235294118f,
-0.09827001960784314f, 0.14317550980392157f, 0.6123030588235294f,
-0.09919276470588234f, 0.1407148823529412f, 0.6092272941176471f,
-0.10011550980392156f, 0.13825425490196078f, 0.6061515294117646f,
-0.10103825490196078f, 0.1357936274509804f, 0.6030757647058823f,
-0.101961f, 0.133333f, 0.6f,
-0.10257613725490196f, 0.13148758823529413f, 0.5960015686274509f,
-0.10319127450980392f, 0.12964217647058823f, 0.592003137254902f,
-0.10380641176470588f, 0.12779676470588236f, 0.5880047058823529f,
-0.10442154901960785f, 0.1259513529411765f, 0.5840062745098039f,
-0.1050366862745098f, 0.12410594117647059f, 0.5800078431372548f,
-0.10565182352941176f, 0.12226052941176471f, 0.5760094117647059f,
-0.10626696078431372f, 0.12041511764705883f, 0.5720109803921568f,
-0.10688209803921567f, 0.118569705882353f, 0.568012549019608f,
-0.10749723529411764f, 0.11672429411764706f, 0.5640141176470588f,
-0.1081123725490196f, 0.11487888235294118f, 0.5600156862745098f,
-0.10872750980392157f, 0.11303347058823529f, 0.5560172549019607f,
-0.10934264705882353f, 0.11118805882352942f, 0.5520188235294117f,
-0.11064982352941176f, 0.10988088235294118f, 0.5480972745098038f,
-0.11403311764705883f, 0.11018841176470588f, 0.5444063725490196f,
-0.11741641176470588f, 0.11049594117647059f, 0.5407154705882352f,
-0.12079970588235293f, 0.1108034705882353f, 0.537024568627451f,
-0.124183f, 0.111111f, 0.5333336666666666f,
-0.12756629411764706f, 0.11141852941176471f, 0.5296427647058823f,
-0.13094958823529412f, 0.11172605882352941f, 0.525951862745098f,
-0.13433288235294116f, 0.11203358823529412f, 0.5222609607843137f,
-0.13771617647058823f, 0.11234111764705883f, 0.5185700588235294f,
-0.1410994705882353f, 0.11264864705882353f, 0.5148791568627451f,
-0.14448276470588234f, 0.11295617647058824f, 0.5111882549019607f,
-0.14786605882352932f, 0.11326370588235293f, 0.5074973529411766f,
-0.15124935294117647f, 0.11357123529411765f, 0.5038064509803921f,
-0.1534023725490196f, 0.11357123529411765f, 0.4987314705882353f,
-0.1543251176470588f, 0.11326370588235295f, 0.4922724117647059f,
-0.15524786274509803f, 0.11295617647058824f, 0.48581335294117645f,
-0.15617060784313724f, 0.11264864705882353f, 0.4793542941176471f,
-0.15709335294117646f, 0.11234111764705883f, 0.47289523529411764f,
-0.15801609803921568f, 0.11203358823529412f, 0.4664361764705882f,
-0.1589388431372549f, 0.11172605882352941f, 0.45997711764705884f,
-0.15986158823529412f, 0.11141852941176471f, 0.4535180588235294f,
-0.16078433333333333f, 0.111111f, 0.447059f,
-0.16170707843137255f, 0.1108034705882353f, 0.44059994117647056f,
-0.16262982352941174f, 0.11049594117647059f, 0.4341408823529412f,
-0.16355256862745096f, 0.11018841176470588f, 0.42768182352941175f,
-0.16447531372549018f, 0.10988088235294118f, 0.4212227647058823f,
-0.16262988235294118f, 0.10680517647058824f, 0.4126875882352941f,
-0.15986172549019614f, 0.10280674509803933f, 0.40346037254901984f,
-0.15709356862745097f, 0.09880831372549019f, 0.39423315686274507f,
-0.15432541176470588f, 0.09480988235294117f, 0.3850059411764706f,
-0.1515572549019608f, 0.09081145098039216f, 0.3757787254901961f,
-0.1487890980392157f, 0.08681301960784313f, 0.36655150980392154f,
-0.14602094117647058f, 0.08281458823529411f, 0.35732429411764705f,
-0.14325278431372548f, 0.0788161568627451f, 0.34809707843137255f,
-0.1404846274509804f, 0.07481772549019608f, 0.338869862745098f,
-0.1377164705882353f, 0.07081929411764706f, 0.3296426470588235f,
-0.13494831372549018f, 0.06682086274509805f, 0.320415431372549f,
-0.1321801568627451f, 0.06282243137254903f, 0.3111882156862745f,
-0.129412f, 0.058824f, 0.301961f
-}
-, 
-{
-0.94902f, 1.0f, 0.980392f,
-0.9471745490196078f, 1.0f, 0.9791617254901961f,
-0.9453290980392156f, 1.0f, 0.9779314509803922f,
-0.9434836470588235f, 1.0f, 0.9767011764705883f,
-0.9416381960784314f, 1.0f, 0.9754709019607843f,
-0.9397927450980392f, 1.0f, 0.9742406274509804f,
-0.937947294117647f, 1.0f, 0.9730103529411765f,
-0.9361018431372549f, 1.0f, 0.9717800784313726f,
-0.9342563921568627f, 1.0f, 0.9705498039215686f,
-0.9324109411764706f, 1.0f, 0.9693195294117647f,
-0.9305654901960784f, 1.0f, 0.9680892549019608f,
-0.9287200392156862f, 1.0f, 0.9668589803921569f,
-0.9268745882352941f, 1.0f, 0.965628705882353f,
-0.925029137254902f, 1.0f, 0.964398431372549f,
-0.9231836862745098f, 1.0f, 0.9631681568627452f,
-0.9213382352941176f, 1.0f, 0.9619378823529412f,
-0.9194927843137255f, 1.0f, 0.9607076078431372f,
-0.9176473333333334f, 1.0f, 0.9594773333333333f,
-0.9158018823529411f, 1.0f, 0.9582470588235295f,
-0.913956431372549f, 1.0f, 0.9570167843137255f,
-0.9121109803921569f, 1.0f, 0.9557865098039215f,
-0.9102655294117648f, 1.0f, 0.9545562352941176f,
-0.9084200784313725f, 1.0f, 0.9533259607843138f,
-0.9065746274509804f, 1.0f, 0.9520956862745098f,
-0.9047291764705883f, 1.0f, 0.9508654117647058f,
-0.9028837254901961f, 1.0f, 0.9496351372549019f,
-0.8991928431372549f, 0.9992310588235294f, 0.9477896862745098f,
-0.8936565294117648f, 0.9976931764705882f, 0.9453290588235294f,
-0.8881202156862745f, 0.996155294117647f, 0.9428684313725489f,
-0.8825839019607843f, 0.9946174117647059f, 0.9404078039215686f,
-0.8770475882352942f, 0.9930795294117647f, 0.9379471764705882f,
-0.8715112745098039f, 0.9915416470588235f, 0.9354865490196078f,
-0.8659749607843138f, 0.9900037647058824f, 0.9330259215686274f,
-0.8604386470588236f, 0.9884658823529412f, 0.9305652941176471f,
-0.8549023333333333f, 0.986928f, 0.9281046666666667f,
-0.8493660196078432f, 0.9853901176470589f, 0.9256440392156863f,
-0.843829705882353f, 0.9838522352941177f, 0.9231834117647059f,
-0.8382933921568628f, 0.9823143529411765f, 0.9207227843137255f,
-0.8327570784313726f, 0.9807764705882354f, 0.918262156862745f,
-0.8276820588235294f, 0.9792385882352942f, 0.9162629411764706f,
-0.8227608039215687f, 0.977700705882353f, 0.9144175294117647f,
-0.8178395490196079f, 0.9761628235294117f, 0.9125721176470588f,
-0.812918294117647f, 0.9746249411764706f, 0.9107267058823529f,
-0.8079970392156863f, 0.9730870588235294f, 0.9088812941176471f,
-0.8030757843137255f, 0.9715491764705882f, 0.9070358823529412f,
-0.7981545294117647f, 0.9700112941176471f, 0.9051904705882353f,
-0.7932332745098039f, 0.9684734117647059f, 0.9033450588235294f,
-0.7883120196078431f, 0.9669355294117647f, 0.9014996470588235f,
-0.7833907647058823f, 0.9653976470588235f, 0.8996542352941176f,
-0.7784695098039216f, 0.9638597647058823f, 0.8978088235294117f,
-0.7735482549019608f, 0.9623218823529411f, 0.8959634117647058f,
-0.768627f, 0.960784f, 0.894118f,
-0.7630906862745097f, 0.9583234509803922f, 0.8916573725490196f,
-0.7575543725490196f, 0.9558629019607843f, 0.8891967450980393f,
-0.7520180588235293f, 0.9534023529411765f, 0.8867361176470588f,
-0.7464817450980392f, 0.9509418039215686f, 0.8842754901960784f,
-0.740945431372549f, 0.9484812549019608f, 0.881814862745098f,
-0.7354091176470589f, 0.946020705882353f, 0.8793542352941176f,
-0.7298728039215686f, 0.9435601568627451f, 0.8768936078431372f,
-0.7243364901960784f, 0.9410996078431373f, 0.8744329803921569f,
-0.7188001764705882f, 0.9386390588235294f, 0.8719723529411765f,
-0.713263862745098f, 0.9361785098039216f, 0.8695117254901961f,
-0.7077275490196078f, 0.9337179607843137f, 0.8670510980392157f,
-0.7021912352941176f, 0.9312574117647059f, 0.8645904705882352f,
-0.6968856078431372f, 0.9288737450980392f, 0.8623605294117647f,
-0.6922720392156863f, 0.9267207254901961f, 0.8608226470588235f,
-0.6876584705882354f, 0.924567705882353f, 0.8592847647058823f,
-0.6830449019607843f, 0.9224146862745098f, 0.8577468823529412f,
-0.6784313333333333f, 0.9202616666666666f, 0.856209f,
-0.6738177647058823f, 0.9181086470588236f, 0.8546711176470588f,
-0.6692041960784314f, 0.9159556274509804f, 0.8531332352941177f,
-0.6645906274509804f, 0.9138026078431373f, 0.8515953529411765f,
-0.6599770588235294f, 0.9116495882352942f, 0.8500574705882353f,
-0.6553634901960784f, 0.909496568627451f, 0.8485195882352942f,
-0.6507499215686275f, 0.9073435490196079f, 0.846981705882353f,
-0.6461363529411764f, 0.9051905294117647f, 0.8454438235294118f,
-0.6415227843137254f, 0.9030375098039216f, 0.8439059411764707f,
-0.6372167450980393f, 0.9007306862745098f, 0.8425218627450981f,
-0.6332182352941177f, 0.8982700588235294f, 0.8412915882352942f,
-0.6292197254901961f, 0.895809431372549f, 0.8400613137254902f,
-0.6252212156862745f, 0.8933488039215687f, 0.8388310392156864f,
-0.621222705882353f, 0.8908881764705883f, 0.8376007647058824f,
-0.6172241960784314f, 0.8884275490196079f, 0.8363704901960785f,
-0.6132256862745098f, 0.8859669215686274f, 0.8351402156862745f,
-0.6092271764705882f, 0.883506294117647f, 0.8339099411764707f,
-0.6052286666666666f, 0.8810456666666667f, 0.8326796666666667f,
-0.6012301568627451f, 0.8785850392156863f, 0.8314493921568628f,
-0.5972316470588235f, 0.8761244117647059f, 0.8302191176470589f,
-0.5932331372549019f, 0.8736637843137255f, 0.828988843137255f,
-0.5892346274509803f, 0.8712031568627451f, 0.827758568627451f,
-0.5854668235294118f, 0.8682812352941177f, 0.8260669411764706f,
-0.5817759215686275f, 0.8652055490196079f, 0.8242215294117647f,
-0.5780850196078431f, 0.8621298627450981f, 0.8223761176470589f,
-0.5743941176470588f, 0.8590541764705882f, 0.820530705882353f,
-0.5707032156862745f, 0.8559784901960784f, 0.8186852941176471f,
-0.5670123137254902f, 0.8529028039215687f, 0.8168398823529412f,
-0.5633214117647058f, 0.8498271176470589f, 0.8149944705882354f,
-0.5596305098039216f, 0.846751431372549f, 0.8131490588235295f,
-0.5559396078431373f, 0.8436757450980393f, 0.8113036470588236f,
-0.552248705882353f, 0.8406000588235294f, 0.8094582352941176f,
-0.5485578039215686f, 0.8375243725490197f, 0.8076128235294118f,
-0.5448669019607844f, 0.8344486862745099f, 0.8057674117647059f,
-0.541176f, 0.831373f, 0.803922f,
-0.5381003137254902f, 0.8289123725490196f, 0.8033067843137255f,
-0.5350246274509803f, 0.8264517450980392f, 0.802691568627451f,
-0.5319489411764706f, 0.8239911176470589f, 0.8020763529411765f,
-0.5288732549019608f, 0.8215304901960785f, 0.801461137254902f,
-0.525797568627451f, 0.8190698627450981f, 0.8008459215686274f,
-0.5227218823529411f, 0.8166092352941177f, 0.8002307058823529f,
-0.5196461960784313f, 0.8141486078431373f, 0.7996154901960785f,
-0.5165705098039216f, 0.811687980392157f, 0.7990002745098039f,
-0.5134948235294118f, 0.8092273529411765f, 0.7983850588235294f,
-0.510419137254902f, 0.8067667254901961f, 0.7977698431372549f,
-0.5073434509803921f, 0.8043060980392157f, 0.7971546274509803f,
-0.5042677647058824f, 0.8018454705882354f, 0.7965394117647059f,
-0.5010382745098039f, 0.7989234901960784f, 0.7953859607843137f,
-0.4973473725490196f, 0.7946174509803922f, 0.7926178039215686f,
-0.4936564705882353f, 0.7903114117647059f, 0.7898496470588235f,
-0.48996556862745094f, 0.7860053725490197f, 0.7870814901960784f,
-0.48627466666666663f, 0.7816993333333334f, 0.7843133333333333f,
-0.4825837647058823f, 0.7773932941176471f, 0.7815451764705882f,
-0.478892862745098f, 0.7730872549019608f, 0.7787770196078431f,
-0.47520196078431376f, 0.7687812156862747f, 0.7760088627450981f,
-0.4715110588235294f, 0.7644751764705883f, 0.7732407058823529f,
-0.4678201568627451f, 0.760169137254902f, 0.7704725490196078f,
-0.4641292549019608f, 0.7558630980392157f, 0.7677043921568627f,
-0.46043835294117647f, 0.7515570588235294f, 0.7649362352941177f,
-0.45674745098039216f, 0.7472510196078431f, 0.7621680784313726f,
-0.4535179215686274f, 0.7430987843137256f, 0.7595537254901961f,
-0.45074976470588235f, 0.7391003529411765f, 0.7570931764705883f,
-0.44798160784313723f, 0.7351019215686275f, 0.7546326274509804f,
-0.4452134509803921f, 0.7311034901960785f, 0.7521720784313726f,
-0.4424452941176471f, 0.7271050588235296f, 0.7497115294117648f,
-0.43967713725490193f, 0.7231066274509804f, 0.7472509803921569f,
-0.43690898039215686f, 0.7191081960784313f, 0.744790431372549f,
-0.43414082352941175f, 0.7151097647058824f, 0.7423298823529412f,
-0.4313726666666666f, 0.7111113333333333f, 0.7398693333333333f,
-0.42860450980392156f, 0.7071129019607844f, 0.7374087843137255f,
-0.42583635294117644f, 0.7031144705882353f, 0.7349482352941176f,
-0.4230681960784314f, 0.6991160392156863f, 0.7324876862745098f,
-0.42030003921568626f, 0.6951176078431373f, 0.7300271372549019f,
-0.41730117647058823f, 0.6904271176470589f, 0.7271051764705881f,
-0.41422541176470584f, 0.6855059411764706f, 0.7240294117647058f,
-0.4111496470588235f, 0.6805847647058824f, 0.7209536470588235f,
-0.4080738823529412f, 0.6756635882352942f, 0.7178778823529411f,
-0.40499811764705884f, 0.6707424117647058f, 0.7148021176470588f,
-0.40192235294117645f, 0.6658212352941176f, 0.7117263529411765f,
-0.3988465882352941f, 0.6609000588235294f, 0.7086505882352941f,
-0.39577082352941184f, 0.6559788823529413f, 0.7055748235294118f,
-0.3926950588235294f, 0.6510577058823529f, 0.7024990588235294f,
-0.38961929411764706f, 0.6461365294117647f, 0.699423294117647f,
-0.38654352941176473f, 0.6412153529411765f, 0.6963475294117647f,
-0.38346776470588234f, 0.6362941764705882f, 0.6932717647058824f,
-0.380392f, 0.631373f, 0.690196f,
-0.3776238431372549f, 0.6273744901960784f, 0.6877354509803922f,
-0.3748556862745098f, 0.6233759803921568f, 0.6852749019607843f,
-0.3720875294117647f, 0.6193774705882352f, 0.6828143529411765f,
-0.36931937254901964f, 0.6153789607843138f, 0.6803538039215686f,
-0.3665512156862745f, 0.6113804509803922f, 0.6778932549019608f,
-0.3637830588235294f, 0.6073819411764706f, 0.6754327058823529f,
-0.36101490196078434f, 0.603383431372549f, 0.6729721568627451f,
-0.3582467450980392f, 0.5993849215686274f, 0.6705116078431372f,
-0.3554785882352941f, 0.5953864117647059f, 0.6680510588235294f,
-0.35271043137254904f, 0.5913879019607843f, 0.6655905098039215f,
-0.34994227450980403f, 0.5873893921568629f, 0.6631299607843137f,
-0.34717411764705886f, 0.5833908823529412f, 0.6606694117647058f,
-0.3445597450980392f, 0.5793154901960784f, 0.6580550588235293f,
-0.3424067254901961f, 0.5750094509803921f, 0.654979294117647f,
-0.34025370588235293f, 0.5707034117647058f, 0.6519035294117647f,
-0.33810068627450984f, 0.5663973725490197f, 0.6488277647058823f,
-0.3359476666666667f, 0.5620913333333334f, 0.645752f,
-0.33379464705882356f, 0.5577852941176471f, 0.6426762352941177f,
-0.3316416274509804f, 0.5534792549019608f, 0.6396004705882353f,
-0.32948860784313727f, 0.5491732156862745f, 0.636524705882353f,
-0.32733558823529413f, 0.5448671764705882f, 0.6334489411764707f,
-0.325182568627451f, 0.540561137254902f, 0.6303731764705882f,
-0.32302954901960784f, 0.5362550980392157f, 0.6272974117647059f,
-0.3208765294117647f, 0.5319490588235294f, 0.6242216470588235f,
-0.31872350980392156f, 0.5276430196078431f, 0.6211458823529412f,
-0.3171856274509804f, 0.5233369803921569f, 0.6180701176470589f,
-0.3162628823529412f, 0.5190309411764708f, 0.6149943529411765f,
-0.31534013725490195f, 0.5147249019607844f, 0.6119185882352942f,
-0.31441739215686276f, 0.5104188627450981f, 0.6088428235294118f,
-0.3134946470588235f, 0.5061128235294118f, 0.6057670588235294f,
-0.3125719019607843f, 0.5018067843137255f, 0.6026912941176471f,
-0.3116491568627451f, 0.4975007450980392f, 0.5996155294117648f,
-0.3107264117647059f, 0.4931947058823529f, 0.5965397647058824f,
-0.30980366666666664f, 0.4888886666666667f, 0.593464f,
-0.30888092156862745f, 0.4845826274509804f, 0.5903882352941177f,
-0.3079581764705882f, 0.4802765882352941f, 0.5873124705882353f,
-0.307035431372549f, 0.47597054901960784f, 0.584236705882353f,
-0.30611268627450977f, 0.4716645098039216f, 0.5811609411764707f,
-0.30495929411764705f, 0.4673584705882353f, 0.5773931764705883f,
-0.30372901960784315f, 0.46305243137254903f, 0.5733947450980392f,
-0.3024987450980392f, 0.45874639215686275f, 0.5693963137254902f,
-0.3012684705882353f, 0.45444035294117646f, 0.5653978823529412f,
-0.3000381960784314f, 0.45013431372549034f, 0.5613994509803922f,
-0.29880792156862745f, 0.44582827450980395f, 0.5574010196078432f,
-0.2975776470588235f, 0.44152223529411766f, 0.5534025882352941f,
-0.2963473725490196f, 0.4372161960784314f, 0.5494041568627451f,
-0.2951170980392157f, 0.4329101568627451f, 0.5454057254901961f,
-0.29388682352941176f, 0.42860411764705886f, 0.5414072941176471f,
-0.29265654901960786f, 0.42429807843137257f, 0.537408862745098f,
-0.2914262745098039f, 0.4199920392156863f, 0.5334104313725491f,
-0.290196f, 0.415686f, 0.529412f,
-0.2892732549019608f, 0.41076482352941174f, 0.5247983529411765f,
-0.2883505098039216f, 0.40584364705882353f, 0.5201847058823529f,
-0.2874277647058824f, 0.40092247058823527f, 0.5155710588235294f,
-0.28650501960784314f, 0.39600129411764706f, 0.5109574117647059f,
-0.28558227450980395f, 0.3910801176470588f, 0.5063437647058824f,
-0.2846595294117647f, 0.3861589411764706f, 0.5017301176470588f,
-0.2837367843137255f, 0.38123776470588233f, 0.49711647058823527f,
-0.2828140392156863f, 0.37631658823529424f, 0.4925028235294119f,
-0.2818912941176471f, 0.37139541176470586f, 0.4878891764705882f,
-0.2809685490196078f, 0.36647423529411766f, 0.4832755294117647f,
-0.28004580392156864f, 0.3615530588235294f, 0.47866188235294116f,
-0.2791230588235294f, 0.3566318823529412f, 0.47404823529411766f,
-0.27804654901960785f, 0.35186449019607846f, 0.46920392156862745f,
-0.2765087450980392f, 0.3475584509803922f, 0.46366760784313726f,
-0.2749709411764706f, 0.3432524117647059f, 0.4581312941176471f,
-0.2734331372549019f, 0.3389463725490196f, 0.4525949803921569f,
-0.2718953333333333f, 0.3346403333333333f, 0.44705866666666666f,
-0.2703575294117647f, 0.3303342941176471f, 0.4415223529411765f,
-0.26881972549019606f, 0.3260282549019608f, 0.4359860392156863f,
-0.26728192156862746f, 0.3217222156862745f, 0.4304497254901961f,
-0.2657441176470588f, 0.31741617647058823f, 0.4249134117647059f,
-0.2642063137254902f, 0.31311013725490194f, 0.4193770980392157f,
-0.26266850980392153f, 0.3088040980392157f, 0.4138407843137255f,
-0.261130705882353f, 0.30449805882352954f, 0.4083044705882355f,
-0.25959290196078433f, 0.30019201960784314f, 0.40276815686274514f,
-0.2579012549019608f, 0.29603978431372546f, 0.3972318431372549f,
-0.25605576470588237f, 0.29204135294117645f, 0.3916955294117647f,
-0.25421027450980394f, 0.28804292156862743f, 0.38615921568627454f,
-0.2523647843137255f, 0.2840444901960784f, 0.3806229019607843f,
-0.25051929411764706f, 0.2800460588235294f, 0.3750865882352941f,
-0.24867380392156863f, 0.2760476274509804f, 0.36955027450980393f,
-0.2468283137254902f, 0.27204919607843137f, 0.36401396078431375f,
-0.24498282352941175f, 0.26805076470588235f, 0.3584776470588235f,
-0.24313733333333334f, 0.26405233333333333f, 0.35294133333333333f,
-0.2412918431372549f, 0.2600539019607843f, 0.34740501960784315f,
-0.23944635294117647f, 0.2560554705882353f, 0.34186870588235296f,
-0.23760086274509803f, 0.2520570392156863f, 0.3363323921568627f,
-0.23575537254901963f, 0.24805860784313727f, 0.33079607843137254f,
-0.2320644705882353f, 0.24290676470588235f, 0.32364499999999996f,
-0.22775843137254914f, 0.2373704509803923f, 0.31595566666666686f,
-0.22345239215686274f, 0.23183413725490196f, 0.3082663333333333f,
-0.21914635294117646f, 0.22629782352941177f, 0.300577f,
-0.2148403137254902f, 0.22076150980392156f, 0.29288766666666666f,
-0.21053427450980391f, 0.21522519607843138f, 0.28519833333333333f,
-0.20622823529411766f, 0.20968888235294117f, 0.277509f,
-0.20192219607843137f, 0.204152568627451f, 0.2698196666666667f,
-0.19761615686274508f, 0.19861625490196078f, 0.26213033333333335f,
-0.19331011764705883f, 0.1930799411764706f, 0.25444100000000003f,
-0.18900407843137254f, 0.18754362745098038f, 0.24675166666666665f,
-0.18469803921568628f, 0.18200731372549017f, 0.23906233333333332f,
-0.180392f, 0.176471f, 0.231373f
-}
-, 
-{
-0.960784f, 0.988235f, 1.0f,
-0.9530946666666666f, 0.9845440980392156f, 0.9984621176470588f,
-0.9454053333333333f, 0.9808531960784314f, 0.9969242352941177f,
-0.937716f, 0.977162294117647f, 0.9953863529411765f,
-0.9300266666666667f, 0.9734713921568627f, 0.9938484705882353f,
-0.9223373333333333f, 0.9697804901960784f, 0.9923105882352942f,
-0.914648f, 0.9660895882352941f, 0.990772705882353f,
-0.9069586666666667f, 0.9623986862745098f, 0.9892348235294118f,
-0.8992693333333334f, 0.9587077843137255f, 0.9876969411764707f,
-0.8915799999999999f, 0.9550168823529411f, 0.9861590588235294f,
-0.8838906666666666f, 0.9513259803921569f, 0.9846211764705882f,
-0.8762013333333333f, 0.9476350784313725f, 0.983083294117647f,
-0.868512f, 0.9439441764705883f, 0.9815454117647059f,
-0.8611302352941176f, 0.9400226078431373f, 0.9797768627450981f,
-0.8546711764705882f, 0.9354090392156863f, 0.9773163137254902f,
-0.8482121176470588f, 0.9307954705882353f, 0.9748557647058824f,
-0.8417530588235294f, 0.9261819019607843f, 0.9723952156862745f,
-0.835294f, 0.9215683333333333f, 0.9699346666666667f,
-0.8288349411764706f, 0.9169547647058823f, 0.9674741176470588f,
-0.8223758823529411f, 0.9123411960784314f, 0.965013568627451f,
-0.8159168235294117f, 0.9077276274509805f, 0.9625530196078431f,
-0.8094577647058824f, 0.9031140588235295f, 0.9600924705882353f,
-0.8029987058823529f, 0.8985004901960785f, 0.9576319215686274f,
-0.7965396470588235f, 0.8938869215686275f, 0.9551713725490196f,
-0.7900805882352941f, 0.8892733529411765f, 0.9527108235294117f,
-0.7836215294117647f, 0.8846597843137255f, 0.9502502745098039f,
-0.7790848039215686f, 0.8808920196078431f, 0.9478665882352941f,
-0.7764704117647059f, 0.8779700588235294f, 0.9455597647058823f,
-0.7738560196078431f, 0.8750480980392157f, 0.9432529411764705f,
-0.7712416274509803f, 0.872126137254902f, 0.9409461176470588f,
-0.7686272352941176f, 0.8692041764705882f, 0.938639294117647f,
-0.7660128431372549f, 0.8662822156862745f, 0.9363324705882353f,
-0.7633984509803922f, 0.8633602549019608f, 0.9340256470588235f,
-0.7607840588235294f, 0.8604382941176472f, 0.9317188235294117f,
-0.7581696666666666f, 0.8575163333333333f, 0.929412f,
-0.755555274509804f, 0.8545943725490196f, 0.9271051764705882f,
-0.7529408823529412f, 0.8516724117647059f, 0.9247983529411764f,
-0.7503264901960784f, 0.8487504509803921f, 0.9224915294117647f,
-0.7477120980392157f, 0.8458284901960784f, 0.9201847058823529f,
-0.7450977058823529f, 0.8429065294117647f, 0.9178778823529412f,
-0.7424833137254903f, 0.839984568627451f, 0.9155710588235294f,
-0.7398689215686275f, 0.8370626078431372f, 0.9132642352941176f,
-0.7372545294117647f, 0.8341406470588235f, 0.9109574117647059f,
-0.734640137254902f, 0.8312186862745098f, 0.9086505882352941f,
-0.7320257450980392f, 0.8282967254901961f, 0.9063437647058823f,
-0.7294113529411765f, 0.8253747647058823f, 0.9040369411764706f,
-0.7267969607843138f, 0.8224528039215686f, 0.9017301176470588f,
-0.724182568627451f, 0.8195308431372549f, 0.899423294117647f,
-0.7215681764705882f, 0.8166088823529412f, 0.8971164705882353f,
-0.7189537843137256f, 0.8136869215686274f, 0.8948096470588235f,
-0.7163393921568628f, 0.8107649607843137f, 0.8925028235294118f,
-0.713725f, 0.807843f, 0.890196f,
-0.7101879019607844f, 0.8044597058823529f, 0.8874278431372549f,
-0.7066508039215686f, 0.8010764117647059f, 0.8846596862745099f,
-0.703113705882353f, 0.7976931176470589f, 0.8818915294117647f,
-0.6995766078431374f, 0.7943098235294117f, 0.8791233725490196f,
-0.6960395098039216f, 0.7909265294117647f, 0.8763552156862745f,
-0.692502411764706f, 0.7875432352941176f, 0.8735870588235294f,
-0.6889653137254902f, 0.7841599411764706f, 0.8708189019607843f,
-0.6854282156862745f, 0.7807766470588235f, 0.8680507450980393f,
-0.6818911176470589f, 0.7773933529411765f, 0.8652825882352941f,
-0.6783540196078433f, 0.7740100588235295f, 0.862514431372549f,
-0.6748169215686275f, 0.7706267647058823f, 0.8597462745098039f,
-0.6712798235294118f, 0.7672434705882353f, 0.8569781176470588f,
-0.6677427254901961f, 0.7638601764705882f, 0.8542099607843138f,
-0.6642056274509804f, 0.7604768823529412f, 0.8514418039215687f,
-0.6606685294117648f, 0.7570935882352942f, 0.8486736470588235f,
-0.657131431372549f, 0.753710294117647f, 0.8459054901960784f,
-0.6535943333333334f, 0.750327f, 0.8431373333333333f,
-0.6500572352941176f, 0.7469437058823529f, 0.8403691764705883f,
-0.646520137254902f, 0.7435604117647059f, 0.8376010196078432f,
-0.6429830392156863f, 0.7401771176470588f, 0.834832862745098f,
-0.6394459411764706f, 0.7367938235294118f, 0.8320647058823529f,
-0.6359088431372549f, 0.7334105294117648f, 0.8292965490196078f,
-0.6323717450980393f, 0.7300272352941177f, 0.8265283921568628f,
-0.6288346470588235f, 0.7266439411764706f, 0.8237602352941177f,
-0.6252975490196079f, 0.7232606470588235f, 0.8209920784313726f,
-0.6217604509803921f, 0.7198004509803921f, 0.8180701176470588f,
-0.6182233529411765f, 0.7162633529411765f, 0.8149943529411765f,
-0.6146862549019608f, 0.7127262549019608f, 0.8119185882352942f,
-0.6111491568627451f, 0.7091891568627451f, 0.8088428235294117f,
-0.6076120588235294f, 0.7056520588235294f, 0.8057670588235294f,
-0.6040749607843138f, 0.7021149607843138f, 0.8026912941176471f,
-0.600537862745098f, 0.698577862745098f, 0.7996155294117647f,
-0.5970007647058824f, 0.6950407647058824f, 0.7965397647058824f,
-0.5934636666666666f, 0.6915036666666666f, 0.793464f,
-0.589926568627451f, 0.687966568627451f, 0.7903882352941176f,
-0.5863894705882353f, 0.6844294705882353f, 0.7873124705882353f,
-0.5828523725490196f, 0.6808923725490196f, 0.784236705882353f,
-0.5793152745098039f, 0.6773552745098039f, 0.7811609411764706f,
-0.5757781764705883f, 0.6738181764705883f, 0.7780851764705883f,
-0.5722410784313725f, 0.6702810784313725f, 0.7750094117647058f,
-0.5687039803921569f, 0.6667439803921569f, 0.7719336470588235f,
-0.5651668823529411f, 0.6632068823529411f, 0.7688578823529412f,
-0.5616297843137255f, 0.6596697843137255f, 0.7657821176470588f,
-0.5580926862745098f, 0.6561326862745098f, 0.7627063529411764f,
-0.554555588235294f, 0.6525955882352941f, 0.7596305882352941f,
-0.5510184901960784f, 0.6490584901960784f, 0.7565548235294117f,
-0.5474813921568628f, 0.6455213921568628f, 0.7534790588235294f,
-0.543944294117647f, 0.641984294117647f, 0.7504032941176471f,
-0.5404071960784314f, 0.6384471960784314f, 0.7473275294117646f,
-0.5368700980392156f, 0.6349100980392156f, 0.7442517647058823f,
-0.533333f, 0.631373f, 0.741176f,
-0.5302572745098039f, 0.6281434705882353f, 0.7379465098039215f,
-0.5271815490196078f, 0.6249139411764706f, 0.7347170196078431f,
-0.5241058235294117f, 0.6216844117647058f, 0.7314875294117646f,
-0.5210300980392156f, 0.6184548823529412f, 0.7282580392156862f,
-0.5179543725490195f, 0.6152253529411764f, 0.7250285490196078f,
-0.5148786470588235f, 0.6119958235294117f, 0.7217990588235293f,
-0.5118029215686274f, 0.608766294117647f, 0.7185695686274509f,
-0.5087271960784313f, 0.6055367647058824f, 0.7153400784313725f,
-0.5056514705882352f, 0.6023072352941176f, 0.712110588235294f,
-0.5025757450980392f, 0.5990777058823529f, 0.7088810980392156f,
-0.49950001960784307f, 0.5958481764705882f, 0.7056516078431372f,
-0.49642429411764705f, 0.5926186470588235f, 0.7024221176470589f,
-0.4933485686274509f, 0.5893891176470588f, 0.6991926274509803f,
-0.49027284313725483f, 0.5861595882352941f, 0.6959631372549019f,
-0.48719711764705875f, 0.5829300588235293f, 0.6927336470588235f,
-0.48412139215686273f, 0.5797005294117646f, 0.689504156862745f,
-0.48104566666666665f, 0.576471f, 0.6862746666666666f,
-0.47796994117647057f, 0.5732414705882353f, 0.6830451764705882f,
-0.4748942156862745f, 0.5700119411764706f, 0.6798156862745097f,
-0.47181849019607847f, 0.5667824117647059f, 0.6765861960784314f,
-0.46874276470588233f, 0.5635528823529411f, 0.6733567058823529f,
-0.46566703921568625f, 0.5603233529411764f, 0.6701272156862744f,
-0.46259131372549017f, 0.5570938235294117f, 0.666897725490196f,
-0.4595155882352941f, 0.553864294117647f, 0.6636682352941177f,
-0.456439862745098f, 0.5506347647058824f, 0.6604387450980391f,
-0.4535179215686274f, 0.5474821176470588f, 0.6572861176470588f,
-0.45074976470588235f, 0.5444063529411765f, 0.6542103529411765f,
-0.44798160784313723f, 0.5413305882352941f, 0.6511345882352941f,
-0.4452134509803921f, 0.5382548235294117f, 0.6480588235294117f,
-0.4424452941176471f, 0.5351790588235295f, 0.6449830588235295f,
-0.43967713725490193f, 0.532103294117647f, 0.641907294117647f,
-0.43690898039215686f, 0.5290275294117647f, 0.6388315294117647f,
-0.43414082352941175f, 0.5259517647058823f, 0.6357557647058824f,
-0.4313726666666666f, 0.522876f, 0.63268f,
-0.42860450980392156f, 0.5198002352941176f, 0.6296042352941176f,
-0.42583635294117644f, 0.5167244705882352f, 0.6265284705882352f,
-0.4230681960784314f, 0.5136487058823529f, 0.6234527058823529f,
-0.42030003921568626f, 0.5105729411764706f, 0.6203769411764706f,
-0.41753188235294114f, 0.5074971764705882f, 0.6173011764705882f,
-0.4147637254901961f, 0.5044214117647059f, 0.6142254117647059f,
-0.41199556862745096f, 0.5013456470588236f, 0.6111496470588236f,
-0.40922741176470584f, 0.4982698823529412f, 0.6080738823529411f,
-0.4064592549019608f, 0.4951941176470588f, 0.6049981176470588f,
-0.40369109803921566f, 0.49211835294117645f, 0.6019223529411765f,
-0.4009229411764706f, 0.4890425882352941f, 0.5988465882352941f,
-0.39815478431372553f, 0.48596682352941184f, 0.5957708235294119f,
-0.39538662745098035f, 0.4828910588235294f, 0.5926950588235295f,
-0.3926184705882353f, 0.47981529411764706f, 0.589619294117647f,
-0.38985031372549017f, 0.47673952941176473f, 0.5865435294117647f,
-0.38708215686274505f, 0.47366376470588234f, 0.5834677647058824f,
-0.384314f, 0.470588f, 0.580392f,
-0.3815458039215686f, 0.4676660392156863f, 0.5773162745098039f,
-0.37877760784313724f, 0.46474407843137255f, 0.5742405490196079f,
-0.37600941176470587f, 0.4618221176470588f, 0.5711648235294118f,
-0.3732412156862745f, 0.4589001568627451f, 0.5680890980392157f,
-0.3704730196078431f, 0.4559781960784314f, 0.5650133725490196f,
-0.36770482352941175f, 0.45305623529411765f, 0.5619376470588235f,
-0.3649366274509804f, 0.4501342745098039f, 0.5588619215686275f,
-0.362168431372549f, 0.4472123137254902f, 0.5557861960784314f,
-0.35940023529411763f, 0.44429035294117647f, 0.5527104705882353f,
-0.35663203921568626f, 0.44136839215686274f, 0.5496347450980392f,
-0.35386384313725494f, 0.43844643137254913f, 0.5465590196078433f,
-0.3510956470588235f, 0.4355244705882353f, 0.5434832941176471f,
-0.34832745098039214f, 0.43260250980392156f, 0.540407568627451f,
-0.34555925490196077f, 0.42968054901960784f, 0.5373318431372549f,
-0.3427910588235294f, 0.4267585882352941f, 0.5342561176470588f,
-0.340022862745098f, 0.4238366274509804f, 0.5311803921568627f,
-0.33725466666666665f, 0.42091466666666666f, 0.5281046666666667f,
-0.3344864705882353f, 0.41799270588235293f, 0.5250289411764706f,
-0.3317182745098039f, 0.4150707450980392f, 0.5219532156862745f,
-0.32895007843137253f, 0.4121487843137255f, 0.5188774901960784f,
-0.32618188235294115f, 0.40922682352941175f, 0.5158017647058823f,
-0.3234136862745098f, 0.406304862745098f, 0.5127260392156863f,
-0.3206454901960784f, 0.4033829019607843f, 0.5096503137254902f,
-0.31787729411764704f, 0.4004609411764706f, 0.5065745882352941f,
-0.31510909803921566f, 0.39753898039215685f, 0.503498862745098f,
-0.3124947058823529f, 0.39461703921568625f, 0.5003462352941176f,
-0.31003411764705885f, 0.3916951176470589f, 0.497116705882353f,
-0.3075735294117647f, 0.38877319607843136f, 0.4938871764705882f,
-0.3051129411764706f, 0.3858512745098039f, 0.49065764705882353f,
-0.30265235294117643f, 0.38292935294117647f, 0.4874281176470588f,
-0.3001917647058823f, 0.380007431372549f, 0.4841985882352941f,
-0.2977311764705882f, 0.3770855098039216f, 0.4809690588235294f,
-0.2952705882352941f, 0.3741635882352941f, 0.4777395294117647f,
-0.29280999999999996f, 0.37124166666666664f, 0.47451f,
-0.29034941176470586f, 0.3683197450980392f, 0.4712804705882353f,
-0.28788882352941175f, 0.36539782352941175f, 0.46805094117647056f,
-0.28542823529411765f, 0.3624759019607843f, 0.46482141176470587f,
-0.2829676470588235f, 0.35955398039215686f, 0.4615918823529412f,
-0.2805070588235294f, 0.3566320588235294f, 0.45836235294117644f,
-0.2780464705882353f, 0.35371013725490197f, 0.45513282352941176f,
-0.2755858823529412f, 0.3507882156862745f, 0.45190329411764707f,
-0.273125294117647f, 0.347866294117647f, 0.4486737647058823f,
-0.270664705882353f, 0.34494437254901966f, 0.4454442352941177f,
-0.2682041176470588f, 0.34202245098039213f, 0.4422147058823529f,
-0.2657435294117647f, 0.3391005294117647f, 0.4389851764705882f,
-0.26328294117647055f, 0.33617860784313724f, 0.4357556470588235f,
-0.26082235294117645f, 0.3332566862745098f, 0.43252611764705884f,
-0.25836176470588235f, 0.33033476470588236f, 0.4292965882352941f,
-0.25590117647058824f, 0.3274128431372549f, 0.4260670588235294f,
-0.2534405882352941f, 0.32449092156862747f, 0.42283752941176467f,
-0.25098f, 0.321569f, 0.419608f,
-0.24898078431372547f, 0.3191084117647059f, 0.41683984313725486f,
-0.24698156862745096f, 0.31664782352941173f, 0.4140716862745098f,
-0.24498235294117646f, 0.31418723529411763f, 0.4113035294117647f,
-0.24298313725490195f, 0.3117266470588235f, 0.4085353725490196f,
-0.24098392156862744f, 0.3092660588235294f, 0.4057672156862745f,
-0.23898470588235293f, 0.30680547058823526f, 0.4029990588235294f,
-0.23698549019607842f, 0.30434488235294116f, 0.4002309019607843f,
-0.23498627450980397f, 0.3018842941176471f, 0.3974627450980393f,
-0.2329870588235294f, 0.29942370588235295f, 0.39469458823529413f,
-0.2309878431372549f, 0.2969631176470588f, 0.391926431372549f,
-0.2289886274509804f, 0.2945025294117647f, 0.3891582745098039f,
-0.22698941176470588f, 0.2920419411764706f, 0.38639011764705883f,
-0.22499019607843138f, 0.2895813529411765f, 0.3836219607843137f,
-0.22299098039215687f, 0.2871207647058823f, 0.3808538039215686f,
-0.22099176470588236f, 0.2846601764705882f, 0.3780856470588235f,
-0.21899254901960785f, 0.2821995882352941f, 0.3753174901960784f,
-0.21699333333333334f, 0.279739f, 0.37254933333333334f,
-0.21499411764705884f, 0.27727841176470586f, 0.3697811764705882f,
-0.2129949019607843f, 0.27481782352941175f, 0.3670130196078431f,
-0.21099568627450982f, 0.27235723529411765f, 0.36424486274509804f,
-0.2089964705882353f, 0.26989664705882355f, 0.3614767058823529f,
-0.2069972549019608f, 0.2674360588235294f, 0.35870854901960786f,
-0.20499803921568627f, 0.2649754705882353f, 0.35594039215686274f,
-0.20299882352941184f, 0.26251488235294124f, 0.35317223529411773f,
-0.20099960784313725f, 0.2600542941176471f, 0.35040407843137256f,
-0.19900039215686274f, 0.2575936862745098f, 0.3476359019607843f,
-0.19700117647058824f, 0.25513305882352943f, 0.34486770588235294f,
-0.19500196078431373f, 0.252672431372549f, 0.34209950980392156f,
-0.19300274509803922f, 0.2502118039215686f, 0.3393313137254902f,
-0.1910035294117647f, 0.24775117647058822f, 0.3365631176470588f,
-0.1890043137254902f, 0.24529054901960784f, 0.33379492156862745f,
-0.1870050980392157f, 0.24282992156862745f, 0.33102672549019607f,
-0.1850058823529412f, 0.24036929411764707f, 0.3282585294117647f,
-0.18300666666666668f, 0.23790866666666666f, 0.3254903333333333f,
-0.18100745098039217f, 0.23544803921568627f, 0.32272213725490195f,
-0.17900823529411766f, 0.2329874117647059f, 0.3199539411764706f,
-0.17700901960784315f, 0.23052678431372547f, 0.3171857450980392f,
-0.17500980392156865f, 0.2280661568627451f, 0.31441754901960783f,
-0.17301058823529414f, 0.2256055294117647f, 0.31164935294117646f,
-0.17101137254901969f, 0.22314490196078438f, 0.3088811568627452f,
-0.16901215686274512f, 0.22068427450980393f, 0.3061129607843137f,
-0.1670129411764706f, 0.21822364705882352f, 0.30334476470588234f,
-0.1650137254901961f, 0.21576301960784314f, 0.30057656862745097f,
-0.16301450980392157f, 0.21330239215686275f, 0.2978083725490196f,
-0.1610152941176471f, 0.21084176470588237f, 0.2950401764705882f,
-0.15901607843137255f, 0.20838113725490195f, 0.29227198039215685f,
-0.15701686274509805f, 0.20592050980392157f, 0.2895037843137255f,
-0.15501764705882354f, 0.20345988235294118f, 0.2867355882352941f,
-0.15301843137254903f, 0.2009992549019608f, 0.28396739215686273f,
-0.15101921568627452f, 0.1985386274509804f, 0.28119919607843136f,
-0.14902f, 0.196078f, 0.278431f
-}
-, 
-{
-0.0862745098039216f, 0.00392156862745098f, 0.298039215686275f,
-0.08982329931935286f, 0.006456462026374212f, 0.317811345363286f,
-0.09337208883478411f, 0.008991355425297444f, 0.33758347504029695f,
-0.09692087835021537f, 0.011526248824220677f, 0.35735560471730793f,
-0.10046966786564662f, 0.014061142223143908f, 0.3771277343943189f,
-0.10401845738107789f, 0.016596035622067143f, 0.3968998640713299f,
-0.10756724689650915f, 0.019130929020990374f, 0.41667199374834085f,
-0.1111160364119404f, 0.021665822419913605f, 0.4364441234253518f,
-0.11340168241333767f, 0.024661032165175043f, 0.45340494373553814f,
-0.11218083286206953f, 0.02893408342513443f, 0.4625615488616119f,
-0.1109599833108014f, 0.03320713468509382f, 0.4717181539876857f,
-0.10973913375953327f, 0.03748018594505321f, 0.48087475911375943f,
-0.10851828420826513f, 0.0417532372050126f, 0.49003136423983323f,
-0.10729743465699701f, 0.04602628846497199f, 0.499187969365907f,
-0.10607658510572887f, 0.05029933972493138f, 0.5083445744919808f,
-0.09331021241785199f, 0.04876183815739442f, 0.5194176988394122f,
-0.07835474816590786f, 0.046122625387451474f, 0.5308542052536591f,
-0.06339928391396372f, 0.04348341261750853f, 0.5422907116679061f,
-0.04844381966201959f, 0.04084419984756559f, 0.5537272180821531f,
-0.038521432139711945f, 0.04442266469293837f, 0.5642553393014402f,
-0.03670853298500114f, 0.05801929278187287f, 0.5733198350749943f,
-0.03489563383029033f, 0.07161592087080738f, 0.5823843308485482f,
-0.033082734675579525f, 0.0852125489597419f, 0.5914488266221023f,
-0.03153129829389241f, 0.09893905159962488f, 0.6005293276463081f,
-0.0343354355690541f, 0.11482906748563648f, 0.609876451896847f,
-0.03713957284421579f, 0.1307190833716481f, 0.619223576147386f,
-0.03994371011937749f, 0.1466090992576597f, 0.628570700397925f,
-0.04274784739453918f, 0.1624991151436713f, 0.6379178246484639f,
-0.04481715070521909f, 0.1759049571056668f, 0.6448154071591428f,
-0.046767936262570996f, 0.1889101388823944f, 0.6513179151390789f,
-0.0487187218199229f, 0.20191532065912202f, 0.657820423119015f,
-0.050669507377274804f, 0.2149205024358496f, 0.6643229310989512f,
-0.05262029293462671f, 0.2279256842125772f, 0.6708254390788874f,
-0.0545710784919786f, 0.24093086598930472f, 0.6773279470588235f,
-0.054902f, 0.2508731667144937f, 0.6816883716056075f,
-0.054902f, 0.26018975245027703f, 0.6856111906136431f,
-0.054902f, 0.2695063381860604f, 0.6895340096216785f,
-0.054902f, 0.27882292392184366f, 0.6934568286297141f,
-0.054902f, 0.288139509657627f, 0.6973796476377497f,
-0.054902f, 0.2974560953934104f, 0.7013024666457853f,
-0.054902f, 0.3067726811291937f, 0.7052252856538209f,
-0.054902f, 0.31608926686497696f, 0.7091481046618564f,
-0.054569855699292166f, 0.3242898013180376f, 0.7124610866487674f,
-0.054171027288259704f, 0.33226626783825564f, 0.7156516325766821f,
-0.05377219887722724f, 0.3402427343584737f, 0.7188421785045969f,
-0.05337337046619478f, 0.3482192008786917f, 0.7220327244325115f,
-0.05297454205516231f, 0.3561956673989098f, 0.7252232703604263f,
-0.05257571364412985f, 0.36417213391912784f, 0.728413816288341f,
-0.05217688523309739f, 0.37214860043934583f, 0.7316043622162558f,
-0.05177805682206493f, 0.38012506695956383f, 0.7347949081441705f,
-0.051379228411032464f, 0.38810153347978193f, 0.7379854540720853f,
-0.0509804f, 0.396078f, 0.741176f,
-0.0491349568627451f, 0.4071507843137255f, 0.7454820392156862f,
-0.04728951372549021f, 0.4182235686274509f, 0.7497880784313725f,
-0.0454440705882353f, 0.42929635294117646f, 0.7540941176470588f,
-0.04359862745098039f, 0.44036913725490195f, 0.7584001568627451f,
-0.04175318431372549f, 0.45144192156862745f, 0.7627061960784314f,
-0.0399077411764706f, 0.4625147058823529f, 0.7670122352941176f,
-0.03844675882352941f, 0.47358739215686274f, 0.7705493529411764f,
-0.03721645294117647f, 0.48466001960784316f, 0.7736251176470588f,
-0.035986147058823525f, 0.49573264705882353f, 0.7767008823529411f,
-0.0347558411764706f, 0.5068052745098038f, 0.7797766470588235f,
-0.03352553529411764f, 0.5178779019607843f, 0.7828524117647059f,
-0.0322952294117647f, 0.5289505294117647f, 0.7859281764705882f,
-0.0313725f, 0.5402018875658186f, 0.789119092570842f,
-0.0313725f, 0.5519894378290925f, 0.7926554628542101f,
-0.0313725f, 0.5637769880923663f, 0.796191833137578f,
-0.0313725f, 0.5755645383556405f, 0.7997282034209461f,
-0.0313725f, 0.5873520886189144f, 0.803264573704314f,
-0.0313725f, 0.5991396388821885f, 0.8068009439876821f,
-0.0313725f, 0.6109271891454624f, 0.81033731427105f,
-0.030686504498252627f, 0.6239180159928419f, 0.8134800106205788f,
-0.02953605546507661f, 0.637723521737313f, 0.8163561698742561f,
-0.028385606431900596f, 0.6515290274817839f, 0.8192323291279332f,
-0.027235157398724598f, 0.6653345332262547f, 0.8221084883816103f,
-0.026084708365548565f, 0.6791400389707258f, 0.8249846476352876f,
-0.024934259332372546f, 0.6929455447151969f, 0.8278608068889647f,
-0.02378381029919653f, 0.7067510504596678f, 0.8307369661426419f,
-0.026665552705750366f, 0.7201084854266825f, 0.8336130111508377f,
-0.030692146324308684f, 0.7333387006472386f, 0.8364890237215664f,
-0.034718739942867005f, 0.7465689158677947f, 0.8393650362922951f,
-0.03874533356142532f, 0.7597991310883508f, 0.8422410488630238f,
-0.042771927179983585f, 0.7730293463089067f, 0.8451170614337524f,
-0.04679852079854196f, 0.7862595615294631f, 0.8479930740044811f,
-0.050825114417100276f, 0.7994897767500192f, 0.8508690865752098f,
-0.0543243452656655f, 0.8093630740305169f, 0.8543239793739822f,
-0.057802422583076996f, 0.8191017189004424f, 0.8578020921678604f,
-0.06128049990048849f, 0.828840363770368f, 0.8612802049617386f,
-0.06475857721789999f, 0.8385790086402934f, 0.8647583177556168f,
-0.06823665453531148f, 0.848317653510219f, 0.868236430549495f,
-0.08245599373837956f, 0.8578084103145681f, 0.8701036085744033f,
-0.11909686111324025f, 0.8667817205372278f, 0.8686080886509812f,
-0.1557377284881004f, 0.8757550307598874f, 0.8671125687275592f,
-0.19237859586296055f, 0.8847283409825469f, 0.865617048804137f,
-0.2290194632378207f, 0.8937016512052064f, 0.8641215288807149f,
-0.26553302715502125f, 0.9026409960498816f, 0.8629490074850659f,
-0.30057390134620043f, 0.911187417841895f, 0.8655130429914547f,
-0.3356147755373796f, 0.9197338396339085f, 0.8680770784978435f,
-0.3706556497285583f, 0.9282802614259218f, 0.8706411140042323f,
-0.405696523919738f, 0.9368266832179354f, 0.8732051495106212f,
-0.43386584145924617f, 0.9428081693701252f, 0.8717897639601007f,
-0.4549143991909399f, 0.9461317001943552f, 0.8662506400840699f,
-0.4759629569226337f, 0.9494552310185852f, 0.8607115162080392f,
-0.4970115146543274f, 0.9527787618428152f, 0.8551723923320084f,
-0.5180600723860211f, 0.9561022926670452f, 0.8496332684559776f,
-0.5391086301177149f, 0.9594258234912751f, 0.8440941445799469f,
-0.5601571878494084f, 0.9627493543155051f, 0.8385550207039163f,
-0.5800668584292327f, 0.9660728516641083f, 0.8359774258320541f,
-0.5983462540385759f, 0.9693963010935857f, 0.8376391505467928f,
-0.616625649647919f, 0.9727197505230629f, 0.8393008752615315f,
-0.6349050452572622f, 0.9760431999525403f, 0.8409625999762702f,
-0.6531844408666054f, 0.9793666493820177f, 0.8426243246910089f,
-0.6735201542774042f, 0.980392f, 0.8464029045932956f,
-0.6947733420448079f, 0.980392f, 0.8511259690250876f,
-0.7160265298122113f, 0.980392f, 0.8558490334568797f,
-0.7372797175796152f, 0.980392f, 0.8605720978886717f,
-0.7585329053470189f, 0.980392f, 0.8652951623204639f,
-0.7782885274576924f, 0.980392f, 0.8709115804190044f,
-0.7974311152536634f, 0.980392f, 0.8768936963087709f,
-0.8165737030496344f, 0.980392f, 0.8828758121985375f,
-0.8353990946608948f, 0.9813855320133272f, 0.891242498620913f,
-0.853807049407915f, 0.983686573095667f, 0.902747330120914f,
-0.8722150041549349f, 0.9859876141780067f, 0.9142521616209148f,
-0.8904041643798827f, 0.9882352915165185f, 0.9255942420873363f,
-0.8993772540094933f, 0.9882351793505613f, 0.930080917762425f,
-0.908350343639104f, 0.9882350671846042f, 0.9345675934375137f,
-0.9292490417620138f, 0.9903519556804415f, 0.9436056906327613f,
-0.9679599494660566f, 0.9956308177973706f, 0.9594418370893697f,
-0.9990904468949905f, 0.9990904468949905f, 0.9673948853461224f,
-0.9938123206509195f, 0.9938123206509195f, 0.9374855032963872f,
-0.9885341944068486f, 0.9885341944068486f, 0.9075761212466521f,
-0.9891171203205709f, 0.9847079893059514f, 0.882954871665155f,
-0.9900518820820471f, 0.9809689422600464f, 0.8586510658667726f,
-0.9909866438435234f, 0.9772298952141412f, 0.8343472600683889f,
-0.9919214056049996f, 0.9734908481682361f, 0.8100434542700058f,
-0.9921568928672992f, 0.9708277767807161f, 0.784126871617359f,
-0.9921569331317006f, 0.9685269883577208f, 0.7576672640595238f,
-0.992156973396102f, 0.9662261999347256f, 0.7312076565016885f,
-0.9918070264488894f, 0.9640061421312186f, 0.7074261899475284f,
-0.9907754776791468f, 0.9619433076077379f, 0.6888604162201971f,
-0.9897439289094041f, 0.9598804730842572f, 0.6702946424928657f,
-0.9887123801396615f, 0.9578176385607764f, 0.6517288687655345f,
-0.9874876143566166f, 0.9531259764897456f, 0.630431253475807f,
-0.986096404715532f, 0.9461697509019893f, 0.6067803348127022f,
-0.9847051950744474f, 0.9392135253142329f, 0.5831294161495973f,
-0.9833139854333628f, 0.9322572997264765f, 0.5594784974864926f,
-0.9819227757922783f, 0.9253010741387202f, 0.5358275788233877f,
-0.9805315661511936f, 0.9183448485509638f, 0.512176660160283f,
-0.9785571151317216f, 0.9109192967647322f, 0.4939020283846548f,
-0.9765176290163489f, 0.9034414121443467f, 0.47622688653939166f,
-0.9744781429009761f, 0.8959635275239616f, 0.458551744694129f,
-0.9724386567856035f, 0.8884856429035763f, 0.4408766028488664f,
-0.9703991706702307f, 0.8810077582831911f, 0.4232014610036037f,
-0.9681369441769997f, 0.8732357848772184f, 0.4056866743982491f,
-0.9643980579101176f, 0.8635141466474067f, 0.3892349646114899f,
-0.9606591716432354f, 0.853792508417595f, 0.3727832548247306f,
-0.9569202853763533f, 0.8440708701877834f, 0.35633154503797143f,
-0.9531813991094712f, 0.8343492319579716f, 0.3398798352512122f,
-0.9494425128425891f, 0.82462759372816f, 0.323428125464453f,
-0.9448745737626966f, 0.813580230728528f, 0.3124489777119139f,
-0.9402010116374271f, 0.8023639676475951f, 0.3021670456964163f,
-0.9355274495121576f, 0.7911477045666622f, 0.29188511368091874f,
-0.9308538873868881f, 0.7799314414857293f, 0.28160318166542114f,
-0.9272115884592539f, 0.7698697653777618f, 0.27359016194241303f,
-0.924029411764706f, 0.7603232352941179f, 0.26658947058823546f,
-0.9208472350701579f, 0.7507767052104736f, 0.2595887792340575f,
-0.9176650583756099f, 0.7412301751268295f, 0.2525880878798797f,
-0.9144828816810617f, 0.7316836450431854f, 0.2455873965257019f,
-0.9113007049865137f, 0.7221371149595414f, 0.23858670517152414f,
-0.9079179326501371f, 0.7119888941390766f, 0.2315218653002742f,
-0.904357030656664f, 0.701306369763206f, 0.2244000613133279f,
-0.9007961286631908f, 0.6906238453873353f, 0.2172782573263816f,
-0.8972352266697177f, 0.6799413210114648f, 0.2101564533394353f,
-0.8936743246762445f, 0.6692587966355942f, 0.20303464935248902f,
-0.890135681872749f, 0.6586229429937721f, 0.195937261121023f,
-0.8875346414565826f, 0.6499529556618228f, 0.1898683142052401f,
-0.8849336010404162f, 0.6412829683298734f, 0.18379936728945723f,
-0.8823325606242497f, 0.6326129809979241f, 0.17773042037367434f,
-0.8797315202080832f, 0.6239429936659747f, 0.17166147345789146f,
-0.877815862745098f, 0.6140716862745098f, 0.16770433333333334f,
-0.8765855882352941f, 0.6029990588235298f, 0.16585900000000006f,
-0.8753553137254901f, 0.5919264313725491f, 0.16401366666666667f,
-0.8741250392156863f, 0.5808538039215686f, 0.16216833333333333f,
-0.8728947647058823f, 0.5697811764705882f, 0.160323f,
-0.8716644901960785f, 0.5587085490196078f, 0.15847766666666666f,
-0.8702035294117647f, 0.5475590196078431f, 0.1566323137254902f,
-0.8671277647058824f, 0.5358711764705882f, 0.15478682352941175f,
-0.864052f, 0.5241833333333333f, 0.15294133333333335f,
-0.8609762352941176f, 0.5124954901960784f, 0.1510958431372549f,
-0.8579004705882353f, 0.5008076470588235f, 0.14925035294117647f,
-0.8548247058823529f, 0.4891198039215686f, 0.14740486274509804f,
-0.8517489411764706f, 0.4774319607843137f, 0.1455593725490196f,
-0.848673294117647f, 0.4671282352941176f, 0.1437138823529412f,
-0.8455976862745098f, 0.45728588235294115f, 0.14186839215686275f,
-0.8425220784313725f, 0.4474435294117647f, 0.14002290196078432f,
-0.8394464705882353f, 0.4376011764705882f, 0.13817741176470588f,
-0.8363708627450981f, 0.427758823529412f, 0.1363319215686275f,
-0.8332952549019608f, 0.4179164705882353f, 0.13448643137254904f,
-0.8302195882352942f, 0.4078434117647059f, 0.13217958823529413f,
-0.8271438235294117f, 0.3973858431372549f, 0.12910382352941177f,
-0.8240680588235294f, 0.38692827450980394f, 0.1260280588235294f,
-0.8209922941176471f, 0.37647070588235293f, 0.12295229411764706f,
-0.8179165294117647f, 0.366013137254902f, 0.11987652941176472f,
-0.8148407647058823f, 0.35555556862745097f, 0.11680076470588235f,
-0.811765f, 0.345098f, 0.113725f,
-0.8080740196078431f, 0.3327950980392157f, 0.11064932941176471f,
-0.8043830392156862f, 0.3204921960784314f, 0.10757365882352941f,
-0.8006920588235295f, 0.30818929411764706f, 0.10449798823529412f,
-0.7970010784313726f, 0.29588639215686274f, 0.10142231764705882f,
-0.7933100980392157f, 0.2835834901960784f, 0.09834664705882352f,
-0.7896191176470588f, 0.2712805882352941f, 0.09527097647058823f,
-0.7836213725490196f, 0.2585931764705882f, 0.09219526666666666f,
-0.7762395686274511f, 0.24567505882352977f, 0.08911953333333342f,
-0.7688577647058823f, 0.23275694117647058f, 0.0860438f,
-0.7614759607843137f, 0.21983882352941175f, 0.08296806666666667f,
-0.7540941568627451f, 0.20692070588235295f, 0.07989233333333333f,
-0.7467123529411764f, 0.19400258823529412f, 0.0768166f,
-0.7391767843137255f, 0.18200717647058823f, 0.07404843921568628f,
-0.7311799215686274f, 0.17277988235294117f, 0.07220299607843138f,
-0.7231830588235294f, 0.16355258823529412f, 0.07035755294117647f,
-0.7151861960784314f, 0.15432529411764706f, 0.06851210980392157f,
-0.7071893333333333f, 0.145098f, 0.06666666666666667f,
-0.6991924705882353f, 0.13587070588235295f, 0.06482122352941176f,
-0.6911956078431373f, 0.12664341176470587f, 0.06297578039215686f,
-0.6805074509803922f, 0.11687795490196078f, 0.060053833333333334f,
-0.6694348235294117f, 0.10703561764705882f, 0.0569781f,
-0.6583621960784314f, 0.09719328039215686f, 0.053902366666666666f,
-0.647289568627451f, 0.0873509431372549f, 0.050826633333333336f,
-0.636216941176471f, 0.07750860588235323f, 0.04775090000000009f,
-0.6251443137254903f, 0.06766626862745098f, 0.04467516666666667f,
-0.6136913649751099f, 0.059786774105627376f, 0.04543821455211589f,
-0.6018580949253296f, 0.053870122316882124f, 0.05004004365634768f,
-0.5900248248755493f, 0.04795347052813688f, 0.05464187276057946f,
-0.578191554825769f, 0.04203681873939163f, 0.059243701864811246f,
-0.5663582847759888f, 0.036120166950646376f, 0.06384553096904302f,
-0.5545250147262085f, 0.03020351516190113f, 0.06844736007327482f,
-0.5410219798296445f, 0.026251307171840824f, 0.07258769484519471f,
-0.5260663992311696f, 0.0240079891502823f, 0.07632657092660074f,
-0.5111108186326947f, 0.021764671128723774f, 0.08006544700800676f,
-0.49615523803421974f, 0.019521353107165247f, 0.08380432308941278f,
-0.4811996574357448f, 0.017278035085606727f, 0.08754319917081879f,
-0.4671127882020719f, 0.015107094915631003f, 0.09077531345386532f,
-0.455148071471673f, 0.013112964725519002f, 0.09276947245909949f,
-0.44318335474127407f, 0.011118834535407001f, 0.09476363146433366f,
-0.43121863801087557f, 0.009124704345295059f, 0.09675779046956778f,
-0.4192539212804763f, 0.007130574155183001f, 0.09875194947480201f,
-0.4072892045500774f, 0.0051364439650710015f, 0.10074610848003619f,
-0.39013769849135815f, 0.0037389347229546776f, 0.10049990825007929f,
-0.3648998979446442f, 0.003271567882585343f, 0.09676094913058408f,
-0.33966209739793035f, 0.0028042010422160086f, 0.09302199001108886f,
-0.3144242968512165f, 0.0023368342018466734f, 0.08928303089159366f,
-0.28918649630450255f, 0.001869467361477339f, 0.08554407177209844f,
-0.2639486957577887f, 0.0014021005211080043f, 0.08180511265260323f,
-0.23871089521107478f, 0.0009347336807386691f, 0.07806615353310803f,
-0.2134730946643609f, 0.00046736684036933476f, 0.07432719441361282f,
-0.188235294117647f, 0.0f, 0.0705882352941176f
-}
-, 
-{
-1.0f, 0.984314f, 0.8f,
-0.9992310588235294f, 0.9812382352941177f, 0.789388705882353f,
-0.9984621176470588f, 0.9781624705882354f, 0.7787774117647059f,
-0.9976931764705882f, 0.9750867058823529f, 0.7681661176470589f,
-0.9969242352941177f, 0.9720109411764706f, 0.7575548235294118f,
-0.996155294117647f, 0.9689351764705882f, 0.7469435294117648f,
-0.9953863529411765f, 0.9658594117647059f, 0.7363322352941177f,
-0.9946174117647059f, 0.9627836470588236f, 0.7257209411764706f,
-0.9938484705882353f, 0.9597078823529411f, 0.7151096470588235f,
-0.9930795294117647f, 0.9566321176470588f, 0.7044983529411765f,
-0.9923105882352942f, 0.9535563529411765f, 0.6938870588235294f,
-0.9915416470588235f, 0.9504805882352941f, 0.6832757647058824f,
-0.990772705882353f, 0.9474048235294118f, 0.6726644705882353f,
-0.9900037647058824f, 0.9443290588235294f, 0.6620531764705883f,
-0.9892348235294118f, 0.941253294117647f, 0.6514418823529412f,
-0.9884658823529412f, 0.9381775294117647f, 0.6408305882352942f,
-0.9876969411764707f, 0.9351017647058824f, 0.6302192941176471f,
-0.986928f, 0.932026f, 0.619608f,
-0.9861590588235294f, 0.9289502352941176f, 0.6089967058823529f,
-0.9853901176470589f, 0.9258744705882352f, 0.5983854117647058f,
-0.9846211764705882f, 0.9227987058823529f, 0.5877741176470588f,
-0.9838522352941177f, 0.9197229411764706f, 0.5771628235294117f,
-0.983083294117647f, 0.9166471764705882f, 0.5665515294117647f,
-0.9823143529411765f, 0.9135714117647058f, 0.5559402352941176f,
-0.9815454117647059f, 0.9104956470588235f, 0.5453289411764706f,
-0.9807764705882354f, 0.9074198823529411f, 0.5347176470588235f,
-0.9800075294117647f, 0.9041134509803921f, 0.5265669411764706f,
-0.9792385882352942f, 0.9005763529411764f, 0.5208768235294118f,
-0.9784696470588236f, 0.8970392549019608f, 0.5151867058823529f,
-0.977700705882353f, 0.893502156862745f, 0.5094965882352941f,
-0.9769317647058824f, 0.8899650588235294f, 0.5038064705882352f,
-0.9761628235294117f, 0.8864279607843137f, 0.49811635294117645f,
-0.9753938823529412f, 0.882890862745098f, 0.49242623529411766f,
-0.9746249411764706f, 0.8793537647058823f, 0.4867361176470589f,
-0.973856f, 0.8758166666666667f, 0.481046f,
-0.9730870588235294f, 0.8722795686274509f, 0.4753558823529412f,
-0.9723181176470589f, 0.8687424705882353f, 0.46966576470588234f,
-0.9715491764705882f, 0.8652053725490196f, 0.46397564705882355f,
-0.9707802352941176f, 0.8616682745098039f, 0.4582855294117647f,
-0.9700112941176471f, 0.8581311764705882f, 0.45259541176470586f,
-0.9692423529411764f, 0.8545940784313726f, 0.44690529411764707f,
-0.9684734117647059f, 0.8510569803921569f, 0.4412151764705883f,
-0.9677044705882353f, 0.8475198823529412f, 0.43552505882352943f,
-0.9669355294117647f, 0.8439827843137255f, 0.4298349411764706f,
-0.9661665882352941f, 0.8404456862745098f, 0.42414482352941174f,
-0.9653976470588235f, 0.8369085882352941f, 0.41845470588235295f,
-0.9646287058823529f, 0.8333714901960785f, 0.4127645882352941f,
-0.9638597647058823f, 0.8298343921568627f, 0.4070744705882353f,
-0.9630908235294118f, 0.8262972941176471f, 0.40138435294117647f,
-0.9623218823529411f, 0.8227601960784314f, 0.3956942352941177f,
-0.9615529411764706f, 0.8192230980392157f, 0.3900041176470588f,
-0.960784f, 0.815686f, 0.384314f,
-0.9560166274509804f, 0.8095345490196079f, 0.3807769019607843f,
-0.9512492549019608f, 0.8033830980392157f, 0.37723980392156864f,
-0.9464818823529412f, 0.7972316470588235f, 0.37370270588235294f,
-0.9417145098039216f, 0.7910801960784314f, 0.37016560784313723f,
-0.936947137254902f, 0.7849287450980392f, 0.3666285098039216f,
-0.9321797647058824f, 0.7787772941176471f, 0.3630914117647059f,
-0.9274123921568627f, 0.7726258431372549f, 0.3595543137254902f,
-0.9226450196078431f, 0.7664743921568627f, 0.3560172156862745f,
-0.9178776470588235f, 0.7603229411764706f, 0.35248011764705883f,
-0.9131102745098039f, 0.7541714901960784f, 0.34894301960784313f,
-0.9083429019607843f, 0.7480200392156863f, 0.3454059215686274f,
-0.9035755294117647f, 0.7418685882352941f, 0.3418688235294117f,
-0.8988081568627451f, 0.735717137254902f, 0.3383317254901961f,
-0.8940407843137255f, 0.7295656862745098f, 0.3347946274509804f,
-0.8892734117647059f, 0.7234142352941177f, 0.33125752941176473f,
-0.8845060392156863f, 0.7172627843137255f, 0.327720431372549f,
-0.8797386666666667f, 0.7111113333333333f, 0.3241833333333333f,
-0.8749712941176471f, 0.7049598823529412f, 0.3206462352941176f,
-0.8702039215686275f, 0.698808431372549f, 0.317109137254902f,
-0.8654365490196079f, 0.6926569803921568f, 0.31357203921568627f,
-0.8606691764705882f, 0.6865055294117647f, 0.31003494117647057f,
-0.8559018039215686f, 0.6803540784313725f, 0.3064978431372549f,
-0.8511344313725491f, 0.6742026274509805f, 0.3029607450980393f,
-0.8463670588235294f, 0.6680511764705882f, 0.2994236470588235f,
-0.8415996862745098f, 0.661899725490196f, 0.29588654901960787f,
-0.8369091960784313f, 0.6559020392156862f, 0.2925801176470588f,
-0.832295588235294f, 0.6500581176470588f, 0.28950435294117643f,
-0.8276819803921568f, 0.6442141960784313f, 0.2864285882352941f,
-0.8230683725490195f, 0.6383702745098039f, 0.28335282352941177f,
-0.8184547647058823f, 0.6325263529411764f, 0.2802770588235294f,
-0.8138411568627452f, 0.6266824313725491f, 0.2772012941176471f,
-0.8092275490196078f, 0.6208385098039215f, 0.2741255294117647f,
-0.8046139411764706f, 0.6149945882352941f, 0.2710497647058823f,
-0.8000003333333333f, 0.6091506666666666f, 0.267974f,
-0.7953867254901961f, 0.6033067450980392f, 0.26489823529411766f,
-0.7907731176470588f, 0.5974628235294117f, 0.26182247058823527f,
-0.7861595098039216f, 0.5916189019607843f, 0.25874670588235293f,
-0.7815459019607843f, 0.5857749803921568f, 0.2556709411764706f,
-0.7769322941176471f, 0.5799310588235295f, 0.25259517647058827f,
-0.7723186862745098f, 0.574087137254902f, 0.24951941176470588f,
-0.7677050784313726f, 0.5682432156862746f, 0.24644364705882352f,
-0.7630914705882352f, 0.5623992941176471f, 0.24336788235294116f,
-0.7584778627450981f, 0.5565553725490197f, 0.24029211764705882f,
-0.7538642549019607f, 0.5507114509803922f, 0.23721635294117646f,
-0.7492506470588235f, 0.5448675294117648f, 0.2341405882352941f,
-0.7446370392156862f, 0.5390236078431373f, 0.23106482352941177f,
-0.7400234313725491f, 0.53317968627451f, 0.22798905882352943f,
-0.7354098235294118f, 0.5273357647058824f, 0.22491329411764704f,
-0.7307962156862745f, 0.5214918431372549f, 0.2218375294117647f,
-0.7261826078431373f, 0.5156479215686275f, 0.21876176470588232f,
-0.721569f, 0.509804f, 0.215686f,
-0.7188008039215686f, 0.5061130980392157f, 0.21430192156862743f,
-0.7160326078431373f, 0.5024221960784314f, 0.2129178431372549f,
-0.7132644117647059f, 0.4987312941176471f, 0.21153376470588234f,
-0.7104962156862745f, 0.49504039215686285f, 0.2101496862745098f,
-0.7077280196078432f, 0.49134949019607843f, 0.20876560784313725f,
-0.7049598235294118f, 0.4876585882352941f, 0.2073815294117647f,
-0.7021916274509804f, 0.4839676862745098f, 0.20599745098039215f,
-0.699423431372549f, 0.4802767843137255f, 0.2046133725490196f,
-0.6966552352941177f, 0.4765858823529412f, 0.20322929411764706f,
-0.6938870392156863f, 0.4728949803921569f, 0.2018452156862745f,
-0.6911188431372549f, 0.4692040784313726f, 0.20046113725490194f,
-0.6883506470588235f, 0.4655131764705883f, 0.1990770588235294f,
-0.6855824509803922f, 0.46182227450980395f, 0.19769298039215685f,
-0.6828142549019608f, 0.4581313725490196f, 0.1963089019607843f,
-0.6800460588235294f, 0.45444047058823533f, 0.19492482352941176f,
-0.677277862745098f, 0.45074956862745097f, 0.1935407450980392f,
-0.6745096666666667f, 0.4470586666666667f, 0.19215666666666667f,
-0.6717414705882353f, 0.44336776470588235f, 0.1907725882352941f,
-0.6689732745098039f, 0.43967686274509804f, 0.18938850980392155f,
-0.6662050784313726f, 0.4359859607843138f, 0.18800443137254905f,
-0.6634368823529412f, 0.4322950588235294f, 0.18662035294117646f,
-0.6606686862745098f, 0.4286041568627451f, 0.18523627450980393f,
-0.6579004901960784f, 0.4249132549019608f, 0.18385219607843137f,
-0.6551322941176471f, 0.4212223529411765f, 0.18246811764705884f,
-0.6523640980392157f, 0.4175314509803921f, 0.18108403921568628f,
-0.6489807843137255f, 0.41360988235294116f, 0.17954617647058824f,
-0.6449823529411765f, 0.40945764705882354f, 0.1778545294117647f,
-0.6409839215686275f, 0.40530541176470586f, 0.17616288235294117f,
-0.6369854901960784f, 0.40115317647058824f, 0.17447123529411765f,
-0.6329870588235296f, 0.3970009411764707f, 0.17277958823529416f,
-0.6289886274509804f, 0.39284870588235293f, 0.17108794117647058f,
-0.6249901960784313f, 0.3886964705882353f, 0.16939629411764706f,
-0.6209917647058824f, 0.38454423529411763f, 0.1677046470588235f,
-0.6169933333333333f, 0.380392f, 0.166013f,
-0.6129949019607843f, 0.3762397647058824f, 0.16432135294117647f,
-0.6089964705882353f, 0.3720875294117647f, 0.16262970588235293f,
-0.6049980392156863f, 0.3679352941176471f, 0.1609380588235294f,
-0.6009996078431372f, 0.3637830588235294f, 0.1592464117647059f,
-0.5970011764705883f, 0.3596308235294118f, 0.15755476470588234f,
-0.5930027450980392f, 0.35547858823529416f, 0.15586311764705882f,
-0.5890043137254902f, 0.3513263529411765f, 0.1541714705882353f,
-0.5850058823529412f, 0.34717411764705886f, 0.15247982352941175f,
-0.5810074509803921f, 0.3430218823529412f, 0.15078817647058823f,
-0.5770090196078431f, 0.33886964705882355f, 0.1490965294117647f,
-0.573010588235294f, 0.33471741176470593f, 0.14740488235294116f,
-0.5690121568627452f, 0.33056517647058836f, 0.14571323529411767f,
-0.565013725490196f, 0.32641294117647063f, 0.1440215882352941f,
-0.5610152941176471f, 0.32226070588235295f, 0.14232994117647058f,
-0.557016862745098f, 0.3181084705882353f, 0.14063829411764706f,
-0.553018431372549f, 0.31395623529411765f, 0.1389466470588235f,
-0.54902f, 0.309804f, 0.137255f,
-0.545175294117647f, 0.3061130980392157f, 0.13571711764705882f,
-0.5413305882352941f, 0.3024221960784314f, 0.13417923529411763f,
-0.5374858823529411f, 0.2987312941176471f, 0.13264135294117646f,
-0.5336411764705882f, 0.2950403921568628f, 0.1311034705882353f,
-0.5297964705882352f, 0.2913494901960785f, 0.1295655882352941f,
-0.5259517647058823f, 0.2876585882352941f, 0.12802770588235293f,
-0.5221070588235294f, 0.2839676862745098f, 0.12648982352941177f,
-0.5182623529411764f, 0.2802767843137255f, 0.12495194117647059f,
-0.5144176470588235f, 0.2765858823529412f, 0.1234140588235294f,
-0.5105729411764706f, 0.27289498039215687f, 0.12187617647058822f,
-0.5067282352941177f, 0.2692040784313727f, 0.1203382941176471f,
-0.5028835294117647f, 0.26551317647058825f, 0.11880041176470588f,
-0.49903882352941176f, 0.26182227450980394f, 0.11726252941176471f,
-0.4951941176470588f, 0.25813137254901963f, 0.11572464705882352f,
-0.49134941176470587f, 0.25444047058823527f, 0.11418676470588235f,
-0.4875047058823529f, 0.250749568627451f, 0.11264888235294117f,
-0.48366f, 0.24705866666666668f, 0.111111f,
-0.479815294117647f, 0.24336776470588234f, 0.10957311764705882f,
-0.4759705882352941f, 0.23967686274509803f, 0.10803523529411764f,
-0.4721258823529412f, 0.23598596078431372f, 0.10649735294117647f,
-0.4682811764705882f, 0.2322950588235294f, 0.10495947058823529f,
-0.4644364705882353f, 0.2286041568627451f, 0.10342158823529411f,
-0.46059176470588237f, 0.2249132549019608f, 0.10188370588235293f,
-0.4567470588235294f, 0.22122235294117648f, 0.10034582352941177f,
-0.4529023529411765f, 0.21753145098039214f, 0.0988079411764706f,
-0.44898078431372546f, 0.21376366666666666f, 0.09696249019607843f,
-0.44498235294117655f, 0.2099190000000001f, 0.09480947058823536f,
-0.4409839215686274f, 0.20607433333333333f, 0.09265645098039216f,
-0.4369854901960784f, 0.20222966666666667f, 0.09050343137254901f,
-0.4329870588235294f, 0.19838499999999998f, 0.08835041176470589f,
-0.4289886274509804f, 0.19454033333333332f, 0.08619739215686274f,
-0.42499019607843136f, 0.19069566666666665f, 0.08404437254901961f,
-0.42099176470588234f, 0.186851f, 0.08189135294117647f,
-0.4169933333333333f, 0.18300633333333333f, 0.07973833333333333f,
-0.4129949019607843f, 0.17916166666666666f, 0.0775853137254902f,
-0.4089964705882353f, 0.175317f, 0.07543229411764707f,
-0.4049980392156863f, 0.17147233333333334f, 0.07327927450980393f,
-0.40099960784313726f, 0.16762766666666667f, 0.07112625490196078f,
-0.39700117647058825f, 0.163783f, 0.06897323529411764f,
-0.39300274509803923f, 0.15993833333333332f, 0.06682021568627451f,
-0.3890043137254902f, 0.15609366666666666f, 0.06466719607843138f,
-0.3850058823529412f, 0.152249f, 0.06251417647058824f,
-0.38100745098039224f, 0.14840433333333344f, 0.06036115686274516f,
-0.37700901960784317f, 0.14455966666666667f, 0.05820813725490196f,
-0.3730105882352941f, 0.140715f, 0.056055117647058825f,
-0.3690121568627451f, 0.13687033333333332f, 0.05390209803921569f,
-0.36501372549019606f, 0.13302566666666665f, 0.05174907843137255f,
-0.36101529411764705f, 0.129181f, 0.04959605882352942f,
-0.35701686274509803f, 0.12533633333333333f, 0.04744303921568628f,
-0.353018431372549f, 0.12149166666666666f, 0.04529001960784314f,
-0.34902f, 0.117647f, 0.043137f,
-0.3451752941176471f, 0.1153402156862745f, 0.04221429411764706f,
-0.3413305882352941f, 0.11303343137254902f, 0.04129158823529412f,
-0.3374858823529412f, 0.11072664705882353f, 0.04036888235294118f,
-0.3336411764705882f, 0.10841986274509804f, 0.039446176470588235f,
-0.3297964705882353f, 0.10611307843137255f, 0.0385234705882353f,
-0.32595176470588233f, 0.10380629411764707f, 0.037600764705882354f,
-0.3221070588235294f, 0.10149950980392157f, 0.03667805882352941f,
-0.31826235294117655f, 0.09919272549019614f, 0.0357553529411765f,
-0.3144176470588235f, 0.09688594117647059f, 0.03483264705882353f,
-0.3105729411764706f, 0.0945791568627451f, 0.03390994117647059f,
-0.30672823529411763f, 0.09227237254901961f, 0.03298723529411765f,
-0.3028835294117647f, 0.08996558823529412f, 0.03206452941176471f,
-0.29903882352941175f, 0.08765880392156863f, 0.031141823529411767f,
-0.29519411764705883f, 0.08535201960784314f, 0.030219117647058823f,
-0.29134941176470586f, 0.08304523529411764f, 0.029296411764705883f,
-0.28750470588235294f, 0.08073845098039216f, 0.02837370588235294f,
-0.28366f, 0.07843166666666668f, 0.027451000000000003f,
-0.27981529411764705f, 0.07612488235294118f, 0.02652829411764706f,
-0.2759705882352941f, 0.07381809803921568f, 0.02560558823529412f,
-0.27212588235294116f, 0.07151131372549019f, 0.024682882352941176f,
-0.26828117647058825f, 0.0692045294117647f, 0.023760176470588236f,
-0.2644364705882353f, 0.06689774509803922f, 0.022837470588235292f,
-0.2605917647058823f, 0.06459096078431373f, 0.021914764705882352f,
-0.2567470588235295f, 0.0622841764705883f, 0.02099205882352944f,
-0.25290235294117647f, 0.05997739215686275f, 0.020069352941176472f,
-0.24898078431372547f, 0.05774749019607843f, 0.019223529411764705f,
-0.24498235294117646f, 0.055594470588235294f, 0.01845458823529412f,
-0.24098392156862744f, 0.05344145098039216f, 0.017685647058823528f,
-0.23698549019607842f, 0.05128843137254902f, 0.01691670588235294f,
-0.2329870588235294f, 0.049135411764705886f, 0.016147764705882355f,
-0.22898862745098036f, 0.04698239215686274f, 0.015378823529411764f,
-0.22499019607843135f, 0.044829372549019614f, 0.014609882352941177f,
-0.22099176470588233f, 0.04267635294117647f, 0.013840941176470587f,
-0.21699333333333332f, 0.040523333333333335f, 0.013072f,
-0.2129949019607843f, 0.0383703137254902f, 0.012303058823529412f,
-0.20899647058823528f, 0.03621729411764706f, 0.011534117647058825f,
-0.20499803921568627f, 0.03406427450980393f, 0.010765176470588235f,
-0.20099960784313725f, 0.031911254901960784f, 0.009996235294117648f,
-0.19700117647058824f, 0.029758235294117648f, 0.00922729411764706f,
-0.19300274509803933f, 0.027605215686274574f, 0.008458352941176493f,
-0.1890043137254902f, 0.025452196078431376f, 0.007689411764705884f,
-0.1850058823529412f, 0.023299176470588233f, 0.0069204705882352936f,
-0.18100745098039217f, 0.021146156862745097f, 0.006151529411764707f,
-0.17700901960784315f, 0.01899313725490196f, 0.005382588235294118f,
-0.17301058823529414f, 0.016840117647058825f, 0.004613647058823531f,
-0.16901215686274512f, 0.014687098039215689f, 0.003844705882352941f,
-0.1650137254901961f, 0.012534078431372553f, 0.0030757647058823542f,
-0.1610152941176471f, 0.010381058823529417f, 0.002306823529411764f,
-0.15701686274509807f, 0.00822803921568628f, 0.0015378823529411771f,
-0.15301843137254903f, 0.006075019607843138f, 0.0007689411764705868f,
-0.14902f, 0.003922f, 0.0f
-}
-, 
-{
-0.941176f, 0.941176f, 0.894118f,
-0.9362548235294118f, 0.9362548235294118f, 0.8873513333333333f,
-0.9313336470588235f, 0.9313336470588235f, 0.8805846666666667f,
-0.9264124705882353f, 0.9264124705882353f, 0.873818f,
-0.9214912941176471f, 0.9214912941176471f, 0.8670513333333333f,
-0.9165701176470589f, 0.9165701176470589f, 0.8602846666666666f,
-0.9116489411764706f, 0.9116489411764706f, 0.853518f,
-0.9067277647058823f, 0.9067277647058823f, 0.8467513333333333f,
-0.9018065882352941f, 0.9018065882352941f, 0.8399846666666666f,
-0.8968854117647058f, 0.8968854117647058f, 0.833218f,
-0.8919642352941176f, 0.8919642352941176f, 0.8264513333333333f,
-0.8870430588235294f, 0.8870430588235294f, 0.8196846666666666f,
-0.8821218823529411f, 0.8821218823529411f, 0.812918f,
-0.8772776078431372f, 0.8772776078431372f, 0.8064589215686274f,
-0.8726640392156862f, 0.8726640392156862f, 0.8009226078431372f,
-0.8680504705882353f, 0.8680504705882353f, 0.795386294117647f,
-0.8634369019607843f, 0.8634369019607843f, 0.7898499803921568f,
-0.8588233333333333f, 0.8588233333333333f, 0.7843136666666667f,
-0.8542097647058823f, 0.8542097647058823f, 0.7787773529411764f,
-0.8495961960784314f, 0.8495961960784314f, 0.7732410392156862f,
-0.8449826274509804f, 0.8449826274509804f, 0.7677047254901961f,
-0.8403690588235294f, 0.8403690588235294f, 0.7621684117647058f,
-0.8357554901960784f, 0.8357554901960784f, 0.7566320980392157f,
-0.8311419215686274f, 0.8311419215686274f, 0.7510957843137255f,
-0.8265283529411764f, 0.8265283529411764f, 0.7455594705882352f,
-0.8219147843137254f, 0.8219147843137254f, 0.7400231568627451f,
-0.8183776862745098f, 0.8182239215686274f, 0.7358709215686274f,
-0.8159170588235294f, 0.8154557647058823f, 0.7331027647058823f,
-0.813456431372549f, 0.8126876078431372f, 0.7303346078431372f,
-0.8109958039215687f, 0.8099194509803922f, 0.7275664509803922f,
-0.8085351764705883f, 0.8071512941176471f, 0.7247982941176471f,
-0.8060745490196078f, 0.804383137254902f, 0.722030137254902f,
-0.8036139215686274f, 0.8016149803921568f, 0.7192619803921568f,
-0.8011532941176471f, 0.7988468235294117f, 0.7164938235294117f,
-0.7986926666666667f, 0.7960786666666666f, 0.7137256666666666f,
-0.7962320392156863f, 0.7933105098039215f, 0.7109575098039216f,
-0.7937714117647059f, 0.7905423529411765f, 0.7081893529411765f,
-0.7913107843137255f, 0.7877741960784314f, 0.7054211960784313f,
-0.7888501568627451f, 0.7850060392156862f, 0.7026530392156862f,
-0.7859282352941177f, 0.7817764705882353f, 0.6994235294117647f,
-0.7828525490196079f, 0.7783930980392156f, 0.6960402352941176f,
-0.779776862745098f, 0.7750097254901961f, 0.6926569411764706f,
-0.7767011764705882f, 0.7716263529411764f, 0.6892736470588234f,
-0.7736254901960784f, 0.7682429803921568f, 0.6858903529411764f,
-0.7705498039215687f, 0.7648596078431372f, 0.6825070588235294f,
-0.7674741176470589f, 0.7614762352941176f, 0.6791237647058823f,
-0.764398431372549f, 0.758092862745098f, 0.6757404705882353f,
-0.7613227450980392f, 0.7547094901960784f, 0.6723571764705882f,
-0.7582470588235294f, 0.7513261176470588f, 0.6689738823529412f,
-0.7551713725490197f, 0.7479427450980392f, 0.6655905882352942f,
-0.7520956862745098f, 0.7445593725490196f, 0.662207294117647f,
-0.74902f, 0.741176f, 0.658824f,
-0.7453290980392157f, 0.737177568627451f, 0.6548254901960784f,
-0.7416381960784314f, 0.7331791372549019f, 0.6508269803921569f,
-0.737947294117647f, 0.729180705882353f, 0.6468284705882352f,
-0.7342563921568628f, 0.7251822745098039f, 0.6428299607843138f,
-0.7305654901960784f, 0.7211838431372549f, 0.6388314509803922f,
-0.7268745882352942f, 0.7171854117647058f, 0.6348329411764706f,
-0.7231836862745098f, 0.7131869803921569f, 0.630834431372549f,
-0.7194927843137254f, 0.7091885490196078f, 0.6268359215686274f,
-0.7158018823529412f, 0.7051901176470589f, 0.6228374117647059f,
-0.7121109803921568f, 0.7011916862745099f, 0.6188389019607844f,
-0.7084200784313726f, 0.6971932549019608f, 0.6148403921568628f,
-0.7047291764705882f, 0.6931948235294118f, 0.6108418823529412f,
-0.7009613725490196f, 0.6891963921568628f, 0.6069971764705883f,
-0.696962862745098f, 0.6851979607843137f, 0.6036138823529412f,
-0.6929643529411765f, 0.6811995294117648f, 0.6002305882352942f,
-0.6889658431372548f, 0.6772010980392157f, 0.596847294117647f,
-0.6849673333333333f, 0.6732026666666667f, 0.593464f,
-0.6809688235294117f, 0.6692042352941177f, 0.590080705882353f,
-0.6769703137254902f, 0.6652058039215687f, 0.5866974117647059f,
-0.6729718039215686f, 0.6612073725490196f, 0.5833141176470589f,
-0.668973294117647f, 0.6572089411764706f, 0.5799308235294118f,
-0.6649747843137255f, 0.6532105098039216f, 0.5765475294117648f,
-0.660976274509804f, 0.6492120784313726f, 0.5731642352941178f,
-0.6569777647058823f, 0.6452136470588236f, 0.5697809411764706f,
-0.6529792549019607f, 0.6412152156862745f, 0.5663976470588236f,
-0.6494421568627451f, 0.6373705490196079f, 0.5631681176470589f,
-0.6463664705882353f, 0.6336796470588235f, 0.5600923529411765f,
-0.6432907843137254f, 0.6299887450980393f, 0.5570165882352942f,
-0.6402150980392157f, 0.6262978431372549f, 0.5539408235294118f,
-0.6371394117647059f, 0.6226069411764706f, 0.5508650588235294f,
-0.6340637254901961f, 0.6189160392156864f, 0.5477892941176471f,
-0.6309880392156862f, 0.615225137254902f, 0.5447135294117648f,
-0.6279123529411764f, 0.6115342352941177f, 0.5416377647058824f,
-0.6248366666666667f, 0.6078433333333334f, 0.538562f,
-0.6217609803921569f, 0.604152431372549f, 0.5354862352941177f,
-0.618685294117647f, 0.6004615294117648f, 0.5324104705882353f,
-0.6156096078431372f, 0.5967706274509804f, 0.529334705882353f,
-0.6125339215686274f, 0.5930797254901962f, 0.5262589411764707f,
-0.6099195294117647f, 0.5900808823529412f, 0.5236445882352941f,
-0.6074589019607843f, 0.5873127254901961f, 0.5211840392156862f,
-0.6049982745098039f, 0.584544568627451f, 0.5187234901960784f,
-0.6025376470588235f, 0.581776411764706f, 0.5162629411764706f,
-0.6000770196078431f, 0.5790082549019608f, 0.5138023921568627f,
-0.5976163921568628f, 0.5762400980392157f, 0.511341843137255f,
-0.5951557647058824f, 0.5734719411764706f, 0.508881294117647f,
-0.592695137254902f, 0.5707037843137255f, 0.5064207450980392f,
-0.5902345098039217f, 0.5679356274509805f, 0.5039601960784315f,
-0.5877738823529411f, 0.5651674705882354f, 0.5014996470588235f,
-0.5853132549019608f, 0.5623993137254902f, 0.4990390980392157f,
-0.5828526274509804f, 0.5596311568627451f, 0.49657854901960785f,
-0.580392f, 0.556863f, 0.494118f,
-0.5773162352941177f, 0.5534796274509803f, 0.49104223529411767f,
-0.5742404705882354f, 0.5500962549019608f, 0.4879664705882353f,
-0.5711647058823529f, 0.5467128823529411f, 0.48489070588235295f,
-0.5680889411764707f, 0.5433295098039216f, 0.4818149411764706f,
-0.5650131764705882f, 0.5399461372549019f, 0.4787391764705882f,
-0.5619374117647059f, 0.5365627647058824f, 0.4756634117647059f,
-0.5588616470588236f, 0.5331793921568627f, 0.4725876470588235f,
-0.5557858823529411f, 0.5297960196078432f, 0.46951188235294117f,
-0.5527101176470588f, 0.5264126470588235f, 0.4664361176470588f,
-0.5496343529411765f, 0.523029274509804f, 0.46336035294117645f,
-0.5465585882352941f, 0.5196459019607843f, 0.4602845882352941f,
-0.5434828235294118f, 0.5162625294117648f, 0.4572088235294118f,
-0.5405608627450981f, 0.5130329607843137f, 0.4543637450980392f,
-0.5381003137254902f, 0.5102648039215686f, 0.45221072549019603f,
-0.5356397647058824f, 0.5074966470588235f, 0.45005770588235294f,
-0.5331792156862745f, 0.5047284901960785f, 0.4479046862745098f,
-0.5307186666666667f, 0.5019603333333333f, 0.44575166666666666f,
-0.5282581176470589f, 0.4991921764705882f, 0.4435986470588235f,
-0.525797568627451f, 0.4964240196078431f, 0.44144562745098037f,
-0.5233370196078432f, 0.4936558627450981f, 0.4392926078431373f,
-0.5208764705882353f, 0.4908877058823529f, 0.43713958823529414f,
-0.5184159215686275f, 0.48811954901960786f, 0.434986568627451f,
-0.5159553725490197f, 0.48535139215686274f, 0.43283354901960785f,
-0.5134948235294118f, 0.4825832352941176f, 0.4306805294117647f,
-0.511034274509804f, 0.47981507843137255f, 0.42852750980392157f,
-0.5090350588235294f, 0.47750829411764706f, 0.42668205882352944f,
-0.5074971764705882f, 0.4756628823529412f, 0.4251441764705883f,
-0.5059592941176471f, 0.4738174705882353f, 0.4236062941176471f,
-0.5044214117647059f, 0.4719720588235294f, 0.4220684117647059f,
-0.5028835294117647f, 0.47012664705882357f, 0.4205305294117648f,
-0.5013456470588236f, 0.46828123529411764f, 0.41899264705882355f,
-0.4998077647058824f, 0.46643582352941176f, 0.4174547647058824f,
-0.49826988235294123f, 0.4645904117647059f, 0.4159168823529412f,
-0.496732f, 0.46274499999999996f, 0.414379f,
-0.49519411764705884f, 0.4608995882352941f, 0.41284111764705883f,
-0.4936562352941177f, 0.4590541764705882f, 0.41130323529411766f,
-0.4921183529411765f, 0.45720876470588234f, 0.4097653529411765f,
-0.49058047058823534f, 0.45536335294117647f, 0.40822747058823533f,
-0.4890425882352941f, 0.45351794117647054f, 0.4066895882352941f,
-0.48750470588235295f, 0.45167252941176467f, 0.40515170588235294f,
-0.4859668235294118f, 0.4498271176470588f, 0.4036138235294118f,
-0.4844289411764706f, 0.4479817058823529f, 0.4020759411764706f,
-0.48289105882352945f, 0.44613629411764705f, 0.40053805882352944f,
-0.48135317647058823f, 0.4442908823529412f, 0.3990001764705882f,
-0.47981529411764706f, 0.4424454705882353f, 0.39746229411764705f,
-0.47827741176470595f, 0.4406000588235295f, 0.39592441176470594f,
-0.47673952941176473f, 0.4387546470588235f, 0.3943865294117647f,
-0.47520164705882356f, 0.4369092352941176f, 0.39284864705882355f,
-0.47366376470588234f, 0.43506382352941175f, 0.39131076470588233f,
-0.4721258823529412f, 0.4332184117647059f, 0.38977288235294116f,
-0.470588f, 0.431373f, 0.388235f,
-0.46905011764705884f, 0.42952750980392157f, 0.38669711764705883f,
-0.4675122352941177f, 0.42768201960784313f, 0.38515923529411766f,
-0.46597435294117645f, 0.4258365294117647f, 0.38362135294117644f,
-0.4644364705882353f, 0.42399103921568626f, 0.3820834705882353f,
-0.4628985882352941f, 0.4221455490196079f, 0.3805455882352941f,
-0.46136070588235295f, 0.42030005882352944f, 0.37900770588235294f,
-0.4598228235294118f, 0.418454568627451f, 0.3774698235294118f,
-0.45828494117647056f, 0.41660907843137257f, 0.37593194117647055f,
-0.4567470588235294f, 0.41476358823529413f, 0.3743940588235294f,
-0.45520917647058823f, 0.4129180980392157f, 0.3728561764705882f,
-0.4536712941176471f, 0.4110726078431373f, 0.3713182941176471f,
-0.4521334117647059f, 0.4092271176470588f, 0.3697804117647059f,
-0.450364862745098f, 0.40722786274509803f, 0.3681656470588235f,
-0.44790431372549017f, 0.4047673137254902f, 0.3663202352941176f,
-0.4454437647058823f, 0.40230676470588234f, 0.36447482352941174f,
-0.4429832156862745f, 0.39984621568627454f, 0.36262941176470587f,
-0.4405226666666667f, 0.3973856666666667f, 0.360784f,
-0.4380621176470588f, 0.39492511764705884f, 0.3589385882352941f,
-0.435601568627451f, 0.392464568627451f, 0.35709317647058825f,
-0.4331410196078431f, 0.39000401960784314f, 0.3552477647058824f,
-0.4306804705882353f, 0.3875434705882353f, 0.3534023529411765f,
-0.4282199215686274f, 0.38508292156862745f, 0.35155694117647057f,
-0.4257593725490196f, 0.3826223725490196f, 0.3497115294117647f,
-0.4232988235294117f, 0.38016182352941175f, 0.3478661176470588f,
-0.4208382745098039f, 0.3777012745098039f, 0.34602070588235295f,
-0.4183776862745098f, 0.3750868823529412f, 0.34402149019607847f,
-0.4159170588235295f, 0.3723186470588236f, 0.3418684705882354f,
-0.413456431372549f, 0.3695504117647059f, 0.3397154509803922f,
-0.4109958039215686f, 0.36678217647058825f, 0.33756243137254904f,
-0.40853517647058823f, 0.36401394117647057f, 0.3354094117647059f,
-0.4060745490196078f, 0.36124570588235294f, 0.33325639215686276f,
-0.40361392156862747f, 0.3584774705882353f, 0.3311033725490196f,
-0.40115329411764705f, 0.35570923529411763f, 0.32895035294117647f,
-0.39869266666666664f, 0.352941f, 0.32679733333333333f,
-0.3962320392156863f, 0.3501727647058823f, 0.3246443137254902f,
-0.39377141176470587f, 0.3474045294117647f, 0.3224912941176471f,
-0.39131078431372546f, 0.34463629411764707f, 0.32033827450980396f,
-0.3888501568627451f, 0.3418680588235294f, 0.3181852549019608f,
-0.38662023529411765f, 0.33956123529411764f, 0.31603223529411767f,
-0.3844672156862745f, 0.3374082156862745f, 0.3138792156862745f,
-0.38231419607843137f, 0.33525519607843135f, 0.3117261960784314f,
-0.3801611764705882f, 0.3331021764705882f, 0.30957317647058824f,
-0.37800815686274514f, 0.3309491568627451f, 0.30742015686274515f,
-0.37585513725490194f, 0.3287961372549019f, 0.30526713725490195f,
-0.3737021176470588f, 0.3266431176470588f, 0.3031141176470588f,
-0.3715490980392157f, 0.3244900980392157f, 0.3009610980392157f,
-0.36939607843137257f, 0.32233707843137255f, 0.2988080784313726f,
-0.3672430588235294f, 0.3201840588235294f, 0.29665505882352944f,
-0.3650900392156863f, 0.31803103921568626f, 0.2945020392156863f,
-0.36293701960784314f, 0.3158780196078431f, 0.29234901960784315f,
-0.360784f, 0.313725f, 0.290196f,
-0.35832345098039214f, 0.3112644509803921f, 0.28835058823529414f,
-0.3558629019607843f, 0.3088039019607843f, 0.28650517647058826f,
-0.35340235294117645f, 0.3063433529411765f, 0.28465976470588233f,
-0.3509418039215686f, 0.30388280392156863f, 0.28281435294117646f,
-0.3484812549019608f, 0.3014222549019608f, 0.2809689411764706f,
-0.34602070588235295f, 0.29896170588235293f, 0.2791235294117647f,
-0.3435601568627451f, 0.2965011568627451f, 0.27727811764705884f,
-0.3410996078431373f, 0.29404060784313735f, 0.27543270588235297f,
-0.3386390588235294f, 0.29158005882352944f, 0.27358729411764704f,
-0.33617850980392155f, 0.2891195098039216f, 0.27174188235294117f,
-0.3337179607843137f, 0.28665896078431374f, 0.2698964705882353f,
-0.33125741176470586f, 0.2841984117647059f, 0.2680510588235294f,
-0.3288737450980392f, 0.2817378431372549f, 0.2661287450980392f,
-0.32672072549019604f, 0.2792772156862745f, 0.26397572549019604f,
-0.3245677058823529f, 0.27681658823529415f, 0.26182270588235296f,
-0.3224146862745098f, 0.27435596078431373f, 0.2596696862745098f,
-0.32026166666666667f, 0.2718953333333333f, 0.25751666666666667f,
-0.3181086470588235f, 0.26943470588235297f, 0.25536364705882353f,
-0.3159556274509804f, 0.26697407843137255f, 0.2532106274509804f,
-0.31380260784313724f, 0.26451345098039214f, 0.25105760784313724f,
-0.3116495882352941f, 0.2620528235294118f, 0.24890458823529413f,
-0.30949656862745095f, 0.25959219607843137f, 0.24675156862745098f,
-0.3073435490196078f, 0.25713156862745096f, 0.24459854901960784f,
-0.3051905294117647f, 0.25467094117647066f, 0.24244552941176478f,
-0.3030375098039215f, 0.2522103137254902f, 0.24029250980392158f,
-0.2999617450980392f, 0.2491345882352941f, 0.23767811764705885f,
-0.2959632352941176f, 0.24544376470588233f, 0.23460235294117648f,
-0.29196472549019603f, 0.24175294117647056f, 0.23152658823529412f,
-0.2879662156862745f, 0.2380621176470588f, 0.22845082352941176f,
-0.2839677058823529f, 0.23437129411764704f, 0.22537505882352943f,
-0.27996919607843135f, 0.23068047058823526f, 0.22229929411764707f,
-0.27597068627450977f, 0.22698964705882352f, 0.2192235294117647f,
-0.2719721764705882f, 0.22329882352941174f, 0.21614776470588237f,
-0.26797366666666667f, 0.219608f, 0.213072f,
-0.2639751568627451f, 0.21591717647058822f, 0.20999623529411765f,
-0.2599766470588235f, 0.21222635294117648f, 0.20692047058823532f,
-0.2559781372549019f, 0.2085355294117647f, 0.20384470588235296f,
-0.25197962745098035f, 0.20484470588235293f, 0.2007689411764706f,
-0.2491345882352941f, 0.20253788235294118f, 0.1986159411764706f,
-0.2466740392156863f, 0.2006923921568628f, 0.19677052941176476f,
-0.2442134901960784f, 0.1988469019607843f, 0.19492511764705883f,
-0.24175294117647056f, 0.19700141176470587f, 0.19307970588235296f,
-0.23929239215686274f, 0.19515592156862743f, 0.19123429411764706f,
-0.2368318431372549f, 0.19331043137254902f, 0.18938888235294118f,
-0.23437129411764704f, 0.1914649411764706f, 0.18754347058823528f,
-0.23191074509803922f, 0.18961945098039215f, 0.1856980588235294f,
-0.22945019607843137f, 0.18777396078431371f, 0.18385264705882354f,
-0.22698964705882352f, 0.1859284705882353f, 0.18200723529411764f,
-0.22452909803921567f, 0.18408298039215687f, 0.18016182352941176f,
-0.22206854901960785f, 0.18223749019607843f, 0.17831641176470586f,
-0.219608f, 0.180392f, 0.176471f
-}
-, 
-{
-1.0f, 0.933333f, 0.8f,
-0.996078431372549f, 0.9291807647058823f, 0.7936947254901962f,
-0.9921568627450981f, 0.9250285294117647f, 0.7873894509803921f,
-0.9882352941176471f, 0.920876294117647f, 0.7810841764705883f,
-0.9843137254901961f, 0.9167240588235294f, 0.7747789019607844f,
-0.9803921568627451f, 0.9125718235294117f, 0.7684736274509805f,
-0.9764705882352941f, 0.9084195882352941f, 0.7621683529411765f,
-0.9725490196078431f, 0.9042673529411764f, 0.7558630784313726f,
-0.9686274509803922f, 0.9001151176470588f, 0.7495578039215687f,
-0.9647058823529412f, 0.8959628823529412f, 0.7432525294117648f,
-0.9607843137254902f, 0.8918106470588235f, 0.7369472549019608f,
-0.9568627450980393f, 0.8876584117647058f, 0.7306419803921569f,
-0.9529411764705883f, 0.8835061764705883f, 0.724336705882353f,
-0.9490196078431372f, 0.8793539411764706f, 0.7180314313725491f,
-0.9450980392156862f, 0.8752017058823529f, 0.7117261568627451f,
-0.9411764705882353f, 0.8710494705882352f, 0.7054208823529412f,
-0.9372549019607843f, 0.8668972352941177f, 0.6991156078431373f,
-0.9333333333333333f, 0.862745f, 0.6928103333333333f,
-0.9294117647058824f, 0.8585927647058823f, 0.6865050588235294f,
-0.9254901960784314f, 0.8544405294117647f, 0.6801997843137255f,
-0.9215686274509804f, 0.8502882941176471f, 0.6738945098039216f,
-0.9176470588235295f, 0.8461360588235294f, 0.6675892352941177f,
-0.9137254901960784f, 0.8419838235294117f, 0.6612839607843137f,
-0.9098039215686274f, 0.8378315882352941f, 0.6549786862745098f,
-0.9058823529411765f, 0.8336793529411765f, 0.6486734117647059f,
-0.9019607843137255f, 0.8295271176470588f, 0.6423681372549019f,
-0.8980392156862745f, 0.8253748823529412f, 0.636062862745098f,
-0.8941176470588236f, 0.8212226470588235f, 0.6297575882352942f,
-0.8901960784313726f, 0.8170704117647059f, 0.6234523137254901f,
-0.8862745098039216f, 0.8129181764705883f, 0.6171470392156864f,
-0.8823529411764706f, 0.8087659411764706f, 0.6108417647058824f,
-0.8784313725490196f, 0.8046137058823529f, 0.6045364901960785f,
-0.8745098039215686f, 0.8004614705882354f, 0.5982312156862746f,
-0.8705882352941177f, 0.7963092352941177f, 0.5919259411764707f,
-0.8666666666666667f, 0.792157f, 0.5856206666666667f,
-0.8627450980392157f, 0.7880047647058823f, 0.5793153921568628f,
-0.8588235294117648f, 0.7838525294117646f, 0.5730101176470588f,
-0.8549019607843138f, 0.7797002941176471f, 0.566704843137255f,
-0.8509803921568628f, 0.7755480588235294f, 0.560399568627451f,
-0.8470588235294119f, 0.7713958235294118f, 0.5540942941176471f,
-0.8431372549019608f, 0.7672435882352941f, 0.5477890196078432f,
-0.8392156862745099f, 0.7630913529411765f, 0.5414837450980392f,
-0.8352941176470589f, 0.7589391176470588f, 0.5351784705882353f,
-0.8313725490196079f, 0.7547868823529412f, 0.5288731960784314f,
-0.8274509803921569f, 0.7506346470588235f, 0.5225679215686274f,
-0.823529411764706f, 0.746482411764706f, 0.5162626470588236f,
-0.8196078431372549f, 0.7423301764705883f, 0.5099573725490196f,
-0.8156862745098039f, 0.7381779411764706f, 0.5036520980392156f,
-0.8117647058823529f, 0.7340257058823529f, 0.4973468235294118f,
-0.807843137254902f, 0.7298734705882354f, 0.4910415490196079f,
-0.803921568627451f, 0.7257212352941177f, 0.48473627450980394f,
-0.8f, 0.721569f, 0.478431f,
-0.7960784313725491f, 0.7181856862745098f, 0.4749708039215686f,
-0.7921568627450981f, 0.7148023725490197f, 0.47151060784313725f,
-0.7882352941176471f, 0.7114190588235294f, 0.46805041176470585f,
-0.7843137254901961f, 0.7080357450980392f, 0.4645902156862745f,
-0.7803921568627451f, 0.7046524313725491f, 0.4611300196078431f,
-0.7764705882352941f, 0.7012691176470589f, 0.45766982352941177f,
-0.7725490196078432f, 0.6978858039215686f, 0.45420962745098037f,
-0.7686274509803922f, 0.6945024901960785f, 0.450749431372549f,
-0.7647058823529412f, 0.6911191764705882f, 0.4472892352941176f,
-0.7607843137254903f, 0.687735862745098f, 0.4438290392156863f,
-0.7568627450980392f, 0.6843525490196078f, 0.4403688431372549f,
-0.7529411764705882f, 0.6809692352941177f, 0.43690864705882354f,
-0.7490196078431373f, 0.6775859215686274f, 0.43344845098039214f,
-0.7450980392156863f, 0.6742026078431372f, 0.4299882549019608f,
-0.7411764705882353f, 0.6708192941176471f, 0.42652805882352945f,
-0.7372549019607844f, 0.6674359803921569f, 0.423067862745098f,
-0.7333333333333334f, 0.6640526666666666f, 0.41960766666666666f,
-0.7294117647058824f, 0.6606693529411765f, 0.41614747058823526f,
-0.7254901960784313f, 0.6572860392156863f, 0.4126872745098039f,
-0.7215686274509804f, 0.6539027254901961f, 0.4092270784313725f,
-0.7176470588235294f, 0.6505194117647058f, 0.40576688235294117f,
-0.7137254901960784f, 0.6471360980392157f, 0.4023066862745098f,
-0.7098039215686275f, 0.6437527843137255f, 0.3988464901960785f,
-0.7058823529411765f, 0.6403694705882352f, 0.39538629411764703f,
-0.7019607843137255f, 0.6369861568627451f, 0.3919260980392157f,
-0.6980392156862745f, 0.6336028431372549f, 0.3884659019607843f,
-0.6941176470588235f, 0.6302195294117647f, 0.38500570588235294f,
-0.6901960784313725f, 0.6268362156862745f, 0.38154550980392155f,
-0.6862745098039216f, 0.6234529019607843f, 0.3780853137254902f,
-0.6823529411764706f, 0.6200695882352941f, 0.3746251176470588f,
-0.6784313725490196f, 0.616686274509804f, 0.37116492156862746f,
-0.6745098039215687f, 0.6133029607843137f, 0.36770472549019606f,
-0.6705882352941177f, 0.6099196470588235f, 0.36424452941176466f,
-0.6666666666666667f, 0.6065363333333333f, 0.3607843333333333f,
-0.6627450980392157f, 0.6031530196078431f, 0.357324137254902f,
-0.6588235294117647f, 0.5997697058823529f, 0.3538639411764706f,
-0.6549019607843137f, 0.5963863921568627f, 0.3504037450980392f,
-0.6509803921568628f, 0.5930030784313725f, 0.34694354901960783f,
-0.6470588235294118f, 0.5896197647058824f, 0.3434833529411765f,
-0.6431372549019607f, 0.5862364509803921f, 0.3400231568627451f,
-0.6392156862745098f, 0.582853137254902f, 0.3365629607843137f,
-0.6352941176470588f, 0.5794698235294118f, 0.33310276470588235f,
-0.6313725490196078f, 0.5760865098039215f, 0.32964256862745095f,
-0.6274509803921569f, 0.5727031960784313f, 0.3261823725490196f,
-0.6235294117647059f, 0.5693198823529412f, 0.3227221764705882f,
-0.6196078431372549f, 0.5659365686274509f, 0.3192619803921568f,
-0.615686274509804f, 0.5625532549019607f, 0.3158017843137255f,
-0.611764705882353f, 0.5591699411764706f, 0.3123415882352941f,
-0.6078431372549019f, 0.5557866274509804f, 0.3088813921568627f,
-0.6039215686274509f, 0.5524033137254901f, 0.3054211960784313f,
-0.6f, 0.54902f, 0.301961f,
-0.596078431372549f, 0.5457135686274509f, 0.29919282352941173f,
-0.592156862745098f, 0.5424071372549019f, 0.2964246470588235f,
-0.5882352941176471f, 0.5391007058823529f, 0.2936564705882353f,
-0.5843137254901961f, 0.5357942745098039f, 0.2908882941176471f,
-0.580392156862745f, 0.5324878431372548f, 0.2881201176470588f,
-0.5764705882352941f, 0.5291814117647058f, 0.28535194117647056f,
-0.5725490196078431f, 0.5258749803921569f, 0.2825837647058823f,
-0.5686274509803921f, 0.5225685490196078f, 0.2798155882352941f,
-0.5647058823529412f, 0.5192621176470588f, 0.2770474117647059f,
-0.5607843137254902f, 0.5159556862745097f, 0.27427923529411763f,
-0.5568627450980392f, 0.5126492549019608f, 0.2715110588235294f,
-0.5529411764705883f, 0.5093428235294117f, 0.2687428823529412f,
-0.5490196078431373f, 0.5060363921568627f, 0.26597470588235295f,
-0.5450980392156862f, 0.5027299607843136f, 0.2632065294117647f,
-0.5411764705882353f, 0.49942352941176466f, 0.26043835294117645f,
-0.5372549019607843f, 0.49611709803921566f, 0.2576701764705882f,
-0.5333333333333333f, 0.4928106666666666f, 0.25490199999999996f,
-0.5294117647058824f, 0.48950423529411763f, 0.25213382352941177f,
-0.5254901960784314f, 0.4861978039215686f, 0.24936564705882353f,
-0.5215686274509804f, 0.48289137254901965f, 0.24659747058823533f,
-0.5176470588235295f, 0.47958494117647055f, 0.24382929411764706f,
-0.5137254901960784f, 0.47627850980392156f, 0.24106111764705881f,
-0.5098039215686274f, 0.4729720784313725f, 0.2382929411764706f,
-0.5058823529411764f, 0.4696656470588235f, 0.23552476470588235f,
-0.5019607843137255f, 0.4663592156862745f, 0.2327565882352941f,
-0.4980392156862745f, 0.4630527843137255f, 0.22998841176470589f,
-0.49411764705882355f, 0.45974635294117644f, 0.22722023529411764f,
-0.49019607843137253f, 0.45643992156862745f, 0.22445205882352942f,
-0.48627450980392156f, 0.4531334901960784f, 0.22168388235294118f,
-0.4823529411764707f, 0.44982705882352947f, 0.218915705882353f,
-0.47843137254901963f, 0.44652062745098037f, 0.2161475294117647f,
-0.4745098039215686f, 0.4432141960784314f, 0.21337935294117646f,
-0.47058823529411764f, 0.43990776470588233f, 0.21061117647058825f,
-0.4666666666666667f, 0.43660133333333334f, 0.207843f,
-0.4627450980392157f, 0.4332949019607843f, 0.20507482352941175f,
-0.45882352941176474f, 0.4299884705882353f, 0.2023066470588235f,
-0.4549019607843138f, 0.42668203921568626f, 0.19953847058823532f,
-0.45098039215686275f, 0.4233756078431372f, 0.19677029411764707f,
-0.4470588235294118f, 0.4200691764705883f, 0.19400211764705882f,
-0.44313725490196076f, 0.41676274509803923f, 0.19123394117647058f,
-0.4392156862745098f, 0.4134563137254902f, 0.18846576470588233f,
-0.43529411764705883f, 0.41014988235294114f, 0.18569758823529414f,
-0.43137254901960786f, 0.40684345098039215f, 0.1829294117647059f,
-0.4274509803921569f, 0.40353701960784316f, 0.18016123529411765f,
-0.42352941176470593f, 0.4002305882352941f, 0.1773930588235294f,
-0.419607843137255f, 0.3969241568627452f, 0.17462488235294127f,
-0.41568627450980394f, 0.3936177254901961f, 0.17185670588235297f,
-0.411764705882353f, 0.3903112941176471f, 0.16908852941176472f,
-0.40784313725490196f, 0.38700486274509804f, 0.16632035294117647f,
-0.403921568627451f, 0.383698431372549f, 0.16355217647058826f,
-0.4f, 0.380392f, 0.160784f,
-0.39807766666666666f, 0.37885413725490197f, 0.1587847843137255f,
-0.39615533333333336f, 0.37731627450980393f, 0.156785568627451f,
-0.394233f, 0.3757784117647059f, 0.15478635294117649f,
-0.3923106666666667f, 0.37424054901960785f, 0.15278713725490198f,
-0.39038833333333334f, 0.3727026862745098f, 0.15078792156862747f,
-0.38846600000000003f, 0.37116482352941177f, 0.14878870588235296f,
-0.3865436666666667f, 0.36962696078431373f, 0.14678949019607845f,
-0.38462133333333337f, 0.3680890980392157f, 0.14479027450980392f,
-0.382699f, 0.36655123529411765f, 0.1427910588235294f,
-0.38077666666666665f, 0.3650133725490196f, 0.1407918431372549f,
-0.3788543333333334f, 0.3634755098039216f, 0.13879262745098045f,
-0.376932f, 0.36193764705882353f, 0.13679341176470589f,
-0.3750096666666667f, 0.3603997843137255f, 0.13479419607843138f,
-0.3730873333333333f, 0.35886192156862745f, 0.13279498039215687f,
-0.371165f, 0.3573240588235294f, 0.13079576470588236f,
-0.36924266666666666f, 0.35578619607843137f, 0.12879654901960785f,
-0.36732033333333336f, 0.35424833333333333f, 0.12679733333333335f,
-0.365398f, 0.3527104705882353f, 0.12479811764705884f,
-0.36347566666666664f, 0.35117260784313725f, 0.12279890196078433f,
-0.36155333333333334f, 0.3496347450980392f, 0.12079968627450982f,
-0.35963100000000003f, 0.3480968823529412f, 0.11880047058823531f,
-0.3577086666666667f, 0.34655901960784313f, 0.11680125490196079f,
-0.3557863333333333f, 0.3450211568627451f, 0.11480203921568628f,
-0.353864f, 0.34348329411764705f, 0.11280282352941177f,
-0.35194166666666665f, 0.341945431372549f, 0.11080360784313725f,
-0.35001933333333335f, 0.340407568627451f, 0.10880439215686274f,
-0.34809700000000005f, 0.338869705882353f, 0.10680517647058829f,
-0.34617466666666663f, 0.3373318431372549f, 0.10480596078431373f,
-0.3442523333333333f, 0.33579398039215685f, 0.10280674509803922f,
-0.34233f, 0.3342561176470588f, 0.10080752941176471f,
-0.34040766666666666f, 0.3327182549019608f, 0.0988083137254902f,
-0.3384853333333333f, 0.33118039215686274f, 0.0968090980392157f,
-0.336563f, 0.3296425294117647f, 0.09480988235294117f,
-0.3346406666666667f, 0.32810466666666666f, 0.09281066666666668f,
-0.33271833333333334f, 0.3265668039215686f, 0.09081145098039216f,
-0.330796f, 0.3250289411764706f, 0.08881223529411765f,
-0.3288736666666667f, 0.32349107843137254f, 0.08681301960784314f,
-0.3269513333333333f, 0.3219532156862745f, 0.08481380392156863f,
-0.325029f, 0.32041535294117646f, 0.08281458823529413f,
-0.32310666666666665f, 0.3188774901960784f, 0.08081537254901962f,
-0.3211843333333333f, 0.3173396274509804f, 0.0788161568627451f,
-0.319262f, 0.31580176470588234f, 0.0768169411764706f,
-0.3173396666666667f, 0.31426390196078435f, 0.07481772549019614f,
-0.3154173333333333f, 0.31272603921568626f, 0.07281850980392157f,
-0.31349499999999997f, 0.3111881764705882f, 0.07081929411764706f,
-0.31157266666666666f, 0.3096503137254902f, 0.06882007843137256f,
-0.3096503333333333f, 0.30811245098039214f, 0.06682086274509805f,
-0.307728f, 0.3065745882352941f, 0.06482164705882354f,
-0.30580566666666664f, 0.30503672549019606f, 0.06282243137254902f,
-0.3038833333333333f, 0.303498862745098f, 0.06082321568627451f,
-0.301961f, 0.301961f, 0.058824f,
-0.29965419607843136f, 0.2999617647058823f, 0.05767058823529412f,
-0.29734739215686273f, 0.2979625294117647f, 0.05651717647058824f,
-0.2950405882352941f, 0.29596329411764705f, 0.055363764705882355f,
-0.2927337843137255f, 0.2939640588235294f, 0.05421035294117647f,
-0.29042698039215686f, 0.2919648235294117f, 0.05305694117647059f,
-0.28812017647058824f, 0.2899655882352941f, 0.05190352941176471f,
-0.2858133725490196f, 0.28796635294117645f, 0.05075011764705882f,
-0.28350656862745105f, 0.28596711764705884f, 0.04959670588235297f,
-0.2811997647058823f, 0.2839678823529412f, 0.04844329411764706f,
-0.2788929607843137f, 0.2819686470588235f, 0.04728988235294118f,
-0.27658615686274507f, 0.27996941176470586f, 0.04613647058823529f,
-0.27427935294117645f, 0.27797017647058825f, 0.04498305882352941f,
-0.2719725490196078f, 0.2759709411764706f, 0.04382964705882353f,
-0.2696657450980392f, 0.2739717058823529f, 0.04267623529411765f,
-0.2673589411764706f, 0.27197247058823526f, 0.041522823529411765f,
-0.26505213725490195f, 0.26997323529411765f, 0.04036941176470588f,
-0.26274533333333333f, 0.267974f, 0.039216f,
-0.2604385294117647f, 0.2659747647058823f, 0.03806258823529411f,
-0.2581317254901961f, 0.2639755294117647f, 0.03690917647058824f,
-0.25582492156862746f, 0.26197629411764706f, 0.035755764705882355f,
-0.25351811764705884f, 0.2599770588235294f, 0.03460235294117647f,
-0.2512113137254902f, 0.25797782352941173f, 0.033448941176470584f,
-0.24890450980392156f, 0.2559785882352941f, 0.03229552941176471f,
-0.246597705882353f, 0.2539793529411765f, 0.03114211764705886f,
-0.24429090196078432f, 0.2519801176470588f, 0.02998870588235294f,
-0.24198409803921567f, 0.2499808823529412f, 0.028835294117647063f,
-0.23967729411764704f, 0.24798164705882353f, 0.027681882352941178f,
-0.23737049019607842f, 0.24598241176470587f, 0.026528470588235292f,
-0.2350636862745098f, 0.24398317647058823f, 0.02537505882352941f,
-0.23275688235294117f, 0.2419839411764706f, 0.02422164705882353f,
-0.23045007843137255f, 0.23998470588235293f, 0.023068235294117653f,
-0.22814327450980393f, 0.2379854705882353f, 0.021914823529411764f,
-0.2258364705882353f, 0.23598623529411766f, 0.020761411764705882f,
-0.22352966666666668f, 0.233987f, 0.019608f,
-0.22122286274509803f, 0.23198776470588234f, 0.01845458823529412f,
-0.2189160588235294f, 0.2299885294117647f, 0.01730117647058823f,
-0.21660925490196078f, 0.22798929411764707f, 0.016147764705882355f,
-0.21430245098039213f, 0.2259900588235294f, 0.014994352941176473f,
-0.21199564705882354f, 0.22399082352941177f, 0.01384094117647059f,
-0.20968884313725497f, 0.2219915882352942f, 0.012687529411764736f,
-0.20738203921568626f, 0.21999235294117647f, 0.01153411764705882f,
-0.20507523529411764f, 0.21799311764705884f, 0.010380705882352945f,
-0.20276843137254902f, 0.2159938823529412f, 0.009227294117647056f,
-0.2004616274509804f, 0.21399464705882354f, 0.008073882352941174f,
-0.19815482352941177f, 0.21199541176470588f, 0.006920470588235299f,
-0.19584801960784315f, 0.20999617647058824f, 0.00576705882352941f,
-0.19354121568627453f, 0.2079969411764706f, 0.004613647058823535f,
-0.1912344117647059f, 0.20599770588235294f, 0.003460235294117646f,
-0.18892760784313725f, 0.20399847058823528f, 0.002306823529411764f,
-0.18662080392156866f, 0.20199923529411767f, 0.001153411764705882f,
-0.184314f, 0.2f, 0.0f
-}
-, 
-{
-0.054902f, 0.109804f, 0.121569f,
-0.05643988235294117f, 0.11472517647058823f, 0.12618256862745098f,
-0.057977764705882354f, 0.11964635294117647f, 0.13079613725490197f,
-0.059515647058823534f, 0.1245675294117647f, 0.13540970588235293f,
-0.06105352941176471f, 0.12948870588235295f, 0.14002327450980392f,
-0.06259141176470588f, 0.13440988235294118f, 0.1446368431372549f,
-0.06412929411764706f, 0.13933105882352942f, 0.14925041176470588f,
-0.06566717647058824f, 0.14425223529411765f, 0.15386398039215687f,
-0.06720505882352942f, 0.1491734117647059f, 0.15847754901960784f,
-0.06874294117647059f, 0.15409458823529412f, 0.16309111764705883f,
-0.07028082352941177f, 0.15901576470588236f, 0.1677046862745098f,
-0.07181870588235295f, 0.1639369411764706f, 0.1723182549019608f,
-0.07335658823529412f, 0.16885811764705883f, 0.17693182352941175f,
-0.0747406862745098f, 0.17370241176470588f, 0.18116094117647058f,
-0.07566343137254902f, 0.1783160588235294f, 0.18423670588235294f,
-0.07658617647058824f, 0.18292970588235294f, 0.1873124705882353f,
-0.07750892156862746f, 0.18754335294117647f, 0.19038823529411764f,
-0.07843166666666668f, 0.192157f, 0.193464f,
-0.07935441176470588f, 0.19677064705882352f, 0.19653976470588236f,
-0.0802771568627451f, 0.20138429411764705f, 0.1996155294117647f,
-0.08119990196078432f, 0.20599794117647058f, 0.20269129411764705f,
-0.08212264705882354f, 0.21061158823529413f, 0.20576705882352941f,
-0.08304539215686275f, 0.21522523529411763f, 0.20884282352941175f,
-0.08396813725490196f, 0.2198388823529412f, 0.2119185882352941f,
-0.08489088235294118f, 0.2244525294117647f, 0.21499435294117647f,
-0.0858136274509804f, 0.22906617647058825f, 0.21807011764705883f,
-0.08658256862745098f, 0.23321841176470587f, 0.22083827450980392f,
-0.08719770588235294f, 0.23690923529411764f, 0.22329882352941177f,
-0.0878128431372549f, 0.24060005882352942f, 0.2257593725490196f,
-0.08842798039215687f, 0.24429088235294116f, 0.22821992156862744f,
-0.08904311764705883f, 0.24798170588235294f, 0.2306804705882353f,
-0.08965825490196079f, 0.2516725294117647f, 0.23314101960784311f,
-0.09027339215686274f, 0.25536335294117646f, 0.23560156862745096f,
-0.0908885294117647f, 0.2590541764705882f, 0.23806211764705879f,
-0.09150366666666666f, 0.262745f, 0.24052266666666666f,
-0.09211880392156863f, 0.26643582352941175f, 0.24298321568627448f,
-0.09273394117647059f, 0.2701266470588235f, 0.24544376470588233f,
-0.09334907843137254f, 0.27381747058823525f, 0.24790431372549016f,
-0.0939642156862745f, 0.27750829411764705f, 0.250364862745098f,
-0.09504070588235293f, 0.28258329411764704f, 0.2525947647058823f,
-0.09627098039215685f, 0.2881196862745098f, 0.25474778431372547f,
-0.09750125490196077f, 0.2936560784313725f, 0.2569008039215686f,
-0.0987315294117647f, 0.2991924705882353f, 0.25905382352941175f,
-0.09996180392156863f, 0.30472886274509803f, 0.2612068431372549f,
-0.10119207843137254f, 0.3102652549019608f, 0.26335986274509804f,
-0.10242235294117646f, 0.31580164705882346f, 0.2655128823529411f,
-0.10365262745098039f, 0.32133803921568627f, 0.26766590196078427f,
-0.10488290196078431f, 0.326874431372549f, 0.2698189215686274f,
-0.10611317647058824f, 0.33241082352941176f, 0.27197194117647056f,
-0.10734345098039215f, 0.33794721568627445f, 0.2741249607843137f,
-0.10857372549019607f, 0.34348360784313725f, 0.27627798039215684f,
-0.109804f, 0.34902f, 0.278431f,
-0.1101115294117647f, 0.353018431372549f, 0.278431f,
-0.11041905882352941f, 0.35701686274509803f, 0.278431f,
-0.11072658823529412f, 0.36101529411764705f, 0.278431f,
-0.11103411764705882f, 0.36501372549019606f, 0.278431f,
-0.11134164705882353f, 0.36901215686274513f, 0.278431f,
-0.11164917647058824f, 0.3730105882352941f, 0.278431f,
-0.11195670588235294f, 0.37700901960784317f, 0.278431f,
-0.11226423529411765f, 0.3810074509803922f, 0.278431f,
-0.11257176470588236f, 0.3850058823529412f, 0.278431f,
-0.11287929411764706f, 0.38900431372549016f, 0.278431f,
-0.11318682352941177f, 0.39300274509803923f, 0.278431f,
-0.11349435294117648f, 0.39700117647058825f, 0.278431f,
-0.11380190196078432f, 0.40099960784313726f, 0.2782772156862745f,
-0.11410950980392158f, 0.4049980392156863f, 0.2776620784313725f,
-0.11441711764705882f, 0.40899647058823524f, 0.2770469411764706f,
-0.11472472549019608f, 0.4129949019607843f, 0.27643180392156863f,
-0.11503233333333333f, 0.4169933333333333f, 0.27581666666666665f,
-0.11533994117647059f, 0.42099176470588234f, 0.2752015294117647f,
-0.11564754901960785f, 0.42499019607843136f, 0.27458639215686276f,
-0.1159551568627451f, 0.4289886274509804f, 0.2739712549019608f,
-0.11626276470588236f, 0.4329870588235294f, 0.2733561176470588f,
-0.11657037254901961f, 0.4369854901960784f, 0.27274098039215683f,
-0.11687798039215685f, 0.4409839215686274f, 0.2721258431372549f,
-0.11718558823529412f, 0.4449823529411765f, 0.27151070588235293f,
-0.11749319607843138f, 0.44898078431372546f, 0.27089556862745096f,
-0.117647f, 0.45251788235294116f, 0.26951149019607845f,
-0.117647f, 0.45559364705882355f, 0.2673584705882353f,
-0.117647f, 0.4586694117647059f, 0.26520545098039217f,
-0.117647f, 0.4617451764705882f, 0.263052431372549f,
-0.117647f, 0.4648209411764706f, 0.2608994117647059f,
-0.117647f, 0.4678967058823529f, 0.2587463921568628f,
-0.117647f, 0.4709724705882353f, 0.2565933725490196f,
-0.117647f, 0.47404823529411766f, 0.25444035294117645f,
-0.117647f, 0.477124f, 0.2522873333333333f,
-0.117647f, 0.4801997647058824f, 0.25013431372549016f,
-0.117647f, 0.4832755294117647f, 0.24798129411764705f,
-0.117647f, 0.4863512941176471f, 0.2458282745098039f,
-0.117647f, 0.48942705882352944f, 0.2436752549019608f,
-0.11741629411764706f, 0.4920414705882353f, 0.24083023529411768f,
-0.11710868627450981f, 0.4945020980392157f, 0.23775454901960782f,
-0.11680107843137255f, 0.4969627254901961f, 0.23467886274509803f,
-0.1164934705882353f, 0.49942335294117646f, 0.23160317647058823f,
-0.11618586274509804f, 0.5018839803921569f, 0.22852749019607843f,
-0.11587825490196078f, 0.5043446078431373f, 0.2254518039215686f,
-0.11557064705882353f, 0.5068052352941176f, 0.2223761176470588f,
-0.11526303921568629f, 0.509265862745098f, 0.219300431372549f,
-0.11495543137254903f, 0.5117264901960784f, 0.21622474509803924f,
-0.11464782352941177f, 0.5141871176470588f, 0.2131490588235294f,
-0.11434021568627452f, 0.5166477450980391f, 0.2100733725490196f,
-0.11403260784313726f, 0.5191083725490195f, 0.2069976862745098f,
-0.113725f, 0.521569f, 0.203922f,
-0.1134174705882353f, 0.523722019607843f, 0.1999234901960784f,
-0.1131099411764706f, 0.5258750392156862f, 0.19592498039215686f,
-0.11280241176470589f, 0.5280280588235293f, 0.19192647058823528f,
-0.11249488235294118f, 0.5301810784313725f, 0.18792796078431379f,
-0.11218735294117647f, 0.5323340980392156f, 0.18392945098039215f,
-0.11187982352941177f, 0.5344871176470588f, 0.17993094117647057f,
-0.11157229411764706f, 0.5366401372549019f, 0.17593243137254902f,
-0.11126476470588235f, 0.5387931568627451f, 0.17193392156862744f,
-0.11095723529411765f, 0.5409461764705882f, 0.1679354117647059f,
-0.11064970588235294f, 0.5430991960784313f, 0.1639369019607843f,
-0.11034217647058824f, 0.5452522156862745f, 0.15993839215686273f,
-0.11003464705882353f, 0.5474052352941176f, 0.15593988235294123f,
-0.10926574509803921f, 0.5497889215686274f, 0.15155692156862743f,
-0.10711272549019607f, 0.5528646078431372f, 0.14602060784313725f,
-0.10495970588235294f, 0.555940294117647f, 0.14048429411764704f,
-0.1028066862745098f, 0.5590159803921568f, 0.13494798039215686f,
-0.10065366666666667f, 0.5620916666666667f, 0.12941166666666665f,
-0.09850064705882353f, 0.5651673529411764f, 0.12387535294117646f,
-0.09634762745098038f, 0.5682430392156862f, 0.11833903921568627f,
-0.09419460784313728f, 0.571318725490196f, 0.11280272549019615f,
-0.09204158823529411f, 0.5743944117647058f, 0.10726641176470587f,
-0.08988856862745098f, 0.5774700980392157f, 0.10173009803921568f,
-0.08773554901960784f, 0.5805457843137254f, 0.09619378431372548f,
-0.0855825294117647f, 0.5836214705882352f, 0.09065747058823528f,
-0.08342950980392157f, 0.586697156862745f, 0.08512115686274509f,
-0.08342950980392157f, 0.5899266862745097f, 0.0811226862745098f,
-0.0855825294117647f, 0.5933100588235294f, 0.0786620588235294f,
-0.08773554901960784f, 0.5966934313725489f, 0.07620143137254902f,
-0.08988856862745098f, 0.6000768039215686f, 0.07374080392156862f,
-0.09204158823529406f, 0.6034601764705881f, 0.0712801764705883f,
-0.09419460784313725f, 0.6068435490196078f, 0.06881954901960784f,
-0.09634762745098038f, 0.6102269215686275f, 0.06635892156862745f,
-0.09850064705882353f, 0.613610294117647f, 0.06389829411764705f,
-0.10065366666666667f, 0.6169936666666667f, 0.06143766666666667f,
-0.1028066862745098f, 0.6203770392156862f, 0.05897703921568627f,
-0.10495970588235294f, 0.6237604117647059f, 0.056516411764705884f,
-0.10711272549019607f, 0.6271437843137254f, 0.054055784313725486f,
-0.10926574509803921f, 0.6305271568627451f, 0.0515951568627451f,
-0.11580170588235295f, 0.6341411176470588f, 0.05282547058823529f,
-0.12379864705882353f, 0.6378319411764706f, 0.055286098039215685f,
-0.1317955882352941f, 0.6415227647058823f, 0.05774672549019608f,
-0.1397925294117647f, 0.6452135882352941f, 0.06020735294117647f,
-0.1477894705882353f, 0.6489044117647058f, 0.06266798039215686f,
-0.1557864117647059f, 0.6525952352941177f, 0.06512860784313726f,
-0.16378335294117646f, 0.6562860588235294f, 0.06758923529411764f,
-0.17178029411764684f, 0.659976882352941f, 0.07004986274509796f,
-0.17977723529411765f, 0.6636677058823529f, 0.07251049019607843f,
-0.18777417647058825f, 0.6673585294117647f, 0.07497111764705881f,
-0.19577111764705885f, 0.6710493529411765f, 0.07743174509803921f,
-0.2037680588235294f, 0.6747401764705883f, 0.0798923725490196f,
-0.211765f, 0.678431f, 0.082353f,
-0.2200694705882353f, 0.6818143725490197f, 0.08481354901960785f,
-0.22837394117647059f, 0.6851977450980392f, 0.08727409803921568f,
-0.2366784117647059f, 0.6885811176470589f, 0.08973464705882353f,
-0.2449828823529412f, 0.6919644901960784f, 0.09219519607843137f,
-0.2532873529411765f, 0.6953478627450981f, 0.09465574509803922f,
-0.2615918235294118f, 0.6987312352941176f, 0.09711629411764706f,
-0.2698962941176471f, 0.7021146078431373f, 0.0995768431372549f,
-0.2782007647058824f, 0.7054979803921568f, 0.10203739215686275f,
-0.28650523529411764f, 0.7088813529411765f, 0.1044979411764706f,
-0.29480970588235295f, 0.7122647254901961f, 0.10695849019607843f,
-0.303114176470588f, 0.7156480980392156f, 0.10941903921568621f,
-0.31141864705882355f, 0.7190314705882354f, 0.11187958823529412f,
-0.31987692156862746f, 0.7223379215686274f, 0.11464772549019608f,
-0.32879660784313725f, 0.7254136078431372f, 0.1183386274509804f,
-0.3377162941176471f, 0.7284892941176471f, 0.12202952941176472f,
-0.3466359803921569f, 0.7315649803921569f, 0.125720431372549f,
-0.35555566666666666f, 0.7346406666666667f, 0.12941133333333335f,
-0.3644753529411765f, 0.7377163529411764f, 0.13310223529411766f,
-0.3733950392156863f, 0.7407920392156863f, 0.13679313725490197f,
-0.3823147254901961f, 0.7438677254901961f, 0.14048403921568628f,
-0.3912344117647059f, 0.7469434117647059f, 0.1441749411764706f,
-0.4001540980392157f, 0.7500190980392157f, 0.14786584313725493f,
-0.4090737843137255f, 0.7530947843137255f, 0.15155674509803924f,
-0.4179934705882353f, 0.7561704705882353f, 0.15524764705882355f,
-0.4269131568627451f, 0.7592461568627451f, 0.15893854901960786f,
-0.43629417647058827f, 0.7623218823529412f, 0.16385976470588237f,
-0.4461365294117644f, 0.7653976470588235f, 0.1700112941176469f,
-0.4559788823529412f, 0.7684734117647058f, 0.17616282352941176f,
-0.4658212352941177f, 0.7715491764705883f, 0.18231435294117648f,
-0.47566358823529414f, 0.7746249411764706f, 0.18846588235294118f,
-0.4855059411764706f, 0.777700705882353f, 0.1946174117647059f,
-0.495348294117647f, 0.7807764705882353f, 0.2007689411764706f,
-0.5051906470588235f, 0.7838522352941176f, 0.20692047058823532f,
-0.515033f, 0.7869280000000001f, 0.213072f,
-0.5248753529411765f, 0.7900037647058824f, 0.2192235294117647f,
-0.5347177058823529f, 0.7930795294117647f, 0.22537505882352943f,
-0.5445600588235294f, 0.7961552941176471f, 0.23152658823529415f,
-0.5544024117647058f, 0.7992310588235294f, 0.23767811764705882f,
-0.5633220588235294f, 0.8023068235294118f, 0.24244552941176473f,
-0.5719341372549019f, 0.8053825882352942f, 0.24675156862745098f,
-0.5805462156862745f, 0.8084583529411765f, 0.25105760784313724f,
-0.5891582941176471f, 0.8115341176470589f, 0.25536364705882353f,
-0.5977703725490193f, 0.8146098823529411f, 0.2596696862745097f,
-0.6063824509803921f, 0.8176856470588235f, 0.2639757254901961f,
-0.6149945294117647f, 0.8207614117647059f, 0.26828176470588233f,
-0.6236066078431373f, 0.8238371764705882f, 0.2725878039215686f,
-0.6322186862745098f, 0.8269129411764706f, 0.2768938431372549f,
-0.6408307647058824f, 0.829988705882353f, 0.2811998823529412f,
-0.6494428431372549f, 0.8330644705882353f, 0.2855059215686274f,
-0.6580549215686274f, 0.8361402352941176f, 0.2898119607843137f,
-0.666667f, 0.839216f, 0.294118f,
-0.6758942156862745f, 0.8422916862745098f, 0.302114862745098f,
-0.685121431372549f, 0.8453673725490196f, 0.31011172549019606f,
-0.6943486470588235f, 0.8484430588235293f, 0.3181085882352941f,
-0.703575862745098f, 0.8515187450980392f, 0.3261054509803921f,
-0.7128030784313726f, 0.854594431372549f, 0.3341023137254902f,
-0.7220302941176471f, 0.8576701176470588f, 0.34209917647058824f,
-0.7312575098039216f, 0.8607458039215686f, 0.35009603921568627f,
-0.7404847254901958f, 0.8638214901960783f, 0.3580929019607841f,
-0.7497119411764706f, 0.8668971764705882f, 0.36608976470588234f,
-0.758939156862745f, 0.869972862745098f, 0.37408662745098037f,
-0.7681663725490195f, 0.8730485490196078f, 0.3820834901960784f,
-0.777393588235294f, 0.8761242352941176f, 0.3900803529411765f,
-0.7863132352941176f, 0.8792768431372548f, 0.39876927450980393f,
-0.7943101764705882f, 0.8826602156862745f, 0.4095343725490196f,
-0.8023071176470588f, 0.886043588235294f, 0.42029947058823525f,
-0.8103040588235294f, 0.8894269607843137f, 0.43106456862745096f,
-0.818301f, 0.8928103333333333f, 0.4418296666666666f,
-0.8262979411764706f, 0.8961937058823529f, 0.45259476470588234f,
-0.8342948823529411f, 0.8995770784313725f, 0.463359862745098f,
-0.8422918235294118f, 0.9029604509803921f, 0.47412496078431365f,
-0.8502887647058823f, 0.9063438235294118f, 0.48489005882352937f,
-0.858285705882353f, 0.9097271960784313f, 0.4956551568627451f,
-0.8662826470588235f, 0.913110568627451f, 0.5064202549019607f,
-0.874279588235294f, 0.9164939411764704f, 0.5171853529411761f,
-0.8822765294117647f, 0.9198773137254902f, 0.5279504509803921f,
-0.8891969215686275f, 0.9226455098039216f, 0.5387155490196078f,
-0.8950407647058823f, 0.9247985294117647f, 0.5494806470588235f,
-0.9008846078431373f, 0.9269515490196079f, 0.5602457450980391f,
-0.9067284509803921f, 0.9291045686274509f, 0.5710108431372548f,
-0.9125722941176471f, 0.9312575882352941f, 0.5817759411764706f,
-0.918416137254902f, 0.9334106078431372f, 0.5925410392156862f,
-0.9242599803921568f, 0.9355636274509803f, 0.6033061372549019f,
-0.9301038235294118f, 0.9377166470588235f, 0.6140712352941176f,
-0.9359476666666666f, 0.9398696666666666f, 0.6248363333333333f,
-0.9417915098039216f, 0.9420226862745098f, 0.635601431372549f,
-0.9476353529411764f, 0.9441757058823529f, 0.6463665294117646f,
-0.9534791960784313f, 0.9463287254901961f, 0.6571316274509804f,
-0.9593230392156863f, 0.9484817450980392f, 0.6678967254901961f,
-0.9630908235294118f, 0.9510961176470588f, 0.6841981764705882f,
-0.966166588235294f, 0.9538642745098038f, 0.702345078431372f,
-0.9692423529411764f, 0.956632431372549f, 0.7204919803921568f,
-0.9723181176470588f, 0.9594005882352941f, 0.7386388823529412f,
-0.9753938823529411f, 0.9621687450980392f, 0.7567857843137255f,
-0.9784696470588236f, 0.9649369019607843f, 0.7749326862745098f,
-0.9815454117647059f, 0.9677050588235294f, 0.7930795882352941f,
-0.9846211764705882f, 0.9704732156862745f, 0.8112264901960784f,
-0.9876969411764706f, 0.9732413725490197f, 0.8293733921568627f,
-0.9907727058823529f, 0.9760095294117647f, 0.8475202941176471f,
-0.9938484705882353f, 0.9787776862745098f, 0.8656671960784313f,
-0.9969242352941177f, 0.9815458431372549f, 0.8838140980392157f,
-1.0f, 0.984314f, 0.901961f
-}
-, 
-{
-0.909804f, 0.980392f, 0.831373f,
-0.9051903529411764f, 0.9779314509803922f, 0.8242987254901961f,
-0.9005767058823528f, 0.9754709019607843f, 0.8172244509803922f,
-0.8959630588235293f, 0.9730103529411765f, 0.8101501764705883f,
-0.8913494117647058f, 0.9705498039215686f, 0.8030759019607843f,
-0.8867357647058823f, 0.9680892549019608f, 0.7960016274509804f,
-0.8821221176470587f, 0.965628705882353f, 0.7889273529411764f,
-0.8775084705882352f, 0.9631681568627452f, 0.7818530784313725f,
-0.8728948235294117f, 0.9607076078431372f, 0.7747788039215686f,
-0.8682811764705882f, 0.9582470588235295f, 0.7677045294117647f,
-0.8636675294117646f, 0.9557865098039215f, 0.7606302549019608f,
-0.8590538823529411f, 0.9533259607843138f, 0.7535559803921568f,
-0.8544402352941176f, 0.9508654117647058f, 0.7464817058823529f,
-0.8495959215686274f, 0.9480972745098039f, 0.739023f,
-0.8440596078431372f, 0.9444063725490196f, 0.7304109999999999f,
-0.838523294117647f, 0.9407154705882352f, 0.721799f,
-0.8329869803921568f, 0.937024568627451f, 0.7131869999999999f,
-0.8274506666666667f, 0.9333336666666666f, 0.704575f,
-0.8219143529411764f, 0.9296427647058824f, 0.695963f,
-0.8163780392156862f, 0.925951862745098f, 0.6873509999999999f,
-0.8108417254901961f, 0.9222609607843137f, 0.678739f,
-0.8053054117647058f, 0.9185700588235294f, 0.6701269999999999f,
-0.7997690980392157f, 0.9148791568627451f, 0.661515f,
-0.7942327843137255f, 0.9111882549019608f, 0.652903f,
-0.7886964705882352f, 0.9074973529411765f, 0.644291f,
-0.7831601568627451f, 0.9038064509803921f, 0.635679f,
-0.7783927647058824f, 0.9007307058823529f, 0.6282203529411764f,
-0.774394294117647f, 0.8982701176470589f, 0.6219150588235294f,
-0.7703958235294117f, 0.8958095294117647f, 0.6156097647058824f,
-0.7663973529411765f, 0.8933489411764706f, 0.6093044705882352f,
-0.7623988823529412f, 0.8908883529411765f, 0.6029991764705882f,
-0.7584004117647059f, 0.8884277647058824f, 0.5966938823529412f,
-0.7544019411764705f, 0.8859671764705882f, 0.590388588235294f,
-0.7504034705882353f, 0.8835065882352942f, 0.584083294117647f,
-0.746405f, 0.881046f, 0.577778f,
-0.7424065294117647f, 0.8785854117647058f, 0.5714727058823529f,
-0.7384080588235294f, 0.8761248235294118f, 0.5651674117647059f,
-0.7344095882352941f, 0.8736642352941176f, 0.5588621176470588f,
-0.7304111176470588f, 0.8712036470588235f, 0.5525568235294117f,
-0.7264126470588235f, 0.8687430588235294f, 0.5462515294117647f,
-0.7224141764705883f, 0.8662824705882353f, 0.5399462352941177f,
-0.7184157058823529f, 0.8638218823529412f, 0.5336409411764707f,
-0.7144172352941176f, 0.8613612941176471f, 0.5273356470588235f,
-0.7104187647058824f, 0.8589007058823529f, 0.5210303529411765f,
-0.7064202941176471f, 0.8564401176470589f, 0.5147250588235294f,
-0.7024218235294117f, 0.8539795294117647f, 0.5084197647058823f,
-0.6984233529411765f, 0.8515189411764705f, 0.5021144705882352f,
-0.6944248823529412f, 0.8490583529411765f, 0.4958091764705882f,
-0.6904264117647059f, 0.8465977647058823f, 0.4895038823529412f,
-0.6864279411764707f, 0.8441371764705882f, 0.48319858823529416f,
-0.6824294705882353f, 0.8416765882352941f, 0.476893294117647f,
-0.678431f, 0.839216f, 0.470588f,
-0.6755090784313725f, 0.8364478039215686f, 0.46658952941176474f,
-0.6725871568627452f, 0.8336796078431372f, 0.46259105882352947f,
-0.6696652352941177f, 0.8309114117647058f, 0.45859258823529414f,
-0.6667433137254902f, 0.8281432156862745f, 0.4545941176470588f,
-0.6638213921568628f, 0.8253750196078431f, 0.45059564705882355f,
-0.6608994705882353f, 0.8226068235294117f, 0.4465971764705883f,
-0.6579775490196078f, 0.8198386274509804f, 0.44259870588235295f,
-0.6550556274509803f, 0.817070431372549f, 0.4386002352941176f,
-0.652133705882353f, 0.8143022352941176f, 0.43460176470588235f,
-0.6492117843137255f, 0.8115340392156862f, 0.4306032941176471f,
-0.646289862745098f, 0.8087658431372549f, 0.42660482352941176f,
-0.6433679411764706f, 0.8059976470588235f, 0.4226063529411765f,
-0.6404460196078431f, 0.8032294509803921f, 0.41860788235294116f,
-0.6375240980392156f, 0.8004612549019607f, 0.4146094117647059f,
-0.6346021764705883f, 0.7976930588235294f, 0.4106109411764706f,
-0.6316802549019608f, 0.794924862745098f, 0.4066124705882353f,
-0.6287583333333333f, 0.7921566666666666f, 0.402614f,
-0.6258364117647058f, 0.7893884705882352f, 0.3986155294117647f,
-0.6229144901960784f, 0.7866202745098039f, 0.3946170588235294f,
-0.619992568627451f, 0.7838520784313725f, 0.3906185882352941f,
-0.6170706470588235f, 0.7810838823529411f, 0.38662011764705884f,
-0.6141487254901961f, 0.7783156862745098f, 0.3826216470588235f,
-0.6112268039215686f, 0.7755474901960784f, 0.3786231764705883f,
-0.6083048823529411f, 0.772779294117647f, 0.3746247058823529f,
-0.6053829607843138f, 0.7700110980392156f, 0.37062623529411765f,
-0.6028454901960784f, 0.767704294117647f, 0.36701225490196077f,
-0.6006924705882353f, 0.7658588823529411f, 0.36378276470588233f,
-0.5985394509803921f, 0.7640134705882352f, 0.3605532745098039f,
-0.596386431372549f, 0.7621680588235293f, 0.35732378431372547f,
-0.5942334117647058f, 0.7603226470588235f, 0.35409429411764703f,
-0.5920803921568627f, 0.7584772352941176f, 0.35086480392156866f,
-0.5899273725490196f, 0.7566318235294117f, 0.34763531372549017f,
-0.5877743529411764f, 0.7547864117647058f, 0.34440582352941174f,
-0.5856213333333333f, 0.752941f, 0.3411763333333333f,
-0.5834683137254901f, 0.7510955882352941f, 0.33794684313725487f,
-0.581315294117647f, 0.7492501764705882f, 0.33471735294117644f,
-0.5791622745098038f, 0.7474047647058824f, 0.331487862745098f,
-0.5770092549019608f, 0.7455593529411765f, 0.3282583725490196f,
-0.5748562352941177f, 0.7437139411764706f, 0.3250288823529412f,
-0.5727032156862745f, 0.7418685294117647f, 0.32179939215686276f,
-0.5705501960784314f, 0.7400231176470589f, 0.31856990196078433f,
-0.5683971764705882f, 0.7381777058823529f, 0.3153404117647059f,
-0.5662441568627451f, 0.736332294117647f, 0.31211092156862746f,
-0.5640911372549019f, 0.7344868823529411f, 0.30888143137254903f,
-0.5619381176470588f, 0.7326414705882353f, 0.3056519411764706f,
-0.5597850980392156f, 0.7307960588235294f, 0.30242245098039217f,
-0.5576320784313725f, 0.7289506470588236f, 0.2991929607843138f,
-0.5554790588235293f, 0.7271052352941176f, 0.2959634705882353f,
-0.5533260392156862f, 0.7252598235294118f, 0.29273398039215687f,
-0.551173019607843f, 0.7234144117647059f, 0.28950449019607843f,
-0.54902f, 0.721569f, 0.286275f,
-0.5473283137254902f, 0.7195697450980393f, 0.2845833137254902f,
-0.5456366274509804f, 0.7175704901960784f, 0.2828916274509804f,
-0.5439449411764705f, 0.7155712352941177f, 0.28119994117647057f,
-0.5422532549019607f, 0.7135719803921569f, 0.2795082549019608f,
-0.540561568627451f, 0.7115727254901961f, 0.27781656862745097f,
-0.5388698823529412f, 0.7095734705882353f, 0.2761248823529412f,
-0.5371781960784313f, 0.7075742156862745f, 0.27443319607843136f,
-0.5354865098039215f, 0.7055749607843137f, 0.2727415098039216f,
-0.5337948235294118f, 0.703575705882353f, 0.27104982352941176f,
-0.532103137254902f, 0.7015764509803921f, 0.269358137254902f,
-0.5304114509803921f, 0.6995771960784314f, 0.26766645098039216f,
-0.5287197647058823f, 0.6975779411764707f, 0.2659747647058824f,
-0.5270280784313726f, 0.6955786862745098f, 0.26428307843137255f,
-0.5253363921568628f, 0.693579431372549f, 0.2625913921568627f,
-0.5236447058823529f, 0.6915801764705882f, 0.26089970588235295f,
-0.5219530196078431f, 0.6895809215686275f, 0.2592080196078431f,
-0.5202613333333334f, 0.6875816666666666f, 0.25751633333333335f,
-0.5185696470588236f, 0.6855824117647058f, 0.2558246470588235f,
-0.5168779607843137f, 0.6835831568627451f, 0.2541329607843137f,
-0.5151862745098039f, 0.6815839019607843f, 0.2524412745098039f,
-0.5134945882352941f, 0.6795846470588235f, 0.25074958823529414f,
-0.5118029019607844f, 0.6775853921568628f, 0.2490579019607843f,
-0.5101112156862746f, 0.6755861372549019f, 0.2473662156862745f,
-0.5084195294117647f, 0.6735868823529412f, 0.2456745294117647f,
-0.5067278431372549f, 0.6715876274509803f, 0.24398284313725488f,
-0.5048054901960785f, 0.6692039215686274f, 0.24213739215686272f,
-0.5026524705882354f, 0.6664357647058823f, 0.24013817647058822f,
-0.5004994509803922f, 0.6636676078431372f, 0.2381389607843137f,
-0.4983464313725491f, 0.6608994509803922f, 0.2361397450980392f,
-0.496193411764706f, 0.6581312941176471f, 0.23414052941176475f,
-0.4940403921568628f, 0.6553631372549019f, 0.23214131372549018f,
-0.49188737254901965f, 0.6525949803921568f, 0.23014209803921568f,
-0.4897343529411765f, 0.6498268235294117f, 0.22814288235294117f,
-0.48758133333333337f, 0.6470586666666667f, 0.22614366666666666f,
-0.4854283137254902f, 0.6442905098039216f, 0.22414445098039215f,
-0.4832752941176471f, 0.6415223529411764f, 0.22214523529411764f,
-0.48112227450980394f, 0.6387541960784313f, 0.22014601960784314f,
-0.4789692549019608f, 0.6359860392156862f, 0.21814680392156863f,
-0.47681623529411765f, 0.6332178823529412f, 0.21614758823529412f,
-0.4746632156862745f, 0.6304497254901961f, 0.2141483725490196f,
-0.47251019607843137f, 0.627681568627451f, 0.2121491568627451f,
-0.4703571764705883f, 0.6249134117647058f, 0.21014994117647057f,
-0.4682041568627451f, 0.6221452549019607f, 0.2081507254901961f,
-0.466051137254902f, 0.6193770980392157f, 0.20615150980392155f,
-0.46389811764705885f, 0.6166089411764706f, 0.20415229411764707f,
-0.46174509803921576f, 0.6138407843137256f, 0.2021530784313726f,
-0.45959207843137256f, 0.6110726274509803f, 0.20015386274509803f,
-0.4574390588235294f, 0.6083044705882352f, 0.19815464705882352f,
-0.4552860392156863f, 0.6055363137254902f, 0.196155431372549f,
-0.45313301960784313f, 0.6027681568627451f, 0.1941562156862745f,
-0.45098f, 0.6f, 0.192157f,
-0.44882698039215685f, 0.5972318431372549f, 0.19046535294117647f,
-0.4466739607843137f, 0.5944636862745097f, 0.18877370588235293f,
-0.44452094117647056f, 0.5916955294117647f, 0.1870820588235294f,
-0.4423679215686274f, 0.5889273725490196f, 0.1853904117647059f,
-0.4402149019607843f, 0.5861592156862745f, 0.18369876470588234f,
-0.4380618823529412f, 0.5833910588235294f, 0.18200711764705882f,
-0.43590886274509805f, 0.5806229019607843f, 0.1803154705882353f,
-0.4337558431372549f, 0.5778547450980392f, 0.17862382352941175f,
-0.43160282352941176f, 0.5750865882352941f, 0.17693217647058823f,
-0.4294498039215686f, 0.572318431372549f, 0.1752405294117647f,
-0.42729678431372553f, 0.569550274509804f, 0.17354888235294122f,
-0.42514376470588233f, 0.5667821176470588f, 0.17185723529411764f,
-0.4229907450980392f, 0.5640139607843138f, 0.17016558823529412f,
-0.42083772549019605f, 0.5612458039215686f, 0.16847394117647058f,
-0.41868470588235296f, 0.5584776470588235f, 0.16678229411764706f,
-0.4165316862745098f, 0.5557094901960784f, 0.16509064705882354f,
-0.4143786666666667f, 0.5529413333333333f, 0.16339900000000002f,
-0.41222564705882353f, 0.5501731764705882f, 0.16170735294117647f,
-0.4100726274509804f, 0.5474050196078432f, 0.16001570588235295f,
-0.40791960784313724f, 0.544636862745098f, 0.15832405882352943f,
-0.4057665882352941f, 0.5418687058823529f, 0.15663241176470588f,
-0.40361356862745096f, 0.5391005490196078f, 0.15494076470588236f,
-0.4014605490196078f, 0.5363323921568628f, 0.15324911764705884f,
-0.3993075294117647f, 0.5335642352941177f, 0.15155747058823532f,
-0.39715450980392153f, 0.5307960784313726f, 0.14986582352941177f,
-0.39500149019607844f, 0.5278741176470588f, 0.14825105882352943f,
-0.39284847058823535f, 0.5247983529411766f, 0.1467131764705883f,
-0.39069545098039216f, 0.5217225882352942f, 0.14517529411764707f,
-0.388542431372549f, 0.5186468235294117f, 0.1436374117647059f,
-0.38638941176470587f, 0.5155710588235294f, 0.1420995294117647f,
-0.3842363921568627f, 0.512495294117647f, 0.14056164705882354f,
-0.3820833725490196f, 0.5094195294117647f, 0.13902376470588237f,
-0.37993035294117644f, 0.5063437647058824f, 0.13748588235294118f,
-0.3777773333333333f, 0.503268f, 0.135948f,
-0.37562431372549016f, 0.5001922352941176f, 0.13441011764705882f,
-0.37347129411764707f, 0.49711647058823527f, 0.13287223529411765f,
-0.3713182745098039f, 0.49404070588235294f, 0.13133435294117649f,
-0.3691652549019608f, 0.4909649411764706f, 0.1297964705882353f,
-0.36701223529411764f, 0.4878891764705882f, 0.12825858823529412f,
-0.3648592156862745f, 0.4848134117647059f, 0.12672070588235296f,
-0.36270619607843135f, 0.48173764705882355f, 0.12518282352941176f,
-0.3605531764705882f, 0.47866188235294116f, 0.1236449411764706f,
-0.3584001568627451f, 0.4755861176470589f, 0.12210705882352946f,
-0.3562471372549019f, 0.4725103529411765f, 0.12056917647058824f,
-0.3540941176470588f, 0.4694345882352941f, 0.11903129411764707f,
-0.3519410980392157f, 0.46635882352941177f, 0.11749341176470587f,
-0.34978807843137255f, 0.4632830588235294f, 0.11595552941176471f,
-0.3476350588235294f, 0.46020729411764705f, 0.11441764705882354f,
-0.34548203921568627f, 0.4571315294117647f, 0.11287976470588235f,
-0.3433290196078431f, 0.4540557647058823f, 0.11134188235294118f,
-0.341176f, 0.45098f, 0.109804f,
-0.3390230196078431f, 0.4477504705882353f, 0.10841992156862745f,
-0.33687003921568626f, 0.44452094117647056f, 0.10703584313725491f,
-0.33471705882352937f, 0.4412914117647059f, 0.10565176470588235f,
-0.33256407843137253f, 0.4380618823529412f, 0.1042676862745098f,
-0.33041109803921564f, 0.43483235294117645f, 0.10288360784313726f,
-0.3282581176470588f, 0.43160282352941176f, 0.10149952941176471f,
-0.3261051372549019f, 0.428373294117647f, 0.10011545098039215f,
-0.32395215686274514f, 0.42514376470588244f, 0.09873137254901965f,
-0.3217991764705882f, 0.42191423529411765f, 0.09734729411764706f,
-0.31964619607843137f, 0.41868470588235296f, 0.09596321568627451f,
-0.3174932156862745f, 0.4154551764705882f, 0.09457913725490197f,
-0.31534023529411764f, 0.41222564705882353f, 0.09319505882352941f,
-0.31318725490196075f, 0.4089961176470588f, 0.09181098039215686f,
-0.3110342745098039f, 0.4057665882352941f, 0.09042690196078432f,
-0.30888129411764703f, 0.4025370588235294f, 0.08904282352941177f,
-0.3067283137254902f, 0.3993075294117647f, 0.08765874509803923f,
-0.3045753333333333f, 0.396078f, 0.08627466666666667f,
-0.3024223529411765f, 0.3928484705882353f, 0.08489058823529412f,
-0.3002693725490196f, 0.38961894117647056f, 0.08350650980392157f,
-0.29811639215686275f, 0.38638941176470587f, 0.08212243137254903f,
-0.29596341176470586f, 0.3831598823529412f, 0.08073835294117648f,
-0.29381043137254903f, 0.37993035294117644f, 0.07935427450980392f,
-0.29165745098039214f, 0.37670082352941175f, 0.07797019607843138f,
-0.28950447058823536f, 0.3734712941176471f, 0.07658611764705886f,
-0.2873514901960784f, 0.3702417647058823f, 0.07520203921568627f,
-0.2849678039215686f, 0.36685845098039216f, 0.0738948431372549f,
-0.2823534117647059f, 0.36332135294117646f, 0.07266452941176471f,
-0.27973901960784314f, 0.35978425490196075f, 0.07143421568627452f,
-0.2771246274509804f, 0.3562471568627451f, 0.07020390196078433f,
-0.27451023529411767f, 0.3527100588235294f, 0.06897358823529412f,
-0.2718958431372549f, 0.3491729607843137f, 0.06774327450980393f,
-0.26928145098039213f, 0.345635862745098f, 0.06651296078431373f,
-0.2666670588235294f, 0.34209876470588235f, 0.06528264705882353f,
-0.26405266666666666f, 0.33856166666666665f, 0.06405233333333334f,
-0.26143827450980395f, 0.33502456862745095f, 0.06282201960784314f,
-0.2588238823529412f, 0.3314874705882353f, 0.061591705882352944f,
-0.2562094901960784f, 0.3279503725490196f, 0.06036139215686275f,
-0.2535950980392157f, 0.3244132745098039f, 0.05913107843137255f,
-0.25098070588235294f, 0.3208761764705882f, 0.05790076470588236f,
-0.24836631372549028f, 0.31733907843137266f, 0.056670450980392195f,
-0.24575192156862746f, 0.31380198039215684f, 0.05544013725490196f,
-0.2431375294117647f, 0.31026488235294114f, 0.05420982352941177f,
-0.24052313725490196f, 0.3067277843137255f, 0.052979509803921576f,
-0.23790874509803922f, 0.3031906862745098f, 0.05174919607843138f,
-0.23529435294117648f, 0.2996535882352941f, 0.05051888235294118f,
-0.23267996078431372f, 0.29611649019607844f, 0.049288568627450985f,
-0.23006556862745098f, 0.29257939215686274f, 0.048058254901960785f,
-0.22745117647058824f, 0.28904229411764704f, 0.04682794117647059f,
-0.2248367843137255f, 0.2855051960784314f, 0.045597627450980394f,
-0.22222239215686274f, 0.2819680980392157f, 0.0443673137254902f,
-0.219608f, 0.278431f, 0.043137f
-}
-, 
-{
-0.964706f, 0.988235f, 0.941176f,
-0.9619378431372548f, 0.9871584901960784f, 0.9373313333333333f,
-0.9591696862745097f, 0.9860819803921569f, 0.9334866666666667f,
-0.9564015294117647f, 0.9850054705882353f, 0.929642f,
-0.9536333725490196f, 0.9839289607843137f, 0.9257973333333334f,
-0.9508652156862745f, 0.9828524509803921f, 0.9219526666666666f,
-0.9480970588235293f, 0.9817759411764706f, 0.918108f,
-0.9453289019607842f, 0.980699431372549f, 0.9142633333333333f,
-0.9425607450980392f, 0.9796229215686274f, 0.9104186666666667f,
-0.9397925882352941f, 0.9785464117647058f, 0.906574f,
-0.937024431372549f, 0.9774699019607843f, 0.9027293333333334f,
-0.9342562745098039f, 0.9763933921568627f, 0.8988846666666667f,
-0.9314881176470587f, 0.9753168823529411f, 0.8950400000000001f,
-0.9287199607843137f, 0.9742403725490196f, 0.8911953333333333f,
-0.9259518039215686f, 0.973163862745098f, 0.8873506666666667f,
-0.9231836470588235f, 0.9720873529411764f, 0.883506f,
-0.9204154901960784f, 0.9710108431372548f, 0.8796613333333334f,
-0.9176473333333333f, 0.9699343333333333f, 0.8758166666666667f,
-0.9148791764705881f, 0.9688578235294117f, 0.871972f,
-0.9121110196078431f, 0.9677813137254901f, 0.8681273333333334f,
-0.909342862745098f, 0.9667048039215687f, 0.8642826666666666f,
-0.9065747058823529f, 0.9656282941176471f, 0.860438f,
-0.9038065490196078f, 0.9645517843137255f, 0.8565933333333333f,
-0.9010383921568628f, 0.9634752745098039f, 0.8527486666666667f,
-0.8982702352941176f, 0.9623987647058824f, 0.848904f,
-0.8955020784313725f, 0.9613222549019608f, 0.8450593333333334f,
-0.8921956470588235f, 0.9600150784313726f, 0.8413684509803921f,
-0.8883509411764705f, 0.9584772352941177f, 0.8378313529411765f,
-0.8845062352941176f, 0.9569393921568627f, 0.8342942549019609f,
-0.8806615294117647f, 0.9554015490196078f, 0.8307571568627451f,
-0.8768168235294117f, 0.9538637058823529f, 0.8272200588235294f,
-0.8729721176470588f, 0.952325862745098f, 0.8236829607843137f,
-0.8691274117647059f, 0.9507880196078431f, 0.820145862745098f,
-0.8652827058823529f, 0.9492501764705882f, 0.8166087647058824f,
-0.8614379999999999f, 0.9477123333333333f, 0.8130716666666666f,
-0.8575932941176471f, 0.9461744901960784f, 0.809534568627451f,
-0.8537485882352941f, 0.9446366470588236f, 0.8059974705882353f,
-0.8499038823529411f, 0.9430988039215686f, 0.8024603725490196f,
-0.8460591764705881f, 0.9415609607843137f, 0.7989232745098039f,
-0.8422144705882353f, 0.9400231176470588f, 0.7953861764705882f,
-0.8383697647058823f, 0.9384852745098039f, 0.7918490784313725f,
-0.8345250588235293f, 0.936947431372549f, 0.7883119803921569f,
-0.8306803529411764f, 0.9354095882352941f, 0.7847748823529411f,
-0.8268356470588235f, 0.9338717450980392f, 0.7812377843137255f,
-0.8229909411764705f, 0.9323339019607843f, 0.7777006862745098f,
-0.8191462352941177f, 0.9307960588235293f, 0.7741635882352942f,
-0.8153015294117647f, 0.9292582156862745f, 0.7706264901960784f,
-0.8114568235294117f, 0.9277203725490196f, 0.7670893921568627f,
-0.8076121176470588f, 0.9261825294117647f, 0.763552294117647f,
-0.8037674117647059f, 0.9246446862745098f, 0.7600151960784314f,
-0.7999227058823529f, 0.9231068431372549f, 0.7564780980392156f,
-0.796078f, 0.921569f, 0.752941f,
-0.7942325882352941f, 0.9203386862745098f, 0.751710725490196f,
-0.7923871764705882f, 0.9191083725490196f, 0.7504804509803922f,
-0.7905417647058823f, 0.9178780588235294f, 0.7492501764705882f,
-0.7886963529411765f, 0.9166477450980391f, 0.7480199019607843f,
-0.7868509411764706f, 0.9154174313725489f, 0.7467896274509803f,
-0.7850055294117647f, 0.9141871176470588f, 0.7455593529411765f,
-0.7831601176470588f, 0.9129568039215686f, 0.7443290784313725f,
-0.781314705882353f, 0.9117264901960784f, 0.7430988039215686f,
-0.779469294117647f, 0.9104961764705882f, 0.7418685294117647f,
-0.7776238823529412f, 0.909265862745098f, 0.7406382549019608f,
-0.7757784705882352f, 0.9080355490196078f, 0.7394079803921568f,
-0.7739330588235294f, 0.9068052352941176f, 0.738177705882353f,
-0.7720876470588235f, 0.9055749215686274f, 0.736947431372549f,
-0.7702422352941176f, 0.9043446078431372f, 0.7357171568627451f,
-0.7683968235294117f, 0.9031142941176471f, 0.7344868823529412f,
-0.7665514117647059f, 0.9018839803921569f, 0.7332566078431373f,
-0.764706f, 0.9006536666666667f, 0.7320263333333333f,
-0.7628605882352941f, 0.8994233529411765f, 0.7307960588235294f,
-0.7610151764705883f, 0.8981930392156863f, 0.7295657843137255f,
-0.7591697647058824f, 0.8969627254901961f, 0.7283355098039216f,
-0.7573243529411765f, 0.8957324117647059f, 0.7271052352941176f,
-0.7554789411764706f, 0.8945020980392157f, 0.7258749607843137f,
-0.7536335294117648f, 0.8932717843137254f, 0.7246446862745098f,
-0.7517881176470589f, 0.8920414705882352f, 0.7234144117647059f,
-0.749942705882353f, 0.890811156862745f, 0.7221841372549019f,
-0.7479434901960784f, 0.8894270588235293f, 0.7207231568627451f,
-0.7457904705882353f, 0.8878891764705882f, 0.7190314705882354f,
-0.7436374509803921f, 0.886351294117647f, 0.7173397843137255f,
-0.741484431372549f, 0.8848134117647058f, 0.7156480980392157f,
-0.7393314117647058f, 0.8832755294117647f, 0.7139564117647059f,
-0.7371783921568628f, 0.8817376470588235f, 0.7122647254901961f,
-0.7350253725490197f, 0.8801997647058823f, 0.7105730392156863f,
-0.7328723529411765f, 0.8786618823529412f, 0.7088813529411765f,
-0.7307193333333334f, 0.877124f, 0.7071896666666667f,
-0.7285663137254902f, 0.8755861176470588f, 0.7054979803921568f,
-0.7264132941176471f, 0.8740482352941176f, 0.7038062941176471f,
-0.7242602745098039f, 0.8725103529411764f, 0.7021146078431373f,
-0.7221072549019608f, 0.8709724705882352f, 0.7004229215686275f,
-0.7199542352941177f, 0.8694345882352941f, 0.6987312352941177f,
-0.7178012156862745f, 0.8678967058823529f, 0.6970395490196079f,
-0.7156481960784314f, 0.8663588235294117f, 0.6953478627450981f,
-0.7134951764705882f, 0.8648209411764706f, 0.6936561764705882f,
-0.7113421568627452f, 0.8632830588235294f, 0.6919644901960784f,
-0.709189137254902f, 0.8617451764705882f, 0.6902728039215686f,
-0.7070361176470589f, 0.8602072941176471f, 0.6885811176470589f,
-0.7048830980392157f, 0.8586694117647058f, 0.686889431372549f,
-0.7027300784313726f, 0.8571315294117647f, 0.6851977450980392f,
-0.7005770588235294f, 0.8555936470588235f, 0.6835060588235294f,
-0.6984240392156863f, 0.8540557647058823f, 0.6818143725490197f,
-0.6962710196078431f, 0.8525178823529411f, 0.6801226862745098f,
-0.694118f, 0.85098f, 0.678431f,
-0.6922725490196079f, 0.849749725490196f, 0.677200725490196f,
-0.6904270980392156f, 0.8485194509803922f, 0.6759704509803922f,
-0.6885816470588235f, 0.8472891764705882f, 0.6747401764705883f,
-0.6867361960784314f, 0.8460589019607843f, 0.6735099019607843f,
-0.6848907450980393f, 0.8448286274509803f, 0.6722796274509804f,
-0.683045294117647f, 0.8435983529411765f, 0.6710493529411765f,
-0.6811998431372549f, 0.8423680784313725f, 0.6698190784313726f,
-0.6793543921568628f, 0.8411378039215686f, 0.6685888039215686f,
-0.6775089411764706f, 0.8399075294117647f, 0.6673585294117648f,
-0.6756634901960784f, 0.8386772549019608f, 0.6661282549019608f,
-0.6738180392156863f, 0.8374469803921568f, 0.6648979803921569f,
-0.6719725882352942f, 0.836216705882353f, 0.663667705882353f,
-0.670127137254902f, 0.834986431372549f, 0.6624374313725491f,
-0.6682816862745098f, 0.8337561568627451f, 0.6612071568627451f,
-0.6664362352941177f, 0.8325258823529411f, 0.6599768823529412f,
-0.6645907843137255f, 0.8312956078431373f, 0.6587466078431373f,
-0.6627453333333334f, 0.8300653333333333f, 0.6575163333333334f,
-0.6608998823529412f, 0.8288350588235294f, 0.6562860588235294f,
-0.659054431372549f, 0.8276047843137255f, 0.6550557843137256f,
-0.6572089803921569f, 0.8263745098039216f, 0.6538255098039216f,
-0.6553635294117648f, 0.8251442352941176f, 0.6525952352941177f,
-0.6535180784313726f, 0.8239139607843137f, 0.6513649607843137f,
-0.6516726274509804f, 0.8226836862745098f, 0.6501346862745099f,
-0.6498271764705883f, 0.8214534117647059f, 0.6489044117647059f,
-0.6479817254901962f, 0.8202231372549019f, 0.647674137254902f,
-0.6462900588235294f, 0.8188390588235294f, 0.6465976274509805f,
-0.6447521764705882f, 0.8173011764705882f, 0.6456748823529412f,
-0.6432142941176471f, 0.815763294117647f, 0.644752137254902f,
-0.6416764117647059f, 0.8142254117647059f, 0.6438293921568627f,
-0.6401385294117647f, 0.8126875294117647f, 0.6429066470588236f,
-0.6386006470588236f, 0.8111496470588235f, 0.6419839019607844f,
-0.6370627647058824f, 0.8096117647058824f, 0.6410611568627451f,
-0.6355248823529412f, 0.8080738823529412f, 0.6401384117647059f,
-0.6339870000000001f, 0.806536f, 0.6392156666666667f,
-0.6324491176470589f, 0.8049981176470589f, 0.6382929215686275f,
-0.6309112352941176f, 0.8034602352941176f, 0.6373701764705882f,
-0.6293733529411765f, 0.8019223529411764f, 0.636447431372549f,
-0.6278354705882353f, 0.8003844705882353f, 0.6355246862745099f,
-0.6262975882352941f, 0.7988465882352941f, 0.6346019411764706f,
-0.624759705882353f, 0.7973087058823529f, 0.6336791960784314f,
-0.6232218235294118f, 0.7957708235294118f, 0.6327564509803921f,
-0.6216839411764706f, 0.7942329411764706f, 0.631833705882353f,
-0.6201460588235295f, 0.7926950588235294f, 0.6309109607843137f,
-0.6186081764705883f, 0.7911571764705883f, 0.6299882156862745f,
-0.6170702941176471f, 0.7896192941176471f, 0.6290654705882353f,
-0.615532411764706f, 0.7880814117647059f, 0.6281427254901961f,
-0.6139945294117647f, 0.7865435294117646f, 0.6272199803921569f,
-0.6124566470588235f, 0.7850056470588235f, 0.6262972352941176f,
-0.6109187647058824f, 0.7834677647058823f, 0.6253744901960785f,
-0.6093808823529412f, 0.7819298823529411f, 0.6244517450980392f,
-0.607843f, 0.780392f, 0.623529f,
-0.6059975490196079f, 0.7788541176470588f, 0.622298725490196f,
-0.6041520980392157f, 0.7773162352941176f, 0.6210684509803922f,
-0.6023066470588235f, 0.7757783529411765f, 0.6198381764705883f,
-0.6004611960784314f, 0.7742404705882353f, 0.6186079019607843f,
-0.5986157450980393f, 0.7727025882352941f, 0.6173776274509803f,
-0.596770294117647f, 0.7711647058823529f, 0.6161473529411765f,
-0.5949248431372549f, 0.7696268235294117f, 0.6149170784313726f,
-0.5930793921568628f, 0.7680889411764705f, 0.6136868039215686f,
-0.5912339411764705f, 0.7665510588235294f, 0.6124565294117648f,
-0.5893884901960784f, 0.7650131764705882f, 0.6112262549019608f,
-0.5875430392156863f, 0.763475294117647f, 0.609995980392157f,
-0.585697588235294f, 0.7619374117647059f, 0.608765705882353f,
-0.5838521372549019f, 0.7603995294117647f, 0.6075354313725491f,
-0.5820066862745098f, 0.7588616470588235f, 0.6063051568627451f,
-0.5801612352941177f, 0.7573237647058824f, 0.6050748823529412f,
-0.5783157843137254f, 0.7557858823529411f, 0.6038446078431373f,
-0.5764703333333333f, 0.7542479999999999f, 0.6026143333333334f,
-0.5746248823529412f, 0.7527101176470588f, 0.6013840588235294f,
-0.5727794313725489f, 0.7511722352941176f, 0.6001537843137256f,
-0.5709339803921568f, 0.7496343529411764f, 0.5989235098039216f,
-0.5690885294117647f, 0.7480964705882353f, 0.5976932352941177f,
-0.5672430784313724f, 0.7465585882352941f, 0.5964629607843137f,
-0.5653976274509803f, 0.7450207058823529f, 0.5952326862745099f,
-0.5635521764705882f, 0.7434828235294118f, 0.5940024117647059f,
-0.5617067254901961f, 0.7419449411764706f, 0.592772137254902f,
-0.5599381764705882f, 0.7404070784313725f, 0.5916187450980392f,
-0.5582465294117647f, 0.7388692352941176f, 0.5905422352941178f,
-0.5565548823529411f, 0.7373313921568627f, 0.5894657254901962f,
-0.5548632352941176f, 0.7357935490196078f, 0.5883892156862746f,
-0.5531715882352941f, 0.7342557058823529f, 0.587312705882353f,
-0.5514799411764706f, 0.732717862745098f, 0.5862361960784315f,
-0.5497882941176471f, 0.7311800196078431f, 0.5851596862745099f,
-0.5480966470588234f, 0.7296421764705882f, 0.5840831764705883f,
-0.5464049999999999f, 0.7281043333333332f, 0.5830066666666667f,
-0.5447133529411764f, 0.7265664901960783f, 0.5819301568627452f,
-0.5430217058823529f, 0.7250286470588235f, 0.5808536470588236f,
-0.5413300588235294f, 0.7234908039215686f, 0.579777137254902f,
-0.5396384117647058f, 0.7219529607843137f, 0.5787006274509804f,
-0.5379467647058823f, 0.7204151176470588f, 0.5776241176470589f,
-0.5362551176470588f, 0.7188772745098039f, 0.5765476078431373f,
-0.5345634705882353f, 0.717339431372549f, 0.5754710980392157f,
-0.5328718235294118f, 0.715801588235294f, 0.5743945882352942f,
-0.5311801764705882f, 0.7142637450980392f, 0.5733180784313726f,
-0.5294885294117647f, 0.7127259019607842f, 0.572241568627451f,
-0.5277968823529411f, 0.7111880588235293f, 0.5711650588235294f,
-0.5261052352941176f, 0.7096502156862745f, 0.5700885490196079f,
-0.524413588235294f, 0.7081123725490196f, 0.5690120392156863f,
-0.5227219411764705f, 0.7065745294117647f, 0.5679355294117647f,
-0.521030294117647f, 0.7050366862745098f, 0.5668590196078431f,
-0.5193386470588235f, 0.7034988431372549f, 0.5657825098039216f,
-0.517647f, 0.701961f, 0.564706f,
-0.5156477843137255f, 0.7002693529411764f, 0.5634756862745098f,
-0.5136485686274509f, 0.6985777058823529f, 0.5622453725490196f,
-0.5116493529411764f, 0.6968860588235294f, 0.5610150588235294f,
-0.5096501372549019f, 0.6951944117647059f, 0.5597847450980392f,
-0.5076509215686275f, 0.6935027647058823f, 0.558554431372549f,
-0.5056517058823529f, 0.6918111176470588f, 0.5573241176470588f,
-0.5036524901960784f, 0.6901194705882352f, 0.5560938039215686f,
-0.5016532745098039f, 0.6884278235294118f, 0.5548634901960785f,
-0.49965405882352937f, 0.6867361764705882f, 0.5536331764705883f,
-0.4976548431372549f, 0.6850445294117646f, 0.5524028627450981f,
-0.49565562745098035f, 0.6833528823529411f, 0.5511725490196079f,
-0.49365641176470587f, 0.6816612352941176f, 0.5499422352941177f,
-0.49165719607843134f, 0.6799695882352941f, 0.5487119215686275f,
-0.48965798039215686f, 0.6782779411764706f, 0.5474816078431373f,
-0.4876587647058823f, 0.676586294117647f, 0.5462512941176471f,
-0.48565954901960784f, 0.6748946470588235f, 0.5450209803921569f,
-0.4836603333333333f, 0.673203f, 0.5437906666666666f,
-0.4816611176470588f, 0.6715113529411765f, 0.5425603529411764f,
-0.4796619019607843f, 0.669819705882353f, 0.5413300392156862f,
-0.4776626862745098f, 0.6681280588235293f, 0.540099725490196f,
-0.47566347058823527f, 0.6664364117647058f, 0.5388694117647058f,
-0.4736642549019608f, 0.6647447647058823f, 0.5376390980392156f,
-0.47166503921568625f, 0.6630531176470588f, 0.5364087843137254f,
-0.46966582352941183f, 0.6613614705882354f, 0.5351784705882353f,
-0.46766660784313724f, 0.6596698235294117f, 0.533948156862745f,
-0.4655135882352941f, 0.6574399019607843f, 0.5323333921568627f,
-0.46320676470588235f, 0.6546717058823529f, 0.5303341764705882f,
-0.4608999411764706f, 0.6519035098039215f, 0.5283349607843136f,
-0.45859311764705885f, 0.6491353137254902f, 0.5263357450980392f,
-0.45628629411764704f, 0.6463671176470588f, 0.5243365294117647f,
-0.4539794705882353f, 0.6435989215686274f, 0.5223373137254902f,
-0.45167264705882354f, 0.640830725490196f, 0.5203380980392156f,
-0.4493658235294118f, 0.6380625294117647f, 0.5183388823529411f,
-0.447059f, 0.6352943333333333f, 0.5163396666666666f,
-0.44475217647058823f, 0.6325261372549019f, 0.5143404509803922f,
-0.4424453529411765f, 0.6297579411764705f, 0.5123412352941176f,
-0.44013852941176473f, 0.6269897450980392f, 0.5103420196078431f,
-0.43783170588235293f, 0.6242215490196078f, 0.5083428039215686f,
-0.4355248823529412f, 0.6214533529411764f, 0.5063435882352941f,
-0.4332180588235295f, 0.6186851568627452f, 0.5043443725490196f,
-0.4309112352941177f, 0.6159169607843137f, 0.5023451568627451f,
-0.4286044117647059f, 0.6131487647058823f, 0.5003459411764706f,
-0.4262975882352941f, 0.6103805686274509f, 0.49834672549019604f,
-0.4239907647058824f, 0.6076123725490196f, 0.49634750980392156f,
-0.4216839411764706f, 0.6048441764705882f, 0.494348294117647f,
-0.4193771176470588f, 0.6020759803921568f, 0.49234907843137254f,
-0.41707029411764707f, 0.5993077843137254f, 0.490349862745098f,
-0.4147634705882353f, 0.5965395882352941f, 0.4883506470588235f,
-0.41245664705882357f, 0.5937713921568627f, 0.486351431372549f,
-0.41014982352941176f, 0.5910031960784313f, 0.4843522156862745f,
-0.407843f, 0.588235f, 0.482353f
-}
-, 
-{
-0.976471f, 1.0f, 0.94902f,
-0.9740103725490196f, 0.9984621176470588f, 0.9459442352941176f,
-0.9715497450980392f, 0.9969242352941177f, 0.9428684705882353f,
-0.9690891176470589f, 0.9953863529411765f, 0.9397927058823529f,
-0.9666284901960784f, 0.9938484705882353f, 0.9367169411764705f,
-0.964167862745098f, 0.9923105882352942f, 0.9336411764705882f,
-0.9617072352941176f, 0.990772705882353f, 0.9305654117647059f,
-0.9592466078431372f, 0.9892348235294118f, 0.9274896470588235f,
-0.9567859803921569f, 0.9876969411764707f, 0.9244138823529411f,
-0.9543253529411765f, 0.9861590588235294f, 0.9213381176470588f,
-0.9518647254901961f, 0.9846211764705882f, 0.9182623529411764f,
-0.9494040980392157f, 0.983083294117647f, 0.9151865882352941f,
-0.9469434705882352f, 0.9815454117647059f, 0.9121108235294118f,
-0.9443290588235294f, 0.9797768627450981f, 0.9091119607843137f,
-0.941253294117647f, 0.9773163137254902f, 0.9063438039215685f,
-0.9381775294117647f, 0.9748557647058824f, 0.9035756470588235f,
-0.9351017647058824f, 0.9723952156862745f, 0.9008074901960784f,
-0.932026f, 0.9699346666666667f, 0.8980393333333333f,
-0.9289502352941176f, 0.9674741176470588f, 0.8952711764705882f,
-0.9258744705882352f, 0.965013568627451f, 0.8925030196078431f,
-0.9227987058823529f, 0.9625530196078431f, 0.889734862745098f,
-0.9197229411764706f, 0.9600924705882353f, 0.8869667058823529f,
-0.9166471764705882f, 0.9576319215686274f, 0.8841985490196078f,
-0.9135714117647058f, 0.9551713725490196f, 0.8814303921568627f,
-0.9104956470588235f, 0.9527108235294117f, 0.8786622352941177f,
-0.9074198823529411f, 0.9502502745098039f, 0.8758940784313726f,
-0.904344156862745f, 0.9479434901960784f, 0.8729721176470588f,
-0.9012684705882352f, 0.9457904705882353f, 0.8698963529411765f,
-0.8981927843137254f, 0.9436374509803921f, 0.8668205882352942f,
-0.8951170980392157f, 0.941484431372549f, 0.8637448235294117f,
-0.8920414117647059f, 0.9393314117647058f, 0.8606690588235294f,
-0.888965725490196f, 0.9371783921568627f, 0.8575932941176471f,
-0.8858900392156862f, 0.9350253725490196f, 0.8545175294117647f,
-0.8828143529411765f, 0.9328723529411764f, 0.8514417647058824f,
-0.8797386666666667f, 0.9307193333333333f, 0.848366f,
-0.8766629803921568f, 0.9285663137254901f, 0.8452902352941176f,
-0.873587294117647f, 0.9264132941176471f, 0.8422144705882353f,
-0.8705116078431372f, 0.9242602745098039f, 0.839138705882353f,
-0.8674359215686274f, 0.9221072549019608f, 0.8360629411764706f,
-0.8641294705882353f, 0.9197235294117647f, 0.8332178823529411f,
-0.8607460980392156f, 0.9172629019607843f, 0.830449725490196f,
-0.8573627254901961f, 0.9148022745098039f, 0.827681568627451f,
-0.8539793529411764f, 0.9123416470588235f, 0.8249134117647059f,
-0.8505959803921568f, 0.9098810196078431f, 0.8221452549019608f,
-0.8472126078431372f, 0.9074203921568628f, 0.8193770980392157f,
-0.8438292352941177f, 0.9049597647058824f, 0.8166089411764706f,
-0.840445862745098f, 0.9024991372549019f, 0.8138407843137255f,
-0.8370624901960784f, 0.9000385098039215f, 0.8110726274509804f,
-0.8336791176470588f, 0.8975778823529411f, 0.8083044705882353f,
-0.8302957450980392f, 0.8951172549019608f, 0.8055363137254903f,
-0.8269123725490196f, 0.8926566274509804f, 0.8027681568627452f,
-0.823529f, 0.890196f, 0.8f,
-0.819530568627451f, 0.8871202352941177f, 0.796616705882353f,
-0.8155321372549019f, 0.8840444705882353f, 0.793233411764706f,
-0.811533705882353f, 0.8809687058823529f, 0.7898501176470588f,
-0.8075352745098039f, 0.8778929411764705f, 0.7864668235294118f,
-0.8035368431372549f, 0.8748171764705882f, 0.7830835294117647f,
-0.7995384117647059f, 0.8717414117647059f, 0.7797002352941177f,
-0.7955399803921569f, 0.8686656470588235f, 0.7763169411764705f,
-0.7915415490196078f, 0.8655898823529411f, 0.7729336470588235f,
-0.7875431176470589f, 0.8625141176470588f, 0.7695503529411765f,
-0.7835446862745099f, 0.8594383529411764f, 0.7661670588235294f,
-0.7795462549019608f, 0.8563625882352941f, 0.7627837647058823f,
-0.7755478235294118f, 0.8532868235294118f, 0.7594004705882352f,
-0.7718569607843138f, 0.850364862745098f, 0.7563247450980392f,
-0.7690888039215686f, 0.8479043137254901f, 0.7541717254901961f,
-0.7663206470588236f, 0.8454437647058823f, 0.752018705882353f,
-0.7635524901960785f, 0.8429832156862744f, 0.7498656862745098f,
-0.7607843333333334f, 0.8405226666666666f, 0.7477126666666666f,
-0.7580161764705883f, 0.8380621176470588f, 0.7455596470588235f,
-0.7552480196078432f, 0.8356015686274509f, 0.7434066274509803f,
-0.752479862745098f, 0.8331410196078431f, 0.7412536078431372f,
-0.7497117058823529f, 0.8306804705882352f, 0.7391005882352941f,
-0.7469435490196078f, 0.8282199215686274f, 0.7369475686274509f,
-0.7441753921568628f, 0.8257593725490197f, 0.7347945490196078f,
-0.7414072352941177f, 0.8232988235294117f, 0.7326415294117646f,
-0.7386390784313726f, 0.820838274509804f, 0.7304885098039215f,
-0.7352557843137255f, 0.8180701176470588f, 0.7277203529411764f,
-0.7312573529411764f, 0.8149943529411765f, 0.7243370588235294f,
-0.7272589215686275f, 0.8119185882352942f, 0.7209537647058823f,
-0.7232604901960784f, 0.8088428235294117f, 0.7175704705882352f,
-0.7192620588235294f, 0.8057670588235294f, 0.7141871764705882f,
-0.7152636274509804f, 0.8026912941176471f, 0.7108038823529412f,
-0.7112651960784313f, 0.7996155294117647f, 0.7074205882352941f,
-0.7072667647058823f, 0.7965397647058824f, 0.704037294117647f,
-0.7032683333333333f, 0.793464f, 0.700654f,
-0.6992699019607843f, 0.7903882352941176f, 0.6972707058823528f,
-0.6952714705882352f, 0.7873124705882353f, 0.6938874117647058f,
-0.6912730392156863f, 0.784236705882353f, 0.6905041176470588f,
-0.6872746078431372f, 0.7811609411764706f, 0.6871208235294117f,
-0.6841988235294117f, 0.7780851764705883f, 0.6846602352941177f,
-0.6814305882352941f, 0.7750094117647058f, 0.6825072156862745f,
-0.6786623529411765f, 0.7719336470588235f, 0.6803541960784314f,
-0.6758941176470588f, 0.7688578823529412f, 0.6782011764705882f,
-0.6731258823529411f, 0.7657821176470588f, 0.6760481568627451f,
-0.6703576470588235f, 0.7627063529411764f, 0.6738951372549019f,
-0.6675894117647059f, 0.7596305882352941f, 0.6717421176470588f,
-0.6648211764705882f, 0.7565548235294117f, 0.6695890980392156f,
-0.6620529411764706f, 0.7534790588235294f, 0.6674360784313725f,
-0.659284705882353f, 0.7504032941176471f, 0.6652830588235293f,
-0.6565164705882353f, 0.7473275294117646f, 0.6631300392156863f,
-0.6537482352941176f, 0.7442517647058823f, 0.660977019607843f,
-0.65098f, 0.741176f, 0.658824f,
-0.647596705882353f, 0.7381003137254901f, 0.6560557647058823f,
-0.6442134117647059f, 0.7350246274509803f, 0.6532875294117647f,
-0.6408301176470589f, 0.7319489411764706f, 0.6505192941176471f,
-0.6374468235294118f, 0.7288732549019608f, 0.6477510588235295f,
-0.6340635294117647f, 0.7257975686274509f, 0.6449828235294117f,
-0.6306802352941177f, 0.7227218823529411f, 0.6422145882352941f,
-0.6272969411764706f, 0.7196461960784313f, 0.6394463529411765f,
-0.6239136470588236f, 0.7165705098039216f, 0.6366781176470588f,
-0.6205303529411765f, 0.7134948235294117f, 0.6339098823529412f,
-0.6171470588235294f, 0.7104191372549019f, 0.6311416470588235f,
-0.6137637647058823f, 0.7073434509803921f, 0.6283734117647058f,
-0.6103804705882354f, 0.7042677647058824f, 0.6256051764705883f,
-0.6071509607843137f, 0.7011151764705882f, 0.6229907450980392f,
-0.6043828039215686f, 0.6977318823529411f, 0.6208377254901961f,
-0.6016146470588235f, 0.6943485882352941f, 0.6186847058823529f,
-0.5988464901960785f, 0.690965294117647f, 0.6165316862745098f,
-0.5960783333333334f, 0.6875819999999999f, 0.6143786666666666f,
-0.5933101764705883f, 0.6841987058823529f, 0.6122256470588235f,
-0.5905420196078431f, 0.6808154117647058f, 0.6100726274509803f,
-0.5877738627450981f, 0.6774321176470588f, 0.6079196078431373f,
-0.5850057058823529f, 0.6740488235294118f, 0.6057665882352942f,
-0.5822375490196078f, 0.6706655294117647f, 0.603613568627451f,
-0.5794693921568628f, 0.6672822352941176f, 0.6014605490196079f,
-0.5767012352941177f, 0.6638989411764705f, 0.5993075294117647f,
-0.5739330784313725f, 0.6605156470588235f, 0.5971545098039216f,
-0.5707035490196078f, 0.6569785490196078f, 0.5943863529411765f,
-0.5670126470588235f, 0.6532876470588235f, 0.5910030588235294f,
-0.5633217450980392f, 0.6495967450980392f, 0.5876197647058824f,
-0.5596308431372549f, 0.6459058431372549f, 0.5842364705882352f,
-0.5559399411764707f, 0.6422149411764707f, 0.5808531764705883f,
-0.5522490392156862f, 0.6385240392156862f, 0.5774698823529412f,
-0.548558137254902f, 0.634833137254902f, 0.5740865882352941f,
-0.5448672352941176f, 0.6311422352941176f, 0.5707032941176471f,
-0.5411763333333334f, 0.6274513333333334f, 0.56732f,
-0.537485431372549f, 0.623760431372549f, 0.563936705882353f,
-0.5337945294117648f, 0.6200695294117647f, 0.5605534117647059f,
-0.5301036274509804f, 0.6163786274509804f, 0.5571701176470588f,
-0.5264127254901961f, 0.6126877254901961f, 0.5537868235294118f,
-0.5238752352941176f, 0.6099195294117647f, 0.5517875882352942f,
-0.5217222156862745f, 0.6074589019607843f, 0.550249705882353f,
-0.5195691960784314f, 0.6049982745098039f, 0.5487118235294117f,
-0.5174161764705882f, 0.6025376470588235f, 0.5471739411764706f,
-0.5152631568627452f, 0.6000770196078431f, 0.5456360588235294f,
-0.513110137254902f, 0.5976163921568628f, 0.5440981764705882f,
-0.5109571176470589f, 0.5951557647058824f, 0.5425602941176471f,
-0.5088040980392158f, 0.5926951372549021f, 0.5410224117647059f,
-0.5066510784313726f, 0.5902345098039216f, 0.5394845294117647f,
-0.5044980588235294f, 0.5877738823529411f, 0.5379466470588234f,
-0.5023450392156863f, 0.5853132549019608f, 0.5364087647058823f,
-0.5001920196078431f, 0.5828526274509804f, 0.5348708823529411f,
-0.498039f, 0.580392f, 0.533333f,
-0.49557845098039216f, 0.5779314509803922f, 0.5314875882352941f,
-0.4931179019607843f, 0.5754709019607843f, 0.5296421764705882f,
-0.49065735294117646f, 0.5730103529411765f, 0.5277967647058823f,
-0.4881968039215686f, 0.5705498039215686f, 0.5259513529411765f,
-0.4857362549019608f, 0.5680892549019608f, 0.5241059411764706f,
-0.48327570588235297f, 0.5656287058823529f, 0.5222605294117647f,
-0.4808151568627451f, 0.5631681568627451f, 0.5204151176470588f,
-0.47835460784313727f, 0.5607076078431372f, 0.518569705882353f,
-0.4758940588235294f, 0.5582470588235294f, 0.5167242941176471f,
-0.47343350980392157f, 0.5557865098039215f, 0.5148788823529412f,
-0.4709729607843138f, 0.5533259607843137f, 0.5130334705882353f,
-0.46851241176470587f, 0.5508654117647058f, 0.5111880588235295f,
-0.4662056274509804f, 0.5484817450980392f, 0.5093426470588236f,
-0.46436013725490194f, 0.5463287254901961f, 0.5074972352941177f,
-0.4625146470588235f, 0.5441757058823529f, 0.5056518235294118f,
-0.4606691568627451f, 0.5420226862745098f, 0.5038064117647059f,
-0.4588236666666667f, 0.5398696666666666f, 0.501961f,
-0.45697817647058825f, 0.5377166470588235f, 0.5001155882352941f,
-0.4551326862745098f, 0.5355636274509803f, 0.49827017647058824f,
-0.4532871960784314f, 0.5334106078431372f, 0.49642476470588237f,
-0.45144170588235294f, 0.5312575882352941f, 0.4945793529411765f,
-0.4495962156862745f, 0.5291045686274509f, 0.4927339411764706f,
-0.44775072549019607f, 0.5269515490196078f, 0.49088852941176475f,
-0.44590523529411763f, 0.5247985294117646f, 0.4890431176470588f,
-0.4440597450980392f, 0.5226455098039215f, 0.48719770588235295f,
-0.4423680588235294f, 0.5208000588235293f, 0.4858136274509804f,
-0.44083017647058825f, 0.5192621764705883f, 0.4848908823529412f,
-0.4392922941176471f, 0.517724294117647f, 0.48396813725490195f,
-0.43775441176470586f, 0.5161864117647058f, 0.48304539215686276f,
-0.4362165294117647f, 0.5146485294117646f, 0.4821226470588235f,
-0.43467864705882353f, 0.5131106470588235f, 0.4811999019607843f,
-0.43314076470588236f, 0.5115727647058823f, 0.4802771568627451f,
-0.4316028823529412f, 0.5100348823529411f, 0.4793544117647059f,
-0.430065f, 0.508497f, 0.47843166666666664f,
-0.4285271176470588f, 0.5069591176470588f, 0.47750892156862745f,
-0.42698923529411764f, 0.5054212352941176f, 0.4765861764705882f,
-0.4254513529411765f, 0.5038833529411765f, 0.475663431372549f,
-0.4239134705882353f, 0.5023454705882353f, 0.47474068627450977f,
-0.42283699999999996f, 0.5005768823529412f, 0.4738179411764706f,
-0.42191433333333334f, 0.49873139215686274f, 0.4728951960784314f,
-0.42099166666666665f, 0.4968859019607843f, 0.47197245098039214f,
-0.42006899999999997f, 0.49504041176470587f, 0.47104970588235295f,
-0.41914633333333334f, 0.4931949215686275f, 0.47012696078431376f,
-0.41822366666666666f, 0.491349431372549f, 0.4692042156862745f,
-0.417301f, 0.48950394117647056f, 0.46828147058823527f,
-0.41637833333333335f, 0.4876584509803922f, 0.4673587254901961f,
-0.41545566666666667f, 0.48581296078431374f, 0.4664359803921569f,
-0.414533f, 0.4839674705882353f, 0.46551323529411764f,
-0.4136103333333333f, 0.48212198039215687f, 0.46459049019607845f,
-0.4126876666666667f, 0.48027649019607843f, 0.4636677450980392f,
-0.411765f, 0.478431f, 0.462745f,
-0.411149862745098f, 0.4768931960784314f, 0.4618222549019608f,
-0.4105347254901961f, 0.47535539215686273f, 0.4608995098039216f,
-0.4099195882352941f, 0.47381758823529413f, 0.4599767647058824f,
-0.40930445098039214f, 0.47227978431372547f, 0.45905401960784314f,
-0.40868931372549017f, 0.47074198039215687f, 0.45813127450980395f,
-0.40807417647058825f, 0.46920417647058826f, 0.4572085294117647f,
-0.40745903921568627f, 0.4676663725490196f, 0.4562857843137255f,
-0.40684390196078435f, 0.466128568627451f, 0.4553630392156863f,
-0.4062287647058824f, 0.46459076470588234f, 0.4544402941176471f,
-0.4056136274509804f, 0.46305296078431374f, 0.45351754901960784f,
-0.4049984901960784f, 0.4615151568627451f, 0.45259480392156864f,
-0.4043833529411765f, 0.4599773529411765f, 0.4516720588235294f,
-0.4037681960784314f, 0.4584395294117647f, 0.45067243137254903f,
-0.40315298039215686f, 0.4569016470588235f, 0.4494421568627451f,
-0.4025377647058824f, 0.45536376470588236f, 0.4482118823529412f,
-0.40192254901960783f, 0.4538258823529412f, 0.44698160784313723f,
-0.40130733333333335f, 0.452288f, 0.44575133333333333f,
-0.4006921176470588f, 0.4507501176470588f, 0.44452105882352944f,
-0.4000769019607843f, 0.44921223529411763f, 0.4432907843137255f,
-0.3994616862745098f, 0.44767435294117647f, 0.4420605098039216f,
-0.3988464705882353f, 0.4461364705882353f, 0.44083023529411763f,
-0.39823125490196076f, 0.44459858823529413f, 0.43959996078431374f,
-0.3976160392156863f, 0.4430607058823529f, 0.43836968627450984f,
-0.3970008235294118f, 0.4415228235294118f, 0.43713941176470594f,
-0.39638560784313726f, 0.4399849411764706f, 0.435909137254902f,
-0.395770431372549f, 0.4384470588235294f, 0.43467886274509804f,
-0.39515529411764705f, 0.43690917647058825f, 0.43344858823529414f,
-0.3945401568627451f, 0.4353712941176471f, 0.4322183137254902f,
-0.39392501960784315f, 0.43383341176470586f, 0.4309880392156863f,
-0.3933098823529412f, 0.4322955294117647f, 0.42975776470588234f,
-0.3926947450980392f, 0.4307576470588235f, 0.42852749019607844f,
-0.3920796078431372f, 0.42921976470588236f, 0.4272972156862745f,
-0.3914644705882353f, 0.4276818823529412f, 0.4260669411764706f,
-0.3908493333333333f, 0.42614399999999997f, 0.42483666666666664f,
-0.39023419607843135f, 0.4246061176470588f, 0.42360639215686274f,
-0.38961905882352943f, 0.42306823529411763f, 0.4223761176470588f,
-0.38900392156862745f, 0.42153035294117647f, 0.4211458431372549f,
-0.3883887843137255f, 0.4199924705882353f, 0.41991556862745094f,
-0.3877736470588235f, 0.41822388235294117f, 0.4184545882352941f,
-0.3871585098039216f, 0.4163783921568628f, 0.41691670588235297f,
-0.3865433725490196f, 0.4145329019607843f, 0.41537882352941174f,
-0.3859282352941176f, 0.41268741176470586f, 0.4138409411764706f,
-0.3853130980392157f, 0.4108419215686274f, 0.4123030588235294f,
-0.38469796078431373f, 0.408996431372549f, 0.41076517647058824f,
-0.38408282352941175f, 0.40715094117647055f, 0.4092272941176471f,
-0.38346768627450983f, 0.40530545098039217f, 0.4076894117647059f,
-0.38285254901960786f, 0.40345996078431373f, 0.4061515294117647f,
-0.3822374117647059f, 0.4016144705882353f, 0.4046136470588235f,
-0.3816222745098039f, 0.39976898039215686f, 0.40307576470588236f,
-0.381007137254902f, 0.3979234901960784f, 0.4015378823529412f,
-0.380392f, 0.396078f, 0.4f
-}
-, 
-{
-1.0f, 0.988235f, 0.968627f,
-1.0f, 0.9839289607843137f, 0.9532483333333334f,
-1.0f, 0.9796229215686274f, 0.9378696666666667f,
-1.0f, 0.9753168823529411f, 0.9224910000000001f,
-1.0f, 0.9710108431372548f, 0.9071123333333334f,
-1.0f, 0.9667048039215687f, 0.8917336666666666f,
-1.0f, 0.9623987647058824f, 0.876355f,
-0.9969242156862745f, 0.9569393921568627f, 0.8602074117647059f,
-0.9920029607843137f, 0.9507880196078431f, 0.8435984705882353f,
-0.987081705882353f, 0.9446366470588236f, 0.8269895294117647f,
-0.9821604509803922f, 0.9384852745098039f, 0.8103805882352941f,
-0.9772391960784313f, 0.9323339019607843f, 0.7937716470588235f,
-0.9723179411764706f, 0.9261825294117647f, 0.7771627058823529f,
-0.968319431372549f, 0.9212613921568628f, 0.7610150980392156f,
-0.9670891568627451f, 0.9200309607843137f, 0.7462514901960784f,
-0.9658588823529412f, 0.9188005294117647f, 0.7314878823529412f,
-0.9646286078431372f, 0.9175700980392156f, 0.7167242745098039f,
-0.9633983333333334f, 0.9163396666666667f, 0.7019606666666667f,
-0.9621680588235294f, 0.9151092352941177f, 0.6871970588235294f,
-0.9609377843137255f, 0.9138788039215686f, 0.6724334509803921f,
-0.958092705882353f, 0.9126485098039215f, 0.6608994509803922f,
-0.9550169411764706f, 0.9114182352941177f, 0.6498268235294117f,
-0.9519411764705882f, 0.9101879607843137f, 0.6387541960784313f,
-0.9488654117647058f, 0.9089576862745098f, 0.627681568627451f,
-0.9457896470588235f, 0.9077274117647058f, 0.6166089411764706f,
-0.9427138823529412f, 0.9064971372549019f, 0.6055363137254902f,
-0.9402533333333334f, 0.9068047450980392f, 0.5953863529411765f,
-0.938408f, 0.9086502352941176f, 0.5861590588235294f,
-0.9365626666666667f, 0.9104957254901961f, 0.5769317647058824f,
-0.9347173333333334f, 0.9123412156862745f, 0.5677044705882353f,
-0.932872f, 0.914186705882353f, 0.5584771764705883f,
-0.9310266666666667f, 0.9160321960784313f, 0.5492498823529411f,
-0.928796843137255f, 0.9174932156862745f, 0.5404070784313726f,
-0.9238755882352941f, 0.9162629411764706f, 0.5342557058823529f,
-0.9189543333333333f, 0.9150326666666666f, 0.5281043333333333f,
-0.9140330784313726f, 0.9138023921568628f, 0.5219529607843137f,
-0.9091118235294118f, 0.9125721176470588f, 0.5158015882352941f,
-0.9041905686274511f, 0.9113418431372549f, 0.5096502156862746f,
-0.8992693137254902f, 0.9101115686274509f, 0.5034988431372549f,
-0.8920413529411765f, 0.9074971764705881f, 0.4973473529411765f,
-0.8840444901960784f, 0.9044214117647058f, 0.49119582352941177f,
-0.8760476274509804f, 0.9013456470588235f, 0.4850442941176471f,
-0.8680507647058824f, 0.8982698823529411f, 0.4788927647058824f,
-0.8600539019607844f, 0.8951941176470588f, 0.47274123529411766f,
-0.8520570392156863f, 0.8921183529411765f, 0.46658970588235293f,
-0.8426760588235295f, 0.8883505882352941f, 0.46089958823529414f,
-0.8309882156862746f, 0.8834294901960784f, 0.45597849019607845f,
-0.8193003725490197f, 0.8785083921568627f, 0.45105739215686275f,
-0.8076125294117648f, 0.873587294117647f, 0.44613629411764705f,
-0.79592468627451f, 0.8686661960784314f, 0.4412151960784314f,
-0.784236843137255f, 0.8637450980392157f, 0.4362940980392157f,
-0.772549f, 0.858824f, 0.431373f,
-0.7590158235294118f, 0.8545179607843137f, 0.42645174509803924f,
-0.7454826470588236f, 0.8502119215686275f, 0.42153049019607847f,
-0.7319494705882353f, 0.8459058823529412f, 0.41660923529411764f,
-0.7184162941176471f, 0.8415998431372549f, 0.41168798039215687f,
-0.7048831176470588f, 0.8372938039215687f, 0.4067667254901961f,
-0.6913499411764706f, 0.8329877647058824f, 0.40184547058823533f,
-0.6782011764705882f, 0.8282972156862746f, 0.3969242156862745f,
-0.6652830588235293f, 0.8233759607843137f, 0.39200296078431374f,
-0.6523649411764706f, 0.818454705882353f, 0.38708170588235297f,
-0.6394468235294118f, 0.8135334509803922f, 0.3821604509803922f,
-0.6265287058823529f, 0.8086121960784314f, 0.37723919607843137f,
-0.613610588235294f, 0.8036909411764707f, 0.3723179411764706f,
-0.6005386666666667f, 0.7987696862745098f, 0.3673967254901961f,
-0.5870053333333333f, 0.793848431372549f, 0.3624756274509804f,
-0.5734720000000002f, 0.7889271764705883f, 0.3575545294117648f,
-0.5599386666666666f, 0.7840059215686275f, 0.35263343137254904f,
-0.5464053333333333f, 0.7790846666666666f, 0.34771233333333335f,
-0.532872f, 0.7741634117647058f, 0.34279123529411765f,
-0.5193386666666666f, 0.769242156862745f, 0.337870137254902f,
-0.5036523529411765f, 0.7621680588235293f, 0.3324107254901961f,
-0.48765847058823525f, 0.7547864117647058f, 0.3268744117647059f,
-0.4716645882352941f, 0.7474047647058824f, 0.3213380980392157f,
-0.45567070588235314f, 0.740023117647059f, 0.3158017843137256f,
-0.43967682352941173f, 0.7326414705882353f, 0.3102654705882353f,
-0.4236829411764706f, 0.7252598235294118f, 0.3047291568627451f,
-0.40861180392156865f, 0.7181856274509804f, 0.29950037254901957f,
-0.3944634117647059f, 0.7114188823529412f, 0.2945791176470588f,
-0.38031501960784314f, 0.704652137254902f, 0.289657862745098f,
-0.3661666274509804f, 0.6978853921568627f, 0.28473660784313726f,
-0.35201823529411763f, 0.6911186470588235f, 0.27981535294117643f,
-0.3378698431372551f, 0.6843519019607844f, 0.2748940980392158f,
-0.32395213725490196f, 0.6778927450980392f, 0.2701266470588235f,
-0.31164923529411764f, 0.6735867058823529f, 0.26643582352941175f,
-0.2993463333333333f, 0.6692806666666666f, 0.262745f,
-0.287043431372549f, 0.6649746274509803f, 0.25905417647058826f,
-0.2747405294117647f, 0.6606685882352942f, 0.25536335294117646f,
-0.26243762745098037f, 0.6563625490196079f, 0.2516725294117647f,
-0.2501347254901961f, 0.6520565098039216f, 0.24798170588235294f,
-0.24382947058823534f, 0.6463664705882354f, 0.24936582352941172f,
-0.23952343137254903f, 0.6402150980392157f, 0.2524415882352941f,
-0.23521739215686274f, 0.6340637254901961f, 0.25551735294117645f,
-0.23091135294117648f, 0.6279123529411764f, 0.25859311764705883f,
-0.2266053137254902f, 0.6217609803921569f, 0.26166888235294117f,
-0.2222992745098039f, 0.6156096078431372f, 0.2647446470588235f,
-0.21799323529411765f, 0.6099195294117647f, 0.2675897058823529f,
-0.21368719607843137f, 0.6049982745098039f, 0.2700502549019608f,
-0.20938115686274517f, 0.6000770196078432f, 0.27251080392156857f,
-0.20507511764705882f, 0.5951557647058824f, 0.2749713529411765f,
-0.20076907843137254f, 0.5902345098039216f, 0.2774319019607843f,
-0.19646303921568628f, 0.5853132549019608f, 0.27989245098039217f,
-0.192157f, 0.580392f, 0.282353f,
-0.1866206862745098f, 0.5773162352941177f, 0.2835832745098039f,
-0.1810843725490196f, 0.5742404705882352f, 0.28481354901960787f,
-0.17554805882352942f, 0.5711647058823529f, 0.28604382352941177f,
-0.1700117450980393f, 0.5680889411764706f, 0.28727409803921566f,
-0.16447543137254902f, 0.5650131764705882f, 0.2885043725490196f,
-0.15893911764705881f, 0.5619374117647058f, 0.2897346470588235f,
-0.15301829411764706f, 0.5588616470588235f, 0.2905805098039216f,
-0.14686676470588236f, 0.5557858823529411f, 0.29119572549019607f,
-0.14071523529411764f, 0.5527101176470588f, 0.2918109411764706f,
-0.13456370588235295f, 0.5496343529411765f, 0.2924261568627451f,
-0.12841217647058822f, 0.5465585882352941f, 0.29304137254901963f,
-0.12226064705882361f, 0.5434828235294118f, 0.2936565882352941f,
-0.11657049019607843f, 0.5404070980392157f, 0.2945793333333333f,
-0.11226445098039216f, 0.5373314901960784f, 0.29642466666666667f,
-0.10795841176470589f, 0.5342558823529412f, 0.29827f,
-0.10365237254901961f, 0.5311802745098039f, 0.30011533333333335f,
-0.09934633333333334f, 0.5281046666666667f, 0.30196066666666666f,
-0.09504029411764706f, 0.5250290588235293f, 0.30380599999999996f,
-0.09073425490196078f, 0.5219534509803921f, 0.30565133333333333f,
-0.08750470588235298f, 0.5199541960784314f, 0.308573294117647f,
-0.08442894117647058f, 0.5181087058823529f, 0.3116490588235294f,
-0.08135317647058823f, 0.5162632156862745f, 0.3147248235294118f,
-0.07827741176470589f, 0.514417725490196f, 0.3178005882352941f,
-0.07520164705882353f, 0.5125722352941177f, 0.32087635294117645f,
-0.07212588235294118f, 0.5107267450980393f, 0.32395211764705883f,
-0.07151074509803922f, 0.509804f, 0.32887329411764704f,
-0.07335623529411764f, 0.509804f, 0.3356398823529412f,
-0.07520172549019608f, 0.509804f, 0.3424064705882353f,
-0.07704721568627451f, 0.509804f, 0.3491730588235294f,
-0.07889270588235288f, 0.509804f, 0.3559396470588233f,
-0.08073819607843137f, 0.509804f, 0.3627062352941176f,
-0.08250678431372549f, 0.5100346862745099f, 0.36947284313725487f,
-0.08373705882352941f, 0.5118801764705883f, 0.3762395882352941f,
-0.08496733333333333f, 0.5137256666666666f, 0.38300633333333334f,
-0.08619760784313725f, 0.5155711568627451f, 0.38977307843137254f,
-0.08742788235294117f, 0.5174166470588235f, 0.39653982352941175f,
-0.0886581568627451f, 0.5192621372549019f, 0.40330656862745096f,
-0.08988843137254902f, 0.5211076274509804f, 0.41007331372549016f,
-0.09065741176470588f, 0.5224917058823529f, 0.4173013529411765f,
-0.09127262745098039f, 0.5237219803921568f, 0.4246831568627451f,
-0.0918878431372549f, 0.5249522549019607f, 0.4320649607843137f,
-0.09250305882352941f, 0.5261825294117647f, 0.43944676470588234f,
-0.09311827450980392f, 0.5274128039215686f, 0.446828568627451f,
-0.09373349019607842f, 0.5286430784313726f, 0.45421037254901964f,
-0.09457935294117646f, 0.530104f, 0.4613614705882353f,
-0.09580962745098034f, 0.5319493333333333f, 0.4681280588235292f,
-0.09703990196078431f, 0.5337946666666666f, 0.47489464705882356f,
-0.09827017647058824f, 0.53564f, 0.48166123529411764f,
-0.09950045098039215f, 0.5374853333333334f, 0.4884278235294118f,
-0.10073072549019607f, 0.5393306666666666f, 0.49519441176470586f,
-0.101961f, 0.541176f, 0.501961f,
-0.10380633333333333f, 0.542406431372549f, 0.5093428039215686f,
-0.10565166666666667f, 0.543636862745098f, 0.5167246078431372f,
-0.107497f, 0.544867294117647f, 0.5241064117647058f,
-0.10934233333333333f, 0.5460977254901961f, 0.5314882156862745f,
-0.11118766666666667f, 0.547328156862745f, 0.5388700196078431f,
-0.11303300000000001f, 0.548558588235294f, 0.5462518235294117f,
-0.1152629411764706f, 0.5470976470588235f, 0.5501733333333333f,
-0.11772364705882353f, 0.5440218823529411f, 0.5520186666666667f,
-0.12018435294117648f, 0.5409461176470588f, 0.5538639999999999f,
-0.12264505882352941f, 0.5378703529411765f, 0.5557093333333333f,
-0.12510576470588228f, 0.5347945882352942f, 0.5575546666666665f,
-0.12756647058823528f, 0.5317188235294118f, 0.5593999999999999f,
-0.13002713725490195f, 0.529104431372549f, 0.5615529411764706f,
-0.1324876862745098f, 0.5278741568627451f, 0.5646287058823529f,
-0.13494823529411765f, 0.5266438823529411f, 0.5677044705882353f,
-0.1374087843137255f, 0.5254136078431372f, 0.5707802352941176f,
-0.13986933333333335f, 0.5241833333333333f, 0.573856f,
-0.14232988235294117f, 0.5229530588235294f, 0.5769317647058824f,
-0.14479043137254902f, 0.5217227843137254f, 0.5800075294117647f,
-0.14671280392156863f, 0.5183393921568628f, 0.583083294117647f,
-0.14855829411764707f, 0.5146484117647059f, 0.5861590588235294f,
-0.1504037843137255f, 0.510957431372549f, 0.5892348235294118f,
-0.1522492745098039f, 0.5072664509803921f, 0.5923105882352941f,
-0.15409476470588235f, 0.5035754705882353f, 0.5953863529411765f,
-0.15594025490196078f, 0.49988449019607845f, 0.5984621176470588f,
-0.15747813725490195f, 0.494040568627451f, 0.6009227450980392f,
-0.15870841176470585f, 0.4860437058823532f, 0.6027682352941176f,
-0.1599386862745098f, 0.4780468431372549f, 0.6046137254901961f,
-0.16116896078431373f, 0.47004998039215684f, 0.6064592156862745f,
-0.16239923529411765f, 0.4620531176470588f, 0.608304705882353f,
-0.16362950980392157f, 0.4540562549019608f, 0.6101501960784314f,
-0.16485978431372547f, 0.4460593725490196f, 0.6119187843137255f,
-0.1660900588235294f, 0.43806235294117646f, 0.6131490588235294f,
-0.16732033333333332f, 0.4300653333333333f, 0.6143793333333334f,
-0.16855060784313725f, 0.4220683137254902f, 0.6156096078431372f,
-0.16978088235294117f, 0.41407129411764704f, 0.6168398823529412f,
-0.1710111568627451f, 0.40607427450980393f, 0.6180701568627451f,
-0.17224143137254902f, 0.3980772549019608f, 0.6193004313725491f,
-0.17301041176470588f, 0.38961894117647056f, 0.6209921176470589f,
-0.1736256274509804f, 0.38100686274509804f, 0.6228376078431372f,
-0.1742408431372549f, 0.37239478431372547f, 0.6246830980392157f,
-0.17485605882352942f, 0.36378270588235295f, 0.6265285882352941f,
-0.1754712745098039f, 0.3551706274509806f, 0.6283740784313725f,
-0.17608649019607842f, 0.3465585490196078f, 0.630219568627451f,
-0.17693235294117646f, 0.33956123529411764f, 0.6318343529411764f,
-0.17816262745098038f, 0.33525519607843135f, 0.6330646274509804f,
-0.1793929019607843f, 0.33094915686274506f, 0.6342949019607843f,
-0.18062317647058823f, 0.3266431176470588f, 0.6355251764705883f,
-0.18185345098039216f, 0.32233707843137255f, 0.6367554509803921f,
-0.18308372549019608f, 0.31803103921568626f, 0.637985725490196f,
-0.184314f, 0.313725f, 0.639216f,
-0.184314f, 0.3075736274509804f, 0.6361402352941177f,
-0.184314f, 0.3014222549019608f, 0.6330644705882353f,
-0.184314f, 0.29527088235294113f, 0.629988705882353f,
-0.184314f, 0.28911950980392154f, 0.6269129411764707f,
-0.184314f, 0.28296813725490194f, 0.6238371764705882f,
-0.184314f, 0.27681676470588235f, 0.6207614117647059f,
-0.184314f, 0.27066529411764706f, 0.6176856470588236f,
-0.184314f, 0.2645137647058825f, 0.6146098823529413f,
-0.184314f, 0.2583622352941176f, 0.6115341176470589f,
-0.184314f, 0.25221070588235295f, 0.6084583529411765f,
-0.184314f, 0.24605917647058823f, 0.6053825882352941f,
-0.184314f, 0.23990764705882353f, 0.6023068235294118f,
-0.184314f, 0.23390992156862744f, 0.5992310588235293f,
-0.184314f, 0.22837360784313726f, 0.596155294117647f,
-0.184314f, 0.22283729411764708f, 0.5930795294117647f,
-0.184314f, 0.21730098039215687f, 0.5900037647058823f,
-0.184314f, 0.21176466666666668f, 0.586928f,
-0.184314f, 0.20622835294117647f, 0.5838522352941177f,
-0.184314f, 0.2006920392156863f, 0.5807764705882353f,
-0.18861996078431373f, 0.1994616862745098f, 0.577700705882353f,
-0.1935410588235294f, 0.1988464705882353f, 0.5746249411764706f,
-0.1984621568627451f, 0.19823125490196078f, 0.5715491764705882f,
-0.20338325490196077f, 0.19761603921568627f, 0.5684734117647059f,
-0.20830435294117633f, 0.19700082352941178f, 0.5653976470588236f,
-0.21322545098039214f, 0.19638560784313724f, 0.5623218823529411f,
-0.21722388235294116f, 0.1957704705882353f, 0.556785568627451f,
-0.22029964705882352f, 0.19515541176470588f, 0.548788705882353f,
-0.22337541176470588f, 0.19454035294117647f, 0.5407918431372549f,
-0.22645117647058824f, 0.19392529411764706f, 0.5327949803921569f,
-0.2295269411764706f, 0.19331023529411764f, 0.5247981176470589f,
-0.23260270588235293f, 0.19269517647058823f, 0.5168012549019608f,
-0.23583225490196078f, 0.19231078431372547f, 0.509188843137255f,
-0.24013829411764706f, 0.1935410588235294f, 0.5042675882352942f,
-0.24444433333333335f, 0.19477133333333332f, 0.49934633333333334f,
-0.2487503725490196f, 0.19600160784313725f, 0.49442507843137257f,
-0.2530564117647059f, 0.19723188235294117f, 0.4895038235294118f,
-0.2573624509803922f, 0.1984621568627451f, 0.48458256862745097f,
-0.26166849019607846f, 0.19969243137254902f, 0.4796613137254902f,
-0.26551323529411763f, 0.20230682352941176f, 0.4738174705882353f,
-0.2692042156862744f, 0.20538258823529404f, 0.4676660980392159f,
-0.2728951960784314f, 0.20845835294117648f, 0.46151472549019606f,
-0.27658617647058825f, 0.21153411764705882f, 0.45536335294117647f,
-0.2802771568627451f, 0.21460988235294118f, 0.4492119803921569f,
-0.28396813725490194f, 0.21768564705882354f, 0.4430606078431373f,
-0.28719770588235294f, 0.22076141176470587f, 0.43690917647058825f,
-0.2896582549019608f, 0.22383717647058823f, 0.4307576470588235f,
-0.29211880392156864f, 0.2269129411764706f, 0.4246061176470588f,
-0.2945793529411765f, 0.22998870588235296f, 0.41845458823529413f,
-0.2970399019607843f, 0.2330644705882353f, 0.4123030588235294f,
-0.29950045098039213f, 0.23614023529411765f, 0.40615152941176474f,
-0.301961f, 0.239216f, 0.4f
-}
-, 
-{
-1.0f, 1.0f, 0.988235f,
-0.9993720473856209f, 0.9995290155228758f, 0.9824263382352941f,
-0.9980905114379085f, 0.9985678227124183f, 0.9705719264705882f,
-0.9968089754901961f, 0.9976066299019608f, 0.9587175147058823f,
-0.9955274395424837f, 0.9966454370915032f, 0.9468631029411765f,
-0.9942459035947713f, 0.9956842442810457f, 0.9350086911764706f,
-0.9929643676470589f, 0.9947230514705883f, 0.9231542794117646f,
-0.9916828316993465f, 0.9937618586601307f, 0.9112998676470588f,
-0.990401295751634f, 0.9928006658496732f, 0.8994454558823529f,
-0.9891197598039216f, 0.9918394730392157f, 0.8875910441176471f,
-0.9878382238562091f, 0.9908782802287581f, 0.8757366323529412f,
-0.9865566879084967f, 0.9899170874183006f, 0.8638822205882353f,
-0.9852751519607843f, 0.9889558946078432f, 0.8520278088235295f,
-0.9843908823529411f, 0.988235f, 0.8411377647058824f,
-0.9846984117647058f, 0.988235f, 0.8331408235294118f,
-0.9850059411764706f, 0.988235f, 0.8251438823529412f,
-0.9853134705882353f, 0.988235f, 0.8171469411764706f,
-0.985621f, 0.988235f, 0.80915f,
-0.9859285294117647f, 0.988235f, 0.8011530588235294f,
-0.9862360588235294f, 0.988235f, 0.7931561176470588f,
-0.9865435882352941f, 0.988235f, 0.7851591764705882f,
-0.9868511176470588f, 0.988235f, 0.7771622352941177f,
-0.9871586470588235f, 0.988235f, 0.769165294117647f,
-0.9874661764705882f, 0.988235f, 0.7611683529411765f,
-0.9877737058823529f, 0.988235f, 0.7531714117647058f,
-0.9880812352941176f, 0.988235f, 0.7451744705882353f,
-0.987927431372549f, 0.9874660588235293f, 0.7377927058823529f,
-0.9873122941176471f, 0.9859281764705882f, 0.7310261176470588f,
-0.9866971568627451f, 0.984390294117647f, 0.7242595294117646f,
-0.9860820196078431f, 0.9828524117647058f, 0.7174929411764706f,
-0.9854668823529412f, 0.9813145294117647f, 0.7107263529411765f,
-0.9848517450980392f, 0.9797766470588235f, 0.7039597647058823f,
-0.9842366078431373f, 0.9782387647058823f, 0.6971931764705882f,
-0.9836214705882353f, 0.9767008823529412f, 0.6904265882352941f,
-0.9830063333333333f, 0.975163f, 0.6836599999999999f,
-0.9823911960784314f, 0.9736251176470588f, 0.6768934117647059f,
-0.9817760588235295f, 0.9720872352941177f, 0.6701268235294118f,
-0.9811609215686274f, 0.9705493529411765f, 0.6633602352941177f,
-0.9805457843137255f, 0.9690114705882353f, 0.6565936470588235f,
-0.980392f, 0.9672429411764706f, 0.6502884117647059f,
-0.980392f, 0.9653975294117647f, 0.6441369607843137f,
-0.980392f, 0.9635521176470588f, 0.6379855098039215f,
-0.980392f, 0.961706705882353f, 0.6318340588235294f,
-0.980392f, 0.9598612941176471f, 0.6256826078431372f,
-0.980392f, 0.9580158823529412f, 0.6195311568627451f,
-0.980392f, 0.9561704705882353f, 0.6133797058823529f,
-0.980392f, 0.9543250588235295f, 0.6072282549019608f,
-0.980392f, 0.9524796470588235f, 0.6010768039215686f,
-0.980392f, 0.9506342352941176f, 0.5949253529411764f,
-0.980392f, 0.9487888235294117f, 0.5887739019607843f,
-0.980392f, 0.9469434117647059f, 0.5826224509803921f,
-0.980392f, 0.945098f, 0.576471f,
-0.9794692549019608f, 0.9420222352941177f, 0.5693968039215686f,
-0.9785465098039217f, 0.9389464705882353f, 0.5623226078431373f,
-0.9776237647058824f, 0.9358707058823529f, 0.5552484117647059f,
-0.9767010196078432f, 0.9327949411764705f, 0.5481742156862744f,
-0.9757782745098039f, 0.9297191764705882f, 0.5411000196078432f,
-0.9748555294117648f, 0.9266434117647059f, 0.5340258235294117f,
-0.9739327843137255f, 0.9235676470588235f, 0.5269516274509803f,
-0.9730100392156863f, 0.9204918823529411f, 0.519877431372549f,
-0.972087294117647f, 0.9174161176470588f, 0.5128032352941176f,
-0.9711645490196079f, 0.9143403529411764f, 0.5057290392156863f,
-0.9702418039215687f, 0.9112645882352941f, 0.4986548431372549f,
-0.9693190588235294f, 0.9081888235294118f, 0.49158064705882354f,
-0.968627f, 0.9050361764705882f, 0.4843526470588235f,
-0.968627f, 0.9016528823529412f, 0.47666323529411764f,
-0.968627f, 0.8982695882352941f, 0.46897382352941186f,
-0.968627f, 0.894886294117647f, 0.4612844117647059f,
-0.968627f, 0.8915029999999999f, 0.453595f,
-0.968627f, 0.8881197058823529f, 0.44590558823529414f,
-0.968627f, 0.8847364117647059f, 0.43821617647058825f,
-0.968627f, 0.8813531176470588f, 0.43052676470588236f,
-0.968627f, 0.8779698235294118f, 0.42283735294117647f,
-0.968627f, 0.8745865294117647f, 0.4151479411764706f,
-0.968627f, 0.8712032352941177f, 0.4074585294117648f,
-0.968627f, 0.8678199411764705f, 0.3997691176470588f,
-0.968627f, 0.8644366470588235f, 0.39207970588235297f,
-0.968319431372549f, 0.8604382156862745f, 0.384236568627451f,
-0.967704294117647f, 0.8558246470588236f, 0.37623970588235295f,
-0.9670891568627451f, 0.8512110784313726f, 0.3682428431372549f,
-0.9664740196078432f, 0.8465975098039216f, 0.3602459803921569f,
-0.9658588823529412f, 0.8419839411764706f, 0.3522491176470588f,
-0.9652437450980392f, 0.8373703725490197f, 0.3442522549019609f,
-0.9646286078431372f, 0.8327568039215686f, 0.33625539215686273f,
-0.9640134705882353f, 0.8281432352941176f, 0.3282585294117647f,
-0.9633983333333334f, 0.8235296666666667f, 0.32026166666666667f,
-0.9627831960784313f, 0.8189160980392157f, 0.31226480392156863f,
-0.9621680588235294f, 0.8143025294117647f, 0.3042679411764706f,
-0.9615529215686274f, 0.8096889607843137f, 0.29627107843137257f,
-0.9609377843137255f, 0.8050753921568627f, 0.28827421568627454f,
-0.960092f, 0.8002310588235295f, 0.28235341176470596f,
-0.9591693333333333f, 0.7953098039215687f, 0.2771246274509804f,
-0.9582466666666667f, 0.7903885490196079f, 0.2718958431372549f,
-0.957324f, 0.785467294117647f, 0.2666670588235294f,
-0.9564013333333333f, 0.7805460392156862f, 0.26143827450980395f,
-0.9554786666666666f, 0.7756247843137255f, 0.2562094901960784f,
-0.954556f, 0.7707035294117647f, 0.25098070588235294f,
-0.9536333333333333f, 0.7657822745098039f, 0.24575192156862746f,
-0.9527106666666667f, 0.7608610196078431f, 0.24052313725490204f,
-0.951788f, 0.7559397647058823f, 0.23529435294117648f,
-0.9508653333333333f, 0.7510185098039215f, 0.23006556862745098f,
-0.9499426666666666f, 0.7460972549019608f, 0.2248367843137255f,
-0.94902f, 0.741176f, 0.219608f,
-0.9484047843137254f, 0.7362548235294117f, 0.21407168627450981f,
-0.947789568627451f, 0.7313336470588235f, 0.2085353725490196f,
-0.9471743529411765f, 0.7264124705882352f, 0.20299905882352942f,
-0.9465591372549019f, 0.7214912941176471f, 0.1974627450980393f,
-0.9459439215686274f, 0.7165701176470588f, 0.19192643137254903f,
-0.945328705882353f, 0.7116489411764706f, 0.18639011764705882f,
-0.9447134901960784f, 0.7067277647058823f, 0.18085380392156863f,
-0.9440982745098039f, 0.7018065882352941f, 0.17531749019607845f,
-0.9434830588235295f, 0.6968854117647059f, 0.16978117647058824f,
-0.9428678431372549f, 0.6919642352941177f, 0.16424486274509806f,
-0.9422526274509804f, 0.6870430588235295f, 0.15870854901960785f,
-0.941637411764706f, 0.6821218823529412f, 0.15317223529411772f,
-0.9409453333333333f, 0.6770469215686274f, 0.14794349019607844f,
-0.9400226666666667f, 0.6715106078431372f, 0.14363745098039216f,
-0.9391f, 0.6659742941176471f, 0.1393314117647059f,
-0.9381773333333333f, 0.6604379803921568f, 0.1350253725490196f,
-0.9372546666666667f, 0.6549016666666667f, 0.13071933333333335f,
-0.936332f, 0.6493653529411765f, 0.12641329411764707f,
-0.9354093333333333f, 0.6438290392156862f, 0.12210725490196078f,
-0.9344866666666667f, 0.6382927254901962f, 0.11780121568627458f,
-0.9335640000000001f, 0.6327564117647059f, 0.11349517647058824f,
-0.9326413333333333f, 0.6272200980392157f, 0.10918913725490195f,
-0.9317186666666667f, 0.6216837843137255f, 0.10488309803921568f,
-0.9307960000000001f, 0.6161474705882353f, 0.10057705882352941f,
-0.9298733333333333f, 0.6106111568627451f, 0.09627101960784312f,
-0.9291044313725491f, 0.6053824117647059f, 0.09258011764705881f,
-0.928489294117647f, 0.6004612352941177f, 0.08950435294117647f,
-0.9278741568627451f, 0.5955400588235295f, 0.08642858823529412f,
-0.9272590196078432f, 0.5906188823529411f, 0.08335282352941176f,
-0.9266438823529412f, 0.585697705882353f, 0.08027705882352949f,
-0.9260287450980392f, 0.5807765294117647f, 0.07720129411764706f,
-0.9254136078431372f, 0.5758553529411765f, 0.0741255294117647f,
-0.9247984705882353f, 0.5709341764705882f, 0.07104976470588235f,
-0.9241833333333334f, 0.566013f, 0.067974f,
-0.9235681960784313f, 0.5610918235294118f, 0.06489823529411765f,
-0.9229530588235294f, 0.5561706470588235f, 0.06182247058823529f,
-0.9223379215686274f, 0.5512494705882353f, 0.05874670588235294f,
-0.9217227843137255f, 0.546328294117647f, 0.05567094117647059f,
-0.9208769411764706f, 0.5416378235294117f, 0.05374858823529412f,
-0.9199541960784313f, 0.5370242549019607f, 0.05221070588235294f,
-0.9190314509803921f, 0.5324106862745098f, 0.05067282352941176f,
-0.9181087058823529f, 0.5277971176470588f, 0.04913494117647059f,
-0.9171859607843137f, 0.5231835490196078f, 0.04759705882352941f,
-0.9162632156862744f, 0.5185699803921568f, 0.046059176470588235f,
-0.9153404705882352f, 0.5139564117647059f, 0.04452129411764706f,
-0.9144177254901961f, 0.5093428431372551f, 0.04298341176470592f,
-0.9134949803921568f, 0.5047292745098039f, 0.04144552941176471f,
-0.9125722352941176f, 0.500115705882353f, 0.039907647058823534f,
-0.9116494901960783f, 0.495502137254902f, 0.03836976470588235f,
-0.9107267450980392f, 0.490888568627451f, 0.03683188235294117f,
-0.909804f, 0.486275f, 0.035294f,
-0.9082661176470588f, 0.48043107843137256f, 0.034063725490196074f,
-0.9067282352941176f, 0.4745871568627451f, 0.032833450980392156f,
-0.9051903529411764f, 0.46874323529411765f, 0.03160317647058823f,
-0.9036524705882353f, 0.4628993137254902f, 0.030372901960784314f,
-0.9021145882352941f, 0.45705539215686275f, 0.029142627450980393f,
-0.900576705882353f, 0.4512114705882353f, 0.02791235294117647f,
-0.8990388235294118f, 0.44536754901960784f, 0.026682078431372547f,
-0.8975009411764706f, 0.4395236274509804f, 0.02545180392156863f,
-0.8959630588235293f, 0.43367970588235294f, 0.024221529411764704f,
-0.8944251764705882f, 0.4278357843137255f, 0.022991254901960786f,
-0.8928872941176471f, 0.4219918627450982f, 0.021760980392156896f,
-0.8913494117647058f, 0.4161479411764706f, 0.020530705882352944f,
-0.8884274509803921f, 0.409688862745098f, 0.019223529411764705f,
-0.8813532549019608f, 0.4013843137254902f, 0.017685647058823528f,
-0.8742790588235294f, 0.39307976470588235f, 0.016147764705882355f,
-0.867204862745098f, 0.3847752156862745f, 0.014609882352941177f,
-0.8601306666666667f, 0.3764706666666667f, 0.013072f,
-0.8530564705882353f, 0.3681661176470588f, 0.011534117647058825f,
-0.8459822745098039f, 0.359861568627451f, 0.009996235294117648f,
-0.8389080784313726f, 0.35155701960784314f, 0.00845835294117647f,
-0.8318338823529412f, 0.34325247058823527f, 0.0069204705882352936f,
-0.8247596862745098f, 0.33494792156862746f, 0.005382588235294118f,
-0.8176854901960785f, 0.3266433725490196f, 0.003844705882352941f,
-0.8106112941176471f, 0.3183388235294118f, 0.002306823529411764f,
-0.8035370980392157f, 0.3100342745098039f, 0.0007689411764705868f,
-0.7984621176470589f, 0.30326764705882353f, 0.0f,
-0.7953863529411765f, 0.29803894117647073f, 0.0f,
-0.7923105882352942f, 0.29281023529411765f, 0.0f,
-0.7892348235294118f, 0.2875815294117647f, 0.0f,
-0.7861590588235294f, 0.28235282352941177f, 0.0f,
-0.7830832941176471f, 0.2771241176470588f, 0.0f,
-0.7800075294117648f, 0.2718954117647059f, 0.0f,
-0.7769317647058824f, 0.266666705882353f, 0.0f,
-0.773856f, 0.261438f, 0.0f,
-0.7707802352941177f, 0.25620929411764704f, 0.0f,
-0.7677044705882353f, 0.2509805882352941f, 0.0f,
-0.764628705882353f, 0.24575188235294118f, 0.0f,
-0.7615529411764707f, 0.24052317647058824f, 0.0f,
-0.7559397058823529f, 0.23575576470588236f, 0.0006920588235294117f,
-0.7494806470588236f, 0.23114211764705883f, 0.0016148039215686276f,
-0.7430215882352942f, 0.2265284705882353f, 0.002537549019607843f,
-0.7365625294117647f, 0.22191482352941178f, 0.003460294117647059f,
-0.7301034705882354f, 0.2173011764705884f, 0.004383039215686248f,
-0.7236444117647058f, 0.21268752941176472f, 0.00530578431372549f,
-0.7171853529411765f, 0.2080738823529412f, 0.006228529411764706f,
-0.7107262941176471f, 0.20346023529411766f, 0.00715127450980392f,
-0.7042672352941176f, 0.1988465882352941f, 0.008074019607843137f,
-0.6978081764705882f, 0.1942329411764706f, 0.008996764705882352f,
-0.6913491176470589f, 0.18961929411764705f, 0.009919509803921568f,
-0.6848900588235294f, 0.18500564705882352f, 0.010842254901960783f,
-0.678431f, 0.180392f, 0.011765f,
-0.6722795490196078f, 0.175778431372549f, 0.012687666666666667f,
-0.6661280980392157f, 0.17116486274509804f, 0.013610333333333334f,
-0.6599766470588235f, 0.16655129411764705f, 0.014533f,
-0.6538251960784314f, 0.16193772549019608f, 0.015455666666666666f,
-0.6476737450980392f, 0.1573241568627451f, 0.016378333333333335f,
-0.641522294117647f, 0.15271058823529413f, 0.017301f,
-0.6353708431372549f, 0.14809701960784313f, 0.018223666666666666f,
-0.629219392156863f, 0.14348345098039228f, 0.019146333333333307f,
-0.6230679411764706f, 0.13886988235294118f, 0.020069000000000004f,
-0.6169164901960784f, 0.13425631372549018f, 0.020991666666666665f,
-0.6107650392156863f, 0.12964274509803922f, 0.021914333333333334f,
-0.6046135882352941f, 0.12502917647058823f, 0.022837000000000003f,
-0.5980776666666666f, 0.12026180392156863f, 0.02368280392156863f,
-0.5903883333333333f, 0.11503301960784314f, 0.024298019607843137f,
-0.582699f, 0.10980423529411765f, 0.02491323529411765f,
-0.5750096666666666f, 0.10457545098039216f, 0.025528450980392157f,
-0.5673203333333333f, 0.09934666666666667f, 0.026143666666666666f,
-0.559631f, 0.09411788235294118f, 0.026758882352941177f,
-0.5519416666666667f, 0.08888909803921569f, 0.027374098039215686f,
-0.5442523333333333f, 0.08366031372549018f, 0.027989313725490197f,
-0.536563f, 0.0784315294117647f, 0.028604529411764706f,
-0.5288736666666667f, 0.0732027450980392f, 0.029219745098039214f,
-0.5211843333333334f, 0.06797396078431373f, 0.029834960784313726f,
-0.5134950000000003f, 0.06274517647058839f, 0.030450176470588217f,
-0.5058056666666666f, 0.05751639215686273f, 0.031065392156862742f,
-0.4979625294117647f, 0.05428686274509804f, 0.03244950980392157f,
-0.4899655882352941f, 0.05305658823529412f, 0.034602529411764706f,
-0.48196864705882353f, 0.051826313725490195f, 0.03675554901960784f,
-0.47397170588235293f, 0.05059603921568628f, 0.03890856862745098f,
-0.46597476470588234f, 0.04936576470588235f, 0.041061588235294114f,
-0.45797782352941174f, 0.048135490196078434f, 0.04321460784313726f,
-0.4499808823529412f, 0.04690521568627451f, 0.04536762745098039f,
-0.4419839411764706f, 0.045674941176470585f, 0.04752064705882353f,
-0.433987f, 0.04444466666666667f, 0.049673666666666665f,
-0.4259900588235294f, 0.04321439215686275f, 0.0518266862745098f,
-0.4179931176470588f, 0.041984117647058825f, 0.053979705882352944f,
-0.4099961764705883f, 0.0407538431372549f, 0.05613272549019608f,
-0.4019992352941177f, 0.03952356862745098f, 0.058285745098039216f,
-0.394233f, 0.03967735294117647f, 0.06066941176470588f,
-0.3865436666666669f, 0.04029249019607842f, 0.06312996078431365f,
-0.37885433333333335f, 0.040907627450980394f, 0.06559050980392157f,
-0.371165f, 0.041522764705882356f, 0.06805105882352941f,
-0.36347566666666664f, 0.04213790196078431f, 0.07051160784313726f,
-0.3557863333333333f, 0.042753039215686274f, 0.0729721568627451f,
-0.348097f, 0.043368176470588236f, 0.07543270588235294f,
-0.34040766666666666f, 0.04398331372549019f, 0.07789325490196078f,
-0.33271833333333334f, 0.044598450980392154f, 0.08035380392156863f,
-0.325029f, 0.045213588235294117f, 0.08281435294117646f,
-0.3173396666666667f, 0.04582872549019608f, 0.08527490196078431f,
-0.3096503333333333f, 0.046443862745098034f, 0.08773545098039215f,
-0.301961f, 0.047059f, 0.090196f
-}
-, 
-{
-0.941176f, 0.8f, 0.552941f,
-0.9396381960784314f, 0.7956939607843138f, 0.5470970784313726f,
-0.9381003921568628f, 0.7913879215686275f, 0.5412531568627451f,
-0.9365625882352941f, 0.7870818823529412f, 0.5354092352941177f,
-0.9350247843137255f, 0.7827758431372549f, 0.5295653137254902f,
-0.9334869803921568f, 0.7784698039215687f, 0.5237213921568628f,
-0.9319491764705883f, 0.7741637647058824f, 0.5178774705882353f,
-0.9304113725490196f, 0.7698577254901962f, 0.5120335490196078f,
-0.928873568627451f, 0.7655516862745099f, 0.5061896274509804f,
-0.9273357647058823f, 0.7612456470588236f, 0.5003457058823529f,
-0.9257979607843138f, 0.7569396078431373f, 0.4945017843137255f,
-0.9242601568627451f, 0.752633568627451f, 0.48865786274509804f,
-0.9227223529411764f, 0.7483275294117647f, 0.4828139411764706f,
-0.9211845294117647f, 0.743867705882353f, 0.47681623529411765f,
-0.9196466470588235f, 0.7389465294117648f, 0.4703571764705882f,
-0.9181087647058823f, 0.7340253529411765f, 0.4638981176470588f,
-0.9165708823529412f, 0.7291041764705882f, 0.4574390588235294f,
-0.915033f, 0.724183f, 0.45098f,
-0.9134951176470588f, 0.7192618235294118f, 0.44452094117647056f,
-0.9119572352941177f, 0.7143406470588235f, 0.4380618823529412f,
-0.9104193529411765f, 0.7094194705882353f, 0.43160282352941176f,
-0.9088814705882353f, 0.7044982941176471f, 0.42514376470588233f,
-0.9073435882352942f, 0.6995771176470589f, 0.41868470588235296f,
-0.905805705882353f, 0.6946559411764706f, 0.41222564705882353f,
-0.9042678235294118f, 0.6897347647058824f, 0.4057665882352941f,
-0.9027299411764707f, 0.6848135882352941f, 0.3993075294117647f,
-0.9011151764705883f, 0.6807382352941176f, 0.39454013725490195f,
-0.8994235294117647f, 0.6775087058823529f, 0.39146441176470587f,
-0.8977318823529412f, 0.6742791764705882f, 0.3883886862745098f,
-0.8960402352941177f, 0.6710496470588235f, 0.3853129607843137f,
-0.8943485882352942f, 0.6678201176470588f, 0.3822372352941176f,
-0.8926569411764707f, 0.6645905882352942f, 0.37916150980392155f,
-0.8909652941176471f, 0.6613610588235294f, 0.37608578431372547f,
-0.8892736470588236f, 0.6581315294117647f, 0.37301005882352944f,
-0.887582f, 0.654902f, 0.3699343333333333f,
-0.8858903529411765f, 0.6516724705882353f, 0.36685860784313723f,
-0.8841987058823529f, 0.6484429411764706f, 0.3637828823529412f,
-0.8825070588235294f, 0.6452134117647059f, 0.3607071568627451f,
-0.8808154117647059f, 0.6419838823529411f, 0.35763143137254905f,
-0.8791237647058824f, 0.6387543529411764f, 0.35455570588235297f,
-0.8774321176470589f, 0.6355248235294118f, 0.3514799803921569f,
-0.8757404705882353f, 0.6322952941176471f, 0.3484042549019608f,
-0.8740488235294118f, 0.6290657647058824f, 0.34532852941176473f,
-0.8723571764705883f, 0.6258362352941176f, 0.34225280392156865f,
-0.8706655294117648f, 0.6226067058823529f, 0.33917707843137257f,
-0.8689738823529413f, 0.6193771764705882f, 0.3361013529411765f,
-0.8672822352941176f, 0.6161476470588235f, 0.3330256274509804f,
-0.8655905882352941f, 0.6129181176470588f, 0.32994990196078433f,
-0.8638989411764706f, 0.609688588235294f, 0.32687417647058825f,
-0.8622072941176471f, 0.6064590588235295f, 0.32379845098039217f,
-0.8605156470588236f, 0.6032295294117647f, 0.3207227254901961f,
-0.858824f, 0.6f, 0.317647f,
-0.8572861176470589f, 0.5972318431372549f, 0.31610911764705885f,
-0.8557482352941177f, 0.5944636862745099f, 0.3145712352941177f,
-0.8542103529411765f, 0.5916955294117647f, 0.31303335294117646f,
-0.8526724705882354f, 0.5889273725490196f, 0.3114954705882353f,
-0.8511345882352942f, 0.5861592156862745f, 0.3099575882352941f,
-0.849596705882353f, 0.5833910588235294f, 0.30841970588235296f,
-0.8480588235294118f, 0.5806229019607843f, 0.3068818235294118f,
-0.8465209411764706f, 0.5778547450980392f, 0.30534394117647057f,
-0.8449830588235294f, 0.5750865882352941f, 0.3038060588235294f,
-0.8434451764705883f, 0.572318431372549f, 0.30226817647058823f,
-0.8419072941176471f, 0.5695502745098039f, 0.30073029411764707f,
-0.8403694117647059f, 0.5667821176470588f, 0.2991924117647059f,
-0.8388315294117648f, 0.5640139607843138f, 0.29765452941176473f,
-0.8372936470588236f, 0.5612458039215686f, 0.2961166470588235f,
-0.8357557647058824f, 0.5584776470588235f, 0.29457876470588235f,
-0.8342178823529411f, 0.5557094901960784f, 0.2930408823529412f,
-0.83268f, 0.5529413333333333f, 0.291503f,
-0.8311421176470588f, 0.5501731764705882f, 0.2899651176470588f,
-0.8296042352941176f, 0.5474050196078432f, 0.2884272352941176f,
-0.8280663529411765f, 0.544636862745098f, 0.28688935294117646f,
-0.8265284705882353f, 0.5418687058823529f, 0.2853514705882353f,
-0.8249905882352941f, 0.5391005490196078f, 0.2838135882352941f,
-0.823452705882353f, 0.5363323921568628f, 0.28227570588235296f,
-0.8219148235294118f, 0.5335642352941177f, 0.28073782352941173f,
-0.8203769411764706f, 0.5307960784313726f, 0.27919994117647057f,
-0.8192235294117647f, 0.5282585882352941f, 0.27750829411764705f,
-0.8184545882352942f, 0.5259517647058823f, 0.2756628823529412f,
-0.8176856470588235f, 0.5236449411764705f, 0.2738174705882353f,
-0.816916705882353f, 0.5213381176470588f, 0.2719720588235294f,
-0.8161477647058824f, 0.519031294117647f, 0.2701266470588235f,
-0.8153788235294118f, 0.5167244705882353f, 0.2682812352941177f,
-0.8146098823529412f, 0.5144176470588235f, 0.26643582352941175f,
-0.8138409411764707f, 0.5121108235294117f, 0.2645904117647059f,
-0.813072f, 0.509804f, 0.262745f,
-0.8123030588235294f, 0.5074971764705882f, 0.26089958823529413f,
-0.8115341176470589f, 0.5051903529411764f, 0.2590541764705882f,
-0.8107651764705882f, 0.5028835294117647f, 0.25720876470588233f,
-0.8099962352941177f, 0.5005767058823529f, 0.25536335294117646f,
-0.809227294117647f, 0.49826988235294123f, 0.2535179411764706f,
-0.8084583529411765f, 0.4959630588235294f, 0.2516725294117647f,
-0.8076894117647059f, 0.4936562352941177f, 0.2498271176470588f,
-0.8069204705882354f, 0.49134941176470587f, 0.24798170588235294f,
-0.8061515294117647f, 0.4890425882352941f, 0.24613629411764704f,
-0.8053825882352942f, 0.48673576470588237f, 0.24429088235294116f,
-0.8046136470588235f, 0.4844289411764706f, 0.2424454705882353f,
-0.803844705882353f, 0.4821221176470588f, 0.24060005882352942f,
-0.8030757647058824f, 0.4798152941176471f, 0.23875464705882354f,
-0.8023068235294118f, 0.4775084705882353f, 0.23690923529411764f,
-0.8015378823529412f, 0.47520164705882356f, 0.23506382352941177f,
-0.8007689411764707f, 0.47289482352941176f, 0.23321841176470587f,
-0.8f, 0.470588f, 0.231373f,
-0.7992310588235294f, 0.46843498039215686f, 0.23029649019607842f,
-0.7984621176470589f, 0.4662819607843137f, 0.22921998039215685f,
-0.7976931764705882f, 0.4641289411764706f, 0.22814347058823528f,
-0.7969242352941177f, 0.4619759215686275f, 0.22706696078431374f,
-0.7961552941176471f, 0.45982290196078435f, 0.22599045098039217f,
-0.7953863529411765f, 0.4576698823529412f, 0.2249139411764706f,
-0.7946174117647059f, 0.45551686274509806f, 0.22383743137254902f,
-0.7938484705882353f, 0.4533638431372549f, 0.22276092156862745f,
-0.7930795294117647f, 0.4512108235294118f, 0.22168441176470588f,
-0.7923105882352941f, 0.44905780392156863f, 0.2206079019607843f,
-0.7915416470588236f, 0.4469047843137255f, 0.21953139215686274f,
-0.7907727058823529f, 0.4447517647058824f, 0.2184548823529412f,
-0.7900037647058824f, 0.4425987450980392f, 0.2173783725490196f,
-0.7892348235294118f, 0.44044572549019606f, 0.21630186274509802f,
-0.7884658823529411f, 0.4382927058823529f, 0.21522535294117645f,
-0.7876969411764706f, 0.43613968627450983f, 0.2141488431372549f,
-0.786928f, 0.4339866666666667f, 0.21307233333333334f,
-0.7861590588235294f, 0.43183364705882354f, 0.21199582352941176f,
-0.7853901176470588f, 0.4296806274509804f, 0.2109193137254902f,
-0.7846211764705883f, 0.42752760784313726f, 0.20984280392156862f,
-0.7838522352941176f, 0.4253745882352941f, 0.20876629411764705f,
-0.783083294117647f, 0.423221568627451f, 0.20768978431372548f,
-0.7823143529411765f, 0.42106854901960783f, 0.2066132745098039f,
-0.7815454117647058f, 0.4189155294117647f, 0.20553676470588234f,
-0.7807764705882353f, 0.4167625098039216f, 0.2044602549019608f,
-0.7800075294117647f, 0.41460949019607846f, 0.20338374509803922f,
-0.7792385882352941f, 0.4124564705882353f, 0.20230723529411765f,
-0.7784696470588235f, 0.41030345098039217f, 0.20123072549019608f,
-0.777700705882353f, 0.408150431372549f, 0.2001542156862745f,
-0.7769317647058823f, 0.40599741176470594f, 0.19907770588235296f,
-0.7761628235294118f, 0.40384439215686274f, 0.19800119607843136f,
-0.7753938823529412f, 0.4016913725490196f, 0.1969246862745098f,
-0.7746249411764706f, 0.39953835294117646f, 0.19584817647058822f,
-0.773856f, 0.3973853333333333f, 0.19477166666666665f,
-0.7730870588235293f, 0.39523231372549017f, 0.19369515686274508f,
-0.7723181176470588f, 0.3930792941176471f, 0.19261864705882353f,
-0.7715491764705882f, 0.39092627450980394f, 0.19154213725490196f,
-0.7707802352941177f, 0.3887732549019608f, 0.1904656274509804f,
-0.770011294117647f, 0.38662023529411765f, 0.18938911764705882f,
-0.7692423529411765f, 0.3844672156862745f, 0.18831260784313725f,
-0.7684734117647058f, 0.38231419607843137f, 0.18723609803921568f,
-0.7677044705882353f, 0.3801611764705882f, 0.1861595882352941f,
-0.7669355294117647f, 0.3780081568627451f, 0.18508307843137253f,
-0.7661665882352942f, 0.37585513725490194f, 0.18400656862745096f,
-0.7653976470588235f, 0.37370211764705885f, 0.18293005882352942f,
-0.764628705882353f, 0.37154909803921576f, 0.18185354901960787f,
-0.7638597647058823f, 0.36939607843137257f, 0.18077703921568627f,
-0.7630908235294118f, 0.3672430588235294f, 0.1797005294117647f,
-0.7623218823529412f, 0.3650900392156863f, 0.17862401960784313f,
-0.7615529411764707f, 0.36293701960784314f, 0.17754750980392156f,
-0.760784f, 0.360784f, 0.176471f,
-0.7600150588235294f, 0.3592461568627451f, 0.1757020588235294f,
-0.7592461176470588f, 0.35770831372549017f, 0.17493311764705882f,
-0.7584771764705882f, 0.3561704705882353f, 0.17416417647058824f,
-0.7577082352941177f, 0.3546326274509804f, 0.17339523529411763f,
-0.756939294117647f, 0.3530947843137255f, 0.17262629411764704f,
-0.7561703529411765f, 0.35155694117647057f, 0.17185735294117646f,
-0.7554014117647059f, 0.35001909803921566f, 0.17108841176470588f,
-0.7546324705882352f, 0.3484812549019608f, 0.1703194705882353f,
-0.7538635294117647f, 0.3469434117647059f, 0.1695505294117647f,
-0.7530945882352941f, 0.345405568627451f, 0.1687815882352941f,
-0.7523256470588235f, 0.3438677254901961f, 0.16801264705882354f,
-0.7515567058823529f, 0.34232988235294115f, 0.16724370588235293f,
-0.7507877647058824f, 0.3407920392156863f, 0.16647476470588235f,
-0.7500188235294117f, 0.3392541960784314f, 0.16570582352941177f,
-0.7492498823529411f, 0.33771635294117647f, 0.16493688235294118f,
-0.7484809411764706f, 0.33617850980392155f, 0.1641679411764706f,
-0.7477119999999999f, 0.33464066666666664f, 0.163399f,
-0.7469430588235294f, 0.3331028235294118f, 0.1626300588235294f,
-0.7461741176470588f, 0.33156498039215687f, 0.16186111764705882f,
-0.7454051764705882f, 0.33002713725490196f, 0.16109217647058824f,
-0.7446362352941176f, 0.32848929411764705f, 0.16032323529411766f,
-0.743867294117647f, 0.32695145098039213f, 0.15955429411764704f,
-0.7430983529411764f, 0.3254136078431372f, 0.15878535294117646f,
-0.7423294117647058f, 0.32387576470588236f, 0.15801641176470588f,
-0.7415604705882353f, 0.32233792156862745f, 0.1572474705882353f,
-0.7407915490196078f, 0.3208000588235294f, 0.1564785294117647f,
-0.7400226470588235f, 0.3192621764705883f, 0.15570958823529413f,
-0.7392537450980392f, 0.3177242941176471f, 0.15494064705882354f,
-0.7384848431372548f, 0.31618641176470585f, 0.15417170588235293f,
-0.7377159411764705f, 0.3146485294117647f, 0.15340276470588235f,
-0.7369470392156863f, 0.3131106470588235f, 0.15263382352941177f,
-0.7361781372549019f, 0.31157276470588235f, 0.15186488235294118f,
-0.7354092352941176f, 0.3100348823529412f, 0.1510959411764706f,
-0.7346403333333333f, 0.308497f, 0.150327f,
-0.733871431372549f, 0.30695911764705885f, 0.1495580588235294f,
-0.7331025294117647f, 0.30542123529411763f, 0.14878911764705882f,
-0.7323336274509804f, 0.30388335294117647f, 0.14802017647058824f,
-0.731564725490196f, 0.3023454705882353f, 0.14725123529411765f,
-0.7307958235294117f, 0.30080758823529413f, 0.14648229411764704f,
-0.7300269215686275f, 0.29926970588235297f, 0.14571335294117646f,
-0.7292580196078431f, 0.2977318235294118f, 0.14494441176470588f,
-0.7284891176470588f, 0.2961939411764706f, 0.1441754705882353f,
-0.7277202156862745f, 0.29465605882352947f, 0.1434065294117647f,
-0.7269513137254902f, 0.29311817647058824f, 0.1426375882352941f,
-0.7261824117647059f, 0.2915802941176471f, 0.14186864705882352f,
-0.7254135098039216f, 0.2900424117647059f, 0.14109970588235293f,
-0.7246446078431372f, 0.28850452941176474f, 0.14033076470588235f,
-0.723875705882353f, 0.2869666470588236f, 0.13956182352941177f,
-0.7231068039215687f, 0.28542876470588235f, 0.13879288235294115f,
-0.7223379019607843f, 0.2838908823529412f, 0.13802394117647057f,
-0.721569f, 0.282353f, 0.137255f,
-0.7208000588235294f, 0.2811226862745098f, 0.1367936274509804f,
-0.7200311176470588f, 0.2798923725490196f, 0.13633225490196077f,
-0.7192621764705882f, 0.2786620588235294f, 0.13587088235294117f,
-0.7184932352941177f, 0.27743174509803925f, 0.13540950980392155f,
-0.717724294117647f, 0.27620143137254904f, 0.13494813725490196f,
-0.7169553529411765f, 0.27497111764705884f, 0.13448676470588233f,
-0.7161864117647059f, 0.27374080392156863f, 0.13402539215686274f,
-0.7154174705882353f, 0.2725104901960785f, 0.13356401960784314f,
-0.7146485294117647f, 0.2712801764705882f, 0.13310264705882352f,
-0.7138795882352941f, 0.27004986274509807f, 0.1326412745098039f,
-0.7131106470588235f, 0.26881954901960786f, 0.1321799019607843f,
-0.7123417058823529f, 0.26758923529411766f, 0.1317185294117647f,
-0.7115727647058824f, 0.26635892156862745f, 0.13125715686274508f,
-0.7108038235294117f, 0.26512860784313724f, 0.1307957843137255f,
-0.7100348823529411f, 0.26389829411764704f, 0.13033441176470587f,
-0.7092659411764706f, 0.2626679803921569f, 0.12987303921568627f,
-0.7084969999999999f, 0.2614376666666667f, 0.12941166666666665f,
-0.7077280588235294f, 0.2602073529411765f, 0.12895029411764705f,
-0.7069591176470588f, 0.25897703921568627f, 0.12848892156862743f,
-0.7061901764705882f, 0.25774672549019606f, 0.12802754901960783f,
-0.7054212352941176f, 0.25651641176470585f, 0.12756617647058824f,
-0.704652294117647f, 0.25528609803921565f, 0.12710480392156862f,
-0.7038833529411764f, 0.2540557843137255f, 0.12664343137254902f,
-0.7031144117647059f, 0.2528254705882353f, 0.12618205882352942f,
-0.7023454705882353f, 0.2515951568627451f, 0.1257206862745098f,
-0.7010382549019607f, 0.25005729411764704f, 0.1250286274509804f,
-0.6991927647058823f, 0.24821188235294117f, 0.12410588235294116f,
-0.6973472745098038f, 0.24636647058823527f, 0.12318313725490196f,
-0.6955017843137254f, 0.2445210588235294f, 0.12226039215686274f,
-0.6936562941176471f, 0.24267564705882352f, 0.12133764705882352f,
-0.6918108039215686f, 0.24083023529411762f, 0.1204149019607843f,
-0.6899653137254902f, 0.23898482352941175f, 0.1194921568627451f,
-0.6881198235294117f, 0.23713941176470588f, 0.11856941176470588f,
-0.6862743333333333f, 0.23529399999999998f, 0.11764666666666666f,
-0.6844288431372549f, 0.2334485882352941f, 0.11672392156862746f,
-0.6825833529411764f, 0.23160317647058823f, 0.11580117647058824f,
-0.6807378627450981f, 0.22975776470588233f, 0.11487843137254902f,
-0.6788923725490196f, 0.22791235294117645f, 0.1139556862745098f,
-0.6768162352941176f, 0.2256055294117647f, 0.11326364705882354f,
-0.6746632156862746f, 0.22314490196078438f, 0.11264850980392159f,
-0.6725101960784314f, 0.2206842745098039f, 0.11203337254901961f,
-0.6703571764705882f, 0.21822364705882352f, 0.11141823529411765f,
-0.6682041568627451f, 0.21576301960784314f, 0.11080309803921569f,
-0.666051137254902f, 0.21330239215686275f, 0.11018796078431373f,
-0.6638981176470589f, 0.21084176470588234f, 0.10957282352941176f,
-0.6617450980392157f, 0.20838113725490195f, 0.10895768627450982f,
-0.6595920784313726f, 0.20592050980392157f, 0.10834254901960785f,
-0.6574390588235294f, 0.20345988235294118f, 0.10772741176470589f,
-0.6552860392156863f, 0.20099925490196077f, 0.10711227450980393f,
-0.6531330196078431f, 0.1985386274509804f, 0.10649713725490197f,
-0.65098f, 0.196078f, 0.105882f
-}
-, 
-{
-0.301961f, 0.047059f, 0.090196f,
-0.33271833333333334f, 0.044598450980392154f, 0.08035380392156863f,
-0.36347566666666664f, 0.04213790196078431f, 0.07051160784313726f,
-0.394233f, 0.03967735294117647f, 0.06066941176470589f,
-0.4259900588235294f, 0.04321439215686274f, 0.05182668627450981f,
-0.45797782352941174f, 0.048135490196078434f, 0.04321460784313726f,
-0.4899655882352941f, 0.05305658823529412f, 0.034602529411764706f,
-0.5211843333333334f, 0.06797396078431373f, 0.029834960784313726f,
-0.5519416666666667f, 0.08888909803921569f, 0.027374098039215686f,
-0.582699f, 0.10980423529411765f, 0.02491323529411765f,
-0.6107650392156863f, 0.12964274509803922f, 0.021914333333333334f,
-0.6353708431372549f, 0.14809701960784313f, 0.018223666666666666f,
-0.6599766470588235f, 0.16655129411764705f, 0.014533f,
-0.6848900588235294f, 0.18500564705882352f, 0.010842254901960783f,
-0.7107262941176471f, 0.20346023529411766f, 0.007151274509803921f,
-0.7365625294117647f, 0.22191482352941178f, 0.003460294117647058f,
-0.7615529411764707f, 0.24052317647058824f, 0.0f,
-0.773856f, 0.261438f, 0.0f,
-0.7861590588235294f, 0.28235282352941177f, 0.0f,
-0.7984621176470589f, 0.30326764705882353f, 0.0f,
-0.8247596862745098f, 0.33494792156862746f, 0.005382588235294118f,
-0.8530564705882353f, 0.3681661176470588f, 0.011534117647058825f,
-0.8813532549019608f, 0.4013843137254902f, 0.01768564705882353f,
-0.8944251764705882f, 0.4278357843137255f, 0.022991254901960786f,
-0.900576705882353f, 0.4512114705882353f, 0.02791235294117647f,
-0.9067282352941176f, 0.4745871568627451f, 0.032833450980392156f,
-0.9116494901960783f, 0.49550213725490194f, 0.03836976470588235f,
-0.9153404705882352f, 0.5139564117647059f, 0.04452129411764706f,
-0.9190314509803921f, 0.5324106862745098f, 0.05067282352941176f,
-0.9223379215686274f, 0.5512494705882353f, 0.05874670588235294f,
-0.9247984705882353f, 0.5709341764705882f, 0.07104976470588235f,
-0.9272590196078432f, 0.5906188823529411f, 0.08335282352941176f,
-0.9298733333333333f, 0.6106111568627451f, 0.09627101960784314f,
-0.933564f, 0.6327564117647058f, 0.11349517647058811f,
-0.9372546666666667f, 0.6549016666666667f, 0.13071933333333335f,
-0.9409453333333333f, 0.6770469215686274f, 0.14794349019607844f,
-0.9434830588235295f, 0.6968854117647059f, 0.16978117647058824f,
-0.9459439215686274f, 0.7165701176470587f, 0.19192643137254886f,
-0.9484047843137254f, 0.7362548235294117f, 0.2140716862745098f,
-0.951788f, 0.7559397647058823f, 0.23529435294117645f,
-0.9554786666666666f, 0.7756247843137255f, 0.2562094901960784f,
-0.9591693333333333f, 0.7953098039215685f, 0.27712462745098027f,
-0.9621680588235294f, 0.8143025294117647f, 0.3042679411764706f,
-0.9646286078431372f, 0.8327568039215686f, 0.33625539215686273f,
-0.9670891568627451f, 0.8512110784313726f, 0.3682428431372549f,
-0.968627f, 0.8678199411764704f, 0.3997691176470586f,
-0.968627f, 0.8813531176470588f, 0.43052676470588236f,
-0.968627f, 0.894886294117647f, 0.4612844117647059f,
-0.9693190588235294f, 0.9081888235294118f, 0.49158064705882354f,
-0.9730100392156863f, 0.9204918823529411f, 0.5198774313725488f,
-0.9767010196078432f, 0.9327949411764705f, 0.5481742156862744f,
-0.980392f, 0.945098f, 0.576471f,
-0.980392f, 0.9524796470588235f, 0.6010768039215686f,
-0.980392f, 0.959861294117647f, 0.625682607843137f,
-0.980392f, 0.9672429411764706f, 0.6502884117647059f,
-0.9823911960784314f, 0.9736251176470588f, 0.6768934117647059f,
-0.9848517450980392f, 0.9797766470588235f, 0.7039597647058823f,
-0.987312294117647f, 0.9859281764705882f, 0.7310261176470586f,
-0.9874661764705882f, 0.988235f, 0.7611683529411765f,
-0.9862360588235294f, 0.988235f, 0.7931561176470588f,
-0.9850059411764706f, 0.988235f, 0.8251438823529412f,
-0.9865566879084967f, 0.9899170874183006f, 0.8638822205882349f,
-0.9916828316993465f, 0.9937618586601307f, 0.9112998676470587f,
-0.9968089754901961f, 0.9976066299019608f, 0.9587175147058823f,
-0.9052372294703552f, 0.9779564511584385f, 0.8243706029943159f,
-0.8867819030220032f, 0.9681138613764159f, 0.7960723730259015f,
-0.8683265765736515f, 0.9582712715943934f, 0.7677741430574876f,
-0.8496495152292368f, 0.9481330038699588f, 0.7391063674947f,
-0.827503374481646f, 0.9333688054588851f, 0.704656989519581f,
-0.8053572337340553f, 0.9186046070478114f, 0.670207611544462f,
-0.7832110929864646f, 0.9038404086367376f, 0.6357582335693429f,
-0.766433500561199f, 0.893371185783902f, 0.6093614727271444f,
-0.7504389784273724f, 0.8835284391328594f, 0.5841392873691418f,
-0.7344444562935458f, 0.8736856924818169f, 0.5589171020111392f,
-0.7184499341597195f, 0.8638429458307745f, 0.533694916653137f,
-0.7024554120258928f, 0.8540001991797318f, 0.508472731295134f,
-0.6864608898920662f, 0.8441574525286893f, 0.48325054593713146f,
-0.6726107669334224f, 0.8337019757621676f, 0.46262336775823965f,
-0.6609226131327606f, 0.8226287485193524f, 0.44662884562441313f,
-0.6492344593320987f, 0.8115555212765373f, 0.4306343234905866f,
-0.6375463055314369f, 0.8004822940337221f, 0.4146398013567601f,
-0.6258581517307751f, 0.7894090667909068f, 0.3986452792229336f,
-0.6141699979301134f, 0.7783358395480918f, 0.3826507570891073f,
-0.6028608203088908f, 0.767717433975006f, 0.3670352497919524f,
-0.5942483973806011f, 0.760335491638489f, 0.3541167722685378f,
-0.5856359744523114f, 0.7529535493019721f, 0.3411982947451232f,
-0.5770235515240217f, 0.7455716069654551f, 0.32827981722170857f,
-0.5684111285957321f, 0.7381896646289382f, 0.31536133969829394f,
-0.5597987056674425f, 0.7308077222924212f, 0.3024428621748793f,
-0.5511862827391528f, 0.7234257799559043f, 0.2895243846514647f,
-0.5439550917001388f, 0.71558323130337f, 0.2812100917001388f,
-0.5371880759214683f, 0.7075858918019468f, 0.2744430759214683f,
-0.530421060142798f, 0.6995885523005235f, 0.26767606014279793f,
-0.5236540443641275f, 0.6915912127991003f, 0.2609090443641276f,
-0.5168870285854572f, 0.6835938732976771f, 0.2541420285854572f,
-0.5101200128067869f, 0.6755965337962538f, 0.24737501280678684f,
-0.5026633222411251f, 0.6664497167745533f, 0.2401482529207051f,
-0.49405089931283536f, 0.655376646400758f, 0.2321510702883018f,
-0.4854384763845458f, 0.6443035760269629f, 0.22415388765589855f,
-0.476826053456256f, 0.6332305056531674f, 0.21615670502349513f,
-0.46821363052796633f, 0.6221574352793721f, 0.2081595223910918f,
-0.4596012075996766f, 0.6110843649055767f, 0.20016233975868847f,
-0.4509887846713869f, 0.6000112945317814f, 0.19216515712628512f,
-0.44237636174309725f, 0.588938224157986f, 0.1853970432864374f,
-0.43376393881480757f, 0.5778651537841907f, 0.1786301843767869f,
-0.42515151588651784f, 0.5667920834103953f, 0.17186332546713637f,
-0.4165390929582283f, 0.5557190130366001f, 0.16509646655748597f,
-0.4079266700299385f, 0.5446459426628045f, 0.15832960764783535f,
-0.39931424710164876f, 0.5335728722890092f, 0.15156274873818484f,
-0.3907018241733591f, 0.5217316928630087f, 0.14517984643150433f,
-0.3820894012450694f, 0.5094281418974407f, 0.13902807094872033f,
-0.37347697831677973f, 0.4971245909318726f, 0.13287629546593632f,
-0.36486455538849005f, 0.48482103996630455f, 0.1267245199831523f,
-0.3562521324602004f, 0.47251748900073653f, 0.12057274450036828f,
-0.3476397095319108f, 0.4602139380351687f, 0.11442096901758436f,
-0.3390273257408728f, 0.4477569299054316f, 0.10842268983622093f,
-0.33041505968160295f, 0.434838295512997f, 0.10288615464932324f,
-0.3218027936223331f, 0.4219196611205625f, 0.09734961946242557f,
-0.3131905275630633f, 0.40900102672812794f, 0.0918130842755279f,
-0.30457826150379347f, 0.3960823923356934f, 0.08627654908863021f,
-0.2959659954445236f, 0.38316375794325885f, 0.08074001390173252f,
-0.28735372938525383f, 0.37024512355082434f, 0.07520347871483485f,
-0.2771269282081088f, 0.35625026963353057f, 0.07020498468017158f,
-0.26666894126117985f, 0.342101311518343f, 0.06528353292014033f,
-0.25621095431425095f, 0.3279523534031557f, 0.06036208116010914f,
-0.24575296736732213f, 0.31380339528796836f, 0.05544062940007796f,
-0.23529498042039326f, 0.299654437172781f, 0.05051917764004678f,
-0.22483699347346442f, 0.28550547905759366f, 0.045597725880015594f,
-0.3816098233847023f, 0.39398373507705337f, 0.4030446360993551f,
-0.3840705692600429f, 0.39767485389006435f, 0.4091966576630016f,
-0.3865313151353836f, 0.4013659727030754f, 0.41534867922664814f,
-0.38899206101072425f, 0.4050570915160864f, 0.4215007007902947f,
-0.3914528068860649f, 0.4087482103290973f, 0.4276527223539411f,
-0.3939135527614056f, 0.4124393291421084f, 0.4338047439175878f,
-0.3963743175261149f, 0.4162045320589903f, 0.4399567465918657f,
-0.39883522027675045f, 0.42051091577848393f, 0.44610861128021734f,
-0.401296123027386f, 0.42481729949797753f, 0.452260475968569f,
-0.4037570257780215f, 0.42912368321747113f, 0.4584123406569207f,
-0.40621792852865707f, 0.43343006693696473f, 0.4645642053452724f,
-0.4086788312792926f, 0.43773645065645833f, 0.47071607003362403f,
-0.41113973402992815f, 0.44204283437595193f, 0.4768679347219757f,
-0.4145182796461964f, 0.4468081565446947f, 0.4834788548760093f,
-0.41820939845920746f, 0.4517298051706709f, 0.4902461413461384f,
-0.4219005172722185f, 0.4566514537966471f, 0.4970134278162676f,
-0.4255916360852295f, 0.46157310242262334f, 0.5037807142863968f,
-0.4292827548982405f, 0.4664947510485996f, 0.510548000756526f,
-0.43297387371125157f, 0.4714163996745758f, 0.5173152872266551f,
-0.43803604331672713f, 0.4774805129366699f, 0.5249965104447574f,
-0.4454184378180438f, 0.4854780154691743f, 0.534224385914932f,
-0.4528008323193609f, 0.49347551800167916f, 0.5434522613851074f,
-0.4601832268206778f, 0.5014730205341839f, 0.5526801368552824f,
-0.46756562132199475f, 0.5094705230666885f, 0.5619080123254574f,
-0.47494801582331164f, 0.5174680255991931f, 0.5711358877956323f,
-0.48233041032462853f, 0.5254655281316978f, 0.5803637632658073f,
-0.49155263873598226f, 0.5371430112748081f, 0.5926583935083312f,
-0.5007805142061572f, 0.5488317894956187f, 0.6049624366356243f,
-0.5100083896763322f, 0.5605205677164292f, 0.6172664797629175f,
-0.5192362651465072f, 0.5722093459372397f, 0.6295705228902105f,
-0.5284641406166822f, 0.5838981241580503f, 0.6418745660175037f,
-0.5376920160868572f, 0.5955869023788608f, 0.6541786091447969f,
-0.5469199892040675f, 0.6068927089272528f, 0.666865428650438f,
-0.5561480215495374f, 0.6179662222415807f, 0.679784422933624f,
-0.5653760538950072f, 0.6290397355559086f, 0.6927034172168101f,
-0.5746040862404771f, 0.6401132488702366f, 0.705622411499996f,
-0.5838321185859466f, 0.6511867621845642f, 0.7185414057831816f,
-0.5930601509314167f, 0.6622602754988924f, 0.731460400066368f,
-0.6027457888132203f, 0.6737914332483586f, 0.7442269110358488f,
-0.6138193021275482f, 0.6867105844068394f, 0.7565309541631419f,
-0.624892815441876f, 0.6996297355653202f, 0.768834997290435f,
-0.635966328756204f, 0.712548886723801f, 0.7811390404177281f,
-0.6470398420705319f, 0.7254680378822819f, 0.7934430835450211f,
-0.6581133553848597f, 0.7383871890407626f, 0.8057471266723143f,
-0.6691868686991876f, 0.7513063401992435f, 0.8180511697996073f,
-0.6797231035826917f, 0.7626133820135156f, 0.8292803820135156f,
-0.6901815088658318f, 0.7736868953278435f, 0.8403538953278435f,
-0.7006399141489721f, 0.7847604086421713f, 0.8514274086421714f,
-0.7110983194321123f, 0.7958339219564993f, 0.8625009219564993f,
-0.7215567247152525f, 0.8069074352708272f, 0.8735744352708272f,
-0.7320151299983927f, 0.8179809485851552f, 0.8846479485851553f,
-0.7433943934130215f, 0.829054540177121f, 0.8948006037679945f,
-0.7556982796650195f, 0.8401282103667435f, 0.9040286361134641f,
-0.7680021659170181f, 0.8512018805563666f, 0.9132566684589342f,
-0.7803060521690164f, 0.8622755507459894f, 0.922484700804404f,
-0.7926099384210146f, 0.8733492209356122f, 0.9317127331498739f,
-0.8049138246730129f, 0.884422891125235f, 0.9409407654953438f,
-0.8182899339607481f, 0.8958028800645715f, 0.9502453531164833f,
-0.8392067445270285f, 0.9093371392542401f, 0.960088336617846f,
-0.8601235550933087f, 0.9228713984439086f, 0.9699313201192087f,
-0.8810403656595891f, 0.9364056576335772f, 0.9797743036205714f,
-0.9169484148295987f, 0.955705868305041f, 0.9861579514817953f,
-0.9538602304608881f, 0.9753921490583561f, 0.9923099730454419f,
-0.9907720460921776f, 0.9950784298116713f, 0.9984619946090884f,
-0.9744873092420462f, 0.9744873092420462f, 0.9270234143318583f,
-0.9560315583016963f, 0.9560315583016963f, 0.8974938363265909f,
-0.9375758073613464f, 0.9375758073613464f, 0.8679642583213234f,
-0.9170615977193497f, 0.9170615977193497f, 0.8404931390177027f,
-0.8949145710906945f, 0.8949145710906945f, 0.8146548367007419f,
-0.8727675444620381f, 0.8727675444620381f, 0.7888165343837795f,
-0.8519876873835621f, 0.8510762022779806f, 0.7648012022779807f,
-0.8335316226926223f, 0.8301593917117002f, 0.7438843917117002f,
-0.8150755580016825f, 0.8092425811454197f, 0.7229675811454197f,
-0.7970702735105003f, 0.7890018834106215f, 0.7027268834106215f,
-0.7810752684454911f, 0.7717763485326472f, 0.6855013485326472f,
-0.7650802633804821f, 0.7545508136546729f, 0.6682758136546729f,
-0.749085258315473f, 0.7373252787766986f, 0.6510502787766986f,
-0.7343154501316902f, 0.7213252532504636f, 0.6362757626022031f,
-0.7195506611290563f, 0.7053302481854542f, 0.6215112873501589f,
-0.7047858721264224f, 0.6893352431204448f, 0.6067468120981148f,
-0.6890259843331534f, 0.6723451392648004f, 0.5909869843331534f,
-0.6730306655175545f, 0.655119604386826f, 0.5749916655175545f,
-0.6570353467019555f, 0.6378940695088515f, 0.5589963467019555f,
-0.641040222860408f, 0.6206685346308771f, 0.5437657161158874f,
-0.6250452177953992f, 0.6034429997529032f, 0.5290009271132541f,
-0.6090502127303895f, 0.5862174648749283f, 0.51423613811062f,
-0.5930552076653802f, 0.5689919299969538f, 0.4989372076653803f,
-0.5770602026003708f, 0.5517663951189792f, 0.4829422026003709f,
-0.5610651975353614f, 0.5345408602410047f, 0.4669471975353615f,
-0.5459816388371387f, 0.5179229821000844f, 0.45216750594446076f,
-0.5336775957098456f, 0.5031581930974507f, 0.4410939926301329f,
-0.5213735525825526f, 0.4883934040948169f, 0.430020479315805f,
-0.5090695094552595f, 0.47370207163840083f, 0.4188735094552594f,
-0.49676546632796625f, 0.46016781244873217f, 0.40656946632796626f,
-0.4844614232006731f, 0.4466335532590635f, 0.39426542320067304f,
-0.4721573800733799f, 0.4330992940693948f, 0.3819613800733799f,
-0.46200048854224385f, 0.4217116390124473f, 0.3728777906086043f,
-0.45215750504088115f, 0.41063781194752963f, 0.3642650231696171f,
-0.44231452153951845f, 0.399563984882612f, 0.3556522557306299f,
-0.43331448829164265f, 0.38933353803815574f, 0.34703948829164266f,
-0.42470172085265556f, 0.3794905545367932f, 0.33842672085265557f,
-0.41608895341366803f, 0.36964757103543017f, 0.32981395341366804f,
-0.40686343129881103f, 0.35919183285819767f, 0.32120118597468084f,
-0.3970201340468586f, 0.3481183195438698f, 0.3125884185356936f,
-0.38717683679490617f, 0.3370448062295419f, 0.3039756510967064f,
-0.37733363702997946f, 0.32635353954295376f, 0.2953628836577192f,
-0.36749065352861676f, 0.31651024229100133f, 0.286750116218732f,
-0.3576476700272541f, 0.3066669450390489f, 0.2781373487797447f,
-0.3479565813407575f, 0.2968236865258914f, 0.2695245813407575f,
-0.3393438139017703f, 0.2869807030245287f, 0.2609118139017703f,
-0.3307310464627831f, 0.27713771952316607f, 0.2522990464627831f,
-0.32211827902379586f, 0.26729473602180337f, 0.2436862790237959f,
-0.3123534587792868f, 0.25745145877928677f, 0.2350735115848086f,
-0.30251016152733423f, 0.24760816152733423f, 0.22646074414582132f,
-0.29266686427538174f, 0.23776486427538174f, 0.21784797670683403f,
-0.28282380201635243f, 0.2279218020163524f, 0.2092352092678468f,
-0.27298081851499f, 0.21807881851499f, 0.20062244182885983f,
-0.26313783501362703f, 0.20823583501362705f, 0.19200967438987238f,
-0.25398590695088513f, 0.19839267526757198f, 0.18408813863419826f,
-0.24537313951189782f, 0.1885493780156195f, 0.17670590100817618f,
-0.23676037207291054f, 0.17870608076366698f, 0.1693236633821541f,
-0.2276867835117145f, 0.1693236046339233f, 0.16148060463392327f,
-0.2178434862597621f, 0.16071083719493606f, 0.15286783719493605f,
-0.20800018900780967f, 0.15209806975594886f, 0.14425506975594884f,
-0.19884806469093963f, 0.14394606469093962f, 0.136333475252044f,
-0.1926960431272931f, 0.1377940431272931f, 0.13141198350136266f,
-0.18654402156364655f, 0.13164202156364654f, 0.12649049175068133f,
-0.180392f, 0.12549f, 0.121569f
-}
-, 
-{
-0.960784f, 0.964706f, 1.0f,
-0.9524795294117647f, 0.9567090588235294f, 0.9953863529411765f,
-0.9441750588235294f, 0.9487121176470588f, 0.9907727058823529f,
-0.9358705882352941f, 0.9407151764705882f, 0.9861590588235294f,
-0.9275661176470589f, 0.9327182352941176f, 0.9815454117647059f,
-0.9192616470588235f, 0.924721294117647f, 0.9769317647058824f,
-0.9109571764705883f, 0.9167243529411765f, 0.9723181176470588f,
-0.9026527058823529f, 0.9087274117647058f, 0.9677044705882353f,
-0.8943482352941177f, 0.9007304705882353f, 0.9630908235294118f,
-0.8860437647058824f, 0.8927335294117646f, 0.9584771764705883f,
-0.8777392941176471f, 0.8847365882352941f, 0.9538635294117647f,
-0.8694348235294118f, 0.8767396470588235f, 0.9492498823529412f,
-0.8611303529411765f, 0.8687427058823529f, 0.9446362352941177f,
-0.853671705882353f, 0.8613609215686274f, 0.9401763921568628f,
-0.8487505294117648f, 0.8558246078431372f, 0.9361779607843137f,
-0.8438293529411766f, 0.8502882941176471f, 0.9321795294117647f,
-0.8389081764705882f, 0.8447519803921568f, 0.9281810980392157f,
-0.833987f, 0.8392156666666667f, 0.9241826666666667f,
-0.8290658235294118f, 0.8336793529411765f, 0.9201842352941176f,
-0.8241446470588235f, 0.8281430392156862f, 0.9161858039215687f,
-0.8192234705882353f, 0.8226067254901961f, 0.9121873725490196f,
-0.8143022941176471f, 0.8170704117647059f, 0.9081889411764705f,
-0.8093811176470589f, 0.8115340980392157f, 0.9041905098039216f,
-0.8044599411764706f, 0.8059977843137255f, 0.9001920784313725f,
-0.7995387647058824f, 0.8004614705882352f, 0.8961936470588235f,
-0.7946175882352942f, 0.7949251568627451f, 0.8921952156862745f,
-0.7912342745098039f, 0.7906191176470588f, 0.8888119215686274f,
-0.7893888235294118f, 0.7875433529411765f, 0.8860437647058823f,
-0.7875433725490196f, 0.7844675882352942f, 0.8832756078431372f,
-0.7856979215686275f, 0.7813918235294117f, 0.8805074509803922f,
-0.7838524705882353f, 0.7783160588235294f, 0.8777392941176471f,
-0.7820070196078431f, 0.775240294117647f, 0.8749711372549019f,
-0.780161568627451f, 0.7721645294117647f, 0.8722029803921568f,
-0.7783161176470589f, 0.7690887647058824f, 0.8694348235294118f,
-0.7764706666666666f, 0.766013f, 0.8666666666666667f,
-0.7746252156862745f, 0.7629372352941177f, 0.8638985098039216f,
-0.7727797647058824f, 0.7598614705882353f, 0.8611303529411765f,
-0.7709343137254903f, 0.7567857058823529f, 0.8583621960784313f,
-0.769088862745098f, 0.7537099411764706f, 0.8555940392156862f,
-0.7672434117647059f, 0.7506341764705883f, 0.8528258823529412f,
-0.7653979607843138f, 0.7475584117647059f, 0.8500577254901961f,
-0.7635525098039216f, 0.7444826470588236f, 0.847289568627451f,
-0.7617070588235294f, 0.7414068823529412f, 0.8445214117647059f,
-0.7598616078431373f, 0.7383311176470588f, 0.8417532549019607f,
-0.7580161568627451f, 0.7352553529411765f, 0.8389850980392157f,
-0.756170705882353f, 0.7321795882352942f, 0.8362169411764706f,
-0.7543252549019608f, 0.7291038235294118f, 0.8334487843137255f,
-0.7524798039215687f, 0.7260280588235295f, 0.8306806274509804f,
-0.7506343529411765f, 0.7229522941176472f, 0.8279124705882352f,
-0.7487889019607844f, 0.7198765294117648f, 0.8251443137254902f,
-0.7469434509803922f, 0.7168007647058824f, 0.8223761568627451f,
-0.745098f, 0.713725f, 0.819608f,
-0.7410995294117647f, 0.710034137254902f, 0.8157633333333333f,
-0.7371010588235295f, 0.706343274509804f, 0.8119186666666667f,
-0.7331025882352942f, 0.7026524117647059f, 0.808074f,
-0.7291041176470588f, 0.6989615490196079f, 0.8042293333333334f,
-0.7251056470588235f, 0.6952706862745098f, 0.8003846666666666f,
-0.7211071764705883f, 0.6915798235294118f, 0.79654f,
-0.717108705882353f, 0.6878889607843137f, 0.7926953333333333f,
-0.7131102352941177f, 0.6841980980392157f, 0.7888506666666667f,
-0.7091117647058823f, 0.6805072352941177f, 0.785006f,
-0.7051132941176471f, 0.6768163725490197f, 0.7811613333333334f,
-0.7011148235294118f, 0.6731255098039216f, 0.7773166666666667f,
-0.6971163529411765f, 0.6694346470588236f, 0.773472f,
-0.6931178823529411f, 0.6657437843137255f, 0.7696273333333333f,
-0.6891194117647059f, 0.6620529215686275f, 0.7657826666666667f,
-0.6851209411764706f, 0.6583620588235295f, 0.7619380000000001f,
-0.6811224705882353f, 0.6546711960784314f, 0.7580933333333334f,
-0.677124f, 0.6509803333333334f, 0.7542486666666667f,
-0.6731255294117646f, 0.6472894705882354f, 0.750404f,
-0.6691270588235294f, 0.6435986078431373f, 0.7465593333333334f,
-0.6651285882352941f, 0.6399077450980393f, 0.7427146666666666f,
-0.6611301176470588f, 0.6362168823529413f, 0.73887f,
-0.6571316470588235f, 0.6325260196078432f, 0.7350253333333333f,
-0.6531331764705882f, 0.6288351568627452f, 0.7311806666666667f,
-0.6491347058823529f, 0.625144294117647f, 0.727336f,
-0.6451362352941176f, 0.621453431372549f, 0.7234913333333334f,
-0.6414453529411764f, 0.6177625490196079f, 0.7195697647058824f,
-0.6380620588235294f, 0.6140716470588236f, 0.7155712941176471f,
-0.6346787647058824f, 0.6103807450980393f, 0.7115728235294118f,
-0.6312954705882352f, 0.6066898431372549f, 0.7075743529411765f,
-0.6279121764705882f, 0.6029989411764707f, 0.7035758823529412f,
-0.6245288823529412f, 0.5993080392156863f, 0.699577411764706f,
-0.6211455882352941f, 0.595617137254902f, 0.6955789411764706f,
-0.617762294117647f, 0.5919262352941177f, 0.6915804705882354f,
-0.614379f, 0.5882353333333333f, 0.687582f,
-0.610995705882353f, 0.5845444313725491f, 0.6835835294117647f,
-0.6076124117647058f, 0.5808535294117647f, 0.6795850588235295f,
-0.6042291176470588f, 0.5771626274509805f, 0.6755865882352942f,
-0.6008458235294117f, 0.5734717254901961f, 0.6715881176470588f,
-0.5974625294117647f, 0.5697808235294118f, 0.6675896470588236f,
-0.5940792352941177f, 0.5660899215686275f, 0.6635911764705883f,
-0.5906959411764706f, 0.5623990196078432f, 0.659592705882353f,
-0.5873126470588235f, 0.5587081176470589f, 0.6555942352941176f,
-0.5839293529411764f, 0.5550172156862745f, 0.6515957647058824f,
-0.5805460588235294f, 0.5513263137254902f, 0.6475972941176471f,
-0.5771627647058823f, 0.5476354117647059f, 0.6435988235294118f,
-0.5737794705882353f, 0.5439445098039216f, 0.6396003529411765f,
-0.5703961764705883f, 0.5402536078431374f, 0.6356018823529412f,
-0.5670128823529412f, 0.536562705882353f, 0.6316034117647059f,
-0.5636295882352941f, 0.5328718039215686f, 0.6276049411764706f,
-0.560246294117647f, 0.5291809019607843f, 0.6236064705882354f,
-0.556863f, 0.52549f, 0.619608f,
-0.5548637450980393f, 0.5233369803921569f, 0.6173011764705882f,
-0.5528644901960784f, 0.5211839607843137f, 0.6149943529411765f,
-0.5508652352941177f, 0.5190309411764706f, 0.6126875294117647f,
-0.548865980392157f, 0.5168779215686274f, 0.610380705882353f,
-0.5468667254901961f, 0.5147249019607844f, 0.6080738823529412f,
-0.5448674705882353f, 0.5125718823529412f, 0.6057670588235294f,
-0.5428682156862745f, 0.5104188627450981f, 0.6034602352941176f,
-0.5408689607843138f, 0.5082658431372549f, 0.6011534117647059f,
-0.538869705882353f, 0.5061128235294118f, 0.5988465882352941f,
-0.5368704509803922f, 0.5039598039215687f, 0.5965397647058823f,
-0.5348711960784314f, 0.5018067843137255f, 0.5942329411764706f,
-0.5328719411764706f, 0.4996537647058824f, 0.5919261176470588f,
-0.5308726862745098f, 0.4975007450980392f, 0.589619294117647f,
-0.528873431372549f, 0.49534772549019607f, 0.5873124705882353f,
-0.5268741764705883f, 0.4931947058823529f, 0.5850056470588235f,
-0.5248749215686275f, 0.49104168627450984f, 0.5826988235294117f,
-0.5228756666666667f, 0.4888886666666667f, 0.580392f,
-0.5208764117647059f, 0.48673564705882355f, 0.5780851764705882f,
-0.5188771568627452f, 0.4845826274509804f, 0.5757783529411764f,
-0.5168779019607844f, 0.48242960784313726f, 0.5734715294117647f,
-0.5148786470588236f, 0.4802765882352941f, 0.5711647058823529f,
-0.5128793921568628f, 0.478123568627451f, 0.5688578823529411f,
-0.510880137254902f, 0.47597054901960784f, 0.5665510588235294f,
-0.5088808823529413f, 0.4738175294117647f, 0.5642442352941176f,
-0.5068816274509804f, 0.4716645098039216f, 0.5619374117647058f,
-0.5049592941176471f, 0.46951149019607846f, 0.5596306078431372f,
-0.5031138823529412f, 0.4673584705882353f, 0.5573238235294117f,
-0.5012684705882353f, 0.4652054509803922f, 0.5550170392156862f,
-0.49942305882352944f, 0.46305243137254903f, 0.5527102549019607f,
-0.4975776470588236f, 0.46089941176470595f, 0.5504034705882354f,
-0.4957322352941177f, 0.45874639215686275f, 0.5480966862745098f,
-0.4938868235294118f, 0.4565933725490196f, 0.5457899019607843f,
-0.49204141176470595f, 0.45444035294117646f, 0.5434831176470588f,
-0.490196f, 0.4522873333333333f, 0.5411763333333333f,
-0.48835058823529415f, 0.4501343137254902f, 0.5388695490196078f,
-0.4865051764705883f, 0.4479812941176471f, 0.5365627647058823f,
-0.4846597647058824f, 0.44582827450980395f, 0.5342559803921568f,
-0.48281435294117653f, 0.4436752549019608f, 0.5319491960784314f,
-0.4809689411764706f, 0.44152223529411766f, 0.5296424117647058f,
-0.4791235294117647f, 0.4393692156862745f, 0.5273356274509804f,
-0.47727811764705885f, 0.4372161960784314f, 0.5250288431372548f,
-0.475432705882353f, 0.43506317647058823f, 0.5227220588235294f,
-0.4735872941176471f, 0.4329101568627451f, 0.5204152745098038f,
-0.4717418823529412f, 0.43075713725490195f, 0.5181084901960784f,
-0.4698964705882353f, 0.42860411764705886f, 0.515801705882353f,
-0.4680510588235295f, 0.42645109803921577f, 0.5134949215686275f,
-0.46620564705882356f, 0.42429807843137257f, 0.511188137254902f,
-0.4643602352941177f, 0.42214505882352943f, 0.5088813529411764f,
-0.46251482352941176f, 0.4199920392156863f, 0.506574568627451f,
-0.4606694117647059f, 0.41783901960784314f, 0.5042677843137254f,
-0.458824f, 0.415686f, 0.501961f,
-0.4571323137254902f, 0.4138405490196078f, 0.4999617450980392f,
-0.4554406274509804f, 0.4119950980392157f, 0.4979624901960784f,
-0.4537489411764706f, 0.4101496470588235f, 0.4959632352941176f,
-0.4520572549019608f, 0.4083041960784314f, 0.49396398039215683f,
-0.450365568627451f, 0.4064587450980392f, 0.4919647254901961f,
-0.4486738823529412f, 0.40461329411764707f, 0.4899654705882353f,
-0.44698219607843137f, 0.4027678431372549f, 0.4879662156862745f,
-0.4452905098039216f, 0.40092239215686276f, 0.48596696078431373f,
-0.44359882352941177f, 0.3990769411764706f, 0.48396770588235294f,
-0.44190713725490194f, 0.3972314901960784f, 0.48196845098039215f,
-0.4402154509803922f, 0.3953860392156863f, 0.4799691960784314f,
-0.43852376470588234f, 0.3935405882352941f, 0.47796994117647057f,
-0.43683207843137256f, 0.39169513725490196f, 0.4759706862745098f,
-0.43514039215686273f, 0.3898496862745098f, 0.473971431372549f,
-0.43344870588235296f, 0.38800423529411765f, 0.4719721764705882f,
-0.43175701960784313f, 0.38615878431372547f, 0.46997292156862747f,
-0.43006533333333336f, 0.38431333333333334f, 0.4679736666666667f,
-0.4283736470588235f, 0.38246788235294116f, 0.4659744117647059f,
-0.42668196078431375f, 0.38062243137254903f, 0.4639751568627451f,
-0.4249902745098039f, 0.37877698039215685f, 0.4619759019607843f,
-0.4232985882352941f, 0.37693152941176467f, 0.4599766470588235f,
-0.4216069019607843f, 0.37508607843137254f, 0.45797739215686273f,
-0.4199152156862745f, 0.37324062745098036f, 0.45597813725490194f,
-0.4182235294117647f, 0.37139517647058823f, 0.45397888235294115f,
-0.4165318431372549f, 0.36954972549019605f, 0.4519796274509804f,
-0.41491707843137254f, 0.3677811764705882f, 0.44998039215686275f,
-0.4133792352941177f, 0.36608952941176476f, 0.44798117647058827f,
-0.4118413921568627f, 0.3643978823529412f, 0.44598196078431374f,
-0.41030354901960786f, 0.3627062352941176f, 0.4439827450980392f,
-0.40876570588235295f, 0.3610145882352941f, 0.4419835294117647f,
-0.40722786274509803f, 0.35932294117647057f, 0.4399843137254902f,
-0.4056900196078431f, 0.35763129411764705f, 0.4379850980392157f,
-0.4041521764705882f, 0.3559396470588235f, 0.43598588235294117f,
-0.40261433333333335f, 0.354248f, 0.4339866666666667f,
-0.40107649019607844f, 0.3525563529411765f, 0.43198745098039215f,
-0.3995386470588235f, 0.3508647058823529f, 0.42998823529411767f,
-0.3980008039215686f, 0.3491730588235294f, 0.42798901960784314f,
-0.3964629607843137f, 0.34748141176470587f, 0.42598980392156865f,
-0.39492511764705884f, 0.34578976470588235f, 0.4239905882352941f,
-0.39338727450980393f, 0.34409811764705883f, 0.42199137254901964f,
-0.391849431372549f, 0.3424064705882353f, 0.4199921568627451f,
-0.3903115882352941f, 0.34071482352941174f, 0.41799294117647057f,
-0.38877374509803925f, 0.33902317647058827f, 0.41599372549019614f,
-0.38723590196078433f, 0.3373315294117647f, 0.4139945098039216f,
-0.3856980588235294f, 0.3356398823529412f, 0.41199529411764707f,
-0.3841602156862745f, 0.33394823529411766f, 0.40999607843137253f,
-0.3826223725490196f, 0.33225658823529414f, 0.40799686274509805f,
-0.3810845294117647f, 0.3305649411764706f, 0.4059976470588236f,
-0.3795466862745098f, 0.32887329411764704f, 0.40399843137254904f,
-0.3780088431372549f, 0.3271816470588235f, 0.4019992156862745f,
-0.376471f, 0.32549f, 0.4f,
-0.37477931372549017f, 0.3236445490196078f, 0.3980007843137255f,
-0.3730876274509804f, 0.3217990980392157f, 0.396001568627451f,
-0.37139594117647057f, 0.3199536470588235f, 0.39400235294117647f,
-0.3697042549019608f, 0.3181081960784314f, 0.392003137254902f,
-0.36801256862745096f, 0.3162627450980392f, 0.39000392156862745f,
-0.3663208823529412f, 0.3144172941176471f, 0.388004705882353f,
-0.36462919607843136f, 0.3125718431372549f, 0.38600549019607844f,
-0.3629375098039216f, 0.3107263921568628f, 0.384006274509804f,
-0.36124582352941176f, 0.3088809411764706f, 0.3820070588235294f,
-0.35955413725490193f, 0.30703549019607845f, 0.3800078431372549f,
-0.35786245098039215f, 0.30519003921568627f, 0.3780086274509804f,
-0.3561707647058823f, 0.3033445882352941f, 0.37600941176470587f,
-0.35447907843137255f, 0.30149913725490196f, 0.3740101960784314f,
-0.3527873921568627f, 0.2996536862745098f, 0.37201098039215685f,
-0.35109570588235295f, 0.29780823529411765f, 0.3700117647058824f,
-0.3494040196078431f, 0.29596278431372547f, 0.36801254901960784f,
-0.34771233333333335f, 0.29411733333333334f, 0.36601333333333336f,
-0.3460206470588235f, 0.29227188235294116f, 0.3640141176470588f,
-0.34432896078431374f, 0.290426431372549f, 0.36201490196078434f,
-0.3426372745098039f, 0.28858098039215685f, 0.3600156862745098f,
-0.3409455882352941f, 0.2867355294117647f, 0.35801647058823527f,
-0.3392539019607843f, 0.28489007843137254f, 0.3560172549019608f,
-0.3375622156862745f, 0.28304462745098036f, 0.35401803921568625f,
-0.33587052941176476f, 0.2811991764705883f, 0.35201882352941183f,
-0.3341788431372549f, 0.27935372549019605f, 0.35001960784313724f,
-0.33192329411764704f, 0.27689315686274507f, 0.34748212418300656f,
-0.3291038823529412f, 0.2738174705882353f, 0.3444063725490196f,
-0.3262844705882353f, 0.2707417843137255f, 0.3413306209150327f,
-0.3234650588235294f, 0.26766609803921565f, 0.33825486928104576f,
-0.32064564705882354f, 0.2645904117647059f, 0.3351791176470588f,
-0.31782623529411763f, 0.26151472549019605f, 0.3321033660130719f,
-0.31500682352941173f, 0.25843903921568623f, 0.32902761437908495f,
-0.3121874117647059f, 0.25536335294117646f, 0.325951862745098f,
-0.309368f, 0.25228766666666663f, 0.3228761111111111f,
-0.3065485882352941f, 0.2492119803921568f, 0.31980035947712415f,
-0.3037291764705882f, 0.246136294117647f, 0.3167246078431372f,
-0.3009097647058823f, 0.2430606078431372f, 0.3136488562091503f,
-0.2980903529411764f, 0.2399849215686274f, 0.31057310457516335f,
-0.29527094117647057f, 0.2369092352941176f, 0.3074973529411764f,
-0.2924515294117648f, 0.23383354901960787f, 0.30442160130718954f,
-0.28927327450980383f, 0.23052715686274503f, 0.3009613725490195f,
-0.28465964705882346f, 0.22629794117647054f, 0.29596323529411755f,
-0.2800460196078431f, 0.22206872549019602f, 0.2909650980392156f,
-0.2754323921568627f, 0.21783950980392153f, 0.28596696078431366f,
-0.2708187647058823f, 0.213610294117647f, 0.2809688235294117f,
-0.2662051372549019f, 0.20938107843137252f, 0.2759706862745097f,
-0.26159150980392154f, 0.20515186274509803f, 0.27097254901960777f,
-0.25697788235294117f, 0.2009226470588235f, 0.2659744117647058f,
-0.25236425490196074f, 0.19669343137254902f, 0.2609762745098039f,
-0.24775062745098037f, 0.19246421568627453f, 0.2559781372549019f,
-0.243137f, 0.188235f, 0.25098f
-}
-, 
-{
-0.25098f, 0.015686f, 0.168627f,
-0.2474429019607843f, 0.017070078431372548f, 0.1706262549019608f,
-0.2439058039215686f, 0.018454156862745097f, 0.17262550980392158f,
-0.24036870588235293f, 0.019838235294117646f, 0.17462476470588234f,
-0.23683160784313723f, 0.021222313725490195f, 0.17662401960784313f,
-0.23329450980392155f, 0.022606392156862744f, 0.17862327450980392f,
-0.22975741176470588f, 0.023990470588235294f, 0.1806225294117647f,
-0.22622031372549017f, 0.025374549019607843f, 0.1826217843137255f,
-0.2226832156862745f, 0.026758627450980392f, 0.18462103921568626f,
-0.21914611764705882f, 0.028142705882352938f, 0.18662029411764705f,
-0.21560901960784312f, 0.02952678431372549f, 0.18861954901960784f,
-0.21207192156862745f, 0.030910862745098036f, 0.19061880392156863f,
-0.20853482352941177f, 0.03229494117647058f, 0.19261805882352942f,
-0.20499772549019607f, 0.033679019607843134f, 0.19461731372549018f,
-0.2014606274509804f, 0.03506309803921569f, 0.19661656862745097f,
-0.1979235294117647f, 0.03644717647058823f, 0.19861582352941176f,
-0.19438643137254902f, 0.03783125490196078f, 0.20061507843137255f,
-0.19084933333333334f, 0.03921533333333333f, 0.20261433333333334f,
-0.18731223529411764f, 0.040599411764705884f, 0.20461358823529413f,
-0.18377513725490197f, 0.04198349019607843f, 0.2066128431372549f,
-0.18023803921568626f, 0.043367568627450975f, 0.20861209803921568f,
-0.17670094117647062f, 0.04475164705882352f, 0.21061135294117647f,
-0.1731638431372549f, 0.04613572549019608f, 0.21261060784313726f,
-0.1696267450980392f, 0.047519803921568626f, 0.21460986274509802f,
-0.16608964705882354f, 0.04890388235294117f, 0.2166091176470588f,
-0.16255254901960786f, 0.050287960784313725f, 0.2186083725490196f,
-0.16132225490196078f, 0.05205650980392157f, 0.22122276470588234f,
-0.16239876470588235f, 0.054209529411764705f, 0.22445229411764706f,
-0.16347527450980393f, 0.05636254901960784f, 0.22768182352941177f,
-0.1645517843137255f, 0.05851556862745098f, 0.23091135294117646f,
-0.16562829411764707f, 0.06066858823529411f, 0.23414088235294117f,
-0.16670480392156864f, 0.06282160784313726f, 0.23737041176470589f,
-0.1677813137254902f, 0.06497462745098039f, 0.24059994117647057f,
-0.16885782352941175f, 0.06712764705882351f, 0.24382947058823526f,
-0.16993433333333335f, 0.06928066666666666f, 0.247059f,
-0.17101084313725493f, 0.0714336862745098f, 0.2502885294117647f,
-0.17208735294117647f, 0.07358670588235294f, 0.2535180588235294f,
-0.17316386274509804f, 0.07573972549019606f, 0.2567475882352941f,
-0.1742403725490196f, 0.07789274509803921f, 0.25997711764705883f,
-0.17531688235294118f, 0.08004576470588236f, 0.2632066470588235f,
-0.17639339215686275f, 0.08219878431372549f, 0.2664361764705882f,
-0.17746990196078433f, 0.08435180392156862f, 0.2696657058823529f,
-0.1785464117647059f, 0.08650482352941177f, 0.27289523529411763f,
-0.17962292156862747f, 0.0886578431372549f, 0.2761247647058823f,
-0.18069943137254904f, 0.09081086274509803f, 0.27935429411764706f,
-0.18177594117647058f, 0.09296388235294117f, 0.2825838235294117f,
-0.18285245098039216f, 0.09511690196078432f, 0.2858133529411765f,
-0.18392896078431373f, 0.09726992156862746f, 0.2890428823529412f,
-0.1850054705882353f, 0.09942294117647059f, 0.29227241176470586f,
-0.18608198039215687f, 0.10157596078431372f, 0.29550194117647055f,
-0.18715849019607844f, 0.10372898039215686f, 0.2987314705882353f,
-0.188235f, 0.105882f, 0.301961f,
-0.18931150980392158f, 0.10803501960784315f, 0.305344294117647f,
-0.19038801960784313f, 0.11018803921568626f, 0.30872758823529406f,
-0.19146452941176473f, 0.11234105882352942f, 0.31211088235294115f,
-0.19254103921568627f, 0.11449407843137255f, 0.3154941764705882f,
-0.19361754901960784f, 0.11664709803921569f, 0.3188774705882353f,
-0.19469405882352941f, 0.1188001176470588f, 0.32226076470588233f,
-0.19577056862745099f, 0.12095313725490196f, 0.3256440588235294f,
-0.19684707843137256f, 0.1231061568627451f, 0.32902735294117647f,
-0.19792358823529413f, 0.12525917647058823f, 0.3324106470588235f,
-0.19900009803921567f, 0.12741219607843135f, 0.33579394117647055f,
-0.20007660784313724f, 0.12956521568627452f, 0.33917723529411764f,
-0.20115311764705882f, 0.13171823529411766f, 0.3425605294117647f,
-0.2022296274509804f, 0.13387125490196078f, 0.3459438235294118f,
-0.20330613725490196f, 0.13602427450980392f, 0.3493271176470588f,
-0.2043826470588235f, 0.13817729411764704f, 0.35271041176470586f,
-0.2054591568627451f, 0.1403303137254902f, 0.35609370588235295f,
-0.20653566666666667f, 0.14248333333333335f, 0.359477f,
-0.20761217647058824f, 0.1446363529411765f, 0.36286029411764703f,
-0.2086886862745098f, 0.1467893725490196f, 0.3662435882352941f,
-0.20976519607843136f, 0.14894239215686275f, 0.36962688235294117f,
-0.21084170588235293f, 0.1510954117647059f, 0.3730101764705882f,
-0.2119182156862745f, 0.15324843137254904f, 0.3763934705882353f,
-0.21299472549019605f, 0.15540145098039215f, 0.3797767647058823f,
-0.21407123529411765f, 0.1575544705882353f, 0.38316005882352944f,
-0.2151477450980392f, 0.15970749019607844f, 0.3865433529411765f,
-0.21691629411764704f, 0.1620911960784314f, 0.3913876470588235f,
-0.21937688235294117f, 0.16470558823529413f, 0.3976929411764706f,
-0.22183747058823527f, 0.16731998039215687f, 0.40399823529411766f,
-0.2242980588235294f, 0.1699343725490196f, 0.4103035294117647f,
-0.2267586470588235f, 0.17254876470588237f, 0.41660882352941175f,
-0.2292192352941176f, 0.17516315686274506f, 0.4229141176470587f,
-0.23167982352941174f, 0.17777754901960785f, 0.42921941176470585f,
-0.23414041176470587f, 0.1803919411764706f, 0.4355247058823529f,
-0.23660099999999998f, 0.18300633333333333f, 0.44183f,
-0.2390615882352941f, 0.1856207254901961f, 0.448135294117647f,
-0.2415221764705882f, 0.18823511764705883f, 0.4544405882352941f,
-0.24398276470588234f, 0.19084950980392157f, 0.46074588235294117f,
-0.24644335294117645f, 0.1934639019607843f, 0.4670511764705882f,
-0.24890394117647055f, 0.19607829411764702f, 0.47335647058823516f,
-0.2513645294117647f, 0.1986926862745098f, 0.4796617647058823f,
-0.2538251176470588f, 0.20130707843137255f, 0.48596705882352936f,
-0.25628570588235294f, 0.2039214705882353f, 0.49227235294117644f,
-0.25874629411764705f, 0.20653586274509803f, 0.4985776470588235f,
-0.26120688235294115f, 0.20915025490196076f, 0.5048829411764706f,
-0.26366747058823525f, 0.21176464705882353f, 0.5111882352941176f,
-0.2661280588235294f, 0.21437903921568627f, 0.5174935294117646f,
-0.26858864705882346f, 0.21699343137254898f, 0.5237988235294117f,
-0.2710492352941176f, 0.21960782352941174f, 0.5301041176470588f,
-0.2735098235294117f, 0.22222221568627448f, 0.5364094117647058f,
-0.2759704117647059f, 0.22483660784313725f, 0.5427147058823529f,
-0.278431f, 0.227451f, 0.54902f,
-0.2801226862745098f, 0.22975782352941176f, 0.5550176862745098f,
-0.2818143725490196f, 0.23206464705882351f, 0.5610153725490196f,
-0.2835060588235294f, 0.2343714705882353f, 0.5670130588235294f,
-0.2851977450980392f, 0.23667829411764701f, 0.5730107450980391f,
-0.286889431372549f, 0.23898511764705882f, 0.579008431372549f,
-0.2885811176470588f, 0.24129194117647057f, 0.5850061176470588f,
-0.2902728039215686f, 0.24359876470588235f, 0.5910038039215686f,
-0.2919644901960784f, 0.2459055882352941f, 0.5970014901960784f,
-0.2936561764705882f, 0.24821241176470588f, 0.6029991764705882f,
-0.29534786274509806f, 0.2505192352941176f, 0.608996862745098f,
-0.29703954901960783f, 0.25282605882352943f, 0.6149945490196078f,
-0.2987312352941176f, 0.2551328823529411f, 0.6209922352941175f,
-0.30042292156862743f, 0.25743970588235293f, 0.6269899215686274f,
-0.30211460784313726f, 0.2597465294117647f, 0.6329876078431372f,
-0.30380629411764704f, 0.2620533529411765f, 0.638985294117647f,
-0.30549798039215686f, 0.26436017647058824f, 0.6449829803921568f,
-0.30718966666666664f, 0.266667f, 0.6509806666666667f,
-0.30888135294117647f, 0.2689738235294118f, 0.6569783529411765f,
-0.31057303921568624f, 0.27128064705882354f, 0.6629760392156863f,
-0.312264725490196f, 0.27358747058823524f, 0.668973725490196f,
-0.3139564117647059f, 0.27589429411764704f, 0.6749714117647059f,
-0.3156480980392157f, 0.27820111764705885f, 0.6809690980392157f,
-0.3173397843137255f, 0.2805079411764706f, 0.6869667843137255f,
-0.3190314705882353f, 0.28281476470588235f, 0.6929644705882352f,
-0.3207231568627451f, 0.2851215882352941f, 0.6989621568627451f,
-0.32218413725490197f, 0.28758217647058826f, 0.7038833333333333f,
-0.32341441176470587f, 0.2901965294117647f, 0.7077279999999999f,
-0.3246446862745098f, 0.29281088235294117f, 0.7115726666666666f,
-0.3258749607843137f, 0.2954252352941176f, 0.7154173333333333f,
-0.3271052352941176f, 0.29803958823529403f, 0.7192619999999998f,
-0.32833550980392157f, 0.3006539411764706f, 0.7231066666666667f,
-0.32956578431372546f, 0.30326829411764705f, 0.7269513333333333f,
-0.3307960588235294f, 0.3058826470588235f, 0.730796f,
-0.3320263333333333f, 0.308497f, 0.7346406666666666f,
-0.33325660784313726f, 0.3111113529411765f, 0.7384853333333333f,
-0.33448688235294116f, 0.31372570588235293f, 0.7423299999999999f,
-0.3357171568627451f, 0.3163400588235294f, 0.7461746666666667f,
-0.336947431372549f, 0.3189544117647059f, 0.7500193333333334f,
-0.33817770588235296f, 0.32156876470588236f, 0.753864f,
-0.33940798039215686f, 0.3241831176470588f, 0.7577086666666667f,
-0.34063825490196076f, 0.3267974705882353f, 0.7615533333333333f,
-0.3418685294117647f, 0.3294118235294118f, 0.765398f,
-0.3430988039215686f, 0.33202617647058824f, 0.7692426666666667f,
-0.34432907843137256f, 0.3346405294117647f, 0.7730873333333333f,
-0.34555935294117646f, 0.33725488235294115f, 0.776932f,
-0.34678962745098035f, 0.33986923529411756f, 0.7807766666666666f,
-0.3480199019607843f, 0.3424835882352941f, 0.7846213333333334f,
-0.34925017647058826f, 0.3450979411764706f, 0.788466f,
-0.35048045098039216f, 0.34771229411764704f, 0.7923106666666667f,
-0.3517107254901961f, 0.35032664705882355f, 0.7961553333333333f,
-0.352941f, 0.352941f, 0.8f,
-0.35524782352941175f, 0.35709327450980394f, 0.8030757254901961f,
-0.35755464705882356f, 0.36124554901960787f, 0.8061514509803922f,
-0.3598614705882353f, 0.36539782352941175f, 0.8092271764705883f,
-0.36216829411764706f, 0.3695500980392157f, 0.8123029019607844f,
-0.3644751176470588f, 0.3737023725490196f, 0.8153786274509804f,
-0.3667819411764706f, 0.37785464705882355f, 0.8184543529411765f,
-0.36908876470588237f, 0.3820069215686275f, 0.8215300784313726f,
-0.3713955882352941f, 0.38615919607843135f, 0.8246058039215687f,
-0.37370241176470587f, 0.3903114705882353f, 0.8276815294117648f,
-0.37600923529411767f, 0.3944637450980392f, 0.8307572549019608f,
-0.37831605882352937f, 0.39861601960784304f, 0.8338329803921568f,
-0.38062288235294117f, 0.4027682941176471f, 0.836908705882353f,
-0.3829297058823529f, 0.40692056862745096f, 0.839984431372549f,
-0.3852365294117647f, 0.4110728431372549f, 0.8430601568627452f,
-0.3875433529411765f, 0.41522511764705883f, 0.8461358823529411f,
-0.38985017647058823f, 0.41937739215686276f, 0.8492116078431372f,
-0.392157f, 0.4235296666666667f, 0.8522873333333333f,
-0.3944638235294118f, 0.4276819411764706f, 0.8553630588235294f,
-0.39677064705882353f, 0.4318342156862745f, 0.8584387843137254f,
-0.3990774705882353f, 0.43598649019607844f, 0.8615145098039215f,
-0.40138429411764703f, 0.44013876470588237f, 0.8645902352941176f,
-0.40369111764705884f, 0.4442910392156863f, 0.8676659607843137f,
-0.4059979411764706f, 0.4484433137254902f, 0.8707416862745098f,
-0.40830476470588234f, 0.4525955882352941f, 0.8738174117647058f,
-0.4106115882352941f, 0.45674786274509804f, 0.8768931372549019f,
-0.413610431372549f, 0.4612845882352941f, 0.8788923725490195f,
-0.41730129411764694f, 0.4662057647058822f, 0.8798151176470588f,
-0.4209921568627451f, 0.4711269411764706f, 0.880737862745098f,
-0.42468301960784316f, 0.47604811764705884f, 0.8816606078431373f,
-0.42837388235294116f, 0.48096929411764705f, 0.8825833529411764f,
-0.4320647450980392f, 0.4858904705882353f, 0.8835060980392156f,
-0.43575560784313727f, 0.4908116470588235f, 0.8844288431372549f,
-0.4394464705882353f, 0.4957328235294118f, 0.8853515882352941f,
-0.4431373333333333f, 0.500654f, 0.8862743333333333f,
-0.4468281960784314f, 0.5055751764705883f, 0.8871970784313725f,
-0.45051905882352944f, 0.5104963529411765f, 0.8881198235294118f,
-0.4542099215686275f, 0.5154175294117647f, 0.889042568627451f,
-0.4579007843137255f, 0.520338705882353f, 0.8899653137254901f,
-0.46159164705882355f, 0.5252598823529412f, 0.8908880588235294f,
-0.4652825098039216f, 0.5301810588235294f, 0.8918108039215686f,
-0.4689733725490196f, 0.5351022352941176f, 0.8927335490196079f,
-0.47266423529411766f, 0.5400234117647059f, 0.893656294117647f,
-0.4763550980392156f, 0.544944588235294f, 0.8945790392156863f,
-0.4800459607843138f, 0.5498657647058823f, 0.8955017843137255f,
-0.48373682352941183f, 0.5547869411764705f, 0.8964245294117648f,
-0.48742768627450983f, 0.5597081176470589f, 0.8973472745098039f,
-0.4911185490196079f, 0.5646292941176471f, 0.8982700196078431f,
-0.49480941176470594f, 0.5695504705882353f, 0.8991927647058824f,
-0.49850027450980394f, 0.5744716470588235f, 0.9001155098039215f,
-0.502191137254902f, 0.5793928235294118f, 0.9010382549019608f,
-0.505882f, 0.584314f, 0.901961f,
-0.5083425882352942f, 0.5883124705882353f, 0.9027299411764707f,
-0.5108031764705883f, 0.5923109411764705f, 0.9034988823529412f,
-0.5132637647058824f, 0.5963094117647059f, 0.9042678235294118f,
-0.5157243529411765f, 0.6003078823529412f, 0.9050367647058823f,
-0.5181849411764706f, 0.6043063529411765f, 0.905805705882353f,
-0.5206455294117648f, 0.6083048235294117f, 0.9065746470588235f,
-0.5231061176470588f, 0.6123032941176471f, 0.9073435882352942f,
-0.5255667058823529f, 0.6163017647058823f, 0.9081125294117647f,
-0.528027294117647f, 0.6203002352941176f, 0.9088814705882353f,
-0.5304878823529412f, 0.6242987058823529f, 0.9096504117647058f,
-0.5329484705882354f, 0.6282971764705882f, 0.9104193529411765f,
-0.5354090588235294f, 0.6322956470588235f, 0.911188294117647f,
-0.5378696470588236f, 0.6362941176470588f, 0.9119572352941177f,
-0.5403302352941177f, 0.6402925882352941f, 0.9127261764705882f,
-0.5427908235294118f, 0.6442910588235294f, 0.9134951176470588f,
-0.5452514117647059f, 0.6482895294117647f, 0.9142640588235293f,
-0.547712f, 0.652288f, 0.915033f,
-0.5501725882352941f, 0.6562864705882353f, 0.9158019411764706f,
-0.5526331764705883f, 0.6602849411764705f, 0.9165708823529412f,
-0.5550937647058823f, 0.6642834117647058f, 0.9173398235294118f,
-0.5575543529411765f, 0.6682818823529412f, 0.9181087647058823f,
-0.5600149411764705f, 0.6722803529411765f, 0.918877705882353f,
-0.5624755294117647f, 0.6762788235294117f, 0.9196466470588235f,
-0.5649361176470588f, 0.6802772941176469f, 0.920415588235294f,
-0.5673967058823529f, 0.6842757647058824f, 0.9211845294117647f,
-0.5717796470588236f, 0.6888893725490196f, 0.922491705882353f,
-0.5780849411764706f, 0.6941181176470588f, 0.9243371176470588f,
-0.5843902352941176f, 0.699346862745098f, 0.9261825294117647f,
-0.5906955294117647f, 0.7045756078431372f, 0.9280279411764706f,
-0.5970008235294118f, 0.7098043529411765f, 0.9298733529411765f,
-0.6033061176470588f, 0.7150330980392157f, 0.9317187647058823f,
-0.6096114117647059f, 0.7202618431372548f, 0.9335641764705882f,
-0.6159167058823529f, 0.7254905882352941f, 0.9354095882352941f,
-0.6222219999999999f, 0.7307193333333333f, 0.937255f,
-0.6285272941176471f, 0.7359480784313726f, 0.9391004117647058f,
-0.6348325882352941f, 0.7411768235294117f, 0.9409458235294118f,
-0.6411378823529411f, 0.746405568627451f, 0.9427912352941177f,
-0.6474431764705882f, 0.7516343137254902f, 0.9446366470588236f,
-0.6537484705882353f, 0.7568630588235294f, 0.9464820588235294f,
-0.6600537647058822f, 0.7620918039215685f, 0.9483274705882352f,
-0.6663590588235294f, 0.7673205490196078f, 0.9501728823529412f,
-0.6726643529411764f, 0.7725492941176471f, 0.952018294117647f,
-0.6789696470588235f, 0.7777780392156862f, 0.9538637058823529f,
-0.6852749411764706f, 0.7830067843137255f, 0.9557091176470588f,
-0.6915802352941176f, 0.7882355294117647f, 0.9575545294117647f,
-0.6978855294117647f, 0.793464274509804f, 0.9593999411764706f,
-0.7041908235294118f, 0.7986930196078431f, 0.9612453529411765f,
-0.7104961176470588f, 0.8039217647058824f, 0.9630907647058824f,
-0.7168014117647058f, 0.8091505098039216f, 0.9649361764705883f,
-0.7231067058823529f, 0.8143792549019608f, 0.9667815882352941f,
-0.729412f, 0.819608f, 0.968627f
-}
-, 
-{
-0.988235f, 0.976471f, 0.909804f,
-0.9866971176470588f, 0.9724724901960784f, 0.8968858823529411f,
-0.9851592352941176f, 0.9684739803921568f, 0.8839677647058823f,
-0.9836213529411765f, 0.9644754705882352f, 0.8710496470588235f,
-0.9820834705882353f, 0.9604769607843138f, 0.8581315294117646f,
-0.9805455882352941f, 0.9564784509803922f, 0.8452134117647059f,
-0.979007705882353f, 0.9524799411764706f, 0.832295294117647f,
-0.9774698235294118f, 0.948481431372549f, 0.8193771764705882f,
-0.9759319411764706f, 0.9444829215686275f, 0.8064590588235294f,
-0.9743940588235294f, 0.9404844117647059f, 0.7935409411764706f,
-0.9728561764705882f, 0.9364859019607843f, 0.7806228235294118f,
-0.971318294117647f, 0.9324873921568628f, 0.767704705882353f,
-0.9697804117647059f, 0.9284888823529412f, 0.7547865882352942f,
-0.9682425490196078f, 0.9238752352941176f, 0.7420222549019608f,
-0.9667047450980393f, 0.9174161764705883f, 0.729719274509804f,
-0.9651669411764706f, 0.9109571176470589f, 0.7174162941176471f,
-0.963629137254902f, 0.9044980588235294f, 0.7051133137254902f,
-0.9620913333333333f, 0.898039f, 0.6928103333333333f,
-0.9605535294117648f, 0.8915799411764707f, 0.6805073529411765f,
-0.9590157254901961f, 0.8851208823529412f, 0.6682043725490197f,
-0.9574779215686274f, 0.8786618235294118f, 0.6559013921568627f,
-0.9559401176470588f, 0.8722027647058824f, 0.6435984117647059f,
-0.9544023137254902f, 0.8657437058823529f, 0.631295431372549f,
-0.9528645098039216f, 0.8592846470588236f, 0.6189924509803921f,
-0.9513267058823529f, 0.8528255882352942f, 0.6066894705882353f,
-0.9497889019607842f, 0.8463665294117647f, 0.5943864901960784f,
-0.9487123921568628f, 0.8403688431372549f, 0.5854668431372548f,
-0.9480971764705882f, 0.8348325294117648f, 0.5799305294117647f,
-0.9474819607843137f, 0.8292962156862745f, 0.5743942156862745f,
-0.9468667450980393f, 0.8237599019607843f, 0.5688579019607842f,
-0.9462515294117647f, 0.8182235882352942f, 0.5633215882352941f,
-0.9456363137254902f, 0.8126872745098039f, 0.5577852745098039f,
-0.9450210980392156f, 0.8071509607843138f, 0.5522489607843137f,
-0.9444058823529412f, 0.8016146470588236f, 0.5467126470588235f,
-0.9437906666666667f, 0.7960783333333333f, 0.5411763333333333f,
-0.9431754509803921f, 0.7905420196078432f, 0.5356400196078431f,
-0.9425602352941177f, 0.785005705882353f, 0.5301037058823529f,
-0.9419450196078432f, 0.7794693921568628f, 0.5245673921568628f,
-0.9413298039215686f, 0.7739330784313726f, 0.5190310784313725f,
-0.940484f, 0.7674740588235295f, 0.5134947647058823f,
-0.9395613333333334f, 0.7607074705882353f, 0.5079584509803922f,
-0.9386386666666667f, 0.7539408823529412f, 0.5024221372549019f,
-0.937716f, 0.7471742941176471f, 0.49688582352941174f,
-0.9367933333333334f, 0.7404077058823529f, 0.49134950980392156f,
-0.9358706666666666f, 0.7336411176470589f, 0.4858131960784313f,
-0.934948f, 0.7268745294117648f, 0.4802768823529412f,
-0.9340253333333334f, 0.7201079411764706f, 0.47474056862745095f,
-0.9331026666666666f, 0.7133413529411765f, 0.46920425490196077f,
-0.93218f, 0.7065747647058823f, 0.4636679411764706f,
-0.9312573333333334f, 0.6998081764705882f, 0.4581316274509804f,
-0.9303346666666666f, 0.693041588235294f, 0.45259531372549017f,
-0.929412f, 0.686275f, 0.447059f,
-0.9278741176470588f, 0.6807386078431372f, 0.4427529607843137f,
-0.9263362352941177f, 0.6752022156862745f, 0.43844692156862747f,
-0.9247983529411765f, 0.6696658235294117f, 0.4341408823529412f,
-0.9232604705882352f, 0.664129431372549f, 0.4298348431372549f,
-0.9217225882352941f, 0.6585930392156862f, 0.4255288039215686f,
-0.9201847058823529f, 0.6530566470588235f, 0.4212227647058824f,
-0.9186468235294117f, 0.6475202549019607f, 0.41691672549019604f,
-0.9171089411764706f, 0.641983862745098f, 0.4126106862745098f,
-0.9155710588235294f, 0.6364474705882353f, 0.4083046470588235f,
-0.9140331764705882f, 0.6309110784313725f, 0.40399860784313724f,
-0.912495294117647f, 0.6253746862745098f, 0.39969256862745095f,
-0.9109574117647058f, 0.619838294117647f, 0.39538652941176466f,
-0.9096502156862745f, 0.6145326078431372f, 0.3915418431372549f,
-0.9090350784313725f, 0.6099190392156862f, 0.38908121568627446f,
-0.9084199411764705f, 0.6053054705882354f, 0.38662058823529416f,
-0.9078048039215686f, 0.6006919019607843f, 0.3841599607843137f,
-0.9071896666666667f, 0.5960783333333333f, 0.38169933333333333f,
-0.9065745294117646f, 0.5914647647058823f, 0.3792387058823529f,
-0.9059593921568627f, 0.5868511960784314f, 0.3767780784313725f,
-0.9053442549019608f, 0.5822376274509804f, 0.37431745098039215f,
-0.9047291176470588f, 0.5776240588235294f, 0.37185682352941174f,
-0.9041139803921568f, 0.5730104901960784f, 0.3693961960784314f,
-0.9034988431372549f, 0.5683969215686275f, 0.366935568627451f,
-0.9028837058823529f, 0.5637833529411764f, 0.36447494117647056f,
-0.902268568627451f, 0.5591697843137254f, 0.3620143137254902f,
-0.9014996274509804f, 0.5536334705882353f, 0.35986129411764706f,
-0.9005768823529412f, 0.5471744117647058f, 0.3580158823529412f,
-0.899654137254902f, 0.5407153529411765f, 0.3561704705882353f,
-0.8987313921568627f, 0.5342562941176471f, 0.35432505882352944f,
-0.8978086470588236f, 0.5277972352941176f, 0.3524796470588235f,
-0.8968859019607843f, 0.5213381764705883f, 0.3506342352941177f,
-0.8959631568627451f, 0.5148791176470588f, 0.34878882352941176f,
-0.8950404117647058f, 0.5084200588235294f, 0.3469434117647059f,
-0.8941176666666667f, 0.501961f, 0.345098f,
-0.8931949215686275f, 0.49550194117647056f, 0.34325258823529414f,
-0.8922721764705882f, 0.4890428823529412f, 0.34140717647058827f,
-0.891349431372549f, 0.48258382352941176f, 0.3395617647058824f,
-0.8904266862745098f, 0.47612476470588233f, 0.3377163529411765f,
-0.8895039411764706f, 0.4703577647058824f, 0.3356402352941177f,
-0.8885811960784313f, 0.4648214509803921f, 0.33348721568627454f,
-0.8876584509803921f, 0.45928513725490194f, 0.3313341960784314f,
-0.886735705882353f, 0.45374882352941176f, 0.32918117647058825f,
-0.8858129607843137f, 0.4482125098039216f, 0.3270281568627451f,
-0.8848902156862745f, 0.44267619607843134f, 0.32487513725490197f,
-0.8839674705882352f, 0.43713988235294116f, 0.3227221176470588f,
-0.8830447254901961f, 0.431603568627451f, 0.32056909803921574f,
-0.8821219803921568f, 0.42606725490196085f, 0.3184160784313726f,
-0.8811992352941176f, 0.4205309411764706f, 0.31626305882352945f,
-0.8802764901960783f, 0.41499462745098037f, 0.3141100392156863f,
-0.8793537450980392f, 0.4094583137254902f, 0.31195701960784317f,
-0.878431f, 0.403922f, 0.309804f,
-0.8753553137254901f, 0.39807807843137255f, 0.305805568627451f,
-0.8722796274509803f, 0.3922341568627451f, 0.301807137254902f,
-0.8692039411764706f, 0.38639023529411765f, 0.297808705882353f,
-0.8661282549019608f, 0.3805463137254903f, 0.293810274509804f,
-0.8630525686274509f, 0.37470239215686274f, 0.28981184313725494f,
-0.8599768823529411f, 0.3688584705882353f, 0.28581341176470587f,
-0.8569011960784313f, 0.36301454901960784f, 0.28181498039215686f,
-0.8538255098039216f, 0.3571706274509804f, 0.27781654901960784f,
-0.8507498235294118f, 0.35132670588235293f, 0.2738181176470588f,
-0.8476741372549019f, 0.3454827843137255f, 0.2698196862745098f,
-0.8445984509803921f, 0.339638862745098f, 0.2658212549019608f,
-0.8415227647058824f, 0.3337949411764707f, 0.26182282352941183f,
-0.8384470588235293f, 0.3284123725490196f, 0.2584395294117647f,
-0.835371294117647f, 0.32441386274509804f, 0.2569016470588235f,
-0.8322955294117647f, 0.32041535294117646f, 0.25536376470588235f,
-0.8292197647058823f, 0.3164168431372549f, 0.2538258823529412f,
-0.826144f, 0.3124183333333333f, 0.252288f,
-0.8230682352941177f, 0.3084198235294118f, 0.25075011764705885f,
-0.8199924705882353f, 0.3044213137254902f, 0.24921223529411765f,
-0.816916705882353f, 0.30042280392156867f, 0.2476743529411765f,
-0.8138409411764707f, 0.29642429411764704f, 0.2461364705882353f,
-0.8107651764705882f, 0.29242578431372546f, 0.24459858823529412f,
-0.8076894117647059f, 0.2884272745098039f, 0.24306070588235296f,
-0.8046136470588235f, 0.28442876470588235f, 0.2415228235294118f,
-0.8015378823529412f, 0.2804302549019608f, 0.2399849411764706f,
-0.7980007843137256f, 0.2758166470588235f, 0.23813949019607844f,
-0.7940023529411765f, 0.27058794117647056f, 0.2359864705882353f,
-0.7900039215686275f, 0.26535923529411765f, 0.23383345098039215f,
-0.7860054901960785f, 0.2601305294117647f, 0.23168043137254904f,
-0.7820070588235296f, 0.25490182352941193f, 0.22952741176470595f,
-0.7780086274509804f, 0.24967311764705882f, 0.22737439215686275f,
-0.7740101960784314f, 0.24444441176470588f, 0.2252213725490196f,
-0.7700117647058824f, 0.23921570588235294f, 0.2230683529411765f,
-0.7660133333333333f, 0.233987f, 0.22091533333333335f,
-0.7620149019607844f, 0.22875829411764706f, 0.2187623137254902f,
-0.7580164705882353f, 0.22352958823529412f, 0.21660929411764707f,
-0.7540180392156863f, 0.21830088235294118f, 0.21445627450980392f,
-0.7500196078431373f, 0.21307217647058824f, 0.21230325490196078f,
-0.7462518235294118f, 0.20991952941176473f, 0.21153429411764707f,
-0.7425609215686274f, 0.20745890196078431f, 0.2112266862745098f,
-0.7388700196078432f, 0.20499827450980393f, 0.21091907843137256f,
-0.7351791176470588f, 0.20253764705882354f, 0.2106114705882353f,
-0.7314882156862745f, 0.20007701960784313f, 0.21030386274509805f,
-0.7277973137254902f, 0.19761639215686275f, 0.20999625490196078f,
-0.7241064117647059f, 0.19515576470588236f, 0.20968864705882354f,
-0.7204155098039217f, 0.19269513725490203f, 0.2093810392156863f,
-0.7167246078431372f, 0.19023450980392156f, 0.20907343137254902f,
-0.7130337058823529f, 0.18777388235294118f, 0.20876582352941175f,
-0.7093428039215686f, 0.1853132549019608f, 0.2084582156862745f,
-0.7056519019607843f, 0.18285262745098038f, 0.20815060784313727f,
-0.701961f, 0.180392f, 0.207843f,
-0.6985777058823529f, 0.17916172549019607f, 0.207843f,
-0.6951944117647059f, 0.17793145098039215f, 0.207843f,
-0.6918111176470588f, 0.17670117647058822f, 0.207843f,
-0.6884278235294117f, 0.1754709019607843f, 0.207843f,
-0.6850445294117646f, 0.1742406274509804f, 0.207843f,
-0.6816612352941176f, 0.17301035294117648f, 0.207843f,
-0.6782779411764706f, 0.17178007843137255f, 0.207843f,
-0.6748946470588235f, 0.17054980392156863f, 0.207843f,
-0.6715113529411765f, 0.1693195294117647f, 0.207843f,
-0.6681280588235293f, 0.16808925490196078f, 0.207843f,
-0.6647447647058824f, 0.16685898039215688f, 0.207843f,
-0.6613614705882352f, 0.16562870588235293f, 0.207843f,
-0.6576705882352941f, 0.1640908431372549f, 0.2074585294117647f,
-0.6530569411764705f, 0.1616302156862745f, 0.20592064705882354f,
-0.648443294117647f, 0.15916958823529412f, 0.20438276470588235f,
-0.6438296470588235f, 0.15670896078431373f, 0.20284488235294118f,
-0.639216f, 0.15424833333333332f, 0.201307f,
-0.6346023529411764f, 0.15178770588235294f, 0.19976911764705882f,
-0.6299887058823529f, 0.14932707843137255f, 0.19823123529411765f,
-0.6253750588235294f, 0.14686645098039217f, 0.19669335294117649f,
-0.6207614117647059f, 0.14440582352941175f, 0.1951554705882353f,
-0.6161477647058823f, 0.14194519607843137f, 0.19361758823529412f,
-0.6115341176470588f, 0.13948456862745098f, 0.19207970588235296f,
-0.6069204705882353f, 0.1370239411764706f, 0.1905418235294118f,
-0.6023068235294118f, 0.1345633137254902f, 0.1890039411764706f,
-0.5976931764705882f, 0.13256405882352942f, 0.18808123529411766f,
-0.5930795294117648f, 0.13102617647058828f, 0.18777370588235295f,
-0.5884658823529412f, 0.12948829411764706f, 0.18746617647058825f,
-0.5838522352941177f, 0.1279504117647059f, 0.18715864705882354f,
-0.5792385882352941f, 0.1264125294117647f, 0.18685111764705883f,
-0.5746249411764706f, 0.12487464705882353f, 0.18654358823529413f,
-0.5700112941176471f, 0.12333676470588237f, 0.18623605882352942f,
-0.5653976470588236f, 0.12179888235294119f, 0.18592852941176471f,
-0.560784f, 0.120261f, 0.185621f,
-0.5561703529411764f, 0.11872311764705883f, 0.1853134705882353f,
-0.551556705882353f, 0.11718523529411766f, 0.1850059411764706f,
-0.5469430588235294f, 0.11564735294117648f, 0.1846984117647059f,
-0.5423294117647058f, 0.1141094705882353f, 0.18439088235294118f,
-0.5388692352941177f, 0.11280229411764707f, 0.18408329411764707f,
-0.5357935490196079f, 0.11157201960784315f, 0.1837756862745098f,
-0.532717862745098f, 0.11034174509803922f, 0.18346807843137256f,
-0.5296421764705882f, 0.1091114705882353f, 0.18316047058823529f,
-0.5265664901960785f, 0.10788119607843141f, 0.18285286274509804f,
-0.5234908039215687f, 0.10665092156862746f, 0.18254525490196077f,
-0.5204151176470588f, 0.10542064705882354f, 0.18223764705882353f,
-0.517339431372549f, 0.10419037254901961f, 0.18193003921568626f,
-0.5142637450980392f, 0.10296009803921569f, 0.18162243137254902f,
-0.5111880588235294f, 0.10172982352941176f, 0.18131482352941175f,
-0.5081123725490196f, 0.10049954901960785f, 0.1810072156862745f,
-0.5050366862745098f, 0.09926927450980393f, 0.18069960784313727f,
-0.501961f, 0.098039f, 0.180392f,
-0.49703982352941173f, 0.09650111764705882f, 0.17916172549019607f,
-0.4921186470588235f, 0.09496323529411765f, 0.17793145098039215f,
-0.48719747058823526f, 0.09342535294117647f, 0.17670117647058822f,
-0.48227629411764705f, 0.09188747058823529f, 0.1754709019607843f,
-0.4773551176470588f, 0.09034958823529413f, 0.1742406274509804f,
-0.4724339411764706f, 0.08881170588235295f, 0.17301035294117648f,
-0.4675127647058823f, 0.08727382352941176f, 0.17178007843137255f,
-0.4625915882352942f, 0.08573594117647063f, 0.17054980392156865f,
-0.45767041176470585f, 0.08419805882352942f, 0.1693195294117647f,
-0.45274923529411765f, 0.08266017647058824f, 0.16808925490196078f,
-0.4478280588235294f, 0.08112229411764706f, 0.16685898039215685f,
-0.4429068823529412f, 0.07958441176470588f, 0.16562870588235293f,
-0.4388315294117647f, 0.07820033333333333f, 0.1646290980392157f,
-0.4372936470588235f, 0.07727766666666666f, 0.16432149019607842f,
-0.43575576470588234f, 0.076355f, 0.16401388235294118f,
-0.4342178823529412f, 0.07543233333333334f, 0.1637062745098039f,
-0.43268f, 0.07450966666666667f, 0.16339866666666666f,
-0.4311421176470588f, 0.073587f, 0.16309105882352942f,
-0.4296042352941176f, 0.07266433333333333f, 0.16278345098039215f,
-0.42806635294117645f, 0.07174166666666668f, 0.1624758431372549f,
-0.4265284705882353f, 0.07081900000000001f, 0.16216823529411764f,
-0.4249905882352941f, 0.06989633333333334f, 0.1618606274509804f,
-0.42345270588235295f, 0.06897366666666667f, 0.16155301960784316f,
-0.4219148235294118f, 0.06805100000000003f, 0.1612454117647059f,
-0.42037694117647056f, 0.06712833333333333f, 0.16093780392156865f,
-0.41653223529411765f, 0.06605182352941177f, 0.15986129411764707f,
-0.4103807058823529f, 0.0648214705882353f, 0.15801588235294117f,
-0.4042291764705882f, 0.06359111764705883f, 0.1561704705882353f,
-0.39807764705882354f, 0.06236076470588235f, 0.15432505882352943f,
-0.3919261176470588f, 0.061130411764705885f, 0.15247964705882353f,
-0.3857745882352941f, 0.05990005882352941f, 0.15063423529411765f,
-0.3796230588235294f, 0.05866970588235294f, 0.14878882352941175f,
-0.3734715294117647f, 0.05743935294117647f, 0.14694341176470588f,
-0.36732f, 0.056209f, 0.145098f,
-0.36116847058823526f, 0.05497864705882353f, 0.1432525882352941f,
-0.3550169411764706f, 0.05374829411764706f, 0.14140717647058823f,
-0.34886541176470587f, 0.05251794117647059f, 0.13956176470588233f,
-0.34271388235294115f, 0.05128758823529411f, 0.13771635294117646f,
-0.3374851176470588f, 0.050288f, 0.1361015882352941f,
-0.3325639411764707f, 0.04936533333333336f, 0.13456370588235297f,
-0.32764276470588233f, 0.04844266666666666f, 0.13302582352941175f,
-0.3227215882352941f, 0.04752f, 0.13148794117647059f,
-0.31780041176470586f, 0.04659733333333333f, 0.1299500588235294f,
-0.31287923529411765f, 0.04567466666666667f, 0.12841217647058822f,
-0.3079580588235294f, 0.044752f, 0.12687429411764706f,
-0.3030368823529412f, 0.04382933333333333f, 0.1253364117647059f,
-0.2981157058823529f, 0.04290666666666666f, 0.1237985294117647f,
-0.2931945294117647f, 0.041984f, 0.12226064705882353f,
-0.28827335294117645f, 0.04106133333333333f, 0.12072276470588235f,
-0.28335217647058825f, 0.04013866666666667f, 0.11918488235294118f,
-0.278431f, 0.039216f, 0.117647f
-}
-, 
-{
-0.988235f, 0.882353f, 0.843137f,
-0.9863126470588235f, 0.8738946666666667f, 0.8327564117647059f,
-0.984390294117647f, 0.8654363333333334f, 0.8223758235294117f,
-0.9824679411764706f, 0.856978f, 0.8119952352941177f,
-0.9805455882352941f, 0.8485196666666667f, 0.8016146470588236f,
-0.9786232352941177f, 0.8400613333333333f, 0.7912340588235294f,
-0.9767008823529412f, 0.831603f, 0.7808534705882353f,
-0.9747785294117647f, 0.8231446666666666f, 0.7704728823529412f,
-0.9728561764705882f, 0.8146863333333333f, 0.7600922941176471f,
-0.9709338235294118f, 0.806228f, 0.7497117058823529f,
-0.9690114705882353f, 0.7977696666666666f, 0.7393311176470588f,
-0.968627f, 0.7893114117647059f, 0.7280277843137255f,
-0.968627f, 0.7808531764705882f, 0.7164937647058823f,
-0.968627f, 0.7723949411764706f, 0.7049597450980393f,
-0.968627f, 0.763936705882353f, 0.6934257254901961f,
-0.968627f, 0.7554784705882353f, 0.681891705882353f,
-0.968627f, 0.7470202352941177f, 0.6703576862745099f,
-0.968627f, 0.7385619999999999f, 0.6588236666666667f,
-0.968627f, 0.7301037647058823f, 0.6472896470588236f,
-0.968627f, 0.7216455294117647f, 0.6357556274509805f,
-0.968627f, 0.713187294117647f, 0.6242216078431373f,
-0.9683963235294117f, 0.7065744705882353f, 0.6160324411764706f,
-0.9680118627450981f, 0.7011919215686274f, 0.6100731764705883f,
-0.9676274019607843f, 0.6958093725490196f, 0.604113911764706f,
-0.9672429411764706f, 0.6904268235294118f, 0.5981546470588236f,
-0.9668584803921568f, 0.6850442745098039f, 0.5921953823529412f,
-0.9664740196078432f, 0.6796617254901961f, 0.5862361176470589f,
-0.9660895588235294f, 0.6742791764705882f, 0.5802768529411765f,
-0.9657050980392157f, 0.6688966274509804f, 0.5743175882352942f,
-0.9653206372549019f, 0.6635140784313726f, 0.5683583235294118f,
-0.9649361764705883f, 0.6581315294117647f, 0.5623990588235295f,
-0.9645517156862745f, 0.6527489803921569f, 0.5564397941176471f,
-0.9641672549019608f, 0.647366431372549f, 0.5504805294117647f,
-0.963782794117647f, 0.6419838823529412f, 0.5445212647058825f,
-0.9633983333333334f, 0.6366013333333334f, 0.538562f,
-0.9630138725490196f, 0.6312187843137255f, 0.5326027352941177f,
-0.9626294117647058f, 0.6258362352941177f, 0.5266434705882354f,
-0.9622449509803921f, 0.6204536862745098f, 0.520684205882353f,
-0.9618604901960784f, 0.6150711372549019f, 0.5147249411764706f,
-0.9614760294117647f, 0.6096885882352941f, 0.5087656764705882f,
-0.9610915686274509f, 0.6043060392156863f, 0.502806411764706f,
-0.9606302222222222f, 0.5986159215686275f, 0.49650113071895435f,
-0.9598613333333333f, 0.5916955294117647f, 0.4888117843137255f,
-0.9590924444444444f, 0.5847751372549019f, 0.4811224379084968f,
-0.9583235555555555f, 0.5778547450980392f, 0.473433091503268f,
-0.9575546666666667f, 0.5709343529411766f, 0.4657437450980393f,
-0.9567857777777777f, 0.5640139607843138f, 0.4580543986928105f,
-0.9560168888888889f, 0.557093568627451f, 0.4503650522875817f,
-0.955248f, 0.5501731764705883f, 0.44267570588235294f,
-0.9544791111111111f, 0.5432527843137256f, 0.4349863594771243f,
-0.9537102222222222f, 0.5363323921568628f, 0.42729701307189544f,
-0.9529413333333333f, 0.529412f, 0.4196076666666667f,
-0.9521724444444444f, 0.5224916078431373f, 0.41191832026143793f,
-0.9514035555555556f, 0.5155712156862746f, 0.40422897385620926f,
-0.9506346666666666f, 0.5086508235294118f, 0.3965396274509804f,
-0.9498657777777777f, 0.501730431372549f, 0.38885028104575164f,
-0.9490968888888889f, 0.4948100392156863f, 0.3811609346405229f,
-0.9474052352941177f, 0.48904300000000006f, 0.37531705882352945f,
-0.9456110522875817f, 0.48340411111111115f, 0.36967823529411764f,
-0.9438168692810457f, 0.47776522222222223f, 0.3640394117647059f,
-0.9420226862745098f, 0.47212633333333337f, 0.35840058823529414f,
-0.9402285032679738f, 0.4664874444444445f, 0.3527617647058824f,
-0.9384343202614379f, 0.4608485555555556f, 0.3471229411764706f,
-0.9366401372549019f, 0.4552096666666667f, 0.3414841176470588f,
-0.934845954248366f, 0.4495707777777778f, 0.3358452941176471f,
-0.93305177124183f, 0.4439318888888889f, 0.3302064705882353f,
-0.9312575882352941f, 0.4382930000000001f, 0.3245676470588236f,
-0.9294634052287581f, 0.43265411111111113f, 0.31892882352941176f,
-0.9276692222222223f, 0.42701522222222227f, 0.31329f,
-0.9258750392156863f, 0.42137633333333335f, 0.30765117647058826f,
-0.9240808562091503f, 0.4157374444444445f, 0.3020123529411765f,
-0.9222866732026144f, 0.4100985555555556f, 0.29637352941176476f,
-0.9198773137254902f, 0.404767274509804f, 0.2911960392156863f,
-0.9170578366013071f, 0.3996410653594772f, 0.28632610457516344f,
-0.9142383594771242f, 0.39451485620915044f, 0.2814561699346406f,
-0.9114188823529411f, 0.3893886470588236f, 0.2765862352941177f,
-0.9085994052287582f, 0.38426243790849673f, 0.2717163006535948f,
-0.9057799281045752f, 0.37913622875816994f, 0.26684636601307193f,
-0.9029604509803921f, 0.3740100196078432f, 0.26197643137254906f,
-0.9001409738562092f, 0.36888381045751634f, 0.2571064967320262f,
-0.8973214967320261f, 0.36375760130718954f, 0.2522365620915033f,
-0.8945020196078431f, 0.35863139215686274f, 0.2473666274509804f,
-0.8916825424836602f, 0.35350518300653605f, 0.2424966928104576f,
-0.8888630653594771f, 0.34837897385620914f, 0.23762675816993467f,
-0.886043588235294f, 0.34325276470588234f, 0.2327568235294118f,
-0.8832241111111111f, 0.33812655555555554f, 0.22788688888888892f,
-0.880404633986928f, 0.33300034640522874f, 0.22301695424836604f,
-0.8776620784313726f, 0.3284892745098039f, 0.21891596078431372f,
-0.8750990065359476f, 0.325413522875817f, 0.21660916339869282f,
-0.8725359346405228f, 0.32233777124183005f, 0.2143023660130719f,
-0.869972862745098f, 0.3192620196078432f, 0.21199556862745103f,
-0.8674097908496732f, 0.3161862679738562f, 0.20968877124183008f,
-0.8648467189542484f, 0.3131105163398693f, 0.20738197385620916f,
-0.8622836470588235f, 0.31003476470588237f, 0.20507517647058823f,
-0.8597205751633986f, 0.30695901307189544f, 0.2027683790849673f,
-0.8571575032679738f, 0.3038832614379085f, 0.20046158169934641f,
-0.854594431372549f, 0.30080750980392157f, 0.1981547843137255f,
-0.8520313594771242f, 0.29773175816993463f, 0.19584798692810457f,
-0.8494682875816993f, 0.29465600653594776f, 0.1935411895424837f,
-0.8469052156862745f, 0.2915802549019608f, 0.19123439215686275f,
-0.8443421437908496f, 0.2885045032679739f, 0.18892759477124182f,
-0.8417790718954248f, 0.28542875163398695f, 0.18662079738562093f,
-0.839216f, 0.282353f, 0.184314f,
-0.836716931372549f, 0.28004618627450983f, 0.1825839019607843f,
-0.834217862745098f, 0.27773937254901965f, 0.18085380392156863f,
-0.831718794117647f, 0.2754325588235294f, 0.17912370588235293f,
-0.8292197254901961f, 0.2731257450980393f, 0.1773936078431373f,
-0.8267206568627451f, 0.27081893137254903f, 0.17566350980392156f,
-0.8242215882352941f, 0.26851211764705885f, 0.1739334117647059f,
-0.8217225196078431f, 0.26620530392156866f, 0.1722033137254902f,
-0.8192234509803922f, 0.2638984901960784f, 0.17047321568627452f,
-0.8167243823529412f, 0.26159167647058823f, 0.16874311764705882f,
-0.8142253137254902f, 0.25928486274509804f, 0.16701301960784315f,
-0.8117262450980391f, 0.25697804901960786f, 0.16528292156862745f,
-0.8092271764705883f, 0.25467123529411767f, 0.16355282352941178f,
-0.8067281078431372f, 0.25236442156862743f, 0.16182272549019608f,
-0.8042290392156862f, 0.25005760784313724f, 0.16009262745098038f,
-0.8017299705882353f, 0.24775079411764706f, 0.1583625294117647f,
-0.7992309019607843f, 0.24544398039215687f, 0.156632431372549f,
-0.7967318333333333f, 0.24313716666666665f, 0.15490233333333334f,
-0.7942327647058823f, 0.24083035294117644f, 0.15317223529411764f,
-0.7917336960784314f, 0.23852353921568625f, 0.15144213725490197f,
-0.7892346274509804f, 0.2362167254901961f, 0.1497120392156863f,
-0.7863895490196078f, 0.23314098039215683f, 0.1477896862745098f,
-0.7833137973856209f, 0.22955261437908495f, 0.1457391633986928f,
-0.780238045751634f, 0.22596424836601306f, 0.14368864052287583f,
-0.777162294117647f, 0.22237588235294117f, 0.14163811764705883f,
-0.7740865424836602f, 0.2187875163398693f, 0.13958759477124183f,
-0.7710107908496732f, 0.2151991503267974f, 0.13753707189542486f,
-0.7679350392156863f, 0.2116107843137255f, 0.13548654901960785f,
-0.7648592875816993f, 0.2080224183006536f, 0.13343602614379085f,
-0.7617835359477124f, 0.2044340522875817f, 0.13138550326797388f,
-0.7587077843137255f, 0.20084568627450994f, 0.12933498039215693f,
-0.7556320326797386f, 0.19725732026143794f, 0.12728445751633988f,
-0.7525562810457517f, 0.19366895424836605f, 0.1252339346405229f,
-0.7494805294117647f, 0.19008058823529417f, 0.12318341176470592f,
-0.7464047777777778f, 0.18649222222222228f, 0.12113288888888893f,
-0.7433290261437908f, 0.1829038562091504f, 0.11908236601307193f,
-0.7404070784313725f, 0.1795461764705883f, 0.11726252941176472f,
-0.7378440065359477f, 0.1767267647058824f, 0.11598096078431375f,
-0.7352809346405229f, 0.1739073529411765f, 0.11469939215686277f,
-0.732717862745098f, 0.1710879411764706f, 0.11341782352941178f,
-0.7301547908496732f, 0.16826852941176473f, 0.1121362549019608f,
-0.7275917189542483f, 0.16544911764705886f, 0.11085468627450983f,
-0.7250286470588235f, 0.16262970588235295f, 0.10957311764705883f,
-0.7224655751633987f, 0.15981029411764708f, 0.10829154901960786f,
-0.7199025032679738f, 0.15699088235294117f, 0.10700998039215687f,
-0.717339431372549f, 0.1541714705882353f, 0.10572841176470589f,
-0.7147763594771243f, 0.1513520588235295f, 0.10444684313725494f,
-0.7122132875816993f, 0.14853264705882352f, 0.10316527450980392f,
-0.7096502156862745f, 0.14571323529411764f, 0.10188370588235295f,
-0.7070871437908496f, 0.14289382352941177f, 0.10060213725490197f,
-0.7045240718954248f, 0.14007441176470586f, 0.09932056862745098f,
-0.701961f, 0.137255f, 0.098039f,
-0.699141588235294f, 0.13469186274509803f, 0.09675743137254902f,
-0.6963221764705881f, 0.13212872549019608f, 0.09547586274509805f,
-0.6935027647058823f, 0.12956558823529413f, 0.09419429411764706f,
-0.6906833529411764f, 0.12700245098039215f, 0.09291272549019608f,
-0.6878639411764705f, 0.12443931372549019f, 0.0916311568627451f,
-0.6850445294117646f, 0.12187617647058824f, 0.09034958823529413f,
-0.6822251176470588f, 0.11931303921568628f, 0.08906801960784315f,
-0.6794057058823529f, 0.11674990196078433f, 0.08778645098039217f,
-0.676586294117647f, 0.11418676470588236f, 0.08650488235294118f,
-0.6737668823529411f, 0.11162362745098041f, 0.0852233137254902f,
-0.6709474705882353f, 0.10906049019607852f, 0.08394174509803927f,
-0.6681280588235294f, 0.10649735294117649f, 0.08266017647058825f,
-0.6653086470588235f, 0.10393421568627453f, 0.08137860784313727f,
-0.6624892352941176f, 0.10137107843137258f, 0.08009703921568628f,
-0.6596698235294117f, 0.09880794117647063f, 0.0788154705882353f,
-0.6570298039215686f, 0.09750077777777778f, 0.07878987581699345f,
-0.6544666666666666f, 0.0967318888888889f, 0.07930255555555556f,
-0.6519035294117647f, 0.095963f, 0.07981523529411765f,
-0.6493403921568628f, 0.09519411111111112f, 0.08032791503267973f,
-0.6467772549019608f, 0.09442522222222223f, 0.08084059477124182f,
-0.6442141176470588f, 0.09365633333333334f, 0.08135327450980392f,
-0.6416509803921568f, 0.09288744444444445f, 0.08186595424836601f,
-0.6390878431372549f, 0.09211855555555556f, 0.0823786339869281f,
-0.636524705882353f, 0.09134966666666668f, 0.0828913137254902f,
-0.633961568627451f, 0.09058077777777779f, 0.08340399346405229f,
-0.631398431372549f, 0.08981188888888889f, 0.08391667320261438f,
-0.6288352941176472f, 0.08904300000000002f, 0.08442935294117646f,
-0.6262721568627452f, 0.08827411111111111f, 0.08494203267973857f,
-0.6237090196078432f, 0.08750522222222222f, 0.08545471241830066f,
-0.6211458823529412f, 0.08673633333333333f, 0.08596739215686275f,
-0.6185827450980392f, 0.086480045751634f, 0.08689013725490197f,
-0.6160196078431373f, 0.08699266013071896f, 0.08842798039215688f,
-0.6134564705882354f, 0.08750527450980393f, 0.08996582352941178f,
-0.6108933333333334f, 0.0880178888888889f, 0.09150366666666668f,
-0.6083301960784314f, 0.08853050326797386f, 0.09304150980392158f,
-0.6057670588235294f, 0.08904311764705883f, 0.09457935294117648f,
-0.6032039215686275f, 0.0895557320261438f, 0.09611719607843137f,
-0.6006407843137255f, 0.09006834640522876f, 0.09765503921568627f,
-0.5980776470588236f, 0.09058096078431373f, 0.09919288235294117f,
-0.5955145098039216f, 0.0910935751633987f, 0.10073072549019607f,
-0.5929513725490196f, 0.09160618954248366f, 0.10226856862745098f,
-0.5903882352941177f, 0.09211880392156863f, 0.10380641176470587f,
-0.5878250980392158f, 0.09263141830065358f, 0.10534425490196073f,
-0.5852619607843138f, 0.09314403267973856f, 0.10688209803921568f,
-0.5826988235294118f, 0.09365664705882353f, 0.10841994117647058f,
-0.5801356862745098f, 0.09414362745098039f, 0.10995778431372548f,
-0.5775725490196079f, 0.09439990196078431f, 0.11149562745098038f,
-0.5750094117647059f, 0.09465617647058823f, 0.11303347058823529f,
-0.5724462745098039f, 0.09491245098039215f, 0.11457131372549019f,
-0.569883137254902f, 0.09516872549019607f, 0.1161091568627451f,
-0.56732f, 0.095425f, 0.117647f,
-0.5647568627450981f, 0.09568127450980392f, 0.1191848431372549f,
-0.5621937254901961f, 0.09593754901960784f, 0.12072268627450981f,
-0.5596305882352941f, 0.09619382352941176f, 0.12226052941176471f,
-0.5570674509803921f, 0.09645009803921568f, 0.12379837254901963f,
-0.5545043137254901f, 0.0967063725490196f, 0.1253362156862745f,
-0.5519411764705883f, 0.09696264705882353f, 0.12687405882352942f,
-0.5493780392156863f, 0.09721892156862745f, 0.12841190196078434f,
-0.5468149019607844f, 0.09747519607843137f, 0.1299497450980392f,
-0.5442517647058823f, 0.0977314705882353f, 0.13148758823529413f,
-0.5416886274509803f, 0.09798774509803922f, 0.13302543137254905f,
-0.5387154509803921f, 0.098039f, 0.13456330588235296f,
-0.5356397647058824f, 0.098039f, 0.13610118823529413f,
-0.5325640784313725f, 0.098039f, 0.1376390705882353f,
-0.5294883921568627f, 0.098039f, 0.1391769529411765f,
-0.5264127058823529f, 0.098039f, 0.14071483529411766f,
-0.5233370196078432f, 0.098039f, 0.14225271764705882f,
-0.5202613333333334f, 0.098039f, 0.1437906f,
-0.5171856470588235f, 0.098039f, 0.14532848235294116f,
-0.5141099607843137f, 0.098039f, 0.14686636470588232f,
-0.511034274509804f, 0.098039f, 0.14840424705882352f,
-0.5079585882352942f, 0.098039f, 0.14994212941176469f,
-0.5048829019607843f, 0.098039f, 0.15148001176470585f,
-0.5017918352941177f, 0.09805438039215686f, 0.15297175686274508f,
-0.4984085411764707f, 0.0983619882352941f, 0.15358689411764703f,
-0.49502524705882356f, 0.09866959607843137f, 0.154202031372549f,
-0.4916419529411765f, 0.09897720392156863f, 0.15481716862745099f,
-0.4882586588235294f, 0.09928481176470588f, 0.15543230588235293f,
-0.4848753647058824f, 0.09959241960784314f, 0.1560474431372549f,
-0.4814920705882353f, 0.0999000274509804f, 0.15666258039215686f,
-0.47810877647058825f, 0.10020763529411765f, 0.15727771764705883f,
-0.47472548235294115f, 0.1005152431372549f, 0.15789285490196078f,
-0.4713421882352941f, 0.10082285098039216f, 0.15850799215686276f,
-0.467958894117647f, 0.1011304588235294f, 0.1591231294117647f,
-0.4645756f, 0.10143806666666666f, 0.15973826666666668f,
-0.46119230588235294f, 0.10174567450980392f, 0.16035340392156863f,
-0.45805505882352937f, 0.10211478431372549f, 0.16109158823529413f,
-0.45549192156862744f, 0.10262739869281046f, 0.1621168823529412f,
-0.45292878431372546f, 0.10314001307189542f, 0.16314217647058826f,
-0.45036564705882354f, 0.10365262745098039f, 0.1641674705882353f,
-0.4478025098039216f, 0.10416524183006534f, 0.16519276470588234f,
-0.4452393725490196f, 0.10467785620915032f, 0.1662180588235294f,
-0.44267623529411765f, 0.10519047058823529f, 0.16724335294117648f,
-0.44011309803921567f, 0.10570308496732027f, 0.16826864705882352f,
-0.4375499607843137f, 0.10621569934640523f, 0.1692939411764706f,
-0.43498682352941176f, 0.1067283137254902f, 0.17031923529411766f,
-0.4324236862745098f, 0.10724092810457517f, 0.1713445294117647f,
-0.42986054901960785f, 0.10775354248366013f, 0.17236982352941177f,
-0.42729741176470587f, 0.1082661568627451f, 0.1733951176470588f,
-0.4247342745098039f, 0.10877877124183007f, 0.17442041176470588f,
-0.42217113725490196f, 0.10929138562091503f, 0.17544570588235292f,
-0.419608f, 0.109804f, 0.176471f
-}
-, 
-{
-1.0f, 0.972549f, 0.94902f,
-0.9953863529411765f, 0.9642445294117648f, 0.9394851764705883f,
-0.9907727058823529f, 0.9559400588235294f, 0.9299503529411765f,
-0.9861590588235294f, 0.9476355882352941f, 0.9204155294117647f,
-0.9815454117647059f, 0.9393311176470588f, 0.9108807058823529f,
-0.9769317647058824f, 0.9310266470588235f, 0.9013458823529412f,
-0.9723181176470588f, 0.9227221764705882f, 0.8918110588235294f,
-0.9677044705882353f, 0.9144177058823529f, 0.8822762352941177f,
-0.9630908235294118f, 0.9061132352941176f, 0.8727414117647059f,
-0.9584771764705883f, 0.8978087647058823f, 0.8632065882352942f,
-0.9538635294117647f, 0.889504294117647f, 0.8536717647058824f,
-0.9492498823529412f, 0.8811998235294117f, 0.8441369411764706f,
-0.9446362352941177f, 0.8728953529411765f, 0.8346021176470588f,
-0.9399457058823529f, 0.864590862745098f, 0.8250672941176471f,
-0.9350245294117647f, 0.8562863137254901f, 0.8155324705882353f,
-0.9301033529411765f, 0.8479817647058824f, 0.8059976470588236f,
-0.9251821764705882f, 0.8396772156862745f, 0.7964628235294118f,
-0.920261f, 0.8313726666666666f, 0.7869280000000001f,
-0.9153398235294118f, 0.8230681176470588f, 0.7773931764705883f,
-0.9104186470588235f, 0.814763568627451f, 0.7678583529411765f,
-0.9054974705882353f, 0.8064590196078432f, 0.7583235294117647f,
-0.9005762941176471f, 0.7981544705882353f, 0.7487887058823529f,
-0.8956551176470589f, 0.7898499215686274f, 0.7392538823529412f,
-0.8907339411764705f, 0.7815453725490196f, 0.7297190588235294f,
-0.8858127647058823f, 0.7732408235294118f, 0.7201842352941177f,
-0.8808915882352941f, 0.764936274509804f, 0.7106494117647059f,
-0.8761242156862745f, 0.757400705882353f, 0.7026524705882353f,
-0.8715106470588235f, 0.7506341176470589f, 0.6961934117647058f,
-0.8668970784313725f, 0.7438675294117647f, 0.6897343529411765f,
-0.8622835098039215f, 0.7371009411764706f, 0.6832752941176471f,
-0.8576699411764706f, 0.7303343529411765f, 0.6768162352941176f,
-0.8530563725490196f, 0.7235677647058824f, 0.6703571764705882f,
-0.8484428039215686f, 0.7168011764705883f, 0.6638981176470589f,
-0.8438292352941177f, 0.7100345882352942f, 0.6574390588235295f,
-0.8392156666666667f, 0.703268f, 0.65098f,
-0.8346020980392157f, 0.6965014117647059f, 0.6445209411764706f,
-0.8299885294117647f, 0.6897348235294118f, 0.6380618823529411f,
-0.8253749607843137f, 0.6829682352941178f, 0.6316028235294118f,
-0.8207613921568627f, 0.6762016470588236f, 0.6251437647058824f,
-0.8173011764705882f, 0.670588411764706f, 0.6198381176470589f,
-0.8142254117647059f, 0.6653596274509804f, 0.6149169411764706f,
-0.8111496470588235f, 0.660130843137255f, 0.6099957647058823f,
-0.8080738823529412f, 0.6549020588235295f, 0.6050745882352941f,
-0.8049981176470589f, 0.6496732745098039f, 0.6001534117647058f,
-0.8019223529411764f, 0.6444444901960785f, 0.5952322352941176f,
-0.7988465882352941f, 0.639215705882353f, 0.5903110588235294f,
-0.7957708235294118f, 0.6339869215686275f, 0.5853898823529411f,
-0.7926950588235294f, 0.628758137254902f, 0.5804687058823529f,
-0.7896192941176471f, 0.6235293529411765f, 0.5755475294117647f,
-0.7865435294117648f, 0.6183005686274511f, 0.5706263529411765f,
-0.7834677647058823f, 0.6130717843137256f, 0.5657051764705882f,
-0.780392f, 0.607843f, 0.560784f,
-0.7773162352941176f, 0.6029218235294118f, 0.556170431372549f,
-0.7742404705882353f, 0.5980006470588236f, 0.5515568627450981f,
-0.7711647058823529f, 0.5930794705882353f, 0.546943294117647f,
-0.7680889411764705f, 0.5881582941176471f, 0.5423297254901961f,
-0.7650131764705882f, 0.5832371176470589f, 0.5377161568627451f,
-0.7619374117647059f, 0.5783159411764707f, 0.5331025882352941f,
-0.7588616470588235f, 0.5733947647058824f, 0.5284890196078431f,
-0.7557858823529411f, 0.5684735882352941f, 0.5238754509803921f,
-0.7527101176470588f, 0.5635524117647058f, 0.5192618823529411f,
-0.7496343529411764f, 0.5586312352941176f, 0.5146483137254902f,
-0.7465585882352941f, 0.5537100588235294f, 0.5100347450980391f,
-0.7434828235294118f, 0.5487888823529412f, 0.5054211764705883f,
-0.740560862745098f, 0.5440983921568627f, 0.5010382745098039f,
-0.7381003137254901f, 0.5400999607843137f, 0.4973473725490196f,
-0.7356397647058823f, 0.5361015294117647f, 0.49365647058823536f,
-0.7331792156862744f, 0.5321030980392156f, 0.48996556862745094f,
-0.7307186666666666f, 0.5281046666666667f, 0.48627466666666663f,
-0.7282581176470588f, 0.5241062352941176f, 0.4825837647058823f,
-0.7257975686274509f, 0.5201078039215686f, 0.478892862745098f,
-0.7233370196078431f, 0.5161093725490196f, 0.4752019607843137f,
-0.7208764705882352f, 0.5121109411764706f, 0.4715110588235294f,
-0.7184159215686274f, 0.5081125098039215f, 0.4678201568627451f,
-0.7159553725490196f, 0.5041140784313726f, 0.46412925490196083f,
-0.7134948235294117f, 0.5001156470588235f, 0.46043835294117647f,
-0.7110342745098039f, 0.49611721568627454f, 0.45674745098039216f,
-0.7082661176470588f, 0.49196498039215686f, 0.4530565490196078f,
-0.7051903529411765f, 0.4876589411764706f, 0.4493656470588235f,
-0.7021145882352942f, 0.4833529019607843f, 0.44567474509803917f,
-0.6990388235294117f, 0.47904686274509806f, 0.44198384313725486f,
-0.6959630588235294f, 0.47474082352941177f, 0.43829294117647055f,
-0.692887294117647f, 0.47043478431372554f, 0.4346020392156863f,
-0.6898115294117647f, 0.4661287450980392f, 0.43091113725490193f,
-0.6867357647058824f, 0.46182270588235297f, 0.4272202352941176f,
-0.6836599999999999f, 0.4575166666666667f, 0.4235293333333333f,
-0.6805842352941176f, 0.4532106274509804f, 0.419838431372549f,
-0.6775084705882353f, 0.4489045882352941f, 0.4161475294117647f,
-0.6744327058823529f, 0.4445985490196078f, 0.4124566274509804f,
-0.6713569411764706f, 0.44029250980392154f, 0.4087657254901961f,
-0.6687425882352941f, 0.4369091764705883f, 0.40599758823529414f,
-0.6662820392156863f, 0.43383341176470586f, 0.4035370392156863f,
-0.6638214901960784f, 0.4307576470588235f, 0.40107649019607844f,
-0.6613609411764706f, 0.4276818823529412f, 0.3986159411764706f,
-0.6589003921568627f, 0.4246061176470588f, 0.39615539215686274f,
-0.6564398431372549f, 0.42153035294117647f, 0.3936948431372549f,
-0.653979294117647f, 0.41845458823529413f, 0.39123429411764704f,
-0.6515187450980392f, 0.4153788235294118f, 0.38877374509803925f,
-0.6490581960784314f, 0.41230305882352947f, 0.3863131960784314f,
-0.6465976470588235f, 0.4092272941176471f, 0.38385264705882355f,
-0.6441370980392157f, 0.40615152941176474f, 0.3813920980392157f,
-0.6416765490196078f, 0.40307576470588236f, 0.37893154901960785f,
-0.639216f, 0.4f, 0.376471f,
-0.6370629803921569f, 0.3972318431372549f, 0.37462550980392156f,
-0.6349099607843137f, 0.39446368627450984f, 0.3727800196078431f,
-0.6327569411764706f, 0.3916955294117647f, 0.3709345294117647f,
-0.6306039215686274f, 0.38892737254901966f, 0.3690890392156863f,
-0.6284509019607843f, 0.38615921568627454f, 0.3672435490196079f,
-0.6262978823529411f, 0.3833910588235294f, 0.36539805882352944f,
-0.6241448627450981f, 0.3806229019607843f, 0.363552568627451f,
-0.6219918431372549f, 0.3778547450980392f, 0.36170707843137256f,
-0.6198388235294118f, 0.3750865882352941f, 0.3598615882352941f,
-0.6176858039215687f, 0.372318431372549f, 0.3580160980392157f,
-0.6155327843137255f, 0.3695502745098039f, 0.35617060784313725f,
-0.6133797647058824f, 0.3667821176470588f, 0.3543251176470589f,
-0.611149843137255f, 0.36409084313725487f, 0.3524796470588235f,
-0.6086892156862745f, 0.36163021568627446f, 0.35063423529411764f,
-0.6062285882352941f, 0.3591695882352941f, 0.34878882352941176f,
-0.6037679607843137f, 0.3567089607843137f, 0.3469434117647059f,
-0.6013073333333333f, 0.35424833333333333f, 0.345098f,
-0.598846705882353f, 0.3517877058823529f, 0.3432525882352941f,
-0.5963860784313726f, 0.3493270784313725f, 0.3414071764705882f,
-0.5939254509803922f, 0.34686645098039215f, 0.3395617647058824f,
-0.5914648235294118f, 0.34440582352941174f, 0.33771635294117647f,
-0.5890041960784314f, 0.3419451960784314f, 0.3358709411764706f,
-0.5865435686274509f, 0.33948456862745097f, 0.33402552941176467f,
-0.5840829411764706f, 0.33702394117647055f, 0.3321801176470588f,
-0.5816223137254902f, 0.3345633137254902f, 0.3303347058823529f,
-0.5791617254901961f, 0.3321027254901961f, 0.3286430588235294f,
-0.5767011764705883f, 0.32964217647058824f, 0.32710517647058823f,
-0.5742406274509804f, 0.3271816274509804f, 0.32556729411764707f,
-0.5717800784313726f, 0.32472107843137255f, 0.3240294117647059f,
-0.5693195294117648f, 0.32226052941176475f, 0.32249152941176473f,
-0.5668589803921569f, 0.31979998039215685f, 0.3209536470588235f,
-0.564398431372549f, 0.317339431372549f, 0.31941576470588234f,
-0.5619378823529412f, 0.31487888235294115f, 0.3178778823529412f,
-0.5594773333333333f, 0.3124183333333333f, 0.31634f,
-0.5570167843137255f, 0.30995778431372545f, 0.31480211764705884f,
-0.5545562352941176f, 0.30749723529411765f, 0.3132642352941177f,
-0.5520956862745098f, 0.3050366862745098f, 0.3117263529411765f,
-0.5496351372549019f, 0.30257613725490196f, 0.31018847058823534f,
-0.5474052352941176f, 0.30034623529411764f, 0.3086505882352941f,
-0.5452522156862745f, 0.2981932156862745f, 0.30711270588235295f,
-0.5430991960784314f, 0.29604019607843135f, 0.3055748235294118f,
-0.5409461764705882f, 0.2938871764705882f, 0.3040369411764706f,
-0.5387931568627451f, 0.29173415686274506f, 0.30249905882352945f,
-0.5366401372549019f, 0.2895811372549019f, 0.30096117647058823f,
-0.5344871176470588f, 0.2874281176470588f, 0.29942329411764707f,
-0.5323340980392157f, 0.28527509803921575f, 0.29788541176470595f,
-0.5301810784313725f, 0.28312207843137255f, 0.29634752941176473f,
-0.5280280588235293f, 0.2809690588235294f, 0.29480964705882357f,
-0.5258750392156862f, 0.27881603921568626f, 0.29327176470588234f,
-0.523722019607843f, 0.2766630196078431f, 0.2917338823529412f,
-0.521569f, 0.27451f, 0.290196f,
-0.5191083725490195f, 0.27266450980392154f, 0.2889657254901961f,
-0.5166477450980391f, 0.2708190196078431f, 0.28773545098039216f,
-0.5141871176470588f, 0.26897352941176467f, 0.28650517647058826f,
-0.5117264901960784f, 0.26712803921568623f, 0.2852749019607843f,
-0.509265862745098f, 0.2652825490196078f, 0.2840446274509804f,
-0.5068052352941176f, 0.2634370588235294f, 0.28281435294117646f,
-0.5043446078431373f, 0.261591568627451f, 0.28158407843137256f,
-0.5018839803921569f, 0.25974607843137254f, 0.2803538039215686f,
-0.49942335294117646f, 0.2579005882352941f, 0.2791235294117647f,
-0.4969627254901961f, 0.25605509803921567f, 0.27789325490196076f,
-0.49450209803921574f, 0.2542096078431373f, 0.27666298039215687f,
-0.49204147058823533f, 0.2523641176470588f, 0.2754327058823529f,
-0.48958086274509804f, 0.2504417450980392f, 0.27412552941176466f,
-0.4871203137254902f, 0.24828872549019607f, 0.2725876470588235f,
-0.48465976470588235f, 0.24613570588235292f, 0.2710497647058823f,
-0.48219921568627455f, 0.24398268627450978f, 0.26951188235294116f,
-0.4797386666666667f, 0.24182966666666666f, 0.267974f,
-0.47727811764705885f, 0.23967664705882352f, 0.2664361176470588f,
-0.474817568627451f, 0.23752362745098038f, 0.26489823529411766f,
-0.47235701960784315f, 0.23537060784313724f, 0.2633603529411765f,
-0.4698964705882353f, 0.23321758823529412f, 0.2618224705882353f,
-0.46743592156862745f, 0.23106456862745098f, 0.2602845882352941f,
-0.4649753725490196f, 0.22891154901960784f, 0.25874670588235293f,
-0.46251482352941176f, 0.22675852941176472f, 0.25720882352941177f,
-0.4600542745098039f, 0.22460550980392158f, 0.2556709411764706f,
-0.45774749019607847f, 0.2227600980392157f, 0.25428686274509804f,
-0.4555944705882354f, 0.2212222941176471f, 0.25305658823529414f,
-0.4534414509803922f, 0.21968449019607844f, 0.2518263137254902f,
-0.45128843137254904f, 0.2181466862745098f, 0.2505960392156863f,
-0.4491354117647059f, 0.21660888235294118f, 0.24936576470588237f,
-0.44698239215686275f, 0.21507107843137255f, 0.24813549019607845f,
-0.4448293725490196f, 0.21353327450980392f, 0.24690521568627452f,
-0.44267635294117647f, 0.21199547058823529f, 0.2456749411764706f,
-0.4405233333333333f, 0.21045766666666665f, 0.24444466666666667f,
-0.4383703137254902f, 0.20891986274509802f, 0.24321439215686275f,
-0.4362172941176471f, 0.2073820588235294f, 0.24198411764705885f,
-0.43406427450980395f, 0.2058442549019608f, 0.24075384313725492f,
-0.4319112549019608f, 0.20430645098039216f, 0.239523568627451f,
-0.4295275294117647f, 0.20253788235294118f, 0.23829323529411767f,
-0.42706690196078434f, 0.20069239215686274f, 0.23706288235294118f,
-0.4246062745098039f, 0.1988469019607843f, 0.23583252941176472f,
-0.4221456470588235f, 0.19700141176470587f, 0.23460217647058826f,
-0.4196850196078432f, 0.1951559215686275f, 0.2333718235294118f,
-0.41722439215686274f, 0.19331043137254902f, 0.2321414705882353f,
-0.4147637647058824f, 0.1914649411764706f, 0.23091111764705882f,
-0.412303137254902f, 0.18961945098039215f, 0.22968076470588236f,
-0.40984250980392156f, 0.18777396078431371f, 0.2284504117647059f,
-0.4073818823529412f, 0.1859284705882353f, 0.2272200588235294f,
-0.4049212549019608f, 0.18408298039215687f, 0.22598970588235295f,
-0.40246062745098043f, 0.18223749019607843f, 0.22475935294117647f,
-0.4f, 0.180392f, 0.223529f,
-0.3975393725490196f, 0.17854658823529412f, 0.22199119607843137f,
-0.39507874509803925f, 0.17670117647058822f, 0.22045339215686274f,
-0.39261811764705884f, 0.17485576470588235f, 0.2189155882352941f,
-0.3901574901960784f, 0.17301035294117648f, 0.21737778431372548f,
-0.38769686274509807f, 0.17116494117647058f, 0.21583998039215685f,
-0.38523623529411766f, 0.1693195294117647f, 0.21430217647058825f,
-0.38277560784313724f, 0.16747411764705883f, 0.21276437254901961f,
-0.38031498039215694f, 0.16562870588235298f, 0.211226568627451f,
-0.3778543529411765f, 0.16378329411764705f, 0.20968876470588235f,
-0.37539372549019606f, 0.16193788235294118f, 0.20815096078431372f,
-0.37293309803921565f, 0.16009247058823528f, 0.2066131568627451f,
-0.3704724705882353f, 0.1582470588235294f, 0.20507535294117646f,
-0.3678580784313725f, 0.1564016274509804f, 0.2035375294117647f,
-0.3647823921568627f, 0.15455613725490197f, 0.20199964705882353f,
-0.36170670588235293f, 0.15271064705882353f, 0.20046176470588234f,
-0.3586310196078431f, 0.1508651568627451f, 0.19892388235294117f,
-0.35555533333333333f, 0.14901966666666666f, 0.197386f,
-0.3524796470588235f, 0.14717417647058825f, 0.1958481176470588f,
-0.3494039607843137f, 0.14532868627450982f, 0.19431023529411764f,
-0.3463282745098039f, 0.14348319607843138f, 0.19277235294117648f,
-0.3432525882352941f, 0.14163770588235294f, 0.19123447058823528f,
-0.3401769019607843f, 0.1397922156862745f, 0.18969658823529412f,
-0.3371012156862745f, 0.1379467254901961f, 0.18815870588235295f,
-0.3340255294117648f, 0.13610123529411772f, 0.1866208235294118f,
-0.3309498431372549f, 0.13425574509803923f, 0.1850829411764706f,
-0.3278741176470588f, 0.13225649019607844f, 0.1833912549019608f,
-0.3247983529411765f, 0.1301034705882353f, 0.18154576470588235f,
-0.3217225882352941f, 0.12795045098039215f, 0.17970027450980391f,
-0.31864682352941176f, 0.12579743137254903f, 0.1778547843137255f,
-0.3155710588235294f, 0.12364441176470589f, 0.17600929411764707f,
-0.31249529411764704f, 0.12149139215686275f, 0.17416380392156863f,
-0.3094195294117647f, 0.11933837254901962f, 0.1723183137254902f,
-0.30634376470588237f, 0.11718535294117648f, 0.17047282352941176f,
-0.303268f, 0.11503233333333335f, 0.16862733333333335f,
-0.30019223529411765f, 0.1128793137254902f, 0.16678184313725491f,
-0.2971164705882353f, 0.11072629411764706f, 0.16493635294117648f,
-0.2940407058823529f, 0.10857327450980392f, 0.16309086274509804f,
-0.2909649411764706f, 0.10642025490196079f, 0.1612453725490196f,
-0.28788917647058826f, 0.10426723529411765f, 0.15778517647058823f,
-0.284813411764706f, 0.10211421568627457f, 0.15378674509803933f,
-0.28173764705882354f, 0.09996119607843137f, 0.1497883137254902f,
-0.2786618823529412f, 0.09780817647058823f, 0.14578988235294119f,
-0.2755861176470588f, 0.0956551568627451f, 0.14179145098039217f,
-0.2725103529411765f, 0.09350213725490196f, 0.13779301960784315f,
-0.2694345882352941f, 0.09134911764705883f, 0.1337945882352941f,
-0.26635882352941176f, 0.08919609803921569f, 0.1297961568627451f,
-0.2632830588235294f, 0.08704307843137254f, 0.12579772549019608f,
-0.26020729411764704f, 0.08489005882352942f, 0.12179929411764706f,
-0.2571315294117647f, 0.08273703921568627f, 0.11780086274509805f,
-0.2540557647058823f, 0.08058401960784314f, 0.11380243137254903f,
-0.25098f, 0.078431f, 0.109804f
-}
-, 
-{
-1.0f, 0.996078f, 0.890196f,
-0.9995386274509804f, 0.9950014901960784f, 0.8832755686274509f,
-0.9990772549019608f, 0.9939249803921569f, 0.876355137254902f,
-0.9986158823529412f, 0.9928484705882353f, 0.8694347058823529f,
-0.9981545098039216f, 0.9917719607843137f, 0.8625142745098039f,
-0.9976931372549019f, 0.9906954509803921f, 0.855593843137255f,
-0.9972317647058824f, 0.9896189411764706f, 0.8486734117647059f,
-0.9967703921568627f, 0.988542431372549f, 0.8417529803921568f,
-0.9963090196078431f, 0.9874659215686274f, 0.8348325490196078f,
-0.9958476470588236f, 0.9863894117647058f, 0.8279121176470589f,
-0.9953862745098039f, 0.9853129019607844f, 0.8209916862745098f,
-0.9949249019607843f, 0.9842363921568628f, 0.8140712549019608f,
-0.9944635294117647f, 0.9831598823529412f, 0.8071508235294118f,
-0.9940021568627451f, 0.9820833725490197f, 0.8002303921568628f,
-0.9935407843137255f, 0.9810068627450981f, 0.7933099607843137f,
-0.9930794117647058f, 0.9799303529411765f, 0.7863895294117648f,
-0.9926180392156863f, 0.9788538431372549f, 0.7794690980392157f,
-0.9921566666666667f, 0.9777773333333334f, 0.7725486666666667f,
-0.991695294117647f, 0.9767008235294118f, 0.7656282352941177f,
-0.9912339215686274f, 0.9756243137254902f, 0.7587078039215687f,
-0.9907725490196079f, 0.9745478039215687f, 0.7517873725490196f,
-0.9903111764705882f, 0.9734712941176471f, 0.7448669411764707f,
-0.9898498039215686f, 0.9723947843137255f, 0.7379465098039216f,
-0.989388431372549f, 0.9713182745098039f, 0.7310260784313726f,
-0.9889270588235294f, 0.9702417647058824f, 0.7241056470588236f,
-0.9884656862745098f, 0.9691652549019608f, 0.7171852156862746f,
-0.9880812156862745f, 0.9679349607843137f, 0.7112644117647059f,
-0.9877736470588235f, 0.9665508823529412f, 0.7063432352941177f,
-0.9874660784313726f, 0.9651668039215686f, 0.7014220588235295f,
-0.9871585098039215f, 0.9637827254901961f, 0.6965008823529412f,
-0.9868509411764705f, 0.9623986470588235f, 0.691579705882353f,
-0.9865433725490196f, 0.9610145686274509f, 0.6866585294117648f,
-0.9862358039215686f, 0.9596304901960785f, 0.6817373529411765f,
-0.9859282352941177f, 0.9582464117647059f, 0.6768161764705883f,
-0.9856206666666667f, 0.9568623333333334f, 0.671895f,
-0.9853130980392157f, 0.9554782549019608f, 0.6669738235294118f,
-0.9850055294117647f, 0.9540941764705883f, 0.6620526470588235f,
-0.9846979607843137f, 0.9527100980392157f, 0.6571314705882353f,
-0.9843903921568627f, 0.9513260196078431f, 0.6522102941176471f,
-0.9840828235294118f, 0.9499419411764706f, 0.6472891176470588f,
-0.9837752549019608f, 0.948557862745098f, 0.6423679411764706f,
-0.9834676862745099f, 0.9471737843137255f, 0.6374467647058824f,
-0.9831601176470588f, 0.9457897058823529f, 0.6325255882352941f,
-0.9828525490196078f, 0.9444056274509803f, 0.6276044117647058f,
-0.9825449803921569f, 0.9430215490196078f, 0.6226832352941176f,
-0.9822374117647059f, 0.9416374705882353f, 0.6177620588235294f,
-0.981929843137255f, 0.9402533921568628f, 0.6128408823529411f,
-0.981622274509804f, 0.9388693137254902f, 0.6079197058823529f,
-0.9813147058823529f, 0.9374852352941176f, 0.6029985294117647f,
-0.981007137254902f, 0.9361011568627451f, 0.5980773529411765f,
-0.980699568627451f, 0.9347170784313725f, 0.5931561764705882f,
-0.980392f, 0.933333f, 0.588235f,
-0.9799306274509805f, 0.932102725490196f, 0.5850054705882353f,
-0.9794692549019608f, 0.9308724509803922f, 0.5817759411764706f,
-0.9790078823529412f, 0.9296421764705882f, 0.5785464117647059f,
-0.9785465098039217f, 0.9284119019607843f, 0.5753168823529411f,
-0.978085137254902f, 0.9271816274509803f, 0.5720873529411764f,
-0.9776237647058824f, 0.9259513529411765f, 0.5688578235294117f,
-0.9771623921568627f, 0.9247210784313725f, 0.565628294117647f,
-0.9767010196078432f, 0.9234908039215686f, 0.5623987647058823f,
-0.9762396470588236f, 0.9222605294117647f, 0.5591692352941177f,
-0.9757782745098039f, 0.9210302549019608f, 0.555939705882353f,
-0.9753169019607844f, 0.9197999803921568f, 0.5527101764705882f,
-0.9748555294117648f, 0.918569705882353f, 0.5494806470588235f,
-0.9743941568627451f, 0.917339431372549f, 0.5462511176470588f,
-0.9739327843137255f, 0.9161091568627451f, 0.5430215882352941f,
-0.9734714117647059f, 0.9148788823529412f, 0.5397920588235294f,
-0.9730100392156863f, 0.9136486078431373f, 0.5365625294117647f,
-0.9725486666666667f, 0.9124183333333333f, 0.5333330000000001f,
-0.972087294117647f, 0.9111880588235294f, 0.5301034705882353f,
-0.9716259215686275f, 0.9099577843137255f, 0.5268739411764706f,
-0.9711645490196079f, 0.9087275098039216f, 0.5236444117647059f,
-0.9707031764705882f, 0.9074972352941176f, 0.5204148823529412f,
-0.9702418039215687f, 0.9062669607843137f, 0.5171853529411765f,
-0.969780431372549f, 0.9050366862745098f, 0.5139558235294118f,
-0.9693190588235294f, 0.9038064117647059f, 0.5107262941176471f,
-0.9688576862745099f, 0.9025761372549019f, 0.5074967647058825f,
-0.9684732156862745f, 0.9011920588235294f, 0.5042672352941177f,
-0.9681656470588236f, 0.8996541764705882f, 0.501037705882353f,
-0.9678580784313726f, 0.898116294117647f, 0.4978081764705883f,
-0.9675505098039215f, 0.8965784117647059f, 0.4945786470588236f,
-0.9672429411764706f, 0.8950405294117647f, 0.49134911764705885f,
-0.9669353725490196f, 0.8935026470588235f, 0.4881195882352942f,
-0.9666278039215687f, 0.8919647647058824f, 0.4848900588235294f,
-0.9663202352941177f, 0.8904268823529412f, 0.48166052941176474f,
-0.9660126666666666f, 0.888889f, 0.47843100000000005f,
-0.9657050980392157f, 0.8873511176470589f, 0.4752014705882353f,
-0.9653975294117647f, 0.8858132352941176f, 0.4719719411764706f,
-0.9650899607843137f, 0.8842753529411764f, 0.4687424117647059f,
-0.9647823921568628f, 0.8827374705882353f, 0.4655128823529412f,
-0.9644748235294117f, 0.8811995882352941f, 0.4622833529411765f,
-0.9641672549019608f, 0.8796617058823529f, 0.45905382352941176f,
-0.9638596862745098f, 0.8781238235294118f, 0.4558242941176471f,
-0.9635521176470588f, 0.8765859411764706f, 0.45259476470588234f,
-0.9632445490196079f, 0.8750480588235294f, 0.44936523529411765f,
-0.9629369803921568f, 0.8735101764705883f, 0.44613570588235296f,
-0.9626294117647058f, 0.8719722941176471f, 0.4429061764705883f,
-0.9623218431372549f, 0.8704344117647058f, 0.43967664705882353f,
-0.9620142745098039f, 0.8688965294117648f, 0.4364471176470589f,
-0.961706705882353f, 0.8673586470588235f, 0.4332175882352941f,
-0.9613991372549019f, 0.8658207647058823f, 0.4299880588235294f,
-0.9610915686274509f, 0.8642828823529412f, 0.4267585294117647f,
-0.960784f, 0.862745f, 0.423529f,
-0.9603226666666667f, 0.861514725490196f, 0.4210684117647059f,
-0.9598613333333333f, 0.8602844509803922f, 0.4186078235294117f,
-0.9593999999999999f, 0.8590541764705882f, 0.4161472352941176f,
-0.9589386666666666f, 0.8578239019607843f, 0.4136866470588236f,
-0.9584773333333333f, 0.8565936274509803f, 0.4112260588235294f,
-0.958016f, 0.8553633529411765f, 0.40876547058823526f,
-0.9575546666666667f, 0.8541330784313725f, 0.40630488235294115f,
-0.9570933333333334f, 0.8529028039215686f, 0.40384429411764705f,
-0.9566319999999999f, 0.8516725294117647f, 0.40138370588235295f,
-0.9561706666666666f, 0.8504422549019608f, 0.3989231176470588f,
-0.9557093333333333f, 0.8492119803921568f, 0.3964625294117647f,
-0.955248f, 0.847981705882353f, 0.39400194117647064f,
-0.9547866666666667f, 0.846751431372549f, 0.3915413529411765f,
-0.9543253333333334f, 0.8455211568627451f, 0.3890807647058823f,
-0.9538639999999999f, 0.8442908823529411f, 0.3866201764705882f,
-0.9534026666666666f, 0.8430606078431373f, 0.3841595882352941f,
-0.9529413333333333f, 0.8418303333333333f, 0.381699f,
-0.95248f, 0.8406000588235294f, 0.37923841176470585f,
-0.9520186666666667f, 0.8393697843137256f, 0.37677782352941175f,
-0.9515573333333334f, 0.8381395098039216f, 0.37431723529411765f,
-0.9510959999999999f, 0.8369092352941176f, 0.37185664705882354f,
-0.9506346666666666f, 0.8356789607843137f, 0.3693960588235294f,
-0.9501733333333333f, 0.8344486862745099f, 0.3669354705882353f,
-0.949712f, 0.8332184117647059f, 0.3644748823529412f,
-0.9492506666666667f, 0.831988137254902f, 0.3620142941176471f,
-0.9488661960784314f, 0.830757843137255f, 0.3596306078431373f,
-0.948558588235294f, 0.8295275294117648f, 0.3573238235294118f,
-0.9482509803921568f, 0.8282972156862746f, 0.3550170392156863f,
-0.9479433725490196f, 0.8270669019607844f, 0.3527102549019608f,
-0.9476357647058823f, 0.8258365882352942f, 0.35040347058823534f,
-0.9473281568627451f, 0.824606274509804f, 0.3480966862745098f,
-0.9470205490196079f, 0.8233759607843137f, 0.3457899019607843f,
-0.9467129411764705f, 0.8221456470588235f, 0.3434831176470588f,
-0.9464053333333333f, 0.8209153333333333f, 0.3411763333333333f,
-0.9460977254901961f, 0.8196850196078431f, 0.3388695490196078f,
-0.9457901176470588f, 0.818454705882353f, 0.33656276470588237f,
-0.9454825098039216f, 0.8172243921568628f, 0.33425598039215687f,
-0.9451749019607844f, 0.8159940784313726f, 0.3319491960784314f,
-0.944867294117647f, 0.8147637647058824f, 0.3296424117647059f,
-0.9445596862745098f, 0.8135334509803922f, 0.3273356274509804f,
-0.9442520784313726f, 0.812303137254902f, 0.3250288431372549f,
-0.9439444705882353f, 0.8110728235294118f, 0.3227220588235294f,
-0.9436368627450981f, 0.8098425098039216f, 0.3204152745098039f,
-0.9433292549019607f, 0.8086121960784314f, 0.3181084901960784f,
-0.9430216470588235f, 0.8073818823529412f, 0.3158017058823529f,
-0.9427140392156863f, 0.8061515686274511f, 0.3134949215686275f,
-0.942406431372549f, 0.8049212549019609f, 0.31118813725490196f,
-0.9420988235294118f, 0.8036909411764707f, 0.30888135294117647f,
-0.9417912156862746f, 0.8024606274509805f, 0.30657456862745097f,
-0.9414836078431372f, 0.8012303137254903f, 0.3042677843137255f,
-0.941176f, 0.8f, 0.301961f,
-0.9407146666666667f, 0.7987696862745098f, 0.29965417647058823f,
-0.9402533333333334f, 0.7975393725490196f, 0.2973473529411764f,
-0.939792f, 0.7963090588235294f, 0.2950405294117647f,
-0.9393306666666666f, 0.7950787450980392f, 0.2927337058823529f,
-0.9388693333333333f, 0.793848431372549f, 0.2904268823529412f,
-0.938408f, 0.7926181176470588f, 0.28812005882352937f,
-0.9379466666666667f, 0.7913878039215686f, 0.2858132352941176f,
-0.9374853333333334f, 0.7901574901960784f, 0.28350641176470587f,
-0.937024f, 0.7889271764705883f, 0.2811995882352941f,
-0.9365626666666667f, 0.7876968627450981f, 0.2788927647058823f,
-0.9361013333333333f, 0.7864665490196079f, 0.2765859411764706f,
-0.93564f, 0.7852362352941177f, 0.2742791176470588f,
-0.9351786666666667f, 0.7840059215686275f, 0.27197229411764706f,
-0.9347173333333334f, 0.7827756078431373f, 0.26966547058823526f,
-0.934256f, 0.7815452941176471f, 0.2673586470588235f,
-0.9337946666666667f, 0.7803149803921569f, 0.26505182352941176f,
-0.9333333333333333f, 0.7790846666666666f, 0.262745f,
-0.932872f, 0.7778543529411764f, 0.2604381764705882f,
-0.9324106666666667f, 0.7766240392156862f, 0.25813135294117645f,
-0.9319493333333333f, 0.775393725490196f, 0.2558245294117647f,
-0.931488f, 0.7741634117647058f, 0.25351770588235295f,
-0.9310266666666667f, 0.7729330980392156f, 0.25121088235294114f,
-0.9305653333333334f, 0.7717027843137254f, 0.2489040588235294f,
-0.930104f, 0.7704724705882352f, 0.24659723529411764f,
-0.9296426666666667f, 0.769242156862745f, 0.24429041176470587f,
-0.9292582156862745f, 0.7679349607843137f, 0.24159913725490195f,
-0.9289506470588236f, 0.7665508823529412f, 0.23852341176470596f,
-0.9286430784313726f, 0.7651668039215686f, 0.2354476862745098f,
-0.9283355098039215f, 0.763782725490196f, 0.2323719607843137f,
-0.9280279411764706f, 0.7623986470588235f, 0.22929623529411763f,
-0.9277203725490196f, 0.761014568627451f, 0.22622050980392155f,
-0.9274128039215687f, 0.7596304901960784f, 0.22314478431372547f,
-0.9271052352941177f, 0.7582464117647059f, 0.2200690588235294f,
-0.9267976666666666f, 0.7568623333333333f, 0.21699333333333332f,
-0.9264900980392157f, 0.7554782549019607f, 0.21391760784313724f,
-0.9261825294117647f, 0.7540941764705882f, 0.21084188235294116f,
-0.9258749607843137f, 0.7527100980392156f, 0.20776615686274508f,
-0.9255673921568628f, 0.7513260196078431f, 0.204690431372549f,
-0.9252598235294117f, 0.7499419411764706f, 0.20161470588235292f,
-0.9249522549019608f, 0.7485578627450981f, 0.19853898039215687f,
-0.9246446862745098f, 0.7471737843137255f, 0.1954632549019608f,
-0.9243371176470588f, 0.7457897058823529f, 0.19238752941176468f,
-0.9240295490196079f, 0.7444056274509804f, 0.1893118039215687f,
-0.9237219803921568f, 0.7430215490196078f, 0.18623607843137255f,
-0.9234144117647058f, 0.7416374705882353f, 0.18316035294117647f,
-0.9231068431372549f, 0.7402533921568627f, 0.18008462745098036f,
-0.9227992745098039f, 0.7388693137254903f, 0.1770089019607843f,
-0.922491705882353f, 0.7374852352941177f, 0.17393317647058823f,
-0.9221841372549019f, 0.7361011568627451f, 0.17085745098039215f,
-0.9218765686274509f, 0.7347170784313726f, 0.16778172549019607f,
-0.921569f, 0.733333f, 0.164706f,
-0.9211076274509804f, 0.7316413529411765f, 0.16178403921568627f,
-0.9206462549019607f, 0.729949705882353f, 0.15886207843137254f,
-0.9201848823529412f, 0.7282580588235295f, 0.1559401176470588f,
-0.9197235098039216f, 0.7265664117647059f, 0.15301815686274509f,
-0.9192621372549019f, 0.7248747647058824f, 0.15009619607843136f,
-0.9188007647058823f, 0.7231831176470589f, 0.14717423529411763f,
-0.9183393921568627f, 0.7214914705882353f, 0.1442522745098039f,
-0.9178780196078431f, 0.7197998235294119f, 0.14133031372549026f,
-0.9174166470588235f, 0.7181081764705882f, 0.13840835294117645f,
-0.9169552745098039f, 0.7164165294117647f, 0.13548639215686273f,
-0.9164939019607843f, 0.7147248823529412f, 0.132564431372549f,
-0.9160325294117647f, 0.7130332352941177f, 0.1296424705882353f,
-0.915571156862745f, 0.7113415882352941f, 0.12672050980392158f,
-0.9151097843137255f, 0.7096499411764706f, 0.12379854901960784f,
-0.9146484117647058f, 0.7079582941176471f, 0.12087658823529411f,
-0.9141870392156862f, 0.7062666470588236f, 0.1179546274509804f,
-0.9137256666666667f, 0.7045750000000001f, 0.11503266666666667f,
-0.913264294117647f, 0.7028833529411765f, 0.11211070588235293f,
-0.9128029215686274f, 0.7011917058823529f, 0.10918874509803922f,
-0.9123415490196078f, 0.6995000588235294f, 0.10626678431372549f,
-0.9118801764705882f, 0.6978084117647059f, 0.10334482352941177f,
-0.9114188039215686f, 0.6961167647058824f, 0.10042286274509803f,
-0.9109574313725489f, 0.6944251176470588f, 0.09750090196078431f,
-0.9104960588235294f, 0.6927334705882354f, 0.09457894117647067f,
-0.9100346862745098f, 0.6910418235294118f, 0.09165698039215686f,
-0.9095476928104574f, 0.6891707712418301f, 0.08865812418300653f,
-0.9090350784313725f, 0.6871203137254902f, 0.0855823725490196f,
-0.9085224640522875f, 0.6850698562091504f, 0.08250662091503266f,
-0.9080098496732025f, 0.6830193986928105f, 0.07943086928104574f,
-0.9074972352941176f, 0.6809689411764706f, 0.07635511764705881f,
-0.9069846209150326f, 0.6789184836601307f, 0.07327936601307188f,
-0.9064720065359477f, 0.6768680261437908f, 0.07020361437908494f,
-0.9059593921568627f, 0.674817568627451f, 0.06712786274509802f,
-0.9054467777777777f, 0.6727671111111111f, 0.06405211111111109f,
-0.9049341633986928f, 0.6707166535947712f, 0.060976359477124153f,
-0.9044215490196078f, 0.6686661960784314f, 0.05790060784313722f,
-0.9039089346405229f, 0.6666157385620914f, 0.054824856209150286f,
-0.9033963202614379f, 0.6645652810457516f, 0.05174910457516336f,
-0.9028837058823529f, 0.6625148235294117f, 0.048673352941176425f,
-0.902371091503268f, 0.6604643660130719f, 0.04559760130718958f,
-0.9017303137254902f, 0.6584395294117646f, 0.042291176470588165f,
-0.9005768823529412f, 0.6565171764705882f, 0.03806205882352935f,
-0.8994234509803921f, 0.6545948235294117f, 0.03383294117647053f,
-0.8982700196078431f, 0.6526724705882353f, 0.029603823529411717f,
-0.897116588235294f, 0.6507501176470588f, 0.0253747058823529f,
-0.8959631568627451f, 0.6488277647058823f, 0.021145588235294083f,
-0.8948097254901961f, 0.6469054117647058f, 0.01691647058823527f,
-0.893656294117647f, 0.6449830588235294f, 0.012687352941176448f,
-0.892502862745098f, 0.643060705882353f, 0.008458235294117634f,
-0.891349431372549f, 0.6411383529411765f, 0.004229117647058821f,
-0.890196f, 0.639216f, 0.0f
-    }
-    , 
-    {
-        1.0f, 0.996078f, 0.941176f,
-        0.996155294117647f, 0.9922334901960784f, 0.9257973725490196f,
-        0.9923105882352942f, 0.9883889803921568f, 0.9104187450980392f,
-        0.9884658823529412f, 0.9845444705882352f, 0.8950401176470588f,
-        0.9846211764705882f, 0.9806999607843138f, 0.8796614901960784f,
-        0.9807764705882354f, 0.9768554509803922f, 0.864282862745098f,
-        0.9790078823529412f, 0.9741641764705882f, 0.8549018235294118f,
-        0.9774699738562092f, 0.9716010392156863f, 0.846187183006536f,
-        0.9759320653594772f, 0.9690379019607843f, 0.8374725424836601f,
-        0.9743941568627451f, 0.9664747647058823f, 0.8287579019607842f,
-        0.9728562483660131f, 0.9639116274509804f, 0.8200432614379085f,
-        0.9713183398692811f, 0.9613484901960785f, 0.8113286209150327f,
-        0.969780431372549f, 0.9587853529411765f, 0.8026139803921568f,
-        0.9680887450980392f, 0.9558633725490197f, 0.7937711960784313f,
-        0.9659357254901961f, 0.9518648627450981f, 0.7845439803921568f,
-        0.9637827058823529f, 0.9478663529411765f, 0.7753167647058823f,
-        0.9616296862745098f, 0.9438678431372549f, 0.7660895490196078f,
-        0.9594766666666666f, 0.9398693333333333f, 0.7568623333333333f,
-        0.9573236470588236f, 0.9358708235294118f, 0.7476351176470588f,
-        0.9551706274509804f, 0.9318723137254902f, 0.7384079019607843f,
-        0.9530176078431373f, 0.9278738039215686f, 0.7291806862745098f,
-        0.9508645882352942f, 0.923875294117647f, 0.7199534705882353f,
-        0.948711568627451f, 0.9198767843137254f, 0.7107262549019607f,
-        0.9465585490196079f, 0.9158782745098039f, 0.7014990392156862f,
-        0.9444055294117647f, 0.9118797647058823f, 0.6922718235294117f,
-        0.9422525098039216f, 0.9078812549019607f, 0.6830446078431373f,
-        0.9404070784313726f, 0.9043441372549019f, 0.6757397254901961f,
-        0.9388692352941177f, 0.9012684117647058f, 0.6703571764705882f,
-        0.9373313921568628f, 0.8981926862745098f, 0.6649746274509803f,
-        0.9357935490196079f, 0.8951169607843137f, 0.6595920784313726f,
-        0.934255705882353f, 0.8920412352941176f, 0.6542095294117647f,
-        0.932717862745098f, 0.8889655098039215f, 0.6488269803921569f,
-        0.9311800196078431f, 0.8858897843137254f, 0.643444431372549f,
-        0.9296421764705882f, 0.8828140588235294f, 0.6380618823529413f,
-        0.9281043333333333f, 0.8797383333333333f, 0.6326793333333334f,
-        0.9265664901960784f, 0.8766626078431372f, 0.6272967843137255f,
-        0.9250286470588236f, 0.8735868823529411f, 0.6219142352941176f,
-        0.9234908039215687f, 0.8705111568627452f, 0.6165316862745098f,
-        0.9219529607843138f, 0.867435431372549f, 0.6111491372549019f,
-        0.9204151176470589f, 0.8643597058823529f, 0.6057665882352941f,
-        0.9188772745098039f, 0.8612839803921569f, 0.6003840392156863f,
-        0.917339431372549f, 0.8582082549019608f, 0.5950014901960785f,
-        0.9158015882352941f, 0.8551325294117648f, 0.5896189411764706f,
-        0.9142637450980392f, 0.8520568039215687f, 0.5842363921568627f,
-        0.9127259019607843f, 0.8489810784313726f, 0.5788538431372549f,
-        0.9111880588235295f, 0.8459053529411765f, 0.5734712941176471f,
-        0.9096502156862745f, 0.8428296274509804f, 0.5680887450980392f,
-        0.9081123725490197f, 0.8397539019607844f, 0.5627061960784314f,
-        0.9065745294117648f, 0.8366781764705883f, 0.5573236470588235f,
-        0.9050366862745098f, 0.8336024509803922f, 0.5519410980392156f,
-        0.9034988431372549f, 0.8305267254901961f, 0.5465585490196079f,
-        0.901961f, 0.827451f, 0.541176f,
-        0.8999617450980393f, 0.8237600980392157f, 0.5379465098039216f,
-        0.8979624901960784f, 0.8200691960784314f, 0.5347170196078431f,
-        0.8959632352941177f, 0.8163782941176471f, 0.5314875294117647f,
-        0.8939639803921569f, 0.8126873921568628f, 0.5282580392156863f,
-        0.8919647254901961f, 0.8089964901960784f, 0.5250285490196078f,
-        0.8899654705882353f, 0.8053055882352942f, 0.5217990588235294f,
-        0.8879662156862745f, 0.8016146862745098f, 0.518569568627451f,
-        0.8859669607843137f, 0.7979237843137256f, 0.5153400784313725f,
-        0.883967705882353f, 0.7942328823529412f, 0.5121105882352941f,
-        0.8819684509803921f, 0.790541980392157f, 0.5088810980392157f,
-        0.8799691960784314f, 0.7868510784313726f, 0.5056516078431372f,
-        0.8779699411764705f, 0.7831601764705882f, 0.5024221176470588f,
-        0.8759706862745098f, 0.779469274509804f, 0.49919262745098036f,
-        0.873971431372549f, 0.7757783725490196f, 0.49596313725490193f,
-        0.8719721764705882f, 0.7720874705882353f, 0.49273364705882355f,
-        0.8699729215686274f, 0.768396568627451f, 0.4895041568627451f,
-        0.8679736666666666f, 0.7647056666666667f, 0.4862746666666667f,
-        0.8659744117647059f, 0.7610147647058824f, 0.48304517647058826f,
-        0.8639751568627451f, 0.757323862745098f, 0.4798156862745098f,
-        0.8619759019607843f, 0.7536329607843137f, 0.4765861960784314f,
-        0.8599766470588235f, 0.7499420588235295f, 0.47335670588235296f,
-        0.8579773921568628f, 0.7462511568627451f, 0.4701272156862745f,
-        0.8559781372549019f, 0.7425602549019609f, 0.46689772549019615f,
-        0.8539788823529412f, 0.7388693529411765f, 0.46366823529411766f,
-        0.8519796274509803f, 0.7351784509803921f, 0.4604387450980392f,
-        0.8499803921568627f, 0.7316413529411765f, 0.4576705882352941f,
-        0.8479811764705882f, 0.7282580588235295f, 0.45536376470588236f,
-        0.8459819607843136f, 0.7248747647058824f, 0.4530569411764706f,
-        0.8439827450980392f, 0.7214914705882353f, 0.45075011764705886f,
-        0.8419835294117647f, 0.7181081764705882f, 0.44844329411764705f,
-        0.8399843137254902f, 0.7147248823529413f, 0.44613647058823536f,
-        0.8379850980392156f, 0.7113415882352941f, 0.44382964705882355f,
-        0.8359858823529411f, 0.7079582941176471f, 0.4415228235294118f,
-        0.8339866666666667f, 0.7045750000000001f, 0.439216f,
-        0.8319874509803922f, 0.7011917058823529f, 0.43690917647058825f,
-        0.8299882352941177f, 0.6978084117647059f, 0.4346023529411765f,
-        0.8279890196078431f, 0.6944251176470588f, 0.43229552941176475f,
-        0.8259898039215686f, 0.6910418235294118f, 0.42998870588235294f,
-        0.8239905882352941f, 0.6876585294117648f, 0.42768188235294124f,
-        0.8219913725490197f, 0.6842752352941177f, 0.42537505882352944f,
-        0.8199921568627451f, 0.6808919411764707f, 0.4230682352941177f,
-        0.8179929411764706f, 0.6775086470588235f, 0.4207614117647059f,
-        0.8159937254901961f, 0.6741253529411765f, 0.41845458823529413f,
-        0.8139945098039216f, 0.6707420588235294f, 0.4161477647058824f,
-        0.811995294117647f, 0.6673587647058824f, 0.41384094117647063f,
-        0.8099960784313726f, 0.6639754705882354f, 0.41153411764705883f,
-        0.8079968627450981f, 0.6605921764705883f, 0.40922729411764713f,
-        0.8059976470588236f, 0.6572088823529412f, 0.40692047058823533f,
-        0.803998431372549f, 0.6538255882352941f, 0.4046136470588236f,
-        0.8019992156862745f, 0.6504422941176471f, 0.4023068235294118f,
-        0.8f, 0.647059f, 0.4f,
-        0.7984621176470589f, 0.6436756666666668f, 0.3980007843137255f,
-        0.7969242352941177f, 0.6402923333333334f, 0.396001568627451f,
-        0.7953863529411765f, 0.6369090000000001f, 0.39400235294117647f,
-        0.7938484705882354f, 0.6335256666666668f, 0.392003137254902f,
-        0.7923105882352942f, 0.6301423333333334f, 0.39000392156862745f,
-        0.7907727058823529f, 0.6267590000000001f, 0.388004705882353f,
-        0.7892348235294118f, 0.6233756666666667f, 0.38600549019607844f,
-        0.7876969411764706f, 0.6199923333333334f, 0.38400627450980396f,
-        0.7861590588235294f, 0.616609f, 0.3820070588235294f,
-        0.7846211764705883f, 0.6132256666666667f, 0.3800078431372549f,
-        0.7830832941176471f, 0.6098423333333334f, 0.3780086274509804f,
-        0.7815454117647059f, 0.6064590000000001f, 0.3760094117647059f,
-        0.7800075294117648f, 0.6030756666666667f, 0.3740101960784314f,
-        0.7784696470588236f, 0.5996923333333333f, 0.37201098039215685f,
-        0.7769317647058824f, 0.596309f, 0.3700117647058824f,
-        0.7753938823529412f, 0.5929256666666667f, 0.36801254901960784f,
-        0.773856f, 0.5895423333333333f, 0.36601333333333336f,
-        0.7723181176470588f, 0.586159f, 0.3640141176470588f,
-        0.7707802352941177f, 0.5827756666666666f, 0.36201490196078434f,
-        0.7692423529411765f, 0.5793923333333334f, 0.36001568627450986f,
-        0.7677044705882353f, 0.576009f, 0.35801647058823527f,
-        0.7661665882352942f, 0.5726256666666666f, 0.3560172549019608f,
-        0.764628705882353f, 0.5692423333333333f, 0.35401803921568625f,
-        0.7630908235294118f, 0.565859f, 0.3520188235294118f,
-        0.7615529411764707f, 0.5624756666666666f, 0.35001960784313724f,
-        0.7592461568627451f, 0.5587847843137255f, 0.34748211764705883f,
-        0.7561704705882353f, 0.5547863529411764f, 0.3444063529411765f,
-        0.7530947843137255f, 0.5507879215686274f, 0.3413305882352941f,
-        0.7500190980392157f, 0.5467894901960784f, 0.3382548235294118f,
-        0.746943411764706f, 0.5427910588235295f, 0.3351790588235295f,
-        0.7438677254901961f, 0.5387926274509803f, 0.33210329411764705f,
-        0.7407920392156863f, 0.5347941960784314f, 0.3290275294117647f,
-        0.7377163529411764f, 0.5307957647058823f, 0.3259517647058824f,
-        0.7346406666666667f, 0.5267973333333333f, 0.322876f,
-        0.7315649803921569f, 0.5227989019607844f, 0.31980023529411766f,
-        0.7284892941176471f, 0.5188004705882353f, 0.31672447058823533f,
-        0.7254136078431372f, 0.5148020392156863f, 0.31364870588235294f,
-        0.7223379215686274f, 0.5108036078431373f, 0.3105729411764706f,
-        0.7197235294117648f, 0.5070358235294118f, 0.3077278823529412f,
-        0.7172629019607843f, 0.5033449215686275f, 0.3049597254901961f,
-        0.7148022745098039f, 0.49965401960784317f, 0.302191568627451f,
-        0.7123416470588235f, 0.49596311764705886f, 0.2994234117647059f,
-        0.7098810196078431f, 0.49227221568627455f, 0.29665525490196076f,
-        0.7074203921568628f, 0.48858131372549024f, 0.2938870980392157f,
-        0.7049597647058824f, 0.48489041176470593f, 0.2911189411764706f,
-        0.7024991372549021f, 0.4811995098039217f, 0.28835078431372557f,
-        0.7000385098039216f, 0.47750860784313726f, 0.2855826274509804f,
-        0.6975778823529412f, 0.47381770588235295f, 0.2828144705882353f,
-        0.6951172549019609f, 0.47012680392156864f, 0.28004631372549016f,
-        0.6926566274509804f, 0.4664359019607843f, 0.2772781568627451f,
-        0.690196f, 0.462745f, 0.27451f,
-        0.6877354509803922f, 0.4599768431372549f, 0.27512513725490195f,
-        0.6852749019607843f, 0.45720868627450983f, 0.27574027450980393f,
-        0.6828143529411765f, 0.4544405294117647f, 0.27635541176470585f,
-        0.6803538039215686f, 0.45167237254901965f, 0.2769705490196078f,
-        0.6778932549019608f, 0.44890421568627453f, 0.2775856862745098f,
-        0.6754327058823529f, 0.4461360588235294f, 0.2782008235294118f,
-        0.6729721568627451f, 0.44336790196078435f, 0.2788159607843137f,
-        0.6705116078431372f, 0.44059974509803923f, 0.2794310980392157f,
-        0.6680510588235294f, 0.4378315882352941f, 0.28004623529411765f,
-        0.6655905098039215f, 0.43506343137254905f, 0.2806613725490196f,
-        0.6631299607843137f, 0.43229527450980404f, 0.28127650980392155f,
-        0.6606694117647058f, 0.42952711764705886f, 0.2818916470588235f,
-        0.6582857450980392f, 0.42698964705882353f, 0.2825067843137255f,
-        0.6561327254901961f, 0.42514423529411766f, 0.2831219215686275f,
-        0.6539797058823529f, 0.4232988235294118f, 0.28373705882352945f,
-        0.6518266862745098f, 0.4214534117647059f, 0.2843521960784314f,
-        0.6496736666666666f, 0.41960800000000004f, 0.28496733333333335f,
-        0.6475206470588235f, 0.4177625882352941f, 0.2855824705882353f,
-        0.6453676274509803f, 0.41591717647058823f, 0.28619760784313725f,
-        0.6432146078431372f, 0.41407176470588236f, 0.2868127450980392f,
-        0.6410615882352941f, 0.4122263529411765f, 0.2874278823529412f,
-        0.6389085686274509f, 0.4103809411764706f, 0.2880430196078432f,
-        0.6367555490196078f, 0.40853552941176474f, 0.2886581568627451f,
-        0.6346025294117646f, 0.4066901176470588f, 0.2892732941176471f,
-        0.6324495098039216f, 0.40484470588235294f, 0.28988843137254905f,
-        0.6301426862745098f, 0.4029992549019608f, 0.2903498039215686f,
-        0.6276820588235295f, 0.40115376470588243f, 0.2906574117647059f,
-        0.6252214313725489f, 0.39930827450980394f, 0.29096501960784316f,
-        0.6227608039215686f, 0.3974627843137255f, 0.29127262745098037f,
-        0.6203001764705882f, 0.39561729411764707f, 0.29158023529411764f,
-        0.6178395490196078f, 0.39377180392156863f, 0.2918878431372549f,
-        0.6153789215686274f, 0.3919263137254902f, 0.2921954509803922f,
-        0.612918294117647f, 0.39008082352941176f, 0.2925030588235294f,
-        0.6104576666666667f, 0.3882353333333333f, 0.29281066666666666f,
-        0.6079970392156863f, 0.3863898431372549f, 0.29311827450980393f,
-        0.6055364117647059f, 0.3845443529411765f, 0.29342588235294115f,
-        0.6030757843137254f, 0.38269886274509807f, 0.2937334901960784f,
-        0.600615156862745f, 0.38085337254901963f, 0.2940410980392157f,
-        0.5970011764705883f, 0.37877723529411766f, 0.2936566470588235f,
-        0.5930027450980392f, 0.3766242156862745f, 0.2930415098039216f,
-        0.5890043137254901f, 0.3744711960784314f, 0.2924263725490196f,
-        0.5850058823529412f, 0.37231817647058824f, 0.2918112352941176f,
-        0.5810074509803922f, 0.37016515686274515f, 0.2911960980392157f,
-        0.5770090196078431f, 0.36801213725490195f, 0.2905809607843137f,
-        0.573010588235294f, 0.3658591176470588f, 0.28996582352941175f,
-        0.5690121568627451f, 0.3637060980392157f, 0.2893506862745098f,
-        0.565013725490196f, 0.3615530784313726f, 0.28873554901960785f,
-        0.5610152941176471f, 0.35940005882352943f, 0.2881204117647059f,
-        0.557016862745098f, 0.3572470392156863f, 0.2875052745098039f,
-        0.553018431372549f, 0.35509401960784315f, 0.286890137254902f,
-        0.54902f, 0.352941f, 0.286275f,
-        0.5459442352941176f, 0.35078798039215686f, 0.28596739215686273f,
-        0.5428684705882353f, 0.3486349607843137f, 0.2856597843137255f,
-        0.539792705882353f, 0.3464819411764706f, 0.28535217647058825f,
-        0.5367169411764706f, 0.34432892156862743f, 0.285044568627451f,
-        0.5336411764705882f, 0.3421759019607843f, 0.2847369607843137f,
-        0.5305654117647058f, 0.3400228823529412f, 0.2844293529411765f,
-        0.5274896470588235f, 0.33786986274509806f, 0.2841217450980392f,
-        0.5244138823529413f, 0.33571684313725497f, 0.283814137254902f,
-        0.5213381176470588f, 0.3335638235294118f, 0.28350652941176474f,
-        0.5182623529411765f, 0.33141080392156863f, 0.28319892156862747f,
-        0.5151865882352942f, 0.3292577843137255f, 0.2828913137254902f,
-        0.5121108235294118f, 0.32710476470588234f, 0.282583705882353f,
-        0.5090350588235294f, 0.3250286470588235f, 0.28219921568627454f,
-        0.5059592941176471f, 0.32318323529411763f, 0.28158407843137256f,
-        0.5028835294117647f, 0.32133782352941176f, 0.2809689411764706f,
-        0.4998077647058824f, 0.3194924117647059f, 0.2803538039215686f,
-        0.496732f, 0.317647f, 0.2797386666666667f,
-        0.4936562352941177f, 0.3158015882352941f, 0.2791235294117647f,
-        0.49058047058823534f, 0.3139561764705882f, 0.27850839215686274f,
-        0.48750470588235295f, 0.31211076470588234f, 0.27789325490196076f,
-        0.4844289411764706f, 0.31026535294117646f, 0.27727811764705884f,
-        0.48135317647058823f, 0.3084199411764706f, 0.27666298039215687f,
-        0.4782774117647059f, 0.30657452941176466f, 0.2760478431372549f,
-        0.4752016470588236f, 0.30472911764705884f, 0.2754327058823529f,
-        0.4721258823529412f, 0.3028837058823529f, 0.27481756862745094f,
-        0.4690501568627451f, 0.30088449019607844f, 0.27404862745098035f,
-        0.4659744705882353f, 0.2987314705882353f, 0.27312588235294116f,
-        0.4628987843137255f, 0.29657845098039215f, 0.2722031372549019f,
-        0.45982309803921567f, 0.294425431372549f, 0.2712803921568627f,
-        0.4567474117647059f, 0.29227241176470586f, 0.27035764705882354f,
-        0.4536717254901961f, 0.2901193921568627f, 0.2694349019607843f,
-        0.4505960392156863f, 0.2879663725490196f, 0.2685121568627451f,
-        0.4475203529411765f, 0.28581335294117644f, 0.26758941176470585f,
-        0.44444466666666665f, 0.2836603333333333f, 0.26666666666666666f,
-        0.4413689803921569f, 0.28150731372549015f, 0.2657439215686275f,
-        0.43829329411764706f, 0.27935429411764706f, 0.2648211764705882f,
-        0.43521760784313723f, 0.2772012745098039f, 0.26389843137254904f,
-        0.43214192156862746f, 0.2750482549019608f, 0.2629756862745098f,
-        0.42837411764705885f, 0.2733565882352941f, 0.26182229411764707f,
-        0.4243756078431374f, 0.27181870588235296f, 0.2605920196078432f,
-        0.4203770980392157f, 0.27028082352941174f, 0.2593617450980392f,
-        0.4163785882352941f, 0.2687429411764706f, 0.2581314705882353f,
-        0.41238007843137253f, 0.2672050588235294f, 0.25690119607843137f,
-        0.408381568627451f, 0.26566717647058824f, 0.2556709215686275f,
-        0.40438305882352943f, 0.2641292941176471f, 0.2544406470588235f,
-        0.40038454901960785f, 0.2625914117647059f, 0.2532103725490196f,
-        0.39638603921568627f, 0.2610535294117647f, 0.25198009803921567f,
-        0.3923875294117647f, 0.2595156470588235f, 0.2507498235294118f,
-        0.38838901960784317f, 0.25797776470588235f, 0.24951954901960785f,
-        0.3843905098039216f, 0.2564398823529412f, 0.24828927450980393f,
-        0.380392f, 0.254902f, 0.247059f,
-    },
-
-    // one channel red for 
-    {
-        0.0f, 0.f, 0.f, 
-        0.00392156862745098f, 0.f, 0.f, 
-        0.00784313725490196f, 0.f, 0.f, 
-        0.011764705882352941f, 0.f, 0.f, 
-        0.01568627450980392f, 0.f, 0.f, 
-        0.0196078431372549f, 0.f, 0.f, 
-        0.023529411764705882f, 0.f, 0.f, 
-        0.027450980392156862f, 0.f, 0.f, 
-        0.03137254901960784f, 0.f, 0.f, 
-        0.03529411764705882f, 0.f, 0.f, 
-        0.0392156862745098f, 0.f, 0.f, 
-        0.043137254901960784f, 0.f, 0.f, 
-        0.047058823529411764f, 0.f, 0.f, 
-        0.050980392156862744f, 0.f, 0.f, 
-        0.054901960784313725f, 0.f, 0.f, 
-        0.058823529411764705f, 0.f, 0.f, 
-        0.06274509803921569f, 0.f, 0.f, 
-        0.06666666666666667f, 0.f, 0.f, 
-        0.07058823529411765f, 0.f, 0.f, 
-        0.07450980392156863f, 0.f, 0.f, 
-        0.0784313725490196f, 0.f, 0.f, 
-        0.08235294117647059f, 0.f, 0.f, 
-        0.08627450980392157f, 0.f, 0.f, 
-        0.09019607843137255f, 0.f, 0.f, 
-        0.09411764705882353f, 0.f, 0.f, 
-        0.09803921568627451f, 0.f, 0.f, 
-        0.10196078431372549f, 0.f, 0.f, 
-        0.10588235294117647f, 0.f, 0.f, 
-        0.10980392156862745f, 0.f, 0.f, 
-        0.11372549019607843f, 0.f, 0.f, 
-        0.11764705882352941f, 0.f, 0.f, 
-        0.12156862745098039f, 0.f, 0.f, 
-        0.12549019607843137f, 0.f, 0.f, 
-        0.12941176470588234f, 0.f, 0.f, 
-        0.13333333333333333f, 0.f, 0.f, 
-        0.13725490196078433f, 0.f, 0.f, 
-        0.1411764705882353f, 0.f, 0.f, 
-        0.14509803921568626f, 0.f, 0.f, 
-        0.14901960784313725f, 0.f, 0.f, 
-        0.15294117647058825f, 0.f, 0.f, 
-        0.1568627450980392f, 0.f, 0.f, 
-        0.16078431372549018f, 0.f, 0.f, 
-        0.16470588235294117f, 0.f, 0.f, 
-        0.16862745098039217f, 0.f, 0.f, 
-        0.17254901960784313f, 0.f, 0.f, 
-        0.1764705882352941f, 0.f, 0.f, 
-        0.1803921568627451f, 0.f, 0.f, 
-        0.1843137254901961f, 0.f, 0.f, 
-        0.18823529411764706f, 0.f, 0.f, 
-        0.19215686274509802f, 0.f, 0.f, 
-        0.19607843137254902f, 0.f, 0.f, 
-        0.2f, 0.f, 0.f, 
-        0.20392156862745098f, 0.f, 0.f, 
-        0.20784313725490194f, 0.f, 0.f, 
-        0.21176470588235294f, 0.f, 0.f, 
-        0.21568627450980393f, 0.f, 0.f, 
-        0.2196078431372549f, 0.f, 0.f, 
-        0.22352941176470587f, 0.f, 0.f, 
-        0.22745098039215686f, 0.f, 0.f, 
-        0.23137254901960785f, 0.f, 0.f, 
-        0.23529411764705882f, 0.f, 0.f, 
-        0.2392156862745098f, 0.f, 0.f, 
-        0.24313725490196078f, 0.f, 0.f, 
-        0.24705882352941178f, 0.f, 0.f, 
-        0.25098039215686274f, 0.f, 0.f, 
-        0.2549019607843137f, 0.f, 0.f, 
-        0.2588235294117647f, 0.f, 0.f, 
-        0.2627450980392157f, 0.f, 0.f, 
-        0.26666666666666666f, 0.f, 0.f, 
-        0.27058823529411763f, 0.f, 0.f, 
-        0.27450980392156865f, 0.f, 0.f, 
-        0.2784313725490196f, 0.f, 0.f, 
-        0.2823529411764706f, 0.f, 0.f, 
-        0.28627450980392155f, 0.f, 0.f, 
-        0.2901960784313725f, 0.f, 0.f, 
-        0.29411764705882354f, 0.f, 0.f, 
-        0.2980392156862745f, 0.f, 0.f, 
-        0.30196078431372547f, 0.f, 0.f, 
-        0.3058823529411765f, 0.f, 0.f, 
-        0.30980392156862746f, 0.f, 0.f, 
-        0.3137254901960784f, 0.f, 0.f, 
-        0.3176470588235294f, 0.f, 0.f, 
-        0.32156862745098036f, 0.f, 0.f, 
-        0.3254901960784314f, 0.f, 0.f, 
-        0.32941176470588235f, 0.f, 0.f, 
-        0.3333333333333333f, 0.f, 0.f, 
-        0.33725490196078434f, 0.f, 0.f, 
-        0.3411764705882353f, 0.f, 0.f, 
-        0.34509803921568627f, 0.f, 0.f, 
-        0.34901960784313724f, 0.f, 0.f, 
-        0.3529411764705882f, 0.f, 0.f, 
-        0.3568627450980392f, 0.f, 0.f, 
-        0.3607843137254902f, 0.f, 0.f, 
-        0.36470588235294116f, 0.f, 0.f, 
-        0.3686274509803922f, 0.f, 0.f, 
-        0.37254901960784315f, 0.f, 0.f, 
-        0.3764705882352941f, 0.f, 0.f, 
-        0.3803921568627451f, 0.f, 0.f, 
-        0.38431372549019605f, 0.f, 0.f, 
-        0.38823529411764707f, 0.f, 0.f, 
-        0.39215686274509803f, 0.f, 0.f, 
-        0.396078431372549f, 0.f, 0.f, 
-        0.4f, 0.f, 0.f, 
-        0.403921568627451f, 0.f, 0.f, 
-        0.40784313725490196f, 0.f, 0.f, 
-        0.4117647058823529f, 0.f, 0.f, 
-        0.4156862745098039f, 0.f, 0.f, 
-        0.4196078431372549f, 0.f, 0.f, 
-        0.4235294117647059f, 0.f, 0.f, 
-        0.42745098039215684f, 0.f, 0.f, 
-        0.43137254901960786f, 0.f, 0.f, 
-        0.43529411764705883f, 0.f, 0.f, 
-        0.4392156862745098f, 0.f, 0.f, 
-        0.44313725490196076f, 0.f, 0.f, 
-        0.44705882352941173f, 0.f, 0.f, 
-        0.45098039215686275f, 0.f, 0.f, 
-        0.4549019607843137f, 0.f, 0.f, 
-        0.4588235294117647f, 0.f, 0.f, 
-        0.4627450980392157f, 0.f, 0.f, 
-        0.4666666666666667f, 0.f, 0.f, 
-        0.47058823529411764f, 0.f, 0.f, 
-        0.4745098039215686f, 0.f, 0.f, 
-        0.4784313725490196f, 0.f, 0.f, 
-        0.4823529411764706f, 0.f, 0.f, 
-        0.48627450980392156f, 0.f, 0.f, 
-        0.49019607843137253f, 0.f, 0.f, 
-        0.49411764705882355f, 0.f, 0.f, 
-        0.4980392156862745f, 0.f, 0.f, 
-        0.5019607843137255f, 0.f, 0.f, 
-        0.5058823529411764f, 0.f, 0.f, 
-        0.5098039215686274f, 0.f, 0.f, 
-        0.5137254901960784f, 0.f, 0.f, 
-        0.5176470588235293f, 0.f, 0.f, 
-        0.5215686274509804f, 0.f, 0.f, 
-        0.5254901960784314f, 0.f, 0.f, 
-        0.5294117647058824f, 0.f, 0.f, 
-        0.5333333333333333f, 0.f, 0.f, 
-        0.5372549019607843f, 0.f, 0.f, 
-        0.5411764705882353f, 0.f, 0.f, 
-        0.5450980392156862f, 0.f, 0.f, 
-        0.5490196078431373f, 0.f, 0.f, 
-        0.5529411764705883f, 0.f, 0.f, 
-        0.5568627450980392f, 0.f, 0.f, 
-        0.5607843137254902f, 0.f, 0.f, 
-        0.5647058823529412f, 0.f, 0.f, 
-        0.5686274509803921f, 0.f, 0.f, 
-        0.5725490196078431f, 0.f, 0.f, 
-        0.5764705882352941f, 0.f, 0.f, 
-        0.580392156862745f, 0.f, 0.f, 
-        0.5843137254901961f, 0.f, 0.f, 
-        0.5882352941176471f, 0.f, 0.f, 
-        0.592156862745098f, 0.f, 0.f, 
-        0.596078431372549f, 0.f, 0.f, 
-        0.6f, 0.f, 0.f, 
-        0.6039215686274509f, 0.f, 0.f, 
-        0.6078431372549019f, 0.f, 0.f, 
-        0.611764705882353f, 0.f, 0.f, 
-        0.615686274509804f, 0.f, 0.f, 
-        0.6196078431372549f, 0.f, 0.f, 
-        0.6235294117647059f, 0.f, 0.f, 
-        0.6274509803921569f, 0.f, 0.f, 
-        0.6313725490196078f, 0.f, 0.f, 
-        0.6352941176470588f, 0.f, 0.f, 
-        0.6392156862745098f, 0.f, 0.f, 
-        0.6431372549019607f, 0.f, 0.f, 
-        0.6470588235294118f, 0.f, 0.f, 
-        0.6509803921568628f, 0.f, 0.f, 
-        0.6549019607843137f, 0.f, 0.f, 
-        0.6588235294117647f, 0.f, 0.f, 
-        0.6627450980392157f, 0.f, 0.f, 
-        0.6666666666666666f, 0.f, 0.f, 
-        0.6705882352941176f, 0.f, 0.f, 
-        0.6745098039215687f, 0.f, 0.f, 
-        0.6784313725490196f, 0.f, 0.f, 
-        0.6823529411764706f, 0.f, 0.f, 
-        0.6862745098039216f, 0.f, 0.f, 
-        0.6901960784313725f, 0.f, 0.f, 
-        0.6941176470588235f, 0.f, 0.f, 
-        0.6980392156862745f, 0.f, 0.f, 
-        0.7019607843137254f, 0.f, 0.f, 
-        0.7058823529411764f, 0.f, 0.f, 
-        0.7098039215686275f, 0.f, 0.f, 
-        0.7137254901960784f, 0.f, 0.f, 
-        0.7176470588235294f, 0.f, 0.f, 
-        0.7215686274509804f, 0.f, 0.f, 
-        0.7254901960784313f, 0.f, 0.f, 
-        0.7294117647058823f, 0.f, 0.f, 
-        0.7333333333333333f, 0.f, 0.f, 
-        0.7372549019607844f, 0.f, 0.f, 
-        0.7411764705882353f, 0.f, 0.f, 
-        0.7450980392156863f, 0.f, 0.f, 
-        0.7490196078431373f, 0.f, 0.f, 
-        0.7529411764705882f, 0.f, 0.f, 
-        0.7568627450980392f, 0.f, 0.f, 
-        0.7607843137254902f, 0.f, 0.f, 
-        0.7647058823529411f, 0.f, 0.f, 
-        0.7686274509803921f, 0.f, 0.f, 
-        0.7725490196078432f, 0.f, 0.f, 
-        0.7764705882352941f, 0.f, 0.f, 
-        0.7803921568627451f, 0.f, 0.f, 
-        0.7843137254901961f, 0.f, 0.f, 
-        0.788235294117647f, 0.f, 0.f, 
-        0.792156862745098f, 0.f, 0.f, 
-        0.796078431372549f, 0.f, 0.f, 
-        0.8f, 0.f, 0.f, 
-        0.803921568627451f, 0.f, 0.f, 
-        0.807843137254902f, 0.f, 0.f, 
-        0.8117647058823529f, 0.f, 0.f, 
-        0.8156862745098039f, 0.f, 0.f, 
-        0.8196078431372549f, 0.f, 0.f, 
-        0.8235294117647058f, 0.f, 0.f, 
-        0.8274509803921568f, 0.f, 0.f, 
-        0.8313725490196078f, 0.f, 0.f, 
-        0.8352941176470589f, 0.f, 0.f, 
-        0.8392156862745098f, 0.f, 0.f, 
-        0.8431372549019608f, 0.f, 0.f, 
-        0.8470588235294118f, 0.f, 0.f, 
-        0.8509803921568627f, 0.f, 0.f, 
-        0.8549019607843137f, 0.f, 0.f, 
-        0.8588235294117647f, 0.f, 0.f, 
-        0.8627450980392157f, 0.f, 0.f, 
-        0.8666666666666667f, 0.f, 0.f, 
-        0.8705882352941177f, 0.f, 0.f, 
-        0.8745098039215686f, 0.f, 0.f, 
-        0.8784313725490196f, 0.f, 0.f, 
-        0.8823529411764706f, 0.f, 0.f, 
-        0.8862745098039215f, 0.f, 0.f, 
-        0.8901960784313725f, 0.f, 0.f, 
-        0.8941176470588235f, 0.f, 0.f, 
-        0.8980392156862745f, 0.f, 0.f, 
-        0.9019607843137255f, 0.f, 0.f, 
-        0.9058823529411765f, 0.f, 0.f, 
-        0.9098039215686274f, 0.f, 0.f, 
-        0.9137254901960784f, 0.f, 0.f, 
-        0.9176470588235294f, 0.f, 0.f, 
-        0.9215686274509803f, 0.f, 0.f, 
-        0.9254901960784314f, 0.f, 0.f, 
-        0.9294117647058824f, 0.f, 0.f, 
-        0.9333333333333333f, 0.f, 0.f, 
-        0.9372549019607843f, 0.f, 0.f, 
-        0.9411764705882353f, 0.f, 0.f, 
-        0.9450980392156862f, 0.f, 0.f, 
-        0.9490196078431372f, 0.f, 0.f, 
-        0.9529411764705882f, 0.f, 0.f, 
-        0.9568627450980391f, 0.f, 0.f, 
-        0.9607843137254902f, 0.f, 0.f, 
-        0.9647058823529412f, 0.f, 0.f, 
-        0.9686274509803922f, 0.f, 0.f, 
-        0.9725490196078431f, 0.f, 0.f, 
-        0.9764705882352941f, 0.f, 0.f, 
-        0.9803921568627451f, 0.f, 0.f, 
-        0.984313725490196f, 0.f, 0.f, 
-        0.9882352941176471f, 0.f, 0.f, 
-        0.9921568627450981f, 0.f, 0.f, 
-        0.996078431372549f, 0.f, 0.f, 
-        1.0f, 0.f, 0.f, 
-    }
-};
-
-
-// global_variable
-const char* Color_Map_Names[Number_of_Color_Maps] = 
-{
-    "Inferno",
-    "Black Body",
-    "Kindlmann",
-    "Extended Kindlmann",
-    "Three Wave Yellow-Grey-Blue",
-    "Three Wave Blue-Green-Yellow",
-    "Four Wave Grey-Red-Green",
-    "Five Wave Yellow-Brown-Blue",
-    "Blue 1",
-    "Blue 2",
-    "Blue 3",
-    "Blue 4",
-    "Blue-Orange Divergent",
-    "Brown 1",
-    "Brown 2",
-    "Brown 3",
-    "Green 1",
-    "Green 2",
-    "Green 3",
-    "Green 4",
-    "Mellow Rainbow",
-    "Orange 1",
-    "Orange 2",
-    "Orange-Green-Blue-Gray",
-    "Purple 1",
-    "Purple 2",
-    "Red 1",
-    "Red 2",
-    "Red 3",
-    "Yellow 1",
-    "Yellow 2", 
-    "Copy_texture(always at end)"
-};
\ No newline at end of file
diff --git a/Resources.cpp b/src/Resources.cpp
similarity index 100%
rename from Resources.cpp
rename to src/Resources.cpp
diff --git a/TextureLoadQueue.cpp b/src/TextureLoadQueue.cpp
similarity index 100%
rename from TextureLoadQueue.cpp
rename to src/TextureLoadQueue.cpp
diff --git a/glad.c b/src/glad.c
similarity index 100%
rename from glad.c
rename to src/glad.c

From b798e1c0eb833a1df1f3d3f188b88ac493f76f18 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:12:24 +0100
Subject: [PATCH 22/33] grey out frags with selected meta-tags

---
 src/grey_out_settings.h | 69 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 69 insertions(+)
 create mode 100644 src/grey_out_settings.h

diff --git a/src/grey_out_settings.h b/src/grey_out_settings.h
new file mode 100644
index 0000000..4098ce7
--- /dev/null
+++ b/src/grey_out_settings.h
@@ -0,0 +1,69 @@
+
+#ifndef GREY_OUT_SETTINGS_H
+#define GREY_OUT_SETTINGS_H
+
+#include "genomeData.h"
+#include <unordered_set>
+
+
+struct GreyOutSettings
+{   
+    s32 n_flags = 64;
+    s32 grey_out_flags[64];
+    std::unordered_set<std::string> default_grey_out_tag = {"FalseDuplicate", "Primary"}; 
+
+    GreyOutSettings() 
+    {
+        for (u32 i = 0; i < n_flags; i ++ )
+        {
+            grey_out_flags[i] = 0;
+        }
+    }
+
+    GreyOutSettings(meta_data* Meta_Data) 
+    {
+        for (u32 i = 0; i < n_flags; i ++ )
+        {   
+            if (default_grey_out_tag.find((char*)Meta_Data->tags[i]) != default_grey_out_tag.end())
+            {
+                grey_out_flags[i] = 1;
+            }
+            else
+            {
+                grey_out_flags[i] = 0;
+            }
+        }
+    }
+
+    void toggle_grey_out(u32 index)
+    {
+        if (index < n_flags) {
+            grey_out_flags[index] = !grey_out_flags[index];
+        }
+    }
+
+    void clear_all()
+    {
+        std::fill(grey_out_flags, grey_out_flags+n_flags, 0);
+    }
+
+    /* 
+    Return the tag name if the tag is greyed out, otherwise return an empty string
+    */
+    std::string is_grey_out(u64* meta_data_flags, meta_data* Meta_Data)
+    {   
+        for (s32 i = 0; i  < ArrayCount(Meta_Data->tags) ; i ++ )
+        {
+            if (grey_out_flags[i] && (*meta_data_flags & ((u64)1 << i)))
+            {
+                return (char*)(Meta_Data->tags[i]);
+            }
+        }
+        return "";
+    }
+    
+};
+
+
+
+#endif // GREY_OUT_SETTINGS_H
\ No newline at end of file

From 09a636689d874e349362ba03b97b23e0fd0de962 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:12:40 +0100
Subject: [PATCH 23/33] add comments

---
 src/shaderSource.h | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/src/shaderSource.h b/src/shaderSource.h
index e963fae..8473e45 100644
--- a/src/shaderSource.h
+++ b/src/shaderSource.h
@@ -70,14 +70,14 @@ float linearTextureID(vec2 coords)
 }
 
 // https://community.khronos.org/t/mipmap-level-calculation-using-dfdx-dfdy/67480
-float mip_map_level(in vec2 texture_coordinate)
+float mip_map_level(in vec2 texture_coordinate) // textCoord.xy * textureSize(tex, 0).xy
 {
     // The OpenGL Graphics System: A Specification 4.2
     //  - chapter 3.9.11, equation 3.21
-    vec2  dx_vtc        = dFdx(texture_coordinate);
-    vec2  dy_vtc        = dFdy(texture_coordinate);
+    vec2  dx_vtc        = dFdx(texture_coordinate); // change velocity in x direction
+    vec2  dy_vtc        = dFdy(texture_coordinate); // change velocity in y direction
     float delta_max_sqr = max(dot(dx_vtc, dx_vtc), dot(dy_vtc, dy_vtc));
-    return 0.5 * log2(delta_max_sqr);
+    return 0.5 * log2(delta_max_sqr); // the faster the change, the higher the mip level
 }
 
 vec3 pixLookup(vec3 inCoords)

From c369c55fd1964b1d2ff7f4a69998b5f99172d62e Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:13:21 +0100
Subject: [PATCH 24/33] small changes

---
 src/showWindowData.h        | 6 ++++++
 src/user_profile_settings.h | 7 +++++++
 src/utilsPretextView.h      | 2 +-
 3 files changed, 14 insertions(+), 1 deletion(-)
 create mode 100644 src/user_profile_settings.h

diff --git a/src/showWindowData.h b/src/showWindowData.h
index 45b4e76..4132b4e 100644
--- a/src/showWindowData.h
+++ b/src/showWindowData.h
@@ -24,7 +24,11 @@ SOFTWARE.
 #ifndef SHOW_WINDOW_DATA_H
 #define SHOW_WINDOW_DATA_H
 
+#include <vector>
+#include <fmt/core.h>
 #include "utilsPretextView.h"
+#include "genomeData.h"
+
 
 struct pointi
 {
@@ -225,4 +229,6 @@ enum global_mode
 };
 
 
+
+
 #endif // SHOW_WINDOW_DATA_H
\ No newline at end of file
diff --git a/src/user_profile_settings.h b/src/user_profile_settings.h
new file mode 100644
index 0000000..3ddcb16
--- /dev/null
+++ b/src/user_profile_settings.h
@@ -0,0 +1,7 @@
+
+
+struct UserProfileSettings
+{
+    bool invert_mouse = false;
+    UserProfileSettings() {}
+};
diff --git a/src/utilsPretextView.h b/src/utilsPretextView.h
index 29ec32c..13247e9 100644
--- a/src/utilsPretextView.h
+++ b/src/utilsPretextView.h
@@ -233,7 +233,7 @@ void my_code_position_handler(const char* file, int line, const char* message=nu
 f64 GetTime(); // return current time in seconds
 
 
-/* 计算数组的 95% 分位数 */
+/* 计算数组的 95% 分位数, 最大的 0.05 */
 template <typename T>
 T percentile_cal(T* data, u32 size, f32 percentile=0.95)
 {   

From d871090bc7ce99cd9989ff7197214b60ecb7061e Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:13:43 +0100
Subject: [PATCH 25/33] yahs sort mode (unfinished)

---
 src/yahs_sort.cpp | 505 ++++++++++++++++++++++++++++++++++++++++++++++
 src/yahs_sort.h   | 118 +++++++++++
 2 files changed, 623 insertions(+)
 create mode 100644 src/yahs_sort.cpp
 create mode 100644 src/yahs_sort.h

diff --git a/src/yahs_sort.cpp b/src/yahs_sort.cpp
new file mode 100644
index 0000000..31e3afe
--- /dev/null
+++ b/src/yahs_sort.cpp
@@ -0,0 +1,505 @@
+#include "yahs_sort.h"
+
+
+
+
+// Graph类方法实现
+int64_t Graph::add_arc(uint32_t v, uint32_t w, double wt, int64_t link_id, bool comp)
+{
+    link_id = link_id > 0 ? link_id : arcs.size();
+    arcs.emplace_back(v, w, wt, link_id, comp);
+    return this->arcs.size() - 1;
+}
+
+
+void Graph::arc_sort() 
+{
+    std::sort(arcs.begin(), arcs.end(),
+              [](const GraphArc& a, const GraphArc& b) {
+                  return a.v < b.v;
+              });
+}
+
+/*
+创建索引 idx 向量，
+    编号为：节点编号<<32 | 正反向，
+    每个元素为uint64_t类型，
+        高32位：开始边的在arcs中的索引
+        低32位：开始节点对应的边的数量
+*/
+void Graph::arc_index()
+{
+    this->idx.clear();
+    this->idx.resize(arcs.size() * 2, 0);
+    
+    uint64_t last_i = 0;
+    for (uint64_t i = 1; i <= arcs.size(); ++i) 
+    {
+        if (i == arcs.size() || arcs[i-1].v != arcs[i].v) 
+        {
+            this->idx[this->arcs[i-1].v] = (last_i << 32) | (i - last_i);
+            last_i = i;
+        }
+    }
+}
+
+void Graph::clean(bool shear) {
+    arcs.erase(
+        std::remove_if(arcs.begin(), arcs.end(),
+                      [](const GraphArc& arc) { return arc.del; }),
+        arcs.end()
+    );
+    
+    if (shear) {
+        arcs.shrink_to_fit();
+    }
+    
+    arc_sort();
+    arc_index();
+}
+
+
+/*
+    判断v节点是否存在权重大于wt的边，且该边指向w节点
+*/
+bool Graph::is_mwt(uint32_t v, uint32_t w) const {
+    size_t na = get_arc_count(v);
+    if (na == 0) return false;
+    
+    const GraphArc* av = get_arcs(v);
+    double mwt = 0;
+    bool found = false;
+    
+    for (size_t i = 0; i < na; ++i) {
+        if (av[i].wt > mwt) {
+            mwt = av[i].wt;
+            found = (av[i].w == w);
+        }
+    }
+    return found;
+}
+
+
+bool Graph::exist_strong_edge(uint32_t v, double wt) const {
+    size_t na = get_arc_count(v);
+    if (na < 2) return false;
+    
+    const GraphArc* av = get_arcs(v);
+    for (size_t i = 0; i < na; ++i) {
+        if (av[i].wt > wt) return true;
+    }
+    return false;
+}
+
+int Graph::trim_simple_filter(double min_wt, double min_diff_h, double min_diff_l, int min_len) {
+    int n_del = 0;
+    uint32_t nv = idx.size();
+    
+    // 移除短序列的边
+    for (uint32_t v = 0; v < nv; ++v) {
+        if (v >= min_len) continue;
+        size_t na = get_arc_count(v);
+        GraphArc* av = get_arcs(v);
+        for (size_t i = 0; i < na; ++i) {
+            av[i].del = true;
+            ++n_del;
+        }
+    }
+    clean(true);
+    
+    // 基于权重过滤
+    for (uint32_t v = 0; v < nv; ++v) {
+        size_t na = get_arc_count(v);
+        GraphArc* av = get_arcs(v);
+        
+        double mwt = 0, smwt = 0;
+        size_t n_ma = 0;
+        
+        // 找出最大和次大权重
+        for (size_t i = 0; i < na; ++i) {
+            if (av[i].wt > mwt) {
+                smwt = mwt;
+                mwt = av[i].wt;
+                n_ma = 1;
+            } else if (av[i].wt == mwt) {
+                ++n_ma;
+            } else if (av[i].wt > smwt) {
+                smwt = av[i].wt;
+            }
+        }
+        
+        // 过滤边
+        for (size_t i = 0; i < na; ++i) {
+            if ((av[i].wt >= min_wt && av[i].wt >= mwt * min_diff_h) ||
+                (av[i].wt < min_wt && av[i].wt == mwt && mwt * min_diff_l >= smwt && n_ma == 1)) {
+                continue;
+            }
+            av[i].del = true;
+            ++n_del;
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+int Graph::trim_tips() {
+    int n_del = 0;
+    uint32_t nv = idx.size();
+    
+    for (uint32_t v = 0; v < nv; ++v) {
+        if (get_arc_count(v^1) != 1 || get_arc_count(v) > 0) continue;
+        
+        GraphArc* av = get_arcs(v^1);
+        if (get_arc_count(av->w^1) > 1 && !is_mwt(av->w^1, v)) {
+            av->del = true;
+            ++n_del;
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+int Graph::trim_blunts() {
+    int n_del = 0;
+    uint32_t nv = idx.size();
+    
+    for (uint32_t v = 0; v < nv; ++v) {
+        if (get_arc_count(v^1) > 0) continue;
+        
+        size_t na = get_arc_count(v);
+        if (na > 1) {
+            GraphArc* av = get_arcs(v);
+            bool del = true;
+            
+            for (size_t i = 0; i < na; ++i) {
+                if (get_arc_count(av[i].w) < 1) {
+                    del = false;
+                    break;
+                }
+            }
+            
+            if (del) {
+                for (size_t i = 0; i < na; ++i) {
+                    av[i].del = true;
+                    ++n_del;
+                }
+            }
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+int Graph::trim_repeats() {
+    int n_del = 0;
+    uint32_t nv = idx.size();
+    
+    for (uint32_t v = 0; v < nv; ++v) {
+        if (get_arc_count(v) != 2 || get_arc_count(v^1) != 2) continue;
+        
+        GraphArc* av = get_arcs(v^1);
+        uint32_t v1 = av->w;
+        uint32_t v2 = (av + 1)->w;
+        
+        GraphArc* aw = get_arcs(v);
+        uint32_t w1 = aw->w;
+        uint32_t w2 = (aw + 1)->w;
+        
+        if ((is_mwt(v1, w1) && is_mwt(v2, w2)) ||
+            (is_mwt(v1, w2) && is_mwt(v2, w1))) {
+            av->del = true;
+            (av + 1)->del = true;
+            aw->del = true;
+            (aw + 1)->del = true;
+            n_del += 4;
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+int Graph::trim_transitive_edges() {
+    int n_del = 0;
+    uint32_t nv = idx.size();
+    
+    for (uint32_t v = 0; v < nv; ++v) {
+        size_t na = get_arc_count(v);
+        if (na < 2) continue;
+        
+        GraphArc* av = get_arcs(v);
+        for (size_t i = 0; i < na; ++i) {
+            for (size_t j = 0; j < na; ++j) {
+                if (i == j) continue;
+                if (is_mwt(av[i].w, av[j].w)) {
+                    av[j].del = true;
+                    ++n_del;
+                }
+            }
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+int Graph::trim_pop_bubbles() {
+    int n_del = 0;
+    uint32_t nv = idx.size();
+    
+    for (uint32_t v = 0; v < nv; ++v) {
+        size_t na = get_arc_count(v);
+        if (na != 2) continue;
+        
+        GraphArc* av = get_arcs(v);
+        uint32_t v1 = av->w;
+        uint32_t v2 = (av + 1)->w;
+        
+        if (get_arc_count(v1) == 1 && get_arc_count(v2) == 1) {
+            GraphArc* av1 = get_arcs(v1);
+            GraphArc* av2 = get_arcs(v2);
+            
+            if (av1->w == av2->w) {
+                av->del = true;
+                (av + 1)->del = true;
+                av1->del = true;
+                av2->del = true;
+                n_del += 4;
+            }
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+int Graph::trim_pop_undirected() {
+    int n_del = 0;
+    uint32_t nv = idx.size();
+    
+    for (uint32_t v = 0; v < nv; ++v) {
+        if (get_arc_count(v) != 1) continue;
+        
+        GraphArc* av = get_arcs(v);
+        uint32_t w = av->w;
+        
+        if (is_mwt(v^1, w) && get_arc_count(v^1) > 1 && get_arc_count(w) > 0) {
+            ++n_del;
+            // 删除互补边
+            for (size_t i = 0; i < get_arc_count(w^1); ++i) {
+                GraphArc* aw = get_arcs(w^1);
+                if (aw[i].w == v) {
+                    aw[i].del = true;
+                    break;
+                }
+            }
+            // 删除原边
+            for (size_t i = 0; i < get_arc_count(v^1); ++i) {
+                GraphArc* av1 = get_arcs(v^1);
+                if (av1[i].w == w) {
+                    av1[i].del = true;
+                    break;
+                }
+            }
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+int Graph::trim_weak_edges() {
+    int n_del = 0;
+    
+    for (auto& arc : arcs) {
+        if (exist_strong_edge(arc.v, arc.wt) && exist_strong_edge(arc.w^1, arc.wt)) {
+            arc.del = true;
+            ++n_del;
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+int Graph::trim_self_loops() {
+    int n_del = 0;
+    
+    for (auto& arc : arcs) {
+        if (get_arc_count(arc.v) > 1) continue;
+        
+        for (size_t i = 0; i < get_arc_count(arc.w); ++i) {
+            GraphArc* aw = get_arcs(arc.w);
+            if (aw[i].w == arc.v) {
+                aw[i].del = true;
+                ++n_del;
+                break;
+            }
+        }
+    }
+    
+    clean(true);
+    return n_del;
+}
+
+
+/* 
+    删除歧义边
+*/
+int Graph::trim_ambiguous_edges() {
+    int n_del = 0;
+    
+    for (auto& arc : arcs) {
+        if (
+            get_arc_count(arc.v) > 1      // 节点v存在多条边
+            || get_arc_count(arc.w^1) > 1 // 节点w的互补边存在多条
+        ) {
+            arc.del = true;
+            ++n_del;
+        }
+    }
+    
+    this->clean(true);
+    return n_del;
+}
+
+
+// 构建图的方法
+Graph* YaHS::build_graph_from_links(const LikelihoodTable& likelihood_table, double min_score, double la)
+{
+    auto g = std::make_unique<Graph>();
+    
+    uint32_t num_frags = likelihood_table.get_num_frags();
+    
+    // 预分配空间以提高性能
+    g->reserve_arcs(num_frags * num_frags * 4);  // 最坏情况下每个片段对可能有4个方向的连接
+    
+    // 遍历所有可能的片段对构建图
+    double link_score = 0.f;
+    int index_arc = 0;
+    for (uint32_t i = 0; i < num_frags; ++i) {
+        for (uint32_t j = i+1; j < num_frags; ++j) {
+            
+            // 检查四个方向的连接
+            for (uint32_t k = 0; k < 4; ++k) {
+                link_score = likelihood_table(i, j, k);
+                if (link_score >= min_score) {
+                    // 添加边
+                    index_arc = g->add_arc((i << 1) | (k >> 1), (j << 1) | (k & 1), link_score, -1, false);
+                    
+                    // 添加互补边
+                    g->add_arc((j << 1) | !(k & 1), (i << 1) | !(k >> 1), link_score, index_arc, true);
+                }
+            }
+        }
+    }
+    
+    // 排序和索引
+    g->arc_sort();
+    g->arc_index();
+    
+    return g.release();
+}
+
+
+
+// 运行scaffolding
+std::vector<std::vector<uint32_t>> YaHS::yahs_sorting(const LikelihoodTable& likelihood_table, double min_score, double la)
+{
+    min_norm_ = min_score;
+    la_ = la;
+    
+    // 构建图
+    std::unique_ptr<Graph> g(this->build_graph_from_links(likelihood_table, min_norm_, la_));
+    
+    // 记录修剪前的边数
+    size_t n_arc = g->get_arc_size();
+    
+    // 循环进行图修剪，直到没有边被删除
+    while (true) {
+        // 简单过滤
+        g->trim_simple_filter(0.1, 0.7, 0.1, 0);
+        
+        // 修剪端点
+        g->trim_tips();
+        
+        // 修剪钝化
+        g->trim_blunts();
+        
+        // 修剪重复边
+        g->trim_repeats();
+        
+        // 修剪传递边
+        g->trim_transitive_edges();
+        
+        // 修剪泡状边
+        g->trim_pop_bubbles();
+        
+        // 修剪无向边
+        g->trim_pop_undirected();
+        
+        // 修剪弱边
+        g->trim_weak_edges();
+        
+        // 修剪自环边
+        g->trim_self_loops();
+        
+        // 如果边数没有变化，说明修剪完成
+        if (g->get_arc_size() == n_arc)
+            break;
+        n_arc = g->get_arc_size();
+    }
+    
+    // 最后修剪歧义边
+    g->trim_ambiguous_edges();
+    
+    // 构建染色体序列
+    std::vector<std::vector<uint32_t>> chromosomes;
+    std::vector<bool> visited(g->get_arc_size() * 2, false);
+    
+    // 遍历所有节点
+    for (uint32_t v = 0; v < g->get_arc_size() * 2; ++v) {
+        if (visited[v]) continue;
+        
+        // 开始一个新的染色体
+        std::vector<uint32_t> chromosome;
+        
+        // 使用深度优先搜索遍历图
+        uint32_t current = v;
+        while (!visited[current]) {
+            visited[current] = true;
+            chromosome.push_back(current >> 1);  // 存储节点编号（去掉方向位）
+            
+            // 获取当前节点的边
+            size_t na = g->get_arc_count(current);
+            if (na == 0) break;
+            
+            const GraphArc* av = g->get_arcs(current);
+            if (av == nullptr) break;
+            
+            // 选择权重最大的边
+            double max_wt = -1;
+            uint32_t next = current;
+            
+            for (size_t i = 0; i < na; ++i) {
+                if (av[i].wt > max_wt) {
+                    max_wt = av[i].wt;
+                    next = av[i].w;
+                }
+            }
+            
+            if (next == current) break;
+            current = next;
+        }
+        
+        if (!chromosome.empty()) {
+            chromosomes.push_back(std::move(chromosome));
+        }
+    }
+    
+    return chromosomes;
+}
+
diff --git a/src/yahs_sort.h b/src/yahs_sort.h
new file mode 100644
index 0000000..8affb60
--- /dev/null
+++ b/src/yahs_sort.h
@@ -0,0 +1,118 @@
+#ifndef YAHS_SORT_H
+#define YAHS_SORT_H
+
+// likelihood table
+#include "likelihood_table.h"
+// frags_order
+#include "frags_order.h"
+
+
+
+// 定义图相关的结构体
+struct GraphArc 
+{
+    uint32_t v;      // 前31位表示节点编号，最后一位表示方向
+    uint32_t w;      // 前31位表示节点编号，最后一位表示方向
+    double wt;       // 权重
+    int64_t link_id; // 边的id 
+    bool del=false;        // 是否删除
+    bool comp;       // 是否互补
+
+    GraphArc(uint32_t v, uint32_t w, double wt, int64_t link_id, bool comp)
+        : v(v), w(w), wt(wt), link_id(link_id), comp(comp) {}
+};
+
+class Graph 
+{
+public:
+    Graph() = default;
+    ~Graph() = default;
+    
+    // 基本图操作方法
+    // 添加边
+    int64_t add_arc(uint32_t v, uint32_t w, double wt, int64_t link_id, bool comp = false) ;
+    // 预分配空间
+    void reserve_arcs(size_t size) {
+        arcs.reserve(size);
+    }
+    // 获取边的数量
+    size_t get_arc_count(uint32_t v) const {
+        if (v >= idx.size()) return 0;
+        return idx[v] & 0xFFFFFFFF; // 取低32位
+    }
+    // 获取边的指针
+    GraphArc* get_arcs(uint32_t v) {
+        if (v >= idx.size()) return nullptr;
+        uint64_t offset = idx[v] >> 32;
+        return &arcs[offset];
+    }
+    // 获取边的指针（常量版本）
+    const GraphArc* get_arcs(uint32_t v) const {
+        if (v >= idx.size()) return nullptr;
+        uint64_t offset = idx[v] >> 32;
+        return &arcs[offset];
+    }
+    // 获取边的数量 
+    size_t get_arc_size() const { return arcs.size(); }
+
+    // 图操作相关方法
+    // 对边进行排序
+    void arc_sort();
+    // 对边进行索引
+    void arc_index();
+    // 清理图，删除被标记为删除的边
+    void clean(bool shear);
+    
+    // 图修剪方法
+    // 基于权重过滤
+    int trim_simple_filter(double min_wt, double min_diff_h, double min_diff_l, int min_len);
+    // 修剪端点
+    int trim_tips();
+    // 修剪钝化
+    int trim_blunts();
+    // 修剪重复边
+    int trim_repeats();
+    // 修剪传递边
+    int trim_transitive_edges();
+    // 修剪泡状边
+    int trim_pop_bubbles();
+    // 修剪无向边
+    int trim_pop_undirected();
+    // 修剪弱边
+    int trim_weak_edges();
+    // 修剪自环边
+    int trim_self_loops();
+    // 修剪歧义边
+    int trim_ambiguous_edges();
+
+private:
+    std::vector<GraphArc> arcs;
+    std::vector<uint64_t> idx;  // 用于快速查找节点的边, 记录每个节点开始和结束的边
+    
+    // 辅助方法
+    bool is_mwt(uint32_t v, uint32_t w) const; // 判断v和w之间是否存在最大权重边
+    bool exist_strong_edge(uint32_t v, double wt) const; // 判断v节点是否存在权重大于wt的边
+};
+
+
+
+
+class YaHS
+{
+    
+private:
+    double min_norm_;
+    double la_;
+    
+public:
+    YaHS() = default;
+    
+    // 构建图的方法
+    Graph* build_graph_from_links(const LikelihoodTable& likelihood_table, double min_norm, double la);
+    
+    // 运行scaffolding
+    std::vector<std::vector<uint32_t>> yahs_sorting(const LikelihoodTable& likelihood_table, double min_norm = 0.1, double la = 0.1);
+};
+
+
+#endif // YAHS_H 
\ No newline at end of file

From ea7e968d3abb8e518042e2a2be708460c7dd7633 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:13:58 +0100
Subject: [PATCH 26/33] small changes

---
 .gitignore | 5 +++++
 Info.plist | 2 +-
 2 files changed, 6 insertions(+), 1 deletion(-)

diff --git a/.gitignore b/.gitignore
index 3289690..1570b61 100644
--- a/.gitignore
+++ b/.gitignore
@@ -32,6 +32,9 @@ python/autoCut/models/cfgs/error_model.pth
 python/autoCut/models/cfgs/chr_model.pth
 python/autoCut/models/swin/mmdet.egg-info/*
 python/classic_cut/show_hic_density.py
+python/colormap_colorblind/*
+python/greedy_topo_sort_test/*
+python/linear_weight_test/*
 
 subprojects/eigen
 subprojects/glfw
@@ -77,3 +80,5 @@ hic_figures/
 auto_cut_output/
 python/autoCut/models/swin/README.md
 python/autoCut/models/swin/mmcv_custom/
+
+yahs_archive/
\ No newline at end of file
diff --git a/Info.plist b/Info.plist
index 608a4e5..d9961f5 100644
--- a/Info.plist
+++ b/Info.plist
@@ -13,7 +13,7 @@
   <key>CFBundleIconFile</key>
   <string>icon.icns</string>
   <key>CFBundleShortVersionString</key>
-  <string>1.0.2</string>
+  <string>1.0.3</string>
   <key>CFBundleInfoDictionaryVersion</key>
   <string>6.0</string>
   <key>CFBundlePackageType</key>

From c81b6b0a8c0da2d62544728d5bcc938d1fec194a Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:15:20 +0100
Subject: [PATCH 27/33] changes

---
 CMakeLists.txt    |  25 +-
 PretextView.cpp   | 682 +++++++++++++++++++++++++++++++++-------------
 install.cmake.bat |   2 +-
 install.cmake.sh  |   1 -
 4 files changed, 515 insertions(+), 195 deletions(-)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 0a9d104..e800b11 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -1,10 +1,12 @@
 cmake_minimum_required(VERSION 3.17)
-project(PretextView VERSION "1.0.2")
+project(PretextView VERSION "1.0.3")
 
 
-option(FORCE_MAC_X86 "Force macOS x86_64 build" OFF)  
+option(FORCE_MAC_X86 "Force macOS x86_64 build" OFF)
 option(BUILD_UNIVERSAL "Build macOS universal binary" OFF) 
 option(WITH_TORCH "Find and link torch lib" OFF)
+option(DEBUG_OUTPUT_LIKELIHOOD_TABLE "Output likelihood table to file" OFF)
+option(DEBUG_OUTPUT_PIXEL_CUT_FILE "Output the pixel cut files" OFF)
 
 set(CMAKE_C_STANDARD 17)   
 set(CMAKE_CXX_STANDARD 17)   
@@ -14,7 +16,6 @@ set(CMAKE_CXX_STANDARD_REQUIRED ON)
 set(target_name PretextViewAI)
 
 if (APPLE)
-
     if(FORCE_MAC_X86) 
         message(STATUS "Forcing x86_64 architecture build")
         set(CMAKE_OSX_ARCHITECTURES "x86_64" CACHE STRING "Build architectures" FORCE)
@@ -114,8 +115,9 @@ else()
     message(STATUS "lib_deflate_static found: ${lib_deflate_static}")
 endif()
 
-
-
+add_library(lib_yahs STATIC 
+    src/yahs_sort.cpp
+)
 
 # set(CMAKE_PREFIX_PATH "subprojects/libtorch/share/cmake/Torch")
 if (WITH_TORCH)
@@ -131,14 +133,16 @@ if (WITH_TORCH)
     target_link_libraries(
         sortlib PUBLIC 
         ${TORCH_LIBRARIES} 
-        ${torchscatter})
+        ${torchscatter} 
+        lib_yahs)
 else()
     message(STATUS "Libtorch: Build without libtorch")
     add_library(sortlib STATIC src/frag_sort.cpp )
+    target_link_libraries(sortlib PRIVATE lib_yahs)
 endif()
 
 
-add_library(glad STATIC glad.c)
+add_library(glad STATIC src/glad.c)
 add_library(utilsPretextView STATIC src/utilsPretextView.cpp)
 add_library(copy_texture STATIC src/copy_texture.cpp)
 # target_link_libraries(copy_texture PRIVATE Eigen3::Eigen)
@@ -161,6 +165,7 @@ else()
         MACOSX_BUNDLE_INFO_PLIST "${CMAKE_SOURCE_DIR}/Info.plist"
     )
 endif()
+
 target_compile_definitions(${target_name} PRIVATE PV=${PROJECT_VERSION})
 if (CMAKE_BUILD_TYPE STREQUAL "Debug")
     target_compile_definitions(${target_name} PRIVATE DEBUG)
@@ -169,6 +174,11 @@ if (CMAKE_BUILD_TYPE STREQUAL "Debug")
 else()
     message(STATUS "Release version: flag (${CMAKE_BUILD_TYPE}) without DEBUG")
 endif()
+if (DEBUG_OUTPUT_LIKELIHOOD_TABLE)
+    message(STATUS "Debug version: flag (${CMAKE_BUILD_TYPE}) with DEBUG_OUTPUT_LIKELIHOOD_TABLE")
+    target_compile_definitions(${target_name} PRIVATE DEBUG_OUTPUT_LIKELIHOOD_TABLE)
+endif()
+
 target_link_libraries(
     ${target_name} PRIVATE 
     ${lib_deflate_static} 
@@ -180,6 +190,7 @@ target_link_libraries(
     copy_texture 
     imgui 
     glfw 
+    lib_yahs
 )   
 
 if (APPLE)
diff --git a/PretextView.cpp b/PretextView.cpp
index 978f593..b6b60a8 100644
--- a/PretextView.cpp
+++ b/PretextView.cpp
@@ -99,13 +99,15 @@ SOFTWARE.
 #pragma clang diagnostic pop
 
 
-#include "Resources.cpp"
+#include "Resources.cpp" // defines icons, fonts
 
 #include "genomeData.h"
 #include "showWindowData.h"
 #include "hic_figure.h"
 #include "ai_cut_json_parse.h"
 #include "frag_cut_calculation.h"
+#include "grey_out_settings.h"
+#include "user_profile_settings.h"
 
 #include "shaderSource.h"
 /*
@@ -365,6 +367,16 @@ QuadTree_Data;
 #endif
 
 
+global_variable
+Extension_Graph_Data * pixel_density_extension = nullptr;
+
+global_variable
+TexturesArray4AI* textures_array_ptr = nullptr;
+
+global_variable
+FragCutCal* frag_cut_cal_ptr = nullptr;
+
+
 struct
     CustomColorMapOrder
 {
@@ -561,6 +573,10 @@ meta_mode_data *
 Extension_Mode_Data;
 
 
+/*
+用于显示在input sequences中点击后选中的片段的名字，
+显示label持续 5 秒钟
+*/
 struct selected_sequence_cover_countor
 {
     s32 original_contig_index;
@@ -724,12 +740,12 @@ setContactMatrixVertexArray(contact_matrix* Contact_Matrix_, bool copy_flag=fals
 
 global_function
 void 
-prepare_before_copy(f32 *original_control_points);
+prepare_before_copy(f32 *original_color_control_points);
 
 
 global_function
 void 
-restore_settings_after_copy(const f32 *original_control_points);
+restore_settings_after_copy(const f32 *original_color_control_points);
 
 
 
@@ -781,7 +797,9 @@ Default_Tags[] =
     "U",
     "B1",
     "B2",
-    "B3"
+    "B3", 
+    "FalseDuplicate",
+    "Primary"
 };
 
 
@@ -792,7 +810,7 @@ Map_State;
 
 #include "copy_texture.h"
 // define the struct to store the ai model mask
-std::unique_ptr<FragSortTool> frag_sort_method = nullptr;
+FragSortTool* frag_sort_method = new FragSortTool();
 global_variable auto auto_curation_state = AutoCurationState();
 
 
@@ -1684,7 +1702,7 @@ MouseMove(GLFWwindow* window, f64 x, f64 y)
                 u32 testPixel = pixel1;
                 u32 testContig = contig;
                 while (testContig == contig) // move testContig to the 1 before the first or 1 after the last pixel of the contig
-                {   
+                {
 
                     if (pixel1 < pixel2) 
                     {
@@ -1699,8 +1717,8 @@ MouseMove(GLFWwindow* window, f64 x, f64 y)
 
                     if (testPixel == 0 || testPixel >= (nPixels - 1)) break;
 
-                    testContig = pixel1 < pixel2 ? Map_State->contigIds[++testPixel] : Map_State->contigIds[--testPixel];
-                    if (testPixel == 0 || testPixel >= (nPixels - 1)) break;
+                    // testContig = pixel1 < pixel2 ? Map_State->contigIds[++testPixel] : Map_State->contigIds[--testPixel];
+                    // if (testPixel == 0 || testPixel >= (nPixels - 1)) break;
                 }   
                 if (Map_State->contigIds[testPixel] == contig)
                 {
@@ -1716,7 +1734,7 @@ MouseMove(GLFWwindow* window, f64 x, f64 y)
             if (Edit_Pixels.selecting) // select the contig
             {
                 Edit_Pixels.selectPixels.x = my_Max(pixel1, my_Max(Edit_Pixels.selectPixels.x, Edit_Pixels.selectPixels.y)); // used to select multiple contigs with pressing space and dragging the mouse
-                Edit_Pixels.selectPixels.x = pixel1;
+                // Edit_Pixels.selectPixels.x = pixel1;
                 while(  Edit_Pixels.selectPixels.x < (Number_of_Pixels_1D - 1) &&
                         ((Edit_Pixels.scaffSelecting && Map_State->scaffIds[Edit_Pixels.selectPixels.x] && Map_State->scaffIds[Edit_Pixels.selectPixels.x] == Map_State->scaffIds[1 + Edit_Pixels.selectPixels.x] ) || 
                           Map_State->contigIds[Edit_Pixels.selectPixels.x] == Map_State->contigIds[1 + Edit_Pixels.selectPixels.x]))
@@ -1964,15 +1982,16 @@ MouseMove(GLFWwindow* window, f64 x, f64 y)
     }
 }
 
-global_variable
-u08
-Mouse_Invert = 0;
-
 
 global_variable
 u08
 Grey_Haplotigs = 1;
 
+global_variable
+GreyOutSettings* Grey_Out_Settings = nullptr;
+
+global_variable
+UserProfileSettings* user_profile_settings_ptr = nullptr;
 
 global_variable
 u08
@@ -1982,8 +2001,8 @@ global_function
 void
 Mouse(GLFWwindow* window, s32 button, s32 action, s32 mods)
 {
-    s32 primaryMouse = Mouse_Invert ? GLFW_MOUSE_BUTTON_RIGHT : GLFW_MOUSE_BUTTON_LEFT;
-    s32 secondaryMouse = Mouse_Invert ? GLFW_MOUSE_BUTTON_LEFT : GLFW_MOUSE_BUTTON_RIGHT;    
+    s32 primaryMouse = user_profile_settings_ptr->invert_mouse ? GLFW_MOUSE_BUTTON_RIGHT : GLFW_MOUSE_BUTTON_LEFT;
+    s32 secondaryMouse = user_profile_settings_ptr->invert_mouse ? GLFW_MOUSE_BUTTON_LEFT : GLFW_MOUSE_BUTTON_RIGHT;    
     
     if (Loading || auto_sort_state)
     {
@@ -3347,7 +3366,7 @@ Render() {
             }
         }
 
-        // label to show the selected sequence
+        // label to show the selected sequence 从 input sequences 里面选中的片段
         if (Selected_Sequence_Cover_Countor.end_time > 0.)
         {   
             f64 crt_time = GetTime();
@@ -3534,9 +3553,6 @@ Render() {
             // selected frags into a string
             SelectArea select_area;
             auto_curation_state.get_selected_fragments(select_area, Map_State, Number_of_Pixels_1D, Contigs);
-            std::string selectedFragmentsStr = fmt::format(
-                "Selected fragments number: {}", select_area.selected_frag_ids.size());
-            std::string start_and_end_pixel_str = fmt::format("Select range: {} - {}", auto_curation_state.get_start(), auto_curation_state.get_end());
 
             { // draw the help text in the bottom right corner
                 fonsSetFont(FontStash_Context, Font_Bold);
@@ -3544,23 +3560,26 @@ Render() {
                 fonsVertMetrics(FontStash_Context, 0, 0, &lh);
                 fonsSetColor(FontStash_Context, FourFloatColorToU32(Scaff_Mode_Data->text));
 
-                std::vector<char*> helpTexts = {
-                    (char *)"Select/Exclude area for sorting",
-                    (char *)"K: exit",
-                    (char *)"S: clear select area",
-                    (char *)"Q/W: quit/redo edit",
-                    (char *)"Space: Auto sort",
-                    (char *)"Left Click: select/unselect the fragment",
-                    (char *)selectedFragmentsStr.c_str(),
-                    (char *)start_and_end_pixel_str.c_str()
+                std::vector<std::string> helpTexts = {
+                    "Select/Exclude area for sorting",
+                    "F: exit",
+                    "S: clear select area", 
+                    "C/Z: cut / cancel cut",
+                    "Q/W: quit / redo edit",
+                    "Space: Pixel sort",
+                    "Left Click: un/select the fragment",
+                    fmt::format("Up/Down: inc/dec Cut threshold: {:.4f}", auto_curation_state.auto_cut_threshold).c_str(),
+                    fmt::format("Left/Right: change Sort Mode ({})", auto_curation_state.sort_mode_names[auto_curation_state.sort_mode]),
+                    // fmt::format("Selected fragments number: {}", select_area.selected_frag_ids.size()),
+                    // fmt::format("Select pixel range: {} - {}", auto_curation_state.get_start(), auto_curation_state.get_end())
                 };
 
                 f32 textBoxHeight = (f32)helpTexts.size() * (lh + 1.0f) - 1.0f;
                 f32 spacing = 10.0f;
 
                 f32 textWidth = 0.f; 
-                for (auto* i :helpTexts){
-                    textWidth = my_Max(textWidth, fonsTextBounds(FontStash_Context, 0, 0, i, 0, NULL));
+                for (const auto& i :helpTexts){
+                    textWidth = my_Max(textWidth, fonsTextBounds(FontStash_Context, 0, 0, i.c_str(), 0, NULL));
                 } 
                 textWidth = my_Min(textWidth, width - 2 * spacing);
 
@@ -3586,7 +3605,7 @@ Render() {
                         FontStash_Context, 
                         width - spacing - textWidth, 
                         height - spacing - textBoxHeight + (i * (lh + 1.0f)), 
-                        helpTexts[i], 
+                        helpTexts[i].c_str(), 
                         0);
                 }
             }
@@ -3687,7 +3706,7 @@ Render() {
                 end_contig += ((f32)cont->length / (f32)Number_of_Pixels_1D); // end of the contig
                 if (*cont->metaDataFlags)
                 {
-                    u32 tmp = 0;
+                    u32 tmp = 0; // used to count the number of tags drawn
                     bool haplotigTagged = false;
                     ForLoop2(ArrayCount(Meta_Data->tags))
                     {
@@ -3724,7 +3743,9 @@ Render() {
                         }
                     }
 
-                    if (haplotigTagged && Grey_Haplotigs)
+                    // draw the grey out mask
+                    std::string grey_out_tag = Grey_Out_Settings->is_grey_out(cont->metaDataFlags, Meta_Data);
+                    if (!grey_out_tag.empty())
                     {
                         vert[0].x = ModelXToScreen(start_contig - 0.5f);     vert[0].y = ModelYToScreen(0.5f - start_contig);
                         vert[1].x = ModelXToScreen(start_contig - 0.5f);     vert[1].y = ModelYToScreen(0.5f - end_contig);
@@ -4369,7 +4390,7 @@ Render() {
             glUseProgram(UI_Shader->shaderProgram);
             glUniformMatrix4fv(Flat_Shader->matLocation, 1, GL_FALSE, textNormalMat);
             glViewport(0, 0, (s32)width, (s32)height);
-            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "Auto Sorting...", 0);
+            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "Pixel Sorting...", 0);
 
             ChangeSize((s32)width, (s32)height);
         }
@@ -4387,7 +4408,7 @@ Render() {
             glUseProgram(UI_Shader->shaderProgram);
             glUniformMatrix4fv(Flat_Shader->matLocation, 1, GL_FALSE, textNormalMat);
             glViewport(0, 0, (s32)width, (s32)height);
-            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "Auto cutting...", 0);
+            fonsDrawText(FontStash_Context, width * 0.5f, height * 0.5f, "Pixel cut...", 0);
 
             ChangeSize((s32)width, (s32)height);
         }
@@ -4655,7 +4676,7 @@ add_graph_to_extensions(
 
 global_function
 void 
-push_extensions_to_opengl(memory_arena *arena, u32 added_index = 0)
+push_extensions_to_opengl(memory_arena *arena, u32 added_index = 0, f32 scale=-1.f)
 {   
     u32 exIndex = 0;
     // TraverseLinkedList(Extensions.head, extension_node)
@@ -4663,7 +4684,7 @@ push_extensions_to_opengl(memory_arena *arena, u32 added_index = 0)
     extension_node* node = Extensions.head ;
     while (added_index--)
     {
-        node = node->next;
+        node = node->next; // 移动到最后一个节点
         exIndex ++ ;
     }
     for ( ; node; node = node->next) {
@@ -4677,7 +4698,7 @@ push_extensions_to_opengl(memory_arena *arena, u32 added_index = 0)
 #define DefaultGraphBase 32.0f
 #define DefaultGraphLineSize 1.0f
 #define DefaultGraphColour {0.1f, 0.8f, 0.7f, 1.0f}
-                    gph->scale = DefaultGraphScale;
+                    gph->scale = scale>0.f? scale : DefaultGraphScale;
                     gph->base = DefaultGraphBase;
                     gph->lineSize = DefaultGraphLineSize;
                     gph->colour = DefaultGraphColour;
@@ -4851,6 +4872,16 @@ LoadFile(const char *filePath, memory_arena *arena, char **fileName, u64 *header
             delete[] Contact_Matrix->vbos; 
             Contact_Matrix->vbos = nullptr;
         }
+        if (pixel_density_extension)
+        {
+            delete pixel_density_extension;
+            pixel_density_extension = nullptr;
+        }
+        if (frag_cut_cal_ptr)
+        {
+            delete frag_cut_cal_ptr;
+            frag_cut_cal_ptr = nullptr;
+        }
         ResetMemoryArenaP(arena); // release all the memory allocated, avoid memory leak
         auto_curation_state.clear();
     }
@@ -5441,6 +5472,62 @@ LoadFile(const char *filePath, memory_arena *arena, char **fileName, u64 *header
         }
     }
 
+    // 为添加的功能初始化指针
+    {
+        // 给pixel_density_extension分配内存
+        // pixel_density_extension = new Extension_Graph_Data(Number_of_Pixels_1D);
+        
+        // 为 copy textures to cpu 分配内存
+        if (textures_array_ptr) delete textures_array_ptr;
+        textures_array_ptr = new TexturesArray4AI(
+            Number_of_Textures_1D, 
+            Texture_Resolution, 
+            *fileName, 
+            Contigs
+        );
+        // prepare before reading textures
+        f32 original_color_control_points[3];
+        prepare_before_copy(original_color_control_points);
+        textures_array_ptr->copy_buffer_to_textures(
+            Contact_Matrix, 
+            false
+        ); 
+        restore_settings_after_copy(original_color_control_points);
+
+        // 为 Pixel Cut 存储的数据分配空间，初始化，并且将数据push到extension中
+        if (frag_cut_cal_ptr) delete frag_cut_cal_ptr;
+        frag_cut_cal_ptr = new FragCutCal(
+            textures_array_ptr,
+            Number_of_Pixels_1D,
+            auto_curation_state.auto_cut_diag_window_for_pixel_mean);
+
+        // clear mem for textures
+        textures_array_ptr->clear_textures();
+        
+        // 添加数据到extensions
+        std::string graph_name = "pixel_discontinuity";
+        bool is_pix_density_added = Extensions.is_graph_name_exist(graph_name);
+        if (!is_pix_density_added) // 未添加pixel density
+        {
+            u32 added_num = Extensions.get_num_extensions();
+            u32 graph_data[Number_of_Pixels_1D];
+            f32 max_desity = 0.f;
+            for (auto& i : frag_cut_cal_ptr->hic_pixel_density_origin) max_desity = std::max(max_desity, i);
+            for (u32 i=0; i < Number_of_Pixels_1D; i ++)
+            {
+                graph_data[i] = (u32)(255 * (1.001f- frag_cut_cal_ptr->hic_pixel_density_origin[i] / max_desity));
+            }
+            add_graph_to_extensions(arena, (u32*)graph_name.c_str(), graph_data);
+            
+            push_extensions_to_opengl(arena, added_num, 0.05f); // push hic_pixel_density to opengl buffer
+        }
+        else 
+        {
+            fmt::print("hic_pixel_density extension already exists\n");
+            assert(false);
+        }
+    }
+
     FenceIn(File_Loaded = 1);
 
     if (LoadState(*headerHash)) LoadState(*headerHash + 1);
@@ -5633,6 +5720,18 @@ Setup()
     memset(Meta_Data, 0, sizeof(meta_data));
     MetaData_Active_Tag = 0;
     ForLoop(ArrayCount(Default_Tags)) strcpy((char *)Meta_Data->tags[MetaData_Active_Tag + index], Default_Tags[index]);
+    if (Grey_Out_Settings) 
+    {
+        delete Grey_Out_Settings;
+        Grey_Out_Settings = nullptr;
+    }
+    Grey_Out_Settings = new GreyOutSettings(Meta_Data);
+    if (user_profile_settings_ptr)
+    {
+        delete user_profile_settings_ptr;
+        user_profile_settings_ptr = nullptr;
+    }
+    user_profile_settings_ptr = new UserProfileSettings();
 
     // Contig Name Label UI
     {
@@ -6289,13 +6388,13 @@ global_function
 void 
 BreakMap(
     const u32& loc, 
-    const u32 ignore_len = 10 // cut点到开头或者结尾的长度不足ignore_len的contig不会被切断
+    const u32 ignore_len // cut点到开头或者结尾的长度不足ignore_len的contig不会被切断
 )
 {  
     if (loc >= Number_of_Pixels_1D)
     {
         char buff[512];
-        snprintf(buff, sizeof(buff), "Error: loc (%d) should be smaller than number of pixels (%d)", loc, Number_of_Pixels_1D);
+        snprintf(buff, sizeof(buff), "[Pixel Cut] Error: loc (%d) should be smaller than number of pixels (%d)", loc, Number_of_Pixels_1D);
         MY_CHECK(buff);
         assert(0);
     }
@@ -6317,7 +6416,7 @@ BreakMap(
     if ((loc - ptr_left) < ignore_len || (ptr_right - loc) < ignore_len )  
     {   
         fmt::print(
-            "Error: original_contig_id {} current_contig_id {}, pixel range: [{}, cut({}), {}], left({}), right({}), smaller than ignore_len ({})\n", 
+            "[Pixel Cut] Warning: original_contig_id {} current_contig_id {}, pixel range: [{}, cut({}), {}], left({}), right({}), smaller than ignore_len ({}), cut at loc: ({}) is ignored\n", 
             original_contig_id%Max_Number_of_Contigs, 
             contig_id, 
             ptr_left, 
@@ -6325,7 +6424,7 @@ BreakMap(
             ptr_right, 
             loc - ptr_left, 
             ptr_right - loc, 
-            ignore_len);
+            ignore_len, loc);
         return;
     } 
 
@@ -6336,7 +6435,7 @@ BreakMap(
     {   
         if (Map_State->originalContigIds[tmp] > (std::numeric_limits<u32>::max() - Max_Number_of_Contigs))
         {
-            fmt::print("Error: originalContigIds[{}] + {} = {} is greater than the max number of contigs ({}), file: {}, line: {}\n", tmp, 
+            fmt::print("[Pixel Cut] originalContigIds[{}] + {} = {} is greater than the max number of contigs ({}), file: {}, line: {}\n", tmp, 
             Max_Number_of_Contigs, 
             (u64)Map_State->originalContigIds[tmp] + Max_Number_of_Contigs, 
             std::numeric_limits<u32>::max(), 
@@ -6347,7 +6446,7 @@ BreakMap(
     }
 
     fmt::print(
-        "Original contig_id ({}), current_id ({}), pixel range: [{}, {}] {} inversed, cut at {}\n", 
+        "[Pixel Cut] Original contig_id ({}), current_id ({}), pixel range: [{}, {}] {} inversed, cut at {}\n", 
         original_contig_id%Max_Number_of_Contigs, 
         contig_id, 
         ptr_left, 
@@ -6419,7 +6518,7 @@ void cut_frags(const std::vector<u32>& problem_locs)
     for (auto & i : problem_locs)
     {
         // cut the fragment
-        BreakMap(i);
+        BreakMap(i, auto_curation_state.auto_cut_smallest_frag_size_in_pixel);
         UpdateContigsFromMapState();
     }
     Redisplay = 1;
@@ -6508,7 +6607,7 @@ void AutoCurationFromFragsOrder(
             AddMapEdit(0, {start_loc, start_loc + current_order[i].second - 1}, true);
             current_order[i].first = -current_order[i].first;
             start_loc += current_order[i].second;
-            printf("[Auto curation] (#%d) Invert contig %d.\n", ++num_autoCurated_edits, predicted_order[i]);
+            printf("[Pixel Sort] (#%d) Invert contig %d.\n", ++num_autoCurated_edits, predicted_order[i]);
             continue;
         }
         else if (
@@ -6554,9 +6653,9 @@ void AutoCurationFromFragsOrder(
             start_loc += current_order[i].second; // update start_loc after move fragments
             if (is_curated_inverted)
             {
-                printf("[Auto curation] (#%d) Invert and Move contig %d to %d, start_loc:%d.\n", ++num_autoCurated_edits, predicted_order[i], i, start_loc);
+                printf("[Pixel Sort] (#%d) Invert and Move contig %d to %d, start_loc:%d.\n", ++num_autoCurated_edits, predicted_order[i], i, start_loc);
             }
-            else printf("[Auto curation] (#%d) Move contig %d to %d, start_loc:%d.\n", ++num_autoCurated_edits, predicted_order[i], i, start_loc);
+            else printf("[Pixel Sort] (#%d) Move contig %d to %d, start_loc:%d.\n", ++num_autoCurated_edits, predicted_order[i], i, start_loc);
             continue;
         }
         else 
@@ -6565,7 +6664,7 @@ void AutoCurationFromFragsOrder(
             continue;
         }
     }
-    printf("[Auto curatioin] finished with %d edits\n", num_autoCurated_edits);
+    printf("[Pixel Sort] finished with %d edits\n", num_autoCurated_edits);
 
     return ;
 }
@@ -6573,7 +6672,7 @@ void AutoCurationFromFragsOrder(
 
 global_function
 void
-Auto_cutting_func(
+auto_cut_func(
     char * currFileName,
     memory_arena* arena = nullptr,
     std::string file_save_name = std::string("/auto_curation_tmp")
@@ -6585,64 +6684,42 @@ Auto_cutting_func(
     {   
         u32 start = auto_curation_state.get_start();
         u32 end = auto_curation_state.get_end();
-        if (start < 0 || end < 0) return ;
-        fprintf(stdout, "========================\n");
-        fmt::print(stdout, "[Auto cutting] restoring the cutted frags within the selected area: [{}, {}]\n", start, end);
+        if (start < 0 || end < 0) 
+        {
+            fmt::print(
+                stdout, 
+                "Error: no area selected. start({}) or end({}) is less than 0\n", 
+                start, end);
+            return ;
+        }
+        fmt::print(stdout, "[Pixel cut] restoring the cutted frags within the selected area: [{}, {}]\n", start, end);
         Map_State->restore_cutted_contigs( start, end );
         UpdateContigsFromMapState();
         Redisplay = 1;
         return ;
     }
 
-    fprintf(stdout, "========================\n");
-    // copy textures to cpu
-    fprintf(stdout, "[Auto cutting] coping textures...\n");
-    auto texture_array_4_ai = TexturesArray4AI(Number_of_Textures_1D, Texture_Resolution, (char*)currFileName, Contigs);
-    // prepare before reading textures
-    f32 original_control_points[3];
-    prepare_before_copy(original_control_points);
-    texture_array_4_ai.copy_buffer_to_textures_dynamic(Contact_Matrix, false); 
-    restore_settings_after_copy(original_control_points);
-    
     // classic cut 
-    {   
+    {  
         SelectArea selected_area;
         auto_curation_state.get_selected_fragments(
             selected_area, 
             Map_State, 
             Number_of_Pixels_1D, 
             Contigs);
-        FragCutCal frag_cut_cal(
-            &texture_array_4_ai, 
-            Contigs, 
-            Number_of_Pixels_1D, 
-            &selected_area, 
-            auto_curation_state.auto_cut_diag_window_for_pixel_mean,
-            auto_curation_state.auto_cut_threshold, 
-            auto_curation_state.auto_cut_smallest_frag_size_in_pixel); 
-        std::vector<u32> cut_locs_pixel = frag_cut_cal.get_cut_locs_pixel();
-        #ifdef DEBUG
-            std::cout << "Cut locs in pixel: [";
-            for (auto & i : cut_locs_pixel) std::cout << i << " ";
-            std::cout << "]" << std::endl;
-        #endif
+        glBindBuffer(GL_TEXTURE_BUFFER, Contact_Matrix->pixelRearrangmentLookupBuffer);
+        const u32 *pixel_rearrange_index = (const u32 *)glMapBufferRange(GL_TEXTURE_BUFFER, 0, Number_of_Pixels_1D * sizeof(u32), GL_MAP_READ_BIT);
+        std::vector<u32> cut_locs_pixel = frag_cut_cal_ptr->get_cut_locs_pixel(
+            auto_curation_state, 
+            pixel_rearrange_index, // pixel_rearrangement_buffer
+            Contigs,
+            &selected_area
+        );
+        glUnmapBuffer(GL_TEXTURE_BUFFER);
         
-        // add the cut locs to the extensions
-        #ifdef DEBUG    
-            if (arena)
-            {   
-                u32 added_num = Extensions.get_num_extensions();
-                std::string graph_name = "hic_pixel_density";
-                u32* graph_data = new u32[Number_of_Pixels_1D];
-                for (u32 i=0; i < Number_of_Pixels_1D; i ++)
-                {
-                    graph_data[i] = (u32)(frag_cut_cal.hic_density[i] * 100.f);
-                }
-                add_graph_to_extensions(arena, (u32*)graph_name.c_str(), graph_data);
-                
-                push_extensions_to_opengl(arena, added_num);
-            }
-        #endif // DEBUG
+        fmt::print(stdout, "[Pixel cut] cut locs in pixel: [");
+        for (auto & i : cut_locs_pixel) fmt::print(stdout, "{} ", i);
+        fmt::print(stdout, "]\n");
         
         // cut the contigs
         cut_frags(cut_locs_pixel);
@@ -6651,12 +6728,11 @@ Auto_cutting_func(
     // generate figures
     if (0)
     {
-        fprintf(stdout, "[Auto cutting] generating hic figures...\n");
+        fprintf(stdout, "[Pixel cut] generating hic figures...\n");
         Hic_Figure hic_figure(
             Total_Genome_Length, 
             file_save_name,
-            &texture_array_4_ai
-        );
+            textures_array_ptr );
         hic_figure.generate_hic_figures();
 
         // run ai model with python script
@@ -6677,24 +6753,22 @@ Auto_cutting_func(
 
 global_function
 void
-Auto_sort_func(char* currFileName)
+auto_sort_func(char* currFileName)
 {   
     if (!currFileName || !auto_sort_state) return;
     
     fprintf(stdout, "========================\n");
-    fprintf(stdout, "[Auto Sort] start...\n");
-    fprintf(stdout, "[Auto Sort] smallest_frag_size_in_pixel: %d\n", auto_curation_state.smallest_frag_size_in_pixel);
-    fprintf(stdout, "[Auto Sort] link_score_threshold:        %.3f\n", auto_curation_state.link_score_threshold);
-    fprintf(stdout, "[Auto Sort] Sort mode:                   %s\n", auto_curation_state.get_sort_mode_name().c_str());
-    // compress the HiC data
-    auto texture_array_4_ai = TexturesArray4AI(Number_of_Textures_1D, Texture_Resolution, (char*)currFileName, Contigs);
+    fprintf(stdout, "[Pixel Sort] start...\n");
+    fprintf(stdout, "[Pixel Sort] smallest_frag_size_in_pixel: %d\n", auto_curation_state.smallest_frag_size_in_pixel);
+    fprintf(stdout, "[Pixel Sort] link_score_threshold:        %.3f\n", auto_curation_state.link_score_threshold);
+    fprintf(stdout, "[Pixel Sort] Sort mode:                   %s\n", auto_curation_state.get_sort_mode_name().c_str());
 
+    // compress the HiC data
     // prepare before reading textures
-    f32 original_control_points[3];
-    prepare_before_copy(original_control_points);
-    texture_array_4_ai.copy_buffer_to_textures_dynamic(Contact_Matrix, false); // copy the remapped data which is reordered according to the Map_State
-    restore_settings_after_copy(original_control_points);
-
+    f32 original_color_control_points[3];
+    prepare_before_copy(original_color_control_points);
+    textures_array_ptr->copy_buffer_to_textures_dynamic(Contact_Matrix, false); 
+    restore_settings_after_copy(original_color_control_points);
     // check if select the area for sorting
     SelectArea selected_area;
     auto_curation_state.get_selected_fragments(
@@ -6705,43 +6779,52 @@ Auto_sort_func(char* currFileName)
     );
     if (auto_curation_state.get_start() >=0 && auto_curation_state.get_end() >= 0 && selected_area.selected_frag_ids.size() > 0)
     {   
-        fprintf(stdout, "Using selected area for sorting:\n");
-        fprintf(stdout, "Selected pixel range: %d ~ %d\n", auto_curation_state.get_start(), auto_curation_state.get_end());
+        fprintf(stdout, "[Pixel Sort] Using selected area for sorting:\n");
+        fprintf(stdout, "[Pixel Sort] Selected pixel range: %d ~ %d\n", auto_curation_state.get_start(), auto_curation_state.get_end());
         std::stringstream ss; 
         ss << selected_area.selected_frag_ids[0];
         for (u32 i = 1; i < selected_area.selected_frag_ids.size(); ++i) ss << ", " << selected_area.selected_frag_ids[i];
-        fprintf(stdout, "Selected fragments: %s\n\n", ss.str().c_str());
+        fmt::print(stdout, "[Pixel Sort] Selected fragments ({}): {}\n\n", selected_area.selected_frag_ids.size(), ss.str().c_str());
         selected_area.select_flag = true;
         selected_area.start_pixel = auto_curation_state.get_start();
         selected_area.end_pixel = auto_curation_state.get_end();
     }
 
-    texture_array_4_ai.cal_compressed_hic(
+    textures_array_ptr->cal_compressed_hic(
         Contigs, 
         Extensions, 
         false, 
         false, 
         &selected_area
     );
-    FragsOrder frags_order(texture_array_4_ai.get_num_frags()); // intilize the frags_order with the number of fragments including the filtered out ones
-    // exclude the fragments with first two tags during the auto curationme
+    FragsOrder frags_order(textures_array_ptr->get_num_frags()); // intilize the frags_order with the number of fragments including the filtered out ones
+    // exclude the fragments with first two tags during the auto curation 
     std::vector<std::string> exclude_tags = {"haplotig", "unloc"};
     std::vector<s32> exclude_meta_tag_idx = get_exclude_metaData_idx(exclude_tags);
     LikelihoodTable likelihood_table(
-        texture_array_4_ai.get_frags(),
-        texture_array_4_ai.get_compressed_hic(),
+        textures_array_ptr->get_frags(),
+        textures_array_ptr->get_compressed_hic(),
         (f32)auto_curation_state.smallest_frag_size_in_pixel / ((f32)Number_of_Pixels_1D + 1.f), 
         exclude_meta_tag_idx, 
         Number_of_Pixels_1D);
+    
+    #ifdef DEBUG_OUTPUT_LIKELIHOOD_TABLE
+        likelihood_table.output_fragsInfo_likelihoodTable("frags_info_likelihood_table.txt", textures_array_ptr->get_compressed_hic());
+    #endif // DEBUG_OUTPUT_LIKELIHOOD_TABLE
+    
+    // clear mem for textures and compressed hic
+    textures_array_ptr->clear_textures(); 
+    textures_array_ptr->clear_compressed_hic();
+
     // use the compressed_hic to calculate the frags_order directly
-    if (auto_curation_state.sort_mode == 0)
+    if (auto_curation_state.sort_mode == 0 || !selected_area.select_flag) // sort with union find if sorting the whole genome
     {
         frag_sort_method->sort_according_likelihood_unionFind( 
             likelihood_table, 
             frags_order, 
             selected_area,
             auto_curation_state.link_score_threshold, 
-            texture_array_4_ai.get_frags());
+            textures_array_ptr->get_frags());
     }
     else if (auto_curation_state.sort_mode == 1)
     {
@@ -6750,7 +6833,7 @@ Auto_sort_func(char* currFileName)
             frags_order, 
             selected_area,
             auto_curation_state.link_score_threshold, 
-            texture_array_4_ai.get_frags(), true, true);
+            textures_array_ptr->get_frags(), true, true);
     }
     else if (auto_curation_state.sort_mode == 2)
     {
@@ -6759,11 +6842,21 @@ Auto_sort_func(char* currFileName)
             frags_order, 
             selected_area,
             auto_curation_state.link_score_threshold, 
-            texture_array_4_ai.get_frags(), false, true);
+            textures_array_ptr->get_frags(), false, true);
+    }
+    else if (auto_curation_state.sort_mode == 3) // not finished yet
+    {
+        frag_sort_method->sort_according_yahs(
+            likelihood_table,
+            frags_order,
+            selected_area,
+            auto_curation_state.link_score_threshold,
+            textures_array_ptr->get_frags()
+        );
     }
     else 
     {
-        fprintf(stderr, "[Auto Sort] Error: Unknown sort mode (%d)\n", auto_curation_state.sort_mode);
+        fprintf(stderr, "[Pixel Sort] Error: Unknown sort mode (%d)\n", auto_curation_state.sort_mode);
         assert(0);
     }
     AutoCurationFromFragsOrder(
@@ -6779,9 +6872,12 @@ Auto_sort_func(char* currFileName)
 
 
 // set the vertex array for the contact matrix
-// copy_flag: if true, show full texture for copy the texture to array on cpu
+// copy_flag: if true, set vaos to [-1, 1, -1, 1] for copying tiles from opengl buffer to cpu memory
 global_function
-void setContactMatrixVertexArray(contact_matrix* Contact_Matrix_, bool copy_flag, bool regenerate_flag)
+void setContactMatrixVertexArray(
+    contact_matrix* Contact_Matrix_, 
+    bool copy_flag, 
+    bool regenerate_flag)
 {   
     glUseProgram(Contact_Matrix_->shaderProgram);
     GLuint posAttrib = (GLuint)glGetAttribLocation(Contact_Matrix_->shaderProgram, "position");
@@ -6838,26 +6934,27 @@ void setContactMatrixVertexArray(contact_matrix* Contact_Matrix_, bool copy_flag
                 3 2
                 0 1
             */
-            textureVertices[0].x = !copy_flag? x               : -1.f;   // x -> [-0.5, 0.5]
-            textureVertices[0].y = !copy_flag? y - quadSize    : -1.f;   // y -> [-0.5, 0.5]
+            f32 half_len = 1.0f;
+            textureVertices[0].x = !copy_flag? x               : -half_len;   // x -> [-0.5, 0.5]
+            textureVertices[0].y = !copy_flag? y - quadSize    : -half_len;   // y -> [-0.5, 0.5]
             textureVertices[0].s = !copy_flag? cornerCoords[0] :  0.f; // s表示texture的水平分量
             textureVertices[0].t = !copy_flag? cornerCoords[1] :  0.f; // t表示texture的垂直分量
             textureVertices[0].u = texture_index_symmetric;
 
-            textureVertices[1].x = !copy_flag? x + quadSize    :  1.f;   
-            textureVertices[1].y = !copy_flag? y - quadSize    : -1.f;
+            textureVertices[1].x = !copy_flag? x + quadSize    :  half_len;   
+            textureVertices[1].y = !copy_flag? y - quadSize    : -half_len;
             textureVertices[1].s = !copy_flag? 1.0f            :  1.f;
             textureVertices[1].t = !copy_flag? 1.0f            :  0.f;
             textureVertices[1].u = texture_index_symmetric;
 
-            textureVertices[2].x = !copy_flag? x + quadSize    :  1.f;   
-            textureVertices[2].y = !copy_flag? y               :  1.f;
+            textureVertices[2].x = !copy_flag? x + quadSize    :  half_len;   
+            textureVertices[2].y = !copy_flag? y               :  half_len;
             textureVertices[2].s = !copy_flag? cornerCoords[1] :  1.f;
             textureVertices[2].t = !copy_flag? cornerCoords[0] :  1.f;
             textureVertices[2].u = texture_index_symmetric;
 
-            textureVertices[3].x = !copy_flag? x               : -1.f;              
-            textureVertices[3].y = !copy_flag? y               :  1.f;
+            textureVertices[3].x = !copy_flag? x               : -half_len;              
+            textureVertices[3].y = !copy_flag? y               :  half_len;
             textureVertices[3].s = !copy_flag? 0.0f            :  0.f;
             textureVertices[3].t = !copy_flag? 0.0f            :  1.f;
             textureVertices[3].u = texture_index_symmetric;
@@ -6906,14 +7003,15 @@ void setContactMatrixVertexArray(contact_matrix* Contact_Matrix_, bool copy_flag
 
 global_function
 void 
-prepare_before_copy(f32 *original_control_points)
+prepare_before_copy(f32 *original_color_control_points)
 {
     // prepare before reading textures
     glActiveTexture(GL_TEXTURE1);
-    glBindTexture(GL_TEXTURE_BUFFER, Color_Maps->maps[Color_Maps->nMaps-1]);
+
+    glBindTexture(GL_TEXTURE_BUFFER, Color_Maps->maps[Color_Maps->nMaps-1]); // 绑定到颜色映射的最后一个颜色
     glActiveTexture(GL_TEXTURE0);
     for (u32 i = 0; i < 3 ; i ++ ) 
-        original_control_points[i] = Color_Maps->controlPoints[i];
+        original_color_control_points[i] = Color_Maps->controlPoints[i];
     Color_Maps->controlPoints[0] = 0.0f;
     Color_Maps->controlPoints[1] = 0.5f;
     Color_Maps->controlPoints[2] = 1.0f;
@@ -6926,12 +7024,12 @@ prepare_before_copy(f32 *original_control_points)
 
 global_function
 void 
-restore_settings_after_copy(const f32 *original_control_points)
+restore_settings_after_copy(const f32 *original_color_control_points)
 {
     // restore settings
     setContactMatrixVertexArray(Contact_Matrix, false, true);  // restore the vertex array for rendering
     for (u32 i = 0 ; i < 3 ; i ++ )
-        Color_Maps->controlPoints[i] = original_control_points[i];
+        Color_Maps->controlPoints[i] = original_color_control_points[i];
     glUseProgram(Contact_Matrix->shaderProgram);
     glUniform3fv( Color_Maps->cpLocation, 1, Color_Maps->controlPoints);
     glActiveTexture(GL_TEXTURE1);
@@ -7533,10 +7631,6 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                     break;
 
                 case GLFW_KEY_V:
-                    if (Select_Sort_Area_Mode)
-                    {
-                        auto_cut_state = 2; // restore the cutted frags within the selected area
-                    }
                     break;
 
                 case GLFW_KEY_W:
@@ -7558,6 +7652,10 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                     break;
 
                 case GLFW_KEY_Z:
+                    if (Select_Sort_Area_Mode)
+                    {
+                        auto_cut_state = 2; // restore the cutted frags within the selected area
+                    }
                     break;
 
                 case GLFW_KEY_SPACE:
@@ -7643,6 +7741,10 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                         }
 
                     }
+                    else if (Select_Sort_Area_Mode)
+                    {
+                        auto_curation_state.change_sort_mode(-1);
+                    }
                     else AdjustColorMap(-1);
                     break;
 
@@ -7660,15 +7762,28 @@ KeyBoard(GLFWwindow* window, s32 key, s32 scancode, s32 action, s32 mods)
                             }
                         }
                     }
+                    else if (Select_Sort_Area_Mode)
+                    {
+                        auto_curation_state.change_sort_mode(1);
+                    }
                     else AdjustColorMap(1);
                     break;
 
                 case GLFW_KEY_UP:
-                    NextColorMap(1);
+                    if (Select_Sort_Area_Mode)
+                    {
+                        auto_curation_state.adjust_cut_threshold(1);
+                    }
+                    else NextColorMap(1);
                     break;
 
                 case GLFW_KEY_DOWN:
-                    NextColorMap(-1);
+                    if (Select_Sort_Area_Mode)
+                    {
+                        auto_curation_state.adjust_cut_threshold(-1);
+                    }
+                    else 
+                        NextColorMap(-1);
                     break;
 #ifdef Internal
                 case GLFW_KEY_ESCAPE:
@@ -8496,24 +8611,28 @@ resources, click each entry to view its licence.)text";
 global_function void
 UserSaveState(const char *, u08, char *);
 
+global_function
+void
+default_user_profile_settings () ;
+
 global_variable char
-    profile_savebuff[1024] = {0};
+profile_savebuff[1024] = {0};
 
 global_function
-    u08
-    UserProfileEditorRun(const char *name, struct file_browser *browser, struct nk_context *ctx, u32 show, u08 save = 0)
+u08
+UserProfileEditorRun(const char *name, struct file_browser *browser, struct nk_context *ctx, u32 show, u08 save = 0)
 {
     struct nk_window *window = nk_window_find(ctx, name);
     u32 doesExist = window != 0;
 
-    if (!show && !doesExist)
+    if (!show && !doesExist) // 不显示 且 不存在
     {
         return (0);
     }
 
-    if (show && doesExist && (window->flags & NK_WINDOW_HIDDEN))
+    if (show && doesExist && (window->flags & NK_WINDOW_HIDDEN))  // 显示 且 存在 且 隐藏
     {
-        window->flags &= ~(nk_flags)NK_WINDOW_HIDDEN;
+        window->flags &= ~(nk_flags)NK_WINDOW_HIDDEN;  // 取消隐藏
         FileBrowserReloadDirectoryContent(browser, browser->directory);
     }
 
@@ -8527,11 +8646,28 @@ global_function
 
         nk_layout_row_dynamic(ctx, Screen_Scale.y * 30.0f, 1);
 
-        if (nk_button_label(ctx, "Save"))
+        if (nk_button_label(ctx, "Save") && window && !(window->flags & NK_WINDOW_HIDDEN))
+        {
+            ret = 2;
+        }
+
+        if (nk_button_label(ctx, "Default") && window && !(window->flags & NK_WINDOW_HIDDEN))
         {
-            UserSaveState("userprofile", 1, 0);
+            default_user_profile_settings(); 
+            ret = 2;
         }
 
+        // invert mouse setting
+        nk_layout_row_dynamic(ctx, Screen_Scale.y * 30.0f, 4);
+        bool original_value = user_profile_settings_ptr->invert_mouse;
+        nk_checkbox_label(NK_Context, "Invert Mouse Buttons", (int*)&user_profile_settings_ptr->invert_mouse);
+        if (original_value != user_profile_settings_ptr->invert_mouse)
+        {   
+            fmt::print("[UserPofile]: Mouse_invert settings changed ({}) -> ({})\n", original_value, user_profile_settings_ptr->invert_mouse);
+            ret = 2;
+        }
+        
+        // extensions
         nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 4);
 
         if (Extensions.head)
@@ -8748,7 +8884,7 @@ global_function
     }
     nk_end(ctx);
 
-    return (ret);
+    return ret; 
 }
 
 
@@ -9033,7 +9169,7 @@ SaveState(
             settings |= Grid->on ? (1 << 3) : 0;
             settings |= Contig_Ids->on ? (1 << 4) : 0;
             settings |= Tool_Tip->on ? (1 << 5) : 0;
-            settings |= Mouse_Invert ? (1 << 6) : 0;
+            settings |= user_profile_settings_ptr->invert_mouse ? (1 << 6) : 0;
             settings |= Scaffs_Always_Visible ? (1 << 7) : 0;
 
             *fileWriter++ = settings;
@@ -9578,7 +9714,7 @@ LoadState(u64 headerHash, char *path)
                 Grid->on = settings & (1 << 3);
                 Contig_Ids->on = settings & (1 << 4);
                 Tool_Tip->on = settings & (1 << 5);
-                Mouse_Invert = settings & (1 << 6);
+                // user_profile_settings_ptr->invert_mouse = settings & (1 << 6); // this is moved to the user settings, then this setting will not be influnced by clearcache
                 Scaffs_Always_Visible = settings & (1 << 7);
 
                 nBytesRead += 2;
@@ -9936,7 +10072,7 @@ restore_initial_state()
         Grid->on                 = 1;
         Contig_Ids->on           = 1;
         Tool_Tip->on             = 1;
-        Mouse_Invert             = 0;
+        // user_profile_settings_ptr   = 0;
         Scaffs_Always_Visible    = 1;
         MetaData_Always_Visible  = 1;
 
@@ -10203,6 +10339,47 @@ UserSetSaveStatePaths()
     }
 }
 
+
+
+
+// default user profile settings
+global_function
+void 
+default_user_profile_settings()
+{   
+    // metadata tags color profile
+    meta_outline->on = 1;
+
+    // background color
+    active_bgcolor = 0;
+
+    // color map order
+    useCustomOrder = 1;
+    userColourMapOrder.nMaps = Color_Maps->nMaps;
+    for (u32 i = 0; i < Color_Maps->nMaps; i++)
+    {
+        userColourMapOrder.order[i] = i;
+    }
+
+    // select area color
+    auto_curation_state.set_default_mask_color();
+
+    // invert mouse
+    if (user_profile_settings_ptr)
+    {
+        user_profile_settings_ptr->invert_mouse = false;
+    }
+
+    // Grey out settings
+    if (Grey_Out_Settings)
+    {   
+        delete Grey_Out_Settings;
+        Grey_Out_Settings = new GreyOutSettings(Meta_Data);
+    }
+    return ;
+}
+
+
 global_function void
 UserSaveState(const char *headerHash = "userprofile", u08 overwrite = 1, char *path = 0)
 {
@@ -10289,6 +10466,30 @@ UserSaveState(const char *headerHash = "userprofile", u08 overwrite = 1, char *p
     // save selected area color 
     fwrite(&auto_curation_state.mask_color, sizeof(auto_curation_state.mask_color), 1, file);
 
+    // save grey out flags
+    if (Grey_Out_Settings)
+    {
+        fwrite(Grey_Out_Settings->grey_out_flags, sizeof(int32_t), 64, file);
+    }
+    else
+    {
+        int32_t grey_out_flags[64] = {0};
+        fwrite(grey_out_flags, sizeof(int32_t), 64, file);
+    }
+
+    // save user profile settings
+    if (user_profile_settings_ptr)
+    {
+        fwrite(&user_profile_settings_ptr->invert_mouse, sizeof(bool), 1, file);
+    }
+    else 
+    {   
+        bool invert_mouse_tmp = false;
+        fwrite(&invert_mouse_tmp, sizeof(bool), 1, file);
+    }
+
+    // END of SAVE user profile settings
+
     fclose(file);
 
     printf("[UserProfile]: Saved to: %s\n", path);
@@ -10411,6 +10612,28 @@ global_function
         auto_curation_state.set_mask_color(auto_curation_state.mask_color_default);
     }
 
+    // load grey out flags
+    int32_t grey_out_flags[64];
+    if (fread(grey_out_flags, sizeof(int32_t), 64, file) == 64)
+    {   
+        if (Grey_Out_Settings)
+        {
+            memcpy(Grey_Out_Settings->grey_out_flags, grey_out_flags, 64 * sizeof(int32_t));
+        }
+    }
+
+    // load user profile settings
+    bool invert_mouse_tmp;
+    if (fread(&invert_mouse_tmp, sizeof(bool), 1, file) == 1)
+    {
+        if (user_profile_settings_ptr)
+        {
+            user_profile_settings_ptr->invert_mouse = invert_mouse_tmp;
+        }
+    }
+
+
+
     fclose(file);
     return 0; // Success
 }
@@ -10491,7 +10714,14 @@ CopyEditsToClipBoard(GLFWwindow *window)
 global_function
 void
 GenerateAGP(char *path, u08 overwrite, u08 formatSingletons, u08 preserveOrder)
-{
+{   
+
+    // 这里粘贴所有的original contigs
+    u32 tmp_orignal_contig_ids[Number_of_Pixels_1D];
+    for (u32 i = 0 ; i< Number_of_Pixels_1D; i++ )tmp_orignal_contig_ids[i] = Map_State->originalContigIds[i];
+    Map_State->restore_cutted_contigs_all(Number_of_Pixels_1D);
+    UpdateContigsFromMapState(); // todo 检查粘贴后会不会影响 painted 的 scaffolds
+
     FILE *file;
     if (!overwrite && (file = fopen((const char *)path, "rb")))
     {
@@ -10596,6 +10826,11 @@ GenerateAGP(char *path, u08 overwrite, u08 formatSingletons, u08 preserveOrder)
 
         fclose(file);
     }
+
+    // 把临时切开的再粘贴上
+    for (u32 i = 0; i < Number_of_Pixels_1D; i++) Map_State->originalContigIds[i] = tmp_orignal_contig_ids[i]; 
+    UpdateContigsFromMapState();
+    return;
 }
 
 
@@ -10677,7 +10912,8 @@ void SortMapByMetaTags(u64 tagMask)
 }
 
 
-MainArgs {
+MainArgs 
+{
     u32 initWithFile = 0;   // initialization with .map file or not 
     u08 currFile[256];      // save the name of file inputed 
     u08 *currFileName = 0;
@@ -10870,7 +11106,7 @@ MainArgs {
         if (auto_sort_state)
         {
             if (currFileName) SaveState(headerHash);
-            Auto_sort_func((char*)currFileName);
+            auto_sort_func((char*)currFileName);
             auto_sort_state = 0;
             Redisplay = 1;
         }
@@ -10878,7 +11114,7 @@ MainArgs {
         if (auto_cut_state)
         {   
             if (currFileName) SaveState(headerHash);
-            Auto_cutting_func((char*)currFileName, Loading_Arena);
+            auto_cut_func((char*)currFileName, Loading_Arena);
             auto_cut_state = 0;
             Redisplay = 1;
         }
@@ -10952,7 +11188,7 @@ MainArgs {
             {
                 static nk_flags flags = NK_WINDOW_BORDER | NK_WINDOW_SCALABLE | NK_WINDOW_MOVABLE | NK_WINDOW_SCROLL_AUTO_HIDE | NK_WINDOW_TITLE;
 
-                if (nk_begin_titled(NK_Context, "Options", currFileName ? (const char *)currFileName : "", nk_rect(Screen_Scale.x * 10, Screen_Scale.y * 10, Screen_Scale.x * 800, Screen_Scale.y * 350), flags))
+                if (nk_begin_titled(NK_Context, "Options", currFileName ? (const char *)currFileName : "", nk_rect(Screen_Scale.x * 10, Screen_Scale.y * 10, Screen_Scale.x * 800, Screen_Scale.y * 600), flags))
                 {
                     struct nk_rect bounds = nk_window_get_bounds(NK_Context);
                     bounds.w /= 8;
@@ -10985,7 +11221,7 @@ MainArgs {
                     showUserProfileScreen = nk_button_label(NK_Context, "User Profile");
                     showAboutScreen = nk_button_label(NK_Context, "About");
 
-                    // AI & Auto sort button 
+                    // AI & Pixel Sort button 
                     struct nk_color sort_button_normal = nk_rgb(153, 50, 204); 
                     struct nk_color sort_button_hover  = nk_rgb(186, 85, 211); 
                     struct nk_color sort_button_active = nk_rgb(128, 0, 128);  
@@ -11034,10 +11270,10 @@ MainArgs {
                         }
                     }
                     */
-                    // auto cut
+                    // Pixel Cut
                     {
                         bounds = nk_widget_bounds(NK_Context);
-                        auto_cut_button = nk_button_label(NK_Context, "Auto cut");
+                        auto_cut_button = nk_button_label(NK_Context, "Pixel Cut");
 
                         if (auto_cut_button && currFileName)
                         {
@@ -11076,7 +11312,7 @@ MainArgs {
                             nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 2);
                             if (nk_button_label(NK_Context, "Cancel")) 
                             {   
-                                printf("[Auto Cut] Cancel button clicked\n");
+                                printf("[Pixel Cut] Cancel button clicked\n");
                                 auto_curation_state.set_buf();
                                 nk_contextual_close(NK_Context);
                             }
@@ -11084,14 +11320,14 @@ MainArgs {
                             {
                                 // Apply changes
                                 // Convert text to integer and float
-                                auto_curation_state.update_value_from_buf();
+                                auto_curation_state.update_value_from_buf(); // todo 这里还有问题，因为设置了pixel_mean之后 并没有重新计算 pixel_density...
                                 
                                 nk_contextual_close(NK_Context);
 
                                 auto_curation_state.set_buf();
-                                fmt::print("[Auto Cut] cut_threshold:               {:.3f}\n", auto_curation_state.auto_cut_threshold);
-                                fmt::print("[Auto Cut] pixel mean window size:      {}\n", auto_curation_state.auto_cut_diag_window_for_pixel_mean);
-                                fmt::print("[Auto Cut] smallest_frag_size_in_pixel: {}\n", auto_curation_state.auto_cut_smallest_frag_size_in_pixel);
+                                fmt::print("[Pixel Cut] cut_threshold:               {:.3f}\n", auto_curation_state.auto_cut_threshold);
+                                fmt::print("[Pixel Cut] pixel mean window size:      {}\n", auto_curation_state.auto_cut_diag_window_for_pixel_mean);
+                                fmt::print("[Pixel Cut] smallest_frag_size_in_pixel: {}\n", auto_curation_state.auto_cut_smallest_frag_size_in_pixel);
                             }
 
                             nk_contextual_end(NK_Context);
@@ -11100,9 +11336,9 @@ MainArgs {
                         }
                     }
 
-                    // Auto sort button
+                    // Pixel Sort button
                     bounds = nk_widget_bounds(NK_Context);
-                    auto_sort_button = nk_button_label(NK_Context, "Auto Sort");
+                    auto_sort_button = nk_button_label(NK_Context, "Pixel Sort");
                     {   
                         if (auto_sort_button && currFileName) 
                         {
@@ -11121,7 +11357,7 @@ MainArgs {
                                 sizeof(auto_curation_state.frag_size_buf), 
                                 nk_filter_decimal);
 
-                            nk_label(NK_Context, "Link score threshold (default: 0.4)", NK_TEXT_LEFT);
+                            nk_label(NK_Context, "Link threshold (default: 0.4)", NK_TEXT_LEFT);
                             nk_edit_string_zero_terminated(
                                 NK_Context, 
                                 NK_EDIT_FIELD, 
@@ -11145,10 +11381,15 @@ MainArgs {
                                 auto_curation_state.sort_mode = 2;
                             }
 
+                            // if (nk_option_label(NK_Context, "YaHS", auto_curation_state.sort_mode == 3)) 
+                            // {
+                            //     auto_curation_state.sort_mode = 3;
+                            // }
+
                             nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 2);
                             if (nk_button_label(NK_Context, "Cancel")) 
                             {   
-                                printf("[Auto Sort] Cancel button clicked\n");
+                                printf("[Pixel Sort] Cancel button clicked\n");
                                 auto_curation_state.set_buf();
                                 nk_contextual_close(NK_Context);
                             }
@@ -11161,9 +11402,9 @@ MainArgs {
                                 nk_contextual_close(NK_Context);
 
                                 auto_curation_state.set_buf();
-                                fmt::print("[Auto Sort] smallest_frag_size_in_pixel: {}\n", auto_curation_state.smallest_frag_size_in_pixel);
-                                fmt::print("[Auto Sort] link_score_threshold:        {:.3f}\n", auto_curation_state.link_score_threshold);
-                                fmt::print("[Auto Sort] Sort mode:                   {}\n", auto_curation_state.get_sort_mode_name());
+                                fmt::print("[Pixel Sort] smallest_frag_size_in_pixel: {}\n", auto_curation_state.smallest_frag_size_in_pixel);
+                                fmt::print("[Pixel Sort] link_score_threshold:        {:.3f}\n", auto_curation_state.link_score_threshold);
+                                fmt::print("[Pixel Sort] Sort mode:                   {}\n", auto_curation_state.get_sort_mode_name());
                             }
                             // redo all the changes
                             if (!auto_curation_state.show_autoSort_redo_confirm_popup)
@@ -11181,7 +11422,7 @@ MainArgs {
                                 if (nk_button_label(NK_Context, "Yes")) 
                                 {
                                     auto_curation_state.show_autoSort_redo_confirm_popup=false;
-                                    printf("[Auto Sort] Redo all edits.\n");
+                                    printf("[Pixel Sort] Redo all edits.\n");
                                     RedoAllEdits(Map_Editor);
                                 }
                                 if (nk_button_label(NK_Context, "No")) 
@@ -11205,7 +11446,7 @@ MainArgs {
                                 if (nk_button_label(NK_Context, "Yes")) 
                                 {
                                     auto_curation_state.show_autoSort_erase_confirm_popup=false;
-                                    printf("[Auto Sort] Erase all edits.\n");
+                                    printf("[Pixel Sort] Erase all edits.\n");
                                     EraseAllEdits(Map_Editor);
                                 }
                                 if (nk_button_label(NK_Context, "No")) 
@@ -11216,7 +11457,7 @@ MainArgs {
                             nk_contextual_end(NK_Context);
                         }
                     }
-                    pop_nk_style(NK_Context, 3); // pop the style for auto cut and auto sort button
+                    pop_nk_style(NK_Context, 3); // pop the style for Pixel Cut and Pixel Sort button
 
                     // waypoint edit mode
                     nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 3);
@@ -11506,11 +11747,49 @@ MainArgs {
 
                         nk_contextual_end(NK_Context);
                     }
+                    
+                    bool original_value = user_profile_settings_ptr->invert_mouse;
+                    nk_checkbox_label(NK_Context, "Invert Mouse Buttons", (int*)&user_profile_settings_ptr->invert_mouse);
+                    if (original_value != user_profile_settings_ptr->invert_mouse)
+                    {   
+                        fmt::print("[UserPofile]: Mouse_invert settings changed ({}) -> ({})\n", original_value, user_profile_settings_ptr->invert_mouse);
+                        UserSaveState();
+                    }
+                    
+
+                    // select current active tag
+                    if (nk_tree_push(NK_Context, NK_TREE_TAB, "Select Active Tag", NK_MINIMIZED))
+                    {
+                        nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 20.0f, 4);
+                        for (u32 i = 0; i < ArrayCount(Meta_Data->tags); i ++ )
+                        {   
+                            if (*Meta_Data->tags[i] == 0) break;
+                            int tmp = ( MetaData_Active_Tag == i) ; 
+                            tmp = nk_check_label(NK_Context, (char*)Meta_Data->tags[i], tmp);
+                            if (tmp) MetaData_Active_Tag = i;
+                        }
+                        nk_tree_pop(NK_Context);
+                    }
+
+                    // grey out settings
+                    if (nk_tree_push(NK_Context, NK_TREE_TAB, "Grey Out Tags", NK_MINIMIZED))
+                    {
+                        nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 20.0f, 4);
+                        for (u32 i = 0; i < ArrayCount(Meta_Data->tags); i ++ )
+                        {   
+                            if (*Meta_Data->tags[i] == 0) break;
+                            s32 tmp = Grey_Out_Settings->grey_out_flags[i];
+                            nk_checkbox_label(NK_Context, (char*)Meta_Data->tags[i], Grey_Out_Settings->grey_out_flags+i);
+                            if (tmp!=Grey_Out_Settings->grey_out_flags[i]) 
+                            {
+                                fmt::print("[UserPofile]: Grey out tags changed\n");
+                                UserSaveState();
+                            }
+                        }
+                        nk_tree_pop(NK_Context);
+                    }
 
-                    Mouse_Invert = nk_check_label(NK_Context, "Invert Mouse Buttons", (s32)Mouse_Invert) ? 1 : 0;
 
-                    nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 2);
-                    Grey_Haplotigs = nk_check_label(NK_Context, "Grey out 'Haplotig'", (s32)Grey_Haplotigs) ? 1 : 0;
 
                     nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 2);
                     nk_label(NK_Context, "Gamma Min", NK_TEXT_LEFT);
@@ -11597,7 +11876,7 @@ MainArgs {
                         static int tree_state = 0;
 
                         // Tree tab for checkboxes
-                        if (nk_tree_push(NK_Context, NK_TREE_TAB, "Select Tags", NK_MINIMIZED))
+                        if (nk_tree_push(NK_Context, NK_TREE_TAB, "Select Tags Move to End", NK_MINIMIZED))
                         {
                             nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 20.0f, 4); // Adjust layout for checkboxes
                             for (int i = 0; i < NUM_TAGS; i++)
@@ -11610,7 +11889,7 @@ MainArgs {
                         }
 
                         nk_layout_row_dynamic(NK_Context, Screen_Scale.y * 30.0f, 3); // Layout for buttons
-                        if (nk_button_label(NK_Context, "Sort Selected"))
+                        if (nk_button_label(NK_Context, "Move Selected"))
                         {
                             if (!sortMetaEdits)
                             {
@@ -11625,7 +11904,7 @@ MainArgs {
                                 SortMapByMetaTags(tag_mask); // Pass the tag mask to the sorting function
                             }
                         }
-                        if (nk_button_label(NK_Context, "Sort All Tags"))
+                        if (nk_button_label(NK_Context, "Move All Tags"))
                         {
                             if (!sortMetaEdits)
                             {
@@ -11633,7 +11912,7 @@ MainArgs {
                                 SortMapByMetaTags(all_tags_mask);
                             }
                         }
-                        if (nk_button_label(NK_Context, "Undo Sort"))
+                        if (nk_button_label(NK_Context, "Undo Move"))
                         {
                             while (sortMetaEdits)
                             {
@@ -12083,7 +12362,9 @@ MainArgs {
                 u08 state;
                 if ((state = UserProfileEditorRun("User profile editor", &saveBrowser, NK_Context, (u32)showUserProfileScreen, 1)))
                 {
-                    UserSaveState("trial", state & 2, saveBrowser.file);
+                    if (showUserProfileScreen) 
+                        fmt::print("[UserPofile]: showUserProfileScreen = {} \n", showUserProfileScreen);
+                    UserSaveState("userprofile", state & 2, 0); // save the user profile settings
                     FileBrowserReloadDirectoryContent(&saveBrowser, saveBrowser.directory);
                 }
 
@@ -12134,6 +12415,35 @@ MainArgs {
     // and memory allocated  by new or malloc
     delete[] Contact_Matrix->vaos;
     delete[] Contact_Matrix->vbos;
+    if (Grey_Out_Settings) 
+    {
+        delete Grey_Out_Settings; Grey_Out_Settings = nullptr;
+    }
+    if (user_profile_settings_ptr)
+    {
+        delete user_profile_settings_ptr; user_profile_settings_ptr = nullptr;
+    }
+
+    if (pixel_density_extension)
+    {
+        delete pixel_density_extension; pixel_density_extension = nullptr;
+    }
+
+    if (textures_array_ptr)
+    {
+        delete textures_array_ptr; textures_array_ptr = nullptr;
+    }
+
+    if (frag_cut_cal_ptr)
+    {
+        delete frag_cut_cal_ptr; frag_cut_cal_ptr = nullptr;
+    }
+
+    if (frag_sort_method) 
+    {
+        delete frag_sort_method; frag_sort_method = nullptr;
+    }
+
     ResetMemoryArenaP(Loading_Arena);
     fprintf(stdout, "Memory freed for the arena.\n");
 
diff --git a/install.cmake.bat b/install.cmake.bat
index 72f4906..574aa5d 100644
--- a/install.cmake.bat
+++ b/install.cmake.bat
@@ -83,7 +83,7 @@ if exist app (
     echo "Removed existing app directory."
 )
 
-cmake -DCMAKE_BUILD_TYPE=Release -DGLFW_BUILD_WAYLAND=OFF -DGLFW_BUILD_X11=OFF -DWITH_PYTHON=OFF -DCMAKE_INSTALL_PREFIX=app -DCMAKE_PREFIX_PATH=%cmake_prefix_path_tmp% -S . -B build_cmake
+cmake -DCMAKE_BUILD_TYPE=Release -DGLFW_BUILD_WAYLAND=OFF -DGLFW_BUILD_X11=OFF -DCMAKE_INSTALL_PREFIX=app -DCMAKE_PREFIX_PATH=%cmake_prefix_path_tmp% -S . -B build_cmake
 if errorlevel 1 (
     echo "CMake configuration failed."
     goto :error
diff --git a/install.cmake.sh b/install.cmake.sh
index fedc78c..05efad3 100755
--- a/install.cmake.sh
+++ b/install.cmake.sh
@@ -86,7 +86,6 @@ CMAKE_OPTIONS=(
     -DCMAKE_BUILD_TYPE=Release
     -DGLFW_BUILD_WAYLAND=OFF
     # -DGLFW_BUILD_X11=OFF
-    -DWITH_PYTHON=OFF
     -DWITH_TORCH=OFF
     -DCMAKE_INSTALL_PREFIX="$install_path"
     -DCMAKE_PREFIX_PATH="$cmake_prefix_path_tmp"

From aa2a41b59f8ed6e63a48893362b41101ec3904da Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:30:11 +0100
Subject: [PATCH 28/33] include <memory>

---
 src/yahs_sort.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/yahs_sort.h b/src/yahs_sort.h
index 8affb60..3ffa9ea 100644
--- a/src/yahs_sort.h
+++ b/src/yahs_sort.h
@@ -5,7 +5,7 @@
 #include "likelihood_table.h"
 // frags_order
 #include "frags_order.h"
-
+#include <memory>
 
 
 // 定义图相关的结构体

From f742a208d0fa40acbc33661eb03c4a40069fbd8b Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:30:21 +0100
Subject: [PATCH 29/33] update readme

---
 README.md                  |  26 +++++++++++++-------------
 doc/image_mainGUI.png      | Bin 137187 -> 778881 bytes
 doc/image_user_profile.png | Bin 80079 -> 500972 bytes
 3 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/README.md b/README.md
index 2b7d25c..343f95d 100644
--- a/README.md
+++ b/README.md
@@ -16,7 +16,7 @@ PretextView is a desktop application for viewing pretext contact maps.<br/>
 - `M`: enter the Meta Tag edit mode.
 - `S`: enter the Scaffold painting mode.
 - `X`: enter the extension edit mode.
-- `W`: endter the Waypoint edit mode.
+- `W`: enter the Waypoint edit mode.
 - `J`: window jump to the diagnol line without changing the zoom-level, which is usefull for selecting the correct place for a small fragment.
 - `L`: open / close the Grid.
 - `T`: open / close the Tooltip.
@@ -50,38 +50,38 @@ Enter scaffolding mode with the 's' key.
 
 ## Select sort area mode (`F`)
 
-After click `Auto Sort` button in the main UI, it will defaultly run sort globally. If enter the the `select sort area mode`(by pressing `F`), it can sort only the selected area with pressing `Space` after selecting at least 2 contigs. 
+After click `Pixel Sort` button in the main UI, it will defaultly run sort globally. If enter the the `select sort area mode`(by pressing `F`), it can sort only the selected area with pressing `Space` after selecting at least 2 contigs. 
 
 Enter the select sort area mode by pressing `F`.
 - `Left mouse click`: select / un-select area for sorting.
 - `S`: clear all the select area.
-- `Space`: call `YaHS sort` to sort the select area **(NOTE: only work if select at least 3 fragments)**.
+- `Space`: call `Pixel Sort` to sort the select area **(NOTE: only work if select at least 3 fragments)**.
 - `Q/W`: quit/redo edit (currently will change the edit made globally not only the edit made within the selected area, so use this with caution as it can change also other parts)
+- `C`: cut the scaffolds/contigs within the selected area.
+- `Z`: glue the splitted fragments within the selected area. (NOTE: only link the continuous fragments.)
 
 ## Sort fragments according to link score
 
 Usage, first endter the main GUI: 
-- **Right click** the `Auto Sort` button to open sort settings, 
+- **Right click** the `Pixel Sort` button to open sort settings, 
   - `Smallest Frag Size (pixels)` (defaul: 2) which represents the smallest fragment size in unit of pixels to consider during the sorting
   -  `Link Score Threshold`: $S_{thresh}$ (defaul: 0.4).
   -  Select the sort mode: 
      -  `UnionFind`: sort all links higher than $S_{thresh}$ and then traverse all the links by from high to low. 
      -  `Fuse`: traverse all links and fuse two chains.
-- **Left click** the `YaHS Sort` button: run.
+- **Left click** the `Pixel Sort` button: run.
 - **Redo all edits**
 - **Erase all edits**
 
-## `Auto Cut` according to the continuity of HiC density
+## `Pixel Cut` according to the continuity of HiC density
 Usage:
-- In the main GUI:
-  - **Right click** the `Auto Cut` button to open cut settings,
+In the main GUI:
+  - **Right click** the `Pixel Cut` button to open cut settings,
     - `Cut threshold`: the smaller, the more likely to break one contig at the discontinuous locations.
     - `Pixel_mean window size`: length of range from diagonl line, which is considered to calculate the HiC density.
     - `Smallest frag size`: the smallest fragment length that can be generated.
-  - **Left click** the `Auto Cut` button: cut all the scaffolds/contigs globally.
-- Under `Select sort area mode`:
-  - `C`: cut the scaffolds/contigs within the selected area.
-  - `V`: glue the splitted fragments within the selected area. (NOTE: only link the continuous fragments.)
+  - **Left click** the `Pixel Cut` button: cut all the scaffolds/contigs globally.
+
 
 
 
@@ -166,7 +166,7 @@ to remove the **quarantine** and then unmount and re-mount the dmg file. And the
 
 The score between two fragments represents the level of continuity. The higher score the better continuity.
 
-The link score calculation method refers to [YaHS](https://github.com/c-zhou/yahs). Thus, we call this function as **YaHS Sort**.
+The link score calculation method refers to [YaHS](https://github.com/c-zhou/yahs). 
 
 Reference: Chenxi Zhou, Shane A McCarthy, Richard Durbin, YaHS: yet another Hi-C scaffolding tool,
 Bioinformatics, Volume 39, Issue 1, January 2023, btac808, https://doi.org/10.1093/bioinformatics/btac808
diff --git a/doc/image_mainGUI.png b/doc/image_mainGUI.png
index 3cdb4d7082a43ec881cc05185a58b7a09b443069..badd4072405d4ee1468c86d1e6edbdf5c2720d05 100644
GIT binary patch
literal 778881
zcmb5Vby!@z);ByexD_eIWpMZ6P$=#$gF6&=cUmZ3+^u+VcgjG4;?Cewyi9Q@4j=b9
z=Xu_9z29GNt}BzZldP3VvUZY{-`Y`XDzaGUB<KJD083s@N&^5umH_||n^2Koa(LsL
z1_1z|g{`Edn!KbWSk2AZ+Sb7e0ANhCFg4whXJs5RGcz?E8fRfbck|YWjEvPV4IYN}
zK_{TY1%0L&8HR?NL^zuWT77^b_1;!{%vkVU1N%@FrQz$e3A3P4`ir5JkDk#bf~9`=
z(9|UT9L(ZpK)t-%dv!Dfgd$)9xtg&Cz?>hjL@g~Y3tZ9^|Kow`j})1UXm$duL_>2!
zZr}hcfrg2rn8g1gfDIVX<VDD%a<etFFg+25u0+21e(e?42Sr*D84?*12ZX4FQf2f;
z_L4=$5rK$vTo1T_a98Ug8%EP|OGfh{OF5e%Q%2yS7h*I%=MHBvXekfV7jqcX4^v3I
z1@|KmCp)Z<L=+R(Je-n<h?O2#Z6^~sFv!X2fN&SHPb8t#jd76s4mpBG6&)`I?-$kR
zO<ak<_#$fG-ueU-Ds=SxJoyU~^G_BgW^nEE^Rxf`^K)=I7boY2xv9V@7C->g;sNhF
ztm|deqpcwF*2>BN<`)_jfCwZ3AiYq)m-r4y`oFXckP(3JuY3>y@YxoC_@6o|FX=xa
z;U)gV^Y0WPDFT4<@`U#ieF{MTQyW>P0O5aV#HN=rz<W(edHI)A)56Wl%E|qMvxiaJ
z0OLyrnv0yCI{-jH_fG`MYtWv)$X~S8f_Olbm4qyu9ofw+oz1P-eH>l>kpmF%5qcpV
ztvt-YK8_Af?m|AI)c>j>^g{oW%|Q+RR}~L?QEG^?8d%cV%?ixN&dJV6{SF-r28*~^
zS_^4N$^3`?<w=zKgNKKU5C@01w>P^t54*FQ4F{K?pdbe)HwQO2+e-~LcV8zDGaoi5
zcbb0-`5!q_R_+#Vwk{sF&Q9Qe<eHf~dwPgcQ~%S@|Av46Pb(kW|I?C_`+vIi(m{@Y
zN;tUKIXV7U_6w`XKe<9`wmw!4dQ!HIFJt!7#yc)<&bK1}V)*|R{hucPixu);tbCk&
z|IPYeMgKRew!4*^q_g8oqaN@6&xHMl`M(SQ!zjY>Pv8G*EdD*v|H^$C>38TN9RE9O
z@6dZ^apGPklFU|0RqG{vQL}#nNdM*MMIru4Um`NoFRZ~LX#hYRATRY^%Ln);7tNGb
zy8D@B*fL4JVfeVA4!rNi0(SGEVk~s?fot|>?}ws9x4!E|xccoM{6-Du1Ovb8WdCAI
zG=C%zJq%@t4b=H3QJGubnm+UST2nOj^3vmaRklI1?)UHG3X{h}-8Jp`9_PoPp1kT?
zrS-YYz45O0n~9aNQP`0Ulj~E&<YT|cUPj=!3d?7u!q%T;osSfru$i2#bf??Tf^NI)
z<5cV53c{yr@3DE^RcCtH7XHkuY_|D%ls`_Te9D<0^B!E1$9Kj?>WgpJbc1Z}biZVN
zCU2^2zxU?P?jLz$9VBGo<vX@+?PUiWQ<}(#5~h6XPwrmOex%Ev@zLbe=c0AgnFod9
z(dA-tWg$CJ`^Vk{{Fu|-pv;l{)^LK>8)DUB(%RjAGSWkwJemD5?^(co7Q3eI`sd%P
z9lqwE)twfljJb)G?B04Rp0{b-&s~Q)+54x-d8SS`wWMQIlOp|w?q`%^i$R@P)vaqb
zEw7)mojW%$j~a6I8sT>Clg-{l$+<l4v$NI<MXxxojqhySZxrLX7mP3cdBn20mas_2
zMt)M-m7WAnRSEAUT5CA=JSW+YIZX=nt@;mNS-Z4;&SXnBF+*-Ku;*Lz>#oigqp|nm
z3jBtfywox6@cr4V5k5W1HS4|R<TjgZ?Sn0vuNsij{*29^wG_N=AMw|}P0xRE-Fh@I
zTiN;tLF4fO!gy}aV{&G826rHhWNBH?#>{w}GjAY^%-9fxz4Bczn{^PkqhFG|KS%!X
zqQyqES@MdItsfqm|M%Hn{+YoF5udwn<outYM%bA-caGbR*ct0uGnnzOGNV?WZBOEk
zA31q~;9_G%CVnaHyV$K+#PeJYD}SxGPOdX0=_W~^@4s08MDXIhCvPjdJz8LYY5B8q
z?IH1+*&J;W+QnEd&)zS8k-k4|@Vqo*mgk49P7`j1oSWq|?b}>`CmxPv$NaQ;U1jN>
zbr;>?wuRw`juxc*)0#2s#@l81n2KLsU*0z{<De+tIb2oGftdLB`qKQuJH@zxZjU<%
zdsB(z+3SVQCJ)5!Gm7i@2C5MMpo;pFCE=qt2JzXG{u8!)vSNa*4&SY<Gr07UIlJ!`
zbai=qJ>`x1?O4fRguF@227l6iV_9EG8jOFuUMIeNmqs@O#VpHtmb@6y%<mCD+s9rv
z%d2m>a-RVqDBqkxyS<5?)*79RmagF~OL5lup+{LzX}Elgonepb;d}Sf)8$gWI=zz*
zY}|!oR9}Ll9i0~qSG3-Hpq{z8v3QGB5I?v6rf|PJT>Inr?98quB!+C8o?V%B<$!d%
z2%8ukzCx~+ugyK-5<Y1x^SxX4tl#n-kY(p2{QjIu=Q}d5u2B$%F%IKL+`tMZO^}}>
z`3{0*XVDFLYqHqmo^_pR*^8Ez{o-9uC*>Dr7$i~QkKcT=eVzI9%Mn?6$sf_YUdhYq
zgYU$@z4wG+YU&#aR#ALZKYp@z59S2z6jc`H@VHy+Y_R{ZuB&j%%@R}!o@O@|_Wfg*
zqLkKmtuE~AuuC@0lb1hE>FMH$AYMfAi9NRlu><cwp@1dy|AAP-Yr7n@C#*OC(O?uF
zJpI42vxM$4hRT^*UVAoo7H1~kS4Cx3=?My`C66`S3eIk5=EGyNn4L7f%4Dif?Cy{6
z%%v2s*q?WIm-Ym!%jmV|k&wx4{u&uF$jnr*8J|+?$;=$jylwY=RLN?;?UMg{mBn-w
z*U2ZyW#K2qlK6)wkXfMm;;FYdGe4jIR&bXvZk@s?KZp0@5Bc`ZtY(cv+}uxxqa$pD
zOjLww#W|~<iGI6-Xz*#{-H$M6fvk;6kAFd#pJ@xjw9R?Dpw*w|%l(0ELlxhCMU}3~
zx{_RF!ozUg55F@qv+A!M-Z6)JndE1tuW;OirmtJ}WKLSfk<`M4P6+cW#PY}W7j!NJ
zEfX8L4~`To0%}p^L+*N<q3-Fa2|4?8F%HYgZ?x$GxU=%Jc>?X6^Or#z#O+M(=NXzk
zW0iqZyYHgxeSPMg4>OWU*|Pfy9qSgX2i(76UccUXo!ez8NkT&U_P1-hpeDu65(7n|
zr;Ou7Wpy`F`>T>WC1$~bj_#JlKDN8$yj8mh85sjDgVro-IdBy=56`S$p3k~s{<q7^
z6k-Le&kcAoEN?Ax3Cs2y=;s%l=^Kgf-@Sr+^JrG?vn;1u<m4>c>C_m{HkB&dIgp|K
z-Y(T{Q$lTgn*Ol*7iP)3!PQESqUm{|angvBGCUSAf%(_3h7N9BqXLTCvQjx2C1aN*
z1v>{PAEtKar#*Y?Zx5*<3=ZnV!Fz3nHqum{$?0kMRi)2VGM{?i$I<F<XbZl&7@x*$
zWmE6sJM>f82twOvHOW|g@J#JInIygHR6j}op7AH^d#)Sy2uH>Ez=9~h*x#bEo52PF
zi|(v>?<JRpGbcy8<W>e4Q(VuuVfJ7*sjK~sm$x6;#?fB2LGUi#qKK%4PD$a0U|#Ph
zk5BEHdY#sV8z@a)c7r4pZZ(ST6L!@?Q&jfaZT52k+RHNsEW3ByRhI2n%8*Qj8qf1Z
zZM5g{vDpfnk5{{o*P^C6M@`-G&A-+@z{p%%`$7`3K(Q(C8$W{AyB|2N?+wc6#|*#g
zq6hgJFPL*w6M-d*m=HL_`T~`DfRv>+`-rTTW6;x(%-mCPqy%&nB5;}3S3`{g)DXnW
z(qaeM>30I$rraQUfPMs0<d}8w!--_-q!<ufQCVwZ8i9YxGsk8XkdJ@S2tqP6Z^s&%
z;kcXKa(5XhK$1e1MRv@24L}chIqIT8-?OtMh=Zk6B{s1`qCViUAea`5@tVRjbtyt3
zx|SKnZGdBHa48m!ghFqyZ7<L~;4%)=z|_*oOCK5<LbG;~$PER7nT7K+)k(EgtX?B9
zVb|5&YJE=W*=dcksP!Qu!7rz-v-N_};tM0*5)IdY3CM3FkD3dVo)hQ;LrDNMPnPdr
z4UaVk8l$g%PdQ}YA=Pn6`*zl}pbM(kVOk$;+gauiRHGDs9Vjy*uS2`@$_PM%=cWV#
z%j@yHtqTZElwtC&n(7EvpvB<>4zMY;5ta~P7HMmkNpkp{3~r4ssckvZ?lrTB3Qi+Y
ztSn;W8hbY4U+pfFw+LVd8Aj|f$+qfUlMX0$?yRx4Ybmr|ZTpqvc&dEYx0E8rmoByo
zO`Ndx4a`YNHdwlU6B2B86714dQTry>*sr}NS){#@e*B=*!4j-5zw%pDE54_0pZ9nz
z(%?sWW#!WK_KlO*n$cZ#a>%`ohwNSV>(E3Uxnk{Xc-qJB#mPTwbVOKOJwH`j^_}Zx
zaa6G`0Axy=(5j3`8M1k0-{KSLRdgml<bKy$=FHT=ab`N+$3S<f61i(S$WGNDH~si2
z90f)t1X;U}trAK?yh`KFg*R%zuoZ1pMo_2Z%2#vdPm@hHqr_3pcN-es5-Vj8=bPuR
zsm<}Qgo{S|9fE;lroAAsWP{T2(~*>23?W0BvvfbTJ$Ed?=eN3Q#|(&!(EF4sd1Rym
zB~>L1%*2wI4;DzFC341>o74KCYLMGx5LoUsyjk99uu&fvnJEwFC%^|Kx&3Y;CNeT>
zm~Dh5A#x&4g+b4YyI5uKA|gahCB-pe^pAF8;HjJs`ihv-z|$e2Q6g5}Il>)vN&J*R
zTqR&Rz;eX=ktz>Zm#|`5VIn*K&0p{=><SH_5d2luLAakM>iiK@P-+E~66~<m@%&Qm
zsAn%)n7f@s3B?sB1iazSC{ih5jKTGq{jvx+4}aCvPI_Oa_c1+Rkstawkm%-u_x!p|
zyq+smc|`&p{-Z5C%3vpAm_Akq@_ozRHyP%yw=%IegZ*@lp%%xcbO$1s1u-nt=;@8A
zpyZY?{6vav;yW<!%u+q|7=LwCu@1YR($SN&O%8e1V<q=~O98)4JSrG@>+BHs`qPTu
zhMaU@j<&QKa&I&KLmhmrCz%AkKs=uQV>v*4xls8P%gWo*h%<wj0acn^@`avUui9l-
z(PN_;tJ#$6hkEtmj%&8#%vU1<H?fE(>B6--GE)`-ltg3=#g#}hK0R`Q%MY`QDGTxQ
zgTXkRo`wkvZf+PolU*ijjW|`U{>GVzhk+bn&HZs$y<Mq!jr5Lb19BxIpLQ`uY95x~
z#zuRy381+1KCGRd|E&JiO?Gq--xs!9@SVHTaev&DG}!4j8>)i79^h8#wQ0sx&dp7H
zCv%Z~Z(Y10j7#JS|1HzhPW-lln8(OqbvoNy?;}U<ZNrOFTVFa0C<odo5JOh(YPjGu
z*O=iK3#*k+f`y5^y!b@6v=&DCXLuPnoJlDX1}f(0n_($ijJv4-0%S{1pj;T8tW0!p
z3T|Ut@z!_zKgBut`w;I-7O{~lT3ILp5M#oPw%c>C!vaKfpt~-gi!H~rlR~P8e@g{L
zy|gK7v=&!*sH#$OIgt`Az6IU%0l`!(b!I-SyX~K50`c!@s}xUBkWsfnaBV!f!1l?p
zAr#_R*@$EbcZhPALxlq%h6?_#cdmqKQ&*VcQ9%}&)1x(1sYsWA>n8Sg`ce@%j-KwN
zDS<Vc9Id8a<fQ_v8zkoDwjvEOt(Q%0C16f!-q*pXlc7&<Xv%ye2V~rWzFu{3RZU&u
z4k2GfA?g)PvoR~|(8flK!8c5#>dXnjH!yCfyGAEq{rDs4BNW>(VA{Fubc?cSj@ABs
z*qiu%oI{1tY^6k_3;KR*EMpxU$A{#^&d<e@Ym%gTwtn+wfkdsGt}&~egMA;ucUivg
ziC#C0mih9qsUfxM&(ps{lymB;Y!SIFN)c_ILTAy;7Wu*GS*uMSs*pdRaG|YZU)?Tb
zqJFZoqrceIJ~WTxB^713f+EA!>~>yrDM_s~tZk@iliM6j<~FJBIjPE$YTrB*$=z%a
zJ4c^<uBy{`5V2sW;<S7GS<|RPH+^|=3N0z#ptdEs$C@$w@hHllUr3w&l=5e5PhjpT
zDJM(lLNrNFQS0utuMBi(6AvJh=1tk`_0G9pcxBi^mu6X)IId*DIkma6!BE{@s#_?%
z6i8G3uvo72X_SQ!iV8#a)r*5nz1IN)a0Ho#>*c_qZ^o1q<vG-7Xu<=&tnkuHQGaU#
z3o>o1mm|>P(Ddtz%VnzG+J=%a?x}CRZUxc38}Z!>#J>z7>7P5$X8mHE+6p5p{AqtU
zz)bZ;7u$0ZEd>aHDifGuhM=N9u)bvz4-F4bi-Z_3Vqfp2qHF?$Xl)(=`LrJ(8^47Z
zN|BZH(#;^BBUPdCg3Y}hD|pV`ZT$^Rrt9tuX{o#l5}OjRFzO0LbS*=dUHl*6QukmO
zjcLDrO1a1p&F;RmW=(AYZ+>R8Q{5)%J4cGgJ&J-?fr(6~;R&0P=?);vvk~0a(&Eu~
z0N-ShAS*U%lC}x`1hi|`AsRTraHI`XzxY63<LE0{$ZK2i^WVuVp-!>5Eb+(rB>%G$
zE(q2%*?p)p)(}o(>ndrU`k<Uy_x)S*o(kkam;%s9#u+~Ft8Q>uZyY<2B-j8=PAF((
zc&^&l4yXjz9PwAR!w=rYzf#}Lausc>bX3U3iKvXtp2GZIAgErKSEb$l>&ERoZ}vi{
z{EFB%(K^6HkR;drlSz_6wU3QiaJTD+0f}|%cKdQ0Sx=&kW~0h#N_*l;CQTO3uz8xp
z4xM7_EB>zZWP`KE5+3i?LX$KN^7Yjvd+STb`GZBxDtY%ideRD-LAf~|W2Qy&b;9O2
zx?fESwFEt+1J~QI(*eOsv6v+7l9a-my`#S}P1bSt<44$m?TUy4T?;Ig3RcTew_VQC
zj;A@9zYmCLoB}ueh|}J2RPTSiu=B7W^y@~UT>Rj*IvvwE+}VF^uHYvXoYSz^YWn+9
zljdi(%uGegm%_bByWc*yg!de%v79~EvHLa`vg8jRvRm^u;murxCL>G=L%Z+zK2om4
zhw+CJsEQvOH~;hsjFoQ2GNt7xpyB6W`cRLu7($1fh6jn=EL0G~i1w^Q$;FonH6Lc4
z01|lQE>gCQO89@72M#kId&V~-MHT#6?UjD5;J^CTEVY2yalKd^n^nWFU1t-R6^ORW
z2H}BF6qJYvPWOVsfw4amx!V%^8W2>OLTE!o05t4QQCLDXZf!FSy=&U*d>XqzdxttN
zt|%mw5aa?>0$Pf|W=YRQ@JWZt{hj3z7sCSiV>F?J5$q085fjvkaZr%Eb8!GP#$T=k
z8+Jz(hv2kd<AG<3I)DLtbg$2az=J@6CahCi6=-U+4=%Z-ET<;`8{?JwmnZelY#bC!
z&Q{cSA$0xWUtihqZX@Mi-(e81z%OM(KL>@H=v|?4r}lS1BlD7-DwIKHuM-S1KDA$;
z5_D3ClU8xX>Ts&wPhQX$k?&M%K^K<y^mm7c;Coa0a3|^Z*uj<Js=oyEbvjzhp`KE}
zx4U=KOov4wBX(K(`Jl+GjJR&Wn8tEG*V=hBW4y6(&knm$v_F{~h$gKyv2%v9Ja(cB
zT`knkGlgT_7b8zPMGzY@5;DDMeCdF0_=!OKf<))4Fd^NV{-E8+9C|F`Ml0iYp+~=z
z)m0y@05KD_Dt_|`Mv7YIR86*;P3Drds9U>0%Dxqic7Ghz2?s3}vF`)bx{m>K+zbP8
z96`%wJ)}&Xb;`s8Co)_psUyX0TJ{r#H-GQSa8JB^gjX~0g81+3XzsZ>(gxSQmE?|p
zDEiOg(8!io<$PYfq9nhf(#&?xn5?sFTsCzKIPPF8>MkxCz`LDW6_I~)XKx^;tsgu`
zrf%yS{4Q8H3_6sO_^xdlD#wO_CHA9tBQA%Cblf0e!8*6Btjr-G(KnQ)AC#H4i&QVF
zw$~{@IYF+Y0|<{(rCQKxE9g@|gqlYXQADf%!m~;a!BWsHOiwZwjnj_CKVn<}zOB;-
zx8Kk_{N4H#St|a9Q1KJ^C=Y|>8zQ<XhOaz;+IE@}For|KDY@x$yHEQH6~+W9z_CEo
z)jw{`a}$L)o2pYASFJFG5`)fJO~r`~?Gvk{V~}VN-zCU;Cj4PRZ%a&-{w$8W98(fL
zUIY&jf-OsOBN1h$7k-jJwEDRmQE*h%<L{DB3+|uMGQD0eeu$1H_(m8a74rI)tgSzB
z^MZB<1-vyA7Q;EB_GS5?eL+JQkU;wr2z)<lm%XcfiyV>UHoL3$F?Ixr*p@_1!3`Nq
zW7>&ih`79X!+f}(`1KvnVo$)wT47(WFXM^gi-=8aoJ^kTy9DiKtDmOPM;9jSuZ}jI
z+4dFR=QYbAx<k^J->i<$Z;V*1W)%~>{v`vVneVqiP0kEH8-DaU=0o4E#a$p_I*~b@
zWhSYn_`KZuvC%6EMolv^_bPU7!k&)Kp7=|F!I8hoKilU03f_Qwv&gNjS8%H$=oq8k
z6)M-MKU3wmy&l)b8_;<3Eyam!BX#CrE?^u{rD(8KB^9Zo>g$|~d@L?sM%$wP{P}nM
z1?2e_SA)rommRe8zI28Qn(xe16P(2>ix`S~*ZAv2B4yGpL@YTfM{0Qaho(Xy5lR~>
zeU+q~Akz*|0f4PjbqNwQZ+`CvMe6(v<X^>emr?Msp_)N~eD#0jU}j^bf7}wOn}ko4
zkXXqKz&1~D`NUoG-DMkz9SGh$thYjn#gzYrEuPeGE^Vr8?KY?{FA$R+#b%qV|5bcf
z<J2Gqa~XWsmLPHVWgd0Zyh;+lk#mI#4sTpmD;fD!B`G3J%HF-F#gWEmS65}WqocIX
z7a^xFf?4LU=>-=W9Z^o>eMOeI^2JwqF@#q1VPYaHv|gQsfmN~#a9lwt2+DIAKhr6X
z$)KTl-a?_E1w+ZIbO837xjTAUkz!{t1CiaOVoYJf2%sphp0HVjyaW_)>kG&a#xF6w
z62u>{FDJa&WN-%g5;Fb(1W0X<{%u4)K!*sBzL_vxdl0bAZE)(Zh_}A2jxzGqVb_1&
zs!hw?N?n+X&GBvY^@CL@usK;X*3cQse<}=;xtMdYD;Ds_vUv*kz#2f1K6f)ob%a1?
z{CtWWB-~wTR$Dzn6UF)YZ|1sNub9OSeT@y1x5SmwfA%ETrPXI~KYsiO!($i|nw_7G
z-)`c=E{XiPu%o5ow0c5LVmEZZvQs+u{`+5ut*6#<U!m{CrsTHuQC4ch&9L@-dTQ6G
zD}0`Suc5f2jr`Zsm4bY40*0Aeq`&veDeNs<{7FJ7$2Ua)jbPw4zj=3kF$ueM>)3(6
z%ar@)S0{go=WcUqEasJZP}pQv-Hnd!ap<zIvRksqb-u`~9UGf~@D6tB?a?Bg_39R@
z1PfU?8y=nfGrkTLG`o2h_vRL8t|*vG(1@4KEjT)!!@i*nU?sBPuhPn&Irbtj{UjHV
z2*5HUh9v=5Ku_ZeXL)^CsUb+>(S|C=>6gNEr{?-#WMz;85rGRgotn8?0iv)74@ew9
zZ#d;OQp(T;-G_CjKtC4aU-mOm;eP5;4MB6tsq+H=MYF26tiM8-0z~5Y7^1Dq_hV4u
z28b3lFy6623+_H(qrou@?Qo^m@>Y6cHc``vcYp_Vn~{eaO!mWO?@^{ff+0PaFis@!
z`=K~uHGiAIG9`(c<~_s&ANFsE9YYFM6v5Z;T>1y8dv&XkxJ9yd7}ik;sOFG1ft<4p
zo1ZP=3!WjkK8SU71@R+GK8$vRA{I<x{%oogg%+=ma5j?4N;o=T;aPWNrk^_M8SCq1
z(9fGh)Jid8AxoambL^pYXpBx~eWLpvHr(l4>+!)SdasJKg>BZ7UR3P~vkkjyMA1WI
z!cAP_=23?E5T2T^Y3Hvop3T&rbC^Oz6L<KV(f9Lk4IYwdhJT98*|MC_nrnI2Tl+*@
zs|>|VQDH2ivpug#l8ykl{iM2z({a$Eu#SIquFlLovJ0t_P+aHTtt<Of7D7||)nOBn
zV>pssH`MX*{F1TFn<kcmTOMxUw|nZcc+_GXv)x%8%Uf`6tskp3Vq7F1p#XQk4Rc4H
z1Ve*cepti`Jf;6|6&y=;5;<w&Rp-h7vODD*m!jIZ7uf(EI)?vM7hVk;;~i$QU-Rvj
zmF=1KfbUkikE)^)4UcLRdQ-W4OL^Alm_%1{VXt;D_2dQGgDx*^+e`cOE$x^sEFu;&
z2tmj$7Dza~>4d&a*qk6};^=1dNX8qtrIdk+j{yqq9r3#$S1Hu<?kV>z+YeF)J@{9R
z$&KfNY)~l?pcZB&1zrJQ_$&&e8W<L@*y~(P?}6=t7*uV~3`9m`Y$agP?>FNz*97`(
zwjuxl9C*vQDqV5K0Ls<K@hCCE+K&xwGl!JK=jIszd?0Q4IGWyu<sgIQZz>i-r00uw
z;va~whV^AS703l<_`y?h`^E{LK+2IlacZj;7#Ym`(?<B;<s14Cp+ihdK%2~JLX|j?
zp{KeHWdYuCTOBg44^ML1gX8LJ>`Q<+)iG+WL?G-jeP^#qsp_6>c$uETCX`oRFBX&T
zv8GA?dV<&f^qgeraRnXi)0Nq3-zWc9MD+-fRx_k@eIIyBvEuwoNq^VWb2-M8&WQNR
zEG%^h`);!#Z>KmnU;49I5g2p_U3^+HdlOnJEOz_mD)o=c7kl+UjEjOT7P6tJ%3!8d
z$@QzYPzbss=63!3wwlYnEs<)keI5&c4SN-&(hH7XtT*oa0*okC*EET2#>EN7992hb
zxcRnP^H?|KW2uaPJvroHXmz_L&uz1y16vv19(Q}mf&D13Cp=#kD8>nR_DNxR%U{3M
zPW2e2zBcR;Ay(Dqg{_;KG@Y_tXFkpurUJu^)sN7j;;r3y%(#@cnJkumTa2RmhZ^7Q
z#Vw!heAS|#ZRUbnS(93hiMocWHaWa@e-2||tZrOidFP``tG{jKC2n?a-co;gGC4Bj
z9~j5I0`$BA(qLLb<TvBL36ksRq1*`5{JGvL!1MC<H9{c5H3UNoHry+V_srhc5oHM5
zA|u~bndi2VLm~<sybC}&n1})lzk<+kJ^|qU_fV|as#X2Oo48$+=m*?Tve(<uDAZFR
z&X1dWjQhzN$Jphs#EB1k?WoH9`<J0sS`pA+AaG=;1gIZ?Po*54hlD2eQ(qJnpGXT6
zi_qi~A27aPdospY0uYVx-i1RN6APM|1I12?lhi24Rco&@{IXbVk7(X{D`l@Hj@`S=
z#wa1UmaPl+VxT<C$iDzY80#sG55GYJry$7sy}`_Ix7Y4de{n-#2H;<~kt6<yN;K2U
zJK*<J{Ql5lDi0(gmYPVSAz*?)cdXv6|C}IC9niO=sy((qLjb?5UP%~k*adH^iKvH`
z^pU=8e850k&#_Wwrh`_ztXu6K&srS+vu=rTZpDq~xe3SWT#==wEed3|)ZEM9xhh=F
zxS8pSXxig%JR#XZCjOdKYUe4MeJ@s(UqKs+S76*HIv>6AclE!p{sC)R3TnFWQjlqG
zgu#PMY2}I_^u3)W`}2Dor3t!qo&!VrF5R<wHD-n@tMwVzBa>S@<i#4!M(_Q4FF3*A
zj5788=6K1tiXI)I@nZxrlYiY$fM{a#S(+#2?B*wBjCvl=m+L%7^sDZTif*DhyRMj3
zMu*w{eBYQEIs4_a!uig&g_6aKjc81A-5i5ox5HlwWOi<sV#!<*7v`(8A^0QFn2kWt
zTsdpkKxxCajkBUS-Y<nEhWF>Q-CrJpzfJ=2Td-4h)y{a=fYz9@61CdkRg7)IKnhZs
zBBO?gRJ&DP!=JbYuy#MMjp$e4uNxn0n##L#(HEBuIBHp!`<a!EaM7L@e>i{V{>;nH
zRyuH^QY*IGVJuYo;Y3u%$(BX8ywi1@{PBVHd}^sj$oEf$tQ$r7aAVyKrqpyLfobkc
z<Dfeq(-hYSUcqnsHpPG;pGnC+rQeyAmQHb7wMz>xcgb3)v5ci*cZ|m+4lRI=%NRON
zsuc3=j}9_te2BD|sB;-9sVHdnF~oY|n6DIxIi!Wah%lnIFMUTE=x|H`Cer~Bob=;k
zv?z7B5CAa^9f?pkK<h&6?ET%F$lM1=?4l0=^NA`X|8h})b{#PRpD~%>0LTI078H*I
zadOixs3bSoL5~Qp5UfqDxW5(rf<ypRE`QL}wELW<Y+|IVg|JcNSC}lYnp3Mur$WR5
z3lQ3K2*2SLF$=!`_><8?vfQty7_U*bH?9f*zj6~<C_080suFnjjU{<Zym9+v*5p{Z
z5UYZv84o9IofH92o9=8u*c-GEspn|i$$`vez|b^Hra77fXj4uuZz;+)IpL4HCik(<
zV+dChr7r{0rvyN1l_F=Tz*u#NUWgK8)1`?^7hQ*u+7w<&7J?fn8%qSu54HS4EMF2(
zL*y}Adj8Rw%3}PpA9AgoSoxZi2~VRo{$&6HP0!#^U22!^jMB5j+bZmB0{YeL*~XkX
z_iBbtMm`CmYK0qHt2{4u{K_@7Zv364k?(L=Gb1o>J4J=G`F=u-o)xCC!~vDz)4n0x
zac?-aoE>*lNN(hzAO764N|~zYdhW1Kq!oWscfaP!sKh+5VE3}olX1>q@{7;c)IUV8
zj=KdXO<5PbdAxF#G41ULDhlXv=GS3;o12<!(0)&cXy1{*kG=YwKy1tdOZdz8<FGDA
zV{YS*-_LZreA3`G*B<rL#D__(?k<ThyuV`ycc-;3=`qgPTJf))`m2AfuzeQzzA|pW
zbLia9^U5Tya5k1NVtR+zvRZ0&)pwkxTj&~r=IQ0?f7ifsJTPCE;l*#Of4V$kUttvS
z_MC5^A5I@;Rq?BG&G!Y1Mjsih&g|PE&uM&w8M|(v^a>SDeC7IXI};&ykY1A#XPY_`
z_<`CHRu!4gQIw~;OcFurBTltCD3;y(SKL{qW9+E^3x?J7%Sn{lly=DUy6X`yBP;Li
z?5xHnoJw~1)ew{&U}%D6|9Me_a_`QRxV2uwgFU<>Zk!6%4Z#<MnS%!IBDX#&jU1&m
z3rydJRAF0}_)zwfypIk%N3a_^<j1!B5`;iusvNeJn?anXvC*g`Ux32~py^+QKECFX
zDb@a2&dQskkOyG^&~$+rpI^HLO1t>L`4qvUT~Gk4X<ZsrdX5!FsPSZwuHE4TOakGA
zRk9KmkiiL;(Q$ym5RJ{_Q4mYp*BQYmuW<wo(>V1yfnP!*HVmbgEY2Eta@mQ}5lCij
zh3Uz6?jTt?Sgtm9P16U2f_Uf~(hQ?R)9F?M#+u$H!&}x21h%WhYEn}vv}<F$@IwM9
zT{!n}i!?O8K_9??{fy}Z$u=b~d_aJVqh|Dlea@iE+f9g|WK~Z|Najo_g~V#TbvgVI
z^r4vaV9aX}@A37&aI{tP1$(FZ(Hu^FMJ-nHUa#u&SpBJ{LiMKs6P81F&!$Ra-C*JP
zc?f69Hiv_mNlk30>QYd33L%?J@o9?op4*Dctemg#lItH2J#Fq+yq#F{BKh+>SJ#rF
zJuHdp3GZGMSEE|Y(fLA#Csekzt@@aTe^KT{WZBVNusyYt(%7U~mvjkc2VYM+=a-$R
zl5e~-C83xqt$(GUVhdwU+xovO<UM!*TAQ=9ck#v}RIaSKV97p1#ooE>eD=jGE*l<0
zkeFiPGtY(9z*1{4UTotgaw4>Qh*SDvx<+>+kL>_aV%Yb(6+AV>5D-r2TiOJm8X~7D
zgn$kvA*B%21VuBn$1-ySI||oSMKs5_>stZktdST^i6utNN(wBi<<Nos-xv}P&Y+;R
zA2VS>8XyVO&0|ng(GUYYYLhrtnU!r4UWjC0Y?77$YS(EP6hXzbJItWQHWVmGI$3;U
zbwIcQQQs6AL4vfNNg>9Xrg9f}ZSReCBxtN5G|2-%;-D*af(X1V;+#HxW7AxX1adB!
zy9kj_9>(=@LkNrS9dN6pKea(BF`-}Ds|X9aRjpS=<&QiPqOv@Tq_=nMMnN=FI0@7I
z<!vmpAZ~z&R*z>CT?9Y}55R@w(oapkl?~T%nbH6`G;Crmq$WtFdG!^P{2C^J6PcHo
zdYZ$*pYJ?e>K1}<5XGvhwG+<EWgZ<>>|mR3CaQPLG%uE48#@Q;u6!;27O`zgX&#^3
zv}w_26TYKvi8o=(z!YThLnA>4lILXtzkE6!wP6~XopiqThSNej7a2xJ0zN+Q@-O#U
zFm>Cc{#{PtQ@&K(ONGRKnl`N1H1Z}xhmUW)zz!eW%nh?s7U{|!xiPqqVt&iy^4jYL
zYXy#X7^^s$UGv*_cPr@ps&Zl>wqYLD{H-!-J^WR%)eEgaZgwGGPrYsjp7Py+mvc+^
zk5_}+w0~K8gz(C8{>TKA-tfz6+k$(6v92l;5z$QSt?w6|l9{u}TuU#h(Y&LiBa~`p
zm<sP9{%x%Brfm%4d@*Xjd8i}v=-_P-v#q_$pL15rx{UKxDYgMV?U~~`5i>`uP6C}Y
z`yFLDj;I)UR>%}W5%S;8vYNj}pXu#kOpdJ1IwBKwo+etPmF??Zro|-It*KoT9tqf%
z8jF@8OJ>F+*ExCXDJgp*AqG~CiTF1mC8(PSPEB%cGa!5bDixNw{MU_~PLaP6PXO!7
zD0*=Pg<D3hxO--1H4qWiWI_mD&Bcm-1!LOh4J<860?G<X9)IMj;*{%5keE&;hko1A
z67_2iRso%%!%(3pLhb@SK)I#WnfL=HD_8@>hq~po=Ip9$RksC50b&xVn|iaVMSkXH
zB{WZVqq8=00}4g*+nWIsLNL+Y0BRw~HBA8B0_i@DfxUY0r}LPa=06F%D&GMF<1BlT
z2(S0%RQt(*tW>zy;RT0(TW`jcDk(8wq=uhO#4&u%p_FTB%_vuzt&NujU05Ww4+Lt{
zf{&)REJHKyl_g$Lx|0&pI@BYdkg!7{yV-A3^<%}?GVYJNCu5J=x&vkU@_0CO1B0`9
zeZpM_gp-pU9;|Q7ZiROBd6RW{d}=HkKFW?|f7B7{4UftfV_lBWR3s!iYb}`#UbqX=
zp1DqKS_}N(Zkc6DW_#^!gtoDJ-sY7$x-fJfcbF`lcTQi49Ui+^S$^%x<$P;$^|#VS
z#t5QR-r+SQP?5TeuQkT@$?BpLgi)j&Hz3!2Ui_qBtM!88E2kP(dXiOq@@n?&`eJz^
zxJ_)K3~$w{NHZ=vU!X3!+-c)`@X7Y9K$3ZI`=CX-{Jtz_weRHOszCtO^zp<gw^_bG
ze~KEF{_{9Xm2N#zFA1fU{r3LyO^Zxi&vvxDAm3)YLwuRq%G#;-kD^_ZAF^k<2Gv*T
z=Qqc!rp2?CT0JQ%s=a|M=dAQ7D9yg!eNN1o*N%I=-4q=KRKKHS(49yhPuhA+9(wF;
zbMtG04Q=~<ew7(vI9D?^;`>Owda57IvwscIh<@Bn@kt{$o$m0eccM-;5x(8Pu2nj9
z=B%+i4GP>>cQ;_O2zh>d1)}NM2&nx{JUO}dMd)TaZF_CuNRx7ZZALi}kSx#6ajoH2
zCfz`QuTfL9NomB?)LroKQ^HazSw1U+M)JI@I3-g%2UkxA5M5l&!lDZ7*^E{g@rOSt
zYz|h1ARDUaDPp8Hn}$rNufP5orfb8_usbSo>W*#c(hCKuM0!iqoeIK{6~LdX1mx;6
z5=YrreZS68CRt&)$`t%Iv{(`VgZLbv-at&mXa-;!P71j$L|T5KdI0h56=}LY!~hU)
z8rLW2_+yu-8Yoid3H*cepkQT+34is-DOo8yj_EWAh)PMnq5AXPa}(WFdT&=VCxa&N
z(JuYCbEBW-tC2eo^B1cxK{SLcs-chz;=de;wAySi|N2X{dsa^B+rM?15+06!@CF2a
zlPWM#45ys?o)-n=kM*%FQVRkQ+Xd*{(P^<K`F`JRhOo*s9hCq5u}q`3u~FX35^Hy1
zB=%{3xm!mC^n&vmX585N#H_kd-kUW&SatSPCX0)nSMm86c;<Mzdxs%bsL!=#ZK&L#
zR$6h9)7RSV9GS#^o<@K1)}+E?o$k2s8~h4$ZHH(U-T5b)-~QW9>X>1kRJYlU)QEfm
z;@a<X=SSH2j}id(8cR3X)1aZnR}{<YHmjYutr5;BwNFCZH-GoxPsYo5E*sIuay_5h
zXV3@DKUi1(_#U&mIv34b8SJ?Ayf((i!Ml7!*~yH8la}R>j@~n#MH2=^J%8~MRNVXq
zec&vD!n+aV?WNy$q1zi#?cOWssrnGmXl%jP2Fwu=5LQ{*-$vS(_6Y*e-p|+b^;W0y
z7OTNNh2R<oJ_-IwSfpWJM}<l$z=#D?>veFaX5(|dzCnO(S!ke&6=0^(09YRzxZk^;
z;B_S$io6a=Ua<Q0?zn#MlA~nvPut%<yl^ji*IYWz5CLic3xeBcX<-JWOH4c^<AA^0
zI*K)Yk;Gy5oFfyZQ47Ei)Gj|?M>it_1-_b_dn6EM0Z5{#LYWAr)PsQ3#_eHb+7z4?
zDo!8e+?P{V(&h9yi&RU$0(b~=)P3!HzaqE)H8?Gr3IbpPUN671sg?%_AQt0~Tw7d)
z&7<xk{Q@hJw&l4$B82pcLLyAh?XsKKac{N<>~X;sZZ8gb3S|g1ItgoH<P85dwzmk{
zYp9+EEQ#3CYyGQ?K`$4C9F30Oq<Ew69Oc=2TRN~eHm~d^-IsTFH;y2Vw<TFU!mm<K
z+F#t5G98hKqBKDNvH^DeOS3K76`|F-jd$^VRzodGTkBX1$zB13Y01H48@nT3CmiJe
zIY|8oNi3Nyyc2Y!L)8*%J^G5n>m^-HMKWo%UbXSo%xh}zx1(V5g`OjfY#K=>?cjCX
z0P5XcRTZz=4)&0)hCYmdaX5;2Yt_1a@0WG-9LFfuDmowsAW{&!Pf0+<_v&-dHH(@M
zKwiVw27iw9T07*`EAh}*rs3%4?!88@(};;#d3F}kAv6Hu6r`q1?FU?O%$YhsOzh3?
zbZ)V@7w?M2O$zoV>-oAMDB8&I;~N6{De(Bv3@8SNRSO`yNxCZ*y2Z2gCgznrF7+tJ
zmdj=Zx~Z9$+j1C6TIBMFY0BJ7G}9qmZ>eu4A?RIdZ#J=e-1{liA0e}sl|}$Yn4c5@
zF@P1Xjm_SQYJDH=dkqa#rk3?>IYVt1+H?c~IIejw0nTQgp0Ft>C}iHSh3f{%g1l7-
zU=I^O;*=FJqXI#;s~~;@EVuiJpTsFcFs6zp^$N=46RALT*dC#{EN930c62XrgEX5`
z!4W~bvFH)j_w|l&jb;r;rKh^@wwSx@sk`7%6`)Mt`j0w$MWemhXi?bDyKkGL-(i1Q
zd6ztXemxg8iiu@xtbXyL!@>`;P5b994$@D3a`F%<QGg%q*~cS7k|%844J&#Y!)o~r
z>i=!YRsIeyziuHR>Fsh2K)LXmOs_#)Z7md@Iav|nZ1C`J^vvP6`(sD`b1jR6E!27(
zH^>x`gRjSDv!-T1u1L6+__B@&=pgq3(YTzM=|urzHwl=Tqy-BQfF6;fz@ZRnRHxzy
zOJ5+)mbvfPPy^`YJ|V)hq{QuGAs0<0X38a#UVsbsi#nL&Qhv1j6%u9}rTrKFjsUCe
zyfD_YmJSSNiot`cC?1zrWizK<b<SjE;t&<Q4Y7(!L4#>XHX&=jN;kRC_=~UzQN(TE
z0U1)E3NGyRr&nv50z=d^z$h>V`A~$Z@Y<XZ%vD$Q!Hy9PqY&0d0278McIFbAIGvs?
zBd$B`mr-_{q&{8@2!!P|HfGi=^@_grP<5(jHIkg=iT@_%rW{92Gwo#8s_8F+7{u~l
z1GtGwHl<bf)xNirwC{$Yi;=AFtl3P>cUj<Xwz4nH$ayl;JQMCzn(kypf@c9t1X7_k
z^9TZAowcEFK!cmI`%MNYx9RDgUwq=0iRYB)lp2vac9xW@6U8BvgKSEx)C=1+7_n~~
z)BEm9Gp{^tIoQxVxH%h|X!VF&lijK_=a#=3n-(`L3o+<=lQC})1`%slDCKCx;^xsw
zC&n1|q}KY{?nEX<nK+rTF?owsjk$G2a6FaZnbgjo+bE9^2oMHz&tJ*T>$10KYqi(G
z&DcZ!ZA1}0>il!RpBlir-$4=ZA?C&H)yg1yh$n+cO14P3IcuMuesElk>gKMcCBn{L
zmbHGQXt2imupuoYl3>xXR4Gf||0@tz?B9rE#Yx8u`Vr#XsE!f2U+YZFSuYMtt%bSW
z3~IERnzkM(CIKiABO@a-M-P#QlL>eDO%VteF~XEi!o>|9R8b&+66xY93?P(a8GI|~
zn5f8{N+Sk)ji^3}S~rC5L!`!<Q1xNuhJpZ)aL)w5Frt|l06W$86M{!Vs@SYc{`}O|
z(h(ny{O|(<tr8!8Cl3&p(4m$8Q2tOQo>DwwRXo?KA9xl$FQY<06$oOh>jE32f+c!<
zs3vpF-5@OJP!fy}K2BvWdRA;KYFd$>m8bxnyC{mdyR8LbH@^wRrTp0#3*Fmj+iuAg
z3gl<FYQJDR{bfWY;u1$~$x~Ea`YGAW98;x(A425oca4-F*94&L{-%|fioQ7#&;sB?
zBLg0UG6dvL^Bp1qK&J)dAxbTi0=n<2lo=5YcvKCpOv1dTRJfv2F|IFWZFxG(i%YTL
zbptJ_(%~<7t`D1z>PDNlCi|{sl%^us_hju;MNu>4hueMXH*PgO2&D|0(jqWD*}wfA
z(3gGlC}njN=QIglrjq;i?!LVNZ$*NACzFhV3k_vU=B>~ZLYMIp%-~&^t$#bo^88L*
zW#diW^X=8~+cFB_`nN6*OA2Jn;|L+wG?Fv}aTUK<6<qZYm<(tG)Yn*-23_TIi|>9u
zp#>hiMshH?FRu+WYz;FZ2v+!smNhWar#OqwD*RR>4vyr=oKVP^nq;q&7`eQt0q(JU
zHD4`uZtZ4Xxq2Qi-BO!nqiP}o%Is2B`-Mm<{Pfmu40B)bkDbK|Zk3tF#(QJ1<+aCx
zn3ta?zH{-}0_huzZ%K=vs&rcq?>#*n7HR$Olx{i?YuWyIZ(ALjr28KI7}qUnxV&+7
z0+%({s3NU?8t5ifx4)@NYrR`89_;SI-&F)2R37QsbxBqJq7iXEyD-Wsy9si|*gAf*
zf0_U5(NgYWW$rRXeo`)Yj^9~BQW8e~`(`XhuhFgF3H>zPwWnwh!Fxj(HmJ*mH9;a>
zmg07;aD|rv8{8^3T=1{KJ(JP=ouL<Kq)w->@ufT`X5q)^CsBhJr1oyy+UF{gI4OuD
zXuLsA?2oQ(uGy73w~wd~g^cJSiIsfZ5n)ew34d$#2Oshv(|*2C68)DqHmG;LvdJ@!
z-J3PAu27-B)pmBSNso=?OWh8WHH&{F5k5Y_+D<)l!Z{fncU8V!fDAx<a9&mdRz>&N
z4A6ms(BRlLW?w_M$h~!8eL(4V;r$XiP-OPzKi?**`%E~LB}Q(lpcplizdwx1pT&ox
zbG)f(t%H26k?KQWaQTv&>L#XFu)i80SO<hB8D|P~`cMNnn5j>Z;287Y7Kz&ciWsP0
z-&Ik$e37QO33*izi&ac*{6~Ei-~$nV&kYi<14kZfz7p!csFQcxB8wqF5mDm+ZWA*A
z;!)#Y9a!AU`Q9VRSfOp!*h4~19W8p%LMi8XIi%~96W++pWT%uwrVLC`Wpr2>66XlB
zDs}=IOn*yc@74F=5=lG;CWSiszu)_tb*`6gUHIeiT{EGsb$?KOz^pZ5aZ9l!7dO5I
z*A^G)Q)|{fcP!|mF1z4|vpig7OZH6%wiBn*qO|vTdRLX$z$ZRu`4uAmIZemq6j<lC
z3~rGlMZtM$-DT2+Ni#}abxQ52%uM<0xlNaHq8}f^tuYN9%IssR{c2@*vua-F$La6>
z4pwRT&<ayHTCSiwQ&GVv_k`-Qz1<qonmx%2>+kM4CX(%L@lRzmxBEJV`D5Z#hVBlT
zWkP143l)W%c^g&>I9HJS_05<2n;;jH8b{GB2UAIWHYJ(%;Oq6C6ho!LDf9^-yGgU`
zGl=v~vhd4;e3dm`Mdcl>aY+cN79}4c(QvlwnMkQ%w^j|s!h9G1d~0KqRJ&_ANzg;!
zHf5QDJ==FYx|l7)YSHaXH)jvF7ibNp`VpN7D{dDQt2K)-<XnnbfdBW;ypW34MVIz$
zyv}}g%cCp2$7h{scw|U)_pE(%WBxjaFI}7aNyA;|Vz5nw%z*>?;#W85_lAyS<NbSM
zmZJ;1!MX!4Klj*eUIA+)cqE+g_$=r&Gd*=I65ej-G31m~-!R?0IyVO*f>Y}L;;_S_
z9QTJG`<!@ery5t=zrV0E?|xT>mGH=0tzLK=?GIyYZaZcRiZ~J#wGXu7vk9NinQk@j
z0o3Wil!a9EboD7XXCisEL^t??f<B&<JR4s+(=HFyb9IzD9>tS(meT!tA1JBCZQ@Z8
z+m~NG<6Wix@Oi=J&-J3583K#veB-x|LQ{{(`^z%Tz2{~1pi>d0Yn&YTq6lXJ4sz=0
zYxLNF-^bY(`=W5VwC5M2_VsMMpUK@pE)`d**CpGxx^c756qY*G5HmifWlN{-s}ppQ
z<AFNHpFX?_pPQzxx`cccU~h^lBgSFln)D@$)NNxVI0x_hxDntnRPEok-fWW%1mhuf
zgrKjgB(|z5SS2|vZ8-TVx_G6W+3x-zPgkTs-I`X6`Q9k!>MAFIC;aJCHhT!>A-Oq9
z`_p@#N_)me9wF<}h6e7$`$~_Au^}uZDn7Z_`a6QH22(jnENoSRA}pxNd>s*Q&<Mz3
zk8dg^fhe?F@i_kBM6|RaAG<u~!%L9T$YMgLYy~zk(CqOsbRu8D*dg|&x8~J61+~Ga
zn}(2Gt0S%gaxBP9Y!wk7%AeMJJ+{%<af6nOY^Sbahb8WsmeCnWsJO8L_@HXZw_2PD
z@cPS_9P<P@IWJx3T2OIOs=VSWNT4aN1S#_TXIj!7V_7J`yyFthVPx@!4TRj!0v-xW
z#dIvL)`-Qzyv)KK5Q)04qaY?e15g`i{-(U+%2iqEP&|<_F$-O(BnPLyQkYH|*#JTj
zjcP4VSyZb91;e6y?ChiNH|Hy@bZU(q|E}AsvT>t~M|`H>4t~BTkM8-1q<+Mtbhb%A
zZ#Id4eW%?6M;UBsTHoiblo??bwA*iRrTfTjz5j?Fnq4CrSNV-k;i;Xv%9;7=p%*do
zh06GF2TDw<#QDN*jptTkIL?KF1V?WhV5Wo!pQD`io9;xL4)I(DXXSMdE&<KF2*&PD
zLFdPxexKTVnWFnvd;8h!SXOu;#!0>59)@TOszJU`n(TL_J9%M7!@hBZ>hFV!)<uJJ
z&yGM^NpBKPXa-GM$a{!C`&r@|L^YR|-B_0V*2x|qyWiChC~KTt<tZr9Rp0P2QCKe5
zlIc~2`?UDP=QLX_cg)yfa^hWmUHN^^&i@nX<bK0yIc-Wv;LzyL9|k*Hm)3GlN6en5
zr^-)-8#B~)>n}(=F_z%dl``EWd|dM(=!{ndo$|iRfBxAjR^DC(l{qTxBqF~dS$>gZ
zJ-?Q`=a(=8j=k1%VG^lsb~`)Aai@z)*oAL9w+h#=J3J5hIj5Hls!xDT@?$xHm&F-J
zQa?F4KVSkrSPml+IiWph<BGp7GXW#5zaE;r`PtrryL?kqMY^e;&StcSbjbx^bBf<Q
zz<WD1x-3+boPs|sV*h(Q+1g>eyqeIDUB0=#Krw{jhHCtqauqh2<AHVAjP>qfc{9;;
zyzXqJv#fdSj+C=FDM*M8>AsEoyT~TDQ)^Ff7xVS1mX3|P+wE;^W$5=3lWFOL;QN4q
z-P$XPY>xXYGViTLZTswHy@@LWhCt1fWZBhK0&Q5g#W7JD$IF)rxdU=Uii+3E9iILf
z>D@chI#j>)>w_MVFPdwEe>mi3xTKS~<!Hqw{t-#Tc%1oW-;nGU0C5#`2~AC5VcCjW
z1nkYE+LSDwq@Dv4`Uc3%%m<c3;E)KVvXZ6+IT<BvKw49ncNzZw$I@ANMcKVye;B%@
zK~fRv?gka4B&EBh1*BsDMWnkVC6(@up*sZW7`hp{C*FC!zxNNAS@&A^b<TaRy+7AJ
z<-i4#S0)yg+;z4l3%c=LE6#GahM}bWwbUQYNNbffx9Tn2LuVg`zjoWx1;qGG3dS`*
z6d(eRv1J9)&HQ2;d1`7uV0=ud(8bCRe{BQ6Tvbl2X#ijxQq_dc;hg?DHHlD`i?fke
z451Qn@VI5QZ<MDW2DAdoHXiXmW6%Kj09E6W-y3BtH>h}9Mtgxa?SCou=-njeoV)GX
zF=wtHN@cyeTH&{nCmW}&okzm^!H;A|<p;Q+t3B>LI3ogNXZQ1-v(Ab=L}cEJhOL6~
z?JN8c@YkxP+oY@|CHM}h*OrLd=CZYQj3zXwwEO*(>-9_Wg#By{kRm>%+l##;HlGUH
zbG*V4@WfV4RsqZTkBBojv-_}p_Vvkvf$`2eM4!ohY&52x|3zSd>U#(-YG@~(1FCR(
zi(5!xdUe^_m(gOmtykaY{g+4H(c!d-W6H>;kvr`WMMzoJ#T#dT-!;cun8~C`o_^$T
zggJTBF0;%+#Xfe~|HnzC8K2+$rhn9$a{bSCGVij-U|69K4>vVTB|q3R)F%Cb8fU0~
zL1g2GL2P-ZvD*IkkAOU-d!XnWJ=5noFVOBP58|!^MDP9WuRpMQ{zaKhtYa?X7xO4z
z@)SHACs`$o(nJ##g1k<>7P2B3mJPnRtf&^CVkG`A3xGYl<D{51Gvch<9Jjpt@dlXz
zL2k|@9p9MXpd6cC4rbPJobmKU*^s9S8Z@ew(U+Dqz1+<!A*L9wOrD`tZOk}Yfa?if
z5N|}JOBum(|EZgVkU_QV<nr;V8Ug%!Hl2S<0SCCDBx&Jq9I7d;;{svU6YfJ&O9W&7
z`5Jien(I?W#IT9?oG?C%>F3+TR1O^`37%tW?KjX0EZYl#Rxi|K7m3bn9S6;`XCcN_
z*>Vi4Du!~u<Py<g8e`n)^*kuSCQ6kOLrWY$Y7)2P%MJuR{t$g>fsqf4PVQJ9cLNvg
zOI`V1jh8YWD{;|PckTIKh4Tg~T-O8Nlm9ToJHE?95Y;QTKkB^wGcElVz-*7$H}pV_
ze#&2@l212`-8xO+8GdQuaTgdc&49BmZ9FGP_(T*<X`3HkQ5+LIU73GuP>n%G`N0Pg
z6z@0tr+RiWw)UmPR9^EM_(xv)hFgWH_xgin&)DXE()+j!A)s%GZIaX<)9m`X<v*0?
zxUB2z<<*kAx^~T`KHQdD5weIst<W}Fm8v0AgGR?xaahgUllGsplULuYj$pgIAYmVf
zXEitF|IOSyPW&A2Jvu6K0x%aTp1H2*?Jg+?t&DV_x_0f;$qn_i`t8c(-BVxZ!CI6Q
zm+tr<GF2Focw_u}f~fAZnd6`A)zLL`$J5ix6<^%`N_|>$<9*E-)Xa?GYxh8(xG(M5
z|2Y4Dfl*Qa(Jve84*a)B(RgU}plY+?WQk(yn$Va#wPzT;4_@qj+s?CF<9ukp4#Vda
zXkNedK^Dg$C*q#;yN^=s))?B;&mSXHc?G*+b1W<iPMzS%l8P<e4HZsqrVzj;+X^X=
z-EL=jwAHjO?G+H9tzChh_Dm}0n|8N3U{J`<7H(Nd@P*#C4V`fq+d@FHTtFg;af5~X
znU#|Kk|0pt1Bh#qfAeBMbOeV<-ZH+t(t%6)Bcr><$5o?v-`cnS<A{K2L0;*La69J)
zA)H;&?k9B5o0|ol&djvAsA)QU3j@Bl`y(WC-1tPO{1tbx4h+1F0X#Id4VjRhY&m?C
zsot?Y7}Xy{J8mjvglcG%U-O`qBn+Xw7OCj~W39TNrH$jiI%G`zQyMqvxIMrs1EKT0
zPK@jCz1TBGSWU_to6K?Nv#>@};j7;mMJL+JTw!>g%*GiebaP_xU3L+}BVZ?ZJKD6L
zcW)+}J9|!=@78%#-^L${I`i!Z?Og%eZ1y<xsfXy`@2-494WO-`Q@sTEJf=gjGm_1&
zfkTeE?^lyz<Kyw^+7?#tOS3A&HOl3ZYUs^$`On72GagGsyCy44+rOTb-Cd=3KgK@v
z6{n}05Vy;9qo2zh8)}Fadf*;t$UZ&rj;HFIx$vSC!Indm-dh3TR;zz4sK3zUeKP84
ze0*`Ve60U(VUNh7Rt6c+;oEp?9!rOG+%KqQDXkYQ(V+dfY|B5hIR8<-^1p)9MHL)@
zD|=(xn%$ZrYJRo#>Hn+0A;O0pU8PvNRi(+=3s$c>MBC23w3wu==swZ2YfyMD7Y+Z6
z3_b%gk5#7ALUP66Z?~O}wNKvR3OU97rltO4tIbG5BcdU@>FasUOiYmKg(4st18Zb2
zezRt5#|1fRWogy~!^7NuN}jc~ANaGdvhgcTy1nYY=#JHJ#g5s)8dWvZ^r(0P5HbPM
zp-qD&Fh2J)J~89V<zK?vM47`uDpAN7?-?qV#5Y<;-jeMBAJG7BYP_OF+2VKcsr178
zgSa$WWC(*=OJ80*-IvlaFBoaC@0dt!s2rC$Ro7DiJT#5S?yTiPa87G1BB4a_gt!eO
zhWQ2Ws~R=&k^!6mApjY`#jb{cy(sj>KBcN3CQyz{)DZB#h-Tm#%#w<UA?FKFO~|sr
zkn8aZt4KDX&zEW0kT@VA7A?(&2^b!>AOISpS#D-9)2kIEJcmm`=wfOV`0nQR)-PXO
zo2NzQA~OCbMU?WE#(e7hdl^@FukoF6*`4&6Z}L*Wlo$R|jW2ixo$sr^l0k~WI+cf?
z?CaOBZ~CJD%y1%pi~bU$;~~M{*U6KC%xoEH%Y1;qCz}lu`gb=MN5I3A!~q%Nu=BSQ
z2_ZS1ceh(k_kR1L9x9w*3EY};a{(M9o<!^g9off`DFWAXSmYglseOT#4D}d0dqhOU
zM4$Ka-@1f3u`u2EV<%a21ca3iR3Uze8-HhBH#L&pb?N1agcFmS5=*||8R6FmPV!r|
zW;A_Cdx;)n;=Zff#AdC@C_r%G$%vr{^O}>)Y@8`JDx~NM7Vn4K#jQHLuRarees$Km
z&zKk#FBQ4Dy%%Y;S0!fV*s4F)7Q*jvx8`k1Jyxq$CL*HF+R#bqCtObe9#+%qk<ekj
zB+jWvo6Isa7j8(m20ftWx|?_%xyl;i?BcWBxBVABB;M)ukFXte{vyC*PAjtXbotoo
zU-e$v>1Mj3)cJoO??Xv#;`j-+WAAQnBal#GWt^Sx_Q#LT$u$?>g&h<j+D41Z!ctS;
zWX_y>6?h>n&U_WShVJm4WzC7idoOo)p{-9OzVu)r9!v<~-CB>}Tqz+4r5%a=tZbp$
zX{YdNs*f|ul^k+gyn;d-w+*Ve5f{#m{@Re!J`}bZLLvg(xtYjv){IGXs0`4F@|t-h
zr?)Q3jxm~Y`Vu=AGxGOi)Y-okgnvk)x}SN$&e7n3RiyA;41!5mrF+iARr*LLs#{cw
z6#A2Hsi3#AK^n1`CU3E`KnoR+o`~WPBPC_LN3A&n!RJ=o8yHLK7Dc&K-*2T6?(rlW
zo(VMC1d$(TMKi~$TSj5K)gaFgoB>6GhI=v95G(Ihho@l}nGV_O>PHXH++rNp`WmRx
zXx-Ng1|BpOixD(7fpTiPetNW5o1yLxE$<VmWvqM-utKmxCnIb=Y^}P=#{&)$Zmawr
z?COEkc+ds_51U3E6Y#E|_}UJ$mq3X|nuU|Biaz+fXRnyo|JTGZ@+$T$5Mtu9%{Y;L
zA385p)$>uNSfBwt7niLX&zUaaMJVZPKYSY0K-jF09#t$LUIkVIDmf>~c>=YP>dZ1J
zG=XA#ycTE{7{_GT39qmox$_11Dvot7$#|CgY25$de6a}S*6gjUNCuh``SqjiDZi(;
zY%=Z~4Z}${52|d{cHOkw>B`4;N`4a);fML|Sc^&MAjfZy#;$qI+g*3D(5v_cIoSlp
zf!6*YxvUcH?5S9>77akEvk&tkZiy#kQRgwgjzzxn;M<`|A$D%Iu)_&AS<|VMN1f0$
zY)m?5)ZgqK4l&x_E<5E+;BC0ii|?0-N9WHwl<|J@n<ldOcK3pPn)6rw=hPDT-O578
z$J$^zgT2uYs)#(9vwDvlVcYLTN@k57#o}8(hMre^2=iklXWZj7@0`Zs_1VpGk1?i`
zx?$H4(MY=?!i`R=pA^?PQvkFO@L?X@P&XcGVKaQn!>kD9={&)6!Xz6u`l9l5XVM%o
zxVNqHxRzI|7fm7?s~2>BPX$3d_5}~!yfr}BnfE1K4zto(miv9$R`i7W<wc)7>5bGc
zxvZ;#&$@GA9+EuTRA-LQvYkbU$Kqvp1R6uZ!8ZQxV2MAmIk)S=Z$yGEis(9b{~>l!
zJXgKOmhH~`sgdy`DI*Bu*REqRco>~n%Eo)uJvM%H20_Ey>?##jyw+a4D*BfH+e_{-
z(Ta+@_YJsj854^ssH*-IV*hyc+`TaE!z+>^N}&DI*-)qC4I35QOK4j(uvAg+cY49o
zADWw{=)Lb3%JuNYnzDE|H)@>sh9&`LHKu>4;{2Cq$6Kc_1_W`Q^(wX(#W}P#-Et{j
zpF)@{`se1Pc^GlOTrSRWvA;)SpD3%IJsg~HxjARRP|TnEZ1GJF{B+Wpo{|tR7;pk*
z_gQ)f%v!ZQZ9kJ?fi;s+vs;#MQ}xOwz$#F)Y?=k{+Z)}jg)2AqP#`*mQbN+pwNM5O
zEk<oMLZ`JC_yl0W5E3vU`4tai!y#Enco9&0LmMBA!SprqC*brImx!W1VOC|-Rj4G#
zLF<l{dxfUdGw*pxCkm)%wdf1JuWgv#QqY?BndzMe@ASWH-sdnP7ES!;=F-f$%xB2z
z=SRat#4NtIurmo6gj<{@{(b16#OxRC>0F?OMNWNNbx3?qNYTQiZ;p9P0xYw)p-<Xj
zv%qk;qJ6&8?5>QllK@b*!%PCG-^-Iu*4h{zu+LepIXe73#kA3<BSq5+`h%97rLrP-
zdbRxUp%?F4$YBVxeeeN6yXx{f8_d0`!GDf;ILd!25Yj$AS;vqcKM5OV9fuRU)SQmt
zQUA(mxqrN7l_ABA+1a<B_#sDfXDKZr;<)~pCTe(0A{P=5y3|ns4`)HH{p+`v8-tsc
zt@wMfRo<usES->F^eQhF5Zd1V{L6opCfcn99W}7wd5s<8ODVMOwH``lejj)>{8goL
z;q(6c=}Qm2=h4DKP^$KQb7Y%7uL0plyqJ$;^4%dRYYnPMoL2P}pS(_Plt<(66#T;4
z6IGAnYDTq=-`^}b%-c{JEs=I8<z+t(>WtqyNe#?50b`_$t5z7EGT!rE8&FXjK2xBV
z6#{)?O^DSkHr>xj>m58-->x+ZE^$8ha+#jK*C!M+c9G6WfIzx-1*$}^wguG{k40UC
zB`XHKcDH{1l*!GYUijH1VNpbu8>@I3LZDZH1EEopcK2|3#b|1TviSt}Fv%SpT&l)C
zp}#mYG@|`iL+;??_=;~UeWgVT52b$fchTllm|#51RV}gkptHEK2#s^}fUkFu5ics#
z*W^e3m?Ul42I&Etjr@;75)Tg>MmXba&ED^|XU|tw{A&Bd8On#7kN#nV3|~f$qY)3M
zaJIAQ3k$f7mvbRw+%s0~1GeHK=3^Mh&owdp$F=LF59p6~hpdIVM6D2y92Irs*_QG^
z*}($QbOZSV*QF8wdo~IwZxWkGfsT*&go~-ln7G0&3wUW-vubZSz|VW9<x%@$Eb<R9
zBq^;JlV8Q|wNx}N4q8h~d?Nuk^k`W5koU{L`~+hKCjbveF*oJrjW!yjxeUr5`pOuf
z)lx0cQWm!GmN>Xb{9%CCi0v#Gwb)egEA$r!a?`w&s)LCM{pL9V(nUpphzj3dozryW
zHHl;yXl_>$r94}HuWY5mQgKGjXeHl;#71)h>AAQ86h9c>WXc%kY}nyOldPspPytB2
zE3ff<Gt-Z2&`iF#XaRiZ6rLL-3J7L3m&t_HO)dHJFYH}Ww%Rmt@G47vhCKa&0N=}8
z|GLXu!?*a=`AarS#6_eu&%Gpb;UyYpvW%f<;C<)WG!eLQy^}cbpHaTe>k{$|da{oL
zJX)o>0rT(sLKA&*esyWfj#>4iG5=GLips%gDw=h;%43vm|McFUny=2`%I@lq=ae~)
zw%MV2t{$iX*bj`UFfoL^pvQ_~!VYWdMx^Mc8^qew8+xuk<cZSSKFl3`$q37pJrkd*
zgR87f02C9Z$`Lmdc8y^bSFn~Z+aEvz`et6SerrzVyFA=<oS2FRCyEb?ZZ}soU>NDc
zg_t3|%+=u#@AXI%Acn80^L7r&!RLY!l`s{bUx6R3V}!n2rK@CPH0k>J4BfPp_O|PH
z#x#ikuK7xdhqy6*SWeT3`!Dwl^q^EwZ#1L!toARr7ZW%pt*wt9@T28c9MG2|`bd4}
zDwg?WtA%vPr_lCa@q*0%=SbT56_*zndwxn&y#)+MON3eEdT0Bfwq_zf=V&hhy=y1*
zFCcKCT4%o7Y$yqNb!pdT3MLCN-cfl~*C1?R3gkc}bxuN$Y8;3-UxbUP_&I5XCYeH=
zK3mSl<!nPO?Wy>Y8!n#yqt~BAhwrZ|M(nQe(f-wIpHC^fYsh(BZs`r}2xr`k8zb!P
zq@yuQ(Y<n$#io`mj@o!_kqP8T04aMYArG3JVa;16?9k~%%*#U^#L`I2p$@>ef~tjx
zFU**m7Y(rCH5%=;b)bO8+)v$VGKD?Q5h4x{3L#)`WM^f<p8nc(n0zzR0FA9=(4r$T
zj66^ZxxzhdQ^k3gNlG0$wAc~&yUc&AJXRbi{rC2+nD7TrFq-_mi##cp?H6tqLneI_
zceH|p8V)qM7ZbQ@XrD)LO#sx1VBIQxxrC7CXp})5$;>>smtgI89@C6j#OX#YGaP7&
zI9&(?aaK%)GnaHD5*f4;<PqA4>HwfZjI#TJK!?Z5aO|68K-A>p0rr2Dnf)RlI!0ju
zj=zzf{iQ&BBP3#lnw$|HuaSbO_<8KQOHju7A38AEZh|2I%Cq32{6zdLG=tTWAI+!}
zTuDj&A{Si_584=ZNDG0bJsd+^3C(!Ko2QB%Bk$0hh$})pk|IUDhYKQutC=p;H)oC%
zSP`m;1JR7sBRjhSynG}d<=S51QwqBUjQvrlwQ+s8G9SMKom(4nI@_=Rc9+ap#*L9G
z^6@_{(~=Zn`e4%iu-1*~S9wdfH0#3yC9sy^FMwGq(0y=&Dv>{$TbB(ojHY{n_C9IL
zrllJ272B-sYD<H0t`zdDo|9-`XZ8@EIWq6=-B--P`yE>ucKjf=`#JRN0VhCag{|KY
zMadL5H#`<jlqzKS>XV}^7{~cU4bu)*_0~SGMeEm#-?cwBYEn`E0*#GSuELA2W4`#g
z#9?R_ZK^nl86LT|hzomT&9(RxCZ|TD)SB-$?IuERfJD#SQ87;_isMaPY{BLiD)rI@
zwziwUhdZ_9U1?sb`W*k!G^lO()5dJP>^zp@q>?+L9fItEDk?60nU=@?g25N|20xq}
z`i0@PCdX!%&yGTwtxVH#)A0a-%^(DNoaM$EcrnV7e!Jpk?z#9Y-lO!_&~Y`=I?;oo
zZHVzsh<2N)0^C8q-~BkpVt19sZI|lpDSL`KNd`b0C$Sb@BeTiF&cBRtgM^U*52<@S
zH^fFvRz_(glBYFay?^Fe+?tib!t>XM{p+`sRx4oLohRD$T9Qc9lSBtJrnhx`%45qr
zEu}I`CoHz28vtczBh9Z6T2VIv$kSiYi2g%17qRx<FssZwo&U5vRxxFJpjMQd@2A4n
z&B3e*;V6=@mEV;GxWhp3CODjdTMZl2<Q*C_Rq%1Z9v-GSs~HYY-H%%N=g(CQ!`W&z
z0H%tzjt@nJc@_*^XEJuIt$|_KJ5jxE$2&!>#ywHh9Y;o<vlAfxE9rYfUW0I)9i~kq
zjMf%0<`5bV=Q9=z0biy_t7;skLkeDl;%iIYmo<H_rKu#34gH{3dA7CqR%uM4Q!o3o
z<ZA35j?LYKkylv||EdNQyd;74d2jbpS=N@H6c{sXwG$&KV^il*7_O^az3g4Wr8r2Q
zzCTEdwCO^eo*4&Ue>Q59_74uQ{pC3c3S=Xv_M2T<=UlRNi6N2YdqWUCB(wLq)o3KM
zep04cDU{ZZ-vbha6v{(x=@ZEHCR5}GbO+rwrCAK;^b7hu(fj;FKCLm1e<vhfa)P;k
zr9zv@yDnm;sv3f~XhZ)&Tef-d&!v%mQ)Zfa5vi?*8lqWQZiuViO2<%=lxYAnG{}&t
zh!K+R)J*Hz@)zW=DQbaJg+_xue_Gpa*y>nJlcJobQ*jme;lqc_h-<TF_6Poddn8!7
z2!GR^%@o=wD{Od@Un9N=JT-MgplD9vw8T~BZisyla%xZrxDln=FVS{f)vj8%@g~~h
z7$CZE&{=r)JKDZ|2|R2TaO#lPO<M9L^zOb?7T&6C&0oAv+iVv9(WJnXp{Wp1kwX8q
zSzYajrnS!`Xu?ZxP+sDCN-)noj0U&mqvr#)Iq0DrvgmHHVCqvVe~gPDlp7SxQP;?7
zhjy*N&fGxu41u3JIrYB0f~ePdi=jl28-kNx7yiXTb|d|(p~f8n&`$lj&UM%}<K2=s
z1<5q7^}^9y`NXdXsZpuDzY*M0vOJ6y+Yq<z!ry0+yzk5tSD$t-XI+=we;ix{Zi!4u
zYmA#3$X#%mR3b`DrAmAHSJh5(RAzPS_zGk*%C|%)`6CdAd&XR{{WL3X-9h8YxxPES
zt9#Y)B2$w%A9;!lG1=jVQ-~FrJgCu~ff+D|Acm`$erT8Fboao<Pk;Hl(Iw{4FQZRf
zcLM*7DY8W=A#PSY(o?&Gc6aokU3q$UcTMbQ<PMn(t)_ob!K5iDK1*GXXYitv`v&6S
z#yQ25rL>y(KV`0S;n>ZbHQw~5xk~jH>3_=HCtSWa=<gL9>dkvnVfk-RnjD5ja4vcE
z$&5Fi8MzVcZRWS6kvs%!zuU-`T@HXduqC*2`-%q+qv{q&#|MALb;3!BGnjo>9Nl$8
zU+*_d>3PX%?r2MuO1!&<2|?tx9;V4=!S7bbvM<X})#ZL2m%)v<1q;XTvJ{cJ&nL7a
zGOe{PvAuCWI#w6Ke$UqT8bc)TBwK@X`Z3>cg466J=B~OH7mc@=Ijr*xU#V@5V5|+p
zq8}d$9Z8~Q#m4`L>sjozsQ?kyHvx5GD%z1_V*vU*r&4P+@K$p3?pua_{^ul|{^niV
zWVAVg1|Tf~#`A8g7Tjtka^AE7?=mB~!-R3t*bG7qm=s|*BzYKAx<JuOARB4jk4;0S
z_4f=GY72KO0UB3TJ{{r)wtpXN)6H$W9vj_8x%GR;omknK20H*rW=^9yqKq(bz}v6Y
zc^5K<cEcR6Sj(A6(YDdtF13wTtbb94b=lh{hIw3dcp{ili@lyp@nb8Eo<;peL}S3-
z*U`Ux78ujpL$ZiWzngM%2fa9o_CFjo*2GfDRL2jU_4z-1zAi94oj901MWsL0D<Tj*
z@xfN8v2ud-r(7Q{wyZuplS|od;cj?0gm<-!k{a!^#0T&XNNhaw?L>5!i#(wZH10gS
zoU>v*n_x}ahNfdsI%Nm)0KJ>&)J{l-sq~;U>98{LzbW0@lkMb82;fyiDT8%f@Qjv8
zlp5p-K|H~TSr*Ph{3(O8?$$wsPe7n;1~@vEN>Dhnz<XcPx}O^`QJ)7Vz(;%=NvVq&
zK-p3(luQ@0Lmky_pyMoV0}{QVhP^`1DR{+F@RRT!3$Kh8oKeQr^`GzU(p1pMIQd}U
z`{?}>j8Ft%n~1>kF6@0ssVpwIZwo8k6>)u;-nn?f`?>Pxb+fTs{e>Ib_BwB{;A@#D
z!zsRa-+Kg$C)FK1Qk*E=Asse_EIX41EU$q^`QICZ$8*vp5$Eq^DK_f<cI?fuM#wx>
zNy#zDkpZSXbKz<}tp&{aTZa4D_6sBw6b3Vy7*Wwkl{Jb%VG8q!tDp4}f5U!}aVb8Z
z_=>LRVf0qbVSD7(n%RA=-WmEZ^vibkH<oA-CCvv)s~E7#XZXCutwfqh_UaGM<b`Yr
zp_xT+Uo7Q!Sq2T^23lzuT^*m+j>gN|jjvhb%Z=sCBU27<4e~0gXfrj7NVEiAXyElz
z{z<)4y!^;n>6lyG;#2y+W5NPQvuo>QatQjn)kI(Ix6Ahu3%{J^*>Ia2rG7Dph=}$&
zeVdj4Hx|&jTWIs#C&<x%hJd51zp3qYRPu+?)pEPsvhaP>0fHM{vG@6!tHrmRp6bLT
zpUMg@CAPV$b_AB^&D_HCkdVI=g?Q9R11IGz{ZaU%D3W<BUDbtVmG-o2#nmn^=(aBE
z14t%I9)zOu<@);?0&+LzHKdMk>Pw#4EbQ+4W`>ZPaGe(1V}~eD$7*+xJZ(msBTM|?
z<z|)?^Zm;f3$8$0f2Fp&uBY)<^ZUA`I`{Ie_#$T}Zz~_V`(gNmS=rSEUP!xL`)QQQ
z%Pd|Mtp_#3wr3c0yX&tFe-#y}nXuqzS+ox|y9(Rq++hmx3%jD%8e4_Ojx-=^Bj}vX
zo|6W`3(nxvCINjnBKJ=%Rw-h4`XS7Ituy}fv+=@HVu(KqD&6et<2>zKj-Gd&cbVu6
zwbg#Bn-Rxd9PF=rAO7ukRnV^>8}w^V4yZa~<cMgkF^)+P%6^&pu(Y1kLJ`e&-8#om
zD8<XcAFKK!^o9%;Z{_p!stj#!SFCfEoX))VaHH@O{=qHUPU<*|JG!R#@)qLu*S`VM
zLW>T3qlSsIFDA7u-gyK;G8vP5^{ZOm)r0sf?f%Jz2mwU;wR(VdX7ju}ig@lilR7bp
ze7n#;Zo$o?lN6Zd#FsQu#OR7O!H4?rfYZiOOrpg2l}HAzDM3GkefOu-4A=85>43|>
z8bVJe?ahr9IBA=xm^Rbh=5_BZOj1qPde5K;`X<o`MsFn&B{6Af<^ky3ywMICS`d#e
zehfgf>*A)G+rK0~ycfPY6*!5N<I}LlPY@%=V32Nt9HO158B8mc+Wl1Dce~R5^7cn+
z@4gI@#W0E7I8N%Ub*e6n)yB%?NT(f+i;qX%hv-D@_d(UI_bZS&>^&m81ygU4r?c)n
z-5P9&!U3wkMfk6t=>eFnfeR?;Cg;>V_m2heq<4AyuXe$9nq#!hP>Z5k%7(#@Bt|~M
zXpZ!RzCIQ*tKs|~#Ga|QIcIRNOma`KhtTQ;Y$~{Cu-{|9g#$zFg+-^VCA-(xM^Ip3
z`3rS|kapQMuRG~Il&{GF{G}e)b5*Ip#7rr5IsT+$0D<s0MV$p9--~1gbSKPlnjuJ6
z%s_rzx~?6Ec282WX4@a1C3GvFt)Tjh!k5wf>!KC;H3c>dcmXB9(aVV{8WF)yr%fMr
zWH(Ue&hh-|uXeMoPQb^3Wm47Dn`6;oi9HL?lUj{N8Z4o-@3A>C#|>-64ro;d6|Em|
zq5gN&uR}WwyKS}gK7aK%esx@o6=WkKpy^D#aM~dvm?!vkp589VM9iYdjH!&CG3%yb
zHTLbzv;oxiKk%T*D>~8Lup-`Jy%|3wA)<(~^?Smh*%1!z&crv7kV?&%0mTw^EVSoy
z>ZJ$Qv(f*vg@e1&mrV}RL5pYEt1noeZxD&c+-kS)0Z4qlT$vWmSEznqFb(ERmO~v3
zQXBKL%FS@^_kZ|1JdC-`f`@gATLMvqWt4!e!jPgDY)hb0rr)(yti8d_0d;NN<<)zk
zEH7$VDAv164Al%{$BDT=+rHdMS?C%CjUc{be=>VLaOUMf@%82T#(M^v+|T{nD&+R$
zV0u!yVcmsw3!cO`wdSF{#ET%1XRl5?yVlEs-;c`*R9T!=E$j2CFBcyhOt{(OuqCs6
zAhuMVKgkE~i|<&$1r*H8du7K}q7dZB6a~7q;$FPx-eRfLh}8b!Q0lqY8lmXO&s|9m
z`tVWWJF$M6+V!HBrVip3HtK$nFJ3l`reRWxTR?&)Xf?!Jm1w4Tpi@DIK+oeGAA_-Y
zL}>8UXnM!A)NRdM7xBCO$p9$-em1J!2aC8fLI#wqG~HDJY#*@&(7p|i+-Oy69x$|)
z)X6V5I^lB0GN1ub31ne?Xv`_BfXC@0ioM8q+1`%x)?&$g-p*^ht44gDCQ4?5pMX(<
zDmuM$Zw-D1te9tka5zwrTt;yt0z3a14ug-GKN}ZPinfW+(M`R*6HRFnagh0ok{Q5U
zx5Qt_tlfg2@#j@I8gmt$>#xO|oaDOaa?cU~3hsRH16EnVh0$=$&*#gYQ(h>PLq~pB
zgeQCQ;h#N$oSuQijuv(n6>QA&l$(5$$`nT5H~C<JIxYCk3u+`hl?`ExWT^7Cufn#{
zyV+YO!*$HAs#$6+p=n;|lyv8cZ+^j+_47;$J|fP{5z8jtOB5m+=fvBD{Fyk4ry2-q
zX2dZ962i`Qj0$A^gF-22SvNWFrKZgr{3d={@llR3SNMLu`=(lty)&7B@xx;#5~&UR
z!c=%=yA=9VXMuXQ_<!|R9UHcS;^MZZeJ;KSFRnxfk__wDo*%EX<6vYzWllvCR|i8k
zzw>-5{yNrB9So-hDh#?vw|v?bN3%I+l{-VGQCRZ*y>pX3jg4aY$Q7Yt6+=)>Ey(MO
z@SG_8rw-c6e5D>&uZ{N6ztZ<Pz3JwZT`&UDE!1L+|Jrv8E^{=@tdeVenakRGFRxsM
zhShmwZf<!cSOxsdZ+0%pwFxBvzs@j;Ni?!f$2fE`-jAkajn)3A&nb(PJR8O8iKvm{
zaVw9H-73z!biF02@V8;0TR8qTH(FPT@q~(9pkzk>sY-rRvvre|4Dq0mly2p>_k)mD
z1wjp>O-z?IGIpwb;!l%8Os0>F9a?(bQdL?BIA+qUyCWuu-Om&?U#E`NQ9NwjyDGpZ
zDwv%o9sWHQW(JCW6`8Vl1Mw4Yov;<eg|B-R&k-o|^8U<`^5*7M)}Q~$r`c7P6y4f8
zI55y&+)d4ywEWD7aM)x-tbGE=%+|dA<>=V9v9>R)fr>&;J#?G9cL^=JH}RVX_CN_l
zIE@+XY8W<*r9;OTcAV_faVm5yz-B;%dELD8dW_v@Qyeew*GXOj^If!gx03t<1+7A0
zL4^OU`%q>no#~O?z6Tl)WS+=r)t{vP+fGU-=T0Be)XmTHcY9F)h5Z(cH)u(%1Wln%
z#0+Nu!sL{S4#YDA|I`S&_1s$r$yktT*X-imXzu^QX0$PLpNV$vM94xAoAKjwK(iXJ
zOp0d-TAwfcTI-DXC2ZCoA9cp7mf@!w`ldc#A1?K~9k^AsfxIFvKZTBOmnUFZCI|Nv
zXV)uKU5|vGr(U3crKt<mA&epPMO?1+UtSqJW|JdYR!7w)a9$-A;|1dXM#FK(^#{nb
z)V@6BSg^9sM#PbF=8RG7qml|oxgIzQM*-jjkEBg(b>2p5G-`2pj3C9h8BhDi68)ZJ
z?%%J>_9)!)HKx?wfb|viUK!MavE>)1B=@a<frpJDPiWi{Gj{LhrEgI0K)(BWx0<T8
z-fv4wgKvn!8?^GB0#Wl8w&fPo{ElSO^L8uQ?2}L0)74dW9<r5JanSR;+Z=_do&vL9
zEhKzo=X=xjbq5cBfUDcIz|HS2`1c)pc}mSST`Q-5%vSw98514~xH35FxZy5M>(^*G
zZCaUF;3e4aB(=auEe-f?Y6|@kmU)Yz;~A3J5R@A>MGh-2Egk>Nu^O(nh&`puI+~iX
zWrRV-liPJ(T+J2o5{e)J{>XFhb%&y?Cv#b3c<a_^Io0GEG1rEAz!h!OAchYy{`czO
z(>G2m!x<Mx`kga_2cPdMZ)P9N4L%nutd=(a(0fMk1qv|{kLy40X-t)!I6&X%m>MSi
z1ITX)xB)Q7i*3l^D`E^uh5VF+w#@2U#&?u=)FZxSHBSw5VELr?^gl%@%%Wgw(hgg0
z*4q4QR;zib-HgK1;<0J~#-rCOwl^gn;2$gvvB;$g_fNVp(c^p<`S+Gw#|sh`it|y-
zZ#U1Z4VE^9*AJH-#GT;cq~$f*wj%sXsT#2cYU$O*jpB865(B_Sv*^Tm4^xnf<+gu<
zBPaPyP6ur>DGjD=S}r-Cz(H#);bdb^N6%GD;~PFV*1VySSpHYCIU;?`(NNQJIJm0e
zH0Do?GrF!pnnKWzg|du1(J0B^$O7%+v&EDh5ogm!MJ<j_XzA&y=?hSjOrqN2>IUyN
z+e@+}S=lf7i45t*zeSX99N<zC62CLwyDz>v%cq0yYt?0IR*X?BH>9TSg4C;GN=x?D
z9EEIG=yeALnqm+!@oswA;uV!DI;@;j+v^WAX<w?jbaB<D)Rf)$xpHQT4ij(0to>A~
z8a|_C%^A(gDss(<cuPRsol0sh+LyiB3rXne|6&$UWq*{in)#{Q=Z&uP4+}>>_mOO7
z_Qv7LyV0pY{VZDXT;ByLR50m1jk>I$_P@R44FoYSIx^S4HEC!P7pQ<&cU|$!`55x=
zsbxraqFg<=RW}=c5ie>ukcG|(BSsJ2x}aiIa|h$JM?qf8eD2)VU=!+kT};WN1asg$
zZL$}2<x51aP6dywUN*1TJZxf0u};O6`|zCwEjZxkM{xQ)nfhd2?WVmdB=!Xm>E!_c
z(=>?+eTQxV(W`(;_a&-0NGk6GM0H76QkawThYX=gE(K=rtvT|ic31H0XNOeju=$ys
zWX_776cc3!z8DO}ota%?ZRah#XpkpR!S1#F_=^aJPwV$bAZ|zOj^K-9D!h~XWSW#l
zRrTvgsyZv@;f|*UQT%I;^*=#hO$Gy@V^jvk;liet_T2zl;=qGBE0g^K2X1AIlp3b9
z;#hc;bU^@&^0<kpY8NAFk@^)!y<TX&tnXd`rV53zx=FAwcvvBqmWSOh0HB0g&v|G4
zbPpSurBv*eBtqcf08BfZIni7oow|SjbG>(+uWPLqFC|Jb-*tNjPetleM@RF|I0?Ki
zR(M~Wtdue^JcT%>nh&A~Y<mYjY|iy|Ok^Rx(-bRo&YA9&Izam6b>3;iveU&$@rNur
z62fU8H@PGad;o7l(k-^W`Wp05;Fff0^Q^-neJIQM;`(t6nMHkDJUwvR42Vqpc7^>*
zs>}he)`F0UEp$WDr+?<#EHv9q9{-i1R0ZPkiKv@xQrE%%zAfXoX!?HtQE_=PclA8R
ze``rCihj#&>}YmIATK-iF{(tyQuT~_xB}jMHMGO%XTcxMW)ieaHl&_@C(adC<D2gW
z!;Qv%6g!ZBPD)bvC|LXQ$-)UT<h!wFeJDT!XJN;SrP2df7!g<@Xdw}@pQYztwtTx{
z1vC(Ywzy`?2Y$+0!Y@QZt%<&P-}SB+qEH(Bj>}Z6a%xa-Soj^WDi^5tG0}&2n9KuF
zUSZcudbL|Fe0EYdHmT6;m;R&A^VkK{yioh-*LQ4Y{e+319gPw5eIhmZr7AYH`;>*P
z3J65I^04C86QWJHI~(&@`ADq3MDIux1<6U~WvihK-8*PX7S%^*{7Wvm8E3wp1J)?_
z#i3m=*1gmIQ|dwEu*S(m8qQ}cF_D_4quHPW{ZG~lt_+OB*JPM&^J^_K0^B|>=QiiR
z1L?Ycm3r3o{Cs4pkWp;83_;tNmFw^P*}nT_rQrq*<1BzbwW-xF&{sQ?wB`t#5MFl#
zWDa)t0&0E>B%<ClRKp|qprPI{XH{f)W?jBN?tN{0>ZONtnKQK3zkx0Al+UNf1h8(&
zHQISDS9qBlhbE#E`+xJhh~(u;q=!~y7ZmT<edVcHNe6exr@sUt#>^*pL0*xBua3(+
z@QU*5Cf0n0P|>#iOkz(u#i4e%R!Ed+R{Rb(R`zyQ_MAU?WUM$X<yPHGP>k=B%(dGY
z2ktNe{CKmsohPwS$8#c|M&r4Rw&wIhm=rT+8WA6SIg3@E=$CQ}qyQ+^G|7#?e%*Di
z5G*SEPX58){AnLfg<+;SKT>P1vh&r}l+2F*GKUeq2t!<!7k|1n;)~V&#>HX=>)p>N
zj1TO6X$gOcHDtISG`Q*s`KsOT7t&wB#hT=ZJXQ}b_UlJHQSRFB-(e%SlPPgBb2V!e
zPw9HY4f43&EA>-{PF8>^fYsJ(iO*?LUMl0?u;<{5Jx^JwHB9ZXT{tP)3jPg0Au=sD
zAY}|hwyT6Da#r!1cW%5qR*7C@5*cn^4SV5q&~cXI#vaKJ%q#7_Q#)O!46@;hPHb(>
zCR|*fE$^bBpXu;^OLB}M<6iHzzO+fn>=yh)lN2;aAPk@uH?FF}siRXn^IKR^T`#Wm
zh*UsaOwg)^B!1iRKB6_#-?IC(J@^5ON<!Y2qxOU5w|ZlcmDNlaEsf#SH<pwZOR$oG
z8B*BF0ESZ7McEc2k9CxRFlt)bIZs(8mPfm{a&GC1of;=eq|F9NV?(m;abtR4D#)oV
z1*|w2rsHJ0kGtoA%(MPJy*|!0Pppm+{);x0m@&^j!tX#W6-tWGome;D$|b|vpRh<V
zt~{xy)g&=;EievGiSN2^iOI4m7>(Z;K$#H%o1%eVwdt4@FKP?`&|JGFC6QMs-j+n)
z;w<pVSar}GAOw~Ml~GWSOqNX0*k@bjy_J=&$OUkR;hqDXM)J%1Wlj8B_qre$^}N|`
z&Hjzz3NEZ^k+Pd@1e|rs^tDY(%B~v#h3@Or>XjC;9YW8GjZbvQ>+Nx^cu=;ashEVm
z5k8ewaD!dM_>2k6yQ;tpTxg<Wg}Z%vQi-7!>4zh+qbDI~O=_h2g<DZYvgjsiNt<EW
zkHr~^zi;ne@rz)Jlygf(wCUUI39MXewDA9K_1Zm!{f^&q@H3st!e)Ff>a<W%QD&H2
z*G8)><SiqYm^Rlu^Vy?@!>nwoL#$d?TcoftkQvDSevabK^-G#ZZ40J*Al3d!F1Iai
zO=Ok%m9ZXV(p29_wUyO)EWke2ba&gTagSEGjksW5?=OqQ-Q-&`hkCmK%}CLoitej=
z7DekD$632kpXM}OqUlnu??I3c{4!yY8&#rfU+TSMAr?^3MkpzZ3tzXS+bg~>Z76Dn
zVldH6%m|<CRE~j&tyDz$&y$;dU_Xbd);fM<jqPK@_Q}^Xzjj~|6|b5oYDx_OPqxpB
z<Q6a{mkZhIes9s=5fc8Z*3xruMU`7|M*C&(&h~yrgotz9wRAK|>#+8-t6=s*B>XF~
z1Ux)&Q?Wn<i_d{OB+nAf(F_ov#W2zPpEFO4@baK|y2`KtCC1U<tl6h4kpZSfr+Dc=
z{g_w<_dQUqO9?qE*l+|O@`|DYwv((b5J9iq@%5+gt1}sRpPkYBtos@A{Ap&)CMU^-
zCyco9kti&_q1LJ1R9%K>(ooWtDj>lj8Cd?=^|!ckoxZ)Vj-Y9)9lU{vhf#g0W@Y$W
z^xQEX=>YhjqD=a{XBn4nFHq^to{xu&4Wp(NzK^zO{H)KuVQ41N8C#nAPK?w!XkJc^
z7GRB0;m3SyF-Dnn8tI&Zt<pLtDJrrReO!tFe3{$4JT5JXzJ;vfPlP}~h*utLr*u5+
znZ-AO_cEI}p1l_u+xu_c>;Vd!;qcpVC)nawUwVIYW&JU{eOB40Ryh4FD;_E$0qy>k
z2%ZsE1);Y|`@B(_*mT@u?Lo^<62!>ID0|^&K?n(;FkY3%vH?YG&}sBjY^{0x(Fq};
z<kpJnsUglqaa{V&mTB@0>Ar&fITx&L#}%G3F5J?vg+kOX2`e~yAA8N$;e%T{kahU&
z%LS*^K8dY2h9($f@&gc$<v%R0xX6d8r^*#ElbnPaf^rIKW!$2mvjB+B*6S0MR@Ht1
zmW)Bg_m#%A`Hh4gi23Gj;{?pF)(*@6nOmyDu1}`kiTL7|CTA!vj*<Y?hVk$b;{{f~
z5?;6J@2uZ+`TebQH(`kTn<N!=;tF|YKe6uNWO4nIBBRnH*goFx-~GZ8Y+=%LvBgZX
zao4bpl18g!mP^rSP4a)=#qx*KdHr~dpMs!f?Zmykb^oI2-*u?oxX^ZH*n8+=jcrLN
zPn~7TgT?RXlta7Y|7FhjSjRM!i>9e}#xk@&U%7O7%5kmWD*U4`L6V!L*8A#T3A(`y
zSgE0fqy36)p!J>#vqIqoMVe+BT8v2{U<Q;#`<e<Y!o|Dr95-gw^GtS}YUhX}LvwtS
z6I48f`OFWRTyQFjS!6@57Wl5J;d~`R0HogY_dobeW!1_hrM(1o5J?sxmd`51bbU64
z?nE}){lr~5f~o2LY@?ElY!4lYz-@()e#%FB0JXGRPPL_5+8^_+oNH#6^KcdsXL(Yp
zl7|ZB52SUIN;uSF=V0`=p^0asR;V@<7wws#fh(Zx)56Vro7lRX3QSn!#<;$@PtFM5
zDp;B-L_zU}fQ@oEIN;2Mp=mY_aYDJ_zaKD_a`WAaX&eVY8Uo}?d8WkA8cnl|F)UwT
zmTAc5J(8bGJU~jqmh=H#{S$ar9DuojBNr(Ysn=!CVWz|f1>5SIxbWQ1-;vWvYq7a;
zu_aLig%sa)UJ2E&RJAWktTCE6<Xm(SujCvV{1~_0tyAS`@VA90Q_0QD=3TD_v16%B
zA??qk05*Thu0i(xZ2|r>(H(a<Jce?5Y9aYn>c`Ndw!I%D3&`i-n6{|4s3<nlcbNQJ
zKl`fM28F|bDx?&ciQ{8u6KWa<&_mQ^rtNb+C2egI`cwbLufs)kO})1N7~ic+pb+>(
zZK2{@;B?^5xKH>asz+D8oy=&MX+)2)x#3!hWfYot+i8Arnf@^(=44&=*cpwvos)*S
zh=ze7Df5@UZiZ;vZQJpLe)sK%g7FAY8YG5nM;6^1qa+Dd61&Hos$Pc1Y<U78?+~j%
zl(I_Ksx1-z=p}m0gF&X8GkXMJ-F|x+^G2NKiOgJ;GW;eM8qz1fSkp3(nN3r`IBGw`
zrUG@OE&8i}+%>=O?dlHPgUN;__MFQ?QrU$&^S+@r=;mRs!v~9(7G2m`hv+f^8z$9Q
zF&j_JpVSlXG3+l#pBm5adgIM)pycHfjHqV;e9n7tZt121`TG!A9}-pv#6F3G`sHEf
zu5Av8X34|3fH6U|>DvftVBOSwyAWqGzVU$XiVu?e14@-Ao}PFb28}wI?M|)dLcCW_
z_iiPjzgyHnLDT4rziUTnk4-Q#w9|jAiN1cV$1SLO|7<f4oy^ofu0u4$a8AwRVY}^0
z9l*TObSjc1nm%P-)iLMYuP0bSyQ1@NY6NG&<ipBs?C_l1-nV;1@T$_I`e5YXUdKG8
z0Az;Ex`;BlEZODs-Fjfr^{4Ey3buSsBW2yE1pKruQPT}_g?@)GaLRGFR%8`xv;SS+
zG9!lVG2Gi)cTtR=1;U*98Wx9fynRVz=637DOpx(=v`oKQZ&2in=6d9^$9abfZ@$R$
z(TwwG#l<%CqA~AXrB7ewW@(DT=ZT@_F#FL~Z$wU;art@<NBvTuUwRQy4Ls5v?U+>C
zb8$k!pgm*m^>}EklbQ<T@gO<rC?MmdZQu|w*TrA+jjvvO{XSXtc&<GujJB_e(%B9q
z#TFy_28-=hbU3@35zLCMb$=h?f2TIhOLwzZc5pr0u(pwtQH*k_HgrlgdRkYp;#+Wa
zxb|72W;kKhN*5xWRXS6fyhA($X>Fp9J*Lgzk1P&8@#{zs&kr0boEIuFxI#V7Ql3uZ
zcsfPXZ!>WXzid{rU0!mkoh&+u@|70d9nS*fb1{&s$M$$lPee=OPO!r-2;nH`L8YT-
z|E+1E-RyUgR8tu}VWUi4!Wh60i+IL?Pn~=`nj1sfNoGMejk*-YT0rJzCFPqxsxrn}
z+#w{+yM+f=*yN#S{$~*tNsNLDKP6qj|E7$0l%x-MF)VgEfn0+FTQ#q9F+5t-Qz#cN
zh)$!3N|x9K-`m6Xu|E(gRJ^tWV15Rs%<+9#zkbxu>LZq3H0O3QI>@mPo<3PBHU*$x
zqQ49Y?b$*bT3zV4Z=K?a9*dtceWdVg1;@o0&<fsDC61JTeC;wRwe4c}F6!DFh@ULl
zacXM!<t$^OM8g(<HStH9?`=ywnP&zQ8{BS(4CBh^p0bt+=TAe;U<@Twr0D^m?2~2v
z)U@XB_(TZo)@LsX<vW*XM33k^Tl`uv12ieQ1ylQZvRJ>H+B-$Qu#2XaaJ7^uN2BE{
zq0%ubmiJai52|MSUBCQmnUe+9s{6sAQ};CC%9Ne^IlP9*I_6|e($P@O(yqh^P4Sui
zVigD`9JrHxMG!*<j9OjjQnYKpkOCL=_Qyd2Q(||XW(oMVk~Rf_5E2CV3SLI77K|i=
z%D~YD)rT0K<UD^rO3B#@hs+hb4}A(Hxfuw<mPi1hc8ko7GJB+$qeXvXj`gCA9wGHZ
zhFn}bkSSv9I5A#qL<(9P?=8C@UqgZLKCZm#7XO1415T1lgnSCI>~H_;3a|-8V#_)Y
zJxkXwk4fy7w2-}u((hQcoxQvudRS~b*wR?ARnnf?m)XzU`^_Bxn=P6U!yIw6Ft)7l
zs=I(UgU#RKl|k<jg5D;^xzj#kspD+$p|zOiu;L%4hyYITwGcBEXf?kBey4Zox7gB^
zw6dbIVw6bY&=!+ei|1Jp{LuVTAN1%wsa59O8PLAu(>?KPsbYs;u%U;)^FS`7qbpXQ
z%-8sM?X!Txb^B+2Mwd!|UOVY6!?e;6FNv&yibC^NT7}kYd6Cl=Ywy5+66WSA7X2Je
z&f6+c4Q$VsCz!Qy-^qGAFm&6O8n*kbNhOdtgtuoZ@Ps;geq;RNSUioYc_zM;Dik!(
zcQWZZ>4ZsMoU!=0*NiMxqdA=666W*R6W=!$lxH`<u(Dgs!2vK|7S}mgw=#Q$BnrxW
zPP2Ur-WgW4B$g9k*7=C;VTQO}$TRApUh1+g^}EBC;Trq~3;6LOtblPq+5$W3Xx8Qm
zsg!AS5z&#-RNDO%1Z2W{Eh*wcpGl0%B)-?RDWlkhik{Lp^=!Pn;#<uQ$qGL0MS=s?
zt4uL=+pDyu`>g(2XF@v11;aLb)CuIW9`Yfuz%%Tf75SJ%_RLWg{QOMn908g}bV2ic
z<>Prm$Dz^$+eQaPcL8RgVE=3OsK~bm6rX?wVUe8CGL})QlGUPPT5!qrKqE|R)XLq8
z?zULsnops8U?T+M2*BcJ&8bGI{f<{A6Rx~Y|AL!;P~&7w6sTn|<8r^mNt4$0Ai(|P
z0gXTirtn`n@+*>(WXv34@Ok8`Em^+gY<YswLLS|&Bs1%n=>pH?_iQgs<wANcKq#(D
zy85=1!osF7+h&_hS`2H$X9wlY{=7?-%LK&72aOj_yj2hU9IpHVrnSM0g@p#54?-di
z@ala~TM?fb3}JAD#yn?3*}*_%v6Jn|%O@!W{T9xr<J|uhjZ|oWl7i2pI^fg|ahj5@
z`_bNL;}!Bb;eEnm3w3C4NVTrT&h%R%Azr_WXpX4wTImbW8^yl0J7;0@`FmQdK&-2c
z&Y&e#^N(m)Nw@WR$b#10v~JAgkMK22#j>c^Nwo3`ouXuy%4_h0EBqGxnET?tFu%t5
zls!+QR`&mDTMwU))W)SHS5w3F2HGAd;eHZVnv>vo?Wb6+6x|h<Gh930PGX<=ugr&a
zr=jEWU6#B7!U>gG)xta#^yh=a>3s97zf%e5VQk@;mGjK)+InHsuT*b+mdH)3QQYdy
z<6e(Q#&J0VwfmOLMhJW&Qee(=hvfb69T@2KckPZ*YZ7s7FWZdghrB7qP=vM|^IUb+
zPt`{?HDo-SZpINDU^w2897cSON#xH=dik0Oo9UZSr*^?zhEF&A14Rr;2njWF2GkO#
zhh4%_uiG65QKEzF4C4>0JHPhHbEpd#;!XglGIM|4%qT>zckDb&H66l10L5U$)fl%y
zX!o-~&~-FF>P|O^&G(2=8s-e*Ck^qhLt6fhHISdT9A!L|2g9S;;eeL_B8NB^682CM
z%74$4tUlNtmSJlu`Th3b1AQx)vEmT;fT{B>r0<Iy?vZOK6B?_Sq`nP))HU9>Wu`pd
zGl{pa%YhrqyZW$s6OWPoN|QuEFkVdNh_uITMcDcaGD|iZIGT=;8)zs%tqZ_c2gb_K
zmE3zr++f%EjxalXkP|zc?>!DxHgXe>b;<-Z=TKVCl{pmo!Wv^h#yJ^&$LmuNYekW4
z_>C2Ll|246{{I0&LA|~|&MItia1CpuR%itn$;dYg4%>%^hbyuU`njmaRKQX9cVub`
z2l52k7Uq1tU0|fv4#brPjGTMkg;jl30BB=QD;O>o(+;yx{ty)~f*&3H=>)^Ce{+SN
z{r7{HYyJid1ju!vRbb@6d+jHu=#`9-@BQx&%w?DV#%15oF$lDUCBj_XOJ97l^(jR4
z)r_)0|Ng{9ml%N5MnV^&E}mBa06+jqL_t&-X+%n`#|Z)>t>;uu1#c&xIHrxt;ct}k
zj-#Ba)v4M?O11c_@p|(znlst)F}s5ndBt&Oca+a@LTlPElQQ+>R_}h4b?oNVc?nbl
z$l576dPQ10wOPEhwcd?5%cNGl7f*WKk~%Gc>}1vq4GhgQMtJT^4U9z1is`frS7ZLc
zvCEmgT8*-(Y_roV1U}9yBQEPJoth)Mz(}B5tXVT+?k6)(4}8G=&4vx@D|W;Zon7HL
zMjsoKvfme$5*WE%W>7wK{>RMJga84Q?<wsh7qbCB2XB`B`3hh}yV~C@XYDvccOXh$
z426t*jLb^Y^_Ej#TfrFFy=SjE^Be&YCDT3Do~FvcAIOY=0?Iicpy1BRG5R~jOhIP&
zG68fU9DUT$Re_Ofue;X9nG}m9A;1D#_-}jtX%#+3FzxkjnWLGSgjJFPr?4Q#Kqe_O
zOnH4w%C4Y%iQo$*DNw^QTLhdW06<h<drkL9i`-CDrW6ll9{|aW$|qC66O@0`isN=@
zBqS#oB>+N%O&iynGv4;bP+;UUU-+`hlPN44Q#nXs4wwX`<sEN(Lj_}Gj}(gk@tzNu
zoJ>k#uB<SrasgsgRuMb70!OF^^-W5VPdoC8bMA-U8_F25z=)CW895to^uxN7rnEf!
z|Ld(ZSb#kG#1qX^p8Ss`>39dgw{y?GP^R9-<%i_ottljw-6*~5g;szO`X)}YdFO-S
zM?d~Q^Ru7-(q6`4_3OmI8XO!bi$*=qyWkUMVq#+cdZ&O9S3uYf=amMGT=<EL&7B(C
z!ouSo`zZ6sM?AdZ*2k3wvrzsJ6);lG$H;Zp|K5D%tKaY!S@%3DU}W<?>*ZtQ)fMIa
zTz&S3>buPZz!^V_C76~zjTtTGI>z>2U3R%OXU27Zxb?Q%#N4?wlv(MI$9WeR2}&f|
z-z!Z~-!4r$TIp8iw$kKX*~0KEo%%=#FKu~VZC*z5S(6=KHH?vJPqm)h>fNuFZ}<A@
zkw8$97MGUK=zFeDmH;{xYP}nAl?k$_Rd2tnf_}^V_qyltmjHdxZ%T{jzSO`-{H3<I
zH4M`=EN%@dOUuc}NS`(cH>%UB%wZ*8!7SJgvt{2+4L5SBo^h>>zV#~K=p0cmpWyvf
z4UC|e8q`kzEt@w<k#?ua<#4X;1r$ApAAYFyJFxGTE$;ICG0Z7&HDKh+U;DPXzNGA*
z&!BXIQ{K}~eU%v;C|9)JA|)_BMuMHSnvW3}BWHf(e5*vy2<$+SpH>?Rb3NvCOlir~
zlRr4;>%_rPJ5I<k^^{EZDs8WT2$|B7c_DwuO6mKK_nxKr0tA#lPQa!FJoG6|DwQ;E
zJ?&JpVSTxSbij)bd=!V>VwETj@sk1-j(xx}=HzET%b&eH|MEA!WiGq&DtnIaj=TU4
zm>O?*)ho=Q2OnG(e|kD+-t~dA&EC;5ffzE4CEqATfjiV^O6(UvA54T50XPT(SpY?-
zhc%NWmJIP+I<BRZ2fMg2<z>xhfs#mB2#8@Fu1i)hNyrnxgt+({*|cdx6=38l^O-Mv
zMa&PeRw$U9$V2r=D4(~=k@LOeBfuYU%={nkdcW<D<d6T8897=97@3o)nxY)AQ=f#K
zsRNK;ZmTEJXU;$GtWaR&)vte>*d<eDSbeOKNo4!pta;Dd-=O!R+-7n+rnGuhwsSy5
zTm3sS*ny4{_JP1Vz>fEA+qat+pZo%nm~VVj=C!W&%xl>TWq%eB@>YNm#)8+p`jvW5
z_9@GiADHffB}RbR0t+M<uD<44^Yw3hyX-Rleg6HGr>Jj~=h-e}q^yv@kGKFM6H{3;
zm>Dwc{L83s?h$yoUeB2jHknoPG4eT?W%~19Zt~J8FTcaVr~b>6&9M)9fW7DcbM{Bh
zHL$MyeXLzz#O1~K`=&RZX8LvXhkIm(>imy=V*V2GHQdy-Wlg+6^%IzAvpw>GB!5sF
z0^{z^UAxTt&yk~cjUhO1#PMQYV<Y45kj8+10pX{g{$?!_%bltJ@lQ9JOTYN#`8qXn
z!Vtxf7Z!L&p7?~vS+kRv%=*F?zhrK@>8}R=Jdb(wBQ?Bv<C{zSo8Me%zVp5B+cDBz
z3p;j!kua%5bxrI@J*!&I*InOkq{M1*dCxUvRyc6=^3+_a`D$Fu&Oxy_{r-)D>zj7!
z&`8=^aiV67;7By+&^B0pHF*+myL^ysT35+n4CWD+-(ll=T<R8=dbA=V(x5nEbZun)
z=5H$wNE01Yy86rV@AALruB~R@{aHVj&k6qc!E@ffoYvJt+6;t~ROUsO9=r2H!=7J7
zSB~F=RV#80mcJT*x1*F;Zt_9_n`g}=)5HYVkCwU}Q2_n5=8;<O#!>(ywfxQbpGB>D
z`{Q-bm$wA+IMLSpR*-cO<~g3b14i=Fk$7vU#^;eZOJk2VyRMT36gRKMxvo*yt$d7f
zADOK`*Qx0VxeQ3nuZi-qoRTW}-;KKN%PGIGIsNWC?=kxxu*K>bs<ow^`H-~1??1+Q
zo-(?Mdf!3S>QwJ~r{;)N3yh$|dFVq<G>?AN!{uutZ7%-QXXLQ=wu<L~lJ!~7e45=M
z>%aN=FZ`Fe@y~zpU$Svt@cieLUE{}h|M&am^2@KNxXy)oz(_DrHDCni>+ctf1-})3
zU?iK(OJV<ps=!DNU_<N@%xLv!62e4OTDBG6`R3PEFh<x(e1;sn-y>x;0n<~~FNr{p
zb#gxa&bPeIw#Q%RKi_n-Is4pC*nCZiEix>>9Y^2iX!Ft#SzW4K*T1;zH)dD>1uqo9
z1>u>`c(QriBOl?<+MZwh`Af|eS6*W>ik}p4kd+Cop`o-H-812RNJ<d{h>?VOA!q71
zfCoHrfpcdTi^-}_ERz}e6zNf26EcxSaLf_sUxaJ6-7rQ<%D{crjhHiJN-GsGrS<D8
z<$V1!my$%;dy@dbNJ=u!jb+U{-}x5vcQWJV4=8otCMWTE6j}m35`6+|O6IaqjuLwH
z(Fj12NTw|V1;FI|^FAc`dy9U5egGrI+@ze>56({_{`+C)on(xJmFLW5`S0C)^Izq!
zW3M^>xCdE(9cM{7oRfk*EbI`Nv(MT=-4g3Z?<{7q)~@wFq+NV`e&D!$d|YPC0s#|f
zr^><mS~++RhHwA(cg>Yo{nlQ-_u+?`lVA9}V3acO#ND*-w%9I!ul~>)FfwL7e9n1R
z$yw_|DlWhXA6%KiA&spv^``#-ondJa!X~q7z{q!{Sik(&;bZ41@=bH#fq&;0t@_Ed
z`Ot?yTH)`bU0}o&0;A&PFMW~uyMXyS+IP;ml`KM+u8|{ozgB$^e>eVsX*T+;^Cja~
z2oaADi?_b*bh9qdcL7+>JOAU7JxsUificW)(yLm`V;}TD^Dj^NCwu$*Klrg_i|~Vb
zKmBR{VwslywB!kx320a=R0Bp}jNz+iC5Bc2f!5?8;9lU{oJ>RV(YcgiT=UhiogWuI
zhGw*Q?&~RM8!crWqbNI@(+9r#{!+_3incV9N7%UDO!{`?1{H4QAL8xi9!>gIWJDSi
zMTBO7k!Wh&d@iG!bJl3?Z%zRuvtb&m7CxdE#)*1zt9QTo7A$Uxl{gNzBf;{1&`wSZ
zX_o>7SSUT(vu5N|>)nX6Ols9TT4{E3TS^J={gUb*sA5XX=`$AQf~Bl$yJd&1f!ob<
zIi}yfd(3QF&!(T&pDQNVGV-yzT-HQO`J93}gQ)A$4ye*jT{<;KGzE;f509v*i}$_n
z(dOCDe0qgkKKr>#4fguOLR1S#7!<FPt$1ew<V#{I{QjEjY>~l5$A0_mYhL+^mz9~-
z`lM{A-*)>QWmT-k82PStzF%I&);r46*NHI_+%4}IBV`5XM~`*~zfmTuQ1to39Rfzq
zmMN`C0RwB-WIX5NiZIukFmKUr{Cxt>*KrX2p0m!j%m`4u8JHFV0$@sDy6Ua3lPRr$
zf{yY&|K+clE3dxV?kdK~IRP_TN6R<MzrXB7_P77f4}S80=6gT>iDj?AOesoXz>ef&
z?|+PWVHL*6Rlob41w}m20mtw~bJ(E=np6Mnl=<i7%DCZAe=?uB^uMfP4`=h!VsE4c
zf=E{<6I((t1=<{P*g;~b+|PXXM?Y3Or65u}*=?Ouo2SR8<ZM18Q(7{;HOPk?Lv@dt
z7@M*{5&l|;Kb7}bC)Na)7%%ugLD@ur(Xm0#;~j5(Lj^E`^Yu$U`z0F>6If}na;61d
zVS)=_>GacH6>drk#)!<1!Ir@k7|<(#%ADGh(2j$m6u<<W+V>{p`$W6)&-=)kRe_P2
z+=N*(JY=SnX5qurW)b=)dSGN&>?iquQC-8rzWZ(w6XO&;w}pukg)Zjm_R+g~?X}mN
zuYUcTVM!~lwE~P_LJQv!K_4+Fqj8dbxBS@<4n6b`^O6^N3>JUEepZRQz8`H8W5j`H
z?s_d?ggCDhBPD2(4rkpsBA=X`G~0KLnoS!<6w@aw0cUCLmWT^5vSFWf<~h%N<~-}@
z?yXzRe}DTs1^_o~GOGrR+;GDm&3}FQt4?(DN5)G`tOdu-+chR%B#=934r;C5PZ&V}
z%l`0(KmN(!L#CPV(1)IAp7g}WS0rh~r^f^Ce?JQp(Vq!H1vz!xys4c2$~yosQq8$L
z@_gpAo@ISx1uGTwj}yGl?~@+(5Ch}o=D*$&cH3S%_~3)g3t#YDJHFlUhd){;?S2_!
zV(#Svs$gIG19iUSGoLqi>A6=6Pk8)e&BIT6XvOU?UojPt7U9BT?z-+obD}iu)XoAp
zY$sDE>;ikc^4te?;}Q+4U*oA4Vn!+Ng<Rva!`WXa|9RNB?!O<m^LF}jt8HmTMx===
zFtSVQk_`c~qOHn?bir>gmsdTx1@G6|JmT_<(=n>Cr(T;DtDJiESc?0A0BjM#x&dot
z7d)9baDHJ;?e2H_h$*eyY%$bXEIxNISv_p61GH9?;Je|^r3;MsrO`cK3JK5$*+k-e
zjkx))sLh@Q=}6jfqP!ICS&3y<o0nT{YZUp_Go@8DV#}J&nJ8?m?PymCi$AY;<+k%z
zEdN5G?%cyS+}^Fb&G6=Rk-e)<FB0r0e*HgS&s28Y3=E+QTL`2@>VE4Y3wOiMg5Ls7
z0VCmAwtf{q9#8w%r&MHs`47y1bTS-z$ie0%FMi>C5x3oTyLI;LcID3B(8<P)8?4gQ
znJPhH`{9qAXD}`04-b0a{ms8VwX!|<C%{iZSQr@@Hm^VJH5F!1P{dwv;U_DSRAY?n
zkVERXzVm(74$8XqLv}F*9UX$fJKiL-SOH)Jp9p7s;6r9iPQb0Rauoeix?AVqS~$J~
z#z+V-a?{PXm~%gQiIn@3wr-Y1A%NkO7e8NSZSG%IUu1dC$3AXuyzv&x*yzXai57Xi
z0wo@L!U^W-Pxf{n`xAZpdp|J0yzEyN7%7Ok0T2N@0|v=UUif@-<l%?=quX87mwx4|
z=IZNjux*B6GAI*TIi)QMxY%EG{+B%WIZ`%mH1B%<ht1C2yG>Hy5&#eZW=K&E+A+n&
z&;>{`GlSeHTRvJ!z{t1&Cj#se01Dz-hxGzfl$kl~h(paArPO5iv_D+`hd-I~F1%Q*
z9Bd)>nmlq=eCrrqWxoGCZ?7;vhBEO@r=MY_v@n<xD3Tr-Fw=X+1d7Rzigu%BHmo&M
zJNM{WiAkb*77IBkeEZG0=blxe_@yqVzvoPo-Lp$(Ov-(XctFL%2ZF0NihRrkjRYK#
zPl|DI)`!lq{$gBQcf^5ua2Gu$T;BKo51Bo?c9<RkfTP>@n8A%}LkC*q#`}2iK?k}l
z?a}RI|BS%Wn{WPW***Uc{{RZu$j<xWp&@@1%Q~sX7<u%g9%0P^l_k)RTQx3VHqVU<
zo2B@D&8uD+jApL?{S8)G>~0{#r)eyBfB?B*z&8V|hJq%YuyFDVo@efT@59P&p@gO$
zu-d#qe8J3wKOzcd6wSd<i+K@@%Ws$%FfxMSf4~2ud8R<Mk1C$!%U*JF1)IZ}n+p3-
zs|JklE?jWoMKX=$Ikb26d5}zK{i{rA1%tnS{(C&PSN*?NRJ@<n0BDysIAL-f|1$J}
zp+Nr{&t}a<7k}Eo@CXZsi*59*XFN?{-<b8E6BY-#9enUXWw+>ipH#lpe61kQgAO^=
z{M#unvkV=7;_cga*m*6hx&;G$^#1od#`^d9>n(pXupq+%z~oEyZVqca;C(#jqZb&=
zdj-Q`0u5hu^7AWzK;&`Jr#>CZ)FKan8BBBqgWtDL@Ln8#`^TI()Tde<I;E(uR^DpA
z-l)8no2jgN&oyjZua{;i@4NSRF@eukWJH>{0was{{OZXqcz?M8Ba0=+)m#qi)oH0r
z`ovnq|0bbdRuZQ4-J8=Ux?XK~8<NEU0^g~PDpuS$ViPv;$0>Q}t1@66>Sx4L5QVSx
z8!b2ZZuqMhwuTQ|^ZL^))_JXZFIFA8<#b*G4j6IFk#U(#9LNkve<RaKOU!caeCxX<
zv?>A2b@$6+N<GF%H9c3W<yK4Ql=<cN=gjuq@<pw0?X_zMOsaw%KW8S?&+)y?w{^9R
z!!FN8$9xyRy<=m*2xUN;nAAMFMmk9ss4M2g(KGnX=mTh=KMjIKJIC0(-Ct&E+2!Sq
z!L*&4BbF2x`Nt<c(LChD2fH@UAF=1p4(zZH-uc2}*1pCL@l)k&*(tK=l;8iskF41i
z|4vx$qiB4Xb_u(1jduHfT?`YawB<UW$+MsJbaT{^N4RM7M|R78>eHXK=49qCmQEh@
zKq>g1TB+beA$Y!=Ib#mRAF2T(fHm)a-`Qrj5VTZ^WtP|V@>=kn@%A@Xn9@Qi`;iMj
zVeY!;9`yPqC$mufS`-!p5)6yg@W$7^Ce)M`o&3i?`=$Ax@BBdEg?w9ReR<MDA8ela
zw5P}kyW4f+)%XU1TbEq=Ma$A4Uwn%Y*s*iU3twRFeb}L1MBm?=Zuy(J<nv#UFOr-G
zbcnfuZ;b)@^H{fLoq5Bn|IG{qOcNoi@BiOV&9AS#+M12Y2+Wu*_L$?eJNvPZez-NM
z1c>#{_nc)J2-aNIl*~|JW=r+z7gMAtFr+Y*laG;p3&sGL;Lto>`ppYzovo`sBmj6M
zuY}4$&Y0?Y>QkO%{^9YD_O--2Ww-ZxKX{JG=4Ibj`ORrTp2TN}29bLUoI6eJ>C@so
z7(RaS=ge=fmI)fk7YFP!6JoHipr4#jxdLxQKke17rye<<Dk|@Tq{x2z+uvNls3G5T
zFStkytQ)oH&&*FEv}a7_f7S$c&=a4$04W@sifb->@}f_RDRsO1Ie!cTM!1QeoEv_B
zy+w8FA^0e1pYimk&gbb)U^k(Rb>M`%UhR0m{`;F#UjE{W`Ug#EsfU@hT7VsM%ze!h
z^^U_d@dx66`ZHfJcip|P+r65}E&3qf4S**3Fs?n~;SaN8gacGv-T+3((Vwro$MG?b
zeuPZ?J<@;4p5G(Z#W%nEzfwx?G;1=0X8pQ#c9%c-5h(X$?dm@$04{M4Q0WT-oU4H`
z#PNp{PdMH>#r9w0+&}iQpvC-NQg%N5p(t$d-EXfnD@WPyd;eKx@9td{D1R6*;>K6N
zDaPAT`4VFM!OzQcpZyH0sP~r(*bA5(W-6^3U<8=)v5$Syj45q1;ePkKuX*~v{>%LQ
z;nLDqzc26zpn?9%{IO=uh&2a?DJuF8?Eydu3jhQF1;GH*=7&H0v7Y^I>)&MEy0zve
zJyXhXiv#-G;~x9y3Sa_2in$Gk^vrF%*BdqeouFrY%zgjf{;tNz+IP!l^PFctQ+`j@
z+bZEJ>ir))JCwabe>_zzp0IHPCeIhY^p((ObMV0jS>T0d>kqJ_0J?4!7|3{x<9keN
z@tnJ%F$u8a)Ecvkfhn!98b`U%_skTOk40f+cKTYQz(_OoV7_KovHt_3ne>Yh*NsDq
zQC1XXRO@@4R1CHvBNRpbTD!5TucrivLN0O4qn_Nj&*mFXvEV5qHmx5CZHk{!w$KL}
zR|14)x+*Vv;6hNr(Y`3=%yb9FI@YU`+Sr!R_h%Rb!vPiT*T1<OqVjE3t<+lYP8cv!
zEuY1{K9L<YgF|Z;yV!1d9g{%#cmW{NC&qbB?`KiplD#^z5nx7Z*BN~;!5E3c^pd|>
zVRQV#T;Pw{J#S3{;m^9Yyuwp0n9>3`wyH9%kO%@9^xuT^4fx>fhR#c%6U_fcYF&M9
z3--U@@7lR98Zfe6PPvb`_rh-f_3PK0<Bs(dMZqfOWkTqxt9~a?U{trf23`OAKbVOC
zLnD~3ncpa*U;5&c%XYb=i2Kuxf3^-~QNEziy3bKZnL`daXkOX0ea8;-na_T~D(NT>
z$L#FxK1@pfLk>QOOQGQ|DNO$)pa3NgCb%Aa{ILe}LHGUpqbn4$E`uFAcN$E|u&enN
zDcT6tfD!I()9&EkUH5y{|1aitIb$cx388$8Om%GvFh;nClI`m2uQxaR>Cfi2JMM5U
zKqT>=Q0ME{-f)AQ#1ERorO-!_xLcqNPUAP~Sq7C_01@Y3^eO8+o3c`p_ZDPz*kRV}
z7RvW(fx7%&b-MnBKbU*AZ4-zfpB!m9E0^Nz-s(fo*RJazI6_`|`4C|@c1pI5@#~O~
zQqqOrU3Z=NY{_H}JqpGM%oS%!%P~lZPg{y&i=cD_%)xhwO{4xnATFk~)`;!%kQ0tG
z2g%>SF-IR2st_bTY@Xj<eVw^eAk^>w^ha~TvB#P%GMDpU?RsbTd04=){8hiZ*4!<n
z=1+flnZT4j&)<p`_DO*?ge~fy_dfDSE87!KJl<gPEiCMk&yve!%56@l?)L3FE$e`Z
zW9wbp%%=S}S-K%~@)K1r`tvcz{C%i=$(#4$I`y$T)W7JPE>r_X=qJCs>bL4&cUeWQ
ziywBRZv$4sqVk8I|KeBXhd=s>|5C*{id+1D01g1K1Rb7}peUvA?NUl(P7RY?gF}O2
z#f-=tQ>E>FcDbXB51Pv(pA%1bu=&R)J)xq{qxeTLPM-LB!Q>Fmsh|2U|4g#7K>n~y
z@V|gh8GqPk;|7`2I;CRwzdsJ=m|8pQtdCezV~6OS#FXE??sY&g+Wc==ahNE?eAJE|
zJLj+S4)e@ifEjSC#%8Vu!;IdEYp=b|jx`0bmwKgS|I1Ccm_2H@3n>5JEXEPc9e)74
zg+Vm7cXva;2(tXM{5YNP;NvZz!+6*$N3uNIpz<As|F>kS>vvaQGhZZt)Ps*d&g`%E
z;y(8|vcgQ7i?~%|%pb(;!aoUM6-xYifiVIg0;W;0E~Euio0ycDKJ^3S;KoPh7<|@T
zB@=sLf%or}Q(jVGk}nvSx}J6RITi?Ziv#NMq<?sV1z{ZE?@!14i7yoD<M&0xI0tC<
zm%rR(qjL`sXOqULqa`=qOMgH<t{=fjxz_>v+j04*qmHoShd-|Wd{krejW_<qTzlR1
zWisYDVXAVAK$v>r;!k~6<JYbAqD6Y2-7q<gyFn9#%VSq^e0+?AJ(Q(dP>0nes%r%~
zk&fD@3-k33?L4cd?2W4K?d0Rep?30E>1kV$5ox-Nk+F(3gTJqo%hdxB{pOPVasB5D
z3bMrmM(UZfawe<&^^E7a_dV1JV8m;0`9_I$WdI0JVBMc;-ppdVW!FlNDXl)OYdaOn
z-;MH878@98M)_Q;-sP=WJ-^shy>VnP)Z}0@vaKeu&aTJ8$5_V{p)bM^ar0d{W294K
zO*KiZuIo{`M_7<r__JQP-`LsC@kz6BxKa;6YG+DoTIU(*7ssXlTO<7s8;CHgKe}~?
zS-;<A`5l&(6<M4Ke!n$VgZ5)}?mLmM0%qT6=gC%Dyi}Xgs?8Bu`Li679%^B2GN*0O
z@`u`>tQHLzIa-S2=W6F^tKnlZr&4W)bu)STCl8#}!q#Aib(kW_pD)UzKTC0hNt(SD
zu<`C5``}~EQ~&87{ZVV4|LQkam>>V}C+010dY!>8dn4h>D}QUg{XgHaFG0VQR5+Fn
z`WW%&^PTViK>jp-YL)skIXO&!>+35RBmVe5_{mSr*T3~WySqIp7DQ5m%-ddnYDJps
z1WtYBTmNHTe)97y7~s#te~!cKFMjRoQv8agAZk=Wis(d7pLxVXPB2e>@{{~gYo1^F
zm2a4z{_I!UE!-<sM@c!SF<{gBeazFI@{i_-!wv~mO8PTJS$(AdhHrf5duDVuK-<DV
zf*G(xvR6zI&nY{;Khi2^Ml6z9nXvLA_jr!m#rCkskwB5O*ekh-oH^p~L(Lmb{eND(
z2L9eGaO<p(UT8ji<{6<%=ddhLq`v0OZ#R>e($b4QtHnb?pbsXt{^1Fa3N=L+mXEvk
zn=7so+vj`sr~h5x{6AiQnl&5e;?_E1(g`*Q^&?aRMqD0O{O)S=-|9Q{LU6}C@7>2f
zev#R_bz9A(0E{rU+6nX(fKPxQ6zqUgVF3j+JM&L?@NuEac4Pu5@=diDcGPLFeN}~W
z+2wobf5~j%A1loU0=_))36HHXvFFlpgr#!U*&nqGgGWB%B>%0N=Kwxem;5Oa5r7sm
zQ>}(CeDO;Heg8b491!3cIO7&T<qrTd7hLi=lP=~flpJA<5eu$f@NqF{9%V37SZ$03
z5dDc*hL>Ib>-qjjxc~r|giF~8(qEJ6=fC*nW#$KBW;7CL=Y5Vk(meJtkFwyD13#*z
z1uXg5&wo`>&!D0Nb_;oi0WL0Y*c6y!1L$CJuy=IAtlP5D0x%~Xf4o4Olf<xC>!MdZ
z0$l!3^ATW$dzUet#T-+wJ@oINU2%nZ%Htnr0ivpf1a5=@od*hBe%^DRWdZF#^!cCE
z@8U~7qs;}CjJ5el9XQ#mabRtKIXFZyVU862V`|h@ae3GKc=hjmMe&!E+)&idV$M;0
zHd#UEZoj6kZnbh+P(m)m3cI#&x3#mVSjCjyikT(H7@0}-)iuXb?^(7cfu%`R?IWd;
z+H|5H)GN1<K3gx3X74vDBfoMyZj6yQ%57|HuQt>y%y-l=r43DCrYro<Z0putX7iQ}
z_H%hk3YARB_e5|`_R@ID9_Gcyl`NI)hcNk#<#T34N*%sVyfGnR&#S!`E)REfZ3?fq
zyW8xsTp1b6JIZm^)y!*;w*z=Vsp0Nc9#IWpDy2`~i0)YjW#WEgdZj1;0@l@_He)oB
zzuTDESlr6`HFAY_y`YUf1M(ADSx)&gaP=9}ns-F=d1d^{62r$FFW-r&iH6M4ROZ_G
zD_Jq+dc^1-)bnaIRC`|<O;0@k*xO9HrH&hXKdaTpQ>W5%swq_GjHi$utd_<6^#!GG
zu%FIHse9rWBh&gmn$-?}7$ZsP##poFrHf*$&B`i5qhoFq&(FVxuG@eW5Fv{9X0gSv
zz!(W1AL87mIM;2lpJO}ad2<^ZsmsSQH|GcQ4qLN?jfbJY$Thip=1xBD2{z9I4?aAY
zciC?;tv<jy=dyiflXioiaNMzFX`A`+i(gzO<&US>ZYG^SE(0dPuNXktcEH~qV2;+#
zdz>X>t|O@h2a8{rD|@BnzwORD&5crCV$Q@VkI0AojyU8HbJWrIu?h#S;V2Cz0)eT5
zGrx>bj{jM^W*`5!$4IH#XYUago2S^B)4`@83uR^U^51diop+ckw6i>|Ng|_(<I#_N
zxE6N_#Z@G`(?rie31Bji?hiNKX#V(@n{524=^67+PkfB^x1hC`4aC7T31*$xIe&v#
zGFM;yd+W<%R^_E->dE>T(av>$xcQb_<zM6?wXe@SM1T;ChYgY!JARA$sDPPpy;ual
zyy8kZiJs8Te6btExIiX7bG_!MBMvk7)edCjgKX@-Qk}U5OM}K!C*=I=tFAV;-f@Sm
zXG-R)rezkaPhdt-XfVD$4m<Q<F*S}f2OoH#S+{nL2Q(?&w05R%-L}L0O(vhNmooPD
z+hsyVCSX!B5mu024=Y%e&u(}8mmrIROk)8GaUa=I{;U8Jz!1uuk?E}-Jug+yn%cro
z$iDk;F^@qZt++Pm*rPn>`-ty44+gfzPk!+$^REJj()J@rsgQ-*qJuprFP@9Nz`6eW
z$~XSo^MRszxDGNag(zU-1MYuc5_)yu;CsmSy(E7<n}q6~(!bko6B|cN9D;xHr<}(;
z>XGuBP=7H*+MPRh%k125IJE&&bFjj~LP6;5-2yhQyY~0?Hf2#alumk*W$z<w0Q@&N
zr7V6GF8svBVcC{l+qh+ud5Fdr|NB=KA@GA|1apLU93}Hkv3g)}V7hMp`32Ghf8rqh
zcrg+-Z{Ek?+#3Zwz89_$=m0PRi-C4v8fmTiXIOw4cTW}2=iq}6($4({SpPt<*Lcpb
zE_UtNZSJ{ytKGE^3j`)!&?iY)S?(I<u<p3yPWMyHrxEjTu^((7oWELrGNWhM7oepU
zF{P*R=eNJTM(^7KaK!>0wEpo=c)~o3?$3YytGP^L_=IE~_MsLAi~v-j{KxS;CJJG(
zz=}$0p-dkFXt??2Tg;VWBw;Qp7+|gdV1=2g3&wT7F;C?OdEE8#Bc<~+cxg4CsGP$O
zJ5=M*A!h#r_Oom&m`gB}=v%NVZqe8Y8?jL{P~`PMfk7L^^n;m3U|idyF&6eAAU)5J
zzKT4P0!#h#8lbynGAbWfC@|=lZvYzcakFjvb}Ks=Irt9p2YjbI>X9eSzk~jJ{_|@E
zAi^Fw;)uh`*r5J6Vdt=@kVP=S`hp$AGiE$zUdgKuU_x!1`qy9odW-GznDdG%Fo*LL
zd7e&wV;U1YC6!l-f1LD}>wOfm(^K<@@^R_?aTep;ji=>hiSekuLBJ$IkLH+!z#VqX
zxPDt6zp-DK=4wV!;m<X!p7qQt`qR7fYR|*nZ}k{ez8+%)``F<--Td`yrr$(WKY#gA
zo`;R=!TkMgXjG9C%wswKZbU}G_D#y9OuE#Ux~8;(((%@@d9fjJ!s&oa-YXaU<1%9q
zJM;;8jft!m0Z>?{_32-){tFANQ<-Z4mc1SR02k%KmF<ml-1BCZ-4oMhjW%3Wa|BEL
zy)xrFAXYpq7gtB>9rht#UTP<QdGm+Yj>gLIzu%3e2#cjPjn<Zp`uHFR|GZRfPGLuS
zPH|`S-8`soJqM(?8u;bTIsZ*-y~HyIc;Wf^<G6Gjxn6Hxr4KRxy5Ra~o^_&pSnjiC
zP|sW7oB$FxFIZEU>YpyqU+K5(kS5S?{r##;4(jI#t<U{q4$p_D%0FsS)S8m@bR5T(
zuyw7f+*XLDf3{VHjwNn%tUxK7RDbr5i{+J4A5-56z8~dryk86NTJvLt%WqQP+<;6&
zPft%+V8l#K+WC$RCcWAS?6Ij9-W!)L>Z5z-qR!W!UvLcIUM;ppBjbuc|7OqKoHe07
zF&GNWR?l(a*;kLlold=Tjxkbg>{^WHS-a)&x!dToicgKpYf`QLY=}uqU%|S`|GfR@
z<vCUUV&fyay7fNGF+pHtLf*-GM(1*KC*AKjn?pTdq_UXcmDUH$U`|o}_RzYp-5Fj9
zPAS`{f;6gnh2`T7Se1zA9sxE90T|j=Y|T26Hf0t2e2~*N4M;LUh4gXyjT3X%o^;79
z)s%c=pzt4(Q=o!CjD$SpOpH#Mk#$4n^w+(%!udM>NY0egd4LNC{>%u>STj5*<=V79
zGG?uh62K5Z5kQ8F{CBWm1o$Bx!PSF2Sm;fOeF4=iFBSy8J8*zb$mhK}lPZ-?*u^8D
zMN#<<t{IThOs|XD1-QWxA1N#_)AGqNErv&WK(snJV@C!l<qzQC01kkXUMUIzCG=KU
zSr&E8f?|?`1C3yjB$Xd!<aAEoz^?+Py>O)7qw9bX<cH4^fF+il_z?kYaR8JWO~YkN
zQ~pA9ojBABhDv@?4%`X4Pc)nW+Mj?%<WC(H61tA;iqncub^<)`G2#VslRR_s`=CV+
ziwcz1y_)QZHy{V@ejg*HL{$%AlUb{Z4FAHsT7hoo$x+&Y|9tHncLB3QD8B23W(LA~
zd9`w%e&PCcE9vW{>2&_}$`5;nwP^U`lUeyH8Y;`yjjIzl*OL;%AHskU<oUr5ony`8
zxwzzk(jNsspjKQ%l+UYPUSZEBOfL2E2}`q-*IKD>wVfhM*^Z?toB6;C_#m3qzc@Ow
zjMMXqsxFQ(5=VKvKGBI|qvKE~%5Ejkde6Jw{Z`6sMMk8dFR>1;rUYoEF3TxRJ-G$%
z^Np9%I=HU`?6^3{Ev`FG-8w!FG78p#^=g52y1uEr-RGTGwg&646u*P-hQAKA*6{v#
zOL6YiyK*=1Qq-keX6GaTV<d?QEdeBh0!Sw1C)fEHaV8|20cSf`)0LdF`(9{#Osus&
z+A*3H&|CifAGP%2Ja$Fh)^2{`5~ww|g{NrjY7iJ9RT(hSKd@dqqRW{Bjb$C*SRO*l
z0MPg#&}wU%UBJ4*1~VNf1}AronvqTGDt6C|PmCL>_Kh~ZTctepn`cBgMY<IAMS%^@
zJk*Q;1(fC}>n%G%3VP^@01;CYd8_=lzC5&WeB0|zm4n{$=Z=}7cS});RvvZ>Aty!(
zX`E6RK%`GV3`(sjIZn@vU|-*3!~lfg<em5gfD%|P!WAqiGp#z{2;GG#0THQ8+JYW^
zGI=#6kOSreAcPXjiJtP2jA)MqJOo(Gsy%=SLR{=U4=%8U**eWhp-Qk0&GV&QWxCcY
zmQYbU{wWX00$I89ynq|pV}r`W89d;N1xi#VGUS@jALT>+w95#VNlD(M13Uo~=@sZw
zP=5U~F*dFC*r58)31FJjE~#GaK1Y_A>*88oJIP!CpebO|OXUzZIWlC+;6Amp&p_oA
zMaw75m3GE)o>9BjXcu<14{O)&h#L&_#eTJOuj&i<gyX^>a8hlz3i(2fb&7D}@l4!D
zt3Mqwe*W{GZGB}>2N(hWDL)EXImjlEj-66`?D~NaOb-yk9Me|5fAXzz&rY*t-+gSF
zmHemHGi|orVG|HEd>B##0h`SLXa$-18u9@nev>g_d4&NZ)bYatVt4J{-E4K+j_>bZ
zZD-cu48r8u;Uc@)$6f++CM&iL%&mgPVCJYO6!n$kY3v2ja0Qu*2%ZDyMw2ls5lZU=
zolWGmnL_K2DCBLkUM7FK<52Af#T+;*2BBLE)sDSz7bnE9kJvX}ZT+?odu5*0>sw*>
zgKN|{OQxEQ?kV*dXWq+S|0vgX{(PG|uN?borC;jjfg&x>Fs(J~keocX6ZU8$;5zwY
ziz2kL;jG#Y+OU2R_HMX($62{Zy=SE`Z`6d6Yg3$$tTEE+e{2X{si1AM@<*37I4E6)
zz8O>6^fo>|CVj<#Wjgmaz`AX=yg0`Xdw=6x-=(bElo&=SjpM6zO|+D<kEZPA)(>H2
z)|%VG;x~G&E--T7!G|fYg^#{QGwwt@%wFN3)v*;?Ed$dbL5IiwyxRTvk8W@arxXM1
zjE`qcpPY<m$E9SK5*=p4fIx_8u^9TqEPz3glyW{NHbS3(2!6vjK$(xy9b5y&{5(4&
zCvX6TzW@kq+rRNOudeVhvQ5Cq`#y56sJmi$2oUKN0FqaJ!!S;iKcGlTWdfSa=r^XZ
zl8Q4UWj)udsVoVa*Mai;R37niGLZ!!k{5%80C3W$2ve$4U*AICD+q7^ETEp`kyKqN
zb4KYb*drha)(Z@jJ^?6zDX?aWQVPK|Ny;1-V9m(z(7f3vSR<529b7ke;7L+o0>DI&
z)j=KmRMzdc-YMI#suLiLs!9I@v;hFY+?Qpgs9eArk13*Op?mzcV2a{{3fXc5c+iGk
z#kW9D2~<JGfLrbv!$c`c4qTt1ZeoHExJEkxO=e3w>5Fm#>%#E*5i^+XuR4iT>+^o#
z!h3S$xbRF|{<V$_UO`ho!T9Vn+kew$>+1p&It2WDeD14X_brUuA4pqmL$JQ9?RVG&
zRO_P(`=5gYgX$?>U-9>^u)L}PBOm$b`PR(Ya*=n~M{acULAVsc<QX1w<zMypv*p;@
zu>R9ZI~Y&%%m;e?@Q%`HV5HTZ!kvCMnvtE~wm<0_0Y^RmGoEh%R~Dut?HWRxEm9?)
z1@4~zSU$J<@06ccQ-YPpTKZp|7>E3Tda#AH-j$_P!cylMtqH>vdacNaH0-Et1sH)d
z?l<2Pl;LvuT~BUd_u1g&)*tP(v(fjol``CXRSg))Y7O7662Fytk*40*>>n?e!x+VS
z6}CUy#=>RF_eJ$@8~3b`O;mwAt$8+$7hxtUt388ifLU0*!rl!RFU(GVu*~3Zm$uVK
z>g|A$T27cdU5{@5<r47E$x*x$`2NYj6wyXIla#qvy{Bv{!33-y7-?iZR$kMU@vGK4
zr`;C1yxlP@zgqXg;`^_4a_)4dw5By=4<?-D6o2W<=e#u;d*+fcml##3_dbWs1^)3M
zSP!=bbMqV~uUM1jw6S|^*SHzkFk*GVq{Cz$rX<7G&B1)#?^<(PYcln6ZU7jm?PJ7M
zu-1`}6zp9ily*YDY-GPt4x%}>6K;?njQ{%bSKG;=ENxh4DU0Cy;RXth&QwUP9g39;
zAn#?cK0Nfv)2EN`f|w5lnO8ztkC`caVPr5PRWjuSZ~!O*ID!%%N9NPxQx;qx4d$O<
zkmO|xH#OLA{erypwAY3LBkwuuqf*?9Q6WV<AjqVE1^^kBlR4F6KxVw0zm%~(lLAdL
z_Sq){0&tz+*d_Ld{$S&mFuF$*qr#eXg901`3Mp^u2M~fn8|DV}o5AFk1t8R3)k6h%
zNQ*!QOHWa3XYPvpa<D8EXgjQnN64}Dv!C%)d##;6XMXrRbIZ+tvt^-p0`%ZHz}(3L
zbd)lOor3J=9lE<hX_x8*n*=7$tQaea0Wn2j#BfjT;y;4R%gK)TE?_1Mz)OOaDXkI<
zMK@;!+PHxEv~}y(R|Q5~ympUaA4m*d03!fo)%v<z(3q2-_uP431^Ac1x5&0_+uN#A
zGw)_w1x1{=)}T(GnAXHnO_sH)!UTm4<B2NH?82TO`GrkDEM!w+5B1HPbMs1O0SjYf
zK)Zgt&{Ou29hTQh0Y+$VP=@su8$8ce6gcz|Y+vWZsA_f>a_PaebZE7{Efv;eqFV1w
zlUc<gdg}^a&rxtv^GHJD?%cd7EjN?2*1T%t>Gr)@vuSNPbw2nWgq2upQ5_b)lh<15
zJARw)u5KPM;!KydE^q3?CLsP{N4}cuUAaqr3>(*@s1tpp8qg3${^a5MYAb!9nL0Et
zBdvqAernan$m#$_kkNR>92n<$D%fX6#qwTLa=z}r7iWdA=WB)Wt+_FdjeXU8xp-~&
z{J^xfqu%>o3ab;q$e1<`Vum@>D}Y4bM`1b%_l@p<4xK#r<hbxX=P%0x%uxE3%Xn9I
z(ySliySx0lKdn2jHRfRQb|5AHH-R<0OS_aum~shiuEG+)QUZkaWDqtlcX}SFSEokq
z`|HR#en)D*m!7|fo%ifATlU*zi>TH={dv2(R9njej9@}5E5Eb}F-B%4)F<^DrgL8J
zsps!cRyA7uPIBe9bF4pRwUhfoTIcg);Yw%En0Jsb<yQ0M680<`tyL%2_tThQTSWgz
ziv`S8led#)*m&cA9!vS2`0Eio56Z7vJA>(J{jN2))k;_IUMMi~=i6_bd&WOK)22P(
zUI*37W6|!X<%sFj*S<<03GGo>|M<s0ny-EBn~PTOs0*5ulB1x75lVioXy&u$1Da(a
zG)>2<0RX{l6W3T=aoycV$qthvrO9VXd9uU;q{s-Zf4zK+;FQ@PcJJ9^KKjw~%x*bD
zPKlY}IS0g9T)87cl&%BX^*^m6rE*RH4J7#E(-!V?kNRYE&IBr~MoK@=K_Y2T0HRo@
z{;r+GcWtppTA=#`7%6;A`vZ3{-}Nv7BTst5<4d==x)9Hkl0Q}X`9e%%K6cTk%<XsF
zX#p<l_+Iq|0KzO;N+zeMCl0=o+NGq+;1;$G<Oq8P*;{jBx`EXzz!U190>mP+0{10q
zGbPZ8>v<h%q6J!ruXyx>qL?uND}<y#F^NhG3^_5wq2F#TFeS_pJ<mkxS=fk+^e3ke
zk_B6#m4%{M|A&8gym{zDPMp7qQ}VBWWBI<L^9||HiT6=g{@%4)RB+<*=NI-2aRfdr
zb8bOkWX~P}VPbmq4S2vvzi)h)!d#qoR69kFd!L^BoM)Q*9dnGoNPB+4$38yqDBg2S
z8Eu;@3S8t1wPfCCRmR80^c>2^>PFLc<E<<SOiS6*ulbyi(>yvUpJ55Ti-O%XFAPW#
zw6flxI6l2eeQeI=w9|VS`{=7DOv-8M0vNt}+TBb`!{*U?>9bP$v(C?G!&4j;nDy!(
z#r<k>8`pcrMyYh)oK>c^TGmppEq3kNV>WD9*J|}wTspo-`4((8@I9Fu*LTnGK-qUE
z-<(q#m-F(eIV(kbTHj~xyR_LdqK;qC#~+&p{p(~`Mz!@J{bjxS3tzAGLbZ2?bfaVA
zGI^VkiLsut@A7(iF2;Q}UTL4FSvNAc7{ztVXh8yuzX|D#+Ig2o_l}yLOvd!h7R+oy
zIxTJPoW(z?^fFixStDx~QSORwp_ye#s=bqCk(SAiF>T}>)`s3j#*w%+X>42@sihwq
z8FU*k<5t6NeL5)tHvO;xx0Y_P)$;GW<5rvWtKa#`{JYWf^UJ<B_v%-^#opL=-ve6B
zJMwf4Na+YDQy^;A{82~^Ywg|$iQe%`c_}&oE*65PO2nZmA*hsGh}Zhr;fEh)Ui#7(
zTa#M;TyS{&iBDc^{&vgX{Fj&AIZT~-pCN73eirm$nFMI5`CQHdGClMs&=mz7y;~-W
zCKv)fBk&PZ(1Jwhiz3LhS83M7%$Mg}-bR%l6#+iFNnBVjUV3lm6Q;{7n+Nbm3;)?U
zWkMPH03cma08d)%m;>*1FY_dUBKJP*FfUX6-L-3%`LIkV?UlmUf=gO_*}OrQN|j*;
zG<X*5DA+nZ?js6>N$~)Jc*ikI#w8apV@CM^Z3IlfS*DGUI9OGfK(k;K$Xuzu#7l~C
z16W193APQy(X~0HnH3X*fWv$K&4*Gs7OCRC>Pg-4X+vGSpn3uxlu!cv&>&~br2)9H
z@Z{s)B5O7eo>9VHqs|19T3}7_U9iu@%41>t!2gbWo>ngU(_4~V5fg0G7nyK$&tGy7
z56sNTo+t98PJOE53_zUX7HO}J*0+yMCw8FB*g2DEYPGA9J<=Dwd*1ayv58JS_0?wW
z+O=iz`0>LZ{n-5YM?X>17ur=3X_a5_<hi`5#EbU`T|X-ClZ#jWVwdE#m-T}Gom=c$
zEQj(c*IKRDdNsb2tFg`DJ*4^=(~68p!-kfkeTSPJ=c!BF#=4+<+;1mMe|^1so^m%#
z0V3s@)}C`xdb`aQE|*Rny>Wsyp1+=*Ds8dyjZ;5y{UKWU@*WYYtxFe6(u-XVCTFU>
z8w)aNc3kOvE8Mu(2rH)ZwP~+}DAyM4)g;=~=+X%7JR|p<z4=WWV@qpQcl2r;?O5RD
z+5NkmB@pMaYq`laT<`6zt;+S!@oE`;pc6Xqd+xc@HuSs;-#@>~D|5bn;K7H-OA;Nf
zO+A`Yhl_7G`&#z>8_vWR5&szXxzCYi%jV7IxMLq=hK9m-X^fALo2!0%wN+l<cH13h
z&)&VhT)L-=ErCY+EB%F?{46#SI3ARCd{%=>|GKqudWJ$=K49b+G___#rs|R=yLZx4
zbgM!5f0^2owMzYjOq&sAWTp*A^Ygw^bj^YhQeXo*p`^zI8m8USy<z~RWKvCO4ms=~
zbK<cNG$ShKfc^G0`)u4`<J(O2C!3ux=UnhfbF2CmCdqI%pHO|00s!zAQ;=U3%&8T$
z^UIot(ax~^nChx}SrcR>tv{yJs9UDwgM|egteRo5X~txd4bTL0bL`ZDRYd@-D3(|^
z_Co0oeqCsrHCLu<N$Ld{Bw6yWC<X-3L|;0iV`%{gs6x=X(bh$vB1gA?1hlX~6HF-!
zDhTWl_)-)gf^Q%I8v9(7g#+=5dWOAX*YwJ_5NWL68r7Hc1ipO$qts@AE}jiX3l8Zz
zAPXQ$LB4^g8Ra{$!7VxMH8DC8C8o!$o*lC18LG)<&bE(6dCaK17)c4$1$@VrD-&>Q
z)(@HSF`3a){J~7QX&o=+f?2LZ4nD~GjC#NW?r-+ncZ<7ScBH>vbM1BJ_S^5YzD91p
z<BqcXo%&JF{8gvQZ6)7M(MemWOE*pYBv8~i&qrqz&cGMj%t$NJ+`ebrY+AQ;Y>g=8
z-AenHDouABsxSFgWJDUbD=rR<5x+^4M%SP9B~v>_J-OApUpx2i-PIt0sK1$)i_B`h
z8*!FNt$HsP**8<qFmt5Mq+hOa>4U6IIVNQdCObB224q4lsx|9!l}l&p-5JoiGj*+<
zEB&P)J<DLKJqtp$OIeJ&aeXFpvJNzD)(@iBif8E591&_tYjiF*chdcivy$inBk}AI
zQ7>pxvNQO`*S}U1QrNDna2b5<8{ab5TytIc&F<A$Nr3(`IyPY$BRFMG$jKSNNK)oL
zW(AC-*A800C>Bi6ISP6L3U3_VqqI+pX)%$NV|$(Fb%Z@YK%wrK5Gemq&=YWKkF)zs
zO2~6B<rU4t9&)^S;j^D%@4G)J=yPH#{NYc3GT-{%56$-NyX-B%2mqF($}Py3OcGNP
zDsyTwZ$XW`t|i1=pbW>RNXhPQzv5#?txs_Q6)-J>38%dBf`QVn`oaDn&a?oIqS`;O
zrqW)oGiQbmpHw;_^Kl6?D$_Ir!)Y@yF26NuH;(QBMus+TFw=V{=0OY0o8hykw4r7m
zFR@^T03v`4YYGmoxa!$Itc9WKg3M+l`;^)MIFiG!4nT~6Ac6;`%$l^oDi@~5r)_?g
zF`_j7kdgl%WR?|Jkt~^agM5Jx9LmkeKaIUFAk2a_7Icxn905%QmAwFr^qAq{5i_Rt
z_6hJJk6AI0aybDpI_J3qggJjj12WxHO%dxRvlpNI0(015hxn^o^ZfeX-(bG{m9N)~
zvl?$Yd9KE`ciUWF0;^R~UtcnviLqMc-OM<YmrZkTCz^xhJM%2s%{6Mpe={<2bsrz!
zJ8vec9x$@hprEUB)W>>qt9L)@`otl>sQYFd+7z{N*uTLg>Tl+#<dv5tj8xA8t2->;
zb~Yr`dN*L@56hHRy>TV1PMyA1tKOY1yPN;wNr3mV9bg1s1^6%?C|Q%)x^<^E+-)pt
z0KPP^5-}J<Upi$K?Z6NJJhAnj9cIhE$R^(Fu13TV*YUiOy8GkcJ7n+ny%l}dA7^RK
z;|ka;NTXA8L`;B@_&<P!M-h*+s0Rh;#ZUg{CrSC;xMTa@TydrO+pV|7Lm-P*3)fd&
zf$^ANU_^K|39k?RWh^^omG0EMSP-xw6G#a?Z&(jW`HIMmjtVFlu`C0w^#VkQWdOL~
z7!oi>W(7!~pf6(HNda?QE}k_nCJ_~-<nNVoAI8F9V%AK|3RsZR|BwR@GAEsQyvmm?
zV1X68ckVHFZQE*YyW=i1hUqLR*9n-8>JxYXAb^P`7)_+<gTYWLR~55yVPVXb>}63e
z0&)6>w3DDD4h}F*GW~sKTtEWa{6T>RlT!X;Z+KcjPXbUvpao%2K*?0r10#|u(*g;)
z4m$(J2xg>URLsugRJT403}Jo>Ss^0}EWr{Hn356d?Y{5;06+jqL_t)DF)8rIGDoz4
zvkVZ$n^k*Uux$ib5C}zG0V!Pj?76N1R?Mi(9+iV^EK5UR6EdOR906izJ94mQw)DGC
zX00rEr2GM`052%d`K0mpXzG?8N{bzn)9#&=W$3^@5i}z^uz;qR(h^&R^a4g^RhOBf
z*hZ>HT42zW*f+y#2TY&ZRu33?)FU5m8EQ@j^-O(rRFqxR?=wSpcSv`435bGpjM5-O
zOG<YP64Kq>4I(Ka4AR{oAcAy*l(ctvzwh347k|w<4<~k<v(G+z|8{K4N<e$-{etdn
zRP+9^r>%3HAM&1$)1IY?G>J)aX049=Lm|3oar_*|$re8@_mSQd;`E;3Fb{F3Z%%J=
zpf4_)Z+~l|vl*9>&c`}`>GEMafV;IoC^u_5ETy0}ug1z^Dd6YrL(zjk;RroKLU(QF
z7qe}@N&dGNdg_?fY<7113e_?AQ6Y5k0xjAJQS2a%f1Qcec$IR2W&Agjj!f#m7D)C0
z&)?0|1^%{xX1$~1S?Vbi1+a#bT>EYsQe{VSFnIV))lZzkSenyuDUG#x%-cOwyoKU<
z7IK-$6BpY{bSg8<P;q64AOgpo5E+?UEVyETfoc#0P<49bBZT;nSw?v{ZTR?-U48FZ
zr}cr!2_%t<(Q6_%=m4Z<)BrgDt7mBgpf+trPh6ahh8{EqW3TaJaPR?0l3(LJvM0*~
zNbs3kGMBd%{Xj#fz=!fP!@*lJLG|blJr8Fi+cyX=mN>HGg(<FzcS3})Qi&flP&Is!
zp8>_cD?3%4_(I)|AKtF^<}{(;y_xbwrVR)a39L|W1}jEOgdA!t`&a-}Z%yuF)=p`=
zIFlXJKBdrl(`PkZ4p*C42NCK@lieW`dR_9!r-JFZ^ZPRg6YX!2B|t8k1mt3kRH+z`
zuga+k#29t35RMms4~Z^O$CrQ4@|+N!fWOP%BRKVZ9Y9?aNEVgF%qNgmsivqS<ojsD
zMTWY7tw(`?8Z!F~qdr)E<voG%gT9#g@2{W9XSAg0B?8(q9=jJxxr^+VM10wVbn+&O
zdlaJI$-iWXo?5f_XL)$+3Qra*7P;9v<J&MxW@b<&t+dZieNQ#scGj6~Fsw9zpDANH
zr+Tt(bx8FWl+bV*Bxs39R^rgmCbiFxAD;>{Poj(z>GbzI>EDb{%%~fES0tjUwExKR
zZ#K#B+npLI+463{BCkJl&)vZE0%r%R%FP({yQ02Uga6H2^mS%hP-Txg=c#_BK<8hF
z75(<aZt5dN)Qm<|OKGd5q(XyLp73v$)`lc~eMUqnL+wGZt$Jik|GRho!8-1?v3d|_
z2(zyp)|J9DxAQ+gjGO9ODpTSjwnyxl<+I6zY8qM9ZKO)rhBI#`@}&L#(s*V)oHYF-
z6fL;QR|p(ZrvI$pYfd>>;H~wn3C8$MKsaSm9S1x<idO8G7xDtM@e<R?FBqS1BzZ=>
zlami4R6s^>`a?HfIsx0X>6>EY?5|7t#V2D}@q0=b*r%#`CGKvCV@efx@Wec>bm2Oo
z<i9W&%)UZ()QqIJP!S^?t|d`g7cYK^#0Gq=!zQG;!66Fq0b+}_S%+UlXh_7~n!M(H
zc5K~uzec-VEJcSbsP*oi`Y6O3;S4Ui|4Gb6fXA0UHWZRV{b`~C2YH}TmL>h55C`Bu
zCZ__haW#8K$xIEPSXOQ5>HEX%CrtL(bD)2$voUaI>iLLclC5QPrwvLwj%aBEO<(SK
zl;u}91}2;$DLZ(~?5Djb!}&CWlf$npy<`*ML>X*Z+V5bT>8EmDsU={&P(#WQ@)y0X
z&9Q<rGpYOxLYAGzC|_Klao9}q4&&IE(V6@&4ZeCkyYRPy(B;}Vd+vDL4C-^zu{M=Q
z&Wa22Kpn(dp>5a(v7#DWgL6Zk?6Elp)Lin0sV_6wkSILEPUnjGN{#jErcnD7%y)z~
zROIYC6mBuB0EutQlxXe^K6t`>Ee@UKxn{L=4RpT3>X&wKrxR?LP!IWr8a{IOtdiUU
zMN-&B?A*vIX2;7wL=Keq7k*(cUCUAT@>`W$2ZRy$5B&*XDA6EK?SO`sr)Fk9rPnt-
z*_(Fz#%8QFBQYbvXD_q0EB(ZlyOlqzv90ud>9>`yBZgo<q^*qbMpShhnlTBP$kvvX
zzq0-B{iAEe8~hj2$q$e`f=dNa=HFBBbpBI9S*39>Pm1j^Y*$ycRO=e$7|?1ZuWsp5
zbCz7~w3~Irtt#!kTv*@@{O3MS#a~6L+VC^*{N2@1%i!#6NX87MlcVjl#l_KJrts9~
z&b^sI?!V3bO^^1PE<SN={k?tft<|Z#t*1V{&k2sDuLVJ|zD?zGeF~}R>1d+d_)>P@
zeB_iRxH4t@_nA`EJFFaRs=}MFZVRI(&r-kIKRd($v?6&S1K3>#cGXxKQ-j)#QMuKK
zlwU4(zUgpuQc2ure~5~btY%Q(FK*Gee>s3ub!?$VA5!NDQEjpLdWVFPeSYVEa~hHW
zgRqVVJ${T!5XZ@_qusvvBE?CB&bBWOj>U(@uKoL7kLiZzfi8;<!zc$w<EP`Wb@4BR
ztaq`6N<|Z>3WDO{VzMvtAks-bY0&6R3*`}1L?AneS;ift>TLh1g9U#eKVJ%5nqB=5
zGbJhYEMY`-eVi4oI!Rnjx<a@ItRxi=cCzqV$v1|S%r)CI5o}TQA&CJ-d<9Op`26%+
zfd<%E0GX;CXcvl#JsEmswloq%2`~yK5t)sIE77xHVw1~KVA1VK(d!q8F|SGGLj6B0
zrtb61SKPv+{aE_$)2(CKaG^8t!?|oAj#)d^(X;@5C-n3R20+K6ylDRuwdSxH50*zP
zj;#eGa_FN)MBTg+E*mEXyPmk&k9gcL6$L0xKF$PL7{2$peGY|SF8>5u4%@!DLyTp#
zo~-=dC9qk9@{G~8Yh64(QfekKu!Oxv{Dy7nd~8JguAblb_hq-(1PVeye&EhL5HSQn
z^Uxo0mFh#%)+eChE%%xJZexSX|E3xo2HrJz=QWLyl4V_2nNly6s9{i{8k!6s5-Fw|
zZIwrO#O(3(^*sAZdxW$S3|uy7_A{q#<<hH!_Rp@ZVgq=pFOPfmS(^4~T6m&qVUvdn
zCd3)nH9s32#S&S#Un^kv^D#r2KF4@avo>sabQ<oJHd3NUGVYPP3q}O|tT435B0#)m
z3(p9Bcu9KuMQ8r~l#FQm)T^fM$YBEqKHT%bKChj*<`fB0MjGDQf5v|dDG)cIC!1h`
zt<pF+$Uf_K@AFIH-r9^8LF%J{j%kW-QuY*TIYC`6!8x1QBx5}B`=k)jk9x#lNdVNb
zGEDgc`K%xhO{ugiT8Thr#1Zivo!(pj$dUX(>o4#cy+YVd4N{up!uGqwtK~%o@GgH7
z?V9ca>!F`_fHI@%dqwM|X@QPmAfuW0W}g^*0Xf(%@UsVJnukPTheo%erZM_8fA>8h
zOJO|KphPD}sR|4X2j!#R1i9$BYF7qZmJM1E!BB<+@$q(|%wu3MC4e*t1$Q)`tA!Cv
zWI%Tqd{*^l^7uGA`TCbXhysuS$T8E;l`+}iQWSK^rV!kD>&7<)8l~J$VtMu3pL@Ls
zeyk16WU*DxxQM&6(tToI`K0u+RZQ1Yl@K1H#2pP{xkd57hbD=6s2Dg5_@!KG75es?
z?||B?ooVqVlTuVaZFS`2>~9|31pirj<(Dpz9DHcn@D<3HXA_$-;21!)9Cm%>an1AF
zPZo|IP0-#>HDjY3$zGE^>OXZA;!$LspTb7$-t{cWK-;6m!&aan6z!2O^?0#)02_VR
zvH$0%Kx|x##ZwLH2wa?zYCSZMgImri3~z<0fIynGyVgMd?WOP6%6iaO!ku;%OC85&
zxRUeBM?KUN{kQ`b&fY>r<(@W`G|P)?YbHN+Mgp0Q;*@_adtkCHe)WSTsz<*(y85K^
ztHEQWI;E3@BB9hBTCC$>(7y8w>%neSy>s<HFfBW7<C}`5jGc!GX<H?Ch7+sKj3UPg
z$hPjmik^MOVMeCNfdJMjUDGC{>et+~JU?om!+81p{Msek678rff4~gdjZUx7^dvkU
zM)0B2uogb(;*T5kr@DwMIADVuW*VK~vE59;&wtzdN<SLEx-9VC*?)*;1c51=c%kG!
zFY)3g$Dr%}Ky*DpRD$)#eR51K*0cKE5rvkcW~x}Uk{uWOz`%#fl$W`60#jk#MvfxL
z2YTH8oUX5$I_?)9;DfI}xBfD^n);BDJJtI0r_x`^@qWqEwVsDR?K}Jla5LWbX48j`
z!UoB}0}8a5?$csK5NO1Uj+6L409yPxaI*e+qyW*@eO$4{#uWdvj`inTGS&E5pWwki
ze`)c{LrTe>F|^r_876ccUk<&vQ8h6b_689`(hWr-$3X&IFF}1B0SBc!Sudez&KNL;
z&><W4ovjNRS3aI(cy!}_DiKx01v@|IcJX1$cs~lb813%4RS2pVjANu$Q@=Mj2E-@8
zkz|IP9aE&ow6=jvCx^jB^u1mJ*+tVtmH_N>Kte8?o2kU*B=pVtEf?Vg69pWdrTR=D
zWy`Ngtc$UZGoB`PlA8@gS8y9Y5igsZt;bu8!bdM#m6x2s2@WtkU9Ep{Sfw<CbZ#<V
z77k#iOTz@Oy}G4RFP#}$*Q7f0?_kQKZ?6fyglDV+d^bq>B+eCr(X?QJIZMH+=0)o&
zw>kx@dLqJ*Gg2`*>2!mvR3)L!W)a3oHdJ`{g2+khQok>m&g`l>H?B@{JZT00Z>%4E
z?Ad<~a@(6TAFftgs^R6tmzEhGfp1gbwke(n3t@?Q`@<9LPA>*|#)nac-2ZIYD7`sH
zE&dc>RaU>IW+5(Q8BsB=X}MOX#*=Z}C&b{=(U5Hg+l&!(OSc+%Jm_luBxj2D$nD=b
z2@zgP%W&yOpCt7$f&l0%gEvP*si<GDqI<gxc_&`bKk^#f=lR(Q7k5P?$qo(-jwai0
zmc7v#8g~p-WiF`Bn|sKxVQ&{kXF|LDL$~Fr%mt__|K_6E*{s9OmLVt`{gfh0i=4cd
zp7~t^rTjbuKm?ho4MI|I#Bg4J$xE%zUAs4g;$rDwva|x)n#UV&rMT$;%g5;C6Mug&
z(c7{=-;(A#NLf1h8iYAhk@GPxF$AQfb`{!uQc%dhe@G`4gz<rKN|D6?M*spyzzHcd
z7x{$GA_eLDRkWn}(fcww)-mptB325oT{&8G?Gr;_VU718YX}fqa2?tjQYxxL-A(GK
zb~k1aiMfAh`Qt2AARBUaN6#eOE$tGkP#I}buRr3f^lV-#1E`JOM;CjMRX;od(t?Sm
z*MAiM$mnTs9HjcI{SzL(-I=n>cI(xF_(j#DIXE1MqxS6;aAnFv<|5nw!PrIb(iyv+
z1}qnoE&noO+q$#|+HbXMg?CEaVgN>!idW4sdh~%3O=5cV>-v{t;SjA8KeB3{f_Z$|
z<g;2W)k8=3fr8t(;-_s0D>_<RxNC^#!c8k+GBw6~{Pv-kiqQw7_cE(>CV<&CMjQc<
zrqQK?sJTqDLY<}zzUb!RtA9If(z(^|eQ3}LF@7g}GE4m%O(FZ0jxZl&!YGsoA~;ZE
z=$XlFD55OC$Ie>B*U%`17AoFwd5rMc1Y~6rn20)tBR4yyP`4Zsxv<_7BD{(6{*X~D
zKL@z%Ao?yqot`W}JENZ7&!HO4iG)Y~qFWsh8;roHQ^&kl;>V~yTmHrVknpin1M}*|
zLyxx@Gzv+YZ(z2n-u0WnKPG6;<Giq+H}>!-eh8KxUYiBU)pYl=>wB;Hisa<K02p9E
zuGN1=X2gp!%xIF0cm<ciPrG%cu#$(sP#a&s3BDJ@@PXI__uLgHyWHVl-TZAon87Oj
zOm<){0y(R(d<#1GYix{RT0*Z7aenH?c7P^9#3=`w!7O3c9|cq5UN_thrBu6nPxU%C
zLil8B2{Jq%W!`6cz7|&)^#z2!9;6@?)p6OC!s_g1C;ee*XSgIb+i;)hH6c5oL%;=t
zNNB9Rf2Yd0j*mkzzIFb6+IJL2y{U#2rqqn<#G$@p)=U0nO#jP&{QUwpw8V*mXyuWJ
zQ47s&^)4P}ziLxppQ(-Ba*Brh8Y%cz8yh8zYg(JEFT{GDH`1r|UQ$4=R%!8_wc%^(
z9H#bW_p{v-L(@{$w($i?_J@>&zi0PWjwV-+o~-GH)RybiSM!ygbu=rhBf8*bynU)^
zf5Ee#i|BksYF>NOqJPgEj9a$V!DH!Dt~Rzb8)6=Fl^-ttN`2EIzTMnCXciOqu#H7{
zf1}Rt8F))1*<SHXljI;C|Lo<Y6)axk@9fI6Q9~SAEpID<;8)dn<cxhC1F`UEoeu)s
z)`tdlyPZP2^Kr=b_@jp1B5OZ0XA(=L=u?raK|dJ9^3(eoArUWN=a+kl=k91U_<4;e
zbByvdeA*>`PG;&4Y)G&7+4;FNdwUFk4iGSeQ3s<B+GG33O%{@$OsZ++!|(wS`PuX4
zk8d;qfSmlWiGnV~21Qc9+6&3l+wQl}xtwgRA&SNg{eE5k{HkP2_tj;6Y`?pkw849<
z?s}1A83*|1Pl7V2DlBCl-@n7VRJ!z7h`^=Lwn<tG%Ng{Lf|WX1KosCg&e`V_iO9{t
zN=|!A%l9Ilf{SdrJHPvIz7bBj6rsdLy@;e2!-;E8ecytQPX>^2#U%qVTz^w&T}^oK
z^X&LS1pMU%i)k{51XfYf!Re_2KD92LxK}U#DeZe}YOH_WBST^pDQz^0%9KQPu<6&1
zSgz=M;W};0<B`Tao87S#L1QidJRl)kquOJMysYs*&qgElK!N%&BiG3jrUC-6PQn~n
zJZ)sI=Juzp;GvS+=Nb`=>iG}bXP!m(6$4N9vhlCT<Ih`Mpj&7fe|>fybmySc*JZf+
z*(sOZ*XeZ;5BDycl_dQRa|E;*Vsr4C;G0fd7h!v~^wrVIL51#XRqP{}50XvNAQ$OF
z{KNkN;xIBxB#R0K@)cdYdO1$5H_>rQeC5{A3q)&*gf6b<WGiCLX#Y{bKuY99XXioQ
zJZJkUh2O&CFstVB1T&JUdv<)QC*mlKS%G(JbvYV(>*m9l_sbcHa_k!LE_pdiEe#mC
zHA|!ysj#M$ow~|p9q?(mWi9kQVFOqvyWvm87n>kdvkr8C4$2b+D{@i|0A<2ehcLM6
zpjw@DvKFKLzb7caxboRqKDd<|C!3;}YJtYJb>}8f06a3T_r<=&#w`JZhbuY4cJ8Ap
z+i$>Tr=Y}{b9l%+Y~1vsm>KT}nfDwgx9oeP%X;*XVH(2p0xJVolU|A>8-4(H<T1Vj
z>L;cdCINT=vI#O1jT6HLamwCEgrG`2?>0<MT&poa=XPt?SWC5#&(i7*MaEL%p76@a
zg5bc-EwVq{_pO!1QU3n%5FHN!fG<hs?8}>|;1sTgx!LBq%5sxR>kfp|p61j3->?7i
zmv5Syuio74d#zSj5*F3-eVH7`=q;uoD-x}!hq0gb4-(9G2Ik{tBgGja2wfMvaogxO
z{BM^hJh3JoSnBgWj4bNVG0Q>Ro85vE$gff}S_}6M5`ry4ZKJBTQA2sQIKtw*wJJ9~
zHm6PcEgo%j#_R7I8=h^6M1Nvh3s1hPli!X6i-)1rm_+P7-!LQhdSvqH6WB}mkSNVQ
z5}^Y74d4T8W;yO7<ASm;z;8{wk-+d>*jzjtdc2a7GjE+4CF^?I2j$@lrbtirbLVti
zEN0#{Fp>5fF!&S^SokLkgw+k^0doOH)iWMJ%=0ps((1@xf%tf|0EN>OP`un7;EWES
z)v<>M-*Ap9gGLcH0u;#!fR(>iUj&JqnMGb$qF(Ykp!OMcZ!Q!Ask9&xF;)$9bmhNM
zHbs7?S?m1oXK~}fC7*8O^kL{IK|-~(eD4In?9k9PEaZ%77QYllqf(P25cO~A(q^&k
z6q(DScQ2Xqm}GhH!}R6gyqGLC_(wvj<S1B9a9obicl`RNRNyt_>sK;s`e&-|XQT#;
zuj^!~1zA^$>IHMz`RhM+if1MLBcZ`U8PeKA3!6r4a%mYvGUmM#&AYP}I<{s5xX~kx
z9UAtP3LgbACT+T~wT4}l&1Dr%d_=iokZmthD86vs`^QaH5m34(NBZpRPI^hjpX#|B
zDGznc+lYStS{w)Y<!tkY@v>YLcgmARIIhMtNj`<IQrr@Uifgt(*Q}his7@paVIBw}
zMTaPTxwLEQ-_&ury!E>D%ukVOx7k<8Z~d@+dm~R$H09kfcZM(&a#Dx9=s(=ASqPr{
z*Z8z>L@7*rms09(-@p0%ki*Ap@>%s3>Ixss=F%K=Yq%g>nH0z1+LoGA?#fD)`C07O
z8J0nIc`x!O=hwZj6!SOTuFjy9mXwwuavI3FxBKy)o2vr&f3Sf^?-oKlEb){xXt2k)
z-s^dpf?u~SV21W1>Ju#K^ESN9n%DV>l&r*|RpgN!%!NYwg3YnyxCZkSd3g%~dv^P1
zmtAC$6JIaNjdHCiEAkksJ4uXdQw37{8O7mMX~Vu=!+coo;voI(Ut#j2fCzgR-b;!D
zUSuMb`NM=<JW!vAvRf`0y(kOoVH>J(z8+W+Rm8q-6o1r~z{rT4P5cM3LG^{D)VX$g
zB7-LC_(EBfzyOmDQa+#>8!|kAtW(+h6A+`r8TKK^PW>Y*4#O^i-L9r8eG}saF=YGd
zy5Yc4=|ZEF@f3gil|UsTMYac6)T~zWQ5>8)9hWP10p8Fr)$aAuUndljl9F7(HUi85
za##e4;9C<_D2Nu?j0vD6=?el(w`urvNIwLNyvA=NoZk$8f}Oj-7-9!=0Yl_Grt1>`
zo8D<ndTp#4Mz@qkk$CR;ob`B0U5^na%?{$M(*$hq;nVS>cHQAhRnqC5`$1|*biwwU
z)?wFzF2lI8FT2<VGbDptVKm3qKFG8~LaDl(8o!=XibhtxmYxtyIw6%M+lq0$Nl3Qf
z_0an9V`|@GyPwRnG3Dan_lkmtWB2Ht#(Vh)DiM$C$jCXHs<UWq#<rBbbY#pkEHG?z
zz`9AiN#{3_PuwX6%IV^2m1oBbXq`^(@w<j|=O)k9EwWGMI*;iP%IpWY6X&JDC_fu!
z>+L{6-IsLd^n+^lH~X)TduDjEtnPwkj?D5}Ai7=Alr_9Iwjf5oLF-Czioj5Xe)~#O
zc~{M&8gA1PJ~HvWH}YNczju1MgZZpUUyh3J`RB6?@M5Z;&8C+&WzDx3?;MyW8m=td
zKU8LX*j85=`|M`ld*dUy;abCQZN^a*qbHC-eI0x3c1JZ7C?Hws>7&M|Wll68Lu!qJ
zY3YeEX)I$uMsiNG^Q<hV_WR}3S~_`_g8NqnT=I<UOg^Ng@ZIh9LN^`3TPw${yP6*D
zW2<LK7`--9aY14_mL#@g4oB$4SkaA)-{cXo%P)wJ^4Df_av^>MnRDD`1WarH_)5Bf
zXlNpbFhkO7KGw!l*=}5TJbpWfuTa+>114m%%?+Jrv_49oW2p1Wor^ttS8!3sF%8z&
ztes$^$wx2H{qtfC0!HPElEaRB?5!Rei*RP>lL4*L(+c(=X+RC&JOw=G`QhszMf_}i
zwgr_xGp3)r-2h_P<;Mt~IsFVhC{uw5!J#{i@q`B*vUl0r?akd{VRn(=IHIM*thbVL
z(jH@+fx2ka1CnLun@tMv;c`_y)7;^zsLk;2nxY^g`=DVwL4$EJ4ESI^xxfbVj4r^D
z8U#)+`ugn}haf#EWU4@3s8a#GQBqQZc@Zp)ZbSeFsf657yMoOCZ5O`x#4Q?Yy~%#n
zdPMqQd*}iX0+7=`GZZa3NGnXbsC)fhrt=LKAHzKEc8g*Xu~=eoU(#DcRMfE$;$l81
zTbVA_zic2SgCc;!j)kkt8e>qn__d)L|Bs5HQ5`dq$Gu@!66p(;%nkzzY_)K<>GCS=
zUn|b#0iw2uJevxX=viL(n*sFUv0AM;F0@*d6W)KMo`(2M$b}D9v27kXBcJ~-ivLbA
ziNkoeGsn_TT?NOB?rn0$KejDMU)Qy(QcSYM;*2>>ZvVV5$YP}aO08Xk078G8cql7<
zc>UVB-Fssx@@IRnT&QB|N~k(dCNSGy-22NK`TX_L+~>Z55aJbp;iVng=r!#Wz9j?Z
zxq5ob_h>?ppbp#5fs*hROm`UC7m!PlxfwCB8_zRF_Am6v3i!7r@p<tPJaPump2l$^
zGfaFwX6<;0f88@8LnIIDXSM+&nwrLdIaESbEmE8TMamwD_*M)Z?d@;M(&AEnXXbp_
ze}(`EMdlE@@!TKG#d1~Yw|56pJp<c>&2?gIA-N(wW29~W=z@~0=!}QXM@NbRvg&VK
zQcfojCekd{ttRu!)!R88#>>D)k55mIhU&mEDTwb%${R7};-tJZKni@^3X-RW>#9Zx
zEl3t(X!u`X$>R|{T>(L~)CsR4H^deqpZu(m4FETMKBQER43rJeJb+|)z>joK@nm?a
zLQPa$qNe%Ekm4^#UO*62q-|+0b;cVC7%QNiuIn0LZ2zlrqLEw>frKldGJzqMz|Ng(
zEVosE?x$|0;bIm=r~5q&r^iaJ9^|VO)k=s4EhdSqUC7)zU*N_~)re^qHde)`Mm$EF
z4gWBihm5o@Sy2YkNIL&Qm|rSQi2F4QqNcqJGS#9pZS^ALl0X+P5D}MW@L!}vm(DNF
zKe>V%t%S?-wpls{qOJ%Bi*m;GB4Ld*V_=)X<<)>8?#|=islZO^{a<?$(`adeeA6~J
zZs&MuvdQMUUKk@Awq@?rF#?OJW@?KS;)Mu5OWVmr#vi$YQH8Q9M^8<sM7_?}u!I^<
z2f@4Jt`f$WQ&k%f#0wGDjDXt&UVvh!pOpeuDUpz`8>&C+#fs<pJ4jga0jjU82I1*s
zGmzbp`;<`;J*yCxBa&d`?Qo$pWffYCB<eE&4UUu@iS(*?8&e>V1rxNDnof-y{-sl{
z*EYV{m`)7>3xOvj1`;HnUgN%4#ES=ldr;s6$ydm5vIA`69f0VXs{rAuVCvx)_cjnT
zw2gA8I@`W&{6RJQjm=dM#&E4zAT`>rx=R}*5{nL{AcC3clU-H9(1Nf5mGqwP7JC1j
z`ZE{TMb^;V>(*k&S>H+@m5pXor~VzW#t3+M+AZe|5ChMJ2-bppklgUUegIJnVWfUS
zG6r6{jeyyD8|dPOyWPmEand>XoY^?bcF+ykaF$=nqDt!lwa_wrpr@M1bF|R(o$6WB
zqe3#O_vPPVsdRG2s|+^jOINs}96*KWI2xo%f|7}-Yp{6GEt8Wu`f;l|b=BCrHkEb7
zKvHW~W2>qwZftX@TCO3ZLs$7{393S>ax7xc=-hk)Px%N}d`Q&Mi8gRCS4BKi80b&5
z6V+hcSzR^kOV&Uu=UD=2D*J49d}=&dRauUbNi3HPE@SJn9p9;Nf^PE>Al0|?R$SJd
zJ;^#vW9}-38;*O<@v_+iwPJ(T!m|a4uop6N{nBq`l5AWvXggIR@WCqw-_h-{1Akpf
zSLHxwc(|pD;!R9OS>X?&IvGHqTAY=i8A=F^#FUo+FzLdC+e_qCv~Ha8fOx3zCZy1{
z1}(oA4GORoBgYn7xv<z7d!qxIow&p=Y~svTC7j6yLE#@MI0AamzeJ;?vqQd5pwD)9
zypdD(8TCX#+LHK3N&%8gGLr;EfY={b6*)Xmf#G>WvdE-2w$|gn_Dx#G1=&kFxZX3w
zn;zZLCuuLD$TIhRd@e_1YPYt7%GYy?Payf}3a&tr?0RAI1<IBe2xF-|=6hMntZQeb
zdX!%~2t(v_ybN3Q?xX+k5k;ZAi-8Hnr8ou6`P36k4%+3<KM=DY_=kZAMTRj-A+`@u
z7%iam(-n@Epnz<qI?Rf$N(>_bnB;uwR_7&4%+)W`Q20h}f75#5_<O2LE4c7!Sa?KR
zW72&L_SsL>3M_9fKbo~rSkW=x9d_Nz^T`GwbrlAv0z-<*qtg7CIlQ^bd)`QhLTR4q
zqy55GG3)L4^WzGh#H~{I10_E!H+j2=b^^RT8UB6UUPNpbG8zAKSA(qFNSH4kZdxo1
zC_Wsw*qYg|3Ob6^Q1RG(FzU)MH1$gnt{#swFjqISluNRmYgbQdZ%Io)-VLVZ6R0Ta
ze64bBOhz6t@^TjSoj;cS;)gdz92N&tKYlRNEIW>&9K1Hu{#(u&EN!)nz}Xl{jnT5E
zIk-n$>}F9E6huy**>zXX4;OIBYaI}GRB7Xmd|MHOu(1~JdNd~%IpAUPEUQ$z<g7fN
z2aRC0%*C9JeQ{b9P@Z&JRw+5?{S!?1ZvNadz6C|0gGlH$dy&Uh!`P4KKuWK)#&$2a
zYS}M9gq4?<oW41r^LA;uMC9fG*R=;_`4K`*w%R<97JI;_V$|pzNdHrFWsqvJJF<W1
zHCcyUq-JH8WpsH!b4;Y@s!oafg3(VX)lllaeZppefWvb(Lo(x=ci5kKjx!~FQc4PG
z*S;rI5|B#{PSn5cP!Tafl&4_jEqqMfUw)YhDJc&gQ&}kY)$d{$@_QbkPXS$_+i(n5
z02_NNFb<GpVV%@cl2Kdh*!!lnnO431?|dJylUEQ?vpLQ&A)n>7mG3KOZK=D9_Q}!L
z{meF1o7w;5D5>-8d4BO<y%Xby_JJ*4&@!5E*VwZg>CKkRdE<uujhi-WqPQimrttG0
zB&^~!n{95kvE<7hX(Y>^YwwOauK!>Dw(svjG@pT+_Nohu<sTu7TxE047_IE$5vlsc
z(nr%LVYSHnPhpZJDg8py^u2?1vGkhib_5@ToyiH1tsz*fha$P(A{W^+A_$QR_;N30
zLE?sabL!7WZ3`Zc)rL*IKc|xmlx>X~n4)6M=wFwm=cc&2vyU3rMS6qe_;9%J=~;rY
z#2$`6@guH1uIlLvXVg8ceeZTPMpiMb3NP?3rQl&GI2Twuo5d?%O$QHuiaVnALE+Xz
zuw7hzewMKL?)UK;cumjF4P6D^h;q(Aznx_zl8@?nV7bGEbRf9nrXN%zG335QrQc1T
z1n?_lPvB+8BH51(q7B3J7yxANj3MQyZ~Qg)f0lzPYLrx6*tu0x+#^Mm9#VKJ?s`WC
zrRF!&{TdW?ii2+Cq>+C-KQHY@gdlydqa%-3;sRsi=rO3HrcZp&)`M7fS7#8MTPPl_
zw)@JlGz*B}1eU)2bsO_i;LFAh>u*Ydbp+mJ6x_9&Lee|>f@DkAvS!65ciQz=r{{NU
zJI0&00qUwGxi6>YW8Bj54YT%BsHGZgjm|$s{P`xr`>*NP`+#tFxWRp~CM;P{K={q<
z53M7Adj(Yen6#IayzK>T$c4$CmymVOGD+OIK$fSJzF+N}tCm%6-=49E;vLUP$o~sn
z`TS3)jbA4T`F{`pae)f$leyUsx;}<`!4<0g7OV{t`&7>*Yi}I+>V*%K(^hRAOb}cA
z+^s5U`wHSN?AUS3v7C%>x&6>+96KX*)qJ12FUb;!!<eqW4^2;o1jtJBrV8}pa8)s5
z-Kwbnc<Et;86by&|HAq68^QdIV`Kf3+y~B2%P(^oo~Py1D}SJ_i94Z8tKbAsCcInR
z*KoubT$R5uCvjevDz0xD!oZRLf?iDx#<XbvVttD{{<GU6o0lN@hV6H&ZhQo<sx%2d
z2D2s%mce%K0NBPmLOP|L=q8%jgIFoV(-DDM87bgr|Jc$)m8+sDgzeut3=dAVxgO>c
zya(#|IX)eQB1#M~GAE3ukG8$QI#tC$*N;+SD<OG=B0XCp@|AKjU-h5QbiO<(=15Gz
z;83;xJ;PYimRROKT<e2@_NBBs{`2MEf;MIp0c@o(@u6Be*5ilUm?U2q!F{MI7B;);
zR&^^sX8*sz>V4Ck!5P%P-zMrd#<H~Iex$PKT=+|`S0b>miZxzaEb!Ka_*nj7FJiR%
z8dmb2g3~LX{76;Fv9XG->|6>h7Af(y!0YAtB{OD<gS&EFmO(3&#L+7z`@2?g4geKu
z_3E;+G#qg_Y+KYX#a0>Tj*W`WxC+qEH8h_0oBc&3#$6^gfPH=20Y%P<%G8bxFqixH
z+yLtS0;~RyI=K$*HcwGIn#2SM<9~~2ErYeO)BPZ6n0$rP7=zrOG5G__tZsCta>s{g
zg-yW%HBd0k590W&q&;X&fg}mS4TGpVlChn#`kfqHi->e+s$Q4=@M`6k$61sQ&Y15|
zOQB@j4Nv8m)KOzomPjWddhE`h8XE+{Y9e6XXS@wlZ0P#yG>>rfDayBbWw@h-_9a~^
zvqe1C>(Ze_+WUE}9rOs$Zw50m_oHLpOX0SmD8h+@S^*x-(XBoPa=w`&qSQZpRjav0
z?)S=Go#Iwo8il%8cfZ`aP{byOSy3th%F*qn9xp!A&C>4zO4%A%#MdwqK4vcSSHpJ!
z?6V*h)ZXEAe8!*y%`Jt^ouYgn0)Xx`EokE$oY9tlfh<dS?<w2eWFPaK$VxVUpg0zJ
z=l;1&vo7Yn>5!arUo1U#RW>__%O;2{BpMcCGXx&kn^)?9$Cyqr0VIHBz4^>MHl+h7
zy1PQ6Kw}3(9-N*Y#|a>lVuR{FE`8@K$mQcgMca)2HrwAg%w4{61!AL(or+T4u#lyN
zsmK13;Fqv@@O7dAdY`f70c?qH#BH{Oi#h<)_&6+-rZKR-Srk+6G(B8SNtyIYFFO_K
z)}54iop!7EqCl+bo~0gAxb7>zz%BS%R@uPi2+XKhd2Y%tLnpbK>4>4FlAl^Ay<gvU
z1A6;<U{YWG>@fD_Wu6YF0D~GC`OUw>927t5fy;r{q=wC-9&lViK|ztxY!RE8Qq%3Q
z3~`60z@^6Dy}!kjlENrxtX$w8gbmz742J*y=FAMbx;A+FmL<8dPK*q+&K(~fA1<%7
zoP1jizR)jiJrzD*!DRFyLfh8O5?^W88kABJJ^6KEWaZ2K22tjHtull=<ES%KEZ3S3
zU<~!L#`IMW<!Z;wa{XrQfhb1KH|=|DeaIOx9E~nVFM_qR9OMx~G%+GHRQo<DQ8%6^
zfk_s9_I36)K<8?qYO@-U$WE6ZEX13~8RNqqhvFn#Q7oO3I7F|;<DMEmj-E)3B*1C0
zY!u;BSp0&0APt*v{%=M6pN0eN7z<}j<tWrm0MwS?EgAo!=j>IICuNDscDUFQeSq_@
zXhcz`>iF^!P(OXVm1g3vKIFzm!WU*KOAJpZ%}ll<guc2$+YK6&HuYBb|CSu43`hWs
zFfmI=E++mbRkD?8<87q|Pgm*fY2A9c!O`~wx!P})JfcTe?^mX{0(^2t4FLuasp>!6
zSQSRk8NOU>?DCG!qlR9+O>c0qr7x033Bm~x$+hkl73g!gwOWGusqPkCvPdDUf+aWJ
z8KVxRFlixaRu!u=CYxD~t-3X;uCXKL>8}G8)j~u5YC_+3elg1uZ!7Eb)_=Gi{zHZQ
zzM#V(<8X#;0|dR*RZLZMjToP$Uk-(sPkS40v&XELNKxS&>%SJ6fYS?CzO6^pVlMok
zo%WBHJl=SGy->RQe!u*rB1@#}OmeVbRKtiz$}A6vMusw-&yXk-Z@}+u1<R7eLr0U`
z&ec8hdYaeQfe9#}5P6VG^4%Uw--84kxcy1|GzJ|$c<)qZ{?42q14a2F`rjUzu*)Wx
z|1%Hvh$x!ZJ3!&~B1IL~IaOEj3aox3o@i{UXsw~Rnn`&JRb(0X*&jl*3Lwt~+sz$T
z2wpnj_{G~y<<9Y!X#qqxpW^^_|KjR-Y7qBz$=~obY}+ep_7BWEy;^wqu%duk3zFSO
z7Kk)fAvWo_&#=-9i@<o}DTb1q!MCQgB#9Z5T;o1Ui}4$Z&1kaG(NoJ}(j{Mkx21;X
z$=+)}KW#TQBN08I?%LzcgRp`5<e`8t(Y76BN&FrU9Ii&mNB0PJ!FK?R5=69*l8pe?
zsHvR>@CN4S;!vFj9~BYcQ$X}0nAT~9WcpK4fC9roNHu|MQ9zTn#B#C0aIadrkTDX$
zq0-e5B8#Y>A7p@vzKhH^tt${%s#^2rVKeu|rTeL|EFCCAXpJF5m}|t|EihRD^xVxB
z&F-sbwr>CVXSI6Suc+IoTK!S($!b*u(Bk3`h1qBsJWVc-ZMPCbbqTcGkY}X%<<;LV
zC{DEp$)~suQ~#kL)f@gWYGlnYp}BnS-m|S6fF_=N6EPC{`j7~g07P2^w0#eEqh@14
zXtKLSJ9o|~yG3qsmS?lnwf8Mo6I>P}1naF6^%?~>T(Ns4H5Nllf=byf=}2#70}G+l
zKH1{uL&(<($WROE)<tKYhV&3H?zxY%*{9MP=VSZF0Yy_hVO=<?vQ0|7KfU^c3y2<|
z=GbK{@bzo7HV<DgvL+qW%h9JBclI%E1m}kbcYL%*7GUh<S4RzIS#+-z+XLwgQU|$R
zDTcz3Av(nhnSm|I82F3>eEA>gQ0lpGy%y6t8Sd4Fc*-*fjf7a!sxi?&-5uIY|3l7~
zt7kxdK6OCE{B_fnTirXe=nZZ3#+>4IAJ<S_`ZCiZ?vbl~lGPu43T{f(1iM0|JM7kR
z0h^^*^Dd>#a>0QBu-kiC1y~X0Vwk}sm)fw+y*dU*y#_fR$SVH#3ys7@$a<7L!8Mw&
zibpul4f8!J`P4V{^eQeSH(-8pCfL7V5ywJ5c|-zaYe7T7cmdfHON%ZDNn0G3rvBs}
z1~=J9E~cPcC>`-vQWB__O}GA7eA5A=QTy&%4^5X`7;ZXXIZFENYR2>zExytU8pA=f
zR`tO*j=(2zvc)2{*zLfB6_S^ZokFRcijl9wxTXy#jMD7guKH2N8_O&^g&l9hpA_0r
zI_rGw^F6N1pcF`1tAW<V_h{=o;}r&65m4BEZfwd<hD)3@mOXMQf0U%YO)Q`>Jtnzt
z&~v-KfzWPhs72e`I`SXC1{$wx>%FY+d;ZGo4e{3R4t{J4mrFH9=_`LHMbSxssC4cc
zT$E`|VzzNAlty4jv|zv<y8($f=It%D>5NKeen7wVe6b<1+Pb~t{&g@0>*qGovjmB~
zv?-x>?YYRA?9x^I@jr`~;N)<@Jjk1xb@YKCjTtuqUN?_0=gO#t&R5Dj5v<0t=pgJ(
z&su;GKLgdmX*A&uo2^jk%Df^hQQe<|5WvcK77=sQb)J2D#3>P=s?X{w&kR;f#yFjn
z)90oi7Tt5vbRnet{Bo<lRskNm2ZoI1tgu-&?BX$u&1Ml=RtyiLV8>Tez>_&BUYp>{
zlGMuscmV9HBS7HelRP5c>L-iZj6_%Wtheb)tb>af$b6sqKt7f?PzX~Nnl@`j0}xZt
z3pt`)vn8$COjrBC^O5O#y$!bX0J5J^0z;an8HCyZ!3;_xFj;8R4id8XYD*j+{K!>}
z1g*z_p*w|kGB*smBpfKg=px>+$AGUjW25X)27^@N2!>(IkEsGC{s!)@GZaTLONN{J
z9BIBQOqPunTEP(E2M(b@f%Z;$oXW$Z6k~s2C$PC2X%M0^uaTcE4m}s#2@aw0<wld(
zDEG_?%{sa6Wfz?N_!9NPAh(q)Rjqfgx^%Hp4%ziB`-{InzHOb*eG*t5GV&5K-xiwF
ztrwmuZG)`F)fWhuXnQcSOyt&tDW}?qY@Wps_E^jno{T20Wlj1$Psv(b>a2LZYizeJ
zHoe|~v?$_mV#CvR68k()OzO*k=Z;I8_dv$L^`EZFP%>It>-Xd{NYwF8YRvQzZF@eg
zes~6(?XdA)e^z|BW9PVN5ybv10P@}U`?bktXKZ-3bGE0Li?wxd)jx@>o;>4GP}V;W
z8!vUqhJe(4&XU=72<!g(NER<{UQ{Pp8vkG<5Z${q@5fggPm*~9dqDAyNZO~GounMX
zJspkRv1}9TRAI&nRwc-+(p`s=eew;q;D|?g@h|#S9-fx&<_8?|_&NIi9V6%LH01_I
z7|e=&DCNDGx(30J7d{^fG{m`P3}X_^*coR*2&4khjmrJ#TbbD91->7{1)r5keUor{
zT);|yk&15ac7^sj8KMEuU2K;#7|M@pv%A`o!obwBX<+jg^|ZrVUN!Z(5U4v?tg%T^
zoSi>^XOLRCB%AatL%&-D5b2M?wf56}TJM`&EH8N|%dOo-5;g{U9ZWSm=;d97&+jSC
zJyQW2onXrOE`_o9T@u+S<=+@^^pTkQm;g3>KhV$oSEN$Hz1~A-`N-PxzE&7gq3;W&
zHE$zi)e<ASZ3X!BMEz}pD@lM*7lY9<`Ej%^ck`>OUK+|O(Y-sm`~+0vePp+UN2Y0}
zP8}eF0><aUwXKqLBkt052N8+MWw7{!9tUzA^;2=D;Q5^+#VNuk$6hlH8S+YqPM8+~
zg}rLs@=lx?d7fQ^ig7ys8mMC%fLi0wj^&?r;o>3(c+@d)d5PhZk!Kk3_Lo`o(d++I
z^qes5sYJ)^-SX@`mT#9pm?AAaWVBy22FWof7&G!}*D0r%7PnHE6G6Y!gu0eS_bk-u
z&zV*_+V>JL%JxukPqF)l$oIl3Lf#M)pkX4+5(26ypGrT*NLq3mlGm2zFTXwN&UzM6
zvBo7a{1#D@X76t~-hoQB4Q0J~jFeCxmeuDsl!I3>f0xR8%RBy?(zy+Fp91%^Lnunn
z*l#I<me?gur6l@|O-3?=a6saLX|RpX@S}@rn_H@C$|$Cf&MqxYeXhyyn<<Oz(hVW!
z1psEjr&Wt78!~NkOJ`K+^_R1+`xgrx2C^pn&$Y0>j2E1JA!YC}{yZX>CO_yrS;?CV
z06)$TBwNnJ1^=Ohy`La8tp#{2UDU>4Qq$BOuPD1+W#MS!x(pM-J7&1a#kR@*mp&O(
zfU1q9%+P*(E`>|36`YR3oU+UnTNYHWc%4FF6OYt7Kh0-2r<O;5Pn-*;98~D9!#n?U
z;bNxsIat?qlM<wCQNB&3|Lmy8eTX}@r%mTKnQOw!k=qc}wXZPhLp<l`EBSAudagLC
ziXwLF>G*eF27jHxvxATB9j-_9T1UX~Z+Fj`Sq6tnIx=$&*Xb;ZMs2pFz~mUqMGWGZ
zTb*y;vfQRS9m?ix$a?-svJS1c!=Iql;FwT15|}Xf`Nz~KKn#|vY5b<B^3GD%Tyod0
z^^ZxQ%9bHH^FJ3!+utXl)Es23xiPAz#w0EuKcFur`6XRDCjMc|OU4zb@4ff1XFClw
zvJk#YR^meoITnzC5U&n1EjM~hLfKzRs8~1lZ>hyQIE&A@Z%{Iamqkv0U3s;*G`b>G
z!y}_M+ujq=p2ceBpp+`x!TYMs+j^wY>PrJh+Y67jRmR7#QIeU@LGN%z(eAL`eZAxR
zHP|Gr-}$VMZNUq5DrNNZ&9AW>Ja@KZ|9mg%UB^X|4>~KNl_YsFg9*y770C6I##Ez?
zhOP39`nNe23dglC;DaG2T{E#gBF<8k`$1X@6!U1f$cVoSbTnmy;vl~jCIXEn+sSei
zrleARq?dDG(idtF4@`LYA^OWp8I8nx{pqwKNOtUtT8ie7Rd^6Q2>e@Cslj4mPDASf
z4TrE*r1BfG+U*dEKlk&%@AFR|<nGAIS!Z$Oe~@V!gpBlNYQ{T!`9X||@-!+wTC~wi
zYs2?beT>$oFfus|mH-)^>mLl_BqIOKN(gOzxasTpp$>GjPdD$q<n8Iq#ZOX`fWEJ;
z(bS?g>&V)xObI583*pp6|NR3pE-@HwDS(o5tBF#(D`BsJazDT}G4UOQforfy?yG|-
zZV*xsl8xBp)r@atY5gl4L#G0vqGZfl1B0uOn5Q<Bevn%;rP3Rm`Wm>SYP$NG?3(O2
zwXAiK#5tURv3MseIXiWF%tEU<BNFrIfu&3}oo)^RFDUt=Q2botkX#I3Z1OpExLuT0
zrrO4#&!vjDkW;oz;Odq9NPzaL|5d~<x1Lp=BXY=DA6{O|QPajRpFT)cb!i~0MrLUk
zF4ZjmV9VqZOUVZ|WAiB^_Lj31>%59YX3;P6VSP`;ga>`!l3i}$tq4-Vo%Gj27&lsM
z&t82d$syymlrHrbSKC}F4fQ23wR&1h5E3ccQm4N@ck^gH<m05tyuhv<X~TE3r!`GE
z!BZR{)H=iSqYbZv?Kv|IDOf0<W<L(nt^MMYc)d}HrnqSBBZvCT&`Sr521bjFMv`tN
z-61qUwq{!p3%(dmnWrXFQRrJl)U;u)EUWk|<7X1r6*%fobC)vu-p+Q}Yp&@sa%*1k
zetsf>mY^iSyv<x=;(IiE(j4#fg)aBx%_z~t@1Dy4SyqQ@zQg?;q`gfcb8ew>f8S;0
zu%n_@>#t4*=X3e(=24x?%V|J*u&Lu*nwrmv&M7}ijPJL<u-_mc?6k{wg7L9)+fqf3
z0<i+kU263Dn#mK*bd#rrrc%R6Y50Zot+P%Kgqqe)g9wXs3qOvhGQaFOeHxl}2c}VL
zDOo`wSXN<jh8@i{)1~e&edX{xMOWIZ%v|{$)Z*&ryk1j<WbAY&pZ@TBH5(3H=9gz#
z$LqPQ_cVwB@x|hKaU0cW$~bn$ArA^%`NJbbI1DzW@91&*52y^@_H3GHzfn?)x4SX(
z68le}(p-U7k>Wxn9AhWblh*%l>T6F^hkwvafA=)?d1`bQZoAY_nkVvOW+P>68q8iO
zO{X))?A)cKR4q1GmH)A|n4Qd?w?>KHbVFTaCtqrH;W;FG;^P}?+4s7Qqnas5Vda@Y
z{@TwfiiGVdh4jfds%qb=&9GL5iO+$^_bRA}$Elu{AdDu!Ui89bB(ucy(-U35YqIeU
zC{9G*5gG+XI~xY&8QI0lZgu((Uw9+Pr|a=4_(_G`kkI^GxR+-<?9Say*lX-r?fcOR
z9K;9$ke);)6ban{DLPp@<-o<Dp_mD(<egz_XQQnqSXQ}RzXU`RkKo{(V^gD3Yf4)t
z*JEShXk5f$-5Pi){wHt=NYaZ|Ca>ciL`1*4JQ0N7iEgmT?5m1t?2~XWI?*K=MY&cN
z0t!~mB2`QEk;MlX!|w?tzdw6wmWP8vJW?v(_<zFI)z*#}%tmD|m=n=8@DXXL+-|d}
ztkC#}m_YY8>^DynP(*YlKzY%0TKu%}5Hq5EwCC(^C*7ugqD?H)i`tS6^&g~&ADYqB
zgG>wSMsKtb-V?}ORo+bgbMhr;#}DL0PpygwvS-mUj$Nu%p1^G9gbZz|M4EPO!lD!8
z+6+y3o+joj7OzBsb8A4&6UP7aJG4KDZW;3^Hst1MXtz|vVD72GrvIL}Ah#jQ(n@OG
z`TUpP^KmAB^N$5b8_If54AvZcVsO)3#eR3sTY2ApMe~;piMbkW$vP@M*mLD_=a+Jw
zpG+N}1{k6qERII?=gu(s{rRcW^(WT$Ip=-ndScSLD}t!k9hi*&aV+>r%Cf7_DHxr<
z68Aq%2pM-X2nqdYSiIq_{1H?eVo*5DIWqTC+~Mp)?fz;P<U!&XZO(|&Cwenz?l=C^
zcq_d@s>AOAg|<C~keUBFNI`nhyp@vk8~W6Y0uQ2Fz+7VqVgKK@C%GQlM`S~*`C3*Y
zxG5KxP;U4?ClfId8vat#R)hA~DM2DbV!HQAqmM9DQAV=AJmLRh0n9P14Mt-Sz3f-y
zUzy}vK&%k-;oLoo8$|x4l_2$h<uC*Lh;BJ@OFhgR6|o|6T;x`({}oanEHr_kCb0zy
zSb~@lrumcPA>{o@cr2H>grVJ`7gHR=x8<ett+wh1)WT)nv9<}@2RN>kFV8QfP9>-r
zo^~VO|9CpybKHveo9k(jH(LlA-)vcn89WK@#1oBkyU<Kg|ECScy|Ar#;atHH;r_9-
zS-i5o_-z~xT#+j^^u^N$K}b0}fLwc_Wf*1iHA7`%@vAw}=5hhh>uu#L#b*ht?p=?C
z`-Vt-)86Wz7@vsfMuui@IqNwNL|=a;oPW|wVu>dSzK|r|eiD{=ZnTd^BdO-5eowP0
zaD}JZ5nD!}MX?E&e9^+GsGLsIC5DHeC?LOFM!-q)L?Sf_YgcOINo!?Y-s)e**pB~f
zwHbIiAai~La*|W#7V#XPYv0aI!~D0CCFaTsjppWcDoncmX$hxXU{NsUn7aHBeIkpY
zIFz%flL*a!?5W#b%ik1iZn(&cI=h~1qr8X@=nN5zNX<28Cj2UvZ{oi6q6YHVR$IV-
zlwOC>bt1G|#kckjYPZ%kC6=+s(;}7<vwmIA5lU~SnR7wDr>pv0!H<j(+u@n+cZp|c
z!^(H%Sa{27+(K^oA6y68upWgK{4Y0nX#zim!h`%=fdXTG>(YHuGmy4cg5t6rkzTt|
z?vw(1iLl)|^4x6MKR=x4aOui5rF?|R@|Em)AYP;?YDV%@{%4XN<HKL#q>1ko-s)d(
zx}7}mrZ$j7*<-Eln|T)^fZbz|zWrI@AZ<-Ov1UI0cDUN2yyA8SveAg^LhsVbKi-i2
zdAHy%xPPA!hF7VrOWUBkgcS3zj^=*!{GW;>i#_r4cYgLSIWlJ_QOL$4X<{5|Dl3as
zPu>tyFp4Ca*SZA=H0r){DdzY;Onrq{TVK#^a4AlUQzW=cad&qsMO&<Rao3jO9$Xu|
zxVsg1cXua9aC`ZE?|W;#KOlGAn{&>dnZ5VSVU5nf0c}pwZ>XxXPB&7<;jQ0j(4Tdd
ze{l`!{_ChyU><YtmB!2w9MrkJ{jMY-Y-jl{S0QsFCqPVttn+g8LUYJj-@h6ue^v8e
zxC6zAk;8X8U6WUwK=Ks!5GbEyhUS+S0z>s0nPIoznrW*$EiH6BMI|z0_SrlpNz+Uz
z2i}(<zyIzL7^t&y`BwcsGAQxPNzX8ti+WtLdCosNzIMW&!|qhZyef;NFu(6OgqMQ0
zZ#H(WMqbr$&CO2#9W2b?@M?M74$YZsLQA}9VarzK_r-OB;qEf&O?_^P+d15sO*@N?
zd(sCBn=}pJELS8;F2P4$br@}CNm1Pp?=oXKUz$UwqLt&vr!gW#Er~F;LB{ZD>G$|2
zS8=iWe|YbXOxejiUv`Pv*x7Z;zg_dld(A7%SXr+fl~M#G?+NnI3lCi1Ng4L)b@u-2
zp-C6i*zPs#pH0I&4$Ub_6}!GI-6#*batIeuqqKmw@~HUHHr(bDA8oM4Xi-5d$o>H*
znKZUk%F+CMWY_HGST^#TIsK5F#=pigermRyV5JzLL(lQ4x(}AR;hr_g2A(_BUJz&O
zACz#<E{MJS?Rc<GNMD=HbX_L7{9yEBvEoaEu7V5-1(Ip+K+<!0X$6a3<WT$0N37!w
zy^WsZzvv;x^9-*ECPicK1vqqkZNzNX%1hUio{i$)<fV{ji@5_j^Mq}6&|hZd;D|1r
zXx)tHnv+}Pj?>d%sZ5TNY=OgJZ@R_fLdnEHM|teUFbc$#ti)W_Kx?JHw4z*(rBQaf
z(abg()}%fe^R0VB?Wz24MOoPLsZqeFjz{T{8|J$G^~rF}v}XM>SKe%n`_kgbgkccw
z`B*glS*2K0nrGyt->uk3U3~h9!B5(b1J=IdR6=sW0=E(!<YEe0J|lUzf2#~l{|i$o
zux_X0FFM>+vJ)|}52wVnI^q0<1}8-mgkuXDUlxp<(CTMoOaI3ApjfmiiM<G_h+I6k
z+ME(@gxV!UtqwZ?R3&z|$1}<bF;O)KXQn2!EQ{q#Ud7vQrxtA3#+=~C?*0pLx3Ye$
zhZYZBr>%<YE>@vZ=CwS@Ix7{8U}{Nra@?li9&@M*#Ku%~ZLzA{!zEYUoD{H5V9Qp-
z`^Db6zH9SK(vi@Z(U!ogE<ACz|26*2JpD$9W_SCO#+jR*9#C~M5^KW<vqEK2aUY&u
zF>X99Ab{VoTk~=4i_&_zj*DmGdYs3J)?2IFP4!dpANaamNzmAX|MvTv<4}{*Mlo{w
z4@h5}1(bJ1L#P+1=~H)*Pl@5r+}AyI>gy22FV-fcxoJ@$eRmz&vF%qu?oBH@PJ<K|
z88@?n%q$L-^CK{x^Q(ffE(Y#qz(BOFhAn&N=YBk`Oq|@^yvhL0yU5s0IW#bfE(vc{
zr=aHm=6eY}HrUSE3wwhoJCAkhT`hwLi4?M(w=37be}Ae-bVIRg&6dRtH~s8Adm$f^
zT6_$lQ^wnCR9SoKY(#{cfP{Xs@$YMR4HMzjJ1MV|9<Fmil4aWMX`PrqkvEr5%UzM0
zrN*<*8n5?+yshiTcHVv+e(+<yUH!)T5_D&8MlPZe=0GE9H(jT1Gg^8XL+aL6)nO|T
zI7w%g@7!goW3-|N;<avLgV?S?k>}*HbC$IHb;b(XTR{XdRZ{Nj^7d83n&c3)DM9mQ
zmE=QNjBuS~l9CPjCkFuKi_d83aY(hvVrh>vJ-@i@vj0k1Uc(ok5Uh)O1&peR{+BZU
zG(j!QQ<HK?$h*;<9};NfK!5_>_g|2C)9|pMLWTSd&+tJuL=v8FakDt*rA1sa)Nk9t
zb0Hjek{MET7Il2bwi4_k5Uo_l_R(x|+?&wI^~S4&>*<s5L0TV=<uc?fftogct$W@j
zY08o4yc{hG-<!Pl<<>UP;XYjrNLj8%%{%=^HoJ&(6X;A?nri#gW(z{!fw<ZTFPR^}
zyVaqk#WxmzgPGgMPMAxeXoNe8wQ4gC&W36zc)Rs{dIS<s@cOVgLY3NWq0Ii6!6IC<
zbsvJ7oR=V8jTbYOcel=E@hE(fvVNQYC`sKXBn8`qTI=O1-`Who18guf9BvJHeNyag
ztxs8mywK#?22s`9=nttwbxMm<o_d}lDc}NsGk6g~X*f>BVmk}D8-!fiZ#yasZ8co9
zo73-Jm}Rq~N5LDAruZ*rOEL+bi!5obJ0YQqDF9j}7G-3G`Z28b=V(MJL-(fuE%D;7
zKOak(W0slsUXe^`LPD~>bU$5Y-00%vEZAtSSr~Khi{kkjF6IM%nLn~IK!A`+6<_Si
z=BI!!m9-0mGhE%12d_>C`3ETvD5f;)My`8b+S(K3V2?_pKzdtXM0u4hHz)S6{KxP1
zdUY4-%2cmU=>I~tXm-sr$A(9KrdiPy=9j7H<dE|aO75mW;?v$9Rpb3CMlq5~J%U08
zCt%ZS>T{RBXSMO5%&e@y+(Ush0!#0746^<nipk%QrgAtN7#z5_HM-%`|J~;bM<z=k
z$<nMMOxo)mT!a~+ECg>!=ruI>qWpCcd@T34BBRvlPa9PdoO>G{msE}yOCh=pEkS}4
zL4FNi4D36*nZUnjV|X-bBpjL``XN-7VKraqPdlSkEqG7vWh6vy`k(5}ri>lgcX9=;
z?}UnJ`vg$hX*Bd3g&%7~^5v4_%MV_xQ)282ifF*M<8K{PK$jy-`my{dg1q>wF+VhK
zecsip*fXZ~D36yWzeZJfnRUY4ogC^=ySPBpG_mtGCv?iUc90zvKA({Qb=3Wd>Z018
zH;Fgz)1+X)g;PF$y>goLVc}Dbo}>R&xiQvT<Umk1$X>Wrn%R*pjk@7on0P{{iLXif
z>m}cEiH!-`QVu^#xU2h-BX&ow@!KJ(hueK?%8$SrEvvV$MOY|o?sRF6ee!__1E>r&
z`_7TxMQ@`kIVCIEo@1@7<BEZ(*A7cN_>g&@t8Cv4ZMUCyj`bD<@gBB4?*kKF-X6QI
zkE3JrEKat3><Z^itQDTLz0<_mI|y;QP*q?2XeSf6jCnaLNzZ*${hrS}ecZ-JL>L#G
zvt854<>U#-H@q_Uf7AFY1wcOfel4S(S;O%awQE$F`>RxxJVN|(DkSLA8?{GRbsUj*
zrwFG7&+*Nr#AO7rfb2fQZEoL|H3wRbfYE$5v@lnM$B#yd4wq}kxO2N=xrVjYpGm%Q
zqkI`cV9SHhv64boEf)a+0)YEiMQ>v?3)7ISMUwwY2But2$`T)kYRloMFyzNhIb%hA
z3f+9z?47X%98>cf5at6wmj95#eOX@4vs#&4>G;dCG=vc2Oqnr&pJw6--=glZSRd1w
zOip7vH4Vi1)b_Rc00xuu#K8J#H8XV|M%M7(-)+9@y3h*NTkp>rV@8#+q*2m`C$7bW
z3107qx?<zrotM>aZyy(nH2%;l)#@d7WNv0e_!se+3{b|ulY5q^ImX_((|Nq(sEJwm
zp&(npFI|}iEYKa&7^ASV5M~!2iM3L0)+q@moBa8Es^Cc+xe)jb_D`LEcj$!Wzw200
zi0v$m2mR}%rBE)6!;P~liO_1yCRJjBnqI?YryiZzb$EgmZTcOx)+0EX=4ggx^%{yd
zmKXx55cnv&=G&{|+%rCo8T2lB6#8xU{6u@Lg4MBAYpq!jUHH|w3&*>uDq*dY%FI*c
zJ_5?W0!Fb&cA+kP?mxKY1W%M>Ch9qAU?876f<@hf@qwE|QXng`Ac{cM+MIXXE+cNE
z=<Tc<G%a_VTD%8^%qO-uO3Zd>hftd0wyqXht-opqrZ1f7D2xPV1TwKLb%ia-TJBup
z5m^H4YyJP4Q)Sa`u$hZnF=iR=EcWHs;--q{;%|nju%jQkp6pb5XYFd=Z;G}QWNy4X
z`?#&OPb`MQ)Jr_RmVIpTVfzIm&ZAB9dsDNk)u-=`WAo~*3Umc3+T8auw2eov!ncp#
z^s&0+56-n=HT=OL98q{r0X#`<5|Lmh>}qoE`nUF*k`^QNIq@`ss#a93>1m^3LJs(d
z_O(amqaLxh^8>-<5s`(T6R|qjozU!7!6U~V21-pttG(rHk`%r52hgLs4Kp$m78Yfl
zO22%@Y~LxfJ`i)-B)_diaxAHnl<w|a6NWsE^)W469+Ai=awc9W58Ov7#c|BBO9p%-
zJfsZoe#|C$9H)aVVN)%m5klo%p@pRrBddQ3FBKMI=Vgn1q%nqEdc!Ek7Sy#|N(CQ&
zPbp})S(0`V#e_mSO3rD`<_Y0Bf~&noMGZ>k%Dz_vqEdxGD~_-Tni_R#$d3c9F>Fe1
zQ`0-NkfzFZ8C12=aIz$(4kiix6FXSdMJ#v3Duq~;N$a--rp95+aeMvKD&Gu(vrGvd
zB-$bw7js?6C%L0C%ike@`8pYwekY<@#jEi9RU~UtIVrZ1GfL2&DE<gn34vc+azP(?
z*aBP~u$1VUqD;c}Nlmoy8_MvtTFGCO0y}2jR!rz6o1x1D_wfsRz+igNiehs%LW2>X
zwb+Zp+D+Xf<!=$cS!08IXyzc)0q)1gjwefngI#P2j(?AxWNkmcxKAN{Rt7(1`x!)g
z6<pKS`Y2Y=8YzE#rA$0fhl@i~twJk$B<3qZ8Ana$mk*}1`^Ww>puaGV^P3CsZN&VX
zjD1Pep`^KV&yS{+FM=V|eXiQa!~B>l1|w06HqOGq$6aBzBcw$XoyhB;*6G;0WFLn?
z<Ue8p>O<dvY)XTbg8p4ujcAhrnD1(4K8+vLG6MRLn~6CrUGZzD%rH|MRci7H?;+MM
zB3Z!2)tbs1eOHqIlBw^r#W<?m{t}A|Z{1H&%#zi~Zn1cS8xn`4EK2<#3RJ1WN<l2u
z<@Hit&L%f4<6#c@g`zJvWsu#Qont+a)e@s~2_ER5-H1McznAH}*DONsljhScut%VD
zR3)tj8W-lit&Vge{$n&Hs5OT{`lrZbuF)_F&81MSWMOQ9v`FQAPDIv0U?(?UlXmv2
z*loHGwIgkT3a_=tx>n0kPHYEW>w8pv#zDZYQu83&I^(WV=^&xGBo(tNn%9clHF_x7
zHSCJ?wf;EiKX9(nbNTDGBlWYYq~(7hAFiMukCLEe(nAmVpEeQt57uj9v&ccr9REDP
z+UF)Eueu=H=B4T7zi?ch_$iyr3*wA2KE}=Pz^kw!ZrwuQ)BSJGl-|Z_p6PfJkBYue
zY%uySF#dUCEH}ks|42Ptun_Tg*3oSrwpeMTgP*YT!3=F=j6V=ZpE?jNCpzPhJho?d
zlblyM{CHxQ=9(M>MNPkXunEuHNb<HZLYQ-dg(Rw<7F_z^-eU3H8A*0>Cn}RUMS%J(
z4cq19?9otgiMO_1pGPe{^DZi$YC7?RtHfHZnIfYKiLLcT!AEEJYW}1w*S~e`(Obw{
z1k3z9fOkX5)ni}1UrgV@#vXTt#c2bmmf3@Bu3rG{-Xe|Y=(!HbT1vBeNB+x;-t#l{
zW1?{FRa>D?>?B%^2)!T@gn7aAbeq4$BGiAmqKHfHrJ}L(y3kIU%>Am3tYO8HF$M}N
z?p!vS*<7tWLygv_@k|yXxFN-ynS`dMj6*iE$T`;mvBg?gtE!LaqX;=mcRY;mJZeFq
zBfkfwo1cqC>g`P-cj2I7YJ6s_W>~a0dCo26%lB!lL6Y^Ix)?nhq43ure0jPo%h9@P
znI7WlCx{%EUE1q>_8*9eQg_M1MoqmPRDxuPwRf7Q=28&-n~;R03(^DWO}bDw?w#YE
zgg27D@li(N-F`p6h2O9=4{dPSYb;Y+SZ@9T=_SB7q2xkOOF3`mYm0u|^Mr+KA5Bv<
zvm-PqpTOe4_bzR<v=>MMaJXe2iTA79fd!9GlnZbh3_*&HmODPXLR+=AV*=bSBo9M=
zTTiCD-ejcz9mTVr%DFM@_F}cM#-D+iOb?delr7S&oU?<`PjVhY1G@$#IVxLnwS(T-
zYUTkA8`5`+YY!g(5w>jq3-&WcU8w)izt(R3RXQ?BSEKEEN49D&`1c;hHz+pzi#^`M
z?ofS*ldg{1VXHkv4WBoXW~HCP`Shi*nq3lo8!3}<uc+8f>^2hOrp2vD+)G84BQ0-V
zcRWp^e#oqGk9r2H(gk;dKRJWnc#~v*hsrlIr?kKmNlYz&_LR@F&cNX(d_Lj4S^G6~
zdmOVTlV3ULg%^U_`N~?7>)X|65U>|>W=Vy`@sD%!=C7tts>ypPiV5p&dGa1T(feUi
zo^gS+p6O}xS-DRGEeA)9P?V5aU21-HWc&kWy-)p*Zp+Fccn@GG38jd)Kk%3P-l~4f
z04fFe_*Lhx1WBbWUb#ZbG}Bp@h-{eY%LyY&;p$alMW<SJq(0+`emA!?aZygu=xlwp
zSO0W;l{N_t7GCHiwM|GLVZvTei!Oc52u@5u%+JIvSWSGvyVhU5n~1wlbPh4d@n6s|
zNoAnfCzIJ2782V}!q()y&aR<YXsi1)nSUw3g(<luLFfOtCVkTe{XorB$Q?l%A=(vY
zkPS$K&c(i_&(7)@tL-u$_-j>o1b-rGKr86X0bxqJ+qa%8CJ=+Zbfz#=K<Ma3gm*Ey
zx@pY^k2M{=_Go1lV-vo%&`=Fi{s#QQpzZU{B#Ui1f&@I|5q{qn0Y)U?p$@Wdv~Q(z
zMA&)1W=#N31oGsyj_u3$2(>_a#|w>`BQ_l-gUu8HSxg%3Vn^9bM~WAV2h96>Ft)s!
zXsLPqXKsp&O9lN67uPC7^;i7IJz~vZ+T(U<*~zZ-FLREHW1djC432yh`TT9JX=RUt
zXuX@;z&yv~r}7pPn-pM=5s9He-X5xRxH!@7$19{}q_3xH$u*oN=(*PkpwR-ktO@l#
zA_@jL{t>j|+Q`w!{d~4+$oo<jq=9BVfP+VcrPwZ=en_dj^g;2~Pv~kYAuqt<TYS6k
z)BS+35Af!f-HHi#$y)LRkJgys2DVr?9?5_}M(*PeQ-u_M*IKgli2`i5$e0^mbVJg`
zEJakV)K%8YqsT899njYlaIEs=#@E`M2P3bccpGlm1Qr<z1#Qvzt3|&<CB#k^VY@hY
z&1!7@tBvPR3L=+}a>w{X$+4v;xula;KX;~1y~7q$kPZI{@F+k>*?X&>|1D0;-Ce~z
z!gn{|i4Ott&YxF$y>4^pjb{p?ewa?>V3LR2VWALhy#B`+s&(APlQ7>6W$`&eQMc#=
z_aeYKY^$%T#4F2&Pc&1BqWDJGFYi(Q_`r<q>{tZi<{!3hRIfUKa)$7e^=DSl!If-Z
zl`#i9r_kfh>bWKeGi!C#U>Dr-L;q6@0?1!^v9vPcXG~PVIKNaoeWUJj!RF-IVYJjn
z=2(_hglYE7n_^neeL(@L_NsgPZh}arV5x_aIZI9F&RW93(jAL_O1}M;9hYns!s%7i
z{%7A<)uvlxh>ul&@531y3{H%fVlFKpKcci8$jZ<9T?cluXjc-S<o7V`a7|1<TLs~c
z_Xu^KM<{)NkWTbmH+hJ%3BC1<C;mbgVe-qgTe(+Afjx;Q`GWv87fDxD<^7bQ5Me7x
z7j}S>_Jj30PY1pRG~$p9f>s>U<BNZU8R1JL+SDmZ!DR{~Ixy8mFONe>Mq>n3XT9&8
zN7oIGAdw>bTZFVOa^5}W;0^oEjzhYi^)Ip$Mhk59cPPFlMyH=U)Re+0-ifHLfWR{n
zg97(&?f1M#gwXSYoo}61z7%p8A9Eji?tc3GtHvbm>>=EEEBo#j*nTv#0}jNUQ6mx}
zm~^$ioVc}RuZ%t2TKVm8(Q8X#NsE1n<Uk{;crR(mBMR4%H(KuRUOVa%A<E{wap#bF
z!VKC_c4i$HM&Q5;RYLkyD1>n#0XBV+u=zTT-NpZBx*8C2v~IPI7ji@jiFCN#LlQPW
zh@M^ER_2n;S?X6htS}=Deu34+k%%~>ulz>XD}LM}2fxC&4fbEab{kZf?pBdzAWbGV
zw+|eW$(YgkM&74;*t3TR>rQ$noh2Dz7YInNkx<jkjsw_rIOZ9H5-FV;!UGEO=ESm<
zN>JY}s>8fr36HtF*|DoG*L9eO&6NQGhavLxWsv*!ujG=gjycFWy4>nXin|sv7HP<`
z!+^{5gW^JBEp4qFf3HeCBh)r2f#=Q{>|+l9vCQvuo=N$m@Za2j$LY-K(UErP;t0^8
z7e9;1^qUT@-#|0AyI#S57&{;bLq+bjCTVj#T*bIGv%$vB%5W0dGl~L6k^Y*e?3RJ9
zCfX$b(Nc<D-Q`uY++Aaz)egoqJO1%7`4ILcyt_LlMn74Bbh|?bIIW!bC}Ar6pE#+{
zu0B}`Dds*INhbe!zkW41@>gQFO~oABnDuDhJsi!)XZEVx%vc8`=^7pgcV2Cfu}D6!
z!Ef@g-}W%%Vci+s;TLIbL}&2{0L<#lfH7^^p@1D}U<=oWA{IFMG%&88Rj;*gnlxD}
z*C5}#@xctc6Ns#B2l;QIjqwp%ZLHe(Wwfs$hKZtIQ%sWv5LrR2Xl?Ag+P~Q!Ty4$a
zi2qDoY$0Y1e`+bTvNL++2{&})G<cWSsNpffo5{zQ*%XkVg_Mnl`eXNlhs|Vesb^Fz
zRw%@sv@Fxve?tI`t$A-MMeFgVsoj3n4^Eq{bEgEPnQ_ifrgQfUVZE6p&1uTeeC?m7
zlCl)>Dw=eh#vP#bNiu6~8P9QFlMY@=j#(u~dE~pw%<=7NFjfbVhhuaYLK(nMn3t;d
zz}$Btl2ZzHip51GkBp{3Uh0CHDJQ#sRNlvqjkS6fhW*HN%!^2P03Sx9kn__-ne!lP
z!RU%T1n;%#=4TZd6egh`Ll+s8Lp;4f-kAdq^k%<C<N1H(T|MI8DS0)(couTT3CT>e
ze)2^7epIX-Gm(YJ*MR#y4_A`ZV~1vf6t$(OJkWIg0AvOCv+eEPL*cG-M=^KTAOqU_
z9u%__s+e@bZt$Jxbwjes`Q6&-&viJQmP9BiZ*JvMG7fHubMJizz@Lm0tosL2G+4vn
zeVynyxAD<aBE_v2d+%7&Qp}FuuG#KL`ba>Lku{~F>6#Ifp!vk*h13WKXUwYc!IgC!
zp;#JO(Mal%(l|K4ees9IgL_4CLiq2P5Z`(<uj^S-3frag<6Mod!MtsD8vft5Rd9q)
zO)U76QfOSk)?B!E*&kgIJFKX^1>@0P;)_tM3;kMMeCV3{4F{p+tfgU@x&T>feXUf|
z780WFL!pPRjpD8GNXZAd_b@dqlAT{0SI0~!-Lsj(cgyG|Ra&w&c;QDv4U<qq$~Vpv
z_1myugy^yl(8LZBhkR01*Yu>;qv(Ai)B+gmDhbTCB=0Pa*P^*XmI!hhKtT^73i<_b
z8M10>G24E&zayeD+igWlbX~LI1BYTnb36Kf#>n1<nuCkSMa`|qx0HJ)3f{aV&sxU>
z1>k^BUbjCF*kfKanG+EnyRUh)$l?Nl7pr5(lnLYml<`6pGYL=rFp*HJbwq>)9nO{b
zYee_I`c?{UK)urL@@CJoB&z+4Z@VFQbc9(kkxMq+N(J9EKcEs(wrOC2tTA?+(iy^M
z+8xlVAp=#YknZ`1FFo|HsWBV_mFrWO!;($bo*O8-J?lnKe&%w5HlYWnYZ!T(=!^d)
z(wyK3`#Db?9BWS8pg6Vm4-DKJ1XjAM`0v>|$sZYI9}})t>0=ZdW)my@%78K)EST{;
zp+S+S_2C|GtWL#wb@uEr=pIkp&?LN~R$Gph0r^H{7u_0iLu%%|dCtniQ5+A0md^N|
zgCR~%UC)K^2xqjsaW`pU==SHEOmizD>M!Zg@4NLX40qUJQgOvH(+6-dMm?J$9tBI_
z(KUCfAHp>dj(ZLDGf1+4C>H8@k%~v;%rB@##-OC&m-kviGrt`-R&xA`fa77}@R_xZ
z*JBS8JC#SM`cI$Rtp)z4(i>kp%S{<eOj-JYN9=kO??ZQk@ioAlDOz7ni6E3*gZyBT
zDBen=MarDal;bg4=}^=``ex2UEbIN|&(%(7n24shZw=qM&ptViQAfy0dE2j;VvpE2
zH78koBpZBvwxTx&pY{l6J?XfX-SNl-F@DO)czdumhW@+T#`nZS0-dmi2_=@yD!9`#
zvJA88c0L$3H6%&L3dq=Tq`<!7i2}Q|^!k<$tMMl-rNhiz%GsnoKffSx1H~gIae{56
zsJSai!y8M(zdp_7(1SUgs|d%vWoDMbTI^Bp>qv414m1&(d;f0;&D|`PEGul!3p>XK
z)p_7z+4b(7eA)Ju#Gt>(mHyEBaffB+@;#bsSI=pxVv~TK!-E9HIMa;_N3$V8Dy3ms
zadcO7t?jurhu4ol+}XSE#F<I1Z$epg7LUqqh_%^(Po6=>zl@MR+u)lX@_f^4Zl_5x
zu!D2k{b+FbC(v5diWo}-iY;}he(+|ioWL4Q*29Z}QQK{Nj(bNw`5uLY7^4AG4)kKt
zC>GDjw2>)v&c5K(o#Qe0$Tl&16yT>m^;RKPW4K#T+8N(2X{tl^PqwRb3-4~zj?sQC
z3B<}F37`%NV~a15d-%HGxEV77_3#*}SH2#y5Rzz+gS1-=S1f_Zl$Vxjsr5{Mo$`2F
zEp{U<+L{k0hDcO#+y=LkuOT)E*&hMLf9>3^F&(Xp@{51QXK5T0WG*y6&5JekUcqPV
zF3!yGk3^INZrBrdISokz$^?94$27->+@)aB=VV$_LmTG`*grmmFWWUxFH~^<Z>v^~
zTV70}wTSnyiGL2mm{n}wM?ZNkEdUw8hgQEF)`fXm@-t3HCEn9$yiC@{KBbFDW79)r
zUO1AsJGbgP8UG4K-!+NdQdTi9EYuF5{6HatDPkw>=0#P|OT9xltL=%&E$2m%VZa8r
zHA29Ae0E=`B2M7T+MdC6#A4h1M}j!V``atKWnSJ1oDHSyHOT^H=V-5|!1yH>10A;<
zX$x5I&HLlE=u|yDcp#A%|KSh8rC-|;GZZV6Qk!PK{=^YWQX0!~e<FtD)xJ^I1G2c3
z^OKC2;(6%4WO4z>7IfPL2s9eiI*(#W{sk!vm$14nhQ2pM#KvN8C}#@JacY=}aQfYX
zP6X#x$bwDVW*2Ni^xol!1n2A>)|!rxnPb3}8#(JP0zbcQi{_Vmvxy`wJcoSvsV{o+
zb;%EbxL(pPx+w?wQDV${1HQeS6YmcI6HzLkX8HwCCuKenY@$$fmeZ1;?X7&+mG;`8
zoFEwPdKO1LixA+}xry7dsUVUp(3~#LEW>>W)1AH5y6Jpn5s~$_Vq`sM$5oe>d<Z*y
zd>bC1IB;(mXNp1TlIzV-Aw#i(r_H3~8evP=Z(*R|)kp*t7P3+C-7eYKV=6bTsCR^;
zM|nGIn6e4_^4V9>fu`yAjMUsOE=6%Te{%QInX^HV;zMyL;UJ{wEmZHBy@}$Z<KB_8
z)L%B@SOvubq!~p9pCHKzm(d;zThKi16eXf%>C{5tz3iF^-|&7#(faH}=+d1B9yz+W
zBYw?mS*I8o+HuzajF}%>mth|`;tYlhl5%ZAFvZ2V=yHL$4{~6bzE?Gv--6~Ur=0{U
zr^&}x{O9sArMV*|h1PH~I&-t}O^J--wa?52jlLqT3A161@`=9EAyx61O~c5aX}n{Y
zw5vvpS2Tc%rOst*FIi~9@S-;A0iQ=2N94Ccx9@j_c_%}1l2O^53CdoY{J*bJL(|gE
z|HJ-g8mb(1En^zer4dcWF5d}#iqAh{afmRX`Ml7Fqqi1=hNbp4yJTFNpr3`MsB&Uy
zaSdh|2lsn;s}@`lsQ&z4-^#VRu&5uaan}y@%n1uleKlGZDTD_a6(#{w<)F>p?p@?K
zr}o0k1jw=%K|t%8cY;n%**kTnli}9vX<pQ)Ki5-{U6_x+r-<JpY?u_<rEKp}{E`Y>
zI17yd;(@r?PMMa60fES6wMV!DbU|dlecM^5-i^2K;f848N_MtxPP7(`Ub;AG8?i{3
zYt$_eOuwms2aO_T%n+>(<EeX2SxCiIgIATSa9S`v5Hsl5<70iY8yXipIy2MU-_J*A
zoD%^WC!^lI`P>)8jEO$MRUNPg_%P`Hs$=I*{E7b|*6vhsn1D!vPfpGQEf8eY@3*k(
zl=gb=(41^7+H4ogBp_cx%|&!A(+mJ2Do@f6U|xV}rpeVhrnUm$4mX<{;IznY&Cgpq
zs452T{KQqw?$o~!!3(#2J~br$@m1uNfUC;)SG(Kv>q_Aw_>g9C3NcwD<`QNA_^f2-
z$;Uykvag&bmHDc;|BSew1#EYX-{PGzIldTiAZ+Hso~>`RY8`EFI@(1@u;YsPs^`7w
z2w&^5WbQ^F6efT}gVD7Ggz}36pIXu4eAZcG-9YS8{vSIx=1EK&VlS)BO$^y2(s*Uy
zfKlCQ$cw?`(f$pWHcHGj+-EM_SYA=5a=Q0njw{XoJhUPrIeiDtaRPvFDJ)V@o^?H_
z@Y8Etv|FKe^-v#s9KX%Ns~F7ACFrBILZ+jiaA5`8zY$7EEgj53xS&=kIHEG!1DRdg
z3jtU8?(KrQ@M3UnF9+BwUa13*+asd(9nCCG%!pF#s+1gghTAiIL)^!GJotlAvD0FX
zeY@=fTGF0!pvo=}drNE!9?{h1&p=gP$20`V7)?>$t<}CErC;kX6NkLegEO2g|9*~p
zeBBS+Y~Hnbw&zhnQSilTbZ_@!vu+?81@4?wIJ(Vri!=qHr#0ySs)Ok3>`pxzod5aT
zkla8C#4+|rWdzkey;2YRW8rcy$0_i)8HLk&ojj{tT*>_JBb(dRj&@!%#Lerv4f~WU
zszwMWgr(%t95=4{*EfPtb@dn=q`ud%C-9JONh`o}mgyFbi|80Y<`51C9#!|p=sFz7
zwIVR(K&PMbA7m#4^fHKeSF$$WkoOYVY_`^GzO|hcwj5{;i0f|}0jF7Qux>wlq?NIm
z7uim(>;U8e#_IbeZfD~gA8pw^kclM1+to7>$a-)hxHhP=QKT3SIugR(aHP^fP84p)
zM3=-g=DixnIC8;M9SR4LQhm|9V~y38<9ew7seyo~UH_C|<76QY_=Twe28*u&gsZJ^
zKVEB~%9Oxhl9|g6enr?ry7i-WlGZ7Irw*Vktl7TKb{>`<mh)b!2&;7H_5ZPPG!uXY
zWXHK7lxMmEY<w(SvRq!wQ`ntWw6y4Fn?(1r`$buB-yvF2BWtWHzK>|WBlX$}WZ8KW
zW(ZBWh`D0=I=oW>+va2o$<x8cKmHs(oA}g(`0}C-^oVw;iidJa%c6+q7GNEz@V~~i
zT(MQMHH95=9t?^qO#BfJlOEkhtwXy2;B?cuigj#Nln<tk(5#Hw-;a6Q)s`Z<a&`rg
zf_@rF)dS$)Ftz}F0+vG6i>g5T41m&F87`EAiyoY*0aFtA7Y45iIwMs;IJCu+bX^wb
zn2*s^yEt$%hf5FC0HR=W0p2u9_d<;N3+xXfi?j=cZjBk8jKtEnkxJ4B_=72G+c$Aa
zhq-n^xF(y}DAx}df~pC9lCT7ZrI5gfA<fM}oml8;SNHU%H}XDw)V;5Ywuperx|zW|
zCAO-Vr_UPxkSLEnZD8U&FrV+5jQn@bctU5#)u`xt+)HSd?F#d)z|g;<sMc6i$}NMy
zvLdE``rm|>xl+0;;8Sb=6yK$>BxyOciP5*td-wxhV9rUMJ2IY?>a<G{;>nzw5!RMm
zL8<c)rHY(syMxDx_+z$6mN>(;E_$(}v0&_3SW6Ag>US+cSYm=IUR`|`--oy#mv)4A
z#lm!gN{kM%zk0xH5>=dQ?&x;42b`MS_Bz8Vw9Mt*&V=`@X6tqc99*%4GS#z5U8suK
zjJ>>6pKw0z%hWYanI3A#CSH?hfP71Ia}%Xa5}T*=0CAd}{~SnO+Pn|0oybFFG+NcE
zy%lz9!cz$L=4ch?alV@6okfK^Oa8h4ng29Q;Y!OPIXfl8rPi2C5hGd=i3%51OrBZ7
z@F!mu{3(xeA3ZA0zpe7ZcdtP9$oBWa!p4)`^h-CV;`??|g}!!r^lmv+^ZP??5`ee*
zx?#lMG{v8+B=%$2za|So^IaN19W$f6zvPR0)H((Qjq>eGFwj^~gUl7wH4O-lbjBk$
zR$QK}2h^~rV^R3+`Q7A!q<?&n_7H|(VJcY@D2e!IYozOG_7Bo|dreA@(~Y|wtK4~*
zCoisizP*|MN<nQ-CkqN*(U?}hOZLE~o=yx?OHqAcRW07N!&Qu@j5Nf7M=7Z!2I96j
zNV(hzYXoo)i1cY}T3R*jgr|24$U5R@QViz1!m{Jw1V@TQ2+ourzli#HSl1Nohh7QO
zuo{1c1GcG?5P3=pi+537gw)sf#&Oo$@ZIKft8{5@Y#JxNw)7cDW~dfo!wgQb8p;7y
znz9HS2~TSn-uefDxE=2ThXaYy!cZN+p-%9Nly<ZJSup&h!xA6<kn8ic4QMb#-4MR+
z?8|LL4F;9TQ0$>cQBn-)x_6@Ri9x?9>Pz&;^&Ku_9F`NV3-<GsgAEtfYR*>06vLJ1
zC*SqE6=a_xm^^$m%V?5t;zz3#b;tpU<0>l0Z=EMV3<gtSPP!roq8GdAADO=p|1`yT
zcXZP${00p@*Lz%%=L}jyk!sNFf-3_V8noiiH2j!;c(V#{2^_{;A{yhwJIqqreXS{@
zbp3EjOcr7(bmC<)X!9q3!4!=_N=#<{*=U_LgMnSP6FZ~0IIzNcxPE!%^Y{m%Kfka{
zqsOq%g_an3i+cpB#uzOuP|B)Ajd-zFYN>{+@P32>V_3l4yC0X~+?TBx0NDm~lMP8T
zS?qs(?X>6BTg)t$EnVff*+^la2j}}^<Y1eg-@se4Swd@{Z0Xs((Qw4l!D(QcpPas3
z=}rmxE|2}m9^he@*;$Juix;@V){<Z2=gXx9bIZKDWokb~pqZ2ba4HjoEuonZ-g3LM
zy^zHNC*ni41MwCk2Hi5M^p7}qKh8ug>7E|IMDB&>rOS7eO1cC9p59s&8NjpkU?FzL
ztH$o46600F4gmL^#(~uM<=K&)F39hTD=D<Nk5(z<O8x%R7E)Ie)ngnE<tAPdOY!GT
zP3;_d_y{zXhHyvh4R@)*`qFdXFH?nG{}m|4nd~|!tD+7X{`God_g`LKwssxg9Z==&
zUFF>=Gq$51BIO=Vb87yL0jOq7M*rs(sWe&qxPU*BwoBsj&nJHT?R{W;c9TmMzm*;w
zkb#Dw)m=oOkQj-hDmp=X0WEz{u8WT1i;d&JREYSVnG{p~8H41ZmdUJ6lwZ6nnTwQ!
z2&cDE1d~PJwqFb?y@_>51)^+f4m&*J*U%s(MXp{qmzRrfHu6Mf4lp*;+s;SK<X8s4
z7`v7gtzXyhu0LS8i4x^sEOuxtjDrT*Ok*7(Y~rcPfJ7PzD%eS;)*2h1qbDoAXIDJ3
z*HziXGh8zrU2^$YZuhLtPlMB~CK6!wX&)Hnc83eJGqJ7+yBu%<g@<+cNO*5wY`J3^
zZFIEl{6wKNVXCvE`ny4AQb@X@_Mym4n_D3VZb(!Eoqr(=E{5xUQgm=_7G3F}R7O}s
z0cVvGe=Y#t4w%fTA)epb32}^p8D&3}nTS|??6_n|tPFsb&t|HChOVQ(8Z~r<*-EGH
zz!%QY1EenL;Rfuh!lDSx3zUw)Z!i(oQ?z-1kHAQrj6w(%k~kr&&oP~D1pmJKs~P?<
zkv_Zoud?>`)D&*UJUF7EvvcHjW$U;_Q1Wneq#~ct>kAgC{6q|GB6UMvOsG~cs-Jl{
z+Lz1dGZv0-G2xt1qS0kewc=nBeXW+mtcm_KDVi4;#;QXyH)%6|_)g}orOf%<pJyD=
zo!xqNOvx$vUAX$z4iG*%!$=R97e$Cx9_?d6B;bq%+fN^zzgxBKyn|+-PGG`fTiDr0
zB0)k4C+j8JvJJb*)DIMG$%Ra0hT0lD`dO$Y#ePu%k)_EcoJmBc3&X|39_P4?n<53a
zg5Lf(b!Mn5$a1uPd{tIpve+oOgMY6wZ#@FCu5~JiGdNY-c#zzQif@K6_n;bK|3`Y#
zv2$Xx{oq#x@dy=YYIl(B@Yj`<8l%EY(j`9}TBAY4YP9uujXg_0O55>xD|U2#u7M`S
zbyhz9czbA~kEXewXt=U?Yni&C=`UNF6;sC}57U#3!&o!pSMjyCXCZ|^g9?}b3V;*g
znKQY5>H-g=7K32!94Mky%g**|7L3ej;D_s2RmuRHd(PHM^rJ3OigLNLQBv)_7z*8C
z3pwQTCT7d9Wx2liEkpJz4j>G<|4<HdnN2eSu4ia&#_kmqg6uoYp5~j=+%1)78++{_
zqwL-Ud+V3sFBc;qnhT5KL8s?jHkqWr+p(Ohg@E`7YzI?J=g?eybJ15(roDSbBUn-o
z=o{3p`Ctz(+%65U?FiwO(OC+g)*nLMU)jTR{w{ZkyV!eo-(-7!q^<V($rE-jIL!io
zO67oj&nm%Oqgu1Jowd%GTkR}6ZQhjtYrXQD;LXEg;|-_DgQ+or!%$R1VMDZ@3d7n+
zd3>3#mC(<12iP<7N&|7z6XGq8(XW&>6c@6pOcdMw$;3c2HA%t6Cr%63&N=M=ddd~v
zaX4?^+!Zn)5;>uMPhm?C9Am???N@K<Pw}3g>q3o&n}{o>P8pcJEXqNtn~_a8W{sa_
z@HV$4lCy`2WG-hNCYdp4JKw~oEz@NXqQWqg|G;o7FIAp3=WADJ7z;ihcgS_7x7k<l
z1S8@4&<E7BtA576Z-W|tb+}|6=X6UR1+C*4WEqHoDsD46F8Yu&&TqAjn2e>rapgg>
zWRewe(Z7Qyi665-0O^<(>HaV$#XsZAu_EmtHbIt~rGl6nuQy+of_|>6$mQk0kJ5yi
z>qE3)3`XYt;hzjuA^R<aBV+(vJJ<zXRRiW<3N@HaKbZa$qS)vl>RC`K|H79(8#VTk
z=qT@l?fZ(R_efEuo8Jx}84OH+L2#G2I_f-Q)Lw343qnCkr1JR$N9SxZj5+5`-=2hE
z9TqG9by(1fRfiO(Q1S77akDmJ`ZxBhnEle$>Dl;3hnr2O>AarYs|kuH-h;Z`#a`bf
z+p<<-;i(VMdaJi*S()|{l7~^L)^(8UTkmf_B~!C=rk^=eS^9Fmx;NSZ@fM<oLmfTD
zSjP?EPQYRE>Uw-(BqQE=c{{DGSBKi6oEz?swN7G2R~FTmzGMNcDt{6O<lum_x<&mh
zN*z&XQ20~sMr}eKsFQnZw-=P(m79|vI}H`l3XXq8dZTZ|LFtU}06`Yu4--b$v0j%3
zh6bFqDKNx~ijkqP68d|5$4AiOGRV1}n`hK5t7?OnQ6+0APT)1&r+sJ4-^t$znoUhS
zdk99vFs6<Mua$j>!AE*59C^=sFmM!pJyW3EEz{g);83Loy3W{~`A_Uc3FmAa%NLv7
z-=1vv+3&e;9PiYq*rhEBV5sP9-Af<2_I>t?URB#@6izG~y<LUbqyqCp>kxczo59|`
z8kX2y6SPA7^J~dRj^0t16-}<wH+e`z3LqNyvsDIDnho*9H2gVq_gkp!v7%{j=E~hj
zZ7=s;0ya^%YGFQ20j)j!=Y1jhAJHmNm3U&3d52m@r?B54s+0X#EhxV9oncFIOjm+l
zF6#zs8o|2~i&?j{etFVcGbKijZ{{7x2V^-1IvB?*RP?2(4*p-?;fxM)Q{^jpiV88w
z@rWEQ>LK_YmcnwBDgWc&)`awWddxU6b@L<oV7S1V;n1SCGG(hzcI?){4l~6Ju*El$
zo+W>g0fXhj<f*v&3M!Cg8T4g-)vnYHGk{w@#Y*o^vLhh`J`lmBW1Avo-$4vYKA>eB
zT=+j>>Q2Qk@v6ZpY>wY>vS2ba{>U{u^8Refcx#p5{5yBa?+8)aSVVt^cONQ?ne@t8
zU0h@1@tjsasb&ao0D!Bu*$TtqaU#Jx6gv#$ms=g;fnmr+(!s6RSZ6ljgTVAKtE@xe
zb+K_d1O>k1=ofp_UUT%tfg71CxYulEKUwu%yBb@46Dq90SYfO{hBT$wer!#_QQ8VC
zmlCaGc)||Bg7XfiSL9xO15L|j!u%pD=Njd7Q6)HWQj|!)n6e<;W{4qS%?e$fGS;z4
z9-JrGFYK0uo&7^#y)Iv9?Lcj&nrW{vO2_Nx%y)V72LS2*#O>2*%)Fc#OoO$XzE_X`
z5jmUJwit+Up*5@bL&k$RP%*{j3%?(GB|hfne3(jp#5fP|vT@vlJiC_V{64nOyT67V
zdIji!HCBr*vTMw!i!%{5#Y6!A$}eN|`}23MV5`)X4Q%?4X~+(UrGBH|zMMNO{f^#T
zN4QTaSyj_6RCr&B-La|br#2PtAZ+~xVnOdXIV9*uu8x($#vp9ldf|V*64E>sg?Sq}
z^SDOtR==q@C6`;~s7`dFL`fqFZc6{-+{F{bvd~`peQXX^PnaEZzk~xqpZFEv`|rn+
z>in7F*B|`|<1DlNzx<v5?5=@{)ilr$gu8H*{f+~k-P7_%+0rkO-^M$w5Jc|gB?0N^
znYX@*XHyFH^Q&MmH?GA*8>%|Eldfif)>HOZ-3n=J7h|T|QQ*&F4`%#gyd>-G@v-Nz
zR-`w(kFs#U#dhgu;F}kgA^6|}+}~CFy9d_+%j^G$e!G1F4;GOYhQ<@k!l$E{38+T{
zYy`O}>ecL>QU1T~M`qSEF3d)OCLw^pKW=VqXzH@N5TO^$R*K&^O8d0Z6;|yy>Xa9V
z`)JB>=R4z`KwO_wbiiJ)J>K*nff={|l-WeYv}c5HN@l23=#b=pSDY+m&E@^vvWp-a
z*L^ScfDHS}FyYZ8a3B^`-4~6m0O83G_5wMyieDdH4SnH27%oS6jE&!!)|ytluV8&R
ztmCWj|7%Vt4U8`i`+Qt{B@shyV#7)S5i@bl0DrAeA<9*fhhRkcfhEPQxCJ&wUtv0^
z(~|>h#4fdE>Z?o|by9S(C547mJa(Xx5oXj6QBinb3_D4$7R~h0%kEt%HRR9?XIL8`
z`~;`r*27gKd+Mfu$!{XiI}S-Efx%k#<uEPHP(rB<x;Qs#m^(nB=8n1F0%qRI)c*_K
z5Z7U&)JfwPO((%~iz~HZnuX@{-#vZ=c0BzKZWx{hy2Zv>(CFl}TxdwY-1+1v<kPg6
z#W4z>l(PSX<D=B|VsODACE`qjhgpOp!{DWfMKSm?J_e}lUw3+4OB8A-y$=cILYq~*
zX~z~qC2Ec}w{}hb&JnT0X43nt_t}oBe$1Qs?*|#x)&<*~P{$tD*zwYO`-ZP=cn`-3
zYTf1-&JnM2bj$B)KUh)r)9ekZziw-a2I209_=>{lFq)AcmrYt>WP~~+!p@=8OaFCQ
zcgZ{Oalv}R7z_4xL}Pf&F~m7y4jGI<WDUi?x3iFB0ph&+e)=M@Fb_cif05rc@5vSH
zykQoOPX1sVo79KRV&;cn&aTx&cSQSo7%A#eCsZ$f18vg=QfK{+y+)8@(I}VR-}&5X
z%=P9}==Qtn2@7U)nOjY_yMyT%{hGSdDNgB_95qme;mGsZH0u*TdiflFo5R*1Vo}py
zcv?m4P;u2;bk#}~8L>V~S}j4+Y}l?mnH)tT>c)+YnO8fF8^?K^!qr<{V$2P@-igcq
zgegr-LQ&Pn3s|k!{mfD@fEkF|9?fuqFJu0F$L9xQmP+?x6NEq%o1ByFw(Q+EX*i%I
z+eQ3cyj9#?;`dXJ0yVgH&z66@ute0df;*>EGyr~J!d0v`ORAx35K<8JL0>GZfFUi`
zFC?p##JvI#ab)+&1B;g<xDw5DM*Tep-W=cJY062Zk>G?VatFu$(Xg41a>Iz>$?-_k
z^xhp@)M{C`>%jF=`8@-hnDYc^+V{TezF}WcqYIUGj-h;C6IP|Lugr*8W_G@-f0X#o
z7{C1g9SI{oBcrp9B@#wGML{m>A&vJpUj?6QR&EB?PD3|ky;c-$DtBD#!o$$KqjPES
zLuxB1*D8Q4=tKg0Fp)ngA0t(TfgS)Pn~@Xk%WztQivnn7(zc_l4tgR&U`>OE(~`3c
z!g896XLSvRfw-QkyZ;C@Bf9qlV!rfTBX0@RHx6x9GZLEa>S=seON_#OGP9+)Hg7fo
z+1H{;oqg7z5yB6$ps0*X!Cn4P4ilR;zy39>kx94YEyinBvj=17g-?Ald%ic`Fr%4y
zq00DqK#)f`P^TnaN6v~%+_2TnLG6GA>cBd8iK@*yl%pQzu+fNEW#JqDiU4dAa;PJ5
z`L7=+-#p`DUGsJv(9IhtiJ;oZnBAN$!fPURcgqowv5|HYy*`LpyY@nxushFLa#OwL
z)co9@YL1X_uG~^J;S|>G-Udqgk%t5$4pK=Y==xb<Lp9dNj&WNx=EfXmopF1J)Ao4@
zcljR>W%)nez2qY_;5PFlhzW%~DCg*~55qhMo~{F0j>XmR<~O{<|Dz~13VnpuJRezT
zcmU{uf;6;tvgr3&k9`^~4Pv<G{FQdc7#uaD)IjNRd$w*So?EJVmnh33u?SIN9Wun(
zLiZ7#s-Kuq99BM-u;K}<==DQ*moY{YOlSX3fKvquvO`5FU{B=8Fnd3|c%EzzQpxj8
zSNe3k|5^YP$2xCc$>}<kA`pp&4O%~dUbN_X%WU)4p!r!QA5uGod;DB~Ut!4;=Jfw&
z0j$ILM0xFN13R=vkV4Bz*L?HUc#$AH_vmK|4lWhf!5N1ZvC^Gun7~ST+AU<4nvfPd
z&aDFtqX<Z&<5@Xz6OYaa>GMqpjCVI08s#Q5{bjY6zH>V_M>6VkzxVnuyuW02YH|Lg
z|E;n`Fjs_bhU<~nqWy4ek{DcxXb}KLeRZs!cV^7kF*oX|WZhj`R<gt2;DdkbtH;az
zg?XCCV7mLW(qH?N$8QK|$@+a3$&z*OFW-UTftK@R+lSOzpQvKxbN;&-5#5&mm>;YS
zDQ`zFF7Ya{+UClM*cU?)IzJ1HuLllI!_zGif@zh!w5O|rM;V+dE6s>RY&GJ}ya+q6
z45;2EFEGBe;MdKQ$*XA&8~cdHUVd|OPjBeqi0R;7uXe70UE^3=8n)Y2Ii(bm(8nD3
zz(ge&AVK!#2F}X4)lM((=MHxgnB90_-8MdrI&fTMX+W{+@c#1-&CqJ>Nf`ug*u;b)
z7`l?NeV9!31pq%GC2ktp2NEA_H{g-OTu&zjhSu6H`rL8f$;ncAzskQnOq>5{DE)M|
zW<(am_UHZT>n<rd4>sp`%@L>bYKMUWbPTKzD$`A=NGfrVSNv(<93Xp-qoF$dL74%}
z{}Oips7oy~P6%9GC9R->Ez2RP$rnl$1gFWg!T{8J-y5Q>RMDNmRTPt6RUk1ixXMGe
zns<O5YVzL%5ZKKfx=BhqH_c!`-I#UaE48Mq{QaaMpMA#=5$Du>htKcYh{9UoJC;`I
zXsgx*RCEGHaGMD)6~cG(*QKT$q0BEGEr+LO2*5$A9Qequu2r)R^=Jj#cqMxZJ!*se
z++G3f(AhrTr$>^tn)1@=<S{S5&ZBdrKvY041ksfmX{mr4>HVXGqQ)XD2LF47It!e6
ztQ3K)tUsi^4vs*SdRo*3YaCnh2Ev-@OwRnkPjPM<8^pDc=ZlR0iD3Ihl!q3t{!S0m
zBXLe$C>s3~HTHxh9_{VWG8Y8BA9e76*0cba9AVCzKz=I7cTAG*+Wyce`l5PrAk4Ei
zXZ8znY9V))HjRL_2AXcVW_D||%d2Cc4K@NB(qj(pOco?O?3}E*w|=TEH2<x_qmy*p
zf=_GKJC!B+{78>UuxRF@f39eXc(9R{z_>Fl#okf$G2u#T31^~{^iL+cDPk@{JMv3E
zUCm#HNckV{ylLJS4TK3T!0<JTlllg_CJuI_-GRqx5<-i+kNW8`>71hZNke6SEVIeq
z+r8ycR8`?IPg+1b&%0A*4#Ic9-RUdIu=C+=jqB->6e+5W<My45+TZg+*a0ze@*}Nm
zsC%4Y%g84aJbzJM;`}B9iznDsTQ7wav=F_DIxu1}tooa0oOqEQ@@#b*wQZ@ahzB}(
zM>V-Scpv~TZhqt6-C2$=-Fj9$e-o>??0soefs?7)I9$3Z2B^^inPIbAG>CzKg8^ed
zk8!PY8`E7fv{7TtU@3d}%p#1!(v4-a2n&J#=FX7+uf;zR+jKYX{}A<+VR1CU*0{Sn
zi@UqKy9EjEuEE`Hae})9cXtgCJV0<KxI00=<$dmV@89X&nVv40uBua~bGhgC<)YMC
zEgHG?-09qpYO{anSEwYl!m44W!r5fAk|&@K7?VfR;yVBP?OSRd9Hfe2<M3EKhJyzV
zy=naY?<x>Z@FeAVt~rwiMVMV%e>@$$cZ>VfW1+WCG3z#9m-BE#11vUI$Q^0QPg>+9
z9@NX}tp-s&gM)WW<BnO!>obEjl#M0jg(B}ro8J%jn5wR)_FFHN+_kvyWYrgV2UBS8
zMQ|TG%j}f(qcSpnHt7hALJK01b2_U={-?EVCYRQqwl7u5WU(^DMdeGVwKc!dznVLA
z^FAotKaXye^yKAo>uH@+C3Q1{`z3=z>|F5gx2XpVR)j_;DRh=)CzHbBfg_2=w@Vt#
z#g=-P*(nM?P9G8k4Rm<F=X)%=J0*#5DsZ^Y3gINt(athkBzIgq>3>jv&){QhTZZ6|
z4d%Z&j*?vnEph3|=5xIcB*9fZIxI18g9ZhRalMw68}Ps<Bx2?#K^H&?!$#t8s^P25
ze}Pq|Q^4#*r!>n@bW4XMJPW<g3FI!L67GJ)tF*|s;ux1^wTt=1?Dk@{8+r{8K~bnp
z{9`^%Nm)KcWEUb{^a)$Y{5Gh4_ZVB6LvT6`gj(3i={Q;L(dg$3fs9X8)ZH7F^CTWs
zh~D{=i!lVG`D$LBde{){s@zLL;kU&6v}9%|G`1-+?+Sv^s0m5WS~#Y7(h=Sq`+?%Q
z_zkP7)I77J^rfxQH`(~%ct~J-Fe4~e5rhEx_uZptU;Oy<OhlQ>=xQdBizt&hHr3&J
z<0KS?0UiIxX#AAJftcGZgU}@o|FfF*CcaHu*4sZJL!$ZA6ZdF#o{OI5Q7Gf#1L=UY
z<9*M}RE{||)J3wS@2Q>>=_+JjjVbA$(yBSZ4)x&#(h->M?i=9+(hV_c4KZQyTr3Ph
zmY^RwLo@swDE!>$qefrxEo}_uM<Y<-vzeju`Jyc4ixniWU%)5<{oAk^W#TZdPUg49
zC=n2zlF#MCVWCnXL7q?Gnd!aDv`q0?FX!}>&DeZ}>-9Lwg?cLEzl>gCQ$HMOVz#od
zs=x#7wRi=tY`RzS8S2vUeYxcE=Er)M+Wc?px;0wj=J)PE`4NdOvm`T$SAFp#vR8X?
z-L9*duH$#Ghv#y2y%OMM`)~RO)gHfN2bCxZPV|9uM?u>erUm}Fpq0`K2x;e|T!v|l
zB13&~F$%p=UjL~F`{1PV)vlQy29$NSx~!oSBP0rTX=&b;Qa8HNpEBUa!|5@|U~bX8
z-#2d2xZKC%9?KlE(m8j?-yuPbY=Q<_izsQpBwKIP2`v!KjLmoZB=u+=ha7IXsl2r|
zkf_%m`()|p`=M#$!J$-rZY%nfqG=oEgE3o@AJg=HY00U;B)#H!F|(rY|BWXs+*L?9
zz(3F*DupMx<eqsRV(BR({6VcW{u~<B#v%DzNs@hL?dnV*|Cn~r+wP{1Qx>dp9!jwY
z712B>_nVctp#d2fPeG(okceZL5$K3?QqL6~9+{SenZ~+cp&pr|{ay851sj*wt?Ity
zRU<H`boTpg*U`ttK2RpV$G`~DNB1VhfYxWF5k#g!teYwp{C<6<aPl3sS3a-$7I%xT
zDCFM*-kkd$$m6@qA0hyqU&AAjIU|?)jwWPc+@_i6<WaGLu+q=e^)6cV<!xMm#U<%;
z{K_m;NPYwwA5V7VXahppP9I$~=4-!_-#Ty9f0YvM+<sCbm}6QLfiSb`H9fDUbVGje
zeD;GA39AfJEHthkk`=eg6?@aFpw)!(1S9W=Yj1d~2NzHP^-7NocMQ(|&H1sqEc64a
zDwDwn<~c&agO17bD|0_w{9=<(u<aPXe7I7&V#5WX@qqR+L;_7W^m6+p`WOCKuQ!+>
zWO@5v24Ay;>+f!sFQ81!pEtp?hF%M!2iCn>w6aGTo1D#0zLcih(;^>g5QeVl1XQwc
ze}{4n`_vo7w62oLwZPK#ugqb&zH#PW9lAg6WXSDES;wf^Z89PiTF=iMPvMOz<vkBJ
z3-?MQXgZseKRi<}C$%%dH&6+esPYweKz+dD^wXXM-T|F=1&s`P?V}2hp4yNU9^0)#
z2@_ON0tl+Dqu9@`)c#D<uf9ll$2|5W1bp=P|9~7O96#C!R}c6%wrI)3eK?awGU3AB
zT;<Xu&&s*M;0dle#Br&JEY@WpCFy$fpjxu0Pxx0vX~gIZd{m1bNzziSvYj30z}OuV
z&>N-32wbA?%!z=^;IBE4qXykVFrA2a1KYwcXjvSnE@fpS(c)egObOqD<Sj*0ogeY}
z&m`PL2<E%~PDl^51MWBDLS{gLXnWaaHR2g7s;xF8FtO7K6a7t6a1zxVlA?6C`gpBQ
zooj)sqN8XbRa?7U7n8aiMC@h=4<zbI$2*<l{H?nFV+Rabfsu3a>_=SL5KET?(s<ZF
z{A9bbt$CO05vVEf_Y|IhhrLl0J8&0*;x3C<%}058|BLD-^V<w&r!eQGEs+&Q471I!
zqjxwdCFX&>cO+8%+4R=yxioUty-UDGbtX%|ldi$dB~9}9<G*;f!GJZ{p+4{5)RGLj
zfH)7f8;2MO=`mF(#VB%6a~6?R?5@g`f41}X&Y0V0sCUdG^4*a2_m&EKj&=WWhz?87
zM``DNlDYF%E^46Xf#JV+ZR=AIHQxW&5~ePk5B^PNfTKqQeJSQez;n&tU;Ke&=`IT^
zVD%`Q?P4K3k~k}%-26KI0L_d7T1<ko?7LEb0OVo@8|Qi{Prky76GQymy*hCz=vyFc
z73AMbI|?UMofr6s=v(gD!86{Gm*}9TR_j;Bl6Kilm%iHr-^dOs8=Jj)w<-0!b^3{+
z_PXs-#?Y%RfuIuu1h}$t{B4|U1=GiKr?a)viRaz;`KZPvosHsm-3tc~BCmp4+T6~E
z*zT8epcG#4n(jCls4_95E%JU+d-0Tgvy%GTgU-=x<J;qgkmbTwD+#QM*MUCY>00`k
z+nL;x9i~5#^2b^z9as~-M1sPdVtQ12pU$@@#Y+&#UVr-jY)frxJFMDIP@VaoWltg$
z13*Ln1GFg{3rQ{=Z?aj7%1N*nYJ6Z%x)g{}IqSq!_Z9P}rpjSSI$WfZf2^tZs!V8^
z4MM*IurKL<V1mbK29Sj^F{3B8pUhEU_d-j^oHcYzpK%4=P6!12?jzNHg0t6ixeotE
zoz)X>@!6cc0m8mGDmz*BhME`6QUi$pU9jzS5#W&-hd2Qh61wkteYXD{0lYYbZ8t4_
zRX!}|mNK>`7cI5EroC>|A8wfSkGUqZ&EcP%M|fa4@!JK>N4*%^T)=k3{qY}XgW0hO
z3FuT?I_t{<BPscQQ?&e7`cazhAF80p7a?-BKio{cv#PlUy8)`AMk-jaIaSSQ*NPF*
zr&#`P#ZC)bBz^<J&4we(KaYMsxmR=rp-*?+j#SBM>~U?-(UR&6e?otRQ3_U={{<z{
zmKj0db%=x?5@GIx3ui^Sy$Gr)2bUiLG`=MJpP3p-TWe3}mS)$@BuV@3`-4ZXYQ~z!
zX)b71cIMMOtddaB#@x-aNtzS%SRApX#Mj~$&zWbkUYJ#k`u!vRIA4Zl1w~DhG})>^
zmd<+@-|qpJ24(5^&Xr?)wk?+A)1i~P#<C7`s<z*mxm+`7ixDg80-v?=1yX`Nn|U9W
zJjlQC9sN^}O*4~3$4IA6bdr<!(!XH;>->J6a)!JmGwJa->a>;c`8Yt9eW@dUD3MV-
z$m4PCDv$H3OPb)HR1#u0Yy5LrP>rL_aL@j;XKDZ#bE#VZwfg0wY^L_Y;~-IXl%|h_
z3Ga%eB~dB7O&XCDn9++vPp!FZ+`Z+T)Va3e>HmJuT=k8}KzstWX3lF<!yMmv{^m+}
zpbMG{tgjoB-a6JctQVc|(xvRpfp(BYm42vxVQ0NNC6MY_tSNNjxmMYU`D3<i=169r
z<`5>H?@6bTIkFRKD^<6~wEy+_PePxyj5vQi$%aoP@)tEv`HxQh3EB2rrm#L=@m7~D
z?^?h8dh32=iidAl`}!)lvhKi~nU*d%<}Rg>$>gnzzZ}Jv#%)5^t1^aP$(;JfYRndv
zNN55zS$^`a<cvAk4z9&>zMS}37&>=jN@8j;FK_1W(^nJ+Zd|{Va`Kh25GHMM7k<sV
z{LO)vf`V6=PU+`-Oh9)jsq%}K+z$U=h>#WLR8>R1f$tiL#0WE0>8#z{FKA!Troexi
zjw~*YAktZYNTncN11G}jyYX}z1$)&UUnx(;SCrVdcb^Zp8%O{2hdU6J)I^@+mTbXs
zX5W@tdZedHUWH7pP{4A*d*|Od8jSMeFMmLxiHUMx65om#eoib<;|$toM-si4@l|L{
zgT54x9PCQ(On=2O1r^Vv*Sr*hq%#!-j|?nS6e+!arGBF`rgh8Mq6I`1b=0{0z(*_K
zao#giVm`ENHp%X{_f_I*Url@c4|weydC(W|Pb8_Dzeq8}_EFp&?r~h0VncKR-mCF3
zoM_f$`wVnq3teD;l!#q&|KEwy$zDRg_lbqf6GadMxh3nW+t5;D{yI%1@_v<KuK%ka
zqt)EfZVrg1#id<S)U3#Gofn^?ySsb1donRQt!Ftzf-1#C#sFrd_iy=85=ANHpH5{t
z`l4L-;Yi*lB&k2a{{PeRZ5?5=6T&8QE0(TmsM2flO<seGU9`MQ_7_xk6jqTMnGlzK
zjCv9N-p$Y^OSXjofdGAO3Oj-%f3P!Q(T`K_Ho+ggo&@+!LY36+sV<W}f*-R6lZ65S
z8&M8OWb%+TUAq{>W|M=d_$59G4<V+<LFmAA1<JS$n0c=UloVeD{stvpp_%k@>Nczd
zp&vX<6wM5JmgQoxKuu&0%c`v4Kb0ny#l6FzPHWp$RFV5tAi7cRzT$o}Yi9+O{Q?RW
zfoPISZ|>~1jx}sXsz&_)Cbu%i{w*+@G*um~hS-cp*B{?~Y?{!MsLm{uv4lhL+=UU+
z7Q%5=Ql3Ni8!f7f#*d_*V=POo&{$+Gf6Dx>rX71O5Q+rrswi|_ZPu-qzHu*!I=rk=
zp0bYPjr<#U90F4XN(4aCw?m=UVw9B-V#f(mC(nBt5Dgtln)KLh4;|i01r$kkrYh6@
zR0Rl3LyvTL`mq#cdAOp;uMz>M944Errzie7`tIJNNUc{N;;M#~EBbUS<7Yi&Y}V(<
z;h&%P%g#<5{L;xg6rY`#<)dwhTxXX={)}K^?iOQf4SFiNtL}W+!XWR&UB)}rO6*ne
zM$D5le@AaUQt4e0L=#Z@%*sq6jut8L;wJPQTu@%6QfFHEvj16T$=yr8L7|c3qfU4B
zO6J0lx|%Et9pWMSn`(NG?uzGYEC-LYEpRe#;{Sp@)R~v;z{@<AMohiW=dimLU5pT^
zg9U=o&C23L3btl}c&&c7t&iz(L^Spvx%;W8v|kurdJ55-K8nNZfjhhXD*MVtjj#|b
zvdn)(#9jE21irl*1l-{>vim3HzGs=EA&J^GoB@KwFob`Mg<{tAsmO0r-+6e7JUZD6
zRsy!YTB40(G@MFE&D=(98)@P}jigvIgpgyEo-g~E#KrQ5LhmP$C!i)*?jiMn;|eI&
z#vCw8Zj&w(R2e>r<F#Rq2w|xern6i9s6*tdU%b8?_s7mN!4mF9E(;xZzj;J{-G;p0
zBN}fqK^UA>i^iiLZj57t;$*s`)BHn2746pz4c8im+VDek-Q$yc!BO6FBTT}d$eZl9
zq7OBcN>d-S9%N$ZDsq<_0{K4k_KSTW9?_E%v&Z9l|H+YP_1b@$bdiS?A($u|?kTCB
z|Cqox>$ClY{G2Z+K3cy9TD)!Eyb}ywbb?$9z29<}L^k_$>v@PEFNn@g!yfwq5(u4o
zucqmg%1b5dKN{PG@{&%Ev<{oz?rDCz#KqZ>eYaI^D6vo2f1ivD<LsZ4D#Z1e4MbrV
z(09G30kenrh!AkidMlzoygJKqnMH}2yXPjGXL(IO0xExp9Jw+WKz)GCyc9@Yac#ov
zSGpn@$o8Pqez*lUu9itd|8^pbyGO0O$jxM?AA>+&5`rp>pT9nwcIa&gVo0Ot`+4E*
zvkNNm#qos$K`1s3125(L<X^|^AS4^f8l1S?xs`-=tm|8ED8viY-jm#XbILTW$c<rD
zOWIdOVlsrB^VvO$FpP8V;9awyu?l7s*-?oSD{qE#RxSVs8h0%hR%1XXK48lA43Uz3
z#mc#cPR%#<*yW|TK=MY-d~?4jr?%||gW$ZxgFdP=HUlGE3P;JRXd`2&LcMEC^2x<W
zRNFuda_gu~v^9bZ$tlOpF_mw0c<kPvv#5~??IvAr7_0}%UOcl5=Om#|5tFJ&u~5g6
z<|88?d9EInboDmuP&g9o_R#{!>PgfF5^$D`WWTB8)%CL1pVB|UYc=VvDC7{VH|ut-
zhwIsC=&R;H%>^V)|F91#SEsL*i@#h;-@Cj)>{V-K=CK|YQ@#1Lk;SEvzvrF&l5db0
z=}QmdH+-uo_x3Z}ME)pSzu+;x5P=KMf1pQlDYNEj+j<lO3yeE&!p+QMepmp><au%!
zBEGX7P5J~xw?D<28<oT&G>*yy=kZcp{M-7<uK_WN6*C7SF>pVIG$%p$LOrx3oY14l
zf1Z?FV=5DMgW}pPLsg37ZHJkOAtNu}c+Ttnw&%1ULgGWW-`kL{(%V2HfDPF1bMm9K
zV9shrl14Eb6X(67?HX%Y31g&D#~H%Tj6iQ{9#bCaa3xLrekPK?1CBZ)@J)MXuX~be
zdZ<Z!h9R#mVkOy<VFY#yu4okLdOZh#pB&`Q-2f`G{ElCRr&t=h{mBCcX;tA-%IAn|
z_p<!6@qZniI`TQl&G|huQdO}w8-1weP{EaWh~w1O4~w5fZSb5PX@@Hky`6Po^zhAy
znT*`V=D8P}2exm4f73xyEK2+Omk`!Ep-@@Ej&@u}#)%z$J~nC|jfSpq?6D_i+;r#Z
zOkxH3Z@to(BzwzfjM#JwnZhxdi52ngS&3s8p+ii7h;%P1s4-QXr%>jhNTRmqbyv@G
z^m+m>$f+^T7xBKtUSbB2Vg_E}?}+JlpzAFbL=TZ+|CY%@^E|6}6X!{X0OI`wkWxjg
z2M|MLUMnVR1v?Hswg$R?UzI&sa%!4JAV>tvr@lA}9P(0cJ>2g53Yd#F5)X3wXq)zZ
zBAZH1S6ct9GFZP1E?V(kfXwAjPmzQ_9T<`a@g-zD3q1{*)TFP}kUh~2vq?63^RDmX
z-1ToEl)#shHwvE%{{tq6t|5)PaFO%w=rp=sw;vPYv6aoDU%30>e1IN+7-Z$k9=b2b
z8vRWD=(swjMvY>dPe)(5zM4aI*Ce_2oz``yf{-<m&g3a<3Mz6w3YH2-8e=>HZhO+d
zmit|Z(mm1?GlhgQ*~j?qewHf&;uBEDy7nKs7cSG5DL(5oO84Dg&xYEjN<j?8<qHFD
zhZy}t9<3@y?xP2{mwBmn|GH7!80B9{n714S=I8B)0{T0Y`d22OrQaC$6pg%=!H$O-
zfYWE&z9RzO`vw&~rf}s$2hQ*U#~|+FK8KImwY3H=I2`AE?CI+Z+`R<}mhQ69$5}D2
zEtlGq29q;HBfuY*<l`{2Rq8X;SA8)1F&_D)d9Pfv=f7ipq=8zGsm%!k>;e3LbBSe5
zN6=IiM2x>0v9I0S$HqgADLaQNm``_us8vmA;+K#I;XNqvAafz#cM~|NDhvve_HyrG
zkMkKHm#EJq<qiMX`PRvMwI=D={0eh8JW&c#(bN>>Zf`CM;#&@9u3LfK+DcdNFpWP7
z6-3pXTvqp?sOg%+cs7hf&%PFBkch|60uy``?BoyDR0ex$phLjh=~D7Yt}Eof9LvOT
z-di0ws+o}u`aACwu#R33nsbmjsA4^0RcUQ9Q`eRLS%ta<oy_9WzA$7a8t8qDgGK4m
z#FehZ=8-{C`lD#PAAt4**N)H+>Mza2@}6?xMI{A0;a;7lOT7dYNWM%}+bxu8s&+bv
z1&IUEyu^Uiq-0lesLct#c(EMFeOU)ZDZy#1b0UpB;lnSGSKimx@@c*^i$(oz_JkBW
zN|DN+3qb-!<{Cq2GD4-jv__C<W|*XSIZ4q8N4#p4DR!%CX6F`Q;Y#~Mx?dJ$n3yvO
zKuoIUkxeKTuNyO?Hba=<Gt63{^GJDkMb&uHRb)lFQ2_Erw1G>0F#OSbv<Bdngz;(|
zK6vKCrfUsQaJ1sF1rNo%vt80(<^rDX-b2FO>jF|*RdR<*UU-itb|7p}$n**<Q|9J3
z`nsgpzo7KLNj--wpEa<Q0CGr88?BD@dErWzaZN2fPy>VQ&sjVcA!+%S$8i!0F!NM-
z^2(vo8LH=`RCOd>ERyHTt9SzZ(4CKv1>Qabb6_-7L&AVSh|8fg(skj7R1T-sjSDF)
zI}AQNm}wEwxtTFsk>>9^CIe3D7r>}0`G%}O>+gObqrhJ|BJt~|lrfe~cWvQDkxTS9
zP1PX#Qc@5Uiz<n11!IQzH|c3?wd>Y|Q$SXU1;E1US+0Ty>Cl~$WA+Lsj~r8VeZF*K
zj_4L(6Mx+OK&SCjN9=b<Vkr%2y8~>!J!%B%eVh_8y3j`&+(C;ZouSe=-!jye$g5BH
z>x<e1(}A6aW@@2OgvAAgt^haq#r)E+Qj!(YAsIrd5&2|$T}0xGq&mf|4~2!gv@ZoK
z#$DrZihj8dev^bHhUPB8OLuWl1Fzz55$`$|MPo36V!VGbdk+oT!io$N{>my_0`ErA
zl08*I7%hEck+G<21_HPn4U=U`?-Uq$P8i^dsrt<dAq}XV`=`5%Pog0T`c~vyNyWo-
zW{-ESGiLA9@`)I=+)P5YVP}2b^2W$!&M1JxZA>sEm_D-nNzS#H`?fL*!j+`Ri`MAR
zyoZK87<+BBywv02+F(z=iP8wKPrH)ft@e6mj)R?(mA3>qA_|5;zu)Qkd8%OU3hP%Q
zRSZq){wRk8LFWFQo{#MG%pf;Z!XPlDw$~CqYUQ6i!w;e}UV5S=d*5Oqq`gQRw4B8|
zJ*a1PN%-e{O5-1waw3f&2EEejrJm*K(2H*iv^QV3WA*ac_M-2k4GR)FE_3ih-LPed
zJsKLi=8k>A^iyBtI%tZjUg=SqLB!t;!7nFse5oH|BJQJufp_x_7?Vfb{|<WdFQWCi
zZ(9QEFh(z7ji?~hCVg^Zm*1Fq2R(8UIZCV)q`WKA%A|!G<BTzgeypGgQmMv^QfUW{
zd@0Kl5i)hwe;)tk>I7C(kI9WoJbw-Xr9$+<ZGzwJXK;I2k6hKN+D2|dJlsw`*l#g~
zh+h*gHz(=SJfgEYa$N7VgLUG^P6b~E^I^Q+W(7U;{>!@l_jGR4k(OZRoFEplkFE|r
zG}Y)fdR_Z$Ujw4Tl20NAH}Q3(6U7-YUKU|CwvYeSB}RSVm1I_{zxx9B6v_Av-kV1v
z@9|5=VT*s20MQF@Xp!$RO-7?}!|S4=V0#*RC7vQr6nV1_ez&Xhw$G{33|iJ@6eH38
zOyAydlg6GlWX{h6`F2NlxGnT-|7GSjnkmg`*Wh0<XPlA5!Cr3y2v&MQ+KKr`{tUzX
zNhD##Y643%dhdEla#2;hS+oDR3tIAU?H5l!HRIUK-wo=-M>?7)%om(lW0vFFu!JG1
zUhMAqx1VXP=bruM>`63HGmu|h+<s{utRLC8o?DZm8sR^J3r(22fbe!_GSH+WxB5eb
z*-y-vUMCEZ31`{V{3J^UFbAxWZo_Hkj)0{+-_CsyoNc4m@++7$5)H!B<7L?!_i_06
zlBPG^)Pw~o=FdoVCf>yL8_oybH1vy-{S6BFhAOOOGqv@HWP9vc^(}>jg5p*hC6X$r
zDati)vP%jh3BGu_zg#g<4&^H^6ksj~=wx9O-z_^^ldTe_esEI_UW}wXTsrTrjy9x=
zm2G|8jK*t$Y2nI^6OCg?GcQ@Br|z0TV-Tl5Szohpg{8P`)hvA(2{v2IHu6%w<3^Dd
zEsmwC8nr@Wdl%%ZjrREK^mG6zZO!}hS{3HZ9`B(69smnXxxrzUaLEMf<qS$12bXAZ
z`9h7JK83-zMIECQ-A$o1UnZN76wrf#w?lH1chUX3(v~5b^Bc0rmJkfoz*V3s*id<M
zKoe;{?L7;dhhJFE^C=_??Ogs02E`xAn=UgBi>=J->{@0jZI3){*Pqp+s>8;U5nlP_
zd<Z%OT!AlMv;nC-t{m@YY?*0luhGijfD0A`KWZ>YT~yVWrYj{jC?EyJi6GGjYxFk6
z0#`iTmQEWkQqWcYPaUd|KUUoZ=E-sr!_8kGMude&p!bE=X-Sd7iz$bkv5$vRFS3#>
zv_Vvt{<1y5TVqeTnS-wqU*z~hWX8Vr+HvBX*QA2N?6}V`J30SAe12l`jFsqd*$<hR
z7qfILqv6%`%AaZb>TQl@xQukoOn1at@U7rH`Pr(&>}hT4LZTPEvK?H((kLe^wTB!s
z`z8Tu3)DgfPdSLcjbV&B0HHR@vuD#Tx2Kf8oWRZ%B&EjRvI6_|{qi<Y@=;>uLP-;@
zG42uoPL+B&oMZtZ&3=tfObUyHLYy=7F|+9jHm0s+lb?CKo?<xtH^&<aoQhVA-4qaQ
zGxfy4kTKL5;f}rZz2Ktqv=T2mgaHtPCz><~Ng_tsM?~$eJpb{}aVEE-wszr<*2W)>
z$HDH(+Rp2)HpdPAEuq7014Obct)MM<YrP#pxxR@n`6%LU8QMclMgB9UW`w;0IZ$vW
zo44Pq%I34tDcD)6{jUs{V2z<TCaoG)uz&_to&s~GIU#7TkWTI3%Iy8aY}}!|$X<Pc
z?=O9gKe4mnw+6S$eV77tQJ~DdHXAnIXc?jK>;dBXHCJDL{6rV$dU}O2F~*rywC<8c
zHgFIGR=KJ0<~*#sL?3%L08x_>2DPe(>iOz2>tR!Bmuu{CCtu@o29yuT!oCpzUc3L^
z*eUW{k69rG=miy?TE!mPyk+4Xq~9HMo}C~!A`9CGHRO}Q-DZJ~SIZ&_8?p1|VCJU9
z_m?=D2)`oGD_rgd9LiJ^_TK;IIW5k5A9);s?R#&Ec+qh)wl~pe!jGoGG$9$hMJ1(?
zE&=TBbV0fjR|^0kqC_Vt0P@0*wH(&z`*8$#vNSxl{VEiS3|yGoo_Gh5Z*WH}iJWzN
z-bA<c8>^7+{avu93Hcuny)f+XkHOfA9TF7FN4XJqUbx0_5#9s`H$Vt(i1)s_w`LRO
zfyeWmmoxHb3*ox1>P?{|GkdV`G|=ud`gy><IOU%1q|`t`51pbB*;3dP01{xB#7}B4
zijHc<1d1`XYN@^%c1@vQMR7~VZe8BU_Y7*`g9}!AT2jCaam*aTS^B7}c#Jy8h26Nb
zx^Y(k<ES!5{f6<>&MSNh7LNGdB=zo|3lm3r{o~~~Pe?qQGfKk&Y7-pMxO^Idv)vKX
z%U~M2leY}K!t<aP>77Mi5)MCBHvIE}Ug?7KMuEU@re)YC&}4!{+(W_1=E)6D12QnH
z)>!o%a>{zrk1b1yxQ2{mIF5SI)&|zZaG~g<3O1Xf>h+h)9g(Pt`1KQTEM8?M41uh8
znSNkm!3kJ@dGfVZP54b95qS$q=#=QUEnN5!tEWJ%HgGb=9Ka@gHgeCv(1pQT0PCfy
z=Wu&4KiLRjRE2;$0j=`d^7%EZjZBPLs!n(6wx19iR++w!?ICm@E26!8)hvf^H2O-Q
z4xH#cKB-*B*DKdZdPZSJ*X|}%5Sl~5)hg^^NbA7l+prAH!mH;_Lif*C7hq6#J;mt^
zs>sxguA2VPE^OB@a4?0%(SAAD4hz*?I7MWxnfi!zOLXk-7Y;CQ2;2n+s+pu^i;#b%
z!?zF~)V<vnVV;P!hs1#_u?Do#_qGoN2cjLms0UwRGIe4i$gU21w>wB~X=H!Xc;&%z
zqnU&3=z<M)Tc|P0bAMX``k1VGK-q1w0hL1+=;qKrO$f(_V$$MrW<9_>^No2XWSog{
z%*%E44pFwGDV}fHnGE3;m8ZS~rPvd5>`1!-;cWoObwINn-+?QjAeJKrK$!bAPDJ2=
zUs(6B;+=^$e5Z0=^b?k8gox97+YW41X3F@-&%&?M=QzbpNqB0fjnBXFo85K&&y6EH
z^NH^-0C3me?dQNOO|ICY{(*P>x^8$xdXWi5!>9h5+v}!^GB&rjl61Q%z`Xab#K_)!
zaFjcY3JoAifkAqK)^80N#!ExeTer-Iyf&A?l4OZ&*A@Kqb3HbQD<=TAcoOy&+hd&&
zda{j`0}2DC%esM?Ax*We`iXaRJ<NN05(`E6iHuRuxlOBiNm-$cv&+uT5^Z7mJl=Ci
zLZwfqtzxvCdBcZE){U??j#atdaW#8O08tkXvPKzh{YnVD2*L~>)L!3J_D|@6v95Sl
zdkU)h@^SZ4IFW6(guoOWVT3*pzPhkp41yA|mlO~*54)@?L!1j${jWfdYHOag25KNf
zw_gW3(+UTmFaGSW3m4JA(87}x7?a=U-knWghhKb%?wvQVCFaB7YH6{<U(J=lm$y_l
z<O7|p=JB;5BaN}VMnx>hRGUR{;g>tbm~4-ROD68(PXHAVBF|+~0?>kU?HBnEd26{?
zFkyQQle(pzZ$|Mb{!(-jTL-o1qy9y3fbgSWAZ@>;K&R<CWAGMbCoen5J!MlbR82Uc
zQ93_q&kHEm!$($pcSA5DJQDk`rnfmya&13A4{Hwla<vTGz?jFBw-&5$zB1V9ZhEq^
zIy|%ZWx@(kSZT@}X|=KCa5=gz9oqWR*PW!qd#Z}t=61lZ1yRHbkELETh^sf%#N>6)
z07!iA#k1!t|A8_r2obVu6J!Kmqf6<}pMg1Ssck)Pl4GL3EdYB8)qQB9j(q7qHt#$|
zZyYcv-I<$tr!%<4vfU}OO!@exa^=Z9(Q$<$8nl?5)QWIn!rbr42B+g^#f53Sh<1u4
z0=a3WQ+7V(bv|W>Iw>`)IZObvrJ*}Qcu7_*LgK+nx`Q8afXb>nB5ZTz5GOnuVOrfn
zi5?y|?Zp3%_Wr}Zgoib<pR+Kt73u_r@!AolB9txCNy)FzY$|-p#p_aZPsOg5;JAJT
zh592GYE0^CTC|@xTyB7ubZRj7bqeYBpCveT@744`8{^_HyzC5Aw!?u{@Ug%J!mXcE
z)v;6Kx3`@e-4Ts4h|pQMTxoB+z2wqCAqes4{X@HdLU$|-rIAJ7^>q#BDgfwL2yPsv
zP_a(SrVB9Luh)&>WM+?pOFy#X4l|mE{-MVpH&ujs%k2aan`uV42AD2B<0Uz83d9_K
zck<!siyGakZm;dQ_`|WL7}|4zsN1^Ms;IQjPB@d`cv@N$r6cY86tHkWeCO8n^T?i3
zI%I*y5NNbQ!g8DJZJRJALVL>Icd`p3F-gk70M2B(ncG#7-uo>We@U~+{<ln~R$rHi
zW7U1MryWf(%AM&M-u_y=omh3(mL0+NiBwn5STV8xnKr()NQm|pRluL76~8VML%UWb
z<y}}CL0`3Wk4Xw;X9sN}7FG#^p7GJzjdS|ZT9T=CnOyWvm+7YLrJU`8D0CXy3SI>8
za8Y6i%V9VVtO5Aq^*60_**AdV>JCIyl9VP>(+J!p!Gw}I9XDE;UyQ$8{fKNs!W0<e
zvTk81nkpPER3`*@*Uxft>CPfTEh=mX1aM-AFK%m#NNYQJc*W=8^?FUm&Ke@1f#7X&
zW3IZwMcXNy^<E7w4bXpq==Ir4K#e~}TEntJBq!<OqGa1(;2@dU-sitQAh+61pu}W+
zS7Wt_bX}uOHoi4|vtK30FBPwqOgYo<LWs*UO;fU=`XlpDedhm%HoX5#vfq)dQn=B|
zs;}w`9h9X2gso%txlr|L8Cr%fT)DTy#eIiugSc@kg)vVZpG_aRbWTCZE-e%mggYFq
zX$cC-s0gdgqeYu4C!b^w*`~!<;MsW`wXQD|HB)S+3%%3<o#`-J5k06OhHTab>I|ni
z2JiSAYQ@dxYxBUBcT6GM(#L@Y(afP1x^ez)s=0r@<vQT=X}>hkDU?;!U9dna!4I{4
zbv%LFW5^r0uo;BOW)1AOX+nsDGoFhWk)=&Dw#p<!ph{u&ko(Rkz;0$*lB%G=R-$Ob
zMP63>Iax8O&)1AfDtO4L{9H5R@Sv=<>RGkN&wD#cicZ;WEZbF1g9ICd4dHQ{rKRVs
zf=+<}@2&}&4G~5Cqw>U2F>anBnMZXk^8Prb|2a)xA&t>;c~A9p8Crh{GAYH5jtE0c
zeHvUO$$K3S7W#C50h>Ayi6JhDu6sd@tN;y87{Vh~O$rFrhpL7ADO_o&vx8P8!lmC{
zWR;8dakCW+pL!+wDd2@HrUK%!%1?F4AO|f)CiTL)>Tnbcl6&RnC5qX#HHFBt7K$1@
z!30D9F+<J;QXZgS-t=!$62F-3sJBe_uhys5<E)$d-BEG^>I7xW-Qk*wafooDz_4f7
z=t?bmpp!NMg`10UIX44TC*G$%K961OPHSmt!%dn%Ev$1#3K(%2kg+PDtKP0Mp6mQ!
z4DTnB?sqhFgNXY-9l%14JUYeAR|x%;VaD0}QPp*%g8H%H^14ETp5jr$E8PGcTfwPD
zAzi6is+z%Ov0K5~kSqWDpY7w0Jt<tJKSHo>SkQ*4COSn^d6Sj~7K9rdI9J<L;q(_J
zEx1W^kVbIJ@au7Dkg-OhfmS6)l#Pu|KHKLDU-Sv@Gzp}Q&q?8n)N9w0AZu=|RkvO_
ziT}3A?%<%%V{y3L9`WHk9i#F>WGgi9Nt$>+EYx73u+J%-jJhC>OU_F!s-T>TVBMsx
zTtCQGZYi)#4o~5F8t}9bq3P3aKwB*BCp@}Q99mOJh#&Eva@j^tjQu@UB*i&q3GQZ3
z%{ncdYOZ?96&=P;57MuVu%MhZ9OkbkIu2ZI_=82vpBKTUL3#y#L8QeM9wgrXpI)UX
zejka#Ow}X;T`8c0@>wb5&51dc_&z1I<5UZyxEg|YOp4Jf<wK5MnKq$GcYvQRM}OE<
z5|z}!&#&2q4Y3jZxnA(7T}?(>HP{d!ATwya?$HZY>(Da%Kh5>wzQQ(1+_?5Xa~&u<
zjY|4fD*5JbqE5Q1+^PLj4W!N?VPMn_XcP@Q;UQ?p@Zr{tI>s_**lJVi+*12b;-zdD
zCCVD^pFd<oum1|Hyw9TYJNfZxqN*Uhz?qF#`=bsT)Tbyo;98ok2>V~9&C%>SjY40c
zy)R@+QkD69!LK08AZ7-t+glHO`(Cw=aeo(Hu1BXq#wAZ>A#X=O=w1ZUW+>86?bEA+
z{y)1|SYM2a?))75#sgt%czW@!(V`7{Cc&p|5NL239_yJ>4#meVIW5aK?-q=F`OkEM
z!~sEYnmj^czkB86>zUqNTSeb)U+8*-q*6+Y05b9&oY*lIC%WFwV0LwtC7_$rLLs3o
zzd|vQ#Lr(1O2P0%FZfKM;kc`P8Y?e^P7_mR*+FZNXG&$>r}u&a&BP9Z5nBs1e1rmJ
z-z)Vy3h8T}eB(Y{(`?JIZ07lmNZSKjkwHASU3O>ffA@z77rkE*ncqg~QfCopL=BhF
zq))YJRK;2)KT8d#o{$=kxBro6kN~gQqh$f|L?G{j50SCtvGpN(a(>DIELyaSs4tKG
zT;YCnRkyGG0<u~vpvPMifOFMxA!o`Le6Al6c$fE-+<jLXp4c9D?-PWQ`3h3ps=;|l
zoz{P977C}iB&4{biux1gK+TC!{*T-u5q^{!^N_GK0@Zc4P`yTgslfLtf6zi38&wp2
zdS)v)V?#p4Fq@Tl0bOd4mMeHID1ihDFa~*SCS&5F%3{!in}ML?S@a@QE&iJY5_s8d
zmO)Eb*5@p`p9D*#Wh%<@bMYfFJ<hcrKW`c?>We-WhrX|Y6+=VUVa})J>dl?~b6CcR
z<sS7XT?$>yv*H1Qq#yF5>l%*+T?)B;BTKSuW6!--DILg|5WoeXSnIIWIavi2Cp_0z
zm|NTH7VYAV)ssG5xH5Ua+gBAbr3Uj5k7Wl`87?)qY>Es`%1<qnfV+$0`kM6|8M%;E
zBjH~VqXm+@ptP}V&*EgarB-NkwA$)I(>T5S{Nn?j?7Rl*2Fu(=&q46<ozUJN@%tl4
zR7~6lJp%#dEr%WRM)ktz-{|yBx$gxL>h8&dLOo(No8NlI&)_=S+7doOjFyh&t5V|0
zg07Rzf&*G2N0G84E5A@%oy@ICaO&2)T^tuWrz*+pE9`nYW$ZP2KwW?ID8g*}HEN<i
zisFU+3&^}3<a(7p+7YbGF$%lQl+<JH(mXiz6TyI21NXjlf0Y)@d2H;U#|qgJ68u4v
z1m+#is~moLtCseA=suTG#kHG6ZPC2r3+e8b|9UIJ3Hb@@uYc?fY+9s$X;wkS`yE-*
z?6FEXWVD4+9<0({=djzF%uINiBF*t@{h!CM%glk%6O3+~S9};TG95XM{pbW#QXV71
zzbql~iMgA9el(<bqze3sUzG|Vg1tXQ4(CIzuiT-Z2@ZxMlDUH}`|faIX0MvoonRHv
zIpxPT%$vTlkPerF(hdNNl}nuvd%#~MzbxJ)yGTH2Q`l-FM?Z;e<TWg5lRr}rpV)nN
zGk6KYZ7T*|><KdXVLpocuUoFmY_5MPR%OKUvX_i=tgvH7Q}yZub3mbrmNc_RVV%@-
z3RP-ZweI<X<reH|hf;TIR~alUHR8v|r^}Z6;YZ)Pw^Hdn){D4qtV#RVI~N;0bLP)|
zFTP`gWb&2vm=!G_tNj+Z1{?5zdo;QWV4B%yCSh<zyrxWhzcmS+S6r-N_vW!Uq@-K3
z@aJqrOz1A7y&~JlruyVuJ9DfJyL#&6&l$hWw4F{H7iOo6;Sk<%TajSV)^EoqF05zR
ztz||gv{|ofq~gFQj#dWqz$sv)V+=!nQnY4T5u<PLE9Kw~Y>ncVyuIb28(0Q>>&$&L
zvTv0L!D9XA-Ij|%wM%$xb~+qOw}04w`T#`lS?X!P5t2Mf-=`M8h4|wlMc9&xeX&98
zU_^A$MYAFIYnHFNm=c0wHIyEm*CewRt=|ITu}EeV$ktO!R_IiqgCQyKl9&{_lUj_p
zhSVQrlB<b+8HOp+JZM+L8Me~(B`3(&YrgI(>;G{>9ZMC1GMtH&f#u?(Tc5Xo`ui8e
zex(VSAM6cwJdG@RJPjN{fOp~gyJVPjTbUdhMyB9><E(I=xob!?DjXDo4Yh|CI3_93
z&HL;!bjI#jtaD+Egzip;Zwzx72I|2gqks3t)*RN(?`qK2ODuUry0IuXisg&6A6gic
zPEbWUJ!=#VMe2+T;VqYXTl%deSQ4aqVd}W^yCU!kkr1&puOea2@R&EIjzaPf`A-z#
zh;$;MK8#;FKoTZJ({v6R+$$1)i@c3v@LZJEh0qizlc(ep?V-RfwQey-j<X}ogs|>R
zno*!)4(Fzq2mlxku7604{6J8To`BGW5l8ZevHvZCJ<TvjiYX~yl2NQ#V~cg3&{Pm{
zC~MfCR@p7OpP+!IHDvF|h&mUE2xLvdjGM8Bu|4GKp3Fiw9eaj7*4cT4(f|gR&)N5L
zpovb;$!~`?sqcF0+zNI&R6g)|WJhuvWaXaTPw=Vx7RSzH|6H?E6oAgniabm53`L)l
zd|C--wXUswtZ^P9noDtr$7{H(xQJJ){z5yLU7(pE1}NW}R5nMc>87d4%PlYsF>I=Q
zRIKwD)eX)cE%zZo7C<y(C9>j^WTsgxKYFD7>TqQ{Wvf%(iKhd_x<ut=Mz{92zS>MU
z-2{S(-D*iPZF@2)JA+|yUS8$vnnxVEiruA#C7Ru;P@*vQhGwgn;VkDLH0%y0IOA?$
ze5p+%k0d%ZO948>pmEf%D0rM82?j{;VezOI4%doq07=0w@TpLCOlU~OqQC!iOkB-_
z<dR*eRd(OhzJwhW`t>SWEXoF78Q8`G`s8p1AF*AB8~tKXsfL~4whFz;n%N`^CE=+B
z%!6PMPUc5fNy0goLIYz9d{v||*cL^Piy1Mg(awtf67kSZhH&BH$9Oohb|U^X2C0(n
zOEeWioMZLo<&MQ>tJV4jGsJX<MT}rwqme4AhR9_RA->#dK&jRUcNYiGV|B!@vo4N<
z6@s0kP&5CCN>rs}mPNtRJizt#+kxQj`B{Q$hFX)6ry5nMu?q0_Lk@#Gfm$#55(QvU
zEua){e~<clii`8RX%_R1^Ory5tHOGOGy4dof?*;R;5$9g1i!EFc+gxf<b`}xF|yxQ
zeh+3}zB);ySyYu2P4yR?9q`U&Q*r%{t%AM(&CpWAtb-`!kAxicTw^McvTEw6a`eO;
zCIv)YMWE$8Y(lF+F2A;y2g1J#+BRLW9<3_NShTImxSGvdjLR>lbJKUWB#km_$kMWq
zV(1~{(m*zk&?*<@lM{QX|1>2k>C5x*$qF1jwRN3V&({HoQe+$mywK%lii8_~*a`NA
zibunG+iF`(C@3s8j4;C^dtz2?P=tr<Kg&e=4ibNJVYVi^5_!-yc{F-bD)c5CI*v1i
zXgewd82GT5CF2kQIlM<l684fCCLh^T1U_RqudPG~7a_YMp6E78g)kg)A_e?nqbwWU
z0IV2L6b^O3*9l7e%}}frE)G#R(h^RQ(?YjD&lp@F3LkFYB#(Y_&=2l!X{$^IJi4_s
zpieL80ou;XIWxprR4C&R96g0_CdLcxf`WJqy72peHw}zgD)9__4Sp)2347sKNjLyn
zR4)k27#Xgf{!O38`d83oxvOpwvZAyoCbUDIxnKA{Mr934a#;l+D_)2`+GDt-Lipet
z{Nkpu8v`p7_q4D)Z`eypbql)Xq#GBesUivdAmlLg66t;canCGPY(odf;+d#KAzg?t
zxHr|ykAz$32KD)8W8~}xJ$y1|?F>vazc6n2C0GI$xUsKU;~|R)oPF3PEsM@nDnB>%
zM1uT6X0Th0$T$;&5NoNt7=!T!roW8i4x4J5SHH;kUmbzjT$c?tNr72;-vAI#+)^xJ
zICMp_#S^a|Fs-;ySBg(zD3;ZfDW9bdV({2TmXT{m_$sQJu!9}nEx!;@YzWbJN!2dE
zr&5uTCuB@~ACaZ;A;+`e&pqw*7Dk1c+b}|$#E!wUkr45ii0sV5T4lwvR71yOF=KSD
zt3Rd)BcY}#hz>G%Up_|#7YKV3&kl)>6U4;4nh6F#AF3^4i3UA@@BTtFF?m{k00IDj
zqexZazi^XG-&f$Js&mTj%w?@J0ZMsX%C<l1Z7s_I{gDUIoLb{odjk*wpwQhD24Wo7
zGC!jvkrKWc2V@O|Tn^E6DQfymVtUR5XaH8|a4-wRiVE-#JP>F%dACtCN)}VIfzGsi
ztOT){1<e^*5;`gg;^kVj>qZfSte<Fw@-?3oXZB0*h%EMlita3|I&|SAb$OU>4V^;*
z=qP5fBa5z&95iDDaR?|yqa1&5)?aG#3+0)s5Nc#mM9LIVRF1Hd#$>7OigG!!;*uG&
zA&W?uRQ+{h?zkqE!Xd0X?_z=(1Sw)@WPVDiqtvYhToiBJ0Oi1Dd1z|se((Mq0@*~?
z?6-+v9@Cn12(w1x-vI1(mc3a+69){Ac_H~IGPTEp@~b<T(UmNPMCf}l3_LHbbim&E
zA)fYaqEM$Jt2jX^o2f&<JaX+r&%P2;-2t|fEr+LDX4Tf@ml^3qwuVYi8tDd&;2O1!
zRWj|9MQ<3hSy?m5%sF-fxsxnOz?fTlb+OT?NBj4ystdz|SkEwW>2ApdN|vT{qgj|;
z{X8tiPGy5uQLsJe;W0|Y1F@*PnaeWH6o_Oz{H*+v>m<$(WrVm83_FJY2L7=3#N#0%
zEmuSJS7n!OVQMw{Y9?@|c)VuZ40Etda(_S3qdar)fT(#~4LSlVI}NmPi6c@oBP=QI
zo^YU83?L4Gp>u>1EYvKUe0Js+qY!ixIr|Dm2XUUtnA7GKGy(;O&HFX-xJngEm>Yp}
z*kPhk1P!fdbmZv&a{(X#<N({SzaSearlDlQqLZ$ZkdP<M=0Z=&;9<~M2W4OpAIP3y
z(IJ{Ng<b_ls7F8Lkv3^vRlkKc0+=D-0g%hvEaI3X7voYlE+5K<pT9`>Lg`@vq_)jU
zkZMxs<}I|iL&MQb9dHU=h7M&Biy%tiC-cOzWo*{~wRd|VKBG^<uMz!>nUq4aC??au
zhViBm#i-LrwB!;LSqvE5zpxNjuo3Olvg8-q>k>wiJ72?Ju!m)hl3k!5*2$b%i7u<y
zfBm#<GH$gSe?*#0k?P11e#kD0LI<zA{D3-#oO^_6(-Cb85bbnmO%L&Dx=+gPB&(Cl
zb35~^Y+xpDh%d}0=T9C{=+1T!r5SAnZ3=uRkfrGs@uEQqC73>Q9b8kkke2U+`scnA
z*?75r)IB!A-pm^d+&o*8Vm3w$_w_#9ldLQC?!Yt_0`5z%bkCex(%DVncgyo;mO-o~
z3a%_{oud~p>e)tnp?=3Vi1Is|^X-YNRodgG%NEO1b8V}(M>ZHLHuY^Mhr6yhJeSym
zbvY|;&naT4S3A0>b%em|mfa8ZTSRjQ!zXB-?5Zor+bp~lcjJ&e?;|nriw!%H4497p
zOUf8l*^M4|HNBP)vrw`CG*d!wX{U1qSq-=S{2M(-OnhqY8S5hf*|?Nfa^T;=&~U|z
zL!oa0^2T2f#hyxy14(A-gBGQx{$e6l8;3;8p(Vo2h6EK5qYR}eVwGfc2i@K)!Y|pq
z(<f>Jx%gCJs<`3Gww&2FN8iN@g3&D;P%`?(y`iaj0VR>YD#@r7#LpJ|))wK-$zCnO
z_>h|wd9$<XP=yE0aiT&9b_Q{I;Y0`ll<l<u98oCA_JCPl@WP;=XLu1RAT)%cCH$>{
zK93l{MW0tvMxgWuOnq#9noIyW16&e4CK+5$5-ldEkU97l4uYtWBFt3oA#l(_I`e6#
z6YfDKV+lfS_&8(zcu!<KG#_gl-Y;B^#j)h7aBxY~7*Ml79c+mNmx{J$XekayHu4l4
z6yus}_v4pwOe|vNO(2B2lx<%W6w-Hq1s+W^A$8j<X+QVokg(13i%keVb0W91Dk-Yl
zKa0gQv7T#)zpmSaRO4@dsz6F-Bk<nveB~hXuM%iXF)Jt5gM&{|7_}0jon5;&9U9`<
z<)Q)dLc5<{Xy4Dk2DqHSj6H+^?gxqO#V3SUlZn(C!*E%uBPz7r6r`87zx3O78orTe
zpmH<7WPizLxX$lwaDd*ZZY_k3Hn%R_n}TAk8fQ34gOtK>X|(eE#LyWH?q(gEDAiI%
z3D`^go4uQ&n_4<;M8<lk*bd6sflPY5TY!Hih2&k+8X4TrPo;5~o!ug<tYy8E;eJ>$
z0tVJ@nzpOzAh(KXM1e$kk<C!Y3mF5<vJee08NqSGjyOP;7@0_zMLcJYNd%%E*%Re=
z!0qKqg!PlZ;thiV_fr?OE1Jy00Z5~bSEFNCiur-hxydf;E~Bu6+o(ps`SA@Sk{{Pm
z3=+^vQA?6DTFcP72Ed7*hD@ZaBDciO4jUU$if^M91}b5S^D^hts|;s9mnEF?{E|dy
zCL;&9wNC>OK_4-U9)J7k7&-6-B@F5)7A3`CJoIBe<Z;oj85brYWK=m-ZdA~{;tEi!
ze2T#S)J5}_2(sy)q$y-@03*>XIJd8_T?a(D^P*;<@G&IAkoNkE-@tSS!T*L`K@Li$
zMh*K1Jy0-(9O0H^4}v>@t$=S)_p76E<lrc;|2AA<u_g^^XPM5h-!;+5`s>R2<;TtO
zegHJxPbvfqdWbkctwhvIaikm<hwsr(h-}{2#b%u!joOhoXjhms5ZMv24~rujh*~+{
z<rD<phh7CI&@`)kIzD!PQT}{^HP?c4TL@f;zglPQn>d64rLZl%plit;!~A876>Z#E
z*f83~mP@f$+x>IWnDAT;d^yBcJJWgQ#<eCSdwnL*2xX|PS6;=IEcoS)N~KGB-)XmO
zr`d*v<$B)Y&9t2QCLZ_$-}pyQU%U83fMOMFA;o@Md{ok%9p{x77wBVu=j?MKsF%yQ
zUtGU1Vq2WwWIC}wc2uQ`iTB&T59{x6u{Dm0X65-D1^o8gicUgi5uW+sJ5oVlK@;L~
zsl^$(5M9aqnSFXBo7FI_Nx0+oSQVCpoX}Rz95uZOYf&*?ts1i|Hr7*o5&-Y$4-@xT
zqh6dfF5(#xm*J{&H!7<lV1TmZgAn?Oq{Xqe@@@P7k@b~vQFdRn(j_tEP!6GlfOHNF
zJ%ogcba$6@cejXiiGWCqw6t`0H%NDP-N*O$zxT`i!WZh9v(MRQ@3q%jkFZwr+7Pw1
zZPsL^v5?H(!LOzR^Pm{g{YdZ$k>iq0E2Jg&&BL_|{eEH~UCvJu<m;$UzdqHr4<7pd
z(#o?M$3i`SDdt)sg;gZJ9D%Kv6j3Ve&aXAuIrKRML`#=$a$&R0nP3^{Q1#P|V0T)9
zdZfEiw<PtY6GIstd~SSF!@5xOE9XD~!W_k8#rd!zjdDJy@pUIV>T&G0mXXO?6#Ix&
zjg$xr{v{412>}3?IRwY(0WSC;7Lj+KPPlS!kV9~oEKOFJLfj%GqfjWQ>PAaHnz3Q?
zB0NwU`lSE;I4Ix(fBczsBip<qBLwF})EY{g@WlJ;jW~#Y>Uv<K;lIH1z#AvQoZN#P
z$=zo0bBAFZ{Y5lvvI6e!$%+LTdr~we=W%rAQe|rGpPdk~a2rVp>R%CE91cT!oYb%K
z=Be5KK)<m>B*^g<oemgV|1h66u1B#{eGA}q21E4kd=B62>Q_WQIG?Ko__)#hnIKkB
zpQQRl##H_2!HXEYN%-~*z=lD{!ia;4VP6|kNBBddZZ-c{8D?LZE;}bMDoM5&Dr;Q8
zyxaVI=EB|9hR5G<S7}XLH1%8DpbUGgBrE14pY#c4I{s8jo?}oc<R73T<Keq0NI5O2
zw`vNHW>V48qZNjpr52yv?C&eOmA3u&ye9|q7H630X`1zjhOfmR{~@*#IfDSeDk_2H
zYLTV!E1r7rYUJ6Sm|W#>oDH@X{l2qs;y0auLCYeK(m|2D$p5BWbp9T;y=}A<dcS*A
zNlW*=ZC;fY_qj49(7}aJy~&87Fqpig<XPHJX4qpi4?J?emUV1g^GW$&Y^j3i^X`%Q
z%mPJY3ptv=gi@TT_Nltz7sR6$!L5vbVck7tV3n}7AAV;BRKK>ZKf0q*p8hB%Ig*FC
z?8ea9$gFo)kmY0lM}zDLS~Y9sZ3WeDjT0SU+~;^iyXh&Xa%pCMh#(!`_>ibub1kFk
z7#T*Ng_wiWZ$qEKa@=y$+29-iER}zGJDY4Fcm3kz+w-LEeMzA7oVF2bpj4RpOAV4W
z>KK|Gtcjtt@S;@^!4Wc__vCy2VG{k{I{9FJl+nxCHi<+}W-q*V|BAlJSkx2S6$SKk
zu&nv!q)(H<Y14DluYUdq+e!m)FIzaV#cBnKi=gUFG7-*<rXh~V$K;wMh#E4a(p$UK
zngyAyQpghFmIgj)@R_Tz8ZJfEl$X_!?8ByK`#N{V<9FZrSI)5)k+}$e^!sT;mW>-?
zRs46pZZV}AIFHTgk+qc%Y8OsI%bFXYV6VZI9hr3Nz$yZ3nF-7YEC$Bg8FQMx-6Nc{
za=xX^uLr7)YMF?6OZpTMN^NdU3X{WMan&*pY3f?!Ze|y*X@%pDj%oTVC()bJYNuyh
zrBqkGDF~+z%EHMY*LuB-Pi?WjVMLc)MD(XkTQ4ES0<L@bh_1+aZxbPpDl44LQY2rq
z#pd+PFNUhV?+!OS8pX~;!kyILaikqwJd3C=GEXWTbYA!Nxta_+9iqI7GM0{ZpEG~W
z{cmM)F~$7B%rPp?4zP=0$q7+{d_p4n5QT)rXoU~VQ23422N1pOHjZjAkPlXH|C?NE
ztVtGAQUw?E-5UzLS~x`~v3D9Jq|aPpejV~+sb)l=Rt-(FtBykdf@qqph!yujhI#h`
zB}yR9eit9M9mi*=MS?yw<k`q^<O!bhmA)Sw^LK+;2qmg#l5^T;ly;X??U)B_oYsS8
z>v^_;iC9a_(_`oq5gp4xay`1+D1LFXICrB}C8-LgL2)iDG9(YxyRro7#m2H@=GV^6
z6CO4y`MFs9LYJy2bH*yP7q06Ve^bzI_ASce9fKP8D1seY4pYBgxPGc?r}h9F{(X65
zwuY*ytu|#lt`Yea51(hAXZ%{Jyc3KHqKg9~M&i%gzI+`Q5+G+Oz@1!hNRLxsj7wvs
zR4Q-YHkbHwptT5?HHFOS{7N6T{oGKW*>B1;uvg+*7LitJ!Ed3<GhaZ{*$p*9DLt=}
zj}1>~QiDu=$OXD*ELf7pziH4jpox-uqX`#jEPg%kD1AChr^qE3_)a0_9q>|Z5Iyxa
zrRw*8D~|^Z>N|ApN((<jHcGdMGpMRMWF4$F{1m@G8+!U&UyI5QmJ<hQQJc{EXEzyz
zP<oSHlshOA(yMiREOuECPRM-~#CwpbcJ^WCK}KwE8ar^B$KFaS9>@<f!zacc9RlPi
zYG0PT_r0}N8&-;s02O=g&4bUEDQn~I?CROK!6}o<gmkA&{K)=b^1-jr$aIiqmEpVq
zo4He!YikT&m2TGV<8x&3tOheH@(k{S1`dI0t1^aDFVOJDAF=4ZMp`o`dkh{pG3S_0
zchSs$cPjleB%|mtVoE(G3htgn2L}CbpUM?{8v9wBEa^!kgMexOW0@M;JIn!c)PZmi
zs{r68L19zL&}E}hvWF2z2R{%0u}p}>-`4DC=#*Kla?|2UNY7zUvbUEPAA!J|EpVke
zaHwoRAo4PI-gNLs24&uD$WOkb$80du_jMFu)2IvnY?D@$>hD32yLe{L?XrLQo8bG#
zOzeOI9qnxmqOee0yCaTou4yfUfL^Ieeqh(`!Uk3Iw}q;GWeuNK6NyebbH5yfKQTMm
zQRWmaGSn+QtFrMYN|%Aai^30Y7<ptyjz++ayT%=dH(Wmprx^@m`bPA|hmH>3Xluvi
zWJ@M;tC&5>pfA&EJs*w5x;#q@@0zP;z!vg#D)miaLEbU$D5L+<6;3U&+YIag0v=Qv
zv3-0~%cTm_wp!V*57Flpa;3cz5uSOV*Rl5MFnHjU<xkkZYvmPiO2bX*1{L`sxm<Ol
zc=c+hi8ts6+a#RGt@|d`uLnZ&c{ke!IL807h|(FN7iEib<qN)~H$<pkYhqQr%jSS$
zOr@75m|6ou-ycn(?gUKHF(wuU0x(;7MF>WVm*xsqHj*zo+tlC}<B$YR7&8O>?|w9C
zZe~ZgrcZR)-d!tG>}A@%@d0n16Z!DJ`xEZr<%8RS8K^yf1eEoxEoOIptl4IW_GEZM
zW-$L-3SRA4KA1x4ItvLWWLxf;U`^GD<Ke`8EG+Y}Hc=2c&?6dvIKIim{g^cOe7I$+
zGV1)>6%BQG)p>M^l)WUU)J6UatS5vh9nMloInlW_mG8Ht{@8dZNt3w|Q=DL7FfFwh
zW^pwMKEjlacGhnx0?IOCWZm#*sNq4YR2S@e5<4ft(uAPb-<^mf%N6@&!%ydH-u)z2
zp`h2Dsr}i99Z(l3R@~8}Q7+bg&-qu?qSLir-Q`#af(Qp8_BPynT?1J$sa9)gN9U<5
zX(zWH%KV_Gwe3AW@MMFitDEli-$j7(JiR;yvH3B7dsU=b0u{p7`BIPS-I4h?S1)OS
z-=*nUV8?Xj+WbmS*TCoq)ux70{J;{<F(uZRr44XUu0VQc?*=ag0E2=JA!7SpQfMLm
zAmrP+;y=w@-}V#VhOE%gJj|xpN?24ExXS?8t17XoCy?I%qxfJF&~IRA)t}+<)U!!9
zfiRtZ^7KYkuR8>CX%ln)eFi7_SGr*mAeR!->&@${cgXsq%S5S`6Pc{GSO0<44^Smi
zk{p@vm+<Kpy0$~vrt#=)wbup5CQzjgc=KQNcaI|QwQb4Ow$y$jd6mY2ioG}cn#{5(
zSN`%lw%*!*EK<S_JI&n2Xrd`4a@m$j{g+_=!=xN+@)bmR;aWxN4dWC|jGXB73io6#
z1yzj>%}a-uMW<#24rh32a<+a2*$vUM{C7GF-5eDChfX-I^8B02r;ST-HwqtDsdK{<
z4C1I&p}@yA_)0?)k?{EH?d+vxg3-S5OVL&e`V&mmnPFLGZg^d9@EADt)kA4oE<p+o
zW|e%jI~&b~Pz%j@?ikoonf>}<^+1LNsQls?+Mwq12gV(7mkopC|2D8)zv_+|KC-9y
z+`eV1%Uhrbq%k&9wcjJfI4b+p3`jY^L2ww5-9EGr#JroMI+%501WKEIalsk3R;UDs
zzV(p)|2<%M!k7SVnFcLYnarADU#zM0o>=TUyyio5AKrViec`uSCBX9mY0ni^>lM6l
z3hcOc&$A7!Cg$Xt&;FKr$qOZ21RfE6$O80*xs-gr+T6}zvfyghn~$(b&v1OlJMguv
z(;1!(`rBf2`X4DdK57?s1rNTRA<DFckHJ7f@nAQ2{u)BRdHn<0g^B8M$W>NezsrkE
zwC`uHGW=S(a?zMZQCY;frRVIB7hhp(kz0@LV;t<5!_g2iS-p0?$N;#dUoar4jQ(RH
z#nJ-{zWz6#IAU^NZS#Z2&x$kIl*!lma3<{t7NhTLBKfcX&G-eRb5xBwECrR;cv&=D
z)9e}7n2o+CZeujq0QuX_q4|)1*Y)@8;#Gt1>!%g^k-@+!HB5fs`TunT(CfU?=1@I*
zrlR4d20!5U8txD()%d3ilgtWc+I6nWVZe@$Ntm75`zeAQz(62dv38{Tc60nQnq;uk
z--p*BA0prkTvc{wOflcNJ{q=)Nbpd7dvMd0>nFu~U-RyJcebvwHk`pd?yb+Zd7*jk
zJNpOeL4;Mce34B61EckgmEGX3id?i5#+FK;PvrUK<vV?4{1|)XC|5ED_zy6}$Bv$o
z7iiU8xD3V%Pn*L%>o2oF41?F2@5TR7(*ohqsN^iQE;vG*$l6bauImRbDT;xti1LJ+
zx>iifFI_z;`A#nUzx!Y(dowWwOZRQxRBo(%GzL}0(<UoHCJH6Hi$y2vH48EGj25$j
z;<dfzE<)T!-`gr;fRd7h$gI+B8wg@r(+`Yd(;1$s_m<J9#VQ|diIES)41S0znZNKM
zpssr=n}EbokYcUwbMNtu$JyPm*itHw%Pf0y9W+~=I4>YWai6l&pA@dje0*;y`fqkr
z?4dD$7GdX{_d%{mCs+AlI;JXBE~$>VAR-|bjaUwj=JACZkvg_(9Fx#C<Jza16qhgy
zWAMDDy%~-g&P9`aC{WDoS>GmFt`8Wi3$&zik)$O+{kIx32bkE%W`k3-+a0o%2aXhe
zd{`A3HFdM@N*u^dLvJg0{CR+@A7*4yk1{FOO)z4mHSljKGKl_LigKilij@n>+R;c;
z9NmJf#oqgua5Mf-t+u+U{L8`KGv0Z&j`3S>Y(*yP{4jw|?1gcI0f$`i*CG4Y)1wir
zO}{PgoYMlNqFr_CEN*LOdwbkN_2z_)T)mX^@xyK{vNMxGsPRh3uBK#fNv6IqjcdGS
z2~NoWl<H>9U(of&j{tp<>g(y7MEq<&!XrNNG&}U3NHX`_YVTE!S1wY$1qW-+YY@n{
zsHNlGo$FJRhpG3s>$Ub@L_52$r0u<V4;s!sCB(gl444|d98J$}23v<53q)3)U3@5S
zi=CmpKLia;pMCd^R=_WKmL*1MRjV3vcBVCz^;*$VZB~_XxWoKu$62e~BE@J_HuH~s
zfoey%lmvB9tk_66sC4D^_KaX@HbAkX<GuPT2)ZoQu8mMD%a|?sz@3i<p%<||;rv@W
zjKbe!7nc6`7;3xL2+Pvs)=o-Ad}W?e;ZC+9QC_r$AnRGqf7B40=6{tljy)#UL;pN9
z$76Z+-48XUxp8yN2sJ)}?QC3IaV>?a=^zI?)gzh-M5>U@=3wMoRz7v!M0NhI5K+>`
zmgw)o;GhKejL5o-5d!fRV|srw3VNk-trkO^QVm#T!i1#|$uKy8co=~K8!k=NNn^`1
z7`K}FO{k)-oeE{+R^x}pKHJ0w?(%j!fy)(<?M(9*iyveCe9j@~%Psfg&J>tmL?5SA
z3jESD7n(N7Zx0py_!>kC5t6_BHePKJVSYA4>4`sa{`&YhW_Ei<?^jqoY@fpSDx{Gv
z!#qJQ*bev4`kK99$mNWooGy{C>s`}%dW(n>lvVc+1*GtUM9Sn(>E80$63RpG_&35E
zJ@q`-NOU`rPd8_wdmz3Bm-B#z@eoFcxR1{T1K*|)C)aOhX49D+&(DLFrd?d)GbH22
zAGdeZ3#tl0SWCTou_7sYeha&ykv?bj(4A(+Ec@SfE=Xxwg-xf=gf4zOP|E5^M{Cd)
z9-?6ds6~JnrMlo#Z>YeLDt`J87;k+)qs2E}Aot8}e*q|u;uI&XKcf<U+2f>(;hs>~
z_?8slmXK_IyR@WbGCKH3%{gWn02?5Vv6<16`FkBw1G1Z|*gv#h`*~Iu;h%-tb8^R2
z<^@3mrK-tM6`M=O$;Hpv196-jm(#6p?e2M*NY)fO5#oXqGvzYHynAHyJTw(7XFj9g
z@>1T%L%<jTqDS*(uL){AH|F+Fto|fIC<T}B7F|A5n#WBvd0<6}_$6lvB~)K#Xz|^0
z(%_dy<uu}P-Kb_AK1*a}?YIi>n2r$CWY5wRrYzEtv-y$9-_o+`L(l(gd)nLAaO%+8
z6mqTlW(fqKVg!pKYHYcAcngZxMl<%wB&bf(aHPOhnWysDu34Qvr@b0%q{e8BCtQSx
zktp`n#pn{0Eia^<ds8j4Wb{|)c<beT<zCH*Zk#ZEVo@GC(si4_RY!YD2j=KdKN&WJ
zB5i)WFpqU&cr%LFm>W*;0ZxB`Olt8tC%dfq;KiI!8sQGjdNPc^fP42_V$h|=t19WL
z2G$(rvz@2gSGMG(sWtW23q;%TO|HbH$mmzXk5mu4!*L^9OD^c09$^i93yx1j&KIRu
z7VL`ISEobQ%1sYd=!9l(i0FHpS=qG80~so$Bc*Pcpb2>zehLPBHFne_cVFTXc94jF
z5m|WZTs|$yzTM=XyYM+KBHM|yKD4Osq9I7w@}A<G0Kim>(G)>_0b#KbcS--$axy-f
z^tz%Up+JB2j_~_G7S$kM@ADhB<F+uhKr}0X*f_Hej5YsI)HJ5!Fp-9Jt4Ww^@k~ux
zb4~xaO}4r5VWnLB<c!dkdPy)Y92}F3&GQFxZ6?8-f+PEKu8`Ac`U4|$;o#KbvKiP;
z!vEV&A^zU#;?L4I2AzHG4`%QTi&efn_+~|C#SHEzf5Ji-lW_j6O4f?ng3B&xMJyG3
z9VYm6;pr2?$;1nPJ<6fVO<prWQ263&Q!g)ZUvM-XlZ1!;Xl8Hee0Q(H+mhj`wO*lu
zn?;&o-OXygmbLEMcn`}(mu{QJmBp$2jvmiBJp-bbIIeg$$|V{+an1LSm6Wc?d2OHE
znsqnA?ueWL8v-WOtMI#)CpZ3wlib3C&qa3>sA8UCIdP98VMpbD&*k2Tb9gN5ige=H
zH5Fi`93Y;~T^aDK*27`hcDE0)<N9(4m<gi(bzGs-c>)W9q@crLO`{8`V+m|MRN`YN
z0@Yj=+mM$gaG6=fBN8KL_(Jo%`GP%8sj#@;dAHk3&+U0rB(Dr0ZX=>c$^Ea>2$4u7
ze@ff+33QW?+=iWw^+$q7@xJ(HO7`z^9|<h}QUi~VWc2+$#zL;tYTzy^!v((ojnnBd
z*D1D8@@h#=N8@e*y?(#D7Axd76}(%r_^*!LCj0TJJ2Bl=?6{dNu`r}}u7LOe{n>gq
z7c1y|y6_FgCBe%W5DO8h9*V_F|M9P*cxTvir-ANvVF~Xutvh6G$CEN=G4syt5UCer
zJZ)deZ3G;&V-ubKm1MOn?ZedRW&&xGMN%Z#l}Jo2Z3X4TY#WL=PHbZFd|&(CdoTIB
z*S_?~i&}O)4)QObtfUQzLi6oq;9d76*u_<>dkk2m`PlgjkEW_GF>(C2<{<4z+1v9e
z>ybW%gH4PFpMy6x1n6jwSy*Qb?<yY?r9xp4&F0ZjCcBVqiukKnc5YI}x43QIC8O_}
z-wWm4S5I@Gi~B2>eE7oV2x~Hpd`b9mLosJVHki=IUH(BAEjBswOFyxPUdHS~NvUze
z(BjH;Sv^`Qm7rU-lcGtec(Im>*-d@j(hg@;l4(Me#*3v`f4B7Pjvig>L$X2t8rNa>
z>UScyf7&}jn$)Ia?I`Q)7T<nhuH@kS(}6kXI+;%fwu-s$ce6M0!{hdqE^)YS^U{-n
zvT4(N^WIONCb}2-F+zoW<g-|QE<d$CiowQ7chpLIL#gi+elO?~yxGI=`w0Dc{4HM_
zrGH8Dmlr$}#ih#hi`%XD^{I8RRnD}n*Og}X$S4N05$bOVe{gYu<x=0q7sbjhSyri=
zk9zxjI47}{5xA5oU~<g+#?1@x&e!`T$J^qRj|(coj3Guv;$D4hj<}^in_M%>pYVhM
z%the!XB*N@N(ju&GtZw6PQkg`kpx^a>>027glRN-ICd5w8~FOZt8%nS#2hp)Q18az
zjrRgwE~w;?EiUB@D6ZWYc%zi$<4W^OxWxXN>=g9@N6jzWFUD?qm4ts>%%jqZtFw4`
z<PXs4zrP=vPC9eRJK;%2@UlermG30!F-U&PWeV^nEZEUl-<A0TMe>_-`@)KWya|iR
znLct0@*Ns&Mg-F7@OYB#?XjKgb>&TP9dy;->6*($hnM$Q7f4u1G#>Tk6LzL>I8v7I
z@Gou_jiF&6wQRj%;TmVdhM#53H>SxIW}c<Xk)8i`m${aU*11|aDr$qCtA&O&zJk}-
zF8raOU;Tqa(9*~7p$e~T!~gerq_ri>(Jx0r*q}tENMW!U<a$mygq)3W=VJyr>XPT~
zd*?PouVI}OA#OT2Iw9xnsWhT6w}`rs3$%~rd$Bve-Al&T_g2^K`3W8Yfh`N!wDqt7
ztC7Mj2x!%UUZei&^Inm%<;o}6uHpZ3`t2QYE7A0qy~%@n<EPCPbS3ymA!>7o+#5&N
zapwkg^}f>P(Sav3SMqAE2+zMzZ$V1`ah1(wI^&*$4xUqb(QCiFEsxSEm%0evx=+*-
zFQlqlcAg~R<ieGfQU2E{mNin2BZ-r7$r?$e1vo^d5{Wm=H$KSnr=x26t^^4$8UzD{
z_4|%+edAAM>wC2{gKV+B)|i@ug5s^H(95uyj2u3VNo`65<xs<c?vj!MCwjTsXPKL3
z>%ub^(YfN=1+S0QDd?t$r_D*PWePD3%8nDkbzXxg<ErHoVDjEAk1+k;N=bG?`K`U7
zxa=xF!8nc2%2hfrc~B(w^H2kiVK6nuVu^92&G<eNHwt_X_K<r6BEQdPK|5dp+U)ap
zLCBFY4q(8N6VxE`Ui}Nohi^$E&p87>tz>vF>we`Y_Lvum8y)uZ(r<TZz-d}|aUCt5
zt)Rvj^3$eheIdXL0xkS}y@r`hbL@nCzlzdf>F;nAbp&k)DA_|dLN#XZxR9YV$Zgax
zOVY4wnLpVl=4M}$Gy3kg!(}Ide+S;lA8^lVE9X${OZ+n5ENvO>j`cgtqseGlcN@jt
zySfMD^lWpiHsh0tE>;Q|d($b!94xD-!S1R(96ZaHW!56f|Do)U=L*&OLe!T2P}H&D
z2A_daAU*?Co{b%WN3URKzu=<nL)Ff&v!@;u*L=4u$ErW#ODaGC|E;ncIE#qMs`5Ru
zhSY%p1p#Sm$~*r=hPo<Cp>h>$7@SVHrDG<*0vdQ}4BBXI)HJm!O)#G)jzYJNZFneq
z3q{i$JS4qpI8ZtZ(=2Iiy&ZhxYN?XJP<F-{sJpqRHz#(c&dSG|OLxZWiGFSj@2VE6
zG2$D3t<QM1-$i;Dr%23PXCd3j7_=_gCU^Id>IpQ0Gqxa%@;E*AJ6LVk>SBMJlk?5u
za&bz37q)a7S<DoPiRw{(INKwbjhLd(H8Fp3u4fRPfZt`iiw7Ju){Titcn;-5{*ub-
zR2j<UsoXm28rd?ux!0A_!NT-lEJe^K^wU@b8L0Vvq+~+QE^eW_(>mhMoTO>O6U>kR
zbx~J%XR3}+2AX?V#hT6a1bCBrhoDQ+8LxF36KDRhwITCJ>+HG6U_<&r@Ge2mQOi8G
zkk!+oMXhQS7EGBYN7CQHcddrg^Y2ZJ^A?J-BJ_}yzwj|&IoK&y^b2YF4Bk+M$RlLo
zluW!zEDGt7WyQglgWT9CIHc<aS)|sNWw9f+Wz}ZM*5WJJ1A4dW9y5fG(HAOM`edL_
zf2@6KcbREOnq^e_d3o_F_?Xi1_yu%C4lyGm;ICW}<c3MjID>xM?&lX_ce~ckGt@?w
zH_|_IRP!m}c({D8ud^!U;0ivoK-fc1Abo6le7XtJYImu>OE6d9jUcV?>Bn&|<V6n7
zx!Z5_o+FO=-Sd>g4q5ZALJFSXbz4#<%W@9B{M0UZUv~f3ZPVmE3k5w`l!M-Fzt!tk
zQLbrljem-yzk%nTMOv1J2lqq3!=Z4CVKG<+m*n{f8YTfoyG@+?s!?C;7E^}7h1Uxl
zF5(Y>t}ej<b>Dpljd3)f0YNzYqN<n!-}fNM-s;B&uiJuh?XTrbvp^6hatl2B=14#C
z(+bgc&x2cjr%gA76MlFCh|lk;<BW3qvZXX1R~R3Pl>#Llq%?s@xMzKhbVFStcON>A
zWjdyVY)G&iD&H#ra(>W3<TXu5=-paDC+v6~=X|<5i{xr;6jRsz6lCK$F!O?o&hUj~
z>fFfi*+xZ3yYqc+^<^ILrjfexmX@lzve-D-8P~(kYKwJVkyAFvA_;tQ+dhcz;*w2U
zHMqnQ1DX${lA8HCI4bPxCYPv(_xX3M5n~S*p~i3K=YgyW?ClBU;%QXLaAYY&pNopR
z4+S)uK%f14LLl8j;$;=B0+p%Zx6~XA$9WKq1NNOstgN;lbO9)N&pvr(y-v2EF?oF9
zoE8^J_8j%YEO94&`Hp#yMcKBLVrI&{Wquhzx+4e&4Qy%Zks&nRA2kgBT*}-oRO_?J
zf6`*Er5rPQE)Ss(>P6}X0Fa{>EQu}$I$=s2tK2`vV?RGklvPJk#m#?e%w@8Bv}M|x
zEHmbwTM<ubm<x$|-3>-+q$lPyijHAapD*I9AS1$l>fk$0egn=ZYZ$3n(G5-D(1uK6
zCedVI%pS8ma*uauaD4fG^~)Y_U7I~NZOHc)^?Uzd#&oHlPstDlMqlvSneJEiwAf-1
z@rM<Aw6tfCx|QnCM-2kVQ0vo$pKm9$e0**_b~ybNd<c5t@L!{PiXj)|d0tspYVUVG
zh<mkxhWT3b=72Id(Fv14oP#{1d$#<>RP<4?z0?B*D=_gK#Z~)NPjo0|Jr0d}nQ6d2
z@Lu>)G1pQpuI$H{YN<3_kIUTSkfSGv&&>m(R^hg-%XT4*KfTNGZFGl5_^la(vtDoP
z{3~*EHB8str;BpG5oO;q{=N#8>l}1U0-`4J-6-6M*9>86Uos7hg1cqHvq#T26(hk|
z23r9Hq4qe!-2@$eI~Xmb)s4}Luiogs+k2~dkjxh|V%~iA_GWx^G#uRO<Z{D)NTBuF
zvwMF|;Y`8o{+j>lxbo?_=-nLjVvD9{6}scj#|+M-Bil2ccnaj3`=<e%et1O6tuAbe
z@}RcWH<ECtKYUtjf`9K63_tX1*~UfX4l!HBWB80{KFF${e;Q)I6+h-D5|e$>NLJ?F
zQC-;lw`~(39|x0dG9=hh`C?W5CijFUww7+$Yk`dcxh<WZn5UY1)f0%HriZ^+UgIQ8
zZunA;x^n4<Wi*eFKL5|!QQk`LhMjtlUB%-}(PDpFwTVA9Z-4#IJU2`M>b(yQ3d+&$
z>WNdIK1oDJli;{WIdH$y5`o$UM=3el|BNl!_~G^fz;Db>T?wZLg*xP#xnh@Z`ajj5
zFhzi<=u0?1Pg+D1w;S%N9f(eDJ;rwN5G``zjvm-szP_TN^LLYyi2(JQXf+X6U#-`T
zfr}tYP<DOUZ1+9p@gaYl{g#!@h8MXQ*`5i!(()4SR}51NBZR~R3{Q8Hwb+9Rw+p|D
z-Nn{x!>fs_kf2|VwC@^!JI@v4(L6i=Chv}hR`OQg{G(?5rjvqi#Ziox4D!gsi|rqZ
zXK-y_-Ae{MxvW8UvWlpC=7a<V?Z1Ls4>uF_%IzhmI2KPQA<GP?7c%>`EC&@vAnNY9
z=QOUs|EO=dOk5s$wUx~@Fkx|IIV}EMx5L+XrAHHeANy%LgpnVUnl`>uX`A?HYQb_=
zLg?h!XBRbQdszQ*$MBZqs-Xko@n(y9I@^7ona)c9eZN~tnfg6_l|}`38cujL=ea6_
zt*fzZ5=m4jS}zJxA?#W_+Ccq2Y)B;Qa^6&0*Kw~l7W>qkH7S2h?MU#zO9tL{ecv_g
z$H-%_UQI(yD>iWp26bRU4Ug7NR8~3PnZzUrze2I6<B(0M$$N8bjr2r{Nl-6&9qRxi
z`cVib_>pZyq_I~M9Vwvn$Vmr3ZVbv;@T*NK*-2i#hckhq!p$L@RaHo5JxSA#SyR7i
zg5N)5{=_M&JmNkYYHMtHd*EZOd1jwMig$q{s=(s>7?$%2BM_ZJi_4!$e^4tTg;LLH
zdnq)HM#~6w1F__Uaq`0d5@#+KAKuYp4zwHR!U85anSny4pDv`BHhcDtGRjwr*kv9r
zUj&#%BoO1UqTe>o^4gx~GEGMeR`GOi*8Xkcwe2t)t7bWLhrQ)wfc4+oWq|v!ol@-f
zd!Vgdm@2kVLpYYxGuM_oHUo=7nz!?C#P9(}HOuhN<pl#&qnDp$-$hLSe0aaNn0Z^#
z+9G*OaZu!kKU<xvd-ZC{lbeU8rAkXl?RTzEzH{GaaKbXeBKZY9qwQpY8zIEeWPU%K
zz2V*;w@vKFiLOmDALdnzLY?L$ju}YOaLK<Tt@(ki(d*@71uLc=PU8_9VZ^%<MiW82
zb@1pWEu#0nQAmgu&KIjR!Zk$$ADNcCgYnQ!Uu@p(&ULhPN{NL;sSSHq&g^ko^EJ(1
zSLm>94#M#5`t4+ds_-|RHk{HB6{1SMHECFZ8ODWsvkIX<N00~2m=c1SO9_)wzyVWP
z(|gIT7U5#tCjPvVBg>78bx4GQd=7b&ZovaO>4lA(n?GC#-<(av8a4>0M_4B(18=9M
zpKK<;RVa<f0ili`cz(Y)va7;qwGm;RXQ30k#@P6_z7+bIl8AEk=M&#E*H1)^m$Oy_
zgYzt>v$i!4YSkQ`wdA_o{doqJuAL3GD#}eI9jxqy(^vl!GJb<za{lFUBL635k4W=)
zi8*#<N+id%(1CnP`E(tB&m2p@>T}YBAlm%3j=TTs>7dDaL)kHS<Q@Q3H7z17?1j(I
z-<}Wd9cGZmAN+Pk^4ZK*Ld^R-4-iImrZG83f<adrBq0WPrt#F7lCoDTD(t_eWCI5S
z4p{Y~=YFd@&R&V0G+usQI|H`URi+OswaXtn+?(Z_6_xXL{ipoHTQr0ZTBdhLw4{Lt
z(E-JZYbz>PVf>F27&T(fUw4*F_maqP^Z7LG@jduzBg<wJPDzu2TBenaOp^|~sZrS*
zFjCbZ+I}4y*35-ndJRUBJK=mmcDusM8MwH;mEpjY={kYrB3-&H#871oSK)oao$<kI
z-Khw<fL{6ynnmY`5On|q24^nm(N?tmt3ru(9CQmC?MP2|3_oG=-0};`A0y+4@C1LC
zrQ{_qzJp)R%_E_A#~afG;ex2~b=ez&?;us*{U5VN?dgrAx=^s3*sN`3Pl*eTVeUgG
z`>4%&IY@Nq6nZo&=SbhL$q_=o4!P!zZ;m8!5SPNM3<?iEw+Drw0Sc>j+bjR$W_2&i
z+xvk$e8M!Qak2uX0S<Sz;UxP_K@~`JIEWnM?5`W_gp|~30_X3SI8ekpz*>!vo-!p|
zyDYtSvF6XV@kBp96Fb?t-F&z+kwUc8&D$c?t-aC-@q$tieGDY(v^iOZ9Zgejs{ajp
zsR79Gfl9`XgJLVL#pGKymx3K`C!y=0qw{kA*3kO^(gd21Vk0;_R_fMITB{7a@Spo(
zXvk>X0bzWi2X*A6AKvA}Si8lR6RjRqFcH<RvF)EGZ;1WihJ7yB5VBRM6x!?NIz4d`
zPM?&Vf^?fL)9PViUAP22<LeJ%>oAKAlHwx2NF^1`K$xGk_g_QDB90Za)t>e(x`PPh
zD~~nPw#N$(){l6O&=1wM3^MnT=PNR!bsH2ofOb~ieEg#!T`prX2v)t~-PH3Pb3kb)
z+q6OW;1HqTQrGI7uA63d)G&H_`icHS<;6kGH4!<3OCEOt2&(E=Qht){N-tBm%QfFO
znW%Mpxw<&@J_h6ylJTh4(ewO6yN&CPC_AiB$-=pl*A4ucNJ|lOpX>1ove*Zc+4XU=
z$b8=a;n?u_#9mK1U4~)(m1E;JE~0?`az6iXE{d{tyv6%yQ>YY8O1A5p^UbTRNf0v^
z?f!<}O<3)(?9X3lq-d_8!EP`6<Be?5Q>?ARnAv_6WqVDh8o5=n>c0>dS{SZ5t46wL
z1Q6+}<bq*{p(q`Vvp;Wetp39v$`DO*a-?9aTQT%&ko)p%H<tGFDaEg+eK$M8N2zsQ
zkKoid7O@s61Mal2zKJlk@B}9Id&Kqe4Q@kI69e2JZTqdlikv>31q$p_OFM<+n0bH=
zS(PV8K{f4QvpQ_BnH=t+fW{^plW_g*6;Chneg{GE_37dAi=|FA)0pfJW(vSqvSeH=
z+&duxgsCxA0UuLGG~3pP(ww|`I`%*vcbog=!YU1^JFO2Dd`BtcWe1P`6GqU3cl~TE
zbeh_Wqv`SY<lNi83l>Y+!&;tCC%OESO%Q_FaN=&4yqo92dxgzDdB>uG0&@m%cfhV~
zq~dM?I$V;z77xtBs(!dmdsKFWFPDyETV{OS=<NPIUSZN`w!1U+<tuBkcAO`Y&bztQ
z>Ua4`u63=lR?#c%4Vp*{Q^q6Bn2uVu9CXapR}VtdiMZS8*oZ>x8$z}hvZq4^OQzO2
z<c6V?n}IqJyhf3AmN5xtS*n(1y?d9)SGK9FZ1N0ROSlUPUX6BPWH+f7MXF1B3G{PZ
zFR?g!t*<ey-=U#RM-u!T90|Pj6vr`H!`ZWR-yoD6Q*J1pk~QwwpGD#$Of=gfVJq{^
z)SjG2a6R;?FBkeIjmO*=J-@8zl8xnsqtA3>Dp<CD|2hIrrGuy0suLX7oW3%edT~9U
zN|t$&X+}D{?^$e3%>sdxZrZAuH|)(dqdEpFF3#NN@aE^o7$-x3xsdgiGNa{%<F969
zK|(VtDRkS%)3jMf9DABrDakU$S7{qiva&URX90qE)l<XZIfBPGZrQ5Sk<dfG4=LWG
zX&)s}Qy_t2wnRi!Bg6qgIzAkPW0nWI)@zV<v+qkTo1;MFr~)l9FRGAW^p8QAB$7}m
z$VwySqnybIcB71k`B~hIcta{_u!6CR4cr*Dpb++yRY`$KC6E^8mq0atJ{{WRrU%dV
z-Yfd|m)fM_1+}PPo~za&mYng_H1TEihgT9RaE4Iyp%;xfx|H!Rq*Ou_Uu>iByv^Z7
z=8~a|M*aSK85=-D=JlUO1X1afC;<_{kl-qZ&_mM2934%U*g(Mf2>LbT`t(lIqSX2c
zU!!;nzhsf93P2?Of~7Gw-?I~iluM<2DiT)JJdQE1PRV=+gCkE_lKsLeO`QD#2vPl9
z)PszVYBN2*9M9(>`0nEkmYfNrH+8-T$^Fn@tS+F%8F;V%%%f-cJKvuP8U;Knlua)J
z<~jeHEy5{pB-&E^k-qCqnKu?RVuRjp``F)tB20WZ=;1jrr>`zOzGqhw0|cmFUcSTV
zMlZK5Sb#4g6Lq(BpvzVru|QrpK_Jr~8;rDnh`qfG^<dvfIvRIWVVPHyzi(|(7ODl@
zeZzKe*U5=4#fjb|DPKc`(o?}CjfSdpmpb^CS0yYby4q}aq3ky}Xat-fbMD21r=let
zdGh{0o)JLni{71Qo36KP3akUzEU?%TR|0wtOEe<Qxo5fewhBD{HY3=Hw9$mvS~H#|
z1UZw(Y^0GFyWa0ousSan44HjysSA16zgM@j+ou7C)5gEjXr?Y1sUZ^IVE%UGoyU2q
z%a`71oDNgrLYcx2%;2XZ)37s+jeOi6^r*3Lsn6t`$m7Zmf68zXMlyeo$B`=eIZ%Rp
zKO;@I_@30^1`p`Z<leg3-6lq>xdoHrWk|F$hVlIwf(8G+C2%%ET{}EGf+YgEYV`(p
ze%na_ZwKW*|Gpbz*plEc1y=ZlCG><6St?~ro>hl+#(M(k(L#Rq+X$GsKqiT~RLMD7
zE|r5Nd<$o2niIo&Thwm|#<sgf#~l*^g&0(NDQpX2nRWE2Ye3O_oz9V5D^Bc2@7x3N
zvxe`l6vxl7x5tK?&!W+Zd}-3zH)~SR@@~a&Hpf`XY-PWNC!ou2E`;O&9yaAD0A@_-
zHJQ!_L=+%$!Zh+|&lbODk~SO4ScbP~!)3VJ5V8E>lb_1l6}|KkNq<d6KI$z-iM>O!
zRI(|n{kEF#!*||9_EF)T@;~Y(=zKXdD?KPW?xoZ^=mF0NwV)~^g<f?Az0U2SSeqf)
zj9kZg$d2>`fA+AK+$n?(?!AuyLV8?%%sbV=@yQ~Jy4BBf+w!(Q?<vQ2%`z~^4$#My
z?aKuE=xhh+<w-7eA1?cOJ?&BMU$<f14cj4V<?P|n(S=UVC;Yy4+WeXY>3KlV29YsG
z?=(h!w|~BIehvKq^QR`ZzF=ROM&WzBXx1<h83<2kNFBK@!^)){w~8b9h3wUI8I;fh
zwyn3nKl7+E_~DCz8cvonK1RqcOXO@Mm7;AtF2ymQk4y7?#aJ5hw<P2|DQZdVY(^)Z
z?+>DXYZmqW2r_DT?!KD*8*OH)vM_odV)G|DdYhxVf!8fb@_-BhX_rTUrR6*@Q7Yzz
z=dJW=u#Ub+JfpgvC(Q806uO0AfcLW#Q?lrAObZw9^PhVbEa_xC*dxS2%XGFqn^A#{
z@lsrvU!ESXwRZ=`Xq!{fa|ZSn&8a12AXN=ON&Evp9MmKzDIGo0%ERc#^EP_hZPC-V
zOMWlfp~g9!4sP~V*>f3>zYVryj9UF306KmCqN}fE3?PI+2I+-rZW(kd$-Sg<$R-~B
zhLye+8B>dj!i*I3kzmKr60+~x3`-hS&hCFS1MrM*cnhjK_ol7pyp)DZ6_^DF^Jd*Y
z6C}vUMGH)VPdpXpY#vnahiGWF&g$wJcAxLIzE%Ye?v)HQ9E3Tx{HYHF3^(plY6=D#
zGPAuMi57uL_4#$>9`1uO4d<!*UXcvndgW$na>&=9NA#W)bx1|>BNt$8Le7BpsN=R6
z4Gt(`N?AJPP63W{L93Dj=FEi_83Ah9ys*4z(}R7RN4@3)z}n?j;WE2PNqyCDTjLe{
z;8W?$k%#VNHTMjO*|pX>vwF>UBm#6zRu}Eow%7hP9ll*|KE1<^Be9nfbZJ3@#%~KL
zjPu(=?}ojg;JVxaBAak3fXk|_v|%KjGA8k!qmbmUIdhX`9#{Sb(WPAlQ0M3?9@hmQ
zm+n1@0aT_T4oZ4DBv8xA+45-nzIo3ZFo>RG_Nv@oPlhHoCNRN|F~zvuiGBQe`D|OU
z2Fg3Y0&7<~uui1Xz+?!|aTOoSH=zpHGFm*FoB!&E7q1X*TOOI&&CQwDrQr!mIfuHb
zzS!jpn8&NeFUu4x74HZ}uXr}&VSL@6pa00)yx6N985#4KSXM`{mD$VGKJgI^+};Qy
zRmR-oj*4dFAW?c>1oALmSasjYf#|#Fmj`t$GZN2==bb#X{ye^E#5(}X=uOw@-pEOy
z4pm{yc89-&S5eTb4gZ{Ma;<LtkzsdmW>-`@2Ii!2sXR+d9p+3)T?FiC#^LPB(PQ<z
zapA&8dS`!}m;OI9no~eigTQzWJrKRPRIwzn7MTf2P6@xp!kN!D`zD1D{j@oXG5T*}
zsY_aOb)<VIj6---5@in5WH(T%=Xc~lWHlL3BORtNMPC^2lsL%Z(=y4vY@l8SGU0_A
ztmhHQ3O|swdlyp_{n$JQAMT4=<p_JY9LH>iS0Q~Lx@$f)?yC&4x5X6VtQfBdIY+xZ
z8nc>#W#zs_G<K=8z2tWZe#WV$s3IV!AkrlAg(_0TOjx-oS5NJv^xZ3drCIq5Ptu9>
zPfjXd<;N+IQ&bSY*BdQuqgpDaGMBLGS&?9(=fgCI4+ke(+w=VQ;&5&e3J|^J^`zk^
z{4_@Rs;)V;A`RNb>KF?mbBaro%RaujD|Y&#b!b9|duWjaxM<xDAHH_Ta%Q{~#u*Hq
z{joaAitc(s!f6yXve0w{pJ7Ck78@KqlodcfC><|7gx&vRnlJZ~7E=#qO8FJoh;VW_
z6;_%n0nQ0l765CUgs5uz=Thm)XhfAb$&wo~j!)`ma4RQa96<eYG?XvyfWhS%hNg4L
zU$*yr#frpZaBsI6jH4eI>Grbz$|}UX&R)LL&(Y$cdW?r8ppSus$0ba@moLR**1)j4
z4n8j1XdSfV)`n_*t3(|yoBK<ht6+H(?h56Bn_R3C-wh%}0`ca}3(>o4xMGDR-u?|0
z6=u4Rx99~`X0{QfQdvVjntYmHqc1W1bW!S8`Y+DHXQ!mMf1g}K!2$SR|FLZ%H@HZ!
zO;`I-FRw0?_Vy^g-K_%40hbp)L#GfhTV1mk3AtSK@Mu4XdUMBoms~{l6Fn0gu;w9<
zGmT;KXJyzDyav0kg)EI3$PZY0w)~Yddf99MN^;sqwc{8XTG8nE(`?pOQvGGAF#^!?
z(82F_{TWV0>@Kp-mCYM}Mi9rm$LZv<cDRquc`zi24o~<P8Uea|j~dv)lw<cJ5BA^W
zou_~Rz_PmwHuVs3bP6oKTghF%Q3N;9>eRA;9p*@=^V6WE7=dLF&=FV+3_8h0v+C5e
zkv!X_Z+}rk1EMF0Y+tZYrbJz<l#Vy8Ia_SrIxzm-Dr*_7fg>j_tY9`4FbOV)kh~7n
zMok03<O%&}L`}odzM36zF-m<+4zz<3+0I1AZyf2>Znl(cQbj~iwoTe*3o4qc+827D
zzRH@_z>$9t(!>Wg2PX{PYa%LVWI{t#@UhNB=y5fQthJW{W6q03jrGvPiLKw2>q^6n
zBh0R#(vqPaU<3(HY~U(({-YA({&IH%sP>pvxN3JRB5ni><p6^Xt2OnvCX@Q>HZxc_
z68VtOCXHu%F(og2Wx2KI<hR1sI#|kR-HZPz_0Y2;!`Wf%4S-A#;V*|Ne&bE!yjy;1
zNPq0{H5zCuTm}f!&<X44Keat~Ew_n><6y%?1AT*WgWG4wIKFmiSWP7fhFdCysrQZH
zI7mj{d6a@!)ZWo1<)IeuZdKlscS)JDix!8R*nUAU0OTLE8Me)nRJRPJ6iHB)(LsHn
z(w%&^$iN@ARM8(T3N$n;YrU_5q&#wYJEJ;{gUgp!b>y+gdHxe+@4=sD=~c2m{k`0(
z_o{>F!&c|k->Nq=mK7O3`>oEGj>UUpE8QC{*VH2C&P|pa($PGBc(&dETo^sFo{0bJ
z1>pX$Iu&j3)>Vx=(3+dXCN&w-4FuY^YM6C6!B*Syu}#{ncG3m@C+3&(4Y<5}5V!1r
zB7Sf{{ov=g4y%kZvb&reD!KCiC~}7XC~_*3`MjuA9he!VdCRNb)E%-JajyYu&3__g
z(~tTtgI=irtWP?;bPApBxu6VTL7#j^(x9Wu<4;p*^0$AceXXptq|;+>Kw8!AuvYud
z)w@Fl$%lVUi{UPnQoU&Vqlkeg)@pG3r5S`?MsYDy@Q>ULa4%VH21aoM1i>KJiN}Vq
zJ3T)g$(Ppxbr1H|lx6)m`rgBP7Tu;e>&lJ;zmC4(mrLG#R_bTmrl5aI_Pio40X{BD
zgAg+UOuD@a$p^Cm#Gg`~fL>ZJ2bV49g==&sIQqToH;yFTq{tn>X7V@Mbh|T+z{(!F
zF=p(5=(UUMx1seURRN1QRp;ItUYUbA_qcMwX2iof|J+;LEqDiJ&tRO6hi*`%Q|k3!
zC*>T;%enWD#4-c{$DGKxqQNG@^1$2%r?c)7axD_@L{#ap6cILFBy+IrK&hp{(fR`y
zxHj1QJie6Gt@+x1bj0(MJ&yX0@3wG-+M50O{j=7mOy$N9e$y$U2kY#4vE+yA8uY|?
zxW^CcN7|9soFBenr(^s$npd;ZUS0g<Q#r8Uuiqg%GMK_F{`Dbo<v+6%=iE3DyP~`1
zYUtLt_%w$S&c7E_X488IZ-&cY;=k?r0mlPtQ@miB(uHZsRfk|Y5g(JU$+H(UE1Ly)
z|MUs~ynEpqB*{4L%z|ZpmEjps^b-E5E}BMS<cLOAKL0aEDH?aYr!B$3{Yb2k=cp2+
z2f)c5H9Gvx#s$DHK_`KAI#vyXUG`WJhP99f3KT%!qZ|#M?b|1X(8H10pLY#=ZzZFK
zjVZN}E%yf=<@2^)L7xu7LSx6vp}wo{MA;}_<N);59{^>2jMiB+v06Mdm1Y!uX5eYd
z{yvaKR%Rwlk;^Agk$b28;HoS?`*G9$wqNl<ts|UQhXDN84;{{FUh=$4XleGs)xEnT
zHy63pDf_+doz`tOQMDQpGryz%1Tg<NIg6U8+_(0Q_l{%yV96Y6RjacX+B;_pS$6mY
zG;l^0;JqfnuQ%i<z~LVlw7-9PQSkNW#}qJNfEgrj{JotT*A~k{k>*b{^}y_G2mz6<
zJ1GX?g%YW_gF$Q(-e}Azr<bz#XGKc@R5kk&{CJ^jKK6!X*X=_P!Ogbnv|{>Uo7ZaJ
z+cySVK^HEbcI|4n(?_YI%10u18lD?S4U<VYr}suN#%><t(c9M#D9U4Tc18S!;XD|J
z&5>`KauUQo0Bnk?yV!p^7QYp*(g4EX7YQA*vAwPP?y=~bubBF_2)fW-T^r6B1Dx~w
zT}FN|vx?SYDd*qr%FXHPjKLcqspE3sR0_L37!~fIh+>pyrLhO8IdFeY@3~*Un<^^<
z3Xp%9o%l_kx0aw#X|XQ3+C|M<%Bs6c$ILa7E@0{)a}Hd)RYTKFuitKrd!P!ZLq1tX
zhBoo$hX{&h1{8cqK~HtYk~qdj&{TEq2^o}O3UeAwY}?VobC8~Wkbq>Q6$HlL&qKVp
zJM%!R#;ZR159Kfx-*CzEn(^3g09KfkiIUv$##qkTg1)^3TQ;t9KVBVOfHZ7G>5wB<
zx=m*(q>;N;PXJ&1ZmZdi=;%)}b4;9C!#QwSN2fMM>W(J+T{oh)zE(&vYy^k2nq)Nl
zo(LK}w>WA)CslCRksRTLdwMiJ;oN~oalBXc*VmTomp=TCu90BrGYj9PwcNAEBxo<;
zB&5{SaB3cc!MTcr6)V&v@N1hs$}-0!)LhCM8>X~3U$*j}2n*8K`97Gp<n!)x_sY57
z?I6BSQ&p%-PGoZw-2*-7U?4f?CjcoO2~uvFiGO~CVo~+{qI>U9q{WcZ=kbn4$~-I3
z4|XJ_HMmJqjiki^F@Q*YP<iKtKRjZdnzLwY+BCSpmdET{t5Y2FdPw~jB+<`nc@~ur
zPI(AatUi0U(CXxl8?3N{cCCmQ1{9!TCm4=d<{bfh1o9(!lTgre585&*>mEY@zIB2c
z>5g$HGH#{fikjqCAMBR}8P+vv$&j5_0B98F#Oy2|KiQXJQ{%TEobL(<H)H`lJAhNf
zieGY$#hw5RK?be5iX{caqDF@G6|Fnl@z$vUkH%R=W*q2Tg~hnXYUm^gd}5`Ouem)b
z%Ar7iBRm*4X%9aA8ST0;F8@?r91GvD>EE<4{PsBC$)v|J{Sq)=0wZYPeKzteQ=J7>
z3+HuoBN=foH;g+dSoqh5g}PT3J9S1$I>H~@bG5@0jPd4zmRzPraYc?Mc;a2%4${o0
zMyLH;y(IMjYP9~UGd2GT|GHBn(e=h)*BldLlaYS?@j89MSx=?PN>nPim&A-pn-Xg*
zenr38i@WKU<Gi!h16GI8R;IJS(XYoa!yCsbWdA>)HgDT&45<E-L|6vGYt^_zp?~uh
z3i+<TcoAMxRI_yi+XD;T=6JRs&3BETN7yflNlzV)h5-}+W><a5lw#bclkxcw6^{en
z42(>$IJL5Q_wW{2z_<B!%4<9JE?bOv{MoqUA1oMgAecr>;oR&8|H7~ztvna!XLDPS
zfDuNqN?$CXD{7~%-HLsSwTS~-QD3Ly)7Plr%zIVbI1P9$1IiWE#|+q<rLg7TD<#Ii
zMBEk6U?H&HjqP5{cy=1YDx@^X$g@%Ol$1LF{r;Bxe2FNf5K(`Q9C>^_5jsCJJUonh
zf6r>5D4h~5C1g7z%tbrEBf;_>ec%Jcr@yyq`Av*c{oq_=RmyVnRc|fd#7?K8!pxah
zZ}>`r>Ce2Ue|+2?>JmFEtCx*VmF`zog=Oh{0&;o({J#eW1=J+uxus}mz97>zWG?C@
zgxY1jt-SSeT*T&|sE!L+k|X`bN10}8&V7;|F>lk<!2DQ4g`>W9vHUsnhbSt7WJy+_
zch2j<=fXNiX`(QdE3?@x=C7s*w8$p%Z^BHKV4B(Px|>e1=iUDS0YU!0UAEWxdm->P
z>RYet*@^rlA%d}3<&87dY=)8v5F>1zAsVG9Z-S|pnslO;&6jDV38v3Yu4y~<&TvA`
z25tQ5O=;;dRlnPdfL``*R(dnfh28UjcSKhZBYH5pg&2WZ6_5ciXQd4wFurp2ZS?n-
zHft}}ox`jU9BWG-C1yf=Xp0<w=-$(5Z@Y&DSiik*`nUE3&U)drR|W`{KzL|<F+jKn
z%@R0r*9O!Tw(_-tE_ojUx|5ICkIp^eaQgiNY%89H*!lo-E|-7!{0425joZeBHh@09
zm%q&%760&-ztEizwB|eINY)5?KWhf*!)^CG&i!z&{<7t7m`sB?EohHugW0b2i{5aY
z&=k_f1t|t{3}Z|mFb==+fj2amRD+qHk6d*v3xa2anEJHXrGDk5=Qn7lY#U3{rB_^I
zs;O}ao6R|h&F5$X%nN<t`~OGS;jRxb_l6zl13L3RqZ1i|j|1#n*9VN$1KG#J`_4X&
z?@x#^t&bCg-w+^1D=8XYxtyJ@vub6MHNn!wn#KtCxiY2I`Ciic?!cV)r62s-aiyT)
z_dlO{8};cl?vMe92+X(ofVuPEFM0zlW9RMqfamga-@BPcc<jpnk^u9W-qaQ5a6BLS
zcM9ebJo7NKs10WG({Zc!q<=c&RobXxF8Rc_Z=@|ucIX4z3SvZGe`u}z*NuN5XpYDL
zKP><6jjw4SIB?%?v%ok0`8--MRPQtqBq+34#uz6RI)^~3#7HY2_(qFdqicF}^dD_b
z(&+PO=Wj)f7^5BZelxDMw^!|ieMT6;$4Hz7MX<AOM~pbF?Rcp`Lu`hkTY?#Gu*p8b
zTFTK<-g9C^D+oe4*>YK!VA5XK<EOO|U*a@j5Bw)-`)V_9zFNW?9IS#BIc7Aztn%{W
zo>PH|wk*^WVK7wxJ+6&ocdl<M1h8WD!NdqO7xnk;^7HGy6G4$vtqG$x89h64i~y6q
zaM-hX#SSB*`kAE}8ylfX))<)2!f_|*5C@@^F@yB_^#-oB8l2R>)iPRHo23{#D04%W
zBuPWm874rZ1|~ppPaC(Dx{qg4EJNt0v55)F7pfEwRajGKfHwgbD4m<6B?||p{m@}9
zsqpucwvYAmthWDBFpV^pt^4MYdC2-}=WKmDjOH%cm-47jgWuW75^Z0%mHL+~U=6hh
z*MCMvXY~G%c}=jEOeah@8P!SOE+Z`Taoqlv?MHeibwRudCeo<4j|OAL8i3k75q~xY
z{n&NdW=X_JOq`~Z@x3@|=PRS~F%HI9?znQ*GQ;A~hDtged}74N!AiFfBS0P+9T%K*
z1n+vS?}rR8yl?;8<1CneQrLutb6=Pe3Wxh>Cl-*unKe74ZxPs;|1A3@f$1xK04V};
zM1~w|fn0g*9R$a;+5o}!htEHe_TnAU2oR+GkZr&lIaPoAX>1ba&_<ImPqE|OYk1c%
z>VsXnFuAqElKI@m{b)5i0!9;W{_`Vr!yg_LWoZ9-)q%UQK(Bu1GVVhd|Jm=}M4Psc
zYomL14u`#;W%Eq1g|822I~>cq4G?O+m)it#L?51GzZhTo!7YXag+8wKH?$>AW3378
zaM{3kU&f@A6p-Q`@csY`8MS85u(KW7B2T`sQJdxT`RYA)q>Ilxp)POStnj$Hg3Wb3
z!cLjZf|+x8r2X=bZlz~lZY22@u{om8z4Ocl9|tG{<K)IaK2%4_$+Dp}^UgDlZrBlz
zUFg5y@rU-|S%4rK!qV69GjJ?xi$DkrA3I-V(lLmkrZK|15ZT8_K9!^7aGZ*T3YDNS
z!b?a+jC8iwpn(Mk>QWN}njRqVw8XJh4Qr+d-^|ZsJ+X;2aKiqLpZ~UD4&P&kCG@ci
zPi`O!KpO72Y=L;L;HzUR6M`TiFwWm}{9$#^J@UtV4~;^jcf|gpT_@K0F9NB6xdL;#
z3~)LRvtD|iMVQa=yqbmcPkIHNa^!vux6|yq>UZo^eI1jvFpqcKA$u~h%3>XU8A^<0
zBA9Cg31y@ihd3X6)A4lJK0C?i)5pi2T}QwC<HKTZ>}O~53)p1j!%v&fW77%MiDwA3
zN{pC&9651(Xe0pAE{>fx(kLINY1<L6)5M4ore^)-B_2isIBot$-v`?JojTWli0cI{
zUe?PFBcrm$$JzOM(pzH0sGJ~Qv%(ERV<Te3$(+<|o9*Tv?X#yBJ~Uev;#c{<ZS0S1
zQL-ld(F*<C<fpBloAjMZV<b3fwk{w6^87&-;&n}uHAaGhg<iO2)Gnv(H;QNV>`bsw
zIuQ-7#z<mdh`)h*>V%~bIwyL&OXJ6=UyS;>Q`eeVA9yjZY1@hCGRzuf=}ekUWz;lA
z@V<kOkrF?viBM>o7}2H+`uUbrh8D3;S-s<Uqk79Zpi{(%lXqoM=1YlRo}I6UG|t!M
z^-do1Y}PXtU^Bm+#Kw=Zp-sJkR(hKC9s_B5>Y@8Zo_DW#U^?%f@oJ%Z^nr(o3wy1u
z9jEW=PUHivyFC4~n}`wY@I9YROu>v51VZ`%GghOiEc^M$69j7zN@;D!<K9=UzLlU^
zVH6;Ih9h3B(_aXeVAAUEEUd=)QWki<l8Fh}gO>rmE53ch9}NZf@LK}&D~Ig8Q-cs4
z>BrJpk?#Q8iPV}<L0O+?L83I}1>qNnn;n?wIh=)BAZLsM1m5?u`J;(!o=KO{22#f;
zjny>>$YE9qq`~g1me65nhXM9QCKB#`<nOeBv4!vx!pBE!{0V@(8ejq3v8(qGc|(|L
zls9%tJjDWh2t;9$=h1(xBec^lpzheQ4I)CDFGvyWzMkOS(K{@iPls}w5LX6Z|I^#=
zr9U#+gR*6SZ<#~(+DYtgICB48O#1*t-QOO2mNt!y3m*mWtAH}$6JW*SVLD>pT`3x=
zH?JehNO8Zxd=&^BI9)z$pPgxE7V5rYzg-%3s>?Lc8iD@=2%k}B5PC9T+?;jvL3HLZ
z2h|C4<$VZTv;3Ffen_$wjKm4M)Zts@&)la$3=km3D9sBtUYPUPj1R=vkuLA9VlzY)
zZhwJyaTfaql=JmAcOEq9c3H8IwJ>(Co6k^<d)Ge4<ir+!CMMaG+mj77!?c2D1SH)d
zOjg1amstRL`0(FfrpH;BzvHrnOnlGhd2yEp;t)QfKEp)#n_qi`c+NpS;yn;y(p+zf
z5M|-r0c}1bjTBjz6A#~q&Oh;R-o-aFXXJTay8nsi1<^f!Fv;3ys7qWX5pyEmBY39p
zyqX30t%9bT-pRW@KFIW+vZ*i3U3;@;qCPzS+<Gzh-~Yr5Li0;|A7&*%>>Is9wehx(
zXXp}WuyI)`v<Csdh>`ZxzE;XO@2VRLfMY~6VMnKlk#_clwk#_FoHpKEJm0u|g4cC}
z?5Aw3Y0ZYz#F)?+0d7)b#BA(IY?~2u$u#XeHcLKP;kIdvcO3UssKP#du#*_`1A^Hc
zVTWBaI7l;5l3WX#@JQblSx^1-p#8S0r&<5WF)!~U5B@w6VPwO08k(P^N#0>heIYS6
ztjh2T53#G2zja1=^GrW)Of>SbWkt53Ei%Ym^0IyMy7p*v&+6G}g&5hw>)d!eL<22o
z^=Rv^KbJm6<aOC!FTcpfQ+8bLNZ>5f&V^$uLUoKAsX^f<Rz~9!#0dPc2@L?&Vicyd
zm>7X6Ef6DN{{D{j7|S#KtZS=zH$&h_T<qkjt=4HarBAcQNGxDx0OcIj?Kro)9di6P
ziyPi;%9K{KsJVy_!7YUIZ+y)mw3tZ>JK+wB!0vwt(|>&1J=2;3W<^5_1j6(s^I729
z$i37hgD?!{qkhS|^^MG8$;c4&opRKE!bFx<SSIi5k02F(bn9R1d|N<pf8{0TQ<4SP
zP6GsS-@WNiPNN&fLwP7yOB@(R)LndtcN$;*i{I(4*8hZ`j#a!E{T!1~&4h1V53@4P
zd|!YVfFSd<qYj{9Cd#Y=gpDt48lh|MxX+MK()y{f%13_>4&T?Zpz=GPc&q-F>2Ekm
zfB$7aW<Lw3Sqg~t8C-hBwX~H#M706W49K*zk6X<j|HHQ9I?d$F&Fs_VclSOKSf2xu
z7>V@P2uctmrDDzGS&4m(6vnqJVnjO#+H?*_=1UDsS!s(0G4PHr{y=+4oMWE+!Drvm
zaNX*{HNU%`PB?sT)^;*BUvVFk72mw>PBCx3ljn74hnc-IK(_wR&3Dn$FKx0aQrz2r
z*A;XglOr(CD9J>#L@-_V>%TlI$RH;q1I*>Un@xOa33_ebc%S|DzO{7y?;oTSU$GBs
zQNBWZyYBpoYktq9=?it&@Z+tg9wjsx&9FcmyoWw=)pd#(0T^uxdSb-wg49iQy($Z4
zAN&%=xM`O@O{a+wDIk>ff;ovIzq=O1V~$Lc$aKvftpsq$E9gSwX4~r~z8ywJc^Wxi
zcls=uX&9_uznNC7SR#sZnwL@Cyt=lLwa|Ncy>4bE$_h8Mv6N0VO*mQA9Xn`y`9JxI
z)(l2L9_P)p!wR2nyeBnA+DVLfg_F~MGm2;R>;RK?n$l_p53_nhiweQXPW)!d3wm6m
zzLj+zD;5k%H+etEM@fv}SYYjt8Zpu@oUelzNrqsMP|n0bZb^OJnjuDHN=1xFB<znt
zBu4COtztBTfZe=?cZ4EFnnAS1xG=E-(<u;8K{$n-wlUr<4)z)*P_nE&vYB^*VHf>F
zPrpd(wv0M15n3JxuyE>tsl4a$4qy-i*f<NV6lk14g9JN-ooFt|f+6TSc+VYZ-~Z3v
zdw@xko#mmYbI<g|KZpG{YFE43U9Gf|00V+cFu~Xcfx!k7F7`DDjKOfh2+7DK$pQf;
z$)tzta4!bIHDC<bmhmMC0g{k*C2i8?-9M+9>FMbl?)z5HnK@luvAepuy8Ao3|5SC=
z3Ez3osZ*!Ec*hx74s2SCFVforz1z_4pmhiZIFCN_@=%NlG7Y~HuAxX<(DdwutCqD(
z(ELvM*dE(u`(YA;5j(jSGwYx`8(Q^HXb~p;$w!|;@cQ8x173B|H4K@e)UKw>vK5R4
ztzc+fu>-4}XW?&BYgxN#baS4WF=)KnGw%@xepq@jFwOitZ^B<P<J&Oa4BhesN!*V9
z7F_kv3dAWf2sEv7?!WVl<qnNPI*u=`Lujc&3!}%MzhpfxjLDGXL<VBMy`u{2N9oH>
zt!$K3W$Q40gyl*ryCxPRF%ROGefDeBx?w5d)I0@{6jIw01q@&R#ec(o+4J{bV{K;u
zbw1xt=>xFb(yr<Fygzu?Es9nxXO3@J>u<f;gzLew<@%0!DUb}BtoragF#!5qaBpQj
zPp|}|#TlPLTAe(B=jo$xs~4tTwtoWFi*#$pXO(Lt-#wLD-Lkm(z{j3BaN9*Iw!ii1
z54wJYa_w;n46hXoXxTyqgH6OE;<)V98`e9G>)|6$zhnuDgW@6i#z(*|7ztaW*T;F-
zx|n|y&7@Jvj6&O}WlX_HSk3rR+e|_^H%3+xXiEwWqhQ1sBg@Nu!MwH@aqCai&!<~E
zabNesjT-muhy^2V{kw4SnmT*txZ7>Ho?ryXc;#2e5@LcfZo)Vk_HlPJ=yUBPl=FIa
z#<3WgGzMODPrB`>>km49QMWq_^=ZKf=OSZ`RCn>+jo316SpK#1`Gc4zv(OizF+K#6
zymrNh!eX`5KwOc6VLiroga}3wFbBx+38Ix7mQ<}39Ch83qBm^+7BRnI53QaFJuj1<
zPo`ib2c~#7mqx6S`a!d_7-^)%mDT|uax#R#B<B94>ZD|(uZK6(jYuv`g6>ENTd+#T
z6^uABUdDmY)R3c_%vHp4ipv^AyD-sxa<W8m)42EQbGZT65b$P4_Yi(5YJc5qFNnWr
zKe|S_KTnLTQTp#y1|ehN;;^SW@S|Q@-SEmfbZ-ZOky1dx$WY-qx!^Uo$l%V@5f3gT
zu6AgsTy)yU8r5FL{(BP7GuzlgAmU8~q4rC9P4mmQDECbvx)zLN;Rc%lre{eoqTf+|
z<L1PG_j-1SNm4CF?5?!@`Y!j!L%@7i#;Yk@X)PGpt5#GC)*&g($r1!46pW;bMU_jL
zwtL~~mfZ7;z)qzM_v}jt?nT*CyS4QE_I@^c=JY;_l5!D@M2V)<B?t%tf`A7CWFQtg
zyDGo3p^EFvmapP^wW)d?FpxXYP^^_zt%tyBNAK{cMQ#ZKf`A~990crw5idrb-qg~Z
zdGC5zojh|?wcw_13Bd);k7%85lwO*Gkx|NzQie<>_Cxz@*L9FlR)?F!6RRbC$@AzY
zfCwWP@oZp>o96h8QBA0#CIil?{ZZ$M{bP3&7Zss8;*mKhgVL=E6C<}g_Uj@T0mH*1
zHVfxNZ*OlvFk+uGCZ>9BjyfOBy6op(>#5G`MVrnusUz3hI`*$j6b>+5*{Ogzo`F?t
z?!cEn1;R{uSXOAftPA)<`tz`k%^@VRJ%z#4Uawh~y&m)T&3auO0mD?MomNBD_L}O{
z#;U$#9a)Bn>pag3u(aKSaI#TvtCjU7%Yt<sF%71jC^U7#Qqx7+bMUCZ_wFdk^Vz8(
zjw0f9Wyp{|sJwr6`1{~StS1zVP}n(bI@~(lg(}gsT|DXM!s}Ubqd%(abk(`Kzm|+r
zrxuKC?Uq#v!ryMW3<a5WEBN|Wtpkr1tFAbk&xRKkTJ4MO*BD=Q8P7lWv`0>DTF=!a
z)?~X<js166tb)8-!q%ztYwYyzBp&T!v{i;8)$;OyU?hoj3r&+lfbX{0EJoh@-uLw0
z@s3}#dUNjXyT|HP`NmC!Wlk7~W1%R81pz@IP6Wt6T<&bE)uTr&H$pIcvD^)H;g*h?
z_0{bywU<VmAQ6njIV{qaARq{gfdJRF)m~YxV_gfAxaN(^pa4(G5-^+Xc3`QL;o>@H
ziGf42@Q=MnDan1N8$Y*rM_%`$9=}mA3RA}EId!@gWxwLU&kKV#7nV+Y;hwiq#xRbt
z8ilU7%5bWXNy>K)T};!}^t>mpXNT|KP8;TpOn42`vX_amGnwcHSts7(8cJ8tJZDxj
z&WD~`yv~HG`4|7-Wn&+6>sjR$<l&t|!9wq#Z5Ah8J-B;LA?5_fO4p5Bj>&0nVP<r@
z->1&Zi9TC|>;Coc@ft1uOv`6__ptB+eHjqV`%F5I#;9L|?r;BBzukSH_Aw&lX4@7S
zR(-^Bux{4UZ>L}1;otsyj-*%^@0a5UbEa$Bn4jlNPIYsf`!td!D;NH7{QEST_3&;#
z9b%@-upXZ8PbZym;HdY2oTnUbjyv~G{LOekS+BwaHO>AHjrr{NWV9Z5J?DC!=$h?V
z*?;SN<}pYfGrycC8v(8nS8{hmO5|B}UcLPE(^j2td+Rs#7gGF6?erSGH$L<R>&Cqg
z_m88teErfzb)x}o1YsbK3Zaw}1O$PY5g-FG-!7}-#!<Dt*>{Q3g0;xrc2%YDF;%~L
zO*JLPNX(-m%?Sd6z&HrlEk?$vD~hsqEwCVUMy1^rrIw{hlNn<qNPp%fkC$fhy!3J0
zevE>VQ}Fe1>6o?hez^JFpSAwZD!u0~Tv4~2J()=IwcDsVn7fuMEm&gMtp+Da`1zMF
zt6R>U8fqelCb{{38LGhNg4eU5G4YFC<kQ4NZj<iikQO7?Yc+LjX=x(#j8m!JYtSX$
zMj<~SP#UM+c_^zbMtIC)?93P=MZ_BMOB!3|i^3hEUwc9Cn{@e=8Rz--t9xGWSKxZB
zn1`ElSi8+a@*3l-<vn!*v8KmKw45v>HuA=qW3jm-9*5p2%JLl%o5hF-Mxq!psY4JD
z1ZEWhQ!tXw!p*=+L6vuEaH9onDQK|Nx2}s|WL5`P=oSPP6oGLCBVj)FqJsEly#d!L
zhb~XNJ}{zcbDVmTBp9K6Bei<_x}oFr%D=1^1J1wnV&oVFBXOnW?MjO?XB1o)W5&hX
zOEPDY;GW8<*IEcrfyKiH7;5%-ER$xcPs;_M*R!E9Ng{L6i*Bd!otJADE?rlJ^^!Wa
z98AD7jXhbESet?o#%1HUGx&O~S%<5nLHBr3wkvhbHUbiSeR%wWynnSzHVP3pHcJOC
z8iNSvhg+QP6|NN`7@6)ENb7=tAg~|^kb#)1-B3%%PpSNJL9N06^0jNbR(uiIm#b#A
zEP{~*nG4cQL111G7*{Y7wNPT*m<>rBmm~>B;>2fB8ChX49tEQ^g0C+xSnzdw{2o8t
z{O*StBgs2D0@Gst2ceh@*qvGvaRZku!AOuYQRbNzHeSz$2u7kDVX4ar0lz(?n_z^&
z*MnG$IJGI)<{yD1uU*qZe_HRDY4w-yHW7?W3w>!<5D)|w3IQ?@5uQ!0p6Z*v7jkK}
zeZ68~P+VJ9I~OjfW=3KsE!3ot&I$rEjzH1`Bcs5nHJ(N(>sLm9#8V41kASy11Rpl7
zE%be_dg{=cPPIGGdR$hE^r7O#fFoXEg4M>Y1sjGoud(;L=ZBF~`JJ6DRVuAmx00yY
z?L_mq*X-K-Ubr#%Z@M}de0|*ZlU6>+x|1}TN2Q1BJ5w>`A8cH&6Z+`Yre>sRixIE9
zaouxrCCPWj%?sDm`srf_F(DY!f$qOXNnX#6#w3aB3hC}GmqzaQC|F6E_z}>8k=^R9
zWg)_G&vWloudDTywfG5&utmn!Y4*CRfPJJr1>MI6&1bKFoZs4dz}~jgdNocv<CLw{
z_EavHAIX)Ly<g+dF)wAW>{Qgr_5MC{R{FGC-d2}Vw?s<rJ*yS<<VQYa)%pHMKYUPk
z>ZzBW?0x0m{wnLjD_;GYgBxK_Y6yr|yMA45wTcpBBrJH6Ul0%kriK6+h=n#x4Nsp|
zjlH_c7W1H`AE=P-v|802wbw1lx;C{Dl7<BVK_DywT-P?UHFf0Vu&?tVbv3&k2n>cl
z@`99g%2NeXb9qFDk5hd~xYj{Hz!v=1Q`?by;3WhQ4U_iTEmfKkzsY958u(D93k*r_
z2#kW!-d;uJ^LZ;!aRxzY7cn{GMz)hu<vN}<Fw<r39nOqV)}6!B^LfT~xr%=|#CRxT
zP4>cQ67Qup7^K-$&|vSJ*-WNbzJDd`U;DFT_Klb@C2?J;LD;#p%=nJuV_<bs{v;9s
zYYyfzs*U(CUAVYQHyen#j^OLq2Wg9u8up1xbGA=R(qNctLXzn6qGhjMv)1t>u3cVq
z%f09j;JYnqi;+8*w$w8pd&DC0H-GOp4&FYgKl$@N=>4;Ad5d-9j{EK(s~20>ud1Dn
zFc8NAQ3?wJf<UYYkbziIyK3#&F{|Ay;)=B3LJQwE(Be|6a`T3&bXP<$66>f)V}gJn
zFb)D-*VfY&wR!BsNY!XfGOxm6wC}^@)oqd;8c);=-D0>S>ElN|wQ!S6XSiPZ=Wat7
zl+99sAX%w}mv$~NLo#~W&G{M>BVC5aTCJ{@mWrwcCT13bku+u%pBpB!3^~X6oBw5S
zw}T_uue{D{O!jl@;{G4zG3TlIOxo}1a&-tubJ!F0+v{3KeNP|FwzX?4)=io@CMJ`Z
z*L0s)UelxNXn|=)3#@*5=I*Bs>S;7PDhn-p%|z4pypC@DvX>#2gEnLK9N~Y*F&{bV
zhbSG=XpZgS)$nH;zo_anDd>-J|B9Xm=6$DgJ-_VtG@ZOoVRfeq<tD5J1;h<xJ9%hw
zn|XA3xB7VB>z*!eKRV4Ljh)$k8&qYjr+2Uhop<o+@Gqx!>R=u@Jsa#t8*54iA2dl>
zw@vyRu)OR5i=|xlU_BV_@#0)D=gA~`J=(nPVubjVI;}RrsH{T}Qh?yUj==OiDBYy+
zZo!09sSaiXei!kqGL6%Q6D8(cbGjeoT9`2j`V0qYcM<YvyofQ_E(fb31<V<&;dNLP
z<uEsL3EXoo!uU+PkDM1Bd}l>17`ZKbRXzX2V^;Tm_UGSqu(G8j`1-VmT-p}|1c7NH
zKn7w_Rn(EQXDoAex>dP;eb<UNlFs6#eB-LxZHgt_G{;EV76b%=a0qZ+JDP8(W2euC
zD=4`GBM?=Lktp5g__;ucqAf3%ErG$C^Ngclw1_cA?9;#tx69jb!IJ5y(&~o2nEVz5
z4iK=rjf!H;JE%P1iPJjV0{cG)@!O=5*Rzx40LET)PYNrkH$DWkJ0tB%i*vCH*GRbr
zh2SMcOOghJVr#<qj&D$As3rS)K0~AnsSzQ-cSO{J5m#4QA{dEi)TA0gKoFQo1js;K
zPFLYd>y(0$nUylokgV6+s#wgb@|DXX7@5hD6`}=!#YI2_Ba<Kds01UEr!?9Y{1{n_
z6mP{1SwD=N%IzTddMQNkbuxS_P~tdq>Znt7L)R|dEUVK;lI6?Y9y7uVw;gDfXW>>u
z1S3O3D;Ep|G$XAHSGY$C1>??iPe;+z4@0rnv%|O!OZlQ@-593MIPzA&^jZ;NXB<#Z
zwGC4+qToua$9PeQsm0*yO;|{ZU}UPpli0>NZ@RD;S<3mZ<<k&?M*-Zd(j7(LgVe!y
zTcQLbTH5PHJYV=%5TtllVvKmfB=-aXK_FoWkb$_7gA0q}CsYaok)=#VHDKDCh2=;c
zfpsfauc(TwYY7`QAx97p1iTPPnqb5pWZ8@MS-B^Ik^V^4y8R%^OX;lIhP#ZNi&xai
zThFK>e$!0?-Lm0e{hpPv9hXioCZ${5WDsgsI6;FX*MU88_DWj#aJ9UrPOK$6HjEc;
zH{fcl*shCU#A~pPdqLbG8rc(Z8U=%hjI<_v%&is}w?ZkdPh6YkJ$XGlT38fxU5OJb
zFS_HTL)sb*ffnBKbo}U-v|xlnx-OX9*?QHo7->N;(t(Sum4wF@VbJQ*N<kGe{l04J
zFDL6TKP+bDeiZBRxP6`0+I4xmqPpM#ojJBXZWrdVyxZJxyFLyGd`BcoFcP#gd*vN<
z;o>!=&W;|$J!o5!PY@6U1c68pxbnDKTg|EEBS%#>lTy`MTQzG<wX(jX@^Gb9-r7-H
zSGQD3*0o6BmP!NxK_CVMu&$lhSW+9uj*rw>5QUg96;EQb9H#wA<Q2h)Gpb}@Z!RIY
z>Vel@=X~>gbrfb_Sc{Q)#Lo-2tGm!JR;|PfMy?}H%CV)Tc^saZ=+m?qX;FYP&U@GE
z*%6D8C`L!>7zY6@7^ze%ss}*`Ek>4MAxXhV7VBjY;mSDek6U)%p~M33YP=B)Dj@T&
zW%%y2+yFAR1yj^9Z-#p)SNsr|x?n^H!to;1ZW78el&n4X*vFN+^Nn6b<(?oQ2nYg+
zL7?~SpQ=;$-KSEp6v@F#q<|oB6$E->qG|6ov|6q_^^{7<x|WzR6G8+5LBJaUu50HW
zxL+-2{mswKnzS1vQ_I#2JQ-wvXx69C&HD5;eH^t;rw=h|$9H?Z0k;fxYZV5*t+v|-
zdCWk#)c#}nRm0!47V2mmNG}6=cJZEQH4zMYd3hj^uAc_sE?gj)_3M7<`ZX>*nv61?
z&%7V^IsN2UcgkvgebcpRr+&C*_PF6RFHAh_MZNCX^P0bPJ1oy&pJdSR8j}|8=sfz{
z$Ehv$Tc_&oj_NI!2DD&J4SrpJt<zR%1V!eYYi#eRGy;#W<g<Z;B%5^EgX8XEF0W2W
z3s+mIXgtg#pC<-_F4<_au0z{jyUOBRV`{_4>0Yg+_O4x18=I>Gb5D<lS-*L{1E$+L
z{Mnny>6&@p@7(YC_A4L7eLe1B>oVtH*s^Z<M_re=E_3eFoJ?z9n{Al1>Hx~-ywYQ@
z*X(5ka^@3QuU6IW4&1NcM+w@#48YDrs}cJ%VgGGq#2>dl>Gvb^X<GPhYjwNU`$WGh
z%=evME7;!-p7UkUap}g6I&x%16%Z6&*VV&4euHQWFkJ`h<olw9@(d!c^V>D-cgx#q
zb!|h@iY$-s7$%;#J6v1JYIAd4Re?|vTozGycx0()t*={KHx$AvtJ9nAZvObrT*>Av
zCZ)#3>LC8Ojg`J1<#F71o1P2f&}8qwuFLJZzORqOYY+RD4Y;$Sg(r`GWiMR5p-Pyq
z43N(GY2U->x;a<*OzXL+pAp`_e*K2Ztip8`-gAvsS1lLQs)zSoA=OhmJD}slI^ujH
z9y{K_+G_*LwYeWN-W6Q?_>InIh5zdC@W7<o)UkZnHs7nHhxvHU?|UY`E4jWf>1T@P
zTtB!U<Q{<U30jA+zh3z0bxMydk9r@$I*GSwPt!oq`Q>yDZ?~>BpY3}LmeI+4*Xa87
zJ+GrlgFf#dm^u5(_J|?rC<2dB8UK71xTY0TS?kEP%-l;m(dYE6>bXRWnfONFnDHOi
z)jcSu=1NN{4R>pt2loBERVd&ebNvoj#L5;1dwab>-qpHK=6LIR^!(y)O-GPAb@?Di
zU2l-Q<K_uMyWUH4j#la|wbW_EVlnb`<AzGT_q{*Yd&fI|$?E61yYIG+T!gY*ojej>
z?%deg(~T=vRjq}`?3UNsZ^>^#KoAfF5`{qT{72xoQmLaSPYlszu5q%i4Z$N91OY)H
zp$KqYD`H(ca^g5XfchUA+^=zO#}71qaOi!W{&3`a$B#mNpG+lq7CcD2&!5c0r1yWi
zp8nPTUvur$*U8*u1?Jw;UY2zmO&-vw>#!d+&ALvOXMdRZS6|ok=)SUDCf?^g;=!X%
zng(7W9>kD|AC$yI*T;Ig@MBLGj49N?L|k-v>jM7hy7d^aT|2%yFOQlgdp%~JAm^IS
zAbB-5etE5N1T#McCZwJNer3%2_Axi_ndf0>H_P#Sl(P0R=5J!dIiTm0*`ChFe(7h2
z_uB|43+aYq8Q$0Bcx18;z*S|k)*Rzijz8xn6LIIM7v{Q*RUpDYR{#5C!o_H?S?M*u
zt^T0Hc08?iAh67W0b(X%)B$ss{ZHXT*}SiDH1q3or@6-b?19Er=M8gos@JW&7kzGJ
z%<F#TooFHUWIovOHS0~vxqV*S`)i*^=KU7eUQ7x-B{@a)bA(&`r82g;#H;!_GUqj~
zb8c&TSdX6TY?F!4p4q;2(e}q$A8}3BWgStTYXg&}u>(cY44BNEhor+yx~!&?ZF0^M
z3#Y_<?IxJ9Yy|5E6NN1Lon&3W+QR3?j;D2R#6Kj5bd8cYc8ul7{&TG)>o^Mwef_RA
z=dK-d`<x$#4$^L~f0X_>m2sns;~i%GvDedVK|q%4Gx(j_^}A-auj|&B>Z84m2If~9
z6D)^+{rPzK-08f2I9l&S%+C&lGTd!uQ5VOZ^>$i#_W_qo9>On7Ds!)->m_Zx?l<S`
z^syYvu&;Vvnf>HBlfGyDz#l%#`dPH@W4>xybRHh<*UY+gz94yYUi0WwmwDgo+-%cY
z>yVi^=sm<Zb$hit>OGTQo|9H<J;0uz3!w|=3+Iid->V(7F0-F{?wi-V>T$ZK>u|a^
z>NS&IdmUc)?0JLy9<@HVa$(*ntb-}&7&0%{$`0NUS8{joXJqNDI<H=S`f01ax4rF~
zt>ad!u2S~k>y%(JSqsSitDKq|)Ks7}kee+|sH-nuQD<&FW#w({OChy{=gNyOs@*NP
zo>@7jdbz%jPb;JR5d;JQK_CeT)V5WsQBf-!>*~yH!4}(#FFr5pS`x-iXb=Ph0cQlT
zuBpbZ$}O#^bNAe7eb`Y`&SL1}GsDsz#r{oe*6B*PucKxT6Sd=*kIBg|JM(hSr!`%z
zjJ7{olh?egwdSa=#-ufC%{=yVwnf^xw=_S8S(e(Pyyuj3Io_iN8$V<QwE_<{-R$M5
zwZ@IOy?;&`RlAL<YE<fK<LHK^;aGsxVG}q~8<0uUp^v&Pv+umeq<doaQ|AwIv}43F
z_I7#AsUN&<Ca1c1pY`qH1BJf9i>uJ?#I^oEO}$x*^SKjEPC2y(o&6r^XQGy(JvnJ3
z+VeR5Hv6mT(?{kv@AJ2lRv)jMWoY`JS%HSF1D(vPlc~YDv$Li0%Wx@zZ!HR*>R@JS
zs|soja$Z7gqKjHb-6oHmJLX8}djkbV{K02hYq|!hS3jdw^ufN~kOcOdjHe|qhfQk%
zdmv`E#lCNs>#7Te;7T5v`D=&PO7^wJ{B5pxrsnd95QHuk1gFrI1}MN4#heFb5v@&a
zUYq8*IbKe5m}@2Po7%G)UwfW<ZBOk~8ma-yk;3K%;`a0{hDtd;QynXgU=y0ie7{jE
zoLcb2tBv;>nLK2W^1Z3&yPnHzj~dLHLBul5$7|$^Xogu4?*UpQ@*Xv^_wWqx9*SG;
zdf7$?e+m!;@JImxuQBoIW?g({_%32wEK4g7_MM4z=%lr^`MbHhYgvr295v%fAJ<W)
zD%{Ym!Ll(*ixE5SRzv%LyzViJ#+2{yn>Vgo3bkv;R;@LTz0>{Idl2U5`p4^hr!&tu
zu9r;uMtyIvA53~Y-1Z~=T!(+du0{6u4c9eoRdiTu@H1kk*9ToU=ZwbMsU}*M?A@$D
ziE!Q8cbWSsr~FRWXfd*SbY1PjTB?E9_)-qmQ){ah)3X9Sg5okP0$|-o+W02Z<6=JZ
zX1z{(8IA?V!5m+E-*g_cE~j&ijnlm_*WB8(?<<_@H2bLcL~ixz@)|d%-+pCv8GU5?
zIau)MeY;5ypF{3P^ik8N%Q_v+_O+!5^YNL`f+@~z=HonI-Ia1hr6Dk*kekoH{yt$o
zzCZe3I8Ln%?#Y;gG_pTIpH-(?&37pGF<iSzmzhl5^f>FI&Zn`^Yqa^zq3<!T6Rzfc
zT_2Bb*Ljcgf{D1=_mRBU-hU@rnT$Vfb=sdRqloLN-+1YQ8Y}%StHA<0(J@9y@K&R#
zjuhNoG<31~C~w__CO}j7GjYeNB>_(<ZepAgH%WD@-YdmHGi;rg%mF@NOx$GUu?dPI
z*(YJW*;F}d=S5V%Nn@oB27^og%zTS{ZXb`hu?Rc28hGY%)H<}y7jE*k8R?|@&}=kR
z-eX;ex?S@b3{!UzE-v`b)I1DQKFW11?6`W-7v<PGjbRgWs)x0%fOW`B#J^Q*s5Z3U
zx|m-UV}@(u{@kNdQwt{3bW_G_X?zHBUUspQ=u$%nsfV?bDP`S<@EvtAWefNKYV=*2
zQH3aj$~s_-@U5;j;j)I-RHuz66(5j2?o#;X*I$}AO(P$fDy)av-Mzj#Ap-xw&n^Dz
zYop@h)G9;TkQezZZA5A_;a!o!k!8*23e}-3KBkeM_feIV*ryEKp)i4@eqY%({*gvr
zWPZ|N(Z(aoTLqARpS~>WAWf`~{mN%hjvo$ammkaedG?!2T;~ggeqChVGI7puyrOz<
zS(SSH@uShBsN4#G0LQZgw**~`?<U+F@#BFXy5<L>S2gy0VL#r@y3G$BeZ9;z1V8L<
zZS?(dkNRWK-Ve8WtH6gYO?29YYlc+_TbOhmJZff)&SxIAVC3l0Bi530{`?hn%Pl7l
zvgmpaF3ib^moNJ{>EhVk7zLwpdCS_D@S{D6L6XM?#i+m(2pN}^7jFDWZdWQ6^VcjB
z_I#8NdcQU<-7Kpm1m>k%8l9qe{zs|9WL{}nX7xDASTUAC3wyP_Mn@fAE2%weZ6ExO
z4~y$4bl5TXVrV<PGwtR5etSJTNshZhfpDh?mrD3Pps+N^9^w4Um((riPWwR;=sxRO
zIeBE@*$q_CmMQF<vK4pzX)<D&X`H>8Ya{zN-1@-pBqn|{jGA`#cGXI0MeXfXRILr(
z8(6g^{MNzOm%z8^i}OZp!MU7UpPX|~@2Vf24E}y{ePxJ1w|D7K#wlN5F!NIzhxGfc
zsd`P^Lrj_i-m6L26J6fkMITlNkaQE=4P;8lkLy*$KZEkktR=qN;t-6y@7fE)79&By
zn#P0Zw`c^VJ*s^CuqNYS^w&f%N+ub2C<&PA#V;5dlh1Xt-oOH!N`xO)_SU{3>2FW^
zyk93Yrnv$3+L$}hD%ZwCz-xo-x9M<e*=Y>j<`mb!DmDuBJ!o@+G24TnBm>Qf6t&f$
z-LaH~f6%>_B^XH;AcVl=&B4b{8$yvT1S;ubLA7_w&;}{0!g@(H_Zlj<T(ATI?P{Y>
zA{qH8NI^qqUr>^Qh946J2QB1V-dI)*_#mf1p+|uWBnN3|2T~BzuJwiUthcp)URW(c
zvAzaPZ}_`~iw>HmcX2F1V`+C6bWsR`g`Gm5Ea+}Su)xGRGN_aNV4~@N6T<U0{#gPZ
z3M*EZR10)k0u<CiVF>G?;HR)sRP}PLUl@Na;mE!qC|AN~0&uh%X5GYz^rD=VARIyQ
z0d3j-4gbGR?`yBQ<6ooXNFoANm;rPCmk`i$)aPIKYn=H(E!;Ic|Kof_YdVsAWbpOv
zon55`BT1$;O4^*>5mDAZ4|R=#5v?VPOC_uONniq)<8c1`WeENHUINVPqn=;h0)qkZ
zLhQ*4H)>9{E4v3VNbF^U{vLO24BF19`MBrb!gt7#VsWBO&{642dO0l^sbMR49KPv;
zeCN6`9og4$-i0PsPMW{w{c+Fj)>dYHZhLq$&$Q0BAsFdFFtQn=Qr)!rI!e1qU$30{
z#AmJwtF6uTWs65v!*@;!RwrwvRYjgwy@T(eUK4_mlI2RR38j=3>>0<=<aeTDo~}{3
zXw4QcYo7Y&)|Vu&U2Zf_>-v@zGl*7Kljcmv+e7XP;}DFnR|BrJ&b<1e17@(_Mq;Pc
zfK>lbGsX`ht#$8LuU|b;-RFn9PSj4IBy*{hv--q)yvCqxOlHt0U4}<5!H8YLZ<aI9
z^@fY)VmcV8-aJuYz|?~l{&Jx>5ozhg>)Yzgu?>@?(C6ch>BZ|ipkY09yQ37~215tQ
z3sRnaBhxeoMy#pD!f~SsG?}M5ke!Kxp9v<P2QFLo#mlTi(|2vJs*YiEPui^@dd-g7
zEFoALOts0MW&6CRO|A=IjM3!K{$cG@w{F9nfBM=eM=o9+Xf2zV(3c>!fM_mmY0IPH
zoy%J)yRvLa-nAB$eiqm+Y1Jm`)Y#Rg<Ye%osspn{vy2nHe)pZqgt_L0sjjC#*8P#z
z^!Sw{CJiw8Ip@4K#jJ;}C~BCaY2Zt9POcveT&KrC)4`)|*GZ#Lx8uZ^s8n}V2f~hZ
ze7u;68?|n|t}5Gm2bVi|Hd3$}$zWbC6;i5N!}GJY(idVNh*lPs3(YFlHcA(0Ey9Ng
zDGdt|3KJ-VuzdF8+_E0gQV8_#zzSr)Ob!+wxqQZ&TNSw4pzwhL3N0|P>V<%WZLkO}
zKbTmK!itrRJ`;-9_&m4pZ0S6tlNKtLkOXa007JnDeL0sPSZTpRg@TdV9<<7FpZR!?
zf*+P+KQinSu9HUA%|zi01tQC+gQ<z@+A^l!7A<NpYw#gKs~K9tuzuDTrC@|}fyrr}
zST|<mkKbC$Zd|x<MV&o+(sJLT-5iDKp53~@sVtv6u5UqH9B>`rtk-UG%=-8oa<00$
z;L+o@hxy)vujUoFpP=w9hdI*4J0XMlQotG<HtDrVmotCsb%XcZ&UHDON^fJ1A6Y64
zyx+K1kWOutq4Ou{sMo-tH0?HOmK%sP7?`vE90lz&^EtgE!juo1m+m{(p()5`Wz=NO
z*HP>6D(6;zP>i_u(ryiQVKGvGCF?v|jF>pF&zH|%8?dg@7`x%dJrUOrEf~?l{tK6`
ztFxz%>Ec67Y`woUpKX(tgMv6Q*F5uj6z4(SQ@R~f8On`Top;uclfayeG^(G}VKGAO
zUavMZvoXy0jEecBHH*CjBW?<oCN|TWYoxoCUqW0oScPK`Wp_(6D&9fM>9r(i{!#Pk
zcJ!JwYCW@Bj?ZOwY00t}X~ErSwpjz;tptnumYO%g$h4j%zqw)3#dD@n_pavN#jMwl
z2EY6Ey0l=Voq?4RT$A?SkAplFzdDnC-|PKj-m_z4?-wmvPQoh3&Rb+Y@_m@dk8@gM
z-RfrhT$jyx!uKZQ4K=~1+PeW|m7}WzJW{J6k3ZVN-y_SwtsU%*ucXcmTpRbtJ*yS<
z<VQYaRrCHwKYUQ#fMDd_hu?5;Lr(k==zp|DFPSqR0#czD1v{hLpF;l}*K07V!i#F0
z<n}NHt@9H6O`e9|$yrZP^kZBsiQy#Yi{BWzeboAu3v=K6Xbe+M@&-e|X$;-w6otsG
z?XqgL_Dv}8^6X%<MavPURaimPyAUwcYp~#en;K@@w}9w2>z1WR5Bbv&vS*Q(0);Nz
zH0oqsgW#YCcY>)rEKYFD!<3jp3ML9TXu*-e3!2(VIS4ZvJA0}Cfk$(v+82mGa6};k
z+qK+Q(JdB)C6;9g>L{!&sRpb#v_(lCLdPboNm8&z>46@mzJF5)H_&bmoA})7p+JSg
z8~$V6mMgFQGFi~ZRG{StHV+gAu}#*?MBz%K4uK}}r$HlwaWc_87_ZTFm?da|aHI)K
z5?aYDrQ2#_<A{|n>3^<Os_Hm|pOWH1fbR}WB2~bnH;P~+9xCUnl^w*|pp}<cjLg@#
zM$sFm_o!dJbfvWnYcHlK_WM-sa#QT`qh(t6y>R0<R}R9+(&~81T+`~Cv>Ois+^^Ol
zO|3P$>Xd`)nRuv(zZI`%$88TEKb6BR(v{Zg`i58w3_~+lS2zc`H>ILoE0b4g?4PtN
zEeb{$14t_gOseC;KbRD<ldHFk`LYbJ5>9)fORzjW?O=I2Ickz!vlhN4o67Li@!d9c
zixHD-C+%WW%fQU*b{<4JWC{eF7{ul^^W3dX`g}dQS+as|>Smz^hX4RT07*naRLMe(
z4f;B#w9H~8pfEMloq|reY4ua{s0BvX?$vFzernUI%cPz3aPhqY2HkQge;{RZqS;=K
zW%V^p^Uhvbl?qFaivsfpNfdxEQ3zs=r?!p>Qpcq7%;LwJ6Be8=ahZX7jhmNd5E-<l
z4`io)jB|~PS9WD}C~&P;onZt@#Q5^HZFTzihNDE>tDQx=XFlilzB{c&I*-%GmD`-6
zHB$-!F128U0b?mRXv3us)9T6+EJpfnrfAK<l-l>>Zwf|uJp*5}R)TP)4Ii`=mY6BO
zx~!VzJ<CNG+sM#QCh}4!LAO)1tmwl11QP`rmJkHjD2&KbdmS3}jI&YMI<#(~@WP5s
zf;`1F)W2`7!aB0pq&F&9HyJYp?dW8+z0|k%SX+my3%Ckn?3oM%7jzLuVG7-8Sym%(
zsYTi-<Y-V(gH%8}bmK*{0ByNq-9Qthrwf{`gb(}fzJ&{aQ!uiU?ZChKVc^+*#z8H9
zzMCnaVSKFPrM_UooC|vFO>G}tPmrT7Z|2n&7501PI%U6SmN(C7DY*h`BWl&ry3Zct
zLD!>?x-9$1;B^f43up%{%XA)HMj!R{X&%RQB}wblI`<<t_teT7gmJVMTOM{%6QqwO
zE?0M}>I9xSowUX^=Z&^*)%Wa2y@xTcn{}JlXZf6h5xToOf&jlktQst<8LUIQH!AAP
ziQ(rsEEazIaJ`2#>y2U@oZb;;8}oKP3TC5Vw6%2umV_%-pi?I%fFG{*J!qJ{m0#U%
z_ea4-*X2~!3pXto>A|W-YdmTTdtGOcqh8DW^0K{2?w9@A4|1Pt3itJNyQSard>>Q8
zv$9=PM~<u<?8lw(4%7zT$)>hwkTOB@kS6;(-Yl#8WL|SRH|GHvu-kCUnprNY<s#x1
zFg_mk20Gb2I?!{%DtHA^2@&0J=DU+Fg(hV+>GkZm?cp^h(>ziz(%pBZ<%XR}UywXz
z{y5LwXyf%JESTv1!7DkHx9`=Nr(x{xS)WrmuXcEkZg=*s+*C);9zWoLIqlt;m-|#|
zh`VW<Gml@QK!mY9w{O5g2yW;LU92nX>sE}BG90uVg_0WgmTo+`ao%SdAFln|_y*L<
zxL$iGUB<6pLGGV_8LlYLg?B~iHXcFRH1p_n)ZV8pc%;d~5@QwLTu#ZJ-^@ena2=GL
z*O}LCE!XYq^4zP?s=?eh>u(OdZ`EVtB^c3pbiMlMMYql~%2DsJCRKk>n%t}pwFP9*
zdS)RX`3R}np1K^D#mIZ#`<~uA-tmhTHs|iXdloQ?j7Dx8>E>tDL{{18rQKVls~~(k
zTQ8t2P8wj=<q;r#5^cFtA$MjK0d8D4Z>EvjU|54g97+53=ZpPfxA@|n#YFl$U|JTy
za36Q=(fv2a+FmE|W3r>gGS+qDhIV&m-2h|R9+Y0!&9a1_`CB-r2{c`#b$b2Etia~1
ze_xP*eY6$W+$yBU4Y5c#M~vnzMn2tSFll#d)(`yQb$yMlrkM0^>zp(}=t37-Ixq8C
zLD+R()IoPtET^wC{s@H-x(sP&VqFxrSbafPc$MW?XFoXm;kAQKXt1ps`hP4z&^`{l
zV$}iR2J1GHRWBGDI$5-%9O=^Sp>W@&7qGLBjV>q0#YBLQ8>cQ^J~CmJsxDpGVZMgh
z-{@!DevArLx<j1S+TBLm+5Pp!tv|ZoZugz8^Re$>4NU4YIZB_L`f0z$_k3J^*7Ms=
zL)hO=STtarSU?OZzvMQ5xHr#02+#eNz5Y=!8l{Y;oj7rd_updrs{7GSFXwX5vhKAG
z-_h9VBZb{$e7Pkz44Qav(p6v<>lzg>^xDQf3HKnHc5_c;p1ZZDuY1k2AZu5UJi0BX
zqZ|HV`k?D2E!<J~t$8~CG>@Ez_&?A$gMH4L*zg|b0N*hUD3nl9g;ux+M<}Nd4|AM3
zr}D_>m%@&%olaeU*m`t1ee^5u_Uy2{8%?_INgjjH9o2W^B>ORL-C@VTZB2-q4z@?y
z?a!(CECr!SKd+pp9B*1U>theDnkjgZ6fXjknqyJ)g6jt9W76;Y{;%==Hu2q`$M=6!
zDeQ9{RlD|jtZzWP1cSWO>}8Ywn`_3E+#QjUc~+fQFF*aXRp;B@`c3@}G=7cjO3O`3
z7b(>-R5p&(Z=AAWx>l)HR1tobO%1ZB+8(E#s3h6r&=3a4Maer&lW8xFrA5Ji*8Ax+
zhMZF!Y+7l~YNng8&`2ZRRFzf}km_sJN*Zpjy3j^iDyCskfxzQn)LI{{umb6EGmMl*
z+@vmCWA&iHWBtZGT6%O5H=&C<En&cZdzQN`1f}g%_rQocv?O6bcgFN+L3@g+Q$rAV
zxai8j!ft7E1wrBA`brBT@KM~i(4gzCL$_D7(y)S@bJvfQfptlT!O}sCj`Km6R<!z{
z6$mXxQV0m2;)gbbTugMaMJo|nn54n9>B4e^iB=wkWmrGJLL>{}Hb2yR2=Z<vFo>A?
zLE>R;0w$YI+LFYo3(O#mYpKvv>kbwpK{a4?{XE9K?z(BXm^1}Bw`(2+DYFQ9IG!m8
zUegesm6ldiSiJ4pcwTdAlrg}}hjtlPf%!@0Minmd8tT~Q_|_kjKxI<TxYIjg67<h<
zok20;r@v<FQ*(M5KG*r%-0#ma4q<8c!j1DegS|)?S6WeitKNim?UDTe(P7rWsGdWw
z0O59=gLrAMmkILQZfy~yj9;E8yat%#%>8N=+br(CoGgCqeF*!m^y`n?{Ykgu_3TXZ
zyU+~}eVwtDI<OdV8Y`Vgj?p0y74If2MVx~<crB@dr`5xETz78|F-DfH@1+zh-g5X(
z<d>q?-ZN;M=6YZ+qb=Oc>rt)IW}R{#4Fa@wqcX%QzU!InF?yyopZ$8L=b%0^KdrPj
z4IYMH9d7p*X}<Ek*{bYL$CXx;F-9oese_5dfAgBi-iqThzuj&@+ZhIH9$HqTV%$b7
zfGl1{lI912p!1~x23i60D$3Nvhhu*K!!gQwr(KsepPV)wakTF=hApf?8HY89xYVn4
z1TODWW?4nhB4*#ERUSdd8C+a5e<&cKE~9m%n<Tp40yDR7O#vYT*Ek3$Xo<qb)e@*6
zk6vgkK?Ma2a8X4ydB*0TWd*H4C>*ika#-%Lu)T&Lrw77#Wa4=S@jY~KbxX}2Do>WR
z3EI?x4+d`6!O)qX7AAC+#gt*+ARM7JiMC4VqTVd}Nx2GX;eA@8&_x$owR}2}2Ignm
zmLP{NviJho7sj9<tGCK^_ci~m3%sshGRU9HAh<UKN>L=O!5g-&Y!2=rH;<q{;nt2t
z*b{`sGRpj1&nc#BTUKfQTJU~t)8%M>0(%a>c4z5+3qsx&)}kf2_0x}Ol61XwRF%#5
z2P$xoQo0*S=@2PN>6GqnkS^&iIdm(X(%s!H(nxoA9h&=~U*G#%>#qAB>p5pF&NH)T
zX3r<~-m%epg@l3oG2cZv=9Y+^f#D!_pc)}owByI`A0KCzyq(2<AAi4#dl{JXMUMUC
zQCl#D*1b@0rM26|vNHz9+UW|xky<~p0f{o?&02I|GLBYuvgo1<++tN9LjbXp9CJZ|
zm?%EO9Bl}TL8Cum4etcy&JU;Zlts#MPA9JTROOETt1FrKAiBkxhnmp>8+)r;cuoY&
z&QBtg%UIpN<^rV{=+_eT0$4=n9ODh)MLQ?oy84r41j5B)|9qKErB%wkKyymOhBEj@
z>4%wspgqyoj1S9mg>PEN$9*k-+4h5mtd5ZU<Hc1N&ms$(c&Y#)(^;pKg^jxyk!oF0
z)!vd38U5t5<o}P<!LMB${&Oy4;dWDGJ2fPa+x4!*cQ1-pkTi)R_U{(}^_n6)xRN@_
z2(BRc=Z|mIy~dw0M=&R?g0MaXR7{RpNw0Gj`9&idD2^l2hYrF{;P$y%#g0<?m(8^A
zn3Ox!<CoJ@>_)`3sge#K++ioFg4PZbBc{-7iq`8)>Vm}WJ#tRn>DWQQg`>6rq#N03
z5b!E$?dsd@*#!i4w`k)Jm-S2pgM9HmuGBRiD_n*X>8^$xAeemC3)K3<5kIqC6wrbY
z5LwadRPCm4iw*+|1i>^MN=iZItZZmAh>u|TGJ{Rg!Qh}!IF{X5!i6C7SJu($cl|}|
z^?|`s8?>M|MDT(NsB!dNemSsV(Z~x-Wkx{^1lb21`&m(ra2+$Fg!3Vdd%`pMi?K46
zKf_@nbVj#342%c~u^SGC5k+^bHyR>6wLBg2NQOP5;pA@_|JW*kAsl9dZkv2iahiXC
zXjs#w+`e^(rdHbc_TS)o!(SgFtnE~o<Xo;;#B=YgQ&?#4XN(Q)p5oSB%~GxZ+Ej9$
z^FzBGUv6m3elJ^0-tQ?_(y2i$ZUBx}4xV)t9F$BmB3r?l7uN)<v5>FWe`gWhkqq;T
z#s*OqmE&g$l=vn=2P$!@wg}2+d@{qqrrk$csT~1Zp<$*|Y)$ODD)~j*X?IehWt&f6
zQf+CVY8<Mh8aI<O=Kq9!i_baw?{2;6{gh3@N5f%%a_Nr`j!h+)nL4}4-V0M$w$O@K
z{5K7Nnv4VFm49bg=pdzSMaXkB%l;dKv-1rOAS(Ski4DLcx*sbnR>{op`nGf#<%|a^
zI7>qXsh64LOvzUyV&5_vY)9kqFqzK*9uhlDRgKkXhpJoVGC75I)HH)!K=;H#J|6N`
zfi97u1KtpPK}z%XO@uM7RJOcXNY$&E_iPm|;3r}IL+rH*)+?ORn)Yh{`w*(nSaul9
z$?KNWhmNK59D!ZdmyOypmwV22XdgUaPW6-0AChzS*}{W$MzhhuJAC187hQv8Iy@80
z3~jv&hmSvIB={9;rc{bh#aWZrf7sPoQ*4Y5Yk;-d9OW0iHuWzHb~z|Pl4Q>!b08X@
zUzN(YJ-D&0D2?Yv(COepy>HK0ZUsI}3-jOSTgZS^T9A-v8eC1A6a{9fq<L%6A#Oub
zFVXkN3`;I=N&MZAMAduv{pLL?Aa~L~xkOHrY8)T$nAtUuiXmMI@X;BKeS=FDUYWod
zKDkjEs%K<BD>?FSVf&DxLP2{;qXI6U7nMq6|ApfVXfJm^&c+Np^EbF$!kUKYK@H+t
z%a@YMM7rvH#b(f^d3r|BQzS~RR%Lk+@d@-ux&C4H=mgLj%tERfh$?W13FKfFp|I$@
zRY|({euVh=?0Q^=iE#!%J#2xq1>Ln~^Y$o{l9d+ZWWo7eQSi|{4K7C+MzMU~Ea-Ml
zM;Qz0H;jvKV>r?9CcS|s!?;#b4zs2qqaRv$0O7wF0akM~{56Z9Fmp`5K1SC6#uMb*
zk9IPTA-t6FG0-Sy%9b~Eq_8H$5COa|+T**>&nRNWf{v%pht)%gdrJ1vS`HT&wABj4
zI`x*b2#xuL52)bc_L9`yansHX@k=(p7u!-V{i>LS7V&n!Md6`Bp@Bgl$;v<ere_V9
zU%MmfiJ@PJo)P!@n-au-`=uTEN*#5GKk<lok%A$oL5DcuMdTcyeHRsx9qjS-(PoY*
zg=m{Qp}uM<->`a(O6hry9``{~Iz}r|XH<kaW<&lPy%iv-q;FMd`nLlB-kMM!5)5Gw
zxgj)t&%ZJ7#S?6|kwX>Z=ohynWfn!mCNHwhRXD7jvgalcn@}NW@kF_}5*QcsKZ=mt
zp+^!`QQ;6htd0AQ1_H1)KV0$>X&v(paipW*1d85%FxcSU2*7c?hGVA<wEUtW+U|n6
z&TipL1SD?$fgOJgCD3oAqvWGw?f=D3`)9J1K2S7o$LH|;Pjvsn2mIA%38OfLyWnfV
zbpRSw%R2LTb=jAyR+F;qvjVW|!1?`ncqO-!BUS}H6;Z866xbCVyzpacYAkn64{mct
zn!Mb9_KjsO)>QG7QlZ+Xl0im#5lO&60=OMuMm0OOU7H-Mil?U&Um9?QQ1(m9p{TAw
zEU=XOvAW|E=!DL_$r(G~qcJ=B??<V3GyjbsDi`!O1pDVT)+oZf_WxO9z%C=}2k)L%
z-d58SGI(JK?gk<yVd=-&Xr*2bC@lI*kGOm$<0VK$KoiV-eQRN^tn5<9fCKD<npmrn
zVF@aBh#1*w?j|z0RrA)wpjX@VR_yo|RiEIw129JN`~3T~xJweQO745n9OdHl3O>9C
z0m^TJ$?>RvzYmyUzrQn_ok45-UpW8%_-BSAfeCvNl%BrTK)@^eN^Ydyfo|Ppu@r&d
z+VkqlM47o#b4kIX;az)q4J5qH(l#_NIae04UhnK#X2R}JT#YH_ph6&c+cIJC!^vz)
zjy0S|SEfTePfaPc<*P<#JzK9hMn{b*ng_3JM?Xb{q{SkSt!8aRy_D>}rh@rW;C@Z6
zhg~LvD1J+AVO`VoEG$vJUHVMMVRy6~h}<rN-s=CP3jKc<B`a&S!l!@xT+&<6{T0a0
zQPdJw49kPJaF$eG+di<zLIsu_OAT2Zj0PE-9{vm=lx(>lhjkh%F-2!k`{jp5ta+2N
zr|uVxCB;1cb^{fTW_wDcp)Lb{=TZwWp6q}G>O`0;>kXNB0kQSp?bHMsik}Lw@|YOl
z28#bXZuCAs|93+U)C}kmamzgY5N24D{p#w9LqD`w<Zub}<ghom2+ReZmA^#BB?ZF5
z%*|sQkg-=-i&W+nM?R|<@3Y7E*=T^P&s}R`lC?X;QPw8{>B_;J0q4mDt2U8U&WBDK
zFm`60b&QlTJ_2|ufr*OSZzwm@roOtJk&{&Jc>rf(Z3MapuR=&JtDqjP?`MgEaMgU}
zRW->PXE%#GL*A){EjmX&1?$&3`lqL>-x`7EW6L+Ym6ocTL8Rqoe@C>^j7{iP3bfI}
zbnNMh4)PTdevS+{s&COMUBP`8?@7dBedo%X#u1fhgsGbJy6Fp9BntI}4Zum+?CIMY
z%$<e|Fla(8g_1*bA3J8Yt03b=@@V2`bL<xBIa-{cZUYKJofu72aSCY6pyoaI%EB*a
z%sE<?ss;I8g}Kh@sW1{yjhDCS14RpkJoh6dX>J>L8ik53qZl_toxE@-q7h+AH1NXU
zzcGvk!#G7r4CQ)Ehc)D`yL2AvjUCY5^Cy<!-WcU5pS~_{XeJbM2sH<^-l2GiWZh6u
zO?nhFKRWz!DE!diiZ{l|pn<ce+dzI)n7Q%?InzS~tZz7+{@RN1re1c|Oh72bW6fas
zRXcwgoPP82SuUQ_kqn-umE_{}Csofg-owm&HWw&xsOnQu3T=vb0sZ=~XnB39Sd*xn
z%vZG@RJqe6JDQst(?3eSE3`277Gx@|yU8gtT`1iYq!sG!7U=YeX50hEux~Wl-yPn_
zELRVxRqD?FMj-!UY1AfXa;`|h;<7Z3mQbc(c>K|wSQ6*5h_?n}Ab0p7lovma#qVJB
zjnK!!=iC$qIK_RVoec87MFURoU{p)#7&DeN%3oXuz9}Qz(0O_Rj5VA`nBZUW*xpj;
zpcGVKUN(ixw8DsiBS$#0V3G!&ko4p&e|n$TWyk0z>cg)J@!L&%bLlQqBW*{8)<Z4V
zhq|fq`E9mp%$4WDs=SZc4y3rhkGU6vQXXQ!+~++zMKYYMZ#&@UKV076IjynXL8G}h
zUCHlu<f_7|aBn|JiCOUnkr>Du1dDqiL1+GbX^M9_3VK4%FRh`s?slw5YCn`{JXs+p
zQ3OUg4v?f%`rx|zba(+vG3Kmp0hkM+NS;?3Uk`ahs*>n8kgf}{PAJP7SSqO~HceBf
z8qoMPpPe=gGcXjWgeSc5_lwqHDf?PbtZMhXswtd!b$_iUNdk+S6>BzkOhj<ys%+qg
zBVTU9sL<UB*2>0w(oMH7G-@wM0_PSRKKTvf+?le@PQmu0_WZb#V`oH#h|bNSdi?w4
zOTB*nry?S!8<g!KN8bbjls2Ax`+a{>5k_3WOEKMamAVIia4x%>509<!LBd>5n#J*3
zx;!bMqf6y+m0Oe;?^BVL(`F<oVY0o3CoC>+imG_EMa4VqJO0fwQf7?*?Qj@f^{B#w
z?)<m<X&3I^mq&4p+nSZM-bUWKzK{>g@$uVc4fBeYWzt2PnzlB!^ZJJ{#QtXp5Zm*Z
zAH!8{s>_qfUx_!V(e|8f=aXE0Rd3PIp0loyZ>PmU;CJ!<S=ej1m<J?KX_TgSd5N1M
zG09<7w7AK6%m&z|wA-YxZ$8wX!&mDpVuVGj6ZR*W%bv|)v0)<3FXX_IKdp~JiehLB
z=~A`h=XTlMY&>>eW6whP?q4cXZ=)6{*c2E>9W+u>BUCU8F8qqwmk?QLJs#BP7(2Ys
zzW0xCJLS0T3_CzfI<<Mh_E=Q;F@egclm8ThVgKWz4xBg1-DEZQjjPd`;;$@vsaQJi
zg=DEjS(MI_=|LiY!ml*2<Zn+vvJWHR#!EV~S=Im<OoNu|eJd%&Y8Z&R3N+q#ROyzr
z@0BW~3lBxpzBeQfi40Ov20D@FK)v*tnFmLm+zJjOw=31Ty}fjV#O%nR`&2B@eG(~f
zzl@L8qUYXhAzAz)x7X4J!byg-Q5M--KJdZ@zF_J?-x$B+E^M-~9fE8h>V8+I=12PO
z-%%3G{nSKsd1f;b2|ZXaDpSPM+S+!SX*?X<TJqavAAj+aq>Nwt?DG|?*P2uSRst~d
z$(QJquB^=Lp-t99UkvjM0m+kO(Y}D5$E#+vZwrg|O$F1HzffOAG3qQTn96>B+0Uh`
z4jj?Vbw%SnS(1lHqH`sAw4J%rZzv9QFgbAt&0Dn^)m8E%iwo!gAFBNbQRL8k)81r|
zyRB-K^_io2Xm+GNi}TuLP>JEfJ>VAX+XZU1e&udJez^`B&n4j5yBKro`<4o7oYC*u
z>^XNSrs)lf#RfM&jiG3^xGvY=1i_N8UnnM#u4hmi=lj0U7Kk3R1}Yo~FK6d=@;+gb
z<khwiJZQ#_S@#f+CF7iMzI_=QbOtouad|ve7l8_uP1Vq4uDJt0F8p%2?ISPqpZZz%
zI`TL5(r<;PbVo)uwJswsqNi8zR*wUjj$(kjr$gN$mHzqa@Eb!K+q=N?9dI!5<6Fz-
zQk^cz=}S2uyS$<;`Rv_Xly=W?NJfW%D0XyrL|82LlEXxkMC36&4J9#m;d}wYt6t-!
z>wQk{hnAmAGZW+Evz$hXjL%M9GJaUNNl5>*K7)dEifoMGz|v`VYVZDLXzzf5TDfLI
zC3m<rZ!)ABu=ohC<f!R^r1P2>G2?cBZnL6<MN>sEsDYFGqD_#$#{VW1etA3at47Sd
zS-q@Cx&F{VDppoj&9w#(Q>=_T#>=K`0<rn0`Td1kTKBcML>~{0Cl9J@>9!w&v;Q=n
zacez@Jjb=)z$hgG9>a+6BIn>8DkrRE1W&8=mx%)kTry9T2l3M5#tzW~hbQ8r<i{@+
z?2ptPq+`sA?iapwTbLy-!x!V;M?a9BrnKqeq%p>g8|G)3W_{xyvX_??!U-nOz`s1F
zi@cfhzfz-c_`Jbo5mWwY3j3VM0Puw|1x}c(>#-QeQtX%GbI%c%oPp6CSSE<E?CxxS
z?$n|UpVB+MG}lVp-yZOY#Hg6&xxi(Pa|F5}0rV&5`jw<4szETfW?M>#0QzQbrt-cB
zsLIed@<@5JAc4d)a1kLn&xsB4V62i7nJS$M29bvuu&~k|g=sPFU3#Q&HTSPGkF)~L
zXP}AKJW`;5+RFi=CuYM2hf4qIOBOTg$BhFEYK73y(l3f3Eke$`4=xi$Y0m2U$7K`8
z63pVo&H}&a^a~t7*gPg#1XW_oTgT}8KW%)9!bf)#jnj1=t>6{CmXkXSrujo1ZV~dQ
z02(P-0z^)QZ_h{fzkQaDyg|)z(7la+Ny76u{nmq~*yAvxa}cTRC5OgS*G<dFi^DPz
zFxwD2<pq(M;O&N7U9p;-ps6$dNN$FB^k81GTjT3)7D0yuO51L3OibJIl6P}^PUSNO
zJPf9iyJ;81YAx6i#Iy<_dfj+*4cwl8#_ba>7C8gh*&)hH<Snk9#i&#}Di)ROVT652
zP~=dq=YFz!I18&}8_<2^Ev|Wc3>}Q11QcU6kmF;yQ%HVJ0R_uXr{^?MwUkOXmC2WB
zbC4$SAVQ_8`k^GL6@P6uOWJ6meV;Nk9poBQBCP77<%bsXxyOK_Q9t&Z37T*DtC)!p
zJ3F2L@~g@EB}I~(qXZ4^RE4_7meo6x$1_FVA-P&tt8;d9r?qhZVf*vehm=zq+<=SP
zvOz*Kw7K&W$TYe3%q-SDr)~pVZbTbP_Vkg2I@Hi5_dw)DZ%rC)*CM82mXdB~UsI&u
za1MyPJ8mxf)dj#T*v6kKy->v?TsKTlQ@NDlb-I+}DZGDIPk0B8fc9(T4$uGn6X~hI
zK~G!Ag($zR-Q7d?b|j)un{Oir)%0a9b+74@tVz3@NL>{~NX`y9c~|6Gv5ep{39_Mu
zDh~Y@YM{`M3wX%`eYPZb<(}^TqsrU;gx7t&OZ##FB2cA&!fyWBDlem0jU3a&*1N|u
z@_R-{3R1PWx+iS;H5jbXK$rWqj}&v24~7-~$mNHJyk|jCE#I`_%+hj&*Z0j!S1-2Q
zAMe%^tnZyvwV%RDYgDtHh6Y9l)ZBaM7eW+oa%vW<405!ud8?tqRIc)0Xhe8PYNS49
zfiLHazgJg8UOV%PhXdK|s%k2%7+ms(FlrJr^a6&*S+{P%CI+<<_Y}^SlRK}n0O`d$
zulL;AH3s$ItSiRj{p(lOuSOV&?<Jd@q9VbU*jmOFT6oq1h`+%W>l};n(b3ZbIm$e#
zeEs~Z)^LSso`;!bMH#<KYYwJu+<SCxnnC2?{c<gu!kqeSmOFs3Z^m^7_puJyi60m?
z(6rTr(O4oUkitV5Hl_u_8d=E!u23Zczi4N^NK>4d6J3~Z(6AF)jwC;5({46B_3@cb
z)mQMlE`{qlxGLf59NofM0$<XPvUh5*eM$)3Y<N8Bzf!ODCOX#lXy%Cac0nS5$^ps%
zu%j;f69f#Op)V~yG#7oRjQ@Cpy|;R9CD-^`XPXOScfuhRdX|2gc04mMFb_87A-S$B
zStan=aCDn;)>wziK!4QqX~d(^y0ppFEPB#c(UgSas;5}a@TVo$g;d#u>}92O>H)-w
ze?q2PvCFMhMZk6%L=i6XN#WqZo)#xZ)?vi4kbI}+$KKfCuBqJ9`@6L(`u*msZ1w5p
zI>jXgRfX+gDZ}*TDh+u#uZ@o?yxhIw!2Q8+ofd1D6IvgS0#j$<G7?VvJ*uCqb^%Ij
zLxIY1&`^Gz>s{z3+}T9itaqlxwX8Q8h*9_OQNQGTm%N60NX~i6--g7IBlCEloSi;-
zChc0#>yH0Glacd)|7E4(_tU*}XEZP|lTi1CtyC;JPyt9n90nG3UBin$U8W#-X_Z$8
zOrprUynH}-M_~K=bfjzD#^5ZJMDQM+d<#L_q~%ImNrc_uCf$22Sbnss>pT5ZOw32r
z5I7*U@Aj{%MrQz)LNK*W`eE<8<vYx;vslvNqts7Unu`HZWMAWg6}q*TljQeNUWC0k
z-kL@Rp|mZ@&pe?Y&g%^IDBc}^!Rd5P+bQhxE-taY83#9R9yiieY2BTdx@{%SCse+S
zx9!XUGF-Qf5=IkhecG62zjoZ`RHsv=1u0M8xM>exJmFHFxzglXEMG4N;jJ{b9n!mO
zmd{Nlk3VoFzUiAy6Kysm>QPvf?Az_W$J5Kx>way?+RU`PP~)c<Ie$elSf>PszqLAj
zAn|Kel^XAei}_v*i@6~@Qn#tuW<=0z&$1&#=;TttT%_jYzV@WpV%03FF(#@nrAs2i
zRL5L=ZjQZ-le$6IP95yg?50dF8uI}ypmu;9H9=4)QQC3DNKVFV|KZX_XfF~MmjESi
zik5+e&r)q7ax(=)Zj@HG4EMT&Lx48o#rJAfTgmE``o~0~n)5GmHAB22hg~Bxc_r^r
z&epS2Z>&8~zah)HOc59J4hv`Y)3w2TM^$4^3cV@y&M4A|PvgF9_?00{GA#UB#Nu<K
zxq++@vrtgaLkNSAZ!9nGDTY;#Pw_fMV{Mko7Mj|@FN;2jFZuwe|I^^%F)khgZ|gG1
zPl3x7Eku%m;Ln(2f34>5$xUw|*4olGB~2hhT_Z*u9TfBR`F;$%Be#8+Oky(xrH?=K
z*t0eE_-1<56MH$9>E2CK-bX-i;76z*Sk;qbL9v6jOS<}frm}g{cwJtu6w^_nV^(gj
zRQX1AAGiJjv*od~&D>n<??71wn&ffqEnZuM!V-I@)U67Hq3B32xgqonGo&?X1QmK`
znsNp+{U_7BIpbX7yBQ$QV2M7zenv~X1)~92{q&3F$SLRCr6qasRk5z;)m;?@E{i!3
z6V;8Y7hJ|B1UD%{@l@D0)N}#Sbg8R;dF(<diz^Q50w+!y5jkNZQ}abXayPNm!0)EA
zx^dC;C9m=FXU>_Yt^FK0Q+9Re6VGeQn{|o`45jqG$&mGs;oh|I-PGT<lYF4^i+&3(
zdzVZK>SPt)LEN1#!<|N4g{yhUo*X$$h}9GRDCFD!o5@#Y^?OPUB8j`{5W3sW{fBY5
zOPFPpAv8qAlM)X!l{fcaOh1HY2A~A;v>~%_@FgarVBnp|k&1qOoXxWgfomUTA|x>)
zYzFIXZ)^Ld$^o-W)@iw%N}?wlWx*=K_vl|jmN}{)*_j%xe%jtPTrh6s|6t4j<j^yN
zU+5v;WmWCHjJRXQUmUAAqjAZpB>#Uh6sA;P%54;VIYS)ysLuy>X1;c-ZO#=qh)krb
zO^=Ut8}u{kTgA(Mma<Z(U9%h|oh{QvCutmuU+j-hDdMne`B{qN1UG|K+q2>j>0p2;
zSWFu%%~QgwR!uuD7*WliI&0{7c!~9y2G_Ult5>HLqV4oCNwJ&(_*DiSKYGBQ{8?c@
zJiA96<*4aEV`suBDO9i9^@5rK*UnnGv^V?(EP$;l*JuXj3AARo0lxSp9I)exq^0X}
zyvl;3g!_P~UYUw~pcWa6^^4f>X|u^g0nP?%HEr%G)?>fuJd2{`w(Bf-(Uh0Er3}c=
zo32aS@6~J^TP}oGJ9)c;y}jk03+@P_^c$xv(sddMbb%emCQdirPLt<PD6PY<jRkv~
z>mMv{JHHiqk1LNp^wk!<r^I{0eDTVRA*+<yJFlc4ml;EwFXiq`kPMX_pa1IWBh=Sq
zZ*;%tH^mG(cZOZ1r8|vaXe4(n<k+<G?XqkB%Drn-WL8qO;8?wYna~EsW;8S~Dv?B(
z&tEJmj2X$KEb?rrFNyNHMUE0hLzU2V@#Icz^u30S(u&2|Ac@oDNTef!OWtkT%-8@C
zE#OnNVo?4_$m-sBPLL^n_S@g_SIt)6Bi0%$r&>w<2*b2gV$SG8@Z%%couF4Xd#X^@
zapLlV$*cgr^eYgp<bD5e`RX9Mae7gzMVIHAcAW}OobYC0Xr3D$#wV8|mAH7YZRUs$
zVwoI!tZuw0hAimpug{g1>3`;P(u-c;=KsUqFYlIu@5DUW4&;+a_0eMQCq&y1$rDP^
zb?(DE%5*GVw;oaJwUl#<vnO9fHs|Kjyko}T_aM2|hbKc7pO9|#Lo+hbggiQ!A+}na
z_@|Jpf#rla)TJF^PgfwOB`ZslSsr`&sBk&@i}v>paK$3Bo=bYE;RL*4H#U(Zoq9Pa
zLLmlqHe10>o2K*AofIJQp}O3ukNjq_^qJZ(X*@6Pe7Nqt#Nu%S76;1efTBJ^<cN_6
zKnm|Hya8!`^OqiIj=28}ky^StfuycL_)Kvi?;+mD<x_<d!$rFOJ{#UnQ}O(}UwQ=%
z)$=_O8BF;eKcYIm+6z*Xu|v5kJnkUw&dXWb8k`uRjdM?kY<c*U>_1JZl=5~Fam4vx
ze^5@IW}K^ht_B~cBLvp}KAGrG27e*UHeML@IL9EmX6CJ71&}lv{c^R7>PAM~WS!fr
zaH<~|>Nhf_7@7d^ZB#Veb8T{|S&M+Wbm)U##*70Rc(}_fVj2z&3=aAphYk`VM@<s+
zLs4q{W$@{jFR1kcEDg@nd<JTuYcCqO{KLBX!jos0SNXiVB=6%<!ZU(+*IIP^Gju!h
z!Az6grV$Cztl?~rM37nNdggsB7A~goVQ7bh?zyblyH4-g<kEY2-TkrVUI*$R810WA
zprKv{ie=~2sOB-eIxq%0j%M5Vt`bDZQvzsUV9sO}&Ho@_8aDmf+qcILtvrQ&7`IC^
zRiP?%F67hjk%H(DL8f~RF)Lm~gy1d}nz}SU`=Ro%Qo8Eu2bV5)(%OM~yRxT(`zTGO
zAveeq@e-r~LDZS}<BKa7l-dqg@%_a?q!75^dfShFcmU~z=R6Qg`e#(53h~TaKEH&j
znD`1d&ND}nSfLorGF8BpR$yk8*t*ZrHfP{AbX7QK_}t>b_!f1Q1qDjN6|TY<)kB+J
zwDS#u2x77Mgw80{A^s--#dV<W=x4N58xe}-%nP7tLX2KBu#ovS1|Om)Oim^ysMiT!
z&mzu9CfhgEHYNWmOWkJ^?mRW)JLNiH{;jy@^jcL7=8urLs`o>iu80sgs3xq?Or2ze
zC><Ke5xH_&smpyhU1v~s{ppL*E_tPBO`c2aRW}ljeZ&&A%hvYcy5r{?F16kd7eAYw
z7Roxg)kv=SBkpUrsNUcTGEdwsuj#S8@k4VE2ktIOh=1=c(#C2z@+1J(k$wPa<?_Ik
zPPrQgb3dx%#aE~=f4jNhDq;HhWDOW$P)$cnpelXqi6Nfb<(F`VmB5W-+mmqL^I>#D
z$$Dh=esyjstejOah0UG7+AkWZEppbTZFeas&wGW&FB*z-L+tF&(L4qNu!vWBES1L`
z+H9v++zj!fIb<2-ZyS09lRgsY<ED5?w@V<35{gG2D17P=f8CLoaL`&no^5ddwG+H~
zUD7bKLJfb`1hjJe;^GL7maNH|Q>USsEGn$l>zHXll^(o5&9cLXT(})k;3mK19XGj$
z@213kMwH^}$G>`e<o1BbR(eS@H49`*;GRYcph$jU@1ny0dJW5OU-lG#I5fYbiRp)?
z;&O|`wpZJBObAYueu2oIqNRzKqn1Ae6y^8HyAt-5yct;a%89z{Cg=c{tApzsQj-Go
zava??T5jP8LFA6Pm(%(X<Ez%d8wPB0{a-+q>Z9<7m+q&Rc=Mbe#|!x_-^n_dAgGW#
zI=Nmbj?*?B#I2v;2Vyae#X#Cn5Z}mpS!*obTz1`sn(8xjUVpaUB~zPYTxO-g&D=Z+
zZfoJfH0_pQ?wqwve6o!quDe18YO8F#K;%k({dc?Y16RLHORdQo6+(k!VX0*q@{5k2
zdX4l}AGwcH|LS`d5?!7|Q^lG?0fVTtsz$2HP-ClH1R=;53Um)5pFqEOjrx|K>Z<^7
zm3I~bkz)`^7JS%!4qxP*UN2vg-<Sf`biB&0t@+j-^_H-A2X5VGdN)9woL9nbI57U+
zx8*lW(|xygA_p2_a2EvBmXP+_wURDrC~~O`MUBD_w+~&uV$@`~Z#xfXESK5ohxG}u
z=&rdv!o=bMA`vS_VKj>r*lW=``m=@ue(&WsL%JeWQ;t2E_IB3JH%xP<5J1ACHpqx{
z!}dzn`aT!8KotyV?aZZHNDzAuUR`*_^1`^k&1gY958V7(+P?qtq|zq?VsRieO}$||
z6Zsu%3^ULlfXfy>0@qLYzw&@gJoCJs5Nn7@OKAkA)tt_)NoRz9An2SFS7H`bIyHsa
zfmB+j$`BK*)ml*`OQOxjB<6KQg^Q-D5&e-@6>INuVZ>l#e58`w)2PzHrSwH=$4_9F
zehDWdo+blYz@>O)?X?-%R5-s^z_?~ueOiV<7flKcA1<r}EO*sLhFoF+Qpelu_lhVa
zo9P~Mnv3^8p5P$8Ft+5Ui(P(b7JTE}RAyNu$JCE9niW=TGD6HT+P-&8*PyzlUp|4&
zx7{drO(<-;=RCl%K_X<Yi;64U9*OYAFIup9LzNXZ3c`2w4#DdfE}0UK#y|c3bGPSR
z3>XWpyI@%Aso_{bYMBR;2X2J5ZaCQ8O%K-V?z)Be)?FGl{2uZ=85zmE2DB3u*%WO!
z&(-b(3DwL*HMpg0PW+)q<dggCe(_op00Tu9>ouN$w7zzrOpL~kv7R>){?mM{e><i^
z<l9}x%yUT+^9BkQ(S?_irJ4Zmxgf)AU$x{@&f}A=+~Ht?hT73TjGT(fXZdHJd=%_M
zEaa2*&O!$pCB!uzKXnJ+(3N-O-y7JbmuVP-bi7sPzjCdKlzs2%A8K2a2Us6Y929Vi
z5FP&W=m=TjO-pq|@!qho9o#&40hNbP$Qqd7Up=4UKwOdPoYCO{T##-7MH^&9J!Y-8
z@1u?XU2)T+yh_O;PqUaF>J@cncjiPg(@_)URNLigvaUsuC5f#nqut!R?bCUO4R#n>
zcqX9Wc=N1}fT83U)gXK)ogMCq3YLc9FlQqMU29pWYK(lq<SlXsWL9*I<Ft(N?Mq2k
znQ9+skZZ?%f-I^g!r5Qkkiy13#aCZ0FilTvps6=w0~FyGLq^^-bZ80Fs826W5$ngY
zunh>H`taIgH6nNKN%b60FRz<2;rwO^tOpo+;U9D+o}7||eFGQWdP7}*E$kJDwW2t7
zteH$YZgUt-4UFPBlfOq?)OlBrD7x~#`1Xw1bnB|@r`oCRfs@yon1%%;dG&r6r{H}-
zi~bD*eE0uoPZpGrcR29Rv?tw?h{E#Sf^PA0x=J5E#;I@y?%9C%{MEw8jcLWb$K#R0
zTsHsC=)o%ZDC8l9RM}>hlB3(cE|n@EudHo66Fc&Yo@r*|P~w0Xs({fnE5w&icYYS>
zZA#G(Pg?mwJy@X*uFy!JilYNd5qjK>2!FjVgNv^q2hUuYS!1TMHZoU&-D1*xw-<Q<
zy@Wt@jSJ8ehDf*&ftL&{zJy00cF6bUl{9{>71J?~aXfa?D2%IU2s?SxyS_y80Qtc<
z$-tM08Su_dTCl54`FCJOjuC<=4YPWHqpQ1W>YSRg>^dryifGi$d!}hSH(r-2+)($W
zVWd*K_4l&widf(S11Oyy=;jHF!b`3C*MZD}y?D-Vd0vCiM@T!wP8qy8vVVH2_g4PQ
zAUPl)*(;pTeNocq%}fY>n7eEGDkyxIK!~r(ibCwgQT^_7<mJzG?aS*(A`9icd;LfN
zfhk^$RVZ_ll8sh3$eYW3P@`UtwN~Dz-T49sB9cQv;y!xZYHc6P3Ec)3Ao-iNZLGgo
z`j*p%X_M~Rk}7S*v8Unk$9=V4MAwZH1`$f6vTo0@LtVBVD$6?z?aS^+5^Gd5C#gC$
z=c4)Zwh4Kr^>j_QuF=il1t)bO92!A2m1a>vG%&7S|86H45X+>b@~m#`?No|(bKq)x
zXcUIr;N&|QkCe@V!>y==K~5v579&L5gXFyM&{X2{I$#=Rv53mq&Z$qE5gQXtqmDQ$
zHVAWG0Q8Y%!=!~WZDD=$tHe!IUaYf-QR$S6p0Na-{EM;Ais!mTi`ZM^*M7|T?}x8U
zHb0nK&iwF0Q%ss_Z_4-`m@D_fmwMd*%N{RYB3~=8_X+~H&&1hmNwcxob#AhKOudLm
zg$C&5h;jA(z<%H2lBu92Ori(@HqWhxe}J~sYTMi8sf(x?3K9P3nj4M|alw6+-)S=U
zY4_fD)J^B%BUg<F_QFf@8fU$!)7k{%Xag-`K4TV?rEUO1Rw8aB@k1MmPZGppP)Q4-
z+6_mUS9%}*;Y!h<?q;rfPDmQNsb{ExDfr0_KI>q&Al*4+6`G#-OHtMGX*7r&)l3^`
zxA{;$LDL57Yfie-3(9nLfkG)gK*(dE9nT{HasXA@g5w8hzoNFSzKr&YQPmuRlD;P=
zDd6T%antTo)`G4=zRb}tU4Q@NG<E)T`iWU6H)`g|QD=;|Ai4C(Se5qyDfnjCU5UnZ
zQBH`ZtCqbU444?&xSMCbiZDaF4hQSAjGHIL4K|Q6dxGkzzie&j7R1Nq=k?OmvW<(%
zUl3vqNWgpgjT-<{X?&10dn|YEV(g+>Qf#e8F^t6K^6d*%jnz*<BEy{5fC#R%8hgZ`
z+wya@UN@awlDj_cwS)=yc#0PF6{eB`z3Bgl!yvq`;%B!d;rhKi=zfH>WZ_ZE0oVt>
z&EbN7er3;XhzHivujc*!XsjAd%R&XTt@UeemRv1lPC_kos(vU;dscFIig>|Fds9T;
zaR*wY{qe)dF1hzf=PT^QInro$Vz$eS^cj}0FGZR7SBH(-F6mH<FOFwhPvrO6+Fp^&
zdmm|ua=7iCv#Gprs}nc|)>iogAWfNSSw<#{mO&_TMmtsZs<vWreV1x2hA8sP>ArbM
z@+4T_*T~#&W|xZeM09H6<rxK~@F7j1>`@;S*vB~=o2a+Zzdx}TzihG(qW!YVfkuIC
zJx+lDT@~Q-T~8Pycm)EI2r2}qnHsObik^*@vdO_z^UIZaxaDY^CxWGp<4Z4oWCV&6
zH}5s-`RT`j6nB?4yo9v$_hw!iZF4R+yGB;-ps$!^-Ai58#%G*YL_ElM#p~(FfeAMF
zIq8S|cXXe;uGx-x6D0Ea+92tsQ(Hibw@sZB@+WAOcN}mK=0-3f5M<9QH^iTjRgd6{
z5#o*Co>1jC(TuE;)WXfqAfcEVLQv`*O7+0!N;cDXB__n43_D}s?*b-rB@@K9VL-_(
zu`)MvyH4%}Et!!*z3wn;rVn{`qeFXrvT|^SJ*R$2-Ye6&-;JG&sy2R%@kWhpi$!s*
zE5Vve&g+`V;*q!ZZnGlcxJ_AzT%LTsBE@~H-Z7JzlssA#|6=ZZJ<4dNkR+e6tkOB{
zt+#3U%MOIbIrDcN8^M+rM;YJw`F^XGCQeaZm(`dMbaaP(btoCzB?M0E2dz;xXG%CP
z_{XY9@^jl3%PxFhi6*`56_}o4`yBs)P}#bg9$MyH{>!<v!A^g91rYiAPjagfyVKhU
zZimkw62oMx%O4NzWlGAh8TNS|kjD=dzKn9^l!$Udobf}>VT$79i<cZ9VpE8Lz}$)8
z@+~=TAPPiS3j%A&(L$R(VlOCoPL*f)sNgVhn!yC^!odXbUtj6Vd_?3<yGHmps9zK(
z4Y_+1Q8^8H2)n&TslhZ1nZ*cvQ=L<egoMv=+c8Tp6Esx*6VCeubmF+4mC-9^pQiQT
z_x+rRL42ML$1Uw1!F)|97J_LrLPHZ-H!E)z^hj&_#af$oZNd%W>^_J*)OC5z?lKV)
z(^<sC?0#xvUC`B5wN@^0&`ndO=HszKNFw%hmzrWi+0tzxuzs%xYp~3O-Y3@#Gy55G
zuIO+PQRY-W``hN=)Wu)_EQt-KTcRkwr<U0FWIumlJK!i72po#Pnn%GHQ7N-AAR%gt
z6h>dZtHyO@(DM3*jbEbO?C%YK{3V$8wR>kX<bI-<`|N}a=Br&R@zWUgg5!?-G09tx
zEFOj`ucb^L`9_?oeAP#`$-BlbH&^dtk#0s``@qG$ILHEYvtq;An>hMTs-q@c0%KBr
zy0b}ly#f4A4d;yJdq6w<4ED8uu6;!&g1DC}xDcAOdH<8zoQMS(L!2#dVTVNQMHhDq
z=#{)4Yr!ltkg11`<jk?n!@Khx;RaeU9o=UmYWn>e;QC<*{^g|Nf|Bqegk`*!`&U4@
zeaxv()`4oA6Ksdo*tu|pvq-3ZOz*N4ul-kj49b^wGx6{pArtL)H=i)6#1|Ar<_K_n
zo7!OK{B6~Z<u>!LU|;w*JGj4xEbDEsi8sbt=BeP3-0=HFdccz(gm!#9hLRcD{se^{
z@U)lB_|5l5randB)2Aj)_=NS~J4OC^mGJ#8t6yjSzP*`B%u!~7r|8wJLvkO}LCeQR
zgqU(UlLezAKMd2OYqc~N?!{GkFWI_Yd9dD_cdTt-$1>9H6dEkVy`^#5<~WK>48OV}
zhUR`00g#ZnDqCaJ+XMU^{sf_w@wi;FZ<_l!bM6Sz;Uu^w7CG`yAIqWj(+>;IwB)oO
z6#J*foGfgBBG&=ivJaj&b=*KzF?XW-kqWnm(JhpPfSnIU<ta~x5u4cH%PoE7pDcP!
zF(BWZzkrvZf|zy1ZVcpc^_u#yWz>ZhOY(wPTyYTco<XyS38fWdmpH5wzSm$v@^y)J
zZrVKpO`Je<*ITY0jH0w)Qdr5?jw~jWpJ5_e$x6)h4SG4UGXkM;5Ooeht0y|Ui0M;r
zc3RI1(j6k9Ov|L6_-?q&ftP)ynPK}RK2y-`kBcdnN8b?vGMechu)~M9l{+~YKq*|$
zAAM=zg-I#h_Vp0_jy1B{9pzRn-xfw!b})2wRn;X9A_HO12^<LHI!N@6@k29kUfmqp
zOK`}~4rnhN3Onj2AgB;RJe|&{X<a<HV_NX(>7p^8bBK6%j(37Y#V@&hKX}qG++mQT
zTvULqWuAMbZ7$EO5a2`cRoF#|F1eLGP}fre7b#YhnwSo6A@?%Z*QvvLSASYR?XKa{
z@M*U&u9>vVsjt`}&LQ6`u%NFsE%+^gu}-P)!lm+DN$8r$w?o6>I+d1rgwJEES2~wf
z*j05>8u!H;8D8&<u=?!f^MnD3^==fips1kCV0U=?z|(8#M?yc`RM~>qa?)~3WDB>_
zCy|-TNz#c%#lr$ik`vRTZH`;%>-Z*xgsUvg#cHAYM?sLF-wXd~Xli~qs?Q-0f+NCB
z>~JL_Nx@G^=?OGOKW+Hpg62c@sao^WNv=6ab_)nlC6gFQ#4#2n#%?=9Tb*GEs4qA)
zCDlut-&Og7-4MBOUo1(QeY<AxJu*yBkgpyDI4b-G-_{r?hZ_`qEkYghvkVQ5z9?rz
zmy`f>IJ}!Hd!HcaxO^|{$H8<$6N~bLf>9~WX;N8Zl)YYr3eT%$mq}i|Z=W&@sBRxx
zBN{p%bspARnR}Y$I0zi;Rug|{Tez%ls)`;VLEF6{^^4{lOQ`J?6j-{eI$zTDR@PXA
zNsM)*w4iEjD@dt;c+6CaZW2e_WsH{uGu^#hkm5ZLv!zDd#IEaZfmyGUKhba9H@6yU
zL4>7SFty{)GfXXKL?lZ4I2>bnOtOH4o#A6gI7q$^$wl2aDwmakfkanhQ)w3mPZQ4}
zDsS6y+v{F_bLj{VKv@pCD{Ol#rMjY$t?6eMBqjc<vH&0iW#-s-U2<%->t@R(ubX;8
z-tO4aVQe)f`Uo1b7V_X%g6M2vuxp%9?zTl-k{<h_A)T89h2^!y5eyI*<?x{kCj(*G
zMiOg{tyt?shfjUD`l!wk?n$`h5*S8c|Cv83s#aE6Eb{np+VVT`@JKbYg&NCl6Aabi
zz>-bz{O@|*ELEdDoQ94K(E!@VVYJcsY>G%fJMZe?cqq*Mo)_s=HgfoLy!Kx$0BZk_
zWJ5$43-3TOO{1NaFsC1hMQ!EIVHOI2HZ%ghFWHIx^XT9O1|x5LL4453=hbEv^c2W9
zh%mn|P&68Tp|2sr1Xy~(3^qTSo|QtQX<w~`*n3}^Tmv29^>zk{-Z~8>nqcD5OeS1N
z8R38GI`7f7Y_f>1L`M8-?LWk*F~Erso|NApGNVm2q&*F@obG@n8+~)SdMw~Gi7!P&
z_0pl`$-V09v>>3p6vK5Fjv<XwdREV6(ML~4Ykp|lXo+M=&xG*B^I6~BsnbP#n{S#;
z>9JA^RWSEYNE8Gi7oT6?7)>TKNz%2(wx9>%D@nSZwxHH?(kOox;B;7aLDqEoGIcY{
z^8hore+CY<pW*jI59-qiwi5$t{i`R&WRs%uKLbQ|G+DOUDpt3wn1GxMmJO{}LPpYl
zyqPH2*Vq)(9Rc?pRW_n1RVbOZc|5AeqB~X^1?)h{-CF#2i-*}g!Qv1anwrt?G<&ZU
z;Qpu`r&&MceV#<tG9NPt=;R=}qZap+82nfZUvt{Q1uvX1p@I5i1=>XNY%YFs^~9|=
z^~e=LipfS5HJ#3jEG@JIUR3%W0OW??Lb0|^+~_eeExv<=+4k;BmyjXr+GRpytj^f%
zkLTYjbkG$X!m|1HLtmU^7V{-@h|90)J0(C$QY83VOgo{Y3vV{XsdNDu3C)b!#7Cyi
z|Ap?aZ#h#i^9#3TC>n7ZZz^^1H<qsA+befu>a4@we`k7}D~IjrcqpefOZYMoj<cIT
zO;tn3HdOE1u28L%*@Vz5l3DYb!KQ!ZM24o-&Rc^hSl0$ouTrW#d(mH$)mYzdGct1*
zVHF<3u&g-KDu41L9^(50zMsj7bLp>MNY5dK+{M6-ESku%XCA6IO)46LoG3>owc;t@
zDgIo1;s21$Vt>eH50#}mM}6ha*6vC~pcUz&$QY3^_q?}mt=^&}j#^WC_Z_<RyG!?*
z^880rb&umQR-=cd>dn@$wg#O2$@1-vyDt#DgC^qrWB-LKl<F(qP>5@iBLa4&?2*#q
z-inT}CiZrupQv6JFY;pU_{i*4A5crA1Mc^FJ^Py(SYMWBL_29}->01$zSwp2@I%N|
zVQTbNH@tn3H~Ngzo$T<hu$kJ)<n`~-@96xZ#r33>C><ATBHB0P5nnYYWyBsdV}s)$
z&dNB}0}7~7agua}6Bg>wu@bcmPNR&_SC=2m<SrTs-g!rPFigqUIB&P3vxq0>Y}$49
z?0(DR>E}`1PMwqo#EU{Idsu*4*&9T&T|OXik3>l$HhCeYosOYoSG_t}s)YgS75EDU
zvb4u%`3a=J50CShjaSPyKHBxgU$Bgv@s%vY-0+zGwdFD5nZMqnBb$DqZJrF)V|p2d
zV`a@4siuoBd#lR2lm@Mq_ITfPttg~(`(#S%Cu)&gMVtpSqbu6t*b(nfaWTI6eed9{
zxCkV+g~=HJY0iHPi!4r4M4Fsje!pzwRtV@#%LygZ`xzF>zV?Y_J0pT+eArm|*+-ms
z=NB!CJ)H^PZjJNe_jdiq8bfUZY4pb6rZ1jiL_CdIXUvqBVI!-Pq`Ezr7iTw=BkF2J
z@k35iw8w?vRefV+UTf!Nv0l@c$9A{))x{0ibMwXr$!_`UO$od9c0`AkszknOVRs2W
z2PiBxh7xl-rJoBXIjWo*`Dq^nD~M~n62h{>GcLGB@Tv|dW-u0OfkoE6srAqH0o3IU
z{!77z60nkq|9hPD2R;scXHhiyQ@m<1ZBC9FsMsA^fh0QI82c5${32`qz&~Y_3i{iC
zC8Fq=ZGrM)BunSYzLi?fH3E()aA<A4ovcRkFsWd&8!O%pv$mP6vCN0HEJ|U3FD_Mz
z@I_0yaHoSO16&<`G5^;ZS^zK^AYbH2KQRxdb)YfMU3~CVe%|sscKlrZe`XQDoA?3R
z+8w$s;ZT}6-@{fMspxC#kVc81#i*%8NlbJegJ=?yq?NvVlfGAM$K;+f3caWqR~Cky
z-dT2hY5e=MpTm4jJ{=JfFY#|A!db@OGWv(^q6tgDQ1%JHf9ct0-WJIijC$MeFxc2^
zuw~ZKuTzt?*HS$Az7Qxrtz5GzSrppP@QtVz7eO*Td6Zah!D@tgsN3@>Aip&)a(#Tm
z=ms4#Ox_GP7pzG#;;K}ltVEqEUj%zC|2k<_uC#3KE45xlMHQMqDRzJ$z5xkqx#%l=
zYC6e#ncD*g6NSShtGUs2bnyM|^bl>$gyqH3^)CW?{3FkL?`jOpi#oI5mm-p9|M}ah
z!hHyyJ!e(Cr9v(Lj7q53P<6b5vu?cu{DncMoK{qs#oEjU<8emP6xWSh3g_-WxqkSU
zajZzv89p!qCFhl1wK7Ve+6w8#)fAC@&$nIa94=TYpP~7jLtZV}v`N>FtXa1K5hVpo
zTLq{cArQ5G?f-W&Z2zpoL3HGd0m^?*^^pHEW1D>B3`N|@U|Ld>_N*LrY;bY0HT@v7
zOiQ|Uxy$at+ao{-U;SdIbFZLDwZZ@l;6XDQ6B%!<Upcyi3Eusj+b~AUnNLbwr}0C(
z;zeN0*fM_QcV)QXdN>^cFS-yF{oE(#kIN5eW(NO!g1*6(=w?Ph`}erL`}ucBr@VN_
zOiRvyJV*5#@c2=MmQM0ALVGYeom_JJm6+~VlX6w#s#BOe)vGZo+=yt8;+TUUZohr$
z5XT!|WQWs@>I9t<XaG3-|2$N~aDP?R!+o4mO#f_PQ2fv7lGkQ-y_~297N6Yv#Ki0$
z)+T$;`2VXZwSWk8a--sx<g~-v<Bqvg^Ye<i<6nWcgUa*~#)4>Qp$an4097Ck4M%Ab
zgU*C&*Yz=p)Vjal>Aw+eeYP?7r|Gr!YPymC`z7_y0EsOf&ILD2+t&DQrcbr%pu2@b
z&5twNRAbW|E?6?;xh49|?*Oojvse|-_G>TP2N1(<IsgUrw%+!_&-`ey*G}cDd=Kg`
zMyv_qMqh`$b0apKaQHn*---gbJQ5Y<x&L;S1IBq6<uAiJJgO!o_uqRnJp*x98*o#n
z6Q(VdG8ao(tm6xyS3D&_w6#36u|ox)-*d<xh!re^AIa~<unvg$09D5de$3T08gq5p
zacG-UCZ6#bSM`6A<<1k_(VV>e0RKKL1`x1^v5t?+{rCQ&&rAv!IVFnW-^&Yw1pss;
zyjsCBu$8Vv0yZEVx3Qwg0iC%$JE`&H%-HFE_8YutuPCQ0mZOrYF>37s!<bNJm_(Fl
z@udN+r7@d-*+ip0!}q$<b1rj7Zn!q}>B{y!XQiU~no?D6hMJgx0`0h>gUMohpvKRu
z`VM;}H<_0%Ugi?)mh4k+H!!u$S$5p=rxdCH>X&9eWM1prw9|7_q??Uu;Aq?c2Y!0<
z@2A2;BLjqaHxy<A#lJ`Nq<<JXQVCk}k#K-1q_PGmvt^JXcLDaZ;|^FT!m$36v7Z3K
zWe=D9gMI7Im>zK<eVJG4vGy1Vz$me_Gra&dP<0={1jhe-tyO<m4lVPMvmM04f2%;i
z_Bk|7WF}SQ@)Z(ZHoAH)IIxb>rj}Z~=LTF0m%e)zi&vjKVtij=+_RX61B`#jk?g{-
zh?RQAJcL_;S;0>aB>K8^z|pYVwJV2#NDJv^AqkUwrbvB4sEaOrrbS|uysWmbAs?2!
zW-iU5ql5rx<{(80(zPi#v~b^1C(37d%fk<!WQoHNuL5C6*~$-o<n>)l9K=u1x84_z
za;p34vSQz8!@IRySy|lG%#1Q_66M@{He9S{RMwo$J^J=C??SWrmioEv#~=4Ol%QTO
z+I4!WuEM56;#6<NdIWDYuP-^hz5IVC7!PP^<_@;T{i<+3jWh!JJGh#24r#8J6PsK!
zJZO#{^TT(^7`^V;PxB+?erTGf$+x?`zfuZh!Q#c2T5o#A;aGqq06fTU0f$XqsK#k~
z$!>57*P~H*{!|44`n4(<w;q>rS@NB)5+d71koxtL#*+GzX8C3Lz_5m985CK$%tMx1
zTaMiKOD?C~A+2{ynvlywl?ss#gcK1gQ^6i=;g|ACn8w6EyYVas2Zz5cJ=JA9PyPH6
z(x1s%#aeXcuZV_Mk$-K_N8I|vk%Y{JOP3f#Wp7eWZp)-%jW^;EzwX1Ic09{t^uBBe
zu8B@n(J-;v5aKmaSIdOLd<C^B>eUzoD@sL8J;LUF9ovw3*(<u`U3Dmf{Eq9_hCnRL
z^B<plVA50E+FZvF`o61J*=PM|pbA-&JlFX9>7TC+MZyG5gI1{-RX%+n3;&HQXP98`
z5&pE;WoWW4P{y!?FfMAuj{Xv(_V`Q3=h^0IFdgoV0RwX9Xw1d;+S<D6E%OJeT&8WE
z7HIOo00&^1)<2dzESLYTuhr!x9Vn&vX4yU2|5jGC5X7*%6N9K5K9RQ7zzsAh@#XE>
ziV82<P9r~Zg&H%GW}lHmIxV*yO5lTCrev%R>JuFnu<@P(+GZxGy7%>tG%i_VK&%n%
zDi=Hg2N^cKwD{?B!EO>fDa0(isc$Eih`VbgDs_G&zE`az>d_czW@gP(<0W!9$M@88
zq%c%Lx!UM?@zqkZAc4YeQ|XMGlyWjn;`^bA6lK_ZF-YV$F%q}-1XoBG%Fp;k)_1X~
zT-)6f2Z4sN$Po@ML!7@KE~==UY=3DjRpY<)cORkuKf^r957)|d>HYn53T~vgB$)br
zZ3C6Kk=(h-{~6`iqrUU`PfY{G__Xy@Kd$`W_yBEA=UQ7WRFzdXXXuy}{vW2!I;ibv
z>HA55Vuhl?-66QU6nB@P#ogVZxE6P5fnvekDK15d7pJ&86z$8scix%j|IFmf@61{G
z?00vg<V4ui=^ctT$kGd)Ov0g57lxkne;>m_Yo6GpX$Mly&H!_;fhXb2@ocTU(FLuv
zvQj;12foc<$KWK|Dn)cI*kF{OA(DhNIx*V{RE49`il<|I!tDPX;qAn1h|M`N6QtLT
zhlbp`jez1dTI)@`b$j3c^DA~@dMhOi4NWN)-g@$hmQaQ?0wt4T&4hW80AXG1x!3Z~
zgbyrsnEqY5Xip`7J~ph+@O|OR7yY-Y+^}1#RD!#txaLf3?QH^|prWFbwFUQttI5Fj
z*`~FnHViF4GikI|(UC++SvhlIf#)TsXYY(V@pdpIsV?sDSSjt(iGJI@?$~Q`3fbx;
zDA1YdzIY<J;zvlCmux|{rbYRep`o~Hi<?Qh(s&y?Cr-{;PD>~Y3e*F=iE(6cr}5+m
zx92CY2xk$oJm*t?qbIzNy(tQIlSgCrD5MEtZU!9Zx6{Ou<Gzy{M3Q%EDW4s}I8B?M
zpUK5sk<(7TJj<bfKR1xtweV}+k(LAdE;p>|w93;y?<a$kU2Im}Zu!9PW1bBz4|uMk
zwkoApXCGy52@k~_o2<i+akIG^!im^iMN$?`l?In34YNN-cFkjt9Fm^O)0@F*MfF<B
zL{r~*=m{C5eseVz_??9aEL-lx@LHU+&KGE^)NC1dgwmx;OmK@ctTM8mrau0Odo<Q)
zs1SNh%eI3)PwvnvCJud0lT>q)ED{cq-eymYae-@H+GmKA*>TxEa@9NkFh2OJ7ZzJ&
z>P<M<^1CEtlq1si6}Csn+4+z0`vjA)Qkz|<?Ip9^7wsy6ScaBm@rImUS$3zd95&w!
zI<&s;6s9t?C8W%iCQGsS=!vk+eAlJ=owU`4PgRZrjf32YSH3vnt#qfQ_ZSUdr`?gJ
zU-@Jpu&K+v)AnOm8*SDug6$QmqE){seKP8MsF{CeunuRujy9jK+M%yCR#T!y`$oD1
z$kzrFC?sax6vICwSy+jOvxI!kN^+T<Wp{9OFtk(`lhg?s+nM2?WdWE_=5ae1l2|e_
zO}8}k1FQkxm$pxwP>~Lz0n!j_a`JD6_q0O|;FH4Ygfrz3Dj-sdNG5oChQ*|00$|yQ
z3ILvv=k<f_pDe|(zr-dO0TOIbWtZ;EUBS9v-(LX~0scSB0;YfB?3P!NgpfBj^bDJS
zNRS1Ro1M2bA(;H&I=mqhlZ$;OoEllA^MS(0A&vfWLiVOsJUu$(%Gz>4O$OZPnn-(0
zAVnSFA2jrAY02N3OV51oP!efagf)3^|KsAv*aHCe=EAzUA6y*OZ}|Z|!#HB@wO)mf
zY4d7mZ`Vk|UBi;Cj-EI&%gsa%E#I`n+PE7-P`Ec=xVfugd?3NAj5e+gdQAI{Yy=Ti
z2voxz!Dun668ja)Xv+R$F}9rVaHU))|H5Dp-k5doA;;z@2ku53&&*uf@9Ui+hJV+L
z-Gb}z#pB-km_weZXH}z*=Wh4yGo6)V<r6+tlbB{@SJa4=-^=S2Vz_)u{p$Tvt~cWg
zZRpjKy;ynr-KGRVP~~r96z<3$l|?lh?#+^jW-O^aK0U6)CeOs!Y#8VwdSwx16IuoZ
zrq8m83EAnI^1V?oslUA3W<X>Jqpu)BuDJy~1sl987~>QqMT-#QjtUtqri1rx)SH$>
z&@}-8iYi%YdOxXeMt4W+VPl}L!a=^mCc9ZsL*0?sM&tK2c^B{+7D(#bk7a?yNGF+1
zeE`|f(TYmL!Z6M;xW>bi!{ob}*2C!}4uHL{&VA;M^U%U#)uzQZ;;>@n@+m|Jz*lIa
zCk4tk0nRY?H0y>bsvLY3w#76F-Ic%32Br`Ya-Ob%v~M8b-YkTe+6v%LU~Lx8-;ag>
zE&vsVgWVLvm*pdFqtRcb;|v!@H`qKikHnh+*QXwpD)9=(xxMMJCGt*UjD|*SQ#zxb
zTp>xW8+}((c-)C#U)w8pl(|&lt(c}q`9DM<e8_#jqwIdF59bX6KB(_zqMjL)O>r;3
znC^gJIt}X0PHEw88?btLVIixgG@+F4yjE_1;frw3d!!S00fpP?@4KrJCH2x@KVX@e
zLxgR=M>ixZ=d7=b9P540he)SicQ_4LRqoeSVm`E`I^0<dJ1XJe-sIV>sWZ|DyYab6
zuE9|?nVh&9t)}($yuDJCmTa5<J<=Sf@9caSx9HoCI4R(en=oPB3{9k4%)ORzw0TL#
zF?U7FpOSuK?bY>#5n(N-ASM7js6|WHe8d-|@If!mWDCg0Mlp_Mpr=N<fLn;PU@I9c
zwF#}1GLbILns!3WZbW%@W@)AG(Aq@4-DG3}Xn<-b+gD}+g5OUv;9(4h61n<t3zO;w
zHI8VdI#Asg6DSMMw*E-QMcSiYkjj`*+@I415H-zzN2FFGgeVj;Tq5;l7fa;;gRx2D
zxKhO}5n+l<)+Vt$*K>#?ra`Db>K5q6R%W@)w9&xtr0v5eM41xcuv0*Y^QBC0qnYCX
z6*%v7%!GB~ITU<ogOj%E8ud8&bmg1B&1h%;lhT5ke=FnvHUeX!)#DeURGBs1F@Z^q
zh)hf{R56Uz;p;h)s5FO2S70Y4YB>P?%}a2*TQjYkkLP)+_5lRP;kv_G4Z?EbKS@-M
zeqk|}uiajqrE3Vr81mxgRO_x!BO1mf#dS~cgscK6)0*)anuW)O4NKlyPcJL;hpS8e
zjLMfUPGrI2k4f<_)e3C8$cMk!ruE?Lf*XE&*yJuaxvA>L@M=8?9G<IsZ=G~E#bL~)
zV$4aTl9F-fxvn(epl-T`0oA6khhxhcC+yO8#pVR#h(6;N(IUhZb1q*@gu><3j{!CT
z_OeY%S;%cOeM$>+kiTqHi&{L*Q4mgu$yYzlb{YFGz279kWV(TyPb7^%^{rrZVxSSN
z#qF=<d7BzF9x=st<0wGZ$#2C9j`bWLZS>K>jGXn<bql1730#~rA`V(Y#7E7=^+0Rd
zlEFDfj8Hcu)MNsiEkp<FCNL$l!?N?Ze0U~Zitf*Ydt$iM;ZT-fVgNNLkeF{K|ETE}
zstRxN4nRH6%qH!@=s6O%N}hzrhSZ|}Jd`{DqG{h?4&*Vk`Mp@b6O_9-M3T`i?jLLT
zXhb_#<JbNa$I9pCCaXy_%w40%WXuZXXZ&oI`KLN^mgdFtIJ=l@RY8TxQs%giM{?JV
zP)6<E)&<SP+}iyo%uuV{u0Z0na`(HZ<aw9qpRVC1cH4v~g^tW)U8Fa$bNhg%e1W?X
zbEHgvKcb?FN&&Me!-7-wse(@!A{<;LRZg#8CIOLtbMFYSiqp5GOl{<k0MqzYwUlm}
z42Ktiq5JWCTzZ`L3YVo(<t^2dn`@oN|6+P3bZ-?9KC#M?i-DZ2SDnA@ccy$e1seUL
z50AZ3Nkn`3>j3;8$6`8ge%biR8NomF@Wux_qvc`csPheT$J-L8On0x=z5~Sq8g3j_
zWx_ghO{U|lG$dx;${ZoTe!=vFr-gZS6ar?<i{JmQe(|k)SsJ}TxeagaN*Onnf0858
zGG^B{S9>s6le;&_R!=-p`fx(PQ@Qvd^RVvMBcil?K(FQF4F4w2P`??~3{E~7jyx-a
zhzv9BNNb`%x7K`D13vbxZF8L0<#U7J&rgwL**fvZh_8cuc-E;&DTVDy*(VBh0BQW`
zHBsGkf}VehLf398qE-n@zODb3Ca#gt2X6QEWGU2*S<Z6IUCrtU7e?Y%G|?2C=V_U5
zF@aPV+Qhmxib$=&k2@<?_VS^D5J}6=WoGOW7niyI_AVmPZP8IrtqY+;GWB59O`-V>
zW_z*VsMq-Tu{3GCvyz=|`Ph{8`K~+Pj<NBh`=u{`vfmfu^mGo!O$~95qL(Ultg2#`
z33C#*4jrjy8Jru<+cE2_ubtB7Apgg6I2(r#%0~ilK~4G@2JrTa(m92#@N$TJKW6&%
z65ZY5x(}XaLy=o_f<HPi!eK-nrwW#M!w--Ei*0glX%hWv<St9s+9uT7(Z^+g{3s{(
zpX;W&e~ReE4IsyFYNcqiAYu@DBH_FD%2Fr~!~<E#V$~J1n55LMbxX!rG?iEJ_2`Xf
z0(h`=gW}~f`SD#1wc{U?m(|DpI)^Wlz9P6umaM48$2+zKWN+zw43g7CXA2FUZo`R&
z_M}~-)m59mGEm#I-Z<~x9ZEbrdyqLq{UJJrM%}mMhG0H@nGlU8rk^Vq`RVHsbNfBj
z67YNc!4Zo{ZEwQ%yrrm{WO1&QyR)MwRc1)|j<_VDPT}`n39KX2$%J7APp1V=dT4l9
zQ91q6vxfgz-cCNoL`|OG#e%yAZz{wzY<(~toL^(uEuUBy=yQ{S&M+cb6>ab`Nc_SP
zE7y}eBP%@FsNkyOwDU;hroW~i2TAuC2aYxdb~SC`Th}B=uSCHoyfAiTA}HaAir3C!
zu{kCiKc&e}iMGtz^V`qH_&%FHDm3BuZE~;8>L=gXG%e3ST7=Lg9jgC4;t+LVQaohY
z^oI75;5LC6-M^cABoT0KYn*~%nE*C|j6!30e72;?pL#M&vCCi}58xd}3j8MArg`5x
zhF|G&62_j&3-yxsZ(?G_LnPH->nyxbc(p}QF|=T0h-RV9{4tCl!dJxJVdwzYYjgZ%
zoE1o~y7^AQg|HaGbUS*465d78WC9C<Zwc=yrdzorcG<{Fw9~>1Z>y_LvLKwkEd3rU
zm>l)n0qp`+6+B-Sn(Iu=E%E|8rGbYB`d!xQbeT6MfZ5>uV7?)SRW}X>)e^04{I*B3
zOhg-Y;pa)h)@AoJ!PthRKdOl~K{z}m-#+7Np?#!lZ(Ob!&_s_Io8->ZK>i)ou0_P*
zxqqRc$y$T?C@P!ER_$DU{)a%lSATZ$ACI(tDi}?BQ{cA^h&EC!z7ccDCUdmY_?oFX
z`#CJjD<96~c>Vz=RunaeYVfDiJizaTPVUju_6K$kb4f05NRoQ%e5=jj=K=g5FRt2t
zr95`xGy0eKRDRO4mI1~p+Z+{r%&eC770zk6%+>kBEav1Z1uOGkT%{2O!i*Ot13p^L
zF8_hy^!wml6B$XUJ<;wpTYQLvMEEe#)EwtF;p0p{pJkA<P_?k8&0#U`8S-a*8Z0)9
zwvp%Vsg??g_WSVOC=Db{7{HOZt=46nxNI<Z%RpQIH3^fd1CuigNIsn!@jd?n2$S^1
z@fQ}BgP;Y6n9RW8WT_LwuR<|Z2HGpKJwn3uA%Rcgv*GOYe1e<e0D&cw9S~vE6y^Zg
zuK*yf_)7(0hC#j`L@e8fDUrN7vP$E&7-v(ssPGuxcWNvf{urmxj=yw5*U<zhLBNT^
z$g0`DAoq+p0^+4Q1Z@8BH3#DZkwUD;-D!)glEb6L&TNnswL;iP_2Le^VtX;GT@pK6
z7wR9(wPxfyvE#=vs=ID5Xy-*77GQdIHQdVuo2SR@L{(d8bL(=qG~HG51Nb(D!$$Sr
zhuUU*mD3CK(Km3~dOz^;Kul0KYKAw|+`={{cX_<%9G0H<@1{^MM^ug{M`z>MvB=Dd
zk>gtPXM|{7f^Uy%aXi|K$vJzMZNgP&GO=4CY7iIR!d72z)Ojbkfa$iGKL>jqa;I*&
zCFZyKbyv3Zi|Wlrd|FdxI{bG1&`a+314XG`Ci@}{me<SQ)^Haqs_3s0PnNCT@m6sj
z-h}<;o6gbUHk0ayqe`2Lfu%;P6d`Zsk^S~1#kulz%o0ssp>0>>(S)q1-{HNhp$IoS
z(4RMJ+EKXDC*4BKl(Smw{=`{$Wou)pKYCed%F6S-(6zgOPYGBNwQTUKw7yU9UHes&
zGlpm4$b5ksETItq599Pb?clY%5nH}-SIlu|0Vn{ZLru1Sd<F!XG==Ch_F@5$C!-NL
z*p(F%wNaJ7Bm#~cX4ta7tT3bI%l<`80*gCun6wb^v(W~BWT}U@6=To~G~u3Mai*lI
z#D=Is%Jac#7Vzfow%fTW;m82wAAjhyoN~r;-<7EunSR9>=rTN{)B12athD4WghO|R
zhYD%saGW(E0o4LqIrw=TE1R?_#zbk1)F)HJmaL^4r`qST)yey`r(~)WR4j@7BzGC?
zV+e@j?s^=V7%mV2otom}CEek!M=Te76Dt(xJG2e7IVs5>Re}zb4@JibbPSk14OIW#
zRoUQa<TJIbCjh7rBj3t(O}A!)TBxK_6PQh9ElEf8=`!5~2bgy3^_^yMIN6BFe<OCc
z1`cHB*IK1+8>9()0AFI_XK`*}@qrEu_hKbi)yciyGc6iZLO}e(6s9WLw5gEB)%SNe
zhmUXI{ZU<Cf$YsuaWhLBbUe4s8|p_74-VK?Hd^Jn?)~y5?a($2g2q@`I@)=gm^0wF
ztRA_GO}didj1kb=JyhQ-Z2QH7KF`PWq2}UA4=Kt^?3f~CEBrRb@BChj{0DVlY?la%
z&R{c{DEn74`fAK!SbiL@sV(MM14#@yj)&@Awg^Ovv3F0eEXzU!z#~Z^K#kKSMR^a{
z?1k1>X}Cx@SB9sBSDRUc_5>2Od;~<G<%$S78kz!=0n~sv-Aar+GmnsQCIg#KJ|I32
zen>1#ZV>bXS0@LMG<p(7f$xuQ2cEb4K`=kqg}zKq9&z)9Dbbr&Z`CV$c}F?69_WH}
zkuOqaa%O<%4f=GMOo9wk0L9{mu>1jnso<j#PqJ_>LM_JMYg3^#zWLxM7=w*0;JL|b
zOkOJ8ci}eu3Uegyrgf$-bl{yW9zBS=elRK^c(0(s6<`9dP$jP$2a%?*7$!v)(B7Yd
zyA_$rxHf&mczgRzl{)FFbhZifkwNjvrNrWH|H(*77gWG(!-!8Dw4nH!bP1`<<iA7t
zWi*uUpsjIfuPdh{I*eqizT8!}I(1LqcY7;~orW*X84Pa*FI$o023#zC3`;)<yv^AL
zQ=LR+J~2x9&2C#7s(6*eFqP-}LeI5PkSaC#Y1q}Hk9|%UZ}lqj0yn&eV5iFII_VAh
zC2KAL%hluVJ=tU9KONqP;Br?faLg;ux$OhWZeTIm1q#!Q<l+Z|PY}|=R)hidyNu%S
zHoKsv1f~E^U50MyuKrdLDHURc#B2uSOSk$~a#C3tr!i<C25XUy(fgrN7vvJL9T1+g
zO*j)w8;E{KSSP7D#8oHQ7#$w!2jbPFB_Av{^>(jN3~e$?11@0%V}NMkMKDwYR3(-^
zKMjR)0F>dA;p=2W`KToD#H2#65I6u5CMC%l>&j_aIq{qlATf;K6o~jNLZ|L@#CL|C
zChuPio5MGf)6%HE@JcNfP#t5T*kvFF7qfGOn}LNI&*ewwg8B5LI>3mzvNo-)v0aX7
z4AOgdmX_sqfTFoi@(~9o{3kq*aqizo^YMrB<04|$Dq8%(OjV8W32<LcHnpnU5%AF2
zk|!*boKy;TjTy90mi(sejdXfU6y}K4%B?<RV0$8BbQ{<7jBjE3XvZ*B@lhyXZDjY#
zx^#7^^<@3YuXe6zy|vxSjnAWeqdivn)J+}QPQ}UTf=#w=Me`rY+`&|8y|P8ZD(Zgy
zt~text4#A2zF)7>XJ6jl<x%|9r@<p%JL9&~PYKo_s$Ewnf?c=Jbz5|lSoXR~oJube
zV(v&-1<WL*dC;MgasJGgxitHwWSTRhy;qq2>x}$|A%m#Wrf#xZ^g|775W^!i>^@q~
zTRr3kYN5u>!6iae*6Fa&PQ|1H<)mGFr^-P=@eguAF`hW7*)ck5YVOMeMgo%pkQQF%
zbH|zVERl3aMx^i-@K{)oHp60>K~W%Jy%<kmG(;9@QyFl_Z-<?hn8R}AiCi^0G6r|@
zt+Hm^+#>5wZ&GTYMgfBMobnMJ8317u<uurNKm0bPYZ4(C4_)1}&o-Q@5S4BLEj>7u
z0l~)c{G68)<ZB%i3m`O&!U!^dH!WphViP(j3phFWU52M>aksQKKU2#t4f&iMR51pG
zzuj*8ncOiG<15>eAKkuKwzzpOE8w90VNOJKOZ^IBo1i;(r`d6)9!}g?N=%w$RiYa=
z(A7ZQy+vqYYtJ>4mhL7@tmAn7IJVWReio9RJTAAPEPP#T!(N}-_1)8B7BK}PZM&Fk
zsT4mD@^&wT(TVY^+I!7!wVVIo(RXj_7es7&ctCA{{oB=rx7(Q#?HMO-x7g0T73a^E
z-nAzr>8T6!e*K+idA$`$@9#!#;C`Pye8s0a@v2@XC5L(~;gfSezAaYrW1`Fpf3fs$
z;Ow@UD>IFW<fUjByzadgkes}r1>U#-$ut~5i=hEeAsQ@Q3ZpArg)uSN`Fz^`LVp(F
z5sf>5P4S^lfE=-!?-&ar3)708(cP3Zi9|5ufcMfdY|y4>5w*bJ4RftUaj8>W1}NKK
zc&j6sV3vheG8=^I>F6^e@*1ND1A9XNx_ndb5m=f}2KC7MLzv<1dEtw7^1v_x++b=_
z#ql(7FiLg(_gzYO=VC39(1$8`us!AEXl`1J_X2kwx7|W@!#vl2398uJtl3Pe`4<Of
z=^K)JPaUJWqIPCugF`j90aN+ay8U1RgIOIdjpxBeRtBfl$CkI-K~#eBYjYW%qOK{@
zG?}%nD*Vx(H7#y!Sv~mS8S6LpSMx2r{YzZ3Mv!D-8RYkZ;Gy#9Suv3^!<Sjmj&B0L
zn6Cki^lE}1lY5}y3gQMK9(>MQTT_oA;!UA%8fB&gC&L?0jeZoHweWoD8Em~(=GHfw
zTxv34rmGm;_hrdxt2zrwT60CLcX76GcxSg}vkKsXZ4L=z32FE9T2r(r&Iei^iIRux
z5VpnIs9W%<PExC9Wy3L#$`9g~YvyF{rVfStm#Y4GB1w9;ud7|hf#U?H17-2%UjU!F
zVy7iBF+=!iZ}Q;8+z*VUwFz63o5X-;E%1^UAK-gI82v!1XIj=D(?BuEmrrD+gT;@Z
zH;)V}ybP*qYfqfTkW8eMH)RDkBhC=m6+tafbn9N4j4Fjw7l52e*4Kz%=)G}zlnY29
zznPP!pecZx5vj~86%;XK17}Rlyi7Wft@{UnwW+tPP?A36Pq_Pm9u9F%CS}zXg~}=>
zASoEC*>z>Ty;yOX-@Nq1<4}uye(1<qxyWT`Gr<Ad`B)-8kW8ey`yLx>Y5IK0T_C8&
zkwk5`;yNa=g7XM!^!?p4kj>TEdk&`G&<@l$4orFBH(mUJsrP)BVZ6jKad=&}BttzV
z?^1$0oI>4A^OjAgwP4k-4%aoO-jDd=qV89-v?-(^aDZ-oG1}9MymiOi=mG>??Eb2c
z)o3r91}E9huR&Toi3|VZ1_}PqVzE}VOo6CU^77D%R1F0axa~sWZiPI#H&h+?dRMjY
zdu1JT>+0=Xs_82D>E}LMs?D}oxr%zdkn_(!ui4&ZJV3uq;dd&koad^t2xwi~3cHcp
zy5TOCPLE}d#pN*AOaUbUj(j@m!ElD0{G1};5qQ0FrSWa-Z_ISYc_0S<!W2}AxvI>-
zGD__H6^%I<SrU>21E3>XHlj3=ANZ;SWe|{!XNee?FbqYbWjlsMfb9s9kJN;NapYH6
z091k;EU1a3`g^GmetRK|m->*&-N4E&L4b8Tu2}vf1_yR!SZCuRG<`E!;nr+oD;VDn
z^P4LK5VrfiH|9!w)-}(fKrY9PDg%02pE^p4U*8+0R4582{^bz-<&Z!xdL+3RcRWcC
z&rmlNP~DqgY8sa+VA<-)l^8z}A`gwzgtlrse4SSB<PT2z)HVa1ecLdGb5e6GP_*|;
z0*&@VC{^RTimymoWv{-i1p_VL2W4HHtV_<zSAB7vi}kfznZlF1NrqE=iFLAjN5&4m
z-iTYVhmMal?g=Jn#4Khe@5-iR6G#0K6HYW#J9RLs6hh2gd2@}s570|2CI<g2y!x#r
z{CQ#=Fx7L}&b7|-E~E&gfTnExW)!2U{%aJ++j{)Gf7%mUvy_c(85-n&{nW-yFtQk&
zaIBHMJ8o(&c~*~cG~NdsM-M>)z+v=`m|L|pjr8K;21hiVY#KRXO!XtsJ0+zHYEyiH
zV-FQOsUP|>AJs&U9MiObBkp|Ly9^rt026?&=;IU%Ta8aL4n=x=mFoQvg4L*kt2>S^
zh8~QM_;UxxngkWcZj|E~WjvH@;0|K~i97_UJz>0tWl%CZ{~J4}r9^eziSTZJGIf;T
zH(d%_>_XX1N=PlnA@K<=JqJg|`g#t6ggrq$eb(66e0pylPg{}wsph7Kfy=MJmt(o6
zZ0_TQm1nu2X_!(Kd59PmYkI$MP{5ZPp|(-jXxWq&<zInnrkg=pEuERJ-BzPoFx5vY
z%e7^0n#*tNSav%@!7BekFI!ZfY=x)rY*m`h8<ZoyHBN5-DSkr*d+VeO=@+C&YQcIA
zA{v4Jt5LNKwXVNvY1zqlwaYQ?hGcYDm{S<N1uD>O=p3&;eNfT*n^FBQ*t7vr#V>Em
z+wlfW58{s&I@6X2?8~FZCXo*6%&}B7O6cg?O6_qi6I7R#-VIlqTBHx2Wu;?NWw>pa
zINE`z8oE3rWL<N@)T)pwwUOqr@bxGlhH8n2DwIuu7scT|RVy>Fnm79MpWt%Wz-H7c
zM2^XtA56F%(*Z1FV^)Ya8pOv`1_qjhAoYUY(-o$|IBIdSC=VV@5=V)NQ&1WHBp~g4
zA*$|ulw^TfR>|4ZzSM2TTmj;4On-+EG3n(M<36tq0uNJd5Wo}~V|Iq7pW-=C7p5ap
z#R#%Pa+pn%bTf=>f>WP;hC+E4PIXK|IO0245@F%UEX@F8NpRFw**@Yb#Tl|=vvaJJ
zAH&9w2=F9|{4MkQ(LfPe+X7wG*Rv?72UD$~SoX}A#lZ%+#U*v3J9i9h5TN@>axvYy
zC+~`vKp9MI92FBU72P4GziS~gZ;9UcpC4o1QNm_64c)x85;qMQ{Z#N_%5FT6Ui0ek
z7=)VVX090@tPKoL#dv8%W2a&(n5RBaKqonvy^}UdzeYRfJbbLCv0l8jyV$AJ$fu8U
zj)|?$3%;t@X;j2H?R6SaFXs<wx0K%yLEFsHA+mu@n@^3r7H_PB33@80CmO@uhKbCq
zR1jRxUFUbsmMU2jb9HiA<$pts7Za9mN$aRCv)*tX7`fo@;#YDdlD||Z=NYJJC9!VF
zDc_x}a6xY!-}+jdLs8of*SbB#^jd|QZ+@`6R0W#ZA%K3cVpJuimU5H=r*_U(t%AJp
zg{iV}R1tY36U_*netJGf9fk%(Pa)7tLE@Tz?F$8i!wwp^y&ZX)k$8b00Jum|%5mtR
zGOL|ClzI~cGtqmCq)$NXLEKL)nfP=p(Gb4Jb~G?0fU?bffs`&v@7p6tOk#m3m=Hh)
zVgQ)@B>_;1pL|gcq0vp&t;B2koJpal_t|}<E8XLX#>>bM6i7=~yQjJaAZ-ZaMif)c
z2+k(9I3OS=au!L!9ieE^m7itwGS%2_=#5+kVPFN*VDnUuD6$BO?yG$_uxlD~4YS(u
z@$(hX#$lVWwhR4zR%?z3!FN_rP3)B|2}vL1;I}bi*0$^XGAOG~66^3!eJ?SaD!pD<
zLC?#;_^sS8*NdbLFhiWRyngIk!1}aXVS80NM_w8}6hhw(M4|Xb{A+0;Uv6J7PV-T3
zJGk2Rs&>YL)ym<w=^}gxw{xi;4R55fvTE76)}wIPxLMcgClFrcytTqy%l0xxON7@w
zQvjsoLqrHBq<?`y_Ff&kcO+roGlw~nq<&7R-1cSuAXW{HR4y^|H9wGu^P7~5MFSPp
zcYSbj8ni)rzAd???fxiX@Zwu)W1JPismBv8D`&JWVv)eXAJX>5!O6AlZL{Pj&vbLC
z4&EyEVynYpjC<G_J~zvPZ-MIJW8YNcb~v2nb%cKn)!9R=a75Ko*2)nf{-<cDTvbV<
zKiB#1F8%J!-NX{fzU1N{6EUF8;z1F^T1HvtlwB4_4b6$D{$HbLB_@Wpa~6g;cWrF^
z%Yv5TcRFk?b23NQMWI?+O=a>I92Iex$Q_S!EwL|qwq1a`tg+4cSy8@4ZJvf*!l9vn
zAT^<@6$~#U$L`jv^yAwL+VK7NkD47rOXNtcexBX4$jFbwGa3sZc(%`fDRkulLhdap
zIdRRQ3($|Dfnq3jXp|TTym|{=X#z+KxVrC_+jkbm-f;~tn}Ofg{a8ehR%A#(Vi#D3
z?>0}XHUuDw$yKmgaDbrpR@k&aRq?FW!+OBGLij?|bXrDiwon1c=GWsd&WJf5VsVzo
zIRTc=RMw(BX<s@)a`aYj&+dX)L!PQvsxY8(S?(cKiOBV13?1bJE^6qet-|i4?8tof
z5je5l7SI0+0`l#FzK*3~|FYy`cyWI~27>Bx%sc5Z0#QTKp=~3*N*uv;-94?pQ|i^o
ze@HFrwh^DheW9O~2296MTd*B8O>fNI_-C1zg#_&81R>gAsT^nf+zQ3@Dk_^Vt|M2B
zk2hRcz8DDhp)j7c2PuxX+xgoYxCCkI<(rrMN~nMDzyG2h;kOakeeqzrH|nhS>cms+
z7?4wZn>~lOCj2jxlwb^7S);V$e349ql~zez?}DM)ctsyGDETO^1Ft)OZomIn=bj~i
zqQsbVKY?!?`G)kD0Xw;>s2~UAqHDOjp>`0-=Z{$C!_~JQAFrP2x-Y2GJe6C4jc;-^
zeshO^1_fSb2@3iZls`k@o)i~1gI^psJN~_#%tDKpb9hea2jxBmv)ih>ndMrRjZe7<
z30f?8mR{-hdMqmtqm)kc7{|P~qi$57)M2s7RV9z{Itxgnv`1Ma;zn1aJEu@)cM8ii
zhfehh4n4`W?VU_#4<PTB7cKb&H@;p=zwANzac=G3lxL<&U!e0&kGsoF7*bv15<W3)
z=fl5It{XF-!)N}QDf2m0-G>X_5p(4%Or0b2pI!zfz9&m1wj#ZV*Z2GDL%6#-ff#F3
zLzDy7kTx(PgT*RrS7&fowhFysn!uA&3-ef7?$EYMft(CvK%VVO(C*Gc!i}hn{50>g
z8Y#xh=QPMv+HOL1z~QCZY`JiwK6ZFddF{_s|F^%K{@XA%d0tISr-8EXbssC-?laz$
za*4w045Rrr#F8Om&^_LByjezoe=w$h95$<c#e4j@Tk7P}uw#2!^YA?VF(MK@d#x&R
zP)l#&w&+P_qUeLL$@Tin#b-B*#`hEER+IdfTz*@bDWA7j*D_Ni6r)F*H?K^=0;njn
z2=I0}E%r$|89P!Annqq#>Qj4TO#26{`6EtrDU+#L>AftD6b8z|JLQr4RW~x4?^!WS
z+?!I-i)qpfk&UF&)M9!}!2GkWr?a%csvODWWrIV%G<fMh%X`;lsomTAu}mF4!pp;6
z#BXjc6|+|O*UTqoNdU|8&H~CYG5`>a8Cqi|Wy;L%-dJ7Cbp<GVgGbCDB;b<C8k8X)
zX@NG?LL8-p#!`;SB&-*1#%s*=v(RcJKpr67szLx(#k`O>aXpSvKn25C%2GBOBRj#Z
zUsT6{<O{wZV{dpM^5z8J?7=&QKRxxzbz5{n?Di1M-uF$FN*2)Ga1NmH%?WyK-4q?F
z>@k)?p^Tl**pNw9@(N4TV;lf%nrWiNQPplmt%v~0O8=6BN$8qR)8%{ipMd<za0#VT
z)FqaWRc>DS*hOiRbb(Bi@ds6pkdDk;k{9v=sISc3V4MLgAr1CPi(<!W>P@U{$vhU4
z=@1!_ulW>H-T}3-zwmIl(m_Lqith&6p)thr4H3QUhYf@u8J|cqS3kpnj1Vmh(ZUgy
z1M`cU+x=c$+5l9_Z)31S-QG;Mz#)L`TW{|mh{}E;lAG?<Zc_Ynvj#LFSWoS^rS-1e
zQve8M*zQqK5ah_7?Stl3+b5{qp5wB$R`X);Jo&rrtNA>0^zSRT9xv~1Y<Q>=4m;SI
z;l6QekmAc7j)0et^RUFBLR*36UcOqJ*RlS=`ULC7E$%<4aK4V462|;6@rTTR%S}8?
z9tD{Li}H`biMAb2o~xXHhoNv(fotKD{bRav0UfG(6Hnp%U*=>`_gO^4Pd|f7hj%6H
zprMXr_&-S8=ANO74zIpKOYFa^^x5_Bie37y5qOVwLjS>bxyolYv}wlOEZU0vy5j(i
zcW-NG%3Qc@{;8MWHrvt3lr7zM`>}`ymHRLht)BB)JG{P9+lsiHy33TZZ7!cy>paN@
z6oIC^n&;TrK-geT@3?WrD+GvS@o!x;XoF|fYCMd&?Z8w(dNf>{yx-b)t~jh(#vzqa
z^*I?&?QFO!GXjq1y>;Pj6E;}GsB34pBUiH05jiADB4>f^LI>qse$6~X-ILRy_dncV
z^*aiML}ho0oL=(g*cddp3?K5baQ%*fQ-B4{2M7-SaOz6KUA{RDz~?D!hsgufAPmUC
zp>T4oJPgk8i#j_kz3@F&qwS+0P1Ux6_jO<@M=eFdwT9rBU_zvO?L3g7YXeK|$OdgK
z966ONTLA$RT3J`6l8Y)WGutP)u}YE!sn6Be1@N^)%^PUc)kp~yWF`ac7c;*?S=)a$
z=%eB=2G0eCR<+0n^D{eWeT0hP#VgM)=VAv3*z%moI!(ptbp#eoXrGEU5oa8GD5*M-
zc3i#b;AHsi;RWV1hM3^vTAOj*;+8YRuEP;Vww{_^4WsXHTuZc$ttzwv)ZE}bGG-G~
zcxqjy8t~Elb!cxT<ivs%#&wa{un?4onrISaTn@{boJ+#R1O@&+nl31x>DdeWdO}#z
zLs8n^HZ~GDtkrpqy!V{Y#KkAbUn7DbI5BF``7d`m)~XNgrfF1q{BSkE*39<Ws-r$!
ziPrDm?^>mKOqx5Ru~q-5qi<c%iT{u{m8m`~|7cQbP`pbG_jA;@dwz)chbgrcEp@}e
z2OsK-W2oc)F3fLJoFS%uepAo;XSp;<wIZ((n6Tr|qhI;U?7Y6paI2Mm49~*VT4X^Y
z|2NC(#=yX-dn=LWkf26~CC?L=t!A8OVIdC7caMIk-`$Xhb`k^sy}q1T%_kUQaa66;
zD%Ue0XPK5Ql<6pPQnGd=hL~lWe2ytmNB@Se=M87zK4m=mUD-sr`n8u`7#1jqKTsM{
z;|QoW4^YrrSjv1<IG-6<2mW??r>0-QYn<gvSK>kHnT6{SzpQJrP3GCIB~T5(Y~Vsh
zqHs1>4~eGxj)*{+>NL1NW^g!6z)y30Ow_!wc(>(a#4BKI>B><_R`Z%&SfV86eeOF-
zyuAn?KRM3LFkz0pt-$ljEAUU=L3G<x+Y#vZ=!;rCl|D;Y#6N|1#;|)4C8SX=VNh|a
zzg1e{h!WhS&43WLNLIbDwWU_c+Z$LWcS&5f-{s(6xqm*Fl55hfbsGe3Q7+vbaAe-^
zU_d-!Nz>SkODkFUrP#xV&UT0X9^$x4lHzeXmkK-C-LKI)L(zX5kB2F$nmv2Ya!Qqx
zIu)YEFn){ce41Hc_V8kQlARhECEm7VKQrTEBAH(Hl23CKwqQr171cm;8}<{kL$FzD
zjqzz2^OQBS^z|lc;@>Zo=leES=<D-I5vP;i_R48^kusmnnKZX@L}bLoYrs>5n!9<H
z^`7HWtC|Q}Q9#W167Km1NvK8=T0&c9`e4*;4V`3{oWgjBr%kex@H3m+NCcXtdbMM|
z_nlsY*SSb#6RD$~Hetkk_Q5^slH9GQ^xX!;GIFM~*S2^;a?NmU5U!-8xki7l8U8@|
zTP_64)t8rG&{}ph(KJ(FFr%p1f!hu@gFa<6lP*7S=tR_;Ef+=+NN4l9sYAHPU84Rb
z8#~r88yM()i7=3ikKI_K2uxwkI8GybeO{stUrtT-t6*vClczLgFy9F^FZrSiQQZmV
z=i{=iCn?dFT%anjbfp4G+5SDXQJS7<PzEbKbnhJ}@$NR-)GB0e)`wQ?E=>tk>Og~Y
z8W#vPJ6lqDTn6CBY-a_zgzP<gYKP;-*48r{L4K5>TpXk|v$~z1UIb>VT$iJ!iZ1;$
zbG{1u-RHnvF^`|ys<K}rXOAOQ79pwP$839*nj(0N$+JW|6Gf$|@HcAOExoDvK29w+
zjacojIq|zU@ZWs-2y+YFZ+7Ap^gPuLN}e~#!*=^rv)D6^dG~d0hM|OIHGlB7yu=dD
z&ZPU2`cFWKMVkS-)GhPar>c}$C>5`esldC(Dix>hF_(C>)B1YJ!_m^(qu%ZpD(Lks
zQ*_66(5%`bJ@c8by3(ugq@i#%?BZo-V|%&OQ^&c0ut{0LQT%FcFdZ+j+cyQ9S~S4M
zE6>$PvwHC2*Mu2A1Ld3M%@7Cex3KbaYw}F1+gxn(Z*RSA6oco#8Y(%TvwQ5sTeXk0
z?MoictH4b>$|OIpr?byQ+He?a|4k@muAa0PC#ysank)Amc1w-m;;#rc4k?VU@t#G+
zF9`qiQkCvECp$~rA6k`f7HNl-Kt<z^ULX;Pzf>AJs~^uJJTQ&i80%;WXWGr#aO_|P
zlO;`iW_GFvhL2O~j)wDNI#UjA9702+srFu4OKA(S)G9vDUOi2)7boK<od!d8*QpKX
zG6tq0S$<X|vEuZ?sEwii7RNjy|2IS>A(c@LUdF1fZY_#suQxEX85H*FGNeoRvFp^j
z)VWhewV%yXg0`vh@5Rw=<6}=&hjwgAaiHGFNp1FNWwYFk{<?P^|J?x|ok(9|pl629
zChrFs>#il~|33@h`=^gV^Sc>Y6OlhVsK#t49!|+#N1h)p=_s#Zjj{iw2<#Dm+T(kD
z)L1yajWI?P)Civ&xIK=tswh5=Ba(87z6=US+oZ23{cUsFh=RD|95`rW%*CXq!N&~H
z*PXThtqPZvagui&$6=Ks`XXo3^l^Zz?oysBeQOf+-h6guladzIMV<UlCEqbw*!rCT
z2g#N_7%<dD5V9rM)S8=Kh_Yyv$AI92D=Vc3bjkfii^OXZqME}sU2KrCl>jLPOcFex
z8KNACnm4TGUgnCQ@(XK}8dNtO_l41g4G#Y;&L0&{Tl!S4&?zlD>=Ala7yQvZHvEVy
z4r{bB?RIa-B^jT4d0OBXiGSXo2&JfduJe}aUtyYlm8<vg7EKR_9zhZzVS>Ap@k-QH
zRR4<Ct!nPFk&d6<B0qiK-$5~#7M-3Gwr*!xR+&Y{H<YtKq<KSW2kF8fqOI4ZmWqIA
zD)__FJpW|=-LP%u!avndGJK6H;32QqXT0Y3|BPrRH(vJ?e*Ti_rf;S<`c>-GHF2`u
zz{@(}7?RSuxl&s(+#V9ukSugsj6HE|6kueuevv5pvdexQe;z36-&P$Y6Msy1ENt|g
ztwY!R#Hu1>)BSjQF8SR?i!4$nwf`PSVbXKH*uyT`^^`BWi=9)Z{7>?-OjToaw!AI%
z+?V}0pgYH9@Fv>zMXL-g{H#Z}Z<dKAK*O}2q{B07%Kxy^;=kp(u}^<`L+V%MSW1YH
z3iKR#%1Z##H9h%L$7;%|=skZ7)@j;~jpFQvnOOZA_+3{P2EMPq=ACD4yd0_|ZOhk)
zMOUu7mj7-w0yi~C!s?ygW>RXq#3>LAVk>I#v6NOa9p-r1RQM<^tLQYWtB(ivhvQ{t
zko~J4uowNe+qH++(=N)c)W#d@l9jE=^CTObq=wwSqOPutMIJ$Eu_<4vQQa6SETK-2
z)`@qYr(K@_&1fRt&)|FbSmmSd7%+bktl_9n$WS0^{KwKW^0)f0JsAY!Aaj2ehv1|x
zzq`!O(4Q>09!Oj7h+T4+7Zb|Dkzxabl7BK{IA(;Yx^TfvqPLV93Yu&1@KBdG<-9|D
zoHJKV&DM_QwKb^x3pNNIGR`HVs6PF46ZDi-Xcs!|E3400jM?w~n!}TIT<yJ*J3YNC
zxpaN0FTynOSFHA+TIc%UQAv{|Enl*ERN}j#1k9BFqfjnon5&A&HS9gw-nJBu29J%=
zzoTmiTXLT(ytp$w4UlTM=SKH>Z?VTyygIMLn2mSit^{F`$t5IQew7wKBPh~|zp=~~
zU(}PHN26|z;WUTJ{&&O7YFOOInLhtdPUO$mybO*>;VQa~+<kr7$cHWom4@(sq)F41
zCsYfpOkW=nS|$Sr5BD^YaW!Vd5{~_t2=ITI(|(<WP3N|W?@?>alvy7ia&y)f*O4O|
zG|l+f%%V_d$=a=F;tpZMMR>CTC;>E8kKLuX0Mg=MJ07I2um||IDTXp(@<^Iq{1Q6v
zu5p>w^f)eNfhTjqnt%23PG)Uo-Ob9ucoTt)m0F>0ps~_-zCWX2NVJV|P)1W0rQ4dU
z%5JR<dAl4aYd2`M_+}L#LDK#XC=f@A-(Vl?)dk03!_bn&9BI!XLqdVt=W%SYRm?8;
z{88UYNYK$!fnbL{-fhLJtI&zsY`hrKn;sXO=?NsmSitQBucDgcv_!NIqaWo!XYHgT
z;Hr*EFBjA-{V6xQj=qsKmgkCuC=;w_tYg=$FuwI{eIdV$>TkvmJ|cB+C$XSW^brTA
zM%u5M07U$Kcj|^GcxiG&Y9Zd^L>neX?gLwHa#6UA$KiPIC9*>TxjOcdW0Q!SUSR}D
zOcC;!Q(n7LczRNijt(>Z#~+Wjx+H;A)o26wpESr2<JdL~oG0~;i1I)Q-*{hQfzmvV
zx=!yHPqd@F*f_F(?w7f|v}O=H7+qJmoo15U!``SS@Z{f7yB#?0UiaPl&4a#00N0--
zqZF!_Eh!n|Ps1V9N=}4u!Ts40A}91QGH0X~WOWwfAU4e73r(F1;Wj=UY{HC!z%%}g
zrm`YCsjaM@vOiZ?4)b3B_F|2!o*HNGCKL+E-VPEG`C-Ij{Kz@!&a1q4B#Ns_hCAI6
zs-NXeBxh#2n(faDwjw;&5GJv7PvpAbc5G4jRH!7zZz2Rwf6#kwbNdxIwJ3r@m(2lF
z&3+V2?a}v}fx*w~O;iO!^_NZxTAQU${bDb@3Q}n+6nmJr+!lEfgaoIA0yhf@e)W`m
z<Kq`M@Jm=7Tf+FcwZ&z_AYaIa`JK8F<LboUv4cdR?*Zkdd&dJW8_#LQa(})xKn;t{
zbu5}LGnR!fNSu30sAtW8tJ2umxoZZmZ$s#*qaS_EK$DxQ9<o+)YhH(=<(Z2$67rRz
z!OQq|trNU*cRdwY!hdKnd??JcXaDoH<*kA?M_RMgoPz!vUf%sp*$u|e!V;UI@&VeF
z{mRmAU>WLAVDOl;gYxeE67f1qkn_JCHq2O$Z+VVbC9ZwzJ{@eF>sD|bMzeeTS+;=w
zmQ$d%3Eg<4_Kfs%4>3SA6(|<buckd0I0pwH*Vk>2tqWCMp`IpZ$-<8|OaxZ^j`@?7
zoexDUe4B7ouu?m)nYlHspxb10sU9t{o2xS<jY^K5<18XS@M=^&7Uun!%5$!dE(obH
z;8p#L_iwT1pZ-6qAJ>s&4@D1?iDyx7s||>{K6LFH-jPad20hG8%w!TLDvpPFEgXhF
zOi_jpj=2^}6BT0oF?dKH2P;y%ZBBbgd7EIk^mpgO$?L6cfes=5L&ts3BH%(R`L*;f
zH}w0LSULO_=8T>#$_B>1ee<{PzdT;@!|&Axm2|fd+)I|SIiWVUw0Wj=)v7n;@-lx;
zUm+W&MfP&T>_b#qEHVUojIRk$QnUE?=+)CgxJG@4AON63(xWvE#_axd<}u<giVGxv
zvcR3kV8uN5na|eRW5zNfr39K}q5!GldAh?2SdVz0h4?7>Y$=&2Rv~QP5Lh}xm#H=L
zlMz%fej17)VBtEfTLFZr{d@h_-UTjZ&NBpqri%<EZqb4G+f-P6c0X2ym(1$n(=&S4
zp!{WVH(zpEa1OeSAIwQY#nwZaH5Iv`tI7&|PC-YcKh6ZHcM86c_k7%xlaZ;tAXyQ$
z9g4KuSD)pf|D%$_^S=BCH&BeaFw*MaP=VpQ&cY4q-`)VMLIiWekMStpSiv*%vlhzy
zXZUWf1Xf5GdU$HCuO(|B;vTO>)}Ll7q6`_Tpr`?rv16RXGBXMPZ&##x!NkO4nV!hw
z-#w}-J3eYtInd^%xDG+iY0bMp)-#M>2?6rHvV`8k4C5=LT`Rd^BP5I|6IR^iIz&tN
z1_5b+xb99{$EU~By>hDp&f(6#^loip*kMKqck=vuDFrW~xTI8J;`rOM=l^cX9$Ls=
zmjLRW*D)>7d+6&+j?+{9Z#@v($-a_+%$FN=Jqix(h;aFlwpaw|<e?lbS9Z_K@S5h(
zJ7;|`TwLeFZ4Kv}!^(6{ihX=JC3_yZV8Pe$7w<8)o@o`KS3RcGOCWA?s4N)Frvp+Q
zFIIE&c3vN~)%~={PrkU<Fm9bVsS>xB(4mug><jw4`LsL4LpXgk3yL$L*t(y>ME?6^
z2LF75CUSA+HH4!5Qtxu@v&~#^&B{F5rjhI{PYOdeW}ZHW4+;7Neb1IW*HUKnZ(Z0j
zKKOub5Sh3CknurqGkROrg~p0HbD6brN*uX5(=b<Tw5dniP9^BH^>{Oi-&{)s-*9yn
z!jfk9QFt8rPyU7I%Pk^=58taN_xRuWum}ZSYDg_7L(pG+YUb@*=-@=>UJMiEz1^EZ
zu17S-_#~ls^nGdr+--2VVGVbGSChS>4;NDyru&|RkI&rk`RBgreuXnZY1cFLDbN_V
zoA}{Lrzda(S}DB_`yqAm#dt~3l&2;K(bH#&XxH;*0<UWwW-M5<x}B}T!lrJ$`e`+0
zdArRwmS4bxTDUqBbGhZ`h-sKN^iA11lY$FZ8*77VQ~SKeM`YK$B;<04KmNc{Bo-mj
z8PC+9_J-CTauu11PQ!9oiNbjsQ+yJ-R5e$C-K~-JpY`E1pMI})>Yw_8^<NzaxtIBf
z7OiEE57%%#OK>gNueOC%bT&+P4omZA3UvgP9*gnnpn<SOFowBFMXT>TqgyNO9-jlo
zk+*m??zFmC%CT3_vxa$7Yv>P}xFWLG8<y)$ry|4W&YcmeiP*$d%Z0Kt<{aML`H4N8
zcuz1xkN@VP<v%RwMypl%ol80RNNx%j>2tL~;l$%UT|V2!-^qKL6BPI~ISb7vQi&Ec
zD5Qkiz&4Gxm*qC65q>fDE9xZw%44b#RV8gVDqYy^xVc$>DCq>NF}^e*m;N;5=9*&D
z820WUt2Ez~h40?Jhu}^K8UJ_T`9H&R?-kIZ?;uMs=CVV!M?b>S159j;Z|hfRgp@%q
z-9Nq30$U&VRxWnTLYgFOA=(m``|8eT_0QV+#7fo%fd-da7hk8`p0|F9t`w`iwYab$
z>&F#NM3qR?^^S-Py6Z8Krc1C2?$PYc#2WXID^Qr2i$_>Ehj@Kvs?t`ROj@=T=Pr@L
zb8K1`FxLuxS2Xn121wTx3FF;<iojL*=-2{~H2)Rx{U}$Txfd{_kuU<H&Hb5RMgjuR
z4!ME(vT|d0DeraLMYu8JJZ=I|+pT|~nDpUWGb_%qbl(%e`ZN)j>#8aew?p>xbqq`N
zV}CWuYsgf{6e^Lg4u3=NGS=<cRJK>uCj6`Ut`#u%HV4sAY|ul;z)>0V-~dT%btpTz
z`S>VA+zqyOSvw@;-}W->mAUb{^<hC6+&dE180A+YC6xZo+_qtJ*jpQrUiQ&HR5sG2
zm5#>CTHqXqb++Dm^gUccVG?5{xb0*ABfJ-s6JR8-<6uH9|Aj+{!<Lzj0MF;1EOj%Z
zN%Mkuc5lDuIjwE~6%pR@LpD)$m+cSyxL6waebrWf_{oBQN4)RNw3`k{TxGDpd)GEN
zAZceq?e{$&*4(3f?$dN2)~|1>CG14M@~^9;KrI=Ewf5fS(6NJ$+k{0$QEx?4$Sd@|
zL$q7t<pz_f7)3fFoWE!fNfcM?&ab9ztjByRB^VPzj2ZE8W-xN|Jejw*P_yP}*3L{n
z-4R?l+s|XSCQ>Qa5imUYEiBFm)2H-er%xx(%qNB%#o~PsT1tklXy3IOR^lYzuDRY~
zE;ftl8SgeDiYg*}OxzeH_NPU4Y>9&Xg0)ZRg8u7$xPz(pJ=fU?+iU6oDus!HH0>ao
zaHe=#x=q7|7OQ+Z5wPD~Ssk#1KNXkwtz_hWx?J%808&A%zJ+5l{f|$4QtNRks1%?b
zdxuQlV7D2V)dqLT8FMav-RsRcXS~8)nmZ-|Bg`S!y!G<D@0hXw?tAYuKfm`^b7jPx
zVC9MRBE%g1yvE*A_C;Y}gmuH2r=D!bUM$qY7NmYVsOR%v`MUL{;a`tBhxgaK?Q(0a
z7j>U&yf5}e#Gk|p@X^N}k4w6w*Kw(ZyedrP0wdva%Ke?>p2`6uV**Ad+7wSj79<WB
znNVM@wJR;YqjhWhOlOS(NSDjgm7Wg^qaJxh2+JRq8()0upyl>#<vdSATa3g#Fz&N(
zqg79I#I|{%Mbg49Zqngd07Jg`jc-{YP|~wq^R~AbCUr;Kcj9;VEAFQ~Mj@^E>^}F*
z`Gw6xzyH1Y?vL)!<~0zO<Q35TQBw}E;m6I-T=4a0o-y~__rRjThe|yJnuve(u50Hs
zU%&FL@0iCbXtDLSH@z`0#)@k<f)oTQQNjRBm(M)y6!Xf{PMzBn;Cc*ApT!WdUT9C4
zl7JD?Zj<E?i|)0n6?{FLF@O5Y<MLzvP*DMXdH3g?b*8Nni<ggm`ZJ0_P#GaF8kh(Z
z<J?nF>2H4j2lKr<eq!rS`U8OTw>P|33yJtKgK{V@O#VAW7{UfA&SHc`a88R`{Gwx?
z>uW?XzP9OJt2BbphA5Ya2Xp;Z3$1Fxf~a!KZ=`&@oBa~dw_+tCz+*<5R#S54gN29#
zN2c|g4PXnKX1$XZZQcPhawF-$MV3shTeK;kbEl<pEV+I>D_Cro6-ifjyGf5^RbQKp
zL4v!ioGeRPb+1F<L3UK<63}BkGMST5Uz=1L!$d+wHNwU)u~sXpfP9k^838A?b%|5x
z{dNzmkt?z;>+TEJUsC``iba$X7%_Tq%x1?@1YqYK0_e^>bw2jZb_Dj{`nK7#|DfeX
zd#ckMv3`TT<E!NoXWso4g%FJRfB)!>0*?dmV?y9kzbt8ZpB=Ie=@1Xc4h913uKtv@
zC>g<hm#ACTJ-GPl5(r{ldfBV`BOdpz>Xr*GV#%~AxVc=i2uLYLQ16Ox$MAY6YHi}u
zBH%}u;?r~6cL==r`kX+7agzB60Q~W%o-%)a{yEdF4QQMt+rNJH3v*ThhaY;VVho*R
z=T*vm?8&DVSY-NjSK>S>u93QP#efmUWGoQEv>0H-FALX$xS9;pklej~^i#KlFWcX^
zDFh6<RIZVn%fj#9`BU?dEC`dxpCh8uF^BzwG^)dt1dNb&&FWR=9L=$ZO3=u}^UU+l
zo4fD3zbGXNK#^;&c<Y>E2CgwtD5?Yjk{~!ta(jto2YuKn$i(po#gS{@{?<fSF^|Y%
z=bN(JNwPw!#B(jS46JJY_J;T5xznS}Y5^m-IUbd(zBL+Smy&#U%e8FB9)V=*dlzpF
zxL2MFx;2;dN;p|8CCvxLmJyY%-sJ>QZ4O_%N}dFKw~+2Wu@2dfIP&`q^TF9?oNiv>
z#nM}hB|UX8=A3c*8P?UEtJ58mfDzK*MjT~M#$c4C9(nXHiq-hrqV_jfj<LUka6iR3
zSc<D;fRQOFsW`VkwxHQRap0?;=KhC%Ykqv!&;3i0=ly+s=C9xLZtah?K8+yed*V(0
zLE~>FDZI9-MA^79y;d@&FfBDO5^j(G+W{j8x@Y8imKY;T35@t{Yo6CH2jUbn>$k=E
zEvwetnO{roepO>Ui)y{Ipdl2+h~kmf!eWGWR4w>An}esMNneWAKonyteQi*h2<tZ#
zUQ{Dk`rx~-lg8t`38QNZT8}@w`<L!#<dH=WE{J^B4Df;l!@dF*A1ts~XhG!pA0NBX
zjB24$3$&DQvV?SPQe$%93<>;SbMCw*Vx@r*7V#H)R#1*hNx%p`C|AjEHm)O_MT9Q|
zb7dqQqXuGv<`O(NOof1vaQb3@^DSd-g61exX<&ph-*feqDnA&Xu=Mf$_j8|X|9sQU
z1vedN661H+<q4O=b<EHANBDI11v-U*5#CFe+%&}rV*RpgW6Jmgzu$9p6BvPY@Do9L
zn>LYH>R9b)S&KN=TJ2r3+E7d?Yf&Qb!&(r`nu_NR%6dfL`ZN|M78nVHB36S@eH(XI
z6LJ^TE)WGGXu7i^!GZ#}T3K0B0EBeNvVvG503itH0bn=+g4l9oVFIv%tE~WlQA`bi
z2E)5{1bJ)?_3@`epk7CpTm|wRmh?>ee^8oaBLam0eTb8Cj06WCeCL%)=F%DSvCn_r
zJhOARlondd$o{b)_KMtPoqN`)<{g(T1a1H6=3C4z`K%rt9+&Hgz;bGIC~dm?yJa<|
z_$&fH@JBzDmG55#_U;rQH72)aJpxW}yM<K<@kw|OP^Vmfu|@zep;g2fF#sr4bfv}f
z|Hlvg)dDvnPfE-3i?@Hvwkzq+Yvsn`^p~HsAj!W-cu&l>r6T4vUr0qe=-!1aMv`KT
z_|JkB=YK2!jQD~xKX0M)VF9E^DF)gN(%uhK5-{S@5_gMuC{Wz|>ATD=G_K=*>K{J*
zL0A0Tu@Epam#~bJ3l-nw6{pODmX!oXXwLa(pJlB(Iggr~+4FqiYu_*cc~BgWSOb0F
z`s-{wqNsAn`@K8vl+wZ@oYpm2VB~!QMnb;fs|Ad}j5t?Xb<L<cYn+BPQGDz>?$$z>
z)97*n_yqY1Q^A$sZx|TCH4+~E9brxX;rCr{<7#oOkR`pJ-`^AM9v+j+kzPCIyh6YT
zKzEo50VCn`#r_6RXMK(<M5tU~q))E2q6knL>%$G}*4niW7P`n+d}7`3@tZ%ZHBw;7
z?BDaxPd(`bDG!_<TMqd@A-9DGhZPH`&WZSu0PJCE1YiU~cTW2VTxq!&BdvO;P0Nk&
z+pN#uNik6>)nb2qYjj|wSf2RLTx_0|l8*IsL&X@0dNI3q?GnOMf0GNF`WGE_RGx4n
zs{LHU)E$%eFriKGEopNl5q_WX{~V@Y{7R;&aw(BSzO@R$)^CykGb-?LQW@2{SBu4R
zwc5Vi%f;e=1u9xGjw%g|_yTuV2gf8}#0~bBUHp1;oZKQo_zd5@XRkdk_fM^X5x(g+
zz3#PhV@EZB5j23%1cK1oJ^6&=7qI6m-}*PJLA~IHcj1al-()qwu*w-$aKa!6yO%(x
z|Mah)lF)Tw0PHA%PTax-Bp+*%I6-VwS^2Ff{0pE6z$i>B)na6HMpg}SaV6B<S}{yz
zb<T!RbK7J#t;;e|$GUJ<fZyKNZCN*?_!L-k5Olq>SHajNd?$$dv;^rw+DwOp?0_E`
z#n>QNdYixxVqI8^4@puuPLWn+bo9#27Ql#zGJ;_2B*Jr603?u}vaL(5v@AvND;(p}
z#}>3yf5fNNTMpA>&04$YH++Bq9a4(<krN-TyYy0X{4qz*wUR*K|L4bVv|t7Iu>k3o
zl?j_`0+0Wv_q|uHwdP&pv7!I3H{Phg?14Kd2e42Fpa;MRWjME3R75a$h%^&=Zyn-w
zS}|MT?SOzRXiUH)fqQw~l>|ndwvMgD`-3G6N*~U`z}+fV=#9|kWyc=lve_fg_s=&G
zcWU8yl$WrIKgtD0igl$0-vE@tR4I!Q;=TNhgw$aw1{m?Jc<%i9FAT0Flc><RPXGVU
zfAgUQMHJgd7Gx}N0e3dnNom0sm{l4W@!h+{&CR~%avYDGi(NFT|M-{3%<-Xbfk(al
zJKr^beC&w=PrdZF8SAl3#m%fwO<=_5#_~9K>-b7O&2o?vzz-*I5A!ArjD+v=8|5zc
zGSFfz=^1lgaoQ^)UC&kv7&+y{<1N_W)-&Y*BXMITJfYloz2oiHLKW-$N96LCeW*{Y
zA8z}`H_g+}KIdP~JExqhrEt3S@WafWef#7xYhr=bD`PTNJL7VrRqurL;yX(OMxJ^4
zS<|Qe)Y|TRV5wHm%0804;z&BM7{Q}JS9h<CF(U1rMaM9y)pIoKsVxV5`9s{kq_!<;
z)>dl{Fh?{zF!ID>kJ-(9t(jL$ee2e*H>(scp%BHHDx#h6(NBLS@%i5J#*59dN56PM
z6~zP!BM2xPRR|E}&Y%6vWF;RNkT$SeoW>0f2rDqvN0W=qDGQ0oDa8ShpsSEJ#e&9$
zE>t4;R;b9xpDbW_h8BcETJR9J!nr4i%uwJW<>69jOn&XxNYfg(GX%k%f-e#_)dR7?
z>_8J+u>;W3pU|QW|FkUliXk?pSO^@oZChUQdH_aTgU&namC~#_&HgM_Xo25xeDOlW
zG4Xk_O$*$s<=!GpaT>4Tbfx?TSRj}%zH}W`8yLA&fs9$m(|>&;FfIXd<H#3W-uvLM
zt%Zy4($D291{k3)LOFpSfU})4f96|ili(d~w_#Z)OzU1+ZR`)ZWdKBN!Bjd%6U?=P
zBfgp`kMDqYjGtoH_X#kFCUlq&r2-?tI3TxA5-xYBPfciSWFthsa&`G_s>gE{2m2Mg
zysfuiF+=8kSK~fQsE5W(xxBK$+!S*{R$2fddbs&V5s$GVnNe(*4!Jp+7#a?&OJ-BF
z=sptmE?F?(CJWb6SceeP0>bKJ5>Un``lsp$01#>ODR_HEd9gMDFu}zW=c6x_=6Bk!
zqT1&v#Rh5D+&~|}I)vxJeb$sf2TOXyfBN74F)t7^0_XqZ<Dam=$fQ6gTwTGtNdX_Y
z-MaSjH<{BVG=PY|^tB)S(ERe=hs>&VedgfaQ41n*uS>ucZpeCiJI%-exup~6lpD*~
z>n#>2QS1}NhRF(C>Qb6vS&I-SrbU)4e4ezK_l(bR#FMl+3{<+l&%Da_+bS!gU~Le5
z15xv<TWioppL*rW<-_m2-hQwC^+G~#=N9NKm%PCo`{JYgtoHjW7Z_pOV7<e0;ZANs
z-vaYG-*qWe31EaZmtBKt9CP!P<E;hD<X(y~&Zl|+M*OwzPk(WbjhErKF?`NAG`M=5
zU1QSMC)MW|x81YyjQ=QLB>b#}`bfB*Vtx~+3#AME{QFew8pF>Mch0wP(JRlI3sC%>
zAKhWjk-J-G0qgQHCJ-o{c_m8N$9IVG146A`FT7y=*Rw7PQ#rs0ytZ>F^PBo1YustO
z{@0OtpiH97cCAf0?ia9~n$Z~Pl?NCHDCbi-zzE+BK=6dX{Vv6QvU9WA@7F~gO^Xqq
z(@S^0bKlUSgC#EqIRS5=SYRY7{(|qqu^3Bw#t^q&rf%k$oYu<_W6sM@IyuskzS6)5
z?bvqs;dbop_5|(K21dU2?eCf^FMqRrUjGkZHEtl|2*tq<zZW-JZ3YEZJaZf>D@&is
z1x6UR+L-$__TuuKcY?fy<6`y}SY8r;j&)nGFPblJvn=V~ebp6q{pUYJlEny$Y>*!y
zocCzQUh$<v|0ZVTB8zRi-zInNfDv42VIeX-Hfk+KP%g>oS!*3*#I-Bx$oSSv93QzY
zE|%CYOLA3g@@BdsIRKE>-C5r;Mv|<(<MvWTK7KQqIKbT2(7;Gmnvu^b_Cmc<y9VH`
zApvq!Tbb<s?yugLr-kOy-Sqj}Y%p~9+dneNd_?&OY8geJ`|{Tk1BKzj3qPOGXTSVa
z^XDfXQ(#sElF|NP*_6CQRxk*PiGhL31~x`aco0HHHf)&qKm3P3n(eaY2~XCf1<Wai
z(6(?^FKE9JkUNgh0Dbt8KhC)z2&bbRKl%5&t?4T6NVXn!xIqJjAlaDb+YP3)+MjuD
zyIc$W-tyd+^M%Xvf1^E!Hr|ojwkn{f%Iof)+o&_?<NKVzp1=F!pDb^}8!*a(F8~fK
zP@;&1MmCc*Km6%kwyiD={;qwu4(_@;@-FUs=r^`KXuq5y%YgN2qtg&Zn}vY0q`|*>
z_<d0h8{vK}_RxSiOMt_g*+93h4vqxDLxAsyXT0;@e<9(T1aoCjITj<%>=I!BaY3N_
z^;~hm*Wo$O1WyolI5$vZ@{7D(@9XY|9x}Uk?6PrQ-hJiU=iaG%oG>uLcj(+oxisM;
z?(M?v@$i$rTERupZaz)^+xf12q7OYG_c8vv*}MOMx$B-^`pHT<_bo>L$Bmyf?^G~a
zfB~NvU(sq}p&UiNwc0IT{EET<?_z)v`q^gn&67?z!3MV-6i*mmAVABszR#VCCH3S}
z&zRrqTS7A);=Svp=^uzTk>@=hddMMmQ%T<dbmAJZ7!Vq<(Ro~be_)_LGQexGz(|)g
zkxo4BSZhH;U|+@s>fybi8Mga{JsM-5P}?8P3)WkS#f(I>Mq3-pk7EA-D7CA7DaB=R
zZE!~x2Sa0l0G{#yL}v64K*R#cT9o4|YZ{P5{}GVg#v+k`Nz!ux6*6DvT-4SJa3auI
z7AMnk#bc%H`ipRx;OK2Vfw0TQ-I%p<5YuchbxErfmqQjM?z-|>7hVDvx@38hlM5~9
zhKqJhsr=0Fex<5UVl&&bc7u2)z^F&oGy*1Y8OCpTJ0N#t|JOf!WI^Nq>nCoKYfSma
zMt~=)5&pFbuplPODRND8^tP>LuVN3}_uy~MK8+2;W|<ZU0$Kj)9P1AYVCh<?>fWf>
zN3WDipcfx`gjpvA4mbC(rKP{_kt?pp6cqIS-~Qh0*uK*OIk+pE;#*s%Z&UBNM}kHA
zIQ<=M)UAhaQB0-Pc1_}sW0Z%o#*<pF;ksui#>)4|ty|5@j}MeJ$baP4BW(FRha=`0
z<{O{J1e!9xUv$Cw79?v|-y$}NZ`DY8zD34T##En(FR@GCXP9UMKG6YKHYo;Czv37H
zj6et^#?%(#OZ@4v$IYJ~f5NVT!exXvGUWgxdxl0#e`l-Na_C{!Kl`>L4zq0me0A^Y
z;XgcLwOqxp$aoC(0E{3sA`Tez#8Xe(ptE6qx!?4olaGI?Ir^v<>pk{bK@-o)+{d&2
zK^ncRGd}YE_bqU-6b44<qXamQ3a-oA3IPu5*B{;W?=~2&YXfuAV1K_|+nsg#sSElO
z>Chf_a}mE09{%trKbsRGa;+5dEk<tt-Vf~BB<vE7e(W@tqkxe?&1J0b5zetr;F~^w
z5x@xi<9(d?vX^Qdey9RLYrN39#m#T5Blqs#Z{s3j)mutjH~4#+;`|)7ZL0x%0vLgx
z1cYa<WL*zWZ@Bhqvq9?vpNa)W-~;8qOqRrq3%xz`rvNyl?7bT6o_h9Ktp)yI_RGig
zxW?HjS-=hs7FKL?bEe;}=6S$@@Yq)=zNDpa(NknRx#c1!fOo|LBffH>>z~9t<QstQ
zec;z-|LAa}OCi2#=3T7F`BqrRGJdfBW<UABZyz>~KKe&<>@mk_->}fdQWzNN5twoD
zYc8<$#{I#n9}vMO_EQ8$KSAr8leJf8T=0qS_9?AT{C&}G?GJwOtNZ<AdFMkz2W0uG
zea5<luFH9D-(uunZo1jLOZ&d?eivE8$3A^aY)lmRcJXUoZRe(2zW8OkCRq$H!Z>w=
zERA1wti~?wW7wz6?Ta-3v5(oUeav5;d`feo_7+%n7ePEP@&o$RVUn%%urE?R+%E3e
zwcGBCSSvB^<6_LMdtA{;z=%u3k^aNHLEmGYh(ev4H^yi8V62}W`SYKxbq#@`C$(RQ
ziUZ@9n{=KTAG7f&kw=Zmg=|ksG7R;lam_0u!&$R&)0#Y2b4g`2@2wmTxZvw!T3>0q
zBmk|`_+x3MIpN0UsB#GiJ|RUoz5_Rp7V@MMr^xP?Pn}RHK%Ad{vFC-J!Ii&b&jBk_
z*tELeT}nKL%Z=l0I30z$VDu=}vF>YVU}WFkz2<-x<Mm8ir0D>FP>3$Q`1N+t9!`f=
z-Cg(J7rB68qIt(<3pE;C8A+NA2n+~j2(bzI|0H-B8Xby7OJ92Qi|u3|^(-t*nCN_B
z!}_ZSA2L5<Q>OclNcX>d<Zs;1#N#v1J#T#}lPHcqu7Ldt#CE4PwNbGkTpEI%5diyw
zbI!JlfpD8$It0N8N`86&gL!Rt_uY~F2(vFZ=Pbj9zYxI(78>8b^Ct@c1IqgQ58Mz5
zj=6HNAo|2-zmO+<IqoG#o6FvCkxR1R$O7kcU;e6j{RQWnla7CxUHB|Wqzkm|-|xN8
zKDYnA7-)~&bf~#ruCq!Z1bCmi^^11HSxP>awj5x@<t^rjzyl2`7O|X15q<cV+Cbpa
z?tRx@WrYlWx-c-p2H@{M@P0qZf^!6VfRH}DSuS6q++6r+!sl#IKKGTcg)bNVI}D5v
zQ1?}@JYzxrZ+z!_R_M3jhAueMe1x1IyZN)$swN5;`ENHW5VT@Jgvl4^u-L!kwHI1o
zCGNZQ^N);<*&v7r`v?@;Go6dH7@;iuG6U#^H&(7Z`a8kl@6q>;i#P|6x%<TcBM49N
zDU3j>5Wz1rfuGj+R<FdnnUE#Olon=LtT^<Iqz1ap_<>oVQ{M(a#AGI|O|z^X1dM19
zk&s^?y^aXr0YTa%fNJd^-U#XCB1@oS0IUcY(MES#0oq$rihChzkxBZ9@<OQrS%VA(
zKx=BS*II^j_Q-*gZWEuRO%^3sTu|1uG?u$${lM?>p@Xu5C=ZOS)w}2(T%(vB3UZ#E
zv_a!>6-EpXtTzS)IQ-X-{M~{v=`T+|Z9Xf@t|9RsYZ?F*1mFZj&t(Gus#Ebl#t!OT
ziAh+{5Q7E2V_5>V3t+;cla1-67hPapsrVZR41EHCL~GC|1Q6f5^DgtV`yNo8Do5|L
zSHah{YB${i3eG?Km5Sx@@;O1H%j@Qm@b`>x`+Il%#A>mVmbeuI^czuJU~hXf;SavL
zZ@%?Q=0op!*Mea53EAduefgW;R^axh=YDg7Q0SryUTu{?@@~ZZ*rQ<0Xp-WVG>Y6p
zy&Pa<QlKQD>l-e7tpw5?QN`P9C^s;+L<w~20T_vGc<2Vgu-Cu(JS%(;-J1K!TIyTh
z|DhFTt(BnOrzl_q!TS1jfq!~GtvhEO{K?OK!7#TtLO^)s<(ImjMI9;U<C+iY=jBis
z7-8Lqz%EQ~eechsF3o8#KheDI{PQf3dh=~x7WlcXY+wYfag-ij{<7m1C~dg<{3GK9
z3MmLESqCN&^Ct6tIChAihcy!b(VG;XCcM5V1{guGdCBW9v_e3?+^F-MEbdUoLioFX
z|6a3dRlc%FRGyV^%?S=_@vLr~tW;ggNls8KFyaJhrScFVJaZoVukgI>%0eOFCkjCR
z)brb|B9Xi9j$vQ~F#2!acfGqEc|>92i{JQG-o3ZI>5X#=Zh1EneiVv7M)39C{y{Bc
z^4B;N$@BWa2$t@E;V#&_tCn>QZid6_1y=_11y&-v_Utvc-1a4Noxr1Vfe{o}-+19`
zEN~I`K3uyv9?&}Ij-UP9F#jb{Mv7GDoOG((xSzFP4e0XW^7X-oA2tvC=C|g(S6?L$
zRKXg}-H!uCe8KBh%{y;ae+nz8xH9}BK={A?@QwxRCuafdr!902Fp`q%ECk(C5^lF@
zpF;5U>_pZ~ws%O_to=%pnj8pmz;~q;y44UrD?N7>^U&2=8z5JJqQ>fC+Es2E*3%6Q
zj9@K-Hb_0wfZT{|m<^&FZs*SZ!)DW(0ehL42$<=I=~*p4KL6FPhcDaTXeGGV00?oz
z0$$K~+(zKaB22zO-EAV{1_d;_{z?I@otA<t3jrU%4Y37|d&x^I*x>HD<E>x4-8}vL
zbMB{Ie4y2dO9eC$qKJ3|_)OitbC=cPA`Ke-o24yr?9oT%ZE#qmJSuI2M<0Je)*jDW
ztpqk(Xs?DB5@>@__Bm&qZZ~XL60pcaTi@4ir9Kt~SUj-7iz3Qq1ADIA9Qb_4l?fYG
z1lI^KUUrNuNkUq$`?Uc0^6lR?934^KOO8C!u)#n1gqK;(Id{(;qks_>os`cr^Dq3&
zZ1jGjcZ84uqA!HsqrIU&^r0^)sO?kF&c842ZQQWlTsaHCu}O}yP|8RM;P*3JjNJB(
z+arT+yZVxj)dofgaLR&LY|kUKWzYb|-PooL8>}h$-P+(2Y(Csa!oUbZ#$%*eyg|a3
zlaD{%_-5a(k|<z=?;)^oX|Z{nHIK_fA5Tl5>?R7X19*M)+uw0XN<M~x5d;%|b=|cK
z`cIrN4eJQ}|I?SSHo8Ng<TY~L5f)s~utKi1*2z)@S6<=8Gn%+)iy~-Rw{g8W>*SO3
zgr4-*;2ZsIq&pGnK25@31d(B3FJ<l&4-wWMqW6rj_(2Inu!uuD*{4Fl2wJFbeBFh4
zzzhAKwm<v)^M>a;^VC!FHXvxMfA!zKYuL!uBe6KJ79(Ecre*q$m4{FdtyXm8<Iy|7
zWgr$M+-sE!GysuSS&Z0g0xsI5iH5ZZ)*-E3GE)`^f|bYoKisd(g6Ap&ms-4A?lYcE
z$(>Vb)@_ysY}+C%7aZyHu<Zc6$qeliK$33_;L_F0u|xdm=u{t5d-aeO6yckv&!hxy
z{O9-IkQXz=wfhmxbKkt<4zuIA-S!d|CWEU5j0pTl3lISeanz>qE7b!Z1$3y7L!)w^
z)winGTyyCq<^;{xK0Pbp<InGXzznB{&1tgOI{m~G{M&Y(yy@1jm?w8WZ#Ah<^_Z4h
zqf-Qk0PKe69RwZs{aSOU-XDEGJT9P}f6JG?VgM6-Ehol(fEI`{eV2mvqW<vh@csIP
zW#SjV@lEUU!Oc&6)8R3jJgfm6XyeNG*7tv8wcXwCBe!j}>oLl3M4z;lA$kw28<=+y
z;QMtj|KolzNib0kFmmM9!v#dYDQ^s+T_|xN90u?@L_WO%b^rM1zu0vI+L=DZjpcrt
zV$K&6V<f5k*%81TCB6c>he12u`*w}l9I;rYZ?N{m%7FM#KDjlme~tG=;9j&>T>xMI
zK7A?-jBGvZFe&A{&L+b`!dX@M`B*d1Hyoj{`xz-euqG;p!oUcYNH@Im>ILg%>ta&?
z%+P_oc}f9%H@N=fn|ndRITTc0DFI(O79*@--~OgctYxkLEc<1>#(c>d7Oidsw&C?W
z{q0*n_@N!w{Jidbvn;Tf_gPE$1emh(g%`xXthEps=<Bm<d%igYjy@FvMwolAy8N;Q
zF1P^I9{&CB<W{iXt^<7+U5pL)-2Z_2=6AnqR<Di+()@Nb&x_;0go1*1ctOFVa`}P3
zqwbe_t<mKK{__?KjQDv<Ie*>xud-|Aa31DN_C?ICjJ3>pcinTZRi+Bh{b68)y%c*}
z_L0Oj_2*PqPZTf$5E^{%Et=O-dHZbgG3H_|F}HTq*)x)td{ZE59F52ttWTEZp>3Z{
z&U1u;5o7~cON0sU5C8VVABQhXK{9av$=ro<B_Qj&6qBkPV1#|iJ1%=up3)^_)059U
zYjAU%(>$}~&`t7uaYEjnobmHV0!NTT#F3j1!tV|+pZCmql=TGfn>b*s%dq_LDGnHM
zX>ijH$mT5bxQ>hfi^At6AHn*M*t)*_1W+9%735xvA?HeKG_81p0zx{n#HZ|)MY%>G
z0X*GBTAMAV4b>7DFQ!e+yvZDhlYu7ZYTjK=9H<5`(x2OIhMxJ0Wz*dcKRCy(mZ6#K
z%sD5XW-lCba&VlxTgC~K;**a*UPjK^W@*&~`64X}3K5nVACaIjycuLO@K3Vl2nzz3
z(9ps_n8-$viQ0iAeqCRa@QsaCm=MM?p+Zi;8#gEr!z9Z66&mXR0Fb|^uoq_kIBoK<
zAc6_^+PA+oQX3N?E5g<&Vx#tVH@tUFlQ1d{0-E8JjmD=>fN1a+Am`tH@MG)Viu;(S
zzU7jO7r0pfSi%$*U#4NgS_h35|9!C_`TBRhmlv>_N&gKOUSJm%ei`uXi?@GM)*lOn
zc)0hvUOqj2feU$~fDu<F^$;K}?Ejkkd@Cb*<{$jzXL84KzzPLkBexP^6JgTgQs^rZ
zxQE{t_n=qEr5S>qFx69xk#O6>zY!SW8Y|3Szy`h-cm?qu-uk9DMDmII?z`bUOmP9*
zqks{Ymv4yoz9|HxRy)QBuC!PXzxSFeElA+n!w~@{?z9m2`2-k%6&K`O_LT#StdjqD
zHY04Bd_n-p2FD4UaF?}Je7WJ;Yv$a*kse_(0?|MI=?{6rA@1Rq9ZRRM1qj~&pb{>J
zpw+7eB?P@l!0Sf)JNzdJ7(t`!`m5fNx7pL)Mc>*_^EM`Eo&m=B#Afr8pS#tXuKU;P
z;oJcu?FybgDeHuGxn7!7><KJE05}lJTjF6wQ_{o&7y)Q%(WV2i1Xo%BB{rUj(l|Fv
zxY>fZPI-ZKLzf+Mh%4f(MLG!FuDV-iHRPw|t_A8)tc>jbgBE0Q-~wf}DR}wtj>T=L
zYz0;RgOy9~x{VfCz?~McItW-#`=R$psCnXX$3(V*K<q!e_kMHt{l7M8JD&vs=&_pV
z?oXM_7@(A{iLcI;76oBt0-g7Y)68YBe~q25?FIcgp!hHUEKT`?QWRj^NBH}<@4D7(
z34IUuKKL71aDL11UW8qlPRW8VyX1`v)*XzWxBzt;^Q=)mAjOKXP?-(#cYl1RIVh{4
zRq7{UDmTW6E2rETBM90Fu#G_0CnxCe=SBoQD5&@X23&1?{;OY$RSGB-7=dqCEFysR
z34uJCt0<5_jB^N+SeryqBLE}&71!p!eE0^lUO~=%LP-AQZ+&~tg&IN{+*Z0^!hTxL
zqks{Y2CKg7uewse3XyW=DGC^wziZ3|=bmjrKK~l$m5eb$d0}7#pbLwQIL+mUfA`zG
zxep=#2d{sZ1?vC(-d|ZuU@Rs}1xD!8uRRYz>4`LJj?nD=+*iJy=Q<UOY1W_4H7|*o
z8h-Jsp#HhcvnQ-quX)?$)-9KR55X8#K#xh|nVE%s0^f^^h3Kc{JP8<KeTM=_cprcZ
zw(m%&IXtxA@(p(~SHJbLy!9aC=chC#?hyb~Z?qIzp|#{eC7+M?xSZq!#R4PDse`P&
zB3@ja8!{9R5#sxTYO-SiNxTi_8wi2@+q5rYo{XX}Fyg0)Qvi(uMqEDH12DWUG$)i3
zV<fN`DIOSM|HIrH2GWrY{L80r$@AVodsz>#-$x<kGhh0$b<bH2FalUizHq(Bc~~>K
zeGzp6_+N49oATZpa)sMuVT=p!Facl_Gb9X9y&xr(uW4W8<Q=3V-X-qAqbLa&VegF9
zs=wxA-GntO&*11)uRPN#s=D7Ck#l_V^S2i563ox*&cnb+i-NCXF)|^gt(>(O$+v!S
z`I^U-&4D->XrpkYl5=OHJhN-(&I2Q+Shrg)Pt>s*z{t`4qvp|v<Z(e%RbV75LEdxn
z3EK#?e#>SvIO|6}slC1yx1a9z?|$F>@h^}0m#p;!LR1Ltr=R@dZtH_LFOhy=N%3_F
zd&Bg8X@|R*0xl3+;9svxLAWIK5IGYjCTA80{(_0g>t8-~i%o|nJ855y7QO_4Vfw^p
zKX110-D5f>Oq*Y*wU}d$I?{p*;q-UQWk=u_IS{(hl7J*Uk^t;K$XyW|CIBJYjvEcg
zwS)85O7kV`mznE;7XT8tv7yaQ3m-u*!qG6%mY>~o&s<va&<2FLVJ$@fC<N=S9qx#L
z<^2*g(Pp2%cgLO9-#XV^T6gs0zqmZ`{6_JT#hQCJ?jA?L4m2>rgqFr_U%TBda9#OW
zHN5BQEAxOI@?l|uAe;?q6xA~@B2jE!u(zo8bB%>73s|36v;l&lE#ng!OSr4^`@XxM
z1dIT55`ZuK&g&T%L9-j-Pnghv27vU*nYp8>{jr-rV;Ab>03+v~eXjMv>zB=<|KkcQ
z3#jT7=FxAL_U8%59_wGW+NB@;*MBuVeS>+|vF1AW%ro+?@t*lc!y4Ue_Aq(JA}UN#
zz=->fE|(j-m%ilaye#w~`jRg|MM#4${DtP|zgpm1Ij^uK`eA_)DdS<?F_jZ|D<Rp8
zKq!Rz2=IBf$q}p+bguEmT4@Lj0rY6q{4u4Cd7E5hsbk1hR;OuK%!?d>(v@}w%M+!u
zF-aK9#2aVrfTaW4?+Ev&Z~>+Ih@;_RV<2to>`{Plxt~&#Z2f=&q|xZE!0Wr0h-0Ny
zJFxK(xfl!j8ZNcqXLnzh=~Ud1jq46EAHMGD1y)Xe4G7;K6tHmL@1(UY;ax^$bm_f}
zr*o!5Rwua(Zn_klld|>Rh~LtsV8;LDBOkIBl73k~(%Rw9U*2mkv7zVR#pj)4E)2ya
z_~|e2GvE2yT{c}`hx&sy-)EnBrgb6bmxrYk8(^PsQ5Ob$i9>*f{=r<v^tCPg?#cy5
zoDbP@ff1}Lap~m~W6@1t_@el%k79QAMYFk6AmDXZpb>hApWMbqz;$O7l?sgDW)Z8L
zF#TBWOzwZ^w|SSbqQLSm%384;V8o4cekIki7@-cvQpVIU5kKbM2Y;P+8E^!^Eew2<
zC<z$he$-r531GzaJ?bczqJR<RC%z#cfTMhbOkeuiSMz`o6rNCgLaFLAnumy;Qw}hK
zwZ;3cxms()c}q0v<lE*O4-?i?0CPUEo&+RyYjk(-h%JX&tCsM(%2}Vfeup*Lpj<kl
z;1Cv!QdSZ$a?<g~SvNGUPL5bued05px578(xZZvVNi~mPjfKUvPfuxI^+j1wyIqKX
zuhQr4d6UW|4V2S7+kq0oY+`>eP)s?q*<ufsC?mG;Mc%;gdQa=ZxgxIEMv@b3IpT-~
z!s_sI6$^~eQ(XTDmk}3_4v!T87e28r`RHdqXNC{tD+eC2`4DsMI|BKGpPq9c7@-}z
z6kK}K+F)%SrzBA;V1#eJTduU`zn=NlGq#2;Mwsu<);{8Vt#f?hxo|U##iAqfvsVPD
zU`>YH14;r$o_KDT*|dJ4YcD($yyo0<T+#N(zW+vzT~2Y2i+m5gZJkze6>kMTJuBtQ
zFUZ@7zYpPl6u?g!7qiMQ14T>b$|yI)cWKRtH-#|K@7^w;COn=3u%Sff6Z>Q0>xR8p
z{NU$5H^0<;@5VWQj3a-cG0FiW>=`FjhYP+wDPUwqpi?a@M*Qb2_Po$@6<bcR>0Esr
zA^40OB`4)KVo&<_S?_8$rbw27+{}L`emZyEU87w8-2LY0(sCX}E?>~PK+fo=bNBsg
z+Krlw_L<%6F?FN(I`rHEf7`S#b8(g_A97{5`Zxx6v_-YG1kdQFwHX)j$JOBKE~Y5j
z6gTdZ-<9dgN-CRvFflVV7seqjEoE}g9bMV}(ZA=P`)N2YB*SyhmsCc$%tF5@D=nGC
z=%_4|2m4G~b3`p%Y00Udskp^RBQ1CarB%FseZHxG)XO4>_JyxnsEyAi<SRe;e%^xc
zytB@<zVj)EK#vbT{5ydcS1&;K0T2H9rkm%qH%X7%9E5aC<OnwZ<x@9v*)WL^Fg^^7
za1Y?%%+pS_+SQm@+SnDc7=c_TjT4&RCLrS1k3MErNkBIL(YDyh_5<&Fr&%R)RiDu4
z2Jj%TZ$RV<h9R`N-y>~5$N>@dhyJ*B@cSGsW?mH%igKNHqsf6Ub@wrI9qkDO3M|}w
z!qvx3x85ef9Hj?&a8YpeTi;?gRDN3A2K?bqf41p@;VDQDpRbVnlB2e5^OG9_#o_-M
zL9i>^9Wmd<U4&1CfDwei2!q2!P)Yo^viNc3lZHiLl!XZ!*%uVQ2MP~VaqqZR79$AL
z(1KxsquEQrqvtp6lj<-0#ueD_A~x<k4+84He$Tt|%#*oJ5N-m9`<ggh=iR;kx@#gg
z<w?K@3%(Ctf87FCF!c<K@Ev~e-S3htdj2bg2l%Z=GuaU?MX+uIM8M@8e1M{aO<YfM
zrS<n8`s)RL-=h}L_)KQ8d(v^o&$)G@Y)7B^+~>^hee>F#2wVXE!rD#<Xg;a_jljl#
zO5QVWJ+>Zxm@9Jb7zd0D^!1w$=^ckHkM8{WFRc0bTmiEuXmy1*3s|lEtAxYh%}Y3M
zt$(wD9GO%M5DCAr^gsx1eJF`*Gn!M{y1UHO*tl&L;0SKB07PtT5gXjCMZbff;%(Ag
zoC7wLqXj@fV-cVSL<|xKy6{qh{?}qA#|3^!C`VE(QQDM__#%KaZt%qt0w4vNl52s?
z$bNfX?|-_~ciOvpY}}8Y?ha`$k4s3Gl9t@0nNUFa%ir)ibHUkXy2p%toJl?VBac01
zesI^_0tok*Nm<pvlXd|tfGlIOAc3bH0!Chb%(3R0%Pz?)i+=a<TW>Yb$d%T((he$)
zOj>O?>7_3**Ijn$Tw3BTd_qFjy$AM7;5uV^ghSCMdi5OVobd|lG7=zz@9rKc6QHT?
z6D}db^PjUu+beem1UU>-xxk2<<H`X>02a_r4ZH9n1`%;lC>w?{b~u+W1QRE2nNL{Q
zed-HeGzTNBO-lhrG%QJZ<dOxJ5v=Pzdh=)We$#%n_B;ys=SH=G5#A??I7vi%aeo*`
zZ@uIV)*YpPpE(&N94sfpMC`Ov1q6ht5HJ!o7FPlop&h=9+;F>!`5gs}po{}Bf+ns{
zXmS5P|MQ<UjtGPnF^^sKmP^ex#UOjLGXMZU07*naR5skBfZ{hv^OkSC9AJbPY^TdY
z*e4Wbh@XPBW0(M{PzGUr;1iZsxZUyTb?2YAz!HUavL0qF=DuN<C$9Z*zzE|I>)Wsu
zC`x3wi-jC8!ujb6{*HBwPbg8Lxr~sr+7y1?sQQAj$C!hf=T>?C{(XT(G37*+?O&@z
z8Gd<7dETgUg1uT>t?LcGX`g>_zzFNe9lPh1LZaT`<4-+h9(nXHCJGoqF#Qn;sKejx
zGuyYDFMQ?eRsin<QlpAT`d>cskp*5BXsZv5_-*GL*#mq2uv?~j21ZuNz4c#T_f9*u
z`p<(UAF+=d@oazd-s|mt!0j0rkCT8AXI1OU;K*1X<vNjl{B2+Vrk`}~oEWfJclgA-
ziDln*&uWCS-~Y~Q@+|wvL;v}-TrL6}`$YS(p53x(lYiNs#{na&<vt#Tg)rz%{I#%q
zFT3xWb^GX7utUmaH{bTf+0~X?am}ZQ>fiHsjomADb@$j9BQ28cv}*6@Txm_p0=+jv
z5y#J8?Q`1CO413{U44a~%cZN-(bX5UCg3wNpJA<EvEugcPe{JfrvA{TR=ex29qzX~
zay{<c-Q(!bUHlG{D+69pVmQ6K@8{!MM!-(jtRi+ekDoRubzT<O_`t|Q9{JA;Z+uGP
z4gG+3-Z?8{i%aX?fjj!;yL11Zf9~$PW70j+%(p)IIbC_h9H}!q*I!!fy0A9Dh?GJr
z?Me&(z0Yf+XaquVw)I6XD%e!a(6JN<6B|On4PSV6gC;k${<-$iPv2qzj;IYfWrZgs
zGz-{(Lce=NzA}IGV|yN0jK~MP1YB_haL}ESP)w7qb4BCQ;>ruR5@Gs=EMES9_TB?p
zvZ~tmUMJ_ygoXxbvV;qWAVH!e5hbV~!c&5x0^Z=w?~QT4@xC#>@!s{`>o8$Jl7ay+
zT*L?nNRTX|f&_uahHjwa$@!e`|65h-^g657u3UAhPuGM#RkiopYp<~P3iChzbFwLq
z2}+eP=reU<0kCTI>WoWJKu}Kvw^hEyhMzbIBybjKep>e(YmYrw?&f~lPd0*Ik@dR2
zDaQvon^#}hM6s!JfP#OH@3Hcr+_**oW7luYROTfCBVq7$|30yMaGp3YEYw0d^+at>
z1It*_+<40m<f}Cv7<#UOk+2T2-%ki6#BEqiaHEBSAzZXQ@%ulRpKIY3llPo_k{BlY
zXPlaEzz7wLfn_j`zVP*HX24b7s@H#33I#@_PB;JhnZL_u*GJ^j_D+Edz7T+KTuPt?
zqSODc%P;p&+3)v1^pIjCY_VQKvH2;DrR|;qfrKx-^pg4Vw{O;_v(v^l=#*8Alce{@
zZE{>X_xV@yyZgRVPj)|NK5x8L@6YPY_b09powmS{b5k&rNg2QR!L9Sn)&U~`8Wyw=
zIMSql6C<*?(wNw&IlzJ@%Fw3K0&3cVBLw^&v!HfUbIOg@EZ18~572@uF2;G7A*})i
zMu)`?(Y;_cOpb|FqH~B1+&(s8TEr5uE~j*STE82_ESVN4V1bcw7#B+0G;Idfy=kwj
z^*@Y1onnSG3A|{A=>agZs8fEUQybG})$iS6V4isN5gB7utZuxck82G3!O!m!fFzJe
z-~}=)D;NN!?#@>8-V;wUN4#w|69VSRe_eOI+)EDHK3Kf4LvcpN%s#vCYA$;JIR=*z
z&#hXefSwPUO+)=Q!2JRRx^K`zlHluXJ`w6eyek0VA#Fna0M?wx6(=q^L;26Ie9gQ$
z&~Fx$5g2jzC<YkWT~@m&mO}~GQ2u%1rNQ<MCCRcrJ`NbI1->m~iIhQt%Zql!Wh$<U
zd_n0A+a@Ll2-G_3w3M4e|9mmPh=0BNp2Qe&=d*pHWd6$a-;5?fN#fkY`U!|*!5S<k
zeylwI{9FPfemRQy9s@?mhjlXRWnb)C`IcMFs@K-q!>rG*xI}KWG{*XXL=v#@!`pvi
zo_Qe?%;LF=85Xcs#b_xeTyydHxB{JR1cm_+AfYcY1{q_~SM_X!ZacYd*A+Y4cZ_Gm
zmn2|>XY|ujQikaPZSQ&TH#Q3&7};~>Zr0k{&ztY;jo2gkvQci`mT3OlTBVp^#?*Cd
zbki~4^aM%3NIuWv!;k+q5tN4U$h#X#(lxw07k=n`#TMN=yriaq5nO3?izOBl#)^<J
zg8OX@b#RvpYtI+Xees&>XX4!A?(d9KPO@v)8x&&-7ni9#3yU#=3oC$yP#%2radXcD
z4~B>BZ`=l7DNs6;`yYP9#yN4vai9J1k6e(!(0E!PPf~1_W2EmmF3@+7Uy>^=3>7d~
z@P&EmhVQCP6b~(w(@%PrjjQS(N1uX0UFbsIPm}FCMszK5EjlnbAm)0zs5OHY7{Q<t
zS6T~NW`p<H@+$RQJ=RF6Em*zl3vs3845O-7y4-i#8t3TTwW@9k=C*18Bh+e}21b}1
z_t+~_DG=-KPs-&|j53bWeQnwz1VsLdOE0pQ<F@7heC-;ueNUHFOk8>C#dae1^Cq4J
z@j#ql8HI%u;=9So2N-dApmqUi^AYbPU?k=W%;n?zLw3jb{KGSUGPnKW4q3Ns3D(EP
z3swgJ5{kqx$%-(vaBz7fDLGM?qc{(R&BA+*J3i_|8t?)uHiW)(7I}UiERwF$rj8)l
zF^RbuAf4a-{&g{61Yhu<yZ(C9=7==@1^S*;XRKPVZV4skYKSyg)L~f{leq>)P?Es9
zV1d(|u(%;;x9`W)uZOSH_mAJ9xHgae_DTQH%=bgZ<a&36$u%E=5rUNyfH9Px+<BMz
z-5;K|hn<oWun<dL$FeCD7+I;nfft;2Zg?v(Q0@^EW%fIYON~i^;oWjc_>Wh9$|_*}
z3s=js4eKZ;Sk_{770Sb+XWw<-1L0x&JLw%w0!9cdc<4a~X5@9%bvGE8O`*g<VA6a;
z43;<246X2*x%$oGpm#W>T(@P=*b8tmB^Oz^!fH|=!zo!6F&@GeL7JtM8<&#1skL2c
zCdK4P#~sl*fDp^{(0AN!we&2|bpioojnXJ^0FVXOQq5B0kBJpRT#+fcfST4l0AWBr
zEC3@$gk_vi9*K>o!L>FHiRiHDaoV1&<v(h-&PB5RRugw9u>5e}u;~;ifg7t4#aVFj
zj>8T%?~&zpY)#<O_@A$>eZyRH>km!uK%aGi0}BN3g?{?y%Py7DayAQbofMs4{N}gJ
zhBs5sE8tAK(od*dBLY>%)ZS`($+R|_3TCeULAvxY?Exe9p#CH+%-xvb(z#FK8?77c
zFRLppSB7GMkz<ZHTngyfZZjw=af}Fh8WWTaunMup^u@(c@sIJn5MTt$+P}a26JZvK
zE>cNI%ml1FL#Z?{0%L^OP}u~=2yS0vf}6pAyy8<Cr$WY}p8!SxN~=(|R$zo@_uWSw
zt$0*N*(&*KH(c5>r|~ZP^82TsmID7Sd&sT9nUk-VBFgO<3I#@BQ2*^GK5hXP|6YH7
zd6f;0>mT#K6O(P<J@@nvS+}gdaS`vy-)JlfU7%4`%Iww?ekR1%q3)r}l_X%qH~#YT
z@c~XB7{MYMpZKAKiba^Ktqrr^^YEJRH;S=MYd4w|D|QG^<^3D$O?o}aYBTSQwt|F7
zF))fL`ql)D@LuQA6C?p60G4<|aDqpK#$cS_-Hq)7xyC8=-SzGhmLzP+P<VIVbK2?h
zS~1%|kaV@o81aD-w@z?%qy1yR2<LGTj;rucex!IvPb+2-*TO;sU?oNe;~ii?D$l}z
zktEldKfCK!Vq9eQ1!b~~pHE#jE2am>F)qDH>p>^G%gxt^&&f3WxbLAyGFZOu8t0nP
z{dS*8z{r_$r+A>;D*D2h{-tZKpXCP3ollCN1ptrtkYdCKM&!TPG^H6M#26W`GB8p|
z`NL<jkZbE{OK$-OjBM7Pa+!cN=6kk3Z3WKx*-Fx{^*~RV7)#;v68=sqN51Kjt}o?y
ztpkNyH!vc__^TT9w+3M&uyT(*@+xY~|05SdMaKdR@s-oR{>^M3^)cln6Ew=Y%~Gr*
zF32*+VWK!Tgc}sHF-8C<2&jeI9jvn1K;cq{iFwyub~bN~06+jE4?pvC)GCzn#1t4<
zjlqN==%y2Y6AUGjO-_;|fdp<L6T8lDSN}YIhF?kp1=JsxWiCF8O9xQ%g{!Zz!JyqS
z|1;Zygv}Ry;Y$n{!If4qj1hvc;a)0~m>*NpeD{a92^d)wCbz$n+NfHtw4AcZ-_%A6
zns6}5T^#{G3CbChq=n&p1V*q-#%h=6+?SW76vPKD?>;_Z`3zY$v9X6dl3ZyWvj1Dn
zdrm#Yzm5I=BaIhNKmS6;amFVs)NxUUGSZhJx%T+n*S;}x2zPgsCzP14dY6VkzUQ7f
z+tm=~l7Nv5p@8qYi@SX06TkmhJcY7cN<7LHiVI#J9-qtKykC72dIr+N(zF!hdMGU5
z0Wi|Oc%d1V)q0~Acnx9+Obak+5Kw7>6R`;zTg32C><EIeH@5?#$ns9}AfO1qN&BL1
zDc(m+i(;5eNWqRDd*YCc^vbepL8loNH~_eURvuSf)}5C&SGFAb-M+9(m`Z~g5ul*m
zX9~1Y-Xs0RZKSH*)__C{mhWtX%L9xwH#G?4pSEtn0E8yh-qgE$L5Jp7xxrFgop&k*
z(DqV}vf0iiudiEguD$tYvsp@L1QyCdx#{}Mr5`hk6nNT~SEWq(ynKk$&9GLMF6%ZE
z0%yimhN0de)1kf?*Y(6LandGmq*Z+ZgU2yOIENo<f`+m-K<UYM)&=}KNY<30TRW}+
z7>T));9Rk;w73?Dtt(P%o@=z`NwrNX<(9u#e8vJj<7R?b4KZQN1Q-e3gS_&p+>pq}
zcPNDbBT3ey+15YXEn8~cWrX*wG%(Vu^$Er*#foq6<LB|z9~HCajywHBGvAZiXKMvU
zs8|wXggH6J$YAXW3uja=md-!tOatr5xubRK%0hvW#d?RYkdaj=7@Yk07b%8$c#hx1
zHetQwO9;}iXZULYBYt3B(zC{9tsN8IF|4Rzr1_Erj3g;Calh+aV~uNmYSCPT;0EU)
zF+f9k{C7{8U)+7~tiy%Ik*Re>z^uy6vmYb#Oiz#mjQB2X>B3Pd^a@etB^d<Va@*}0
z@%YG}_2Ee;zAM9pB<E@w7|FJ8%m$1sUbN79^9a{o&%OAP4Oq^4hkX;?K^_+49e27&
zF~G<>Wt?$*#5jZ4h>Tw`LC3`S2y?`ja8727d+(_y`$_EgB!f{-;zs=}#h(c!2^axj
z4qb0C4`FzS;ZG>I^WJ6Wox;P`1BV0lGfw6>;P#5cat}YQy{ZF7CdbFksQRdyjFB8g
zt)H%;7RW~jy4H)CR6|9t7o|)sz#OslGDfzd1y5T?yV+%zU1xP$(&Bp8op&}LQ4ruz
z{vwO$Z>24w{UGg7)S;?7NwWNK3PF5`{(~$K*<=wv10T9h28Rbsn-p3;F!I`KZ<?KV
z-rfQrC^z1^*PhmeSLln><q`V?7<uB^KP3jdS}En?1yVAG@{>DcrLF+8PEJx_=p&>Y
zVv@y$4{eepNx;bW6f5D+FTEVSEUE3je$y>x-J5IeH7r1IceXH4Faa(Rv^+_Y`t`9V
ze`jDTxDBkkzMQ}aY#~@fq0AL9vR>~lXnZzFTEJxkM)E<yldSf$Ejp7dc!~i=URwRC
z`HoyIIq?}$ETS*@Fs>+OTNd(cefgT}4eoNvCH@!^Dkm@epcMYQ>@w?CC^uk-uz<z-
z8wmloLV=N^4n5eMc-(OS6+!C0v+l!>TmJht^ZZMzW*x_3a!|1dy2Ka(m|QNGXueRs
zfBmn|DHw92-4L#L%P!Uxke`O{zqniRn0_~#jpEWIfhj&P;+`=+KXcJXGwAg$uR=b*
z{=J*cx;3vCa!JM0vN*7ek+c%NRqm6<hvh~Ka6(FR7#hU?0JuP=1a9yfaAUN$-yXBB
zwdDJLVq{RjhvJV2Bmop@>F%&`G%$g+ZneZ3Az#NRf#EQ&_#%x01YnXl!EF|?o@_99
z9kg*rbU!{P2sYkgOmBp$@2;<<&lWKOI+rf9fDo=4Xy#VUEu(S|HY9+8wrpu@5xCG~
zi;eYXn>KVSrSv@d&_m726+34z6#V+#D<g`ZDL6Mm51~-}*H2z%x`J`{6$Qlnij@9M
zDgy!_g0$-X1GqPnERY+np)t8Hqrdcz7-MWO8#Er_u3+aS%j62-K(oin-2_-J%vvs+
zq4$Te=MAx2%IQkWy`#kdBe574%+0v1S}#A`$Yz0j_+Lg~nl9<=vA{?emjt(1-?;JH
z87!JYfRP;(XYAt_e%Rj2{}XdJ=*};cOS4Z)NfS!7fssNL!byNmOgZi9hQj=BKk+g5
zbLMkC0wYw;2Tf)wt1lS*;Jy(iIY8uZ9(~;0?YUhk6d2h-j7waug!05wayj~|WP_xW
zk2}V^<IuN-mzkSnoCA{=@s)^v!Z2o_e-4oK%1bZG03VZpk(d>-pI^S;-Fh=jUq}$}
z-tkehQ2VM{5t#m4JM$zd21nS&TN5zSuDNtYu<j1)lP@q*sGFB0gUoP^@5;ma@R*~I
ziMr0NHDKh8-T||Ng6bkLMkd=k^*+u9SUHys7~$MU&r_@oDe-;5)i!%GVg~{^uvS1%
zL!3mU7+~ZijiE<G#?Wk@99WYvUWd$Lz;fJ`I{m`ohbWd(#3R6ue|cwOTu$l+8zz(_
zU<9uo%a$$)lNbCOcVu{nC`M3lvo;;lo^cumyFNR$cZ4xAs-WDg-<*qK)^NkxzYM-1
zR^KkO`?rvLyLD&mnr!Dt@6IRb1!H;o$MC2?KzY_&lX?bY>uHxB&=f`)GOm4BEG^|}
z(KEGq)1X<hq<a>~>dv_U_WWjl(ycw7s|R^oFAwSajEzs3E*V!ju(r^8#^NGzk0kNS
zk(7qxy?P!H#sga~z0WQBKB#iH!JOBA)3NUc`ZbR)XcwD)+A==E^=;^)61E93ZwKU2
zVPNB+3<sOc^2I6OGQ8FnBK^meo%T!WA3r_cUH`k!@O=2&-6K36{w}ni!t0Cq%{|$U
z`Z29sK6-L(j2rj6T%YdwVr?|gx7n<1n63XR^*(#|447ws`<Tt<fu|my$*%0Kw4m+3
z^76|wH>l+=i{vwMS)j;EDldQq8@-Tbjs-GG1Qhd}=Mx6Vm#@8kR#qtr{Ku^2f`0;k
zm)C#yW^1wREUB?<z9X$XA#D;`%3|Q0_q;nJ&>E91N?YP;^!D}3QgorL)n)@g<eSZv
z7U!w+Wm0yAlc{fUNg9H-VYv}n!;wcQNx%peoa7sm1NYn4oO#-*8K-=A23W;jDc{B2
zic#Q8|G<D{-}uM;?@2yf*?7TxL9y#gHeh7q$WWAumsIB%Fw(Ed?BpYkFee>*jDK7E
zeLgcrphII#7)mz4Y@a>%PypOB!b@z>aV)iBf+BKj3%>5^amaxO7y|f)g5~6{JMUz0
zM>Z&zabJWX<vq8B0wc#BahN&wsH4K$nD5DI^v@AjZq5qmV!2M)S?r5Y{_hvQJOdsj
z^;tGL`c6CUXo0CPUmqCp&jFC)*3TEl#J>Ik9ZD@c{c}m*QEC;t0Z7WT)Z-2qX%QeY
zF(P+LS|9>qOlhNq5PW@H^AGN`I7dKsm>zr|kpdaU2*3#K!RBz1#g5!*SvOl8(>R)P
zf2EevGX!&FY;Z)^$%U4{fesd+T1<}*$*Q-r-F`L+jG0u7lh&RNJ7-vzQ_5Tar^+DU
zWq9-YQmbB->s!0JO{)~@)U{jgs{lv(w0Le;-BPZ!rggl_bO;<8QQNZdAZ`eYbmE5W
zv)4*<hBmk>#XzEcLm3qK{O@1+YAU@peVsDS_+Ov;gf(#R)4n8S<+Zomq78uDq$zK0
z<?U(8!<;m@X(+|k5n$3G_DUzL9+kOE$te3#nnO;M+n>--0uZ-PYXV%v4Qh=*FzBTU
zVj4;%fDw0(Vt|oYoH3r^Y;GOe1#*2-j1iRdz7T`$YXVfb-j_mvk)(A)(wM}3I@H&H
zuf7hY+Q3M7|6IR`O+j3XP_o5^<GK4MxmFBiz5*jKC2_aQde|50i_0<QZQfhfe60DM
z6bg*&B;$-r6pJmCY#<eBlK_dEWspSq5SU;DDP|Aj%QMfvVD}0f3pehpyJ4Ax`=TUZ
z1RwYKx(@}W*rb?+h*NlkY20tDn~7nV<oeB}tAEyR3(Pk?LB7C9p*?cyiAnJ$<NIaS
zhey2Qi1>b4u|I$Rr$5ayuyo~_>)`9YD=k+h_n87Q&Na-rtY2Wd<09KhlD7m5IN5Ii
zAfN}iN-^6UGolz^B*wa<4B6ayG9REbAnvR$<1#GzkE^r})SQ4Eb(ma&9dopw#D4$p
zpa0VQ@lVfY9Opoi!B`9!p}xdA!8pU0)V?TXFy$Z5_Wf2pN3Xx}X0{}W=iFWq_wWn4
zdMv}q1z)c-MiQH?A}3Hb_e)<JW9h8j*k_h6nay&F6{1oK@$g<rI?ym(p}%PZWNSA_
zD74;NV|wPct(GxDgTMUZi&2*)wbmIunyki$Cd`g2wlkZg{8`l1Zj~K9k$|I={mgft
zYOqod<vTz4Z@cL|J56g0>9a0<vO?fmgCBi2G5VVou8qYQA)xBV|Hp?jEVXcdar>`+
zW&ZraOEZ(M^Kb0R;XZRQMo1TPafH(KYd3t$`s?=<1o!?%Zp9vc)MX$2Oacjf%eC;r
zgANeQ^pums!{+;<us-|e7wvIezI+mZMj((cgL37=LMllJ9{G`TGnqha`hO+bZAvM%
z<?3v}2#fugB-J?vjBtIW86zQr9xHRwv0@k_NqPCllPKdc#)#XT@3;3}=A1N3!q2N1
z#s~^<EMa28qOukkBPbv60UjzY@L{}6%1x9}Pb%KXPw%{IRy7I*M&5DAUzw9nIAK;E
z{J8bDADic1Oes;^N&HSvYs{I{z5n5oi?sn*=1$p9faCuoV8qFuE91db0e4XD#3ac#
zf07b1_Kr?Uf&AGoT{R0ZiruT$kFhwwl@=Q=+&ZO#oXg*;ltr3!g^Vb8y4)RtN;E1q
zz%<MbeMhEbg+ojc`OQlGc8VzdZfcUH9@ogK*#aUxot6R8BmiJyR6vTxHb4_hMhLF1
z$;C8E(F+rW&D}&lih8+qa>{0<wM-RRBaIJ@DptqFlnX9<=i2?#zF?8sKvu-6Q){<?
zAF**71w7HdZC$MbM~LMzX_juczyc9$v;iw7C#TJ*>S^0UfYrNBJX%53Pp~YGu(qH5
z`qi=`f5Ws3-0D!AmMCLHOv3-V<^~(*Wqe4RmqjTiAItaw0WecyinPfM7w*0I$MY~Q
zaH^-L)f~L{KIXi$&$7q-d&6M;>794W=lK@JY_=F9<#wgz=AGnqP_iqnvt<Ah3n&XP
zf*<#oFka$HD`bq|dJ7j^p%em)>>z_0)?cAy0}C3&tomP<XMzQ_4vZu-Mt*OZeqkH%
zJ1GWB3>b0G1=nE#iz3Sx0=>cv>K_>N56^s0vZ}uJ+uv0zkCXx@n<d~(MmfG5DZ`C-
z$$$g#jdJ+S`^P#m!*W?;G|wmQW{Ge0Rjr$yGK}jA1x7GZ!tys1eB5JPkR;J(TZwNH
zOBO~8!u#Uf&e#L#KiL=~xa<A26mIzfBkwxqo#x2H4)wFQ-`PXG^vWytJ9AuDLLg@P
zaYDvU3m47?bFsu9(YU;M&3dzZmmSJxz3A4O83iu#fP2=frBHIrXYPx%H)xJ7Qi_S=
zlI>QjUV6z)5YMc}WqW~<3xa)RSmb<xkwRT*CA}}W&&7>yD7-sI9eGrAzg%s`NUvPt
zVXdBzIA8=n|A_PT4-0g0TYEIdh#S+<g}}&zo#6`rG=KzSRA_8k3^0-eE`RSww;AHZ
z#Dp?k_K^!Sj8B-;{`0D@&0LZcL-O`t-f5nC`k9$@?#r6jHz<zGA&RSdwmUTQ83RT*
z8H<w(v*!Q#{1+6PF28=7xk-^<95Ax6uh+D9cWEr|*B++BG-(Yxrhe*yUS0+%ggID-
z+tstEa0|qApcTHQ&A#08e<@`ylx<kby=w8=tsWRzD_2TbBi6d?BKH?i!Sj_QMJ%x>
z7RrYt;0E&T?`zTWm&`>=lC?CG9}2^_O3{lV@pE6j+7QdbNw3iLE-emlrR84K*hYu+
zzm$UUu|UC?4H(H5V}$&$B)L)w&sG~mmm?{;=Yijt2OoXRoUGv2fG2<@|8?!PW=smQ
zY?1^LgkX@%g9ZIDZ#z_hr;m4sW<H^?KKbVtEaec(-AgY#U;ff(yYj)$Gp>e`7T~ly
z)~KNnBZk;9p;c*2=~`@zk+}j!!nzjnyHdbNlGV>`vY7cpU@^01-Db1Xc8jd<>6l;E
z`3j8CPch4BzyCs7Hyg1_&pkI9LnA51$btLrWzIbPbU$1B{WdWu{_yN`vyStAPUs!$
zmX$r5lMt8;UD+jBIo*B#gEnT!tjyCtV7tXO07<||Qj8JCRmKB1hQ{ty>&F~0(j*W9
zW<Z<VDUA$DSr0IRX#PNPPvCVMSe<m*T(&P=Y(_S~7!d#>AjHKK;gTtBjwUAs0EqJ4
zq<Mzxr<De2wDY4B|DwD%6PTUYQUEO)?`^R4^mm(>KL8^FMOr$=ToF*vqVi2B7RE^L
zM!9*~GN;X?udy#(OScmMqTqgV0X8M2CT_!8JH(oi`>Lf&7nx%ZKh&()VMp`eW4|>|
zz3_tCOkk4!4J+;kqz>GUHMhxkvcQz%4?E1f@8lCQYWVM8zt+66YPC(<B0%wff96VA
z@}*o#xWsFvy#N0`|5XKm?=YhRT3P{^#4tetJ0`HEQ=rID|A_JsPz5le`|+={wZZ(&
zMHibLW##V+>xwUa>qaq6di}%J{T}O%Q2s-19@c5o8l&^iMU0UoYt8Gv{XMhhjW@#L
z+uuFfcz;4lWZ1gCtdb%ag>WdbH5HHZH8*`*Zc%45o|0l{5D=HZxG{lMd&MOeWt^&Y
zU?eHVNVd2ZD63GihY|xu{4|vUM*Mpe``ve!6#^sVnY4y+_j~wp`MAIDe)rRU77C0c
zwJTP|SQjS=j1?HCz5vj`R)6lbSM7VA6k~+F411cG0Dv*?hwoStV<aiYh#zAx8!&=P
ziV*=!i<iu{z;<Pu3l|ga-t+y_DWc0&&(#9Le9QlG<y-E2^3fAi3K+>{Fcz+5d3TOK
z<~VceiN|L&L$!es)*ShY14d%$Qy@k%V}$d#{Eo#5BtQ1>H)_8Bi|3J|7+?fI7oaMX
zAIK#+b9hWJV!J}I7a6yG3B8IW?W?oxWs~$tNx%pJ>_gz5IRJLBuW$0pS?=!+7(vn9
zEFgZozu$C<(aZg(<wmPqz(~2vUC(cBTfk?Ggk#>^maKmFYPG;t4~!6Scg?FQE2L^Q
zTA}wr!La+DdlqyVi~qk<faB0&8*2f4ND|x>E6dM+?Hf_cWUl+$%RXk=CH{>GF3p5R
zfG%QApfLBv9~9hiEFrg3z*k>b0HFK}Ej<rA_@MYgNfSCW`CNkXnTgw%H>9|^PMazh
z*8{6F79Y7J2^b-WC&4^J!D8fWS>CcK4JA~H#H?SLaKC!}w`Ky_hj}C|NMajbuB`(j
z@gQ-|=X17AE$xRZD}rgWFvZoI3!IB`&c(<eof8*>$NlDWRLb}^S-vIFkFzZ%tK~|o
z*vUIvj1lS%`1-!nPT^~Yp!EO3)z?~Z)jwG%V`K*@fIt4xkNR2I?|1$BJ_Rs*a@KLI
z2pOLi2pIXNPhW1Wv-||BwE4YB)}T&clw9_a3*66{PXYiE(<dfLv1F2fk<(5*-n?C7
zUMSy^yFPz&6rQie-{Et_`(<N{xWZ~4v!YEXF+#KrG-C>sK81f-F&|-#U}4)N7g0hx
zQc2ZcE~W@TwgW<@q;x`HkH8K=$xlG{mIXc9z{nRX;5cj%#N9UyjQFugIM>qAYJrjV
z1)VmS`=rV<+P5inZH@kAOzPTh$C<b+HUPbt8Gug9q}2ZVPhM&kNZF43?C$%_Pwu!o
z#flLV0B~eXKoFo2`5*-R{Otey`;1`5|Ms6>G#ljZ3^!-EA^Vq4eJsNT9l<$2cijzU
z-P#Q{ukIyXW_WPSv<qk&lz(AB7|ZU_obTM7wTtDmSb)^OUU@~-W#>1=hzl8z=brgq
zJ09Dccz@U|uhZr=X7M=}zzFX%?{z3i1{I_Oytqv3q|lP|iKl+AcntT1S(!qB5fs9o
z`P5}IeeE*CEfHmk34u=ufEtPqj5wdVVVZn?FA%u)*B`$$BWW1uH6ua#0MG3{d+ctR
zhs-6;3YIzZAFlXhMqRT7<xX<HHJ^bIjH51BpxtnPK=Y!2VtpLOl*<<waq|Gy&j3z5
zf4;o@m(@15n19Uw9(&JO8(($p4c1tIw3Hi0YRLXzjdtC4zBh~M;_?PGV?L)`z9a!7
zto<lUC>SB#``|<NaJDt1_3sT}yI?kBB+Re=yZkNS$8ah~?(<1cP%1EjZh?0LdHu~d
zwKq!5`$_v#;t3H~C=}kE(@%Z30Z<8Lz5*lE^H90x!+0VlxX{J5xzE&NuDj?%AC!Bp
z<+CmU%poRes6!!9QcRDSF{;ayaglWb@uhruQ7(xwVsnyY{Pn;ik7-=|jY~G`Gs)mJ
z2^fj#anKn7j^lNp8sY;Zvi4DokwLj|?3R1pF$H69tpg*~=!C8C9s@GWYuEGQ8;6J0
zaBHOSc)5R9lDSc>I+VLjs*xXaTQyv1-LY<!DJaGW@eh*RO!<~NQ<_{}k~^(hmL&oe
z_?Jy$0dufSASk&nY%th7B2NqC!1^gkPCw~H8xYk$&m`<NI`=*Fu(|J{NBq-P@i->M
zi*;3^Gbk(AsD@Trl;g8s{;E`Ca_@oWBrU|Nz&%qaxnhjC^ei5*Hc0Dx>|queSo7oJ
z0!gBL+LPX&Bwz&g2`rSD!~(D4qw_1*e^c%ih#QcS!{2s@Ipz3c-B0^TIh|kU*Is+W
zEZQ*@e9Q&Hip92oUGj|^zhh&j%*sqZvH>F-h6ke-wn=p^Cu1a5cglXtZMPeg5#^Gk
zg?hfg$Xj+<VX)|DVai6B4Kaek{9|JL09g9s^7W6o@3Dm*=c)&c4A}*D5*x{t(|tNO
z3hsC8GZ!!d!;M%*;X<3dpj`7FhLQ~!dEu2cmVM)2Gp<-U|Mt)Sl%e2>SwlO<$YL?I
z{`nK1@UyYs?|bk;bKfJ6&m8X(1MTvQFA`HPHD0<CFRogpjlym2=d91;k3Leg@yv~0
z2#oMdvCuzY-+jzkr=K<}4}Rq1>W_3TAHLU`us9erEn?}2eZjk;@y&jwt@An-OMx!M
z`Q+%Z1x6Y)FIhKOq-73~D=k<CtpYAa`i4@~BM-IR_;6}0ZBSskCdDFgz(|8wBrrw*
zT401Y5TrrRB6EUut>rfq_fon>+-yxui89|VS6YKLA7`Ycqf3B}fDf?*+QrBi8Gw-i
zI5cH0yYO6d@cwU|spV78JZrvj^A9Ymq(k7(sQQC+xHhBiEv-#9So=R;@p1E(6+6vb
z$Hx0#{^Rp9oR}0VWy}V#zu^66nxo%-$V?i(;G*kGVt73F>`SIgeM&n|s4aOmn+d=U
zLnSSMP=GhERF*C1Hve+Pr)=yKcb-pt@_%2xp+g~nIDy1NsRS^B+r3V)l8SMqg`%Ap
zMleHtx$D>W3#|REe`x0W9=pB8oPW+)Gr!%JAOGSG1KS~#LVywGDO?EkD3*^e>o+NW
zi3~Afk~DYvz)0-6e2<k1jG!1N21={II^WQRXO6mtR<FH%edbHoTpul7(imJ_VB~vp
zrByzK&)c<zJ}n*lEoL}FJ0)2zV<G(6YpyY!4RS~5Nxr~{yTsl@?o)OP_8|DsNBPJ6
z7Ye{2F=DpgZnm3C6tLIa^xav>FVQ=O+b>@L$-Z*kHx2GzLcvO#I#VuRl7JCn00Q2H
z>pTn=FkC>g0V5+a#2AJ#vbgjh!hU%R{qB?_E|9Cg-zqe(QqueT?o#pzuYFSn6Wa@D
z_vQ7?{bt8SJ^rDY@5Pj3zUT>PgnWUKBzI(cuH53e00SWmg%I9F-pNox*HG>}@6Jcg
z|FHEw;f}e_Vu6txzW)Oo*qFK77v5n&(J-+1=3ZO|-k4#a<zE|q?~~EYwANnjTEmQK
zJvy*?$i{f<U9-XLvTI8JkQ8Htyck!AF-pv%P-5|y7_SNiMgWXsu8d>(yD}s(OG7;m
zY*OO6h5+-Qv~K%BWZg#ovGoealYo(=7?nS}_g?ec-#*2qS%o%32N*M!aTn%N?8bl*
z7$cM8<7SiG+Am+a&}CJX&s;3Hr1!Ndg`1<BxUoD}UskhtG2`%R-oIM+N^0|5&uO*F
zH<$OR7BE5`3j#(8X_Qx_=o*)zzE)+Az4ppj%;!_q!w)&woTA{WN%Etg{@gtM-1AAN
z6(m_PEzbKu;`wiW?^d&V&6-)~WB&DEu7tj`p(X!^=eN>ja^VM|na&1tIF~Ol!h{%$
zv4E8U3peh|206BIa%FIzNgx3mqZ@DiK02jY(khFKk88oe!oe2;y<h#!n`SX{2!M&-
z?{IN($34F`k3RW3|8o1CdZDC8S?>#*x*NWmX<bX2Y`{pxV~og`>5F0~#U#mkgaCAa
z3+0j|<&Fa|vTbT9!(oSry>!Cywm5f+f$&J0fxv<mmkgnzX`L*m@CWY)6X#+U<@ijc
z&qMj8`unj+f0M2l#t7GaRMthry6^=+6QAKj3V2>F$pwsXpBPZc@g(c%Y`_RHctW8M
zFS_7D8@SLvw?V$r+3@(n!f{&el@2)QKy%(1XZWY=_pe@eo!OubwG;a7;`7flC^~%s
z;QYGY=Wq<+9sM8YzTb|UetN!#zzEL}!5Nd>dk}Q_>KnhEb;HKBCmnZ;`Ngm9G2=Dh
z{~C~c^((KLUEh)#qv)I7k?EP(BE+e%Zk*Jv&2qOiDR)Z%6;q=Mm@W$&0<+J!*-~3t
zfjJ`ZLnq`aN`ceU&#!gCCFS+Bj$3fV#@Zsdy8^GvUvz^2d7ZGaN^r>q8wF+j#IO`*
z=?y7yMMO)mc{J)+Q%jR#U&yVOZapk_TDbVCRq0%`O!WRXff@>AJ|v(+;CqwYVj=(Y
zip$J$4?q~!GylI8v!+i$3T^Bar2!bB{<I@t$%Oji|NXm9&vKLXCxO&o`}WN?Ph5<_
z_Be3Qz04&a%2Y1?<j#A{eNR4aK@u)-uC!X31&nMSw2YC{PdL^*`ShRUPU=kycI~jR
z+x*kz0!H*a_!0v~lHL=p0~m=}O>%B77$dN67&mqZl)KsD9~^ho5we_p2SuKd5sldx
zJ;0&}C8lU#JrZ-J#kC)J-+Roy0qb}|@7`6{-(dP>r5eh+Pd&*Ty8r&+VG}Ap81rHZ
z?u-k)18kg@FIpUBjBFNw@nx;gatR79=2Qf>2)@&uoFfC6{r1`0{j{Hy@yplVpuo0$
zvySuQjFV3?2OpSO*Gd2*v3{lup<79g6(cBv1_#E>$wwZpID;qKEMmY2Cz33eV~R|G
z!jq3Y(wuPgQ5nVh;q5;$&%6*fE+cq3LCk$&-uuS4z8y8@VNHcQ7v3RXVhS|p8rM(9
zcdUy2pz-c}QrojP;F|C`=Gp%6`_Gbb*{&`N`w3tSNV9IkhD2ZlR^Ca+9&7Ho|9&&B
zv9>OiYk`<T&?(HydDr=(Cnyyd@!9Xh3L<6@!aIv0RA>ZA>?C3+g)%bOXa40gpAk@y
zVz&C{fA|vxUzb4uR@b2v6MUUK*l&mKt8t(EwQp)Iueml~DaDl*_h4N4(An=Zd+fH`
ztXnev5F8ywoiC&>6c|a}<4oP-Eq+#(;S&PPE<tR+jIkzTql<q<T8t8C%MfJRB#?=5
zome2gU?75_NVtz84Q{}3?;Z-z1oJ?sA0hVKwQ@=7Gioxg?KH=0o`Pvxh>Nr^9|w%|
z5A~ZV-1>^W-r1P~Mu;&IbKM%|QUAT|w186wj^X7%&qr+Rnoq4!i8*#_ZJ#OWK8ED6
z0w9){Y=eqR^m;?49=nwK?cFnAp8f4(Hj4+I+G4<4F<@j=lj7<~oQ+Z{QLbwVq((rs
zVg%~=<(gcWBtv1sz;A!3-1F0M9e$>9yA{gl=$QGhuYNsaf{Fnn7k~IXx!>4vRu(J*
zaHYe-#0kI!3WZpC3sIWKtPQvy`DXJq?aJXklYkMFkl(ub`xzE~&{v_}hvk3h@h1#!
zmV6-%ZiaBZ;|o9sfK#Yk-%f!nv83}ijMNFM`<O*3`CobI#fm30+iD7?&X*L&B&pw%
z{G)&GM?aPV;$^?6_IpzBb>9zjc+tLl?P(R};c?uP;ijxyk_5o~_KynwdB?93H?{8m
zAAH{#GYa>g-F3G%09)wbF&9Ym$4$5X(7Y13=A*v2-U^=!?h`7^-M#r_1Mq2C7KMxp
zfHqvNCCLZ%4B}?Q7Z#EzU&2MVpQh07xLU#G3W5bFmZqWXv|<-?$+_puI?S^N1w6)H
z`RI89b#|T&>Tq#V{wUw~-FH75Og3y^0@M?z-U*D6F)7wByYNCAC&?Y>6Gh6Gul{;Q
zAl0K4#P)=A5L=g?eujaDTR~p{l0W=`v!j+^#Gc_AC-kWUGh#|k(!o-Ckk~=)BJZ<Z
zpvYP8KGoLartkmItO_<$Uec<41sFlZ^{}7x&7ct|(bQlj#wIO0VM+>oSR()<FiF~a
zI?T|U>xF$N?u6<ztymc_NA$Ajx<(tU+q#L;bpRmbHP+j2({q!_am6eF9MOM(8wj9?
zV=OiZ?3f-)x!tlSZGr}S-%+tFkY*|TCj^KnBU?7^*Pv%*v!01s6aqjl5b!}P5n3Ej
z#DW}p$J^wh>|=W0_Lnh*FF*a&z2?q`9~RRlT~DqPD5Sp@7+JYug}FkTUAI!;8UNCC
zH<{;O`HRhleuI(HCYA~Av<MXH3kvP41PZ_Y#s>9Cs=Zs3mVoel9xDs>vraq7-1G3G
z<|hJ!c~`sD&j0q;pU&9)hVF1!fS+>2k%oYxG5OLp*O}EDH$*q81c1i6F$7<56NJ?_
z^^A$Lh9XGyt}AxVNb<FBDv<hycp!BWvj#a`u@YQ+at!4QN(tJ^7Yr&cx!^+?iUtSZ
z`)S>G-u(lDwFBpve^#xYtytlBQp~v+GmZ5g1|E)C%=JzfPZ?L8(2q%N=+eYK7t2Th
z#YZUly3+F9m4z&pB!A}j>)AuW?hDKn0?vlYJklg7^f*VoF+gzhAIA{pGyv_5n>MEo
z1b>rU#rp1<f=dei$tLYcu@+BGgBZk-fk7Dxacx-Nhigh3<Xpx-p&WbUI}B^$P;M8i
z=_!xZ<lY_9(XP1C4UI@bD`c1F5r@9bg8l9{pR5<Imcotrc)sA$m-jP#x1apO(>A!f
z|6BsBvVM;3We7BX^^G?P%njxV-7INMk88GDe*6>DyJ>yKnt}2VpXB(XjubHRfMPs7
z;TOC9y$UVhKZE7E-F(p##NIL5=&>iCvJCcI?L`bd+$kZ%3L<8ZlNd9e_wc&!pj@BL
zh}(PRuIAjc&vuv2`ux?ee{JrUYo?fT5jTLJk{iH~aglAjCi9pF4ffeLY8~M36J`}8
z^1}fmtV{ZPht2XODg6k~+#<1;FVQ%&qt<8WSODXFPZ2J=7zlwb1f$JRna^AkGxnSm
z;ChwzU45P>3YR~hyDKj3`RANz=PGwR1$5k!(HD$a^lKRV<oTChF<%j=wM=_aUt-L7
z`uY1B6Q2vbK2Q$gaW!Fx6bO52_SP{u{iJuvW61#-r+g#D@FYwb3?ncgMUV38+AaIP
zFm0~i4j37hca3qa(VC?2ZWB{zM21&d2xn!Z<jQ+%KOY05=b}!<D(5S~YG5a^MxI_4
zEbcTNt*JQ6bI~T%y64c;xLEa>z;(6mU(5GF_fZX3TC_pgff4`ZenSfB{$Rse3k|<=
zk3B3xpO0&sSb1adJxKSG6dqV7<dPHKc~mrJMv^sMj7EkfB-UD?un@aj8wUjHJQEts
z5`UxSu8FzEap%HMT;Bi+uo?1&iTgV!httX)|6J^Q3>YB~{cmGV{I}eYZQPvN0Hc7x
zIzD6tu&KLVi$avyp)Bm_v72k(?F0GW{lJ5U;Fo+yP*C8~!!1T#K3GcLs)b0nkm5WG
z99(n6d`Kq+VB<c3SfM2d`B9!<-FJUl^~sTxBmpC^AYenpg!nW7cHxE=Me)6IQTBoW
z80x-2)`P=pzc<&dk7bkdSQ5zN6iy6YT|JpB4lG=Dla=6kXT5i({ldWN(2g&aGOHCO
zt|vSXJQwSvG*dQmO|!yUq@v^<9v(4E7SFyPu}c8iVnEfw2kdXoIOQZS_taOEFW<W5
z2lBNVSKRu6SN)6e1x!HI4JuoUT#x{q608)<;e5qyIC)3$AMSj7CjpnkdSd?V%FAad
z*^+{<lIN9|UKF(;!?g~U<e|7ejP*6{H@>5oA-LB-rW&Tr-n-!j;H+8M@Z7P8e(}ZU
zGoA;ON4N(F6+3KVZj(ZY4VAjt29Tz&uyn#IIgF`<`;c$me2bJjDeG+N>VP4XM*zZ0
zmn@!DE;cc|M~st$={f+Y?6TXg=7hr!H;g$5%2w)HOTwaQT5SYlq-#-^8R{Fhu{@f^
zAQ&DHYeNES0l{4sg4-*OHR{>qlVIL9Si7E20;ErlPl;WE8z?=myj*FS7?p*Nz!1g^
z*dZwJtt+lH3j|kB2=6t3%!mSk3(-$st*n~UOcenv#L+-tj|^zsnUI^WT9b~2OHFHc
zr)AtUsXcj4rqnjg?E>4x9y#gwBh5Ldo-(W0yyLgub&uxThb-vcs`hM^3#|dUGTVFK
z-OWeNI#YAk7N+C<k3M0(`=i@!drXc=VQAY{?#9$FM;~#hIsc4zXN=!#^bCIgXL=`O
zflUmPo+aI8rCf6zcHlwgfW7y$8}wU$_Di{md(?C<XcuHQZ7x0k{0xRD{fL$O3))}<
zcD+^Co_50V_WIb`ARF@)w{~peL%~3RxL>h(*`?)uyhQIME&)Tapyi{F&CJo^dCGb2
z`_3@C@Aj5ir%>W!%;2mlnFCQchwk?PPrh~Ytuy0=p9yOuTsekg5@{a$-BSkEsard^
z>nP8~0<uGkS<3NKF*Q?juCdR+drm&dDs~9s9$I;l9t%>~kw~$?2<#PDETQBAMz|mA
zrVGR@X8uA#V;9yuD3ZduGR^}MJpba0JkC~04qF6+oY$nZTM-!VFg{b)Fql2}gbjpJ
zKK5+jUgojiKAEvr;<}Wp*ALou<F^w)v%Btlz#3Hm@}m5M^_Zmea@WN^N5zah@4aW}
zo!i@<2Eb?i5<U~O<%iCmZB-2TaFtw9g=<RE;))i4EtG6)J<<}B?d;P}n>Ajr-oa4A
zXLhg$xIoG|pIuIw?^JunSk#Q?=LO~Gev}<UB)}S6An~5SoX2{ad2)GRXvX*n<A^;`
zDAXHeFY(es0UY_+9lta~8Xui|K=LNA{hlj#v&nyY=bZwjo(V6l|E_QgxOHkNagbs~
ztaUv>tjyst4Dw>na|n!tZ3hbqmV<Bnha&vhKfgG03)~E|FADda>ou1B=%?B*Yah(|
z3>V+;k>LQsT<pejm(godXs6<gxg)7hR()?j<X|!RkDGl$zsQF<Gd5n6I7ycQpn@J^
zgTQR|z<{TW1>qRzZs_(L{AR3e^IQ(HmcQcCi>&T}v5UQ6%qvGnN4xoJjNYVhcLMVU
z22Lo<F&GiJy_P#q{xI)dd@Xm}eKK~jCjo@<g*^+`IhHNy4mogt0}$a0V>IpQV#xUC
zn4jq9uuRO~%r&$reT7Rg%77sk08yx0z<>ZZ-7#+f<jvkI1hWCN9+5ucam8O79vRks
z8q|)>)`;<lLl3okY}y7^KhIsU;(!s1ecQF>8dtoi)(+O)GTu^s{C+R?9$WKu#ex#$
ztg{>E%b8~>dG#v(OuN>N+X7>xly;i$*KXCoh~GI(f-h^sR7=t&CEIdYA(tXpJ7e`6
z$}jG|*Whn5m!ROoqE2(?EJ<<!6az+>;K&2tnxWw1Q*)S|{8-qqL3#ff@6l#vHoX?>
zh(A92tRWECGtWIAW|Q&zv=cLf8l$je)5k_AABh1YC|RG>0tSmPo|*M>dBY?Oqa#$>
zaGY2qzj<s6?J2x3)~EnNKAo8Iu`t7`)lT>V$F^HG+d`KOCCXyo8ifVU+YUTH3e%-F
zb^(j8n2f2NS*&64iXsIz&&s##qJVmPScORJ8tVGY^DkJCf{ifl5RfEbg!J5Fjmr6i
ztU_oDCzO-LHP(2!<nH?)%-Vk8^?(U1HgRW0TZDZ=`pr^OqjZ>L0&AY_vVrpeYdb0q
zg@0{KF@NM?hsw3Y(c19OemUqr6dG7G(XZ4U-~eC!XTAG0ZZS)6$?*J(FI%94=lf8(
zdSZit62f0N`xmj0eEQkvtviR%8kD-@MlELX>*s+xwJ(Z2<QB;Ox!89U7W?h9m+ISO
zl?*;m$Gc0+A{M&zw{z8oWyS^XeXo5UU7mn>xRr5&VhTUxd+xSd)H<7VJm**!u~=uJ
zj$e7iZBhUXh}D6$urIJ!o_YRxTgM&c5^LWy(CpGsPLyRRk6d2vocm-GiS>7Ap+>(_
zE}m!i{-QAD-MrufXJ;tCuuOeUa||0dz#;#gzycqJ5|-4V@Ge<qo$h0SV?9?+Qr7j~
z)tI*B{c`8#`jhqq7_qT0^iJRk3+6$)SRy0+!xs1$mc`u+NTCfgphv4%50)*WYl!>7
zcbEqV?2b{5xlQe<xnyc=RF_XGO|#k?hKU#v5(2gZv>-M(yEaaUb0J0C5*Q<m0zjtZ
zMk>V`;U3y%>3)=h{u$m}94KpVR)Ie)+K_cFS!U<3=}Cd~>3(nNY%ya3oLV|t?VNGR
zIcID9*fXO9FUSJ$&zf6a*E86o_jy-YUBB(X{bt5G#;%71#QjL`1aU=Ziz$IBlcVG6
zuTI5ZNda7Eo^+x)UhaBOPWocsiNPVUp~g**#y;2(PB!ZK|B@`TH;M7nCQznH_u5_W
z{l({<m!U|Z9DvLiH}K43ZF=v64_mN01k`a4`1sR*G%vpJl6g&FG0ZK-6qNe=Y77kn
zn!EA<QZi=o4#Rdr!OnYh@c#QL)=Z~4<gIVbh*jaP<()(!v{oPx<2s4}3zDim+<Ofh
zfwc_E0bfwg68M}s6axlahcbpZ=^q%B66xpWFR#7k4%tuU%bhe=@4wI9834OW$i2y%
zwF4|QpOL#_ryb4t??1=1d41xp<DRrR>n_@tIfYnimtUM|z%Zcx1*8U4Vl9xbtdy1U
zGJ#F>>A?paptkMKxD+trl?YqGC1ea{E#xa?Vgc%LC5g3cXt+Q=v<dx(!koI&@2t~^
zO%gs^%(K*$dF>H_9WZ<WT=w60Z+pEfm;Xr}xQ4O$iQoTz783!5A*@KA&s>7xQcPJN
zyN>4vmXd29lwqzeJQEl)F)#7yD~m$Q<;Q>bd-IFC@0qy;^BOT&i0#93=klCMpfA)H
zr5>(S0C;%re4%Wtm5w{|h*|GBb2&gEZM#s)$o&Q6_-j_)Mfw&CVV=#Hus^{PJhr!>
z?_h!OOw;Zt5CJ8LHQ<9U<V#telRGflhL{OHV;jZf*{8$KT$xb>+LrfWLT(hgw9oOM
zH}d@A)mIAyjh8(mv-*LuwLqnmc8nWaqn-fvJnz6#8D<^Aca3|YV7#fV_YnX9KmbWZ
zK~&?94dHo|{xBhY&?DvokbN@&)%`h*ym*J1bJ<t%zE0`gVV>(~Z_7~1lb-PtF9po2
zxUZ(%-2bh6?K!hvA)PPG<?K7zBm7=*41HmZeUS_u{56Sx-hIa##uu--R&KXbYfg9E
zeTMr4#z>4enzS!sj`rPMkI0QFdkfYw>|Yow5vKbO2)uUt6~>Y80?umx=LAL^hIQo0
z_n4s5V6JBWMSC&t`@$X&K-&Se?i^z-ZZUB!&iaIV_+vYDzxiicgJ{f5l0(%Wr<`!S
zo`>1b5_O@U*%!Ng5o0*}A0Je6`Ld2+UV2??s1QVA4M#hLo-eqTcd=h{IcreHbN{*I
znemye?p}p_`oM@~vkwlKj?Pr@^$E3Ww)iE5+^3#)ezibM2RgrMn5s>kF_PS?%n@5H
zV<fEBW)0Bm*R2Z=&)M%i(-vR(6ae}>5(d^|LM0fvFHBrX{+9i8?t3g&MUpFtq(Om=
z>U&Q+IfGH*?!jgiixS)rVM)R!FD8EA!`OLu48<~jF#W~ZgO5J04e*05NzTt0Fv0=^
z3vDbc-XYq5Y=KF}EP&WZ<3lrPQ|Af*i{O;wkF&mv!$pQm<371R8+jD4ENXnAuAlkT
zWmfUvA5Z$ubMV<OeZ~4V2k6OHKKIqDXRPp(fDxA$n^Rn-g>3+{2LH{A(pH}?efXIz
zU~QO|-}tz7i?P`GVvof(0CvOw_Csaud)TZEv40PqgL~9hY_?gXI%^_7jr-2V{l~xf
zrTXv_?oiH8+8392q2NCp7fqou-%rQ0WaI5fnURm|Bx^`qzT^_zWcdoYz4zGNE{M5~
zdw%}w-<YK=N@A2H-3Q?1=Xc$0Py3q`7WPM~KMsG}!5NB=FfX2ce2epZxj0FFUSWCw
zH(dGn3OPRwi}Z&Tg9aBNYzP=LVHL&X&2{UoyOr=9in@0l`%W7Rpjj-2Y{L7rYW1tu
z+8ZmaN(+i(EVBv`+OvP6$+SvYPdpJ~d=MnOL&3sPumhMhNkI-Hp;5pEpasB5OIN!E
zLST|`j3D7C=Sd4tLi%YonCe6QQ5t<0C^9Km21s*jhiUHWu<6)=wsp6eu|c`qlA^6q
zWguVNW!Zr4y3R3H$d7wj5Jv2g*3Nb_s14+}d<@ruV1=~Zt}CQyQ(H92J(++JVxqJO
z<bhd3U1^tvVrtMl?>bVhZ!|V_CIIAqop_$sYGZx(Ll2q9|M;wcrc~f`Woc!wZ^SHI
z)S>rUV2KX)bhVm8|LP!flHA1rfMgRCASk<kDC=ny7NhDvfDsrYa=>qnR_p@Y&a_KG
z5)(GLFi22(;6fv0Wx8uz90IIK50h1Cp@Pu_Kq*W(UrCTpKEKi00;VUMbn5E_Wzum+
zA89Rga{<+iAt=HblVVI3lnRWc`3jbpzM^D8VAUQ1C(b<m^e8AnzKn|~w(-x++!c#!
z=I#%B<lnV^$ahpY^R!dVf&1<2E-%|B0qtQ3P<CGc?%#9DNtR*aA9vqLkC6?)ICb$|
z#JDu>6Gi#|{lb^6Tei?h#ifaTGB-0v`vNFJUuIjUP_}RX;7578NJJ_0_m?XeeFXIK
z@BQOn++o*`9AixHk_GIQ0))6OcF)-U!Ot)JjslS2=P}twMhRGL!^HB}<*~FdrXG0s
zQS+3Pn_<6(=>ZGhr@cjJjqK8KJqpL$etCzDv6f^pPkPe+$Ctl4!{~B(u#RGlj0?A<
zHGoS?9<;*`@3_;v@TX_Yf(7w)fJ<Nh+`3wzmh=RznqPL6tD;=;k5_&wr=e7?JYwfz
zyN3Wi?>4~5k%zrK!ws;@!?O2&DK}it{n>({fAjH66TuF^(>paMGB4*VDB-WV?uLxy
zxDjIv@|EaeA`HcdbH!dICah~a<!Qjpd$|4u?9L|4?O*uDwX<S8?k~U@CLO{Yfs1Qr
zaLBQD9rI2LDDyoZVIE}d5K3z7I?Tq*483TCX;}vXpaW#EKSHvt{TUClt=*{4gK~+6
zkp+O*CqH^&qOnO5m;oqq(y_<bwNJKwqpXx4<3i?ie-FyJLIqt`v)0Z-LxW~oF4Bx#
zYIUZ8k$Q{~o(s&#YF`R<n^kJPM)aIC%X2}@4PfoJmfj)tY^X=x8p>TOWn{b@>hIlp
zfstVa6zpi6Z82Ue0oHckbI)0u^L$>`*aXHR1qG?U(I1gZAAC#aE5s{UDH^(yNhztA
zIAbz6JRpn2w&<XMV&LtE93(4<on`U3*eY$YwjxM3mUYg}3>FMAsxS9*mV2>#`^Ts^
zfRMA+WWxN0e4slQ5PsUE?=fJ61r7mGQCvU=#u5kj2(2jglo6H|`1r<(1BGk8%0$Y`
zh8e50UA54{ni5Mjm@*>@PE0UnK%wVfdMO&qfjW=}%G)HCK}J2bfn&jmi-FMPM@Cxj
z07?kQ2%ru2*(=jB*(Haaf;DDrbBki&kOSUocHL!HOQmj=8yl1{Sg*onp&fDYfz?gE
z!lF0k-on3r5|=qrh>pwJyhTBu^AUgzp1qi`AxGguTcF&gA6S&Vss$$OA_B%aVIc{-
zC?+sRo>FjPl;iozn4UKjaxuYjnxL|jkLQVUvDk-2fz>ka-Cty3MBn5yzCfQo^02q*
zR@s+B^tTV%u`qv^7Pcs$apMI`$rr4d_LRj^zOq{HfwL@%ff@2HB>(&Si3RnRv~_X2
z9sjW4#8t@J*Wa`XNS-hMxs2uImkAee2Z}v{!i~i`i*&{>+LHI34TZ}O%i<&}FW5wE
z7Lo)|mo~wgoHj>cOy8lrgJHKu<$por02>K+zofj%b-b9x67AVLDr>HV9cFOzkY#6B
z#)iNN72K|@dG0u-9gU6efeS0#HNg@=_?(a;-LX!l6oX`13f@MIjnjIT$%eK;>~k#Z
zePWI@i%~-GbXX+QI@TlwV9EuT(#x_BCIrHzfGymYv7NkJJPl%F@Tm8Rg)rBF*r`7O
z*In!s8jrF9_TV1OF-IoE?7$tE1&`FGi<d4k`|Y{ARfzAj{r1*U(<wo5Q@BZE!7E}W
zydZ`As@GmMBVy*L`Q$P{3>1Cu>}u0yU0{^jh<>5J0REa4^u3AChGqdI`<dOvde~9V
zZI^sevjJut@9iHDAhO2%?vKxymoyiSs~>4kdaF~wO0znzQEQ1d^$o5x4wnlWEJ9(t
zG4EjwiQ*cUIBvdSUEt<)`p?OP3jNYcuV_=e%DV3Htu|enT%Y&JhyevS8+|=RHhHt!
zJs&~wiq)g-6CTPys1uCL)Y@l}>5z|V8p4iAvead*05A>%CG+Z3W~?=EH<YhDB1H)b
zA}0VN6B<L7ELm#)YX7%dh92c=QGnY)DFJZ*_|zYtwqv6kS6CY)fiWmO?zs2Yss=rp
zTj;x;0`^Sqt8%Ac9)l$l6BI%F?y=JBr|}pz9?w1uqjh2iP*;>FuCC1A!~=^7>qqK(
z^x=n_?(XbB;LP>vNDOge3O1g1f*|913ATidIi<c~4a_>=<yTglKd*XumI1*`zL77i
z?@^=yKBt7ID6?Pwp!Ef;DknUz7!q{ry@d(N-0dspFkHd-B_^z&+~%M0388H<4oDKl
z2Ie-}n)Pm{-hEinFgsvlGnd+R-%G0uYsY*ArUuNGy|s>5s`U=6Ulg3U?)m*Q&sw*9
zu3ohH!3WHCFXQgRvw(8ZrFF;MCu>%$t0_OOECD{L6LSZ^Eo@NUW5xrNzkPk1N(M&8
zrK)I=s$xD!{HZK|&D`F8LT%R^X}eO(Q44wk-uc~j-+k8JKB*;7I_?<zoF<(w_V`1O
zJz?kR*!8Sq*%z^n<-K97VQ=@U<~kT=t5&@@<GL2Wg1N{Ea~0zebzG}4;aRb{VlFXV
z`YoTSp3OE6@$SNuW1Lu^{Tjn0`*s*_j6b};+1A$mdZ)VucDm7(a$_W!O<*tI^WbkZ
zrllCoq-D%HQTtUjj65-{vu)z%0h<JtWxldXYdGJnS$G|NK%T6Pwo~0<=QV15Jfd+t
z)h}x`_rGHG3~o*&k6dei=GDKl``=wH=z+25fWQQR72-Xu)qF#*AQ+g1?m1IujBS5h
zb|(p}MY}Iru|i{Nj~#n?rf5_84%gC-8JUzvA;$qmaGf<cG+-uCbdRKBj7-Q~(tKr%
zxN$R|c8cj6^Sds#GKpVVNO|Hp6n?bWdbvAg(}B7)^YL>o`l?pj5#7hUVT{noObD;a
z@4m0tt9Ik>B86Tq<!-slRIZ1*U)Hs?5O|1<<`tK2iC2(hjgYM!W55U#{Iz0?uz+&H
z<P7M;vlH6{yL5$qk}qg?X8_(6n$KLP&-QHj3?u66FUTrg?wBGxUkZM<m_OC7M?URT
zi)G5E+_}=k`p|!--1XFdO3KqQDPK^;#w3<kES)=6?lw(&rnnDM&3D7qv6N5R6IWUt
z3vj`N)wJ76%3WP|y9GQZ*(j!WO|3cx0|EdI1>eMIiW$<T>ru*cP8ulF00^y(CV|=!
z3l?bXZIm*sMIgcO`b|of0!47K#eGRbKFFBL($pbfMb}J8Sx-z88z4UIu8Vjkutu7s
z{6`2TK044Vupxhd)?Qe~KV8eVSKU%&CR^GA!NpZm8|4@KMu?zu>o=+(>nL-()TS25
z&@(h55CTEjFeC;@liF&0aKv;eSYRt4pW>5ri`~<w7%U6L>X}gcIF~ZY%5<tdHVNQ@
z0i-UH0o|x+Q(M6XY7qm(3Hhg<)oIJ91X!UA^!_*?l*-B@O>fY);rm$b_sW&4mOkg=
zokrP&f9!3oz5Q{$R-dg>%H_-&wv4X%Kk2D!1!%~^Ef#MrTTTQaZdWi`!~rDP(kGoO
zrS8?tGj=bYEtI55b8~O+CIMe{#z?H~DtRp1o}(Q1rN8I9o*)TS%D4Ue7_M9szUTh9
z9OfBY!-r|g@jEP2K6x_kF;?a?-sS=o^0{9n(-c$xT;-`mp0Rs!A0%n)ja^g7F`hxz
zN&^-cX)z7@?HD8F07eRBG89twdfLjgKsctw?&8L-A#H3IG|RYdp32${y=M8+EnxOM
zt!7cxVs4vPz=&%F7CrdOt+s4uEy$|>u&o;yGTX_ocB$e9g;F+3rIvx~_uFT08@t2h
zb-QToPswt%5QzaJTu-28=YA(Czgh?im89ek1I8uQp&G~M@_EU&Ft0}4>-Q+V1!8@e
z?U~|Ss2Fsmh~?$Z)jtbw0orp=wyw=$V~h}R9V&tZ0A!2+=u)7=c9>A0?)HUUW>lXr
zHzov1*fdhqqwt5p!7kLsxaie#Q@<^PL?8)y5JSXj+O<)d77GH#$due`@kvnj23dPq
zMvBh6&6G5GTR;}_Ndpb8md$db1^dE+Bq|p&I<Q%a<@~LfU0$Vrws!WYoy0JZG*WJf
zYZJaxMg)KYpr?CDmzW=72dEvJ1prNqjG0ACyQ~|mW~{RXDA5(LZyMCz3&rRe6JRl{
zKIzu?eg$NwO`8R{0Ak>B42BJE$Y7$h!Xg>dz33MS=i242jAORH(rwwLxF>@{0&Z-8
z_-VN_8?o_aEJ&jMY!uqkD#lCF`)D(-|0!DwBmpC3yFra+v@MS7`K~`v-0^R;zpr3m
z1U3YUB4n=D`*Zc^<$AWt!5DFM%ePNm`da@?+MnbqcRmVx%}z#qgIrmae7;un1eJI;
z@=*eo(tz3Coos9PQtnxfYbm$yN~;`s*7K;n7I1q-fRPEYYArA#)=fS5dhPX9EtSZP
zUF)=0ozK9*+D(0C`QimN(ey*|{xsJH<UY{J98qpyWT(ze=GCWCF&!Ux>hYO&Z|I#K
zoj&HE!|Z{5-<JBEIa`4*udQBfhGk(~E!k<s3T-q>2Pdjk-q?NIpfHykB@{0gD9EUD
zr;*ENXsjL_O9Dn>epQq5sI}vBf%0Z^1u&P-OSXl1t@T^IJn<Ir`_Ngx!U`cCLpTOu
zMc63Hd`uC`#<EGENugG=@lw@~(gO6&aI+NVqbTghO}iKyxEvW5(}0*FPO)x54JpZ+
z6f0vAR)`aU9>}z;++lb$>6)>=L3<2<V@k)Z+bn`Oi*@4MNllHU=WG}cjyZz+Ed5J~
z63c}E_l13;Oe`?c1~}83&R@rI@dX<LLE+yh1^?js^=2%H8Ia1e{6xKVSQG9Su)hrk
z2uR21k<uw3AV^DhhrsA=kaU3bKsp6MI;9&N-O`;RoeI(d?>^t(@4ept_TR2+JJ<7^
zbDneV`$L(f<Fz!-DFBzSV?9l`Ih60~42WrYycCe&Q~N5yO`k;%b7z-M1+v`fnFyh0
zzyBzS1)}^^9%s93nCXR$n@PYMI$EbA>09eDjZ|=SEXN-pm3*(nr^;zic*FV9(9t-V
z;;^WyA6v`Z=S5*bV2wfqILc7i8su!U1T>SEXHf}=J=DjPH>ubCSC5C#>4<J#rn<SX
z>q_0sYU1aWKkqcFDhjxqTi5~Bh`Wy?{Vid68-Hgl5jvGe^Nv-`DpF@M2Q+koIMFy7
z;IiUefY5Gy%!1)_ajBx``*m?5o~=A%mqR2ykl|GQhd>Ligue1mJ6p&eef@8O96pA$
z%1M`y31~onS9h???@L9n*_qNHgkr&{qeNCnl8Y^^UP$)AN{UNk)KXuHbU5*5{80+U
z{K|E)_S_cSP!niQrGhNA+ah1zucY8yiJ&O+mBQQ2HD+Da9Ben`A4W$~=lkWRg92Z&
zv#GxC$Z+g=js0MX69#&Gm<YTs6OjA<^z*F~!QV_Yzu=|ikVF_)on5Xp8Pre6_T`d^
z1kV!L?o^M94CdO`B=&2Zm#p4|24%&umlo3kayiRZJ1A%zs~P7vrPKi9q4cX`Dn`7B
zl$-b^3l0gcW(}?7qgT*q?TlZiBLirPgF}+TZ$vUcBAoDiN843fdrOcxjDm5S3tC{R
z5@BI@pu&1%%BVuPBP=3Xc?=}P`ilY^o67>iNLVyjyejFR3oY_v|Eu|<=0yiW{+f0)
z8N<oelLUYkc_kJd<w<2qlcqVF4XbGC1_<4z7?5@*m82Q;WWDqR;*yxqB)nGxSc3yz
zh#R^VP=Okp4Tr3dL>3K=tw}jDgPPGp+Xl#fB(D;3WRY0dO>zi!>1#uGvvjM6Ff~2(
ztrNbV{_4ZBiv^Q{q+*&|nIH(yYmt%2n(BC)dOx%U?C!*uI2s|k)18tJd3E09E-kso
z^ENYV7Ih)}Ldkbuz8V9}(VrH7JZLr{_CLi@VU|1m*_{ws8qwQtt+FtQY~*7~Gj1NJ
zYE6({*&h#Piz|Lx^)>1|LLMu%Ql@df_;!|FzukGcp}T=y%|M^<fii%u^u7?eZE{<C
z^<{chqm%>x3U%POlu!^_CPTE*K&t`UE{XTtS}GC)5$<st+#kMw{&_7Q_op@2{wb?w
ze653kwj?;kHh(eCb;#hW3}5Yyc(GrV@}CN!Z-%avMYASdP9a>b?Jx`aRSx}3lNIk_
zC6ir}0b8}d4*#}uAG0zfmj2#}R?*$}!qP5%BLR6kVrQRq)e|P3uianV?FKt)OqDfi
zJBA5RCK`9LtqOAtWZW&iq20}?+URMh-9R0tg@S>cUmWN>kwZ1hWvT;0j%KE|dX<ax
zXRz&h9y-F(<#!n5%SBg&sqAlQ{e<JwYt+w{L5u68NhX(D6jt91*e&}gj*C7IyOgO^
zUpp#~%nE;S{#<eVv+;7{wh?fW<rqAas{N(snU}-2gL&ChpQDV5*Quf#-$YtUqP$g$
zMzoc0`J2Zb%|8yXnC~1pF6i|&&ii0TwP|!$u+yWsLPj<t2g7OonzcJ@OaoZN7<a`$
z=S3V-<cm~S-+tR$WPo~e$hd+4>(TnRHMfH2eHmV=(|@$pUj3jB9_pqjacw>Eq1xv7
zB@4z&;Rm*XUBEHAO`Llkz|YPW@8Yq^+J3VQiZ?gD{wLvuGl}{FZb$?vNlkq{$DvL4
zgrl+JpwetQ>WRpnc!-xl&M4M^j&(K~jUOAzu!9W(OqWaUF`|LE3Rfd@La10r_W_ki
z>*SOpF8gb}nF9j~HVs*1><?j3gMvN27)%4(lw;(B(a1oVm*f=-XNM#i6Pw)k6lSd(
zPh7dl*-{g6E>ougj0i3OOp1ujhdN5yj|)fU+Lh4ws~J#$gD|DDvA}p!rg>GcAzSKb
z9y{F<WQ86NkepQOd(&N&nv4evEp;{qTj4sXga`px8hqQ~1ZL(scvtbI&RTqLhmy_r
z#vK9<(kc;upJ#`A49mg2t+0d*?_%1y?vian{H9rI&?M2D4Idr~gi1xR1<=#52a9Ll
z<=}{N2B1$C8{2Dk2~OpTrladDtDOH!{!~XlQR%>+X^i=*NRKa^6CJ|F@?_LmEK$98
znh9Hua%{35z^clsT6=F>0?yl`vShOVu6HQ99!CRrMeOE>a<9yEXO)}XZz!>|)~|if
z@Ensk8r`OBOSoT%$<5PgW;pd5ZTjFoq0@9VEcn7Uz<1ZfETc;8bFVlpXU*&gtN!8G
z#chM>uI?q{oC^x5_6n6SK^VCQh7#cO%qO>e(-4^eYXEwwN&4*usPU!G4#;{@y=ISN
z;JGdhB=dpu{z!PY$qy63B(4&MhGvtKLeB|TBHXVL7Z<nrlf-=fUBcKtG3CtBaUnY=
zokv;JRA3PEKS&a%2aS=DQujYhE@O!%@%Hv{i%4oKJh2<7<AIWYrIzdWU)&dwwvQ<%
z=ldd|WgoUX@Uv)|Mx8ER<lc4O_OD$VSsVS`y6f&RHs^kk=!A1&LYgrVyW&Lk)#OAg
z`eI4QAefGS_>*Ec4hPe&Qf%Fy4ng0Bg^<M8V&y4&pSupeIMs#BI{(ADjD7z36du<A
zS_>tuAvPx=P<dVT*OkmbIuIBA3Juf(q5{=QEnA<w;^SN}!S0B6<faOcz3{A9^)YNV
zFe$kugIj6El<$(&qIw;Q*(68;6L~>a9}Jh4sMKzo^gsaVKx^}eQU~EKkrj3^`bKKP
zOUBLz_FsAD>maSgkKo7r)s(W<D&>%a9ZFei<;HMM3tbFkAs>)I;Oo6{g8Ye`p%Xly
z)|w!+4Bv&8Oh%@9*v*1v61|&01#k>tO{-#zH9uU<zkwR8Y$?UIk^qWLCFR>C(==Nr
zLDZQl*K;){wuvaO$`0CV5Ta+qK?m?G^qwoUwIGjDGnes!Z61FQHI3=COcugY1v^n0
z?bj256P{N$sgy?;kmzgI#f7(yX!;hVR)jaLT|B<da^-MSrGY4dVQLQb{=Y8^)5?zj
zXblW&DeJk2cYp1V?O&}#>;yJlno87~DA`vyS^h9N#qper_Hb3@lUL#^9-1S&xP3hV
zb<~C@+RyJY{yVL!T+yr@HjkexS_fI+%$_fjrZuK);7Rf@f2z7%cDi8khQ(+2A{DBC
z9Fkl67^F#t-9=9P4Mgw9)LCQ($t*KUXnO7>qR+P+!A8bRa`vjZ_Fgpf`mhkHQm2ws
zzEgj_n^Uaq+^XW-wpLeE%uJJ!cK`Lp?|0(h*f%==mONPHqJ8U<`~At-<Nc|nPa@Ak
zM#A-fKGb)5xLND#67{GjkXkFx;L3!@PvUQibf6njp8sFKX&t`%-sI1<c}=hLi!W|)
zTnl3mPKfgv_j&fb8e7mzHR->kI}|43Xd~7URnF9P&fv2BkgZsiBJ87buZ$(3kU8{>
zE2+}TD7c9fPx4q`+Gz)1lSiU6e~+VQ5dUjM3HhTuhbtx>b4%7b3t5|WT_(F`rHuwP
z9{FVuL8Su;A5{@Yv;y4N=n?y3!11aTHlR&f5CZ@h)a<RutcUI?Tr0}iE^A%%uL|UD
z-knebEDjIMQfK{r|Iy$DUf-Wl@_RMhMEUV#E?889x9@TbH9$dcS`s}^HTe<v2w)e)
zsmw@dCom9fz&x2zaUuf}r2XvaA8AI5kf&q?BA~l(E0C!+<JmzV;TAG-)4xC!!qs98
zvpZy{^nuVR*}l~4glNLnb0StSq(?5kQN{>BiX?@7_Tn1Y<4lO^_aO<Bb^bWYzqSyn
z-bRNcS0Q!}bb^l3g41n-!UED@nFZlv-c4>Ik15)<dhvC))y5`M5B$5HpSq^#sb-kz
z^o_XRJkg$Z-3IScO*oAoQH`mIagPhHgwD)Ox8%BJlb8%M@VHgS#Mm=dvw<NLsq`|&
zR<zu8?4BNm)jlh$?Zcc#E$`Jv^s2LW9rN_6?^7A+go9HVzZuYG56XKQ!LNtGzF#WG
z1{-%P({iDD^RQ3dyh@odR=b<+m&32KHxixB{ukS+>#ft%TE#YnDSmA>?3%Jt>>@(9
zs69246(OTMCNSb}ym~eb8EccpbH&z<uD)`e13{zXT;ro!Bii<+d7&zuR*ptBbC;D}
zIN|r?8c~{LY4ijug{PX9zP2#+d-b&BDp)umA`eS3NeS<_;&o-;ABX4IH*cQyP`kY=
zi{A!6l2}j)xxBo7{hR7yFJ1q(<or&}i3pam%>ic+Ub_oJY_A+j_Y2B)O?WoulIwG$
zXp=pw>g4xbRoAYSs=HeTmH5fIZ16;kS^$jfu!SyzrRANv^(y5Lxs|a?@dMP3i#_<7
z>{Sxpv}z{jkxV5zW+*b=j!+q9!d<C#{5%6)>rt8TjbI8)@g>fy44n<t(G%2&6gm<U
zfHQ^0RRde11>saGSC8B48?}+E5RzO3^6N8dA37GwVf(>}&u3*}Z?e(17dyaddB`ss
z2x6)Y*>o)-l79eUH2_AIF^QE)bqRpd4U8cvAoPl<=qbax4B!-%Wzmh$G2BioFEhMW
z522s^a!!2H5qCqwPV5NI;OpCgklRg_zAonEU^bCa!P$3gJ|;^~`-kQA1>{m{8^Z@G
z<wGmQVg<_90oQdO^Sk+Q>NMUMyhLw$+eYr432eP^&@+Sf(E>YaPdGQDJv=vTQYV0X
zyb4(7@qa+w3tV}r7%6ru-2ISz4pGW&t326(sQL&Mqs<>B?xJP!O_t|3sJ`I#l){Q#
z(jrRG#df!&RV?!-Bth$WzEV(=XKJI?o@e9A!1JT=q@v8q3#HT&quVz3W~1whqI&?O
z@`&c880T0O#CyV9NIb;*vP^`$ci6?S)OLor#fNpyG;Q?7itP}D?b;;kWKGei*OtSz
z@dps$RdW4o$0gZe&ce7wy)39#U?0LLkHm5zY~S8Wd*$4Sh1f<zv4|dM=iPq|MHw<R
z+bTj`GbA=PT%7P{1_!t^_vAWtJ}%7noE7?qTXS4KU|(eOd^e=Sy}s51Lt0&D9A+lP
z#6^R~q#~EDR<51PCT^!QZf6<E9httoX3t>}V%6Eog{#3KoUaCpWT|Lrg^q-rcpWza
zXPj0WmiXmkT6wBld`xRS*V8>F?rHs*&i<|_{ok^&`T)9g7}Zode?jDUjk?bb?(S-r
z9YYN0Sj4rRb_WI38tl-FY!NzLJO|mDt<`nF?tZ-tf(yXzP^l#0;&Z(+6lh21mv=$v
zoM2S+R6epS%^KR@M!NV64%MP=z<vq2Smj^35XH;j<t}<N;$K4~2|0>aR=T#oz0$=f
zt3;Pq8JV5|GOD6mq#OK}y}&w-qnMOKVws~qhFH99V6Vn-X)t#(5Z2;-sUweLw;$MU
zKvCF#LgM}^K`(@*AZ5XV;()z%qsAPdT9Xc$5(IF34kMI?`oVlRl`4zs=NbU-qJk_!
z7+3w+Op}*mH~0I{@~QvH*cro;0HH61ie8U2A4)KesE=E(IkBZad~eK#hQu|Eu}LyL
zr_>}<%)WEhxJ};cDOxM`ok_4?5Y%J!%}SO2O!Q(lvQAqFo+y<1g<s3&RO;PqnKE~C
zmQFfE<KSHBt`F9z@9rSFob3$VDwOl@8tX5-A1nFgZEhuTY2$Z9l>OFoEyPjD0<k^v
z!SU6XinXY>HIjd@4z<?E6teJ@xWE`@G)8YL-!(3C&O8=Z;C9;@ev7wraY`1OTx@HY
z<tGp5v`p@==E3Am`(Gpuj%b}&O7Lo(1T37Tf56$U*i;=!w4KD1&zeRgDpEXGbqHEp
zGBEtcs<@M-K4`YF%9BvL=0(mIzq|7N!<4~V=P#d_XUXf$IxRp)zvFVZx#rAc%eMDV
z5YH8%aR}Q&vPN0BN$W_utHh1vv6IY_){-Kcc^}FuW=lBsN6C2Be2%#O_}S>s^x8-9
zdho6XQ8d@sXzk`irS*eCkr8x0QgMQWs*{h)#eIz!JC>8I3)N~}<nMnfb~Q&x3IIAF
zso5;Z=j4nZ?bdlQBr!M>5`D<Sx2P3sQC=17YH%RL-HD*#tdQ{4=BlWn>i2lyt4&dd
z7zhNUoA(0AZQ(SqLt2@@wn%-4=L87ojhX&UlZOufxQL<`N)l6oRiNlZu9}bBc!Pxo
zw6i6OdoL`khSo?=tFjv0L9gtIUhzu2c3Gleuj<V8nm0|A#Anz5va7I~WzYjKh&?uT
z3Keg=r74K9B&qoJ7F#&jE4<Z<{5^O^tW7_Zai)03_@quoP&aee=zQn{skY_*1E>S!
zqr^a)GlgfrTJUFrqHcRyyT`-QQhOc-&>euAg?E!7dU3_ZoC%L!lfJOeh-VOc-N%;G
zLsh!Hbv%WyKb+$nzA%Iko{zkpuiS9*F-pq@>$&JB-KJD6+c$yDKfVZ*wDeQIAbZ-V
z75TdCY<6_bQ5<KEXz=w};%k^bdN6YS?sPJc!9Tqr%oCDm<oS$Tf-ZGYeQ=c`;+kt$
z_-KyL*>+_+v`u$xI*>HDf3+gPkzLnJKf>uuZJKi3R4n}qDpI+Pzx!YBQuIZ+NamtP
z@i7=RS|bpSXElCyJ)CB3HU8mc8*Bw&cD@1I)xs#W@OaZkqfaKf8%Co!74LRlO|0r+
z$`cvYea86=J3mbKR#vBs=1T8bFf{L)5+f)g(#L@)+MVD1WN(0<u~g`UYvm^^h*S)2
z3bGxxoh@{koGFTo8cuMmgI)*xLVFM}k7zvA*D|25g)naST%=j$=Q<aV@kl+`HIe0s
zTObfog67Os8JI|TNTSCuQ&AT(gpchnP`0wp>^o4PQXxdW#uZL@l%(iV%y~lZEd+j#
zWCdC&noZ-1$=w#gN#TUdYF|DRh4s~u70bHJV?^2(YmP)=APFJQh05Xy4G8_^L`(gX
zC5ZswRl7u<Jm3zd8?dwnsYVF_!fPd?f2_dIc?FFC=QGHtbASm{Z;lBnn{@)r2r*`-
z;*s4m_e42q=p>1MXp$c)N%|mhrbSV4{!yqfrwLA82K|F0YCHyqh|PDV#$8S)c!Yp|
zfBo8B_mdJy$We}F?R!vG0=7}8lt$o`0w?QE$7b!w#*~_C(c~)wxH1@WHT1U<8_?EE
zol$N`4+_aj;+jU&^w6aRxleVRG}b2>FCWJvS(NaBSK>aGOY;Mluq*8T2Ev7uY`(2;
zT&7QR=eE^8+20U}PEMgl`#c-|^<*Q-Rk)RkLFP=x%MQ6m3xQMGQfPz!f9BQbP6HX9
zlfL-XvBJ@|c_hKhPk(*W^+!(7p~J`Wi(!|Ql%PxM`%gxC5#;r!ij{WHUN|@(vp1MJ
zw0=_FPpf;ErK~>mUvP{pZ~qyK@iSPcJWFk~y45(5U`ry);~immka;MUanWSAkeaW!
z=NWT<;*{gKB?*G&=;TNndMLXL=Pf5ky-~C>&TwV>xf;wt`aWxAdhm=^+eTEQKt1M2
z&R}nwEtszDHEJzhX!P^Z{-M^UElEdx|AgaJ=JL(XoAyS{ug67s;sKFU+rov5w5;1j
z1`FyH8qE*j7HD2~2t&)aIs`VKvc_~l2IjWpEk3Wsno+a}DA_`xmS?+NDG!ai4&7j1
z>=r-(P(dClJj=RI8JQg*p|8erm!CG8Wdk~?@<TJo45nGHjd<lLFL9e<=Xe>K&`~qt
zCT58|?k@qvoX~;*BQN@#eCrad7B}fuw>NS@<|d0&v2tj}s27sw{@i&X9n0vDpZ;Uu
zfD&3Zw{XDzuO^xoeF{INN2#ZT2i3aA7Zx|80Qg{8^81JmU`R`s?k(<5f29*q)@%Kf
zm;-(WVKX>Qj_82M!fDI-*G{`Lync2Hz3<~K4_Xu!zmSXDcS3Wq9?e763-INzD>G=i
z@x$FZo1~)#aEWl`L!X*v4mmsTzl>?Zz)l@=tK_YAuC%8;4-TE?Cf{)(kN7vJO{ls1
zEh@C_7N?3grboX-pN|=2c=!tokDIDc&7q(~ojoA7Z!qUOn~l6BwO>ZZCizvUtIh!o
ziLPqJu5#=;$ka32m=r4^n=3+W^dC1qH03{L`CpJ-q3G^k9rbjVMlvoA&E~UAeb(5F
z=C!VATbKD(+O@}nEb@<EzzH*U{<^m=&`iab(tWlMaTu7AOrI-)7@_b@7521K{`wIq
z#QfY8U(n30iU|+R9NNGbkr<r!6AK}`lrN@@r$Tq*0;w!|e@7WEYtjhAZqw`?)tiIu
z7CDbpgjy_JD$p1Omq+IS@;&5TSvXmg&R71m*mfbUf$!?L#&Ak|Dhq$!^B!3}yj?Oz
zp2-OL4=oCfK{6eGFP@p-6}yJ!?P?YN(o-C{<>R9R$aCJym~qxi=gCK=>@dQ@38>IE
z^W{eT(QQD;=s4~CSJbTYtu`Py@GA720Y&LSg3i&MB)eVv?O&MZDHxfX0|1D^)Ju<M
zG)HuE^jg%v4ur))onza^?Uudu!(j0Z+VX?&Jff&+;aL0bph2DBk9D4OgWGP^)tnWw
z-a7(dhtfw_n949s`aYw8^H&}=sG<thx%Yu{cD|%;r@Nw1en(V*^lF-W2nB!w4y~qc
za_;jw(4mf3Z0s)0-~cEkgP6@5k<#?8rxt6r&o*n2$;;8GlBJiS6HUG}8*Dp^*MYK)
zv^<8kPDhOLrh>0N)ir^Xt?Tccw5$ubjCAl0C3DBy_8K6JRd=S?VL;}+E1Whh;mW7+
zWe^^Ot*EnpJx^Z!&QmsythTbBU7qcDxirBsU`Gk3pKyx<IAWB^AYgp4%Tga0kdp2~
zw3tNrCwe=D*WX3C{9jkAkFwI*3aYd;X>sm{Slr;REipH$7sc;sjp!eCZQn0v36K8}
zbN+J8wTSk-Y>Ldcc*{$-;<z#$laqs~z1DJ_l#SP-$+slyzDGKQaAYkLLms=*L)`Bf
zSIxWQ4QpMir*qsoJ<5cy*?ukx4|p;t1&e1lgf0KzFkP<}yt*j2%1SQAOddvPda}iA
zT3o<)GT$1rQ?*h4n{24R*ebphjWN90QMbBu9yIz5K}F)=T(FJGlsbKk%(=g$qmpiz
zacNeXUw>LH**E0_Eu7}8xJ!#B+ZfM$nA;jM!H@;#F8^_zqS>TzQQp`%b3glC4Jh_w
zp+Ce2_}#)!C0c3|W-(Uy^Cdj84+KP4BX0*Gtvpm*O1$FVQx?9sis&-%kqbLVGl9Fq
z&hiC_Vd@6I(I=&r{gH9@sK?peXb9u;k@Sl+IgoMR044Cf0+$+<5%}|v(pW!LK8BD>
zbjmZDa~>xtqMen_mQmba*eXhoDFGZnfmsRy`1==kvQDkKWN=y>tTI~e0j0ejJSxE2
zyf!E0`B#$`YQ~z^8$bi|0S#MNx=I{cap5#TUV~?mwgJZ+6N&-Xi3DI5$_*xHn(l%6
zv{go4sRfcRht{#vA#;UHe2c)lZ<}F}xL0wDb<Nn5^pMxJ(h>c5wIPRW_0{2W)79si
zM-pJi5;cHS3#h(U*{fRD+UaY@;iu~O6py*JibumtGSrWsYQG@uSta)?`oE&^or$)K
zzESI`p(CmmG_K-B5kG{xPeAcno^tdHx(v>>mEk7U-0Yh3unQ|y(b8ufa(|}XCVD%{
z?ETi2n|xoaPzG01u^6L(Xid&LhOEBrsf=%$^*@U6mtEKSc`Y`Sgd`p`&A06(c)av@
z<~MuT{&Rr)&z>y=9^WWcb4x>RJ@ASvpTgrEJ7ybfuLNebt74+Akw#|zUzAwekK85>
z|Nfm<sVW$)Gbwy9Kd)u5+sC=UuKNilOq|ND-W0gTlr(W6Eyl3pFzUeVhC<^7it3N6
z3TAH%0}U^R{BhhT-V#|_$&UOUo=PlbNb~r_u2DYfC~-M#G8cZ0W8T;1PoxY4Qo8{c
zHH+IVh4NbApKRDLH6D}kguvX{Sbv?vQi+9SkuhMkLJ>RZTl}cJhO3ryM?wN#_9l#x
zpvR@nV^}gECIN%j#^|~h8Q$3%Y^)=tu8wW0u*Z{b22#SBUuRuNP=yt}+C&r9pBGG-
z_B<mCAqRk3MnHCwPDZDgGU#ry2ciMn=Zca*^Byeg*-Kt?$UH!xC8BLcqNQT1{khS%
zhQ22#K+Jh}86zR2$u5bCBSwvy?9i_J`7U+q6oZ6f_9J~fu*aK`*Vl^+4wl5=Q|2td
zH1BUBR>9*cNxjB$SL=iRY@;_6YUDbtb;eI-nj?QMbR$|xvKrg{n9sE`M~=oW5VBz`
zrhW33%G*ovw%4ywH|&%Ol-g9|_ee1ZnVi>Xt{W+7ZL&5mY;Au#G2B!u!YR*YGGsJV
z#p7Nyr&)kV3|x=&C>)-1Y0#Vq<w7LF?uw~*R_D)4b;Rl;obW6CEuj4-^}8#($00S`
zE~)ewWWD_4iIk3~A!SN~iNjZ?By~Jv%{il~+^BH!#!pwTdW}>4y77u_2HMGXAm^Kj
zGbk{$Ok!@O#O;&a_YxD5NnmNEQr*D56W@^J&F)D&ZtwFgcKmyo!Gpz_{?`X=pvAx1
z(Q$H5ildqjFei+>%o)6u2$vt!nwawCt3Pk~r97sZvAKLCUIu*7V14=If9L!SynF7)
zW*uT|R`cVbe&p36@5>CJ4U@RGjP|eMXGZ~S<tvOMvs$^4t*&t+DM+^NnBkl&Y6lHT
z*AQ}T&<g-TrIr~2xOZCi3@W>)#m8D`p1@B_uFiVR1>px+%BbMKd?JDaL7BuUCef9!
z^*YW%^TDbV(h!nUqu`S26ry*Mfa^?ZLe`Kd=fn|xfDGkS1_=QOiIH8XsT3pc3}+dX
z|GQ)pf?=B1#=0B5WmL%2$Qk(uxCkHs;q-$9w#)to<C13ZuO{D3u@pKI*(Jb2;R}U}
zFS&!G@Wy6~G{<oC2lZNO>LjB9I0l?o)PMve!6LJsjytHOc@cRAQj<sDDm?)v52XhH
z#sE6Fmlf;m+F@`4BrAk#?J$<GD4b%Njn`I-`@4JqrfS8%3*Xp>TKAw2ZO>I=M3Q_2
zA_BMy-<tnmfgwt}P>coJ_hfBg0zobH#y3G5W#cyaMC@Hkerh<K*YjDy#u)NV%t)~>
z-u^eZ<O0`oC_rvaaOz!7#afe`zEP^fk@Pv&E)6Tw6R$Jd1^&?-unH~B$Juyn_<v=Y
zgwu8FjVXSY1bNlFhk$cjQGFv~{=d$XXWX3yz03G78yY8C6w8B$4}A;4=IcHq@2gBA
z>UJFFj40a|@2}$H$Vi$SPcIAqz6fcp!b7C2$6v5SYGH&+2L;(UHUvM<62~qzPG5Ej
z(aKHNava#4%L`X4u?0q?&f2iuh!(CILwB`J9qPd~eTz3X-wYS$UZNrLCT-{wNCI=Z
zXlAyWO6p%Q+A&Zau;e%<TDu?VIT@O`IA$))R||`Anc3WdF-Yzteg_j;euQr@hFKPy
zV&Ay;SJm*z0?-PX-ge{9gygM}jcFVw2!ZwKbpK(aYvh|Q;2-5Qs1UbCN>8czr`+>0
ziIQvw=tw@RZaNkuqfQVb06t?~cXq;Gi-qp-GTN;ADpFYVgGggO<=X;$jgmc51)uMK
zeab?Qg2rLnsRh^Qe>P~seB&h!N6-5Nmh~DM(3yHLQ6a4^sx+DK?b`vDf^GnAmJ|?C
z;i4NVA)lB-`%G%syRkmB?vqsHS28{+fPhJYpX3*z#@grSY8}Ll_|uFVjpKI`PL>f!
z0HLRnb83v~*YWVvPuW$$A&giEh-9T+2rHKJN>O0sj1!TraR129*pEK@nzTVwn5xa8
z)1F?^P|{l2h*?)QB(pVeWr_=~Xl*|y*axxKdO)>@7D2O5+jF2s5F*7hCtbb2o^A8I
zHP}&ECNM5sxm{Hsc(egq#uC=d@g5i&yJJM55Dt}aESucNl~e2F<5T%oYCWhHtlRZP
zCM?irmJD(p=rBaHmd6uz2d!!N69{7YY~Cm9d~aPu+d-q&&3edj9?OgsQat?f`#lZh
znXy&y<1U4j^Rn=wYq4?TBe`W?KZJr@Lu)_SF1SgnQ}2K8M&S`w<5`z6A&sU6$9HG^
zO#n}08VH+Z7>l(oqKD&4zj{~fE3yxH^^;`7<n~FSz0O@jC31RA#Jx9A*Ru9$rNbdV
zx7TI^S`xZBql!!`J+HsiEzFDs;d~5=(jxoKXpid=yJj%l+wt$^N;zr2jbS$DY%#P5
zc5Tdb?Yms>`+nI=$?((wbNZJt^kW6;VgtZE=y-tx;b`)lzZ3LM3pASD`Tnl(GiZtW
zS;IK9f#BD0nG$%K6qxHRCyvZW02QvO6W(->2_Qkr5dR)sSpA1-D%eEF?1!p-mnGp6
z(?(C4ID<JlZ^R;1-)Wwc5VYgek%Hj>eqbrOP#BoV0z_jx035<uet2hyzmOkjfl%dP
zQ8&AErC_9+#DT;KmWwqJlo0w=G@<9HV1G|t8oE+G_0j-s6R<gjSHlDi_lM3>=$nqi
z!l1;kV5Ct#O_W>`_E4^pvQt9Wc&|<ERx|=6p*WC=*KD2B=K5th@o&ot!zcHC4k=L*
zgSCDwchlZKEJbL%2Sy(QxZ+Exbks-`v46URRCoXwcRnYx78D>ag1Qu{0F0?VQ$$z#
zP440<VsydRMg+!^PQU^mKIf021Pcrb64e@eGAj|@Lvt?KLe8@WZA=|wvunmyFBJ{(
zQ+hU89i`5P#!f<547G<?Uu=l`HB}m-m~~Im@M@SB-qHw3?&#HNix8WSiJ$kSQ9Ueu
z=O@?bxftqZ^9b{2VQ!WbxpH**hBC+k5e?X^^xW%(tLv@EWoGiG-0lvC5Us&#B_WIp
z>9a)xDcmHbG~2G-aq8M}0OKv-`9)*FMU_4mKKAW=er~lt*MC!KY&fe%xOqlD-c>Z6
zmoL`c2PtG?7RjC_C5d7nYS^5pbydJzE|yz+q+(`o^bBw&bBRxKY*a=9O$6eiNJH|D
zBf&sk^AG?Z4S<+ByFhaffXxEt-^9k|g?qu#`cO`2h>NL|`xxJwhW%f!u9d*_>b0)`
zfJ!Vo=<ZO8iM=3^06J1Cv@6(XL^k^q$(3ph8|Vh)Xr(0&V1-ojzd2FE(Kk}ffyraN
z&zNVZmx=(`Jpg@5Bk|7@-awVg5%$dn=;7&;c+|~+2#M%^>7|@(KAe_P>T!?|3n1QW
zy=`N<vFWkTL~~t@nm~Gm9w74BtZ?%;@yG}1o&F3Ug$5_VR+`4WUTT2bOrH1;I-CK!
z3rr*k0OIB6&I9MaL3#3|<nOdE;Rk5#9|0#=xFoT|r50~>T^4MG%r4E+fB2S9l=O}9
zjffotHDvDGXqUN{uSzP%J8XMXS|8RdN~~Ou(SAQSO`pf+P13g>&9F13T8On!ZWkK4
zHvAfJ$S1%3L&noYNZ2`?U;f<h*wNhyHM%^fqCG5h$Pp@@U3GdKA}<){J0h+w7`9%(
zBm)Cr4>sEjU}1>IT1{#{D-)1spDNGvD?YLN8~7vc_dC&nD9~l*ffU8fP;Xz$`je<7
z(~|_7m1^RL>c;Hpg9a9>C_OCC8a)S{s>kweYVif<;WoDlki60XLdm&@V=j8mANg?J
zE|qVfjQ{4FR@XPpgvFdRE>7myyGGsrmF*tR4iy>?S)Ji$r-jt>!l~t@8<9?bYdkIt
z9jV^F9Z&LljV0ut3~OL!gFSSx>#jSqnW4@XS*a${WRS1nJZUP)&Pfe*-)lHH)R7A(
zDvB+1rINd~pHgA4sK&X-G|`)f8GtUV%lZ^;VbPuRvH%DTZ-E@=EtA}b8ql>&!Zuvj
z?21gC1e@8g+N!@KO}}Ao0{}kKfrOkmiF1@&XO&uiy79fuGX5zH@F;}6Z{Xf1swv3Z
z_OJy*Kmj1h0>x|0$^3YU0CXE{?(bM?E~d6+@!VXO5v_c;PxaU9Qc1@8ZF^klhFx5$
z!kvW7F(hnf;|i2Z6e9s=MN=J9*-OnVLRa78;>c^Ms9e`2#)mX>zjE%3+sQ^#`pJkW
z70Och`y`*Vw>jCgJ_KX>RV1<k4iYkl4m+&K6Oe*?v;cDQz4i?e(mafa@vj*ya3*w%
z7lg*>^O4R#quE@bBx%#o><>jg>$EXd2vPwltr?t3M?=2()slTn@3!FS+2biN_8FXb
z&J3eZBN%_tEhr|*S<U%<vr()^v`|rSWm#@G_)?oQ?e^D&%ibxQetF+}DIuIqZd0@v
z`rekqEXLm*e|tn8zED_x^fOdk;Nv;9Jr3RZtxg9S{wX&ctuWqkp9PesynR%<3l=H%
zy^D{7s1IWw_id#b`8&Qv5QhH9Ez7G{WWyPdYH@)mtQdK5tYB%=ulm&A=Xe#(RU*1R
zKVcY6#<nJ()jULgY8Stgj|VY6wmLR<GM+5`5j|+4={76)Eu?zt)}?nP#^=+>VD@4J
zW9%(olh~Aj3e{>@AV;0)JOj&?;g%B4z)(!EtL{M5het~tj}h7}4nkLv%hgyv+V;a%
z_zCUxQDXkW#qaoiHuJ&-(p=wLax&7-y2<;m)$Goz+qE^jT)JdM`MSkPzuA}EjbR^3
zsTvh<2=!jib{kSy4o0k{m@u%leKelk_R->#-_m*tj+#-r7@XXj{$0At3!YI7^|fTe
zRA9wHuQ4F2p+{!|ya%|MH{ctP0HFjJx<g@8AHotfEke<cqmybWhUzOJ)W8DSTK>M6
z8VEH=4c7qrsx6666JTlnS%Jl}l)zX<C8_R|0tuWRS*--WGh^c^r(>*cvCQW9l0h6r
zR3nb{LxNLSYd*pX{{>DPx8H0&K%=b2lykBp9<3aD7Gt0iai{^=1SPz+R*+gvMLsu>
z?8S<oB1h6EwN4P-H0VV4N}CX)$+OqL|1tCJ>oyY?S|}DoN*GS^H%N7%m~|c#sadA=
zEkuyCrWZ3z*w=jIO|+wcsql1Z1}77eXpxy$3DD8RCNx%AEdWEdgj9weaaErc!qO0>
zTo7JG$s*n5r{qkGa}dI7d4Gm(!Er-+d_O96=5>1p&SB<ReN4dBWAEv)=e1F$k(M7?
z+{XOKM)ZYdSO8aNBO2sD_Dw)%oZ2V8s{Pnob`&jnoJ|4T`6NcVSDk8Gp-WRDF7{I-
zkfY7{?b6HEm8o~%f1KKxKQm5SclEsRG;(Jii0Yyl97HoOoPZqs5IhLiqoQc1G-RDF
z>uNq%+fPCGxkmc-v7sUUCT&c+ti6iE#a>IQO<!4a=$dHv`~Df5*aRL@dX)OE>#3kF
z@>_nh@{4}R0UD#`N#nKf!T_e3@Gcy2+_wWw5Ph~g*oJ#v!v;XdRUmi|u<QwKz}W=w
z0;r+*Pz-l5w>Y^myj9uf!_36;0gJ41K=bX$Zh}H!KZmCIl~Mp%uPi?R9q^-aSQ14l
z2SA`1t@wARXx+rn&kbGW@SPYjeiQe4uqV13I@##Hk3^bVCs+ZfUF7kiE_JNoEfu&{
zk!(i|*BreG(_PwaN$w2>M!%>OtyvSufrJEoZF#}n(-v$eoeh%KLuUuPdcT&T=Cuy1
zDNVSPaz+e$lXm~OZ8iYfDk*cg$klm>M|e~7_6>4L(Sp{=CmWHm;9A)M7Wr9HuwCRO
zeNO!TYULsi?P^Hb<H4T=0LloeC@ZBCCdTdcntrm76JS;OYTJh7+eaY<Wmp43chyso
zIX)MNz-0|qF2R*zQK0!?BzQMFB|BKZ4Tc<e4fdWzsWq>0oCp>a*uW689vd?pcjq4S
z)5a)QRMz2xbNiMaC+Ck=-OeYAN|!m^eRusgCrbyrz80S=e_FnCA+s5G2o@-M@1$+?
zvYwQ<GR}IO!uJyxaZt55SSSHR5Xi)C9??QFAzx-Y+BpXl-V7)tt*bTu5*RhIDt0aY
z(@p4m)`_`#YMSgpn)9XqXB@x6j0x-MoA+ZU;z5IDKG9UCJZDkb7?2RPq66h#rDb-o
zn>Jbw%5df<oy_9<Wa9gBa{Rtu-K~A*BDliP4th$-vZjqN<CV`?9VlvP`;SpQR|5R|
zj2CuVYo&pC1EA(jSLJW{G=l76&tr(fWMJ&Ty=*Zunu>(T`@;#)-ss@@8cY3)XBNZr
z1hSyx60hoYYe7*)Ejd1oJ`lZvh@_Gs6rO`dLQo3-Lo32N+uKvKNKWc&jKM;HChLsP
zfnO`GzzAYBkEu-A;jTC*e2v`JSHX3BTM6Nf>&Ir5$*B=l00X(3=Fa?Dzc6(3F1d}$
zQIqm<D!?hhAw(?)?5yvwA20xhn3I~fUkzw^Wjfk9?i8~KGIaE4;IeXan8e{9aRek*
z-HYTVy^~tIMQaQ5jy)7HUTImxr_8b(=P1J28ZmD)pDDt2*xCz)_V+&CdHo&wywc<;
z9i){#P5SJCc)Jt2X;%hfA0K_s%SQ3)ZBQjI=44_@CZGIy%lQaNQo$QPOnJ%&ctz22
z??tT3b)UibD}u&D5z-&R7pDcd)p81#yeM@gt>Zxz4WE409ODID*E-}wv&WWLXGhcF
z07lrg%liFDAnCx6SFNaz&iamvZdZft3wyi`hit<X^<=Twy9>R+GSTKO+K+_Kc<Z>K
zR__&K1IJSJmQov89oposLw_DokBw1e-<K^2CN$S_9Qk?ER;#by_lbSA4M~(|lQDKD
zwU`(7E5QU;|I9;hLIOk&zBIn;l(WpK89{HB7%%2c{XTWyyd<<@SUq#C+`?0(;#QA`
zgrrrh+;{R5+FbTD)rRQk8(Em59KK1o{XwXYjSbFU_8e_$d(absbYbG|!T=)Bf_C>F
z03b#1MusIWM(R5>4Qu!VuR-twiv*?<k%>ol0xPyGa<j|HJW3M0oRH1xSp+8TUnI2z
zNuj$zSib^xThhvB@p&aHpND=U)(FxKS+a=4WhY>_SG~q@!ejcQf}6nzXFTHXa&xhR
zN#AatDxE>m#Af5wjBZ2dDQwZ}5~6&i&ue8DBNO>c7hk}iGsSX*z9toi3c!h;=F$Bj
zb&^5s0A-9OtRVX0kt_p_cY_kaSf_RJ+YC77g*r*W4fwCRmEI=n556rDkYtb~10j<W
zFg2p<`s9Er47o$0fK*LBX@G5lz3lBMC@D%CCbbN<qbJABbPstUz^*(-SfPE{OT=DY
zO?rh%a%1{{@GK(Dxf%JBrCp`>>BCvAmg774ihYSXRlW7g0fOBNtx1940PDc|A2v6G
zXpGE!Y3BHsF!Le*S;DT}ZPy)fGxdO~?cq^jIRVk0{WtFntx)qCTW6t`gYj)NVm<?3
zUV}f$e;L=-0B5(Yj=Scm_RRW0W8FEN$A;_Gdb$^E1<6;-of1V|!@08K0uoveRzo4C
zEo06Ah=?0({baJ|8}s-MP$R>WrF96cM#9Z&zKXr&{%@|S%|wi4{G6E2eD*hTp4B>P
z7zb=CVjq>m=MHj{kk2hW$_%bGkhR<foa#JCEVq>LFaBJc7c~C}=&p3x_VkmEh$#fP
zPJKGYy<5{DHALr8J2aLFZAe(cbjVJg8d68!3Oyg^9&6)kr4rG7lk%0l>w5iuZOQfv
zh4|{<v)_57U_-ngKSrIkYC^vqV2GO3Wb-68MEh|GHpJ?gRjj#unb-<r9hD#e_&Bs>
zEoXYE6YAg!8|PripT3c(FfaS#u6j5Zj=9B);Y#&PGR%L}hl6fn;rK8iJD1iw79`N&
zSezt@i$rKr439(4qyW#bi9hCTpiFW0I)7NrA;>6XsjCwag%6idR1*^f$^>B@Xwq2A
z^#%<A#J;|4W=+GceFeJ0-ULcw@8cj<o{>d98VeMMU9o!yKy8D-EK0~@05$ztp5&=i
zE0VWc1yq$wrXiK^mmcP0=lDkwwV`41e!<8}j{DBi@jX~$k{gaU2GLRkOtg5t<@JmI
z^XXs9uR59sNwX9xO%Yf{!a#Ez&X@)rN640l-e+6`K>$<BAiHv|^_*DkTDyYAy(pft
z)w{O0U$%A0X<V8C#hk1is}8Daq^U35&@f7+x0ZkWVE<G-_u0e&m~sQg!GW`qx<xrU
z&MEnLY7YcTOIq8Tj-`g2R{g%ri}o%a__7}<7kAt$auSmG3d$}lZ39*Cs$LxlkU(?K
zs{WdMZ73k>z;e^3k8!s36WA(k1s{g;@mO?(FvIih)hsVp|A3o)udh@aOIl+VQ(6Z?
zUUs@a%v&E*e*V0AV^gkaI;i&cOQIfC@wWp14W?Gvzc_7ICvUc|PS>2zeiRBB=UAS{
z$(~2hUiBhHzl$ZhQR(B0Mcz6X^7MDKPLJ?ox!$zT9p&2E>PGCffa4TB-<vgMuQb?p
zWqUMMIGYeS9%zlu`JM+w)KoPMitke}h7}Z~ZXUJNFyGwG`xf;J9FbEG9-P~iZTqla
zVf!qKw>Rmo(LMT=P_vy<T!Xg`<Hh73<#!Ld9@ShD18ORLEu?bM``AbB#8*_zD1*M-
z`Fb~0HTeI#j89W+i+8`x_#VEoB%y9X<31Sp_}a<ut=DtNK?BG}a4EMOUBS_I#a-B3
zKv+5rPJn-tge!It8Q)4H-I(6E<=IaA=+^UIn1Lp4d+7T+gRa6z;m<FQ9$Flil&HG!
z;S5X@5>nU7MscP%5c^#ThZ3`W7d5zX<|dKCOaG^457T>Vss+GA51_iX81xJCn$2-0
zot%_4n^YpPHjB?mgX2CKH>_$E-{PQ?x^WdRVsy5BBNjzvf3a(m)^~~&Ov#|-CB|x~
z)*;i*FKr=i2@EYpO1RXfaX!j#It;_!Asq-b0FqkBoLMJ=l8-kO%JndBtUg|SV^TMG
z8;}zlCW6!oQ3V>)o0q$3aqKcL)WIBYTA;n<3z!s`I#8hqKGhT&#xAe}d6;DtNe#5+
zM;QMp7(qlK<TYWa_2xB!`FHETfDFJsbRCN`o!P{tjHi%3grEXMD-Yl)BnMz8^^28l
z0|B*pf2V1pA1MiCKDZSK;uC!Bs1fXWwu=EL=w48q{r=Ta@-rE%+)dm(9MClR<bLT#
zc%=#rHDKrkiISnSf&of~m84stxE$PA3yHE+vle4C2lcEu-ws!!mg`y=Q#07Nfh1vk
z0$gu8EX@1Z9e%0{oHaJ=)>ZRVRjK8OX--FWaT`wGEU<|zui?eM@4@&ZKj^*dRN)Vp
zme^fci&R_KD2-3R9~jCAor3-AuMwl|U5q)x9Eg*-$8Q&y5bd-#XBwBB9@brL{7Xh|
z$%JpI+-_B>%i}4u7<!T`wvfaaXl8Spb9_|v(Arz?;x$;f!Y$7#NyFx%(ycp@l~K~`
z@twjJa_&9goQyDV_oEU|W9B7V@L8^(89(cJW!O8|gGl`C{(WPpb8mFpZ$){7+VC(j
zz)iMRcQNgf3uyV#{20{)sq-mHEn0nOt6Kn^QrzZ#v!!^uK2R<1oDfOA9qO~nlQH@(
zJ*Rq#hyBE=y<-K{A^TVDW&SYt$v@)qObY*l&Wn3{XSp>7ZxcATFxHO4nSHjtd3gWG
zN*+O4)!c)n^iVSn{Wn`Tr^c$n^M}nN#Vq@O@0a^J^qhmt3)3P*Jta`dW}H;uzDTQ5
zi~#OApFrWW8cah`#)o;AaK|u?s;@9O0bor4;<oS+hsDxm$PUjFZ88+FA2OBmpxfFw
z&2F9(Nk|AMs+%7*7s<1%>sjGLT)>~hL6CjEAH9cgKVUID`-L5T%VrNVN%SHMoH?xJ
zx7KUi3}y+vNNe~ST#S<$aUJfA-NGtqZVE>x03y!E5R&Ar+eDB~mUP2j*lz$6H1aXS
z1P&SM^Ksg1n3MaUxO<8-9|KPjtm2mDX(<oQ1Zkj4hH!IAwr{^R`<o2kR|aVyL>N$h
zJOpSM2M!^I0B=Z`yr8$5@aZd4;%ZI+`tESXp@HrifVv@)qt+cLTjM@STG#C%WecDP
z!KsGI$Ko@btpx&Y@|EFif7?sBPSj3Px*BLwBz7AeRs>uTKMO3AN0f1s>h5y4l}6Fe
zo`!YM)@%^(BOFVfzjZ&0QXy>zvpdSaO|cgmz>0{dTx3|&`M}k=gBo!WzT93pO-cMZ
zV~IF36ENPbV175}(N00$42?h8y2qr-41W<EvKq`~@|na~01s}j)AZ83uX)2R=jT8B
z3AJfz36n14`$l-8%kW&aPr#g*bj3C6b)GFkMO9@<8eT;i*s1-?qy2R1^<4H|ZADtH
z`O130f`sp>pY6hkQF7v(+GJO;TRtwP&F(72UBsDH*3{R0G_*ad<!xqd&(nFcH9POe
z-^FH5v*Nt<?~Fc7rvjnVnDRn@r)zWk5#3d|*i~IaJuyRd+-5w)AERe{+h{;<mvfTz
z`1+^FR=AWoF6+zvrB1by^vs()zBe$CThFlUR?VAM1d^SmIyAbofBy7N=G}mdQ6jB1
zqQ(JKZiAAmjYx=)dM%tvcsK*(i?_dfm2Jg8OX$;s;2;3#^l?K5pNvcB`5?S)_#^xH
z1iTC^;E+KJ2(krg*RBr0YeDo3;S-~|Av5Sk;J}Sl9iWa*5+5f3Cq-zHfS)x5Aji&K
zjf-0bXDFEQXl3qwYC~Mh{K6*2V_|F}L=MG-ihdg|GSnbHPJlnajDgHyq%1|g?Sh4K
z!f<o|Ax6A{51K2B+^CrWe=$~F_86}eItk(Tu#I`1ZEcFA5Zv<Er&qS!Ge8ye*GqoG
zZ#c!jzBy#t#S%qFm$qeu(}N_URHglpOqTmPp*_D?j`}o1%hiqYytm(>_i1+aanJ|j
zbpL!trjARrOpgN9yxm6xQlpf~9l%u(eCxF}z}y@|)$xOTSBCFdrBApm>HO{Bw}_Ex
zakU)PJ$B>iKzdHA=8(i1&YaaxU5uhjKHnPV5b-xRy&AGHbpno;Vu!G=G#8gf{8sK=
zY3o}M#><8PgUW?9qpH=D*?n)3m)|nfiP!KM)=O8{W(77RJ31{|27rw2b+?ayg5+Ql
z3C(?<m)p0Zlezw1&0Ips>!Gqv=p)+^_VKb-_qr7mC{ISdw^m2SSM@7L@~oYdE2=z_
z&tqeX*hSr1b5?IoQ{cg&??UY*uGGr`1%Z?1c3fPfdw;2um)fjL>tWe<Li0ksAjDG~
zKh$w<iDS=@!1DtpJkgb=ZN{4hVn4e8V-%Ke4sx>f?rr4H|C#R;d7oK%H~;NL!1D%F
zjIpbtj<mV9`yH`@kl%va#Dwysn{P=Vf2MDStDk7rBaX1?CPst5buv!wbx7}!a{PJI
zK^w%06FkAj4d%`f+WIhbGQ$MN%!4D&ofP{Mk=`U@nDHs#LO>+~+m7wm-Lr!41~07p
z2ayDy@`DYL$kOG37Iks-&n%Eve9E9Jh5l0Lq`49&h5?=wz`eB-`o3+Ej>*IYh(su0
zSe1x_T6(YGskJOkQ@=|QQsr+cPBJJP>%u^EZoi7K!K?^uRxB$)pJyUE>kt7D)!$wd
zld9<QH?=`Mci)*7bY$Sn(xgg<0+l!>wuH?ErFU*@Lbz}y3@nL}DeI{WT93CFi(k2W
zju!zDR&OYhf5_{+C{FR4du9+?J(c_@>@!^R17ESJZn?l2wBD=KJ+6o9*kJ9y3bs;{
zx@TDnJw%<3(sZpR*25}K022jLmBEU>D9&C0*EC<A4t3Zl_SKVBd58pW5~|YGu|{=j
zS;HI&5p^g0pqJ;*=i+d*lwHR`jBESKWX`+R@movwmw(G>Hge(sm*atGw{tCVf9q*Y
zKVCG*@cg`+$<49xBC2lBxwrTDPpQ3@EjYRV@MSlY&q(YVTHW9_-oYU3tA@nF9S)yP
z$|ei5;&}0W(~jf1)--7JVo_K9OF^~W5nFQlYd;Hyi}?Kvjw%-c`6L!>zjuv^JI}Lc
zKP+RZsPM<^TPInijA^C0%%;vi`3w^WO%`yREYG==9G>hd7p6XQ5k3u9GfjTy@ZyxR
zcuo6rtUw<O{%nxrfSG1nSFd{fcq9Qx$T13ILlR8I_GB>dBZvuT2G9lqmNw()rQBR3
zu>euJJ7&ZIFolTw&{f6Meh*&xXv9_Dkogz%`j6O%RC>#Qp(dh+Mf^X=NwI>^AR56C
zq6U2kDTMJVOueO^|ECUqWdteUS*=h4u<4`kAMo!QRFxOseP<9!|9_Zz%Z4bwxBGi$
z=#p;f?iQp$QluM(kPxN2V*u%)B@_v1X^_sLOFE^yyN2dJ-`{;bxZlA%*t5@lto2#z
zpl>kZP;pHzh(?Q0_$!o(riV(DkAJ5(gA0lO7)-+Gb{=*si(~!^b^_rO;MC|p#Bc_}
zEH!AXNhsxf67tvsqYfAcT5N{#Vvxgetd$^~ymkOtz}xwxoyg(b_O7G!1hFz}pc)h3
zTZP`g)i)_<;aGQ}=0^)@raNi_GienlqU_}8OaYkhC<y0~39Y@!8s+)vS`NYs<d<vZ
zo&APDyg155(OOC5ypwym#O?{H8`#E%4T;l%;6-I0%C{>B)n#J#cZ%?dgpmr;w=c6P
zdgaZ2+x}c_*Yv!T73VCac1H33*C5;L>6DZu<!fFRlL^_-C|RL*thhBygFk!E5h9Dj
zt8Dqja{rEk!Q~SB@_`xK#dd#ctFs!x8i#8ipddN9$o|?y5Jwk)6`huCED9qh-3Zyt
zY^F6(*6peM$Hma!wd5!ssCbcQyvgy&Lt-yegPnct@`(NZy^o?54iqddbohK#3JcRq
z;V3UXZT5Lu)^8gkiss-tMf#IU2}Jc>x|2RUCz@_7m6ae<4Zqejn^cq#)Q~)qph7px
z9_F{gJ~S{iwssTfH>hS?UAwTXw>kY%QCn{tJjlliWh0{_6g>D|h72-IB$nxQu(nnX
zPr)ZgcNXY5_z;(z5WJa>awnn~u1W3~H(bR_$DpEFAtwHZzCx-t3Yz!JJksT^47iZ+
zKt6e!^VqI}mWSker}=LI<LQ@pK-Kf_(}4(XC9F%M@)9nEvU5$<vuNaT;%kkU?Zgc`
zS)0eA#1wRL;pM3rk4vV>Q4GgQ({G;Y;*)*9KDm>$g%4ySk6ut45ODc^C=JZr)Dngy
z?vX8ic)0|J+x-y~KTxo4G)v&%8l#0RPbOf)2{#!W70Qsvs2MGyc#m-`AjmSU7?`gX
zKN@jHF;$a5+i2k*USk2U00i^&@})GP;-?I5_)PVENo1^KoBBH8;=`Duag;0<Tx@3J
zCnd00+Ii9AJGmuXSfbF3dMja`I1K$aqZNLQ^p#7sVrngLI^jG>uA;&Hb(OXBsZ2V3
zevK9F_fn2@9*`|3py_}unhwFm1M30Eog2KuW~oJ!%x<9jXbz6ScTcEiFh-!XmQ<S^
z#{yvKmn0++WUh{)Bx?`L`xU*?^a|@r{R{XqhCHNv7-cR2Dg=@5p^?eXPa0a!!onL0
zjFAB;A<#_ZoL&#+TVJh^*sh<S4FcAGsOv4AGj5;uotM`GfK<#t1~i-9<$Z0!WeqpQ
z1!y}%jlG~fWMn93Iu<S~aPpB=pu~&qw`Sjf-rXCemWK4hvazuihs7UW#Od2@MouQr
zEzK*B%LDp${uLhW<lh&s!527!d0w~mnc5Ys-S(zemEUz$Tp+dqga~V5`H{QK!|899
z$Vp-0lx6vBNnYZmVUU2(r02?o0;2nnF5B#^#_5Hw!Rz!S)~P$J_pPwxoBdCupO%}!
zQGBWK>K{H=>mWkMWtF6!isvikrPsV~Z6&-+Z2-#i4At<&u1nH|&Yc1HmDYBKjh?+<
zJBPxxlkBQc?dDYE{Ppm!=EqwV3Cnx#E>rS~Y}b!($&Hl<WDN;!<n7~DW(TVsM=879
zB&?!TW^sG@sHoWl#aW_Dd15SM+Z#T4j1<=M?(KRk6r1`t85y5k_Wt|+H#o6}j6oAI
zct8xZ;a4;HzE59wavnd$56vgY18Vc@FMP$v2!FI59O<N%X24YzdLDJ!ncJq}jdtno
z>gQb{Bvk!mzs6iVu9LVgm^tpjWO$acFmzNmK#~twnVdn6(3MYt1bvk24A|XnM$TLJ
zt|AxN=mXnI@9Ufv0MGyxi*7u`n;Jk5V66H;(rd~(kV)w9msJ537PC&`>SGN-4-ZmZ
z!WepWTT`-0f(>@?Q$U|`97SR!0{{~UqU{LEK98Fqi_m@Qt&t`jTUjHsR^k}0;_Hb<
z6rlo=)(4f{+T_hIlQukHc!rXJOE`m1B(UC`z{+(@6&m`1X8JeSf0L|%N~SEcXoH|g
zt7>`nlcrgeDq7(ulvZie(sxG$A$7+7cr#cJp>otnzB-QOU7Uzu0ye=FFj5YTS04%z
z!Hy*O(`Ueugi-hkN(Abm%#c56*%Px?RRF3t&TTMY!mS!FL0`A-8CaUpL_XpVkgfWb
z@a+3c{|g+_Ktimj4p-f(DS2)*eW(%4V;H(3XXx*3RC$y1CU0&*ztqdEc9?9Ug<`@Y
zG&(Jesg3Z7{;&8|(G2lz-F?E1y?3S{X5*}_K+NSAR~yZsjyV6%5QKQi1{?UXbJ1-P
zcG=F*7dak`BTYv$S#O330a{+zaj!;+v>cym`WX%U_x+A{QKUB2F;o8@m?e!wO|(sw
zC}hcqPqtsrQ62!N^0Ybrm+6hmiKR@Pgdeh2y{vYflX0BDHx??}!s2^8zN=UH2XQQd
zLKg2@hC4)RY~T9h7^=#$V?9Q??J6$HH_>-F*?SX5BSecdsniAsp4g9jZl1aowa5?v
z_@@;gvVwa87811;6AMWM`A)Bax%?z&8A_8kdyIVW11rea@*A!J6POtBr_8!uL$CN@
z9)Qf9<^F_1WDaXUuFCc0yRiJrqM7=*Q-R5hhK-&{qmv-N5#e2r&RvDcZr3@&;ZWup
zv`<Z+U56_|yVHIDethKzqTawh7OG~D@h+E%(Eqgt$lOXJr+X-h(6+<;4&Z|tvq0J?
z9e}cA*m(nyVR)O!XI`}TTrptYah?1~Vr~NX3S=POL8HZi5U!zPgky7;1LDl%@bbRi
z0K}a`w}Z5$wv`%jGIUnssk7=lKjs?*aqt720j+2Au`v_xUK0=;CHeqOdnD;P9WkT{
zEnZ-*iDFzSOnpi;WPHR0&%f(vS?ZkktZ2wE*etP*PNduso)s768(D$Vc0Do%W9g^a
zFeBeX1mkGDzTkfN&_KR?j#FFG)agxUcA~N%G55jGv-!@%;E<wl(4^G<#s3s>B=F3w
z96yeNzodM!Sx#@7WN&GzlWFkFUAQ(my(;rjeiHVa-jux*a%@Zb&r+Z8Odr%%e-6wV
zp;)7u5HjzX8WI)4_<^t0s!GI05QsW_FDxZo1j&!N2#V`u;{EGF{SM~^@E%h(t*EeY
zIbSc%@%mel4BaDSQ`qSQCgZhox4XcEK1JBjCA2=8!Irk<sVV8lCVxu$f{YSp^y5@n
z0`*ppx$$~>6?%2`(jMBs)DssKTV43wS1pBq-EJTs=cyAj3Q1ucqco_nZxd-Z>~N^`
zx$88ib!cx}Hv5v>9J-ck`++-i-oiYhT9XJM{fH}#3kG8VHO$X6d#Ug_I@9rt(wTO)
zbLX*9_}I|L9kCH|NHF8H4R8<z1SkW1H7LVdM&{+?y?FSVxYEB4z=LJZB6cwWut`dh
z8QO+$V%mV-T6j5-7FGbhbQ0`1CH?gP``o)QhvdG!9C3#V!blTiw{l&eLu%R%6JkY=
z=Ad#YMTUNwoFXYzkhLjykDPwECZZHcumrN_+Jxe4?tuBzVCpBx6R|A{hO&jq3cRB2
zbPVSE(@OhEnUxL~8$X=}qB0CJWbQ37#4|=?#uV+2CCQnF95D9{gg`h-vwic0Y4F-}
zd^tbGyh7r}CqzyA9%|dYrA0X#R1s`d$kVvKAu13KOjx799!G7kLXJtBRDH43n!}ru
zrOAH!RtA(BQytY@{f@}FW@oAOg8v&MlJ)!0&FN1lz2Y0@ymeI&dF<?HxDm^}zwGM4
zREm?ws%gv4QTy|!27jde@?x(`r3x~J<2V7tMI_gOl`-QWHiUQ`#@!f||IvLgpw7;C
zi2K`!5RL0feoSAe_;M=>32&Q%Czm`7+`Puf-qu%5-rQ)19gYvy`b9Wt_8IIHefiK}
z$ffjilD<ltY1`-MU|Q%ik<KYoi-Lsc*8fi8Lv>6B@(hE2o=>AUgHfCB9t?;kmOp`<
zY-Q`aa6I3neZmN`RBrRIx^@9O`MKO3oS&x8$pD2#-I_m547!4){LaZxKblHsI<KbQ
z)Y$8IO+17JCaC=o3|&3G&FIH)WO-P9m06!RDu^d=LZNWV#zR=~DoQGZuAbIxbiHRQ
zNNG15+0-w)bp5jYD8lR>PBT}+P(DsG+miErEnvObDDgL;b<-OGl|zsi=uIc)S_DC#
zzriaA4dUqq09X+jY&JkZfaDtJ&ChNRA}Ic;>=1xxhL2y(Cm8cFnQx_KlcH0V7&Qn6
z#J&T)1qk&OA|a8Vt6HVuf~_$g&Cg{G`~M0N{l(7)&?=QI&c7mN4zD)Vc}b9t(mm)S
zP9>2vzby&^2-|^*(U8N-*iX>^CjJ^y1fDA>Wf@wqW(DGY=9Ye2nk+iK>v7LXf)UV<
zLds8%CNmGD0%!sx&6CI9U?{|jmSUozs1c#aB116lQCaDpl|FuuPF@0QzN-3NCJBIk
zvy#h~T#z<<Ho~W|HiRU{f0WIErzE9|CB=vWB=k_MV%15PTMZ;@g}6<4hMx2^X|*ug
z93bkffiOVo$0h|r#ZO3N=WcHgl?a230$9RGHE(w=E$|xQD@Z3tf5y%%gxZdg7zRs*
zyB|djn{TqC{NiVAiOn9$a+tO|w^eU_GBwLsCf_tkf3YbkXdq;XYuevmD9frkI{wzM
z(tOFR-sTFoXwSd1ZFq3~?AUoc3obJ$?M6(xtp{K2CO?Q5I+_jZEjW2N&c0t<<~$6>
z52>u=m~S>7x5$oCb}JClHubXk?4rY>CEUMJTVE}#daziSy8Jt>-SAIi)@8J+cF(oZ
zwjWQ7sF<PRWP{E^+jxn;F;M*SZ=n4MRV?46s=g5x1yKfy+=e4XI-}p1<RWp*ab{!0
zqT_Dt@#5g+*S_KO>o=;E%Ap@-`2~%#<0Ga%H<ev_T?Q&=q}wcy1qO`R(c;|3wS<3x
z6YVDH75C4~v`!b94(^FWN&<hR5^I00WzKtw#NI9CZ^gwxXB|w4k14u)?~esbN+Lrc
zL{vkV%Yb?`NFHK?j6*z)%kbki7nEV>eLm3VZ^6BgV46*TLsb8b3~XaOmjm)D^V2GM
zbHs@v8<uhHJfdB{U<Z)(ASXg0+%ikC0DHQu!4OCoeiIrJ1So=lpdi;EzpuTxJ?M1h
zk*e>FX3&uP!syB?x`^j#W9Z))FuGb7h}xV=9rbyh);^P9hGKeCSILT3F#t!1zuUB(
ziTxb!E(Wl|=<i$EGxJmkPHLPVh<QD>vQ(^D>SLP=P*4<zlp;(gaM&8PToqYTF`l~R
zNn{S%umm711`Oz;s5$f`?={V_1`jjt0riE52CphRdH%%cW;Mh+>+`=~w}=+56!T|B
ziFm!@QK<>P%2A7`e+ZAF`q!(hJtT&I>|)%evUwcaH7Uh+3qO3|n#|5BY-@Fjv49af
zPF6z|r)v;H`OpN<Y72!@M&*rTrL9WV&C!-2%NN(V6AvlHO2qfn*}&0Rf6L+PTlTqS
z0R>b=MPaUr!<0Y=rnhZ;qUr%dh=)~t)&@z`jCMu#Zayj*M{^(9TQ-G<8|P#Xym0gi
z{)Gh3lZEZu24Vu5Xe6WHXv&PFBhHbs$Q4XqbU2+861?nlyteE5MuX|B8hMDn7(k($
zzYnL%2n<&dmMlX!Tlaz9Pab|5!ahk?nku6ro=S)a=T-<`5{Z(~Z}q3mbBeV)qbW3$
zqjWln!<e4G8lWRu2eMNR1s_Xq^yxE1!5dy8A<O$ut^M}6#j>K@@DC_JLZ+rB3cOH|
z1^^E!A@u9Vm5on$^Cs53YdpgMsj|x|E`SaI(_HZunarXr9#eqP`INY^Ki?6CJc?+?
zg|3B^(9WWpqr=eG21^Wx7)155aLHJ=$$NJ+jI@vNPcytUJ}^NNh0>Wpis421G<<UJ
z1uj#CoihtyhaeK>dXBiLGkgQ?+V~7E2)>(cL3{GvA3m9UTxoJ?w3w(|{;E$(S1HK6
zva&BdxWKiXc9&qH9<<=tkOyA3%z_VBemXD=n9|xPR)7-jP(z{XbcE!%@Zd((yy+R5
zboiDGxy1<QU@U1fYI-mqKWom{_))#!d!x<ww3-*I#fdF5_esXKk011E?B&h3U{P<q
z`rPk_M)X!LW6jF@-dlM%P}+I8oqTea`wo~U`n-tH-Z^v?7cih(+bdDW-f6hAr5Q*9
zZjLP^nnVoP8*0(<bmcwM3YWoZyw8Wn#y^6~OrHwtQUrH2gsTB9*Cy%A0=r$VHoqc_
zj<3hO6p_T!P<ja%KUouz8RCaBjs)m<F|OQ~*HB1tOMaG+a4N8uxPA~vERN?=a4o-V
zKT)%9`S=GK7$h_~c19*%>U8{@WGB}@qGB&k%(mrKLn5QGfm16(JtwcLEKmz~B`7F~
znqN$k{mN&|(8f?a=`Z`Smb^=-Y?|D@a7J#h-hOqUwaaz$_BT{QHIK_c40UJ1o<o|r
z&pBD7a`{TbQd7l^E_1xcSsP&UzyCP(z$O`PLSFePq(N?6o*hK@;FO7suYA;a!N@RR
zGK|F;pA0k4H9*MAens{4hhQqYOc%LXRZD#2AU>};vRtI93>at035bn{{CWp;l8FG6
zHu>krUVnF6)R`+;KuKqd(M53AoOAk((Op!!3FsgmqsDZ0vb(4+vQ)^@EQr-@ogvaB
z7C(r#LAIXtM97|`&VVxzUNt_4JPb1j2LrCpt0uRJ7n8rC@IiutI)y{HTC6N^@B>wG
z-;(=p@s0)zB10mBYUZ8(8c3=pxs$WRC&w;!0A58Oz-t^I8|O><CnhJ@MY>jND$`7`
zJtNje#dD`88z$NJ<ik?sYEdN_<$s@Yc3C?Q=v+qTQU0>Vu@~QClxKA>r^XQuJbjg>
zC*y&DADQDYFvd~fCi!f24Y`lgt(@OKK7J@(=A-)jmCu!3@?9Ye`s_|Yw%^lYvHJY(
zq^zx|z_vJZ)7Q3}k0lT8%;HYO&{y~l!Bz4H^bMR(_Yw)X7Sxr$(l^TBR(gSjqk0pb
zxA9BPr$6k+VrqFBxC{IaGaR8<)GDcXAaa`&M@4}8>`hb-SDzE#?BtKEBO*i@{Dp&=
z(_0_O=e&joR@B?_fG3@ZqCD9zgR1I1{lAYgR~&GZDbEKc+1@xU&8O5iy4iuF0(rU+
zLqIiCaN=fvd`mYPZurB(Y`_xa=wHiREC;Wh7(t)BJ@xUc5F%mL+h?e4k}gL*CU&MS
zRQYLI59dfqeDmhSQPP{7hU4Lo_x<#N+`^MqseS2g#`m@Qg-;#+n-s;QWWQN%Md8L|
zWEhpxd#kvyW*M~vdu|GEe4uM6<PT{=6_jT=(yJ&y+$us6bx!1Pe;^?*5rE!vE#&vb
zD**3VlsEAf@HILa6$pua7l_We6TZ+%dur*Tk$TH81rq)gH~RavLI63!QTIlbUlUb1
z{rTPmSt#O3#<+1mSArwI`7>HLsx|g!NO|rXu20c%hSG3Mfp7x`WU$#(1W}YX!%|)-
z|CxcbXWOMjsilc^ICEd^d#w+zswERp^e{z@@MHQ~M(Oo1*W%QWBXrQ3!HKOT{-*s%
zH_=Nvy5Dv~OEQM}2JLdf{UZp`GbL%?!BX*!6`NViMFZ%6scijHTcCwBccA`c$&NSM
zSoXiG>3K`Xg+Fg{OwW7la$JZEw%&g~OmXn?FF)8B+=!$Y*K!Q8Ag*%N-k}zE29M2q
zN?3mREmcwUg+)ZaGOnw1fNcs*VG^;f<J#x}^V&@?H&`&#aG^<IVSt^-ZL(%1M!Og0
zoJCrtGtB?<6RM$yo3_N4F$LicEMW|ax8sOr?abz&srDA5IvP2hAH&&8p&7yWT%Wbd
z7AC51=J!Q^MYckDGWYBnadx5?d#)z*m2RDP9{!sf_w9@N(tNG<XT)0<>tFM_UJc;)
zcR1#|jY+K%qhEb<0ePQ?CqYXqO5&t-H+v@xY@aT(o<4Z>hVqmT(L9}>kD2bCw0oId
zIlsV3W~7D9=)VwN3VP@t72&j*+LaTv9gemvzunEAnn|(y5)kL+l9Jjn;r4(_?dGSp
zN-g$9*-x|Iq&wd)8du_!=Z#wX{}0o&M6O3YsU@*CVA<bQ_#qXiQoI&Lkz)M2;2(hX
zoE(4((2fiSW&nTK8FEIZU@C?9v4w;o>A|SKlaXp~?X^~m7kkM2L^WyB-b5PUTw3JY
zi#J+nw0#;Njnp6tm9wCDK+XZ2A(2GK{|Ww3yK6`(kK%-e4oo1MkgO<+JP&T|BbxBk
zmF|d7BS5y|+4s+hSkJZi{ke|)o_00IB7qjt3&MA{R1QFmL^NQ?la^f2;-TOmCzf)$
zYb0P_dHL6@ksvDo4AypHHdmJCs@#EMMYK7go_99q^0g+0hI4+#q9g*=N>BX-snrJG
zvXNmM9ztw67<y6Hy8N9abkA7*8XQr+Pd*1>x?N85jGIZy6RJlHZWf%&#9A80_p5s3
z2@?xqb<La$j$%PxKlmm`+Xc1YzeOs9ZKv)3Gt5!QKIVUJNX$)q$p7A2cQtDR70jFI
z8{*s#5M2Od<ZE;59vp{dS90(93?y@udrD=08H<}!rZ%}M7&JK*I@M`!T|Ue_@FF>O
z#`QJN&pAye6zC`N50G{ANi#7#@vY}GQ&)aC5j(m$`Ipk-1)g==CL;dnT2ys;>kRUF
zjez`VN8-rK8{av?TMDxRZN5i%$#ElaO)2bFtyeUnTX8gEJ3*fQV)XAY)*^)q1sPyf
zdXSU1m)Kq_|7=>O)EAHYwn}0wt#S4)G3kSD2Q<CrB;d_R2Zz!AnKs`inFZB@E(meR
z$SJ9kpA=Km)2GoCEWXu*;TA0O9G`gR5rR7=Enc1!mu{wt<d1Rv_iX5P@lm`Nf6W{%
z=U>@O#2a~dQz=|hnXoryy1EdhC4DZgl2}4)q~VBx(Q@dk@bmZIvFv2>O7ZBc@(F%s
z7A(a~U5@7?@Iv_^)^~}5NFgg+F_HhnvM^q|z+iaJNykSPMO%aI;i7k5Q__1zEQ$Mg
z>Gi=9WU|D`rujw*{VZM7fSo`iX?|9gZgBug3P5X(0xMij=I_^<B{4${&tT+Q%2$kP
zhUv^b!olYNbQutUj@($z*nk*<Oxc0bRaP1RNC5U?pC93G*m7p4SJ&jcLJGAo*ki!F
z8JItzs&Kd0Gna<ZR(ms~%$xuiWnYE3-iv)zLsl}t#;;cXY^M}VyMLK;Ul-2755&(x
z30bOgQ0hC;DC4m+1N7g1>mwfI=YSRG0N{HRCjW#WDY@c%kalNmDz{jrJztrMZjn#B
z-}(<x_2PPoHJ8COfFf8}synT|jioE8-lT^#H*RrN7_k7m2?&Z_KJ>LS0X54=k>W8~
zP)*1{RTuFjuQ@_u&`ryGVaN|w*p($J!0!C4+-8f+yKa1|t>WZ^tkA{zkEA&(BiF?;
zNUYE4;Pr%`XBiNWSxtFDUcnz}{k@&M$1VmL#-Hwu`-y|<TAQMb)v&tz6lSmTny1UV
zNdPUd!gF&Y%o(?<!n*o&uW_g5PbW#=0HifLKe!r7Y<aIRuCI0iN%S`fE@x5A>aXyv
zT|RN?peFZ~z&(uwgBV6wJ>^j&#hg0YzUO{VB~RfQ`ZM;nsAP06`m!WBg{I6P;EV~^
zIX>u$VR^sBtKa&!sOQ-kKD>?QrRQnCeo??Od76SK;XEmZ+kMP$R-hO9&)#5#9i}A4
zyrWC%6!Tc@(NX&+V?@pQu&v%m<wQYFp^_!cB;;$T$c(lp$5zxsV|kGo;?=gFby4~@
z2|U;{qZ^MMyY;0$JGM44t#yyk+-h+Y19%bLrYU~MiTB5nsJqDkZv{Pm;x&>~87gkb
z8Y)&^YrLgq(gTa7jt(e#lOkrGvTvL>hZa5Ll`#kpDaS&UhJR~QIA^dq9jg|FjF(=z
z2dqEXsLpR_^(f397KkJccn$E!6R-lT0i+ci5;+rEpbt03{bg-h0RFPdOk~jkG4Z$S
zdOKUFD%^E?C<oO>KfD=`ZCJ=H<%}(|X$fn{*64lyn=+<r%0M3gip4aAkrzl5Y*Nh^
z7!WH}Ym9G>&AL+FFRq+Q7Vb*%J!*!UH0H~&Edv6Bu!?bsdqWcGcU7Ka20$pix2Qc0
zj6kQ~2gZ}sYo&RFzvRk=2^IZ5O-&nh(Jexr^!lBxz`l>T;mPz$DLg=1Mdg1m*5nvr
z@g3Bm{7=TD@$VP=q~Mk;r?Af)2(9cBv6Q9#Yje3x11um<Ud`3ysgZBS1pB7LqUTn#
z$_O|8UymYRB@rpJ5lYMa6k^zQF(Tk_E;_i`IEg5cRNnNL5Ix!bVv4jBifophne7&8
z1a2)nZd~zd{FZ4<<J{6eW!YyX<Yd>5cFuIkuHRbQ=H(vfqE^j{L0R>${a@^gplION
z<x|ghE9K+Nr}bP&&?TRKs+|QoT+&bdqnVESRBPdB9!eZWFE`(Vj;USuLm%n%lY)2R
z{7zg11Xzdw&O;{dop+9Ms431}ZL`Ad0ih4qr;e{^9DO@hFh;Pl9_+pDLa7^nXO@GY
z<JA~#B4Z*??uQIl!+!A9G;%<lvWtXo&dLMRLfaqs$v@|}UVCRBg~eD%-{KS^_~TIB
z>pxz~NdGrB{G?0OB^Y-56VRs=K9^`G&?36!IdCS)V)>9jM$agXZ-BcQEff4uID@He
zVW@D$7R>4<t3B(1Z7GQ$K*D>k6M|__QJ^SpZUtf3dD<yCTg0<yz@=U595h;h4x}LD
zV>U9nbgpSqFx$eb=<?3?UicYznN=I^JOhZ*{nbuH%I^W&ld@?R0RBAlKm)3!Jce{9
z*z$`-D|?}m%flN!OQLsS&#W}o!z??<1n7V9lw?~IWvW?mzG_H?BlAPi$Jx5?@t$Ra
zhWUr$KB%*LVgkOhKKIJPlJIkYEkx#D<xD~Kl$44d6Fc)n++CpMWVAFrm}uWk!Xu#=
zGI;+-0wBbEw2BJ9il5Ffj=)W>x5>0|55$^_<2I}dl!Wr-UO-eOBil;qy;%Vg$O%OR
z?!7{}l9C)z+XQfNdKHG>FV*_S#)xZ8)Mc4>P5m0lyl92i<#b{iB>L8G!yA#fK_D0I
zGgvUCq@$<J%n5*-mIbuzkWx||n&=6aboi^B(cNHNVRT*e|5*Zu6ni~Cfox9=^iX`T
zLd8HH01Idi6&dN6LNp4&FRXB`De4bLylQr~+SK+%3`1@OMw)%1E6XJmFU37)rZx*n
zW<UR0?a`2g<oMptpNJ!7O=Mogf{CTi+iaO$ohYy7jIQ%bv#zndwykkj27Eg}{bul)
z%9c%#yuavfS=$LQk9QVFG%A@|jEc(1OW^9ma+R49h2+1rz_!q?WEwabe5+r=Q~Nj*
zq>$^rqZ(r6y!Kn&yVbADlG<8`O3J0H4bDKb$_x6W@swJei#UZ#|1_67UT$w1&=YZ`
z+baVO7f*zRR|!V7<^nkgms-_c9~|703=y@5r~8rW=`U4buW9NV!B^a}7XxO$I7*wl
z{G=mc?w20mnj_KDoNMl?GbGcf5;N>{`DGp4t&XQHhmF)^dg1sxB%K^HoSPRgr_YNE
zWulk1u1=z%4f`t7#@a1fQ4cbISFI>gwAnLaZr3?D$l2Y>C&~^%iBPI~9gQh_d`Uk1
zu3`ywT9Ug#&#DJ)iFRE!2DT1N=d0YbKpZmOIepLv`|Nb{BcvVJKryd0$XQW3PO`+@
ziA=;iWe8mhXrT8V6ouZ$9n>4@%GAKDpVmYFPl=dUzLxA4NtM8~Ox;K27R-ZnOi;v-
z#U#3qYo2S0E{VCNJpmozULk8H$#gbTIs-qfjI&Z14q^wGFs7lXqV+~9LzrVg?^+l6
zrSVC$$JOPk?*Rd~st?iUvaj(cORP32HnoP`RRg!b|0L0%PaXT_uG;1@W(j4blIT%n
zp8EBc&#ZnhlgyFOFY`D>(L8@5W=JIP=B*zEJ4v9TpiY&d0LzG~KkxEcz)|*n6EiH2
zkX=BMn)mZXjcCnO8t3tf1$JjTi8WO4u%yeZqwFEQK!(fi-;Z)0|9HhiB{9;G#i-S1
z%FmiSTff(@8c~!l7$q*1?=GFi|L;QWY7|~ZTex(_s?$Y1k=ZGXiMe!?_iU7Tz~R;M
z<oI#f-?VC5D*ZHM!eA=WT)-{mnR3;JWH63lY?&Hx9Ok9WMc(ohx4S&cXTBAeVtF_g
zbgO+t?LFR&6zV1+jEwpxpEl@&R)l;^%VzzeYr~JA#hc@kTFgyaF0wuO-Rxb{^R8Vq
zh2&`2iuN6&hSPDGOBb?P{EB+htA;TgnwlrL|CZ&+UiD^1oUpICj9`r{m0%5v#0}pG
zTf7qv*`R<gbmCdH()TW0<&TJY{y*IOGog>;y&Ae?8X9ZShlRC%OyD=T56?`K1IV}0
zXl4gl@FN??NE$Tgw3}1D>R(}YCzm4Xx1YPrT1Ya6%e)uTzfUV^efLzu?waG}y<qGo
zTQ}nTk|ah=ctpM_En=#yMUx?smGUad{b;+yV`cZSC-pzdqy!!eLtWDUk{dbl6R_PF
zK%Y901hywne#gSMVd>2>ie8ba^p1ZI33VXr)y-O}^-_O=Q?;;_H;c=PfFx#KH;H)W
z=CnMNz%mX!68MuZ=N+-jXh?E*iH1+0@#`eaNDvYsQb+F<BrI5bXg4%by9m`(tgpSP
zioZ(wB$vISe)e3x%ohM2cn8K4D@5#goxKh&z+{)kbkVOhM^qR9gBe%C$ewtGjX*T0
zOcBB43qKxsgE=5?2EsUdn0`sRp%t!5reapqhZjsS^*->L5I_RjIp3(_En0tAMB$T4
zz@5chHX?3Rl4$H@XH&q!rn=~c(z%S@=lw<gRTPxeRYA!;HDo<gaQLbi6oc}l70$&!
zo``{BHOAkkr0Fqt*qk7lU*Ow+Tq}CvSuq7ezr?Gkt$1eEy>%Iij|t?Sj%!l6rK(m=
zV1AX~M2(q!8DN_m^*4qVE8b-CasnF1I%N+{R1-I=yD~fUmmvL>x^k~D?OW^3uTC=e
zb%CO(gJRH6Yx3ykQ8+<3U*2T@zMB8Hyi_=IW~GPtZ~vC-q?;pz*MDJRP?~L+sd$n&
z$o8=FDjse)vszhHk81Rtg>m_GLhnE5(;P}fY-uG=jpDrugmPlBej(|B)+gVzm>NzU
z6vM&ZM@$==CGGN69Elr!ISTy|hFoMBL-TDqdHmC*F?tuk{MbE8ZXQehJW<<z4_o={
z*rC<;#^v8|q!?Nm5Ul&(vb5Q3cWXeTAP72ra1{J-O3L-f_$OEISuL|4;=JBVs*jB^
zSl$liOh5yBcPgNQ%3dxXDpqc(RyzA{S3CA`c8rE;cv9Y<bhN>djQlCcBp%N{)w^N2
zIqCqd?Qk#+0SqGpbHDj<{RXyAnu4BcW9>-Mp!tJIM9!k+A9%9t+(F>+_puGFJ1LGu
z((kenEs#On4jFIw)Hxb-N98r)XFaUGF9mdG5j(`{(xsCa@1y9-61VhaE-n}b?Tz$Z
zc+{^bv{VcU7g#O7zW-urx;Zs{t7A_&(QF}bjXPToyn4&a_c7cP6K6l5KBUtW^@5Ik
zQoq&&j{*4x{geq{DoqIHD*u>?=x(nZV)})B@l!kSVl6txfDYMF*wT+_6lFKM#7?<d
z+FZ@s83nFJBKO#%1t>-bR7oF*Iu2rAH^Y&sY5Sw4gJIrd)$cXDvK1Xr?4ht;(qkk{
zje6`e#8w4sQcgk$vRjyk>`5iE##a-^;2RA`Vz^0CKtU_3zTr1R7ebpOy`PE1F+vtc
z6kyFnMzox$roXukR5iEM0@45I!*Nckp}`Dz11w0P=6}#ds$fT;AjHb991B^__%J5E
za>n6k_doQRxERy0x_^xuUje2QsB@@ocrVCl2I4hWXbPAZ%$VG*fJ)SvhN$X~F<h28
z2^ZhhIJS$B8;NFi%g7ie&WDg~&(*!f?W;=)qaV)Sa+sYK5HAyg8V^41-PUIiv<}K5
z%Q^^ExqzqIcq;@JX5c$9haUs_z16ed(52D(5@Lo_uuntd&g#b6kQsjnuBtc3`)*q{
zyjSLvvfXzmZJxYPfaD)i=CMb_#Qg%o4iEfzxGSZ0c72|`+7?+GHJq9pw>BvUzMM<D
zG%h5+6Sl%=AQx8qI@ZvfeRCvuZ!%xP8H)&owj}zBCN_{ah-4;|PB{o7tk!>xXfD6%
zS{sW|OfmD@dm5s_aNG5%s`^-{W&3C&f>F_Rxq!#(+Ch5Lx)Bs~+J<MX-%{o5?CfK&
zSnh6c;^On8jX>kpIl-~yv0y`hqVHMz%dtt28L^m^-qrETDH6^5qgT~R)V_Dtp6#3r
z<8u`+^$$-pFW=Iod_kKNRoFCTCEGMIhv$SHWe#nGKguBZJi}F{U_`5+Iw!7`;Y5e;
zRDhy-ql8$IaE&D__?)i?Q^$dC>rQgEB&}vQ+b_=Vmqju3aC=`n4;IOzKQ?;^;cDM<
z`6vr9vFbIL|Gco*agg^iz3Ji0@9dOU*SE{VHJHQla*xu~pKD^JIOpO@iV`gjCs+4M
zlQ>H%8A2Et$j#<6%^kaZ)TnS$(Lo~`@VghVrm(|V<x0<Wt~k7sXZYGqZ&VH5u@ZRi
z??S`&#S(8fykEgmpRa#rdvV_hk*`9)U!0$>p(l#51u+)XWPhZ`-pWj~M6tL+=SdEu
ziTNuGXenB(b#SpCR47pcHzLhz&C<0s+?Iz|!bm(i3(vTl`mF5x8_`vO`i2l>2v8nK
zK(f`wPXzjLYcmAF8GL6{e@oAQ`FjP%G#icbhGnl=eHyJ;It!l<?0k)qj6`cL_W@g)
z92Rl0luNnDx1Sov_9pTkHQlwwiC^Ony3FxLv<lV53>KJ`TqNW>QG-d}>xqYmN6O$;
zpLQU7K1Z_IMY<h#?6u^8IPni^6l!2QtQ9y<8;8l=wii(iX6&w5PCiKoD*%%WGATk6
zW03)}<M4?ue2U25@ctC!w;LuIv{M6!q&X+w1nUV5VXf5|7)g9L+jUMuP2hi(-rM+1
zTi<AeV(SA?==t?*#Ldb0QjWsSU$y@Px8k*k9a5`C8mC%=kw?Vd#lmQ!w0-aH^E%ck
zPRsmyxwgc-)D_#z(^%7+tZ&D!JO0d~x=kik<?dVDSekLv)Mwx%-q<5#e&c;@66>k0
z(cI@is=ILjoEG+S)Kz92IucvD9i?mu1wiFUNd@VzpC95I+iXr;mrebrBL6<M4pN}E
z<o&&zPgx30wh<lF68(2DyTFh*ou~x(zb=61l>j6gn}y#hg*!{-9nBL5kL#)(^TmE^
z2Q7}KL2%v58l;bXN?d{>IEi6gE@LKM4uVMynIAQ8MmyF(#ts6aco!0;`0n=Z(@$w$
z?E%U2FZb*5&6fQ(Z>V+wzy%(WQ*hI$=#FFib-b>Nk_DE|D>Zf1x9NDKB}XSFV%zj#
zv?Z=6^m!vcC2dc|oMFqi8D=eFr|tSopzilt0p%cu7Gx5Mqv+w~yWO9go+tB;HRB{T
zKhrba9(q9ZCq2vXC1YX2@Ofg~)QP7_H1Ls{ltTlM@+$r-U99h^_ED<T<NnymSi`C#
zs;k&<DY9cP)`63ETRZ#s`lcUYiBkO}u$`NhlM4P@8WH#Tr?ny}ESpk4nkIDVz<&j}
z?z<<;9H`^&GuP>62JR8@)E+^P!)8lPK|!V93>`hiqFUHBlNt_ebic)OFfC{TNAKP&
z^|3#W#>;E9W%bHe130z`!`kv)NY2WE3@=(3(`SwRH<G7F1i&5P=@BOT+^uVm(G}Y#
z1+`l@%U&H<7j6w^b9g!CvVya2#5B86@jkU;4)6-q>lLrHoQx>mcE{WGc-!W1l^<)9
zXeXy=rtXRR7IULiaeheSokdKsl>IiJmZ-dzHv<DMN1_}ueeS<7H}H@yuecANdhjfN
zq+Lo$Gb!9X<X=`?JPrhnr_V@uU6v!R)}>n&B<j5FI~??joERS0v?0}%6@jsD?R_7d
zz%&($zq%Ed9K2S(%#~-+USGWkn!KoRw~w3T><+`TJ^gcEpgdkn0Syl$r;eJCo!RBH
zyJpINHMSnE?B=*Ao~Q1JF4#*qN8b%bORLp(q?58gqgBkwClLx;`@0gqfmcmnag0~-
z181);EnPJ2w5<GDsKdPK<k;dI(ty-@gSpuE$reF>F+R|(rz*c|L7}FZ4nQAAh<UsK
zJQ)xXbuwkq_stl`7VPW(cs6Eq>AVYImI0nK`JJLwdPM9mnSmi6yqugH?@!M7VVsQE
zaIz=To;Na(+$|1$0B96xr>rx9R#KU%QQWa2{EM~&FbB#}h+0u#Zj=BGqsRN?4K5>r
zk!|DOb85&!NqX^tugQ~~jKH?#Y~C#?_++vdmfX@@E@}Wdgn{fK?5yBelaSHKdDT1g
z1xz&!2AU(KS+r&YKX|Jm9aLeCy8q@Mz!>hi=dQvCk3;)95bDUO(Rx6qfYCpm8LSR|
z2jmg}wQ`=dxg&*!1K@TPG2u~px$SXm2@j16qnKnELFon&#*f-|YX2;*zmrtz&$c2l
zd}n9U`KL!Yw^)tYNKSQr!ZjexSZv^yV2=Ijw}j`MR8ivr01vmtFqE&hjBge#CvIhK
zVznA)FGn~!IeY8yl7r4huzi$Dbe-8#-RZAWk<e^Xm%~?hKxQi26-$}$g@9SsIGH5R
zY<4xZ^#_$0#kFgh6#pXXWv@o|@K?uL@u%;K_t_CS%*E<>V7$*qz0&+eI?FZ>W|Lvx
zVVg|ccMI%p3~g?gL=8F1s5NJQ2xoo^b8~-&Mz+B7?!@jWoAj-D$1=c~(=sFuVg34M
zG#rQevZsKcR_(>Jd6>iieY3oex#9LxPPIQ6`&M_va`$>?p}@_Bt-{RZjzWSYl?z;N
zavZi;Tt`FAkr1(!9q+r#;j^b$9$1(Ohwh5>DH%0+X4~mhUYk~W4j&)l4rdGhYlTj5
zaqFDYh_LuJ7OdhRF_hPO%Au}U^3(3V%?vnc;+G>Tk)&^J<rRLO3-T@^mwn=^_ZmQd
z+2;vta@3#&ziA$`deOx7EB2eTq(KBg&v`<fzx&-B3zG~b=Jj_)l-^{ZDDZwG%-EX1
z^-G6Zv*6U}rbGljZpVpHK|`AE?#3rZS|C8)FZEdF=tarl#Q@s#-IISEs{!YSt3(F;
zb`hrjGpl4d8)I{Swkb>=?s#UC>)6#b74CN>pMIlmoK?$$@VXKt8E3ISq&b7C?*FE>
z|0qUszrR(Kn3On4#Pyms;Q$gmo!~Y!AHHlpR1g8d3A>1iG;GnWZ@+p5hI+TgBKg}+
z*uupHAhAI(ci5f%SR3H{vP@}eM`P`BmlPMUrJc<OVcHk(-(=nyc<JdEACg#~i06mV
zfhG4%S=n;n9YJuZrPII{w1*|!w5HpT%ku1X%k?MYp6UBZ;r9Kk>**r0oSjM=5!?r<
zyI!7UulASjy^m;TT>*Zdm`ElHv40^!v{s*=ahIH&5`V6+z7`$h9=j*J{^=(5H5hO7
zj$e};nAYJFdSB_6JH>WFaUWkIVu`c+KjC|3|JlCaX57XQh7KnxYl(eOlm>J*Hk|%&
z?~nf*lcGt4Cz6kwV66yE$eqXOw}$q6x9jj-#1E=|>wSc=R<=r#i^**(BEMhz5u-zr
z4tfaQ0^+;9Fy={_qa5=obe%u~WO+TI4`eTI9kGc3NB}Y*s!VG<1V_enBbKOFwE7-}
z9vYpgMIvCKx=CNr3e{wu(}B?2QBaygHFgo%R}eOlg{~#-xiq#Fo)Ul+k$J$fIzNNP
zbrTq;a=P*{!vb{6L}3}gy9blGh&&XVxl62Zy3+5+<xsfj8&NKV6a}8<8NxfGM(j&;
zGfXWl4RO~e*!xP!PY}>oJ`=^@cC#3#{K6QLpZw|LwZ@#zTGroCBZz#KArb<<tjReP
zt+12tGNFjNASsn|kd6DE-Xrv5n24-8od0oC%9X#Y=1})KII+SkWC?TpZ%aIaM^F--
zDkN{)xnbU?>lEb+wCOEIkW!x7+sB?>X*c#;x-8&$!KN0UwvC0pZ7StI{f$#kmTVMf
ztI@m)&K=?mmb*RPFfz8`z;&D#Hr*@r%n*wIQ1r3w&+taGc<i<;guT}f{=&qkV9VyR
z+8$~|{MxCwZA$Oi?)r1wRIT1>cA!F4$y>F*LEgg|B^Keg;z7;Fo^YqtyNd2Qy+0TX
zd)nEz-sB@Rbc+LnX@?<8GE454r^`<fubN_dXoK1CC2`Hqq7K^=8k5?|&A;AbG~;YC
z{rr3=oU|PYs2HZclJ**uCk<Dpblcd>ZW^U?1BU=FYj(|at<g<WWsY_`o?}}~IF9*N
zq#pLw%UxsTB><G!k6meh*Sux~ve&tr*Tku2Fh7<v0_@RQpuRE>opNDrh!6c(!200}
z$*7)NkkWPgW7wGRY}^?3=32k-b62`l?NfpZk4Jpq?YmN|+j8iQ+1*Zqe_||`tA&AL
z(fT18ugj*Bqbpjk?XMTyjtx2s2A$PV$h+*j4gH=AbUh($HV?4-vkxj%<CCqcYaUm}
zrPFFt=SkWh_|bWUMZ+Gd#_lyw(C^HC4O};dAD?MpOg|+r3o46`I?sNoKDZDr#qhMk
z+KzmJj0?mtSK6GitKR^^02tkT0V3)TOQ-OYj`(cKFIpT38)kekl)CDz={R4tTKhvN
zO%tg&{9nexQn8#^RgmP_xg8GRvVUS#CHtkpGvJXAIu_J={05%3cq(WDbcV*rY!Ct#
z8aiIy!7t687HCQqE{<B|Xw9ykw@tUHHzjvga@E)6UMla7c4L1NORk(CN*>bR@U4UA
zQ&MdRYuG<3?J~0;C3_CH5g)+MKPKIG)lCp^=vP#c^!4N#%se%%A$pYPEX<&h9<Zl+
z=2D+TacByEm8JczpQA4K59e5v^bk6laC9j4lEFH+t<m3PCIbro9^BHbJd0>N11h9v
zc21mFWx~FDJl0=`?gr<5ImkJ_G)^!NwGwO`7pr?oqjLuqmL(lVj3ULdeqpEV2rEwJ
ze`i$1%ZxhIDjkA92qd&F@4}vl>BJ$a%aN%e!gTE{0$dIMEaI-)JmjsC2SnQ>oM98@
z2Y93q4}k0xkaB?1)>Kw9Q7$BqudUzmD&?2JgQ}^ozS~c${77g2p{740S@y&Eh?pO#
zFrD&G^iubv*i&z;o~HCnmpz_zX7i(12;fkqIqtB8P~lZhq>h~{v=nS;t}X*5N=Nu<
z{qh?%7wV@P#tUei!%2K0_J)OL>;;^M$Dd3PCRfJ7nAzew?ph=KGgs7q<j4XR)V&Lq
z?U|XZ$5xaw*xh7G(J{4C>N9TLVx>gsc2Sm$sAVfC{d9;zW=Q)flWgE=IvuqU%}`G9
z%%c_i7EaYUXY5fhlg{|G<TEWjDaJfzsZV{~NNjCUwnxT$aWwZ#bNJwNU$W14^uE-v
z*MhOtq$2zx%h|)_*U~*>LE2El11x59j4<-ahCk%_zCn0cB1$jmO?ok0KIj2l{w)4w
z&nE2FrsGk_YpxA$<Um$udiK+XJd_UnWWrw<s0Bqo9h2~{AUEGi#BG!1ILx-aVKkd-
z1$&#gE$=R`GXM|KP)AD5F6g~R4gCU0775;M#~6J-)#rLhCVRrzp?b5HynBBdVAi*}
zdTURTi~^wh8TtESbv39ily8l&+56DWv~F-)TaQnHoAfLcf0<?JZQH43_B=AS`0l>C
z5NgPn*WLg=Ja9XvuwQ)R<4I-qV?ItO>28yH#odv|Ne-CzPA7rx9ASa9i@}fnJkl+K
z8vTz~E7LTO7}quh0I01O@;ql!!u3nZdr!B2F?#Fo@)ns%O>y-)S|>(qT2DC4+!~&@
z-@HtjaG!-gaQ>))@?F%jNOnp*PsEqmf~2uhu_~sOdg6Q@4$ZVtl$;-DrCxeX5dO=b
z7B&~NRxJn#_fc1rA5H+$?ULuNN!%eagy0podLTG0;EXVP0ifXg3*Ma8(E~OOJ8aQ3
zU%6g==v0t|#F1V;9yFa{G&s)rpJ45j{H3{EERwp5LJhfJwQn-LT5VbH(w$p0ds$GG
zrsLj`e5rH#8l!j2qRLppFLcXr!ocKUvBA>fGSV|coK_iBCW(_X78V>rJ_i&-KahA8
znM1o>YGDFXN=NCzYm^4*%>Qa+U?QLSD4YOo`*r*c`v8E28XoBYybKj&91pR<>K6Y%
zI9bi`r_Tt6GTk>UJ4TKgv|*3Uaav;qFt3JeuuCyANSb58%#@*6OB#s)KGYQ4O{q;n
z7$%=+pVtTURtx~alG1lS_PUQp1a&ekhzyzl<Shxd=^73;u{7E|>vZpFl^CaV266D@
zJNaH2d~M_^^adlwzGwiRrj*dqy%)J-C1#aY<P#kguNkBbG)yFc#0`{8q%q+d$xQ>C
z(G{W}k;162s!_e^Z*Zisa{AMyiLXSk2^FBY^KS-}+9dlh03T$`BYW$$R%7LJFL>XX
z#~4(34~8vPg)DS&QNWZmGYujF^5Vxl3rv5RHk7`s`#!*w3@D$Ts!;zUmkV*BwWMSi
zW_iznPZWDh(h+ajDr)tv97m_v-A{ch0?UO{1^#FL3D3U2X+p>X?R9~%%af27y{~*V
zrg+0kkR*&4zFb^S(%pr2<C>erk$BV#anEf<6=P$96?`fr1?q>}(pl1-6aLswKm_-n
z(5fhT;YXM0w&D}z7&j{RU=RP1);<7bRu1TF^tq&Y-V3~2Ts5)>IO!-4jk)OUb==KV
z&Rh9EA*^LjvyNvmU3r(xt3LhqC2XE&WC#D7VY50i%;)sgH<(bOx{I6B)vIvTzWJBi
z3H2r1pc^-{atA)t4jpW~f1z&(P%Cjw9+VZq@y@;%o79_GkB$A|xZj6=2R)tjb8<&Y
z1@k?cI0a}Emm&*my*#$#rXLbzS1!ZQOYiu8KDaMZc{QsKS{n)CGk`YR)H_=Dy)1B-
z*Mb&OCEMOPmUPK%ygkp$UfpYd=%&%%-tKR*$+U)(2V6zjJ-P(5Hmb)kA3Q}gv`#F2
zk{|25__YM=#V1c}N!M@P3)B5Tb25kcZTmR-*56JMj56gIDF++&<fG==-(BR%uT!OF
zk6o&?Xn#3jMZBTRU{q4)CAbU58&pxnY;l#!px)t<jcI`6M)O}5a6so4CX2hm!^clV
z96cYGeiDD&WegV4kG%7k`IDtN^U~N_>9Ze0yt;eje2MzH?ZtA=-nCLMg>94OKT(d5
zQ0iZ7H1PA34OFgfa>1!W^@|*ELI}LJGW17;^Clb{*f?ZFL<!`OLWNtCuehklcgd5u
z2mpwc0E9cw%BUU^jFkQVSW-~{XeYTMINiWhv^BKdk3k0hxUt5_bZF@iHVf&9(M`|N
zvR1N3mt^_C1G+s?2x7wSS5+_|_D!QD?T>WR0zfn#I6&mHK>2UbfAm$5H<8M+%Oppw
zuCJsS&5Mu&1lQ!Qe`V+fKrs0D3eou#oJ+$vz5<(+<@EE-v+V~o+`BP(acN}%J_lWy
zsX#{1={_#TPHeAnV%EQfYYj6%7<$<#9{`u8vS%p)D2tyg_atpIAS_xS!h2APrf?+U
z!^pr%Sf4)_g)fZc-Q#qfhI|WJ%-4oR3N5@?QXoPAMSZb&H*gcD_|28Px*;vZ|0o}c
zK|r9tXCT8<zDee&X!HxM1O6))>^`*6Fx*(BY3>`f1^+tyJ6|rRh4f^b%>2P+ySW@5
zryA79271^4goGW3wGaJNjEv!3je~udx5{AKY4#UABm@@u<5;@`Z24c34q-6PoIW6W
z-B&JyzJEnDcGQ@3_IYJ{8g0ycoO!V%X2{i}e}%$bA9(eL5{2OP`;oYopJ^RyKA)D(
z$@!K7H{7yGjjgh!tc*e0f1>>#>JP}M8kX)AEp2tIXX6!u@t61|;FZ~Wx;QlebJ!1b
z7QNM%&BawM6bonS%}X3t<1Iqi*lOFlN5)bDK72jKb(2V~tTE`&&TQNBbm>dPmxrQ`
z$C<)L++$<~-Ws4}n2FK)#R1YB_>hFiUgGE8qEzc$?5F$=-Lhhur$^EZP?glhFa2g4
zspHD7UV9o(*b|!YD&-YU0KS#v&!l;g9uA*fN;95My~_G(?XMLAp-<0D7hn6u7bG^>
z)a$2Mka-^$8=U5mi#<SObbT;On)Cn1)LXbU`S@?V3%WZ+K#;D{f|N)}s-)B?5s~g5
zAWEl%k|NTb0|bUNNJ~g}cMUdRXP@uyT<1D}!q{{1ym!azCh*zZB=>$zoq~4(<O<$z
zT|UpBr`sYr`$&CM0B21xo9P@m4I)WjX9B(RRi*dG(+lgV=L=u6bnfU{$h+|s!o86S
zsx#JU73cc+I=NLGKm+^Y3i{=jU`ra#leI0+J^d#&UC#K^a2}>3QB#6JfA!*bhhKc_
zL1_9%az@8pOZkiOZevM<uVN@lWYXGkNpNgZ?3V>w3A0SbB$_CXP-X#;7rxd<z*C<7
z4DxHJ8>HJ9W&%|jva=;$^Vg5vWtexHEU*VcZA=iKZk#5330Cy<<+g}kRMtegyfh3X
zpNO_UD=80U8Pp6Ad7#Qo1YB3XiT(FoCdQo?7v}s2`>vW2)(`vqtQn`8L7lDfa|T;1
zHQpn2uK2Ed0F<|cwLT&i?-zmoI2%<zFVxDxdjOC&ar#7xlis-OoT<s616MB{tvZ2<
z*2;7b5h=J90B{_vQ*|P+f)rj9P|zR$J3Up6`WQD1cYDyEn~aa`44{-`&}&Oy45<nF
zGwkt_NBxE<5;$9fv;G_rA1cVy3dYl*co(EGGm2tQ#c@;VCSmt&MQJY!mbAaL{uX%y
zekWR*!FiE>{U~NGSz1y+F6ztLr=P9fVyfliT=yTtA4$D@ny!LEs0#^e3Vn>m*0`u*
z0V(Te%{tKmGt+#_N9Wnln5hZViu1$i&Jv4|RugiQ9Vw^wZz^|w6<ADkf1S>jcvbd6
z>%#E9Q|ec<h47`+G{eV65LH#&#EaLKU^k}zd2CW$olWQdzGhki(OO}H=*M{8`OI|O
zS>D%aWaEm|a`yJO$^B;N91FUSpy#rH#EQT>{p#0PKN?+d(<0iYszQ8X&_@Zu`kO?2
zS*$;v^q#x^cKVB!zd9vtl3DdA*Gg!qmhr!Z2Vm5y{?XvHzl3hs3Bw1?sdh103-8*{
zd*n`~%wO*<@(}R^>dLpD6M|87%#YMeXzy_#48f()nbnWMmX~Yva4{U6)yDmFsXIgG
zoO|md<=;1Rq_MwNQ-wY3v;E$}HHby;hnd?Qp?I)3-Tp8LgqPCjF;fEW`fW$x2}e}f
zz1xuI?ql-eQmR6bs<#Uq$(;hD1bi&3x+#!#KbXZ6abJ3}&Eq`iXD!C!6VkjsVmz&a
zEd>e7&IG&>YFF*ie`CgH05R%6mm2kL37AzDLjIF##GHLk$cU+q1Aouvgl@bI{fq-8
z=e7<9SRb_>%Qyd$VPoDJ*ng4dm0{M;ptFohI5ubwIp40N=s$g0qnOVts3_;PzD81L
zXoiU@0sGQuAGf2roK-*M>HIwDqwTe+uNRoHR|X}l#YzO%GILwPE8*vNHgOJspl|$w
zHBws}qC?><4AWNnE42g6WAr@&)#o65SMvt0E`Nj41q_}Hfpq5`-5>Hj$lIt{G8^mc
zFWp)u1}wPE!+gSkuzu1slIx|%GcZL;Yp=veF;FCc8aLgnOmrs&8tRH6FU@MspMVID
z!bU1=E|$5{E!|6SmBY@u$Bke8p+<m5<MV_zE=PsD@PrUJ`8z>m>b8ED4Ai=~^yd9d
zg=qR}c2^Appk_A-VK_%Pwv|R-W~<Fc<Uj!aiSJgx4@gz&VPo$dcoWg{OukAO@twW7
z{@ImMsCL=J<I`8MxcRhhc*S~o$)IXAh{<2e6w|(-HZ$TY!5re7Dcw1Us$RrRmldE`
zqM4dGnO|&HZvB;kMJVOQ;`<Wv*bfc--^@e)cj1jPX#5wpo;j%8D=~d7Yv#~EEMTwm
zW}?-#MLPI}d3C)#hf8nINrXm<Y*tmr^89})OE<&USFuSEgJ`n<O69wUj`KRZ*x*(}
z-SS<F?t2(kT!AHF&eNqTEs?!aru*0zQP|7Pe0+W80$qzwDFF;v|FnSng9x)uPSi37
z{K~%lmR8Ecx(;%V6$MpjWkZ?zk@pE9*3@X2N3~|YQi;jedj(qo!J$KaX$fpt5E&9U
zZ@SR^{IWY~WsdcJqezNGcNMjanlmEqo@J4@<-bYlVJ`j89i{XA#g`IHaluUYT)gHL
zE4IXXt5K4TAaGjSik!L~|7Zpe{5||n4g>?RPEg7GcYbN~VxT>_qnfeA)+XIO5$xnh
zI#u*L)L29A9L>Q~LI!rjo>qlxcGK*|e$9dugnE6*`0@avUrh><5%`Z2iX+)O8qCN_
zvY1Zl!peU2*W=Lv?+EoWPLt7icLsofP8*KDp5wqp>GFA-``gl;5|9I8qaki-*s(7E
zYT}3&dt5mxR93<R?`$ry&u}G=tMm!Wf3yot(FHIdwqM%0a}d`S5Dj>qv1+yr6^9m(
z&k}ah;ei48SVT^LS?9vl6EzA?>#u+Kx$sB`$=^e-x$y|gQz^~7S0y!_oVpj2jQQb_
zRP&wT%mBa^ljeFU69IrE^x+S8j7km!@>+o5*WQ&-ZDIt@4m2+mC5z|yOn{c^Ch$ox
zGX<xg!w^{NZ%rY2l~>Jkg31vWVz{#@`JIv|;Vjz*2W_J^E+7Gc??EalR&-|FvXcw3
z#<F}V@L*jF(QbmALZ{kXv@>=ufjhu2fOms@rWrW|o?L~Wadp!xd9iiO7@$gN`0{<r
zDiLY*C!7<o$g4DTr5#XzXDn=%IFK*$iP$3;?`Y>t1-0n`c17TyHNAM(kvNv?ti{Tp
zfv*ak($X~&3lRJD;jWZQd@SCe_7G3CK8yqVzx>61G-gVb-qpcIXYPUgYO>FP>$e3K
z>ilKBd5M({+PmqFV}Pn*_gK!~2fuPrOp))|u!@5IG<^Bcf3&Be!5{%*=t;uR?6f(H
zB4)j(nY80VL>j(b3K*4>b8`||GfSj`$t@5rys>&7{AHG-=y0yt!Ol-w7%jpkkEPZr
zO$9xQyjkwJAvI7Mc=CBadwJ;$pV1M-kvKYf1*A8xCQR%^+>ai&`cs4RD519cp@o=T
z7MBVlfNlVeIQL4OAoaE)k9qJeSZY9LY!|a_xzx%YkW4aTfyDLcTO3}b$9~X}Igi4C
z!y$`Tv>}iFI?YG?^q3Uv@KI>r<!Y)-MYuUUXB$;$yA!O;qYsp(<o`I!>Nb-enBWIr
zP`aB+=~DTiK&JHqr0O~u(JJ(cLpVtX%<<F%i^?8QyMTxFtmnts$AqNI?Sw|pyA@ut
z-{Ko#GmOsnfm9+?MY5J$&N27RWaBvJwvAM5@`29t^<nc0)pLim+Z|`aE>9L*9{<e4
zsO&2ojlpx;F=jE<no^c(IiZwhjwCyN>O@xyn3Yo_NTwhv%N%&Z#sFX7T9wf-1>@L$
z?2DkQc_LD5W465jM?}#wuDbQr7oIX1;vYnj&KUV0uM9??@<!6<Uo(R=XzG7k2a%NJ
z@Q7Hul)BEp?@QhT@E98e<lJGsCu3I86XqCYh^)w&=?AiyAK#O+s$lENAc<&HQ#~8f
z$bA#{<Tsi?@2z$aXJ<||oHcqKn+eO)4zc+2g-BhWPFmS=CawEvNdbT%KVh%VI*uYX
z+t#^(6aT0YAVq?+nq3ZwsULZU1Ej|S8$hBUL&5_RkAFXJ>y`h|D3VPu)x3tD7FU(I
z4-rgqLMw2fbZE|n@0|#CE@D7zf{FD?d?m*W$j{XQ@0v^ml=sY{^0d~I0tZSD<)Hf$
zxu@;_u%c7w&Y>@QUC7Js$3J<I+V9c#I?uuXAr=0wiuLw_sagCk_sKY-gwCdbJhpq;
z_s(9%LfG`M82@=BL{vh4^VfbqZBY6b3H#iKm5tdS8t#$k$t*mk8V8A)XEs0QYSi-}
zggR{t|Njuy3O`04tg4DbmgRHQOuA^ruq)iY5-66xwidmtvCA!WKM}NzudpCZD!Zn@
zJE9~mf>iw^<0AsSYeEJ?IfBE4({NruzjUCNg5)!uifA2yCKDS~e2FJull%RU5ag$E
zg%N~Sj-x!<&j;E5o{_3_!4r=CJxxj_^6<{e_l0J)7xd9B7M|)lM56=K#@sl$NGkhW
zOT#D-dat5mE}t#A1fubfWy|-KrN&eupzE`m(ZG>2$up{58W9l1nQQYv%k^*Kh01K%
zS4zdDWczAjh<^L_i&X7X{rt;MA>1{ee*5kH!=NsgA)j7~3%iB6Qsbc=i`Fs#$<vuF
zgt6C`7g)UdrA5WnmW{heE8kmyb0G~9JY&>rQOvOJ)`=tXTN*DCCd&`{meL2R$n*_W
zdT1kaK$Yy{Xs2*tAK1DUz4HKo>tJy}6IR#-Vh27?B+zpHKsHOwNtz2Fk02rjt;+$(
zlm?<q*1$C!QV2XkBXa;-n*KB(^o1;tr<x8RRo22Ru|-GQ`a=K#K$^qo@k9Wn#R5XD
zfER#|MAm5xTi=noyVvDcRs$4T7K-8V?A&(KFJcT<2;50xQmOHt_xiYa5Pwc6lgp5L
z@t$6U5NXi=OhHio1#SW4ZR`a(1E{pF@c#GCDMYmxv_kMNE(-kmb6}&L5nJjWk|g`o
z!JHo<xEuZa50$2MTpkBX{E_!)J&dX{fFGgW8fy@WFx@|nN)CLxKhcnXZ1GopIne-J
ziXCMzsbTMRe_&JKU1d7@@;}f=laNx!CyN<Hj%CH{14onCq-VL1T8+)3L9w9Hss~5}
zL1leDM0Kz>0c!<mC#Ygz;)ye-ze3z(ihxU$a>TI+Hu!$LtV~e_-J5R<?TeJAdH+1G
zmgJ6hb^Y^gnC<mG6N2i%93+H4u3DE+Tt=}XM21+5m}R2rmya7mXvtzSZ-6Sw^*s*5
zzZ#p?fTFNqv`mnkoq31KfVs;v$uOHby9wlPdwGx7Sp*|i_1q6N#CSEKoDSB}4@Jz@
zr-|s4D{%VHn}tIvh_XN@{^_u!;&C)Kui>71dz(&Grsb}1b$EP-2pj)Qe1$!CELb0L
zb&Fy4c%41#-&Bd@5)dWqH=E#UD%c+{X;;qe1Yn=UqK<3Y9)l8>2hV!xc~AVrM~I!)
z^FItTvp><}SUl-sZ8mk5+ZBy(sq#Jf*-H?;WHcR8^OEd={D;n|QQ9$o)X{q*FYy^|
zXt$5l|AbF@7vF%=z>|;Yg^{2OjjUrQ_bG8%Q}^{0T&WQ20T_T}14cR|LS!4+w-PT~
zQeti`jaW&fTr!i~4FdB3Y-}|IaEaUFWd#|L8ZaVG0KxpA-89zRYaGqBYX0<wHjz?9
z*!E*CjUMn^87Wi3)1>owDkh-FE>tSDYpZuibSDT<4T4(NNJGOiw3Pf%5CIBcV#GE3
zg~Q9rMu40owyVPcfoK_qKB|_K+|#@3<_*<QI)f+z3<P);5McvVHp;jI`N*YjFlqqd
zT4W-gnogw!bB-Rsv$+n;z{thcijsY=w2K?9o*p_>9%g+)(xTD>05XJq7Za5!UpZ}o
z)|-zDmoqZ3=0NgQwdrVZO>LjUB0~kOag#<X%RHOeeBROpJU;`A-)wYNXoPaZRB+9=
z1xPMU#GW>;@s<4&QI)BfzlR^?wr^n5MEP2|kmNrMtMaU`ZX=_D^R_matF-N9<a+BT
zNVf{?hB*qm@Bh3e3JLjytV@RcPXx7}3a-1v3O66J<G$?w3}{UX-W06D7+%b;1W6qk
z)@KxrG7QX54JJJGV{WazmmEweJm)90QNbR|{~wRAewC5q{&7up<3aq9!kYKNzfCTt
zlRKY2r>4ifulrNxnsVFc50~2O3e_2;<fJKwV_$Ay^FymSr(d{xYK52XIum&05rq7B
zHXj*iY-~DTm{`(>=0)u%eaT~q=N0*2wzBT}RTW0L2E8qka(Np|p&JS(ax0mB9*^E^
z4!P7vcB{$fmrA#wySL?!kD{v$OGSU)_ERxLIdVPiKcs<P6n~c2cg_|n88~QGaZYQ$
zaKW^@xd7hW{=iP9zmJ&AMlLwsk+TvXdtm<aD*j8S^PmAtSM`RX9p`aXK5#3dqF<t_
zLQjv)kfJ>UQDG|BD3gRLi96b!e~5u24#2UKq_~HpPtLvZ-^uNbWPQpYzc6->jltX3
zs7qQ4>%UlkPp&#9ht|F7V6`b=pUO%8`t?D1TwE(<o6lDFL$(ooLT@$15<%}nk~@xd
zk!Kq+nrz@3LrG&6qA}-k5JVA2y<CXjS4s;Er0r(wWckuuL6rf5TlIN4tA7fbtLAil
zi0I7Vsn&Tu%s_Xd(I(%yfKAS_DeZQLG0#`RnQ>mvc;KSD!Qg7nPq=JY7`v@=Eob=V
zW)sPShYPa=6jk?`rh)(}V-DJ}t_928416ExJL3}%#KTlzo}!ezpN~n|&|dn+n!-eX
z#u^rn@Q1?Vbz!@nw_cDLKcy~iArat}iexAy5Kzz@hW{$)0WF{YJWSDI^g%x60@zFH
zJOqNhV2gg-^9n-$F4?Fy#fPOgb}nmHh^>^<$ALE8w*P5<Fdc5%$_q#Jn>v8%?bGT<
zdPA8N{m)DA83x`Y2wFyx@a;6R^M?zG>*ES>sNw5`C0)`uV>`ImiJ}1nfxnAt)umZg
zaE>SG#l-a?se|S>Vu~g<VH?-wRKr`j>5#(p%i5`Uf5q<Bi8GP`gW&<AdUy$??7HEk
zc89_M)5*HuEM7}N#D^befHB>}NyMr9fc8%7tC5`f8TD)lEWj0O(e0w|;rEugm#?UY
z!5m(LUo#HnD++UGh)#YLE!yX9HYVtHM-#f9obXq=+Gn}7w3<M`B4I^8xu2q59+l_@
z-WCe5USBsH<r#iV5sAOYKFRSA_l;SIT6@9Df?If)0mIECfynQ@yAc1rdC4Az=6#FU
zYs*;?L$z<kH`M9+y1@%JQk&?5^p3rI&-ug5_!$p2M`%ILz;4EZbtjuIR=S{$S*<r7
zBksECt8CSQcwL!FeQF5%%D@5OT&+KDM=z);+%-SXmcPP=TSD$OxFcd_@3%Xy;@Fln
zceLnYDOH$UiMT&BG)|_bwQC4?4>kLb{C3dpl{=~_c5K*wy6$j9N$#gEU3S=%33<jb
z>Kknx`_$SaHp$R>9m*55IApQs+Dxxgwug~$ccHh_YanJQ!F~71&PD-iaygLvjy0R!
z`^BmJ!OVR8>I~KTvdt@>b*D~Tw9qHYTALsdJBGZ#55(*DO$aEr>)Be7hRF*=Q#Q{q
z{XI?&3+-w#rp?5W(%V{ku|m{)MLEkd_qhfJWbGyF#VAcdB<0t(@=ADsuSkHkD=kB@
zmr8-uVBvS<fZvMLI8+F>ExdKkKjetGJZ;c-qrqJU0Sb@MpQv}sidenE8%Bzv#0x5(
z4d=@$s+3S9NmV186B~Qpp$7Fqj3^2v1ERekf`ZpAigAcGx|Pp+CWL^LhQVp<k_;~m
z_3Mv-BRtngD-Z`R;%C?=<JK)a&6ax3ufRCyZib_9&^0OHA3TJTyyxpLWmU<uk9Pvr
zr4?iJ?OU4K$87VAJc;#G1gI`H7C%thfAra?j;F;hP~Oz^e+XXPLuX#z3G2jj&4$YF
zr?6p~_kgNr%Q3x=4>sM}PM0K&@)=qF^SZDaeI0jvYsv26u=$q<>DCrHXCCLjqP|_p
z6NJ1Lf8_HfJMJ&_@K9hXE`Ifkar@S}s7BNMrZj1FrnVKE<hOXD3O%_o0}lguu&;w}
zU44dMIy>!(us#o7gLtzT#m~32YUybUIzi{i*8vjPG}(fum75N+F885n%73QzjI9*o
zF>}j&0MNhwN+-cwuC!24Q1Uo39>`&WE!hkT2nfLMM60)-SgFqj-8a|Bllt`cOz&IX
zQ%OCcgmEon21XZ<A!n0+j{yF+qsM`tXF*d0TOGmIKaOTkzsA+N1a1yV;b!zco<__2
z$2Cv>>t~pYA|t9TfrI<D7rxFnluCSI0;-Bgu1-6?uWcbS!(VN|U-uG=GC-EMClrQY
zj*ts+9rxN7uI3#^cA9?Mjuz`W^KE5(3ZZMu=%fU0&<lgFj<<+mN&o@dlo`QaWnK7-
zfWdXv?U9tw+~w8opw7G1Nm&cbo2CFG3F0W3s`jmvH`zh%5)GrHbb2_hoZo)ueN#xs
z6<OQ2p4WkXWV<+~Qs&!xM%>sv<&H~_3(7^7+ebIJIT$~k$GTu)%n}Uz;}^%vcjSvr
zQYz##m(1yYTQb6+FDx>F-9aaFx=wqT^W1X*4_Ug3U&k-_OOy7N>)fj#UeP#*Htu?S
z1Pn<v3q|>eh}zgyhOzXXc0?*&9%u!YKwPVRRIk3##iMJ@1HR^o(Rh_(DyoP3qZKa%
z4ES>`w2rxMgv5ZKJs@qLSF?o`ks9;NUZ?AObCk#{62wvg``K;afUsm3C0Mb3^hRT<
z`PU<4jEp>&4>Shk6bt$8%R;9g<lULUCaxF^`Z9UNN5W7cU_(rxpVlg{R$NZR34*dQ
z@_>qAt-@{T!x5<fBAJ6u#wC}s_spvO)r0Fy#M7T%o0h$KELObEZ=FO!)uo&_a9!!-
ze0u#0Kmp`W_2EE+CvXSCrZw?kY;z_#={ou>B=k#+bRN%Vy5|(>9uHFeS`_*ij9>GH
z?*pA3`v|dfb8k^};daDVTqzj|A(-O;O6eCJEbq}uLZ!s14iSK(1)m95^+?wH_u)T?
zle^s|cgvu>aG7`A9_uA1kYkvMsP0Qew%1R_+TSE*lP}+~t7d79pUS?GuhhCQu8-fj
zalX}=e3|?-827N2!PkakiugnF;Ku~Vzv;ZS0V_6ihco}F^j1OQwX7^vOuPYnn{i^k
z+fE-6g2lEsQT5Du8Ud1LrPP(fhPbll?#n8h@3vT&D}H=kUI_oAeV2x&h1!}R`ZF5+
z)Z4ma9aSwGwq9TZ?~4xZELlZgu#*Vc{-s_ObN{^MyqeHKbZMbA`P7L1%g@K2;6s4w
zbUokM)x0w?lGg_<>vQOS8>1=QzUQ*=lN|25{+AWwo{3&CYHvK1b@;#7V!#$7bCHpq
z?M=w?xjZR~*5>m_H!gP(qNx4kvCn+AkJA^@<B#~bu>y^2;#uEomkC6kC8ZbVk`rA{
z8r|F{n6Hx!_?q(%9IZQjU^emLt-e@pJN07jxEpclX}KAc>evk7a$ZM$?mVuj^1I}M
z;?Ub2J5Vx&@?F7)gr$rOy9uI*edDVE4%I{tFRS^>Y_U<$kj+Tkwbxm)Je3#>9MxrT
zsva2JC>(HaaD;96ldy`qZppFUWbmW&x>M`0+rk%Kt%kCnUsH*kUG8^Ou&p(HtebAI
zlz^l|&O<`_=Hqy2g#x$bkPJ4Qfw6Z+?HdtC`5vQj51j*IDVaDM`%Ep-d}7w0aId>^
z?+>6yfBltu7MrNo*YRSXFTPuJn^`=X8x4Uh#2$7W*$KP7^d{xbN2s^4es2KXJ!S-z
zR$8eHo-pV?sjp#AT)OD~+k1SXV`jd3u7SS1f|LY`_1TL*u}rJ+^XbuEB)rQZ>-qYw
zaHM=FZ`2^jn8c>rh;J276YU{7M?M%~8Ja-{l7h0)*ou6%QyjYtv-UTBkQlA_TDCXN
z4pQLeWLU#5a%R*2Im#-uTjEHMl?~QYH}=HLff5aZ+2}jxK$QW-iK=`@O#lLd`w`!$
z&Ja5dSa_rH6@G(1wX(bxaST0(N6}<JRe)xRiwhs$Z&bq@MzR6XQ~f-gnvNi1g$HQ>
zh@#p-$%(m*SSa=fRc_jXVDFceHDQvfMSY9ObhNPqP{r`gC(T#S40nkWaTuouL0Yra
z9=6(egTMwUPL7{GNfIIqFm{-b$g9MP0!hGktteD$*boFYOqspCya$4Pn*R7kw$##T
z!pGB)PPqt6Z}}QW`{;yYOp`6<vV_X>;gv2RnxNo&aKmf;6)&@&+BA<HsS%NcOgAZ)
zePW{*iMHnBLF=J%3t1QLSe<VT)qPznXuqzJm38^R(c7%5=*a&}pg-;zw^91Ee|qo0
z!heNA0UI2nM!?jet6H+SQRKc)S>O9z(vx@Z{ng5bFNUOwfePF);Gq2=`8CB<yq;oy
z+4Cu?Docpm*RCmk%ZkA^viF;1$iN?$r?Cm-TpeOqJYxNmz_znSe#>tQhwsJATsV+@
zv2=d~qEdr;8=*BXr_T$j%}kwgu`t<>F1<<+Pdpmabfme4jCs0K{>*lGn`;{ri^thm
z5xk}>$zF0_UIg!hqxI^GAMO`x4e#p{F1NN>#tES~esiE$y;F0ar4oamsR{mM8v^DR
zF#HT9cAfI34vT}26z^V9{Q!4)m#Bu^pB}pWgMQUJJ`{xoi&_KZkmYp?juxnVge%Sa
zTV%*>3VKbdy@WgyD&es6<@>&}=&S62tAB45F<5_s2~I|6KN0)`(I3w08x`<9Ob34}
zGbL3e?v8sjYoQgJL1tm=9}T|TPr+|c6NqYRKbrCu3IuTt1}L=r8)F^uKsmIMCZcjs
z^x7_d!~Bb1i*hZzd@cTPW=O{5ccT98K%+FjVm1$5<|B#E#DFX_MzHw*0Z-rX9#E6O
zLQt(u20cU9Bjmw5m6%;!DT|%%FTOX#xi0)QK~Ea7@~{e49`^Z6{!jj>Y3ZhLrgfLV
z1D|cT43wPj1O7vI244I4S`Ir$OVS3g0oaavc992f1P87Q=YMv*C)0HmWN0(ixPW9G
z{W3-Ldsz~&CWqXH1!U&5I=#*$J;0#YYbxLB)K>n+Wj&AlY#Nw*0=ZJa20q~m%fV_F
zu4^tw^D`Z5cFqZA+Yyt=B(2q^fmwMcPdaol(7W5`sBd4hFFzkLwe4597-GQ6r3#c%
zLNqZLg@9u^F3q$lq>90k7=PKG8Oqz=Ur0_&uFLBrGDeVr14xw|Ys)TRs@|JcK8iZP
zM=P~Iy*953uD%Vbt+oqNXJEXcrb?qAA&3PKc9U?Z!Wce<GVVMHYSGR0*h@>D8RWx+
z$hsRGMpIMz#r%maA6yLLd&F^NcSuVq>s^kyuQDLJ-|x82HD^{3=wL1Cj~8^`w^UxE
zcO@s2VPX8NS>qP}0c$j-SQ;(d4A<<6!{nQoWtDVSlud-bU{}0({6$Y}Ze98hr^eac
zo?djUX;avLgwt)X@Uadwf1~0p;OgxQ4J%zgE8%QZA}A(!H}0~1Mnr*=7S7j|VM*Pu
z2ypGydS%{WI;tupW_Kc44L<1#B?TEk-QbFceB|oml>17PWJWp0TqC3zd-(7gc?zwp
z434K#EInM`ofcj;%i!uK?(pX&;D&OOYf;V>3i;^R0o(u#&?dzTn3L79u`&CG*xNr5
zfCDGF&&tO^VWxr$&a|2^Eqm*tcl=-})Z%mo_5hn@_jhqAIe4N-ad!CETo$0YZQVo?
zu$ZSuYVNSgMeN!WY7~D7YQ<4h1Jz=8#MIpjjkzR%PdFK5E>hQF$`YZRI#Mv1o!1!Z
zp`X75&G%$<hZx-Ap5NQb@fOKf-j@C|H@R|Tf^v@JF}(EKQm8v96E;g*ZxOc1O4_xR
zCWv1Sru&83o66^~O#5}jIt1&G{Rw1k8zG9_Fd&~Cl+bZ8SqhjO^v*n=_h4#5@VWR~
z>6Du2RSR=`$a&=%QPD>ILAI(K3mwaHb6dKq<;&&rF?6yu7_E)XuN!1Wl4UnkmJj`F
z%6B>p`pjP;He4iy=)Yx3un1vps@U3bk<Hn5!Y*>T0lf6n>-L_OA)o2zl!r_z(6wo|
z@NoMl=sY5~QK<|~8vC`@-b0O^OmSNsW>9vu4t<<6yEsp55y*SPu|JAkZ6K0A`%Z3=
z@-9oZ0@-o(4g9p6qBortg^$^JVk6D(fccFK^f_stS+X%%q-%PHO#;MpDdm;USOl+-
z9JPu*0OMuVznK|4UzhDTB<FR#_EH7+5f<v`d1>N3%;L*HVouUAF)Nj3pKnTp!Q;z!
zXbhY0#fa?v>QIm|WGAxyR(KiHWa~;CXOPA|V4VH-Lx&GCUTkc#19RgNi20fMy^p)7
zGiF%zWV>Tq7r(0GIB5Fg{2JNjF<-H49-=%7x{o8)^wDD6<pDwOAf^Lld6%9Jc@FPY
zIO)QSHW*XxYzPMl<6Z3|XR7YzM{zO7+_;{_vAOV30GIKlV^r|V<1RIq`R_xA5ylLd
z*CfR2i$%KO9XDWwowK-rsOxZX65I@;ahzfRT*T>&8xH<=aNpN)W|TurgE?5ytL(WK
zZk}j%7rZ*Hy8E`G8`#&>BdNCJvlO&8qGCz2e19CGd2$?d$Ld?TkZ|A44fd4RM)ey7
za$YCg&-z%j(tW{Pj@}ubB*@=rRPc2KA4=o%oH;F@4R=Zf-|piS-<h3M1wY0iG8u2-
z=GY72FtUD0vA-+UY)C!3)@q0bv!dVhy~!VWhsq4u#s@FB2=d>|jLm7a6uD+AJ>|`X
ztW|BrVZ|J??&*#u?g|CotIjSp-LO{SL;QvZH*G4pUGL4Ctl@jXY}h%|Iq%Bi9pOKr
zWJf=31FQNgsPi-$IGgmCUhB9SJ>CK-4(OGGz{IZhDl#+Y@5<v@LK)zTeEbT#*Tw5F
z%GrRTIpF~Ua?=9#t$`A-H9+^-TbvJgid4xSjy<znepX(3?c6u&&OBetrsQ7k_qdlx
z*k0fR^!ygo7<WL4P(@M%et8<oc`CK&{h&ThL3MP($}zk8qqNvL4&e_p+l4Rg*gH>n
zG<Cs1k!qtIVu)U8_btqc0}#gD#$}Y6ZQrLeR;Oen)P&<5yO!_;=@Vha!yp$+-9~Rl
zKP8C<Xu?s?oVUF-4?$rc2#q-p`&A4zWjtjl-zl4RdlIe}ezYMu`aReZT&<YMYV<D1
zTr7jm$y^P<kFQvQ5z!vw7yS!2<4eI65gpn{y*G(Qy-twN9A7qc&Ta&7q_l(@V*zF_
zWh}1X?K;Sh1OBcj;&2b{oK$DOeA5g6`OqXc<TU*JWws6~^D!HficOuV*|VG?s`d6I
z(^gIQ<=q7482=pZ5})L!HG`T@-W_MWTdWiLQZDpuJ+qmOK4aUT72P@GMm<?fYhyyM
z>v?qF9}yeGVSi}&*+1%Yp+l4=T-51>j_tPZMa~(z`|f5W)a>1|U+`(h=6P;hF<Hvi
z$#>4OyZxaybqy`um=89ueQL33^{<&{fjKb)1_#~?XrWlhtN0+H6h}>Yv69s}O99k$
zO4mx~-F{>B2HjZ*mrS9~Bew}@agBd|XCe<Cz6YYRBj?@9CVMN}+dZH}VnSjT){(3;
zx(ZcOQloSxtbCOxOZ>=1X`!TrXFJk*h~CWdWsdXyj(9siWCM-q(fG9E*JzAt9c|!|
zSgyO*{cn4PielsKPTgPUKcV{}m^RCf>-~_k(N^jd983G;Y583|;pJ2-Xl$elFG0Y3
z%!~*ZdO{~;C7kuk-5QF|x#LutLlDt}K?>j7oebsP7%*m5f`P&v*I{6yEAOg6&k&T6
z0`qP6=xw>>;ii0z05pzbWnQ<N2?%2!3$Q$(Az?LOG*Uo)C&vjnVeS~O;PMfOVx`c`
z|6x#Qkzd)t6aG{n{#{tmiEi*&t{k`2%@Ss5ChVBDi@lV54%iPU?u5m3b_mgnv7$?>
z+*gSC7a*<}m*mH#<i@0!n_P6=jImU`7njY*C}yIADEM+Fr0X(4#?e|nFpgxxQdkjx
zM!DOjGeVz6@me|h^i&XYB=u^z;o@rDCp`DQ+0wN}3CePCe2D!T#DjYpo-HLL#~NT*
z1SlBcf_H=V810i31++){3P~SOC{lTq`4}h!5Xrv_xlA>BaBYz0HbixH+n;a;pDem`
zYHH>{Y`2D^eX9C6Z|r)kZ+IJAZpz6TSznW+E8Jv<bYsU=n=S7XSZ@P5>y&%g#x^({
zycE9P(G$v`%l1MpP?Ow8J<GR;%V`fsXk!XjFHWcDPoZB4X|h*(n+bndDqM{Q%UoYB
z-)5b!V!qEPtXxEsJU*5&&ek{@9&GVm=GNmLJj;84oK3j<tJ}9XwioB@z?!(%mE5P>
zPh8#K7U1hy&W?Ik{ym?#R0nZ|j2ciu_2u4{tkiX!y4>Y==AtJ<`0_%6^W}5;mai+8
z-AAWBTZYJ8Q!n2;wcruuJ(vtZomPb$&fM|QwXn8*5>La9w5$maBxc!0#+5lH{hohr
zo8%<W>Wo$|GQaTSWTlh<g4xlJzTn|S#MoH%VkkvIPapnh_E7rCnC&sj*&7-&$*u>}
zfBh>#ER-1V7)AoaArnYcdWh7E4SLYEGAg^Pg@&=gTYfWP6HbDQ6C)9Z^#XX9p@OGh
zIGSPWsGyNHjw@wdwwOdv9}c@QsuB^-QDfpdt`epKa^&=s``KtfHv=`QY$_(!dw!|O
zwf2cb6%bp7z;tODI`A-I(sI`MZ8G?Jw9_&qMN}I3_vm9O<09r!k+S<yZTT32?smZ|
zd?>LZ%~22F*+>9Lw)@)EOSB^~QJG~v4Hzf$du9Lo$UD};XpbL{gOcBa-pYJ41%iMl
ziCOVpq*3Oe3_GEoF<LC$tXjQ=ohu4&7a(LcalS*Oea$_Z;)h|8wDO@#c4F%7r>;3V
zQp?EU(zig_oW;%y1(>mN>pmPET{|Phw(}phyyG2&`ZG_$1NL`z*k<Pg?q#<n_uhXE
z8bLE`CRAyh8*CjFl=<9?O2DrhCvvhmvjr?lPX!AG<k~O%H{nssJSi5y|KLRx-kMiA
zEdQvwA^^&Qy@$K=K9rR)lf4y94_R@@E&gCyyD@lK`ZN2<3EM!u)z!>|FyybR7v-~l
zOs&4*_4SI@Oj~HD;4K^BoiF$-aId*H(l$8IjqThZKex*I9jV?G;k%<_9lNZoSC+5s
zU3l_P{8jDW-jH?dW1>jDRSgNg5)axPvLsu&JyWI+|Dj>--gZ7NB_FtJ;Vgq~F8#kO
z0K#xlRjsf%HPWqMa}nJ3{e<8?j}-#HNR4qSk<g@!sVx9On#i@a{$s0Z6bu%JeCoSm
z9r25SP{-L$q<gJ>8@Y5_8(WhbG9pNUhdI9OT4r+m67gi**#{JR+y5*xZNgX_rm&2*
zzwa`hQ51QlqdPx36_o`qME%qa>HP*@XwQRfg=j#`L_QTg-tclB#Szs`-XT}2>OlQq
zZBDyGdc`RqBbF)kJa$kgDD)EZukVh%ZA#C$6)re1B7lr2P0MpyRxc6&DA5!IePl5_
zf81VXpn18%uW+^s_!Rp%HYgX9y-XN<(Wg)X>o;<9fZvM{S^S>LIFdlb!f|-QG#fdZ
zvmdRXj}-zv1x=}N<#~6I4Zl&4sF;QwoOHycNC6=4w!02U@F^Hz4*hkzo_q6k1>J7y
z=e3L`G;*j`kVi%c<CxLUO2g<`ZxQG)Bg^1_qyEjhcRNgM+gd`dJme?RM_upKMlW)2
z$nSkWl#?Hgrok*=`JWV@b?CyQ(F!3unBe$<%egZ02iLTRDVMCUBuQ2UzcuSA0H;c4
zQZ=Sy7x8sDH=M<*Bnu)t?x=oM9~GtTy<Qqu#xJzIHEYYHn5*`if(hNQa-j$)X0U}D
zuLZD<A%Fz_$7^eZuujH*c6xmx<c^^$i*)H-&BYUt1(~VtF}5+OPrhLlK7_w3Wtz9M
znfc>VOs&6B2n3oQhk3;5!XU9GR!>P8*L_G<@`P#VGJ!@f$u7p&1AL10l4_F)l{;d_
zzdwaB#WOz214x+yWV-Q{MoNQya6P5-5C&BlkCT7!He1yJ9#Gh9@WGmwUK133w1^B?
z%Mh@3lb$iz>(l%6K88L_0488s>QU???4|<s&EglL&Lc22bNF}tl4ABHMIw3i>qSR6
z{4xDE0Zro>F6YK~%c<$dwvDN>0iB9$(E}>1<D`G6Ox`cFrrsk7D=Qsq4H8X7GM<>l
zynLpUUXV#_ZuEs@2><QK9liHrSa@>U4{ZA-_FI-8wRw$<hsqNntbO&qZx1I_IVHp4
z$&p3=yG%E^mlEmRT)3*Saf(V8EmpA)@Z#Q{4K9+3xckKje(1Bmd&99`pD*Lo{{Ak?
z*(L0h&&e!Dh5IFo;GWO7g_!pquaD0ppv%$OBPGO*&d0K_q|mLUA?y+{-ZueD+5Lak
zba|vU!uwOnYuuYo2j8yD-6oIgRFLGlkhYZd2y7lKw8QtpwFW^D_h3*F{ruC)(}LOz
zlH<z-$j`+E2GNe2%=S(QBUCX~lR<`cXmMBplinfi_5rMz>VoNRlIAEU5AJS=czHKd
z|2s!<d25TXYux8<T!AX^&?s2NENRvz<XA0OUumi1V&O>!$MSuYi?`d;z?1dnSQ*&(
z{c8-W&YI>n8QigU=1L-EfWr}@jTnaCTixE#cii%BM-$kXnV||3FtVLZ8ofNNoul&x
zav^svcZXOk%|Qz1LpVzX$^E+f`o&8axgn~M(>9#<Y+d0_Ed3}uxpwmS_<C9wr>g&w
z;mdA4jxxR^)r#_$!?5pJ*AgiUpQrIw9%z9T7!97vM0WnYX~$d#<k(bR+x_F-A`r;0
zr;+Nso2x<*`Q%ROnE79f*!0ZYoURNP>4ssh(D#$u3yCI-YaO?fA@_BiO~LC@Pml)Q
zfysx<*YhELp(-x-_gzp-CvnL;1;iwGw+>1M_-Wm9<PMfu6~c$7MU}Z_V|kZaiFuj+
zhD%ClKX{`r_}PaqKURb>xYoPBphcBVA*YiaRJY2ZYA;*v8nl)YiWO{*_4k|?B(yR!
zW^JV^C;sD9Z*FE_bo#A-r+amb6#+R8sTt?(%(%U1Tf2Y&kYuET6&ubU@(i{h2yx&Q
z9#mN)?}f?s8+;?TXR(AJ;tj1&$WYKHsqW|Rh{YWUHg>!X5*WB)RGf&k#3!Kh=Xo&;
zPk{vhy+i^$a%MZzJR$@x;V1B+80XqxG)+t0s}P(@1OgTX<XPihb2|Cu)l&*EQ^d3W
zu|h<`n7G*1s$b)=`tzw!9O455BYg1DswBxuFOWBxesSFBBZv)nLLdg><82=(uKdn*
zQs6Y+%wlRMUTH}n4%8a&Jh>%ca8pnX8N|`0rvNGAHO9nXcgR)#&@H?(4>N`t><vYj
zA;%?w>wW>A@l_W<Ht)*7O7cO|F|Q4l*Gp9UTpZBu5W3soKP?FX;Ha(2Rq?Wlw@xmM
zn}t?bC>V7RchP3{mKoh;@067r8ySF1Z4}08EWaPX%@=<6R&(uiqs#v#GPOY}?ua`G
zd2{S=@uo~_A^OUMfur1aQe7l2gL-oq<zFZvqvPUz{BtS^=`oJxMFr5FWey4D?8aM1
zb<*|3)jS5P<|8xY!X<^X8->SjO+;w1jo9hj9Z5d5i`z%Ry7OD`ZIQ!&*qKizWe>``
z%z7QmBL^>02j<C*fPHfap+%%URctHXWWC4ynLBw}J)-gF)bjYszOta?8yuUB#@%bK
zE2MGd%-&`xU7Up_w6(NdM^;rka~VY&G&6S!KP-3Zo=Wxn!02@CG;fCe>vMRDl84Sh
z{&Zk2XQ~)?TF+h^v8*dRv^n~05fXB;gkOb+gI_gM>iiQfHueq|b*eGclfhOkStBGZ
zul+l*tJFQr+eY2~GEha4<Py3N!Uf%pu@1Ox%#a%0=<U<}a`zrWZ4vZ1c<NW~%$34j
z8HUR@f9l4KkfrTvuB?;!7^8Ty7oHx@5|r0T;}&F~_m6utr<oSr@z@Y76v3rY8L~I?
zS)o{_p@kLsvKmJ#L^nsCxcy=)eewM@9L_tApJXg#d1=5(YVj*z6nAec^MRa|e(=S$
z+ew6Q@PK)OLuELlBCvGndS$;8Jm6;^G71rEwA*2NXTTE+AersBl<nBhkOEkxaQZB_
zT|MlR_fIhU&`G9P1g8a)gz*AWw$EGItM%5X1X2}l1w-_JFzx_e>>+NNutXTVe%=vO
z(Na(Q3vP`@RiWSDPaGbX9$SmDt(PneyT%3UYLPCKHZ%;%2;NYeDX{)wSaW?!F%MpJ
z5AyX6el2=mx}bZhN`f70J5Fz3o1hQDX#3<c1aJ%KPSoV4n~ETc>S=-q4-X}KwEPwa
zJk;1MZMv601<$=_pIKh7(bKB!3~wXKcY-wa{7~fcJt*D0tL58U`moVx+_%tequ>qF
z<}H5-LoH*=YG|#xLkb%1$Z%JDX18KCL39akxvkx=7M_G_X=fA~r62fzAN8mi<^$2S
zF!ZVsu5)%yEyCxLPIH3dUgGTIi36~&RAz|UpTsDGHomb`hL@uoje`rXRzDG)teGJq
z!OVYq74dEhGdTN*8p(cqgdTPOT5<9b0?S`g?SIVRF32ANFHGVcu9BzKRC7arvX6zY
zoH~U*f$^$rDtiZ~Fri6%1RYwv7l3!6!yreTv5&Cb^c7rTz~fo^dUE6EiN0oG5l|B6
zj^FKLRo~|14a}=m@i-E3=&fS6<3qor2BW>)xt@H6?LSD*8A5BbbeX4xHSNzfZ;^3k
za0rzwG%i?tT8WJ&JP8$4en4$q1aAj^D=r6HdJ1L{3sk0{!gF@0H-TPz-R9$7o`56-
z*PwJ-Jm3LUE!e`1MW-PlKn#$E-jj8PC+Dac`Tkbm7L~Dgc-I&(`wdC@+xx>}qY02>
z`uOQyx4e#)YGc{MnIFeSmW{s-M)NraKaN%|00#*YFU6$GoYP8<-TXB{s)geQ4+8j)
zX3J46QJb;<Xrxpbi<1pHxZ9SK&s4U{G;RX@ejIaS%nZ?mUFCZ*(<xU6&%LMwTL+gN
z+^b$yYR~_B4-2R+Ki-k|{#eRzR6%`b_@_HGqp6lyFcXsObnL9^WRkFH@~)@1GOXQ*
zRnAZ#L&7y=3n}YNc9@&%sSrQ##NnUU$mgHmJ>2VQdI|jJWLkF21J`VpS3D&(Vv|h1
z#MEpJfyn+`Khb14PxI5(q6B<)NMRhlb9Xi<CDXDAL=Gj)7X^1XGt(G`rm5O7a-%u7
z>q^DS{3mQJo->>L{VF-8<K7XA*2^&H)!>5HO=o*&*GeN#R@Yaij4ZV(_=x}26>}K$
zKhf<q@2U@o_R#n=3vbz?5tiy%C&6*}Z3ih)$^~%Xb%6P@2H-l07Y`L9mKr!?_u$1K
z09&w{G87NVQo^P8lLHE}d8*N~8KzOAvhmx&RO@5@;K!q^?l)r^v6nC+?Fu%#Fl;K4
zl7XG!2>3J}=iP#VqBr*tPZf~EZZ4I$EEz^bb&%3B$VP`yx6C8T!$0)tDW0Y53Wu&2
zyv8-GQpckYGz;QjYu5ZBU$d;=io?b2`ZNevgpKt>;{CYBm@m8MsHmr6Jc7V0B7=tj
z;9DzrL-3xA!uqJBKyOA%LyIo#H#mCby`%J-5;Dz#yi~1i_agld2hw-4NF#U`Q_Wkf
zo#6c-BYeq#I}Z<H4Gnf39uqr4WT8biv4AKBA|XW9zOd8h30<WXaxE%F)){;Alr#YI
z)Ji-tGhr#!lU+a02(H_dhVSpFy6kAOa`gVPIp;_^U^OP~^%suG-i0i$Z;{fYT)Htw
zTN>ZK@sB?0y)0-Kp39I<fAorIH2+6huaDv1-zaFi!eNly{&~to+|OL5q9zUVFM{SL
zU-0g>|JDkg`!z_?H^jvsUO1)3Xk?_2Nw`?0y5&&uThe|PFf4!va7x+?cK+YG^Q+BQ
zIzI>6Xn<5qQA;y~|0F+i|An>UH^AXPaMOqVUpp^Gl}shdPK0_?NLWcoNt(`~<sAam
zO;2Q^@^5Rtp|g*ba_{zMATFlC=935Wv$!H*hx`MZ)8by0qQ?cxhACU7c9v;V*u>0d
zrQ7^n6T3(}%c7jzK&0jnP*v+4g3^pdP_`}4U@jmaKujW2Qg2v)l_cVWo-tN=42L@!
z*_tc-N}<X2>?qJ4OxgQ(<!iwWPr2ZTdlNi)lq4S1{onKlk@6oCwWCnm@&rvcub$2M
z2bo&no%8a&eHrlu8?uUiV|j{dKtJrIISx*G3y3zdr}GIcBB=sH<q24UorVT*$5;X9
zLwz5A*W{aeoMzfkDg98!kTydpyISwP=)I3!y~|LWfePTcE=<wH6x0dU8++S9FKP;S
z8r_q~@(M6EhLPx*0SEvyAnBGc4j9Fz|G#EbKDCH8z+bvgK@-3TF8~pA8ZaCwym*Em
z3ASA!Ly(n|%!&cX0hE<4eESpX>~=s%5Re~6*3H3Q_`(gx8Nf8vZR@~VuTSp`FR-zW
zH;Zh0<Z5dQQEH^s#1X+UvhF6eGNlBB5!WQrJ{o)GTr;BIGyrrBH-+eZ%U4VeMFDua
zE}2T>yfs<*-1qG&hxC|3mxiz60%}4%53L)DT^?iiK=<Ep$t4svX*6oETBm$@oPg<+
zaQNyqlrCXoYmVyp<7Vz!01NZ1wl`osaF}tVV|n?BwjO5M<ke0tU0)XkN%nRe4kBX<
zdIO*}iqya>**RCci^asa?n`QHr(n?`*Ez_|TqMi>OiwVjJzbFUj-~86n+wUe@vpv_
zfz*{8bW7gNU==^;#a(D<s`a5F@Y}4mXR;@=vU~E|X|ykPUg2{30Hr=^TtfoIi4%x`
ztd-rbX1d^ILymlJZ_f|bTh{HG-AcK;B{PbS?9vuVh=%@Np_&>7(RSI~aPDZ<F@f$@
z{&@h$n?)fbXSZIaqb5$}1DQqg81J7+v{-?ppu}-J$?Bd2r~BuaD*-bYYU|2{7e<1%
z2jM`u4i@L#j>@I+!@;nE0whNs-WU;r0HWd=s*>oDYy%;9hmfHDgsf#?#)Dtr>H}f{
z>WDWEt9vgH8fZAPK5i)#AzVU&=qK<=<gZF{drfM6Atf0}t#}BeBjE(w$%5Wobpt|4
zm4B-kr5f{=Cj5Pg@G^r^{=iAs?|$>0{I3B`+3y}TN@kquP@<B|_9T2isPz|q+-?6M
z>i~GP*F=r8YJM7S;YIxBn&t}bAJQ}L9F`|g+cHJ8d&Ohd0%JN3wMqaVDBS;tF`E^e
zFHi#{+)~=r1EFjM1I<Y2ft)Z-9$Ljxwv&P85hkFC6T}9P0uR0b!tf0V?+_9*jYq&)
zo}gV=;7+T6iIBcMRvtr1ujUpc{4vR;j<tWxAW6vomYc@0^ux>WMg0iw{wxWR<SVjo
zidjGHOt#?hp%PA)H}G>49=kNjx8sy{A`b}P#|6#29Z&x?1JQ9=AK7jTk$TzA**GIG
zs!fNo(wUo)?I^iUD){wh`3LGGlrkj5kflJ5x$IWx!A!1rb6;v3;dZk^fHj|K*2DeR
z6CQTkGfqQt#xynsa@Mh~i$K+pZupU-Jo6!@<<f=Y*p?}SE^p*~R>CBKKHV`(sLr=a
zKnFG8__iF5{ki<!cl)p@JY{{rT3@TOasP=U3hui(Wn@PYRa9Xyntg1Usl<-nx9QE^
zc`kuYOS4?(J5KlGg2zZ&3u0cB+sFKGVq;E3H7y{nS&}|KCv!*u79TxJ*nPm`I6ss2
zJeJg$l9q9V(7K(Y5$vY{0DFa!U=#G@FkFV^vT9}=U+9ic;QP(T&?ju*a(pubFWYkN
zsqQT{pz#G%uHThBpo^UvMQoEVNko#+`H({)raeZ63<?H3d!&lz4oqenLe%iu;u%=L
zXc!*Dq|9q{N<pGJR(g@%<N#0%Xyz4oKsAsQ!Td_e4w{Nk+~wIKsI1A76$vFO;FmHR
zB%c89<oW$>+zaD<#^xJVK<{YXp7jZm+ITUp%I2BDx(!lp?4}5X{JsUjHezj1E~ZB=
zBed%!FeOz~1nDb1Ic12F*{k3Tk)XBAJo{u6XcD)VxTN5gJ%!V@#g!!8^zj|X*i*$Z
z3c4uUu<_RhB0dZ<t=3;ZlvQC*FDbicarbnowRoy;Q^kfHC61Jd8P}I6>jr~g4-I2a
zFeuNX9>Hhul9n?x)=^ed!#@L1urfRu+D4>1)Uo1dC_ZT!m|w`gd9PcAT#O#AG`}-2
z6L-YUY+%EkRTli2NR5Y;oualY4{gUsqQV^3Ukokz061{}9c*ZMGg^O`GZrK5jx#qA
z5H}N(Sa{=BGg)blJ)yXMT+;-i$A3h?r~R=s>+rU_vD|BIc*+`H#IFm-G{;=}e=emD
zG>}80l?HR&fm!XWa&lxA%}ugG^B&vj<5k;CZJqdQl~xmSJ@}_-BrO{Jf7ka~I!jH}
zX99nKQBN7^gcsjXQe;7QM6fCnHSx#1-49km06lRW=L?)<GZ2V_cZf7Bocx&{*f0`?
zw=JW&?Z$ySrpMURCQVC?)BBO}m_NE1cnwp^p}}$kk8(?#jk#fT@FpyLr*Ut^SV^b|
z?{KW*?aqrA(727Ka!<YRLT0&zgzyOnFB4|D;7K3~9!()o4Lxo!fNESGiU)W`Fs8?a
zBl8UQ2#4`O%(?(c?9%_k(_8p8;s0N}+Zf$FdNk4?U4oL*-Cd%9G|~g4k?uwikZwi~
z>F$u0mTm+D?>^t(z4u?(-jD6|I<MzBhnxeI3|$%pfseF?P$$ImAA~;mp|h&B065_o
zJ*f*`O1s4?*wU22OucyklGF=rO|*%tE)(lvIfM`0xjf&5K^YKIq_d%~LSicHfhuC%
zwo5J66PDHr+|IDhQH)@U%U+k(<FxQItZu|=dsdn(;*j{2KMtp4XII4^lngjHa<eNu
z!$gJ#)H@B_T%=o)(Gb2W(%W!Nt-<C^x+&j7PUF#hA=I&mR4x<ZJ&td|SocZSc8b0J
zf|!s@@E=osu3|}XCHFrP*Mh325<@MC^i;^d8Ge1YNmgh3erkt<fW`SA3;~XuTz?)0
z%JRl<O!T;ru8)~6%(QRG+3R=~80P=>E|b0czSzb7A-`^%$661#GbQ&QekFAx#UY$M
zZdV($9~aB}>-~bm(a?wI4qPms`C0jD4U~!w^S#VXwa0oN4=8kZ!Z~5@@sqo{N=+~_
z#>`8-X}QY6-e#;&a07!tKV9i_Vk<MFr*_40(Qpl;ud`<Ny~Sx1pGylR+HoWP0rDU^
zWnWXg*#W|8v!dDb#S|B>#KpP$=F27yQ&PnNA;QWtsCZw-?FN>!=&L)SDh?i)6MX`q
z07C;I#!McIHW$3Sed(i>n_5tJAX%}ME=)lI3PMCQ*Q&IEAUGs07Qp~w1Drmp;3#0h
z>yVxhl}2?29#<uc#sp+c8pNF?Aslg3`?d`xjt3?fq9-U};ROmCiCfhmJPOUZF*-P{
zYRxRn;LzW-`e30cbO0sM6T31w_huWA1SA)U`Wp1E;EBMW5Wyr0FgobXl+LjvZ=FsI
zK?wgyE#uCM9Knf;_YjUiMA9YDAj1FFs+N1-Q3HrWURu>1P!FT}Eh=oa=7njJ;z`qc
zVQpeqV7aO%Y*gb0U}KNkW<4B0_>-F2xGZGvFW;4QN6{7eSjK)9?!A02`EoRx%&>us
zU}79`gfK>cu9C@{nSqQxT~sBxD|@Xxh9k{volE{VDp$padVI=>`{f)T2>94lmNTry
zGSOVVP?ml@vAyv#F)6HEE__uiuo!FD<T-nELGGx!tbZzlA}==9B3N1-1$;FJDth_z
z#EZzpP~GSs%f|Q|?29SboYj)E`|tY9bNMWBSOlDBV3vhT4W(nn8JE}Uy+qoFD(7~Z
zoV+}Cm-on)Fn1+m95VP=o8MhE=*N0aHva=+rGR(*GmDbF@P){ORl#cU|DIG5%U-fc
z(L^gw{z`IC-{wG+2gCL+sfxOfmal6nlT(!LTi3n6O}-D2!uRdmjMJ-6o+>)Kd*zFJ
zpK|cwH5Iw_XZ#g68?B>u?8t>(O6sS=byhdsMumhE>gvmD+woDUhhq{{0}RJ^G+sIM
z-7sU$*2a(*)2rluvV82k1=jEOX_RtsGw=aeX%@~0&8}lfFI(5P&{Z%mf0(#b;=!;e
zb~r#9YPw`l6n$+Nw3ag`(d;H2Ro;cmGM?b*#}fXW=1h}hvLAQ2{mABC^cDDJ%_ivW
zCJF$!iOMI01){KkL3v=<V?>v|JGGx<=L8A<ya{NG8Li0+q#6R25e(oZD)Ze$v;kET
zXe>e=ETiBe%wlS}`(xsBgeYupL7L}Yg6}XD%Tq$#RReWbzcC;pA3ozH)OoX6fc1^l
zh~Yr(zoe3W$iNsH-mh&fU;r7a2@q+oQ*1*mc3phkq}7saE^a(i6!URrMvxIMj)=U*
zRFus}2Yq{&-e!^%ckuBWbn$)pAi6vYg<tPmle<vsR>7ry&kQ^}2k8uXykyq1m28aI
zj(|@@tHglerwjq^(l19o-F6K_-e2!b8bsDYDhxe72iLZ_eoMuQ?W53i;$E70w5btT
z^UT@}%xG32i{Yp*vo>md`g<f0c!fK+09eqF2rcFtJ6loxiOiw2yXW4?qI>u3p^t4v
zmN~OCQNQLLryfd9dgWxzGv$4SbJplngL(&bpY<vwy+#cI#LhL9#cdw-EoALI22mRR
z{W#+=;&i$921Uq=P#DX5S}@=-=bh!|>Cm@^FBi#X_I~NOHcwOfrH{_pb}EV}sNac&
z=M4du6@mNv+lLWa=W%dPL}U<<H}={Vj>6<{MCV9g0Cik@ML%vE07nU5Dp&6QtGd!6
zV~OmcYUF6oc&@w*%(M>?=TL^lTj6|~?N}l!y7!-twsa?h^PW<ctZC<5;nsDLZ5>`t
ze!MN61w2r*1%@4Orpx=bPt>y7i0f11wK~lNDh+tXX74ahppRiGr_guzKS1e@6U(iV
zYCnEIFfnlYEx9@?NH%JtGYKG&4gcgF^cx?zX6Z6T7A1zki3b3lzwW?K?zh<&mYIH`
zFmq&jArEGcjl0Me&)68ljFhuF^9$6@b3d64)bN%WiU8q~NSGZ{Y-JlH_>APR4`A}W
zoswbu3K6y{2gK?Kww{&+F(J?))amMRl&%-A<j1l`@#4hr5Wh?uF+Vd1VLvtslRTC|
z0Wpn|M!T};-$!rcj*;<p6CaY!AmVe(y=F`DLtqMm7@HKN7T<IN;P#Z#$#&NbUqw##
zZmimAc*%JR6&wzu=qSmqbZn|4;-{ZHv^J7gI^V|Woy8>wC0z*<LNbdtp5e5e6Uixe
zS3AtS^*q`hL1|JCV$lof*M2CmvGKwQ#rgx)wh-6k>NO|P=gJko35~oKdl9ED-{czY
znFg8stN$V#KVIHYF18r1i*Fw`Ud5)*2ELqp%vovKy7+DtD*%qnp?0WV)^O|zbYaO}
zZ;ZjO<jFo6Uyw$wVsh^T^|F^3Y>jl+IvAY=dF$Ex_nX#M#MpLje0NN%b=P^WR9sj+
zqXbWF`Rd8za0-07jqa0q{_jgy=I<MrPwl3yiI+w0<?O;*{7Y0YM6)`9NB`2&^xKbs
z&v`h9@Y)bVVEbnembpSK;-Z=k5Qx&xiieGdyOo+sCLD=P5&Gx!)w$_97lFeGNIF>v
zf}v+%>xkddQr6A@CIA*3v0J{<8mwhn`x>BWp?Azq_EsONqz(fQD)G{eSQR(!@>g)T
zn3STIP6(!;XnD}Ogzc<azo_A3v8W>u4y%{s(<*ODn19_~tL*YZUWL@e-jJHa#Pt{D
z2ewv#Kn_(+(QJXVBVkoEI02EHX1!S!PY!R^>gIHgLgC600Hq$JFYiMT0~!Mi5Js?V
zx5l~W3Ap!V$V~4oK`$yArmb#}t->)GpNz$30!Esd##joy1ttQ^5f^>pfQc<hxmHf6
zg{D*}f&OVl>yfE?u`{4mOFIBLB#lC@><O*yQliLakC*92#ChEqVAZ1E{GydsS5}){
z1B`??gl{>Kjhu|zH)&-40jsMnC+In(UQj2Kq3rd{<Xgjc^ep_2mmR|`WuqDYs|6-!
zo-Dp$@HRcYa%_I-(5k!t(!0GP&DwOi<)d7)EtjK^MKO<sjd5i$r{2%_*?xw-w(aHB
zoP72Fg3b1hewJJp|JM$TgCq>_)^VESxP8g<|11fM{FHjQa1sArkoIw!2VLW^3kM`j
zp{kY*-*wPX%%)y>B3P61^(`(~_FKf8)S7I<vioD{G;x_I{>GhyqkiNutmlI=hs<Ff
zHtVU=PkYZwlD<}NG0N+mXX!fhiDBVxb&Kk=o>{%~%`lr+^BnCW+CpRHv@?wyekG23
z{Nj^}GRX^N@n^+;BfE>mUTk-y?R?ep3)6J)7mQ!HKlprIk)sOr=@LV=3c>&+(En9|
zVPNi|A|l7OPlc(-D~$nYGW|fxHV~(U&t`ZOU@vYBIu$B?()Z8Bx&FsYoD%SjO~HeA
z;|Yci1d9`~OS%b$qV)8-numR0UDW~-08*gX2+T*_U;vB<zy~Ai69hsE<4Q*8UWupr
zyKQ{?+tk^HqKV9#=I${<kN#bFQBZsvl?j5s76kxMVDFA}_>?0ZW+W3wIfeo7?F{M>
z1&fES!}eJ5ug^WpP*G&a0Og2!h~(={U+duWY%k0An<b8A(usHn35B-G)vb!Q!}1A!
zo?0g$Ud$y3#n(x}(;kKw1ZAd(p;nMe+TAVpn)!5m(!;vAgdWvyLHV<H-ixVrX)1+$
z%_nX<?91=@OrR$BCkUYl10ohcY!c<U<LeltUrG4efy`p33W-qao43-Ps|feGQ7GPW
zzS3d+)*3-$9N!IKv~N4zd&c;l{C9}QH;=lDqq_S=393B&wm9ju2Y9!@#@D<ex;|~}
zOBU`HCG!E+xaqy~%XIP#?hVw{RG1O{Fxss2kkyE%w{n?cvKTopwI<}jBb*ykB9gA|
zTJ<p5pBPfJxa&@nxsnt_WP1M)Sq)Nqtd2}({%x>IE%o9oF%Ny!@ZODi_lx+1IJelp
zTK{YNGI)xgpZBqbbIX{$VBZF$`~KN`w&PBBb@lziV}CcW5^t}9s}nufPpwltlT`JR
z^WM|8_2WZyHT-T+Bsn0@3u%ie7X2%6VoU$><fUGkcb2lZUs~#42eCOesJCR?b55F5
zx%+d#OWht)`#st|TQb)d#-(n-#v=HX?c)V85gDgG=QEvaDr+qRp6PXfC#S8<J_pwP
zK?}${$2_1$R48pvd&cown`*$9*_?6FrCVdx9&zKEe%tK(C8}|~^VUiKk<@=-lYz@l
z{~o*lY5xQ;zEo~{A(Dtj##fkG_Di*7^_@jl)D-;DHZj!czlD?^7fO+SHvJOSw7XrL
z+n9S&oeH)T%kn4S<H#RFYKqF*WRSpkauUn&|D=l_e*`p+QCZquFEIYLop=76et)}0
z{Q$^Uh#%#Qky<gPp8{@bBu>-9zzajeN(ZMSAqo069_S=6Q<)#TI@&#x>Ne7Y%(kd$
zVOuE)7J4E_!MEt3SGQo5O|No(D?xz3dTR@D&alC<n3jeokOT$-;PJl1sl}k6wDLK4
z7AQd7v}LmZY6RcL>ySI_tG#j6`XPM`BfxD9VfeMq;epK}6th!PLFd%^gTEHF0No-|
zw)+ja@gZIC-uQe<Y{T4!f)s+DPc9)@cGJg1-;|(5wg3c3>Zb$hxE6e#Xh)pXn#v=T
zFqpzl7IcL%dB4sOh|7sFN{eF-3^gN?-7^^9HY2;OyQ}uWE+hTc(+floKsHS44nSFC
zRj3~?GN~*1^fzl;u#*)6W0QFIJqX4NTKw~z7tHrnD_HgQPjgwLa6&6Wx6*lvL5R<O
z8EHpjOR}D$!~VLj+kteSLu^R032TsvxsiQG_}8;)xe14Nx4(J^WePR1&DKBqY89VN
zAor9M1=y^UXRi2vYFMdx>iADXLEd3F?}s(Rbn7m$6ytGUC7ie4Jd*-wO_4$va<BDS
z&i>}O+N7B5mnI}+d-)xSj-kkNH(DstG0VKceqm_BVBWGww$*X7!}iJL-$Tf<lCPjM
zAN}3Nuk;Pf^~dMMklw(vg#p%F*>uAlOT358Xa5+TeVX>IfgH|<v=*_FI^sRmTi46;
zI0H0muqeM}Z0yn0-d}&p3~!B$nj6*Mp0fy~5az%N$EWKeMp^-s)%p!vXN+2o1Zz4B
zyRWzZ(?7yw%9&8y|97?AJ2`XWj+X+;Q=dZ_MdiI9kO}rUt48W_Ha#G4PzgXbKolGW
z6EAT$ttnsC17;!qMzVlPwxud~$aUe{3_aCL3w+oic-u&&)~^_!28KGkl|ebOy_}dD
zsJO;|-VQxlrW0$9TJ?K@AN2_e0OKy<6&X-g<H87o=}q}8%0bF)Ac>q&HfhRvg{g~Z
z6aOYuN&!?1gxGf?|A4-?g)~d7yfz78&uv0Ym20T(=}FWmVex$FFp^@9T~xN#O}L2_
zut3Avg-Jd_8kZbhmJ>@FA3{CRlnTI2XF?*0ME6)mYhbtx<t@f?wOe_zPN$y&OhKFt
zOXEd3Pi|>Mbzorr3pU9pz&R;}rK_Nw@3sh(IF`^ROa=v`fw_QCx}1k*f$AH8)mCYe
z(bG}Oi!u^P98}ykeBgQ|xYq3YgrLya5j%G)pb6r#o1U(-j{RJ)j0TBX<g3i;T>JY{
ze(Rh0Mv-Ron>@cC{!K-TXMvrkeT}H{6=^=xVk%VbNqgtp9`!iZ##yV|J1aTnbpNR#
z!jjJUU;v{WsyXX`A~kP)na`ZarcFSjj~`R6=dqDDuC3FGp?$!4+!BB>k>e9X${kDl
zOti<{vidul+*|0U-(l^GajZ`UbJ)z^J9X(MM~l7ZeQR%%%@|y-Jvox+y%sRJ$3l66
zevdknD(KZa9x5#$tSRiw9mDCqOq=}g*XJwe_N+5wAhXkdR(YZBWI=*lNa17yd&${U
z#@SZv5MbkII4XTr^#>7slY&aD6Oa|a*hld@+EXUsG7f-`JS&cbc!GMt1e^ZQ7K7Rs
zI$g#5IgBR@$`1+CkNVtYc$H?slzn}Ggaa|X(AC$N39e(<sq}xuqnbs{$(BHBQ##p`
zc_XGN0E`JzS8Y^ckX_6KE=g#y22r*xc!)a2RBNqD9AJw<fK06t5sCm1Y$X8_=9tmB
zn{w3YD9J>Ejk&hM2oN~|gG8D%@Bt(YBk>wvkNnc<#gYWyEB117a>H~^!c}xkCK)>e
zN*DomT(mU#NUuD-+7LzE8-T$dR*G=9{H8BdRfX$XZVkGXTP*1Q3SWLBE1A78WLAI*
zoWN*ATj5$c4BsPkto8$rhucXR06=!y^}bEd2PpsbZ1<qk@yTT03Hkj^JZ0+Kj>DzQ
zv>OiItwR`Nymm#p7O(adXQ`|eb1Ek9jd4Tl-S*vDi0a82%FasW6NLizPabpb4>iul
zeeVwgd}pf_OoEVfD$)pOVtPf3Z67bBty;sSdU~Cd-C+2^B9WE?j;t6|1IRZbBI{N|
zYCbv~SNU?o8&1h@dd%P)x1MR-Rt*X66iCOgTXlJyZM<p4;eN@Q4etLQ(eOj7r@c_<
zYp-+(C%1f4j*2RW)VtA2ViQ;*3My}xey1rNe?_PEmT`Q0xY8X7)X}`UF-fm%aGrGI
zaPM6X+ShH;hmHh%U!0>&GRF!ltwvjc8f@^UsF?o@(G)!tKp?%(n;~p-Fi1+wCklX2
zx^d0Z)X@M~EbLQkHzRs_(79BIa4rFOOd(&-4nsj2VV_;QSr*XhQ1SBf4)rl&5r3dJ
zQ4|c#3Yy)n5pp?6c^OcqW&GWQ<0Oh!jN-V_n(9NILj&E-q3^Jq`Ue>X<U4xC>jTpR
zy{6C*7{fQ@0NESkza+P~oDS=nIQa9TDb)Za3=;eTnaY$~O$AQ;rf(CyQ_?{bq#4WZ
zo-g*ol|3NRp-dW~5cY!!=3UP*+$WYe%t{jS7FqtB;a7*vu>$s{sOjV)Th;k`h2kvH
zOH6&x7Z?Vw8;15wf|MeLFpT%DJgTUd7zuV%0ECXjFg8Sv6Hpb#VhCcK_TxF;*$K*j
zX9DM0v`iyAIUQOy-19<-<<wrecjCZH$k<&I=HX2;$5rK~sLWg=ksspp%`LuBq8Lq{
z^IW+cJG`~i^tI|*wr2l&$L<VHuVZse>XIeLR?)J{l0bU5W~c+R9!iI?r#>kRIPXPo
zYaxGIMz|mS%}&4B-igQD!qI80Z5lHsAW6LPzE-xx+3ffZ>#ZMYA6$}c3~rY^lRdn(
zAAX24+#!aXteKhCOhyfCchG6yCRw{nn4wIQo2&-39}K_mu!#|G8%ZCMAYL7h_`P)+
z<?uf4zvnTB2Fa9fuw3`iLw9uw*F}|FSY(k&h~OVc=574Q5k9CP@44j;h)c0;WPJwj
z-!`iA+Dc<6``4zSnrKAC{G#s_tHdXDsjh3bb%_YpBFT8|NSLJf;#0QPjA+wr@UOzF
zH0m<`_b3!mnln@sH<q#q2@%_@(LP!<!5nDIpZ9wu08)~aI)@cB5Dv*;9A4IFkQ(Pj
zcFo7FIH;YbTRv_tY^u6aI=x;R-+HmkM}IIiyS$b=gNZ4FSFNXftWLD=dg-UW1W7RK
zK?#Ftc<||b+Ey+=aWgQs;NpNP4GVJ<1j9(fbl9{C-Aq$~#;#)Ie7L!4jul$DtU89_
zH=$eip*HB}+Rm%`92T$DOGB;VNWkxb(yf0O#5`FRusF#9v=?r6SY4;!*uR+qdx~1M
z7YFTL)6TLCTRgk_XpmN|D^(!~P~<48{qXo*d|i&m`9kvMw?#9YXzoQlX^E=mVI?bq
z3(bptq1dy^0_0D$gJdbK_vXQYG&jjzjVXThq^7xb72>c2$FV|j`ofu%zp{MXrz<r1
zT4+)t>i)g?RgxVucrSTfb96y}Muh)6_4Wgyuy;!v(vss9n7Sg+s{dlK<irsfwSmEF
zW$d@%*;gsGm?Of27*cJ(wDlt`t>AhenqBRYu`*EIynZF_+kJPfsz=;DR8747UVreA
zc01DRUcMTe+WvHpXUr;459`NSqv1|NEBA&#M>pQi#QnsYyvCF;ylaA*U}LeRy#F@a
zW3|OsRl$|dVc&<xsB^oy;lRcu_bSi6a!8Vw|0ApLF6$y7^tE|&wjJ~O&G?R>liNBP
z?vk<%DcwRg=IO(R<^TS+w%(k)R-r>$%+#f~4vm6vtE$lth%J!NE4jvNC;)S;uxj%t
zPKM9TJ8YWPm=^7krhf}6(rQ~h{EdtmDDM()TNxRRKnDdAzXb1rVhLf1%}e+I{8IFj
z0{T5v=Ze$yk8OaHT8W8Bzrr>)>+dY&2D-tx=x(`k?0R4jh6)fuAI0{|>YaesNKe$E
zCk`GESStvyak<cCVCzV#lWhj4CE(cf$fcvakT1ZFE-+_K-50^Zwu*)KiP}&D1$DO)
z%5b-;2=c)MvRbD6a%{x3DCrosXa#_T0-9ChLRPD91fO%=<(gu$IneFOxH*E>8$PiJ
z6hc^^;xJe(&Ebl?QWjs}*xxx<X6ALr$cC$gzF$!<eDS~$*;;iFrFZ>@I{J;|_?YcB
zrvA?)qhX2Cb;@vH4*jPr->=DkCDfWrf2*6U+kzOM;gqb(t95Yv3b_9)ZM)O9@3mRn
zVGO>LOpWt#oqI`<L-oq=MWKN?GCD3f5SAS~x~|5Lma(2>g97=vveF0fc|2$OWawI!
zmO4{yy4-3zyx&3``0}W;?5>=m^zT*0vibEE42icm;B~W`{yF)P(w@*GF>Lz49~;~y
zk^OW*8ELH)$I&nOE|xH)wzs%?-aEvrhJ#A~g0lCMiD`^Pe#?bpVt{Ex8!|qZODiZQ
ztD8q_kMrnuFksz@%7W?ZnxTKj!`r6AKQ1D=h?!#Y+Fj*kZa0JkQ7$hY-yof}s<X%z
z&RVwa?S=lKLEEY+fa~Qc%1^W!*9xC4=JC_AVWA^srWikQCo8GyQp3jlh43nF#nA>)
z))pmSpkUVHqE4cRI$vJ1nz2k#ox20s<=`5;ty;IEk1nRG`(T7&0ATKXEjq=nh=PX9
z4NxNu!s-=umF0b(Pn##kD0*5;y#&4OX*Q9uMBEW`V#P3qh_fN**O)hs*yt_*Gz^rb
zbnL34-Z&J{kUaox4D%cV5c|~tAI4~l>;Zj`Pp&o5l#C!;1|VKwrW}mcsy`E${04t{
z$`TLc)Yv!SE{bI3hZ?j!S8N3;7vNcpNN>gfQSp{E+5xXh(Dy+q93Cn{C=hlKPXgg)
z=qW)*&gX0nJOaAU0t8TN82V;X%s#!hE-#;n7V~wK0WG`+#OC@ylj_~`+tb;YBTdv>
zKEMS3H&Mc#Nt>j$Nxu8I<e2Z<u|%#a-evCE9~1;(f0*gv;GtkDUUZm9=Uk4;kQR~j
z?y5KN{GbgKo%5mJROSx#VU=jK1e*RCSHC7X@bDFY(Wzf9vox+OG3^mv{9HHlVSbpd
zXfF}f0xV20Kh6HHvF-1^$cy~$B(4|gGp>6fe-3VeRZ7b$<vb^!@&;}ML>t6}<0MIi
zPmu(ckBP(hqtnZ%YBJAP_fGM%#xwV>i$fn7onMKW_3vihEb(6pRXC<(>vzFFF3N=h
zhH^c|Gi#6x7Ub}0q$Mz?qQ>V@<#D#ywlV$M5o(VVF_eO4uQ_5mh5|}ZU%B4Ka~@9C
z<2_DH)KA#X3I2W(cIYy+8AzM26h<DZ-H%^71z-zMg5-UJ{{;3t4Wu5;8y-%Uk1=>A
zuvqI|%`^&Ky8SEZq`Q{e&LCficiz{*`Fr^dx6!ZqPnzhTINp1*rq1zO(=HiH;6>+9
zpmIz&ty5>(_wtQ*%^{Nkd)n0#=6{Y$d5+fqf5s5c|9bn^?d37<s^z60!y6FKCOke<
z!t1_q4T~L1hZH?lrGb(wKEOCdrQwj7hv$pWHIXndue8fyq(pVB^t*kscO9i-_MiU$
zmmK`>FmB{H>FjlKaAm@AbY3kR(gE-TWASiT;6V6}#CxV6L8KP8I5x#K!v-5#o9E#n
z9#<<`=t5D4Ab+++pfDPqY#DYAzyJxUbdr08zCmR|mJ1i$u><FIdonYS4AmTF4C(-8
zb>7?eG4-J`LkvJ(1kJ*%cRWvUVo4DlwnhkMA@~=Vppaqv)oCreTP8AIcZq0V0+0cT
zGk}VFQ-UtRhh-84cmNd4y-{fd0j3>lo!$tDO$3cC<ftNe>q|lcpUtOp<=vZZ6u~(>
zL==k3kN2U^4iI(l9##Mo?Qq(1r--%+8G3XXmRO-I4Sbd%vBhr*Yx29y>bt^t>}fkJ
z00Ru7mXKhpH78grMFvjFX;N@i(TV?C=sImmos@*g;x34<HuhB4ajr6{dAm@&KaJUN
zpPY%7HVmO)72ps7%KiB`?eMZXe({`B3<P=cfw^jPX7v7<KbCphuc3o_`#KF72X1?@
zdg<%gV4QX1_d(OhQVolDrh5tN_U6Pt`Sgp!{%Of}qiDyj;k?)C%Wj-jTDp8Rp3`v0
zPTNLrs+CPz&V}+HZ`5i~%_I2SSO8bmRx^$>@+g}B(G$&3(X2a`rwJ+~@^flfyuS9j
zGr0)Y+9vU>SxehCTA1fiul)M1gGK6{GA#>yboP4O5!ueivZmd^rVPsZz{gP{sy`(w
zG9uNQX=JQc0~b1}s<+0SVsOL(pjfLQAW%*Pbu@nn9#VrX*izweet&m{i9-P3f2~Ui
z!*Yv4{lrg#!KaZqMME~=!$(eS4J=WjhhPApJq5My&t$h~!DB!|3nzd8l=yp1DlM40
zX8osYz12VO3z0JQ`d=gE*qdL4faF-9Hx@<+kzmna4$cf@yk&9HZ4|y0C;k!ZKJ^hZ
zP6)r%`_GQ4UD5<gv`jYO%>u5FoGU6xQ40*1{K+QK2PokCvTSp6%;Jpq(;Imp)rh(Z
z(GJrz_u!k7*j1z7)HoULb(rjr?J{TQ1_7ETgjKN|0*7w&Q*I~PrzREoSRi?TQuD6;
zPgtr`fOH-QOw_@6Kx@!T(|xUUa+U8HK_tU-{f((vt%~VM`OXI{@62HRQ5T^oQ2Vrt
zTUSgoP?&6K`MugdZLTRZE9xcAV5h>MZeTHD8@H8=zMtzerT#NZ4qa&rE^mESMUGG)
z9j>pDV+LIj%hi0^`tF#r0odoOf%Ld2B&A~d&vjMF|Hz5o#=~7;81DvVOvH7vZP4|x
z$fw;?XSeY2%+~*9qcwi(@8Xr`1}Dt*Lw<kATcvbF<%#4weQCDu?4sTCWlDN=>1M)}
z<CBAM4##~i82l<RA{71Z?r-mh)_(`THP0_Q*S>Ljjbw*dkPxnr)SxCn(A?TvaP*DN
z@NS%>IFO!*Dl4O1)n4V|ECj^xn@M)F{Vd|E;R{!u#Gg#LT{=jm)5OWA&@(>Sec{S=
zAldL}q}*h6@IPBkTPY4ilqcPK?G)*mNIkAF>P$%e+b|a@g$r>Y4h^;V{T4<5K;xFe
zfQEPRj-TV_!Lm7U_L%?#83gbd5G2uXgQn(q&L%lM*nCLNobxmPg`~*0Hh}zVX^^m{
zr34{Xj~(nMa@owmS)$kniBlT@d0%grYwJm&Qv+m4nv#gql+{%Th+uDJ>UB`dp!lBi
z^Ik_#4@|U`5TPhZkiqsc5dnaAh6FyVEncgKL55ESCSbPd(g|6Qf+0T#9w0c8ZCe^|
z-jML+o=SL17lJ@x6%wp~2T6S7C9!-c4kFv=1wvu$Z->wi`h0G&liF)^BU7JffQ(i#
z8WA9L(5;AtEGL~+0#4guX;BzQb%fq{VFR{iWDAEbP}9<ph&)KvhvuB&7Gs}hXCUeU
zE^pG1#WEFQ&2JrRBT2i_w?B_AU!Or-gB^2kfc97KBgd!nA6?k*eX9Mua(IHOU%kls
zXk@+{llAJGX?LT0wKiW+iO=27hPrnaHkxfQzn{k<X0F5YxYg9Htz-M@KuOivzy0*+
z2UFwo;-A&tduPGvyM95xONpyJg1@dMCiZx1^M5y)BTgrO^4dAAFYd=d?M(i@p`Jn?
zXXweX9>B0z={bjK*6MBBCw6cbBL1+R%fxJIzE=TYBprAE%CD+*_VsdXo86Tt6RBMw
z)q18IZZUOQSfbpWt9H=A1Q%B9W;n%wqCLD5C@&<eREbtk*?uYU?_&R%QvgcSeZ%yw
z!{gPv_l2}FYuj2T`G4;ioCl+Qg}s}s_)kp__P<nYydo1uKcAdzi+AEYp$d`0Yszq;
zKwt!9P(U(uf!S`eAD|U7mJeyd%nO?T5kNC6dq-uXD{SundyRiY@ELj*19JQ4ck?%n
z)vJLl_{K(-a_2W+V{wbSV$Yd2ZsU*GA%uOT^it^Zf3pw(MC7lb%}FU@!dPYtx+?q9
zYFeC!z6n?-@I#2j1`N%Vk2~@n*kEI9t7JPbNl-Q=iHn=3i|TPl{7F%jyY4~5zyxcw
z$E7SX%#Dqpc)e02bVMUFvx7#`3H9cwZB&u~spv}#9dBh|hyec}xX2I+ngt0rR0(~<
z96r>ua@>{8fVVb{uED#y!W9ZYTZ^Al5!I^^iRfO~jzouOaG=TS7cl(7B@pivF^wDB
zXOFWb8UQbSG8(|PpxI(Eqn&HeD=~*#(+m9RGy1i*FVK-bcRPBt9}@h9!KIY)!H9dL
z*1hC|l7eX<Fo9QL0?dNjo?GDNy^*W15DmJ}!y<P_Hy6I&Lv?$KlJew!ujO<((U3cD
zv5UfZ)!6N9OR*SXbWedQI`hCcj*hHYKXP+U^WKO8u!c-03B@M<;lO2UX@B%Tzlr#W
zB=Z8!t^5lGMYHz#_*KJv{F*asXdM#YICZ;nWH5L+%?#uei*}@%?h2_o?~Kg4C$bu*
zd5+f~UQPd2s$G1CHn?nt@pqLdxl;<OtLq#5)4*Z$%UkY>nryGj(qM^r)0jDAoM_NA
zLc(eMQBqia<tmLCEN_nl_{{{@D=!^-7C-8#3wLJfqu@YMTFqai7-R+n+$?l>s%uF(
zhr|RS;&8m@ovPT?>6PtCH8|o>`J#@qB?{5t;q!azw!EHk{CxCjuQN$c%w8`#Rc)@$
z+X!>_ui?Ai3k|X$nTg|6up0&(Pz)^-#Uu#U%`SX;VmKIPgVoVBqc;05k@Qi%z+Qzt
z0mT`?Zzn>vtCqmBKt~s(GKDVvS_Kc(hGL|_1AA=_0QdAwzEk(HC?~K-L8b@*xREi1
z7WIe$q?}2jtFCbTWx!1G09%?3i}#nn4am81_bv87RI+||)Hu+uf8-WCmIB5ANI_X~
zh*u4Pe;S$p#UV&C7|XeA9LqLf%-oQ6>Ca9;+97O&rE`x<KB|js0A2xsV(=LJ&aJwR
zNnO*nEJBvf^;cL&3Kz;;nJO$RI0pFj0M8>D2L>H6PN@D|$o3_J2_^>MY5ax9kuOa+
zV)D)wzh|jMK#kwd8Ji7=<N64v0Vqzx)oMLS6rgubG^L!E#COHt4fwCxX1;a?f2J+R
z_rXYs%eu@0$WM4kmV+<{n6KMn|Ilsk3~fnlYq%;%uj(D-amR=)w#KnxD4N^(+=cZl
zwx&1h@du6e3PMVSJsKHX!~?f}GI;H9Z~tQ9l11r=A`#?lcFqqY%BJ6iZg4tmpuUhJ
zR_8OPIp<#NK(jt-tZeLy-j}F5@RnX~NVebJ3g_eB|I)<L--;epNZK|-#NoK}wt^#V
zNn<0e52bC~!MZ)N`(c!_eP-SDB8ltvz^R+=xO3dEAs{(8(K~L3-1*F>+a;#6bD{mV
zTI1#=g_PHhx_`2B99q3m7~%X-y&1!_*L|)4-~Kjwk;Y$pT@%%8>xvn`VlC~(UABk%
zFD%Onp5D5UJqp!X^OIzcSC!=-XB(;PkF!etGF*P0WNfW<I1a00wurEZ7`!2MPP6a^
zBq35@PcAG|Ad~=ASBXeDqJg3+*%;mbFAHGGz3lZs9-tNh#Xka|h=&KVSLvYompQ(<
zp{6|ul?RUcxpCMI{EJ>SB@9AgK|5(>vKHI7{3vq!8Gvi`8D1>rIvG~!IUJuuNoFE0
znaYC^d9aP841L?$!PR0!lTL?*K9-u`N}nKi$(sQ10Ax(PV#Th<1T5>(Sqvk5F)3aG
zpohseMbt2?;hQ%rAmEHA^%JDvm%$4V=UebsI#j7axkZ9I;}L8N;sIGq&)*|OD9g|n
z#5cqw0m{`4+{K|^O)`OHDIQ-QKq1tB7z}~(Ij5*L_>UlDIf9EM)`P~5FEB#{$Oj#c
zf9!Wj*`WB=dux%MB3LuSjE&V6ce8x$&Z$R2%IvN1Ffb0OUO1nmO|SQAAYB{~qs9&<
z?Ee}SRo94jImvEqHgMBX8M6IiuGZy0I&p~rsj8{A*V>X++)f9nTQ5si-Hx+=nO*5=
zDDBB@zsvi;eM%{j{-OxQ2Rk}Tvqh`IP$!?<buFW&Iivj`H-?G_g(+i}Ci;c1J~n$_
zsp`oYgW8W;Gp3B3+heBjUkwNQ9ox0|u{*BoxnXfao+hm`A$+f66Rm2t_tfAs><2Wj
zC4yC#-$BcR<>E{OjblEM-`{+^MNV0mF#uciZocZqg<a;sm`?Uy6=V5mhgFm@ugsda
zs+WAonT$JVJK3M@!r~=njyLK=qJK6kg<N%Ux1nDQh@GtPn?UazJUz0ow#(#7&9SN)
z9(B)j2T$E_OfL#<^ppJG{2xn<df&VH-JDO}8H<;HJXIu(Hm%Mm^@BU*LFnsPY)~2s
z5&RE?A9Q`)?$bh;!zL7A2E?m{{emTS%LW3``@o6gYkV(9#PH7f=3ZUq{))pQ`@EuO
zLb2t1Je06OHU$s?z5(rpRG3#LxXGyyP*A#@2@q@oKLqO#Mk3(~ah_njxSN6p(S2#$
zTy)2G><~ZuzJY@AtR*HCik&D!E15A%{{);(dKPq~{ARL86vK}+@i_g+prbDUR>;HO
zOnJE~_wfP`62<^f@F!Nd7$?D!faL1?FPzx`=7V0ocOI1nm_V$->x6sa^Ka$PiErPB
zdNAqRkV&lXhGJPOz0UD&$AbetiE00%CSaCylSCV2tuGJ1!BOb4#ijm9<_EU@_&anO
z)t>wlQkm{G2Pi34uA_s?b(=$@tDhu#7Bk<J-N~2-Eb+?c_!9qHwbM5Ok9KmsI0?us
z4n=nPYMF_vS-6HwzfZ}0*Y$B%uQgX*cnlYd@+j%A?QFb2=S3c~#o7A$Rdpj-VW?a%
z&~Pa07B@leX<Dq_%0uDE7H_!Sd)50+z@76yK%zU^O_Jg%E_dF^O6OGKS<TN2c~Oix
zjQB^xTa9aiaXKVvMq;AIksjvckX6T_nr53ny5%PsBUW-Ed}o0)CTjPOhtb!6+YszL
zYm@93s9YqnCRAOC%MbdwUA*D7EafFeboQ{4VpqqdW{rYOaWm01PnYyM=l<CTF`S=|
zYow}nGk-#anGXznMkp3zXXk6bheZ)p)EcVN<Z7rJ&X3gd_<@eS=UDcHe%WYU>Hv$q
z%sSkU9O}FB>Q^Uc@R{9y3SQeHHv>TDgw1t-2C@BdWW{KWcjhBLog}7N!0t1D_E-nk
zwND9zk(6R*(uSrbgZ`6?72TbIDIfV;E7OEXdaFUM0SsfSPy)`TvL|97C4d+gm7n$B
za6{V5%f)_vS6)AnAnzN~)Atb&#yQnWqL*{zsFXR8rZXMdive@kS?yLzZ214)j;LF8
ze|-uz3D#U7s=r1xvS5;tw{2G`1^GovhOAT$NI6n;D21JX7LJc~LYkM5k>*d%ugcCM
ze;1~X#cngXU}dlZWRJboF~que0g(N!(cMIBULdJt>pH_bjsh%C0_v`2V{Zh}<{+x8
zo(@mXL5y55yPGDOOXwz?!abI>NlsMDutNKV^%(gQ7fi$><be&nK{i*fBOy(iDp(k#
z2zZTU6|$-2=(j_b$fZJzF67|I6)6`TBR)djxBBxtroWc)6^PNlu?dWbeW}-qr}k%X
z9h2BjeLRi4R`P>|%```}cB)wau*8=m?Z^Nw4!Q$MJ8F0)3H5_fa0;EBzvXDgz{sMK
z=ML5TZ3yFR?eV=`rnjcfTj6v!cD0Y3d%prAFu(h`Q9I64sioQVHQ6V@8bH)tKt#2x
z7FG84{WRceo?AF(J>P$DcQ+t0jPI4|=wzN0=IJ{2KK!{;nZm=ZL{z>gho&9*5FIA&
zV-x>FU#rvqm-@9u`wY5mkaiBK*?wE$u8CpxZY~5Rn8|x57+P_?upPP?e2hP;#`+-C
zobjUZ<W>AwSa#NEcfgcZM<ws`c0*~YjP{1m_PT@9IN#yQ*YJ%glDk~kQB(P9rE0y%
z-|+HXJ@1%%x~nd#UP$Y_{>!2Xg=wvz%}>|YOZU`4=WdEz>PGY1_#+fiP0jOv9*h^&
zidAcqW0yt`KCCL7RMLGLP4J!_xS~9-DXe;L8K$mFr{hYToW{~BW!+aZT{HTsD>`6c
zME~@JMM!dd@Jf8Jv`4wU(3QpL7XvQa!Ih%W4H^{4O}~T%IT-pj|J<3-<o-to2Mk0P
zw~D{+FF|&E^c`THGkf-NejiRH1x`iLg>a-K-2&+pQ&2p3p<&jwuit~{EbI|_6xRZ}
zKodg1OQ;B(C>B4YI5}tK#e92^J}hJeK}%1FUCfMzewqA^SbCuKpO(~CB*0^ibf}X9
z5xm3%e#^>@iK(x{D~iz(xbPd*@6L<b7zK(ZJrvS_VkG?oQa%YU0>PoQm=@Tvgy=>G
zaAhd;EnY&RUv;f{Qm^|PBy%~UGq123G5=-lYn#P!K@f~AVo76~O%y5zRXPaAW{KHB
zlhp(z1e=1gZH%446MjaxxbNS%1uKC|A_;#)*E`|BP9ds_XTD@nd?DD}R*8<k89x}0
zr5SmOmzDqbevxe=R*e~c9U5ih8)az7+!k6X9g73~MRDH7jXSxxqpLoSb6b;}XHKzb
z6lA65ubL6NJkJs!w=(a`suQ`~n8Wc{#l}H{3=;nU%}aw$;a}}N4{pRQ&ws~ZiS-ah
zhf0Ni_lDtWWVrslBFa|URhR89InLW#XkR!QL|?U|=?SJ$@P?>j-;E;I`lF#Zlc8pG
zW8snUXJd6dq)swx*wj$VYS(>eIqcbpy61D8b?3YD{crg|n34yg!v!T)<?DCjBv!20
z)g<SSZK6eJdiO+8fparqxV-$^VJ4b&NVf?x`HP`*3M>CD2v@ASn$+_{Rdy35Hgacp
z&wU_<z~Ihx37aPu3C=DH52QjC0MagMk=8)~@_6};Bs!fWZIMYrx2WtmHh-Pzi1THg
zo8&7AVV`-IM2{mhoRGsj3<&lgydjZRcoUnHhls5o0lEA^U^q8K_Wpu}B<=y_6wzWo
zTMT8!{sGDK6%=&cI&2IpsH8C9VBLHHpG*J4(J6~21Yw{|NxOJhrFr!Ew*tzhL__-Y
zQFoxd%mfS_UnQcier**f3d0$bt5nBu`5fmyuEr61bSYDlv>q``VAhKu6h;7J(1o|=
z&hRBK1?dFhq}_cTI~X=PP)uGrs+Hmm3;DGvfB+Eh{w-{jGgW-?N|pf)M};u%Vxw7u
zHl*Iz)RGp+VU&st<FF=QCZim;Xvsvb<x{B(zmwLY+u~||=Bvgg3btEd6D<ml7!fyg
zJn}v)_Jtd;>KiR=R5l~ERlC3AZ^Kn3CqG!^;=C6!*JY}y<EYI)`@I#%NeNN%^HO1T
zjmaik+#Y_9ytq!m2eBUEbsqIb&g#`~_&68%G(_*9)s0JdZ>?VC|A07PKz>Kbxp6}+
z&u&fnRMlA*5Cdppb>+@B@?^?4!m)`|K2Fx+g2Og1CyBF`)T+)m?X*a>O0cs$R9*tr
ztP51!MKd$#96}o0y;+wMaRRE|Ne$@6uda0b{qZa)jBj^*bM&8(x0DV{)ioRXobRW{
z^|%@RZ>)Cf>&>a=`DPDx22*<4mJQva$Q{dvIoBq>@$w1IGY50$ppA6N(`~Fncdl?g
zkhs;<Z*V*R$iTP%l53peOc$j+n2Ss~u91KAzNW!#wiY~?#n2u!;9qDHIKu)l=#=vF
zz|vI5_L}vkPYaQ%s}BWXlIy`281ONh6^&ssa!?R=emMwj%!zCa7afmH9{q$$^v81D
z#$2fKkG8MD`6K<L-#LZQFgN>&DoKcr3HIyXZpQG*Vmyx<O5-#)+Plg{2Fz$reHug%
zAOIL{CAV*rVh$aOI5)R-tS58rRGkfw^^z1@<&EI31PgnDVpG3UVFdmYBf?kd0X2hO
z3UgELO}Fbjs&&|(7XSrh^HCxww7*|VGr&^XJD3taKsn!DT6b;$z~&c&$-QMS`v{Il
zH{jDZc->A?Ok<(Q9k}z1Vuzs%@Xh%^$0gz-D~OUq;F6ocOf=W;T!@@Z+_Boyu8@No
z|FwczFh8!whr8POSHo#_T@0Olt-qDo&t8c)G-m32Ty?dMz22E{%;J-*u9l#^+K}wx
zj<-Q<B&+}NN%}{d-$ZL(VtAO*;%}6Ve2Z+_-@{b;9QHmsxK*ySy82Ri2d+J|Ckp%F
zo{z6ix4U=Kau1w$)75`;x}==#YWRD%up?+TnR)BD^D}PqrF{J-9+Sn9J$*BQ|GF+3
zi=;;+$IX?ox+We^%f8iQ*<r-)$}N1MzMqy(da@+KRL@zd=VjS}|IMzm&k5g}<Z+`_
zje%#M<mVt=Ru_lerV1zY{kp12RN0(@N$JDZX#s6tJgeM5)WUqi-YSJYFQ1}Sk{91w
z-Hhe^RS##;6Q}G}edLYeaea(KMoAR=MUz`OK!xQPt<`jlT|MQlY%at<Fgla*bA83T
z#^lu^+>{1EdIGRqx!L)fQQ8P2LN|Be-xzu>Ve?1`Yo<E(Z^|z^OpD^R)C*GtQnM;R
z{3wW%Y|R&Qzy=c}fFdoqtE6QC^e%(*<VT-{EbNxvD4OB|fkU%Ytvw94XwIDgkx03a
zF&L+~)w{+yseOw?i)-{H=qF!QH*ZfIeKYb}1rL8?x4Ow!)|QZiT2oFmTk_hc*UsH#
zuWYO`IJh2Gbc5cJ*OKYsp6KqjVGCQ*Gok`sZ-zsNQ**PH=&kl5!$6SNkCp^p!}Oc!
zU<AluMxRb`KWPz0WEFgTEF3ax4cZu$OkWKmbl*`C1I|GXy_EoS1*q<*(sDSQsXw(|
z{TO+DerUqoMC~pM<Ew5MLRNhE#<;o3Y54uhlLni9Ak9RF{HiXf%9B135N0Zb4F?q^
z?`HprrR3$0?+U=>_LZyKSde&hIF0?fxaJ?e;_sE^<zs$#?SJ3ZR{ElQq+{CKT!+WC
z9w;*ZQPdhNulU$O(Ba0qeqyUr<LbLgZCs7U(%O<vpZJAuO#f-~$V-B~w|>{``c7fF
zAyFVA{?LZBJD~MSO5NaAvZ3rK@mhtmT=H<2_A}p%o@1hWYilUqYyS~1b?jQOm0q*_
z)w&#3-FLqe0p7`_CM<8;S@u6wtO^sxNPu-k@knGNw|R1QXTMYQof9n4!9&P>T^Ei0
zT7T=y5#afWBOCE%drn~S-l2NX1lQI?#5r|Zf+x|Z|B4lUwb1mU+XQ#%^G0`g;qSk%
ztD|fj<1Oj=!uh2Hk1DKVGgf0}kloY7(^9TZ%f`C^p)h7p-5n|s<naeA0&fgG7`J}c
zT)F;U+y$i33tHYH!nOd(hd-7+Q2+T_?}e|S6xB?9Wwm3itf**?L!w-Wl%MFzM`0sh
zaqf!q4w*!~P1wyA+5Ix8YjsCJyny!rt#|nX?0dH~es4}IT(AdxAe^jNr+5}`#)v}(
z+^IpRDx}M-6fq@58Em#dQ8M8$t~zy#Qs~R?s?;P+<~mg1nsiA$I=6s%^SMWTz!zAK
zWj>tN>JS5i9tda^;Q+~T7o(%Q4OLONYOCR^=RWdCL$=ZYLchvK>y((-q%m(&Nd$7?
z>=QSTgwJuJi)M2?Q{-|6$BFKIjmRiOJ~x@aga(W=MXjKfpNT7tqQQiVd{3Eot<h4#
zG0y~aZ@t4t+b4YMmYhf+hKoL6BL+laUU3x3UN<8PyG!p&xiil;@t8O^?S*KgybENk
zmaAK3_Tk`fuHjU@<&i1RK<>p~597r8T&Gla>rzN-oBh`!R3y5UWc4xb1M6=Zk)w9U
zY59MR_V%e8VcXaK4ALSRm6#wc*5CdnG^1kOhSFl~C9O#4(5SS^LYYs#`6gSAxzue?
zC8zU?%qOq4^_?W%*tb^Gll-hkpUsvZ@Q^8|%Pd@aD9s5lAE*7DcfAn9%C^OdTU$Q)
z1!WE@hh%};5lolmJ_)7u@e6+k=>OCRWH#ns{A_sJ<^0M`5?VkWPK?Z#%^kRx7W5;s
z_#){$L%Jqp6DbHxcviFBf-OLm2}Pw3kVH#bdC+&AnrNkZ8jrpL7aSal+ibT-uyLpe
z6f8hzNx~gIRBe^yF<0_(s{+^ho5$|V2+n<b^o{HF@paPp)Y9T1FA&20eZ6FLCl~RK
z@Eq~V%9CKUX!fvu^OPIV0fUtrkm?$6*J`9h3W6|xg;b#El&MeT&SOA;oxi_x+jz%f
z50WIB$^wWG;N!$_j|Ht20D&>Pf&>u+iS@qsc2YDa?dL3fb&YF>!Dr;6iAo&i>u<61
znv=qS)$<0EY(x|qu|&s2p0J)k?=1`bO}f!9P4uq~cUP@o$q|cy@ce7Qs;y|P{K}Qg
z<P#bs_yzym$0cfrnv!wR5@W4Gb44mt*}{g&%Z@6@WT<k6R9^r&FByJu;5fB%yh27I
z>TxaNA^CQk&w>;fN<MrblcPDJz}Q)1B4Y>I?Y~5AO%7I3lonmf^BnbGb$`!OQ`F{p
z`t_K0mXfFb-zDd(O=gLii34uMjfMKaWJfpaq9y8sK%2Rp*{KYSz)Ifa;3tWv?A0am
z_J!58-i~3`N;OGjQWBfp{D1QN{p`Yc@;js_vEzf7cl_Q(jj;E2Hr9d~%C;E;4u5fl
z;Mz}5J}0X@Q;zH`mT3S-8DGAdU+&X~NW!%CFcAjVSVBE3GFyUM)8Rj88r{`OHtRb|
zi+~4-!{~X97ctYkAn8bZLa7;<OvH9{g$xpmu=gZUFVadr;-t7o7c(O$;7y@o1w|8=
zG74;dM8&1JvlPkjV~PF4Wby$>qVWm<sEmEt8jWS`e!$B=B%=N!msI8_7pdiI3`ba4
z@Q3mQ!or4!FXEgbImv7&(u8}&7KPS1C_6w2%U`UV#7JJyKt$MBOf3irjG>`}T>Yg{
zGJ#|m2WCXPvK9}#VXzqYsBb7RWjKv?Z+a}YO?cRA4KHaz03-Q<DL<$l2dE5g130Eo
zrGD`2Jubj^KDz!fOU`oi3Z{ajSJ5q6MVJn=S1}!92;ct1Zmq!z&UFzaf{oq(RE&U6
zX;F?i>HX2ca(sa>ndY%V0@ecNOW5T^Jwddlg%r<!CNFUMLl53=c2G&!{kBfw<5#9J
zO#Rh~XW`|!Xsdg5b@QKLEzKV3?6<tGu?NIvbTkJ%W9{7yiF6{(+V>~Z<LR6qt;IK_
zk-;#uni~`N*p}1PL`XM|>qb~bhqCFpC&gHH|F9LXQftB6_++!oq>_7IwS6SdZo^BT
z<6nhaI}R_yMt4t`$$P=;t|Nz@N!NXePrser56|Y0zpc)R#G_?Iba5?e;GdpNYL7S7
zQ&AlVdakz7@pAbcLHwVlXNxvd{rq~!rT*PFwY{VB-Tor$V=u_bq9^hz>VLWXShBVb
zFWE9!wpwBeSZl7z6I}1G=|&as>CkIB&m=Ffa<||8ccgCec|=lfAI&C+t!?Rg-!pyt
zV*Y$MQkESB|FD$QUim&&{MCqAB$_D+?y}yVq_3BM>m_S-bz}zdKa)OpDc5b$;Yca<
z39L9<pmJfQ0i!7&^F+T8bS=nI_!B#<h|Qw6e~KU(u&ByRaRDTHkHx#5I8B!ZzIc;I
z!)Aqf!KZ_65z=3Y5HSb%YyQl-oR&95pu7G*0M9@$ze)zg(vT6qE^x(xDFO;`Bjpm{
zfD$n_D9aXL20*1RS5mM;P7m6_&fSIsI{-OwdqoTm+6M>$e88YV%n|~~lW>palJdh?
z0i>`vCdXpJP#<-WkOfy(9rYK626`YKNL8^_h+#rZbC@44FO-;3yDoMJI&ggt0VL=8
zO0h;@#2_1t6~K+EfG3zafG(G@MpO=9#eq$(cKsRY!vQAx319_C0(hkgf+7L#pUI7c
zXU-1UH(U$ikiZ@hXry`ot%yNl@kVH3Tj#dreu87$rAj~m`MNH_EWn6wMCIeAtvYw)
zh9s-x*Pm6LtFC7n#>lE=&9ePj$Mu`^eK1a+v*>8U7)f_QhGCfx<Snf-j=EiW+UW-O
zY9V89+_kwa=cC~CJpEoL_aT4)xY8<XReEKsaLql_D&Irq8YJ#Iq@9b)A9pNm%rRi3
zt#|uw&HxC&&I=jLVT|NGefi-#K|@Ag+_CPifxLOrrw@#rOLtp(PKhycs<E38ejCO}
zp03vPvfyS8x5ce@9tZDx|9f2p_TKRAJB-0)eb1Q9U|dT{_J?uhHs{w{1HB8y`BdB?
z?w5kbiibhAHp@4ft>AN?`IJ-AFMRgXm&)2Yy69be;$z?74t(_E;r8<E6gwv}kTFrc
zO&>l}sAsOMpzHNY_104B=ThDcEsWtRi3wm6;8nO{NSm17&`WXSx)pZ~lkO>eH=lL8
z!Om!vE6RB5?We(mm+v{f#MO1<6t=_;q0i$p1uU1_tgejo0|oC6K?s2f0zY=eka#Mv
z!wm5S0TVDN9AiY_Mwl+d7^%BmKoX4gLxC7DI$S_=eel7OAZhNZv?Tx}bggxwx-rlb
zhhr#^gaG9_Fp8Z4*bo9EFgvydAOOZxWvmC}Z~@R^mEd+t?bYRWtF86{SYU1dLJ$PK
zBX$Pu!%U%n82bS<2KsWs#TLvF%8;CkFuCr60RuzCfd|N>)KJiKg2ht@sUR@OF<sP8
z$57ER+61WJ9|6&wt1AI3oDX%O?TG#17%wn01d`w~tSCLSakHhe=m+qEJ{w47*8ulr
z=K>G61T0cNx)~^-JNW^f+6OIn4FPn9-L~o{xV(U=YR$RUa<UUs<dPvI5KDjwz{$Df
zfi-d=gJD6>Ag;7388b$VoignT$t(;#^gYihPmGae>S@+-{ic2&_zMl}I?SS@tQaFE
z!#Zxq=`D^iLVOX%<64f~Cat6H(_}USS6ZIDoHJmA`45bd+mh0rouoN@9(3aaBdY+e
znB$y?nVUDjC=b^rOBt|b?KYZ@F=D_7uCxYP@N?K1Zh(<2mA!(x78QRd&tumL)O_oj
z#NN5qF-C6chtDeS#JmL_TBR;FZI9Q$DqkXL6Z7$sZ*r_!128}W=*`XRG}xqjiu^+X
zBV`5Dv(+Cy+X{;OPNc3?_i>%u=lN0?x}YWi#6XIISQRB1+q*Kh@5x|)qIexcu_PQj
zL?A;&fC;ROo`95dxuiNDh-squBDjoV)f@3da2r)qK9afL0u+IfKzt7X6yk;S`-)*A
zP{WM>j)~%o{2@33xq$?m1F!+01S5ZW#FXG%=%%XZ7#z6sLN<UAfD&x~xbCV7_<=bB
zs8JP&6Yh(x5#lbvUByCq`s5_o6I0|^vccf!D9#E<4r~(iLEw2}qF7uHfEM)Al6x`g
zt4N2Tn=l~>BsvW6fq_kc1*xq%`3+D*oD%j6fExt`M+Qt0$l?OI3siyq0ubctp(FZD
zU~~giG`DjJH)7~!NHBK5np3sG%0qSTT)vh`4j8G+by-hfmSigL?*?bM1e0s6A%QCe
z&QRYu&#TJ8wkhrI1<iUPDC!!|!WeN4dw-0fh0)+N1NO)&3?5#07P7lxjEq;S50Kr|
z?}NB|d(ycOiwpShb1r`}a-}u?s^m9En%<aioV8jS$M0P0fQ?nx4sL#|L<3x-Mutqo
zSvtXd6(^ZndSAjq@PW&?XLXh~W+}6Ar4_HWRmT`%j`Q|YG2P|$WRqsoz&l5+;HAAV
zK2}l<JeSN(v1#?R0LU049W5L>Rg961E3G8W-R|Q__J`Z8mHhO6QSVfn*RfqoS6X(7
zZt91R^>2~eq?_wVD0^eW7+Iw1mtHp$v1wP}juY}&t1h7SuJR?4_E9@d{>}M(Xu!XQ
z+^3{|LW~iNIz0u@tBMYd%L0<I2ib1NfGjT4=FdzT;Mx$RlIv2Ef!>0X!&)E?2h0%M
zJzcuC!q~3_qryN1jObyKAs+xk|Mgk~91mAknwu$@C_4fOhBE%c?7;mNuB`wIa2ZvT
z3od{Sf}WF{D=L8jB!ZA5Q$_iYpS9dE7#t+(+9e3PKpVgqlLoO)#Qt#cQ3P%b#7L>i
zRT%BFPr`MT3m~s!06u^b93zlC2}TIa3S=h%{@}I>h6|$Aai?__$_TiDyDi6*k=_6-
zwxzE#ff40V5Or?<f#Qjj)c31cC9djF4uM9bs`QJCF+iAs01slD5CdkYW2SgmC>7}p
zeL27bIpu=xLJSvNWC6M;v-AiUauJm8><4Ej0(-o)C6{0q0=57|wny$TTtIe#Cq=o^
zYRE@3T$(I^5wF?JZh{7~21W!k3-%1h(S#<aI-E48oT`$T%>Fwf=5*qr+w^@ft&SL#
zaHW++j1hBfoOmH}d}5s5;us^`=hL|6$Z_{KqGg}~14b^ifZ$Gyo7oJS$Ym92&(12p
z3&$yoXL_sNa4|-Fmpe)27PT@xslCndY&3AxIu7=Bb_2HOny(gmZGA_`m6jIQbq8C)
z%XgaDG_=OG_#o~&bz@O;O>~gwu`A1K6xCt$cb08UGR|h%-Yj+5IFGydQm+%N)Rk7$
zO;6l|`lQ_nc4TZ|m2ldMu=7_wK8rpUR%ZLrXRg)#`WqikbV+nA;{tv}_3Zx!A;VQt
zM=LeE$MT`xtOw<hA*8DSt>-eFuquZ#B=TB$j+j9cqZ7Xmy1n2|HCz#VDnm?De#!l0
zWucSS<oaVEqfL~S?o+94b%0DX`j97JZfNL_yH_xt+G=vY%~|(N)@jR|jJBn`zn@f}
zwHKF{bLbf3qUB4f$3K3c6%utB;>?)Lxj5bVWo#cgNVI8#8w1Fh45JtqS;4;}H!OK^
zr8SLS((;D7zJimY*CXqfh{+Od*RI2|w$Xn2?S^;4^B`L&^Y#1E?@v!-NQL+X`Y){i
zDle49T652hDOUfnR;f1>EU&0Bh<ZqjIY$bb=VFYAF;dqGAdazOu%`VZ8JJ6bF#%K;
ziRXlLbp^xU_SfXI-_vZ<bTqD>q;~B%y`_HomXvH|L-f4=ZlpdNE28<5e*67mzpZO?
zJlJ-woB%k68mE!3D`rGZu{<#TcVw6+z&gfx8YCaA43}Sqe71lVq_T|l#{wuAIL^iF
zC<z>aO+h>jD&7)!002XsRn-R&L4qZL>yp0848$TK(EF*YM=q?yu7D8%;Gk+9ut6QX
zvvkOnCFTgA46F`h<TrJmitSPnc!6{@#WCcCwL)E;R!{6CK!*!4R{%?3M}z<kKnLIu
zvJn4+iqFJuK`$^zs-svZ)JLN3PWQ~QKllwJguv*;Fd;4qaZRWbcU%?CbpSNb2YGq|
zREW(&UB%*-%Zs}(YDRx9Hi_gbsUCDs9RyqtU3Hy1U<5FvAVBewQ9^0yQ;n*f(3KZ*
z;fAZEm}6&c#WNx9i{hi4YobWofGLy*C}F>{BY?83V=@L2cZhq2D`xku-P6=-DZGQ+
zWsn4+ZYb4?nI~A8Nw(4Vz5%9_l(jr@@^TLOOjgV5Z<ESd9?mllYG_5V>Bg@!)EYVo
z^<m|r+kp(Staijb#H!My^M)+tJ32bJT($4(ZCo2sU?lI~2Nr{nd9K#Cj=dlJ`!3B$
z8K)=B>VwwNkro-$1&xXMHkzX+>DP|?ZSt(!9m{c&-QVs<j{cbZzDUS?e0UJ-@89(g
zyM5lVry9>;v~KV01SO4sd{<D0p7u4TiGH)#Tm&81a}w2y{XT=4H(K1mG0NB$0sL>B
zM^Ci6{QmCP`@spWKYl%S%)eH4jBWI|U)J9b^M-wA-@ne$e9_<JpC(bB<hNnv`37NL
z(9r$Ld7l+w*VwjJ#(rD9adlfB`xdtAVc)Nm^~Z4XbB&!%LMzh`wE&}_XX%~-WuNN1
zc1QBj3#*4YF!zhez|&0~B%Yyg@2bxx563xAIm>5T9ga=b=0GklxyPu3_N${gIECz6
z&f4)S3gx<L1zD|q+ouhG%&~r1J*?8N$Qa#IHS;ifeEDdDcH@$t&)>Ja_>V^y$bWeF
zOxLdH7z3@*wIW~KHKDceSq|rk)=$$h=P5T$ijyy?Jm1=8&#=DR*pgI+<Bzq-r72gj
z(J?Bm4Cj!I`^DO_V{v_aCOc16^S_erE1t$I6Ls%t9?ZLGoW5A#Ar86nULQjb6(4G+
zxf8TbT0vW`v#JGshiY!y)#9dZ?>u<r<-2YiFH3$)TuI_((trPc;5_H-Ipw#15Ad60
z=fgYV>KW%g!*!bJH_UzYm%512WbXvZo6haNXB*ag)y`E7;U4wi>XGr$lGWuO=NNNt
zYT6meW%>Or`J(rOf1Z8+ncz3J5yn5t8vEPw(0^aI)<?>bIN#U1(YIBFXRiNc{9;@o
zKgq6zzi-EogJk`%GT~9=*K6Lf)QiE17J^-17}vLHy;Qv%_pH2?p>C4@{^7S>EA~mW
zMIWQb!?VeMku}tPC@o_-@h|vg{qReNN5gxD{it63yo?`wA7eKmnVzu!hB^(~=A6lG
zTaKkAUk=KUtp2b+*MFzUX>D46Y-<a;I8U8%`n9qqgA-&V3;WUYJo6{=e9_P%$)}1V
zx+nAXGJugzpzr*+7$ZzX_*{%8yD|JuMDwiMhtK=p_>k*sctOKK(;l#B-}r4ef}OLA
z$!~w(&ZlkBGcOJjHxhN5j_o&jjuhCwE_Wg|O=|pPd7{U{zSz<5)*rB<RDMpv*24BI
z2V3h`_}Iih#Fi(_fAygHzx_NMC;58u_rtT(|BY*7l3M6nSe9)#7)(|!)CboB4PI%H
zm(0K!gsG1Y+ql)9(LN&s+GFe2g)DJ><CrfmgAmuAvNpN4_Nn*i*<rB1+r0d+re0E9
z-Qk(5zfoDQ&P=w)#2lk~c-MtGP8v^W%aYy8);@h>pW{|nDF4;9W8k#s$Lg77Yi&l$
zno*nU$|TR8ooK#Ee_K1YrCjuygm=dE8vXp&&xH4!-3$J?@QyLvg?hAlxPMM$5Z=)X
zry%1(c$^-F6Hq9N28>WA=TBNETD@O!;4FG4p^<lwPlJLcn1!8s&={@KCx2f?V2sqP
z2qI8wMMm3gwPoe(nj&FnK$S6D<?MRW0`-UeaK%P{!aA%x=j~o0LwGja^UGR(wxL$Y
z6Q*z+w#j~Yo<3Nj46^Y~^p9J+Y%NF_ZThmaZ!($Q{65l#$>hr$mJ9pOd46-gqiX|W
z1VDqtG1<jk9VO|Q#5jMbkC=v74nvLm=-pjs<zpWs26<YVa5lt1IO)qjyXtfO)H(7F
zwXL{gGW;tq6SS%f_8l4S-2@JoPO6){&YhIDfa%(}HpoLc{xQL2HLt5vA3S!QeO<Qo
zL16a>bppDC{T6_tdZ~xHxv$vr;pC13oajH(9qr=&i?Th*S^;G6&Z8g8%keP3?m^{6
zfAop|(AP87g|3jB-)$M?%ESxNZ~DV;z?G_WL9NKfKjcS7&cSv;hL(U4OO=vXCCbNj
z<1zwps7lusZn_-Mp#zjh7WATBm*0RX2Mp=?0(2pD1sK63+1`?De7x3cK~vY(bZqx7
zV;LW|Xo%~8p7^KconAOc8IEzDeEdUa<YC|HWS_)u+vj(fGpse#eb|PTBahX?*4hsD
z`9=P)?&!P2<Zzkv$La}NyZRmWLHE5^e^{@T3_qaEFhhR+v;3wT`)&1GY5RSuRZK9Y
zni}ImiNjXdkLtz$?e9~9UWavtV>QQSO@7^0@2qu2%URtA&z=VRckep6IcDDrc0O9p
zey`j20Sl+XdsyRuuBhIa?mE#=lAW{uS^v3~rvkBA6cD|yqxUd%kRQXs;P_ZfxP}|w
z3L00!`mgQ{+6h|@w><qkBg{K#+v*DQA@kL>;g}_42wSxy({26x%K99w!H)Y|`f2T2
z9k!)BeevfX{yFq>a;SwCrD4#J+d%ZnKgzHTpP#UP_DNhb>M&y-$LKppxo0V5*Tat4
zZ_dW)iT0=y-CMokZ<|Y9)fv_iwrPF$`-D8y=C>8rel?bnhw~(phi#Y`$3xa4cc$qU
z8SR<0{*d3=vC>z^thR8A3|F;<*M+iS`S5-Wb6%aZ{&C#zQ`ldVEIdaZ>LQ=LldL`K
zzh58yp?}=Ze%X!#w?jaNJS6e~0<;taxTpE){(~3XT*l@X)~@x3_DtSr+x+Glhbe5E
zzo8z)&~<WY-k^NEn@gH|n2g~_aaEH3u&5C*gmkEdLru-mOb+gE78lW`y{|b(vg>K`
zam@C~Yxji@j6`*IB4cz@3z%$Np$<P;9saSnee#jGUUuz}#bmWI<O-z^$12;m^Q;2i
zD{FCy{^Nbp?w$uPJqTkHQkKy#Y$@!kuTKoM;k!G0Cd_baeZIQethM;9$sP7-vQpN6
zYNUkUPmWog>_^W{s$<GpvOe+KBs7qgW7J7Ic28O!Ya4kuX8QN@Q=UW~lI63qY{^gE
zZ#_N;8h09+?}qbw%WG0uzsZlii8=7q804xCo;_0?atqtmV&7A_-fGHaF3gd0f!+O!
zVsQVZ{h+I7y5{E8;kOJW0Wny_VsasmJ!iI0|Jb*_gj$h-T?>?$KT{_%+i~`Jvcvr0
zaqcmGbB=VZZ=;45uv#C(XN@yq9;?e_;f%@PZ#hmf*=Wc5LfP=%rcI8L?AZ$YY4^V6
zwXMk+mbv<C?NB3m>^Jh#f{icyVtnHsLtc{A8U6}o=9v9<nIeCzUOP-)>mSD`O&iuO
z@`REgi@vpV&ey$tTUs4QN17MaszFVQy^`dh4eGUim@Ks8OK*9i$1VT3Eiz%VFo2Wa
zc0ah?(Vwt?Vat=Wg<RG*N`>XGuAlYG>N0)$ZL)9uv*V$B)Pt1aFTZ~BUHuN_<(Rd@
zz9l;r-CCPr{b3!JC$6j=V@q!9`;{DF4#su9^N;08VpmQf-daNN^>&~|hd=zo|LetH
z`)l9tq<GJV`3MLT>a+dPA-6Fl+p%b#M~@!cDYwo0im+E+dGVUO(H}{Ai01QuPg<t0
z59%U@8z~WE3vN&<GA^2qnP}mv1XD#(Mvg4fq_Snv*1Fb_rlW$c*9YJK0j+F!S*wHx
z!54q;OTpWZAG=9$v|RLeR3GEcNApkm`%H3WalgzWA8i+OuSUm(No9}bLk1G{O)F1H
za#Y1~NJ~9^ob<JFd)8zz24<0uHtCZdb2o9DRR&*2v((}1mHlDN{gZT0g@#*N)hkU(
z@UOdHm){1tqraoaI8XBH3y;MA_3Ojn%tUxx^2?0N%Xtjg_wW7x?7e50RmHaNUC@DU
za#Rse5X^!Rb3hCjw>e-I^LFQR?zzwV-1q%>?}u~GKIiOY76a%uXWeGpCd`UiF`%f3
zAgSr5>BRg0)#|aRRcp=_x*G)S!9&lv=B!z>X3d&aV~qa@L*3XhV>GU+-NGLhxzE1%
zLY@yK9WS5k{%nn`*-7UJ^E7h!a5rkyR$3FBoA=Fp_szH8sWUSQ0jr~0;q<B`a6+a~
zkKfHI0VlNE?q}DAeP?a3bK$n1EOw2;a`r4J_v<C>7=>l*SiEiaAOLO7FjjWHlxvg9
z80ca3l)L46$P_B<+c8FFIL5+$qg)bkt$`pgf_RYd8&Hb1K65U!jo(XJeYA~MQa8=D
z_l?&|wHN4eucTf;89au<=dx0qx4xuKtDp1XIV-~!3eRPZ<o>0cJtJo*Z<JO?+VSwZ
zupT?kmTSXvNp*!~;`3Hl{9IPf&eP}30A}jz@P2l_tGD(}iq#uFTd}eoIUlYE<=ksi
zlMf2@UD!v3%7y2+rYDx28ca#!DeUWTpE}~-94lpBqFkwzbdT^_=CD4BAwOcvE#*<v
z-_<GvFk`k+O3>xYmb(QD7b?foN<SBb*V#U6*WiHz+?Kk&x9X|Tn%68vS)KBVEcjue
z`(fcCc?$}V?m`~Z?p^D>tk+s9+wpK)udtznNbCqztIfTN@S`ny)>ZQOE{BjK0qjF6
z_MXDB9o;W=E-F;NZ>PpgKY@XT2-$d5BUwZ8Dd}1Hj@Af(U_7#h(N=PT#gufdl?l(q
z-x2k#CF;$rm1~c!-OH2>@4fcg%$l!nU8h+e*HWg=unlX^iART5((~}o+Jdn}Pgybd
zlmeLcaxITcP4q!W=A-bQ>vEsAb>%nRN8y>a-*%iEAOLmwiskN;xu50mqn&w{w0B~4
z@=k{i8PYB%)VLp{ADEjUCsb&B_Lc!lA%eGPY|lFvmMhi<E8CTG8U{j)BgRu26>FER
zZ;V<0T_QUP3}pO6#p=(LD(!DtTjJyVhSHwP&Rd(~a&di`+B)~QSe?br7rJ|At_|xA
z>q*;BDj${&&$Az&<2Qs(f_zaJoh4CxT~e71{aB%A;M<Fzy-@j%JZo3iSEvo;+>5dh
zV_;#`n9PqZ`a8mX<L1wu>udX8|Ld}BtvzOb`_Y<zKl%)RV23>>W>0i<XJfMnGgtYJ
zKJU7`4sQf#)sp3I;Gq7p4$UaLI&)8~Qu?l=#CV|<LucB!p<atM@5#p=<F+3+Hrqr(
z**yFFi|*C8-p(FdA3J8Do)kW6g>usLPRL1vTWF13|JEgi7E7JIPmDMxJ!_$R?78a%
zcl2S07AoSNdErGj`}H>y%ar0Cop|8}ey~tK&j0#5H+Yb|Kca-`i<c_Ag<PN;#JXTv
zZ`Kk&u4;5^>Xk^Zd5zn0_|QD&G7P!0F1ux2&TD23vCED-xMPnvOyQr_enVf)|Hj?*
z;KOdUEQ`8=(%uci){&DAaeM7H!3+D`3r77l%{$L%{t1QZN&D~T4%~O|g3WpGA5Xa1
z@4wffwE%$T-+pz4)}yuk5Z`Q>EPSrI>6W5Or1dc;#ckVAey6?bj<k=uRxWLDefuki
z$%+O=n~|rM{-I=TYX|Ns<Y}ZA;(yl+$s{jsR&E&*y+b~syMX_m-U~&RY6_Ry!{?om
z=Y>M6wJO(&FK0?P$tV@tC#MZz+p6V(#yD(PzRKr#sa6=VMujus9OW9-F3bIrxd#d%
zc^WB9n^A=T20UI+gb^yO&4(k)@IB3|R%etxItP(AXO+UY^&430Ey-<*c-2oiY#`<;
zERGuGA=gX!Y7ndNMrx1;Pm|oVtCcHfb)E9(sExiYJd5rRu*9=uo=+ZC@|J7S^AkR;
zNg>tx>YfdCvMj>YU7_FFq|hb5S4ds?e#3fy>U{*2r9Uoqg9Q>7qNF+b=OiECG18|!
zjAWsHvd|C9WuVb2j}gNEU>)12Z)08Y>V?|V(Q`uv4{`_Zx3Al7n{8ZGyZ2~ZSBvJF
z&%T`J{_*s)QdF+ZpTwFvagW{Ho)gAPnb|+S%>U+lT)6m0_t=w9y6*&#B#p1c3rjqB
z+6l+I?Z<ADT3WYvH_W`vEnKuXwM?NC6L;IyO_@BYP!adhr*qxCk35!Brla#n+&{+8
z5-u5!(=ap&vSw$(E9Eq}%pVoX5BC3DzE&xbbom!weeG_)_kMq>l(r%RH7dmC>N-3|
zG*9VU!dL?^A?cg4zKZ*r-*}5!IGv=nam=m{&xgE?)ob56<-+s)PO5|BAz94z@L0H?
zCVz4N7Om&~d-u+ZBoJQzQ|;q1(pKBY4UtuNY5maRdMgZC5-|^*bLz=1G~QV$8Ra&O
zpGBHK)-~<C!}jjDBM;A!6TbfDTX*M!4>=6+Al}FL(PP}Xr=MD|Iq!b(k$dQgCkvJ-
zbaLv+C(4j>>(nBEyw}gT-7WgDc)xtNXqot)-Z^|uxjLWa^(*A=Ez@s-J@pNEPbo*G
z@(1=+{=vExO>WCAhA2F<Ld@zrWxj`3L)K(W7y!t_5CRLLA0hkAK6(S@IF4aek7xJF
zWeQ_9SUGG`bYFydg*UXh!M<1OgQ#<f_)o@XDU|e`#&ssuQ*7+pdD|w{!LdSfWU*`0
z>g=lHVL#w;MmWUYTL1AT3D4R6c$bY!m%B>wMT0bdL2;N~E8D&OX?nn-t90ev8K+S8
zyO*Km_dP~+<t^NH{Z_h<-+JBG`Qn^6vUMkh(#oFgz|O`l_-wn)n6jRFMO|G%M!Byn
zVVH1HuGLFfkwgpSR=D%lqud}RTTP;RS<2K(=<X>7ZT@%PNyuNFRIH?9#vGSaxVunJ
z4~&qt_W5U=mW|JEClT=eHvMMTq}cBp0;MIYDwm1r$g`v|=H=EkN*05JZVG9clm_K6
zw!4TyJ?!9vWO-Plu>H9fbyja7V8qJ&oVGA@039(f(x}h`m70Sr?z!~|j3?u*PwiR{
z0z1dHtHwiE=Vn8PN_kRxO#dcX8X%a!XhSGKx7k`=7h<}h7=ZH!^@n+u`PS&t^UiUb
z4b8av+cCEHNs;lP6dLP-jENo4ub;c(!t?UL2%hb_TW30e2%~-W+`}EP_g>yx%AV2A
z*%n~ry^lV&pLyFP2Gt6oT)6DKb6sCCs%aLmr2->XXZN=AJ%FO6Lf#-GvRtuvxgXaR
zW=x?m50U5{BDV>&rd?3?N!@jhEmWxxAI&QTuCxOq{$M+R;`jBZm%@7jAhZbB=po=_
z4c@J6)rK_^xN8(9p+$LGTH5&)pASXnLL%Wk;D<g=e>kkCO#&^}uO46#c!9;(8XZH)
z1=vEq7HIX#jD<D#2FTz)Uv;Cd0f_0{x0eTvIA7bRm(No}4jA7q-4;-!g-~aBUg<Y&
zfT)M(Bc}}8HszkF()G>CFVjQkc+LuiE*q$0KeuG}D-|-cZv>wEPAl?fW0w!&$1~+7
zp5Mc_*xYSBa-`dA*ictnqdZF*OjvfqA1?lJi33DmxM-2`C1rrUxGsLvuWRJ(S1Sch
zyw)a_U!TWpElRrQ)}uzcQ;s`Eyw+L^=+(--^Me$TTWr2rj*^h;=YH{}yYqpEd{w>V
zeR;}p$NDwd>SnFR6T6QTA^o-Xg@mW}Rnu>Ab>dUHk}!I}=*hpk_W3fbpDC&{z{vQW
zc65gwbbxX|$)mG<o(=0~KM5G&xEW`#Un&)PU!~SjtlW@6gJejPE4N}KGGCso#%~p?
zqmS02X;aToF0dMhS4|Ryz({;P={ey^Ta*t>{#QQNR(#F+{#}Tl>dLkh;<wY}!mhl3
zSXYhaz)@Qc4-cj8uWDH7D&(>C-M0(ekCHV^(Y>YSROWy}1pof|mtSSa+mVwFc6;x!
zd)_5)zxTd-^r>g^j(7EeOU^yZZ9Z&RUbXi>_PG1t<4>~3w^Y=L(@!|g^XX|uMx6yl
zj-PUbL#LHCrxgQ6Y<;8AjNyLbCo`O6_A$!ZY1DVwEPjf-lvP?QnxsElrQ9L|`YBA3
zfU$1snjR<wjC9Q%y0(^L^W!3UnhugNS25%8jqSNR(u-lP=Z;*zv0ejjygou{(LY@Z
zjC?0yykW`0oPOWXLuthxBmeIoew)+&Vn615@S%J7iKlXk8KX96zyNpUMHl#UtrCX*
z_tn?9o)WH;=)eQ^b(8klC+Sqqu^D&W<GzxTZ(}0skBzC#v^8!w%G9Z+oRr5H;OhUU
zYj2Rt>sr8yi5|C5vp`6D_F@Y{+I5D|CIMZYVSq!6ChomR%CTX}A5jYNEEQsWdPvae
z-7{IhDWx71t|*~HA)D*I`+lLj<&L|gY{*rv6{-^^4u!=@#~$tY9%Cv5jBJRDY&<6I
zd6LEv0t4zsi%n@|i)~}!3h<*<o*}rp$I2SUU$cPQY6&$>?eE2e8t%hl)>HBK)2zFS
zwI%613XI*y?~-&X@z~-eOT1MSb1;Mw>-Se&>J=1;7qlPz=DP*%mOJnE0v7<oFQ=Wa
z+&%f$tdBkYtb6O-_X<_z&vj-zVFh>j`R962C%lmtY-Y;hE3$+m1j*>FwsI#NePrHu
z#P!91k$4^m@5w&TGPG|$cgTMGc`ugud`Ezhw6Pu6ZDYMN{TbJ}sr)7xcnFMG;Deq5
zHSrv&Af7W8dF{Z6SU`D(tkk$t-Vu2j0y+SmR0x!A)Av$cD{xwqC|(jiF?`rjD(gW7
z0VC#dK>~V@E@<MZ6(CW+Oo@^KG}`9|zyuP@YjG6Fq3=<30CJ#2^8ky^ao&RxI&T05
zb&$v&S}iXTs7HH#6@Ux8PN)X}gKP2PSh2j(=l}49^#`D)Q3+5X?*Ssf#seI>PtAbd
z-n&H&@xX|IBD4|k2EuCwPz3S-iTb3vSB2}VoG~?lP;YV9eZk;Ql3@bYkplux6K_DH
z_k;o>iB_n5?3t2OjyapWImP&oq-#q#7OoNYJ;v819pBV|5!QvBl!s)00nq^V#R#C_
zpD)jLpML&DR*@H8vtq?k*F)i!Hbh_~Y-hONN8w;DI{PdE7`es~x7~A}`||7gZij8h
zxswH&zz-WOS-Q+!ee<op52hS8$xYmKS3Bjmfc6hP{-jsvop;8mZm9CvgmmL=clfYe
z;jvQolg6Z7hZ3+)WiS6evG|USX*vsx*nQw3jyYnofS|4IP<ERJjARRFhjo&kEau@+
z=vl-1I<tSyX{Wf+0>wfi9N5)2-RhQ0=~)h8{4r(nB)8qze6O7l7>Q*bdrsSOPh_A*
zbrK<k>8J2q*rxTRWMV$XBHHA(unpzz8=n;S1NDUSa$K*;g?N33JbGNNtA39jHOifH
z+R0^Ahk@Zzh1@eb^Q05pxG|&cXW|y#;otvywGVAg0l<J=cit(l?8kFIbN4*_Xx{Op
z0}GX#X7m2#E9zMfuDIYlH%Q?CV|x7AS?-N@-px78cRXasVE2nDN6K4b^PF@1c;|x;
z+#^psop+oAX=6GrjJ$&v%Q1&f&YROhV8q5%TD^=nlFrlSj*<Z*l>)b$<;`5t*x>5L
zTJ{saCM&CKB!Z6C-+f)-9$^0Y;hP`a$WiHlWs+Ql*9MmRA+H+VX~B{eZusE-c{;PO
zOlS6EU?ho`rv9Y1+kHcj&yIC(H{2dT7D9ft`G&P^eI#F}N4gXkp~8(_@ImP|Xy8D%
z)rc*#7R5YQOh3q-`dh`koj?Cu_wI)uWluBlkn|P77OMrU7mOZ%_Bkz#bL?l%cKfko
z-L_-LD8BAMH&Rv>p`|Yi|AK`-$ogurd-?S@-4cbh>8ADH0~rBuYMuq#C!c@r?tS>N
zqG4l_TFeCBC}mDRF$qSC<?=j4?&WKz1UplFi#7LVVzjW<>8aSA-)r17tXft4!jux_
zTB;Rd%u{ZWcX+p5cJ{8w@d*^$({I1iTV2NIiv6Cr`)+Q^p@-xZD+G+BExvh`Y@i1U
zF$sm5ob(P&sEUCRlE5`cv0ANo;3z&wB^gpjEsMiC;{EXoX@z_E7y*@O)@00?Sm-WU
zwoEb28{CVpzLxzynOhG(_(1P{0pKuF-bnxnMgStrwM&;TcdxwhR#t(9V6lfneS}0j
z@SFdBm;14c=H#?7hb1KnqAfPtEWEu33jxakXOrk~c}(rQ*B*JNiUA{e7eoil%Z7IS
z?jaHgW&i*{07*naRPsymEIm5{j6}ECXn$JYtnUXvu<fOB4=}QFX}t#{GSDdA4}{dx
z7!jA*L$R$ZWQDm#>khx$1Z?&oECq^7UB5;NVtc4O-XlE#F!1uQbE*T+4SsVDB1B0=
zwY)1762W_M=rG$Fg``>~@Bj}j0E=D<L)9Xnp;g8qfFFFLG{D#IYs!ER{$Bb`QgiAd
zq!yrvh1SZrb!5I*zzW*o^R;LTpwXgs;AOH}3F&M5^^%gTxxEg+1|<d+xKSaG7|ZcO
zzf=hTB1Sv+q`idt;(qJ{>fmj{zAl${iNdS&>#h5!Kcy@fpt?%&7)krK+aG<sZA?km
znHwgp5vA0(vC9YHGom<0fzEnlM5vT6=g)T}3+*EwZu@Pw&GVLE+yka#!5kCoC6=@#
zM2*+lxNK*CGafan%Ql>d?|Z=B6WyT)?w@mbg%UlZ<b-GAyZNQO9g(XblyU!j`4#ul
zYp=V3GFG`%7S`46IkBt~!czhsL2>=x%8PUQJ@>jE0%1+jY;&$9rL}r&TPMI6nas-D
z@w9D{LYC?3nwc})`=88nTMX`}p3}Ue387rW6Ye=;_!jO$c}c`n3NXSorF_HddcMJ1
z1nc-DYxgIg`=`Quz7s!Q;_tAX!-fvgcsM&V*9N6SU?e!~ca!=8e${EtZW7?17CtAg
zXMOzc=y)l$ZH4CLG#<Xxy2~xc_}Fjao_?$kQ_hVvB8=0QUw!50yXRh-?N;jD**CS9
z!q4HQjgj0K8T#N|W(1G<<=0>Pb(`^KhEF?fznwer=qY(R6oA%3UeZ=?XSalJCX91P
zc%R=aqdttZl4$Jc(QfL=C*+;#tjCB|$Q*pu$tQZ5$&RrNfsycd(tfFPXVSHHj4{|#
zo?)x-_9KzDyn|NOH*0O^>pe-*3_|Ssa<)nHLZRzB`fYYazZ7a$N6sy8UM&zKH{wP|
z>e*1PStNZc=>)1$42G<oOZ1%$jv|fNxp0f*@%@j>-H^e3i-t)Iuj}6b>3V>*ts59w
zNB!S(!CN^s#~veg5{qO&k|dfZ!RPjSAMlGx(s?@;Thn~I;Cpw|-|w{3saxXo9=zW^
z+4FNGKt1@xlUnGL!vSRr{!Dh#EE?9w<8QIqFn8cSdwT^XmQkOJp?m4o*E<D1bY;8o
zin#oObF-EqcRuj2PyE)E_b60XA;!GWb#@L0?(9=f&a*<Z^9a+kUVPbpmb~3YigXym
zm<j<SaoL~C@46MsT)+NNcmy%DNggBgNnJyolt6v6!c5ptogEKhUGY8^s?4!~$uSW+
zYsTI8W<ibkeE9pQ!w+?PO~?o`04X=l{Ci@~5JF(=H*qgnG>-A!CRqQ@e*F#iH3`s5
zpw7-)SfTB==k9JamZswDzW(-G_rfc$DY@i|)VuFCULfNUhv%K|EHDyQ_{$5=&kLvV
za|TAjzFFV<Ewb!dDWGG;k1M3y7qB64yiM_(+hnley+U*h0vsW<RjbbV;~Lv|k^qQ6
zc#O1?rvwlLQdzt=GT@3piJY)gcz-DDl{`13D|D6euxL_vt4x@z%yGg{^;B4{3IGSd
z2fz%xF{Hd|RTwLBiy))}gj8uC&y!XGGF%5$l6NG?MdA;(|LLQgNaXtfJmJ2z{RPsh
z{s!eYAtV^o+^A>MGxX_S;~E>9eH-x}A+JiMa-#r(RCv!5d68%h!jr_eS?#M*9ZlTZ
z138(SQg4f%S*6?n<tZrz7|C3@Y5eWzyuK*`Be7*Np`WlwX6}RNqaAkG$(?!PFY-bT
z0;vD#x*L^CCl}DJlgCJ{5(o1kt!rS!H?qhdEJ1rK%!xjyR4*nJ{3!ZEB9zjt0_-#3
zr*o4I*w1_HE?-fn93x*UMCAg1F3kw$nww{2m55u77$GCG-8_#?y>sTgqq(HCSnpw*
z%Gpl}$HV6zhtfisrmoR{{+3RJw)vM|=I7NY6&NYyPG$jKO4*K-K@Q@*CK1LZynp8O
zatH-DAw%h(kO~1KVL97J0lOOPT<hQZ@Q$$i7TXr{#4UDxvGaxI<@F7WSX%**Pder(
z?>T12Yzvt2U)Nmk7B61tdT8D&<OK!bak4!BA;#78+wV}gy+u}4&K8f6-6!ng`uFYY
z0BHfCF<`a%D5r22{{ZCOf3H2=cH_2LD<{11rVr<vcB`Q>(7O1Xe2}rLfD!JS1lS}2
zBWZGqfy-&nP|9(}V70#Sx_UyV=^IzxfR&Bnq2yg0dfk>%*ZM9K_6z%-v?i>SSM-*{
zGCBK_+O(m0Dimu^_?$`4^HUsKBm>`pLk4Dnk+gYqS<~wE5<0P19~C?QQ}p4+eouJg
z1>HGD*3D}y6&T4_Ih~*A5?P9J7RK&(mtCAyQ1dh_$}s%JtcwZRbe2Lk*)@Os>s7L%
zkdWW5wD2c7_-9=+%N=ai27B4YHo;*Xj+MC4#s-WKGU1|grg|$&tZd>uBUs#^xK6W>
zx95(x;WrhYD+G)rxvGcPZYulX7>8+UUAZzZ$4DWeigRg-!9s$X^<ObY&B`Th(-bT7
z=6~Y|<xz?t6ncU@`@&1|%3GU+E9ss;<*`!wGi)CMA_}~iVqhd319916zfojh9T-x%
zz({AtKzQGu`@V(Js;_JEK!-j6-z8X7uMwExL6E?}g73qC5j-QR`YJDol8gg_wAN?x
zdQ|pS;_(KBkdlG}K!mVX03;qfP+0&6$P%klC*CL)f{Jq<;7~b06Q6ip%>o=DKS%|@
zgp!{35HJCt0q=!3MvWA*ByTs$BysQQR+HmH0E&gX0??q%eyaqTP&OBc%6ULXAdB}*
zQQLYefoPM!7?Q;|D+dZhVt0sVvAQYWs}{gU`G(|uvNlP*yj?h`XQ-^|vB3c&>x0KD
zcOLSct3R6xFoMS-UImO*qw8nnCg!$2BGeY4q++`2rd#Cwy)5T2p|nsK^^ylySMd8e
zm9K*zJpaF5d3jz~$}i@9<!-<4fueUjQeM1!@4j1Jxj$b&T?)<QP)ydZc^7rzfg|MU
zzfY8>fT=VFMml;+Ji7mN9RnjBZCfeV;dP5=$545vlT)NhYeE?10EO#yx89Zow@Rs#
zGF=IbV93*~^@QZjA$5k=3G3Mq_X~X+g?RXN{VW3q^j9d_3v$BgaZkp{ze~TdZ0V0)
z=2$zM=C^1^&7<Tog76f%?VkJd3vHxF$kTI)z9+t;kdlCrj*j1>Cd&Y9w_Hz_Bw!@G
zf4JY#2n1mrjKNlnN)(!zkdbQyjI3I%&~5@oHpUzy8SWZjyqnf#55&Mo+P!{CPnnK>
zmub(^eLS}ZSlc#ukCC)Va#Im}cqNA)qA(;%WE<1_ihujSKc0vW=lmYO(+=*0qo(8>
z<HrlHyz1S^^G*Z@LT?q4|6V5nD$QCuxU7436YYUgC)~b@8-LN+XS#l}k}=v4fRXQh
zSn9UiELU-A?Tfbu?Yp--WdHrVLYr9J?1#z+lx}ze;1N&`bp#kGr~XZ;Y?5$QNT^y-
zU+1lWh(8-sQeANwes?t1GlDd^KWfBeheW=gzutJ0PoN$ikN3TRfBrdcs1*BHuMzj$
zEPms%8^Z5W`_tN*28_gsAgz2ywp{^?bmYF>*W~s<okD3<3%F=*YIe()DU!J6O7e<$
zPg5zf*k22dkoSX-R_)&`TEBK8A4yMztRSzD2R#5p00N{0trAGWHWSuLo);<ykkDM;
z=)nlWY7r_60$3nqRjb-#q9bMT3Zb0$8WDiPKA;EShEMt~?+g!$wC573QV1@e+ryO5
zv_s{a8#1{@D&&dc!*fXqMqSlPhTf$10Mw9dy+QR5a;rsNBOc69*ev$Bhf$--@NC=y
zuM?=3a*_bl;JE_%XI6V_`YOz&$~7xwSDV25pJVd%a_1r6f%>zl$OmW03xuS$MkH9r
z@;M~yBi5KJr(NI%^k17qoW$h+dG&R9`O*MJ@EE}ZVM7K+j3;H?#Q-c0r$iVm!eJF7
zLf-7M;|@8+h-Log>u&HNFmq0{|FC{dI@eXl);Tbuegja7sVhB3;<~%?H=(BS)@5wJ
z_SV}DYkuU9ke;9Ys(XI+E8(G{`&|u;0IP>#ii<X2BQH?s+bG1tudBxn4PMUGc-eZ?
z$h=$Ke((M6?RV$6{sRUTR4|9R1qq84hA_{obR#>E28_J(zH*E_`E-{&O=p3Tu4n{f
zkWi7U6hf<2VUavA(%9+-$;;P_>uhc5iu)zi(be^*=|b(^-z{F|`pOum6wkG|uK2gr
zS!~<Y&sgk!@%gU4KRHGQD*1Y`P+GKOLo#T!_h;L<pSpjS-UF;{8$2+Q_Bfl0;KTE6
zHew5R=^6QPV6kNS%MCYq#Y0*@VA+6$Lrf$V!%_^TMM$_Qu9l!ZREZ3UTZ|<%L@eH?
zXU!@Z7(qZG)<5p6SRb@V+1Ju4D{u+)<s_{R+qS;<QH1jhUMZt~LQul8p`%27m9l{C
zsd%$hvLZq`)J%WSFYQW?j@I2(*CAk3Ns&_}uOO^@QEuTC5s#C!`;q?y%kXhy)(Y#C
zEf*M}uNXVzT%b*i(l*7MFLnGBV$9RpZpTRyhqnmvXbBHQ;_C~}JUuU`%bRoFb^mx~
zmR(lTc0&M0LS~_(i#P9v_rm+(`zDkF;W&6deCu67N$;?bP_<GyMo63&C;DY9HR>D!
zPlc4B(pmb1&a|^L-*fzoX>;wpk38m3G!&u{!-u=`mAivJEyv5GP@QS#{5)juPun)l
zL*zLq)E_Zml=N=W?#J;oVB~Hotv+dw;ZD1l7y14m0HP)(M=W*j=_+7^wabrdwS_sy
zuU#6~T{%|#18%~9R=2En<T+TbP+CmhJr&2fwMifW-lkeZ=~vc2)<Otigq$Kik48I4
zLU<_yGJF}vhQdX)$g84N$uetXCETnqR6PYStWn6T4A`Re8o)#28P=vsp|7fxccf97
zHLC?|c;G=`201(cDF84S`yRX@q?Xoa$a01FANpfn9{#t+DFoO8QCGF<#OuV9Zd}v9
z)-^S1-E5x=4QL@`S8rJ*Hm+*%&yfLaG60U{4{6i0^y;Iw%1Z@Mg`6m><;lV`W^$;I
zi-c#OAN61D?p34Y>uS3R^_hlj`hUxvhpb0775U(Q{^m-*{@QcZ$;*<^SRsXd6!P><
z;_V)J>S_1kC!dDJ{XW<9s?8{vLtupU2_Cjm-zaNoOZ$4yTwU(`=+~ikIl=g1Ki8_R
za_j5Ur?>m{#nZBDhdofKz{n}bAM1A9cAQ=2x19k-Si|6vcrHBQngs^K<HsS(a&DOL
zqbq<B_%kn$>D_UE_<4I*ockZYy2A4v;qA)-Mp!HHngjS|W~6RpJmebY;B`snWy}0c
zc<6m6?&)^gHdo#y)adlv{$3Ve+8JOZR6fD4u2f%C3N&AzvG$F}_dL4BA#yAU;CVlM
zW2+?R_iV*`bKSh4aL$R}e<2>;%Y3v~V=6nA)Lz=$RYZP1#!XCbz5AZK>;C%+dW_J%
z-(GT&>n%B#@94j;y+M3V#>+FVFYR|)_?@Kkehz8x56<&#(HG2*>w?mNk=K=z_K9bo
z%ihxXbl#Kk={Bci3#III78tSbD_o1v(X>cbD79}GgYZfi)AdrGyVZ3oT)p_M!3vYq
ztZy}zAyXcu`|!`K2g(6Per8>>-X1pux!2oGgB!87bps<ohjis`siCyECyXbFW#f3K
zoWo6+1ieWFGyHp9M@;%8y5)|$6=OFyE<LgFpn+0kgfZJm*!B76pBDs1h_%1{xN%BO
zHcGiKhWW(XFy072P_q`U<pOhjeBSpzbl-meeI8>*S#wkM7ZvzJvNQl2hy?Rk52X=_
zkLP^wp%N(Wl~@D@@tNmebjzd&$2xQLsI8Q^Y`hY#4fRYAi$B(QU(NqkaYEnoeO@Y0
z$$RsOn#YWGn-5ccvQ%JEhY(bVcl^zQ1>W+6zDZ&*=obJ7k_@L2@uD9W^Aa*mqm<Bj
zALRo`BkuRAEUjn{%F8{)G!9c68FLWA^8%sdl$ra*mwE4tGF;nFR$8O=-VfMkVq&fZ
zV*Rcp(OEn3j88xRf{(ju$Jq9gVw;c;+m9P7U}CVh{(?D#Z2XWThj&Tajj@ki7Yi}w
zzOn*;wi-FYoum9BJlC6Z-V^xotdA*+HCjjjlYD=81(Zsq03*EXBMv!OK+IZ?o3#27
zYo2@Y<-D+X08@-L7)1b_P#|y8t37?nQC?_ok#e%KT{#)v15cRO-hL-M<j3B&W4Cc*
z6=HA1mcxC*@oKEP^}Rs6Q{w0@T(ro2AZ5!ZpM9P)7OgV8Z+N%GXqA}Pr=Nexhe*J}
zYoZb$wn}-5GS%i3tAl&r_sAn?SnZJi&UW5mJGb37+xXlQedWQ#xMs{WD=t5v7xN$1
zPzx6RkR40e3d-EsVcT(j?uy%n;_1Fe|532EFgZC1aS_r&0-zo@W^{NcZ=cXx-+jL@
z@3<WxLHsT|Z11)mJI05c;x}VF(;v-#?&H0`t^Qc}!`hh1JkMcD5<jqif2&I4NmlhJ
zx!VaL@yS{Lbj(3Xlm?8<y!$@))i>Wbz6%mX5+k1Zl6z9ue611hfAEnXOG)!$c(2X?
zBlH7dY3QGknlt-K(H71-KQ37+z-qpGQ||?%f^I6@1B{Ob<qc_UlZS)=lAfBwJqV$2
zCu?*LzynVaKsg^m3t&WwG!J$tTvm-N_Zn9yajmA7UJ9#4*eVEc13;iv-!CAE_tsz%
zY@dTbJ*)cE$P+{N(6u%4)M!z8JTm|%@U~#QBS^QY4*kd^K$n8Im#*{5OTuAE(FX|8
zCgqX8ulDiGsMfZDa8{ac)Nt>0(n~o-Ds+EU<*8L|SS@dpUIH&{ZUopMd{^W02H)-;
z6+Pr#hQSEN2TJO$dLi1zJ;_DVXF#py)HOcbRxbfO++)R(l|J7}ReL@XZ?qxgX$>AD
z@*wF3MtnCG{j(_mBg~UY`5{O)9?nrAF9aVtMsw|wrAyuWAAgeSmB=*)j8tijhY!Ys
ze3W=>a%o{aW5hho{g@x&%a|+Eh<SbUVZ)RkYa8*4BmFvre1fcEWc7Waym@cG_d#mz
zs<ER-t4$fhEv{$&j@MpzPI%9d>J<(cUY}U^$R`l%PUzvxF}}sm6^bk5ojCsDtFQT-
z0%5;$oHpRS1HdwTi_OIo_w_s$=Ku=o)fV`SkLG^n*E9Ipm{6Kls*j`^_NO`nj1UeM
zxoFJRqkLNcvgq4IKQ3`|w8np+_s<;m+e@YBl}r*+F<=BfihOxHiznV<^Ud6Ve*HW@
zpz1u&N53yzwAh1gABzu+`!Kv$NB0>wS17L-YxN?5A2$m87{Akwc{w?mmv6q~PKVr3
zDwPV1pk#-4Nh9M|1rfPVKAG!2*1I>*f@?6o;#+{=nU^ik<t+}`e_vn57fATips|c`
z7Gu>2UIrOv+cBfH9qB#9=quxAC|lpK%F7I&N~R}F663<i+D9Ep;pMDcxP4LiEy-;J
zPdVk#L$q#b90&YiSa#M6FSvzzUv{2$(Fe$V%nisrHXk6r0QSO<lCb=<FTU)wVI}XD
zF>n20Z9~5QKY#qAc!vJ44vEAZGqpWOB;WD9%H<T&@Xa?*)iGGPkH7fB^R)Im-U6y_
zzwKCW(8|0^ev)i|h)<zE7O1b^`|u+_KZSfTe0JzT7}w43*{{FpbuUTh(tr^Q+fK+j
z#sqUO@0vLU<E6!#Q$CtIR~|R-C+0;A@6{1t#Kt7^d!IoA+=098=C;<@9;keWHix74
zFkK3{T=?F@z6sl4z=&;`%gHgav4+wLpRId;-Sxnd70X?JfufZ~lCQ737IfGAQ}+OK
zMK>_Aj^08!zz9t^@tC8$(k5<Bp;+`N(uhBtMkwD<!X*(x0?Y%1;um-ORV1pyYBPc1
zfvGuKRv1a(4UAYUu&`jg^2S?k)=MwvFm;ssU%&rNQj_wIeKr3ZAI^vqY5ijtEfK@{
z^mEU<(@!|w4&`kheEbPnc7Ns$JLo_k12b;3U4Rl0;ppYp-pFaYowqI5<DP%mK?mfy
zcUzfoYl7HgPd}3f7*N+AuDCR@f2|G_^;h3K!vhF*48i53V>5~3>{z%(k$Q_jucHn-
z)a^EYmsEz3w&9Y_bF&}PmSP5{DRfpj1QT=p%-cOfW!H@!HOd`x#N-@c1&c&1fnf~8
z_rds~JuqyvC-!`Zg;u*qDcjHrZKm==FwQ1O;QGZ;N9J9A=K~LUz$ouT%7IdV5d<6J
zyp{qU@a%w=q0mQPU3|_|hhWV4+vFlX@9X*577=#h*OyF703ku%z6DGvh!8^l2O}8b
zQ;t2x4I4Vd4rRC8BYZb@f<%?~3lIo}F3z2M`l(`SM_94cExBFrYD4%->wmi#0qTGM
z>Cah(0>ByqT7L=bMhNG$H6BZh#V0j4FduThqlEBz<wY0dDH(YFzukCqDuBlOMyWq&
zzyNQBM%fTVq&;ad(8IP8G6W0Rn0PmM4U)(;3^S4G4-`6BwXPKfIsVu)vm`Ko=&Rx!
z;|ZZPjj)=#_STts^OluQ14dA0AQT_4_e2lUSov_vJ3$D=mA^@f^)%`zFv1w&+a5hC
zKfFSGFF*Ic@|v5iB$jDHSA5OS@wa)57%<}9wCD)E%QhKb;2q+B`a^3yM=IK_)H&yK
zgXnVvpztBAbc`@lc%=bqRQ0aO03(cD0SF{jZxfJDj*!g#bg|C)1myx6EFTAXKU$U3
zgE|PKWgrPTKUN9!fGp${VXT@2%2eyVc!Uu4s%cfTtCiPBy~2dyoiboxt&}xb-(rp1
zCQp%O@2NsPc%HNh_@EtjKXQ;%*JKp3Rl42-Dea&OK*%atF7lkcdI^}(Z@xc35z4K?
z^F`Nsph)uu$>yPo9&5UR5#JM~{n-?N5fo_u_Um8f&6{6p-M!=f2eS&qj_^WUQ!R_y
zUIIl9pLB=^t<6}*dKixiEYU;4;NYLLXS-M5d^>L~Wt}_p!2P9I+ubYF!s~3^4c8&g
zvvxlE)H7b89F}K)FBwW4HF*-dWf8_t<cDLejzeFTQW1Uo*_ZQtQu{vr2Kq|?R*pVw
zvNs4Tr4AJN$X>BwT6}Fsfe{QTj?{ar)|{9|7&<-t#8d8qsi)`g+>|c_j1Y<q&m;JC
zdtVt35h=qpCV9^xVIV`uA##9pB_VU3B%YKZh+OjAOS640&P&fbC#x*A`@QzoJL2*F
zX+Mi@O9e(uRw!Db3e1uL6Y@$(#~yKb7TmLE2+ya5eX=r)<*TOOlI2A(EaTnmApY0h
zjg{pai;r;*P!q!)4B$cnL_I+v=1L*^CdlD&#Rcc(fev9EGw;01eWp1IJpwWc27i1%
zM)2c|r=+n2@Cp#rS)lDqg&-Ukgdntg-*D@#Zo%TEsdDh~M;_sJ-DM|V98d(lG$cR?
z!W$N%*!v<Z3lIV?ieafywm&4-#q%vP3!WVG7lc7wAwx*Y-2Koa9++i6rZiv#BS!L2
z@vf3|JXR0$+T9O7l9#i}&a>?ZFv4{JAcrY$DPVXEYH>a!#`=e!&UKG#ti<vs<p4wo
zrPVCi5@V9(3LV+fqssNJR4$-^2Mz1&-tX>##TwrO2Vw+}Y;2y?Lk8-_=hlbkL)yLL
zwsik4y$6^p7Ir6JFTKOcDOoNsva>8%PZBs16BfR=-gQrW*#8chHk3Fq!@5ebyZ@vZ
zx+oQ63Ki-fe)gHW@1aLsT|=F#t*-G5O;{dlq06P%GvwaRVUZ1RvqBaQJ8!>zUcxQP
zA@qRm8^O3?$%Tb@oJiNoVA=8d+jG1!nB-^BNLjFNxy2Sz0FHBbNf@y}Vx@!C5(;j>
zBA6GHaWRwlmaHA9oBR?mBUm56a10W&XBHb)FA58o1>)K!5p|>NJ9xi+a{3+N0A>bn
z2ZWYgcGw}$(rwAIW$upq9&~JtsBinRW8IL!gWMhyc99j(TI+2qlLU;Q^ux-9XXbzW
z%qSY4(K|8$ld%xLN2{MP^miq){4!9ob8d4naZ}p?I9zN|5XaOg0XH0HSmxe#?|oSY
zYJ}e1S*M<qy%tZE7hZkUD|!*OdG5&v9^i({I?ago%(&}bFI;vNFapze>3L^+>p=tt
z5}9HVgD{A|6+atRcjQSZl}Z6dct^Y|1W(>|JodO3U?sxdkFxwE!Dz_*Vbyuw>8I&k
zXIk}!+@SF!GD5+z{kG%W7E*NYuKGhpm!$G<yz{PGAixZdE{O8yopD-@@WOGF#LvrP
z3*nyk!u#R<hXiwa|D%t2rJ)_)dym~Dxam7pJLB<48NNNdI6l+7^Tjvc`QLaeY%^M(
zDiWZ^@3>>mH}$=&$?)QO?Tt62(Cp>M`)R*8F57SP&6Bfc>3M!oU2U>{9pXJ85Xy`g
zn>XlPFD_v1YGn&;e=+4qH(}=;^NQRfA&Pk@iRcFeVCH9(6#yKjD6;Z=V*rqNr7(t)
z`YjC@0VFzE;HU``4YC#lkY+yQ8=`JJRtCxX$_T;p>YHx$9)29dIuxL8*PVCv{b%KD
z+Yw-dK7aMix7>bv?dj(REDrhZx7chmZv|>~+m=|)D01SyvGW}PMg|Iu!aEFsJ|uud
z#xAt~#J%#)<3acIKVQhUWqmNYVf!}peUh&?=)0?5DXVCWYd--<Y1XPFmr%yjA@dY?
zir_g?QK7k`ebUP07g6XB{cch6_38nA1cEgAMC)rbHc7lrm@FT@N^88YQve^Ks{lyw
z)bOFEv~~b)w90FwqEg`&nslweh-N%G<YCdII=QA&-W}e{#J5cor|xTEusj$8kRm`v
zAB=AUp|b!zNS01GEIc}(Rmu@U?vK?1PkL7L&~bT*C~Ow(;odwq?LhOw_g*PMyro56
zA@Y)`73jis9;i{9)+Te8vQcfuTcv)v!fWX{A=*=^`}${6pKJ;qBjtWa9Z~wzcbft*
z!gCNlD$E(eF$jPh1tQkP9qFT$0!C_O9Cg}p$NRhunOZY{QJ|vS!!rag5(~F$Wmr@0
zeCT2K`4{u-kl*5shoUuJyYLJFNa34fEr<bXJS(OvOcF}BkO12OS|IAgf;lGEvDl**
zuU7I2TzJ+QS$+<HB#F;L9*EbQ2OfLEd&gM6jvX`F%Vr_SLLYd~5Uu&h5b)p_cZByq
z7Qpx<$(X^Oqa(lwvJGDMw%%J^@N3_Rrvq5V%LC;<-mLJ}vC%^sFaoc5?&+s^4<hQN
zZ%|g_m5f(@zrKCEMLF_n_@18=zkydlSCZw|p$#_(=)zMOns(Nik_9)<xfQbLwYSXl
zynN1ylpm!6Biv`qR$F=5HugB8YQ_QHIdkO|yF`5I7e6e_#`w?38t3>vqb!jhZoKUd
z&+qaZ!V7=6_#u?>A-@Vw`?@^3@jwEkvoVTM-+mMK%v{|5*S{l0Xu#Ns5KxokvCTUV
z30db)GiJC-c~ly~Z|=X(MDHOU^YvDSdozCdk1-GB;86!lUfN6EN}+7Q_;_Q^9QV5T
z^m-XR4H`JehbXrBi@suB!62tWx(Q_019g97p^##oL3pmz{WQ?$RkiJ@=U+;!2B7A*
zmtE|ii8%$`39@NO7#<NZ1D>AwDgFj2n+A-~4;M~7Gt2wZAFqf%=DlGg#`%Ny+c(cc
z=>f@~c>h6s$BapM{v;1N%5(%6f!}5RLjEN*9qS4p3f_a^cTW98^0@WRdvat_jvI9Z
z7-8OIPK(C^@-H4w%M}W1H_0Y^FCh_$<Mw+W$TO${7^$qSaSh@}$uZJHzZ+M#xq;#p
ze{Np4VSjb+7w&<M03(HK>^{4(^+26~kp=4u7`gtd_t*U8__KT;?XbthPWRQOBKTNG
z)QBPdy%yzop@r%GFL`ss%s-3OZ!VeU6}$)sD3x&gjVW!B1e_qW3G~FQ=R6R>Vv^)B
zf;Dnd_$zWV002O=hdd3TWeCS%SpRzC%~{5dV-O24mgS-4MCOchPszHQg*u4<8-91`
z#Zn?<l=HFWJJ-PAP<PxGlza3qK+Z$|c+!WrqC86bQ;t6-$NHS(087(vzr!tEp6mY1
zxp8AgJCw8G`$FN1f)wSa5y}ygxxqw*M8DmB-~D-v7iF+2nI@qw6tXy;1dMPT;y&l;
znE{Yuf&qikL6C;gz`gtMLk`NbBI7z-neR+c+*+Och#5whi>XkK5!!b4DJSJ5#3SMP
z3$MJEQw>b##pj%rWgI#A__MRTB0r~~ewff2A20SBE3}ivfI)<kB0P{0NgiXN9d8B%
zE(GXOs1#r%tOsz-g#YmP%L<zW=$J%UE)oXL2!QWud0vFTl%3-n=Ua!ELm1Ck28C2p
z)64zg@=N^~3Xi|=%B$H>b{wB{z<v^B_7019A&)u<s}K_N!IVQMC4!}duK<8cdWTJo
ztK4}K9{9c>lrwnDASl+8N2K2Xcagl8_(qJd%zWXc7kunsI|k^=H-v?qQ7O4VtZdqr
zb|J9G!a^Yc?QnfB`|7)a9(?>s?;`JiB>Z7M0fdYN0E#97Bg_$(op-i`(2Qj(<?a+v
zHCJIBLfTAW*Dloj916gH7dZR*ym?_U^>1Hy(OFX+G2KJz2ry#T6WZh6M;^_3k0HFD
zbj;CSz_c=KS?lo70tif^a)A*)@0_Z6XXMD7>Jai73(VILYtZ=u2E)EV*}z<mHEcKK
z_P|OdU#}5XQLk@x<%&iRFl6$72oNw^%pOvz;WeUugnB4GeXFh`PJK@SBy9p#di0c+
zpa2E^@nvcT_46PANz~U!S=v)Pk>&odoE{X!GbEF6y;8YIR8BkrAO|^1YI;|>28FXC
zcL%@&$=WGT_$>elLT(ZAE0d=L;6dTA1WEv)0E$%BDl}Jva-}HQdz;#jp^VpwZz~7X
zWCLvo6PEE*(Q~SPqh9h_0SvKxD1aRvc+tg_$3vw>N*CI)Mv6?lXW9g;7y;IhB%bRz
zUZb2OebiSQicn7+@D{awfIJS2$^|{_x<cEeHKI^?JGZIG2k@9E!IB7`1U~ttmtPi`
zxAr|Gm5q;K6*uXigWWNc56xN2;MZ^Z`(0U&M&`ux#aE(e4=E{GZ)m7xW#KysM{0&0
zD3_VX<B+jQ^;sQke~KI<B&Z)Xa%*?~nWy=1H(`6Q{wK_@DXcjMh(sTRyaCGMTkpC%
zkLR+w$^}N~w@c1FD`&l@F8EZ|PV+2A@sDM?@lkd^+ok~{7<!?cHNt@4cJXS+5Js%4
zSma|&VZ=I#tl3qB(){xC&+&2|hB4$9vGp2d`lJIgfLVAy6r_M~rBJEBh}E^#h!O5=
z@e-lD$GLA6#_f+c-sGwTo)b2)rmDi#YF&c>5KcY*SijF7_#2+%dx0HBX7LL@3tvn9
zj43>DkZohSaOxR(axI}sZ@cII>^mAcYJ@xg)KhZCA@w{VSqHGi$nH%D;S&XD@?Jw?
ze!1r68E%pCG{iX>n3oSf^e}IL#5EX~0Y-j2|7*8KAvQ6vJYGgbremNCARR_#;Wxwc
zqBjT);tEB&pq^3){lf30PYHQCu46tTZ)-T7tv;0U<_&^Ag?CVh;%~*urU4^-kAM$G
z0FD>}g@o?nVtwBsNCv3<)IXmOi}`WFH%QpTn9_7UluPp%;T%9N&kTReJL26C>J2bE
zoYN?4p|{wNsUyJ1M1g%%1p0*pm^bsDd%XcG#{s6#JLA+G*^lE%{f#b*Tq13f8{s3{
z6b+?b-ok?h^>;P$AnB%`z6ToGnqBXnRX=@0f)<ctWYCa7(!CYtr;vf-IWefFdv_!5
z0oJzeP+IHg1(xeELNn6bYVVU0$E;hexp+^Y_=8Y1KJ&th?ox&4i3v}NdlDGd%g#SH
z8_Eaf2TSraj}a8<FykTPgp&Tx*H6#69Tw&LC*=6B`g02F2Ld00UraB_@(7FPBx129
z<P?k#fDm=#ZX6R~b|xxTW=t^GF#1?ihlI5o;kshN3JPnekdl}=&Zh|n#1Osgf^)Oh
zYgX5zN{mU2<s=I8zu5UgzzAgkJNDgk4=bMCqA<Vh?)!4w`mw6N=<KO^j6NPFB%FiM
zO`?tfBM3QHPCGy6UU7)3Z{{y>BStA)!Q{zVWjYpcc#L2L7gI+SCZQD?z+M=idgAW8
zdS!lmFRcEOgp{}(zsm(iu!8)h6a}FIpL&wU9mg5>muoDA)@y{kxUD@T7-d32Ft%`e
zY03wfaYB2AR%Qrs(`DUeR(jM4Xasl^6W~&uuY@vK7c$R;mX;hRyxX7T(%w_dtA(oI
zI0Qp`j+E5oGJ-iLyb=P0og20F*6!l7&diyMZo2(W_vze^^XdZ31=RJ|3KY6g<%qrB
z83Kr4EQkG1eSoNcoqls(qDJNl-Y*_tB&+`G^qYLyLTw<A&!Gu|e-bc)_Z-4PD7gHQ
zdo0VJ6B!4lC_zzm{fyf^*c!J55Hl9IItq*+fHL2Pq29P3q1&+hj1^=EWjtpRl?#j@
z@Z(7n65k3SB0^V4gd@a@G9(l-gaQi%`S5&K?Jroe!VMps>(Nq9-5Ka13rUo7X@|<L
zXl`=71rF3}46R(T8l*y?fD}5)EkY7@h;UkM@)W7nI@2tGz`|=0n;fq%b*e81V8VVT
z5xYO+D%!ybA2Lf`8yehM(1AQ2<R9S}W4=vm#DKwll>}b7IRqkPlAsHakdg=R!~+um
z4^mt~c$olzP`Bj=A*@!D!e~JrWYKx*sgOqqo}=UhAtAhPhsxlIvbt`SFHd-`hGh!d
zARq(6bER5d9?b$Z3_QW>gK%2_D}XR;vFc{OWwktBbZsL^=G7(-1gRdpOVkX7^=k4V
zx+<%BxB&ujHWXmQKWqp7G$`*(?{>vuDVKM2o!?Y|5kTE@PCMC;^|b!NlKX>?KUIFD
zcm2Fp%6!$*+~|IH#T7YurKlfeKL!yYp(sZQ7!qM}@u&_dDfCs6@(%?tvHC+|KEC#r
z8HvUnlv`iGh+s<n(>x#mld*To2`7pl-8F2gH}<+tYbR@BNXSP3?6E=Cdmk!nslxas
zQMtfKnk778r0_5f3EuK@@n5yFoQ^387-2q*Lu$sxLsm~HyD{FLe*Q)89ct%0yCruD
zp+O-+XaPh)DD$@+zNP0K;H^R`MVW2o%LPX08+hM1MC|-;zjc4Tb%v{}YjRr-9bnb_
z?WvM&33(M#Y;0ubFbV?*!ax#PoKSGcKQR$TEB4@u-#uO&$Oy5A6$X3AT`@6k@v0@~
zNJuEkuT%I`46kD1+c@;#Nj?ND=RyNTc=#$Qv44O0WnKobYty`n!er@;r8H2;>M3;_
z^8evAGw;4v9%0|OSh;zEfD3ZiKuJbJVIAy`7(U#Crg8YIBt<=ajNT%Yh2r;S^%E+K
zkXaD@fe}xVH#(lqlMgz;F3WB^3XCL;P2?TGIPzJB#F~TdB{tZlOu4`a`hhDnUSkj|
z$#aM>TZE?z30=i?x6I7Z12G0$#ly8oE+eE?%gQ==IB(h>Bjxf|VIOvMKmLYFt%p^1
z{=9G8R-<!0fJ(i7U6kWlRxMsq5-?(*^15h8cbz{?5AgnbD;MUfy5;UC0Y--Iu!|ci
ztI>{rRCL>>BKU+uAY2f`<4M9rXuVCG#W&>YjB+zxzyJ^ablr_vi!JKKV)yr#UE+l;
z+|K`e{f+*1X}}040S@-lLQFUoLhtz86;dpRl%_OCF_vTn3}Z{Y(m2csUJWd|Ndg{j
zq$t#&G$4X=Nx~e*Y4^>t+Cf1U5)8{_VoYO&C+_{^$_On?P!ON(EpoydoO?hD^!F32
zd|;~2IPrKXCCB7kfYKS)>acDrfagT)O;ZdQNqc6(kzq|0_a~M*NggIx_~G`QL}~B1
zkjDsuBFt4tST2$G!P>M`LgqkO>FzaQf;({EeY^r2p^SGN3aeH}M-(Q6Yr5pTvpu+g
z6+22#z7HdBO{0E~88qC<5h$#!Rz7K4E-=E_IPs_{Zs#4h&#4Z^9R@xo!YCx=(Me*)
zxfZv3-0x|7oQou^n|Xt9OC%MA83sTkwy}B2)L|h})yMsRzq=y5%<tn4PS~L&B4%yu
zEk&Qx_w(j`kq2h1(S3d+fr{{bFyDlGLO_F<pLW@K7kBPSC+6G=!TE`&pK>cyZ1@o6
z@o)+kr)(q#DrDZC@}lD#o-IX=h2_h+xC1}P*Rz*Aabk)E5@kwo+l<=E3!S`Ylo9tm
znj3a3iNONU1=x)#2^a}OB3gN@%Y5h(JJ7xbxF#pk4~rMMmtTL=5vDPTlB|>s82M)D
zl5D7m@LJw6?>Z*j?g@io)}D5r{s4f9J!1fD@X+Gj$5aY15?Za}K^F`Ac<KR&#)Pr~
zr9(`G=EAt_PxYIX?7Ibv-Qdj!X026AX-m@BOS&#DM*`uVnjd^9t@^d|XPW>MsE3%#
zOagW(7<*-uasVUVyF#;hYqLBz32y+Hg7%;K1eWx`!cT&153R+Z0(f~4$|{pjB-65L
zff9t40zkmqqDmpJAeM7HO(488S_Mpa(uqm|4}`@cK|E`1o5G0zuy~IQyfEb5(Wty1
zkk4Zx0JC;rAJ-_bg!c7NeHEG$sRP2n1y2<NLjZmtf31Xcj*~M5PzJDr+#}ReCBP?>
zKSiE2>Ib_XKFENPn%>nuTo=FyVY_%v!hp%WWRn0!d`Fe>CvBZ?(wM34k*gRDtKC@T
zgZNwHnOK+i-7DYVC464?F@OPB#;_<fXo$=4d%JD6cBh}1yN<#m;QdMXo{&)PqU1G#
zCzvTi8|L<qFtA|FC44ME2FgI%^8D;q!eTPiQ2hUh7bct*206CY=Ukv4>kAkO@4Ke8
z$^GHVEAtf0ga$@wmqf;Ih39xCj1c0)+muNIM*cSAR<}&K5JJ+DDMPa3v&p>e9Q@Q@
z<h4*Q@06JER>zYirX*mbSwQf=UX+PKVC4y`HbZ<UAuvNC>?Sf$y+YNX6eslY@}|bJ
ztZ0+QI`@D#z~~R~jQN_}W_S)8aUQGs&4=bHYw?=7O8g>WeTz}KzzFqWIe*x}2j)~=
zuY3OAYp<6aAumb+zV=MWk;s#wBAg5Hte!637sI2Fcvha7XG_XOVr6MJ%5V%^Fs4El
zM{$mv1T~B2?5(x+;`3za6nGJH?g!obx*4~zn@4GY8{XJ>aKyyCeD2h<d_r~J5%a;<
z0*N4$(gPH_^f#AXnq&AvPD13Um=}acJ6XA8@aA~u{r7!>?^r&M?_bLAc;2#W2qjD;
zf{>7T@G#+hF~4E-8Mia!Ex8si?2Bbw$UGNPk}L-BMTC#AEORwp$+7%Th%CHqkoW0(
z^gl_m=U%%_@E}BZZ}ww3Bs`zyF~YO*oyE}vxHip0%ks{I^|N0tFhXeaLlpikB<9S&
zO}|;8urhp^okLf2<+Ka@cVfrb#xf6MkWgCn$Ps#{Ycx+a$YW%nypr3h<$X(7*DfUc
z?z^hWuCA}E+SOI{pwp?-oY}vKXGT}uuhe=<^%yC&zV31xR}Y}GX;j$0p9B~gK7NAh
zFBkl8zW&aQ+G<3HdTwJCd<s1*ix!>&F`*;^+z1y(@*!aL&#}MUFx@A`O=8NaD`sxC
z8Z+9R*Pb(i_>B)ek#QMiF-Z!g6}AV4uU4UVVC+I>l`^5_M|e%45H<%%368sFNIVm+
z!Sm+N4-fnOFve=3xc+G%0ig#7)s;k{MYWyB8UY1ysHh|K5h0L5g0TSPNCOm-+JiuY
z(hv)wn7R@eAtckw3e6PLl*yCa#K;<sGI(j=o}EM;0Y*YF%j$#~oGs-nN_7jfz&Qki
zaGYEDcnf2`yuQ_KwuRgtW#3l{g>6sr$l%^s{Ka4qUNk63lRT5+`<4rga6RvUcM%eb
zXoNfiChQzxzz7Ex6Cp4F5yEg<JP%5Uv7<+Oq2q5F#~~0D*THWr*80lQ5J3UuO872d
zN`H0Hg?X11QWVga^eJr#=~*dDo_g|$JYZzR=EK~DXP@PZ#xC_M<kG*3;>@0R@=0M;
z{+XVh^_+X}lesyH6@-vj=*YRSgFmq1JzG}1v85PsfQbhj(ovHMCzi1|O`EUcF<%Za
z5{8De`=Vr;C9vonDR4s1uTuM?5HJFeo<wQDNN7nY-L)qh353Hy=!)r90T*AlhiDAX
zl>&@ddB*V{F25v45ary9Z@%fC`{#>REW0IC2A(%Dk@ylpcSEF%ZL{V~{p$#I#+4mK
zP>aUdl6oo3v|j5&6IdXxhhDWF&}eG+Sn54&us{bX>{N#HfF@9@EPtVjifVxd<mOO_
zErBS25R`2Z&r?xT>pe}(3k6UFFvJ59C?oJx5#b#I8Tfz~hJIzdFF2>RdCw7DTdkz(
z<TAk{Bm<V<wV}{mDx*vM{RwTwy($E3;4y+1NQ+{C8v$sLuY^!tHMLc4r94LZC_y{(
zLEVyi0VI|3>;T}%<O`|LGgVeqc%Xx4^PVBPAK(maFhB>d8J<rX8CTV}R&#LX{&=7C
z5U(&$0M5qlF=BmK>NX7+DR%v)Bp<Loppe~LfO)7`OX~x8Q9O}nz4VgzQcNm$^kgMm
z-!0dp8pSdONo+$Ro<5$o#KFfq{k6B>&g1>V^0W~vV7yx!BvaI_%mumAEV=*Vs%!i`
z!n$mKeE=h@Pc7|672J2rVME-sbI-9RWVgsuv$T$cVdU&2xfAjHkBN{u=J^_)O9Mu3
zn0cF5qK9SaSCp$LSwq6e2XC7s!U7&k_?VJ_kw#6G|0ZCBcNY>0+-IJD(TCx+<Hpa0
z{9&Eu^nP8S=pw8VMq2>J<hMf}iTSwDFwok>_Wnoz;XY8vref3)V8n;JI!9ji?dy%*
z<96jzeE-WYa>Bx-$-;b_KAe`kQA0A&h36!6ECzgq2>zT%1ca0Za0@N>>9<Q{;DuLi
zNXwQJ*7~OKP<9_cfZU7myC;VO`6S~0mvw{t{guD;pdkQh#>fo7KLd>Pml4`;0Y+r(
z5Yl@ee&juXmP_exlr)y?T%j%BE0o6lgkVReWacljJjS4`?M9&o6!~-rJ6(**0Y>oN
z`R%0_<;~y7P5{T@9GTW191p*@Qh<@~zW-4<Nrq$}B@GzCV-L^_*&!x!VG>5p+GXdT
zpZ#hc=*_uufDw2m@>|9-_hQXy*Cfno7%egfhPiFH=O>z@?tbuLc0GArmVgnqEtMLB
z+BV40g-}{es-vqF@}1$Ot<Bxr{2o}}0MQSMPBLidVCCV>7+mGIsQXA~dw_RZqcy6K
zVN)shh=J--t|_GqGRIE>jF4jlp?+y&lN&50dq-$f5`1L)=ofbs9GA+<WHW`XiV3AV
z);rK3C6vALq6^sd)?F|q5KIk<OA8U<59$wAvVXblw(MeX<fKC!36deg1ikgq2j0TE
z6kx<(l=3Hy8B7WVW>pZ={+H7(@OeS(7+V-9;_-)sV1R2li#x*GUu4yh<ZjG0Nk9b0
zVR&GCk_h1w1~n$Y3laxIFb8Lya-vVj7nVbj{WpbZi7gzg46#LtF&a}^n5M1-M(&fa
zKz!Gj(qg2N{{aCwi9&E8JYUFT#D|Mg_#p)4n6QdLh{w|g;2~uE;<DlIaup_w9fU1F
z8XxOdmXU<Yi75?Sv^+d<`S^DlF!HbztaDNntmKElm6$e$6kGn#k0^&@D>SZw$tT%;
z7$OegkFZZzmZgQ!3h!5JACH?~U3_6)v3K<T9#L4hBtp3eFhc0RkN{QxboJG)x;9tH
zpQ7*y6BR}QK@4!z06M%2rR7m5gHc}SL+FWV$+BhM+6=)M>$6hFQe025-@GTnW0BZ9
zB<3WnoCui(A)wO#%**k)@H*-X1<!JT5t1=tX~?}|BGLTY@4lxnjGqeF%}+8Lm$%<t
z35-0gFodtjY8WD1UZ2edyFnU{P}-ar1NpwN{v-h-KkA)dbMvikmBt5hu7&AI`ws7W
z^a;&{ANwl9br&N8j16Kc1f<7hOa0v_AOP#q&eHs^#J+94rMQM3SxeP&>Rcu-j)Co=
zugbZOvP~_kJqTIfq;OMGn6?QhA;~(x2q1(eSt<DyK1xC<KmqED7JZWtAV^D7M)^uY
zcFOoLTr!r?B<S;t5QeL#j#F>#&_S6{S6X`^fCW$tJU=RW5fUo{Zg3880S^rT2=mBj
zT2ZgiSs4WlVXdGRg{bmGo%&CBD*zGh?@99ylu6nyMqWufn*@eXcQ1h!t5*v2@ZKPJ
zyQmGaw$6B=wD>$GLP_L_BEh^pM3+8r^$&nYZ$B^L)sppwp$Td?0E$nvE<gm(1tM%1
z&)KV{!u8ep&H^J?6zh%nx;yeG_SE=p@$$S7GF>Ummb=c|Gi^%pfjtx6qF9TE@#pPa
z(iRKxJ0E-~FBA?uESAojZ#FFFGOUHKn=w;{1{npXokX5sy^XyF?0mV~G+^Y9S6%Ck
zWEN{39V*@>r1b@iG-w^E6?jW1t$lW%;Ep)-(6AYPpIizry!y6s6XwflBrC>KAYPBt
z_^)z+5tM=d^2^Ke+LIP8nSzAcBu7e2Nx+DL_PLACIKyovAS@<~gW#*+)iIbs!N0EP
zaLHl&w##97mA@uqwwT~2FVJ%__lJanl+e+|c)v7YWbQn9hb#A567fz6HAZOZkX~2&
z?iT-xu_xZj)jcYG_`5i_nOzerl=&VBWd*1PUwpklvG^ObYm04{h({hKIpHF;Gu9)J
zD-$EXw7xk=27iYn82ka$#9n0EjNH<lb$Vu8#m|Etp;EHmuP&bE$n_T!0PDLSedLqY
zqX3V+XX5hVZ^C%4TGi|}@88G4ztk^YA)cd`>nFp6q?rFn9_!RM<F0!oTYQ_XnlRc#
z#ei<rxioL;>;>8>2N>ZU|5^%tz^s@E!#n5057i$Zdhjun1;Vlba`cr&S}6q>39n~A
z4H&@?34^30VooE{LRBPleKhwocdtBNVk!q10q7vNuH}rix`;4Bq>-3dhX@mgEE*C9
zsTeOB88Ffy5UfVu!)oanR*GlK<QU17IqjOxY!`~hWK4F`+8*diK5%WV>vpdj7+JUd
zyD`+hzJQUrtABJomakd8=Ad0C`hKzG>zzSaH$}mR&Ph|whjCUz<#ZeqI4OZNj1USt
zEJh%hyDc``%ri)MqL?=Xp=wy56a>PTOG4FQM2QQJ^*D-Jl(Z~vESg)58sT=_c3XdK
z_@@vsLb+3pKUUX`3or3X7d&BLqKrb%gV>tf%Gfpy7$Nx^iIS5j?AP#o7z&K%CuA9}
z<RPK>{L^(eNMOiyk+w1<Dm?4t+|UJ_gQ+C8XUwd2B`|Weax458<>`RAI((mfe3%An
zqiqYtS551qLcj=2Aql_%&e8}^2bB762t4@mtFLoBACj&qS78!bp#d=7tOTP;4EZoA
zNJ&0jDJ4`$X_jkY8TQi@vg825Lnet}crTQ4n{P6t*b~FT{h*-0l8bwV6vu<M<d2Lk
z!Y~nIxES%S_`b19^xk2zy5TnrDLF&p<m=Q|2pHjf5@1T3ZvFcmZvOXc!**a@cA1pZ
z<l{g|gKKvvfWz#PQ11eT0O8rv2&*UD_8Chs{9zqw*OYraZJrLntT9`S5{R0yv<f}*
z;(H-*hixne7-2p-<Ameg*fFEyYO}xb8u|RodEWDZWT8ppK3iBjT?vdlE<lj?U=*(T
zh33RK{0i460VAwwR|$}gg~_CN_tf#JzJVVljK>6cxW|P2kR#>9EsyJ5*T0u5B-H@&
zb%@f05Rzn9YkY;uaNfbv@0ZC_q0zRAh%x{GKmbWZK~xPKxw-dru+SB(`6hT<`@`79
z)JGVC7FqrH93zUY4(QOPMC<?_01MUqdb_6O?I4I0{NAHO^9pgENw^LXGOtY@AAmZ{
zIgkfcbUk6WAPdQbCrPjV0wWaeirgS;1Qw8ty}3!?iONHGkyOYF#1gCnocKH=Qt|@W
zRQK<j0U}hEkXCqVFb7$}bFTHdM&wNbNa9K7y(fucXK^eOqD$T;GUnlTb3>yq#(e-b
zDyx;NL?N`iM~K!lygx|p-cm0RNY4l`QrV|A>miZ>WCR4YE28n1Y17<Je~=SJxA)KL
zffp$)``6&1vhfMuSavO0ROg0oHqgKCOy`#Nx7>B!YTZ=j1NdyhA+komgW9)a&mC`3
zJl-V5W|$9ykXHcvSapYl@ef8oT|wM~ITbkpB{ciAo&UQkY+)SomwOs8l63#|0gRXh
zqHT{j<RG_CLe9P$6k7LFg$+Ux9us^Mz({P-j1m@QY)s_<BPf|!17j*A3&GFAQ^k}7
zj1bQDz+ESJ4-=bL<Lj6+2?qslim?^m-sUN?MEipU-jqr)$jWu_nHUMd1LEC$wTvE;
za$L|pGm;DM2d_g8vCslPJa7AHzzBRJz*ib2y)XLXKXuQ3YCF#~e8|wO@eU!b@J0wJ
z4xx<U3}Xhvv2qMC_?Bw*Z4Kz(FY}G*dioDg{Xm7iMm`S-!?VBLcypetX=MO>rYZjq
z^@jASy!xMtLgoJAs3T<9vXg&DH_FJ2aJ5EwcK6nrz2C%%K2J>A9K~3|Gv>{A-}5;%
z0N=Wrwoy((3>jmBCxthK&|eTni~h%x<*zr~l-EZwPsDZQ03(!7%L&3c2p;y^1>bwa
zO#r^wdn1+;x)K;E6qcNCh4lh2?2yvJ(v|~^^w-$^jR3Tm@Phm7i+p(%z6<a4cz(qY
zj__xUDF|7S+#)NLbX|*&tI+S2t6CMJxn`a4P|TfX@D)Ex_wS9T2MPfrN$b_d^JLvE
zEwu;M1u!zA_fj{1&O5#dFV0yTy(bYEDYc<V<u+BpCoQB0@jG#8g~BB5#I6nF37hgX
z?lWAWXTR~L2R{B)R!1<FFf%A0jXcWkm7p1`N;}RL00e;s00ZmV7~v1g#{0#95fq4L
z%VH)b5-($EWn`8_EPl9GTTD|sA8yltkvkrEC~?8%+O)pA`=N&w!}SYaJ<JK=&R|AE
z0$d_SavG(**ErXPUEh_!2r>0xVqzk0HSw!s>S|zw`e9^<`H!Ghj26n1;|V2`EM&D)
zuEHdY(~awQQh(4cz6+8z$4Q|9ZV-r3G!&wa03+0Q=z#|)JlkG5Rb$~rcq|xZgbRQ*
zl44^`^QsbUPCfaAobtHTU#(CWrers3sd~j}z3Qg5o(wq`wf|sTZ@0}DpL~#<FtJdT
zbd4P=28@vW6CpOP^W6_VP#C|bts4Jq<eh+Cc;z)Ef}Nc`#}4z8CL%Yi|3Dx1D7G5P
zDXbsN69^i3`@|M+aoKWzr&%n9p`(sDd~y~HEaZh13q0ijBeWY2Lfp{-t_l(T!JJ8W
zs2>7BHom4SfsuF)5AT&WC#L1BDF+w<z$4s7XX(RFl@~@ayf-xxEZ|m3iIk+=E9F_6
z1&p-n`ju@0G6XV^oE*h}^?<(KV}ejs8BYyaUyG-}qoP?Bdl2Qk?S+7YR>kq}Dd3?e
zuIrkEDg-3pA<?RDy;9bdfFRZK&>)-?3ZE*059Hp!1H-~@k&mOfsnvrc016fIxBw{O
z{2GBLo&Y;^9}=SDsZk+-6(Sdj53^+k4VwQgzlT3A<(#_813o?5Jvjh8+Vl*B%feF+
z?~#6k`gluLLTurc!qx*Ust;Ntz{3-uN4p1&2SI3~fDuA>HL1;2dj95mB@ERwx9K?T
z>!Z3G1vCwm2TEr>Mly|9|9|lihXBILp>o%Gt9w(G5A6BiQ^t)M?J!g!3<+yU8ex!u
zVTKXB4Ibz5di;$*bF9NVLs%DM2mw$(n0$l<+7;iYD}a$SfH=wc3NZN3GG@Y?GKo$+
z`bdFLJLa5X4JB+9-k3>*5)mbL+_xytub**S-X%$jcHKg}W_~CwJbVa|6jPGo+RBit
z4NLWyl7Nu~KQ4D8h7R<(wkAmVSd5T)Fw7%F7(8l{^0-(X@p;BHVcXz$0bM?xs~m&v
zArs@WW47AL=e&%Mr+J))<vI$CgvKsbAN=<9irTQaou?wqGU1keE}nbtbLX9QssN0&
zN@dDVm$913?Uvtzu{BfWL$7bUhi&N>3<Sv=Lx>~f&@|w|uCZ+~V1(RZ=bwDCpI7V}
zwiur<b}j!;ALS^$RK|gQ1Og&Q;bB4BjZBYFEpOD*PdF|wpOBRcw}4;FQH;y5uCPo;
z_Axs9jY3$(-y@22B2F*|pwnP}fL?h+`4XRhp`=@t3yk1BKUHCX$x&L2kmUfQ2xDjc
z0oWXdO72QvBq<EAwGqIG^@49EBxEXbQd#@#Tsgo9;ogu9J4?(J|9SOw*|zvGsBaFB
zkv0L?Jru^Zv2oK*zHWVJz`p(7P`2`YE;;P^Pu1o^z{pS4rtbT$-yT>Oz{qy}R=SVh
zdfh)r4lpvedGVT4_MGfbjM;q;*Gq{LLEvmTMO?D|Y6-ASVumWj1mM)nf=&65Ha5X$
z^mgO?vxKOlJ(o=w{H~QXCe{n_A_Oyq;LsvD`+q7qF#-pQ)BnElM%7!96*mDexE{-q
zkWdg3+muBauM+4rDOaDA0-Des&??0O?%i8)zT4w~!*F3~ir{tfu}9~{A7w$T)4~JO
zUnffoLd(Qf?9_{5>77qL&c>8ONN(2q<$d#B5%OTP;+UvrWksLWs9#9bPV)CKIlCoV
z<~|{%0VB7FiTiy1*V(pkjn%QsPCNKS*kL{F-}T6&?)^_PUMBSWzg%*0UYuVTt$)Av
zy6iaRUD}xD8Ml<Od0Lc!oVwqX<<&z^KkWqz#(S&!v`XV20cd#SL4Y;OpQPZ%lEl6T
zYddwNJ@bus-sR>gXNX-xS!nOQc6Ub}d~oJR@OQ?2V$SEy&mOVwl<W80V^??dq)FLx
zwi9}c*tlLUw`bxo^QK8^G`Q~oH%V^t69haIW9G3I!%GS!X=Yv^+}v8F%1XrntnAY_
zTTl3Gc7JPkKUpu~<%0sbGZcGgQI~zscF$2;k96mslKH;Edp-5S3+~m|-wF?9_pvsa
zFlJ0%$O99g=Kt`$|F(W}VJ74p@v7hb^b_}+3op!3h9jKbe9t}ZyKfeFVCOP1oCr?u
z$wTUnM;;-%i|<3ezg}|pH^R#C7g7RjBSAAJJfRjT2hLJ?a6o{EF=0LUz*A3V!tn*;
zp0`B|S?+i2<jHRDD5m);)$J4h$&#FReBzNuxIM>@&sd?|CM&e|<gUz*P_xQo1%xnz
z0_M2GCcCk+2rD+05a^zq^-nLrQy1S@NQ@o(UO7&iDPv=b#Pj_$|5zEWrp&L-KQAw=
z0>Us>c}=pC{>`+E>--9=nWo?5^L-Fk)7oJ7t<s#*Q;$fP8|Ksl_SiEu`TED7f9}J4
z&`#=MeS`)N9w@J@$!_$>kwyE1c+~%RT1hoO{>aZY%*BiiqqBZ-LY|eQC0}2>V!59?
zbDyEb?XmMN?xdrpSQXjrLyte{UjN{I|J$C|Zne#bkv_a#ws8ASXwAnV`^GMp)!NV~
zGy?)8xp)}+5OWP-P!5q|;-Dxj1m}4_%q_e_Ye%?G86%E=^2HbKk!POGjs>2Z@y8rZ
z4b2+wt99=xjR(SGh4Zo16K|=PXM${u*<5Ac4DXd=_B_@`w(v*Hh2i@T_o<h$YW!2!
zCiWTIHTwSX1gX~A%Wv*W-8`3{uXQYZUzz)6cs9nqI?*-Cv!kv~0Du4sPr3$Wf0p#S
zO5P)FtvW6x{~CRVnR4nw@iRTeXAm;0P3_=$JUF6O@IB+HQ7NTftHMlpAcf;T4+%L*
zRG$Q40GA4Re*hkMP)7G^suQ3h00FS3N+1bgfEoc(+NnaI1UWsZt6GW9o7({l??Iw^
zEt$Jdk}gmMa07q@Z<4ka%^d=5@EqZJNZQ_{kX*FKgCM$x?c0{yvJU~ukgtSm0a>aA
zf~>~VMfEY)lbqg%0c(F5Rwvi?lD9~24Y>*l`<gBMJ7SE$yH)8O^3K#7{=NJ1nrC8S
zOc+(FkF6d?Qxd_Q8)wP2jE8V6C0%33DBq-a$h?_ER$qM1>bGycZpAW<p}u~+at$qw
zV-uUjx_v8L`?VCRA+-(PBmDW3@Aq7g%L|O!c5F_Y?AtS*(B_HoJI~AxvRsx|#_>lT
z?fLYiwU9OZzplC7duSn7AtS)UvJD9yak_X#D6?*Bys5`q{Q8ht6Du`FTlJfHg1T(v
zaNk{b8Sjo&D9kuyD#|=^pb^@O`FFGop_0OZ{hz;G<(jlM*?44Kf_U~q9wWCs_<){G
zZB-kXoB7(>KOujN`R(-Mk8|U<+s@Z#e<=I!*Gi_+H;6K+h3{YA64&zlQ+3Vukq12M
zGkI7RF8wjP_EH~r<$LK{(_4zZbMoW^fRw-7e6ugfTxaWcGB6_WMZm~+^5CJr=?D4%
zgOv*tWQdoQKk&g>vs^Fn5j?w%9qqb-z5BR-x#BW*bI5>^a6Iwd@Z7ZmsrQq|!G1EF
z!kWGi(I3c0k3IQ}tCt5O-zMKMy~zAxWZyNv?c0XmVSMp?*5@Z5bF`bV!wz;?hqeF<
zGw;3EHEaIiz0lu`0m%BMkNC*nPP;JA^8E9!zH+zfTE-J)A@jOqzn}76;emrUbx8L=
z^^BYI-n$w&J>6yUI7fDyqi|98{o`?W#xIU{JB=F`7I(8`^z-cO+5WjtJMLIFe*5h`
z0P!Ewr@I~sL5Z;dee2g+UD>2#)spF_9CWbw=w0&wM?7VgDy--?GAe7)Iy+urp<*2Z
zz{r2yaHFf&`d6hjtx0Q9e+&uqp6uOkKX$B}dcrSq+Dbcad+<RA@7bvJZ=yi6GxTjh
zk39X1dv(qn{~U}}dK{&6FL|<7DikJj0XanXlqXU#kd*u2J@Ux2&$`MrHnZe3CGiJs
zXZ|s-SQ}qf_z2}w+9#4dI0py}7>sTMV>pPA<?|OV$bO&xV9FnI5nOT3*<ODU7O~{(
z@wdgdfl&nSaJTV0=jEyi>tLV$nElpU?(u&<@2d5V(32RS%UZ}YAp^y9$3rsy{On7A
zIBWyg3?Dqi%fhi&5C&cMYpg=d_u(4L9PsOlE|hL47Z@RY)!$~!6n~Q`&-cunN}J;T
zrrp=5?2;dsW_c9uwdLlU`#xkpB;*a`6uw%U%Q<hq*|+0tDSM6PO!|zu)r{y?Nk3Jg
zZ(Ui~T|;BWW2A?SH0|4_9w><)vA$z#JTm<f{tnw?eP-uZYu&S`D~w6Xa$mj?<~-AL
zmeQ8EjW(}wJOo1QJ{+fBo~uQDA19mVxRqfWKYu)5TidyBxb8E@THV~6`O4N{e*^vD
z{m?g@Zxru~F&6iOKSch3iuFPGywn5HrkFnp&xLi`^O%fHd%`+t4|Ofm_#8MuV6*00
z-Z{_X$CJ)mfATz#JrnJKAL5--2jv+9ynpK(e?j{n?nS+fRU0$m`C|Lwy9)QY&VDn-
z_>S$p**V&2$C<B<zqfZspRgXVe$!XsoN9H2eQa&DzUGcx%kx9=y2~*e57t)7(q~q0
z*e8^MtY592)NAj9ExfSZi*`cXhkg1P**v}<W#aldhRm^`M{)3)vZc~C{Z_co-pWLb
zxq9}aZl~VE-2A5H$)U7Ltu(3J#s-Xxi<t7HXR>1_96iNhMH18P_deh)Pfj}KD1Ykp
zx950k=(t>xbrgexuvx@)Cr&YR=K~LA!|O1ho3PT<qy<A`gq!ar=bn|vK!VWYiVA~m
zi*-`S9O^0NF+!Pr_T1eaamc~!W{Hq5#E&O2SSS=EDX6m-weO^X1X$$E`+BX?KfG?&
zU3PNEOF0`8F>8tA3t{zfIRH(hWJlmb_)Md%Mk=|lJSJkob@~}uhOn4~w5fTF*t5aF
z;#$6!7-f=r7GjnWsPAHoel<^%`C3Iab(LdDI&REbD>MKsfV@IPo3_^bf)Ndg>#&eP
ziIqkj^%%j5lh~<zZy`PY>@0;Mde`l%kPe3*a*$^L2`^ADD@XGP3Co)+K42y}RiLyS
zGo0gDr=E0zS61-bh;JI}sh3`RJ!>h}ksKpdAGt6tlAC_qr;kdY1X#fXjSvb5oOlXh
zg%*<+xN6si2cYdFA4ljSR9d021UPb=zWriJE#rF@`kj`e2hTpn`8lVZ?6;*RbF_Pv
z>oH>2@%+R|ra!ijb!pt!R*o%#1Tlq^JV(0HW28{>^0Y*?v4yj>yBv>^gZAB90(rj2
z$UTaQ{PEn+tjg|fi9HYs5nMv@_4?I~u6~u6TIKm5XGg}vLIOrRSmCSIcOVAAD~tdl
znv{sWhrBiLlE@@HSItuTP^MJ?WQCM+7-qByRIlJ$XI<2}D&_cKJOg9^EWq;O^}#$<
zr7a#M-n)ccAPQ%t56|BoFam%f>y2UU)ErJoExb!wr0nrIM^t~6lBtuNool!?00tf`
zc)*aP9qK)RFkY>$ubzi>sZlY{0ZzmkxP}#tzARveEg#;)MQl3#)C12Mm4}!k0BV}m
zChk)!prw9kz0Ye>rQ@sBj%s<OP*1D!7WGk?jWLv#f7lNGDP^5EfhWAhrYIjUUf|<n
zqL2Cw80_bL!apU=`+vFNCJ#<<Exg9BE|LXs6l&*Ng<HKzmh8nSo(m{XXiF4~@x8H%
zfApzmweEhOt+rT>k!-<Wrw}kgn9yRBBoo1N#~vd*8$FD0fBclg6}Bbc5CmXkv5ZAR
zvJitPqmw*F$Y+PgaZJg;h{B59c$52K$>OYHjqBi{ki}v$o-aPm@4kI{%TVr;yi*2@
zgeT(f2+xFxQo@tlhEQmQh;YEjWt|~B5iouuWDmI`;RECGzFHoHy$p!g&5#S@usZPP
zByL7-Oe5YC{6u_hu~1snck*$^c)l~N`-8c2-Te<g=J`;-E520=UuzyAVR^#V;Wc96
zW8r(_(79oG_S5)HLgZrjgP|ljI0gy4L4J(GJs0yB0bm}YP>ozmE{c!{n}!?#oht7N
zzISv0*D2ye9Jb6`>uZh)^$5tnjHR(-M(26s`$~g9g?PrGSjoG{ndCW)i81?<^56ab
z@=G<otG&S`9{QzHxgH~7y@12?2jOI6kRmL{{=-i`<@1lk$Gg&F1kaSTc@zNs%4rwm
zcs<b;WGReo<8QPaj}gMRF<-`1uFS;sd~+)`hEYZ{1~H<mZy=ObjR!{RSG6cNNo^tz
z)m3@_;^*l8y<zu&8P!-T@%kG$u%G6xYCDv(jmzhp+=zc{Y|phWzt{EZF_M&Hq%-4a
zQxJU8+`duv!F<KULQd>ezzvAx#9w~>bzICVo>6Yagrzx3;vH2#%p<Oc<{nGgB!(u5
zai?q&^U@h$1Z#v}O}o%Diqvm3>&2J7au(18@FWA4!1QJ;*sWZ=O#?<kFetue;_kc3
zQhQQ-*o9$mKzd1RhmaByR(Zs@#2Pt?==U?FfQ|`a;>A~A%VRD!Rba%PA0}qAp+n_?
zu#+E;cr7FmU<3kX%zTxiFtPdo5_guu7<x+9H?O|=ww+4cBHSa~pllBbB^+VN@V-i;
zjshdp_lu*B%zoQeKM5}%lqC;dUW6VBW5I5#&{u>V2#Eyb2<}*E;nI%+oP>*v&m<Zv
z#(b(mbj6g!D0c=Jp-%~ug)%%Otovr%eXk4K@$lnMDjDHNVKJ|+Xwox;p=50SWGoF*
z;=BF!-p9{<v89|<^ySy{ef($}OLo1TFL%rM1?-6d8A5<n30&U4Jv3UW$sFze9R)`0
zdOTcq+G=a>-Gfpi7XGaq@0gq@)*p7fD}j;clzja~37<(s+fF~>_`FkLKCSR%IlxGo
z&{#^Cjqsk``<Wg<U5L#amOT)mv>K!sStS?N3{a6Nr@@o~LiDpz-lv4n!eavP0q>30
z_K;gW<@o?OLb$6|85Nj)Ppj(f*`9}_LJFQX0h%5-QHh=c{(RmMZY56>lD2a{545y<
zs-V=Y8PLZkJ;(b4AfjpIY9&lpB6NkQ;yxbqz@tRxA@9Y(^}<p}*xn4VqWQ`I9kxC!
zS34kHrRM@90qDrIo8mo`;8PxL0ynsxoFg@T<x#R)c~H~_JWT*tpq^D3j}zuC5ALWf
z)dFz%4qK%7YZdrI&J%tExXAF!tq};dF^AIf58i=4#nyQYdR7Zqt8E`Mc8g6xKHypU
z%UfoT2@~8%DS+R8|AWN*cwrw|7$55g+me_T3tuQL#|R#pgyuCvC%d-TmT`=SGm65H
zkTJ+VfTu`E8!|8ww)3!q_V;-b!ei`DmsR<93m3+R{cplq;;EbGjUzs1rudwglH@C<
z(4<gWX*os;Db(?>N2wW85-<{%vETirwBLEh?R{+j!SdKd28t<@+v8R#$)$+wAcdif
zQTC&#jd_zeTo=Yd7!AcV>&2HOH@*=c&i?KUFfv~9_X*<FLs}|tjO%Wh>GqP0dYB9{
zR@AL@SKWH6Uni3C&S3O)qw@I>;t2y9o)g*hvoGg`)n@kri>{n@ejYE!eMr_$0&fV5
zT0DL_0*u(r4%&AgFF)9Cw)Nt@uNMI4Kp4LW!q~i2iflY?-g@stcjx1e$*W0TM#21>
zq(`v2h$=Bbks0Vi<O$}En8+*omm8-mFH*aQX<VSp-`wNMi!R7Z*bVT8PQ}7_mkaK7
z6c`ER4m{r{?7X8lgvBT>sXx+m3u!!VnCCF=s<h<m^HSy~u4nvWv=)P-CcDM=C<hqf
znSWDasK59=@o)Oh<_iNxh%$i8$QoJKkPeK*b(Q-&o|nqKzPtQ}(F4VRkqx7j-K|)c
zJ-}S?t;^L@b@qO#p|rYzk?l(ETbSzKUw%oR!VRVg7jzODVy*Gt*Ib{NNEb!j?=HPK
z58)?{B$nhbBuR<0SWFtlrwougxDjYRsD4N+K7cB;AtWpyrc9m`9`gI8c#Kdm$pW5y
z6VqjtV0k^#7UTG)X~0OD^##`x7Z$-XrdSY466^q&ur{$o_Hi8~4CDD?O0(wQ)PNBb
z<Aikr*o!@O81Li7EYtTfK!U&p;M5q5H7wgm<lCoTu2oaqxBT`(iCo`_@rxB2v2sc5
zU5wHa<pHVyuB1_CfDr&o##TsJz2W7AQW%BC{g3{`0hT~8sxYi!pP`JvMf|A44s|28
z*ur~+A#kNpl3-5Rq;Z)hm>0`2V)w?I;ly@tH#<gHxND`fIan6{#I;7i`<sNK#&$)g
zm1A4iQ0Myf?k6Q=Pajeb;f&<!s}xSAO-c3`4+y-tsqeSfo_Rtf0tS||aZ<zZ9_8*|
zbhg4KL;xCq17jWVl=nwW@-Q}Vv3ryYjNrvKNCFn$5upJ>Lh#;m*YR#|#iEauHOy5c
znvC1lmB0vjt;lbcM0mCwt#KODg966h`!E+6qf9x#2o`zgB;>8K@R9L7yMLG015GWf
zeb}cZO1|#3B@#-dG4tRBVW~W*peji|j-u6vh|+I>06+qYw@Cql%Q~JTBwQ!SI$IxF
zOK#pq6$;A%&|uUPPZ7-zfE9$ZYFwVl5#STE>)s?QZ<A8R{|0=~_dz=<1W>dzXY!$V
z?-t^vc*<-6BU%A603#Is3*f_t)Dj2+*wUz6BDL~J!IJ|Xs7Cj0)%|#0{sUt893!fS
zJSGtL<$39&H7Y}!0cHrx<@?Q-(jWDK5qVbdAOWxeJR)aFTVs>L;#B7V!&ZJ%-iAv5
za-MTj07jV8&p+eT?E3!v>{s3Mv-7R;!#;xd|JPq$t`JD=;qBTx?Qhd>&RWo;K)UFx
zGu##mV-gbU!L?HE#(aHPhW*&H0Hy8kF1<)rmbv%DfaDhOBH#S5)NMIzEw2U`4k#WI
zp@NsLTIs{4#pl+S$4GeJ@jGnmPB<o)ALD!><0^P06vyN?2r111I*I?%RFMdbVBqkF
zD=y7bo&qAI@pvPL%X4^Y#&bAqPZBVKa<8Aj=H5~yV)TKRPiQ#DGhyM5Cng@LG2xMm
z@kvMN+|y6>p4wcG5g)lriV+?ocmu!_hD82@KhK=$`Y6ms5_JX`fwv=^R+wkggMb3e
zj-7J28=>zi2CI_BJP{Yj0khx4J-ycjVP!)(Cv0~ZF(Eu2-v_1pV9D=z7BLn>D&{d_
zWr>Dxv96Cp;yv(>C$JrGM6_W2GmjX%&b9`OR4MudvI%)5k=yAbqe`uV+ikOrJLG`<
z^4dw68<n%AEAj%evj3p`TxT9P#T_tlFT1_plI$I@nLJHA5B*;(hh$hmxxfhXE}lVn
z@_hN#*Evxlm^*hEH`ez@9F2l|GA~<bVQWKI0wW6q^y0xpKg7iQyI68gYykGIcyvPL
z##9b4LYS`KYVC++lQ{ove2@5d3``g>Lh|*Vnm1duUESJKq3|lD3&=OZ48f|n?vDI6
z55lx-L(jytGU4$>iV!hO-hUe!eY;?>y!LVpA~v)(meZbPD_1BSW=*QcNI7+Nm)$sf
zpc@!jhwm+pA)FLStI%YO6$6RhLV`&(X5|Bc9QQx=cz8NzKTP~<ClGkddKdu<K`W#c
z^>wbdloO!~?7v=ldETN$JO1xquMRKu`(c9X@MNjL$OxEth5tcmVuVuVK`~?}9ecDt
zR4T)q28<9I2rrZ*G8f_ST(c}Vue#}$97`yaj7KV#a!AFz2__$SK-PlXuEV<o3q5kC
z#I&ITBT3Ir+hLMX2qzJt8yJr}Dvm3$L-An1lOZG+n~{TuN|||*&!=JMV5a|kz2f33
zfp|!vG9Wx2@9!=p`LRbF?hliAdd6M%CNh8VbwdUZ^2}fyV;My^7FtaSihJtl3@}0{
z0j&E%feT>h1%0Q7DK`s>eZaR+ae?I;)>k1tD{DEdzHrel1O#$TQdj_(V?1MGC41Vl
zl?se7cLB7<;ll7f+h3p<)|T%{7{2#WS)pMK9uaQQLp4SK7{`qnt<WHK?mES-<r|I1
z1A;vsDkxNp5QGRP7z+D#uA^H37~V}tyeAlag!OWgIa(d%03!(aCmu6JYsFSRyvGf)
zLJS2^t8=)7^|MYnF+2ZroNw^Iue~A9%dIPcktEL}YZIUd;7CY}J>onQY9gj`fDyi-
ze~}U?7UrlIOU-=oFZXVt^;zzj?qBo(p|l#C8r_PO@{mNB)}q}euLgurh%i-+imOi^
zD1eWinu}WXT?iHMAV+IsW<F=F_xU&!KFd5qNVdy=!fLhYK0Os0!2=^IT-m4A^9Lky
z2l%KVX}LmE0hCn81Ef`9sf<Y4zIsKy0Ere?A%GBI0-%CK@gb7|kOH-8jj5LB8DI#&
z2<^tJq@^j7Ys7oTs1AMDzC7Sc-~Kgj<#L{dTqOEWKj?ZW6DmwQDsTB!de$iRy24};
z!i%zvD`i}wEg`=M0k%f-J<nRD=VI(OF0Uuvp<?W==~7@MR&M(SZ-9T|dA-yoZVJFi
z9R4Zv77hC$?xSgEo#_)aTN%K+|G4U!9D{%(CLJV&V6LS-p{1A`Eh)5JA8+Hi5wDhC
zT|6z%lYo$ggf5H;?-CM;$J7<c*W)_vcifj8PYR`F<=GC@{PD|6F3EeYrxp5^+;<_F
zEEjsdhUZZrqDYLX9AE@~;@20GSFUePH5mFJqa;z9K~@ql!h79&kKKH(LU=ViT(BgM
z37!})>=@iCBm=~C#J@54yZD@`9$bSzLq;GxaWTSEkNkm2#(V!idv6*h%T?ZaMrGET
zwcRUSEeNeh2(STx8iCDXFqq8-Yy;lF;K3f_rMqY84>MopQ-7G}nVF}jryrUzHh2s+
z#@NQDZMWSRn_YlFED~TL5Ex;!UEQ_ztjx-+`TtJjTlZ8{L}X-U<XUmAR1q0**7uw^
z=RN2DE*!37!YnTI0*pYXxBPsl!R_ab0LY7eG~~N6d+l3uc(OpYUx<Kg2si<ZAgq|?
z{F*Z2w8MEGr(H|iFw4u;4_Sc`(!>e>;fI2~9Hx-ZYtQ-~@oDhX(cliHHGW?DW6Uwn
zc7AH&4)#fhaMv$N>m74feqjHB@Mf>U_&j{RFutMzFc|QLIgh<~C!rHy1TCwVYM+1!
zX+RJJlIV|Um=}NW<~O|lh_F>uIFodF0wd7jci#I)Zhs!F6K2ym8;J&-)U&^dhE9MH
z_bp)!;o2u2$_63E-_e|gMpQS!2S!#Fm)-tw6X)yVt9+{i-Q=A|8hj$)>%sT+*-5?t
zEBH<-LUsz+8l9HnahZp^?9@qpqqAwzf`=YD9b9|uRjF|%;pMpO6Y``r{B)r)k|#yI
zys`izUi2fkw>|3_!E0ao(j(99fiF33&-^Q9cit*#v!w984~p@4<S^r5k{<q{B$_Ko
z9s7a9?f>hY@6yzKcydN1_{QrFgE^$j3XHHw{;DL|IK=Y<wk)xCOSwpVM-SEgxOSfe
zjG#XIM}Phyca~#uxx(-{2i;OV_XD5>D$<znvVi(45;@UBbL_l+@vHY(w1X?{LV7uB
zMhJ<l_x{PB9g-ZeL_-BeKKxg~cfa?bs|wQizM6EBxwQZK(SHo?k@QVm_>t$xKZclO
zf~k;4rC%%^j5M1?3#d~Aj6@H!qlW<&-@iJ#YfRp7ZKxLa@y~oV_?#S7W(kgAlLXuK
z5r&<ETTH8Ai#RSMfd?P^%$)(g6Vd`@V%Ca0EKnXBUx1pp%~PB?qH{WsBtGEHyWaPQ
zsmx`ZKcggI1k+Fee9=I^)Bi~UCJ={S1pFJJjpGOVzA5+pfA;6WcST!&4k9ilQvgPO
z`jxLbNMF^a?03b?Uyzi??#B8N4WcL*i7jL#XOqm``E|;A&s^~Kx4zkJcA{x$4?bx+
zm3-?281XkDm>_uPAH4U#tU|OtuaucU%mrGwxVEm%&>uP=#b+8j1x5fhU+~P^Vm0D0
zt-+=^I%jj}`)fMDi<)0q-`H>(M9lTk&<QX?+Bhjqo@bFp#6-u}{^^?s=4VLPY4DvQ
zcA7RWzV8e#?)@O2na0S49f1e}CONzp65xQ3k^lpIXZV2wtk6+8U<2l+<Wyao-zlxT
z00U@#V2X;cx3eWM1Z@rZ837a#NIHLF$(hFjj35A}-~=6~>lJ|$J6qD?&;c;za)4M9
za6&LG5~m@e_YUZg28T0mC6iT{&cZB~bIPvg<T0zd;0MG38!ngdcm-$gu5AJ}b_HlG
zpR9;4kZ(3=nc!1odRjn-(#^|H$!=9zjrtw|gqmt6fDpksZZ}X}?RG)D1zI9Fk)L0w
z1l5b1!DUyj2HOH;&~o8lU3o33j8WLshm{`8-z@8VTumkc7>R0YGbf@w+m09dpMK-N
zKhQFPHgSFvo^<^agI{{X8yq<2(+@L{pSkOc{-Z<p@F)mf0D72dE(99>Ex<?%Fth}5
zzzBj_n6-&Y2$N>ifRUQCSN`?ee?yv|hb3a;04iHQF<}aQ{D2k!hpuZ{+oB;27<u5)
zM-KTIA?+JqC;tw&+;S)z1R>rjAc3{h>OlnVf9a=wDtNXuaiig@-}q+m?hpJqQ*UtD
zWlM5`E7N}hxc=SS-ljPxoU78zPNFda0O$94FNfK_|BQc@1N1-n%fHGbz(l+Jw0=9I
zzRLXdxEEk#TYI9P_>mtu!VBYf27u?He0^ZL4xuIsfLVOc{<ufji*=E2;xGQ|-#LGs
z7C(Bu`NkWyH~8u3<5*yXy|Vc^@e_0cjIak`FT(e3!OW!5md|LO_*B9?!8yWHZhf*$
zR9%{1d_SF*pIvWKS#OGzh4uzo19^tul!?1j($<IzAJ_iub8Yr#(R7^vBWQ!X@pZ2`
z$O7PuaUK`q_H=yTQS;&50V5~FvkUPhen{H$5y#(SSQ_GEnILLbnmOm^U-3!@+E`q!
z2#)^8_r5RKxp))|N-|-dWzUQi0`?M7?J^%2*`Jj#N_9J^m!)wl0h2h55x=g*eRc}q
z(+FqklM>FJIUAffxzZLG>9oD=G>AIQuiN;T*(PjU*mir7tF^D{HN@>pIX9dYKX`V1
zM*?S6o!?%5$eBB3`s;C09aw8Co|(p^czHfAfRQ|@8ObXPFyb?@k>t*~_ixLlsLz;K
z92nt$|DWF(%t)#+DUc57bIfi<1BcE3;dg%5Emr=!IE|6F38=s%j~`eRk-SFAz(Nu;
z#{0BGb4zd6XFucVlJI|-{d8B#qdiHgn{ks)5-@@!0?v_99rXheIUJ-#=^bYZhB)|x
zz^R@e06GAi08y~@I@?m?Ms*gGJHDv_+JyE?oD>=P|C?WXn={Ab=g)@t9oiU24;d{^
zoCG%SKKNm5*M85U9bA3D;^3if&_qHbDJ~>QkKFr>`<!+{TpIe5ezgEN0Z0)SZhN{+
zj=kdL2d_9(=5QH$`C)ziLa?%OSXDea(IScU7D=PPI1&Cx4!yrGU@U8RiPXGb_`K&H
ziGJU=A8=sbkr#UKbd8Chk*aw##Eqk99DXNh<j{U>$40es*l2NJ90f1QE6r{!?+KVi
zz)^KHL?s-NAOsA=MsPIvjv1rxqQ6n^|8;<bR-f?RzHbid|I_dOu7g{nzzFd+WlHN;
ze({%rr)iU60qM<;e&Md5tj$<?Y3Yy|j;J}b?>_jT`xc^O$@1#<3IGid9Sv;k-Xp2S
zI1P?yT>p2{xQm-d-z*I)v_&kW9cD!H^`%Bl6`i>UlNM1x5J2-iI^2mmowagb_`ExU
zm%QNl_Vds+u1=WbKvK{@SRql=G=ybDy#VQ>QjkvT<z4SSVh)US0HT2Pr1b-mh#wHR
z8D~<BxcIVq*^6EnVEzPi7n2&|*3GPK1sK^8AF;hF@dyJ(NDMfl0EmF8Ea#EjJ!$m;
z0tPyGYwEDjX_Bz1Vtr>m5bg0P(u8@}GY5|79?cQLl)#0#lgk1<09DYe5C|ftsf311
z@VPv6(4+;}0VeZ`3xMIwbZK+EAgAd&a*j@5NX#wD#DcUtaDMLmjtJy10EIZtM~U*Q
z3-}=T;|O2{vsv>pt%Y9@?kzYUB6>D0>Cc(jvdY|6JOLi6ivxh983I6J!?ID8)U9YT
zE0Bh^VOFaophP9g_Xq%yev6*~7}=F}%Ssz*Hmk=ZT(bZpaeVY703&e#b(}dZpGF8K
zG3Qxet&2*p^KN%{TTZ!uLjn|s4+=T_<iHDU%xGY)2DC@G;S183Fdh!=y#HBXOjJ+^
zX?=J%4#g})*JOLn^&Ri||6I9o8*L6W-!9F7=z76ijt{SPYx-MR)2=<bM&qPA$$SoU
zec|(882qI8SAPWo=s+_wD)EhQ)vwCu7>DD2ND~}^SA=JbhWG#ZUy6VGLG-b6`W@$g
z@{>BKzB6)A&E5}TFS9b>XNEQ~{cM4`?FP{YUv?IVga5ZU{smtJZ-2`#YflsM)@TmA
zQ`*d1+7ouVu28?Xzx9_7_B;4oL=z$|#C_`soFRDT&(EZFLKg3P{}CXOcbLQat+)Nk
zAwHDnXq3{o?E)H42zErrBkAszaNOTZNG=-iF?q{PH%1>jO>l%q(llfM=jh95z~m^V
zM=d}Dgq=7rjt0JM4%o8@2C-NDRry(r3a*eR<KcIH|M!E7E#ZS`IUGtq|GCc%UhVDU
zG57Z8fAu%9LV!_oXlPucQQ-qi(K4dHjem_k=QoG9>!&XSzx2k}9r2yQd*&xJuGoKc
z5|YLpnm`Dy`XM^+!Fyw_IBwpG=F7MM5RZO`NoO2?`wt&;e6(M_-Logg^j!2?;XQkP
ztaG4AXJJVi_XuhF=9+oOoXOrMZY(8PA;4M#RvV%LQ;feU&@fsC<AC#&IExzO!M^A9
zuX%N_Er1aLFF!<qkzjjE0%LWZt1XR$a0x=i>-BoDIwVjbZXWgNH!k=7Ia?le&Z;qz
zM_K*IBW)h+ro3*)3BU-!7pm^2{9&XifRR2@Jqs|>uK1|Ln8lBi0JC@gHOXg2l6eyr
zld_n5|4=f{5gTRP{ndN5F}l+g2lMoLnY!X|)<68Wx+EQT*?a#wr-Kea0W<k!@7xAY
zhgsO>%o5_52J&FdN7C)T{M|<a990w8bfGGW^tT^4bo&F@-Hcn9QGx!IU;Md4W`W2b
zGe)SF8p8wAj~c(P1wa}2i{YR<%Ixs&Mlt$G>)3%l&nzTK7Eul$R&-NFI_mSWPk%;B
z@!^mJj3hDkq{crDfL(M$0O0gVnJ9Ze08CBJP%mGW9~{x|@iS+glseMt7NF~g{`zl^
zZ2UOn$C<dPRYOCVyf{r1|1iMv^{=zCL;v$X{W?cWd!8nxV{!Z|CXz6f6b-Z$vjNeu
zNPl5s0PrXpP}K&Ii%NSEZ%vH5F-QOhzacgFsDn$V>a2}tTzdR=#BGGwoY9wQA&F52
z<fDx{I^5*D2HY~!e$l??`#>Txj!BM|4MPq92@n|#fJcA#_a8kn2U$Gkr&s>O%MO86
z(1m(F@W8hYX^c>3Km$x_5n%R_T8jcpmUk~#)Z_8#5)E;OSkXAq-+ad$PDK5d5D!UJ
zHdFqb@$wh{m^;X~d^ili-uP*KcYZ+150H=p4Fa1%G@qh2!LK-Vx!XATl8u%Rs`L2R
zvANBj^R2z<^*`%OLfA8|Y&1D6LdyOlANz#!chl6tIeo%`5H|GvfQA!Nr*WoSFj2z6
zLfnS#gKF=0{`q@;RtNX5e%VW%kClT5ts4N+kAM2k;Ebd^IiTm;<?sPu5>pX`v;&$}
zHpljYDXl%tTNeRFL_d6Cgg!b1Sj<S1L%*Fb5NSc+dqf}L5n7?Gfp3nQBy7vGp$V=A
zzDA@Wf+mRbKO!?&PIF{>&Veo^0UWaeBdX`toVLl7NvW$2bE|T?E)b$LCto2lmj!4t
zD^plJn~_-@01yHoi1Xo5YY6<93H^cyh`<CEW7U8X{DI&wy)1B}Dc>a23Fqnn7WfSz
zE&<bAvog^zjV6gU-!6~sP~&O&vPDT}T@oO1fDu5FTaBlJ#S@iaN6z7!%6nFtG?>@|
zKmp|8o<M#L9mcMzj)lgEBT3solK_mwY3DNcf8w*B3-Fa>fP?WuNbv)v;K$9vm+rYY
z7ThJx9nX1==HXW#;pG4{KKjW|IWt6VZq#>hjkLI*BOijS+swfq`S>S-FWvJsE5>Q3
zM}-4eCs?l@kcrij#`Du|y(M^+_(0Of?GuR0p`vDsBLM*wfc<EIAHvK`9FJ~s(Vpdy
zH|ji?_Z+}uUN^d?@eUfTo;(%&yl4uL?*|Sx5p1%<TX-Tg7ovhRfG2;bH7(903_J?v
zeOQ~L;coRC8tBn_xdUr7EwxVjfqML{w07aYMRx6FR-Fl1#ss`Tl%F8}AINv3N#)Zg
z@b(DV7=K6EfL~wu%2yA@8fEgHy`|3^e?w-LIG~T#*RNyTJ>}A`zHi5@LPVBd8gKBa
z&)wzL@o2r0q|^a;-uI^;iseD9Jg)QueHYzt!Dk|X1z`8kLyyVa%GC$SqC#KxB56Nv
z@43Q9VIC`Lh8DBB@Nx7pad1SB&l1`j-9N&g|A8zN#PL^tS$6+C<@IuD&PD?o-tT<(
z?*$8sm*yEi4t)A8H$O>p0%ozpubKKF*n=RT-xv6;0eHmZn+3+(KYa4j_A|FD^8i{S
zRwv6C%NZQNcM{0s6+iys0Kcly01uA84qtHe7R{Ro0$E`0282M!?hrjRPXcWHkTh;A
z-1&vOHBWuce!46B1ok)AM1)p($#}v_N7Q#9bw=QaeZL<V^Xmf5F4Nxg^AcKw{&n%<
z70pkte%VU|=H28+clU9^PP{;S5FyYg$iX|n^nVo3?YEcb1lr3!*5;n@srJU2KS{%0
zoAxkXm@jXC=F@{`Kl2&RTr^rM%w>LH3{K0hC_t?yekVA$-7E(y;^mGT59QOPxKF3R
zIh_fvE^&?-3S6bfF5Y%R0eONAB+AxFlA*+f?;C_Et<SGL9849?*Kh5(lYd-hYhLi&
z!)9m51SwWFJnaH@SCPbN@g@h+lMYPdfDt5g$pguOXh3CH{EpK~VT1BPnP7UuYktZt
zj(%Kh9Dn2f`-AU3`27Is@^&cit6usNw+Zv}V1fKU-u<4SlHg3wkIQ`=FoFcaSHJcR
z=j($F=2@8>Vv$CpBf3%JIlf3fmtuMe2{Zs9p9z5B0l0!O#sTJ2Zob($pY!X-LXPca
zQ%&X_Ov~-hc)CEE%iSS2&cfpYObnYxBsSPQq0MvC4cE!u@6E9e5y^wL-h2PIf=ADs
z3BG>s*AEz`Bw&R2l<|n_`Q@*D{SYHVJ(CW$qxCh@RRB(?cKdBYS{o+$2a-yIVZU4l
zw$FIl(+;URv-!l~AN7d_I}C_EUy&r(GoSWUZQ!3Ir(l;hCB);P_G|au=X`V^4HswP
z2%vyX<f?!jm@nVohhG-p4x7P8C1J)v!p%3_pnkai06g&H+Wl$8yZ<}i)!2CWknQG&
zA2|~|@tP}KG&VDsD2SSGNRv8@s%euE;7vP6^YkTi!y8}wnj`k2$%F3&b^nvvWCPTo
zoq=XpRKk%X&+q@s4;<KsDLFf^A&<Barz%b#vuXSO_a6-EI=Eb#J8U1Be3+kZyzwR-
z@W1|0zt9(7{_0nQ)8{TYNppZs{H~Z2L%wCeS?0}lb&IQSyJst6)PGY43T#X*P$uRS
zl1#=~Ji9WNpy2>*qK6n3hbvk@Y!dGN=6ykJW6ObvZ;~$+{K;5gvy288;1(UsI`B-<
z11U)Yb>m<Spo4n?z#3AJKm3%ZX!CvapoTbXOuqHNcieaL715J?Ig~&Ol5r6Y|Lq??
z?!ZwtdTeMIOLZNnL$Bxth<X4dedew&9Av}z1a!Jw{Ym}ez#1ZBV#dvhn$H0Otxdc`
ziWxsVabP9y*}USTrXv47Y&I?M{Q!!axfpw1Ttf-~jBLrL$ZpfX0hzs;la`3qzox*D
znToVAFinL9NedL2UlE`o5I~W_Of)8(!}85-rzNs*>at+x>;>gBf1vTPbMB(z3W$(~
zN=Yi!03-VXWY892Yy(!5)NlAK;qaBXiWrumE5HdFBLvc#9}+;5rq&?9nYx^*<KTR+
zS_{p52|%exyJTD1B~HUbKn2<(b%7oDBLP^!PY3~R5`YWLW?>=>vs}}n191re9QaP5
z9!^6>-)ZQBM9$k+LY}JHG9iXz_Z_L8i)c=B#ZAsG6DgEvQ?a5t&c`?obp=i6pQLrZ
zU2`V^7-7ytP$YUNPusu$&_e-cVb6#UoYnZdQT)eKZ+-HC1U%m%YaYCQ+?t2R?SgFU
zzv?9~4W6fivZx>ib10I^@Mav4`D+{6BxoD>{3~_*1qnPv4`V4C0VD*S00Pj|{^V!x
zQoCMzWDVF<d0)Tp{@{M`eGhBx0%&1fVLha6&?7DYIw0*#eUQY3uWV}0d6o`aIY|ED
zr`~$xK-uzv?|>(JRENC{eW&n*@P@2Co1)R;@(H&dvagBqlV}$p3D3HSFcaE3q_cqO
zP9(>pQqniwbVG3c6Ry!(d+ULqf~B!5^e1oj5%_lS4_T2YXI%JwV(BOonsY$O8p{{}
z`1x&VM?}}$O*!<%+!>$}>n44N=7k9+k*C!IZCmn&H$ZT}Z|mdd&IKQo=_?L|vj+BC
z0Dsqs-zGo5Zh$uMyZG}XPXt1u!Ri5j{;a3}@ByzEjR!4A|B&`F-_)K4KZk%tPkZt$
zNA`S{mUiCvt#4`VKX=4LkHz8o%9qMCY2;uWGq&%3e<&n$QR~hLtvME`Gfvy109{tU
znSdS$Yx!f0yxB_uz;R#<j|4CDFPa|^3W)~#nmzcJwGZ^iR5TvHr?lo_S}iK1_)&ot
zpZL@#4hbUp>EY8Wnp=MTEx&wdKgU?2Y=m*xTL9qx&{LiqTr1zA)+Xi$1lW8(ihka`
z-oH)4j8DAgnnT&(<B+{{mH>FjT!coofy8m?Fn@gCx3!1XyaG^xz~`_3@|zD?@L+5*
zZ^6@}g|;ef{HJTK^1%+~B7iRwh@)J-|8Sphl;^g(;Ow4rh;Kgdfa7nmZo%FbUfuuJ
zh|fM7upLbj=4i^If0lF~)dD0w=Qs#;`hsxOhq?NPnA0KxPoyEx|ImUqeNA(k11-W=
zuKziYc!?|+MAs0aU?!L`0Ua5SHeSf{??1tL*d>sj_;J99tgQyCmB|t=$y-AJ+Vv}p
z!@>A@ooTMfk6{6f<Ub0!)jr9;Ch0j23K*dl1u$}J=fDUH8P3T4z(Vk!fB(G>N%x=o
z_r6`WIA0Y>>f>?12>D(oM&h-vdU@i42ZQzDzx{~YtVPWx`SoHWhuV9)fJ8B7du;I|
z{%61VB>@$mZI|@meH<{tf{&Us4!MzrN!n<!;3GAGv=PipT0jE(6)$<QQ|mO_yh-sX
zpG_N5YEc{dwDsTp*4vLTp-Fi$4*t(~{J+kgF}@$NhX4C_e$O#I#7_c7EH5_3m`sb(
z0LF*I*f=w7mX@pUq>_yk4%kg%-H#W&2mFC?|MXpVIi@c9p5I9aCP~gOqj?x;6MY`%
z10x#8|INnj^U}z9rKE?lhOY>m`@4Vqn2XLv{$c0H>k^0))l{Lc|3skUKZ_xa3png!
z<LjFlij$&=riDSKZ()XC`|4K&H%m3y|K=<{-vi$Ql6+6orU1u`ae)o(e|*>PCH9%6
z;Se9Q0RC8DLxhx7RVvR>tu;%BsXrX)v5`hv3$3ko0frh!snk6VU~88?>w7f37<(4j
z5OH`E{dVkq^h$f4`|M{$A1D6yHODw$!6!zX<Y_bmHWwU}_!7tTIns|;T@iwF(KK<t
zF?GO!#0z80un`}y3SfxCD$MBpgf{d^AS2@jtu{0?e0niH|CfLBTZso4(R%Tn{O|wI
zJKP4$e@8ydb@(r0thWmQ{{QV`pKu$%c5z4feLDe!Ug}<#Eon3YH@5{^i|z#cc@QwI
zH6t@sb)1|l;AaC+z=0If%1|de4GWx&12i;rKnH+;9}$2Me2TdEq|;geut487rJg-K
zBhW;@2@V`l-;%bex;PL*8YVo);W}5wcC!Vf@Lmt}Tp$Hi)*tgNf)n>$0V33$Iyt`}
zinFjPjShho^UL!=RaZ1d0DMAC64hOjl?wtm(DI<pDuAjT&;&4OSM3ZP%&S}pi5!F$
z-KdlEJ)(J*;LL6bEGf?^e`m%^^~CwQXnNeF9h27i0~#jWOad@M(^xBC@Vq;mpimMp
z?2~}`lLK9Rg>bN*6rk&jgr24)X!Kn1!Y_K>^Mb_~;U_;m-ynPiybEPS0~#e9Y{8Gj
z1rEf}3Smt^Ga=~!&HS9eQ(?NR(}1}k4vYVXUq9;iq+MR0(>?>OZ-AnW7BE5@X#e6D
z{HPN!@B#F={NOV<l!X6{(_~5VeHE7%&*FS{aKQPK(w<EcVxX@8R9MI0-{O26T6wI$
z>?H^YPrOE^Y%!}9JwUbicEw~cz6b%|5M1jdaJczDzWvwQuI)H5|KksR_(;~6(LrLq
zlYm(;v}Uu03*rObsl9emh}&y?;;t_QpB1l6*ltSdzBF?vd6Z8yJ}C>R#DOvUlYb>Z
zDJsZEA@n(*2-5uXIzUD{B8lgMH|2Y57hd<uR|L=e;inx$V-Ndo@pTVB{4c4%2t52%
za(GU94$R~B6$hE>hHvJZ0_o5cNi!+tr|IVY_GkQ%G*n)GkQAX!O!&1wgW#?J06+jq
zL_t&oi7WwN2%u!vfMo~-MhRD(y4(rrz-PCk2Vg6p33G@K5=Aw{+tq{l761bCxJdz!
z1MraZrk{EBs}lG1tSzi5X#V(hBH!q^PAbcuqv63?5)BBFGS4Ec+KxtiJIvR2X>aG7
z{~}Kx7-7$dAiz#l&etV;a@=^RD2<EX$rM0{V?{y*zSW%M*xkJu7l5F<){TFvK>;Hl
zJ@a5N_3rO~WB(nmeUqcwQ*XbcXX;JL#wU#sEDSj3^BL5tY-wXpHfwmw%})w$@(%2<
zsr?V1`t*_HI7r@8-bsg&YzCeq$GtGjs9fSB0Er_Q<1c^hYX_?0NL3-792bzH0B~p*
zXg7e6NocVkBZXsX-|f=#J`Nax`NHH6n-$Cgp^AyL%B&7TI48mP3U+;wvdY3VT0N*Q
z%pWSJNOZ7SK`IEQt17A5(~_uqSdw2b6-jnX$pfZ0PC~=Vv1>z8xUx{Le*BC(jE$2h
zw76oITesnX&Bs1{r`t605E?w?$+ujJ8o2|YKx+nuI~o>*sy{~uqt{<|y_43Ol{p<Y
zU`Sh`%>o@!>*rwF0upXF-*ls^Gt%Z9<YJ?k{9(9}wn9P<2^KabX(8#r1Z@E}=<UL{
z#R#%_jRp?g(fr67ko0-*UmlVaQ8FE)dC^9gUIM_#MUA_Y8ZQ>m7GZ2!*sIqaFo6_%
zQouBTZ!a9Jf;dS#KRsj4`j&g7eUPLhpNce=V18j<zw))OA26|u>mL{6O8jVmnFnyn
z0zgFlU0}L{!}=s}ma=$fOf3f$Y*OeenCL|gc{ILg4-7M75OYcQOJXX{j77BWX}_~D
zGSP1-3;q2v0i2Koiw5_tPnDvNgL}UIPtG2;z2KnX$v2B0Ixukb5CFhN+R;PbAet*k
zk+J!V+XQ0DV!5R$f}{I}zQ-9)+R@5!S4{^TUYP{ip?9qxNP7ZES-|APvjj*Z1<d?N
zy_hSg)7Pa<`1qN#7WLrD7-J)7CW{VUB;6=a4ry#mN2388UN33gr`~v@({^EuPwVh}
zPlxEx1alNf!?Br-^FI>J?>PSsO=;ET;8!!4=sqJ23V;JY`2ISaB8R?DoXuj!3ZMX!
zRX94Yp@{)#q0rP^sL4f|9Wt{eUS&pF5;cJYu1__;?A67il*7ZODLD+D5?>OgQy$!_
zE{fo^I=BaLXw(Ff=t|i*I1jZ%c0<!&dS`1Srn_bYM%3gO-UYrp@gBq{%<@k08IR4q
zrsl{Q2^rL#50JTv9Ii_wajFauBePzMv#vbSQ7#hf9JcO>CzzdK_13{zwH_=>D!0}G
zOJ*^}rF>_U$J(0Otb2Tz;D8<NmWtBv%bXVRyY(~T3(ZisQ4jt)ALn}_ss59;KLF6Z
z<w-XN*I)BQXZN4E3to}$0_k3Oad=h^lHo6s=0Ky@rht)I&4(rZ;}8ICUYvTv-(wng
zVV-d(O(=bH8`8)?JK|g4`De}7KX5+uET8BVEltc3B9)J(DLgQT=17X)E1@8pSJ@*N
ze;>``+u!}3=HUmO5JudZ7>ytO{b?Ocb7&7pgLYR^s7s<hy2c@Vf>bkWk7<nf>1h+%
z+2KCuM6e+3Ha=!}j&*6$V3vpV0}Yg;TyvNcni6ro8~k$odrZ)<o<;-f8Hd)lXg#A|
zGwR1J@knS9-7Ad?_*)wX9O5I@9T(_h-*gf4)OA{?;jgZ~>I#|IS`D>R)Xq>lVZ&*o
zqHTfzh?NnSchWP)E_|fT`*HE$jn6wR(0h-(#w=2&pxGC-_hf8l4Lka-5C}>N@Ss2X
zoM#Jsm~)^%9e(#c_qcrwY1ZnyGLaju9mYdqQu$WxF)T1<K6m$*>}RGc-wk7kgLZ_2
zFhR<m76;Vu{Op$-0@8o%j^`!<XB;M9kOqCb!1sq~SNn?C!(yt{pO>R~_&hvk7dL93
z`Q)2!bb>bkZ??y!-t6<(W1!ja_?a^rOVQl&{f*}3%U|}ggG@IzHk{Dq!Gkt8n6SHD
zTFhC)S@GSzHUq+#>{&P~V4sJD1Jm|~9vsX*{OF@@4;%+RqxlRyvj*k?&Nc9x7Z>)#
zV}0uLcWDfX?@~Z>8toG13bbPY;8<V6KKhOm&WroraRiUGi>%?UyT5$IN0Y=~+_^?o
zWA~X)y4jf@M|);bbz;oGL+wa2I6Pl?$j#Hf)}mjw0VAAgv1eqw2AZp<wI10X(vLq*
z_k1GEjS>k;thVoOwNsx@iJ7)};g?C&PWZJpSr6>>hWwZ=$gFQafss5LGj9BRo^9_}
zUd$B(Go@7kBb{rE^sD2O^7R>-q&S0k4x{j|f9=0JFj2*Z&vz7<r4Dhx2n)*ZzyD7T
z9VoH^0`y_A=Wsr+K)!JdL|*xiLO%X@>9)P3BXmluG{$@sXwyj@Y}3?9ntsRawm-Af
zi@fp00uVz$d+J^9d%u$;^~Y2+FWx0FwTy?Qm6IKhqi7kE^t+cn?C1B@OTF^$pHZqy
z-fin=er8ImDO<cXfej5E@&i6VPXc~9(5xs+=wMG^0tb{%TSDtBfJqa8L0}4+7k<Dr
zR$0=n&IbslxU_M{e@E4s)CyxZ<d9u@A39VMkg>Y#K#F}GD!9-ppZjS68hb6jBIMx!
z4>19hOI@6X2jGVU4e<TpOkzpP1rt{_r2~{f+hZEd4b`J8zJ%+vfQ>mlE2~XA(&(rQ
z9C7QH4x|8e2$<d?-i*KwHuPHpjVKQjV~XsYveyNm5LRT~VNdU-b!bvm{x#+4oVIUN
zg9T}-KvOhR7FAx-{O2;7z(3v2Lz4!K#C5HGYin!gb-;EqE=uAv=!(XO0V8n{dz~A5
zO&?dMJfF22TkYOc7wdyg`PV$#VR>Ozg?en9-wLj|^3r&S(_FALS+Aqi&a8DFM7%V9
zHL3hO`TeACEHAS)X!~LXg234Ni@*6>XL>H}+h|9VVUDG=dMEW)zn^zYkLW>kKkAt@
zzWPyCTArQOgE5GY5g!<7>YQgryx6##BSmnnD(&g0Hd-3}CgGiLHq5`zvu)?sWlBpz
zDB}W*<k|Ld%q#00*vXp98qsfHWINam-txx(tuUo^OBbfJ20c*QF*Jj!4+p0B?y&GT
zAN>bsGv9t^xyAt_q{GP?Qe8#podQYYrISgh+ggkwgWGLe20^cG>(-BSXbkvj{;2<B
zi<^)%rtDqP!Qdo~2S<n@|0(D><PI3|;?4Q`j{JULf(oF+nVG7EKu74@{E~(nCaVA`
z07VEoV{!gPoaTt6HUT-%{=grIbG|O1U<!$N{dSJn^&azBgr=@eVh)WF(s31d(3Ibk
z(oD<Wh@^O(2`zyQ1V9OVj9|`-0Ek1rn628ChKMuACGf%O2Oxou5eJH>ZBq9jRWL6Q
z1G8BGAb>0QCLz!^z?aZ*yUI{}XhNXv_)?**fJ8L@z^s_wTbT#GNq8?h9sDK_Olc|q
zr1_6>CNXq74;=?!#QM7%8j+?QFk<;n>a`u++P$Atx)kloPJt-Disc3ACT7oIgkz%Z
z@m85&LNl0vwFCSXB0!k-ZA57@N=;0w_bAn~s7r4YU<?{CvMb@7vVLzhO2MiGTKd65
z4N5DcH>wn=+fe{*q!SXhGGGJ{D&+h_fv0xqi}#&TU=+{3SR)1jjQodB-BSP~;{uE%
zIX>=exOX2LvlDNA!|M;s%3#h1d-k1bb0h&HfB2^#I^qnm``RX_WI|0-bxck<yf>Yq
z@JS5t?tUYflkMt$hJwNECw2RqH-0HFb3s4~4d;Sp)+}jE5tny9Q|ph{aT^bQ-Dz7=
z7?!H+xY09-e#P_tP{2%So!`V^xg3)N49F}N00r6>fClcsU%z?A?qW_LLS4YX9)2@$
z%8u5D0E));j^b*)XqnTJiiJQ7nNE<ukNuYUK<7V1bA@d4C}~5Nx}t}sgvv6N(3FV<
z0)T__L8A4?<*8@Qn#^x`PRpmAlk}R(mc|O2CM~c8AYxA%BOz$gl<$xQ)mLRHj`I;B
z-h}#c4=_R~$;=g+CbV%Ez=S%IpW5I;Yd=+yuM(1>f#P;r`awP=024iTnlaLF!K|2>
z|1lZ@>P&)<k+U+%aY|YWorG@Zp#m65r!sV%Ykd=T2#}s)^4_)^TkYOY9z}{4bxHx_
zSF4)mm$V*b<?A{vCu{sX51KFwjTtR7<!%0Vs7u<nVKiwsujE<Rtn^8%ch)?Lc>Se7
z9^VjSa8`WbhWt`i78irvjrCx&Djy@qP`ISOy%nGPFuv{hnsoLw3XF7?x`;gq6kv_$
zH!xy!Dx9wu8Y4CqQm@#z{OMP}@*vJ<<jgM3&g((M$jz{Ukw|qW@@EnQJcvcHU~tEC
zNMRcq|3Ub~AYcz2r<mA9CM1n1d*=fqE<x*0F&=zs^nPE^vv(R!at=6OZwR<JFTWb@
zq+h4|03(FD%xV$b!M?x*cQrtP7kukFNCL!gni}|_5-2jKO>1>+%bC&wu$Vyu10X?p
zqut>U1oCY;UpMnLP74Hd!39Zp$-Gxn_T33+lejCIB`x3uQ(6S4<svY`oWzq4AP@Q0
zw`<Y{k@iUo6af$cuyE$K1n!^#vR4gFZ@PFgf#r}70VKr1mkD87`4KM!Y(gLe^I<MO
zfhvU1!MYqDU$z*m%eM&abbwJS51hfvAXl&|U@2+-a~VzGpKj-&0vJgj4A8aE7`d3}
zKpN_GJND8kKY4GDv$S*?zdFu6jX%gYz2RpMv_*F0tKmI={3p&|5ba9)Hb$8yY4slG
z{wnHzm;#JJ(-?tgqTdT(<S>njUq>kr1xC<JXS33hZ#tTwcrtMmV6GShF!I%{(}y)i
zp7iY7gBj;gmf?n~Qs=`iy73wH-WeB-ktFBq-O!gwrIWsW(hWBRZ+z`f%9)*<s48r2
zZwK#s?;klQ(4Ch5iXVS*@Z4uV>qxRc`>VfkPP>l0Xg%=j9A)@@XHx!zX=!9>J&@97
zuq*qITat>Mm!0*Ba};ZbHb<#JT6=qDeLGl`(^e#p!s$4y6!s+NWU_>>Brk0YKYyde
zT3yn=I;@%1VG>6%1Z>q(YaI(&w1aCB0}LHzl!AKCtKR(xOkUGV%N=W`P$z<4FY7<Q
zo@o5=Z(WB|i#kY~)_jF?I9a0iqvdi>pKei(B$hkTw6U>Z_Jr+sv{E(YKifuwr=5lq
zc{2jq<zr>@?QCw!`J_zJ%;>-o^8yRYi@{rd_Gf}CPn~kn07E|b;lGx)$oDN-DloFW
zv+3-^(w?Rc5BDT-jT2&i*Ecs(4>=AB=<voI0(O`zHhMRC<zXM3@o{c#JGg3f>7ZWi
zzRgZHka6G7K|y`G>7~r9X)T^z`_vRS&)CXf^K!IIdmgO=&k6qQOxn6MGWb@x_Zj=7
zd|W22Ka5kf9m&(`;a|z4DNT-s9D=XO`8q{7Q&IXRYXTH3%*yN(8lbhRG(S|I{l<PU
zt+^C{0SyyphDvGoH22LYZbQz-0Ym^A9FQQb6Z~`V49yV-Y-lb4C}_ypG<jk^3vG>(
z<^j{xa3;7EzceqPMs*@Dv`+}m2|Ip2R2Be6Q)LkFCxTBDG(nuRcGZz*b*<xN;w#>q
z>R6LM5!!%Bta%)+ySB+`sDKmVP#0*yI$9TP^ph$$-z1bn+lj;TX_dbtdO3{~%2T?k
z=)!v(z1LNrd0k1rCmIw!MxbD}aGdrBoy^rnJ<^0A+th*C2zxh-tIxY&j%Y_fD12}y
zA!$u}^!$2o#brweJgw!KL^sPP>Aj^duK81-9Uq#1iM{1$d!UBq@=kbnzf>D<{xiGJ
zYD1m)K7ZMZNiX)_uYTE&J15*$kI#Pb?%)%5e$IXtz{mk5M*GGe6TQ?l?Ym5S*G-xJ
zTF^~;((ql*N;@^YR)U?4wP3qh3Rc9s`NAdkofv=XmzRZhq=SRp5+Axa&8de3f!x1K
zT(|M&RtEI4XJ<Eeq{T8Pew#gV_%x25_RR9IdZJxJSZuA8mM_o5HoCvIw{SmOe|Se+
z%h!Kr_mp+16=8mszoobO*p;;Yz2(*J+D)2vd3N*u?1c-#37PsfZ8z(yC=LDk?zr<g
zdM~7o(I`v#em;JD=o_vJm-6-B+kG^>^^3i?t3BtegtK?*YTuc?v$$EV)a|1DCNAlG
zhp`vd!yheSE<7h#d9x~eOQw*0-p<y)Xj{S@!oS2J;MY-GLU;bNfEj1Chjlk?9EUk%
z|GO#P_>th+><Kc&ymfgqxc@76yO?+0`z3dMsJ?a>7<uaLcVx&aJ<4&~_@q-JzqiKZ
z*`$zP&dE8Urn|hf6x{Lb+k!iub(;i7LkUj~P5<Vj{}_DzzHcTcPI!Ye^|$`Q&jnXq
z5gx$$(Y}1|*Mg6H{F6Ft2tQc=U9b1{Ve^r0>d<IeQa3hPrsZuDlU>u|j?%kfj2CNv
zv1v6*<a>>RfxYI=A>aJmoE)t`Ie5{JJTF+1TC;^O-*Zp!kxzX3(6^X2rg-NwwU!Up
zevOAbzsYX$I<vJCT(Q`h`5B)s+*2;06Nar{^bBiK{OvadY%S`hoOa`s_YG|X_q4`c
z*b!i%Z``yg06Z`c<wlzZ9NHoT0~0vqFUw38ni!x1P6I<@9MHp=&*C{j@5%y-8!~N0
z2>pe$v_E75i}iaSP(&bpQ`##{X_2tjmS;mvGzXHPWg`9p-z4}VK~n|LB>WImIsOx)
z?);yiok4usqC+mOn785{EfN49fD#Al2rQXf2pzM#%IOdBn#5OmMhgIzIcd=V_~3H|
z6JpTHIdT_`08&EwgkNT91~XtG^#Nn109RD5qmA0IDqtpQ{<C^a+BM5O6lZ$lIPDLz
z&?q`CNK=y5=8QI#gXH(4^|0UmeAUf*W%<NiljgHHx`kzj0$p+M<MJ<_<wt>}zRItJ
zd~-5>wWcLw*v^o96{4BZ+IYjYPYkfOaLv_M9T4jDnKQv3f9S)(woD;fNc%R7CZjN~
zSbgo9R&R?lN!Pr-r@ZpXLfZ}TbIS<2^bi<>l1$a<vP`^6V`N+F-EL_{->7&s3Czq*
zwHMgT+GgTOM8|;}Atx8^x$HjL3_o8E)lEp#;`{F{ZQL{7B|W!vN!JYlD$5Z8+@!e2
z>3KKK*F*3m%Sm5%l#JtVyUVG6sX7^loy6^5jgxa`jmTSL<eJKb;Gui(afKQ%!jHb2
z(n@aTXuLU&8Y4-RM~d{NFM4565u$(Uvde-iPF)_%=&-{ApvY&xaCh*j&);Q_GF@>J
zclDK51V8-Lrv%qsb4_d-^1e^bp1=2l9|UW1CimbE9*T`uJZw*aVKTtSjlpf-8>1ED
z_GdmLc*0dz1}9fn1yEiQEG)DRELBY~$?r)^<b5CbfDRighbo>nrg)chFgU5>q1Vms
zA>wxMV^Zna0SStzDHJdtBkT43@StBB69m>*ca?_5{7S`XVPL9iZb|++1PnNnSc-t(
zjL^)Lv`7Si5bBbiBTNZ60Z72)6u`mU@<LEOzlK8(2ND2qpee#cg0pq<+N*_Wr)T7J
zUGLB!L8BxDL;y&X7Eq)W$lrk!0xz786m6hqw2`jkh@G~yfDxhtdQcC{VmVC?`5ghc
zs7hig1X&abZlBQF-2!INu4%<<C~Zw9!JK~*>J0Fr;J_hwHx(?Othjk`S2RK!1??2)
zk3{b&yKuhl2$lIy0gR-Rius)bBewIScz~H+`+VNETjScjAB4%+d|H~Ei{A$S-cP~d
z9nzUaI7$KISEaw`w3m;@80{xFUUzNqG@1EcURrd5ge%L-Ze6rmudS~KACM2gbLTJE
zBX>>vHjE~tIIpDD`%nXmA7f1c#vrC5>jLxGGt%z>BgawecC0&cz?EOeC(YdaNby8T
z>+*m}b8L!ZuE-l0iPIQ?pan1z<6~r9ri)gzu^)E8gwm?&F(`%xt&soq-~P6QIl}`2
zYyX}kHqV_u?^Jfb^SuY{apo(&N$^pUH2{jdU;Yh>&>aN^HD2OsmxaMK6}`Chqx#Iu
zVzsMZKcond+Pt`13r<SIqu21}H@-f&>H6yqWy7H<=J7s%_m_jOeEl22+J&=6%wMIA
zDc+4UFw%`Al+DeD2swZsS-!6URHG<jL@59;vL%P-b!|4yd=>})&Y3xN7O=qo&Rms1
z60|?ivcQ=+rn3MI01Sw>x3MKKL{fY$A0y6Wmgb6S`Trnppm><l0_Xq;aX|osuSF7?
z(Nf+m-y$=W<)FDMkVCwN)54I-^^}<#fbUSj^Kyt@m5I4231!U5gyk;gy=0~fa3qlM
z20Gib0%dA)sy;(q1Xuu4pb=A2z66}L&#K;NgG|fp6#xjqfhkH4=t3LNBEftYS|Iye
z^?N~T=qQL#hcHoys$eS21rX#`n39CwiN*QE`Oo6@``V9nC#}v|=Akn(t9HfWVSv2=
zM$)Jn1ncV;f`!WBz=07)w2bgMQ>%Ur8CQ#Fb-DNbF<I)D$9=o8)$V;XzhVCFrvL?f
zbX^nGvP{Jk;V=b^U)|HZKA%Af{c!63{o36ZxSpXt&%fh25>kA@p^Bi5xm&9R|8)Pi
zg3sLb#b9%5>(G0Cq<tGklL8nyLbl>T1`3!d?hH{Aqc8>y7%`2J&Dy@xx}9IRWVIkM
zg9&(%254tJd(4~P8`gd{Yj;L{jATtW>hbdKN58$J(x-W~k+p4Lz{tD5|Bd~3y!K6w
zp4U9{HtjpH6&7y%uoG_|F$|vhIG~~IPc4|LDjh)x?TuZM)Bssq)Fu~u_I{jsZDOlB
zboa}%Jh@U1f$DHLNO1KNGl+Hr_Ry$bU7L04YVpaB--hT}eW@q+W=D<pG=ohr00akQ
zb#0C>YdIjsPTHRGZnW~Fe9M#jsE+{N5hxqeH)YXsPiZNSa;;v(vAEdzr~U-mMqFsU
zC%fO+xn7c;!$wsH0-_auM$BfROQ1j29;+8^z^3A+4oHk9e)*I^y=aq_W&O7)TK>{&
zeo9>ZRa6^n*ENhLxO;JTifeIqrxf?1f#UALrNP}@S{#CFad#<Fpt!pi+MWA(|L<fU
z<t!s3m#sC|l+c34h}t`TAu}=Wi-4=6-2~)APGgUc6}nja1zW!`53FtAjs@1tEl0+i
zg93k?w<3Qq^ZpgmEi|q$6(@{y1Ro%$FX(<@eC6%Ec|1%UC?kfd+Y+brb*AW^FiMmZ
zUx!}J6Nj(H|HVl`-7dp(7B`(fXTSKdb~tw~Pzk!0E`)I+`vy`7yOx6)q1>KZ?4}+f
zoXCnr%ZM1NxgsmOpRkT!g~cgMtc)|L%UCa;s@b+~P|>C5H}J87bOSz6)5ue$U|as6
z(wmL)O7n=OyjPMjshMFntNl`Yl_H(5t58cQ+)$=M%2*fZc!oN6J;ZyWKWD6&5|h?3
z+@?l%hwUDFkMCJ%*6MjR+`UxL^X&+XRzNF2c}7MZ+~2bHDMoSLa}DdJ;kCV$Ctx_j
z7J&@?fub|&MaFE_Y!csJSQL5*Y0)<xW(jc>uX=*;o+7_wwst1Z$}9Wo%ufvF*nElo
zxWQ}sMKwty)yXCyjR&>P=kIOWOpLk4fwkBDn^UzjLW1an{?+zwMOgAMjfKLz9Yfag
z=YR;$j4X<{Z`mguqD0vVpPitWm5j#)u{)PXWC)r9q*A@+{;c}OLch<TPIl@BAGh4g
z@`Jc?QX%)CxuWup!jG!;L4{M30)%JnL)?Dr{<X`_50e0n`I{-}ppt^14)I-lTVBUE
zvNG=7jFwx`?}isEF+7fO+d{{cy;D=7bdu;*Of#7h#8it3o|m&eIAUb;2O-n8a*w{N
z2bjH0Dwg>TrVCOH1+5Zw`V{9U!DJ8TSMmHgL11y{hc(+;z*|Gdx^t}(GGOvU3(SDo
zo_6|VP^8oLOrF8gSD{|`*&WeZ)AqJg=FP5m>N$OwGe?jd&`8L2!5*xqc$!dV8F;Y1
zTNwKH!^`S$QpZ&{Z{(|A64QM#Ws7(slX#<-;cc#2esh8C25-=-=jco7os;PIkMH7s
zv0tMbi4;9EtIFfTdK@Jdq%?SZa+G3tjpl*eTuWv2M1UIhq`rlXgWfhg&;?;Cx%?o+
z0F<>@D)$ji7WV08NJUDA8N$PWb+R}-pPO6b04ZjfClEui2E}K3#V)971;SE~7Wg65
z;(#4IAEsNhRF-}re8Z-RG#u)=QxYKF{W%WcM}ufKqn*bmBTq*3*RpM9%EWL@cw<mR
zmh#hNh^U`@H&?qH018BL1t;GygcmCFE}Cm~Ubi>a2@#lEwIXKE_D<(!?zlVe#$P*F
z#)>~4itOO(7sP!CgUcam4J*Tc_W^rNzrHz)FW`}5hOj+-km|E`&hvr~Yp`Rn@s5$<
zI8^vYf+o5y`M*wT^}9ceQ>h0<g^hby*1j6uSxrZ<CvY+1Hs=R_g*y43tpm>WnNhz7
zLP!foRn2c>ub9h2qu)k6y*fcAmB{-qTUGliWbZE|C5&ZEXkK<(r`FU-Zy0`l*bjpK
zeLCk|OG^J!-)v#F4pa9aihWqNCUkH;E-a{aN^jjF%y0ORy@qST00_bE`py%@WKLPK
zxm+<=_65$L`DpB)IQE49+Rn-f!z`7GgXO{vgF;D}*~;x2GR_m9&0rMX$kAxpZ`<AU
zzfE2ymFL3wxm}RuR5jy2ls+Sd4eN$=`hUKYo&Go(9}5$w(eOc9wm<2pga1vkO7<Zr
zw(B?s5n$eJJ4Kr~9lDgURqeWGjePJhec-y2SfK%*Z%{9gB4L@$^DSd8?a+p<KrH#z
zztlIHf4#@-rbVnKj&@v6pq6v&YrwJD+Q8#r$MMs<=4KePedfjT?m(F|^cv^YNzBb3
z<%seZ9UfIS4#c_Tn&Mo~sI#&lPV5Hi87~TPYEe&CFnLok+5SZ`BOX}txOe`Tr@`B6
z3elVVHS2UV&}F+Ntpw<A>HE97Or9XoPr*m=tq|@H%g&3#`<b{4FAcI)qvd8Vj&IoB
z&EkfJM0%%#&_1GcHnNpvY^RHlBs5ctQ72v5^D~M@Ne^?_;fO{nj=!`A^6yiP!feIv
zK~g%qJsb;}L_-x49xn~sNec<cWO$!Lhjp>;sU^1>vGR#BG{^%)W`|VaV38J|)_<2d
zzIF}<zQ}A%mAv~NJhvU}JefAJ#)QF2yp{8*r)H(gHdu!1w*gtf-=9S@s1Mx@9BZvg
zqGVf>46iLhKl6E~$HKUqOBzNt2yb?1{0XKb*u~`}6ctF}o|EFMuFoHS;Uu{3_9(@v
z4dZUrK65m4BmPt_NQxIUcBK4WJ$axyPKSBOQBjK8FSbQo)7m`fe3?!EuL9bm26Dtf
zem8lc6V~yxCl`p%f6bZtwGMtQJ^_xkzzyZ~Ag8Qn&bwvB%~R7li;VL-`81lNI&GX9
zT#MdNMedaU<XAO->SdjlE}`%3^wVR#?(bZM&Np|0H8lx?`k9<9*exkiZnlwc7<hhC
z=JyfHo#0tNVE&~yu4U0+&R8@bS8+fraO$a%`?D=pOVB*yw9b*<8tB<*bI<y`)nB<#
zthB*aKpnCy%H+(tlE41bqF@->XPPFI_F7HF#(uZ!d%qL8`}+`QalQ?#*XVBJP_kuK
z%B0_YKF<3R(<{KV+V-1ei|1LN!uJ3<pA-5JGrMvr(TlO>c&Rkwyaq?5QaFups#nrJ
zH(H>5W>mIiIEH<IeN*qoj_G>^@z|T$Xe#uC+<Kq<4mR2w%L=b43CbJG5&bRI3YB2c
z_kZ(S=pEecO2cWnnit#`!7w^CrCMA$nbN*+>{@ki{b095Wg4XIFlebA72!8P-uaJg
ziTHUT*q6G^G}O1<0Ya~ngc=oMN2UC+Yv)-D?TeH15uo~?y_J?4ZtIVgKJri7@GWu8
zW;Ret-}s_Zc+uWhwnr|NCc_QnA75Ik61e-c#;6N$9WxYKFBj<3*Y?Lg+oC&mwuSjP
z<MUZFu&0JqB3t)u=iz8enh$_f<NOn^HI^^QnDkWEjGt*>a6dBtiX4zGXH9ZGm<z->
z^z+=uXYWkTyYgU#=Cpgu83fXwv>>)11#FvRwF*skzPOESHx}aW97KyA7D9y$jVh}w
ze=dIb5`au1cvTl`@xc$H@vN;8!svSGV*}mUwDo)b1={Y6LcLq3JogttIOTKh&Ce@W
zi!~z(YiBo?<$1?$M!s)u+W*d<91Rq(s{rqaYqabd6km==3JVQ9__l5Jc<*$*s9fK@
z+-R}+Ie=w7gpBLaFLGG1oi;@~7Q6DZxcuzX_6Jvh`Yb&N%BN61G?S>HbaJgW<@jmz
z8HlS0w_VQ!7;vQozZ)OH2Jol>MR>vdcA@pV)XY)_Yc!ZZ!L+4_`pxkbAo+JPrcS%?
zHa8tAJ-vMmX80W!PSS`%MU@h{OI-~%1fm(oc2REuuK|JM=Pm>eF5Z^+HR-#w=O{HO
z1_z!^85HHCkC*Lv_Ac$VNIY8`w`1(+am5S`4+79g+6{vJtzy@bUZHK;pEP=Rfvvs2
z9XZ7VgxmfOpRV0V&9x`-tWT}X`7z8*J>{1^&IQD7m8gWYxh6Na7|i$AU7T`}C_3wy
z2h{y|V05(}5xbq>4MEM9^F^G<@b;qGUoap$r;haj{rD#!=)<QO&Ib|=Fw770Ib2zn
zsHBoz%}bRss}t)XOl^KJhVg1O=X<W)+1bU($fD37uiKiNhuXL0mX!YH!{$1#<b9((
z4^0b8jx{W9J>nH%MnF~+Q7oS77^*2A2`htvZ9O1)2vC9h=M%JQX>Qu){E^v|NyoAj
z-;lQ@;K!``eEHD1XF_he_k1UtJN`K$Z+%@1|7UYjx~=Mh)qfGe_dI(4M04wyaB0VJ
z=!eLf4}IB?x?4c0B-y6xx~14Gi4oUh%2lP;pOaq=ZjTxBXSThv#WhGh3?C{-E)Qw6
zq35oFnSAAn6P;aZAZrWRbw|aLbNfd!Bv-ZMNF>STuM&_)*Fau~AH*%oc8~8hc1W}7
z+lt@tl+qlgU7ht_)?Xa1VwoTWOaK6YTy8rt-KuCH8J(3bjXwyA_RA)CI<rGj{80^?
zP}eVEW>xp!5v*nUMYfFQzNRi(?TmxmD}jP`3KLe3{n_w>0A&<RrPhwZ0SW~2Y~gB}
zRQj~CSeDwBBt<5;MjV8X#rt!KP)<c;b4+BJ%*%d|@V1f+-We9#oIBGn))m~(lG83e
zWiOBltHh*T5#j<);>V9~05LnFkKv0{hTC6$3FTx+q<}z`<WIy}ly*};{mFmMpjmx)
zY@t8$yb@qI9>$*^SIYipoP%`5G!c+x%P*W4CI6f=mckKanl~Q}g>|!Rxf$Wc6@4hD
zL_F8N@N{1^sYE)Ln+roVQrI4qJetBbOCJqx#GPD`#5f^nfeTMd7W0F{F1<&o500gC
zrDpeI*>z@@77rTfFSMnR>A>W9C7R3bm^91n`nQKVEmQvUD{eo*i^CKTswoK;(>3?u
z?R=l1<7~bn54l??jYXzVgD>W_^)HmyvNWPVA_}956@joSU2J&cIli~Q6gAqt6J!de
z)c|=vs&@t>2E7NhrbhKGUNnLxCw0ToORlhuqq{9=j7(#&g-RixRitD;QW<2dbbo&D
zOQ|YT=UIO@K+Y*1=5Vjn*_g#-+#j*<8T;XvE7?Fo+d#sSK4=@tJ=9qm_CtMSz})}I
z1ymwRSW|9roRx;!7{GK)TJd|q@z1a$7D`^-3c?j?I1cr;s*<B#9w%-<G|&lv1e;oO
zg=DQ{CBo?^3dF!OA|P#-SpFU?=4~G#h{KN01|(}v$k>=8Fao}x)uKP)NNSnl6jEOm
zv6#<+f<9kJn#%_UYHvrp2B9E;lv%4v4~>C}nYVc$_U?6}J}@eWBp#eiqdnRB4`{9b
z_bMvmZlLNSrCt<f)=dgYxnnNuD_zV{Q4vxTXY%1itr}II0avMrcAs>dCmohj(*sbj
zWRrYH62)~6jVV3fC_rUrr|r3Q2-ct`L35>pjsrcy+;273bG-je&ai)L^d2@=-_GJ>
zi{H^E;fx?4F%}3fU4+$U&i8kLKEU$hkQniAu-?RH?@!mAzBKz^9;SJ0P5m(cXu5mT
zwdLN=JO|8&md}{`av&TjNcFkm+$sGLnD=>-T4b-m;M>n+RTCCuzo_4e5YZK^2$0Fu
zs0~md4swD>0TRVP0Xq(*?ad@-N-}PyvK0K3;}y%)D-yjMJL08%tLQ|b_!D1uVz6?y
zIgZ=|n8CA}y=ov5esuFeX72tKEgQQRwHAe$8bpn@0RENFuQCI!B2VMe(WTcrts*Mo
zzXp{oC&y+gK0FdpisQ4HdiuH;Of(1zWSs~7OvdbqRIBj3UZ2CHh)Nn>BVLJ8vW)e+
z!6N%(L$y%nZ*2PaPn*B3gHBYU>+{gi<$YEl5je+)s3p{)R&>=NJy2x}Q>;n%ppBf;
zsto!1OAoA60Yw!BeX*`Eaq#ctF=##n<AP&ZprpCa%&XUqLhoDe>@;wQ)X@7d4nLU~
z;tGCJ;DyIb0D(4B_Z2YMQExz_w$WL=aHB3Z^{EyhQ&QnrXPbs{y*(=vLngInRY_Ka
zeJmn?)SlUR*2p5wMQXs;G=RFQrR=qhZhGlET1SFaT~3gR-S|d7L64HvgSj&D@HA={
znA|<m$)*oW)uuUi2QFH=6O|uG2hHMQWyglff45)h85f$B{9U#d@5>D!P_>>Ktw7T|
z;%n_V%3XseXb(0VkpJ;4y!0?IfGt7!sa79Ty=9W~<?^QO@=!EY;Vm)?8F|*u>1<q2
zzM^k5TOj!Bv_ekL2%%zTGdq~}Nl`h)YC2nwFCSG+w%~Es`+ulYx^UddYhKw!-msKg
zfXVIg^6p`%LtXFilu_OR_t8_!g_jYB=y5qXNR|GDv@hK`m+!6ciHeusU&^22Rb?lq
zXvO?hX{|MBc;L}6sP1?KmDr*HJcE5}vEr~k!q6SpV0tn^oV(;R!pR^x8oAtgZqKWq
zW-z~N6AvX4en|Gxyfx<Zgp`Z<KN!3047vd!vZ(Hmpr#hg$<*bwzZZYMI&}^twHA~^
z<F)?nZ!W?%nVa(&u}w6-Rpz?Z94X0xkHw9B^zY-9B7vNk6m}eIj40(59ZGs${ML45
zGVLc~0|VoDM<Zj?2HdYvS<EgQN2_rU2d=3|EkE$>A&b^4!4f}H>f(|Gu8K4;v*zZV
z`!H?2lu3yIXqi*A<}ga6B$5U4ApF$!2XWDZQs0oV8kP9*vN)PPi^o^FeWqQp#&$8f
zCY>dB9>j1p?+|p0pv=LKu%yqLD4N*@BTsWBnHd<gT-t|_d1O(^AByiIWL$vD9Ci>p
zQMHebaCq9u6AYfggIZk4g$^SZqDYRMb>*HZH$}YYmSMs%!%_3lRh->ZJTv^D&jge-
z?(8}v_A*1^k?#a86{-e9_-bAX85#qgKi0c8?As1CGropJ#t4~FS%VZ{0hGVIc(9cc
zm_Mxl7QCA#J3F?CACn$cx3hXqVoYRV&t2uP6Ww{@8RJW*xJh2dstHKRfsOet_SYtI
z*J>%}WKO(193Fety(pRLON~h5RD^f+5N$<z(a(p?d)N7y|B9Wjb7gy=wLRnr>=o!<
zN~M~m&gH`~3a*ITxM2*N%a~gmyi$nbzx6bkVWrat@Uoc?T6x_p1C-09ugkU3p30$<
zN^gx10GL-?$M4^f-HcXcA<Z?gJk_28Rv#h0>~qNUjYgHmMr(pWZ^uIv?6l{VfmfII
z%?<t%UPolx)(3sAWri7(u!tP^#_`DT9+c*do?yFo5nx<AVSj5AS0&T#QV}MGOgOx<
z@O2Rq+>ukBj}+~R<<_B#Zf>MQ+5U11zB+@HGA-TKxO=;{uaL;A;4%0Ra8kLN8OA;r
z5zq)0dVraq_9{=GyHquPJBQa(A0CLDb~pnGGc)@?AWNdHF%ojs=wxS<c&dJ32_~<H
zFEtmQ_?D!ojQ}pKwuQOA)TYu~na6MB9iWl2*<6fg4S%OfMQ0AhzdIo{WO!o7&7>>u
zZG69AXrHr??s~rE{a0c3(+}tI2f3`{<Y!LguFA$~g>mR>hPk?`r`>mUEas$gl*`#_
z?)s5A3rCe+#6c8t5;<){`F}wsB3V(z$Xj#5tHVq$0UN>I(sT~2pfBuVwKjBXb2Y#G
zORQbRdI-3JmcU87(lJoy0lRq?u24qOf0MFi`I4>oDu&(c>UAsbOR(bDCutXNt)_FJ
z5Z-cjg}7)7C?7c{Bla3yt($8oG|M(Xa9R7LFCrMJ-h<Q6b+io1N*UIE?n8l^tHa@O
zo0n?85L}NWz)ilxM~Ce3|G-d&vm}VrXvLUfUhJ+M*k<9Hyyihc^1dQi63R{gyiSZ!
z@+vskV|E{XI#9}u81~Wmo7~6}6GvT}&)+6CRg-kVpnTBdz{NVWLvG7cX2y=P<V9pg
zoMK7fq0>~*YzqXWvoS<-0ZielIz&(_qGT5e;lh#J@Y1p&+-e;Z=52xD(cKz*=w^YM
z0B|^Xsmf1hJ6xJ1>Y~Q~vF|VQGz5&Ss;Tr(3_(5>xld6^pp2{&jvXK~JE>`h!pt&}
z#nOj(NkL5u9x%D*fULuPF{DNnmql^M+RH*anUAJ<tL9Tev1>aUCsLH7o(AIVfT_4L
zHRmBu2*|W!cSMD;>rNz7vSC75^Zi<Xa-r%Edj$Cbz+7rWU;AMC+`3MApWqgTce-N+
zZW9wdbf{*_C{m_DH!&R=*{vN}RZ3y!p2zB*UzE{{57ZpYvT>ATq&M+GIJ7S{$ba>h
z7^umRD!5*D?+rO(ASMsnsV@Fh_$wG4$MqRI&g|Fn#sJXP^bm_~d6I4kRtG>B^Wf+d
z(2fVfpni`<Z6P)!$n^LAf#o5-4t@I;xMAFa?B}0hWWAvF>MfmPSq>%UAp|(2H6c3P
z%&DgHgHxxM<!RL#2(E*ly=QTsqWvDX-|9@yRkLhk;$AVdA+RaMU_R4el+D#YZh(}#
zp)Xp492PF0F!^$sFc6rI>I7>bc4rlKmxnFlGn8yxqJ*$+%$1v#C2Fx3h6yf{GuPU>
z8kb9=y;I>3#S-RAecWQ;Y?M!5%Y9fWHbj~X?1E5;U4QgN5>iW&S7`M+1Bt+tU<6Fb
z&~$-lkb~gi*|?Q50Bw3SPau+yWt40d!4aL0b$q^_S!n@5;gSQBlz#q&x{&2FDIJn>
zIOkr5)L-F5zyM7rI)LT#GEhPJCJmUmGwJaSlb7}Dg|A6DRM~F0vWF%*)Wmm54uB%A
zO}Yq-!?i2&+szeQEJWp3bm<qT3#Q9lEk1{x_1Hq8-98w%&kh2h{_P%q2wiDbK`pl}
z57kM_;QJSWnbp-hB)Zt6(OHBG>Tg(VPr$Y3Q>PBbLtKX)PZ%}k>)}Wg9O(6gUYq88
zo!C0Nbjd&YPgQ~Agj>s^(z93nrc$<@@v!+PO9b8f@NjjQu>4(_+O<d)TnZY5zST!`
zQ<awM=D9g4tpE9i7}6`n#plcBXDK;usWl`8$)J~}^hRpUD^<YzC4bz0|62xIj1Xpy
zw^VBnviyF<&C|Bo)C1Dnp(jslcCftu%Z%vKW19AMkPSt2$;6`OV#U|D2EMZk%9Mx$
zt@`F0Wny_%(U><d#i4|TU_9+fion%CG`~t;S}hPo9jW8X8q)GUJxQLHB8n#J!gXAD
zqi~JZMFn_Gj=$=tWs`-&yGR3+WqRbt(_+=#bH)|28Nj@-V5{U*BBK2^6GH|!$@87D
z_D=WYcxT+#a7{C9=|ofx+j$iLn)(tDmO!}!#=9xSw5nkRkz=-e;+z{g*@rqPF2E@m
z%PGQFmVhOaKtR^G?(#nc4?yw=Qo7p_C+z}HO1Ht&o0+z{vR|{&d|`{kKbXeW=xoCr
z>inhcIT@4$9!e7ayDI-IMq(WAQ|GN<e~Y2hslKjKtMTti`9fCr^xk-hboBT2YwB_h
zFhcMg$SRDzpa9y6oU<a}mv`yufd=#|Y8i_Zg5yBRE)sP4Wi*ktZyPpO;j9`64w<}Z
zNMGvUMo|un>^dPDZ5;o*TvxZ&(7ddb9jt4~7YEnx4M#Yt()@XPHQZ)iGp$kBPhoAx
z$<0m(q2p$D!yElKP*gj+V(dO{Kt*mw%vK<3QjKv70h>}7rJBO)<oV7i*Uhi8xP~KN
z4p(;iK+DsYo{5(}A{`EzvASH4*?hjjj4?{O1_xpjkH)0HY(PdPb=}~p6=+Dno?|^s
z;xEZi$>DNn6{i_$<$%2kSG+jmzIxMd=$66yOAU7;*<zS?&~uOZiA+Ur^S?{8zwWh-
z-NI%Xr<L&bkXbdnCT9wX_c|2Tkv5;~_|E)fnTjozZS{8}E^X{#vzjyBFtR+`c>%J4
zvnbJt3;8MTnpYSdsLeH}>c94n9#qiQ_1bvTrI9^z%Vzo07c)8EV9k%*E?*Ye(kEpT
zdsm2*ZXJ$Xz`IncNT#T9QXluKjJvmS8|!}RSx^t24nSrz+cg{Mrxzrvh^`DsQFoyZ
zgC}%JG<DlXCZ$vVT5O#>F+@kvX&Vt3g7R{D0hG8x-s!_UDRpDJWIT=Ri(?E@Kq<sW
zvksv};8Q7AzR@UdN!!m{J26iG#&N>WK9Z*sB5CK7I#`JB6EEZ0=D9ZYN3*J|Y^c~?
zXp2Uj&dN|Own2|#4gn}g0|*!7{k%a&T;e~EYiv+F^BquJ`aC0-+b8i?G_}?0q%?7&
zLTg#TRT70Obah&a-28KKESGXi$Nz4EYADB!8VBOG5}D!G6GJ^+>uhdP!SL)HdytmA
z8P>5ksYZUYGH6Sl6uW@`U(k?Bm3QxW?p?qw<*sk0yA4MAU@QMnj@p!^y(gXBbnIie
zhLL{m9f#{V*-+p>C4P{*Kkv_#fF4mMsqf<9T0*xJazSh{8!qu=a6G!7msd*plxCWs
zIsBh-`cS%DU~ZX<+J5S`as81Nu*;j^=Kfq~UcT!++2}3FnLQItQTL~(yEPjZX=JXF
zzAv!ii_3S4ja?)rP2Evz_wI`mc;2Y{?y+!zG445Nk<KY4`Tv+A#2I=LFj{>6`cwx!
zLI1#v=<SrSL|K|p{GF<U#x;i-+I9ei!U@?Q?djB|>fB;I2FHj3jN2c;dakOlpjVs1
zdq6Z!c3d(E8%UmSuWBG`r2j4z0AGmI7vrZ=g&=9AyDfzoOc9OBl87AifxK`hh&6Bi
z^LT96#P~#3ck(LBqp=}H=DJ%p#rr-XBR!MZK>#!AEzFr;)sEkRCnJI&>4dBMlj}i8
zWrhmp72+22fmMD-dx!ZBdZIkY=Mq5v8n#U)Yn@s5uHv~!ApD>~5r28<Bu?fTbf~}i
znt+1yBQ^a7QMM==fBxft$d8QM(*kyWS%&?)TTTBHPZ_%TNsdn&J(f6qihp!5#N|ag
zn*Al7NQoG*lHm*vq<8g7Na@0=54px+WLOv#NwDteo{|Z4C3SZ#^t`K|C@)v%pKN-Z
zjqEZ)jPUfKH}%fHMwE<gRBzoCD38Q#!&rk+{qYk$!BRqd)yn1PdAnaq{*Bc`d3zaz
z8STYeebH%V06O`YOb1G21MXF?S3B6s#*^_Dp~@r5hIMsAGEA|z1ChL?JN<TY-mOG|
zEMvWuki%Sk)kJ}Re)fgt2hv|4Sj7#G#)HGzEeM4>B>?;YT$fbSEy5#e>VA-h;H-3E
z^k#dTU%a;GE2u5}g_qM^+y99SHxr<!*`Fkw#17O$+)S}UbHsb5q5(t#La)%b=&)XA
z#cZko?`DZvICb~1>G_`JvS&RqAhCetBQsPf<g^A(rgN-5sVO6B2<<3a27O3cZx#Bp
z^6rq1+Py+m*|&#l$|mzC9M_pEIa+ggN1T8{K?xRmJi&g%*)VTU(2uKH)GKLIEW01A
z$|~r?X=DgZ!u9&PudsvJ41)VdaI5)8{X9~>G>fuu`2TNN;z<$bpO&<$SZ!qUU7qbZ
zmi~{V9Mr7<;>=U2z+g#fgEzBNK^ZdHVfC_c7ip7*W1dTl(!uzEv>y<tkj{|KIsts^
zSrRl2T0`>9BtIW;eG9zkk1$v7`Imk`XpB!hwDanYnjq@Mp)cM{tF7~lB6AaG(Nn9f
zaL~?Y6UuAxU!#M2WZsmrT=(5dZd)O%h<k0X!Oqasc+6QN>Yc0otX7AdW7C7m&U&5M
zow#$e&O|{#181fkktSaMoHq=9Q#CCcDx+318hk;*7eyh{^SFXNi$GqEAfp{(D>B<X
z8NYcP#TC}dU7nl2dqt+8dj_LnM<rzYpYag%IoI=%2JaqHj+uHU^qQ2+nxi&)v2-VU
z;q9Ch4-XyW)Yk*R!igM}0DZGE%DLeMvJQQq2M!5SLkQnU@{4h8&fEJ=|C^Im5I-wl
zl>yV)cjxi+!<(4;xZC0Z0WUE+0a5m?HizZLo%x5GFNt?&W%+Vz0W68u9R1>_q2fO%
zRyDRZKu8O4XtUXU7CuKVuOIUW`5zuVjuHrCKw53BlZcrec(Xw8c3;FY`NUyd0j{cI
zkyZ#JB6t)E#)5wIQ2@y5_XVWA`h<g(wgKtcuUUL8Er@oL2-aBz(!4{@Z!X{V7)xW_
z?+uW_g4E@;cL3TKz4>y)b3#7FTzK!S<@XL@aRn)io=dO$x!n$??uJxwyjRYjwUe_(
z+$53tq*#$86|eJ_A$@UphMX54e2?-DhBZ@U4H?m<nPS%X!VK8UD&pO5_sx|&)>3@g
zo2J(7r<1-;{Z=sb!9H;-U_bHE5biu@2qa8Nt+ow6ShktXtV3|O7dge=U08ND`uf^j
z*Q5(EJySpQUtSHIV+NgvnO6V$nc8!+SInC?^u*juZ;x!P%*v_!$a7KbsR-kw&GBbv
zyiMPn4^hw?z8p0d_zyH0iTwCzZBy98S$t0RxOEniU2XO<yf!rv_QhetJNZyF>*3ry
z-p2IQmK?i)2(vN3*46s?zU7{d!<2Id$D;-`brC*uS=Obc@>S^I{NBF4;%zN$#VijL
zbp3}nL1+e9{^9pNLgJYUnZEZlbea0?&FYw{_bup^S)%WQ@mx<nKSBXo^<_Bdy3A(C
zO6~Ow@-%*nU4Thg*QJ_nJ+$KY$7yHXIPp_0Rn#B#&~uuzH6=p{C&OQ_KE9g;oi5jj
z2}z>2J%yJ$rNf_1L%6TPkb`kHTz2oXn#hsVc=eoQX&dDUV0QK%^1Xyd>Nx_FVHfJB
z!qHFN>Stk<yId2*2R_#%8&Y%If^6@@$ZugIivM0CdTxpgy-wF8WcbZmh-)AWNE~Ud
zk?ZEvqMj(?Z<r%qO$ochWDy9DdZcmrV)#zoA8uq<P}aLIWEP8pkQ{BTI*a5za^+F0
zCLn7yx#?algs2e{vrv=pDwHq={$pwguO0)=Fv#S{#8Cxn1@Zf2WwjWe{Oq<JqHU})
z*-$dp$|HjbA{ZzsWv5JrfPD5uKm$n30H%B+>XAnP0D}OLg$5tJ;#YMpUIS)RcYR@{
z3BsrFXBJB$2o^h7W_sHdVe4o5f_FsnnK690Bs!R^b}^}|+*;s^9$C|Tb?(mzi%SAB
zLr0g>;@#JsFZIk;LNK5?OJlp>Go}63xBn{P2$mWTi&&?lwVu{k;uKjJYw<KYCnX%p
z_^o?6{eM7#hWX&h+%bR6X5PV)+Ust8)LA4)P`04vOmM+|KDS*t)Qdq=d-+nWA(#o)
zwZJs=Y<KT>!U+WZF9KXXd$!jVI!T?*l>;1Pfe?rn6(-cHWO5cG6-Uk@i}^n+;0w|R
z{E%41Zv-#^;J;eXAnRYWwk>1k>;Lltn2DcYZFcD<N1_&k@T||#Tm*i&gG^NT2xJj(
z9iA@9E5$GGo9mLP^YPE#{7lJWgI7$aFY94)De35;TM)b1E%Jvm#3?rkwUH&3)-38_
z$K)3qSV4h#2C*RU@#YFd^dU_eTS%jC7t%E8r~#P}`f%MNv1Y-&xS>A?!IT_@dBvg+
zUM<EZa|A`xm`dAeMx-X?Vmkl~<n&SFj1vtJ()m|6KveF9Ft5Q!);(nt5YZZMIihr4
zO^sT&!zfLq`zRK>Y*aG1uzW{*ORRaW8^2G+<)AOu!(4lMC1I<)Co^h;OYv<1GfojE
z9z*0x7Os!YaAYxmJ*iDk)0!s+8C_G-l}-u$$Ll*z8Xe*DrG#^UNWsekq>1a6%nc1&
z%V31R^4<3;O*3_Mh^%SKK!_F1GGI3|Qd=Tc_BPU_D36s}kDyOcg87$4zK3eAZzfig
zQe2SfL$ijxrDGlykr28J?>HW|Z;oJWiPrTQ9%3UTL#5%iTg7tnmi^nm_IvmF%i!})
zPMZI(S1%3Rxv|%4QlRX`5&b=m|IPh#rzqn+Y+Z6c84{ceD~D)WmJ;S}rGBHNID&sK
znDQ}p+}XV_UKg2WaU;=y<u%CV2z!=V6CQv72{ori5*I}QWAKAi^r`F00Wr_}%Jr;$
z@V1ATLc&bs+Z1phfJv?Z^7S0GlqEvNNzH7cGXbA;c`18nJKF-alt}n&;y0R5ApHiu
zv!AbAHXGTzIqq!8att^vnNOq*CcA3w!>h2C&Yt~57^G`I52i^iLtOZ|xva9(ZjCny
zWd7;!NqjHD27|HBCe2BY8aJEG{bv269Yhe1FqoF$$flN24G4k#dsfWvBt%NV*HtCu
z+<5$s>lRIG@)|kr5Jp6<`E-a}I5FB@1C<R19OTPH`0_{VF(93%kXb>>?R`YVQ0B+4
z3g^sR<F>))C;9h*u*y2^V+vV9a}7WqX|~rDUwNZRwCXdXYrVIU;BDB#j~$(g&<QAa
ztGW8tH*^d2C=_eQ<ptvGpq2F#f!Gx-py`b{`Ay*^7s)o?uxLUHCQpLa@r+bUR^!eK
zm!s-+ExtLueZdlxKXV}W*S!Kvq+sm+J~J=2wb|L-uFfN7I?H!|-JyGae&k-w@O9_c
zb#Y&Q4~;4aCKq6)1`LGI!d_c&=E&uUyo$DrA;G&h0SP)gUsV<RI+|GkzCM81PMciW
zk~vC|G?bi2yHS+tOC74FW*@f*>3flRX`l`;dNl>#eHIrSK?3psG#2{KC62KIqT2B%
zz8fxptcYEBI-%PqFkK-#ax8OQVAcwcd1<LH(2PzE$YPg}3Y|kD23q}Iy_=8ynlOW3
z0SrSzk>+^c`t8LW52N=zfIb!2_Axzy{sg`<G!w5dW&tm0R#^!uT*?h2S32@wGDcU-
z661PiiA;^4SQepNpW2CZaP>aIK?YIX;|4ko7Rr8+u3U5%=(z@qvLp`INhfKo^4c8d
zsXx%bS}T!kt8Xg3J74TKjSR)->dq3Zd$hKyy77KYm-&rFaods7c|J)#|6a?<CdnQi
zt*nWg69aKVDO;h#GHyr!&i112@ufAGyOg;VpZ~(N6OvYE^F8{H<8P;K>?h0nQ`>**
zHqLLlZZ|n8_QaZ7dOQCWuV1)e=ll8lG`A+?8==Fh%<pes{C~}s2kh>syG(sM1FhKz
z52E<~{4`neVnCd|aBB2JhZS*$vh2~-*ba77ne%R5@`f`Hb%qOX#_U&uaqh22xTl3Q
z*g5WMe*Pa84s+B8>ACjSobOP!!lv);8+Mw|4_R={-6kMQzEW;HG#$ZrD8(hHG-^0d
z$t%pyeUf&hDm^^uM5X6NRg)^`)e}-3P}bas2-T16)K~p0f;1G=Xx6|E0P`gKv`{ZY
z7rfP~DLN1@BR}*Vmqx_cdnh+tnYTF#DJhuAjWnSEC%^|MY8)<FGvOxpHEQ$_<#8$y
z<a|_+wfdfu9nqf*K*#e|#>XIJ?d_@e7oCaF#6MdnL~LF*%i#OyAkiALeZq-R#xtiL
zGW4UNr$;ueAGYLYpr)<-8qE@W)^axQKbGP$dJ2tb8>QQVvI3|>?*UD&gWld`AfEIG
zK3X--zERggqYoZ=2pt5Dqv}APdD6df2YX+Py{IBd0?bRc{9~ccba7EsjDgG|HJiT6
z=UqOK2|mVHE^@U0*wYqy>&t$vOg)b3T`RAd4v@azw3*=Bd)hsfkp5I%f4`Sg)jvOX
ziNg4cXNgIAgHFE>8(;xph5GULIl6pil>&?=)bmpENFh_Cp>3>`za_d}o4fq~cbU4B
zw$9%Gw-=Z=IR6E@Zjm6$C*}Jl^iDJLwEYZovaNt{T(6q%9KH5?Rer+_MHk*%sm0XR
zw+^hBE{NW7^#p_}9_2Qu<XEb&k(N3a9v6X^5o2=vZ`x_<;gL@>ce6dI6{kXI>^_|p
z^3%E?%=YcGI^tJ&(o)0Y^gmOtqD<)u)nGt*5f%mSLi|BFiOV3H!Z{j8h>aQ`R@DI*
zuZmi@ngQ?A7dLznZ%+>xe8&B8r)xLPdM8F3{C=LX0sPF0TwAF&n4|UQWW7ip<Dd)R
zsj5A679_;}&U*WVExS$9f6;3#e6_yb@y3Bi>|Hh6<XzV{DkfB$BW|?YM5QXR*Q3?#
z7!AR9586?|B4U*gdy(>YD=cv>Ii4VKM`gJHj>^;p{=tRu^Oa7o09(`jmJ+f{GT~3G
zsH!G-d9j+B)XE_+u|BDHa!^H}NQr8Sw58RZf`zHL{7+W(m{wM$pNMfueoCZtJBVs+
zr78RZ=u2otzu7*Ko(qm(!h)%4KxWXBEK<sI%5l?9r(Wmf7C7QD^~0D1p>d3<LHuA0
zGQr8~<l)Q&8kPj>wB7$*uvjdGg2Ae;z7#6O2a89c`9ipE(oNr#99rLh8;z74&(5t8
znZn@{Mu3EqQo)fS;IOexAvVDY3kbaP)(aB0;z1kPw+5mLGJB)m5%tpOfD+rNxrUx;
zaR;!r!7|?Sc+v&*Zsr`4GIQj^;BU3rLA!UqKz7@=vi%;OuPH~G_-6Eipe^_SoNaKJ
zm#VYH@`O;VUU%Lq`}5z-<WDq`s5W(z$V%FhsDI@*oIxq|bDJ3?4bKz0q1k;ofkHtA
z={6Mie1|vw>?7Ci^EZ_(s$d|N&_5b>M?&|OAV4Y#PEkHGRQ4-U_Im)NT3KdT;Qrz_
z_mlRd%(+NAOlEIW%OI3vBitHzyh}zG=LGHzIa}Z)wDBHp4MWy+-_>hvz8XS3!ze_P
zVWB-==gx-gbS3Lv7U489-=QGh0ES7zu9S@*_mQ#8OBt!?-d#tG#6<~9Z|_LaB62v@
zBCFh@@3f*zMa$BvJ)tj0Cn3n+G#UunjZHX>d{rv-%n`5iFd!x^oq8|5DS{cfDfW7G
z4BJ0l-xu~|Vk9_rnVffWe<h(ZP)Op>=RU3a2$h5tsH^_o*V=L-*-)6SRfC}Rh#uMb
zQj@6lc}x~c8&j=i#_}*KZs>YyHW;HihRC17AS%9T*Qcpmxg*b%O;oYlpQ8}%EdqB%
z>H-=bz<ZbN*LXfAkIYLjatI~>-wsm95TZMJrfO$eN~0E?hY3J_rcY|G>^`YoNUfY`
ze}a<7;XjbUiSDIXs`tsO6{7kA=ts-b7H*h#b7V;w?6RTdtQlMV_6$GUi9_FGrLH;&
zeR0nE^}WheO_F#9qJtVom?pPsmoQl~y3wC?D`k*#Jp543TM!W$g$}28B@%joQbd1v
zs_l1hvm3G&t{b}`{M!_Gc#H|;M?j8d7H90pCE<8X4VcwO<#!@i<2o(devO6N9Yak&
zB|HzU_CNc5bsE~|LKM!bVrKyL!xw+E^hcSEcRJrBWB=;#kC`i4jQCZhHRvnEtj1nQ
zP~A}OKiHdoeBjYJ)oMnlW^>h@<+GPeNcEQ6^4p+aHzAwSf}px)JhxyKFC=@K=Zbrt
zfNMT8J<<vY@mTO2VO0BVS^Cf9?P?(56ys_9FVUhk>gf&l%}?eh;^W4@a*eqxT;mvO
z(`xLE6CrRg`~op7M+O;7=<EO9vQs0nALE~a5RBsWofjLNeL`ypGU63?BWEv_2ppN0
zx8WMm{E5DPGLn*-xD<kL;ZlP~@eb+W5Bwlrp9@LUkoQH;Y^j0>q(@g<S-G?Vg<){&
z%cxkQF5Xi0K5pDUYsL|XY|8JP)xp4`&3R@}r8+MhgFToI?i0%NVa&zhejf!pJ~%m9
z5<lPaVjR@IjVuc@s%yGw7iexE+$j?Nl$NLZ<+z8w$p}n6O+e_W^LC9HN{D7Yt|}L#
zjF2eEYZ2R9*U>?(ZNfY86)QRm&j-r+V(vCE_g%+1_h$|ZJD(vJQ{*wHg|sCy#KXRr
zoqql|Y<A?4N$J|#QH0BA4Um6|l)nb>*5>(2sFaWiY<LSSqs1k4#fV!xmt9x6s!jO`
zl4p3O=9dME<jq?po-0>9QlD~zthsA57_1AT%qhgfGb|?=yrp@E3!jO<^(!b>It6KU
zRX$Z$lhQ#!+HfCMa)sxV=J0VpunO|y#yRqb{>3UqG)t0ya@~(<ytB=&_ej4hp^W1H
zOwTVbpXcJcQm2iX6~wIv>+I;~HddSzV;9bEciRL#q^m6Ai5NN0Bp%<BY$^GV&|Fm7
z6<8PPY%nsw%2bY+Z*dc>Dirc8ScW3LU)XN!v|H~cEwxh=-lCEedY6kuu`+MV6-;uX
ze;KmtZ#jXSQC$1<<liJ!Xct>xkqjw>s8}m9y(Af6+>=k~d{lr?P}bEKsWmbYqSjCl
zGS?_14XRGcFC?F#?be@vamd;vWzq+?jXwP|$1V*;tPj^8B^>a1l%=*W(b3T<&uiGJ
zW}lvdivETobXig5Pk6_rslInPlyoUwi8C+`;l}l^j3T07aqxaSA}BQ4ec@j?D2st-
zDsi?VW(VyR=;2B;sa%MhO<#nan*5r1HZfcWZvZ862T!j>f(b)7zLS}w_n{m><$pcD
z{J#lY&G3Quu@~id9a2ew_6I)c_fDSzAsDecAP-h>lGHscKowr2qo(=N%G$MyEcxS~
zw%bUkGgDlY28#`K9U2wg8_n(iH?<kF{+C>lkd*ARST>$~tFxS$hiQYGc>LqpkAN*{
z(}Rod!%+%l{D1-RVLj1|35I$#2T`BmI9R;9g~YhIBNflG%gjVDm_v})U8Xj!?f=Q8
zMK8M^Rgir(YR)e&dXHK}Ho_`4DNV6Ay<BheoP#^Rei~45o2Y4JBou!il?7XV;%g+|
zE5Z>RL$TPZK2Jxq@sHLNNFDjZnM8rSl>e{A*8<3MokqYt(nq!Rz|4+z53btRz0^Mn
z^9E)uGTy>YGV-23k8^Xv`$)KLglrmfG(iEgjxSn_S0_KtUZM2$Raljk=sz=42oYEm
z_0on>Ny%@~iy~2auiqMcq|0%<5Ti#Qkw_T`r2hIRzpST3j;3B5mQsWckEMeN+>c;w
zXn6<9!FGTlDXE?1B#Nw5Rk%xG5uUDzc@#aRPC{f~av&f@*X6Ie593rqx8D(cpr8C2
z6C6KR6(Y6tzVa>Hhdt_U#=}dy@JZ(0Y}E+h7jlGsXHxn_O!o8>Ka^DROdvi^m>B_|
zoc+xd+^jYpz+HM<ZT{6e@n85am|cz9{NF;83@;fvEEKYviYDg5+PvU!FL$ZVm$;KZ
zd>tmK1w~oE@}w<GN2bXl>qn>&QDFcyrP7M_O{C8!-|7>V1n3}WYr1ki{GoJ7q^5hQ
z^M_Vp%H_(|*XKiih!mgP!BCUVkc7Glam2loe88GCcHiiBCD%rB2cq<=II$CY2Wy~;
zjmIIhf@C;L*PEZx_`T{%%@JAlU426kw~xoknsJT+fDUy}6<!`M8ROMef+X9CX)Of5
zh?$EQ@kPjVO{du0WnHWC@o`=FsVcR9VBD)+tf%3fF$*tuUikkt9J}UR@y{@#Q%VPc
zE4Fr_k>9YAznGURUE7On7b<k1jIeh&1g`R<1}X+Xv@dlz9_y0AzGyrHii(KCrEsGD
zRn`#O-OVTdA5Lz{C9LoKhRIsq$VDgD$MOf-2R=bsBQ5fiGuVcs9%7h^540_hZIo`(
zfus9GiM8qN0<8rn;Qu8uET-!CUY(rlEc{guj@X#OurHdn?bS!&r^$LUh*juJG}zR7
zJfw46N*f^#3ob|=McqMm8@;GH%zDz+fZUridz)+slz5b16+*+H#Eutb3k_C=_Wx5-
zI`@_ww^_=Xck3~fO(;K!rAkd+YpmWv<k}hCeE6SC7$q(#8Jc=f*6d-74P2{nJ>RU*
zZVPtaeJc2yrn{Fb|6jq%ZqmEuIO;s^*6warXmFbL;0SrIW@lvB-PF<@y<&z23!tK3
z3g>Q%5^!rKWt(`tX+Z#H3t!1w?zLTOD)A}%|2=xX-$u=-DV56_^*-b<#?bYQcQy7c
zX3G;k2B10&{>+{iRwpZ93*%1^e6zCiAiz1LXHD=U^2LaPdn4yOH~hracFxTAatW(x
zL8>&qZB0$@J`tvScrWgKGB$lH?sL^FxJb{>%Z@nHqk8x3r4Z8Q%*w%(TX(4xdekpa
z-r&-+$>SbTfEhwzCp5`6iKc3Yw7?S*5s6GKAE-h~1jk-px(~e49L`OHN6Vh3*Hv4i
zEhL^Min(#Q{siO`=NgGvttLrpZ8r79+L{y0(iXC-!2y`K%*W5*j@M}A^GP4*5{8<<
z*xVMHM3cG}4-UE@2RJ8y!j&~#^!%dD8ju`J2@sr*_3$krfw>w(XE|t``6Q&>jlRQ+
zD*_}GKX8umE_E_4xmReF@W6g-sxNZJa@mS;<TAs3zm#ZU;`=aO|9L>-<T#0#O}j3^
zVVmPgXI>Peb#VQ!Y8646H*nK`)oxuGH;;N0fZ6*P{aC1;3Unw_O4jjU+<-x1l~=-R
z#gC^o+%9%iWk)dmaD*QcKbnH|b}%sC6Gu-OM1Vw=?UuH`)MAV4`!K`|_0K&4w``(@
z5O*UCQ<KuU?`S&++O0VA$saU>7w@_sP6kRdI)=)ocs3ia;L^RBT-oD-)z)?on*Y97
z-|zOU-^#yc4nGr*%}c!5DH!K3MkEfMc}bl7N)X*$6RNg_WiD~?P@gb?AY7v-$w;|*
zq&<1kSd_L<pF7KvMdzhWpr@r4MBzV`KM2yUH?-pw&VutF-4+esud!<PY^a~`X!K5V
zV&_JffM$ehMCx;sk>AVrbNL&3#`R62mAN8G!g8X8f(1C<?=O^(<4x14Ix=7c#_u+V
zrS}abmLyP~yBOW{eWntQCQ9L>I%Tp@nINdV4T9kp^o`8)VFW`O0Q|d<ck6D|@jUVd
zg}^po{02}2fI3-#Vvbsf^bFwow09MF+n-KX3{UQ%Ae7nQ1`a=oJwV$=v}yISf*t3Z
z%92J;YFQ<7wC7v0#BiQpnAsFyhVOKY4WylLR=MRWM|Ko`X$)Ck2~mIWsgeIIdeBr*
zjcp%h7W4fW3`hWh>5fXJfyr!})vqwq#E*pAJA7Y$qKru>TPRN?2`&^U0p%O{T5{S+
z-GNPMh_+4@RIfqle%|;@b<N>#qtnBF^-%Il{nqb7wGGW#Us-Cp%AA`f=YGw4>4Q4N
z21#`T-`sxdec+4IgZ4n-x!ycnAAKLHG!}jZKL5@hFMp=ct823bxvS%0Q)SE)76{>o
z-lm!B9Fl`*W_lD^D|}$=Ctn~P!?WG`>Kaz|$tg=<MYXr~P8G+6fP#8h;NnB7QXsWD
zU*oO#T5!qLV8K7(Z!>}bI%wQN0vinA9oA`uo|nQfjg0?6i;M7}zR_d!oF7X}O<APJ
zBa$ERJo;z*1i6&tv6<QHl-eFrmQy4P)Q43z&(&a9FGtmAoy7mW(ufO<w9Q9tI2n4z
zU0OZ@ShwSXbM6E$cWp}+-Wr9`d~7H4-}_Ptqu#*C0Q_l;R6d3_-O_KrH&3^zm`<^L
z!TMkVc5SU`gARdeUY=GtNhA0~IBmmlF_4F<LkfCkBWoPq%@D4L_0HEpA$>D+it?HF
z*yQQ(U-Wx8@Mp0igec5`!h(`;00cGsS)9<IlDFTJ#!BP$EL{DF{&fOpj?NJWO{s(A
z6*)2NJL)q!{q%5d6;_crFh&?IRMF+cbeIYC`CFVe&%LCJ`ttt2QjEmVUcx18YQ%b^
zdIW%6>yP3!BN)b;-24Uu-H+)%2aS+8$`x)30_l!xPR<-V`D=gunO0`~tU(JC;tv7;
zFt3`8Z0s9TMX)dQV3v$APqQv|(sJ?bAu+nj#_Q$u%WMuPbopk@ucg_H=K6e`?1qEz
zl`iq4iV|$R_t35yu4)=7Y_YYLe;gct5;O}AFkzh36XvxLuXG}ga*r1@te1z-b0rTM
zoYFf6?3^`@c<T2Sz=*92mn}rk-s&MR&K%!x*KNA0daUd9e%PWfuRJa9e<k#>xeLLt
zjP9Ck+=odmA+$Skn?DK-XHtSv)8mt$l^yg*>89G-&2M*F;F$_Z;+L-f0jWQxjJ=lx
z+mu*uq8%R2@AaX+>?l3oBzlcvFg7B10lu>_t7(JfDcp^f!wv>H3>kR}B9z!9iE`5}
z6&Db=;2rV%;Pp7HS+HM|0NIPyUg!#^g%p{U*-6+59kp0oQkF8E6NI`uVB(Az8Ts<u
z2eiJiu6aGg7r^ZH<Szimq+wE<R*AY`BsN8JQ#a?BEqqZp3kr~B(N*JI3`H4v*7Zu&
z@uQ{j`Y668vt*8}2i#pzcP3rQqz;ft>2^!%FTPibHD#;_LT>4sR42!jbcr2h>FNqG
zdjMq60U}xFbVOR_e%;wiKs=aDyaKoT8)Am(a>d3imHz>#<i=I=PVb^H-{8;dO~Cri
z|JYlfLXw6M*tvUahTg82!Jc4KH4|>6Yi=I)-xR9$YGb5mr@wN0rk-<i>mKcE?R^OH
zH<CrJzmh6M(8ndUwT+w`P=fabXCcjw^dU)XgE^&fGyiGKo~4=Dvlbt!kqMQKn?$#M
zZlVy<&k3SLGCJ|*l*M-s8km@3IrSVPJ+tVeOq&^__x$Nb_H0{7MvEJ3(|`ILy(31-
z2%kHg>951)<X+zQ+I4`ev|So;o<~7J7+G$eyu)3PHq<5mfdyna-n}3?hQvp@78fO?
zG%~H|C&K<|-^jdfU!gC-+n6|11_tZkoRh)O#o#WM`ie^;eEpven`?`t-!W|sNU0HQ
z#g(J}*m+H&pH+@~q%CaX9ECVzW4#br|NY*>bgMrfFF62@yzBlFtisp;-~%uc6558<
zd=-(h_WS|lxX=4F3rQOv;v+0?WKQ|#0H=v^(S89`%q9GKGg3_mKo99^S@AOV&0K~X
z{6Q3Z3!`U*`OOpBDo-2_zSpRTpyiV7y_N~pRVwKy=|H)wJZ}pC%1yq?8^o{Yg?WWA
z4$A$pR7SQ{3t|2L0sBA%zi5i!d&IOtD4TXwRx3e6V9;(00CJg9qS`vIcK{<x%A){A
zj53$5agK&3X@793rSW&|*1ot@n9`aOs5#2O$h_(|twXJ}kcHPsO5cud?cOKldmNq*
zVm~?#bToc-TeJVTgCXtc{rHBYljWG0;C~y@v0yYA#5qE{I!E7QN-N2HUNmm;J3k7<
z0c}aS=OHM}7&KsHcA;|Me0@gzS6;wKQvHkP<4b{yn=-FCzc7$9(D9|xWY?UvZ79IV
zHP5`QXC`9YH$FZykwxmoUo4!XPi@}w4Advs)dqcTFu+Kjbx+I7=zRUPPYizkr(YYe
z*|Ko<-1*>-|Kcx$jm^+LxjnaQCyTkI>G##z)9-S!<UdRX*ckIDJ)*y}w4>8Fd6$tz
zL$1?0b{oIlnE$!gy(W0dEjOoT|DN}MAb9NYGl_8w6r~UMD~rLB%z0Qy8dIdHwKPWJ
z{4urL@Vx4!FA1LWtlO-Z*z2GEm%k1k_~-A%z8l0t)T8q<O_*g(No$jRUukjjdS~My
zulzdAC-3jM)AGkLeodydsxmcIZOFV-%S0An1WwZd7XT~FzXrbn2%JN800J~F#J-fJ
zO@YZQXMRf?R)7ou5;P@n)Q$p2Lz*CHY>*bP0^cA!BQ7iuKnsC7IBg3havCNrAO!PV
z_!R-DMN@@xT|oOp;0oF))L};2DZF=DA<jX((g-M>`v2K`>o7TvEbTX%QEHJzmTi$`
z*|N;G%*;H7F*A(^cfQ?c_j~R>_x^GB*>CT{kYUWs3}d#(C^Ivo$g;4RyYKsp>L^8J
zWfpaHceT{VR%cdZP)0^Zo;dFTS%8#{s<!UD=nvK$xX0?2OD(KIVr{BgS<v)!%W6gc
zS<4Xuw+mz~(I%@$LF%c8pzR#t-zw^AOnZa;UXK#!BVdU8Wy}DIEMU`TWl~<N>*`~%
z%H|`+h+E%>`km{03Mmh~mto*}u+Xwq^=8WWx`gE6zCTE~nKVC|8|wvPC*FUWBBr5X
zjbi6i7Bnz&?BR!+t><o$@>XuV^KSFmq8}1?h?E7qi>bbGd_TLJ@)KjkYjWgtJsSbX
zuXgGiQzh4lg+hMt^?dO+z9xUFcgKy%2ryD8v!U?&;m^@b_rs0ixkic43M1wqz(}`v
zI4ni57$F`WuC#EmMLGDiLZkj#JQUjf47dIu6C6`Fd|YYy?@ghOHGPDf;(*(O6l#C@
zwhaj|GIxia`#mP>)5a%Ch{cA!rn=fp-((XrWzs}bt@9dLJD^3?sgFWSORHJ6Mh|h>
zGPC5zABAf~TR&tn;DyZ41pA-uf&eHoXV%PMHRid6FPoR&cq=J=s1l}5nQWhNRQo*R
zMN5}TDC8tO;W@{T9c#vojdl~@N-dWFyFnYkTxclgODe`ZmwTbiU@h2_CT?U#*Vbm$
z&7W4TGW87&S(O=x%!eO%fZ1}+Yzr_Vzzd^vz=$go_6$?aDXv9nOo!Fw+TpnOZVSwg
z+iqj3WbsmL-Pib6W9LH+j3oDH=u*;+=g{?Rq~$^w0SAl_*s=>(PsH>Pzd}%SEJZ*d
zYRA@?wiYZcgmLs;1B~bvV1Pvk!Ow9u1!%#!1t)NUg&Qe>0o9}BHmg~}aX=Cg074u@
zQ5vgHg{(@jK<N@#($j@Yt3F^v)+9DYhw6183T=1kcenCc3lFtRe3I72b^$WtsRTxp
zD$shDfD+=LbT?z^A{Qz2M{NVLbhc=t4LG8;j~F2cPm2IS#E;=ttGlz^f+HnzrDA~*
z0lxr7-LeYSAlbYiIu)bkN(}HueZYlQk@8q865W9XhtwB>zE@P0$y!GDg)1#tmqe|9
zDz;uyuIteHv_J44EY_M+CQmY(O`B#WZZyGu&#+bM*8A7o(qh=yf4^w4S-xULsP&_(
zi*YT}UP6&<3RXMvpndliFf!Lv$x5_1g4X~TS-f<ayREE_>E>3t>tT&e7&q389h08#
zrYt->Yhjnb&^9ThH8nSz4!!53ZOD|(-cDZ;`PxsmAResobNp(XK1RNR&tSD5)sAM7
zDbvX0lW>3Qz@Jp_h8vSY1V;RRXDWct=NOLEm3{<#04{y*@Epv+QhhgCq`=Z63*J)2
zck33fTCO!W%wlA)_WrK7Kh`U*a9wRqG6y6-DAA3?7;&SSzvHSkYfRm!+JfbY@^_~r
zHARg8--r=lBx-c>oE{Z?J**iv-E4gJ+;wL&ZORl2goNeu)7TWf_tD2z$lsB|J(OQ2
zcwWd1O|Y-o{AN5)KjApD*(TGHioN>Q+ve{tE=)=%VXb_qkCC>)VA83_A8TfAwpnD8
z-4dic@aW^_voF7j%)^C!cHhnHwd*b^^KkK@$DcGGe)?J3d|?@ea=!QsZdl)PNy}zK
zmP`Gv%rS=^V&-o(*Zs`s_}DYgns+|<Fr&N!oe5xe-WGGr5eFR*3NoYvM%)G~>=~w-
zQ%x;RrnWj9PztLnI1T_Jv!-umjz08YD^Lnv9|1<zjYuZrxi0s}*<h_5g)-8e(gGqb
z2-<foio;P|t5_qen@#Pw8hfOl1WYfI%&JTvL#F@<-akMGg0NS~JydHWU_lH}!SbR?
zU;!~bdV2sfWOX9of!H3}1I7v0TmT{b!>twIONl_2CY8fN1p<(;F06E3Dr*w}6~Go0
zO#mYRN@Drs)(UGB^~qYVs9%tE`=z={6_<qbPFau;7#*?!<aJZM5@Z(1Js;GgO?R<e
zkYX9)h=A-R(oiJmy6unF%$|0+_!2)+ta?jSFa0f*1x))|S)T|5A=rDltWyw}0@47e
za2+>cB}Kh?u^4gv8!pFO*P-<d7-^Bk<StuoWp>|XCo^%vgtU(jMT!?*dDVRJ^*44t
z1q(2)xEKK-EJOjO>u-tHfjP5hnWGOmDAW~Zp#dWY?6s%aZRZ_b_p>?z5Z5;}nje-d
zu{`;r#Y+TurFKc0RRe*{8={NQz!3ENXDWWR-<20|1*z8MLLQTJmgNOJ<Qq)a@&W2G
z2Lb*&<<_}fyjO{MwAS7dGfKSZhNfJyp-h~7Slg_wuQ#J>stcMY8V=7hAI~D2`y~EV
ztt|QT;c<g++Xyg{&FJ?dZ>GRVb#;|FLR#5aH~1RaUD9&=X%$-Ijn>EiRB4T2O%kLf
z%YKx5s0U0g;7B1eH2!^7NQkhp03!&23L0_*yh}xFR@k)o#F*akz(cmYpYG0ANMm%P
znkx4*@5eA<+Roc=XEqu?&Wu((lo^|DA^}RW)=xuYlVL;H(AXH+r_E>1G@H+yF3tW*
zGjsZ;2`<FYO#AkSMG{tZ3QYUNES0rUE;N+$B^6_y%e@R#nJ{axZRc&J`8!z_U{z-J
z^y!hnCt&9{-~V8r(?_3vZa2&W)z4h&J>|G#Ezl4wn+_Op`NN)JsyW5AD5f^7F4s=*
zc=&<)n;o{9A51pG0T_V+V66)Rw)xh6%Vs>2267}Jrx6JI9^_QfAQS=^X$2h6_oY$6
z*RAD)+67D~m-R(QyELH{J7bi9613TgpV1*<e}$|z04EUs6Z?ZWAu-F2E>kn6+E&vl
z7gLT(6_0~@991h1pZZqCE#ZWiBIVlbHVIrPtCmHFKoBf5AbtZtv^B<JoKy+CA-)KN
z>nea9doFMUs}yI|g5?H4Nz8p0F1`eW0J;cIA=XEUHr&o)(*hT3Wi=w6q*p+V14pd&
zNH^9lF~y1=1?4T4MG5YU07|eDDOMX$Zyz92s{VG#B1IW3C<1Roy98_R6+n~^S6a5g
zVe}{0b!dG8M&JRDJorE}f6H~s6j%;_yJ(TAuW!&kpweuy`R1lf0wAB5qbO9o`{74^
zvOQ;90Y*AC$EpSD6*?Vv<l$zkIh!YSv(SK%xdL%!O8J7l+04y04Fz*}F2qJ*e}@GS
ziY7rTR`S53Klb$B&DsL66iyoWK}hi3K@CE4IW^(<)fPRk(b7mS6dGt(iK*6QHztJ$
zxQ+lL;WIQJU_@BWxd=#oxA9?hW|GDn1Q=<SGF>k*-()$`SyU?FsdzcLQ+Jk6uCAd_
zf^}_YdqmQO%hr4dfz!2JFAL~8jb9iIr~OJ;|HIOJU?d;&FszMyrVYk32y5fVwT))-
zm{IXtArDk7UM4ICY&>I@Ct>SHm`n`|s(if<eA`BVkt`onuGupIMliv~<r9K&M=;2r
zeDS5#{#q@!S&j&TeC%O|nAtO?yQKK>xrMLDa^%hU?@Uh`nj5o@G2#a(=e$co5D0&A
z{tFeH%>K1t|NOjRF9xv!h#3HN3f%a&OE0qkNkvgIh~P{l(L4im*h)T&Yb20fA(Oec
z-hI!!_r({s9{wP3q~Clb+a$2*Si9Dg){Zu{W2#J<%5!h{rld~tV&0GWG`}I=lp3bK
zMSY_jo6lOx$=!yEnHn@f^%^_PdCCUI13~A$yYFuH-)&beClQ)}_x<hBv>Ih-g>5%y
zjydU=qir34K{wrXkNIZV(m0sZ-FM%v5B~VOhPeinAkfi2{cnoeI!wDXf=5>*0(4>J
zT|N0aree(V)rPBQAU}})Tz2ho<h{S{wmTK8<0n7ap1bFibYYMa4wkEnU3S<$snvAA
z2&U4!2WasG1AIqqs_gqDKyOq{t<`c3Hk!?RSg_;v=AeD{4pxy87-4N=X+zn(R2t)9
zaZZvd9L5RRjOQ@E&Css%teKx{maR6^rcFv(n<?af_xL@~f(Y8kmGZ3_%=fNR<I^RO
zu2}~CSZ^?3Xgrq-80iw=K%5X_gOtnCqP&!EXODG@g`gix2!#GUDrW%-0SZ`jlvk9P
zwpO{h(mRIhD)Ln*_D5-nEJEtrtlKQCK1v04G{`yycUQ#a7$xAKiT0F^TQA&Q6-#@l
z3~LO5B7i470%ibMtTl+*#tMZt9pQ@00v1Xj4`f|t=^QYptzKY_zU9tR1OTYGRIbBX
zHCF_r0A6$n__3sXAUwoU=~gV3(sF<l02YBS+#L1vfg|LPxy&jztpZqj^?k<5#F4XJ
zsZ_A~YL%^&FgqRj-v|FZUAaQMp6fcaJ^>>~$yx$s3P<n+_loyhD^LU?UdX&H=b4j_
zJ=#=wG4#0BDj(~Ax%xW27cmgf=GS|R&-JQ4>&rqEVA4;YqAZ{X5*T4zM~@yG??x&i
zQ9P4(!ZQQV@)g{;jAGIsue;8a_9!le_HpC1??Z8G^O-Y}@~)8v>&<uFt2l_6ySvTz
zdhgxR26dXr+6ljk8!aK=5+JTyw}*P5GU<D&KKgspuzac3<tQFLte(M6qu<q|#-QUc
zLIpYk$Y1Kj17XdWD=`OAqHAnvw8|*m%}u7Qt0>YPDLieBWI4IFn7*gHlgNXDpbRD5
z`|6HiwQ&t%Z_+8AvrK$-nfTM{;xfDc;F^^+^~LNA`wj%%l!fJ?9w*~!_>n!(WRl{P
zs*i5nAWwxr|1g;x6>%LolCRM0$DlyY#5jfZZ^^QiX6&Rod*AU+1y0;pxjF-vBY&=3
zy~d0gT_a_M!dc5BJ3De=Kt_OX#0W4lAVZs6X)0i(U4j1UDr?Mnr=24Iw246!agX(=
zw1yCr1_@20b5A{4ruE|!E`9#h*XEvwAF*i?=gHS-ap(e*P)I-ch*U{?ESrrI3#bnW
zxH{be>3-+y$R^`7xy72^=N1cK<d1*7$^s;*7VK==ciIE1wW%B<jn#Z1k@O)OAv>{U
zPCxOu(5^Edo__vC^M=5kbcFWWug*JLnzdzigLT=pH>3qrqU!WNR71%Cr((>b`krfg
zzP0iFe(>&Ic;@MTl08SD@dwT8LZ$=v-plN|(+){pOb3im;HSoBQ)9%8YLfz?0mBN1
z>2rzZX>bj)jE*$a%LQrzP*ev3U1l>M76{D2e|eBHu^0&o#DjH&{dVs>Yh{}RuoH9-
zj(i3#s3EyrerPAj5A8q`--vn%bzQ&cXOT3l+_Oowu4eLXq$;1uy}9y1raqIhBYE;W
z0z8XSffGdnIIv1>Y1ey;>nd&d5aMHnQ7t!FJ$gyo<SGjQp_2F=0xYl?alzGzEz${4
z0Z7ot+E@*0<+UIMFL|Hkhr0vv*3{|U(WMqCF5pTFaHC4#3%`riJ~2+H+XjSJL4vbe
z7g_=^unOr`nF_@U0lew%kYz|YX}I2^AL5l%mUt;P5POX*TCg^uj}FM-Ud49-df=|h
z29?)10K@=z;4Z6F@uRvG{|)emvRxgjqi@SyQX!B;b@b{T!vux8EEu9A!nL0ERsk;B
zpnF2LB7r|;0*7z|K|oOyG<XJV<b%ZXaS9v$HTMqrY1y7f@p{ygP5rABZ)Lpve`EcY
zPRp8gy;}l~Y&W>;3owG~$CH0{OoG7tnp<vD43U-b$L9q2SX5wH5v0G~aFg+ss<=*>
z5((ui^c@X;YlG#}{q}jnba`{VzR&Kv$~EkQq~ZW0pDXx!*t?LFIgpTwXF$29Wlg;)
zs~c@b={*A!Azo58-{ZQ`qs=eQIz7p11d!v38*aAoMBIDB+IQ;lKeJZifo_@)KKa!A
z?TM!>e-yk{2)|huAlCS?0_NEJApmUE-V6=WR53u@{xsA3NVNu~s)Og(*{StiYbXSW
zkxGFJUZbt8%WPDe3`G0o95B+>-eW2S;_ESnu2p`SRQ>d?5kI9%AIB)+qMtW-j!dCK
z8uAch=9K>Uoe9rIK1X4FPxbEPB1dp-4fGL<5%}7m0NnBD1Ks|HTz}X+3L6jRZoAf!
zbi6@WS(nCp*=>2uHqCeOPhD-MtEkM>mKU22@mHh7uQhkeg{ody-r1FlSD2cK<4lc|
zc{;?mA=7HpI#$`Y&vfgwt3P<;8^{_?eGZ(+=DU~=d0DWJ`KZ&~L$*o*j8CvX!7(eu
z-)kD`%_!~9vjIl1o|rhMHm`C*H#UPkhVFU0_6zZt={k~Y-|OqwnB~Qr<XSwm+?-KO
z=CgNV#tIAHe=DA+sJ^$ocdu<@LDxruugAy5p0I$OeE8vtJFsQK1%fSKb@Od0-C_|?
z_^<=^ODOl!Yp=_~<h6vfOnwZdpgWWPkq50RtZMiOK|U~I6Z-iFrm<N*$6@f-ND#|0
zTPl;LEA9`vh3eD4OekytoHA*m1<Rt!x2u1*+;g9-3RXmAvFT`~{o?F1t)<LrS;74E
z`kPYb&-CI@&X-h-d8Yl(G+!nHoOE?XZRkcT(ARG-I4`Mqz1*Z-dd>Ao>4lUK1&mxR
z!Oc&e;3lfiVb3ttoZ?zOTxs<O7>T-<^({T@*%ZP(hOLuP{TcyMqDH2F>5MDTUYEdu
z#vTd&1xA2YSnCgMXaPS;B*+ICsZvZ3tT?)4UC4Dn2ZE#%Po!1uuPo_X3E|4h21*y`
zLc9%Pcn~nXt+6ErPN;0PfDvMol*vMa7#*}de!?izC<~Jg1#XX7kq}5d2Fh3~4(i8B
z1X3FoS>;@(1Y4&b`qkamYAWhRDW-=&4fPSQ!-7757AkM?8^u3SXLMbzvE&xZS#*Hz
z5RAQtm?%`OHoK&i56FRgtr+YJcET9M`cm9?-zYg0%UuCwx@ARIr24BSjJ4J-))ED<
zr%c+8+UQ5Ef40%}@K3Jm(E0+5oFpp{cwR@i*Zk9!*SMed=*9?lOust!Y*ROSor?o_
z#6Mnfb&?`Xy?_zsZ<WBf{sJRQSFSZvCuA0IJfY)R{740ieDcHhp~C2N)w^*E=3_sG
z6}7J%5X_Uw?@B3E{pS30O@(IxMma1-;K2#B?o${D#&rU-lOOk6g-n8b^AZqvAOs&l
zrISp?F%!iU$FDk9S_9E{zps2p7Jt9gY_i$Lesa>eZ^6oUYsumt<C@>u@K977rVJGI
z?vUS&Nq^TB>WU)&{-W}PrH{DM3LBNeN^^>-sd|;}J!1}bh#xDLA`Hh~S&X!H7MU>u
zFNup6RJ6zzEKJq6!MMgc;FN6!qrDBKjkS#prdCQj*#ILO%Ghiu_sO?y#A0NSo^~pW
zkxe$9YR)?8#H5})|MIJ5;p=ZCrTYoE-}=?LXT_7yB)j$A`%J5}^3svj(3G&ESR3&&
z2?5%)k!QJO<KK;jWS+_T(^WfVYO_nAL6;EwPHAptO2kz;;iw~Q{0&zo-@u5gm*Yxp
zL?*~jr6YVem+GU6<wZ9dvl4i4AGi>Ks+D)|-4>X+o6ojBhS3aL`1+eU2hH|@kuE(C
z#xF?32f6j$`vsmQZnFG*{Q*YUYy<M3g=&2vqY<ILFh81`8l<J$|BbcZSK4EIX^U~i
z1qA<mlS{PBI`jcU^T*E*+Fd4>Da?s53oF+SmO$eK7DW+t(mvyjpl<2v1~X2Y$pi6D
z3^p)=_H>D!S=jUB{@h%S8XG@73K+RuoA@;X;2_4*zBhf(lQx_{z~5Xs!C^;)#RzpZ
z$)as^ZJnKu%t@XPbAvk9u34q|kZ5g3nJ~&07zt`o*maY4qVL`4k*2v7(q=h+VSOu<
zH2>M8>r=KH?R518ud@k4OFqmpF__o?{gVXh6GtU$eENAt&T|`qE`cKeBdeQZuCJW1
zpST_nz(|J#`V|5&SaVp90W<(doMlI^EG=}=RE?>TFh!Oq($?#g<p&@Gz=R8KE+oO$
z32;y|sxlT&M1kG;1`{|PU}Oz(PDH$iF>pn}-u11<T^Dgd2x^WM2*3!S3V@2W=un-!
z&se$uU^rmJ1xmNTi1PG`SHb$j62OSwby%fhDNc0Dy2@ID0G{;av)7fmNWe(9tU%oQ
zR*Kt6z!eLS=t!V>neF#Tqq)dg!RdZ0#?>k4yDa2n-GZe}vA%6%>uN@Tk=PSR`*(c-
zM*iiuze<Q50#Au+ET8<f8rMzRZ9UIkh9`XF>1WI*pMT+Jx90#OHPtmbPYk{u<9+zH
zqTGXPI{5S$zn2YfKqImJh2u41`wZ&JrM^_a$l|5T>^)`jXWjkDZJ+$lm+5_kf1po3
zSuJ}B(|lK2{`E}Gvjs-R)YVxF6L`-cU4Q$XiUqhln4I+6@g%IdWm<b6_E{)tagV9^
zVxCv3>q(6%0UnI?9Ke@x<X^gZLObwpxHqQ12=05d4${Yg3^t#RPYEh7Wnx9(KmZ4f
zgpE-?)7*C?8=<HL^wkM{u5zNKKdmuSC$4iTM<xE66RgyY$v^@l%U0E!DdRKa_g()e
zGrGLj?i*d&Kpu1H<3RRjxYV5qFp}<Gn1kh#$24jUFVbEgV5CR!-YScF<&Ko^ULvIf
zS`u?7{|c@h$vH++hQkP~T(#Pak(Y+y0*v%`-VTRnGT43O+cpA>4AQ6vff37iO8@|J
z{j(zvGxHP&B1mB-@WDJV(dV6ZvJJ%j)wkbUvum`(QW1i30y)l_G2KksXo3~!!=OS4
z8VT^ec-b=Zp-i}cShCc#D2QlfpINTW)c0qX9kw%@ZXB(-^2R&w+VZrw*RH$Rfby7^
z!x%$oPJQ>oBE>^_%dAjf`7lCAPMNK?*xW4GaR)0j_wykz|MHt}<64+pBRKg}&%dDf
zD>Jl7+{(|MbpHOwpV;zA7nA^FJhr^TY&vb4*?ya?4ZuvDf}&eyUhSg2w`Og<`S|nC
z%_5D{(T5&l$J4K?UBVwU|C=Nn$u=mF4~%^F<yUqbNDI<yZ@q1vd2wMdIsQ9aV1(yG
zJ9B2ul*_VBta~?nSF;Ing!bC1)vL|d-+gc1|M=5LAIJot3=4Ga>iOpT?*$xvWH(7z
zKW(c`Ju!CNd}RFasZARnefn8^)72pmZ@+!^HS@NZV<vd{5xOs1{*z(u5r-wyjUM+r
zX`_wI_VefId2g(bV_jSzVFmB9-eLnzAX7G|={A4v{?Z)>0~oQ-b7JgyZno((>(jbI
z^SV?*N=E>xtLxXAueFA}_t8i89dg&Aj#0n}8m+hj<JoLCe`_;w{CK(3Q2;#q&a`SA
zmzuZUd*6Ke*(au2K!;CJo2oJi2zQ#hm6@~oEX|McRtuB+;yE<SZO!5(OU!3qeQlq4
zIxDSgff3ig+L~%JZ|+>H!8?BJSZjgh)&k}ofpEW9Ki<<E@m-y`^26zvYU4Yc?qj5{
zi6fA&Ae-%qok0NTHd&0|qN-H!Hd-2Fr6H>Z+*o0OQ7%_hP^o+&w-bv)X_YmKr6GXC
zrW22ZynSFsl|YXc#qt0EvDXEZRLHef(;63yz1Lcr0El$TC7l)&+g_FOw94&CtgaX!
zQX)`9_=UA1v6ZQ9+-@Pn>mWe8h<52CmL|kQ>F(^d7Am%W@~D0-FrYj}5<kNlWZis`
z&OZoNMs-72oS-P6kCz2e2(Aur6T8-8twOLgvA+TFt7LxPm&aO$s7>1FY!-l|{?I@E
ziD8n9hF&v9ea)A}NUj>Uw(|r0#|bbwmPo+SDfeSz^u2NfPz{B-PjXgb>jN+X-}aBc
zyvW^l{CN5GH(I{aCHQMKeEkgZT&opB@1xH?kH;Z#dEK#5z(}Ql5d^Gdaw{}(!UXX?
zTN{+Lh<Sn4vm?Ic@VVbDS|mQ~6Y+g(+@++W<D)F^C|-l_Bo-6Cu;0}O@AvM9A8Wr6
zi<Od8nU~OCU}XHbv1W()^Ud@={s%=c2tj<KtaFzv`_a5Bx5Ym!UJ~c+{BiZV5<d+S
zhAmn8li7G`j7QHF7^$tPF~5|O8+)E0-SNOf;?2GbCfnZt4AVE=#KxP1cO9)gO1Z|I
zbO^qyskz0h6u*v}<qyS2N9_@J+;+a)d2MRT1^-ab6VE<x_dWD=H@Q)rGi#QCFXH*a
zt0P!??d^9A`;{QEhrrz%0qWt|sRMun>V$vdIV}Cr&ap+}5j&(1A=|@XH3juMieC+O
za<BzhnP|YfbvD>neRb?-EeO6WUbM_i+hmH(p6cDmbWDQnX8Id&XxUG)QVW1JnUqh}
z?@Y=M=R8r*J0JJtzPtUci*GwhdzAuOgYz$&x|xFnU&l2w79#KwZCXoe#EXUjBiYnl
zsClEF?_jSzZhV4@#V);&k9;G?$l`V&LES>W|5Sa8nqR>(VSG_A55F_5FRbi9rtys!
zXkg^rKip}q{Kpf_h%2oS;Dd)t{D<FOcwU@|bq~Bn!oELVdA0kQmm`~&trdLyfW7y$
zOli7GVT8Yt;QP^M{vIkQJ@)WJZGBPo;L_`@_dZZeke}I&S5!VWO7}hTm<36Kc@Eic
z9~(41n4ao4Aj?%Z-D=BGJ{3Lv!i(nhx8DuT!sdwp(`eu^c8(YqC@k#<2}bd+P7D;T
z12kPCU?dkacpn&HL&{hZb0lb*k0#hX5B<%86t0i{F&kh6LGQsDGr$ZO6rZ?n|DN47
z?(Lz+pA2oEdJaEuf0s4s_}!w#=5|@U9HY5}AUzea(SKO&Ac#BZ=p)Va%{Gh7PG7OM
z%EYA{8_@&z*~<nvbsM#)imCT4ZRTHG__7uLN9A+r=@$Qi21a<E2kG7bogCQY`ex^`
zo^j$+mg&5n=ii=uIubB+<->px>iqA^uCUL1YYAzh2tZ@uYp<D?UVqDO+}VurUzY$H
zKgJ024S~(Rd+cTb+^G6oI-nSR|7qnqAMEa$J7x=v&<Ez`feP4;5Wo@dHY)*uq5ocx
z{^&E$T434b9d5@Bc>_7zp4CwF7ht4Q8tF|Pas#C`rbred)(QhbKGgT^!1Wc_LZ<)_
z;(PGI08n7PLF|lPff3~bL?B{)^hj&im37t{1YqZz+KE+(z!U%$K#-Vgt{w$btu)PY
z%LSkki(Mi&T+*1fr1i*x3+{Ixhyf4*(dW23EIR6_4-CQDqlAFxDp#U1or;Zwdo0LW
zoM_W+1G2lAAZo9OG+ChFN~=UJvU;TrP5hE##c#tBhJfsjxL&NXxX{9#mbEw$xKdUt
z7i8RTpLNTK#fTfR%#P~=Fv54`pMUeqr0@2(-~V9ll%U*c*XOcU6Jtb?+7&~u+8ns=
zK9)alf^;{+C>;fPKrFmpc=c7wGh6<_W!8~(V81<gH@j@Vy|p&+F4=E9zy0qdzv2Cf
zp%^Ae=ena@ff4R$pFQZ)wh7>jt2cN=$?9)UKW$o?TT^Pq2g`;1&K4K}EV$tGQ{!NP
zU*`>X+-1QVKaF#&)Q>*&V1pa3;Mnqff@zdluKm%2+Hb88h!LbC4m!Z{Tfubq`+k)}
zS&q0jlP7NEk`s>bKL9!3i???KaC3zA0eCoYgf$ZDBA|-jPk2N4_E+AF@rEvcAs!tF
zFjy8NLoo(n8vLo&oQCE`Q!O`r1FA4QIRF4a07*naR3%(wpU@y6Etg>TBd)ZDVg_Z`
zCYCmn$4h}WX=j>wJ-n9yA{F{xIbftox-NCqW$P_4lKJS2<cu1DSd0;X{$%BZs4^qz
z>(>Yj1~8)Cvt8dtfRPAbWXsKGo8#pkDoEeTlHxWAjPpe>p1bd|li6ptT@yB11W<kZ
zgAdKuix(Mrxi;HvvyIg}@)=dM#qN0E!K46=)22=_8&8>HtuoN!4icLT>YpbthD{ud
zx^H#h=SOSo+FS2P5;n8Z+IC*d>H`6PnBW@A4S$oiYRy`6%iZ@{%{MgQuv*wkf=&OP
zT-#B=2zdbrcHVAVm(?CWl_u1i3V4p0I#w`@fg?5qx83)EwJi9V1pTvTOgHa;@~L_1
zxfc@p=kljJ27wU-z6*BT-XM?;-piW$dUMmA_gE{OV7=J@BRj}4=I{d#2v%V=@!x#+
zeSI)W%pMDNHq-Rned0M24=1SohgKLHD8_%AKnxwqW2x|>05kwBHsU=3d3^0czYWeG
z|NC=h!nkn)y|%Su3Mhozhdw!6LU!iF9rr(IzL6jyiU7ZG&&I~mCo~5Ldf6!JBY?5Z
z1a$d?W+B1}=N8Jpmd*J<10(bu_ZmUs-F!pyxI~&cdo0+=Y%+CfLKg0eAg{dpQ4|J@
zpo#d7EGBl_X-E4m15`j8Pnnzq=#l=c0H#;oSQm`SZ8Tv_nB#kDt`T(A2X$KI$N&A0
zKem=Zn@!u;Zsh&<p;_9IyoXWXNVdQTYcwFRTYFe<o`3ljY0xgw9PBZZWoZPkhLw>c
ztXA*~j_Z@*NU1iyLv=rFi2Jiy8mfM;Z|$OuD(Sw}5N?s*=5B!$fD7dUJaA=20Q7Q!
z8CX}82>gfzM3+TIhctjC=<l!=5qiHhFfP7_UBC6m0UB6*U^!tyi#TARI2+??ObgZ_
zidh1c$bWQ=HoNt*DxodffLu9i!Ga}*`UU~0K>9Sq{^*3ffGk~#O9Hrp>n(z!+k9%f
zRPMpL``YVPaCLwT7lXvc6_MpksVqGJc48H)Ulj@>5BMX&Y)?nK03QW^7x2NcN8`sl
zCdP;bSJ;Hn4_tH!M9~jZCM(%)1*f-t&^2|@j2&H_H;WNl@c{nu-&KY>2)<|9bgt{r
z`T&eDM#QM_V}`hK!x|P|i0?}-V?>z}DP^5{^f8t<bM<muwtTsHQx=^oWf`)AloFh4
zE|&+v^*s+ilB76+($z7C9g@VGVi{tUkMw@S7oarc6V?LPOYpy9Ww_Nid2<0q;0caC
z>|irrR-Qg#5fA^cR7y-c?yy5b3>WgDSatJV_nMU}e(aaqom9XGRzE>E!dPmYdi=3T
z<A<UOyacgEf`rmh&>|bY*IML$`ITIc)>^O%t9ze_0e9nF_aw!yoH}{3-P5q=!ka*l
z*emib2ipwh!)oN27hbZ-@O}iAcM4tX6<&Jn4fDk}-`FvjAZwoe_T0l-sk?fZV>fFL
z7j%=>pS8r*9XtkIr3SC3`tA7D9=?A?B?{==8EVNkP+Hr77Ng7Mt}?K~hFOOCYjUc4
z%H&$AHOGxfAqp4$;OlNK;hMQjBRCl3+8m6w3Zac$?stVi=?UVe`orrp2MNC3E#SRd
zURjEzc+xAY(+aJP&SJ#xn>*(nfm>mXbMH$o@rk0w-qo4w^{8?q>G>G}WQTQfyE|N5
zY31jxM{4m$fN$GCTa38&Mu3qBVB{cKBp|2@5*7jo!t+H4=}$cR$hdHidT|kir3RZU
z0uB~ygI^{sGF(ChH$yO}2pieN1Zk`Uf)|{5T9OGk%q@ZeOu%1w<y9N|0gz%pX>d7H
zcE2t*dT+n~fu9sV=R<Ys3CGFIdy{w?Cz-?;VS|Kl*a0kl1yR5V!u+#OIVl0i!G*we
zx7`sJhL5QmZO)g5SEa_x5hnRJX~R*i08YeiY3PSwGz=I)ID6vJM_4yuu1$CRSiYnm
zdh7`cFt{|1*#IK|IsjQdt(A+Qt8Tg_9!Q=|*00VxJ4u`Hjd$L&F3kM0?wn_Ej>Z#r
zWI<w{tQD}s^It1~43}0|>fjO}NX%J;<P9>rCqK^wHwnB)xNQg$pvdi-w_(Jb+E;TC
zpvNcbWUkW&#Q2?g(uvl^m!I!-ZT_C_+x+?I*_;nFFmjCE<@s9$1p_?KTjTQ>_i(sy
zaDso=e>T^+PVk9!qN>jqei$&~^5S;mfk&UPZWl;njX34FW6ij+>of*QXFa*<=3C<~
zKm1J<YsJqMeAWk30c=>aSW{p!(Fg^|3$9nVI|CdIbE}mNFhYEki_SVjpvOAF!UK;!
zF86s~(ET_uXU{(6=W+LITwva%0_Ea`3-Bb>#&@W?i&arj05DX2+0fhVl$)b=31yl(
z<d>c`S-=9WsI<)YffCwa>b#;#77>y;bhJx5SnE8ltN=sEh;@i{xg`(<VV-qUg++(X
zD@RwEwg$PTl1r%`S(-R2lS*w`$lEUWS6Hc33h*e^u?hEFDqE(yh*Ltn^vMD<0!HXd
zvEF53i2zy<d&B`L7OWw*iSDCFWw7?pVra`qqr<wp>Q(G2KnsGx$Ahv<AZ6?515NZn
zw7?4AI_|SeAPoQt@k(s_B?<y6AeC{jRwT@4@d!l%_ANl7a@-r`i80bq-(+K$aR0?}
zD>H8F=wSnl`0r{uVL+;PDc5yqeE~+`*|(TID|Cne=wE;5UGwrQuh=*>p_w8Q8P`4b
z+|wMf|NaSrcobW1xZ`fCV8e4O7YKOP&rdLu11?<OliR6Bo_Z#s_Antmz|GZCn87jx
z8YMyf#pj%vv{wkL&#yckFw(B~qaxt$cjxW5vkEtUJ-F|^<?j2;sx@mYpj0<@ygB2<
z<E?_1pYH>?o4Vufy9NvxxlF)F7we2i#~pE45=h16zrVE5y!z(bUe4Gl>%>K8oo*F{
z{EO`8?)}@NahG7sALdbzu=2*ORv3ltZO8{WgW?i=8&*E>?(?Pa6m<WEvI|Nt@am^&
z--ZH<qp;YSqiSpHUJ4HijvoK}v)T`S=ziK`f7X(Oit_Zs@vB{Yi!}ej012A%*TXzD
zZL2|uJYGWNAYT*~Fg@4@ea+-tC^tYL)w=A)q!57-`dzKql<fW7TnZbjd_E(;F4v}?
zwmzQ+=UgWD`s3P0Gigk%o)J$eH)0NUNV%p=>r}fGPf7)hG>gR>EhU<sigHt3T$TW0
zI#*!ymsV-p^?kJz@y5v&bUr8@2<O^N*Omiv(&Zbup4|wb3{p3`M*H}fMR|6mM{*C=
z2xQ_aGS|NFjTiw&avi-a%LQYMxZ-HiAOH-~%Wu4Co?G}zFgd5+2t3a|`R555IaCA#
zefL9ui`UuQ+GHxE)i`#{So4ds&xmXIxD1cUr55oZ!iddIm|)szU%8DDWp`Dq5HzfV
z6zq?GosU(nzzFqVvB1V7NKwEDmMVMfk|;3uwV<enjR_{@zO?}x1uT8OmqkH;2+gC}
zQ#V%pm{V*uK2QcQ!SiG@;S)j?+!C<K^eG!)gnQ=ReZoTKNm;%46!yF^%?Ch<A_Ul2
zqy&jg){S@Goe-A+&9!q+Jt>&g{>C!ritDa3_0lMIgd2hj&pbT|tiht>50_sVDj-K#
zPZ{5zHuV8WGTuJnit4Zf_VbhMCXr3Q15gq!X82>Efe|z`ah>NA;3J@tPrs1xpE2|a
zCV}V4yb7ZzU?j>~0&tf1B}l{=dQO{Y$Ze`BWQu>5z^EXQ_MzOV-SY6mHa3&12Vm^N
zGfqnYcIfki0ttfv24%7VM)28<;DmSG(G@q|9O`G?U&~qVm>a{vVkC^oNVV~G<51|w
zbZ+bhqW?qD=0My>syf>Qjur_dAjSxI0suu>MVV=DlEnpJg#Ot$8QPc<Qv_;lTBq$?
z+6PQv-GQ}+UM@+Fq_rt)k{$sNScBLA>bS2GFj6V7qg&@~ia#<=fz}%puLM_IHZF)j
zgIWb|Z;~r5s8UuH7EqD(h>g*qdW&UM0=NLV0O~P|5{W4U&d_J5mpq_2A{X<+{^(o(
zaBtPFHL*hAZbyS$cgfnYSnU9|sAk)L`rXmksvz#wHkL_`EL3ohr5m$V70Y!|w}L(w
z$x^RJ=~hr8aHLdmGzk7qIwBju60Q@ogte@=r_+oRSdvMMk%7`qwl~nvKZ3#}j9a&s
zrmHU3b!dG7MmC!^&ALzZzdyk-L)icEXJ5$j`vYs?8TGyP%XN0Nncx5NV)@fv2l#Ux
z7e<dRe93-aou=#pDUKW@1rncLQCz!cUtH)XnVn=^edzxCCZzd%tGhr<#3_>!6=bkX
zh1Y;z3!`+vNU#jbO20n;oCKvM^7t02Zf+6t;)DQj03ZLCE3Y!m4)hE*kiqX%79&`I
zpya{Q2l%w@)?1ka_S#E4%sN3e&yl!#k3I8jDE~?q@HT)&t4|L<@szc8c4_dK7m3d$
z#)nTIe)5@l=!qx&q=a*L<B1cJJpo|pa`!_I8@w7oSoHFI9HG2~D=p$jV(s+Ha}t$q
z?t1XyBx@te!e7H@L-3F&wqz1;-WC|5Kk%!qVvk$(O(Gb0rvBB3sJnS<*1bDn6*?Zi
zU2{7u4pdl~fljkAT~U^hfV`#;7>R0=b+23RUWF`xMkqD{sd$P^h54C|<!ZoJ@sz~h
z&6ON5Qld4sQ$Vhf5<!n-TUCnHwzj?3OcwCsljGa7iHk9qfSTXeT+iJ-=2E^zJb!hc
zLS$I^bk~DrGU>~bWh>3NNn_%Qn87**;x{p}$^|r!7N0&s8|Db`Z5t9`WX=vdC22Ug
zN02F^zP@aH5PV}95Tr1znqZzxf8z$|bZPAbH&;Q8p~jX*QzbvxSabaRn4^<S%yCVF
zyDK&gK|25RQ{*x!`97Y0UNJnR)#fjRXs83c1g%EU)W>~K*yf0>bbo-6*)wODlcjwg
zq$ps7xG2FOglNbGL5G9opP3-^J`vl3=hq)5nep0U<7uJ52yJaSXLEyfhwoyDbX-$m
zv4SfoN7(=)XoC^Yf%@@ZO#RQj^s)`8=<4Hm=1C`*O*dYr1x6Ys^8YOmB5GqEHdd4)
zX!gBw&EOMr6=C;AAsaK?KwW>!%?a9k<UjRzEQ+T2)!4pZ<@4iDD-zQA!F}$zV3&mS
zHVL%<dc)00F<}7xn6trm2#X3VD{`fQ21a<ExX%L^tgendPuxHSH4smi@IUC{qC<kh
z|M}~cq3=8Q8wHFI;1gFHL1OHFcj0+S@r_o?eD(4hZj{JO8)CVQAjZhd>C=+{3FZ|*
zF3;a5fZy{3qD_?5hEL>U6CR{&fDwY(;vU2&KtjO6AOS$l+hPj=d@HPf>_zg$JzOZb
zR;Akb4%I#R-;jLt0bT=vHs6*J%6tR}++8H0eXD>G$%8CeOAG-JfE~E6!jhs`AVH^o
zS5yi#k$Wv-W&m73ZH+DV$RhxxsI09ttyqW%Orb6dT<8c{U;}}v^45-#J1GGftJeq!
z(RzUEtBR^JTh?Y)8u_unfN<#r*g|{~3%n4+M8sMIh+zR1^%r+wSiY3VZI%U8Xj|9W
zBy|Yzfl8FGtBF`9+Gy*$SfY3*#7!yFv9r(Z6~K#)Y0}qzk+c)Lq=nBUBZdh5bpQxQ
zYe6C_5-dlELDD0Eb&)h2YbJ~_%`4W(Vni+_)DI&8+&Br|G66<xXZ!glxE@4Rm+yBn
z*iNqN(E74JU_7F}v%#@~Z$`Q3<=5Y|-=N@i|M#4ko5>pUSU<_~O4r|hr*%>3UkCyR
zSgPE5&;8crl%EgF&OPN`c5At=s!=R1ct`s3;;XOu*`yqCbdtq0$}xYw>RRibFClZ^
zk8FXFog`pC^nm>ms)B#G^x7NZiY#t!0bu<8q6-ofD=7Hfi!Yg%-+CvZj&wf|5VdG6
znIPUd3K&7L2TM}=1Av2K2QFtFvA@Et82ctHI->Y#@>32aGc4(GRmGU!cJBjH%v$WK
zvw+CO;$8hVxQOx=;6oIln1U<fAQ9`2y`$?FdB_W4QNsQQ*J)hG)$%#Ar1c5Qa`s)W
zyt@V{0zbZ5);!!hb2*=st?(|`wp;6bcUcHv+q68F>6?QVdZ7^C=zIzGvsIjNzz7y2
zZp;d0Tzpqrh0>=&8>4K;Wu@ZbPSD;mU$k`PS~F!rUB5VU<{-h>o0^-%1S)o0OS2IR
zYU-q@<6LR^{q^5zKRwg);r<LWz0dVM-%7dB98*W|_1r<ZLLQ@30B0fhdocSpWWdPi
zP3M@YTKK{UlQy(3CXT8}`B7b8Ha@>T@0<iTEpANk&x(aYJ_y%UL3crD^<1Y-5WZ#|
z`3%k8yDcz#OQ7l#8?S3`z1@JS1}O>{K?^J_$lq_I%@y=}i{_itqzcwA-gD4^5jGdN
zeF_3rn5PGWM23wM8wa%2`%7W2v|ODCcF?d2^YzPyhu9iyz@Th^5m!HBfVLOfY&;h>
z!#?2)d1grP^Z)awOOjl9xpGm!2pVzke2{1|Pnly5m4E1vxFmoPFTC_Z0x&}Uy?5Wk
z?7izwoF)(!JWjjKUEuz(x(Qk&Ak@C*mfKyn_|a-0>Rn~C>(i%Sd}&=f`Dxjl4>U01
z`eL7_0LgM~o}0xaHYpB=Phr9euK&khF0(H8{JicwN?`M>fHVSUhEbGjkU#wOD(h~;
ztw9IxyRVIj<k#Wio_xP@MUpnriANo7w%B}gKfAToxa_(cllq?xFoNH7;*0o%5a(H0
z^8u1_?uaoC1C8DF;dFFDqv6!&k-iSz2;fSqTXVlb(?1rsLvF0(Hmh5}LzOn4-~xal
zRRShTB-n3Lyb)YW#jGo0{RAW^E+wvpAS$`6;&-_~l(zb21!S+XuDQDVz!6FilcTz}
z!q(SN-)hScjJ>S9!~!Fu##Wog)iKKvtUyX-ZNjw{Kom!S43>11;0~2-3q0s(U1)KQ
zehJudRwe*4rMfp4pM)#`E&wJa>TkF16CkHbLEziPyAmj#R&cvjDb2*zRcq<I?p;<c
z)wQOlt(7XQMF^H61bFWeI8s_sZ2_Au{U#m3VMSeyY1eq5NIslmjMy8@!=GH&q4i~d
zz!1Xg0lM!Z*HC<8(h<D<yB~gRo_qOaYt<E&|FHcPaDBT(g&tyATzTUyNlGCIzE6@;
zQJY+feEr?`icR*e`Dyj4NH;w65jUs^>RTkBY8060)A2_hZWV9*G|o}Npp7uf78t=Y
z6$@3L;A8)C^>uN9yg#>ocj5U-7TDh@Hp;E{J?K}F)p-;!lCBbznjU-lS+nHFACq{O
zsIn-TVL{HAqhJQF<Op}57t3k`UehO(h1kzUQ4}x|+y~(v4NLF_#pC-%^P|48F<xbu
z<u^PM{2}0p?@54p9t2C`6&2`-<5%@HGNTGen3fM<Bo#RB-ko}^9QCpDwfo(e6rwAw
zLIg$%bZQS-w+0Fr>5|1mjrh$lU}VT{cBJ8g8iDm47#Xa=4g(7NqnJQ_P?>m3=gNJc
z?uBpLkXVe&-C^f`-_!cC@reqqm@i-iW)t83L3ds4((eE3^*2gL8r!7fZ?r=jqcRD?
zVQ9iMrFm#UGi~=E!+?=42}%BP^|hAyb!Pc4f0(5VO!8&d-VmxahM-|EEk<yk6AZW=
z7F&RzfqN*Zy-x(~W}Ny<QNW1P;IXkU<o`L$s^Z00Uyp0*<pPYP>pOGGcOmBT|DQi!
z8Yvu(0!9FF*l-6aD*g`E2=Bc6b`mfWwRsLR-Dd;&yNk|G@{Js2Wq=Q5EJlKHPY5oK
zMM|zTSinfS{$WjpRaKDwze_F?z!vV_A_^F}PXOYlUw##=%KkoJuRYCfJMWlu?Pggy
z{U8eqx2Zb&l#`PzSP*Ow8)Zdvqp2-P)XqEdAjKJ(zjZ>r2nQ~?=DH+dOE$m=?;s#)
zm_P~v27ultpMPnA!{N56PZ+Gg54Z6RB#-<Y;t`OH-ugby#vRc>bh!fs7HJ1Jx2}uT
z0hoYGEi6U43BoRrVAL21{s9W~0dEp$LCg>+76YU>9<aPppaO1Wy1IHydxJn*1i%6Y
z%4JCcFj1lnZL898rG<@8gM4_Em*ZJoZK-|$iE05dO>1L-@0js5rm;~$&n1K_l{Rl>
zg#zR1cbR|;tXO*VFXp;Sjv&+yzj2=h5j4G2u}fTlcK{n)Zn*&N+)QUno6T$craFmP
zQm*pU)uH$&8UP!#-Q|k~VOM+-2`RB=>CyNRY#qWv1nO#UiCc~o_W|(4Dk+oeG2)Gk
z6<9P_79+Mp1@I@=b!dIrANY@8&a9bc|GoA|iYeje@y}6ux=SvP2sZ8$dx8s2KPAb+
z63fl2Z@D$;d(Zof<tY}V`JgCZ#3@IFwGqYJ;es_7pqu;q{Vy)C7SDcNxMjLQLEhbW
z+Mip*>ey)fct5+wt-)HDxJ1WyyWi=KQNRdpqOhuEZDRc*A2G@RHt0J<9|10iFM>;|
z{!k1UIWviO%>@{tUzf`b7%>4ep;}oE0m{H{2MGY>Jpp)F#W2?HxdihS@^ARn4lUzY
zITw<}$Uu*o|Lz3aay<M{03(Z+EH~3OPIS}f_m6(N7$biA$a&$6z|s}67}+R)79+)q
z5n3c=z~*kb=Mpb90*n;Svo5rri~u8r_C!+mCtb{*)TP6(@{Je)M#9G?@2OESMlLwx
zRLc|x8zflyORv2VOwRf@0zKLYGW}?Wp*@9mnIoT}0Z74Tevs%Brtv{0HVhcS$^a%B
z7fV4pLjrw2ARpHY2^cv-g4XQ=g4U<xLg$Tu3mt;G2LnIn%MDr-FyfEpCR3;CIUSqi
zmW<68fWr^69?2CLDV5f3rCfOsh_Sj-LK(RQVS`Le4$R)eV#;I-j4+SbtcLj#-gkGo
zx{_;#AmJ+Pv(G+B0!G5(X;3x{7$LsMug*K0%L#<LxUauW25AruY&&l&ZDJ*;^N7Io
zmkVGFz7JlOl&6CUjNs~m{^M7?s;bhuad6KQ%cBY75@U?e@mzorf}8rT9H_@>5G+u9
zp@Zc5AxH#N$NGjfs9f$Pes%u2X55%D3E6Pdz(zW3tf$L`TF`Gff(O>PptVyrzzFLn
z+NVKa+3y$vqp!aE&b;;R`)2X7Ws$(3U#8H{Q*C?;z3(Ia8~723Ta3skkzndQ3cOxL
z!1TUYB4u)KMLOVu3sxSp3W-@(v@`%L$nBNBF_g8WZ-KSIklQQVT2+l3t#~0Vrb_WT
zaI;1J8pRWVngxJh=|P=!<7*8TBv^>xGRqObM<@Oh;BmC}>BWr46LEpp0XYCYpl<me
zFTqkn`2j=9<VK5FDb{6{+~YvKXvj)=!~!?^#uf>5vFWNCvT;vj9aAn~r9<w*dX%SE
z@xo}QNE(YJayij0fES>oyQ@=xkF0A1lF+WR96?xGEU=_U*CFDLlu39x0*u(<$mq}d
z2#j#!2)Vc3a;_n;_(qBe<^<_cgMxMU4FV6^_3d;7A3)rP(Xt5k>BpabvaWLdv|P_=
zgP0bUca%#L&gnb<w%vN3_B<1ks?HZM0*~{L7hjk(HVE@S{bHTM260x(<h~A?zt!9%
zS5wwC>}6Mgeda__z(|;7JmUpV4e*6)P)AsOKd-%@b4}&a!;V;u;Qkg@#qf845%72L
z?Zghm>e;7UfDw2wtoqqMW<t!*D9d(N9!~_{FIps*v>%vn<zg(%m9ooQXvZeSi>j2X
zo(6q~#s$W=(E6Hdf5I$BQt^k*CHGJOBkAsO#9}1f2o3%9d<9=`XYJGa*)DlSiT0IE
ziY*f5N^9uvdZY*a9f8G*mYHdrObM+fQwzGajSZ$&-_G^bl~(BBj3g$Ezz8ssFdPH>
zBMKM^^X*ShFg3;{04hP%zD<7#CYwzVZjpRymMfh<U3HD0B<;^w@O2m(-*2p+iFgY5
z*bfRP!+;S?tFKg`V&AOW)m7-g$gJrEiat7FbM>`^pLg8<kh=vfp2RjdCBb3?pb5c!
zn4mviFvF85V8jWTT{ZaSC0<65*c9J%_dUpyWbJUKsgY1-_1d*|V-w8l|3*N*!?yFy
zEV;@WJEksFTi(y_pXUOMTqARS-=&+YFA5kzOYrm0KkXMV!p8i9Gfs0&<$MgfXUMjq
zd;lZC`4H>_KCE{X@Il;;aalgkT!4}5ZwKi5G1v|Jo6UbPh&>lvz<Yp^3JErTchLpX
zpiFfA5L#i1MAj&P5uTAx*#IMy8!LbYw-RVSrXt$Ijop)S^|nI3N{2J0+V~FV`x@zc
zenz09tKD?TB5Q3^mjwd=B{~2l^bP=45Ran=z(D{2F+e&677)h+ms1d6gbRv}+bhao
zkr9viLC|B>1r=#y%XQ+DRLV-Exg{2S-PWc0aoy#DyW1Ea0x7EGa;rsP31A2o8U#v*
zDwW>Vp%CDzJKi4&@317nQl&&fMBIH5>x97Zt&K6u5aOm-S736FW$RLTKpm_^0B`6k
z{RD7<5D0QF5MYXnc~UG;rMsiu22}_6uofiB!+D8>w16YTzA3G&(LKjpVREfpYvPD2
zEjyeU{aIgu5qI+d_OoY9H+;*cO`Vb;Om}(Qk$7TQ9yr4K6{V1x>csCEmKX3={UMYY
zP_mdcBgSj^LjH91<_j3Xsux$i=|H1&{pry={U5S`4(@f*m2=m#0Y+SzL-ya-?5rU6
zKB3@o=K~KbE=jV&4)@4j4dtxOrcJZ%vNFBzT!0ajOs>$LF3RPZ-*0ygNOLBZ?E#A%
zcilhIrzf9#-VB%cZ1Afsd<*)%FSv~iS~L6iH<0H(fU%|z2p#KzHOp2vm?=^=Af5u$
zpTOwwhuu}x)`gb8|DX@u^88REg}OooMrsv&eS`{k1ePpaVWv)<V6EwL?IrIez(}!p
zuWr1cNa>}ux5U(z$tqcZ@ld^^x!#Xo4z8Gh_d{(p?)PbZIbSX1iP75U7vkErzKn69
z-^U0rGKfzo3K$8qOrTpIfBuERbyI(djR~8CVBkIc{^I7!5eqGWs;c`KnvEw<Hs>gy
zagg5s_!IN!(|=FWieBG<5ylNoae^8;0`%bC(QMbQ%lg&Dosho~;oAMue$Q3Q%(cZa
z3K;S8aE?|Ou8^jCeRX+=g8;yG(OGAt1xC1!U3c2S#+E^|&nGn3aLGUnj%I0#Z9ZeV
zv;mXj1PvxIQaNgLf}iVbEJik-qWLWsNkKxh60NvQ30HIo6WF}wLiqqjqT1zoA~?Yf
z1=@r@+2<)&2FvAcgU#-2fhK@pLCOUf!L=AJ#=;0IEY`drJ^uIS%-ipOU@b-}1>pU!
zi+^D)o&0>fe?gGLT}Oyfto!vzNwP+|Q}-tUsYlrYBNPQJJnX>z&88AU`0u+{ZTyt^
z6JICdsl)4J1Q;1!&ug%I3owF4b(@J<X#i3H3UtXr0}zAXSZqL8aR4G%w@`}tVbeOb
zUg`x!=txj@^1B!v#2tY~sr*`NDS~B)zzke+=~`7yg#|1CK>#KyRcE6B4C`{LO>VOE
zt-?)JSvA4n6;NIM>X7R#2mmK$#o`5dSG`codPP7_U)&JD5y;l3wy+{`=ky(`5C9(l
zpi;SL=$7S6JWh$$A{R)WIx6G_i*E!jx47pD^_#L?a>-RSR&hwyG#LW5TWb<2rgXK%
zz+B>vKt<ghX8iau1>s5yUJ)l|gF(5jL+d**;*b4=F{8{8ha74FlCW<R+WmjJ@)|qu
zD2Me1dxD?^0oKJgrwGRFw;4RAU2E~M=k7L+nSUJ%fCk4qir1L2>83W|xSxmfd;ud^
zGUN6%I2KrTw)VyMp}jDkfpmy#ysuZ>aC0mnhrhW1Bka4*I{CyTWh4|WuD|V0gC)2l
zc)El4-P>9+1C;rMtK>COf+D^nJT0;Mg6^zx0Y->TcjZml0VDJaC6`0?+sAA%dsfnO
z@%zm=yy<g_;YI-VbOeaZ>OENNQLsM<U~!|>B|xN77J-FAVQaJxoMh6^RDpU^t;>aM
zF+#tKYHG4_dzNY(G8xAWH4o!fUL?LROI{&guC%(P%mPT)Q&wStk!}GZ<>K{*J1~;%
zeDZ<IY|CzF`B$x3Ba6%05nv>X2OsvW$)b3{W#t<&;!10f9!nH3f_t8$4ml{P7t59}
zSL}j2lhU(IC=tVqwLnnli7=GSl24Aw11&`Wc&4<ZgM=%or=Nc@zR7ae*Ee8<`vOe*
z`7uW&1n5Snu~z<Y5yGOG=5M};k#Mi#Uu0@#?#3?am?<zqc{YP=s4%7XiOn4rK6~%B
zYg%Ar8*QGBI^<y8lV5|e)-R7f{frf&yDRyMF~WEjlEugy@4RQ8dj17>n;9LM1KdY0
zln-Ep_SqQI2OppzaQHEY9g;9d=_6|kf_pZ`j`A5}<d%EnN^8+#S9QX1Cuu_-y8pfj
zY5Z`pi&%pyq+$5G3oo=jxBZL=2Z%8eMyWPx!Th-ZBh-Pr3*4COC~GQwf(L~WemiJX
z-gMX9RxshG50~?7#&fvbQ-Ax}pWt!*ZM)FQJ7A=tSymyeeSO~yzzV2Ru|W|0$JQiW
z?*oo2<esWs?y0!mB_N?&Kmp~gm4-kIOIpv$wW-ww(;=XP_##+q0GM>EPF!e}$c0r`
zt73)-B!PGrJ7l3j`v4)%rB=DFJ5Yo+I{<Ff7u!$u*#%f9a6D@#fz)HJyX4YOeeF<R
zmD%<m*IEz&4d6<pT!#UM5O5xMV1PRSBgFzZxF^~HBuXGYt`ZbOJ$)7-<Si<ZcB9;7
zb=I#Hpdx@of!ZMeo)T^BRkaimWHD0}b0uakZUBFBU5ADnFj75cBQtg4*pyE+iqBw-
zM%9h8zyxk8!@f(GUVDR$<3IygjQsk%vrTotVuWDdH^|yE9fb)B=WagR9CyTFNo}J@
z^QZ!e`xc}F9b*I+VZS*q*<$3AKBXmhe_?YAmm3{=XL;XxWocT|5AE_?fD!f_zdZM>
zq}X@#8B28Ek|HVw-zQS4d*;Q33WgqwErXKN#pmQF#z?Nf2>r($?VMS&%ud^FV{ieS
z%9;pp<gSPQW?gGhwou9KnZ+DRMW8tDGIpr@8;t3h@BW0nJFX2kCWWYVjVrC8v>3@2
zoN{BjAsv^ltT$68j227T$5iHU^UOgkMvSa;N7sx}e7UuzO~7Tf)}X2ljEBl$_%_%g
zF^iEZ0VByVu{PL28NwUKy0Jf&jk$E4wPFMq$z|knC=&&Yj2k<~TrA+i7qC-3+R=Zw
z;_A@&9N9FF0ROzxPf3bnLOc)L%)o#;@);TeYa{#)5<$IRdhK;9@bIs#kH84`$0l;O
zop&;OF4#Hg8F9T^!jDF|3?m5hn{qSO*%>oa=30L#3K(ha=`vLz+DS2B<DjH|qCEyc
zhGj{VD=p?9?iGkX;S<_^xI-hBhEKyC7>SB8@{s~TKKS^P{;j4xj}lm<EvO9n07kan
zatpJ&<|sf100Tsy*bE2bbnrZ{zUfvw4{XN1Klubk?$(C(i?6@&tB;@K>v}(Fl?Le^
z0g419WetjoF+zaOD{j0cVa;;&q}r&tJnon)FyhKHwzx0B`elv+)(7LxxO|`Keh5UL
zD`@8G8H%GD&*cjGJgKgKF3pCb|08V<><AEJgy8Ev65@9VtiT$iRKP{2EHpX=SX4`h
z-zF=KVu6VczyL%*y%!J!3ku?YIJZ?C0VKKvGFYn<9c#wbh!2v>D}g5fBi5ak%7MMK
z5M*8N;i$1Oi;6}Cb_bXM@W84=7h?ke=pdk^SM~U@Mto}!fDhj3(n?u<sQ*xjtWa=;
zMST|BpixJv69S;Y#TNMhN!seQcGXr{Acx@d9ZhX9utetsc1K{{DL}=>G*KM@C_Qq`
zMf?*09PYDy^=b=@5LcvIF%Wut1>)#G#&-Ppu?1l<;)Y=`j&2>wwyog?j704Z;171#
zW*ak27OwBg`uF3{K2I77e~hBOuUFrEn<0K06cu9x#e-{ZxjjL-BB`;y1Yb*#cjtoB
zPgOAY#JC~sVXnLV&LlUf0}YHs#TX%o`K|ZfAHOAk+*~5}C79E60%{HcFv7F>2e}ai
zR0$HwEbo5!kxd3bfxin{fuQk<LY*Jiu~2{!cR$P*=E`>S=9w+$Y#!;N6n+-#<3i@2
z+@85b^Ja7b3Dd%Wk$ei}!`>bDd<)%Tq(5M%yT`#eb_k@!;$omF4Co%{HdED~Phg}-
z3MRdZ0Yr=uEJn(t)G`8$q#FJ8|5_>wnvvk^>;LmlcZYo2Mu3rYqm-xXQ5zqsKU13i
zn@WQwNLU{rT<b3#cjVz_D>0xxVOD;{jW-!wYB_RD9-1E8$dcvg5K9)^d*S9QNb4go
z0t1ifHzwe0MBe}C6N6SAChTnb0GC#4BSI{Xfe08F28<w*0Ywe-k%{2tsDlr*X7j=4
z<9<g0BXt5qaIX-g?-pq@dC$7dq`Mv!8)q<GX$2YFY+Pxvq5SQI=Ot|{g8@hf8nMm?
zFoMMymS_IPKg!QG?lQ3G2yQk{J^p8*0mgH2rRA(Y++6XGhwi_h)fn{CIKNgK;2-6F
z%dJ6Yo%C}vW3x>YvJtBY0S`e`gOqBc7R;M1FoI?y024t>@x$&@jr@`BD8bc&9k#PB
zO8je$O-<&q>u$7bg`Ykc=LmN&V$YS_c(O^EIw9G`TCVkulv_VW0AQp=-#<K5Il!W=
zvDLb|>J<QD<AumYR;k=lSyx$r696M3Kmr%Ay>gT+NR*FjHDhZ`yDUU7(CBQ{H?FQq
zZeU`dOGh7|5D&h-CI*Dm2#|o91wv2<fCrW(rE0H3b(AZ<2}jbf@^FBObDL!g>pO3$
z4@8Om(0%pw*Zrp7+zYW$Vs5*na3D=#2wnlpmUaz@bt%TZ_vN?143;r{)axmukmO>m
zRE?8n<MLG&3?bf#PGaDQgl-NP8L=2~qnP@*J^&-J*j6XR-Qc2>^x~_pCEd7NtGU3p
z>Ysl33-PFZ)>nPGFTegqxmJpOH<_#F$xZI$Ns|)l!c8R>neesQ5J1E~h?hd~#-}CA
zel$1XX0#7Tbl0%Lo-=Ev`x!sx%as=Q_YW7HpA<ud@9`zqTp#Cs{Bd)gaITq}iT@*E
zT9C(d`7$}?VlhJBfB%aMk}SVz?}=xhvjNIE$HnhyCmhFV0=@c{;>o<YFd?m6ye`Uj
zqiPea%?8Vr7W)$R7X*c8Kj9PZ+;H;+pcjnc#vb*OYp%=6wXI+GV4TCRc9rP+A&bAk
z)@;r8{!$49v#l%a-Enm|9=;GQMpE_L^?&3zv?H*hsl`l?f=PcZMmof&iJ>%I0xZiF
zyRAz~Jmv6s0-Y<162DbLd&e6_>%KBYq2BSz)oY|2Tq9m4W-&sa*&Xy}E*9#&4rRac
z`7O%#eejJKG+^Y`C7<^G*I_ZA{&RQOx!?CP+%`T_X3dCXoaV{p&Cd=$G-*tJShB=i
ze)}EL;EHJ|)}fu5MvGv=_;GUcR2<h1qtb1+n{STVfB%H~sS{T(Y+j_dX|Hw37jc=|
zfT2OF8)nrfv^g)i=_Um=jhSfs*RV>NGI64xBvbBK@b#6g&8AwLCQOap;+Splbt*pf
z1PO2DpV%jS?*37hHSG$rmCBdpY-x6*6(6LhUwF~H{`R{DZ5lRkzShsYt+zBsA8>%(
zY<~UicjnFqAGVvSOkQev8xIYRmC7B9tJi7v_`t}U(iFS<vB#`OlIEXQo3RQ`1|yF3
zh@)=xf0E*EoFNxz6XaSej8+QB`OmAbHDzMvtBY}05~InJCYtk4JH`EsABibLU~Ffl
zLf<O2`TpH`=Nhaof&?(qu0ECNX$0xuefKfDNPz4U0{VQp(%M0FAFeinbeEpnm*0Gw
zkc=f2uHw-A_X+D7EM$DZkn5Rf&b#O5*O%_RtftER%S9I^Wvg#&G?!`<pN`l_UwFo8
zW}^w??LBg@5U!rimlmfkC(wtoh<PZ)A_nb4+_4480VDN#zo=@o)(1z#*C8fNkY0Y{
zP3t2%n9gs4aBezn<Ah7Nu=&r+u8_8Prv*kz)y4s;d#_y=B;@1AZ4W(cKKT4|GfIq0
zBk!!rLZ$K@PB6%FX^q45GA_(5_b%B?`ol9iXqN>BHxxm-@&5bG;_nv4=Q`JG6z}BN
z!w)mlH<@Mws-ukwC>o@x@(q5fEbM$uPuyw!S?`*&%3HQdCgS6h-x*hzTO0Wu?6=F4
z=@9`>Hq2bGy-YWd%Nv@_Br%eX+Vt)ReQo2Oq)X2!OmINERa#Fn0daD9zg)k9$6V@4
z1rmj6Xa?JKzxxBW!U-@>hBEMCquwEzH&|<rHERXI#a=;sdohpiI#oxpsIZnAeK!s`
z;ojHyb@3rsn?Ss;xT$hPc_<Fz>XOS#tPU0=^5CMSo0ug~V7;?_=&RG6yS0Hn@R{Pb
zEvh5^Ly%|tNtJ!|g|+MMjr+3wR$iM%nZDbw*K}X}ww5P-_0gUS$j-TS?ZrtSS$ALJ
z<FFW^a79_S*7(uZl~yiu*VUWp(P@#w8wS7S!L=gOI&;a_(6Gjg9zE8${$yImaN8e*
zDNY@K)DdRO&1c7N8H-9RD}xkVtIs|4<OKJxSTFt0U#~Q)#fS6Vgbv<!FY#iD-){Kd
z>u$Zn#*quQ;s54+>DIh#()SvE8Rd>&pMP#rjIYnW{L0)bi^(8iT{3CnM#1E~#TcRD
zDDZ`tAy`#F!Eqyv^{Yv1Tazr(Mwj<S0wbBc1MmWjOQnDbp8%l#di_lXMUNl>6#V9b
zbCVQy$j4YRrqH}C=9r(!by<*}c=maLC+{aD*U7cmg=e0gROiy4elpkVc}LM%;*SaJ
z?h~%GF2CtUYwhRKQ+>~{xIg5;{mj-|Y++-aaNl7u{MbtX^o*%nXR(g5*(LI7kPp|y
z>|fydEAp&?4+zkvHNCa1+e}uhAGc3(V>8glsCgJvjNzNX{X2!auyX0XEB^K1y8}<s
zr1{LZxNU8tp5rJhN67U4Q|*g0ts|3sjryLHDBg{~_syhSzVc-I9m{oG{C3<qYa8>k
zxuw&L7x3WAOS0i#gJTh#Ye5-Rs`VjQ7r&P-Uu{N>AEj^AI{8FczEo+vmyKfHO7)E@
zk$dspmL}8It@v~mV)$f1;=Y5a>TuVh`rn^1b9ILu{r4>FI;^Fl)g|qj821kGukIZW
zD?89>QDc=$yZ)NRxu$P|{2^VPZr+nGY`qIBzjDP&Q>%5nQtoS+*T~c|wc5({T@C98
zX@mVOq0Lm!FWtPPe2{nsY`>~QJlyzs{b%K;r#es9pKR*n-DwlQ0Y8ho#ieFis;0BP
zHfL0m`RtuHZMlW-zZEZEbk|*%_5S<6Uu_dM**wvDA)YIXlV!`ogZpMt#h_-LG9?i-
z@Eb?V>lf#oW2SAKC@>_?eGflu-g@_4KYQf)*fC?xd1szszWU}H^WY=vHc;av@ciAy
z7bp3xedNg}&C9RFG|*UVyT#a)ijigGbKV(em{~JsB($}7=~8pWb=R#EZg^LJbK!+%
z>XbwwbG_Uz{qge4wE@tg7N85yI@`>UyDpz#fdBi?e=%+H3mHadopP#~8?byq==eXE
zTx#kY8(24;93!dt+|B2hvrbJG{6G5SQ)c07uO{UmEr8)4etB^`0QBo`y=5MHl%VRg
zGB~uc#`D)2SLp5c-ZS?;^pHJIb)RIIJ!7W5*PsR;*H+3^#~&`el+y&7JaLld$S)Go
z_~EzfF4k7pnE(C1ejn;1_+S6+-$LgHWsf}cP_xH^L}52VuS>4FO4b6)UG?$fLHq4z
z_TPK&_-{_gf7J~)n6EX*33}+snap0h?QRY~_@IRR{J2Tr;%8rek(3^l-+wP$q2QSB
zu<f?CFYkWvfy<K9u}<@hu?!Mc5&!w;KL?YOe%FnzGynX%-&$ebN1uLbZoB(#n`hES
z6YcYKLO)lK{@(q-1Ma6i*64Zu%kTfe;Bp~I|M$QDdt9&|ymt7(2W!lC3nt4C_7d~g
zYp+d8uMk7}fBn-xCapt1NQ>}_>#w)@yZY83){=|QJvT`U^%H?2*WYrRSQZ4T3WhA9
zkCrP_EJ2M#n)aDYmD`b>w%bm=&7Zcv*{pSW8r%EqzMDBlU`LRyzwLJOv4A54sEiwR
zpC~YO*PV8<Sy`X{L+ewE;xlO)+D)Jp=JG>y!qG>Y-90dQ74Mo|{|E*ho2jZ=?)PH(
z=@+T|vf#J!AP_B)pt!F}s<lKrwoiS6TM81}fNuw7smt#R!d;p~DCWwRs$SYIQC`M_
z`$Pat90My7>bocEQ?2$}(AT`}BYlPFZ`irLn8Bae-LK1-s;>UFwXT)napvyV9qF}y
z9(5O~@`PO@4RT}8c`%>*?H`$!C`j12AF`+r_Gg+}{aUDLgz`NC8;w=4_-?u8vhS*%
zmDamrX({t=`-HWMrM_gn;`X_c4c^7ALp(9kxDJt>->ey7N1I81+;x5j&uyCixVrf5
z6YGv$qr61g3KI1q#Ix(9gw(|nkdVf2*QU$ktPxzFsf$}${^N>ZwFUiqvi1ku3-`gj
zbN)}i{#Akt+%{Qq{KudEY%R68e@E5=Npr!C&Htls>?mpHJNn}5uO$G#)z<0Kw%Kw^
zbK1{;t}<Rly7=^S&zWame9>jIM;?*OMg9Z!-p9ae{P<IhXCtuo|Ni~&Ef3-s^!X8&
zUM=N?fBNlj;(WDB=lE~``Cq!XRy%)Od+z!{`|fM@-)A59Gk&~Md;?Yp!Tj6JpKne&
z?znh9ck;r^FPkTxdB$`x(Aqn&-wD!|bG9&}1SY)l>Z_sGgZbTWf1F6Gs;n~q{13lN
z3Uc2lrI-J?^s>->5qm)R9=_i`vCsL#WtYoZcCAg`ddsc!{7&?<*z<=Td)&PC<{Js=
zo6nqO&N=<Gq-VWM*6){!_om-r#GZ7sfFc@JH-EbPukj%E2#!hubI6SY$GttCoO`(R
z>Z|2iE&ScwbJyL>5r-Tc?+TVn|Ec-j9<d*cXB?mtcOR*$hhK&7WCFrpQHJ+7NW9N3
z-SK1OA)WE->f_7&eAL4Da<sgk&K<e+ku{a;T%oMXLKob*E91K5(*55qzx|%5EVaT1
z!jG0K_7`if<+Y_yM?5Wb8G7aJF~)yOIe`Qb{c&;x*KR(JnLLY3+TnQ>X`jQG=c~?4
z+TBp|_+#wOgU8_5@GdYPnd8juT!`n46<4=hv_ecz*%RA(`CTSU-(2gayd&-fa&@J<
zw^;7N`Rn@YdKCQa(z87V$0ghH1D!v3pRT{GUAXV9s7zF5NY|fWed+QK<@Mpz*Wfq}
zRNwu2Sv!`O%?LK#uiqP+#b(i$U)nnU^<OWF*Hu*C+upm^wlNEY4bfsGOd~8lIy#v!
zZk#z+=Ixk0`-IlaWAZKgk$elswJnvwp=E4#i61d@v(3zr5(;CE{Q5iZ+5qXU9+)cJ
zVqvoE)6yT8TT}N{+KiOx%h4&pU#-mICr_L#ZSzwSgpx2!k398^!Ga)2)c@=A&XKwF
zy3H7Q(3rmTx*M#fP_Rsx#R!_%f4=J4NK=02GRUuse%<%TV{uI?|C)au73BWq*QMS5
z@+*E;dp>){baV30j)|wk)PMWKBI`D*M?wjoxYnwB#fS4xLY!AhJ0T3UPld+%^|^C2
z>(Q`>{J@c!Aj$jfA&Z{{vG>i@$1%#4R*eF2;(mZ<=@V8<H{W%ypJevhV^_0J0N`*L
z&`jjHhtaXJiXtGmPb^3`2ZBnv>-$Q8z2~mGxS#Rkoew-@-v8h|Q(RMPW34!CfGAg5
zSV&+g5u|Ea16(9cI_}*kH2SW)>n^io<*Eb=BA18wJ3o_Py`ibmx^{Buxg5uj8*AMh
z1j}Y>F|ymvJDLOb+S3;ItwCy}CG$Hy&)^0d3$YvJ-okEx`WoGPw_UY?O$4l5f3U^^
z5D1&QjKRFQb6nQ=k-Bfb``&~=p<LW~`|Ye-x_HLElc%4T)t1&QsHL?@z)6L*Ttn+~
zb6I-%L`<KDo_NaoT1Ru;Y5ueb<;U#>8i#1wvYuWew_D8jAca}Ok>_E7q#u0zX+rih
zPCVXhCTkH#k3Rj3{Ct1xekL4)^%4a2{zo5A$g@_P{4oNE3X%Nr3~B@pM1S2ktUq<P
zOCzNs3$yrqwAr8j7JAuC$A{qSZS5Tf+rZuqR`J}avq!GCBxDJrc3rFRu660#TiOQ>
zgyknKSSBi;pPs4@e!g7J^HF!*s8X|L^;&_la#z;34F@nFP&<UMr%Z0cSnqLZW?hAe
zFA@4_(>w8b=99kjWpef9KKI0X=@tJ~ta{)nty?m=q5uE`fOTYp%uD!%%d&1g{~`$(
z$cwwMa&0Ed%Zg3o+Ey!-!%Z4}aa5}7-5mfOC8n~v+%!uozgXpZl;+%x@f#r<F3X_G
zl5W#pL_dqHTeP-@RvVwDL;^Yp;a!K^uT|HSn@)kiW$JI!nikWtW_T%1O`0{s^7kF>
z-4e!YEOj3zA91AFcB{GeKHZ;ZUwPF$Dd9eXy?VKmE3d9F6<ux8h>VTfoSD<jug^Hu
zWw+n?%Wl5Ie7j_ko<*A-dmj3@s$%g67n$+;PWpu48~&iaA=xUNxy3hp|GoB*mGwL3
z*_U3nMcK3bUc3e_W_`j{)Afom(xmUS(-<Tc8s)?2(rd3zXcwMi&C-=-+{P2k@dq8E
z;PUg6ibcgULty~-QN9HNg|X6r-=uG0-?)vXRB__bN80)J|Ni%X3t$?P)Mlm$07k!*
zC2nwBQdx`~y8pgX>ex1^iO&UY-TUymcx41HKT9r{gY@RR@0ll`dm)~TaPj9-?qEL?
zB$T9xDHCP^%lpRo5d+gF+*{$6ty#(;KKcBiZ)JAGK?f=p+q`(zSKfHbJi9P9Z{mr4
zC);g3PkYz{<H_t#Z@A;G1YR|MG33PU141s@q%dpNsJxk`qnX~V^{QL%hP7h$cv9-H
z3Orh`ecmcqH|y<^#Xp$|8;`eZSPkoX-x}|}tZr8p{2i>z{SIpFyL67!v9`V4jFokg
zBlyxf@uj4XvFV8|x%O8JtQspzsD_5M+V96K%;`T?R-@#>0BX~i066-tv{LoQ^&#r$
z&#kCDVd?&SaOYScc2`s<+&g9St?eDwpOLi8Mxaqv9{sTvp_TgPM)3qSrbXYgN*y}{
zM$2T*l!{--M~sndw9zxW>-8fS`v-{cPO13Bpv$sc-{rv?otXRquQ`cw*s%7)_DXJ@
z@$(1sF?Rr`iDTw_sPW6V^I%!hX#)_PvhF(GGgyDqd8KgmraA^Kz*6NM>TB7)hi<%5
z^&7r$&|GPa03*o)#Nfk02zcyahgx6u!E3}1Ksb(I6Q&dmVf?lt0Nifttz<es*)sF6
zGk4zqp!xc{?-QyZ=D^uvq&Yf5fcf0QSIo0-yk<J2L0k$`S2IeC^f3nEuTRv+CM1lB
z#vKtcCkPxEB;px-zi4qV*@OWjWV-#n2hBI%C-&VH-f`P)%wY%Y=YGbI_)dQ6`4=oO
zft3XD0MH`qblL*y7jag$m^CwA#+@uvTm$;(2P@>-#0eA31=?T+fdW@9>d2-C@aUB{
z-x45{D2(9QG$<HYodmy`(1IPeHwWz#vzS;RAw7XC0bzpVUUvVZ!)EM636^Kd7qL%K
zzzBJ`UtBQx#2f*L_)gD|=R02;b^1hlm_-VCzmnUrAXN!)J74wVXE;bNXtV#)Yl&KJ
z<cR`C7QX(5dGv|L<8BVzLLLQ-y!HME=83;Q7i`x4p1<W>bJU>+$F(Ws%fA2sKmbWZ
zK~%Y1rZuwIz*jTx9pEJ3Bk#@*+iY#No4>WSt{`3q0DvR+9Nh0rj?*^X%$$9~u}N9e
z0V4=0c<<Q6Gqw<cISJ6|=D~TV$-SLir1`{i<30Z7`ycG{#El;=&e$%6Jx?ci2!3z@
zlbA2W-tlR<v;h&eU{&$Z<4>7ovIJt@<BICo!wyM`!@>J<<&C!}kh9_yXmeGnjW?D+
zY^nf_00x7=6`OPBzz3gvYC%`t(Qdt0JMXxY+=FayU3=iJ;i{W%NpgusoRBk4JVD@Z
z;ymXba8vQXV^1Wkal{uw7~}|n5$?)dAX}G?5QM&t9aCqM0I_eo?}4PXkIgye&W<6%
zMHwy<3yp$8qih+FOzwxxa+w6TPSBf4nf~S(i2LYop9ZtM6V&v#O=67LAnMWp?=NNJ
zk|2{cHj_O4?Q=He^YM;1w6~b{`exJGqF5^Wk9P)(xgu?bigYZMfV@jV;LB<%B?M@(
zY4nkR^tkXMy;Eh;Z6y8+00r8>`W%2RkX3mqOlPwK&I|OwYOPEBODtB5v@gk{b=JBg
zQyHC_DghrPw>QUv%oA^=v$Zc^y#x>t!Rc*m7Yc|fY86w4p!DKjUEun1rFAKm4BsZ}
zlB};krBxOB9<-Rw(n<;Ri4h}EMZimi;?}_bwluX1EGf6zpj}<Prn9|M!RgCwoR_8*
z%WQkY<<B@NHsRj1slL^eNJG0)AlCv}r5^8T&~q>FB2O<|XkL(jxk*5W9;c}-?lKLn
zy=D{f0~eifQi3phq1?pYasS`!9AGY3(zB=*;Iz|DJDL-ZI3h`T0?+_{nxnN~)N`Gx
zcq9ny*`x5i_2XSo=b0zTMP|U_o#%;Sz?WZtW9RCYb7q@E_uJQ&592k$;uj*Y9A8~)
zmBxs<5+rzScphJA%;lM>7-c7EFT-5+33pI`e`%rl;QjXm4%S%S4kZ}`^E>Mv0eWz^
zg=KYhZC%_AN4D=VdnlBvf_sP{Fan5!%c<?fTkazz6@L3f{+sT)S21GN#*^XsaDQ4Y
z1sI>m%N~mU6Gv|^DHgGB;#mai#ck}hx85F)J>%Cib@C+3Ck12GkjL?ae*UQQUBo*b
zEFQ`cd>OGb*-!X{3$c@acC_a3RGUnHACVFYaUFe%TH8r)ScZ$?@u4m|5&UD(0ialw
z#;u3nUb^!J1h5sAroX}0S4#P4Lf~1ZYdh6-+?VFtZJ6I)kU#VVf`uGo1UD(Q{Qx`t
zF&H@?oDoO`Dn|8yIoJUoqy1Hb^farAIxSFwML>u4mcC2qs5){@_krPD>qu43vTWi3
z!)#un*2h%!am{~!UEcoQN3MNw?GMG#t>-?dn`u8<dNjv8_w;_bT*or(x%l-4&kJSD
z*`q)xaYv0wu<wQN+q%?}_n{w)66L1-?OUWPtr1`(Jut%Hqp^f0c2LtP_+VfhgA6x$
z5Ec^rIE>i1;Fs19!iUBkK*XXYONByqOQU~F3CAZS6i0Ih-^OT@#Yqe%LggtEFwWaz
z3qug%Fk%z((Py8TWvasu+8zarz}SEE>F3ru<LmFfOQ;e7=`YScBQ6L|xaj?0lk}IX
zud~)Ovt~@UL8Yf}5<Srg`uw>zZ$B(vY`&MU8$$4Wj0Cy(^38?5S6z2M^hkoW3`+$c
zc_$PNtCygZ0A2{35yCs6H_w|m8GuZHQ;cQ0Aej><jJE)ktAqAZ0V4#OKSA2|!Ffe{
zlw<tah%r|HeohLB*~0Tf>z@80Bxl^wxSzZEY_pN#--Hoh&&QvAVPiO8Il#R_QNRfK
zaS63TfwJED;G=kzQNRdR1|NU+xfRmC`EH{BW$ty4U3QX7C~Zjf%!9-jV*yhxfqJzz
zqi$^I<73Y}Yr%+MnRLGiDn3C%f1b+@+ssb_7bwp?y15vv4-pd^`HjYpPuT3fDc{mh
zKKpz;J9CNW$z1jc_e!hRt}!(dYElmY#Q_?4p6>o!e%{{?KKaCa^7)q%l>B4`Bh$2b
z-AXR)90=g@a=b~}m{T^MY7X9W5BnUM4~&%$$h!Q@tB0O=($?+Lf}5(D5-uDg;3^1?
zke~QM&8ojdCezhbl?h|<sX$~b$sq2DwGRudEoRRYD3S;k*-U+Z0K&fg;Rg#s1GsLc
zGN&AOYyxQ1+S+E``}kwSx=7s<^&GaDw`Keux!>3NG-};kc5S4JQ{nOq)e+ymwc~5;
z`V`#w=DSY_yjDvac;ji4tdKX~^^erozY%cYYE&&<`i|kIpdt^zNWQf-v)xQY-DAGe
z(~V^&?GS%y^@^3U(upZXu<<CCfDfyYUffMd+ZT%uEJv*6hpasSMBF)TAmlEq1|Vs-
z!gs(1g4~x%5D$2P^+%aNhYo=w02m$0XUpkxTv=OTEe=X`9Z*FJpDhOvQYHWc>Jl)4
zHAuIBYb&uQQEQ>KDp6k`t^<CQl$Thm5UfZnaMJgU!de9PES6*)R8=a!=(2{u8V0ME
zSb5!t01&_tEJvELP|<jl>Kk1n%aIn{Q~lCJ!E3C-!u>lz-#OeHb_g_r7#A#YetrJw
zW{Vj!k~;p)qDAJ5Z@xD_tzISTk4iIf(iF49{H@~ODR|(8Z@g{pdGtvO8USM9B7(8)
z((?ltDN(s&4mijxP@FI{JbmK3g#s3QNVWJr1ow_m4#29!-w*JeK1qr=LE$_7fOmw~
z;Tw*k#Pct|CSK#HB)$Y|DE2RWyO$~-e9|<z)C5q0*9jx~iP9Fl`$ytA8v>Rt@K-2)
zlxp4u3BC^g4gjG_W8w2Y&%Z3ooM)bjE8kEy69FB{Dk!5=NYKCCym_JfIKVC3jarwL
zn)mFTP^<tP2+~hdh<Z@U8bL)2^6j(xuAy<*Q1ap#jg|#J_x8>QAIb$+y)1S13|4RJ
zMIi*fn6bd38&G6R@ptqojIbd7NEX<%0e^?WUXagm<*+z8=aiG;`xw$YJ3H-jUL?Mg
z`9*9y6s}P6bA+`mmce0)crH&iM}QG1-`2X0U+ph2lFhvoLf%yTp&OH2go^`xZ&AQV
zwxH)g_j9EF{*8bGMvUJ3ww@AG(bKN)SdHabyQQo#oP@Rg8;Oz9*^EH5JXTalak)^|
zu1t82Z0>Kc^0Gf!{KHZ+e%i#mzH2BpqDq-*+?P!5*Z07Kl67-KePIA2!2qtA+|2qA
zTruiCVP?_tAy74&37=t)D$}e@#cOZ9ZC%oYX-~OwFscVj<DX4Oa1-P5xFZ@xA8Mn*
zCLtIw4`C}B<53$HSBE<;QhX9zGC2l13K((uVZ0v<2@(zCe#F5ChQ>p1Wnrc-x#oJa
zl|T(NzB46)781PB5mp!bYV)z9H21<bRxV%E@r74jlO@FKQI|v0Pd)xv>lQ8aVqYTn
zcdddZu2{7yRW^H#jUiSxZVsgaMkoNdfR<U%rJHLAO)oV2*hKwSK_Y_|11=vLhi|;|
zo;mEm{mqW*LnefI;D23mxz&m!e-tp{>RPLf*=2IS;3x_hao3r{|8x13?q~dnOkj)I
zvu%4=DY&tUX9%2d@3>&|L7YIAls}I-^bordOqnT%i?8ba$I8X0U3c6;pw9xkLH5(#
zIRGBdvqi%G-<*F=<YxHM=btz4zVp7dPys*-qqT}DgqsBflSj!P_yV;Z1~yTaF?{5S
zC(P#(?)Vh8sls~efW7y$RwwD!EaD%0^2Ha{l^|_ElsRahy+U17_-*^=_x<fr^Qi;|
z5NpnXDjzl1{c?aN3*UId3K~+a!+x1UIrp_qQf++m)n8}nm9P3X<a%JB8&@`(DU-(y
z3ee;OBe`lr4b=1U=kY+bG1&D2jI3ExZ`%663ff1JrMNGcaEam>SP(=YMUTLWFv7B=
zVsy0?j(6g!O5fWO2?^lIaFJCe_f}Yp;3f-eAWbAo1cnF9D955iF1IMB|7F=1i*JP4
zh_(Juoz|kH57@xkgmM5d9kMEcs?>g)+OTz~98@lA7W!bVM#S6T&I=%<n?7~6nT`_O
zQ=dC8z#aNiQ&%M`U**+#n=EAFoz)rO$#Aq7>98PG%x^yyBM8pBnmf&*2kdS3*>&dx
z;WFbEMtpOYNg?HtXI?O0{P3N9PR<n?&(7i>`U5ahESESna>uX>Jc3-KMkzPBc6dfu
ztUvb5-z88_<~^|}foCJ=co?C8fdvVG3f2x-c83ub#0$k=p_mkOy&9I^rQI$|`ELT2
z@?2xDv%eHIuo6wj+hL9RhTKuT`08t>v3_*|Fyh)tb;PYH?oInc00+wjf&vn;|3U#I
zjIg?4f3uBTB!}@uJg@gY`q;YG8Y5+^pnGJ$&Z}>}&4MK7o_ewcM*M53&QX54?zTHE
z_!gCK`lg$h!?h2Hdft9HKpiZ+UwZX*)2eTppkM!T!JRvPwa|c(!L}IjAB-E5p%yNN
z$(5pj5x@T<=Yu~2@N9tdVJ^sX>5T(Muo&r(Rc%?j+{H_&6J>*ZDD&jfr$Q=|Zyq_{
z^$pcn^>^)J%^6iICE5tfk)gV;sCILGZb1(sVf7cX+widdx-?`TBd)aEXbkW%U5t^i
z<^f0uB#cS;6lpo5-AzD0Ov7;#h2{>7DOLoyFLHvwu)6#-HV2r-&z?EmYEq9HC2b{X
z9Cpi}H+3L9$4$@|Uw>;gjr_9ioI3VUpyAruL{~cQdhobb+Sh1NJHq8n(0{GF#t{pX
zuy`h1L!$|SKcEb8P3Q~0gAvR!9%zuE`HhA+nopS$_l%|z_2E(qiw(5PH<IN=SYYaO
z^<vI`)lCX|*(U%<m!IowE+<4-R=DyA2VQ<dt`wxLm5vY`02uI$AzU22s7*&ILW7Iv
zgXPeex;klXb=k2c&J7ziG~S^3Th7t5n3AOV$2GLbn3q`GY@+9o5osg$)blUK#|yU&
zSp0-hr!+wb2ncBBXofamTWmht{Y*M`>79Az#g~%OT>{S-%||xUydO1s-vGyWe!Lee
zR;@PQYhGbpl*!n;dV@zcvb@*%rmwXIU{w?(-m?Yvd2VDi7y;JSHZ)i(nh!qt%xdjX
zX8ZZf_hc7STmvNFol2DlOCsioBdp(I`6r55s8KuJrlGOXTByA*tD<E~7bOWEq8Jn2
z$+2Tco4K-1n>B3{JD*vz&`e-6O+e2dSFF;S^|ci$xIS>?9hf_Nmf3WZX?}9bbKH_G
zmn#uRXq3*GInyx5cs~KG5k?^R!ws078=CKEU-oBixb}wP=rc)!0gMdAJ+9Af*2|q7
zE&_0~G|+^g9=8OAxXCOui;;Xh3&y)i{9bcYhX4=>O7*Sm761(}0^q^#B3aJ$iifin
zBz^8Wdn6d(y0s?Jk#&@yq{Xr*!L<^+B!Y7rJ4ENbI>*8TvOtTze-r`mu$CgE!_VLX
zPw7~tSke#TgYdgj)+ViMnoJQ^E2=Mcog%7V{nV0e>#GnTf*USZN3kqd=pP^mbpQ+$
zwKhxRvy?_9*pRz0S*1Xfv2Mcpc&o}$0!rj+i#}m>GTbaiYR1bIuKHariyZh3XE|cs
zjCII=zrLT_Z#~z{m^RH!7H?V$Z-TXu#<NY|pVey{%%bH#N|XOf^Y!<OtY$0#myHo4
zVNfNOG;-6$JmJ~nTBJl)9F>|22r&7+&6RQp7LpUi_qlxryk28tleHH8{Ht%o)2Z&h
zF}DvQ@Mo;}MR-T}2rNpmPOTTO0bl(-LVm3o5Ihk4K^U>`z>1wbQ^f0MN^i;Pfw&oA
z1P=gT$6g966u#dC;AO(V+pUuIF&4U?fAzKPFLmVd4Mb^#b}}YC5BQ=Ey{9O*t=67l
z)ta@IPs+58%`_*rn?Fzcp>bBR1%<Yi;#C1f;0Ku_+&>CWLDyTZjfHY$+o=5npbi%6
znG$p36<Lql_c}!(ERJT+m~Q!N_H9_Bpln7t`pfte&+xM^zB0J69WLbf)lRLC`E=1$
z==Y!M-Ed<v)WAsC7>)oVVWU#`X)ENpV511HBw@X04tD7~*R8d^Sl_vBf$_u~$(Q@r
zur@PI+tA(lW;#wok*|=~uAvx@4YAD;U}O-VdM?07RQK3Crej2-%0{LC|Lnbav@F?G
z-g$1$d8aqjQ_U(>l}bX48bA}KB!mPAV+_JV5(WhJGNaqQdi84e|GoOZUcI_|wFmGB
z76>Ge0b3Hn491q!Mj#<1EI<f>h8kYIsyE#!=giyR??j&a;>L}PjL3}0jLaLS>fOwY
zGwria3@7$?zP)L)t)Sr*8CIeeG-<enK!&t@o!b|EwU?GYXb-@Lmk75Dix<Hs^wRX?
z)vGpc2*y~KD341bc0~n^%ETD?nSb}LQ*jQc>o;6Q`6YCHEc27`P10>v50Rjb!p0x9
zF!z`D)8zce<9?4Cf7~|mRBl=03F|j(4kw*cEG9QPy;Fu*f16=#WvzFA|5^8yw0(SK
z65MU?#-X|YZZAmYv^M<L-7@q!O}?!22`qY9FHwt5X*~-UDYpYgYTAY8f1b@EgCCHa
zrGBu^s?yJc<M97R-?wG_?hW-kW@Ca3iX(zB3J5{}B^kJ7{Tv&<wPPRSwSazEUA@%`
z<z;+5DW}Fua@XXRG7YgoaOH%1G{9cV{7{{$7$vyY;<p7%wDbYEz~z=@bcoqNJ_csX
zDuHz(=E$+y;TZ8s07qb^;Km7-3Cs|W5i*iYeJ_rQ++^Xt%ei`@96(G%@iPEMB#P?D
zy_94_M$1qUs8Zk6`KNxmV;KQ>aQC%;^(nvla{qnniWehrXIJbQNxOg3+w=na1zo@(
z0c6ln^@*W-!*mp51#qdRI4m8-VX@#9<5ziC&aK@5i%_Bsz^?pF>GfC}3kdW)X}v%1
z^KIVx{z-G|<gi_cF3(@yeUCCiwGQ^o=8hK=%$4UE*w!^gu=rJc28QW;((`5BwXxG#
z(oe@`Kvt4FN0YXbw7h?PleF%n@@~D_(NNg!p0tmo<%4oes;_Merm83eS*56Jqa!eT
znjIl4xu&)2mwT1g2~qdBX=Bc7KF>Mt4v@Zt2dx7$(&i8tBl2swG+<=X^_l*Nix$(+
zz@g5YMy)Y%O_W9-i)bUuYb#9tu(brNlFOI3Phwrodk@2=G1MBC4G3-x;XcD;iaK^n
z-eXup>wmK&*0ewIj3^i*evM@KJxyRF3o$wAMluY=UC_*WLOR(Oec`%(r+pF+<L@e_
zfCEKv@ATWh_g@Y18^V%%>7e{O6r;hvTMSzMzj)WrTVK3xyzBI!bT0*yUaZ6P73WwS
z9r+iyliGB1Gbz6~zFZB2?cRzyr^v$)%R@LwRgH5#gpHP9=nXw+<0kiTcb{?VMU_jl
zt*Cw`<#C;$2NoR8$UUn*C-qSrUoQ=$35=wD{;!h3Bkoyo0*v5XM#l1?T=oHezz`XU
zL1Leomc=0!hz#B~jTs<(w`>i8E3ieZK~hFZzzUZvs3GHfXJ5g;b)4TL0Wf6&8;&Ug
z1EHeWAUzxF1NMh%+dq7c4V;pxF7`%uzhz3aujAGY7X8T}PjT2De77#P)D{MP02qR&
z^L-?M217o~l=_CorU2QnJuRz6;LnEQiikjBK^5YPI9E}Oi@@aoGp(D~+|X9y^P*Ug
z6}cR1Xm0z`eNO;K4}eL!!X;TvdanU)D6u9bi1dN-F)eM%e|oPbHz7LC%N_t1OrWy9
z^CC&2?xBDb;>pw$z+Tj0>sq9ub7q$H8?DAQYNc(YP0MxPq-UHPTQHw@q%?h7+leXu
zSPgLhx_0}>+<P%Ru1Okd%Hn;1HR-Xg*@5m$03%8Ji1V%!HTF3DrTu(Gp5V3Nm(dT5
zgc;2IWf%S3FLxGtnKquJ?;F$FTjhLrpUq+n#EPiOy9zd~p**h~w~n;AfEXi-WjV*W
zKC>8q9@^B3E%&YvgL;axKLpGAe@AV=@RwWE_bjiiMU5e<ALkX|>dk#~;bQZQJt(TK
zKQHY0%QK<?M*LdJ@O$2Y5r)2hbYM2?(P(K}3!S{aWW3><zR@y9+^GPq`=@{QUDmyU
zJ2vSPmdWeoW7#EKo&B?a^K+Ie?vADLnRc++S!J5_Au?-!{(AleuK#%PeRVVtHiuE?
z6nT?e_k^{zI-1Czq=i1Sos+blquEM1xpfDOOu8N?*UUa=)pZ-DAqR}epQ`B{v^~}c
z1!ee#IRc2lBEM3Fbs-m980iFr;6}^tA{T%nj)=;MVPair)vUoC*HnNeBQZ;w7u2@m
zT;R^Br+5?qFc|BtTP&C#Fc!wfc_<0i?UsI9<Gq**1a-H_hYa#CMk)&QJ`~HZy4^%j
zn-7rT++lHyRFaz=k_AvS4jZgpEECuu75N4~>f^Rc!QSf%5`PllT?|6{kxUJh>qsX6
zIzzEh0B{bUecHCUDt`b*VA5>vXm_g^AzcARB<r?JpPOQmR5VZcVFwr?;Yze6rp&hD
zh3t0(iijb^Jl3V}p}-N?H_ff8X&or=tb(t%1bV?LVt#9Ve6DYrS%DENz>@yVyD7_w
zJ?$AcE#I|Me9o_deeLYN{nRz{(8JF+&MQHjd0w{RF>1X!zgF_wLe#b4*U3%o%)P7~
z`*}%Ga_2?H^C(B!tYOZ<4aK_ZXafM7>3W9;rm5iTd2pq59_M#e#_WQxCyg1E_W8Ri
z`dUXhc}5h#$bx2e6@U@{f@YFh+@f_-{;D5;%XfZ<jf>&N@$B{MhQNez(s#f4+Z0s$
zD^H4k;mc2$cfapfPkyh`m#E-M?i`47zGO9Kx4<nPNLgt(`*z(N`g@8xr=Uk~ZEnlF
zcXqF8SX=&f^ZeaiPkEAYmGzk?J+6u@Pd)XldEkM&V~d+6FtYCV@HEDUzFUeVF(`Ws
z2@67KwnmDzMFQ{u3>hd`IjpaZk{B5RKwuGU;8IKM511uoZG6Ev4k%$`j|hMu=0sJX
zgKX~BAb%$=$oM|1HUJx(k>0kcHUK36LL87mnUPqs#1<Ke(PCqt2sp89m2ugM++ks`
zx9)Rf*tfFl!@A}|7L^}ZCWOsW3=Gld8=GdJ_#przE*=WG^jB+Cta~WjaKSjCJ}$nF
zu0P{7up)o&e)M5&z7Z2dU=OL&l`F)SbSOOoMm6Q?44u^kRP}MCrDAB00QAja*X(u0
zx)I>Pc|lwfm?MBI&F#ABj=`?3^xDAHSzln%uwxBH@IGc?j94L8{Li#!+_d}^AAat{
zj{|6@aW74C;8E+<H2X=~Rvb_%X`BA?QP)OZb%Ok1EOr05itp>80nR}Hk%OZ{8%t}X
z&!*lT7tI)14{fggDA+Vaavl*TjmX`*k^&=9_A<wAnAB%+JYO2%*;W7}3!0N<0wd78
zdU_dUw;6O_m)zos#U^P(FWJo;_}82DMiRDY#;r!#8UnQIZi34kf4w)o?s4;)Z+L}M
zkz$H;6({5u6er}+1gW=9Z@cnQN8rOBc<c9?Z3Xxuv8(UBAN-*CtIz$_Nn2rG+{*xM
zRVAkj;Ky&=A2c@Nx%u&rA*`R^u`pw7SQ{{VIx?y-#$a*}k+2u*u=a!X=lLDAFrTO0
zS<k(B`aExS9qlx14x`Q~fRUZeoltd!wG~=0rGsg(w!*Hrq~bYUCapKE{K5Whv$K0~
zTCs`wGQ-NlwdeZ96$)NkamMeSQM3`Kyz6VxpYzJLrya9fGV~s4r+r@r;tL8iT@C0j
zYD`i0+c4SUxT6Wqm%snMo=<M~o!hk=){no=r0=uVOS8SSWpJ|5ZgAb@^7xE<{W@Ue
zu%+PYVg-<Fkafkx!Cgd2?z4s+z0Qr9Bkl#zbDJfm0U!zB32wG9wqse>da#Te7_mAm
zSfaW%Dc`xh;<I<4-IuIQ#3<<;dd7Oo?yy35FPyGE{ac2L)(rGV0;Hk7{5uIm0oIm_
zCjlqqmu0-L@ko>pXacBF(<g@g_Fl^ZBW8QYy7&SJ0lcAaVr!7PuvC<&gX^hnF)I%D
zJ1Yl_5Mu-o17M^h=15I0uLfGP_r=^H0ceoWZBujB(tQFQ9Vwvtc2%wrO9DJpwxKqN
zJ%XDx0Fs{Kuw0Oz=bjiOuz-dFuF3+g1S3h`4O7>eIBfl9JJX-A^OIw^lwu}bin@NH
z?jx@LGCrrR8OR@sRpZvaQDe+Pet&;q^FW`$bx4-F#%;xYmrn;~_scl14tNG$P>f61
zBfI*HlP2jUtgd_4a(PMWx;lLNpS7%4#lun<gPMJJOR5k5d?l^xmnrx?Yd1Dg*N%S;
zpXQkSWAx8dz4cp^-5WhRGca`L(B0jQgER<&bPp}vARsV=fOI2`ba&?<DIi@UB`qxt
z3Z8jC-*e7&&L8mn^z3Wzb+3Cx=)e1x^I%xS-9spJzvE!FrR^%|{6#^^7_m<u>H+_~
zl<JG;rUBou{wLei-~(#v7&%E;cqBx<V{VniI^y2Op((c*XFD!xH=j;*G|AS#ziz#i
zr00cFo_d{ZL5-Vg@DHo(rtZGC^Ix-Do{nqVeyw%b`kr@t3es5~NQ<I57xLQa&IMnZ
z2QS&1=X!uQIqdVeLM)y;k!1mF;iKU`Ho>HWt6nTQ*N;2j+c03egYEhJRDq-GJEInl
zYFz)+&3^_xqNiT}Ie2l`S)!L=9GrzvSquF`F6?n!W73I!*1g<w<>DsuQQ*t{^JA=q
zAl>8orcnFcU2k5(j@(s>aXM8rx%55!p-rr3!?eqtEorCVdH7&5<bEb3W-rW1&!=PZ
zip9~guj(-_Me0b7K#y<Q-_i@>qnS+JR<Wnp#a+C$G<6dBSMlhp%o_Q3`?<a^`>YOB
zmV4_f0!_|eqa|EsiEg$Dmx4W<1t?Wg1yb6Ej^EWQY${vDLCXM1VT$}moU%~}G<XVb
zt$f1XiytFYgj-v_W9cOFVY+XUTs||A71GlOXhiy&zx@#dE((9#4D9rgo0lo>5*;7Y
zScJmBm}OD1Vj#rlbvpfwDrih70E_%BO|div8WguO_#vg%K8q>9x-saY(sPlfQBIhB
zLQ?yWXuL&>1Nw+cJ@*AVYYYHrRnP(7?RWsGyruD{rxA3Su+=re`KX887rghA{8QO}
zX>I5F!og?IG>SJsD*ig?V2*W{IwzXyi}O#lSiz3Dqpu$*&ziZD!pw*HLy^Z`(7WBo
zVS<_?a<B|hn}JBHU2Xinz)B0PLS|<id&vCL3>?vzz@Jw$zRQP6dEipk<^RD+6?i?G
z-xaVfO>a$#v~(=rCQiz9Z=f{`+kSh=R-+o&g}z8SO8@$|{a-4zxrNV8Tt}<X#)I9D
z6OY#@Ok+uloozM0ZD;w=uKMyh=9i9)u4SBZa>Mmpy2MFzb`6(V1zOwMutvse9&=00
z$?kp_?W(+Co4%`K-eDh~ssrY`!~&F(8URY>f49HeJIaaB7zdbJq@8@M%Di+7xT(hZ
z{Zj43le#=K4u4wuPp)kO^*FS#GB2}1{;bRUcLdAJ+0VXe!H=*1P5!+Nd1&x`4v>;E
zqdLCroQzUg_#(|CgHGsXzo1e5bhV+mLamgiGXz~-mz;n|Aj<oEQERdIo3O5EMa3pE
zx;vgTWKjGbr-k@EmB()$_TkCJ8A~MAC=zfLqweQFe*(*znw`+l(a8W5x7}>WT)T#?
z=p~WTsuf1TDikF{{iwb{j9)})RAPD{D1b-QR^fa?k=V6TNih1Fv|k=00<6awpKa?h
z!@Gy>0%SC1qj6fdwvHK#c9Y9Vuz?6EsjWW6ck=>hOpKl+m#|?mRQUe3yIPyPBBGC0
zJke4tGbqbu048QoN@~qyO0qN{btK~xla*-=BhpidiYVr7aI%0UBa6Xcv$u--lbQuI
zVsSROKboLzYAD70jM<w_E5ec$IOuU5@0j$!Zs=P_PJJo-$m~a_NBdPFSL5JM#!Z4~
zW(GZ9xCF0jO|Yyp>_kSmm++CZ<#Oxs%Cphef|B(b>vu|KRGMrngiSO4L^7xoXrall
z;NO8V!Uzcr>k{s}Y?@kZfYJ)amf_T~!1((_Mz`f)Y8RR|JPrA_-YaL6_~LRO-ofTt
zfD)PTAlTYQE=Kb+RUDed?=Q;V#QGl7_s_TtUMccbX=83q({!#q7Foso{Ye)xF8zC9
zUVC_%b-Evbhj;YLN$lIxohcrz=BeMpm-ic*`JF+qv(65eI$5uI=XIY^g2d_^HPD@C
z^Voo<#Q~aa#-q^!<{ReuBc}fm8688_UmKPdsZsrSWu8uWI%!wg;mIwmN~aOuG-AH}
zC~FdAEakDE-7vzv_Li|yz_)hK<LO;tGCI%D5X^3w@_;ATndY5+O(#S1Gz7p2GYIKN
zsN)F2YO(!iJf@%K4p*#U&e4Q#QYM%W)UfFX1(2~wP*wnSf`blBN$1S73+u=%J&@Co
zLV5X)N9aSsu@sgKmK5rY!WF&~n8JCIAYj=;n_{HtU3Sx1_lUo!=EnwH8^6md%$+?=
zN~RZ*Gu*MIMG&SSAlP<I{pamOCO`#^p9?@xLcXSii;5$C_kr(0ZSUW45Z`-+MtOJn
z{jaNJ*!zaA<`81%e08e;1Z(BQn0%+2zkIjl9+pyM!TMt(PLl1gNu|REiBtsi*i+ta
zJ+rbi($q28y;O1QrY)?2YovQ!wY&1l?z<AarH)W)3^jyRW+zw9{!OqhxlwKUE~-Wy
zY4N6f(Jtp@(=6b4kqpp$Ry^S2AS$z8yjbC6Jx3XAl=A1C55VMHNV#d3$?|6Z3C--}
zLCX71>|bNWxyN@$Z5M9IM6U*QXOY^hqnsR1^Z@#d8n~?s-q=pI&+KDXYd@8zhHs|m
z&XDT!F|X!>blTS+y<GT`5^jPQ&L&2%-5!l4#u1Y>5=|!x_->tg=48;_nhS3%MH35M
zx`?m%(K`!QuCCtv&fa$W{BDnV=|5{e^BMK$S7{mkeF{92&Dr5T>#dpTePyjw|IK~6
z)j@AXbAEjO7Kub@4dM`Wln=7jmA}@=IFUmuvUNh^!sql(EGU#LAzco3V_9AcD2`bK
z@)sF0iqijqkN~a3m;8<eRRMzy>PC(DEeRe2^XedK6{b);GHFZuE(Cb{o%gOIwy$y}
zgL+quDr8fjA<)N?ZH&FbI$VC*B?G%d6Dhh(1e~nCYT#)XH~ecU5j~4A*e=>yjUAFa
z9>OI@M@2PkP=tS^e(5mDuhi2tG)DiED5h|m^y`M+8u2PCnksTP(q^id40z*<O0?GJ
z@wSaulxGkG4MpAAKFmlpawd7&(9aoDLMELYi@J`8PM+DG3hT*!^z5KNj#+&9H{r-m
z{!G_$iG2gJAIyq7WC#>S$~b?V#E#Ft)ps#zaJleENZ<~PY-6^bZ!#6u2dA)o3V6Nq
zJ>|Hj_=ijU$0qssb`JiCU}=f2BNQckdQt%;Slmj9sDv1XQldF`s*=(V8U}_xbJ+bJ
ziTWuy&5E6KYL3xViOg4t4$~bD4_Vk1c&0t^9Y}QH8Ji??9#mr45o=t(AN2QXjNHd%
z*NE)^49T*0UH;5_KwsCAFa&^b${s}qi?xZ`qbbdG7}SKDF9+fH_xaU!pIr&c=DToV
z#u2C-G?}aX9r$-yq{|j{_h_}rvi^B3giP$i`&ZgCA9Q3(B)NtKr!y-)JQ%6QGWjn+
z-SH&Oxdh**w+4zklXD&XC|Q>sXuA~O&`YVDIpF-Zv~7EX!sH)lzNe8dY+f}hCZLos
z1o>@9BtwN*0}(<7!AWtb0nX7B+bp33@)SQwnFCAyRM|4dNW39r2g%Em5MisVz<%*t
z-E?!JJU5(omVRkng1R;WXag=E@cFH*!z(;qL|cBRT_88q8u9r6VFjQ~$&kgR#jJ^2
zp0ukz0jw5CP>R{$qpYEX4OAF~uIpYO!L#&jA0O-SB+G=tPPc0-xQ<YkFhaKiISuCB
zAOvtm=Eu<|a)*ld-ivO8v5LP{Q0wTid8<|OGEDig*nu#>Bm$Ql^WwAD*4Nlvy-?yE
zPJMegV^DfI5Bb5aJARqc`p7u|x|;GH&JGtt)pT?CJM=<=020o9NSUOfD2(>YcnclV
zy@cDWO*aw3aG1WkTcKtSLYt2QYYU4)m}11-C>{ID13HwO3qKl)0(1{!>{cmt6p`&m
z6_3!fc^mO6Pfdal*1@!qhk_69=JP$~FUbl2jfbiYAUaa7lcNGRInVeDN~Rq~*QK;F
zo%(W5eS|eK6!(??d!LLp4Ym`OLr!;bAq-0!K~`_cekGvIbwqjKQPvuDd4AS&WLL_v
zig$J;6Hq#P>)+<4x<4`;r#3fVavL5x*KDk_d%BUX;HUWATNu&o!V*f@t7(W!f>4oU
z3v*(b(qO2jN=f~EZ%TtZBw^7+seuF<)v~d^^NY~`L5$r-Z1$5rx+VRtxqR1e6x^D0
z-!|)w;z?qc?4N}VR7gsJFZhY{cs0#eov`{Dm^J@ZYoP_vaZW+2!)G^5{u%2>Zhze3
zoItZAr$RKz+uUzJsJbl^(~hwccxQ$tQcZXo#S_5UTfx=bV#JU+%r)FtV6hB9MHN1%
zqU5t?1q@yw!J(J|A^`41Aq%35E&^IZtd&P%2=ErjwLnQyG_9t=P-V)DTyBWxOJNh7
z9e{pqhW3eoAFj>~cDRCP3J}a`{f*Q1n%GAP^QJ}=btPz6%Hac<P&#j_V$y;9f@$1`
zn0Bfr0=mv`^19%7FZiwGs377gIb8sjk~#o-3JMu4o8YjQPwNZXdjr|aF9Z%4lkNx=
zi=7Ma&y8=ht3Zm*GXY$QE_2VJ-(D8LrCCCakzZDUeG_=12i#T}&&JDE8D_;Wp&#f^
z@@^5%6rnbto}|W6=_3*f#|=l3{4{fa1VQEZuAfKPe?iE6sc8DNNhI;)j`#1?e+YN_
zXA)L2Sq-q=oV0$bYfkNtQQOwfSEmW@f~7$4qB`92)E)rSvmC?SzJz<}9^IAy-zf>h
zl!@{LtJkh?pL76B2fw*hWcQ^vyHgU&Dg<KD-YL~E^9xSEijQn8ce`rI-)x-nF89(W
z8(kBMl$H3SI!WmxCbwKq>BDlCcq^qJfUm2#4DuEEITGByi#IZ{@>GRK=5OVS`+t!&
zDen8uJ^bTqnvJVKhwEXf?7O~uTl(9<CqIizI>`sw^BaOz{ujSW?yW@lW9z~XtS<a0
zt8jDV+xb2Cvg_#H`z7vf3cJ9Ve|50Gc3(Tlf4?Ms|MySsJE7yOnC8jp$^@OqpLdB!
zNXy})BP^tda+d2|pcBK#B83YZIt8f<PO)!(*CZS`D3y|(u%DxEGcubvq%ts$vWiQB
zy~F_v<0RMsDi*olS;{2|K;$SUaFXSyGxa;BDaV3RysGt50x^kvcq)ok7)3fQjt0^Y
zM|HYP`4}ho6yL>${x$$&9wXK*<%qFo-bMPE$b|qv033jkD4#=AS<Q8e1CX5i7UaRr
zUf2XcX<)4K{65SERaSLDaqnp15Ys+oh$H+;KLh7Irj+9CgMFk&=Z;PWV2>lP2Fys|
zptEM<KbHc${3Tv9M1@RS@-D();iOyF@wY{raP%>Ss#x$Gc?Ul#2_@SKd4m#w#l2BP
zOP050{HQ<x00EXb#|R)>Ov>6Mt(*v^!xu8td99k0+}%Fpa6#A>1-PPia;&Nm3ZbE2
z0#TePut_))J+@tAb<??rAfM^>EA=x?JKp@bG<~(9Z!tC?<XHIL4ju3Tb$%#J3gmWV
zwGA*`+@0d)LSgk7nAx*EUCCqdu%I6{vY4Lh`$uu`7*@OE`~2r>r|reDQbfmT{Tqq>
z5c~H9Ml7AFW8(p>)jx>nuWqpflv1R<2CoTyTslPF>))mG`)|`;p<5IfpnLo={ov`!
z(}vYFlAP>yVL1IKU-a!lOIDrVJ1u^tZFb&aL64;)4&O{Q9pqbADu|#KzT71?_;wcL
zXWLYFwK;DegunYeLzLh1-#*XPZutw_)_=3FL%#e?ef{s_!^(8Xkn;(RUm{(|FD{pn
z|EBwWSb`w&Kp<&j+?`O6&CTWevf>Ez?KP6MgyaYf^5NxID5>xkMO-1|jLuG>UrP6H
zkny!()HllonP_RUPFM1p5N=DChpX80ufke~LRo!NCclK?gyqdG=db<EEINNrUH(E>
zlD^Zjk{FAYCGjc-c@?j>LRB1k2uH28$HWt)SXrP%+ff9Cc^Z_q6u)8VfS7j()IzWI
zCMuY?;oX}kWl=j!+qd}g(twvq-+zX^cZ+V+VjGEse%O%v0b+6i1KA!z36bto9rO}J
z9?2=l6~`MV)4WL|5E|VlCW=;OVCj_%#q1pMER1(~?Ya7;4gHxK`;&vAJMhk!*MBl4
zbWI!co{&G(sWRbujsWpe0zqy1&Zt2d+t13q1>EyaKn{b5FY*Hd2Z~Kp*9NnhA@dp#
zhfO%URLg|Ee9!jz>NM7b>=iIcoaAljbM{gFJRv_LL1beXIWFpcp@G%MmJxwB0AGYK
zz={&2y%0fELkbidV{RDfvwv%FQqk#9HD~ZTC}sQ+j~%QNF=VkC#uNS**>IIAHvs8;
zdqaF%{thkDp;UZ-8M+X#XyLlvhkW?_msHc%jzyKLh4aN-GgE4g#nIJk>|1!hJPBTz
zlV5pPtTd#q3Dl?bkug-qy0u34UCt@zpPTv5&UQ@<^fJ1@c88^+_L~d-cIeR{FdDZE
zQ~AIU)juBs^Ik`b;{Zkq$oQM^;*y1@qV%MdB@9&NAxsX3_HU?MLvtczgerOdd1NWq
zG$KN{@1MHjv<=#NXbJ~<w*B7Yc1T<=e`?{kRG!z0OKu5G*6Vrx`#63SxU$yKm(o+%
zA^AMrQg0OU<M@`wEOjM5z^r?-ZtPQ*@Pkpf<z4%h2*Kae$E$PBNlHi-b9gGtkyl2S
zLtg6qW4_ONc%MsLGLuZ8?Md6#nB|WqIkh3&(9hh(?<!g4dm(5d6emD&7%erID>1T%
zDIu*&vMz5e!%UlWUIB)A<1&d)@d%_=;&dCpJG%bf$|v77zLG8-Lw;dNK^+4IkI8`!
z8w5XQF)S$l)*J#5QhkuTaRJLJlL65du88Q-e!oD~LcAhmho{rIgUBLriFg4hD9?>J
zR8ur$qm@zEoLu3L+YkZOF6&Ig1@x81Qx1H)c-zWD@vXr&P)_@<IW`&&W@k+xKb!*J
zQTzeG%m8EuxEg9HAV6@et|D{Afn^@+hz<^rp(98d4>Py!{qyHPlE4-4J9=g7=BgDY
z^S@*OLZT*ug>^xth3DjPS4`%g&73oDb#8zX%wfN<*efg7Gj{7L&r{-{e|mAfe$8e4
za_kv1(l&xHEtzEnA6MhjPqR&?jkrIVIcRi~;orJt<*hfz*|pE}A5pwztG)L+6WLm-
zmW#Y<A|)R7_CD}4r4TUodToh&Ie9AYbv}HW?pxZ*VqmN&Sqp6Met`$|=(3aS--i5r
zI$m+;J)@(Ew(WEXCH%xBKMawmM`pdAJGHk&b_;L371G$We9*otY`8R{jurC#CpoY_
zJrC=){7W`18Ea{~YPTX~RYtR4<FOQEgu>)HrjE=4Js7r^#vSV9AlL*Spj?%;tY#S`
z6g{6%A{RH48v)+mwo65fUp%k%m^=NCE%aGcF8gfu!jhw=w!w3_@f{`mHO1@K2d%pJ
zJ6&;ICw*7lu!(Tl%YgnIWibGg3y><xtdVPX2@5)a#u%me85Q!f2M1kNc*vdJ6&?No
zEHktP!LkA9GK`FAjnt(;!E=!~<~|hN(Kn$#`?>Cn09<SXvwCQ^sIDE(G)BubgJ0TQ
zfSBOd<ite>R$dq-9^A=IyWm`a!oVqsYe^)?LM5JNJ55D19$g6?lYcaf$}?6kb)r02
zqpqAp_iqXyK4%%<WfGSvP~N(3BAvvwlzO6+6%MpR<4$TCVcU)XgE`|gE>o#vi?}3w
znRT!)R7O2?_A=VwBrZ2bz=kAkf<b!MmV0KSLe+q`oU#}`Qk3$F{tcpNb#x+V;VArJ
zy7Maz)+Hz%IK_-7G><ROnMc0)Ehiq&hjCH9R;(LK@aEOkMJ8h`P>yL}#3gHdog=?M
zrJvLK9@=-6qd?cPn53E93XA)svcoIMl^^<+!C&q8d7_nY1W%J!TT<jNn(%UwN&T(Q
zx~O`-lx$`@Z-HeF+JE?qI4;)GC$ln>y)%=;VAijTvf-f6Dy|``uK2)8(WxLiKe^+c
z8^ZM4I}aVklwa~I{c^MWZF>Bx_7;gfe{#~;*BR5Qzod^GQ+V~C#HXV{5ThHTSK7hn
zWra~H5f&@r7|t<_-xhR@v5$mDvUSPKy;HwvHZOd9Q{$9KPJL%Zl=ji%XTK2Vh>9-V
z)&8*2mR;zIkZa2PM9+?)!jtgSPKRfg7n|log9HxmtG}naK!``zE99l*QCJ{Pk>2iM
zoqDh4S37z8YxDPu+7At#S^D-Hf2}+u>e0vzHm4z3>EPt_7KcXT<`lOKUmk0PM&`#D
za<qFuEC2wkmV%>fTC(AMl=*zuPANAwObdWCX85N!JLBT)31-txbR?MYq~MMu2+)Q1
zLNSc?eP#kNBP0ZMw{FL$@bureeinq|2=cGo0-!BSsR#%ah%f;>qSeosgmO_L_Z}a}
zMr!KLP>gfckrZYrS3~fLF+YcI5#1ZH!Z#@Mq|+vvX1Qb7D9KC(jg56HvO>E+2ZV}~
z$;8h$ssfBoIshUwA9v|#F)q*8pV5E|KuY63G|inq8k|WWR3e?#jQpK6nUwJKFserN
z01h<R)&{N)3d|bt4!8oJ8Yx8CZa2x-q&NlxB|3V)kdgsj8SH%dK!$^C3ZOzY>)Mhk
zRJ3>iNKw0CU09+J72<LeJr?nx*HxI@lwLsvm;YEjI)CIT3Qj8J%(QcjI|R~4-%tf<
z(S286wMp`cBYb4l2-Me{xHDFLUY4n5Tg<(V^SXRBUa&wrYPMb>XtY23;%k4~7RFV`
z#(c|_CE1fgt%1D*U#+B3eQ8l$e9sZ;y3*IYjI()1HcWZOTG?jj(Ga?AhXCQ35!7Dv
z<gn$kf3ipaXbohi<UTP5)n;A83@&b%noI4Oe~(i03#xZ}{NWq7tqxW7T22>BeH=Lf
z_#mIp?~W~R*B#*Y3cxsVmy!#KbH%JM3@?!ZjnKTPMGP1-2d-ZV)aXp`8#ZZNj#z9c
zGY%CQPTl*p;o5#pt<8MY=j9UE$L4;s!1tFgNk!2vA0^P0rV^Z=nc{M;oJcuRt`dU-
zu+P@A%mx3lAa^333$nr{8hm5U4MG2~@s~Ax<ez_pe|YmZ&nEZTwerB6Ai&dFXPH-?
ztz7Diui!e6gjDux?~nTF5<!x^Web#Gn9atlry_IUOU1kUlY!*UK(<^XGBct6&ckjd
zKO?Lxmn`a<dFPZ5@0n3he#>!deUYOXyufdOg^Uy6b(ahBqG0g9^^tbB!7zDgOBz9S
z#UU3F71TMfEr#aPs1DHfMzlG_pZJ8bWng}m$^eApo{p_S?i7Cj2${$AZ8{Fx!8*Db
zTZchwcFD}e@fJm8${>^~PjW?C+jQH3Ogmc}Nv2R#%5#-dK>iHt37J*aOw_@M`c@Kz
z9(Vf##y$MQOLHBUZOIteuS{MyBX?jM7^>WU+}Z2MAudg?VFm{m0-<P<KpRGQ!0S`<
zRl*sj&pi;@j<S@KE)U|FTthmV^q+=`J18fOg$YpnQQ^<rPq)nEt^+eVy$lU+O*cci
zrBM>_x<nyT033QuAI#^COlf}CPgnqgqc?pUPC4dzD<<YeJ(2c6YO4-{KcyV7aTLuK
z>^H9<9`2}d!b9AaOlkJHlm@S#n4HUd`%}3^8)lw`5x#RmQ>UH<Y3G?T_d!cOat(P#
zA@eW3_R3%6H8W2=wF=GMq=JqI>iya&tTWnn#=k-4Uuo`(yC~Hf>?}-Iy;_jE;PxU+
zyShEi+20<X7#O_uL+;Mq2b==FFFRF$3{L3%sP9fb60QHrK3)wf@V4-LXgT;tyV&<4
z(aW8_%u#Vt_Y>n}pnL)H|7Z9<1u=KNzCW(Lyt7?T4VNUwH{kjyZg19LcYTf=dW%?I
zyy$fJ@jd*G?~D}8?-@K*Uda^irStCbyUG4@_w?gl-s(GT*r_nV!TYa{GXr2({ihoF
z?s**I2}<k#!+><I%gSX9u3}|y1dlF5KTqFxNhY$)qbkiUZFUJN8Lua7#~m)}j`-x*
z0ND)1BHtr`pYVpR+#5xyRV>$q23V73gG;uGnb-7vT-P!b<#909isIwTkunBg48SyQ
z;d}(b&fpfl!oXZ}F1RtF=9Wc`DiNg(f@|y0U-m(8ZA{QGYmmsO=2jBp56TiYaI*1w
zd6+x&=S$Bm3;@Pt<m^gN5qcy+wL)FV2u_-dfGoF@0==BY)0a8~3)x9@>YZJXi~5Tk
zkOsFngI2kyuZW_=@k{k5#%!SlYBgVPHEAMK8{nrT_aBO&h_DZDXqsP0f~WxAhjajx
zTTD+vFPrK*4@YMJ8}lk$WG9rXXAPwZ<3ij26RoJl>5{MoO?oZHN+FLzgKLL0>1x|s
zLNSIsHF>IRwWau#((blzwN$GL^kvwtov@N5YS<7pM^!9AdY8j^yzvqu`cyh9q)cX~
zZ12Eb<%0_L5^@0W87Rs<;yo+t(n%-!+u^^{)%kuoeTLG15%UY#viD(H?Yi95Kz{!8
zXiREjbF1|&>1#(XuL;+-PuF;zzG`DID3UM}B#_WfMGlWF6k}b}Iz(p7Jpa>Hr8v68
zPt;FnSjLyL8_1jNKfK{~_wwPKWPY_mttnY9Sv2N4*7~(X(Z$e22(uNTrhSqCjv-uK
z<UA7crXL=7C7fgrV@?I~b!uO`gx9wZ;W6BIK(HX&H22!XLSYfJ-SRMZipA4&5AMV&
zP%ertz>s?4UY0~lP+YtLj`my0*@)s$6^=p$dNs~3pVk97k(5PK1Hh_GKsFd~-ZnM=
zX3Ow0lZjp^o~fx1#ur&MNwtm}q9hpcAqo^eZq#mQ)s%+n3Gluq`4OR;fgP8EfG6+P
z+>*lj8f_`)c7c7)jyM2nWe^jFHy%9*@n#Nhw$|mQ)RKB#Uz#(XLKL01!Z?rz2&Ro4
ziz#da-~yDW=OcEkPpohpqic%*cR5%HJa?eD03;Dp!)NGk1?Gi=N@P&y9^(kMq-(U<
zOAs>!4lFHh-JmjPRXnAe2TH$PN3hn*K>^fszU0T@%CU`RPkcWDEf|ZIs!tff*|Id^
z^dz5Kt-w5}7`tG*gB|@lTuPQf2e_#;=)kH@%n7a(tz}(lvQqZ_2;&I3p#Oi)D^&G(
z5ghAL)?I^ZjV>sBeiG@HG+C{wHTOBr{D<PjjnDx$`U<t2XAa6qBI(nh#36)<ruGA8
zfRcWW26o)57@f`{d7dFnZIu6L1o6qDN(|hFGcIKLJUFktMsb6r$-JZXcugkA3|s0H
zbuHfAudSky@A5rn!pEHD`~$VMwU9k4{_gu;5G96?v_hWFZY?ptr_1@wnis>G$VyW=
zpR;QQkIrmJ@c+&QuRO17r}SHzAV22|A2@P{o)8H}8Mk7>83nF9Bn+$zs*R96TX&8#
zri+R36h}ypTY@N%ZF*uxR(us}T9brCk;$IOs?|^?-Vqhmx@_lsdHh}3YUaTgE{J<T
zbZC~9nJwWCN|KzXt8O*fBm(LdN+OxzEXo_R%gT#^z_4aSNWp+xc!FQL=vCfXI7cTV
zAf}!YZIi=XQe+T^@hywwS7P;a7&Q(U?1Y6({wgj4x&UlEFcASDcml+Us*tJ9w>Tg|
z_UmFl*};6J1%Sp|r80yf(Z!>s5J&xq3P8+kNN(J?RQCU50pQ*=miL(L&64@Jzy2`6
zi5>?q7B5o@n$+}g0lZsT;d4ZiPG_+S{tyO!<q}Y|co_>I4)MFecnrnx_bC2nb3FlA
z*J2rh{Xs{$eNH|EM^Jzl!q&1(1dH4_2D!(<lNE;{aw9eK+{>5$lV8=0>bCS#<8H0i
z)&}A(|4<xei&jI5uqpzK4m1M$?TWRn@htuZf|Ph`#F6x-4AZf{iT<BlzUBN5yQ56r
znh#WO>Lv)OxVvOzsD?Kd+~9;(g&F}-rWt3k!B@qhJsZ<k#h-67BYZ4?aW(&>M`<lR
zbfRmkiVB4rCa+50IhdVrKyVT=mCpZ<bVJNRn`%j)QZRp66f3B1bkg2qy(mi>h3Qzp
zt6cdfgsGNx<UJqsArg*Bgp(A5bJAh&7)t@vVqhL{PB7!JY#9~A>ke-Tjl~2KqRP;5
z<YVv<!Pmjoey*+KAo&DR^p9<#sA;-Gx{Y;Bun<`5=xQDQc8>0l<J*Y~UI*@C@L&iH
zN-qK*L1ak5JmF4YXo-@8GFEXKW9kSz+zN}0_?-_p;<v5}PxzJL2vQ{pQFz);h`r>e
zSAo+H-O;0=;}DD<!E1rhz?B)oSQIfUsUfOHXk@M9C@$2Kz772q)_1wr0D`qt!rlC>
z6Z045aPNB7gZ^E8I^Itb?pQ|%T(Z$$nbynzrt6UqmQan3<s+;iL>`65o3=K`w)03E
z=fw4r@8{V}P;#7>YwI6xs)r;j*A7!;JxWDV{&{Cof@*I||7CyT+$|=zY3j=4hV!60
zyN$PE5O<u_wXm+El>hfANClA>92lTzAM2H;s~!|g90+l;M)#YV|Hlt`csJE}D17bZ
z&bnMUcNrGQK||YD-B4|E$6G>;xF$hQ_fy*M<;~2M?<C5*#*?V5-M)XAg(4yDfjwD<
z)SENX10o2fNyTJ$IOAz0PU`H<&oGp(O;y?Aqq=zw2DK2?>(-j4RNv-^ht~zVqE6p4
z{rft=uAgN0Y3`b?CF<84mZy#JUGbhgUrUznz8{KP#Y#CA0z4dlXx*EBeHFNoPwJ5H
zt+iDps7tmy$h^1F{;d1>ON&+3uCf`n_VmK5h_7<RkfS@i#7Q%9zU+*!Th}sz7O`5(
zR+iU7%9pRG402o5)FYkEkMQwi?7+M&T~=Vk5L^Is1>fyUO%%v~N6ic;2Bjpj$2xHV
zQAa(SZ3hSlfD~ImH8sJi9b;^!;LeyIJm-e~>c2Sw3aB<ou;`gS+?YDwT6MlX1Jcn!
z76hTGConh()<wJF`dzXGTOR2zbgfFb-F;S{#4>&Rn+jIK&?axB6r#vz-~t6F@`Rqi
z$)sZb0q<wXRpj2|%0}(<V2?{_5nPo)(M?|iS=lE*H~>_1tM|?p&c^%E*&?+G=xb>s
z&cqN-I1lghm{z~b8qp2<JqWv`TC`LG+wm|do$3~A3mYlbxkc|!7n39#6Q26aiy0a1
z^Q$zV<N3&edTf_yXTRO&pp<V&>bD_FJ{Q0DF1nJbjBKTRQP3ZaknHT8sptI4kBZFk
zIT=dHq$m<ooi1S0F9rWwxvCTR?{y6UbL%c9cc;37`nQ&CZF(84t6zYZtDk*drW|Ok
zE~^^gz#k(IG-vtTj5mwxaQih4+hc`3>9>2e6SLf{#ZRw5b8oDT52F74&DCC;aYVPu
zIph8w`ys?RH=M;~<*>!r*naHv;Ad#%VD@4es;RRJN1J{!7A~sDR=*$k>TiFioh5a*
zPwKz5tIUZ%e!F%-oirV^bHU}Y)pTb}Q32$iuJg~G`5e~q2lu1whdNP}zNBeWn{cDh
zAAc|rDERI~e<sh2txDF2@DcmTz00?XgCFsqcq|hY^6?LUz9{*G`E|KSg2T@B0pIrH
z<z4&uCNj3EE{u21F0Q0=#%$y!<tEo?B>JySJGvDRgMMI`6dn#{g&802jR^4cSX+U|
z*2(ArBhPhkwDIMkFNYhe<P8jd#sEgh-7wZqF(q$otCiP4OsiCcyO#aZX-cb`)is+1
z!f@vjJijg_IK;<b_p#u&<+>e1K^y^qbLbCZk=S!544Vw|Df6bcxC1hVWB4+R&}!Z8
z<L_B+Z0~OOfi-;`LQ|*K>p7bVjKk0PM4@XQU%@kBNP#38Lcg+#+8J}rK){s9nb?q|
z!KuyJ{U*i1M06Z6q9&I(rk_du$=mWB!9>IYjgUp#kSv3Kv~695+2b{UvU6t6@3_oq
zjm|^$aJu#93M;MF&qh_<VGv`-H%4F!f<$gV1HxnC`D2gf4#J!kb^0<3SB5+aH&NiF
z%nNw2>!~3ucZQ5nAirv@DV@qwhIMJGOGv3^`{nchc5(O)P2wao(4|+=wp1I%OVvO<
z?&|{{a9*EoqA!YQBJ_VE;4klrck`9$jF!ztMKs>zu^8EV;n7Z{FGZR<{GL*IMGZ-j
zZ{%m|m1j-2H-501Dsx~Dc|Zb`z1Jqk9EAVlf2KS09CSTgu{?amCKvp*9rpguaH3V+
zh4`jMr+vDAlZ~4*Sm71ZqZ+%~&{pLAW>?V(rXN)1u$7R}Fs4xznv<8%$!-0Xk^{M1
z!_6|KI(~=ei*DBGaFjKrf)+s0@4=z?1-XVofxXmglytZMfX#^jtO*YUy!}`@AVB5!
zGQ$gs`iYLp3XdAwX5fa*8Fv(of@32Z%rWE&|EX2Uao;hX){5fAo`HGeF`>7f{Q2`K
z_|eLb)j19K#@@^xvxS|b4@2%}x3Xs}%R|89kR<*OWKR?12&g2jrFe>&1pf&l{($<Y
zWTb;D#SDRmKo4(+M{^R}gBw0G!9jj{a1H<yzZyW{u)a;Ny}<TY<g@57edAf%g0eb%
zlBLnm7^Dbb29S&~0liy>-^rC{Ffe9WMg!bCP**(VPTNakgDvVN0Ny1As|<<vtV4Hj
z;7d;d<$F8?sN-c%Qq}l{>T<_`qY=3=vXd-qVwpKg-N$ATMtyHZ!6f)dPvs&Gi@0=^
z<~Nm0>96u{xgoJeEjZ>^!h#Js++H8lOe5xNN&1(JkPp+MZGw~&l=EnHtXlI^B9a1>
zuu$Cm<%q}+{}#E#mY)gS?`4HY7}}+7FwCbHzI7oL;s3QE!eckjnb4XfVT!WGg#}|a
z6+5^*EXn!8J&sXKGd-uN&DKw;a`j$Sa{Q_fVHUN{{{Mv>b6!>KA<9Vg&iP@{y{|rs
zKbq2%TKA*!8NZUvJss65kB${}>K9KSK-%WMZ8GV!>=#{|3|h3?J%CzC>C(D$g7Rx&
zVl`Ky@)l{+31K4!Ky8fgqpILxD;S26hQLj`jI@e<cUFi>gaKXAdQX&)3ec^<E*=Ww
z;WYZb(NYMH7h)WgK`6_I&hY3cpoIWP-{{e!RPsRwyU=uu;K{D4g|Pr?I2n0K&=!E=
z2zv!e4yOZ9!kL2!A_Wpr4`+}mN)RECK%4IEqmO`nl3sw*2^KOIt#xk>U;uOU%QWOE
z<Wy@(?5>ds<%BEz$drZ_*=I3>>_-rUwtl?W;&X<T@{<Ihu0hVw_&H*f$loA$AYgI$
z6OcPJ^i66H3Xr=bXoWF_V%Z4J<~R<L-USBGFGfNW^26$dT3x<^5ep>j&3wKZ@<D0f
z9(q~CS13L9z(!P5)cBRD11SPyRXQiee#<B?f(Elr@oKUqc>gxE^KM&nh2Kt+tkpcP
zav+V<796+r>vQoRh(~;zFK?t;jzOM#e#SLkqm10y%m4|x{Tw2HiJY7y)7j{2ez-LZ
z6<We#ZBJ-zW7OOz`Lutt(JD!48l-A;w?_n#NUT%8beAmk*D<^e7r=j<d=t4fp^_1#
z;V2b<WxJGFQDbgwpKhO$5H{Lu=T_2bBFSTvS8uzVALxMne=Y%jScHHzb@p`;S#Eeq
ze}R-^*1^|kVIG<85Sz1&VY=sEHWp(xv!21ySMI0z85bvf`*s!NSO@Yq&L0<FY)?(}
zBr+!+sAnToR(I69d!=*f%Hf0*=K{NawjG(+cLMhe#hu(cWT-QBDiZ`iX#_uCu~a1^
zKWt0ruSv#-Pc(3502LVt1(gsi$`J7qEnqtd$5z*~k~?ImV}nt<JlOJ}#4N1qM?|_g
zrUc9#cHzdgALY2RGbHVCLT~_@G=8ci#^~Fq7v^5kMz-q!nqO*IA4Xv|^IQN&R9sGI
zv2QWC@m2_V?<L{N$^r^)?L&;(eb%nj8<K+EneE@6HCBj4sCsC{493)(>RKvlgrBtv
z{ibWlCUKw%BZg&6pbjoO%u|#7NUV<F9<m-$fKUPn-G{coZ0xY5R8C5gMy*It0X1OT
zzFat3Mld7xGLCMhlnrLv%}frnHZZ1kQY&i%5TvC32xafsbi{^Na#YClElZe_GqHVI
z6G^j?`I&#-LfW2xJbF`*;Z!2i8H!?Jg~v*9_+^Ps6Hm5$@6Uwh<C(>4UPjuHzPi2j
zIGhxZ1Md{x6xzfcSBcsymP+Tk9+UdVo$LGQUy75Rsl)%WZtxB&oCbd}C&Hc5)`FjS
z9=reL8hLLXuh1P7dX+8?YLsu_i|Kfl?$y;Q_P-!GP<ga>)GGgP%6;+pl?Uxcuyglu
z?;>uDg}=#J?+^L?M22nhQ@Ov6_%!(v<AHNlP^o_(GmWx7_8PYZcWQz_?ParhZ1Ykq
z(;YKE(@`!X9}eVmf5*XKC^Rh(TG;4tIpQx@u%wjxKCAr|3tK4X&XLemM~vWcG0^X|
z3Nb|0M9RNAs@zgbVYAj%*)=1=CoX;?Kaf4FBzBl#S)+jiCN8{dg%$pwA{A&UzqN4$
z;a41(w8X&yD%;SvBUGx31sABg9aBR482PQ=QV=+r2)um!^9MvBjX){G;7)CdI~u~L
z@EtmP`!3Q4o=FRfC#q}azW5pUa8%DthSDTuU?iFmZPQrSTfxi~Y;CXeiU`y~oKxZ^
zdJOUvZRc5m8Bx*JS%ugB2o^S9Aw6wFT3kVpFF*wVx4V;`o%B(lg=Z&(F1P_Mih;L|
z4tVoJ^z7p*Jy4K<u7tBKN`04DzJx2N`<8vf_3m57CprMNOzYmjOLMBjEP<W6Sb>=`
zFefDT2;>Qf!MXmh^_w}YPbVWJE>tqkNk`JA=eih!S>H|+Oz?iaYho^!I5+PZpf3uB
z+~|?_0(J?U6rxV29GK`%G-R-^-XdDmK3yet8)!J9^<PPKXR_Lkn(MrBhG9}TtT|=S
z``TpA?+Y|tQfGO1Qy_1n;^WThjWeHb!>dM!)X4c6`-MmTdP&QZ;7;<^y7Em%9(|PD
za54hz8k(>=zm~}FYee0tA}LE;t?^RR*TO7;kB|AC|HeNE=a_zPYeG#(8Org%iu2l>
zxaysVw<auj#-)ip<a-@-5r-5NCsBMf){P*s$dPZ}AlY7X4^x|?RnSR~+F@jwM{wgu
z*vJg=$s9ir^+$bW>GZd4EPl%NecB{$Iu2=KewfbMd$iF-p0R4ef9f&S#7F=u#)oV_
zSG~+?qn&f*`)atIfty2RAjUK3rB6jrx3MCW*-G3a8~l*9@Y<f)+T&Nkd9YYt%?p9I
zR>Z^fIl{vNOUy1_8J7+KCRbd3hhV#8>?at3-SRc3slCT&bf1)<?<w7{VRI!{C?SF%
zYw_ED9h^Xb0>Ehcg&)sku#v3m4^P{4!NM>NId6M{tqNVG*)g8FYpDuLqXf4JYHTmg
z^=OQX;0)e-DB;vvN?n$xS)(JsU~KeL{W)sjIp+o@{G(50>2nRkO9aRkpyd3QFq9WA
z%BJdbdK(->f2CDd=bksLI9H=KQfA2D=y$xGVQa*HrqQ`Anew9|`x9+P3`hkq<{tSk
zVZ0&`AFid`E08q6Hkq3LSfMX-zpXWy$`h&tr_HhU{r1`0h%_2DS{k{ULT<LYLf;@G
zOiNA`5zsLqNviu~<L=jIIHM%E?HdcdG)GE!MG?J*OpYjiCfQ^+r(lnB$9!R7DcKeF
z3M|+N%ZTZ(Hc~g+(Gf`wRsG6AK_~Wb?0Tlq*N0&Nf1N(c=+I#ir2h0)pDf^pCs#PR
zX;T5p*17J$aVNY4kWdbj>|b@~c7c!fX&dFsn*9o^UNrCZnFY39NzvI+1g!+LiVzi=
zTta-BTe6h$gqKAv4|I_y@)esy97nt;-p$6Lse@T>r`uU~PmWTvB9+8E-~gulyNl8j
zeRYw|qvM?)^;@&qnE)%&H)MC!it$wSFM4)rvPVOaOg@>2-$5^3z6+S2$dgR^_g?w8
zTN%GB^_vSCat1vjIxOA`BC#a758Gu)p>LKuzS`#VX=7jc>*i<7`q5I%)luDA5tk=c
zi?vKBDYdKCv2V8!>aqUMzIM3}R?3~tHS1N)V$!x&P08HT%x3mfNXN~|skl_!c$kAy
zBE}x(44Y=H6j=D%)Z2}%w8#1btIL-)C<b{Tp{-QKJ+{mQ|K#aEG^q<l4z|&Y`@eS0
zQDLEY;6jSsN$9?Q(^VY-xk~_&i0ETKF|{>~6Q!+}H3EZYB(H-L|A7(={frx4>6W!|
zg@;2p`Js<inh3+to7n}Auz^(Wq~g9LOoGpD6dWCQoYi=Z($Nk@-vk{|E>K6XUwx`F
zmGw#3>r$7eIWkm5fNcRHcy6te>0DAvE*bF?p9fG#yccNessN19-+%;Af`GnM2w4r*
z#N3SiYwnDu8T4@!&m@>V0kl`2&!l0Pc7blqbl?_HLf;@ci3_0s@M<&)!F&-oa<oLx
z_hz2lNf{$QLG@B)DjF`n3wCB#sNx{?h{mMsIdtHL>JO_qgewDqbimM~R?6&wp%NWY
z(@)p9&JS{)aNZK@-4e5-Lh#W7$$J|4ATmTWsomF*vf4l-{Xh01;}7Wp$m{DMM;HDq
zqRCWGLc3_Gx^7ZXcD*dT<}snm_p*xEn~=p%zl(F#tOrdTohwE_Q0wAG3K4A;j0y2L
zPr3r(8(CP#E*sBUBT1g&6K*9GX(mr_+$)o@sp<^XQ?Kxdrn4Chw-;>ZWIZnx*pT$B
zIrJ}IW8FvzG6`?tK-#BxHGVv{i$mOhjV+qL`x{<s>g+F`s#Cqm*sHCfnSf2RuNA45
z-MN<-X&O1r6<4;#>;6~dy$IZ#W`Cq@EvWyu%#?o|_mp2%YOJ`NvKZdv)7^%xpGETz
z;Yk(56A`iI9L)uiFggfckWe&NjWx^a$&F?!sN=!o-{1j(MtefsG(2ke&ivW(^1}Io
zgYv_FPKgZFOecz;Dz6t5-kq~Q58we?h|5&(5<>I;$#=52^JQV?d_!s?vV=~BcSe9k
z3A<{?DshBw6a}DYp4MQ<_Ui^Hln|h9`ic-h2+B?ckO4Gc-f3b~k9D^#w=U7~v)`RC
zNeG=3`Ylu5#StZbc=?JDR5gDa{f=o0VNUF&fax9~t88U;Au-b_M-`t}ww=oE8flf4
zW@%kPyM=;C0Q>1?ZtXNEP=|xbVk&US0FKUYN6W+NnTSt_ciZ<y3gt9@D`2$rl>&aj
ziQ(L~Kp;!&EIGTv3Bd}5vegGe3B8R@1mg$>cUtie^P1$0agzo)@M4~iL?ky{%qLVi
z8qD9sLSYQz1n5#q;>myfSK-uO%?kj3?HYU|3&WgN!LJ~MmWVn9xB+(QXA965RPZ11
z4qsaEN@$@&N7c<k(rm>}+^MZ2%z3nwB8j@nus;bie|#ItXcdP<BsTgm<~ljason{!
z{Xc`N;NA=KpB7a2F_YIK>|RI_MU6<eFUCYc<xbJusw1M-0O>zqD|kQRycy(BL{-fk
zgHbn{x)5&_*WolL!)A0qXTB+hzvef5ze)A?sZ;XI5NQSX7Q}SNZQENl9V)kQ6?;~^
z?oJ3Qd|{Xnyd>gvyg{5PF`XL@VB+b_TKm2x(0~#ypp@U~LvIj++snn9xt@O0NhpeM
zIU454?0@?vqZOPZVc3p{w3Q*T<HNX3y*aN4w49Mwx!54p6Lac}y(KRO8rIr&Z(l8)
z{O{~DYs-^8Gt8Ggu~NKlI1!{>9msj9RVExV-F)cf$^Aw-b*f!@rYujE?5T6?ibY)X
zA4u$z!QZKJw0FR(=GR~L`H@)AErmg}rdHfoe0td}E!b|+UN=W^<MAzwYy=~R!wzVV
z+^J;*&lx5%rwgh2Uu(oABRLKb#jZ=rEETJg-apAGtM?anRUrT(D!~YYJt8OAx;`+t
z0*%bG<_~0<L;VX}_Su{;8tR{5lvI4&sG;OptVaX@Rfh|wP|N>eJfZmEv8lVhE+P*O
zY)8cLl-i9#qixF#fvD7mOJP{_Y@Ts{UgB`Y_8F4HQi$74<71%dp6hl-N?%ue^zrv&
zR~^ry^X85?qdN`xhWNCDpap93uvjP^92fh!;>Ab{7vRIz4!?>d=w+l;W7;q;gDcja
zWo1(2SfvLcHO-8*!4@v#7m{DmzeA@2tFZ4C(BLYHl59#CJ7J&0dB2V>k}HAYzyK;*
zD#b!Zq<{`X8-FWII;?mQIvb^N2;Nr<-@1GRuAG13*<~O`L}T}zT~^?Thkt(guK><V
zt5w`n2a&OopV?qWDI<kphNM7j_6gno=U|)b1KNGk_ckxNnSr|ZcAXuo`ne$OYI1qI
z4rRr=Q>nTBNyA}J*j)ytdzGOF41NYv167dZC*L1^LW1Wajhx?2e{L`=KN)}9rp9vR
zy|`<{dz?gqy87eoKwru&^JEeJZE4W|tCKx_YkrA!F4X;5vHiUcFpJlIi<L{mas8X-
zsHy9_?%!d>S?CDNB_FY_vqtmi)4D$A_ju@|NjK*`RFY-!1rHT&ac0J^IC^Rtd9N5!
zyh8*3TapD02%)~xwoj%h0{SuGf6WdUTel87a9E-4^1`w<P6~aQZ?b?Uh3mv+$a-?;
z!;@S87DyxZtvE_GsF3d2y2oE2d$MX;)%w-W{Y}QcqtLTy>Fhy?=o&WU;p}g|t}jiZ
zb;bEOtZ#9x?V#>oPrCo=83Hpws^fNyGb2Z6W9G$`_fv&!e*0c0C6SnTng>~$Pjdf`
z^f3*;=K_U)UDXJ3ErY=SL{7_2zlGu?@io5?&J0HpyhW;|IS$z)fi-8y6#X@OU<L-@
zV$J=lySfq11;6rJrH8>{CMq17C5%@I)&iQac*FfjNvkvUhGd&$KpK7e!c=shx|NRG
zM0xva#PClj4IBm3n8{7>6UZtOir?{k8TK{sB<<Vy%=HiiSOLpTE2rJU6wq_d0SUtr
z018`5C-;VkuxBdO+sIYtQw}ur-}<Pp2-0Y_fNdjvR%w8`B!E`7BZ{4p0Zj3(!rCl6
zUbl^bKXQk5tZ?fWNI3vx?M@*|@K)PRU*nI;35kGC7pTIWBa({`nl0)~7Ap1%yWho(
zrg@V57Zw3nA9sk3kG=c4$QsT9;6O*mqOHD_LG0kIFnDlpS(?3z@3K<WCo$ExmMQ7)
za!{r9jtj1crK)GTaT;L%_HQ2zt~I|P^1>fQ=_G=CD-r2d2q^3R+oBM<h|WKZ=ahf<
zj^g6(bO~s=fllU7l55}7gF=TP-c#bY%)M10emCyFn^^Vd|5je=Ob0HNFqzrkRGslr
z{ubQ3n#jL-Ao@={E`66?bCbM~v6CpUy=?INsRj*3>g<FUer{>6J{7{;A(~sNcvdfc
z_mO8zLcC^qGSWGSw0E=h?rw@-sYzDMo<d%mEemZw=<f1G$S<UpW3G&%_RjD)e^9ty
zc3I9=`bQI*Qa%-J_Cb2P#t~xdKNom<cK9gL!>e#P)pRVNjn44l2|GLKL>FJf3RbeJ
zI%`-)X}J49ME##B*Dz!AL}lQ`9V3~)<rCl43S*X@_72XR-{Ai)Pd=m!xw5=23!S!u
zP3^=hi(FLf;0>Vw8jW<lwXV?vWAKQ|hB^P!ofM8jJE`>F6p%)}=gZBh^bDVKX6Vq!
zev%>$927a$2tWbgLYOk17~0<$?oKqYjr^8zmjYwF^ar5Ll1sPiy-7H*Q+LJ+#RxJa
z6db!7UWMScs_uMA0E^T}ae)o!sHrO~diAj)=+j4CS24C_=bD#}+Z{qvj;6|NtpVG}
z@D3fpGxala=6at8mIugaSy*BF?k(1XrMn6dmGzMC7rJzYj&2tVdR0BBvNAoOP|d*F
ziIZIls&c&4%%-RRbK)schI1sE{RzawpWp56y`lqmcL2g8>6^w!<1@X!gAae&#}QC(
z2xtR*+?gVsh$m>ZyUk@njjXb^j8Wz1x+s9#8;Sp0`gKci%D#32xDWkUN&yRdcGcVd
zi!rX}HEC;je#$}na=Akr+a-Eglgi$?lhDha8G`GMCpm)gDp?=Hn)zR`VNZldL7oS5
ztmipZ`Lp2o?!!haPEhWstiO^CZk%{88A9bHG8zi^2+E6tceDLTxb8zvbj7r|7`T6y
zp{iDKmti8*JnX!?r#t#AMi%s+7#Fpw<hI#xYpQ<S_B<c&KhEu_&<EBg+CB|8(R!!6
zx+pydf@HFw_Vs!^Ax_Nuh{YE(l6x_WQ(v&omHcr|^EJA-vGKlaC)NwryM*b9*uvHp
z6S5jQF%y^^e^M#e%l)exxP720&ua*<94*?#<to^P^mXFfM*B`iF#DJpji`AZ=&dz2
zv(nSB+kJU5;9WgGd#&zi>F(pwwbT6m@&DrLEWDyzzqdaOEhQl!-QC?K(h|}=w6uV9
z4WNLOl!SCkgLDtl-61td_t4Gon{&>0t#_^WUwH2O+4tU`y{~J(HusI=<r`VtnD35w
z48vO^Z))Zw@un&uLnXDp5!<O@FAxfML3&ZcqF~8w00{FT1CUvY8Qh21&|}nV8bZFs
z8ok?g@P-h@z`$f+BKI3>tDrs7T;;H4N}xifZ#LwJ(aXGyla#oI4@CdTb&VK8OQH=_
z3K#R{^g!wVidcDCV>>FvYUtHy$54qNf*apJg(Ro4(B{YA&cyDUWEB;S{wcMzXo}U0
zAo;WWSz%|rSH`F^KU<dT`F>CDMc!%Ykz=D3qg9n$X9w}AWVMX6?SA>TNXlBd>te)&
z4taUfx)k*g8V9Q5kxJu|@CViKG&sjLOn*N?asMP}$KxK!8s=Y@eOb?X9hoY`QH~12
z0;?<~aS3nJ2L*;zYh-Nw+X|dn;yy==3|8p3S^6K(ODb$g(5rg{ZUpO1(9aaMB04v-
zWI^x!qs6EgZ}P2K?uX0};Dp?A(94Lk4l!kf>lBmVEJ+PRF&}8Cqbf*2I7%?4{hoK`
zqM8{0ZaVVYDbj8K04Vlro~fJ0CYE2qA*1c2YsAGVH(L1+;jiWaEA8FEIFx27c%a-}
z`+NV<bNr#iUsw*$#q2x)CI-7`_<06hYqQVhluUN;=3WyO$a&ftaL%B9x@11u9l)z9
z9w2qYd-Sx6VQ67377UASXXS|E{e>-AEJ*F^%3OZIsN{-0^QIPpGovCE*VnQ0U+%`u
zWyMg%9@6}aORjFDxc1%(f)Ir-i7Aa#e_V2Bbmp!!O;dirNipAQIkJmE*-pC@A0JsR
zY*miactMj<7~B5c6Qbl=z=k$3*qR`xaEgQI#woQg5;y-yRK<i~QIfNqphJ{w>_+(H
z0VE$4w^3~ulJs0^dD;nstyp@*pqxpc`)k|FB~EGq3IHA_kZX(I0Vap2o-~-QFab<{
z(o)FRt!9(P{Joc^_Lzo6=4&GAM_ijx{f9W-=~PFy<t7M0e`p39AZ|(Nq>9W^NQo%4
znV)Gzmlt<A1xx?|0LW!toli}4H<UQT%N|?KPqB`CEc{=Fp)g?_;erz?Ro&m(da?`C
zf)&W9Z7J$k?Qh144Fn^Q@hU8jvde45k&Y!l_980`>(+iL9BmD)UO-f)P45o-NxzEk
z{wvblyw*-OHpf9s-lmEk7`xMcVnz1;>!PN$(9iejcU{E8U&QWkE%&zCS2@uO0EL5h
zEu?DfZn!B#0Ueb%9yTV^>=^zXEjDPtH)QuF@*9D10>_yqpE=%Ap9$2&`0GmyC4M)W
zZPVi`)hiH~hz#ZZmD{3npyDWNnRUQ~5N2|2j^u+Aw2?W23d=%7c2&zL@5x4HfT@m_
zic+kr<M=sgeG3>kpAkAGL_KK&V#r^na|}g$lzF?jqV`jdLMnnR4VBa|`Ny<T0f)+(
z$Wty%U4xgd_YwSBCq<kU0w5*L3Jk#}rqxotw=SU@#fxS-b?373b@ra=UuPM)(s`SU
ziRL|=g5s6zyzbs&7a9oMlmkVfa$A3bYGz`2Nv3RtPXo!pJY^c033;J4GZ8kid#^>3
zM>rEv6^?{$=p!KbmZ<;}z#s${2_@u84oF06_)*d7mwP$YdgM<>TZE4XpM_3;>q8<t
ztEFa(%s!7&ByZ>gw|EfJi}D>6EKo0yZd%wWK6@h*5LbV*t*0h#p7HLRU>G~+0BJ2G
z(!@?M!j%7mlTrVmQm)E7u1OFfz6UDPBq}fS*Kcx@KB2RO{-~fr)@#O6fI<b?_#<m}
zh@DPi8}i7S8bO1ZY4NWWwCM`#NW^RM|G@>fygBkN{@<kbk-9_0_&2yMafiKqj(#0}
zh!onfLm&$}bsA$p4nlkgcK4$F={b$;L_1T&QH8MuWllkS{TBaAgENYjSy;-9NrR81
z#`#O+-9Vve|NWN}oDN~{5JG$o?Z>$45V@Q)Pfse<Vx$<NS@^9T?!60Uq=*w%+DlVF
zKgv3Wn+Sjxg9A4S2mnq3!Js#H*KGzwTu3@EDcZRwk(s`tDiVdEq#(&c5UqIq%4H@8
zpObP#FV+WuUV+kgMKgH2HigM!sR{)D)R@+s=t%_Trqmk(NQ?`AbuL6|)4d5aJO_ju
zfZw9|lSi6vf<S{i-&03ekf=6d(}t@|ELJ4muOzfmipwrIw-apl1ck6_r{eP5;?3g9
zOsr{ZEPAcvw0BVhVqUU;f$Tevy?f&aApO(DdJ_(8xINK$?B!%&8~8pS^svaM`th&q
zpC4(=@1t>SgYPzq_c2)P@x$ICEq%Kdvu!K(om5|H^+N14R+zWZp803wE8sbR2s%-M
zhCo3hvr0%YpB;mB!>S(^RD3O1A{>2q6-)jYHgiQRUGjeL9<4I5cjHE9IYXw8PN?5w
z#KGx3qf1nfZa#FGC;Pe}Z+~f8Un_wl;c*CA@wXp%0qH<^?eYDH)<sdk_0;~L*21dh
zPh(R!{@r=LE7$OktlG<fbZ*pWANzM1YvQu8yVoE88wO@XQrOT;kA5(`-A5#Ie;K0-
zcD0ZZ3>qOCHxILK+TBvS;4h6~hA54AnA+IjzXWk^T50}nO(hE>B&NYX4Jl^_=3n?P
zhLrNgiRQHmv4~;g7*3W)an2IfSSA5pY$Ra!^5l+bDe@|iy@Qd;PTPExCU)#;{qz2z
zc?npXiw8?PpeQ$g(yw|^^D+tL<{gjfFi;pRLUDemw>B;{*;SUrajV1vH5!`SZT8gT
zV4oau{<m6$$x?1oh~_1xmr~#;L5L`jM$Rd=)3Oq<Mj_=rm*Nq8jvKS`wAWM2ouc%8
zk<*=a$K(sG^!maVRd>XCU(A{)6dw1Fsuk2lrudGyXjHeJ2rm_YzCN!JAO`d|H?{23
z>=G>IwX9|HJW>)$Gf`E21#S<VXad4h`I%b0refd7XNQbB)~LGbtcE_F#CLwE)qEys
z{zfb(c&5K!)F6SK6Pyq?kNm1K@CC^F$Ncn|`P*fE1jXAQSD#v9!1=fPpuCiRaa#tU
zzulxdut^4KT&xjsnYVwS^6h?8Obabua9tBHznxCqCV*gJi${z7F<hL*^3}0S1(kE^
z6TCHQM%xiJ-?<8DKEYb~qOn6DhMn=oTfM<r%mZmE^TQUUqqm3EW<9r2OpML8^RZ7z
ziKJ^!f52U|M%u3Q<D*VsJCooCu|EY;hBCuPc3!7rIY~r9!pia&hoy^uQ{+9$wS=%q
z)34q}`W*$~nZfEDEcHZ-ubM_L77<eOReB5|_1}VXnlh+hS(|bToWDKl8f|p1ZV7op
zZ>dxQxW7F)2y^?$hAz7!wl3*^?^6f!ZuDU%SeS~J@KhkbUzc<I2`DB93Py4RBSXZw
z0Fil*UP;471^8CIOhdo&z=b`XD?$6x5PWhw`Sb>?Hev|4G*oSt7VjnKO$-46;03}B
zq0<4VC}b2-yqY?N7!Bt@BBkL{q1@~;ac}Wqpsc0+IEJd7@7Y!qJn`om@kb-mdv~m(
zT2AM3%6!8?b(3tYs?B`lQEJ~^3b%L2<ZMBq^%TGxiw!~j_A%#eyq9I&y4fg@!E(oh
zj9-BzIta<A5j95O1!&89Z2j`+Idl8<?Mg!E(+g|tDlLTkv5R{n`}b1lmaN0tjUoI#
zXTSSi-}NY+GRv-AaYv9=dPveP6h^afMde%c`l(Yn@Hw;P%EA1m-!_VuV2?S?h0SFU
zb(Q2RmH?}$?PEnnRc>79Q?nOhcB>?*P^}(h4u|fwb5IU+R6Tk=LG{1$Zcx9Oy+Ph*
z)qvq^{jyEi72<ct6$2HrqwUwuvl+V=3C2sZDaR!I^fx0uFIlyHe;0Ps>$|1A<gH_F
z#7Xuxp5MOu7*V7xRXI2t&i@b|^Vgqkj%t5I=wytMVyp$Y5t6?4^>tOTF@cHD9d;`J
zyt@<Ft(SHW$(4wt=hf?rZ1V@a;!4Vhy21#m0LVXXVzCwXE3xD|=1-5@!$Rw*n=u?F
z*2QHZzT>IBsF=oS59S6t&RsuqI{x=yaAR6AJm6Qj;#qjqbrx_cr05`9K$ND1qR!6*
z!JXd_*!L$YKTs$l2H}`cgG=#ON(GHy2H)@-M^6u4%%rM~|1$7JLss}cQpKRkQhYR1
zOi|z*Gu4+6RJ!>HPYuTK{%~;Q55(RL@&}+`BdquCB@HsU#q2ALf@Wr+{>U`?=J9-J
zo;9aF2K+dLzL%ob1|>4N7cmkPj=8;Z|J{!~D$9LkCXGnE*Q55!2oEY|DtqO)Cc^LM
zq~jx!LMB0IOb>Q<Wr|gz2oJy;RE~xSHiJDvxJstoQ%G+Ciji2;#~)k-@m^uQz%-eD
zB(VBC5eKP20uBDOdrd!ZZ>nvQk}X5$imNzpuu9y92I}Kei__JoF~ScB^lJ~Xkw9K8
z#>J!C9OitVE%>s*Y^-kOVz|GdR&*_r2xa9_OnF8`ta^0Tyk9UYeB~|q5W?TdfI$3v
z1DaYw%J@PkZ=zbnZf*VZnxONehSXV%@9Eu*I=kLhb=p5x2^2gVFUKZ=PdrLyvu-FY
z>iS7b_uiiU9*^_A&D+gj_cj&ESjt);dRtSriQF;hb;;-J?>b@9g)rnfyPUc2l;&58
z-zxs7a(9155A(_JJ%vAnMvS2c9eR(JNL`fJ#SMsleD1hOYYJL`@_D`n9{9Q8sv<gE
z#ozlVGIj&f^?zH2h*zVfnoIxJszSs^-V0(O;pqLAJ)2ho@x(kE2B_scI|4SJOEat^
zYN<&>%+P!Q#+E2NGKsX5!rB1Tg1FEU7|WuV?3ib!%q|!cnHLje91kG-`2EXBMdtxX
zr-TA?2bBi-s}z7n3BY}+%C>(SFLwVffDlZE7!dWsesy7KRbU;0Gfd#|S*KLUX%s($
z84Y7y$xyaOvOn_Pwg690Xb77hk!l&9FTMl_qgcU|ZTn?jdXaMQ*n+xWwxOguAK}OP
z&w&&CKNbm!5Cb;SUt{*d2Rm+}S+s?4#=eIS*OP}<BWJ$MrA+zW`Qv@Hl{^}+@p<W$
zV$QvXcnRLBe@>}m(zFr;v#S5-(XkU?sIsT$d2kgzi&LQJW;9C-BQ)J_EXdlAP*l8b
zLat8hs-MQ94qEJ&{W`OzgiNs~u^&zA!Lak*xa~e6sdggcU&+bDa<IWuNp`EP1Bv%&
zW0fj^=UtPHNORV`g{R(r+VWHfc6!p9$L8SPe7ObouE<p#TaqneI|GiL3Hb|M>&8BU
zrB@k~IhpF`&4+%k7wXN0`)>L^fIVuMuJtu0umJd>)vhisj$?F<r9mX6lQL;&h*IxG
zQj<tZC*ooi;>4_U`Ho23?Q0lXJWo9Y4+V8T<&mVvc5I%UZE^<|NuDO^`{dMN@24-n
zNX-AT*@{g64pkwbZy~VquC*Gj5*E#7Mo_|NW?5y{W(z+}%HcySk?K;zjvw{Y26-=k
zgHPh6?V7e6#zb~K{QdR=`rJ#MC7QhjnYW#3KWV7uHGiJ27S~3YmyQ3lwm`R)G@pK1
zW3XPF*SFn>|39N(;sy(6PVZ0UtNwWtbJ<-LMH<MYv}ltFi>Tt3B!%hcjjDpNwOK#~
z4nLW4zW9*_8UQ-+tb?}fTy1=GdaXqWG7>W%%Vx?80P6atLBAdW0KG~jB1%~+;qM9>
z7fFrgs|-jVarZ020gDKoZK?8x-4>1pnXf4Y%~1DIl;KJ4m?W!OC6gHJG`gp_UH5W^
znnDx(4EHP4(-7x26m9t~o7$#=-;9fhfpzIx)gMK9_TT!e2;EwWHW?tK*!O|Q7!As?
z-|G!AAZXr33thL2u-jYLHDPsxv@u0V!>^K^Yw-D68cu(hay=_k{CGR?{P1ic2JDBY
z<~%;^pjw(t`~dly`=5wj=$qdr-ob+5hFHm?ODSDcFy7}|YItGk+zHRf)7?zxO+Eh6
zlE=cBxqXoIQ%SHbE-qe)ZhD&ZKc2zCmOUQU748*r9N-91otF4A#L(au44%BC&evp_
zE;ir)({h^14`qF?it<7%Dj1ht;d1@Zg9eKk^*23{#{yZOdDN;%?jS&xil2#%VT(TE
z?zH>tP6k{jQ8zi<?UhKo<!8@=@tX)OO+)a_D9qpu=JYh{bGH7Dhw>o-`v*nfMaqa5
zW<ywqrm5WN?5W(#l%SFq2Lm~KMM#YbiCMZAD@1FTtpj6@oqu1I^vbMxciaSCqaNz?
zV@rEAHjRGkx`@~!08vPMT2VlX8O<F{E*nK?)Y;`~AzAuiBh}RZ-4{f!F3lG!Bct!K
zU93YQFmM8)pZR@&k(pnJ$d4oW@vI8{C3vo6+G&?e^g)Hp3px?Vi?6AW!)ky308oT<
z@J_oZsj^R!5u5@wQTiinQU*;^-fcb!@^Ar|BH2S7i>QtWD03@}8jix=^c~RrS>L<o
z7#btAXW98`mGF}tOF<+Y(@g`wd>`Sgov>U14?&NL*@6Z$2&*#~ye#^$_yVYCLT8I{
z1kT&)79_<yEp|bfKROv+<empx*CVY7?(C$X7tiVJ$jEM5$OFa!*EBO_Nc4CT*Lz>1
zIp&X$D`C)cXQ+b%(Ep^xy-&gi)<Gz&XzG7$c(GMCVfYlZ?)_Y6kBRcIqu(_w0h@m7
zq2*?3cKf)i%54{0h>WCRy0YTl`GJd#P5Y|lx2FfP5)k7o;oDQ*e$+cYjg6-r<E}eT
zGX1#~|B2O|wYD6RC5l?ihx>??$1qW8DNl{z9;drK>c_lQ454WD)qjx|b7<xTw}#|X
zrtdoZ)EB#Kr?ezT_fK0`g#NWRLPwV#<`8>tTbGO-@F_!mbC42YF{aBL!>)27vdJ}E
z5sql0o5tXy#~{-6#4*!E|G{jbCX-X%fsKAg6(+6gqKg37aM0!9Tu;T>hyNbJVi(E~
zeBx(laVlCG9UKL6ngD%9zDS>d_oZhUbNgN5UkF1l6_7u^RO^D1ciA7JeABp>oOl;P
zQnnEA#6GHOcDbm0diCuq>-iUMY2MN{`GFBXw}+vwo1W{p6Xpm=^tfbjbn{OR#JIg9
zIII4WM9FRKI_f@pL1s}^1wlF1=)egT-pe+X`-UV%L8+J)9!U)=Zy(AFLJA>~#)Vb%
zA9yD!T0ku`OEOUB7HP2~ii{nv)%tH+LMm!_CiI@Tue5<^k*Hi-W-9AI045~z9S!Q^
z1>>;CPc~a1SRSt!Q|YU;mJFJo;n*>AE@o8|fCeB3u*obhYK45S0z{2ry_-pJYIZ$f
z2a7v>KRGroURo5hh+9{Q&OIz@Z5k?4Si!&A1mrL|#~_a`lhO}F$>w(=xtWCikt=f;
z<uTk#&D|7?En#jl_ScZ!k6M~&=U2$S8SI%iC3bsDs;GUGllKXTiGE`weSO`OK0uT!
zndY+6>I}~bhEI%Ywo(i}5hFmyuDpgC4ME5r6rKVTs1Ckj57QmXdD=GrVXlWQW3smX
z!&|#bkt$w%B{gtNKSOb=Re11ym*27B1$FR66Sm`t00DRauj{FDCn!X!`_iHV7I1qb
zZEHpB62%MO!=mDYGSt}VyL<`T?+Sb*+>s96dd8y4zW4v9EeW5+4M0UPRzMrOeq}nW
zzex2aOFUD$?}r$i2m#4f`6n702kp!g1uOW$LI2cF(Rgcwj_`u2Se7giQ;tJ^l@ptx
z`WH|ArX^c89GPKnT0q?FxJKqjy?>|jOzfTi2u>Ax54I0xlZZaT!nF8mTlPzttJBe`
zB_1fgm8A}}BgTLp?~S(VdZ93}-~ktREw`_&=ntqoiF5MIIGy9F7~LaXJB;RTDupRb
zhpVuM2Bmq_wkkLr8aVKJXOLOyig`!)`2v>Zz<r>Y?fT;A!$T>M^y`n#FT0!<0j$K$
z`|83j&OBM1e0z5PmHY3EX2To4SB1~lH7H|M$mJZ0LhWPOSa}lly4;r1KdBFtW-RdS
zk@pWCBiG8+69awL!zGFQQAiAlBQ-(pSbFPRs%Yv4w7qGk1~g&<QsbmUb<TU@o)kR4
z-85gVSa-u|!WG|wB5gu2uuECj`jiyB5k;K7yr(iQDUx%on<}D7*bb0+?MdVz-+v?P
zJu1^(ltM~x(>Gi`_?D#mw`0_|P=5{?vH_{uV}bxm5K6oP=8aLk`1G#kD1mTe86Yhi
zoObV}%`J5NJ(|}Yibc&ENUr^xoY@0nOOsm`)emo$2@#ftH_wGUPcZjw-G0A$ih9ka
zhgC{PDW-`6whr!cG`l*l7J>zg+}xJ`fdzAcwmPgpy*Wch4>sl(x|Uw;Yd|Krge58b
z_e0QJzu+Zx&ea2X=RWxu)Mw6cvLA^a?)6nviQ~=I1v8*9_$J$&S3~+D=J|9M_Ag>B
z@zNv`DG1SZZ+f*WilB1G7)}n!{Vsrs4|KukAoz;+^DFxCRh_?xXObmL%~@m<sO4_A
z;yLvQ6hx%))AyFM^FP*mzo%Fx*n9LLE)Qx{6M?DevbeyF$znE}BT|f-u9g<)ivWZ_
zAA~5?ER;hM^gA|eETh}u$pT0H{;=w2#fv-cfV0`N8fxf66TJAYjHGK%XC6KH!F!DQ
z`DP+hjQEOv0*eff)VG9`zQi=Og3O9x+)~^j?mQV$;1?OJ@9e9ky-TSPjS(xsHA1(w
zyEux|^22j<6e=B`SDAqcM~|&$^((J$3oJ{q-yJFbn%zBm|4NXgD8ybI<pMiqG@pJq
z-!r9k=D=iGn(yW)+vZZee0M?-s<;}rAE!bbU&AYgLI;q~1k;wSBazeju#A6xN1AXK
zLC*e)Sw@^AsS1J-n(p04gd|a`eE81EQup`=S*%q8031Vfw4`cguG?~QK<H!|*p|lF
zONNvQNY4FsQW=O>ubDxQI?;$T6S)O49SPq41xT$uEKZQg!Fi1@+-pK3$mL>KQ!Mnp
zodEUSv0?ZNH)S33+*q4<C83qb1Qp<icuXd2n(;D(>tN89MYzQeVDZwlaQ6cG4-(6M
zo_XO8c*71S^q$GM|9Q6{_)__z8?PNok^ZBUy-`I2r1cI&D4!Ez6HAd%{S-AYD($m;
zbBDNKNnm{a__QEbc#5kOQqjs8b55s2(3yU)Y2ZFZ8voxQ65;l4BD5GlT*s3?ywUe4
zAtJ1A4Dn;zI|CoM?cx+E)2?H%b_P~u$>;16nj~<nk%s@l_rmXyc3$%*hc1K%S0I$+
zeCmnf)rTLu%%Wjuf}jamB>Kt%`Z`*nZMVY@7TpUA+a^Y-ZoSfuL{&0iBSXjc2KBy5
zh7xt>JXI1S5Pr*TF#Zk1oLS#!vM2Ze)-iKGb(O=@oafi%W!I-%9q>nWtSNXzcsak_
zC+MWQ3lwyW+(qm6bXz1miJ@csVP-D&*Xuo-x_r}sz*_A`VN^5ukFYOL+ok#w+{@|W
z>U4&mSGpaf8;8x+sdA11ue)+*{%Sl9;{8cugOBUuke-cn9zav?UZ&SWOg*!*neYDV
zHM>r`<kU2t{~S?v!#`&_EZg(I_d^~%xly!L4IfZYQj2~&xPGsRf=16QzvHJ`Z~i*>
zE>5efU^n2+#MqDQ-Do9yApML?TH&U0TSws&`iQrn^p)r?YM}R)O1a>y5q6E`C>P{I
zmCX+s)JkY%cb*0*nH)NbT#e=?G@NzB^c5=7%Sc5zHpj}h@t=ZvaBGhcv=}4g-FpHs
z->k33gPAs%!7Kp(zzth7oLk6z7xXrFSf>9j;biWTJu<wiLZawwgsFQWqP6(w?^&F~
z{10;X>bG_Q+@)DoS2m{hv7iB;%}?$=+j~ZGd(<Hd&i;w7@9=Wk9D>`VeZ7kiC$eQ@
zI$q;vd5FDD9Mj^vDcrFXpn<`2{#W|(^Ia@npk}Qd(WRLOkVoL?>oz^)!)Hg%{Zq%z
z`@Wr$pmj?1SIjvxV0=<`=yM0uiYmpp5_CTagJ<6rXgn*Qe%yd6gYn6NH#ecyW)H*F
z(lZl;8q&^&^mcr5xTorGav#(WCr#Qvr3NaJVYHL0=+|hNsBFIW*`*xLpkB+nXO<~A
z+&^L_?{*4!6?mTIH{bB%z&(hFm6|LP;s5&Pw^QLiO^AP5+I3SpQwMwqLV?h-)3?kX
zP*WM`8zH=<Un7JgfoL#Or*+>GnKOras<ruk>Z8I%Rg$Z0#4Z<OYGvPlS;cG@E+o<l
z?`6#CT(wUN`O?sC_2U3uTG#EPdDdgx{dp)lb&ZtI4KD7e-ShdyPIV`uDk_@N|B8C;
z9r)0coj1@%ry=d38x;5q$uGiV2W^&j!MD$!{lD1@kzHT4a;|S-ncx4qt7h{+V2;)&
zx?&}r`;KqHD=R)be}2*lNe#1`!}%`q6{%`@<$c^}tA;C={e={U5v=1iYT!{ib<n^~
z`Yts1xwH`h2b0B!@9A$}8d~p08?3h;@<+S9YQyXCc72PG+RtsFn-<scYyXP8_^Vdp
z7?3+@*U2pG^K5WAo-YuR5a1|AJ)H+byoH>;({v)D;jj?wT{%<<q^sX(<UObY18FX%
zrkL{GAY8(NR#xPYG&&w7fTU7GxB->eP8#)fZcM*1Nk&Qo578hG5yqRGlypl&GK5%-
z#0<<=!_T2a4@QB6)9eLqQxakX6^;o5Rfj$ClJlGq(~%2q)sASEUl*-fb5c<b@Z&oe
zg4ew*3rOn>C54XjLS0qWWqu^Yz{&lTD=}s*9#O2Cjx9LmZ!<`Q`ndKp!W=c08(04U
ztax!@)%rymz-s=?xc(4X3cw*wwQmtPLAdpAW&natcF+bN{FsR&6=$xXZNXtx%`v~r
zR~*xLTEEGebu%wUqEGf$b3AVfp71uT8J^p`=&v;1fp$JfdS#tRXJ7ii2W@#)2R8Vw
zI+}YFXOIiKrsu$0yt714a9@U<AsFRozdG%nzxFOeMh$>V3H-|c0ewdqi38<F#4C^H
z;1Q=|lTYs!HH4i1o-fXce=nuhFb(W~?6hQ0G`k(_x$Zbw8~9yN1^}?cD$C32>`G1u
z)cdC&zFZh=&v;sRzUKA#=<{KuMY`Vi<~#e&VdnwkP98e7Kq}YKk5(^n4+=nhtMDuB
z9q#9QEWQ4Ml-(cWbKFuS)geEx98By(Yn3+Ujcr?gURjWd4LLX=Ui>+`sgsR=BK4}r
zvSrk|4aofFd`C2vH?fG)>Wn|b`J2TJ7jOg?^_IZa`{Oi`nAZ=36+40~X5aqvf20dB
zJG@owl|Q(epZ5ozL-t_VssLtO?t<q9jpyhVa|sjwfM3Y_!SEq!{wo>tDr~H);QMNL
ztZ#xDjlm;fJTHgj#gWI3UGRR#&e=lya44^Z%}TBL^JFkz&fUdCYmS{1=J{<C^zk^a
zd(+e+R0_U?F3sQ#{FXt|b*kmP7(@&iWcB2@R9t*xtpy~)T6t?P))w@?@yfZ^^ndL6
zS5ta+m{zHx=69xKKIWxbCx~ctRk<KY#XF82!oW?H!#$;{#4TWXiOjCeF$g&gz<e?i
ziE9jL?Xy&Dh2Txsb{f~n`!<jNeoHu-_ygE{L&UJ^QGyt_$PGtU{ZN;;4vlE4KYZVB
z1;JRc(wYcm8*Wm=dF^D8tGR~6kW9NXq?FsIv>ss<0!VbyaTGEelCe<$sw-H{W3-){
z$aoDK`!?vIQhv}=C|+Fv%#qZpDhZ)|4RU444sV<=4&&!)frN*@Ffn@-`OBfIzDnW|
zh!w7b;(`rA{*&O>Ju~%0ZidBkUVJtsPECDxOPvnAjm1mrh}+e}PKiYQE$|#bkhHzc
z{lIon@M4WhEk`#d;pQ{KbUL+%=Xq25G0fKcp`!ZCJW-r8=JBGi5Q0(F(J6a$3-UWZ
z(7o{Jy8OpN=tf`STJ9nX_#K0bS}k)xHXfPhU^Dp9{vv3<p|-)lOhtG1&-1<X<5s#@
z)nuno@l#UpQ=Qk$BJ}>+wZYBOB>S0geH+?wqu+k0^x6rHe(lA{2*6xvP-yb%aFPCc
z{T45yKj==n9eN7aF5I@Kap-83DE$Q!bqGwe%R<nbj`=7m;vxjS5t_O}VOl^)M&sKO
ztUwIS+2+aF8<#VpdKHC?{M(y@=?dvJ{tIP&Q7Uin4(I#pET2|7>Rvk<NYEpspebdE
zBvDu~68E7IqtGXDSJ9bn{s}2Z_!al#bo%B6#X2fjgyAUd{=WKQi64aR-1QLnY?eEi
zF?Ow2uW=My))%bXHbvEZV}_;IzBi>PXHJY*S?i^U2pWN~bP{5g%>|Usnl}f&-m5lT
zM<<I}21oI1-~``s@TbFYN%`qx3wOspp!Y8qJx1pL2b}B3{|B6jaSBmmP|{hFtPwkZ
zlz$BEG$Sh*NxpW!qUOK>V~<q5qys>v2(^*t!Jeh4{BNp407MmS?139pSj7<oout_{
z@}Qtu6fsaFBUluK!GXO<+89WA!)!oFy3)aC1yI=5V2n90YodMgt8?)@7Ry-gB-^qI
z@t#-c;7`&DTao5}rF@_Gz->T3uUXlNB**q*ag7Q9cS_rFnL*44FvSQ^eM1estfPy^
zfP5BfMl%-FH@fLO#XqF8$2e5DurL(6H=7G^KN92)O!qzX4#=Qvea3?fqT4{ABs|lL
z_EBHxy*2N5cWli-pF7viQee2ThiAOJ^g+8l)K}3-w}_bfnUucW5oHFZir4j!Q1=T$
zboZ4SPCL?my6{I@M1cDK7|KVvf52{jnHEw@g`%|foCg)ifnRiql!l)<G#Lq-KP8f6
zJe+hcx68pa0c72m7hR817mYA@6@MVw$P-lhf_p5%gl-xJ4VY<4OAS6=$Rt7%xjP+z
zLl8{`ZJ538#fBnv1Njud*5rOVNh96+dAKX*z(`d;6?LiEHFKBf=1+>eB7!k899J0q
zK}67OPXfl9vN2g&m^ETj?mi?TB_``?CnY=fNgEd5EXsKZk5*lxdqp;Q(NUh3g#g(9
zc)wW}9yUWT+CIbk%guZhqNE>I5OhH|Cca|YvC~WzuMdJzOpjW=M8M*H+;X^tbN$u>
zDtTjY?_0C-j8I#0lQ4J*bX>J{USNNaU1I=!7IR$~>A|A*O|xW1oFMGE63QEVO_Ec`
zJ<sDXiQm>6%FU&N*h9cvW_+7!BPAb&dh)wO{0jHKytW#k@IEIz3GJyZ<@e4`A!v|9
ztt48YO+@>c?g9I?S8)VDyv$EsXuyp$UrZH?^y@)SPO2TKH<;?d#%vq*uOD=!c*p=}
zF0B^|4dzL?5}2_@7$Bpb2#&ZvDc+B=k&FMdV%aU^mMiB`VsUS2bh&Ig@iz6qZ#~j7
z{3iv8kSyej<sy^CCkRie^W3p3glLcK(s|X06|Xqz9hjc<8{->Y@kTVu9E$SbjB2kB
z2!+o;^J=KBoOrLI7TZ^g>6cc8{O6W=JZ5<)#MZRAZ0=1r65et)CeEx>=n#z753B9A
z%O8W!PgD+nKj9f^c^A!a<tAMg;~kPk$GASFbm5Z)JkbN0rEa_N(p-vY=>vx-R>giD
zWtRbK_)9XezDZuj-$h7&n*Xjx(T{<Up?F3ep5fSS`mGBc?5<sm=;;lQB_;ysQ@PA|
zq~YV#5{QM>AqVE2hpRi9Fs^QtNMIU)8Y~D}oj&onpJSkJ*sfnAdx+|8zOC`_@UWBf
zbkrt*_aid?^4>L|5|usF&}n^XX+=Q4+LS+-bD?J<cs=N15*ZIAI{o<~=NW5e|C19O
zFAxf?g=rBv94>a<-+AY>o3})bB6<8J*nCRQdGMXtj~db~G$2dcap&9jn61<|ZWTTO
zUS=l|HAsCD8&U;WJtxZ#9dqjkm^&jbt=^W3<K&4!Kf>i`6ko_s4dL2zj88IMlJ}1+
zEOb-#zT(9^cTj}mzgPG3q3ub5TS&rCTdh>-ssVMnahx)a#7Ahyj^f%`qh?JANwzuB
zyUub{@3-Q3!I$X4kQL9nX=UvRvo06-SR2&i;7bUlJ`^6lkdy3Y&V)n{;F3Onaut6g
z_`LZHFWCv}dJ?Z{ed;Y*5QoA8f%Yd-o|)7LV>tyzEb*Grh`~bClX{YzFRUedAIkpM
zy>*?=M@T42fhaFBnTJ1u_UwX~=9q&FxCi4~etx2v1Q~}%BKRdTxy!1Az>T@+v2>8u
zuj)6_h<eQMXv-9x@?l^dj8ASQ?<URTV<j=*Ow>KDEOb4X#)~AQRi)2i0AR5Q%Wo1Z
z&Zjv_F6pccE2rX>A2!-WWDk=)zf}5<&k<lq0jOXZnt|sWK=_x|0UeybtaW-22O%8N
z={`>Rc{tCQ0><Nd&rRAagCj!-w|B+_0UxXv6D;)m_br+#4=l@Kx)G?jUJR&4)ST5B
z_m!7UzXs=?=*^aaINXghT$Ovn5xj~p3~agE=3YK)kuC?u%p$tNblwKNUVl==>jFaz
z_a4SvZxt>~T>}cp$A_4C6{HH+5B-m=>O2vo7_k}NGq8A~+J9gFN6sz0>7NvL14v&T
zaHrbP_-w0xyhZ4(-g^Z;<9Ccgj;Z;2kiX%)j?0c=28e2ZmVRoCZNq2k*;XYP`-W5;
zv#PCyCWGXy|Ip-`+56J>QLtL5`nkWZRuy#Iv*xXASEX(TE4_$ZgiF8teO+=ylH)Vo
z%4$_W0^^pxx(VJ&Zd)(OX|WBuCV^Vz_0ybQ9$k_9b=?<f!}RN0IvVVQHWb(F$$+TJ
zInTeJC)U1)E9++M&zPF}d)8t6D&_bU33g<NS`Xg{cgtudw@zxdX_UTjgsW%U*`AN8
zS<X)Jxb>jQhoEiB$;-qhg7RfeLfe3$9eVn@`ktN$FZ43)z=Qe9C%r4qpRkb5o9w`H
zf5d)U-!7O(sd{<lbt)z$YOvfB?_c1$!`0Bo&44*B52at90pj`<b{=DRvCi*n<&xt+
zR=KCfn@GMe%AK~VOjg(-oejZXB0YQaeSXR^;EAU*x~Ics1nPZvpFb~rc>H)Wym0Lq
z>}UjrMmhlKb1s9p$;8n8(1|mDfD|%6DKNdsUu2`i>|Ze4wP{*mz{w;()YuFVmno`B
zj-m35F>pH^n?8x;{U$ge$*m#AdeaxpMO-qL%}BBlVe#qg%NVzc!q@ZNt64*4)t-}9
za#$S_J=VCSV&-#Ci+y)(G4Z!*Q|3;Tw<|1vz^P(9e2eer)<9&#-;^3LPeS5cc)ffs
zKm|m;`hyXsSxQ_CP6>agb-Tyzc^1<0%^5=y90a6D#+{)zAnhOn!C&lbc!3fbSFtMC
z^GGNa<UbXCV=?!xNxmkn*;8Fbl+}aRytl0lK8Wc7>9>=c<~QDZhPCD0LNS-cv9GlM
ziY(FqP{1ev9|$ogLT2^|`97hOvyBFp4Lf)b5MW@~|Mt!a(||@|j>FAv9y0~#oID8<
zN5Zoi343rzWR(MigV4c*OXnX&2xAhMYBcC%x55N4)iuArLP}=>^IO+Wjw>P0xw~yr
zfH2;syaSS=C=ndAViaqBNcvoXA4-o~YKzss&=E^Vg!ED#`2@vg1JvdN8YJ7-{@~Ps
z2MMWnmJ`B0NvOeSpiYDzd|M4k;MHdO%5x8fDA7Nlp?(WJ>-Mq^<~9<6&h`(EOCJ+g
zR%Xc0V)kQz3qXoWx8aOV_Jg(qIifQfL?f=ZwGL6U4VoZ^<&|te&r{c3rGQDcO8sDl
z6J+pjqfn_r6nY(SagyxIRTQqvr9#0*Qech>ViAK<wyLdq4w@buuXpe-ea)JEvLnfo
z(ow&yA&tFBmP75&@nIH6;rgg=*T?)~BfDo-1A|u?>AMGl6OpcX6iU7cTEWxxrS3R=
zf+MZ58E=gh^H&zQJLPI(RPq7OgS~%|1Wa#@&E1yALOJ~Sw7fc>;ufKQXn}FES4Muf
zMlIZEYBd?1HzpTzM)uDE+gq7n7(-m0%g-doe!+#u$#Lhz@|C-#yc2U1ETW4xF0Spz
zJW;f)8eRfy)Gs5*#`N7Te=kt^bnLe^O1<6?`76~=M=T#<<kosGLnbM{iZb?b!MLa(
zxj)~Kp69_!*f=4s@`7M|y<*%LwH%kF(i^TW<dQr-;h{zsDC-*Vom5WpJ@PPG%wXU6
zf#O^2R?_d2K7Uzw6n?_IsF}8gX>x<!vo3G2lJmHe|J+Q9(<rG*l2dmODcw_`HU3PK
zsFkHhNs4eoxyp_>r?(X9W(3KWUVX8=^85I~gYU4%#7gCz*O>b6^diCfhLP;HRpJ|4
z#>%2=0rsx+<-gfAq}vk$5J%$1`IQ%_0u^QY*cps^rHq;~b-Fg}j9CuUY%Gm2C<^(t
zW>ZsHnW;#md8P@ZA=o@`h-A7yVG@6C%Ug>A#)CoOSVv)z^%DsQx8=_F5pU<TGxbGt
zgTD5+F<u~R72$jG@DYLJywGEdlPoqqi5EF=g_5Dj-9w1a<V;B}Jdr=phK0r8h6=IL
zpgC|9-gARFP+T%uv||9c!*crrw&6s{jB#guZ!G`{8<;V<;M-KqwztJ!?c}2cWHfRa
zRb-g#HQt7D06sb3klMc%mK8usJ3uDgxE6^->b6w_ez1Hn7jUg&<;mBvM}4699nmPq
zubjc4_48FhGmlVlcg;gthKgOx$C>Pm8-=_VCC|P-f&hv#L-u#Skmy_9(`!fdPeE6%
z6t%Am|7=XV<dYoelIA$Rej@iC9^&u&#x+v^_t%*>cl3?aqT>&1Ay6S9TXPU9B=U|W
zUC>I*iLp^BTwE5%RC1v~-;8UodgMj~eBD-A{gt~DP6ql560@GzJLHlapYo~Z_X*F3
zx9pU^(aLt%E`9cj@m*?p=l*@!;M={&aU<A>oq*K-Oi=aZihnW^ec!cNQAmOm#m~AQ
zwfat!7hk@8_LnCVpeD*ki?7Jk0zlblAFIyFT^kn?*t=M8&017Me2PlGZi<L(qy?R7
z-3YJg0|AxSetOk}UHETa#3a|zNWCYezo|4}V;=PzIEBwGct=XF3m6FfWyp!1kbC<F
zS?ez%^5z7ceKELL!D2I*FGrsJ6zFb_&GUQcNpj>QUSaME+eam))zjW*nuT)=snmQi
zt1y8*7;!_%HEa=McTYxUSlHVrY24Kc!8Grl;TEL7wwNZG6F9eZE2!GoI6;j{Lqf%a
zN*Gt5&CzWk-9OeTdb(^DGioZLSp{9}hsX9aabxc}I$*z0YG>2QuL2htZe(2RRtuAr
z8+0K`o$VtQF_(~jB@%;I^>_pWrS4@VQOt9uwRk`E*vr!BKGCcLK%XJ<^y+KOzvW)N
z%zmAs&{}E~x;uTVrdf!&u0%nD+>?YPS7Qx0u`Xa~W5g}%+UR~=u~N8-XZz)ps5Fti
zC9m3mA#qO53&Z=EwVmO{k1+!US7%|4M5V^$LvUdiHO&j`DD=WHoEr6-p;31{Fz@s_
zfQRsO_>TMatNIfB&`SJ!Rr({r4^~@gUkQsW?otwhrsm>c6bFaYC$A(aE&d3RWO}4D
zyG$|uy1SHS)ir{2hf{B&YXaKj0q-q`++@*)V}_FYUJ&*Vu4zc&Pe^pp!k8cwU^M*<
zFq#BJ9v~wvUzA!I|A2_X-ESh^SA41x3>OrR<}FafT)h=RN|&S#<PdUMW}+Y+>5GEL
zj_4^Oj;GTKakA7;&N&8cwGHbVqv#%RAkSXxZ(fa;HaGsT=!wq9G7i9Snir?kg^V)w
zQm;CC&TK#f`gU0{+f?3(^iz`+N3=tdoDfN@Tb(4#_3)RqDM@eZ6kdwKUvcI?Dmvax
z9#|i-ahvlTnuOUlBzmW&7=fH)DU>_44xI)=LKFvlMzzZ3L)W~jZE4>6MIj7z=i46q
z<67rmq>Ub<?>JAF^#nm%SHg|hE8YuTgF$H$9(sSC&gS&ZONuPK%O20rygsu%uCk{n
zBkvv<nl=GR|1^4(_>DHhH&>-!(JNvvJq==jKHe?_KeR~D{>=O)=_PQv#x<r{banC&
z{OVBH<$X6_6)#79mPm_Yls~n`KI8T$&8g}(hgv^1(*@^67WnysWU1X+b4?9=*B=)0
z<MXHbE7i<6zuT#SncrpTITP$bUb|U5*g2i}7nH=BH`2dagQ8RfbzW8DXM(Oo2=C*{
z(<(=!20I51HEfo#UQwv5C~sCCw9Gnk1}&;qI<}mJP7+d|5(yLEe`ISO0@X~9#@a+r
z@^B0FnTU*8C~ynLMYtboyF8MryL`^$D=uP;Pe?Y^7(W#*3=-F?FUdAjpr37iQ>r#*
zS!Vk6Mx&Fk@reJ#+)muHH};7|D7)5_OSt*LO^g+c`a#t4N<_;e*>YG%CDU@6TiA)a
zvtiMnkg_vA@;A-h6wL{!fRxxHgtg~>qw*V$J7)YxFq3Qw6A#_}U>umRJ3YuHzNdQg
zqjTvrnWRN#1;Inw?^toTp1tGegYNO1;Kt_b^O2$)_-?FeTQ3&5P$bf5kB?><H70|O
z8zTYTqTo|6tL5Br)Be=fQ#PJ7k;WV6s`NeR&lbCE+N##0=@`&)R@O!rwa?@wP11`@
zSW_UElfOwLkM@^2$7SzGC`<_fPt67e_04FRkmaYp2+3pz!cpBY8b#eP@vPmBWy6DL
zUcPkU`L+Uz7qRHF-Ebr9%`{9ud|&mYnTan6b93OUt|}}(u;`%+r#SdfmHSX#6jnr;
z%&5vYFXYES78>k2fespOXaBJL)~;ahFPkqOqqh&t!x~qm6E*}=qHS#rtx?=cI#gWk
zFuJkb<`#Ax*GE4!J3yToK?SFHXv5Kxs+sR7bt815p0Yh~_XRTHy7pTSrlk^w(wj}O
zZmb*yf4sAOHIu0~>vHZYz=47qb_VFqa=aZ|i|b9plJbXghNl$4BGt~Hmnd|bC}a=K
zi0fr|Vd&qp$TkmJlGz7xkzrRX^F?kk8q7MRH=9uwF=32VvtlL5^^S~j_u+Pq6E391
zuKhRF@87FH^~*TE6pu(&8aT-a^bng%{muS7kl#@p;5+wJJ_DqvLm9voV>bTrx3?J9
zYv)h+u-+B0zFS&PE_Hw26GF}1DJ!NXxm};ESSp+od~ax^(}ql+RGcO(5F4ezt9EHv
zE4=oH(d^H)pAqFBdm?q2ly0fP+5&|l?6V@&St_3&(aAUF5~rF=)<4|zIa?%=N&hrF
znAYCSRT{n6X%b)|?JLz1d@UAu?c{d#D%A$X4B9Fh7Nx!!^89ImO^snjSZ7er_b9#L
zwf$s>pC`<?g#rzGMwkpw7SFW>jpiMHMZ6VCMjKV84~ODWq5@*Q*5aa9u4@_X{jJ@a
z3l6CToc8DR`CN17wnKe^={e6oNl}#8^(tal$CT>W^e<J(EBD!9(#r4Gv&%v;-sMfB
z-sRb5oU0^!RZF3?zWGp8W9;olZF7#igTW~8V7wjtAP&pQ#P|KOFE0vZzI0moK`%FE
z>)L8>VbZV3O!laVEv8^c^!T3%yHPHH`efR>XntYH{Cr0~3Ci93bUCGb(R9x7>)GU%
z*7`qr)FOgd|L%xAwsfzd7o%_TzjJB#c%xT#Z9&mIgk$t|h6WdmDaD^{{j`xymrkYB
zGd0U1e6MX|*^=hUN72JX%FrIxecj14Md3o!(E+LfRVh8tIdtwjyp|xsKSVO6W5r=X
zMoJzJd-$Hzz7!b9WBKS~IL=g$|B!O582(K_<l95^xVWN;$9{gj*qH_TJZ6UZ3Z2}f
z?RKaNe;G7i1VGw{Nt*X{cQl(;Z!I9{)57!7)2&sl?XidNt}98VI9^P+ub+-@@M+td
zjc!G1)tC@$=;tu0>#@*L@k4aWsV^kfKF<D|@4BBAFmD4kPPSF~adiXhp1}qkwv!-H
zMGwnM%)i}>qR!(^Zs#$tEwOPwx8NM&T2bBeAG5-JH7sQv1FF?Xh4m@~kTO3!Y|bq-
z9S&`E?B3h+-lqy_x+`Nlm3ep5_~PVhksh`dylrkF>G#=z70kMRseP{aATDSgyCJD}
zB0Lv$uEQQ{a`9%4NpBRVqDrF?S=yq4vZwnesi@Owo2BQPqOo7rq7SuRckGSHaXM*^
z6#j4V#7!P5%o2#0c-)AIC+`J2%BS1TIz<Zp&zqtz)*-r9kdQ2>sqH4D%H`fH%XB2I
z(u4w8k6)qKXSol$_E0zKZhBZpN0*zJCT{4tc{ua#;qUnuk|=%+NyZ)+vS#ms)%$GK
zpLz#sJc3*%AFE90$$Q?9%_4C55TSxXms45&&$gSQg~r!mCE4H0Fwlv|-Hg1|RCvqk
z+eK>6ZoIh$7=#1EK+xTs=#qI?P32<K7ErU}J^?R<pP%QHuKoVW)x&U;3~&9La@C~h
zVP{XXqC$Plj_9J~o^iL)f9qdroP%^EdRzNHwI>Hp_5;86ZORb;!AL;iYSSC%X$(Nf
z7bRiH7hA0EF9E(r_Erq$W^<?9^O`n-!-c!}Q5F}}*sENN%ej8+FK~*H!{R}qGl%Lt
z%<Ej;s6ZmkQhEAq6prR0F6Z!d6<w41r@Y5KC<fq*L-Wt+sIoMk4#E!npH<2(^=*$-
zE%xLmz~*r6o3Dh2$~~~br0r7ye|ggWt|A_!$a?xv_hJ9C5ji+PO}7G_^Q=-2Y-KJ&
z$?CM3U@@2W8jYi@12iCipGU?<ArA*((U~Flw@~TLrG7~{Xw|&Y)ED!IoInO^&IXj{
zY(oBGDtxLbKrgF*6zM>d!A2;#TRwC)UnfS}i0=n~u(BT6(VLoEV3F0hpQ=(K0wpdi
zGx&veSnkxbZe%aJP8AG2$i-Hhqd(9wYviQhKX*v__w`GRMg~CJ5GU5yb=9~CpQq+K
zFYK91HD}uym0<0G6)l*d7tqz->SXJ9za^)l$M&KS4Qgv^>&t8On))-x`LoRXU`v_J
z=%B28T1Aw}^;hY1{&5l%=F>we0ppLY<8ED-NG=j}U4%j64ADJJhr#}97>Xc3972MH
zYJ1V`psPQoPoCuXCyzt5;b1hteGE@G!=O}Uq^xzDtx4B+cINdx@j9s!kcIHY)%s;{
z-QJNCiPX=WreUtN$%UFn0-^%5pN}gJ^r+%pZ4L#fj$O%*3I;eaYOc6Zc;2wY*t!eu
zU^T(_l%zwJ0f9nHw+o<r*W@SkF2XL~bJl{o{R<6PGiCpjqU~)uCy`F>`w;wBP#;b-
zQd+AmJN8r?L^-Q?Gvs*~t2|;C%Hq9xmW18L;0fxjZz<a_7*}L~hKlN=Vu<yI$(FL~
z;n_6=^|=rKOf3(}`2##EM^iH@h3TgvqIM7$fc{{%hN`-w<6R75f(duFCvBCwGB6;w
zz&M67I}o|J6D&Z^Ip@hPA5r-}T}#OzW$lG09c>L&rU+3t;Rq+(R*UPW=BO(V+B{0r
zZ+B%{DAJru%XDoLwR7uazY;~vpxI&q4XF(M-ZT@2^B2I!kD<IWdH0N{#FHO|UsGS+
zKs0?_N!eP~VWG!yQoV-}l?3*qE4zt8T!We)<F!zF_S||4@LkKS&e0(XGVM&aH}yB*
z{e&arY4XCT7sdMtXj2&&(=^mTUbSc;Iy!((AwbstbB1-(%Pko(>qLR>*xYC$uqcN#
z5&<4KK~n^&xOcP07pOprl!OKdM{$d9iS+Us`tYv4Eu#K|c+u`@h`iBImlwK1T$Y4#
z?_{TXat^^8@=Li_1&OL8Trc&_+chPNtHdoLY8pI~_Fja|g$aEkxRf(#MEY_ZJUz?0
zabnE>n*Ixb-Ws>E{g}A>DpX6T>z82p4_2#1r->o}Jyvee;iIpe9mCymNEj$qBFnvm
zvj02a^CKvIPY&&A1faF{=PK@W{pl@hZ-3-jP4pZDV`PfrbR^7O08PWI9p!${g3dhP
zDb|OKt=AGj8f(LkY1XJs3tRieKa>PGj`2|et!yhdTDuCG;~;%-eBJ?*N?CB|sNIAW
zq41EoNL3j7Z+?a55BGlaZeB{>YMMYqXKa}GlDZaDE~sa3+%*d>Ss-1+SQ@HVRPdW6
z81<zvxy~FP*X|2-z~4ZhuJTWN>RNXfW<_JL4(>lYD&JG2qzlRC9MR8W>5OZn3Gk(w
z(9x;yf*Mm1=MSmbsnN$AQ!Th(SZO+Oa0&taG`<BnNiGr+;Vin6L5%8#HecZekgp2W
zN8539y_(n_AE{({Q&}XdV3^mx;8^K$bxu4FoC7L&dBUm@?|?0mLU=dLY)5W5IWt7S
z2_B6OCuvnBB+N50_3N3_SC;(U&tn}O&<@iss0*M{0e$uhf`vK1;#2%YO9_9@_EYdy
zSxtfmEllS?NbSVt8dAUs^v>=D#HdAT;E|^fhlCfiG^Nm49jvsgJU6xd&Cfx9m>Lk#
zyomR*uh?`rxthYuWg@^v`u!sos<;7EZXj)1+{A_4tswg#>(>)iv@-MnmJc6isi)}Y
zBhA5jSMvVv;+hWQA|=WuG8ll9e`6K5ciz%{8L32!Zb<Wg05?I%z63Tnmy{dh*qD>_
z4E$|^UTv~9q&h9yhD!nz-1w%kbs8<FA@+VUI^gGN@dC`C;cvtPYs8CZs?h4k23{Qc
z^d;bdP4@_lo>QDC2i>AzfAC64*vCj1a!;t-0>K_-rn)OqoClJd_l!$J@6aD5{Yk^8
z0rp0y!0@^n0Kg^cF#y#Jaj4U3#3r0n*_BkA*9f~)@!kZ8;Bb3Uz6jVZ0B<z0J22uO
zHCC#^a$`f407)2aPho#jE;{dQas8r8<h}k!H@;3_j4HTX=u2+Q($+yAo+pl*Ansdo
zs~%0>7XplMov<8gdz-LHt4kEW{YRcyBih<KbmyY)O94jC089Rgi!aK^i)*F-xc|YF
zYOf3D&lBg)Ju4&J*6lmRTVUGg-?&165o1D?07g_ETq8}JGDVyL6<}1^Wn|Z@aGPF!
z1vVT%P>)B{Rc*rOs?<;e+-klW#Cvggf$Yv)^|8vY(g2eF+cB0|wMi-KLStH1Kj$j9
z)p=Q&o<1K@fR&Z$*Fsu_&T&~6jzJ!x3x<NC5g=V00Eq_xIp-W!%vo!py~-UY`k1SQ
z`dU%OQr}(0j7=-bS?g)oxxM&}s53??G{PG9?@miO8!*B^H(z&+IC0X%wD5d*u62vp
z`|&3k?nZ~}ueuU+`Kd-{sHJ*A?ECzS40k#VB{0Ip_&v*(fc=rK7VG==cJc6IPhv5b
ze!_<3<>{$aOK!`ssZvY82<veqNvR(ZE83t|C%9JB{YXi}tPhQ2O5fz9U$#lDSz1R*
zJ5{S5`l-`;dseH}E0>SnpH}B$B`{)j*-Dvo8{-(CttunnB%A_H-()T%ZN(VrYea0E
z_6>JKwHAcT3hywVh=4^x8z9my1(_Jw$fnu|1&m083z&C#0v^fUkYGdvY!QU@fISj{
z8ZCc|M+Dn&Vh$VZv{^#cS`>mKZjS>ep$9|^aR_`!{}9+G*m!TmCNA@mpbN@F76^fs
z82lBm(g1dg6!=TL23Oh_kv22{1W*RBvE6_JFia<oQ^$=saZJWY80y2wY8e1H)X)E5
zsUW6|qy9AFw6zQa;Do44B-E+A<kjxLh;PKPvYb?<MPQ^i7?G?V_Gtuted)sari_st
zyZ$6LZ)p|%+9}uw)Ni>Zn}%4^fS5Y*gpwH}+1ME$U|q5v>yvV&SYRa79}}m}JXzcb
zBkV?@{f&SB)dLc|qaOjCabptgs)zlrhaSO6;+`~M<h0qd#NtJns>-URN{c`U?LPK|
z!ccvj4mOP$JzC5^d#;#q;uQSGlT58-h2G!$q1f=^OM3Nebzce)uMPy5#Vu#Et`hI`
zZk$Z_y6nl0sN_J%H9s`Y;VW5=V$;r5?qbubhqrGFa12tFRv+w~aLvemr%h6;4)fTz
z=JqLBCEmA{uAQCVU6rnTJ^e%r@Lf>{Mk+KKiUCHZVBvDh_19*U<jZfq6^}o=E_DJb
zBS6xj)_@Tf`_{j^PE0=G__QjK;{M=oAD4@kG!H!Rbk@?FZ_HrUR1z4e#TUD=-75J!
zlphf_!Ohi2G}!2BD1H;Fl5JK`d$0vm9}c=CPWsR|p4n+Lj<=x(Mskfey&koE&)-cf
z*qx!~td{F^q?f&%4Wm-prW_k1m3}XGLk&qYHp=ZN<p3jOjKrZDg`4G3ur>%VfZPEl
z2pJ$E!S+C<sWM9%KgWG61hpmrP8e3Q5O5<|2#84bhPNf5W{7IF{7|PwiNU?6GXgbT
zO(`2AK5W=iO^Wo##ytTNCpO=e+9tPD#T5dWkx-vSe>VUZUMV!V&dvH5$|{*FNK4un
zAzLIFhT%K`fHYuBxB%d=%nmR=2A~on(%ubKQLs_c)6HPylt^&I;Y`@zi9^MV0~?gs
z6_;wh!rg5$ocd#TV59}Cjy|wv{3HA#3_!}0><gBk;Kq4#&&*<DWH)S#Y?8o8AHsP6
zTJ?ef<iHrAunni-(ACNtZ;;RWpL~oFzuzZT+<J>t@1(~goBfHk>%~{!eycl|bYCnm
z!c2a6=IP>+^Uur3kG4X7`N-;&`#-E&k-;!0)8#*Z{Sev|V;l=$rONaCdFO})03VD}
z4PfMi<0gn7!SMR<VJ+fK*e7}7`7L7ns9|YEAq$N4UbtX>+8XGiefz}P4XtTmi+w0`
zPE#n*PNqi`zr$c>Sd1OnEPtoPmWkob-j#)xwUqoyiBoI9h*j;U|8|V+Hk9i^mL=CX
z%~fs%$QkQ16wd^JeK-VBm(p5*V^9Gj5zI*h)i(PBM<p;~)o-3XQH(mpT))H8dbRSg
z%6)g0%hIWQ4Qo4xV#Z&#nyR(3+~>eNg{m!e;WwgMz{vjt00960>sL7E002M$Nkl<Z
z?7ereG|5@+neE*<=d<IDW@JV}fFxv)M+O1H2qT1q)e5_p-Q~u{x`_Q}BX%P;VmCJS
z-i;d<_QGCZ;lkpDK*)g*LJ~+I140svyfp5c(ah(LX>a!VWp$pfPM_-P>h7xU>T@b*
zzSCv0GT(eNt19b#@_FCf^-*W^@@K!mke=~e@7L}G{OOsYDVe-I<KOmY{le!y=ifEM
z+n#%5f89IzOP@F*!c*_N@5sJ?Czs2aANt<!G50;?$^OH3`>pSIm-*+<9&PJnrXPIW
zcbl(!+EXVFo8Nis+sqfg^3`L<IdId(3+88j>_<$YP&jh_qyO+p^PazY|B+)+cAow9
z&or<5j&D1b^=<EZkNNb!+<)x&D_-&<^YRzH@YwO+{-ZxOU;gR?$Bw_~`Oh`4`IeU-
zJO23PE9U?CPrq%3!{M=Wr{zF<)Hl^!apleJw3?<+Dw$%jG<V(mgPz$c%zvG$=wR=f
zX>6Z$yV>tp13fd)v-9XZz1#1aonkq4ItuE;$uV`U+c9U$6)QXEDz$oGJA%pz%$VT8
zC}X-g?CH&xGu8a+@7H;no(6r+Bkv~XjWTX#wJ*+lJW5&pzMeF_-qibXmb^)x-_yHg
zd(L-DYuGon6yKBRyVvWQPN8Vp&8{g|OQzT9n?a}NsZw*uTweb)pW~c%O}{@hy=L1K
z%6iXge>5EFk1@ku-{eXKQ>|AtZyK55U}*Z?zA09`_i?$LH~D<t>|NeB<*lme?ze3n
zTjv_)V6SaP%1<iQ%PPNPN;Q?EV}owbZueSUQ_*#uy_PBJ+(>h?e6e6krM&4%=6<^;
znTl5STtV}E$xVI8-Bx*{-oTb$*{+#fbKi{WTc&+=&y=?7rh9Eyb@lD}ks6cBd9teU
z+Eyc_A2S-6*4`DH$NfqB_rcc9x0#{(pmwHi2Ix{csMOTQ-GRwVN3Zzi=b7*Sw%54k
z9R7aj<DW2p_GfQ5?Lx&=$^}zV|M$jo;k@)QlrElp_ub}a-}J`AYuv{DpZ~o1-9P@5
z!1){f@m;TZwRz4re7(zl`1`jX{g`>*2mj`9f0>PG^ZE~b`0vaIKKzluqQCdG-)^4u
z%&$9k?lbp)*8K51-+g%e7k>7q%$e=&!+kcodS?Bfzx2z}@jzqw#vJm8f8hJfQ}4Zp
zEJxCxzWdM3KmGH+969E8Zol<b^HV?iCbxU|JCA+}((}K1%bU%)vuEt}pZ&rY&9DB^
zTg^`C_<W7LKmU_&HfPUxdAZ;d|MV&I&iB6Wu<nzz(U%>dIJN_%wOy0*)wIevuUyPi
z>v`{e5Ol5o-kCg~W}Sq|J=u<R?#D4FT&Mi&uA4Dfs?|)d)iT|Zo+9?!v60$2>$>%o
zME&csOD)rubv(PSbUNmI#j|lNBm1J)Zel;VwhlB$D8xKR;2rU>xzkk<%kO>L`{w>n
zeAJfpp-+D7u*}>nU}UI|91KQL5*_)_=f89$YuZ12f`Adf0iv#l_)q=l4{Or9bEI8e
zeQf^TpZux$>Vpp*Iqu(C1Q>bJU3Z!vf5Yp&k4ax)pAm@h#{x5^CBMx6JTHF1^USxt
z{AK<_cKgrY|JUa4Kk<*pj^A?AMe~ztZ*c;c`HVoIx9Q_%qz^imDtc(V<%i#BZn|*Z
z9XkBo+ut|8{2RYz+IqO0hSZb+r=Ai6&LS@nzoC1X{^H?P+1?5LaGD0HMyZ$QQA-Ax
z#+(MWC}poY4@SRAUIvqqj4dGm+JOv;lkE*UHjn%~FSZ1n@oe`SA30voWOWkU@{cJ#
z&K`f@lG(cDW^-N!wyq|*NdY6%o@?{8b=kbu&=>I?7`3iN+qEelkWuxpiTTp}ahwz|
zGS574TA%u#Gt=^H+Hv&V>b6Y}W0efv;0c{pUkeio77*#S^tRJnvAR<;9RVke?W$=W
z=ou?80wZ@`;6rP-b=d4ut!%2*qUi}d8R#F|;3fs-b7@z=$kzFWY3_GSeXC+hDyJ!6
z0|PwQbp>#cFyI5QP!4I(7q9^aBLlsi#$l*exo6J{;E?gZFQflJzZNjj!itfM)FcPE
z01i07rZni7L0N6uZJJ_TAkDSw78og%DypZe^7UclK!52aFM~4WZ5l9g?ykG74yw{^
zNBSwMgXxQ=`W2w*d%yEl=C$AQEjEw;&qqJ?Df5Sa^hah;t=nsuHy71Ued&ClJ^>tg
z&NJ>aKl**&=fBKu2LU5p%@G(cuYB=~%qw2_!eieDUw+_0`#l1{5cPZQSjD~r-=ccH
z<;BMrNKFGqwi*rdpTGIXO|4e-l_IcD?FHodl5bJfPrUIB=1F(naU>r*|NcMzy1BNy
zdkBo&a??!~;Fw$_M&5V7@BQXu|M*EZk0lnez47~AckK8-eCpHY9e?qc$BqXbsFcg*
zU;o@s>z?v}kWRO2{@ri<rn!9W+F^G80wUVZqBfVNEj)AQH~Dv<hlrkTN18DCY)@j&
zbfm%=JO249SvT3pG^43UPKV78Cv|e<!X<YWozJ>z!{l~((*C<{z;K>19>;vn-ygi6
zx_Z|8pNanSZksV!lufOvg%#WNZFBwlRa4iygT*}6TEk^a{Vz5Ql3OS~kHG65C-caf
zmb~uS)zrIZaxv*V<36lC81p;Bfd!u1vdwh05N7|%zUJ>8bLU;R9bUT3yAA89|IN9G
zEsT4uobwjhM!k<GWp+AC>Ra}qhaNLG-*NF!&O_xbvB5hcHNeQc?KlaHk;#|*&6@DO
z;dS3*&YwHF627kwz4v|SBj)p8_`JEkw`V^B{Q7!>*e1_->QnrO?DjOq2rKIVF<7lN
znXYO={)@l$TUuJ>eXLHN4;tWQecZC*>h6G#5I$7?;7{JJ4{2|epDPa;UjIGcb&To4
z2hRKd_QQ516d%|Pz06+oEiW@qf9g}sqhEW>+<NoPM?Tib|JFZyr@3D%t>R9T{S$Zj
zLFEPMHz?mc$NVgUCuCf{s0qWoDM%OYB{C0Vga_H3civ)x%9}@~lQ6-Zr@hPO1!U+C
z^s*0FiMVXpi(Vhphm&K9i9=bS4n`QxB?XKu+J<?r<NhXLz}j{Oy?F~+LH?k=Nvhh8
z1?-NpoRX?*o^qE31?Q=2)ABlCWLWVGddsjaco$&6#qiDwcYj$?-xDt1;|EP^NZuB^
zVIJ1VvVF}M=7W9zTGQNi;l_&Yxcb!rBM04_sWnRG;2I#uNI;0kple@mT2Q1dMh7g6
zoOB8B05V~jBFfEifD{-Tmcb!FqYvvuAjR1W0zle*vnAHXK8z0KZHp;V&}z$myK9Qo
zvZ>We*4RGKd!#F4y;nAD5P>S1t5GJu)Ke5Aq^P=Dk_(`tCH9DAiD+I~tf)+`*OyyP
z&-Rf%60N-YK!l+pktJ8p)C9&1HQ#Oba`t&S(n5v2K=ytcrYcMn%_+qc0U*ii1J0Hu
zpoHIc^ZF6hr+wj6{ha2t(pzK4Gx+acIk5Zal>Yg-pLnBrP5?{dp+_GzzxXS^V)lA@
zbMv_^bLF5dy-O$Rqe4OVl*;_pm%PNh=A|z=q7QaH`}r@J-}&RWS;h!#grfTK&f9M@
zFM8f{&Am_ZfEsrZ?<?lm@A&ig#<mEFZvcRWr#<;zSIyzC9~dDYvYiuC;yb_fRpzcc
zZa>WDHiE#27e8N%we;;UnLhDPpEmFP>kpXy-D~Da_uOlK;Jd%m?1))1xk!uaKBa|X
zutxj~wK$GUFlxT(8^3|wBkA$Wm(8#K_V0?Zb@ZF(NT%@)<@^_7ft)*g<C_oG$lKrZ
z7svFBp57qu01m*pdi}bsr`>LwhaS9i#HbQf*GV|$p7X9PQ4KaOf84*5^VP8OoslRk
zPtduCA9>u|ar;d{rw)%zzQ2;5Q~BqIw2S1t7R;lQ!v?Hq!!P%Ws?4fcg9hBU;_kto
z*=cN>eHK#6+W>6YY2T1*5Hry#-6C3NQ<{8sCR+KF!XJ9@QN5>bopN^lhgb3J^1lZF
zFude#iSz9aWWQ_D0Plzl7)eO^^SILTi_oao4OH=aMSo|NpFa)sq5Jp~Pnf;^1Jja$
zJ1@4ymh#`G71gjR{0YDSV1yN3E|mmeJpVb*GPmDyi@8rg3Lis3>5)etGoSn7mkg_y
z9(ed+^VruO56WilyzMq~&)s+F!{#1yhqkAs2OoY!F1@~D?$?Ld6IZStW;?Hsfgk&!
z*B_(1lc5O(QuVV!jE^E1Gq@Flx_)NoD60lJ`Iz8?1Xo~DK1dFQ$Tk+%z%nlp|I1+!
z0kR41B3et=H9>tCHm2NUH>edsWd-HSjz!Zzrq5`4I<2`JFmhTxGS}VD=zG7dIbWq}
z*&3aFZOcX0Q0#&dZna=<2o$Jnd5a!8WA|4UWsvaX#U&FiuP_dj1v>P^MgXMXI1Gh@
zs|U834S`Cq6<~~%<ZjD?6k?!sJ6fnBaD)Vi0tiCgB<I$v+^AT_K}RmJNB|_b`0C1F
zk83OdFkD69o(nfyl#TJfP?T(%vzC+}aL3B6>2IwtG%bLUf!2=;Kmi=di7`7AD3Zr*
zmJ;WN0;cjhj-go@HWL_;9va);Jie{IdFk`M$r{0L(SoU$usvnI{>e{&+I;DeM{S?|
zt>(Y`drk9A&wPftaN(S3XuQ1qh0izDiSGmS)@rrQ*YrJs&ME>px9SZoL^|%O@}Su?
zfA@)hFn|B?f6(HJ;005Z7d1WiS>I?uh?~V4#|5wLvx$pYpmjy;AQEhft*wUH+1j#W
z%byi_KJm#<nZNwthpc-n|G8?lB0%ZKEa>4s&NhsZfq;wi=gwL8TEG9cx0~<zj@MZ5
z#h;fk35y+<#JB{mRjcN?&-w;~8!Z?kL9%l_-#lOV@>f)L*M5UNA(y|?(yIhoz4V1g
z-Ah0A_~Yg?_kT{`OIK{2ci(Y``MRe+&9DfKgeQTwzT=&?j%np)$CKAU*nBnl+SgY@
z*!(cfIti0|ecGI_W>#gBG9&7Ay}55i_e;N5#1u)Z``76$$<C}&18IJCtx^xCxh|d+
z88DL22v`&tK^!LAPkz#q%*$T%0)Za41g>NX5}Q>`Ev+766912%{)~Cxp@)yGbOL1f
z*YdjyRX%EZ?_Yn=d_;bXUGlF?dF4xAY+m-lZ+53<{s#H$g~kpLglny!M4P_)&_m`=
z-u34Kg!avMy!w^q1<!eQP*(db{cry7U$#03%#a<7RRhbsL>8g;`Gw!K32quzYxwn-
zhV7s}3>#B!k{cv@P+39wmObWH-z-zkvY_BHbw|B+Rr-u7PnK_W8rahc{2(@y|Ct>p
zFcSB(G3Xv;-hxrSpnr}YzB-LAUlaQ7^g5<1S6aBD0^ooN0s{mWRztanfYrg=k$54v
zrGkMl8Z$VEJ%Xz(4@?m-A^-yyO&*Ix><$4FfEz`*aw=1%%5KSpmP^$}+3JsdRiV}F
z1Je`O0dPV*49c-d^EG6Ganh3<uuVKhNnd$kLyRm-g@thiEqLj-I-X3D?MPB3lv&I7
z<-W^1ldlK>(kl79*dm(5;y=j)P;?sB2opdUC76shjWKfe&bw?Q&nRxob-6-g-uu7*
z5C2Udpy6pF`0xMtfBak1)mZtj-|`dYu8_q_!B^in%HpDH*LTflzxYM#N{Yo;aU{m)
zfB#?pyeSj+X2Aql!NMtG5&pe^pi7TEHnk)GjPHHzx0#FQkGtXJB0!%%k?UZX?ce^Y
zSD0`9rso8f_3))f&9DE?@0pt~UNk@V<{t~pGwtA?iDkq0*tEpLuUEh9rDC6az2fkM
z>xFRv$nu`Q`fKx*2X8ETn^tyqJXsBd%~zAHdA+5F%@5P8lQ6m0tI_DSU%I+$Zaues
zqJR;<T`9UIrsre~;!3Np_XLY~c6a4UOR>r_U}U<^vNO1+T|91XroWS+0p1a72SyhC
zC_V|Uv}O$>J|Zs2t=65l-)`=?>n<@RwykenRxNR^r;lM`Yh1p1%}$mdeB=?s1RkbE
zP-0csD@Chk1z%7a=z}URHw3sfa0!;jS01=A0r&fOh1&g;d+%{sXa0WbpZ~?mH<^)-
zfv<o1eRc)bMa2=RD%djZx}sHMtkQxF;#_gzZtQu_e%53Wc3<BNcI6m)ammpM-F_0k
zi;m$%Wm$E7)8zEOM6L*sJFjuMs&o=2xU14tTI&hwL;sjUZoi*P&P`jP9#mFZ+r0^L
zW%^uQFQw&q?plmHs+$5<j=S^x#tDr0%Z+2(KSzn9yfl=7z7ND0=_*E4Plj`1gaB-i
zn9mi(+!*vcfCN`k1pyo$J7i#8Sh2<wumqswFj;a1F1G|I2+Xj*0wb7P<^`14Ii?sL
zWdSa@HO<L2Q(3HrzVf#f^MiS!bJ+#!gX=AeM8^HT7$p|85jzE!UVtJnpmD+FVv2az
zj=Z*r@j^SuTWNTK+d*LJB^klxK8$Pfn!ontTB{%ygvTK9Txk)E9rlO|%3df}H!j9V
z{p?L9Cmj~_58VTPegA7-Es)A%#h_nUJ**b)36v^}=>vELz?4_}M`QNPAH4lNW_zn_
zUMP21xSnFBPc}E@XO(eZ;~+1FO-TlM3b6g4$#X|@?q;)P!2%cyd{<zIpOqLVQSwm^
zd2x*kC>EA_Vh5Epj!6I^%;{lC!1TDn_l}qp{&{{_S+FdgC#F2#P<&4TL;#4~|Jl!)
zPu>4{TPD}zrWbI+rJj6#m-H<{0wntI$38CIHq1-ES&PAfKcU>EM}GN+dFYW#M~(&U
z0N%htx$}<O%!RY(6!%B$E3u#Y>fe1mlVR?E{>xuBU(&)i_dOI;&Z5VH0Odu?isQN<
zcTmg7n8bT_a=sczyEmbXu=!z{brL4`Cg>d*cf~?mwM*nh@11!UfnB<MJr-A_j6oNC
zy$=ZAJh0!6gHz;6D{MS3+Mi)%PM(V!?3ye;iTyM`yOL;c9A(aXk4G7M>l{10Bi0U#
zEZQ-r0T`L48~Y&wopSa!%d{c_kB2=JsETg^VinX00GeQ$=R+$&F=8!xGAL23g(qCt
zv=zi*<0WB{4g+T<T~AhiavBKwNSPc{#2Bf{MGA>?88DJufA~ud>Kp(0IJe!qCC)Nu
zm9>n4Wmef~J|E>h6eUlbz(|z(=FzDuZ=I9tY4ZBxeND>~bR2!R3BF!XaB<x*rD9Go
z5<Hhv01_}jx}sJTWCW<zJcBm@%u6z|6U-Y{2Z_7@7W`t&CJqSf4id(59k!_}_6Wd)
zjUA%%*4>uMs%V=SBwc|sRmB4#0DD=1+nKXr%*QpBWv2*8;XFp|p%^U|yir{wm?T4D
zV5mG~=^RLI&Z%w;7uLuwL5c!@iao#}#Tt=OyrBKQSSr??mJG%c+#2<V9qD15EI1-r
zx}BywWPej1V?@9R`rbNMxBYu?y~VCgMKNUnLf9<|oB|NxA3Db^S-ICXL;R<^&Qm_v
zQQQZ(-{LrBaT?v00bazuuFG{6v-){r#94vn_~Y)yIWAdQ)>-ekS?5<F=qeU=E!&Qy
z`Zx#}iDFb)wmtJ)KRI8WcI`<6>%;Dis5+UZt0a#jcYPbcc>Hipva%)04x76ydcW~%
zpSRUvaxBrj#4lwGmc<z9ju|8Ld*(_jakX!J1!;bEEu-93*lnGBi$0&+_5R;D%kq~S
zw#~C514hC$w8Xh(TxmtAHi-cf_E2#3UDc|xa{>N}qsSar9w$p&<t<Z=KJksbbO#pU
zU+o^4VZ#rGjhCdG7B(@Dqi)7n7oY74>YAlwG!WE>lVi$%uX8Q~Mw00ZO7%alg8C+@
zn*Dv8R2`G$t~+S!q&&;+M>V|o`*>QNQFPOn>!PsLcTwe`F2K(?7h6QFds_RGJRU{2
zLG=TSGz9>3G~4I_fXG-50|O?9RA`e8wvItu#&Q@WxJ<=xJ`h+@kYSy8Bmg2FGeO32
zzz-S9tBM~2fO4?cF<T1$jnN$gdS8GB`Ekv)ea4Gj0^5MV-hBZW16^Mc_(41q7%b$2
z@!+g4Ho&@MfW5+T%7VoLNWw3vEI9#2F#gl7yciNgf>r~3Ae&riwTA+Fv<-4IR9p<K
z{CS_tp`idHp)s~Rz=EONrghp!T|cEA7-`&eo9(x%V!r@3(WmH$7%i5qqqf&|uBhLR
zm_c+o;1mJvYr~G&9~AX&ESa`+iGC_#?ZD2d$pu=w*)vt`x0T-obH_zlU9NR85GDzX
zEIQt8XD0k-);n(2`Gm#WnBtRqf9q*w1brUS!=#~>``~DfBiR$*nVhdCUSUcq2)j4Z
ztP|v}3K*FjS4pci&N<1V_uHb^C9Rtz%4ZD17-_c-%t)-i!NI<%X(7OA3ckLqUW0(?
zdCJ_x@s{Ri*F3tPH}A52?XNG+IbWP*`DF~-<{goGU}PFi=Oi*l*1b9?%0poqdmwC}
zo#b<mUfVObp4~pNO`LZB(cmcKB`L+YiFw${Z?`%(ZL7?Z(#f=P)7pO5hkm)k#?<b?
zp4n<B0BgXfa8OzPa^u=gt1jbepDSlo`gHYWzni6jxELdQGGZ5H$gPc8VR2|7&G+b{
zA0X)aK)?vzvj~`u8!hWTt1A~%0v#;VVhlE*LtJAG1yqm#Q3m4}9*whgb8xL?872T3
zMSuvgI;t|N1B3vi6jfeF!M{gZ7%<W?K#AI!`VqE5NAX7-^MpD8GjPdd0S?6t!CVQ8
z1>=6-8u~}3vQx9{5id~u$kYY4G{uad{JP|Vr2>Pcy{ls+l?O1=85PICi<mC}BTCd|
zlUO5z=Dzk}oXAXS0hOT)B75HX@!zxqBlUA~rKL7maLX_GxIKVb=g$IU$c>%)rK^5v
z3CQ9=y)ZD{Ld650WK8ZJ;A*X8nLsTCvS+)hG0;{&lmt>W#WZTjMcFEV5t`y>efpcW
z?(JFcxLN1pls{}98)rSME-P;1SlaXyHeX$}u2)%K*u9ZvorKA~O8sqGU5n1C{Bh}4
z+2(_mfDi>+@A()bWw9!WF;dm>EG?#idHOTTBCN%vNbagMKf97^byl)p1H2<r4~+N)
z_HS<*Fmm&g?=d^YvcCzb+m7;3c;t!e=JtzcQX~813Zm2qb_xWEGG3BWLpH%pO0%mg
zH>eN8#uVq~WsEF3p<kUA(pghb-=tI0OsaKgATD48*-z6wy6DHE14c?Uxvv7OQ0gd-
zM_#U#h&f_?B?%aS!BLU%+s2<z9uW+zJEa2wDJ6jp7}LF&9~Juxt7Kbjf{tQ%5SV_T
zGQbx|y|Mc(fQ&k<61mYD;2KNkTVf`_LLli77=}%ND_yy}!eU?Dt_ft3i><!mS1A9P
zb2ZC$z~z>WVWM1B4E1UUY!ZUVBO7s0%DCymT~|&3je@3j_Qzm}>K-cgL!n;x+;6Fj
z-d@wjA3+EG{XKy~m5qxrLhyC;Ru=$8YF_tTffb3X(&?Hlfg&VaX^|KM^b@S1Lc3{(
zRqr~fz$WWvOFFe~$#jH$#sa@&8I66eC(Rh~q)F%BS?{=6=hNA$lOt6S8$ZsK_i<<{
zY`z*t`Ei#K*Sj|C-bk}f!sL!y*UOZbHUMy$wyfh?#-NKa!Z#!R4lr^W86$Ds^Yiwj
zW+;J+80Rf#mHE^B>{_KBPIFy6D>eognY_6tfswR#<WujvFYO{mVgO_>?9_MT$^=~G
zCGyZikC{8~yk%))Sd<BF+^f}Arr1+bQBWU-jVaE}3yho+RmR@TpuUN{lB~egX&~zR
zJp)Eo+jZ!>)9aYlupq#oZCyzbcLOW8bw4H7R{n&mDK9|z80FB(oPdo=9al|VYal1~
z1cr6oSrzm<l%*YmJaI~P1cY=1P{8P*{JLB+RmECp3qS$<0O-JF7hnYCR>cl!iRpox
zEE2A@x^l7Q&f6Fza{FZ6f9X2T5sSo2a+xI{gmU|mjdEc~kSe3DX^Sxe8j~BxAsGrN
z$SHs}by2r<5vG8&uti{90Gv!RM!b5``?qo77OMmF(-o5k0D}aphxVPluw_9hj`ajI
zK_nrLNK-oBE|26&%hOklSTkY~5$lCiRzGz1T9%RHu}1(}`leP;>@L71nJ$uarDcUk
z>7QBexS;bv#`=^tZ|!0&3u0`7VJ&RFy3P)WbDRc&`9b%FYj4szdH9jX&AD68PIbRI
zX)AJa@w+dh*Vgt~|DjZFN1cbIMfkKWP-P6}wTJ}Qn|w1~zkby;<ZS@P$h2?BRqJxu
z=VMfP<9rrGt&@CV_u)L{tTJ!X&o0-NEA6&9U#+^`jrrZQ=h7mI<=wV1z{q3+Ghk#=
zBg^kcc_^HQ8)JD*dL^86T}2r$ap}^P<6iP%6LXh0`}H)C)^QlrhhbxibJHfZL1ktA
zzqkf6eTM69Qzy3Jt%`nj5*SI-M-cA0whN``joE;l=)0v@8qGfLmK3W)ANT+m&P5jN
zjJ6Emhe>!yu~yNpSR@LL-o)(^djd8{<%aiBW&sL;5HL7wEE1K8Yp!xRFV|KB>k5nG
zlm$D5gwY;jdqs8Ph6~0D00n?XY0M%SDGrHs?**WtbHvKP{TE;caWQQC5tZZIX>p7^
zfHgfaCvg9T8?H%UWHe^6c(F(ZUcCZ^07G)ck_Ajg-L4od*Q~6Y@@M<jJEVVNfDj8t
zAlDdRQump%z${!~aR}YR1_H$DNbd!KR@FhrjJtQ-4+Hf>MSAr5939}Q42BT_?@8sp
zbX={Oid>Loz=&1Ju|L^6@R&+gdLYQ{V4B`DC?jmXx>9AWxU#T&Bh5P56kue<-Ff1w
zV+;;pTgsKzPGj5b?OwGpM#AFaoj4_A4Vqg6X?}Li-I(lxg*3psEdxdtQr-L)ZGtN;
zeE8QZ)xZxWHzAD+<;GWGOMa<W>0WG|s~X~3{X{>IR;jmXqkYvc5f~sYAiA5FyKxyL
zW7xT%{FBGR;;T&N&GzTlfa^o9q;E-GO=JsN?HpvRVZ22FM!d2-mm$lJ0V~ug`@?h6
z$kj?C=9P8X{S-Wq+!%Gg!MTEu&a2b7Sag%XNU(JAAB_U?P6I|3(Qn)>2r~9Xslx#y
zL%E3}!SLva8PHY0a@<d0P`AfrK*wbkF+vE&4GW_nV4^3M07d}}>3|?GP5J^p8e1?y
zil(i(nC$JgTw=a71wiywKH!9Q&GvT1^vR$^@bxxq5Tz}}3E5N3l9E^_B=UQ}jmKaq
zR-L=8kyh#|2&&3}<xnC1h_-Ejcml{P5qzBl;L;NqLyVMKQS(FvUniaktP}FkUf3oz
z#Vlz#FbFvX=nVAS$_W%}?O(NJZOWfBcibtEOZrwnl<HOWmA2_G#giEj8$>`>O#qj3
zkp@K7RGb(3aa%D)uB$t?1b(nzm2P16kkGNzr7@6`E3}*!hDCY$`}>MfQ#0ia1V*Af
z18n1G{BiS-Y2~@|{$k>q$NS67wu3bg#8jBpSHUGDe>Cbh&^*5R<B)pVdu9>+1zi(#
zZ#cQ5>cr)b`#0=8kK6+>2-5KV?3dHOjXoy#m(>w{2(VtyFSlT1nq1T3NKMN>*`BER
zCa;^e&lnsD7|Cg|0pE=DJHKfHM&h!iP1AupeqC(ZqB_6fTP!q9-$D1fD`(!{ILn&1
z?nU!2`s_-3TW7_Fo;Ds9)uaEqu(EeGHf#DO(Q+*RC9!QUV)Nx$kpUyIbwPMRn*fX~
z`cMdOxpx-&>m9e>6eVk7d7K-k#40#t1t&RPf>eo7?+L9Yy{S>N6CDh<VK1a<leM6-
zPx7&_7t=c19rSS(Hl{ci1sF-HO<`kgo;nwKS<h4cth|pre%0K5)49-WL46Z?!6`o3
z9QO6W=^3U^iyGDij6~7zrOVgNZ5NV@gXUO^QDliL5BkQHR^Jz=&H@yQuhAICp8y0Y
zm-4195TqecgwzvYfXk(x&XF*b!xDj!0odUT?M1nu>MCzVph*WXMIX}5rUI497*AXa
zVp@1^p2S7~q>zz?a{1Sn<=ta3r~+Xp0|6a?Bh|WCAJ#BGGX0){eQUoZBR>JwIp0=b
za2O+pc`PuX^5y=?u}j2zuz|&g0!peobp>Wu><!5Z^Tc!IWt(MVtKcGv<FHYh*B<vy
zY{=hpcinASIF@-MeO9*>>|XI%_OBe+7SQ&A`iFj{Pw8vohtvd|0AdfuF-8DJFoF=b
zr7VyP-O`uzZx4h7Y)g4^&8Ak+5|>2h#I!k$Y1TV#*7?(@vw7RR3=lufxIeE9N#+lm
zuO?aVy2}o`H`1(=FuB*Q(@m~_n&EX__b_9yT&vlIb`BUhtuaQX^-a>_ZC!o8OuN$j
z>{_N>*|pI%z&j#Uz(~*vk!j8G-*oe?nw(Z<lE*=9&y=DEU!OOB9I{7wC@g9md~El?
z+_K{Zkc>mtX=Q}1%z5<Mp1Jkx_Owg0;}JFBRysr(FA<B3l#P$9haR|O?z;Q-LtrG*
zg;^%6fpr_?C%Yl&BPDE1ac-T!2=XThGRd4MpYO6-C7hy+gP^`iSgfr0DH@n(UNohM
z*bm1EjKqCUEjmY8R!LFrt3~TT-(AfiN9D39ixJSm-I5*^9ataMSne?dg31THZ~zDa
z)B!nQ5MVSfi)FzRrz?hmi)-PYKss%W>@L4MPI)b{IZ9)<Ow{S|^E~EXw>y-vJ#T>&
zClAJY{v8PXs0zI4iFKk~Yp@59;oNnB0FAQ%B?|7YSSNY8^{OfOdcM&xdsmw_m^<#m
zh$E7d{IFk$rGeZqO8U}CQO5`tPs|d>YN5b^0+b`=XvkC6W4`1CX5voDQesp_b*pmd
zj?B3l8zQIbf^kzoS7RmgS{}QlreNt^v2zB}M_DWw7%=VY0yBu8LPN%B`+UPPa@ak%
zdSHXgR|RTVNd$1vKkb9AjR^wy(>Pa`e&m8odg*gsI-)p=r4RR&7gk9_v0w<kt{^a$
z(c}8a)k|Iy_4L%o1<BV(-yN}hWH=JUl^8I>eN9Tk>TD~>#GhI3xS;cZqUdZ=&l5#w
zyJHP-uko#3&COmou!sR~9U!&8bHnDV{(L8GJLukUZA`OH!sK@HpTuA0kfd_Xi$%4H
zHm++PM%(NdRE&{<i!mZsS`ED;)5I8AMIWTOj-(cpU6WhYvt^H2J#YKltpm*=%Zbg2
z=9OpJ{7HV3EGye0*RO)H%R6Faz{npy_Sw<@@@+q4b@7bnzF-lZ9J$6BXOA45v9r!W
zGB41vrw`%n6|dxqGNGAe#PFZ@Z>Pby(;4+mCBg)K#xPpz>2;1_f5W-|@}j9ub9>~^
z8`m};S&uw=#oTf0g+pMZIqXfP2m8U(xMbMWau#`BCfQcTage*FuxAkG<^@LLo}Bxi
zY4hl3naSEb^~9C;w0=NN>Ti{cF5(`mlWmD}KTcp|o_4RYKa%=zTcjVHzB_Vz1#<&X
z!!bE1uOg#8uC9h7#rcr+93waY2C*ne;~#5O-<2z?qCkg|016Vuat!b=3JT+EaF11!
z(Ve<#DjU~K7=ycFesCS-mB&VR3#yDw#k`{dZkhrT07qaekV^V(3y{H`6Mzdq0H^qM
zfDA#yw*>M4lmH%pLNqFhk+QvQt~{}S7>}eVw_P^Kx?~!vOcL=)2vm+>uC2HX<2+yn
z43)lI#gdSbRH`;iPI|DhQN&V#>0;x1=p4WZ01NKB(6wc!*sJtMdV<-3eh6reETp1<
z4%$p=N`72>x!5b@?Wuib<!LGQM?=8PjlNM_6`jLP7_mVBd+IyVJrUL=JFd-O<<Ot>
zQyV$eZ2AW60*v&>j1l(zi8|;5`q%*@oF5(RD|Sg1d_BxvWY9ZMnl}dm;CAL%<mT^>
zP2A>gnH*P1v||z&5!@&j8{vlBRTxUb+8Q=r4ZCiW&IQeRgYFI2rZnp$Ol~KC8h*pR
zYZsA^?>|y4A!hS5dHvs=%i1omp)qIJ_OWKC>4Gun557*nS84zw=?T6*P4>8tuVWs|
zoSwO4(0E&R9L&<s!w-MW+;PXvhuP)<&(r+uI&9Dh+gNlwEP9<&-?VQ@o)t+0BY%1M
ztL9rTJjvv)k2<55KidOD;s8d1Ue>O`N&ZfL#7A`zb+ywm=PHY#m`S$BUv}6-!RdRl
z{r>Cc**<AT>v@2mWpC(t^qf>)#zO@k*i+&`Bvr4QfVl~7*mzv`M^jQdI4R}2KJ>RI
z>>0$ld4Ums*_*h%u4hG1-z=-Yq+J{>D__{NWZAOAu3uNZewo6Umi~Nh+ZpPo-H$Hc
zI)2?dVjXqQs&k%sg1;>EP4IQ%LbSvbz}O9tfDzpVHYb3&bz#*gTL5Grqj*n127%IH
zUyv~T6Pu$i<2pgpD;VenMBp;Zf)!XE#tAo0824Sk_OlnZJn%(8L7Qs@CbR^y5C@|o
zu%e}y7v!yrMbQ-FgZLsASWx+}8_p^Q2H;Ce*AmbjbE|cmB>;ltxyKR!BO|`&CQHHD
zOA78@s#@@*D?lMHX2_tex?yU_B^a?rMrsRbr1%^b1raO6rQv`$9W{GyC>`cy(B`+G
z6@|U_fxra;O&GZ?lLQ$&LoDqW$(WQAIFZx&vRqrC>$2L6K1VXJkn)NLLO^(2lcD=z
z*E`PjIe}OX7^zC<UA2?ADOJld(V`OR*2ZvAdts$`?GZ>tS+YhGD4p{~xy0(=)=aL*
zNPr?RegJ$9cH3rKI%>*<Tv9&}Tc+mA*An}rRI6C`WdJ;YDS$Z*T~ikm2WTa8rDfH6
z>`%};5MTtxNZOP8UW-f4dEQa;<PDQA2pCzU?6`H|mw9r&>aWYc?Z0N7+w;EXr^y<0
zZ@9dZI&sHW^&6&lg0I7_!wn%ovp?>_yz(VEpIu~}&ny3?<qzxE$+j>CVMOk00RX|*
z8LuZ57zqMC{JNUuQheGsroW6y{YKee_WCGw#GPl+XIGlPIilz&uJ=Ney3(G9XGPM$
z2s-${)d!;iBhwI^v@0@-e3PqTqFl4;JYf%oc~I9?%bnG?9u0(zm!u^Mo0unE_okH}
zG;vBp8-n^UY)o-(njto*tTePSiVTk*v~1j*D0!32<Np}jWPK*7qpWOC191W)aeqK<
z`bSF87(?G(#TXIS!?F@ua;1d99IyaD1XoU^s+bA5cEX@;6IOJkq73>rc)5)GBe5s|
zM=-9pTYa<D5KBYH07h_+#Y#UCEQN}IhpX3`W?Qj9z)WDJaCZN?zyZ~fBVLHE1>nGt
zkE<&T@g-yvYotF0J>Y3<pYz;Fb#x4th-G-FEC2<-9gI~2fCZ1xVL_Q%zHgd+8KQLz
zprzmTTv%B^MBoF*^KzdB3&gTW1X4K4hyx`?-LCW@hN~FZTnC~6;6niNA>ar92@uH(
zEGZ~pE8xa(418E{14f9h0kG%`l(EU_Rm>7tDd^jQK)CrLRip>kMqGH1=qqBN&>w(B
zw4FwvkGy26;1*4Q7r;ng?z$W>wR3*UUXL3r++)!X1Yf5Qw#q}(&Pi@<qr)wY0hl7V
zG=l|GQk@nM5m*HirXm1k)D#0p20}F~14gW7kN?TufyZ@{9XYCju=(oIoF{89?A}PT
zPW*Du>&4?__27K1iF=MVU+wefW(;m}jFCkb(9Ex?><YgI7JYURMEo?dNc}oHY1_Ob
zRsoElx7ZmYC#^Hm(4Ivf3TcpiT{3Ligdt8nt#iEiC%Dtvv<xc=bA@N#m{$7c9Dn4A
z>*n^0$ps@`R1-mc7&fN%ntPh;*5wZGxO;`5vKE!=BwV-5*r)wb^?8yPGi=g0n6~OG
zsWzk;7>S~bpw-Ap=_-zL(RW+ElXnj=@QZ06LpyGl03%3U1>44;&oSa(zz|{Ixp@M(
zu#5{`TOcU6jOfk{71!?w_}G`B9VQ1r!_FBoA@maL$aqhXbkcb-My_1hmkTP-2v1%N
z_W%O?BN12~7KsavZe`T5HVhI01(j;S99&m<V_?BvBiS5~1EWG#e0v^dh1uUT?U9%g
z%1>E<36#yRPzAlD^O8qk0&b`j;{z59K*LZB?OXxIhT5v_k$kM@<$@!pwz)Su43mLk
zZ2+p^wu<vM1_=!ZtbjeD?b6P+1#BFXg)#>Ucn**PV3HSLG7z{;*(3FfR|nO2U<cp@
zeTKdWqE1`8`z?FkwpCeiwFL`@RIS6rX?n~Xfl?$3qzEv<#nyp9sfPLscVp;(t2i`!
zZOJP=6XV1K$<!vXc^y~-n*@+%D4+ych}{DFsH`?-j1jBiV}GK)1CL#}#sg_x1hYn!
zBwDqbuMYJ53!0ZETKO6a54$&<PSW^XfibeReRhqS+svvrZJy5<9BGk5x25;H;<Dwm
zU(-SW=1tQU*KKCcWKzeefiyq6;%sMD)^rW<Zc7>%fqDF{N58ThFcM^Z&L)62IyNqP
zca(?1NnDM2LzD4r`Nyk&Lg8PLO-)e@IPpl7Oes;uOENWkhFd7nVN~cvjbHR1ZYLHE
z;FVOHr<Lm;V?lYAVXGyzstO$!G+CYW@jY!!(U(IR{QQ$z|9uwrp%$c*MUPG1!%@n4
zLX-20qIJ8pNEz<hD8F$~{NvQ8zn`4^LC4&=?03Byh%!F=x_?;(Kh(!3Dae=bWGiMp
zrzZ0+x}Uc9Y8+!7V5He;n|^_hePVlw@gPIDbvcEiWWWau0Vn|&scb2TxeV|(;%0a-
ze@RAnSRAk)aOFgx?=u2CNH9cj&vixs#IEXqWkAdf+;LG>O>73@U38VNJ$CK1ub}9#
zATYQC98^?(TP&7}%0SMt_HEfJ7ZwRY#949B9fK`-U0W8=(%utrqOru7vO(3=Ho%->
zm%!p%9!PW07C3_Z82CM4sHeI}mRW&IE#GPe<S9reJ%JY1^_2xNkVk>mx8;)T>fRwR
z;#eb;Wtk<qcBoZ=4qSmQQ<Xa{^aT(yggFBHM7n^5G1M=*b_w@h04Rrm2*8JRJEm);
zgA!%v8j@wzsQ&<tEK6hzGTD=o1F(g5!I-J5{RaZDh@S%61dymU)V-(}Bg7v8+^W$B
z>SNrSH36w42TYU><3@c>{dIs5Ei&lKh1hmQ08~m+9QVq!rc5ptnDvhHpPwdI+P*{M
zmS@>#Y25B?C;K7p@~4fRX~&tbcEE%M!tO+feK4(@O*`(tHzsADHiw_gx5)lv|A&<k
z<U+_Vca(WfSovu>XBm_JmtUHMOF))c<A5<}suj7Q)cd-sMILggRh7Mf7$a518p%?y
z2G(&;%<9WYN$1T!d41Uavd^w{=x<tkm<uqcFKG@L<@;<}IcYi`#-!3$f%x>^0<C3a
z-rK!uF6Hi8rif$L-nYGP?*GI`?ePzN@?%)D%m*$%v>Y&Utd@`i>z*iuh?e?OQ636W
zZ~P=5UTIWhQU>HSw$;fLWxO2bjXChWe}e1k$bN%0u%|$1+Xc@*d2o*DN1u>Usaz?<
z{?W%vth2Wp85luA#+x4)xzcW%^P2RM(mWnkWe~QCi0Mz9`rMRv!6x=oOx-<qF%B5<
z-;Z+_l3fs41CBA$8VacJ1?h&BU}J!Y<Nz?TBSyh>F$#!TfdL#tIP3~o6@7sd04gK^
z4;Ts<-M7v+tWh2JOc>Yu1G!<+H5lFr7EZz~7TayP>8dM+1$khGz^ni;uz-j508=`O
z{b7*|E>tQjjtPOxYg-i?homl$09Rnh2-wn*T*wMQP*$FPXJFlg^~A8K$mlOFh(G~Y
zAs!nfuh<>1UJ5e!lgGNL0wfTi08mr{sL(I+d$B@@{UHX4jTa&oi9QrbeHlygGH8zU
z@p!}4mOvimCGN;b?zeKd-xBiz0K}F*1~UK~C>yW>mP}iK1iGb8FV>6Ko|>2)E%gT=
z5AL!6Ltt@q<#G%#2gG9}=8KaFxrk%ZRUgxqnpiW$4MA79@xrB6Pp-TA{DWPiv{M`y
zNH|;PJb<I4zQSD=F41t2))HH%p>x#FIqEF;1u_Y&&ln@qSC7Z}4zwM0%Kk+CR-8vI
zNt8KH{gZiNF$0QhD~?f;DQ$D8uzO>k@4_T>hOQv}9bDcsx8Hif_yLlW!WfBbJj7Mb
zvi%sh9E?F2Be_o3bYXIe;fd?Weg=#z+f^s}`emP8%Z%e?7a=UuG4sluw)gNdZJ)fB
zcU#ho5m#rRE3Gi}$6c8Hu1*6{9t!?R#xjy^V&8{>Zu9yW#90H01&*7#+9S$%v5k!J
zkLv>-UUgBTeG5cmtYx!{3S)IK?xO0fBQUEY<4jU3;etM{!p0QmrWs;`%8IKO<DLY}
z(+9KKFi%<YJad!O=d3oR>3oz`C@rnbiDO01H5p|6EH&4+^jlFzU6MjUC$63-<2_DS
zT2bnAjgHbE>-@lQa>o7Jld*@7`ksvR1A!8<aob;53RS=bfeC;T<&xr0=)7fn0CLDs
zuOAyf1Y<g>E9M34jIsi*S6~pxibQ!_R|HVez-!BZ4wz9DP=UuPumlY2j6U)cz<WzZ
ze-h)A*cQB5h>yXdEkKN_01?V1H5DZMjEwUHgm*!}`6Y-vm%?x;%7D?avU#qtsH|2T
znSNQp;spv&9&G>^qE=#=^mH7+18}1(S6YAt1cfKiIpx}5=n@{;ag&9cFxVQjZO{aq
zkkLf!6V3sSkN`{E%q%ZfMNuGwjU%FbutU%vsV28rurP)Kkti3|43!d$9+y@GYX>NS
zHG+IDNPP8d!?H(k4aRZH;1R$=UpkNrAcaJGV1?jbi#DLIJ=|+)8&(ebEp-KiDUdob
zVtRemuRbL8Bv%#Lb=>P6v^J8=eajf(T}qFx+vUwNj8@yw1^>)?$IUvwV6`XyvU!ax
z(-h+7`u_8AGh+N@FR~prU-g$a&pXb4U0S!@y%E%|s5)6k+mQP-emBmVH?8}=@_)0s
zcU92%h`M-*F$nPA6I;Kp1tSA3f&ds{5yiA`$e{X`Iku?Bie>8Dq-&%0<uYvwdOn4f
z*KQqX9;(GIS`d?^qZ(M(Jk2k~Dr}{^Bi0Ix_#2#UuU7*MIQ%gd^x>U76z1=Xph<jK
zDdZ<%=-RXwsCzigGXZwjgy|&8Iadx3uUuox)brko)+pD`lV{%koTuzn=k<e`X=<1I
zxbo{GY)o-(+N3t9EPuIa+K#gi;w;OrqfOGMU+zubo@92(Olk^Phf|p11V;SNr+J#{
z-UXA_B(1k;^-jtWHQ(fQLHp>tqu}cUeYEw*#_z6-=CDEVo834IYePozV$L+h#wd$n
zK?3L?0eV>QLYDM_zy%EQHV%l4<`n@XWicSS0>LXVNhnib2y6`JmI<RgEQG3z{7uyd
zi(yNk2{A~DVnAS&$4Czt0ehq;zyKhFWDWB=Q4(l_jDRe#RS39V7W0GlbOms@SS5fK
z$VXw-{J?Y+SRD2QtP#Kl*cAW>03+<%pzOp`Ax;Ruj`9LlAdro*A$f7x<uOPAhrGF$
zb$2CTWT5wo7f%FmgwX?-q?s83*`;sWFx3gb;kgXEaZl1-$HJjrWTqZlxAJ?g&Aiwl
z)a%85kvsxY(3f?m1#qE%fFS@q#Bk|LX4pQ~&6Xt2i;Y4(4L~rtHp5++1xj?RHWD}^
zkOP+q`_d~6&%6qUiBr;fwz~pdEC@w^Xe^X_VuIG|dcNdO?R<aMJ8sr_sWM9e8c5RM
zmDGGEpc#=4x%p}z@G_xqkw}89I8)fY;dGKlI|6c7x6h<@UX^4w&H_i0tb3la<CwoO
z2J@O9VAHDCw#>fV;}By6V1(mxBsz~sS>98kfiyq6*4+ini@l{$>y{+h`guoWz(}$>
z-l)>U9tsY5b-x?cuuRxQG!Qmk5~<KvXxPNumu;!dO_@k8)yRY{j<RYnJiS#|)ZP0x
zIze}LNq2WEQqtWGBGTP5fb`HvgEUH)bPS!6(kb0Yx9sux{r7t?$8)gmYkk-K$^TLA
zsvz#9gpoAPxd~>}!>nR9;e5omq(@AZ2mdz2BgPMLZaNOx?g*o%?wL?`936_QRjsj@
z@<#yJa7XIBgg$+bo`MNG*2HzvAa@-UuNE`YZSON1GxVs23xLmD%Jm+Fv&!6WW9}o3
zbVxcaJ;t*D>5&mgL?L@H79X|bGy0WEPV-WMv9U7H7%-<#6!xhSot1Rg6&;a8+5|u*
zm>OC2{aE2WxjTY1Q2Nan$}p;uOj5>No_PivB@FPR8|gzDMv9rEf~3h&Fq&IOapMg=
zIw0J}{#!P%d%r({v~edz$*BvVq;SI4cH>Pypw~%%6HY>xy7=lQ5;<;?=vgbPIrpsu
zWj|>x3Ov_Ugw!xnL=zwXbKU_xBs;{z!G=)Za`pZ9b57wln|KAaW(05_5aSgP+T2Xf
zneMOi0CeafT$=OhA0Wv$e!vkSZ|St`fjU%>SVEg`aA%68%>Z-z)J~HwkFsBd&<?#x
zqkn~b0n^#>I8FS!j-18VH@_wiS!T_83^3x*$8>Efh|iObD)OWgC-T6FL<Qe0H<kZn
z8QBDShJRgqBnEPb<^y-R{I!}$Lv8+%36Y^f1Xe`_3$%SC88?_M-yMH$$Ep3J6#M5#
zDvG$_+rcO#ksGt$hjUnm-!+PKC*K7xR+_0+#Wgmn#;pLFR&$glLM+H*R$~0}W3ic=
zad<)Gb)c}L^7+%)@CW`Ro^P3A<9KMSqw8sQOIW-sC9fsBjI)X+`1hSwdEW-=_S>XD
z`34lU7w0RtJm;{v-~72AN9pyNhz8ZDr<`Z~qC)1gCC@AP-i#-R6%GJpS*LizP#(Ee
zxYw*SYdOE0`!Lh^8r5=1bCFpI8Bo>=J&4W=&({oi>LYV&T7PZWK=uTB;QU!tiYBL(
zkvx~j8l6E!_ZlijJSHGtMzHEQYyyNxQ^D8DS8o_AR&-kW3sx@jIBqS<dRr&n80)ZG
z1)0=p10dpn&qo8Zm!XupHhbFCwK|4s#v<_Of~Cem^Wv-9TYBO%JQL-XcYs$i2bugS
z=wkZJh{23Id>3j}_JDx5Ewy8KM{Y`<xS`hc1JRfRl3s)T*TpBU>W?z3-^kKM`<N1g
z2TYm5(ehBQY1yRbr1K(@BitgVdR#-k7Y2n?8{ME+kO_FTvCqF3skPgeImneP5;unA
z>y5}v5arc55cShd2>|HqT#9AM=M&#|Y`xDhix`80J$3--sc_}#4Mce<x6wN+tGe-;
zV!@C*ZoNX%WV<~mag=s`qLuM@%|%p^&T6{|M3=ThbP>>z8BDhoJWA5{g{Z5Unv#-v
zAaX_J;x7`(Z%1luW3eOE#q7)E38?JXAE#n&7Y{1>A2L&I^;TlgdgOjvONH&M7P-fa
zhxs(+2%FPaVZYELL;PLn4F+j#+9E=Fbc;XgjMF+QE(>J+(Ui7L<{IRdXg#lsLv>l`
z6Ib5#9k1My>{yx_*&TV%6Pze4mAyo|wpd8tG+u!Xi83NWR1id*iuPre)+6?dXiS6%
zdYP2Ur=%0&Z6AUOJj0HQa9I?GXU0hKSJR9f-Sc-_+}>$om$Q-wHnlm<;Bimm(`@jt
z5-fvmReAk-@koY*omZ8Hahx&;0nF<BDUn@MI_g&iB-jvI0Jh^vn=nlu{i5D+bSuV@
zpp;X#HYaD@5VHCk>6>U{P24t+w=p+fB&FpdS#+BP=2`2&R~Jg1q~Ol_^`tf~(oF`{
za+bY(Pzw`221+_^-xCCrH))hf(b2?Cga{3Qj%p5}*=tx7%j;rY8QGL3u;JJ+Gxh@{
zK&D<Ncq|AW-+CcJajZ=ywz4;)w@0Y#t$q>yp@Ug6Gol7hNi-R98@LQI(1f<95*8j^
z8Lc>dV@Gr=WjBeSc-qyW5`BG32}1+m{)k_o<y4Kf$sP(M<;Cmk>?nnI=~{uvOEF0)
z@16etEPzMrm7ppUfHvSwUxZy*7eJ>m8m&eWNp7-#JdC5BDLlcGu;k%;J?&(SBuyj6
zY1>E341zKpEghrXoUY&4owOT>LdLG8N@>+u%beL!Z40oAkX`D((ARrHHdJTM@)U6K
zabb-B)v^TlvSl8_&Zb~oW1m5IR`tzJ57&oo)ojn=?-hCey~BGRyl3wHCSSMD;Oznd
zObO!~<F4HesOWCQ4u<<S(AVTfMXQ2zJYnw9lp0@-hH%|UfOH0#mBBilF}6zhxuMPs
z>Eom0|3pb>^V03M^A^NWO^7jR18paLK)P>Mlt?OpatC}C^~2Ee0A}_>P0RD&ouqo>
zel(R<4r(*Tnp-xhSP4uHmZ#gXy?Wo-?306XBv;SG9SycKzwIFZH-#$=ijL|A8aqQ5
zl=E;u(w69O+)BC;2MDSRcG~<j&}U$(G;nSR?c5^Hxy?JPvV?jdmf#tMyv4bRlZH43
z69Gt`LhZ~$YepAT-GUcH<SVct5mm;T-B&cG3i6LCwPGmHHNpjrIi5|8Kg$F%1?=-m
zH3TS`Sl<guwj6VqK83OJ8G#HRcpKcuTNXk5uXJqMn|STA@(u(RXhDwfbk$7|gRz7w
zB`>(Sc^i~Hs+CGQoG^u#2P8pw?B<~cgaY1;fJ|LV5km)JxHb_rkRj&Oldnqcd7=}{
z&Go&FD6toqQJ7!FBeD+!&vL->+hq;{6s~-WlLEeATlg_^4y)6AK~(S%H>)<#GYl&`
z#8(HCysdL;(U#);f%HoNNlth#m=07nz}X5YcPW?cVaNb`P;grUNq)E+p&clgU6GzJ
z%({az%h2$+OyKpf8yB8mW@|UFoF)L%2$Rs)9MZ)j>Q2KAfWngV6ghB-qkd1&fA^W1
zetX3o0JG@e5&67lw;Z@xeLN|SzQP!0`J)bJagy-b;#SVk^Nm>VcJ8lU(c6sPN$@fY
zHb?^9CrT7+HQRAYXDvCfpXB)PS)YpU?Y=h&Jt;Tu_wVr9lAeN)28p(nz}i;RcO@nC
zYs-oU;kNy9V7h<?JpG!aQ4mW%Pxj1>dSs!B0V!`DN`sQ6=ep?NC@W*LqJ)p01Y@!f
zOaD~(fA*eiBO&g(QtZ@|%}uuENu!m4IpGJ(I%r;J##r(RQ+aUFSE<)K$du?GcWB`v
zDv@+E0!FbLZ!Kf$`Q(c#cj~~~A0Y%yI2+VN6_ZFLiaJQQozUCBllb#+8I3`bk^ICM
zdmszSXVpzZO(QuE@5<vN!6_-ioU@(~^-V)OBp<!5<vnA0NE9YY)ny7|-bP3(ctl&<
z7Mmy%fC=hI2H1Yfdi%#knrH-|{>n+wmqyy!UddGzVH$-*x-qa;8)apUSV43BM*lTJ
z!C*SFzRaW>EF)Jsj{F!y)xXw&olW*yj0kD|PgiVIfnWl_80BWI#ZARL3AtfmtBlS$
zR_8&-Y(sJsTqY;D8&RnR#r}j_N`K4ffNvIdN0%#Zg08HQn`uu<K{T1@1khFBmQes`
zw7DhfeAcn_WFAOZlIz+`s$k43nj}(~Uf`7=94H&UXm)N{Yh6n^#VO%5k1i@9yO-gU
zseD^i&Oyo@@+PkI2yMat{ocwsK2W|uQV7hsIwVt%QBb83<L6>NZaC8+aKFlH_ratg
zfN8z1>*rkOY17E}wuqkH2n+W+o9m9_YYR6nBQLK&KYkd=@du!c?6FoS>A|fENUlKA
zOt131$U(8(X|Uk|E`-QrbYB_V0@OUMo6j6m&Jd0CujcxWZW+{0!BEQ2^nHo;qC0)R
z_#|D}VuxK_?Y1&KS!B9kH+aUSIfRhU$z-{5{_nj!hY2krw2p2Y41Hic72V}<&iQrf
z8SqV+Y<c=h^$(a%IiT8fQK-o*d*Yf5$xyd+e6-&4?5pqK1noU_=hf$yt;QHWuInI_
zv9We(q(KtWp>#}?*FD0guYpwhzmerATwYUU=>-W7{2*BXk_`0gyCArP$jW8YaU!A;
zA=?>$?nt$|AS(<K85PV=h=$&z97$uw0}GA_zk+cP8&NMk8WR&QzM<26-iF`~%H^|L
zuXebZlSPLCx!<_CX%SVG0DKY#LCl0No6KsPNeE$xGk7xLPrk#8`00>Q&dr8}W+n);
zq0Yz9#hAdY{zn*LD)I_W4Q0538j?C*xlhzOP&nkMvM!_pJeja=qSZr~dPs$~pXZRy
zStCGS=8+(z$90zF2&6`6jkpFYg!TFB?Z_#y6EE<l0DDU+o{(#-+ZhdTkOIN%K*9Xd
z?6sKQT7Mi5<u`h#P2xy83%6KDqQNZfN_J{UP<<WBN4^O$mkWcyikf4XF<GSO#O8Ci
z!q;0$-Zbu48F2Z+t2D=_@Bb<7P?N<RbMPX9-KGSK?W$T7$Voy0uOtx7_9T#pOq3GY
zDrx3h{(4ny`S+T=G@LZV$H<Z{R&~JN$s%xcQ^lXU?S_f)CDpqR-WoRH&6)N8f^HfQ
zNGGm8nSkrpJwDL*GZz(7&$0B2Q~MlIaFwTT$UO_AA6DBj`>L<a__6=L0U5H33JHFQ
zbz!Q%g&P+tIaHRIb3EEgCETC!yO^opjfz+AIu=rs{9dwiOmA8dz9>~p6<(4d-ETv-
zdW&>lvce$H*deS<f(T>*P*9MBxKe<{r4e5gnk(D}+-_>W_U^(sh;UGq!TPAydb(LI
zjyg#iUCiO~<*Q9d0r8HT5-(*VT}*;+bd<5uzJW!|eYE?VT{!FCv4Uw}EjP4M01RJ|
zIn5d$eL|9Rf8K{C!i^5v$ye)6ne8DZR3E4r6@ZPX=fWRz+gLBm+Sma*6(po1GUCZq
z04(X~c35QQrau7c!eNf$RK%|J)~SbeYk^wV>*^IZNXUwwocJvU_GUZ^s)n@J#Jbq~
zg#z73e0%z8GXd#R!a*5Fu=IBYPFP&-rth{xHo*l;2`ZGrgD%4b8Hv0fNp_wGRN_rA
zh<Ixd*s?;~qR*w$kH!M*jFbTT)SD2;$eLqUq;)<Zds_<BeynPe+l5-j+?m*!r1|sg
z(T?@M&zCxwe`m(oWG&1JGBrSgR^DhoXI#f+d9GDg=t%lMU+4Zh8LT?XG8*@ReCvv}
zn{i(VnB}@vT>k$>RWoz)r3d7^r;nLD>B;t^i9~{bekydOT_1&1msVEu`KHc8Ui|H*
zrZb~uwdOVpXGEC{wVjy`L~{wt0dkY@e}sqIyW(fg5Ggo;TN%I5wx08zFyP%<4_1Fx
z$=rC@8E<s{K=c2`#Gw`UZ}?4b6MROVqK_R`$Rc|kjaGPDE*OJxy#o^0ZKOUhlG4M;
z<L)oI#3donyXL%B_Ck>-zzPehRSgDc@M3?FgvM=0n@H;ou?J#+cBEMT*7k2|DB9vY
zqNxvOIzSFUW^Mq`Z?J(5CY)L(suKf{H>B$sZZ#=3(l5-vH4S(VPZ2MG_`m?rtsXZy
zz~vjZ0l?L`9sh{-oiu2;pD{mBP`CvU;pb{@2<}Io5=fcgf<z!n(xbxwP)u(u*=+{w
z!4AM);|Fwj7T-La2LUx4@z+Xyh#}T{OMzx$Z({iSjS>u~Fr@$^*pN5%eT)_Zkt5Ba
zBrWG1MN4Q(wmJHt<(}?SB7#;6ov#UzNR!aj<Cm*KJYtWz-$k-%>Oed|tKf#o+*h!O
z@pGjfOk%ias?gaF<fpI<_nokr7byfns78cvaI|!TsEwVCN}!%xC6cV~Qmc&J9b=-B
z-<w<OSq0ig<@M_F@csP<0&cytm(y815D}<37PA;XBpUDZudzM+Qf7B_Ut8=W!NdcX
zru3fX)QCK_Ue_yj&u>lY7`x{iU_=O8$JS#E(7;f-ipwl(D2H}Tkf43vH}MU0mvT)M
zmnK!Mkq0K3*BnW?@j_|U-r?H$GiMWYa8Adm_|R0jw|`a{+ne$>#vk-O#29(p{~_UC
zj=^T9luRw_8aJ9JX{{_^vm&Z@8&Dw@SYs*4w58FhOK4@)UbL9=%URu^b7pZjn{4Px
zq_juX;6^w&k=Y?#t7OgUr|f!;lRQ#$!NGBqtP9DYB~%rFg2G*c<Ow7%N`wHQJ5)Y^
zHfiZdjB@}{U&FU<r-a0ux9w;oWR4-;F+Xbx#rrsb^sjupO*WL;wn&$A$}8t`(@8B=
z%_q<H{X&A{J$@=k%WFVT<bo`NVHY&QexBu);OesjQ3xpsMe~n0yQWIYv9;{Y^|Y>z
zwe!4%52?_5Q^0vJfC?0a%|Pt<XGtHoX7b$FItn#}D%`5(^F<qk@EyYITWKU|qAzgu
zK9Ev|^!FuiZnJf8qOQ>ndIK}MHO-_S)hZf|#xWRx2ct)igOR<Oa{8Wh$w^2rggPLv
zX8swhV4i@H$)V-I{GrX}SV3@Dd{U6J%b~{S0Qw2E4|vOpwhh>K<<T;4QfG`D*R}g_
zBWfz;2V5gOk`<+h9RxmHFR|QOK&0PYgRBbQv$~-RI$Fg1Z0?aN&yD%{A5{Ei)><c4
z0B7%s539K9lAIaqo|uE~|4bI9+Qf^V2QvUG7xk$}Q%<^A$Y?$!v9*)M)ltOJ)#Sl2
zQL*78X5aO(=s~T%?kZP#MZEXh!{9_6wN$-DpvF<pytUTrBCk}8hofpPg=$v`f~RYT
z9<Kz>1plRTn`M4dY3^_BjZckw*F3qq4ddrYDlRkPyamHv*CRIWr%z#Y8UNfeB`P25
z&t2ZeDxGC(2daVsH3<6sAq0DmLhm~4#%I!0LnL{LeQiap=J79(12qPg*eT~BV7Z4a
zohuU_TytF*1SWk-+I2~kT?6S|=pfv)pc59-XtuJXh84XEg-hN5@-D^wo?^D!X-1Gx
z8|Y9V5$Yu94;a$he<F;al!6Qa&B^F(D5!K#x_XAWt`A(`sB0vbGwaijbLg4fXU&Vh
z53>pS_6dE7nQNHr`U^5me&$g|BXBa|vEroF4t<;1`8yU+I5jZe?f!=+Dh3Naw@C#$
zS2vZ019pzLQ@J=&cu3=@<9zqlyhI*AmKEC(0TR5*a~Nl0i+UH#V+pPTvwonG#53n5
zP)8KjWY<BA^pTFSiq>y304ZF2$wbr|=z&<Ea3^IU1yh(DVB(;Tpwi$e$K`IH8P&)b
z05c-h3jTsfx@GkgG4M6DID03#91F)Q->e_sM)}hsML(ok2jlO!ZYbzHPveh%7KgkO
zIQT-?kiiDWMiW&J#U`;5FrVv(9vOy9!r?yi3V8Qu#45SnoMQUQ(cFRcbIh0y)2MLe
zBeYv)AaXZlwL_wx#BRa+5zZFOl$BSk3h^hv&2pG5^_BE+eeULTadp38KiqpWvGNiX
zzr-tbH5^Zwn$t2Zsx9j$R2T8a?~yS!2`>6z=j-+iZ|2T%dIu=05iEK({^!B@{=D0{
zNLJmG<6xw2knCIcybr%DNKN!gd2H2S3M+pzw|M<>-iCtU9kUIhZtK7BNz{pOf`cjE
zBb;-?<%q0kZ&WF8q@khq$a6Juk4&Hdv0KA!wqarHyKWmwp4Y-^4j2Hm5dW>JDDBxF
z2=Yu$iq6LuXf8#oUyj&GAC{wsgDt2hW!|D`eL1RXl+F^wzy)H`uQ+sjElBD$(I~4D
zQ!$q)eAneM;=MaVSAPYp0azvg>=Xoe$VGkQ4_+H%L8$Zz_!mONN03RtTVZZM=}CFw
zG;LP&ImBhzSvb}53W%DZmV(U%cj_iVP)7+!+bZh-D0xHrwIY?G43nb4QxPB)x8{ZE
zmQQL`WdrE?74{s#l-%kH{o#Wo=jn{oY1_4&FoX~Kw+2bq3BL>L*va4Sfcx}q$}p;4
z(TF)Bspp$zeZ*EKh2UP*GUypTg4LY&A$7wOA~zPz`f|~60$yb$bm}D)&GerAPQxM9
zwy;Rbip8&CaERT^<QojLh1ua#`hNzxS7j4@&^+J(O%?5JKLI=z#+FBpcAtH;prvL&
z<u!Dzq2@!qq(KLg4;-@}I=q&wAJDm*$Hb!@k$DxZA@c3GfvU~pzP06~C!--c?PCiZ
zq%9&#P;p6~K0X}5JnXIE>2I+s4-6SGG*Nm#-(WYe)Nth9I8aZa_>l(AA}WsfYL481
z=>$m6fo#$Ot-N0IE=B@KyZ~+_xc_++%oyo)!bO7;CB2s#OhYFo?;yDoA)<+kGX1$Y
zuYJ3FmCzjZI}P_+0|&{F4lOS8NTn*O8`M)1b%Tc&ZVl`x-e^{IViWqqi|uxe_eu^4
z`RSwpCI-?!Vj?1H1T=th&?NFCf^Ik>g|t-)0M-1BIwk~FTG<4+h`Nad*&`u&A{vAR
z$2<b<i;L_<w_ryArZ|z!kYX)v2XqLgG%9)g+gMWOH560>cASEM=6zZn0`b=BR#z3l
z6e^&jLVp>k4?R5?(=z@$1!%&%KmD~7J^FhZCeH(EWqcAt1RHr|N<?^1k7ZzriYGHA
zVfl1CfFp;E=s>E5aQ6H2Ydy|}1hs3>H_C6PexGXa%;`C6W@g^C2ig7uUE}pKD;k)|
z6pW+~Ywka$+&Fy9h3^(DImWGZ_D*75^(iAg+ook3$D2PsZX_~QWh`^80a{2y0d&ZQ
z%1tI+Kq!WJ?p$1}ooU2|Pw>G3gNag$UV^M=^6u>?{j4NEudrEP&2h(%U)2u3diN~_
zCysSpSvOTUOZhWX%s!V5?3NvhEVb_3&WV`9Q*<t#n@;imPtCUqQy>uXnYm)cPdgt=
zoN+7K^HE=~kZCnYu58%%SjDI}@QZFc+q8zV+aJ?52(hi37&eOQsb+ox%K2H+VvO)6
z#CHQ$%y5$n*c*`{C>xP6q?^{>Rn;h&s-igc1@trk0I&um2h2l3l@UG7biqV4|K0Kn
zFZu1DV#<y_T^_IT1gU?{2$PvwxDSAEgbBAf=U0;JR|N8~0EAE`>Zl{2sd)p;Yon%8
zGmKc&t%ZS@gd>;;Awwzwz3R=PRUCjydq+2>nfZq!C@vnsDUzjN1SoHwx8a61Z^Su(
z{D`u*8K{k@aOKLwaF6;P>(6Ue6bCYtZxB0BOt=pVJvtN+hm!t^#$T03G~7zhUj)%?
zygcX^4Z1_K50&q53_G_yKp~v6-%6uBIe5i%C%Rc__Kq;x8r-Rc3dM~(nCE?`EmYN0
zhSsk7#>}KP0N(Si+v)7JOgAGO(`{>=8U0(6xLM{6lg#kUm%rw@1OmD)TRNn#WOSf3
zD?XJ{9~xfn6*nBGGX&sPi2&pgUqkZ%J~oHwmpr3W{ue%8c&?e`DASJea=nB@rzp!K
z>+SQc7T&e$8u?$*=`wt4jO69@a8Kt!8~jMN4-aK~X`_c<c^2mVo4&~co1qRP3&TxT
zQVg(2p+ZqnXOqUCm&RvnHlqO(<67$_+m21}^v0yN8G^B>)Tu?@i5G*j4aJt-O13K7
z<syp9Q?=x*|7OXTJb6b=6>OI~Nk4yBu`Mf5Gk#vjg1;i`=}%UL{=Z8@2S}%kCD@yi
zFs1;f`m8^qae2<93;wPY5wYE}hI+_>wTAEa5t-ywYJVUEQhO|wtP=Nvh?RM_t@x1@
z_J2peOP#2L-kF4?FIS%?dfFgb0Sdcf+PXcDFRP*FX5p8eBvgHEBxn^I0ai)_fmv>7
zcWX@*VZgjFeMTJ~0&@mH?v*EoONez4kQarv%6^6GG(Q}Kiw06~IkC10&d5J9|A=GW
z!2}3tbfNM>lndC-8_P_`4D3Iwc|BzDUA+r}p~poduXx%Fka6I4doI9-pd7td&WptS
z+j)d&&NWM^Kud|@L&Jz%RULiZI{WHkk`smw!3NdrAyRE4(&|4c_>m1SSZ+lF-a}*W
zWcok%{(&y@8sQ9l05ahHB4{GG(#caYHCG8bGsg&LPCKOThH4xWY%}0etRa~YCs-S>
z`*Q$Euu$;I<sTf5(mQlP`Av>+>)8MbJ2`rZf_D|gf&fHb@KaRm?*i577gjlGB_gY$
z+CZx>wI_h}uMO?sK-!##pSv|ojWZ9y;`6u`mpW;-GeN#NJ~%hAsh|Hn9)>%irbe%B
zpJIvL8)a3%zaczwt<gq8toB&6X@;0Bjh}gVPsrVpm$Qbc;*q?tZ}^2m0<&_xmBl$m
z-$j)H*Bs$7H3b0#$r1dM@^bRj&U>txvuyWMs!b>IZHyWk`}PlgweuGdm&a19(Ys;p
zn2!>}$fDQ3IboGWY!Lws@`+~o%qOfZ%^`MTfs^6Q7-&PnzvMl4&9E!07!hMBUwB?_
zKGI)&{B3H<J2s*sd9y0P$1)+x1cN(9OW9uPVU$~XbS7-K&f>^VdvzzXcK^5S{*w(}
zihrw+HR^d65=h!V!fV`O+vwk=yatY$>-L<{WoOsg<(+@nF?jp6Uf2C1j^$un$^Ig~
zv|%KjsHFta%<OHa1Y+6RKxB6DcQOFU!H}btnjh^Sk?v+OUXR|t`}V3B_y9=gbEBm?
zEV3p%Kw=aC;NG^n8tg%g1km-lx6va3^CHqIC|kbL>bOkM;~;pU+~4f@eJ^RIBqUu~
z=3Q{_u!;EgX@o&)czA|zK{`j~0*x6$3Ib>(k~62XsX%a5xB=Yk_%b&7d9Qcy<PppZ
z=qCqw)fz&a^$Bb8>ME95to4?gb+ZK3lr9V=We4F>u6{1NSRetm%WFXhNE&SVXj97t
zAYg~lcx%(aHCb*tVePukAnQmGUg3l`RuikO_&k{$<b7Z|>5!~^)qbsv+XEHpJJpn<
zAE7N_SDf7aAQFINv|5knew?H<6{y<?DSBMY97Q_nQ$3X6xrg_+=qF&rq2&M!r52K8
zS*R{2v!PMC->*}fEJ$c>62l_(@YdT4;~N!>V+9%&kjx5i)owWCKcW;mdMexbc_{Yq
zGhXseb?T&{u`j!!)E0GsUxMY$PNe%OiJ=CR$spbB`LG*TNS1)w%lY%q-&O>U`33Nx
zq*p&2S7iUmutL3z+<{zoqC7ub<P8Yj>6N<4A}>h^Z4J7o2ZH#2PE_3)k*h_`96Ow+
zbUtWSnRI>3^*KJ*XEQbUbIlcS>RyYM>tWF(eos~Z6W&jzb)&A}x>uzB4}q<T@<r46
z5=D>~{6Cof5={4(cIK(OKB^)juk-v!%;6E9h$-N`=4-E?db|>DPop|86kcx0uqig6
zU>iAQA`KqHf!@^h8@%t2+qA<;@gEbf8N(s{>56O<yKU*XVOmBj<EtgzbCZ#2hiVQ~
zhoxgDgIzv<KBrReL2s7iS81s(V@~ihkn_B-dN~@>kFV`0HIvSqOHWK}s>-*%OXH!U
zlTt>p&UdKdLT7F%x!``%YwSbNkd(&fLcN+4NHyvHVVV1DG)Yae8|dmFIUJ6G@P-LU
zhQn5rAEts}9^pEj%Ko$b{L_NZ1;cIAr&C6BT8&WRuD2g`#Z$^~psAo)u_oia+3``e
z<HgPfNCY6eHPxqO<A^TUATHJ_m<0(1wSz~nLFqWvvWX#~*n&7Rdu}$de$BOh+&@}k
zF=#a;p`n4-DNS;8U<@OxnzOLpW|?%4Hr{?UPNZDt_Ah~#Gkr0B&~uDa#Bj%ljQ=gS
z+-AvHKTC*oQ~15E(Ybq!EtZ6=rz;ow&R|8JNsSk9qV8Q*6;(&0t&>Dsj$mzSY6SdK
zL`qk0$m1ANV+R|`QS-0+zWawMJC2}eUYHtfR8@E;uh^h7F4RL<#YqY=-~utK%ADzh
z9*`Zpdm$kHS3H+-t#(U;an3a^%?dOojAIib=Gfn(NRq|(3;)PYXxtW#59)}jtm{=9
z3+|1Z3A-7<spWs9Udp*kQGtke$q$QhOWoAKsZKW#W2Ktj1l&*KV=BMORIG||t<vJP
zWJi%hayIf~GLB17n!sc*&{?I^Y^Lba?ySgwSJb?SnFtg-%9Dz3oGna8FAc_6htPcC
z`2bMYv>Otp$pdU`NYCrnyqWKn-bW?*9&Hia#xYTuN_^W}N2bjvcpbN?0pq3-_A%H%
zQpVM%=2-DX6%RJREJw%?wd96v(<u1a>xFv<bEUp2@Eakl6*o8ekn2t}0CxHs*#Ue6
zM<d96l8OO&Ya(cIUGGRb&=d8O{1L&>cQUG!Y9(4!K;S|q>2_S`-WfcuRk?THjcY@=
zU@k#DZM;%yQ%K#h406q%s|PNNU~Sk0^Bhp=8MblMsun$#k#Nj8rh4WxY$yT<o(K7H
zToBXjYAipBge;>UYeG{R0Caj{mOae5O1}`TM<2?t{04n5ZJY-vJg8Q`ZWGMc^I=Id
z>>H5o1lEoGMa`<N(u^B!4e=1lU^?u4S)Mh)_i`@MVH!F$lm60JyX`URoN048Dq=78
zQEb!iRM?j!RHjzi->n$2(4F4As6^l92th8}q5N;`mH|b?n1;hU*I)I@r2nFYW~w~R
z+7wL<eVU4?NHXT~vyh-G>;npT;b(Ra2HQR<5pS}w7llU+i#iqsV(cNvCS9|pSw{jw
zJlK2ywMXGdHk#GVrO~Rl;419rB@}-oS0@u#q}jks4dpMP()?8e@<KcFBp$be3s3i>
zupC61yARj>n7MDu1teM~z0LCQsUN9W%$dQ>As<Kg9aq&*Nm$SkSVo0aK?(b#3<P<I
zs2c?=#99LBv>*i(1yf{-d~Svdk@;FLjxdSCXCO^5Gbj{JRQn@b&Ru7RBEH$U?LG2b
z3|^@Qtl+GNZmcid+k+fkB9zu@SJ#oOoJV)?hl9nrm9VNP)A{yx@QpjWFE21V{dZu;
z+0N+_Kl0fpKYJ|$4nf)!=^H0jcN1A1gH*LPP9tAWjGUU3Nnu!ca4qJAvE8s|=0l@?
zpH3Rj2{_n%`n%_2@pE`c^_gcoT+0VkkgcKw>#GE75(Q{xKr|V0=$w>oXlj(gX^&X1
zQyymEr=8<v-iOth@|3|}WcMwr+%4AKbi>6^^B*+6#n={u;dhI)Dpu8@;kkKHeIcZL
zsx!h<&Itjnl^0`r-&2&n_Mq~|wSu#r=0&B5BqBgM1PiHHxYzhu%>F57?h<~oIdhU3
zbEnPS+IrakB2HbG6z+(rt?bm`H2&CSl2Y56YI#681y{1@S_<lE(aPdqOiYCTiALZP
zFlm-#^g1xvI?!Mdp6z4$HE&jE1yB4A*1J32<}8H-6~EZGaeu6oH?YW>o%T>Tf+SW!
z7fgVt4E+Rl!_udq?`<2aBwak8ttgmk{Pa$X-IkE3x))3dTtK6tyhC7ggvrprxnZk-
zTQFV42C{ar>>E;gj1-a$i%gt9?XHa_lSH|%zg{}llA5#N#Dhyv6ukn1=|Q^m_~vLk
zxNQ9+^rTBm1D-_lKEZSSe-lPa>oD$eD9%Mv^iF>O3>^|Y6)m>Bp$5D}%s786R~>X>
zc1OV23w}a76LTs}i;t+VGPVKqF_I-)laeWvtLkUX4*FF;SmY|Pg7!g7FkK<GIbPrg
z@Gqq_C?=@6m!$6W7Xod*%+*Rl46p9=oIvh$4~c=N8Dhs4K#pX6g=gNoI+IUq?@j;G
zj5{}LFGEA#@wPJNBv-u2a@u2DWrnZbmsfCM87t|WhT3Gu*7kpqF<$UShpcMgfVsSr
zgnRU;+#gS4yByttq21fvL&Hhkisk~6vr<ph22qvLCdM^fCL|KlwVM@fJJFy6eEIER
zc=yv`f}VBYN~ti;nTB0`ha%?208Ob*S=h9nWW_3i5VR~SwszQ5X}<GFrc2v`S|>Y;
zm;{xKx`O{3F~6N8|Mfk?LWSd)us<5rvA0=+1NtO3HD2sV{V35F_*P8@v)Ly7`dH3P
ztHvu=V9dv%mQWsN=UYgD-MUw3oVF{z-&XqR@F)<yt*{cXV`8k$(uvGj#2vZiG}1aS
z%~cUorkgE!ephAV1bEwkXt@!BI<M7tOarWOsbb|u3|G_`xaGva&SwaeMxO2W#apKE
z9>XD_T1E)gle7*|0ddPLu7);>cv~!|9#A$lKKBMR!lI=5ct!kOM~~6}yu!t1unnzd
z`-El=IRpZNNo5>x8`(g3@CN`P7?D^L0L<EdSMEAfiFSH=t!NN@+UC!pML#J9NM2gd
z>*|IAuR?;s`>(~OhE+Liy_VBs710nlbk)_i3adKv)v&94yZ3<;E`wn^W;*GHg`5Cs
z^F<b0=-pma4FM8a@0V<Sr3dNUYpj5$JUgc(s(OMhq^IZ2cR>!`>nnBc1SIEq>~fu<
zP9~>5L5QnVM%~ncX0c5-KJQ}euYD58jhX%>1McK{o#^j5r!#$)-9|Gqk-*Mk*{71u
zRrHz{=Vf$#;YEsx=9&KG4UOTMR);2bc77>~Hb@QUS4ndn^}^fhbv+vTGCv(q53l}K
zr7XQj^L~p2rT=vePx4E!)Y5iXu(q1Dyto)0h1BXW_v>D>r)eE!10qyhY*qf+{$LYC
ztg6jx2Adbo<OvH?IHOq8s62A#2m&$w59fjnRq1-4j_I!+w1&pVl?9ZmB{w-vFyUjf
zo+^IKhD9dB6H`_cL@2xd>P$FcApBTa9qhwR<d~~+IU7}ZO}Cwczhjzt-Hw&SaICLS
z-cyAxy0=8(!Fgcy_g1@pKK1(bxg<eE1)DmA{;@J7Yt$eN5bU_3JRJBXmM-6lnCq1J
z^sS!USvzK~>Juu)8346eYttakzC2Z5s_!q2>sZUE<$CSAp?RWZn6p<OhFPL(Cre3`
zXD@>~!Uh7dVGR`-NZ<xFoD&j`poK7RqB4xdwzKhZhH7v%74x(c-Vd;u^x^9bFnWQ$
zSI#}p_Qwck^og`_3h4EUE0}Rhibn;2g4R5m0C5VMv4sAbSao<TQKIn!Fg}OA2Ty7f
zxQ8#WaNFG>)4Y}Thb87;wd(9=IVl1XVFZ8zESjW;80`vHUGyCkf>*v{nMvRG{8M=}
zXFB|+A&K8E-FB&qD;>p7fwzE&mUZu0Wa8~AJGW1GKa9)wkJV_&(xDcz3}eAm9qqPG
z$!vo)4Gzkgd371dSRT|j*BCpG_IeXP9C`MANF0yW+&0GP)QtI>%&v$s$1Xn^dW^OB
zG1=sno(|hTKU!R#)QGfqK|LMR)@F_pUVR&H-jlA1%ln%vPUNB1a<uoH!!9-=A-`<!
z*_G)1V2E9v>3}&g()UeWog>eg#pU|0P=sLImjK7^_eF<wiaDHg#@RL;bn**7hV-ZJ
z^btrFKi`D?2uvQgL#kl7{#%{sQ@lqB4>mbJysBwXG~X`jBZ;~w5Yz4nA%|NK$_wrz
zik+IuvAuQQRWVr@_a{I!@-oFfzP*;Bzt@)vy2)<2RrJn?HhqdW>vZU37s8exr5A*o
zEkW6Hs7{6t{NJOzvQPdGQL%$5s2jMB_$`f$T>1L>1tfa=WkYda9(h?Rxuyq-tPP5c
zuU5YpMRr1)`r>JTG#CicFnWM6te+s<#aP$3R)Sw@%BzhfHPV>-WcULzHz)zpRvc}t
z+D-K>GHeofH))LMh1}IV;Kn90EW>IUIiEckm~=HT;8k@te0RD%U82u!B2hlt$bR;b
z>om$Hejk7ogjdxb>E0t=#!{kSz^j2Z5EwE<EXu)GG3UsAi$oI*!-qbN?-20_yGY2g
zE<GURp|H0+%~mg2WnGlM(lh*c_bM+jSp%77$SS<SL24>B$z~LVLX+Gg5fbO9x0712
zZJZDl#UXJez&ROY_8fi0q4|1emfWV^+t(~F9grsXvjU&Qpy?IMKIPtkcno96xRGqh
zp9$Uxp)Mn<B`L@5p|UJWg0D0m7&=bLjvgFlFEq>+0z4@%D$2(HPsrh~)2dWQ^>WEJ
z>{pQ5QMHF^H3_5}|3btgkCrK>{1^^hDq&wA{slOjwCTuGj<NUW74YIU>KB8DP~sgf
zE1J{07ZaLHt^a|RW}UrM&4-)*5emB1xYP@}Y|8hYcDcw<>zZgkT{fd5ctA?t3XY%_
z{4XH8QJ2p_Gn3yOe$o^+A#E5PLZ_b=5%OqH1t(a!zG0WfvPgbHEz+CY7ToWZ_*P+X
z&E@v#f=a&#s?z*k=V1lGWQ=Z$7pctm(7=1GYLg7w;3pAFoxyZ&fX?kUNyMDjPuN^m
zM-vIa{lxP*nIa1oiwQtTN2DK>Cix1Jj<d+I0<dzLx5Y4lLao2i9E^t%an_ev+5v0;
z4lq_A&8r)f(+o7XMZ^OHZ@oWAF(b=5E?%3~+lXT++?(YXo3W6HE9(m6v2UZr#HcmB
zj3Y!fGUB*PZx?@LFRdoGam~Iai^7ztGfx;ng2BO)G@G|5Q?(~L4CI3(BJLWs<^<$&
zqDM!aZ!~)2Ht#%^qtP`5!anx%HcYZSW6->X5=&Q9_P~xki_s56s^d1qE67%26XUHQ
z7P7Uf4^U#uE@rQosm($;BCEc*(mZ%57PkUhxfl_DklON-{ZfFZr=g*m3S8rt=~?$L
z*bLPR_i+Op&pfy6ctg(Ocss%>&uUduiP-4N!MIrFoV9wS_k%ev-+Sx9*sd};u>IXC
zlg|8(?y_-$e*^st+*l~wDWQ59m}R{#$C2f2gamy%JC}X`9!O{F?KP%CoQ(eNp}1O-
z<`!{H-dcMqEs*(brLAt~zaoOme?>vxhfxeEf^gJK>zTPh`0zkcr?2dsjnCDRS2e*u
zH>W3_?&8ypoI4j!sQz}In4AfI{hKyHb?Uqs9@Jj(U_m9wv;Nj3S1u#;VN$Au(X(!T
zg>N}?8>u+2F66>yu)vOH#+F~g(Xgsncb*djy<5kP7k4$Ok{#E-4y&v)+?KIQw4Yv!
z2hjbDOUKRZI*z&5z{~340iyK80x6^mJXd3yrf|8t&-0~GK+aolQ2)SMrY#vtZM<Nj
zxHuZAwJWv44wFtiA}-FsY*kgpx<F0gTb;VreV^`zTf{iKxTyS&yZ}c4Rxo2x`*^&0
z$dJC7Mnz2O_IKKPKvJ+Z8Q#$wL{~>O7!v9La#g-Ku?EN_rIfX>tH5F5vZ1tsm;31w
zTRq|?5nv1ezpG*Qc#5iafa92)=JaM7kBH$3k!%+YVrSFud|>b!?NElRGty5LkficP
z_TAy_be_GRZ?N^@=wsItp%~VZzsm9p6u+E{M`3bZLD>e7Zm6*lwIW-Y%(E$A85s0@
z)+|qzR$iDDH}LD?t65r#4ac|E$8j;N*CX5_nTEm_@CfkUItr0N%gVO1=H4|fc~ykt
z<CC*2`K}uuSjIsFNwK&siZ!*~MVHJjhr8QU&%8(cN_&TKKswxc-+u8knY!k3t?exC
zmv;&}dbLynn;dEQZ5A)t=+Hv4z1;2C*HWj8@Ky4oX~h&l<&hIxCE1owUwjz{?NeH}
z;^v98_jqs3)qNEymL(SIt7v7*f5dpZ6brD%FqB73w-Q|jIG({)iuP0{dphYo626FZ
zA*BX-nfyl=QJ+q`QhncSpAT6`y&S4jHr)jl2@n3Yn!o)ymr8W|x!KM1?fW%cQtrCQ
zU0P#4bQ0gG%UpZY%jh(DX1iMUms7n@v&SEK#ncRM^10PaG%nTx&I|6l&+a+he?PTY
zEfOQj-3|F%lCQf|v^Qft9Xh{MwOlP&H3~CUzZ`N%!LM?8sb=AOcwbuA{k~KHZ`H?n
z&2hNybsJ&68X|1Q_}8ue#76)gwL3GAQrM!J^$Y0>VRb`kdBaS_ggjLjE8XI`cU%y{
zXVS~(d(}fPwZA?QymhK=)O&W`D-%b}qgkaKiaVar#Htmqo8??`Rq2o_5&&T^dJUtw
zfj&f984ZG)M4WA`EEXt@f#BjowE?67PlgIpf=SoPL0skm8<{c6MtcN#+Baiy&=ZL-
zvy+nP$b!Brra$Px5F{AV_fneQH4+Z&IKi(tg6X+Q!a>{$6cOJ;{J-7gT88XX>|nE(
zcY`G#tkh|^LOCGlt}@Xkq<4hK1rlCqJI2A?6|fDnFZRrE8A^@SpQgOvZD-Tx%t~*q
zzW}Itkq3KDBBUnhvFoo|3F{xr+^hBHiI*(95jd<PnDToD{8Rk>l!BbbXghA6CL1fn
zzay1Lj!`%M5;(ymKJdif4)$(r`pNY)rbZPPQ&S7zzKH3%c{$AOsO4s3f0<vs<-ha`
zG?NYHkPLjW{kgn$e_FE{mveb^u`rB?U39W|x?4`kf(!mxw%xmW74&#1<=uzhEzWxK
z;hai(o*1#{?~K$-%%YTV`rgH)fA0Z(Y;6o9DczCRo>1C!{aB)a3mn+cbxW^XD-8U}
z6i&9zEw`CjFGtju#siiM7*RN2oT!S^Ej-pH0*`GW$!!{M|M<{e#%v-`>2NC<@FzGz
zWotLaGL5|7ZDy0RpD>z_%8h`uaMxlj4M?Z)L%_Dv05+u5b&t5K!{tA0@jZaR^T(Sw
z9ZF7}&n|qxOl8Uz^TYT>ngE5wh%d5#ug+e!s<Z6d+WyM$HO&a;`_A!ojR&0BcRi;c
z$AIW=R|3=S&xP9m)`(w?vyxpM=uYG*bse##W&AxjkX5Sl)0?!r_&p9QHhWoJ7s8fu
z8$QXL?}Rrq+)w!4{FWqdo_xUE@qfDbLN_SnXq&TTb5ZyO)hT|qocF=l_s+H6b2O|t
zq{Y$D%%Pw?IZ%PTB$nmFkPK<SAAIr#5YM*?m0ya6*5g+hsksWJ6oEjx&~!+mM<ch1
zCJ@@Nt@}OeG93s&24F)diAWQgiAZol3(BN-YlP51tP0@%M@a1J5Klm1h=`IJ?2UnR
zSk^6nwLfR2c{~n3B@F_+92+>HNC?3U=HygW(1PM(VE&-&2jFUjeP(0(`$b2{9CKe@
zbzYszhS#}ZP12368*RAU7(EYwv2Q4i4I^i7l!ZrZ6P1HP3Dk`NSRr7*EW-Bj2^xUr
zYZCaa@mS*Sk9*yw<u`M8wn{v5>}<$2=r{@=CbBwT+6Y67KRC5L>t52=W>bFPW!bed
zd(6Ba>5V_c6yfc<jR^YK+B(-l58i%L8o{HO65-xoH6UM@3#7w`Lt;%a?O=<d^-s%{
zjU*v!LD$_8SZ7>8iyBS#paIkSY2syP@$}QzF%?fX&%c+~Elyvah@T=#oVxzP%Xl^p
zy$ui1^&bn;^^h$|;NXkZF16HihMhq?rR~l`)#-#^p`{lAF$=qeIEx*w0SCM&!uvwh
z_Ctm^m5mgt6~^T(=Y-a3X*vH@&Hm|fLrv1z#Ikcs|4P72zcgHj3V4rligmA-l%4#2
z)+y`1-o8Ppd&N=q3~}lPIPajG)r<Uv??m6^^2x~gT%3a&i#p&gap)GeistNB#~Xh|
zY_nDm6r$4;QvoBfJ*;jRyy(aJwE0MPO;i4D#`}G-B$j0ZeemeYpa!cAy!C;NX+S~4
z>U&q0?cp`YPXqs2_T4DjUhCYVWpM+uMyW2OO&_nHW=9{Z6gHom$cBH85K&B0{ge15
zsZWf9<V??#h%?f<_QM*z7#Yu=V!HUvrsy*r(@Io~*B|1%GlKpe?>PuEDi}S$e7l7t
z2HWrMC*0FA%^t4T`Sq$yC063=5(grBC?AIIhN*pLc+TOsdam1q#IC&QOJJjTNYl#>
zvg9}HD6WrRa-O`8=#s7u)4G1OZzE1LE&r7|d)!!0?|R-JJbopqNq*gWKVU1_3r`~A
zlFa`6ynFVN?<;-UV9aViOZR+nRVprP$~{7aasNkN%IRMb3g_s+-~Y)HtI>Q!*-X?_
zy@Yh9n6&@=k7LPdd;@>w;+Hk{W$t`!wtY}m{>fCZAxuhVfp#80Tob_boDG+RebKEi
z7m5Br4nYuH`wR)w{ON8pOBmA&#LRAz4k56@+yYSph|R5VAZ^vW@3@r)0En(~!bmfp
zk5D|YxIri&b4{s2Sh1+I11=FYzmA^wF{oU-dd($50LkiGIV-HFWB$Texc0+qh>~p8
zSMz)vYOs*{D$fkU0IFO-mNu6oEuko{3$$S$*)TH6vx6aZ>>l07tBcMc5m!B*2~g@<
z@)3*vjB$WuihjDWzGQ)o+ZxFbg_{gdt?1*Y-?r<Px#z$99kZT;a+}>|pUk?h;P6~7
zvT3Y&)Q8uE$JwuyTuqDQJ;3gtFUo9fDDZqzfpou5A(Q8{XVRU$om)E9H70#;%7QMb
zKgse0JF$CR-mbPUlf(p_{*-d+x-Bso^?tP;QLiZr7!SIhapnPV%}o2hJcRDjQNQ#x
z%<#qI(;>_$4ErD*W)Qy}N)X&qh4<&Yakm`S#1lK_aP})~+1bsC`?A+*U~oKmd#H8V
zF<8Uw?ym%uHrrgj4W+&pLr=KFQd`An6WW}NT9>S26H@59v+u%P6c82+eK}T<DnbNM
zefVqtvVMlTaA`c&O=5?FG0QekB2U%f;M8{gRO{T&_7E;!?N1-e0(N|Vi}r8Vf|~}T
zc&2N3<TfotCQ@leYoo{_w-&$5d6k~5X6Q<s|Fe{S_Orz!Y1v`9&2;?A#ou-Q+`&5@
zv~4xWl&1l~Ws(ekXnGnV|4G^8SN`R1uk#brWnVV*GruIkCX(E#|3&go@~#(8T}SH2
zLAA_(t$wOuH^eeyO<l)3y8@sp#^fyP)~#*IA*LI(mksQ6+vrr)he|1*azGXG$&I4W
zVa0!P+R6*9o4um=Fxg*}{|hPp_n0c#>tg)oyvt=%%SLow*{-j$s1x%MAoDD$x6Z+O
zD_XU+xIQm#K`Mj(;ADTjbsCJZ4q|jslud9Wb_S)Qn1k8jI<%)yUqFI7!TuEPhd(d%
zw{sR=#<s0zA9snAQW-*_ynF*Re0+TUUKE3hj2Rhx?(fHQ_C#LBR@_g=5R`fN`i}d?
zC}c2r3Y9@Ivtxnl>t!9S^HnB4{I!xBq>M~{5NqjFUUyLEtp+y&K2Ic)0qKjEgl-t6
zk(m?#0MG*br$5bDgMd?L+!f48VLm;@%7L|W{hSQ@+ysCY2~26iRtWWq_SlM0-z?^U
zbU+DmWKCEU8Wdw9(=$R#&H~Ka^zD`M*JYGACxaLy<9(s)Zp@r{i0X)Yo8L0jZ+!TQ
zbxg!wKNAkk3cn7VzXKCQJ<zxJ22g+c6TJH^^tRsO?^mI<!1a~ejpvdUJ`(C9d3qk*
zbT0QTU6$d%0emxXrkmDm1d&E;K=TLF9f$-qSv9n^*_nb4b&H=<(y78HbzJxWp}zar
zKu|G17#B@;nX=Lw#ex%t_jMBP5qii<LO6c%@s{$QQK`tV9FNu5Nf;MGKfWM5%sl!e
zJRi^?w%nePF6;U@tJ@L8sK8f(Q<9OLyj#rS_}(zEOfgM%R4*$-i;@0%Z^QR7)5Lp4
z(#CHkV|M2i@aPhwk`9@oSsCCXw019l**mT6*v75Jak^*@Ex*cP!N*BH@<;3{*vt78
zox#Ha@T)_V!&<PdrhhfX&AlmRX2!9%lJ3PaohbkkPm{oVck(6A^*e06bB@Rkx)IQU
z0<AFMGg(+`zV%&cD>pc63ct7_O9*q+uy{9mv#ghHSkH#|8tvqXZ(KW4QZ{hwzL79E
z5nB4H?X!h>k<$o9TZU^f1-~<O{c@bx!q@rrPo+<qMN4nTkx24sl-*xpf=v96PL?*<
z>u)puk{%XIi)3&P5I)eTsM5uz*kz$%LS1Nt&%e<*T;D35JZ+);aHnp#SUwQU_<T>^
zPua9OAR5E;H2z54aJaYf_<1|UO8oBk*1hDa-?`9c@6=#(?7hn`ffXChH2S@iOrPP7
z-&H}vBh-fy-IrR^j4ZhcEHTXjVaFCRf=$<+*Y*`Gl@3Bky*$oUV*&^FZh@?(=t3Y>
zt0J{b-;8u=MZHV)vuF?BL7D6>y`hqhzs&%s{@zhtR;)-lXE4njGLRzJ90JDJprOnz
zY!^?63Oh&TqnOMu<gma~djBmC4Tu4u_%j8RFGGCq!I+%^HopS+MJmp6cF_gjb7b^|
zELPPG$+Ucy`RuAj@~iM>igC>LbwlAhj8FXH4vVcl3gMG)g8jDBbFQ|-Mt)69W8GsV
zqC$7SRKDL#5v@0j+pBC;25oc;y|a}E;QWrHq$Yz1RE~mBEsQ3G@3tU`nctIDni?GZ
zYy>N{k^jJ~4Z{|_&m_iJ<SA39llRR0d&O`1tl-6|vrmgW3S)|PEy{Z+E{D^1farYL
z9tFPl&lghqS3GzM@sFV+k~ush{ewfhSq=^pL=~PNgdvH^agx*O*jG2}YeLx)J)gO7
z0z!%kt+1tjIY@ZBgJ(!93u6WKr5HpsAK3IN-=R|fgiLwRkk2|dnK{mgL2un0?f8fG
zyG9E~qR>W>T>2gm42di*CMr1Nu2Er)-G22ZlQ2Siq>1*c3)oq6mdB%v5Tol@e>1Pi
z-Hnm6O`PyP1UM7C2%x*8ll3nd5K}7#%NghNAHnCk;x)hAJH&u%`-xxG_3vZ(OYX3A
z=js1H03Si%zH3Vrs^>7x+5P~K7D*=K`!&|`i8r1o-*?wo>GN3PC5+Yi@9LK+;gaBd
zA~C-K&DJ6?GK@1821Yj7Y_p`=4m)6fv(t9LruILt{6nOKPqx4a8;T<jKG4>aiL?(A
zBjq|gm6euhQ&(nMTAH*O(ZNz1Cl0##kS$wfmg--H>gd*?wO8g8kV?s>>_P)0Tg~0f
z9Cvtdz>EW=G66>Xv6&+YE+lZEyB~bmeE!vZKRW5YkS{F8*h&4s3~-X4Nfi#`k9<EO
z`7|xv#yp>L*TlSqTyupSpQt&ETT{G^x{4Yjg&h90Wkt<R+6*~F`WurRvKO<A;s_Md
z$4D+rqjX(&61VBnjE6X1n$>BH^y$zI>sDJS0VCSZOIxBwX0G~$CG-fus1nn=OMppD
zwM;4Lln#rQYmYQBN-YplB7me*+9F~W=%osn0btRiA5NRVYIO*lATRM~#a2~FYa_8G
zOkC`Kqo+r4fEfZhNJn~r4IH~;$;9Px+9r1I=L$%a-%WZk6+5I!0^yGY4HGnAs*P;q
z<6}hqtdjqYZuPN8jDEBzC>(7S%yB{1=L&JkKi+_m%1X_l{wqToQT<{+S0)o$T$$W?
z^Y)GeMqYURE%WGOe>N@MXw?KJvlw_xYE{;0^x6VP+?H3bKb7*&(kb8t=j#<R!HdQS
zehOe>vQNh@KRyY%#Xw@eJ{AeaE`AkhE9Vx@6zqMoJzGS5;{xMPT7z2s?X69wapDwP
zkKJE$eu)_qH1$uAz*S>IVB(E&8*X6a&ifxS041E^eERImugpCU|5?k(d(gZD_^!P4
zf^)653E)b-Ow4Sy@kU;X;5HXvBz&f1EoCz|^7LrOKBX~<fMx;24IvFDo|}Ay&g`1f
zTB7}|t3Ew*R{Kr!&ATX$!uwU9!Gb>z%IXK;RT1|5J4AVL^$ZoB4KPAkhwMMk>=N>Q
zN%;J4zBRYs_W-wPDeBpA8Ny@Y!iT!@j@n?o^~~`)pV!t7YC>fSj7(^3FvlKtsKFPH
zPq_jk1MeBF3EnrX1AM^th?kLWP55<ZbI*FA9-e3TCA`o0mkZ%%g|!F&h3~lio~r$<
z!MOKAF7*^DPodAKtU;$SQY}EFqq#{w*Q!l9dzqS4Ev$U{k?H%V(EHqc+FZ;VJpyg*
zt@785Z|7)$J33b}bWIuozHLQ~kztzAp@5NtWrr4t9X@=&yY`02R%8o|AOUj1k%!x|
z?$V}t;SWEgl*`7Cjl;HEY+;=QPMpxF15tqL^1&2A3poG)KmbWZK~!v$^g61vU(BCx
z{`S$wI;;kVPFGMABHAhBZG|NI)|Gni3^{?@ZmTWLtaWFaDU&8yr|@hrC>y8oI2`{<
zo3Hmi{7}+LEw0$uYoURWEd@{=dcZupQ2;z@mgHQRvIRzvQaMX1^o;Gpk3VS-98m`i
zzn<6L`m3D64N5g4;etA}n+rcJ_wT&_p)^sxw!zb<PL)IVZO!_#X31`5t(B~4*Ss;`
z?`zH&-{IJ}*~+~?xo&_dBHv9yx{bNl#NzE39)YMiOt+?5+FE2Mxh~}p%vvhdMi>=&
zoPOdg7-{8NG18nga-$edZq0=^;r=@<U?n6Uv|T=i5-C2INgOxA5`9sU;V}PUwm|8m
zXGjy}RNiLH@(=B(li9`vo?}VI1!+dVFBI#36A#rW4K41}LBeVC!~qyCugw_oXQ=qa
zt6I$TiFN8Z%#h4gXkAs5U#D-ESg`i+$8V@*g*1^S)#-g)l4QQp<txM$21vIKnsE}M
zu8ue+eK<V4mdGC-KYxu7hs1rY)?%mp5Oj+%-MnO#shu{#Op=5gM0(~9{+QG1@XH`g
zxDLX63dg%=#c7P>a>xvq8UF3ppUZtNQpK)cxzsljd7Q>bXTKQ1n9$OES2fgVe^qKV
zJ-W39Fna|P1Of~J;Nbn&ElrHdgyx3R=qMAg;20!<IW6T)WM;jT190#TmHqk-F;m3b
zvsx;YT`9(m-OmZQP%>+BYsG@4X2QBtZN22B49c_GAu2~p(WXNq#g$2#$^<9_utIoe
zf^pN?p>(x+7qLIH?|f<wfExOstlHxU7!d%|OF>ZL!XPa9>T4S`Uj9_qHrl54>bZu=
zi?&Gh<VMS!#jMyK+iYu2IVRZK`+Wi!`SYWHHtl_Ai(pbrV=A97CmeF1X%Lg%nbHCj
zgZRyv<B`W7)qA7g%$hpg>@O$InAdS;nV8#^t5#bl-uRfn)RIrULw4I?hXIpOeq8vz
zM-1ljl2Y@hC!Uf&8#&q593OS?L3XbU7)OJN_X&x0>JJm`L-=dsIUAZ8Q>Vyue3jk%
z(q7t#*8RtyeQy8ZKQK&y7VNPd@l5#)5VXfXl%@syc+O_qZ?%P)CUZVGREMG1B0KdT
zsr*;oc*~w;@Y^t<vC%p(-+iYYEWqT-=E@%Y^ZD}!=76x(P0i-17hdExC-OB$E|<xy
zV-GvT%+^`UCjgl%Zn()hP4~myy|38<6MSbe$T#13BNshzO$UtldjiIbbK16BZlUW;
zGl?^$#s*?;@SBPCf-mQPZO@e9^+Y<>RHwj$Gmkma-UsYlVg`xT^3*G@n5Gqh77J(1
zy%K&p^~gsOJ|S@&4~4(EgyDCCx!!c+IrhBDZ~8;KyS0W^H8okEk?+Xy{0~b8O-#A(
zXn4QM`EBz}H!)l2-?{?J>crr8<uNZymoK;HIL>pdr$SM-z)0K~ckz-X=DOSN9B9^_
zyY6hAyAK2;ZrroOGxy3HZw^EsEK3-iNlW!ibM6ll{eR)rza#|%)J>c?!ECd|T(k3b
z+gfnIPXM^~)z{zHIdS2x9LBb>p}}mt;Re=Ph7U}-wD@lANce80-TnO~GNJvPeIL0r
z{uQ7Q9tHxBIJM{bryhtkv2yil>kR(A4?nW+9M(z{?JM7-)22+0isCh`ad^#|JEy+=
zezd3AXv6j8`y~KYDYHr6KAsugH><S{ha%dD=GRtQ&okw?euBmYpoMt|-d}BI$+BhE
zKP~UNTtSzi+L23qE_=3g_3&KwOPje%--U*{2Gg{9r75YdGL`z~b&2WSn9%CrIc^u&
z#d(Qy6Z<93ieYklRNBeDpEMo1w*n=JJj8WrA43_f+KX3BY%pD`o6Y2zQ*3cApDWwF
zL-_5V9jS|b5$8qrrI0&2x^{%Y7|v|DoS|&t5r3l2BN54xg2H>~LboZYvSIjv8N^cY
zD9gtvFv6TK_-?V8y<YP7zR*MAuggM|9bOmNl$lT7Z0d1o?fam4brBeG15)VgP{0WF
zxs5~Ihc#1R#4){wDieu{BM&;z+NgDPr|UNxw<n)}Q3s9Bw3$ekQj8a&@t6nOYV*yk
zq(b`CQnM4_^70#Rm^a>jH!Tyt(7?!0V?=!1rZGw_+4a+nCv$nA4w}2~w4;lPyxuD_
zUhjVP$-oDX&E3W4oGIzc)Q>U+ysM4*BTqeTw$s6o!ySjRD58E2Z;w3nj3m|qMqyOC
zLWG461%)Ux(=wu%>Y3(U+tazt)o71H)EvgGDLz=3(rVNuH;VkSU|MVcYpN1zl2Pf`
zhOm{|<W1D!Ae}U&TMNA27g<M?Q^aJ`;Zg5?2aFVItb9O<dYhLvnc68jyor%o!NF7K
z=<op`m%7>10~n<1cP{0UXDn~_7q4nIGvr)(IQs367yFGe9S()Y6h!*YtwFzzeC`Wf
zUOsIpR9+Z5oiYp2CE(I(O~?T`{zIyz8B(riheJ0S8&>l|+7%NN2epf-)GG5+6XdKM
z@Qn9jM^n3Xu8w9$g@Bb73ANxry{bwQ_<9z*RSuv7z#>#FAf!Wp273#p9TQjfUS%Ow
zjmAi`z(|~iQ-8aFX?$JOqP3FPI9e?hnfYp0+i>*Gv+mjp0An>ulonqsIEA;&h1DIV
zu0n0DuC)LQZ2|D1ZQTNgtS=I^5$zO+Isky^*NShywPEAsPjyZ3UF#E*h68X}d6gWe
z2O1-#+Uw2RV;6JcVF#yl=lNG<O6$qTOnXVNZ|sr7^*Vvg|NN7S%w#be<LIgzZj@8^
zDs$!u$ES6`4MY3x2OmliVmju~L(Dc?&W$TRB^)Q@ZEfuq%=pz$E=+qy2y45uw}#nw
zpnM0wL>!9nt@!X0PYy7|UHYkFQvT$ev)s>=D*{D#-2Z^tXOG>@-n;EOa6StsWH0{b
zC!aQl?!TWEPzuMV`we6Anp<v9mn3am7#P`Xqd5j4yD-r{+8IU12uy(q-Q3b@uDbDN
zbIjp~W(|zEa$)2ie&GIQlQ|p4j+aZzwY9CyJoDnq=KYU9vF{K+u0-bX{^h5a_(3W6
z*W7)#S+XDyD4QcD@EP*&n2yk#`Sgn~Q<AWzw%BY_ou&3kT2nBKd1WEB@6L2&Cj8N7
zo-=<FGu$Wkt3A9Awcpre<}6ukIKZqkeY&3{@*Y7pe2XG{$h$OO$`%*_oZN2fEmNxf
z`rGf!ZTH?k5PtG8M@4EAxdd28z)AwdcwKnb8D_db%Q*T*AjPfs+?Ny`$LB~&6#C)H
z0SGvKh&2$(YC%mvzc4@tuz(iLUc2v__MPa;LlX)h_K9bokF@x4_Q@xj^=Hp=$y2U8
zqu+f0y%nnD{2xaMZ{GF5LupN;lb^LWXZ`gD)-CM=5J9WX0Xwb^=Iz19p0H!#!q%j#
z4~(EO(kYXH03}u0zx2wy<WK@l`HThM?`5s+W=dU+S+cUltTQQoZu6<9n78mF(AM6f
z{i)2==Aki?3#gPyPgqwiItL85kCR-^xY?e$avdXpk}`qr6*&k}#*Hz&B|uIcE9}{#
zMBnru7B7iQo>Mp$qds{%5FmK@y|)I+C|TX#*}wO8f&YuSI|kQ**c@)Y5VPK=zM1J6
z(SF*IxX%O_8LAy|8;1KIdCa2zL*@75r%J2gbRB}i5*0{OAo+kq#q=pt%pQ^uK|;b2
zQcp;5z5dsCTu{mtDVtM{Jz5enGm_%aJV8}^byJI#oIo-OaK;fELgJrz?gcGKeX$~F
zIDnCeYLbF-0Y>}+kYqwy2+1rYS^N#QO9wFWiKI3hA=!b{*6dm9njHjw@Y^R2s4u?u
z7sHxhlfz~SNh~DR{P}WuKK$e}^U&i@#Ez>=m(MkhiJecGndbEw;hE+srW;!$5H*MC
z))aAZ66#T9jji5}uhtaDb9~L$NVm3-rnCy(c3=7~-?B$qVkxe@03)rP9r7bmZNUco
zg<wvqyRE~55H$_e250E_kpLX90D?dZ%xmFfy{l6iB?1$O#}pM#%<JpqxLiO41P}oM
zPykZ4<C7u*j985l#g~ckMLK>1W)ST(QvgG%<m1GGEdnY!R0qE)!<x!c7<)XIS7#9m
zj<_<ZTg3|85YQrk1`}N!GWk?rB?-?mOhR>74Vlj74w9N$nGeIEe2ugM0d~q|A_}5!
zrP4}hS+dan9xs2Ysse_8i5RR<Z%?NMI7@Xv$Ir+C2kmc8KJw6%?mYkM>lPSk>5&$X
z=8Qqagw~dG=9oEi);D|Xu$`&b{ya=iJ^zB)Zkw&`o*&aNfM8+fnajgI^}atpn#90G
zO972#v;+VgqG;jb#pbPd|7J~dVcrQQ9d)2h-6)3oE<0>*CVAk^2V$0gE(z%`B^~Z0
zyeSuEB|=wz`;gp*@&CQFt~qy*#*fDdIsX29t!#vj=5Dr$*-++s5a9CDQx8mg1d7nQ
z#|%vzA<2E^^*5P)BoV%C0vtjxES!h)4a`@c@x5%t3Uk{%gN*06I^x2^zzAIYi_bkP
zMbn%ygqiBsO_`N)Nc^)4&a>J<4&XUS%<pV9MktH9JN@|MlID=`?-neyb3@<Pk>oma
z<Ft{8=N<B*v<6<0V4Q9T2yF`hFeLT2)ci!9o30fw@<Rd`;oJ!z13<!jMx~DfMmU>s
zmU43jX!qPJuNoNcyr*VNpC%w~@1(u~j$C!)EoQZttd87%b=~#k(EONVt%it;W9$KC
z&>CWW?77R%$>X4U07noG%$KqOMi?i+1jac`oH1W~?e&4+4Q9{QnRKR|!zfVF2S&)m
zd`y`%(d@qSP6OHx1BL0v0UCZ7{zNtvz{dxnR$y@Wxnvw$H9uAzE|Ky3g^LVlXsl!$
zCtwe%)bkxBFbD|MD~(DBlXzX4$D8D2p7XOKzDod%J4s{3Uq`HA-eHSmjlpTGg-vy#
zHN&&U`@<*J<+?Lw$Pdug_MGm=asSTy9|*jjX#2ps5pwI%{_y^wt#rekcj+CP$|tFo
zCd64Mo?zE3Sr{(>H`<MGEAzICz5{@ve&51#?NvAKZ+R{;rPZT3MwqU-X{99Psw^-P
zcZMAIoi27D?-6kC^L&7jyc<z0U{pr{-bc}t)-X-MSOP|1vY&IxNmjkm@9rnsv^}5$
z50V8CFNHZ9ZfMRq_2i@vH1T)d_n`UuyYKw$_MS})5=d-JePSbVhw}KFkIC9FVRFi;
zi;|v?_~Xw$Z~pq;-~8+Y_u~;5agAq_bBYcf8wnf<Qy3WW<IxD&b;s@fAbY6fFvxg6
z_RMqU!%si68+gVNb^5R)J;s$fqNJsg3eMxlXnsVBjN9xdE?ZPT`x>&#5jBT#YYHib
z9vyJvHnmY@W!IWwYJUWZ<LQq^@jg5PFzS2tJ&JoC!|Bug9&7KVbZs52I0uZhb@f~S
zAZ@GKEEv$!A=69R6YvkhoDwFs@X3KT2LMFir$VN(1bSdfscvG8Wsu-oq+MD+5azL3
z<&UCHKtq=p<q$vxWHnM$79hs}u%Uu0lpg{daR3Xyaq8Zd@NEG&(yF#P?H05>08j99
zQQfF^qU9mY62f7O)d<vS6}aI59TjBSdZc}#w5_W!Jp<z(iN2t2+R&u(0et!d@&F=e
zlVn<#E}|*phH4rnTgUO0_0^_Z+CbehWr&1esk9zTW!hoh-g`!x(t7@t*UjTkKW>`w
zF_HjAY^M{GxBD*Jo3oEUCS@F8{=y`~3=Y!!*wx-%6684Z_L*m7#IMGUcia^zu?(~D
z!n00K+S9wr{`k<J41h=+VQOY?NoPafe!sv<s)q%n5HLFH<P(w@|1c3BlC(3L(GcxD
zK-wf>E$!tiR~q~XV#3HL(!&%-W56f&%rHXf6K9=orFFE$+$~Zh;4z1Fr8F%ZGn4Y7
zoLLXIj}c61t&&+E&Q#(3KTLe2;=`tbH{EC>bF$1;(Po&rU(Em7<N}N!g?{>p$60X9
zA4AIH+(G@00Itqd8-1oV^9*Pa<&Tjw&S_e+NWX{42S!{Rpx%i`9T}<F5Cx2I9=`bO
zGX#1}bVb<p#B<;iW|Yp;S=gBcBM!eV&ZIHF*{_z@*P2VtJkx@VggJWSop;UCFT7;W
zX@rHr7{WNyp+VuyeG!%mWdn>%n>xi@aK>p?lfbWwwQ}8UcO*avKgQUxMSCPnVPJ%K
z#$?k?H!)|Qd{Q`hBrxI<@ol-_%+swoZWqQiQy(My@AV^t>8~*PK1Q0GR+;+7iTa8p
z6DIl&Kkxv3ItF7f{db2<&H3Oe;2pk%0DyeLG}VKTJT|~f@#9$k=bwJ6A7t-&23YHu
zzs9sS-?q7%Zfy3Fwi@9+@or)*N4-z?+HDszPo}sWeWQ2j9n$)7d>z6mkNNZ49W{5Z
zjC5B_4%L<>3vQ-MKN@jOt<9zc0pG4pyGQjkM*1Zb;xyAoqiw}H$L<Kk@gK%+i;FEQ
z^a#NFC;}se9@Rv7V+j~RVu6poPkd<EIJ2>Kgbn{{*(X2lh{LUoO&87;iG}NQ$RM9j
zNR*tR4YL0sC(Vub-ES6t`@Nsk-tV!?PUe8U_eu&!s^QvOZ;#xNjb~tlPVOPe84hh>
zN(YS0)u!@9ZN8%DxtCu}vgt}1e7Rh3#;I0f#?OPA?(LFJbJC!GTp{j9^CMDZ+!kU6
zQ_Fx!2}RjiO--F&NYosrTT{d#1%y4=Fh!LWPE-7y`3MxpGxM=p8&h^kqP@RBrnCxe
zUwR~fmXWB#UsnzoX>ALfY6A?QA>yb{OmXXwTmkkfCR3lZMYt~!v!^csfRG+-i!K2t
zn4$taVZUZ2RmBs6Sp!Jo?-|-v4_YTGQp}dXZwQDBzDv3VVgSnYIp%YMuY^WQK+-M|
zcmn~*sCd&U(4^0TFJk;lyMuN%Os+G{GWk^|b6YTNdld$VgOl*CQtDPYodSWxki`i*
z46JT}Dv$+$bnR6iX(QCSJk`Hm5HmLa)-=lejLc|NG}hbk>XXI;AWxrs1YsI)e@Qc+
zdR!`F^0^mZk%^zDBoW(TJ62LE6F`_*O^gFt;Q#W=OOk*EHv)fs|9$hw6HjR@1MLX*
z@MufGzzx$=GO7Fe+tFI%KiYK{v)^8Wj5(JV?F*!$ed6?h*#dkX9d(Gz`D_)K^N9jc
zV3eMA!m%j~@~^)6#tL2e<>J@ivJ1{j(X>X0C`#xG!6V<_h+h_3><DW$uU=u+Uw=-@
z-Wq9ad<i-lZD0g23v*Us!jA${<31g&G2d2W-rCw`t`N9`x!qiVk*y_te$1f<+vn1i
z&wJ$BTko)b3|yGI9=iWL>&wLb=8DNP81YdK?KvZIRt(cMcgd92!oZZ49~ZX<qkxh1
z)?L?}eaeYRZU4i~w^{!mewo~#dfYJs%w&Z8=pT%FN9;-W5Wsi%e)C+Ay`nupnE-8$
zc)k!&<V@=mO!fN&Cgnq!03+0oa3bck!+vQvgFmcy0lqw<2(5!(Ui{;fs3_(-V^Sk+
zo`3%3&y&WO%uyO6uF#PHMw+BK(9k%^y&i4V_$s<c;4an|e0uBM_w0MX591zFY3H7H
zauTq_+P_OC#qbFdCP2hr%6wpyWd{IZz!(1+Ks#=?9d`&9W8Y)E&wS$D1L%TJ5l3%H
z$n^10I^xjE!LkRW=9*_;e%ZYCCV)(K<eS9JF6YRlc4x&dz30m1=cm7|RX|C#^)XTi
zVV;reNwKDp9035eTCJ0!#>hyH`q-)q-bWD_8KyZIOTY+rKQZs*(*miv-+0H}evp;)
z=#-$s4x4WwiJfDU;`nj9q!8Yp|COot7?a0I@+zz?g5=57+7#k!JxpxOE>e5KrlOvB
z?s@a(JMV>)+fC$or=5}#{>tlbTGi*22nFP8N+VxWT7H{Q%{}aZ{rn(%pAHz=aJ|`9
z!YZ5z36`sGy45U^)QfYb>(cleco&A$*3~!PD)6hx1r_wl=A%r9@X`2)6dAV#oyt)u
zENTwp))XdfO2lA`+tfytH7cE2Ln<wfXXX<*(oFKqTQ+<t3ki%Adi?VJ{tw@%<zBJ_
zMx;z&aK_y$AOsYkLK+!p7*t7%qO(=zmlOv%Af&WHAV&fe0d$ZrYyl#H&2jvTbl?v}
zHu3>O(7wPqIs}udWp%r$sHu<zia?_SGh0ZZR*KQnqB=3z1%m|#=uKkG*U*LzG4~bT
z(b8eHILcHmAWoZ1U`?D@Yki@#3mAc+1E5hRV1z=*)28}pf4|z`Ok<UIwVO^^6t-F~
zY8SwWbv!T43aC@{Kxm8ri~#n~zwzc{q^53y?MKz5Myo~BErBI8MS8nBtmeo8`|e{-
zJt0-|<GGh!6%*`9vs(UGNH3JbmQ$*6V4NmO`{956!_QNgasXhL-*mHCv21zL{tYb(
zz>zS$_rXWzPftD-4z|BBk9)~^=Oi%-i9ovfo`?Tr*NDpqFoO1HrNEd=1!UkL7zfvw
zC~_JYF75QGQ_YXhJS|D+ia1QfzW>oDE}6Xol$|Z6^K3c)_6d_aH{JC|Kgiy1E63_b
zA98R?ILz-WZ@Ja9x3n3)f%Rgda$Z5oo-u>5`@o}*Co#N78yG=r{E~Cewq~;Y4$)Tp
zMYQ&IS)2N3KK$&0bFIU1AKb|W7&%v(6k+B%rjzjx!#ITrUkhlqFjhWA0b;}>H9zl1
zd-zH1{pGjZVOA|$ni7twphH671c`iL#7VkyUgg~C(^D@9_=_+XjTQnR&=S@8s+h<y
zvVH28KeL~lf1XT53)u9?A73xn0*HV(K4JFl${TLdnhM@!eq5&aalptDJ>M&&ImLM!
za06%HzPTpq#;U-}e|f_?6_2ZrwevqN`)Nva957N|S!w?DSHDcEi_D`9jI`*P#&@2-
z3NbN-)r&CQ`p5gM50G#;YZ#52b*4{Cib7ioe=C3_J~7@rcfO?z7Y`r?Fy<5G{#cqb
zVft1k>TbFFo}^&r1;0tYiSoq~V!ic`_og_PM}YDy`8@%&!))Q3(rEGlnWTcfABEnZ
zc|{M~E$7HCgf6tcO9g!O3HU3mt~M2VH&xdJW-EuS@-?_%xu-9ve}}O)RLyP8(poM8
zBV%n|MzTHZ`-;HGFpc<F0Y=yi{o<kvQZ}SW5+PCKlVeIYHZ+>QlY~Qf6Z?iF#&E{&
zDC(n3T)NE<s;D>_pEw~n2)eSOq`-(9ZyF<vnZGeABrt-y^-Xu)ZP=*z;1uH&rHMm$
z)W(bOe8@@oy7f{7M)DcW;mNyH2m5JK@z0bn+ku#=e_-(Q4#(y=ius;yO%WFdj35ct
zC?>{mP@((`6snxz=wrCdsJRKJS=)a*27I&L#~q!$W}2KWjwQ;~!CfYOa6U#VYk0@V
zOjGQeBa_k!n`SusooP$m+K5WuY#k`3YQ06JbzuM_9Wo2ilW+=IEx@Biphh*Ig%~k|
zdyZiLf!QngGgSg6@E3ylswx2nSgdlsLwMKo8?XaV#Q6(>cWHf;@JSHi3>~1PL%;(d
z2xhZL15t<jc5p{J$|MizD1tB*Z=Jsf@q_93O}-ACvWuBtsf0L*hiF4-XPYSzBZG2q
z4BtsVXajM|N?$sqxdLO{cNo7^Oserj_0wir4H>jVFgw;Oa0o3R%wzS-KK?#(zJB^i
z$ES4Z`Pbesk39aU>DFPst4Y%7V#1Yb&dCc+oj%3<&!1kTb>cL5f+Cj7*TrvcxY3kp
z9R2;<6!{{&=<G8!t^vRjqz*t1nAmZIAQk@feCn29&gIwL5UB}{oqn7XvtNcG4P!b?
zQQFop-C@4Ngbx!YWnp6bgfP)>uD;g#Iq}2efDwFi;F|<;-$0n|Xw_gI_?ra_ObLt8
zBl<9!rnG2E)c6B50XP9XL7U9k&`ONcUgq-c&px&1h-`roe0$)hqykXWBOeUljIeTf
zi&=MCV6LiKOvztdcz%*uDzc(cgl18g;((FgUMurpVqAnN6JW&U*;3%&@geY!wTd?W
zyJ}BY_n_%Dv@D{49#__siH+tb=bf8!R>p58jAus+m#j4FPM>5oq|lTKlW*=R9GA)O
zIAA1QUcMQBdFyTSm$%=x8bs;Rt|`C>&vb1F7(w%C8~IA|iT4K24w^JkbmYMY%Jkq?
zNl}2R2rh<wjl=;XJfFY4=K3U%l>*E4{{HnP7pFWcXjmcW=o9>upUB5hSlb1_3ja0B
zSF9XJ&X=}XKwg-4Fp!|YH=G&yzGqTjq0g6`6}x3!px^pst1;az((;luluE7nA}~_m
ziC%+>Yj1DWb61(PCt8ELH!NKlPkfC`27qrw5g5s2q;g0zR)7&49Y)y(efq_hwJAzg
zqx6}Qe}CyuQl!il>R@n#4keEG7+-q+IVlodICKA#)N<n}O4_KorBxddsdjstQu2>C
zV5AHNL*igh`wIz-;8)}U*)fgUMEGj5)awXGdN^4tB)R(i^-{?4!(~M=y4;~PoVNIT
z{ER@<9Hv`SNOLsQHVkyotwREk1gqD9r$|GOfIn`l#B7DRiC+LmFj~uWut2h|NaJ?|
zqTZW@>P@!}y)a+nfXYJElS?`2`t7g%T*@A;JPsJ?=$B)1jeBjqq#(uA*8c{O1f~bi
ztpgQcNa%kHE(Dq$0gym?3v?()r}ioAReA+BR7>-O{0?j&9l(Sm%EEM3Spp;>hB|>h
zs>Ph?us#I<Kx78X=5u+2cEF6S6qB(_^}&!M9pFZdm_XfPc5@F=zoKR0r&qon%Ei!V
zF=ci2HoRLviz^?Hggms1`?3<ilAs)ZFZy!6EpNR1nX<tKrd_~Bg`CI(il9l;-6n03
z1i-QPZaWDWIWDC;_!xQQ$;afEMPsA90F-FWbhb)^L}NQ;$|Uno7hjOFm&b?3<+t2s
z>LqdP10#*n;6QK*N98_YV&{$oVBm+ldz@wCw<AoCKK-o0G*y^zbiC^h+uLBy6r3%>
zggIHv%KC&)hASjp-y#z!VLEjGRC7%*AAff(;7oG*IAG-E*WWPDzw(+bI%<Csu9M$m
z0T_wv4?Zsb^4GU*vmG-R^G;V>f0J1*CJ;=^Y=IG+N}nzujDGtRK5K;EwFu~irRJ9`
zTPAZ_ccy?3*#aZm$XxU>2Oi)z$KLxIBQA2rv}rOwbV`ag1My1*7``qON`L$GQ<({r
z)?0saA6*#^_}*#@X`dXHy6(gEI$<)*)SYznk*=75E1FfGe(}}7?`&`4fDs&nFI~Re
z7KceE&g^Jn`Jfh-9q_O6oYsAml?O7fDZofrV}vz)v4G&I(xmYT6XVz1a(fCt!lj8b
z-FBP&#eDtU_b!>e#sMQdhyQieHA!jtLHbM2Ki4!i)F;K!Cj1Kd#OlO<PuRy0Y2c-N
z_{nGH-48x6Xpw|9^89j!>puJrC{#Twv}S9x@2SRgq}_=mm&{qQw_GMSySq$3-!i@b
z@i|fiMv}*AG=Htu+)O~rbqtYnC(x}uq;Hjl_kD-7q06OZJy8G~6lGG*U+)Oa)k}~m
zYJIvcyYGnmJ(jNN&QaO4(;t&;@<pYkEXV<_{N^3rg|@hOudFA=6bjBWgTQRK9qu>R
z25+8LOUQQV;#9#KTWF4dHrZ@5^U2$9+T1U{_trouC9C^8`}f{1P$8YWW3npgK&mWm
zzWD63q||*gcTD;{+CRn$Fv8)3Lr9pgk&A?nPd-x;)oi54d_p3E54n?0oH=8LwZH49
z;r`V(-!jj>^h!8*;P;8z7|zuO(I@LHP#^u~kc7r~14h_<`s%5L1V%Vua?p$-=7P<B
z_^|Ts2OpY0KlyZ2mb}CK^^$jCX$pvXL1%hsO;c<!YLrJHY7Wz_DdOT7u~B7>a`(n{
z^~LcVdior*gOpL5<e^I#z7&|NS@hixX2yEcEu*<oPMqC)RIF*!A`JxRN67WV#nUR%
z$UDVA14)Au?h1uMg#<>NiIaiK^1N~V$+MU-l^tNDheN&&_karl1z-aAH3BDkg(&bJ
z0D_x2G(iAM$^~cy6kN-7m>@m>It4;t?g~C9pn^Nd0X%THARMZ&z(%it5Shvnm;%58
zo>8qm58wvrl-z5dgfvyZqcSDvku9$ZF*5)$AX}z1A|U6$9qkXo=}T90vz!|@7~0z@
zCO5zc{R2?x(C=FHyG!{UutNKxg<s6K*YWaa!mQZ>LOM-#L#<_C^+<yTQ&{2{3xMp3
z*BE*Ab@P~j5m_p;=Eukfm5FJDxwUE182RT*QtjJ+tn<?y4?bw5Uy=ljpbY>pf&+4&
zzWU}{YYX2G^Y3At;^U)M_S}8K&)seJ+&^$;APy;fOv?bmz_i70Ls%-GeQcDu8kl=;
zigAtqnJ{6lDeS+2v&!Yy-Dp;89r+XoB=|l_qP~H=7u)0u#+QPotwU*yxI?(#hFtDp
zsyiPr&iOj!FkhEmaBd2qg(<OHWL5_aAmXtg@WCgang?YL%u%+$2$J%EEMW@!Gzm+R
zGM~OL1oC6b={JA4F3DLt;n@Ns*=UTg{!TgOD6{dL4XG-UU@GG;g0;3lW9U1w_A?x?
z*PdpNkZCO98RO>W7C))I$K(+pVwj@L!-dnN`yB_2a1Q4@?i1&F<_}*hJ^`X2c(-EJ
zDnHoXuPMNY|4g&rnNU?@F8lF$*1VkG=_N}6?DSsZgBC$?&j8v_2!{Iavp8U6xlEj)
zl@cb_C+}cPU-^Vd?8~pe(GRlsm_I!3@IzAsR9!;efdG*2y#Jwql1~kMoROmNK60e$
z<*Zm*Bh#L3?WR}E(h@WQq}^O9p@X<H<VdxxSdVb3zHV#YajcDn<3F_NJJ0zDe`wY6
zhvq1J&VvchJHnl%!tpL&E?4%yZZFPwLCr1l_g7mnP+u<P=9|ZFSH5NCl7G10DSRV}
z8Y8*POa^(z3NW(w9=n-+_t-t9#rOaDQS;GfpC^U8mjV)G*tiVa(BuX1AHThP;G>U}
z0(N7=goXmrLva+P^3KN_NwBa#j3O{HIEt@H!{?b7Umgtb?yzx-J!UpxJ~7A4^>C2C
zUWTLmbY=P*+;r)RabtG`qUJE&n!+(!W8I*}h~KlK#)v-}`QI})J=L{iPNFCOo~=D4
zbjzB5xmceGzIH)7UPxf1pbc9?3k!UVw98I*jkF^G2LfOL2jq3~1L7Ds00e*rtzwFk
z&L_YTKm~{x00=;lN=d;%fsYZ&MRP<ZWV9!%60m`02}QZmkk%p<8Y>WSN;+`0f%-!`
z+Qo=~2AZODemiWOO%b$*GKsf92gQq_VQTAWl5HPAPPafAG*^5A5a|HmQG3Zh-hh$X
z$urc)U`_!?I##uq5}5+&5knupD*NoYt2zC|RE-gQj6C}ElSb}kEHHxqnI4UU^T{%)
zzTW)n-(8wwO6zZ8-rf4pL#9r9eP3fFS73y>z?U9^Kt91ByyAwN&FX|XC*n9a03-k?
zyzuH@%u9bsJuAQ*yhJ8qCQZy=YtaV~T-)M+5g63(Nni|6IBI|5^843JfDu1^KKF6J
z2xa<lt`5c@e~*)7y7<)Nj<#cd@56sKpM3G9OOmz5$nHDsXb#+W?}7a81}gzDg&iR=
z+aTtWBOH%oLd_@6egEh3tE7z<Fn(NI959lPj}cckSH=UMG!6B7x1=H3_pQvQJ@M@G
zVlD@o7k=_1wB~mRt$9B_(|Z`|m=(*FvH?b@1AykjGtZEwU?BbM>cxi&zC^=7(3%2_
zgwF|R#$EQ~^HatlY#qamkGpO)cXJCyxZm!|`TPx;Q1i7j;((DK1ioB*>+LS<z%^T7
zg!j)82OS_YZCed~X9r44zCrNxWxa0!Nqq1oIaT3)6_Qb$O{+LZc9fT!5;WqG@W=VO
z)_SG(7sF*r%NH!lrme-aV`l`?@gGL2m4y<b7^(Kg)ic}xMqIu5hO_T00wej3Vod(A
z0*u5t5|48J?Y|VzK3I31by9lw|NicZ0eijJ6pnJzjw(7T>2ZXO)$O+0G9~)zn{F}7
z5;ig2%OPK2BrLU*QfWj$+?Yj4--NTXsqr@)=`=<@`uy`Gsa#i9ob&Z(UU}8LCa3dJ
zgmlS;XPvG??jX8Nx+v#(QTcNT^Vdr*`SZz>>7g~BI*NIRcLbv5utSWna!L63%o<vO
zltQaEx3yxD`I79eg;8a>xHaKA(wmHt9?#L1LMikb8R_|oYJa9<nyrRKlO)({#Frkc
zrnI8QV!VerQ(8C>?GaGnOi=+OU<MBj3$!o*Gkg<R_ylQf>o!%@g2==`X_GksK!qBa
zTLN5hzBw2l2d)4vSnx&Vadp}lXn(jm07o2AAD{~PEoh=EL^wW9Am`Mbbd*O~q(}P&
zpCR}Tp+xeNyH@~Cb)~djBz0XOlOli~R+B{e=uf#cBGhFIgfJcuV?xy(9o?pB(E=Mb
zUj9s+y`JgQ81!@`Oa)X4?BX6l7imfCv+IsA&evZvk3RE+=`fY1r&Z4Dwg1CMNuK~I
zoY7C2G{OAK&r{9f{_VpL%^fn|Q(Y!!^#Vs6<pYecSI0~)j5?pf`(|hI3a0Z_H{EKT
zL;LBu515JPr7*79TQ|$RTBgKWy5;VBO;d&%BM(WF{{4g|eWu^Qu>y=lITdG|p3}Ku
zqD(05u+7$1S{yT4XlgjhR%0X%w1~1o;OgLt$sf)(Vfz2B{DYp^j-=qYI<C9>k7n_L
zMM-J-*6t?D2#4%9FDaZKVUt#V0yHNMJ79ma!R%QnYt~Q4J>&J*)6W{r89QR$5%Pmc
z9wtmf_a!vp!f{c*c}D!Qqtc`c%N7`+tSD1;Jd;=@2!k|h3NX^#B$K<16ZEYS@F{8U
zDYf<H=kk41iMb_@7A#t9Zj?XII6_O~gpi*a(jpLwh7J}m91S-xLOq=O@yUd*(MoCW
zr6cC}V+l<DNtPtSfTYVen)_GB`*Cf_<qB`22f+fpSNg5Sh%`vFcPiKS1FhRyPYWQI
z@}u%BTE5!MoRaK2H!58*Z0wFe9RDGg7H4~|%B7A%=V8Cp-7jG&Ss3ss)3c<XXSnZ&
zYkb-F6@igqn!Rj*5hOb+B-t4zZ1zT}-QF*C*N?mnS~#EoJI)p+Dzo8@n9t~j$&M{n
zpE#8KyEI1lutG@xAO#dAB*2gai=$(u_PKRvGfA0j*qx(5mp5Nv#FdlHH6LIkTu(Y+
z<de_82q(9{<A9NL<BE+`w0tn-<r5pA`-&PP{*a7``?YNowP3*#Gi%oLq_K};JoxXR
zq{R9`T$~zoR9WHtWBK=3kT&u`<U)+6Q<uh%iL=Y<<{+2;7TV;JYs=bx&E+6kNR1J<
zP6{;-x%8({bq`lL4j5@_?U4q7v?2s-K;7+Krc4gl%LG8w=<0aVU1GA^cmWD!0v_11
zRm=H0z(|8s@c|+_<-ENrp=|;H0`TBK7tE$vkVP2*D*!pr0KuG<ZbgIPfHSqK+W|5F
zAFZp}OzngkF$bkNqI#X_td`a7rn+9>4(&_qBiZ{QP*N(uhBm-(?CMfK^<PC#hXqDj
z)z%7uAzhuSUqDDroqTZ!JOSjvPYU@UK%KU>u5ky9)K3p2>HE3@$@y~m3&AW_Z(EBg
zt*A2l?!I#jFcN5tw6qC05+exh63CXLF~!Hozx)j6>&cSg?@O}(<_8`!^~zHyV1#+&
z+<*@QpB67!Vto^`p76nk<8FLSe7|Usx%sYp2F??Ho<L*dY-x-nGn()f^50iq>nF|W
zK3-!a85rr6`4wj-BdQ`yX`mG`lqs#Kyt#z=S|MRg7^MI4O;95SpigKz0C?5O%o*oK
z&W}E23ykbyL5O|*BJ4dTdokS`N9UX>jggS~CeD=q_lm1i7_@P~$PItI$1MDQVM-O*
zn9}ky!?fls$~g@S39P#~0vx&J?t79nM~*mXe*sS0BxT`!cKz*lTE7lS5ivgs@x96g
z7>PUQej}}sTV;ODr|}GoTzT8=X2p_awmV^9WRkR^FB32ler7N$#vH~GYX&|>nDFU?
z`L&VM81dVHsXCrde9MO4`F=X?UwPwA`BQq;eTKu)<N7ye#WtDjES1)p4~$gkeKTAd
zBP;cOm=yZXj)g%StL+*KV-s#eYkP~S(tgX|6NS@^wcmyKevGv~WzwF7%T}9przY<Y
zGs!npnj$bV)ClDfpA9e)S52nC$O@t2lS7+fSLTk}ZetEVaDVr6;2O1w^_h}Cl~eV}
zAvJcKg>fixz{regQ*{tJW1viTgR?(22yuincs|@dbxXnmV5DW7`xx>2#C;)wkr%X)
zdF1)$t#qjG7>at<nLge67zuCe-}~SrNrpT{l7f@JUJ73L5Cs;3aqFt^8Nx2CcpbhG
zSh!@RS$F!RNK+MYYl=DQ*89hIzV5fMkn6!OcWu4TXFQ!lx$nPQY-X>wPS!&RXwojm
z=+uy%`>YEZ<_!4pz()VtNQDGO3O#;q{ZTIOnz~7K<K843z(`wXhv}5_b+jego7x13
zR9Zh2<pL!joTxi4M=!vFn9wC+^kH5JFd?`Gj=d{nKC7*%!_+iXn{K78lJj+d2Ka*b
zBfyjj+8ux$B?29qmrHY_u1Y=(1j6VyKuVX=kf&C_N4vBW09tBlD=nbXCIAFrgfh_Z
zsFovj%v}N4fQR%beRq4O?0#2SAf!TNbPBY=1Qh)NOd(%Iulih4DXkKj0aIJhg6Qq+
zR^J1u*h;_>X-A-K!QQ7_`Jhz`Q<<kWVvPQ&pFYF%O4Fr`c~BpPP+6^ye%ZJ#5rb#m
zUVE6+PB<o|_0PWeih2Cmr_5>rQ2;3Y@_$h-^FUVHM5!k=)|>zJ55GuZl3_~gj)xyE
zsKyA5Lu&beefCOfC+CH0WHM%{81!@2U(cL=!f|%apM3rW^TylnCZ%;h09m4ZjC6K&
zncrS>oq)3dKyvA`y=DrG)HMuk%8Rzf0VDtN-@nznr@|D<(QG<kgfTcy!gOek#8H$9
z<!phGZMK+ej*+HBn4Wx2^C>ezaTFzNhrfg0{y{!Q1i1PX2aH@RjgcQh8YAS}bC;bB
zjQ%jC(-`s7GKaI*ooVnL!Pzv-D0gW;`|>NRh2qY@`^(43o)JDqZj(8xufP4yWlMYY
zzgKA!risi27};&d9n3-V_DLxf@DaZ+u8$;$14h31YQDMa!H1K|U=Lrdz1#o%yI<Jn
z+>iX`yYJ1da!?*c0Pq(`xDUR9PiT8#j?V`h{CIc2|6V`RT<q+A4qR`z<8HHX@nZWs
z07fPXat|m@GmUrEwfGp(_b7_czQTM}nBc46yVcJ(5*j1nb^tsB-fSz&99WZZf&+fK
zZkgM>N>)0;+H`)J@p{i1bQ&Y20xK{{Ss`#3jgc<BkJEwO<F(JlcIQ6=&27!vdsdrO
zdN#)T`7{6C7E6d4fe{5pUU=ov{>v`;C)<uWTWl58o~**Y(+Ae4vF?0*z4cO6i2e3Q
z-Jg8iv1Xesw@8V)`+*0|2m1JCLL2HJa=}?=rX+n@hmsd0E#Om+(Dm|!P1t|^hhL{S
z^7~ef6tBJc)`2W_b+zVKKfNSH4fQ=q4c+@t@@8oulWs1OWB2uD&GLh+4dUNkeNBqw
z&}?a?Tzu|1Non~(+v&q_(82r7Gkfp3M@rF8KKq<`N&Zu^p<T53Jo>Oy$&x$oyWf2D
z$)}M;&5?<+^UpXvCGMT~-#7QmK5!ImqRsp{r=F4$Rp@c*lzMoT4(gD<UQ#L^<^WV4
zhfDHfn44ZKJ~0AO>nAGA4OqIk;(BnSRG4dVJPR|9qgl#u+II#Rfw|!J3;Fc3D?)WU
zpYr^?sDMMxiJ>3o-}~P+KYXm*^X|jHMxRvN0Tn6)9(0S5UMWU;kC^B%_RyRtmsUrQ
zfDS+rs9$l8N-Jb$Nq~qmMc4yEA%X!$e&*9L!}vWwx}{D)5o@-}CJbWnaRL~DGwh(a
zKr5g-;RGCDrm{+)2O1x0Gd0?_0^%fys;C!W1MQPCfgrs-LEFlDyQJD*Vp^KpUGj{s
zJzd$Q&#2I0OjKT3leYBz#PmJmq*Nax&%gYNdF1iOWRgr~!~}|z$TtkcxI+^sG@5_<
z<<C=W=)W&hTDRPDkEs@z;!~ZRb6<ARk5ilxe=&c)x%rN}{G^fhI8^`TPk$owZpkKH
zpMCKq^V~}>n<EcB*zC5`PL|2}pI2NZ2i3_=(fvYCKK?iZ6W%A7qgUN<llehfDEZJ1
z+iq=+KkBHIBJPuUv3C=`9@6bM{EV#U^W<23&ITL!!2)l(%)kHUw|aJZlEU-(ar%iT
zm@N{($A4-3R)!p96L;DPf&Z3pJ?)Zc{_iWUH0|vjNl~YtbfUpDP?-Mvs;kY4glS_x
zZk7Oqi_SUQKCgcGP-l&Q{KaKfQ^?P=P+CJ*-*mHmfB511?76!+XutjZAbWr7z4w~W
zKK(o;{LljrFhAOD*Oc%(?!C`^oUpFozvIzdnml>3wCt)3el>i;K|SZTJyag?K4He~
zKQF&>fJwjOcH5Zajy%#&V(%Y$;z{!t`Gg!Pnzz?p=74=u{l26-v&NlG+vKqNKYsI@
z#QQwx0pFE>`yc<9(w%R<`_5c<%WX;FJU3G&Pc?sk*`-P0{FwjEH|7Qb4p9X2{*nvM
zw?ck?JpPIP`^u{a#@vtd?+=bSxBubBo6YwLpqzbeRZZOhFhc$d#gAA&f*<kwYpyk&
zoib+_LGuK{9I$UHs04omK8a5NBfq-z62rUECp77<xb}KK$lmiEyzJr&%~Y*VpH{9~
zZT?HoZwl%AiAEU0k!ay{b_|-yLl|O{IdkL_XdkoAj2ZUZ>*v4uuDi__0{gNd|2fJg
z-$<sl`z7%y`XzANukpv%NSOc;wl0O(vyNn6igk_L2y}IK$`oDlJAdSAFV-{`Mu2a`
z(q$_P+7$Tz^FI!>X@7s!e+-mevbw*sfA8&fu>o^;46XyQIoy2l*`TdYvX6COWRuO}
zrFi1HpQ*-3Tz<kip#1u$7wcn?d{9BE0h_QsIcaSi9{&8o3sOF!VNDP|;5hsatIjW6
z{DZa6=?>X`9oX``_?$D%gwQ78%dfvSci#VypVZzLO=%4d$#)ADnp>pe95t?S>g`b`
z(*{e-@6N54(NH_gN1{EyUKev0Y6PO@Fm6pDRo5r!!b+Kgs4hu<m(iZ6vI^DCwXvMy
zc&?4Jo_U|UdgRm+X1Pqwj5R8R1nyYt6B;8e9qp#0M~=@0d|1-{tE#@rY8asL(YmS?
zj<~7NJ?67InmeRLP$pk60XR}WvDRvEbW3Uy?TiMQqiSyLGF5;q0xHld!6a9OfC)$#
zg_Shz?iS!8lT`p1Rbn!9Al<FF9%+ko2(X}BV2t*bPE(`yw62t+Q`JjZy#i)fC#BNx
zs8xMvnzRcv!AulXrgUhIKvnfMrc>orH`bUg0V(*&sIBa?j>UV$*l?UW%7rk^1XG%J
zbM>WS+nZP0G~?w@WlcSZk!>!0aeKV}ge3j`VtSr-{Lv}h!TI`Q&pc^Xt_T1IfRQpC
z?E7@UhbB*&X#VY=|2{%v<U{kvKR;s1mEPaK<>O<-HSUySk2afbG{^n4*TqYgni~YH
z{PetYELii^*WZ{sWy;I-GbwT6$F4hWZw{WfZ%PzC5Z?aV`zhh!0Dm0AaU-HJ(%o$}
z?BneChuf6N@1ZnCDrFk&XXl?MUu>yB8>V~lEg7b4HAXl~{^FtwQnV`K%BMco)z5x>
zUJ78wJTjkQ+GYFg%pv>F3kREfpLoK2^5G{b;r_Vzk-o+VzOfqRZ=hD%496aRh_sj1
zmru(T=DOSNjAXjc6eIcc6OK*lC+4)!rf@zaCd&WOrE=cQTJtGW6UOP*s9)Yt_qk|{
z#A)R)CRg8dYf_c8^PhfxsrCKiM=V{w!u&yCfKTk<H{NI?bM9%Wf_mvRMyLaG(5V_D
zE#^vrp<%5ezfSJsG)C|hg3k~rtT95^f&1)jcHenNZU>0@#s3kYOcbFp6$Wk5enJBS
zV9BT9)EHsRcvkimkXR)U?~eN(FkgN1tzW&phYv)X=We=5Qk?&MBt@k8F&;FfS%WxR
z?~;`P4UFk)ZIL#sG${pAkHtM}n&BQ7QESScJC6S_E+(Pa$byUjypN)fk%A0mY9S+O
zN-MSY<d7&QYbzvabk&WwBppoLOJSQWwlF6gbwqL@`gODHtHPZ3X<unp?7e%kqz)T|
zKiqs<idii3vFSNq(p8vi@afT~pV49G13$>#7lDz%Awkm1ZoK%|hv;}Yg4;4A^@ViZ
zA8wI-)s=-sKV8*sz4!-f7dPD3BI9<rM|k+lkD9}DYl^tIO>I<J!{387w}$aDo*dE!
zW#E(NV{Ns)>W{*39av!0_lofs_dNP-80tP3e=|dsxu(TCQ(7H;a$YXKBKV*H41oYd
z>L%7%V8Q|o(j);K0hFk$uC(AqwM=H=W5j9B08q4vNnrhl2w>nh)FP8i4U+<}3t$9-
zc}==%0TS(E7{QE5DB=VO^cRQcj%Zg~OCXgFC^BJ6y;;4oO#q6RwTUv@1fl?Z&@Qx5
zFb{?lH{@81fH62<?~@&P%ze?Ha%sKbNFEbj0a#OFYUMz@RqK)Vy0$Kv|Ftb^y!@%1
zG;N@%<zndcv<2hWCqSrF%-ntV+|8VFZ0eqTC}3pj^mWYt`t{FJG)8c~e&?V5B<<hA
zIbU}jfceJx!ObJ#%*$O5Jghz(Z2_#I9Cedt(u4`pOgJ-zng0F4MOHw_H&;V?pUUd0
z&B43xVV->c#Ym?2b}}(@)FB7?S?ql-&ezF!+VRJjjW^uTmf`-uT*U-XHqPQ*-ngp|
z0EB&vxIFZW{)OqmM;|vIL};lG1&l;FIX(K2gJq^C*_r`NWSGeg02)E%zBKUb-(=2)
zcJA%F!;84{{s+yM0?EP@=iK^jnSguriKmj119bf4+_Nm;krbgHQH<Y1_utR#y#2N|
zX@{iiuekB16fGK8&LPr5z<-IORjcXaEdy(+R{-Qi=bU3^t~0|Wu~(eQUn{^Rj<HT6
z0Hy;4c)t3^TX~x^$^{s~d3cnj5Po7X3FQ;>cA0>!u+{_5;BT+FUS<#529mQ+KSJwc
zcbU}-Q#xRTwCA6Gs#!NQmYDEn%)^>a;j-h#;O++>HlKer-zIMn(+l%>?%l-RWA6GJ
zSW|ca7(M|aM)?6H9cJOe8Xcd0@udYG!(eX!jNn(P&M)5luymQZRwf3+gup4nqnN<*
ziM4n6^*8!K7WlsS>@x=dtH+;x-n=CrCsBj|D}t<)?+CMI%!%`JJDgg^uFg!a`Rw0W
zgS`^Q=n>DPPv6gyjy8d_b=tEQfsstde{D=R9(|0A#yqXbXEt~rMPOu@R#hQ@ktoS4
zn)@eh4ByiRF;l{JDo)bFo6=kEzSn%aV8K8!jwyMD)V0GBUuc0K@fD^Cjg9tW8@9<B
zC)IVp-g}unc1hkaHR<rn1_3}NOtgb>2ooR*V9IdNA^U4HEGdjIrE`8BPV0WhZPwz(
zhkS9)*NJ=Vndi)3-}@W41H@*PjaqmPGC4pF6q3V@TQ6%WFp|&4EQjG->#{^K-_xxr
z;^G*wQDv=l1Da=t{q<fP&pZcn4bMK7eT?|)cukoG2aMojq`R%tn#vNd*;L8z2U->u
zd|1(9fd@1$Is`mG7Q?{5hyaCZX>6d?QQJ^ufewI#N`VpWGKYoc2j;ghz11)|nStXR
zrsG=#e<LuJbTO+|v`Ran-1;uD=DP??u$++0#K#Cu)^QNuuQt#|9J3SFr9PAkwCNT+
zaDEp606+jqL_t&lP%0A_I9><vv7|-=&a_Kwq^hsWwDk%!5paT85I__(Pb$RR0nEV}
zd}n9iV7^P>3{KhGqy^LRL$Ux*Q1ICOTQ_wb>tDs9J6<#U1wNK&d=8qompSgx1HG8w
z_SHB3Y94y@PiCdGbfErzY0v0COvrJaIdg{jfBx}T1B|mEmocAx`K7^MLtS}h&eyTg
zkFEYp2!=0ARI~(qVn51R1QS!gyY5DTX{q~tKRxG_qh)$4ESU}C@p1OSF<K)qAdf%d
zaDz}4XBM<o!t^8gBG^ynaKhyKJ;~-9cns|fnLVbR_dfiH`BY<FC_3%<V+Fu$V2iu<
z);q*pP3_yp8*XS$J?`kF^B%rg@ZkuPIZUS>cZ|&KY#0vKnJl_|5@NRJ+|y1LIGM~$
zMhoc6`Cq4$6=nSZ#_$bypsg`zR?lUGDSWZO4ggu7TEu|A<*s`Sf_)Hc1ZjD+AEM~N
z$DS}BeEf-xazX-UoOqno7IEp*oyF$;Xpf|_!<r!MNq63HN4rkL=WFWylK_S&-xxe2
z_(a)Nj<Nyo-gx^R*PNWL`1QfRPne#4=@s+pn{Oor<F9X>iIc5%M%Y0&ZM|Ju2VwJK
z6cFVbl<*g2Ui!t?Ugy?QV|}gp@mXg^0#%EC_(4w1Q~3xkZ`4`@e8Fc+RGVBru5tc6
z0iK?f$w3HUq(0;~h&6`aBg~Hb6bE$RCyD3K2N-zP?-0m0KT2Ce@9$q-a#0Gv3uyNH
z>ym*E>W>3P;E(+0m4EP?C=+1onzO~*am0HXzD|^O4L~!13u_PR(zoR*fZLE?S65ct
z)jdaX@na!;b!)RWH&v!m0ux>ISMQBd`Qfe9{v+-TITqSoY{Te|fb%gj8a%Dh?Rl|w
zHv;VYiol2)jRIfu1xC=QxI_-nCkl876A}SP24q7~%zif2s9v@vR6o09s$u(WwwBFQ
zNqc!4QXHf6v1L;eMLVPRyx+b_d61I2<Gu&Y(&ftsvaCOQmO15^qovtcIS|IC2kEgW
zNhaEW^wYMX0|%M0w|V{TcN$ci3rTE}&OGS^LtDcXcDxsk5C6t&*csYzaX<?b5-j)q
z`O!$}syJYTv7*0h_P_n^d+Q(-X(`4ie6V4ohJ@IM3H!R?`U~+p9TU3{<)<qnn?q~5
za)v8zJZv6^tM8eX8#RaN)>IJ~$#i7%PZu>e`IkCaLh<<+9E-7bmkF?7Z8#UD??!Cu
z07g2bJ%IEh+7<u>fD{$FV!EnE=B)apMZtH!t4#od_6ks~!rNt1t6b8Pn4ofeO!k)m
z3teJ_L6wT=5`(f!Ajw35EgjnP;dmVa$f&Q87Kq}i<)Z{}!?7!a6h`1gk3fztYX%E5
zS>^Iaf`$j6NOgmN8aa!vnIKbMGVw*Zw3$KxZ17J4kq5#NJEX3dvd%VXgOphlU0sx`
z^!RM3l=&|lyR%>GR6QN4yAHEriL#bONTJ(8$LOEh$<r;!#GV(QExiCr0*@SBaK_1I
zr)^TT0KSz5_wTR0#<a8qrpbD{IxHwczEYXVI%xlW%!$XO`qF3;Q|@;+-z+dn0F}TI
zM{%aKkO;p{z5$wB0szLPkGpanz`Sb|ee~Jq1}ET|P)SXdx$Mj{QY6*!tMl4hZ<%-G
z4+MrGAP{Dx&{o-PryZ@f37QLkkpBbD3O>OYgaOMwJ4`S`*_X4|&gL7)8R*dc=Y@-t
zw)QIX)C(_~CCips#wccramvk^Dq8}C04#vP2|Xo-`s)!4b^3Sd`R67{sK=dMkY-2X
z9FuEd8cF~U#y(EZeZsU7XG&j7hB*Nk@fpd?4UFVCW;g9R_tcXN=B&bmV{AY%fHgp(
zt>(_vd`X*3{SMQc@4RQ8c<%Y6;Dh9EVYi)j7zo4pG+G4!H9q0M`YfFV-Po`u&~TZr
z`E$gcv{&z>v*g!hi;XvS!S>3#7>xr2>X7yatSpyy+j?SVvqm>aoG~AkKa#(F_;FI+
zY<@7VFzR90hsg&<?7Kx{G(~fL*n#_-S?eU9OKC5>g(seSAzY089j9f*xV))+Sd&1D
zVux+Evd=5w%%e}8I=`an6eR>nx_$Q8UH^9Tv)Frh70<l*vP_??w9h%u03d`90HA4d
zp3HaoybCgOW+v~~-+gD=#W0^abCy}Ya;3C%BuJ2;8Plfe+<$zMa3SqSbLM>kT#OCG
z`;m9Y%o)=YIfDyYHjh5_j5Y6=6rmrRZ?dsL(<@AXP}fKxba}{+PaH5pdocUPyTlQK
zsXrB;$XVc^ozKhv>T7SkZQo%o9c2O_{%F@-T#)4rz$0*d8;BU<rme5r)Rqo@Glob$
zD$zWbIA8DU)ZVAM#&owdn<{{M304$=kx`kFaZzbH{=>NFK(Uc|j{y6=A~2HoFh&=U
z3ortcd#kyd*#pf6+O%)B@kWt_v#CU5V&&>p`sjC=_dgnXQ^lt7*a#bwNG*IMXLfA@
zMW#=gVz%6Tv!u-o_HmJ@VRH~id~~tziA`Yc9bs#eLlAXLnLNp=?N$jaaKy&ol{enh
z#_^4$k2BIj^qnyXGyGj%uFJFur~CzSI2NUj?Waw5Unz->t>xqq2XpIcBfjP4o2wu0
z=s%c~jS1uMgCx+9w)o(aPsB6`7z=}`jNSNh=-PCnjZ)Nt{dC{R=^wV1Ki1~iH+K~W
zj6DC!YdVPRBxXp3-Pqw6j!k=5Q-nG=;NAboW0BHIejD<+AI*<QHiy<+`xs?HAlI_S
zdcLKt?PiM1PdLZCetTETHuD5A<O)qua~QX#aM0W@vj=gT+NiP$-QKmSyyAGSO|zYS
zUkhnUEBo5lz8nr1X=;`EEA3O9_6I&aAow=c5xRhjj;3}~A#ekNxnA8UA0Pq@aBSWn
zfCH!NXk2t8_AY=4<gtY)PnCd-N|ob41^@^^6Mz?h4FC;xN5r?SmKL9UgFx($Aix^H
z80y5w34jdk02mS9-&WHOP@;BqS}hUz2}x*}&<B7KKoAIxk?zWBnceEQW-ulQ=;>Bl
z*q#B3;Af;uK2Xp=sgYSO%Eu4N_g{Qv^N*K51E#bDW|V4-@e8xpuG;Hq?A8$zZqJ=}
z7+`*d`}FSn@0-uQm~Y4b#aI7g2YBDT_O$!won$hI{bCd?`eBLoiC>6e_`RGHFS1GE
zfDt0u_oDp(6L*#N^yvtOHcZN>eRPz0qqwpF(C4nVzQODdjOsAK7;O<)gEmo}&QCrg
zcK)~Dn%nPvU?2;Q=rI!v0~zLYn1M<{*3%bX%{Sl4S#ztHX<vQ&y-WhP4U`dggTwBh
zoO7m3mQ}|kvEjXXMt*lqvgx688322>-EvFQr**xloV)v){G35PlHe0QAz;j67Ru3K
z2kb8i`fc1#D}07uNk=e(0Tqyv-)MvNU2xL1RsF^XNSgqN<*Fm>y8xi($b%2mnRWA|
zu>@$~Tn3m%J#)nv_CX6kp(n+B2h>VOn0Y!5K!t}o4Z-++NWx-h?S#jlGbTcJFxFpq
z=N(hiUt;GT4FUWvID%3C-UlBUB<^{MOw(ByfxHQgfx~Q=$8Ws-u6g>!my*t^Zok@~
zwRO@FM??zLk+(%s`Dn(}$@h!@JOUu0iRA-Hu1)E#oU;KSSno5Xy~Fu6YCQu00Ln0e
z2qdyjd5-*ZI`zO)c;lUS%?mnXvPPqbc3*JDsVOF;qSC&pu_{p;oyjsk4H{e@fBw0B
z=U~RkC-?*x$eb2Jh3TlPtK0lej>~<)NXFvqlTVBkL=2aOKU4gg`UHS?<Ut1{ts&OW
zXL=tjlEn<x))t#>G5}1vyp;36BabJwg*Y@=rpt8rR^q9IHFI2E=H|msKeeFJ7xTZi
z!OTsRDJ>#btzND5{F&9t;vEGba{7tK31k}td0B6i#Xr^+^9?w|v+kZH)&+vA=}f)Z
z#*EWHJeL3?y?U<uSxb5s)oJY)O=*qO2_2zkx3;yIDrp+|!XiUeorxgXP-Wy3zi9bt
zGjmEHERjz+#k|1?!22iyBg3>Basfv0zj5JNXJpkPq@k|3;iio8@WI8&*zQvOjdITC
z7sAIC4Uwl_cqzrNh@Z~gvnfO6eQ!y7IEf|~7rs*Va!FFX`08sG6yUc{TWh0z?BR#{
zL22(f8~_63N|>TTYRcDS8LBL_Y_7gZc2D);4HI@_cb7y^ru4w0kC_j(*>MyHjBudj
zfN|IX`^9dw+@^?wg1>2WWf$`54h7kKL<&`Zk#Sq7kqpYgF=Yd`SPNTh+R~!tFx{FW
zt_X~z9n!oa-3E_qnZ_Dh%tb1WK-35nHAbRFW%$CvrnIQ35+GY9s9?C&%CARrbEh@O
zgnyB))?i-(bIOwTXPD4}kpknZTvs$d&=A2<J7%)l05b%Z_@=Y~KQO;lEdZuR07R8Q
z3w(P(UD9gkkhv>-g9N6uWQt2=W8Mp(1x=AowF47fIE?pAY4M@K1Q=8zh8Bdzhy!xa
z3W1y{t@_4l)6&#o8f2b}HUW&#2ed^XwvYfn<<@b0iCHS9(0HN=vu0ZjAhc}&l3?`z
z`>%f~W?TSpW!jg&`{#eNjNAY7zyEbe=AtX}<=5Ykl<{l!XF6cSrN2w&k{lDyg+*O2
zJX4O8r%y|YLRjke*WYNebx1y0niiM~LQu*FKavVh{D7H>MmA<=ecuP0Z8XQ69ATC>
zstgx)(_McY0F7K&+?Dgs5n`<RAV4@R4CvoocSDMU=5RXqo3jaK$)cFvE)7?lWdo2n
znmK!2bN)#u+I#>gS6qKn(pc=e;|_9czmH2k<P~rM>GCjfro`+p=8<8#rz6fLPV(Q^
zx`@jI6L&B9uSCi{Zq6|chV=wYT%B_4QCc5!B8v+HQZTDIr$vEn;Q~10A;1SP;=o^*
z#_d--^gfzWQ)hPFZD+IFw%f*@&s@G>ES6bw(3q*tmRMuI#ynFD=$T1*^o8dG!0Szc
zB2gyi!sSexIz?6=jxrMyT1(-0`o($tL7i#m2*?9~iX*gs9((%PlxQ@~4%D+AH|J#F
zY<u~2GRLQ@PiS&rdOO`(q>X6)gn>TFAwL>LnEVTaM}9rrv-bY_-uv<q^Ja<$7|-k9
z{p7+(b7J9g`0evB(g7nbuWufVcLV?>e1oCd%$y@ciWU{1%$iC(mmPXWd{bIAdO!7R
zUlDhPTvNxQ*ssh-z`f5$gQt~wD~l<Ia|DXO$S{s>F2D$zv>)v@sOp$W^EhZ=<1-sV
zGiC0k8(S#@Haay@HRN!MYBJJM3nVpznkt)DHWAqnk{(DkA(e$4TIxa*0|}2deXJHP
zULsZUZ|%o9PMsW;*li=62gbEyu_WJ+n97xq=3oP!?m!bR8#~Epx<noDkdWFiMhYuj
z4!@C9^39aR0VAmO(og0LX_QSi+|VAJF(c5^+he8MK9@uv>f7NBd{n*phK=S&B#Nn?
zX@6GX&{Uk2#`5%1Ntu>cOqbgTM9pEkHAP(9rZ%dq+*-Bv=4uuo(vWy2isQNVu7gZ^
zK9+rqWYYd{nz3%$%>g4V?R^3%sx9D9QLFD)VtZC04G)Oz8d@Cq?WmBsDKshY?*Y|M
zs<Qw{uM&xmXMquzkL~iwQ6)_hr>()a5un5Y5IxcqQ8H_~3Qd#d&TdmFFokeH88k#_
zNE^V6exuRi_Kg4;P0L&4C!<`p-K$LVDrvI_xB*B3gejAWuU_p9Av8|_ODfgBa+Ob=
z{gSk9myZ#^6~F>aXo<ID!5P;^uB%lCnlV)xhaM$ZI{zD6=6Lyo#t0x5S~gYcf42aV
z;}1K~luLuej;TN+m~D)!olgcw)GtPRpkX6lByyyF&;1YBC2`!5hYk9FP}-6b%@O0N
zv7sCQq;d&DF#j+JopWOL?|bjItICYlGNDd<++ogWC_?LdBWWf0gx1MWngV`$|DGPN
zBTbY|=4>Fz{Mj<sQ!fUzw1gDiDZvhe#1<@CEdMEACGDqO88{!_(bJ;n);_R8(xGug
z{9mP+pA7>Mv$;ENyN#JTWwMzdh9zUo*#`|9`i^ut8anRzh^yO$1B}hva}TS<;KHJ=
zFUiM)4+6nd-tR|yTGMk#nZN$mcarMlj4@Xl8JQ9i^XdFrqP31xI$HiSG|n|O)t0f1
zZxgg7n3wtAe5W(zw^^M{m|L`aF!4Nl=1fzi@#xb#rb*g7-$<M1oeyL=Az>;F%@^7b
zMS!N57@|FnX!E8Ub5>nn8X}XenIy)Ixk38|Gtg)>pwUksqR8!60Y>V}f@cQvMx22+
zTz|HiGI5d}3mDdbVwi7Qykv<57CB!7;yI$PnVz;O51=p|c;L!)SI&f-S2??`6jOYq
z{Ip<oA?_UR@`bO=?KWCFTS=1%jT+WC)(aM?U!Mq|dQTc`rD94~>DfGRpS|r_D_jP@
zF{2h|>X~)4{nlGaJ7to<E9%fQu6DDQ7yq!-jvpEmQO^;YVVJhV`U7)Op>GfUZj$!I
zLRpvie$fv$f7IEXcJWSJZ}u#Ez6KzvudkCpK&724Kr-HaSbN|-6SpoHll}JEGh=Yt
z=V6c+pE-4P$v&EBn<FeuP`@Mc@SL#j&>#Z10yI;w`rczTL(qo$<clxuvlumQuB<iT
z%5zzwcLi1k`lKn<rT0aROlkEOO=+zO(>rW^aPRXlFtR2)8y~j89HxSY%Yun(F1(MT
z#)un@0$+0hMhe`7tcv5q+R=tH&17ecSrrnKn6lUO4{=P_a|(_FMxvbbVFs*Jn|W-v
zj`SGEeMEBVbe5UIjN7%z>{XUSHLA&-7D7}b!gENPxHZ+MBO_|A4Yh%3gK&9KWrfoe
zf2SXTkp@QE`sAQl%)BCvzY!=TFp};ak;};B0&;Z9R9_X$<1tDO7-{R)UMj&%b7rb&
zGQbGH1mHovv^xL_%H^K{rg<k$%msV^AXqN4(!=}-fDZv9GG|pKu%ff0M@&5V@DSjE
z$*dkRAjt!>4#pw`@Di9u;<rp$DZE^D_Q~ltgcb+j$!Y;I4K<afS!MT0vjnh&HsY_O
z$};(lX<pqf-zPB8f_Feid%)xebirH}<@Si7QC8Pr+eqI4RM0G`v|x<D6qSqCNT<SK
z7Se8P`Hy#Cq;qwvDVIi2cS7^0Lif%&eYwWf-S>%k-`FU?skK*HIdWuQC9`Jow;+_n
z{$@^?n=$}zCJ)j$)m8~;k_2u2Fzok(O5*?a7|pGWc4k6DF(f!!K)LLFhuSj?6+cw_
z^J&{qWe-Puq2I%WKKHT5#|&UY0)QBfG03*fZok^9cBL~nn{Bzn(YJic46hCU*$DT|
zjY*ifm}~vv^0O_+&)?i6=j)9V-Dfc4qWm&4&NRviX)~0|yRi^W#|lB?FqmM@q-`xa
zEB45DR)tJ4_Nop2;%xzpU`ne-nq6xHVSc)D4DV^j;v96eYu}~yQmA*=NcS<5IbXPJ
zm05Rc(HcbD-0*EH0wZywlV^As7?~-_l)7Qu(B;`&o4wE*nnJW`Z7d@W7<umHSFK8S
zCNIB{7`M?}FaE}SxcVAC3=CI2qggJjk_F!_GV89lPNaX4T$qfR_9to%<JJ_wNVini
ziy9-Dj&Anp{Pi9+H`$k1%(149Kv$1|_%%rSZqbS+Gjnp|n4K<wk*@Ae(@|1x%~au3
zy;lGROpit}%DW{E4A6m(1jqp&aDXZS6cmp$Zm~&{;sH5G3kZ^6n-6G<RFp}pKzZ5)
zJb<Y{)B$jTmMR3JvRwJ`K>{UsF$2t)z;sv8&Pst8)MMMDe3-W(o{})j<p2@1muYFc
zfSQUPG5f{9_xnXXw6$D~0>p8@X)or%=+ELW=iBh{@~3X<I#v^;QfA0-SYIwp8cb?+
z0Fczmsl1*wr};ylDg}Pw>>jGG=r>Jmnj=X@_er9;L=K+&0Y@N>Gi#(*+JL13p<$Ml
zh+$MyuKBL6H4_A$jaMc>(81jQ4)u*2Dn9p`i#gp0_<If)mhS2cZ=_2%QgQ58(H22N
z!fD$^)y+Pdci5VurJnBDh}*+Qjmt=fH8-s^4ULmVx?0oG-(#xv?Oqco(=*|MMa$&x
zW$LJOnKjrX$4oK+Djy@#3KTF>A?qb25-J%fVC4Jn7n@nLXAJC>MtUGtNLY5Vtf7nq
zjTN3$zI`nuFw)l7Ey2Hd&D(t2TFgH@Bfz(<2#gHR*qSgfvdLzf4O2~pE4Hb%S(0R!
zF(^LFg&WoEONs+Vumg=f=&?ehxd9Gu%oi+JVrI>nUa$=)QjJTNHkoy&t=Y)BV}>R8
zUL(4+!gEOFachbLbh#vu`n1V)CJX$=MwR8qr@J5POwn|0&m?Y{0Hvuq6bwZRmr9Z(
zvEK6;&!OsZ@kNahHxlFH>TBh#X~0OPhyBr-sisajfEGZ}0s~wO^D=2n^vI-Dg+LLg
ztJT?B?>A*Q9hLc>3NgLA+oeUK-#GJb5SY>;EsgpKI4KvHA)g`m3+d?SwwflGaq8AR
zPRs#JYS6|Ic!4$sU<Ka@G&BHVAT&(S;Ha0mtmYL7Q!^^JQYOAA#}OJX6}8ned!)Pq
zKsv?H$BY(KqVn8*U{*^KosBZ>s`~lY-P~+VY0);DUhTjXSEV#zAbg}?E(|kNXq42{
zNjg+X$D5B42aMEeEZbITuIYq`v9mr{=%41i46`umFThAwi+sWeXlkgHd9$`YX^3bX
z1Z0(H-1-EhmC9*7dHXustqCnOM@l6@N_=g#bT(wVcQiCcY{$mHpP{~SL&cAQF06G;
z$kvR;a66`DzuE%8E6potVj9zIg`BLR_N?hNAPe!WOZRNV?codA|6I#z5-`#@3cyIN
z?H!BxGCfa>X4_eV?R@(rC{U)ePp?cF)=g;8KB&i<*cu6ts##!Yy@V8;5Xoqc$9SoG
zG_{Nvqe8yB5H$*R&EsYCi|u3t__h^+kzra3LjfZ`lVPYPjFhy&P%njSq<gYSJ5>KO
zjnAh}@;b^!s%I!we^+KU*P(zBd}QRi0nc=-(yh~>)=M^h&Ngj$W6oy3Qp|%ws*a}X
zd$#Q@rcWJ#@Enq&rLEaii=h(+XPFB$Mv(fcNtj>q=ZCVY8X7Y;(c!noFUJpGo9@}f
z{jjpd%$yXMDvO$%e8wZ{eZ%2?oyoEvZCS4Qls+<f3&Bvxw52eB#J@}{h3c1;&l=99
zUuZD&%5Ha$nC|&%!sgR<e+<^-`(^<nNIN#Q22RQGr-6g>I)NRnt6HricKlM5i+PR>
zeg}-e1P3(1N5>?kh24&J2Ox*j!f@k;ZxR^k6DHPJ%?q12k*-4d+XQw5{zUNc5tu^4
ze@MgRdO5)jfDmlncUY(Dg8&Uq*t?XsUVsZ4C+%tvKnG0ya{P@*LxT3<{{k?lMrOaz
z7U^#5FyjB4)}As`iB=5#6Ci>aE#-F2w%`cua^_B|b^rb6gZ9@45{}`U3G2?bU=pUq
zdIWq3iNb!FZE=M(MgW?upB2p~z)-iwyI11|fC4bGdUeNO?>x;Q3$O^R(%?$NC|Ema
zj?~r49GQGzREu#F{?<mls|O1odw03MKZg3oh2wL1PmaCO8v89uH_rJ;;l{1&xbRFy
ziZc)zBRsDYDn>>+JCk-5I?a;iRx_=kc3`g-em4C2hT~jVD2a45MtY?wm+$%6?{6me
z>DF{6`AL(_I)y18XL2qMEX=fxOPbrvbghN4OsiU&>{+ox>$bD2-_%xWzaXJs#KUmP
zp3dxDRKKzrdzU|#JTC9rdW{>iu=&kQ*J`%o*xKGAKXrxCqAheA3&m?Jbi1-0<FTHe
zZ`&FJjEwccE^3U7^%=^oeH_pW&Dfi}=_cmHqmD>v&|m)gwt4!6mr}wr4Jfo4`fzN3
zGwn-wV?G@7kZD_r>2e!^@ElS~+?s;<(bLgBDn3RsJ(P`S2@&26P12B>JY%8-o}%U^
zw^@$Pg=A2x7`y9CX-xb6%yjJS+8N`1BBDoPWJB{=qb@I3B+`2%eOD=<GGD2vxWiDc
zBjehcP5!t%`F>a8%FU*2qy>m-mk$XPLjVxa9Kc>S00ZCxPRa2lQmOk6X%e7Gf=Q-|
zYWbs3I@06N97F^YRKN;lVwmI0!<x~ONvc`_5}nd4!A}PuL*PF|K1e!)F!GYoK@!y0
zE6tA@p%?AV9cqW1c}oL@1k~A|F!fc9c_@JzmC{l`5duDaN1IH`C@=m(FcU>4G*xf_
zPhY62lY5!c>aLImO9F@?FQ&3E4`rRi)6Rr;1!lF-G68&9{>?WkVf+xl2y+Rj)rT|s
zprf_}%pu{7L9aA!0K0kxjFf1e@T*c=(rtl}68VSe6L`j0S}hZ;lRlZ$VlJ6@YhKGv
zfc(%%E4-wx-ZTi<3<D!0C5Tn1*~#@88|oX^q62b$!{}-Zr<43I)cB2NIUU;9RfqhH
zrPB&26my6^hXJ)*))D*FPCZ_IrDc}c9xijV#<o!FhVh1)q-nU0Oil78fzmxYaeMe6
z=ji_BnXVEbjE^JE?djIq=(e;lwJ&Rvws`H(+Tz*NR>)^Cn|iZJ3oz0r26um(ggxY2
zq(nYP%5|;OH#QewB%8Wj+CuMZUAe{U@QuJ&21W|i>#_2UDpY?8T@HSzeptMupvA)f
z|K#mAZJGb~g@?=@6KC1`k{4ciwEwb8{>lEBv&B~S8n+Sg52OCQ&6N1_-FM&nKgaz2
zjM+1jq}h8U!G-<iOvsIa+$)+gYkfJBT!U#~E>)mnum8B`KukFfg{53#+og6`TOI!q
zWxn|594ET@%e7s(=5ymQ*u&t?^(P$eev=Q%2FQr3i*o3HaO8B`MM9&=l`~YIIm8})
z#9?OV?Y2uPXu+aI=9*h?vy!^uK2UD24!yp{1mht&O51DZS5~U+Zwhwp;l5jIk@exo
z#j!@|AoK0UmU<`)^4E)FV7hd!tU_FA59Ie5TacACNWcbhocG;&Of=6jPk6@Yw_CGr
z{N0suDTBQBnMm}RabrBK{dI)}%osVjRe63NyY)QL0{2Y0v7=n#EK@=W;7$0?E@NvM
zQ@Z+d##}%_In*Z!VEuAq?&=P=pRsZ4kuvyATbUQy?)q;F*B?JW_cp)!%XqmyxHQxW
z!GQ9|!Id41TQCl`or%YRF{CYw4|eFeGEbBbmBPf4^bX~cMSE6B4~xMjvn!NCI{H8v
zm~AO5QT;G-R2J=^Wb$I3rA&<dlIm)k-j=5VnLqkPIgBywg(#EOQYY=9EH}P3LE?`a
zYqz!uCLd2q(5qlhnP=w7Pe)q!jwdbuQ3wA)j4h+Uytr}Z>e6se`)R8q^13?Q)x}X3
zCDUe@hRh@N(=J!OOY2`L$BhR}(rP4?MI3HRBw<yl13&p`tSwXbjISF{Oxe^}M&{rc
z+VzrX5NTccc3v0(H8i*no@t%~M~n(%;MyK$oVdQYxp8Ame)<A2pRB!Lg!S8wOi-75
zCe<vvj)Ks^HOUyeHZa$e6GgOxF1qK6`E}&x*!}j~NSG_1d-Czye>b@G#JXqf{PZpl
zb&xK2UW2v9oREj-BcQ=|0wKhak9d4T1pl#ygJ+XG)aTZgokxX{*Y=J-`q3^^NV0$w
zFmkm%sx|)*W8vnWvRQkM=mX>6zdvkyR37V<wB&(kBjwO8=D?O~`!)D(a^;4v)InWt
z&5_sj$4^UL{ye#L<)`7^&r6=552dzkZv0(2)XP7&CdeO*`QZD4{`>Ki&-!E@9TDcY
z-<G5In|)^VzJ&4bh`He#1K7d)0<!UmxdJF4xqHUw18smvP9Daebo|yuDS~IlUiBNl
z9b}=M^pC5n%NC?Ri6ZS!>Z1l%7vqCJ5wuKb8|4xY1?{5^!8#z0F#aQ@HHD$~I^m>&
z2ruKgP=M>Va}MwNyJW$lAn`c)J89aqK!ZlUMVOBOETXfCPHVsbI{>$EV2`$q)v^(|
zR8gw$qrQ!8wa%Pc`c9Wf(PUzGwdvCsmRD8Uwc+NIwZXmr{-iYKgSA1NogLlNG&k?e
zg<E5;cK@B`!gywa*=J1!YmWJJ?=J2gxpm5K_Z(9<?=PM!h%~f=y50N5M(7XWtXGKj
z=GskLA@(0-0*ENf)kQk|-9RO3Yn2?ax_1y^ZvBVvDI;8$eU`j7@Qk}}k^QEsMZr66
z@VigB5R7d%2egfQ{)-~|?cN{$d(6L&$`=lE>FEdm(N_NB-oKL1jf0B|m&I>ao*(DF
z+wQ&EAuT4fmRvcMOMmF8n_Jg6%3+*&H-?G){yc@_X$#*x)`@%ny0T~s;jCBtK2D4`
z@vKY8wab-D8wdlO?&NvYwX(j(u1&&eyL~??9sO}+$5fr;IdW@`IMTYh=nG}AO8s#9
zz|v(r7)S0^%|IGA2E0Fl)gGj6*BF-zK&8#}iF9r)aqsG-9rpXJt6QgBT^Qr$_9i|0
z9+j5|W*q(U7(Yxr*6N*|EBWi$mEr1jzq7fzHAWuJjotmyI+Oqb<Ig-)YY*erC21HV
zw})c1AonhdTGKXx{<t;me&=%a$1<2H-o?QgftI+nn`znkj2mrXeEjE+dfjuAX&YQR
z*Oqi+NZL+;`FzLI^_4Pco7*SxoVYbjSgQb$e#}^D?f3QTEUa&&KAHA`;Wrwn)WQ1h
z)qJxS{O8}*%^1*M%H~;9H3|uy8LrF;^F(^y+3vkXKOo9?@$NmqZ=SgzZ}4tlk5!>(
zlr)$ct(7nd&#-&b1#2>R=g`;STT5+}64c}R&)B%<*M*UWdI_i1ez^P1`i1=Wfs3;d
z5r3!)q7{q*&yj7F0%)VF$F*tjH|v^|=_%@&qAbdBeT*xCbndzJ(@-Y!?dI6Ejq&O1
z=#*knRb)QDjPQL_J97o6{dW8Ju1*)0?tJRfMqOS0=(yB*8EOqgl{1`SthJ$6PIfJD
zePkXvb1Yr9GP?qXNWI>Av(5aEKeVZSf7O2sq)q`wTtnP-Af;}i!ou<2f49I78uRzF
z)|(akAk$ySy)?M@#~>W<UU1=b{LYmqvqKA_T!*hp9R{Fq8{C3)?F?ES{3X2$59f{h
z9nQn=T*$REJWf$%xjy)DF8@$reth^ostnTj;o-U&a~^QUZPLVvX3mD|n@#3yWVV{S
zxs?PA*TL`kl7#%=<4;U;ON;sJi!UYdDVv#ILVpgt;DKAeJp+M>n{XVzAy=mR?ZVnS
zC7CK_hua|A#QGC1hw!K{m)^f-nvRV*hgvAyKL0oOuH7!&uaA5Gd2(^|oP6%%8&^iS
zz2V=kyr^rwX`}Lo*O*J=)*9^yj{|93xJ%D(|K5#JwjX1^Y+H%<r(1JakvLqr`jCuf
zJ%`s6hdWHklnWSPPWb2A#I;xXkbDj1JE+8eF1%{(t>3qy;@n#B)46p)Shy_Lt{{%_
z4$=g36Z|61&4U}G!I0!NNqOXQ>0SALz5YFALw=aM=L!YmG<Z*$j;O1xwN>`|>yqkq
z@&477>xXe4uE(`KoX-7r_50)M;-apuUU!X3<HDlSxG;BhX~S1H_JgAn976cbI5~3V
zxVdrR`CKUza&>Wy8y}Z8Zk+vgxwZyn4<3?Sc)GHPv!IHeH<;WdvaQZIV8+nB8ysW9
zwS_#iop92^)b@#kY54H%#|NX}hXnWj_;W}O4vS(?hzW1UN(o3~8A*zV*+DpA{xv8q
zc|KhjWmSoB;>LvEjwqu_{U(e<vm@$rKN$lU0_3VxIAa9kioCj5V(c8b@?ok1Fj(SV
zS_CkXD=MV{l4t{Ux-^tQsnqPopDXchZfFa22W@qXyAtb1g!U3n8peqcq<xNg$6=f?
zrb(1Xde#Bt(vv%=A4X%)C)XzWO&sM=CiyAP-BV6b+Tc78k7iPZ_M^m+2lCT#AGC(>
zAcnRt{BEI5J6`@wpEc7mYFy{s_%kOAGgrVL_FLV$+V!C;Yob()q;4^OR+lIIybRIl
zQkf=e=n^oZvBft6K53|f0p=NU;8lexfnnnY9!3Up&O8Jo&H4!1N0o^pf-rt7z^!AK
zzE5ixfXo%^!dX-Jt_kKaSReK|Q33~g0SXaM8)+Bi!`uQWa#V)Kk_gjd_+%k3nnn27
zDU)^*S~vK`K|80SoDPfe&ESe$U#PcNTr{5gN<DAH1}&nG^gl>%GiC87=v$D%_1j;&
zxt2?bu3ePNy(>4UZSb#agI|}+>rzpM-$VDC*&$zYoa;#tPFaLdzpKmTr~KsHgbb2P
z@24S*-%(d`y4vIVJ6L*lcl6bj>FRbR_`f-abP1e<6<H^SS}^)9WiXE6{=2fdLT)@=
zzg_vna_`2{9|!WeblkhLc;~|yb~XEDlFp}adBnN4`S;Z0$d%`S5z@Q%lZKRCT>?$?
zFmXO}>0SPzu700}O6%e?%|p5lwav~I?>yUfzxA%f!e3oo?iyEvAOHWecOGz(ROQ+~
zGd($NU|DuaO4uc50Y$P1q98#<k?S?0VnQ%*z3A0z!fOsFUel*sMG!EHiX?GC1q4Ad
z5@cC6hsiTNJ>B2)oSxHrrlvYqr<v#MZ>OrOPMz?te4cvW*Rc)mC%2FMyncCH&J<+t
z^1G?q#{cc+-83kIa5+nN=LXAgd0lyK8kX0s3r_tqD2y<<=PuU|KM&XQcUWS6j!lds
zzFHfE|0e8fzB?wc@0;)}7sGi<F^nl#rYp~t9jx2!6V}Wl$F;!maNDsQw=L&$)x&&l
zy}x~WJN`{4SBLxUwsF5Xb<6y6X6~<a+ue39uWN@}?%MB;H!k<*{kq(?ZX2%i>-Fnk
zIcaXwukGpO)0esJxQ;1ZPr9E$9>KOTKkIR4xZg$|WA6A$=Sa6`1Qu3pcusKR60G0#
ziRHQEFZbK%*Wu@J+l5Vo+xvOkI=>CM%JtXz$7+yOunyO5e>vw}y?#6VI+(}R<I0%6
zjia9~fxp7@v81i_rGMrA%X*o@`sTNnd4g^Bm-*{B&oQVqtE06T@r9AkUG@osPk7k`
zGea1$OUwAvXci*`Q%rJ$48?3yXx^Z@$=^&9xG1#aMiLy9EVNwZE1ke02OelnhowlN
zX&CVqzH-?&%_zRG&y~`b4qkBbCwH^Rm(xACq!JjcM#20tgD;itOM@}obX<MB`7pT)
z6<@6R;O>8+<BoNIS++K~JbW-uQpCfT7p^@VWJ$jiq`x|Gb^HBtdD(8)UcO9_6SrmC
z*%w+NSl1(1?gk57dD+?)#quyUKiK|4+Yvm^v$fBapDjPxFb}!#%Wi%f>AtREnU1-;
zA^MEDo&y+mia$Iq_|jzk+#^;_M`yQP=l+bqT^93m2$f%TamR4Kaxicb$9$@(#k3>1
zF9+C|=!Svy@mTDb3Vcb198`1r*OkqOd>)6~)G>5@=bEq^tlQR2U4Hl5pN8!_f4x8N
zv-aoxJYn;N*5S^Rurk7y@vAx}4m6S7<X9Yj=)3IKQe^FT1-H%C9-mWK|Li_Px%^tR
zoK$Q-bJgXyIafZnj>nQa=H0SJng;iayAB6mhFzbyuX)zdVuS&}`86Kh7j^X@CJ@dZ
zmdE7IJIB*a?aGip!S%u4A)91f#OpT`uQTo%&+^6*x1kqb^k&;%K^C?U{0Up`_Iuc|
z=kl{o*3WAVlYjnj9TUG0<hjOt!~Mf>@sYOH<=4IB>GE+s2Y4qr;MCqett?wm;7_5~
zvcP6_`E!i-h^gy3dH5W2HF&nYALQ!08>Gy|Be-q^_fkCIm=umFx;<lRn)=X$^C^Ca
zz%ThRZ4O?eaeUf?me)TVoOJD$KI6C{4tTD|x^~>ZaoF1DhB#cQS+{R=orZyACR)A3
zU?F4U;n3w91PhmF4MPEC0?TYLd0c1crY(u*vUH3Uwks`ous-X;4I8)My2N-kHV?L?
ztFa`XQ}L|G^N6u%CLoenktu~X;n<>e4qdu2QBYyV6yhhfwKmOoKCzkfCqp3Dahb_f
zCo`33>)6D&J$^XY&gWGgWA2?Kc&&2xhHBvO?@jDCQ?Nbdw1bux!Dn}t$9l&R8>$`W
zQ$3C&aKN0GpI%s{E9ECpiE}2bR{z`>!L^rSb|LQu*`8J1{SBj>RyfD4#H4V4i#5S=
zYyx8tv>3t3%6A)z@@YYFy5*4HPWNogW$ryP_?*jQAE$rD+JJpx;xgZQn)#fygnJ$x
zMLRojKjt>Ph2}a}*;cmQZEGtj;E%sA{rP;G=F02(=vc5Um*3x~T<$-Yx;$K-Ed{rC
z`M51pE=F7*R)<^W%67jw<+XbPVp<H(5c~j>&*15Ol^z*z+pugV|J|4Sgdaq?^3pkG
z{tcGla`?;r`F!QM?SoUco9&Mx`RysTJja#KflL%63n`D=x#adUkDp=epN^^LO1IqQ
z<1~yH&sQclPGx`KkGpp8eU9@yhPh8C@DbYjwd-BBNT$9p;w(m1#RpAOe`-8+)B%Uu
zO7=WBJuRe0`XcX}vS27Ae<kv49v?Rg@IpG<Q$PGNcEgoc(C2Vizx|6s*m^B1j=*zw
z-pL%a|Gv1C@iv-2jpK#(`kVh~;&_OtRm$c|wi-UFeIK`t;PW)wKJp(HvgN6V<@i3_
z4&rq{m|%*c-TppweH=va_0Evs>%6ZHV_kPwPv`?-SXpc{9)(@enobcYWZ#AHoJ{}=
zVNk99`Y&Cj)*0o4a3O85U&`XyjP4A}eE^U8`W?n%3kKNdVGy*iyK8O&P8i9ZD?Lb5
z2peA8XjS6BT(WGxAGEGZVgS1xysMG?x>;YX&X;OkhoU$KtFd3g>L`WL0D=%FTiVU=
zrcLRABUo-GDTZ*s1uuxS1WV$aX@;P{ff}3mg30xGxwSD&u)=<g%MpdC#JDjOHCPTU
zG!hgVY=W?X0f!ic2`<C7PEcV!CW`U=z@fRb!*0(&^b?3<lfu`&Of4A9nuOS4-H4$q
zmpG7-z_u~GFxxmS6nZd7(S{EwD0(nHONtg(80@fdS}@Pl8B3aY48q8iAk&0$D8k@9
z8L>rhtr!_Lv7R2oXP#2sVz@Tuo7VIDRPb7muP?J<y??#S)qnR6G>XHf1IPV1j<udx
z`fE_eZL_g_aMxkL+inPqxdt=+ItyhOJZ@ZDa~-e2ZNlox=Z8MmHODb2WE>%^?n*AJ
zhqytN?4Qa`6Bt1d?eXST(}5U96S$Yb1m1M;LoZzWQy9y0=csr!4uNVsqm4rFrG;`7
z*RH5z`j)b%)(wuB#l7iw26t~@In@}WV?DX<LtI~viTAcj2VZyZkR=NvvpqOgJslYJ
z{6JGroNC<`egzkn=%%2Wsg}nJFJ4Iu67_nnn5x<La_U;MWzZ~{-&;-*vtrw@m#bXI
z)Gz|*!K4i@nCTZ!A3|7J{(8+9M<5%|MtsmIuIKfYn*%BwtLwr`UOgq68&5)%k7A%c
z>SHRbJpZMjo;ViTmMD1_T4z1v%XS^Ahql#A8%kkBVT3WHhBptHI6nMJLG-YLwpJK{
zd#<VHDjTqUs-qM{Qn(&9VO<lgO(2Xm@uM|9cTPQ%Qv9I$NDGXKQ3xIQRIm;H(mBr+
z6f5FfjsdDxys%FU!Eyx`xYjP@fh8JzSeaH>b}>$Y6+j@Q(Cmcebo6zavB5DbxKKpl
zc1@_4>nOM|AUzXBo7T=&6CWN&xfDlW<BIy)dJ(I}HWlveAkMU>;%0Obu@^d7H{!FP
zycSrnP)O<SX*bP?v(;Ur;OiVzX+uBSAyU;M@xG(M!~Epk^D2(wDG_ArHg51^>yqN;
z%DHC+zsmLRiG>7(&Npzz>qlSTg3t~*CQ%G7hE`C<LjPJBWK>46+O*CHjN#f%L8NW!
z8I}9JwX@UkgVGq{-_+albe-Xwv&|{z+EouiNN~pm??Lt0#$cOk{hQZ{N(&>dk5v&y
zo>{rZ?6}hsSEr^$5-8@PFp@-owK^NDDYCoDUM7KJj+bmz74ncCwypnw8@67{$|t~^
zDpRhn+zPFQ`<bcF6SBhgVK$2qZp-5>+dynsS=@#x>Mzpg!k|Geo@uK3+eid*odaw8
zhs<{KdJ66i2TfYw+NM-tWCHdFj1=V3TsWUxUrU`z7%7$9?BA}GYjr;cHk%=g46fe{
z*D_EGXu*Mhyw_6zp@KkHTFnqX5@RC}9v~FJH+y1mz%;|6lNKf1hSniW6fyXl7A163
zLU&oT9-&ae_#W00!{%i9Y4=-p9k!!oNMa1ZtGnCG_?BVQ(oJCm&kHE71McO<AfB|s
z)fQuaP_*C^NlP27L}2OBhUH@eBUUttqYNgBAWRfn*bZx@0?U(T#3)H(8@jBELs;o-
zN~A4D&^}s{Sj!TW-w6Sxi`Fhfuu!d$Fk(BPKXp&Qy#r0)nC^sISAXD%^K_3>h}EF$
zXsPc*ZhJTNg%B?Dtwi!!gSaxnrcN-~ylJCZuyC8OEM{PM&~)M2Qt!e@XydhTRuJHM
z8D|_B#LeyP?lmJA?1XPeSR^8@7QCYHTvZZ*AOUwhug7awu&sss_U}i9<X0Yy7d3(~
zvH`wu7eVPQBykuIA1f)9NMN>I6iURP%?cz?%j2b#!GHq0)v(#zz(y$@2y;8N^66Ek
zd%H#O_1!d0&?u$+ux-Nj;Sj7s=u<xI!Gy<KDZ)rt-CCBLK-j*?y+z?`{g;1#zVH&1
zv(^Nn9H@_47|A69)tXzW^+(CO)Hc*&yL{Inef67ukt=ope-Oe*4DVH2)^4(KK&-X#
zlsM81H&oU-1mXxSM;K>>iPj^{j6(w9q#3aYQ*c?uY0|oLrIiURN?^Uwgy#iXjKm;B
zB;p8q3rmjn?haUcj6igNl?dXEus(_ubYGQ%c+pC$5?G8dwg)Xf+9+PY;)B9VC#*L{
z21f0Cit#`oywFvbOIG|qIg^M#Lf`ZhOsq>ViXCWU48qIgRO}E6H`c-h8>32?;&o%t
zY|OTs0*VdJj%|7&j?k4>v0Po7G@Tm+g6}|cWgu!cSd7e#BXn-|lILX#qInYp@B{>s
z$>AZ>1vktih%rO21O5kkg%S#TPSx{Aq4e_S%k?f8{Fo|#Y2i71fjBG&U9h#`BapE1
zhp_xwRvQAjj+Js-#CMLmw-_n*YsPY+%;oKa4|#K$h5DYKKrN4#d;=!=Tdy4X2vmB2
zF>D_W<H4jo%Y#WjzC`PUpV*);e)c1*tbF~``ZR$t_~B&`dq0V%Peb!T0*o;-i2=lH
zLS-<2uHRuVoVoJnTUW`4=6rP*yS~u2N*D<z8+Ppq&pCUS7J9AbAWj^D1YIQc_rqcY
z79d0G;gU24aRc#0tY`o+BuQ%#{G$lM_#VypP$3DoSusp$wZX4$U7+^P1-)i;W510H
zLZO7~T6=qJaC3?hbY*2-WYOIgtx2rQEYvk+UBbQ6*3)4{kQT%UiElw!?VTnG_g=9s
zh!b#?#d2(-76ReLFz$&JV^Ac%$Y+_1SCXJREGD}3q5xv8UeaO>twpfVE~d*ZT8%W}
zBaE>m<BFuMUg$>4y3|6>o`u~ehQYN$gb@vn5&GOr?i~ojuy(Mu+MX1mzIPV77aLU=
z$<^*c`>uTTPk`rTycO;qc^`r6EXEk&v+6K@H+jNH(6Xxjd!`mvl7PFOM<I;V>e^gM
zK69aLK2O)ZFrtUf>K%$vUKFa=YSmOL0_7YpZqU8d2NQoVAlu;5COian<IBZb`Dw1z
zerc3?{e9@#oa>mfZnp4+XP9e(u(Glp7p|_Tr_Rkl)a8~^UNM0@H{(2MrJE59I`zSI
z3eQQo5h;cG+}Vc0$Ow*`KDa3;H$^?TUG4L}-2NA`eXWI&S}~<%LvI#e_T(1zu~W#t
zsJA>}_r82><N#p`LdR(T2proXP6w<)D1OjVghAM8eah5=j}($q!OyJ~2Hk8Sad636
zcR-l1iNXhx6-{stX~tK&6f~@>Es81D1r{_5Jdd$x>plyX9kejvx@HVwCE-4cLQMCf
zJ~KKvZYY%4VCyJ5hV{<97A-~+j8kGon{>eSc8DwEaH*9JpiW@~`U$q7cuVV*_|RC|
zIt6t#;a*3}lg^f;8E=9p#DMQ`4Mq!>QiKuPas8Q-1l;rL7>+jzjsrM{nD`u3j?1cg
zhnee|!&rdrS**Tm-LG8ju65hWtv}z_@a2}@NZZq$8}Dz;6XVupBrQe^elvzuGlUYw
zPimySS-rN7AozM4uHD55BiWvf`3}wm``hYuAW&1bwgs~3r$>PIh`JX>>N)uOYJ4fR
zbenk|(#D>34|&C06dJq7YJE8bYI(fO<_Fwz_DLgc8@3PSN~=EhVA#H?kCxSEE9Tk*
z=mdlj4tz#oF%tDxwvDha+@M*Z0d{LGjMVC^%Xb}$`q-&eKg+2%AEU6c8DoU!?&h@{
z5r1IHbrxJ@(OnkBkx882Ofd}h(gMS}u7Zm#s3Z^_@S9c}7-V9s5egXAS_8`{crYd4
zE-QvOAd_&VMPVX=xE~ZrQix4tV|_rFU{Llr1FlbrD%_6ccJ{(`76hBtt`0LX4A&>H
z{-6lcKEEd&P#wZZE6N_}N62eho-j>KTj8>cuC=V-LhBMNx1tHkv4P?3AO(W8!}V56
zeAFa6yUq9@1HiLy5NKdELaP+Exig6mYZPN7%~yXz3AlHlag2R+<6MdIE8$WGx?TNY
zu5m*<P~yT|*PL8&0!yqevdv^M5*f<6E{Nx46D%TAW8?Nd$NOCuKB^nV$2?)Kv>M@=
zv95v7lE%5~xyC?e$qp@{8@#vGy)Y8&mo8vIBGLUW=Axh%(&)H1YDr<oOSR6=To2}r
z%CnK$9mReQ+lRRVH}ZJPHV_+DRwMDOx3-FBz3q|uYEP8HNYqxrZg8l+h?UoxVuX=O
zSgzHIX};@Ft=d)I-i&tZe2l_s<3JIuK1Q~TAm#@xMVf7(b&4Id9BGCn2;EanPNoBy
zGwz6WtwrI4VhL?ZAUe>^mQ7RQM#sD^Gd>J!j*MWDwj^nRzjwwOq3B?(Nf=z6uC*YJ
zP(*2kxI!_*T5MpD*$O0#ErRx0|5q#@@k8*#R%b;O#2|@5z@Td^CyX#}2jYW_AtnhE
z#TW*IkJHVTt$VA>$oABPnP`DOal{s3d=mb*Vh_ZbZq(h3uSX?}Sd99Ug@Ahp8piQH
z4__MB`g-K<rEZ#~*<~-zcFmzDfEI|LF$$)PM^tAiTe~aO%q?(HINx0(D^-cgok4)-
z<v1(`+owL>;eKz!hXAy4_FZX&ePvjb?b`0EbT<sr5`uI{x0KSNbPY&JHwX;UHGrUW
ztAI)~bTgzdv~+h5G1LG9dt7_3Z~ysR>o}gD&&>VYbzSFq<<N`{!IlUb8zipHZI-(1
zroR5+j(co8<_1+;Us;?R=fozfPt|X145Yzo;&`WQ*<+6R*0j~3gU4tlPOpvY3@BjX
zl>Rb!j9ewg*}SQe-|3BVs)VmD6;}%}{&`~BI}_~)nH3eJkj}ugv3I5V({W7m_VidH
zL$S~qDzlIos=L~cVZw-qF0XE_pU%*hm!t?l`CFb8VF$-<k{+nC-V#f-KPp&!XT6Cx
z8l+3O{jj|)4w|C>JM0B;GlG{pg|bamELuZnZ)>9N>poSJ*{?KCFx$s^42y=5!p}#N
z-$IJV*Bnol<`<TLk(TE!W)7=?MFeHGhH(O*XzvoN5KHoMHPbKWxIN@_pSpUF<@3Ea
z3Zs)G7?@>6fby}ARaN;h$`d<dF{#;cL)AM&Pt@$)la(5)M6W=qRv9^dd`w~7Jo~wE
zE)C4tBK4X%GypGKLQ0eA*BWH_ESjV*d&NkKc#Q#p3?n##xlE+Kr6~kuip<#p%b1}R
z5X~t`o0l~T_T=5qcb|1t8W#-mIc7X`q;P2_6;#;)DMJxAe2f3C1rR-d6E8gEQ#SGh
zfXMNFkf2M`H5{z(-MU?E8|x~*wL3l;)DOYJsEm`8DsCl!FgS~+8s<Vsy*AzU24cs{
z5*<CueG5$YZ!b18NdkQAKthbrfQwuI);Clt;<zv7s{&mdBL#MPJS8)mefbo7H|Q{a
zwVods(fKSjdU{dtNNznBxxCDhiev~;?9;t{e^ri-Uj9b(cu^cqTP3()!mLxfV*aTk
zsBp&9W%8%DZ4?7_Gsi5~d`fp~_&Sge>FhSR@cVogRlX_G-|)MRaP7UgWTrq|K=Vu9
zFXeTE&+c^MA|vAqet%&&H!N=EQ}X6%mVF6uXA`N#j1>&weI^>NGs5_=wk76x?O6=J
z{1%k?{6NC0=#tH8bg&&8%hj)zQl+Rq@ZuhSUMzmjktla-*|X7n2U?dfZiRk_vvgQ7
zjcPCV=jiDki-jOH>-p#s=Um6WTv*WsR=1~tj~w<i&a{iD6uhCEH}liPtMW-Nt_ey~
zl^w;nDwXV7oH7;8j{*QGJz;FW9Co}OqIaxk(=O{#mbvp2p<#-u!aUD7RSa1|U49CD
z_>IjCz9J2$-q2LIvd>y3AV?C7W{Fk~<%ebT;jv;@$<J;3noJv&a>hIm!8WcB#l<4Q
zei=0b4dm*^b$~LbQ>fUpC~}kbYFEX-UDx$dwV(;l)OA!=G(b1rTA#8ChrS4f#+(l*
zUz(6==%^Sx<pYKv>~T&`tURu2@MWg4{+6JVGD#Wx<Sc1<tY#8Az%PL2|1rwCf!>GH
zv^~<+`sXX4S42Rz*o&XwaoppUZvx}tiN55NM<>l+n+G-IIUPD>N?4CO%^hm(OJ-F%
z%trQ4<|frjS!gr$`*Y&c{aiowsIkHu$>(65Q9$Wi{=Gd|lediA=v(I+eZz+(`ei*d
zMDuS_RO+(SC@t7F_tnaA5P|dWg3j^8y_YK_oR|{8#4J~|Z3mrQ2M|Zm%!s({*(k%d
zZ<^o)tBx<8bJNP@EtS;^lA3W(1Wf!qm~DjHF6P{~4^M<lk<+V{>n-SfBUds?vN-;N
zW8VH6qOd|j%j}`hlH>7li(wTmH&xym%89g>n>le%Tq9y)Cxzp4q*2vCp`!ye>&Bz5
ze{}C7heoO0Y{I{f5#}AlbUvVPJkl@$uQ0;WA5%(R$;EWOGL9eh=!&IU|5{|76_N2_
zbBc(nn*g6k%d`N09FrSCoNi}W5E_vD5O=^j|IF*%%RV}V4_^4Jf`pdr*in{a4}JAX
zjgu?4EZr_%#IlmUFb<HM_%isJc}?llTY3wGHC<K_-?=E-U`6N%sZu>|%cXv$!}vHo
z-PRaw|E8IT0obNIFh?HA5Hx8LhB;INGUk*s!^u~qNR`Pj%c{z2Zg1Rp;m^zOUIsU$
z?;{1~1Av?zPkGL8c4y#9s%eV`Z@wG9`5paFV&`iqQ%<c`VUEa1H<G<r*g&Jo;;ULJ
z4jou+$&%8wfCwDa;a%In6jA$QDeZ)j1^8|0@)gy*0Zh4oeCX*#$s}4l*-&fcr}n9%
z_YfM8pIysFT1Vo^pl6!q%q&`3YJ5>Pwx6phjjq9QKx$@N|N42$&}&fwyFX0tJ!xMs
zt=mplYNL0&Kzy<<`o`3{i<(85=6%Ucg+`aV5t_z<oxX7MMw7TrV%I`@$wC@c=f*J}
zJ_xItz)FEo0%2s_3C!BbE)grlaV}NpPNQL(l+rs72_~p9wY9-e8|qT2<f}!(L9*18
z@D$a##W%*S`8{f-%@eVV6ex@$+SHC8T+>FBD#lf<L?4IonZbsuSo1UCLR^&sD!=^h
zD{G%tDJ;z<aH&&mLG*dA79W(P2pT6dd)LSx@9tG?#y|h;4#llMg71G6u;LmTl8YdI
zzxm$0jJ@&u(T~oyji#;N+!Plf<u&hDUaC9zP^<`wV<)wtlVz4^{aZW+0IW40JX0xC
z2StTT5)aP<4?UcJix@w=-AUE{0`GcZ&b{{vuqq$v`1UJ0sI;13zvn#csF%Z{85Q)I
z-*x6%KyI=2dQP-72sjA6rU$1$Kfx~3c0rym)q}1%NK(5%B&?R3Yi%xF)+`BWO<FB}
zO`go9yb*lXW?tbs(b``3gPEqA5Ppr@&~|-=3lEv}?RU*nlL_)}0+Q$($g<)<5D3T7
zH1cm%Q$m9@pF_N_ueb_Qw>L1c3$pq5E1Mi!d7PGiH)-llj;y;K4%FH<=hO+8xseE;
zy9-_nOl>Dv?X9^(gMRI|H80UXWDxrAnORF^?jUcfEGWiOro`Ju;hh~;jlYN#C44q>
zeOvOzailNRSf2#v=>WJydes_l2W+TqP{{J<g}IQ3i?L;#2p2sqSzlsGJr!9MpiQBf
zL#>Hp%nN)fxA@Ih_<fIf)i@Oe##*WMDBy!!(afK&G~L7JbprFhlZTn`8HbEM3g@e^
zml=-$g!4!xFz@Pterg1`78l^Iu6?NS_MPe2p_EU5A}B&gV;nTq#2RC{qalc`T0GBF
zhlAd5r57Dk2IM0c;7`!sG5JR|YB37Tz)Uc!KJKYd`HMh8?Mbg`tP;|uYpWItd>Q9o
z?%ZG<FNX+ag0K$AA|bAvpd_$qK3Ui*e-=t`wQgZT&~FamxkFL}$_`F`f0Zy{x;Z%^
z*|b9kXA%w~Xwu+LIPzJT&wyV1pi}EYhv1g+FXK><-b=3jjJm8H#$S%AmJz$UsWq*G
z-yNi_6A<22$*U0hnU!2Bw#NezPQISLVpdFbk?@g9XWNPD@4cA-yYk#c-IEq@*V9U(
zzQ@H}4I2kg$lmdl;YR1Qf4QKtpt*2X!n6A(otN!DQ3pMuwsgUqg(zJoJ(!GyMNmXS
zzy|k|hPDM1Jx|3{J5t*08E|9%zDe%9R?oumre*yB&Ro9oZKu3#)ejsu?0TZ6*SO7G
zo>FbMT5Ae(n|87sveAUk*=-vM!QJQZkX}t^m7~Oxu#oUu`ImgbZDAA6N3n!=!}YE6
z33^Y68-BDWAj&oGPOSzm>k-@jWG}(-nOcDlsOdkSA%-TAf&4eXo55uRB;RQSHn8l;
z9lWrEfnPfK>T#7nqSB~os3lwTe0BqtZo|%R{-Ydjbh+7gy-Eh*HGT2>`6))Qa|bhK
znEFN6_En2~<x1eTN?ltx(GFpJ$3}+2=}vi<df+NOU5F!r^ujfi6*A)}#~6rY2PBq=
z2A*ULi3uTYwtw!^q(v24E_*Q4ckas$A`pm29RUlYMZo=^Z&gRm&5p6Am-$M~QjE9u
zh~~voenRFpiM3?7&pg~J`af$+``pHIrOp#xCJr~)HdQZ(7-XE`!o&ZOlS}+{=Z&(B
zDvYak8J30{{j$MlymXN$=E*bQ1O*w`AA7rW1z_NSL%9xG@V2<N#4P6eBxd5eRR>ub
z26iJH>EW)^(|m+rAwQS84GY%l%^0i3{pR+OTMsiSm9?uc3}JP^&groE>uKlq((hXn
z7!M|CvHN`)2E4OX#+I4qIS0T^U^MaB8#Dwvb$GsZH2$664_vezdyn}84<1rF6N021
z=89$4B+MoThVzx}@Wtt6O)pk75^hstRO`L9-`}lFeT;4-O92lC<0uj&*y1u~k6VRX
z;S8!1#;2z;HjG0*r#`8{ED#WCB1>g+i-+D$J;qcRHv&EBPOEx=OGhcp+|j_b|4v2b
zMZ?(I=a@A$I?RvaYnfiGpcfu)qfO(n&q*@?wDn(z=1G~huK6ZHFAF%$98MlBO0Vzs
zh-wD!MgNoc%-@xB1K$qElBMA!ecND%PUSe8eyv<`*}HoYXCdG*!uJa-5HuII8EJOo
zPvXwBbixqd)j-e0wNZk!M7oF}JEu|72demCBN1TMhJY+?$b!Q$tz@Y*K6GeNrtT=g
z->~hyVXto)p6T}0(&m%A`gZAQ2v;HjeNn-YG7tYULdP&?V>&WG<rstio*`+vg{46?
zdz%Lzg>B%LyUot4l4<F$b~zYRpg^bRNNyeG#q!H~IhR11wxcJ9<bWbCB4c6Z@eMrj
zi3dZP`+`uy%~*dC3!e~g&f%4s<@lNWH@nBKvEQ<Vs$K_6d(bnzZrpxtkXEvcs-s&=
zTpu|f)@1K{>JNjo!W%P@3!dn#ZT>I>eCaS1XWwH`V#p`q+NN3BB_r#4TBvV2g@NAD
zd$6mhT)Li!*6qG=v#}1yxaQi0h^A<eFds-vY_w1z6>n>%fGZSVjBU)MJLGw64ta<r
zBxE>CPB2-7PwE{FF0=uqP)O6mB0tFlyAl@xVvau92I=`D(;Lot(Yw>i!$tSaZEjoH
znJ8u&S#1MNz^-<Daf0NE&itJWiI#TUu7=DuQ7Ca;?Kr<{=k-ez1O3kM8)A!uLyXkc
zgN=b7-jEb|!9MlhgfD3cqnn5}KhyYh@)x(LEQmTWp}TiBXqDGItuIDd3pp*nF5iN_
zyqGtbp&od{8Dcd%1#v8{#IcJ77UBXUPD{Mn-fx*{Eh89w&;1yl7PqI&)KRK{1(U;%
z1&Gx*F?rBc^xyDOgcWHgbUplrAt)hW10yj8>#$X&GT&-m8Uw&DtSWh@9?}?lbw(u8
zrW*qvC~}kSH^mYhC$I3lxcbCO>J?<O9%@+`$`$dT;{q#)1Kv$RSXaw@C9xi)gZbid
zY7)hBfMd8~jQ6jIPNG!pP7I>Zj}KXJ{2gS-Ri%};;u1T?!>tGhLKLl=-p|2~ZKL@N
zrzZ(Z$OBD<L!8%YtQ4$};F$54YisLRL8}5-k{q&J83^!@sfgSpG0}h73f#OJ#3#eM
zz3my^ZrW1n{5EqT5Ih>&N%Za7X7gQro;G5PbcJ(FSX<7~XU;#$9-pU@k}*sHO&{C}
zQ#w#3C6MNhSD&5B$-Yn4az7nz?9J?2iu>}jo_h23TH7zs@gEpS9Z;jS?{pTYm}8Yw
zJM?q&1R7@$;A*i|V$p_Mvr?F4q3PP21|u*f9X3*LKarxoNKPL7oYI#Sp0sl*WnsQB
zoiu#P?6j<80eM;K`w;GDDHS9}s7w4iipcVF7yI7nfP?z$c{DSNe6Z~_-Tr;=N4<v^
z!$7<0OW=m7Lwv_8Uc9leR4UD~U4G0%n!>B<RI4zQB$XoWdV-rj`Qu3MJglm2qfgzO
zvCRDXGhAWC6PaCT!zE?`0e2I$8j7DgRQe3B*#?UsnvGL95-o6QLe6`RBRC=`uJ)_-
z*<av_ObuW@#Kj^u4id<X2F_9%bH1+>|4r6q&`kN7)2-p6el%$6bHAbjR@Th#*>LVh
z4+Zs<UbsFlz*d+UTY|fEzX+jvH)_@0j$NdlArp*pL>=tDu?cuJ=$mgUgUH(rMTS09
zbPHl>_cZf-*TXimT}h(vS=;;(#I5a)*`O^-n5e%<@Gwzf`(@mS#=_AMLyj#uRE>^D
zEmWVc)12x(ualBCQ)k<7#B#*@$sd!oC68{H)0B;#?YVxI^eeMB1a7Q0A30GDzbnR#
z{_r0kc_qw&lb4gnP&GvL`eW#KD<yf%|NQLb$e&haX-YtG7h&aQ5WaC2cw1njdoMjV
zyp>zeNAHKAN@$3#X-ugSH2Q17kN-IO2V5BF;U+#Sn!%v@qTh2qe9Zix9!R7-D=sS-
z*!{Y?N3E`vJub=VZH?)F`Ci*65AaxH1%;#%oz7qWmv<F$MM4GEGI`_3R2B8eNFV&i
zCnOnH4Up8M_5vqCl>EQEs=~$$RD!CpGS<I?{FfE^tRxSO4o^xZrgvuhPm6(Gs(7`3
zy{aPZGHLrCKlK3_XUvCiGCWpx2Jtzy|FQ^=lH?nq{aZtHL%rv8tcTa|!L+QXqqd3E
z#<itpAseyt;$A&f2Dk5KmD`=W4z|vC!lmCDae*7w+qt%U%@?*1p9y)qS_g{%@<pT}
z#uS_efZd~Ir>P1vCI_)SeL&;wSJyfQRNs0D4Be-WFNAG;u^jrJkE2rlDdqr`N+eEw
z`eezp^4zFecoJpi@6W|XJ;q>hU^f+aNO#hASWU^-p_<yd4W2;zw=g!2PZ9|GLfo3#
zZdhCMp}o(HPa7>mo&bq-<3Q=`sH<O&&VSg@bdx^=Vxt|)-i~{tR&2oJq8^d$g)H<r
zE<c{-{3xRE=H;Z{@2j$6Cg%ic%Zx61Nz~b8ur=e|bmpkHG3P@zdG|w%?d-P~e#6Ox
zA7i?|A>kGJ>GMZ7cD5O||6#UU*dZPfNlmqYA4LLd#3?FNYQ`teP<-?JLIJYAKW6<;
z6ekuZU5u@UI2g6l?^bmo?@-i;W4)502}dznVon3@)~%i>1;Fg$nU_!c)|T516CUf>
zL3~b~iG5whv`krLXOcVON_Z?YOtU2ar5v$w8>DgNfM(>o^WxS6#9f8Cj5X59w9;!#
z8nOs&i}m8o(8#!1H5V{)WUqbPs#oy0%L_@8|BB0MZ3~6rw#kY}0N-1c*#ah9G~zUM
zjlh`u&XghHR$b`2_V2W0PXB?duVHn%z^+iRrlGC5tb1EHXv5+Sgd;{FUjl$JzYTr=
z?W^;aY^s2ewHQ?^^VEWqtoIVF?Ts>`*jY+M(uMEpG{u(wp%H5_4ZOYCf1b7vwgVNo
z_Fb|I7XS14Zc}(snR@08!OBgdjCle_$h8zVUFr%o<kQ!Gy%%3NMmeFlH!{xwTr?_$
z;bhb0@2S;{tppKUX-p0Y9r;y|o|TR*3~gygQ72lVxUI`{l<%^;3YY{$b%(N8wnaDG
zCYv7msWBpjFhqo+e6n<>oyb+Lo1|dF>UC8?U~M*-HTSf_ho3ES5F068Q6~S1AHp-$
zTj~{UzZQ@ypM<9-Jv!m`6%@bhhid#S#qB()hmM}*1<vnA3Oe4NsnU0O6+fdHmw3Y9
zqV@i+Eu(yVz?v{r#mYm7)x*cPh_^>5qsSJWl_M@I82zKnO>+v~b4S}(b#0Ez+GV}o
zsd;Yl*QxI%e+W;q>7#0~4(*&KSgd5kO7_&28ld-HIn|)aHEtaXJlvxjx*iJ4uo?7S
z8aDO2h!MvAETQxMjwaY??TuH7_q8BShH#>c_=KpKFk1f`Fx*Rf?&;#pl$R|F+4D7>
zWx_tx?7OlTx%$l#44sV33)k$l%iQuKQc7HK%?#oqD8<^|ug=nT!kgPofp>n|prnL8
zA>OY(A8aVYDwZ)yiY+P{Mq)8)cT{Wl15tws0{rin%<JYfNs~EUWScz7Xldu)ZJsdQ
zgQ{l((hluJ4u#V!>=*+*YR$9=$|Hgnr>-{|o{#qtF7D^t7jh4c_RoNwYyYkU_mbqF
z6R=`X>?*V-B{dicRTSH|rPKlNM56_XWG}mxvud-jLrZ>M2gn(Qw%pe~$>dNm%zMdn
zw>9EKNOn1|NiuIogD5v(JObGkF?HvoGUHJh%PF}Thfv)0mfOq5VjI)UH|hWfS4Tk$
z>HEo|`jRf4<1dpR@^)1Ap%tb!*!yosrFN)Gbb#GQ6Q<e0gM*1es!X4$72D4dCPT3S
zOj#iXq`+?H?hhACeJ4$*I)w8AOOq)bAXQ-P%KqtJG>XnL=A)MYB-*_2Do+MEnYxpU
z!}KmUf-WZ)X=eL9`1jfHQ8=WGV}g#{@hvl%SP<{T;!fqqcLc0(o$s~s9xM-anW7sQ
z_yjwyo8^KdZX}!s#M$zKM~rNfo=Bf?BYcDIRNqwDy2byQW_w#OM%$jR)tlPzq^^s|
z<MW?2$30hq#SB3o9YWrG$ZNy#)tmIm*_uAu+Lhg3ca<69nYGO|dCLZVtFw{qe?!{`
zbzXAO?`{A*wom_V1Qvh}EP3rEZOw#T3VizX0)al&sOMiL=5+>BC_RkbP5V|Qovf^?
zQ#ExL|0bcl<?{`JY!0tP*_0<<TgI+-^4gy7c_&@Gz1$ouv+YzgSO_pEI*5!DOm^^3
z&G2?JId5dTE8nALTSksd+nJJ8Ie^FH4a8@Y?`tL2$=w95D@df1cyGX)L%8ro>^R(}
zRR?9KPkxVq9I^EI;FkbkzPqtq&9U58-xGixivteWPNby+i&EH-1IK(`z)$X1TW6~d
z%jUSfU7nebz;^4Jp=0gp&(cT%Xw&jdCQVitbF$zRX4!|Olh}0_T#C<UsB$7I*Vvg4
z+37uznsVAz?HXU07_oKc{_&q;ykl*&zKc6>a*c31gVMU&+Qtb$gU05@t(1R=YWRb6
z+*TT)e$1OizU_O;TcfRcapKJGo-9g)O%w!0RuM^hE13tSG%_O^R?B8rn<cQV&LqpE
zjAhDsz;v`C_xdM|FWLvJmvZ-MbI<V^Tk2{MEzhNTNBQ%NNO%f{KhViuv{p1vI4RHt
zpBD!V#rw*9cF6jxmfK<sL66AHg2$@xnEkfANPZl{;KlOQO8b!g=1u!GMVC)FdlUbh
z-Spu}3i1OIA;&02U&GJljKTfMn#A1YV$mB-aic75^^{ir;b`XzKAk1ns=K*!fCo9*
zm;=9<0>kyOYX)Wl(h6;VLH$Vuj7#hKI=3~$y0%X2rm{?q{C`WIRZRNGH+{SlT<Sqw
zE>23jR%~bHQ1gql;^r>F$XwX)iNKr3zaex(47lc?k+0g?`>Z(EH2gcls*IIVYXJu~
z5mfD3*fnkUCvPG<;bz8-KA(_CrXzQ`!QA1aT_FqcdmfW|3b|o<dKzN3o0x|qIQ*@1
zPBTIq{CxJg(zy^XI!B{Qr{crW?Ikd0UpY~Hf)NniayR!`J+yT~a%6}eXKWiN5P%$3
zhYgYi*c2a94}VAen51)))^eSm6kD>f-fTgn^bPeN@vqz=SC8(Jjn3#X+vE;+FofF?
zL$?<rmsx$J#bJGC+=)`%JZ3eM$?Oe>W+XB;E`5y@8uT;UC!PRoxbGs!B!Ey?{^bK=
z<$k*<XVFU(g0@@N|2Od-PCftG<@|{B)It3`u|6j1`2J4X3t|;P7E44~GBcH!56c&T
zB2f7<K6U4>oW2Tl+ALIPz!w7E4{;#2Z<xe<S##qFP8sJqendzYp+OM2=K)JV`l8Rw
zl+(ntwX#7jH%29B-TQ$5QVz(7oPLCunn)B~r~o~lCScwe_0CTI3xe)gK%?w$_-H*P
zQ}dJM<7oBjJk$FiTXc=g!@DQ?8T_X1L7iVd%=0-7gp5#-%=Zxs;aKa3Zw5PUN+NPB
zmfq>R_A^4hy#|Jd!R#1)D<yW1gM=_nfoRtxJm`QHv1ozfI${<Hd)>adshX*b2v{ug
z8R#_1)YIH$&G4p|zrTJaW>OVSu<SV%ckscF<M{*O1=T!M`!dx>NF~n`nP3ag;TKW!
zFv8xTiDgF7W~+RpwW~i8{ILACiviuR<6h9(zZ{D)*c`Fe<>W*F$B`SuhZ22k-lVT=
z+l!PM()5*a!kMo(nX!|-DL1fPh-(Z)7BSt!D`4$YP9n>#S40w!@=)$JlA!fDja(FM
zz;^a0A@q*B1&@2gLi?kZuY{hjJ~8={0-Y3deim%qw#{eub_s0FDelIkkByOae|0Sd
z!^e%Gd>pH(ADb@8330Y46@ZFX9%*z6`Of$}Fs)&*6DQ)xWz~}R+MoG)!}5>W#e@lS
zstpYfC)-K73^rfOt<d=P;?zZ@siQ;Kmdy2;G52dYYT@uuP}PwSpOsP<O>X{osIkR?
zW@+AR=myl4;}3tOE#16(?wDZNbdyv(L;~`h2lKawAw5uA^1CG~=G%p;2V%}E$(_v(
zNu3Y&Q9koy#xENb-lX<bT*`SOU7uSsxdm}SadEVeJy&FvekT*XcR<hmWu^#HNRM$K
zeM_I`?WCyBP;uLRR_TMj%bv7LF9l^9S-u|N>d8&JtmR!u=VDu~7UDXMQE=!PH9~hL
ze_tfQF)sjJw+rJ$L&ofoLqi-FTRIpp#X}EHe4I3-%JSi$8ui%m*J)eDZMPHcqbiG~
z;JZb;)vKLceWJ_x;LfKXFlHjy@ZF9i_Axy>s=b&NMJc0Q!jG$b^xCOL%qy2&TvBqz
ztv-P5I~DhnN!m{+rLKlg?F%S0z`k7}E^<NW4n_0p;1j)6zD(DIsKQa{gHM+D_^gP4
zW*AF=Y~Qh8e#*q}^!u8j{+(%VIAptjeWY{{*DOE=O%13IWgemm?6=0(OQy2ujsp8f
zt0ShE_;{nDJ?<Q}(tW^gZuhR0nGKg?DHrKz9Cn%4#+a%vca5?bVuJgPQc>*IPy*RX
zA9!VX3trHbN!%)Nf6rU=sIRS_#hMaq(HbpF&9mQipVfA)+m3F}MzomgP)d+JF3z)$
z_mmlyiRGxdI>CC%NGF}k7$SapA)HV}%~QX4qmACId+Fz}lfmCDf4kEM9Oyeu96{#W
z|K=sL#Qo%xnFp2ziE^6;zuRvHQCq3)GRcTR=U}?vT#Uh^_nRj^9~OC;XcM#eE&Rj$
z0bz?8ln<6bTY_81K|bC_2WP;BwgP3$MI9_XE@(eq6x_z#R|q51l{cQRMb;DP9T~Q!
zLGRYS<Nc(|Y~ri}<$!k_@93H#=r{1d2m2q3(dIYe`J?VVNB3xHXD}g%!yEOiLl<70
z^Ge$Pk_H66B0hn67;6UZ9ooUJs@ulVVH&A>pIoIDiA5)DrtJj%4#xM%Fz=Mzjy9iw
zg6@-CQ46Cv1}rf*aW0M0XZmMfo<%`$cRncNmFd+*`K>-7_PF6IZ%BAJ;BUvcZwEm{
zbOQ?J3*(lLt7+%4BgoQ1`nHi3pd<;>DuYn2QfCk&J{!|}l=|UgV}F5B|Ir@Gb@Ndu
z+sx7OddrO<kEf{yS<}q>Csu6&qKWG3C*jC55|H)cXrqm;=2vo+o$@2cqu@irs>Cn-
zBhm(Llqy=v+2nG%{Z}*f!RU5LS^n9FVCz&X#u1~mou#98H`DRC=$O!>C(EFwUAVok
z5l=~C@R32};aXcEZpq(l7XtO@c2Ga8f!z$5=rWoCR>!Gf`EA^f4AQb*F#|O)o1PiR
zHIOJ6Nbu#!ZPLus?<~2WcwI`=FQB54OBMQ8MrCvbiEW#|9m0xB*VX0j!jMu{37C3S
zP&K=NEaYu~`Ln~gfK>m(p<v<mhV!F|SUY-5;}2J1SM=FiAND66uzWZX6JI%cVoVme
z;%GoeY~1`iuG7_c>tl>MrP9(llDo)A{P#<M1RY<wR^%I!iTYso66C}(ljo3<bOr-6
zbt@ZUeki;#%wzwS%lbv&b&If*23|0ds?sNaXR8>ZExnL^8ybXC-YN#jlk~pgQU$D9
zUKIkqlQek`a)2uZ!7(REVmXqUuh%_#X9}8&op$>%QjNIT%k3;?w=KWC>Eb5+)U8c<
zAnP@&P1Hk~=I3-vGbev(3u;;qc`{Tuy|BR_`&RZTFg*2jUieI3gDl?|`HuDST~7_d
zfyr>nG<;u20L4@)0WG!a=K)|u>Kn_Qu665Rwyqpd422B(3{VbwW<Zz2v}Lo5ifI30
zq6ul#l!vNe;IhI=<6kjL+;n%r)X@@3M6hFWT^osYDdSdkK1dZT#!bd1k;Yw!Apr1S
zoh1cGQxL^6q}{GaK6fOQ(-mmgV6e+{#t@VQgd>Fv-~HJnT#T~zlg~6-+>zz6oR*~|
zm>I=!MHS|L7b0X~bm;OvH!y8d1eCoF8N7syrFH4wx!b0j*79{%_Eki)U0Tr{>G~%)
z*!#An#B!600Ed{KQu~_9iuqUxnn2_Lm<k8&=s4xt(^)y5Xd1%U3bSaoHJ0}aZq1Tg
zt!y1?&%1qiIiWRse=jecXh~0Np=Q}Tbd{Os0`n`q`9Tj12e<p(l@Q(R7n?$Hdnk>u
zu`t^71isA+Hbvm8yM{Fv(LlZD*%r)$&P*mE{@JoSKGE)Y6?LL_t2FJhjO@v~Rra|n
zr67J*+)<dC2N1c%m?0rFWwMDy!~HCPSj36V&d>NhiV9Yyma|L;cNxxn5c>4kd4*It
zg>K^s2dyldTVXWb(A~;)_cqd2x!+Q9)S=$@v~E2S(@H#9bN%Z7*WFV7vs-wq@KLZL
zbAjJUCJoa&)VU|AxW$)%5)(M+w`qK*@-o%u_Qm%0Gpw!f=a&~Wvn@11sJT(gEie<A
z7f^Yhk~m*`jlffmiW_YDPsmpO2W~b{%Ds8?Kb_S^+J*~My<Jz`m!KR6mAa6H%8rei
z-c@<{usXGIy_wkJ`9~lRCP?C_*cg`;^wGU;w?wXS%r!a3j_~tWVpCCPiJo}v=Ns?Y
zm_0NH!<mjilP6m`-IvD{i!V@p(-oa3AAS5crVgXTxsN_Ha}r<NqsEHKxYrb-A7i+9
z+ZsCq!>QU2qESO7?g>{hg#B?<@}V(b`e;Vv2u>6vN%b%VjV#?MH7N1z-)zQ3!Mf6E
z#KOCY@jQR^-qy7Ft<4f;$MYz2fg4oDUfMBY*40g=`#*v{?Rk+#&A$VAm~(SjL@CMX
zclFD7q=j}$I9hY9_+Mt1la)_*j~(BT*Yx92@E1FOWBMC+^o;GuCROQ8#+w*xnJi42
z?;=?aDt2&@2ZkRezOy{DY)D!&*zNSgNzip008n`O))oZY&pmA_xSnY&p5^aGQNb?r
z+{i?z0i<@h0w<xJHOn`J+`@twrbg&kWO}b6EH!}@Q4C9AP>7uP{y_0QQ?I}=1rB1C
znmAIJ!E=-`n$F`}UeB;Fv!r!0re1(a9F-z89FSH|4ej@@5T9zXTwW|aADMoRojNPp
zMAQ5ayeD*51y@?`YUr@D2TkL`W$$w#urElN44do%_3A_`0lEIrFZ8_lrr>A?I)Ax$
zEZ|Y-h9`R4YSO8l#w}I{_V!uIO92*9CjsfuZXv#6H?;?X>430pxT<X|FIf0A7h+Bc
zbFsb%-^@#@ipEMIRq=S!4(QzTFU6b~igrD_2en$wS(#hx_zaRbjSwz}n+P}EA||*K
z<+@vLzhj`3MM)4(WB%s1_y-S$2V|QscTobeJyqzoQ>4*VWo|f+i8``2liQ4yfgbLJ
z?oi-<8Oq<|b9c)M%O-f=yzp3xK?#Q0Tj91A+MXPCq@WO)%5`*in>chl(~vhw$qE)Y
z7P5YMrBGrt6UqC0tAa$aW9EA7eNh7I${pv7Pm}ulefBiFx!f|Mdqi;iGZJZC<J(ZQ
z`QtV;Z%VMnB|in5*Zy{flNeVyr^3{&t{_WLZY<?DmPOk;<7~=cnZ&{h)8kCq6KQ)J
zMH-{RBJ@R}oLM3$2<!WiXD{f@vfz-Emj|HFJ^VJ$X)J?fRwsDW-nxOjkVe$6OXFPb
zs_>-OMfkZHc3emL-(A|(r_aN)_;4}nc`tm~!^wmWI;uq4iwAi=jCTLCBYol5nt$xI
zr#lf=0T}{9ansr#l%FI-#YxMVxy!6gbtvcLEqM)~?Iq7Ij%gx#%Q8|G<o*iFDMKXC
z=~HZLMHM#J|L|XWu;Q_%kSc|$O13X82hS#Kv5zcI^DH|w(?Ub@?!}_p-BW0SzXlD&
zlZGXved5e-L3qk#QX@h)Z7A4f<2x@bZ^?-}#tU^>_@ExODXHwPx83)HxU3a2yKaNM
z%aMm@L;0%t-tM!zwb1PnUNF6<K&iQwrb+AQ<4AZ(H9P)8`q2xDc)Pa1O(vL(QJwW@
z06r_jVcDJS)?J^$>_0OT)YK<msiC-8zB@}J5ssA^)O5?&$THiWRuhaV-!JW8YsJ;w
zf@1yo21(#z!2WJuIxESHjME$4!8XL8Tn5?zQ1jjS)F$B-07y?~{8eyiz(vYMiq%}R
zyzVW9%Eh8H4kE-;D|0In+&8%mI+M!&x=U`t+;A&qnu4qX?zAO*S@;9H-;_=L?fk-6
z9(pPQe$~m*DFA*%GL!e((V~r_T|HCPIn>&8(O&)zc&G3YizIh-<-t(?*2`y+7~P<?
zX$1IC*ZyT^_B!kRjpy?;W6)C-0MK&(-hY}~bM!O<o*Ai96ufyfO_uH3IhRvAn8nOV
zPXO%Hh9O>bvmota!pbKpRj$tOqRfM<XAu(WC_YM=JZ^Ig-iK|3XfVMFYPzgS*&;A7
zJ>q}*u1z0`%WJoz#mz~ITF)D?l)t;okfm>;!TijXX0UuSiQm2M&juMD`Y4BxCgH2U
z?G{AwlKu#F>AjWi^p8yj!skPgu}Fp#0U;)RYgbj$>->4{3s1oIGK8)!{Tba&*($a5
zcT7D|;5YUuC%(YL7+Z9y`+g_Iu`rkVoh2~5j`LQx<7_0THw?L_?b09pBqRPA>t8IF
z0(N*XW8lA>FdM{A%}1bWwU}hYQ%xcFD@O8haWo7~X@3BWVi`YQxH>zHBO8O?^(3zJ
zqHJ0E6Efz^D)s?Qw4C)Cdmfi$TeGY^`!N2--(x5*?7{RWOU?Bs)y*3??Gcj$?}FVs
z5$Gi>wLH_pRT<L64P_Kdlx-DtmAVXsznWkql>o@+R2dkm2-rXX{wu%ihVD)8<}EKm
zgQwB@AH%};lmHe-G9O@UvhD`C=v1!|%v!mN#W+3k+jRfKO=_!NE!|I+5Sb-J6xJ&U
z<WoiDah9WKy5)O;KWt(Z0R@HqICqC^5Sj!!IZcjFf6N6yQ0tmGy%>eh8$m%8Q==f<
zjo}%wW_Tqp=KR^K+%T_CeU7uDh^At^#>6sH$2N~iq8TRZa$HU2zff=`OP6i{ZEm)}
zV%2Uc;lkxP2Ux(ia~pIeg~03%f60N5(<kSu4Y^5(8$v7b-?zxSg)00%E*r#88W~rW
zPkd+_H(^+n@L0qUjiME}v^Q;)#^OR4lO->^DoBtZmQ1Ay*y^!+Gm^22e*kO}hpq&Q
zvT#}iRM&=d?m!w-2MlbVMI;F)Nc3mBA9U8e5W7DF208?rv~6YhwycRGZi9Stl8%Cd
zFI2l5uC~FOvsPf~s&}nO@wIUsOiuR6SAbcY8|5u)R|#HW(xbUxG`DneX;@dFi;8fm
zv47U{VTbTrF0W0TpvNzgatL2TwX;F|rY<fk4$u6r77~4iPRiawN?`crtM3831eapP
zOhGFmWZjc52DARy2dOdL+wT%^6s>m}S~6q*qMTrw4<S=l-##uH@&D@_3KdLt^tc8Z
z^*$RNx${dz&hq+hcPrC~F1PG>3oo#q3|@I>;T+W6`rWiI1wLKF<H^;hHnS2WeP)K0
z@~Jw894M%AwJE+Rffe=xLPQg|`SXh@HPTsjuft|*VIl#I;*j@U{~(zJ<jr6HOchYv
z&(m&S*MOk2lk@>!=s0`r`qKbU)>^L=<<j9{+&@MF%D*kVuP)x_B7T1pV{vxBZ6jAp
zO?7vwxb4UV{<zpjTE8ZtUY{{#+FQ4J;>#KOwf*KzpUJe&a>LeEHeWfXTHC4(f3`kx
z@a;;|Jld2iD`M^cj7EtC{FVa5*Luz2*%zj*wLg3jx0N;0f}om6Jy8`kFgESmhjqFs
zt8g>sE(c%VSWm?>nB}sQzYvxdb>ZxlZ3e{^(8;m?1>8O$!)0X#zMXXknkuH5-HnMw
z@$q{DpJ9<rt6(SF$~&a(1m~68o)#wN+;~9tFXnBl$LjC>?-9?1Ws`j3x~PDizZ$bK
zcIeec(mWc^{|HSP@McE?N#_|$v`C47eJ1~yruJw<$QP&`X>)iO1k`l66r6dFLNcM0
z3lKmQnzZStHSFw9D~UEV2W}t2qL%$)ooTGwOpiCYfjyrHor~K)QZv^??(D;lY36-e
z?@Qza30D|3fchLS8eTP_f0<dEHLhsvMt3O%CYx?W!!g(&TxK(<0qH^XzwIETg^<*I
zKBY*+U?hq7mevvRDA85UeFcSK(LC3(*6^HA8Ny;1*h&J;@}%0__C`qY8`;a=O-&T{
zbgW4B5+Eg&QUbZ?(DaC^S{5G|{*Wx<v7Aj@7i8H*%DCPHWInu2u96_>*66>ECZkOU
z*Er#={3~;>ep0=vG)GbYC2k>I3X=`|TORO&;BkA@VCnfAxpBUb$qwRqzW(3paS4}J
z9|N5Rj)HBKEw@UJq@=8Epl@nKuXq+fNQcCnGyrjVuO)Ep3Gp>(ea^(Asq)K?2Wcq<
z=%TM#Q%GWB@fDg0uFJT6$bA@<2SWE&NoTSMj%v5>m0o<Hi-}ycfalefQ^BFK=NG~-
z@=^o^uP(qfv~v+s_{UWpk(n21xE;9WT%5Jdli?CCe-fH1Dy{!T4YcA1vbjIBSTJ_0
za#b1i$F3zVV!DHmY+bhx*L;?E^q_Y6_VlYQuA0H^U4?M^wT;JU9^br+^2+J62M;z$
z4e_$?_b6@;eD3z>j)dA37h+Ah(Chi2M~*}tojm97Ax#1Qh4zzM>o44-Bd10@JvRlF
z27j*H6@5P?>!zKZk^5ZB?L2dFKtWQEvT^$KU5^rt!0i|v?HAXJouK_cRM8~tMA+<1
z&`k!b!ZLGt!iFy-0?%wyP>>0u0tXAW;&aXW7bVC1P~joD%1=)4?^mSmQn_J+^lvZp
zWPObTj;dQ5aYLgf0D!MyU;5D==zSXNcGw^L)d+4&c1PN|Me`Pi=LoB|_TUkS$4q3T
z1&5lE;O0FnxZ~wx7(NTnKce;bXdNp;_>)n?>)!j9lsV8W3c26Aj)&M~y6qL<3^cUT
zu@j2z99Rj&1lYAalxy9<s-*8&smbG9Hx24SyXoo|xA})*kl>o%K+7Fp5ImOlL|^vG
zm0$1%UiX|s&JIv|oqU9Lxhfy;cu?S+%f;pIGaT-UOM!Wz0M>{iR)JBJ51}6p#rj|%
zAPm`=&I_(R*+mRz0}HA^h<xPLCCG;J8k1u2O^(OL-4@uIxNdER-vxD3!&Bqe2%zSN
z1H8EJ?{;1y3Yjci7Y!S2I=OnHT?1F>Czev-?It5WE!zYJ8w*pLh4x+6B7SgzhRf=l
zItXE#ThBR`kqQGr(r$b)y~OhxYgp>h-Klm#sx;+W^B`z}DDEvu@FflIE0TsHWg%yk
zdwboQ8?89I&zH)_?Awat|0XPUZ&4gp;|uVfZ>xfe3U2er3A1wduJ#3O7Vn~S7)V@I
z_Jd~+yYYea@dm_pvS~=TRAr3<FFMv80c+9}{Vs!L`FH8(hQ5cd(MS<Tq5JT7okRf~
zrdQM_ZfSW|;(DFcc*jpN^T0nXO|SQj_<bqGzgS$4*a9-Y@wQ3M4s;{zMRisX9Q<P!
z?sQ2*N31daXoLUW4+@4>hhkJM<~U#MpMD}<o`&u3ilDl_x;dZe35y7cI-8{SR2(?;
zkZE{XpED40=&9jy;7bx>4u!bORiTr~`}1`u`c|GQe_Sdjey$!x!&Fzl>*OlzC`xQ7
zrrzlxkw5rggMr4iwh1^({|$vKyPUDSsG$dD$Z#T}Q9h~}0i9HDo~%_78c(Y$u&~Tq
zp0?||iWn@pMLDbEV4u_0hrz>Q@!o28u4)#yk(TGC+L2O9Rci6MhM=i%6%(b=T{Qn?
zn5Znc@Z{(T6R<}pxVbOS$PR=#m}FOk4QX`sQ8j(!vE3nt746!|JBxo{Am^?4yYURQ
z4Vy`;;ahX}Myn|!J_5@(A3Lf#ERXoc3Llx3U{TBm|Jp=TJGolPhNow~q9s(eON0ao
zmn>UaEH{xsMg59{Y;sjH2Yvi*^kq@p&+cr^jjeJB6<iXxd8%e1SZ$YsPkq{OwgC}J
zxlx@;+{JmU)VeGV5=vk0l?YjLmqBZwdm+<{6Uh$GkUq8%Z4Sn-p!Ysq)Q&O~U@}GZ
zdhHzFUT`(axoeB-0}5|Sx;@e+2{Tzh(U{|Z7Yd$<=*SGT13<gu>FVk7f;mN+(JZk|
zQ)AQd0mUJ+mjDS)G_6i%6%AnKs*zL~z~6yKK#AvqoM+v-KkkvVR(Y#VFJHvXPF#mg
z!7IwCQB&Jn3RoqHU3x$ARexhel%9DR2cvk<lB*-+FrRdFpl3#~JJb&L-2TSmq!1bT
zf%`EiXXL}E0*%@N$V=J<JF2gg9ZQ)oSU@a|YG~Lk;2h19@EMdCqd1{=y8Xw4-=qpk
z?%awCxGOBzJ{+kod;DDFXc)d9nxO&{z3N|a@l|b>xcDdo%LQnV5O?T(x2BI83hRmt
z8&w!&n)=l-b92ztx)_B<pcSKSLQFUnyH5|fqQ$+OyZO&E`cKdKrJzkKr;Q#AIJfnd
zm6?wp;{2Ki60MS3f4sf)4Sx2Ve*L;W7gQs0`oV_JaccMIF0h^Ivw7f*>`tFFcbj`K
z+O1r9J%~!onxwl%`#jQjsbOe3ddp)4<H7BI3S=~P({Xz$G8VgSRw2DKO3sn)S2`*~
zH&V~c)HHfM;|p?_#;NDN5VI~nHnnL<K&2hpkJMKR`Q;w_6=d<p5_U50-H8s7N-Pn?
zm0+Z4(*&j4k%n&nGc1N5=;f+1{jNp_cngp_$p(?8GGC%R3%w_|x?i~baeRi%h|9FM
zJC`F>-JEa6Q-V2nZhyT=_$$2>kwyE_s7O`DL;Iacm7t5yCa)_C)Q(SCNd%s(a0eO7
zZ>Uo$=&upn=Pd=+H*Sx)`5JZ#`OJcnyir{h3qdEW>51(U>7e<b#4PUn<7thRcf_IN
zFAF|~xEXL@2rko+Z{%1MO%RbqFyiwyl9$Q2bo?WME>=sG9r#N$XiL*AN@<nD;+|9K
zmfWJ`o>5R8x>j5%LkJI=!W)5v{hpoc^WeK3&@RpY#G7cM1zZ=7?)%7Z6DJ0>T-(m|
z@oF0q-k0zT`V%{Z)3&Gx??Ou_9KLGe%E##@jH2vmss`l>PBHVAEE7zrJ+YX7d>v%0
zcA%a2^~Yjv4!x+$#kl1a>dl;2EB=V$wmh+A3EU14`9(_J%!PK)y;!76ZS0F04XcA}
zYs;ihN7CO6PtP(nMT=BK#oT&+54BLX?Od0qD-Dw18=&W=wFK&|vaK*n)uNzV6qR1C
z=&22Vyo!!fvs%In;p7p7k2SCC!D5_Y&)xS3yS)k`k|><I>W1pTsr*29Q1#}t$6qjQ
z$n1-JQ{Y=mONdduZOL>dC+8tLg4X{EF6FHIuOQmD|5KC;D}XUO*LEsZpzCEOE2YYw
z87bc5wdtMP>MXb`fIb`kZEoJUwEa%Ep#_aUobF;mMsM;AZ5V^zAA$|ZJ5SNQ`?%y+
zE$<h7YAkvY+(<q`?BZir%&=Vge}V}&mr(ufr75xPGFV{Ni{+t%wk}vY4oEimE>em~
zQ6dYaoZ%l!Lyx=UJI!!lu!k7+?Oj;O7v&YOm+0glxW71VU+J|MfS8rj)+JfoGSvjy
zLFlYL9TNnmzKDwkrs56Kfsn%hPN2J9sEuAn!Ti%ELTe)pFR3~Z;zMnJ1KQpI2(&O5
z1+A={?W!_SmRj|?!Yz91aElgKmiGElz9f9Z*I5{sgyD?TAKEIKG4$n{4GnZ^ZVTAS
zc+)sloS=di8+$~lQe`O~MvtpJn%+pQj<#CjL4G?m=l3ycdqt_JW5b^t6QdZ`7}nKC
zUwt1{0v40Tk@Ut%kfm~htny}kg*e&AG#5j}BOC71cn-n}$*xxX;<Z1FbbKsOuWh$p
zMi`#<NWKg#oJ@DIF|;_X&D-wvuD@rs5OpN;xRPx)cC08Q-F-h+T!0Ro6gSOY^qPL|
zGoZ+cJC<TP0p1aGaQv3H;8%PoJW`wE9{(jy6uW%eozfO7wE|e9_J<m5$qb*Sd_Di0
zX*pi=&pGjPi!*9{Z}yG!kru!~M1gI3$%C4TexX9i3Yqe>dFDd@3$Jd9!zIRrg3X0u
z)weGrOzg$)C33%fqom5?oPVulw0RaapL+^Lm3hZQ$TPAvpiuj=+Ht(5RBKRd3uPwf
zmlZpXG5pOhuBMLG(6`j@RQ`7ty^`fH*++p_kv;uR<Vsr$NfpF#A8<u5A2h8u?|Ae7
zYRLF%Gn`14w{aQ0pUhp`@pLy-UYfmE{Cd=8I+F<3rQ)w1L0`Pj++e5K>R(QVG|rsy
zrPQnjX5J2adU@BWizZUMgG}KP%PwZf8Ef{&@nq)Y#L_ABz|ry|nC9*aEL$~SNnn&J
zy6b=V5_a(j4@~G!cwss^&aeeG$I`v)ZxQ1DUWm@&ud|vW%NKZ+(EBl69ygq570B>1
zqSkBI>iJOU?poY(Dp%2#po;7z+C*+Mz3{qTVN&>;$iHk+knoN_F);L1%Ptv90tus0
zZ6_13BU53t9IKJo>nM9HX+3Is8n*A@{Kflk8|?W(zLA>MlcN_mOqOKb(fa@5z>qyM
z@%`}U&R{Zwb<VZ$KwNi4Q6j=-<${SrvNM#N`TZ;5_Tq{1VuEs4)UCefP0&5zP=$l_
zT++X50Yv<DyeKqgcx(lR<#Z{^P@%bB$Tf2%nl)#VwC_@O1004$cK$@l9Nl00bmP;C
zP)eIxt(;#>QA4&p)6zY)-nY%V0P{?2;C{z15NN-+;IFqr5DnqKj`pdelvD+q+<jtt
zJk^_|SMO6%8JDOvs>^rT=ruTr7Rx^5UtWKCWjvh&KMAXdY!FO=W2)s1eTJGwW-%HR
znqWr}m8s1!N5hT|MbxQ>yK1;6in`y9mo{7qBD$7QJ>5^pGd45H{>EaPF>NCRl3g!}
zG_}%m4fnz5s9Rg+v4?-;@^`6+l&Zu_)^?tINjT|$h^4WKE&d|tt6t?}TH3fQoV>s(
z)B9<hAy(~*69Y%{zuk!h3QkqP3tF%Y(M8su0Q8sc=$JvGX9Pu5mrE&J&0C$*!n|YP
z|2xk~JRN4EYR$Y^EO>2dhvt-YH;IF*^D{%IULcX;VA6RO4iQF8g=zCILN317lb1<P
zDgW1rg{Uy={1gyM&K8Jyp~m{P%NCV+p5*fes`gF%P5NKrDE_~}iWU9awK^uog;!K@
zNdb}5nMEb#O(N$%w1WnxS|_S)%8A&u4P-CUt-zl(wCVn4>H2?ZCGzY4zC0hclZmoe
z53j}x<NnM(jvthjlhCPHxO;Ygr@t5D$X`ovA$_GI#k-w|{^#e~^Z)&y?J48PlHY)n
zdZpTaTm40s{(FJGbE64zDuaQ6lv4cPj>G@tb9|={i6mDkH@Z>VTjvL*6tWl?r$}+W
zuKq7;2h)E@#120DSNg-p1m0v#0IfQ^hN_;r!cCl@%3nj9{>QhnMIGYw=s64J|I1aN
zO0HdzZsxae$+oTflUXsZ%xE3Xp2$W~RMIHG!akAw2=*_>OJN!IGa7UDdHdczO+9tk
zS)DIc8!*LV?xLWa<MyZwBlY<wilQe|#z{+;+TBAfQ`eH_(1g{KF)|8A=V!dUV{S77
zlJBU!eW<%X(^SBhkXY!#2N^2K6b+&(5Q~A$v_n%obe)_8u5!mC!gkS!ls7KTU3E;o
z{>x?s2-(s$V_w#g@+JR<rPEkCP5kCaB|VRm!ryu(v93Yy<8-C(y|(=Bew`P*s6tt^
zRxQ_f)np|2NHeY6Nrh}jY$!eUl(m)WQ$IA3n5UYr_t^V6RTtA8{$iyu|0|@~_|25E
zjIQZ4+TaY2k(SNO8lSEW9C17M;fZKPe1_V)SweOX_Ns0gvbXE)^veaM;5Fa)5>l+_
zlH%wp%D&QxWfYLL(6IEl4P|9_+>j+?QG&$8#IiV~aYWqb=B!?QdXdHdOd=-j^I6p-
z*lyj>1g6pZMAOc-IVSN5p_mrO$&(Z)#1p0g)^t<2?uld<dB*A*MauvX9OC6Is{K02
zwuUDxAsce7L$jyOp@Z{ExK3NkP1YvOlNKRRTUVcXvMM;Sv<U3QQPotnD5j;Ohn#x^
zd78w%_+px7+&rgA9<;Ob?wR#hin_LPf3rGn!QT=>LJzQqiZD<>kL?lhIVjT5^Al~S
z53k_}GDc~<JE~9fvPElxEG|R*I$a*gt9R&$fb}W8Ei+WDGZhm|63JoUwI!dg1qMuE
zyq#&o+t(|i2x77eIBAnq0I8YgX%sA-yqE|*U)*6ue__5RPndHSW-we;3z>D_c!-Jk
zKa9O&bYxxIE!?r~j?uBxv2EK)1s!*6+jctc*tXdn+qP|e)z5jKbN;+{-`}sQF)E{K
zXYIAun%A83nhU@cX0DC9?Ehq1ZscX0Aflbo<4c9uC}TY~yfS0{EVQ1s<`nW|zZTKL
zNBafdHO<k>@C2C8z_r#uPb-f;VxP&~lC_mz-1dWc@m#W9_-n7LOh;4Ak2i%Q>$NtL
zqMc^Gq0=@am-%nSf;?#OHaM&+4Cz?Q_3j>s(ALJuK^csdP1PM<%UUj(sf~==gi*t^
z3k@_u89?{#BYZN7QJI9D%_;zXg0)>A(dIMTP2*9G5l*}MWgOW;oo((ATo&Fa<#}C?
zB~~~)&(P|kbxA&KB_W32S5?Xj6jd~_%3zubd67e@vUr2zTq}ZZd`?K--erz3s1l7$
z!Eh*4LLE+De{pFRUo_l_7S~7Vw90!-mhg~!Hl9<a31DL@{^7Y*rZ);gSy`;f97iae
zxp>9(B|zaAf=E5X%`e)%n^|2Km}|km-x{_HYze+=As8K@(_T>z5m@9Kb&|2u1tzQ=
z$Cl{OU7=S#$R71X=``4GI0|9+C`692MV~FhEsiI(eBra%Z)(`t34&SmVDt^d&|#~p
zsGwe`v|6BD=oCwXeCe^|N2~hSU0Rv9cu-t?>plZJZ|k4koMMD*=Hk+3TnrSf#G7v^
zZ-PLAT|LOS0t^TiV!|&+$&sJ-IRcXt?)8SeFWmyc_y=q<zHcNk=Ok8*YE{W^767Kv
z@+{V}EanKEV?y)Jh4gRpifMi`_d@bW1KS0bi$ogj1LG)P_J5bo-&ph6WU!U9_bK;=
zdc-$@9XGRGyRSroyr7FsWIS~SnujlFz>hj4cHt1Ru?j9qe4)}IVu#w`yo`3b=T8Bw
z+XT3UkMnd<1@;q%FQ$Bj0)J}eRfZ4n4CClWpB0b$3GqZh9BtcyA!(J?oR^2MGjVOb
zge<6p_a<P!EeW^&LAgwZ216^&k5=#0n1<aSq8LhvuGiX1>8~u|YXD-H191`VsUN9K
zCXlHkhmwym5^wBFP3avefNn0lLdnG#qfL;F%7JQj(?8stOEeSg>}c?!cn_9F<!rUe
zXvRgn|L9cvn(N0j13FdI!1^}+yLPF<dG;4bw%siS2uOGy+DHL3YjC6E{ZC-Zui2`j
z(4}D;<?mQYypiQac^OT%CgF&&$Hm0hptpGl!w7S;8^-=hwVAm!)9phS3*-LL>;>Fk
zLxw#b9TnC6{7Icq`&=&eynP`~P<{5n@=S9Q!p-M*hXJh?IhE%4=el;Tj>9qn-8T4y
z4G(R<!otiiVh1`GX6uzp+qc2PJDgS91xs^|!Ck+Xr86pp(o_^f@3t<g-jL&|`jo1U
zPYxq@LR(C{#13I$k@I<OO*LtajcChPzGb->F$iBu>pWzDaWb{<Ht!zqEC>icYFY3R
zT)#1RuEOaN>%+BhX*BX*9A6jpKplEMy;<k~sM?byTL^Hg@p6=M{=#{`W7+410uw!O
z&k+84t`U<*t1k9GbDf4ZeAS@CM^E)6s7S%E=dMPvXqCN_QAv(})hRaFw&O8;{;bRX
zclm-fja}BG*MNddFyRH*>t`%s`81QI>*taDR3yaNHZL8srPCn}NQ4<e9}&{qwiB1j
zcMY?Q(m00Gog|@S1X#r&tDS=-Mo`Pq%HDFOw=)Y~;QTFi$~dlgSHiyXJ}Qr)CXPgR
zz}H#fH*U;GY1AWLQJRPYJV?{qSbZ?kiiz!1WBN9zDYI18-TScg2in2kDueelENaCR
zZiK`V(@FGSVooCZfdH7HwvDi@iW)#f;VsVubGyUQ(RrppVHV_j0j)R_>{PIXcvsm{
z=*${wsDyP)A7pg$)$0iREV!cZ^xscxnlQ=h_v?D$1EJ~jU4>3CwCrpR_;vYLVrq#+
z6n0a6U8kejlp=GIB+-K62sC~O5ao0W)%ZOvGvR-8TTD*IHbN$)EL51q4uOziI6fi7
z4!V<wII2}6MDC`V-XyH}N-3d$oo4J+ZMH?D3ibQv03-xlVGkW?A()dq3By9u%=I9%
z(Lz)0k;sa9l#pL|nR$|MSrSZrOTs)uHGyQLAiN*OIJhX$tRPC*oAEkgeT;s1-S_(6
zH9Aw?mMDQ1)0m`3&WfuzAqrww;Uk(9uuNXMx~8bKl&ib7oAvlK+wz<Tf|t4tp%!qz
z({EOL9NZJLz+o*E{L<_j+PX;+D?p$xb>j%9?6`iNTwGRnY(NCXK^{u2eT{qQFRrq4
znrSHs-u>)D*;s)`_@zpSPlUGdRBUvE(&(aHd{oNtCkj%!yBW*eoiqHp+O}v3!E+It
zh&0jf42c>#%UdWNABC>P&WYq#!xzQ{rU9o~?2_^IgU&E1>=L;MRcTh3$p+<)PGh!G
zR)NYl^@(<1EA^_A;X|nX2>eQ?5`*>YApjn%dc`@U%@^KMI(LLwh^TUIpcUiGf_not
zRsF}_{)mg!6DFx_b#sv=ZK0&6*J`5LS4LZ;t^5M~c_LfZxhWITL~QV3bX|QxUkRbz
zO-Ap?%2LMRY=4Cwh>FZJb9OCc4GsDm%?CAfP{zp#<AKm<XZZfUFX(~6z$`gig{*)V
z8OItE2*mx6%gJRJMX+A*p_8;?!byvtINLpBzrX2)lc7C^j%b!n0s``2*a0n>UdYt3
zV#G4_54{MOB9K~Rw)hpQsZC`RXZtSL^SB+hNumqtK+{{M-~?DXA0kQ~z73XnnnX_k
z05S;SL=66<g3*7^(M^;Fo!my+GLEU46jwA8o#AlF*Gc(wvvmv?YOFnoSIsOp2-+zT
zX4fzet{nW>BrjE@DW=@vr_Zcb_T+Sd&kS;E_gV@@#b8AQQ)6cNqUre$w%Iw@+;6Vv
z$rO}BS`b?g!@=H(5ni8|iZI;7LJTm#)U|HhTMt&b9V#6t7FgQXI9Xp4%GOw|l1rmg
zeLd87>k$Aq!v<X)8*?Wtmi=q$c3)774=mNgbb&gl>Tz1;12hb2-Db6&etXJFk`SHP
z?+Wo1b|7-r|Fv*Uo(Hu$$Q9hH1q|5y6EdZ4X=rRQ+JopEcxT>7yVFTUHYh>*97pF!
zIJV!3@ta7yvxQ3`J4(DDWkPiwI3c9!Z0fK7Y<IOVz?jd6yzOo64pCK8s=eKvM)Ms=
z7B)YWn-cUBe+aM&SVr6IKnO;tDf3>YG+zuzKFa((0y@cbWQU2?7JqGOa;+%Fh8|E7
zJ`pTcx<dTQ46!D7?Ge;)5*f#DNzsY`AE9#J96*KaCIA-fw*FKif*4td2-IM9-PB5y
z*HPNUC|ZEFCjV7OI0(Eq)TuAX4VIKPL^=FRqh9%iZwA%9iIr{tE$w0B_5gHXoXx1R
zt3f0J=G?TuD=xB)8NLuoUxA&5$>g<v3B*W?4_a{mePioZGIeyL5_awYbQ7*HuC^xs
z3WzKj!8e(hRHqPDF*qz}?0HqJ>u7l4M3C&tC*31!LzwO3?K%xC-3VZq(Q9ly1av?8
zy+f%RZNxe0FJJmb9na+RCp=mRmdni#8;ulrfx}$aIo2<TQ|<$rCDq^UTIYo<LOu<M
z^cdT__K%(fDWEC-5Y@H?d;R5fzUv-b<+K1?9ki@h;A!$9E=O}!I6=>K<CK98vK^hH
z__8-dwTXpFsXDbMbI^8ipecNM3i%FeJE_?qG8(bZ#Zr4dq`mk>;mfU)rfkhi6(&Q%
zxVRV4zj;lrgoM<5=WK9=y*2VQuRedMm3oGKX*rxk(HyPs#2y7d@i@)8z{JYPV@7D-
z2VhVgNt$P&QO`IZfE)7{VC2pguDKCp8?Yf1eK){$_2x7*Mt&)60!B@oEU_xZdTGG>
z8f0m`(?9U_-XsTTS71k82UMlZP#iG#H{EVt4Wq+vtIP35wO0hjEuZtc@LVrWVEnAb
ztf1Y~H2c)E&_)8?sb}S*{{56)x(No!R@+K*Zy#xAu;+!NA3=*0_#N7Y{b59rxYMH=
zMESM_CROs-ry<+|u*7VP+*$4EoC47x4q9Ip6@3zgYkuN@<)VLy4_%lI<j_`IG=M7R
zU=$>lh!*~8F0XB6s|fpkSyAn$@nv*KgBWI3HUfeiQkXD*gl5)+PQcb``U^VL2UYSU
zHL2tXMkq46P-=EBY(OQuyk=cB-Z;DPX-*kEq<+z3PtIk%jb^=d_7ZpmTxF4OcmE@S
zOyyA|RXirLK_&=IxJ`3K-W^1Rv^z>hz{#TYe$gT$xFE@|PGUv)+FtN}oQ&_KlPLYb
zZj~eiIx(WiSY2H2A>vO=YH(mfwK7=|p7N3cfs+2Lbf$~y`xikba3grDk*RdFSdmBG
zk>~wpRDB{YddA=AzBl`Z!(KDbOMqT+;P`sO&(r9dZ6JfrGdK9sy$m`VO=w3DhU<E9
z@a4(ud7?-+u`}HcxSZjcthl}%o@gU$8CQOLoJ1&XuGYGMQ9V`aQ^oLHC>DWog})}{
ziZ+kx`&|ap)z;==n;DwHL3?X)r*p)=qn5WpUB?58XLQ)h+l-x?L0?Dhlu1{`zTAoq
z>Per-h)RT&bSIj#82q>t6Cvah9pUez*o4&T`n95=>ru*QP3<_E=6=7+bR<WUxvOdR
zn7UQBv2*qsgyXcJF%S8;-Im>8$uDKs5M06;Z0Vg-`@1gmN(fIO`4#BKT&t6>@0qD*
zI3hSqX18=ZZt7?0?GwS~W&?mGFtuv5RZ16Ym8i0!K9GNDyZ|-4^pJe_C@{|b^C<Pt
znehP`Vz>Np;VrA7%w#cLE8V0aqtq$tJXEpQo90R`E%B&a#@-ZnEUdifbhW9%z%^28
zEkiI&J=pyB_c7<3&^0<!WFlCeJ$N$PP%q>IA#^Zp!&MEw4QsJW_&qpM#{y9}1l&oy
zgIs)xAtL^uSoADu%E+Hh#%x-VKL^6=g9Z~B{Xskik!q4DkU2EgcP|yRWO{PEVBhk5
z<3*m>41L!j?S_bWjr3$l)FA1k4)fqh-;otJ)!?>+oK6N13Ohq<KvoUYgg{IHuC%<1
z+4RT(@6Ll-RSK^Lc*t5Az*7=>NI>uM<*k9?!~@W15TS+|){Jscx?YNjhkh@#&vHfc
z0uO33I~sCu;Y`g(TA?{dxNPtmYHO+TLpuhsxAio1kit!Nd_;`4Stj%%yfXFJhIt&q
zgI2gW<JyxZbn3xdN|6BA3N~M|gml9e7R0!KJSgs8Z~e~90$e+Z)o+fZgLks7ou~ag
zV52NG?~!3J{oq=Y)nis*MLHqmsfN%ZM+iiSf*QF3iP*lD^sU^&GHH~7am86k4g-&d
z_%bcy#2mltU)T=`HyWx4dxzVp3<C)(8J*<4c$AMGdP{^C@y{F2g+3m8L-}rL6SWMV
zt~)GSjf$$L!zcDqE&}6v;so2}8%+XHTBoJ9b)uLt6|U$qakVcqz4ZD4b_KswpO;mS
z?iaM5-Z?eDTaU~W@L{QGVfUb6oa;ZTk8PeTah312z$jpE^|`~L-;D|*cJko#3xAQ#
z4#48%%emBSU++Am+cTy$xn$7ieVRX5$4%1BxF8;^%G5Vz;sA&%yCW}GlmY93(}MX1
z>6vGWZU@rADp&Udc}uHGqu;!zaw`VM;tkOu?oH6pM#{srn+@SSnT$%q_jidDiz)6F
z+PsZlPE*HS<dGotDQ?`K()sjP`tAiAqSNkdJ5VFpnLz!TETuLPEOApXESY-W8J)H*
z9pH(V-QHmkAysQP486v_{_Ms<>~6Tf-QQC-u&ZO%rh4d<f+6L}(qf{-$8-ICcjv)0
zk*9dxZMEw9mfBC`4T?!hihcOa!oQ4`dw|Fdb)qk=f69#-Tq2Ua7ibvIyxfjJae~E<
zlobS4jh>_ZTy55PD6qI;J@ptuEVK$01JJssGD$;cl#v20wcBtIesa(Vjy-dT#6~~z
zpiqr`7^lXtDugMyCaD9H5h{>bC|ws#kUvqTGwOw@Vi-{Mh(w^I05KUzpb5DG!J3QR
ziX=3qxpV?%C~S4AuQ$q&2QdWn{EDBmB6gUt$Wbs&sbMH$c^Q6XJup?M#|qM3K2j08
zhr%gYRtq|JgmLi3KnFD(2aU2W3X{2dG$#rEM#&iXT<4q`p?fd#hoYg2##jHw5kq#;
zvYGrm0IN})qQ(|k<u7B9ir7AWuVDz;i}E6Rf&?q`s|ICedgYrU2MxvuCPT}-Jp`x*
z&2P)#>RxLrVC!JWSm=_oo|pIR{(TvL)OLFI*{UHgkn4GKM(4iZbKR=BJ^eqSy(=W)
zM==bM=-OS^3;ZB140adcKL+i*qy>F#0w7>!RXuf3z1pjuDbQr$BAys8ml`rg?^oz(
zO8`xT(#y@EQT_|jN|m}j%;CkI$C_}~mbp)*>weTK!yeiNz+-H^cH8z79b(vIH{KWC
zN~$uSwN@&4k@)IQ&v`u`GEGu)LH{qA`cH2@M|@%p2DI}(5zeRzW9N25XDlAliBU#=
zW#F!?xlh)j`QP0X|4+kEVD{TJV$T3XGECo>Ya87F4->)ZVEf;h`2VucfAj>lKb`9a
zFQvqh&0M50Oe7H~A!`4YA(V;>z7O1D;6GCt8r$IQ9vA!z1I+>I01Q3ObOvIm=)l7N
zcg_E6Y7!$mFrP^A@~*}1XIaX>{Ou<odrlU4DDr<P^8fvfAE)qt5{Eb)s>|~K<trzl
zdnJDVltKIWZ(h)v<sUISxegA)=qH~0vf4i=D*tsSe~uAQ_#c9OBr@7qn~$e?pj@xR
zM(o`l^AxC;=bHA^18H-aUYT@Ev;;;aL>23Qy|<GbOk$P#ABlk|_|^Y%`9z?5D9~Dw
z#x&Nzyo(MxYiFl~l#<Vwq?N7dZPR~xchW{GJC@4VzHTo6{I`&+PysL$N)>Fw6s1Rg
zMilm%eexMJ%Z830%DuyTgkE)LxBLY9CiGqB&M;)j!SDZ;*tLU9DTWPmGWP>r3=bNH
z?B6HCBsS1QM$28f$-74@=8rpy;rA0pZdR!tSE&>ZRCbkEg6b@>b&=2_@SkU&fDv*m
zBdQ<cw)348iiDOnz;vpepzfh^CF2ZhjJ-n_zp$uot-f(khJ?vnY><5PnJ&O|?wM}C
zCdrbK-<X^t0WqKugJ=>pBF*~XA=S%erRTKWIQQt}`0Lst#jF^l_~2OgWwG0zxQ0Ih
zr0iRtDKn?b{m{p^3+_$hlHaEnw~pyN4C2f@wcue=V&Ld;M8(a)tXs$F@0FYpAWjHv
zzr)*3j*r7(kU=E#h=?@p$C321u*fGI)1R^(+P-%kAD12<pC5B<cy+O>I-LwEH<f;0
zQ$Bdvo(f&IJSbol;MLxpVX#p`n+Z!#Vt{}H_4g$rf`<Nw7XdF6WB{1}pkJp|?TbY7
zq<8PggOWIYbBGpK7h%M?rl|i>%1p9F&_OkkPoHi~X&a%Jw;bIt_ZEsH9!#9@!@6e2
ze$q=Be)6b#F6u%c_#DtYS3>jMOMv(J)t!|E3OXL&Z$y2=#dLD=V5IHxV4-v<OFxA_
z8DCHU85tZL<ey&5tiel(#R|X8P_yt=&R1=qQ$n{@X%%EOUl8c-TO+M!FtiyNvoU$x
zf`lB0b|8-1+<c>1H2jkL8JV<@szNB%w&txl;6ddZo=5EG(-XF_g_RN!sfrbz^57wI
z)}tD$9$&3aGX<w$A&4Va+Y>PPvTPkE8qA%6bu|<Y3=vw0`3twVDT{|B>3ZgVbir$*
zzInw4^OfRsi1~>9yAl(H`#|m{{RA9jwFmyEyyv62CP7(d>sq!k=afCThV3AP<!TcV
z>)_+!>F(Oj*Q|wux|v!Hk|ZUn;*3#^NYTFV->4oEKc_7|8~yNm0GU&gvHSwj_2#oc
zwLXIS&x7)x0n6b<E)K>9_TDT%wQdkkj)KEr`ES_&Mmct+6fJ*_EfG=Dy;^{p#gE}B
zlSHx9N788scpjRGR=TkwgPnOjyv~@zN8-1XrGLR7s$K~fJ<gus6r}T&WU+L$Ig4%G
z4Q`udVX#&r<Hv3emABt)S0^k~JJT>zIu5$am_zb@!LoiH)siE9kW}z|2{<VBTc=-(
z)c-iLaT+^AQeXYa>#R}E_c6Y3^hzd7-*$9yd>@iEw?6<cR^>MEek@CYuck6Y_;DZm
zJQk&inanY5g@?iQ9W=Q{LZb+n|6rql#B-;ls2BFPsR8`VZ(c+oH10FC*r~9;{-e*!
zN2NOMRf7);xHqxSd)iPb>BB}v4yRGV%yC9nvs_2Bp|oK8ok39n+-%CagL|C)?w|*O
zI)B%`v!&7!1DfB{8?VM8e?=&IZdmz+j(tFSs{Oso6h=gGO#2a8!^8Lptc9n|TI@6V
zsG#E84E$COuZs$y+fLbaX{1G1T*u_bMEe_StECIy^^fP65gY2`LP9`)`VoGWR}j}$
zY=_fq*W13>-2;W1w){xvUmnIk4{ALlsCQT#!ph*RlE!Hblw#x}%9Y(JE%>irAs1zH
zB=h7m--ca$?kMgyX9~JvKe}<{6hbcAB3Rp=v&(Fh**6}H1%Kc@iW8w9D-d$LfIA)X
zabp=vk<A?YXX~9wsxkG258TXR<zx{_0uUC~u0o$N5^>r)qu>>b8TnqK7c7Dz%*7k}
zhrQeA)w5ksN?Mz<_1@T>W;U`JMGu<xP)SvyJTF(#oZbobp0UpODfBfG&L#st+Z8M^
zrbAlZvf41a{^b1oii7g!AIytvfhHDPjlDwYg0Vt-!}=;DSZk0Y?sbzrp{STTT^lT(
ztqs*M+%QkM@RoqMBHnQ~OM2&)PV2r|f;A_avi{SHyV;c5&YwD&aT&b_Z~;F(oh?Q^
zHn5J0<51KhG5T7lFPW+)b;JMLJV@ZKk7C1JDh~DFd?;&S@mcfsy&E?g%CNyw&?T%7
z>-`3{!N#nas51+SDRt7FWmGfMc32N$PE9%zuX;%8uI5H=-ozX3VTKg(UEuc@vEtCy
z%~ux3H2&6C^3)KI%AB$J0OZ#*+xGa`@Zh?4NwaC1q+ijzM&l#tLlb;M_Oz%VM^hz3
zeY=iIU+frCYesK|@Y_?kj;cH6wx8a@cX5V(F;sT+-4PgTyM3>&xJ66M`n5@rG4~)*
zaCSR`B}w{^i}#=T@)ZqKC~|fxF?l8EJGh|z`mc~O<><w#V#&Q!gk{U3g5z5wL=|l@
z^4X4<u1`>Nme=EzvOPMUUhH_GHxbT7m(zn4$4EXM`f(47-%|EkW!uU%irw1);c-UF
zGvUT#d+ZxU!2)EQ5b;!P6*}e7oVG{}u-018U3j;dz1<KN+b_EjxU*vlsO`;$mh0>x
zR42_T7bSX)D^d=JEgitsWQ!8-nu<e={~$2sU`&wH<Qdo$p?bm|!;xtp^-@$9O~Y(G
z*uU{{7j@i8@s{Smm0DP8pfo~I*>$6MH*#_@D*n)$J;*KZVe-CcegHHJDqDs$9x0Q-
z>BXF!VN|xB3EsW7jy;jxt}so`6gY<v#?I(HF*ne77A^Lr%dX72a~L!Qm)UqJO!0?j
z+jgJR%y}51Mu&&N88VO7FQhl=0F05}9XfB%XYM#irUcM*=ezAXEQ-Dyck?={v?F06
z-U}6=sgluPOBhTY@xD0mSFThb@l*`z%H(K{?C}lFUtuS_oYxA`X-;N8I59J2DrnnP
z@!pPpcH(}0Z2uh25bXxALU8_e#nVOOIoW>gyEa#fFQZM$MaYzY)b5lKa`y<}mh6`J
z`+~QO141m8?u*R?R;~vwN&8ui|8dbRT&}vpPe0FRbEPYgA$${4tS>Y3I4>r<2rH_>
zDeW!pit6r)@Kx@tX$AI<kJhFQg|ckxVt3@j%jDhCEujkv{P#+Z7AHJ;N3+HmMloM*
zTGhtu69KFBRnwL8V*&@$fqX9w|AZH?9hI^%NCf`PH#X4@PX8&ptUkf9>E_;*i?r+V
z?HkSgug_)#M+yAxD%SX!)naYu4sq`%AC78McP8%>!2Ra6m)$GP^CSPeJzWSz=X;&y
zELz#Z$yWw2kSm7ozo=POnZ1xq7tcfNhr#0+Ha>=Nb*suGzioJ9b*d0t8jg?cvh4b1
z<Q3`w0{2>zBtds|iwFxc>odL<;+Jm_3^}wneSld|KYQ50jk=r)md$2tJ!U$k3N)aP
z6#VW3FXiUAFaAVQQo}GBmd5k(zQS*r2TVSz>bgrLWu*d~BgGS|7&>jQe-(+QN+H8|
zTeQN#VOOlh_zN;Y246<}b2mb<nexl#!C@pr<zM@#3R_?AgYLBJTM9fyK5^@u77)f~
zrqHQSw0kmOG(B}1U~IJPy5ChU;8#2*b*ex5eB5hS%rlx7Q$HPM$SZlAR`-h>omt%O
z+=paO=^rLsR+pg{w^ygXl9brrf_NC%xrt5*F~RNTeU(L@5}Dyu7`%*dpO>ZBZll%z
z89k`}!W2PbdUM&_dX+U)FXgMh^Mar<IvJVlWta)KL1tX+eTCNS`+kO$m5+&l$pw)|
zeIWgD7t)>=KsFoqa#ku?BK)v45Vu={N=jyfxM&0=noSVWqmqI~r6Lmfsi>;`0luPr
zlcYWK*3jn=oe$=E=rjHM+n5%Waa5UU7b>8?qSHa1M<mmE76hV*Q_RM1c^kOzD|!(H
zjB~vkG<a?q2!5u{>IAo+u$F1}tja_m%ZZcu>bLOj;?s$BoIvG-K6q#cb)m_~SNyXT
zG`!3RQKw)X(*FXY*vS>4n11UnHyb_K5_C@8r}+-Vy0eOcV2*r25AI7$5>d-T7}hM3
z_szG8ZRRIqn!m1<^=QDY2SqDT6B3T$UIC7fls6)|B<`T@J>k2@<+^7QK51RpQa%!g
ziukFMt#byycF#E4!&<(OtVCpL<!m?+Clp*;RUO@F?Klty>rG8LR`iyLpI3DSH>L3V
zi|VIqyQ7#k5Co?Y#Z$h5t`qQGcjzCeqLCr3`vq<82sUbfjd8@8<>HV+mXh#2#1xq;
zi_bfD4FGQD7+;JLzRKQWrefEjq2V$<@_Sre=!S(g{qPLrdJ6S&Qv90Y2$D#$mp%#D
zU%>bJH0pp4u#R8%*$%opKKbyFRcu!We`&WE66y^-GM9=VZz%VZ@{b9iT=|w_GH=%L
z{fpSBBq{?CvPd3xt8%9nzs}1=?#des$=WGZ#GqK>+q?*(0#9(1pD~!f1UiwAg&y~F
zsl)vC8~EPt7XY-H${-&nz8|_lk1P;L`1+m;M;3eAZcp%Jy};+0zx3OJsVX@cgKOq}
zFyUu33F^4hyw9Qi+oqJfGQe1xB-RqsWkw#h5<WMe_84Bx-g&GbYTqA7oErYpk3_GU
zmS&|d({2UvkEOsLyyq;~&A_xhzoCUp`lk#S)?^$t2lV?Ad=<Q|xzVfWdS&0sIBRg>
zDffmT{M-QB*7rJlpff7ZTW6Z^yr8dFHHUqRbUk-7p)!JL^KJVuVmCtL6I*nVi{Auz
zdTp(@?nKpn3*&e89>>8NsFJ#9?&cirm1>Ds&RG2Ust3)Bg>{$0`zr%k83+CB$E}O{
z+Mdtf?<?R&d|p?CkJ;Hxp{h79Twbyk2=nJC?r(uYO8N7Y3HwSuZ%ig$HA|oDW!?2m
z*@s$U%6^7~&QbsxQ}5SXR{Q%ah9-)IM{QYiO|*JArt9$q{DLIh`A%K?H{Ih_Zo*&U
ziZ~8iP~-D~`h2L|NWDDRvZ_m9lI0WrqqWzRUoAxoVv}-++#ZmM@DyV+y>*MJYoK0S
z-H!*NjxC9_T(dR6l0-M2H*v2E4{dx7HSdflone(V9fto7K_{1?p_!)p!%<!)VZ(`X
zmCf7SsOC*ZuNO}cj#KsfX3FE|tYWBMw1L2y_j*;!p?KZvdFmMZQG1hweoyLlLO?VO
z=E8TzHhsSqGQIkFbRY%m5oyaUeffEn|N6x1q_)nVv!biGEhoMP(wkF0uaBTpx_4lk
zyU%^Sm_=gQb6-5A1VLM@qR8vQXIHjL3F@LnS<(Gj6u9x&(1bsPGd?cOtlnUclD(Zz
z+?(x*5HQ8zG6@C2Co$hMVxg1XX7n-%6<w?m32^1+<eE2n7@CfOr&GNgKYc@wvm)a@
zJQyI5;{@<K5zI=>Uji|Cna|5u)I4P>QveQ6s&*Y@di2YykDg)F^b4?~(-L;}3z`Cp
z;*;@%s&In##?xhs7K+n{GD8WCp+$uGmwzg&|4+$%1`Jdo6%(70oFg3PVzY5p*0Kqx
z)biie=Xb2s;;EYa06!^hkR)XX1d@9n`ew=v%I4^UKbvX59ZVzfw-t6rs)CaTbs|#i
zN$)>o?7B+>g1{qqY4HwKHw4_%X=;P{H*T7rH!N+M#LQ>3H(Wn7-Nfhj$*@X}W|!X^
zc<o+4<7Lha4ev0z-N~jZ=Q*56^?uv9(rTcPVO;7%VC=Z{VeC<nb6v62%hhu_To=_O
z@bJ!T1=i@818_`A;F7<fsUyuwE8x5xT;-=9B4NlrBtW%fjjr%yb-hksx~P<lfXH9A
zM6-|XfxqDW<+pZ);-z_=g4Yvo6+<ZL)D>0gaJpO}>^A>hqp~4E#U$C<F@pB&$NQDD
zb=C-1vbh{ENSKOJ0`ac}uOFLmj%sL}c4NoS{U0$XRW)K=pc6gWVqdn}2)&KsCJ>T^
zS+u?u?O{TCv|+qSKO)>S%&djRiPva+(9zDn6rk-JGCk{W>HL73ncu9B9ld$G)EBA^
zkgKWKB7S6fLL7BjI`P`j94b*i?aE*+A-Cppnkuju85R+2pQ}faRhd#zEg;qa>`EY$
zzM!)P?T)eIy(U2eheuTv#ioB46PfhmVnZNu5pvB^AEaPfGDzwRmwngcgAbh<l&p<n
zt}y~$LU#XRJ$H4{FS=-X4<eB3Pl?2MWw%j)FJx!tdm4YILBzqe{IDfUB3sQRLQ#0%
zONm8BKXra0_jcB5?Z(@0O4)GCZrx6Pxf#00XD<lj&E)hRg5VEw2eg5GV+ZVh*ZA~a
zV-2#HxK&@^a9`ElbXn$Bn4e(5Mco92N;%X^t4{lwisKG%+6N+f`V1C{tEsSs#9Q<`
zf?R(bQ@$w0;;pcd4_Q)hOy?1q(_|Jmrisf9oKoT>>XxZg#C}$>XAKmw*lFk^cj2#c
zz0?^kj>R%FKpF0FoCHVtHnyeXBj4nTszU6a=W`aA+F0seAHxF&G5di*orAdK@$D|x
zL7rCPVN^QfmV-XDzXy>~_W^>#v5BUj{OHi^tr03)(I=>djXTa<mg42_MI|@!rf6h2
zkl%5KYP(4;I(^IPDZ1au@3WW*CVnX?jjkA9=(61@usBgsT63AeqY(2a5*bx}B9l2i
zwPdO3b68k?uVyPUnO<sgbnXO9@yubwU~+^t7VV^*pfS?iPF;Mo`a0T@0x8=IwBv6+
zw9e2r3=4QOtLSqWbC6t)AE1fsIFFDukfi{#%3(PI-Y)OD)=diaJ^b#OM6ApP2fNol
zwI`qSsnjRSmV)~Yk3pv6Vz7HkDd(5IiP8v$t`^F5n(eo5gjLat-xu=8bUuBA1$3bj
zi-qUdN99zd>l#iU0|~9!!cxp`l??qaNw`eY8)zZ&@ho!&`YxU8HNVWyV)~^TIZFh*
zA5uK#RGauxC%&z5-VMLyh;YhG#e}No;8fyXc+0$PB}0h_ZbKygK>G{IHG=~`?GK%U
z5HCTD$HoC;Vpb6b>^4GO%#A(H656C^6xREbiHt&cy;$n!%JKYM{OH$>;Va6N$Z(Wf
zv56%a>N|RBp&)?$wm(2Mk7%v@r@*rS9_a_Qd)`01LnH?4<%OQo&TIOyrdAsB(f4bG
z)ZW7pTXrvaquP6<`gy&|27D@x>bDS!%6E^kIEbs;UzY_Rol3mfEQY2|sb_nR?K#LX
z0Skl<Ib1J~DOd%z|7TiW;UJ59a6Z!5-H?*@heGhg=$6K;fAS%sz;BGnYoWE2Ae*^l
zWiCB5$TYgLMOJpcv9{@l9?6Y~W8ZvEaAoonnXh_xY`U)z{vi1SmJ|y6N6j1DTy?yj
zDAZf2lQiaO^c&mUZ_1EpETlV;M-Ww2HeT+sbw=vQl63!x$^1`cmt%rbjYhiXYz(l{
zsL<~!S+&YKst89Bj(oLt9a|+4vk(l$o;~6Z@cw$b1wl#Dd714~EzV7#|L9lO;NbZ3
zOJBz$<Qc7##|2gph1BD$UBzqvY#=;N`s3QgX|Do_L-p+QLfC+H8z3;OCFf1{Hf}$z
zd<zI66jOA3Qlg)G+w_W@^;@vL#Run#K6b824i@G;z3(4ZtRbMYmBVK_$TIjocKA6^
zIpp^2nO;{uULhM6_kU!Wn{!1Lrko0fqHvRp4}HUPh~+wY42Y29d%TrC##&e_VcUc5
zvG<Cs7Kk1If>rR)AzrxA<|sLtva1sF23I{F5S%(|E2Xl;EO$F2MUPXIuXj$<hA$Qy
zoX$6T_w|3cQ?@4Z@x`;@;hMkjVghYM3Uo}&%&isl;c>XZdeFhxHBx{-TrkG+j_bz4
z^?|-;%0+z=*{=HjMzFk}RA%W*MM9xZ=gpQw#OTS-8Js&mpAPWS2;*OQtFML%boLtw
zW%xHxqLIx{*e&nl;&TjDT{AT{O%mw>wJ+FA<{ww}>K0CsocqaqIDBBJfry)3QY=~u
z$T~>1yD6Nt{ppRXZ*GkFw^0w-6~(`6g6-c@M#WbStvYGcG{%Sf6Q5^AQ_NWd8a&V2
zP)HqT<wDNxZJ(#0gg6z;ZLXBY+pQjwj`i1-%5UwIYZgz+_fy~b*1;2$?k|VBcz=33
z?C{>34eb!<q<r_f{6Vre&^QGHjOo_THGBRYC>4sbU+|0OIW0FPmEe)SCIqKAncT22
zx)Wf=7qL_Gqr1jwSdY7GN1*ePueMB2MpJKL>StWVUq;uLXtESqMLKd;nh6z~3DBW1
z1|*askZbPWb~JL<3E}4MFM$n8qAw%WLn|@-&`^^hy(sm(uO|#u>PWnGg9H^Z0VHA)
zM4q^Cv$to42uOXJtYHQApi!7(t;UN^10<pc$M*w9@en)@gX;LD%?fV?AFWD?)@UL*
zU^ZOajZh3aXXfS>r}_pKR&9cXR^b`%??3-A>PrbHJ2QrZ5vD&*w_X^_90}3E?>Bc(
z{LR}jXSFl2mk<_SBQE&X5QAFx-iu~^^xZ54d1P(8?p|4DYl8wHuJj7)Q*Ak?yKbDs
zOH{e(73GDO`f87OGEG+{LK#1#IhA!GAlmR^@mB1PJgw~`+$#Kx9?N$S)l{gLykd1+
zpb{0?jgu7Td1@K{UQ~jCRtQC|GS#wS4}{+3v8H32gA)UuCk7_vyR8y<d7ic*@9wg7
z^RWPD2jxRiWMW*x#CLZ-Ty<tsZrFuEd~f+9_!ubjjWa=D^J&tMg``ZiLx|F_(@G}Q
z3*#f`Sig_H?}DZlNv+^DV;NKK`2HLm8uxIDW4X{K=-z%IgvHue3-`jpoR0}Ivb8jL
z>4g8FEMvE-5j%L0=kwIROID5^Cd|O6e{`^0Cs;)nPC_PCu|l5F3MN7)g*vSX=w!Qh
z*7LscIu7qz*GHD+C)#-J#tonSqSrr+7;#`J6w$wdJ^Ib-eyB?wZzc({@!=2u0>oNL
z8Pp-B3&pQ)Yl=|lDh@238Ltzn<78TbLY#l+vb2!>)QH8Pt%ruK0QZMrh;xh=z+zjQ
zet*nBibgtA2Pr{Qc#n<0`c?<|4SE*l>xvN)macn39;tH`GA+|J_>ZcOv9i`YhXbJs
zyJZ@EHFTe=i|=Yp{I^z$?X25#L?CR|xXMU8s^x*zg*)$xP^Y=;4ueb(#}N&}eHJ|e
zG2!B*&t>l3x`Zk@TzmQ7`6T0rY9OQ27~U5M9os6d?!D3_+dfGJ>1E&)cKNS9Qen21
z*xJF0cPl==Oz?Fc&i<kQ-cOL4(EA@A982H~OGnOqg2LKE<2nS%y`Imis3;?#{-6oH
zW(F=40}<r@P~5abk;q9!uDLJHU&NBJpR7@?Bob`Ms2`S0TF&pp9;u(;B|6bZS5#4p
z3!j@}I;p-#k3<~iwC;s`*z5Tnu|6N?P7uKS3;rRQjO-tN#dlH4AQ<r$u=!E7QF~Tk
z5Q=glTHXZsi35#d`Eg|MDt*DJOZWM-PRx>GTW8bxn$8eaO(@dV@KyOcw`yEs4P-nc
znbB_;=}nE-kqbAkIDSNW%9+_YwMVaNsCkb?zBaQDgo(<vpY*R!4fkZCLtjNjC{6<&
z!#BX~;ZB7}($uos50!>mWtOEHPgf2$+|o%v3=zoey`g`^m(;TYCB^eo@D{rtT>Ylo
zks{*@-uL3HHdlUe6NxLDoo2|?KkC9Uua8o8{j+tFD2$sYeHsILyAvM>Q`UZ-d-kWP
z<VzMYbiLdy|7w<;I7=vFKeSYo{<iV0*LXH<h`_i)?P9}q!sz&5VO0_R0==d${^{8A
zlEw(BiT|~{_3=96W(x1yL4b_Mz=-q2P#Qk9!22Fl7fow$JJykYVHZVZ!n7l%7Xoja
zX@b;Zr3>V7w&WwowuTRgdGJtCcqry`+S7T+Myfi&)v^!jyGgmXf=tHZy3dEtcQv)5
z)31`^Gmd{3@BExaayY}{@C74z%D341p^-opBQ+QR9iefiq$1_mlNUoFIe41k1Fn{v
zDaQ*Ncg3ww7xLXzJ%P3KS_$(+OaOh~Z{F<s*g*PzMlz&SIc9Z3D{6fs5MrG{06~xt
zb}uj?W>G!!;kYj%G@13XM_@+qcB<Ps#L(UH>XpxFJNKnnz0->Q5&lT!q1){asOEjs
zpX<PJ`r*FoEoy!dauId$u|Jq6dXyK4Au>rg^fJzaSQ;q=wd&&oXn5Xpaw(<D-*Iy=
zXeV%l^Z_ajeHSD2>Ge#i_E)bPQxhp{Hmr)#D~u{&`mP5jo3S2(Q*QeOKJmYf0s(D6
zRLh<OS?S309cf|Xskn9bnH&GDu|Eg!I36-VN8(uwXu&LvTOsfiLxa@GDe~Vg!D3&I
zs5(qFx)8>$lIzsY$~)pfHfLcOm-=iOqyjxv!mG+T-~E!Y<9!!gu}la(^xHDm6p*H3
zqyctOR@HsGe0&!~J{G#Co^uPtclMx?wNP2!lDd#iL(g=iPvu|PU8aUMlmzhB?>A=@
zOxN@Tej{g)=3;Vvj2hiMpm?Qv!aNuiSR{Fp8jJtELQ?`S{USCv6g(GvmjQaVc3!4N
zWseCTjP1r%Q*Z~ov*aS>>Q<kA2R>xEMJ~HwaTpHz*Rx)pIMOFxOS^bm%PGt2G_SiY
z_4<pd?{5iRHu<<^y5F`qJ&`nHznAviD}JSOyv*Xp;ACcp$=2rnc^*RIzYC4kzvgL}
zdb{eJM|Y6r8p`!_`4+xBc4Kvc{`<sFv*g^BRvx=r9!qe^mSGr$-)|RXj_O|YoKO!|
z{4R#?q9`<c`n1!&Mz6ZizyEV#0@YnV+(`q#cjQTW|IRX-LbpbO4+jHk0(6E@*SYTV
zXM&L-kF&`i^DpCEl@hFgi9Wlx9U2=DI|r*N>`lNkzJfwKG;`2?#_4hLAIg#>SCyJ0
z?UFia$USS}7~|yvGLIy$_?83q7$F+VV>7kmq~>7O!7^BddBJ)61oICYbGNTfm`s5p
z!&S-^Y>*Wt6?N{al9A_e`gC_pDDP1q8NXrk$xkq8mX1+*kcS)(s*CZv&lwH6WR!g$
zv<fMZTZg>seI9!%F9%iw7fJm4<17j}t2c{q=^{A3TtS6x61M3zV=ounDLPe>1OAcr
z|7?i(j|PhZI)_S%MkF@B;D=4TOP7zm6gn_YXgKww&{pfq<;d0BR_xntKGG~%uqGwl
zoHu=Uj8X1u?zed}iZ#b|o44nz=<KeCDfZ-s#g{VNbESpn+m#+?LZ|NY)-|9(9etW>
zpyB}Z+Jya0cj^Fg<(%DfyP~N0%~>~2e%#(ntV}0Ky7N;uw~jGIjU(zT{nxtwcXx<V
z1~#KtT_+&n2-FCuPRe}Z0sjHy2+_@wzMK$T5FbmdO_oYwa@!?DcCe23?{ODzT2Z~|
zDXfzZzlPTBYn^qzM>Sn|JoQ%>IdPftZur7{jraEMQz0pMbDu;C)h41p_~Ba2brsuz
z{^q&}HZ-eCG~(@kXT1zH7G&FP^QFA}sO#f|zvLx{i`j1>*?1gxi+{+Xb?gU2!<Ssw
zt%)kd{wzP(l%_0qGIYMS{V=2^W)7QTz>~g?c6{Q$EsSD0p8>{@_x6KgA#;<#1`AU^
z&`H}NWhK=y$^V-LARieE{lxhrv<a0MK)9;&L$}pHpkmoDY}q5l#DaJfa_WaN5;~R+
zN0B~4Jjm;@*V?|}9T}$45me6^jmPt$E10L!Vy42%Ndv*m=$z5SclAt-iI%gW++>bV
z5w|b??Kzy3&sLkgtg_}&045(g6$X`Bg{1A>Z#@+{Xg*|^<sUE>uxK{^VKv!$x><d$
z2~F{IFsRfL_u_;ufuR0mFiDK?I|(qNdHYXJ!pjWcSdC}iRIl@r>n3y2(in4aZ50}k
zTzT%fnS9!=vJx8DNnnz30cvN58eg%NeI5=x@P{D6K&f$DD-Jh;ofOP_PwKObVEM#d
zGaCg9RLpafj_!!uRwZDH>`Sn7XY8l}QbB8Kv`Q4dc$y%rQ~vLIQS2*n4yq4)omw&6
z>(XCK8a`2@7F}XR5*uwtf|HDBH|(G`*xTJdbM0M)lM-Kxrgk%%q}Ssys92sg+Q;QR
zhZ==FkiJNNS%f1JIrp(>=;Vhl0bKu9`}^uA{>P>h8bT@dNr)SnqyVx>kvcKImQvNU
zVxn$Lu;GE0<ARU$pgqIHd(WqGK2V|JWmC0tJ{o*cyr|W`f3n1La4sLQsWl+9s`2f`
z*Vpr7Y3RVK%D~lp`pU4lS}dozCdM#6itGb;9!|@BiqTo0s6we?jS{uW%D&Rxh}++b
z?lhpc$v9@mOFui?=SmmC`~Z|#X>yCxU7!vCWJ{MUuG;ycx^EO9bw5EZUpu|w;Uj}M
zXGj%QW%-zTTh*9R7t(jW?#G)>$_0Gh(@9;c4JQ0}uR8vY#A_54@AcQ87C^2hR-j@M
zV$u4Q7WrEya)OtTR48;&OfiFypmBo!80<R-AMh*#tcCU5cgF`K>vhW`JznHR!(Tji
z4G)<~RpEsjov$0}7B^~75-IqlVx8CPEU@F=Cs+$5;#u0-$P`e2g@e9C3ZcIG0Arp6
zrH)LLOgGRMs?@rRV4RQHM-AQ3KP9wTdraJSpTpH;hf+P)mc3#f@WcKWxC{5p`?lV8
z$I-Wzh$;oNnSwWkMD3KH*&fiq19D*R!hATsY}h?(J)I78E%13>$I){Gy(eYlr=#Z2
zVo$d*^b6j?GC%$<<$wt0V1sn^OW2KDjsCb|5)xnV{=wV^^5v=`<dbCFL(djhW@HAD
z1xi4mRe->MEISYfSfzZx<JHI?Vcf>+T+=t7CtD~=-!{9~!MVIYCTMJ!LjFat1qZ<b
z*Gk6LY<}>zFjP;GZh95vx7*;QRtb(l@hOlhRTJ|?95b2<B^Um$T-nJT(^akK^$?G%
zLXtE-^edW30Rfb0_*^2AR+llB!Sv)Xuob&gD9g><D2@>KxlOatRv^=%m-yJ1*LnD0
z326TE+VQDUT^S_5wyXWSfdd+J99l&xGO{qClpRk|*@)DMslaJs@p&*~+CEqTuma+j
za?y}LF1-KnP>g?t@zLIS?r&4ER4?mw@==M#A_`GFEQkz!+BRbUC#Xz8g))tio4aft
z5uq%1i;0zuv!XXQ_V_H?lHx4l;Gs+bii)Qo^?Zp{m`KS+Eu%3i?et#<4+)(5i<nHD
z!(d|aQMJ_ynjdh9{|WT3c)@b92Z$d0I{MjqbRl1CY2%|IQ-YC03MEKV{RewT8X-*=
z)|ipgNP9Fd2+U0%-)YZ|J@Dq9d+xeXN;98_?M42TmLP&gCnD+(O`WF3Ms~Q{Oa)V?
z;_BT8W%(6LM~Zgb;v^g{g>ps5qAo7__QDHYFkLx<)f2elo6bL3wmnr1>*8EW78-kG
z3|m-BKF38VIS+5GES!g2@&OV5`EP3DGN@l8k(k*+89&}#JCU4u|4=A}vBSPGv}rgj
zXd;P9#J`v|{v&6Y@q;h>W0&040;K{f`*``kuA=`_Fq_GOYXxEdD#o$vTu&F4(W3h+
z@9;m;(^sTF^aFBDblg(^=@b-R>D0=}=Bcv<c}1(xRyZ=ofP`qZB>BHAR{!NF&D;RS
zBne8qnQ`ItzR08)J>+SB6=(iOv-L|B&Djf!vxvn4{QP`NR`oA;^*>iW=?-1b!=zL}
zRWaXYJPr8&bAbZL#gHI7mHF|^<b%O;&vzNrzmC(tJ*bwU58%N!@w$HJ%gm37n{rgT
z3(0>H5xj`(Q&ImORM;w`{bSi|dvyNi-Et*RAxLJxqm@uXU{X<tjR`nfst<Q@eQ}VK
zku4pU5DDrl7|T!%ryDDBa-^KYr`8QQnx8*1XXPhQ-`ljl{>>KA4J;o@U8=S3o|*|?
zBQcV-#$92qxoq(q0p+Hxtxh!!uE8?nKWBGUw1P;9lYYpRWo4NSn11-5dXyy%SYp5g
zshB#ejcs<dsF}0oqZ;Lx7FAuQ4KpRdI~8|os4J?AQw*y3$m=V<VrNq&pi$2+RFP1t
zICdtH1}DI<Vn<;(8Ga4g7?-f^%Dn$ZLjkqF4D8us6(iZl13hw-GEt&R3KyHDJwM?a
z-QkcEdbE4sK1heKsC#5b(03@<2a_{Zp8{RI$P`^v=VGU#rp$r1^jKZiDJ{L~PYfR&
zE!8+@uTps?nt*cpsM2NssYa`>qBc`LLEMFtIT*gtUXfZ!5_#zv1E6)<%+7xD=C#%A
za#Eqo*wBwDb;Q?ew)19gw9ZiC?VwJQD7#frazbz{lQ9)@at@xoeXKt|jySQOF2TWw
z2~owK=&aBlR4r$(G&MDymHBK=nps*!Q##F&CB(@xoSh|CqR^~7z*GdEU(UGC4f9jB
zfVq?fPwB{5EGU2JVENWnXpH1LxQQ|H#u`@N4I7*y>Yhn#2r(X9*{-qDfT9a|reevj
zKv+^F;m=2o<i_v=Bo^dhE`s&Eezz<$FmqF3t&{f(U(1=%61ZZHp-aZ;69;ZmyT%G<
zo;BC}TnSi4H`#rO0}l9hxS76X^f;7)c?*AIq&SHjR4t|{6q1<mu8ZR0y-XiqdKu9F
z1gQ&^^?k{!{)>MjnJ-#n?l_`BFqeWH2f@_i7&F~7OZBG(D3dRRO3yB4Ks^^AQ^$|r
zon-FlVTU)sDkWq5G6UlohEj&K#SUNO<b~f%U)L?)1hx>|eDe^6U5lqjh?y-&sG&~F
z1NXzof+))iYs*8pWP@4PHR^-Rl!ba;@pT1Tr9tmOE`J^F?KbKx>+JlBSNuxZl%=3%
zoV9(~WH7{O+|>J^!cNZQvBXH5#P3C~1Uidu&EyfqbZP!Ur0Dsr-ws}^wRQHfzxg<H
z3n$QMJd?To+X+Vq>@)A!x{Jfu>6OD>vlh?kM5TM>6x(&WYD7lEnfJ)`4PH&Mc9qw=
z*w;ktlArne1tDO`ND4do#$-tfU~$_JKkbht0L&;A2FzRcmn>j`{6b+e<Xw~+av2tB
z&fpQ2U@Zfzjf#e-rzzMaCT-~jRt|6J1-ycdq@kTlJ|g~B`X#mP<ogBJ;|(|7Rd<v9
ziCVMDG7__(4DXcjql^(3g@`YywR{O~<K{}z(^dT88=$zq0KaKnxYfyp^_pk5-+O7~
zS5VaTup37>De_j4N#!1l@?Wg8TufIE@4>G?1=?qb@|ZRHSsd2I9!AREnP`)vV`NH9
z;2_)MlX%o44itV#Y10aR!Df;#FB5;L#|e#d!fh;m7kztz7$G`OQ%b2pw;?56Q6WKB
zMWH>MFczRT$j$s{3FS-B?q9GA9o~&@$s|ad1$wWUVrfk!_9ftDy$uXA!KEoM-*C7T
z>2H`zsA$4(%#)!&1Hh%xKl&tM7WD8X>H}*OvOx`n{K>*aJyF@N)>G37AO!tH-)&Hd
zhyY(X{dH=xQf;^;!@h=rRO-d&2<VbonhJ<9MhAWp4i5y!WM7hsU(A$9z{64A$|6Wg
zONVTU17RbQHC@W4l^kzIzTdflQb7%jrzd>k7W{1-PZi6&mY!pP%cXEZyB;OUC9<t3
zYf3gnACHpm*WLKtJ4Pxt^%F;!9diGgovcOd*EB&l+LWKt04pgpDwb}c2Z%sPa1kl&
z<}}N=U%EqIhmioiH)NX6t@#!w`%F<dVJPwVR%M7SqcTM`5SkcApkI(^g*03aqg?iW
z4Y)_HSml3ySvXm*nb^WzneJWgSk_IgWsOM>O11pzju$}op0McLZ$!<?SLk!`$Ry8u
zP_BImJl)5?CChMaAH_3Jcdfk)^9%wbdIGbq%Vqxhgj4=Oi^7x}Ks26if>57zg52=D
zR;t=Ee}#6-=R^8gDpCdPGuCyJO-Atvme`!mr@l$;k>Ufe|BtGFijTX0{>S0Q+SqEe
zjT_s$(I$<ZG&UNev28SIY`d{-n@!T#M&thX`h0))!F{}E`<VA@W}cc}otkxDPDRV#
zYs}X%;<*XCMH_T2G418|WogsCq66WL++_9&Njf55=TjdMC8M=YzwNz;ZnHPK)M`<w
zM)9oD89DhqNw=>04!<>b-27nZkT_yBXL@{&{VPJ)<(Vf(-XFzCNcdJR_u8=X+>11S
zdus3f{fy2=e(vgY^j!;YEoJ8?xEr5`6MKIN>GXB;-P>{8B<|;7JbA*-?)G+-wS$B2
zHA)=O&YIjnxU{N`+nCKa$20Sx;i&KgE`kjjer_ZbOy5lDMQOySSKV>CvR(ww@-Gm#
zc(Lt3k$ya643@d1iCok)RK$bX<CY*c{>)G6rT&yu;*F>|^;A%9r4NXh7)Y_|8VG@y
zZ08))2rIc!q9;@8T>HCMrHFOvd=99SNsnJM^h(5?SJD?71mnP45o<}I!beiE=PP9c
z^QsLybwUEXfpQSOLOR-9+5iD~*L1e1xHFE-d1ZdsB4m78D(|w%oBX#8O4nI)9O2R=
zc<~0(sN{5)091I{WUjR-u}TQU(iL%s=24V8a&QvM0Zu>b$Pec{6UGCiLL&u+Ckj#G
zS)^18Br1}D4@QajQHg3Qbw72gfm<-IEvH0?!nh+LRA)Ng5Q&2n&woZYE+-!#G$!0k
zJbag?`!&LXOQNpu%e?@7=4y@<!YPtua#OF3{;umK*G*@wngO)*810!dOP+c3L%{|3
z&uTKHlkNGd>09ldxHQQPdN&Ig0V<aZldosx<cunH_%I?&)-uuF??Yb}N2A=9)CJP#
z>bm%}NeLFZmNDD61Y{d$FVQF_J69%&o+eM5s}_q3%F-{ig#O)sY2252zrjK_?hshi
zs&HHyp-6C|Ymu@#2#(8Xv1PO3n(i%3$mvKNB8DlkIpQv!04|qtmDFX<sgr6>3ZLBV
z`xKo!sxoNkek5p%N#jgWMiUA;^DBB=@aeaAWb$@3kf+$VK?LDkeazjC_ssa@aGE5%
zsSr)1nklC_`DRm7l07^L{SaQ-bx`|tgu-MOt%SQBUBI7@rx}?luO}M}P^2c3-bQly
z<*d^98AY8S%Dw(rK<yKk&JKrNXJZ2Ro&{!k{#iV{$(3FuI)WjYlf;%kJHqmo%B#Q7
z8L>Z3_W?D?4yrBnqr8zvLq4X~W_<{l6`m<tB<)nPm6HQ<6W4wy&K;`7QmhLGN?&$-
z<c`IBC$ij$C`V1t8bhS#DJ2S`HpQcDFKx=@XJwSNtR9O;<4TeoL|(oX->zg}G+FwR
zM4|NE(m42_NleLd9}?RU(1CD*7Eeq0nFs=3>s!t-mqdxLq)#`3ii7(~RD@eB&*^S2
z0saGaLVlaVMj4=)+p3AcN~BnY6|AGJbeLJ(Dk}E>N;VL|{!h|bAE*HFH1#(_Jf|Ej
zls5~&GM;hI{g%k%(?cEM!Yqw6pt=YGM?_-Cdn*NijgbNvaS;<$xKxMS5}4CrK=wcA
zC&YE><kMQ;O<#ly)Ok_j89I2biI9ooS;tdY7x#I_311W_j870#u7_>T`%gEo)BAra
z`^YE8?zS$I#zWg4I-q4HS8c6|$9{(>I<SL0h@na$)o9vq9Ua$tQosWJvF>yybBRn|
z*thzykjL@V)_3cN#l|zkdU_Y}!{n=pF={cJkw+=9TEnD&R1Oc<xL0RR)3Z$fR+A16
z%_pDeN2@~3blG!muVn2D{%+P#kSKFt>f_3b01tn0+f`%SvUuKHMvuIdg?$hiFGhe<
zjcpl<L+gI>ymLte|K#_XrA}3?dh;NiR*sjO9hc3IQ1OcIrLz~m(^;(lJog=ms?cE}
zkrBFrS?jXyW?F9c#sj;$K9hCEns3b#tfCir<9Wi5i{KA{9I;lVPEYF55OHsIIjF)k
zm7=j}@^tmYbxJ4Hb*p2KE%v}*siQ^VymlAWCb8|}>$<Adpth<GTUpT0H`!PE)k|R(
zT#~oQWmuwhn`!8O*_LP2XsMi2P(x_bCi{0ITO2gucm8(pN)<S0j<_3FGdCC@`|?#r
zgg3fl%s3l)L7&TZn|tdoV?oN{gnjk=%C4|2!B&NpIKyu#>E^Q4%vSbICN;O227LZt
zz8`2n2Y0eOOu#o{Ui3cy0Yo1IqUwM{qG)2eXS_7{U*N_d6m~{-C4Ckv5!Ok2WvX6;
zSLE;yBC~R0PVx&5Mmgn?L_g5(Y!l^w%Htz!_dPQ}(#SHFiW^-#nV;u%7XTxO>C-14
zjKrveipz*@B5}@=uqd^~w&>OTXYSPD4{GNqk%%CSmO0c)@}_@Gf|u#Yg8q^GkiPNn
zk~PQC&JQWaO#$H}1yM1(z!3-B0wn>X;wq8m0}wKHxjvzshi33TMtC_3y#yCKhm-*W
zc`C($ci<nPadY3$?2B*(X;GF{u|RVHX$dnKrt^9PYpeo-yt2<tTk#Bcu3)lK>pW4y
zCVmoFW!T?Y-aGq@*5fhDox9v&_FP2z=XNgZ<yE5eIE8C&T;jg5y=(`vs!6-nh-e%2
zT<ZWo<$R)RKUc;yuNPL~)F0;AO14%98y#smdW#RUOx->#XDLfoA_pl=>bp;*%9zG`
z#{&fWr9@Whf7Ti|b|0}Lb}Jsuy_KysC%(2C#Jpl#9irGV#i4ZDZt-zcdx|?)+rM~@
z+sIvKLglri%5+n!<2UOBGMdr;zYPXcAro%!fog2<zeYMc%|g6%3r&!XsTyld|4D;N
zT>=Nz{l>qPRC0IOO;gp>P<?4`dti7?tz$Smbz*gzM5n3I=TNLjpF_3o9YW(}!JGN9
zv;bJdcG1TOf%}2Rg))fnJw2G(s7LZn3lMdCbqh+8TcePUHOw6g43L0-(})M;J{9v3
zybp)fmt#4{q{c-(ru*j_V8pWm>9knMf2nSW{i6f|!2F4n2;7kv5Yp2id%i;Xo?+y1
zLqox;ss4MgZ#6(8R|boN7NB)j^2EdlhA2$XZhr;B{v)##w-zcHG8A~m><u&li0~J$
zA~gu24<h?Snkvp8N)uiM1S0<HbCBHx?W0Y4<@tqD8tDPrs@ReS+qjJKM=8$wf|K1)
z<8Fv+<_{nkDMfEl{;4JZuyHgBD5Xl5sQAI$QNCyO_WabLZ}X9udyA2{Ss%u*Q>XIl
zy^KK?_Y9??4JvdciL>VQN(y8r1jfb5)X(WXuY^NlLQ8qi?d{M6rTYKX$TDGFfM3Vj
zMvuA7A`Fe;CVp#CO~NR28^xl5f!i=2M0h&8Nw04t{${&K{w}o;MAT}w{skNjFYIJ}
z{b?PUwnZ2wZSH1V3^69^ho&fe;43ym^P&Ro@83-NU8QHIM(;{)>Bs?-Nwxi10cj{X
z1IXQKN4fIWKUo)suUst*{&mov7<tHV!|uXorY@~{hi_?Izo+-2A>eC)e3EyY?Q8@@
z4N4cBj5gK?YTl%b_ZC@R_+#rSA+XOx+h1rk5HhBUUF-GGSU(lKJ-3i25y|w%`e|OE
zxiz~K?VOtr$rjn)d6aB<G+9BJ2R<V5yG0>Uv|}}uUOvo=m}dlEvRe1t;pMj$w~hFd
z5a%0ug2*JT2J-N4nnP_ydyBa5`a`w3<08|Gn0e6tnko}SBx8~wisFjo6T~(XtwD4W
z38;n<$16loM^+=8IYjhvo%y{acOf*2@o|1Zl~>N1-Ih!i`<X6pjw<FKfNvqdm#^3o
zaU6bOIco6Vb(J=C2=q?ENmzs}CDTn%BOKZ*8v{Z<5O!V*z#yPO$L=qN69AOac>9n;
zb(heK>X=<z;X<;IQ7luF33rFLNJ)ure8(N=F8?i;0_aXu%9dvwU)%zr9YFGGlquut
zn!wm4P7b_>c_u}f|BE=pEQ69hxFaG^wa$SSExty9(nR?c;2|+FM`;31J=GqK=34`I
zxU9P{{UFg@K_cT^1EkTaxCr^~LM{w*B=CG1vHl^Gm(z$QQ*?+pCJM4y71xkH8%<r4
z)T5J#-v#U?ZcZmKR1TLa_i-w{=j!<=jj(Zz+be!@rWQ4=z5u?vZwSVCvvQp~uTOD9
zRo8rKI9sKyL9V(^ztUudD^?9kx*PzZd1@K{`~$1>YNjg>4}E782bHfaETtZ`jdm&g
zuwai`1Ub8xG+gZ$jZxpvIU$XVf+H5o#lKJ*TUd!WD^Htao1#U}Wrrfaemw2GOetP-
zFAw2M=C~YuwY9M_F8r9Qj@G96t`kqHf1fPCD<o3xxvTg2B_C)iSH?tk_C)C?F=m@K
zKLA}xPkC)s#|fs_-#4}FwiCM!kZ0Vy?2){MAAwRUl(D>YgUHkTF-SwrDy4&_opbNa
zaM{TZdN&?4JwMCo3F00T`B$+uZh!eN%H^#dz<`+P7gX3)Dt=B-4h63}(;wAbpP56(
zl%0R9D>h}BCpA^AmLh5}Gf?u6eU3WbV7p<A^3YLd-&w42n)28l@^V8I;WDosQ9L~1
z@=NaJj61<h|C&*F>k?1xSSpstzeUJb)Gwic4)&wU;vb@k4)IK1P4{Og?E<ZIA<cpi
zP-SotO=6=Cno9k1#U0P#GtCRd7#(t=t{v&c+mu<~5>3VGTmo}vFecHlgLYrRqEj~b
zsO7o+j|`;>r88^;$oOPNSf4{B@{FM@H$E-^h^2)DCljA9#*GMzPqTfY|Km$C5;AUl
z%WsZ}9TQAM9DL7=D9ix=(Vidpqj7BEQFtSC<o@hIRHl|1QOaCJ+~_-D3VX-Dvu){G
z=Mns<*S3X63Kf3v2R+mYC6WR8%$<vSAzI3|CTk!H8kBfyE;Ol~@co&E)avMEz`xdr
zd|v@q3(8*-DiS|)ZmDoj3$<%&N&GpA7sdIzGoX+Q%@*1hJ-8aTpJPmiHbG)okvd-E
zL!gNVx4*~Vhpf`iGlC@pNHXSjUe>wrT=nwKX>rca6L$}!a-nUnQKC9+3&wiYzNGGd
z4KLVL(^Qt}=qCjo776jPTu_3+L{UjG`)sJUKxP&ZFAix*TMW5avomlk+Sam!Ykz3f
zAZ2{gAnI;X0+HuNT7tG^Kd#!GbXs9~2HFO73nCII{gPfwsm8j;lK0sPjos3Qe16$x
z(UK!I`{@Np-sF;{@#3<ZO8=U*y?s8fZ*KfJtQPG6Gj;A;DNRbEE=@B8t_sI)9%P$B
zYQ_5YmqPW@B4yq_c#DT`gEM-!+v)9W!T6fC^!hYdxvVZNj8>|lwSYG3^5=<o<KdTW
z9$57lSShvd@hcGj6-t@JURV#h*q9vbL<-Ip&C)$~U&sp2dx|wR^>AuD+_pz(8T<uX
zA%*{3&|l(P9k_YN-kdURo;MvfwrsR_=DBzVA@e_>B9bcfQDfBRGJALo6T|L_0-=>g
zekxjj=-9zUD5FRM{3Y59i8MYA)S`%Wl(08I9F%~CfPTkBG5RQYbQxTerYMDtvd8dT
zQ4d^$9)_e$#I#@+C4HB^CD90kMCv;m@I6`tsenvQC<<Hp1hot(hZ-2@o^6f!yIO!P
zMk9U?LXuK68Ia%t+Vp*Nr7{QBG*~08vyG6?e<zrWxha{kBN%l7pgf`-Q}_p&lf?zK
zpauchUCooJdr01lkKow*Pm4~s3C1iKWuKiRv9n51-T^^{<p<;UDwk>ce`A-)x@KqG
z8{21X(eiyuwmK-sY#TZF5NVY1ii(Tx$dZ?{C}4>OAr<!}#A_K@{M*+xe#z;1lD*A&
z&}!=a>*f`qG2ftY$*bVF?%aNPz59<Pee7Evd%uOGDlW}eToi7KVv<1Q$;Y5sxdCMM
z>`W3(WjXzfWFiz}G6-=m)eLk4wWe)J9+?OdmzU*0=F=0eVIOq1L;Iy*mgX7wyXki2
zm+z|G@cbQV$})Zmi;iVF?!=<~LCCC=r`9q2vc7cu0A$9QlT+i|)^2~-l!WK*SMK4%
zYeljBq)^lObe{FI)lK3Z!%rf1j||7J(DkxEq(FnMNWJ2v70-@~*F3Yv86`^s{t4|3
zfN&~?4lje2J>P(Ug2P;Gqbf&<2@$h&f}@l97N2Tpe8wp7Tnt-CD2s4_c`HQl*V*Pi
zJfPl<ESE@LANu>BK+y=2C5sdJKdn83Z<YDL>gFuV3M$#uSl8Rsrx6-CMRR<TJ56xO
ziBO5p&pp(l8){fpRpQ63qu{mR;=Tr_5R4V<6(mOb7I9G~wY+%USy}W4+$xhaVgXS^
z_P!aE`b-^}J|TW}R9t0pusWg)0iz)YrHDD?=uf6{Xzf-y7Fg(%mMse?6(4FV`DqP6
z^dhDriOU9*oZ(@VwDgTGR3eEad!P0npkf~s`DBP1jj=#8A_!VUUjJDUmr8<bkiG|g
zQ;WuuY2t*%t$Bm%q%3U8XfiP%B@blVI$A`k@fx<lxKuY6p7LmH|9H>qOkOlnYi`%x
zX-gKwRWumq!*8+UYX8YNxSMg2nfCDGuAA0g3q0(+DjNa{XVGLE;#jMvCxJO--U24a
zH@F0m8bsPZO-AV5q}{)M#a1%RQLdAjYA=0P5@AXYB(f?&a*8Lc*fit54KCi)g7_3+
zTkg3hpx7Os&QCBrPAK-ONmiuI<rC+zqy}BxF3Jg?wJe>GjK8TfnXVW7)?px8gUkBk
z#jbh1+2#S6G-#oB7h5kKkj-A4povV_&VLGX)ry|O{9@28EA|Q0M-wXfz`#-$ME|Fc
zkMLjX$dq2<2ilo44U*%=r_bLSpMy}Cty*%Y#S(+4D(=L{wK{V8@`d3htrG?)%fKo<
zrBd#(W_9vh``iIMvlDoPSI5J_0bO#`O8T{r5k@Sp{hG*cv6O-41@75<K|IMt>xj5U
z=__JFQ?q}O@dJRuM+OptLP%+O_$1#reC)-vL7e(r!;(2ncX9cr#kiiSKl6iJtJpxM
z0%`XdIH7gM1lIi+iNMvVV;*9s!U{|v3gn|6?^=&%fhFNvCfgfef>%WyjH?MWs-!}o
zpruOu)DsYldaGj8CuWBFHc%)NPDwFOG;cc3Ho1s!cw%+;l}SP=PCYI0ehYR71x*F0
zE4OhhffLzA%*{8m+<<1K;T`z<ujj%4;3|s$;HsKTdvbU_L`n8Jf#nKqowbSSY8_)W
z8|uy!Tg<j^P-q^3eW%&PuFkLfDdV%AFbY)#3K1O{gU88$1goX0&QcpmQM?i9Belwy
zk><>9!K(KN{GzBcIq4G(HSX`+YIBn@D|O>ScFV<Xjng;Dz%W|w6pV*fp29WO%rk`L
zTPC+#q59@L?BPbsFQ3QcYz|Mg9wQ|)rk*?n-!Gat*o~`)*IY9eZuYek`!AQ{bu%w6
za|l<eQiu9ZxX2`K{g@DZ2UkSxyYwF$o7bN!n{u25i%Ig)e5r`i?@uAkO@f1(H_=QX
z>sKQ@i?dm-*(8H~0`iIp@sz}1>YzxB_mVk(st|;~cz6Wtk2RFnV)b7Ds{HecYuf@l
z$KE~w+Ke`%a>l?AH^g!gh|5eSK6!NggySRQG5Oi-B}?>|K!=%kHp8U|U%C9XS-~|)
zRV_-?Xql=?dmJoMHOE|ZHMw~BnE*yi_-EccE5>u)Ab14Wzl}joO`n#{aTv&f(qc~j
zy)z1%^G5}=6b>wN$yGmtVI$B4><>lelgRQC#gk}EEwScM_|!|ZKR@oJ22lLeE|ux4
zeXAj+EJhf*hDq({P1%pn&8vR_s*!J8t#X8TAde=EZp&;`gRnpyEUG44NQiz6YzRXf
zL@Oy=<Xz4=w6$Cqz%OFuxp4rlb;*6?w9m$aK8%(w-Ls|-S<np7;$-%o1ch}JVs&1Z
zE|Hn#uP&I{!v(Gen$|C>QJ9w2MRgqVf7`xWzv*kCD$YxloUa_!LbIA(1i0yTD^EF>
z-ZYScOyXs+?ULMmgYO*k7^<9xl=-dD-6o?gKE1KOSszw;xJ~Y~Vv5H_^4<3?lx{s)
zR~Yp_UGY4oDD)yJa`9qXhKRUNhMk|Yey_Z3{%mP%dR;A8;fg6-n&Gw{^!MZ(U7S_z
zz+_;v*PO3<5Lit7Jjjz81{Um~ho0Ox{~aq8OIQ=(A!xFx;N#pDs&5i`_*-t2FL_Zx
zYq6bmGSEWUfTfZGbAi;p<hinVGINTTV+xDL?ym{IE!s0lM#!Vn+RT~N22_4YS`XSK
z+H<ztI63^aA_U8UiS|pcs*!IgAziJ0H$#1^!^8SXp&hSB7ut?!rj|SLX@GG@-ogPP
zH+<npzPe`N*!mxEh`CxNa%ij`0aU-laB{;;rDDU`Xw#0`gSrGt6;iuvHDf^d3Fz8=
z{%H5Q{w1tvwri+nl5Iw)P_?P&9|yAIGNLq??Vd+GvY)3P5(8+nwVHXv5fG3j<r2hZ
z6_yghBawFDfEZ>HII+9;#-bx1tNO&!fN0T?`CcDRP_Wnm4Cynl9%4yqoTW4}(tAAt
z`*J~rhQ8&TrTK<iE|eIL(+_U7zpVt!F4fZDvohv&iPmK2Rq0FM9)^jzBKpX2Bzxiq
zWLq%N($oTKS=IdF2R<Bw0s~)wiaY740mMgS%HU+=@x?-9k-GxfFj=CekVqUx5SKRW
zZ=xMa?c(hgz2&?`dP`PF9&<6jsy=<(8x?kPA?pva(SELma>*Q^8wCyMU-9t}&n!ia
z#s9__y({bQ1?#l;NMZgSntLf@BmeFBk3=XKyKc;7&}jv<yHS0+ZSA;YFEKMP0u{}c
z(OtAl^G;&X2$haJc{Y-ND|*csqFpeSgSWx&Akdxw_vw1l{|KsA?y)d18XH~Rv?Ar#
zHw|20Gl^o2C4aPTPAF~|drQo1#AdCzpLFu`c(Y4hrja2B!rGzTgO;ewY^4pWcgPo<
zE#FQE7wEgLF0?#m+@NV4_UCU()DjDIS#DF!T{zK4S{)@)kqH`4dY?l`7iAMRyDBf^
z$WXw2LFHKK=CYswv0G*3%2?Ij-a%KE2@Aq~Ri$9VS@K{h#9*YXWD_5~m${=Hq(+HY
zzFCpa4xb8?n1Eu;1M#u6HCh<sQ9C;Iu?$!^YnO-wM=>;4cw-0oLgUnqpt@jq_*-iZ
z_(H_p)>y?s<Tv|ZF;lrXW!r$Jzz?~{Rb|(eewYU)0xXa7-4L8>uB1T|93~xh7>>X(
z)N)pc!jUtj4^gR((K@ko$zEjv4XpHmB_}>9!&20PGDBkT&lW&d7GE5Xs~a3e%4+gG
zL;6QjX<q>l5<|d0Dtg?;^g;^RpUNq4U_>7p_C7ZLz&LaNtOwG-$4U}J|N7ffgP`eW
zC0|ciTN(vRX_s1xshxUYQg8Ntg6Z6wm>)MwkCcfd1+cZ}(Ve;Q*{dxIkA_yh-LCT@
zOR9@7rEk4rqv)n-=SjnZ04e74aISK+!}#f@tcX<LPyxaI>ZpMZy4K5_XQNZtew#pj
zwU+!wF~Cc2@yBs;Gqh6l&sqNJwoimvEdQ=~&ckb*3L78n%g&xgg*s15C$@K6Z<Of&
z;$X_lm4~FhqOz-m>(UH=5tMg0JlR`3Ch?46$aa1KZ<JN`%`!^wQ_cX~?XuHV(6%Yv
zt3!482hd=rtqD%5rUkp?l<#NA$5<Q?W}DoY3rIS?630<yi*r80z8!~AsFAmm&yOj6
zvwv}JF4z0JifM3TahWE|W}MwxNwmr)RQxQEcX8ybiDA3mCiw|GROS|T5_HT|E;ZC7
ze-h%xGT|H`f(SN5jk?#<YEbMUL>xuf{ThUJ1#r2M4}lv^xX|bOjzG*NiF6iDc@0_&
z)IXz>`9ML-z{r`3d63)3r!Xgh_nr`MG~!<xF4mhXua!p*M293=CMh}yWCt<^Tg|JY
zq*hV)`U@!)#nVw9gP_AE4y51hm(LwxBJ?`w1l|QPxl$9YgqY5NjI0>DQ8<BHs5Kz{
z!k97vD~uJln2NnmR?y2c{-Z2mJTh}qA@Z%+l~ME*+`iOEDkEsC{ouabDxR#~iqXKI
z9C~EMup5_la>qwo1sEx?Lg*rMFD5VUY_7JaqzC@Jid2<*d_1>8D-F^aG-;y%N<w~I
zxtn|HNE#wWDJ*#C5s9KLLg;Q1#PTt*H-<80*lbe(%W@TNVzpQ=mdLUM^zH}y(9mYU
z=jW2lLFxbZp#0xnl=K{hT|0#zHQhFx`dzY~yZd_)9*ys*jm0?oMX6`b_9d=3Q%08c
zfyGbW=lAKR7w;8OoGFH<%2d^X@U5_HPTkcNCt6JXeXiYB4L|yYd5bTc`rSGWD4W_1
zRgxuj|9etST_kmwrv9&VY`EBQ_HMUMH=WuO?>&e3^+=m})N#5|4B{~QP`)+Lv^(r+
z3`sFtGauJvCC*jpH$Pf2@09$CoEvA=vmtscTDUr3Z}fI_c*hU?W`}C<I=2_T{-&j%
z2YN|DG0xUhc==LPwI|a2Q=5go&h5>-(Ph;;#7y2x`Liq8fU~9PxOpR#$%|y(XfArb
zSXx<a#4P=UhO}Tf#n4F+T{*?QY#WMFcJyr$kH+rYuF0MC_ah*Um_jomwv-2#8(}Aq
zwz$P#fo=nUjYY6V0Rh6Re%7WAVuY>Q>1iU2OmIxkfMqvaAmBn+jB=_?sp%hRLOH)Q
zK{fpVr;es(!4_{xDZ`PJjUXx(_+by0;IgdCFjt1Y0E-ILJ7kp)Zz>&LTwEKTqyEb@
zD`CE@jh3S}w<as9Acj44mD&pQ?~~4PA`O3PEGllfRzuq7g>fkx1e6!Jn-EiUbsC0v
z)^AAb6_ek$U%%lvUPXzkgx-_vP;sJD&^(?O+9wddmWVq)lN}!ggM;>mK%PCD{L^97
zN4Pr;H}Gclb`NpE3Mqvn<xVO+T<FyFfPIDZ0R$BELL+_hnFk^W`Lj;moFq~<xVxaI
z9rrD>&y_9Upc}%a>?*EmIj=O83p(IpjovKr!+$0^?k7BUo_uZiA4I){VQlhHd490-
zz*l{q2$QPut#9VrZ7OZ>NTbYw-t0GzjSV~+G9+T5+@qkBE{{2P{kxTezvp&!3m>fT
zNDf-E*N?U1FUf1=wIoX)-a3}1tBqb+Fr@0sM;P0*)tB6IFf|b{IBvDsOTrVZVQl`s
zHu}<QT~~xSw^&m|E<H5Pi}*f(?A%$!H0HKEp=ODfN?hHSR_oa1>Wxy3O06C3U^y+}
z?Q5U8|M#+e4R>W6|1J(R!iLp%U%KTNjAWI2`PIWgC|)&bqY-2k4}D~PvzKhV5jX5=
zPL)~MchoiNbLzV&y@bPFA53naU9G>s8qBk<4BDp@EAY18BQ#Q`E1~_YnmMn8r#UDq
zGA_hl@TGC6{ur1~6)NdN7npshy4`Zp3;Wpgd3TK>M8AU$#y^CsI@$S6I5}_&z+9Li
z`S)XLNjiK&-=J4HDBAgffqB!}$j<?Uc{GN>jbR*M3OkBaMF0|P%k@w7H*qT$t}BF_
zgr2(Q8>3TBKy(BR!jA%~(mxo1KOu<Y95(m?_RJ`?dv_JE5V^|mVFEbi%@9W5aA*ME
z>ZzPkHT>?rfMMQDgwqRqson~dh;hWp-YVt%M~sN;9V8SH5P}>L6%A;#EzYmD*;gu1
zFbQOC!rls}v2R`3ld0AYA#`$P0f1xpQ8DO-Mgjh%)kOv?pC2e9aulDnOa&J*9w_<7
zhytJ$lOz5|Ga!Yld=y6;ZfG2hijDZJktU-6msE+<7<4cl&yuK`H<2q>H2^a3cq4+m
zvJH5GECa~$k3W}(7y}EG2vA7d<FK>x(Ep`w3H`q}E&hL4PttRccuxmYdeu`WQxOqL
zHG+x?6ZOo*G>fRa&f%2Z@w$)5*^+(twYLT``wn3he|+^K1*-EO9By^O{~LkS#PiOE
zu9(Od4Jl~Mcs()9EK^==zEbBM7zIdxovE<)z&<k)Qi8r4nyB_f0m{2<aPxLRW31^s
zH$Uu3_yWg{*THSAOrvU_bGz)qB8P%oR|H%cE9Q4+BhkpgD<E<tOF6q!B<4rwgltm~
z(K@zwfC#2+0zAPf@z3C~0<@@06dXDo9+(d79M?UJe7LAb0J9ftA;KtHC@`~C+N2Z-
z|C1P^B^Z`5hAkZ7Aa4hLu;=M}$r_@Wkp_R_tS{RUTu4A90pmwnsS+a&zzKdi^^DZH
zKRMdJ4W(}!3QUqm;=#V$O2xEhVtqhtX#@aM`r_1za~5KmcwaByce>Ty9^0j58I48I
zT$uS_^jL_(Z>t+?f;|bOkr-ku0=4M)=uy*L-)+&<X;hFf2ijnvMs%(py|(1>Ezm+g
zIv5avDb67;7vPYegQbs@Nn<2Y6vGpSup|8!W+m(hIoq|qlTyc+Qsk&6IDnC?Tz<v8
zsRT}|UccC4#(W+bBM=loI|Zh6MD*MSel#1BEj4gG;CcFgdJ?=2==hSXIn|5-OBVDR
z9=i^oZIks(^q1iPVn?yusBl?W)Cm3$3kKb{ZXEV6?hI@}V-d?Gf5qIyDMXiDXn9X{
z(KpT(ResSVd7@FNLAiLYDzILC_1zDPs6jijRY6gQ24erBCf&kS-DO0?MOBF6Hm^H{
z2XzhwJsD4xq3b&)AKWt5ryQPySd@uyWiV8!oM$l|e)+ucGe{g-6*CfF_<gQ3ntmS!
z;BIQvd)h3cZ=YUTHChgvJq8N2uL#T_%2q(<p!0WXt^f2OdX#;}f?2xLy<$jraBmmI
zsL<7B!lfLFhfaP_2F+0i4>~52=w8zG_F@g@yTks~1$-?R>w`#l69RmcaOHyVW`QUu
z=vd^+nK^}Vn8CjE_(YanpAeF?)pBR7@tc9%kTy18ttAD_FENXY4gPKtx#4(MscddD
z$z&9K+@V#0X$vvyc=)_kmPv2QSo7Xu!-J?UJ}&ma8&|Qd(jU27*f<nGY_9EqLCJ@l
z!l-Kouv4G}tU8@?DczisK2aU;E75!Yn>MZ<e;xcrUK%6fQp}L`APNdjIiwG}+z6s%
z>J#UW0wytxFahU9qzG}(!gRCfS(5~#jB~v!K&U@KXBiS0zh{L!#!zcg#0Tl5Pb-%L
z9*m0M6_wB6z~DWFKp<ePWE^M}ETr>G{3iHGnPME!%Y~ZE?JUWj4>JH=kp3eL<X?U_
z>7Toe)BNXq7&gRvM@hO;9ocF<yUgNb-x<zFJhbtCc;U`#mlC8hSn}+dye8nit!FCz
zGJo;EwW&VeXaLzDl=u-pW-Baa3nT6i6Naph7nN1<MvpbPm9;i`LH)c>tkJwo?rk?I
zKC1tHL)S;Sd~XdQIcztYRm+ZtV3?Bsvf>JQ!k6sFg(gns5WzY0WKw=&S|E@dLu_N{
z%V+W*i^hS1C|<oWZm?txoo}+$9-TM;fx;0e)BHZ?{a#@P@E-gsRSk-vM4V63Dk$7y
z>&y7T<~*#%|FWFhc<^0XPndQ0iaWQ)vh_n2=k=IUoYa_@GYOkIT4TcvqN9Seke%Bo
z#%-&7&XO~_Gq=(-KDA_pCpbA;58i=<!q(m-YjL7WiMBh(Ywk7<>>G(#GZhmRYct%6
zIPZ#reA>qH9ABq~`>QYQ3dsT=g^3W@=Pp4|eR4~)A`=(nn~Skg^koH5cqJ6dDzQCr
zMI7d$$PxDb=1ZoLqUJN#K}`B-1*vAMmFtw$g@s2PFB}WD(!#{2Ql+FZ#Q$QzR2)NN
zM-qw0vcMj18qudqAnz%eXUyjN1yVV4r`l9?RRA+byP3?LCs}d{koh4`9fGl!mOTtm
zQyc*GSh_}PM7i|DAwfRydDF>=sUxD`?2cra!uz<_>OO;KBcpVO+r`G@z>RYVi(?CH
zBg8JDV;GD&5V82e5xIG9BJbcr!-q%_p)1k3t+2mncQ)R#gGinYs^1fFUpMmFLhfgz
zQOEzJ6c4QMoI^g^=2=xSCj{C^N+Pc#)zR3PM~TQ8>H~rykv`{$0o*XEra)VNCVWO_
zgCw~lRlKLN-&988+^D0tiPE|1E(TcJ0{Yy!whz!tPXE(#03O_$lgczoz7OAjm0GAn
zfK%_+pCYR`8H1(wH(`S{1|k@n3IIO+UI;$9xX4nC#6At<bc0N0wlB2DuY0c!WW3&W
z5@DM8eU)2%5cs<7PO=M(#*94Srz=NdS$a}$pAM@SOzOr)r;#eR^%mU!uH$Vd@h{y2
zJdX_hXcTV62b-3kk;)Tx2Ss|#{A50$$H<j-b)~Wf20gO5nCx)DM4a=<;Y=8Cl(t>L
zU2=^sL}ctKR@GFg@>ZFQAA7etDlEsFFoqD`ZDyid+N22`HesErrEuFTy)0tUfEx2Y
z?rU8)bm6738<P9|J~YO<B(<p(yxEly4hf0CBLF|{GiA*?1%!_a`m|ya$!P=itw=1t
z^=sf~@aDD|9%flR<-x7@^hkd4wOWs8*eK2LAb$MBN?4hw3lG{JDIQIJMfxyePf)DI
zyIc#A<*dbI&OwtMVoo@&=)7c(dwK$ksU+`T(O?eM=Nz0G&zsNtvZ%}NRA{mMNW<in
zTX&_ovhe1!*)i!26tT_kk8d&`cPYh2<D}ZSJToTi))_L!C77(-PxmRd@G=Sq<_tIF
z?5DjuZjrc?r%mY^G$m`CbHZTL9JFbjR+?{BQ7rz@R?z3vuk@IRD&(%t^nvPFTNPVZ
z`Xy&Ry=kk;>@UoJv#Oe=n&<I-REs4k#y5cXYP$;Yn`#a?ZeG<mF$p#pQdwn-3HhT%
zg0`qdZ(Qnz-DA~}(S+E!Zx4Z|jd~O+=TXAuVEa=|1O*EJQ%k*zT%}D}2t}YEidYao
zdWmxc1r*l^fb7`hiMqegXjNVg=Rd>Cam)e;{9d3~J^Pi1a444}4k-|ChP|8vLK07A
zGys@WFq#z!9rS!Z2vL5&(wGQwn(8ev#E7IVg-DtDL>jh=9QFtYasm0FTI71A5elRT
zcJ#(@<uCESDFejzks^>qmWfN73<ZnnuE%>jjxfiKMxODHm4t)z((eq(WRTMD<_6%$
zj(Y6bUc~D_87+N0p^m~a{GN18#L<_;0#?$kmZ=(%xmbn(e6$}l%wLcurnOmR2<2U7
zUJ2mFa(KI?GZ?-cJVSEm)toI$ZY?p@$r&cNA<>q@cda%Scr;RGI3NMUs!*c>eyg5K
z7rqqr>6u@G<3@%b%aybg8#qs+2I-|HIJM|po>b4@`nGWD^sE|Yz16H)=LHmNdHV~w
zzrJ{U{P~RT>dBJfMDnC>x%Q`t`<xb<RxsS8P`V$rkppA5zd)0{9M(Irb)sz1>bi3H
z`0%<o<#o27X0mCKX^_ER7Qy@Vw%W3(<HBxgSw33_ipg*{dwO7|e@ATQ!Z8RJ`^5M4
zizRejZp>OdYu@o*spx)qs@2pHJ$$UhG9JKGkxwn2qadq^Sl4x@=h`m1TLdc16SJAI
z3Tmys)^U27LU|x|T-~?>?!hwa>aVIUxuG!*|A*I-NJd{tCK;W!eJ&?tRS8mBYIJsJ
z<B2)CY|FE?e#$@acC>~6IH^*S6XMxz`?(Q$x0x)dT3dvR{T#)Z6u1d$DA6~Azb3OT
zBJADJWYMso@T+}3V6AozS^3{6q9%TYrW0x%fxR3mkh#Jpgkc70(hr+=jbRK~Z>svb
z;DXff8OU!b8#y)oD^->{5N-ZWuovNl9s;#OmPJ0eTHyquJ#J9G$>-j~rl;pKRLWrC
zs3{SNQtDPhiO7~e3iX=INwNs%m~gJ~Sb5^kQ|~h74zt;ZRy<L%*z&sm;knuj0(>R6
zSCu;O6jWmJy>90CnDqS|R5AXUGr`CNk1`6Dho`MlHqJUqAxDH5d)xftYVrPg^Y_|V
z;zJVz5wLZXYC7?mxk0WUlRg-c1F?x93&1=K4kJWa_h+2#6x&?52P_emT&Kb$S!FB{
zPF?sISj9W96!WIsDdd7UVl2o8mowrVebNXE^{LY=iyp>-CD6u2P`T-%z!4AXfm1O7
zTK-YWa<nYy!T^T(vnMm0b+|Re+9)EObEQe@$}pO})rpni_PL5t9QG2*^G7-D7jkPQ
z*Y68bL<p8=l&Qsv=b1o$IE^`eId#2spWNOUtI+JwY(HH@EGv0H{ttiIceMwy8++;=
zw`GlrYBt|LD>eQ%%j$Z)UjV~+Dn6sUHnFS>){oS59NwSQWn8Laio7LSv`OVOKs$gf
zW)40$S}n?$be|%FOjR${f9Y1TEap#zmFA;JJuxjJ!Trrwy}x1^15t!5YJ%srum&k1
zi=lZ`_m3xsM$f>PqMu$<H`81<(uL}@E#kPU)<%k&lEE5JnuU!{|KqOBXj>IJT^Pq1
z^co%Y2b>zscvas$x3$Zs+3sL&HG4(dWfw{JJVc|QCcmLt${#~uZBBt4MWevUK@Q&h
zI<nWyXMF`<ueIFq`k$A7`9|FMuxa#*abBB#t<EDrD-+rTr?XPZo3lI%*NjHJePlKt
z_H$bFNnWB1KM^UTW1k16m1;jn^bkIgYY`1;@)~^;eB5e$n4=3M`&(pFAtcH0yOdr=
z+Ni`=pe!OmBsC=+QQN=Kjb6MC?Tt$hA0~laD0tA9o~$9sAyt=}ADcEI@Q}9|0F=W~
zlQm|iX!lI;A)8t|UJncyK$gKycZ>ADg7^cd62y`e(sgGBdZIO~<i5f22eo#gtTo`N
zkpmeaeMbo14dP5R@zJ!kiTvH9dT<*K7Z`6?@57$$4H5LJvDe*s<RwAgAkeCH?I({O
zcSf=3+xtoxCM+PO!O8*fSsH*t*&|?Ru319U8`$*Se`HqScL4bt)vQYjt%M;pU+}+r
zN_v4@jI;o1=AbRY^aWH96{>(l5|wxJY&qUSk+=doM?ya<j0!TLR1WnL)xz{0Adhcs
zw8%u4$k~Ze1FRGrPZZ%Y+&Egz)?^wtr}q1bnktyP6)qoDG{Dv7|Ca^u78`UWJG%u)
z=IZ$wiJb_MV$bJnILdJ%u<g7cui^JD$h-dk;~%D$E$HafrpY*q@X69+Sj$M9ul5_d
zptZVx*UwZ1jrE-Feq_;`fJVxco?d#YZM;09=@0UpW%`Kd+KNAt&axR7uGPTDXk)BV
zG$RT}jLDyNp3Edw{vH@k?_Tg_IL`g*n>2m7*Z=OMe2T!|Q~}5O!A^wcwQ0TTAMIu-
zq5GTV$-pyy)V=4b?Mue%qMemO@PRbZn8%97y5DI_LsMSjyL;pQOb~_|9l00RrbQDg
zt;qF6{mEZV1GtPqeiYA7$vQWw&rnL%n6^d#Z-}dqb~<ajeBg)k%)bxl1Pw#~>0Z~t
z%)@XUiPlLn2ndMYSPiQ={uzW2$Ftpu0e5-tCO+<&XQ76F7yL!`&3w0j`n7=j_2K`_
zO#5F5ALT!436z99%c2?7n-}P`Iz`l+9W)O!IdaBPiN=a#Js%)lIz3N*5@o4)p4{cl
zQeCfz;OxFF1~dhk?^+rM7IQs{BF@VKUdtUg{z~#qn#`G!<(Qd}@|N31SdOyp$&ugx
zzNm*<l$(~x%2HUFU&)yzqXAMSi-~4TuBzum9*ur{_{t^>%!Vd!?zCYU!on7@ZwWu%
zDf$VoCy4@y;))z-*yjTH5#&EQWzII>UlOPPF)}2ZHX`q&1R9c5paD@!)`9dbkZCc_
zuXn1myfvf;4Rz*p!?hnMQ2K$uDR!k~<AY<}T{?z-DAMENtB^y(MKNmVoe-Z&Oj@8_
z{^KOXiUIn4a1DmIrw8wKf;siS+2a5aPXD9?NJzu(MN@%WiZC&rXCaF74FpY;crJua
z_o4Z0HZlk(b8`|YwAe}nGFa)evML*}iQ$(kYWg!0gh4zdA$2>H<_-4P>4#vFM2O5I
zc3ytai`wi&-+)FaBThVp%fw~`=7aktG-9q->b@u9m1_dg+K<u_S0?%n=p9fSa}@x|
z50S-81nWS@dqOa&4AF5@P_QvuFhK;>25tJEd~Fu};NSN}#<UoXaT<U;Jvq{VM0bZ=
z)_?4Ii@}TO+@x3SO=>@<ug7eZS&M$HU&?*@8>9O1fBKP{UduleO<n|l8(cSf*D}qY
zR(d8-s%)LtzPIrx9P`=tg{v$7wCGa(Si;Ba{d=?VZDKj|#4+-AUGsqZGSYnf^kbnh
zFx*#r%PaGlR85!P*F2NWt^J^A?NCjlOejWV3#vr0uKgvV2)&vh*EwBH3=6smZ@W!6
zlv7jvwjN1TMwo9T7mfJ*yf7trkxRgmTGO_xSF8DPcn=F#phH5QF3H;6RKj+%PQA0i
zj{HHIJz!Kp=yY*NXW8V_hohAldDlO8N`EpPUcSpT+8saKRPo+Cto->~iL2dRBwe7~
zxI4DrQ^W=JcqUnIgvR(r4UH>GoI|xezGd`K`{g$|+hFD?_^xp8z1Q!%<Kf2YqpZGO
zWjdYaRWdpcDX~#cI>zVq^X?4oe^5Ey<tLb!RqE9J#fp+>V@r;IVROSF)#|wu%v<yQ
z^MH~}x71RVQ_D7=Iva5^@q5(knHNP$(2Z<Kj$Hh%)2{{6lt%04_wL)P23@}`*&Gq#
zmbl#WQ0!aeZI)#zID!Plst`P^x<xaemB(8sud5rf9$vcVoW@2E@5mVGMCYG<u$h~G
zThCo~ugkh#RA49zw#KIyg(o1)2J5tZ3$Y8Nh?Wuy|0drDbN}t)X)@Wav;EEEVj)MN
zRGJ{a`W@m1v&tY}F_hdFb1<Zmz{2d=Ux-7N@g0N>3OnK|Ngi@UrY$0B@t^7$bYr9^
zQoQ{NP>TYJu?2&yp=eC_7954UF3=*krs_=EQ};0J)`gz5K_%F{QB0o=UnYeibzo;$
zAv$V@PevVe0i9E;r9>2)<ze~{nU6J5$@bbZQaC;%cj-g!N$?&*1W9^dT_Tl`G!m5A
zM%KQUdGMSU(!|+M`pxDmU}+uMZ-FS%LnHtKJ1`$VD9gk4`J1Bn{D?P{>03y48;X_q
zgPW@sfa}r#9fTTk@u;vt)FB}!zWzdam)IXIa?=-FXOM{Jw>LWArW_;El#eW5SLv{5
z)}Xmw@sxV>DclxP2(lL6H8rXNztvNk3N&&A`;4T+nUJURfkgBL6G3Ms4{@*)=Kh*=
zG9T}z;~Bq+OT(HsDlJhS{BmZ1s*3fUAEiw87B8fl<xf~KVY<BLeo~@Py!;L}VMFEg
zkV#G06{M;0g6iNrJ|{0s`uqDt^EH+&O|zWpRI0w+hu;cP@I+#JWoYc0bevv#t8GBj
zryf+%`3Z~*0W0l0j;>eZ51Op&-|e~m9M}XC7!z`EhwdNS&Msa)D#Z!kyR2gYGj&Zw
zv^eAx<qcseCjH)2IK<I$7ffbjCTsRUji$5U`wKt5>E%>v_M85kS3FjYcYWr+>t3v`
zLuGU?;D=sm3RO+Le<j{)3LU!MXK#lU^dtCQj~mIF6?8csOCQf)R;&Pv&(Z@GjvT%B
z;F70lM<IFRI+DxNW!$fc)PVn#7_T@qjXJN@ghmzry&okn<L`jdz17tomi>yRzchY*
zb-ND;nV)37Tp+qm2I`iwE~mLM{ILN=|0u6tIt&_UpKE$<wP*ai{+S|2e0#I2gK^+<
zmS)EmA*hBUP~iLTUaxX!jG1DUmRc#+ub&2yL}{3VBh(a;2?KJFs+<!Q2J!IJ$gAq4
zx6%7s^V|h7z9NF*Us@Ydy~n&$Yr?j6fLq*W>)bUOM5iF@gsDv9h)RSA9Iy;HkMJbq
zI)S0nObBlIvTPxjI`{=3#M4_?(Eo6~&2TQ~?e4C^+QkVsik@DYe-dPhNO(xhVHw6=
zgFn<KOn6*%@&iu>hxv+=eDbSRqQuL#KNzmK8ltNN1KLL##Y=>_``VX_g4>V-FHJTu
z1b>r6k4z<uPx*^ajVhjOme_-Clp2d1sGfNDGtkNSs+A)`B;IhgOCC!xQG!X)|I;cq
zSgf0dt5=dkJfkkqiYXnYfPwwI0N$1=08BvxjF})y@RjDu1AmfH<Q)MnG<!(I9d?Lf
zF9l%QK~0FfSPER^i19b@iklOJHQ2uly=E-r)$-(@3AT#@IcVh&P^V3Drp{=NXxO8r
z0|cM>u?UTWO~QI@FmmK)Te^!bo`xDHS`R$^u6BM$9z^d8ezf=8%wO$1Ki?zb8W;gg
z?t@@h2S(hy4|cMlC2gd1YVk9Nhp&-ogs=P7h}U?a!R&m74LMQzzkm;`e3CkR;{l|z
z%muQBf4(QFG?<#5!omprI7(sPHrv!=;uJ0COnP5EthZmEp1=GI4CZr3j+)Zkawy1-
zUz=_`y?)%|qxJk%OevL4y!+mze?QibG0*x@5WK-|Z`f1c+TFS%dw=IV?N!W9H_iXX
z)p&T(p4Ojli^^DHVW3q-3u>v<Zq^sRTiHdh%DjnFlq>)DJji+-2uIoh6RPts+?{*T
zDsa6zDovlS;fDWz=1TV|R(2I8T^;3*x+nx4Je+L&GZvic+A`XUdudGWf<|~QTTj%Z
zgy-(kOVLrPSw%aK(PrUBizAOjR?^h@co_F0TYv#b=Zjpj0wNlTY8HDVmVvh>$!iy~
zhE7+G_kRJzp7+}?<)3a}-hi4#;n<>)3CD6JI&dly8BRO04cvR0K2-^uFKTg>M}34X
z^49v}?e~KgSqBtLYHHEJUS7Af6%f1lo<H5h%wxFV^d@R;s8JUyz0{Zz)GMYQzKZD#
zA|D2VEfb>xifdcOO;QG(nN3_yTXbq`6>UQ(5_&<8aHH6qp8FI}ZDJbHK|d%#pn7eq
zQm`dB$`k>gQEPr#0zD$muJjA7b8%hi^PfJ^tZv2LI|7sldhb-F5U@Jxe1aGiC^3|u
z4KY2Bnz+Yr-emP}n7cu9dMKPHBO@eGu9uh}1zksnQhNqLlq@=`mze;C=HWB!O<?T)
z?gn;MI+YCe<>eA8ASI_T>5&ZM_muwtJ8G;PY;h}A9a`R=Kpo$A75>1#>KL}gIWl<w
zVFOS|DHo5u7_Nv~$5nG=kHpK-Hy=cwVrzuQq^sN;1Z@{XqgTZ_ghPl(0#*bpXH!DJ
z7_o2gkW&GbAqLr6?{N4<=CtVlgw>lC*W6zY7BAA(84;GrVg3&jQKpxk$ZWhoPelC+
z8({A1hJcW}_2j?+t!9heCFS#=y`gprPG#2(=q0Etlio7`-TF5XCXxe_V0j16>k{n{
zAT9Ne{ke~)uTPy_-2?@zm6{aaqctM6Y5*t+UoTs}2;U4OS$iAzdCH#uI%PTGt`Sha
zzb8nfmvb59UOpB=AST1ed^S=%po<Y!6pm|cNeDcvCA~3m?7E=vNU!qT65K2=RdPbZ
z6nzhV?r&Uf&zc$@HetW-{oA2BXjne}S=VpgU{Uz5=pQ$oj0nN!tNw**!Smo#<ipbp
zp0f3eHq!OHKp~L#)8+^TcG@2T-Ki8IC^%zh68WHI#o0DH0O6KxqCA^hIt72Ni81K=
zmAS>&ZT2j$iK)AiV#gO<<dV}HfmPeQ{7B$zLot7A^GAVsXyvU4Taks)C<`L^$bGMK
zF5xWa*gI*>=>>Qiy1Hj49}N8!Kg`kcD7Ak7s>2`Ms!mA8(vGn7MHEBM<PvlGS2Q2$
z{tnKi^p4oqmnELsZ=3Z)IX}zBt)go&LfO4-H^OSFtM>Vdc+FaEZH#;;t5`&@-2iDN
zzb#-Q61WbAuB$%|d(>M=e{xUj@NIhV5dVC_X+9qQ6)Cg0_?r9S0|BV&y2IkxNMCyT
zD|>Zooc;Ks#MChIFzYbzvLYfTV<APN@IplDF=2^8`m6$<`t~Jii?!p5MKby>q-5bw
zomorsFI_;iWn$vrPmV?qQ)6qsQGv<Gnpl?C0t%B=3+-)`Fq)L8n^s__>NHR;Rn6Sa
z9kAOHcaW4$spYqOHo4uRF@SFpcY@XjF++@qMG-3TM6APDpxeF@BtvQy^CNDZfxA@=
z-J>)D=isFm`lD2hDtFC4l;UNt0#w=R2eD{Y*#StsLBdLNb5;W00X-79Z?+R(XDXzy
zhXo+AS7{13;;av+zgtCHL;yenqL$KWu^Sq6mEc08t%_R-7{!C3FGP{3ILt#7VQ#vy
zJOftxUtX@n1f8uKX}))4%xfZ};0VN62(I$T2j}WH1(<Qfv2#n+^**9dS@m^1#oM%N
zIDV*4J7%atO`H7Fj4&;XC?4$~N`^2*3GU-BLX@|koHE?_&;{u^Gt4PO61g#1cY6Pl
zs`$bNqYXz{H-=rR<}v~<lJ`hp+t#8>6K_O+iT`Bl6yYG}4JjQkp^}gG515_K{bm&N
z@2gEw0919s6J{^ODEt3$^_O8$eQ($>JVS@HbcdiKFwz|=B`pflgEUA>*8mEFba$6@
zcXtilHRK@OFhf4``@hHiKF4!^+w)~UtiASmUF(c>rGBnXS+NZgRHO}%hnP&flzWu&
zI-!3tqLF_mm%3kE?mEEeu{Vg1tBCV6nz-AGnBiwbkG^@5fLmG1oYzyE+q};~Tivmj
z+uUg9Xj`CLcc3(rG!8Q{j-tXKzJIHjsN6tZMX~G2o$Qy(kg?&>;o{-q@?%Xs(R7_?
zW7osqVPU^{uGb`JGItBu1+?QkUbfX}*bH7a6{VYS!qw`OY}LX%cN>p3Ubn|l>p-dh
z5toRZr3pUr+k$%?oxtCg?%&a{B`sD5JFWy$7^Py$1stB)>U{yT4XX2X=={`YC;PCQ
z=FuPe-Mn_uukWTJjwdLdw#PRwx87EO(dWeLFmIBm<7lh}%H!+P#yIo#2Kub1tFNr&
zGl!ZbU1YlZ!$b?rDnja-YJ}#2i)T4O5=wyw=QBc7tNw`RKs@q2Js$F@3fcSg9DG;*
zDk+XwnJ))<-OyiiZWdO!7$whX*I!xpl|Oh3+$pW}whHd9Ql&HbR|<w5Xo~+ynF-pN
z4HfN2T3X9`l*C)KZ12rS#?sndz=%%P3OW4QTwQJA`&U>iDijL2c<Eu#(Qti+bXN&d
z0XfS+{{HMd{I1$1y2AMizsVZR-)XjT!CGUOv8d|^s6wWxUcI?Y;jpktx}JoWOHgOV
zG4n9cuySm-so0+Alr7gn;j^!Kjy=lSfAuTC)MZURfPYQ{x-1mrq_LzwQ#%;)YGN0b
zT?1*Jn_=(oEMC`kMrhC<A86~p`}n>2K=tlAp;oHG8<IM_#3rnW{aNO&hcxwc3hk9*
zAS<T|rhu*v>JKr;7L>@ref{JNyhZ06M9O*uq<UEl%4(y!ToE2XLSm-TT7J3GWKMp8
z&ke2oEJa3&<~aa`m)IO<n4@yI#7+@gika^Z-jTn&6CwG>azrFe%&+DEIDwFEqDdF+
zCzM_|shRW!fY2C<7!TYS^{dV6i6bl(!4`{Y30Wlmoa&kU??HcMGs+>D$UBH9=BEXx
zGKcsEQ`Cf44-kx5e4T`b>6FmS%)FE-SzYo`l?g(MpRiywtee6Xg)S%OB0NUV{!HwR
zhGKm=x8sTG3Y~A=FXDk4>x(%qDGA)hx}27&V*u;1gQDjP7ia5yj3tT0lkF-_E)kA&
zB92)q+;f4q5;k4YnX5k{1$L;tGp<kM{G(s4M^t@V1+b2Htb1Gi(QkeS&(cbI>-leT
zK<j7lgZnbAk$r~CtraJOfzOc(Ji_bPfTTXd?<WCX%XMUGuFybgsO~s#wjK5#(YhD5
z5igtXZ4;9<RYiv++aHd-iun`VPXjacZIAs<nQ1xp$CqK3Sk7}+g(>gJ$lw#%j?PoK
z1KP3=qq`lZ{X0L}7-9}Sq^}(Dz4=&}U*I=-t#ZW{8Ady2#KKy5pC#6%qLU?1Z8Gh7
z(~iqmZLH2ANo2$SD?P|n=ZH)}xTw$Ul&#f^VJ^tgC)A>ZAWV2|{!XCOwtHV+i8X-#
zj$dPGY8}h?Yw~L4Y@#$D?N!;Dxb@AqV*0Ylu-4@+qmp3-QNq6X%pc5DiSjWPBDP8G
z$1Fe=-sJe*T5qzMXQ1<cwStfUN!ftbui&4Z!{sig9^tUubCGJF@H@J{vkm<NL@6<&
zjoGgPaJeAp*gu&VASB9G<aHsc!Ua`$Q+fZU(>N*!T;*$cn=lAodTXUH^yE6Fl(#Y!
zVn%oI$DR-UEzLQvLJoC_jkN=fEz3P@*C?+U`Ar7w>ZCCtfg=`%(s+(<V}EJ=!7qHq
zF9|Y!fk}KCV5X_%k-XQ96eAog({u#_lDgjE2l+-PUNb;2Y2N4qnI=9tcH*=Pmd@dg
zXAP5|0(GBfZHyY5Vb#n@eUxD$fDEQV81AD_i8R0H@0!1kfPKL-FU7&mVuHO$zceIf
zf+Y^&<wa58E}7m|9NOf)QG&$CO4M}z-o~{=d-+X<0GHU0{S-cGqT|OBL;m<B>Wt1;
zvp4a|i-qA$rnQ)R=XWZ22Uh-zz+_OZQ5MkuCvCr6qJMPJ`(VLy(Oa^c<6h{7wjPF~
zcZHz+>iKjmzz39zUi?+Y7^=|Xmea4`DR2yr3&IffIQ*FFyB+;Qa^)px6OlFv#P^Q6
zzwB=@v##|%?UYcuSb7N}?L#-|YX->4$;*uvN}kKuhz?d9=BkOmvW;fEm5IKg%<#+b
zfQ}ehXQCI0pAP22hch3QRWD)$UXsZxQ@+*d-=lv$a<Ts`&uBtQbs<l2J2QH^A#7%9
z!%dY;emL~?K;Nz@{`5kO;Lyf+q?8EX-|*3X6)WbzFjIihf`6U3l~WjQB*^J3nzr!J
zLEZ3^gm*^wl<coXK``&AsdlXCxxvJT#k%*oNV)jkZXfV=s#N1t>uy=%;<ej&)7n#K
zARK#NO(QNou+k6FIBJOtX}XGN87&|%fydEve37-sp!z$U=MR|Is^#KXId$O`d22_>
zcwGA#eNiwiPpr%K>ks-^P$|I{JG;1^E!b4Tt~to%cV#{9p9Y3*>$<&+vsZjY;pG~C
zzc$2LpUtWozErBg)SRG0&*B@B^O*@rV}hk?jWhYBpNfUFiusS%dN^#U^ZBsty{pt%
z3pOp@x}r=3MgyUdQ&}k7K+gx(7(H3X%sq2uR)5sw^Zfdg4C_kb7(a74xX;a4Q2>35
z(4et3rK!1{JOTT`R?$2WC^yk9?aUIKdxn?QMH71DjvW_QaW23DLMQn1=774)3$&+I
ztIHy+JP@p)Gh(U>#&q77*-4&FF$)TJw^EJrlcBw|43y?4^T>7N9Ven?4<}nCv`!y#
ze-<s&IrQ86S0|O(Io-h7ug!p2(pQBE{lVQFn1E+PpJ%!rGzEQB;T=I3gvx^!xaL)$
z4uLG#c1g}r2ennQv`=y7>}!pnF$_xG=fXh)i(+G&f6$~AFxKBf?k{OoNR>Uat2F+8
z)S(SjA;a@usk<L}td$O`l>z90N@jx>#giw|K5sIJyf&s5p1h@S$m5aHr^wj+%0R!a
z5lxwe3mMAVm*89jX$SEnS|w-Y9b5KDo1K2})l=Ze)|>cnqopZ3Uh}}aw`&!)@PYoZ
zFu7{x3)k^#NNP{ZWr<7pmRFtoenM&nhU@W)NrZ%JODC+Spw=5~TE{OZ%+~k<tjHFw
zJAc95aO;&K=c6SSxwrT#mK@Pe<E494>@|gSKz*$xYKqROHg-+p`d2M2%9DH|`x!!B
zSN_rbgy*@hTHV}gfBlt1NL*b`Lyhb@X4xMe3>vsBCi&Vw>o;$=;@fs6Q^r+{sHL`g
z#NUUut_SRX-Z^h8@wI-uo#^?pVGQJ8a4Yqw^5%1(dEyaWj0`-`b)8CzvV}GGrTGpI
z!Sl8vg;Is=X8Nxot5I%dV7a|@M5D)c6~}1$&6*8ixLTt;XZZl5bC^e$bIHLvi$v3f
zmGj7v+XBJO#h7ox?8k+I>N|d&pP=bldG>^3<b~0wffF4O75WC-ano+P&6v-ej`;yx
z<yvS1Lfn$MXk$0cE}rrXg4KnqBIo*<$LpvhgyAE?4ml(+zhuEd=T+MDf$$_!!n%ch
zX6F|*DtT#}F^$!5zegD8@e-$L1AQw8Khq{+soi$55KH_}%6)K!!wBRT#~7|#Q25>a
zwnl>qjc~**%^_g1-mHcKwBrwjRt;fug$-mbSsgduNY=<d3a+qp?U1;bokNH6gwOq>
z{g5k^&=%M^yYo!FEoTf9r*wEgty(vZ?TgxktO}o-JNPDkW+D|;Hf+Lg^WS_&hsGEk
z5tlr$)obLyu;pIN9sP9UK8Mb2bQ1o8$1_{x`d_YA?JVEj%?R=?6ME#fYU3XRW+=N~
zO<xaE5-RP;dCOL$<yQx4(CY4)t{3AkB|ycv#>vLDXsN6&Yg<lknD+XkO2p2xT~m_r
zY-k76O3ICW8orn{sob%deX9^-1qpwAd?9i3;F|~KJ2G0k4S6EWH?Y(9eIZCF5RBu+
znEu`1cHN%xxIDaFliP<w|3MUu>Ft`}`!3&P3@XR1&~c?)@u<Y7HiyS=gJR3MJsc#?
zECcUzcbb2cKibhbA)8AWBPS>QgaRF7xG@Ye6f1JiV^4$rxMY1~uMsAWXBR*Vd!wMV
zO=9(jE$2BTT1F&8lG`X_n&Wj%56ff&mVCd~C4HtWzg)tPC8`T8bX|+degH`+J_)RH
zP9D)V^pBY0qS(=P?mfbx+BXv;M#U<u_pJHip?fJf@Ei3G?$usP03#iaLvNW9@05C#
zPcf<VbK~bZFV0wwNRgBOv~>~8emJ|Q9`Pa-_d1l_h2g9WO=H@%bc5Jgm}spsEg0=i
zXm3t^(1g*8-g(!8CqjYz>Xlr7c9V9{gqgWqs3|*nTO9`jgcG!X^*ovO4T(LfWEs<g
zn0U6a*WYIV1n^&%yi^alKL>E?im}}202KglWk>Z$bgUFxUm6Jz>$1^x*8p&X&x5e*
z9=Ye+n4l3PNavv<x_ucBz!%IN6?%m2#e_tPXao=7wY&fCtxdw;j1YF|B;F84q#*V1
zC-YWV+Ex9lz5bn0CN2F<_$en#d3U66yS9J2fKlF^s*oOu#ca-^cN&qxE%*sMHmZI9
zY3kT!C@vwjr_rXyNs(2eRr*n%Gu&J3Rir9I%T$ae3QBZ4e{ZFV5I@TMUM`D`et)}G
z22qOL-bZJ>Ld=Eg7QWRg^$3p!s;EAl#WEVLg}DW6Zx2g7j6V^-i#J4XY;DT%Kb!hv
zWD2D2#LlG4>NcG&43Lf~@w>7qkysCQetLsUbl2<a_;5<yk-643;C;0W6R7a%8yQ9J
zrsWQSzm_^tN%zJ((T!uA)7$&zint6YMn`Xu6K`sFJfzPQ7Yg7)->97fs2)dP54NVd
zkY^AGsu!JX-S|@R70aNT6!j@jc;5{(rUvqoU;?g4{a+4(LNmN?t!O)_F3^}^yW)3r
z7omx*?jAjB<B`pn%V0y=$Bwo=bLR`&hUH&@j2)AAOdzK*9g8Gw{qxP@Lja9KFN0jI
zgb(<e)A0RWV*9tXzpi3t1BzcXdidq^mx>1sI}<yu=xdzkk9UZrE1r&~Gj$*7T`$I;
z&M=)f&jyl(+HSw>lX2cPGCoytWY)4D0G5|z(?z?E@<PIF@C(;6`-754_`+?z()Oz;
zyHnJY;PV)J^qnmfz1bJZ-b&2u-Hv1Qt2;S}-?hGISsRH=>3#9nO(M3B4N1a*D(h3N
zQxUrq16I4sWA_2DXw?Ka&YhTk;kX~nNwo~LI|>gfl$kb$$a{+yDxch1bvp+O`~wy}
zFC7JqIC6WJSn{S;8e2PPYoZi7uOis#s<3P9FPCXXIri*18j27A)`y-RiftxB+AZ}(
zwcm6e_^nA~!(gG&y2Y`SXSVyY7Uqfv?-~|WML$F+as$u>mzlAx9XhgbMDkx*8`LS|
zv6gp4IHoXdqXXXKSNo5*Ee~=?+&!8_$wds3C+Y(k8YzVYjid?AgK6D2tVgEFS&A8!
z>tfJJ9bQK6apN%%{{|YD<MVw<qGu&o4-nvDCwQ*>O-7h+z%*X%LbuQ&9;=MPJ09S>
zZD#$9%m~1eRNq~iF#9f=B3Pg-|CvL*oqz&H)Vdo+mLMOmwZLbPM}FUnc!4DR9{z>O
zv?Z~{^=*cM>-T|Z;Md@1V7%UxXjdogZg^d}arnta2Bx?S;WVH5cXvw<sPa9l|K?hH
zqsYDW2z$iI>~{DK;U-jks)*dBkSGYqn!ZnYqF{5OHQO-h(zaC?JHS0)QZt`-u~F+!
zl6hkDgs^blv9}Y_tl7ZsduXgX_5tT<t+0(F<w`xk+uLMnpc9`vy^{SFuM@+IU-sT5
zuf@v%bY>kpR}}_tjbEI|Gx)t*B561%A87sOLHQXO(%CwL+I6e1v6qAoB|Yz4tt8XF
z3*9}0fRH&?4yz}2{1rY(C8QP9+_?UAsBc-LrClomlX6Ls!My%fpa~zY<jL+Y)q1-N
zA8R?*1%?6i)0k9Ja`R__Nzq??Wx7Qklz7?)t*x@mh`(U%onAQ|q`M`%8C|pEhXOvS
z%RQGqFKjs$l<GLX1xJr`Ob#PXcTTF!xUu)mPud^gSpJNb2WdP<DHck4{Tu-L{XgdF
zG|w^Q03?ss?-`5GLK)l<B!LuU=hwJRzPDtIb-X`Z+V$a$F=(uoPIPs)B_emDe9ux)
zL_CIEl*n16Rx|#XPd)qftj`{0$78t1o-7ZqNOEXa*;l%sSB%3gd-YPu2X+XbuajK3
z2Q#>KO0Qc~`3P)!8d2WmDKt46@;y&6B;||X%KDJdWdi$1jDXag@VI>&Zh@E#en6Ry
zPxq9rX+axvGXqS6kxEg|F#I~uHBZTe?qf4?>6c6{+8}5uj|i;wGJwxo*7nwe3(Pi(
zyd~rFuq{p$a<}2_q?@seN;Kg{rk)9qF-A=Awas5DOZo_BdCkX5Yoi(9{&+F*e;%Sj
zF+hEB6fe?=8c3-eU69bsN9%UK2V3)R&FcaC?Sgn=AN(y#Mdrjo=hc)(rrf^1bynq4
zQm0u+<1n@0=6ptl(`N0~(UhX|K#szK(cO06Uo(Cq;o>>5Y<_neL|Q<+q-H1fqWjC0
znzLsUHCsQo0waF_%qP{cz(5)*Jt2F+3h$v8=F|NIUmP{&;ieO#qov^;yU7-%L$7K%
z>+M%{$)Elfx+RaJAUj%;HR^3OA*p!Pq~DRf9SFruH-29sOr;$9FlWVaf+JVg0km8{
zxTSq!0&g0}CN{G^8jOh?j20sx3Yia^+M}1x#sDRI9=n;GS^vy9831Ddv_l{BOag52
z+l4zGAw9veAHoq>V~Op^BXX<Q#|0nDDY`miQxQ*Ms_byd`wTZ~WpyNmy`{Hs?6>2^
z<id1oht@GG67S7*+Cm0!?+$w7+2jb!7F0!%#{}Ax8c?R75tqra&9=!`xtz$2Oxeh0
zzUbaH1Va-rw{nuNDyA5$<Aby+Tloc*13bzTVH)VZB?sE}O~$qhm~y(SABWC~Y*I~M
zwN#1fpjkmUXDax4rrE2l7%@MW7d)E_v?|ZKrvvd8PEQACtl3A1Jr;rET##bnjxYD#
zfj&Z(g~C%t44v9N&O8cbWfdh_4Z&<u)sy1<@e^SbD+L%~3iO?q$8ADaw{CogAG0|9
z%Q2cs|NZ|SB7WeOYJ+?04o*%yC%kX88<O8^Zg{nE8!psofZ5*gbbob5Aq9r~ym3Je
zDE_j9)|c|jO+nAuGEjJPdeyHTrhoOv=vD@vd*ag|0xydHvw$8<m^|)tgy)6lC%ZjN
zGztQ8W8kX$hdbYg3wIax%fs9};^#2Nf%3nfetO#oFM$NDC6Qe>BTYvAn2VgL>rW3x
zQmg}h&e_&o8Yv=d{eFFnkM0ZE&piAN+fLr<ZBMg>`Q8Pmj-J>(Zk@C8lK@C3lyOPd
zDVVgWjNy9^C+^F5TtNn2*C{-PRi8Bfg;)6Ik4C1qvEZk*J`HGh*!6R;+q(0?g}Ncp
z_?gb_sAD^~fd;#jujf$(>d>&#z;a(UCH~IK<l`6vFkoWL5`&6ab2#sisSn5saX5*k
z>-pE<*QjIbGIGv}-(I9<zx;`y9sHW0EK7@+BRx|DWEL^Uj`_UacvD%fF6Z7ZWA7tG
zb5(+ORfxZubmr&>cw0fFjQ-j;e;GYyN9*9yujIjtZR>n66cez-L&j@m6_&@<hAZ-q
z9{Sm)p};^@?83PdM|vjuG985r*3YR#%n#%)D1a<H;=9|ff^}RahG5gT4Sk2)@R*Ba
z3Q*3Lc-!3*iv_vO{Jerwwu_1D{|f>C!-AvOPETyPG&cM88<LA7n<q;`W#yKvlzDCt
zDN9bIHTq*xZoReyuH8Q^Sxa6(tOn&s>r74l9*94#0iFfGUbwv3TebIaj=#o8A$}U&
zyTy@T(cnN#vLWC^<)yrR&kNE4L;x0}r!}JD&5j#e`?nk;JPJ|mm*JeFWK(pw&TMY?
zRRc*xZFlKiY*-MW(Q_I3y+p&ycg{ojC3VE7@6UKLp8{Sz*IG(I$HBW80YXUQ{+gg2
z(RhaeFuqgKSJ|Gfvqq+zJZ)Q3_8Ionh5`aWR`oXH8RU+%ND{2Du!yiEHJ|OkFYLt(
z(OS?3U&PjqV3$0hc`cT+i1UENO-e~c+gSC)OVBxQvOvpM_;kQ?>@Ygc^<!2vv4C18
zrsK@V`P~^c?c2SsQkqAXfS<^EsTEHuCVWY?3pW~CcB}JgXo)ZG&pP8mbk?%ddG(HM
z-AtBJTU{e-7hHKf;x7glV}xfy`(okK@2jOcjG2~s>!(2ST0G9_Z%#Z9KB?WK*lbV#
z=&c5OUi-F75x9p%`e&k&zx`#DRIkh%k{|L#J3D6}BvwsoQmoFp!9+&+uYcQ_|9`N2
zBEFO4wLdE+l_S(^`5`jd{_&im=42ucKeW}=Kbm;*{$W7!yI@%O7l|$y(3eP81z`;)
zqmv>@Qs@qgQPYU1dJ3lN5sTIH^*+vpL>Gnc&bJ=sqhXKXo138Kf~vr88tCRm9Z0$%
zXc;~S*ixuF+8Dz;A<^<Xnmc&@0y3fxak0>rZ6!8^NC{*dGN?0!N!E%>IFq&u@n{c-
ziQd=<&5RW}SYP$?9m3DF=Z>!kWAb@ufeA%=E%A5lD*6QjS@V3J#9-#CqQ3cx{GCDL
zzGox*9A+Ob!!*_2p14obmNo6#Kb*?LHwG9Q`draKXqDFjLoXEy(|*KckI7gUW=8)9
zrT1=r4ccfZ5Z&t>jIP#ToD)1cqTiXC%hkyPiIms*y5!l-1lJjebw_;mD11x!0#nYF
zuBNZ`DxWgxLQPM^!bbiMYvij^=Wm_3t)sRZ%#w+R)&Em3|Nn(^;wMG~Ke#t!e+)a5
zABI3e%}=^;DL=a$p|O5pc?+LreELmI%auBR#`yHnm&C4Kb>d;7=LVtl@2IcjJ#R{q
zKCU~Pl9}>z(7D>vh4YtocKr;EP8`CXr~7$q{n{PQ6lzOLjJ;x;asjuIOvNa})0N#M
zZG+nwa3%6FvV!;K^^+%r#Au{z{1jo5XMCRFb}MuJm~r>?&G-JorntLdgr}cF^0dA9
zwB<NnD#6H5kK9nL^ABd&`Ui0Z%V}}$tOA`-G`R_LkDd}kjd%%FsC_J%=fCRtwtj2+
z4f&lKl)lqNgd6vq3i$jyy{Yi;TT*ix!nKKscbc3E4aKVSMRrG7l?1VM^-T+eA}&tU
z|66VajuAeUweBT$=zWVvnJsQ!lZKQ9g>wI8K6}&^f?h#$?JsUK0d@ZW7JW?OO86ps
z-pI7x4n=vC6MT8b^-TrHYCd#%%sp=Lo^zVu|E@~Sq(ypY@~=SwS&s*l<-9~E-t3>u
z!_inZ4nEA$S=~{IT+hTpx7^l@4^Hn|2=!+7sEuB88#oy*_KyF@Cq)aCZFLR6o(xfp
zNZFTH*2z=1Es;J8-=9tL{~azOJ@tFKuqeH_sN6*i4w}_W=~Kt^oyo+`N~m@q-B1y5
zJQvqB9FWt03Yy6=NAcHA=4(!#w!GK(Vp7e+VBOaW8noB2yR%5BkDpH^HyC4D(Xtu?
z7yBd=(diY~1y5K$sf<_(B$~KI-bMNdQ+?m0o4t#kT#>OEsF#tnUp8TywpY<fZ#D@Q
zdC~hLi6{&Mh$VHRd!x}g7==K10o^XqGbcXD*kJk5KdHaXB`>w$0I)XPE81Qy(|XSx
zNwk~%ubaigJ_KG`VgIsCD(&Ape_@Mv`LBa}mIo!c9sU64F<I0!t;L@>>qFnROj~#j
zF731WF6Zr3im9q<j5nWX$C+vcHE62F^~pWtjD9QF@b-~A`Ml;;m_k3WSD}6)N|V^q
zIXL`iI9`$0`FrFdfKQd5(R~#k3a>JzoVm8Kfsm~I8FqUqtEp?AZRwfW)N#bSPqFZS
zeLSTWxO5?%gHKoanJL&ArAPZ&sGg#@qZA*EQ8<}oj7MtjQ{MdO$v4w5HPrYgSg>af
zntFuQjE~>4NtVlG#PoEF<S7?UR$Eb_6-;u@^r;uw+NMXz4{Wbc6pr_m7z~yAT8$^i
z4ErF`6M-hz&I8_F6ZMa#wQKHMahQfi#P8TV2L(_NEs4vZDksoe8099~d0w|$@VfB4
z%jq5_37X=B{*k5lL!kZ~^iJ!<<G>6)L(k#ZqKptiHt_h&KHCl+h3P-49*9fge5<3O
zIupvx@eW%-4CtXW)ElwbBb~(xF>)#$3I|&Ir6%UZe(SpYmwp$k<S0b?y~Vc|ucBL@
zGe_eNZ-81~JKA&d{2h!ObU%tCGvmUcQXD~g)<`Q<cI{P6=z8r2T%G#4`I8vYT&>RI
zhc&Kb`dt%Vw9mJ8ey5{$lCf=>!w(zXsZX&8t-IfM&;1bEsi0;xd`ahCWLv#QG4p?|
zJRL5na112+X}3g*??$Sg)3K%;6a4e~Zse$8NwiMZWZR{}J=x>fV^<58Bv#XCQtrvp
zulD^rez6R%;~G>))U=SEs~bIzA^}rR9uXX0BsnLz8>ROQ%YRtDLi{8OIU7=}*-iE_
z_w-y!W1BD^DNyl!?gc}=>xT0CXyid4(QN+P>Xc3#NDr;i!^c1!{!9SF=|+_sOp=c|
zlV=$NvLP<1z=^7IX;I}Qq3Ck|@vLxYg^$aZ`i|xe)ZB~&uAp5+0dkbG{rfdkIkif;
zN$Lf*Tv+l9`>X>dAdENvs?Xj>Z{K+1pA8+z>EYx_z3(m>2jzoo`TMJ<x}0=WlyE^v
z!iE>!XXGKaSdjBieCo+|W1gGwk}2O}@mrk2-FHQ-YDIS*(VB%*z4}98T5FE(X=yyk
z(Mk57D;dfev3gP;=)Wr&WyIn3`6%_!eZuy*+QP`39r0^(^4cFVfH!VMMsO8Wve{;3
zEG<c2ogNE56bjf$15N|UkA7Er2OJUiJGAL4aT5X1A<dF|pZMPo`l|C4EPm<=+fEe{
zd(IwgjyR~T-Ex-YiyKOMWZ#x#)w$t%vvfwZd#=t&KU$IS%woHO1?GhS#Yzzh3~T0g
z{^mT%aaI#Z?AS-5*ntVmx_ezB5bulGr<GIoi+?{PMT~%8HjnQ8q*UG)H6nVCa-XG+
zR-%g9clr-QLITaL90o5(ww)a9x>5{}ryuTf`qjg?@SES^d-5o0Vlr0+e7p04uB%gL
zZJq^8K@%?xG?EyUrtK!f2nmw{=0di;ET3xk_G0a54<egHV!tq^a*i!GCV%gEr96rn
zeA#1eU2K^Uw~;eE{LFlGC9EIfbY>0~E<}E+*s<U5tlgs@{Aw$i8ZyV<H0hMl7TUAX
z_Q`nca*tsx3wV~h2`wFd2`h>46cp0qeNQ{=gXi4#D>>%e0EDdSIqA~3l3{8Z9<&%8
zMU+DAd*IA^JQTBmtbu~|4A1f7IBINOm+}kPc;-#2dAx|P;pg%Amr*D1IVElKBFukR
zb_Z72v>ULUBbzc50atOX)}K4=KVP!<tOh5r(Pr={){FKPMV3ujul%G)j0=YANJzAK
z=7|eNByV;(PD|;lb`%M+&>u}mB?(;A#PB3A1=Z9O3@5)ydGsEu=jf=Z{o8*nIpoE!
zouHZQMccCDw+ZKb0Zqh4kt8c^cVq0yCNtk8_g9$>RyGQBk@dGC-s8JibrCwaEJDsn
zEUFo*@Un948f2mE7(cF*qtnGSNBC~xD@*$1=>>_md{(#BB7B7sO&c2|UP%^)CpI(w
zR#K6w?G_Sr;+1$E;e*2!g?HCT<M6eXv%l0CUxj9uwPHW=<vdi%t?eZ*zTzXUev0~l
zWi`yL=n+0Fu;-1ic^Vjqsf^Z5dwL8vO7db{({b~KqOo?lZqY{9zbdOoXWdP6De(An
z&yQg>pV0<4YQL`AA=p}Xe!Ah1x(C(>V6-;EZ(%7Y_t|u$^@(HB!?jTRn(FC&Foap^
zZq~(ow|sS^OG8rbaoOStdTGf^A&2@wGf9P<Mkli5>rYA>6|_pR`txUs3JE8w;J<n8
z>&v9wN730Gyk>C3mxn46!j<zSZ>v=#6(!&JVxe`;<Uyh`jojJ<nBm^`Y-evtEk1Gz
zH{m`KqRTBny5k1G+!hb(7|4~H8)`mL(QlbsZaW83$dL3AG4Qs5y-$sZ&p<@6E8;&h
zz_lt}@$JiVr*SI{hsAsz0m_c7rO&G!zE`y7@1!%gRuiB4ZS_WBZ5>M;BLofilPeBe
zD%U0>KVd#uD4HnTd#3m`9mzU7d?pwwhZD}|cQIFC@<H%`eL>}2QIySAUcAzj>i7OC
z5~1>NpLzK}WN`dJcL1Z~vRw_Ixw~l~w4k-$h_?ohme_&#(wVJC!dy<8kEv*R!Nns+
zJ5stYdNQ2ztCLv;&ZX~+=dr9oy9<iln(^q#D>HhSwT-wtzw{3*`fh*>KV^uw_$qud
zzX@Io5G)_9beDV*CsO6Tuat8qgsb^#JP6>3k~EM%*Ba1@oBrMEE%}uSp-Xn3VVKn5
z%u@K4irwh^bbw4s?HDLmJzS#x#wym%p@~Om*#2{~FhhVivLpBWVypJT7dBeD7{eGw
zmuR5Zd&XZv&}eAOz-@*-+S=}Xv4ruXQ=zimj=R%sl0FG%Bo&>U^&wd1ccpMwWn&c+
z)%+Q`w^4Wd$Z<Mhg#Y(jTC2I+D+eUj2!4Ufcq~eqxi31%9~W6|J&32odsdpkD4%b|
zFa8Sj4yKcoQcH`#{zxl;{W)Ybc!8>X^iP7vY0v%L4i1Tsz$QaHne-B4djjsr+v4=@
z&3fKZD09#7!V8PcK#7ZjnE`eW#Z(*W8ub*uz@<3suT-N%mXbN|30QUbL^LWPeUR>t
z)3V_qp7M3|v@->w;P+3`)ZfQ<ZlT-RI(%=yUH-3YmL>|z0H1F$pKoJ&hMgozDb2fj
zCXhE)X1Os!02jYR&ZHkNFvB~LrI?E6XrR+%T+q8E!=1X9zChY_#$iiVKfC&86X;9}
zp#0rVe2hvUarH}pxfV(`fSy{8#eF_*mByXZ1h4`ae*#0LA#dVrN2G*J{9|*5_cXp5
z`DMbL+svhzdTQOd(U?L2?`|=q)!jN7@Slai82WaTcgA(WS&TNTcncc}8`}ioTGA<t
z0drEaK*?muvtpE5hHLs-MH&C}<qZ@Q9&%fuMr#D&{*y9&y?18@BnPr6p%tv7OaD9v
zC`5Iw{K$SLxCd6P=r`(2>+TJUzW|qOOY20|KDi{xi%)(Bv~)T0T94)s+8u7KanvEs
z_2-GyoM6hj=h#_T3l{?h9V?A)7T%sy&#d@Fg@bRV3q4LCnQR^jU1P17a$hwYC8E)P
z1uQ0MdL2k*1|0p9I#`?M0NxP3F+vaS?RH}2-mW<4*~IKY@mrD-#-6pAb$M6DhBorG
zX@8dR@zG4WE}{3XS2paCc~*hgy#d%tF>sSzl6d=f64+Op=k{@wQsaR%xkA=3=I>+w
z1};ukMWYUW_aiIaL>slYr^fNr^(KVC!d_JOZsS^8G<A}HHv1T-b2iyfJSaXDk`dP0
zzjC#%sc|^>%$|Y@m!h+Q%ND#{r0JSec>LM~*LISMdK!KX^E+fsi7LI%%5FM?Js!bS
zI@lfR{uv~;4gIG=f0>-1@L1oF*ZQeBHURVUh=&mm-%Sa1u&P8`X!S;-$%$&1Y%in0
zZY0p=y$J^BnffRP`?1vRBF{F*4p3@=A3v;g@XfY76=+l5YakT{9uvR%EM?jEa8rK!
z7kTiWr~*f+_-!>ywC@M2FmI$DOb57t+-s&`ZRfX!H8^i-rlwx^TK@1S<|wJ`o2^y-
z{?jVVA?GiMkIeHO{aUS(aYC(szaoiz`1#>fu))?o8HHb25Yl#9A)Qt+_JbB+Or7+J
z#4aF>4W=40{klnZwj%mmp$EC}E6zLY!&$vy_kqF6&)H?6d~yt$?#)8#`n!od5$pF4
zDKb~rWtE=DE%jU%6l1NN^2C|pmf$fkl4%pPuwYAqI&Z&ZIQ@-|l{6*F?}oUe97%x>
zkEEVdk!&$uU%g3=4X&kz0Y9JINS516Jd85)E%20uJf>m3SSTIG&H6Wun)b?U8>-|>
z5hq<Et;d6oMh}o(sBFmDDNZ*N+^1Y9q9nlkkwlIF)msgQ`pTsuA!ran^$KZ!!yTdt
zz-d*MAAk<RL-sMaK%qi9z$3f`!t#A{S;Z$+H-&MEj+uV_*NDAhs=<SZKP#H^MBKmG
zg7uvJZl}e9veHzps2uO?6@{_q`Sjh+0A_^$1d~736&<*c4LoSA^V<Rf<d<6it0s-(
zvgvE--Z>mt;i_ot97B1}l<Q%|eDj;J@UJl+w!V5Zrq$#PUTvLp2!eP&hc`}jCuc7p
z4Om?Nomv2MwlJ?UrdDC%o&QO`M?4S2L^0Lmk8bMC^%ZA&t=%Oft(ngf;$3Q9Fs~<3
zDxTrz`^RYGWx_aFFLlcOFl9J9hjM1^`zV1UiHY3U<$X%lMbX;2@3MM#g&E<!a>5~A
zZlXFSz2WiUGfk=HM+y+h-%n!YWHf{|e`I6FiS)pEy-6wRN>W4F@pY{6OKHdP1#9)u
zW7Mc{5MYGuplzv+a{wpRqPg}^3J@0Ny5N3{P})jC`J}AD+a=mhUHWc%7wvm<XrL+A
za1zns9eXdOBIhIfd&KLDg`UW{LnV8^Y1&bp!Y}sveh0x)yhKvhh|o)s^w)@i4rHV1
z`#3dVQ?S~}no0Zl(aC<+$VVo4I!bz#>0why#|iT`I&ap_dbgk$St}qHa9$*Dg5h%d
zAG2rv{l^bk!5W$mZxz41j^{W%-pHqH8`MAV-fhVo(T2s%If2oI>LAe;>5k1nCj8u&
z87-X(NwveqS(`7o)Pr6xVNJemQ9}{!#(zp0W#dy1gJxg7si<^ctu!S)-Pz-*o^(&1
z>YtcQ9f&c@`+8*;bxVE6ku#yY$!5K=4s_mBM6sO_h&yxBmEWgu%!oY9u-eocx&3A3
zlv@92b9WsZs|Uv-jWtw#JAgh3@CYxr9+`~`1TKlIrtSmf(!$IsZ$~w2aMeVYR`242
zhiLxx%mMioIdyTuO)EpK(pGZ5uKy6hBh)&0C;h=!S(loe$#_3Y+1L7gD+WYe=OX*6
zDlgETmG(j_l`S@Qh!aQ(U`=YM4kcPLmFEon{&A|2$!C#2VeSn^^HuIKmvlDNO<5q`
zpV)S<X0amNx2R+4q*RB?7vBLg!OcA0hkRj`t*`}j)<#nX$^XK_HtztqOq7Bz{D?L6
zJPF<3C1YAuWnBfZ<I?cdGi-YOMN13ASxBaMiN^Iw$d=z6Q}@(12p|XlT;V&*$d&Xc
znQSm)FAtzm<;;_(dBPXXUzXzk&X}8R+(m>s?9pfS$@pLkz$(p$09w5*vl%3ax2?ts
z7ty>RY(G^=Eu{}=@_BJpFQu}RCY{z5It1J4K;Q5w)N9y|N*I!xFvMAyvy@#G&a=pv
zoY$yHuuriT#MxAh=e|?X;u`QR4<1~qt+bE{1smBq_MWvZwYrR`siE1G4Z29{8uP_&
zp?w^tVFj=oSR0~=xgGHhFtp_(!@fDW-=pk8*JKe@Farx35WDAeB+3(L-Ru<qEM_cN
z4>L^P5ekK;q@F01h!)_7VSL#8T&F5|hH<gh&>?U-lf>aFY1{s|da@j%Jo}y}ivetE
z>b{U7*R5KtmWJ5f_KTX}7O`Hg2UqTW0B?&)-9|S(`H#(04Owp9W<Fipo66bj1YCAz
zzlXziX2cZ`h|Wa$s3jDA7Yv`yL=+7T6{<uX-;EeS<ENZ(KT%TQel*=VImU3y7oXd#
zK$)vklNyKYz8dtL<NcW@P-6Gz-$}micZ=H#F%_`!DK&?Ft<C6xE1e2Zr9iok9;01-
zp>sUpmsZ1K1inl0@-$oOQ@_BhF7E>GuP7y!TlLuMr}E?=B*T85f*jN0oJ6+2&g0QQ
zIT9Am`<0`TWg2(~dX~>CAE3BjAkp9K>z}bJ(4wZheEM8D{#FQ6;ZxL{ydXzqHd~E2
z8V2e%7gylVpbuK`8saO6G})+{Bau3!AM@!{XOpah!z==C0v*)r&+}~{D7z}=i^w#e
z^BfTt1ixK@cNo=Votm3__uCz#FLaKCDf?uHv$%3Kt^;U?cNiF%AJAj~${@zR4M_+-
zgrJkO6K~W%UWV@eN0DTxQ4xT7taLWuu#ups9Uor_O>Ub^vDFN7VgY7Z8f8zdR|$y1
zl);G#l8@ZfUH>&I8%IF=O~vFDmua0NTg^@hqT3DMSC5Hb7T%+>CjEcz0{FuH!FTK@
zbqcpui#&T3T@k-gTJ&|~u>SLP=FU7dxb7Q2H0fK@Yhs7~f@q%SNyn7Z-vqoc`oTH)
zs5uj8{WHM4vUs5;$T;3eujnJ$^>O2VKXk(&ZMM1ptW8(kWkkj3zl~a#RaIdGfa_w*
z(Fky*w*Od8G-=b#f;!C&nP}IzIPb-d1y5tiwO^FzLB}5*A|(s$NY{m}kC!mzShfc?
zF`!`}=db^%>sa!Kj-QD@0pWLk6YOrJV2meHzne*-<g{{`eFgA`vHg>$gTU-UBT+>Y
z4BK?S`|W`eNDv5&aTF7&|5&c>P;AhNWBZga-!Yfw45#k01N)v%XX1u=P6XZ2irQkz
zMM!?+^>ib*7kWI?nhqJ8C;E4(UC-t<Xb>J<EO^ih=r9s{#5tJG6Icl93JJ+{{S<MT
zlP1w%)3tT6wO`FpaWz<zU@)CES}PGV_<YVY&PKu|R1;aw!6;>?t175;!K;)LUc#>m
zdr@Xl<_q6nJnoC*6xvVe52$AVvwbx@efYD<cXZ0c7`|;ZF8n;+54DFs-?DA^h`zk)
zN4ZYcxCvtTi!{Kd#((8i7lIGrJ}(*Hd|dQ)ou5z`LK7O^yG;=I_hWUIyWIY444gp*
z?oEk`84N1z{|z9r?ai%%;5bf&n|JX>WfS~>VD5&|+lGd92IP3V3QN6!P)3{{OmaHw
zAqZtOffiekSJM#IDL|EZE=Ha@15?l}f><5$L@fV(9o?>TB1iBz@h|H9O`<l_CD`I#
zlV{-*e?Le|jOA2N0{B!K6w(<S1BmkV=Xv1|B=$B&8$B<*b(GaaXF?;O&c+jDXjf1B
zTPc|l-R_j-@b@ba^jb2)$>jZ<y?n){G2SzkXhR(f>2d^a*Pcj<km~fWVZhI4N_SXU
zR^Rk}4Kd)-?G;fktdf^NG)mj|;b2J{nWc&(_JinTu2EMJFvf-9;j7bS>nw!=B$tV~
zIjnI%m!G$XuTh7+b*3Qs#gO^eln*iQIUgv#!o~jmhpcHmZp6zd7mDg;&4`jSX%-Q$
zP<l$V6&o_yp>wcRd1oW8P00ZFry@imNCugQTa$w7GzEHS(&tezfkMo%`}JF?<jluM
zyL7*zQHf6!x}@Md2A#8Th2;T-5oN2z5Np}2F_LMNo4CfOv@{?!v3W%vcE|9yuAIN(
z^AyIp7%#~cB34aSP+p=FM;Gr2F{Cz_z4Fl()BBJ2UZ{PA<vpduNfVr$$HURG3FQMw
zgl3<GOjuNSp=@5x<6;f_iw%B1;ywOF%a?Q;Ay)85|Htan`O)5dWMx*+7@PO=gz`Xb
z-B#&Qb2~kj{=(^bMJV$0>^ge%?^0<4y5_J)P)f^M@u}EJ+xY!uVn;DvByT=>K?wTG
zD<MYj%*AS}=}>Pxr-&PSqubCf@%UU#w=~nWhXkSJkaB$I`4fc<BCT!YOl%9(ae#4$
zVb<lLQw(OOD*h;(3itEbIWUL$qkq{09sy309K>E*2R<2foW9VEBU@%_Q=?HA%fltl
zE4nvm6>8ll54t9ZF`8emnO0phHhcc;|HOKHKnov4j5@ScxBV_r>d<Bb3o6{erQ|aI
zc-A~hQ<uX<1LZ^cxGBEaRnux2el;{cYc?KYoA~H4H*ImSsPA$O5wkTBPwpM53b-n>
zc&Du<0`2F*qk8)a$XZsurF%?No{nl_^i!@JRQ~5uQGRboZTqqV8FK+O;}Pljdq{)o
z`7gn*Os?aL>teM*y>eATlm=&Sx+1F{MQXUT*M6uHx3T8V?CS*+(u4kp*e-0fygLf*
z2ub5-0{PDtLJ@SjWX+j9qjCJt78ER48N;#oUo+KGQ{>A4*SoMJbt!VEoH}{1J{KkB
zRd)vR3r-LNl{+c1+Xbu7%{Jmbew<S6x+gc5wc@?h#ok0?nTbGfC2On!UZE_m@tyb(
zEDsVwvpjM%DTn}K7~<ARMb}`$ZZo&TDtYc?7ZLIXf8sM0bvJX%K7mdqJ8zGyBr%f`
zw^0I#(;nqNbU(}KEVK7X)1QC~5TefK{GS040PWwwdpro#JoX1x<-PzA;GJ+{Vlrp)
z-MgOtTQLm=e9Hx#uvcMf%eSAnt6$Sm3HiMDz|VX0c`iulVw=?$b93DpgON>w8VQ>0
zx3VV;K0G)1X7c$FaF3%9kVZj3!@c;uf|{0r_s>3UlBZ*V)QzfYY8mVK3r)piBi)C#
z+&6SFlrLb>c&FQmnf*ThYkt&n56#biZ?z<Qw^)fO{jdGeI<YHgycUoXRq8&7O_duc
z&kL?kiPn{Xg|CZF@d)jCH|l`52ZUCv%5PWG&C7Y`&pZ?^BymBX)I7Xd=bnLL^1dzC
zf1iUt&_WBsX|+CN&a>@CE_HTn9eY6$+j$a`Tk`I~f~Di)bxTAFNs%N|KI;Z2p#QAf
znZMW_l;+r`(0TvPmiSX;zNqj(DwQ?6Ofcf&zRmneozG<Ky2pCr<2G9$b!PWCeFSfY
zJ(lnDe|-*eI-qw;HZpulEk>Y}F#|L%-^&Xl%c+h}<eAWWq4wK&cxz22x#|46-cHGR
z8R-Y~2K!ZOl&Pa_{E*M6byM%lSuTX6GxrIxBs7|)N{qZ*O6-^sKe@lSXYzVlmRk94
z*L<XwLUxn16(O>u&(5ctREcL3NNXQ0*BkCvG@P3OIDeRtmo$~zsU*5Zy3D6a;p*d&
z!aG2Pn2_CnkJya7Qi!41<K6Q*qlaCoUkG((v`%!u2Py(W_~^hJ;jgw&jMDoS^M-}L
zq><ZkN|9-XPs5490k0?y@8E7fs*783oUC#n(LUIODdT<*p~a}Hfl}s=F@#jC3Sza}
z*dw}^Kz83Y*;|ZPf_BI$>?ML{IiG}e@?NRI<Q4UkjmN*P&ztBaATE{;2XD&78;KKc
zJc*QQq!87w2PwRyQnhc{3$x#D?ymm5u?pafU5YDl>nX+$<6q-LnWAZVvPI3bQXB1?
zPzD47N^M4m;s^u!sb`>fc6XuZ@OAyg-Cn2T(FcB5%{GijUcoi?z<_^XcL(AfuDK!<
zVXviV)3?{h+v!h%-Y%2p3@*q|pJ0tb_hSKIBTg$HMHl{6SgJFnq<On#eOWghI|-Yz
z`WJ`pX1ivSLA0>R&kSHZoz*~lseX4FhHSdHGWI9P@8cLe{5fjo(W#)tUIV(q$O5FO
zU(Y&H;W(<80V(5ce7rxSjnHuL0{+&%Kab?~Q4rVueJgT`RYugTR?&&G|7GsDQ^tEy
zM$AJmW5W0EAP|BTPRfBklDNI?-di|SdSBPLn}k)9!l-rd;n!d3gU$@kFIv?AF=AcN
zAR1HtA%dQZ)C&5q<pVM<T>R8HCR*y_KK8ml@(sQ>&KupCuVE<iQ{<ClcLbK=(*#&2
z!S0APStss&s(5YBI=fXE4utxjribpb?XtR|vk17ithyUQGICGzM0R&<$=9srlS8Iv
zFmTq^pK1!NG#baU&y9Do=?k$sy1_j?>=O0Ar+RC_W&PbDxP%tkky_5jc6xdI08h#l
zEbLcKp<#uVdeN4tlk$xz?)IQ`aBY}-19Ir(gqM$;Yew!MhTnjaN|EVrw)fh)=C=V{
zjs@Jb%%9LZi|Zr4Sj}ls^4ySgcCnsNzsqY|QDoKmPxAc<0BfY=SzhL^wt+%iU7RGV
z^AP?dXPUU_M&@+&#X5@$WVqlO%!S%f`<5+P1L^jJ^tF+*LJ|9Lgh@nnQ0Eyhwrl99
zr0rFWLAyoZ?KqLRC4kbd?Sea9?pw7b%I)xN>*)$u<9uW1+#=142+V<yWSF=<YkzVd
zNMGt5$9*QU43@)(;<N7={}dc!jg&G*1^Hy0*Vmb8U*J(~Y?#Enax$y&^(<)ZM7bA!
zKW$H|r3N$jjIMZ$Zy-_yzgGfC;ltkFWE>szIezl*(bH2`*%B;Vr~=Ia*pk5d71k(V
z8A+kYrc`7z_KNUb*z{5Nj_Fm6rfXU)9r?h?jbhLp5ksxC{-iSoQR9zC5euGywom43
zzYlPX_%FF0(|>fIal7}qSbOT)dPL5KE;c+t;<sLLq+7~l)u@U=t({91)RF%B3<=K|
zVNIHl74&^E=gl2D=hDvHX|*+y#d<aIx@qxy#1@=~NYeBg9|8Gq1tcCnH}S0aYA4Hg
z>&mR6Cd=k6HH!zvZlH@cC8l%E@aOLcf9vs5q%qc@ty+CtXNw=J)7(y;8I3VI@VTwG
zI|Eog>D)-t<2Huh#!w_rYx4dnB<m7M*7KJ$wZ?Fw5nsWE`E}FP;dMG93mwC^q{8oQ
zb~d)@1s-QVc4%e(XMr)msedOIxCp$cPc#G@UUc_dw6pD?2uSsQUw>BYfc4*;`WG?{
z)JFj^`g{^f@Sn`K!`gy>WApZAgtxY^p$iNB3fJ!KOdPA0o>|H2Q$fBRbL7rwAMnuF
zd8h8B`Pg#nR<j?LRw{^@FjL5gX-G;_L!ELY^G<UFcQ44Hfg+IS8MqVRSjKf=CLMrl
zEL+r;|Jhe&*bA+acb#496OiPXzg#$3%oX0*rN=~v1G`EddJN7QbMnE%qeD`7s<X^=
zCFsp@=EJ(L;_!`CIgNlc07C#17JmWd+sH2-$K6iF^LhM=ywq0E!K&LN_jRFm+0x9|
z_h!gv#pN|0VLVZ2-O(bDC;?1%o&D$W-mY<VS^(jRLB}h3e9TJhtlkt_O$jFH$TZWa
zvTel{<r4ADtY6<0o!%>%Yb+t~d6#E7xQIvHMJ$fzx|&(IKfcGv!hl9g`ZRdfg@lGi
z%vNUme5RPHl_ou~*zs>&AF^Khci)$#9KXZIIiFz0K6edkv^QL7JR}r4T6CVE^ZH)r
zH#-~mkq%Zeq}_11P5S#Fk(&}b%t6)!1I)(5hk(?|FVPJT<$0BH0fk&e)ZJiVb1~;4
zbr~;FMcWX(ouYSKL&FlAH2R&@#XlNSnX7vCmP7*CO9qwYanBIBF3RPvGQb66RO>tr
z!QYxEG58)jwpfj$Ms^!tiYGTWUv5$+ulKk9xtp18dG_D^-kR`O##;!#bT`Mk(8`aR
zri}u<(Ger>s-}O)apxf$Lm$Q$8Yn{u#_0ws7=F$hMT!~LhyY+c!8)FTWzcs`PFhwl
z=UZ=SA5hUyYUcS0g=q|R9F)NS9pgb}8e^iL1PnbeJagg`6x|Ha(g*%6Z@r~7zsZom
zD$C+il`!jbzCmYXMBFNTTyC&?%joyy#oC3DIBwB_AWv4Jb(10h^n0y&CaYjW0;su0
z{X{$`fAO}t4*gjjPnmByzN`5!hn3RMw&4QTL9>pTB8NR?Cr?72LWtUVtVo?_d3>$>
z#YuaCtlT~SSyXGZnM_NHQ2EAq)8$0_ue4Yk?2<x!HJavz`14^Dg@J9sbxZ2B1{1=x
zm)e_!Di^+=Tnps48=WOIzb@zLz$&4q+*#2DKlNf0#V~j5XYK$BHXqA)=H)tTZ$DbT
zah>rNP_%1rjqN@8DQwj`ZWZ=kkusnxZZS$Q8FuCbJ;`v>%K0`0kbC)0+N$$#zN4qt
zQw}@r>%3)p-0#<pLZ<N1x$hqBx-&vT%gzYew3nHLf1b}bxNB1u7GfMH_!t7D!*b|z
zpy!LPklMvb!2mj$Y4mVX%c?Ab+rzBYd$)-8bcy`ZG}l~$m{={Gu?3E8BaCT{q|}r?
zqhA30^X^g#ZGJ^d)=^Ux%Pm-Pch18?8i%chG&*|*R(#=XT4!(k5<VB%eWLVf1hP1$
zllnLt7%!`<Y`X|M*W4_6Ic;p&0KUc}HrMLJmX4gQCZ>v}1OthIS49knl-Ezh|5%Ke
zBGi<B6BOyI%)fE}*@H0jS5CTqM-hi!(Z-u?=FbKs0+49E|8Oqbll%YhbPj%X2khTJ
z*;>n9TFbUuSYFF^C%fenwy;jNZLLnWd9s&nn@{(Bzn<UU@Wt!8K6qb-J|@k>A6`_p
zH_^>jn#pm^VZ}mavNS|qO2=D46qPQQswOGbB`}iti{D|+MvTG~h!|G_pNFpp_heZ?
zAJU4K(Dr8o<itqJhH@D_w=DCO-S_3<Vs2;p!KQ`TCL!GP3N=7oW7EEZ6o;Phh^1@l
z|F$4`dkpewZbl-u{zc8Vz320j1u+kOd(!ozQ$6YTg8-Czc?!VUGBEkJL&fuI-i5={
z{?cvP<{TMtP{^JXIJ=O5xCY(cnCCz6;bUNPUa`9)8MPowWg16G-KdR$=o{rLmB2*X
zCTdq~<%6B3E(ZQ#xIz#CoIeo(D$!XXFH_Fz*TShNwllTBjNpY)|1N|*-GNn-f^RL^
za##1;!L!YcPwh{Nau1SwB1cRv<)nHcIH{;slIua6W_IcH1%_+qPdm?(ES*TQA62u~
zmVZvy^dr+1{x=H0?=tmie_jUFMSh;apA<R|H)j++0z*LE)3dutpr@tY?+hAlJgEN%
zy_Z@XM?G<V?>6Z8>AB_cUr^Y$Kot}eDt`vw_Wv5^AFpSpo8b%iB|QO$C2DfoxVLH!
z{>T+NCd5oixLD^=H7l(h<1`wjItv~CYTh@A!EyA<BAS{j!oY8+Ph)yN`;l~aWmxUS
z8QmvbDu9;->#bRBorIWmFXM^{C~S8}QPDU6QRtw1j<j!kQM$>9UD|tKDc-%w6kY2;
zH<@HLc<$b_p~^vkizC($foZ--6kT#Z53<DH`_Ab!ow5EHt9bVsO{!r}$k(cJD_ROn
z{P<qe$HnDc#Mq+xkE{<2p53C(%e-Vez^)f7rjLSb)~jVY3C$mRZxIhjo7Z<G>8AV6
zHr6zc$I2N=a8Q`Nl?+XL*l(oqM5~sK)0CN8_UZiTlx_QfZqx8T4u#i)9nEVfPvqLo
zFdDRAJWO;j7M~3^>Po(CHzg2)jxR(Rq_|6brKu?jk&Gl_O73bOMr`z+$UT~NcJ_@(
z*KbJOe8l&4GOz(ZL@03ol{18qK+N|HNGPf~U;+vj3s<<@qSnp<n?RUM@NCSPB>-f=
zXUU~)iloJ2DlsG&3;<TLZ(#awSP|)L{(w)uuqXh`M|csf4iZ?(b5U5aAW@uWZ!HXz
zKB7e~TRL_*S_xBxH!2XN5Jj{!+!9st=qJ%6OwJ$W+8I$;Q#u$ZcCWf`6tGMHet=Q7
z=L4TO+;4xYVZJzBOj~V4L-=|Cof(@<fkSW2@6b!YK7b9zOnWAWjWk127*EWM2wpXL
z`h&Q;RisuXjjqoh0F8L^V<4ccm_$nGrw-4wx#Wod1i5%zf8Q+FccSd!-S)012;W6^
z!k<O*LU1>-N83Bb&PE=$C4U54d8v;>Ii|Pbbmj(=lPDmSr(6H)J)s$YyWI+r(MZ)X
zw99PXDPo0b(CB@5%pIM+)6O=N-I|OsedYEoseB_n;JtcuM~CtG!$y*Rg=s8+axe0A
z;>*yX_deH-e!5Z{T?W=Xtn{hZx!l!k*_+Yr(EOeCyQD5B_Xzr=63-(`-4|_{`oP^S
zSqrQlNKPVRE0UhG^;@|<iRrY?SkQ-}IKFdDgukC9PL;zy^T2-ueS`gwcZ}QeF2FbW
z85obZ7~jYlRSd=%;9@DyYo6I3n|(;0FT_bR04CGQk`T&i1l8dk$pWOib!h(A-6((7
z_ys=becq|^YH!+KR+IcAi%aN)Fp5IC$X^D2{%W^S6Q&18co4T(4uu1ZFD+O4WDPM|
z*LEh<YE;%e&UFPnDNYWy#zJTakJx1DD8)>*bg(&L8Q7{2Fs!I|vPqM$P@~s9EwbkW
z5%1Rt$zt+c<QxnxS&UG>nNUpZzAE$`cu6l&jfDw%Cgf)XynW!}wxBcfvthD?T4-~s
za&U7av*rDa7v`nMK+}KVrOpOSD`7rUvJP5D?k`q9=P78rs7(!GcYye<Qb&&t<KX!3
zOLLoGtpYVrjNnGMBzpsCg>b9lLmrrruYSxOCu}2Ftrk7Q#H9Bl7$FNuY=_OE>5#Q?
z@FE#YmjgPfhWE(<MarHR0Wx{PtsCROJ+pjTPNwMmVn+~;4UYooI~_eUu0D(zAYnH@
zXVWGcP!s=>7{AqNnDsL`X^#D@cN2EJmgglG6AyNEy?iBr3V(mn0ojPA#ds!S=Umif
zTV=VWHub1mhZ%doqXM$BXCw1JdnlC}7Z(Yt(td181Li+kijhj1_M({>T-dpNzvSw9
z6Tf)c_`Z0(%Mfi?XAsa}X3&o0$;qy$mcf7gnEWR}b(EQ{pP>iLaIkT&EVJ|hY%Rai
zyUF_6?J%tG=winrzt%fZQ^XL@%0D;0vqqRo6J^Xujf9Fln9l!AD6+h71`zNgWC5Hy
z^XOSu(|zLimV5ZYy;<OkJ@v*QXH}eAnQ(K1#7ahcz4u!%zt>grdxlpVmk>N?C7ci3
zcI7E5FiWZhf^q0OH15nb$8_53Q3-RHEg@n28Y$j186gn)L}0|x`|xp68fQ4xA4B!u
zT&d98F_~XA5P@wIRgCpw9iAR211*QhCVFkBiDMBYZI9;~?*8%0PoQLbpG6~X6~|LC
z1p3wn!+!R!TJT~|5!dED;;;SW8ZX*Wt6C6L;lV3dgDZ8BZjbLoJ|_yy!W=chn_P|d
z$Mi9YH_QA&je*e_m<|!0xZY1VL|&r)-p+RAHGlpufNrQ-!OXZY9Qqbs6SOj6$Wz75
zLO(aPq&3msHv5?hF6L^Zi?b6AB3WKSpYhw`s$IEA(@Ee|3=<UW*Y`_?=e45N4Wc-K
zJ}G*+`;eX~xJHuP`Q^YvdSi|tObl_<2*x(L0l|7kxV)(vbwPYp$<zV{JMS`WSQ{e^
z^=+mLm~w~K3uygZT#0YyBnF<+IhSaqxQ+UL8s&RgYZ8DrgrQPW%=dL@GLR)c2U3q-
zhXsL`rB-tX34$p9OojF*AukCC!+M>&m~0KrBLkD*BPa*HzVx3V`?{m}XAUPij@rRw
zW`E_F=rF6v>#=~qMsQs*+t5{1fbQrOr#TJkM;cVmFmW_o^czQwdKF36aO9f1Y*b*c
z#Vx;~SsuxAPR$&_QrlG1{u^=DG1X+8s{qP<dY?CtBbMJ5SkOX++NzFa1epC&p+)cN
z4r7=4H;RU)bVT`NTqm@D7ME#>7;oiD!}&SvAM^kS{`5Pwldf`CSbum+rmMy48fkoJ
z!nTt32C(cKAIJ5GyD>-cp>&FuN7{Nd=~k{maVWQ9edwCiqY#l#m3}vRDeEklm_VvE
z!z89d25M-d@kxrsYgxiPd5Nk$g)UH<5``7XteZ_Ywc--(dvZo1^d#UtYx|Hu-y)gg
z+oX?{$p!5)uCe`Lz3K$|p7&BX=~{U*e5Sb?8Dv~oG8G@hSUZ;>Al~IvtkJyg;=EWe
zw8F02Oota(Q*x8WlRO2~{s`sA6EezJ=AHipyrXSj>sf-7_NmvN$uetSq;t?gV0eQ|
z2uD;+4L3DlsJjGup|t;~swhSO_x5+y{Cn`OZO{2{kR8D?-Rh4WEt>>Pa=h_YsE3&x
z$u!Z;*yqcCWOgwq0kHqz_xVs$S6w%ZHCJzlQp~<h=ah^aYNe<eJNb^ZH*AG$3rhHj
z!eejuZwKq^U?vm&21q)4K8#N+{}fNVKzJELRp{)+O(nd?_^Sc`u^}J&%AQh@I*aHh
zJ|4m|*!zQMEBbn5iQ<Zb^_S!wP5^E4qQn(E8%)jElsb>G?egBXuBq2Ip%Nd`JXTq_
zKt2Eq+OVbSGoXKcpOiU~ABO0j2?hjHEk=?@gMkJjS8(Z~7lU&GNKsyXO+$cwyzC`@
zS0hvDdJ$Y<2tQT0k+K4QO;#Fikv?eIVFlp(RKyq>5Hlh?5dPJnD22EgRX!&c$4w>F
zu@xf`f97~9cvLYgGyfD*Mu<jk+>e%xL=F%@oMk!xZTe6SVgN99euBKuEN+s*%#-%R
z!Oj;eux|i5h(%T>5!d(6a2=l}TA!oEUf`0y!?AtBbkl6P1wwNMLEo$*B5X-#o11)!
z!ESo(4NQ_#aiVslGS_uFYhOftW7$)MsFJEWkZ{y9(UiMg{`;?%FjO+WbMgu4I}2u;
zXQ|l!#j~f<8w%=fypjD*-tB`-r>J|{bkm00)p`&_1Xbj>3W<s_cp;B%5>x=PTe_Wn
zWz%2k_&4beyz-&hn+KX|#=c6DD)7FWug^S+0u=6@gK)~HIF!zLp3dXqqG^#$?<}0K
zI7XMPl@`pW8l}A4f%N7cA*a0zp0c$yw|~xw!dvov1@dEzPw9MDfrM+gBXZ5Wa<tg<
zgK}4t2ExSDb(a{I+sdQX1vgnpR~K1zha#gQwjW;g)ALD$S^TtwyAKtgbCuY4lGBem
z@;@eOUP2U3Q}RhO%~~L34>-nwx>LF43SZ8F3fp=aI&OA_7&?7w#1C{8niJY(iC2YO
z{3-7wkF%sNMWF7-R~@}`$5Qw;QKy%NNd)OB_pn2_K~7N)eEzXsu)eA%4~bHWN_uN-
z;^ZN2g4~0a&4m?QJA&=ZyBC+w4A<TfRD(v!qjYaxmxrbH8-1`?>j70|+zxV9|4vnl
zldCliXMM`Q``@=C=V@);T0tT+TD1Dgab?)ysBXsI^`0EyFu6#Y?)p$XCthrng&di2
zlNq^f7n;^~+0NFZ_)YvZw(w6J31H&*zLC42?z0_?f-KBO046|4j8BMlXwkES=V*`F
zD#uBc<Z^5qw-ZK8T(vO$5>~q0B8+DTX@?;1b2JF-wi%5zi@@g?IRL|d4{hw?D<ktH
zArK`e8pfpnU^nS?JR>B&1Z#Clb|M(Ty>cH%=s)1&?Q)4a!(x>mFvFnm86iQ2P$;P|
zFE0L4T@S44Vs}&V1%pUK%L~ImziaUSjCzF8BW<3e8!)Pl0D>=F#C{{t#20ZZ)16@W
zlmfnGSHM~s*NLp>nYfQU3aR>ZV%=}>duoe(jMR&1yGb*7z<)w5^4VOvE>-J*D3mkZ
zpc_WflCg2ZhA}a4h$syc(_VT@P=W~7Wf4@w1>rM?DZeka5(t@d!j{5YW15e&D;3`8
zUt`VA9v*#EKWike7onvX-02e(7hdL_@R!jeP3wHx+LRK$%T0ZcT5`|-5{dfXH=+C?
zF&`Yajf+Pw=PgyBNUTJ*3n4#H>!`9k1`n^@6cq0{ZOjn(5oa*Q@cLq}X9c;}rlQ>v
zSaOXe2H*hVEG{Frgzw9V_~EY(U+vrA1?-N$U7q3=<vZwUq;|-?OHB7~FV}+54&S%=
z-M{)zab_*nPJve&G!eaRNj}u>1?Um!UXKU6tPN|{+PV#$vYrbfLI<)_*fvy3^8<rA
zRs~}lPCnM%=sSouEakdgPK<Po73;NWmJs;=GdKAEV{TL}R%cthC*a+zA2o(K72pb1
zoMoqNzqWW9o?Nn-$ITXElenuEo*ZuAocvYP=0{WtoBM+WZ%BFqEB(wNxFPc2VY(Xk
z@aQMu+|UC9kAvjrMr>Ot>HMx+0g@tTemiNmPp`6hd`%a=oez(Jm}~FMyf}wZL*w@F
zw#&X;$P^Ejt|)~s2fKPS13d_lXWxl3vL7xLE84|M1SE<nOijK-#fQh5ws?qxOs#tl
z%ls3Q893c;=-Zk49;Xt4p^bKMM>C}Q>f?&crbU^-bI*Y0qN-xg#fts)<PR%$046Q)
z(5OD#pSBMv8ZeWsP`TY!jDhbHqBU?f+>_`K5XwO)-d<_1<dT2<FK-jsV4MtVr0_3x
zGO{0P9A`J_7o{P$x^{Vh>9JCAJtJdw$n2nsKqWuWZxnTqA&AZh-`oyocnDSr#TNFs
zQ(-Z88|xl+P(r7y>C;Jux#5>e+W>w1eHDRX8KiPYME`|R4Q);f9FX+e4J-ssX~VQH
zb@j_U3OELw#>-q4uq0fiRnhe0zgA*Hw?UuiJ3#G|?n<s&BET0LV5US&{Z%_*pAiu6
zno%|X!9{R@)i)r2N96&3#-v<vT=1QyCJ!&S+v9XK4!M_Q%v$$J_St%Ls^}?vcM`eT
zHs9kq%rn^`kC<gmJWFiKB7dhzJ-XRPir9&vMblH@k%vc>`3J8eZh1UWu)=Q=w~14p
zz@B3P(f4G3t<&`Hov-3}dCchhN{+gpm+{TWZgWs=dWyq_0_jw7r`U){h%kgtdQcXO
zA9}60f6eC2n*ZKe8zjU|RQ{o*^-rp=0&msYz$Ss>wpDG3J%U%12}5LsP^sOC-u&4)
z?{otxj^ir*c6L>TxKs8!#>nT)$l!ApwyB?t^xDWgnch!ohy;P)h}j#f2uR2869#rA
z`%w@1E!>e#WMV?A-`P|gl(?$*m<$eJ7cnIxK-!It;h0hq_kk_OHHn(CP>}4OvzQ!=
zrR+G*QUtr|JG2POH3T3Bc9ahlZ9)nimn6PjW-xd##=B!>AxVz0&s136=Mr{VObr1r
z01FE%46x0Pg=jegD25HS0!Dux1eav$Dq?h!?C=9RDWUjio1p}s8P^fpClqoLeeX46
z+)-0GUtsktcEm)`2C1*qqY?OBX`^6(HAbt$umYujxD5-Mzq6OA=dzUvf1hmcipTeX
zo5+rVD>mc){zJr~ahFCgeF|<#CGw{26Q!BfWu0^1aq?w-`*(OVZAU~!l3zE)dLX*z
z>h*Qs+vw3c)O=W`?&}n406}LA?FruXVE(m<>Sc5x@~ad+9-_IkbxW>1^@Q{@%SMP)
zT2*qq;+FTxJ#aT%+Wq_V7jXPLuulrQLBc?P!sX;sq_4r-;YzDA-tv{g%wB+-6Nm-)
zSm!>va|NqxF0CBmSfi20r~jHdR3V6jnPebrN1iF^X0=TGc{tvKEs4$Lsf#b;(YtLt
zBrQR3Ez}L;eK5|xNfxosPDxFP<!o{Fi72l>-aKqtYNcfmMGp=A)}TWY1%w5@lP7g8
z2UXqBe|yk3iiefAtr^=9HWcS}fHA<q8eB|O6QL`eQXlrEJWr<&s+a{3<_cyL0QJRJ
zOGar4QR&7h^qqUMSv>QXI<`C#KvvYF2X~%shb3U_sQMJ5^ycVal@6LS;uCkrd2yHH
zzyGH2hn)T1W#J=Vs)f*qetJwDO9*mvB16SafAO*cDi@*L21uK2&Hb&Cwjl`VE^YCl
zz2<i?#KW1|+#eqf7Wo_wHk!T7VcG0=h!7NYRwC^pvY!Vd?E-H4mVRtu(ufdukRu1f
z4>=n)<iwnU^d5tVjdQ6S>2q}|3M{*$3g%-V9cpmYhf!M?9<jD!=1>57dI`gG`Y1}Z
z79s^Hb`D>1Rn6GeBzB2aQg)k(Zv^swXDdGDI=dV<ax@zwGhqTk)&0fiFb{~q@H>kx
z0Du5O0$OV^e;7i*HL(g=|D547Ax(wF#XrI3v^xMWt?3b1tZ2@8*gVvlt5oHOWpjMB
zS9E}J6^jkaQ`D7KP<>EFJFYeg6Qe}>Z6sn_`6;j9o3QsJdQ5#3A)EZvkI{(Ilh8aN
zCL?MUL>CizK9h!M0y}s|b$iy~uOp~?Cw##1NuKhQf|c68N)<AiXkpxLo0K?(Bt6BG
zlFohxPG^y#A%+!oCF5oOYn5<6{8H*pr>&xN%t=wkJ;;*TWE_a=)ZcrVLnwqftchIs
z8$QBoZtZ?Fu0g48=&7(bxc8<>6y&Y`k3<i}%S(?z$z9IV_id|tJ<#D$2VDZf5~{C;
zNBPn-%@hr42~0zqf(O|u9$#BfOlT*7aRXWwZ6RfghMRP|k$sI=*?(}|W~Z#QIgR%!
zeRmh~8l=7of9hVe4gOt6^?f{fzdvK$f_B3<6@rNX5z2l1Xnwe#r#%tKHF_))`x-4R
z`IiA5tdVY47-LDlJIUhdZn9!?F*A1N&TOD@<hew01DwmV`NpUSth6`yzxV4sxU<ab
z<K`{#STG;^FXtlYy}Sq#BgO?_{Fxf<<o8@8jprhThp!%Lx4rxZ5|bv`(FXodFnyFG
zrm|u%T1CP5?FUSahu5&3vv`H@oWtM)IN3DKp`z+s%oDwYy*}}ke8>a6iC6l8i2c!G
z5n(TCa~}!Tk)G~?IMjaiha(8V3w3gu1>jt)l(D5}OP+5-)Bw@|jd!%#*~gPP%XQq>
zco_rA$k|BMRV&c;_O#?|d}8S36rP$9h$?w6RgD;;@>swO+X*8J7s~0YvRnIuS7rvK
zeFC}5U!E?D6D>C8-OCk&35Ca&%)E;I>o#n#L16^1NXn&tIh7lLqma^t<ouxrp9ZF!
z3U@dUJm_!lEU}LMsu}z%9)8$%*=QGGOt!6`DdAgqkg59xYGZ-_AVj$niA(b{oF0JM
zs$o4;Sdj8u{Pim-+op}>lGJL}x<FMzC-S4h`Qn#wZ;#yF^NP)aOrNd$O8PAO<wsg<
zJzlz(O6)4l_tTH#`P(3z=iS?p<-6q>vz1O+M`$YKSbFns<#Ceq@^(JpXTX*zvzB3h
zKmqu?Rb#rBto!6Ygk8_G<%j%~)RB}-sNmGFv-%Z^wF>VD`hS0Fq@O>I7V3F^2(FGr
z?=6-!=K`ip^|BngTF3eXg14q1F4Y*9>wYr)@(R-r&5Vr4#Lr^C`!0IYN1N#xeM)FL
zWc<_~*X?6dU&JJk6y==nL_bx6WAmvay(wb*vHH={*1ILe(aMdP4P1m@=t=}Uwfq%Q
zc?LD<-}^XD9FYQKqhW@aYZ%t5gl^4D6_58t-9Y+LF)0{oG##@&&X57SpGJD8lidlp
zhvKjR(@mR8NjJsJpa+_V7377)>y`5^zJZa<Fkr2I5pmE*Jo1S7eTTOf=a=Hi*5-p)
z`A)F0s!P;1|17~Ym`LvX1&bw@BOe8|YMQv*RRk;t@yn_V*hB7cE}X9m<F65s-1ZMF
z>;#$)N-XfRQl3HG@XbJuj#FNyvT{UPhO(&X=-)k=uiCmB-mkXxz#h+@#EQf-X)}1X
z*!jCy=(NuJ-qvoYdgdRmlD9f;S!fBJ%jxHaJvzA?v=c81@4?iG7`eFJoQY34VjY(R
zu(Ac1cZFy}4Ix7!d`wwX$#WmHRA;<!*eDq(jU)G#&c|)8$kzz11ka<uUQ=+D@rsrx
z3Y1`;ZnlHqCYI;l3L{A<oiH<1t&71(W?EwMU(+nDEL21&;VEIq0V2omHym$2e?vn!
zLAUHiRr?;z%}CpcNNKL&r{q<@4`T>IjThC_Q{msN&COu{B>DxwrC|z>Ew!_!epV6$
zf;Z2iOJ2ww1E2ti0Zdj2S_7Bif)$OX!$p+ZFoP<Z!?nc3Q~iKJ$4(Rg8Yqmrpr9GR
zlH&&J7#ow0JTM-DAe_G_o+g%zBq;)kTd4Nbp`cGj4T`6a+`BlD$b_4EWdPdG290B<
z%pYiWFoL6Wt9xLn^QoxQQfj83<LYa>@FsFJN$M8Mmw2k&5<p2xQ@N;>7e~IlwqH$O
zd%^s#&1`gk)IDdyh^fmyAy-YTKex}4?APsLklS!O1i8pGu)q=$72gx4%s%o^UwMoT
z4CX@m5lJTOF^5a&TnvJXGMDTk1y4V-(meCrul^p0z0Kz8espuig|32$y<^MQY<vgu
zhD-4(tMo^FesNJ>>G*^QD<4`m#rT4%FEyxb|2q{hxNKRv+q6Z0BEuP*SBuv))2JZs
z{e(6X^dG`$g59hR8ttX50B9g#{?xMKb-kfQ<m5A}{bbrT^%Sbe7Ts>eX9woJ6f-7c
zwahDX<D${PxN&K-4LE8RUo8@8=F*=Ymid^%v4;9;IzU#o45r-56JpCCUz&Gj^v?(q
zitf#M5}wMq9~FGU>Z}wQ<f2dnscP1Hx1u%zh<JSA)lK8&b4}c8L$mVlix_vZK=T>6
z2>9vBoi(|EoJN5$o4CiAr&IRbZTZdIiWEzEtTiJg?_1BMpDcor_EOR}_k>PqY$qS0
z;bafVrVn3u28O_`;g{^+kB5+v2elnLt(_F`po7H<#^CW(WBtWwq{d_%K%hSw&(mT$
zl4GaY$0Y)n%aH)X4`KmwKxOnJ(1(NUKma<=G0ZoB;JG(DJvImeLIY5PFq2bq+lHci
zb|*|!7g{?3ogX_9LD9xAINSPC-jWX+YNi&sdWcHHlh`oTe~GESAi=7@6a$1h{gKCZ
zCm~|GW*Ry~LNc`ob}@m&;@)&cK4>7SPh3DIzVDMu08v1?Hg;nyS!})hSMOhqi+Wk0
zEDko4mjGnK8gsZ;<mkVJ&FKuRLmwQ?#X+dfVuA=o7Zab_VQH{_j()@B42N^T24GR5
zP@)F+W7NC@2QnW=<>YDTRS+=J%17PsZb;s9Uy2+tvy6sm_k;{KQ2>b&p$8J}DLhcG
zwHflvP*_zL*mRCdQxFuBe<ro8(=Z8#q>-E;`r3lKydRm`^K_tjw7Mr)!sKf}BDYoK
zaP3z&Y!|;atF*@@z5l13p~AC*-r2d^fVG160cmZ~Yv*s`)&LZDpY;m%Ed-HRG<&k#
zR%c~??&&>!`EUi+ntxAuq3+AdcchZ6$fTD4(_YhHw$*fjJyv%7$1q~A4VI()Y4SVD
zKfYf!OX&ikOZ)Ns8lpt*?i;~12XHI=_UPBA$Z^EP!R1v(AC~jW9nO>%fE+!&J0HkX
zH09V%uB!!yHE)`A085SG3eP(9PQ{HgM6$CdzawRJJy~H9b_*ebT8>WAXF^7Orv7^I
z2iVimW2T4$BgDEKDlqi=M3se`Mo2xLCz<ysmFxkRXwD0QtkBsSM1--9f|YEH=X3?_
z96R=RG^qlbtC&mv=jcv>TYNEC?SYM@P>LBkVWsgLBA|hmUDWVzQDYD`wKI!|@Gt+E
z2m}E(V{`6Lq3`}9*aw7j3WIbR&WrXrXU+CCiJJjF7rFW~y=vp-S5p(Q!pw7-!2WPb
zY}yz(;s9b?*q@|P_ivw-U^KFGs7^4-N>trDPkk~L*e*^sf1uV7yvB>ir3{}r`%ap4
zNfC(1^SRZ+hf;U;T8|ZzBCpZCGgcj?l!bFVNepFNdyz2LVT1t6>j+}KszihgX;a+;
zvPwE^=3-LWAzTuGjB-)j1eBbwa8v|gU8&q~On>v6T^wGTWpTeEpaMz;#c(D7xtRAP
z8;F+2LKHt<5P9v`<TsfWu8Pw=9~@dRo9@l)4@FK_w4rXs6_5QF`iypW4@)s1tali#
zq?rLe!u={g^In}=981wjeCP2oWN0u#y9gRKmDuH#iN(vv*p6cdC;j@^YnZtD(C?w0
zsVwY~Q4Hbz2TPXtI?<a?L^LC1KoX~4n%J36*G(kIUg<YCMCDWzVtpY@&-pb4`AU~l
zKs6yDV`X-q^{a2E3RwS)t8CgPXhF!HI$`^#v)BnL-Vc$HSqo3fEWK=UV>_)_6ttPZ
zAYatXU9?(vXOhBM4o*&UyTLK5X&%E6W*yz3Qwl~HLwAHWLpddlUv)OAn+j!b<2J{q
zcm4JqmB7n?wg$l225phesqy_8+YTk@81SlEncgPo%uR2}UFd3{$~`{Gf=IO$P6CRi
zu$wkt;>)YZ-`LvSj?+r5(zcc_tSn+&$d<$jS!BW}|J&4aOy%xgP{3wVPvIYk8`buE
z$m(lwG)3*BK@Ci21J`{8hP3k1TC5+M&dD3)p#RL7V^dcKd&t6%q@m7~SP;q&%I0@a
zN+|qxslls~nY{w$A@5P@wLZ$wm)UITxzd|+^4HV&GL;!pa!%2@{)8Wlq_NsszFPE5
zLzbh_5dS%jU$e|W-&DCY2>Sc|E4J$5I`St}=tS$T{L7_Ha6&-eYxzM2YbuQI8UDom
z_a~2~^tzPh$~_3MJfv`TS8mN?j>`ph@JeSsSaphDNo&bHc=oEjV&}n8*<3{EBL{`O
z#+c8)_eD`jz%#^i|5T9KX6@;~gWuG+WUqnu)%IZe?ZCf5)IHs;zU`61!Jo~{TIc(o
zYfzqdKi;S1UyFb<EIMPef*Je$S<d?=R|AcAYstI`IgS&-hBtA^qArYdIkd2L>?l}a
z^@*^yY5mRJvu^WnINTd(^=B$#CZAt_%g(5o2gKUZcCRSB=QcbLwPm(b*&Bpv-J-W+
zI=}(vIpUk+i~r0?Yw6u1@sRhqvD(?ZiSfX~t=m%N@K~g2uboscBRwKiSqM4A5J1I3
zcC!PS@F{R`4qqwht^JwX5yLx%%AbPZWUN11g{z!Q8FqOgE^wKvpeg#yR(-!mRC-q2
zBzes)Ga^(>swBUIU~T-SlCpq6CZs6u>nH2(MP*6Cz94f_I~*KbU-L(Nr_12eb5BxU
zgp#J>`0!lSqk_guL5IJ&F25&EhlXym&{12BTz+6h73%%gORDDE2^k?9zqfPcvimmL
zdu&6%X)_Ye$4_tqshEtU=$gcqX72|p&%S<3bp7fb_koPBbG;FaTxF}j9IuZLznbzY
z!=X6YdV+|es6s3Su!!F`G$lA%t|ZP_&!EdARHUb6a~0tyBtwn>6Y16(W`}33h&yJ-
zY$tHlw?nx}KuJO2z${1Uutc$X2;(ACb)zY873|l>D^NE-ULy0d*H8s67q$Psb-9iD
z3v3wN@^e`*x2o~^fW(hK_z(5di_^u6jh3H~$Hq4L#Y^Dx>(=I|tWxV%wEuhcgbW+|
z=Jlaf?KJSPPfH|QE%o||A*1>*Zdzj0U98^TB#LKKpJakzI#$7Ov}_!GP!Ex*c1YSa
zwGZcvZ0mc+akOq6iNJ$($G8Aqc!k?)8B85Xoa&vvWw9~ScIklov8(5Y{x+7x&*6Tc
zEs3f7vywO>%_IVIl{@FP1Jx9cS>vNB{*X~uM>Y_Tff0JK50czjzxh(y-iz61XG${9
ziTSX-PM6Il?@Q3<j%E2_puXS*>FQH&l)NeKe$)t3dkoT+gZNpacZ1v?hD4Z*iHzw4
z4?vTq7DOmUQ<Xj@$?TSfey_2&mflqzqhBiYE>i@F;^h!&Fva^3Q1{hhp?9EMwqrX#
z!%I+eU$aT6@_ayeI|}KR_pEU#%ZE@V8m@KAh@h#(Ax>VNB2T&bBD9gkVN+`IM~)TZ
zY{ULQeohA_0;c!G-io)hh>y(T=*7^x?%pU`wf>UF;^czYUJzXtG%}zc!+Y|MzO&I3
z?Lgmj;tOc%$yL}u-!?m(fxFiesf;>LbBg52t3teTHg9x=8tD6-F4_8ZUE9_euW~bQ
z{)+^gb2<7*wELd^x}cbGa@Zpbr%1#af9C(VD4lSyxx}OSwkb?OP8`dMMy69b3XI`3
zW%uRzoL61%=ps;xOyQE+!F)g#Z=P5Bp!*|1b8I8UhV4r{!@Wq;IagIz`|;&%M;uE;
z8Hw`p^Ov84^vJK&yxT#yNN{$An!ADo1$`SSftOcjs~)s$Ji*hRK!SsoaL(JErb>a>
zS+-5)5+99peC$+~=&PsP_iU@|Wv5x{F3tVV$%c;=B&#kMa6GmE3{uk3tHXm9#GB0p
zgARgOWp@7GaE~Gil^2;u-fj7Q;2`fZ&wS6RR)rST-_*|y(UA^BS8$`b>;oz>?&Mu}
zxV3iJPT85?Jx|AyJM{)5)x9jUX)AbB{*0EAwjaq2LC3c?TRiw#l+KBO6MCD%r;`m%
zO3&uMFY7A@!DkImbIR6wt}gNDR}3juBZE=sHujeRhb3`aebCvSehUKn<8MBpp2}@N
zIfC>coZZ3Ml(kX+<8&1}!n(1Wt*1K{(MoZ0eIbA5!+{MyF<<Cg;slmenct1E?6g(O
zt&0uwl-Q^Xn!AJP&MVw_;SPm0%DSQFh5hYq@%YixVqVk7GpVRdb}S1P_0GNT=9OoU
z+2hm?O^&Bz?EdCiv1i;;<l>$!FDdJ-vxacdiGIAF45?JAZAtcDa=50b>#7@BMze5Z
zg=`~_%v(&_<oN97=^vy{rXPSkr(_gDTbb;A&?fy3%5;ZHn}5|lsTijA5^pAh5uffs
z_}c>?r{nF%4#qumrye&d4T2w@?ETVp?%J21pREhzAD72D`6w{c;+h|<rk-v3&=~VH
zQrEB2eR>|QHr}~C9y8AqeH;2FWMkfacNH?e#ty?G(mWJxE61MaFuL1Fuj6mF2GzbT
zZD;=j>8<{{9Y`V2v#N%JTF9n9GSJ~1O9vTz-n=yV8xVJfA@dX_T8<$ukccaylt>p0
zTcbi+nNEYb)Dp$f1|k#Lg?Rww(QWpid2&9m0FB5dcOB!kaW0Jmlpn_GDN2{HZ>NET
zi0uB3j-JL?@mL|c5JBC!NXt~0(wae+8rSj9WucEXhhrufL=uDXYwpxi4+Pu+DP&pf
zDYPkr3v8csOG97~=s7QGGM<p*|9ypXF?W_;Y}X^j8j}_pi;KbO70&)u7KbrIu>|~a
z$*0AX3SK<#b5wOpjsEsHx6vwZQFTo}$Up0RNj-Y8X(p4b?DvFCjwjmqu)nvl*5Vwy
zAmCw+uqEKx<cP>=peHCkUGj4+li5>D&3TFM(~aL9QzTYqiEePayb}4goTzb+!*}-p
zneB62a!OEUie#Kc5ca?9rJOP~ehZt|+*6*c5!NS)Y!P=j1%L`&b(i`dxQ?m`!=gLo
z)EuZ677fK?6I@5?h1>dFMz+swgmTJ>fSq*5-pf+_2+^UWV4*QqqsEKp7#~LoxV>0l
ziawdovaZ-=E8Y<65#fyWK)=x@E`p9Y!e2!xEh9oQ?y#(f$Tt|WA=x5s+dp=^V##B_
z!jx^=Ng6H0quw8op?o%r*xU)bcs6Fgoc${UV1*xLO~GUFv}b0hQ&b}FOfJB;;P7BJ
zh1$eE;%2av@E6&yHaTVjvE!fC-jnfBF3J>Q22&s=5NH3iME5~3AN;K@fdrRmJp2*!
zy#0I`Y^Q|3y~ShQ=<%>{7I$g(Y_>VP8%07NMv$-nGyr|4DwOsQ^4p!&;*B<7sJ3|g
z1DBj0UQ&zLN$76(3C8skB|sYj&yMcz#ixTx@PxCK1|#k|>uA>yo+=0b9@g8LYT$2k
z|H8P9saZbGm8@X~#`H)Kxe(s!7LJC>Yx{^s6V$lYykC^dn(ZH)QLVhxH^E$|TYw}y
zy*4{wT*PqusL?MWJFjs(()Q9=Ns;W@4=WJge<F9%^Xb57@{p=z8e5;O)BYhab^oy5
z5aKZgHP?H(KhWpzVr0-B<$vAJiayF)ZmDVPwj{Ee@-XCJ?FiF}2)zqjLW0G1xe<#T
zP2z|p5vt*n*rkQi$_mzbcla7|=nUeap;`PmQEqp!#Vo0Ie!B2Ge?{?`DnvC!>h&<2
zrc2>cQR*E(SbTk%-c5e*;?_*D1X+6=B8}%WtM`+=i2*#>3&%9JLCgD+GkKkeq4$Jk
zw#C#G_d)Af9$@w?zm0K%<(A{_7|t_x4w})Se-=L;-k-mx5`Afvs>g0-?+(!pp4pmT
zHSF$(D7l}NMQd|O4z~-^DIit-jlbDVg!AQx;%hZ$`NNr+>68B|9)YV{OIVqdyqNs2
z$($2+8$qwfha(P}Q8X{376uPnk8#Y(q$(VLr0GRPD_cBV2t+Hq@!c`{^C(zQv!M}@
z-_WYp6kw!EF#SZaMv7M1R=vHXHSA#dMWS)8NSB6-_!gU-7NW-6My7vFzgfu~;ys@>
zJFoEKlK(`qk_m0yw+xn~P4PJ-Bd)g-*D>o&9utJ0YzY0vw{_RJ+I1`;ZjtNitYMMx
zcn#emdVzwwu5q0k){wzCq#vm$+RsK6`2!C6jfRouqN>@Js-S&Ps<eW0jINN^hL7og
z8a;{4yV6KpHbQ95ZI84{kW$zR*F%DCf;*_h$dfO79|YE5c3~n^dmhP;S6!m>Ej}03
zQ;^ncubj*Oe-^-<vA>1g8>_>r;?YXT?=NegPlJpjK(?&opY_{e$#ET|(>$<!bF?o#
zFVBJ=l*&eP#610c-b9k;-cB4g?>BB%TNGQ_+Mipkh3~c=@2;`>rxH%`FnoM%mc@gO
zGJv84Z|kl*6WN!>Tj-k;ow=B%2{T1|UryGtX0;5y>_3$#inSTGEx8(Et3<gx^htP{
zwy{?6aiF0YO$t&P(j_>^W<73ZKaY<drWswmUr(W-9yq^(?Sl~mFn`;6-`s5Qk^R~4
zHFAV*jDNy99^heDXG1lf^ViRQgN>e4gB>nVMd3IYc&fnu&trUtlR{8fo&=AHUuyN-
z^o6`w)zSkp3Tv{9Yikj^KA6{C%71>N)*AN|r_9O^6Zf|@Y~%5|H`jDms&4&xkuDzd
z;Y?|{0vlm<;hc%CEwnywiJ`so!Q~#}hR~x@BKk!@OiP@sg_alWd0sAnGMXl{x^>4R
zgDI2KNH0wEgsVMlG&8Kp&oAlyM1yNBMk~E{JtsgmdZP%F){BmeF(BV_^NpZeQ1UeP
zJox0*qqBp)^66w-;o%sjB=t5;T*N17r4__A$WwG@q~YdrMbo{aHQdk-&1<^b+m7Gs
zSLglxuPN%%Ovrsrx!)?({wq25cuI>}VcHwEpW}GrV?HDj(oijDfoRE{Me-vnj~x5t
z52?Pu{OU}PFC6L<W^&(h^a11Wz+j*cjJfN1nWn|lb;Us4)Ox?M0a4$4e@Er3&~kz?
z+B8Ln0ZRm<TQ19BtOHI*3DOnr{HdRh2UZq!f?MI!@6+YFaZ>MVd=m4s;JF`plf)H+
z(ZZ5Rb9HqiqWAxP%{Nc21+FdY0CO|1#yHo|ZwHF+>$+d+IGtWz;*E5+M?}6lwuuSR
zliXoW_?@FnuKhupxx0?s?jOF~Ys*n?YQj1}y?`3|apzw!^NqW`{QA3ai#8NkH`D6J
z$$Gcuou<8xb%fbjEQeWtNaT5MgJbrMtm*rEKw&6_FZ$2k98y>7^F@k?f+vgrRLh=f
zyM?~t-Cb#*-Fne4OJ0^}|LEtMm9`cE?YMrh;4;z^l5u_x=~ARXldpe+j0CBfL<Z2b
zKQ0K#0^V^f&37r}TU?~ziP_Zxna2W?sLR&uE`QHIsIt4LNhv6X3r+u|RnJn#PGocV
zPVfgCdC;-QV)ASXDrdRkB0mQH>5dzm9GharS|#<<#Vyde>1$jo^)KQCefq(HH(BYM
zvkM>$vcNAKc<FA$63Ed<2;QIahtTt!Z-2TU9d$gP9`56eMO#z6=kJ!eSy(ldH=9w!
z^q#pg7o%c$=wAJAS2@2WZhj*GU+T}^YGLlJW0D78OCu7|un!7u$Cz&yIC;5bHQlut
zI;`BSHhVzzkH%63r<RGp%T_dIF=EI>@Rzc%0ocgcl<=@!7nQyq>YC~uXMalfaq;kg
zjJ03N1h>kkwTX!SQk%$_nf1P!oCynv%!)_)MGj+|U=vOTeS9_ft7m7?$HPQCn;Q+>
z4#6aWvIiw*b|kGSbysF{t?U|2eMd{ocJ0a6{6kjMy8E;fZ%?@S(hAozf!QD+0G+h=
z>=xMyB(oO#-Wl8D)peF$77S<6^!R>|2Y=T9iKAybN+9Z$`9|_zVkR-B`lx>%o~3vN
z#ZphAh5K!Ca`FOA0EC{x|9iUE5<|}%ytZKjzpNe83h{2@+PEIHLX?~CXK}#M%VG+U
z9hL2uF$~F3G5n`kN}vhzhKGTQ=c8{%&jRtl0!L74QR{u_oT9~R@!1o9gN*)N^>wa2
z=YXgpG6GTD$M!Gp1JXNHVv7(I{$;e4V**JI(+COSRMD!%T%Jz>5hfKg!i;L6U_EJU
zISrs1{|zM|MjiRqlPyLQbuQi=h4wb3evUdzjuPGoK^)aOTxD7+5Lb4-GW1!*(n%!(
zq!5Edei4=0B6P1FxL$>5XZUV)6UJmCTTN=p<7J~tI^};Zb(!du(Y`LAu2aWW4!Ck|
zqg`i-Oe#<&d7RFvSI0b|V>3BLqwRuu+X$fO{B%i~I`8GvsSeKn*>EOJg8)ytS<f=P
zc9#eEa8(V+4sP=cXWKJ5r5@Oba<<3u%l#QmOjp-W*Pk60xI(%r8fbMW7EkcvckXz$
zLFCCdSjKYn;wk#`q`#-Hz>{<{NQ0+ehFFJG#imL8PI37qXP2__>GnjB5a?u?>0AHv
zT@0dy{yS~60oUUZveVbr&4~CXlkI#RKWb9@Yd}K4+WZVSLKYr@8KYOOFXM4MSN&X~
z#KOA%u^*Gk;)jZP@%5MXHuP8tV?4_iuR8I>D1?u8M-M3<8=FiPn?bQVrCohqhq8qg
zN=i-!X!w~t+#_zeJM;##i^@xsz|nT+k+R1mU}!6B>oH$RF&#Y#M#4N()P+z=wvv0P
zD-UtR^wqeLQ<>88(v*vkxuaK>yBYLilY7qg`kL>?e&1==LCHWd-Ac2_=3cL=BPH{$
zRPnHM%8};i56xv5y;ZzLaAYuN6afX^3Y#WL@wR%VM7K7%IDYA<3(Lk3ZM(b>PL2}V
z!eRc>NZbZwpaefZ!g9T>G>Ck_a+_b7{E~(uNPszidY>;Lz>7V4Z?S;(Z}_y!p(_$X
zzT2pH8DF;~Cq;hFE-Zfthp^Jt`AW147Ecyal#p&_23}koGgSyUq-NrlVm9lrR21z=
z_hNuS^jF4A;syi0Z#oBsG9%wjl-F6?Qht(rs@{DIG^YOXuruNVs%P6D+dur{|E!K~
zwj6}P29+se2L?NuYp`xrQ%5e*el=3U4f<_AZ4<?bNt0~-u><}L7geGB5KS}Rzf-zV
z^A&3rVo+j6s?T~E=Z{EZv3_z&FY!9;_E6`!Dx2owiW0w4V#*3Q`{b#3@?(lY>?j8n
z_Y+r1Uy~YHesmbKzih}AV7i0{$I(n51ycMVQTKJ_DF50s&GfezAJjZW9En5y-;b2X
z7R%kEomDW)$Zm;pfoJFKWy>)aFQZQBSo!=Xc&p_ZOSYn`>p1dj*9hJ-b)BN0H+mh-
z1v_40rkO`@4Xy&h9UV7!iJq<-bs`pc!klg4cP`-OQ+xB&KrD548$89fQiTDJuKceR
z-v#yx`$Gg)N~emObXLmBzDTZIm2M*QsaN~u_Txp$j<EOnnXeBu2|%OHCI4YQdSqN-
z2MNRs@dpLm)PVIx)YfpS>UfR@)DlPyclqeaGCvk1!afv?XEA9a>?Y86srs_;&NQe)
zv%-<KbrIL2?QuXaU3KU{mAqb|#+>2v<{GBV^0__sxtB(E99GndJzIjK-0Amps{fZh
zEK&YZ$aDR7IHe11!#|~e*F!2ZrJ;*TM}gIAdTsWlLfpGIYBi)~Gnw1XDXp3TzMVBU
z4rB<`^3_J2cfV3Uxel^);p8WDBDQ<Nv(ItyT8)KRi0%>40wlyF5nF>GL^D&jpesXb
zN<bqeWleZ9sF-v*mahFm#U*x;LrZNKVn-RIlg@?rBl!2HJD7090HUcO$)JKr2)|W4
z*FT+fczYT#2@Km{(oq$jZ1%1iZI+Q(%p-o*^+w@<WK^Z~Dw9{O-{}&m+fpmqJXuK>
zp2R^!8ZL&P!)*LV|K<yaZx`A5it-9Fhfla}EO%w^BDRTDy6{y#wDCOYD$hr)DGAlo
zLCZvJaqIS>JhkX!P1bhB`#*hkQ=)7HUW=qr;0{zqqQwNWuF!H<nMLOJ*hTi5Ljxog
z+cfnkF50HMyxQZ^m5I*k4P$%HLqd4y*&4=~JCih;m`WEYA|t_F!=2&9#wWCON-FIJ
z6bbVO!;gmtHx3PI<W)Zr+EXDZ<G|PLh~AovTMs(st5pqpQK0R^H6JUnPt}hMbsYnu
zkA_fh5jEh$Kjn$(E8YRLx~&8u8H)KCBW5v1eVBDLfHZM>9F_n3nfhVg#?^CFYm8j~
zs+Ut?Io+y;FC+LxnN9j`25ASjTu{Cn42xE;we<cxxvaZGT@sooaHIk2Yhih)IIyC`
z$<(GQb6#4vE(UZ$qi|juC1(k<77B-xHdO;dt;G#LO&InFp95x)i~M8Kjj<$RF9777
zVP(smz9R;CE#<$?(T@}#F0i&=3p^|=9a&230U1cxO!D^d_llS}I}{1AxhcF3M2nh(
zfn|&C4c<56H7GI5ml&HV6<8VlA9+8|ezm3h`*((5v~=|f+qt)ZJ$Kl##kqXnD3$rd
zFFBKX0V=X7t;5`J3q&HbJW3OvXJ8g3B3-{{HoB5mBWL7`v9^cY=sc!KzT{5=y<RlZ
z6sufSMWUK9ELMuyB?h9+lhH_}R8Hc2v94GuDst~E{?0d#iGf5HMP|>;ewWO1JjSe{
zVec1U8t+=eQd9HAi;1@>f(~vm`~GUo{j1_B_tHyVGe~VT@VW9LFI18n%bSaot2CWD
z?7unW;`&m`2a#07AtN2yoWKmgN*#@jHT=<z5G)3eq{Naic%<EMF}OflqCKZ-@GaBI
z{2c{9`VcC`0ZYKl=7X0_1~sXG`63&G5Gvsq_zGilH6a#op0-5$K}DR1D<dJK;5Q;>
z4GyJhK%pY-Ak39M7jFY>hXb|g%p>3&<2)$9@4IXou!LPAi$p=j$jk|${yv3#j*c+D
zieK+ai{c(e+N@CIZKIVQu>UzARqY}OQ7Zi~q5o+MuqIEwRI<@ljw9Q<`WLnq3NL+#
z9oSr<{gUALtddQU(?kGuZS{0279G#n#(H7!Wh*GIdolW^Y*3<ihw;Od#*+y}lbW8&
ziqAS|WXFuSWR2U|TXpLmgdMts)Y84Bt5b*&8l751KM3cxIm#-!xr~0~7V7ovv?VWi
zVo=|T^nkNym8Q|wakKXBpem=kUWVDIsz&|d?q+|kzS-#B6taC<QrEV`cRWNY0{t9b
zM(5G7NWESigUPHe@-_CYC8QUPl3)z=bB+D5+|K(_xkUxEuPOMIlSGzKtx`mE?Ok~J
z-AZRy{5ybJF<nfZ{$=JWd%k#&3HM|D<b2(?z&Q`>10;+H^js*DTFTVs8Q&_*)QQab
zhpD^oA#5((+)m*D>0&2yIO)Kdfm=U>A<<&Uf1dwo(W=w2=5ss=7kKJVD9Pb=gGPoJ
zhR%S&^uPXtt|o-Pc0R5rLe+7J;cAN?2w@2S26AKghO0c~2Jn1$p~0d+Did;CKophD
zLx{(=vZ^w6vg;E+`d!!nsGMl`#p6P;0*D8Ml_&E2z!bqNKw(Kwg-*Qb>&69%eTA_m
zW(0^J=d)$fL6BR|NfZT|z7bjcMRoibRePs(gmXLyqcXG!;VKOmXaGf%a;X%{b^BBR
z81bnK+{x36Xzh?wWsE~IM97mY0O5`kpv)PTuJOO$SY}{G+ynj^+zaQN38nE6qn+-k
zyk8lPv6b)G#$p;_2+nx(+9`-v1Q2Qmhwab>^qO0}<*b#s2~LUVP1z7B*=bz(?bk*H
zCf|IAOxp;NCWky5OT~$vw-Da(vS%jJB$Z~}nPAsh8p?yWn-*`7BH;BG&~xEpb?9aT
zd#oa8>V*A{W=Bj09f(v5YMD>-ax116JzO=?8$BhW4j+ZBjDj&`N<Jv=c?mWsxs`nS
z(v`JIq<oyT!8Z<-($b`!4qf0g!WH5JGoU7>R&S*Tg}0wTqQDxqcn-KcqUX!-6YxPF
zIQGThYCi{5Jb*b48fSIwxb`Fb=EqL#<-EOkko|6dY`Ddy6uF~t@nuMCEo|O>E6Kl`
zvVEsqt}lN$P0}AKzL0Z4<eKQO4#Xh)MNhl%l{e^{L%-n4(E-#W$TQQ+x`XgvJBpIN
zuF`VbuER~{fosMvdE&hv9JO_3SgI&I?-wDtlelWk2&t7U)#<Cn{K2?iDP<*Ll3*NI
zYNV*KKN&weT*XsY->hHqH*;s0*q+4N-bxiX{_Hqs@V1<VF1TG{va6MOk}nyH#PMvo
zIzqS8rh-*ZX6h1>lf-AM+#38|r@l#Iz+TV*|B~xyef%fg(!r|gtY0zvJSJnYs8==C
z>8joq9s^?p_XVE=<t2C~i2#ImE)Vv%gW?3zTr+<X$I2hQ1R?j=UyBD$pP<7F7l(;g
z#Bc+C(S^hxFy^yHYUU3y?IyqwlexC91~ZfiJqK>D^QE_oHxd>!RjBWpQ#&zFh9Gmx
zI|A6`!PrvMz-F$huON7cM$Y8&lvRM54v5JOrMa4>TzH~B#+GcCOjj@N8YLe@4wL#F
zv9R7Jb(^lw5Al&Or$1EjDE%fnga)f?t*2gTE##%#^oM!ME=kXn8^^zbT{+(7N1@Y1
zmP1QaLSW~Q?>PSt8bRg0rnHn#EozKZp*3ah&7v;fFFmfoPna5$<{NGMTEBGM(%HU)
zdU~(tdRp6W`VMAT5B`Il<PWxX#~o)g4;6Qrz0lu#cv{Ngg3{=9{ORS{XxZZQJp4^S
zWB~^V=8r9kNqmwMsWGB&z|95><9!kaxSu+NofP6W4>XK@-p@M6v02-D5GE|3L8CNh
z^hFRCk6i4B9WN$t^v$aI%;{d=)GE_^ekMTDQyX9>+F|13pc$qfoZxMC_ruVNJCDu$
zmD`x2Hc1(q8dE6D6YM#h2x>q`%LT6*4{_Jur0=x}P;;=~aca`n@1vlh(=&hT@0a;3
zz3=2eQr`W}T6y1v9Yf4l26-o7#Jl9q6#BQN^GUcn1&D{soy=SWM(UX71WxZNUnTvl
z>7dql8*v^Ax}FQeBj%y)b(+#zp_v?jv0mRVr|}ka{tMbZr*ggGz2+4!zIi?ODu2E0
zW4+qD-k$q;AGltffyM@X;z>;m01pH`Xda82BRPP;)Ux1#9$u9Lb(+1(!EyTX8~_Sv
znJ^GwU?^7LM-P5TR;rAL#s<>SXHFej{3{E~08U{49zH^7HjAdN7@!;?Dib4O3G@`0
zLVqL#MChQs0y9_?`iyU?h`D16G7kY1`VuLRm!#A5Mh!l!0N}9vCNw}sMqq*nH0Ea@
zT!zn*JPsx-;1Jpv4pU^77GOrRB7LF~SYbdFVUi1)D2=KZh12<7HPKpBqy5Wwng{CK
zv|PRBjc99hoLZECGpR<=4ku9Lc&&T;#r<kfze|2f<a<d|T6F*;%hHtAg0wauff3wC
z0RTs_*H&v)KXZ+rb`K&=(C>eob?o(<>n4AQv)(~zADCSb<734xe{(_UFujsXgvqPt
z2|I?}_MyG@XV5v7@AMKhM)nWV`MS>3^!MO7(2q$$e|hxZgKoXs%Q@(}4WfKE8`G*j
z{TMH*Ec4pyrq!eu$9=DpdK`615>3^Ix;ZenF(ozHU?(;;r4@fKs=kT8q`{R?g~`AY
z9Bif`7zd&TMxtH^nDlJG`yn{L-8SZY8aZj}0R@aGGZ4XgY<N@a^N_6{hoxmnG^NEu
z1`ua@3NX!gVNeGT<h%pnv6qIM-!Y+&Ej^EFpVM<aZTfp-bAapBIRX|dE%Q^WIFX)#
zsVn*dkwA#l6agTDYZjUyjALl>Fb}jU81i&l4u|M8lf{sg4%(r~(bmFHd9na9R-q9>
zU?h)I?{us#9i>BCg_<LUQTRAPU6@{))}o0m<e_P-257NfJj_6UAp}y85U)Rg6<k?A
z{%-=*p|21+QKz{rng(N_)(Arb@eK4aLV$tV`UQX{1Xu_RF_cRgF%E6CBmyHmjJ1jb
zR8FR}WCIfNr`K<WK5z5)AWYZ=6--9OdcK#`6rpLP@yfU~$yuqioUhxooTDl?r8C?d
zaFag_cgy0hJ?EGTKu4Q=CjQc+D}n32L~4xqX`({8)ai9Ud4i;gi*ArQ^gAslg$7Qb
zRAA@S42s`oe0m?815xubYI<yu+8pw??c@pH4>~rB!udMOa0{&yNi8iEmCoF>>h)D=
zc(t8$je}o$nf?5217Fy_Do(5wvC*OS2>BP*R6vln1f4U@K4e*L3zKD>QlsgC26lDk
zi7y*1@`3GkGD{SsZBeF;V*J_O$tdcm>Sp6-6!cRX{pWz6v7*1bY7E$Fs%X`*2v+u*
z^}Mz`PH`%q4TN{9gHya-h93dCDqGJN*KfTn^)2Jo)O>R*pTI~BrdkM$1j*+{)9IPh
zx6~<pobkF=VXyt_^&Rvxwt}=(r{~5~H9pk4Wh;ZvN-)D?>w7<QwJJl=rkSV;fDwKl
zvoP&rqpQ-Ic^R+sQva>;w9<60+4N^|{$_obvHnqO1hob<+oAZTPVK5XsrIPnZvCS2
z_`PyWFu3vCFRfur7lfB)f7#lr(&+D+Egj?hq;bmY6wmc4fsxuG{d*8&Kq&+hq&X^p
z5C9&U07<A$k)sK%HZVeM4hDiH1SI$(t`b-PNWnF3zLXa=n8Ko_h(s}XWdOKA4U;^y
zI_fatH8NhtKCETdl4VdkqzOMFf@^;OC`~w3XI%)K5a`H5Q-x-~q^1e$fd6@DpVXNL
z`I$y)vjBixfd&ad6$TxMA<fZzLo8<qhyXxA8tJ4R<XG=gwvlL!k<oE?Gg;<*AQ{l-
zKOEN_AqQs`#E!|9@6%qY;5GIEr(v&{vF`PQl*@8d;E0{B*2E<I&;^<AocQn6230Pv
zd@rdHDff&38f)_~6OZp50m3lAh;5&(mlx)@n!Cc}4U#7a--6Ixk7-RGBVpPY<Qo@;
z{(8vMudz>FxvjlzyyUuX=ztNmIZ}*OS6c&niR1L%UW)WWM&=FLF+CR;f&gFUW_4MV
z;6ToH(jzWDhp1}ev(Ta6b~yBNQbUcJm!!9X1$D7)-!=*c>8EbhqDiOAGW!LL=;f;4
zYSXIj)9Y~#uy5wE85@Ioe^j7=ns<+v()}ywxHkKc3*s6Uv}FLZ<{ZS6M^uaErcv2q
z)$24jJ1xd1IyXG54?|Qu=x%Ya7`0#u%BNtZ^p;xH&Fq`rsHVA-&H;VBqsD+!o%!CC
zAW*4elR1LZmwH~OJhpgMj(RWjH0qfi$KOuzdb&ZKQq=39|2Fy4nw5eHjTJCb1~8J8
z`Al!4-``F`52yTe5<Wh$Nj}!+dp*4pd^65%e3<>9)9}h`rggg4>!{bo&3u%e&WSDs
zGp(<A_pZdxT&?>LVHxz@LO~Hj9fzUZ%(Hpb&xx;2@v_hu|2UOv#yjy%r8A$05sv9&
zlT9Ws8HV{CmJR5`A3ytMH8$1Nty+H4xaH?Fy+1h~npl%g0~Ap{I3(afO$-ErAk_XK
zkkQ0J0zRXUku01+w^TN8!8$#Ibvo|}pzv9q8XNOVYtLwwOZp}O*s-RG!aNAsQb{zh
zW+uQ=09Zmt>^XoF4S*&L5~Km(L9LWV8!(Y&OP~c~QGzAK7BGVJ65wb99n>r#2*N9a
z)HJ~}0wI;+s#vVIeUqV`ESrwpsj-qnTPMfIy4LWL$bkQFpOA&txiS$w;OaOQ-<l<&
zn=7Ii`;L{>9<UreD4~3i`OdZ;B+VdgATXjdMlyH@n+x;!-i^t_JOV(?2uw!?2~Kqi
zLc>g}w?|($nCW`G*URaZR#f@bJgTly(d~r}_YFDBk#2{5IS!hbtNqL$ahMo8ytpK`
zR%poBIe76j&aviPtIj8!CHPfgT^I`Zw%O#DW?NKxKm3=31N@=Plb^cA{Jc1@|4@LO
z-bv3q`)u*h-FM1_x7~Z!T87NbYIXJUXFN+roO?;@y7qJke<(G@7Be%j^%iWRjk?Kn
zmR*J&(E>I(`k+w)Zm)@*>6**SiFKM5HNQ-Hwo|KSnvHtzG{L)_?vI;S`leNlaj!D>
z9hs5#amQizHx2P|!0(4w??Y7^w=qRiS|zCHmka=qPQNBg__&UZTR*Ofc6#;C0bAcN
zx8J_?Us9TK*f_v%Wf==dn&oI>lc`Q-C#4-~Z5DNX;-c%Po&-i#$aerE8!71XP&)V6
zdtT>}Jw<nH3m9pgL^cm0bf;dQF<yN0dL9K=X6QEv!E^@Acrz|&d4AW$Ho8uE2RR4Y
z>N)5*u2-kw_hY5)oIFckA~;~kAb}A87LDa433xQ1@wW=lB$vlxmEeg4CkQwY@BoNX
z1Ym+dlUOy_kS8FLhY(+yziK;Br-`f@KobUn7+ILqqQiBiEs{gJ1~g9;%{m;|^LGZc
zN5<jMeFpc)$3P90mgY%o3>GQ?WB{nJ-ULCIk7YAtp;c0@6vP7jlrX?IJWNuxzD3Ib
zhUmnd=C>;FSCXm0qzZ)5Nt{=xfk@};>Ts28jKiM;GxOqv2(&dE$a_W)!@K5m_|vm-
z9tnDFJKsxchRiO`!Iaimhx7G1%);q1GLNcksK>qXJt(u{8b3kn!Gk07GfQH!k`qOT
z!{I?`rs?e*==T`2*TNo4A0x}~F*0--BlhM*8yj`Ud2Q&3suMlaQFW(<X|yfvq~<1+
z>`WQ&-7(l&-Gl|DEttGyUop_>axdXn6=mY|4TD*RNjK{8L3)F*|7o_%>HhEnQ0mcc
zlzNckML$W(dXtAEP&2oo&Q9$b=6JT3YWDn?pSXD<_NF+^N4>lYHo=42n3@4FQU)->
zKIW!n3_vg!xU|KaU4(Nz4;A^?nY4}yu=S0td2ao0n@8DKb94eaRSsq6Im(4J>z>0q
zl%!3Pnrr#K@nljGzY9Dm3ooluXcEROfjN$B<HSs(-@A=DrZb-hrLUKjl-~qKma#4x
z!zO<t1;6u9(s^Qz$tdd^hDSU(UBbyjI=)S!05dB7aIT`EFk5iqn>s}1^xQ@(3328c
zEzU6)XM6iu?_SH|dUXb-seo%ppCI)A(Og_^feWy=PQM-vnt+9{O`58L_DG{%#ZQ>J
zf`jsA4O-C9kfy`-mT!`VD8Sl)^c5oG0BV57D$H7qOpP9Q*3P}oXi3x-KntKmOPd2{
z!?9+dQ};aTK@(dvi&dk;c4(>;0BlHy?EpN=07({S0pI`xDFWqSJVS}5xd@J+K^ZuJ
zr_=VOg%w<jVgY-N%0wC9$btk$78<A*G(;%8C_p1+1%6WKi=>=iO)M}%Kv)3KiNg7I
zobNzNyEhK-y`=N?g#`eN1RNI^;e5SiN-HX0BudR>H;qy+PV_p5dC`yJ`7qRx?R1kr
zunqm*j|~goq8KA?G`#pgK!a;p%F%lbGc(7;p#$-2jC}vl<6_-ujP%O<e&_cZSgeoR
zq|opBrsXAv1Je-5Oj&@cL5`PU+q5R9H<MEnTOYcODcXzW3H3_0TjR$@D>*G~)x^&n
zoRm5b13#0(tBqC<U1KDQO{Nk=M#V={Ko}L>LDESzM!XNK46=U)S$3GYnm$HG+aTL|
z21eZa#ffjqe1p?-H=60QXgFY+ev5``^i;TBB`{Kh`Q%ohWAg?yC|dqRq)$9JfTp=D
z{$B;a-Gn9xHAi?5lIdCj$0K<V0su-&Qv(P7<UO3H6Cl9>4r+a@wjH#~W0U|#a&+82
z4=oUYFhznI(B2@(QW~$|cgvZ(1YMv>Ql6;5ln@-n12B=kNMI&Q`WFG9A<b|hJwcKc
zw5J5k6q?E;K*K=bMQWz-H)xesVLq%_hIv~Y&}wU-$b|S)%&m&avC%{VuGAQ5j#N@$
zM0M6Wz4E=}p&edV0M;o}T8$N$fkk{8QV!TS;3j{trBCD59|ySZOEpH;U%z&>hyuC!
z!*gQijxAlMDfPtZq@Z&n{jE6QB)A&ncnP~{Jp!A+^9K%!Q%^fZHP2H+Xf0v-BT^37
z`p|7mQDY>h`54hFW22=%H-8TkNMY8!ANN7|8KzEy6kzkT*K(}^wdtHMvkGC_w_fw`
z-J$*<&1^)m$&C74jGA5&`Y{DYlF$putV5W&ni{t>OX){r#LxTWrkP=KZAHL@k975G
zKI!kh6W`Q&*XenXS`JcMP4BlTPd5i?|9VQRyeH_VQn_B8hLd#qTcqX(gVX|n1p}I-
zDol)t+RRJ~7=hM?^eF--?2Y9`=`(@pa{xy&bj%JPBQ&2?n3xbNv$HKN6DD8)NJ8*K
z2GFH*^_CA4lmiFXQd0yc<eShep%zFM2Z#z~2!sKC5C~x?L)(L43FfqzpEOGVLg-*U
z2mc_;^8iSoVbZ|81VQk8WU?Z`5jt!if%Zrhz)P+OjTC6QEY(|_pb4%OfGSx4CQX><
zDnVl;sXj&sj5J{GY?2y02@g&ttb<g0&)!2~(~c?YVKPN0%p5hywL9zWRcF4J)EHT!
zrU*=@&CX0qXY4tQZ2~2JOsOgUq}TX3khJfToBSa@F04<=Uhk+|t2r*{2k6PipOmsX
zeZJAxHEuM5JX42Zn1_dk4y`92BRcbfFs0YDadcF#bu#n&nSj>gq%aJ9WA;_j@6~%2
z*Ld+Wp@!YG(rn#4PK~6h)WJqs9Pg2>58cMp9PGswu&J>*&a=@9*20aQTD{YX3mbu-
zN&u$X9QJDRE-W49^E3DLLO+VUG-p<T@FgfP;xvau$v?KwshlX<J}4S3Q(DU~r8VAF
z1yg_$H9xxw>|=yJ4@J?hb$t7mVRn)KhbiRYlzC#)W}Gl@J3Q7Zuk%QYcKsOy7zr|-
z{PYq1A?|?rPhG(5TG@ym(%b@EU$0Sf;+y&|I6Yq@FhoCeQ^7-2?L<}84H&66;5;2*
zK@QpoO=xux=ukk0bbyYQ$ao28zy}D-!8RIg-yS0{@q@fHcST@C(U4jp1Td&Ag8vfC
z;5{@#_?x~+2#!!|Lu!1$1eVn7SVs8;XrfT#1SYkZU-=;+LkXCyzyufL>7#@}f;cq0
z1(RAd-6iS3>=#XEEl$%f3F`*H1>gsPks{22Eui{YfHed|awvzON)`Ew*=7oiNUje4
zj~}_u;A;tCq~TL3$T!DUfBK`O>8v^qLRK-xS7B0XbaZS(`f0QA$4{J}Z@Hoq*2M7)
z{4OO4q;!)%sJ80YUdO3Us~^8jlWjrP!&Q6_N|yC!T=lknZ&g`t>tU59?3L@jxHU$;
zec&<aa6S@KTEjL7RmYecR~4_WY9X%f-LAID>L~nURB-aOj)jmD621&!hE1TTmlbqU
z(DM%aZyR52aTHuNz-Kp~CMuu0D&F)ueocCcZroRMGF9aba=e7uv^tptQ|;;J)z+8v
zn=yrri0NFspY`fzS++iO8dE%xx`@T?7#6#_a6{GCMoT?&yY~7{y3roSXKI5ImY;f?
zorL^q?5nGuNBwqtrqT@hIxL&csQE-wS{2A?8WCxX^zshc#z`;qqUKdp^aza9!FOut
z=)UNpqCY4)UVWv{Lsgil>E+wF_lZA>`RN(4T7sXJ^3<T39ID*R3Qm~M9~R@AHi@05
z?v^xYZfFbw*u}Y-!2Y!ejFfSbmcrLCrnIW7&=$+IPq2q-(cm-bW6M?>^Ev05A_UT2
z`zRZ&18}B&!bHniwe9&viE~;L^rQVj%l4C=>(v<m6IGbAV#vYiIJH3tZZKphl*K7^
zpdb?TkdXvH0MG)C05DO6<Mst$6wuwm30a!q%8!kU<}!VLWJC_y7)>~KFTgYwwK52H
zFyvuEs{!p0NgIF#O=>mhguVTo=B@~U)Sv~zK;S}|#o~Py8Y0pUNgD_u=)pYH9HH~}
zRRBE%I;bhakS_uF03fnNKn0<ag^4XWf53gi?2^cjpdIk7BA<;w3%6VjHLpklG5(>C
zk&*F<4daV4Mph=8G~df2#^Bt-yr4O|Rb0z39k~%rY1!7uy>wofeUEJ}r95W-z!5RE
zbDN{wvX^>>$*bfK2anE+o!d8sA(8llS!z72xAkzS&a!0Ux-V{EL=_teFrtX1*LXOf
zCWSB?rFd$TQcVH})I@65rJv)4+X4NAp}A=_>z3Xpp98i&<Vl|jHkY<JXwW`JY>T_(
zb2(TIRU2)66O2-tIt&Mb9Fhn_-)|nLIiLBH?qkFW$oJd!B&H)UGQYe8!TypM*)k!v
zV$&a2;K}+W2=bf`*~C?@jjlcqd4VXl{3>qW6OW78MVzQU>0%W%oYy_S5#tbeo_OkM
zYxUQ8v=-^{gXRJ%5Nr};vwdqlvSWlmsN1n3^UBfR=YC9kIq^-IRZ!3U(CTmN($eDK
zK$!W3d!P<r1n*Q6anu;0IjdD19w1mj;9_NV4%NmkhT0zhKr%45)x?Qun!%#akz5Yx
z84uuuL0*{_S|p@LEfJ-CLUUOJT_hj@|0dZ|NosQ_@PiK6B~Su7t2OumfqxSE6Ja3m
zAwd&>9J#i}2;-^wvD!pk0fJCNrIjB(Q&1M$z)*&1t)*$4z8}emYPKk8<>|wLd6A?E
zz-KcYwev)8n0ZAO662qiDXj`lsly?68oD?T<{juFl7I;ABgdd2vH<@+P4+cRYE^Kk
zPS<jFTjQEdQ2Qb?KC1h;=|@#hFFHZLmDFUM!aS)C`vmC|uevxrcN*iYdlMQ)IcVGU
zGI#Z(4Nh(I8^4~{!+v?-l#c7Z6c};h&IXL{r=WX1v3I?E)co|xVc61DLYYC17qw|+
zoi`Hy{q#}4(_mXCi&Tpnzvb(FsPedtDFP!SW&5d1lUA>%i&cF#>eb|@Up@>I-w!P;
ziEX3t37GxzY#7r~P|wqMS=e!g2q;laP+CuMSE`@i9i=fMMoMDD<D~u|HfCO0(P8@{
z>^!8G*96G($bI*U%o#7#qlfqRV?Pt8KkWhtA6tS>o`Rm32zI)Wd@Z%Vr7Li(T*hhj
zjTXGdmeccLaSXoE`ZKEb1`+N_U<7_Q;L4*NU{Z@dKVYFD+jh9l2lOA(#Q$8WEP)0B
zA_@R$FQ27pEozQ1O|!ZrpG!>+n$MDEuK=82E~A+&`U|N-Q-vBM#R@<dn9-sZNXuUc
zK#?}cF*4fn3qpY~H9}SqpM!P>L!%0<70{*c5c(IP`K(d}njcd0gpUA@P%9KPmgknG
z*)94o8QlcaU+{rK=j(X@Fw5|<QiNHrT3efBtdes)+$0Mg;2-IH9a@)e79R#!y|n0v
zIKcN(IbWBiv|u`O)S@vG#X+PX^F>6>ODq|#y;m^DC>c>=E9y`&H~E7@7HhN<_+m}$
z5*fw#G|8{`^|0CAp!Zz&4Fxcw3uA*evLET!%fuMfq|gtZr)3PD1A`nd=B71jf$T88
z(>%pVaDp56K`#9L3dzhq<cM|~Q}ggK(kzx_+suWBqQ!+-17-k9g{&BbQ%k3yPKEM$
z&=cln*-mAc@nQ3sboz0x_}P!Lyvhh`zt)Rx9E(n`GS|y<1&l=Dd_8J2Gh8ngAOu=Z
zr$Wi^#VYK-<7DtSP8x5NpwB~L0iT&8hsA-T@F}w6{B^;$)jju!6Shu@sjb^&dwK;%
z)al3|{7t&Xh@Xzu$CfR<?YY}pAqf9$<wYImcO4dWxf_Wd*Q>Jt5@_0L@$jq!8l>}e
zfEz$3q|<fja2+5CHA6B0IjB`ZpB)58^q@3H@+J5w!8$feEfHvvH0e_Wfo8W<$O0fI
zkik%ZCP)otwrDO(0u^+w4p1YLql0#q39SzRDAErJ(;>am4xtZ{<_a`OXf_M&VEQcb
z$v|Zl#Ek%isV(zaQ&h%BL=_-RzP2O;PUI65IUWGAAZ-D@Q_65eujZBG&j<gDz{u$M
z6!mt}pm9LW=X>`b5u3M9N^^n3v|rJhB7gF|RKSQdsWm$d4avcp(n{L*!Q@$b?=F8>
zFZ@~W>)~GeiR<Zu>&f|guMF8p`E|jdYPXwEQl&|+DF+k>2032LO)JhXizS>?h!PM;
z7c#1+<GGrB$Wq+K6oC;gZmTdYV4iR_X+;f`#lwLX)g%fqOn%1M{y1n;iz%oEQyleJ
zk2GmxxQG+7QQ2#zX|4%P=+i6DlK4LL+9stjQp9Q3B%3hmwVh`EDCVKV`}T?>RX9W5
zvfCs-l=rJg9~WDSRk8KNU9$dRc*Hy{P&1q<t)XO`QE4x}c|AACrMDZ916;2z08Ch>
zCvpU8WfaRT0E3RmTh7g?MY1BDr85n+IvVgBk|jU^;DLD>F!dwNWdS^?R~O;`1HcNk
zL|_6-n$N;-830buL>GO5(B}xj51PnYf%XVP>pqWXFrS4qtDsLh@;?)_O=9t<3JNe0
z08k3hCZV<o1I=u$EG&z$&7-1LgMTOZ7+HX3Ne=ZbHDLm*t#MXe!g^pNFLKD=1pO?&
z=Sc-d_)Zt#!=;#MnJZH~Sg+S%1~+=?6TX)OL~1xB!|Ut}05Jf0=9)F#C-lPQ^_SO8
z{?PB%#&vLSy|0IR?I(UChC*Yc*M_F~{c#{klY+k*r>WxNfFDr6X}tK!+iSho>!x*>
z1p0CGq0^XRSu~}U$09dxG2Ih(JoIC%_R6P0<PXbF)^*>}S#jd#i9u9*Xv-Y7aae~|
zT2#*0!#Hzv>u73?QAm>n9~o}-4Dw6?BTl9SgU}kLwE8?0R>-s7H-b8k{Pu0j`8u4?
zBZN8GO+Pd_%Xips`AJCU#W%0#NoYe_hrx1y>(v>W(t?&U9jX^6$3$(8AV*%5#^5*|
zAP6-<3ee1$$0^io0h%5FJDTtpwMyS1@Fzko5^7^mvjjI%Q=|@!j2umIwN2bfA0g1{
zsLjmDJPPQ@L34y=vZw_@Z4buN+!g(i(7Aej30f%xSlYfsq)9DmlAvA$W@uWAU<v~P
z4+fg$D$=ADt^`ub&=RS^345LQ?J!nCdjX6r0VpBO$`pK$;5X@}rnD57db`s9LOz$-
zP#U7WaqM_*zLy04>1$*hfQS<?ay&=+E9HRu!OZo-pY^^T?zNw|?ps&DNIz{p_saZ!
z=l2r05}nM#HfpbqhuE&L8`ysHB+l}965$y5xf!M})@yJRV#nR)bkiC)hvTPj>qDn8
zwT$_Rr+?@`&jyT8MSct_@@a^M1Gc`2hH^?}J#ZjOU?gf#D6W%MgX}i~BP%?K>)>Ol
zpLL_z4W3`_hej3!_~QZnXPCTL?;F7|b9D<Cp}($h8Y9Z|3VH2ky}$3+oH%jQxY;!W
zz4zjq*K@DR)AQP513npCOnQ3JNS`O00}>bkU_cOorm^zSMrdN4Dz!gQbvj6IF4KQU
z7S7C}K>{;~1WlSWi`aIuP7nhKiZq#(#j^|m5t`C!VC>{LcKDpM@Q;2<NK2a8LR=Oa
z9@IvW+8%W94h;`#n9$S~eU32H5my|qNDY$${DrL07YJz3#1^$%TAD2YMKECVaJ<fI
zjT$HbYgBpyE2O8iNk&#z#8RV0zdQj@WENl$%!3gaA&JRxtkY=<EMZ}ooUixGS6afl
zsH-|Q>n2@^j0nD$(uCH+Jb;mL!SCImO=%fbOm7p&0Vnx`+M_)6<UX-^*H%%<CU0)@
z%g*(_9`?%zr*vHR4FniD?c$3)9Y{J=I*kv;0kvsSFSedr2a@=_#FHm*?3Tm@jb86z
z_<eA0K~#@Ti^-ka*33JR-#MP>!OAnriPhG(QSwNQA(N)lN5o7Q$Gy2}9Y&uZaqwlB
zXqtV<x;u?2jzJhVUh^?xw$nz-OcVEeVwPvSAC__Iho83grE%+@1ALF@vYsbJqwq1}
z^nD4-$)o`z`Z_CYn+S}UW+1}S8m4@@|BSHn_j=z5hMB9$R|<+4xZbz%Hi~hp#@ckP
zDyHBcEJ?wMZ)*Nf&y{ID@@N!-Yo#fpn*cQVHv=I#trpE6JSHY~Y?jT=LlB+C`oE5o
z{r%QvTMrsOy^ftI$Fr&n*Q?X;vr${7*({8232@L6ddr*^0g6?8Cm9HgFk}ICP+KI6
z_o0dRf<Ol%WnO|J%27MDNJvLIatG~I`WnF)BJhzz-Ub}B=SC{+c=*)YGA<U6El8TV
z3gS^N(>A&9L^}EtA#fq<Bg;g&ETipbMEW9OT`|V9aH>v!B{i4>V`%v$ftDD|X)zR5
zmPIuKP^RtwM8*-A0iaUMHpSHBcp~Kp1V%EI(V+k<g90NB$Q7V}B>N|-Kjyr)@5qcO
z0bCn}woyqlN9LsUtFP@$T5fgJ)7tppHJ|alR6a&Bcn7O<bE1rXuLE!lqIDaWe?zyn
zua{?f9}Y{K<`?J12<BEN8K^#<`_ZN_prBX1*86(cY=6-2KG%H%0Y=Wf<XL9jhx$Ir
zb_$(7ST=?`2_<c5)3}ragB&kv(;B81+#G0zp_k?f#{pX(x{WEC(rUmAf$apQjaIl?
zt)E;@-z7iagp{A_XB>oSi4qRl@^V}?AP|X4Q14cjt$aPsHkRigz)X^Z@+iPi0wZ+3
z&eM<4iavC1ChxaC4>`?eUYo~Z+z5_B$hbf=IF=BFj}ZbR6cq6kx|0yVQuMlt(#Mvq
zVP4N|dCa(e3sTJTo$hV(rc+uK-^)C$Xsov@*Q@gY8>)*d5`-YgL9nA`z6#nQ(6o?1
z1W+;p9$A=zmEeR1D9~pJwMv@!Zc&RQhcvt=FhWp6(o)(b0M-bq5a?jYLt~`AxGc(>
z#>5KziPYhDgos?e0yAFBPoRSi)ma8iY)P#SB6fI|hqee!ebLVeKPq_$kZF30rnRzd
zvs!f0&Nfk#grEnX(OLUQwl1n!fIlF`z&0`PIo!C5sR?IbB+S^8#3T4aU<7LCeFr8Z
zsQS>82-5Gg5HQj<BNwLOd@l=lk5-_)B~5Cj8Y5xYxMA|}@B{r0alEY8!{O@v8^-mp
z+dwqe!)|H)Jmb1A1xEaE$|H@_hRq{e;vTyRB@<WKv_b>tz#zv<6r0w8GjM}fz}AOO
zV~PtGfz6_TX*%12*+y%Fvln~T@D9jRes1*9<J^7<`$`b5M5Qsp)76taZ5vS-#*=DZ
zO(!r?1D`78)+v9l^y}rhJ`Xv~XI?^-VcH0udD3?ArYkNHfA{(CId-AB5!89)n8|p+
zJS}an6c{mBvvwPXuWOmp9~*)LT(2(TTe!T0<5n=8EX`g40BK^r&OsxKnita46~C7N
zLvrPj7BGU7>p&c2`g{d|1OSay<dNDS5Zn{opwAM5AT*!Fxi1TFM4I1PZMLr9Ff~K!
z3ky>Fgk>`i&2Z&ma+csm10V)$a!JSU(pfv|BOS7Xeh%PC7C=TFj^GKt&_@aL)6^HW
zM+kDz?+5{r@<bV@?Hi&_5CwUm8zPohp*;oQhh>lsK@^5!4jLmaz=%vVpg(-435?hd
zR}H92X|XYIfbXSpzRvI541keQypKUOMq;R1N;1J5aFRbHxXOX6=65~(|9t2w@%10w
zE$)5f$-rgwGlRJv?zL=wBT`_bS1zuD{7R6p5hevc-4#|)qxk5jjJVQdu)y8Eb*nh_
z<Xv#wmKXaE9THDG`II;cXMdB0qBy*`B#Kano@lFj`>96(=G96#sc(iFXy#~I0Hzq~
zSbQ%-oxEQm-L%HT>3*IHwDqCe97K%~o&r*)EOZXoXz9<B{5uKrhRL5{<|o_I&v9(>
zZoS`ABZH>2qH?}&GnLQ_4|y_*erWq`uBMkdd*vlCvJ6u!W0hq2X>hCeQ8<t7)yG;C
z^N`<-pj%twes<2OJH$1we6Be8gqGQ*YyR`69g8$Kf?;?xH#;pgMp#IyF=A(R7{1ym
zrLpUk16;4t`Fd0O1)+8ZOjThoM9mC_95gx_1RfXwD8Olm7=bhM<vHw?c+Y@6n+z&1
z%~{jY`s}>a_CP|Ag`p$>BH8vS`dn$Gl?Q*b#e&rQD2)NGfUlDVv_1-BWti@wDJ|3o
zbfvj2q$SwV(rST&b)+GnQENL@mrmOyy}T&ER8kh;2%k0R!=xRUpJ6215KHw|y-80x
zerJ2omXT6MOaQRy_q<YH&4`sI>Vj%yGXRpP1%oTu3)n~2n;Ijd@I~0>xQt8xtUnI$
zy;P>O>UcjI@HLY97+HV)9CeG_<PTAoH7I(1*Tet*f4^T=@!==-i<|FwK-~4<<AbWQ
zYq{on*fp__hg|n<SYU(#<W;pW@xiQKk|b4~T5l7DbE%IJog@AE9n}7vSFN6P@kQdL
zFS;Cy+SUS==^uXdG4YLWeOolJKn@cwy5M~AnwP!AHR~V#__+A;_1|<&qg44^6Zw2-
z+q5cM&^}2dCRyBeOoNQEehTSredx3>p`behLEC!N7)j#y!PYl}cwvV1ogM_Q`c)=R
z=w~3P5jZK@fQgPu4<-*-=cvbj)byg@D^L6~gmqZ1fRRxEBf}!p@v;N%hp+41$H>!m
z?-bX(>bYX))~OD5#0`u@X-X?D({(}DB0;|uPTx5C0EGBhsOLfJ(b19V&AQ}(8WUWv
zRxu%0`8fnYLQrIt+8_w@A;O>>uoKuQ!S@C=FjnbX1P9fq<srco?#<f!GlDAA`e;e`
zsfhvY6C6Hkf_9VNUTB)o`8wnC(#Hs$uS1)J&d=Lr^Pbu%_4x(RX^P_1gjgZa(l*J}
zq$BuZ3tS<HB293it}Hi;dJ)i}21pqo&f?6{8i-REZRN?=mP8X}*8uuR4HP!Cz31$w
zkCEw>f|y%@nXgqih6gZW3)~4$6jtQ_m4&0eff3cmI?W`mE%f?@{T-KiX&Vp2)-gyv
zzL(0BR+Z-L78j((NL1!ugVZHWJ3Jh4l0OVj3%qL@6c|wyAARZ&0FwK~?e{;Xq7rdU
zfsqj1kjIJk$?<V<+9@ZCty?yW%1Bw1v0x}5z7Eya<yuWl&&-I!M~;do_wL2-Su8%n
z36MinFHu?M7U7@@i=}>_^hgvFP?@Km+AmHzX-7x3^bdRy+7qXow9}W6V_|Vo?A?E$
zgO=M3wKGmVX%`m1_C=nacNPxC4jekX7R5Y&`{55@0noBPyY>J-KL6G0#1l{L(W89-
zKJV<a#LHg%BAf&&3f9jS`~V}id|~6%rd8$lvNNvI1-a^d9weVFZO~1-jm9v?+1hXR
zA=~G)FkxR5q2^!5yJtHIZlg8K{EDLPy^P;s=4UVMj)L#&odyM=a0x4ck(N20-dcPZ
zc73-(6D+EbL~WrbpV%W#I@RBr!N!MS1V(J_2pXr)Lt)=*zZ=1zb?YVl)6Y0by!9o|
z6sPXkVkHwdFhahT$}ri3(Cd1i4tY=9GHGo`k|Xu8Wvi#xb6cJ?u6GXD){4DT4Mu_A
zxaDeHY#Z&~OX-xjUY!BZK}`*Y4A!U8NjeV>5WoPypn!u0H0-m`^x)2p{zm9%T?VOf
z0j&-RY|sQ3v^}8NFP*T14nYtCE(`=$_@Ceig9J<fkYwP{J<GD&zz9JU1)dP-kb!jq
zFe1Sl`Yu5rII`yB1g0A4KwfIQpiTvt_@enOngin*Qt8YcGLsDUfR$nsTDh(I)7%z;
z71GJ(G9nAmCvNBKGXO-6*5F^o0Su9iv;4`fHpEn^Dz>2hX;{A;;Cson>FmNB{IZRS
z`Pmt%A(;Xr>$lINXql7zAqv_-X~eaU@w*;Y8Y5=CpV)sy+<f~3;<kGqF&_=&J_SYs
zu)RDoA}+b`>Egok&l8)cpgQS8&`b@IeFqMRU)_C=_}%Xx87M$9h?5CPnG}+~(K=z<
zR`G#%y~Ed-A~5ptPk&ad&<UQ8aKU-!ir2sL<vvmhaqk1a6W_Y|mbJ7wEUf>xzx=Q$
zTEL61-FTCD_z#b+r493A0s!MHue?H>b^7kE?1KbGs1b2&-;pjbtE-X$Bcc}A1uQ;v
za$SI={|-$sf{Zq;6`SC}Z4RO-ts2I*%?zB4Rs_}Fuq2Yg&;8SLV#k!T`5>OecRGd7
zXZsG%be+cDu$>xF!?`vYfuO`|lj(QvN&5FbBDVEJhQP=QgkI&gFuNyWLzJ1e!6zp+
z^;BcO8^N9m@tOI>mz*PBe%S?L*S1a#5i@%nz{nil+i?g`8Rn$(^_IVyUV)Jy8~8!&
zYfRp%d#`Vt*K<Y3?P|_}G-bOLg71EXai%2U002M$Nkl<ZsLT|a?TEAfUOt23EITM|
zu2*MhZi)&Y*yAaaRn(p+z#JA)5Xv5E3(EjVpalXjL;@iAO@M(RUj_g%w;(E8CdBeW
zRpg6#QJY(A?Jog3tWr}NA=B0pp{9rgM?hSH9ul>H&ycZEQJ<Y_0VxVR!Jd%iOOswO
zvBf~sTJ&8)e<iF7pV1T+>))&{gU%SLNt0VJO$99&YO+YJ5@?6eDSTxsKv0~(Ulw#M
z-|}A~3F8mVYZd@LOpcHC47e(q`n3vES^JjYR68IbvPMH%3CV~`9^go6$?C6gEp5mj
zM_^KV3*=85IShK9=X*(Dgok}s0rp9a5db7!Yu1h2XGv&#RNpJTEpB^6Jsp3$$sf$L
z!+P&`J?sRGsCMr;G!1~{e(~eGA68KVxu(E~54Lj=aPiYG5HEc0bHo@Nkm;d{#mM2K
zN5wqUOQ{vIW%DMnb@S#9vqLKVZy|KP>HD|BR&~n+f{Ksh8f0S;rhLC2DWdBYoM`EF
zqZFo&LGpxE`}E5PFYSYg@iFoIXFnT!%0Y8vm)JDvep>D;H{2*5dF+qAn)sI2y+)jK
z<{6G@S8!tIVbH2!@%7Z+ed4#jd$1#|E&u$&qWI+JuN4GodJXhn@}X<q*+piMzz8d2
z6TEj-GGIg!POb&4a$=J|_2fQr(#a=w)Xye(aMD`i#+)0(95x)os4)`9`0QY2L^o>g
zv-OQz5&E30geY!p_B$6(^YkH1QYe~c9o&TK%6!or1PL}U&7_sy`{Mxnt{3O)W?$+8
zTYoihPvv)Pxl>8jQvxFk5L!)+I)jACs&O>62|5qiO7*)DwB?N|?xoLunt0iB&KD<a
zp72F74q!wtI7(AmI^87y*1u7#r`Oo>dY<IA4s+e&n7`v%%TtgxJ--Yei!i@NlRZR8
zRG?`oTAv^|g5OQ-*{D%cggL6(>>_n8MR6R?(`k+h>6o9PDSd-LyMy2b);=`5#Xw-B
zzTDo!K`z06IrxesP?81ULPzWk?D0s8K1gU5OPcFKT57H|7g0X7Qt-Tq^)!JI=Ai@i
zIslJVI#0*5!o(=(=0uGS;pvbad(J!@wbQf~fe`{a<vi+(gL3tH6T)Vi_yX7h9+39r
zXdXT(OQGZlRe&m`Ov^uuY_#@gxsel30Tk&;A_HE?K(Gk@M|PE$L@C>JTwiL%`~C@-
zOcdY=aZtfrUk$1ralFTV+roV$ovj}|I*oU1R01Of0Mc2EVJE8terd;%PW73(x|P8Z
zOmmNbHQkb>&q8n@X)@3QvkT$`%*Rm%zw2QyU_{mFz|mRhgXGp<{Z2*o^_l`B9_*&s
zte3s`g;G;QPf9@KXTQ8d-2c$SVtxTOH?_di@2`5vm0}#HEzEGkcfKd{nDOSln*hj6
zo8)^Zp`=;oxIcJaf|5fJ^i#!I3Gw2tkJ*0xo@T2)@}74KYA88{`yPBqeDk|EJ3U^D
zA1Rf@-+b_WR!8S5(ak^liMaDO_o&;hE6v1R{+wrv(@#AW3sX2>KX_2wcE_)9J=B%f
zZ-nAA)6)dqfBNWOb!8niFfs!*@JXBuNkcL@pak4{?I?ISHoYu1@7jV*N*0b1GorqP
z6KmLb*}&Nu0Joz6yYsLBt_LnAv)G`H!KtF3zzya8BXcnSaa>3h*BnB@{nVyWQSQ+X
zgP(RqnMMhPxKXXbEW@CH1%2*`vVBp~ifimdS;r`8%~Ep#dorFdjlz_c*Y`eX-TiDX
zgVKyLJ=crJ_8%4F+c)8x)i!hDQNUbvN#r4PvT2MYy}ceybA7DlAtnmD&i1<zbgf8;
zhdiY9ip$QI8X}tmZf@fSMwG?~=hsT1bM5LipH){OD8=_W*ZvMOJr9r#>Koy=Yf@mu
zFQ?LUItSc96$Oms=<J*T0e}Jm3~Oenp#3pYDT-z3e}v|;0FWSE4)INZ4%E!3!xUG(
z3@7aXMEIQGO3R^nD}4#imS*AS1M@$_Dm6yrk5$pY+BXL+FoGu1Bo@q4@;B5(B+c6N
zto&WTVZJO)UO~I1i33IiNa*_{2Qyi$GXuZpIXGP>$U*1pH2^-OLsMG$617-b_e?8+
z5oney9Bau_DpM7)IK3!%B>)t)^YvNy5jjxpoTpfiz(E5>Xdf(#_H8Lmi>Wp!6_{R{
zA(0%=HAcn&kj&0ZWA8X7>R7urrC+?3ww6=n^)=Noye|d3r?vxmdVaiF0K_`?ruuXK
zmU~r<IX7ByACxxhl-?ibK&mlvoELt95tZX8K%DPEgM?<b`sKfyWv9T1>H*7@n~<wu
zg6M*C&*^x&h=szn*If@^49<e#9ox5w_r2q7U790z-Sb=Vy&wL_^E83i2?2eQ?WkP)
zCJ7TE>o_dV*D2U^GM{4OZ`^Tze)U_#$-8#SVmRI^hQ0J9EEGTXsn3cURjFOV`RAN1
z-tg*II?~~TfBpAs#lgeJ{ffAdi{%luWd6^O{&g4aK?5V$lB{jiyDFEk2tNy`gjG3~
z>gYl*Re)NU(4pmM+w2FQagL<(^(ahfasFUP(i|cyl%PMhzDZE^(y9!Q0~CZRQ(Bpk
zk|;Wy@bp{7sDY6rtf$=C&iQ8nHs;+n$7X(NjBG?R5K+uS{RT#I5N^Kexlacma=w@t
z4QDnfZeWBa%6Oud!FCA29DnQa-fL`mJztMJN!t{R16;4p0qCeL!OsWwMtJ}V4Qf^(
zPCDj*rbre-Rcc<;=Vsv$eN?7v;s6lA3F)vLKm<)&jcjTi)M)}3p>GfR0+9d-z7+&Z
z@;JOnjSy;x&_9XN6lvgltbC40XY8bhGz>JgMJ)~)00gj1X^Kmk(xN#n9Cl-1I{GNd
z;qVdb%RqBrD=-5F^hMrNd!;xDQ(Mqnp$12W{zur>cA(ZvzPS>qkCEj@Ry?`n?(0RB
z?|IdMk#=HgjGR*G-I{`D*=d5laX<khycQvknVX#!6@2fKx`s;nnze5oH{gPl{9!$>
ziTtjIg90O}>t+|0BslV&U*0Dgx(2065b-qyMqC*B>`N|^4x;rWM6GDXh(cSh@W$7?
zN<8iCvpSM}`<5Sw-`sb9N8Eb88RVoe3@3w~`u^>2yh@yL+G+Cf-M_uBOJj)-zW%Lm
ziw7V1gG}S}=S{DDwK(srGsX1mteBpj5vQGUa#zxieeyG622R%p73jz0|N38&2aF7g
zA7RSfdw4preh(J0si|b$Pt$~Hvn`KJ@ZdHFQDcOQT$}cVjh3y<G|nFfQhxg5(?(3i
zLlhOjicxJcqyF~Oaqc)wjI5`yQGJZ)^N^D!gO?_W-;H3e<7K2+5U+e501=qXs=#zj
zFClJVgnZ58TS~2u^(+K%>jQmk+0uDEx8+IWdgp-GLBifkh>il+tMdRts>^hCPSAl`
z9{@<85wZdkSrQx}$UtBLKnH;kYG5?0(E0!nK~q`;LTIK+X^9Y&phigp`76*iu0qSB
z4se1dtX5Z$k96Qegn`;2jT-!y;64w{uT>l%qQ(b}?6jP_XAlPsmWc_mf_-JK0?-2o
zj3gjKpoHeW@Ly`O)Cjl`D9MS^XbGmYaJruD!E*`BKy!s43BeC|MG>`JOG7^kAedS&
zS<HV;&=?uXMyfIL#|8LK3F}ita!LC`U<8L~*FyV#BEKR|7-^X-TYHxN=!pY^+~>-D
zB&=sN06^9;F6k4e1W+=nHEU0dNi!#w18(w%#8x_B#r&>^!vZ6!I*a&z(Fe&*KmV;*
z!S^k$kOCtPoTXFfzkKgC(sYoX<d1)P651l)(4!oGKkK4rh?oD_i{*0yD!+g1kK+29
zz9Wvpgn=IwacZp4bP-iEd8$5(#Y7#0p^imN1OK^L(SyGD3nLgH{Q?#!1uTqM-xVyF
zSe99iDpRHBCWy;|8cv3=etP;|Ck66=rj}SQo?wx6!osRyv|ZInr6=Eb%8m}oYiNgF
zCzaOERrUcFc5oV(!$}y@ZQ!(x>W8p&y~@^skuQGzdhxb5TqXNo&-dW(9}!>2cybCQ
zsCRz!12Cfnb6R))=3cRRYEqoB`?Rj){QySDH}<3I6Z$pLc}HM`0Er%gYK$Dk;(RmO
z8aDKE)2hF#Ve7kR?;){k=eDrfedVL62#kLD&sT;4C$jaS+Z;sa>onbC)AX>>8gTQv
zicgxRTUQ0Oh+*KTR^9YxA2|>;FybZH_G%jIQ^0yi0j6}`o5LEQS3BRm*2}BjPS0%s
z0D+NZ2*AcFsgDsSzB$GF-3U76_lh5d*3_$>-vUI6So=m1;s!?M0gMz91dQn4#vtph
zUL18hedE;WU8m<!^lwUIeQ<#5)oGf^!hVzB1`vyu4-sg55GbKZsw(`6(7`zY1ZtDe
z7e}ECe<3yOznKU7cLE{Yw=&S&mCOrm4Vv4c_6UXfrLhud!K4-fH8o(rPR6fL69oqW
z>MH<GXyOjPm1B45&jg?aLwS5eEKS4o6~2`$lfFY}eoLC}Lfr^}(6m+^pbdXhTcn1%
zpb|_juSL+HL-vtOT`WOkqyVralcRI_8rofKT{DfyG)9irisA_VM;6!+@X{7eDZ^oR
zu2W89OP9v^=Kw~#+>`px5lu?Dj|4EX2-Eo!@XJP@S|jwq0~wRe9BDMArc{QV18(w%
zVQ-tNzu)z6ufT{ZuZDN-<~tq`H~#EiQ3Y@wbx45`2adiJenp;h$;DkyzjotI;^99$
z+7)k$IQ67m;sx-{z|;5-;bi}#k3S&+2U}VdM?w4v+qR0c&p2J2cJj$$%ce~bY!^EK
z8k$i$HZvn0d;Cx0p8M|?(=)L7X$dFq+$o-OX?Gh`(kJ)q6?feI8!2#q#`#Z^S_#`x
zc8-f21e$3&02531{q7-gFC51cjA3Zv@nsiXEG~T7d1Bj^&1(W}f)M);9Tay#2v1E0
z3ck%rp@0)ASH0q8l8)^UeYbq;=3C^WQ+DkVm&4)xt{o?cA{@*y5RjOjosoTT$FF}Q
zXl9Cme4`JH=RNCEvEzj8Yd$qtha<<P#RCsLEPe$)8eXQSBr*Ki1Q3A^%Xgo0irBt&
zi=eL#2>`X%Kcqt~j3=JjBksNb0dWBSHn`4<D(Jw-|M=HWi>qJ%TCrpM_73vo6+w+<
zp4PDhYAaD|iQ(VBdYyRwv!3Z27~#0#_;CwAyzS@WD0u4@Hcw5-F?jl^r@-e9)TIG_
zaQq%Ta#Y;^y9Wh*rEpyRhmU*^W*9nEn}Y%)ZuRi<Y&~sS2PJ^>!{<2D*!s}P&k^lB
z*A|!HL~=NtuRHNEh!Dt$MjY|BzKNq8FPhX~VhF3nlvMtDwL9o@n=m~njUee)07lRz
zd6lO}@ttsTm};Y`ObOEOP#J=lV>3%)(_~pJ!`68j!lY$<WBeROArNf(8)A7gIP_uq
z?ilttb{L(ntA4hP0l#tjJQVg__PY`ETPKxn0#1}){d^p#g2^l%h>9}A4UAAwl!b%l
z5t!tm8))16#yO>k{e^lTb^b$C3SE=VHtWyz-)?*{4~4nSV+281F6F^AL#IFy=4nuX
zjQvId^9KXFgME_2|CP32HxI!x(~Tfa4RINqA|7q~C1ZUVm}j|K7hB<|SB*WrElzWe
z%-qL6QM_AytnN@2<>p&NHHdrWC5*?VWth9-_^Uvdf<OlM<s4%eYZ4G?o7N(butKnb
zq5WJsJx6+)+$sW$r~-H(Fwv}HOkzAr4G|>Il}n-l|C%yuTa$z7s5R0AXwLmJHAi@G
zC|9KR2tgPe5*TUi)o5OeS|rkB7UHQL!rz<+npK+if_X0ZYNS>PUo!d|Avm&3-y%rI
zKrn@%OQisFU-{!cKUM&~6yQr`1t!Ic5dk9&Xo)-qZIMWWM0h|Ed^JNk+Yl#JS~3dr
z*&yzRdCg)oj~b*sZr=eveZlvVnj#IzFCdhY&ev%=as_Kc0wX#Q?}uKRW@B)`N&c{&
zTJLu~92GF4ct<~U^gnXLPktko_-%_Qq`-)Ukwtu@|MtTlSQEypB(!bz_y7C}AXL$H
zfTGPSH85WM{O3wQ#ymZ*9M2E{(8Dq&>2KfqLqY8UGdvCcCEobzmVXKJ8SnS*KOjE$
z<*$lYTzQ4K=z<Gu-}SX@1cR=J<9HtPc=PLCBhK7?+S)Vo2kS}@;?|%13=ZNJIvgJp
zVEEtv>LX^BwR<ka2n3vQ+Nt8zFTHXNbXiNF{Wx;$nE3Mb--NT}83{nW5FiS9=oFsZ
zyH}>Ky1YQ-@5N8QKwSC47p#FRIz6_{?1MTMc0c*$?c(;k?)C(}bXsn|O<?4L^Ue`h
zUjBlvEOfw5aKRSX_Z$PWbMxYVe&*lAThMQ3>^`k4sZC=<<?TwWMSS*4UlDuu?bl*D
ze_n9jx#E==TctLr)%mn__t+DE64!tGyW-tfzgcYGvZW(Q3XF8HBT+Y|@F6LcQ-TN9
zhYMAP8o*O0=j-!}^P-9IX%ph;v?QS+`r~H<ZZ;=O9+QUK{V?;>tz8@OnJ(CiqWyHf
zUct9Fs96G=FpuwjP=6a<Z3waf_o9`Q=fen$M8P|K9tt~-{cZ&P)^`&eslWF57mAlW
z`)Pi&1WwlvV5C9;D1@k~DJ^^V=wr{GB7E$6oTI{XV|a5?085RL2DChVWmHsev^U);
zlF~?bm$Z}!NJ}@;At23=(w)+s!k~aOLxXg8gES01)Xd!Z-@D%T>#Vcp!#U61&yHWn
z+H4n)1$;zd^0SXo!6&Asw_rn3`6mS_yJHrURi+NbKHft;&X7s!qi7N!t3hJ7D8P}H
zKjI4VCaq}4nQp*bp24BC&;5yX%ce#pRH>+*f)RortaI~<qL(-1dn(3Bx{K4JC~~Y#
zdG%lYeB%(T9*Qs)v2^-gvJt(v6k+WG4cy*282wTcehK<ttvt$`0a+H<Ya11sb#ba<
z0bp;l7!%_N{LqiG-@kPji4MarXiZ-u2%knwI1N%7!Y{nM@McSPUKE>po6;FqJY&qt
zz^*XHvN1@iT~=G6NpVa|<m3Bi1ZY%|YRjC!eO>G?-IH<J!-ty=+C(af8Dp<8U@&Vv
z{A$CBcB|K{{}U>c`6^`l1Q7JNkcuhE8>kfJKIX#s(@O5sf(DWH$nc!>s0XuV3usT>
z?P&KL7LGYs*?q};+?2)o-MLRlkYuc;m@8O9&);=O>9SWwqLFQ=SgD2@r7(E$c=sIW
zr=GQ)-?vw|>CLL~>0q#tr%NRddjd;px^sRQ<X*Ro`Q@?(k0+H=tk~-yhE?un$BX?;
z6#Z%5|Dks(n8)nu3vcE=LKcA~NuoGwj~PnjfUaQ;Td~bjMUt&hmi*X0E{C+kxz(<^
zj-8PHY8*wkp^A|g%j7H!;c4-p=|Yt|oi}-Gj)+-RZu}gRQOWe)uWau?Hw6KM3kF^K
zvn(pT0(nTOk{4~t>(~8-d``-rDi^9Ut_oT}dYRR%V!jX2t&e}!&=Qkl;)+afyzIx0
zLPI7+>9^q<Oi3d;ZfDjQ3v?$1mBV6_^ACCb@=a(Nd+5^}x?FXGpnrEjlEb`i<JpG}
zz+;3x;}+nh=eZ-KWJ~5L`{C?vDQ@OI>^*nU$P(yr`}r-Qn{qA?^iSVoLJvrOSQu1M
zRt<mb(0t(}n^1GIZFDte6^RqnkTEpz+Vr!6sIngaT?&SeHJiUqI$vP|>q##3Z}>*l
zE4^9v@wWg-qw|%=@^|3oNbR|gLXdn#kX=&KG~a1ya?tvp@q29GZ>4%vOzV@|S8bh)
z;yI(v*y!;>4V4*McS;IbOOD<(35oD1^{5>jQs`HTEJ)TIRl#&1LK!^>rCh@c3JsI$
zxg-v5&CYR6-q>H&v1N59c8kO3n|~gK$;()P<0LWu`E8kWy`3ePnOXzcWh)fd-(8ph
z%hFOucLpxHQ-&CRJ#Fu?*)xE|6?&nS+zT<TgObwL-TJ|>S7EpF9sOHEp4OASKYd6X
zCn@9Eg|h^~jN`VJ^dIKX{PuNNXDq#B#|5P}B`H6pbA-FNTI_JrAMIdk<;?Rxu8z(>
z)yfE-N)Sl1(N?{C7pg&<R=-IrZ4fB@toQGq-c@c;N3>6@pKe>AM)$Rl6XhEnVFF4w
zjM*Ju+YUX+AK#ls{GRkIsElmbdYk&?JGjykM?E_)`1+$I*xCap6W*iRjoFVDCn{ig
z=S5S8R!6*hC>phEt4b6y9!MMX>YKP;*?O@{qdl!1r{?d%2^{L_79D=~9CfGVVk?nZ
zJw!aL%_+JXBOdop`#Hp$Nh%Rjs?L%1n6JveVY2-%b3ONG7E};FC(*Z0Hz6)L;rjh{
z%g;PLx`T}9xK$yuRt!>^!z{(C%b&m4<llNI<<yw`F%7+cf<*pf+Z2!rFPZo?{KVhz
z{iG|I_@Up+(d4Sew`eH$LuGe<4bR81uX-H|PemaTv{LOMk3g>Hl)LvFzasdVQ8n{^
zCwZ)<TW6ip<g?5r^W5+^NRQ5E)YRtC-FXRkJQISFyl<mQdhM^%<tD=I%$~@r_22M`
zN2aw*%km2=+t|JEtG}yizig{1oRd_yV2cxJG;_|9ZuvF2&p(WDT;(w;%>-PD`Z(nw
znv~XW{L4@DaY*D+S8@v(<Jtuq1t>9w4tld4Q!)to2&qKOhR;0njyux^Rso-rI5wRo
zefEoKv!lGo#12GTsJ%QVy2Q*Q#bZl;KC>NgbYq?v*8+G|ojTPO5_r3L^VY<=M@YIG
z|Mc(=0e|6mK1?QQI;XZoG3*y9g{)Nm<Q((Bd4@<c+c`3)to&YMj{!c^=~(Z2XPLh>
z3;ZkIJqt-4A>`BUd&=r8J3u{j97JUecDK=f4o3`tNe%^?f>)xFt74&XI$gS=6S-gM
z|H+Pq+KY#{DLVNOhdHlds!iwph*V${Tv}ztoXD-v6i@!|8ER^=8)xdAAqj)?>b6B4
z2ri$sn6!R+06{ZodonaY^@6mOt~@V3^mqIZ=(7a|G0R`6#eNSIrr!b&H+$wXT=ma0
zrdDL#wS5>UOu4aRR%<b=*cbWp-BB;afwIGdx|Hi+ews&}o`VQ@>k^lOtALCoH^G@q
zS*p@kTey@LBtLa=EAUip_e5th6(lHFyRa};c*W)CdBI%g^qSf1D(o=kP^Key6I_&W
zg->2c(3Lqccb6r-*GC}Wd2_RCE(56h9M;eJ{py$2RSLvp4xMd!K0`omK@LfDUF_A-
zii=05dZ_f+!v$e5TAI$V1(j2r?_XTrsGf=d9R(2W*@GeiLGgr&WM%&~c9oRXdCiWD
zvtYGDGw|9I&a>d;!YValeZ7jueYc_KF_r}2pbgr^#+n))@U&z+{VDRvqJmYn-yeE(
z2nzZ8TUinP8~yQCBo(e*GW~2NT%F5;<O}Xp=aD>;qJh`D628J!VoXAs%&XgsE0|B0
zsq;o>8xxJNyY$T<ug^ea7r(I5vNG6j&3I5%QnD9yF@cJ1v@DwgyxutZ+N*QYmOf~A
zT6A)4r*c1jqq6#~lR@rKw7*KxS&$pZXnG)rGEYdU>io9X2SoDXfK8{1kS<{ldi?$d
zUA<My{Ws}13G?UQSt@C^90|ed!b)DQWgw_~&o{=o;{gmAUvwldVXSKw`kx}<37J1C
zmI8w9*}Zg8HqDNi#Q~p!Ps08MH>q)lq+5t23Lpwd#%~u#^YL+|)cGvK*{|GBu2i4v
zK03@CY65%FME%Pv!}+>qPlRPe8S{V-E|50_liacA?xVGsj`3VOx|a_`&lK2mQ*wt4
z`Cm?2wn)BmboohH-bA#XM3TplzbkG2!CyG+0mh_Q@!#b!^rE1hN!LfOP(<J+;ZWy8
z-!#Hm8Fox<f--7~*O=~D%wAa7@v`{JHiVmW?aM;eqoL`j?3g77Na&|p=%CY`U?ip*
z^_Gq2=!cU!pO|C6j-LkQjiymX)^B+%A`<JT)`N`5LG~QSi-oYCoYFN|VT=g$|K-}2
z6r1bICghIos@ul|e`ho?QS6EOap15UDPe8#SXsw3)h-*0Cp)wV1^w`~eywaF6*%B<
ztwZDaMjR>^c+n1w>k7E9-gl1`VUP}-=-8(3Y9Je?Llc9o$pqsONE`g*Xwb+JA6J8u
zf*ux3U)WI{KyVs2662k@0kYuhIP=)T4Onfn^6N<w6)|EXY&-f^#pE$E4XMYsSM{Qx
zUkbQ(TcP2*b4$V1Opi|V)1afU8_C_HeJDxc^E)BJ1DilnRxQ61;g1c-OjGR-j91s3
zX7h>{EcW5+o?ECiS8l!woCA-C1x%U)0BTvHVMPjyi1&spV<O=s2<YI!1d{Ez`>9Jy
z%o^a%$*=G3GSq>;+uOt$RqUD^+*L6K!n$EWrYO!#^GjWIUh?nMOr__;0=S^iuv$cb
zzmU?q(X#$jCw0eJ_X^mNJK4_o<8Al+(ALb?C#Cm&4%>?qsBu$!8{WD4WS4xm!e&cy
zIU|gey6hTe!Q4q9r|ux$tQhLp)oCo@awlEgT6M(X?_pk@Vig8Ls61V#W>e=zcRakG
zSvq9M-S{e#fmkOBIh?QmQSGD+8L=hLbCr&TN+ZUfIbyaG>H0YF1DoXz$p4*~hXbU<
zSKCOoM?_>4y~N+oQlcze7{$Al<ZtLFvJvU9r+vG}A5%$}bt4#Vy*t+O@_{<v^pjS>
z(rb!e4M#ylJ>ZCbH^G;E@)m`M;3!I(6{!gVDY}F50s{4@nVDE2Ohxw)d4SaP`;=-a
zIrV5jh#Nt{XiFryG(<p-j@qt>rI#EO9Hd5S7GOW8vAriF)vO!{b20~ifG3J}o8=LJ
z=<Oc9*vC^NHCW71i2E&RTpb#K2FDq(_+I*1r5FCOK^3PuKAHp@`n);1g#rSvb*%i_
z8qd+foP))&in8D7WfXnG4~i$k%Fdz43c4Jz2E1MhI$h(4>0E7!#X<{0v`eJV1k}k6
z>4vq5@zQS2JtG4M(3M&U*8J{tzx8y@<E&_R+al|2ih`igzZBxD+~*7Ot?qxA0)C$@
z_HnX>17My7<>c%DNn|Fk2aRE{e^GTLAi@q;*qq{4XI!I4?Ntb|PIO(1!{=baTp}QO
z_lZ2O;+J3#jrgAQLUBx*VlrB^aCG~fAMy$Z%`@6UrJeUEMG_0)CqN<Y((wWD@;J_d
zkdOF(YBSoo$F9bv%nl}6R|WCB#0^&dAp9S9RGp?^)}B^YuZNy3V%Z53G+VRIt7Em2
z0_GwN^E_+(eMnKA!3wWKy2$ygI1LsYvVk|Ab5=jk=_Q>d!%&!53mXhOcNEFvz7+Tv
zEi(p8k|QuUI%Qr;W66&3x9g_t|C>4({~^#c#6QQhJ&jWCZ9Z`Qcav_4$yy8~o%L!`
zfhrKxHGZqFhxM3r<mN9hn%zq;H4F!7gxVXLQfpF);eF6E!HMGSi?L|A_gdF*29}jz
z`P~v2K-fgj(!l`>FWh7a_FDpW?3N}L0q}5ok6FW5Qq%biGxqi=o<aTmodRv;vo*Dd
zFa$AxT5rU3^q#WJudG5)r#`hfEz!YWvUdMSCL-2>m?A|XQ{A_BZqZRYnEkImT>l{9
zyH<YCZ6mWtpq!N6^SaW1c*Wnprzy?cjnpuNCq#3vVGqdOS%Sdyb5I2L?gCKW?pa&A
z$gpzG_IAKtuEz?v&2BpnPqeNfhyM+`P_|Pj=dtw#H;*?l;`z=yHm{XDR{yg4*p!k@
z4a-0c<I~6({U?8Xk+lx*cI<(>t<%#6y<q?7Ht!ixMjmy4IkEJ&(_hp|(g96{f5)B(
zMo2j;-P7ifNw#?B`Uu|Kes#lfW&gyrFJSdF@^17(2zOCUeB5bOicnz+*mJYJjxt^q
z|DC;r*Uu~Y=Mu`DbeVg4@83s>$ha}SYJ)IdmVf55yylhV!vWC5-%@6_)ra#q7N2sE
z<qqE{vQ8o_#)b@8XA^ZsXnZy{jy$RLkxF=b#W|tUd`H%4Aj-UP=Q~e4Ed_2&dKYhW
zDu3ea+1WMrH!FK|-R53X#erJWqUb?>cpMEujn7H3rmW`;VV1!s!p5d(pwl--)&qov
z&JHgqOdUS1q;9@f%)2Fx5=+h^IUd7<`0>-}WWq}J#^UHoyB8Db!T;Szy<&V~7H9}E
zg2rj6#e%v{CAG?b6&MSF^XmlSv>?d4>#~PFA2)K$q=}rQ#S6q)Qc3@vRE((vJ3s}K
zrjcwMpu0l<P7R!cx&spY4wCrApyjCQAj@x)Z`yv(d__@$!3yEcV@naEV6=FZs2xJM
z8MASRwqQ6Fm0j!Af$@Ob&eQ<jRNJl9)hv0}bUfKo(2bIZ%n>HCk{35-{`Oz@W!Rxx
zKpZVIU<<OXu$#Np)y4@!dl)PN;tH+@$h`atgp&zMOha*z;yafy;*pB{P-NACy$FwS
zk+E8He_QfLF!XYGmq)DlOY2DnZ<dkrZ7BkB6+{NxH9lBCyeYKEGQo*&>~e@IrS2N=
za%b+9&B#O<CWz*sJB@B2`Nxff&UZ<2zO=cPApW+W`lB{_ixVF9BNGl0i_J{XR0V_~
zlu&}1rzhQ<U~g6glbg$u;QNhtt~qwTemXvbMdM*L>h4=C;jXZ(X){2QN*v08werFC
zjRQ{Q1r!k1d15IosyUrI<pS}zn$y>(5=9I;7ivR8h&|652XV$@R$tTSe>_CB|8=(F
z`RU{NvQ56h?g?c}n-zjv#V>7`nK<{O9vfdnqMh75U20rfsqa!C)^CKD@9WMF))*PU
z!p*=R^?XdlzWI)&cOIxr`%bv?A45~jo;fB8|0}fTsVV6B!h`|e08Ri}?^<pD|EplI
zMne;D;*&Muk2<2U$`I<WHiq*(ScLdA@iUM9=0ZwrW#tWvcxDw%uoEblIF~Djg@wi&
zZVubAo+XTGUjp8y@N09Z!f8nJK634d<_F)v1$lgP7FPTyEcRDV>!YkMWR4Zbwz5q9
zV|@R{@qlCDIC)>@n$jur(C#fmh#ea~r48bFUBY^Od~m11t<R#fT9R{|y;uO9)2LM$
zJ9kfAU}gCMLG!mGg>voc5f~fUHBrOCPPko*SihA-yxq`iJ}VSF2^Zj&M$ZV_$<-3l
zg&4uDEH|m+M1$-6auFqWV?JWlgvIvivrvD~WyXzqzd_=MexL8)L`pA<fk3ihDbf4d
zS*1V8C}m=!Dh@+q$fie%NGM;X6;W9l^FmU7M7E6}L|crigbaOue`7-Laud<+z+BPC
zxy@ogq8n=#Qo=AK2mW?gW#^z{a-|P7kaC_Xt@y%Oy5*tzt-Bt$S8@<H_s#}&)!A0i
z^nBXVqs)lt2X}{yYWpsa_@VCAg@GB%pO1e&s(UjZDZF3rQ9&@_gmrW*$_O`pkfr?h
z<>(3Uh$tL%74N6RLmhKjH2$w~GNE3MMr;fqKXM1ah0bjf?t4y%4+o)+nW4DqnBk{<
zf1@VUZ}Yuy!>{t;?^Lj*5^zH&*K#eBTRWti%;8b4kV0+2_KN?(WLon;@c5vb-Qy~f
zT;&k_q8>8f_5hH9X)@|KIbde{A*y<5NCwbb%*@~s@jiIX;p)e)jmvcmSS|KzMc0v(
z4*?AYi{Avw@8--m8G@h2)<}<=|GPc7X&rB}84$8Pmpdhug}jsEdw92p2RxP6cH&Q6
z2vdq;zuvSTmSKn|jobqI^}Yrkc^z-<7fRoV^S9Zi`9e;CsZ?)Rq8A7)Y)%g*cA3C_
zEoQ=x&)@7D*L_d9-M?<Un$|*%$SbBKM*+Z3%>%1AlMsQcz*mm@u|(SACsk!jDV-4?
zm#8X1D=A%=nRj<}=buMYumTDZwKXSSfHkU;O5?9iE*uH~ey7YRUPcLUPTNDv)VsH^
zj>ha?)m8O%P3>sqMR<Xi!Q)<QY9WsAe&>-*pK8WN&xmPgzgmtsZfSAJ(Q<vP%!k|>
z@+W`ur@OJD(Uyg}EHPtAXWMEq@8*o-`|-?8_DOx#jU<dM1^Ebddl52AV2xIvD8m}|
z%V`sI-bv<c<RUv(155qzU!GbzJ9n4#NsB-0LQtN;by7&yl4RLQQIZxZLaMv_JWvWJ
zTEEZ<p+j<7JbU$KP~sMqtAaQFZ%n4cN6-Nq>D2rJa47l%uzpG!(BgYIR>&lR-_vCK
z*<#=xR&#$8q|vRvTG5m_^CN(~eS)|)@L2f2+9Dkl`VWbPc{PjRjl86}@~>vU6!E91
z2Ik7g|9Tu1aUyn1kym`z<}whPtL?q#z#C0uCZ@F(A8HQWXVSD}b?Nd-Lvku6VCOZC
z$FU&K{2DU=PQX%Z??1}@sc~->w~(j4_2uzv>6Z#AuYE)pBnn#i*BG{no<+Y^$UxWp
z`c{_im7u5RE|NwyHy2m*9MbA1B$w}eR7{98W{%8vy-kG7^yNZYz4O49mslvS0uzPs
zcE{uL(vlw+b9W6ngm{@>0h13Uf#d*}WIZGWxA0qm?*^yYqT)TGrTXZ+)vk9b9yAwt
zrh`~RE8la7Au|C8hwtJp6^c=o5+w>q2GTQ0eD6F%3}j||IB8xN-H-hT{;J+xs+ZvV
zO-8wwB!<4jufLXdi>S&#Ve-s*i*HJ{-?zBLKc?&kxjJaY|DiecM`of9pK#PeVG@x$
zYlz1TPFF9IyHX-IIewuT0GD0rTvd?ME~8C!O{?ADXhpJf6Q?All==cit+~nc==lV+
zj!d!?6+7s>ocFxz+usU%>Ekm0s>U?)>(X=$y9;jG;8MU;BC~uprCH>gYukuIS9(2h
zXzE+x`>a-o59fj2=#Y?6XnEhaBQ&KtXLG0e>vfz@G1Kk~f}#fNNm_DdQnmtu_Bso4
zNw57|yh2vSCGj9;=Wo7?yf+=@SZu%*qi^VJVvNmrbzo0T?67^_10em~U@|?h$rCQc
zg4jKy!LAz$>X`8F^HvzS-lSiTti5l4y6SZF91da!FbQRn{}WCdi_?2}9L;e5gthgZ
zs~eY+dfa)JsM3EZVTS`b>>h(ZF5IQIXkmaV{q&rRK*;8|*m#W4_#E29q(pt6O=(HX
z2a-fQw!+&tYPeP4Xtuo@qlqxu)v;E*CX+^&F|0u=rH|_I#)O~$MEk+(fOdm)<Q$vW
z;T80}tDRn>?QNe(j6Voh4m$*@5PV#!vBIfHekOh}yC&M$-3_yM=Z<zl|H{~HKFV8=
z71Bg^_i+p%t0d;?j47C=H@4}x=OeZ5c6bUu$K4`<$h5P%@4e>xjUb4J-mD6^Z)8y<
z5F4XI2TcE4@!Va`S$szzzGF4bpP2<${ik$pL@sv^=%DY{A9DlRucs~VbIoozxVVgP
zX);1AMtoeMq05bKpx9ShPZTk4C?2oxM5Ci<DG3r%&M#H33u?C*V=?%IG3oa^WFAt4
z7%GGUUmhr8`qRlUktb|sOfn3oP%#N(u(@a-*AILK5%tc&t5DHOA~cGXc0Up!rTSw=
zZ-^VXUktWm)Y;9zNTdRKYv=p$Yj|ME^&vBUaW40GJ_v|Hp<(uJMx#-%y}^(-H<<GS
zr5iWTVtL{Oz5_o187vGwM*9IVQJbSlsV*;C-94;X8q@U8P-;$~Qm?72DL6s<{i>{4
z<Jt~Qbo+Lx*GnVUBS@=i)kH)iSg@cN+l-4_SJAVYgR0q<zxfm`C8{)lG&^Zq<qtk@
zZQxH&{5O?>-JS(+0hBqoHzml26VA$N=4c0ZWC-e{h*^6Q@jW4Kk@Y<zSqGDnlqzQx
zQMlit!qJ`7*WFO_V2K1?)^>-NOV5xT5lSpuL>Hyux!uhqxI4f-lmFR=gB3e(38X5w
zng903toGKAJ@b$UF*Rh^dt>Xj&Y-6aEdS<RMe;)K&8O5ay;Eb+2D1f)5LAk4T_Yo-
zr<Om5vC6VNU<Q1$@}!ueE%{2fb;bix+HQA$S~-$ud3p1_yYilBoSA1K#?)UaT0JU)
zX{61AHE3i4%Utd<s>yG|y{|ctl2K&SG9JtIrP6I<L@czZx&L;4C7yU4J(Y<gU|ye%
z>WYjnH9dClyF{XcKId~dNh<;k&BY?LU8s#hoho6lOOGrY{%n4L`ilN^@PS2+5QWWm
zbKi#={u@a#8$b%-;*1L-jCfOmMBkby!tVYSbqj$1a|)q%%A2X!olG(|CsvhpD2?0H
zus9VyIY01un5w`Q6JHaFq8gb#MySc}0weK^jXUltY$Hdx%v&{Pa0nT}NSC#<45s{Z
zKFEP!Bn5)?-R&z(F5aoUF=p28Q|LtwcB5d6y2Op=)1;K;+ukm=FEvyBZ<dA31c%0n
z#s$z}gHI9d#qFjGydc8;5x524yhqvgMR9cf#93-|qXgUXUvL^+`H~m>&hwePSuq%X
zWKDBrVjs9|>IS}k(|3&+Is#>N_CF~<S)wTQJy1|0ZoV)}w87h+$Zh<s#{%CT_fIk^
z77=#=h=nlsq}0vFt{NK9|3{M$Bet0Xe}_k)gwkBD4;PTdlTtU`s{ET4I#|!D;GOj}
z6_Y=WUjD386LpcSGi~MQYx+4(@BeramcQc%UAsOfha?&PL5vz%x*$`b>4h-AW?~(s
zW9#>rxFy@R2*OsC4p1l+E&g^jlm4+3;ZX43O7(Q~O9AWV%a^R!nXiGB<_IvYmyVGt
zs?y870`NF*Xt6j_@s~@Jz-^BT@J00|z!L!zY|cPPf}<id64`ZQ#onX2qd;Ewso}qR
znM9?p=3RJC>#0mY-}iykzpt6pJ{ZXhDLEB=ditllCn~)PnFT5#^FQ|q6lVs$-(Ekx
z*%EVTDD6x3lY?y&dsI%lL;Dr9Pf=}ckrv%GKO!j(5~DGQ$<vqpRWN_Z!t6;2O?D%7
z3V4meO*D#N+H8lsl1pK^XcY726;>+x)%P%pwTg!9N#LC$BgQSjS8Yf;j2%S%l|pWf
zbyz6?Ly167&C0*rc5<bIHL!vnLCRqacrvlm%aitBXAGCpqssOAPkwO-OI?r`8Ebqy
z*vktKygZsSq%ymfya<H<CzpK-61hr0LZAsm`L5gC;PNlq^_xNBh;Zr`^Lu_{?&e(&
zz00+zJ%c2P?Mu)?E0(&A@_h7Lwx86tTIJq7J&+8O>@JO{sJ@&>Nrxl_cQXZpVW_)X
z-7k7vHVs6xea2=GDmi=-!M8K%iw#IjxBsrQz+D)BDp@6#+tKLIgjDqedm6~4z-_yT
z!NNvLrf55RU40ZX<N8O3l<~Pqabrh<6%~nvHo6`aWY!QEQ<IEQWo+Wo1p;iRZ4$5{
zy2$87x=d13eS?^2d!GMz;<rcO*_pZs2r-?*uv;Sk-oE_IraBEg)N^uanr|#KPs2LF
z7nDic+&V1|7S5h6p~Sb%w2*3I|BKMjMYv+|3mo~+?$f=;zc!z4M?<eG`QJko{^#Gn
zDRJ}8t5`KfO8JQ@h4mh+M<d6=JoQlG0I(we{aN}miDaX8WMTP3PY0}tai!e7O9H<R
zXB)E)2|Z!-tq@HOAHQByHD?Ijd>>l|E?*(K!dIekGG*oWF0`;~dw<;3w`l5B4q74N
zb^rXb<7kjo!a8g5d&pe+C_|XgM1N1qrXb-UVQ-42Bp!q=qC~>d<%rniat;K#=a7`H
zt%-1W|HzU!29XF^7Gc-<yQqdSeN=WK9jbo$&u<0~eLldJ(yn#-`>zigvN0xMhz-hQ
ze`UG0#>YNh9pKHVHWO=sQS5RcLntu@rn`-C%0U4VQMqN6cn>7iNyn@fi6~=81t<P(
zsqtY$LAhdL8M^4B!t4pvA87>p(*v1lX)~~+_3<9FPani-j=)d+DkJqk0G*<JI*;*9
zHQ`NPfDbPQz_d!1c3MYbIfQ-#7oR=VU!l@L68O7EHw?$Dvk)B$!Ll9|R^>1;^UA_F
z_+<)V?>oK20e>%lqZdsvVNuke!0D_g$|v;r>6iKc=mJ&)R&QQHqiD&N6vhVe5Z&Wv
zmf_#nd=rxsae$Zo%B7knfIq>c$MC-{l99%O(3GOSJmI0bQo1f2u*;)M7-nT8_Jq?s
z1I2h!dZfd3ob#U<Dzs;)#O>l~=KC~q)t$=8%7jbF80+342%E4{iEE<(w|y<7VTc`p
zz?g?V+g7f{7@PqG2r9R!8+B4gcK<K^^y}tCeK5Q0tc<I*wBj5^U_`aP&S{MO2d!?=
zQMVR-h1qIm8u2p-I|Kt28Z&K<&_&Tl{(NcXn6Uk$U7k__u~ty8xfDi->|@F>p}oV3
z{fT5>yhV({aV-<*{;#NaU>XfeuHmf#EixO*(=sjz1wu+%Hi7s%SCBXonuQSxZKqq`
zFlu<`u+RH2dNJx290@e#uhD5N&8vxH<*M8jnD{v}Rz>XG-)DXiKTseypq_Kk?+bi-
z!;CwT@-9$LUH$&Rx=c|kx=W48s(8i^OHT^d+Z$PFIbLws=p0v3aw47LKQ=Iea<>1T
zGJAGc4vB{1IE%M-fN!vv(PZjJFfUE2gwpAAfK12>Ueh!S$r0wXDr|fCk|A&#kz|DV
zt}~3mueTltg9@%6intK*gPG{zDL;XiE6ychQOWeJ1yM2;vll1HkDOyt6vLxI{#%6#
zV`XbB9x@>*0Qi&W_V|I;ir2==#D7qs&ZlyNZ@+9l|2HjC@YG<Lwm?;?GQHkx0WIKY
z+~HC%l$yh945D!-r7GhNHH_4uC`x%JcmGlR5R_9Ztq;mURuh3UoaSF6TW|YtJ63-c
zgc3tbw-zpi7=;iC1QU@!1oGZ3A`8s=#8NspjSIc`7d%xEZnaeJ$|tSztP_l}vpA&o
z27`jZ-uL!0Z@Dcf&4vQR%^X;GOc-PFMfc_`HY5bw>e)<wwkKodq~vvoy1aEgchpI_
zAL`#RR5H#+&)tYLO(I3QZqVnyYzGVo)}C2*YCrb*Qs=h(6^niHdtz;iw>UiQbC*=c
zW>Xv0!_V^$!C|nFJk3)gB)T<EmH7My#dI8@YBXfgnzENUy@seT3ggMPpT{q^sq<(S
zebpot1827-6uC9g9zUCFF0b@mBuNaqFJ77wF=KJ2KZaa{3Ad5cYs}SY_|>5?CEngh
zl@Ovnd>sDk6+*)2q{#gNTl-T7>H)B7rdPJ{Gk*3+su4NvSX~aII<Z|PR*4FC(jPKZ
z<n%rjVhcr66lUQnyJ@!=q2WkaG3q^aMCYr99*7PKPxMr8jv^`*BVKw8o1oE|FACZG
zZBaEBdUAu9V-#CY#mt}PLWROP=gfI(WbRA_{J6}vx=e%=2as2xH2k15eMrTRBv7nq
zQaleMN!~i*-2SKCb_9*fKlRt4yz;qI*X5r;asecH^=|6<^~$~fLgzm9jJyl!ln^3G
zXRnh$yJkXD;VY@S%lU3E-UFW6WWA65dk@pel2kI*zMV%R>`glPx|<KY>F~cj3bY3y
zY=ru?keSTkGr&Rh7O1`up3?`dI}60QFOb5d7ZsPg5dXgU*ly*Devh^VNo?6Eldr7K
zpgQvj{+BBn%UTflf%&{0(QQ64rW@h>$3Bj=dPa+6RN7)JY`Y($$%n6hG2^HY+`-jN
zO{{(O&gcCfH#0xxh^eU~GHC!}8#_ox`)qBHImXlVzH?tO%V%(IYX~~6!iQM9eVv=w
zBKO&NHm=1vYx|PVzKNG+ZP&VYu@!9`{Guah^knh4-LW-83YdZtM(1_|oD4eC5Z*f^
zQAP7n^?sGh?S7t^`L<j*@MA$Zi+#a|NI{=Mmil1m>At+yVD+g%&y+IPi|!QEYzvwx
z8l$5rt;zz=0z7w$1>V{l=!bnn1A8^<>E)q44QmD|>Ei3z*afo+6@6168O77M)ssLm
zJ2Mv)YrgwA=c5M8J_dL_pYI-mS{0OP@$uw(lR6JKOhaCic)S3vZ_I`M4txW;8Qnny
z|DfpfrFhDM6%^O8DRx+d!%5s~#UAa9omcGHb5ubrT1*bvk3Vi<wD(UQ-Q3%{4hTJh
zc?dH{H=4&LwcZbpIR+x(xVbh;>fb9iGw%{Q0`sNMXBFD_+4UQ+$-;wc@BBuTG=W)j
zZlE`El`nF;xI`h{7uRtEdc2)%Du3#m5Hk!fM;VmiiPy#u$LosPYlT+`;}qkJu7b7e
zx9s(I1y6oDX}YI7|EcH*B8gbseF!Erd=~iVP}c)eLH&*C7^w0IRi44o=d!>;1-I#Q
z#wvqf5uSH;6%{3&p<ub91OjH~LgZMWVu6A%aK*1)n7JoOrBg73%LKU5Ivammg46{Z
z922;mv@LAplxI$VQlXDP8nZB@XB@Y0=;dG5>m`LjWa`sRGdL<!YsygA4PKErzmgXZ
z<--cKA;Mx{2&Gc3%bb_5t6tau-<|qw(M-MD2$XRu^xo<|Ja3$8n(8!wbR6b&Ej_F~
zpEWLmJGw!@o`d%yUkusq!WHtPERoj+a8DD!05yv!bg`<|L&5{Wf77o$7@9lhjIA1c
zpBQ<7oCu}*2g&#p4gKO+v~NtS8Xs;&xJZ3GE8g1q=uTzy+pV67GT~rj^2uK#GX}1!
zH;+q+6YB1I=8=5~RRtVEr49mqysE9?R`yZq?oHb({cJtpRz%q?lMHv!q(5hrJ6DYo
z9f9*X1dV-moqxP+>KOqCop~NW;Jcm`3v1MToW}3b<6B|;VB<Fy5po-zcd>a8uwWMN
zc^m)g(jOKi05>M=iwqpyGtl2Zi(n*a%o4haM3YRPK+N&7QuyL=GonCfW*;6yz>v~?
zy#RCwOK!?x2KzxEl;4(bPg6c6cSdNPXhxVmRm$Ax_=rRQLaITJ@sA^g@IliKRt#jO
zLfBc_fed^_89XY3Md7DigkG|yY8C(VYVrjF?E;Is`M=zCmu?z2>A^~y^qyvV0)VDz
zWP^xFq*}}U1Q0glwwVDS;zM!T0YLBIV@=EFyJ`UV#@3f^=!Dxf&DAFC(L8uG@Hz(c
z3u!=saNTVa$Y;c-7FOTO3OkuQn3BHA@>$@YL_UF%!*2_-^UqVc@0IBu1#d)AvYtqG
zKv<RNA&0YbQ-KRF_n6X0(gmOjFZhd9$afsdU`Ud*bNA(K*3`{4Mww(?VR0u53do|j
zcLVqE5sCgT3h+X#_T#n?&F0B(!wtH#xmscXxYUHr$MzMn>0c>i0ka{{9!M@G7&Nus
z`k3ms-5mg3U>kBBepz*cvEJ;eweV0EE=0i;Pv0`cmz)^SCw5gmHJ{zxnkziL5#blF
zF38CbBcH7UILmq89U5j$i6)I5Y?^}o1Z(+SQl-Aw_Iua+yc!yz4|orKG5KP#KYMXO
zW4=Hi&hz&wDAK#_Bp3W~miv95kI%dBGjIQ~3!cck&pw5Z+BT;>DR)D-527~xAQ+WJ
zFVNxtm>@WtZYsJ2<1l`%=R-@7NB^WUfA{IL@<C`V`BJUc{F^2IE_5Yij~~Fy!3u|Y
zql#98^gq!dc%gLf%CHkMv{inlXTNPG;T_0W;_OL{_{PE+s6XZK{(+n9#B>3@5p6MM
z0gD*5pXCLsL50KWp1z_UUtZ)a6N&UzwSAW3m+Gm=@Y@;rO#!WTlm%pIBq1yxTRF-Q
z(oF~wnUErpA?u7NctNr-T_3gY42he-nVn3qVl$FqFH~k<X6}28O}&NC{)oK8O5)K}
zFq1WbaY^;ZTqLC@*%Xr!Ct4o1x1aOu97Ai6n!?b}FI+2l^o?9kOSX<LOiJl?0*>V!
zkmGeVlfXFT!XI45)hqD@g+EBfjfbCG-Y2}GBE7fx7Ir!h2r9Mkp6as;e<2igr8Bp%
zkH22yi-Rd@9ai1!sD`Ex2JhY+pwa<mbx$|Js;`3xYONPSQp{6Spez4R3*f28EvL|b
zrS>fOPvR8S7qEXCc@(7CEkT49xMi<+9wx7|`zxz!T-y+O=)SqNx+PlqFSgoY1|vjE
z8}<zD8n$`<w!=_Dz!yg3EmKaePVyNm!r_{AlxUJ2XiA|th&CAdq9wiwxQ6a`Q6#n1
z+{?hrE9=Bwa7E382xTrqC{PGFsh2tT%eoJM_b*bDhrB`mC{bH4kEf&@+|AMu;VJ^^
zi|)GIjXMxTiOl1Bp|^;*iJ&Rn0zg-${>BuNP+YhAJ@jJYjpNF!^@^-NqrN#joeCa%
z2OSqY*uEU0r9q2#zs~*5{%z3EyxN~6RSwbf6?Oz~5Y-%eYHs`m__<A?s6=qyu(AY2
z4G({x#*kcFJ18>|0kQJ!u%cxnQ$LVcjOn0IS|r5xQzNhYz6b*sp=o@)OF{Rs!?aRB
zvK<KT==T;vgx>$sC3Y_=PuC-UQsE&qpAeQZaY}%s;qG<?p2?CrBQW>nHsEN_)HKu|
zhn=H34I8|Bibh8B->5aStv72MDe;K>HMV5a|Ha%_E@a(cJO)WV3rTYN2s!;!vuQXe
z;tv5mPSJQ_;MpG8e(w-$Kni+m(CtWYM}stwYkSRWq&dj4%wp#{!0s4jI<R5<c=EC|
zjc+$SfOg_pE4qTeT1%~)RqMH8UK{(@M5QdNCb!mks$5_{JKs?0{u7+Yck|%i<JGX!
z#gyRWA=@{sr#0h)!~3$OC5QNQG(0zLr0Cz@+%o=DdQvB~6WwkpXmw3-Moc6HZy8Sn
zbl*&Nx}~>&!H>5vkdIByBt|&W{vfis`QMPmT%e>|<HF@LKaXYpRyA#%NnGN+NNEw3
z2R@41$H7r#nU0u04y86+CL8VlBK!K9_$w|zYUei&j-|d{P;_Ih#JN9tT4<V4A*w|(
zUC^tDH9y6)x9l#d_4b^BuAI6C0-B$>r}zg~d0UKy5`HtX+jA?Micn|C;*C);pk*z(
zV*O?scl|-<IB8u`WxhJE%*K<=TR$0S#L%T%@Q0Gu;=>21%c-YGS~ORBCfv21+kM}{
z$Q>;_wV_hm|8%_bXO)xB(!ExI1gx=y><ClIYG@?ZY((5AAQ^5sxsmPQSvnatjIJd8
zVqo5cm!d<?@wK9~Rp+q=nTZZ?9dk2jj-*5zIG5{kHha?;<dzd#EaHf_+g)<cMDUU&
za~=J70S?TSO$xx5yi`{GNRQQj+w3Oolz+eF4?}+w-W+Hm2r=Jm`!v|XIllnuzM-ku
z;s*;z0ifS|uM(^aE)@)T(82!C%d9fC()WKs0$|vMPY&XTmX2Kob0`&acnGlp_I5un
z{MlR*QyURYJTf7@pvMez{-jOM!pBI!?cBxX^b-6T6<|^Dbe*NHmA=Yn5fR&P2Oye5
z`xN~eBYqk033U!_dWg9v2LjvhaZ<Lt<oPOFv-P+gH6ADGM)g?^==YD)WCj<giR&zX
zYCZ1-ZmvBjgzgp2tdHiJ;B!DYK~HpLa~^@gR7iVtuyDxQS0uKG8OE0^HL!24Jh1Du
zV`I1&s2O~vhR9Vg>zh)UTeEPsr>bMo;8S>GNUV&E>^LEA>oK{Nk6)|V@-ZRoK@)H1
zIQU5!Fu+$(^nMe54|pjI*HHX6&TZrome01y*wIXM^Hq}M$p^oXgYGI-720R+B3C$I
ze9tOFal(5#CHOZW;_oPe1t;GxrSo#70X<JNma(l}MJ6lj<##oOKIY$SuC%~k-WF~M
z-)7MW@$ZNhb0pUnA+#ky_*!DqFVO6LC|}giMpMw-7fMB>@cM*C)oI9J{@%g{C(L@U
z6YN796$F2}Nt`3tA}`n`u+Oci_gKXrJ^e<%Atojc;E70(tV)DgH&PL94RVWrcFrc0
z8vcs<4f!7OW%|%^D_8Cx1`QY<j6X{L?@DgTaTOeP%G9IMW+vFd9FlVB7sc5~0E`G+
zrR<^b*!p3)NA1lZ_OrlYlKfxQ%axB)FytKgraH$Em8p_-wMTE7z~@{VY#)@qXd-eJ
za2DJVCLwdQNjNGAv@TtLV+kRH;yvTzQ7JFCH*YCj2twcfvb$!T>=@{5hPRC7;|6%U
zJlw=vkpo~3TO^*r>G>nDZ(&d8GBU@{Re&!?m^TZFA|o`arGt*~Vh2jVwCI{INZe>4
zLg-Y<2wgl$KOLD&1+zoTfILAZv|xAX{46ZVkl5}l!!BT)hkSH>#qwL4U)tI=g6y*2
z{f&i>S=jIlbbzM$nD2hk$BwRE8uEQ?x|&0wphmS|xH3CU;TFkYPnyODXNyRCA%99w
zb%<=3ok~~`oe4bt$vSSvWv7n#r*fkuakp?5yw~zB|Gug=mL?Q8BAF28P<k4}Su~7t
z1zcy%+HLD?!9PK)^@2QNspxjXboh?zhzJ=)nz|0Yx-0s6BX-{x*D?_fG5w*eqqguJ
z-a=pF=P{2;y$_(6PlUI`Zs{|k6=$`MJ0R-ZDg+Lyz;~5M<n+jk|7V9>j|yDEeR{nT
z;5yi7#jfB${Kk8JkZzvm8bf@{-Nan$b6wQHkninIR2?eKB9aBhArjKbtY0g4CTa)R
zl4ZMU*x_80f``qFcJc-6dzw4ZwR()Ws|<4a+i@1WtqjQv;k$Dc)laHh_fJbJMDJk|
z8v)rk4GhyozK=-Cjbo<(sLuJS&)wd<5~|H7t9NwJ0ajsyYnJy7xSN;lOPd*8rGq{w
z9wO?r(j`wb=~x&}%ibQ4-MmjG1g#;Kmy=*DON{CSP3bF2&%m!%&%NG{n_&N8(;x|%
zA{O{{4C{EzZuJw%Swc(KKlOVm68Qv7lK3KE;Q8r6qh`=fUJ6S1Q8N$+L&>%CHr2ND
z;S@grRQk)XqpGgqmOt_?OmXKGUa1f0G(Xe9t12Mz()IH!vydW2sAx_O$6BW8h@Dr0
z3NKb*_|<iVXy+zM$o)1JKpOgE%1}uSuoz#p>gEUC%cSvAji<i#s#(Zc_=p@4>e1IF
z=Z|=y<F#h6sxEVj)CZEDkto=0DNG>+1mQJ$49>hWT_ffb_&}9mxm%#XRTca33H9Pp
zYc<!K77}<(dpWi`)={?&`hMZOJh0yyX!X^|9bL;W{N$^*(&I_tdV(>qLnC~&?b&7e
z(T^6~!ir$rg|=ojcesD1%5ps)LLi0uS7{)fkoPn#^Nt6F2i@m%{7bmXK{~~vv5&K+
z@vE?xe(7i6b5SsX1g$Pa%rAVDcQ>1kx-tl){t(}lBm`+=G;O{##B5t)eJoUficor|
zm-3(~*SX~cu0`fO{sjYaA2(Yf+-!c3^6iTJciSfLRZ<UzuDAt2#%XPEvq11}+H?^7
zp?J~O(ntcj<*`%%D&Q*$3xoxXnNN1necSx!37llPSfDY#bxL*_I!@!<a=U>^h)tad
z+SjcY`VbfPUkqAWznM2bs8#1@*GM|e+?mm~=gf>W6RYD!rluUjV`xXRIw|`p8<~lO
zm)P5MRK&!%g$%|nM;An`!76SND<CH6kIh+oxi-e4*bxos+(!*$a3%-xDdkiec{<h|
zXI7FCsj_orE~-SA`!xSbNS}U6gcu){--i&r(TEc+!0|zTcPkz0+q{@Z6UwF<%4WK#
zKtMs9i_SM7jFHQ{;GjhNVeDHgq%tUzujPa?k}tOh`LA!2Nf5n(=wlfs?BTRv;aL?7
zG`#z=a!TWaKO;liofR?dI12w-F(El|UCrQSW4slb2i=(C2Kt@RBY%U8p8|cR4?S)O
zYgnbb#c(fa6w_F$4|mu+O9t<}+B22&2?<e^3IY$rZ|Gzmvh2)@yU2eaF4Ht=&mWKp
ze2T|Mb9h%HD=Z3Nd%i^Xp_XptM{~-C27tO%i}@t}cX!p3u!fYShAKeHgcm8=QgRNH
z!E0XFe(9<KJq1KJ*9l4L1;FtoF5Jerc}zET_+xp@IdU>XD(Q0@)_BS{#1b>Mzwtb-
z-7Rgw3nKMRCH;N7pf756;7}+3FF&!GYr`eNtt>V6=6xVdAlD_@AZjmA>(yNn)993C
z+@dA!ZbR@bBtPKc-jg?{E9ydu97?MTSgs#!5UCUw1&ZV+CIsN0HZ~S_xV`wz8D(5b
z$;nC`Q5Aepe#^Tv42!|`_8vGGBfAd$WgmWW-6DA&01pCv^j2TXf?d1jLGSlhJXcZa
zObcMYFWQy?fh*D-&z>t!%^c93r_%I6MnRzayYPZ2=pLmFL;-u$OncmGw&<OPLhQ}k
z;4hdH3>XWRxEhaNrn~6$xu@)dbCd4J4pyZ)^a;lcx7<K-L6DHD?`NjJHTVIKs+&pe
z;zghf)gd|&!ZTpu6ih^8kYeS6U#ZV39dltXaijC9Q{2&JVliWQy*Ltdk$HVk3D`bA
z=;-G5Rhau3lEM_Ubbo6a&W&u1ru+@=cFwnV3%=j?PkB)HaY*+I8+&sraTFlCDF;b|
zcVpKy!n4gmLmB@%Mqbb`l;Lg`fX9a$XKubJnuD~&2;onL#bs2&gtrZM_KCTp{&+<q
z{tWW`T)uKgT%nXmuNp+A=@<}d9w>#DLCM-Eq84F98n{Z>whv#@IMm9T{j~11zb6}Z
zZoLYKurZ~QAP6Y0{ba<aM5L@4<6y72z95S9Dho+I_0@-#@mHfx`9G#H1i_7VK^$8%
zZakKMQW0ItK{Wg7C*S_gYH!^vq;c*RhPp$|AI#PRcezWMNU7wG;z8!TuH!d{SD-0o
z36?c+;_$<d&MA^p9Dsrc=oyzg^dQu4dvp8K*2Ux6mV^^di`khB`TY;Dtx=Ckvk@_0
z2{@Xug~+96!mxidF@Wc+#c{n|=`SIc4vGEBlmK<Pa~|tDN_e2rPPzW+8Dh{w-hCg2
z{@#_=KX;W}>PfC=qzr1cDJp8Q54kTCe>MQa_}7+#Eo^3yKTS%d``#g3n6Dg8zS$N&
zFM6wg`t76%Sc`}X>!_%GS_QG0?u(K6fyO;&o_1%d+uV4?Ox76BK_@Plru_{8Cso4~
z*sqRpYn%+v_sEc;z_D5Lr;Pc&5;03>fM|WuRYX4SC>@=#H16e_^vu2QpNux^Z;qdi
z2wf-!J6&kN{2?i^_~qXGWv6ACkZ}IJ#ooR=nz6=RmM3ti<xZj!qOQU1)jEllm?}Rn
zc~k~J65QMm^M;9Np)g%MBT;<wf@b<X*b-_Wj4WKj6&FOEC;84^N}gz1I^L04bmLf;
zY2SsUlp$efGdOc7h@uT)ou%lqIGj|@f{CWJlHL89xm*YkgDb33bXk-zD@yWF+&2*b
zc78sK<hQrLifQAwpXow4<_mQ<?G$^;dWh5_r#TtV?Xy4Ek%W?(A~Rqs{w<;rlC?<t
z>|Tax^LYp}L{+nO!?Aq%^0WVFttOwM2^KfwhvW|^9?8-ShRMpZyg`okCO_&XKT<N8
z)%w7$o#xL@sG4*tRx{6U4NGSZ)Fa=jjocfrMrXL@mPy~_U&QkB;i`{HN*NXS@UnMK
zNUiaqk{7NYWNP{E7NW+df@BD{)Iw6eDt4@($2S!GJ!)4L{&Mc-wm#)7w6_NOdDiYR
z+{p9T|3Y{wFr;<Z4~<W$1o`!&FfKu)XpU`v4Szsox_M<@I5>DsYuRXwHCZy1Hs6L#
z7^!;pDhvBh4<k6scZMoZJ#(kZa{%vDZCV(wqf&#TGS-wD?-Tlh@K5N&iv{k$J+*W0
zLp(VG4tnJ+!}luErP5Sab2f_Pyt9ro{)5@ow$4`p{8+(v)up>-TFsM<az@^1w|=Cf
z@2{w-yct?m0!&u8e-s9ReR5j54~W+i*g}a#d8Y7BBiV=Q#bQ}3<6EEfq29|~!e=iv
zNirAkTg1n8RB*_jubZ?~W2mzToJR{oinhTLk34GAmgh?C99XU$#YIJ%NX}_e!0A|B
zab01cO)G0k6ReD~iX*!JkV}YFjvhV`+W=|*n9ZGPoBH{CtSq5Jyom~m@N8tf!;JJ!
zJysUNvF%osnh3Qhl_%DFhI1_vxun5({~RZN7=o-*#_b-lA3`keuyA6=#xaE5Dlabk
zBjWnv+%;+9eaAF}GH6@Ubox782vn>Ni2h)P65sUtL)?U<3B-1W!qB`H2pi#O{{8oZ
zfrAOzIY&a^{9^;EQup<`Ix=G+-_*uClTpi|o)K-c%)4XVY@Jz|Q7xK6m;e-v8QFU@
zMGok;59RBivDXeT>|iR<2>;Acy2+~riF$g8TU0&%BL{8wcwSxy*h2cMVeWdOnj8n4
z`MZ(tuA;8E5p`9)*Rm8kUn+MI|L(bwr>!r3%@{yBSf&8I?~HN_zPIu<Z#~ZaBxz3O
zns)3;OV>W3DkP<f(hn*UZ`CbOpHMPMXkGSGG1!1Xmd6PKe*2gA^|3rhg0|)-m0USb
z6hM&Nzn4SIE(m09isDC8w-$L=_><t&-V%>XF5t=<F#K`~Ks36!=jK@7rD`$t!c1e@
zG`?zt!2BkQenW@DyUpjm&z`thr%uU1OK5#wj*fcH&NpaQ2k(MRI@<kNk59Qqxr_lA
z2_AY+l=FatT49Dny<Z|f3$VBdBAXq>CD6~%f2mBn{Vep2Pb%r-F(BY1dnnTAiOTCC
zq+rOa)uTdr|9Uo-&j^KP8vpqARu$o*J>cczoD@%iFzUqi$RM4b72`6VNUoPZD8`jv
z=DcQkUGQg3_a)GU>%^Y+T=Rm$PeuQuDdTAbh9-S?z0K(kY09^97ykE-+`p`JD>;Wf
zqz~01%h8oazb5;<mT+zn<$0p6eEe)tu;tmEg~VtZY3Is}zPNC=!BN6Li`d0=ek+_J
zHBZogf!KoJmb2oA0!fJTa0!5`$yma&ww`ouCf>AU`H58kq~E!GL6kb&>5CkKN)^Jl
zC1ZrKf7iBvuqb&r;!fP<-UdYTwu{Dfnh%*a@vQ25Jo^~aA5neZ&}nf!&&{vRva;BB
z)lmw5d$js~Q^jOw)CfsANVf+|HqKKwUcvxda9seZ^!?5=&vW_ta(l$2>#%0ypNz(e
z!;R9!5oM)}*1ISrQ>?@}py466s$@~%g-O@45nSH=)f=tvG)-?Y<4v4xp)&Z5YYjK2
zZpFS4{aihPq!wjn`mG$67|}&(MB1#(sVIbKu6a%o%cXhb7PHntL;`U)l#v^bEu>f<
zyf98Yc25f^It=(2rqO@hH!ZVbwWeWhE$E4-E{HS#%zPD}mMAiY9PiBBA32JwFk&r7
z{69>+bz4;Z*FHQmG)Q-+beD80ozgvYBaQTcbR!BV9nziBL#H%IcXxOE#_x3>&+)v0
z{m)*Xwa!@aumWf7@S$BiWcq-aSIdRLjgWH#k|L%3khf6Y5i|DvPWZSb)&42@^et4b
zT)W!rkfC3efGE;#w7IWfaW5zIo^rOAGjAlkdnaqyD*Odzl{|00GjoCtWCd<ujs@%!
zzWzGCG_Ul1YCUy)9;Z<fb^2Mi^!s>(S={Ak3R62(i}p{cDV~Kyfzr8&+8Rs4Prnfk
zeqQ0^@#9uaiM8f?CHv-yCy>eNw>?^Z6UrUcXvEF@yc@n+VOXf=?`3`7iryUN*ej`e
z|K4!<4y?e(td%qw%rZEn9!YpQI^9;ZJe^+Kf4my)#X4pXxG^8yX6QZY65}_QK(NH~
z)3V?%O;anssDZEJ=lajPcZU_g{x;7O_S5YajlOI4vN!%`w|}q4$n&z4LnQ`qX0vrX
zJ~fr8_87Hmti(={A~jHXo9`0cSXh5QOer0+x_owVe8h-l?K|ec?J!A481m>v(0Mwq
z3!p?avi8|M!S-4`e^Qm@G_jFCOKsT1a{%6wmuX{f{z;AT#Bv)E?SA?t!0%K-HCZa9
zw+wgJ%`IBE^85PDnCj}1P1yF$g%1YCt#Dv9>zH1qR`=xjOS|9!qr)^#WF_XI+nhH+
za7$#(e2dQ?D>s$wD()4I1Z$QT{%wH^DO>Lj9k%lBpRRpYlPmfoul`4W^Uec=TV2;3
zRXmelVgm`|{}SAydmgTej?CSmD-YWq2f)b||G3AtWXLd&pX}q_ak&h)JX#ca=<%yB
zW^g&w)+meeuKRt*-|(Jyc>7w^*1mmiicIkIdw?*`gD6MTd|ON;U%xiLb<Oj$ao?!S
z9D^UAiF!>p<PU=O%&Ra?syx!Zy^MXT{0?`;XaWCXKJPn6WFjTY$pVea#c_S3{k`lQ
z>1C&kouQt}YJ{^O92`#3AXOi&GOxNod$!%~l!_Ie>$l*O`SDLZ<<B<VVWLlG(mSN8
zffF~UDX+^NqFJj><viwej8CJ}IyOLj2atX<ptiy>3HG!%GQ>**4G{hTFx2zwwvSbJ
z&T2-bpK^L*q4RD=0gF$`hO#n2fqs{tDPbT~RmY_N(Q)Yz+da_i6c<RF;>8~3P1bng
zy=$H0vN@aRy|%FOUb$s={C%izsLVX|JLTs~_QeiqC)k^|QjbZja<00p+vaRR7Zx9D
z?{j1;L|F2>-l{Ss!Qp2}*H+Zm(C*K>UD&S{#8Mo~J`ECR0l<qT2_R~Oz<1M4EZ$Ao
z$&I&#W8r*^4uJ}G9pu-oB>lAOuT})d>o!F4cQtV%cNzq_hpunTWE5i7{nZw3-A%X5
zifA+QB3EzvS>JV7hl%^a!vo)lCFMYFZefaCem9CFde^x?j^5kqhW)rg#5`Y|j=?;7
zYK!1^z#(i?f?|>mr;L1i3a89c>0?_oF|9d~3tVmU?mc(Sa&9x(p%`W6tcnQBP9Kq;
zHMoZo*LxwlX5b?2Oief6|8M#6hZT<V-0PE;!NtKQOS^?;ZMpo@(JBks;EMHZ-;xza
z`xN^>Cl4q%{1_$ngJCTnmm`mxKBm~8WMtrT;`AEY;ggeecWIfP=H=nKT_1dJGC8y*
z>B;aaG5RdL&?YE<VIQOPF_ZhS9V>lHPnmeQeR^fyTL|bxmYMp<f$+&G_3A-I-zi}k
zk8V6)nR}KkZMuq7Gw$BbXzc?{27lpW;pDB{e1dLl=2|m?0(@#da?)EXW0$ch!fN{5
zbWes|MLi9zm$Hj#ReOHba@Gv&Aan!N3=evG9fh3h1vAf!A3fDao-4fHVAeG98qxwi
zUC1<>+9aIX)2U+^#o5<I>_`2PqLPJ!h}aBPf+H5G1AgpaG-wZ@6`G{y$7;HuI}KHy
z(2QFY%3<LY<qCLCoH}49279o$dI)@MHmRakZ_!Pr=HZaIZcE(Au2Ny&Ei-}1?V^>|
zs#nioh%E0gFU*r3=HE5CZ5|6RtY@%~OEGTcbL1l&>VTQZ{3b6Wg~!$L;`crGP4!kW
z>G7_;s}5b#ro9{!P0m{){agbRbpwx<<GG$VEpHw=$$3xY7b1q29D}m$1@<yQSAt@q
zd@YNML3oGR%_mM=V%#MN1=ZGCRNB<o6jMIa0S@}@j0B;Z3l2-WOOa&D-hv*@Q68hY
zZqvLrrbpt&Q*F5LesK$k_e81+qghL^m8Y))efu2V&z8@pMxq2nf=p>FDYyL0@bG3$
zzZ~NoH>?Q+Th4@zgf#>O`8f>#O6_*OdyFnWc}z3mzCGF=leSg9DC?{6^K!9^LyK{r
zlc@+CmB)7PP%&RBIwYM=k@R%Coyxw1JDhSWOf`H#XZrF$NlCd3I-xLaU%>wDtZ#47
z@S&EO+C5b3iyjaP?u^Ua6RKzUg+6H;wN4EMQeP)a;PSq2z4cfO(wv1;2WWuP(t!z#
z;`>TqSp}(M3h+cd7_EOWjjG2?mbwC7mrgUhpJ?J%GbiM(e1|gfVbr~nzSg@f*ODd{
zr;Ckc`XdsT$8xtdj-Dz(FT5?F206Xj@|%tXhWwu&%{_k?aq2zSr&$GZtF3z>O}q%^
z7EAXL$Kw%!cPfo;2(~}bkq(ymD5?*BjW)b7C%$eHb4VIs41@J4XxnTn)H^iqm&xyo
zg;LAk{q<vQ7mGr9$Bubxtvxo7FLC0s`@Yno%p(g^)tBuqHiDRTO5a6!X!A3x&QL-2
zUE1N3T5A6`32aq<5yz&=n>=TF^aHehq*$2#DXQ!@Ue*5FZIfN!sOgSx31$&H;GPUt
zcqSa*d4`4l<$ZE^<?G4GoR(s>G3&bAuk*9>R9Fm(>wIYrYYgD(|GnB2{j2bCi-0Z0
zWp*_U<}j6uJ|FLmPGBrW3;S=7djz1B&-y|VgUQ6zL*GU`v>>%OtYi<ARe}|x8=^KW
z80zO3niXuN$*N$Kdi-)iR2<3ZOGCPk$?JwjU+a_6mzuKa^V&-Pk=BFQPha7t7Ux9=
zYkzWQ_6^)Y2NZ})D{EhxF8-z0Rh|0Cz^MnenwKm+I6l65CPlxD_X5nPvA18bYjc_S
z<(iA`G}a>9rjwn8pdUW2S7B)~lo6Pg{ruj{bmWtUUeZrN3#W~O`COq(!=L{BKrZD%
zJ494xAoU<IMOEsnT{oKdcPzk^)&P6jbm7yluf-I}q7omnfI>Ke*OB=R901y%o834-
zB47mdTa;U^ij45Np~5kx`jgoD(p7qB;y1GKh>kohK6Fkq()0<+_NhTL+x+{^Y;-rC
zg9WA1jB8!Qa#99jO+1)rq22HRJNbb#&a<r08&z3}#<>9<2zZPu*_<?Cj~a`15RMWQ
zcl#3ZglQxif9aE4Zm)j}!(D%4`TC>Z8UJkfiMDOtH-O+)HX5!hskrO?A^C0cMV&bX
zuR1Y&jMCx?<95GM-NWG#k`D&mxTj#tgLuf%R?o9YWD(gqt@BWeQFef@@}*deD3&`q
z(z1q9tHI3A!Kjn0#pbkFEX}{X&&CZkPRg+w?i9Am$AXxo751b6Vm(0#J0GVLyv9O`
zo#rPM6~~AKRF%0OEY`CpfY@INxh(!euVe?W%M{k&cbU%xy!v$-k2#Go8V>sxnWC?H
zwHk71kCfh5t)Bu&GbqU}!ni3G?foH%8iF>B{-p07P6>AeIF<!?i_KP+TJjQPY9y=s
zWiHattUgK$?<cwXAZ6U{#;$&9KuE3B<`XRat54sNkrWW2i*E+u&3Fkb!Y|jC&T05a
z;rqUyO;MT(u`E6Y65CAlHe|y!Chk}fK|`!~(ODNlI_(v9f+SChQ-p@~i_H=HfaYxv
zsytH&K3GxFUp}H2B<C_~9tq{%FToh~!Z#iW8>Wd#c`(EUOVz9mzEm@7tWIM%GR-D)
z+0n986}Wt?J@)8jW<uR7Zk&)mhZ5E_y#3tvc47lI6*-K0|2u}1^;x2Y-2KUVpB0kS
zlH==vr#+TEdf7H&gDX$I#{HJS<eG+)8(Eh~(*&>o{<#=sDV?{h=n&3cl0T9Wk||v0
zTKU}ZNBs>|6uLIU_o>)9vi8x-b=%XoJE2_W7Wym|>i0F{*XiTemg(bVi9cv5hswN4
zt~msj0f>Ut%lCTi82ZzrHGi~<d`;FmXQTq~n@RRkiFmn8Qy%oB&&}Rp5ed!)UtaY}
zU#@lE{&8d`3+y=Y3$@~k?9g~TXuAxbv>T@8vAau^L?e7<KbfhOxe~>wnX=8Z`ns@h
zFjvKi^|Gs$VANOra&6sVs)C96M|;0SP%)kY<dn{DmE7oQ(LiFUv9QR8DRMCvYW2k5
z>`b%OUA@7$i}s;~n8u-fY>AtxzbudkblI7ku3T&OPUTUA$RkfNVc9uVtKOiUw$SxW
zOvsYNIg>pQDP|g;gzgs{=nirw1Vw(JF%qpc^~~Xr`dy|(Z+ZZ-F^vFx)E$@hZp}X>
zZl+vV))d4MmL9Z=+G|CSk)Ztcu_VhAoPg)K{Zwv~!UI5#Ni>$yq02dSY+=TF4?wmY
zr;_~L=JBx_Dkt^4Z_`%wecY2VbcF^lMf5LPN~kevztAKrkV^8dEoayIV;Z%q0UN2p
zFo?Eo+_Xvs%WkZ6#S&3Qh&F7}wIRl2*B-b#cE*F%_k6tT8yTjp3>iq~7xlzL4_$5X
zR3tsRTI6FB;4+EsDgQGJDxA>>MOlg|5-b&=T$%KngC7b1A>44~E2wONyA$dLC-ZV+
zU?TPydHbB-)?sq*XpeK~WxP7u?pdN=hDU?(OKk^fb8C950qv%9{BQpA-fIr35KL7R
zN2}wVu-sLbNwQA){pvnYhd1omqFI#_kMRN6p*b?o%BP%g8Kpy-d@z(mnqd9yu}r;m
zE^>@2Oj;Uq;3$G?=dnX|W%rw{-O-P7;Y|$EG46Ex(hjEMxt||MwtwC<9#(MO(4CLC
zHWXB{DH~sgPKKH>sAcjHPmObe+np&VJU^HsaLYHb!7rUCXvS-P$Fh)2TKajgYRAv!
z@gQZ%bGP;&{-q?PWC)iH;>B*HUn16<p@)ICBv!#_80HyGgIL<T(*R9W|LRL|cRAE%
zez1^8a_!f*+Q)dB7FB%m8cX_Xq{#beY}AyRtt!!DXb7e6G>^sC-Omok;OE)t(aOw`
zka$wbLYc<iv&?%gm2|J|Jo7JB!?5EA?AGRad4%Ix&M#|M#0(jk>AKc7Dwjf1I$xh|
zwjZJX1E;6Q=7Qhpu#R8j*v<MK$M2SI^EzH7y&vPb!s$-@0&6Z~Y3F&fczfIL<pWpl
zE)R6Gb5BJe#@~u_lvzBu-Yq!BPo*7~2l=R<M11s!)0sv{J`)A46v-#`^v)`vIWzA2
z{_*cmPr3B#u$yF=;GO4|1;~6P2C|yvCji70%+jXuzDAp`2+kBcIppw3&9MRB+N8T*
zJsu3rcRe@|(@xV?^>8a{ihE2F_ADk<uz5hNRr*7>KmLfBW3c|&lEsH`<<3!YxnS)$
zrKCNQ2{61htafBSQ<Wa8arpE+<W@<mJ#XA;AkA83K>D}t_4cj9S)<cQ8W{;++b=Ku
zzZi%n3KAw#hl_PD2=~ZO*OY@P-)H$wG)r8+l_*K^C=V4hux2;16?}X2M>|B@y)&L2
ziej!_d^K42-@Rrd(eBmVeeSR|$2U~7>R5g@B645KvYua|ToJ31IL|!kP?kFspv>H9
zK3`VkB~65*jc<SAx||sc9Q8bYeg5*g`xIGyy_<E{U2F2t8)&O4a?{*jxbJJ|DMMgm
z&p^bqotHOORQ!7ZzXq^|oRs*|^E^2{(lNEnO?)gjWJ6<|oJw_?rq7Bco|2)i)G+Tc
zH@VfaY_mwNb!50fhI`up4|n+!fiL8$K{JkAM~ROF8MmejEc<<^);w~WRi{2k{_5hr
z{-|?ryVgDXxzGGEUA2^_!Y|&zTLPk?N>5b%r`owsrI`64WUAjx)g^LS{msMi@NOIS
zPR9qasi%`Uj^{2IH=B-g$qU)?#jk!;tI@A6+HQ}oZ{^07jo#xXiP#(nywwHpX@$ED
z&3sTEE2d}^C@FR7+f<$Vv4te%%x-1DGWuIh-{U%izYE7-skog}0>RqYD4L^!kcq8C
z<ZmtQqvbNHTEA_|2V{LZ+Znw>_2<aM5f$!F{UpMB7QRMEY>n>E&%$0>#^MU;v@)(O
zZ(D4F5CuN8$v=}`Icyk{xsQ4_-}ZONT~Nlan$$syZd!GjU+UxuA`U2suO5Thz7Rmu
zEX-eZ^J*G9MlAz7BF!moe=3>L>h!h6E$#?yRiY6MA^>Qe9J)ta{DvQfS{=vxax9N8
zORl3&F>Sv!6_PHBNXd>iqqmc8L`5lY$jtaE+q8W=KxUm3c~Jty?i?9k%yObQ`Uv4Z
z`bO)#-}ASppN$RxggKj%1I0OY$~>cIN$L7isViqpCP<>{QJ`@oGiDeKZ%o@;8z!7h
zGmwO34N0f)=M+Eb!Zo|UX!jl$l!wh~2$s9K2Yav_lIvR=4L>Bmw{7v8PO~yf1J)B*
zDo^7u_BMw;RglNMt8X8%Iqq?d8@Deb9vuHpK8lO)YB*S)b&M2uJzF7P9$SS*Jwrr$
zxY6KXdK~)M9dp2})p)Y>WsJ#0`A}ad3|*eQym}G>E_qZsv^h$z@z)d(I8r1R-K;90
zr&N0Tpd#onK+;g|;q`G>Pow3UxdG0Psx>{O=v8UVkgM1|y;+yi+&}{JqvEc-pF^=q
zX=&cr=uTa)(3G+2!>?B!0>*W{!W$y8*V?#QdZF9L*;hN#$k!f%y;>;U*kwEOMm&oX
z@Fb{+m=;s~DdE0v%A_v&emn5?lefcsormA7fbZY56)W-9oAMR=)1|(SGmW)3^>XH{
z+`N-jE|sNU=ujt$OdHUIC__cD1%~DFnOX6q)fT5HXxn!t$_;id>U|5!nb}Wkp(2En
zhTRMM=SEzuNTOeUo3!9is@Uv0U_~=9vW9)KIbY;r!&-Dx`f^<(dv2IfwmqnLZuAB%
zdXo|tgK*epJa;y-a%=&G=VTx()_Uo|>W7J>)>hd$8MWy#)Y<t4W4O_2cVl<_j72Ci
zmT_|q=Q6VcopQp@_02zpwA+Vx+l_IkIqcdOcWgEJr0+Wgu6O!5j|l=PBXdl0uoeuU
z7I<6xcZB%As~T+gG~Q&j4iA*dv6air_LU@IKu@xGCvMvAe0R&Q^A<kDU8~P2L?eLg
zm0Y)2^sEQO@a=oKFbH>Cl^UK1bS^Kr=n~-t9@>kYk*oIj{I>`6*7ad`^eP%=c6yp;
zq+bfmw>ewk#QzOr<A+%8hmY<Dwscl7bJDS30kTB1W%YM~@XkAw@^=&&L<?+t_|}0&
zN?Ez7l6*mwWTj}lwoKnho(tncw1&Us%hFn5_<yn`KgI#w7>H`)I#Af-0?=mcfmlEq
zx(6kHZJq)oz&Qn0gd2dK)A#3sVXC$bk5u!vWnOaJyB<SkCV0~4JPzA=TyXD#dN$ev
zuLG7t#9K*!2MhB{6Pat?Ayi_9HH1ZgHdq#M243!KpK^KGgS$w^tcMoG^QRJ29$8U<
z-qjCuR{*KQd-xC_cyO=)pf*PYh!o=;vRSv8${Yd+srP|YMZ+|49|?eZB1Z+b?@n!v
zOO%5{N0R5Y2Hu_YB2uqk4bh88jn24t9HXXgF5DK<I*u&jW^<p0ybN%Qisur&7P&1l
zE<}~Yl>k<igxpT1q!#P4(SN%vkfEI|67X^2L|4kXzp~O)+*1`=yqAtDCGG}A|9&xd
zAG0!=e(uT_N0jMn(;xnxl_5UE1;O?id-hB;b+uM9MXMKX+VINL2jlqGAk!?jEuF>n
z2mb8`4IdR%vZRNb;F`{tiw%8x=Jd^b9RCWpU9(}^^3b*6is_af+g#+~h+cDEbqQ;b
z*-R1%H=I}Y&ghZ(78#TXKKJliRWj4j;Xn9G7_aij-*-`6hxG5eFofU$9fEp9lDtKp
zL6VzgC_PeRyohH|>PE3c=gFG*)J{QHJxf7_TKbQVE9Qa&@=h6-t{ZT82>?Yf``TQ8
z&g$&>l8R*6n>tN+Kjd$7U!^H>ybkM9q(#{J1)JHuZR_g3Bgu-R^b4C^wSINAR-PKg
z+j?rPB?fv12MG24T;&dL9ESCbXQK=U3={=&Qi4z5LU>JO)F937adjPmb|xG~OiW5i
zbKHBw$YeG{)gAs#HpKS=&nzHhtd>etrRr=|Y5!s!2R(=P?jaTKcr8_%2=u@Q3!srE
zffW+x7lb$qjjssb1xhY$j~D>5p~p?aNzRp(fH#~90NBWQ^PKGF83dDt4eN(QCCGTZ
zDG2B$bH^M+ZAf_Gf$t=mr{*#@AEdjtmv1qg3`ulgx2<<!jE@Az=Oa{oQ$0AyU-<Qj
zOoJRiNe6ff!F>2eSQVT>g35zrfOJ+oSdXXXy8W^9h*DEY@@8{k%c7LTfA+ENz^-(q
z1W;UdWSq34BKG$DW^(lY=fv}h<HiooX;{Q-oLP{ve`dT{n(|<GEe)bf@||5cn}hYl
z@m&rlT4}|njOb<)GqA<=X)zIz-ibX?!x{PX<!)9Bb<o}8Y1ts6WlVg$MG$FY{GVr?
zWS-e$PT6NUxEP*ThalzpTe`8F#MowfVuosLOV~#|Ac=fUT7R90oY1EezBCY3pX_pz
zvK;;RqKu(h!hVsI3R<}5_;iyjwPD3^oLX>I(@@7kS?FUB_xaE-JTiNXT(HySqH%i*
z=?NP2N!ntNNZ{J>EADbwT9*svXBumvI8O#SRf+?rNH){9J-Zghu=j8~;V|c)eKrn&
zHitx)ei(v_h$r9upw^yFM9wLT+?`Ii2>^#^Vi#4H*7UnhE;b9w1%8Y2^>)009lQgA
z=G3(9NY_AAL1Lhb1cB9J#ljD-Z#`D&YRqPeD69CZ2%!T63#vom{s7E_8YU*{K13=6
zpDF+e1Wk~O!LHC_5kFpU1L=p4r^fuH*mYl$THEe=Nt9VC@(8tF_6PsBf!h6tAdj*?
z9_$@SJoTGD+t{xf-HUBVnw<cQhAe#s32i9F2x}8H*>PdHo9vn-5AK6zD{*N27@~Bo
zx;n`C+pY*eMNq4u0t&deM7hGi$Pv3FCS)I|08n?ZaUjq0Yt)Vo!nV)38xtUK9=lYw
zoLKpo`d~kaZoI>LY*m_Ft(C8#$mlco_|(t~l}q@}e>GPij>mshgM=ac=9?R4!G~oI
zC3;MJ1^dT^nZ&LHcnnUl-)d-_*cAszMZep;awtu2@pfSy(g<#Uj-=<hysTpT@W*;U
z^%WZ?4D3<)DvRR<LI;KQi}~9bZH=z;P_^Ks*_Na2|5YkpSf#>+qxBjSf11tgsd%|K
zf4pv#nDg_{lRcgNG<C&s)Domj#u+>dr-ZGGMK?@=o$W#|i$~o67Ny(1lbym27e}#f
z@tDJnDn<Gn!Efs7>l#Jry9U9JP03r{k9F^*r>0AB#9Z$AIAQ)_z!YQA)cT!u9QX^P
z!bfu!+Cpy?Fj5(g_l)BO^vjU?pCcotp9)%WH59}7DUwENj11Sbw46Y--hH!gm@L}3
z*loki&)u-0>xOur(6>@{&{hChaRfdMIBzPEokj`!8uJWp9H7=Y1y+}J6gATunE15$
zVNe_JR@)pN8nLMWVNYhF25*U|YZn1;LfNJ?(;CD<>T;ag;Dm7*#oa1VD#Z0chZYNv
zhnm8$4S~glnEtGs;xVp5sf#{O({eC^%16|0gzj)IKtUaX0~C=tBIwI!K?lq(JM2C*
z0+f58M-{?)YmHeXRzR$p2?49Lxn_f8OUSq}rxrL5>{aSg{F37`Kh2jj2B4)IshM&3
z`7sBNN2crp!6ie~?ntX|d(_ya_oB`6;$|b;zbwE!_SlIdj+cCArQ21(b`anK9iX?5
zG3@ZLp!n^Ht*P5(U;$1^?NULDx$A7H3SEGz^e`amQ#I^LO!Y8W0C$Th;WhhYqGmy9
zcBglOyia03|EmqXK-wd<n6IYJm-%KV%S;#@dK+Yk`e1eYY&j$@NlVi)G6xXnBn}H;
z30Qzk1sH@KgF+&k|BB@!#q~UvpJS;<@bHGGQ|Mqj|8%G`V}%WcH{sGi#cMdZD)$)m
zIR626ff2%!Riz367|lp$g#5awrJnwoe#n%hC}*CV#*1JJ83_E8u@^Tlkp+-(UG!eZ
zvQxa~!^^aleS?CdPa9H0qeNg@s!#&A6JcHm0Ax!_nE;992yDh#MhJ!ps-cnw>+M@l
zgp89;ntyKmZx=XC$LSZgqYy-nh~TxMFI1gxa*op%93r_oeW4%-K`I~HbrO7b!nDi7
z?i?{$;}O|>B$qm(L~zIs;vk48oe)b>P=YWF+#dyQ=}I;Y{Ftm}0AeZ&i-b~?66bD`
zLyL`>Bj?reiOf22w>=yvS%@mJ%pj)P5}8~$Qi=|Vt$i}cm2nTi)E-}=4-nek<1!{n
zZBhDY0ItOMR^7LBtj7|}U6BQr@y(v?CM)x?|FXgMU9E$Dh@|%X^saXT8~-&?ha(GI
z%-S1t97X-m=y&p~k=8z!^M?iQr=nIV0&%p+?ke`t-eP#X9toNmjhm(m0r46EU+uXR
zo&$mw)Gs)MRPj7hcb_*J3H`-|1UOUJP0lIfRx48xf(dl~<p)GTvo8nwTyCsdUgo~+
zGuyh&qYfo`0qTTi8pq_~;bE&TNplh0T%6pLV2c&frLa{lm$)L<wVhHAHcLepMmBqG
z2Uj}!&g7mWm!G&&#oh;8q5#ljQ!d#yh#OA@ElhU(;~-RmDg<Bn)n%yL>a5H5?dY`-
zYg7&aKmd!danv`6E*DexR6q$eu3MMNF%myfCxQt_2(RBP=4k8fPh*@U323k)fcD~-
zK?b%64suZvCBGCYm2^LHy%zlu;d+}pG|cHBh%(cDB~Xd=d!{{{m9)XAPtOuIK9ajl
zMrP{B%3C0waZ#}&Ajmk71cVHDV+aBREOJD5ufofNp}tbj(!WLE(xm(g<S!=ZuRF8P
zdM3CM|56i*#c<P!U5@z|kS#Dd*6&|fSw@FJ7#twUvWUJcdIe#|)!#cZ!=$W__w?)F
z*>jqt7E?>|#P{mYh}p2evXyRn*Xtbz@!f>Ha5`O-$fqrR2A_QE38ro({u{yd&Nqh6
zLHS1rpx)1zy4}ZK=GM27nS{0!K=b>McfT}S>>kkOYda~F%f6ern?|C*V&_|7<}bpc
zjYy>T0j@<1DYBu-&9rhwwI<+b2cD`2GqkD{JL|9!ntMX{uN}#e+dThOzieZdDqLkw
zw6}?f=*QT1jAs5zTR|sFeAe;gxp%^ki2js1rf*88`z=8-sVN&)Kl){U{mTaI|B#{{
zA=u*MnrZ5CSO#T)<p$ZIQ&82t9iF-P%;}NmklTr@x{0R8|0qf4dZeIY6pESXM#$8K
zKH{{Eyt4=jPY4}<q++B}UsOD7ZYu%6KOQB(!SV+^m6||P9@#n35vL^l(xBuWkdsD@
znkC1xRGwNCE1(Gc6|cv&s<@;9AqJp{)eyP0pA~_X<8yYeL8)>VJV(OTmdUI++tAaY
z+N(=t32%`;2t1W?q`pDF4CYiNpRw&$nNPo1MqV4VF=tuNX|8YHQJS1pL$0p7k_0<J
zTP4Q6$QW?)_;}{Bf1SBW*vxuA$2okZW{mW*h2K{~?}?F9R2GJ`7Q2G`la*yhF|UyL
ztGW%#Xt%y361)uJ5kbfw$7K)+IKu&jQ*g?q%i1(0`CyX=1(q%9Gc<qH5mCBhJ`+e8
z1GTXmO0xsB@Y(H8T_veUdV<%}UtIiI4`pwHdqFL4%Y8%vTO*nliV$kjo8d5_DL!>l
zLE8?^xT5~+oILwnXIoL<YbA^0@Bx476wa33<L?MEWjUMTZjI2gtX(%swWYN=g0+98
zppSp?301R&O+)Upd#86Di8$Vl>GsU0ZnpP?9vwX~i8CGHM3Unh%tx6U>8sd>C*6`B
z<>B86c%riHYg03O%GvV}k_Xj0Vs0XvRsXLRsQjx1|Nj2(&|tU!G1SibKG}DJAk)hk
zj|S;EFSxYnt*Q!zOB$nZaK@u|D)p>|4e5KX=~?n{`FVjTL8Jx*YYDX?Z|4m{r)gpc
zP_Q5&wBiieaO>ya$oARc@p673Urx&$g4QHAj!}RNMqF}F_s%Ct$WxlwQ1yWbM{3I<
zBOoT*9Mu6s4E9G^C=lbNHT{e)s}GqKdk&{eSU&DpVi`GXiUA6GNXJQ)o@0Sxld=zm
zrqn>g9kLTNwRMfafb~*<<ACBE$3y35o=Nd)n!UDw#QtFHk}rNCf4d4o+CWDVNeB@5
z26-7Fa$6!gGe<QO{^NOEfBeaYIYEFZ1gA=wv9_F3ODnY86|@^Mz_6bpK{3XknJlZk
z8KMVD*Tm1R%?se4=Jza;GYA#ZfRNP?mSNzWnnq6J8?zSPj63BlHykMJ<0pZloPPiW
zYf3swTFGjah}&qoR8|2ss5LIqmV#sAM65Rl`>HD0D1Y&06%}$88_>8o%{Zi<tS!(S
z4leyH;2b4n0v2KL9FF~}NIHfRHBX;;G#r7zqeHiEpq+?T=QNU|2i1Ve#s%L~)nybj
z<}RE<Gv(FqDL~|3N9LGuTCy)O*>xJ06LeW;82ssPig$b8Sk5{O)X9ANfk+g@$IC(g
z=!~Y6Z2s@5XO+$To)*%}4s1VIEvr@8Jfc+m_rtLiN<)+1SKE-^hfE&uA3n6Q?Y!?5
zYQwe&_dsTpx=5v(L`u|Hq4yp(jra|JHeowWCw5ZJ^f?|Py#Cv{Y)V_U?2Rr0gm%xi
z4qhqS$c&G7tA~S52;_)dy<Pp$aGBo$HwKaHxTMsFxxWK`5#^62!6kg3oHT~d9F{kb
zkFD(!Y~3Wh0UQMJ4CMbc41$40DAw(_H+v{lSzTC5aVJfE_^0C%Q4QD-Y1vamwiW>w
zgv^52(o-$|6=Z_N7j6rZ_==F7&}KR^PEVzV-w=b-%mAUZ&9MI{Q<+}}z<s~DkF11&
z^Ybk^m)u>t)-r^hj$tV?=@0U{7Nezzgq$}T&c{z^A+~lSwm^k%)P5uoBnRBZl+B;@
z;r8KTlU@#ebDqFp_Y9znoK2GMG2Uq&ISX|!Ku&J2kTXuo_%`aLn=(NP?t`rf0Bchs
zz=FXlMmMb|P^MAh_0sR^tIMcIHK4<hi?fDZhbPztmduf3eoJbx9t2cF2Ra?y?k{H0
zYi=IU`TqOK?6A9hZj|swY4>$xjWwtQbDFZxxfOp37K8&Z&^tEiVRWtlC=+&TcmLn>
z9TxAJ@Eu718Is4Pl4n5W&@SY|R+_@|55ULuFhb8W*zvd42yLj;au;bdlhSYb-+y&~
zhh_LT5Q7W|rQ)5{)*iv$R>j}q6#t%=Hm(n#1Bs04XTJ0$cd;af>qu!jTH-&*WB9Xp
zo<Q(_;3Ek)(|$fc+5Dbz=OYDB(DxqSUaOGJ3%=a5fPMooI6sgGxx7$}hkX0aL#Yk=
zqlsMk>!Qx+tDG=(vx5y=Y(|f+FzAC^@njqVGQbuKg9D8dKu5KWx4@j|V;>Q}_!1=<
z6cDBC)t&@EP0j%=ez0WH(fQ3P70RcXoM4<e-qgoJWoyk~FRk|-J^}htgf;?+-RPN^
zU;<jmM}<luvpj_T0egUP`*uSnqf5TLmn)4)l?VZTG3xpSgqLTkBS1u03ql9wbq{08
zxfR65Ebgq$79&JssmkiMWxpitk&YuCeGTgirHDpC1cv-lCg6{IuzcijF3IQAkEq_|
zb0QKol&&CtLvgUoPCrIW%NGli4r5nav?hqEWg=fkmO*I`B3Ui4IW@`uJ?lu0@zxds
zSlV_xpK<!6eyjK<rtHRl(}%L|H)>0v03jUhs86jwdNY5mBUgq0ch;)nxQ7o$dCK)(
zPh=zk{mN#vLp{&|i;crSJfYvtZE4?hNDXIw5rmx%@81cs_Z?oHe|umGqivDR#NIkx
znSyih;Cwgw{@)N#mn~4meG4*nim$vqn^=fRlh6}Hi*kIoD#h_0<8`u4=Om(f%FTNP
z7o-UG&}|5s!{@{;qs_W??SVtq1?MC2{suDiRAi5`Iv`aukDJN*Wxz+Q>!K!e(d>%!
zLy&yM{L|v4Rp^#bA%}E1^~j5w*eyUwk1iXJUiKhmpu{>fc4s0H!582Ka8W=t2#rHr
zZ<hN`QEPsxFcB6?6Ow{!5{@F}8?sR2faXC*3l3=qIc39rB}mP6V)RRvz^q4ugEPV*
zV$=iTgT=^pMc3e9|BY!DPOt=I4esW&lhUg19U9mX)sFli=m0yF<^_q1LZCmWZ-5{D
z1((+~7k$|eVgcBvQwjG@a93N+JSz2>dxFhzq;9|EWpVmV`g}narmhE6*oExjNN8&c
zM~o%4+HV=*E?6o>ROz^rBZr60%71*|iR1b@xZ4X5ii3gowma2S1paGp3acb#;r>qs
zZ(<x4v@nT3eZve@&cO>$bvsu5;N1~pUPeSj+^;{in=#{czjo8!<hM<8$N($ROLUW0
zOfX&}0g!($sR@mEm*J^#zJFYe!WY8k_s;ZbSVSf>aA_*~jV11{y!Ky3Gf|xFBlISp
zSD6<)eop2Fj$eLSCr?*Lf-!jKZk7KF4d}ob3r_l%=l!F&Tf^W^Z$$Ep#+9w?fS!I1
zeONN^;o3D5KUVB{Dk1Ap|5as;!S4b^#&}5fyxSZh)CW%rznv11Iqmv0bgk0gV$I3v
zyv5LibuIz7t30Min3t2Fj_f79o2`<*$M1)?KqeH%AC9rH+$f#@CU+}jdxMfX!w$q1
zCC|m&Lw07RmdQu92(T85^A8y`H8pH9<aWF+rG`QUAM8PX7g9Rl<a9EuD1|?$e)19Z
z3Aov6NOJ)A5>5DTjMLNgr66{A0bO$JAO?^#*wVR-MOMrKS%RWVlMkOJ>bA<h#!QMA
zzagr+6W{`kpTXbfQjowMCzWgr9oL<4oeDsd3zC-vxl6>5$wXKNBMStb2MGujGmZCz
zZg22-k6!{(B|6~BaonZAW&i;w-MT(b30fQO`-y7wX3FI2oErc$*eii-7||JC=FKxu
zQz|W_#|T1!qR*((f&d8<Q%e?NLIY^x%LON+`sbnJTz9VIVA4ILPgXZ{rSxtb&eEM{
zaj{l0k6m(w;aQVUf&Pg)1F13wyGQ5JrWut^SLO!))Zm+~yMdz<<EQ{Nmsq!2iT|M6
z5>-D~&B%#LDaS-afa}(T;Jc^_t=`(*mKx}L?PI{q*aCgy+XH`&<?(AKCgbx4c%e-^
zCtzkEhsbTN_ul&B>5u>mR&7TPV^HEt^X8bB0T{;ZZVWpX71F#dU=5!1u!@2fqvYnL
zEyTvc&RQR2TM{`e;{e>rw-juu#1kcu>Y3qi^!Shde;uKvVZR9`qSa1#+S0`E3GJn!
zBcLw8TR~oT>c`vnoIu4+018OSu>t8K8ibXxApo2*-1Hsv`kZ$cJBrp)riI+XAErL~
z8i*##edRQ8h2l0aF<A@6BueR1Nvkw7xZYU%jJiLJYl{l(08uB+bO14U2<%(=DHGFz
zWap!U@xKLOHr?i?5^H2|9X~uba`yf-K8@l7(~o6QG|hyRDzlJ2DL437EGgyABd}7H
zWE9AJ2*fH7NKZw@gct?)KE!X%Wd|Nr4UsAg9=9)S>ZAB2fC8iq%_3r1gHVcq(YFAI
z#-{1n;$5FnJd$VlhLF=CK)5;J+PM1jo)2KVs|Rp5SW>AKi+~m0K#UfA%Bea|3M3k0
zRw9BTo9&@*I54?vtEe7W8Lan_aZYNA6Xg*$44Xp$4&&kwY-6CgK58jKbST$&WtiA5
z7&_otB}>ErMunnz%NZ!!_Vkd@W^yGzopL+Ipk9izHW6>0e6RriD;9J#<`vR>uy7%-
z*m&xzDtZhZ95V$hf&*yd4Whb0lF_AAzEN&T09y4ckEsLXX8ts*FP~e#uwnhm3or(c
zO6Hlevk=@f7cd^TDQw0+2u^FqJowa`y_SC^|F|DB-{h``A#K)S0+ZPP5&jw>=s+(E
zMW8+bTF4)rlwa|+d$MypNNHqfR5p#<SK50FL~rQG?umiOv0rGD7FG!M)+1fJ%LH(F
zx_2h*xkHVSK`WV?tC7{+ngX~IsU5pf#=TYoV)bvnH4rh83w9;Vr$dQUX}hP^sl>Yu
zz4#X`nsHck7Q-7=f5qoi8MmovC(nZp;VpnwQW&4zg-6HGSixkOznkHgx6caWwOKPb
zjcqaqo&#s6f2W1iFzY4IhDQiNq9Mp%5yTi-+Hg~;cZ>hB$6);xqi#lyMPPwn_js_l
zmi|WA0-6SEEW8aNT%eyTVv*=XW0|d{t`soFGY@-$AVb7LT4=5aue<1}t1DEdk^%XS
z)s`x6M^!|xf@%P?04$qm0pA=W(1GL~CjjI9KF|sfug@v2NpbG8NZH4-_;D7fHUDv?
z#z^0+0O~`X!(%qjOd^h=wqH|$4~ZF|pQOYVMmUr8fcve>AK!Njl*}BJWB2wPC7K`F
zr(n%mSZ#r=aQeyn!UF}R-H4d<-=9Ggg)dD12No>c$^-k9eHqJM@iUH*|1mu1fJ5aW
zt6c}I<@vmngB$DrL>(AkAkf)85&BZ`XSs=9oSVb?Zc8*qF-#AvL4Iw@|6{ue1|YtT
z<-m5cNUQubcT1i*IeD_PWpb!5*(qzS{Ct)&0tfSc*0s}MEMag9&6Sm@F4FAzlNpIs
zqv`Iu39Iwe|0^EWS6N_ee8Qe8^hYOf60P7;i%l366+0FLOPM_MazSBWkR98T_gymx
z8z2ghIdX|bg3x{MvpprhgrH(-0vYGcS}($Iok}&dY<}6NC_lp-OFWx!_e1og0sG4@
z2<TXbw-5^{^+Z%+!SPyzVW~ZR>RUFnH+3p1N;i|42n1LUh{%w731lX+7rhw`Z5x&j
z=?<VUg7bS1gY5VI2T^R3iR}nbIV;ja@tjgRF^sdE5F2k<D023Dbst)lc=$&=`1W*)
zPX2R=+&Am>0QS9jC<dVn8n_e0mJQ-ZA;LDu%V}RwGX$AI1Sf0YvWgdNWA1wYmdmw{
z8tjA<5!s~Cf#>(?8C~2Sl;tXt0g(;5BnZ`_$ZwNz6hnM~4zOBIM`D&>fs9tr4!7dK
zLogrh36vpv{cbP3bMsP6#Iey<teUy!tN#Wc;Uhj#aA;8}{Y@moH2mkphHo8I5EMid
zCG1pI026UV^buwrl7{X$QeWIdGwu1cRE~vy^R6zBYex+>6{1X;{H8b@E9d^@Q{S<<
z+w1pbU_JZpx<>}4s=Z?-K1{$X$G6rrPGrXSdOTS+mHZQ#Z_9g)EQJ5L%tU0|Kl3UF
zYi=On%fr2)P0SVba<pnZILLd2<K?6W8nH@JdM>u-P4FKkZ7hRE_=aJ$kZ6UX71?gI
zm-mixqy28SuPqtdU#xFt8tj|r!yP2|4E%HG_bu*0y+intWRbf8_g7nri{F=rtxV}J
zqwzjZeYLB+45*jY*W(5<Y&GBRM>Ai`<`Zf6Ut}pnPgm*qneES0lJX+(;AnRXTF<O-
zT8xXRUaD>gkkcyJUC(MW8}B<vpL3k6P@oa`CWL3E#7JgnYn%*kG#jZU-m6?YDd%p0
zl(}7f-t~S2a00LQ>O{jL3AGi++1BedG3TUcQiPF+CtcfVTUT-@1g}grbpeP4L1f2t
z3r@qfr|;*`<g1yYR@^YHK*f1czQ04+cG8cud8ECadvISOtCXv#3>nLliMY_K<zS7S
zQBF}QUqLotW2ifoJ06CwE_z6FP1hRg<_CYRhhLk-L1-ZvQX}MNzw-DUKFR{v47&n|
zV~P6^-@K86bfynrX)E|MqG<-@OyX1pVJH}^MPn?UX~e{@PX_XeqM`<d@*{y!!I(-(
zt+1V#Q3B;y3)=$sjsra33o%B0cUzn9;k&!J-m!M&h{r7MH>$;P>Q0+?n-r_mDgH+E
zyOHwpu0&r%;!yZGMM`}|JQyYlccgCg?HBsc25z=N%#XwBnenHNs`7Jhdzq8J9*8o*
zvwvs7=ULVNf=T4ayCvqYRmN_&xzQP9M1;i*++$7T<V4YNl)KQwyDx2u>xjRd`nY*~
z`&abHFYv|N-+1~t*S1VhOo-RlH)WS^`gaDgF0|vf>rD8vM!JDExaT_1$QYKHnq$}~
z`rQd6kOlItA4-nf553lutV8dU<AM?gkC$gKmaQ{=CLP)~eCaDKekcSybBqj|ydoYi
zU8hnciIg(ohEGOG+p<kLwnh1){EG>ZSRRzmW6m0K|Ep38cOvy{04WOlx_1E1Ez|x|
zr;G;x!Cp{)w0AXj5A%Ko9GRefH<{^ohC(leHKZf#AX@pzemFd=I5bJG1Dp??Q4sUP
zI{q$R713nidM)6+gvg|s@tWGFNM$uS`703}EEY*t#IJ^;D5A0xPjlF~a175u=(XF`
z`_8S62nU}xR#BfnbR}|lzAwZ`$YU?+$Vi5wzTnc*|G?FF@5pTxkDbTw8}of5v#=ZZ
zNXy}z>$gy~K)dH(q>s`!71n=~-9doW#jjAMq0l1e#ysH_GRDV9x`IlMHyJsDByVum
zjnbI+N7C#Gfb2Mrxr3mDA(f$bOc)C%(F;->LNe$JqO^$t3<!{8NsxuiUfOjO2XxNa
zlx}jWBZnfa!O2M(05}B6yTkBNx(VTDv-4`Ai`wFH&NtrAxdC*Ab|4rJhR2a(-EXLy
z0?rj&gAZ-%^``h-*HCxOVzJArmiFb=PjAM_^mFlgFP}mW8%Mri8XDxP8yyY720A$6
zJfBpb8raA=FE-TYml_)dsCg)w(omuqRr2Y1<>^sK70M}JS@4W=S?!!{B{g@?GC#QS
zM?G5N2npcsbT2mk*=EtnBbGB9mSV;F|9J3alcwZfD>1hI^0HtAeN@6uv%v9N7yP~|
z1^*aW#8l2!Ukx}7{jfOk^}A3=F4~$$%i_5;Ugq=FY3qe~-F7gZEcYgcb;7ajs}U?-
zDz+dpNACYZ1rQ_QDW@n6l5`~^J$CJ-i$>u3fF(zWoTH{skO~!KX7DX!B1aDrw2r8P
z^Fe+?&595Cf^WW?ijy7b4?BN*FB*=1gi@^rU#9DiqW~mJ(pf_ZJ{r+c@U}H!)FgDk
z814gbCTO(7cS~V-9aKbkOCmEJE)gO_t8)QiQAFVwWdqFO8>dw!<9PA~2<>~~P$~%g
zLhwu?u7sH+QGG1}kMuP@2fmUPL{3tOQVU3ZMF_ksdr-$FlDLnPO2DUUh$|AL!u-iZ
zV4;4;;jU8V+!7<92Z(LueeW<(k$DLlaka_)XQ1Xi*PzFS{s9L<U74Z=rBXQoxVkJM
zeV}Y9rHsEc3)Bi)5c*h3gYPg!<^|_q_(pOU;PV?0GD*{;ZSbs=gP-eT>OGz@B^_iT
zIIlB8fq_-Q0_XgeGn6HAt}P7ACk+lFV|vYW(9|)S!@wHi2E*3@qMXt9tc^zoWBOLr
zBlyVq87AMm6sIbtE>Ff?+wy&TXZts9m{7r9?_Haau+xG)ObsJ{H(dGX_j$oLd9~pe
zoJ;}vbiuc*(5bf9oGPY<r8*ZG{MA79TQumv#ClMZ`~Fa<BB3fbd#d@qI6PZL*9p@o
zx@!;OO#Kz2?gmERjp0UgExgg<KY>i?^k_L*!&Kd6n~>{<F;n{ST}9d=dZ>QW#T-o-
zmUo&p(kibGH@e&P3zFmV|5;vvFc@{%Fz~7BcUK}C7QYF3(^vcvqMMOU`hHg*11#8(
zCE4j3uKe+nYSgltxBypZsc%7ZpaqT|Cf=Y4^N$15`I--9f|}VZYu!TI^()=Iak#5l
z4LD4g5)+b8G>H_ICWmp7eXI&aiJ>15_%4sPzH#1=#0PDC3AtHw?tvyu2}zJH;~@3v
zM`M4{Y40luu<=@cBRKrx9sF1KeCtVDWX1$@3T5oWGo5xsnmv$@XXUpME=cP~?U*-T
zEuD~uT}GmWa%$lxvq`1oSRr3UAdvuU0A|QUc|t2bd!v>vXNVUQ5YgC%{q%GBWX-b!
zvH&2Tim$g!veKNCN)`(8EU`=H)chRkZ5D(r5#CVPZFX5It_8&P02*c!#-NiziA<HQ
zSjulYOALRrK=MoaR643Ts4z$H;u4$p{#;e$z$v(bvcXwf9&GUlcAYfE_GON#%uBoq
zoPnO~glsIKV<9-*nRCC2T$<Eu^iFs(#PK$*3ybVoW3f+qV^U>irpzCG8|_%F`$!A#
z^}x89yZ`3)Zk*Qtc^?J!8!<zwHAw*r6=e_7)SfY_zDO}T0qVEW2r?M(F^28PF)@E1
z1#Zh&IhoJ?!!9NDQMUGkb8&mu&rC%XFCyo=6aO0f+zr^IN7Xyr)UWV&jS2ccj1Wb3
zrPIy;IVZnvQmsyf5*3dJ^C%04w^1W4s5G-AF6@U){`IF8IH+BqM0(uX8jqTCjV=?C
z_M5a`{WrtYu~w<G^n{}e{i9RrI*9m3k#MG0@6a<8*1#7DCv%aKe@V6bL>%WIL|63=
z97UDsWP)uqW+%om@?~7pI{BP>!ZV156BmGk{Hyn8e3iZdjyprfb(ekOu99gt>!l=h
zP0?z-`G*nTQ4bd8ihBjF!Uxw6!&XIuE~-QFjkIsA#bU6-+n|QoDqS)&+Yw{Pj5cye
zy#>Y@qb{l@t(+$0f@g{a+JubxT!lRgjDwPa?xfCJBVh?$E~h=+jas{rbEF|lAopy2
z2#y27?=`8=>z}w-G%{fh@AoQ0=7{(;QxD(<7>l2ZAAaa$;P1BpP`4d2D6oQ4g(AiN
z_GmLP(-9s;mp~~}sba$jD@tieJrJg$cI&yYXgSJHhm_l?t}Tb|NqwVm#2dm@h;GLT
z8^$#K?yH0m&qR|gdzAETXhKYgHs2A*QExm@^>cx$PEQ%Ctl-Eq`V~Ufm7WN0v*@X?
zJy-xE0f_59eC~AsaE0=M&K4&fn9>?(vK{s_SqTtQQ5enLGF}_UwMkwh*{INd@80L`
zzTcC%Yy)w|2$jQfjJ_|qy_4TzufdeITh0@U$+O+%_L$!S#p7L)YI@e9keGdX+oAHO
zS|sB~?bGWghbJp0)+$!aP;ZjWhjQslXuaa4gRA3$K1{Z`+EZKS+ECLk^=Co&FAep^
zB;FfdX4~^*C2wkLcY@+NIrkoaqnhjf6068<%bP30>2<SNCJgTH54ZC2Y87Plj3uU>
z;E}OIh2urD?qki=8WRtk61f_0t1a?*dzcOnMuXL?eC&TU%Z)gV4E6oMQL>W8QD_C(
z94Fktp3hoS9~ZlIfI&83G;o+PQJSLCN46Pm&piR4Gzy=ay$+<>=(2O681LGlrX$?N
zwh7OmL25x4e0u{YL?LirzSbg^6<Mvu966CVpqt~OOWJQ{15IV5+G3X#9rY?S3X(xA
zMl5}wQ&{49_NiE<OTpvq3Hn^(xC{dP_&lB21o?|y%oc;Xj^h3Kes{wfJ^-e!+^8b=
z|GM1!lwGMVO6`vS25-JU?D0a%#<D6!3{GykSKB4{#?MKyj10Ub$C~tY;RBc)+6lt6
z3+JK^R{4nQ`l66-WLo6lTV>JH2XL^7{1vn#r{Yfy!~lxv?nE7!vd+Of_Xbl6VRnBB
zsq)i+C>gJJ$dUY1mY|z9?JvhNf{6L*c$gGXU=r<H=EpRrYIa7Q5W9aM@!*oh3y4<J
zkD1!f<n58!5b;;JUJ&<oS#?<TngI!ha435GBp`D6XUgwFuq(aJSLZ;4Y*i>@m$1o1
zlh)kLzzscY9+|$zMMDK8ZuypgOQ?wR=~nPGY<f6#xR-Q%pm}2NKXwiwb?C)F`u|rZ
zvmM=%6If*`8RkkFRpRb<CXbXd-_jgOI_8^QbA^o(df*2oGWLlc3Gxn!ww|DMn)0JI
zWBhB}+zU#iGU)H0{R<kAkFOto&19LXUeNv{^-q>_b#{J(FD(|JZLc6Og`kgBfdZjP
zOv>x<V^))dVJL|$JaOFLkh6cXF2g12_k1<2i)n*3+QbR#F718s^kaFX2te+DB&J$0
zY^f#6#_sGeJA+78@!0GtoJ@#RYLrBX<Gy*if@(<J6k1NGfS?C9MbxyErlAtLvG7bL
zxMe}oLG&#;8jfj|z7|ZNMc9)@laT)8iGNZ|O#q)A(*OU+dJC>LpsiapcyMTo6?cc=
z?!~<nDa9#n#oeLMLUGsP?(V_ein|qeCqUq)=X_V*cz+-nBiVbewdR@<q$bDS{;IHU
zNX_9?H4&Ks`7H2sY?6*cg|>-6K~n(IunhfL!2IXUgZPoPUiAlC(?=yB+Z$6V41(bw
zqX>Ti?@mvEEI1x6G%-c2Lhk^jh}AgjOnH7(Fi{>;k4+AoBz!xC3!W}5CLG8$kX%pW
z<|D8mltDyFQZ_`in?L8=vKG)zG0_QQkv>^JQBee6XoolCalpp}E9=4wkbs;pG(MQV
z5&im>CDl$pdp+~(2-<FFv#-3{zTYN(nVbh%z;HoaiUdfenc%oqqST<S%bf$X&GN1b
z`EvkFqu`pf;3(=Npux9r=Fe~t7v&qd(FlY1S8T5tK_fxzM_FS}Io0eiPQcGs!lJo)
zFwOnwE!6ZnAVjK2Ug^6^RP?|)t$=d6<I(}^|Lu3D4a0uF53W>J>vZME@1MAHt?uPF
z4ewF!GMqm5cdYY!%`~+j`4Z2E<{iDV$N4vUTj1tbm9S}AYDeN9LPqPsZ%8%E0q<-g
z5edzLq<YUPjQ5|?U802cZ%)L>{gcFN`|6V!|IXGg(7+agc55>U^zR~qJ}GNRTP{_f
zBlAl5h}5;&kse-EW-lP||4G7<tEG_T6Aupv9*gP`K2((QC|Hi|QMo-p=3gcb8|1ac
z2LpSKc>9v({R60hwJV5UK2WwitY4BM0)HznQr8>Dx1U9EP<e$9GyCh?A$<MRhv=w;
zLa>bLcvG%JgAaqZm)3Gi)<kPHKVL{g)x%i78TWocRT|nZ*7k)hlT3`ck&J+EFGp_3
zuvi1pt4ZamWu}g+#X|xDTE)=^lnB{$-UPg@<~LHtNs4I2k!eyQoEr}m2~81f*T-j|
zCdI5#ZoExZ(<Mv*X#^l?xpWdJl$FG21Ng#I-^!p@U=wl$Z~#a_G9J+>LP~E3vf2JB
zsX!69spgPo0gK-{jJg+n;6DR|t>J~_EVJu1wPhe_?-6`ajJm~98-nLC7z?gI62fX9
zSl;U^a?6kyedRjD-pOPAQv+tv5elFUptAb)79R_B3K1Za+;ei<xA6J@pf;&{SNp#W
z_>vI=uJrX2xvTae=<MxZ$$T!dLQ~90197|d*6U?<KXcd^eL&N^`+^(x=l!)cYmh#u
zDy%1^Fuabv<R2MeQ%_J(IN{;(?@2K&&!LPf^M6;x_MeQtCQ9nE>_}s`@{TS}HmgR|
zfLjHQEKaL!7_c6@C6)wv&NRbg9cTFOvb0uAS3Ac!{ohUI5HDnN(T}ex0H~{c4#=)e
z{r4;Ox}L<uxTnftK++VqX(pkY15B&DL0H*xjPBxbb5H*w?EMcXw%B;w#5b?V!TCoZ
z2T)spcmOng$kk~?M{KqTJ*IZ@rXOlz!ZY(pm&w53*2>TUnaSZ_Z#&ASbhT6waXw_j
zi3kE{9w!5-$rL|OX8iSSYNZ)Lh?&xa>cgSZ08&*N=#)MvAqor~`+89xnoPjsgG8nv
z6wd^$_;{hq@OttT>jssJ-waA3n^J>0Kx_eMH|iHI-S10&bQtwIqcQ|w!a3)znmU8H
z5yvCEv-VKbRHUO|Vu|{6oL+btV0lEs3mkGrxP(Ls$b}EU3S>5edn;*!+8)@DW8LSc
z-cedKZJOlqp2mW%NgoHWB~WAo50XO){UeA+(I@axdovo(I(f~g__zFa0Pa<>XM_Xp
zcl6)=oczKpJyE#gh~o(ebE){hrLxnGvdIZ^G8-SUEqfZa8-8;%z$wm(xJc^tnDZ==
z#k$*m{^TzYpbxP+E*3#tYB5+`#Ey)q(osrX%N->A5s1se`Fc}fk<EuKcJ&*YHPLlD
ztgAP_5BZvh%ErS4x^j4dJdHc}-*D{>TV$bETFOV=!F(1@@w@EmkY!$qCPl4Yj+@%M
zVem1bc7*)+uj8k^Zj{Pj>ei3t|7zH}f&nJ`>eOw^tM#Y3imxItmU+rM@-s!#zehe@
z<jkJ2NzF6=vu((Orv~1y`!D6Sve%#{HQK;NbSoiZciS_3!a;TNS%D;QkxipFzh#Cx
zY5M%S`mO2u?SK8g_|Iz~E8gSO_o><5=R#w>9cs;D_f`IL{*OIRFQJ<>ytzsw3d~pq
zu3G~?#Ukxs<g=s2#0s^*e;Y~xVf=84Ac@)-0)EOVaijj!kScv(#p<qNq2D)#!KM@f
zkg%Lnl>0mBT$gjaP;tpG`OKg^7V=gCtdUx2;ID#<&nhu%%zHJmz*MC0xS?GZp$`l{
zkQ9uHXui#olQFu)i~o|8SMgL0mhy8=C(-CTiY(p$Ac2Tj0{)mP)F0s`BAnokO4xP#
z0QwxaD$(FS8ifFc-r*hn2no(c)R7R)c)+VM0SKa@dGCJMfWrm6J$i$sipEmCRi{v!
z`6xjQpaMwo!U0NDLoTQU?xdlgj;RwVc}7$i49xMW#n(xp)F<iaPrHYyavIWuU*O#i
zU;`JMorA^rwBHiEQpg~D2Ez|&6IX4QOqMojqv@=H&zc_GJst(OEwmIrFLc-n%*qfX
ztt5yjZwCU3y$k@@IlX8Ak@wM~K}qrrd!f|5SXkev`iKJ5;!SCM*%(RQ>oSBe2MZt=
zA#7^A0W`vR-&vX4UDRKU$R@AVso&L4x^%gV=;!t&p^z2cwDxz!%DTA!Gm`6i*?10O
zeQuiUwEX9|&8;d79dBd)5=0G!iMQMO99GH$aTf+*V|;r`LI?T#g;81$+V;LAzxq-3
z&aZ4DE*pAQgV!DmEP+3E*gkv>gz?z{)UshHCXsqSENL8hBC7xQ{<9b?H%6Fx_7hn@
z{m?o?`S72j!B^W&%Ub)7EepqnZjtLO*LADo@sRL0yE>CPSbk$+!+riK7W3f!`3v(*
z=m3Td_0Rn&n2lhy-`LUWF}XMRc@?`;YD=2k4;!{D(ZcW2IIs4$S%e7j^q=$2!P|H0
z#`d~vw-SuT0q1f8q`8h&{_&tlMkUtmQMnJSG^cPgo<izsH$x>Y^Bmyfd(nQ3a$w;E
z#jgu&uLMwI_ATy`AQs2qqi)sf9m(ZIB03|E<9eQVOghD72eUF+Y-of9q?#C1nM49Y
z016oya_N%L#SSWoZyDF6fmGoG4x`<Mcv88vF<T_>rXgeNQS>BLZa`R74Nf3PF(M*E
z>kVxdV5F`N^cqf>_tqf{pasxa1rYpdDaFxUG4aAF9WgBf(x}U^F5&}tK=g8+fr45$
zdgsURx);W~PH5c)7dqip+WomhsGkCjgE|yKCO>YB0FX?yK?dK_(!k*^%LsX;@Z4^$
zm8GU2O+k?DCw2^c@b}Vy_uN|51S#L~ER#E&;9dDAVF8S8?~lYVu#)*@$!(69l_=iq
z62A3fI{0}|-ypE>5&a7&@qM!c$sfpCsAd9kERH}7|6s|!$lmB+jgk#OU3(W*;Nk$=
zG4RMgTVTMZG*7g~75G2K%y!nN`BpBs(1R=JrQlYlmd{>#Tl_q<!hZ>IGU@V9@wG8Z
zWb<;@-7{>@YzWGJ>{(dLDe(9e7>mj#rQk^6-qbXn$L_tT*WtAEj%F+|xmsIya7D89
zIM<bolsp}}g0k0gym0%{mwWS;8_z11T3zm*bG<T{fr*UvQy)Jzk`628>G}}8m!oKy
zi<8<@8Q~ozHQMPYnHN}P=sU*BCz?e!Xn%TdZDzKr&DH&oR4#cd{Wc`EZ6hNiD=oe0
z?pzdi8kde(iSJ!EPxXND{b*<LsCI+tP&V=!%8Rh^Ib&A^d4)18*%sHD{iGE`kie}Y
z_`i(o_ql5*{l#ZhHh9_0!L|otq7lavb<d#=`h3_OClg#18(A-+N{@c{zS;HcqiXhA
zcbloc3}HRHCl`?3;<n`HJ8LGEPm&RvRy#Z!!feDn3gaVKFQQeCPi9Cm=5`A(><i{%
zY$1a04|VYAm)fK6dR>7%!$=NiDJY;g^%K{OpAk;*52rQZ_hHL=b>4@6z>rg#dxJ%v
zFpi$MUV+1m1GZOwVu9`b-LLNGU0;Z5uyd8{JH41>Zwi15$1``1Tkok&B3=3`0kcMV
z^uym9;@%P?%NrPZDgPEc*3+Ry4UMF_iN~y?-RB<;ty=#0kY7o%29j<^G?n!9UYFD=
zLCus3>-*VMHeUH*wF3Fa$a210>l>vF{ARUdri`SFbD985l8gO=@B4;L`$hqukeKR^
zEgj`1B|E7;N?tS73uH9Psce;(!c$+Ez%#(8ixe@4TaG;+GIe>8whG>DEBZ=lhagg@
zC!HZ1=inH+Zy$j^IGI?0wIntW#&O@UptB1nK^bBwt1Q3jkFv?5VN~M;FCPP&<>Xfa
zKLD8m;;58}Xl*rP)mll@0R#lOWwaCwVLHNs6bkJQM_w^gj1uU`-{AG%jDRV&$iWyp
zGCt5ZJGg*2{Dat7eUV>yfz0p%a*HIV2~h#R5)4G13hXRFsnS-8;vajs6DDgv-()IS
zcAyw(yyFTP#RMTlAT~yQr@D#;aCEee%6N{FFe(5H#(q0VZz=Xu933Q1j(tX{pgfew
zm>+~cYjVxNZP|yk?|&GC|A3WlqLO)=rA_T7$apt2rzAxouc~Mj;snb`cnZYTzO7?l
z=OurcB^V#f_8B16J+Q%uj5K~}C;RRIZRAUH@P2mPyUGe7xV<1tddTT;ye@<sPx?&w
z95PRiky5CCP&&Q5=p<w(%S?N^YYcN0_;VfM#h@e<?#ce{tix^Yxz7KYi+W81K;3m!
z)<yU8)DpXrul=lHPj>w&D^}_1FG1H=(TgoHGRAX;)B)Exn2g2Pi2&B#?*3$R&!QNV
z`o~*WC!!t|x!OI>L<y|rzVxGqt&Og31<%xDkXET2A0p|CZ!8DD3mlc{hsqJwT5W%t
zKo>u&-Z+=X=gX@XS!PJ4UjTWttE%Rm%EP)^m<8qZop5vGx%aWn7%V0ObFZ6Hsl*dN
zold4uRvv>Idb%bT7`s{KcgZlh|5-Ghy%FC(tpH|p=DI{kT}Q&@8YYrlBmY?mNC>Aj
z^+gF6vCpw`=L0>Q93Te3IW%n&o&5t7{&%cqN<||W&^IHbEB0lOFP#HuD*ZX&?OgCn
zxS=x*CC*nxU7C=ko(+WGl3T1Jv4&tdc_T?KMksE@AS=b6V^w^$fD4UBC*A#yl(*Fy
zf=uxf$x6EWU*7P@M&#$!baP(P6Oswkn6{#ANb|FbV+90|a;VFZ4iqDTBx}nt;edDm
zJlY#2Af;=#2R{aIN@LD!i4&d<0K8DWOznL89)tAVn?@F(T*;5M>BOB4IzH6du=0Tn
z{EH9<p{@4i17{$(q%>OVT*`A9o&@lDiU%&>0)VY=f!eyFW-Ky~3`Q(r{e}?erbFp6
z7YZL77{qFv!k^<AfQ$0M_)9>(R@HfZ1dedP_rda-<dIqhmc{(WKONOySp#_ig03|l
zZivE~eONi!Anl9+AOX&Jk_jb|P2?go-YXI8_<D-68|CMyede=<r@USEe)nr;?Yh0K
zB&mgd0iDC$kNaU9mZT#A4llb!;&>5zE@4lJ_S<t<%L#3VxugQk_falc9ejczS?)ez
zjCh&gMD1$IKL6KEZ~|?v^vMA8BV4B!q}Gji_17O?1C@QThe`k3u+DP=^N>4jfzfty
zv@+{5SOApg{>nyWrkgy6EE#vQTd$P-bI}IW_j+uhi{sBX|Lrc1Y|n9KLk^IGB1=(~
z@AYSmxsXf9G2hsJ*Ub&My-u{I_^BVBy2&AZRGat9j@kTz`~2<yU_da=S9#S>ZU3o;
zRMk;O<}#x|Gq>4j^ozl}r~&x+0G(dlsBHmmSPYYnVWGe)f`fktWb527e(Ctf8djN2
zYC!B~y#?c)>_Ld}{e1QH^}!BjJzRQ0!L1XHhC9(C!F*V-*P7lu-SB5xK0~{aP&GLg
zy62Jd0emfo|Fka=ffPSO7U4`#hln<qnA_t45kZp3^Qf4i*M?Xxky&AO;%EZPlkox3
z!j#E>y-Ub5yu`y*U~1o>@E?*otO2-S?A>$jKEJ6au0T|IyT*$)yEi{CVhC1puKomK
zMr)S1DywYoH<I^Q0jH%OXu**Q2I(duE~R)|Z%E_gRFJlh0NA@^;Syp*@t8sd_epaU
zQ~M){cFx2inD3Lxe+nZBK#ikjVFfxWsK@ydve|eA`6v%{w8bLMQrw?<q9{*6lD{F+
zlGoTo-h>})y+VpHiu#IV2^N)D`TK!|;KO2fPR5Y1uM)6)2qX$e9l#zy)h;{MQKF{n
z(^!tTga~~Lz!b+#282wJ!NtIw=5d2Cd}A6XDA%zp6OZAD{to}7buOJUq3A|}qYA-A
z0Lh~km;CUB|BXl}B|0|XN&Kj23<ikH5e1S*#r9e(3RZv2U{k8JFszpmEDh<T>1UAi
zf#y>Qq5CK<H<&tFCvlw)TKTeWX6;`CWR5cPJv-^>K7H5=5xIg91^&id=)xZ2FgWVU
z6~2P~pUA`6&$r2|IW3*Aqm1MHmiPGFcjI2CFkFl|V93zts%ng&^SSBq^1?Fm>=~A+
zkOJvGe{C~wbHV+v^?d%a6=qA6?7n8W4^h)0_<~P+a%Tul7PB#8z*JBbftp==``>Iy
z6ir6&-ypmwzCQ2c`8$6rU&el2%1&4HzjNL<!-tr|_G)j3sNph_*-*lAbn=zA8)cjR
zrvul6l?M9_D<@b3ZgmtfTA7TLL`3+RUC}#l>65qYe+--nZ+&~|D!N`XqtowEah9~?
z5*+_<!cj{=>1bx%m`t^ZxKZ1h2ni<Yx9_A+lPSLR4e~DhS5u;}=nkWmb*nJov7q;&
z2Nr0ilN|6vQM*G-$?gy+Gvs_!A{^n;;dqfTFt2$23ajP{!D~otW?&}nZol5^5c1=<
zDc8LD`;I<8iN5EIgBT!fL;}78$T7ajRk(bDLyL2osAFuTMuck|EHSeI55<W7gl)vN
zD~Dp)=EKC1<In@8b?858ER%(lQt|vfku^lDz|uhvz^LYyt7bCd-aMEb8P3jE_NAlj
z%UWSfv1NDAKfP>I7rRX11QsDwQ+zUrr|?t&1mT4e2q6y0NS*Fpk$FtpXQxI^@O@=I
zVgL$X)PXGD!UU*6opH-1w>lsRsO&u~$AokgSpF?NNF*wn59`QCII5!CLdgCXW<XxW
z%e6voHrpzZ+0ZDhDLE$v5InT5Y1LM&Iy=3oxwFd!fzv2OQ=F&0Ndn`X;hrq0zsfQ0
z_RmKqivGF3&R46g85nBe+2*6xH8PmIj@KGlwT<MXCou#1P?Iq--E1Tcs-ZJ+9r<{C
zGU{t^{PvN9{7rBO=*yY+`Na4Y53-EU{AqN~aPKAb+&9{@*T!&Xv;f7X!rX7C$xl`n
z0~Shm@85i$WwW>Mzem1IF8_K-FM98Jog)V2dd+F+Ac)o2wpx!~Xn&u%AY~kPd1H7k
zf0Uw`>VKsnW=!UH0ry4Z{)<G9)61us$^yfe`!2fH5<b+M?CUOnSezH{>t@)x4K~tT
z?~m8h1^?dmdNCn=Tz~)X9fq?|6C!GO8MZ<2P_<(H?Ee4s%P>YtTYlY(FO2Z0=*!ig
znVydHz&ig^uy)OT^m#>;FRwI*LYZyB4<nLa(WB~sYPiGgax*}cea%B6O~|QS&(l^s
z$`X6$-BY)F`}Fga&<2CD6#M*BGb@V_i1W-NiQ~0;quF>r*-^<cd(Q@r&+zm>DEpUA
z*VnYzO)9d}L;laq#v7P06(nhR)b_3gk^(oP0uG&Gv~B3%ALm8E`rFuVpW2J+JjNPo
z62At%ts$OTKj1wwu?70X@R5S@sMi#T6%--m*UEn6{f%#eC3kPhWKbEGX1pnV2w37^
zM0{4$d<Ny8kEw_oNK-gGh|Ej1B1M8qxKXqW4l4-;-vd(hq9Nh}7!hj1Rnj-oecTdH
zrs*KnBQ6+pz+#m16_FR`2(|k%X=F-Y9OViQ9uB$t&(M$0aT;qy@Px~YNa@9@mW~-h
zM6MXfsELagT1-U~qoRx&k_-k)1dqnu^Z`h;grIhW6H7*6!qe{XExf6)nmM={#7-Uj
zACE?XREf;Ykx3{f@~_fn#yJ!?s-Ba6e<Tm#q{0U@P`qdoBuFydjN`dyi=tk-T`23{
zgXQuDk&RjYj-st@y-~w9ogAukEwiA~PckrM^(eghT2FPugdWhNHW&F5m1f%CERlKd
z4F`y-ZNN(S-x807i27EgM_8LVa-oL9sh806b2s`l!6#cSaJbEJ{>Qq_A#%SfMiE7i
zfeBt^^vH|iU828ly6pJu+y9EnTs@9xcP;xsxUw5veV>QvCyrPDe``KOhxM0vz6jQ<
z4g%|Klu(e2A#{L`odo;wBJcc3;`&j$>!_y$`{h!0%mOxbhMHbKZn}LPFB4f#9h5T8
z+DX;^A_mPn=d1F#QD0fL5PhB1kEB)kmmO)D<+oSHJNa8o?5<brtHFJ#c3NKiTW1#+
zTQc9JUUxI-?kn^aH26l}@H@=iboW1o=lTZ7T;+D;DVAjDV=L*g#b5tR2!}n#WBYFC
z|3}aHUsXstjMVwxHh0EgCBtO9>%MK>QX;mT9dVfE7FJbgw1}(5{)2K{>P9$M<stR#
zl^ydN@D=_8`~^Ph6jk1celwvrCI`L~Ck_r#L4@DsUzjchyc=em#WT(>FZ|A{Ph$)m
z`Z6b&7_6G*67?lo(7-uURsJ}SRvAY;E~;<Z(XLymW)}~>jCz1T#R+i{^+zbkTeduo
zw85FJvl4lt537c<6|piVZlsA8AcMD5t!-)jwH97>Rz-sXeXHbVo({1G&XYw>vTN4B
zk<Y^Q1CSgEh`=DZZ7Xd`_YQCDN9!V8c+J8zQtb6qqL6ws>ze>%svS3B!DtmA7#Z~!
zmZVVw>52?=4HuROio~^ap<|*2wxs-<kXE+|_ed@fI%HjP(rCa8PybvVRbQ#0X*_X}
zjz=?JNq}xdGdGz8Ko8kPZ7DE*c|GayOkvSXJ!>nibt)c9yYH(`brg@0-;+n&cgyM@
znAd7My}T2dgdVT^pLu3EtBCEd&(y8?wyi8_k*MMH6G`JBO5h->#|&fI-sw1NVkK(c
zLo0?fg_t-wxtvNl@J~j@!g>6dIi33NEAL@$r*J#pKLE>e#qmdWfaAr=gq-rO$O8)5
zG3YSw^LmF3QOiQ4IR0vsJG3$Ea>Vesne)9sgqcFaQ=tn)BOAHw{W%9-ap!T2aMobu
z8Z&k05LTk>4$E1sa*m#{zLL-A<pa>?;iZhvI#*}t<9F=VFO*hMC#5+8Z(K!ISxjjb
zn_7IPtz&$yqfn60yfHeBvQ&KR2?$Z|0yi3!;UY0GvC_n#n@<f};1$PB<j23~cj2s8
zJohzc#}JEbGXv+jL|$yZQ`u~fKdTqx4u^dF%OOs^g}#2k>8ya<mfsP9R}>fPXCbZn
zmPa(V^tZt&^_Sc~R*Ex`U6&mp5|R)F{D$CQ7-&6rrVK|+V}|f3-LASq-{0-hW7-YZ
z`n)(k&@|gqQtbHf^o~X2ye<==Lic0>g*C3YsT?UfJxuGkVu5b|`CsOVSFq>gmkrzf
z4a2ioS#mDJx!X_71v8A-$>dsutf4cl`!6VCi$m1{yZLX+YWdJ3LFcGTh>vV{)e)%U
zKib&V;0sD~fGEoy22|YYT%%jPDE)+}3xrEUTcp-l9(bxa8fi$ZZ5ex5Dk-sf14@-t
zK8wTlAtHKnZH52DTA3Mv%dVm0%BQQBskcUfb1<`bO<Ab!cwL>l{KS3WNQVb2?_~Mi
zgODTn&qs>IqeUl^&~}59m<9BA#+h?J>emBfgZg4Dc7MN@67~!(^jesYG?~4(y!Rkc
zunAT<V{#?N47SysxKQvY5~#?m>tKfBl+7mMjbA`d=z<H@-NRfT^D{O=?BJWf^{jL_
zd5t!hG)NoqfXNuPSDS3H1y-gp;cnDAkp$+-<I~kkaLTd3anhu+37HnpS0@?bkqKsb
z9+9{Li&ofT;Q_dV6aY}oGWBr1YYIC8b*Wd)8(4m5e&<UDk_;@i+b7@`MbmfX{@f39
zNLs%SB_s)gKjAO{zyR-V(@<c{^i#eL`@w|L$O631R_NK))z%cf86q8y<cPri^`CQq
zTo9z2|FQ3YA`vboAue(9!1Dqm#?Na6Wv&nVF0UH^gUul$wFUGpY<4)I+f2oxM@5z0
zH5s>c6c6@1QrrWSI`5Cr%P|+(4w96}(rP^IESWu{ewM}y)U&)D-@u8am{T+)jBPdc
zHM&XUD1__ccAt$W9v?}{;(*hT126#S+qWYC;UjN&mF^^AX`Q>IpO%)0stl3fOom_T
zG1-wfC6%AgRDPf5wwwMDVPtSZc~5<o8hgxB(ZV{~_34=Cj}6Pq3M0Dhhl(z*moNr5
zs~Zfr<`mAbn<}Qu0xHc;g7q-`*N@3Z0-c60aUXb0sgJyx^<m`KPKGYF_cVNKF9AKz
zo|~0I3qKYJ2}_LowdJ+{_Bc<L2tFmIcUt*fuMCz&g^}_b-`DKbqO(`GUHU^89V)=C
z)I}V85PhM>bb>6WT7k>vow3Ggb%92WdAQSMipW0?fAc-JOEyvsg>Pa~Q!<>2wA#ts
z$2j(Sk~O2d&6clW;m>DJXH8i5bt?Wp%xF(Ot(-|jf|Q@$nH(l}JvMlX_gn^W&}3}0
z3mq#t?<uNq2{ry&nj~RAH1T}_`#D!-T!4rGm_dA@`S?r@ey3%_WNB0OkOB8Dhe=ZZ
zlbSsz)OB+2S+;4}4tCLizP{P~%-}YEzb;H08S@jjD^&Uv#R>mZ0qcbcGiA)jZ<ani
z!Yr&Q+48b^>IaMqCF+4dU{bcr^QHOzon5Y@!W24S=;^fK7Tn|Um})VOfGuKscPeJ&
z-=`08GlYVN#Gaq%$J#coWSnOwGMP*JwF2II_IF|^Rwo*<#)>}e9*-C7XD@Jq{uBv_
z-R8d?AEwVL#xayl@79}`K&3!M70PR<d)=f&FuJ~QAe*C_|KhQP(J5w;ac6REvc$`h
zxb%Y>7UkqNbPbbP84r4dR)%_^t|b<4j2nrtcnq4tatS~4uI^Wdm2HX)i7_+{-I6<>
z_Z~UzZ!he@h-G!2>2?KgCe$~6g^i6+2l;NUDezYC&D$@x8MN8>R=&N@DaH;Ok8+h$
zDdy8F#ndFop1!nk?LeSbx9itEZRrRbtC0<JF!`aY%6G-n+W0>6Tqb*4kS_^o)yvgj
zGA&qtQMlfqLj4@3<ORs!$n6gRSHQG7D9&iEA0Ftz7^8^O^60iP^4O;U1e(F{fH}%R
zS){M(9U%3DDU3rTLUoa9Q4eh9d=w)EmSAZ~+e0zp@{(-~!h|do9Sk#D&zda)v_Bjr
zWBOh$i=yOj2bk&2(F2gEHh|H|G`v_0amfrVb1yM7<c(oSR(~&uVVomhjMr<BC=DTQ
z$ZsmMrLoc&oA@Ikqu`m#YWiE0%$Qs-fu!ZMff!WN`zS$P-W}<CD1rnb!u}z-ID+k4
zNo!nq^cyN#V`qA`i2;qQr|EZ$;dr%ngE6V^f{|iEsx|-SWhWrLlOzxAg!=@dKj^(-
zCZ`xE-*Xf9+}qut<iNsSjJ;@OE&A0of(`og(fgoy2<8m~|L{IWA5PgQd5Zl;I)#jo
z`k~H}fE0I`ZL|_Nju82-KRn0KvqlS$T3x}%wxpA7Ot}kWEL>39_=pj?xb)+d+Z4`S
z%w61_%ZzxJHG=jkWRa>Y@zHzK+<o5kH)JKUmF(;tXnT3j``63RFR@NT@P6zDqI>5Y
z>{g@J=qZSG`ri(c8@LU$v8=kh-Wk3s12y;Qz;olwDtpZA9gFas0oX5p0eG+2H2c@^
z^7b#?*cDvM+1dB(v7LP@OpkB6oqhbyAijHChR?O0r*l;xi|iZ}4rt7*0)K|X!@T=R
z^p4Q_4e~KfRbp}{yyWxdVAFA!lXd?Y0pR=V&ZMYrSur#R3CO?O_jc^O<H#B+S5Y6_
zwV8A2c->FjX+MBRg$6(yY;m4_x^B7m&hnnvW1oe<8*expczthq{5hJ`ji;TZJ%FEq
z00<y|y<u}+=(u0HetE9w#D=Ga-4rV4+udYM%`kvafVhsySqpMup~YC2&(+LcW?GXL
zA$4`pI$OZ*VqSEqH61J!(d!PfKYq)%W?N`sUOJ-Z3MqiEo+uMEI0pY}Ju~0m8+%&6
zY}EaHa!J1K7uIr}MO|}uWR8{bwMIyQgDmV1+M-#WFNSw>TGrcNx-NH{J%4|@0ch*X
zm?@!Ils`I=HFWg5g-`vy%bWJ2x*DP$OSRor6<uZ4?{^Lp)!BF$iD0syDH7u^dRu<X
z!T3m5fHJQFpgx&!V*$4I-_~j@petrDo;H%i_2#A4<GPpfyf-WZhp%)Gbd>p%o_w|E
zp4fz_tcK-^0+ClLZ#`WYEY>{^Q}+@H+Vi!N8lz?5WRTrXIf=%tn0eDugl&=Q^OD{@
zS5(YFWLmZA9?4Kr3&O6ExZ2Kt3!H_R+Gw_!4<0Z5I#*`t>@?S?Gk)Jl5kT@9*TyUB
zKy1JVHBcNF8e5G_)gYt~z`fy;t7iZ96J`Z;fP(Y6KK&9H*e@o{i$ehqkPr<3FsHHI
z1-p(NJ4=oU6Gtf?JHSMjBzKC5%D%n=p<aLJKBW&)TL>H1!)*KS^Wj2cm{JlxtMOYe
zv%umRlaHFtQ70Q<1_0XWU*Pa@_y)JK)fp-dFF)Ukl-V<4Bt1swku@S&eKdjxp!yp7
z_+=wZuKF2;T!o}L!S5J@fv&O7U?2vLkEl8fA_S(+2BBjSI=)a1;shRUp~%?2&hZ(*
z$sACNBW)t~+|<2yR%vtwhmX^1p`scgI?vE;A6}bG83j{-I60Ge-iDd8f+UG_C;`M#
z`a1!!L=B1a9=$L&g%gPp;vNu;!`shBGXdLQJgY>Qpqb(K^G?^9MpstI2db{1)W!0Q
z*!Ev$tD#@8;Ourbsn7pyc*f2NcDOQ9Zj_6^<&KBu<51oeA8)B3P`~*3zgl#0@~wJh
zr{X@W*#8WS#X-;V>ABoXYwBt^8oPeFPmg}xchwveas}j^G_}7pF#92vJpCy0w>W!&
z_<AjvvWcFrGGzjL6n&3c+1b%8+lY~a)?|+*_N9xmY%`%R<l0A}B90Lm!&Mze3Zlzu
zav>F^iQj~KSDR0Iu2})%2Q(>&-Cm57(f14XFO4n7FHcGnvgpXo{C@W}M+++G=J}e2
zAW9Yg-GXGDBpG@11YJ8dtcqlBzsms6`}7S<30!fs<zl3m=jT9_@g@g|diC(`WHqPg
z#k%_vZ!-1~rc(F{q91r(;LAx`@=w+s)sqaPc2KZjY!DcjGb{RtvY-Vg9WDA264)Qo
z0Yz<l>6d+(TAum^=|CHtLjy??vM*A-^m+I@-HWvHX7jEL$c&d&HtR)XMv6svdQEce
zU85n$82ASTI%h13v*S5`z2}g9QmycKh2?{M@xDU)^5cBa%A}iL-HVgUzLOl_UO$*y
zkk}{bZt`4BM7lPwaM?U2SR<ZdnmZqb+2^*8x11}Mjt&b<n`=<{-a^@iyW%XRvQp|I
z$WHH_O(xTZmn0qtiVTWft`WW-6Rq>Uuhb-3(B*!E7;1gn4M+W=rrU63U$~Rmj4kbp
zWBaw~lp0}AIXT;|Quto_S8b!OC>p31YKd)Y*)NkoSr0@t`a4}>yW$(g9L|dKg`!d1
zHgFNX2FV1y4<D!l&W*Rdw-yyJs&kX~`8j)P(2<choV%1|AdeLed-Lf)+XIfrlm4ca
zuhSjoKc@m%3t<mF$0{>hTc4hnL_^9`6`AjtPeoLrl$@qAsn^<ERW)VlbLgVu>v{H5
zdusMQo#!_S3(?2Gd+iz`pzE%r8f5gEY!m+Xe*R80dRDE<s=+%-!2fsw*enBBz$NXb
zE4<eV0xg3wvZYtK$ARm^tE=tfUFTp}1Qz2WUYpxV@yP3#v-h~a&t6?p<@?593}Vw`
zqn_BP-lwOgb#t=h9bXWi5t~+cIYEZcM^G6QLtLuM&t)*vDPOG8BaSGz6P+?W2o>kh
zH2b06i~@1;64G8b`A+TTCteXux9_@^JobWsa_R`q!-l?ikKffKO0I#vi9$lzD;ZXf
z!YDvJB)*#>=cVv(4v4qv_v)Y4vJIYtG%$OPSLdhtB#4Xfj6!TszN_MZC2?N9h&^mM
z#qVe5U0_!=J3;EauUo2o=HT(q>R$0K(D28Q(C%f#f)5SFZkL#TBhXEIZ`|RhC))g-
z7xH6D8WL$W2r5W_BH@y$kJlA?&nIX@IhLb=PPm1(rB^|EWbDf0uG!xL#ScF{Y?65v
zwS<}8Ro(hx>{Z=+a;=UY{hG`~U(qK>#GR4%ZhQIQt3L`k?DJ(f3N~W(K6y{f>Cm}t
z>%8=ty*RO8XHXk~Q2v&Auk9rwFg8lnP|x`Po|AtE2V5r}+A#l9;P0{xBdu4`*swro
z?R5d{k(mR?=r#FWflc>D#hMqC#>QHtVh#DYL01Lr(Ca_`ujjAkDBi2+g7*QJfi{&K
z>}VzDcS8Qwowv|)X<qv}B6h)ZpX<VfE_)<|olX=okYUT^0G=jvDNM}P&1SGe49@zw
zx9hnqq*XYr<dx*y@P0%)jhU^3+hg%a+~?Ujn@jce%zbp<?FY~0J~<?u%)WGq`egMr
z<VuVY+*2?etA4HRe>Gh?WeOm9y*GSSy|tokJs8P!x*Fc=T`)Knt@MCSVB2$ub?(rQ
z4t_-Qy|f^QR4)<1gJTY##2#ZITXs#X`{x>CQ_+40*my4`X|dndv@7|rUyinR+h6$T
zvpsSlF^_*OGKAaQt00SAuWe%Ko=@rPUnhht3jEM9GP-6+l$2D;hepGYJkOren%_dI
zy1QOhx-|I98J&lb2-5gE&l)sc{cnjkK=<n?c?WV{lPl}?6CdTO8K=Ep7sOuTJJ_CG
z$%d`QUZ2~#np#<3OkeK(A9Ju%cihhJ$T`VhPb|iFR7G2SB5n5AUmv2i7k|Y!t^2h>
z&ak5a3?#uv=rPjp9-jMOHkf2)6i+Q)v%lCwYkF$dj$ja-^PJ}1$S8Z$pncm8*7Yg*
zvr#|9O_T-5ef(B)B|9yX&wuH`*PTVpL8WI8N3bEU_{sZI3|(CKwYhpkL;~K;jQhJY
zH`<m~*zFF1uC_chc%t)rf2_??J|1`I(6UL)V>O>!bAI1oSoO<%@&{s$TVC?j260|;
z>BQ$z{5gSiv5yXC#df<dxzx8$qtjw>R1(auai1wZ1!2sU@vkes;l+IF<mD<fzal#{
zK6THmta4K<bN_DR9DAEPukiA;espVOr+c*#rVH&<CKa?T9Q%E^<n~b`nBRO5#!Upo
zMojRxlsIJ$tpW*&^wKzc|6tHS?A|rljeKKu%-$7(whQoDPxBl8e@{E?nWRBK=&nW0
zHmO$7vZ^S6dA3+(yZowOFU$XT<`HG|HqN{bP_O&XU&lzYbW!)3uZ^n!)GxQ+bh5k#
zpC|EjQssasSdIAOg<3+g<Q+37o+ljMkFSqnua*$qgH$~x7Ja{Y$X#E=<gEcco|xAn
zfmmx|rj@_Pxc3;xn`f*-@(s`A7HDOgHc8a`MkDE*_vAn#*OGcpAV<O|qd|__>;~*L
z8IML9dYFODhJ{Rbh)UsI@w~DUlB>pkDa>r0L)PCOj<-aVg4h{W)j{tx2!#TlyW$u5
zSptk+Z0aLkzn<=47y3UoMW+|_v&btvFcUJT#K`$yoVg=M$l;WIeKFV#k~}2Pcj@%p
zrJcig`(fjCOH7olBQddN*C;jQ&}r3gVY1ZrNH8#u>{smdq`(RcLR%K_IQc<8e!J1U
zWL8sUTuv20A(?i%*5n{<@ETT~GW>FV=1XXbSP}gXX)6OT_k7mBZ>VCQM13z1GJQ1p
zbqw0&Gqm~){bJj2OGvFsE@&x3^L}47Nt&{HpC*v3s`Co1Ml&i_?ej<8Lyp(S!%6v-
zT_Aw__wg;e=oKZi%U;zE<ze<J1L-?|DqaD{atp*1Ou3bud>}n}$6m)CVZ)9pD6q5G
zQKZ2lTBKbH`;>gvJ^m6`@VLyx0jbvoGRpVl=j61CfgN4Ru;(Y()2YGCIS&yh4f$z2
zb~>w3^rN*XMV1*~hi_p(z*FbGS-t-RGYVB`k}*Tp?OWe+s^i<f(~C}oHyk9>4Q-sZ
z8|#;h@p`&3<_RXRQC^P-I>tJ#stRGnRUn6vRKMc`qA8yJas3C3*1Vg*W3TEz_}Oa$
zM|9KdA9Y?&e}J=%AmBqh@{Y6)UQc$UxAE_uHP|7e57&;X4X0Pl*JsU7B)f^3%S+u`
zyN+#wOIv5J<9=tqO~&lWYb$0{F8DGVb4i?U<GT#dqWZ<oymnhXC%P&v%`j)|H>xG9
zN^C#Axy2~7(%;~pOXadKwLMsGdnQTOYPX}O5MrfoZ!$t8xp%6moSaNn+PY}oR<p3h
zIFt5u$LVu<Px)+aF7?4O_*&-Ix`U#mAQf#H_4ZYB=Vgn&_v7)E%ydha-U1D`n%+im
zLs-Lk6w-C;<Sb7UgU}rr+>1p-YnRpOr=$+q0w_KY(n{&vWbti<x#%M-(aiq)cdeHR
zz9Xp%zqhVo8D;m)`lfpo>|%o%?#n;Urb@eb;bjIsdhUe63|iZ-gcRu4p1+T`eHm3g
z;PbRHUuLnw6kr)}8zcODd-#hu^8amk8bgrkIc-n~9nstHZSgRQZcpmG$CManzU9H}
zzNqaz?nSQ>AU*eIyeRaguF?8r_wN%q-bh>26z?Cj^ERIQzXX(ZQdbpqUWD!~+^?A)
zKMo_@>l#ir-HDn>2F9w?KkSn~1`rWd;)}F>7Fs;-Iw#1q%(W+V=KRY+ZOZF|OBojg
zKd<shNN0FMj&MD66)md_DRTBrIUUQJt=OO+Q_^BBPZr(1yfj)m4>iY~k384Jhejcx
ze@c-eg&D_z^^8G>p+y<C>$I$~MqUcbZ(bRz=~95C_;ib>+<ZY0Xf~W2G72)2AU!Sf
z#hI8tM4BKa+0d!xHOBX~h27903*Teb9@fYfU*B&)c%8Q$nWb{Ow;_x4q6Noa=XsxU
zHB>pyJ$8ssM6O$_)-0zEDCQ7k==xc#5`aZKzt!xk-S5y(sOr-ctpv1QMrFl576if<
z-xLr#I4xOs+7pvDEVBg&yo4M)79sKOMj)Z*;&$3@1O_7u?cajdhq)QUhQGyQ#j7-u
zFKBMShZUT2;Wvv0(f0J@<FJ*n$lCC>TrKCa<VGpBRxa(LeLH^5K`&!oG<X(hMQSxx
z?b=!U5&S3-TD^`$+Gp&UOM*c04WN_o?3lHT4=6uaunJKV!c{19Q@kXy2yDP8*nO<V
zulm;@Bgv=;nzwl#C+?@!8CE-uH{$QuBSzXP;xU24t#o}de~}|qP&%V|E5hcYhLo-5
z@&DZ#|7&@SuD=+a!a}ustTGSKi||}U+LOTHZDf9zi)7c?8=a#k3j>R~j7*@9Wj+`D
zlO*2vXV_%+$Idu}j)^>&kssD7UJ1xwev{;%=;){VUF@uk>uVwU?mFIEQPAUV&~2M5
zDTh&O{~}>)2#9^XtMO0DeGEEPkr2w65s{TNo`aXc!FZi>Aaj+{@$1kzi7aC;F}GN~
zSvyZhDX`f7lR*|<yNv3e<3eR{d&RuEwYzX^-#q;+6ZCf+K<$?lTm~E*J{-)peAyJ4
z!-zNhnRsyvU=DO@bF<|mZ@KfW{JFb;LKa6hdf(ag+>*$bj7r$Ti9uMBS2C?6)DVEX
z$kKIRT2@y2IDzfJbFUDRb+fn7F;E)?z!SZ@RAr1za#yZLk8GCdC5f;)t$k%Dhvp>(
z#`@;AWV-Jujy0G_L3&fB_h8edq7<!zis+U{nZxxVyYCVSnOKjOBh>mnF*Dy6A-K?R
z!MQ@ee*&^>|0N;XOYG>wm&&N>l$*0v8Ba5N0Ox9ylYdulhQTejOrd#^vZT>yP-Ar+
z0xV3&ikecKmi|s&XXN-L?Q+eq_v(Yl!2J)>8{?fo{T>y4NvxXd0Mh6BafzO%ki?ud
z28xf>{tcB7qfRZcuyN%&SEQ!{6q(hhj<BM!N{$+6(!F4nZ9jY8I1%4ITXJ39lio3&
z`6$Oqn07Ul^)4nEb`-V8I<yiwy6LG!z3p_K{Mq1J^v>gWIHpk12IcF0Qs)H`J}S&C
zIy}$<bvO?aVG_3ayfkw3Y&D0BzMakST}Du;(obwiw@4^fMY@HKl$qmW=m;XHmg2iY
zAyvxhNX^SvQ{WZ?gq0=~y4Ut>-+tzN(Xj66r-~gtxqQxy(aOtPgOpwYY56arYLV5t
z^4xvQ&LJd&VsmHFpu!C=gZ#8No+3HfW1DFmr`gMwa;xFbYhl5GETd(kwd3}0;`=`%
zV1-#=1A>fBQMTZ>K7Uf!(IyMPt&Dwryi6b4nVdKO2K%%lk(m(vU@mx>EW;|WoK9G8
zl<&iR6^HEUs=6x^p#pmH+&c6v8r<LMtJf1La!?+nQhe!8CX|=KjaQ9vHE~pdH=N6%
zVw}P>9G8B*Q9d8KAnx--?5Hr0<Q0GDJrl0|f*Y&R-|K~q-`{j)Z>05JA`ILrJ$FY%
zSW(;~wcTEO;emy^SWm}Ye(|1iU_~zQQ0D!rjZrP)TljO>5l-7QcE0z$|B`dJiK-)3
z?H|O*@abH#36Xr%ajG}2+Bs%?86roE7yZTg+mSpyQ3yDAn4LGunUc_nC8cYAUl%N=
z0)23XbJs(8K~zMV$MAg`RUoX`<QIaA#2NSra*zT5(NY6^fpE#O&P2${IE0ZSgpQ^y
z&pp@UqwE#PfBCVwz8{~Xa{O$K^{6I~B%)pe*=ZkPIEQ~1=y(rD-O=jS3;15$+R7QD
ztiy<9cjn9And#iHau<pADDt1V-rco$%)F=@j$=G2;#&Jlz&C5CuH9f{to4e^u^d+<
z?w*AK<)j<tE|n`iP*jGgULjTRKMS>k?=EK%DxihVGpahfMbl5sJ7th%AR(%0<cFnp
zxEL*69d6W6^HDQctf2LT)vDJ2+u2snu~fT(<!arhGg3sTtjS=1{5*Kb|LHH^c%HYl
zB55;oF)nEPO4@ff)pU(hCKoH{LNUY`tuTB2$=P|-P8P8{xMtg!>twNUTyQGb44OUN
zM}aXV_A*FcO0sWCO}&@FH5s!5{m|<Th$}9^Gy2PA`1o@PwvW+No`sj`^ypm}cgXfB
zW%8#Fj9u~bYHB$xa@TB7M<)#IM@Xkn$%Us*xaP72kWcs>9_~pOW#mFgxa0%j)Z@ts
zFVDaK(p}$$70quiA`ZTF9in;$2Gh$A`>K8ewO`jEbUl{0p8|C%;Hd#0s(B^^p|syG
zECk$;;E81A#^I?8B-izxvfQ<QMHiAi&r$Ccp8K#%mQ|R+G^g_?jadqz0v$(i)EQ-X
zt@P!UUZx^6S~I+Xp^j)x;&pczRRwp`gvG^n>9>B!F<g0e`{A>5kD`_L9ObUBS>7)(
z^_Pv(XEI!1xxWg0;sgJKosK(pAJV7ZckD@W2@+x&p(Q!&J2QjjxK521Tta<e*7s>O
zdKrBGFCGKCopbOuD3LexBj)a}&rY&=m0=5Lg#Q@Hqrz9S>$@+!UUg;xvC=g{)}iyj
z_lK_9T|Mn5<y^sGc0-Jks&k4UAp^HP#c^h~E^clXaG+G!SaR3PFQ#!clBNC9RZXTr
zHXnb<5HD-%<r5)~sj-QPw6{phzZ-CrU{+{<s9=?&k|to5*Z`^^rNiwH?;+l6+O-#e
zV=2_+P)89chJCILpoV{^i-%H$0+YAK&5WfRJRKR$?4dNRRr=ZSWyq%QWs6pVbe*_c
zX3;=Bx6g$O?IXX>dw7mi1H8j%(!4J&j2r5HpN}b)U2&sFZWfE~{*5o-yR!G!xPu1u
zq=3IIU+eM2XeYLWBFt#KHHi!fQ?brzZI&Zo6}L>5^1e%7qbD(ya&Io|$uRf@`yoOD
zoy(E6v8#+-A(K_f_b9je8*ExzUaD)=`yusVGBgj0Ugne6=J$@Wa~Q!*>!)>zPqHbc
zv}+rt19J6c6b#5@6|*CHScHtnFS7izDzh!Mp>t2~LYFqZdVMWx#8#3-)n-v(a<7Zq
zI4)J5bKf~6a}T52vlZBPZ{?G@IOpv#)adHAXnW_>M}cI5OxMw}iI>Lxv%X<KAV^%C
zF^!a*O+miAGlWZo0$J?bvmG{WW98ZrXv3Lt<P;TCC1t5;n7Dt2J7G3Y-Ci~FfrF}{
zizl&mk2fs335FpFN;gbF&4e~ytjV{f*JdU&OqYH~-~^(@vrrlK#1XpmL#~3WxWgF%
z_T-w#CZxGo6?$9X@Nh2|U4^%Ec0KBT;k9mh!g$aPP`)Ds@@a96W<NfSoRl2~i>FL`
z-rCf_l=TL>SKRz>|FE6ZK+iO{3}(`gW&{V1+uR1yKC$H)-=;wvtJ)Xjs^ssAl&lPL
z*uox_jSG#E@JE@_xg!$49cHNB&?*U*i2fXxH~1(9xlL{O#FWl!cB?;LtT+5swwA7d
zv(C9x+btzslV4+rwZM_Uh5bll<52^3Y5w>r_hiV_-vBaGPoIv5hL*ja{prD&9s6!@
zy<Q?s+u%r^M%X{lLi^WePX<*gAEs>mw1tP-V4_2Dy0P|0MK->$zm9enSuY9rNlF}s
z^^HRqW*lWcI1!H%H1DVqY`SfVChzQ~MXAlX&vG9ld6_T<Z6AkR#r+qDMqEEY^t#-m
zIO}Tp*3NKk-1zw{fn6-YsRb>c%S!sqvx2C8aHG{IZ0ywEvZzhIPDSzHCwYy>*Dbk~
zTDF|%)LJ$cbUa}ue$dCR;sss7sHW~?Qqf{7H~dW)pg)mj6a1(Bd#o_)lERLsyL45b
zm)6a#sY!AuxFKh=-`)Sl)Oiv*2Y&`M?}B1hRsGQxw|x8}*4!`Lb`RNgv$l$aX%=gL
zoTQrv#%BJZfq>gKtO)X7(h3_kA$P1|wUm8&mgNlZsn>u7(p@+VvM6Dpbz-M0{$LDD
z#=gXfu)|8&-J0^RY?v{4fcNLRgHDn^6tvt<2;mt}JBT%vacsYrlbb!HKKJihVk=b9
zk)(OiFnowh#T8#XggKFAHd?wLqjTO_vYu>}uJ?pEH@Fpt&CRXh+MO+%dfT5t>?~aP
zzv^r!3(<5MsJ_k1^s_Il#Ud<meO*y9y^Riin)*~TNYY`q=-<p}1Opp4e<DkYwm)YO
zq`7xo`EaG2iq2Hyg5KaOFz1i*U>j?TndX}hDEs72JLu(IpQf3!lCGTI`Q{~JfAM!S
z!LaWTq(-S+xtadqwNep0j?Uok$~^bah7U=g&HEiwTFk;un7NXDA41{}%YZ!M{RVQ9
z?rE58RXxsBh`LC$#4zft9gvvnc!+{V1NK#Uu=rw7#lPk_G^zTA4%wkO>d*s+Q2rN&
zRmFB7F_A_bSJwi|umqOJ^h`Pqq53;DH8`0^rpZJ84WuD-T>g($^Fj%S8XiYiKHX$R
z**}y<@;T+>v+*cQ;3%rd0i^vcL5Fkvmf70kcwlTWKKRX}-bZ~RI6lasCmeeMV_N=^
zA{u{+sjVysHI2*H+95~nO~8!J;~#nS+P~kQ#@p=-1A>=XQE}walyqW5;6wqGiw~b!
z5?CpC7y=gHxEslK8B412XT@b=`4Bmk#!x@Kv7`bO(Ol38CDh!p7h}C~-AA6m=VmPJ
z5ZyJijVv+*$SIg2^`#`Sr4ma;+kzK=)_KOyo*(31nF@h%(cOTc!x?XJB5+2Fm3Sd7
z7bl&#w6m~bs_an~uJ}y0hR3%`)IbZb&&ZZ>COrF(D2No}sR_Kf<I1Wq2S-pQuTaW^
zsx91S@Y-bWU&LjeDUCV-4c%?F#@54Pmvsy}5kN{3tb%Ey6GGD2eHMf-7P(l73-FBc
zv#@Bo?1)7Z=BqPC{a<phkM7NOk-`SxHl=br5E;wNAnqNIUvA017s9qMdXv9Nzwt8U
z;=HR#fiL7vl#w2WuP{rSz!Z(45G8s^D|Kok!rv)ka3QM?73V2o1@YUeHmH$P(eCYs
z_hS*NB6AxBc;{VCGkoI?eoqfSO<~P|?D?{lD8rtN8U}Q!!+X4_>j97TIjX!;5Z3$g
zo@Kr#cw_~TyF;J-oafAvKQ+A0W{K>zR{5@3N0k#ArIVW&<trmp25%c<Yv1D2%{b3E
zDTJc$!1g&<-<yrws+Ue%G&4SSuR(y8O>K2}D_eW8knghl+w~E!oPD*3x#cWMcX6HP
zt!GiHyWTCQfPwJcmszLrzcAq4yj(MrZv@)$pSv(uF(4~w6=+T+>r%UBeq_=iShDsy
ztk*?^XS(sf=z7brs2VTaduCu@=<W^)35S$sL{e!G355Zq8>DN%A(U1?x~01%2T)L2
zq>+;D1}Qn)=YP)o?ft&z16+GwYpr|z)_s3+1`AdGZ3wEk08Y>;+{Z{_h>W#4Q>oGw
zhJ(gN4Toz%$L{r5C?B?RnxWlS3lQ99H&?F@x)2@w<1&fGSW5v=Z1crx`UfmClB~v?
zP<Ew68hi*LcEP5hE@2$GrjO+H(``LvbF1}}mRgWZ;<q{;Za=L!1g99W>ezt4<g))y
zI{&N{@gFpjcYFEu3s|A2o02J(yfL5PD^4=%uvdmD19P<*bxx6B3%yU3av|WbBef6O
z45zXpJjhqdm`DT>V3|uuVJIdFn_kjYP)`&-<2BK}q)FlyFo21_q(pWLrQCx~OqGrP
zMhT&q+aWr|9y+&z8&3<hI?V`@{I19Epfyt6x?U0v>xH7?XIP|RtG&Z~{Apo%C>Ah%
z6od!adRnHg*0;&fh*R$VRx0DRrtT^8=o{uT6xms7AwOJ~cc57!Yc<<3nY#EHsrED7
z+t^WIPOs7E_3Dg~sQHJl=s2RJCi5MHKQh2U97;2s*3lAWeN{mO<0tZeP8yE?BG{a`
zp8R-x+cc(JO_=^sYq;W93L`)oIsskp!$mvi&ld^EkLiDZ;lp*S$b#m?W#HYQ#p}i<
z^gD;(v-Ee*&NnU39FZgKw=tbIn*NU(R<8aDFU-N)TU2Wzsr8p-@>kVV!{^yP02twp
z2%gD1{<~tUih<)5rU?CO*5Jyn{dZfqi!R|2GW`{0cBb;-=z{w=7&yB3LFXpqqtoRC
z%M*u7B!SCz*jQf6$F|L|<RLNx&k8w!2z(LncVKYhqnW|Pmy0icS4OC_#0lbecN$Z0
zX|EezdmaJmQb226TfgxbCQoW`tZZFYJ~_>@0d7o99R*vfXNOVdO{&k=VWKN84GkB1
z<Yp|TkyQujX)*kwbH)yCCC-MdK=qjH##T`=mTt8Fk<+uit(-m8qPdGmlaz6Yospbv
z*mtDz_ZmW_H~-djh}K?2DHwh!X@mkFSECbs=39UqEY4cGoO<EY;ycG@YJIily8U`v
z9NvA!yF+1i@cQndLSRhLO84;8oLw_n?~AgypT^Hmzt7N87uFl3{%LON@i8!+DcA?U
zv0(T{<w!m|M+`sS&1zrg2_wkVoBlqu9tqgFHt&Rp1sIJa6QGD|hkmWL(anojGipW!
zfY=Mr$BC;iHcmn_w*P!LCvX+eWYnkS`~(i+EkN_-PvJ>!ix?er0CDdMrb*UCt1>ij
z)(!7C5WXSd04791=WU$%W4k2EAq-fsTFFn$yZRBV;;Rkn{pwP@QKD9E1mLzx3=6h0
zgVswRK^Ut?j&DCbk>E?=350y_r;6`_h(Oo?KUG6aF(|{(`pq8{SXwoNL>&9$aaWKU
z0ii2DA&B}T@}1$RBCc(E09=!2<=#yEp&mr88j6HsU|hRNci4wtL)hza`@5BC2)OLZ
znJr&U*6I~g^a6Jl)$10fD*|@v(?}$zjS(9fi!bU?XkfL9ax~?-+`Y!{^|bQD`e0J`
zTRuD)h<=F12h#j|`Il+rtuO<;+?IExB!vUz{c%ng;$OOGy9jSBIKXKMAhjpk>A1<h
z=8egejozy1o73tawe^V=ijMr77ElQr+=W9r9LwpqGSU@yoee=0`Hq@~F8O6^!JCTm
z8v}SJgE!k@mY+tXb<tIm-cuj*DkEB<(E)#3y;Bbs&Z?_a_S&6jXph8&B4Y~U55yfU
zgPc7;9FZDkC+AHDMi<`X_wRmvteTQd&GmOSN?<HI7Q>1rBqp35Rqe{W-2G#-5e)AV
zl(gR%UNRrBeq_SV|I+P9Y5iT)gK{aVs49OhwQZH7;JZzhEtQ|Q-~9SWwA53A6N-J|
zxkn+%D)0(wcqs9ZS->iZMsHL8x3jfnv%B73tjSN|fJ%BE%~EP?+qn)LCEDPY)l9vL
zkEwO&BOdT`CwTmv=RQ2NaKxdiWsLfISU_G>KKPLBI_m(Nu!IH|(QB;EY*PGHekFn=
z^+(O-?*;*J3YwifXg71Ey@0xF7FR>FnFA8Nn4ku(a$pkA%~aZ0c>Lj6u3hL)M)8MT
zfVEqN9azEaQ+<LEwV6xs-wWm}VU@UwUrS^Lz@01F=-zA%0N?$vH%W~W8{c{0TjJZK
z;OHsr$v~S*34ad+^M!?$3@N`0#*+3`av5$htkNV?R&71FjZPSnoaM+b{a0KFAj)mp
z#<O35ygG6xqOh>gK!khxD(sD7+N3f|+quNDx%n!Nt~EX8@j)H_lp=ePXKh)%mHpD%
z^X}a2_?~uA-fFBs(n~1JWUc-YzoE7rcYA@hF|<LD^p%w^tg>H-Aa*nyld#Yq9y7QZ
zPjsiBl}$D97^6al6NbPE;flSDXQX7n!NJM$zv6;p!@IYmF=^Vjt46?jJ+k!pSDP9S
z|5RSR%NT?fe@q1|vDBtfibX`;ZQ*-W1d+z3q~se+<ftrXCQt?CfP65+Af8Y$g<AFa
zcU^cNo>6-ALTDk(V-_c?6O~qOFk}rTZu;J4I>;9I%6Q7Dt-d|Z1hF!5V^i<O<aL{^
z6;Z7lmVfLXZ!?q{GavIs#YAIL15C{W+DeRZ5*Se{(?ZpckPX@o&TR0A!c7TLA9mrj
z0^Ai;FIA$=&WN_sQgl-)g3jbiK5IPncG4V)w7R@f4YUY<1;iWnE7Q@!6;oU>-$HMM
zj+7Z43O#SR`#tQrahJr+xUf0BI1;G?1D)k8tp}>}*tse<`U3?m-mW<YtZ6dEQ|q!S
zy(ICbr@b5_*%w-DIr%ur!m_Dr!ZHU{4vjgPYviTnkjcOdB3?_SkRtDr6!?}${noc^
zhTOP!iLHfzxgQpI_bKBSfHs|S{M{wFLv$NW2)zXqq<*`|ep+*zpSD~Mf4Af0UqD6C
zV2<jia(@=Q!$S1{F$O;Nj~QBzrs=d@%Q-8dE-PV%hkB{N+mGK33@L47GRQb~9D9as
z`2|>|RQ|(peODYzpCvuShr5p*^`&}uo(;XjN`sTg-Bd$8$&I61u>YKNIV?Ce##H_q
z(o}67;VGE<`v&*rWTUAivOwUZBB~o`y8wk+YVh-WX1{Lg26Im(x>t1|tWFO9Y<b66
z$x`~nId$nplZK6(-wJIzqWxh_At&m#4;N6(f19M>W4rD7nY_5PGw+USOxj5<otVxZ
zjhYqCT<it}^;p@YHAOuddpqug#BOy~j7iY6=NA1BO8p|!=jrH3Ac0iO3a)#s4xU79
z-gguaS9ML^{^xi|VP?~=-ak;H(OgsyijtJPih>ibVptf@Q19s&O1~Qxf}0_{pKR1$
za}{@id0@_S@jA1dN`Fe&m-Dfs`M;BOXm>BmBL^Uwp!<UZI*WXLe{*PIA1pXdxxXR-
zOigJ`qu`4HDW;3>l7_$+ObSuI&tO+U>Yt6Emi6HLv?COJ7?0b=gZ=>3IW#Jb8;E^+
znM>Uu{^V6XU)%4Th+8Bb)Ar6ymm5PC0R}Dn)Ph`x8F?>Z@A<XTKynTbCiW+c_xY(J
zV<e~;k%?kyb7v$<mW-Kz;29!Jbuct>Pz9Fz2<@-m@j;LeI5}?)^#^gwxnDn~wz53t
zMt%Z?g7!fzo7oykQpk_Uh(Y8egk!T|R8XDjMiBjxJ0KM*i%sjxhRK)=@qP)z`%r=W
z$un-Gw1VzP7}YVCX$<Twxq1&_eQL2n2!x4yu;!XaNeNO+qXmcBc4xWQi7|J(H_=9<
z48>Zln9Fv{p|q*qS#z+lo3RN&`TCV)zbb&P@B+R)YIZUDf>S*6_L3vTic?M3ic>F@
zbf$VT0pal~KRey!@9HB4g&$|RAF9kT-<B1+qQe{&<HiI#ozfw+=GAE2s~(31OXcwA
z`yG7SxTjCeI=_csl~^TOqsR>L|M9J4O0M@zz)lHTL4Tw8xbA+RP%dn~+dfo2T>(g<
zM1Vub-y7Sh3<@)U9t00O>P;VX$@UTUnSOCv>j9in2l-C8or*J)bFifh`)<*#UzmD&
z9#&tjdw1IG2QGDK*Hcnsq$vS$dcWf!=;YSK#YAqm_)9gpoco^GK;L{iz`6rZUj!}5
z33g6Gl)F<TP`bgVe=kz^Ecui=PiB3OMzU|}Eh7DD<VtSFKWK^`1P!Fu6PGv|89M(+
zdGI|81>P%kSiJP|=yqdXI=>6N@>pM5?s@hMG7Da~yw(~0c)=q7S50p;lqY3@8)sb2
zLIJ8*aWx{%K;6d|Ix{5Dbob@vG%Nc7OP?3Z*u|@<VWxq;Uku&mw;N|>Wp6k40ZsVD
zb!F4s0X=_w=j}!?*)z|MZNIMd12P|S_35=bvli1lmu2-5<X^w++uZu?wWyJb3TS;q
zA5X25E%(2Y2j(zNCZJ6fyEXoWV-#GkLuqK}a$#eOE56j_cfPyo+P%*8<Ti7Yh2&?K
zWbBw%Mo}xZ>q%{}-(>XGa#F-#!Q{(Gi{6q}YGB-|n0p9dPCEjOys3$n|KV{90ZL;2
z7pj%L2Q5U=gPY3_@)vb=o4i!SnXOJFU1e+<$ciW~GK@Rjnk`BeGY>dvDa0+NOs79r
zTMn(0lv(mI@AS&ln%+xHKHoG!qCHFXx$V<F#ss2&*^5o!oBXs7tVmvIO)x0t?Rfhl
zE_%9Rrdo|`J%VvmbB?@E3^LziL6L(8#(?KAnN7>_uNlqH<d`lL97?Q{9_uOI&1stu
z;D5p<MWzzCj~8J#Vpw^;e|-PVy4?eUtxG4QayqC5RNtY<kkLhs-f!<Q#9QXySV-i|
zdrXZ$Z!{ssJOcJkZ_Z&<r*<4%{FEpHd@^I8hozT2#scfJx|+T(R8bhj?b8RV=9tb!
zZrAWrgK!$KuDiWr<@#$}z~o?=SaV29{1bu~+TI}e9~HyuPz=_odUbjo*0}{^WZp>@
zsR{Z!SG~R&iywbW-;RhCOi!7oygr3yntl#!xRppW^VP808%XgkOs6?aq#jMkYm>v$
zpNY5e(|XByyzz|NNiQzgbYT0j5!=`%%OK5m({jrf*4n<anYZ<Nld)yG4{y!1S+>e(
z+%*ufQw7_W&h6Dkr{>`FZAQ;{KkNXqCj`B!<~po;k}f#zZ@fIzOgmYIg#S(rTnsb3
z^!_WX+~FNi)aq8Xt6A0KP%(99ywSTBw=>n-As5)$`k9OV0oPF1Z!+1^L$Z4B*UeS0
zysTq4$#obQW}Ei`g#xIO9YJ#s1~r=Q6fRQnb7mLf?URYIl<lo!T|<G}Fsx|!Lgrl4
zci%f45%HGy%<VZWS=rx1f;$!dR?jWu-(#*Bm0qn5Iw~n*l+^a6DHQ51aacSf_UPWX
ze7?HDrJ-7&5`5hp^qRt;ZdL60%dL6adIYTS!D2w2@9D!%BIe0BME`y>iPY_P!>47r
z)NuEaTauK6+f9S(H|r`k4qy%vSbg)++~SY8?f$>V9H-&WJI64hFs}M%avV-c!krx*
z`UJEy!Am)pUpA^th7Z1&1HVD3+?~(diQW^?zj>>zxse%|{0{|Kkv<TXZ@<_$kMIC6
z*7S#b(bW9m9r2RQEBpR~ar@16rLoq;+6UP|ha`%7r(UH2*U2oY*#s}}xw?bNGzmZ>
zpH!HyD=xPmI(xBvG&8rnQHu=6r*pnJ_a1s2cjG5n<=-QA*LQFi*s9mVv~o83r}?t`
zxl26aDx;yNjLLndo~8{kg&kGUKbi3)#3+RE?zm*cHpFu^Jk5-psj+B8*csHzcxX6Q
z8GU#3FxV#f58Ff~eQ@4`;Ny^sImttojoX7P-^mwuL+L{qv-aE?nz8B2%Sq*lUckY=
zxC_|ZX`FAX^8owg2urle2QwGA_|_N$vyz;e%NDvm-xX<%llD{c9DZB#dx`5tdn7B9
z(YXw;XT9^$*rV{SI(Q}jysYvK#N&2q%f4Z5&&$|%X{Ad+tIs4$Xl&>+-hv6xf_OTp
z0N8DN*!Fun$(n`ZV~@dum&WZ~B~spJ3C>s?Kk+|Jbb=q-O}N|Zn^YeqD0jNW3_o;c
z^NY0(ngSPk-{;iIm|TDmV~@T8c7A=caRr~5$N%Y>cJ-{-T&rKiS=Sm&QD-45o-YQ8
zQrOX7*<X?O!9vg@3^tO(0lGTB=L>2T#J~z5>;h;xB^Q;~z(;EQ*$nRZkB3?kV|tS%
z2fp7*NK|B<$O_e}qY(_<W|tVMFf@C{xO;bNrCdTX<rXrR(*O@i9hm5Xovh~aRb`A%
zB_9WQGJ9y6bb%f@#vWR0!aP*SgwTMZ7%8xGPcuZN*R8*lf_!x30tC=sATJ(8X6jSY
zT7-R{2qs^yCzVv$4NR|pSY1IS)h9<M<W6aU@84Aks^_t!2;rAGC?l|TS4ewMd)Mp;
z;o!KJ4k^G?y0O*r;KI9v*1u&RkJ;1FN|8gMhgbxp!eyOij$}#DKcP(6@el@>Hq08I
zeJ}Zv4To6jp5%n3#!&MFO|XJ=cp}*ecfOFrH?IW=9L)PMA7m8#XUa##?5$R9i4a$a
zvNe^d#@}qymmgw<YB8R44zs>EWk=qN<I%58_;&fsvNZqx1>zVsO#)0drT5-<d|(|d
z+zvNS)cpP6bl+s&&sx?{HT6h$2jtuS$jXmq$j#`Wb!B9u_unPyeYpDx6G@uO%_>;v
zr1W9H_4Ut^3n4gp*L%r5dW7C@fNZ{Sj_(kbI3VFk9622<k9XDKZ%Coj{+sW;M|r=4
zylX`TAG!K{aqnnisa@$OHZ|n+qhKJJCFANR-3PthZ~t}%CTl*yQ|aGZj|uuKFL8s<
zBB>&Ry$(*+n=NC>9xq9RVM!}#Zf?)t)MJSqJ&%*K2GQ(ae#J7z=AQ<x_c_EnyCY?<
zZ%)?bLKKKe0fNXitJ?CEp=-Z9Bi<%6s6}#r+LHeylx43P`{&X>)Snrk6Fqzax~)rq
zXODSjC|kRkQ!`IGFhHYyeBiq;c{0bHE!_EME#U^(@=_z%U!=-|wPUn2XO%FcIbWJY
zr%<v7l*q}dqdyeLfd^1w482{rB&6A`1$$zq!M(P!PES+H?Xxj@ktLZ3%nAApV_PF%
z8k@1n*q2LV0L}u2c8GJ7dK$iVQc%_2=N}pzwigVxGZMljVKQ|K>^PdJ61K_IAu?Tq
zzE;;7fsNBn63LqfxnGP>Vu@M_%3NfWa%Ii5PSw5#nkO3Z7n3s&WaoD&^;Urs1&e_{
z1TyV5alSVtMO39bcpjUF&EQBW;8F-%ILqD(R#2V<;%3oH1^&qnCmEGP2D*c^c3<sx
z%G9TOo^c>K=MQua-f-ZR@uhJ8Wt@jp%<9Ee8NUm^(^Pbk9sax^Fp8(r9+(qT^-)!C
z9?xIJ-Zs<9{!DI_MHyHe2B&JaA7~W~&Q;1xwXNrI>~!#WdNa^=RWb-A8gK`j(Z4i_
z8CFT{b6(vlAZq#1S)pHKQguJ_M>tqH<4t*{@BapHi&y-o<)?9=Xi{cAtzr(z2Dh{o
z{k{-lcrEDXBi~)gz+N5V^c+?sc7(TrC;58<;QQ7#*NJV?4Ke^>P-?vTRDRTOEqn<6
zdq43$J5$Oi2>AFt<7(;;p#fU8Rh+?S4D1vkJa+zV;4;pe*+v?Lmz#0uiP2=Eq+)CA
zx46oYd>`LKISOVX$IyP6IKm<gwFe8ig(bPO<$y#$^wc16>_}3t(qJBk^*)<H6euPX
zg$;_DQ40|seV$1A#3mq`3d6%OH(2WIm_S8<g9%v?p{T>pPO94gvP7qKHS8(2$daME
zq^N|T8U>ihGkuooBB**|r$o`{DkO1TOp`lD6DB+?LNoqW>ubABC8#kRdlPQ<212(k
zntl2`l+7YEo)O+1pw4Ii)bzYj&C}|c{8Xrw?<T7y=S>Dfgu~N_-L{JXWzOv<n_Cu7
z<dXirQuw_S#ie)D)umt|yETId<c?dn(TsI26U&PM*+wT)y2q@zmAZsz{^Hnf0^G^r
z@=vaYNy+z%e{6Q2J1TGgko{T(RD{vd2<%wKeYnK?T_fSXjrx%BAO?e3S{RwQ<@lLv
z9p;<&LD00YVxdhS>q)=h>oq;XXhl&zd3)|`*U-<qc_&c6-CE)W9;Hh#D<Q?OX~y}k
z^M;|JlvZw~Y0%bJuFVv0-)vv!D$f9|p+Ud|YH9ZV5cK0gWA=iT^YzyJdg*|m*vA*H
zS<=4_C#rOG2>|X4BzN<(PriY$0H7^i1cBb%ojtg-rjuatTlAk_4zCJQ36@2o-&OAv
zeH8^$vIE+cBhGDz9mIBJtb{w$*3|p#nY9nk%{s2CXV1{0(8$4@4HCNii{xn>%u!*7
z!ya(_Y5g)#RTEX@XC-(3R*Ub&?5aC80!u;N4R8=5RgO9|6MPxap-}ZQDD`^!@GK-E
z*s9}b{@}sJ8z;0E%)B*}Ug8zxPpF2+4wHNXl7<aqAM!Hz!iD9bBp89_cS!6Iq$U9U
zeD-mNUQYf>Y`5}O`79zcE+he7=2&hC%>Db1e(p^l#zlGn&eTn+7YhK%bzvOku%Jqv
zqA;!7X<lI;jXgukV7Ke=?=;bF{#b_v+r%Z;?R$pTLC4Ddii2V43ODPu$dCk8YkNMu
z{KaPFjXC?aN6Xf<oJ6JeTlEC3KXjJ`^drT#3Ni+wEJzu}blZFU>eBKS0B6kA83vAY
z30#RmB60g#Arfkc#jFPe#}(6ah{+|I!$O&>`mv%IB7>!cTpd&{36rr^LF(}!zELn)
zDVx_jFS7wR&11%ExF%)VQzSI>0OmXAfj42IzX5tk>CF3SDAXV$<3p0)BN|f0g9&RR
z(Z@L)IkL6-;0G`|97}W%hBlX3t|%OrYRr2${N31_1}opuyWgj66}Uvi(*#f5f3dd^
zzsW<7w(Ke_mW+g_vNL|n_*>1-|4kPx_bZ5_Xaanw&@6~O0e<JRu0rr7RLK3X5SKH&
zvQZ6|{{bKW0$O*EkD=!q@AtQA%?}hpz&ZT*UvdsWEP4=xW9hC(9&xuhnE8jGr6u7X
zQpXhj5XpOiG^*9=$N)m;55o~aRN@XcNRc6wVr;8nkB7yQmJ<F(zPP{vTN#gaI0M7E
z(SuPQSCMXc!S8C27<hlacF!$V>h$+kOyq<}21D&Xhd2YXpao~_NFZdi@{bnLzeUS9
z&)DR0&(}oBg5XS5j_>bh66Ahsl|QoqJ}-UbYO+6z5PB$WwB1EOmpg=P@+IDFkW;C)
zahKY%^=A$9K(DA6x8H7kMz~VtU<{$Z$OF9ZlX2s0HH_1NBqCrHS<M~4?{9a14|X2@
zs&h)o{gGfEwXLP8Wi<Kq4T-$PyVC5}|8B$km;?1~li*ay$4i59MzJTwTM<pSuRaG&
zLv2yxPedU8`;>ov({k+Y%CFSZJhyuAbgAyS{6fw}Q*jV_Jss=-`?`jT>IwmSY=vz+
zZ>R}CINh8sTo0}$7S(mPwcWgr`lxliSlz>!AJ&(7vEWcxXn-HrvmkMoQMb!T(fu})
zpjweR@CMHVfGmxz3oxvHM+Ds=3GLSxCgSfphx`sAM8zFwg2HwR<|rozs*1oputU_8
z+e;6$_eqbl?YnV^OLv;D!4d;X5$-0ycG?j*mxBF`GrG_rVDT?jve+rG+!U9tH$HU<
zE$4ndMdH%gv%UGkZy8!H{rE$x=_VH2gehm^Pc?34mb8bvtbUR&u4Bkspu4c|=YFTk
z$9=zd8Y-v*ii6H)Zr7(Ry(a~$m2Li~U)Rg4laS?kPEGGo7);R<Ip~SbmCC-EGHvah
zzHd>uH)|(0fE8WPAfl#KD;Cl}u{#1J+u<l1XRiU8bwv|t`)ZD&_2OASz}t=Zn$IiD
z<0y+O@KjX8gBwK4r)R#?*U<gg`LDg6%ES=4)>0$(vRocwNGHpqoHK!=O6c2d=DNV<
zom-;^9(w=I$6xRW-0*|gzL$ULSo0ZK*A*|P!hAzB9>9w8gN3*g_FpSWV^tRvG=zE7
znsDL9x`GH~2OzABU?uoxOcL3a3XkXq0(oc@#{+Fv4iHLT`xc0WlpW(e#o~}nihr`T
z3=Z2&Yed2@U7>mOr}vLlj4u7Dj8$Tc-MqIUv}|Yh+o%ylm(**K2U|&%Z~Gwyr*QzT
z(Mz<6=%iU~+Ei*q5PldNFQ4FI8&81}zGlt|aqB|3A&5NPGFG?Mx)C9d@`A_Hne<)Z
zLI{vDvOXuXTXMTrZJ=H*<QT4}x}COeIUN1A#{ExD0Pj_XK92xHr8<Zxw*iD0B`_y0
zr){0)L<!j@0Om(&Kjp$$Oh{FQ%i~`Z*nzdscM1$)!nbKQ^6;_zAfY=DNBeQ75k@r^
zqe=&U4q^sbqKUAka|Kas<STUHx$Fp>AE#;+@@mYazlXwEXqOrJSx{sOc2rdOZJmQ>
zNgqbak#wA^N8A^I=PXf+F0usU$MUa#k`p{_Uq{GUS@-kZBg|`@DYL1u6duz8uITOo
z;`Kv_GP93k+~weqciNi|R3Nb@D=)wOR70ZokLDUnLK2KUFdB(pVjA?m&!<Q~TQnZ8
zqkb6HK)YdJUza|$W^DRE$nw>S+1Gz7`AXhlAem$$^`iQ-Q6`W~A|48DEX@z9Shp^l
zZT%lTTPE4mGdpy>@cboGP*Nn|v0hUM*=`HjV>?gx*#DX`@7f1$FByJ90Q`h)<H&lP
zwM>lftb<|ti;(hRJt;k2Q6qr|n)X2p%b$I`BqjE8YCD4~n$5WQLJ|T5sYHCjNpR4M
z7WU&ge!Du}yiMvvQp1Gz#NDB^(02Di|MH01@83eSLL-3)+5Wdl2ng->i*LU2WlJ{p
z6yZ1}!LJbttD!H~;)_@}$RRZYW9$4;Vh|4Wp%JGjA-}_y@QL@XE#e<SB{0mq18iER
z3N~kWoH-r$UNXS7%9+)^ZX|N9CePO8#Kj8OpWHt?wMOP7imsPq7{k_enNZNvA}w<|
zFq5kp_*g<CHt(1$lmH$%RE6RVwDl=wY`^EUDEI1^o(8g4P@rIH^pYIRCU|DB4ng4W
zy*KK1Em2p?A?MA9BYi{ebJ`pfOv{0Il-AsKnK4Wsb;od0sXVZmK3G6uTQsD)qZ!~J
zP$;2&kU5DC4Pm|-*xHoIba9IWK_5*n-toI>MP8xOoS;&GJh#zDC9=vigc}P#b1jQ*
z{@`U9zs(%t#^$HVKe&Y5;I;qZJuv6UFoDF&(|xs=*@4sKA#Ghk>D!-8c{=B7CxFZg
zQ1lVZ!IKd7>)v*ay#cDKAL+ETMjSmZN!C-xh1Jg-+fsW!Cr>>E@~Ut6c>keIVqZT4
zPh|R%q^D2VNjh7#n;?x}TLL~+dxksL%?o|iKdu0`n~D)bfY+Ebv29;1R;*vVNFx}6
z7G>Axy$6149$Oh;d&)4i8Sq?+>Um2p|I%cc5ZwLMED+eJnw@cmViR$X=E#Ye_A($o
zmKh`#qQ&<GbGm2PtT9IHg*|4l2^37Pr>C+t&1up}{25Tgs?vgimhC0zY~EXI+9ebt
z;0TS^tJVT3xO13J^Pqr>W-z*&ufwBGRN9jISqH0)zY$xnKh(KgRbqiE7a|+~1u>jR
z6&R}6`fMrE$`%x=zg-Nv|J0)D-Z&21=U1Q!Zn(;w%o2xXSLR<KFx42|7R8_1oEl%X
zS=w4C2%3Yl6p6dbk7pR4{1$;_$b*MB+mpAc(OnKIFOg6>`;EZcHSeOY5d_&N_sSq)
z%bp<06zNSA<(JbO4c$~>2#q_w33M6^0`sB$R0HiF=3ar!U@l6MwEUa{_~a`QK|Qa)
zqioSMtdY=+Vv`b;C0tJszOS8HP6S3kp~CJxGA8E}L|efS{t;Q$Of+OSVTN?)8)P~r
ze5=;oLaB#AJo#2wu3@uYu)&T(`+ue|i<>GRE<JJ>h3p4)P-7x@2&x_2YWfBRH~p(~
z4$|eYF1rZT&hbfi$wUm0s*uSb0^V&u30O41(wY)P%6O0dOmNQKs-y%M;L4PS_9yx1
zl2jA@Z-PIV`XG>lz4~87+h%nSp-monnQrxZ=czVX4_m!iCN!zs0Na$&-LtVLdLvru
zd|p*WY#xYKZ!aEkK3<@J)mZOXwYYWSdB}_dhDpyIbQlbt@8rik_3Fv;C(%gDZ}OD?
zr6}V+-vHVHa4LE)`-_(k_XXw~Fxyki!232Q^Ph5EZ|0t^@cG4cOVr@!2U1-I)L0yQ
zbM-nz0P{ef>7pMT5OiXKsd!$r24dM;I=S_0deIIzobe7w5;?H5<t*pWg0S2hrPpDW
zmPL2bhOBT&*pAFSFW4-r#I)O=9`+RdVr~Uh-{IiPwjalbVPjB)AthqzJ(O^Dw68{<
zFOa-jW@woDlMy?$^4HI1`l@A@N??2{C#@@6y|o*1V(%1O3r5l_d{A%rKqwJ=<~!z^
zM*x?Vm?LSC$^ln|0SKMNfDb?UG><Hb#(fMcHh#>qr;roF%G=9|AoD`F^awvsTnsP%
z-HagZHl8Vvfh&OUk?&T22$i6B*-6A9H>n@tfkwXkmWr%lMBnc%@*<$(Au-@3RPwh&
z#Oxr<=l+v7VT#541d2Q)gA6Im1v73nvg3tT{>62+ft1i>%-ELA{_8;qeXVp04|;&z
zN?)E8_bfhJ+j#tu+4cBjXfGiG(wsm|TFax24Z8P<^d*SFYDTu|Xu9C@cV15V+ag&6
z*);ng@11Uc+&b+0=XW10=2>g^v*<npJ`dEQRdkDDuL!;dTM+2;FYRvxAzyJ2{D{zS
z96zhZQf8j7(?We?vL5L;s@|TwF(iyGzoc~BCq~axieE&o{AjQYX=pPANSDl16Ber;
z=F07tdKm!zHIdr-|6j8l#G3wkc}D{#Wr5dxgTc3dH2(JRE#LwJP7RqAkjq8k-W*e2
zU2*rA&+F`C+@ebmLGzhCVKif^gQIZn=#!@qrrs~Kt{B)L8O>NJD|fTUFKHZXF%WF&
zlskh`lmzvz7L=f9{Ckdyv?pxPT|gHHh3%za5c{W%@%TgpD#BBZH11yHu!vtm^o5vZ
zrAp=rv9odr0PbaaWc&S}uyaBc`zgj`fh?JJ)Z38i7Ai{`ZSxS)5GH4jVw(45?ERJb
zo3Yb|rlsSdd~oG2{5821Vugt)SMC63FzH!b&G}l|*>cAJ^8&c4@L05uUD~Aq<={I`
zfmF!Rm|7qbZhGe^=k($F#@Fg<3>w80f}f@FwV5nE{06K4aT?X*S9qAbJm`CJ3M_+n
zeR4XTUp7N{EgmEBDA{ru-jn`(Y)w6&vJ~|<0i+DN2eX3y0bkoZVnC_2sr+g;%Tak$
z1roPvAG7P7<7VjbW3z%r-Xn9Sap2^r#*XJG>E#mj<^-kxcycF3)n9ZhheVSa!()8}
zwv%=oIj3wI6h~?4`g!>c2We{>qwmI>P_5b5AqiLa#203Oi{ZPA2Mcz8FRxhAiq4sh
zr``|(TZcXVi+2JgLSvQ7laGBj)%zT)+KXL2#qAswqBd*uY#N0#n$>XyNE^$%1*$&M
z(rzuSR_Z)6kB#t~qTB%JM^7$3M`wegwdSN4WJtw!v02&YnbPz9{|@|@jE9^ITd~}2
zhX(y?=HG0*0Xpjc1)l#6N%(Wg?Q_<2{+a5B`(DUHcibtb(Te9Q+Quq_Xbk`<KIP=*
z)Rf?2+(nP3WmDy_;);WC*-}*}g&t6YH3Pe!8mY9PTfj=Hcz8J%yv7IxLPAg&T2vG!
zpj2EGBzpe~X~+<UPhtLpMGkA5>6q_(s=(*8C4U#7BAuXgr6q!LGT&LJNUjhebU5UN
zAtavnoe^uiRItz3X@M@6KzQG;kjG<_+C1g>-@s*GgjMdtloS;9iCMufK;MuVu(~p~
zuy4oR&R~>b3>k<xg{8la{d-q7HA089Cy!>H=xm@yfH4Zaq88S8ABIuuHwBWqy<*QJ
ztfpb2UCL_#KWYtxjdnjtlA#n7BCIqW)3D1Kdews_mK9!nA}f8~0SzGs@kNKZ8o-)F
zlF<qoXQc-vi3q$9wgGyMt09XlA-ps8Zeo#dArgRyug5Q<A|`YY&GD%bTm{k#e-`v)
zD(oj`?uw?jvXcJC2w3hzh_-N<5RqK~g$2Xf7yS?;-Hh}mRHVOkMh5-!$i|S^nhYoT
z-}33hKn2@JB@F$!X0Vcd;d1+VflIcT*|{mWJO)@fyySDvJMK#F=At*2Nfb0h0<XhS
zPegsxD-#-Js^_1LsbL<Q|4I$et(1vKn|str0Rgg(*aVLM+42n;G<b9GKh|N!p0YPJ
zv3vN3b^2Wl50HPd!$NDfb`?2Lo&|VqQL!Qsu}2l}x92^o@y=5+amF$2g18dI{%rRo
zyG`(NSjm3B#byXWpDR6z^#<YU8nN$y*f&w6@B0tbz1ixK9}Nq`XkjSw$+sk%Y!Mx}
zi9Y5<FflPA`1|<$P7)1rFgA$X^n$-GPwSSofMS+*GrAyLCRs^sI7mT5K-C4CwLoQ-
zT^2tT%7M-HM`<e|9#&>$_Jf(s@=Gj9e3;h@=9o_z*&=u~wEX#sS+@lU(t*PG$1H^~
zM`92Yh<LlY4_lwqk3bt`Nlck5i%NJAoSph?ledfjLd$MqxK3o=k|<wxgvE(T10(2Y
zlY-tYP_ztkU9l};<$7Y?MFUB)uMJZeQ0%(K=9N#jS+J==WV%nzA8Z;LImV$4SrWl$
zHlf{mP#kbvK~0;>vY8lP&D+8|9y9-Q+SHN*om^*WYTl{3`DryAY=4^V_=Pb6)LZTI
z#4__a3rG?9A>Qvb5z#>Je(I0g1<(68Wlo%b>$d*Y4Px4v*I45y@oT^4dC3)1GN8yC
zyHO+^^lIcLWlE)Y6@0id&yp2&ij}lKkveE~Ygf#?-YUV<`6NA%@^02Eqe_Gt=HFcE
zt>Gq!qby~4t=xbf(#RKPI131}>w4vPzozivHmkwsMUx4g&VFlCXXpQoKtq53y#&9A
zjRmWRcCiirLc-eYUNtFiJT8qUaT5AhIN*r}v^_y0sfGW9FHkfmlxwL)5Fb(Oc-$n#
zfmUY^pxVVw0o92pePB|3tH%E5wPNSJTGL5+a*Qf_nYJP1K8hJ6T}-g9V31p<TCEuV
zh5S))ci13%2DvA|)VkRkAW-OLD*F5B<b1i$XW2@zWeIT`4GpnK6=Kh<0!d`9pU7^y
z@Dvz4vDk##nHqCNUSX+PjQp%zAb^nlVO@6NYZI(iS!qPYf{0PnH2CGZu#pY~>D!zr
zy3knM74|ktlxp@g^SBBNm;w|4i~A~RC}QPU>LK{~=t!yvZozpBCIw4ajUH)ZF}HF&
zFc?CmCZ0zrQ76(Q<AR{#{Px6~dIq_*$7vL}qlDWcNPKhPuLxE>Wf+B;jM!;z@p<Tx
zUvmgpe=o<-U~)qeGKf6Bz8JzoglH%hKbQ?m9gFLL?Kuz^JJ78bqn@Kj*vVQ5NDIot
ze?qLS$E~wd9u$~^`0iblZ<>BKoq>^JaZJh_Wz$o;UYrcIan0`TEYs^TT*5Ejd4i)I
zy*4hDjX0gc(iK)+R5EIEp_9jcZy(!<Wn_6~v;E&e<_tgDG0{I@Jc`}XIWIZp#3|_i
ze*L@u`})&wA5~K^=x|v_e~C_LXvua;_L}z-1Cpm+Dc1nu^1ns5;%M<LVnK){?7q}k
z4`v%#JSV)dHZ+FpG+aP;vePsP>iEr(;ls)cLL4DVkOU(fVgkYq)Wj&LdA>1kt~Nry
zg<zwM5OVMRUuI-BU|`F2NP%gO0rNbW^DpmDHAr7*JqiB8#dx>vqGMjBitbgg#O5L(
z4rUN^=(7B5=&HDZOo|_+s>yA79X-xB|NfYft*T%xH>fDW!}J>uE%Ww$RE(6%$t)Hl
z=az&zF~O>>`ZTsS))F=x+M07`-h+B(NTF6^i%d1tpk{hw^%y3IBE~Q&F8V{SrWIdG
zxqO)IC$!SgIWY&9Jn{AeTPOC_>R>B^*sZ8~%M0Wvp}nKbyTsV5Vte9XKT6T2)1(_g
zO2a2?Rc+cvvV~^!97;)#Td~gxmnXu%+Jb3QDT;;flj6eBk`?Mjbem68C2Kt4BpYys
zD^9B)M%37CXWR7j+ME0rgPTDIfOU8J$3HI3&|kMQILh67b<A-%bmM#H3I1POK+3aB
zsxLZ3sHgECh7B9uuY?kmPn>n95g=dr*otOF5;=yzBQ)C`-g)dmX=fgaIe0xQK6em;
zjyP=0>To3CHVM9Fb#!hSRe!OD(||GX9Iu*q?exsw`?s8^w3uC1;66&IFRR_)=hTeX
z#cls?V$R{WbEd`x14qSFS|DOAVc0aqq^i+xXTsSy|F^dn!{62!ecSo7eaY!p3ej(r
zM8&X44G%#`MYinn;-+cF$bkGuAsn(504ZX@%dbAN*}z(j92aIPZ!)R4z|&@XKpdDt
zl*pFbW1b4N;L1r#P&s^wfLbizf$5-O77ExIXg#K6FeXx!l$uuctr8d^Ts3N>CM1fK
zGDIr@eg3C=A+KlI6R8Mef3GHX>*8f-8}g$lRGL^FCoLZNvC=<4bpL*TTvSI*T7!si
zhve};N#_(1w>XX<E@LK%S9KuLY6USfFXj9l>wrFJy8<O5$F!S@zCx}Q={eAi<dGl|
zPF`!A6;OON;tc|4nq82PS{y`a-e$GC0ix055t2K;4k4XAe2v5x=l=+`N^COA!DDOo
z^IYKCEtPU=7RAtnY2-7O=Okj@Lr(HqoOE05>^fbEt4sLf%e6f8$$h$ifh!4NyJ=Gy
z>!O8^am_b59eo4`5I-`6dgZ|OD7FBSYN2SVvCwVSw&!YBqLG!YHb4L?P`4-5|9$?3
zW!Jtj`>)pyt<6Z997Ol2gy3V9R1bA7&)oGSidh==Zg11S)uwy!00x@XwDj_?N28tQ
zIwXna){6`4O{|L*c(W!d(|$?C1TvnxK08SJ{mT!mJe!f{X(r?7Jel`%PuPu%^1ZoY
z>&fqNBLYRY@gD46Hlc#~z#X<dSy8*#1#!mNB(JKv(I?VA@OSOf)4y6uN&&?TDw&w&
z^XFr4M!;nY-E(OsBflH*lhBV@_xj#%q}T&KV@ALWD_fyLKU4AJ(qR4Y@-3Z{Ev@Y*
z+ox8O*MSwE9#!NkoK2R^@2)TO=vnLM#t64hupt<ReDhCqu3rArX1S=i)ar2ttaIct
zrVBDsi0UB9NGj4N1Y|Dc^bDU;)vo1Q16w`+=)tsKJsONX%`t>Rz~v(I+tz?@$B0oC
zT+{u_9I4@bk>YCo=0UufHS(SK(C&v;_c}BScriExXbF4>KM|N5-545X-u3ZG5tyhv
zQU(>KIHMK?arL8Gk6H!G<GUq2D=w7Wiigp<MUyFn_y$szJwiS4+VJ`?B?}tu3g077
zvFM_mCD#T?{6=6SNKs)o^{;hb|7w8JgAmw;3f)}Pv^qXQu^&Y>5d>FOBPPj_3TQla
zRXcvz92V`6)<zO8*E`&l_$K~2-Mbt^`e{c^UyvlIP&chE5yXSut2lp)VN|gGarPcX
zoWjQmDr-Yg7;?W(*R~lX>Pvd&X=nP3WVvpX??_Y``+YHc*%3Q}k7I}u{o6mOs`$up
zaeZFo#Qwt*WJC@hgz^3OAFyJIf+pl|ib5BJL$B`V`NZwQR}4&i)>uvE-b?N*W{v!W
zs)^e+olYs=pt`Dvxdy8L3Qj%wld@u#Y<(?gjq?=wPL2%apsm^bzNZ)8FUl61_KN2o
zO_elzpM?T@O<<F?_KZ);lC7k?1DtCLaPee+*~%$09=!OPlm;vT{*AQ`P+IeL8IIEH
zC3*>4lS`+a=3J9vxTCZEvpVKp+O;g5vb&uvpF~e0hp^iZ7W{T9E5@HRrM7)t6P_X~
z!Y$?mY-Ll4%em)otMr)CskKF?*;VJ&{JrgE%5=0uE^_^DXRc56te^a@mu(#>`RFvC
zZQ{2cB$5c|ELsw7pR4|jJvx9tODNqAUa|AP9_J%BYfzn)fVN9}OKQyT4hDSGtM`6V
zhun_3i5!f3jv0qh5{GN2y{XPhqo2;q!63#9DdL_SQK$&(y&vu_6i^h%%B_(jz?Aw;
zXtl=ih>42Nqim7;5rz+2m_<H7p!(t1QS7{^9*Wa*Kb$g9$l)tH(=<v5h3k1)kbo?|
zKg3K5sgQbT>;w2=<d%)ca^F;jCyid`JP_wkoNfGlfZzU=Z;kpS$@N8KMVT%^7F<(I
z(YAX0u(}r?4mo+`RB%6CdG$~WR-cnCUXTP;D9B|Kk7;=0sCq|jo^0DzZBzUq-tbg5
ze-Z6#*I;;#-X(1A@g)yzJhQ*C3!~%1x`E@wV<*3CD~=8<yBOG|N?`?DqWcd)#DS1n
zkT~dFc}mN!s*mWYxg+g^nepv>TgG}TI!TB)LIBwePe%7X$^@4FI-H0sG$u_+NOcUN
zL{2Y%>n6p4Vo^F{2=V?q0E2^RbwrZ!(0^f<0*2k``0(HVX}W~b-^ATY?LuO_<Xt8n
z)}>DMh;cvs{r3BJHxFbs?K+eC|7Wh_a08g@2u}SH`K)i)a#Mb?7ebT0(=M1Dvsg|*
z;8>roP!!$B%;=v<KVUyu_n)R)zOL?FByi!nR@MGb*EK+q1>ct4n(lbDBF*-^=A0e?
zFWkguojzK|k=GPUMkyS!hKtsJ+nzC<?G4&VFMg=7Kd|0^9J9XuxhPxo=gLZ<4YJkf
ziC=enOn-U$$q7Ya5u5`&XYMhG$6c{$zy}0(QX*gJO6Nz^@6eh3w0;`1JGpsdUDs#L
ztR>R9H#FT<@qEC2y3oILhGA&DA>nxPe<!Q8G^?0CJ9+6fAqr9smKmX9_&5AI=n^a3
zT<Afr=e*_Br+bhx5Z9wlT0&pRu9rOgjGH~=R7HY@tP+>zf`@XSypZ&+(TmU1Ps2$~
zjGPow$|IYL)}kYxRb%Hr{=Mn3{a9Iwra*?R_;o~GIblPB_BBWt#X^h#zt~JVY1hYF
zX&_87M1i}!VxA%JSB%Jsx`ma;m)##Z;z<oN=~sPUX4A_Gd$jB%`t%*19=YxZN)RL;
z3rERL^IHr#2n$b~Xro%osO3Ifh7o2CCYngn^SO^<dx$eo4ldBnww33Np*G$6e)^@&
zY9E86mdB9iCK$;JfxThEdu%W%)ZS1|-*>*0bbONc3)w2_NIXR0Vg869Gn;x-iKmM5
zB;SLOLW6FW^+#ttijttYSpEc}fh^$sP`@=)Mr?vWOY#-R8VPCQd=dREVMrW^ZOA4{
zf5pz}&$Jn6kUPANjUbzKgqJx8U}%lBMZ=;Mr&q>1pxg6$+4KIRjQ*#+QJc&A8=0fD
z-yir9mck9pt|whO!-ra!ou`Cv8Y=+6v=!^^fOjt>vyJz4p^BTp&J4uCw_ROJBc&@h
z;mk3s@}hzw@sF3ZudSR*qndVg66ODj?2AVPx7DLE=AI2>Ma|JRA=6n$N!lY2VB|6z
z;hL9edbT&ebxkVmT+~bA&{s}F{wdvK@_UO2{?G1Cxjj<i!hhWu4@R}RV(9d9!aRMZ
z*;AA9hTa+DSlK6YWesk*3w~lx<R98+`2hhey2&1a7Q!ydR1mwH(P+=)5v{?rkazK3
zrkVkY-v=Oqq{dn?gaTM35|?To7Hq7f&a%I+GCow28<V4bj!m7!8TRc|V?AQQaZCJK
zo9%%~CmCpw`3`(w0aYLq4~}TZU>%_Qzdh0t_aCf%gw18kxM|Zo&R1=r<A`@v89Ub#
zS#wc{?f7l0(-3SS^TX4S&NkL32O<1H1W&<7?diQ-?d9ZL_CoZ;N->HpT8wY`NOGnR
z%mJdebq9?XP~I<-*a35&##rKoFpPt3v7{q{KrdFI&7e(84jpB+Vt-Wc<Ck_qV#jcC
z$_4xoFVeVr+J^+SK)~pg7_K;WvJ+P0I|NGz!qn$MoM$Z$gVi%;QfP=Lf*Qe<B8~3L
z3j`5i$e%vV#i~7Uw-pY=CWTNwl1B@JxT+O&Axd0S?n7Zu7__G)P~=}Gz~<DaB-NkH
z$`ji#d5OIv`p;d-KaAa+5`dPCUx{R1w@1-&s4HaH8zIZ)M;@;DHBbJ+iKB=;-mJ-8
zGWW0A6om`dUQ2i!PLa$sTJ)-ZUA_8DK0iLk#(JC`StQG)s+B!OVjHa5Hd<=ru#5nv
zamq6e@^$Ug=Tn0kAyC*;A_1nGXX61{EW#m7VsNqJs|05ytyi)VVGhKN!*63I5B%8d
z>0drDSI?;4$%*}JpEdg=I^oQ3NRbRwMMKZ?f9JlKUfL30aFb7--eR`$5P-wv3&#vs
zP3<)6bUy69G;o@;|NMShOonIbN>1Feu6psUb-znk8_-t%;&1YM`EaWL>KoIV>Cgd=
zgC|W;iZHKa^fcD5xAO@*J>)}YdH>$VxLM;vqY*Rb#noskYU>LR5o6aNCduGH^MlWw
zy59BN4cqj;(%vvt?|5D~cvd~%fH{oO$(m4%*^0J*t{>c<f4l?S$9zX;Ts&=AuXP_C
zQ43%wl9r<6RqXyMfpXJHd#=j6FQ3G8BmOE0_0CB;oWB)IfI!zcjV&zIx6DT83!WyE
z`!?3Af_uG09BjNrEr)Nc?plc^&~+W`9m5+1pf}y)iK%AChLq^?epQbAB@i|U!f^g%
z*#B(G&e~^}Gz1KJ_Cn!7K@lW6XASI%`~f}dvh#Ok79hus))xQqv2p4hl}lCP#-C$v
zJ%0M$XVibj%1A9xq9_^y`X-|6PeMSn$`Ssic^AXX&v3?{DOXVC!RON);xCeqb#bF<
zPM?6d#_%1)_;UL1#gYXP9C@OsKr{qxwvL-MhGbXooP^&P5uh-I-(e}Bg5o93<UNH>
zrXBRj;7sV1-3Kn-PY!vJL%`0z9##bHu_4~0Yk)3=SmR>|?7(bZNo*uJhDMvXxeeB@
zna(z342Tf4clh@S{IQSDLVlrAQ!|CyQ=^T>9mJ)P4;8;elQ~RkT=w1Z|0p*QvI2S$
zik6Qq6JkFK+j!C#DLb7qag>2Qfre(cU%%-H@8_&5x(lno|0gF7ZcaCznMXF$2pIi4
z-ThZZjbo>)Ta35$UwkuWZL|qs8?XB}VouemJsbgs%)X1<-sjK$l|mr4SR0l$TfSnj
zJL}W<n=$=61nBnD?J~M*{zIGFWNzzzLU&&E{fMh$V1;0EMaB4V8&E`Hn>!j53jK4Y
zc1(Bn#VW*i<rq>$AQACBUe{qXpzFl7AxJO#oI;u6!HlE?q%3|q7@WcQM>R}^YofRa
zFe%&TDX3j4rJu|m!PVnA&Tds5=5TyM!q<o^91m)BLmnT^+k2UH#_ZdX*>rhbGg30i
zk=xol+#iwtIU>6K{;b#-n4OxYq3APe^DLfwUnGn)2`PujW5N!f(al(xTC8stf;ch6
z%MH7hwS>`Y>ZS<N+`6_=q@&3_-xXY1Ui_S6Q!9VF&q$YtLGH$rZ$X3zGW1UZl!-7u
zWCyNLjIaC-%Y>0Wv>hZ&+SG)6*RMx|w`6bq!eWoAI!ujooSKy`Fs!}1VL1%1r=$9%
zFg3RM&R+R6R}a;)Cn_Ec=G64&?DPC&o=4PYO8{}kD^rFw%t(f1Qj-3!p8_@wKgsV(
z%tI>t<g3hM1$<Ndl?bC>Be!pPK6OgPYVpZg_<BteGK*Pj5$&i5rxc#8NHhd1xQR!g
z(HHzt*b#5v<IHWZ3TFK&B+W?O%zo4A(5<+YH`OTQUI~a0?6BT%hoiPn!|{V6j7vEM
z#~c6Mr2;Lzi1BV=ol%|3h+N%ubAF5&qhxtyqjYl^pX)jLNN>F72Z~v|(QkV^u%|}A
zhI3Z33U~PB<IneTw*2bE0?*mi2bOa^HH;rFYs$Xz>aQHO_V7atH)1-7{)V?+wH$4G
znNO9?xqOslRGUpTBA`9rsCd4iEWNS+n4w~^Y-SOtmUIWk@8~!h)=Q~cCrkO{IwS=!
zehsJtqvf3N)y;+bCI2ML!G6n^Jc%0e(PRVI+9_O;v*uxmg+;5cs^ChePg08PY9713
z>+$=h8qms<3}Q=mjb16x+LE>AlA#8e3w0#vVcd!g1^X9bgo15RIW3t|jX<rvi!!L2
zbr(C6e^`A^+t}bSdhmXxzE5d6;~x*>s4fiai$p5c=fV6wc_VZo<kkA5EH;ex)5(0f
zT99}hwd#A{-sE?(D&qip=qEc8THXhg*f7;PJg<bSi0K;(sL%Ik*e`7c!aTEsjUv)S
z8lQqjWP_`WfJq`Ut~i>aw2#_INQBF(EMuB|I>HUsr9A*Kz*b~!)`YOSFPGX9R6`qy
z%1Kh_OgDu&4%r<})}M`lmTR#kKvG@9;6a2c(;P)dv=p9Am>l7T+HEcOxw0ZtIPTjv
z#>XFC{m#i+E2y&A={A+x*y?p<t|rK&HW?g#;3O$aCOspyCaSebqrnp;ZRJNfMu<e)
zDAy~uEw;Ysn=CgS`Q4d51Ub7QNp~tg^63`9w6g;um($OheCo6R!_#{PQvLt$<F7O9
zbr6m%duGqVF|$|52nS`8?Cl&QvI|kl7MaJEc|?-ztZd2Ndqw}#=l%Wt@3`Ua>v<mI
zx~|7{O^`>ZCly&749ZQsB&7Kw9ZjenS1V|Vtf-`WZH_aR!?%R0zb3$-7sjX+W_Y`Z
z{{yng9S5Um+c-X^1Kws<j9<vh#)o(Z%_cLZvInWMmVG<?oDUHpW4#D%2)hr7e#sps
z$3{458IfMy<5)@lF`?k95l?n$#)TM$tkVAwPoL+aVwESoPDuIDfC*f6>innb;fQ(a
zsoPB7^U#&_Ke#LD&*3N~7ARI&D(JJgZ{`=(I*$={TjzX#B%yNr_-j%)?(jd*IpJ~5
z4VmoyO>f&=FE9-GnlxT?i7F{z&6staI=9-43V%s<&EzAVdT&dK{ci}T5?z5{E?6Df
z%tUg8mt2qLupxb(k093HQ~u6&G7~*o^VOK>b}bc^ug%zb`F``GW0Cto;|hU7cU2i}
zot`QMjxh3RXF)BeFcub%^6p2N5`BCa%Jv7oNN63t{rHMA7u16nBxH;aTYj}gk^{OE
zXXgp?AQ4I(eb@6q6Ch5YocLCA8rMVgfJ4|Jj^-U*R(Bq$CWwOzt$xgWIu!9a<R$Xe
z@3lHj_1pyex+XubQ79-E7pzBqnsjO3$(&wWPW5w|o8LsGfBD_?D$}xF*#h1QU#Zoq
z<KwJBZC1SfXxAeBi}hd&Z$p@UtO5ESwfOgUJe*WRasLq=F5=0P7l~*7^>rCi*-u(o
zTCM_d5l;n!$gAts|4|(?&wh--7H-K&r=Qf(OuIj6lS#&5+p^^b7WS4hX%pTA-2I|=
zU1HEqXAG8!hySPT8lM(rHl1`yu&r&tS?AXdx;ggL<Mz1s)z(M#ZvVBVNv(fv30Dj8
z;m-F|o+S;qfEulZzWtmPN*q3MY6&`HauX9OzN;6!HAUNEtC@@V^hjdkrKl)d-Bri0
zN33b^yNpt+G^^jJVHv9>aLyyIxGou`8L;h1peWVl7eU~-1i@tL6Kb&bp<4<xg<Ycf
z{#fsJD-68%BRE40OfYP2^-8bz71L^6W0ZG$@uWRES2%x+hak*oFofwEB@uPQK9)MJ
z_2!^^k&^U2+ibOn=WqcCe6B)&BZ~!bLj~V_iKtTtq6>pT52A)G+Dz@CNW@@>Xwp5b
zn$8PjE#&h_e;=i(ln7?rKi-_s&!sAeS`F{!KJw6z6BJSFc0eD1<EjZ^1&5~JRop*^
zRsY%nYs#ow^ad{t2H~k9MDh~cQ5`irw)KQRqID@<)S=9|U0~sw+el&@c)t&%N8#BX
zI=EfsqR8O)z_G{4QOp|;^$>jt2w@Zpf(Ons#aDGHD3UJ=z|jZGW=j2cFWEy~`b7<7
z9Jj`w0T@})H_k*AFTJdB8R>CN{}9f`ah#|ZbFQyT(sUI_nlL6Y@<FllU!{<`mblM~
z40#aiQ6^Jt7vutJHB~lQULJ+hCrYYI9s0umZ^gnXC<YT^1_n0km>O#+1u?h}?kXv5
zIX?fsm7*epjO%^-a|@gd!VbxaqrpOjJnAnuD;`SM3jL!qzZIhJ1^R>>17w98qHnge
z9d*5O7oI==tVu2t@jy{PH^qFzAO)rpvysAVOvuMHOH(X|i`k<^E0!Cn0}*6ZTs-$~
z*J-(A3_rlYH3kF8t(CxVr4Z)4XDL8gXs7HG;L$hMY}ir30$5g_%C=#f2u+2$&54C!
z!6=pyy4aZ~$uo)~nlgI!n|DdcReur{Kl#Z*)9IJVUZ|GINqq4JHd8J*K~IFjqWOt7
zs-_BY5)?npDGaA*kQo<>88ApNR}&J4dZMdg4Dro^K%p1F%&A|idMN~^2egoVKsQ_L
ze3_`Y0@XFotcel|?<5`iSf(LHUJowHB`|yEG_gMZhviElWKm%unRuD!;Z(QFfbr@N
z;_0%Sv^+hMDZ<~lyu#z3Ada!8C%VUdX2Z!#Z_UVwK8T9`B_;DLEFA7QqDs-{zZcq|
zI05erb4{1}735krKQy_?V4l2lID=Crm5GQmuHaKH+%jFVn^M*N$6>Gs+x^*pji(;V
zGqg)jH{noVZS?=dnb_XVe!r<%375d3y==*rWJPtx=~e=a`o;3miO~Km`dRg7iywxi
zHr`i`*GZSpcAhhFIrznYd;@zbBvri>f3Y*yhn!F;nJ~Sa>kBPO_mlE^^Zzp%i6L`W
zSG$v0N346NSO18O#<tfdlPzn=)C0}v-R;!ai-cp<G(e4(c$9>mG1aMwY0|5=1*84$
zD-Pr+j5528;e_pEQ>%1A=m-h8JP`qr<AVk_vl<;wC>C$sCy12$mO|f8T$Up834x|u
z0cqS0%xgYY$M(7z_ERxk+vz>FMeyEw*a6U^m;eI(bEcR!f*xQr1*w)osEY?P<ep7?
z#qvea5_$^x2yIdMn+JJi6dt#^@ao-3JXrNrq{?X%M47M_FW>ejhg}uMhg&$W@zGV(
z2)Gr{G3Z%o|3Pb82?VO&vfdh2a?`~F-zH2~Qv}kE1I<O4kD(6QT%EB8$ELSlYvrnd
zJ|xrt>w;jBk{yL=Y1%ObDK*UdqvOO+&lq@F1XLL~!G;=BGzNWQ@)Gw~w0Olo`xi+x
z23G5}4qJ^Hj?|~zLPaPLm8Yd$lJ6_=R)VIuh>P*C$|XX5g<qijF8wHtLffBxp?`V4
z^np0dN8kGJ(uu2=qbJt&1)BX0LyLy)al9XyYmrl^!HuW>&l6n2Gs0_IceCD(To+Dm
zk6piAXn@0i3X0F5pShi6DHG^SRr5=ICUe0J2dI|^kpya!b0ea`Uui1Z|8MO4Pasu{
z`Q$t%wxDg)h<a&CJDqsRl=E}|yd$@1>FVqKI_}W2WBN3k=3H-fw_dMeHfExEsBA9m
zzixnY)pM~aWBk-6>x_6@-?fl~D-qWhQZrZo`hnN}uq}y%$Zk~F{UC9A5{t)1RK__r
zW}vtff(ewBLf=PSvywspbrAXEH5po*Wicv>dZ9+<I4<eSJ7_m^SEs$^dI(-!FVXfs
zzZFr!EowVYfT1L@h*$$v@b|Cd-aWwnQ|VDSSB2k0i}Z(9*)3%F_6ZxzF&?<TFbfSr
zm=-Z7LgOMeDvC;(Ew4xgWuUXy7P)~GnfqX`m@!$%<6N(Xi%uQf9?yEeW70Cpbmi>m
zxI63R;MaGt>#u-9;VfVDYrtR)bWv}+k>%~4N=Hj~RzBw-oFUO0ssKAVCN2Y1*sH-l
z)p43uf+=1HZ~TM+T}lARWEZ%2vP28LunkEuBK9Nfjl@z_OGn)?!|4|8<${9o00M5}
zMFLuYZBqdmjlwS#Nr{j%i8-S@v5+sioWdVgk>4gLE6nwI^rYR;+$f;}+hb_ng)UrL
z>ZEk-56a`((YJJd(P;mP3K!u4dWCF)E+%b?i)btFjHa1W@*2Sm)MODUX($W_Rci4a
zM7GZhP?}M=x11E}4q#+^3R|R`8DXO5(~z+Kwe>Dujk%T*WH-C_%h;j_LKAL$g7s_5
ze(2jHW%aF12w%Nnqq9nC`#j%ZK;WR+f@P}wNKMs28b`0ylYbvBuR8TI*?VVaiJJ@l
z?jHt^x|*AR8BfwRN1tmi+(VW8t{a7tzB;*Upr~QLkkIbd(2)gt8nl$*VI3poN_t+D
zk?V0+!7k^2ys7mchbt={^mNyE*~Z}fw4jn?=KZ0vn6#UX^EdY3#n5f#@ehUTOFdTD
zfD((`uJn^i+4F7)hlm<sZg54?abYHd9XRFGMMr5{6c77@IBJas{`HW!eJJ)5%KCfj
zsCK6+TPRHeIAU?#3Cq1`W4IkjgtzTYjN$p>f_(4D3p*3O#rf_I8#jhhn!};<^9YAW
zb#Ad$I@fqAfnJ<km~&C$=R+`l57Z4iiX9f<Bb*{cP>1sBQad-m)TG5>9C?>iV4Chs
z17!B1BGBFBz(z;Mnad)EB_Vz0hlLrr(97gI?i*IS!R%R?^*<dCZlK)3+A8XOg56^e
zp?A#q`A<Xv<%lB|%-2Y!!p6^yX?Vxjzw0lt3E%jh(4t^$P$WT50AIXr@81vAb_>VF
z-H7VCM+~(fR{ltXtL<IBM$Uym^|*`l<}pV&_cB<bpC=cd+2gv+c3G$}`k{y-l|o|(
z?~030(X0HV<7a-RN}+l|=){q5KRgfi^|WM)y|Xw8je7YrQ<O@^Gxa8jR=;Ak3^Y*R
zcWN*1R#1l>8qlIxNEV%N;c(btJt95Bl^WUAWcQBWtcP|>dG6u-5-X?yIUAe;^JACz
zespQuqw%}NQ@BPpr&vz#y2j%vcjnTHYn7kdpN4bsQ>+8QX5$RyROL-ll}h5!v0q5H
z0`A}!!jU*s*(q#D{7L<T-*NH6N%q4H^lI;2u9!`I$IB_LHdk_m18i>-6f*$5l|qgS
z-xU}jnqo$Ah-*DilPfRpZkp^?&NN=ke)IPSu@5AHvz`J;@V9QdK}f@md7@X9J5w+M
zey(R^%jZe{1IjV(nFmEo5x-|9!W(-?OJ{Xk$wg2|KxVHqDG>-Wm_m>haIOa{lad`9
zw3-sACFJGF4d0?yx*v6{4^V;`0A`Ic6Id2k#jS^*J|PsNuSf6Van5@7PO%sM5Oa8b
z`H9<|)Xg}75<xl9rKc%iG9h*)ekYMA<pnjf6me)KOuWko1zIhkuvI5JF{HR=kEY3u
z)>o9|&iy8ic!g1dYM8whJ7K0JTGY@(5jY4^gnl*#IkA7CK~P<z$(BKxl9^0$cU<jo
zmF%nUyIKzy#~z`$i=|MabC5V^(K8w^Ot8-*AmEramr(GPa}i=WfUZtwAzrIUl8e^R
z$R17hxq)a$<*h{671P6{M$_*y`aX}+v|J6)FO_VUe^uC4w}!E*z0Z0e|5$Xbn~-T_
z&*d(dqsxos*xrl!tRas;-PFgv?*;eQ4yF2gPUlX04o`htXIn#3w_Y*jRDWS{`i;gt
zhkzS%YYVT`H=S?t#(N(|*0nLH<o>wfI%XHI_0C}N$!ZxKiCKvb%y=_zIN{SqH((1@
zzkJ3a?2XK59}7O%F`|AZsj>-jh~Sd!6xY)p?ts7{9M?xDnEJ87`nRF~n@#iAE^io%
zTzzMen-*$U)BR_zVcu<bFVd*r@=cr4|80$HJ;%7-)c-;SO)`aeoe#HKQtqQozpQsh
zX%iEE8xU&#X7$6v9OV5JU`o|0O8XwSoS!TGu@I>iaQ%{PxDpvk1lJ%Ty`AgfMvm7d
zuL53VBVts;cIy>Q$%Qy|zQq|^0hax*GYaqLEw;hQzG<TA>>@I=pI^Q;usOef$)inF
zM5ikuG?BJRZ0i;O`N2b(_p~Dc015gQUwzJxDfI`MkU$+qRV0ObAkf{Tbf54yqkWW*
zHN#ovB`4V{$}cv8YUvT!^#rq0VfU{wJyd)r)t<(`Hd<YsXHf*MU<G6eCx`|5Dy5GE
zG8X$`^s^%hjz682^yPVG^Je6QoetGSzmeldYx_^YG@*nNo4kIUVEPk!(`0=c{H*Rv
z!A7`Q;o*m5yDbPh7~q3*qlLkSet*ctV<jxUgPUb|4!Q72u(1T%aB7_b$rHZ3(-(N<
zE<cO+-qWV8r6he^l#&)Q80-+gyt4Y!A4fp&{nRGBcjvC{8N#Ozm*mm++cUR|>q~F)
zyyw)PyWXEG8}=RyV6>3@EHW52_x{5sy;=f)SPh3~1>6Ps_-|F%67AaRT#$>_pN>;<
z)mj3<J9NQpr-*Iz6l7dp#AvK+>6empn|JYy|1|wLN%VvV9g<Vxcb-N0*{{3CkQaAn
zs40BJ4Hn$@F*7G8-Dn}xY2EO~spAi{$+#*rw~&-fO9nT5%Xr;r5&<O2n^@<+Twrh)
zt`fRwV3uEmILm&S3zp6LWU}Dl$6vugJS#w3i3(-w(PiKi%sW<x?u$y?(|D<>#r^j^
zR?ZC^v4i5^es!neIj)t9686}?HX=qUgF@^D2n}xUm{?zRG62Yf*X&cm++9w3sZQyS
z1!0si(Rpb`cJfGjFe%}$;Om!?CwNFeCmO4Gtx4j<$7`TJeP^>Vj5rqHWQ(TVrtIlc
zD2!vsFp%jdB4dO5f1G{GooAtS2gJ^eO$wjVbl@vUGm003Z`6yF(Je+vIgua%0|f8_
z%?gNs@kyO*vo3A$<4VXO2<TxheMI1U3dlpK&OT9W^b!ezDC99Mld^Uv9<Xw@6ab%5
zjV3iP%ISf!Jkwc+wDXbanry&(pE@;|Ev(#sieEC7K`G7UDL|T_6DJCO5G7#|TUX4r
zNZ!=_>=aA*6)Hl`_~7zqf`mG42JIJXL%G^7cv=5Mc&pjCyHkqutqap&Zua>IwFeH9
zo?Q*r1Ktb6U=9>JiTNC%Ui(0ULhQk!Jw-Q5th|T_OVcT_bVbH{YW=%ywi%^(ESvI~
z7<YlfQ48898BBeVZsSP<Y-iHDmf`Xu>*yeBXv3&BHutLKp`%U3mO%xbyhAUaPTk0R
zPgyR(kp>hQvvr4IQ;%{$2ZI{FX)Q%AWDQw0^u<D<blunq?+Pd0o1o_?3^}MxRoP9T
z3tmLq?)$c_jbpT*+X^n-NLf_+4|MW)Yt=u-s_```DtqQ2q)4ei?9gEl3*jtdnGz__
ziucK3nq2^Qp<bx~Ub`*zt&GUJ2f!`ptgVx0dZU~c0774y6xl?4M1szZ7y&pR-YDJL
z>$%_OZ@*WRg^p{>)7Cwdg;3Td)qJQ|vN3sT(4pEkvJBftXbEj8f&xVu0KN6to%;+z
z;G=1W9z>nY9f!DT&h^F~kNlm|U^S)SOWTzJ>CHw7@(EK2;;{~2J*fdW6NXVXRGqPX
zRjx^4>t<)fR}X&uE%Vsd)LY9?GD|e&bEXtmF61<3BdDa~!xq}+2@?d%c8PBW+Uhon
z`Y62Wz2o>*N1;(4AP?bR?5c??Q8^EMluI4P|BI1%0qVPP*KR9+(^EtHD-0zt#ZzHy
zbzxL`CsPW8qOyy>wUM<sn|JF`VV`m`{&4i6Y!SCMv4o0#&!kEdUB}`kQDStk5YXpx
zXU(CxM0B*@EzsMABs}Q)(PB%D`DFIK_k+88t$0O4T1iY0>Fa=?i=W#ktr?-~YIE0e
zlR+bAVUyq5+<DQI9^?<h_gXxDmy=xX)HbzEL8=MtaF3{D4V@D{@;-1^h;dN00ym;=
zbw<Usd_MG=M8E6`Ld~b|COh%eo)ZVzvK1e`IuzWS5w{$Q&uXriuHJBnc}v6L(eOz6
z%CxG|f+F)bRyUpHm+Zr8Q?L_;3F9f3q~}E$b46|KvHAp(lf2>XttO1c8yNiVdb-xW
z>~H?}-uS1Io7FUqEFZl07LCcGp_~+f>cW%UO{IIb9e&dbVY_>dH>Y4Yy~M$Hm#_+d
z=9*lM$HnW-iy6I7M}CQ)Zuq}=eX`Y+^G<upM_KlO#*dF*Bb|i^u@fYGHC7QNY=W24
z)p5V9wM+3N#JFrz56<yup@}>w<)bNMMP>248XkVJws@iCT*vrf;-3kp;If^1%l4Vc
zuYOk}@z<q^71Pbub#sYw*GHQ*gI1N4hDQt>H|_K+mTg<uce@q+NmZtfb&31_;kt42
zYhB8{?ZQjf7FLyKGa^V;!8l3ozr1$3`|s8z*#I2^?WIDtMv4g@5VMwZDj?14h=}mx
zNT;z|;uy_Oi620wdCK1};LJPxSzzbZTW1NxrbYZ&5(rZ#jpB0l<yA?w>@O2O6gmkY
z0pgJ1;8TFvS$AJKxXb7s-#N~@O1=whsum`SY%USoDcQ>F1Nwo}siyG`(Q0)OJJYSw
zcs;7$!n{)*+-#k46+4AkFkogiVMhPX&1CT_?^;@WZ6BPL3+1eXjCT%Srx&utY6DXM
z)9sq}k)%r%Srj*^CxIfq4t3E0m4UtUcW4ki^?suMpdRnQGq&NuDUAv|!doJy5k3~@
z#cYa8jfJCGD<&-3c@^vq1BG=a$RGR3)n4Qnij!eF=)@7&dY>m0pJ~bLPMo}I&Qwwl
zAzK1bjcZPRdzXmV9DNF`{J46FqvF?3Un|r|A)Mpqc7E2VIRBaeHOLe;HBl`6_v(0I
zd1yBPwZ-)MCAY?gPlF>qcbrdGfHWBRzV&YEu1RIG=JnE+o50Qmu9n&-)ct~F@5pFX
zw01P$gl`!u7NTj`-amPfyZCMA&D4X;KCEA_<K<zv?ul(U#9r1fJo_rJ-eEs9-@IJ;
zZ)xj{a-VZZV0vWX7c55Yd-ZcCvGw@ybG>6t$4-M-lFQc*>DMh*4RqoSn<v$C1g@K$
z*bI{!Cu`!j^^B`@a|g9rL~_U;vsF%wCUIZESk%+epxZm977-rB?Q*3VHr<bkIk)>n
zc|AV&*Ltw_aILWF^9~zrT~6+t;up0Kw382PJ$gG%I;Yg&7gFcxGpNh^g|Mol0w2`G
zeG3W;WNWf)&zrwc?%dH}X_@dT+jTjSIJ7L-8Q)~;_C3;alqmbTHjsIX61vK#(fq=6
zCv14{D!jI9?Tv&!(<3%u@`X$2Lb&Ik5u_eG=nu9AJ_tCNonDbD4hhD78R0uO+iOJY
zuR;?zZ(%sGs}G(ek^rKs_|=RcYI|Z-l~yD~9dDZq`nj47b;r=VD3KkdKg~nZ@A{_h
zVRyH9Y_I?(8+DND=$y#}&s8--LXpIv>u1%v7(OA^rh^g)i8(*C(>&tb%=(G{cD9%I
zC|91alNrFK!%NuV=3J{jmjFJCNfM-==_)R$Ng^}=`bGEe`_e}$Uo+EP-L47+Aw|p>
z1$l6g^YyC~@^5hp17j@Px6Jr{LSo!!*OPN6oAMWSGqIGT&S+5iRv2{McB38uep4Gc
zFU=$#n{s_1v)}~`C!}S`7E*r2`MGDEt^hB?NK<fO$o1%568x(#MsJ9w7E^-&!zq6q
zzR!Tyy7b4oefnw9Wv~5GGnTRM2PSl6`?(56ad{VDOm(>!dMYSa`txam>SQfR;L$UO
zwZczyo}Nm-cC0C$C|~hFUs2v-eV7IR?w9ZU#v3B)2j}=fUo=W3^7%LO;6v0)5BXY)
zt^?L1M7=8uW7D39=*Nd3zic{;$ApdW4KmSvL~+T!w~YWR{*FP6xbCT;^c1Gal$1-v
zYrpw=*T7oi_pW%u;lYcqwEvBQc}LjGC7b5CZ>=-lzchWXtanyQ=Oc+`T=4MA#!VUz
zgzBp4IumxbhlG4we4g+EWUY$<%U6Xg?^;e?(TDM#xp(KRu&m?R7q&)s{-L#rVPMVp
zw|ze2ub6n?EUCb-!gX)CEX4hfJN(`-seP)W5$ff$IaUIBc^BzW>%F&;`m4ym!Y|tf
zDGc&2XaE?9_S;xyCL@*fLxU_-kS!lTp!3_(&>cYuUM;j^H^?zSuwbKd?<x6dn-rm;
zN;j`GWJWycL05%XQixq$QmA~F{nyF_RI*xzW~uNuB8Z9`&-LxLTfSKzUI6%{JzDDh
zM45AS^0)D$=n{+F0ivhh63nO=_L>wRBVK5uQl3T_VPWcPVnt~F(%Y3g_#VnL&K!nd
z{t&D6q~ZByZtsMA#r$XG{06f8tLh?Ji4RaC3=!L4E}e3#H5RPKO>ZmX_s410+Gt(p
z#DLt$W3%%_A9H-UCH>GGnD-Y~uSNUK+G9uX1CbwMln99z^DzjPo~3%5DRMJVNz_rf
zLc67GQFjCv7?U5i^e}{0ax1k*pSXhWJqgJ)-GMJD29g`&^9Y}T$dg_e#Qw=`L^rM|
zLjPlW4T?+TK$ZrH|AZUQm_L1DqzF&@<I0}DVULFV#A{#NEl<Tg1K}G?1ygk*!S{L`
zt<i8m1f!7uZx@(5#hL=@4Z)a(`S_=eLm5r&gZGAY0Gx`Q|7^YzeW|5??5Hag?k`r`
zWIgo8w)F$Aia_nV(9ory-@>d~a+bDWCw5D3+;%2f(oadQN`5{QVv=V!oK6^Ma7z~*
zbz_XQuKc$cR7tlFG;vRE>9<z2u<Q@T%t)q~?lmr<UrOSg61niu91jwaqF$sJ-VuNU
znbkm=+&2{dv4RN6DM?Yv8(IRT&<5u={ZKU*0TeZsC}K;st1aa!&hfpIqLG}rr!`7V
z3*_HDNgZAOJ&{nvaCnaKW2+I^O{0TeghGPA3jK%ZMDi|=r;0CqEM&@_EwYXRP(sSz
zGD(g&&@|FWN3cLZ>#3JIp`elnqY{P%7<C{{q;CoiSnyU2Nmj4A*_hm8XRTb#@D-*N
zS6w9BrtBV<Mp1^wh(?&dtzmwk6eoN_6kc!}j77$x(?^3qoU3#w4x`84Vla0F-@lnY
zW5eI-`kt++Y~Ek%BRKb_M=wwUvES7Mmr@G_mtxmJJ_nyq(~ee^7^m3n`D+C$D6uco
z%(&D*%p`Ujo*}Sm0s{^*bFFLPlEG)`N&(EPgt;0PVP<0i$Tz$c53%r<NhB5=@2W)Y
zyMyI#%2&`kLGo2A$@@X!LnKus!nJdj1!vEd!W4d4;10NMG<tIW8|`A{hpP!1z#^m7
z)C|TV>lfEsZRb*Oi`GwMxWzAXR{cn5uO>eR>2Ne|=UM;Amd<}aaH7R*<-<c0oSoxw
z(n%BKz_+A2GV8=ph8c{rD)GYDi;5V{eI2;0znnjq-dO~Joki3&-0R*0GeI3>AvJ(g
zE~4F8mDkK&Df~SQ+CYRYh*TP$LqNbk#UeID?jaM9h2D9ji;!}1dm~<ed<)codf`Qq
z!dt3<rBdy%C50pN@xc4mmi9)!w(hTa+ntoSY4zh5CFQq26V)*-D!{jYJZ@z0(F%pX
zLtzi~1G~R$Ch5VE1co%vS5hpEE%;DW2(mXeAO6<4=C4kBoRlhz>d@Gl^(G|Vmgjnk
zyHQ-@>z*{ptN1WaS*--9V|*+069lg5NXgI~^MVzxFI(>@CdL&-IuHyw;A45eWL8W4
zeoFf6SQJ4(LAoSD|CLLa4MMAFD)=V%@7Qga>JBKvAWbwVmqetObf3T!c9Xi~Z)8m~
zhX0mr%yFCCsQ6>HC`g5yzpCcLPQL!8F5w?kx=OPUmJB6mlwX}FYDi#e@K(wwXT@jM
zEPd4Pcpp>OlaWd>iI>pN8%>+=@_=aLJt5lx1Ba`Yy_BrnZslExr=ju(P0WlNnWcE)
zWfiqO$Hnut9~OqfU%js7O%WEt#<{B1hw{Zs5)1nG`5c7S^aLIA{L%Z7eEPY|saXD0
zbHiVpyJ6fkkeiq_5)<C=-Rtf%+Ff^C{Fc>VO2E7S-pzU>K@d^Geu||owqicsg7tap
z%G=oMPb=ik5AlE7$l%Kw>|ITzgD5n(BDeRfn|a&GB~Sl{P0OU4?)_^ELC;>;o}zMW
z-;HnIsCV{<CDl&y&2CzG?|*9r9|#aX5(pnt*W!V*_aEA%naL@6fY$+*0%#{l8ZfvO
zMNkg^5w=Wp%+>vimmYNVuC5CpC460=9npLC%8Z%z-lUuv(H61CXAia$>WFzbD1h2k
zSTJ5RU&3@fB+bYjM4Y14r`k*76!Qfjfi~de@>uZO+fI)QKR`<f5vZRpv)>rM-zMnd
zAe{a|3q-uFRS@NdQxK*Us6}KGY!oNp^uk7VFU%P3pp`KSrrsc!7=Hzqgc>pid#lql
zZd4D%G!nKodb<-)R|ibJ>|Au>^OQFE(J;cdmNq8%@-*7IMK_syc*G-`A&-HiYDAr0
z12>98KC3zu<f0BL`aPLuCKhNWV<E+MKPR(jZB&6}KX<SWNH4Hk48VVJM$x`>4n~1=
zqEF1l#GM<<m*YVtZ0+Q1${qeY{Y_nRbYgO*ucfqei4;yPI53vfhc5^;aLFE@dVe#A
zcaRi^??kId`zqXs*RkwR9#w!7NO#av-AUdRILk5Jd%&|F|9#;p7nUh4Y%E31bH|y6
z!_gN`ynAe^RBKG!WjJg|XF2w2nvueX!(c%`PO`SJ=Xjzu-K*WdYoSe+n%{M2^;x2I
zOup~d<|b7{ACVW+n9gV$Jzk&BAMv*@Wl1MC`t^GPjJK0<>Odhfx(@tQQuUd)%pWfn
z^^O88UA%{h?+*#prP5LeH-~=RtEp6PZ2S@a_s{TP`+-8_9(spCqwH)3%i~R=^*2f#
zpY6E=lDfNxFhE#XBUvRk%K63m68top?vUHSqnRxmgz|_6QFPYhn|{dz@0ljU8KHf2
zcnK*raaj3{T7Y?UG?y(pb+Vj|t)mbt$4Q{U$jFyp$MPC_t{P_E8ANl;zOM-)K_|sT
z{AIw3fmg(EyeT|t8ItPi?6UO#yZ|Pmr}!1D()2+NL44rw38*^2u>bzp@>rh76Hh%R
zkGHgY)NG04ga8GZ&|-UcXCFQIU8C$v4Vr`RMJh=h;gpc;oqF${m9-48J0&p*Qj<vI
z(Ehe>W>F)*=>6@YC@(R99mN3&hgyIx0djh$+dxF5qaY>ogAjcSQKN!D@kCmXN?p2<
z)pxCAd>A!==tS!~v8w)*WhI5Z6ei{5zX0VD8wO`8KA=alUkd*A&3O5DieMPJfW)g$
z9lJr@?hNE|Fx>d_*(A2AN+TpxD@Z`#yX?bTP7<%Ij27$;POy;&3plc>a`MRHW-4jx
zQ{xwGuiw4HgUtUhreO|>V=B1qseV0DrH!thg{(Gzx<b{G85(XXBV#Qp!m{p=+djSb
zDHZM4(yn(oS`pf^vFCB}k@Z9(n3>oioF6gjP)vB7Y`VhUU~bFm?>cdR=}xsiPPFO0
zHEMpqmvSqj_ULPl+RnM;_HsbrRxrv~K9z64Zn8YI{QVen*Lo4Iw?xDrfhkQ-T-i&Q
zGs8@jS=BjTkr^PJ>uw!OB@%(iLe2zX5!ayeTnURblN1Gw92yh{Mh2h(9uPKB9}{G<
zdFsiMq9VP}rZL{NcX{S!m0nb7Tgr<{K0*6y7Z-UoB%hKQDG`+zmU5lGe~rA#AAJf+
z`Auj04C+ruZFe@!3uaUQ(!kbhGpvBoA$>$S1FC)I5(03*KNj{S#40F^7N3n<C?!y#
z*_MAi@GwDjYVoDh+HK<}l|-bs*6C|q(?4jgcv*<zMuQ516l?U+hJe7lPJ)$9|Inf2
zjbsHy3bdZA%Ry4R<<JKHBp2MiH3f2@FH=uq$srweq8~Oq%zFa-5c~Q>*}0dP>Vf`u
zf52!SP`nq6ZwmW30cd?Fe&At_Al$zFR;ByBy)^#>%(>Y}EBg6!T*#<36$YOT-!iHu
z^)S3r!t}X4V8(4b&HgUlwnxJ_WbRDhX11B7x1yrSCzYDo9{G4Hn_wU*As9zfN^Dg$
zJ^BYH|0^02#{jn1n0?FdeUm;os5AbhzBr8B!V5x7z+KWbANeA@H_;Cz_w^Wf_hz1Q
zu8wcvO4R<-B%4kf-Gc5jm)UhhU@j&<jaW*>vt~K0KIl%@XRYKh2V``WD6_u@f*)6(
zK(a8%RH9$y_;#u>EOGHI#U}aUaL_bYh^N_xwVv!wVd+lY@}!g6fC9;*;GM+X!qlYr
zU6ycGPm1%kaV71dl{Y?ZUN&tD_zLF3wIBwL!dEr7bnK2mA+YV|I2jAl`PD7evuSn=
zEn!`kmIy{Ar_h&u;1(!wqBgl&KtjYw$`(Z_TTTCpN?TZqy8BrA7Mt{gMW(Fj+bJsj
zuRf+p5t0tvDYYsJ<|oFQQ6ELtnSC8q?esAeu~rK85u1>BouJxXOWbQO!i@}r7pN?f
zcZN~C6QRu%VRWPNBsifr$U|A<^Hq(p2$D^~vUTnH2_vptGMqX@)ivt5AiScfiUY=v
zb;~t!h+TBHdk@|?Z)6C5Q#_>Ks4A?tb?!%y|9)V`oE@sf3=jg!=hPqWFCHQMzX1HA
z-hhfokL8#sB;c?o^PDadCjb`)!OtJ-*h9_o^@d>JDACTK@$Z@R;hutDZVavqb-!x{
z%rS1KKM{>6dHUF=AI^c|M%`9B$uPDGWR9%s6IG?4TETmFt6o92#^u+R_D?nG#EDq5
zbeuh09GK&0eZFXXM%rsqR4eoI@QO8E302JgU5s!`uKxrd`lES$@bBiXJ8iB#Q)1M1
z)VAW_md`z5%W^E)c)FNyH3G^BQaA3Ql%fHXiX#T@x}Al#&<TjbPhnD84C_{ieBE51
z_b1kjzochvKb|CgYZ|;}{IYt_FwCWAoxlPSHmc&2<6>{0!x^N0c<mR;{aIfJ_VIP)
zz0IzH%lSUp2`xVgMV5c_%&#`e@E#mO{twhad}Udev@f(E*lH^=5n!z<+e7=|vXLR$
z*AvSXk-KOm^08pU?=HfMk-L)?*Nl(DCpW#NCmOw(H`QCZjPL(r-7`hS*0_rur6XKM
z?B|ng3Vgd4X3h}Oqy6;@LEgbZ$j2>sqnp=K``_LnP~B<JCOoH1!X|)Rfl}kE2A3L!
zDU<>{53OxulR*cLy5>>TwC3K@&JP4abmF2BpZL%e--#<z{rH5qmGpEVDC$f`!ul+}
zSj9GNv}ZlKh;v)sUxR_>n{S4dQjQ9;Gin`YV{N)wp;mnhK|0I_n$0x6{16dwQoG6F
zw~jT;pgbz%-AMz>aqzpEf^7(gx>|ER5G+2D_rxJE^a-wNEfP9JY>ySmjgAt0<71w9
zYjkN_Ks?M456$yML#!V_bCD*<7X<30v!=tW2fh%K-!ViW`NF#V<PSyc$iB@qnY-#8
zGAv`sGWV{*{^m3Wj~O`6$&jiu8!e(fzwHzDP*$I!+1{Hf`&OphXs=wuK>|ABAAWP3
zD)lSGBimm8e9b*a>&CkC_25I5Y^IFd4K?^v>oamBCh-;680PR*i}m^BxeCZKX2<oS
zHETd;Mzux6p0ixkF1<V4jw0$F#;R)VHl5o?H46L&G2VY)do=FrubTfxn^<{+Nk|32
z5&b)A=8fDb?9o5;?%r?H+8|+<@xHHHo45zz82Z=ktRj-B<P6(uDyRN$ioFc1tyBr~
z+E(u`%ZS{@?EW9e9$)1ju<3p^w<TZY@xfavOH`~F@u%5gv49ASR5D<E{y7R2894BV
z`>neUNxHW7@VhiE6cNC;{ve_3p7`&z)eIU=x|#gk9WP!TaQB3!Hy#3=8wA89m``(q
zcbv>&A@?1j!+6Vk`I~PUCKa5EvW#X~UxR52A2U4Dp4MZ9oIdvv>ZT%M)WIt^0MjFV
zb%^ePGCKta(t3bYJkMyh_M${18^WosG(;~m;wzTALK?{Hj_c%8LyK=k18~hjrvxsM
z&kVUg#3)h#m+}|S0EAenvC{cd33=O%FR$#3k>(dJs`hq_+l<(lLIidHR<i*GHb&Zh
zQA34OD3QuXdHpevkd`R`AW#Ne>ffW_;sCXP+=Pf4VE(FfsCyUBgJ!JK(FemUj(KkE
z#I+jxG;8R9Dxl|vNwLn(ECwE|wfrjp6qc~8m@b_5=X3i9JPr|8J7d9|7a!s9PQGf%
z70^7R`jMR+AK9EAq&drLSc<Q^_8HD4FEx^xLi|4n<9Z)~OD^{5alnz@vxl#=wGmZs
z>#3PaP4$H*$>h`|@15H2J2rJ3c!kSj*rzW_t)pB)=@rNLp)s~}ULH?cchVBkFa264
z{@@ZVl`;jVVq8ybR~{{h%Z`L_Wf|AJca^+y!xs_L5Uf-WbFB+aCHEhC%Z1&9WXiif
zjyKuPc}O6XZdx_>=zmZso`D&{vYZr4N;={TXD`f)kgd6w<C_X*b$i~yz6EWLhjuBE
zJO(%dGfCuxQ^B{*wmQ1R0Qei@xg+BFVj@f#aP>+n7WnkH@XjdImSb~9%jI)Xx0V?n
z-k7||XTKi=hFZMf!#@R2zZ2JS{L04TOi7Hj0Es^d!bujiG4gL1yfs0<SeN2kqqF`V
zcw|o&-SQJHB>A-k3}R=Lt80zEp|7?j*N*H$5dd`Q1Y?~*I$$J1(Tr_ntEWkzF*4~q
zH5<@ew^o|#4$!jTPzp3-g#)C%>>~QC8Ly#_)~Th)bBS)&e#i-ibK+^eR8O$~Kmk$L
z=!(7}79_VnB^*XEgA4Hg;=+P_0DC}Kra{C9Lv!Kv=J0apSapn5yShl!vz%1qSHMSw
zBtxz1J`$W4M{Z5v1<P|-&nI-YhV@4MzW1vmwNP#@n)?sG4b)k?vGhE`k*5je+Tsqr
z&Lg}2*N+oOe_UnD%G^AVPWKm{rEXuVd6Z_L@5R|<eVX6aG}yzkINz{=)wsJMO>W(|
zz}y#q*?5)EQ;?utw~b>O{bKQhFJA=D>J8k<93}-kRJHT9S>9QCc6?fjbt=yP{F+6|
z0vCDUui<)cW0>@X%bfY>dEDC*O3Q!4PDNlY4U)5z<d=;^*}xRivI60T8+#{$7FtGk
zX19VSvku=DbZ+1t+y!C7Sz}m^C9aNC(S__Z<}~}7lGk*!NcOm#xUKq@Ff)YJiyjAA
z{p%k=SPc&Mg!SwEAEM7^TnlfHIa`MPEm`o$wDI{mq<1kubx|?I`;!+goEDgy>gv|!
z>Tk{gIsBqnMb02Hy5;)t-&AuuePS~RCLwJ_HeW<j_k81}iD&jO=#mlu+Y?7H#^>MC
zN0}fN=dfVB@h-j^4UR8i4y9p_uq0Qn;+)H&yilD;L+uW{6x8jB#GcrqQ42<r?opKu
z%rG3iLW$wkx;91Sj2RW~fA%`Iv=@06;G@}H0;mGV`->p<eSqP$sn(MhQ(bYWXmlqQ
zsLKt(;v0gz%svV(Ke8f+GprXTe2BPZ9s4nBHFoND0xN#Z1>%Nyn&=NG7&?s4NMnW~
z|5jplx1i|Qys+}KG+wom`q)JCrStO3s*exWp@c>l9cBiSYy1-1gy%YjvQHipp``>4
zDV8;zWht~ZPLjQJaOx!DlVlhC_9F5E=Fk{E)oBTdM;afWCr5-4DvX2?{-moHCS@eW
z^gv__fBtF(O29a-Oz=d_w<_(@1P0^c{c7Yaq-FnD_w%unfmg>CkBHrS)?r0+kaMao
z@hggTSo6Hlt|i|!cLTe9VQR3~B=ekd^wOHL-_}=|rbc%;5u<_oiB7pV1Z=={&fGsW
zHQAe@eQ{!UR$SkU$Ro`}us9~7DrC=bcPHmhv*iSt+tjw<SzBbwZ;!^(nVre@t=e5z
z#q+>9vB-v(j#;<Al;RJhKMR-JN8t2;i>;ua_4Dz2C?C<?9_}Dy3gqNEAYiv~i{R%%
z+p$^*v&;6}zm9Z*df8J!Zu-IxylL1mK!;1itgUQQYrHx+aNGv%p{>Cj@B3x$8wYtE
zn+0($EO>-YT)8i%27(v_1aJb4l(suqv_5#9(48Y-Fqj|YKWb_V9TY-Vy1Q}eQnSxH
z83J&&Ushu^1iDi($jd0~i+m&VYK6RJN_PSuR6Q!QSsHwXm;NmwZbU?KXEm8Ww(6SN
zPIfpZ?KFDC4izoRu0ERp;9LvUb~46p03P_Yy2g7p1uWq;R{%wPWUf34P~JCHUp_@d
z7+Tyb$yMM^wew8b)KcCU`*wAAStR_$ZTtMA&F`C|5u*=^ATX$b2HvDk1^5^X7^*{~
z%2ze<o}agb6sK7<=za${m&I$fif9TySqx}>g<zEBm{8344S+>jzmb<SQ%I%Y$8r=N
zsm-%{<>6dNJWpPpIT)Nz_JR}kjEDOcibuqn6d-Q*K&c9haoFrr;FJ878h!pza1BgK
zoixytDcYTnoJXpGL|V=Un>Y@cH%3EBE4W0TirF1-`eauS>%4s-sormMkh3bkShY6J
z?I2~N{m3g*9^2;5hJLB8H>ZRA-CC<h%TU>(nTlV=UX=v^;Dq>s<`O!k&8*puoG-gU
z^X9u}cRpolZ=kzp4Cmk!Oy-`&FXXB4v=lB*SevE8j|JM-heNDGCXZ$zyI$G6Z}|#5
zO}Eb3ARNPbj_DaE4#nAvdJBzT2XsET3u2I&l&^kV_?gu%;9$bNu%Lm&-D!w-Wv=&~
zhA9ZHgLa@Pt=Tf@RwV`8?khf*hPnG3j?VT2E^X(gKc1E}RKAhIFG^VbAUjaTn6^vm
zqxh$K7iWi}!?V}3I79TxuVBu4VkEBPN^-e<K7gqICFtJVQhB$<WL}Uv#irEY(H4bp
z(;V@9-P=2P^G^$zS*wRk!RphptN-KG@4BD5<P5aNJZoEO)MJwjVy*C>KkcKtF83FQ
zIG<#L*0>hs%h&>|%IBZrm^aWZS1lsQB8V3pAaZ9_RZIMO<Pz0}I)w(YmD@bPvU&o@
z$VItMPrH-#FhC0nXLG0i_y&L(zXz{rbrzZ>IC{a}4VB(Q;YI8K^hEV!Wmw#a*0(2f
zB~g8~V9_3$b)L**^YF2TqU=$B@abKDea&S~SBu+)p?y#A1k*Mds6~)$&*i|bk>M!m
zz;qD6o;wsEXmYC`|FWtbn9AI5jkr-DR1Jif*HJ`%Al<Rt<O{8F1ClmH4WKCXD@Iq<
z>xh@2-dYE~`rr4Nco;!4BYr441P645(yFncB(l9sU*x{iB+1vVTT4xu^Kqw|B)@GR
zC2#@mKvYDXB}%bwNR;RgAEloV1onnP!Xt@K0B2pCz;uCUnd(5|K4NqFr4hyS=sw_v
zOiUnV%de2QUdye`nMEm(F38*|=h<QS*w8eOIf^xM)+K@S5f*A|Wr$feFI4d2Su{V_
z<MRC~mB~Vw<Z^cDXRwHr&~*v3>LB{!)QBs4!PzsHlFy#vtUzg#kHYW$CGt>~C7G%J
z6kUEbeiAd=7=Gw#&-qwV1qs{!z3kWg_pDI6uLocJ_ukD_Byl<GGrGBp*Spnp^B>F0
z6#sk)I~$b_t#g~hge-A~#EYMPDwaQPl&H-WOFm`;b7;zU&+fkaF?YQiK;l15)N<kp
z8}??r_|wC2e~zxQsSos+jH9b~<owr8<O(_r#GwELXdE9MYk63LII?6pFz;(COv8!n
zicby;CiDGDceWJfh?~yj2GoS#d#Tov`hQ4M&Wp#T)RSkstMzE;iL2hf)nAgXg_#gt
zdm=2^EgQdux9E}>C{S5-#^^k8-<Dh93MOJisM*<NPo3A0zvq(p@qfIQ$<x-bs{ky7
zL#<6;&xa7)xLF@G8HFeRP&WgPS1~c8bMcS+5aU_2wE_i>Y7y~?U{DC$g}HgPq~Meu
zz1k=0xMEDA|KUo(=&z}db*5m%Vd01o>^(x3tuqQp<8tSz7%KrN?S%;?;Nsj}rD#Uk
zq7db;P)r_g1A&|qz&($1_8ia{djtyT`+SVpe^;j`V%`It%><gKzsHJ#x45dp^qlDt
z1R@~Q7a&rXNV7Jy;xRRw>a{3RRq3xN?haKYef|e!x?}k<^Dr`fXLN~BfifgR;W@X2
zm??zQIJqb%e;L4YiM<rX5>Xr%I*PAK11#{{gtb*Klxpr&fJa>tUpUv?ndmLa{;9M7
zh}wG+`j*SlaW-x=x=M3B)Hne)YBS=Gctxq3u<|1W5_x}XytLkZM|^$s;Z|z<Ow+jH
zWj4$SdN<|?g1ColSdI0kukj6{0I!Qx^TL|LM>Bn`$&&?8exqp2vxW#V3LR}f=fN!z
zgio#>5bRbz1&_)!LgNxQE`0dKi|jwS4e7kAy4EXNS<u(EidU4|2llM#t%hXugPycX
zyNw%Lv>pC5EzfYp-6GcoJZAM@dM~+kd+Bs%Zpovx?!j}9`<tlS;1Vp}-OfY<y3;}r
zT0;ouMfW5w*zk0G&y_n0xvyKd&~ePN+vwr#XHJ1+Y~2r6aN7Ei_Wda5X8q-z0)0~a
z&x{H_pIZL(<xHR1ULR#%EsRWYjNIUItpNia#|VNd$Xl|2NJOmll+d8ei2Uh|zgKZ-
z8zwK(t81xXHwDiA&b!4y4F9e1-G#8~zq!F*W)yB$%gHZ(p$*QxkE7c~KC7(8vNX8j
z;o}gC6Q@k~hclXsuS2W9sKirw8!PzC**j3}Ku+rMM}))YJ$=40SGk&2RZNQWw*36-
z0TP+peo$9w3MU;24ZaSA|CubM!2h{WkyMm^c->5nGYQe_^YhxN8vh?<lPPLsqf*G+
zF@aly+gUIyuIb+-)|WSZ|K1a!!lR=*e%Bz<xb1N3ch2T*grRFl8-DlJKbV6`$#28<
zxAxv!H0|^5dOTGmene~{`i4;Xw^Qb4;P#hM=;z;HNhA}D2}hyAaS_wXOSFqI(Yc=I
zX;c2PM7gQ$#5Tf&tv~=<9JxKTjaabadM(hw^e9@C2<yuD8d_*@SHzsK@5e%vge+h;
zO5D!Dt+ya<VLjm`!qWK-#m*m@NHhh)8GfSJm}xs%r+9i^^Q9lH)Yo3}09e<jCFDl;
z^AzorI6`Br(5G4;#q~B8DFOgptyw2uYzEoDh)5V{K=4Jt!x(;3QUqwRH$aX9rtI9P
z^|Ai9)0Xr`^#q<CP?)0S8e+jFi+lwTg%I#Ybi{+B*Qt^C*K6l_8%F)BI&c=)3z^9#
zEIUf>l$CB+^iB$dn?ffeV!{yrXbt3+0i;o(@~bKDYkw?P#OHg`%}WRu$4|x~%o-uq
zd_UBov3dHRW%GCG?K$&&t!<*v5Tved5`St>omicqnU+v^hb1L%U4Hf1vh>e>@AdjC
zil)@l9kO@gQQ?}3pFFQA{SWzg%=Us0LVgH0_P=-{k$2>ES42IZnkJMR<rkVr+e`U{
zPG4cA+pnyo*`_YXgv8f*<#SwR?W5uv?JSCfIu+1-zp=%V{r<bJR_h+Vsi|FMZWmn{
zHXC=Y#-!co^Pa+qq!q8B{y8%_%Yz1+1tp<-#gk2cL;li92_bbi%QNm>=XnezZPZkT
zZnP-+{=H})nvv_SW5MYObz`rC@P1V_ZBFh!oMh$~P)xAd^=--G3!R!1Xr3n(c||pK
z^2_mF|BlebPH0ZY=e9NPh02n$R2l=dy}*{Ny8zSl5q`>P^xmT<*y*sMK4Ql093J`g
zUGlcflF(C2MJa?pv%(EaaIWRQ-;46X>(_L1L1Og>23e`7clzT-dukP_&vrTvVrPmM
z9!PJ?SxM{9O8!VJP5YzyDKJ<j{KLbW1abu(=Op3^!v;%=X#z=a!3sJN>sT1M^_ce3
zD=;0;?)!8)?WgW}gClp2|AY;J$$V{d=|XeE)P~6Wf4hvI#!I$TK@a`;p+Ea%N1ZJz
zl4>ni+B#f8U2`su!<!qEz9kb)WdoN9*-xQSh!TM+_OC99yWVcvYuXj(vB#x6yV<_?
zlUbgnYVk#?MK-@bD>-4PY}?JftGN**dMQ(cFcWrc{h`OvAvvwbT%8`A`}6F=qJtnj
z?DeEv(*gjZr4bqAeJ|Si(-d3@RrAlr6eyDvzCpQgB>OPJKN+UVBtO`nwfwg@zInL+
zwob#(%DC<YN=A?*;=HhBy87-i@pQKFKC3|#n>63cI6EVl@mbI5r^t8@qe%#EQ;MU6
ziHLPS9##N2%5xN3OLE&HHs0Q=cfH>u<oe5DN{So9zh3`3Tf(6)cGLP{@ZD3LPetW}
zj}VZ{yfCWX7%rn*sgPlaILM2(Zs1rS+T-Tr<RnQ-d0Iq%Ti}#g1i}%WCXE96h?vD-
z5sMu*U)xdaD2h7!ZJ2wbU-v#XIZbFYAoKJ-kx!lnihR_eg1>0OOtYyi({j`dJfZ8d
zMy~Bz&(E5x1orM$otLbR3#w4OghJ5+c5E*;anX=M?wOY$MC!02<nC|(OM=Mf?G}B4
zWLR-WZU^q|K@F{PQIHhCeBNay3)Ti&ho9Qs!|0g$2qLRFB`#>$1gpGxkpl@+fy>Fa
zzcRweJjv_>)t#3CjsqKOz|`fjuZlgV{;aC~Y8ZhCJJeyhZG_+HS8|87m0}~?0G+Pj
zI1xD0XV%9twV6S8%1#H6$7ICf)q910JXKhU*{l-9anlQ2;bUToyB0GATSOQ--@#+M
z8tb|WkMIt?SD=>BcUQ}h107@6kb!youB)qM!t+|AN8aBFC_ar1zVGSIf2U|Yp6%at
za=zjz8(#IorOt4oaqDb3HMIJ;w>scMMeLuY*_@VB)ZLZ3)N+==@bdHpL}A~DcCV*6
zTOr*~r9pYxyI}%nC0Da~cJ^>M(bFjT2+T#%nal$0a9_&dsN=UTt6^LV$__q1v9<F4
zZS|G#P>_#Wj>2^!>nG{3-{0Fj!X9jOthqmRU{kK6#wB<c{tB9>zEi_Udvykee;Q&Q
zeAaP<YDr_xqRE@a__m+FFDsavKhGZN>RXUMxxL%;eg!!xNql^>($VHv@Te6QoIAs^
zP9yS}<Y5ICi340blB?z}|3Fk*uFkwWaA=F-HD#m8E{-2A5k6D?+x{e;O3DONIB`-H
zF2C_q`=3yyW9ESJpF%xTbo6{XPJ8SK7%GI0zP|O+-Q>srN7HwPv-!S%--+0J?_FDs
zqH1ro_bN*4JvwY5=s<(2S!z^c)vj4HDEhTGwN=#?JGLkN{g3BuUgXGe+&QoNygu_w
zUaiAU!si^He@?i2LXrH#tdM~zWs|K03O;fiZerpPr&syn;I)`kOEF2UMqCcCKUnQY
z3^}8?YI+6@9S!i_!brBN&ZxKFsG7AG3@7Y-&cFd#0nzu;uJjRn&DIJsR@u`o%p1qx
z7Ov|Q_!T@)aeVa$S>_;2iz1y8m-!-TL&WvrFVV5e73~&9iTIR&nCiMFL8}l0i79S6
zZUC|;D<56W#n9MThXUg$)PZZBUFvB7B*<E8YBkMz>l;v4fkpzZzpE|mgh?0=k<VIi
z+l_dFk4m=J?7_CASevjyD5ru>2Qs;-O+`3495sBvvG<BnUAH$|=h%sFI~76ZMj;xn
z>-~UHj+bOm)eh1}CQ49mBVV(%hE%&bQaplPcEv*+6y8=fLY(Vaprk={8kMr-8`s!t
zgKU7C<wX`bvoc{HrLs%l+v}o+(+<|h42m?3#-2QV;4g{gnSv>6U|p`?JIKU)^)`GX
zMxFNv>N$&soRduDaJl2}&L(hfLlp^9G(u(eyPMza{mJs^$hT{lj99#FoJK1?_nW7I
z{z*Od^O5FtB^YG?uQ{jXfk+|oOXJB7*CS4HBBwv?kP6Xuub$*8T*P@aQoVUor18XF
zTCwA>+rDY0bw`3$IyM2UU=4RT_+6}2UT~HKTff?`$UC?0xvZ!wcnXEUk~S1?xq^--
zz(kyiTMxyapOyBtY{aKQzVLaoUyi5kv*wX8Xs@1*xhSXcX%3HwMq;s?<>lk=?qla+
zyHS-P`R34j@-}t<lTRv4PKO!q%%K{LHENU^$2xtTN@B3N0@(sHy+-(s87gN%qO6lc
zJMZ&&wCD28+rrd;K^;gFeG^M(Hk|(9fd=qoI|;g$tvl-gsefYK`t5|7pgtTixkBEt
zoc>U=cae5HA<20G%)b3Qapu`F%p%g>120~@T})(cDW4w%*yxQ`l4uqG6EbRRQugG-
zJ;kNrcdkISyKWa#i--aR`Ea=vJdgL)pQ85&ylp;J(r}2vno^ZNv_@}233yQkFKsQ1
z6tr=F#=WQ3<>F6dxca@ygCF~0Tz#^@0!-#ib3UTu<)l6sbxK&6QQ-fRpqv}rp8wKj
z$(5&_4YEetI@1shKAlt|Y*DEZ%xfo9pz&$Qoc^E;;ON$Y_ee;WrzKnwt{`=&9fU(`
zlnK=tugr<Va?)<dbiW_m>#cq>)mLZYyAIie0u(|z9#UTcHywbmU?0wh+ps45OAz%`
zXOO;TuP?lRkt$y7p}4~(A5!$oQ));2O#zVj2|L9PW#$A%BL`c6{^nlA^s7r%BGF5g
zG`!5zYob|PXg9;Jx2d-27J4v8dw6+4??XWfJb!`;TG8QQl!PPLK`gf91A?av{8WrV
z18@^MYEg=O#Pi6)%EV=iE)LASfQBNFr@u-`yX*f;8dJiYqYvGD{%6a<A{q{vE1z9Q
zvh8l^u5W7RDAmvpcvixm*U2j~iOdhrhV1|QBmZqw=>0~p@z5(!bFUhQdsC_PUTXAz
z>{=o_cl_cn)Q7UAd}kqB7k%D@bxK;bug=tKHAAa9!un%`;u~=g2ma!f%+xT(8T}-D
zMmBz^49zL|+#1IdIjnUEO^LydhEeoML=o6&n^&ys@H*14`A3eOj_3;Pjp9%H*u$C<
z`~uCu{==4qce!UqLT_d;?2mrmUR31y<8LG?Tu+t7JQjMMm3TN>=G5Qx-%$Vep2OV+
z5whA~cO1$F<NA*iBn)$9QwB}f3>V(!Y_L3Yh2VgRXhojg*5MsYPskvSja0YW8X{^L
zsd%AqRF$^uRA)Dg8?kY%yB(E<uP%qjLL9agyI+G1eaF@8+DiSOA*l@Db`bQGdt5dn
zI`_<|B<IUF*x43VH!{ER{QTqd(L}A<k*2YTAk$q%YYa!E--fK&&fK<}`wUpKmECg>
z-AF#?_HS%f?B;UMWos`k6;i6^BmDj}YR<PT<~6lHCnv!<|C?vob{;DEenP0ae!(<t
zIdswYe_{Yic(p9oG#tEB&7x)VT0f5VoA6U26N%18D{|?k)r*dJp%{ravL4c4x6@AF
ze)qH}R++Xts0WB0&xl_+mzD!`6=P-%P){aP4xpG3$r}1yIR@Hc>ZNtzM4}B7Zz7ZI
zEp+a49YRjDCtlZ}UoCb1?Xd+IKIak2a^tW>Np-gHLkGrN;Y0~(uyXPVm8^mS^##f>
z%r5;IE=4}#<Czpx75&l^1&@E`k@1@k9xt^dJ@qFuT=NF`{Q>2Rt@zQ4UkVexlt!9U
zj9zNW!~ml3y9v(gkV3eLC(}}yL)>U@$v74B=!1?r&UOQP6UTID$ev#_|4A~&kTu)<
zK~~9f`nc{gNE`Z^L#)d8=UorSR66{<u@9E~tV<jsx;196NJPOr3$cxE-Pn&#s1NxL
zzli(TOv<QT);z11sywXCh3VttiamA-{}tVyL<syB%V-uGTNOmi(5X~2g;5s8`Fm>P
z8(G$o>&F8{!u;wQH>wqs9zVt!k0!9hn;p$tR)$q&TV(#FlW42yNoEuK-F7JE_uY+u
z&*cG;FCm}K^^M(l?xT$p>czu;8}un(elpvK{JoKTpKibV-Tioay2JbxFJ83i^w<8h
zTb|>TfsQxHGjs{(_Dwc3Sw@BC$nU3lHx|ES*@%W^L#kSaeid6sp6!2nMm;JJQ2V3l
ziKxiN!s*+@X0+W6Gv-1l!o<{YUEZT%L3gXd^NEUd-(5P}JM#49o9pc6`K8@S>z@0u
z`g5h$0{vOInoWnpn=0#rDQC_2OQ#lO4<ApSv446oqj{nTWK`Y`k&YFRb}aAzE&i?)
zasm_Thh2Lq`TJeWS_(~D+9Z!XyZQ2X@p1cbGyCn4ET*3Qy0wVC>k6Ga)HJ_%{W56(
z<CWe|*at>|P>ybVhLvuWxHkLKot?rL&+D(g9|25-4{hVp$7VwM)kI>;70#r}7`g9`
zJ4Jte85;^<OK~N;`^K($aHqKryPi&p+&Wm}s&#WdSqsnMqhH>$z$ST@8t&M31L~D>
zkI?O07K3~pw`$a18c9DZ?Ox4eTGU?WCW?angMJma`>u@H6>{PL+%9wFN9B%nphu@|
zc~5r~Z>P|5CMh9u9%uv1XxLKk;_f8%!~V65?n{LT%(mCUTai28iTue_SIb3<4b=S+
z`TxQnRJ=9MRJ)!Q3s-CFFv|TsP<8$Bs{L}pNBL(GHcZPRD%t<$lud88!5p4^kct27
zJC^w1I17?HDa~;!w7V$GMrk!(&F<<I`!>!GT&5FIy*39*u#4WjPY`ptXJh1H0RMsQ
zd9Z&3b>^{5*1x5%Zh&DcV6=YfMd2c#_wT~A<f*iFWHW*@wlP%^j4lnF{HM(zEUlxk
z7ZvpPZDxj)o#v)feTkwmg|vSZ`|m1whQ@*MOhYxw?}SII>IaLoKNu0z*b4rW&Z~QR
zD8q?IgjPVdf~z^K+6&^`^)|owM4wSt>o_C(nPy;;G74*CtdBgc=?Z5AaL3ZfNR8Bm
z034{8Gh_})+YN+w{{6xGG4o@cnE~?ln{vD2*of9wcI0+6ED8^wT7648<rnZ4`uHBL
zQnk32^Wu(|`%hOSCHE<f9;HbT53cunL6ZB_*vdSLf*^k@=2o$6-q=7*Z<xNvGpeuI
z%@k!B)G_(QQeJWfk_wcCk=7GdK;5rMtg*3HXOFLl!h`;crs5*t)p~qN=#ImTYni!3
z03{@Sprz>Zgmr>Xm?E4&dEGSbNn5DfYXP;JbAs-qbXQ$-fQD8FeB!Qme}Mgy=(%uV
z1S_Z1M=N2yR??U!8kuUc!<k8H2rXTQK+hu8aF}@t-_<FJ3oFy8(%6pK`Guo$%R+{)
zM%4z)L(&&{pPYsEJg&j0F(GWe2-`41;wc1hw;*>8`vsfXnw6S1TDMvE;oz<*;m_G!
z+~^l!LCbOrjyS<37Ks6q0X>Ehdv<r7qT6`J!(zdou=OeA6?2A=g;svZ0ub@uV3)12
zeM@t6dd5RqU~w|qsM5X+p9FVXd*h_r-ZseF#e*>Fl25Vvfx<`8Od*;@_V(+Fhs1dS
z9k33$=6C4O+-2Lt|48nXZZZZIR7u}x)S?JhP(pd9^LXcGk|u5%64pFpi-ZwMmUBEC
zf*kA*uL;-I`t$6l?G~rj<(7~1j+OFtCNn!%agHOHl^5Q2gcpX5xOF^LG0e@<YMD_$
zKWlgI;p@C{yBnCt9nIN}u<T<Y&9%nkz>N4&&uetNd%JMlu-}O>l2r_&sBFjI*b-Vg
z({3C@WW67|qgOob+@Jh#;O9Qk`@2l!s;<SD%L^%f4FB2NJKFVuE%5FbX76uld+HT*
zt8e(sFmhIbbwm5U9P*<~RrnR(LRFP0KJ%A#(BkdN=Z^KWeg2uZpKa@%P|Z~lcN_bW
zAmH?B)*B1<4=I9k8tu_NjX85+IR-^JiuKJS5mA*_XM+vRHZ%R?;{&?VQOUkZLf7L#
zP=TE@*TlHqz&+^6_X0IyVqcvR>x%Ztx3IAdk)XowzV>gM$4egyZk@@n?W@)*ss_J!
z>q_XfOZ`cAPaIo6LB8|+--IfbuO(hBDv+~0wf0Y;YCc0G?@ui}(QnIDH%2$A_&+l^
zd=uayc5r0K@#&>rHP2`+#|_nv$@nq>aTdOJd^GAUA&xQy&{(HGM-KqKUGYnPRuBEv
z(Sd7JZP=6~6ii1TTT4P?l-=TW6eLu-zz)2j`$DIta9}w$6M~$7fa7cgGaaXX^5NlY
zJM#M!<O8o)u0LGH7na%x6U;KydHmiIB)RG|v5x!Y5JiUjbd(5`>41o>#38a;Hh%+C
zs75`?*r-*JHhUAntfuC5uC4`pdPt0dkXYa;gP#GJ+$T}(d^#vYyN?_ckP-IeK!QOE
z9~_F+3A`fN8Au}PjQO<`7c-+dGkESXCDd?x5es|WpxgjGuWgBD9tK6}9Llz19LIc8
z)jS}XA%#^*Da3e(8XSPA=ZZwcdc4@RO3WH*wIGBYqX_fD-`?Q-tn)<Iay#DXHJ2yi
zqO|P#7!;_vY$arU-N%0NY|G{yeFxIwCS&KtB}_Z9#T%4zscINLA>$yEJWZ##MK&!g
z2R;9#+miiXQpwgzQ+=8E@9gjkcBQYXEZII)Gs2;>Y{uLa*NC^v1B-`rZmB-9B^>)$
zmS*m0Wbo4Bo`rM`hffnwPL4tdOQSz1=yk4#E$zFWB|IUWXm)r^>WX^4I;-e?_e%<Y
zp~LR1G@+nAQ`A1guHUWw!fEK%unGTe+Gl|=Io0T0fyeDsN5>576MlvKRL@s8&zGL?
zwY20!QU|5a;uyaJa~J1U4=kRft}QMOx5TylC(x(yPN>Z!2%b!+z(Qg4c1YP;!1d&D
zLi=iMb=#_%{qtqIM^o|7gm0g0ym3=Khn3!iRE1<@+Yqo7U0?dbsABngm`lHO;b~zK
z4I?M!+KxlqWKm@$HzWA%{Jf*)E@d)Rqfi;OzC)EDr4F#!Z_d9%>^Jcsn8!t9iBsAr
zjyX;EVH#LDeBVatS>*NR)vO+ii5aM9608MHFp8XQs>!XKe1>F<xp?82rp%}JzP9I*
zqlaaM^ky~FQeyl=n?<v-*b6{TTfCS00pKci)pHxOui<z8p6xHcr3WB{tX~VFcsTn2
zeL2{O=){EvZsBfN0^01fWq*Be@(F2hxL3KQKMkDPg@55b67Ve8Swq{*;4DHk8!9uJ
zrN0%7Lk~=r6Ru|$cg)dVp6jxcS4$+W#CzwlusIh5NfEs@+Wdbn%+Pl>FOux=SMh&&
zDZaPRb#j)tG@y89qVmJe`ri@_;<q+GfNg0VL>SHqd%Hd)aF!TQ;z`H-JK-fNG&GLP
zq*m0Yl{JcxLkG9ETufRc^^XQpSLJ$+9pj*;MENPr0;NGl!o`^T<39R)5MBa?!%GhD
zwVPoS-jukg1pz36gyL<gjrZd<h#LAH9|!!b*IZg4S8qdo-l`G)_l+YPUOa!Vimb>t
zn}gd-(<j^N;W_PDv2tq2qyoqMPRC}J!UlqC=ButHfzyO<*JDaJROcc%!f=<AoQ#Ip
zey}q@0#ZLm3-SJ}e2R;JdU+kipi*j{C77=fZ_7UvPnNU;49-fMDDz1<2;NkYyd00H
z>kcMnD#P=fzray;>e`!p`T))WnGy%+;B;_qfNngx)b-+C<}7|pug&L|dLSrok;z}T
z@TWAjc+B*h(Ke6l&n08Txeiau>&$t1<k9G$8$8XI$ifdV2k{G+e9DpnVWa3&J>&{2
z?W`=XLcyc7=>}Pi0lfdY_xf$e=Q0eUlHKmkZWesLBBPqIOjGk+FnU8sHU=u!wdWvL
zz#&uGs1^45v(JHLM^GGYK=*q4$zqvLU|4XD{l2=>pC?Rel4-PCNV|hPG?sDo*)^T3
z%FtS1)p4FxJm^6S&Rz{2d`)n1-R?MHBoF#X08rv;UXz)EsqL;$-ZrzEuGgxyV3C%<
zb7##v`++{=o%CJ;)aVQLNQb)6D6J=$UCZ{i=#R$6)kCz^8o*LluiB^*|NF>OKi|43
z`TX`3ZSrp?2CL7Jt)cs6%g=T@@3k5akW)UwW1yFEAC+h5SYNX@mo2&REiUQL;Mbrd
zL3sX+dK<LedU2K%=>lK6LBFj0f;P69-etO^anZea4tQ=hsV68Nijh=fO8u@!f!LN$
zLFt7ZWq4gcE--2vu|7rS59WK1x5G)lPy{yW%)8@)Qu10#;Z$E-DFw&Yvo9I+*NkMK
zOd>Rp{M-q)vw-4Qo=gFaO|}R2(PSpb6=l>6o+J*joTYoVV4adL^JnaWMssLD?O$Ae
zqiqKM%~YsNc0^flTlOs*Z(iMVR;Hq70ku{CW(iw|Sh+8uCKvI~9J9j|P7Z)+P7EaP
zNA2u6pGGjBJFBSdO5rX!JlcFcg~9{R1Khd*#L>*TCmj&e7OGC7oD}hN<lH6GO+&&T
zl)t9M9pUR?Kt0z9sxKl0yMB!{K}3}^KzMg^nO+c&Q>vWLTj6V)JjabXFCA>Mq2a#Q
zW@i*y>>>SD`TJn>rOq-04k}S)hbrYKsdS8ffI#61;Cy4Vzl3vO3M)(K(yM{fO09mr
zky~JsW72_y!hEx514okp`j+>2i`_dGnjL-hH7~HYr&2j{iGjILI~@)}ZVAQ`Iuv0{
z5tGi7TRTrO>@_8yQKu6MCv)V`s&0iUO=Ws`>BB)Q1nAE@$f*zH4oOCD0stN_AszDc
z-0Efv+!90)u8F(-i^RFt0jf+Q4EKA*K1<9=e`X*-CDwRp#qkjLGG+;Xea_HF?N2gI
z;LRjd9N~Aa)<<#h$|=e~$NF|?!$Nq9L13KZ{a8dRvEeq$QL}^@xrDpt17q{>wq;Z-
z>BCzINwx#!AQA-05a&+l<9|*TEqvcS8itFUcW6g}nu4e2$tN6kN4-g@6?h!MJ$drY
zj?*{uiiZKXP92sOX9gR&tGO4kVHp+Nvsd(0VM#9){|-za7g9t2P-5D1TFg8Wy3Rg}
z!e$uoB3NsA?9SdMm1vRv*uChusKQ)1y5vFf@BU&$D}Z265I0;T^7a+JnR_7#SgdPE
zE|`U%*>OmeoJQ<m^J?j{W0*puIgPK=Yuxp`ulcRvXj07|dAB<*w`DL3UHe|OsbhPL
z@5DL3knJUS^w2G-EwYD~&3W|X{AWf|PPdrL?f!q{mHmJ7ql4nxj#P4-mb5m`qL(Az
zZ+$)wxIH#UpVwXeO^Vp+NQZR|M%BM&Gnj;4wEoS>TN?ca;=j1jMLP(r6}Vj7=9ERO
z*;O3D?=ftwt%)|*wk)&yIV^WaDxUrW`eA<tR?V3QBt8;i3W^o#gkRHdq0aXT1b&km
z{V6p+vC+ag8)Cm3r=L$XBu5~V<c9TCibbDD0@i!$@%YsGpz*<S=ydvWhGD~$4v%R?
z<w5)5b1ydcp1g$Tt&fw!ycFVe$qVqYhQbcok;Z?Z>BxvhrCQEOj8~?rx|&R5?LQq+
zK8SlUmve4YcJ!xTLPOB=cAr8R^M#E3w1eRuJsa^r43pN5)Vty(Kb$-)VH@xi5lv<K
zYSK`QpeRTnVak{=RfreZf#)~kcAS#+%Kga#00s!;O56A~<2zNA7k5-SMx*j7Vz!Qd
z89eGn|E#p7X$RiEP*wi#Iys*Nh{t;KUB4-cZ^;u!HHg>Ahi~7XWmwy$7bw;mctzPH
z4I}UWYHwjNPbn&mQePoub#5gnO!16|5nMXKS-ophtpPE|K&0eFN)2uR?wWG>f&@SC
za5gcWFrV<$7rFPL3i`szQz$_uN4uFvUrR#>DOQEe7BqVTunF_I6_G;vb?|8=lnZ06
z!hmo0K2RSiMDK&|WxDf5g<k&bFdoWpq;br8X>Dt3qNFt<wa+WHI)x&>EQ)yWSE{Wn
zv80vk6-cOBFJ32yi~%RJtGzU*_AJ6X|6`OERU^HB+EmCYX|Mi@F31I(0R2HBr0nFX
zzkv;~t*EPW5UcZeM8<*xuhZjoKqMFL7MwUeOcHm@d3Q(h8ow#2@MTUMBli@!rNdKe
zMeaT;mnmJ+JL~^x@jIXG(lCUi)~SjZ$A2zKzKoYNb#StfkD9Ko*!p`cyyd2DDwO<p
zOKK#<_6a=uX6fp?UkfJUHr}7MVS!Lg)l&eKJfHWXxiEtZMz!@$@wkHBuTIVA0@tv8
zGN*J>-9fIo&O4NIH6Y?6>d5|j7>A``!J!NnVf6P`5+%;#P+A9O2o#bB=t+4^!ST;-
zRz=TRGTOJk3|UNAd~aFW7S<IH`1GoeGVN+hO1g>u&QC@Q18@1X!S>=#&SLT|$?vx0
zs&fQmS?<;&<Myd!E#z6xLo2coD&jY5Ob=ha)asy>vf|kM@M@yV_^%3<FjF{MYD}UF
zLXXLirxa&_9)f2zCVn2mZ?RvLc_G)z&b2!-Bf}`wZ#fYmv-4-|N9Fe)Y~b!I^?=Z>
zc`L=o8`s@Dml6Gka<3^7jHa7aw`*-8eY<5Ja`+BIB#jEr(vx6Q|1hQ4%Ke=na>TC8
zA+mYTBJBHDOaUVfci#eIgq8<Hk7d5>q7WNIGQD&h`s?Ff6oGZc+3(YAbd)dNb_=U;
z43(kc?+U1*)q;K`d5XRDcQVrLCc4z_9A!h%S6qxQ{zpkjYKA@=*}@!_{gsaXtmg>k
zDr;z4hV4EMgS>i>byE>C?;|y{w-aQOv$<yA!0}V5>JqCR@JVQ%)LpFiL}>}Zyk>)*
zDtU_69sbvv2@tl~C(Hgf&P_R$-busZGQFFK)e<z8!!NS3mt?l+kQX47$_aBrcas%&
zY*I1JwKQEc1CEu7Fb>JH7t@BGFuLieGs+7UXxxdenVSy$wc(6<i1G)Qy?5uAS<$ye
zFTljAF947m!3sT%(PQEs90rfV%YK1PLt#g2=%7O(uG%2$0Uq&*sA|3rAwx~RA}VQV
z26zVThX87*M2BmH1OSIB5xg{RDM~gOm#o1v!#jL_NuaFO#^AUPxl{0mdelqO&E;3Q
zUjAv3u3`tz_eyGfpz@-Za5l&AdXwPA=6;+$$uUd>tq@`#Id@c0d$Fs8Nb!Fk8thy)
z?Z>QC%E#(W*MREPrfm3i!HF^?6>uFjymb(Bf<}n>PKb>0UClqNbBFM+@;SZ#N$yd&
zDu@+xN#GX|a-C{>jIOX`DkAy2uaU#2pOl*hQ*CI(9*CUjK6}=wPr&hHtffCt_pxVU
z+?{tY^bV`cYZqjy#fEZ_();Y-=3e1ZF?gh=_zJ~Vz4)te9&KX-NQTmJjq4pqdq*ZR
zU_P!#X}Z3rHcMn`?6p+ds41XIp9m`xH=b>zhW=<wDC*k(B`~p)Z|_j^!go6IqJ_RC
zTf33d+g)9LoyV{4Bd~fZjKz6hN$RQ)m2C2tMMn0=IPiXYT0?2kC3H`DiF{t54S`@|
z0KKeC+4P^?fARLk0yUPq3r@Q~<K#98)};6ld10eyXFS6Ssrx)nS&zQEw#Gb&qBAZ_
zi`b<G6W$s^Riv;J?Z2<MVU8}R@a1b`F2Bzp=wM@Zv7vK?D}GiDYPz0aSNNv-{$>v6
zF#<#-V&fB{NuD{`sf&`5W*bA_3-(6}&)vuVc=Gbu3N{Ke<EF<#*tpHdiF;1<6_D2r
zHc({JTL);iwXv9wImR!#OfguAKWEPsa5w73UaDfA-@~h7udpqHp??)q_iB9!l3X6<
zh<x~utBa9uX!5%`1UQrpL>Fp_8l2i7a+eYXf2~XZ8JF$CS<_$R?(0(H1no}m!iFx%
zQQ#UP8tF{4uO%8$WRPE$v?pW$ZViOyoR`4>&S9?8sL?DquGeXxoCm&7q)fKX_R!y1
z1Az<R5cd=70q`d#mlcRJrLQMXQod{_f&c1OTb!^-$1A=D@WDHf1(4|`vkzCh!6+*n
zqEEaXl?DJfkD`?yk<c=~>)lYHj=q7M_$<IhGaWg^^z|&&LRqRgf~y9SpX*!|rKvI<
z`sbF)AE>f3@U0kqPJRBBKcDmSp((jARRjJ2D5a}-v<n~5_i9b~a1?)A+5re6C%WX{
zbr(-RNdHN6OQ!$ZjQ#P)6J$GpMBMw4{vWr0l1S{fxwn%Z)Y#a%rs1g`m5JZyFx73*
zOE;hY_*M_`%bB-LnMC2dy<Zh448!r8z$tidNbio$OCgs)`|g~?UV6YMD%l9_i)a@=
z_X<4&gkZ%bLcG~x9TQLpS&P7C=`QKE-LW)aRF{~swMXcYcM%PviA!|<2VC$6^fQ#E
zzlq=9#l|jORKYNsBT(Dc{g9+8g6i9oEapYxyNwv;^)BMUS4s{T@8C4KM^$xNj+yEe
z%qXq8;e#L?&0Kz{dgnqlr}9lIJr06}`22FuhLC~t{l2O2#o^S|Jq;5xrRbyBq>dw>
zyz=itQMWYY^!iM5%oV`HcIzKAdFLn{=$|Pao6`g)k235Um}g;KDs7=HgloDf8ZfNB
zN4)&K{|0l_!SM7c^}^Msmoj#uutZM_Vg}Ad7A}l&?a>HLw>r6x1NY)=UK<e%+g6rI
z^HenAvKyDdwX26URCvVrqWW3;V+^@plE|^mWf?Zb+Futdm`G5yasNEixm=?3@l>8w
z=Uu~VTNz2U(%!*e!bSz`QI`gG<)JrM?thV6uIzNQ2rUZbu>lx&*@kq7kBVs2gibbP
zy$r5?;60%U2F-!fov;(x$$>y}VVA`vJ&KiG7rDjC`Kyw&;{=#2NK=-vJHJ}Lb=fna
zaO}+bs%Hm(1gGNqVK_R;GcbfZJ;Mzly!DHH+FJ^WOlx@IC|9&o&002vzGA*5k%;f!
zX{rC8m1>lM^Qk%4{pr}~V>la9O?8k3dav~5lL6fj;_WZ&D(9^wfQ(+uGanozz|ct~
zm;_4vL(t&GNYOQJD)6YI2pWOCV)hC#U>lloryN*~$nnI!Az`LoA%)pl&o6#wMNuJ(
zqG7-UDx;9h@qloT`~YP2d#Kp!+axrL;w>Y$D!jL84wyUu2%*IJq#n&5#4LWm0X>zH
zPY-(?lz!s7dX56@^*bMz_(xy<|11DbWEN36P6A!8(~WT+5ts<N{0Es0_BVoH4@GJY
z)%indbc)7!0^tG*@C2NnKc=e}PDokjiK0c0Qk4`$`mc|&h~I_ke+VB&Q8*>7&?v;L
zpvRoY=lDzR1nSdfob5h878F{&c;E4$7F+}gt?<f&W?S9&oM-yn0zCQ}6e_z;M<?~`
z%^>A=2mXY6MkL~y_3cxN>X?yHXKnt<-%*sWX2uHt>t_c9O;enT#J^=e70d6%;jZb8
z%^|WmGOQ2!Y*E=J&z!V#Jx^9zm(`Z!eHJ&~f4E~||GklvMXakQM1_9qqoAly*LaS<
zXQMd-H?o&Y!(#Fu(Y24Zzx$jpMewd;|9jH>`93{`ZB=2Of0E5|U9`BWU~X~yiGBno
zVZ;|(xCS5(=W3m-rE5|*%bG*B-=x!|9o>#w&z7C&In0Zo2)wh>M!QvK@u$-N33L}<
z2UW_8M&z4W-&ygcO{eroe$OYGKw2}Ge-YU(cSYWD>Cfe|n)+($5zcZ8us%-YE6vr)
zp{oysCu1U4<@Oh`<X_#_EM<J_dwzF+W`z22z+gPh3RWnRZiQg<@PnKD!X8yNuXji5
z{`;`b$ac}z{`8d4%`1Gu(#N2FPgLy4JLXD=vSgu?a+sVt=|*D@1{vXYb+(_XI3~0D
zyP-9Wkz1zCzmON&dAM2mc4K@(3XSt*anRHFYDiq_*1w|$LwJX@`(oQRb*5@Or*x^x
zv_x02CJS;l4dpDM%SWUqM2u9@N?qO85~Wz29u5kJTecngi^){jq6+J!MVuF(O?1;s
zo~*?wwQPXczNUmyx;%8hKZHFyCgk;s57dQPChzNxDt-Bw&?ywLS0L?~zQ&d_k<dR?
z`8m?2nN;1pMWn{J<K?n}S^tE}p>`W!d)KNodq0F-qiS``?h3;<T|43m;vS0#xWcek
zbW8`KvCKL1Mg8<`O!cjnB08|hq%va*ms#Q6t9dd}^5{s<QxN=jNpyKaZU_MSCaXe_
zIs-QT`qb;o-sOV?c=BdMefx;6lB0#{+|*#BxsOfl&h||pLc9QVlU3+U#6a5HWwb6q
zlsclzp}>=_W$aCnry37eLTMYQ_*t=ikCx1%86h7R*t$?gE5aejk)suG=L?cHN`Q7;
zxU(F7zsoOB75x>rFh#Xm;ma8jSd9voM`7X*Sba)dnFx#~c#|F6tN=-}3cZgbp*2+4
zPjOXb@={65#k}q&8c<RRYvAb-j;es_(WK$4e-@oi1YZ)gO&BT?F5xr*iPfT9f2Ju`
zVi4ryVhj^%HOx-#C1M^WT5hj<@i=f$iBbGVq+0_NvH0bm?Zw<nmct8mDh%)-?}1~S
zOX52-ck>Z_m3e?tHt0`AKvayFCPO2L8Q{i^QUIGl-*LQ53LK1`v!0nn5Eit92Z1cB
zInj&wGr)6Ho{{>?D<=<cDr}V0$_~$kNM|NwAN7XuKFG<v3}{D68UPjYr~d)@<H&~-
z{#0{sM_eu#7c*Zz|FM^~h`oc+P6_<@u^;4sJyVxeq*2!vbG0C{i#!!pzlw;h6H0>Y
zJuG(L`<+VR<mCQ!1luB8Bsv$r>3?H=&_#YQ^`-?&xVp(-+wBiG%GWx^!da0#N1Faw
zG>SP)9?21C^s<q$J|BueX*wlodHtVSx5Y+qYGx_NH0u#6p`pfy9{LKvFp+cX&#8$t
zC4|kk`Pvv?Zp^UxM_w(nmsp`K^RJ#I9gon}5^y%p^#Si)wPn;OxA9&P+h|IJi!=@W
z!VAxT26{AJUXO#oDz=uNv!DMC63i+IFo@VZ5?YkQVh-^~SeVVT@Dq0XyH@Q){=-!)
zPfD?V%yKoHosC2+KC}OPpmt%4$OIHC>D3KJ96T5j&gIWsbz9+~nMKTr0o-GXGob|7
zQS=!LEm$raN{mbQSPwhteD^lSyioh=LzT~Ip=;l;+%R&?;mv2m!l^BS7KO05OWst8
z7GwX8lx6#3C*Od(2!Q(}&du+7`e#tZS;ykhDcFB+`uo(`v0}Vk_fA3z3=fx~bY{fP
z_c)?#>do>}yuH^{{jZ&{kP0sd!uxD?-lxN&x#=J@*OHKkp2jx!YiEg1FQ#2j%EDdl
zrm9H%_s*fu1$r<1a#^z;cKDw$?(J)<Sy%9e`XyoTh>oH2{^$qwfu0BaxWgYWs)l&;
zhngAF9R4cFawD_AFS!o##uZgUd2x63X>7~in#c;A$d)Yg9+ikhA3^pi98Vop^Lvc9
zGW7~+-scTA;YLY(k@aomGnMwJeGc}tII%<gB~Ttl&JDy>fM2}IZ?8(pS3nKN_Y#yi
zg7=x=;?}8^#1d^ZFmMxa#&8_N0Ka{5$=?>kEChZ^G35Z@ti(l$FtyB5o(#P-%rq;#
z9H@I_mNvyL3HVSc(_QNk36A?DY_=`#6k;(}I<9>ryOTl#5i_K3haH8VkK218F4!6P
zYF59stu9B>8o7AH(U(~#XEu7aN`5PS7y#luNjQ6CH``pNKbv$^t!@Qr0&)BfQ{83i
zU73jad=VYVn0=wV2KfYr2oarEf5qP10(t@X8$e+P8;Lpo#dl+{9K@}+bcku$!$Wxg
zt^dwE!zdThd|hnO@p9Xn8G27Ps&+txEpvO0k#8nmA7Arp8_ipG`L7|tSeeq@LEYPC
zQ(Qsyy_~$;qN|mR_LiE^<y_YN!jbUv-k=UR^wg110kg9W10&=yQ|NMj6Cr||nyT>S
zwo_{RGJdqB*uc$r%ktc-sxQsNn{VGcZq87BcwD>MAT3wHRX7?5(roR>x@n#G?$+$4
z_}h-QC&_*KthP|Vl-zms5ny!FsTNtrb;?A`Wp`VZ0qegxx13?TFTL>dS1}tJ*eO5|
zxE>GQcU|=Jee|1*{pm}B+Q+B9C43}^&1bqvmVM<Gf)!XjZ}^`mN+1_I?790V)EJf*
zx#_`T8$mDkqGW=;WIB2bG5wa0o2$nD$LBk2FHS_t%kUA^71JaE_VfWiroZDmMh5P8
zSJ<jG;+!s?q(w5G<Smwm)owF6M`FhLVqUU(p4g4+Dyi>~VG6MbouLVc8ytr%kyM|b
zCtChOz0Y(}F?Lq5lH!ZtyuO}JxO`JehK-gV*18XM&g`A=w*Q}ngazoU^pF<4PtHIJ
z`9t{0kh@(Z@CQWY#n3H#iY5e)8LEN)h*XQW)={}>1t1SWVu-X%uC=*F4vROqzd#yh
zX}E*+gMXBqK9>kza<nf|7A9G*H5YezOG7y(SH?f_dX*+UDGZq73>+XLi8ILvSMX?a
zN*VWP{oxKY8c~bL6=SHyueR{H4&bZ^MqZNnQA&sTqp@;pvq)U3Y6XC2dhykq+Ito}
z4Flkv&kTcpv4%@bo&ET{(O_Q;bNddjXVP>$;xhPRNvqggM<bnX2b@oc0w2TE;_4>E
zKe{EC$BCzDx#=Z23-gGdesUe}?xLNKxrLq&Titua>O(b%!{eQTmcG>fI2wq0pQ@bD
z3REkzYxdbI6Cmi$va^isam`lQhgo~ctZ*}bzn<<hPke!w4QM4lBpSLm`Sue7*ThGU
zy%ujY;!~9#gaC)T!Rfl#kL~?YV(@Bb$aJHojrULck(NU}$w4&xDg2b~X>)1MyaU?5
zV)i8z$y()ui#^THI<%!<Jy|yn8@r2`z4>D~<5$irH6?WW$C!EfrYoZZBY{DqFFxJP
zow>E3ezy5VzQDAQx7uj5Sl(RXVcw8Gt>oCJobM&2X^Y+aDsHlxk0%5r;GAQ1S$a>j
z7g7bw$fi~v{~3Y#(DFU(T@SUCq%t5@pqU+cH~IF_OUbS`)#urYm$aBu_FG3Bt*xS7
zEDz0ce1>!t{LK7Z2O^B?9FEy&-y3yuU}V-zNcO)CwJUmjaxh2agss_1%n@p{`-=HJ
zztgaGiNB0<3O>zFrhm9wQ^8PDwt+sFz4{)$cZ;)+goTC8IsdazFa0mn!wDI2ITUz3
zckW(q>6jTKyq6IC=<Fy;lE?i)pii;hlEe3)ssFjAhqmnnh=MsD863ze3$tWVBWu75
z98;6Q51=dumBdCLQh4!D>a&0pz(Y_94yF+9OUHevhYe|SsZ&e+5dR#B<Z2=#fahvc
zPOAO|y*CDNmu($geq7PEz-Qoj!@nsSb;T+{4R!{@U2Fqb73$ofIC_cZASw5c!f_!F
z#D$UranTcwA^8`{X7<o_03o)^v1c*5p*<D%_szxYcAp<{Q~0<yTg8TQMfAQx__NiO
z7{!Ud)$OIQdh8v7<c!o6ymus|tZEIA!i4*!+49AgQEX8loOzA%4^raJ02Fw<OKpT}
zrh!OKW4X;|ta$*5f3NCgCeQ#Sc3PX52o$!U&4A{|c6D1Ynvz7FSY{TEn?#lmnNNA0
z5)+u->3L}=qUh^LYk8^zq$}2%AK8!lC!;-_%;1!}oO587)$??JN}M8jEmi9Z)xZ$7
za0R}_lvnT^1_pue(zhjBphyKYo|s8(a<JO6{OGV8(f?iHDFV3KBX(*HpG96~URl;9
za-7xPt?#)wF4jwGwqt6&7L;z5`-HaL12nLD*k;DN>%Jtar!}^IAk8lO*8N3CI*7FS
z<+u;@Pm<lO#d?<TkP$y=`o=q@yBz!`n@=@Wb-d^KEYZh<4OOm~*1e&H6t<U`c>OBa
zjN9T&#KCd=X*vAK;Y>%6c25Hac1zTUxG^9d-F<BctCrK2rE-XoIvDZ0dDU`MJoC!Z
z@@WL>sqfZ5<k;N%YSyv`d>PC`ym?5!s1YjNSi*nTOCI?PcA2dtXv2=fy}l7Hjm;f~
zj*s2s0u?1N9r({|F7PRKi8=o?KJ`kt4BG+|D)u?I@cWm7))9Bt(a>MJ^57b$P{w!B
z!p2Hu^B7OjUqi!xN*erx*09|$*1tNef8ETmKff}Ved=S{TB$~E=Ct(C)a4NcHJ-@M
zo-W%7W#LnPntLk>Pb!%${5m3h=Hpe=;mNe?SH7CzZT1moGiP=UGur_r=_TQ!S53bH
z8n?MG!%j9S*F*FC1#nRfUKIJc%T1AV^mGd+dKVJGWqr+0-Z#yiW_ulfI@vK$ym)1i
zeP5`9w(6}5&cYAT+mQMa=&m~L*h!#k`Q@q>mCc+@p9$^7Y0SYBbYqO+KlvR+qAR4d
z2oR#6HxyK`pj-6^(Z}X+=jR%43-_!CMA}daMV(lH2?>0?T0Gn+>K>02{CEjvCvgZG
z7bVoK)(*qQ&wy>*la9RlOH{J0MtzrDdTFF7UU2XMuj}+=YNQDUyPcod{uIiuIz^6i
zC8S(t(pLjJplUp2Zi3-2i=2UK0`HS39yNTu9cQ259t-OAwr2XeY{h8p#>BE2_Y=Su
z#-;1pQq=-9`Tdkl2x&l*%I-OeU0Lw;s0IQj3Fwa)K8r7&#^0rsAzLlhjBP>=+)QQS
z+LL4+6(7be26kYA?v|%pzq*WAPBalApI?@S4%fo;V<6?P%&foKtPlh+e$jwyMxQ**
zDrbZ<XB}uVnbwb-R~V$dj*GU4Bk~B#A94)Xxuu?M&Zg0I_Arp;mK|p4CKUHhPmuYt
zoQ`52ckww#^>PCmp^{>eK#zrop=Q5?q~`bpV<T7Z&sfRyX(>WK6M@O+xU6f={ra5|
zGH-NnpE9g;>>(jcE2}4kUPDV^Ft;ZB4W!&7xirQ(w*rg4-9ENk?}xn<OKwIP7@_X|
z_BJ`|ZMJ_C>G7Y-`fE9mX6{*aSaVBmkVVMr%QH}%afc9>=8*TB8Le=IuTuW*5yl+-
zF0=u4%>~$7cY0Abt>w*J8x|rH-oYlv-i3|x?DO)mTQ;=&3f@mE$9pN}bZdBb`hajP
z4Tp8`(ZzhR?Y6-?$kH21hE8qr=BtbSWXyNi$By%((yF^r1{PXoE}?zCnkzo}6_Mon
zVVIV{+|p+=ALFo6^`0;E@1+z*<xM#zKPWV@&aZXJ<?OHpjpZa&@ZW!J>Y*XsTUwVG
zg3LOXwYl3g#yktc%8Z?Gb#+Cr^T=f(tKETE4emmo(kf`xyg?pOTGthGAeuE5E#+O7
zuCs$Onbxn=_Qx)dc(VeJDE4OpZ7MOpZ3gIDqwf<TRrT7LC@j1>f^QB&I0m7ckKwJf
z<o7pm-I}pSmBltA@|#1X^ftZd<F!yWE+gsP8B}1UTEex2_dIjkZo?Y)*{^*>a!8a~
zZ2&we@LJY;!(w51SbinXG6#HFV*UDl8S2QHw)&oJk>x-eo;g;-Dlhh}h~p8UY8}b~
zsc<mZom1E&W49F(gEaJHD#uEK|2%+GfcN1HxO<OBBg_E+6sQdOJsV<TMz{-qrm7=A
z?MTlZbz<rs1vEf^98vxqUboZ*FYfd{WE2(7J`Ytv;lr>y9|w2gw-6Y6F`2mK=Hq6_
zkMqf?CmxoWTPpNG7Xv5JPx{jclFV=5cWMs>0i4l$r4dX*ssWsA<O5MBhEG7;b&dQK
zaJbS|1&KN|uvj)RSZ?cU2|M(sGALmlU`8>-(({bI-Q!RL@OYg??Z3)6EsfHLzg{NI
zhrSjb<t0>xa1w#s!J`GC4AQb(a0V|pK^9Oek`HMgE&-A#L5#flx&LJ@j|SJ=e$Wv@
zLG?CX9p<a_@%=tzVP;T0C$dL*syueVeM5e=MV>!aV)C~0xww(eE+^uK54xI~cz=*y
zZlB3A|2}s*g0~&q6buMTeDudAiL5U(-rV?(@@C8eYf3n4*w<4PT_i)M0h!g~u9O|T
za<9s<+86b=&nspnp~yLskDfW@4ZgFc9+dEiF?RV$-~H9Y&x}W{GiHWqCs@9|L+-Aq
z_|>}|yKlZ(oDgefM$E!EtgOIY4)s&tP{sh-7bCcRIVnxkbobM;f4W;F<ohg-@m_o=
zihx%BN*wc|QRZs$TeF2^*dNimy|(>JwPTkZ(@?SK&Ic)DIA%XD2@$RK`O_`j`4%NY
z_wNrI=r?5$l*dLp0IHEUXtdN&QYdW$7_k|m<*}G+^jRCsadeV#W3M5R)OKL7Um&;q
z%@Ia|NNqvV9tNC+ccPU_!18~>?~Z5M=UHQ~dquIg_5J@cmcF#Az~&QC1O_FT0*h6o
zzV+C!erQVbT)NWt4j}fusH}p?N3Q3!y;MFynTcw3=$NCnG-X)T6$>n#h1aO^$x4BF
zB>A??l4tIgMdgk$X`E}g@<(K@q0M(Y4V+ve9?II<72~RrQ_s`J-`pKaS1-Ycmk+3y
zFr3$nx1<tsba@TW^|}J2S{D9M|I+ir)fry6=e~cXEl1c_kMr%mtNi+l<Dr-~oqtj`
zb5rt<f5A52xEQp9Z+dO{ac?AQ9dQXLe$~6pu^I~sgEu@1GgNJ22yg{50m$e&2Q=~h
zLlVMQvt!(xun&R8L_CFy{X|eB)B-?UZ%Xz6(Zt%4mZ)vbS^a<)GK@N>ApG%T=7lFY
z6oT_Z`3tv@LMm#MjWM7l$i>!J8MM`;wO*fp(>F;!E(=0fEvIRgagdUOFRS$cI<4-W
zMfkJgUt~M4PQmiL_bA9Xz#M8u%a2{AAxu^}g4YlN5VtBi;yilNx?R7a=*K9aXIzN)
z0C0y4GR=`$@jV(T?gcyNllNARJ!(zP9l#;%Q^pmym=uBkw0ck>K1TL@3M?YI>Op!K
z!1#MtOii=5k2F;r6gusthh$RIgP9{Zdr6_rw^XsIG%auTka{?SpY3$@i=3+M1j(Vl
zCsJ$JVc9<s^C4?HtJKQxv>PXra*6b~ta#!6YmV0-gs?09;^ZGQ^TPmUt?S|)HiibT
z;8_sX(#Ma|D#n;9HTX`NzN??~T&zgPIJs}s<}_d<ld#cM!yGq?-6V+jyj$OsT1FSI
zPpEJFcd6g6ow8qlxVombOsl}y??>F~FPgGHkGg9<yoeHwcR2d!=M!Gl;NOy+_ct)!
zg23H)Xbz+~EbCEqmCa{wZTwV9vpe^f0FGvU-<t|bG1Yh>lcyScN=sKEK^JLC_ibPU
zwFb{e!mc%3cFxC%p^(U%y%hGrW37+L&*LXe7r*^eeTSXE&{kt)iV??s?Vc#BiUw<3
zO}iVnB<d2HKZ2(L7C4&06c@yW!tZiBc1_6{zV|TlHE-Mv_q1;(j95r%Du!RA?SC_b
zIXOjk7`^K}`R7uJuxz(Ir&dVIxLM^+B`XSFP`X?X*yr)P)@mxbZ@8&KRxL<$P@knk
z{C;i3?ldv}a*RH1`+(*Ybsi2+{{H;nIF5dKYH@>X!))@<uJCzSl}WQtQ3d;L;e5_u
zivnM+StIk5UFAPJRRxpNE6uO}`WrasXZMuNzO4EML_TDSY;9}Dx1cd{)r3|oJKeMm
zG&UR`Slg*xG`hNq(b6_BX3+epwkkvYtVWa>E3NGqag)~vp}VVc0Pch^sbHQ00WPK-
zo7KG`aAw5hqGN@I$|L(bv2fqFb%F9JbSHN22m*xL**>M6I~%BMd&E_ZXlHb9G84Ub
zs%<G=7`t%E5Gl+%5(j0nZ6*M34o0%+<?$&xDQqykvhBxgm6<JDdewWgibTV;D(M!#
zaHQD+T7OsSIzcQ5I)e~L?itW0r$d5fd}sSPfii887qaI_VYMeyp;qK6<9eSv5HUXQ
z90T)XK$##EU=?#9`%?x%Fe?s-{Rpczwj%rbURlZGM+x;Yf{=k7&aV+A4(H>|7cuJ0
z<WTf_vz<!=#I(m%&B#$6^eW?k1nQLVpq%#1`5fsQs_1s1OJ}9bdymFZR8b=CaR^lF
z)C0vQQxfdhZLOQ7Ce><N2{3$S+$;c2fu<k?&1vzvUq54ZFIWXraaOm3oJ$N-wFyoh
z9N_1r4O1W6#C?u&SID|wN1XW^cTXT|TBiG%b}75eMVQ_V$&Z;@JOozrZe;5eio<|a
z&;+_S_uliRgLs@OP7|&rT#?~<$u2^i$XD{deNtKjOY{L(sh$<4Eyf<-6HOQDO>IWe
z6w`CUrJD^OS<M7n6{y!!7Q^<yxmQ?n8%I>>z_-z4<&un6YLaNFNNRp1j=Z(3Wxm4K
zq-KP^<3;^d6(;htW1BrF|LB`iIC5UW=2lJZ{nhKLyZn~Fy~>*l8JIjVx!XTun2DkA
zp_@aYrN5RLu%&HYzLpJw^0>M6<(ty0Bn(Wkd9EFu-tbc>?`EEMzWwglYjI%dx+dYY
zSyYw*kKV$8f#L-AcVL;Txy=2Od3kzzPZDj@ZE<6GxY}FXjxG@yKJY_h9v@T}=bn{R
z%L&^?%j1QX2QIYb8P>|<aV>6(O7eT36P>6Yg8xL`y&)GE@lf6%_49ILY(8LD{Jlg5
z74rTw*}6Ni?2mPlk?%IWX}nmsfK#4vlK%Qg$vI`7BQX-XKEK!snQBcYOQ>^~x6(*B
z#@y9nc*Um6Jz(~yTeW!;i}sNNi+}Cj$aTB+q5#+AbqA7rcrrg3YPHf4R8KS^)IuiC
zLIzsGGE_Aze8#Q(c+LYqYhHmU3B+d~{Fez*LkUZC0}_Lw*dE-g#juXTj{RCc*wI~E
zK<@3=va(^5;^FD1=r()HoY2M%m}l6|MaSL7s%)Z;J*F}!Sn2%faxl*hedG<MHS$8&
zGAi*=_5SkB@Z+|~v#N+a9l2*H0+qG9F~bhDP6!#xz=)t;(LqSKMnultbkfa2n(Q;w
z&+McTmKQ>{oQdK`$X^x<*?Hs6*GB>1144kw04@oaO5Z1NzlW#KFv@vne|Pz|E$_VW
z4yqyjZ3v#6P@*tp5Tp6ojCq);g^r0~a}6hm+yme~1{h~lJ#<O9a5U46mXMxvz(On4
z{o`aMLw4=lsYwBguitF`d!HIj31yRMtQZt-8|u$<lgIRUZ-jB`zVia6M;rP;)vx8b
zVP$D&B9Y`GioXY>8W^pz0{s}o9x%OI<-7-Oh*qPPY|V=HY{X?_Au@VIt!zY_(oLmO
z%N);bOz8yovxO%~L=nBMmif_-Vx`t2L`CIBbkl*17zN>vZc?33-V#R3m9y=C{X<kH
zd7+6mgeNKB+AGnEnH1>(&L9S0!Ckv33xy~GsD4|%|Frsa?~{)2@1XZ$(*x<-mr+BY
zb5xMC?ok(<(}1AzOFYki_fWrWD~q81C?M)PNO{~O%qa>4(DizPBtMhvSBcnqdUy+~
zf?WIush@gPo041u;bIuGK$83(aSccv&}8JQtQ=gA+w#hZjDqsLv7+KhXmNy|)m&fN
zbVU8vng@MqWHlC&%iU>oHUI?!w}@*jKuGwz&nB4+WCT4QRU&L_ZijA+Hr#%ny*J+B
z6z^_JwhIJ*1h^I7{hSf?<^VE3z$K47%J3!BH)8ia!7iZymuwKk-cRwY@0rm^n3T8%
z-h=W7kJm$oqN`d=zcnbe99)wLb?0tK2N=*yKR&=@fBUQ`G(RCycyESdBt8eCX&1I<
zUM0gH78Yc;v?+60wtSJWG;?<tWSiUxr)+!osYRFsY0s;P^Nd?jlm;h;zE3m`1Q(x(
ze*DzK|2bA*0>FcrI#1N2s%{InBRALbsP`r!wO34t8gR?{L1cEO6im?tZ>fE5+`wK-
z%$<YKw9DmNzPrZXA@WiP8<?Hv-5PoLTTIvDK*SA*i-6*`W~6<K_Vg&M%J%=!^c8MV
z_TSU@?$X_jbV!GEOG_i23kwR;T>=XzAxldrDIg#qA)PMLAt5Q9f^>Jh>+}7+|G{3@
zo_juL&YUxI|N25KG%iRX5`Q!=FZ3e!VMjy7+9`K$!QAKI;C|W9QqUshLZMseDo*aQ
z)_8129J9Q(_`scE>*C;H;$i1*UsX=AP6NB`Htzc=?;>UU>F(#XTSnxS+-3iRDz`o#
zZDl|A_QtJW$e~nQPy-)kPMz0PNHv+{)0_lvg2c^&+<=HzG5VI=lI_+QC;aKOA{87O
zr)N?$<0B8-QQodPeDchL4{4Rgbct?Pfj0rwaj}BOl=r8`0ipM8)&Y9MzJPVeb#RDg
z@<`BmJw&Xr{zqWZ?M=u{!D7P055v)L?9Ged8!;ttc-Qdd3d6%&?&%OZ0kFeS!^7VZ
z>&Fb32*B(%(ISC3lRay7Uaw9W@^BDx`m!xbhcChr1qsN-P=3a@?ct9Js1zuG+7Sny
z4mAH$<L@Ur)}ZX@xf7V&1-C1npU+QF9+3uS8%Zc<2B3Ri9D3wdvBhEXV)AAOLq))!
z4Zl64H0I&=UDP6PSa??t<6Fyb<!07puAyU-l6}H^onkz0qsvtDgz1AJ)(rU0HJ@^v
z5{vWOQ!0OQl_{SFXYY5Ij(k!!K6dM*7eBaKX|kM<mZEE66vr4qm}{Yblp%X~ya^I|
zWtKOCk!~#Q**Z=iSoaQLK2L?CVy-V>hBqQMiACh+@|$a>@iy=N<=(;SnS_((%55f|
zb0(7uc|~nnlkbn+l0yme^NZRbNcQqG4YkNvCOu*-AreKH&VEs;kX>%HvL~aMA<q<=
z8e@k)qq}6!K`CN+2BRu!*%P}=wj6EV(FdJ<7h8b3jA!qd7A7sI0x-Fcr>1iNh%<D0
z;Y^&0rt5cN`jV6<@*^@iVO5lBGeS^1Xr5nQ*se$w9af`kqIBatb4XO|&oCMJZ=g58
zi^fLgZ3Ya?-lW|A-gak(_0D5X?5!0>SjQwe?lbq)?x{0=sjeowj<;kRfU?6#-DeiQ
z#psY4-DK#I0RPOZXacx<$E+jPsKVYVv&LI`<d01zBpDAb=CdteBW?qg7k!##@uHe7
z4GVi2=0*ZzE%w6F4R_*262(e<`+gA04KiCFNUqz<RrQ7<lES^a{I7@S!sZLVWXpVA
zUaG_qPj<Di%`fs0@D+2V3{~1xd)zw^Y$7wjQN2v&j+bsqV3Fj4eyCB*pLpI{j}%sr
zSp2$XE-6P}$&q~+j3tmPr~)n>@U3(}{s$f2%1mJpw=LNaKD!e<yGlE;m_26B=${+m
zy)X7oT<0X9Z`4gLzG5yG<mmpz2UwOaC#_czKgz+^<JNcG#_~At?=Q<Yh%X}2^|ko%
zjf&RlIkFz%gxyFqwOyrBU6`&M-?A@D;UQWTxB1hNd2>X(-^{oMDsBHJhn*F#hiqB*
z<0rA%`VsA*$**vuH7W@LoHy5<9E)y_lfjDoJMBg?NM;b6@vfggv>0)I(59b$+hE5k
zTWXw57kc<*y6pn>n6`dDUs&4RMt{iTp1nOI4E;rY84=L;W-!m~n$(&^ny%^LgV>{{
z$z!xSLJlgzhkHea3zN3)^K7@Ep3b`Qmd_B2aAmf|ZU)l4qZOTy;hV1st*Z;6{t`v~
ze!=6$L81FH2l1tWf2o>pq686dHPDN1fj5_~)n=oRo^vReF+j1oCwhqp{&v2LEkK=7
zWoCoaK2+O7*S&#EyK(G#h2avHU<iBjO3+z}cS<qIX*cvP#`;&&E&D-Q#SSz){`MjG
zl5{Z&jq46G%m`3CAQm|9&b=tdETQy?V>(cky=(3jInMk=K8=sUgHN{%rMnCz1Q&5(
zz9^({0sRWzBCQN;oP6bkjvt6-ayjeQwkxkV@P>4gwjNu)lcAJfpHARu&H~5xld+M&
z%Dz$UcMz4npq=%ho4BQiTF50&mE^`l=ELcTid6~+SPlfuk&qoNyq~|}kb5Xw6#HvR
z=IhvliJr%VT;zWkW@sYByEcDmhAseJ5Gf|t^x&Q-4lemgBNgJRUY=tq>ch4veY#+O
zW9oBPc-`8sLNeUnsxo4j``xgPu<pZ!p4P#iF}Pu7I?AcwTFd<9No{)5uV<b`Q`Q`R
zy&Ue6T`PK7CG)lQjm^qLn}>g~GPJP#^djKErRJ0ykzsHZ({`hH&cY`*uYG>==M|tx
zQ&HP<{fN*i;l#mi%HnbVE8`u%$&0OjVDbcw2#-uS+)Z{h^3wI=uzh)j1k}!&5fd8&
z;QMOw_aFm9$TPh=1KL4}mmCj3Ruhl}Og>$K02))3WO@9z3`o}{VD~#+N|*$WhT$`>
zPha+3crxtmQ}>J3@Jhm1a!fJ+0ldRQXEi|UNjwPmACI3cBe_xfHct6hQ|OO<ba61=
z&5QY^X+bJRfWL9ZGXOsFtZ_Dc(sPAb<M<pcu@QPa;CG~2n^9N3d`z%DCYZqtQ0j-d
zW1#y{R@qe4jE*(^u|&6k{x#3?X9)&puq4>2xMId<h=IfC=pej)f~p3?nCh?7O<@^4
zq~;`ZUfo#UQ1vgX;gxGbxQwugWE38*amDg~9#e}QTYnJ?W0fbDC)-ZJQN;uccKsoO
zuh~sC%;nf^&VyVJQ5V7aUVO%81T$l9$%4;-FVOj2+)>Y?=cC1^D*zj><j=t#bX_A|
z+5MVp8n>3QlhR-OP;)`Q{e?C!^Ud~8vbE74-vZPEtKO}=p~+NVjd_DG5M53g%7v<L
z%Y1h)X);IDJ>!o4!3mn7n@~?*!u`*K)UXNF!IUgtTdR4>nZrz>ckMc^VUmG1JRjS4
zv-ty;P8ja66UcA%dj3FzsGdLkbh~ZGTV)u%84JBiWFQ}sG2wren?V|S$8cZUo2In1
z#ODsA+oH)4as=K~;nm6=hu(jKJa=y+Ct!E=iLMM^k<1!BBXl6Ydvb^&X4|Oo!~gdC
zZTb@F9Oi!Qy;IZV^@*+O>rA^GvAeC2`){@%UmxpxtcSOquz}NxJ}+I~=>$C7jNG<3
zD*qXV>|WN(oo(<j$~dP7JXf7Ya##lq2BF??&GLNXwhDF%y}z+?2NDOU^X#D3mu?Iy
z_EL93X>ue`rDZoNO1;0Rbx}lo=zZ)kl7T#my_BFgT=3Z{oVaYT4$lki!oJWEiFKtZ
z@Q(smFVcd0@9#s;C@axE7Hb*==C1bKUb4ClJ%M&@|Cqm4ZmMC{f~b^BBRN6iyn^uu
z5AhcRUw<$31<^vctvTEAfTe~38G1%XgYdoXI@fluf`Rhw87?@no&7J8m(g+8y5_!h
z9i^u=xO*ypy&g5a4Bj0X@6qC|!EKhBjN)&1rL++!zj-}ep!$Ko_D4tnqqik&jy)Jw
z1yQn!I!~63n)0!#|CIaRHw%3ucrbfM>XY;0uZvNLLFVu3%+Da<LX__z)+ja7mtI6@
zZ&3h>cA?E1MoG0A3zky5YRw=83lVg)!zEVZ`^7`}Iz@Ye@m9$#?D{zgNb^3+Z@@8j
zVQxSPTq10T6^;l865HXVR5VmGL5;J?9$B(c0Fwa0dxEpJTX~F=SFX!nqAddZWzzBf
zPIQ`tGP2w#R%=0}tF&>9o=pmh+aZd1`~Lt0VMNOGm~<TiIyBAxY$UodqDc^42;(YN
zy3Kat3y)6aL_d8@F*AVT89;&t_q%~Jr&M3=*PxL(-Xd@*&4+nE$Ya5YkLh=mvM0Nk
zJCv!b&6~mD4wZSqGkjb4Qv>qyM0n|~v(?Cxt<u;Qt}rT$Q>-(0RB9eV@*PCK(qEC6
z&t94pPW}D~D#fnlO(P487t4zg+nF>DGu@H^&lYhVcO#7C=J@WUoHzFAIxSO>okYM1
z2b_WNjdv`80oHY4VCzh2qTB`VP^Q{^<B`uj43Rf8hMh!AIIU<qj&xI0yQu^%*&~u-
zj%&0EF?Iu75lK6@cnJo`v`2XD#^V%%w=$hIDDFowfx4^_w&pf|W~=~jd?DYH88{=R
z^iSms({U0Bm{HcTSa9q;`AFBgL+^HiC3%l-D?_)v@wS>S!b3w=%sPFC$3lF?^)XjF
z24B5ZtDp4OSndG0B$Y>}YmC+FKQlMN);D|y-7gd9%YYaC<_Wa6VXvR!h3i;Ry^IMW
z_HDDzMp9i}mbuN2Xqfk0*Le!R87xnqpu~7-7ob>nZA$%cPuS;^%1-~f<e?Wur%iI(
zvoL}D%FFQdET<ZcY|Sf$d4x{e{qi&H)3~xKep1@i9HV2%HIdk=<|a)~N3ZNn<V9FQ
z8_kyo?l@leWs@UI@lEUNW8C<&h!);!6`d-dQS?ZUiZGqFbEF`9w3DC?_;5uUbO|ji
z)}j&<;B!w!?~GC^Pbx8ySnn9RYWRMIIZQqqdd#tjnFsVh?-Qd71o2zJTbJk{B43wI
z|Eu_mkkm5TGu0D)LNVK{%!B(sl{q($?pd-Gk-Q|_pTFPHF7BB?_vO2Cm+rT?<L{nQ
zkowPWc4Wn90EP?`-=X;owsEaDu~webLYrvpR2RbXaowapCg*cgLkjQ=w6q*T&+iz#
zQ^ufkJ2OiZP&;}U{<()GoQfSMyXVVmUTfbJDmth7P+<}1zRa&D6U8!rvo~Qcrsp?4
ze>t$Pw8*vb^MVt%{7IjjE8W%+*6%rUjATCkiA7{Ax-CA9<f??%=3?u?P?=?>S9&fl
zdef$n%qJy=)=@PoH6=_gN!35MgHo=>_wl`dZ4@X8KBmJ}T(V#_VVQ(9Stc(8+rs<+
zu!@U?IVi9SNm3y8xN@y40WiSo!3k4;dR4j$plSXwPb^fVyrE<I&MO6ajD_&NSl26H
z&y!#7Dcn}h{h5+x%t7$%dDx_3<Qr*n>I)A-qhzKW6N(vA;N|TM#XGkA*Cx2jz&a=^
z(fwe<)^STrVa9R}41prOyNV8B5e{tUd(G3YVz2TKF6el$K>`%Top12KENHyNsBjAc
zRxi~(0uaR$*JPNj8Y9XBenXSwNYX;U?>F{4u%%jX7#<%x_AY5ZARb1?QioYZtO<h?
zKaTfj5l}uBa5umxis5Pw#-FXs_fm8B0ry>WzkkfFQo+EX0>rQOk80~3Cqn0teBit=
zCP+=87l!!ipjBGbyBz6s;M)pK+*)I}>XK-4{<Y&fo9E=PH&iUt0?7!i{;L`9XWwQ+
zO+Qcu=kU*28-<?G9EK4enm)93w(vq!L_@9gQ{br}QFXPiW9YPIQFuFBunwD+)yy>R
zz54*onKvvkvfm{CI><MOob+a^i;+{Ao}+oZyE8ud#>;CRIvaIRJ0<2)lEH4CmY#!w
zDvKSjnQm1z4N3OqF71r8;jY&bWGCIfiYn&r{nt@fU!g|2olnkOV2f{Eer=BOA$WPI
zQ?{Zd2L)7qB?qVM4gH!czttLOS6j*qtmkE3t4I@3q>-1mta$7C3^o7BOOf?;0<p}j
zfe0-$30GaBQJyJ}WZj|HT>W_qDGC<KZ6u5lc-Iq5N9@J;z<Z{?2z@II4{&L^z#=B%
zetqS|2$CWsMIht3pU$^9j(xeu5b>wB3;^ql5jFg&%Si$ntdNyOt^yGw{One5ojH^D
zbF|}+1{xO+1eG3K3{YZStw?U<@R>>FsK+Qz4DN=4=)h`Wrm&^gmUwqy$E~Q>Qo{sr
z#eg60m~`8^)j#2B;?4*Nh`B4gHRrkp8z_8H+%wlCp=?0Q1An?)evg+Y9U@Jd`x!(8
zDTug+@TS!&!F_w_zD&1Pn><c67omK=h+dkG2j{6$i2m?&^fDH;#t)?1C}%KBYTHV~
zMWY%8pNFEn2SxF}2bGSP+%(6lp{M#(I}}nuhQ)g=Xd_&EQ9~{ZkjgL5hi$Z(4!Op}
zVAiRhUXL-yT(MZo`cr4i11{>zhH^I5DPmE}C%%TVIjLG<6%-KP$+hK$G}OA$l;>1P
zQJ5y=1BPoN28!BV=$zLQE-f%viiX>Z2#iJ%m65Mm%<=w`t*BH}$J5Z`9;AIWLJ^I2
z*Zw?h@-3@{xHLx^$K%qt5?I9{&i&9@OT*ZXzcL>1xXT-xz$SYQ@0_-Q;l)e4kx$}U
zGBP<zsuPd5&1_TWwM4-G)k5)0XtfM-baipVa3;1cF?Epw{|&z5+_|)6YEIDt?G{?i
z)gfI$BcV8+{luea4-fO9=-!0P;j`x&d|AiG&CnyotR!<4H`_o!7YIB(ZTb2Z(rC*i
zRp0Ym{_R_Mm8!t|i^&j$tH#Ojzrd)%Ew)D2C!~X{QF%<Yt=8zh{4N&Be#TZ!!4jnD
z%Cq9zL{YyoHG>It%Y9B?l_r<gd_)I4kDw0=BLu**Ef#lIQVE7AL+$6Af|f<l49IXK
zHrdOWBfi6w)nvUldRR@}Aw`Pu<f|44Yy{GWV!t6;LCENqf-*sI8DLb6o1sa@O~V=@
zp`~rk`x$QWo;Vgw1qX_~Uo83VsJhaZS`LbqhY<tgJ8ZdTmS`}6b0#Oq+Pp`Iq%*I`
z;)*%+Z9f<}lLZ)fEMw-SY^(hyfUeL0RDy>+FOf)^PeoMG;3rv9{F7mAzaS-gT$38Z
zsZ<+){%9v}Z`D!6fH8VGpr1rUt;xZ0wQY;QBO^_aAf4tm-!77g@}xf)&D8=nT8=GH
zneKjDwN~bOpoN@1`}=IlgBWb1!yijng9KS`y`|eTa8vw^?2ysVmCrfG<^NYXzebVu
z01E{nQ=FhnGk!Zg`_dyPVtW*Hn4!&#X-@^`h<+}5LXGFR<NOFGBv}?@VRpQBH_ywU
ztDb*TsIW9XaUtzFQqUPNYg(^KGs5KcRYl$w0TcPUyFb7YC?WP;W8S_eNtA_)M}vf3
zj{}B~{;<vGsn@1Los}PtKC&daUsH>Zt#D=B^c{Df;Af)BvKt<^BMl%p<Gh2jd)P+&
zIwwD+2#yZ^RU;270%05lS5t7=$}S)GQ!n})?Z$%Sc@jEwj}*|T04^e!(s^)ie4Z<|
zN*H2$my{NA4%!F=)5Nbo-8O&&ks+Is&r<BdQ(-2(3%_(dt9Y^U9Q-1!JPKzXfmIv`
z1pru(a-<&#g6KMr9zX*ZLu$z;MO~BTk~7yT7*czYgeCw>LAfYkOY<%Uyc4#DXe!EJ
zmI>o+>;`ixFuX6u?_DL54{xQqRqt8>8mX9K5tsHEBm)-;pr#{|Mo^wS!&V}4y$d_+
zOE`k(7J#q7K5_aPG#1l1;-SU$iWpK(8{|>xK2w`jo$~5#_EDJHh?Ftr^dGjQkVo(M
z%~7$2T}~j}g$^yo@Bhf1##%hQiN<-f78k9~qWc5%7@*+V3>>FOHy;*B-^nci53^DV
zs&E=l1VH+z)n3>sVFJS5y-j`|rl+0VEDDqejX;h#Gzxq}tHrj~aIw?<`DUNDg2~-f
z9;If3x%NvgZv8Gd!k)C>hwVmAGMi28(kjwrifd*Q2CX++68+3kBriK=*Z+S-B!tgX
z9_9j4p`dsj%bnrjX2~2W_j-ErpyFcO{r%F;fj+E+aHZC0shXMd9eeia$J}BoeYoxO
z>CNV_B*(R`nOFtI`gf_tJI8b%H$Q5Y4*eq)7>Gh#gi>&8!w9*%X&If^hct95fJ}qB
zlkJ$yR*lvsF|C8Lhh?{y^UoUB!2@SUQTE9byAS6>3LM2kW}9TR9-Qs~1%PpM(r&@J
z@Fn&GQ=*b&XEc5dcui&*lM$`Qd*0<kEZ0EtF_;L;lR#i)UZ_8jKtQ+R+QaY$%mCn8
zSm4q3$~?D%V(nc%1T21`5HYyGSgcydls{NCdPR~D!MI<Wz%#Qeh@2*;qr1@R`n_(g
z2q037JzH2IGDm#ZfTYkqL>kzKZ(lOTT~7<1gOsQ;VXjzPjQbcs3{0IfVd*{WH7f(>
z5jD*Oi-efLFBZl`NQ`Dft&~m7B|K(iKwb_uLsz<H!U6UK^y#3V1l>abr6HH8b~xFl
z6)stLp>MC`%OLPGC>+t%4rf}mjlp^XLNZEBMI)0=k}4@fHrI{%+2Z2FSqo*FjA+a7
z0^;|;%U~1cDIy*+I5Vvr6ubFF*6G@6J!5SA)uuHvsUj@dE;1O#H#902$b*PlT##xY
zcPsqj#p?cL?2v2n<8ZY+4+v&z_8lWbj8DkPLM(5@OAu4OWM=_RJd-`-Tio_ge|6{Z
z?eo!)^#eKSQb8Pd`<vX9H`#+NE^8?7NlO3uJch;4g`kP?&g<bp59M!bA25XQ?P&U*
zRDtmAkZRW`^^DbHHp<Q4+p!`UuOa)+$_i2@F6T(IJ)XutEKWXEDEpP)HdFfVMXR~l
zUE|j;+;WR%_s!%CXH@W?GqtaRd5`JL>Uid6L;m5-V}bB6h*#LtZP&7Af0B20ca_?A
z5@3VuV@}R6sjQ>Mxii=FK`H6*$(To0HgvuE+BfuKPD&><@h|E~)8+OS<vDnf0Ck_>
zk4}B}9)Id4ZdF4{?<~gk#tg}6u7_~&`xaJQRy;zyg)RhRBo>C7g#aiEO2;q&cf|+*
z$YQadKI(m}Q>tgB5@7M_%%Im+7w;p!TM2ZfO6*!j9!sXzE|LGs$n&PlUpkb2ES2@K
z*Q%{4*h?Kj$0ZwhCjyd@p13`$0OC}d&|_^f%}F5q3yZ3D9^P8TJ3%mMH}yUnEl-NV
z9HB}=)~PBNgVDzJfS-=pSP11*6y!PD02Y=957Awu-paXQFStqrub5#Q5HMNx^hWRO
zI(v+SGj3M#4T~fo+YC2M7*p+@=%ObwXLt%2lRYmiF}wk|5!8QhgR`94(ahhxrxYIx
zy6plnY!1hL=`kHWdAf{!r0NM)8YRa*d=m&fy$78+-q7Uk(F%YYy+CPOAA#s?u4pNk
z(j$v<31hob^Kp@GAE1++U!Df~#_8w__g6fsqtuX^dyL=CD@jqggDaa?J`l1WYa<~$
z9}1%cbT}Y;{tYg4RM^%TCs`$UR3B6|Oh5c<b%gNCts&F`-$JsB7uISCt*@_#P;P!!
z*SG0eWl^jLhXEV-nEg2ok2Ws<6yEDG_MJ1I)qFdm!2+|4%eOcN9DS`6!|d$2i+Z_;
z>z6P!cLZ)jKWme(GkDiSeWUOAT7{)1M^U(RDSG#RlMOrS&$8RzZBnUw0^dgdHNdQa
zCSm~X<~8$AzjO7<P5ShDN&{`r9OyNXw-i+YS1G-4;p4)KiEobEv=Xbx(0Q3h(M{hR
zp@mTknaikL@fcpdgjcrMPil`DXfM?mvG5Y|@2@`F*+tRkX?hdRMO-rR)kou<5Wo{{
z+BsfeA|vnpkkRN*#aaRl<yBQNe74P(Pn+Y%+$!eIhX)OfLpd)jNrT#A>1`QfrHXhg
zU>-xOsZC*9Bgku2jJt@x;I%hsSgR&NU_w9{i)fe%@KbBVX7QJj|JkoT*K`C<fXh55
zBC>(2DeqSUCX#S^fJ*=h)le(Sxd#0IBB6TpCY&3#!%|XvlGAou<WtW~T?c2mA~Xcn
zkEyxzKiaX^<08R~m<aqHP|I5v;~^Ri)Lcdkcn1Um8d#<62~zwZnp*WCQe<o}kd?x5
z&092zbVKEF<>E?Iivl9<*1mj2xY(4YT^!WdtsZ=x6BLd-mY<|nkx9U0^u-TUdaWi?
zTwLBM+~AK8J`&5+eT1FEcF|_<)#2F_YL>ZS$sY1anK0)Uhdu+XVWB2nVH!^vK4w+`
zK{>QAyTHFZ7m}~w0ZTn$Nv!WssyM6J;PyYPNpyss{i}$^ew)cDHEZ6Lgdxt(&e|C4
z0FSVf{8_U}HOfyf7el?bWkO#D;a7M_e*CP1XVGg$-|$toiaI^mc|@2geq7O4+6>~@
z{Og_YF{=oBcO9h+NkvNa#r8|96WN13{vLAEn<Ig|T6L7G_~ju^nfv{>?-8&~iHSw-
zz>Dqnn-u%GqWk=5RhnSFxt`h}o2eQa_r|M|mA1qAY+S*TweK1=%R7|Muu;6rspK;5
zE}+zKMU@n?rS&}gr*X?iyXB|c`?8kZ!W=Vyb_jE^X1>>np6RrRjbiV2H^Uw)pM>B|
zFVy3$J_TI+)gDZSaPc1CfT4>z)jICAwnHDAQ5_Fd5ejkr<Q_GdBK4J|qUe-iiNtav
z2`fVz+*)PA3=qJX(Ciri&Jx*}z`}M5uCojBP_$_aKzKVse$)O8JMmX?R+BVedDn2_
zST+pJl_Oi)x$!Hm%Hp<Urz|Fb4sZ<NQ^G)v(D^P@Fs>9|o;!mPnG{A5o=XkyPD5-Z
z7+}5=SBXT12iF--$l1}A@RRhq7jK*!=61Qwkk&n03_5OP^Np6r20c>lmZr`Z>vr?l
zZ?kahe-;2Dsa3buZ}pN=X{CrV`a%a<p}$eOW&vp;wLy6+bl0c?q=_43%ZB0Bhwvqm
zO<Q(a&Dnzx%TJvqNS)L!FavI`T?O2`x&_7rx6^@V6YaQY{B06MXcwlWD%f}CwnG>y
z9`_PDaCt<ShvBoJMZG(D*WvJvX(zBDM)gwS)Va?LcvtrInO5cfu8d_=5XvlVTChi4
zH-tWzV-F?Fqxii+M2mI8zdmXAK43^C(UI|A$NhCoo`|ZNuzOiGYfm-ULQv4A?)OR;
zVHN(fT*Hs+^=+?|74CMS&^ht(XP?b}qEF$|<pfQyXBC4P93WRec_VLKGv$sNv6(QG
z;b`-!zGp`;gHYMkfgrP(&##Rd9PXyb1<UtS5&a-LTP@8A@zwR1SGc25gdMJH&aa@e
zP)n)Y4%qMeTbdb+J7r6;aTNigWp(I~c%G-(g1eQ)lDgxr(MbKueSQdL+>zdudVJkR
zO(PBIa?q3vb^8*Es#@%2*hy!Wke<b34G>Q}ch41S6%HTKuMoB<SmqB*a)J>v&f696
z8>u?sS4ef8DStotYF=xU;~JoH*j_RyC$)yxT>bHLLGf4Vg#vAbGXITK>&TRD7?crm
zuwuIQyQ{`!QMUMRT#(=VtKmC>*QLMz294)nk2`6PSgv1@x9b!&ygrk~N!JO|6s7o?
z$hV*Co;!EkGU*t&=}B!kS{n4noq|7h6E`+&AaJqvG{?^826a96wkzIK_|Zj0q0OIo
zX&W^?9|_s7vR5ZXju!-8{%~e0mXEW|n?@SDSSBDikHuHJ|L+BWA)o~C58b7n#du`G
zsG2=ol#i)tjh;fI@wVte#Dyezk=~lrp%Wq2>mBuZ&VKs!>^*Ace1$z<)50r_v+)EZ
z)J~ngOip;USe=^v%vnt*Td+Ci4ovfV@{)sz<$)?T#Z*Y(r3wK@>_?UnhFrn{_jwR|
zil-5SC=N#OyoH3Hg{>2yL?j0h(*w}?Ui7?4T7{Dca+m)4Ty~>WPTIs#;4;I8)`$rO
z9Wqsaj=O(@PU78-K8H4HQv%<4TS}<A5;!PmXTZT9+3Wse6`G5737NASCBstqb@uZY
z$;EehOrQOhLC^gf=nu6b>HThRZSNts%tn-zu9x1Di0D<ppJT92r03cOoY_cR{n2!^
zFUWNi!icpN!30@W@<h;A>!UFR^!mAeiL01a<k(9yHbfA^lb+Zj3Vj$PJ{cc!Tq3NE
zrdUU8uI(zm{lwkOpIF?(vUndAro5>%xGnxH?McFKDW4CbYwe~(=T(|>a|Y@Ez8B|l
z>VCU^6h<?SBjcSIu@IWV`bxnT-Ct*r=*r3r>>iObCH1}*akS)eUu4(Vb!h2_MD}ja
zFgX@+FII!sM}&7vZiKf@8o#1wtB0#xa)>kQSIS3gqIa&Zy6qKuQX4myMEgG$@u=r2
zIu2Crxv-59`~30wOnzU|qwVvs^2H(tueCuY)Q{=x&74_@=fK;S|3r;})uDy?ZwE6w
zm#?$T10GW%GAkoneiuspFziE#8}b}z*1i7tig0ndDh^4Z;puKFM=Vx}X3OWA32CH1
z<--`&bl(rMBL2SjQE2L>6DW{%29cCU{a|&)=u=bg37VH*k;I`iVx<&QGPnKsB-gt&
zP%kp5>DP9wmd$d9JRa^0&VtHZe{4r~Bb^V3g~N7MF!CD}Ra=Y`$EN7BJn-4obS4}P
zjz@LekD0;bi-}FC)A%F2U$4dRWEO0RML?(#;mdDf40<$*fsD?eN6a;Kl3d!rie-*%
zbi_yH=Ys2$CDU=a<lo~WA?e0o1C_?>k8o)S<>dn54{drX+8a-o=pi`a?Qp8ba5u#j
z9m+45fHCQ{fGURPSgrL}EI=>EM5>E-tlzg=^{&EuHR2JRp-Y$=AKiWH`(2F2oEPu$
z%y@W`OqkZ1S9pbnPD-`cqd>4yk1ax6K<TbifZCDAf{5?YGxI&{)755A*Fjd*M6GZu
zK!xWru+8jZH`tGqGQE-13-Z7kpfh|UO<3xcq%7JI`1{`X75rtlN?1}Mx0aBx*<65v
zy}tc_FZ{p48CRL;o&5K#>qT~uhOsMC6F6b+c(Dn%SWGkVZ_higHDZ+;Gxu>=8=0DA
zV*El=nPNYEC-o6i_Zim@t;vfD>8RwK=_lL&RS3GuuS5^px&KDwbK*ssMQ5+z5{}P%
z=JGA5=sD|u)WfwsV(~|ZJB@l<%*%K_mpEQvNIDKyED$HciI>8IV67G0zDgR)=tt~<
zIQc(o@$12;pBR%S4^Pe@Q5IBzx&hKqIBLtymM;&$R-<?`4}gKB`t1~ggNe2pRh%MO
zk4`kkaRaNi!n2>Yh|92P^onR=g<uE^DnSby+uPgXNLz=@SkV(iN0a~|3eTVoZc5@B
zz^57_fEfx<V6FmKjtu7*y?_$bV3bgku-8{twJ{$jT>4atdesid@*eI=<S=PF-2ftD
zG<fCHKoynj8<ujjes{p5<rmMc3#=66F%ZOdt-%*asU!kG7*JE&lmStr=^0bM!Jtxj
zNsvc;DJ)Iub^E)8pvz-9;<%?2vZdE<=;u;UmScxL<EijFHJx-uh-^oiV+yHOL1${H
zJ+OD>Qb0AZi59<$P35tvd;RwAmmbj#GtloiUTqENj&dJ2Auqsj;+QOg)xcM}L^fjw
zmL_(V`{>*Bei5_bmSFFlXNPBDc3%u$X{YV#6opk`6#ZuSVPCVfGcNz@;?-p)9{VJz
z;JR!2+S)uyh2|6pH!Hx9j%`i_vFx-k*5v#j`HaRe0wIu8j}vOIOH5i<`F6Z27cb)7
z3@bzgNX~LB*0YAVXs!CCP@7Zp6vzod24i7;ck~hIGUp#c$|7$2gZal4GHKqQdh+iq
zqK;8yb*<uU;q>b)+9dna90Q8$tWC)rHri{e?pM8sQJp*zQ`z79yc~#shmYhdZ@a*L
zvr0-zNhLWEFV&6YyRy`4VfwnszL&dd@gx*=!`b;cTYbIqTIx{){xso3d|B_2@<2sF
zwqx+_D=QjF0y!~(LLa*m`_%R(4V>Oz8m0UeAYZ!@t3sI+HyVK-79YL2Ukt10I{q%_
zx9D!a{{uTXZ&R!+H@q)W%wb@b6=%nALARH75^2@OG`4+;ZWkSHubV8$Fj~M2&o35d
z#<YEw_4lGWg9X3)j5Fpp0M++W=ye?X-O~0aW`wK1KbVv+f8GPcR;7tYrC<V~H{){a
zIh(CU?1GI666^{oAs}d>(FC&8uyR2+&ozvQm9ilMq(kD2hQtQPUy!-eF}`5LK!u}{
zq(5nu;T@nJB+_{h346qwx|CkdmlkdESONM*60>2)wxgEL1Z`PSj_Q*4Pb9>X{ad5c
zD-sw4Bi3sLor=*NELH@^JcS*ZLlKRaj1)JDcVlEIxwa$Bo6w}|a^RP>>Nm;A=yjjG
z=vh_@wfIKHZHSmLj3S$UiXQ%kyS73;HVHhlN&X$uqhh^G5jGQL`*U~Y+q2{3fd<01
z_*HK3YsA??qz13X&CNSw=5%w`Q_LH*?6e8lYZc7Cc}juR<bkyKgj@Lof<k>uOWei_
zPx)w!n2<<PBWxd>AX@|-gA}yO!g2Dn-CRt-_M$*8ardC!2|5QS4q9|=Sv2Cc)|Yv+
zb98fBhq<WX7CQH~56;&5#o2@Hq&*J8LivlgDh0%X%gN3C@qewP$SV|lys_3)5X{FT
z9A{@=M`4ISc}^vg!C!YO;KFZgiJjqx8K$3OwBF_){V>{l7U>Voj@w&<GCxch=JYPs
zw8pz7dotOH>IeeHw^@P+eooo_{}F9-26K$n%iOPL=~$4F@jrGu>8=eywO6H6gVtb7
z-ynmU&$%eQhZNO{n|>E_z6tAfOq#r#G#d+}?@FZVk0!tc0WAD5aSv5M$Nk2;(&)^c
ze4C>im&v3B@kasxQzKnst3w!#wTFGkd%QR3vHJRAE#>LH7g=~DK~sn}KYx7w+iam}
z@<&|C=VOJ*oqlCfj6lkzr+Y+<kET>A7<E-#vi<I$Yw;_LkT#9v-`zk`(xY>1Y(^{$
z95^N(<_sG^$lCi#Pm|b$zA@~n>x{s9=lj2%e0rIzv!A7h2rtZ=J)H?e3=M+A-=_eE
zzzaHPzR)l#G$n%08Rr=E90)S#dFgWrn!}_yDB(6GAhfV%0S&U6Ji-_iA64}M*#MsI
z))*Q3%|6RW=G8nRi${6jWt<+Y7*SQvn2qF9#7(&uM6rve|1pu&^%Cs6f#)*m8hr@?
zUVhm8$0VJHts;N#y!)}bYY*6I?+>3b)dPL%;|_{X+F#zMaDSB|pn8nM!-2tu3#u2n
zye@y~=?@w#?_O(PrM$@u6SwvF`@634CcB%tX1JecPbJuRq_k$>_E5ww*d--XHKm`T
z;kSgp#z@4=Q-n5)X{Lt@3ii~Yg+9L9ZRgjJk)XMjwVpKkk^dSj6eZDeSpF$D!TRK&
z?eYiiVr{CE9)od5jJmG3>^|nlMpA)#65>(u(y!1vpFKJKs_#<yNe6*AJHoE|bvtW_
zl!tbs7MajZPQIJfORc*M6sMM<GzlcpJb5a7axnC<WO(V<8E+F;k6JQvb!Wo!QlfUa
z-S2ZHtK=((ZoiF>0xHQl#o}$$iA3oPlh*UXyG3nc1>QBhy{p4|^GD)M5qR={%4B#e
z`1|N}Ct9CW#97M8{_D0}zOFJLYB7veUK<^^%rM(>)AeURS9FfkSFXtKl0`zTWjG(}
zHqVRpZ!M7Fe`Y>>qa=S483Mb+<oWW2pgb$h!h3J9FTrNX)(v1p#xGkl@lR`;)hc9>
z;QcA|5_*5mN?TFu{&F%siGfl}q}tda9R{Zr5W#v%aFbuWM78P>_SQ)Nu@479kXT_7
z06GE$6f$4Y08(J}t@uW+n;Wo0S9E`W-DLX!slmNLwRZfv{;mOl9k}%haYO27$%*HV
zkq1dPMO+yy6A10%k_~*(nujJGM2LsLy=O=p)0th>)t!l`_yGKXrVLNl2DAyFk&*jH
zN0of&#EE5UhpQeP(F|J`9+Q?gMzrFbWtbA{tUDqF8yc~Tn!Z@6xTGd=^LnR5C{MnU
z((Mw|C3|;cvE)&u?0Fs8`Ige76i@-0#fyWf<zpO$lAQ^%^g<Dt2EZi=E~X`!kBm4}
zeeK6vD!yLm1kxeQvQT)7ilr)j*vE~c>x~T)fFN!qm_Q%0dps_00`WMHa&NMHXxox<
z`+Zc=^4)%YW1&6;*{2(QR(!wQnR5|;!6STM?)v|Qg2M8!BaYDSn;-l)?_$X{Jbt2R
zuvTV9B4xu<nbNO~A$~`fPvDN^Yfbyt7H@}L^)nUql1+)J{p9%aV{18nPLEMFp}w@;
zdSPnzc&w1-&wnN5qJA~+|I?{Ygj6cl?z%Q)HQZ{B8&1>uyHHQPbGu|>m?%W$b$0=&
zE#j2y<P&3o?*&Rgbf_kZ^n$}h5!lr0mdS$coau&#kKugRV7g~Q@C&;vep?gFsyzCE
zl2LgK{||pG4QC*5VtP0x=2NIYLkkLG*vaw!f~DRK41V!c%u}%EX|Tk^T<pYr8UI^B
zv2lndCQg8;rTfn4&;zm8rAR%1Mfm&^GknDWNw8;|*y}R+C>V#;y;{aWRg<^}3-?Xv
z4-5-W4;@yaD~g1!YJI3diLL?+Xyv3gXl(~T4F1HxbK;T}fG#`<S2Ix-kb=Hx3`M<Q
zP(XsMf77;EGoHG{2CAxkCoUS9A<2p)8t#^FR(@eZdAc&gE=k;Y364I54BlM#1XJ^*
zR@-sKd>JFnTMO@X+PNl8^jtKypaKz6VpGKYp~|wH;Tk@p!Xp8(#`CVqlVzK4dJeuk
z1Y`|sIk6L8>L8P)2PBHH7>z)An3VRcaN0P^rY{j8P|{CxS>ec~wjE)WjDp!4S95BX
zd-%8SnNx6?5x<4Dt)jCG-v|6)H<C^#`k0s9+F%yW&HI;DN5iL2=Yx*8r$3dIC;V@|
zYe$puDeP!dcA4G?ck#sS;g*`M0G|}~#FgJml+#;$YW((U$*G~JxjW<W&_^6rkw~_$
z>E0%>RBxxcKW0oOjJZ)#pZGkvUPxzJ4Y=~JuYKpl;ZV{*K6RR2Sjp<+?5sDZJcql(
zUHVgtA+v|7gR*TZ*(zH6;Iir5G_~a?Mw@buA|+0vjg-2#Qg%RC(oX5HX&~<ANZ^BW
zSx)%Krc0{Fq~}q2r$b01D;8Zy{p@R>H)El0a$`4clQ3)Lo0TAz8B9ru)UZHtDya)y
zs&Q;0f|h))DiBN>il&LJZ>Zv{c>rgkI@swAao-I`TnVTEj{yAB(&U!1c9;~B4^Aya
z<N(qDj3P$hhnC);RsiA6zKdy_;Y~JI_=I9E->fBiP~icN0Y?k&--uw5B?53iS6-u?
zRXj?QSxs+z!}v@<5?}`$+(pnK%4rX{c@RZe=ZrsVX|NaTXSD7}^lal`69&GLW+V|n
z5nf0-l)ZMn(xWbih+Rhw2}!`5fSN8%AUr^EH@U)u`Oh%!C_sc++aE>^CQw5U!>Ce^
zCVU$HGa^=)q4kiO#lMy7FKq?43KmSU0rvO92@urTZKoGxc=lVCbVmgs-fxcP5YmUj
zX1UcEvfO%B=KW&6*kBi7Rck`yE#NF*6pa8sQJZ{8c4&E|ao(juE&9=RemO_;nnbv;
z_{hlhFCC7!N5_m$JY9-v0L$Mf<Vj~u5kzG-b8OMCEPDJ8+Txg8*=w=FBNy_Vxt=c)
zw-)i-Vmpm8BgP=a(9T;%&b1Hwnf6Pj9p`Dkb=F?<jXyMRAD7RVLebdT?my^oCr-vR
z7=Pc1!~J{7V4i+l6Y@Vw;BNV(YC0glGOftSls<sJ^=_jnlKjj=#Krch)jZBwN&O!Z
zW^S@CVVb!~ulfnJpAey`q!~UlV-Oh}sQ!6|(#1s+gn7)*>2RFZw+~QS!f-26?dCP(
z((UxfS>M0As_%j9-&n-7AVca=W|JWmjK9hqP(F#KG#C~CF~<RUnKW%E<N7I~09x7)
zUJ#xl^`1?5t-XIQTMFCnL;D$aStTn$)%EaK?)!KP6TNJyFdMA*TLVN^q?xPEtI4=3
z*k5==K>@4-6au9h1p1~(HI2My%1AXoYe8P%nVBu6L#%~8h^4U=^?w`K#NLz#uU~i&
zZGv$x42QS)ePbXnyR0S3KHbEdq9kN^7?P5T5zgDU2dC>7AsYmxKhKkb^NUcQ<2!)l
zB{*DFEsylk%m5*A06h8>jvbu5bzu>_yp5A?4Cjnj&-W$#Y;55TiWaXzr?gHELu*`#
zVE7exZ22pr&o97L%6BNmoiPyqM;L8i*sbEJ6iCB8*GJal23<E|5jBst{;Y@!_)wqi
zwWL_{WsF%vXL>)whBGiMT6}}ig9T57y<fJ{D5tj%RQ{WGqLBkJBeD`A&RP)kaTeb)
z5(0h@nX^^;O`*h0%b{pS7%N0Y7n-*EMBwK=kuir<NZS)+i%W4Bu{I|EyPA|<v+P5%
zq;{*~$Jt}2O4dA%`9i!$N7>J*t+vNW6n@J8;Wm>IXj`u}n#!FX9f*Z|4VgKcnC>bR
zU!(g;1EX;n3ThilW>XJQJ!0p2Czka2^hwo_h~5+z&;15@qw^x!|CJC1VyJ9kV?a4O
zbHi-tl{E)Go|4NSi0podg<ZKL)vNHkuM2xwMl91tyET#!mGGD1bA+!L`)fW*o`6{@
zzaRAltu#q>@Ktu-|5r*B&-yStK83LkDf}sHjI-Scd|#H}f{Pr4jsud#s7;(gmTcsx
z`~C;J315uYVVJ3A3UD~pBcWr6HUO_hQ=9106;Fxf1@I}olzLot@p$Q3uq0!o^%WS^
z6@1^ZD-5PWH90Sst0XEu??=mw{<`fVab{W#e~*s9i@kO=Fh}$1z1&9}Z4U|x+3j-V
zV0%bAV2*apkn=@L;%mc`gX5y!Xy}n_*rWp(<me<we?a;ZJ|^Q<COyyvB9Wxf3|Ruu
zX>zxLAx4%wtd76IqOoqM2E;h09UTnpI;>p_)JJy{R-c2&IoAdTUqJGf<84%l%HDWf
zndg3Z`kk%Y@-OkO6>QZ{!TdC;j0<y0?V>4pHAPK!0Kx9s*Yk~OPs<jOHOQXMZA|vY
zvf6F}0~>XLlWh&KpkBBZ%+8$j0@?jWRtDGV>tg&STT(T)L)eqQOBVj1zDD5iSKU(8
zzH*$0^~tzYZ|@@nF_X(8RRZ|kmq`W(#OH7D`*p#i+xUv<rP|G9N5_{7o^SUEb*^O&
z(x|jen<Z$L7wT(}eNP&wTOsTR6`5Y5BKouqE)fyav*0kQ@j8+aPP84>I_jZ>*<8by
zMj^?foj%E@HZGcFQoQ`7Ayv21n?Gi;T=&?B@$M;yti9zp$MbzYSZqE~?;vW&nO)%t
z7$%cGJv{U`s&>l$K6EvQ(Ty6GY6W2=QKFW-UpdkCG?od?vj*>92iL|pIa?_P>o#=N
zj&2^1d+fFOvazg$L32+;lM1yB<{axtU4tG4kIac}cD(2_9#LuB+*%5utxc|dyXD)d
zI{unKhTaEdsPAVHa_H|a*PL4B@~JG+sAJeOM&Rq7W5EBtmyT<QflPKMgDWli%zyi&
znaM%v*Kt9AiX|{A8Wv4d4%*IWrCi!)r*S&onHXAjYnyb5tVo`cPWU7T36~E4yWluQ
zCqS$l;Y3Ia%4%;!mU2dfs;hXLpw=eaG{AC_@%T&_F>1sCtLm2zQb{(ciCne@-sqf(
z*qq@vJhqX2iwCyRP_L0M{Ik2!Z}myXA&RZd9YvMmOhP%aN27YQ68%#@)d{F5NeHf>
zzPbd5Z!S4{mK%AE^&nB4Y8d^Jx;HZY8s@blvS*1=f*nQ~WPS>BJ7Di?u$e>d8k#3R
zwt#{P@{xrCs2S<(i2DYK)S+Je22>DTM3}I5oo`9ysX)Ui3MK!<)Y?GN1)n+;6r7I|
zyH+6?pi7sPF9ko=64G{6TdY?Mfw;N$?4}xj(*SmObo0-X0EXy<r&;Nr*25;lUvI%B
z2`-68g$ox-!Ns`#9`9^Me<f&0$`K$yM8A%zhu_wP@)1QG*J47$=gPs=7G2t(-2C`J
z%B*sf&-7d;)w)9TYC@>{*EqtfW9!yBsE%L=MB#`>*+x5kR3KqG`I|d-wteRDVJ(`A
zd>C82GI9M>q_bJ<K&H`E$qtMbbXVIWRYt$)B87|4q_xoRJgF5;?;m%PCoJ-2(9%N^
ztrb!7oeE$HCE$tn#|n94`q0dawJFu{=YKAq&LVHB{rviURyvOIZS=NO{!I5LM<ENc
zcfYYO$mw9>{G~Cq__9-~W=J*t1%dM-qsaG0rzF_hzkBEwMTc|zF{2uOMDgF{6lZIY
zzqmEIc(zWx>aRXxZPDVm6C2WKbks(7b{l)$`0*T%`Jf2QqBnAL7@&D_ivm;k(CH1o
z*^4osvB6r8p+PF_ZBf^63q)(HQ0<;>>1qF}5l0x-AGR-~<!(2Qb+EnPqBvB46YN<>
zYbKM=WSl~fHE%dPldsaxy6NB#;U_Htf0T?u7x$P)KABEO{^clj^H#V!Z2j=9Bk&|%
z(8BfK9i3W@o%>sGD8*QkVZm@IJSnfe{JYEi@V{LF7rN)~H9bZaO1`!D4((qZ-qlxl
z%)G8ECOi=0UYPWlxTED5DcvU@*JxY{I3=r~Wn24K{Q#K_ly0C(CQ*Y28J2`twFi-G
z^~n=rCPU>n$HWA!bko)c9upd2KY%n7qhrxaI^^VJeEdG|(;*dF*3Hey!>DD>LKj~J
z=49-#L!empm%X7{)jD(#1;&VOS&#fLVEe00d617LqkoTLZlAdB4g68#x9vf3oq3X&
z?d{WzXSj@c6^(3Yu^b>3<dgi}b|j0m51vhKN!vRr+CHqRr*y>f1sz1{x6o7Nrp>1m
zBy;)jh_uyUBHXv~#{u2eHnA@gg1hIpV%rp%)6}g24A)A)l*oO8#aC?%g!dlZ0Cj~u
zh$%4G&1o~wsP~KxbDmLNsu{?G%OH^P_z{+HU~5Y$_-YG%83#_dhpUuS#^0m!&_&KP
z72R~3|LIoY=Yi_oCl>if^7&ak>JNQ{^3Ol^MLF??f7>kN08&>?kbN@Uk3YN?MzIh^
z3JrTqcv8Urb3dU!mnkpIy}TQ0l>LYCr50#05lS*ke&?(x$EA^XU%@J(^z*KBc)5?W
z*4}=sQnw-#_@=99x{~^8*}!|W;K!f8!I`aA+B!O|1qEjUDn{t;Ur(8TC-)Tp<=b9+
zH7jDDGGFpZxsKe>)i~j+kebzd?TQxJf}Yo;@t;L{V}vF+CYNeoDkY0H{2mF?XR-fy
zt(`xvHEFso(e>^-r=RHHK>Oc=WYc5*d={rz$uh7?dBT&16AT;6!V%NJ%+&8{)|ouG
zv7cI>P76B9eHZ&urhHR>#U+pQL3EN9C;aOe|JqLa@g?{0rC~<C&cK4BTT=f`Aba1+
z(fQiZ*7;!I5FZy!0CnzSq+%X<xXK#(hUw9Jg~ye=8z6TaTTgL7p~rJOTjo`>{9cq$
zmRwS7$-8Toj!TD5$s~=>AquV^02)?8I9kjS5$mlulc&gM*`c@T+<9_$i)tzt?S6&J
zhc47BC#B}8WA~bT%^tWIkRqnE-uC(h74stsC+(t&Yfv6l|J{4oyXZR*V`3C~N2B4R
zJdiHhIi{czUSggbK!b~yaKndui+hd{9BZjZx#iAmL=g5H^bB}DvpQVH++a(;DXJOw
zg#jY&kZI?k;y@d~QxuaM>%_lQN5zQM-?R-r)O|_;B!|PXH$OUi#=BpjpS*>iZbWc(
zo>}HNI_*KCchk$j($`=qti_5MK8W09;3xg4fwgP*A(GXkElo-SCv8;$hbQvAaD{}%
ztLEs?r}H0u=UT)-D$U9KOM8tMme#?BegKsR<oimfs9-|FX21)GR>;xb^=r7_oDqnn
zU+1m07pWMDOzyT#5bO04vtwHM$|O9tmY#1gik>u9vCUu2qOrUgejNMBu8<8ps&K<m
zym++7(soEmtI_zq@H<1T<YL?Ges7$~RJ4Rs_omYYL!evGy{o<mzrl!+Bt{&>LazK$
z!Z<4-ux^!H)V&QrR2{z@yK&9w7-<&UbX;prRp3C^GNAMz70|#$&GIhad>z1a_P0I$
zc$-CeamM8By-eT#z*E-3V?#7I-fMCPm5k^caNC*Muj9O$3f7VhxlA)|3cgu=-Aip8
zK7t;4N0ZB+spCQ&Atd+XZ!YG3UC&zDA3`dD+E@Df8u6}%S$fd{L`%a|0?|z!bzz@a
zsNSbtzgUKfJZ2cOr^YbBA8^|)dnemy&pPI26!_!a^7Y`(Rm9R!^Xwn9@oSw*;VTy}
z0o2r%t9-doW|44S$|9vtLE<=ra4h<}zQMkzL47HqbJ^f67L0jAurbkOrNox9Offra
zfD8I2WBrQzd9WXCt#bCe**%9PH^pE_20tpx@R3#h&RZBuLi`#k<jX{hQ?<XT%wcd+
z<tyj+Mf<o;vGc|ui3hZUafvc#<!YanA&Uzy(J#pno}y<=4O=Y<SK~RaD~)}4-!9eJ
zd<YMId*XZL=YMZgO?k;4csL8GwBiqJa9vp3V3PdZdGf9{7k&EdI^!j=Le0Wyq}kqL
z;IDJ8Ud(HX0m}f{%=T$}fo}&v@CeST7qb_CTQ`4E6dirVK`MYzQB<>mc}>@2pTM8i
zR0hOc&FKih9}CL9)uXXjzN21gM`Z&a4!@j~^AHATGB}=)>9+C^bvOC_+96u=(>l;3
z;dyso=oZqc=@A)RhhDb4<@>XL-1VGnuFG0YTvF<}t{?TxM!X}Vu0l4>R*98>Ng!A-
z`T!)amj&nge>9y{SX<H7wL@^X0>PoUTX8QG_X5S;-Ccr|76`5_uBEuUyA>@`2=4A~
zf6h7If0N67k!S9?*P3IDcea0h5$w)BJ`rt+!mD=wb-D|@9&3iPlVjWT96r%hA4dmK
zAOxPV#C4y7Khkxq`6so2%W}oXCFLA`y6qsU!O11eHhjQwFQTW4=r_n`P&<%#G=Tzy
zrQm=7hC&s^N-%;%p6o8i9|pFX4+>b)7hBaO@v*`t(vG^JBa-6%fg2(lolyu>A?R#-
zKNZ$d^oIx&G6C;{h{dagx#U}|gUX4-rQC&MJj-jA2y^?mwEmv1mlrh_?wLL0SEyH&
z2~q-3ghFC2lQHePM;3SN19(AnoN|}3>b;&wpp?%zcX<*CiPy$LVwbIP6#kjBB$xKD
zU*<2g@1Xv>6Fqc|Wk(;L>eMlz|FwAS*%WY#_#B0JS00`IYkTz%&i3w>L7Wg$)q-d3
zKH#6=Lx(JeQ_m%^O{XhU&GMCwb9-=5UqZFQlJZJMuqhEwk|Mm1N;<hrw*Sl)-<3a3
z&Uj!VOzvDKUbNU6B>kHBUJ<!~1jfane`pP^SlkX?-9x>*Q^U8`kWISsm(}~q&9&#|
z9M9>(6rQnce3Mi1<T@7^u{@Of-IWvTUhb&?>rPM*xcRjkqAknP*&l&MYL61icj|Rx
z%B@MYxGbdqI)t#bu1CR_F(FXDU*}KiK`BO*aZf%^DpMf&dvoqOBqDgPIc3o*UdBr7
zm)PCEg$`kkRd-4!!g}^YU7pFCcb-lq-*A^CUTIrvy|%iuX&s$PRg#^)u%Yi65TZC6
zQ=TI<A|Uws@w2b`FR{FYEm&lJ5ILK_2pAF|fN=Djnb-}}?TARoewKlMcB7of3Y|+^
zZSprgh|vBkvFOv0@8s1A6FBW`UO#B7P5Xiqva~1OZS?sG&$`D5aDt>B{Tj3ERp40u
zWiwp0*Y&Qz4rfm032Vzi;vC$m%4u1$Y2}<^DLu{J#r8XncVs2bCINrR=%BmxNANVt
z&cNr%GIV>6`xu+EFYvF&cOs5I+vh9iB7x>7V!rSDW?Wde8X$+`=$y}Y2#N13k|QGF
z)ZnOKRP^>z06bJJ!R1X_u!(?k9pm{?2A&m8pqgr%0Jc(%6Di6%B!-em1Hc5}Y4*xD
zH1O`pq`L4gxqRwqTv%oiiU)qt8FtToeaJpH|9(G?f5~hnc|8~L$I(w{ve_NTN=3~u
zo?0BcHHvl~-Xy%1dVDsfgUzU+`6CuUF2QrDD`y(%n&y$|%mH}+xOvz8lD#jTQ97uq
z>n;h~Fg*3q3<Jy)6PX=>5B?6Kg=F*-z5psQrr_Cs)t_u}E1h=O#0k&>7+9A$i<RLC
z6*}_IxWrY^K8pdqhZ<P+=)q+&^%APDWg@?^NJJYVNZ0Qyp?jf9;P>oLY7rCZ6()AJ
z$oH$~iSKPH=Hfdfs#g`%W|G#Z_X`2vF_Do!oVJ$GirN6&-UjZst(0zDWszvUxo>S&
zfA<A)&iLostJef8?1^%G>yU6SajR(6Kq7yCYKFG9H5O~S)HISGXqwURNquaTpgfK&
zK$Xywe8Y}|!I_nF1?&1mvA0-ylhj%pEpFj>r0$R{$$4_mgmf5Vmi_c8!W(|FwdC&e
z`3m??(ZhV?-DdQp@9k-I(7se08ZuU$8m6FJ51(%^>Z<2Zm>bfa;qLaNhy{QJR>Jsm
zs(-l*{+OaSSJN2r4-SiTmhSOpyQ!y}GWcB1Ky$p@!iZ8v$UHS><A;_k6BlMZ19VE+
z{~p)7vi?=m_WAXpT7GFGi6O44d>5V6+gss>K9$3{xgGgtI41rzvA6QY<`RBPV0omo
zz%cz}P6tXnT?N;9z;;0>D!>+hvd#03*we}CAW0q_ztWdJp>ApPvPmB{85$ZJ1K<)}
zS+^Cfi5|EV@g{HF79(OQUy?0#)U{>y7t>@6P0o^ddX1#amvem`icKKgqCMu!;6T76
z03^1kj4hWjb=N{B3kON#ItqwRg}#Wg;^^}0+s=dPp66iWAhPHOW~YLsth~GDiw!$O
zgWMafDvw5M6&>1LuwljGnH(ie2+fVY77+tHyoih&2-9)kqGYrK-<6*;3ZDp=3Y0)9
zM=<kPH!S@XfX*qF3@6HAE@CU^;!H<m4V*`ozjE)!Gnd(}&)p-FdMH$h3>w%iIf-j8
z2KN5#cXQ-_ZUt>&4*;lrgd>Acua$AU!6-}p<OE8pz?i=6?)y`CIBqPJN2b(%w|FwK
zlD<0Gh(JTqIMpQ*|J523gI5GV9)}J)0J+*pzn?<=As>LbSlj#!=mDq)$N`7f(^(3;
zs)fIQ`lt)_gA+K3mgE8i#FKGK-6IP~3)2EH@NyN5O)#nOFbI&g3@qex{$9XghM?H3
zc|<v6@-K4GZbs!^<y{Z?((s`JzPckL?HBgMec($(Z>$UwXhh!9YE4)d^ZyD=OhcgA
zA=~}l9naOHk_@OolWih1Z=|V4bb)Ui&K5zwN7TIAwCE-zkvINyc7HB+XQ`-!%LRE5
zdpFkaU9`hd@}ceL+VFLkReYAwq17I}Z;(^;;J9dD=8LKRwm}owhVNYK)0%W}`|{~p
zFw0zJ*Lrm4@YDX;YpVU#TJV0;%Iy~wqnlfmp^gjx&P{u02*(3+aE8oeUylU#Vd_(w
zjQ{-L4{b0R#Ur+wu-D3e5NoC%9*;aJ&njc8erdg<%c5L%(tA}l;vjFiDAi<dGc<3&
zVSy0JFXxYavM}=g*RHKjL9sR;&TojM{9ikie8GxK#G*;J>Fg&>STfX+Xh8xmu+{{k
z)nwb!^aL%t>mV}r3v%74BVm~XP)7Pc-f#B_V4BdLjNHAaY-iEZjEb@DmWvu~w^F9_
z)&BfTCjqwon8?c7i0Eb!?e^S9=b!Io^9ea2;wCbv4Ev|!nG<=|JyWDnU*MlB;Xiw8
z>}*&H?$D$8x*OPIbbHDa-B?=Kn3PRu?Rx&g?!Vs602=ca-uEtiwp27Cfc_AucF)I>
zG0}HOUS?kLoXzlz*OuwuUf^s(&__DEA?=GuwD7V3ncF#bnY++X=QsfCUp}W40Fzs#
zYN(%mPcr10*1(6D^L>||TNSOE%o<N~@DhGO0rQ<EQZmj|#@6tnh65JlhOT85`0fWH
zq+UymL#<dN<rpXmVAXT8{hHZaRyx)}#9meE7LB^Ep|Qpg743Eu8yl?!U-llmPuc`y
zE23pO{)h`N7tI96Y2XrQI)&*+^FyC4otThyrKMxW&IUxu`3dD6cE$jhsj(Z2)&mI~
z>R(gZ&5GQM63;b-Ale$3!iLri940|2cZ+m(lE4TAEIzE?w*WAR91O3dZw)mBp&^?_
z(c7ExC6l3q?YjN|n!;%df>wW7Tq5bcXb`hwi`|%Qf}V;~ufY?EzB~qjK@zYLrw&2w
zX#af5zVRpXuU8A2?LgV{8|DkeRHfzhTiX$8i#d1_7?*)m>eJ~;<xwUh=e7gC41D|Y
zJRBq-_h#3uTh7H&6>@xfew6c2|IC=}x0OiYk+O2jA?}!|YxF-!m#-+KLvopvpYpI^
zchP_2gyR=I%yp*)gI*2r42APsX^_ro+l4wwN=Zf?3H9qr=SZSmw8oGhO$~?ymb*i*
zCgS4tbCy^xibSSf(g``a<AR+1h`|X7gm7@uMGvexd!8BFWbBt@3Nx>sB}uRh>W?(8
z+A@hPR&QU^jvuc5m60x;<ZHW*f7__-8h9vBzq8po{M3TjroMiewl>Hpw=$YXWN`eJ
zlQ#ZknS1;c6+v#fPkJ#>9=ezxE1TucLPQZp9=9x$0&}#~?_f0*+>`7@VItYU=>sDW
zAvk*<70VIR--B-N5QFZ~D>aQQ0Z5#wM9Fa-<T!A*NFyQCNr6LzKM(*z&Bn<+nLeZd
zE-=yc5j^~BN$dcM?P$C)StQ+i8etaz2D;ZN3W|W1k#V4@&K`(rEDv%7^JpeWmd7*1
z=>?hN>jIGX$23mll!&;fzej7}PJOgvu|h#T-%$f&FPKPS;DO^Kx+#Havq+b@lniyF
z%|c+@ZCiA?Z9uMkFk}{y^h-}@chb+xWw2_gFt~wQV(>s&nHathOHukJ^5UZbF5WTL
zL>RnNp{jvJkBO`AaPdWfcRF3a!2xcx^d2&i6>WE7yF`y@O^{7Ah-)0(kbKUPG4m*I
ziQ?!fCP_~~Er2lVC{H3<q~@b0lIS{n1%rfLc?;HQ9Vq%p>ibwH)Ds>ftIj7aaK(As
zJ=;L(-3h$W%?CmzjSJ4n4*yB$)lY7ZILrMAa-FG<3PeXTr^Z_bGHqD=_YC_?RzE?m
z-PV6QgetO*60NryPx808u_Z+&lYIW@)^~ZP@$<zw?wyFA3S~3}&0KLgKb0}cK9?gq
zi^(+{Fb<-S7fyg+sHZROJQV_qFfsWvxx~EA|92cn+pITW3k0`McD-EDa1)8c6c_?z
z$dr&T3}95Sq-T(SQc*KOpinJ#+jgQ@X4~|I=Z8hcRJU3;3gwT7C<q+Bc&+@-cfEw4
z^%X~Y94L@QkhiU8^GN&tZL9bK(VVR2!~dy!nmJ`+sYsx!>bZ;4dG@-fFt=|vgh&Z}
z4?V4I^>4*Ma&iCtr|UJkmL~Q2YST!;1BufRG&9t0(2s@cmKbw=>EdR8_!r1KbQjgK
zCv-Q@aF(`I06GW-K{IEuHL#%Z5<nisy@{_x*4Kn3*woMu1%RkG22<U@Xw5@mwx*uL
zzsOFj$TFk`eW5~dB03^iqwDjB+coLTW|1)r-QVXd7266KurpRq+K^X?k7MP#y=1?t
zQE=l?hQBT%BbWwAjUeVqebnCTRlVR^`bl}Qj5l<`P}_jmj}zUO7!z|(&#62`Yyuyt
zYz#@18j*$)r-hT&Z|k8&U~(abUa0l99^M9;MENlHYfu>}=su}bB_cLiS=Jc1EL$p(
zna5EPo%EE1)WDU%)eT*15K{H{o{W4=bMGl?jMSgO4eRS>g0}|nWG;U&#BW|0J$isd
zn85YK6AynzUq|NpWaG{{kBxz|NF!(#JIt2qKCshH<Prmlmf16Q?6s0?%m*j5c{>T4
zSV)g61WulJhH+_x1SS?K!NXxfzXa^;#bLiU^j!=0^^zdaZ)sBi55zs8aQsn)^|~HV
zUei}t4mC<(A$ntr@-@EcZo)YogKqhMNg<dEeU$h7xAD#DMNzN;+?f?JO$MDJtivHe
zr0pHEx>4|?vodT(yq@;?D(qJ1lu0=8kTu!43DN!9<+5xo`v%KwExqYD&_MFc!jIHC
z_69C8_THpoD1DQ8wrhy=cOv>wR?|W>r0C3FtM4s~E0&B3vve){gnnve2OQqtoqDy-
z_`Pi4pM7mBy8&`4I~!O(ar+?~I{qnGaAd{b({LcQ#2l?Zwwp82A3Gfq^>*llDb8x_
z)?2Hht~Y5Z_MdNltq7OT41p^c$`^21`XF1whnwK<1Yo#~aDN5~AXHfkt879In-Q=s
zQSHhDbsq;tCdra~ACTx>TqQ&|Ggd-Mf7f9()DqVl_{9`YAAWG?w}CbjMub^!vWQu1
zJb>Cn<J#i539yS0CjQw!1s+6osK^mjyPD>hf=Lhid%95ugGS%GQ^bFCs;^#whrDax
zdFL{d%+BnV_X*$&Z0d`U<Zpg3!yu@k1SGV&xgu60q=SX4;ak>1ErBdZd~(WexYr%3
z89+)D1EU!MIKA>0GxcN45NZTDYZSxHNE?V`1bXXx7eN!08e@eehCGIJkPaGgTFVC>
z7kDf5a-3}>Y!W+48PwWD14Bn9b2%JOViw<PL>zZ^emhRv4<17x!85mtd!f}n@*mJ}
z+|L$D<G3ModuGi$h;EP_4G|80T?1^e$@K<E8fdluEm@q>$NOyMg2b$dY#xjLtZgX<
zGQ0aOaecCOEmKkzc;!JEv^ONmvlTlzJ5<hHy%Ghm?e9@H!}4a1|FCc7u1%`@>SPa_
z7tX^XiJJb8`+0Z8Q^98~-pVYgJOAfhpOuIr0d?I<p8ac+U9D)lOLgKs+!_`8;An<t
z5F%@oNzGSUQ3%A_?<ij{OC`)&*X{Vwbi{{O_<QohO?8ap0I^h~`h2~G!mQfl%K(}Z
zhN5lZw<vK>{&&I5F<$)tWPU!bZVM2e_|LmDGTr0s#Iim^jWy?$Q*w$57+Xr}{y-8K
ze}oMsf@CeCt}fxc^mUv!_+WkPb?+a{)L{FkweUQq?W(@c;fJLGH{0wT-*`r9LU48~
zpE0iqZ2;z<S=bc%2kd)lVUXNMGfnnvvJF6LUy}u6N5#O6ZOEOLErA+RTrze*Sllj0
z!D%Kw7#5z$(p0dwuzeBU+HK_PMh?{`7{2_O*DJEDP#h7C8NlHQhm1|I9KL>pcD@0*
zpxe>S$KVF4HRo<DRhe}!r1Bl<q0A?`zDhV{l1#{PP)P;4abRnvN!vrhjO*bG!oASl
zcm`6vzK5YM!e50;KM};_)Fd(viA!KepyLT~flCAg2VEcYSS8f8z<o4uDwZ|uQ^bB0
zQxP{b-CsfPI}636BKY7K4XC`b&=*;n|AoSFCz7%X6JM)7V=q$e6MI3>rK|X0uv5Sp
zfOF2Z`<q3s7z7`lEk4a0z}An*qy+w*ZC2Zyl=s}eSsnlA7MH^keelcu*et63W(v-6
z8xt9gvnz)vSHpu!Ar{f^ROl)M1XJVJ738~dAh_PaQFBoH%oY&Mk;j`*^bIJ8lB?4<
zv8~gCKI|->^F9ziyW?u#Zkn6ipz=Vtu*|-*Qh+a0>U3GQN>oQ+)pDlgAfMZMXLb&7
z)(FGHQv5f9jbC&+F?L`_YtQ!oVGw4#oI%e~-oHu+-c5F~HIBCv!v@H=dA8kh)fKSV
zNspNGVbI@}mfx+zgD1h^R|(xiz~5B_bLJH&5O`*y>3_-ysf5$mmOLi3dglKck{~NX
zSAl6E3HxQHH3b{q28$)+6UOB?<=5;h85{26XaCGqmXdH!s*jO-R#UmE{HXu@NB;U?
ziNG+3GLC_91G>(r=IrO*w?p*OJw|TiL@+@DXw+|$taJnZO=NXITg(%=I%rFgk$i#a
zecUcwA56lB(<iwW-&aeK-RFzFEYjMk!nMi|<6?BK5<S676!S*Hxw(|(cW(oU5zMMT
zH6Dg@ZNlP(BnS%p-dfe4FvfEOBlq!+F$g>(8v$aWs5JJRy2S37GQvoOE*R?tbI1P3
zQMzuA=~O44a0*gGWFaFP5_Bnf4)9HSKqkFvIfUeO-J(ja>%zCd5un6Z2W2?vYBiJe
zbv_nV1Ka_CWEJ~w90y`xE4-6{_P0ff*qZm?{?Katzggiyj9;{~5<G@FqqFiK^2`If
z!RpIkSTXs)<(ffofmQIdnEm&RXvqg+e(%8p#<V<#-a9H7ce*5TxAHUtlp|>#UB9tv
z<Zcm?)0MD5xCPP{ptB!)*-47^gLH%62W;IjErYo+!#@C)E67kfJICKgd?2EA81ttq
zthB%4@u~D_vg{si5N!A-@qd8`J<M`6P|Y0gEI1*8o(*{}Y#2`ENiDiOo;<TOr0_9w
zp*!9S=KCAr{=5K)cx*HvTQ4^G_u;I8(^-7^9=GNgt<w-FGb54%OIckRv1nH^%;RwZ
z8wEvAZ`zm9$VgS>J^l&^Uog+L+B4hXer%3Yep7LK3$N-sN1Lc#?cM2n$e0hgp``dm
z-o>t;fqvTSg8`b>c>GEnS2u`yxBh$Y>SP8Mn(YJ8m{)99jYzDegajFbB@`p>%?J8?
z&ezAw)=D41k?EL&e%(&7_AF4Po<`6p2)iI;eFebMtO2wj{?24}u-<&QB#J`?KnY2p
zXwofgG?3PXV8wdE*&uP(v#j%TDFwiAUtb~D+_g#Z_~*c>umA&xjoRNO!7J$TeH_uz
ztawhf@he>X9_07AnydE1hBW=07L_RVJ%0!t`!-D+CJ2%nufwohiatH#ESa);WeRPl
z6%t91D*oMTmSoe;X5&t14r2WEW_UO<KvY?Wg^ieNgI*#%<pvTwkypY2@xZQml@mYQ
zi|8Z}V0VFPv~@(P(WO(SS9yv%6E`BHx^Q=tYM^<z$tkDucbd@X>h~t85^U+r%{y+A
zbN@=$Iw#olp-wSRuf9m{LI@IPc{%N#)`zQq0WQDRR^!5FYsxCUbz}ysgu-Ct3zIt<
zWjYCEd?CD2EeEB_g!j#Gmce|bA?N<yNxk7G*b6U)ZSscB_G|6SFxZg(qze6)|DCZL
zmEd85r9=8gWF)!urRLAtkD||cjnVi&3;){9s{dTB7PIxA{}%A`#oE!a32DXxQYbp3
z6=3rRy|JITDvf-L9gb!!mRCpoQ1IP&m;94!WoP34zGk<aGP@;bmLn-=ZRMR3><hFf
zBk)FVXrDo5e=p+Te(sY|bvxivvYQ&$)+FD-G4o%Lut-8r;D^;KJGUQS)hX5iVWOyB
z0273RZS_9kRzvS@1ONbx=zy{5^?iFis12~i2?Ab7*o+RW#6QvP<;$OZIWi+wa)IY9
zdUZ2W^%QDO*LY4es$<EOiSQ7;15oalF{K?>6Dx4fdy|r%ef1w&P3m047Lvih*3gyl
z?ao5=+Ju_Gxh?u!0*FSpLFIT7s9_vYi5px_Ybj3AMUw08i8p6K2(MH0?A71zOxKZO
zrk!Nh#F0wX5Ch+R1yHdZXo5c(3IQ}F(*XASU<4c(tq#}=Z`2(@-BV7zFZ2lX=z-a{
zm15~wBEdHR0x)?pwR(gAHJILVK#96x0IjDYC7LCF?Fp!ia;nnp1I>af?mj(G1*hgg
zT@p+%9=PP+tq0;2gy$X?6|`;xXmRajYGiL|g}JX)U+#H1r6XQqEq<y+8`!_*WKMjM
z4_pnrB@QuXbEa+5e@a?Ef&m?KOB@!^Cx0LNb~?J}A8S^qz?<sho;6~CbhA3^QhEN1
z@^Gp6P|X5sCmN4PNHEy6H~gl9R+?+)R+GXo0eVt#o|prQzV>yjHkLM!D>T(H?lB`)
zKeY3kUWi`hJyP_C!FqH@7<1~EH4|2*VtxS+9L*fL8AN36qo2QN$5$fy@T=J)-#WIU
zmv+M&n@}fq?UjZ66LUHH(R~9(0wOC2>(T<>f3EYnNx16~kR>E~fY*9N6heuqguGN`
zxt&>plEZ1NC;+|JKnsov$0BR}xLV1CQH{`QDAK^ahqaq+mJN=xTFhn0lgeDQTiW=v
zopUh>Aj5h8<+lI*@3|}tRSFhWq+6+$U-_yAI=D@6YQ@aEvY*eOK=Y|!FynPJE7EtI
z@+^e0Fodwesz6%UXQ^fe3FF;nQaWxd=nL(&$)ky)Xb$xz4^yd`BB1pH0$#JkK2S*x
zK#mCaei#7W0tuSwzcVfVPVn*EliZ6-+hjU@jQy~8`+O`0fgRGmRS*@u0POZ)_d-6c
zn&R1_ZGL({=r2KM-3Z8J#t}KKdNfN|55?$!2d^Vl6r@vw-_vk?WYtE!kY~r+k3twL
z6RVsCB@Hk#0Rq-l%Cx=o`2oxtvReos;tw8<OcQb{S8daB-I235mR4FSA$5vhMbB32
zCYfFMD(es3cy9YU&HyrgGq!pc$LxbCC&z_Iz3!zLpnoH?>7`Tc%Z-GfAAH24uo$o(
zxkb$X4zwohm<~V?G3O6Dmr`xK0b-;KzaMkhPJehK3|Z^#a;8nQl%CEp(>mC@`XXwY
zUH=U^8oYXnzYfBbHCZQu88i!G0u+B@@=Cvf2kxM6NwF--RLs9IGD^JhbEKjV1tVj9
zj0tzlXn(!sYs&_0np5g)LSj<iLOCAuzdm6fu{}dk_Bf!D+SCM?<-mO-nd)4}UN{<V
zMk0b=aD8{|g&$U%{=+u>2qoA9%F8(Jtt!X_NQ6Z}X4q;7lTW&_CBB9N5UJq0QHeYP
zfw?B*48xCVr-=Z13C!{CGmJaC3?a}t95!ppUy5V2*k#j&#wF4w{Z*Z4bk6g>aw<OL
z!0ZhT1*AZDi;}TmLRkbbun{q``+(3sd~H}-ygrO9mWvw~Iipibxc;l?NB?2#lb_E>
z?VRkf>XEMx${)W4B|YE7!ELKej<9CIsZo(0x=U_rVeo4e%JY~_F)p?(zH3IIQLZJg
z1~!{P2rsAL9MU5UoTycmCFjtEA}9bMKY=2pQSQLcRwZ7B-QIAr*I(O*CmK285ttUM
zQ0jr{aN#{2gcov`2=+wxZj3Hv1+`KyWB3^|$9QHbfS_pr*NNO-Gov}nI`gqp>;;~C
zA!|$Q8HK-fOoV~AR!4$?tycd#An1VkP5Yu$QDMR6c2CAWmnA~o=)%!K<y|&DYd)Fl
zc>BvpN2^>%`LP3LI9$a)-d&EKHb;Bd#k+49oM5{&a`}HJ4-!mTM45cZ1nlTpG5CWD
z=b`b$qeJCafqcET!oP)>*tSTHI^RGG%~8v<^>elBi&h|(*oEB!2&DVxAH?~F{s4dK
zcZoz^+!Fbj(;PM<an%o2>vh|&v}Ncwc|T>&Z_bBKYP)A!CHp-MaA6!7EH;KC^X#~K
zEu$pZ1`JLa&UdB(JTww6*Zdys_tF2K1%L}ejm>$X^}}qciCVLTulvtf+h_%wq7%(G
z&CxyOZepeu6=5+Ao%%G-G4ZprHSrGpY3aL+eS!CU)7<8;W(4!_TXb1}6_;^ao2p`n
zxy6K8UN9az$5=6*xhuy?jBwOUJ)s81J0)?`ltgNuh^M2ukCQmJ&663_hrB6w9QLsQ
zRP^6!Gy!a_W{~8`8__%+d|C+%tt-DMbBz>XCcGRHEQ2X5oN#ndCEL(=1Hq7M6=HLb
zSnV`Fdo=(bV;r_D4g&}_HRx{uB$uLE<QeIscQA85H|^k+F*hRSBA9=ozl1F;YLxdP
z1NXr33t_p3Ls~1lA|S8qE^W+D-NV`2now*Bl+n`chBzk>iv;Q)c%11yIrBrWh}94S
zw}iML*B@VnE2LW1%*ywBtQ+LCwHmL!Ldcw>R^au+MYi{&Jb_XOcVG6C(zyoThnysR
z+j)QfwV9rM_dAKmLIc1?W9or|ku$&u*1(0VSD6;{`U&Ac)WjD3ZIGVhv1-G=_ovV{
z8l(hDeTzkh=9WXoBEhmFpAhloYX=>P;}jyq>p?`QVNLS88Y`>dG9=PoNJ2<<fnU_0
z<Rc-%YHyTA(sYk!eM_`p<jUGATgNM>F<9M^jQ(HU#oMlbgqIIVh(-bm_r#TXwVjCi
zTq#d?zl=ELE{?Bc!pkeHcI;&@ARl@mPh|<FvoYa_@aw;yQ5a~xc5RSWy#mRn*#xna
zZblWA%DWk}+?6sld!ao&7<$E*i6M^x`Q-<joJ=ezyoC|^=~KRd>YriozC;63d5>f`
zn!1p<Pd7r7qDkA&5l4j&0ev6^y^4{(GxLSF)!=|#(cq`^8P4%|`d=!m8?l@+?3h#=
zFwfWE4=9nRg+bNujn)?=XYDihEYo@h(_Ca7Do@M(L7h0$Y_=U2Y($ffLc+F8e6la0
zBNAX21LuUFZU>r4EDV)Kda26LId9*jW{Uo*>f8;s22UMa7~<}q;(jPUj@0}6yU<If
zUS~WB1TX?>Jc3$y0Mz24MpWUo`2!XJYf;MjCnX$$ApN(|1PBultVW>rkrb2oo=FO5
zL1l#^R9j0XYaC?a)#V&J-DF-_f(00h^a!go%r`9>8H@?z$~70~8<Bn!Ql+cmkppNc
zXK|!c4#!}yFLnUv%tHRg1@dPD-kn51b#2Qn;fU&&@zMJM)QxN{f;3t?#bp&Qfgu)P
z5{x?5Mpt7;pc~xJckiMtsQ68MN3isz_an>fXVMNWtA5>c&>~l&sJ)*OdP;S8oj7|5
z7-^Yly*d`HVvA5SWT$B{Qv6wF#U-O;zFf0YDz2rJ<7j%mcqL+2GwfyK&{U{1muaJQ
zl5dzz0P7X{3%c}<|0jDCht!88biA0ENzf{wK0ShcHU@}zTZ}QaNX*xnF+ySz0%(q;
zZ1rbgL{7&9t|W7J5)cZRd^8tKRQZY2^~ctVSG0;0^uzjH-DYLIWqS!b@ev&?ge!*s
z--7rOf-Bbe=U1>iMlY?O;Si65dI6XiGw|x-_{jE)Li_Df#_GjY(*nh{2nb^p!`4tY
z_elRkaE2EaQAbez(tl{I_}9}yN4^b_Tq{eQQlf(%W?I_>B4qIXrg*{y9BB%cko^e(
zcK9~4a58jINYRwYQLEJ;cXblhyQ_RQK6oW0qGk9-@T=*wvkJylGBunX078}wC$s%g
z?FD`bnmC=+eORsKF)W-IO$Wg2CU$4}WH_3n`aZXXi}xOLxW^gN*1XDPg$hB1mt#a;
zG-2;53A1NgBm<KmiaPHv3<#=IjKJR2$#ks3v`PuMlU4)?Nni<TYtKic(5wYS1)Y4_
zZduhA-6^k2zkWk%##V5QdrU~M<qMy#pb9iD%B&t>lqF_ol=*qHlWa}UWgX+R*2L!+
z9m=RMkqyS${r1&vgggyMe$)>&o%YT%a&^FgPTed6J^Q&2m}O8+2p9U*PFf0EKP^#D
zL~YXuJlz{yqN#CfqsbAm;wfr1`0o9){u?yJ^{Lrt7D>kMk&yLI3nQ&LWbsE!tLU6(
zFm0;plI?t3`g794>Se4^^WLJIC)XQbhIzn^Hk!}zq4qY>o+`#$$KXtah1-NyMfl&v
z{|~n{7lzb>aiJC+zHgoUe`&o54sNrADc)+6mEdQ^0+Jfi=UP=r<nTc)Dt@|iI*qs>
z*l*a0)nr~SU;f>Ha*m)VIrTD{zRLQEoVU{b34V_c-f!L<H<>b!glcNKNFxr^hJ4Dh
z4`}6b+xE-la5Ti`+XqZCFf5{3v=o(pgMl2mO+q5%9!$M>If<Z`YzoIHGtdH+)52nz
z*pFSN$9QZ)C<b_RgJ3>K&uH%EvHevoR-8n5*>Fk$^VY-l!{5yGsIdA-`e^CIQX=Gl
zg*O;9lWm3qaqDEMs022o?R+eoeTTz6l7_ijgu1)_Hf0RP1t5RZwZQ<>&=7CBo_SI|
zxTfu3lyNs2fsTIvLIaaDUo!8$d$LH3(<cEWqatly1IgY8$WVRL;Q=Z_xMuKlz@6o+
zj*ai?S3_{zSY)a_4Lsq})y|Z_h)%zSfHAizar&h{2W5jHWenuKOZ$99q%V~y5(ffP
z^5Sw!WKk#Z0y|;qiN=-Hr9crI{jdwf6cv}~PY#3<^h*%~J@@Wzmb`c+ZV3W8di8v#
z*zK86v!2UIMLqz*(6gFoH~u#H2h{aSke4WwaV#dzEHoaO8?hS5xQR!e|IYDlTLD9R
ztOZz`#5|XY-np90QUT%G)J%MdoC2VeyB0V&<-<Vi?)l2qeKh^!SQYB<<4^NQkoLEQ
zAlK)GpmpcgjgIG`)rP;EIU)h;qQsF{zI!`GOC!wI|Kpg<c0a>Vv4W$#1dFFXT9`d~
zBiW9^sM7T+8t?8ew$}4fj?<z(ot^S{^X6m+oFMo`CbGnnw=K{5C?s7V-2JXDOJWqE
zySp*P6H!ITi$?dL)SM{(QXn%>i8?y;D#Cp!UvlF{6x8ydylL*a@cF!zY+tGl$5p&P
zX4+%urUzNDtIf?megT<_h45-8@o9aHUX-q*Y0IM;bLgAipVe8rrND!ynLr`mPONtf
z-?V;R75koM_%7>NwyeAFw?Y!`kS5dCeIYX&?PqDZvH6vOZF~QW%B@nLqH0#nEs1lR
zn@+N<+}a-x_Z$`+o5(@F2Sr32Bf>|>exp~;o3AMIOS!-N<st+j<x<2Bia+^uUdwN_
zQz$!;W)2yigV5iyn<%fOD702r$(Oq|<)sK`)m_SIMAP*3)K!@&|L(R{h25MyFY%@G
z+l${DVW;i1GwOX4ZjbWjL8n23TQKc8%B}dM`jnK8Lr)~}k<u%C5%6lz`>*a#<;itK
z=3)!iIsP{37nO0Y68pWlTL1}X=W=s-?2EP*GN;a(Q<)bEXqDkrUXe<Y|ApW%?Qtx~
zK?Pp0#Z}nB-(T;c3b_T?k;Af!;z0g}MvTe`wAmC2b>vZXh;Hr)-&2=AHD5Zl387;D
zn{71bDr)q$#b+#EIz~%UE2i&u5R=p%_RdV1ofP})o$_a;(y-DqA<mKJTA?d@Kp-}+
zEQP1=YS=f1t1p=a+tC~yfBbi7g7+0(!-Ai^+p%+oUyFvOS1AK;3%1T)8_!&cjn8+^
zvI~CSzmHoB=o2PJk(>l4_|RDj{*S*Yrwb3D8G4BIe)_j()i^!`Q6GBnp>9ygSlyD1
zd#1F%A4C4jGbvU^AWcwV!qeCE_G5td?;F@+a;*uRQYT%2_<CipLJuW6nJmqs%~`np
zK>QM*cPpO$z}U&0rVL<`y$n90(WHzmSK~1re5s4@_6bm=qi2{%>3safGb4+_(#k&b
zS&$v8!ewZp=bC+<0iP<VLhL8|tdRROo;CNQ0TAT}F5>S;ZcuG&LjR2<H|N)jEW0Ap
zA^|QB&2*R*aF%#Sd9u+{3qDvtrIl5-T3lRwWbIerrR3L)s9zn`cOQal=KF4*x$c8s
zQk$Lk<F?mSO~Jhi8Dv2nzk-!Dy2vi;dXurewf|kVKs5j>A=+Ok$niM!BPe6OZK>)v
z^4X*m^mICiH{Xn50-As_^s!>1;`3NT&)Ls3^J&4yw!WhYp*PPos&O$5p`Voy?R#sb
zmbJ?YwCl|peiP)}gvmQ;w&g|Dh;LaG^&X3K?d49X*CH%mT;+)6Jd2bOu-f7w=*N*n
z7%w0*FxfkLS{`3XI<!r`EcIM&FOv5Np7H%$)rshh(l#U7{`fxA=;1F*8kVCd=VfeD
zo7S$cXj5~xp~+6<$4GMt3^<HpkiPCgg_MzGS+ru3#JI?U(i>Za$0={&>}BVdsyR;r
zu{m-D)GjH(TMwt6s)K_AUak%(`*y?{`7&!=l{alYkEr32?q2e3=5h&KPtRSuIU)}I
zh1hw{@~d!<-+^7F=DY#wmZDKKe(#y+rMyAe>x*AYXM_1Nqe-FcG2+Vo_|{{4Tf^D`
zW_zjjb02_lq=EV+F`e|vXzy{iYaPlj;&oWf8R{S1-yGw*i?xsKx!+b`?WJ{EilWfv
z_<LF)DT4uy=43t7bGgvKC_9k&`N;Fl4`T0_`<3t{7W;4@<=?<oQItU)d42nak>o-d
zG|IM9508_Zu#6Ua`I9SS%TP1&tcsU2$CA^%Zne#eXR>(tLOVymkZ^}y#i6L6=#ALx
zi0Z*VZdY$YYQSW2vvaJ;h3{y~*?pULN1FH<T~=rN-Jrg+Z&kppR>u&-ah`mbuc+~u
z-@LLSd`+yM9tiJ-LSVKT&8;s&m`&qeaJZ(l)7L~~fL;_H?i-tvnnQ~q_wMu`vOfmQ
zc)av1oNSY$$D>|cE4D`at07Toh2Hv0kc7$Lp|MqlptCR)&)_rbpmG0}aK}$y)>U7(
z#Yx`IvES04E@p^=@|L;f2h#HB=<DtkFzC^HEjcVMCn!X3C)&d>K&Z^j%l|`pbVu<5
zINdnKu3}H2UZNKPjn?k>4}YAmCKWeg8D}s9KRCW$>GUrjsd>w@azk2U`#B~Z#Ev{f
zl$^2E{yX%Hh5loJNJLVD=?l^oTb%i@1ZfXnm}LcqD8iAe2-FMe!$_*)huIcqo|O6e
z4PJeVccfjQ`0hj59cNO;Mx@wa>3Ctvf`GYf&|328-A8@0Hpa5hd3w5%uw&XKO9Q^0
zLh{?N6Lt3VayFPww<R!Cu3Tn*^25NXJn6N9t#O!OzNVqa6)!@Dr(_yyt2&6u^yHx^
z=P;&qJC9#@Z9eUw_u8zr@%Fd<6Mb6>^`wT&F}<uB-$I12cM#^NJ7jyJ4)r8<Kd!bs
zi~1Dj+VvynpTK8H9xU7CrC(78NdM#@EQ^IZN#3X0*C)mXeZ|I?y(3Fo0w>c#3(h7a
zzSBgO%2mm6iphE8q~d$wTnX{Zu~!))6B%n~vri!+8r5vVX{DRHE%eb1wvW|re`R;0
zpFVX3GCc>m3Y$IhIqG#9R@Ix145|{(66nw7jY#??#RO__PKsY%r#BFk_lb=>){B)_
zIx1e$y{x^)bX5{$&c;~1^{qYSbi_W3UXk_2<C0082T)nND9x*$lew|liO`GdXQOYp
ztF-+^kNaw#x_AZmZG=aqOIiC+SGNpT#5mP@r$V_&3z1!NEik&0|Kus~0EAdBd26ZW
zc~{^y4I`x-&+^?N&(vtlpB!{|V#svoed6j0^{c+RTaQ6ze)`}j%z62IRYzP7p$OF3
zTpC5Ll$t;ma$k?U<aE*(6UwhVSg0hhk8Ye6en_>6+wyt|e`<2Gbv_ATUvf(H+u%aF
zv3NAljMYl^b~{8sIKjMypu{f6QcIn8n3>m9j12Q`Yv-e5jEv3U?4{d--7>v{uEabi
zt$BxcdblpjIV$;Mk*mlpHe2ulstm7xi&S!J_nwVS+252%<vTtys5_Oo2@;RmvIx*;
z(|9>*%x{~D3#i2E>&EgBhUzSLm<GB+|B%-|u3Xx%#Jo#pgJx%m%Z&T<p%yEw#u(o<
zZqrl9NkRx&jwWITzp~pzZ#`e}?l)Q0X`k9$J6`In^@zc$Jk@zQ*C@=>LEI-usY+5X
zbdLq9?egp#d_Z?5{x%l;?`I4&=)8~eHfYm2XfOt!JY6jGtMchOuS6#~c4~|;#b4QK
z49DCtmoCv3)2#?9?Z`>2OMlY<uULVPqW<FV#jPO~Nm>h3XHD)Mq>;)Rzf)G67s2ED
z=z2Ko;#n=?wts@(W-)H$gQ8ypM{q4RD1N^4wsfFuDbFT%N4y?dY6@v}#v#b(7;~Te
z-u=hvuHF!SxVJ%)o4B0)L`jhz`(-Ie)7L(Nd%khxc@v$(XYDSnbD<1o*N9Wi_7T6_
zdK(w^bvyh?dLc?Rj~=PRTfb#?C1;YYRlfs;%CDn)+c4J+c^6+|IOLM=F0w201s@)D
zO+2};Jyu)}-g-<=%ZlA{oxR=v&Pilxi4=B9dAa0x(Kz$(8Zri-@w?odJq4dEbQs?k
z1s_F3s3dr*z8>9UnHnJ^5z^x_`G`NVo%p^D+`SN-InKx$9HO$qc7$HU`NU;s8i-_U
zjmnT?uj)7N(=<-jGToP>EU$z)B9HyMC^45Lu7I{Lm)vTrd*c$X@JKBH-NjZNZDClE
zoX*FKxHi+*DeS<m<!ns<)zyP${-DV(0Wreox8A`;Hru{u(nM2^B~Ukba1Ot$^*llt
zSgSB%|7T%q{AxsagMZQ8dfr{hmqifgrR7<GI{*H5tS6q#+s4MFf6jd-1-HoiUB%$z
ztGD3W3HXVq_l8fOKuGDP!*a2Bh!|i{LT)JW07@us-4Th`*^2Yo>Pp)`&xK}tEJm|_
zKolM=O$kaAE#e53Aa-AT*PgMcXR%6-Sdx5O`*Zr~tl{snva(gb9R9ytztQAUDQlW1
zb?ceCEZ%DFNXfqZ(X1l7S7Wzfz)8k9`RRZf#OJ(b>Sfz&xgjbua1=uCX`eodR`11$
za`%3X!!T<vG??@7MU{<*c@PD|>b*Y;K>y0S7|2@u<+lwtzayXS%%FgQrc7@GGbfK=
zwA{~KoqA~9_Ej23K+ZgQvF-RRi*3j6h3ag}zt2@UPfpBDXf%&8!pWZ$(;`C>z3Rpr
zSC$6cYlHHB^yj1Vbici`bkOYT{LLEEs#IZ%6U{{u^y8=VV)LwY`uU69Y}{Dtzuauy
zr=puLdSXFZ^zL{e6y!CCiwU0zF^cSIb}VezFrR3#fzlM5#CRG`%XNxMxzCSwqOOJX
z=`X_DiaI}6RABpa3PPE3p03d|;@Qghf7buha?CId^~C}zB*yK11(9Ae^`ujiHfzWV
zJ82>X*6QCZ`=u~+n0@-(<wf2$_<r2HLhDQaVG+h1<Q#Giy?-zB#OB{*pf8u(I{Uz&
z{ZNCcs6u~4sfSoB&nUC599rz8CEnag1EMcJQQJv|2JP~zhAUFme>~v2yvz>cH6o{R
z3qd&WZc;ZlK=8KYE#gWeMEiUAygVC=F(<EAIn9p%^pbvQ=b`u%XEi%hI(O8R{^8+k
z#PJOoh25`5C?b8)#r8MGwVUV_&D{^NH3R-0clNktzua<HKCX(Q&9|tM@D|F?S-IbJ
zY7~D&tXB_HX7Zfnd#!76Nu`H)uF~{U3?ja0Y1{6x>X|JoX5=ayra?PV?!`yLL1VG|
z?)M+{4(}Koq3I2?NbU`KZKDP~=z!Bs#5>WJ%Vw+=AshI!?++o$%u0M~09aDz_oHo2
zSO6qz-G9`!u9CMzf5<|v%_;4})Y!-;Z0W8=F>Sv=@qQStq(Q;lJMBbvM&j?Mwvwju
zLGRl}a(XGinbiZ4vszW-Z>^3}wh-{NE=g50rKzyp(J)i^1<%var`zQ`CBF0XRdzX6
z^FO#Sr_#5_AFf-EF}oI}<jwD?eOS&?{OkYZob9vlSFYS`pYFAPtW)K9sXVG`S`%=}
zEa8Zv@V`Z@6LT;7*I=0;U!n02$#pM_NBegeIBT4!GGO$)VFe9;DydP=o9(dfI+JU6
zm1||c(|SSW^q+DXiJBP0eJLr&HMMhc<Ap;2Q8Mm0)6+4?^mVbwcj%8)@|aFO#p6;7
zhWkNt9Ad&%+oW?R<!IB|bY+%TrIk^UaK%_UnX<yfk20G+_J@$F^kS|C(>AD=P3M5#
z$AiywwA+Ci{AW0N6=yTQXvjg-gWogwq5P}4rh$RjJ;L2@;zzVjpYKn-Y%16qg!PxI
zQZNYh&{Ow)=j0I#CK6TOIyBaYan>}1*wQ}}TW{|PZ#_BE|Cv*mA@kH5N=D`ynqJWM
zQHVPn8nq13=kEN&ihB4=wF05*lwI&z*sFIfo98a6H3;2&91DzNuXvJKd6oqm4`wLt
zjgB74>FTrz8Vv__RXQw5zMZesXtzF0r^p-Iqv(o#6BThZ?JTX}El`NDn-_&RM}ogH
zY2$BSL<X78Ez^FXw@eBCb{|LXCtl>fh<683(O%zU8*Z^*VAoxln=%TQSim_OWf<x^
zFV~HtROez#r*w7i;~(=LCbL-}?38cd*FP}HVb(08RW{ePU?ktUT&-sB_0;^CO-w99
z7bA7JyWF(U&ok3>o)a*6E+1>72WTzL653m~IHAR~luSmR4sI7-Mc&~env(G>+O3Z&
zR!<UYEq+jr!qBx7FSO{<EA+_hvoAX6>`1$c0G%<;Ux^U7XNPdtqG}kZ&uf;hXHXI`
zH{5F#g`6$7_-6dppp!Z}cXH2UC4-GH^{)0A<|VQd7<QoQ)q^&j3r&E@F$4155m4VX
z-GgCp%QM-6wH<aMzLdqUTk0RRRKq2t*o8CJ*K{}Ml?T=c-}y}1fX!DgpZPCD?0kAn
zyRzFSR`^L%2D^sH)Z1i~bt8mWMyQ4+jveY;cF>+2pAHOt$7vu;dxWJb6?(geW9)nc
zqiJ#VD70&tXThA5LEJN!Jb%aTOA2|1Z-V}kz_#cOKKnwB;xqU|jV1N?t;5&oY<23q
z5B*&&r-zocojENnf2S2MvIWFJM0a`TW%<8-#&*ZwzWr25QCVT!HjzxQRxQh(QRl^s
z1p&6D%^P&7`iEzcU>8~9s+}}5KWgieYBdg0)hn-QMeUx89!6utAoCL*px+BG@Gsx2
zq%Q0hlpmLeR(*cz$aGHhz<G|E%t5HwQ8^{Pg1_2TJ>ts{YAB^mmb4p=@droxHQ;yf
zhCgMA1hoAo!FNw%#m&iYQ8<G4k3?ERPQ%?tFx|jwFtRxGC9*q{e<sS#yF(OJUY@UT
zpm>|yGgju``lWQk@sDfWCl>q*bzZcov7Zj>#pNikZ9Lr+`(<p|nI~mtc4cSLb$IIT
z51wwUfZcp^4IufdJX(!X+VxIQCHu93Sj;KlSflj*pQw^izF|g_-Lck(l)1^{J@_EC
zD~XBLKV*S(MMTkWmG(#L-sxWUSiYwX)K)vim|k(4yMZ%YPS36+6Ju8&3-dl5(S0kY
zBZ?7gnNHHF%VB=Zd>|CGru*~lyfS&fZ`%iYo?evPdp@(F7rC+Nvwhx?|LS!T`4idj
zhX|3qSfFOStE|{)Lt%sC%hFAcFY3!z@t2@o5B;ULAilquY7M2ozckh|-E8htBWDNf
zoBD@=Xp~$~%om#lcM~&OoNz)HCWbJ$1V1+WEc&IxRwCW4y(l9((xZJ0rufhs(B+^k
znW$&%rYho+HRm+U(Z~0|+p%{hb*Xqu>HjvbILT+I;-tR4C%fKhwnEQi{P9kHjg|AL
z|G0%qXZz!u6pz^AU4XmU6%z8!!GZ(SF6id^$;;eTjafk3#sFJ-02K1us?kfSaID)P
zwuNu}M24&MTZTz;DaDsrzOX-n$w%)cAO%ZfkG*xzlN=Q9Bn+zsUyByLrJoUTG)|%>
zjk#~tt#EO&dm%2AMYTNbOrE9y=)NpElHy%3eC`bq7r9Nh%v^DVvUM{?Jdq%g<}rBc
z*rvVGUzw`58IUq9JkI^yjxZrWN@zUOfZ!(St!C5Yn!`IFx+<u?Q7PrAY$_MZNlXY$
zSS7}|R_Yh$B4=Xz&(!r?M7~;uNs%m#t)#NW^1oEIj;d;xxQ~}DI4_zI@9NG(={Nc{
zEXB1I+>2qa@`u)%kg1_o(r5bPKsBi9FHzxA2qdn`=iiC1f0o0)4xAG2oD90_FZ2i{
zH~xN<cJ|A2;l+DrY1E3H*BYa4dWpKCxfXA<8rRIB*>B(7VCdYS?SmIo*n-QPjQZGh
zT=cA`)8DYbA^WZObVa#|XVOENX_|)gySV5@IPZ{k-!5Or>(wI{K9px?4rM_+=xr7m
zE7#T6(D1a$m|@}mj}L12-yR+7v+cnW#jSMUBPV}gweD%~32(7womFL&NL-Fx0iWj*
z5udUT#-ep%H~0}=c}Qc|$Tn=gEU8%%>r1Wb*=1auIQ52v_A#ORyM7*|GZB}n%}(q=
zra+`==f7vWfI&pxW6<ZI+pj2EVL`o*RQvYJR#n)BS&}V3?P)~>iVTGv&HVL3Vk^d<
zu6_+s@E`FzaxpFGq~LxZhkk4;iK~S+yR6alez8WI6i{~)PGw5mlRn=Wl(;ucqIe%S
z`<Jy~&+7Rq(~0%rVOYu^NaWVCCKlL-(JgwJFwxnR*`k$Ze#Pq9viCl1o9mGrYMosV
zRI%LEx5nHdp-V~~bh9vGFZknOi`wlWzYrSWl#Qu}XRyWsu1pRGJT=h&qH?u*ogxGC
zc>+M&q1-HA$_rCL<Okg-Tr&Ii{1Y}lrVHmqW2wuop8^sB@uZ7Ki)4IJ&54)WKM!JW
z2}_yCOD?z|JO>9pkLZZg3!^q=-3>#JhdT<+b#}#kb))mcd!g7eQKTs?_BeNBT&d8e
zqskWZ?loKGY47b$MDNKr`eqq@&^Lw#C&x#le?xRE*V7dnV_zyi@K^_0&wmfLz~n%i
zrVsA~Aoc`&MPfH*OkSQW|7Nf!n75(<lxuojsYypc*pUnhTo8004MJ_Ltpj><7xDhW
zg|ya-Sv+N3b|gw7bRopS4|J;2BPAfp@m3D^@t;ZHO~{XV?xOOp-Lg%>-EV(C8kcHE
z|ISR_H7GJwHYT_Wtu!>a#1pNJ5P#nOY?Wa!BvEe}hcIoc;pA%n(TE^o$VyXhMu39*
zFvtfZVz%3U+AyH2-+boUdQ@v%3&kZMUYL8~QdFH`>ByFF9j%jVbT{xm+4!XpGk7iN
z^<0qTU}z5QmgzM6WRD^HLt9b(0b{8WKeGWk3o2RjW1^yxIb4hBPhRgz;>Qh_r<JSr
zjrQl*$KAf2`>wwB)HVK09<h?Ko4gEt1Aiv_3~)63vHLa}#Cvn*@y5leOFoeQHS=ud
zcrn3tp+Ar$O-Cu^3)rVyalT7bBS<;noD>nelSPcV^K)D0;joYrDfMA-4zaj__mODe
z)*zj=SH*&lX;((2fD8EVdHv}_m^Jnx6~_}TvRO|b=LGX;6yjI(CERRlTlB5_a6&fa
zuCmH)>yYoRSNC<38OqCzyW@eD<K3i>iq*y(lh3*u?0)wn41Xrbs`pZlDtX!B#Kov5
z>xp^B&l6})JhO*_jk0;x({_plNBk8Sr`py>&8nV_@75T8*q`zrC3ZZ#K3T0kQ$kh!
zp&Vs#t!_7go@ZnK54k`_zf4Ujc7<uu7hjl<#(WmL%)Z%<16|oCg}PFD4-iUzZajzE
z4<Fv~q?1g~9^Di8SNi;^=U+5y6-UCK_c>2CmTiThKlESJnjFWEXiT_q6v`g=lW&FE
zksmOYu|jb?9|_&}7d59iUOTS9V*H4zqoBpPF_B+-`i_2znj6A--FIyJTfARcDWR6M
z#R&ZAlKFFO)lWV1S9@HXwZ%ln7@<MI3km)PzuhYenjdxTaK{+w*Yh-M)ls)*71O+a
ztr+tY?r?sO`}%t`XvF7Hw-T<Mb4p8d%Oz(eTz41PAT3z2)~w#J)nrK!^46HI+^s`$
zOw&dI?SE*WR+%3J__z(&A!$G%Dmk<L@uqph<1E<J{`mTGqZ2RSq9~Zh?Fw!L;ggfi
z2k9I$Yj5b!9xYm5&tGc$d$W8G9|=So8GGd<41-pz-6Y@d0dsmJw9jwDe-F68v}sZ|
zb_s2!U1IQTAW&u_QHF%8;eO{j03hQY;vMmRQgX~Sxwd}ucJs@B4m}2p&>4q-@`^Fy
z1j@P1VD>g<r8GKV{Eh^wT&0SsTCHkMQ-b!vtc2H@UfsJ{O{<zUs#^`4sIlpuXZkm@
zWSTq1{fr%5neGS!eM<ZGRu~vd6gpw7FPHZC2DyvGBzN?Li8h5R!yQA};(it+4H@Fo
z@us9*FQ&eZm`eQCm(_-F#mb*-y0{9wAMs@bo0Vh&8a7z~D6YRLCSVkUx?Ndy6ceVy
ziS4Z>bJIqR>?Yf*(w@>!Xwjk3R7z#SJf}!=-kD>&^axWHFJEpxn=r}xr46oAldC(o
zA6)$|Z{m?$q|eQ3FX)dO`}vm<<fvV<hRo6fO(6XBZrZZN6%}_3S{kk|Xn~*5F~0t_
z-18v4Yg=&St~P3Vn%wv0rwi`)?B30s-mychDEf1k7=Hv;xKV@9Vk@iu4-=cKy?W<-
zJN(0aQn{Q>2Wi>l|IeQvvG0-JZk{i#GzT<aV3Cr&Ez^{#P{IB_pgPlK6@zuwPHDd%
z5EENUYXN|eBittN+XmRIAR)e*m@N(PGC6*SY0F-1UdvUGHX!FJ15^5P<x}K)JjlyB
z1AkLCMS@Ums@<5XD60<(B-yqG;{ff_9nwO!)-Wn>kJ6Xb4tlqQ&}X08(bTS4(^RRP
zVOlh8V#lXzXZW~s%^LH_GcOqHE=%n%m#R&f%1Y4&hG*C-VH5qGs-U3_^?lxX;|;NP
zh0Egidc|>Cx#}kgf%cj)lP2kza|J3cy+0nf=EfQ#)?hpbLU{L$2BU+Nx$~vV6Qs#4
zBS;nWQTQIeUEQ}0xZc!_(185=Kc5zFw!>s@-!5&>44WP;#CYH}w?0<~^#)}JbqDQW
zPZ5k&{~26bj!>}k$ZZ?t@Xse#Pi~39UH43WT6gZ!(thSYtkEiHt_sRR=qesl-wEV4
z==O*Hxcc0;VXgc4rgB<aPGfbigg@nE{Xsow-*OFck`u#HmUR!MffQ?tVo2@PH<lsI
z?$SBDu>-~FtD`N(NX>@z1Hmy53+c68dfvHq&Z#Be<c%Z<%6kp}R>Bh8Ym~B@wmY}%
z(7aYwTJgm6q1T{bEEIJOCqMCBQZ!El<It6xyM6@q#Q$~U*^SWnyNSQh5@!viKIXCI
zD_7|(ejF#};!4n}=zGyu6WbP)QPOMp60PCzLTlF*U6V6tYi^(WPZV{}t?8tTx?f9c
zcq-Nf8Xs%dZ`ORhGd5O&T7ticd4oQJA97%HVj917(dS_SliyxHM~`mjn6o=|id|w2
z-z|aXKFvdp%4^I-DeGj5#{yi3;ImR>t&-yHrEz2P=+n<ixgpp`^BkF)TPm_<vo_}l
zm|Z=G<A`t(;V0iu=8o;M5-4q|$&w|w9tgk>N4uixtgG+#HVLky=>3mInD6G#kGhrT
zwZ@toZ@A`~g!ymwy!qz6k4NO0w<vRhUtqz;b^N@uNx%qhw6IL#4rMrHD_35!qpO=(
z5!vFIQU!pNmO%Z0cnQX88G)Dkwf4#wFpk{!O1ThozN>@Z;d9pM@YoL01iASeQ2evG
zs~laK93i*P_}5*#Yy$mZZ`QfPsit9r2Bvz|z@4?b=O2Gsv)b?<*Krf3Op1M<sziM0
z8M5fAVlu>6HE!5QJbA7H;D(JG%+eLh#lP-0qsM<1<H<Xn+|gWgey_MO#60rGdv7V$
z$fDS@xHiQ1Meuw)p?kLYbiq@)YaGLUa&_-M=7d&l{3O<@@R?Vhk8{uMmw#g0cIJwU
zd&lMB$CNLp+Gq3Mv6~~dZIzXx6mcqP{sM>!)FczzM+-|!@#WU*`-y*Lt<5F+?&+5Y
znGIXFB#e>VR0TatK7I;X!qJ@0f1f4w9C?+=Jz|WkSzWS<ync+4As-Jhdo;GBPHIXG
zdF#mo5C8c^`)n@8NKs$;krZR(p?*CLg7;*!O+h5@d-SaW32uL(vfAZQ7H+xzK4es~
zst)?~OU^#gT-Cj!t%v8j|M9miXmHRgE@lN?1vuRt0XVIbQ2n7N-?60wilhtpX;i0%
zIip=Ob84F=RuJ!3y-BVtp8s%?nYZL8|MFou$2HYsKffj>kb36531-@yMTey=`BnNq
zH(hL62&96R{ItOg{&bqGaW-3U#O3|hFZx*!%l&r8yZ-jN$rOVF0f-z|EyG-QerMCJ
zS)lR2JK-HV_dG7&vE#w%fgr}nR>c^pt|bIk2Nofo{A-z+2)A~`CpgT%YSyeSxAlP_
z*MA1r?l2p4W50fp+A{7ju<ZEDqmOBhXNw&4hktbGsOx5x)#CU>%^R0y!H++hSBB(T
z0*3q6rE}%`zuh@U_gwCrqszn5kxS?5>2}sx(t_z>Zxi_VuhKT(CoL0q-{s>-IcyMM
zFwq2aWpU4?57+6mvPc8*8_mdM)VM)IX*dl?*tD}@f9&a}%~lCc5nL1Ov!(=KC$wp8
zI!Ti!HKbhxh%#dIXfxy6Z`C;1pMdhKpH#FC!6KTPj@+3m$6e3wNWQWXF5Gv=9i~Q3
zoSq3&rkGDZ8>>x6klU5*+D0Apv(siLRnYHfXa=dn_en#3DCc0`5m<(+$F<$nM_L4F
z4hZ1)6&GJ*PR|jl0N@NA{*i(nE;l<gHcDw@kRlD)+itkQG;7j?<Z<-ktFM_KmoBl_
z2#TDMkzvj`>r8Fz0-$A&1T%ltce6u`C~a`<A%pwvx6KxmlP>By_bDTs*InbjBUhIz
z$4}$V-O=UY=wIU;N>-ms%dv;FcF*h9mE@7MSAFsHOD~z#0!UoBenQ?Q?}jz7N0%-t
ztDAp2?wt4Z<g?ESaM~Gn{ZRjK&!1Pl*Is$0Iq`%O5|aGs@BfexE7NZ5^YT}pR~>oX
zOQ^6I7hFB=7%qzIP<WgKR};sS8<XMtu6$RXD>ED`!{u{Fm)@1jZ~Q0|8v!9_?rfOz
z4Jtv{mB5j1F;c5`%~=1s=jO=T#C&&S-@c|*^XC4oxO3L<KWi@9%UY-BN09@6<BUD}
z%kR6wuHkVN9fBWS&pO*`vQiam_`_Pmd9H&&e%~Z3zfm4*_~TDMqcwb+U2`hrtl@37
zhNF!ewT6G9HT>(GHJthiQon0gIBod1OBZzvrwy;+V<%2BV<wD`8-s4lM3v{Rh0Ar<
z!^fy}T!UOa{%=41Wfxv(PK&MK+sua}J~qpxtdu1%!{xnoz<@m0aMG1YvWEZV(MN5t
zVcQt}@#BxUYe&kFxe`mcC|-^>ChJ3Yow>nd1M__>_a59azYG^Ej>4IpJ4Yt_Vd0Nv
z&>Qi41NZ*?41QglBS63Jj@#n|;G`$U$lKDm&x_zI;W4SlmFpkL>+ZR8j^Q+1gIrp7
z46f#0=Qnkt2!PwVFp<wESC4;<bKW!S8?kk)Riy-xHWrY|Eh9IQX%)7Ivm8~wQ{@Wm
znm)Ihrj1)Bv^#UhHt~PI7q}6~O%f7#7f$Wa#awv)Rrbm*%?Hm9dc<tplqeKNsi4y-
zUCc@CPPOhQw`|>LzM3)F%$+~eEA8N&R*>RHmos{r<D0e?&r(i&@=7y#>S%#K%ac}G
zuWn;=>wpL1z#I}!{c@}sJ?_JVeEasvMPRlRAq1THd#z++i~{;zd6j8*LOfvOVvKAN
zFj7*KR~!5~3YyN(Q`qxEQAe^i9qhU(|4_vk;Thv%jI=r>S9>8@2e|9_TnqbT+pzkl
z!y046wT<J&XSFw1_XwV6yR_RYNy{-zkGww0%wDuId^zqn))-Gc*w<PuV8u|$n()RY
zG4GE3+I%!+R{X_7^b6ns0e91SHEoSY6EK1>{gHcv$=~lT9~kkkbKbUTJ#*KU-7Hw*
z-y`nG-v)nb7OqHKJp6PA`@DO{R_4Zw&Wu%aBm*P=H}EoRX#}(L+ZTqJrHV@srh)(?
zE>G1;70tu<Tq!pP@(C*|nfo7q>liTNI_DrCqhgHUqY@w9m|Rx$To3Kk+`VeaGI7Z)
zH%AU(HBwVPcp=sow2%=PJ3qH#sfFd|YQT5saI(4JoWK>q4*8VEO+c95c>hDQSg}T;
z$oYTWFCp@;@4C|p`Lm?ui&kBjK9i9Ai?3$b%g!9zPqSLC<EqN%abA8qsSiPFGT)j!
zGn*Zr*Z0%8M}nY%GMTQQqrsiIBg_2dx#twKLw-qj>K#`Oo*SR6l#hn-w)I<0dW}lj
zXiBF`yB1u~pc&0(#1U!j<_ZZ_4rOjtKUS9CUtNO<B@gG0(5%K+cZT}<9tqwnSBm#9
z3()b{D}yXBLfy{#3^xeJ2_Wfr=_QK65N}3}i;DM03LKd`C&$%95Wj4<Gz#$>4q;-P
znYG6TY%DDybXEB|#-)NA8~7pR`FG0FfV8-@pw7VLUw%r}PbsoAAnpTNNc2aF^xYu(
zn!adKzrML${p_B1>R0l4J8^<dLm%Sy!@ADNDVw^T?^(>dX^&&v={GdQH*eZu$Av9K
zU5~GP2x~`F$4jfNdt}*APA&y1$|7Oo#!OSKdSJ1#UC&EE%6)pKN>wXLixqbw()?5`
zi|a4H#GKyYBwN;N?+v$ObH6rR1TQV4coBO_rI?2X4lqp`HneH{KQFxfj#;~EwSP&V
zP+5V4*V=c3Ch8L}ykyqM%s<TI#bqkM5BH6cl)geEA1!?RPe)Ps8!xOcT)xdso8+zm
zu)cEc*raavEUrz=k4>eudB=6v*`_4>_9uVwrI|SO3p-9o=*Apv+aoQs-5Oi91g7=B
z<|_M*CM##*;w9#7xyUFEIiLQ18DR}&|6WngFTb#tGyxMg_rw@suHCk6vj8<`;aAjJ
z3YU*qNd7ujxZK4{%~-+Wh;f_On52yrB&cMT-MM|MNvl*P--?-kn6Kc&_HWx}%A^M|
zr1q&Cgiy@B%qP&g#Y^qc{W;Qs5qvoN?(v8r#60Ta{`5NY4AWhpER^iKd1dHOS+}mT
ze!cg}sw_Q4{?$Wc4{MB{-+pVJYxwvnpPSD$r}GV|yKPJe*N+GLh}QoNnjf4NE8oyl
zFTAY22-dfQU4K;h#1gvwhU*i2#3Mj@?$y`jIwvzB+i|kSxctHkV#)(90b0r<q(m@l
z&OCRiKu6vMJOXR3@Asd~{+#U>=3n=H-!rebY|zMDe|4YO8a`_LWHWC17nap!?RIN6
z>$P*|K^q*woB6z#yIz<h`QaC$);K?g*0}=YPdoWUTl8@8KcCP3&c^cc>$?8RD<a)3
zz5ecd(jbqQUD!4Jx@!y;Gj0vXIuv65+AS;3DwQjl_2TjBRIlt;b5PFx_Y_sZIw@vV
zkWYQ$mu-?peD<DFiZ4;wew(Ccu6Jc{WFFKMX?d`0@k|xOGbQ7tqv|W_HNM@S4ZPXz
zFS4_DnewvU_9@)g{=KMk#s`)%b*ff3w+-kgzBHZ}axq4t^15rTU)=AKIQrv0-eNf@
z-R^O!YLAn%mvnrTUpL=U1p!>!#25MT7G0Y4vMeiCwX)eQcbwU>B2KF)H&rR+O=-=8
zdv|TuypsDZx;#hQv0j}<=GGe?Oc4BI74_J^eq+DQ{GO@PDQBCDdR}dl&7L>Iyf<u6
zep!y1`*mwKFavM=MMD2114j1jFDE5V?MY<;UtKP*3qn_1e4%O6B3|JFixFastl6?D
zGJZ>b<@(>8|E=Vgub}t&b_z3}<;92kbr!VT!tmFHN$a=8Kll67zwV#2AKW3HZ--)a
zR?bK*af=Z@#>kXeUz?ekYs@L>wc~-2mZx;o;=}S&G&+)ik#jq>HUlm^!#>FP8Q+_3
z9a<$k)Z}j$niqymO1SP{IQyg)X5eLK+e^f>h>DN#>WHak?AP=CY=`U|D~j7MKPNWH
z98JJT<+KXssfYR|^i^KK2#H%Zs%`$e{{;z_Be>K0odo=c%dHk5+8=J~ZE94>oxG1^
zVB}ACTy7d1SIzd<FQ0nX2DJ~rjKTmT<o(^iOHI?dHKlF5%iQ~qHxKK^FZ?Y2Z!|5+
zNl1sQmE8i>j&{qIC}6}rZ<xtlEA8j)62Kg3#5-w<)ke+|l;2v^;iUHFg6`eyW6b<+
zw%lR0OL(MNbLW|lMvaNPjwY^O_KDIgxuoZLHqEzl=9-h*CTdE5I(EF7A)ngeym18=
z;YYIa+&e?S)~<C+bD4bpJC`E`5vusYs3J5g-z;b}<8G&F<@gPDUP2WLtyZ;);*|7B
z5XQi?zWKp0vqEl$4hP|K>YB?16nj3O0Y=72m{uHe?=i13NW*w64Ci&3f-bge-AZnE
z_L}Dh4>6lIZ?OWM6oEbn>(eV%GKCSU(7ub#>tRlA*VbCmBDC8s%YeeQD)%WFx8dhW
zHr8`<75t3xhrj|2ax7f(qo}&MMljS{uMc>T1&bD2!DV_*U}sk*?yp`L{DwfgKsfEL
zbKEOm_Io9CYEWOBz5e~IhIO(zWcK{;&9F~L`Pq-y^WFkR+U5a_?AX4gKtlh6@qkfd
z>tNJV+%gN>M?BZIt(&F!8Ssy2zlZ6^1q<RX9Sy)pT!XzI%zLMvbfUpMCY0YAPI@Q`
za5|i8xNTHEe@@UE-n)=%cs@lI`L>%2kJdHZUsL_LAZiUCr}$8x>wi(#u_AYJD5L9h
zzrFq%&ib-(^X5=4Q|L9kXV<f(L4HbX4aefj)x9GtQ?ABSr8IyS`|#?VP<4SW`0t`X
z8N5WqeMylJoi%)u0wV8CtzfE#EQ-D4hwZdQv!)7|eu=Gr;?!y8tj?!e-otOu$Z->7
zF;zI_Sa|260wcxgcfZ|*IS*VIC0dIVri|@ba=|IyG;3$3*}NrN79vtslCT?X@Xgy$
z^wJouQq9H#LRrY|d7LE?N>osqz@$vMA=Q@ADt|eJFY7{2=vlU0rdE{O(`+dfQ3gtO
zj$DiimXmCaJzTUxX&Q^T(L=$BG!QOY)3pwxqy-V5gnn=otVmd{ODRfqp!k%bZ|p|J
zv8#JrpycP$()MSD=sRD(-d@@vz7<6+2-i@=%iz7@#_#G&E;g;S7e-m)tzp9zD|1m?
z%nDa8dHI(4#@)5N<~nK!@(}w?ZC38U)nDMgm){UBg>Whg7tj45R$LHoiTkt@3S3&*
zQ~k;9W0!|JZme>oZFSTyxAeU>_HJG7sS3V6YVzcqdV{MjmdLiu?Iul@i%~q0%kLla
z0!FqTfwB+UPLJmHp#L<-&glX8>D49dN-GSExSlSu#fU4wKOU*T$SWUzZhD>GHbF~*
zcp&%w<BdGsPPqWp>o#pOBfp$$23&Bu|H$^dD8LBXw76g3KbkB~IBPXY6KY7jzz5Vq
z0Kq+1b~m_ha#U1c1UDT7y`nAUBy?Z`L*TGS6FokU(WotHDk?BSI{>`^zrs|Q*fI2B
zSutm553@M(+-Pv^(PGN6Fv4;`5#rEvY~47nNksuhSnP2rM7v=2af`w`U~KT;F$ZxK
z72(I<WZh9)nnR1_M(g)44Uf|%cf}PJ7`dxYH*;FsrY2i~)b9SfeD<lm$%wJfcffyQ
zvY>ghTj_b$!t}jgKJOHN`Q=o0dHF*nX38!h&Xfz20b!|hrCiUoZCY1aN9D}gO<T>7
zQC}Kd;zf~VtTaBb&O#W6PkG)!VaG^R*(G%?42+=B0Henfhc`is7A}nSMM>3PQsvaB
zDc}E93#N(OQNV~3s(kX<IJp8kS?1~Sd9$Vc__(y3^U}<`Rs!+kn=}@qzSVp=^Ba}f
zHQ{j_Fp_X1j|<GHe8*+=9ys44=ASX1S6jlxHLPFHTyfC_#uwHW1Q-dI!*ANa_(eMt
zS0p<m&@HcPxB^1J?!J$xyy0s_6{=7U0M>=)b`KY6maSZ6h75f_QUINhSQmH}=8n?L
zk?LnQM?44pOFtKs(6sDwdS@H6!)?UKTU20#Ih(%DZC9{{7PeiRHFlE;3i@1fk!juh
zc+IoBtmZNCMSfB|tBfiYOp5wpkLJKT<i9#9xM!5MDbIU`VsD_m4FK@ei?77_2<E;|
zb{sLD^q;W+e?Z?O<VZHY3!-I3JR05yb2#m=aZ^;L10}HLI{UQFrd5mM1^!f28+S+x
ze7XfK{q_~)oc0`Fr)~@w`A)vQv1aMewryN?=K5FPdROiN*2G=s2fzrM`tDuNG2Nv7
z&Nn`M)ELu88f#%H3^3xfsqEUL{(vvTkNg3EFZBn0=)-&f;{~7}%NUm*ORCCpao~Qt
zHZeZYRAqj0-#_oqkps!xZ+CPSBLx-u+x1=TCk6*<)O)QGLgBtoSNG9&ENCD3&$V5a
zI29`X`v4=&a<n#+32O?ohVPNvDaxT=O9{Nh%Enn4{HLzr@VmI<a^LX3;~HKNi@0QK
z_;>~M{e0?|vH34N5BlGpf9}@s9kM{ICV?F5Pp~g3-rVI|8~bbc-=2G3F0^*pCOC^k
zeq1FB+m^>S^_ynTlda*M+P5`5&+QSphHsNlx}seDIN@SZ*6@UeL|iE78s1uKI9A@Q
z;h)JeckY5k`nsdZ@_IRZ6=`@Mh2^EAOV0o5mEyHpH4lW;|9W|lxwPl`2EN><l|TJt
z28l-~3>5|#Dd@N@O#MYo8`eON?JIJ46w0-{V9s90pOdy^ZZlQVGZd$6yD6*hYOe&|
z<rOaq9%|zj#hj~9-t3Y_KU7A1%zp7#rNzInkp<9mWx05F78xONY~bhw0SbG4jgbT5
zY4&4@ud>o=%2Gv^Mx}A(Mfu`^h|gpB9$9W@6AM;(N-OT&e#tK&VkJV<S6XtI@)gRN
z9b2+&OgXOamJAU}k*s02Jk!_>w|-yHy^Fcz{KP%Z$O)4b6kYCl#di^}uN3kpo$pmS
z?i#Lbe|!I3iYr;(Y*AdGKRoe_*(ompWmPwX1v$@KzGAAiV&xipx*dwew^!E7q@%pD
zdJcZ4$=ZC!W_bdTwGVAiS0I;N0*=b)nfEJR9?wKyzzgn{tfZ{2BagH$D<y;4uzs6K
zQC$F44uB%Y9<la#AKBZpRaX!jkhVa3vi07SCM`WxW#s}%SjRVQT;JS((|`nE<P*i4
z9651H>>0`mgu)xbo^0OP9ve5Rw5*k)ER}*<bN`kUFoIl={X6Rm#8|OoJ7-+v)_ClU
z9w-S*lBk{p#LJdz&tt$yRPPkxnr|@zFj6(6qUqh`<b)C)f9o?d^M_>#*SV0UiN(6n
z(u6VM^Y6^Ytve);FFG)yxy;m2V0f%o+BB|XYDqv#&}s<AcBA4lELyqFOqltDS+IO<
z!p4_$xGAbxwUUjuao(wItT3PJj@UFK;PruwHFEQUyRR@Y8Df2LQu79;RpZ*Gx?E2n
zC`JGbG3gOxd-aAb>Z9dm%B+R4xDw&Ait@WSzzA*q$9-4Z1;MYI=O^g*#;u8F;4U2j
z-T*L8kYKz~?dtM_UoI}VJ8ro)ZpkzYR;)Ex2`yT+KDHpcGC0BzR1qlEN_}`~my=B0
z8i{Tb7Ar6B)e(LA_Gi;=aB_Fe9cxypWID8NWKK{(bSyd2WU^~PDh7@yTSEQ~YS$0T
z)|g4(E|BZ14R-wx?iT*<A8zk$gWb(pw8}j5{)ET^3JZ+%KIde^xMUOYkgQO`zzFZW
zLn}R_Vv#ga+!*4=qzV*q0!aXjO?r1f>N}V+>qqnb($#TeIkJ(u$VTSmmJQ9>CpI??
zYgaQ><kkyj0pZaWSrP$mz-IseKmbWZK~(*;VXK*}KEs;mZh<%4_bDn^Jl{g+6Pnoe
z)~i|7=Hc4_Y=CiKj4fKZ-j@4=zE{Vz`Q;v}^Dr=ytX?$wS1gO3?~>&?l-J9M0cp}S
zaU&Sj24AT1!RRrjh8RkOreS(#*hdC!@F>Ea#Qk>+v_kf2(`Q&0P{jpCh*5D$yLP5=
zgZidwmC7+wdB);4F}kZHVEtzH9BUon1ipUT2)b8QfuV6_bLz<_#h5ae2Ep^9B}?PR
z8`_5Br%ZK8Vn^mEHgV0HG_l9Z8I|<Ci4jInDPqj-kh_JAn>X1$`fl$0SkUcQArU9V
z0Y<2Q;DCOnir$w`^aZYl7A0|gMp`xtXz(>{)X)ZpMo^9r-OV3tmbY!sG^^LHH3*_9
zE2@u4PhMP`oYwIa)2M!8JTm%p{f2n=f?s{}tqq#%x0`dyYI%G!(}Z>@P-g{g@Ca(|
z+66OByLk5m*TfBm->!lPfBZUo%VMNy<3<*|K-&~z!&(;2Ha1RL({8!y>Ouh{04)~?
zuwqQS`1+d(zzF@^Lf<C!)~i$73N;w3%(D>Vd%K=-tu*3h%$sMH$z7ihP$dIK=%Yp&
zqgbmD@S8EpxDCc;mf5V>9CN>4V3x01WiFFY;<#FY+fZAn{t&SCNxA;Q%|&4dVMB|$
z_2Ph$x!?a_uDhalLNUw@V<%5ZxDGI)DDmdD8*j8@35%yu<0r(*^fMF&7-2lMYtza!
zYt%@tQEF;@$=!td1VUj2t+$mwtu}x;92rYkiJ`&g3!%5`8J+N$ya=u6=Qb@{Sa&#l
zi*{_vZO+0)OU!g>y;3#+89-T>3IdGqK6ytc%kr+FzAJ*<Q!m7L*pk!sZ@!x&FmFw4
zvmLJMNcvqA3%sHLBbgHVm(#l!tg3lG-eQFBXN}&GPe;1N2*Ah&37l%C$Mak68T>UI
z<%UXfBbICp|90UI);&?UUyrTf$JX#<Yxw$2o8+D{5VX26cd*uQSAJecr=?Ars5LyC
zx3E^9P3qS*mrL6`Y7HlNZAAsWZr!3~f!1(bI>9&lYk0VQ?ssw5a1<b(eDP(o%Ui?Y
zQCl=^V%MtUv<7C#ZEH+GsdZ$#)~vPb)|)xv^KercRuph;E6x#q^S;|}wfjrf*+Fl;
zV@~bV!JL2gS#ibE-%r2ts#UHksEeC|0wbq)>|pt=xC)c}m@#X%S)y2TKEdNUm#W{+
zovYy3S^jNvzLYxX+hnw5>o&{#@QjXlH&|CUYtq;>tXI$GhZjVq1d$eNr49Q2zF)Y=
zz9Zf-fJj9FAz2E#zCR_^bZXVYv~Ar+>4`OBv&yAb05+78P`;GF7Rq9L4Ie+ztlzj<
zGK&mz*?HY!{19$aaXDLxy}7`0X!xkHW^E4dzF%XCHML`h_U0t<RCP3#s))yQV~ODJ
zTefevidoZV&o%QG{%8S{J(7JnU?gan+^E*bu;79k6<kug$_f;r&E+Liu!=zf7ivjC
z?|*(CxMy~GIj-HX!94u@%a&imwQp9Y+9qBdAc1GCkXFvLZ`s;hf7!)0+ZT%4^v1AH
zEXd&Gat*2<XS#PeU3~%uB#RC`GdwsJ8L<EGf^!xwFw>^bG8uA<Y~2~#(j*U&oUwd4
z6t~Kn_S#>zZrV&^>o}7ks~u!05d0-P+Va(_&5Sv7%=e3yS%sSvDcwM)o_Mlps=Y0B
z(+l)H;0|C7-9o!Q`fP$(vwnm2K`Ht!gTBI|CR-p!lZN%oeK+*C`E7wYe}<18XC_Vm
zT8gVR%-I_At&VS^5D01ZJu`OJsl0iznEh7n8sS;<qT&D}JQv>#o+B9B>_^;KVed_>
zK=}Nb;$PRSTW7~-UTrycuVi`vzV8??Qp~;#10w(pf4=JqzTG(bV$Ndo^n2suuDKs)
zop`*tt@k<h=kHz|Ze0Ww6&PVsyGk)nXcyY8c@fjdHv(&3{%DE?Cj7Lw_3mQMlJK`M
zgpm37FMSl31u)>2OV2hPC8TB|@X0Mi{x#0m%4&wbB2afpP;r0}>i^Bnmzd@a;>{Jg
z{??eU%qLT4bDBV>O3;0iVt`=b5hlPSC**PWV74ghy>T-V1oZW5RW*OU^9q-=Ku78z
zp2spTo(H{u)=ACH^}SBBz*M-PsP_TjFnsd2=A+586518cpX_(+y{k>Z(v_I9u)qi<
zxN>Q?PwkJdZ*J9?3xg7*iRyb6d_ck>qrON0Sp0G`1h(9Mc^3tKj~8x|A7;ZJ-CX0H
z&RzVI$xz%O7n_IpLH9k1>#sn&p|ya^n`Hrnb%`VToOWW3;+~Ovl=tDpZ;JIUTpdL{
z78DrSD8_8FnAIbh>MJfnd2LAYqp(l~W(A+wVVe8H0`tL$Ps5k(@Ae8PdTFoo?fJ`Z
zylokxq5>mq7P_5vhUs`>`xqk?S!vEiHk}`j`AnMxX)1X{duUf-Xr-9gL2tejmj!{*
z#pj=E1@``UB~jFz!gZ`>-XHOafQ5DZOiIN8Mko}Ekto+^c?pl}*QsMJyRc`1`!33h
zdLFLLU$9VpJXV@e!DfrJ5CY+Q2i`2Jvcz>DND=*a*hs}CoA-U(&BhHHXbhdF&0(du
z>-<R8E(F>m$Bws~kc3PHyx`s`bwHaHV<b)c(gOlUiefQB8R0Uy2Gvwtm@6*6z#fAu
zxqqXck-LCWAAT}2PoNEf`FYxypWN<*Sl#4x&*a+UU*kJRSkp)wa|?d!{7Jd~!jh`6
zR9gak=SqurGD<Ena1qC5$R}KtJ@?w{2|^-(k!lkB-#t+A4I~s8Bdxb(D_6$K^vlR=
zY`b(kFK&Ul%K9W-Lah9V_3XzbOU$R*RHAtsCSTK(XS(Xri%o604@*V}Gw^qhAgV`~
zc;OFu@&iTy1JCQ)#m-+2d`(stYw27GwMI>tSZGiyS(!)D^{@cHAZ`11vlziuL0%NK
zhTo<&+~?WI@2}zEem%B^$M^8DH9WpQ!oQB>H9Xwb!U7|F8|gXmcibB4%B(DK^4cpd
zH#KY2aF=4o!@Y(h=zQ|USHuekYk2i)Rjp+P3K))9^Vw@TYaVx<BWpLnYJLJ!KTko<
zJBXLRphq{Gbd&`1)1?^0+BmR(--PTFKA&c$eEC((Iw>JR=t4n(5ofUy${zW1gm_()
zu6)7`7|KM@XgveKpik_X;O~+V!rzyMycOfgu=2qT9AHaGbQH<&Qtmo|6z_dJ!u%xV
zmQ)4o-l}*T2V@D-TfE;{5z1^)ZF%^~XU)=OE2L~u*ZkLgcgBE`sI-qh_lo&3M~R?b
zt?K5cesXQ)jU^N}_er4zt65*>!*?*@i|LAaIL2CmWJ?)3M&j?vrKJiaP%Oi3S*D!g
z4VM+Tv2*(_D`(h`>;Pm!1TT31Z8u4Qt3fO~Cye1oo_Wcv(Eh5Lth+aE2o!iL3c%Qf
zm4))%dd-!lUCZWH*6^>_-Z1m$FEkZp<pE{OD&#bIO1S=tOYQGue?A^F-i(?u)ew`b
zvSMFuR~s`FqhYtKJQ~%jYp%ZF0>ymHedl~P@MG}fVV--;<f&$)l!8(PsIb$x=AsMD
z$!*#stJ+@w=ioQYym<?4-B|)dDk+u>k{Dc<9ap`EdGMAS<LV*PV8u!-qdNNZJkRdW
zY(Wu!ep<8EygU4(NPgEPFDfv?ez0e^bF5OLQ$%uQhK~SHfHY$#O}3Wp;rqvamuwFl
z14c@qKf}NXaYkHB1;2Q-@NuVwyPYuIf8BZJ<Q5IBmi~iJzGHfHByLIJEk>9m@9NXl
zbZXNiu8Q?rw#y=6p>^eRrofI9B{=qJ#oA5gPp^!SiEXY>hx$9U4$48e?@(lykBJ>C
z51#-s9({APf5}{XUMF+m8SUe)k$0jjB&N<@WQE3kx^+l!KL$ARe*%dz6+@!9R9s->
znsYmui_VOH2HG@R77&lTJ~E+sSeZP0cb^0+8>}$i89Uu<+O|V(pqfRx=mMldNbeIC
zBEP!fLVF3Y3$Q6nxNGyRCRluE+sbvh{$%UPis<)>O%evoR?Es^==d4(^}f@BVd3~T
zl=a;E6V0?ai^A2G*zckMBT<zDLj38DE8^U~k#0+7mbGSLT%DwE?LxV{3ol4R$0-Q1
z0H7#h{6E;Yr?viZ_Zi1yzM5zIrCLV19qU-*INGZI{M`$~6(By#{PxyMZA_RCCVXQ?
z%T*o(Sn^+rVN*exAP{rJuu0#_YHFFa6scXkin+Q+M;o}_CeR-k_P@OPiM8}{*AC^e
zpuot|#fwXP^H9>ynIYf1wWM{QABBY~_$a@ve?J?1($9-VIRZW0+=R((8qkvd=SzbO
z8=+p^yM~iV@EF7xNw|@7fiZjeMHg82Mt&N!t!7C}3&CnTfgQN)@@cKKU|xG?sKJuQ
z5%pubFEO>XJaAfA?mpuiFvW*60W!jYoe^xJZRQhsiGeU@{tu>%=Bu94d<(lD+q`9q
z0VE=jdNL|5FhcwS;uwSp>wxD54Nh>!1dwsF{L->v_9<I!oFKo?Yu9ZsjT+Rmn=GgK
z<kGR3WmD@D+HJS?yUtphl#x$gEIh)*xCsL)lmpn}2#vy9wMniV0!4T~<3FEjHb`5w
zrZ&<&yLE|a)49A;zMO6*C{0NbZj&mMI$+yeTJ5^+>cY9wO4bep+<xF-myYjgqWrg`
zNr{yVpkO$VC<4t7KlxM}W|u4X$Oi0u3DG;rD$1wt7yf7k5NK(gb=s+>LEXBs+ic+8
zc>hDIvBl;b3zpgQ1u~7B5Zk!N66bRg<w^@s;DR3AH4n6l%grX}^>^R5RxADufRQuM
zAnn%8f?O|ZL&&$`z!pDke!vL*flIeDPwf=92I4M_xs872JHV185K5Ga6~A%QX2t0F
zP}*$Kv7;F8H)z9+P{$|c1Lmep`sT6VsxE6cz9|Gvr*$~RPioKe0!9Eu5aM7_<P+Zj
z<)PtwQoFX+I>^5dko5Kk!>#3yfBi^04_n?8q-|jVenIjO)V|`skt;197-0&?9kD@g
zM6Kbs_Pf^94z2(G8XoS~|Kv41O#x}oIrH=qU&H_Y@+;;*&K@tBHRDmShKqOKt##3-
zqO9S}X;{#Od9jjS!~J%0UKC&?tx}b6G1fwOfMUDVt6L}T-jT9~&zSYC89QYf8WGlN
z?8a-amU2RdZ7XXp@jHkYn6)$8S|^;<xsz$Gwcn>tJ{xPbxczIzJ-<o8>;Y%kYumVM
zTEn1wZoMhN;+b{s*;fY_xXOY8BWSHRZdl)RR6HguNyCJ&9sb}a?EzK@c<_0EO6lq5
z4gnvjISN3mwL`?m@H~##lfV~XNs4u+BP>h4{N@|8M&&M9z9Pn7U2^_;=ESz~f^ql=
z0GA(R9g-mhjPBwE!-_%xEHAwJnk+j~EHEP9Or~AqhNg4Jj;2}T#xaWpS3gI}ohd7r
zd9oP(cFug0Dc4)+8C6t&Yw?A(#e1C+dp6Q<l493nS@%wx{*BqJ`fApwZhm$DJ&M~^
zk=quS_)GxSS6_c?c1o^RT^9Y9_Uxgt)Y#s4f1Z2&Em>VGGJ7Q}C?f#Gy4KPj&snfI
zO;X%zm6i+wmsI;T*Whw%e~vXncg@RJUfk0a89R=e{J9x1HrS8s&2iZU!jvMHRu$AY
zzkBc=tFX0EJlP{JzGO1D?Nr%vCne7s`=#_$>$n=GnUsQBH*acAkrGW<UP479#!WOM
zCw*?MIm!#n0dy%Z>y3I+1iGW|b=LBpbXbRs7Vt4k`@sXH_9=FYJnLN0!!*{IbHqD%
z>D~9t0x96I=WL>THL6xO9owHE%aM5C$K`qA)z{5(xtGIABvbCY(gf`6*t%0-QaRJ4
zL0$8+0f`nPqb5v|JKE0nc~-4nBMbW-rcUizsw;6m0~~qn?ROLQiLU;l03(1R@DgD+
zdF;{IW3Z=VUr<j9dOrQQWT_eY(TF2nuHv`(Xgog#jFiAzivmV2Ir~I&m0}Bo>0d*~
zn=j`sj=PK-9Rh=+9X|4lIpz(;AUL;En<4=t7o6VCTzCGdapiB85dO~+q#_W8aGCV$
z8!w6zwsJjLfOt^u|Gnjsn8rJ4p@PI1VRQ1o5(qbw0NAJbOIMq}4IXKQyj*W1mre+X
zRaG1f{&boZJjY!u!jIwtBWJch-UglzR}IK@`(Iy+Tl`6Ls(`6`uIe6_9QNQ(uY4R6
zXd|4UU&DO^c=F(rZ^O4mQrA;jS_^~l&4R=j0r0?*!Y3>ju=2rO7KH!D-|3w-l#4B&
z09bzg%=>Pg_HUKQc}W2ytz?1ms{xFq_`&qV+n<>k-!HRm<<|Yzo@W6PzwwOeUpzI`
zGBW;k&WUGo=M`P!%AK`vg?a4FF|qfL^~W#!_ln!p;1=$MVUw);INV}6FcBu>zm%3d
zj^3Xz(^`bM-yEw*`wCZNete(T^j&-=K<02z6flBmtynqR9DEz6$hYd^AAdX?%_~%$
z#AayPtXaN!qJ%046%o*e>HQI(nt2k^`jo0U{{GtsTI&vCi=gez=A$UU2tvej&x!wr
z5d;Eypuq{DjWa;P$vPpSPgH<-((|p{c-_^ep4<!hRL~eBH}%z~E|hoK3T=$u`@ou8
zua=9&(@$z^`dxW>-1kFWXwKtSCm9tN7{S#Jz(jaHU=#fG%dc8fZolpt)v6gZiGA%Z
z?uP*@EMzc&_Gy<kK+g|;T`q#Q`<G+q^bvg(rh>#6!S&NkdPm{s_)tN%aq-|Nic#OG
zP5v|5D26SDT%N)naix{21!|w>Gh7Gszp_uETxo^tbMAx63p9E_-${4_Hc6n!q-kHq
zvhm&V{rSYE9c{yPvQ%^Bx%+I!>31K<LP-DO>u<z_<XAZjyuP0-$Z{=vaFh0mtfK(4
z5LDr!tRUpvm?$cUX2nucc*!y+wrFh2CqKgb4@Qm@z}CRrc-55&MMeQ^`=xb=@T7q@
zi)`iwz4^9vBUKb&1XpZ~6JPiS5cKZwkK^3S0aRV0ao)CdtGG7ep6#s<hDC1Ba8Gp3
z8K=di=LZ2VCr<ss3JmzR+H0eYhHNrxk7U3I<DVESVfS0O(s)DjjeAE3c9;YCo_+e_
z>lx;=Nr|q?{4|I2JehDfuQr7R_<5yu_sEl4A<g~_yVBy~(H4BYvEm32>KV$~zsI`5
zvOjYEM6KbSBz)*8a57A84aePBdXDh-XjsFUA8^g#Q%SAil!?V+Nw4AmVqKErP+oLw
z4L7DF)^OIN6gfeRqLKnen0Ijr<*zN&SJ*W?+{Thx!+-i|r2)`kw?f=N<^@Odq=<sW
zHv(EmtWSU^1afwSr34leB}FLV+;Kx7z$O0OVDa*fQ0O@Il#|`h*b&}k)w;DdSas}T
zl9Pf0BQ85uA1EG#`7i{tZ+|!}QZ|7g7ex<8@DtAt8l1pib02r6{jR*imX9SWWrp3_
zvd-Ui(+$=|uq%@z$}I1F{7LK`wQ14Z@-luJ&fk&p#saxm&6G>6G${aV+qy-dMY<I@
zUwr=g34AoJZ2#}C|FFtFrKG@Ax~!BHWtCA%)+Pt`?os^3Ugo@R!M-bN=T7qvS&FO`
zK*d<Wh1S6S*P2sKY#VFFq%XfVZ-1n*B)-(b^xAKr(D56A5!F2b{3wA6!$*%Z2LyJM
zkpeRNk^}PfRaMp(+Z0oyoMLK}D_h!Z-6-BueL@V5JpjoXqNOFv+b<rxM)j)ZH$T72
z?)zd5(8*77|Mr(BpEDI@!LePjKgeHJpa+Bh^bRMP>%;>>6Tg~nhJG?SaF3<(Dr5wp
z*sh&&nJdeRG_~PEeJf|4lBmG)@n_@BXaPX<L#onc?@-JUDSF;<-BqTQ#*w3O;@?M3
zo@B?4nqiM=;;Vmt<Bitcm`g|ej|pD}ygs1XKJ^7LT<_}F*EH}faa{VR1`jdI1ky3z
z>=u|q+!A3VlF2B>YZ8kQ^70O!AN;mixNNzsN6L!FPc@fx?P|_CHD1AwwCvZ}w@1yN
zQTr{F+@5=Fh;<c9eq<tBiI1Xss;Qn{-MZTSKldE1(Yr$7$hn`x)#rd~u1fG)01rL+
zi`>1C%m1Ht1UQTL(Nc0fa10nJfxe6aMu_S0hyqTBg%}8C5zzW{M!RO_4hdYLKfL_0
zS*UmuMFB>Lq44N^*VswK7MAnp&Cy?)Pp8fCFDafuQ*-x~-Qup{6Yy6Od=n?cr{Vx3
zl!G8XY=Q9l$S-1A^Q1`^QvTPw`dI4`emnYR!E!5jcE5{yEG{sDu>Jq#SZsvbaL2=g
z4b~by5hvo0cl3^94v8Cc|Ko4@$;{QVM7cO*i2_gdtEb;ntcophY2A<F03*lOuVwz{
zri;apXywSERqHpazW3cvd%Qsb=X-<#&SQ;2yVh-v7W%~%;Qinm@EJ|-R@`Yx0V6f4
zq+5$E07pl-uX^aocjEw$UZ=G+eS3CxNn*!84;o><S8SCqIjb)J`isLR+ZZhF8vXi@
zpIvQ15O)nQgu1a<@rn2|fB)H4al*RC-Wns<cFX-__I$tvr<-#XPX(epEJlb&b2ump
z7zqn=c1aU>Maci`;b?0?>TDq2`cBI!S^6lU3Nbxk+{1JH4?ixl0!F_q1dI52hVUJX
zMs`tv5v&I8l?EAF**=Y#IN5wTW2S$}`b#F5x=%3O&%OG(;(X*<k#V1AD2`Ai3M6pu
z09^eDm+!b?fHo-s)9tP=lx5M|AAaP1CLD_kjG#?+$Bq5%rX*ZA@h#9UbA&sf+LbC>
z_XTLcISLrvc))|WESF!{%O!~&hYbJ7Ecr2B`N7>R4lqJ{uu}3ZPO!9jW>7rfku2c+
zoA0yvT<V%qtm6i0G^I+uQck;%{hDV7UR}5tBgrbqa)W2A=4~7;;PHk44{1|5!sj#_
z>o6q)*>D$y=BH0zX`?e*8y-h|1K0GvM3z=f+$Fni4H^1@;<T)YT`S57^JBx~YR9ZQ
z07eFX^s%WSAK15Cf1Nbds>T(IRT)ABwCo&J(Kx>I<{K=KFjj*7X%g`J?zvn#j(J&(
z(8o6mB&rp1b29ITh1Rv5Uk>MJcHY+idTa6I?jbh%Xyj-M@VILnDf`}AZi;D5x_f|C
zFAKE7Z?aFx+}x2S88A|<N);P7$i)(I<zazCzxot4$6^)pyf&@dwik|%@yk7!=e|p_
zg4R<MU}UQ-Mi3Z5{|>;2Z!r?pZ&7P_s)Q&H-Fr{cHQevlV{5qLZk6yF&bRo~tFM|}
ziVc$&6=n^m9E9_@AS&t_?trnVHYHob$4glCxrA1E(UG)<huc|FzzA!=Kq;bN5$Nbh
zTEhW50Y=a;zxURg<Hm?{QRUk0j#xNcbY2hplfCTlQDbcUHg~_cN1s2zO&3-fj<DhY
z;1AOm;`cwEFy53e6~xpE-}iqP78n5#8F+nPYX#w#LwuF6wK(^YCiIsOx1n4o`IL+w
zgTEUnK94mV*Sv$@9vV}U_47LmJ^vamQ;A9A2y36)2lPwW6U~@4OUg_WBmmDedlV4e
zDAq_t#gQl_p7r*dZ!&eHFy#|5>wfd6hpn5`GP0(C2$;TKJPW+xWdcSnIJc|4K30|w
z?~NGE#>xIKqkYw*fB1Er74sV(d~Cj&JICzK%r<2t*e8&AihzT2JD+MU=^2bAK%p08
zjWFvw@huuxr37?PB$gpqT@V+<o$uPU&&qAMpRMZ_w2=}k?rf1e+;ek(#mdM9B4{Z5
z)+0~9WVY<ws>k1L_F*-m0tjG!=k?c`)&ee&`#nA64fDf~%Pa`8J8PGzQY~Gsu%zB2
zZ(X?UN)@wze4{32V8l4Yl@=Bw8I{Z1zTr=bEOdTz*PWIhbQj+czdQT;?_+UD5Nh)*
zy}EQU-Ev%<F;Bes?ohKxvIE_+;Kp8A$N+X+Bv)u%)pucf^3~VPQn|bpeko83mt{H;
zWoPX%$2V;d8DoU|?~WK{K9`je>33||VSy&Bq<{C&JuwS1(m^O3vd?1w9HxQ*BYvLa
zYSz$R`&vCy&`$0#r(g+!?7*ie<x7B*=LH58#htHT_ObKiJ#Y*dN!~XFz7qwEkQuFe
z7gxX)#$@v|xzq9m>Gxf8t_`}3d!_py$5P-xk${n&r=4K>^^Dil!|e`b5)|5}N&+GX
z_U;SFxgVvmUR+>=x^P9+zIlDQfRf)tP2x|zqs89~L{DJ&Ki+Y9T$}~kvwY1a^FPmh
zz-h6lxWGtb1@Hd-ZI?y10k8wmBTT0Y@WTC8Ed{VgQ+4sGb>@$9SLRbP;r!p^0*gT9
zQB)jYga!U=#TcPpTt4B$a_-_)<{xipk@twe=2vu$*A4)^$K4vhU15ki4<#@{+i**F
zSx%687FAq@{Z=lme8CLw2MdreJ@>&x^Z9p))((6xzSRc#{!f-RixgMKCoG>%Xd2HT
z5-*1sLq1_CgnKQYaLWZ~81Aob9a@>o^-cgSN68JBvwZRM9;)+yH(;cSgpxI+QJxn?
z2~|jsfDui8pU{+g{J9qmn|??5+eb@#n=Cspljr?L3GGN9woGtka71$q4Kkl#)`)w6
z&wif}4*Xof-ous;Q5q+Fqs0M6sH?}>XPIVA8d=k9HdZitejs{WA0hMxQ1XcY+AgTK
zf8CuI7Z|}$F~A6*QkVw6`<}tFB}|w&vspb(!ebcNO!=dJUU46=2!Rk@wr|?Zx}ON=
zMH?Nh^e9RuR7qwrLO=S<cvFFR=bX`58e-M$nnK`0Tq^kjbjBIuElefp{sGreSd5^p
zwpS+ow_j5-uC!<~!u}pz&bAtw2wC`^o_Xc9xQ#$9xh)_*N0<N$01d*_x6kGB!Q3$J
za+Htk^JIzDIpo$U*}I7=ET11yuC(6y^b;%8AWlZ-kVYu;!yE5?7#G~Mv)q84C!fUh
z$BRSWw64>N0*o}1tCpys?`UEAuG`!rF*o#8koYE{{=>xpAg@0@ag}jR?@P7eWm}s|
zxN*Z3Oqj4}!wM>_nM%53zzEhNXLjxsR}=st0cp{y_6chd{F;Z`lh+*U=RaD`eF1(!
z+x(v%V+7$N{dVu|w<lf0H*Q#Kg{buDu{B%*uY<6LKlaKXvsa2sdC|dG!>O;a0CoU@
zFcoDDPX>$}t~DGj<<wHS!ng1<hJTk7FhaT2DpxYS&h25pt0Qd<=bNu6U<3u7dv3ll
zPN2R-+T#fR{e2?O12};I6w7Sf!+xW=!*>xynxYbG(>=G`Xoax|hM#!;C9BQu%H#fh
zx80&RV1ad?yN+e#-(GmZl+Uq}bJwGeg#|_^1AhMWP95Tk!xH+r*9J=fEdjenth2XX
ze_fm)oZDEI4jC#h6XIulf(N<w^2_YH92JYir3FaDI=*p(`qs54-_qcB-ZLxYI>`}k
ziSClZ2zwBp7A;vK4gBHq(6HT<6`;FU@2-LrQTFZGV>)#{-CTBVH$S;OfA;0U=G)nG
z1r{hkxfRXpat<%^;JtU)Jsj`oPme!s)^FctK?B-US+Pw1_}}-(l{Zu6yrCG1SqR;g
z#-a==>twlwy5^D#ta94quV!km@Quxf>#Q>3r3u1LybfGv9S|tth`{dze23@8jg~sq
zo|jTOED^e#dWvFZ1j=_V4@X>Sji3IdmAmbg^#x!|CGClRefRCQPYE3V*JoZZy9Ju0
zWt6w>sLH2TP~h~^0;_`e#(^N{cgi(Zn8u2?8}<3;*0Q6Ff~J?3tSdvxD-Ydrb6nEt
zvuB&3iknDW^*ws7vg#Ac@6fi5x$vB>26tYc&zxyy2|R%qGUX(n+O4$e8guTMXPWcG
z`-SNl@xqH&uCSMP$TGWP<%;%MxlpIpago3XJoF!*e#UH&TqwO#dAV|y<-49YRc!!D
zLAAc`rEz~&=R}3sm!+Hq&l{$~0wcr~3A^-VUU~La_LsRb1M0#0lYM!Ze;p<-0VqEm
zDq{bVo&y#hsyYv6J;#8N66pP8z(_B}Fz9=J;s^Hphm#EMbs$W{{waYg?yV4t;yMZ{
zip5BjKo+p&=a0Q*%k=;7p&{rc9)nNwmi%P?`r4=dB`a7fsw*wOAY6r2m6km56tFt*
z!6ASHEG&qn;nRAhBj!qRskp#Mof?%DGbB-KgLeJ5tSx?$)kQL@TB)K9z|JDDI%EaX
zy<;nLqXM*t>8W?end$SEgfH9Q#Q{cKfm&zmyGWDoe=I@P<hWUL_b)u7o#JIAqVtcu
zt_gWjq9D=bi#h_nu^8t^C1o+<*8`A|E~W?`=O?)W^IOreRb%tBYkK&}?fGk;Of{cP
zmj;JNe4l)$e6K$J@}EN!EUa*$hATCn-W~h3`Dn^4|B`iO_kRN~i@OGp`N`+AW#KhD
zF18NBZA=G4{AFC;L-oS}Bjp4pq`=&$QJFFmtu3G}`@mi+?8A5bN@;x@4aCN+anpEB
zn{c~*VC2IwW6Ye{^K24?s0dTT<ZCaZg#ti<cGUE5W+?{61e?6A806mPo*T}ZCoqDR
zJlgU;u|azBg_rzG_MAEX=W<IFw$NC*e1&;kn-!l(TO44-l|^4>XfuOF3o$%=Mh}++
zXmN&J+iX!fVy+~k;sPT8A_zyr8%ydN{Qd`Ki2`>!H*n!Lu-U4p4K!xj1a)MCw@NMt
zYt^YLpR~2*L$_}@i<$DpSB6+5QB+u9gg#><7A6Emv`w$gtj{s$cWIIdzMYkUOIMsD
zuB!IQy?}N&B@T?ZcJbbLPR11AnZLOPT)X?`8(q@a5n(49m%i3L>knf)>{c!c2sld`
zX<g;7*(coMp+Q_+@<9wN8<-=Y`tl=u!?jl>Y<^~bH%HGh)}~?eef?EeSOE?0y@<<0
zfB0qEggJlmvKT?o5RU0#!4_Fh;f^Ybx}9~VT$go>yR}Cf^cSSLM$8aLxX{9tR+y3j
zkbGCnjZvUw6fi<R1n;s|ToqV-Vf_^*+@jrc%S{%n_3xF`JN3&voaaRWM#xi60v!ZE
z|EU%u^g;Ko-IK22)4rT$fstfucrxXN;h%kGzWZ)&<d8VFhUaz*`~eoMD2(Jqd9C3H
zve9HuMp0|HOIu%YFmRI;rlV;M_t(1ca}?%xNdhCgc4nC}0?{g^S1R%v?zg=p*Kp?^
z59=qa8^hMutepV)zB^PWnDluNzkS6y$2}cuJCw{KBbj9-d_4SoepD1-q>%*Sj0yMb
z)WQCFka&+(Yu3b78O1k|z$&#28XR*e<T5$X1>TleDb5A3yB|K{f+}76ic|pt8x-Tr
zkKqVdd$)j*Fn_;N;K|EU+{oOKX%5JOp=@@RbsY&Xav)W}z&*Do@Ex<~{a~JcaggP0
zW3GC{WMkp-yI(zI`G_Cop6kh1UblQXKm~4?GRjvlfBMw};bt4$X?-$gg20(f1ye6?
zcJBx*Kma@rz`ToxgK&veMhcw70l}JM@2<T$F1T`wEUo9<ueS3NY*{;Z$MSQriol59
z-hY<`J92YL?y+{A`Pb`%)epOE9s6W~(Ybv))3<k^JT-P|5b%6g_CAGT5%{9ayX8VF
zRTe7{?;l{IVcq)XZh>E68a-u-86#I(dt|LqzI<7;O)kHVZ{E;65CTT{cBagb;*zZD
z_9+$<z#^arv0IUixo=BV#T2%=MiBQ3a_80U^fMxXk$;K@U20uvDd4*LEnV?b=-O=U
zO&TgT%)R{+V~ng45B<0-Kq#N*qwOiGBTI6O(@$<=uJ6+){FpYjAn}L7R2X0ces|!2
zehGWv<#L~gE9@`<4Bm;i1MR1MDrr}Ke%VpyCAG#xm3^dL)0b&>Zr^HFuZ~xEE^a%q
zX#Rd?pd2=3*4MG(N^L!`>%iG32AW>2PKnp}ilr*x$%fTGnT@+rOsVQk3Yg>IB})d3
z;40_QpI?)pvHRVRE6v}BjIs?PZoyrBy4iEzg->CD5g3Q3e{p?6Q0|2**2z8BM+pzl
z#DSo{cJ<0}_lR%tz!PuBT`R7|NL*n_elR)y@%GD2T?GgCiJ<Wh2^i7*RV*qlFw(Gg
zHQT<ZMgZIZ;1q?h1i0zqL}28f0!F?%h`>mGZA}J@Fm4&Yg(1dIL9}a%3yf$;+eN*g
zZR^mgk-=IkOi}SWc!nn<fW-eN*K>Ywd&-J(x%JkVugoV?XUA2QkygPxa_^M_3gZ_&
z5@CTf@tYsaR9UX@O%|m8;tD^AKO8VpqfWh8En9ZW{8r4TqoWBJfiY^<yt!+7<k1I4
zd{cK^JmA{O-#kzN{&L@Z|3h2W2`yXLVCnpT`GE<7k_lCkX)`khAGquG1R)=8xzOAW
z6K%p$A{_7>H$6|i_)1)I(v}1;5?5&GhXqDto?pXru|N_6IR5M9K{3WSlucTya<4St
z`UGFA1r@3Q4BdW1{{)RP>Oq^0V2==e;FxoJA;}*$Oi=wgb=7APLX|HCjA)|}MTG@M
zP#6F#El9gueOQ(B?bF-+j2*{LmiwjY(UvT+L<Kw{_<EWqTw;urgvE#hR0~+W9Se8g
za$`brw_ym!7(wuW0K%uRO9t)%im=J|iTEwVae<1uLCD^}*EFtOC&86gW%;iL_(9t;
z3>czC_KbuIxN*aM6yP4;(D*4+&F5bQ40d6G5yp7`tNJ7aeZ<njT3iUwjw0Wch=1RA
zIYZyveEH?3RrBNhB=#H&y0Ar?O9WUN4%E)QC}0Hb(4XCQOM+WGEYsc@mdNLplOP56
zeHoR~U3K<|<{uh>|1MNiU?f$*2wEfm$-qdRx($-9;V-^A$O0qB*6`p<Ihxk+uzT02
zZ=K+_g<Zp|)~X=^VZVgpl35lWt~I<M&+k9P!2m|0<`q{b!uq_{aF@;<OKJ^wz=-9K
z#TOFaBTU(f89@+atV0NN4i9#)=+d6&E8bRXTYNZPoU1b#Pe=Pc9v$zR-R1F*MFB<t
z9PYYlK%A>F$_&SfL0KWn?Ut*5<hThopuPLeKIT?gj$`$U8?D#g9vWwL>e3bT$o>E&
znJ_T2O8cUh-+Ehd_2fcJmLgdacA^LXFtT;`0dvEZmzj?3+QwD0TlfC(w<pZnwE^&B
z!I4s>%{gbCZu;ebS8vHh*5n!A1lAmS7KBr!760w8zx{8ygl9dCC#=n<&G^<#*Sq>@
z<r-@VvUhjjBDHL)ETXgb2<S+))*yQTJ4F26y(8PY@?w2Hpz;AmAOMs-a(7iqGBbb~
z;(_ekmZdxu%&k}VF>MuV%O|X#{x$e@DTc1IaYh*Tci+^{v}xYVT1Gt~U~0|!bv%Zw
zM)m~&joO!@wgIA)(-B~#nfT}1<N_*84j92&BdtmW0Y=6u0{_1|6M-l40)kJM0@K&(
zk4;jF^DT~Xo0X#Hv&yt;5B2U8NCRMV?ipvAo@Xb@$_T!`SbI3YB;1Z+ZIfQLqLlLX
z+2HH<=^N*nefoCpeDm_VADHqLQ%t5}uMn^uH(ccej?~k9aNn)BB;=VfOA4!F$N9zB
z^TGflX|npd_m&$i&+iwMOpz&YUD@A)%YJ%X`@a5eymF&|zogHLy6-9Jb{wkm;HQrP
zBgO3PWWWf;66CyPquOyb5bXNCN8dEWJ-GMk9;Q?4#y0){mK{#BzOcXu?y3+hhlzme
zU(a6}zRX-}&OWiZEJiBC-P<X|nqcBSv5=L7#Rz4hd0$DvqpDX<*Ri~{9zl~DD-wk9
z;r1LUz(~u6wau?@y2$UA*!kU$46*Mrb|G{E<4{3b1vR8mpCJndtTa-!$dyYeV_G$;
zV>-2I9LizLp#Vleud#rrshAw}dz!Q>%c-4M_OxqOFCj2Fjr4&L7Mp?y><XH$*uY3c
zdstx0OE|)K!+L~gEU#x|%r&W7!@3m<cYYKwLK?mSVzl^#brn`tKH(k<H(5UUf#2PA
z`sIcTPdD8=228QL?jNz{nZI<k87m8xIdXZ%@54<|z(`oInZ0wTS*gYQXdpKGP4WrA
zo(ve_`SAaXsl89WNjyOe5f&mC#7y3OGld04(0qL0&fENI>^U0SKP+67aGg26eVf(^
zOw}$4S|9o6GjZt|izNk&*!f&-C73mBM%xzQFG9-GJ9bEDPkz8iS<RET59nWHU<3;m
z8)GD-bwdaE=F!a4K9kOU`UP#ZN)oJA5>t!jD%v{;JJ?W%O`^#&r38Ty+QfKqV}|k}
z+b*>OfiD|E+b*>ew_gaE5oCv{B!LkaWpvHiWFlNGtWYJ~wy58(?|D8nr`RO$F056O
zo;w_ToiZFS5-#sTx%D`ueSDv|fQo)v{D_D9ZlVIePtFPaPTa4sRjZ#ruX7*7@GVBj
z(_I_BGdjgvWbn?3z4POeC35d|vb5T}*o`c)I*8c;`DNM*Ie(IEv{>^23H=x@WC4T$
zEDeu;S7!63O>O)Y_uC!EOS2Y%pc|jeA@xE5GWTlD`o}ZR#c6rE{2T{dt@*Ql{QSxF
zC}5<j03x(Q!?bSwhJ@g)q~ZHH@q{+gWXuKaxsHY^K#*%U*N$c?3NV6zu>8MK@b#j^
z7;*iUYz_a*qmM}vm0|-CA6vtVzJ?c9xuGQ2@X}I5s+3y(KYk4l&v%7g!`-(NE{EUF
zVkAYtzt#3Xg4S@_jVmogrJYVX(VQ>WHt-rz#QFe;jwLEVzu}!rlIR-()x#qrlz;qL
zxir!-Ozcs9u5@9`O5F9Lbbw_>UQ`rdgz|`sNNgpaaASo#E5|DUF5aL$5Nj<l5^&|}
z6BaE)-uoc7*4C68*jp5X23{5Z;qwR=n|^+G4p5V({X~ZL74(DKSD-9`YfXOp1TgZ7
z6vl9+m62A-x=l>gdI!(CQ~lYdNn?Q{eXO#HpYMZ@KQ*6@$rXTOQTc$ZOIkK>W?kw2
z{)uOTfbtp}Nbbw%Ira+Bxb%V^rkAX7lM(&3XxR!gNfzaE7cMq{Zx!Um2_RtGCM@Pu
z6G(svw^?cFsRDI&+koa=+bw~AKuV*!9k>EGl3uxjWCQ!nX(zQa*IaaAPP*K~s41Vz
zQe&ze^C<KELV|zB-YSjlCteFIC+y>>{<8Xym4>5KwFPQizrGoG?KSRl?C5|IE|7Px
z+Qa8`Nw==%>`uuQ;MpTDk&@t7-+pW6Db6HgxU4ebfdWEy!!vP&dYln$G4ik1-;l+~
zN}I+CHg(2#%QxMi&T)|zBVWk!<GoR%<vOdh$rgV~9c7dr7hhGxEB?Qq-D}5lPNDV?
z6+4t0g#kv0>v!9LL^&;1;7gagXUMIU@BXB=ZS0;i_o}fL^jJtAD*kn+6uT;vF6%ep
z-};<2;utVe%vYQjFhT%p0$GRYm5)C+(`GNS?yOQJs2%(DeDmsvsj*~*1x66MKYCxH
z_G+xa0-g|v>83xv5|_HTzz7o*u~u*a)uu@uQ#~U+Hrd5h6!GIo0Y(6?0I|Y^Cdt4*
zzbb)6?n2_yR1^S0(0TGUIIfx%(7Sv29S<fj0w8sc+SNhrs;?mNjK6|XQGk(xmR($6
zgmLODD*!{zXn(xvd_oh`sCM-PORBuulnfZTK)~H~A@^dR$_nI7S%WyDZ~vk1koq0%
z5Cim!r`}bZf?e)X?3i9bi?G1c`n3|ZmSTx=PKdWNc;qy*K!M={)*+nY(ER>)14hb=
z*{<K9L7ttL7ckQC#P+6Fk8TP1u`EH`99J<gp)+QEXWeAEgoOo0swl|dJ+e+IE-`2Q
z<?*NDieZi}4lqJ~{GlRLMl&B!glh$%f&e1}75urngrYtb)MBKD01PZn*tq!wQ%Qhp
z7l<CAvC|9$+-cpSxgo$h&lTQ~`sIYrOA;8N5Bcvlp^PKi)TU(%`OmGDAn*;h#s6K>
zzz8Z8_&Y8MV8my<*$lL3*3`7qhMY}z*cF`rOyP3}jD$6unl@HkjU1oWF3;j+%WNDE
z_nUfe>3glIrQnE;*t}uowQBVmcd5unAH)C{d3NyIrczabJ_-z4TkrCge%B^^Go(W!
z^m?p3>eV%?v2=JdaHB;-{4)K9B?CqPl>in1ntY0~BJ=Mj3wHm-*E7vHZMxm~1azqx
zvIs+X^Mr&qejFUXY+{l$Z`#DaWY4335kNmIOo~g}Gj~|`gNJHkS={#Kmp?BdWwN}-
zg0BZLMhL{5Yz_b5<B!e4CCiVk;mHP3&h`A(@Ztg^N5dLE;ma=-gm`j-FZP4IhHuvV
zT2;c~FzqpN4`TwMX82yz?~(*YvUcq>C%0*1PHfw@$ZL4GjU}~)yWs0^r|^Wt4vJcT
z!u8`?Yn-fZ=Pg)RVggBcoXZ40K$!}vykJp;f>TaIXk&QY<`%L10B!Q3q5>nnrJO6D
zb^Pg<U$x2<TyrtfP&S|~H_9z2AciAcZayowMO(KAUNwlPMjm*E7bMhPFBi4Rh&_FS
z`cl$tdxB}CBhO#ZeT4%?0O{aSw{6=j%aL>ojBJ)QK}7)%cVE}nG-?p9l(BNvYV+qu
zpOhlUZq-#<{7iN8zyI}s0+5%p@kd@5@~(i8EK^$GD#6+HAlB_wiuRE|zqP+%oaRTk
zzkT=PQD&ar8^PTJ3V;!Sj({ZG7f)J7ieY=R1<0u_Db~Fe3Q4l;NRj0TF0ZgI5r9&z
zVmYe-iVG_g(tKJacUq4RdR0NjvrL!H9nH1910~-P6DFE*Qu^UBz$11m)<!7>0dWD$
zaWRDzMvCfcC=U_$MDSh!BOi_!YppJ*lRQ{>poDs>Vh`3+9J%~x?Z%DfV_9R&{AQl1
zDC-RX7QhPlvn;iV`jc6VJT-WTl!unvItjR5Ui|=;vU|^dfjM=|&!kir9y8;oer1M@
z3&1bx*`<EWP<$1Dku(88zkT5Tg#9A%hXDA(R2X0+nd`=Infd+RD0gztz4m&H?@pG!
zq+)}tk=K&iQlLH(D7?2T={BL{yHAQ@$AFPy_H16j2%7j$JaDb`QSMiR7Wg=UA$MQd
z-CllJn&)#p&GW(nBe<aY>s@{PD(rc-f@DPrWD^qRTtF~{5Z9;T03#`yWUlVf$p8@H
zF3Ts{MywHB5oPV%WpFE&F2S--M+z`PAn!iiI`~bq=gU@aG{1R1aGB%ZY~G-jb&-ap
zf=^fkZOP2C&%AMKrVW7JR@NC|I#|F6LBVh9eU2gSiBG)uE!(rq3cZU>a^qB2F;!YM
zuHz@M=RPpv0}%-ci*TX1zzB;U3n^~Ta8DQ3?B(9p9kOK5Gvc0#_&TSwOk`q`0V8Pw
zF^CCLt7@g#(-9}<@wdj9MXT1^Z~VIRPc^tl^XbEh-<bC&%=9nCo%1gG^*YUT6i`K<
z=0#YH3>`Paj8qH~$3Eqi{$SjT0!G4S`}lxewq(h{sJ_I@ENt*~G)?Bq6xR`P5D1PK
z7Q7SG8!!aM3G&@G@w>3V2rj7Zkl-;)xY{B3a6yT0osEr8d^<%2M$r7~t{4Oz1P~!y
z_6hAbHcV`OvUg=$fqs~2KiEPQgzQ&ec8T9Cd%i;^?r4p>fa<(IHfw05Uw&aP3yAoc
zER!xvC|oNLKLzjvkRnVa35-y-n`hae(Jo(b&wbjpBWtH(jck-#vaJG3%A3~BkLOk#
zl_W5dwSSLE(OOng!PgzLi*-nEJtM(3!-TshVqiERi_LPF9552T&bYi*0(d`wIG_jM
zCGjS-a}ua`g0;%<H&JMb4gKh2yJ-xUS(x8EUyHi+;#_ID7$cNJ`)=-gO@d1)#^_rg
z3{wn}tF4Cf*ol+nmLYd7DJU@F`j;_#@4&>3GvoPRFAYliNLX}W*%2n@w^8FK#x`j8
z+;XF-A{Qw>W$BxF@`aa^YPDW*(FJy1@bhpU1&k!ySOUKI@r=S{g@5DT;?=j_jkP;`
z|8V|x@2W6uC@3)EuhW5`C{5M;`@7QGr<h2cPddr;>fX(Et8<6x`YqWSzIf?Uxzm{!
zw}vlVzRbwEh%M*X8Xk<&;;i9MNMN^Hd^wbC>>mwlIKar1&%X#dB=mQ;K#Fn#DblKD
zn0^;u61j#mC&613b`4J^{5o8~2x~Z&mr-EDk+6n4S6c2BVF65REWpt)|LW4XBfJ*Y
zBLob0N;K|XQI7rkTpq`h6s3&s;ZcTvoHQ%nRHN!Ot){vUSdd}P_dgh1iu$xc>vfGB
z!Fr|iH0z2IP{5}TlxME;0(i0~sVt?fuV#K@Mo*X|>yfKW11WL%<huuU0331Sh?NE(
z7DXrlo!hO8IY|IpcyEQ}&Zf;<?Dw)uZk*f8DwOm-IagZQ0x@=%)jhe?s-UrXKtN5V
zK!&mc08Vb++Vs65es91#{@W8zn<cXD;Jka+vt=!Nu`T*J#TNN?PO$eYC0>d^>0%;+
zYQ@7<ltSbcm-Mnh>%(pK+r~M2jnAe`Gb1KU7Qi8;N%5d5iZ6ous{L3oXkTSxiYRTl
zbP4SxybtP=>n<s?9#9~4;+v3`|4>;e<KA%jC8k5jIvRIc4?pvgT)uBIclNu+v=qn&
zVD;B$UNqaYx2r9R<skq90H&<u9K_owCrcJAM=Ith&ep6^%lzzy>uo*$pV3l`88$Ki
z83^9)XrJ1bkzUaTRwuxEe*M8XAXdrvY16E1!^v}Y?ItjKSp`Ku(^`$V^4$?vS}PP<
z0mU|AjFcBB0};0*S&WfM)4w*ua@=A8qGWB$w$?j46=S4&<qGC^4?URB-ehq=qJR;`
z=ku=(Nr<h}xS`?+DQ-`gupVL@7nI=F-+X^q(r?F=!#=HI>G<!$-7Cy-^~SBHPAKd`
zVe%hsX^&Zql)xLy3mBpJU-s{1+BS_Belrm*Sg}^FlxoY=FWb7-@)z2|0wXmgto_H&
z5-moEQSj^s6N~kf#Q{dH=z5Cjea^`V^?$QK=<CQYEHJ{&BR$5*y$Yz^DHKqfc8~dL
zo*De<G)`kgY?(jZak=%g7`v#GOf7zcKbdB}oVVC6MBF=+F-8a=|2w&P3dgkAnz_@w
zs(|A&e^?f4S3zQoxP{Aq#^Rh87Z{;#zz&`f3o4&*5%;z%J0^X*(7GexUSVR4xU#V9
zc%bk3riKT2@!jIG4h;$1bvc5m`eM#v^TM#nu{zzo@DZ&AKpN(f77dTHLa=Ze_Zx5o
zFlP3mRqoOuJw^c|Q4Ls{V;&oI??}2<OByX1Az@y)eZIxW2cyTBxe}K6)K?pvri~i<
zmn`E;{11eRSrS}7{``yfnoCqzp-Pp?vKYDLCYL>SOr{+kOHhE5;s7IEPCw0bIlXg2
zEd-Ke0|^rcvFXGY@u7|}a`tJbYU6u)Li?7jTxneb_%|@e?=Jx`0%xBv^PVKls_Ea(
zin#<TihuIL#u#yJz+#0Uv0;DM08hl601$BQ6q5O)E=gbnVN7|=ZzUCDgz~DCD{rp5
zqPH~o;z1EMkkh`NVO=&kR~Ci!PfkA~`~)ELT|F^JV=ck>$Hf3gN4Si5R{{sV-I5?5
z#0b06qWq}ui1x7|qfgj4J}V*O=B<I|C3*ZZ`Ryp#oPZ0sJ8!zdT6DRz9MR@{O5jB@
z0#HJT6Q-%!U=mM<cLWg!1pv>d{47SCt3f{x=TX2&vgiF$LK$466^0z(SeQIVYFa1Y
zFHHNkOM|w8G-x3|7Jw79{<uJZp-21coZ|gD{HLW?=K_SZrZ}#&$bW1NKm2RB8+*JD
zpN^(AJQ=uiq^;q8|8njG*5T`~zx$q9x+2ldOVpgh`f{YN;Zg18nc>rMt<|ha;{+jl
zxKB_9nEuVT;md{k4ZndR0RrRQQW9OhD*yc)Z43hd@O`+`Vx3J!MYR~AJYtfCc`~Pf
z!k&ls4H4%M%UWj<VlU?WLEHWf#av8Jt7w6gR9QITLJL<D%}mw+06+jqL_t(q$*5QN
zu2y+0d>=4p?4&6%i<P`$j5t?XmE=l`_0Cz00F0!{5@X-aedd23yvMr9_3Ifwd8&E)
z{o$AtnFr-Yt9g?~R>AAHfBu`bL@2F&!afCaFP)Mq%_Lcji2uQY<bZfItVr6lZehBf
z*2%PN*3^PWe%byx{N13T!_4%Vvx1l*036C+UO>oh#U63M2$myQZj=*nlC1#l&gB-0
zNu>qQ0E_^Vs5o<C>y~EVb%{zW?+pLM%wN95{Ob1G>=;_Kbh-KGkT<RM39&Ya?@?A@
zNam&;0$x&VeYCZr?(I}tS~Y8I?(BP=-+X&MPA<BJjh!U$LO_n>nrUiBw)!D+%TAM4
zS%A)-1Ezhe=H^VfgC$N0R?Eo<K63D|;b!iFg*Hh!J@YPEaR6SNd*+#u!PlQ2B3D{D
zaXX!|DE*2|v3AW`CMw3rms03?Z`3FYo|G2=wOj3lA0#eGby*Vq`o8-TfDu1#sNeo5
zU<75w=OTcS26gL7$>}OTyFH)$#g}HHg3=d;m^U2XU6_2w(&pa-$AFRi`zYTO`2i!i
zQ@Zuivy+xHVdf9!<&UQPfA-D;PLkrx^3k@c{dk9)0S52A_eOXtA-qQjNh={~?{@EY
zPrIwVTO$t$5LOaM0!bhcLJ~%J;k_3y1B~}Q?Yr9E`$tt&XJu#EuCA``u5W(RnUxtC
z5gF+sz8CL%3H!4JMv`fac!k~*3drI|uD}R0?7#Ds&l<u7c@1IdfBwnuEKuOK7c($I
zXYaaio)Z|3$;+T{=W_0``<i3-+sXZmFN2r}g#<<})jdcEEwAwbxkZlqr{9Qvh}^!+
zfRQNu(j7C>Wj=tBlMdR|eD=t_hE?Uhr^MQLJaHhM3ot^)JZmsK4%m6CVdby^;9`{!
z_J^K-$2{}qhqkj2m@r@?emTOf9kSat3In#K)ub76)eRl6boB=FPd~ZO&S)9hT9~y=
zfsv&ORg=vu-on&=cqyAnteTE`D)Ry(Znmj6J+s$OJmx6xri67jknnEI78t3O#=_Sn
zXy><N9X5ZuRKIftMqnopn#Uh%s#T8GoA3B#0+VGdfDz7%8Enin_gfH9fF8j~#~$rZ
z8mL`SNMK~2J$5t49CqlidJ)R#(WjnH*j@-=q*Y-CYV=~tS12t`U~A1{gVo`;pxC~-
z03+nfjL8G`-qQ-KdqEIaCLlL@8?U|jws}#G514O$OXm3Ur}O{zI`xBqk(o25Ti`I+
z%sOURUVqE262%+>-@SQy%S#4~koOhmpF0ekME~fzTW(JZ;1f5TkRyI0jMj56y<)#Z
zExISZrmzKG;ED7AI^VngCxNlk*^gsotAOwmLBPn=DH80x<RT7Buw>_(kjVFg>relu
z*vuPNF4MXtbM~b=9>z#mV4<W=6MzwS%wz&am|1ziF1wY!;qKyVS^9=Q=V^_h)m$ex
z6x%m^=!~Lk`2a@7!#CW|81c6QjK^Je*iNB5w-@`dF=4WVOA83W_c9e2;nF5s9<a||
zmTArYp#DJ6aVLL9LhaHG^xGdEdDMn^@^9z&s*P=$fV+uw`p&ccA8(o<#)x-6>-+7w
zhdIipow0iDI;+ojj6!0;6o=__i=6C$G}EU}mALvjmQn5UVgk<%xBuKSNnnWb-Mw8Q
zLD6psvT0&%lZPJ!q8xMq$<J9TVC0s&?lRqCV%2E^0N+TbF|xj+*XnY#Z|o2o<OFl*
zzKL2G8^m(|_IIwa+JgV|^)JimU!@5k&zrjhV)XYbtX0g)%PRy@z)+ADVN-(yz*R5x
zAuDUd50k@vro#B{wZo1A+;%ool7ts-ZEH8*zWzqjr3uml3XvtRiiFe!j0gb15Lt&k
z(hr+Mj1VRxvCXEX|Drx!BVywUVMV)E^cnLlfj#ICdhInaUEW>1$ebJN$Nl;be>ShY
z`<4X<NXO(%z>#jTCdj|GA;hT&KvMbYE6tiQJ1Jx2Q2|`-EdU`yeN*&1upR*$u9r1e
zVQTl-aVN9;j@!$(p|EH+MdBYcGrx25Ef8pIxosWNS&;zzQ3oDij!|R+zrh&!bomOJ
zAs)tyz8xUVX41P03Z?a?z_p)>i47~p>6ieJpf@r@ESJB(>MC2dx2*RB*4*=lKYBOW
z^=yEVWEvy>Fx$fl3D_{w{ww|UMrl9=U4ww)ie^)?(2@D9U`tE0S-W~|KC`qFciw;X
zC7b6nZ@e7OyR2m}GI;Q=(HGEmdnCSy;;AxRStsh$`bd>2o4z2!BStS-I$#8V;F_<W
ztq(<%;^nW&ckX;hv%eE(WMm7B@KU?(o970A5bfsbAKvXRGOORY0wY^1oYeohB$1{_
z!9hR}g5Uoxmc<+&U8@*EX(e-@FJHUS{I47C9X7cBkE>2Mn@GFEYkaV-l1O_8lu3^r
zz0VHj{NwibY>BSsJCqhQ!~gu!8D{ph#)Ps6XZF?a-(|WYLmzt*oP6-E=FB680}Mex
zP(p?y_>m7_<R33Q-fSNb0_=ZW_gf2Ac&i8S@TD^k_inQ5K>(IZ#B;_qJ55ZeYR!Cj
z@UL&0=C%&I9#My-t+U&#(0#FLLrX@02>HQofq~)|aYv)(K<T8MbntFwiie3pU7(kK
z?T2>_yPsS=M{^kjjC5$mcb!B^z1FD?=?cxtFJ{A0xaHPcr>&wF7`guuv31@`WQ+h1
zAdU!#8f3rHY$*T?uVo92@E*Qe;_H48g4oB`-jaN_x3|n>>vXbm38j@Rv`r8Y#7v<p
zG}xJ@fs@xp5{Ba^>J+!ik)6z#y?%+b8werfHNyJ<+~5TF+G!fJ5m(u7=u9wk-)qUj
z3*|eM7G-{3`WRa;n9pj05u<tan#2wqAuP~s@LLcN<WH07Z)Wy}^i9hbFoHAq)FY1!
z6tibA!l1-!Go~r5+(qYmH`#S}-+4m@xkL;AaPS{<_+g4wU1#2S=UqF)5wHa3J>lr0
z;ThH^1qjG1R{?e}h!Nxn;TuiJEHrI`?8#@JGmk$0gpHT&_IvYm+gPRo3Si9;n#OD0
zI`<oXdWS-?b$U0)*Z0_EXLI~fN5p@-6&N)DSJ?8E>&)L>abaK>H$rLQXbrN<&O6&`
ztoZi@0V4>l5(>;Ow7OC;v~QJTW+GTFz*{cB2pS``wRIX8PyA3?+!x+&xSQPIZMV`l
zJn`vq?^5s9(l^|gT!E27f5U^mf!=HBU(W|Ha_r%U=ItBquOGkj@eTI^BlLR-{Y5_!
z1~L5rXY}P4>N|Ou*v5XQE@^K4#a#x`<XlX}Dzw#*9@pLX{ocT^PG?LYa10G^+E11)
zHFy5zUhk&(`ts##%#J&3oii|kGnYQJcWi^s2)ZG#tI<CE>x*K4h`Ee$jp)btCnMea
z;GgUS2qqs94)qrGzmw^X?IOqd$;TdJ^Yi}kmgBb57r083vtR^!jZj**-gUS7H0_$q
z7!_Yf2V@ZNLD3x|a)JYl%->=&^YzOv9L8k0?x(*nb7#*qXB>Bo_3?b?+Mh_|d!<5J
zMaK;2k6;yIjlk&OUSi?~eX{pI{#e=}gYnQ<^$M50>vr3k&m3}~VN#FR@D2UT_pVbY
z+zrwY5%2-~Lm|olS;*6|L(mnmNCge7R<4qk#sFcr(8Z95I}UzW8*S^`#qv4D>@S9f
z*EWc`53^zqwLi@3f4}BNF*~|-&uhX^rF2UKBmgE50t@B{22c`Gp4gU~Z)&bQ`)s@J
z{qvA?cODinf_OOq2zd~ujPWWU<MgH}=5L>TV%v}j&|L}%*3_u>*nT_dh8$p;#GLXP
z&*1;P;YQOY&7CUc(;?jufRUpGiX1I)#BTs2i<d05TDgNU2!?h;r)0JmBSFB(dmk<`
zKfCXDrlUP#HCzCpnx@s+&|zE6n{B>$!3Dzr(dQMy_ECkh^;<T;2=^!~F@H!q!fxLo
zW?mtU@L2d;?d_DJ;B3DmDYDATt~QN58+^W|ykzXjWW)M(W`mx`d>V}A5-{Rx(W}os
z+U&l~=Dw}b@Je6#-Y@ksiiRR^39<!7IPM$g9BX!6m>7nFm+m*E8M0Pd9_fq+#38%h
zw8?>yZu50HXM6M_4>Hb|^Y$AW3BP|jk%r1F{Qr{UzSuNIIQM_K{A4qK?hOBFm|K4T
zw0Y{a#Bc*7y5WDGudqth3F!eIzVM%SBy8gcniR+Fzw@x|xd0<K-ur}k_N_(3@?ry>
z8i0qIP@?mCLu-fm+7Ip;cFe^m9ym-ZgG9v+jQsbPKdYC0BKXrIz~ZYqM?WZ)j1`j%
z7~%OiOXBOt?Z1=#X#VNBx6Q8}eqmVKWx4^;`-@IE(CjCGrb__T-(T|^gHXHQW(qj@
zhYOA~gv)bA%+Rm>;BJNC%TJgWmv1J^AYdfuI7H<ZFABaNgVK-1wrLs=@jsmfIaU0}
zeC|{^9Jfq#IFbg3>kn_dSsl;BcYL<M2zdn!u6f7*SehZkrq0#qNCpEVT^4k4VVeB;
zZzCuC=K`GYgx3Ke@EY18U%co7QzPb!*Mfi%2ehqhZWSGRiuv+|7YrLvq!VA)DT)0N
zJj2e#L-y8#^$X5AQzF-!By6Umj?f)WbFb~JnR};b2A|(j0VDI~%&}S${$mJ9^w`sX
z^>2><MsOP8X|Hiju9GeVv;Do6?7qq;FhY5~id~&AU<7CIC8wQcwp*A8QgUy*5DngQ
zOy(d6a^|_?c2%6?<BvSThICsjkn&fWF-HD{GB1b`1dLD*?<`Vpy|@Q&Rmgy)8bA%I
zZxJqw0)5^*z3Ec{BM0oQnW_gT&b)(?&rHl!K9J4jd9-GX?yH%jUV}|@lY}qP5W-*)
zop{K>HYAug9qYf8Xysys)9@QCva7GWTz-m>f0qrVwWdvZD>gU+*%&}J3@|?{1)ZQ9
z<f!pK3&S#<pc}jn{!CAiV=DDdR*;TK$r$<lgAdD*RTICrXt&ZgeCX~O={I~RCzyJg
z6rmaGey1Qa?r)M~E7|0RfwBQ}>am{*oF!lShI?Ch<m>qWMuNg37TY)6iLW!ZCw%38
zav<ZVq>mRQ1dhH26HWa6!l21AnO+4ZKU{Lo=WTL3K2RtuFDR2NAHIKoclo6@B$YP}
z-|y>gy<L+J5-0kQX8DS>X3K5oo3CDU;V=e9CX5l<=?aC&WHO4^9+q(Tkp~|n$9=h-
zl;EA9v)^7~nW!(|wU-r!7B=&=sZ*p!^1Ad$p7iccSU*q<j>7|-;11CCCdT=G#RcaM
zJ2!xlo5dLEQvX3Qs7!fRhbhwpwhB4>VT??lUT@Z~YqKGvzIw^|(!`nX&DK07hR7Du
zO4(}u=4Pq#`p%DUF_HFGGoa8}fFgrh!Wb#9tTvUpf6hGhcypl6n+Y!e_#fYkUJJUH
z`dul<e!nI&PM<Z^T>1I4%)HsN63Y6&KfJ-LlKw}z0Es?@twMl38X8N>N{O%Q*h=|s
z&;SXEP0}mS1VCg^zPEC{TWb{l$yr3lZ@XZ=`O*dFxJ`D6&wvn5jKlu6_+xW}^j$FW
zuBodQ2r{Vi>$b>NwI*}-35cP8Un>@fs$;ehfOKU*D6L2S`d9OZ$D)7%`W{_feVTX_
zGN&AMl-W~}6d3pZ_f0pM7HR8N2&@2rAw*W6`rct(7X;mBdhRXm<$t;627x}^mcd~|
zbLl*jh0^-bPo*)kYJ~-K!Z8qsv0tA6o=x?feMJBy<6Ez-7c1pk-@Vq<qKTq&>(RL)
zINvSM=J36DGpC;r)dlq)gC-#TcCRsRx#-kW%?{fv^lrB892826=a-2<fQ?>bLK2hw
zn6Q`1IE{2emSo{5z4=n_wzaKh-rVT|MiSe`o30qw+f?@kJ#(XJZLJ&4>Q!s<T0~xe
zk#s24cw&suy8A2a&KFNVY*<4-EzN_Q?|*98_FRCGgLhkKE<fdvVR_yE)a&M+$6pz?
z-Max+1R~InzLfajyEOm%;uFlao6m7S?GoW>X5uGv&hE1HrskXH9-D9sKZ1aftu~oy
z{`u0AhSArejFI1r*n!bhS7WYKI4r**{{GEte(m2J{|#sXc;UC#KU`{l@T*4x>qXeD
zMwRC^+Um;x8V-ykV>qM&MhM|_Wx&03r!-q0e&HQ&vGH|~`fI*%&M-z%5HP}r{hxHb
zx7sxE<4tD84vYk`bu2I<R*WBXfGq-R#Ba%f5wt(JUbGK}whS6Gxf%e=*~jc_4=4QA
zw{E`Q3d0BGckF&UnsblcFTM-S6aXzh(<h!_6jw3@MySHlB}+_1z(BFuRtpyn>of(`
zH4|WDC(Q^#INEQ_On=~^M-n#EF*#zdy(I>FSi*LGc!di6JClWaO681qc`#d^nes2c
z{-)cMu)IJH$IUmHm#_`T1r+#XGs(A_KVLK0PDwb1A5Ulo(KCu6>@}R3=w$disX@Sq
z_qb$Em2|)ev*Mxc`(xU!S+~w=iv$VQpDSU!1ro^fS~6gS&{l-U@LQ^DF<Vsf%g#Gz
zSR1_l_B-Ye4?pV9D*l^6w4X_k3>YC4825?a@(GOOJCqik`SVXYX_%OOGN($ifwKcf
z{DT8#g~Bj#4tST&-$o<(%`@GYqQmY5F}%Qt_c+2(Fv$0dz%rZg+S_h3g96Y)HMQ|-
zdh?8L)JiH~1dX4|&OLkBbMdD~A2YAK`Ia{uyUt+l{Lh_Xh>z^Gr3znwbI5^9uzq%2
zu%(^R>VL+Q#ZYEeDeoPeFMbPRjF4vkz4kOm9&(UBxz#fw{8=jFx!`%mAylkZs8Cus
z2*dVXnz)P+@3R&34fg{|*?hyff0BXFrEhrR(={2s;V<gl`LN!hn48$aI<uH@J{8+H
zoVs}}AHc|I>i_Pjez4ba{f2w<U_D>oaGv>E0qt9heZfQ{_pSQv0}q<_Km5pFto^-U
z^Ub9}bZR^;8xdEgFPV$c@8{itA!3f0d8YQ6BM!IhH-Gv^pL)hTFD*8|5y4=Gt+x<k
z<lJF1)iVJ`$d6E3gw*odx(ypFSi)oq!eV)W9YB;z<pB0FwPB3b$SKdZJAQqSW$1YG
z^R6fR#s>i-TQ8Vzfh%_%T|Q2Hy<LovYJKPW+S_a>tv*fQXdf&$4Ps{ii~vCFwd+pi
zg40e;xW0r7BODbB{rmp(h<WI-sGqLCr(1oVC^(`Off-e1Uxh`x`15DDqRq8p!z>ny
zCn81yEDu6cQ4XNTUJ9#q$yukz)AD?L>j&4Fb?cgK+6qm4>DTz)8eVYqjbM!cmSC1a
zqHTa8Hhz@ycSZ=WzAk|$V$uBb)mK>%%I)(m@05PZbJ8^dP=OUvEA5g#0WDpfJqoW`
zZ9`J^bVkHV30cNSpUO@Kj3B=L=(EpSrii-+`|I95J;wL}=(zqDzc7oItkAWWW{7~8
zuz(^?2X@K+yYFJo{LIj}835<suDjlX1cbMW=r?xEK?emgMt&^T{Bi*!<i|N+^e17h
z%$0uVm(M#d;kvS)HvfV8)9b`sqKTb#Mf%L8XPjcg=(%kyNq_zA()Wu^0`l(5hB3l%
z=v|_j<+leOe#E@t(=PR=A#5F!e;iBOzdxJb>!h7Hy*6yWP3ey7RNua~JTdA*x?@Ib
z|B7`jW}fttN7ITH7)f{1#RQCao#%H+;Q5cw4j){G6{GN4Ii{H<?zQj#@?nFM-EV$J
zsN0_`fA$lq3r!EizrBVN_1m{SkkskAaI@Lwt7jc$k=?I;|5x5a<LiHS){$nf?dHdS
zv%*X5yT5$Myq!b@dW$*J&9}dBYQi!6ct&~_H{CY`2Jtez^7KQ^{yS|oEG>@v{~^Xp
zGFHm41Cnf*rP=o14G^bK28>Xr19#cVTzb;Ma#BPc;Vx^A`Sz_3n3g27^}i+YZ-5@J
z@nWTIH%jQ-*_KN4ErCJqW8${6OvXlG-_BcYV*dWTp$`e$uovI?*j)G9XvaLdW#}&<
zo4$R;saC&YXe%L^dd=5=^lKXe%5C3a!94Rn&;Lwdy;1=-?)Y4m+`9lbex?n1T`g7;
zx-^;C*TqoUEr6Nzzi<4VL6^hXyctu>zlsSI(sSvz-~8oe^T0E2_%|Cr!vvt^okc6m
z8qI3=d|Md_nMiN10K;yxL}e~p)2tZ@eZ%tevvMAM_Du=ICpweeL4MHl6NN4L*3G}S
zA8ePVP|Iw95kL{ZkzzCi*AZUN#Jq8KAAI5o^O`!9eydhz^y)8MHf-<_M3e6x>@$b#
zxwmv94)JGWH6zfD2r``fS6zI86^i!mhZ_Aei65?6x8A$i5Z$$0T_2p9XDYVrz(BwG
z(M`7|q;Wskzn#DnuMOzEb(g|`1<m#*-E7guFPEd=Z(heQ`*_Eldfc(Y2F>haw&?X@
zyd-0#;LK*m8Zl|TmI@e|FOk2~PWa3)dVN<{x0nL=C5>%-!qG>X-FM!}TZ~=z&v<W8
zU6}dfr|**(=PaW0L1slTbuQikBk9nOxcyhZPTB@Roo|j~K#Jg{vs#7ZyGjEA2LQP3
zET4Y<MG5>p=iQX^I<t1dI+__}q9K|ueE#fVp%chF=-v-16bI%hhaP4CJp6XwpVY^A
z@BR2@&JA#h_}_kWP6j_P;%AsJ_(BkrnAOYx^cweJm)`9-J=UyUH=;P=-h4NmJJ(!v
z_UG)TAl(#pSP(*Dr<?fzN<v6Hq8XM9>ccfd{X6f>Q*T+x@+S??A|d1av)lkWpz|}3
z^%_j2i_bYLA(RIF8LkWM?TmL?U1i8>fcWEE6E~XB8Q$37son|Xd5Z?TI9>eCYo0m4
zBmcA4*4`nY<~M<ZSMtMw0&v~Z+G@k-v8@=bPjBr?y|uG3e;Z|TEXUw5!fERjUw6k&
zG>sAOHyrWa=j256w_7&faN22~J$4`UZ@9Z2rEhq&C5rVMzGBT9^YgoYW7_06?bSQf
z50Yp44JRDWr2?);`VIHKN4x-`yPv#E-o?qZ*phw2{m%~hG0Q(+-*9jJSkKotoO%O{
zAhhnCm_#V6-`xLz`U450Ia>{wz86U}neQ{kJxKUk@Au5^$lwo*=`UY;v2|XPJX<06
z?KM07j1!L^R?_+n&F03R-4Qq@WC8`ioU6Y(80IXZ=;@5fLg$=uTA)uLh*`!o2l@=O
zo8Jii`;(vFVFl;?``qvC7jA9NIOzoUGrshPGva&zw-AB%8+~eYPg>eqC9d9JTBRFM
zs|hmnSzG06uM)eU3*bW6J0x)9pTB-p0;7#S4$lSG_}f3Y-mFt!P&5rs`wE=sSN~Th
zdy1~d@kbpNufhu&AO7^Nd+hPyrdre8(rH*m6tZatjj_IT@wxG|Obq*{|NOow7a&6T
zErirzm;i8;%U?JsXFvbT<Rhu%xBPB^P^^<4?GY6e{ZFp+-}LjWlaDut?myfCPv7?6
z#K<6wFieq-=1$Wn_6!2+m_28KBIuf+W74P)Q*F`@*?;d{&3UJsY}0uEJfjd(cis0#
z0Y4RXbES?u`N+e~q5BN=<)71-=$?ljwrr6|YnRnv2?>nBw%=-NbIF;f*>(P`)qVWE
zTW=G%p>xx@c1QYch_F+SI#L4eiQpXR@3`-N^NRFSYG^ai5s&Sp`{=^chw}F9GJk#b
zRdb&vqq%2-P-N4mHkogFo=umYF&4~q?g_7PpP&2L(+%TIuW@|{?dl(olcq_b=NLQs
zLt=9u+@Xm^Edmhz#x-GL<H?%z24MHIyMN<PGHHJ62^7njmXF$KfHr(1R<2y*ipa~-
z%NY5}1HUq7ZNAXvJ7G0Ot~m7&bI`8QlYISyCFX~BJsS9N{qkpzF#GJVAl}HZ2neIa
zY;nH@31yS!IfY8P;oc|w$>YCwUAUR~+UJjo&uVu&Hnw#dK$SN>TxNRp&_Zd0)o|Dz
z+nOWy-d>cpDzkiTv-!vC?{$YHEKiiE_2&XatvTjb55K5mC2XY^^{lCl=AFgM?KzPr
zx)y$*rd>K4-<7TkPGgS2;mm(t_`FoThC}(^1VWH@B(}HO{VQsJzj=WXm;l69+*JA?
zM@dHlW(F@LuYJ5?o%zw-j~W1*WOnMIyPMOcLE$&p1$RFhb$o9&d#dy;jxaD;g0yEI
zf8}rH_CGx1&%*vD?ASHmID6PN1}%pf^y>ScT99jML!BJ-eFi5wmMm=715dwU?vI5w
z1JHquQZHv$P`ySPKFU@k;LqlBrkQ^cpyZ8X>XOmkujrgIu_O1|-duRX0mG6$`{qaH
zXMcRw&M26xxQd)3A-A~Z{%cHG*Z%r1!=B+}nk2&t^KDqOq18P6;ydQCm;a{nhn{u6
zzD#GN$6RywUjns82ubG{TYQ{v|HEIcR*$c!g8a+|7@<O|SFJK_az+)KE!b+SVKcLn
z)zK?@_}-#L=HA~wIINSzu?$!kxHyAxK^QIXb$9y^Ye%%*Z|P=nZM|T=K_|d<=-lyi
zbaD86^vNf3u=9>DH=Vr$ip{#Oot=h|;o9}<EoIjo<6R~T<HIq;R>g@-$N>iKA1_&I
zo%rN=rNU$R9igx>&{d%mM;fU0%mQQ3>3_zFGyGG50t{q=%*HszF<Y59slWmwKUWwI
zg%98vCf_DG%W2cScHL!I`i;5=f2TOdOIEJ1;bVfz0vKXI=6@%A^zkR=emQ~(A4JC%
zR>oY>w3(UYPm6>3&fngn!AqipnewP3Aw>u$;5DA<SKfHby!zH#qWQO**^N`w)|H0X
zvlxP#gaz|`j;!{^gOB{h-d}(&rzjpW@v^x${PjZP1}BH#@$Icgrt2_z?Vu0}x<n<v
ze#ymlMyLPyC!c*@fK2*ugE($>++x1D;B#jVt2cv^U&~>^vs7PKYfjJrX}$*PnEx4A
zr$N;%cmBd`5G5U9L1(lH&UAm9xc$kN9g1`O#=Gw-hW5*b`wasCaG|V{9T&9StJA?C
zh(SBr9)9yXTe)5ggg#xWXU+$V5FYHLV~+M`VF3jOa>>9(?iH9CFvy&75AvJ??70CT
z&bc#BK4I9v?}J5)%{>o1X!{xNc(?RApxri>yx0En)YCd=%|z5YVej2{H3#jxj~zS`
z!`=U0AiNxL%iE=e;X-M-{CFNt)HCR3nlPj9QI)}S!#kIEV59C&+J1Mv7kGz^MLcI$
z?@X5IoR~p*1UV^#(s{SFsa>lDGWkPkxq;?H14g{xaGapLQ-bcTvHga3*k{E14F>?+
zZD)Pc<j5_CZ#eG;KclVC-*E3_0Pj7`dNF*%3GK&qE9P&wx7~2keL)j8U}g9%**84s
zodcksuWvYgSg+;l8}34B&EITO153vB3AnH6OTqAV-#^Od8*;oR%y?P2=|Vwy>+M~q
ztm72UbB|qyCf_)a)|)QbIukzj+i7Pz$tWnzpa1f>dHIbu6LwMNA^Y#ES?bZ~1m2x6
z!RaRi&4l-+12ASnmjC{sZvt}$_K4qtLUxiS9XpydOgy5$=64=5#`)zv_ZcQyz=}BZ
zfc@-5rl5Bc{ly#axYMjyxiSun)Ya8Xyu8<j(h7?$(cPwrs`7OMI-Gj!k>==whx-oR
z{b-T7?&hD1p%DcRD9Z}2t6k{>4`}<IVvc{|qVuS_H5}tN{qz@_H2ewKTkViu0UeBB
zE<gYCVkB%6O{l*QJo>o#lR{C2>Oxjuq+Anm(GvlfAaoX>2nI;v#M$|c5IbSa2&o0b
zq(Z;|VYfK8O6!}=nrpsx*##aKGDoQ&_JN|*+<lMf75D-Bq!PA>m=c%``K2-AfW~Ce
z&|SS&%cNf+rY=0|G_&`vJKJ>LKWj8T{O{{;Rm9UyyP42l2kyJOIs3#D;>iG)el7sx
zt@l2(<-%;C?o%3S%;!%yQMz=|33;^D1G<MFd+r50A(;ui5tUshLG-U*azR4KEtmOA
zZ@sNafxok1u@IoI30Imb<=rf;l}pY!!|dT3V|NPF?9ljNy2i18dg^KOf}WR7X~Wh@
zE9QdFoo2R*jV)aMy4(_74`!3WjJi;OssBFZSo*H?VZHAXmnY-XV~;p2{_b(xd5(A=
zEYTQ}P~_YTQyQBLOc8W<=p%o)c(KND54b~;F1@V9q=`H`QuRAi?ek=j^{RgYTP7J~
zdS2Pq*_-cZ03(~}9{>2YSM4b#h_Am9Sv`3E-qAtaco$zke9!Hy<9Lq5tXcQ63f>R{
z<Znw>n>V%eI`sG2e!ls;&m9@R_Bd;A{N0oB--xs`Gkg=pR6l5!g$a4OM8xaEX}(Ir
zIn4HdGC<=3r~Ww#|KuNNk_TW60?`ON<2;>ZR9juwwUglP?o!;fxV5-z(BkgyB*hCw
zid%8_V#R|Lx8hDI#kIH;cym8veBbXg&aac~J=a`wUF)B39_V(vSwU5FWJdVjuTjd8
z(2YM0XpQ?D<pM!W9WuP1xGGji*}Bu9aECY((II|!OMlf@UX75q8iUw<acg;%k+G$d
z+fO%y^iaZ=iPLP87txBn<s?SIHjz)Q>G<=qLXNA1B(V6mur{+>r#t3#Z@Jkk4208c
z!6wJNE7jxQ;c$s&>xH>L5%=YG5bx#rgQO)ehZU;@%@E`X_}KmMlkciuERNvzFvwYk
zh<hYG2?mFp$h&md61=Vvl`F=?xA({m#f6;?=?6{4KnA=aWS2x6)y>gM=4ayNPsHgu
zW#*B0{^z3c^!++N&rlStgLQsvVe&~=Fq=K+ETBR|{%$g}hksddv9I_yZd0Dw4%8I`
zGh_ZT$MK++dO6)8K3Q|VZI1ZoKV%kqPOx@6-St=N-+;Wt>(YDXOr$IP9XTe-+~31<
z9bv9Pzv?~gW}Umm%rOIeeEdVd5*HuIbxVC0w!;{df#f|cuc(}}Ngi}(OVr@(gTc5w
zAbwzY-mkV8(JK{_kdrO(vfg6>K<@F}mkR?_J92Eup{M2fREMbXlzS?-Vb40pEdG8k
zzW{zr)5!$QY!vo9m)IHY{*6Gme5A03A7=PxgRS5-bi&^;cC_2Vxr~F)cVaAx!;k_M
zs@|mB?9}~`j6y3JJI;}q1&pWg%(2RNyMyc;x0a?oMU(>zE37NaOHv<{c=M$2@U)d(
z@B0L?OLvW`12)U}^OsWu3~Xs4<X|A6YF3p?mLK=ym|HVJ(Kx28ee~)oIs2b|zw{;=
z1+M)B!1kElv0#czhm_b8#V~8QO|LtnSS1NIP9}QS0NJ~p0@GhPwL_9;I4`}od&UC*
zedG$vgIO}4p@8UZ++u{YB{B<|@b}2nZ#zW{Y%lp)ZjO0i7yrzE(l;r>R*s9DQGAw1
zxe2Q0g{|$NisI*MO1e`tp%c3@LBLyF`9f0!X67R7hL1X(DmS(D6}kt>l0Y)LtCbWD
z-I{%*vy1yL%bj^ep&CC!`aBZFA5wqsObLn|Z(fk^Uysh(>^=X8n{cYHKJ`g7Unh7g
zkCv7a_qXQNC>-x=t#iD)E5JdpL-=kzpd2wB;f^ajceh$U=T3L)=-WXF?pZxMx40Dc
z2`|Al6A(^vib0Y;5_n4XL!oQ@{u)DeD!dBZuSY=LOl^wZUvx6zCnYr7Lpnu*v^q!j
zQ=@|WM-@fPjH^QK#4m<R0;$F2_H#q4R)Xi$x}sNZsBa8ChAwJ;FaOH=c;gN4C1Vjp
zpuF3Cer!Bj>g9YH<dZOo>YBsG?~Jh<eeFs?HYWx|-Km4&q(N-PX{n=jtibY)u3=B5
zWh&c11GKu4p^v$#hY9(Zh0U`e15&e3?v;=wcoXJd@IcNDmWB<y3anZrTOGD0wxu&r
zWVWbAy3z?9_z-%mHdlwnvgzhOBNpxF0k0gvsWOFa_fc18C?(T(!oeFI3&i%!Mu3h$
z0N2#LYtK_(boPDI;H~hz?|H+itenJ#aK!m7_xZDF*XwbDTFW0XM1m~0qTVpM{NI8%
zcg93V4&B}-<{tN-V+D9@M^Kn6X$ssA4_~{V+HfYv1dog>UPo?9X}*8%o|4-k$@uyv
zMPT#!7lK&UM@fG^298tgHEfY<r8d=qf$8%Xmw)5|iKV98@0(4Zf%O|Vom<szxuN3z
zgEA;alY-}sH3JE5E-pH~RWhm{9Vy7ao2JBKW6UsGd|Oz{UcNjMJJ0fg?{=agMv~8F
zc<w;yFY{50W1agRD*dc76#jbeXTUaH6+s_%yY@7J!QE38+u7JtGNd3BpEg8!<eme(
z8NY}NIi~5g_%XE6caLXA==CQ&k(m9RUiqWe+@xgCX*zD*J;IGsUGELHgR!?E@Xj;M
z#c+9yab4F=1{;PaYh(y`R}kGH>5I?1_PEP^kRet7txletsT;&ksI1ac#RjFiPI=vq
z%g9;b7r$$}d|r{b%gLA*UVPuA$5~&6o1|?+4^l>cYF2w7cvL|kd7vr7NZm3EX+pB?
zgoL(>EfG~OWiM?84s#f{$)B9zOzNs7y*!O`<IxHtHiNu-E~nP^iZwFEP-gAQlv1g%
zzO?h(bvRFS(u7*Ol3eJSQQO3w6uew9dF?4zgc$SN4PEfkqFhe&)CWDM?2b>wxpgtf
z`=qV7be>M^T?sY%_)gcd|2ZTvTePmtm=NyquRIN(H8`1B{`fRH{b!$a6GgS6n9@${
zNF<ZrDSgGF(<Lf)xKltp?mVnnac6RYe7ZLc#<e36kSXiP>x+HTi{7=}Exnayx0#1<
zF3WV#o$_16XIEpsGy_e<X#XuaEGC2Z3n0E3Qyv*-Q&72{{2Vr-@v{A|nHKsX971+~
zpz+onXggXRM)8R1N?!&dN6R9|KH(0#BY}?oyt}U<rOp{YJf^OKMXV}T)JjRcwzj`k
zE?w7A2Hrm(0wGgR6QNL5P)4gU_kMMlzn{qgxl3FVA{=Iw^>0BmtoQG?0J8!V{$Ixt
z^smEucHrAjQdGSb3Bk=AUV(Eq%qpnfcY`<qnmwbfygO}%syV_WocuVIA^L2sre*T}
ze3|dw_J(63Pv6{$TWr31m$v^Y3Fq))X4!bkuMv1Xw+o|e#kkm`;d&dIIrCj=yIUiM
z>;8|^-$^l-_Sp9O5(KP|n=-AgRynmKr@;$k(N4Z&=`y336BXwwRyK^cWDd@`1ar$H
z0{O!q0=72UD`y8J+kel?xh(U9>_n5ne!1X?TS4>q6PM@UK_}ON*#_v0@oVs02--9v
zNA*?W+B8i1a#|m2VzBKmIj<JLG}lEctgiijFnk{9fC(WqsXF*#cg#Ht_#Fu1NJxVx
zX)2FEo>DW5bZnl45Y@C+&8K_C(x%?mOA?o1R=tjB>ahN<>s5-sjF_Ohm;PN3ci^xP
zwyPZ`ab{LV<u>Na%V%LECcA>cUl!@o;oj%|$neG8m)S(`{17G!R6Q6A^S15I0_J|N
zGsZ*a4gjgr69F1j$25o@l;`(LTR~gmm&*jLDSo9>`9QTK-XNE(2<#NC!|(*O)hL8!
zo$zFh49+FkAr?a1fMG`#qfWVMW6=>NOZaM9Yw(}g@*1x+cg$gGkIZpV3pCOTt4P%1
z@Dsr)s(l$gp1#sIn}Qzoc^>D#)|okEHM_r454PI5=ctB<!yPi*N3T0zHuB`B7IAJg
zp3|{AE;|>#B@+*Q0ghC^pei5Rhu?@AHkkJsS|ftCUdVY2n5eSCxY|$|(eNAE9rC{U
z%a6bYLhKFrqJ(_014*p0;ZDvh^(f)bVb9J$MDg;<co}R2ZoAyOc68%yFz<Bt^(Psc
zG7Wx$tkdNR;G6TLp{QSO-^YhJ+h6P2A6ljJXl=WwYTZLqi4bQy#KC-;2<8FW9JH%T
zZkuEVfN$&7Z`Z$sFbj|M1)PV6>4Q7<ZjJCa*$;x$L>i_6&Rz0iz;|A17P*KU+?&_p
zGzYIdYfEAyH)Uj#KEs2kY{gy$2E4h&*jcIq^7hAzpYzJh`YTYikl`iU*k4yCEqgrM
z8A}4eRDH>wZK?Jv*QUTp<vG@i)}Ep<QElcO;oLhy-WQMg_#5H%AR3ffjWW0I)|p5p
z$YR~$t3xhw!aM8y8eUTTMuENCqnjeZ0~(@xEKQ|7K{DrmRC~8k-xu?zox9zK<2<j-
zv!@ad`R!QqdY~zh=^D5dUr>c;gz!Cvjtbm$9d;kDCw}PoBYR#3^C3Z>?<KfEg%7Qt
zEWv|T!RkeSv$F42k*S)lZr4_Nvi#taSC5u5@<I4LPIU_Ql;)>H`7h$dXiRS_4~=zP
zXApTF?^WSp{0V`U^Ec>HBm!ia9U%rgX$vYDn31zzUVDu&;}?mX5J9?V_A~xQ4<Wl_
z;}hzh{bI#EuaDhMYSp%v1Nf|S;BwVxMd?WL$RXJR@gk+taP-EOtm|1WNev^2(hwGx
zMtV1SFM;bZFjB##WtE<sA@#>six!Fpy$C198ZP-582Y=FC!!;^c@{eFhtfbJG0q)J
z%{xXPZG@E*Im^Bz>Mf#^FLMb}MtBIfi6oN`!8tre<cTTF4g;W=4x&x;D0EDGFnS1&
z=Dg%n9Gcgy(!fCL5B}@gt{SD_w&&l%jw)8fu1%#P3eG!*+2EA36-&#dChl(Am+J|B
zOAlX`c~dZ5k7$X`l_AF0`%g-bSHphmZy?1&JbyC84uYiGCppZ%tvh$_274ugc|PC#
zluTMXPmfRZr+(NA*pm7A^vF^D;rDErF@8@rnzI=L4U;2p`!J0REN80rYAQfkm0UW#
z@u0UOP&LX1jdb)c8$o)75%-zy?MDF`_u!z<4SErdmbS~Lvji88II4K2C#iBD^O`rs
z*F_VwNwP;>uDF12d{Y^MnH{v9yGJ>?0sX|@fPd07K&`qvDN`G&!DC~58Y4H!O&{BP
zf>4alQ4W?Q2P=yTbq4X_LRW`lH@@i-0gp86wQv4lHGtU7vy<GHxd2!((*xoFQ#(LN
zKeQ7QqkDJkb;rxia47OA&lL^B!U6*E|F-j;DOaVMa{u%ZN6g~zXzw2L)q42N8Z|L`
zSp2fsnk(-TZcpqnlnX7sXxJ=NiaG%!`6^s@iPpHK@O{C=1=&tX<O5L5-q<vS3?yo<
z0^<XC8#cNY+vs;O0g6refnCV_v4L>slJYodn694}Le-Eg)T+Rmyf*64ThDx*Bb-?3
zB(&^x;X3Y1B|p7ssbynD6Qcy|WXaQWA|BZea&-cHv=3TC!%w<D`%RP{&}_6_%L5xJ
zzE(vxY!b!65x*+y&M5Rg;|cpJyBE7c-++g#AgZ9P`4vzog-?4FNrT4R2H%nAfP7mP
zD6Gn>gA@oXy2Jz;UFHTiSf>U2-O#A{s-)Qyxt`9+^ZVdSzT=*o8$_-)$;Zxj=pSbr
zW^U<rJ<7-ML&}QtQA5bphSS7BwuoGlxA$|wk3ca!OmZA_Tr;m>u|oB`oIX0$XzL*e
z6v)Nv>+y&GDThqsb~BkF>B3+hXA($xlSR6Xo4FyR#=lf<50gQ`R$$>D*RMBplJnuz
z`ugH0zKc*ZG26b=j3nKMreO~nD=$_uYashq*k;ll^Q10fpMmHQ^4)tNJ+PpI(6&Nh
zU-h5em}#k@7pC;Ty)%VZ-*dOsw3YjbGph^AI&4E;=}ZJ~)eTImsg$d2u>)bt2g#fD
zD-S?_Ck^wntJo<maOS}`oy?=0%zq#rLI+5?>ieBYU_0t1e#cPQ@7}7zeF>!+#2hyu
z`ZOLXP*B`Geb$Mx7}tk~ZTWh0waC2@_`PwH^81tPA{E1J$MbE8Sky5&BleNhRF4yq
ziEG=WEWe!PpJuEj?{><A=5CDQkKtzx2zPCFsZuz$<r25?%K?~_J|5<j9&UW&ad!On
zwwY2mujjUrZr``d{SeJG!i?@H`s{4c`C>GXae0d4?FOApwjzxkv_GC5!>!A?C>R1M
zU8;UaNbx4!)9Q-#rq!2P{<*&%P^B)FkMXzLD;t4*!PPFa;4O2gDe_Xpqilia=r<eq
zlq+$Qjf3{~K>B8k1OD+a*~6FpTMUxb2CtTmgttFp3MV+m+BSoxr{)>DvEmAy$gb5+
z7vd+DwnDag#!WUi-z?!XQ=Ur?3Ynr(YznUnM}Yo301BcCwd40fm|G6Wg#-FwHriX(
z>SMaYt!%0KYPHKhABDA`$=+M;XSMJQu)v{k?z4u(ou&Skdst55t}rEhl?CVYZ*cH{
zpYDC4<4yaCfOXiweb}+-uvh{Q^I^~Si(Ggp$q_|Q>F;sj!=Uk?=nLkR&;g5ZPu0TS
z?Q5FbMPYG@tIz?RZm&QFQ`@CHUm>xPN>k^pi*u8pTFO6i1jMJA(gX;^T$yfU?Y?{8
z>(JzBh@<dB65t_;ncFANL*c$7aYlr6cg+tFcx`lYgn>4At1BPoo(P#y7)@}{dX0ba
z<70?F1kQernRZ{g`gEJMwS6q{H5SA5Bgz_CB99g;#m@c4;J0#`_XirAR!*g*sK*5Q
zLQCPIcw7?{v#54r&pT$_I@D=$0Xs)}Wirn`VB%vjLVQ2R+LiChANNIlr9(12Z%l7B
z^Z#99OeYweuVj4S@9p|PC%2HntoB~<#H3p#Pmi@oJ~8-%%~3dvk~M{EC;_DJyJ;3G
zsYab$^-_IBz}Fbz0Y_t}XJsHU)f#o0Z8yrIrxE46{Sh!C%s;b1fF!1iy<Ju{2sua=
zidCP7Z^ZG&JX&X&`(~=!hW_voIq<P)yA{#*ZH};8x`=hXL^18BM>!{9p`2dvhZ^6E
zAN@hw6h)Lk`b!bGBWlb@{%ySCLreRRaHnAyQsLQQ&Rn8R?kWXFug#jA&wzf7vD4Sk
zvt56)cmrQTV~63>d6fCg1drGpZ<J!cyrZk_B8}Z>6KBb^dI84mJ1(fCkIvEciBhx|
zyc$%U`eu&nLN^#=%R^+@UQmzkCYGi~+Q6*q`<V%R-ae2({A)Y7g65_b^5yIP6ZMyQ
zpq19=zvP0{Qs(CpB~RZ<vxGR0=711$PilfWa}$)6m!NZptY0!uQ7$?m`S-_7N`2Cs
zug092Zo_}!L21GqBRbOdd|ibejP>J)x+vgRPFk@VE&_A8j@Zu(67OZ-++hrot9?#J
z8z&`m*hyUY>z(BF(15a#(M8%mycsK_1ZvxD!iFNL&VFuhcDhDa|A$O58t}WO{rXS1
z^LOt3^^c8z7mIdn536nlmMKb@1;_)XB2F*KX;EY^4*SjD=W{;F{=|+JPHZ%rCMNZ`
z*)>UzEjqKj={$`zac%uUEHtP~H-4QVxJq+tvWf5_dYJ=t3)#_Kqnp0vnlPaHcqB-?
zc0d1lulVieY8=^_9yyE5(>V$4xov0SpH;>nE<T5x7+J4K%fgB-pI=}|b-19tZxdQp
z1n}bBcNW<Y-VrbPM^QBG>8)TLs8^GKH6<TqE`Xtu5-3+mu+)`sJ}8m(>KUJ$bd+(-
zv8}2yUNZR{@Bqbo4R!l!<{K-O_v*)nk}FmFbjmsILcu+5D*`Ty(L***uH2oKw~vot
ze#`Tv?s>Mwxa0Zoa@r?M^8CZ)?;XMG^zj<~hZ;HCHwmt6*}$2f+@B<8wHZwFJm>XR
zX2O3gf1o*hcq1mc(6qfe2&`jh&v+}85C2!2-dPn(e;rjHcWs-FUu2rQQC*>Z|BXNA
zDJG~v8A52$N=_7-`Xs2UW5=A~d~nbNN6DVVok`ABqg4y)91CW7ayvXwIZUTJy$-CM
zYsQqZtaTX>umiEowXesw)!%aV8@|uBh=11?!`CQ@Jh=DA#h8vsIjesV`$!mq{s8={
zJ3=N^5!&Rlyf$Na;2)tE?P|<blVk>u#uLlzBe;C2zzzIBNgAs+v)3m&49L`F2?@ib
zt<;tX5>K%N3USP6v#a4v+LF(dj615Ow;oq%)Gn8AP#xh6PeopX(;&zbM;QmXa<dn|
ztY&|f-2&Wa8V^3{9!PDGEi<)lvWYQb;jN<osxl~e_t{&wLwCB&;ik;A|8C&`TA}vj
zWAaW~#is?#9=OS#=2j2EiH}`huxxu*64Eb5>7_ztduO08?{|GjdY$5gn0tNnc5tnc
zE6*SxGEi4^#<81ao|LX_q*G_Y_}aLt1*@UF%yDL3KN^8KZg~Rzf;s1@3$Q?zM*rpW
zoqTBn2mh|A7D)b^bF=oPHb$zE`DHT<Z??Ss;vV5dA+1S1|Br&7ge5dtw-&&^|AAY|
z@Yl%938R4R%XT1IK5OD%#}yF6OJ)*;L8fS+c!Qc>Fz36HC?-RTE}howZk|$Rs&@dP
zV^|f`t?)>bjk%2&!i!q%)r>jZRd%;3gO~7_oZXv}`*|n#+w*kVxZBYB%HHL4h!FKB
zXdi1KPmJl;u}qiX#Y*Pb$cFdgWZ5DSv~t3B9230q{p9N;#kI2&qasSO-U%f=DW?+6
z&MS@tvk{-NvI}$m`_aBJXGDKl`q6)c*y6)=N}f9HzgY$d_2k3)gY-{*341)fP9Vos
zNZ9mS(dxN0L{en4-^LSL9be_f`MZz@Y(y~|(G*)&<x*`({t@f7?`DYtT=yntne!Kl
zyt(4ykbpd;RY@`bSW%>!CLQihScrpSiq*;|y*-p{)NnZ(X+~_5krkRoOzg;pIZ~<F
zbxT9$-B~$G+za$-nYy6HG^7}oi8mIF9MVNEp6!_c8@j{RAuv7B?9y&1GUB!rW~$kH
zb|6b7At@jRzyn-1mm!1V)KNqDNa>H>-U3-4cbDzZJf#C+6vK$X1HimYSj0_C#6Di$
zaFOh>mL?(-It&YR2;rcEGAG}<8cRV^8#2kLM8hEX5H!-a0BW9i2a>69ZvgH>Rk4jz
zMHe-%1whw4Aflufg9Q&=FEfD{s*Rra{v8n$ClEd2rsit$P=_2P-WwfOq&8A7XD%O3
zDy6D>x;UWvOkTba>%daz3r|R__Jo>t0!V=tV+iKf($9VXvO1CNet^XYg_qCCVo%_<
z^foWDymRT$#k9Rf7x{qBc!hnB3;aVBm$AKJYmu%*VzR<8d44FKJX>?6mhzD;40$!c
zSyPt@w)&FBofq(Z#c!B%Cy0dmgVW1tE6Q)0Klce9&0jcq@Vy<!Gr07a%h>G=HrsfO
zZKk>c(u=a4_9zIoZH>Ti_PtlfUiOl|2-gBk_<-^s0IZoE+rdc_4@quTYYyV3^F4eJ
zu!kUrTMFa}h$Eo06=;UxiBWJ@oUMp6&*$(ex6(zJw3F)_(fQwXre*b^E4-_y4ZqzO
zc4@U(t?PZZ3w)KgyZ_;bSVLLSfEP{UEgg(QPPDWU*lCkBC~W)v3+D_$YHbEDe)Co*
zXHM%K_5O!xru?*Q5fNkT=Sz-E?3c@rp>|Ky=wl3XE)s#RRS201S=9N7-cCeg-W`-E
zE6r9+F3rvlj&lMzJUwqo1QD7$b()xpX^oQ*u8MIa72oDT+k+VYQ6JOnvJ+UXmCs>e
zke`%0wRt}ccVG=!Jy<>0PWa`#K>&UW@Z0NDj~wA%XiBb8^Q<)J>D_+48K4=Jp`5xh
zV9Veie<6L9*v-TJeU8v~RH_llHT&+H>XY_*xaz9Isb&Rxq$y`xFP^};;HjmrDjrLt
ztz0J(j=9lE5iU66tB<_+pF@eTz(Mhc-{Zvg4e;T@aB{+ODHsTX2zWP-=*qLA^{foR
z&LjiSKOZgiQzq|kfItvtn*(-qN=TLk6s`F^K>0)=>jp?bpPDKil>~^6$$d_<lm_@;
zqXKBT`{m!)PF%;HYIFU(aW`#|MZsAPG~B@D<wG4<sCV(1r9PZ#j94q0T6o-G#{B`}
zs~Vm^L=T}va01{trW$B4Mp3g#KoHB(ab>_@Fd0>YiuV+RCi#|&SKX1b;wQUH3flt7
zB1$~g%fLMV2Dq*uqWxLy#Yh87lB_jo&_(5$ZKT@a2B5%@<*y75*BPqH<JPxt3vYhg
z<iIG=SnJhGaes#XKNdh5T2P5bfM==V^i3ZN96M7HB3ngpJm1}_8o*j(q?tcL4SwV=
z<4mf}hrS@+EfYr8j?dq>?oyCuajNC)y2%fqrpr3m1in!|ROZxlgcdnNU+;d*)e0q4
zQ<5_auU|zxWqCfFva1UkPo!<WoA@t!?zNwv?<Y09%SQu4W$p>wRgQ((lm#U_$t}bv
zOZcUz&1X;+P3`mX;CNR3F)IJ-X;K9}_ZNwFK0fT7dVy95OC4N-h3MS6hW9(vD^pK>
za=(2B?nISVPuO0)F_9b_2wg=oM<frCBxPuCjTD<YCv=C6R1Sv?p;LBvAE?0|M+VhJ
zZ|$3^Esez(HZQ1p+lQYy?)iv#e)noT_hM)nW@2BJWbAcwUum6L2jOs%K(q?q&Upn}
zywz#3VR9K18+bCaU;5?72V{^r&8XX$J;fRG;&a^zSz{kw^gcu-iYvu9%}|oy4@mqC
zp_(~4Bv0(oW(r8W6R7<9IT*z7@DjD<zgCOym+t$O3-I@kf){Ai{(}`a&leZN7fGGg
z;r%a?I+_d_UC+b*P&`?O_hbwJhLAE5dIblaWoK~p-*VRoBecA*owy7z73QTw`sP^J
zzXSO>ojOfp5Fy(S%)?{ZM7grY5-W>$D`fqAi0WWGZzv|h^8xwL5Hd)@JV7_=*qM~G
zH3K8Hi3ug7Mb_0|ly!@>&|`SB0vU5<*}`$q^#Gs0pI440PE!Ly#={Ya+7?;-!NwB8
z%OlHSUfd;8Csii=8-B;6(7JEFxrmhPLXZ~~k{|E@4#J9JZEI$9(E}pAR&e6M1wWXy
z?^$VhKqH#9!2nA8ojPK*!I=k9bH;2DFoDG|;7iG04EA>)uPj<(5hN_%%7b`h?`spd
zK1%=uQfX01=>$mjlhH{Fv0YpuuHnnPh?a|wc}fc%C!ooyJz{mwOyhh@LZddhQ*E4E
zaO&Ux28)B*m0=^+?M^EDhK^9QRnV6B9k?kj;%6RS+1~5AcQ0HK0hL|FTv}ou9#<5G
zt=0%6kJl{Yg1>o&T!s(zeGLXO<T|Qc1!@_0s_nDH{GLjcmrfZRxN>O^Ex0mG{5i!9
zKc7>+VThoPl}MHfJU9T)!iGXx$WM4iIFQX5SF)%AE@Y4k-`S!1z5KEB%Uo>rZDVzh
zwf`HCFR?MtPAxFR(e(8CuOZ;6`wn4_#L|k*yaac*oZX61JzF;*@meC4?c8!bLq%+F
zRz7033htP4*!z~`^J>St?AGdt4#f!98fz@M9^Q5aE02$^f!>-kWc$sLj)`AyytA*>
zotBq`CI^QN<cIwY(8;F{Hsk$3@wz<l+L?K5cF+M=(`N<(@*Gz0=7f(1JC+;jCg&Y5
z*WcUam7xi&4OZW;q0k5=N?g@og+xMxw`*p`8jAi{@<#k1Kh)V;I&x5CM*RE*N-ZE<
zR9K0tN0W3QCQn-1bA~-m2^6PW*C%^q7%Ml?LdDf?qyK%Rk)c$rsQB^ObVJweiWw)9
zVFSH7^1Yl%>}26JgfZQvC>W>?5C`LHkCzb<L~YUKIdw{>+h?O=AbP9(S<gPw5-!!p
z#B-=c{ntppziKhy#2P^|^k4}=@qVe+-5k16btJT>4<T5l1{V*`Emf-l{mk8Y)FGxI
zW5fdZo9U835dg=+7{YBdh8<GUY_+DoMI>AG7uy(wcaTf~M?c8S9s!WwL4Mc;D2Z8^
zEyp!&sgeW#@{XR8S;>w8q#G@iuDuYOm+(YOR3SY3OOy78;^D`Of5*mys?tFj=1Yq{
zJlK#bJNc+|X4M|+ev4fIb+Mx($9ECIlN)E73Twpcu8G|W^Iwkj>v)bC0uRc$Nwp+P
z!&Uw}Dz!069HIZ45E6SZe**d(!lQm6Ugre$pUT#VULj41IH8~IYlB>SBSoGvL)7vP
zI!I~$JbewGeA$<i`pTIPi=i;_X>Q9&!<VD3?89`hx98C*Q(sWcwO=bA<IjM;=Ni60
zPD&L27qIsvyj1o<_c|#s8`XFY?ZjZ;XEXl9$LQmfu+Fc+>E9P$_s6hmZO@wNuxQSE
ztdC`Ul+uMVPTP6DbnWsywVEP81Ah!p>a+M36S+?1@||!BuiZ!j86QCM@~gHl9dl6=
zgxL~4W$ll%7y_i9u)#`g_j7OdG{OQVq82=x16%o}t7c)|7(WE0_8R0S=UL_M9)f6s
z?L+T|rR)el#&yXLe4--t;K&Qk?|hQ~BerEZS+5C0JtapHL`&lpUspyWRjQZKuDP4>
zuQw>$I`Bas7L`kwlwJE~Q$10ZGX-CEOn^u_c{Oq;EVK|LlJCOjHs$QKZ73&d(fm`f
zXS<e@VhHX#_Db=EsQ(*Q8{DenL}eKFj|^%sI*)42yzls6WqlMc&mT6$v^Q9c9moIP
zJ|j`m&@wdmDbH4O0=}_He~QiHE4wNIW_ZltET=y$1aSye=q)W39+1xF@$0RhQ_f>5
zX^2WDgP!2cDwniuFP>w4MJ+sQ6e!TjMgX&vb7z&RQQj8dd>VM+E0c!-YA_(ll81ig
z0J<sJS?0!7W?#Aaaq~VTAYQ`fk4+#RB(*PZ;pPUAROUnRhd?$SsT$gLxEWg#C!0sf
zGplgG0i?bcSvBq?>S1TZA_l$#5d$~?@<K7E+Ha(hx9`ZIl=CZ%W)()@69O<jfC1b&
zRf;?mmF>9mh+Rzug6nhu?<szO0zpJ0vyUH4(MhGPu3k12jTG1*t5KqJV?WcXZ_0nI
zWj+J&H_rp<eGhGB4QTjdg0{cFlt(}+kjGZdRpS0(7K(O6uoABv#uSWk;lfVJGs{Ox
zChot)LaPmuXZqKzC)%ftp&CP6p*j5PQU{0k>Abo0^w;;4qj{8vIlW`GxZh5<6$p6x
z!?y4gs}ChZ*BM9A)X;f;=z#eBlJL%7{*9i$-zl-o2U`#dTUL#YOQeeYD|kOOtYtVW
z4cCfN^4*~C@c2AVSdU(|N^BvhvMrXWd&wpL0AZ8g6U$+~beWDo;`xxAK-hV@IrQi<
zT6HqYY@d^z$p^6<n?#0k7O|o>TGg>MoyG;F7k(WTgxm?CoVV2@8_I=`NOsp7)EYXi
zB98Ra1972Ld?81vZ&f0?9tDd}C;hhgaim^KWPg;{qIP_H9#8(7GhTlH#RJ-`fCA3e
zwK>B#<tk@M26}YQ@gSSFX!({FlLg)`Lm|H>T)cYcIK3Pg_F3r9X_=+bNLeM#88aGi
z56|*uUy@s-C1n7PU+T|)j7+gT|CBk;H7j1ZVLD2!n-LZ~=>8(PKS&7I4+OIAU3~IJ
z0^1A*RIP*hYA;S*(Ij4WLtcjh^5lfd7*T)D#LgXCH$8hQ-JAukVps~4yv>X&CAUBK
zcCFQzN)Rpp(~*$h0{j6ipr8`v_;4<fdCp~At7zQCu}`%a*9{XGoG90(ZL9Pk)xgeG
zK_vx13=&dm{bRW(Y!Al?QQAf`ppQ=`E+UAgiNK#@tkCC1PRY3Lxj}8d@$)VTk}^l!
zAa$wKbwHZXJlKhh%lVAN-@Hw3AC~IYE{!XVJb_k=@HL6#7mh0au4@)Q9Rv$Vl1F^u
zUPw?pfkEM`nj*s)Cugow4kMVVmrZ$(=JeyPT49x$M;V}ggsAGxLX+O?qWDZFO@>i&
zuwD$h<}oeW9tVgO)sVDQ!*C&q=(`yHdTWfdN;SA(*1TDp#*a6mQ;zLmfYQy9;6Ocq
zA8E(`a&Uy?S4pm2jTuA_W{sEAaBbJ$2H0xF=i%Gt_BurT-x+OU1aU>%K!GI>+%7Lt
zc5Wru?xiBD46CNm@Y5S1Wp;5D-bzW`wbK1teAoCBQQn?Gc&0{kzd_;m6cu{Y`wC9V
zobV1>i^)CnZj-!mN35QHc()pb&%Qv2+(pt1xsE~NPPS!V@z_wb=YkU;2GKOi3zTF>
zQd~^O$@eh3H$uu7>9okHZ54>AoX+MoMWALjA<O+LW8XG)oKmVXsZ!r#k<RgfI+tgl
zc9aybGUdzVYtA^ZHzK<7E*NKbtuIf7A`Oy!4O0+06(uKN`iFxNlGrN9iQZS_PXAe}
zT(^}rTWZ?!`4puyuJx-YY!3r*WS4TdN{am2`R!uC@xcCgd*(e>exchvE`mVCq;|+a
z_z_jA0<Y`U9of*)Dv8zF{qY|Zc_ZWO$wTI^6*E;qZy0yY4(H!>?VAN{h+iM*pr5Wg
z{FngJupyFOzh^9<w(<hRj5N6-4F}bDnSStuB0igv=d=C8-)<f&6^tMiUE_tsLlwR0
z5er+*6U>RyK#~GONN*daf}kv(90zR)I{8G;c<hhrmY@ANrQujWoQfWsqv2ItlRu@g
znF&CAeQ;%+Ew5&3#)7<7#SuD`8h{4CoUc|bEH#PY3FVh5?iL|7Qn7)(vk(Iy0p#+Q
z3GW2})Ky4ax=vwU9I3<$#XSw8_c$yc`~1)0Q?aeeKwJQqkB;Nu7;ehD-{%527Of3Q
zv(1$f79O;&IJBS(ZDzc{u~M!LU%prv)&hLqpI43Tn!2HeJWa|bFptUcF#B-5f_Y5@
z=Nb&~d}RD)*5^@$A|Ldno*zcH+SDUIT!LD<%VgTVqdXqF1PW$=OV`29x>ElC84dD<
zNHH4iD>aq2aU%Q^Ck|L;g+d(rU6K`#{(eq-GUcLL)k|$gwAQi9jLD>&zJyhLwuZJC
z@g><36atD$f)$RZ_WY=aNMr$cj}FaLNX6R#!^XNd<+yq>r%ClLGD}P!0_;+ztttKi
zi>7jt!(4KmSH(?@zCx2W15kyr?Qy{jzVcN`c88ZbTxHrG`Ccu$=cF0TQ$(b6a5GzI
zuDM2s6k&4%^C(d83VUxvKGwA^Mf&Ne<yh6g&l|9tKyz$^%fm+us)}@twhX`_dhBXq
z<<1at{E4~RK}gPUIA-AOK8ZE`oTZi#MtA!^BLG87A&e~LH0VI;)$-}|u{tkXR&X_B
zy1+|gkp>BDX@bAlw&Irc_D;XXVX@}1@<nc^S4V=PPVEvY{<|!ksS?erO+xU&?}2#y
zo0oK9ytXQzl5cHX>B|pY9F+KGzt*)LGpBKRFPP+0b}h-o(p%FS%#dC<#Lz0e2JDc|
zX=)+5@EQ&U>L94$VJM?2!W!MOXd9RTw1^28o#Zpp<_PiOG6__;_ViLu1z$Tq&1m{d
zeU%x%q@73yqlSPqcrqD?ELjmem#r+|Yyk8DL_n#@#|s{I_su(WM3!ie!W)^#b@VJ_
z{Wnq3ln?2JmSimbs<QCS3vZJUlaTDf?3F_ZopC7;egRwM)21S?4oG6m1yxrG%T%0k
zj(#nWLj`9Ax;L9^+}gIg*noPeZW@UZd+@Yfv?el4wib|hi-rHiAf&>e9q^{#qJ8Fr
zo@hSsFd2q<i0@U>DznOJ=?9o#Q5h%pFqf}~V2+c<Haf{@142+qke7i?5uWDpGeZe3
z+EOi3N0kL3v!zxRT}fpcOK7J=C^6q}F<Ji^Bnl`Y6Rsf4RO%Nt9|^ATE%_Np_D+v7
z>8(&ypuQ4*bV;ASy%DAhAP^OwU36TmF*@FoDB61S_Xbb!!{M+0%-e!_(2q3`+>x!E
zVzhTpbJW}w?UV^X-X%({BW?4ko}(0iP>pva?B7z3`{rQU@Jr)3(S|44NDeM3tgdjj
z0tY{}c(P5Zp6QP5Qlo+7JxA@jsB~1%3oU(A;eIM*I4Ng(HGz58hRVxh*?-p3^8?ZE
zJH@TE$ci0nHU_;gANn@AD!by$wUSI|vh#VyY$t6%TzeyZ$|3pDk(xqKbw7tvuxVkH
zkoT>Vr)sTXQ*a(D$xgclogTYDr7JCc6armpSIuv5MYW@YKQ0^k&nwSU6kHfUJ#bgC
z#XJi32?%A3Xz!;i9ssce7%b^AZ^zMSFuL`C-+`I*xX>L_QCUYl|K=3;Jp`82xW7EK
znsyLeqbigDi-R=t1+~xVjWUuE^&p-D9e+mv>AQ_AL~BD@5ZVk1R0&VZ8e^(f0}wm8
zv`6@iFo4*qn&$nU9>fEn0m$P@L?_Fb9bodwR4J#tvv67thF`Xeyw?iS5`gzH*E8W|
zF#tlnA4l=msJcLFV9YnVZbV&aMvuY{fJ6VRacqTnIj?MJ0xJlk)XFMjfz*KrY$^8@
zO`J}E)TRJUhqap&%otN=fWZO{0r9|{GJh+X<O9}Y>Pk{P9{lmLX3}9!SX-@me6m<Z
zukiiWaGp(0bqOzD!g;p#!vF44r8SThQaq1Xhk<3zC}m?Mvg1P+_|Vp8x2dLaf_>?B
zFkmFUs4-JQB>hw{Ys?p>m|@#1IFHQ#{hI^MD<J<Lo@s7zLuA(A!tJ!#-F}9e=_f%K
zxO`e<x0+bJRp)Q&@tt7H9PcDYvG|S{zr3<D%4MS2nUI02F*;S*EYZseL1>L`;K+6m
zO@0y&)qH!f;?8M20aL6B?0ihj*tI5$`{OAA7uWRCCGs$7x1by_zM!Sab&fg|_w8Q>
zOI*w`kAwkAEKI~6b|N-(z#v-sfDo((=C2wZexVIc?ZY!|>!9=ehr>2eW95~)a-j-h
z6rk>}x#&nIdOX8s$?Lj>M@BxOE(4l1?>Q*yR2-0jvbwEv4*&=><Yy(1VP+(i?pFeH
zW~$?V%-%$pXqU~~Wr(7*o13*)MN<IsZuCYQAI*v;7}|J`rajv#)O*(`8m}K8L4Xct
zEON5%FGN%wQ_>7)_nd>#SGjiz0T=B8xPhP02hfqlw?oNnEchgy?}}E^8T=g^TrZ7{
z2?*B&r-2+g5FxP5N^Ct;>DZ6-sI6iI8bee6&NV3jEFPwKUvR3$S2e#Uzr?lU6MTur
zLSPBsQ&b}hY<%<AXDL0!V^p}v9Hyug9KyG0S?b00KW|(cBc6yymZD?IRWqoJ;^O=L
zu*|S+)1;L2#NP7Xq{W{Wb1nP+Z!`XC5C7d8**C1R5F!qzm~5s{8%HGlD~9?7{xsBA
z1*UlBm(@wL&UcsVDK^`!!!IJ5k>uUpRV`x67@r)AJ2cs1{Ozg#^v+B;m&-=0E!EF!
z=y>OWD)q0@9%~`Z#T3)p4Z0S;X;YuV8Xpqy%Z634Y46W5Dfm~nu~}u}3ZcuWJTe^#
zEMaFq2t`{>=gN<5?NunOsv6gJnsA?Ttyx}V9g1N(ci*7m(PsKw<_ZPcn**p}9H?=X
zGaJyas1Qi*AxO$xM0eb9^XEDpa#X*p;S97v^;sxB-G(RTJv38T-Dj}7Q?PLOlm9>h
zE-haAg$M^){0kql^vu@JH#5>F_=oYSaU$AvJo*R+Ua00jygop2zh43(=;rm_*}T0#
zphaLpk(byhyP~B`w{fs9w(-V2Q-(pXaH2M25}LNV0pYq)D)8f~Kz5m{klB$0Drh96
zrbAkbrNtA)qbDJH#|CBQu!x(la0R!MH_Abj)l5@4=QbAXj#@UE)bLhoH6{|-cy;WZ
z24~THh>%zZx$wZyO~teqzqwn8FsKgPe+_Bu4-G{KdnZcdsUDC*vL;4^UM-4ju9acW
z#%TpcQyKf!WoX%8^R`t3MoPT+H^lLd4Je(~L462rE<OM5(u(#F?e`-S$j~h>+*LHc
zXGvAdY5Rcw1F&Y@W`fO4rSqQ)^1mxVq#5?B%5U<dQqXoRK1+l$&tGwd_RUH`6_}JE
z>wt{$tIgww?4>VhL!c)8poQ3KgIpu$lBw(ha%5AAu#RA;+Dq{8)wwE)YjH@ze6ezP
zju{C$ZFrw`nNQJq8D1pSf!@f20DcJW7^PCqd*mn8ri$yP^2K4Q^ifJj23F>|R=N$v
z$JWs?Rv^QuQCYJ@ieF1_+G+`}Hh+Sk{N*;I>VJOQ)0(&VYwI(62*0e&fFSFq!!Gn2
z?CI;=XZZR?sYm;b?9z1><~2n~v+A+dfp_IstgxK;5?5`lqzsm$2}&i&vt`{IUx%T%
zIWINi=&S`?tmuQIGgH?jb)!7W$3pSj=dK5ms}2Nt^LT-I4ASK3hBEzlOp&W4zw2zW
z*-3C)+4~Mk5c(g+Xfhd8HUXg-ti}@z3}RG7+6e;7{kMxhjHIjk`PdM-%<WSKyDvnk
zdA%_tD-umAlk~emGR`AqG0Qk(2nrHy0;<U`nR}>aFX(7(V%4rSO$Deo4gH5a!nzu=
zK0q+-x@#776@z>LrB#a-3vFv&zj^Q5J+xiVP=JU&i>VJJ&tlr}a7ODC0rs=fFi8;x
zq-zG)Rv}UA5lpGv*emSF!9G^g_&}0>GPt8iqd@(Z>3Q@44D{$9kQ5Phije_IkZ7##
z_vC3xX%<z0U-J>Ah83p~r;a2HgC9i?YTT*WC^%Ep7nm8%%Z+I<Zg7sn-11i;4TgV|
zvP_tFUFum+NNNX0w%g{&p_7(A%BkEF`0ihIe;?<TZ{=#xu6cAx<^TPcb53n;)nk})
z7Id`Aw015dMBq=V(#mRWq84U-0>_#bwR1epZ-f6$84o!JxN}8kB(VN>eEqM9L1|>}
z{!_<xd#bZ4C7G-uZw!hDm(xNOmV5qL@AldHEp$i_5KV4ai)HDsaP=r%*wZ!1D^~?+
zTyVWz{yo?lMHhI#U?)nQ7i_)FEmY8`q$Za}EM*@Q5@vUx7nPW8*@7c7$_i!pR|P8E
zK~skbWRuVE$o;EW$~P~Gs4g8W<dHUIg!hO?3YKsZPyAq3>d~_5rW}LAlBJlCo-t#!
z6PL^U>}-u;hR>oyj;pWrH_teipw?;k4quhF-sP!Y-Ta?Uz+vTU4dlY!ReEMq?y31I
zHl6y+yR(q#CQ5y&6*WrtJG&%=Ns*;*A_y5Dl7{p($yWa4ak*644cu&Z?nd~e_|5Jb
zR6N&$B5r=GJWO)wn#vnZbxd*b7yBw61eL%wmA~I~g0CS(>C?~U8FQi5*6cG^k)3K}
zwGg@P+?Q_KZdqo*$t*{srPiTpXw!PO&k+LgJLWt?WF#@mDevk#IuubDNd1OQWS?Qu
z#iJ}*?MOfwU_@BtAPqr1j{*Yq^i>cWWFHFMK1!kqp)S)Ufwpg7V%`l{S#VXq5m5im
z4Zx{p{YV@upgbmsd~|-)FsvV8RpSB$Mnrs*{<H{`0|Zr-{)FBX-4$*@1kSuIa=|jb
zP<#j_rQ8Sy1YH_2T`T2Ygq&fl&EV#ckUuIfVuxOpNab2$h8{{LX9l5%aNo=^#&Bu*
zLht~&Y*2Opdy5*rVg_muSUTD!HzrH&M;U7Kr+$ec(JGe%KDK_5%G1$$x!omtMsCP^
zBD^LAI@lm%rH)xGZ&Vx+bm@V1b!y>M->@W7ZztZ@eJRrm_tpx^2xfSm*R#1r6X@Ij
z)$7-NZo0W8koF~K_41&5r-!VL<9s$h!;j{;t9Okk6cZ91^kyYUfQ4pg5zJ<rcK7x?
zkSpHCcp3H8|5#f8tnC;__?Rr4e6}Tx2DO0GFo>}_Y8;+m^Q$%YyHoM9pCDy$5+lrL
z<GSO)WS4p#Y(A1<S@XWL-Zxq``~L@VfFrcerao(Em5tvo?jc~5*WP^>SufNM3$$y4
zH;J!=-@(jM-=#a7p_C&kjU{P5lwmFVk3%;|J!UA|n)xS_UGLtYQbgmhz~W~UE9)<U
zAJof@Phg7b%xQfxrFWW-@?)i+nW~3S&+DSbDy$eyJ4@H{U&d793C&Pu2HzI1Pxp&x
znqZGUR=d8wGc5}g`qZvg;aAI~`D6Xp&f{f`>nTsu4}6xD3*rM}dq`9|!H08ar^Z5t
zC!~01&4W@A9g2xhX~(BfBJ^58*zz}}!XiZKUo+<!I^@0}X*dh=H)5{^k2}J)W<B8u
z00W63(Q}zTqE366rF_RCiv@r>Hb;-2Z8-f+mHsrGN80+i8=oW*%CJlRf&LdZLy30L
zUV{j@Et~OdN-vZ&n`t+j12tzbaJK~uMe30;4WvYGsXH$e)hEY6P?D%=xQe3!$8K4`
zC>g9+bV(kWKww0F_af9>Kh_CSRV)A|<^|OAq`j|9Yn2>GW6GH)6hX3G&$Il7^rixD
z1i@XK;TlM#OfjbXWEh1=>js|_oQX(}VMk&1Z>uBuaT^Pi?LPh!P8+NrUN;a3HRiD_
zQ?C@v<0V!emXDO2^MP7un{pz@XTq%k3Zw><pFS{}BXo$ID3~`7<V1c{%C9|3+PLKW
z?PZ&S*A}K&_IZ;v8tn3A#g|?3dAV6oIOp3VnQa4`U(%boj?~$-{|dKK;!I~w&BYci
zi5%a_C?~$FR&dh7q1Yd(gMxxSy($@^ymjk@<$|t{W0Z%jM8`te0sQj$S$wmB3D@@~
zYja}!1ExXet;vf{CBvPfNV_Q*ac@}iV|zH!<4NaA4aGnE$kEk&+U;r|B2~o(EV$Ra
zf@7X$wmzu&_SzU{Sfa@P!yX9Uz-aG@zSMj1A7Ne-`gppy*rA2}X`PH=C4z-zQRsj7
zG|6#5+L2%&XG2HnhUx;QwJ?P+yzaYp#G4dZX^ZSGo^)2fagMjiDRJ@wY(Ii{#^)9o
zGM-<KPwlvOzre*QQ-<Nmdbs_kW<Nz5gs67N7o*4oL{dZUjkrU8ib-UVJI+oSvpdho
zu1_G+pJa~_6CeS1x{M-eC=3lsQllV)7#XYmotdKNQ&hwPEe3lHdtou-Ckt7fBh_~a
zftY<a-;{Y32qoh*!9+2AI~9p8(X78=$><{c#qlVOI)faCOYb}ZSYU>E0#p2PTCE;E
z%l_-QtZl>{&H{$$UaPiu(=sEd3j{EL6{18cDkl1FWRl1Pc8<DF>jYk29Fioe#Cx;H
z#EPI89x&()WgJ4fqCxfG%#VJ6rl;i*03V#upRPx6Syej+Vg_jZK`_^5<|+EHix3a+
z#tp3g@JqV|R-ha?*pF+_hRjzaN_Rj_eDKjcQ1Dj@DR;!k??An(U`T3c9ztQSvxt!$
z8QxBrvd<q%iz!%t!;<l^WyEWd8k>M`1y<VNl1s7btW@yYCu%qd$^|*1q*cfbmz<mm
zvs^{W@b2yaNw+pyD((GzlwOPst7(oIBUa$w)S7wFs1d+Ut-T^|ojbCXT4cBCeJ%US
zL7OxfoAAZp6Lzi!2DyokcS}x^W$hyw!-Ubv=@A={;h_t_dWqg4#QG(}SFMN`;;Kqo
zyZG6q>|3KGU2|60#aE@KKc4ZzTcbS1?SiuovgB(_RuchhRve2>NP&E{JMdQj?j?uy
znmmAE(rGC>s&XQii{iF+^3UqMPq1gDAJcBUZMOGmy6da`Vr{CQM=OBNH0MvMcNBsU
zeZHyRbU~INa?{=L(OSmQiu6SX*v>QqE;IJkPfBS>$xaA)ie{3j(CyyLe#~NmSr)_M
zUeuuomh9kPa_SFbM$#aWsa>AvZpay8JQfQE1YmCs34>Ahjbg6>EIcp^`q3-L@jlR9
zAkH9a4mv_BFe(3fri0T_V5r+IHIlIjk-Q)sYe<qRg!{7ET#FM=Kn>suur2z9LDvpY
zN4B(RIq_Zi`H>t%#Zeu(qo*LCU31XUx+V(Nl!ilzV$iG%sf?n3_67Pm=@LXXb5RmM
zc&c*%-@j#spn)mpQHK=aOV^}hnRe%jpQ&bFnq;p<htPkj!1?DO-)JLx>`!1V9~)W!
z+Q3r;mP0*BQp^-KZ*%WW34orXjx-hphyytuz@?`|8PS9J$au2#y=QU}WluRDK(Z{M
zrsJd*cVunIT17ysZ0FQ=1ZK5WD#aQvt>%VQ&I+)U7>+X8!JA7I8L*f(f-}?n+zj3@
zyVJZwUNgJtJ#9C$6L$SNejPgnyN}FPEW7ok%<}4+Pj52#$84Bl*qQZ)*GvC$shp#^
zIS~#A&In#_=}gSzEDY9oiz<0yCwQ>bHpU#Q{F_zwslZ?@BK3Cpv{T2rP(*43(Cp%$
z%_FQ()f@g9N|KnKBDo;)`L@T>E<S$oA@MedW$uOEoR*l{#=~)Jv^Cyjm4&TGoiu$N
zg11Jx^H1YHB^H*O{}NM`e<~(4x6iSr{8`XzbnC-p>&uQ{NR=IfUQT@Lty0IWs@}Qx
zaQbbc8OD89D+6aNyeSzUPpwvFH8J9Y!`e(7pG+-sk5-O45<AeX5=b;xzK~-E1Rf}R
ze^LUbAUe(a_RFUYGa-pSW3vC@gbg|*>yEZ>ax#6gp&Qx8j`^7B52lsLU+9Z*3|?-~
zBi4;YPYw5)pVwRCQBOtr#Y3B9ElsQVK9UsMa~4O@c<v)lHcyOgtc@oNX93c#7Rf|Z
ztaf)b;+eS~y4XVXw;-hTd5zkp0+a%oHb4{*ismzWAM*~eb;P`T9_?*`)J$oEK|~FR
zW2Q2bONm(b_#YfZPZ#qQp#WfRZn03T7r>`jKPM@TCu{MyS4B06tx*9rjR<cU%7hw7
zOSw8%J-TWc)r?kCz!h42MrmBw2v_z;>R&hDRpm0>Sa^~NfIb=AGmimd=x}2hZXkWL
z0aA2a2<>)UlRPX*O4#8&8on_Ik4feWw%pTHFKvn4gty~|-th1xbE)#e(J%{zS&A-a
zQY2=dYadAk4+nI<O&FVVoM&jG2wxrR*06@x6~2D81SDLLnex~c!M!IZa>QLh^8N1M
zALLD=U53TUK3cnm|3}kTu(j2-%_hO6xP;=xU5dLq6evz{FYd+N-6`%a!Ci{GJH?7S
zMT?)j&v%_)kiD<0z1FOmxo67m87q>Q)v0&!Xw9WxEyFNJTNAF+AzEwvKgC@ddu!&r
zXkQhR<q0Lltt5%becjBTu)<0$v6a{ISjl%s|1&4=9)BDDs+ZJd68V=X@TqbYxgT%W
zqqy=b6n<|}A79h&y{X+Xn+nRA#n_=Zpg#p9d<Rkj2musLAR-O<T+mP$F08b;gaRS}
zkwym%5HP}5g<1L6sFKJmf|V>LGE4f7Zv_t9S!C)>F2po2A1d-a_to>jBDc<Hg=b?=
zAzcMrsRXfM46(^Sbrk!kl=!?F%N&s!H8IYMn*q@B6azzLWIht=jgdkI13m>e{vl#b
zX9><Z(I{iEc5-QJm*76CP1K-sQ4i_G%~a|!npWl>*zX@BT!Qz|_)ZA_&)*!uLSd!S
z7f2ra>(?7}KskXKNQ<1ZNREg2kqWU81^^uR7mxs{kt|&U5GNg5hFB8Mz+n`jL&SzS
zKd~cO!1ou)-JPqqWrGS&7N*d^J3RNN%GS6#z!$?zII+emJefXw#wYIfxk_tHW8*PO
zfE2YnSja1(;iaFu05;lm)$&nB2#>hOrQzmhxkd?35U$t|0qs7!GZU%pXFRAkN0(`A
zH;e8LX)tcdP&hwc<%MnSj&vN&p*)B`l3OXD9<4p}&8uSBaw$B*ZDfUeDLP@FVF?;X
z_tP0A=Kk|f=Rt1-DD7a7I5#c32D&xq`>4YDJoPSs*Z-O=PaN6R^OdnzCg1+i<uc#p
zbxmy9y=Z^miL}JJFpm#Y6@EAYB)5knH=cW_Gboa)&E4%nGXFor`WqLxw#TT5{jo@k
zG4sJaf{mR^lL_ibzwD1w{17CssD8OkG4!BuyKN|~z_m+KmCPkAEgfdY>AZ1V`6*e5
zRxp8{ca<Bt6QV=rRKOjs65veC+8deOudLikA3RP00?%ys0=TP6;9@;&(}snfa(zJB
zi1P~+LvYZs003@pu;o!Xxgo5SH0e&8u??ISO>p>$=A%#c21vB%G{3(3Al&<_nVF^K
zz~(ePnvt8wqpN@pg|YpBKW-Z7HiA(d#Fvc%;N6EMfB?3%zlaVL#0osj`h@^M`ea;(
zQIEWH!+h+iMJ7RpLIw}d3JeN*4etMr(H~^1g=!N14xi=!!MI6;%O$+Jz)ka$NTg7c
zebZX7$vIcCtA|?UCUhY+3uQVsqGDvp6`N(6gtT5i8|DP7O>3!&U;-$wYt^;0+tHHy
zE-bB#!;f@au><AJRB0?yJoPJJ#5_7)L%zl0jVC(GAVlxYp);ATN1JPRGC-v^%Vr10
z%fC)Z9{e1;#Z!;1-T(?FK|!Rtxv%@WHZT!tuEdi*o6lw(<49v(^{x&<Yh<g<Ht?yV
z*R%uodDcfgrI8=N+D&ynV1J$PW_lL_(dCjESK1WKPNw(?6$8}E9OE+a?%Hql_fzh?
zgQ*7X56VbjHZBDjz0|<IM|dfXstwD7$>`7mre-f%ZJ)Uf-Q#<j@s>s(8y!PNYNDs-
zN)zXU5EjnG29*ih#EQ)j`@vZj$NeX~&+eY{NBC-{R1WB92vP<4jh(%S?YKNEgahBn
zOtDcz(10+p`6@I_iD-%$O^oPlG=}L%0F*p3CuWu|!5h8+)D4i%6Ylr1dgaEAkX4X^
zKd<pGRMpcTS_I)f4~X&p`5-OOc<^~wHPcd5oIcVTg+{y{^Q+TwB4Qn68<o6S29{M!
zZ`T!wtNf-_y@KCmS#;C`tr|sXFRBdC;$8QIqvu%A_^H9hvcg5S1S=AYG7QH$SFUWr
zc{giGUsr`YhCCEHrD6y-=?~;vnbD{aPw^+21DPsQjll&UVWs1kMe~IBt+37t0;tz6
zp>A7EXnW#^VqAt`_GEmrIOs3=8J7EdD?<~AdhvS^;BN+i%zTV%0(DgRP3^FZ>O+7;
zdNzZ~g^NeS*|ol?+b?{EMKYCwRXT+`r3Zx2QG!s;yCXY?pnEa*Ki}HQr>55PW$|w?
zSpziHlYghuFCTRqBVZ%@2+p-={Vl!PWD#we+{;1XV@|S7st(GO&5Um05M5?n6dV+K
zM|E|d67^{FdHhzR)iR4bz^QbltrhZp+FL(URX@ZkAUPUy(AI9I*Vu9;$tSx$+vnm^
zH~PSx=DA^X8WU<N7e42~bRON;g9bL*e1|6UnMeN{Tpl**_)ojK@t7+9D=nnk;NVew
zQj1rE7s5jK-&$bL8C-tV<W2HbqPBW@^z=LQTJp@_bp9O#pxDs4kFZr)jaibx0UuY_
zup6-nFLAZ#((sWtJefqt(<6Bn_Ur)sk$Hc6ndBE$)r5<{0MIINKb^-q?eB_dVk(Vk
zw~0BZp9n7%LcVe;HgTQd+MvzBm_lNMdcOE0!s$r&^oj3mkm5vrLAo^%0qD>^S^T6(
zU8R}MoYRz1$YK44Yg~$9j@*nw4#AKZh95wtTW(u}XwgqY?<OiPoq`BJ+@t(kOIgn`
zH>p>{Ucd@j0a`Dpa-iNOIs6sei)IUw8E9UUvg!!O{90+Ev(%y+;I74^4l*_z48~PO
zfq-BDzW}12<|xXsWrC(q{N>|BKKf@v0F)e}-(Bo5o6vU#SzqETQ!(JHDpc1fj&f<4
z&SgHI%&k!IQ4<^gT`0@uo7J3k&0sIwbfN9UkIu0>&qRX^x~ApG>(L7Q4ZvhGp>9X-
z-=ySyu~OgqF<H^XOZ4+4zPe>~sPp|qp6l2DYHPk|E&ZR@sp~xS{63s{*Gry3au!rg
zY+K@n)rrGydzRov&(37gSwaXsSSn&4f7nh(XkAVP=zrBEAu1t1`dHFlx{~$<>itpT
z>mBJN*tIsdVSZ4bX&R-<onynUh5J!b$%%JzavW>@)*AnTVTmKNt)V1iim|U@@0maI
z@+Gjwu*A$XgSia#vbhdk>e`iL(|OFa>y|`rORlXE6Ri(IB&j82E)BmydP?c7TTRfM
zk|CH;g}yyyw_2}mKCgJ#nhU4>-!el{9aSYSoC7^(SY=f7@X_P4Bq~5v^}*lNP>qCA
zG1HN=9sWr{(PQB;=Pd8r5YOoK$1m#}T6M);iyVvG8D%9JIN*fjcB{lBBEx|bjS*RU
zaoy26((h$;<n{1R120fPn-vH{RA!W$bs4<qlm~d;3v0oPtfT+OCW6R-ltGbt!65-4
zW(+T&B11v*PVG83PNHQ@<zy3se>Tddl}J4?WCJXsu(_pE>?7By*T{OnIN78XMpXLX
zHgSU7C}+^H$Y6d0!$nBpcJ7b^DAaYN#bJAgz898#(;~~B1Jvcg4<n8A@@P(>_x7`!
zHdc#GU=IA|UABvuFAP&<hGO9TUMJD4Y#)E>#vT8YLApB?q2pxyN0_yho4JS(oQv^e
z-{55TQaJhr5&+R04Y>>c@?dJQ^=uKu{JcW9wm0VZ^GAiSdr8>uy%TjG=`<cBKv#-K
zIV)dh+1F~ij1Lx~MCZwy6`K4BChApTyD^@sbfU9KLfEXUur0X=IHcryWp3e-{*7}N
z41kd=(?{FO)A>0}92KdpOFzvwlZ-!8<r4qS6HP?fY0THnCUi9g|E7sr{OICcVte$Z
z)Gnx8ZFushA6Mgg&M14PNyp+Snit=yzBIx#@K*h(Y~t2(A=<&Bs1mzcbAz*GlzEjz
zRoZrPZjqX!X>?(jk77!+PoZe@p6j!gu0^MD%A40%Hq7xV_tKwJ+wW1KpCX5|XO28x
z5~dFs;ub&7SGAVrTYd&B!u^Y}qC!#TIn*C~7HL>a1g5~1Np*&MhgWwzC<x38RR7p2
zq!@Lo3P?y;@Sz-5YJN~=evpS!*X9HiDvDjuc>KC0w~_sv>xOUtf#SA>e3yxV??694
zSL1BG>jv378yx4%q6+ux2BvcYd!}ANnXLsN#W(k1C>4`%X*GYM*rA0LJXcQ^LiBj#
z#5&bvRO`~mm6bfykq*aeY}AOThig~X)J~KK#1p}Qn%!m?XuN(Ks*GTm3<E^Q`pN6C
zj3==CQTI8?QpE^9e!^3Tmp1Uw2%qF7gY<_){rM*vGYga%Ln@{6A0X*>6CrYu)(TU>
zWZ+()#0wS^)?5f?^eUFl3*o#EGB(sffQZ0$>x_r@5{-Y{>0H3$H1?B#l~Ox;ft-8~
zDZ}!gt#w9Rj3Oww|7uQ6$vOmL0f=TH_y=dy$eF=38md4f@6vkwIbNhha$eftSuFr|
zXYIYEI(0UDx_sxZ{#L4AS%$*Id_P34s?cm;_4xz8FUVaER9g2@QMhP^)qFm~(E7Z8
zy|q0Iq%BY~0`eA-`ytS(42MF*qH)3C(c)=oJ!xaAE%WQ8T|E_6!@YwqTO^(LMb0+1
za?uHU_-2d`IL=|_jBCZW*D@O?CLTS>e68iRigjlj7!m2^SRKx{n)Y3RHAWR)W3Fyp
z9<6tIoxg=rA3alVc7pGgtT~pJUEO{XWJq#g$uH+njCQ!3{j&r9nP0NJQh|bDwifg5
zO@1oFqV(G9l_jP$cTT8#<hwJJ&?6JCz|-1TGcxCvTBR$W!}T=yPsEtHV@lEVW{R)q
z(eZAYG>B3CZ#TlGOV)U+IT#opjltz_UsKeJ>7(}WPe|<+{j69O%4w}TK%xw$W1|Bg
zR_*Hl05V3%1rP=IUWY9e6vKP&+eoidz()cs0UQ!QAn%(AnPhR+8!Kmf(JUr{2<&%M
zODDab)S3xs%CNgy91>?j#pPACE#zH!LapAjxX1YNne{R#Jbz-%v{mrozPk9A=gg6r
zOAg$+yQR0UepYms4?{w+Ji@~OT>bLMrve4R>0x@IvGoSK0kX(o@r?!;!}EQbV#o<E
zkcSks&X*@Ww7>c+kZ|Fo`AP%hX<4y0YG?fzzx3i=_n)ZeHh)02DDeT%9ktFJsAxoq
zIa>Ki?3*y-;M-tMk>-01Tb7q(<H}X)*~Ie&=rVu98wPqx^hW{X%Y2Awi+uo;{#2}z
zG6-dAx?hc;S<@ayv#P$MLk1{UHMC$>r{8Ape625oUN#I=Fj7MxUP8Y#rgu*LjX67W
z?!`>+kh!q|f^BzJc9SHYAmT~l5+hHazN)i?iG#hIzVK4%E-_D}s2^!F;S>Jqvwwe{
z5}J@!HmjdbBzt_eWfV6byCHX}0=mcA9e24qm)ExBAf?>AIGD8-ukXK~H^<#OG4l78
z^BolWY(7qGZ3sV3ICET&-I2(<W<7lTW0~K<F|$}58*SXj9yDYEaT;47m**1S$V@wg
zL~DLsoX?akIqn>$)OGi0GplzosoU7f5xs2m9wsG;x?xQD>4O4@UO)yP&$+V`1TkcO
z3%YnWWfbUMWSY!I3M_bw$Jz@s!xrZ*Lla#OoEp!adM&9T)TnA94}3XHI_bKMeMnuz
zQ_=e5s`}#{?);%7TjsPgELtnXHOaqt*!5<M=I1y0_YS@!lY(f>UgSC>pRF~8*oV(j
zFAoh5?opfgiX^F4TzPWTuA>~25Vo+3YPEtp=$v+m>nL@03sVCxv)<b1Es5+hUfYEQ
zD$Xpdjo4PNH`)o!eujJdZhe<8bC+((_|`d%z=1%>@NLHX>h|iVH{q?{bwv1{r{G3?
zkW`>facj#cKSi!$vQNTsN4yuLy+lf=+HGEKIUYl<QoShpdmYae=T5Tpa{&TP4;(yB
z9XTTtY5*K7o1uY30FaGifR7ASy$t{Z#(~JNiol~v$q2aMiTGzd;NJIW;jxJXx&-Ic
zqj^3i^?g0rtwLL&4@$7y!Zk;-GUV5q#13GC!7`S%2=*-C2lc98R;ZLQ+iGeto&MHr
zXOGDRs#OqLGtC%kQ?5yMF{S6VC(sOCLbS+(Yd`%?StJFvDic`7x$vh(+y#zGhl0C;
z_jvs{q=`n126s))ji6&@CXZzq%|z?nY|6nL_uv5T*q<=*U-j@+>z(GYO@8u_g`6jc
z*`e$-LC5c0K}X8K7~>e1DD^&9@v|I`J7nCY8i+yon69W>wy@Yac+vU(eVu=vY_$Hg
zW~-v=8@=^$O-{~nPR?MVk7zXf=(VNPoD6&?Z7|sx1_pee2oD+JKQUN?e;|Wxif}+6
zVy~zn<~_oA9vP+SE^BaIe1fQ$@c`m@y?N2BrRQZ$jUYp1edPOPcPD49L66U<xAvW#
zgt~dR`(@i<PRoY-+k@Yf$OKj_)9d}gpu+LqI2%^hA<v`h<AmoHi@5!#nr}xQ%hCkb
z)Z``wyi!YX53#~en-=G-1^x-LsO^zLiC8HJ2EQw3@=<@z=gb_{8HQmtw|9>GyWYh3
zqtSssd)U~Reej5tm7a^2nN#c-Ox$1QT=CtIvU+1)X<~k?P=E9JYpPKF(g%JnZL<lT
zR{^C_Z9)@*r=@H*)zZ^)MrrMSFx~EiU^e^p@5-pqxeUBkl><zLdfmA<c2_G?b?X53
zL)$kGu?6p@wOIuN&)-cd7MT<5tXKkJrySF=UQ!3z3i~~qbNBtcBcBqhH<&tpAXr!!
zppGW^!}uq3J88v(i8K&U5$*3PJNJ!sv;4`8EtNe%N+Xp!{s}bz7$>V<oufo$?=u?&
zu;W=|PM=l`tk9J-56x)=K89EwBc_jit}bSwcokZm$OxfR6{|Kx<R4`+US8y)qOBAR
zo1hv%>L|#Y;@>!<a=|&zovx0E*pHJ`)5(Y!5QbB0m-Q3=^g63%s#*S33tlHCq8K8*
zQ*`e_9%^wEO8?-Ym}YOR&;x9+AsGM_MW>7;4|W?F0JqaDv#I1bx&_q!_RCJNiOLiY
z+W%fSW9C|1DXCQdQ-mn(dElgR=T6!VXRl+|2)p-`+YsqdRtmXT5-lkyUT@A?TXv!4
z@_eO}9-@z!O$Ed&vZ7N1Kf&~BrAF=n^gVdLM1T0O#49xNA-eVN)SAUrw`$3o`WW-q
zFFlMa3oR!(-r3^@X$c0G3tLf*iW{<#kyS*L%pR*~?>Ks48b}9hd05SVjOLccY6KWY
z#P}mu((j7Ls&A3!`3s_+x#oR=D>_u^C_CMerT-9`3;*3{;t*eO(DbTc-*g_zsOT;<
z`RLJ1E1bBsaFJ-$jq(mgW;9c~zx}!R615v#_vYgr$444?7x~6)JY9CE@|%~iiRHW3
z?BBNgr25j+TC`m4TzS_SkB@vjLnAUQ+P2kW<NU(o1zrnvJA8S5J!n=nAD&f@Gc~0~
zrwdH9^wN@rYz$IuKd>ea>R43T<#;v8OGOo5PdqsO-Sc9I65_HlDA0Rrv6MZ&r(9*S
z@@v;8slw$%{{bm!p_14CymE1fOVKrvSY)5^Z2R+xNBfs=((P@nfIIuO_eE6K@9dkb
zk_pWiqWFyvC)*t9TG{4^4z7|%eXtpfyWRg*Dv@sLvl!&u>V1;;XqxR09taR!g&sE{
zjy&&jdN2fr^l}&Q7H#r7&gs60z02sw29YB0!wrG>fkZa{=xl?28Q$2(-UMu4Emu~%
z-riXC+aihdV*0@nO8C+~#k+3>l2V!@9Ql6`V|PV}ZA|Hh#)0nC&>p!6+^(%dzD4?q
zbz{n*a8;j7opz%i83&~3Z*c``vv1*Ev6<UnwdG-`a;<i{Kpus_NsG__OrwubMu4n9
z5ugYR*@6oD(=Q=F>_Czsi3sez8bIFW2Q&B+i1>DEtc455x_=6+mX^iP$nWkuTi^4%
z$2Mz!0?cAOSs@x*3Vw9DKJE(y*W-&!<n-l2!FQ3{N*3(d?FVH}%9pf{hQg*=dPPZ!
zIJLWjZcoPiHdsReHC&$*dR!0uF+aQ&6?vdLJ<M(2JV?mI^zdx#s_m_EMJVg^c$_|)
zU!>_jFa0eQ{_Ctk3Z}3!00+qi?82)p`!uk_&)ju#)U(@O`p2`IO$9Q(pQye+$R4Ae
z+lrz~_&M1=#p>*{O(Hq`9`IJZ|Nca09sn@`yDIIV^3I*0`_VGJf2LH>U_0;A#Rp3D
zQDv-1Y%VXiOzbzMl$DgXm@q`5z@;xZsNO5g6^^|&^zBv6!H*1)w-u}F#OZ4(E~b!z
z{~E-Ve+*Mw;}e+i@bEk8jQ^gWIa6GjMtE3Oph^-Zs#9uHhK0Sd0-)F@d~E#o?s2g=
zr4p=u1E8?dXKf&<t2pRval4=Bmq*6`gEAUuC+jA^mh0-BRCbw*T}M_coO{{LhoooG
zgx(`QGO3B+)QKs_p!k=xkmK%Di{@g@fos&+pxm>n6bW2@;;VhfmLBXqC3;X29A?mz
z-?%hc0Hz4bV6h{$?xsC_X+G$rymlK5Z8ffoTltV`MFufF?!@wEn7|`1rDZUtBpN-A
z#U^fs^A8@bxypL=r=|QX9G{fKH(=?`h($OQiJ>+I9PK0WM^&#Ei7@*vlj?e3$;3DF
zhB4e@`sqO?Z^%;GNA9c+?^)+Qy!QhjF(g*(J22Ujv`yv!PL)VL)~xWTTOu9mrbb)&
z8?iB`N-AxvH#2phqw?tn8-6g|3^*TR>(2|I5Cl?mc!e$Dg+Yuc<Jgp^V+noWr1EOv
z5{sPEl=Wb0Nf?uIR!H_}+VbKb-7jr^HihAwDEPo=%%pZQiZ2?d*55IOOq#vqWW!Qh
zp`rk|?pr5++QmWs85hY62Z-hV)%!Ftbw6_uAxdpDkA@UeVj{|)&cl_QkxsS8Mjr8;
zFoZBt-`<%A?fMh5(k{%(_Y?Mh4fS}blP|R!fQ)NsUZNBKw1~IMx$1MC&S3G`_sg*N
zd#%)%NHciJczvl}-{N*A9DW`q^|(Yh3!TT4V~CqQSusz<&S*gPUHIF2xM-DlD6c#W
ziSSreI<4P_8)l8zk~Mlubm{#Ih8;^>`}Z)+OVl50^qY5|y2JPJ!YjS1J~r1N_Pyh@
z<LL0t3nZM(SgS-ek>hcfiHl&WZ(*=uGJYUvr*hs}(&0%^A%A_h?Kr-RQT6?=oc2vt
zI}C+KpPodv!P~jnX4s3<>674OWLMsWhfQ!0gxlO;X*Pg>$ON}1=`6Le2RZxW&jz!A
z*2fc@7|P6M$7Lx3u5#m?NyOdTDf57mdk)5Ux0>@!T%goJ^6s`)o759BC)~Kl7cyY%
ztt}gL25d0?D=9Txo7f0peIq=$z(K{Bd}7y3b5&E-0Zhp%L5rdj-9Qod@Jks0_-rIA
zA*yCL0P|HqBdqg@<g@rQ()z+8=_3O`sy$JVY_wyn4i8i}38HxNp`=C}9e;(>vjH<}
z(SsBK#G$$%q@v16mQR4~e@8R*3V;u2VSBF_GoKEHw*%vm&=ml;$|E&FleY664pX6k
z2BS(uQYqk!@*7&_!f(60SDqedav=gyn*jZ54uEQivKJWNZ!$tVf%inm)^4(VWoHNk
z2Yg5^{v=`o`&H`cmx$4e{1~vl+`)VVLf)!|&xOWy&<fBwOQ3*NmU;bk0qBHyXbwGU
z3WQGA@%kB!+907G(@Cs#i=>FN{m@`xT!%BjD+*972>u%^5|jWs+M|lI=+u7$hxQbS
z;JRuN*`Z6`*wf?D-Vgw3ult-yIo_}4-*172!M}<@QUw|%U&i>BQ-WXW6XVnLt~<|?
ztHnQhm?>D!-Oev~%z)k^^4_0=I8PhQG7jZ`a@Rg--P%}TxuNb((Q&VSXi!bBqOU2W
zH|@R;syw((N=r73Uu3zXHM-9VHdkM9$MVc=5<(jgL57ja3VJy?_d19)CPa`|$>XAp
z+cE=C7&(ovz_b2IocqMWqRdYAt3+R7-0&!wtwwe<eaXjCb(f(qSAeTP@gW&Vp`A-C
z>xQyU+PR$U`0|mU!TNt|i9>o{qf(<-e5GmUsiIPGhDFJ~)bED(8PDM|GM4SnKG*{W
ztexuYt2^cT$rLsDN2lTsM&9$Zy=2mn0Kc*;Im&Bz#JsaJ&-@%bKIBXqc#tcC30W@+
zAr-5{Gvcd68MT-)5LFpqqSVLvD^fWZ!deFNPmG`F9!CUeoH2o9qSnY?nJ5VsW8eT$
z<sE>QT<EhuivebzBXSho9+^cv2zD~B%GUfrch|)dFY5d^0q*kvA8-Z?Bb53#zuh0(
zpX}%$fF+S!Pstyx$ri=s$m&d2PRCvpJ!UIraMk}e4Xbw5MLkT2+a8QjE{s%fJRV#L
zv<6PXN-K47Hh~Gk)VtXIL`0wOSz2g%!bf5Yq&N#IhiN6Z8F4AP%x`W23v>NV@woG&
zC5R>@ixLyYIq84N2lztoF=Ltw$=zacl1zA`e8Ui8oPXROG{80`pR7m=T$uaaCVJ3Z
zNo<dP2f=-lq&p?p5y5m}gpF2X8{r#qnS$C{u}XsYjZOvfk>)#A543xUdt}!sF1*Wi
zO6%?PlbW?5J?max2L*p_m%G#)1zv||G7E?pglHKI*3cCc5Q7{Wxx*){w$t`)v_qyG
zQNhdh3l<B`{xibCMpq>XD%CU{f|k>tc4ISW6^$j-vtW{DI%<3jzv%OLdX{g8u#BnX
z4qix3RkQN%8$}0?(wXRLo2;*GeJ1Yuw|a%a2VWULq+oWWBTew@S((^1t69mzMuPBi
zE!%5bJJVdCp?+mtIF*d{r)f*ktM971@3a%kWP$Xi*L)*Bt2seo<+#>gxKv^t#tY(+
zJD-ojLCc}<S!E51fu7UXbw6>b-Bbt<+uj|~+-~)|s7k`|<>XOisW^s>da9#c^mw-s
zK-6d>(&WQ!;$${5bX13S4Cn~5?Y2TuAL5R8(vhQwsckjc9ar!hp~50VLdSRn@M^m^
z5KjcT*B_9!$)oa^kWOY31{D3980Qr+^!TB$@8C?ClvRq)*cEk9hqrmBpwVDE3{HiS
z=T8^!+#z|J+<?EOEB+!4D(bbv8B(~4T}+Mt#{kkBy^w`ZTM_A>A%-*94xykj7~TF@
zgLc<N7Q=yGHR+u6Hiwe9>sNyzq=oiF!Z_owdI>NLLEjj|Zi_*!^RrtyXI9$u6<(@p
zTMRig{_VD}=lwbfK0rf&U6ta18#zb%2jUW`^X@bO66KJ3itW(gYls4xnXm}y5O9P8
z<Furs3Su%BBfvmuawrJzA8FFEhObL)L0e6a=d%i&yy_8HzTgNO?{u?preIIj&@l)l
z#W522z>4Lh_L>wLzk$7qv!8t9u78LWV=yi2V{hm3H#O(r(J6zv^{<9dmlyODTi_jp
zv37SKuDT^f<-O})uE}h==|aG2)BLN<{OL2jm5n}vWQs-T`w9-hXFfbzgMA*P{DH9|
z=kr)u{$o*6Xof)wX?R69&t)}zrRp@jnT6o{3$GMf)3qI)nav%94N~u_$>L?k_tX%J
z(8YBt<`Jt5%gN-5UaX$7`tp!p@s#e`=3%2YOhl0+g!Ui2(%34P?h$xsXlzeWzdSkr
zzy>>=LS<*Z$7-;#Y`o{OKVs}CKWie$029~2Rd8_S3eFTB9;qU&)NW-_I8Ap(Ta%>R
zC8EfxWu=tuR-&y=CHk|T=|zL)eS4c&;PnwEzR`$fBp{X|_v#DJnk}(<oj;%oW6^rl
zN?5Io8rdIVC(q#^3DG-G2*Aat%x)L=H8&06^tVN(O9r<84#L;9@ABlsi1kP47`q~|
zJ}8J<8leb`@r`<8*{qT_Qwhy|VZe+m02<V?uJ;m}B9?0+xPpcauWTa9h~}S|Fa(1b
zqz@4{Y{_%h+(x(;$&GahRu?yR1GB!Y{${KvxThFKj7~~j`G~h2|4jrNfJj(UMHK*I
z7#aSP&Ojov$ELFh5bcCLrfELZ3O_eFkMv&kQZ%Y>rYfcS$v;MH@{^HOe}UVE%qjW|
z<w{$v3uuUP20o5bc-mPh=r^|47Yg;Q8T@qb&-re?27;LV@5XmUWk#Tefs=3h-LMF?
zu>;K?CP?X{yxy(Ea;xegl491T`uR908YkP-%qkatkA9uns5u=<F>tX>8~4|r&yS4T
z{U2xtMiWJ{VKs>fDrn64@OBIhX1(ytORD+it@Cwn2d~}$2Q6I0le>6_Il86=DjM9j
ze)*amfJcpcIHtSHb_v}L8|oG6kBFA-*L*DCz)0-0m7VB{Pv8>4CGFGwI!rXNoZ?`G
zaAd*+^*Hv^{(^6GfYy>(s!==H=lNR|scgoF2F7h)g7)GROwI?{_3oOxf`dxe#TLe{
zKQ%FY>y%2D^)~Y(e;IqTDoQ2AjV=W|Y?IC3!fd8~LF?JqO@<JM1yRA;LM@j4oS0G(
zSBe1gU?I4#N74N(w!b&-!8l&X7RZF+(4o6NAR?kQT$!P>OSCBB3KD`-Jxt<C@E!!u
zFn;9mA`(vHU-cX98G>?vk<m?hv;nr`2{vEk6HYn_3W(THthKK@!Rb4Zl_NBa7!cDR
z6tEZ0S>R2bhwe|qs(=UU1Tc+m9DzvEP9`m@#JvJX3|Gkq+CRcj02m>nXh8X*&w1j~
zP9q&+Kn!4GJ|J4FIee{Gl{YbZP<<cym&QVp_Yi_f(3KHH(9k9xgBTBabg_=mXAZr%
zup8R;iE`^-?VER{6LwPS*KNO6(Lkf?PPN=nWK+~m)DsfI+u-9q3=&;&QxTb}N$h?|
zdCBhkOLC%h8Nn$b<}UVqlCJbbRe_di-2whhomh@74rweMCYhW<_TkfXIID8-n<@-N
zXL6%!eymFXWUk<heKaknX!NKGK(XE-H?wfObS+tacGExFztcm0vQdbRt9!0?>%BgA
z2VpnaVJqPU9vn15I*r?%Nqn%LJFuFbc*apmk~ojQ_<2kBNz=XRn&U?1L>Up6D0&{+
zk5jspMW`8I7eUl_i24(<KELhax+?7hrN$uQGujtZ`Hl6mRozt);OSP^!9cyBDI!|O
z`nhgISP8s0ZBr-Z)|*H;iPd%CpL;wviF?A_3>wzv116d50SN^#4_{J&;`0-9ugjdf
zZSHr?9xoBnQ5DxBUI(F_6S!u$*|>(m)>k6-O)Dtt#OG(})jL^wF;jK>8MmL4%Y@Zl
z-=Z6uNq8LDZDiUtZ{F%C4g=m-zv4hIqrwDgF;!!NKTm@XDV<~Bc<Ek-dFzb=ppZ;{
z%^>ge=l6{L*CyJ$l;T=GaKht>%v4)<Uu{+~mb<OQ<j$!fY0u`b8Bf<4mMAnUJYTBE
zlKR`U;)}^i;?G(J++W^Cpb6Y58fnfmK#IsM(^xpz+qBd@mj^Mc?6ij0p~(;9c!1<y
zw(iosBc2naQ5Uj9Z|QXant*f%(GAbkUGyJkzpOM<cV|P<50$MapHPN{;O^ntILp4Q
zP<JSjXQJVzj=`u60LUp*Uh;li%c7xqf#LzpV>wy>qD*Msm^}n>Kx6kzLonN@uhBT3
zp87WK>FDPv-30Vs*bPx^FtF+yu7F&GA*PqmiggTFFa=Y{U?3Gx-b8eJEZ+uNAg;Wc
zk2+Uf3Q>az0LvfZr$oX0MHVk^rvT#M1<bS>Wov~ZE0!1i@yC3Kum$P~goa*+->#c$
zp2d>L&JG?p^kVuC2b@XqCNdJ9plnMRVz?e}v8e-00lGZq{$%pj@^alZye$`_d$_B(
zWF`Rwt4_z?rq2*f3tWbFT0-l4nnrn`BkI1Ush&{@LNnT4LaV<{3>Nrn*A4bQS1ML4
z6p!$H(|!NkXDfe002evK#mSmjLUFRN5pFfKlVxRfqEr~$@OSAVReOTK!2Y8rw~x{?
zXg&*Y*yW+PY^-?ci}D{C8@VXWr1-IcRqDkj<*sX_OW?bDd&hffu?c|_4;O-tkBffN
zG2e;i5IBfEl4w1zO8Y7M?w=Eh`|YNfR$$pF+xWkRd(vcLSoDUmUE(W!SWAL4Yn>ho
zOz@~DQJ)RreKcaFduj`GzqIA!G>=yJ^Me}7p}kYO*gTcBz(IG!yQTJ4H|Vu7+x8Fi
z2>y*cpQ1c;!)XI+HewdA)t-)o?}^R~i*nj<W}6j+2G-Wr!Tj3mwaI4D)#7JLBGhAQ
z7l|~7MP`^3C=cwD=0%2M42RnX0K>r}<rm0jsM7&o^wC&^!#;d2Mt+OeYD|0I#1oU%
z2AFpK0wKtebwLVd^!-tc)s<b9P!O#oTIo>WM|P$U6GX62wBvgN@q=qX;a32}Jz(&}
zcs{z5s2%40s<13I!IQT<uN3Ub%3q{2o{z{`vs_Lojnml;fQZtx6L-EumTgH!9Ffc%
zFexQ#no|Y((WfQ&ZPea45yW1KK3kyDnO%7dAq~p6WJ(`esi6dagbwZ9`xO>Yp<3>h
z-Z)51CJzlJ1Va}`Ocyf}<$m(5`TF!1V*=_56@UrH9}K`a>#2ghcz&jszog^sj7@|2
zafE`0M&1NSBtjVs;J)}#Nc3o+qm#D-)h?rIY)ZdA4fh>a-|0E;@7!R|=S4sJ8?4H9
z8{p3hUaTIF#=UzH(Ta7+81CU4_#FSs^}A-#(!TWGLU>_!_OVb`k>{>vZI-=~<gE(o
ztgP!drLr|q*IkJ`E^<5j@)@rui?rKq@|))$%xaDYjk)IsNgsC_6(Z<V4UKDZPSWPZ
zSz?KSdsd%@B_(;qfQ<p}Np7?0EJO7~<K4Hdta*wME@b)SYz>VB&x-@eSin*HZ*&)a
z*6|Ehh+JS`EQUt|-cV~<1aKMS@HLPc$Gz!e{sefewg;NNbH7qtRZ(1V8LLt;wks#_
zgU)wJ)nT8y5*V>~Y~UeBNc}5WPqF?t{TGeTb=U5l<F}lm`TydEhYs3#TiVTAI{X}2
z)H8Q~)Ya6y&u62**)Ct$WpFB{dphfHz4=)69M%;Z+&A$PnoD8}=X~|m=5DOFP{gi_
z1%I*QKRgw|UWZs@7?{tw)3B85FZf;<NikHVYEW|{Ma>QNyq?_gAB^pXy<|H1#MOV<
zwD`7KS?c_O-h#sYyxq%OhYg|K;Obfhp4CY+Zy*GNT0NcFXnNVO3ko}`4&Sp&V|skn
zF0FE9MSaHLr4kP)9d~!kX1xaQa~d37lkVY$395R^Ee5`C6NnP~b+n$Qda~kqT(t6n
zrg_r)CQ#G))l?uX3aQ&ZJrmtoANq29WM5}Wm<)qiX<Si@Q;PiNd=D0qTRoZ@bgy#$
zI0e+bxAwSf-K?w%l>4;^SNZ;aJXss`fxnBX<R5=705_}09EWPO8R)40_=!rzV{k`m
z|NhWwa#5R>f4Wj(mHW9)%-F5nnx;x_hT6!rr0Ij<9#l-z7^*xfEtXd8naYh<V>y3o
zy<w@2hIl$2GC$K=X)T+b4)eC=DeLef7sHtpagrT|r<oKrtu(yxfZ-YNQzt~u;?O2w
zcrh3WMU_1JEMsm*N%J$GqKw%$SDRyf?=jC0(9U3Z7aV6;fB=C0h{rUknf5yE$nS)P
zoOoqc=(_?C5z8UbC|GSUIy|Z2Ci*i=Fqp}$M0sNEkT_*Eu5w<ZN{5#y&N+QOE`-eW
zwVF;c07=Fc#{H?x<sLwoY0|?mRry`(kyi>j9M!QOvNZRcv&k<zG)1G3*iXDKrE;8O
z!1ZiA2Uwe7Ig8FuCk#g9mz`r4T=)q`>KR=5X*lf6L=a|<Aaq2=2ejF^R#EE1%1H}g
zC4op>KV3w87^!n26B)dx=!M-5F#(wT(7gsK7zP_U_r~YHU=ayxBE~IZLd^Q(>1bp%
zZB=9^Z3UesqD|%3X50+!fZFZOiD?wc8XGyI6Rx9~qnwD?`17}ig%qFL9)J%7Zi0MV
zzTYTAm_qD!v)Q5jQL*Zq(~Orf+E3wL&Ahx9ZxdPuKQ4j}3ZJ4F5_(65{=WT{-aG|<
zf-*ObatpQwAID6HU55UqAC@R1Y!DJt+Xq-F`L}sYMtmF>gQsp2Ky2RQ5qvljf8c$f
zu}>)(Zp1_M0!U_m<;fjoo0~P|o`@i%uC&w&gE&gm3H6;!U32va?&|xRdqj(F0OvGt
zoLgyGag%iqw-nB5YV9-MoR*b}l3ZM|&vW!y$Bd&X?|36#jaWcQN8GlgW-t_1&)IH*
z?>=Q^%&uM@OnfBELUA#cL51O}(uSs3(^-@qROjyZwNE#VIW6kzwpiDN2bD=jUKu_E
zdW5cx5xS2tn}s=j59V6oJc|6>K8Fo}L$WBhE304N)j3d<uixsFwmU4iD)Og;{EdE>
zh7RP{9SEhUr>N){hN(5VKby)AfEA?2R7^tiB1FR);zV}XzscnxolNvi$qgTQ>|#Oq
zT1T1a>#sUI#5hSr=EO8~cQ}%ArdYq;xPV0HllZ9c*)Vu(_@UOE$&+4Z^pX%HuEU^;
zc&>tYC;&x%e>+#Zn6|%U0?Ahn#$?d0@nqNNDgc&!zCuRfw?*^Yo3X@UW;Mzm;wAkN
z+kgEamku{`KwPLZN$Jm?F@olze@UFdzKHau#OYoa^J9!2keXYiDlH|6SXl=#8HOoC
zaPJ8sUxEC=l*g>R_hvL;fjy@>pNiVpUg%@=gYsig1e<A=klE>0zUP>pCz?gM7zyBH
z(vsK*u63BSv^P2&%pqx+aJ(xqVm#RJmvg)&4&XT6$s<S+1@7}%3PtIXofH=Wr6Sjt
zWNA$X(T2+_5wgY$cTAV}oinhvA(_kCjylWx|L_+n&cIUHUprTbEX@=akndQvCGYsh
z&n<rCnDI4u47PR^rqOkISJCR*f&aPbOovs)a%-5d)UQ-i4#}*#xgJU`2$Y_^MUKc<
znrsAMTHJhUMzqd65K3MOHM;)N-^{^<9GjFf-v#O&VQkpTVoHQ^(_5ofLZv1F6&JjY
zg*%;FH+7yrYhxRo+P3VRqIo(bCwDg9tbB7ytlR&_Nao6B1$8o9(AqOkB<!U24Me>T
z8|NqnkXy?c`f7T776HphUGNU?XDvqoc5)3pFvP4SF%b}rg0XXFzEpLW_I}dFc>V>X
zW|qMAkVJzs1_b`dZv)VEmf@0+<;`eK$+BoiIXxw%6Y;^s`XyHMV;v)b1I{4=@J68^
z)X?DDM18%GW{@((FxqsI;hVOu?d&Q$o6QPCC^dZa!HKgDj~v#0LSn;o6gf)#FD@aE
z%fBnc(PIoKtWwE)8$cES%IzQz-=2(0?#cwaXc4|40-S@f!P93UrvQM?%AcPK=Rgyw
z`WOJoqLB@M>Z>SPk{)5bH8k<qBpAaSA`0NDdYy3<U`F{P<A$CZn@$2zZvQ9@z(b8^
zVhfo05IU~xc7nJK(lUZH`^5k;#_(2A0i~4!^ud=pS;F7_F{jy^o`gqvg{H#|FV<O2
zir^$I31OwWf9&GkM9AB7IXCoyBQk7wPZZJv5}P@r6DUzQpDkwpheg8_9k_3|g%33(
zCkjNT)2EDvSue7jk~UaBhm2PI@ecPC3O?o-Uw^sJ+AA{E;WGLDg@>1rUWw&n18TIX
zpmakwH0tQ~uD~e+Orf6bZl=fYiXc@{pZhMPrcT<x%njWqz6P_+0Vo=LmpGo%jG)x=
zcY^Diu`<58AHM}lYq)-;I(i;9xh}a$ddq4*XGr*H4ZCKtbhu06`~KWjIvO$I-2+N7
zGc#pab6+BHsHD^IJ|GoJ04K<h_>`Q84$!~>vVn!!SBf>;;iW_sf+4}%dk;skNkzK7
z;_I?@N{P<MO1^y(#B^;uK7c%BfW)`_sl@bCe|P{EL`uLs602iffVM*R8*QI~3I~4a
z3PelJc<Qp*KUaKF@MAdJtmfRU-l(e2b?gL;vWzPL^E$3E<#Z30=sMK&&WUe6ZTxQb
zYoNObDzXTiA;N0{VK5P`tq*7oW6s(r%b;XK{gX#8>D{8i9l%tC6o8Rf_!Y4bE~vlW
zo-%D$jC=ym!*GPfSUlmFdl(T9$xT?zfGOY;k#INv0b={A34=p14VOTEy<xf`*OQb*
zEZw<)dDN6qj9D8Eq++ze#@0{cz1Ty_6etNX!2mlpXA>9>&=I{%oCsXC<fb4PWz%r@
zdU0)?8U<M8uo!W(q}dmdELk<m)QhaJMe~>bHsw~ZORs`>T6kne@n;Od#|X}A9lI>F
zd3sSezZEr{l;tjA3Rm9SU5Jlbd?O*8Es?ptg9pnBF)PN*+y74^up;q+{)bas5%bR+
zq_y<JZn{bjQBK{<<ql}=d#9Di#e<ISyi&1))RyD2eqZw&4bd<PBIodAfGeW<m_&~N
z96>fDhl<)N7u#6O73{8VEo_h|T1sXxM2V}e6@cooer^xO@h6RP8n}!s2mIR!>G=qQ
z3SxoA%)g3<{3#$C_Kp3CiUt6x<zpB}iU}gfY_F8Y|0WI{%n>tE;RCsiyc->p*<gHh
zXsq2jq80In=&(xb7|$Tm{RI>4cGBLqh>&15wN#<_gxLZOMSU>=IK3IoiTU!$aH5?U
znjgbNa95aFH33F8KUf1kz_2+R&H?P=JF~@f#{skWHg&XDe?&~a&HM`k2?6YB5jtQ{
zP<b)Fi@LfG;F15<lJ-X#hCe{D$?*yK2U6}!WF8(-J|oV=m(1@aq7C}pyyuX`e{wi=
z&eT8|KXj1z6CY4G*yM`J61k~Zz6p|@4zJ2X{Ftj_STwR@>7Vdv$$D++7szNw1Ime@
zb$IA?wb|7OIQg+(@~za&lK%$%_uU$AHHur|Y6JD}=(`;l4%a`r*<8!E3#&LOOKG@?
zLwtVrL^1l(evOkIA486I_;7poq@p<_lVI&K*ZgJO04BPqX<@Z8?NDEF;!bk#Tj`!7
zs5Sky!RGvb2Yl{+9d~=0$^SlJUcrU0a{t+U29EW<$V*$Rr?q`nIK`!UsQK!68R<I|
z4kOhZ113POZDc+GMD)^e@wQHOkV}PRUEnxHVrZOxI0zGy(@qWc4^I_y7!BAWe-B+D
zVWs|eE|QQ+bP?<rC|WOb#K1-jP-WyH$LKe)MWn<C7l9W1qJz-^*n5kXD4mYm14x<T
zvEK4`@l*axBjAXiz*cA4;dgIMq@`kf*_Ba3vA6hH5ndixP@6Xet5hsN+(w3pO$r)&
z_&(!JaxfTPbhHK!5v_Zaf_i(f0q{d(H23vL`f#kR%HIH-sgIvbLH+|M<%{@+I{}Pn
zGJmzmL869fG9-Y=Nt#dV#=<s0XzrB@u)fTX1r>iAn@?7l&5sXMEn~c;^2sf6Hpwq!
z*Qy`t{*?drg>Xd(V*0I)EM`p$N0P9P#;EVWm6>H<ru$-#E4KmqDOpI6D8oicfG7+$
zI^14Rrr(mA)-vB#p3fdf_C6x(ClDd;;T|2q=Z(+5%?3D&>;oz77-C-~2Ps_!N1b|Z
zyv?;o^`-o@U0r;ZJj9Y*6|A7pzQ3NR#n6w7CXhtrmovW{HLt8xyYus@^x}tm&f_g+
zKe?61daoi%L(d}(y*@jArCRzYw`CdJ`d9YK5-NtdvRRzrt>Raq57N?CV0L=~<QGf2
zVJM3hq!$Wh(v6lSmGOciJDSh)zP@zuuW(oDoPV0DuC6YxIXm=0=R!Y?yBSJRRQyAt
zzeLIFfcpDKi6d&3-LGj@FZk1<*K8cQxtNZz)8$(01(agH?%vKoUP+c6eJ-Tz{wap{
zABo)v>j<t&UiP~mf2&sgN!@g4Z`XqA*OCF~9l5a2(erF*BM!?_-@I<umQhHDBb{UF
z!H&NJP28P=o=T3Fo3aNiz^qtJ-+R7q)N$J<S4~v5+l&SDw6~XZ)DR&e2~sbXk1^+P
zRy0dAz6d;gUn_#n44Jo-IkV`DbKg6>{hKD)`}(^p)q=TFiX~T9$Ro|G(A8rr@eHq7
z_?Sq3j`>HfY{m4OfURZPeK(1lx7DTGg5do5gWr=`$YjdP^XvKVhoaY-%|_vD#lFpe
zl4`<Ne8DeCLl|zhtz526NKb>u=eM<|7FTNyhk5hdKMc&gu1GLm&euvFQSZ}ACL2%w
z+dwybw}|ZQSGRb!-S=6(B^JKiXWNb=64+eQ!6s?pfv6qwT40%7_Gq_Mu`X*8f1nS@
zk+qD3Lj%R&3?(oEBb2p*cbp87=CM)4MntE%0C8yNdNm(Tuk5u6c;1nCCDUn%9=+>h
zJW|qn#6?iZV5dCqB=ymI@$Q>$g^+Io-$Fi?MQlpL#ZRCp1Xvjp5We8?7zbd+;46y-
z!nPyM%%D5PUh&r2Hj@Q`Jh!Rcq8pfc<#|N`E&xi}RPq9M?Ov3(DgiK41w{bY-r_GX
zrPDd%h>`C4$r6_f-1<<gu5cMlI<OUH#wY-9)JV8TZ-wny*JK!eCG>9}Pk_2uS%V%i
zH^Lt+V%z70z5pbdfM11KluN1g<&zBs_anopP;0?_98l^{4KT-@VzqbDWT?M-a5%dw
zz{67IjOLNzMQDgU8RhBa<$-RqoUW^l+dJ^>Rr@)v!B0c`?a8-oZ%bF7_j1d&=MV~(
zh>)UTouhdUgP*-cqLK;2DSG&(7`AIHmN`brcTXOLgB*{ePQq;5RiB>AcOUq^Se{tc
zQfAj@`58q=Q$Egb3Qyip*#r{af8w$wG&}D(!Oy6AQd7@G;7D35?uDWKKSY(?Hx&l)
z)Na{WTm6jAKw_KlhZ^Mm@Gu!-$E})NvJzMo_@!>-B;{hndKz+QwxN-%x3$u&)RG4b
z>P19pOx4{ysWaSYcp{{M6ue~ccFCMsUo<ryRr9eJqeJb7r$N)J5a*d=e5ezz!#eHZ
za~bsg02E<k=YO!w4^uu{pZ}JZn{_T4a}47xsyJhz<)7}#;IaxxF!Xz220_){j(qi+
zG>tB=(GMHB&vHV#cl3Mz+xO84rRTT+4{t9vf?r>z2aEpVZU|dniaOpb7BrVRKet*J
zPH{O-C-JQ}D{NeRc$j=ty5*Pt>0_4C#8PUte8@VvCREeyr5Vh|zV#JaE0m3GxOcIc
z-{xX8_*-i$(d65d&*Qz39uXqo*Qk<EOl;O+7|&+Y^v%QcPahxwaiX_1J1BfE@#|*j
zCKdz=8>s7^kqIT!X+%&6=}&~H_FM6P0Mq{fF@gs{;HwqPRI2=~8%9($OIk36{9&3+
zI9)of>Leb@7^1R6o4gJGYcn*J=R4vpI4uECFxBSlxvZi80m#6`f)tIx^qp@h5O#$E
z<vKT(qelQJdzJ^JpOKuv_Q#BDh9S~sMKohY^-COXMN6RwHonP3?9K1JHDo-B^G$sX
z@Lv5c+FqZ{su4`x@3x1wy7jluwT?IwCOzI-LaKrN5y0w1wp8O(*UOh*MIXExiq+iT
zoiEf}^aVxDh>%UMs%8X6O{~bJt_uIS*h-p_RgBb1p5V}Z_Ao!w{@<ad)!i3e7R2aQ
zH^w8L<d~q#S;^tOJ<U6Udm9E&WS#|7r>hWMRI%G25?A4>xCy3~JEU*n4>M2knW#Ve
z5B=3vWa&0|N1s%A{eSRhqhM73N24HYRHlLH@??^u{#1=-BFA;E7Flo1;XWH4YE*0f
zrYRDlyk97`PBmKQ<fm;Yk-}%5LOcYHNK|>ikE8ACZ`Sl6<zsRiGoc^~{?@_D9e6eY
z*>(8XLiInJCIQz^$|L2_?wIIjCJb(aYU4K#zsvpC>mL3`&1tDS40qdsRZ6+dqUPaa
zK|10L--;H%;dT(O)RWT*Ls5!<)ZkL!DiOVVXx5*9|8$LU<DCQ9)kZ37#B3i?Tt{|+
z7-Aa~wq#!hQ=Ni)xwMk7-)q3M5CVt2-7*6_-C*|M)%<#zR7r!0(0*~=Cpc3K&!~9q
zuqQO^5GP93P#>OGx;^w`6s?sF(w_lDZ?<)kX!{E^n7rf_HpWhSh!ec=-h@KPQiwKo
zKp;+9Zg4;!7XcmlV}h6saBPA_W;4Jk2ay;A{=35Z9pMK%)~Up^?W`6M-O!c{9RgGi
zkhpt3qVhpQ>@SA)UjU$mj_{cv^97RU+oDsF9=e>-b0J?hSQ|hmv|o}~?u(5?UBn!m
zeiNKl0wg>5mZ(&6HNC%?td0228Gu-@6?-lo=w|by%=AHi>_Y>^Ob8JhG@r@TUOvAY
zwow@`#iR&7(fhR-Pe%WiDgIWt<;%Fh+()wE9$Dri#$94TtMFeI5-LzJ?xP{)z=@{)
z#%E0u-}nq{Ro6e6FO?hh3YO-xkIG9Du6hhT%OHlH@kG4?B;_V0?Wr{O;Sk36$?^Ai
z`i@-fq+#TMj@m7@IAO=O`-<nd{I6UWS;hv*6kH0aN!n@C+937owfXlzJ^YT+#lp6@
zq6VDd*3z`NRw4>*eY`PWr4%NXZ1sxjeFQ1<$B*ORi#s3kt6nFZUZGB;m<PscQdaiz
z8~*>vw=}L(rfq3Ta;4g5G~dmv6Y5|OD;u89{+|B7#;!CR3U3dOJrpX2jK~&c-$G@^
zKb53X5+-|@kdb99jinid%D%5zS}di=7Q!IQSSCW27-XACjEP~G{Z99Ox!>+RAJ221
z^YNVbd6(Z?7}p>l=Ev`6+l_%WCXKbMW8P)wV1<?-L$}s)E`o`&KU@Diw-qJ2TPkig
zhJH%2#>|bJ0{|&-QGUm@5WlURB6VJqQLEOg$s%$)CM+JgOB*B+t5y%N#Jm35^Zf1-
zLB4<%kM-HcrPL=yf1FHaTx!WL%@4?^pk-YhwRAisS?M;Ot|fhG9b$vb)SUUp8t>+B
zK1h$M(cQ;5UmR1YbXNK<;MuYLk_%0M!vHmJn_P`gGk6fM7yM9YQn?zT-|CVM>8W0j
zieT*+ONJUFa@YI0ho^;`PY);3CP7Ul=Y>34b=zjsTxv!3U;q!`$&fBo(;z1LhI?EE
zV9QZ3QO5Mw5Va?(s3Xm+Wqxdi?>L+h2uc?6>j&|wpH~3LWCy-JZJ!P9*W=ye4QSA4
zL71E^FrlH(TJqimYE?YO^qKKoHTeY4lO5c86)ehg@swvJB1X#hvY+Pl@iw?Gr3eb$
zVeY`tN-~-Ty6GVS&^0f!c%=NP8&M{f32t7T-HOm#@N&1yPR2)qpb|>_`70YPW+w8k
z4f`HiOQRg>Ki7SJtm(a5AzgoRaa(|)y%aVcehtBUc2#)eed`N-)qf_T`*<Wn_rD%G
z(Q7EX-*5ris^h9UDa=T$814M%40E=C4FupieZx-SUW*rM+zei=otPx&HZ(kakJltV
z`fgSiEC7nWSqH#2Rhsha9KT_{@;&n=wtQ1=n@`#5lNN$&7IcteAcaf+D9KV2D_#xj
z=GF0f5^F}{f*LiYE@9;FRYPa8|5}vlPR-R-RTrSLe1r&<P4oMV5!rHHTpjGY9Og<3
zF)K?obD}=d$OoGZb1kyJH@-9a+c{VaYubkojtFx?DrX;P9sn||Lx(<Vmn^1H27xls
z`im>e64n;$f?H0({Aj>=fcG{~8t3ldaYah@KJ82T0a{FwU{FT{4Z?p--Csb!{z;C3
z#n0zKkrE>JiSl7W;_853fESMF9mP);po`B#P3`%IBFuTs`QE}pCj>UsJ56BwG%U=8
z6x-e7(PoaR=ID5s|8*;n7NVOlg4R-^K1WM|L=rrI1Ejb~0+|y;{*Gq{6if4!=U(2@
zY~qWEk+9Rf$e&oE<53Mb<c^<ug@IS-nyaiF3FZdm`>lSR)I;oTIBI`yG`;0%UGRNQ
zC~=C%>SNaPWWaS<kkveL7;rX75i56W@q|*LQIYBoI0~oBQ?FS6<GRP^bw`h+3N!tq
zl|2Jlk;#fMLorrEeSJWBq}kfeu&HQ<cgRZ<G_S8X)!MG=yfIZD`~=zXL_fhb;UA>W
zBO&r(?fpnsm&aSo)BN)}__`08*JCjHe9DgmY?i_#e)vw3^EGnJ)NU4Ho9Y^80m`Zh
zCGU+V(_5<T@Z8kG2I1B!0JM;G``EfnZ{f`cwV*P)zs$ykMuBRtM;W;~q7qV>H|)w!
zPTLF=-TDQ-ng^4cX|lo_VhQwnLHXS>fg$Ca&|baVxVgZn2~+H)%(mGtB~5{<Bn=8!
za@Y12aNi+V!ka+gk76u$?o*U$d_c^L@!0GC*65dWVLk=N>u&&&A@4xgMRkOS^iYb0
zX+Aa6?PprnqvMCWUB<X*>EpFU&)D(mwNFjH%9U4-oA#QtJqd5xV|J3hlIc!pMC$e1
zEw^=rbiV|ul>;@2QPH^xL7DTPn@xv&YRGyjG84I<Yb%_kV;XKt`g%UaD=XPe$;pOK
z@|BoK{QRb-5GMn2Ws7Q^27Wvzqe;XI;XCqDl;`zT_{%<~P|=De+p8b%=Bc#&b?o-n
z@$lUt&Ls;-h;@6HfHe<-JrEC%MWZEaRNxA{clO?htxIAQ*55&PqHJ@V6{p4AgACnq
zDhj*lIT<!FkK1cYFN|5zXTNk-|AP}&84-+KP{K(yo7&xIH-pk8yuZlI+^I?}?}_V5
zKe|s(P`NjjcBgfjTl-T>QjPONt07`G&y~>;KSfFe?=Bqv^X@fU$Z1K!S=~B~egC2U
z#dB~{*OeP*I<l+KrF_xX<&@zURtMM&5r%-{IkWBh#Dhl%vr6sN?{=Sh@e`8FcQ3o=
z_>^y?1MF+sQbZoxbwdiW`xfF`#8xJ2U(FMg5zY<DMyH)rGyKO++r8{IJ24`ys`SfY
z`*#BrRUdWe@g}R<eAZaae<-mj>Azm3iQtPNr^CPc4j)im2CGeY6k30eYD@I8lS)oG
z$pzZrg4;e5lcHCg)jue0C4Bl&&=#NklnRwmhHX1YA3>5WLPk)B!k4*n`#vo$@j0TH
zOqmKfgl$=U4F*0dsGhJK+%n?kPkzHQOxgw;ZIuIW1Aodz>SFsq9ewdHSdGh`YI&!V
zgI_E!S9=czLL?|Z%>n$e9~~uWQ+{SIc9;8P`tz<;sG;Lyt8yOV5nJsCMSOjd-YniJ
z>J#?wJk%d(j>bTn(IUEX;XyyiC&J=p#bu6Bg;!J{&sZxBjc%o>2f|iZA3M-|moBlv
zw&{UR{BpXd&bewm`_UOnH<`jqmb^@b{x10DRAJ$I$g@K}(9{xeQeDMEEfWawcqg$n
zz@rpXtE8|GEd<~*z2e@r-(wUI>e$g+r`+)zmU>Bn{tA<2BsK!GKqNtxvZIx`zi4!y
zNh)@%pU?B#6TS3+qoJZpig+M8UPuMmA?)|9P)<v7-7D|J>1?1%8r#}zC3uxOw0`HH
zZ>|BGJlbP~SfYln+V4>(+wa33(R!*2`-)~v<{<382uMKfOWLSoNsam4=w99=0a;N)
zptt6?|FL0ml{*tMZ7kTZpyvNM{$Cf?0Gq^Rew>v*q)&Pixj@$_)1i-%3SZJFHW-p;
z#Gk=Rp<P@`{p;1CPCAUk9;N#dVjI*%p0i5HYz$4qKgvPzNQO7l$&KDhd+i^d(#-5G
zP#baiz}Vw9KUdN45PAq@6~%P;x-xOu!$s(xclhzMCo(QQBjmWTlUG;6pS1u*G;Y+Q
z{h2A8JM!br`bm@(V=yCuu!P#E&N}-lAW%*rel#rBK|OG7XCxch)L@u~xomePr-GZI
z3A@HpDD>wxUALM-cHI+y=|Y<`MGWa{oJ9n|jq{<{><7IERGn+%d^rrYzsR^4Nt+?)
zw5^Z`J){Cc)L@Q{=fq~hd=a#l`E|B{u*z26J+MmTbb8_184)NN#xG-Uvl6n`^^M{4
zof?5{Q6o+eT{%@h?BUbJD9RawxAYy-S_9-$o}i@d5e91t-^bi&#Q4l@;h*9VM0wG-
zM$K2*OTEpXKfKAiF+4J_fImh4?egMT5Qln(Gt9)cvxaf=Ox(N?WarP#Sb8gaS|hIA
ztB8{kyNtVimlNeZ5nP;!LBh9V-B>vU%VLf&=(5HdGl@7|MX1xrAIZMeq+Ly3&sRYG
zvg$fBiHg_^q3lAaoweOh2dw^v8Lz*Tf+vQ%-ugEF{ld^RC4=@GpJj<NurV_WnYn5k
z&3(2Xh(WB5cG2mJUYsnNPn+?zMHlV&94r4b*1KZuex^djNVZL`8#`$@e<NbfOR$}_
zMo>dBGeX67k`f$n!>yzflrVO`G;XFiq7u0Rs?HZ@ZMS?Sv$*&s!3OMEgdA3_dz)&L
z+34b|*G?HVqIk`5GHG+Y*-+c5GT#BI+R{Ic#OxMAmI6H2sd}y^gd9mOLoVDT8|$Em
z*~sNhBLtCNHp`wi#`r^Y+8KHXmUvbE6Q^BVme;(;91C}s5C+~2LC}smz#Ypg-6EjE
z*r$Y1#{7LSbEGj}Yt_BW`Rz_+^{fHp;ieH~E~Tw2k_2h<mXDxYAxJY0U)ein${O&-
zuSqUrsKs<7B?Sz_?8I{XfJ0i=*;~c%2L)m5x8k|}C^qWEo2(q2W#&v(?u;9<|95ax
zgU__IyuK*Wfkf`Cgy?qdIEgftzM+5`aYF<@%ARPkA&9eENmwY}Ycd^envu504YqIj
z;|8W-bz4<lnUQT?a}<N&X=nLu%0x>0eCyVXj?a~%`IrEgUrqGx3r@t@;r5=$9R~~?
zOchly;;<%LKb})|+ej@(QK6KtoIerv#R`lB1-(Xpcs`l4p^#f%gla)D&dBSzv4`jN
zXR4hI5J!iLL*GldLukhAJuB}Taum6W)-}(JgFd22jkny_2D6TXZm)e=lE%@;LlC=8
z)%lGmW)$J6PTK?RHc~p0RXVwU0Za9tGf;T=*%!RCAYOnidYC}s)H4~(hMcd(kLvUF
z^xPQVrtcCocB}Hmi&1AsJVV;mddy}H3FXv$;Rt!o-0ND*v77^Nm(_)9=c~@SKmHfP
Cb{xR~

literal 137187
zcmZs>bzB_JvNw#gSXdl_y96geg1fuBOM(Vn+?JrhH3Sw35Zr<<5Ind;a0?z-7F(Pr
z_rCXc&w0-C{xP5F?&+znuCA%6s&B<=X(-}hzr;pDLc&v4lG8y#LZv}MLN)<l{2fVl
ziFZXp!mfBHE32g}D@&*4>F)5(#U2SsDK;YuQ!j0pbZGy6Mpg!=^W!?&AzCtFz`;z9
zf)WatP=+q%SIPj1k+z(i6oGdMdn>8bLPU&>G#bvgZ<KO+b3XM}ediLJz}4XWPQP_=
zN6*mxy@@>0v>dM|<glv#-xs=8?~ru-b0UE5LhtpwaY=+;&=mjxQe@xwjVXcxJnAH+
zTqd5^*6|sZV%A>g)m*_go(<Eiz4(ogq<R2D^#Vra09wUCQBuofETjsW*yF0P-<n=L
zy`CsvrNQ#EUwB7hr6O)0r*Cq;)B*j#^GH9^vel$fkeFndG{$Y1>3ozV0J#^x%xHR-
z0HS53rS$}EiobxvTa04n5K@F}9=LlxfoRc=pQvA~Ba5Oh&_m#ceN6v(|HFg3sye|$
zwRIS-PSFomH26;*fVhUn1(5&`H?LD-+n<~V3Ju~r@Fd@>K5>Uc0Ua3%uQ7*!o7m&8
zXm^tz#1>V*^o`qr$N5<M1iTZ`-_dXT?4X}sia22J1p-;3<9A}IKDLHv*ckn$a{R!E
zGSJB!eK)?}8oz^L9RuQ_%5PbDXt&t*QX0RphM9c?xjEQmC*4qGUp#t&E^j|mXm79o
zqQalN{o?UKLUL9eN-D{0^kq^RPC6veAT!`4kTNPFDP$D%QCV$cQv^eTj$A98wGM6J
z2eyZR$WqR1u7~DMEujYyp+LF!?~)JSaA`(ekmqMsXE4GJ{X_$~*4K0Vlaf(FBr<&o
zkU9<#7a)njDx?C`whn%g3qEmrq~Fd+-WHBn=*EqFmzS6rz;NPZ)DbDv10-3ixJw5d
zs_<7cB|HEWYr2=5n4HM~op7WoH0K`*$@thkf^|fCVeji`=OX(y*_=?zeu#Jwn4v#z
z3X)?m@M2hmVV%*aNdv)=KY{2vA79gp|3qs5J0%PJ#=4cI*LqEi!z?AH1gH)dl#bHE
z(nFq?!kcoRQa+&arg#~4D#JJ>e2H;~TWm$hiCNiqcP5jKF>5U;jBA0x`s2{!#TQKX
zsL?Hx8{jH7B(k=*#hTy&C0`CX9ZdzfB#K{tTI=*1ogOX+4YO3WEJ?wLR?{~JJ<JMB
zA0lR%PPt$i+iBw0B(8D3a`y#&)12OK1E|{p=HWGpEkzyELeq=~SeHV$6vZF;dW5aW
zv)R;89GF>Xbw7&4h=jL=wSA)KgAUo$V;iAXQ7>Rpep2tP86>qvSee%4{X+i3=1Ume
z$G)THq0vf5MU)jgI&?Bf-J3VC_T$r*#5VKUXOJqCOaf!FZ-0CA?9+|F4dqSQ4SU3E
z>7-3507D&CC13!I3VzY+(@Sn0$0qxl?kz!IjB@{{eoq?#I0{?>UJV9=o57v^Hdr<N
z7_pwQjD%TC3QQ6ycFKcHelt3$$=@^d6pxtJm^+!%hVSeQJ#j_Spc$)D5}F}OPi)F5
zuhZ<58kiSV4OE&{){}#n<&$U9id4C>Mp8J^-m>_ZAC<h8HnChb|IFU0(XMm%^)mVL
zE3W*Y42u$reyLKa_Flu7fbo~AuBtLlQ$bDN?mzxZ+SqYLvLDLo*mBu0*~r*zJ&i<A
zMLwFPniZ~Ew@I{N`5XH?-vVx3ZdGn)&l%9uBf3yz>E#QJw!ZZi^g7D>$hR{neu%ee
z+{PQgSE*2`QK2e9DIqOcQGsNXkK~UGkE~>@q|<&Y&`;1Ot*p{ps`QlJEvfiYQOu#*
zBz5)eNc@e_xDl6}YtwQ_lSPxV51x;^PuHE>)x{Cd8{*eyZ&F{YyqR!HaMpR9_onT0
z$!Etm5Z&E!@AueEUzXMLrgUb{rwQNBJ6*j~e@E#Q?Y#fK#Yu9C{%Fd3+egKx-J8Yx
zjSs5#;(_4d`YhwY^d;ey@8Q)!@RZP0(%eSXM(OjI)Nq(}vh9?e$40WtflaoXkuzu%
z;4CMgBDicj=Z;IcLeBcq@TH|txVnpaD@&jn2B{V4FP=@VZ|1yqD*Pq64mpB3iP`Mg
zx4FBYn{ql$zIE_AY+Xl*<8<e&WS3>r^3paZH=Q;8Y&xg|*}iibZm9YWwU=)`Z!mG6
zce`3%?*DAtxp<d|H^B14p;0|++|X3iG~BiEq&2U`yvjUd4XB<-s`jeLykY29$5L&J
zr2X@7&jS7I*0(;(#L~p!#7~Knq!gUnoNw!#=9E`r5A}AKJWJmWy@h#^wmG%&n!h)v
zX*+6dFvo9~D50EAJBU80T^(CPb7NhVXccX(^?_YV`GW$9L6`m`0UrYmKoXZT*D7c4
z&p)3v!!1tM4kO|4i=|V6el5p3a=ZjAOVN#{uC)-RZmn+YZpEj;CzR(zB;g-IR*F9!
zr8=c7tW1Fb8?+NZ0MG^ah34Vj=Fj%#Kt?<SYy_^tI>agg*8p5`gK76j4X6aE+{n3c
zVR6ZEMlYv<aQYd>1|}WiCf@h1bMsCaPHx_TxZErZdsbt=2`8CFvl6(Pj6ZjKd*t*q
z{ISo<s^ru(4he{w!`_Rtt<V1T^Vi@no-sNHE-(H>0z(p^9#6*;n@+2a2Gyk<Dpn1w
z{dZ8^AmxWfv@$R?)=n&UtbHs`+vzIy=`j=J>(N*IVd-I?;SF5|T@xL6d3`yne*@Ic
z!tb+!iSIJf1{g9)RwDWprUn7tV7+jVq8R007T5)b&%Jt0{KP~x>^V^}cGt>pvs>?y
z<6_|p<Ji$RH4q<%kKFpa33%h|v-q<54dr4&RY`5NF=lnV>0*<GWAOT2J>ELykr1H}
zi#s&aIo~w(-c9#+X`<K_6wkj{(*JobohOIqC++f|*0wvx%gI&9eY274Qs=HIa>u4h
zp})29Ru+T7bcQ8o2>bKy{lZOk@IdoGV2gDNycy5UZ^^=vK{r*Gzn#%a(V4yJd%@+}
z1c@k5mT{IojNz2<lrZBys|B{O*6E7KitCSC$#Bcy(todSS|Rbr8rD|7=;Q9y6yRfh
zwI941?Drn(Y~p<7JbG=5a6Ae1q}!I7to?3D`J#e~Q|zWG(U7$QzZ~XP=j`O9=gjFA
zkqcJeiO!@slKk_bwR7ovTd90u$r(VHLG1jY<<|yA(jE25TLsaDATt<f9&02v!r+^a
z{;`~?P3|}G_TbI0u;H2)Ly1J)xsMipD#v=e?`~%Y9Ag~MCxRq%y?h(eHg3<&dfTRJ
zNX+~8izhaYELPY2dSJaJ@fKv0h^fE{3(t1%qlUiSG(#uD<(g#+*OR70@ekvB%drwG
zT?@y)i!1jh_o5i2cwgGKJa<;rlXA)2`P|GqsX>8Oyg_FZQ#16_)Yo~Ae!8b!U0FWv
z??HXRV~-J-Gx#Gk72;KoeFsUTDb)v;F+)nQ;wi<x;;Lz~BmtUGiqSxs5S?qii+6nj
zqmE^RorCuf9)F4Jh-aH4`zHv0s6i)xkk13pVs`**5zC+jjzwsw?xn%4>ze4XpLtqO
z8vaIE7vxUsO#9aMk#)H&e=-FE(WriixYv1_KX4ur>UKwTmxVk}IAUiD=paSm3QN$0
z3Hiim#9)NY;3It)uByrBbGq*SfgAqZ-?$t&=z{zWhGu2rsjk`>T!VDLm=8q`vjX>-
zA-mW!(|ElEDb#A_HZvhDaLS2D_2t%YLIf*!Ll5pDG%=M=<Igy12^c~WpDZ5#I<La^
zM#>KA>PQ@a+W;iw=yynHe_P0Zua|!>BqWpqlz&E0X$nyPX(OBb*F$@{-494eQb@{j
z(t04|KY5rz)B{b=iu0Mw2#X(`lsoG^R&j@1-?d~>%AuE8P^RasUR4<jrBG!6Q0AZ$
zvEGB>V!IQ;CwSUv=<_(~S2HpHKDfiqG<6a_)Q&ivn3#Cl`!n{=i^X|$?ofZ?Cy*Bn
z7g_elzum%7W2CN;t5zxm|MSV;d5l(Z(R>l~RseF8k7571QF)-L!G$-0D|r7w?C&eY
z_&fji0{_($-UMWCfA~S)O!<Ek`mY{4%^>Q(6aViS9zO_vs{2M)aqbU(qB&e{b~{<m
zeJosQcMV)sZTC6Ze(U!KC~mo1u^N3rDi>OIUlQEWeEO4StvE~Hz@QyHp0wjUF7ANu
z-UjfR7_FhPado=y<HGxik$LCU%<7f*)S{~)VewwZiHOC26Uhrpv{VinSxh`{Lrb`S
z7W2!W+On8!luzeLtxi{O8=`Y1Tq};(tK)eh3*b$#r(gOpMZ9&Ew2QhE(d7*dBPYw9
zD<{9|XM>wPGhVT?e<D3xEsqqv|DI-^bZY)sKQqIRh=e?joUC`Unv`-|nly2G5)CHC
z84qB3?J?^$*jRYSyzq6}m{5NyIk@}#cWa9ybYIYIIn`73V6md&bG6y$H=}+-nmNC*
zO^92QByZ=qOrT7=?Jp1!#>Pvht+p)f(vSCdT}Vlyr@=ov;S*ylICv`PFb`kfW=gR-
z$ni=?mh)_T#!81f#I!r4I9J@S>-n@FU~TuW(~TBFQ6@=p74>bMY2aVvl-#J^Xl182
zfsH-<2<COH>DX$2-p9r<o#R-67PPT3nXiHQ7(2B-)^#YMt(>2GE?&f(*65FBy6xAj
zoi34Ce&Ff6&g}Nx8d6<#f4UL!Thr`<EjPL9R(9Rpgcq!Lxta<&%uUz$RAzRKP>cGt
zAsi=t+S){}KE$J%V5;;T5cI9>PfCg|KMkm6mX?6fC_!H@`}DKS4iXQh$7vNlydGeT
z(b|Tw{R+LimkbV76ZQ4n?p6QI3HxxdH@>>q?H$EN^EePwH#C0dk#E?gA+jaxH$hR>
z+P(hZ-eQ)3{V80lpj0D&Q_yXFyR#CraXe4d$GZ;$-`i`S$2+&hMLfHuIgsKX)yG5+
z{~9;+x#@T<DI}!x6$htU1!!u!p~15ya5DdqIH`@6%lm?i+gH&Swuio=LSRG_vQ5!q
zv-^HgM+{%`VPVlcYl3<Weg6x6+NmI7OECpl-Ehud?(W{SO>Zz8@cN6}N>|_7{n?I$
zxAPsAuJ|UK=J!f^-DWnbNV%V)@K$!Wsjv{Yi@j>?0me07MqwKI;E1|6{=MWaZheE^
zm38OHXg|E}uk*qg)$c3CYAF5Zw1n)Yke&<t))95N0`6VTE2T{xr$ZZv^(350(b>mS
z9-)9qC0WJdP>99p49=vFfMmAU6iHHAlH`JV9)FK3VaX72(#Hw*1&RO9WQ3~$mA-+9
zV;31L#i+GDD90MI_{JJCC>a6j=bDJ5QIi19Qrp!%-zqA+?YFl8L_QOyby~zp<?EE5
zbIH<w*)}50L#^~2dU0aj!kxF)wQbY#{>tQ;g0pM`OS;G<#3FyC(XZCDBP4RA!w-_5
z7?PR30QTEjQ;s^z^oA`jH4$nmXFl}lHaM&WY;+kd&VJNrcq;j{`Q4m2A1_xhK9#hM
z%!fq9NjKZcu>8T@SWoYGu4|jnZ(sm&A+v^>B(ZCRGpoLw%AV(cr0g~~-N&SM99-Yb
z?P=M47ne{$7_3BGU?B15BQ~(Nhgni`^6Jq4E6ntcu62)|{Aj=r&EPHp2CzCKx?&5L
zx3B=aHidwac_pEqb4fDFqMn}bJy~6{v?QQL!2ws^;IzW3X-i;tpz8fW@n7cJ@F<0j
zPu>R$x-X0co`85Z(_Cko?JRznkTBgK;{#28X^4637P+{mUZa1QQS~=~hMoydc5C0n
zIO5oVTlf@5q)E1$mKozj7k&OrX(aBqa;-GW%f9Ssv)3*(pnY6zsWr;9w`)UzJFw|1
zDb3UHae&LAMLMUjuM>nVmyNqmuY-Sx8q(!YB+e{2MC8>91l2z{B)sDeF6j3m(P0f%
zTJb$y-W6xRORy@TQK_+Z+3wsEKbUDtTXlqhj0_e{bbBwny3)=_cQAs-f+NsMA(GEj
zp#Tdc$I~|2E-sTBK77q24y<LtGqB_hF5$0?;f}F4D&TMLC_&GFEO6ZKoHn5yR2nA6
zfBrrN<#fqfaN+$+-BqUg2wUrD3G?RF$J1xGI0;diZNr*&HE=T{R4Y~oKI7~@NPJpM
zHgMZOCSCR!sAo%{GSvT}c=dLhM4H#CK~ebrUJ~ZPxvei_)V4RGx{73M7$Nmt7LAOU
z@f6-KFxg?a4pNMFxK!@6g#KaRq0(ZdR|Zf53R~7c)SA{;lyU*+iWRALx!=M#u~=W7
zqd>@l1}eRt88u^io0LeW{Jb_gu3t6vb^uHB0}tQO`d+8OLK{tbK0Hx`Tpw?=Nj_XT
zWHClu-r?o0-a0}ur@ap3TQ=?sR*&_0stPadRjNMRB-NTRV;kFj!tdRYM91e`r<OZF
z#Yrz$gal1oD4$8NX^@4Ux(F4xyVpVs-~WyayM?o=b{f8y^~t8Swr7mSmu0#3-swK|
zE`x}Tzj`q@xg24+en#eZh`z7%jQEt<al7Hy;Wa5X??b?x<N5vMaf!B_H9vSd`2N^1
za4UOfcbDY!Hp|j0N#e?^&fjghK5ezl@yikV!3klA5nC^79B!fOclvxS^U73V$-7Yj
z%ZIi`Gkh$0rq$&3M!S=KMH~s=mUVx3J!Lv3f&{J9Gc*RZVym$<&$Cnxjzx<L^k)rE
z(X*S`4R;?yb6;zpZaaVb>-sW?7f<(7^XTZfPj}}jYm>>x2G3%sE*qYqd4RQEidmyu
z<UM5LsocOgWGv@C1%w5;&Tk|%jz-@dz8?;CDr1A#&ksF@PeE*15`Fgb{7v+<9}wdB
z?IGL#*<#jJ`^f8o(2{3T3HxBOk-ns^z-mLM^0?%2;|7-Fzvk>hg&pOw12*Z|vgXs>
zP1I_^9rxCG47hjUc>%^}ge}I`Lbvdk_aVXUqSi;hopKd%X^$4a_uy~bhd%FQa1gK0
zmO&0Q-ky?g7z~OY-Ps-T`yn2aK(;dR5h#dR+7L%cd$$sDC8J}n*fX9-4C|nqaoVRn
z2)cVMP51f5{(Ze`7t;)#+W7x^7)~63-`=o!*gl-bp=SsZW7RS3*kq&A9~Y}{b~tL;
zIydBValhdwxtSky%`9YG<r^FzU$w%9p(**?0Fh&lv=Xx47IvTCTg`=a?UQAAy>O<b
zs1kk}p>$Uy!6L1t5q%fY;$d^>vQqjf@b|D|XTi{CFzGh?!LRzxKt!jp-p;$#Zv>}b
z-wlmCHc-o4nKKXrHNat#b)lBA($7Dm-QW2q!f9oaAAZl;2XO4|Gn0KehVU``ep>2$
zy2jlI8SlMvu;POQv10G;<}J@;RwI+dPuvKQ$#G!Qig+LTqVcZbtCfpY7lqVTT2s(A
zGO3XJeagpPGjPq~dcfi*LTk#x!~Ml)i>j~p525CEl6t*nz_|jy4aiA#T~OKN4*?2o
zo?umDv<rB{3rrNuGo_5ykio#lf%};OoN2%wIHHpQ<y5iA>WVB1uNKWU58m_VTNyUa
zBYxNI8{^~IJ}!dY>i~j5qkU9bg5JIhdZh-MnB-0uxE*e0fPYtbk(mqWu2&NnZ5v(o
zSKga_M#L93{Av@DVVnK@E~+c&?L=V6<3w0jFmy^~24&$Vlo5*hd<k1L<%ErY4eng~
zlC&4d9OgLW72z47)%G{*Q~RsRsL3RKb9!AgjVq{`w+=unA93{p2)~n-pQFWnpvPJD
zlRuQpW+_C*X5Mn!QAqMI@9_3im3SPh6|(BN#_zg%lf)$>`~Y9hg@J;$>XB!zb6}Hj
z6-sLv&>~3k>@_YWzt;K+{_4&obhDEK&?aloh5QjMeb=~1IDx|dduW!^%idENf5bb?
zIC#A-OS?U!+9Ndm>eZ!mHX!$F&5{oO12+91#-Fo_N#c)C!}ooQhwf+eU(+1-sE1xf
zDIWT2OEqI#_R3m?vxWiHplAZ%Neb07IxD@+$31cjoU55Li*0&Tp>*rJ?UvB?X)e?H
zUIcYelu%@sojT@vLrRf{6yWh0`e=c;9zzVs#x9P>Zu`d>dipD*x`<uRU2sPP(=J+q
z_FS=M`;RX&6P&-3pr`C3+n&agv8bMf`5nfuz(S@Z_jP(Q{iH~)F=>KzCb|z+DrpD?
zMkMh5CyodoOpsFNR7ohaim=y7kOxP7^P$>GSN6MF_BriR%)gBAsdBcV?z0iK(|SGt
zCtZ;uGmsXx@>+X6tw+-&IkYg0eZ;cBq6cm{FP)QEfE4O1*n82zQ%P24WC2&>EuB`)
zR@1B!q?B9sqYFx(w={~-4toQ@H3CctSUnlZjGXxZJpc>fkd1;>EpNNqw)zE!CLnS8
zvRJ1%4}_~F6_H<PHwnajqhiq9HE|K-48~6fZl0AED2=7yDgt{MBFbkzF#KZRmEq~U
zE45_oe_KelD+(@9E9Hvel{iriNjv;gx^j9>8k$~)a0otVEbi%G8&?4gYDyrA3<)EJ
zci01EKu28K-!=k-iu8*A5WXx0qf?iBerqI*C&@*!uG71d#ZLXo+4(H92Y-ZCH8W%{
zgCF!!i{^kS_L=n(5b%_-*lCz@!R7;0p^O89J%N`<go#lMgZF1=M3Vts+vaaApu#~l
zX@7Ldk?BC0ye!M8&s!9|ZD2&TkxaS{^WAGhC_B^pY<d-2hHZ%i>X#p}Xb@sI*|dSY
zW=@ic6e6p&Z;D;T?hUXgBDDk|p|?KKKQx|qltER?wmwihirM6HGK-MD9hDhukKH9c
zEnbZJQW@A^Xv3{Axr(?X=4q${$6`tmhpcVVpz3n{D+T|A6q6k}xA8NbOq<hlbP<!R
zIr9UKEGkVJfg&D>=B(*y8=2*mE%Ky{Wy+FfY?ZeKMe05;|3C5TXF?P#*=mqh4w)@s
zuPBj_+)FcN7WZBaz^sy-+V#fcV3z#U@VSA<k<3q!%2Op0VOM@#w7&DgT_5<f3yCUI
zraT%Im;_7^<AdC?8j?g-qz78moY-M5$FcsPa5%MJR>MJ%Pe)8wb=N;nq%&XqwuyI3
z1YA|<LMKM#`Fjx;EuD_E8RGrG*R>8WRt%tzdt1Ligqj4*dQpG<TT3h1KC|}Q79N^R
ziB&>~Wxa<i;PP0|Ws0Gz_D~3H4V9f#8WdprnFg30XxNK-Jto<8DkPzYZ9S275y}{v
zSsTNa|6tPnPWKXb!I#UjGI-lsYx^O-woCh!SQV*cEGrXHMCcvjWS2)id|4-NIHvXK
z`Eh3>_<&f36O6#zHr8fJH>ipI4r%`Z*d?y`G(B=4z_1xf0%JwtTZ9LJ_%&0|Z~$;!
zrhY67uu))`P3J)x^OGM(=vk@SOUdQH(NidleuK5#tjt~u<<cBuf3klREEJC@@j1@4
z;JdP~r1*qT7WcvzfBL@D9bj=6i@@3r?hE{Jj|mTvsWxrC90Uw3(I&t#2O{oetGcHH
zw36vP?9`ZodF49bzry(9F6_zj_B(qz<a<Cja7QW&hW~XKH$7zO%C`ydfr7+(*xbJq
zV(8J>bGCc>gAk}mKT!@BPJ6v3&-9zzHx_UpChj+B?F4jmd#+XW!0k6g-Gro9!#hK9
zJ|ALQz7~H0j#^mCfpG8NT%N3rn(DUEI&1WTJ&I7UNe?9-?#kxn8=zNdv_?AmH?nMU
zz`;lXyE7#BHTk-p8P#7zq>tX;=mA1i`q~x10qxi~xIh*8m1Y8qB1wWD8Fi~W($brZ
zb4nj(E`lmHG2N53xoAkT=8TRbPpsfwI=l%=i*F7^7Otp7-$6HuKT60hebT`@md7tS
zM=!~}qSEr0|80DmIEBl@R`}IR6A(p-ZiWTbn%nB$6fNJsWqPd{yN|BBuGuggE(v3x
z5h6ZHDRKfb%;}>A6H|pKLdD?q^X<1xK_{wNmP|C^`lo}YFr`u1vMb_vo3C-GLqFe_
z5j)e1sdP)&p1sY-bH_x(e?;JSacwZ5bYIiLYC;g`I}h~48i#i9)MV`!3}1BF6C|Pj
zqbGhH&*{7}gUG!ip&^g7+b<2i)7-!Q-%pMM1^JNuLJ=4Hv3#NTGAYVG@sEpywjjK(
zNZ6r&TH=LebyoQT38x}xdfw1%JL3hKvW)64)8FI-FDFy$%G6?-u=C#frYEQ4g%F&F
zs*DLo@Man>G;?MpC%}SEj>svQ;wB4C=oz{wkKMC-DSp(wE(R5_)wAVofE~roex+@x
zrajdp?@Rs$f6N90t1^gdf5Kiu@Ot@2<g;z4u*u3V+(}=Cp2TMbx|802UiDo>m6)A@
zh1_oVRCOrWZYeuU1{S>Eq8HyMC~s3sR3vs+snnrqZMbtLZ~5tNnUx8oGbRczl2)2B
z()KmJ4@dDF>codx+9xoAF++BaUagrJgF|+z{Q~f16KvS2;?^v<ahuzG<+q#}97#B!
zb!;*f-!N*|I%*Z_v1n=^&7D0@^_4Z?6M+OtT6Jn#(la}3nHj0$NgiC#LP(nYpYM#{
z=-wIsn|R-%ydkvwtoibLSLM7xr4nm;?AhlbNN|s3T-bA4qsB#!%l8^@$)y-A+iHSq
ztQ5FwbtHJX){h5KT^l?dtwNt&=q&}HtFk_$frJeWK++}e;}HHZ{BF~jDE|yj2G$Ps
zGf5c}tpXMWYI;f=TiM>Kx}Wa>(UK|KkY7bcab&b#ZH!i045NQeKZDEJ>5+);P!CNo
z#r=3C;ofwwXEfOw`hT;Y*L+<T%r}pk7aRUvx?X>Fm61U*{RSx=CjfSSab8S%4<fH;
zv%nH-!>}<A{|Gn_gqNvydXD>n+^ZfGmmE5!`5Dm9Te|M#yOZzSpP}Lv-UYnDrs<vw
z{3R-!G59r_*oi!!wdgb2p{$C%-1Pjr^%U>V)@ETxBI<hkgX6iyWy)>^ix`qva9l!q
z154<G4P<rO-`E!Qc{7R`Czm6&=lpu@pUkJz5r9cw%;x7T4biofl}&U+ytDCZGe)U-
zqxZo*!l6o(cOQJNnT{AfIpAq4o*9lKmNyupfV_bEu-A;Zi<o5Tm=%@g@?R|bDs;yP
z+lh&{eGOr#VLH+d)XXFtF(s_O(HM60fvkxHU9zO*bk-?(tV@1$I-{wFj!#8N0xH^6
zRS2t)T9Q)YKTs*gh|J@qRC`>R9tJ(d*vqozENPzXZPhlpVB<>a4=NZdg@`LixbG`T
zM&Z-y(~Hx2XFrBTunEBf#e>B$1{lXcevzd>&B@^nm%`nu2K8Gd5i)%-?Xd{xs_Ht~
zfNec^XMn+Dj{lE%@nOYqGcoRvYz~7mT7Phv8aXQ#zQeumV1;2)3de13k$19r0v=i$
z*&ybOVq@gJl`@UA5k-GHYr|H`i>w5#`k1?4b;LzA7K#d_unP&-4T_~NDa)5RK|R8c
zs=nNSft1%{(hOr>H&Z>+vOT<#=@7d5>GB$!t<P3Y%qQ7QIQNnCrd1;M6`*0GtbtJe
zz!YZS+u%E_(*tYSi1~Zsp#Z7H)C2c_Q+#X`EE0Thly!raUTjn<038V5a1q3$mx6e1
ze~>s~?|6D2)IK(}eqYdQYaQSFttE{3>zhChy;jBk)Pxd<@&mL{<BFa2b|N6|#R&sN
zQe6xqc#c$<K?c3tQ>4pw|MZ|&&t_hE!2f|SE%+w&uOiY{;Yg!xr|s*=)c**me8$B8
zv)_nb&y6IV;$0CWQ`KoT4f`A76SxxjKN<s6O|`~;*2tQP%M8yO2Ccs${AwzG^>t@W
zsOC08Gq3`&ph@!NM2Vo2HRMb+1;Ml&u~2nq0*_(d67Re`7-jCF1w9~oSoG#(GG2Hi
zq>rR5I-N%!W|p{XHMEepU6hp79^bLjz<3zP*(o4Sw50dpkN0ww7x5vLWVnj>BV-L(
zDWrzVeT$bZA)digkp)BA1@kLKBaA^}%IP*Yk6&iX?&X)$9Ur5(Gabi~9xLUHNiMI9
z&eG0UuG_*UGT$2i8o#PyyxBAC{!OOq_17fj;k=llU}#N6$L+FjagO0sg^kCJvqA;f
zkiM4N^APO@i}{X>b`jpn4hk>qP6<_^S_dogzdg-X?&6BPL-=&x!D!nzp3lpko}Wja
z=>BNRu;taL7HVphDtiSLG3ms}gJ~_i*NyX&HZgaVXRfyCBk(r~#=(OZm#0n7tx{{7
zm6Y=NEcqV~S*iACMKq~IO;SzGc4o1iDFJzN9&hr|H5ZMP$JHSXzs%DpJvM4X4n5m^
z!!&cp9P-IVpQx56pP+2iXTTu!Kx5P1{MWz}d)k6xCLDv`UoN)ZJ$-!;kzPmwIw;54
z{;y9_z)qj+L5shj!V5qC9m5-)EUkoLjZQl{XbwOJ#M(7b_<XNYPX4t|W~W<Y%bJ8(
ztk;wY%=tEKc9~Jhjc3H4XW1duw#ZC0xmVqg_MP+F$y8ToOL{{YdWV0gv>@;78Rc>~
z-47XSaF1;T0&Pne<*lm>l5OFi0_5Oe@g4g|fN1Fkd!WDCOQAu{JAlnssC}a#S_xzS
zNX=wSf!eYf+M6dYynB@7rRW4{04p1-QSlOU&UM{BDvsAM2e?5Y2K!x%kg{<NV<!(b
zav*_Ley2vA45M0Oi=5HT3WJfAt3nKGJb|krER~-&;*{t49`m$nw)9?N>N@T7jh;42
zeSTw^Vnx3Nrc3+KI)(|EtAdtnE3^4KewnuRt<aEdKg2=<;NkD#UKpnVzxjmXoY`)W
z%8bVNIacEvc%u3AdTNoH5|iF^a=CS0Nbq=U_uNeTb!&Hq)AfG2=w;3M@gv&1Ma3nK
zOfT#gV|dd`KoRGqVZ6X3z@Vzcng`GnSL{u1G$T1Wu&izlNPoP)hK~rBoBGq91<`qE
z{7pf_D&f(V&ttXmQ@$P>JG*)QrzMM*TGg5hG{BnZqR|dHQuZ5EiJ-@73JZcTNuydc
z1rPxSwmXE49M0|OK7*Pj&E6VIAJO>|k0$le7I-}=b|qK)lW1M(=dUqikih#lL$9c<
zBKEP|!ySez%qy;W#3I@)vFUZDpNl-$cv4hEQny=1SVQO%UPPI6zP=O}u7CHNLD{cE
zWia9uPaEAGE-lx?FV<oyJ}cVH=ezZ$ra^S@<vU&0+wbE&h~XU;1cs0d<iJm?0$ORe
z6ZkvU@j>81E`8uQ+$P@fBXKKK(;X^4`zj|v<PV|;B>k+GTpvX|h(6<cgDg=9>$<pd
zF2X<uxpgo_(aEK~afriI;kw7OPPeME>ZG(qIkqzbPLfB+s2;igOQHW}Ro}QBT_zuB
zUpT3yzSUl!`UH+S_?<{XJ`fh0m;GVQQKovpEL!X5vyqQAmfcoKYUWm2;DNjNB<s+X
z*%_3k47U(iZO?@vLyC=o(Ck#7)5?wT@zU(A{Fy0#q`LqLi!@1XBeeb;$nk;3I{pnC
zK-kFbovGM1%>(|rkeMK17a_rqzI{S-GVW3LPRlAUDjW$_C_mM{NZ(gitPR;DXa9*}
zC^1ROA^Rw77R4Z47kjaAsLyAhU*U-UJDXXw1tm-vvDs!qN}}xhHK^MG0Hex$a3M&~
zIG8k=oV>NW`9LkET3RmmP)I(V$zgmi-AcO-rc+o=3YO$frQRw0FZ3%~t59om*sA4T
z?$CB=Ah7-Gf_*5l;2T^*1pdaW1k*Zbh+)F*CmCuu!yjTZhTwP|L_BTiWD>KThi#-p
zFWE|y>+vT-5EXHs9AcI~!17UV736p*v@R)lk0Xh*%g0%FYu5(OKS{1<e8?{{B6uM$
zOfF0qG2jN%s%kzI7bYvE<32;$aq{YMgxF!&iN`Ncc1?s({sSa^A_xZ?S4X7TK%;Go
z9MAD^__u~XAN!yr8XDFmWX?|-Ei%2Qjj#sQyOz3QZ{*$|oo)rZpxVX~Pv4^AJIYmI
zpzJ;LflCmH=UNaEEWZbPiq}AlP9m)(1H^7(NwTI|?GSw^2~mOst^`T5MFS8e^Lkb0
z4$S?-V%L@JGCU1~=yeeSLHJRO-<P+C-P8?Tbieq)y>(XEF__<7>m}L}fn`>ATSmLY
zjB8>%203SZ=_16|YpAS25;wWtO*=ygXCoxewK|1ztpn{63m^i{GHkJj{%78jK*<`z
z)#ib3JZ6%IKW=H0j1p)<o|{~Xnm=$8^q_*2qNWK+8omAjd!>ry=?b@}oggDGPYLLP
zqJiJ#z`$?Lo`e~FFCkUj3*84c>vBFzZKz2|`4r4Dy*l)qH-<q^LB`SZ%WH@;qUoy_
z<z~6>D4<Ay-yl?D1o|zEu)#<>#Yk)NGT%4v-{9((XoMe4s6VqTng7ix(dnWA7OlNH
z(#q67v3OdY5XH|L>weA2iMVR4v46-;Sr=9KmSicwuw{&CTZ<*L{So)&0okl`<}1$U
z+RodWD(Re^MM<=k_Z$D@r;ud};0#tYNrNkX3!bu8l4xgib(6U9ZIN>|>bcoV9c+<j
znlc!@r<t6OW2{a}q@=`E9o!pOv5eZi7{-3Rq0IS_SKJE!NF|JZ${Wa=q&R!18c6sL
zR`Nf9eL)XHT_|=D#)_VE_P@6GKal!sgTJ7>Q<b>m$$ujGF9}deL9dG@8cJXOqk{jl
z+9wiZU9p$Wlg1tY+RT5i6|MmO3(SA|)>u{kFUJ163_!|O{s(c=;Ih|(+kbLLK%bnO
zHXCE&tp4Vo!0?~?fN+IaDK`!6V}&IT=kNb7>ojOC`{#qRh6_6XNmH7>PPEop<%FHX
zbl>HZqKcL={eMjP0sMyZcYkx};vEbBc=!KX<)9-&SFKf95$9>Bt@DRNC0@5;6`KM@
z8m+<0)7;JaRaQox(Z(AIoFaz}0qGwgE~2?VpVff87NclyI#`I%5*GQT$7_bs!u9;Q
z(%N+m#I9Y9VIzn?zBT@;``F8No;)3U9*PCtwO%Wk?W?3*=F-vJeAB!V@mTTHYGlYe
zP2L2J=Q#{k&`fquU-NbVd2LC2cE%)eI5CK8wxY50i;uN_LwqN%(W#k1Qp#}LvcD`H
zJBohgeyf?bKhE-^fkeV&s*EH0j9kI{E0?V%&Yh;$9~!q08Lf6q?y{Fia_Ik(rf~Ds
zzXSB|G<m3BwYNem8g#^{JJb#V#V*|f3P|XrFbaBol|jciaPR^vim|u?MeRImfq|hJ
znu}RH&ZoR@j(@j#M4?s5&5cqI;&R&})?|9ooTV~?IQ6Wl3yzyN=98t9M`N<9b*TlP
z>ZieT7_1>ziD7V)PTBV)ga0M97ref$>jb<Om3~Uqy}|%6h)bh99iLTSjAL~4%cygT
zXOzJdPzej!n3ebvL=Gtde<m40`zswatwv}5{~xi-Kr@v|dYu2(V)XR>D`NZKGsXQ2
zL-P(W{ynRg|Bost;e@Bp9+v7m6Y+e>%iH~Ub$EL2=~blJ?tkyrV)K~!H@KL&0sb5r
zNmVdxtN7}_{_?sJPNo0pk^ZG%zj~3+l}IARB(+hd6Nl|=xP1B6+`8U+w9B4c=>YFA
zl`%!OEyyA+k;=oaI(2l>^@FMJT-9!sPXqFU(*Es>xi0~B`@S`{rc0$opRY60+vG?8
zE7b#)=Aa2~_wz<g{8RO)b$l^^AZ?F}8!;Jrah*i9gyh~^C-KDmiqmP&{rT>T=4!Pj
zl%kQnP2_Jy$8;nWMB1dmGt#u(LH8{Ad7?2=I{BJ3tGz|SOy}4hKfeW*L(luo&`?P0
z%x$Cn=gBK}pW~K4-i;C#TfV1DS;kS(n-#r(D<KYg#LW?ju4WexVlBZ4i5AenTLe7r
zs-~eK#aTNVY4Wb!UYBwxqV+pFhet8ym8ZGrcV7|y8rO}c#c|8#1+FysA1U>zo_nEP
z*rv$YPwtQxF1mBQg><K@x_=g2miP8}%y)cGn(%_~<(?xi_!JR-=;IgE7blebWo0td
z%fxMkogKqBD+^7@&ID8~u=V4pp<QPo#r;1P<}0Fmxx5+iVvRmSG%fG$xm(`%Eu0<7
z-mSbMo7YuJ`T4f*{>nqV!`CAU{nl-=OJ`#bU()!H7rGZ`=H2_N$IWB1CPb}Z>zO!*
zO`i&XAQA&oYaCJo8wwGB+m^}oxg+&&d4|tE&GQ8H&yd4c2vM(*=@$|a2Cl1nA-)&3
zHPE8Ni8VYOxD@DUrlv{M_jlO0x-OHqJH;iH^0$d4A+%<hT(4?9ge0#&=6dT{BIYj=
zRRg@Lkk#glP!oYC28YF(h8-Jg7rh}wSLm~Yt>NO%*R4iS&(7QqJN^FiDQ|ymYY{fc
zt$*9;@<mx&@tuz!ZAin2Cb6#FjP;M(Re155J1l6S+QlkOiqk3^P$O|nF4<Pd%Lk$s
zdY^pbs?h2uX>#P||M1CVL~-A@dUg6SDcRBDP|z(c!N-RVJW+gi?a@e9dq$*5UGY9O
zOR(vKn&`Xp>BrTQS=QDy2BgU}zCSS<C#zVt7#&_^iJp_=*|5{l!$%WalS%HAN06ZH
zlMauRvv^L>_31*-m1I$4pyPU(8=@;!t-@()aZ|bb&m2~b^ZkNb-$LFg5ue8H{%&i1
zBxY>gq8hl9&~Jya)_;}qGIGv#Cx$bdqv1#Uxaexc<idjOcvI!kLgA;5qcQduuY|oG
zSZ!L3d4hpS*q?Vh;x(vtSVp19!JGz1u0>&tt=EUs)B3T~(e?0BMN5Q?i$IDqd8oEq
zb|~Fl3$&=0KFB89j1vLfS@K(3R$qC%{dn%UoGpDCltZRnMH&+xzE8zCSd-{#(jKo#
zb?g>=&Lz3A>alon8aK;w2QMi6E}r4b%W4rag0?6vhEJ}kb)M2g-Z4_9-|o16`&0aK
z1hZT(!#Bs!F_<%`ZvAolb7OH(iZAo0KM@$^AkOrjsOL4gI&GXD#@caTT27Pog3aAs
z)hP0QngILEd$lUsjrIdMcl9lF8oGqFW_z4keP{7JFx=X=iUocA{9#Kp-@m2DSmt}@
zr>l$Yw83@))I<;WI5MnCT!Uqyea;V(9d~Mr6?JWtJT%oGYe3(f^$)s&CiE7~ZhjlK
zi0|IwxUbGT((bU}!}b;M{~Xe|86OB@nyt;U;(D>h>A!C@T0E)%eR&IK)Aq!O%W{fl
zr{S{zmkK7rV-nO2^96_XW3;YTI`S_b6RmF6R5h*_WD@w2y5WTg8XcPA(KcPhcJ0W3
zyJ3>i;8Hnra&)a`f{@RjxA}aP+!1g1?C8U~7!^>(4#uc~GH@SDk)c<#P(QYyXdFCJ
zo*t53ui8nAASn`jkRJr}R_ko@kS{W@t!zD?jL2bc4;J#o#Pel4{O&liHYPgfjC`Lw
zvQ{iOm6lRbm0Db$Lb1oF;eDhb=i3-RYH6|nlcQ&?gt$jJ?c3>$vJU!fk=unR)^^L)
zJ8%|}&H4i!t(ea?uOD1_2b8l$Jqg{hxZSjsv?OhE+6%eL@G>{ei7ri|JtD#1n|2k|
zO^bhVZrk6v2H*o&20p<p4*sfr)amKI!^n-nVQ8r4;F}wK#v3B{kOP$g98bEK4+N80
z2c?83YnRode@((v;=p`Kx<+&hfi(?sSB?=0u*SoB5J!Z*ALA98n_!i)(aUXhCpMfl
zrY}+9_P6ch`P$!u%Y%*t#t2^l4DgqQJU3HLcM)XrJdQeR45yu*IUWIV&ZINPb^T6e
z%&3g(M4Xoy=cxr~)hW%GeN^O9Uj4zi^lX1J_D55j{ob20o!+~^*15PdyoaOdKT)r)
zNM0?Ezscq@?HKM5sn_f7K5Kl`ddFo^S;K(Dd&nzojmOw8hlNgOdB?04_?1)IxAiRJ
z0RKiE%i>_QD+4sh5XZKlLGt#y+6_?&6*E@Iz&swGRTtO7CS3NuHRl|6rGK~fwFy`f
zL+hK;4aPu;T9s=DN-v!OlqPTa?dlOUw10mllMajWEtn~t5<qk5gEfiKCj}&4Y2m(v
zk%_61o1O*LO_V-#vVXEo(1g;SfjJ1y*w>q$Y?^|-P6-Wl#7lWCsD%Qo^nSkAH`=^K
z6N`Z2^ANK9k^v>bWB5d=$TZBsA8=1*#~ZO}>|_{KRZ42MKO<u5DQanVBw^GleAWVZ
zd7z|jrUl>h7lXuN89m81gY29~!OUogPx!XcNN^fo0!=KhzCl+}^gi+FTO>2`?sT9h
zfYtuVmZoshsF)o8)sbEPZ}GC=gy6S4+N?fa_$96#*Gi(E)HED5=1n1$D+B;b=Bn~&
zFbkidfX1KXv#EK?mQ$g!b1{gdKeU!K2^pctS&Y}y_JZb!V}~^#?Is$|3$j=tfu#v1
zrtw|8(RO1QZv80j5y8fw8DAo5APydM((5g*9fw=%n+VzRZpnk*?eUm0PBx80&M#WN
zcYQa>nLD5BwwqVr(cR?gNlQSP0peDJUQc8JaPOQBk3OA}r0DB$u6MS`iM@Tb#>aSf
zkFbESE|o#ZzS8EvR7N@MIF|^hReKkE4cXSls_Xa6c%!1z$>Xzag$KnjlDJJHt4Oss
zo4$nb@0ke(L-#2`M_Y!D_af;RW}>!F{IusWV!$s+DHR~vWgLmI+f}b7kaFCB6;-^a
zm*?-b^z-vBz9-joL^N-W5!!P>Ak|Vy9?zt3hs?qVQvk*~(Qk1(^Yus2xm(peb(}4p
z30#Q`_!_3FL9?$ATauN9x|J91wEL0k^tx;VPHJ6E^pR05E>;b+=zPeFPDJ~etEP$W
z6Xo)X>D_@@_KQ;gKUesX@;Xlv&ePaJ(ikF=zhaT1E7d8`!xMS4kqLJwS@hoys+w|3
z`V^7T<cOWyhbeww$iJ+`?|hl}z1x2aYyD*Og0Jx5^ptFs%-++Mj|jgfK8p1)iUOwy
zTm2?q9WF$~l=fN<M3VVhUoi9$RwmA^Hl_kVHO!Y-QYYpCz!&>>dYy5BwaOc7_uGM8
z&l3_3U{PaE$<1o8)-zfC$3ml^N3Wsy%D0pF9N@`Ejn$bm_A49hB6VDq%UukD8CZbZ
zM@Z}Unlsj3GCekIyvXN97{Bt{mV-v0Q%a##W~kXIt9m{nwBICdW=smt+-J}(G_mzT
zQmK_T4LP&AF^T#R)~$Y%J4`ij#Tfgl8@{#p(9wgWVy>2GLPkT%GTEkT&>^|;oLUQ!
z&I$Z61dX;bV#)3j_6iR)Td&=uJLRd9)JA)GE?nCC3I!{U6C1OCKM+TP=!xq>ET-U0
zb&F`E$8^e%TSEE9Tr7Uu9(q#V(peDDgJm0Jgkei>%W-v~F_vCe6*`)Y2eCQ}->)Nf
zHs9#*ZcndBt@o<$(3R0GupkV4F-fnKm5~+L(_)mK#=8FH_+eSf-;$DJOXRl*Jvgsm
zrqGqeE}BpD+fUCIRd6(IOc>J+A&k-@MHG^^wU4^jY-#3|8QRk3Fx$MwU^ni$_$9bX
za_~MLe)LQOYFlWR4S2Eh%btIOE<r%D2*i)Ee$lJeO_O-n<$0>t&0W%7IY4<0h493h
zSV=K)!iIz-wn=co&LhQi4Bz+`_^;`C*Dik@Gi)j>i1QD@)y^>bsnkID(?}bSvyUnD
zI)n3G{l&i9t@KYz<*#^zRs#D0#>?q8!tBQ!+V9U^Hxbzm_ge>>`w@WPXV!MOgpaT7
z<@&?=pen-p*%E2aPdf<22vmnEZza-8rB$q(ASeUKM2Vt~w>fz;(LwQ5(Gs#lLnwj$
zk}zVwZ-WBAiw?$Sfrj>OkL8AB=ao;-2Jk^u<Tq2E*C8#hoJBjV<mVSvOJGs|XQCLk
zsIv7h`&~iYL0qS(Th8l|lB@Sckm8`iACThy<Qe5X4nUILjR=VB&=MS~FU42NmxVZ8
zm~Nq7cYBqB&PI!f<qX^6QR=(%C2RAB;zR5nnH4ZR<Z^Daz~7?wW!7P;%zutSe9K|@
zp0o|rlR&e;kv3hs35nzE;!fNHdrW`Yn5wpu!xl>KbhU{T)aWNMfX+>zJlw)CB?H!A
z@*NNIcC-6K@Da|>FkmwjgE9<#H##eC+jc@IwrR&vNb)ovbS%55OdbsCW`w?p98QtO
zvEdaz{zcGb3kwJ-^>Z!`M-Vu?oU4q_W&H4LWK9SDMys^OE3$yIV9zUT5p-@;AEHQ3
z<2+3@{X|4?WyQQFPpwDo5EYf{Yn^9H$C7y7lO>tlgeeYf%D&5f9m&%lYUUA<?s^En
zDHmgPCI*U$C<Ma{qrbvu&Wdb(wwe1)Zm(B0*ZPWF!f-OtY~qYJ^8qVzlj-ob;1zQX
z{qj*)ntB5&H3?fJiZ=&J?j(QhG=J+!8kSF<5edP9e$T9<Y(bk0OOII&5`>HyZ|gJ6
zsN~KU_M4T3vyzs_61Jd3y%Q~mp~{&k*><Qcd21b>+X-qDwXoT8R0@5T1p_e5Zq6wA
zW9Y?rwbz-Pqt664<5-DtD-G*lP3l25+F%H9&<V$2=aZ?QKQe>Za4-kyn{-!m+kT{2
zpJ?ffySC>RdEWlqMlIs3<EN{yh$r=AtZa~%7$c{F;}5$q1uZK+X73g@C8I~J>|pa=
zXUWVuWKzRoD{ZpK=X!o){7hf|Hei6q)B*7P<>asQK|Uxcxi7i9apOncFxGf1zMo#(
zQx^|2#vn`>s&^wGL$mWS#jup%m+;fBSK`%RVaq<3kI?Glmz7sL7fM@!XcE216oz~w
zz(JI5J_VW3&>MZe7vl&NmVA8hHf`vZ_hcn(7##9QJ$V!Bw|&+XvT7S7nv5eCTzr#8
zUu&(u;27w?P?YVu+i5)pAhK2hYew|(xCfe`)wYG0S%<$>DD{&t+3S=6sU#A-65Bzt
z#a%buULdFd8P2aL7ZH~rH{|aHW~m_-<`nIqH<fHGK$U$qiVmPc4y$9_OzcdIPPh=Y
zb$pjC?jME=nvY~?{5QV|>v?Ad^Y<NF#7-?h?%{xQ{?J&@{wtnPU@VfATFC|=D!i_y
zu{l))mpQin;KIg<=`x|6_oX<Cbx5E68pS$b&vOAKYRJpd_yGSnP3toa*%p~J87iIF
z3kcx?8Pg3g9KR&t12tm_c8N*i1|Z_Ro_6b#L)FLZVt<VowAXhvQ-(+Ea8}#L;p9ad
zeG(WsHdln;W!xFFHi`5$MPG!#F9BNVONt<*EsC4Je(&(hY7L9Mn*mA4+5RJxE)D<_
zS^7d}mpQ5#JJvSYDZ25v8=mzLB=ptKu$yL2AXdX_)!HJ$Qkys@jT~~P{3w9FxNvUX
zx0hVl8{8kPK^T=0C|#9ecZx$eLSQjZN6qv2+UK{_B+oPA$E%{S?tK#WJMX&h_)l|6
zghd8Vy6d#*ON;$KZleP&&xS1rf8O_X48OHxFat^QniSc!r0ydvDq804n8e7;x4FJ@
zKlq9Vno8#-w(#B-o-FI`xcB!<`dME^!ax|&lm%Lgux+r&ns$s?w3*lbteckeYQbMT
z)3R+vG3)@Q@F=2z5w|{{Q!o`%pWnmV@5Fy*MZ&UR6f1A$u55&a6HBV(cz#~+x2)Tz
z)3BhO*MQ6%_v00A%OWN%4;jwJ`4Qn^@EOXOS>2c=S~0;hE;Sc6Au5M49wEUbcqK3E
zrD~@iJa!Y8-l!rfCQHFGU}7tRrIa}9Fp__ZH<tc8|2w$9@N_4?(Lq=*_?g+b;4#;E
z5`mg>-Lx&8jhaTEi3WDIa2Qv>a0#x!<UW(rwz7UXX9A`_D(b%7P^e-M@$9*<xr*MX
zGumXfE^<H#RK?(A(ZawT5pW{3AWYia^0N<%%4W)$3h})%wg|UE(kbG;p>yKunU81I
zB$?3hH_Rv2MB@(455MLzZ`7`gaWI~3kD%9`BdEe*CQkx*amRnsR*)dcFUIqrK;<Om
zDt;tzI@lqXc`Vu`kWP@WBYc`Sob<6I!4WLPQvrwaYn*0`38_F8&#<5`_!RnRp{i^@
ztn+8#Ul4KMzXZzH8OR_NapMg~l${~<LMEXClcqEKKRZ;QBDO2#X^H2s7>kW53H{Bl
z<~>tGb|+|F#HIioju8@ibr_G?X_B>ZDQ@f8gd3?QCK`)o=bOb>sGx+opC41?H`E7-
zp;&-z@L?AlHj1?jAYiOK!4<HX3~~ONluqvKM!XBG1t8=PbP6-xpn$Z8rv`rg6&-Il
zc{~9f{oX6EDl_NkrN|BdF{JYe3D6^X2C#R~o2;(N)j~Epj=35Zj&F!3{6CDnWmp}{
z)-4=-1$TFMOK|t#E(>=}aMu6}cbDMq?(Qx@0>Le~F5LOp=Y0F!`@Q$i{oBt|-Cf;P
zvu2Gs#;BfDiS~Kn(Wen^q5@7PglFy{w!UX$y1?GcQg511WMiD-28@Vwe5Laaq+=!Q
za%-lwWe8|}MJ5ARpZ2|$f_PkjjjKWE-%ntA8Mb(EfJnwugDW3C65KzSb%Ja}uCcCY
zXCSLX2}kaG9+%yH5=XDd+`v^v-%a#QrZjjWd?^cL4hw3<I+SDXP+Ak&kaxpvEOqQ(
z3iSi|zf0Zgfgjgt9K-2Re1fvo75cx}KmOSwM)>nRh@CpWqJwPRG!$&wr|i=(279+;
zDq2~?Uw$wFDsGSjx$_c4p+^X1junh-M&5vW)%!C-+ci}diMnsh$V@3MIU)e&VmI*(
zg5c;3w^dpp_xknC!8C-yvDtBp>Hs<nG<b{!3LpLDX#@aA3M-L*7ke>V6vy;HMEL?q
z1s!6QR@Vi+2TrF3dTE*T+wSgLlxbC)nr@EPmmc&8f-|4K>Q_pAq`oN93LLqq3wS6%
z0e}A;Hv4Z{hpa9#XLtmZ<fxRg#fu9f%^GvY-gN(8Xrcg$ijOf+&trbnuDb%2LieMF
zV~}e69E|wIENJ*O;>1+gQo4(Z-2z-=1id!>8(RpN+IF4gf_UDNn0IQ@sz#cj+vAjD
zpG3+kHhD}p=Qm0hytor>-p94cT=tpQXr|jUEeWjHZ=_&k5IiESo3y232M1v5HE<Yw
z#(sN1o5V9MA*zZucVEpPVg$PdIgNO#W+K+Y6!Pfy^SA#8G|D|D&>JgTNDl2mBjD(9
z_0g8B<l>p3ymAUuEE1^-@Ik#ItKD*E#0}IAFHEgPA&Y$^LxRfl9fDr~m4po7#0o8k
zJezw<7?NZBYJPLi)|C6hu1Kl3$qz|6ur*c|N)yr=;br;c<InBq8J44<MUUz2nuro=
zG0YeDl}s>;k9A-Z63s{W30G<b!8Ow2S!1jmJNkhrt9&VjY$3H%-(E}ebm^z({k4r`
zp`CIDUN=}U?u4?1m|E-A(9Z%{wXa<XCwvN_&u(RqG^e8smTd@q4=+n#FGFXNqJfN`
zgwD<Zc@V8n^|Aw(Q_cPkPR#FmZHslxi5?Vk^hW8;OUnm4%#;MTDa3vu;y%SnFNUk{
z>u{VPh%PEX@hX0kU6~hKLiVBc1M&D|uk6z5^m*hDcg;;4z&aeCWi@SI$I9vt4mG5t
z^vG?r2HY)24Pzw}W%O_<8^^qH!(mjoelkOHM0PDy$Nc3c{c)Ovyde(FFCnmah0NF=
zV96LicSLGUIFRKHV!XyjdWJpOrh`r{MnPls7v7_^xOGsUo<vkraqcKt;wCUcu(lo^
zcDu9VT?E%&9$^L35F9#vmA7U6Sjb#Z-@`8pBLIZdBV~<s4I}CY;N#XzlzD_^flkJ`
z54`}Ttm|MlcA;M81lsv+1-?DK7IslD=jfNaZc}Y(tK*6_U&6x!L7^_gk8;ameF`na
zgRVjEIyGi$iq@c(i1sj6o*u=8icepzBf&=z0A92dhVwbaE_+WkQnG12@l<X5o*2!=
zE%y$#NRvqamYa`!{0^&zb}r`VrkCxyDptlx3RDz?CdfvgRmH7(#UxW{JK~s)<g6Fz
z(d(okLk#>*J7}kO)e%kr^c}-UNw`01JH)KPi%nUy8mbHAN2%WpMSfF|HHr9==^pcc
zjPl3*l&CFp`wslF9O(|-ZKk7lVpAkHeT3jyxEKG*ch3I}9vvZuESiEAChQ^kdVIpC
zLcj=d-U1<6_EzBj#65aC7G@*1Df@g3$2&$MF_Uu3y&;ZA{TB@fzhjm_dq6mxQL9Wv
zAe5Dlr;4lreXeOQ(iVWrYjgL`gr1m|@LC{z!Ni!BoRSu+Q1;v}M@YFZ5l-#r<e#hy
zfYAo(`4&nwH%EBd6awv{ejKJ(cD9q5JjyZ2M`cCMeb|+Pa0`AGshY1!)G{{@w8P6d
zh>-jk_XFDoS-HXs&Z6${>a|;+EPIffLe8mFuuVP;e5zqahQH)^hABctgT?hF{i7og
z{q?sJU2DU#k`D8!F*_Y5Q>=uQNJK0o$v4j%TwQtuWCv^3t=c;XDEM^xh!ZWb0DI{P
zOY*Wa8w1LFff)|)6q@bertAt(4zy=rk{0fRg@1i#W<(Xpf$^Mu?zc1?`npGuy<*n~
zznBwfa(};dPmqfrgb)Ggkc=SCZ6QCXIn6hPpj1&in(+K!7Von2Y?Z%eg<b|>K=w7x
zF(#H1`+S&Ow?ylvreI6n=v3Yd54wdUmtIEX)##J-YQGV@;_}uN?H~a4Vrs{30(^Ej
z4oU%ZZ{h*P;={LlwUM_^@fgu8Ki@qgy*Y^7C)2$Zoff1)@35W%OjqEu-_^)mO%9PC
zk$eNE&;7tY=ap1yXe>YO7{?zA)G%XI2V~uHPat|gV*yTm!aCs)05cAvGUD|YDge4)
zoC^T=M_)mJ>%F_!eWxgYZrcb}zewz>`$oh%>$Yn8SXhv;&)aF{w7L$bjf@~&u4+A%
zf`HsW>~KK=hNvl*b=9Dl0!TP@`=gtYgm|?{=ZE9&dsAIQ2d1ohJ3a3_69H*$OkCk?
ztZ`0<7b6uL&-<3wp?tP4oE;R-z$oDiidLy}YddtCbft!07=r;We_;OrBBdVka!P5U
zVp(oY?+ZaBWt(>~QU+I6&No@TinDX<>&9B)A;6mF-q+~w74XDEC6A<ux(k+8BW5*I
zmA^{erFnbtVfS?V*nr*&w9Ab+atPQl4+3;La8=t6tN>BknNa}D&W}1ONE_$FtM=^d
ztvRxU-FJGTJ9POOP0A2M#-76hfFcF-kwsRlP_oj!1~)>&n(_C?p)r^>dpYYC5ZUy>
zWq=8m*y#7rVVVUjbM?vD?2_01?>V+Ct3M~NN*UN=G1ygb$o=kuXd-X41ra)HT@N;~
ztJg=3GiM;sWw{)JMcgbqj=71yvarNzXV%sX)R)kdRo7;j7}Hqr;z?Cuq|gjpVYH8u
zx<#A12K=r-x1p!A#QVG%7-H~acW2nesF4MC2gRX?Li~eRED1yFY8vQZ@2X4mZ!Z7_
z5C>I-G(I+D40%Q2rw3Gn$?a|(aZ;rqSQ`{0qXE~ejMA?yqxEPE@;VN(^{*lC#Z`rG
zcJ$c|*%G}4FXFlDRJb<JHuQ@m(PsE_+`&e$2+udoLOVM@4q2c_R<b<Dz3?wQp|Skk
zUq)Ldyas77C_i!m&JPzm%)hC0*z^yh`C36OIGv%q4We83EpilW)KgA`I6#V`8Z#?x
zXTnDlM2hh$d@a!@5CPg(PrD&WAWg$OaTxxRck`aTg%e^MIxsdfUSGn+A^s9%5pZ*n
z>0&K2A2(WWqCDI-0qZUjiH89;N41ss3>~ox^hWHpp+w^mU6^8(L@)imohR;|U8HWN
zzeM+f6{BRhO=kzL;z=C_(i8^pJc)K<ZjKxB^~id#yX$s4McWHvXXV#`r-qMN_Q<d^
zjx%IcpxoIs06!M84{40tEo3xk&v?W;!Xrs)5iuXViOh2EgxRpgzM79N7rKjg8-M6c
z&>?lCaMUKm5bmJgCYE}DQC$vV#6cNv!6MwoKw7-ZfMjiHoA(*)sf<(%MlRemw!z+F
zMKw$cd>~^qEhB?7*$xoP)jkIhku}%F51oT`9)-+i^H>$_A*$hH!suu3MaC@QwOm)O
zOLv}ww4YuZ4SW5S$+8J_^w~pUynAsa{AxbFkrjl)7+y$nSBTIcLfeVXC~<N?$3RGn
zp$7ySG7jvIg390--~qrgKrcMKJ7b4BiYCrSj*rxzKlm==IWNJrgTj&{H7?j?Bx99`
zwPK)iPiSG1v|+WVmaE4Y6Kb?&dWwUsg{UHCk^)>DzP`>jzG8k;Gm;r6?}=p_DuMMM
zVvlCyQRFSl&OR8)K&QL}e+nMCI7<Ek$d5<auXsiKgE3^_j?;%xxoh36NtT6B;X{g`
z6Z-`LX=_(iM0&b@2*OTIjC#lyi`YH8|GZ&JdCS7kfnf*%a9$Lx{=D}>k%7At79aig
z>h{gO@r&9=L&S#y&*z@{vL_8ok&f2JM+PN>o6q?Y`PsF5Ec~$e!5DN}s>+pY4D<W$
zGR_sExgUeuFEaSO^6ZtTwhpQO@RT_Jgf$JjE5j*OT^u-qdCurhh^STRBrXnv_!8xw
zQ7r-Ht<uO1exitXVaVG=L!Q#YqTZXW4sF(@`+dhD0Ia`-<nMR-aD{?Ki@GCMRUjS=
z_@Ox%$=c)&U2P#+;Y9w402&3oO81y3ZihCjz%RHvT^bXVwaCikYm>}=G*i?g1`-zq
z#RfHh>9MTRr>vSeCRdEV_qY%~YGYtefQ=ekjR}R*dmA_pmPF#rWKNjH%L^+rY!RN#
zh)W=PuP6c>TaGlxhQt}+#X^U%hK=>Ml^vw=`D5Eq$Oy<@&g+I@`X~bM92~UL@Bn!3
zU#`MPrPfd%lJ|Y^9sE_e(kgLFY8-1)L%!nSYf9;O0Dt4MCO66Xu+n_n8>EVW%EG3$
zGET?)Tnl?YX;biR05yx(Distju`KFcTD02Vm&YOmMKUcxFC2YkY-si>%${BnktPQy
zzzDv$J`iPK+#OICyX#=3(C!wpXMF|D))TEdw<#EZh6{xuUo`p91yImB44XE)TtFpb
zH77!a688rjUlTk#U*40glS_8b?wiKF2g>{VXqe;^m1`uZL!vqs5qoko!+?}8FrggK
zcc%(qt(wrUV>2%WOc>&!2235wJ5)1OMuK8-CsfDEn{NER7l6P?Jk{U`)Y&;544~oE
zC*gt7GV-3Q*izWt0FjE6`N3su==bzOkUSshoMdBWZsXx#&hp8FdU9w%WBa)FgcKAT
z1sXililG(joDb{!RV`)FJq>vX=|Wa)_uK!~3E)xrUaY!cut5rK9r^&Z!D`!E4U#kr
zH8N`5Fpcb$4^^9>{fYW4Rq<&tVawXm4)H_fi&tn!z*muCdHk~t2KOS}b+SHXj#X5a
zeE2_rzDP6LJyS_-`4NyiA}*iw;wT3X5RG9EXF7HJF7{)w#7$J$xe(qZ{yE%MTI6Ym
z(+SWY4Z1h|F2-WkA-f{v@%5sLyn$?>f%z-h4v@hQ$!11tXOLeCn{J3g)LaX0E65bD
znN`MeftGQC)-uwfpv@eK{6*d|d6XTSO(Q^H9sFfLnB)#ddPi20ecj_e15wbrLKv@4
zo0t&8Kixq|_4}v2+WU+Xe%u$__7<VighoAF&yh9&th+t<fZeW_B{GT7f_aFq_g*Zy
zRKYk&(BcTdC116(^otws#Cp$W&$+|)g52vTbd+e~U|VG$tLUL|uLKq*I_Rgy!X6um
z;fX0C>?hMulfl3Vnra9m(Q6FoTmI@Qf<l&2%2YJ75u3x+h}S?8r9I`_E!pMuIthQj
zcXJ|{tw1dD5VT%@xuBCiKIBsd=BwPUc_>A+1=wsNx;lKHVFz^U8Vk6s5!`+5p(^k-
z*A3g0GsY`v;>U9x(<!FA^ABd=oZiL6pfAh~GMxF)Pp0C2VDExnv$@{fKG|y;vh0bC
zxR({`C+m?tNl5a0SYQMzUnC4|bwRcIR<=hfZ|RR0N9wU<rF9Y&35palr###mt|>Tu
z4Tf72GMgx+oK5+4JNMuxxop;XgK394D~F6)6#ZceIK`WU+LQt4iB`&VGJeAK7@8RB
zezWc?GX0hF4#F=qnYMvy{Gr7bL{QB1RNLnc;Cf}$x6uzUP<gfJ<WOus9V)U}+0C?C
z21y4%V%U&K!=k7=RKkH4vBT-TJ<_^-4+EVRV$kfkF{$z{2*%)C!PR<1UoL^mHfPUh
zoZq1v#XbNF>!6t?lf$R}W)^e^&RKI~$uOqEHC6=C3%Ko}!KCp&W7}VdQcD6YsIIUg
zWAX+~r^fwrxV3R(?lst{S;6^)$|Q6^qqdxFaF6Wy6V4;QvRdjzT81=-6n2rCzKTzl
z=K~_!Wv?2)`-4FAY)GJ~2nQ5?>t<R(ZC-5{R<;Zxvz#rWB;aho`0eGrP?e6FD8{c~
zh!QPg{5?kj7O{dp{fJpBI1F(xL^`3%B5wzu+I}IvL%^++q$gnU!0jEN;HB+9sT=<;
zwyczRZeiLDdNNBW%*hM<E-4E6m<16YbZ&atr#BI-3E;ZmkC?y<9aVxh=s!k0QjI3w
zRE7AI6W<k!3Pp`umCFok3#|xiai8XoCn#aKohCWHXCU|cP9|KwZvq*gh+N;sdI$GK
z!X2VdARG?cY?o(fM#{{Ov`^?)i^zjZ9}QIW0XaYVQm=2ui38Mho)V)$_ma-c7C}dV
zK)dq=aQU|%L~bCM#02Si^d1vUms~@1NVw0ecCX$J4IYVq*yV!3gpogF28Qgri+8V^
ze-jaE8C12Ah*>3BA7|f^IG8z;bd)@GSj$Or1i9204`+UP^iz%IRJ7ic>4fER5CODT
zq;Vvl+xRkcBo(fD)zc-9U?Rw-rkK5&700oYo;XA7K&)q2c<A}FW}h;v?RSQAo)%*z
zvW&z?9}w-^dE<V><(&GSfi-H7>iUza8Gqk#c@|;xr*)3;@rY<5Ts=Yb9e83Akj|dn
za?nNS2dV5C$%;Wh*ENHrXkP>O@T>TISaYhypQVRs@2NJ#8(ODxQay?vS>zRtLVM2Z
z>X;vpW}eQdX!y2omA{jLOphdZz@mkOCeIQcCXuGkQlw_gTS88PUOu&@$K-S&-y^**
z`o|T)Ce_tyuQ?OBp*6{<=jz&OaQd?Qg4bl{4EOc=ah!9SfZYk;gxB5pw#WVH`I|@S
zxmRVj$Jcasw2@?h>j<~FMNMtN6v7386UH%>nn*dvKD^LRDL=1!tT;DMPY9AgsDNkk
zCBEb4Mg7fT%1)Y(3_X~=KOD&Law+@$9@ldRRN*?171{RIUMKkUs{;3xx|DNpn=d{j
zAgBL<xPW_R7ycs-%ZRCNii8N?fnpd`9CEimGIc6o_Z<#4U|cbC6P#Ze8RM?M$%MFB
z=AHwpDJ9jX9ST<`XnMGiLOeUc>Eny(jc<K{7{$g#e3K0pnSG!A5$b|OgA_JnKpg6M
z1lcA#*TN;^k5(Ie3!TU#Y_NexU<Rv~&%`3MUrgDJW_Z@2RqE9NS3#9{iNzE^CGm7_
z>~{7Z2RS=^s_wu9T2Y??$P5aoH)6l&FczBPqQ>3buWknI0;>!s@Q6)6oVGS}F{piD
z$MPRcJdFE9dP5hH8Lv`CCHYa?o4YwJYH?sa6ku;_;xW7S{U7xjU0TPhf4{pih85WG
zV|v4yo3e<+lFtW9C%;Dj^1Bv*LK;~AAV(=^qHr@niMUr;X4KKJ5%@yD+;6uW023=8
zBNIfDcK>UZpxpZc*C#A*jAmN6Y@BohcATBn5e4((W0pvwg})_~tr_)cW#5KeupIT$
z*v2JlFQ;-=TRPYHjSCt|m5XwAO{wN~tz>*aJO_uCLsAT#6YGHzSX-5F_<%){k9t_6
z&*=g#fdPRj(1+wibF3#@t09fmO*S;w{M(EZz_kMW-OgjN?z$5mOt`yL9e6<q!emFV
z*6aG#=d6oD5-7wUR)cz+iQH7%)wPII4bIR~dYFFdYwv%;J?|`f(e3w|OqC3kLA6*P
z)T%FwmNnA<wG)o81AES@&%h5btln}+oN-VXKelKgj-JDgM8|b>m44gw*nQ>swZ)Mr
z0x%T}-sGwns0>ek0*CMrF$N6OQ{a9|zB;*(;P?I0c5&s1NF27xcI#nK)JC|GR1#79
z!>d0fviyWcm8?$Kz9Pa!v62E?!IOF2!UwF~@n}M1F~wj!ad%!<p^RDXm3Ez^!Y^17
z6*z$~Yv8MoswmvyLAI1pehH+%o4=)c0>xaYf!En=e>Q2c6ODZ$);49qMByF|;pwXo
zO~PtiJiB;hHE+UJRCf1`8O9ll+1f`i9_>sG!P5z&%n^vJ4IBDcC7sgOfqneA=F!-1
zXOgs~rM$$uNQ#rfunES`I^kyNyTz>ABNdNBWB4M{OYfd>?p(BIyx=6~`OEPKargtC
zd9q>dMeD`6A+XDd7$V_KrnCWyiwlbP7$*{1>W~{2>6M1!QNPkidjX}TabHLP>qTxM
zMLV_$Gv!hUy}EK(^#Nqj|1H9eD1nEcaL;6HE;xrO$wY4b@_;Suaz$LkA+EFgmLc#}
ze#KcR@df@f7)suOFn=61J~i>9h<wl{8TOx{Jgxob0=)gsJnvOTc}FPLko_-4%c)`e
z1GGJFYGF30iT5Pb>x1~UyxvVsKRe%mL$FKj`E}mi4=UqDyi>|qO`BjuPl3bcu~_z8
zmwr1?(uEK)_ynN)tQZvE`D1hn?py@84);ov=X@lmHl52qoRMP5uNZ`K;AfzC(}fMr
zqWQ(cmE<+lE_KiC4I&GA3o9w8kn8x|xPK&GM{$lK$)zWTU%`iGK<Op8XC#_wVs?qJ
zUQFgw4Rl$E)i(Q9$_4#%A}L_wCe`Qv<mJOb`NJN9xp7g}=g<9Z&I=sGUi$cbG`53g
zvJ6sE51kpUJUjodMrq`+YTywojm_1m)3gP<MIv(O(0#L9s<ftn0fj1GBH4@0+<X1D
zMKfvLy!iiT8u#KCUkB#V(Uin0DvlpJ19avg?;d^w^y_!qUPu+{{(C1&Czs2&QN-`8
zcTppQBZ7Bc_oL+eQAEf)L!96A#NCI&utQ@bRNDuP3V0vu#-+rHZ9^0cG>f~~*sW#1
zoK!03#_%9NqJ2+sM9`7hOUJ+cBq$3=%bY-czZ0T={rIfpJH`Cl$VmX=JH>n)mAjqs
z8?-T4dpQZj{3TsMbNqazZawKdj*&Vj{`GUWxgU~%U3Rdr3N$571H<~RqYh?k0J5Gj
zHrM|XD__fn_t5#B=-(oe1UgSVU!vm)$0&|M;2i2AfD9K_=sF2lle;{Qz+pne%cc??
zo+=D@?<PGhFzr#e$GuWesexDJFG4$3VuhR!X)Jy9VdtzeKf3I9ZIc_CTs9A;6;SlD
zqF9bE8mnNr0e@xF{ktY$aCm5Ho&|Eq<|d-}7dFXzx9@-dP7{j}mXCeOwTo!{w_D?X
zEm6{GA+0?4%P0rR@(O;dC!TeGvc?!3CVsa~8MzIE*Z!rgMqKW>Q?FI>yVe=?OVZ+O
z>D<^qLri!L^$QRj%J{f;s(Y0@g{9S^zGf&SSDmc;({@D#sJqTSy6w1A-=8RUbY82v
z+4$I=_Um6Bf$svd`0cGEWT0Z!ue(&WqXzb1VNvKYs!V2n^`Bj?<_tUapqiyoL{wS>
z0Y2(#Go>Uw<9{3mUBpQb@ngOO=jt(+50CJ!k1=yWlLLBWcfDNsM9UH*r!hZ1RVYH&
zc}6bBEFRVSmXH6V=}?h+_kU3CO6lZVF4oUa`V6fiA}o44b}REmodYkM4*eE*{w_YK
zTt9Ax|3}V3O7)kiW74kK)b3KSKV7a>kcyP>!bcbZ;vtJtF$=0Wanb=BEB$YS|1`<T
zLj6+aK0oJE@qY^pxFgY62X|P{yuF<N5AngKAz7@7MtR^b+;&BCfV1u5*v;wsOyf88
zgTJ%@yT_dE(+jq)|Dm(<Ho9C1>YZwQ`+dYfve4XaldL&+O3-xXT+U@Xelp~$o{4%u
z`Qi$?j#2TDDE!gw993RYF)wy&_0M+<7ETb!k%SZg#C*0kR9r00Jb^;b)AI^nTe%W+
zc_$sO%_~D$XRguhug5td1;O4N9eKK}`PKRWfk7*ECe3282F+%}A~t8&V6nA_cjEeG
zIDE5QVw90jTkBXE{oOqQN!zDl!qa#0#A+=CY)kg~b1DNJgBwQ3jf)BF9eTx;H$nCf
zws+?H^F-jndjGDZ1kB`WBHGK#PTn<>R+DSD)cH(pRQo1~tx6|kq>H=#E~(}1c^t}C
z(c#e7?DSt|qPUPzME-$g?wHtr+W%thJBWEoIsEAU8r3}FM9RlE-e?0nkVHa<Df}ku
zM~GlAMh<;T!Q1GR5Pavtv$Rwt19}e1uEc;4TDf`WL7AEHaJh|IyY1JXw^j1p5&kT4
zLyLIWFN$)VymVSC|0L*o*L~mhbh9|+B*B#4I=SN7s^=A<#|IAhy4)p;nGlNgl03X|
zu^!8y@WWkr#kECK(8^uNZFF=;0MUla&h71{Ipi+T^maSRcC;+&?6Ep~UmFYi#O+my
z*z#S05tV!wdlz#{@IRiB@8T&XO=#q@ZSY=*?5d8FkW2qtd5BFGNo~u<<!=Jp73QTG
zZOa%f?{g2MqC0$Dk4H+kx%_Ict=1|OVD{oPnvp{mEw+@a)xM{z)p;iZO2=~-i0d0T
zVA)TXb<)om{CcW4-CB1U{@CGD+2-kbiaO}9GSw+4zTf62a5y$luS|cJc{NJESfq~7
ztlYucdy!e$j$9$CRvs5}7&DsZ-ZLH42j~9TQ`!Ix5~pZ;mOZ36KbMYpP7=k@g4n5{
z{jhdX9t?0xfB8pL<{yd5T@nt*G$8h!wJKby6hs?T+Q45|t6K{ADKduc2C3m&aSy`y
zw%)a(a}{WI&#VB~y`C%kv`{6=_;TEQX3yt(9)aC2ru4f5czvh;@teqp6XskNEzjl%
z@&=g&fu9Rn1K`V8(O$L%m9LnJP&UkuBs0Logj?H8t%w_7DFO@hI$UD1M$qIMH601^
z<XFl@N%S^2mkCe^vLhRyH$y=Frrm_<-_%GP@Ab!1_Zh;<RXJks^;pn@Gs|T-tMih7
zW?Q<uftlWZbGWnRjT5|H;NUU)_`?xm;n(%iLhstkDx%vG9maFM0&kg}+`Gvr6u+7>
zaQCV9#|1fAs5`+nT{Tl-3>YzQ^@929YSX8qrMz`-;5d{4t-6|;G_Gb?1$Y#L1sF}7
z>_+LRi-LGOx0cE88~PRoIr%i?)nKAH$yZ=@I`>axL-bceD~Xd)>D)$5>bHI_tS<_6
z{JBt>!huP&+;Dx2Y0jd*b&+nk6@og_O}dVbtwF(PrWqpEDY%buw>eJY4inf&t(hsf
zv@7F|c|H3SAs$!aT&zI2=jd41>9+Y!_iZm;Pd}b6euOhKy<10D<|j8Y0R)M8xj9*c
z?WVG_AFSi7JG~2(PKJ=rPGfRG&pUC8t6m5~SE8@?Mn3PXiMx%Wk|0J*7MDr0HQ6*;
z<LPOw%`8u^EHOMAvOZEWb&;UG6B8*<1Q!RZR<Fa(pfgO5WL*zzfZ0LuJ14OE1`hpM
z<Y&kI9|zlj3K$VpktN{=cY|fOm<FHXkJ%r8Mr8D@)JkCbrvz0Wcy-|IgxtE`b>3%b
zrtYYJc#BNB;R5yUMo87jB*r4;ecqsb_Lq|XD6!MABiU*qg&b?&&A4^w_{O}s)TXm=
zoby2p1C+unWR`c~47$UvUfuNd+OI$321>bfUEnT?%4*s<E^fR=+O4ou@vJZmj#cAi
zf1c8F-qvb%o}|7jU&*@NJ&pOdMlp@_C++MZIkeC2YUz4}@IBU+c=!E2hHW)q@BK|3
zG6&gFw$L7KC1`jt5jp@HFIP}!-FBW62x>G@1mqDlQcsSb_F1lsl-xhOQ>s@J^Pk$|
zpknM<{Axy$LQ#j8qYRtK{FLGSlpHULgte-HJKzRd1g;3<0iXNn06Tv2>Z#Y?DIz>Q
zqR3H%?gdX{JgS{R551a#_8YG|KIEu^rvg~}azuxMZr%ki?rc@vJ9u|}6ELs5ulzzz
zUF~NJ#j9`Y;0iAl%p!s_&vo_iP5f57raMRpgy9GBi!evGNQ1~OReo`yE-+d*r$5YV
z1+e9LK1}~(m{TheHuUDNu<bSBILeMFd#=z(lRz)Gn~j|pD>``iv4QH>m&m7`P8N5n
z)^gg_3qff}Qg^~Zd|zfXmbK|!f$xf0o;F$eCtNK2z18`7Cpu0kp|h7|nq6B^_xPUR
zcQHmaN*B3L+UWz7na<$?&4zRJG_XR*Cm!<t$Uw1+NWk3SZ~f?BdCv&6KYkcRLOIGT
zht^wrgWc|VuwjhwWWeE~1&eD;7pG2N8ac3_2Kz3qq3&nIC!EI`YEgA;CWHN$Ojh%n
zJ9N~dAY;)ZU`dTfSNqGM+o~Qq{uMaf#WqMQCEH`4!jCyqh)5gf@h46du->Y@8g46L
zzDWio{POUc^0MtT#;9MbpBOtx{=uBD0~=5U3hLywslD5-xz}R)Jsd-Pu!x{U^KBJf
ztE1~(%0@C)Z~THyl3s5<$61|v=%f1ZE*iw)%6miWq7`MxLT7iG)T^5%f!c6Ie>KA-
z@h98fBL9527L)ua`mQe`W|EVv6$imV7REJ_l7#oMoAk095ZaMseJZhwiME}{(SLn<
zQn?$u(dnc?IEOd#0BPptY!_||_zzz^!-VY=C^09<SsLYao%+{^uYXz=8_9|L_kea0
z$x#Mr*S-d*)fB9BzRz4DFXaxwCk!cW%q$ox(bPoFh@J{MliGDLEmyRR#c(BJ-5TMO
z$1<e2PDUko^4i<RAdX6|=QOG=5Cv6J_zh~YJntBS@4(8&<G_z~WBG?KJt}`>Fo{7P
z2Y-<+E4Kjw1IOi#K|PA8N$6wHMxQ_$MjS#)z~J8q;*}eK4CUv69f({^7fV&z<BzUZ
zb0?obAJ#uTS`=R6iR0@npf;lJCVw7k7P?Yc<si6lud(awqgP<=Ak;IP91fAyXkD-S
zqW1*oQd;ivs#cs8F446E8i;aZABW|nbBW8Y6#u*La3ubcx$dMb0uOt-Zll-W86UbL
z#B-J>jjxw%r|+@hdc?(=033d?!-tcNiPiEsIdi9y^4C{)si5un%aWUv7uH!akV9Y2
zZTOvO2%`*5^xH2WiBu=uR-3CPLLeauGrlD2FAf5c!a&E~pxpujK_$$A^RiDxEevKv
zh?oP_W+GQmE>DsMEaBh-e}Oev;GaoeJgUcp*WI^J2&oazv|Z1#1)3~I(U(N)J1+#k
z)B-4(r*I15_Vf)jGpr!AX+xJfZh>#v|CVJFszBohZ9|B%NbB+%QMh9P_<E1Vvq(3(
zuTAf4fY&nEWt*a~x8*-kD@glHEtj(gs0lAJPW%L4(LjUUg4bm_%v)+n>Pgb&A3ZGt
z8m*rNhHLEM=<e$aDkidHG!-w}8oJ?ZBOoPkwNfuYxPy}4l^VR!?lr$yqRWJ$BlM3d
z!OHWN8R!l|odavCKc=XVuV5wWbSea%`_DZ`qVr1^m5HmVI_0UjPW^kydlcJ@r6mR_
zplL%J!B7uc#Bv97kshEKItg~bba(H=baOvBD+@=*5w%KV@kw%0ii1hZW6-TU19NFT
zpM||O$PY%}+E8&XIvC0K9s=jdV;@skE1>e6N00*V!E$g2-JA8Cd*=x;XZvCm?+VIE
z?HG~l*Iloa`$ln79STJHgUO$Lv~#9xCTv0l=YS1q3q&;05=^;|QFUeB^<E^0xfJUt
z)ir7izdTfg#*lLTCI=bNuT8Bo|5qB`<1e}y%gAg2QLggV(}pO(S)j>9ay&(DA3FKy
z>%D7r2C5D0{3*}7<|lB|WB{1*P*?g5qi2;-z=DJh^a<^biBI*m0Y=)qgeI4zu!x{Y
zUU=19>ct1&%?XdBp7o31E$9m~2BR>mRhu<N&-j8S0EZS82EQc!K%s;f^JrJxiLfAU
z^s-n6RkzM|#qoL=PBYss3^Uw9LCNiZS+)RBnV!Vzzw-GV;O&N)*Rgicw5FDqzM*A4
zb1i)Eh{<{;pZOkaVnq%EA19eEaf*5@+kiOMcBZ}2B_C%HD2}(_f1xmtxB1Eg^NYY|
zh~8rO-vRjxoWEF}J3nQqF-QJ96*_`ng*)>Xx)rGo{FE_X<n7HsFUvpX3E`YN&`s;U
zD2T*0ou;v(Kl8FY>JG5e8M4zMtOE*`Iu8nJZF5Ym@dZkXpSE_3LpN8aQ^edQQx^y4
zeCzq89m*kdSQ0lC<0>$$R67?S{WCrnfva|(e+tG$imQ#nhOSmO(Qi^hDHm>7K<8E$
zer0suOWa!Dmbc!_ACN(gJJS>!)Itf!DgP|Y$vDqJ-cVQ#$}NTIXq+E54?ds<beZa_
zh`O7Jun*`XoliW?`;p9+msfyO|G<v#&%_^wf37@nLO<tukq9{?!3@4CjsnZzc{zaC
zwwo{cP&3($v_NowxL@*)U6o3JET3iWZ`_x72o;Gz`_ldEBa(yo<dzeg0J}{*9@*K?
za~q6Ge)qqCGcTeS%*+YCK_kd+H%r4UpQusptTNTfwgdubl14&{#QBD#`M#ii{nxHL
zSXd8rB;9c#=wo2-e<5WF=5il#K-u_Yz+#=?&U^NCkEjmV$1f9?EE(3%-hnavh8sZ8
zH|vb{8y?}(Cz%xV)z#Wpm=h+Wo;&_5=@AWbTEMnI)BFM-HCbGMQMg;agA!nSQt0q*
zT4@cuM(ewve!grfMh|A!qy14TPa-sraPi3Bka@f`68l0B;M{l;!Dc4@O%VdyUM08A
z!7%kVn9!`E*pwyjG+jUCMHrR46=mU+psJI!%>v+Z5jijWsNpImB!-D&ubvn6=KjZJ
z$=$YKx(5q0ddVC7qq}=}C)}9br%%c4vL&~waa9U`t6Zzqs9ce&0<vb|$Dzfqg<a-u
z;}n~PqQL5HVKA`hRCd2j$Y+1Mq#5E|y!7k+{qFJkh5GlaNujpBE7_kq;sydwXWmjH
zq+}X7&u&15fxL@%bubOzG}qOOBCpx4zyD&=g<1xyYpgEASSfANs;ceILy9~x$5KGD
zRO#E$!<^~WB<CyYVfs?_4&~uaKBX9;xR>u`jMU3sM#B{Wj@*@%S1Atp%7Wj^>DEY$
z<So~PO;p9hg6`kQK#mAS4Ic|w&}{iWug>KxmX7!{WG<peO6uKn2Vt|U@ybUuWh!xk
zvh2h`-VM1#<XcA46|AfPEl>!^Q(5V^ym$6M2Stv^YO{7$nG2moLS=TS8xvOUw%YjN
zjOR|0qB9F8`9zAvHztCy(>M!CeS^w5Z3m^$Unbs?S*+-N82nZH#sXaXpDX06A%qwY
z7~_kO!fF<5W*-m{0;Ap`pkFv5{3e-j4DJ-|o|+E6DJcmV!V3m}c0EuAY=L)y#gu3<
zI6tKx9(>PnZg)d9v8gbtC9`OmblgtC;jqKTYuQad%@gq#;$o8`Txh`SvSVDP@K@Sm
zw&jhJe+PQYuYiJlcf$C=U3gki>|BZ7R7U}vV~p~XCQK&~d5OtV(l^Mv^6;m!?_C?C
zlyI}1oNJe>#9Os3iEoW_g|I)cF~0^5-*|QGN4V)$&naGr{hkk+@>~_XlOhEg3<$7z
z)3TmR4aT=txa&N3Eg~FHB)Ysq&xyg6{zx=*)3sW6`-I<a&lL?niA4XeU3n}WBu6cG
z+JfQYzxV3@|3N3Q(C@Y=@gGc!|G|X*c3?UC%Qnz4UApOq&&shcO?fXWfL!SvQ{9~x
zf5gj%>8(A#2^S7sJZiKppZNP5utKXylD2Ngv6O1aT5~eVwwg0o0w_fdr_EC|f75G|
z58pF}2l<p*K#O*0OgHMd{Ij&jlE0tFfl@Ks8YEF=eUpJ%-e4RoYz)0wxZ0P@EL%ta
z?hE%9?3E1|*lv8R*f{zh4_rk5d9;!+7ioKWktb%`jpUmu??Zwqtg$YHhq~^-II>t~
zNYDw^xnnC+I3Yh4^-eu2`7NIbbo_<((9Lqtzg~b1l8%meK3YlZS5pGRM*W=n*pQuM
zuy7mT4<=Sy#%*1or{OoL7Ucil00UZOt5NQ#(dM63g479G<sD9i`3wA{XeY7om2ICz
z{Sx^d3^evWsrYwd_U?+5PWm+T+ta50-<SXME}!9{XD{8=)Dzpo<Bc_y)4hxTNP@2y
zzPwMgQ1aMQ4C#il#seYG*{**dFQX*&*P~@P?VOXmZrc=q0(x#-MbXWT<L`epCzq_|
zSEa*v^OOru{9qL@L;<PNV}bbZiQ9u}`3F!?;GP&hJ7cTp<)%M%j{mTuef8dw6b~}y
zq;>stWcd$#`$@J3<5HumAP*##x;irh!#FU*#<8MeU@K7A!z^iha?1J=W#TED;@mHZ
z-)hcpnR+ABO2|<4^iIv~X*}MI{~gX$2Of^yp0F2Ee7mhfK=dC9>+#?ba^yOHXtP=4
zpWfZu<(LWyojdYL+I^i^j~}t{u)5h_HzqOG0BRq362^7)HG1B_F6K+^bGt<m79oaV
zT<T!8MN1^d^OI@iCJd_YR(WWg9+pk0xA3Q6&@b8h74=adHQ&H>ndwAC=(X=nqC7_-
zgNYU@R!(?bPB&c>lUf!$>O|^iM5GMu%+hS)d@T2x)EO4rD56iWefH$NZd;>+U}iR=
zcE^v@k8LKCbX4SAR~qG&9S8Hh&zTn)jt_<cf*St<H;y3u6{=EJI!)Sb=5kF0=CQlO
z+vK`*B4ovtr0uquN1_4MFfhMFE28fzBeuEN?Th-<!%Nz`tI5-5MX<EZrGQpBLq#_(
zS6x0;V*UZze$&_*yd@qnv{BzK31YT$Ur59U6XWndUtZ!X7+hvm8?-NBxqPJ<6Y-U&
zINs86Q|8SPD|IF;x%-UA4<4E}T`ctMl6n3Dq`Qw(p-VW)N;Ae)<82Ay?#8DwGtQZV
zTl0hc_32mkJ5*d%d5q6%E%etop`-Y@hUB;lCn8)m{^v*4tTM?9mh9n|=JiKx)7j29
ze?~vjMC4^zchS`4KgGq)!PRzvZWB5Rq+Gp_JR~M3<R}>4h5E}ZxATS4^`vO6J6_CT
z<aw|36X21_vz&8`8Wuc<FjX1DQtM+MD-BmEZdL_d)Y<da-9S`TL`Cm^20!6C;Im5p
z=GL0k?e+VGSVDg-jif)C97pKyu1-g1^+W$-EcV~IA_<Kw)=42*<7jfiW8?Hf@ja}T
z*?g!?-{4Krl;gb&&|?JZY5T^9l3&~?RjCNA<)w<F-R_{H$E@gn`gmHpha#gDmHCwv
z@Oiqt{U!T`Ix5V%_@k8^lO5%^Q(isni9yuOW*@D?@I}p5g<;e2u8)lfU)_PphDn9D
z-z$<%opqxTsJy{A=i0i5GMt*N$QXr4Z0=siaB)|dQ}I`oA?EGLKJ~i3D~y_5#bc*>
zs=JmJ4_SY{Z9Q`K*Xb!|_uQ11S8x1y_GB=6hr8jh{bhi3aYQz(V;4=O%zq|o9(fHG
z8o!nEE*fFIS549=K_wg51sXbfi|fwE^PG(YW=drdrZBv^VUdPy+}+H$iE8RC=0GHh
zG!cnHerd!9#?1Y38`EQPIq;Waw@<V>KdHuDG7?YgppHqKCrh3Au*K=e?GkpE6-Qi|
z+GO?ZjC$nlMzfbazihY5EIqqUH4?;~!na8<S0onj{Me{;_R<_vVen{9ckp#?vRPnr
z>=I4fnfyer=Mq4YvPeIUbT3)OaFn^*G*xviqqx80q#I3HOt-GJi@mONCc`=C2H0y(
zKCYg02@=?S1Uhz_(njgCj3ssZ%03@s2pJw(=-f^_N_5t(1#^izFi5}fGT@Q;2Q%vq
z*?Cx7xN3@je$In!sNrZ#E;c9=OhP`(Iuf0_*73a8Ri<7_xIChF`TTC^Nd+ol>8(Q(
zF_xEy@_$s~ZCT4z<L#4C#7mbVq5`*3T1i8_YD@G_D<7N2leo;+$Isgquhp%Lua!wL
z<b6quxOv;@PHgWiCl9zkqG4FIa1*~p|9xQ{r1Mzq7_uEpzl8^U5SFO)I(aUbW$LV4
zJ=M}Wlxfy%-Mv$9X`Ig0kg{A^mD#BB$-~?%J@Ao;w&7qMuk`8aV`?p3W$J8P_3?~o
zkWT;Ip_**N$KOT$r@EVcnC!aRL$a!y$5FuL{#XKR?WFd6+$itzQa~iYGUd|AJn7I1
z1NZ9d^3BeqwW|~%g_bZRaCS!P{?y)Pf2+W`V7B>&N~_3`g~92b{QFPu655JuwM3$}
zrg@ma3#S5AO`@9$XIq#1WIdK#RXqjt7%{J<TIE}${!cX)9#L>PC5lbWKdU|JBzcq?
zE+jEW^gr>C%W0mJI}W0IYPVJHX*1!dX}lM2y(Oe4IeM>nX0sN=$RGE+8dZwgaAjn}
z-074Oo2IK<HIC<|SMwMbJ%7F~+8K~8+9e%R5xNZby;ReDu4r{POn3RYU*>4mi)1P8
z)<aORPBIi$M^sn;jqKixZ|PZ`wE|O)+{k!MaT-iau$&Z=qT@OY%`)}Ba+wd19$NGS
z01lh=mVpa*Wz?l(^WtF3aR?Weivri|{!UE&uwRO5EnS)txD~@oJhtF$3VmeG>32W0
z(ML`C4mZ(Sa`Cb|v(~B|(~q&S=Y`}0@8nnr@0}l5whx_HuTO6l=K|j;+N(ZKB2a#Z
zMz8X&B^)P+m#DK0Ur#`nkT#Xg6UUQT&ay52wT9No*N#DeQQvN2Y)nz)W|3b^lh<B6
zd23yeY=bA0N2{FpWhPXzK|-VU{j?(B`UjZo5pm<U^Nx<NkP10c2+C(Z^|1Xaf%DF%
zDZiLj!57QsTKAv`Gn@!LZ;Qfy#ER<SCA+8Yx#b>qAilNZ_l~5co4%2aHh~`okZ-qs
zI;7n-1FPlqYh^40bva@=GG<D{rTb8mNQMwyy06_DX+uF>H%&LP7<2mzZ^{#L{oIRm
zl{05=@vfS$?PVgQW>;^-1A#)fSH0<^m*SWX?`Js<_=iHF!I2|R=p!etoH#RPOqeof
zr4q?{>|{0*YM`pZ-9~XWQ@471bJ<ge?2VB?`N3OmeRE`19GQ?dmoe(wH<=;2M4HJf
zO*Q=X#6{<J5+-<s)%s^W^@p3hhpk>){s!UC(st?eh{?C6<;iZ5>OxlZ<>*{tI1iQ6
zmnu!<6V$0gz{^!LtYN8=Z^k0mYYZf21jFjbG}q3D?(4SGH{$QJ=gHpOdkfxlA1o}*
zT%U&%Z8oxNn+9Oo7G6ar8aw9Bc~Qy*Yf+-4shKXwP~iaJ0dVx&C#o*J3Xw~zyl;Q^
zP@FK7n^k~EI458xo-%sIJoDGXP~R2n&KN~!ux#Wlg5w8>*8|0XkWNw2t2#KWFq!no
zLX?5kQG+`q&QB^I*NPw4MvI?xhegQ6q+6+Aq@VmN>?@ydfSW2Mbpx01S1vspy>Hm-
zy7wCSGu)5){<PX3vXJ~#cB@`ve%ZzeE9Y7Qw|Hk1Hv0p_coMa8pZ&CVt5kgF7C;T5
zmRy?N^aor$+xH2M;)}@#KnnvT;l5`AETE0vhq0d>&UEa_aSDy+Neg0K4EKYEu0fJe
zs>~A$eN%t#r|ps7l0(Un|LUGnBvdNWpZ5bZFI6YHc0y?q+@^~m&k-7Pf3|uvV|jL8
z&PCp~ka!&{tK?Sf%q+|_*;9VQlU`P?V@Ro-7Pa3+mg>YnN08E!Oso}0&hTQ|N%pDO
z068@fgO)2*XxDbSJ8*p{zL+SeE7K*=zg-7YUyN!dIv<i_43WiCTo<>`p3P1V=)M%1
zx?+>roRVm?K5&^&Y!O0q<7O&DS2|n7lKd3&{V>s14gGL8Zj>3@@qy|26X~<s&t&&Y
zL6(9vC$qxOL^xN_-DrQHCwkZ8r8P^*LVSp3IJbM$HicXJmlc8p<rg_I*$IALaC|GF
zN^^GF$&C%$nZIOJ#BJ;!-cj3)WO6N-`CbHB{hwYx|Ij#i8B_eUyz}4!BmgU8vXXir
zni9t}CHLive&g}zT4PGn)aJebTi$)pggs}fm|p&cA}Fw(4{(o6HOrqCZduKbxoM4h
zQ%!;hlqHEyJD19JVMp!x=pRu8Szlnyqy2Qz+~4>e?=A4{vzGmE!8l(W&e%p9#m=&;
z*{W`uJw&M7aZWQVs{Nrw*(!t;emJ_Ja9o}Jm+9RUX)FVpXC7uBtp}{ua+Oy~?hvs<
z{$jg{I{vM3l~DeB|Cnd;aBfy29@eCjh+>MjBCj85xr`|_tY|qD*&9Jp`_F1K4s6Jg
zRQO7@?8A<}oRFiX&M1$w=;o#>Bt<GBQry$d%Q4@Fp*Je87)NMMMUoaiK<Z{Zd^7s!
z`9NcA`!Jj&oX@XZtMJUbqlIBQ*Uk9G)E@&X0j9EKdI^}=co<SQYQJK0dtfM}G+|%e
z8AioxedFwI;H}v0?F%g@M!KAC)qW^E0N5C@av`YR|0K6}AorsuedE>@ZXkJ&Ej=Jd
zLrO^5n6GR$vY9+-KlSCj&2}M)Fg7krhSEm-FPhz`Z|%zNN_pZ!c#Av($A5P91;3fm
za?|Gl;I13bP$bw!5H;kW)0{O?S0RW}DRQZL*?joTAHbu(FSE;F*g=eS&^_EQqC?nZ
z4XR?1av<0)$%inN-LU84tZ{C+amNRBNrRZal`%Ug@(TM@W1u`>Kfd^e1`^NcR<88r
zU_q{Qh0Q7o&seLaL>}#+C6sC=?7zXVr7M=7Tjw_#L=2sKF#}b<rUp#53-nFK4(BU|
z8MU{(mO+70Qs2>D+WAAH&_y?G(P>XZnR(Xb_|jU)?v4C4Y5)E(4O9X%V2<t5!+Ktu
z8ug~(9Kp&THvYZ<rn#$9cfwAp7jb4(*(qm0hyCa>Kk}~-{MQUdbui?#&&ihxm|=V$
zgK~e6QU-!+h+Glxrhadkljf`#`|{#xK6)%_z`hbUTW)n!$muszQXZ7+GSKJ%A})i1
z?Bp`+Z*MUTa4Qme9g6eJ$xae1k~Cg9F<1+Sn@zSR*&>=;?%`yVK;FKhD^<^5IlI5l
zt(G&a1^l4blR35{1V3=7`--DAe4tFVN81XAK?ibVHvCLW`BWt)3|D6D3@Hv3V;PvI
zXLP9^%HQ46^qc_C#rTHX_aym0%ZX|>R6=X|_58lVe4nGEA-#ajO=EAH$?nO98cX?h
zQwTBMtRwl7=YrmnYkhm|WrAuai)xq8;`7EYSpL4EDZmS8%;(DBCjC6*9D1g3fuVxw
z15|zf6vd0}+eY!)kkA}H?&Zz*<vS~+i0&`#Om8@lew$KTR~SXu7?lfdpU%SHDwn^1
znHm<|g|KyfGh5b+6?+hcz}I*!VxIcpDrTi!ka@pZ?8AflXPme}T@}#Yew{y7?fUR6
zzHQ7VGVUh|N7mP2(<oPuzkq;yfPO0DM(<totQC48fp?#wdxvA~_60#K?WL{hYFf3b
zKLmR!8Nt5qjS8+O5!<O6Y6G{APH?`+(v1~Pf13XcnbCKWNvrPbMbH85KdY@cAuR&j
zl(TEQqeKL~Y@j4?<h9dFcC6L59~K89CSFUuNL)<5C!`|6``d@aBXjB7$842*;Ijx~
z19VG&JXt}z@h<uK6~S3sJn1u(ekgB{WIzzi<UqtTRqU>vn(u>cVtn>B7UUzx5eZrq
z``XfDp9_|a8-#w`<UJJzv(B>;^0SZL&JGTS?Y7+7dyJ6lte<N(K$OVMwR?0m#c$H3
zD0L@s@WWqk1OEDxINoxU2RZ$LK@xw~oz%kPc6qL|s+QiKelg&qLivxR;!SBrGs>R)
zXR>(e+8x!dBXUTmT33pDnrWou=g#io9RB_!zEu~j;Ey=w%B4*rR4Lwu4_H_XF1Zbi
z1?P*}R073tcXmo==*vV3cF$u~dBlPwr&52d)-OD)FG7=Zcf{g3CSUK)zPdH*DB-AV
z#8kG>x}hN_Lh_FAE^Rpr9ON){rYlz={OGlJ3|6h>*~MEwE%EwiO8qCz0q+Y#^A2SX
zD<jQ?K%rfv3f|FMTm3sXK>aIZA#Mpvr4Tw=;2|OK8-sb>CQj4LzJ|3CI2gJ*<rHqW
zl$T}sAP`KRwG-b6{`YsPMnH+-n#8o_s?wFIn>2ji9K2)Gz;Ls8iqZ8Z<Hh*vrjffn
zy%U^^pWW6YPYvd4{~Ke(2~!^laZxs^tn*gHKU)#vCO=`-(&LmXwhTYM&J}R5mvR5~
zfUun&lQbc>C`h8DTJne@IbGGf^uI&Jd%u7@ti*=q4M-<)Q(dudR1%G4z?zXvx8k)8
zm)BYyb%(|u`+2`#^L<A0ywTu_=;5MU*UlpO{v7RpZXYaX&McdZT*mr44NMsDPeaxA
zpqH>Zw(ze5N>V|fxTE9TA7nT85FPU$Mc)!-ICBotH->Z74l*zh-Z9Pnm~BH14~~SF
zr(r)7)Xc{s^YX%?;z@G+71mN%jyf-YoUwa6Zg~nfd30n-o#a}ZPQRF<^4Ne{GNm4?
zbraiEO@lXIe92<~%$J}Yk?Ksxs|qb8I*i2LICiRak1~&_KUE?6l~~R+zxcyd9Za91
zA)v7T_z6!_)xc(vzBJ}@nr|ap<@>JrWv8~lCs7Y?G0&6jM&PMp7JGJR{w{U7v|>Tw
z{{v+}n!hq}jUuU={VL^yd_AO0>6)6LGL=`ENym%8FQGp6Sw&Q){JN$VD@u0IPRdFt
zE-pt^;!Bs>SgUq$oqW{Ievn)3u(GI?^h%>WPO%SBHfd0UvQOU!Y6|M|DpK-tjq+&&
zeM*X_iRvm{p|z^d_S82WS=~CxzON^x`$hVQ0uIzB>fNP@P4exSeJiO(&p^b1QaIqp
zmQvD2X(pEgGeJO}Tn-kl9Z6=o>%z56vHCEq!B1fZ=xjO1f!Vln`G};h3JKxWjdP0L
zjETGumw?*~SYX@MAz;yL(k5-pboaRaY6-AS4K60(5-BvJEdYY+6`h^dR5K&1t&vb4
zfDOPzmCDQrc*N>zU7uoaLzVGbm#J2nlF&0+EFxAN51u?-tB5yf!MZA!Nov)-stU`O
z5l8{BAkkOGBp@q)WlxWIS8u?EGGjHY&@IrSwjtyr@K?`i;=ol3G*xtTxw`qyF4@uJ
zY8&fZN&u%%D!B@QheTtofs7sjj{ZIkuL2B!8MXm<pnA1MV4}A-06gqh3&;RW03se_
zB&sW2$N5fI+nmq>g=(+b-3Mq>c`5OwM&-ajy3+uEvM+0o_|TC|xyA4)`x4joi8ty;
zZDq{BfPXn_Dq;?2E=qzp59P<wrcAEImH&K~_^pHbx@yY@(CksW8mp>Z4@<p^03DR=
zkJIX0L#*0~Q(8nQ=5!}jo}PmO!z2q0wfFS8d9?xPvjU?(5eFg;L>w4`1AYt}B1v><
zif~}?wAB=8)C_0_lXugdIwa>ck0E$UIXTeTlX7*55sBi{o&Ay{r!fIKkW_2uHn|$@
zC(&f+iO6tXR4w7TFDBs=X;#9j3FZ=B<IRl*So#D;`t$&jAXX=hhy?HI#)Jv%NrgD8
zlw&>G4>Tc}q=qqRHe&kqrIMOqM^hpIA;CM7R9b<HDrs0!9o;HU<w@&dAP8Z3PMvFS
zwI*4;)*fkEsvGK5UZv}m01TzNgKcd?t+fRpfTjkjme3zp`IcA5y=Y9-uj(2J*y;yq
zYt>(PK>hSBlL=mrYJm&V)T+)4^h;m@UZAB(_M}~n>dT<^5Ro@t8$1?O0xIOAqPR2z
zXf)6eTu^%{87)IwZ_sA=Uay&QSluB25(lY>Isr!tj+9pURoQbj(rC1ZXUSNfYtrhS
zfD!;gQW~8a@wYt{lu5gj;va2*PYr+}07F;GH42cBt3yk-12(z^TG(DtUit|bsm(6R
zWBV*;o+SXug)V8ZKw%L@9EdnDEjhs38;o$%viVV?BMw9ym?RD`@j6uqIcK)fwC=H)
z5+OwE8OrcSKQ@&TQn`i{Y{DfcbsT6svY#rQLJPn-zFwwvP(Cv05-04ufd@1J5~?dT
z+s;L1Qu<}yg9fBug0Ydxz>G=~ydCySNGA=)nDju@0nxO8&`_Wisk9$;TDw&Vnhgk$
z;AtQqT8p48-k)r-F0FRb7s{|Q5%hJDCXvR{1=+IcwDIL<&-*sA^+~CpeSd7c$~MO8
zQW>^Y<u)7Q6J{HZ76ihi%9~16DUUzzYeK65(M~{*)u}(82Z0fnz*gu*<-YUJbBS^R
zM1Y(?-!X8_3eWzrGJU4-g=l})a)9S0qUpnd==GgG_iDDc)Q>H}*D4xkCi{8ta?WJ8
zgZfP_n+gq?z%~Su-UL=y)~XO_^<x9s!=$@dEsX~PH8&KW4NIe*;xiG$r^84lO$e!)
zVK*sU5_}AMf5rs;>|xN%yd+=)7v@_btq24d@dHSvmzi@kE%GDilD0$pWX>>auQKt=
zNOmTzjb;`_!M?!;U{r<LTb1=S2mpr!XK7gsC}sOc7HgBWK;^4kjxmdZ*iWTE7_=`i
zvjSKlk@qPr7x-rqdbWP+zxqPk$!}moTL`d{WovLE`%UZZbN~QA07*naRF}&DFDqMH
zRpsL1VYSM^^od_k*FZLJo&|6Su1iaW#z%P|&hd|O>{#u=R|L?AFAVLhQeQ|%AL_&_
z+6*ys(Tkr+_P9<_*aQ*HBn;mVu4TJq>HQyTf1wLLLA0H99N>A0DB?iGfrta;=0GoU
zfr#%>Pj<iDbwuTr&H<aKE<JgavdkP9us<nt!4pwjTDEbRH{o2JY?BF*1kx%wU#Dbl
zl)TVkO{kQ~6Xs7<fQCS7cFETEpm~yAF{hF|Yv3X+p9RdOpbCIM7B~PN%;B8#5{`5J
zAClJl`sAm9K-?))qJ{?7C;t(`53|MbykBNV{r%FENRqFX6M0G|J{9u&z^n_0aU+>i
zAyOxr>+BG1Mz2hw8g$&*k#u$S>Vx`}Qh%$}UkHsxOh5{gDr9bibA)hyXcEvEplLBc
zgsGB_y8&S$b3#`d>hZ`QGp!!^kWfxC8JMg!sIJaz8Z-&~y4B`-nPSBR%K9p$tx-9c
zZ#8KVob$<~%64Wh<zMWC02HQ9w5dyGQ4o2$R9B<Q?-2hmk@7wuzzH+wlGzy!`g3Fk
z2BkDuw+iGn)zrwuO8>I`Y}PS@IdoorH27>#7XBkC@tpGLFD6)!b`9_V(%@l45eGa6
z@_ml{`Dhz)AmTv8fw6GFoBi4t!$&EiN#nrrljq#hOXg2X{Y;!9FNVomuY}Ld^!;fK
zJZc-_5}p+mN)F;l38FY%;}5`0c}-Kj&HUq3-X}p9%z;o1)yU}_O+maq?s{6}m@dDL
zRCm(V>bPG5bdQ96{6~n4rA&cFLx7_~v-Fg0;6!CY20BzaPT5?mtdo<roZIahGx!20
zXfdeYfQ^LbSY2Ik&P=1w9s~lg2M&})+wkK6bfAquD+0m*Q22gON2Sb;Z~{j|5ttyU
zugc(kJ<$GejF4TeAN--7d`b$uQ7)PQ10(8Qhk#C)02{(565a(mB<$D73A`7laiwh+
zPynd#P7&b>EeC)De-3z*6j(@P{VV_mb87)xGQG-bIT(;RuTz${xvs{gRYzY|_0lC@
zlxE!<W~FN>Cs2-9{j`z%@!H@dbwX%Ve7X6KWF7Jr@u15RPqRNQ0(o=efd+1xH9cx>
z#DQtefkMK=wC>Od8}Ea6j4)f2d<JmftOD*Vko+>83D7ZF?{GfofR8zI_vy^k5w?9=
zPx<x`@d<}&nKA(ex^#>%iNiG}Rt7rIDoB_$VI2us|5Kga0tel$r(NmAFaSZnv=f;1
z06J<U;HH%Z(<`D?;oAY|s057YxI+FOG0m(SfROMVlY?|$x>tGHRc@EspqX~%$z&a`
zGcv)VYVQn>Nfwc>Fw;_Ei&&LbI}Q0Nop<!+UM9PPb^#btiu6Pq2@#wUC}InlIhnT1
z;@dTC0T(#j`|CKLqlGZv32h-X7jf~*6TG7y12+=h!vF_i9}^}bN3rDv@}?<sq_F`M
z;2fUH`kojd38eu5aE>2@mI5OvzzKkZ^L{Yk9N}*})OYR|bvLUHG#He_b)3w7S-$=4
zx(B{&w#u5*esJtHE;yG{-gKr2bWG>F7qxvlbD)s$FrE7|(zX|h5;)Q_qV&^~1JfOJ
zc&<;ScNp1hw$a!%9ug2SMIzEuM&>Y>_OO;lr9zdNGxj)g)HD_{*&&pyoCeI43gH<o
z2s8ZVFOim(KoOx3{B=pB<AkC(NpP>1DN{F5o;2&;r`d55k&ft6F{Q!JA}v!Mb1*lv
zDIO5z*UWxz)?5qoD${~=2F~RD{W5`4gsXlt?Fj_@Sbd#KcXTRWv!Y+sxVXaOVoC+y
zD7!}HVGy9AR&B+MC<uA0ZUKU_S_;<tpf1x&D1CK}v>gH;jnedV!v|?YP!PZa9%h`)
z&Gae@Koas-=~kVY3SF-Z+U6bKo1|gz&fYEAh)?qsb*oi?j<N`Ki@F%ICCE&cq;?>h
z6dDe|Lo;E9Jq5}9k!=Zmdl9J_gwKTl1b~Y=)x-XR#K6o7QLz%00iR6cK?c=fS`zj}
zjRJrW-~`IA&*wL7jM^Ph`8dEc*DIji-PI)koYFHIM1ZT+>{vX~%#@GOQAwlbfS=eX
zQ~;x4fg!6<(&*d}9LOIB__kK*7mWi$FfF<?t2q#@<1vEU&j6F|YL}26tcg)5D?)Wu
zZ9*Yo8w*_?`#GTm+JV^=Ab~AvE9TDqSA^qqHe4-XaWPrySD0C4lP`m2)&Us;0{CN8
zY4)G;&~)@k3u5!<O2>L0_=2Fpz`p~Z5koopi!TS}Rh1H+0UfOKkr7|w&2yDc+7SVl
z4B=;WJ;<kerJ2Ehgf`-g-hH9Vc!3`ZCQ?wZgn#^1x}=o>SRwqI1G?}L>v8nR^hij-
z)u_!~GKs=m3lIX}q7NV_KnGxi^L9?<0xr7MS6GCGq)vd1YlO<hWD7GdG$p$1E(mm}
zIQ=BZtE~nyvcLm?&@S_=I_j7227ViW0I%6Vy8;L$Ia(5~dq6-s4|2TiD}6>VfP6jY
z#6SE%+H{WqA)E)4cp|;e!?`Y9M}>z87&98WyS9c08j>7X(REA(AkPy!i2jU<0|YrZ
zfBw8{ZD}c*9}v>i-0bErSWq-gbRpuvcynN=S<VP_jE(!zI50N;MY$)I16W&Nw=or{
zLzB>T4Fht`RA|$9w_#MhP4{%V-j23m+eWzAR9A-t|1KAAnpbuW2+r25PTrph{3}?u
z1A#Z8ZAWKdc7w3qS1CbP!Vk|GGC5iU%&8!p$V~v1z|FN9nc$=aNPKiD78PVaBgf^M
z<_4E)>kQIMkVa0%k=%Sbbd070&;g}8w3Q~Hf<jfdJ>sf^N)X7sz#LQgMB6e_QUC_m
z<I-*rjS9b&UbTmHJh-MLIkG5L9Fb3bvU~yp_(gDyepwoor~gKnFc}~aXhEw20d@c`
zoQG_U5N&Zt_jSZ|y<6$RW?-0Cp`Gaz2x^qZgxr7%wg3tQd_MN9=@CSj3Iu+x^XpYR
z344onqg6l)^QTUMG-d$EH&;LmjSGb1dOaE_r41A1P#X0Da>xgddIdHBH@48Bo|CFX
z;CDiP(sTd@JyWcV=+8(waG|x;ojr3Vcprwy%yZl#aME_+f?K+Ld87duf>+U{hy&A&
z1JO7z-R@Y_-l@ca+34djV&DOgNM(iVQg0prq^I?)7OHDD4M?f^hD%q8--TrMS_!IX
zFECX?n8XnoZ3{waO<X<!U5Y3rfgItSA7BOlkZOcwrDvAk{4cT<#?8D*S_1hi#FQ`D
zCXIps2HKJoX=N^jrbC&7t@7d1fq5004nRjAt9UC~6Op!{3TZF;<iKvLl?#XfXz=q$
zOH0E-f?f!&l?FmhG+&Jkt5&&{%a^)6?;muh&!4rGyD1B82bv9lh$l1~X@QMyG$C1d
zLYd$moX#<kGT}eVBm7CYP9&=!ouH7?#$~=m{qQLbATdy+bZRH%;y?~~L5tBQfxbzO
z{Qw0P3IrIS)c^zlI@VlzsT{;BWj-de8`X8XwapzmcHA|J7ipC-r?J6p*>t(<lwZuY
zUGJG`nB5Dd#k8tjGXdPUnA%T!(VCd=2leXS0j!eZ0bzJ4ANLa4*{u7dUTH={4+oLA
zbkDrAJDLvzA4*>nS`WnYaje^w%a`eS99Rv#OWLxd>hnw^fW*18XGhd60Lan9huxyZ
zi%kPEqRdKc?}+ZXZu5lF!uj*vlEsT$Lw&v8m!vy(;)FYOCg?kDx>Wak!Mu5T&zSJ&
za)*u_)wZ}MINbO~nQAV*WO4YE+kfzoO*#xZnLVggP==KPZ(%xq@|0`u2%`9f&zEz5
z*^*1F|GrGj>JF-(p5hu4ue@xNy^}k3?J=`Eu5qt;-|!2YZnT8wRAZE<lopdcaOki<
zjJD#yWQ+qNW&E0zm+IXN%yh}iyLs%y$()A1v7z2wv2mk)hj#AXJ7{ukq;-{&ej=Rn
zSFTtt|JMQ2=Pvn`9z8zuzC)$(yHFbZL&H`Oh0;vaIb^p&#Fvgq(wRsj0v?!EjTCf*
znbzBJ!5pzG<ZDof?!EQKLZ{t3yY{-X=L1s?KJ-^@+ThO1w4<f#f(fKJW?#Gca@XA0
zC?Qj~O7{vu^~KlTw9v5JJIshYrF!)tl;FyFgm;_8*FFL|GyG@*2-V99;^r?Rff`43
z$vjBp25Mr7gcXO079n;6M{=2c5z$JT4nPN?cKce-yM#<kk}VgMPU}x7Ed{Z_AfIK9
zP2-}nG4(RgAu!NbTj##|xlc<tu6DOyf35p}{_*S5Oi1&gHsU~TphF#{uk;yns(dm5
zO?BZTBGrQO%OpxdIKDVlMCj5r5618t!ECEn{a-FYXstklbvVmU1bd=lv648ei=(X|
z{|n-AgTNhOefz7WVX08tbRLkfY27+^#fA;8A<L_M2M@aM{`jW`fOx6yzw=Jr%T3m7
zj(_{scDom}yGwm-5C8;7)WS18GJp}voiB4Y%&_P;rd}IYuXgk2&DHIMAHjMbUCD0u
z-0N>CZJp8>Cv2c$Fuv^ajndSp67i08KcJnv_mnIwT81Kj)~{LZuGqBEAy7g`jvaIV
z^!*<N_qa%&>2`WF!bZ4{r%#<y1hU$aG#?1rE0!&F%}tGZmwVm$))u#C{{h!2Uy_NS
zJ8!wk-Lmz10s5pnd+wZDb?FLwMt}ITpS#0HKX9LZ=;LmcKJq6{owg7Cyt#AKwr=;|
zzWxmZDZ{mKPIHsH{l**IB?2~{{_^ks!2pau4sR107q)J>+Q1I)GW6IJzi{vE-4{-l
z+vhpFeB*}P@lZQE-rMB>%7tjP#)B=JuW+;%IxFD$ci;F{?zz7C`s>|oH*U3S0JP^{
z{+;dX6&n%gSiW?LTef70<@J9E*74@s+f9?}&#`4JxK6zAeDK!}A3f?$o<8FbN4;ed
z;o7UObo1uSu`=4)+udLNzhAfUZPbO%pp4@^4gknkEL&C*$6kADTkcuR&#-CzI{OwS
zvOa@*-#_5K|I?r4On*Oo_g(HXrQv&qHN?C67L110Y3!drckZC9J+hS8uJ_8m@!2al
z=htccyYlkOg2|Lnj(2zOcAc8M7%Od9yGCG2-(J0o&>_7O-}=F0V=Z>Z<VRMRDx}A*
zAWfw<7Hr@cP8LQ0g$fjRQi|F%R_bXx=j!GSp5gOlE}TElZN2IWdnq}G0Q=_loo<%|
z4Nt420r=XZpK-hPz3=|&8{aWM2(Hyjz`Xv-%Ox~lqM7qP_v*HHTwJC)-RCb@Ruf3G
z=0x~a5D5wsCw_nm1bJ^(#Df4ZNeKV|N;2~IU;)7t(XaFm$8=lQLkU@@146@MTb$8#
z9!&?J0|InF_<Lvs66jDlDpx2h04yflYsOw+1c!OF8E7Zk<dD8Vg3NiDsB}qNf<_0z
zY>Wl<cwRBX0<Zv(;xaJ<aM6C+ZT=m~!=Ojs)~}Kfr;qw8+qS!I14z;c2~<%=f4{&B
zX*KXBB-k&qd$C#mB*7zcR5L`~(o6)VT9pDOE9}}H4bq*8^fh0mVd$9f!@wjfp*B7F
zn`hnr{RiC_9{z+4u-0BP#gbW?z**acz+5k_nZyhFsMBUlwhVNL-y1cM*T_d``+Ga>
zoR8#$&juO~o)1F);=FvR`gNH=4#EgTnMXA<eNwaX-gzA#kj-m0xsTm_x253^f)=w}
zW{Q_PJNlicUwF~&-v7S)@*|(VNFW-0sZ%E>_vpL?@UhbI4?b|K*RCyE4A0%CKK@a+
zcGXI^@4!I=g;!m1nV!u$cjqm)xbOVvv7FYSXdz>~fR^Ld8@AegM6h{Pzkm6)ujn1?
zbIc&#aP2j2wcZ2h@uz;}8ugL?{HH$QTJ^q?rW8UO_Vw@nv-{)U|FWAqr+L^c=VjjZ
z?H~QvJ$V0z+*J}>hNCKtC3BjZ+-8B{I=$D>nRD{VQ5zwI_GKC~(WD@-7XuzIzWSQm
zy>Gw!!!Le5HzrsC**}!WvjMfV24!b2*#_{qWYHqG?cJRkXXIm6glZI&WPv6TuDarK
zX=#HomGO%<A3S_yP_CuY!rXuNow;+o^Y83@&mBK`GIwq)JHLDLEq!P9yRUuuiy9{f
zX3WP@=1j>u5#vC$zL$W$Et*vH<1_abU=MK)Z6nU>Z@sf!ksgC@aXycK@$_#rS@FKT
zoB7u`ZqYjhAvf?H$|ok1(28u^vD3b#`Pcn1lUmCJmM+_{J_jgxCSQN^9ka6V>DVq^
zvBG`$t~+w4m`HhVkG@0rQ<i~#{oD&~k7T4Te&$m%@;fwE7FA3B7>+(TdcrMVF?bSW
zxICk|Zc7G?CZqBkJ8@PMHPg>`vSi?ql(vEy`mquMH$ERYo#)f{9)Hq4Fb_ZQA-i<!
z#3}dnAN*ucva2o&X74tvUG3)1o1?)ZDS<fUe*5~H?){@5xPSdef8btzbDR6oQ%~#V
z*F?DlShNJ}>-%L?$4t6}Y|{kj489@V=OhFxE$Q%gK#<4$DxtxzN5VNDzgk^y7f5NG
zQ|CG^1cM)o1DX$r21BA(;Yh9mtRAoFutf&-pjQxHmkRtZ+aOHFbka<@a2%KHZ?mQU
z`S1Tf*CZ3K)=sT8QdMEj?qtD<9!GYYMNx5>nE|YlGPA0ylNpoR-j~j5I<m+fQ1ZYX
zceu4uJMY|gz{>aGc7xU21Hi)m`$vwr-R~cA^Cbu&{6oL_-7D^ZG!~xjmhgAiO*cuA
zf|dreEeSEI3+Hbf);%2!38_kwi!`MkxMyB|+3k35w=^-mCRn0z@dE;yo{!#huYsV~
zw(rn=7%0=5OQD6q^lQ6jiT1w#z6SjzCP484sFL7%#oG0*^<3~#p*;z;jhQXXWxN{t
zy*>NfQ_nuf<+2eU5eAd|nb@*bFo^}Bt+3~AjBP+L$2Pn<JDZ$e+`V_)=J?<vNIvna
zr>#!*|HbD&>*_Ug$t=D%cgoK@l4CvfdTMzxFQ(;7m*(yvrcAW?Z=}gcNppyha_6l#
z=Q4F@lHhzxtNV+;{RcBM@<Mnn!4Nw<D?Br$01t|R|I}9swFunvd^4t9d)1Ye<?S8s
z8i?}wua=MxkRd?EP}HLd1_UWT84*s)y*Z@um$so92oqz|eRtfh59B4LnHdUr9FQsA
zL5)8PH1UCn8}BLkFW00(KJegOPU>CuQPU0!%s%<dv*D`8yifVPGNFxt$6!`Y#5izR
zrhNPs%%7VxFMR%$SKa$EMf7yf?YBuId7Vwx$)>L;wGZ*WJonNoHqM|?%{P(6gz$lT
z?~x2q<=%exy&=cGcX#bJfP>bqMskDy4&8a%P41O9-W=4n>otiIes<Bg|4Lw?oYc{k
z^+zpdp);f`CU$+Wn_^v>2+%>@)h7QHYsJv#rz7Z~N5jKD>$gmZWV%u=N)HhU>+SdU
z<X-Oin(faQ-Y~O*m<_vj^JPk_1LOw%&<;TF=p|0zD<P18K#Um@CJM~bZ{D!ZEnm9K
zOa{2d3_s>Z+<erM{Av{zwpO8A>or5zBa@$m)S?ZEdJQIVSFdzQh0Y}%Gn|c@fvu6|
zqEa)4G0o&HS+vkCkU0VZd`{CL(1E!?CLM&(t(6eH;*v!&Lt1AaV}y5G9i48dQMDSh
znp7VHSC6zT)SnWNfM^pln9U81wyjgW%=oNdb*bCDalKo-V7>_%_(qt3C<XxpZomFI
zw^}A4(EK@#ZjnG6=3qFIhi7NQLS=7gW)eQPOS`}-<@^px$UwM~U%y1R;2)ZZi~!Pt
z#S1i3uvES<^7+tlxJ-H><%(;P2}qUt$1D|Pv}ub;RjW)|nDx2)*4x}%@ezK^k(o`6
z?jLr#_>WKz6mUc@uJ}t^`lQL)FTusPd*k}G*kk1KRLOarnO97{Jf+fwXI~Mn5PX<%
zN9FH<iq99`@=;=Te3fRgP@$U{oIRM~RD<^smMC|P%Evs1eol4bK6c~@m1Al1P?4}O
zkAN`g@y`#w45!bYv)TGxd-q>dHDGh`!Ub;K>eV(A>dg%D6LKCKXv`5X<*Lt`0{V1}
zIoS3y{&?bkHOZ8-O7D71^|0^l|7l-?CI~2(yBG!?0O&<()7sUm+$EavM<5vuG2@Dv
z7ta~b3eWmTXeeR-gX1UcJ@gs|UoPgDi#491)qzKWS!STs@M+l6HcYzcH@pZ7>F^S*
zk@sUEExbdZD@rN@-#4^DXqKRjx>vNv6XkALyVkT}{#+_mSn_0M91|GcWrRX+a^v%b
zx1{BMVOqxhq(9;N8BXi>xyP4?Pn6G?;b-Ul^>S|+Z6}l9$wZ9<^t0G~<r-Q#?oYm^
zfO~=_kM}&TcP{*mBOk<j#IwkKrVgH|urCPv0A}9vAZ>%CF9f`U5`%Z2_vC=IbN-!U
zETe2syaPOk`$7P!n6`#*%R+tAdA9I58u~j_ZD}z5#Lv*bkG|eQTgqE1K24|ZP@#TA
z=cX422BwjRXm_7xy2?oznI~bYRLIxkoMsJ=%RYCRKCqbltX{Fqy?^+qRgd`(g1x7$
zS8kTs&|3lz!5M%Bv*!nn9Mh<xnM?_td;~sv&mHbI`AGmSS|vCE7@)T_OZWYsKPB^_
zK6m3SH@MGz{DDD3`mvLz-QRrYJMIr3dDyL#zzP9We)!8@xsORr>OWxH<xKguevf?Y
zelz((1HkO(H@^Erci{L*({unjFnf65L-)A*?!Mh-4Y3=ik8PSw{mHL><KC1RLPuNL
z-F53#?&EjgmHW^#3-jlH`;V@#y3aj$?_Dw}+nf_rAAkB;fvL3u9;*U@MA{De_1&NS
z!rF^K^sm11`);nliWhFb^2lebpD)XF<YzL6LRdm;!jAd8^l@e|;lshKX?<Lwe6_z%
z0{;K;U;e9ResSZLhDLnZq5<}y`#+o$z%bi*;<@K@Jmqtl60fTzy#3LmUoauv*YtxY
zpK{w|Qnf;Z;MczJSp!?NfDg#O{-dw3E51MFcJ0}%(7)1`WMvTYvn~nR7CL}g+6OYR
z!m)j^d~4>)Y~!4SRc1IgUb@l*tiy7$yK+<DTSB>E!c=SXWt-gHx7}jG7TOVnPCg>P
zd*dyK`31ASo-WmM@z4YJ+Xu<`p!Cna_;OKUg%8k8*Inywy>Y8)Whj>q+zAQQzj*o?
z6RdsNW7&?!3bUPMc3`IOz2aAm_!}m8gh>^2)#aCKrg3Lr8mQxc`u-2?9_9-YfCz+y
zW`R|-0X+A-hs;7U3vR-R#sX&cH?3Q1f&gX}_=TapctfU{&%X4s37Yey>HEqTzo1#v
z0YMkDh!5R)yZL%-lg7YiS+g1f-m9;E=}~L%PMJ$QuJM6BpkcV<<{PsU2!WQ3=jpe<
zd(FKm(~#i=S_-5bp2e+Ow%AyP(1&w7v-{;D%(wpHZ~o5s;L8Lged43{o2d)yGT}`A
z!tY)+K;YB(?dLxEd-AcGXKx_qUVi-z^YLJO{Hpk9fJ{7se*bfi*!HE@-f%Cx68M4G
zi2{Fk&-NZTV7?`oi2(*MyMr*>YLq~@@1=wGz81)Cr`D}rC6ksr?YW>2j3+!7XnmfO
z9}2V1o&XJ>_~=Ij9u$F5eD9TJ>!sIT&rM|bbbiZyM_a`GLD1uQrhLHiQ@?q>toO**
z$+nn=E}UjO_a|x`$Pb@hs80y^|M{=~HYenh1|gd75Sme_OJ>DSKmUU5hyO59@dsb{
zoW4T?<Im5Y`nCCZ0F*JCWFiSdK>rh&Lqa}{9TfS~qVNBS#A%xV0h9uK&t$d#tMnYB
zU17|}1hNp}zlaaWUAOQZsxvLUeTVc+qCI()?~ukVPfVaZ^uUL00>HjQk}aNj@kM*q
zeVXu==NsQOetg$?*Q)gFGPYovc2=g1Gnv}7==8#di>+=ZqtAYAw8^M@vg(Xoy77wt
zFkI!BMrDf33=fyL*mVZ&m^285h?(wnn!$(gbJ%m>knP`e?bSB3e@x~c2n@~o@VvR>
zJu|1_gZ$uqcezs%GGBi2RoAZ<3tx#QnRI>W;fL&l@P=$O|J!%|Sz3&wyFx-TPSj^k
zoOY+=uzG0k0d!t&lT5gHDUW<`T%hB7uA`;Hz5ezNGkIDi5cI9be&#6W6$vW%qu^UI
zM{QZ98H|7a*^?4rcF4wat!C?+wOuDc^=&iF;B)cG2OkjdxZ4JUzy61BxF0_LbJyC|
z?zUcYwK>cm7tlFz@{But=&0K+weC$@ueEmfYBuKC7hll_FzxmnJZQ6>m#tl6?ZoHh
zz|mvww;GI=EL>;;*HR61fTq*>0AiN4PlNL^X;{#dK>z)*pSTxZmucGJL#{)sq4!FY
zVWEfx9*7WznHiw%T76W}pg?Gmj;U=#vucvbnJ#N(amAm;M?^p*u6y~%kAB`9$r0$j
z`Ph%$J2GSXy?gJ@36ZB2AmCN`?9_<A&;0In`I{^@Z~(v2LYzH!*0sx*<;Vv|b^lju
zu&=R?{$G9bJ8t`~o$k2qJKBs|eK0x&ln@{QbFA|b*S)(zj;0Kl9|%}%SQGFksgDj$
zz?gR3dCSf2oUQ{3aeRlE?G2l7T()t8`#k{(W~P7r{BPY)1&9tG`9KkSHrlK;AAKD4
zAU*_N{LI7lEE3J=`#*i$OiAv#_2%FMpl2LjY~Q)d((~tLX`b%6{Z{iu`HpP$pH>|!
z^z2}(yk_M}6S7MYNOJ`i*-7XlQf_LMfEpIm%27V|K9QXW@a6F?Gn>uKARk?vB$-e+
zr{{<J&-E~^kV({MKk=Ze7ueu~{KLnekRXzl=3|v<2(a(w{-EswoG=;j;;XN_$A12#
z2^8?28OS4=VQ-ab67LxTE}BGy1_*NufC;971Ppj@?_RfBW{U_Wv<Jt{x3=$a?;kp3
zLc;Y|UuFIW2sZ$cpX+0e13HnPa6-lDl*l@saNK706$0FPtsR2xy|>?H@A{5iyWCHH
z@uW>i+@$Bo&#dp-yU$EieOZOJezu=$L}m)t&AIS1$1LmT1tt*00ZKo5;^+2U-YOww
zuCzdab0!r4PtbNvbRd|)U+(A8V<+q#Vdj^}SMU}f2z~D-kGnVCen;>42PW9t$-F-Z
zP6%BHLpbK*%fj~pK65Xj8?N1=S)hKISRQm&%Lf2L;14_5BRoF(=}*}N5k5Nqth#sZ
z*<->U+9Mq0@mqo5<s-lMpgk|}{>OSY-gtYvyHWMP2l~Y`_}W`<=O$7x_j~XoAGR_2
ziC;hMo_OkMd+*n-T4m23_XPoSrqM)=1K#YGi6FGX5RTv8C&TkK{Okfy0k=%(UC_7n
zyZY8)n#II{*A|>>Y0+~KM6Y^N?<^o2(?0Go&(~@B1p+!T9b^pQTflw%ryu;t-UHs_
z@Z^(k2V2S@_yc7e6R;pZ=P-LZ^1(4ndY8UKr_P*F``S$F%1N|)M8os6SrcPSlAt|&
z>e=UQ3?V<fpiI7fe)55Sed&>h&5D3YO5QP^>$`8e)xO8PJFTjVcgYhIUwo&aZ~gEw
z_sk0~O8dLazC!?fCVRqs_4#JXHlsNu@B2B~ePU9Swa2DAmFoLBkMXpW&SC~sZW>C9
zvSat&Ty>a6dAeHu9e|LR-q_|(nQ&aCkgm;h8_lt6uRf}c3!81lZ&dT``oO;=q5QC3
z>>c_Lpu#`&!HHb{HS1RE1(P~m;c#Dh`)${v7Zhrec~4vdcCFUY=#MF)lZ4<C8Z6!v
z=%9>FeR$7hoyHmLo`308x98|#tAl7-5H{OevZi6oqTY2|@S^XP(-#9npG=)zR$8<?
z3{DT-cejgaK#@wootMow!V<)XxKaX)Eo`W^9@dB5Q-{+1_}9<4oqKn?-}p!}8fX9{
z%w({-`3DkknC<ffI6Hp&q&t1?f_=;ra%jijB_Ys3#4L+=rF!_Vd%9`MRqo-B+~+>4
zS(W?mx=Ybyt}&q>Um0d+ZocNq-0Tn_XxD+m66!y2FX@Bt31pAcd9R))TsCl6_oOD9
zCZkh8=zze0pV3CxICcJvJE%VQ%j}GGKUlnwK@ms#UV)@JGNB4=^@Gw596VwI6Gt00
zyWOR<^Xdf-mn{$Yo)aB>?ZFm7^Y)u=Fd+v!^>=o1uhTZb`_Q4+3nY45d~D1y8G{fy
z59`AK$a!6$l)*lqFfX}z>veXTj|OR(CB;4-qFjXRsY+?JwcJ!J4D2&6{5Cg>*(|{P
z;D<kCGmcxY*<#wla64|ePS2v=51y5`w!dpLfy{*BJJBKyn6KlMG^uDS5FGX&6oA)r
z=pB){UrS}m0CCSr6ONpP^EBrW7$}c@K)t7vI?p+P3-|e!8?Lu!;Edii9F!@OXgZ|f
z$++!?8*+f$myy4vP1ozagLvKnsYLe!G+j_Vqb1%6oU;87t&9gTl!@ShAO<1m5#ASo
z{*p8q`Q})pVdfPfE1%%+$<wEE_mS(T&zv<L3}tg4F6XG;DS!XguU4=L0cyrd`mk%?
zzHrh(`|u6{kF>Po6-~5o4*;(_qy>X`e*sGn;2$$K2&Y^GP@at!<WmsljU~_@&Jo26
z5Js8r>6?YQ4DT|}A*N~9Tp9S(_&hUZJCWl+KF_X`7STVqYzgoXrY%=oVWw4_1HfaR
z$yk76cp+l^<r&Q<KqisVyyI_1!Xd}Jzul54N+HG=|J_B)$}@xZkBJh-gkr>4cH51?
zcZl%>z`(N$Fgc=c1_bc-6Nvy|%7p;S24a+s{X*X%CQELX)*a%zMtZ>ZJMRVGAtFTb
zZp>sd3y&$#K_o0DZU-c-DbT11XhLbgV;mZS&qFafBy9tl4k(zZmmQ`C^3|6vmm~2W
z_tv|?i^%mGuHGzDjViZXLIWQ~fJv_e0{l9N0#zC7cR%^{v+l3I^=-H7{rxg)3H&X{
z3!!Gkyr^12H!9h8<fse*6xOUDbf#>UWg4_$^(yzqI}${*Ggvsu^MUbnKre8wd>=Fw
zp&7-2nF|~$_4d2v%Pz5xCh3pL{HibAYwLZaG^2Q2PV^8e{v}H<G2fIv0gABT-=`TE
zR$xbH3D=8eqpQ73LQ+a*WCPJ;<~GZOL=QAGBdlej=H`gdxxg8=e3<d`V7(3^R>cJj
zEP9sOoKJ+)r92bz)pax@wbhi}Z<-6A{)jXV^AfeHL2GxMIh|`SCDv%255n=btG!_r
zV~_3^S`_sTb`2ud>7ZT~0|azHv>U(!vG^cF8ba=#ZEJP30b+fHmAVJ`oM0D!=y)*e
zWEb=PU@kGIDTpkD`3bXLo_u&t?1h;jhp@f%gf_<03GEk}En?=>tL};XLs|U7AF^<n
zzPy>d?b8Rf5HZ`xOdJBQCqn8XY<%gFPut8g=lQTM&xT*+{q<mhS<Qd^_IKUifAiZq
z7x)PvOkhUv_3wT!H{;JU<B7;l<sh`3-fsl+3J;<w+Y33EA=%h4M1}Nifu0$N8A~5=
ziuXVKybV!A(MxnY{0k@KIrRby<{+K`ZejBde?Hg_@Nw|@u^myL*K%-w)~?1^W}#^r
z0dmEN_Aimf0|Mmu=Yr#t5*{JOkhQB;**@=`@AL6f0}~<yFev0>fZrYW8A7XxP|3TD
zAj`N2Eeeefb|U&ylmo@af&2_S`-7fa4nh>>xt~<v^06Tzo#hn~Ds61LfCeW0phXK7
z*!LYx8^H>WhbG<fYvI{qtb<rKfM?3n3B7}m*D$g_xX%OcE8n58tUy|bwxF$t27QM#
zetIhRcL<qbDv^KJC%a8ZFg$;GBF|(uYuxMCyz}rh9!#kiLbb%pvUBf#ce6~Zct9>&
zx5i!|_IK>wXM)6Tsn~I*X7FFOXt7%@-vGeMk;BL2v^>x)7Qx%Ne7U<-J{*hmK{cml
zX+k_@Qd;Ow!ZV-&hwN7r`tH_i<PRcqgX=e6;a-33ZFl8Wmz&AQHkn+p0AR4-pB|R1
z;7eJ2*i17dW+5QZ2iiMF0EQ=P!Tf=cXGwibD<yzu<nIuZ6MGjRL&S%0&fI2qR9c%V
zIe+_C)dmlh#VY5?)^_JsT5|h?b2Rl}R_0e6=K}J%PD^;s$jQCUz=F?3Xk9|*BDe=;
zRRR&>Q=jVRUb9dlFM}13&Ggp0?<%bEewUJ#p;~~2$WkOB9Y1_XJoTwfUGn{4al%aR
zfMAJ`=g&7N`c}BTaRDCyMn$b6V`=M&_7HudTBckVG+@{1-T_|vwKfN{jkKeua-a^%
z!1QXj%)B<N4$S5@Z&>f<H_vf<4;>On(S4%f1qe{UeTVW*B#7wa$!j_MS{<CnbjB02
zT=X>`)q6pObePuoaH!lrPvJ8CxtY9WcCQ#=4)jg=bl}{4qh{z>2n@o6=3^ha*L_cG
z!XWT@XXcci4-fd5A!K$vOiWC$#tJipXqTAGAb+_DlPTV1DBm=#rzZf(TpR9QBgmqK
zgTna=9b2Lj{~*W{8n_sR+Z6`I%;>{c-=<>u%Y7cRnCoS}0&(v!wL!zu*4~kuD9Eo9
z^B52ElQOLh6Xs){e1L#(pToRnvVt+zj|)CuolI=KBQhp;|MMSzHMpdA+4F}o8Kc7B
z*5{A5gE=tNI1o<xmNb^(s8-=?><_w+q5B-Zr}=rt^Oz91j9ZKco7S&4@PU?rNswZM
z1%>BFPhPt4vx|=tG`P<L-yyzXjEDJSoY&Zs_qwaElxFS<c1@139sTj&IDb6W?PU8U
z<+r~+6nQiM+Qc6h5lAM!Arn?(@vnZU2~~CVT4Q2hunt`s{ciTdT&w`htpFa-hBdOY
zlt73L_62=xpf|R^Yrup3%@QzItdJA>k;AS=p>mPqFf$^&>^-;KDBq8}?E~=IJ3HLd
zFaJ)?<q7v^Uw)LV0uKTJnxRU|tSK$Pa$d9L`xVX+ABy#>6mdwQ{I*=aN%o${G*fme
zSGJ$kvO*|FrdpK68X=Vu+y_WlUqEA!iGqZ33IcQxYFDDWB?n+6WolHZk6OB;%j&WV
z>Q9A){Zwn0u4z$2fdo(1R{^F%WWY#e#$|>U)-I41{!lA&4FE(~ME)vT)my+Kh3ZdQ
zjPNZ=iAM<R)k=%7e_jF<E0$9~gIky%IHMAwDkc!Y3Zb4<q?JuRs3rsez4n59tcdQ3
zHd?eS)dA?>7^+eo5Lyu5H!r|b{@%j}wSs+*IjN(5hyOqR<u6@>%-aUUd@CaN$G3qK
zzO3BEP%dUb^REv+E>Z)`$`ZZ;ugSD8pOPWLH2>OIj$8B*_UoXGwwyU~kn|vVv}NSy
z!=bSd@v+3g95b8u1;jCzAiVA;9{7j>4SW{-$JyuS9t?2UpB>nKNOs;VI7gb7Wj3@;
zV+j5S`1F*M!teUv=R&oF8W^6*9IpP*qZ1?zl`~x4v0Tq@{{&S#sL9L@Gkc6r#=6z3
z>=}GjA=8H<1b$B!KNSPlyf2MU%QnBy;cJEVF@bsJw=Z6FiG^)vB2DBtFc$uW?|bNR
z7)y~O&db;ZSU^L;dL)cl_zw+5eoZja<Ufl8Pi_$H>!R!7j#}byj*Dx>mjgeSd<svb
z<zJf#$L*SEuB|Ug{2ll1Vlgt?D1TD6VE-$N2nPidmw2+!Iv(MyyR`6}4>7k!pkl-7
zRdO)in@fU5gs4!^<|{Tiq6_UkC`|z-LlT@ZGvecO(=}SzR$J2k{J9t8e129!@_;am
zaBc#1V_+l6VtTJ^lLMhnGFaTWb&G+J*JUGGJGaTsvyfn=fIyfUni{nDUyWgigl1Cx
zfN()tn3&8eT0#+l>X`|X`4+_B$P5QWyUuBrmUTfY;CbkIL_i}77-4$3?g`CBJiErl
zAmLvL01tshn`Yet?Z0KCZSV&hb@K4s6LPmpKn>u5306!dY2g<Fp5mJUwaZM5bwySz
zTds9J))@e4lT(`~W&$yJVwLk)m3WdRv07jw06dfx9wX^0tD$4=6=-;TE@_k0uB6(@
zx*dJ$P`iP(EPHgWQ(6@kB!rh~U8mf3wddf`qgDtr+wk_p@vItp@U6>L#O$trHn{g;
z3LoP)KpVa7IqCfDD*Md9hXo(bL1Cvz?jxov`9Dl*#)Ia_?_jLt)$?O5C!D`2q$%^~
znZaYxwue9dQGd$#jcLuRZ@rZ}$8*g6@^;ji%H_}Q`NW*ZQJY)OeS2OL5dJ>;H09b7
zrOv-QwejHH_GRZkNB(@FEwiHDw28ijDO^tY{Al;_Z}3js#hz=zh!RS8q6s_u-O{!~
zw4p)I2hT!yQX{`j@NVF_$a|MVCUvdZ#SYuE6IzrLHn}5AtJiFV>*$tN+XF}5_tWz6
z$$ytI_nV2@j)@uv!u%d;AB`wJKM?Ppr|^9ok2d>y*!u5%_#FGZOZZSg_+bEa^8-8B
z#n2xX_bvJz3O&1gPx8M*;qeBIX?QZtR}o&|lW{Tsx#bwg{FzMYkR`@s_p{iuQ);_H
zVq&RTCOzGN1Up1in86N*mK-8!NtY<EDqc5aT0R_m4;->I_&yNbX1lZ?o>2L3-?PUK
z39*atj$=8(EyN4S%mJDb2oOWG8gB9>ivHxun@jmE=P}7SD~E6Bwrj67?ZB@0-#1W0
zek<Fq4*bEH+dn`nmo3R%FlSP=W1rfFBRE9#2$POg7J$M#II>0oVfb=#acGADgf`l5
zAcaZaDkdQHIY*q-p}!QK*|Qk>q$Roi0<$H63Ma_w`Isv&w9^Nm(eYh!uDn6cuT@HX
zBIKawZSE<DFvOlOY!S$56V?0k6djypPuEmF_ZK1!!UjAK48a2566%Mz9&|f6!{6XI
zaE!G?`h(9v*lfg~4{vei=YBy%Q3(ro-X5RXEAQ`dS|HJrU#Ea$;>UYan{dClzkY!#
zX6*Sego%hp^mRNa+>X(fkI-322pesAj46!;fCqvE{AW!LA6D1b;r(s`+xsBGm5&w!
zKha=>2{Q*G&td-J?+N$N`<Rt%%dF?@A_|=f?V5>~2buK2o8FM0nE4RT0Zz%D1{WWc
z#xs9L6D1H){D|5G@yz>o%pco@G5on&T0tD*AtLfnHxz!Z0Vm<Ug!{}J>}hacn3)~W
zV0yDSpSIL3CM%w>T?e2KCDfyZ_3I?mOW=clo_IeY(x63$2KPC*&UySv2Hr{aG21!J
zX<1DjB5e2kxq&mjozNdNGv18LpPMOL#({DQ^2Nr1Qu%d@)``J%6=GcQ<nJ@b1P}0(
z|9pmz!`~#{0UtoZe}{Y-`P*>5{A-vBdO9S{ZP;AOpAT;VD+IBC7Rb>%{GIUQ2j4QD
zSt#7+rHT~kiEox~7vsO*ClH2TABX2Oe@tfrBK)5F@@CZ5WSuF}qk*AwR1F$q_EFV2
z)=VV<4=RZ_4IWs>QgM|Q$18>K=kRn!AL-+#f|t>sd-L5O3IoS{1kRj2FI(uqjHym$
zQ;JpPj%D*S2)JBxne4kYGjfN*)cP5Pg)$#$ZVW!?teg#X%UPK<H6TL!?$+!YLMbLa
zHFKKG`MamRJ6I@AX(;RVt=G8?D_6Oz<PXA{ArP}K+Z0L_pNSrsSN-z&7w!Dol`9mL
z;%YZXA#U%N8Q5H{-pohvF@@JnOPhe>`f`2nyqzTu@JkmgG))Jq3$I%~Fo-T$yigy3
zV9~nphj{UVc{XzgM8K@-l$=~W-FNd%?rQO3^Tu`NRGyZd6|0@|QAM+X5YeBh&_aVj
z2t)EQThpk>P!0NkH)-+0H7iz{^F3xq9hyCWU*7rkj;&j@IO2M@<+97%b^0LqS*XS8
zLn6^&LJdOwvaC}u<u6-8C{CGKNd?LbC-)oTqyOMNce=}@Sy`%q)SG;9jsA8^LxuJR
zAj=Hbl7))|7-~!lvRxsPJ=i0(C4}&~AS5IZU)|4=a}QX9gIQ)~Rf#|apuJ3eVNoWC
zaGHccgosSyXNIo3`YN+Ozg%hA_vaCy`FJqsGXqOlRz7HGNC@A_z|SIAI8A^2!}m#B
zG8l|lMP>TK;)|o9aK0tW1e(i4bH)2YWYJ`F5BMO$6P`Wzhaf>T7w*q5wf>B!aQNAe
z+;g{9xn6Gp0EpIv)&y+<6Bz^yfadABUblL+-7mr_2O@Uhbm`5v5Uhwk<B7?F@B(%?
zVEgd5v<;_G|6ym6<JlKqvUi&MPb4(%H^Pw*&->!5ujPO+zCU5X6|h8EJSRMF5YO#L
z@4d&K!~5>MT|5YYIoI$p$QK?s?}_J|F~$?mG<EtPV2yQ95VEKf8+1T2MEFiLMa(Dz
zO#EyqYa!vhO*`S4e?Q2Bz!L`TUO?v>A+3q<Lz{~c?*O6FiLB){i98b?sQD1M)#?|k
zT6@BF-Gc=B@tode?mMP$gbu~wpGlr8WjYJ()I^qdK*kB!&qUKc)?it`W=#-+Tbc`q
zahPX<J~Ku=r|`;*{qX*iAA3MH{OjzwyYrTt+^2LLw!OCkX3(F-p|N|c-6}Q?P%ih=
zM{nX7?X`by*)oBHwS%8nBNCzatGapX_4Zuz4D<7oK8rQ+=$*HDPM3$?a{`C>F`!TH
zn4dsE_`gH<m3tfpG_(ogd4ss;Ozcqy5!j$ZGTrvyMjy?J_c{E|2j8JU9P{5Ho-N9|
zN%92mED;asD@53Ee4HRaxsMXZdOv)x@P)eO%FRj>Xf9Z1hjEW*7ydl=@++2?cY^N_
zeB?XC`1w)!&s9r{H&&XASWUr{=<qG?m)tT{$eW5!mG(?TJ4#x|gNC3z;N=?YdLEV4
zisaKU@DUuUlQ@gtBd2g)mRGlJckk^xkW0?`4J~@9Zq$rBgU7b_cDeWVzn`6v3WROK
ztuhnOirsw3E?>Xa*6v^-IzneM3x;sNY3&;KhMcDfu||X}B0M21U$1b82<?ym<~a*v
zSzR4Od}2j({5UY<Vs&E#jmLiVYg>Tu770eQhZ(tV|Lh5O_Cl+KN!0Z*xBtKq%}$(?
z7GQ}6lWW{v3f;;Fof*7ue(y(ar@}nq#K`%7@#rJDk0>9`>#o}DFb8<(Lw8$jd3XXb
zC?njno(U_S-*Vj+J5Rr^l}+huinWiRcK$+(TfKaR@!=B9!s55WP`N`+*uok5km`uR
zQ{|%N`O`0cUS$u=5?(F`e^x)gRoCv2K+fWg%!+)VXdyfIDNM2IIHiv_12_4IL%vc2
z`uD|aq7AXQBY@x*g+Dwlt-_yu`Ag>PZ;jO-v{pCC{{GE(wz;GRg>y22U`-VSk4q&O
zZ4pp9FJ2x{cujrtEe*aXXn?--(8t`T?!Vv4=YCwL`^NLYqK!*6OUtY|fP(w<$R{6i
z*GQ1^{6jEgO$65RVDJW1uucQ(mu!?!dZXHlAWQIo-$<CWIR4p>3?Bw&+Y#Id`ARwj
zn_tPK4<VPqoR1G50HPVOeh9Nur}Yf5A~f;i(Q+`u#0Ta#zkSK-_B<)~*89bbwA^q$
zgjo8W3^~(}v^>(+>4Uc@8@@I_KfGo(Z=Qe&0xg0&i!LGn6CU%YPyQ+=Fp?L6g8Kj%
zKs(Au7oHN)?8V=`W<DaMV-^vB!1F*F-v1wJjS4>K+$Wv|?g{tlkN<}+ngcrdAkOog
z)CuGRT0VaNeYy9O>x7?VF)oBx-hD!3^IoG}rF@<vgnIz#QyNd+$*%2!knqLNe9Fx8
zC<j`vbO=>Af%~Xf2=i!9Sl<G{;Rl+eKnUXfSgf|OcEHQ>TgfK`<*)tzmy9p|nP0AF
zlyydM*yaZy0<_~t;p6wS&IT(b<5$5Lg%*m?vV@FAFh@AMO828h@8oA5`nU-#zWh0A
zI{=Y)1pytvK!{$193t@0XS5)M>CGp?hz4u2u!c**`bAh$?lAzAckP#2oDUmxfY|Lf
z-I#lqxz2OKY%rQjW|sFKIH)mtt^s^PuVQD<IPmi8Z`%8ekO%Q?BfK@h5CFry;{9c@
zHq5%XhdleduWVV%2U9)FXe?rp`o&^!_T1?iL@?a7cYjVG&F|-QKc0wj0D=1ppZbI*
zfd(cj@$(|Yw10Ma9}waJklZJ~;yn`uHYV$yEE7Qf*;7y3#1G#mCKY)8c|ZT~(a+iQ
z;qM24f%ky`mq2L0`v-B4nH(YyT1MVEo@WnIEs_54>_4P6T^_ja9@9|p%(1=<p_>`c
zm^1^p+QgQgzegYbqy@V0?+hVq8DCgqhiBa)Vkt5s+O$o2Ul>1lZ<qjO%?(23da&ca
zLwtJxRE%waDaIMb8$dpq#%m=*Fn%(50FY;F;dy5<&dr)6<k@At9m)?!6ZCx?>9!nN
z_VKE-MH4H{de5f<wfFS8`Z55$$?nTY*72yg?V(2ks9s{Ldpj;DtZQI0RAQD~GMumL
z<}Jz{m$GBF@)C@Z_56!QzPJQ=+4$J`3S9&UNCwqP$bmT(!ah-x$djle!bw8%wH9>K
zi>PMk!>16s67p;0%YX>qCBF@1d3+|A*^J9K;otnpSIt5BFaP^D^|4Bsz}u^bq(<w8
zs5S0SAN{;rr<a;K{{4UbYn{os1#{+_Kgos8cGuM@M{xlc{7$F?twPU*E>~R_x7E87
zs;5(f5dIK+0L{t0qg~}`24BwVfHduB{feade8g05ucBJ9rUxCbQ2vbaldgVFo$Kt-
zO6XdX!`8P@J6ZRG&`Rl~d@aNud_m$>ahs7zXL<#$q_t7Km^J~p;2ng9hW1j9;Ts(Q
ze(?S1Rap%QMWm9@Pk-=rQB!QcSF3^6sBd%UD5_5d;H}dD34<#dsHAuc@o^wiT0BoQ
z*5&FYsu=eSo}qc+{w$o=Y<%xZcDoK~nHuB(4^nDZ<SH78gL<#-WwU|`v`JG{qdwLs
ze~;=Qq%xs|p(fo=1-ev3EIH8YK3&j4g&?o_(u6_bXNzM#W(Zf9_~9R9k*cyn3fIF#
zT0Sbw)Us+ef(_@wPR0PUa2;Gn*d?rLQek_s8TUd^xa_IAe_9{yHhq*wLpDRcX3el!
z;SnGn;X0++N6Z8G$cFhCE`;>Vkm6&7U>o+?@#zqln5}1K44?_mIp=+H(4tX?7YKa5
zQn&eSvo`5DPShpr9y2o=zRyF|6D|v$5fme(=h`CZsh#0=Gl+Zh#Bf{f{k69_e75+0
zxK7&6OfMQEgduoJd&6cJc0Y;~7cPqloQC=a%ggL9nustTd=Z8C;mh&Y{Bd7}77)EC
zc3e;r>n?b~hkJ{85VPfkGVM{cAKG1r=rivz&s?Ya%>BBk&y3Zfv4A}A1e30`z7yE!
z^S)0v;u&Jx4^vl0W3+;xK+Ir!b61~cJjVflzkOZ9ZT)jWI_?RP$7*!nnMCA%b1r{O
zE>>o^EXEYT5A9(L;@J&f_hoSXq#{UOB41EXJ{{MNpJ{bV76ig>vA+5AK0j>^#<Wtu
zL$wAn7~2>R3iUZ$2iFI_L%pV%!9OkkIq>b7>RZNj|K5k&nt%R;)-YQ#e|&*Qn<hSy
zJBFk4nkbp4@n*Wy{=EYxg+uja-G&Vsw?{}d4csg_@%9Wj6PHp%tXcwZ(|}sMlsu#{
z^C3pWLZr7NF5rNws-shDTxeS>Cusy>1Xu)qgwM7X2^h@^A)5|_3dLJ8vjT+X*4Azl
zbgLz>CcCw^goIxluCLj`EOfnlPAhCPW1r4OTI$n&zphsc%y?oD;ltC~-s#S3aEaF|
z@|4mbSObVMeG;VA9@Yv$W7L<C;NGrPpd0m}(i$8Upk1}OpO2hSQANL^Oz8tzslhTP
zz@)FH>&Fa&@+7GCY4EFRNEl$i@js&vDL?>C3|booj*LKq3UHYo<ukRS)+-UOTHD)P
ztXsl3JQe@Y!~oco#li`j*^>BTd}lIH7a4>&?pK+7d@zT~k-o1871Ukdgb>|lj@<w_
zPb?MynL_{oKmbWZK~(V9uC_DaQ7!%|n0M7R)N1AuNBv4Us8?wv^xz{80251X;5>Qe
zyyfFwvKS)ygT@CN?sJpw#RY*5)*wmtrfu=a1_^sDy2gTtw67;7lQ5OPAVFj-0F^4x
z|L`DeMQ8sYSO7M>gKGF%{yu40E5ffA!FA4sg_A;gn2{aM-ZwwrR5?al0S_-^&6i4T
z7oeb2b1nFXkm}78hO2}02nGC#5duiQ13%{mp4K2|vFD;#xy8;~oAiF>hiCTbORWbX
zqD2~sFcoTd_&WQHg}L{I-4PIN2fXp~M0)|&BPGT|Z=yI<-l6!Be=WuMsIi2|cD0p*
zKV@OzgRzS+=*(UZNAwwON;s;|aD65ZfiZeV_%00BmWjU3IG;}*bO<gxO@-xOHm&7&
zjRX1j`=ZK+IKe#uIEJMZ&hZ;5@w|Hz!=duh-o=`9E7WezdC)-_h0+u%oAwSjrulr<
zXK7cY6v@$Wedc<1-gl_bbC4@=itjMo6>5KYJfI74$sXfPLxBzg<TS<tKwl^Tz<9T9
zvdbSlT9??GnuaD<RVyL$0z%<1FS3M_ntIX)eiS8EQCc!RWj&gKN4rraL6esV9}i}^
z0|({66sJalFSFtntw};W!gaM~;8PM((OzKNj2v30s8k5;o7Sv$AG`ZbTNvt9nJU?=
zfXYB1rYyElJft3Db{-8uRh$sH5@Xah1aZuPNRP$=?TXV3K5s^>s?oqmjRvZ$bf9#K
z2G!Xu@KJ38Q%0eAD{TD;3CaR)84W5GYCnY&70KQ()l(zhYFyLHq%^9J4?(5ss+abo
zTf(-`nY9-%P$$g*S__E96e+7GDG(sRpNLiv>u*4T&r9&i(hr1x2(SXbsFVDx6@s>+
zK_IAB?dldkp<3~{Ml&j%0zND>*qH@yfWuzUj;K*Q2Q;;HXpIhOGpufvi`IqvW#(@J
zg`{qf7N|qOt&SNetpmcj6o^Sg+t8>`CuU=EQ87S5JBfNlTM4tv2b9QI2rPh>h(^zW
zg;LQs$*kt2Oaj6V#Dz-ane^!kU7ro-@@L&<L!U?6$Fl;Q3p8Ur6a{a-OzkwOQEQse
zp;BLbtq@?jI8_kNn-JoYO~+5SxTT9mK6y9Nb~97UdPuCF<f#zo@Yno%J(_VKdM7*=
zCvKY$BKeNa6w24g73-HpwURb{vdDvu$*O;{>mBuIS}7s0ZvJ8kx>8$fcD_%GG_hU<
z5v!OOsf=myD+$DA7Br5%M0mzTh*@)lPK3#{G#hd4BeX(NAlNKB=2ZyFTw~u%dH@N!
zo)%zv=>B^x3e&3P%LHJO?t4!>>3UMZY#%Y%F>Oe8^telwEp~U`v{fI00kg4(@BgrS
z>%E=srB}BpK6H(N8)pB_VN-wzzHppuk+T;nUzsG?+95V2!zWZ=LFzUOqbngTRw>gk
zfezOBV6j1=Bv+|L^fJBK*+-?xsO<o(-nLFx(=4ZL)q}te0P0uQ5#ljfvWQsX4+L*)
zwncqL8-PGqE1*W(5Z-H)pMCPcBLf>s!xrFzc7X-@JW8)OpruvSRZqHFGt}Z!!h<lC
zL0U8(9Wp2K)GHvy3@GF^7WIt+K0*hq*@5N?G7|~$oc=TltfT}O_#k2k20v{kQ+@8x
zY-WSn!2*Mre*ugLy$e_AEA8=w%{c_fvcCBX(RNfE!1j6N>eXgO)FWY$S;n;P2VmU*
zyRJ9OsgZQEo$_bhW;;KI>i=+{gS@=sP@%M?2(<ZH5-Y60pQ!*kXx2o44(b_drk3l4
zXtd+NY<(vRnXMJ#>v)_S3Pg@aeN&;#{P6(UuMp*@E0m_(=dvP6xr?3B<qb$=qiklS
zY}oi!*ENi?gwbb1JAt_r4^d!dBh#X+K+C!hVe(oBWZI-={J}>ErGN>(UsV!pdFT$y
zvH0~JJKVdw_bDP$umU-=`Dh^!;%nuTkN_Yk)9>DXC#Q9wEzJ#$rb+R>68KG|?AsX3
z93hCs>g&w+0&^|utsu@n)A2e$D?&Rn65uNE?@)VkE1V0>^_z)<6=-=paI!@UF^^Ij
z108yjDH6%sd@$5Dv@Hg<)Nh;T=eh_XAFUDK15JT}Gi5g3s=Zc@3aXRvPR*EEp=}8G
zE8(4Xi~q4$zs)?7uQwI|yrlTmD4>Fd0ZmAUj-j~H_o!fi1@+>O!F9G&;?>ZAh$^%m
z0zxX*bxCsp;KAg$Eo+u#S|nwoEW^B6(1MnQ`_Lq<59M(kLer5}VY=-Gyxf4roL!Ip
zjE)1?Q}dfEhwg|X4oo%&W~)Fu*?f=2fzs|ksWbDXl@_HMjsvBR2c@PPuD+36mzXu{
zXw1N4GC39q)v~9TfZE*?I7B5d!I6xa%$ioQdO62{2ZVA2YnxG*APaa%$~T}|d3stq
z+%qq|X!#JHDYH)JyJa4P?+3?#g^a+&aYZP4?5U@%+_ba+Xc?IChtNV0imqB30`rSd
zJ(w|J>gCJmuMGS@03EDIUL(h3s8?;lC!`<HA@G2C6`BrtVW{o$KLJUY_%HCF`p|F~
zfRSHFO6}~EnHDRX=SZ3od`s{NiB&b3=@eQE01Lu68Y1uTO<U;$yr~ulKxoI*$$*Tm
zB`c&!Qva#GirH-Ot^<uq6+AU4W7VP&fiPnL`LwJ3^AfdAIx#!%|ENeazv8%FCR%_K
z%4w7-7=Xp|Ghw_FP*l2vOs@PaC-<Inm~v4U?T0;RL};73rgG(Lqw^fsgv<si#DC1b
zxc?DF9EdoeaO53bZlV0frV}wyluwf}W0~KE=s4m)#DS^J0l(<-Xj@s%amM7QSvCPW
zt+5sk)cAE+G$ciP;$Z-A^vLG`Kwzds5~LCMF|7emfELg^AgD7_kHa?<lTZr*7%;&>
zFb6E~ZNs0!K!=>L`mzE!Gv**w9LN!pQvwahu{mE8oyV~~qgi~SP4(&+szAe{dJ*vP
z2gxY<R8rBP_z|)=qjpzGc(0I#0cUf}t8h54l=CEU#jOmPxuq4ULyZiM6@>+}EgFO)
zdB4n>a5(qh<w}{k;CBK)dKDQ8K!G_D{lioX3bY}rU&p)%w9kNu`UJnsBuf4;<U_~@
z0Kj`psUXa$tW2F_6>+Fu+6>OK!mtMnobLc6>7I}`SgkzR=YFtCxq%MV%RRt;d#->B
zThoXr{{`i%7k|*gK(K+bsJ}*acdGw{fCbDLAW<5ex%gqq)v9}C6m&h<kN!p+n8h4e
zFu|asL+|2QO^EOtErr88`!Y_P3>NkC#}l~?uXhu!7vGGTPRCD`#`p93hDKCg4)9Hj
zXfOxb<pMAi5%q2?R8GwZrvKB4BsGq?xznr1dEEsdAnY>}Upr3=`$_0E-vK01U0_BV
z;T3`2G!SSq^t_P{p&v7z{(iIxfv-cCgkH)tfm?#Nm8J5Drj$ymjHY^X-bOgjRuPn=
zJWgQ2gm~?8jeAoa*MSMpG+Z<9sxPMcplS*6NkZdFLxG7^&{mo)9}9)#iz##_ASLkW
z2>KF$4}lTAG=LHMX69E^NS!bvOZ2(dKtYhE(%PratITQ%_}wyFLbHLX8RY;x-~+;c
zP+Y%V2ZW$mK}*sCI8qgyLtA2|V`j3Hv1?S#b;SU)R<sLXplv{F5*G-eT&@EUf^|az
zjRl$$`VU{ZP8$G8RWc7tYunPLNMg!cr~3eCVE=+lzDP^E*pi?7c1}Jy?9*0E!g|#w
z>aP>9M9a~qdqul_0DvsJqD{nsS<L~xLK8|2is(XE?y(S)7Ik`;!?g60`QgI}*|(ur
zHHDqA!xc}IZ$`MziQZqjxHJ>Xd6BPI*>BBM@j)tk$Hq`#<1k@tC2Ew7nD8D85Wy<P
zH56BQAm~pe)4X!DW$>t7wp%eV1=-Aawkm{Y1X`TUiJD}-00Iw2x`tl|=>Z)Gwp>Ag
zH35_({RcQ8fOFmidF`O>;2c1Okh_V-S}Pl^f&b>2qqT%x79Pac1iu9c;Ta*<3beXd
zFC*|Np}jbSQxBROOs7(99WJG{KL8w<N8!5>CbS={?oG5SGqaL5#T@#TKE`?!Dici!
zKRGjGwK!_07Ve`yI*Ep-NA03s_yLIE2O+T2FEm5|2K*sde*^P1@^g(XMFA23X6!3z
z0IGByf**6G>ELBFFrjle`UC3VJ+}oDFl&VVa!)ANgG0=`Y6LP^B+|^{l!koWYB!O#
zsE;2+S(NMVH~h!ki}Yw&yrX-aG)6?VVxb;1F}2#q2PdMjap2gAvtuhD${lfFnsXpK
z6ixG9M9q&lF#S1zlv4^JZ<UM20y?Y>UYIG;pmGTMMT(xZ)6*Jw05UqyDmJ^GHiS)t
zTm(&XM%F&&R)7%z0Urm1WWWUivn_&`y$(<y9YQVwEP^bOEytwi9CewAk@m^otr_<o
zLe*x!I#?I+&=by=A2acsK&ukW$}6E7K<HV@M7>t2dg6+J)kh>CnNFeQ07#IIbvoj6
znp|8fn#SigyDF`++^<lw-s~!aUxOlCS=2CqhuRNNsS-G<!X!#jvoQ0b4{?D2TT_Jo
zsO+Q`4kX%E&xKAC(5>&;e$YP1`XT_(o&A|cLI8;^1dxC)fD`xyNNAKWk3R=_@vVRW
zF7zJ&Rwqz^p9W>%^i2dUzybFTo)~Fk_d)03quo30)82&gd&(AX;0co}_)Pf_CRRkj
zVkSbh+r4AaK)^@fA|FKT>Xp`KHmRxg<azd$ZM$sAg0hu09SV~1<d~b8oWZ6M#mtSy
z%(T1$c#{1XlXuDGw_&*Fp75~{i{(L>hI*nb#ttaIT%U%mPvf`YYwSzcnt#gj96NUE
zqVt7L;Da-fL^$JOG#o#R<qMx5i-}#hoYCwTyN`sK4HJ_p<#^xE$oJBJX9}4Bb;|4z
zI&p6BIz8dG@GUDRombROPh%}P@aOcU4Q#{Dyf>>H3qSq)IZ#8e=XuMK=VgAH@G<FA
z$Zn!i`C90i5A!fwzTZE7@|@qbZL#MLIcltQTI<1;n~Jr!9QDL5-FU@+C`UmPT?T?K
zAD$`+*a)qJXvJ&~VK$|GPk|qZUM&f)5JE44Ej|kg3Cj3>aN`3fYM(1G2Li}2!_GAm
zxK#{hT?oAan9!`cG$LkBq%gGs$k8mnuA3mOx|n(QCR^dL1A!V~#C0?etj~d8hG{H7
zF%r}PCTJgMFCYStQNfSv(h{Kgu*#GsqeasIB|+V^QLmKRO`XK7MvFo}=?80l)JpIt
ztqJ)$&o%mmGk1pg!(`Pj_!ICI^C<GeV=m=<XH-97YU2U`ZSuvS@6yFsh+85^q0t~M
z_oY|;0H}1y?1{QrWgMRk<FCpN_$Hx(wyO+lL-w}R3be512Vl=@Ik-oZOIg+W5t0`$
z(<xwt(>enN&qhL~9S8+MsRD@N39FoA@-bV8zK?l!X0|Y^NIzS&;(a(gbCG02o-X8@
zFl2^lbg|rh3!d4F8Ng`Xr6_;3wyx3O>-lXM?wOBk+=QyL0VuqW8s~gjl;QU|pHF^P
z(r@#xYxK=OW#$u2O{GPHL{mGFl+f4-5u&_S-wue;GvMWL9JS;_{_*7xnP{9Y7Q2s{
zdK^?^Q=^sx$U8SEL5T<4J71@N*1}n9vd>?G0;?3`Y_ap<az?VxT3DV2_u8K=ee3&{
zpCkOtbHDkWm4I9?3yjf_e?Rl{hU+KfzUOCtI)7~GYcwG%zL(0^Lg~Uh^ks*)jd^}^
z;dX}8bxL*{>-){J2IVB)^=jnI5zzT_gMn7Dtn=p2FEAPwOFb3O>#Kw&0`ny=45zbZ
zI|!@@vmVi6f&+-a5gnnN^a!7r_uyoWzk{XGc|x>e?!zp-HxWX>HQ`(WF+y)35C<kw
zW+tROIFpATP*pDt2aDHXrj+%apv`JSQi3<8KGcst1`g+#L?Ifox;f{Nl1+$Lxj47m
z59e`xaAP%?e6eVv>;v_(8PL%AOpmk~nyn(jRYoiD;-Jot`r!wFC(&G|e2Qu%ZHtw!
zeFHzb?xTTmolXY81Re$2BkEzFzCqrsjJC3dhk-1er#y}UYV=XvvNq5TLh!;vOuPV(
z05*P>S4p`y5d7!<Dv7n7P`;Ffj|w2jgGJhcSzEhiF#v&t`lX)p@(rPU@<X<-hSGG*
z#<(}LiOTU2olX2I)YtjpsBwi7MCZ!Nf$T6-uuT767ra=;lgE#qEMsZYtEgrR>mF!(
z{P;1{J(*s;n_8_6%MMekW#wzlqQ?3$6_)*!7cC=EU?QSWV!udop=%R;Zmbcu#-qLd
zT_2CKDRX3yX1rQ4%5u><;D3S887vm41?UjcvqCe))|o9o5uCD7od5#5&N0HZ{UoRo
z<$%?>Jz)yPOg#cM0&=pg%k^mSK=L9~+XY~TMq#e^NGK+3ZZcaw<p5OtLxoVy>bj&y
zV8+LxThXov*V`Z<z%QYd&Z}j1WD5mK`vFMklkl9DDHh;C1sWKrWHt)ThY)*AGxmhq
zwY5C7jzUdyqZTKWzXuf|WGh`v;d*=JKf+5(^em{VD$w!(EYJ!e_`?I*2|vk8{SZ6?
zv;f9Phtsy5Q(j(R%7C~R-iaI&D`m6tYBXb3uYT32Jbn;7z{ILsV2O1-sFyn8Djyy~
zluMb`6&(W%sgHX{x>_C2uUE)i<((@H4sBz0F0MN2vOp5EE6lr?x$71XLi<64Eb0fG
zL{wf5c%Wb2Qp;aji=yj<RrJ(=sf9PYDWyhK|Hah*(}fSE^t0S)3JDLTsP)5XPAdF)
z=jnk+N?~&k%#xU~%co-ZigI}t{PU1+a<hEJl3Y>6%;b2UAhuysEng<rC>O13*aU_1
zg=o0vhyHbI(ka|0-;Z#4g-n{@OgLYD`mnheWqQ8Te(Lq+GQOXMc<sv^%4Ex@nGIV^
zfLgW4az1e{!q0C0b&gM*lR3GFd+DDGo}of!Tcis=qo><#g<?(@B>$`zGaK^t70jGH
z>7S4MG(1P)XDR>sXpRF(V<2aj-U0s{au$t8n0(vVhTladp*{ID{<xGa=5xc*#Us!_
z4$pVa1hQ*z%|8p_>*4+llyLE0_+3BOon2`9;$%h7<lk$?mqNt*UW!Rup}hWF#cdBg
zG8BJ&x+%9s@XY8bL^wrYMF=G-6+$ZD08oN?k0+do@zFqFW~L0E00dZO)d3i6eRM1K
z(JmwOBAg=F=F)}Ar5tm{M#9b#V1qhrMjrv5HmXs~;s>+z5`+a%I=eJGpC|bG_Tn6G
zCPgZivWSL7_}_Ha&KE~=d^&JOH%*63t}weo0}`tv7C#x)PxZ?j&D971jVH7S5E>c?
zlP*lxzz-}Y*r;eTJ>m`hH}InN!{;DERlrlgN4xp}GwDa2v`(Gf!GnrsCZ2Ujr|;y0
z*MJ=e6(!mb77c{_Vvhhg`U3&j*yEf(26Qy3ZLISFaXkPK0xar7hcq4@%y7Q~fG7<L
z{nIf-ADLOQ=vd;>bX&W?IuZSeI1q7QTCP1Z(YFm@WC4LkBAS&P0C<d5%P<oe2MC!e
z(ZHtT8cKzpiT!)`7CL>=xpf;hOmB?`C<Bmzz|KN=%(i<nAvch7?G!!;2%@&&p9Fav
z(3wF;&}D0JodcQIfC$PCBwL)PsY6v6cwnDUvR*(27<fS*uwl^>lrL>Uy98*`Dh+}+
zGBx{vj&S5H?*qXWCu-LHNa_b@0X$gLD<N!K14KeQz948YB=HArOf(qi!0f88yVFdp
zyt$T{TgiqH=W;+21Q;;Tp{)ToRpbF#y|gk2@E(+qmLKhgcfQ|U0w`#BAkNVzItcKh
zOv<ASTPa-khhrSq0W(QC&;vSHiJUb(02llK3;117A7o}~%17NjYCn|XSX%>e+6NS3
z&SgG6svkgOb*OCWgfJiT<Ut3=z8-%q+D06RI52%U&@3%yL_>2R0v$v1XQp3`Ku6&_
z+@(p(DMNv{Gi91H5e=E<wk9%{(ab*rIzlEt%xze49N>UD4gU|WBaC{D0uIl}wg|5X
z&9>D(GxrF;5FiD?8uKfBG7PlHJcl{}Eu=G_50h;rm}^U&lmVy+MBboXw8y}W5>OT<
zOw8V=1u_B|TaD39Le6D4tJi8v*jt>ySrr{(y$|yRQ967<;xehKz~o9<s|0S)bofxa
z87(x}&tikwz7QdcXjuRU0|qM3gl;t&p&qb9TM+5dv~Z4c=|3h@Xp<oF@DHGcb_jvr
z(osKJ1+^7^df>q|v`nUX&=&0xe1qSB58C31G67yG+5z*d{=i|%z!Ck<0vhrFH~@J;
z*j?7}2r^L@ZJ-YZ<dlwWr+^DTYNQQlohUb=hy$bLz*$WO@QY}2IN%otot)aIcdgNL
zJH78;)cUb;AR(D$st{|b%{Cf4_Yv|WME0eU67UeDC7Q|`zEVzUm0DqY+{vWW!L=ZG
z4yHT-^P=mdrrn`jO5^9r2b~Wkv0bGLXXBg|pRFUi(b*ssWd-N_G25)9O=e%2ZS(s<
zQk~QPfX_ktfjTIl7}38_AAPljl3UY)_W8@yG3bnMnzbcJ<=a6$M&Ud<<xjKqX-OB9
zJ*f17x~$CryKIFd)2`*omctc)oh|kIW6BQ-8(hAfkN!j)7$XM&wWSikUT<Kj$)j|`
za3DJ(43jFlT!;fBd2S2kkIqFLh&WIl4h)i2%2VW&FJ;ug17@YW+7%DExbqS-^i{P9
zSDk1aH%2fE&P+dMJ!m7C2{*x3`)E7RSn#Q5^>Ht|qW!=uD_Gz!Fgrq1fu948%_g9$
zT-K4GEcrj%%sR8`m@C;#yYf?BT%q#NfIw&t$Y9RVRI2M{3M5T~MblE7dS#+SgD}_X
zmPu2xNA1hbywg_tgVQ;AE71bj489d_01tDkemPL_(IpZVf(*VNW`?DR4(MyRHx_DF
z_XZHyC)2EcwXLqM-nE_YbhTOp(b}#y*m@k&R-g?sr+3vyv@QCL<2zakt4qgdli;)U
zLG`d7m*xP6{=ke2hjrQ-Xpbu7gk#zv@yrWc)Jq!AD^f$4qES8$?34*FJfFE=n6bt5
zJfSH<J7V`(g_$-bD8eFaQ6K%M-8>IarFh^!v_!UIzq>cMmdzS%A`Xn013fKmdT(R`
zSYo6W{|=0}q9{W_4glmL8XX6E&bO5O+>Wk>!nttdEw;jY;T+TJ>`?t1>e}?`<@9eY
zRxd83Xkg8j$y$@jo51ID^_0x%7nc}rJ3sGKKOU8-Rl29sb)P#qB<P^y_$Q>gI$Zbp
zQx+n&+{{2wMF2%GHS-<`l?bTX*bf1jk3D8Uob!ZG>euhU=^LOy2wejy+NWHIvb{+a
zLNDofNT`eQI3^E*E~Zq}LmrN~#(}j_>CH)+Q`)bRITcRafE`T5NUKW*QV@z+1Q5_s
zFB@jAW5R^#ReX+Oz6*H71a|P*2tWyDMH04^4-HdH;35{UGcX4DFd<x~Spi@Xe^id|
zn~$b|9|mihn4`A(03Zv5c3PFCEdR`phJImwrOTuRAebqeUS@z3JO#+$v<}&tD*%bC
z*%N*AydoU0-81F&wb5qk_hx4B5Riiz7A8{cL-38X@Daj<i+c;Gqf9TC;J*XkafpW~
z&rGq@o`@n2jFtmSmM?V6RxTZJ_F@r*MqJP=&k$a(X_m8M#N8jwb35Wn3S^j0po7S1
zBc-AGH`KL}))S>KF9+r=o>SfeM^GB_S0QR@s4aAQMw}b5#seUclBQ!Mgizemde+r6
zXa#O*Kt@U?#48g*W|+JHio^;)02JWBjPT1Ec;sW1?kg_eAYoLFh&qPOw_b3^PMtC#
zwp#!IVHyAdQ3bQ;2(H0QJ7r1`*98P<_93e~TZaj<I%mRiHlH<Be}c6llu+d&AR~}h
zY4IA`YpaAS4gK=7p@0$`;WKI%8WDtgqFNDtH;~Mw)j=zheOSH2IoEI`_hja7N`u3B
zZ)>YU>LyGuPfLqoU`l1uV+3#$;I&Vm@dJSjD5=eT<~t#c4;mtA8>knM1#pA>J)s<N
zDd4$ygzp8w5C0al6tvGknDPL4aJpwzbO0EHmII9v+7{~Z+W=Gqb(vp>c*r$~x()DT
z0R;T#7igxme38;}%ekMdg~OtPI<z#j*8>|XEcz31V1yh9FOW5otcl7AFhLzjgQjWP
z;TG_krrny5W{>vSosh;&kxDLMqA#Ku$pK`N21Us9RHulblVu7^TN<kdqK<~!<o9RV
zc0|Afv#ODTjxfV|J1%IZTeI(q%rqRWU%k?;Te-r`o6{_lpQJl|?!0?v*B;m2F<{!2
zRs^GJt=Py6H<6$+n#l$^m;fzVxBINr$)d|wt#UVQy;jcB0fKz(ogMD`Pd;hkUjZxJ
zcm!$|r%R;*As2zx!q{pHXh6F_<Rm0$Onpo+*Ac(~0T^IlA#MRCl6Fbs9i3?dg0^WG
zB>7^(g&?1lsYZM*W@rLR`iqm13C<G6`_(rsRuO<GgnETVtQ2@+@j!DfPsD9q5z4L<
zSOL(mu1DZVF5secF=;vg9cVr>0*hVLBw;)-Q9ui%{$#;eH)dsA7ass3fHwqv1avRp
zBdFKvN2p!lCiNW?wm=&}J6Itd=Xihtc_@o=je+{>K^Wz9D=k38gPEk#Q5I>~qJ5!l
zL?`23z}s5!6rf1iXf`}?zpc&iR@VVm)Co_O)Y2xzZ%%NoM%T}^cMCki<0vwY=VP>u
zI53?!Km@yRJpRdOJL?n#2CkD)`~T10dq&rFT=|}boCCp}0p^^OL{U^umSiPcmL)ko
zx9#qJ9p0OnH}he7&0DkP!^~Pee$(^X9%su6mX)nY<wTJpMT$A+oCy#)1M~mwa|<}U
z0tkQvNI*PW0`I-&RMn~4=bWnjubsN*ni%eC2JN24qE;df^Z^baFT|7I{Ea`}6Tib|
zclF8-=d!xKA>R}Ix?F9xN^c!qQ4(BzWm!k1Iy_%fcHy+@sfyI}caJ~p{_%&8x{<?%
z+V20N*)!aCzj>GYU;p7JZupSFE!RkwDUyGH5O_Ty!~#lE^K)HNW~O`P!;jr-AAI0X
zo8W4@;Cl8D^l2kEJyqX>U4H~y)0)JDZWG310y#h>s2e=Em<(}k%%gl?c&1UG&6qUa
zT{~yCfer-T1l2`g&Xg8_U3owUCQe}jupne7D(N2tP#`BpdOwXtKqNIM)6UmOlf$``
ztPCX>tO^njst*1fO?65rD6^!-8YLkNX_%U7mC4(*BGPsUR3P}%e}w%}!-u+C7B5hu
zzYGBinKkJcCR2cfl-StFeNwA)5Pb8X!y<?2SW=3jTTw@2herV*2tb!zOZ|m^P_uvq
zWpZQL(`gHk01p8g5WK)2gk$tq-yG2>3Nq8}`t$+8!_2S%Ln@c2h+y2m9P!maQ|unH
zuyVdai5}~?(V{Gsjp*v*0D>0(3Btzyg6Dz7fFxhUcyBMsTQ~`oPvFKezLV&Q_nfx*
z&UL~>8LshlBx2ED48`-vw^1SA8SJh-`TOA$qQP_Kk9hJ4xZr;{0Ty?oy+m9s_g(jg
z#9iS&`S1R|9XWE+F6k4aojx0A|NV2}KYhQbPg(vL1HbvBpR~zw-)CQk&+wBI{?0ki
z3&ulT-xsd!`|I0iZ_JqckyI5rS=?&D;FHgB4chZ&fwU8B<!J7Y^^SYwpXp$HU%el8
zfw+Hs{kX$l`=9am+4GZu`)9SS`^7Q-{Jr71@y}2E^X8f8k8@xAvrC<L-ut$BGAYM-
zU)O*7>$U!7GqIIuHJ&I-TRv$bE%WEwh%evX@3#069uM_s*HbV9`(N6KR6iizp>V=Q
zWCV!zDdWq969aM`|BREUFid{T{xf_ozR$Gn`N~!KjoF<aM^h-Vq;Jpn$G6LG|4#a2
zeDd7THvQ}l@^gZpW8&N7dVbFF&un~|&iw9J^$=^NR@#d0Ak3~BYO2~bxw`-Mo87dD
z<K5@$HoB)@dBcRr&%fO0{>#sPC?C!&H*fkBcl1Os`6J-gNwDM-!Ie)0{uJzLMBL=F
z$~c}TA+)@_)m}4A0&<p)Ap%=@c^Ps0vLs+bNiq|vRHkqe&n2lXd11J6{ic0PuE;@w
zlQQ+ICGe(5pl*^tof*>PB<dOu+;*$Xl2YCFT|3;l${LrCmO_#=$?OsmNT{bBep@{0
zDH;a7HqCN!*LA~M6$Ehv>}IVK@Q<>o8Pa~x2MN<@0wqwI1Zm4tBH<s<z<K;05D5%)
z=;2IglwhxH=$Uz$t`04S`b~1h+pfFTO`kB{ZQpxP&f>ME?WmDimZua6{NB7OLv3TW
zgustEQ=lD+IdkipK9M8G)jiYnc;LZZ0ZdSj{)LkPB1izt$Pk|-`$p;mi~_BTOr>Oc
zMxK!xX>cIMPKI*K->e@3CrLVH{wf+bZK3?&{#3^R6Pl+K<9`5vxbJ;^$I{=Uufr#q
zpZoQo7K#Z~k46g-))~r06RXcmy|+BGJQFMo@pjqaT`@FgP=lW1YHf66>)E1>+F177
z_+8{`J*PP`Df~Lp#U|p6=MlU?Aj?ilaoIx$w6IFvYj!_;UIN~8=1ISJ`-02~9Me>7
zyZ9{1w&|cd<IjcfDP>E}Rk{4U40o=y+V#uNw1u4ZxCZk)M0=F&Kdis~=D4<YTtPv$
z9p|~=y#Hhj{#ju?^1=wP24W4M4eJ{ZoXGc;@09n#32l+=jAQi11XisdR6`4&DowY)
zHbNLmd#*_qXRaa3-t&mEX{wW{KJdMuUzU9ag@SP~cJ_$)w6W+r{i7+`g&*nC(9=Hi
zH~~CrH7|R6^|tP-?*#S3_rM?jx_`3q<q2Mf`Cv(QbuT<|U*Iv%J@-7-o}pF{!F^{P
zZo;NKM%01et$+5|WJ^&)oU>;x_Iz+pxwmwQ`H(q`>(QEjhW#hUDM()+KaN1;)h3Jj
z(wL}?R9)B4uhgYl7QuIPZ7-7rM}v2%GtYp3XM^ho{GgtVRTuDM=3RtOT$9Aoe!k#(
z^hv{$+@>65l7U$N8<1ZzgLepS!2{arr_F5NCfCf+GtPa&pR8G9tkZW-T$6FpCi4~d
zl=lyw(gu9<YYp!Pe4~G<#wWeEs!N}fb6QRMc%R6#%)Oxv?i1rpR@o|9v7})Yqhq{Z
zzMu4u@&Ey@(HXw_F@!&9580Ocirm<wIR^4`3E$Pc5&dKiIa93!rP{3SDs^%d6hmvf
zFJ66vOt#wd^_V?%l8xcS=`%K=@#EmpWAe#3rF;@;Zr#@HEn}=!(mJ#ewMoq`KPS%(
zEa>OTwfleeT#2hFm$)q^rH3+h`-(UlV=j~-VZKJfZmAONVXoCbC&y*w=D1WvhdNVI
zqNI40da1P2fB+oZspml|lK{-M3i9&wU^Tmv(o$DaR_dx%mIokF5BN8hT<`L8b7dM-
zD|4_6*I#|Fl(`hU^c83%RIk3e)}>3p_F1}lriTp~q=!o;Qre9wm3dT|#({i;X;+=T
zW2Tla0Mjoo*A3{`PZLI?J9hM>66e)Ppx0!=gQNbYB}h9WEl#7H(9`<mxwEJ0B&4Ui
zQKLp^!A^09j~;h*m~ZJJUbOfcH)ZS?V^5~~nVy)cG1s|@vNC&7)K!;QAaJ8kBKb7x
zJ3@VSW|kY&zrQpG8LmS442~T<;>yaZUAjO`jXp7X8Zfv(KvO1I0(I5ue??W58#rWu
z8<?A~2fW$sKX_Qrf<R5oM6Ir>*5&J73>`Ae)l^oxVu94^##(#e$=+TiUhw*&mC4in
z8ZvmG8!4XDH`Kc$;zNZrI)Ry(URh1L_>PcK-M=sZvJ!xegqD&M#css#Rt=heKBCVn
z$AO*%9gNBr&)DS2?=-zHo>&mV)Yb$ZKIZM@nHri?r2&jB%;7TO_Ly&cneg%G`}rL3
z!pZqZ1A)pPgO5$xwQrBtRxWg(X_KGv<@r6t%w1i}r@scD2mvj`_t@|Fc3R4`^QEmU
z$-Zs>+=Yd8-@gB(Szj)EiImv$8!mTVAH)5RRScI6e-F1C|2_QNhU?fEV%NY<zB3fA
z-<k4#y-cx8#S3JO00>Bx;0STA!+Z#T=Y06L|L#BSyw)+J(EUB{d~7bu^!FrGPTPv-
zXYCH1v$o^r!Qeuz|N3XaZ`f7J<~mD)H)Q8R?b-Hj=(m4{!+mXS`K#Y#Y;NxyLb1Ot
zr8-mpyfU#qun7qftS8)_e`h%7-%GwjY(5vu)IV3U%z14Frww%Qm3zuIwUHll{AWu!
ze?n^q=p#nH{V*RHJIC1^3zL=g&G^r7fi8ac*QOdEHB8}i;o}zrI;h{NnW9l9I^9W@
z;7;f*-q!;`H9KbOhI!S>FE^O5&b9yB@1BzIU9VmD)?EQW$HcLt-2J!SW&&`XgilPh
zswI4WuwsQ<xqhQ;{^d-~jyLxFF#<fWTR=!PkZiWhmF`%w#4VgP!%El=j{34~&tCV&
zhabvMqRfyWCkyiP-2=DZ>I#Psmq06RiZmL@noRZ|IqIH${SBE>WxBup?sxPV_>*L4
z_x6AO>5r}5XJ32MeYtg;3C6*KAq|Cq589I1QzpB+Z&+%GwM+u{(18Oq)+V>?lTY0z
ztJm5Xvb4K4v2e7z`=*;T-U5ZHt+Vm++^*lY&As*Ur>;r}6z#&_J^o!cbkIP%Zi9IJ
z>t|nZkKcEX0TvDxmz;C|*T4PJ{ZIe+w{Czy6gvyh&mMaieBst?+U%Zt<82f0c`?{s
zv{1KNq*Jvk{NU}kxjECOXopJzd}*0>40BCmvwFiu_xgt)DP(Vk%gN4g|M^GX7cT-t
zd`#f+{gt1)NAJDc_y(VM?LXjt_3ZNoAV&@!;O@EQ7B^<pNNbC>akT$@{d)KAXDeJK
z9nt@w`bRXD*;6Nf<@zhWT<_lg^i$OjK&e=0+9Pk40g?xW;J-o=MpVK10$UMXeH`!;
zdUuSxEx|Ir-OdVT@%5tLZF8XNbcrh%IkfGQmvGd~tu)uPx7c1n)4if9niDSo_C$xe
z^S$bc{$1`i0UZ##;=O^o`?*G|(OJ4!`M9%v?yLoa1?#!MMt%G<JWQaYrzlw)QWv7m
zj{QPyT<p1cK3r@&@%55MO`Q5);}6FF#x$x)LVb4<f^Jfrz=t3@WAX&s$qvk_rkcj!
z<e9T3WQQ}=5@D*KAm9D{V~?6Jy>81E_nYUQ(~f(&d@CloiDSmNL&s0JQ)h||QR1Y&
zaAp83j-EQ@cJAG4Dnf~&?x8zwbJxtCWkUOZfAY@~j#sIT32w~D5pK|c{%+m29hRFT
zKRe6)?9oR|px<}+u>1KR{^WKaIB0cgze<2->#jX+k9P4Vj~ipcb&WJ0|NhjU-P$di
z1x}7hL*lh3_$>sE@B{Muxt~1zEj#zGzx|_IwP}Mpak|*em0*u%WrYA3ycs%Zfcxnq
z-;pLI$F1J5$^G{E7xlT>&73?@j`*Y9;Qj;Ln$25Hs9v*ai@?i3H%#CoSsH;!V~X6q
zBZuXLU+Ma*O#st}pRN$N+GikUfPf72r{`XA?=4^Hb{#n6%1g`5%n1f!YGoP~@qOgb
zp$5JHT%UZg#{K@em*mHiqCSjrqvV%!<irVg{L~rotloX8`!Y+QDqFrb*)o+Y5@1`m
zWt%`rmH`sl_)K5{(D^d~w$a0fxdYO0{p!!pxHapxxNGOlc4J44G5~Vu*ii#j!v+u0
z^=~o2_u;D5?hntutm7$eeBmeo!eMUTp`)&>vRt4kP>Z4gLDK|*uQVJ3FlasSwdu;l
zX^r&Jf0#s2*+vv_AmT;eeiAA*f9A;Kw7Ut@i>^K!m>gtk4mvH3dvBjl*uia&Fym-X
z<J4ZA_~T*Tosm`@zI28#E4kEbcjh{MwG8(rFso_*{#}W&70chWw><M<t>J}uq0DBo
z0BL{UJ^Tz`sGaUOS0y>AH}DV87SF$~@S*tZh5ffZuL@)()2Szz%;eoo?P*b7e)|Ix
zSh*g;{NwlE<NmjQ`dfFyHS=wEKirR7WFnNI-2?#1>b2`E;oa`tdrad2u(|n~MM0FN
zxKZLWkS-M@Lg?^8gWRmildaj!J9oNc0wql{?b)CRTF`{TLN`bzO<6M8nK^Zefg+B7
zv1yYY-6jJ!m=6Ikj-Nas|A=~b`1o-X{wasRd+O{NxBu`lX)u(KPkkYbE<5f-%|Zqq
zIdq78IEY|{Be`r0Wq-I+ngFyIZ+!TP%1SjV!T;99i(7cG?7fc#3@Tk|MVVVA%?>n4
zU}v%b7h_M9J)-x8;oPa>v+l_kUvsa%`L47`Rqg|Uj*^mMcTm8oS{e~gC-m?fSJ`5j
zk0r}g3lk}v#x44k_(i^x(>lLG{03lb+ObQTqb3RaN3031J#^q8jYpBKH1=v~%g)Md
z4MKylX7d*J@;mRV{(70Gl}S7CnSel(n?HS;8zL<a^zo`M98Q~ODoWg;W5?~-;<<Cg
z%YdK9r8!x(euMaY!foBTNB%vDx(~;!9M>-#In1<0=FcJRPvC0;PjFTb00@NvBtEx?
zRlbOSSS<SrUCEu3p5{^CxeGnbiL0l50(x8y>dZUtO%4WjAvz81vKi2uC|>ZmP~XB%
zz3r77AT!*~Qd^*-xACX5W9iOvo(J5Up3+pI<@4mP=YMzfSzFRQ{=b-{^%V;9>r(ri
z-b?)cyy@Jfj;AyAJ2i9n#G-DYT-)g3u6+mG|NF<M-Gg`DDwCTLRssO>?Qh&JQ>f|g
z*U!A*&Pl5Qjg6hJuC9@<My(0uI2NIKz`Un_zkHeHV6G#_9toLYG6x#X`NIbfRv6uY
z4QG|#W0)dg;Xs6N4k5_mPjOs6CnJXj$->Y+poKvYZq~x_%cq|e*r^f-s1gv8^RfhW
zPpW<%vBMr53(S4A;ctE^w24;WAHVmgG#u;Qh8^48vv0g<LNlgfwes;8JuJ9>g?80y
zaZVQ)!NL2m006|Lri>r!w(i*@@WTE`EA85U5F<plSAgb!`QQIfIR#1(m+H5%fhbr(
zbQ4N9%eR9tzYwNyWS@tU<tsyEDMA4^H`lutm%VGzt}O3|`mX<52=g+0QY`m}`e)`%
zv9p*~F`m!Xed$)LTi0Sb=kFz^RpsJ+mdxMks_R<rL%K9H>1pZmeW{T{e!YACt#|C+
z0y^>KfPbFS(4>jUfxrOZ(2^Fr5#Cn^b1`zaT&np?TZ^UdNA!95I1u-mUA_xPSGr;x
z=*UE&yE@iW?>O!D<;?)^<J4VzZmXTX%%8T#b1}!-;sK7tm%4U^`(nNnu>);gFI={#
zzI&6jo@&24+YR&UQrjg3Ku)X3c&W|xiTa%yZci-g<eX@a_8&g#e*U{B-K4Qa?&ie{
zZ6}{TV21PPecy1uc=9i{d!3sdL<7K|;vawXqn73{-+{6u$m0j1*hc3G#++*qxFCZz
zrc9phT6(>^<N75gz<WWJb6K&Fv;YGLA-7tgYWd{v*s0TY+@Gt}gX9T$m}pRxV+`cF
z$W-KnItk`zyb+2I$b@LiuHA0>#0icu-MMs$Lfc*=jmHl6?($EiVJI_GBJck~)G2I*
z@Vh{$C(ntsFX0+9EFegtJfyskO<%}=f(D>Y7q#eAnz&L_wpn>TXv-5?3i5?mz7Iv}
ziJ>F|9h?JvqzQ!J-+^Dsz=8rr8_P6vF5d=q7-xNr#vB7K{v7?FUuaN_Av%F30!Mua
zt;pZYoXtRkct&m$JFd?(X@#m4<N%;JciLo`m`!&5<#U5sp1<dmCG0QG=goF)^;!3s
zYe4XuTqKopRHsao$HUkX{S)E<ki9EAzoDwGOh7)~2{3#<`rdsUAOhLtAnzP}Ij$95
zqwjOzay&2H2|6y7<B2KfJ3U`&dzYhr=4H~Bc4jQevLd<Ks8cg{az>UC<aM(eEG7uX
zGaLaD0UNExw%vQ(cKIMopE%zA@Z0y9U&3VV=m#2wRyaf?t7l$)Rho`E%kzLExi_i7
zObKB*tnt7<q&{YH6YPoyz4&<5D*1Ela@mS_0O7=3C5QGJ?Z5*>!eB%{CD6)bOTv7u
zt_S5xASbC6fS`q}n?Yz-OsG=THVF&yFKCoDKmn*^?vvzRc<U|sh8%Q@X3a7m5`^mo
zGo~3Z`u+3I8vyVlrXPR@`T!WfncP#I`U{Z2e<VC+i~ePRNPX^?t%zE3zE;UZi<HpX
z-$X^cLX|_#5S;HxSSVEEn#D76gP<L``>xyEjn^zNetfiQjr&YySHt8B^B=zRkadH0
z8x^sOoFjFmtznEgSzF$k01{Bf7)ab`U#O@1^~B46b*1I<FVT548Yu!%^)gSxAwETk
z0Pnp1dUwOZ`F8z}SFdr)*Q|3zGOv2<oA+=D^V<NhV9v(9NEFEBI|T147(1P8#anhR
z`X}N59JpA}LEQ*+z=g|2m*)lUOFOZbi}M${4!J)v6FcYEaiMnl>bc(dyzr*5<n;ul
z<z@9mvpw3T5AWNf7l|5gmjj&w9*LcF4r})+ZJcN;oG!cd|NI{wS7_EM_piTs!fLUT
zj~~GYD^|O^Zd_u=P*IbOyHvS3$VQHz$B`4KE&l|$ILc(ALp^r6O%tG9?(ptE(V{$9
zz`V*65w6P1N))=cPEL!du3YWJE@&U66@d_q@ET}8&KQV5z~?%I(#6CmF&vIgZ#)3S
zoSXILH%NH5JQFJ4sQ!_cqN=9RO&K>zej{5YM6Wj!vTxk9)FNjUju`H86*l(xsk7#2
zpCdq$jz&Ul0j?Nhiok*gMTGSQh;Thji9Dr93xE(_OMd_@=3m06Tq#2{Qd10I*2%Fv
zF_!7ww~;8`CaB*F6>*H~m5KjvDe6{@Ox=dZB4|;TNsjF5P@=R^fFn;SviWT&6^P}U
zI!=N@@`ccE=SJ9gs+4Eskiz}yQr0(2vn+w2t_k2;r08RIeccZtdDSXu;&6o__LOAW
z9`Q$JZM1;~#N1f)1AmV+ffu4&B???c6mcNpK;Ps*nsQrom6(Y_(_L4qN98-gfvXvG
zbb>usq)1Ob^BG!Cdn@6u(w6C1IETuW?z^QD=mw=xdt<%Q<7?~sty|q~osAt!ka>>n
zgkPXcS5DDhvT|F*6Q)&!i5;Qr<ld~vmmzd4gu`^OtM3$1>VO?XJ3tgF;^g~8afFf`
zGjgPz2NWQ1dSdtAJG~uHq$%I_wDIFrU)i(;Isg#}>Pr<>2|p2}>Ai;zTN4Pq6N?J<
zJ@6j@)ZkB2I3f_7S=<^F+Zr;{9tq*pK>)|xsahHjCIbEtq;I@-zWedR4_U+~d>rt{
zc;cnMny(6yb6w1I3jhX@UA(^{Ve!JNSvN{irabLcL@nfeB06EV<f%^Mw*-9x0@P^~
zs6yk?tbW$XA>I?(ByzR@NFW0y0-o@U{wM3ZH*3P=ONY&8V@L>i01Pp)g71(wbF%LO
z5daBJ0e%2#&>lItL-1+N^vQM}umLd1k-x`WC644^s_wNnzp4@7z+5U(_lfYso`gvQ
zo~dv20ZkKQ4*m&B!5=fV@|B{`hyxJ^`V<F(*`;e;W*Uen;y}cK?&AOnEPE@p1v+|L
zU%Sg!tuKt#^S)EzbSa!9J3hkdb_b<oWm|jkZ6PFHJ9n0uQVo{r&Mpa~fDCs1AG+&S
z6F7hW+{>0X0&}U8C1(|pYLaOH02yUUVwWTB$M~W`_pLi`xA3@tg=8fJTr_*8GzkF^
zA-n3<J-Z~7Hku<dPRm8Q1|eRD>bmt3#51yT<ZE%g!uw8gD`d)p$<HZ8e3~u40(?qF
zDr_uKs_-D0G-ix@<nFuNQ8}O=KYhxy8dJuNwfq<e$!QYS(*!=!#6vVJma$wyKSH%|
zhUynuI9$LB=4FlYzu*|rwn&`#iOh-E#Xl%5z<ilsk*g$M*)De-IIKue1@7LPZ!$Bq
zExY!(cRpQVY|W9G(~@~}TYOl`r5)IN=#Z<GcEPfcOYlw+z-o|5)KHCuh*(gb9O>%>
zC<Y1uo>sJqCVj`@95X941o%tL76{1G_0hcGbkDuIe*RobBnWV*lb!pzP1_{st3COw
z+;;t9D@*%~;q19$(?Fq-0EmgT?wB-I<BLX{Bmc0$gLGcza57;VJz|*U49U;Qb?dk9
zFjK6VlO{?F#Pw6%F<l?fasB*x2FSR6t@^xa=Pm&rJ-2EHGb@hs9R%sk@aRu(2xCV-
zB8oT=aiD7)sKHM>eT6^War^$k!O=OUneOiHX44(h-CgJCX1cq39Mc?3cMM~O>8_dP
z=Xsvj@B1%&UZ2<XzV7?JB0;q{+CLBBq{5pPYMtc6&?(Wl1uMV%Zd@&yjMv>JGV{(I
z=T2yp${3|`&2Dqcr-_}n5IIr7?&rOen9HtGp3lyAMwDhm=WgIR3E`ASn^BvBAD6X(
za}>R4;E&>BFn~PWy+#g3Q%#YRLgYSOAnH{kwaH`?@kE=!YHB`kEdG`b%`cxg3o69E
zry&^6{Cv8{%C%hMO|M?)O3Ov^0sVwQHZiBTd1uya|0Q2t?5o(PHj@_;;eY4+#Q{y9
zany<U)jA|Aj{Qs5hslK30}6v0PJ>j6a(?}s_drtMKp12sAGur@5B9sra`!tB={kpw
z-KL_p#d+N5?e^<+n)P;>yr8vIs(jt?XJ-kJ2iMH=yazW;b$RO#U&Xx5z8WZxXf{)6
zbR~)%97`RU89nrU=vH@WOMgsXUTns<=In!cq^JTe;{^Pl%9z>Mo#t$LVxw$!)VXy8
zIii2_Do8zvT;~3waOl-T^X3@XB30~FEIMfiP*}(@BZ2@NUzEd0S+aV!E{u8H+t;XP
zZ`qYTMn?`WKvA}T%}1c&&)J9B8Gsm7D!jn!d+S}mqIvXD6>{NWV{V&@T?Rp8qZ1Rq
z{UEDc`Y^vxVa-6f(<B}belY!Vz@H)AvYQ-RKvewMMII@=Z5{`!v;|^QD@Wfy(Sy0=
zt%%$11Tkt%BUMt$JO>vX^4_z+q`P0-KJ7B*ckF5EB{mJKJ@rT=fF9lBxh30xh3k0J
zi)Bg7L(~g}_OstE%IL=;KeW?KpnG8oIz))|1FiwOLp;d!@bJt7e12RP2^ST@i^{3r
zC}m%8?znjHh;`r(XJ6ooiUWJ%A1>#b$XvW{wR<CIRJYNj0`{3M_Z8gU%>y?B`gtF%
ztWg%VjvI%cWI4tTGxMF*Du=Gbi&!VSu)8&XET^K1uQ8;S2SVaCvR^6dE!yNYgn1A^
zec{pdIL+Gp2%E*zX6|Vv(MVbFoYp*`@Pr|vP9^<0l3dO(jlay=19@ngj13qE8ksCP
zn)wY;k^PT#*wcg_tl@sg(ZkW%S~W*o%AX8%%!#^=Z$$OC>zT1g_fr>_jqCz75Yfki
zP3Snn<&Z<(9aVzTc&h4z{-p3sTi)=?Q-*z~0m1_WXxv!sGBvm~-=*k^kqB*+0W~Z^
zQS>+CAFy!V)vB$=dp0y)_9Jr#%nN8wmLK|d0u(|&Q}fgCWMM`5V2;SdzF1aal<4=T
z;#y@O!_06iEBl8HYMc5jJ8oa70yVJym=@6c{x+S_e``+vfGP7P(8B)D)4DA9kx`JU
z_N~U`-2c2WFrg)d`;l7t#%;A!^>TW)9R<%Qv*%i)1tow9(p14K^vQ-TBQ0fb-rZ1o
z*`x|{PedT=+D9^0l5U(I{zF<aW`D0q-%U%pC%QQ)-0w2Au^lkxxNCGLcCnqqUu=A3
ziH1~rBJ<T61v3nO<P|DrA1C8#K(b0`xSndjmhxP~5g>S)0T<6nqW|Wz`1C6WhZODr
zlc2q{9(EZG8cQ|ljfHXijbSha?ys38CeFfX+G^4%D$mE@nmF(9Sd@q9_e5C1#Ccbg
ziT>DgWlBn6Ph=}0U(7|pa7^rG%kfJR4DD*K|33?W?CbGCVA!hA#7dxo^7F8rDHD#M
z9*~kzQM;Z0hHgjs8rx{)L2@L4-G25qBKRZlqiaLO0PlgA<=&{EvD2EM*k>}Ns#()O
z839$A4`8I_QV~TA!}sca_1ludaDch2&!-04J%1_Qu37TkPbNZ7RfafkRyF{)J_mKH
zA&n5u|61#RPeCMD4%dTD_I?|uK^*P98<b}#iq?0M35t6X75^aAmF>QRqK2NPpZ;+J
z+Ie@YR>l92&8VAW6T<xUv4%7WN9QMF)NFiHKq9u|x0uzN3Kro%pjv^0WTPf2DRu*F
z9%^iNX0vc(_oK%?rqeoU(AWK%9Y<DJzu@>+Gzz+hF+XC2s*9f8>;;!q32B4_7-!xp
zILS4a27O1s>2l~{aOLwxM={HA1zfZ-g$G+)O}WQMYC##TUj30D;Ab4ZWvTrmU@#fM
zM~T}^G2XF^PD!y3GF72DpEnigiP|A`bUdODiR=nJJgPJljvuEny}tiYrXJuNmJG&T
z(1=;Cpxxqs#x7?P0Bpr5Dy=YZEYnP(orDrDVvNVAclw1tVBC>{QI7yZn^B@~NKnI-
z6X4p<`)nH9pcoL2e|GA5C6j#9({{^!WcnR4F$EzmQIPpBW&93db1$m>`Lk;!Ff%sF
zs)bGaWH6$u?()Kdc3j)PPqIPLS}eg`%=tTtzVI}t?oN<~u6#U=UK1yZ?E&{i<1`qa
z99s)V4P|Sef=;DC0b4H&o28gf-CD4fb1P$*%L+BEB*?PLfCzT|m&6JA5BtB^!<exK
ziJC@@1T@zH-r4tA5{+NW?J_qIA}!h2lyq(7kKh6GG%M~-Dj@()nGq>D8y|9rxO9aK
zwwHr8CKqK9*U8SdCngSNW+zc6dG@%X5J;lAO0NcdpGH=S^*{mn1t82{a1m1c=la3_
z(sf&_Krh+B33Xl%SG@2`o(LKCD%ctl0i->q&XgDR<ENSB|LjfcI!6PR^0h+X4otE$
z47~dVyzIXU8kNE9!-k>HC}}z+tYQX==4jI^21ONu5#4jlgyg;dSsRPX0>a(sD)$Tu
zHxUAhT`mSEN8gto1zdb#YY{%9%Avdi7se%7%H0{2q<xAwFi9)Tr@Jg|TWF$s4hSGD
zeUT+?|3GRM+cod`#yt?E#2j@y#R)y0HFLI+cmMVnp%Zd1RjJ*2P_|&@=-*U;K8^<K
zk{CDd$ifS~XmEwH{K-s(*m7cL&5<);wc2qjSS4A*D`kQm_B23Mpj=Sv$wd@{`j;=$
zI!o<v`iBkO#l5s}dI?X=(ZP!>=wJ$Ydlr6q4JJiI+F!$~MN>4e3=mA?Eysbi9;gwL
z@x+Qt3qUX#FBk0>M<_&6NpWA$JoRNVK-|J67n_#6^8nyuCt}ncy!U#rs_6JLzPj1{
zfOOErSN;{QuIMYLYU~j*PTr#RN#1UAw4lC6y!#Uz<>V7;QTRq&LvGncD}XjoD5XhM
zW}EU&+juPm!?)IpIV57!7h_Af+On<EF@CPlNQD`cJZTjH9HSxdOvDOB=5ubtu}Edr
zoW4&7Q7L(Amq+gRv(fwj448b+2h&SYZOVS%5z)%i_sKk}(Zj~Z>oqsUA%DfCz3kzL
zFO-}lpmzG(hcf@g^}qU<99TwnN5@47FS143O5(CXj}Jj^wDp;PBkER2K-tIkrN8^{
z<MukQJQM)5iJq<D7ji(g=Y1=sU~@KYYE_j;36NFTPU!So*($?PS;I_*hoF5x_%rJ1
zi8QDod)05hyhN=Y9Bs$`x5*k+#quy|MhIm9)G43`rvvzK^2bIg8rf0ncoo~J5={ZX
zbj6KNr6lvua}}Ag90hD<4+K-XAn!?mG2LS+T{ZjW0~IeuK;e!+?{UhsM+>MjKD|ow
z;=2XhvRYLzp%%5gFAp+k^AJG&7BEIbzvjzbq;#5lh+h=FVX}aOKqfP1pTds5r<<kK
z@5((Rl}ZJW0`fGN?-{^*A9X7)B@z>JzKE>!^Os9@^Lnru<AEjCE$^p_a8RJ60=dW#
zz{S19ghkm`&JsXTObF8gWpHIo)L-Pn0;XrR%#L?r)Qk@@sh<Oz{ULt3vwG)`nfn57
zn(j9g-<OEw^Zzds-W+!?VsFsuLG*OpPgZW=J6eE+89$=~y&#cVLF!cSsV@jTn-h>s
zAknA;ZFuJRYv>c%6OzIXqd#5=J$nMb&*xjxR8jy6Ak<@VOF-Y9wqou2Q_X?aK26yI
zrWlPaT=WsRBb|S1fQMPRLW!4?0-I?Gq_R_dI4ihHmn&e9Vt}%pWevx15a4ZX78}7A
z<^YJlDnA7j0rBHGWtSu8KNkXUfQ~Ae*n18@GnSlm7|Q8Dt!K7@pA*O|+<gq5t-52p
z{(}TCSwW{y(}Rn}|J^<uT9z(OdOU-!s=#jx9XhE&$!%-C^4abLE>SQ?O9?9gl{i%^
zJG0!<rXAl-6F|sFjFt8!vpR$bOMhj^;YKuil=gRRpAd_aI-qPmI)j$He8Mw!o45hO
zdxSH@#L7ISQq58S;fheDp*q!DZi5%N%?`sVfYqX3YRmf5_RqX2ZvT&-e1|l$BR^GK
zEc=9;lIa&Bk%JNhh!D82H;-|mX_eeb`wP{<GJH~v;i|wF`_s)*S&4uzA0Y28F(C!&
z;Xgf}NVK9}Z3$)JMnvJ17~&TAi(NTeq!vO$Q1(o;RUyOS!R9H5(M1@5Bs=Lv=0pT<
z0MiieQqMKd7*c9F12!va*+ec!Vo0pu=VQPN-M89CwXH|+tT7e69SkM$2SiY#$f`I5
zbP5eKGiAhMmWYZd`@!j<wX6vAXpW{npjY-V3nq^oxAdmvNJ(SlX1c-fV1gm%#9Jp2
zjfn}!1=$I6E!4A!kt@#Mm5l~s{JzW&&Y_c89$(}=Aj!eX8yL(Xgi<!;21ra}K*=j&
zwL;+V1mY2@CjCsDq_Ps>^k##rNGb#+Er8iSb9}J_<+t&(49p`V?hR%`^Uv^J@Ahle
zJ^79-zeC7(z_*9T{mP!o3;ibxtg5mJMsk>*%SI%@YR|V4LLKp8hyBVot-Sa;caO7S
zj)$HL1c2&=a(0*g0)MqfXalhmaDX2m6)EK{-yI@f?hW$*3cyn&!Q|nmfKmcr_|qYx
zdwNPw8>P*?_v|0@Q5H}7k&7ba;MfYHy=8yGEaeMhAl14(N5!WMnI7EJM^;?d5C9!i
zn`sJBieq|%?n$XgOw*KA99wYcSXf5EupEIi0kO{zE=dKZH6!ozXZ7#eqaXl(yf)2=
z%!67|QJ*LqmPkY(lUsESAkLGfmLd@-u=pKrr$fCDj80jk0LzvRAR<u4FXH5lUmgNh
zzSzhZX&Lhx88dR%7}@cK?gUp{VgqV2LqM4Own`jR19M%HIUfWJStEH4dvQ9(n=6kW
z;*m+W9*yWZv@4G*hIy#LTu?G{Pu{Ux1b7)ZqdQqKd@!N7ExSG?BFE=l0Fiy6<6%~X
z_I<PtL^wA65bYIPhUyJ0dJ+)2ttLG(Q~guJ4kCtwI_}+zdu8^8W+2m(tsra2&TMek
z*8XD1t(SjlqvRv?-9!H6Z83j5I#Fv|N{CulS4a@CFQr0mP@c6n56LeHD9DyO+7c0G
z9d*Q&2j7-dA<$+u83;v$R1+}Md9`rTYmjAOe5jb>*b*=)N^Z8gifSSY%m;@1HVxjR
zKY?gY{)i-I<9Fg0Asb{c3F_sF#{l38MO=(R*MrJ-J+lS2*x?c3J<J-?>#h^VrX&MW
z?hpBkYUP&uPpahL`3Jmt2cS8BM8vQlhZ!G4h9sgO?;wDNIB}F-nl~Aq8+i~9o&eOa
zZ!s8vUl+}TN=zuB8uL5}F~XSukRgeUr@tcX2k7e;>no;s2$~^a$fxp1s1m!RJekAC
z3QpA;8PxF5)X0g$efpaO2Q%b;)d-2!klk0{8>LAqOtiB+X!6o}qPHi!`QbDEmJVh3
zf&-3Pm&^7z6gHcHDZ;Mw3QLmygU@Uc2>)V@l@LCyM<Z}zHvPk8C}*cB@6zh8<G&}C
z!P-pU0(_qQA!kB-lDnNDQ9uWI-(L{!T>v1-qM(}HSKb@}ph*^dtOu2o+I&W!0m!K6
zEeVd>ZM@~YSJ*C7up)iG_m61N=4Bq_3{=YURH_Yi+_^KOsl`0fhqBsoxg)aKwve0a
zy~1fk`$m{e(HqOf!7&t{qN4m1qD3vy_%S99;rM2&V88<>B&Y9{mza$eBj`Rd9EpMh
zL)E0FV1)zk4Ia_OnP%}HCbN}BYVnQwvu)BPC_8ks9fKH#V=v<lw5M>dsu$`m+HHmM
zB#<|yhtl??sfZlD%U`)*E9ve|b}I!*uPuO@W_eFd6ehmZfy`|gDG}(UsQ$9DGHd75
z!NO6UBJ=liI-?}Zn0Ta*kn`u=c=*z;0pSi=!_w9cLiI1zY^GZMQYTmIS)*V750Q)7
zT*2$Mq4jSZyQoqC(0<BlfQ2XPZ(q~{HC7A_%DljXsh||10KFTF0RW&F7K%H({4u3K
z@s|%5OFz|oYhf`2@J`59t}l@^CS3-Cbqf{wnpTrlvlSnw6;xf>bgi|qVWCk(t`KqL
zPv54sk(Mh=pNdg9xfy}h&sO%gWFL-n(#JJ3@$See!MPOSN3%k}b_hTXlJR}V8Lyjr
zkswsCg996i^B0bWF0U4I6$8Nt$_yK#FhBuepsZqu^&9*~N8KXDs~N5#j`K`cur+Q-
zD3r%d+-`kfO&sk*vblbk*H~tdkP?K_%M3Rw=p5;)%Fbds{Xlc%IlvQtzdG26kPvHU
z>B&yiXBW5_1Qi6JM#7uGWKd@z-eE6Xa!1Sdfk>1VVDtp!9E=pi96p)d3$*^_C?Q+#
z|836B-6^x)JvzUZ7RbMi;I38zvxXALrTjSO<4s$!#;xkgrnbV(=?97HL4lOmC!8&z
z*jsb9n|NSpUlXOAh(2D`h;U;b%rwcYA<iC^7;Nqz<Q1QusPGZ^7{a{EpuJ0hQZ>-7
z8QyAXfe|G;#tS&Orwfi!#9~aM84uUHCH5!9S*e|IAj-aLxcR`Nj-nnCr4@6(N#bNT
zXlBAD-`&gXMaB<G1we^*kT~kSFW+H|#mhkl9**0GB-Tn3M>Qa<z$9~(kh;g2oR&9K
zhCI<jqWtgcH)lS*Ew?=Obets6i#-vdE>_M`JHN;@v;0s_V<bx4r-@b>n|KFGzV3w)
zFY0OA4Fq$GQF6K#<<d~#?7k!;bVi)MQ?HGiFgCV}sMY??34?nEDLfIHq9nzI34-A-
zK2`Q2WC6%p$8SH;f3zDT(<Xe)>TG5rvEVRCkPKuNHN<g9#kvt5CNCf=tm4$7d?5bM
zbWO1ZE<Ait@F{hcBRyy#HM8I?X>bE%H*@cl`g&KKESIai5sGw9WaqpN+Kts($U#!k
zjKjcyrMa8MIr{$SJ;b`7r-Ml%5iObqa)TRL3)JAq{IJL>LyrB9-<&J~At&yFIr;{h
zaKMcWKG3y5{fXn^@gkfc;=xvp<FbT&6M+I}vh#pKM!=A?nv)2@5%3;B38#lGwd9g8
z03)c74o`}heP8rxYvzeg3v&}92Z(ZC2n8`<Yu8AsEfJ{cNQn|-E(B*G$vD*p&QcO5
zuL!o{g-;`<0j-TN-r@d5WXSofwrz6VpD#g@Qk6V3r8K@$mE4H}uYL0_ef;^8)eRo7
z$yH1}nwDh8pFpfl?Z&M9LH0^kUCfh11W-l;sL`5@Dk+_*xSVx+QoO?u*8olnsEL9x
z0&iA7+~I%v4*3t5{*R2!iXn%dF8-U5x9J^+7||}Bp_~77Z~atjxr{8I2+(KLWaI(a
z@_=k6&_b7}eLulRCo{GbaqZ?_fFa$4Y4UiXd!Yw9&p%^yQ6%E%rPfQnz&srQR93R!
z7}V39We^LcS*l;&b;^DM00)LV*}yS0u!GtMWHYHW_H)!7OHRb|OD_5tj46bfK4C|u
zieLb$GB5xsYh>YrMu}>e5oqyBlo$EOV-X?QU&0S6tP>2L(SQzij<e_rvlXHMz4So!
z`E-5+WT)_eyg8gda6%{oE+g-O3ZF<>j-|r_`BWZ@dx1n_h$=f~NqpgQt~4W+)QsE{
zJY*)<U%>yUmnd&b6hAd@N_vD2S6YT9%-fb7<P5WrA>#uu-y9zpJ9G5GrP*(_5*2%v
zNSK#QN!Nm>NY++xn@Uqt%~uzN;sE4ciXDx9c_dX7Ivr>XU-Vb`|86Ok1&W-Ir&7Aj
z{~JQ}pM@N(+!gS1&%RU)ON66s;EM>Sw2wMG<K`DxoC`Qgi932`K{_3p!@&^n4#bjS
zMrU5UIRSy`P6iK$F6H@4DFO$4^2LmN4$#?nJo=xY%a?jB@J5EF3<uH_YOy~R+OuKl
zY@-xIEA9Xq(34|L$wt@Ybk1AOx|vq!Z{p4(+GO2(>DoP-1KngSAk`|*dw42ask?1S
zIg*d`=DFR3pv@n9+f;=y&GwlMJK!wp?CkV$v!(F+R69zo#&2BhI;jBJ&`DSk9h?Zr
zENt7i87dw=Df`&(0)2<_^Jm2&z>{ZqF<$YHTZ&2@=m(nXM_r?!@&VIo6Ec-W`g<{4
z#F_))3SB$DpAF=!u}JU&#h7BGH(X;%+WOHKeR)w`ca_-QZIa6hUx@M|&7Ly2hQWk6
zWN}YcykQ|J<#VW9&Ec4GrqplK!$C3v&cQ*l?5F?xg>9n{_LWyp@oR)Uy7Rd+y#+jA
zv&!~>u9qF)y_~Hs$rk<q6s3Rd(QuL3(YX@<+C)eWwR@NIg$Lz{;6;ZR>n+LA6~bDI
zOqnJ2WdVLb26aG}x&Xj>(kpFCVoNcA9=!uDaCK~0oE9^{SE<F0Or3v6K<E@&vuy)e
z#o%UGt533HvPWc-(AMXwX39Jy#JtA#Q>vj6^C)-_C^O_=lzr&O>=R5srV)N7m{S~K
z9Y~8!@tr9#{!tV`8RlC=OSvtum_~yUDwt6l0#B&wPMSGB@Q@9bgV$tCNa{t}Ela?u
za_Atk534fT(eEGYZeY1+;~4RT!)(q*pqsU&InHTnmt0M=)D%ViOC*kesI<M^n|%6<
zbXJ?N2rs_%=kAUTF^of#`v>F^M&nTwMfh;$LL$=iaR8p}%X-awyf-vpb2Xc3Fg9jb
zzc1OQQ#lCSs(PC&q7nApHb=McoWGA!?UQvGi6=+vSx|QhVCor!)v}1{)@lX1iFB2;
zDlkQvBE4PFn5Ta~!F{-!$rXhY5}^YB;c-sfu0gPs$c0TH_|5kFOuLbO&D06Y*v+KE
zfV@iA;($r*L9i;-2GuO%BP9`jO*RQo5KW(aIxsk9pu%qJy%%L~ZI)OV!rfaW2~|rc
zEfzlLk3W*qj-{%c$FPMMmeBOHnA#!e*HI`L){_9utg>(APp$jGH}kfnSY?Ere3>1`
zl)MVU=jrvUu}yeOT9=Fmu&qwa;Z=5WwccrmI|5}wR-n1Jte~Cz1p-yAU=F7Gjn%Go
zK{~^TXWjxF6*w0O!V$=8zl(3;^=<M&q3@8HvPO||;IW27ZoIUbGUweY&ii=A^m{)b
z3-<=5ON|`SlR$L%(2C*9BB%E0J1s`DC4G9R$=p^I-v6O(t$=H}1hPsrYfn!s1({c*
z>fUduQFC1qFQx6R?0Fu~#8-dC!kCd|V~rqFL08O-jj-@V^ge`!R2p=2auz_CbIs;E
z0Af92XcxT%(ke=Gti{95wYRI-1s;kCQzEk3_B{HgNP}unu(Ziqzgx7eUe@R0ADjMH
zB^+RLwLMO|<rFj}kGZ;ob^u!5HWd#Aw%V>hv#wInKAs2rhPX=Od&zq5Gn~T9?a{PC
zMDx(GKWA`}5%ELaEJ*Irn6lIHNcX~)J5{Q!sj=3QK}z#B8o}SVVkYj5DUqc{)l>Uq
z>j5Zp)wrA;Dbl_z4XMd=^}xA`@0wRGuB<1ycQlpAFtlqX9x^Z{lzhs{%37^lYVYH3
zquP(%0!eZ5W3h>dvk;A48TG;(%bExZWJ@2FQ|YXvEf1;C@0<L^K3j<8?Y$4B0XTnU
zt&RNIX59fglmK)yB#I>E2nt)3_s8WQO%+nS7`Nk&yxEaEx3iC{<nue^Tny2Kh&4T5
zj>jO=BDFnMJ6pYxeYc*kZ2PVL{w(W1dNVgV-?|{}Dj`+n^^efZd9A|gHQLR&HMxi}
z>`<|;&C$OyW#ehX2<8+Y=J7XBq8vhmBKfEUajt6to*yZa!QQ}ANVDquDv7R7)y6M-
z5}5oQ1XOqu(uN6KTO}f!1Is84jOK<CaId)IhtPD{07#fmHUJquq&a16F3P%tr3$U$
z28JS(3o@?JWa|$(TKT|lIZNqtF>3DfcSrF?Vf9_H1psJ(6>baD)sWkq!Lfx~%wmfb
zh_@b`7|N3dMSTKedi+|VPuvM_ogz+VHaE(7*DC+>ucP4FvGAT2=)u~WRy@g4Fw)|K
zY!J{tj*|8zwGv@UKE>U4Qz20uIVWK!j}(Dy=U$>Eg$>4KbwtCQN|`*k@j@-X`($eU
zgg#szP~(2*R%XR$Aw;Rz%hVv~E4(Fq`=vR{;>!=xdgSMK#|Ww1&XJ=wZ7!^h--eL6
zl^6VYclBj1+2+ba6VZ(012~<oWFo@UybOI#+|#wS37KBfJr;At{8AEJuK$t28=?}j
zIxUQD4Nfx$Y|ffzS0h#g?B_>6XK{GPU#=Dkd+pUR3;ImI#*49YBl}-vUrTq8E?+Dk
z9aY+e-Tc1(ZK4LTH+y|?KG>9G6J}P<6>L2_xZFP5Ipkgndf>7D?#f?OB;n!MTw7J=
z-mUvCeg3^qsQV5HzvJlUwbI$o5Q%I3wXWvFd7&h7&OhB(BWih`yYIGUxAybTzEU_I
z#8YT`nmmrE7r0yP?(UNvS5x%d>nBiLFUH^*7c32)cKLL9lv`a^vfr&Og((_fSzdK}
zji?!7rL@Y$voeV^{+`>W@4fmJ;itu8pSCow!`LR%(WW+3uzgJ&8*IdfiLL$DzSIO!
z#Mz{~N%h6$SMNkVBimOGl`C#KRo5X}+A~l6m_P<7VC>v1(q!C7b#tu3FBzK&#d9b!
zfzB1@230lcB{P}mNnY*}V6>>;i|&$<&IALPo1~Y^0t+z{FG-||kR7ozG9wz*Hb1a>
zlbcBld+-p@AbwAhxoR^0(i|97l_V{=Of!JrOK*fsfr$Izs8;o3(t1dtgi@?<rQ(xm
z$Sv|NInH-Dma072It|GMzA7g$01V-KPyY&I2Fps<sUncfv*_V7Gnbhh8Ys^HTO8#<
z$L=>*DL=Bm;hOciS<w4%gDwFtF<0fj=&<Nh)uAYYU^yyovGunap72U^j#szANY!FV
z4wn+Gpz3TYj<AEXRNmmV>h*f#c`QAjnL7ka2<(Dtt>elz`jK3wallVRyTX14jq_tU
z?lAWguO|gSmGMe+YvD0KPkg8a_!btCnP3`^)B6ci=LUaQK_R<jpR0h_vgVuf0)end
z$Ul`ER^L5=kCg%4!Phzn7(N@`ap^=tzD)sMKZXDGw7jz9@GrCKLt(M|mNp$px`AxP
zj0>zL92-CeyiL!4{OY0$X9EZ>*6M{70lec9j0rurW0lS~naB6UZ*S_v09?2uz{A8j
zk<a+Z=&}R1&0|JPcTScg_-nqg|LF&x(<i?M1M6~|=>o5aQn{Rro1x_mPh>0#H>~x5
zHfB=60V>x@We%2uO!oe!)#s3qr&LH<Vnc4Dh=MW!o2xX?QRG>?@1-8@{j2a(c6Pzm
z-b{|2*@%O?Y4A2{o73`{>sRM{G1Xi`{p{?V-y}RXPTzRxesPG!UoU3P!JG85Pii}Q
zuLP2?{p!DM$T%<Ra2Cg?%Hy^e)?Ia)@BHSr=AFuGt(ISs&_oG9MI>qU>j<9E;dfv8
z=<{mqvywAnJ)Jt!S5^K6$0xDEj;v$<pJ6=*<xQhYPdx_IvdwkckzdLw;CJ7}Dv3Y;
zzi(>?X>Vp!{CmxEJ82vPW}?GWEb(vWbK<&@2*IqqT)KT+x~JMO3$s^v{*Ym<Oq>9>
z80?$fihAcR0wVPY>d1~lu?N;u;uEjFpi9~Xc#zM^j)29LCBKmrv9RoMo0Bww4nFhg
zUx(6-D{EUOptnSp2~JQfgDwE+MVjYkACB%0vRO0v^&wf65)0J}J08)^VA{g@9cd<j
z5nuWJ2JA;Rvxd$<cMa7Y{KshfAvMu$D6UYy3lI8<RHhxWgG66&qHzo{YkSrwNdVTs
zU5On^pxEU6*lM{8)J63u)z;v>IOkXbsn8-uJ{PQhtz%)e5`8xt#!;;_In7guwuOoJ
zgfk6CzGP#JHvmWO!Q3tvDn{m}YLCz#veQI$)`j)k!D;?Ae>GcBx1N+7%(Lo<86Nri
z*!Q~nJg6RWK&Sqm5L(^o+1a5seGJVia_#BF?4<=GkdDC__1}1!jI_J_&9BlGyDa+A
zucdsO??Hg17v^Xm&uP4YU^d+D+1q=K*XLCi_ljNZ&`uCDhBs^&QZ=2p_!|M)z+hbG
z{hv?NQC&El{3jw#af0W~7Ar-_SrI+$Wyi#k9#-Xaq;AD29_8Mm(0A%jyMDJ}=YfyR
z!gZ`<s6i0nndci`pKY%=M7_O;X_+^U{#OK9V+Z*kY~82hYeeb#S<s^sz0dICSVf1S
z|7q)6&sEQ_60d{m-$lECI_j06ug*P&^?m=bNI4}`<F@Nj_BSd>0R2z%eaLb#l0;k!
zy`jqWDmD*ZcL$ueBQ}K|;*b5eMtx7+n6>!I+jo=7jahHrobesZ2~$TnsPIPnjkS|I
z;%9hR>^-x}4Q}0QUGFci3Ds|kIAWFhZE@e@7j6uUyGLaS<uTuj6!qnDxrHK17Z-C6
z4bdH}4Oax0Y9nP`O%|m3T@t`?fLQhNS&(LBJYGBD-~47v{vW9+pBB;BJ0APoDq-J!
zW4#nydT$}p!oxucRhh0uw_;bK!$og6{&Ka-<@4b9HQ-_z@RLF`c>eh^qOXAejwqFW
z<s8K#&ZzH~<1!1RTSIvCyzyz^bop#SJn%KZd8$J7>;_8=S_ZCOg_B2sn+3gnjAHJ8
z%!d0^YgkyKfH&qu`n3Nn@LAh{hZ5gr07aPTGJjjpo-mb}j(K!7iHyEb-Xdi+olz?T
zXfyqZO_fd$D*st#@xCwc$SyNI{gi2ZV#kb4-sA1Mq+I<7=lbC&eO#IaGX}+;aB~2%
z0^pn&fPy5S5^0yaP-IKj3rgjG#b~11F-u|7$7(gJ)C2gQ==!2t!)6=)6<JS~crq^F
z+P;mbdac2DtZ`iK5Fd|LkPeEiC8G!@-ieBxxY(0r#J%UN%?RxeqoAM!<{twm)ru%c
zbCEQtu6#AY%~0a0EQ7H^YHsEqfx0S&n#`)H4@~OG5_l!U9IfLiGqHljwd}vuBew0b
zW``!{;aNYW+(+xFVR?a4zZR5fu!=O=+Azr)FuNo+1z!`iNeY`4h&?$*<njdP1soK(
zv|CXG?_f(61tDW~m_EIraIas$`j-zU_bckxocMLxE<{?8-<H&=@z<)tY>Rv899!)l
zLrOe!k&&HyUxWf&_?*}Ge*4}A;Vs?049b&#6GVcW%@w}EmzP5)S!&Lm41iTEvA>1J
zH`cZ#K+5A9?K(W)%+XDX7uTlGOWtpBGc6elgirHTx40hT(*P`DDc4<pQANir4igoB
zJ@FN<BxLel_TBYa!Y4{@yZ~>tUi@G<Qj>sv(KTqUct8ekg!fkIjLtWn$NN?G^%v<V
zRpqC{ac$>R9x80WE@6o8fE7wsUQcua&8nk==%ye8`*}``F;8LJuLhILp{wqj*aD+o
zwZBZ8{Z0IiYlr0R+XT4>56V_kYMQ?ABa^6~>3xsdFVDD7C8#F%#!$K>m>g{5yb~0E
zWu`H?nc&e?sAju+lhX*VLl=1jJ(t_Z$KCC4$0yHA^rV7Qym!&uxu?$IFThE1)N@VH
z+GHB+em~L_0UtF#hN~as%{*uyGafXk@Q5Z$9`@fI=;aI7k~zekl{qNOo1r0YmS}BR
zAdh|#78Ny2)og$g@1Ct3w#+93t(8G5c3j#@>Z<(m=i_>83B(ha4Ge1TCj=sHf>(%C
zwf^-`AZXq;^s4-;K(*rON;P-IpP|V&xDQ`5=CVm7p}(eMV3+sf0xs1oW+GYHxg<9t
z1%`^R6#WTtuei%2xM^8(xr<vWHc`8?#Q{0Q#k~I(bCYv7D?@X9S7(>{X4MQlY{=%H
znEbsrl~<wYmkuj_5^v{tg+T?R1mq0ARI_Z$bSN!}+NUV$wLHvyj8jt#>6A->y;EJR
zrig);Qu)J3Y?}Y1d<B%6Ar)gR#LY$o2mu2l?b^6B_e_74YM@|L^4vcHL&FgxGr?s9
z8uFX`d$=o${x!>?gCz*X34>{&T-r=%ADyGug-ykt#s-@7<KF(1EY=7Jvei#A3&KhE
zqb$C=V+*>G)*8g=f$!4>lEJ;zll2ty)<skhj{1SXiLha<%MzX2B37v;GnFu8j71ME
zcFmN}hI(sE0mT*_{uQWY`3PB`Q3o+i-vkD4mlc3JsCeoeyTveWFHz9_u|d?-uAu2x
z9f{yiL4TWa^B;jL8I6ct_s?h#KSlcKX7sbKBA=`Dd^l9gSk+gS_y4q1_;fOM{Iog}
z-R+1Hkh>Sk4!a_cSY}y@2(;t6*sWR&=<iCL(bkQ*?d*F}@Xb`u6C$fqC$yyzXP=o|
zxbU*fH;*?iCeozqa1!?E_%yiN_rW1Nrjvq##ni3;p6!<yo&1*TRi1^aRVDe=o@9aU
zGq;C9q8qtW>^|PDS27f1GGNkSH~k%-?)`q~x9)PFmnf$TeBR2_<y*n6XjB#n@_}Fx
z+ep^TGjGuS-3+U+cV`E;jh-o*eBz>w+1|-S0wbx__08Lxc#7@xc$_*Ix3OiDAXTR;
zlEnHv%YC+-m$-~zXnblApRhkQ`Oovt^HQG{RV?JRDU|GKkoVC+%*JtOJ}Z-Y#`7b+
zP>>s`M8*fC`5!Vpn7&OUt}3ZV7g@7<|50=XTr{nxO6EL8DoI@NsQ=qP{^rQlLN~DV
z8O^AY@G*L^|K}=aCQCx#p>vOOjG=%-hN{G!@{au+gl}x;p!N253)zdHW@V?#Z~t1C
zRWbEEvD4t^la{l&fHSXBD~tHac>P6(^N>wvL-%~|X9UA(<AB?adU#=<^}tkBQkBRm
zebc~4C!eDgMGx1;C7PU?(T}&kJ$npATneh~v+?Z^40mrc*nJ9JemM@@pD3pOdY*sq
z-AawPCJgbP&(9ECG*)i(xh_)#Fi13pTzt(HGp{2N_J@*L>^t@V*Twt(WzgB$neG@~
zvWL7<XEq`(EZ<kFIAzJ`1a?gK{Ogx@9)-c3ZFQY{N+Y2sy9O`aWJ=_+ZG{}4t)F$Y
zWe_RA@$m98$z=h9L6uH{BJ;rH6vPWU`~*(OEn0&4L9*mtRG9Jc#g_CsXY~Q+kkTe1
z-j|WV+o50w1{TuSLwa$~Fyyx1rUYf`%Ylc#?+z^N!Zg-j@3tvAMdyJcw|iWM;Fl<h
zpbY|@RkM_rq0|sJc{NfWe&a6Kdc)&jiAK#MhJ-oXulk71m`u|jw{n*bX^(T#SOrmG
z?G$}?gN489`Hjvi?0wibn>?78>%1m(gSHy_tc!E`Jt|m<S&9Tw-CdgO#2Aq8|Jf#~
z)aoq)u)*jLyM+8_oP2#G-y9uqbj~_&Dl#4wbJx7jyc>I;6M3-;o!kEr^V{rscJWCh
zBgPrm_4VE)3J0EK1>E%p&@%jWITY`5z4QH!H^Zpb;uHKRw@tqf$62-QS0OA7lJ+_3
zurU$|uzskF1=l9gl-&)ra68>Za?y6GKYJE_=xiq46qFwgw^qsiJd;XTRFD!RBE=}B
zljWP_+z}izv4DiX3IGGHdH2{RcxE^YFM7?H#qECS<$7dMth|w`M&rJ@An50ENHjcL
z1cmrNV>_@RZRi>g`>)Qohz|dvj&AQ4qgI?e8?ndS5sKC`$MmL`)wOejM-OY2Hoq{R
zjTyK=uEM|$H_ZuQ3Uc`;A78g^ujTPZi;SDXTY~+6Ve*OJ30pV7H380DLkXmuza&Wd
zTDZ8JoP8C88I3m3PMr9P!or|*=;Q*kJ+Fs0|1c3mE{eZr4x%ld{@4|GYCoS|Gn||i
z80Ph{H}R9b=C|-aF~nLG4cr^bbTZoLc+Ewvz@2>km2@kv&lb|gYF*<$qRPK=4TfQY
zB-_P;kJ1L)%C)}Wak@PU*CU&kBy5aFVR-NadOqcB?`u9;Hl$J*^(=Aj1%ol@23(Hy
zJCgg*i?9kc$FT?sQLi3}3Oo3=b9kKd5D70fKaZ6}IRA^c<Kx%ZJBYzMtKK)Y??Aq@
zV=)ofSOT6+QPvLza$%Kz>W@#j|8X1g*sy-qczX8)_ukjJXAj*ER4?=*NVe|rAh#Ks
z+v<&*1uVhbTN3s;{}Qrs#V6VQF8-y%7SqxWZ`Z!G>=Rz)%W@);vFW?2s~nF=Z!M_i
zJ|0t|VF&)HD^&Iw>Keb$0H;-fXu02u@GodCFWb305pOGID725P>z@LsA)QF$?b4XB
zk~sA5R9JOIJ#{v}F}Vrrms`}Sz~;rj#R+!CIG_4l0q;7ANlKq9#~9{?jZ5x%Tw8Nu
zT|L`W<kX1;_HGs@$)2@(2ac~>i#L&UVPy6pr7HTa-4;=y`E^^;QSJPezgvytFe$<L
zUoT71){j_ojZ?gj7ib+f{GUI!dCq+Z$S(A|e|M{!-k7#2q}j`ux>#6BukN5Bv^za&
z7o19ZM5xei8EPf}j3x--9rHZiESCO;t@4ouhpyuTEATOi+^crkg(E@s8(l%M0mtSi
zJ3(QGx1C55snE(DB6Z<g9z&ba08gl|Q&7+2m{0Gn92$+t{dg^s%(v5@r+X^7M{<WN
zLYH;jCF4ApppKjL`IFJ4P~%D(h%l7;rZwSgnj|A@d%ugTUMW+~^3DCYBo4e@N5;&K
z47Jr22q;tFmgWHC{2_w=x&G0h)n!)Ot9Lw{$hs(iT6d<&-gl<stmm?vQ+8>>?~+K+
z$I=$KYhdS4v(AKqnmU~Db-lz??D=hJwr`SI@$GeQ%jK}QEc<YMCghdg#|YrwFg!R-
z01Y}1)}>1zj2lr|m%fXRS!pcr9j6Nk<e1Ij_(@F}_{~2do9NevVuXJ35!g-6XveLE
zXCuxx_MKUIXg%bsY^*Gvp`Eu|F=r<d=C5;*(S1bhH~SlvQrcn{0qEK6cdV&AzM*v*
zytJ?=%!%Gxt2Q+>ZiCzpaEu@rIQw0WD5-mO$lZ#YA*dewl3f#kU>mA;V46vUhK2_t
zL;me18{s^WNXD9zWG)@K!a05}?As*EJt%c|AdQu!%DsKRzA%%cd3a*C096*6{M|7P
zVuY?4R+W`Wo%`2%*t#6*x*W=XEs(A1S`LMW+vK3SP?^mGHSON-Wy>;HE|YVRu*)Q*
zDJ~A)u2U~8svwH%?VF^fvcl8Oa0L^0Yuax4(gWxL*uRf6X@Ll+(Ml1iXH?ljxNe{+
z#9-Q!i~#fZ7nmGByPcxO$#_FpehcrxJq-Hy2Yi2mYznA%{b}21Uj$l#M|00ZY|q7{
z)&+XjcyYaZGE%?J%##jv!qv~e7qJ~NCo)g=@w@(ax@x9Z!Iw%3NAt9YKfus+)1A7}
zz0JoTeDWR-KhrLuUj)PCkBhgAholV;1={|!w?Iz_<)-9EtZ*N^K;G$Zd!2Lc4f|Ms
zMCe?F|7JJdhi|_B`@EoCyN--&vh_}4j}F2pf%7Q``)Q7bL!O&ov=*aeJ!fKtGT`E@
zYcW&g+UseZd*<janBU{bvNhk~_{w`K;o|6GZu^`0chZTxH~2!iTjh8yz8~<6_X_wK
zMR8<;#cks<nHGFIS$Nu8T%E%pG0wfCTIML>_BtCP!Fo-*4SxsIjZPqiO;#E#h!k?B
z3bA4!<?5&H3;s25c?#=g1j~tMi=zm^o59Ppn9GT8($BnggB5jTa0YzN|MN2<`IFe4
zXNTeC5HUiM?Y^Jft;67SuAUqEPeiFO1t72b?QgfPb%%$$kv<WdVAZ$q^|o6CjzJ;(
z*R%W1x&0xxqB13t*_7)Cq5Ib2Qa6reWb#3>^zO-h{&U-@L}AuFunM8|cpn0G^XK}>
z+P6WL@`LiO?G^pslj6aCHtb@1D!PoB<$8~(R%xi!4EgSpZBr(z&U!*-4gLU&{Y@ms
zH|gnf6;^Lh8gy`)IF*elSZgNylojXw^|xCVBU8s;2mfkaSaj~PB&xP-RI_Y1THKp|
z4g735eS8^=T0fIUHnx^3S);=ayB5Hk@8MskoC??K<)Rg;wGgvMOJxtD!X*%t|3@lj
ze|n+b0erN=HPE%|f0RfilQBhWh5gaP%7PEqid?KmPAvEcGE^NlFuMlkbW0S_Xg1UR
z5cI2%5n0>(y;S1yXNL$)w3e=Z-4cQcCChT1vjJla^joT?nc6RtSDum?yvN2Z62whI
z*@&tLpY67kXk75hE$6)Wen-rMsdd4uTYnQ(p8YY)M0PE7uZC<GaV3wVcr<UL*79mN
zw8a?jS`kGihN~90*W=ti3OS^3YJP>|MHF7FKyHEt<Adkr>eL!7C=a?JbANH<O2E2^
z$Py`#+N>E18@RbF-%9n?$JgW0_4+W5H<M7<M4v=qv%FKtn6s##rZxf;O1$ijK*UY-
z8#sYeMTdjCD`R;!n|qPEMgP3@S+}10GbxGJ4Cs06M(0=*!oir$#_<cs9CdpT+<*g;
zF@y(P3V*$}p5oa2^+~wr@Tmz>U%}f`ZK`Wwid}?_R4js$efyRr_h=|+W?_0xRv>Ch
zpzu*{bBFADfFxL>yK=J8<Uq*p2zpwA@N#>y;c`Cnb){?IkwZm_>$^3U)sr<~1RGP(
zTZ4oMI^x5}AW}@V{~i``zhi)CS)}^g1t_~}X#wWy?pkKmn=Scq>GVBBqwp5=nOq7D
z7Ku_+t+gCO55|>1`~2G~45W$!j<)rb**c1zjQ9H^gOVMPAny3hcR>+=VJ(?YH9RpK
zO~1MuED@ez<tP&iA1;|@vo647{@!dqtf9{6x_pYU>)%PsR}S5W8^o#jzH?+~!&w5#
z9Dv*VT;Fh$q|hi}H;1=VtrWf(n9Bv=SD9l#@ZCPwl(qQl!TW3<hYjl)#%LPQqXZy4
ziQ&(K7x=Cn44odlXo;8smr>1YG_QK%z{MgnQ4lzRtQ{r}1{LsvH?FcoVBGzCI`T5$
z)R-gj?{utY^(t2j4MC4@;abwU??2tTQ+7oKsdlxstuGif0?CmRNx4;#rAgUZ$h=io
z#@LB@2b@yCt<DqPxn%<4($<PHbxdz8ZPM<aA~gV~h9D%K*p5Z){>0;U<K?2vu-5+v
zZrUIoAQev$=tgIw3kQeuncwPebn9P!evtRK7=iCJ*d80s_hs0rEtO=CSPzZPFqj!V
zyD)p7)mKYMP1>gN@LvPk<!dSWqvw+oiHm+e5)6dzX~Q%|kx3C^gWsKlv$457g}%j2
z`$%AQFpWM+3Ij4ZzM8rBHdm|$QVQH21o?E0ocFCkDO`g8RvZ4iydn^*EVm8Kc$MYo
zXNqq+n&)(UsI=uz*e2&7m=L$zC^mywR2C{&(9rN`2<D9rQduhA$roVti?;<NdHVn*
zipC6*i4hgJpaJ40T|z43vz{o&Q}ZB8hfK3}*j{FCo1z5Dla~F^vU;!H_RDEZr4`*p
z$BjB+i~my?ihhO3-^d&xhosq7?9j=@tXJ%ADe7UhRQ{MNcn})mbiK+Bh({oaI@(j_
zM1U$w+9D2BmC~l4RQ%iM;=^x>LDY+x`?ii<4!%Fw--g6Lim%XG*(6-VR+hCMGx=@G
zib4Z}zuL$*Q-8kv7fvk-j}lU#3s^Af>3UP1w7<tBO-_FL8^_*uLbrfKEC>Z9g*;L%
z@OZ>`AQz)ZN&K7H_?d!x8*rNZx;JLy$>Mv9t+(O-DlPlvK^d+6c1GkzA4^EOj+L+^
zc(r)s1o3n5trSzA`lt-(vj>{$eMIzEb3PRF-s64jC5XjW)l}D@a0fmWN);{?gZ|YB
zeZu0V9vFE8-fMo9N*x7>k7JAGWLKBP0DZ2b?H3_s5KP80beMo(mHwJMmz_NPF5x{4
zo`l(?%OT=z3F>Z}DyH|6P1elI!6*J12PU$OvcNJu$j>8NutG!S8Z}e+>65ly<;u|O
zrA3kekI3JB+H7YyGocSqWqONiR$@*l=&!~6-PelCtJ|VdZ;_(XmyJ;wdx!+Db@*Ru
zbdyy>tUqh=VOuc?8Zgqu?E5@bnnp=xZ7mt`LJMO+zoyE!wUWf&QQwY?j=p+lj4C^Q
zQ@&BpJI=8PF-*9Pbc6nAED*`~m6~AcxTZpC=yU3}JPM7C3TmFTd$Mfy8*w(2S<x=!
zbL&1~zJ?wPeWzoY^jCA>s&dl#Abk2cc_M74SR>}QctB)C#nhBe1J)_Pw?w5Q{71Z&
zpn^jb8TY$CDb&%^OWz}Y&fOJLL%GwPgGu4^lBuC|B2VNZ+V;)#Yl1~Gj|8mGQ)y04
z`qKHicnFkfjk9qoo^(JX0@1v6D!PW?d%@KnLXJ;z9#CX#EV{QGYZXw{&c%g{qaQC-
zzabp>9=FsoU6y%nw&*HW<A=W-8u=<}{)k(h4tP^kcK2@2Npj`8-$JGElQE{E*V9!O
zYLCxR_<EQ@YvUQ$qn>y>`ntt_5X5(a?Ga;{rAf?V<_Uh9Z8`Mk!upmbz44puP)`&t
ziR0@+YMM$}4UFh-l{auCk^FENKfzsJqsOCgg7)L)F}E^iQYvqUjlDSqzgi9!D*-?|
z%M2~0K~$15!fg`^Xp6Do(qO2{^Ys_{JrXbR?Dt?|50pt1oqA&%`=GpQj@paA<$=1<
zA=B^>WA-gjbySA=k5zW3r(Q?H!V6@|uqpmH{v(%%X&)@_wHm?@$9f(ImJ*gM7T=wR
z_QTF0=XB`16Wd6{yt_`PXC%h}DTZS8mDbXsqv&oj&fQ+tKHLE88&MXJukdr?YV5Q;
zscjs>3BJW@wBq<chX;*JQuAk!qR%+t(Y>w-w|UHZP;?-k$RM_qZlF8Q+EivnUbLP_
zpQEjx`2>^KRbe6*g$38{M%Wz;GuR?R*}NivO6u<VTDP6t>raSB{8x)T>1hx!a1Ygc
z2Zb22NzPYW9B3F$glVvmom$ym5d+5#7g)eSG=My?XBg6yug;?bP}tKM*c~L~yU8M)
z9oV$#?~y4@XJf@7d1vCVm1|i*9fFiY3czt_Syp@GdP#NoiW62e)h@ocIxJ*ZmXNb7
z_Ec3qbLTfXM+|o{7=NsDug2D6{bh}^1*&eTv5X@THY5BUfyeS2L`%zZt_~1G!DhPE
zwXjA1BJcQ_ldoW84OkPC_DvFSl$u^^FnKEE$e=^{RQHv@%vs@;!_LCEcUBZ_X1{I2
zH>~S4vV5r~B=LUkci`{2N9<#hwqU4|>H4Z;CNW8ggk+}DXnqf5+SbBTVzRNJ@^OI~
zf*fXjOS%cC`hF18t@-ChhSCjk{W^3AsL`uVB(8?3kFc(0M+_R17fbNK9YskOwi#CO
znVI5W-D%Nbzkc-3{+j<3buOx;ed<0+stC_eY>q-5(0vD;^u(Od?;K5-vy8Bt;X%m*
z{h=@WBmt(6LIMvth}p{D9~wyg)3N1W$BUEmZuVXYyI;e!7?WmL5BEel5K>=cjaH=k
zghy=dv`}Vcmrb+C_w&b0UM+9aM+#?^8vrMi{=O%IBl|;|bB%<YalW^xR8d(X=J#la
zO<0hI+Acqfo>{lGn_Sr67Wv-OVA}nM)Kx*drHN0*YHFCf=hzqho~TXSzAymyF+Avx
z2>ZJ*Wqbx<D${hFoU~8n>X~#~7~X3<n2}1Ozg$DIu<!HJ%FuV_0Z#Khh`vn@0up$8
zq;N3RSntT^8x=ULKrreqR7{nxD*&t28sv;$lq>8}^Z0^rOC&G=Jg&|()fD)HdK>bT
z8G<QScU)9njl207-DE;G4c9SFABf|=QjT^mU~V9Z883UqZH{C0!3MWiG7c!5c^>>X
z#>TFz5P(iVG<;lYWbO^wD{BcC{d$hP_i=^TBM64PiIRKLkbEho14l_F64Ob>>11fL
z=Kp-^vrcr%F`^l>c=Nw!3IOsQ@x8rxH{cn2yVomx&cR)B84*FJ6|uJG_ix7auKiVQ
zGr9xUo3X`x%HZDKiSyC-<fJP#OIkEygHxE`sED8T+KT<}7>^(M&}(A8W!Sd*Oph7n
zk2j4r-W2z0(!OcX`kk+da505rxXOEOxji#_RoHDMM~8XCj>u-x;~!q-UX_5ttp7*W
zTeij71nb%Z4DJrWCAho0yA#~q-C=N-Kp?ogyE_CD+}+*X_2pUnUB~{g|G{)mb$4A=
zcbyjzs|7gLl{gc$+jEp%LNl%)PF%cdsz<&dSi+Q=mq^n6-{8mP3*L0nL=JE||Cje+
zZ8n*1x*V9*C&jJ!VSIiG7j7FY&bQJ9IK?XhyOp42)^)n^f0gXOccI<h^hr1E1p@=>
zizM<CS)c|$BsIFvBIp~+T5&^28B-tHUuagF7c~eokh~BFG2*zSCe>Tz^5YU~{|}yT
zicdTtFcL^nlnXf#?Y|+4SSX2d+23Sl_rVEb8WzFh&wU;;Mt=VGj*-hwJrej)SH8cI
zi+YW|{gh97tN!ERni31DMUJ<nZ3I&Jdt;9jIy^7oFF&7W=VwQ?ziY+w5}1wBrAr9!
zV1UF7bVzHCe~l$mw(AxO>9$CRm!l5E1WCjs2rI~qnDpy)(yXGeu+tG3d%zC(O!^e>
z%$x5ocfXQ~m}=Yg%P9t3)SG6$4Z5EKd5}g+z{kVn(Q-3!ekP#E$nXR#lH%xW1@E9E
zs4yt+Be=r&4%+aQk5<S(4UL1354VcIuB0;syr12!c;RlPOyoGkom+ZQ#1k;SuAnbq
zJ+rF-otbQIf~ZlOl@NQLCOWX{HO66&fYY&XFNYnG?VP0+!&0Y>+{Gz{SwJu}#Yh$h
zDYdzBa_*v2s5FRuevdCxWIr(o+%ml~&5h??lEM7+iJ!InHbZJ@qL^)*kd{Aw;1tYA
zwc9j0rkYEkpJxy;dQeVe|J0~-Ynfo<09V-K=+lHu>SS!X4WkIkBFOw9SyrR7=d<gf
za7-z}k2EXyH>c{Q-yIE%fDUO+c<WMX72qfmk&guL7}oz>t=rKo3ojCO0qeG`K5(5y
z9+Rq8>mrl+9zQSW(7*NUlQ$W2w*ZSwpS-zF>SZRm=a)$}%nEw<{c6v&2Se~p#?zjl
z7K;PyBDhhBV10i^U;s&U^Ok$b(E`UsTH$K1_=eDhpl<O@nKwk;e|cH(vUo%}l(6=A
z1xzUvs8iT_q@rx_yEPv7PVla%aMbqW)|C<254Q|ccDz3!Y;hqIP1De&7L>+m%TAjE
zwpPd*ftK-q%?TVtXdY~`GtJKpRYI^@Swjq9r;{yz%M*-=F~9Z`A+WOLbH!pW_1VHr
zVNkq``h{NE^yx-9=|)_64D{iB<SrZg2}SRW&z0I3=9NydtvwxW=Z+sBb`JawY$r&Q
zMa9Aeg?Ztu;By3C;oTgxVV@KTTI9;+LRvVo9h%ISWTbpqyEj`_;P^<cU~gjOR3c|H
z=>?adO!`P-WKVi6wzK4XAq<TEZcfiLUOsPNbTJ`d*`K1%{x1uFkrBIkZJwOsL1D*i
zwdoS%7WX^-FGYrk!XX_8-gkVFkU%q)jdDGYL`WHbkMB|4gF^!72x7kyY0_xxWQmae
zt>*qAF`&^;d`Om+8rHRrcVA3Z%MDF$OuO^264}<Rd!t6VQvKd(EpJ6a8s<nCZkaFa
zMYFJ%PEF#|(#X6Q0h0L37Q;c)csgOqC*nqu2}N68afaWcLnx&eenV#rb6-Ybs!DqG
zZb(U|9^|?Nj1~{70~go|Z@Dj2h$HngGox|J8=1Py8PPP1biYZnjUyC1|F{`qnI%_B
zk)C@Kc73OK6S@0$&jcZBE92PE$drpnlY*;WmEfc9W!SE&sdu{Iur7?w5!a?#*u1zi
zTlgwuska=u%>7}tP)G5+!=|>7`Z@eh)q*||{Sa$B!;V*XI6lj9e`oSbA9B;tN4QV1
z)!A<{PzugVE>$zu%(<Y@ox~JXw$9BTbK^`g*8VIe^s`ulf&3-O8+KcwxF)%W?hftr
z?QzhL7<kAZg*C9#&wz=v9&!|78GY&|2EfgPA~7y(kZMeh?u9GOp%*I=zYAX1AFBZD
ztl?!2Z2n9DO#AC}MX0y1yv^R|9`lmK^CK=^<O^bB@l2dXMcN6+zK3f=raq0e!FZiH
zOJF!{8FPU^C*$9|76w94;I{>PC)h8u7ZpfLpJ4>Mk$CSc+tP;lYt6}`K%^*_J|$Ei
z;`u5^`!Ig&=rxH;--4_t79T>am_n>-<9!D&Z!?g!O>9$<Mb8G&onUwP@BQgNICCn4
z1(jOwzN2;8A<;2%qSI!E=3Vg^LhpQ}kr)oG9KItBl(dWvPfxmF+4Ie9Pc5c!4=##>
z1oO8TP`=?bZUGwscc^)Aq~$mwpY~M^MmM+I!+a>cXYl7@LI#)~+q3{o(iyD0*|eI0
z5J$yEx|{qd<PnHDTO`cMfeeLh@9oPg@FaSe0KyxZ=gn<r>=y;k1L$t268_T-{NIiP
zjpIZ=>0<klHB*;)jt^}|P^(5~Yx_hSVX^r<&VwTs!jS(d&I=L&ls>Dc2shIvmyc|t
zyS=ejXy$$5?u!$Se;Df8&DjJ(bAksTDVBV9B~)!)W~xe-nMd!S+yhK^SSLwrAy4l)
z?uFGu>YB^`$P#es3^lv&m=$GPu6zBOD-jD0X%S1Kwp$^TQ(MZmPm*bpP*@t-wBDBz
z(NxDXg)OmLNoP!|mL4%<Hh!YwG-x9vcE@c#-!S{Y6)W`?JYW@q(e3Z3&)ujs?53xZ
zb9In0QDtnP(zgQywlqvmUBGvo-!aR_O@C2dn;)LTr`qTwc+7+ba&Ssv*>%z}Sv*W?
zWalUqTXZ}AVj|&wZ}`q_CF>KY1u!y;mWW}Jms0R9F7wsO<aAh}fWD;B)9LU~v0y6=
z@EyjmtrG&vbC8Y#_WbdL&wv*)#=LiWY-2dkFd4?62@gm6wEFG0S8I5^zg7rqVi%6Y
zN{<FhA#Kt~|5dOt{p$o4uX<a*-tLxBYK;Ca{Opc0Us2XPkqBeF6RH7bEVoFPxl5r)
zSd&Q@tz&Ce$2e5op)Zt_#l)IA>*;%k*@!y4pF4buLE&DEVvRM5A4|ZwGE4`*(&XV#
zf<-gpy=sXCPukMhgf6}+z{lW*qRWH_`q4PuO{X!4e;WS8&9;E7_Q@xSZ=pFo>j=x^
z4<_kJf`%a0Y~&6^<yg|wA%~WY$P#7~K-ncM`m>5Yaqo|9H`)D8&-{)m2J`&SWhDRp
zDELvbcYu}t<%|WIPB?qRD_aoM+@gG_W&TJ=VwYz7z#lp=x#2V$dov6t0!s9O215_b
z9H;Ba3rXOrF8!MwaKpe9Aad6IO5m2?5eh+NNIJ++PbpHuVj$Ex$z|q4f)YC_Ig@xR
z0G@jweLLYJ%LGG|{n#*X&<Hb|`f_)<_-zWTR!nECNkE&RpQtS#?q0xHW{x-jOK}YQ
zDyj*>4mybMlElt1O1hfpK<OZ&Il`ICs{s`(^%mTwmF@%l4myZ2#I=NQC9t-4w22>?
zT9$wf4CN735S_z;n3TyWn>!Q$C;?Iogv7^F_M<@1fFb(7o8N}j_FRX=13N8Z#bp(w
z)|vWfVXD|4)0}djchF-KE)7~0bT-o<x+K73g}Hu&Gb(0$_&diO_Pwe;#ZKd48`v1q
z97)!K0duyvPr>f_jW*5}A%+gw-EGBw`anuPf3-wzFttq8J+YkH!vVeG1=_2mhbhnx
zlnUAPQ=`93vWMe;x`i}*hSOn3z1+Cv(rzPc5^jRq#6C>;^T$7q@DXOur+y|N#Qk=<
zf00Mq(V>*jEU7pxAB#+lj4zab#mG-dXviGE<<ZPDfaRCHKvtVj9g3@ik*(=^tQH7g
zV%9gLm>OTj#FYm3PI<$i`=%k@v@f?+MEOs@ih@zeFaJhk3tL32G_+pvPg|Dw)p)#s
z^W&O#v<^8UEI3WEx?;~mdm~s0-6&Ti!Xp76sbDPHYy@4xh<91Yu#?I)n%~Z|*(ic#
zU6M;dW+PG8;{}!-B0LiB8|ZfjMsW6E<P}gPA(PRiggv8q#FcVfZZv3j*}mDGN?@~A
zIXc1}dJ-dJ6S(anjl*T}f+4>#E~DK#&Gs!MO^>c>7%}>wBcn#g^ourq*?k#@nWJKl
z;0|y0QFKiyyBgH{vEFgH(4}Sh4)9$og#4KLw>xv9X4UWD{(1enInRaFe3npa$e77H
z3fv;%f6KjJg~|;W8tulOf4Am-;Iz519h8(P)`T~gZ-G4UdbU!_vr`s8b#JIn6I}Fs
z+ZfJiBLh#7p84@zwjqi9e42z_jXcYgB)sK|`?oIZ&_(N*VwBE|)oSy-^@|-*B?R5g
zs_l9wT`K2n+8Z#p@udF+9VDywX-GAJ?v(N~{An&S{5HZEvdv@XSMmvSEcXx+X={`B
zUvnA;18O_iD!3ko!IBdRzH$mV<oj7t<o;xTgMv9SQH`!u=Vu3z|0N$uL};1-YcbJZ
zhhauGa=BNLX9N2rlkdgmF<B^4VkgFwDHdmfpXZ9&!0#_XL<o7Zn*Jo{%<mgzDDyDX
z_>xaxg8*|J%AmaX2$A&&a3p~H>k3j{kB5MUkK*BVZvcqO4sr&8)bjR?F`ai<t|(&d
z_N3ke6RS&|YTrxr6bwg5)E=aLOEmet_@_V)LOLG+{1RX`mi-yaIj)=729}pld)5gp
z`1)Z&4^Fy!$aRTFkqAoZaOf#Ai#y;8;_b)nXF)Apg4um1BsD^Dkwg1qG!&f0r!C_w
z^MgN-e8a18GaE){6hG|vXMEHjp9zXV*I0PU9c+D{Sq=mMTkt%mT)DurZ*rH9t;AS5
zhkRbRg>`qP#VwQ8@N{c^_7@kmCY&kYbf(BXjvaM78Ai!6IIp9rd7Ee#4Auo=#fW40
zDqcAF6m}aMBrbrd$ko3q2@UqQGZI%DfrEbv$9Ady4}PN`fdDS=W5{a+1jjU#X3eUh
zfG5mw<)p4nQf5CtAeXS%{wn>0S#p@Sk`+4<kPu-2(8{JCaaiqIMT2q2@^mEndpNBq
zgtdwGnK=y&BLJ59J7htm7wVS2O%R}O3yh_wu|0;?CfUP9!v+6F=y&AL|5zm*{O%24
z>RW_7^o;1F{?gHPIyAGZp)LnFie8LTVn^s}A5YhpD_mV~LbU|hsAC6H?g=g-_$pWC
zdlB0!eCehXoj)+PhL4&&-TMUI6zhYK?y<@-9j@=sPxltX;)a9?<3mm_-kUeitXh2M
z(+pd3t&TF?M_u>DcCHOYM)0Ss52p3~RL`YZ=8FG%C4Jx4Q)S)bsv_`_i(zSk*d*Rc
zh|3Ah*QUOzsHjwJq)9f(mse1KunvT}@h-HM6cCLYYohSB4%h5}b7S!t85kB42{aba
znurW|>3baK+}q6YUimI76@Il{I%Ut@jU?!^$2%hsQ}6%2>frB?Bqn2?GEZGa=OXQ@
ztwx^!uH_HoS1fDuvm%s{k!%#RLf1*FX=RG*Dkp&&^8>nc^9lkCUkN$$XHHu68;SFk
z2A&)~9ieWJi&$g-CHQ!-poKp|JI6j36l0oYJ)44%T$@~m6cVJ_$&Yf8ar1v_$C)gJ
zNv5d&Mm)u|mbUUxN;4P-zu$<ZvNH2>j(LXsuDT&D`Hrv`AwrZC1gmm<XDD{-J#rvO
z+4wEhENhe>^V_;N{5fq{DjgCOV=rvVftJ0{CekPgeWviYHNVzuljqs3RPA~<*}#6Q
zy(e3P8ACJD(??jv_YU8xP;Y+Yf#NFaVrx<yTI3U#e|{9?bINX}I5Nt$m*Ar0CNZg6
zK+@A`-s|@D$<alQGI;az3fTC3$m!;rbdePttg*JD>VLPa5jYQ#7r+U&l43hHvk5Fl
zY*fR)ozvIpX6(iXhXqIBjf6pW^Y-soUhr)@YH)Bdj32Yhz0ijrq6SIo3^q&*p-pOZ
zxj%u`gWjGFCl`zOg7he~<+~pFqOUW+`6|f2%Fb5Wg4rH%%i04=Y+lMGT^5@L6<w20
z-C8+?W#s)Nmoy@F78}ef$St8*V4uz2<6Z#PiO|<^wG(+Wl)tl0L#gqg4e=+GZgbzw
zRlJHGf1ktMRoKr5w|C}E37=tV?0`6vx!sR~rt;Ftgvhc5sMvUBZ{yp`Pm4?bf-{~`
z-^mu|YL!N_<;4BW-l63oZakCVKn{6M*Gz@SjkJ!1tTc9;@shNrgIC{H7mNG5^I|+p
zO9c5LLxd%+-4K@vFlenZ=W5m4SN_+|BGvsw-9)-FnU#1oixN%&i=3{U&)32>xABk*
z<A_KgD`;eYeWlv4Cd=K=zzUB3^PpK($D;e4kFjQ4v*qDs?Iq>0eKX1B&hvuHT_d3T
zu&wjjYa$l^QQDtz^)YTGo!8#0f}%9kV;$#lf=UCDw<F+SqJmGK3;KofN?Jlg>*Uua
zA`u62&OPtr%u6lSL*G4$etG}#+Ql5-(dE=%WX+PRV{bN85G>?7HPDPq7!g6SaG!u4
zW?7sq3?=1R`Pn2BAl5jyi<bw8;_(Iu26F(l35|}~(D2)CZUNGeUhvay*9o`KX`!#3
z*2nQ}{relJ&I(rvn<F`jECUPaoMK9BF^2SR%*D(@&6of+;a)-$gib{&1(xo*e#!x7
z5~*1UNqAyYZL+WR>??!dlZd@$x~YeP;d{k>Ze%MB-<bi4g}0sHT_1GT#!!+`ZAZHD
z_W&NzClLg)m`Ac!Spn0N)L)e);Q)fr9++2f>XYZ0+syOyzD^cpnEPh#A>Crf!4?B{
z^6_!?64!5Y!W&Kt)SJ$>M3|nWu-QP&x=Bfb-dukvFkHt7jeH~!u~bDyE*3#r0+X-A
z6X7`K49pReSS`&-U7lT8=%s!G&G;@GuT4Q0s=n^-(`)H>EAaOxerweWns_aL8~H(F
z;(25*<7T7JhmZq-y~E~vqeVemeBKHJcqQka?nW2iS6>DPj|B~eT@B&!`KCk}q%#I+
zU!<#}RPl||^pANSs7V^8!`_=}-^FVGNousm@^6vaHy`vV;mtXEM_oN#LvKFJ6u=hd
zkAs5HhXSQeJB69v=OH_y?#-|#{+AY%ms)1urF;J4-GP;_@bMAKerX|7iO#iJ)3NNc
zez&w9U9H~y7uWrJrJi@y;O~=QWw-hG$G9Y~&40n=dn{SOL3CbOwQ{o|=;EdC<5zZh
zxMmk^M92_FWZNj)OJ+lZ1)mgC9g_Ji$c6Ew4-2SKpypQF&Nd$b7!`sfB7->|R}WlK
zXE;sO5&gw2_M}mu)n?Zu01)*VsbJ_NKD*2z<<81;k-QeH%@lDj0jDJ-vP*q3Rr$eD
zp1keL`(;*@eax7cl`6M)h;zQ`Kc9IQ=nBcBhtIn}?Hu`3;}HQv|BDmh66go74P*q*
z<^X#3$!Yr3scgoJdcbduI<|K=Lz43N!za*O3E4^0rvDlox3fnHy*fS~0?<IVDq9Xd
zi5d}Z>*b<=^9QvTN~itp<rr6~cH9)%G@2N4ib<8eMo`>+Hzg`jaKWx{2BlIoQmLx<
zCKso{_jlVV#&Jsk{Z}%Nh{)41CkmK^rcmE!ts%FCbK;yD8LfGM1egPk*Nb!0Su}8;
zEQq%3k)QH^Fd<f&dz^sD-i@3{)~2h5ckGR!-Q3xxADDz!;!tkYZf+sRGj5vtsI(tF
zf8NRp|BMDcttZ|V6+Uz4rX6)hl=X0KKz?M((s0+7;7{HCrF;BB$E{2L-OB>|#eZH=
z+4ISri?Fk^U7c{Vx|64e_qv1ocoA><8#H|3(cocC6Tw;M%|Nd@H8w~+`%D6l&CLRb
z%hL-2^=4G7%*UnPuVK4|x>t8@1x0WbL1a`Q3;r3-i%{49&EvELpI<eRUF}Nv(9T0@
z{<4;=`N#U1RScO~>=Z-9USnNHB2Y*O?So`{M(&7L=*tPbF9c*18Xwu4w;G|ENhAg_
zn-2>T3c3ZS8MrLp$AzA=tey)?OUfPKL5I8em~J^7^A~`OmzWYL^*iomv-_bx)q0>X
zu3nhME@@8Y0Pr<UVrr5QzN;GQjnE*_<FQhy7KV1|*8o>rvGl=Fnys5D)B<3za}rb;
z(U_qP{B>3krOy&C6}*k(k{BOJ*JO-tTWxNW%=AEaqriCt3Q72fcwy!$Op&MqzU4+P
zVfV5%dSIGs0^$}3)y6Ej_Xfy!D{Ey+FT;st3wNH<Q(7QOX^x{R7|9GqU!)j{7Q`}E
z+7$MurJw68xu*1j;<Eg)U%_3^Tmk!k5+PnAAzl)2uok5>(uvn&d!nZsWO!=1Yb$b1
z>YNG^46#>BA1*wxfP7dx(RCi-+yP(aP*o^1O?AmYZ%Y)6Vo1yvPaFz$yTDb{I7pm<
zrG@hET0=6O>4;!K7!?mNt*S8165=z`XX<G9cDVR3s`%qhEaIzbs>J?#WhFa}VM^8S
zlJ1%lwEU;g^@$O9=E@Fb1AGwy127K+OTCK3dopqIDjE53zdq~iU`cS5`l4d(7B&5l
zUZ-$=NQ?&;4hezj%=lI;MGstIqas+snXRV(T8<Y~(fleJE?ju6PEn}MmLeF>@xf?^
zEr(gBIK!PZGagxr<FU$kL7VsyJbwp^A@RKxI+Jw3TP>iX&<;JuW<^#lp%}}TxN@+9
zqQZILK!G{`;cqQfCC=JSm`?dLOAg^<$<uhvaRMh7xES}3sY?!=4HD**=m4RJIqPbY
z;|^l`0cPfm?gCG)4v5DS+jDnQaqV~ccdfKk%RdF5io2G~ASRLZJM;r5v@d0cqiO2p
z;wE_l{Ok3{$|eNuD<gL;I?~*T9WS-Jke+se%d>H2-Ne3I^%7Z<b@pr|3N0*C;|jJK
zTiL&@oE7Nrb8nEeLo_cfq%rUc30VbWDE+O5!h)!$t=<tvNUxAeN~Z~{^))|YB|XoD
zV4CqyPqO1)9_Epij`~W7lR%=Z)R4}h?;2yp&aR<+&0ym{%L~&|7a{ShmokvWX3Kjy
z)TVItj12T)gw`0xibW{mHC&aD>0uVhL+zV@q+r64Lc*U3M!<kTP$x>Jbh=NOdStrq
z7uZpmeF)tjcfRikM}Q+@msiJ+vN+Y^DTPZwD5mop#ckVMFW+9^#ATADIS^_#HE%x6
z624Yx*pr6qCyNq~37vHY5N(c{qNX|AvGuLO!Yo>$Ku`33p<q3=A7L~kgn}lMJSW19
z1{~)=2|2Z5Kx_h~@1E<nN>5!37zs5Jp9t|CS_bLY*Jdj@=kE0i?DT5vC1d_2mCbVh
z)7F{r>-Sj<ywS?G!)6A5S}(rtT{}pWyDb-aL5XWZ4YS4M5BY?OZ?y~~8x*UXd5h*<
zQ&-p8h1|7o(D_;VD&yEmjW*Td<Z9Z;eDw{S;)cYjOPG-!J7yYRiXy9>7YWYq$%5rt
zx%)f0%{E<o>jQ8~zC#Lw2HSIi@M(+t%va^+C#q^K3ymVazxgU3aZx|Tx2<Enc<N7d
zqrk@awFrOz|36#kU*Lc1co2X_l3+F!F4pkLx$jsMjmWd)y%SW@SLe-zP5u_GF_!T{
zaTRWesT=Mxo>>nn-xTA0kmZh)Lv_avw_$_HR<&*NcQ>7<=6g+88a522TfL*bHylZl
zZ~mJc=U82UP9d%9AZe1@Olo7ow^<x8^vfrt)(jy%!Yu8FU4#G(kNqijUYMy|ESOvR
zyo!)D)*!-**wdl>YMeyw?=@^|-n&mnOKVCXkvY$Az8b}U#BuA~YK6+#rkO6&H6rwb
zbL5E56NMaMQ^|W(e%u^Ql-A0HPBlJ1{m&}!HX}t&cVb$bxE8|}C`8V8Y^GzNL{v^^
zLNo?=V(q6vmLX<h<$!?tl72XV9el#w^%O3?x}2qrt)3p%e44D0==@ydLD01#>+c~2
zU#?P<!FY-X0v8ueom9u?eajJX+fmEWONNZjX<O;&REXCRCvVh-TzqZBugvACl&Wm+
ze|A_l8N%*wq>qe4&$~a*+m+};W162vfV2%CMm4+E2zaYVk8)8A#r}e&a^it<ItKOJ
zx91V;6jL1iGW1(~GRZVwrQ7l=_`L48%H~<bR;IV~*B4J$W9_14RlI=RRPbeed)9Az
zZd9#SfI?SebMj-#eB32sf9&(r*Q1i3?<zq2f*8fEDd}$5cd9W)k|h)j7!kJBnK7EN
zEX$hc(~04!MxCp0M;$ID7v)HNYr|><<s>>`38H&Qg2euo+}?Urf*6y@d59e#PWlhC
ztNbl}P#pN4*5<?5=C;?rOW~fMX8|m*DD{5jP`S)4VW>?*&o=v==2~e^aW%pQ7>+S{
z-s8jDz2j2VT0i!V0cUa$EGLn)gvG|bif&hIvvXG{6dGh@sNG*=rvvrFM<@qd3tJt*
zpLJ0EjiXXpvrnFKMiWKc`qHw&5Yg*S-VdyXR#&dn@%2Cg+2>W-;ch`LB`Mn1g2S%O
zlKw~ITs=Qzgjh8KB^#*EkKX=%!$TTbSXZut?^V~1;C~-2`Cu@kg;oO{X+2ZLg4kVc
zO?Zu*xfzd#LG{F@&ZDg}=ZHI_vWCV>I-QJkJfbQ8NSxP;U-~yi0V;TP*409LBeo$M
zX88IIlmS-{l!lJJFAh=<>*K9mJi3CR#imhWVdG$Ib9*piXQNa~LCAej9hUONQEM2W
zU;APS_42|)<lo>&WpE3BP-gzESV3c_2?CNCh|dU235$*7S0Nm%g&~J&Bd*wdCtdg(
z5!pP3EW$rz&X|G@I<nO_k_*b3>fGvO1kt3BgRzh%(7-CJw6B6Qt<<E)U(FlZP`_3c
z<T$NlgQ<qSKuTe8^TPF_<`z~#1DJuW)c&1GV()MPNIcnuc!>JPruKjJn`pnxMPW8&
z1>FF{UkSIb8a@@|c6?Nl8Sj6W+ZiI#(WiIrj4^7rNAkdkoy##w{~OWYM!$ca)Dvjb
zm}x9Pqfl4ju>1<=mj33T3d2Dgy}d!E3NHXU8)>@cvmb3OJQ5JSm11%yC>^;8EeI(I
z{()pJeR#m+^odhUAKnVoNj$@}PY?_eA&Mo@IA;G;$(6`F;Op-=&<2ow;#VA=cMt@w
zN@+Y4cqOkOHDPpdNx}0tKk2+$f%7%8FgmP|Xk6OaAQ0zG)jagQOz6ah$iRwjv=e-n
zj0o~4xvGezYQ!zLBm2i$7TB&2zoQ0#ZWi^k)NCzwAN&9kr>kh$L^;LZV@j>Z9HrZv
zUG28^N3OxBo*~<v#GdyK*;$#Qkg<0o3-<w~466PfVWIgd8|>vnkskueP*@gi4|o!x
zPhwd9FuiAj+j2p>fG`RklP<xtG2-q)t;#P0ZMs#G%&X`3qg{y*(BRN;FagA4g_)xN
z=gvjaMVfnAC2VcT<Zt*+IA8vkw3@i8?H7ecQ1n^H&k~*xP-W)Ipd|W?2NbeC<5(8c
zq41R^t#Sux!s_++>Pblo^v^Waf%>L>7azFp0+^+AlQb~4LY+N>)k)>bFjVh`&4^qH
zF9VbvU5sRQL>ovt7y%%1y1i9(lba`-P^h1lf}IrY;+*r6xsiKV?zq=&Ce_tywKm2m
z&H?91J>wG!zL$QC(dNNSLew=?txgcj7#YO3z_2$pmu>DqF({gcW1M^kPbEtsp>Q^v
zU$oRLIm3iRuBI9;E&*aC%(UB}-6bUWIae43?WqeKmc=721kT>b1NHxufVIF$8+0)&
zfiCh=3`d%7R$k;P(`>$Afs;6z@L03P)^Fq1-LGi;-?#ODuF%&+2$%f{zvV~C<wari
zd6_&eUe?C~uPHg{@z8F(bRKE@jwx@dRv(EeQ~}H<YullrO@vKrlDqAxMj0X`=>}T%
zZ*Z^H0-YC#I|~Sh)>J@(AoM&DDC+T4AS{D76Fk}&6`qf6+|MNz!d3KNOFg^nu=%^L
zt6`2CK}z&Wj)c!YK+}pG4v>c$A(rffFCeafCJr)7Fl*s(;{2ZvYjC=AsilIBf;I|+
zV&XGGPcTz;<8lvVOm&Ir$n5yw2=a~J@E}QlnL0PuI6Bk|vtz{x3$Qj4%QvuvCY=ok
z_C?Ias3vQp_{jkoekLfPbAN#gm>8C(J_npE(Ktez3jpMe;xO%2Ptzdkl|mN?1ncVW
zM<~8u<NtY81SPI$xXnLt38K2EO4a6~lr$t=d!pkLJFz$XeN237h|D_7bpAWebOK0w
zhaWq5%l(pUa<#wyXNQ*r@s+B1mn=I9+r%uYD=UdB(-KiD69Sp#`y;;vasdqmAOR9h
zDr7s{1@dKmKSI}gez9ZQH<*hW{;Wwm9T*JJ_o%3~<4xQWiv{WYdx!0fU?dlgLT!Qq
zh=UV05y#2KMuG-Ex)_ceD}vj0*+^6TLKncNSfZF1xl}61U~q(}TH=tdRbmXJ1(Bm%
ztntqZ@<z}DV_0XbeC5W4*~@={cXbm61x___ajqy2plZWp>`!JAZTs|fo4_P7yc!xX
zk(EuO2n`zCXDS?}#GZn4e6V>X?x{beI_hoZD@y5j*nrsB@zNz>W#&GXB>9(*A%I{^
z<Fh25Oy26>{abxG*Hbmij!T2$|K$n)RKDAPhBFkmE(QNtx${O93VrDMv>hJCt?#Pg
z|6epZ_Bs;_KK{M|##aAO`R^FNhC@`T%O<?fzJ465Yq|bQ6sZX+BQc>YHxvq7EA#_H
zN(hHU1P7Us$ErdR7>cj6q;q-+^ui2VBU)422+RdNIZu<*jue3ZJ<x@#3?{F4vEhbg
z?f{@kopzD$tD?+1=@}Y{Rl%%R0QA2uNihQqcI*jpyAdGx6Ng+l#vx_~=Ei&qmqgYK
zFBn+5B>jMzRGzN9VxAQh0CcM+guJC>Cu;B$btKG%<M3{B>R;w3%t@%UhS3Fi!mcSv
zjbs`{oLx_v-Bspwn|emx6~mQF?TfMH5Sz4M5U%JAoQMdmNDexH>64F<8>$A@y8knX
zLqeM6h>NXcMvH$oHEo=l*J?A9liQGgvyAGW7|3?U^V2X@lil5Q0u&GTvsy^tnxBi3
zHBB+rFcI>JmW^sv*yN&O?LUR`U()=)*(V3|-6a2#r8GqRS$u8cyc1>qXApQZOG7%u
zBP1f+jx{7gKt_QyHI>ggN{h8v<82Ll$G`M=Uk1rKxGJHw?+^Xfg6CpHPDp}iY@|Uj
zti?4K#5LlWQyAZVk$R`t@erQAS{=Cb#h9Qgv2EWHQ_In&dU&8XjY?l_T54psAj<c|
zBo^WRF4PdBk6a91YQHqP0-2AKRxb$mZXrZfKrShnkigU-_64p){G+LD*u!}xh?#x~
z3dKPqvR`Gk$RM|^dSf0*w)4~n{JQmD$~gq8B%0J_n%S~WM26rq9kORRYg&gv2NlEl
z)7{IbYvOjnp@?Q9$yeKQ3gjEC42+e7b*FFZZpOn5$ogP9*8>g9nlo4975*QM(Y*tz
z9~G8>zbPl!c)QRyl6@^7cqm=Rk}IZ&@eFAejunBzv}VQ-Sw#g690uD^fT{hZkD$_o
zDT=KkR4mD?*GH%FV`VX`3LnOeXf^QB7s?+Zt61xz0H}Z*Y$%?Ix}j83ZD50*7UO#>
z&B@*ra6|;(WkSz%Q+cdtirmEvQ2HbRL@zUH<=DbHo8hIRMebw8Pei@FE)%RkZ=v{f
zxK(4W?E#}P%Toz9nnviL<bJ14sYWg+LsbaYXiG&w_*V|OIuJP}MPue6<q#McmqP_C
z?jPhqd_reUKq&=fBR;s#(5dg6joDNXTVRCUiJTxPEwh{O{@$)Jx*3izb<Gh!q4YRe
zfPkAP#ST`x-(cWUcZVy2*(RNxAL@s&PSJ9*KT)4)HM0`_M8?iXgwj9ps+)$rUg6Qe
zmBXdD)LJFzeuDYU>1UI8L4V2LBe(xY!f_dLC+sgkxCh<J*$VfXWP3;N%mKM+#{#FL
zwGxBsf_zyiq|H$b0-s}ZMe}B=Pz&c!o+hh^3m{pE*)eFf;T?&OnDFg+-F#segAgez
zxs%hwR&HJTq87icMkAaw$<-V5g_{bRYjQHA9O$g8*P7rGi~d6GCC-3`<vG*5t>p-D
zJN$L3)5{EF`b`Hwo64f4$b|Y05)dvxZ~a5!s!HhqD@IC7G*Q0lN->^CA6*Cr3J8-}
zY_vPa2#Dw=3nE^n#$RGDOSDzJWt6X~Ji)2Q`cwab{E5EOVkRtUX+T8>!>paW)aOrN
z<3ou%ng{c{Qykj4K!sWGz!to3XkvklBjj0^9lgz5Z5F`m$Fo;tt{D^uH@svIwsi%M
ziz5}Qr-!Q00wn0$xGC7{9ov!q7xjPTZ*9w0`W@|4aql-(JzYtz*=-|To41wTgbsOB
zO5Z1-ez(3PqYoV@A>6QviX)}p(nX+bXNm$QezE9eI--plPhWsloWgQcJ4vwoo>no)
zQN{S=sEdm~*Szl!oemE9o^W-jzvtNjRqBmuWKc`2RGSSVo^`J_?KSeN);2mIUCAL>
zqMeBOdGivp)Ga_L%zE$QFh4tgiw?+Unyjx$Maei}2TL?{7E+Y>tJ^3RM=RGr`sxKn
zS77K15a6B^Yst~x@o9QoM$3`me*14(NRCLh0h@$d==3F;|4Dl~BwJGAsr*^bR$~D&
z-326UVU^EtIU||D^6mq*UA1ZD74|!)MmQd0#YM~&2*^9S3I*a)o3J7+jgZ<ZxNHJ?
z@)uDUd*R!kWWJ^exens!=ZCv%0UJDP$+RgYxAELOcFXPbU3VuoMlZ=tq+a6dCVf%@
zMBm_tDu;?voN>wYx%F>BrbI|C334Qjr@vK>;1jZrH<?=PAsP6&^eV#t1__i<M{58N
z&#*udd9L$xo_evN>Cu(XQ{*x#Iy1WL!BObGsT~DOb^Tu1JbBkbyQWgQr=i@~S?V<6
zJ0hp!jgo?@znZ!W*g|T)#W+Tk(EAC^C7N!=FU_QwTy<``Wwi11hs1VM3(8Av1CV~P
z>Q?9ZiVlcsp#xXqyYW%4;h-{o`(dB54ZYRpdG3wd?+;ELxW@R4?<MiGSEg#sPKa>i
zF74M+kstqYY+Ae)Hq<BmPgSt)M}LFOYlh(^LSaJz8G-*&x>jsY|E;k~lcTiHSRg@{
z;$!P}{ns-q8v`-o0*ziZ#ON^%7BAwW+^NBX(yhpSk=kf~o<0}n$Sp@0KECQb`!(vE
z2?fpkyRXv{CjdNDU*5;HF@>#5dpZslshHir%o?TMrFju^YTS?^VVX^vD3SpnhxI~l
zLLeN7ETx%5&uAtuihpIq5Id>&Sk`!uFh9Ma>CnyH^?u=#7n>>sn!Kk|^suT&x>CjT
zXE#vU?RO07tk#;x$USaMQUs^MFqd%NC0Ox&NnRAiQ_d3zlQ=e^#Lw~O>FD3?gcSD0
z5sZc=3P`MN<k7)aSihNNc1KfKth2=<FQoIEOWa<Yt36^ROT$r8A|_@&tInnHWd;Xg
z5G!8g2Dh1Ir2>RinEn>)rkgD%nIZhq0m*<#Pj}nltt{@cSEQ6A#ql%9b=DVue-b`a
z<a^Xz31f1Y>7hXT5C_r;h1ykIr?L>iKaODh!#v$iyGKl`SXLLs1A&;c++M7XD_IpX
zz+gN!ABA50FBbqFP}oc8JRhkAvsNXUe6tU;lRXLkcee&#oC&k_|BgsGKG#hI*-ZPu
zVsZ8bLu28tmCa@IUH@oJao+ZJ*AUrj`_)>*tWh{2&+Kr(`*485d#cq1bymMiM?-^)
z0_)3A>>u8X_i-H+gMiP>(|x1=aw%w1Xau1><L5rHXe3Y50|VEIv}0CH{V1lz#O#25
zcEUl3V|2=#R1F4$6akNMv6l%&dqk2B)NM@?oxsJ~`m|<**g?3@r@0?oA{Nj#X7=OB
zJlBZSnZ2jxnSD;EM(kY?oJE=;upe|51e&9Y+}ne98?Sh?tTmm?yKCp=A463-5S}p!
zl*WdZT%h<}WdI>*xeB!1Qc2JQ%m;LHiISbO5IRm=eU!OZ6W$v#V0LLPF;4L1$aU=H
zp1DW>c*)uGazDp8l`}+p&?1B`^mgwSPK3n8%VD=Ge*e@kS;EmPYL<QqfF7kvb--o1
z)(3;T`vbFyFpV%5qm^jlA_oq2YehB9E5J^)93qbxYI2vETY%V&Y@;ErMLTM6`l{FJ
za3ca1zCOr)5;7Eg1CH|Q)Xmd=S}MT8m925T?!Qx*ln7E?Sy)6q@(jGbQK1;C&LtH>
zHB_z&d&`oTHT*^UkEtUlsU4~x5u*R@HX@-d=J+7zlcnqWq2rjAbN$CJGgFh*ml+9=
z?P9umXoDcJi^tKT$9j_g=S0Ubvi;NDg&N~2zA3(AT_2VE_w!5$dAefXkC)RKX1q8A
z`J1gNC7%0VNyexriHiO#@2igi2^{t-owpM1x8>22pEu{1#XZlnj3w%{8fWITx0*I4
z%6ms-i^H9C+pPCZoSRWzZ6NkPE@$^G{yF~Mt2GZat?2=-ua~yZ9ZLb7j6NFy_)Lm;
z%q~bY);|LJsXq_H$32Zx#cX{7(^onU?CzuBCuBMOUaUz;*o!VVZldqL`pIx$cYp4t
z&=reZ6ka`HibuWMXVyK|!!h8+gl_Xdl_0$SW);>#Lu&;DGOlCITtJ26nams2Vf0_6
z(6XQ4johR(mUv;)*{PZFCtPYa+3W<lyYP#;SsafKMZfiZAhrqV0gHFxo+qK$vjvj^
zzx~H)UMPTCkgkxo#u$(ZQ*6>XP&G|6Ju6t~2gg1v5*ty{LkG2@Qx>m7yqAT)y>Rt>
zAGDZ%YiK+nYqgQ`Kis)6*GRt+a@a5b%J%*!a;2tq@mTv5YQL}ciL?zx{T&?qC7!>H
zac$%c1-}Q%cX8#li1{P-g`~$wpQHv4Bs4Eu^uHKso3TiM#hj}@uRtZF7mnjkEY&|)
z8Kl%9j#|ue1%@5Ysd1&?B$-j|v7g56aDAT9))Xd>rWD=VL|a<)wBKgm&8}XG6sV>u
z%Y###wow9k{4c=~L@%rr)WZ`<9io-mFt+tmks4qC#EWN@=;1dmih}do#hd}33f^yD
z9hc6PpVHOQyljD^KS$PH97yK+A6;IC1+m2S>U=+neq*cH#gD`owJ#3CId<IATK#i0
zO!yNn407_pWMWJUWSy$gX+BHduKv+>197hJcl_sgmoh6gdvE!rOUFUJrt`(yd-t(N
zG3RNCnr}dF3|HQPIQNqNM$3KXdbq@FeSP{*aJ&H4${zpuW!57{u_p5>&tZ(>;HydI
z;i6{G{rMfvBv}*JFp-NKU2PTjUc%j-%jBh%hitJwW3;sFHD{eEjxLgG7{m(>Jr@0d
zQYU=M2d{75B~;FRw0*NRn}E-J5s6s$!@}xuHgo_dSub*4pl2fH)h$z+*H#_nKecYU
zd1Mk_By{s+o$^q}Y8%am!s_bgxkcVz91dVW05N$NHb&!bBLj^RRTwsAuAgDdlUhyQ
zT0Exr&Kg(4GrX6!q`=d&5C+LgUN*mzpy+%Bn<DzEa$a)a5eaa+x;oC%cgK$nr;Jq|
zk#vmA@DNKax5`v*R1<@-=5b>+@!NIq(_6`eBL8Unj9t%@o%j2@NB(&mQS!COJ4ciw
z^%d9ilF7^;B3{mQk4yA1O~Zm*JrGEm?KZZn@`KEgc1?C15IL|=DVJ68&#h0ku_u5^
zex9@BTEBmk^bCHk{U7^GDQj~`zZBu=uP*!dZ@n@$C#QG_*=iE-hV@ALjRS{vt57pO
zFiW^G7>Jfp8cLd)yuNyby~ei(-BUlteSdB(X@d+p1i62JiY?75shbdzVtrG)xu<m(
zzkHu2wCSU^{Q~Vwcq%+f*w@DVFA?|=5fHvV>z19y)dy9C!I_&Ro&VW-$|R)ZS-Si3
zVR9oVvoXU)?U9%ze-z19eE3rrwB@X&=%Kwan>1%_2c{dBr?reZqvi;iD+9{^w?*N}
zsa24-n%wNH<uaN2yQr|}4;Z@BH?x|;ESIgp-MCorjA^Q#$A<&gr<Hl>JUzor@GH;j
zB29rQAu{4M52@c2rDENyBMxq>Q-H~g<zj>v$D@oz5s_(S|IwGj^SI-b$-MnhZ}3cn
z{PT@=JG}RUl27O`P8Z`KmS?Rb#dbge?DG6?eYDAzd&>qteM|k6(Q?SIKfin`t`rv+
zXOAB`;KptB#F8^yd0|6A5;uO)n4{S1W@tQP7R6+IZbal)r{n4ExW(+C|CO26AJ!Yg
z+;P6oESt|g&l7vPu>s^TFUDf-3q_)W$DfKw%J!8ihiHD`hp$|dCYsIB3f65~m9t)4
zSOeZ)*c}VgdA_oo^*)-XnoAKA=9F*VQg4g8(xc45JfD@UvtAZisBV~Wei1LEZSUEy
z<{GgJ)!gKM&+s`JR-fZ7o3NN4``G?06VF#0{n5Ld0lu=f=c>19yKw^62g3f-;ueMQ
zjBjEme;SWB1KwuuN&-ApJ3DWly06zSL*v7TUo3S!_V<s0fB`CGSI8v^jbp#*F7Zvg
zINc-#!B?t$ovw_E$rX2Z48ZXQG?5&3ri=R%itghv{7iA6hvzdE1=P4_cLl)mvP^aq
zsiiZ>M~~;B6!Fea{08!Lbdcp{@b?JvwcwT6cfXUU+u5?GBr$T@xPZ{*smPeeziyJi
zCsmv4p$vzjoK1hpm$RnDkO4D}Ht&OH>>2u=c78dCd}aaH`tF?__}O6vzPzW4QirA0
zLj!;A_mhbZ!MptpZIhR^q0AIkO8sf}Akn@gzDW}fTdADx6Pbf$LXH_y#KUI++J+Y`
z+l?K*Xr3%u7d?e0%^8Ewc+Rl2^xNz2A1H<oKB=aN_RE>y(!rcM@8WrGBJPUts%68C
zz?tLBrDhvZyL$}Q{0!PJJ>s1WFt&qv0aa;3fb0y<x(e!<Ul&6*HByH;AFyM42&X#I
z;g#BH-ppjL5A=evH9hYfBBIP<?Dp2dZ`J+Gf(#&NE6F4gmo&D5wsa+Ub(2kfMXrN(
z1VXFwWR!vBNs4&ssjMdZ?AzCg6^eV+I8}}`rkM}OB9)j>r!xLAVZd8fs*8AR*8a-W
zmh^;<YAc`-$UT)qCHFdo@C~K6InzW_(q(K}2)Q|rV~mf{cE=`JGFw%(?4LfXNRn9>
zHv6axA!ObDCrft;YNPRZ%%3$*dKk@gWSI#a{goEesjzqXItC^kq>W8cZP$070Jw!B
zRYd{r_qXQFv6yQEuOS+o*Nm#~epS_-Px*5Lq<zHTZIi%D+$N}&jgO3{b8670<GA{D
zOhdeYFNgP3NIT5cUlYZ`%{7ltI+E_@_qJEOg(}%aJvT2<oRbE<DEvVGIKksuM)r!G
ze5hbh(3{_j?MUv9iDkeX_;U>A-V>d*hz$kKtH30haKv4jO$|f#ra0^Is^F7T`%As-
zj5XUq^I`w6p#QFpRf0u?Kiy%bllQkiB86=2p+6-#8xNy*CwLATI!l}x-I2^9<3vv&
zK8Q***RzaQ>$-atcpm%=_z2mbH-4SGCI+s>8k}2;b(z@Vu8*Se#U7iRY&QDU6we^G
zsAQzsgNxRkAo!Wk!>r(`AbLS$txyXx_~+aQvqso$RpHg}d-$5psNIKu*S!}*eUsr&
zS&yF_`buyi-Dh{qp4HYgyC>;EuTVgz<GPY;g`U=3PU$=!)(X>j!FxHLha<$wTDLH!
z$HCdj;*Y+5w)4c>Nm&cyZ{|H;WX$nY1ipCg3F<o1`ze$7BItm}AE|psr%Xib4c{nJ
zy@BD^Uyr^!AXVbPXDMBD*Q^|L5Ix6I1A*iVW_GiG;@cRl5&JoI%QytT_WLGkf7C)h
zrdwCnj~B~vKl%w~ER)URiNM@0yjq{&%}M-nZ#Q1nfs|bEZE2U4*8VG+F*z1+2dP&5
z9#j{Bo6tv(HcL{1V$Vs6TnugAEeNrF@K<iH73}fuuNpq~!^Bo`xm9l4`Q2S|=NpKc
ztcn@{c)|c){`a@r?bI5tK^fbIoPIg7q}wMhV%e|HCv&`xlkav7K6zph5QaTkFRlh=
zPz2Z<Y*&K3G(=BCEZ=iJ2MkPhyZHFu24|0R0<OC5CWgk3<?#Q_3v_i@sd}GQJ@&8J
z;7*ht?)EZxO$xm1HIOtJIDcpzx63jpFm`=-zc_3@xwjL!#42czEAkzbGHLo;H))D0
zan7XsbHmBNT^O|&GsOX_adl<!Q2+hFy>3lfqd>QGJZTdwMtwM)7P4|)nAt!x3WX%7
zdU~<!F6litXoT2>7pCy1d4ob3L7jWwh=#F9Et&4Y*ncht;(w7|YT(g%<&h!LvCsij
zd$JGA0)cG=i)>bNT#Cd-80<QY-DYy&RtXU9WLv@g>Zs>(OlplLf`*VI2^bh469aUq
z)1Z90KZ?aTnyO^|xY#FPt>75KzjnVBbeh<fAo98@MECs~@H!D~#<0pexzo#9)m5dv
z6qS!_=i%M)^5rm>wq}}>dqVpAXmw$+zS4Az?9fT#2H+qx72fg@YzsYHF{SIpcRwPr
zMAMJ72vIl*LeI%~6nJuc>6^5V4TU3{eO_~79uW_egNjX^oRjH^7kr<QzouIRo*odr
zlW(94Q0iIU;FcTg#QpWTOR~sOe?ONmf+qB0^M4&yeI#e&|IpX>I={agWaX1Acb&DU
zQS5Hm3!HS#S!eiG7JJ{dk6(1C?@Jw*72k7S5T}c+O-it!nTv>4$@j-aMPJwF+3bnl
zen;z_AX`iE4|wKR{Qj!T@{@)NwlXuSF8ojVrVv77z1UxA^&D;Z)1o4=BF1AWOh31}
zM|Yw{-f6hriw%<2bOkzoyy!U8t?=Ty_jURCU2k>0^}~zcT$}BW8hREgvA!yG?-H_;
zG7egI(%4y@r?V?h`h}jXz>vY+T-y5z67Dr)mmmL7Z@IJOIVdID9rU^EAp#icI5?E<
z@IN#6cbp`2qBM2x>otu&beE4Ye(ZT=Z;(8*V8nEejgBVQ+Fco!@;E?R$s>^W$wDC!
z;x+C6wxsag=6^aW$cA$kKl|O&Mhh5)h<{_4cy_Kt<n{tG1~iE#f`4ip*!YCur1+9N
zs+||@4oY5KPmS>zm}Cc&u(c*q&QG}5Yg!Lne69~z6*>^FyC3?oPFph=LNdjEIdte~
z8aQcHroZri-c0z+9C)-8NfP?8JRPqw@*ZZY8M!%jdJvDDIun^|)DYjbj`aB7=aGo9
z{rv71^{Qg@S$CK%kIetV1$yjbW8+GJa>^vysv!;{=-kBX7<k&HQeNFSVloeWH_1Cc
z^6Oy)K1CeNPh%O1-nJ{VAryqjYdjI15g}fYNTrrSAv+R+M;@35JOChoPKT()UUy%^
z*%(vb6&g;hX+>B>I>FFhrdpT6y5+UV^39Fbk}Zb+b&(UJNKtKdAjU$S!`jwk4do(*
zMLhhl3sIh5(?s8L2T6U+g$7%_OkMowMxRZm7&>1<d@^j-4Csk-!HJzh2}BCpZi7dS
zv%tu9Q<k{etuw1aSKc!Yd%ZNVhdF=eO#HKcd+xD`9!(nO40#EOjr%c>^J&F{cb-Qf
z+4NQF2rXWFy`84aK|%}*YK!I(c=hFw!b7ds<GaEAxUs1TUD(Td+=#aM!18To2=V=B
zhuPx~_!NzI+*BwY)^fdBatC6<e`o(n-~6e2k6DjjDyk!zeTT}^+y&AuAgI~^_j~}+
zs%Hr-fxG)8U2?dp@2}QDzK5nN1J|N}g_;YMUA}&<9N6(kZ&0|&{2NU+M2w#Nc*7Ij
z$HTUy^SxiQUCs2$MT5d<+*cnF4GAN=F4~mI#|x(+XTzthA(ZXXTCVs_X?*zB4%w`U
zw<pA!s9C_5)5w08rh`_bLhIebOcIt_aAcs_W3}1R9m7afC|}+V9a*Re@vChWVXcqU
z)`@&`U@hTnAdyCzj{yzml2YbZx5sj8rw5CT(i*j&IkCO}`i)VeZ`8NY#YG?Nw*CF<
zYy9cw`m&u$T>l>vRLsjE!Ukb-=?rE$!1H2OtC3`@bEEJfnf&4EZ|07>qY5)xI$L?f
zSIR#bjhcR>Ssxyk7rU@`Ug^RS!LwVGa)&dmx?tAfCBY9^hEa=}krQ~FO)J~sOcl-g
zQwngW^CP+#{IH)N*)NO4T;}Nw*|eYt^~qXYol-(D8Kxx&r#K9F&*+hR<a^@<|IU}w
zeZNv=Bty>iZ^j?>3`F)8s6FM*w(V0$ydjoqxf=8KeWNp#8J4Q{07+>Fdntse+6#?S
z9fpiB9<5PmT+&7o9$fkU^-v&Hh;=&e|0C-kxGRgcaBFzSwrxAvVa2v>+fK!)uwt`f
z+jgpA?3fi)Y+G;6Irp~PzV`iv)#h4j%rTzc2WlM1$q`2+XaGCwc!U%YjXu+5HTA9e
z60H`&y4#AP5qI}yP4+G5CkVop5fM{3JH{Y5UQ8)`NrZ+7#bKSVsM@hI)>+?^vL0_g
zi(lDnZtQZVe&uh>n9H5*_yMwbn@9*oRkrcm@OtHzh@J*55oN}Y-2#S}IY^ZGl)o|h
zns@DwLT(qi>?ayOM-{i;Hc{p@tPz{A7q<R#W*{BAe_UE~LSk+vOA%^X%e>cE8X-+t
zE+X1fkVfhnj9CrL1o}(e8!|ak=GA1z7`CIT4w2I>_|_d-{kp_#LFsk!4D)O4EZ=WZ
zaTwQijVy1N${%foL~OxF4ZM7ID4mI36L)+nTt8QXum8vK_weEKR_T*)#w)PPHi}9N
z`g1@16IQKhWWI@|2U9n{Fwy^22mHo=;u4~#&8zL9Y^bugMb4y_`R*LP=k}wu`5|Gy
zOVwgF#lXFP2*#(Y11EnPYQ>Bp)M^|*+Rt9!eQa?$4mH6l+kq~{++@v5305TKQKHqZ
zP1v2|bO`2EPrP)ldju!SSW*pDB!SYnezG>9d$#Ofyx`+1>ec<u)4^k}kP}9gij~bl
z$%AAt-#WyMUi(2YnkKu3w~oa)i(@>;%tpFaX4)g;MAuWCemMT;98cP3=33>JgW%@;
zIlX<-4FCCpyZ2;MvQHzsA%vmVLhqxjn+EK!{%+gipM!T3v*`nvv1<WxoAI9`fzL{3
zSX){TA}7NA&W>3OVIYzcgn<=8a<dG<zxUe|$GamWB*=AA#^jGuPthCBe@c##$@Bk4
zw_4i62l$foOF=n?x%lu6!uMu{S%QUq(*A3SoaB-v3W3LrLPu*~zjMko&yhH?no4SX
zV-xu&`GZJ0gkmab(x+%F<@^U9IHGz4DhFl^EOYTcZk5Ft<&9_fa>=7Q7_||Hm=YrA
z`Tm}oJ11jsFeN$PohFm502)aVv+p8$;ZhXWW&Xn?R-Bx-jRz)Kp5Bv>693l);6Om{
zSNJmPy^!R`EUaG72r2Ee3?UpfYK~c3L*aZ3(Qlqy6ZTe*^>N5Yu^eaObg}N_aFN%u
zx~yugrzZD)5U1P6{LA+M4=sGjrEb6N&ZZfH`}1n#nsK2|G^?MGf#t3|%sVeEMOT%<
z8UDkHCePWAKbzCnJ=j!hbgS=!A}Ma;bKmLHw)A@dm%GOTPt}zEufP$P&46WrYyFlF
zqqkX{A>chj;FXW)O8^ZvsQNf(hUb14t+R*V&<7`8pY)X#Jw~S2>5L&civ%oZM%n+H
zpvTuem;H^1WVrdUH-b1<gKK#;TaN<iT&ZlUhqtZq+<zvx1d`7);+q_nJI}HN4F9Gi
zRBcqu-pNvgiYAL*ttM>TcUVeQe&;F%Ll@H}DGmHAkBcJS_S?-@G@kSb-nl%#YZP$r
zBZNZUmG`cS@g=_#F#8M-E}Y-`ZmT7us_S8aQni<Vv>adjs=I9@OOjirj{-kRSi)m3
zr8v5LbSv<}sqO2`4*An<r<M*3^+p?5EFEeMrtU1BQ?q-Oe;u~o73n!4i0pnjUbtl^
z6r&pAZVNfSr^f9y=t@IfVs#^H%hc5px=yDK1U{{CQ*_2S4EtN+THQ2iJL(p69)w-o
zOBR+Xpc2(^BNczX=Jn>*wqJF-KYVosxlZA~ac`%5jrH?30<7*Ts7pGCNZ@nIN~kfo
zx$-$t?E%ELmkykyd^Qw5p-YrG-LsP7hN=9$CthTc;0BOSSBIqAE69wHz(4PK3T5O%
zh)|3C-Zq5FShofKoSL?bS4BwY5#YTCa+`rKJK}mN4J7U8`0;!%IK|~mV7vKjS&Ej-
zoOERZEMf#4fIQt}mTW($e$#1i#C6qO@@;2^gnFY9OKtm+Q(lZvWM{*g33fnU)e2vh
zo~MtL-N_$F2dMQh&<8jX0^a(7x}<RagK*98rJ*02A3kuu`&w=J<}WwZ`#@uU-t)g6
z-Jbi~3Z3;T=V`%Pj4}zlyn%R=v=T0ft>!RQhu4=|JoDlwEG#D6qkIQ$e=JDoI@DT&
zdAdXI;R+JDHj{sqDC6nc&gj4UVr^Euj@0J=e%_@A4qy5l>Sk4AWquRFY3%j9^%|18
zQuoujBeD5&cuoFB1*IuumS|_B3di?A)i%*9o8R?$5&U(+Cgoh#WN09zHa743-&#v`
zP&w#jN;_;_iwI<DgBH^!_8%WKlL@D~7O=Z-h6&I@G^D(m5-vvz{K(8Y4}&9|J&kaJ
zs9)6=mgY0paaxP#3x#Pk?`^j-^phiWu3`p#pLU{IlSA?z7DzZVz2)+LeSX9`p_>K1
z@_W!?rA%?pJP(rwHVzko-Tp+DKciG&!<;TFJ5^=5J-(m^N3~z;=DddxK3OPr7Rszv
zcCAYGV#*XUcXj4zL6hu=^>u#F8k(!|+M0@fn|8m@DJ!*ealEH;YW$t|cEWv4WGzW`
z^cC5%3JbvBW>ciyJ;<E~R>VUeayDd%4=L~T+Fp>M$E(=mDut!a@HKNwhTnYCz)f%T
z6uuRlv1=_kUWCMl0*gAty6^ySce^5N3sXe~;Gk(lb5)lG)j^A^QB|=1ndc%jLI-jf
zz<&JLGl}wzx-<Y4Qw}@ciD0<&X}L$j_?N3B{PrAXHTqhjy=3$2%z6`oG!d9?3zkxj
zhL_JW@Y^*gEF||a1D{kTr~rCcP(TPa!^QDNEtzSZ(+>HHW;~rFIvau+^ad9VufxOj
zR{yvhUDu7@YlZX6%TYeRy>52<x@&PmW?<E--*3lQpsuMsUSwd>;86(m(X}RTO|;!~
zomukeX>)T8`y)VSBI(r0OKHXmySL{4b)vKyS{%(?`tj{SB*d{>kCQF;R@um#8%%qF
zatwx(kY;-_@N26)j^_gG`@gtvRekpc*3FE%uEtT$`H-Ejm(gpE5KW)MtW}3nxqjiw
z<>A6#`^S%vj`w!sGClr6U>v_+$1acfU`Ri^CIK17_vD5(C}aqXiuAGeerlf&b$a&c
zhQ_iq%ePoct~yZ|MLQ_0u9YHPS9K)!V#lLX$<lcK!uP^*w<nRAt*SC@d%Ib_ePPxn
zhhTp62uKIQjT-+pTfXODiX`FOcN1t^*ONJBO1iPo=q2;9Q>fnH2A5+pvRl_V`d}mA
z3I!Cl^M3M2$$r7hKwrbAJ~fJf7EL!gM!lt9nl6kE`-XT_<-zy~hlaF;oz`>C!<m$?
zrcmqvK(b-1<bJ3bmpIqGO^}6yk>qdJT=;>&hi8I$Ex|wt{}$@Gwqh7ov!?-vzPR3n
zw9wM9M!q11ZH7s1FjigG2HXDM6P^psp?NUxTqj!tUSnYkp~J<&W-=J`$rjjfz&3;{
zsH1$IEDTOKG+u;$vbY8$k%&Pbo-#F|88M^yvw4${34pkUaFJ(dB^W&<D-z+}rN)Oj
zD{6?41%z_{jAW6LhDflPqg8JR8BwnXuGZ_o3>P-My%C^5xU5tHME1uIw*)^8EdZv~
zq~i*Tj#WB^BPlg29$9{lGTGc#SZi)>I|#f1E#q#d%HNUqyV$bqOVP9aL+;$MF1la&
zfHNmYax_+!#0g?<p3IA>yjDxrx7(2`tE39g&#z%!%zPGT+ThJ)k_6ted!QrYkm5GM
z+%-;BF7{F;pp?a^HD(J_C?kK>VSlSA-O`9o_Vw9DzG~sF9o7`GzAS3NTkd|fc+8xU
z(w@ceBK5y{ux*N+{z6}juC?36>_Aj98-MSQzv+D$A6GAS^FHP|g1crkB{N387B5>L
z7c6L<(dqH|ySXsLh?Y--vz8i%+qZCRr?xGkV?>XeIN&j1m^?@wv^FHC*BdZR402h%
z`?r!-<HiFTp06N9ZRg#e3;3@5<~L!PQ<`g)EYVfcu=hqE7j<w`FWkq0JC}hkRei9Z
z5gjZG8{v9~7YOpYvB&D?(ZAjJrndqfdoBWknxm)pcs#nh{}m!w8dCP=eC?JubwQS@
z{l<@j5-?fTX%t1#68Gpf2I|HGbT_$}iDNXyl!6_wvLRkKNI69M(SoyujS?lA%>tQ{
zKS-(1CjwiGxUArvOh#(QL`v1P3*2(kjoisx9m#@iD-#rIfnz-IV3f%+c(GA|3orM_
zx6AsR6dN0Iu!jI8pbL)6y9Zx7;;r9Uhn0z0FSGwm4##}p$9i9+PsKGW#v9xstYw-2
zq16=e1&Prnv>+@0?fzk&9<1P#C}Hq!`5k8ea0C59TfHtzSSvU-Z=YnSWeaZ2iw`UD
zr+FT)03_OGe(hfNB{{Ql-xVhvxsr}nqhSVHL)~ZS$V;ep+F#bN#+==H|GM~Nm=I5q
zD^!~H%`xw8Y9FD_OK-L(azx8SoZq)05HcWR-zjUI!H6F>@BsoFQV#b#?^R6b8Nx$%
zC6<F+O$UW*8VifXayM(5e&Spx*wILm-Hkc;@$&E1Vc}qM{`L_W5oY7>d+ht3E^fjK
z4OyJ;Ewz1X`~r|J+rA6w6ERyanRrY5$p>XRYQ~W(^GZByxgt2XU9W@p%VE^wU-t>u
zG4tFJVh{D|vpny!1{V|`LykrJ-u>~u50p&503PV$>DK^)#FCBWWkrRbmn82~ACZ5;
z*Stb34w5VjKQ=?G)r=%&Y;--yFLB2R#lzcE7o5%6Dch&d({JTxK0MF)%^GbQX9)B9
zGTn)BNT;ZLPW4$rq%L09Cxga<h}>Fu+)mR+{ueu2Awpg-c*<lsczi`ZMRD5v0Ii!8
z{A~yIRz`}m8)>-t_?<LHvv%?2hMKKY&~&VsE!`Ns9bAku7#kQt!>^<V+`NI?x66);
zE$|b-AJ+ALeDBuQ&+I(rSRQHFC=2>$WsQ>Se_e97Z1lW~C$f0E^>*<%B(m{P*SS~d
zj@>%o^JkK^Ve9>x>3GHk(w&Ma8GKidCoXXJ+b1#oX||y65Mt1%#m$mFnNa)wA*^0g
zANcFjgwThoxa4fapVL~9&6na`d^LPua7Iq!e9)o%4@tmVl=a%RzV9E40+ad}k4s9n
zv^9Jq%_Nrlz_(>5i5mf<jOVZu&hxa-W1~+d;YdNuOh22?dG0=<!+V@DbI7XC%hOMM
z@UCr<f$f}mB^0$#er|vHnDLTuTfrNNmytkWSE-Z323UNoO%7h1y5o#GhenG!kx`!W
z6mqb_+iry5e)%&kfN-&C?q`}~_eB;cNM|cy-01N2`Wxr(x^w#G4VM|G`I$gBC48FB
z7||Z?DKaVJPyc7CJ!)Pud5oO+{=F!xJf*FV&G<PUkI&n@Kq{gxlP+b`zO?7QkF7vh
z;tlunMOgUiLGmYQu6Mn`<IU@B2){h@pLGH6i7D}^Qm<H;^5noaj~4L{_-NX4SRO>*
z^>SV{j~K^aKeow%>?s<01zZDW94k_LH1L)tgMtn$QVosWhwRp)Tfv~>)LwUKmKT-_
zsrBh%OR^*|SEXY@#}q~}GkEA2O^Z>3f%f~w@bVH|$*#Gu<W92h3MUZ^0zq?~8R8A-
zHP{k!kc^&qdI$lWe5DA2H=JOxpWRxFIJe^-Hd27b>PoCE^&#<0_V0@IPHlpQ!^@^l
zpaw|e47mg2x#DqtLim-BQbdr6AlcN9O^bq}G{12DF(+Lb`%9z1Dm1Rw0;zwiM-4Nb
z1a)0pxl_d1uSZgCALMUa3-EH`q*~iJ^-ug0<~~~(;iahyqx!z9C1iUH{l0(M1aAkU
zO-lDZ^j-A>-@;~g_<KDy*L$5rPe(Z>k2F~lobSIT6PZz<ebbEH2O>)X=5`zSeZLrI
z+UM%my3eN>Hk|j*8>bb!xczjmO{{`&fi(S;j*m))BBZ=PLASSGf?<}Ml#zl;Ul{Oz
zQ9MzedXU$TnHLEJ{u^-(s?(TL`ks8TKDeQ?R^xvSjaA+%dyVi$wrmUDh)5_B9gNmG
zr4J%Gr+YOlv0tQxW`vu;`n?i?k&(6Fk&t$owad_p$JaZhkHsK6EJ;!s&a7^?J>J7$
z*LwYE^a@=yL*myzvs`PIBiFKTTjy)2AWlFHeKSa_cvH57Csda<<uJdyq3Q67z?-+;
zNWw|;ftRLT(Dl$Yx_^_lBN|qpIdg8W_w7q0Ve)i5;o9rNm&=b6w($11!UKQ%HHY-<
zd?~gkgL!8=)+tOq0#x9GrRq>QV$xhxLnU7njuzmrvYeP*Bx<0L3yj5w5?^pr#E64`
za~zD~>t#Ff^74Ybn(vl<y)r4}YDpZ>iw!ZVx<xq$f9DGr;ckts9=H3NUn=B-$N$!B
zVgwuW_TV&{lH_utKhWl}w3DQQRnAX`gb)K{E4KCoffJ27coF6ypN^@0VhXYY8gd1(
z)<Kec>;i;#C&dS3gIggYlf;g8J5?e?6Dw(o2O%-Y{c^1piw$3G0cLl2C-Dq*JMw*g
zf;hdPplK=~SqR2stmh)sRlcsZ0zXZD{pHzVflTT<HUK(byen4!(MZO3I5@0V-?JtC
zIVi&6NiTA%ZN?zr$Iuh_^ZIDtXUDttCf29tCSI1w^K`u*6CAbvB)WQ6!mitxn-0P0
zZGkD`0nr9ai~no3GnhD2=RG+D&-+Ep&#faxS&-s&^e=<q?}BxHrO1sudmPriR)5cQ
zo8oOpF0c+aCY_cavt~%d#_c>c{u}hCg7FN4K=9{&s{YLX@X-3c24ew**Hf*`66zOB
zyLUnRIlK{ZvBBl#`PZ(N-6Aka-%HAaneUNMv+J8q<||XFG0OhGhDEqF4HlBCRY*%s
zFl5h#3(*Okt@PT!%)OcE&`<qZ&#oeuvm?LSQ}|84xAEmeM?7ZW)~ToC*y;$vuR}l5
ziO!HdCY}%mIA44B3i%!BX&=g6V0eOnWG#+}1W`z1AQCK($D|v))WJY-0f2txPW)T%
zZyMJ7#Jd?>3a!zjC)zC^vb|f0P$fw&u%FKy{{C!(@xcKEsFpF>`>3Xz=ry6b5d0fO
z22hEXleL3&KR#?*h<GtwLzhba!sEk>t*vPb?RwNctU`|Bimii2HDO$Xd;R6n{l^<^
zMHyKvn0~61XW^EK#^0(ym@Ju$*cgnEp&mi_jlqoY4)zEd*7a)xNC!g915j5m@<jpc
zBG1^vh1B8-vO=VChxcGX1!e{^gTHISVM=}td-8=&PBp|{(g0bSzllpP5@VBouZp`A
zV+-J@<A#Q1a5INoPx~D)Q)S>2r@bi=S8b5vlmwn=*yz2SblCcNbBpEJh(Z=9#Vo#P
z`Ry}-{(Akjp8eKfW`-e8C)4vVGxxL;i?QR1bL*d$Lx*eM|4;-HEEw!x@gq{44gRy!
z(}r@ZWPfQHT8dFUH)f?qDzu{9_bWAIqJto0e}@rH7{6IQ-{u5rtOay@|9cA$QYoe7
z{?w(CD2&a-=Bd1x;pIGZzCaa?oY4@Qizl-uF$1H8#L6jZXdjH&r7u<cOTI0IH<F5=
zQefHXG9bsXU-`O2fbPMg!m+JQmunNGVfIj@^fHclt}$Y|Z`vIO5I=shT0Qqaz?VQs
zyXwPq7%6ziq`7tGr&sie{n6ep#cT{g^}j;;|9};*DG<?yyA4Vm4bi{O<-<E32O{S2
zb6)NcbQHiv-YM=Hc<iUhb{OBk@=s+(TmF~K58BvU41@qg#bfCy*@{xtdp{*I)TX54
zR${f#@U&Lol<fC+pf*uGS}E?(7eylHWrcsa*bMJRDxakNu<TfgOYfjG*BYeeAp7k=
zqm?%*Q%#wSMM}L2dNL8{rdnl+7heLe+eZx~uoVSXu-o}$Xs#+QChEB{6zR&WHrwiB
zFg2qTco<g=i`ERV>O#8tlq+0k9TG8%6$f%8D}{^kOi5+iYM4r7^m5jya<FrS@ACV5
zqT$J(r^#GSGS~7Z`&)~0D)<nz<lM3qYCd@zgO@sm7r2}Cl13FI4!+R?sjUq<#nwe~
zQ3=;Lvm0kzmWuE3ROZpC?X20mYc+qn*{VQ?vjv+G>(R^x^`D}7*J?H-UxLI#vg}jU
z{~xaM#YK`Je|g5wCgJ=(5q+;8ZI^EPKGW;U%;MIBAybLZQuOAR3O4VcW$xYM)H&-|
zJy{c~7NXaC4axHAt1+f3IxzICAT7h*>HM8gEz2g^f>{|zMP!%NFh(L(LhtDy4@-e;
z`4!k@*B#{m#S>{_>@6nr`Z1{NNXJs=I_0|LR->*ESAxW>n})<n{<`^AV_~q3t0v-Q
zib$C4{xtpIY`F}|#AJwhvQ)UlunnT=DkM;~>`yjl`#wv$x`k|Y6qCFLSa3{`!}2p1
zI4$rMy%|<kl1&XhnQj~&otkzcVW^p27(t~0l3_<`N1LtIcEC%Vw0K6^j=_-h4#74b
z<%hr#u9Du<$w?O*p8-#mQW9-FOm7^ghAIIL!SQdK;!aUB^e#VK+m*(ej7H98%;~)l
zA`uxB<KNZ$YgGjAvNU%A!seGCp1yDLBS*__V(s<m%~pgfi7dmpExJq0Uw?69BWL65
zzUW6NoN>V~hryQ|;^C1r-S(4@Pw4gyA=>ll^@OasT?8hCmEo!gw!#)T9D$$^3inRV
zOB$CWJzaq|as&6ylh9f=bgdrT8zl#OM7WGW?5P_YwG9c@j!f1aqJ%o<i%$Z%a@V};
zCzz!R#+6(>sMM${Btl{<21%2|%Lj=-4QGGps8CeB$=y&W6fWiw4R??E@eHcaN1B2S
z5^>&szS`N&d6m3p53La{)p3Acw|PMa(e|IV*R+bU0@;<DnfhHAB`5HQ`oxf8WAT2Q
z;3n)82#ewGt^~-R(SRDNRVrGgdg+5imb_uVm!f?@F{2?N;9$M;-eI9Zf|-QS{77+=
zZ?$3{N~;{U(q|p=E-@jD$+npYcm=tG<gK-HlsI<bm=qDR@Wz|jJvm9Si3@W(L}mX|
z1sQL-=1-#bI6e0?ZDPN6d0UyW)DiIz(gu$N-x^%p$d5-T53v7g;<U>WcQFw}f^z00
zwxq<7Untnm5LE+7r&Pi(pAfM26n1AJi$s*FsHtV>w%kLm)aOngN<X9hrJt-u&4t2V
zb=k<pI-p==76nF<%O`JO<!G0&CX4#!V?%>bcU&vx^Vb057NTkcgu1Lss-v2qnP5Ml
zNz%!z;88YNtX7*P`sI-g0x225dN@iipPjruL716h@wY#6Ya-V$h!02(`u4!~0#q@I
zlWrk&HeaAStfd~jMhF15xK^7|(%7mq+<LnYod)ELmf^Lop&ruG+Wy!In?Dt|{i9!5
z)qtk}D~#aGHtBs*ap(~oGuyZ@7Zdc8y7Z*7!1LUi?;ec03Hs|uW(EHD?e6d07ml-R
zbMj^KW%$n%Y;X<I@{;+YK|GN|#pP2d2e-(xZK0_aJ7nWCQ#!;aSvMztb_9^BS+fuO
z!ie4$lv2!N3X+6i7F$m?M|Ot_6OO?CB@42f4mEZd>|=yL!2hQtL#!&SMNwFxh9TNw
zEJpIvKFvx%p=`mH7F0zxdOrvwUK6~A6pXyT#yW!TzQ^v!;S9$?`7-Z5Uaq+`rW<V=
z29n;pK~`bi@m|n(hbUgqm#(!%hMxvVOb;k`Olz>;r0#m1wUQkrFwTeBq#WoD;CM4q
zqmRPsPntUtqsU+|Qan`yLUrn(h~*4f3e{(^Z?AA!5=K=wMb_hta~f<mQ+<fBCjNiS
z7RGsQ`H;k~Ir|n7(TNZ{`i8MaLaA+`4mIj-)v%P3Fp>a@FD-T5e0Md1@*MVv!Y={X
z=`8k!T0yq(B{LocCB@Lb5#R?Y)`{Zw3V*mwf*{uTuB|FH+9+qDYQCmiK}t>w`i-@<
zn^REIwcr}V-lg%c1C`_$7wZXOp1_C>Y$#yS+;94|Lh~=}&{vJSU{3p%y*~6Y7!NTO
zj*pa040F~@);<%&JWMWK&jc1zY!tZ*pn;Sqb$OWV+4&mqP7JAc<!h2#5#so*L>wq3
z2@=0lz|!iT;;S(P{Sgn$QV;utnRM=cNk{CL8;C1a4L@H@AIgOk3c;P~Mm#1MgD&zo
zP~;uEUg>?7$Wi(_4*w5J?*G^ohU0%Axtz#-M2BgO?nAf~&O>g&v&EpCDm2~C?MAcx
zh7TL9uk^4D4J567A?T@O*9D;Y^PXhuw*EkQgQLal7oXLti9r4r5_51+84eDhFdhxe
z;-jPLYVohgjZ|r{U`*;I5s8t)sF8Nm&Vgf!SOn&|eH2JEFxA}_z^R=zW`f~JgbiMX
zL`!zDe^32_w?oCexlRkKEj0@M08p=lJj}9E*O!%whWPRL_KSEM4y_?<wjO{bjTg#`
zF><V=<+ycVupv-~PU{L+b$F~CK|;~IVw57G$ouP8scbV5`Ow~40OX(EQ(*UCANK1)
z1H;Z1g^}SJF<fi(LjUhke1V)_BP?|9GvC~FfeQ=d?%RjguB9JI%*9T_JcDf!yovd7
zTPC()+h?t;KB8F?qnzJUlN_Gh0kY%`q=OF7g5Vf3Oi><-ez^ZU<(VL=o+KODZ;}C4
z;9%>mm}^iJS*kwvPr6{yx*0ZM-`i4L{IWta|L7Z=3lLQ;(jPIh1pTaUjmvwM&yP)@
zUah=VCMbLu>ZJ;nP2>j^2iYRf&Tt+Qpw;tm9JTW<-N_;_aD=T_%6wAW%p>mOEr=td
zOlNn`G^~~_KxnIokt^c?LXX;+;n2e~<Os}E^~i(p`(65enH70*5~Th-g<#t<o%F2L
zNtLqwCJCkhjz3kO(_}%^bYN^NL#X!ucS20;dp=)`)CCyQ)FrYcIBQ*5yM}|YB5F2T
z;Qfx(V5gEqdjJK*Qm$e43N@3)wQ3zIwW5WwAz{K!cl2N*DV93bE6Zf0z9<@~wWp)c
zCWfeC-U{WA^kDQL%>|QM&gpUg9hcaQ*liSgSg^kjjAgt_E3~tX((Jqvn}~FT#dyob
zW}Z3Ct4p%K8ON8Sp)0wsSP!%nGN|Dohe6Zo*AmUMAu^iKbh)hdb)S$-neKP3aogrG
zt?N;>CEF};<iLr!iuDcLkRqE<f04nUkY4<_%GQF^4nD3Tvle#EcV5?nE9T)cxKv7K
z*4y^{8YGr@(gR*t;YI;B_h!vk`#RfLdnv%K<}D)_axJ0DhEixF*UJ*gS35UHsSR!9
z>t|ITK?t(1n}lhKkY}v=kC(<7<ofl=Y3hAJ|2}Ctf-W#RJ%K<JV9kxHJhfBc{4E^7
zKOpPBE?cNhR$p18OtB6BD$eEA=yD=Hst=^fGSLz4i)S4GcNZrYb_*bN&fcP+e0Gs<
z9;THuP^brH{`A9Xc8D}k=!WFB#iH(T+^Y_MSOQ1g=_Z;!@tpF$<3!UxDl<tEdZ(~X
z<0AOi4KNF#RhK+brPb=%;kv@OZA7U_zUxDk*4`u}w3S=<9?De=Z!QTy0f%_!D@Nl4
zCIsNXf4t~eV_OI&Y^!OkEc-<t7hm(e{MnOnE&^Hl1V`&8g=+>uz3yg|S7f4E+v5=<
zsZ~}C+W?UI!Id=;V$OU~W0UfwkRpW3VBEN;s&zBcvyzX}U7DZ%ve67g;V1q_j0XoJ
z3u>U*U^x8`O6^?l7R2o;lL=A#lID);3K(c_`DGrU_euelJNs)gZ%1<r63V8clMWd*
zf#MTdH+&qIT1vKKLuj?TGMA9#{)tvSvK$Dtf=WpXw_E_GB3C%yVNBgjZz5}<ZP*NF
zq8N5CpogXolIFStwW<?sGEnTH?_?cl@mPgcq^dLmHRcfI6l6*f$xW@Bf4ws<n6C7G
z5N1N6XPIOoh>@Z{3t}#D)`4S64=u8_=P6X)C_%$w4E7!v@HN^2Ecza7?idH;Gx%nm
ztZA&(SZ7O&VAo>c9-v?qCdHT}QAnjY=tM$qvD}Yd=fWz|@}(0jux}$X{&+*O)rx1E
zD>Fs-qk!)f8X}?th96bt26>eIb`co#(*9EfZTD{rb72UwTB=Q(Hn`^Nf9O?j$@A9w
z4})=n!`H*NOJ=!#Wsd^kZ{Q!$=-_Qy;Rdus^l5B4lClZwL5QSh;E>?j_249|7C-aS
z@sX!$huy`c=7V`P{mtZHcfk0LC{U3YZlw>wh)ZYfV=CBSNzht|g3yVUegID!nK-E;
zUttg8tN`<s0MQh@p?GoBx$$oDYvSD{8@-w!<Ko_^C{gMoHRC#}x(S$w-Bv6Z$9EBT
zMlI*dBBml4Ue6ZF5D%hm2ezE93TbW*QYLh0a52Q?b$U`T_hX=<^1o)p!{?lNv6_ev
z{N5*MRwj<|)HNEgMOOZ~M1#2nMjCJ+fsdZA8UAma-@RgX5K53d53@Q?2Q!4}QIjEI
z2rpN!jQ_P0{fLIx|9ATDsdwx9?YaB!mzL`QrcKZ&-gO$&@L|WA)!oMKAGT<s_(DkH
zZwCrK`=mQT40rIW3RUT0$~6f!^G&?0@MPI`FKkz*n%-L4EIAEm2V>H;&q>SFPqaaR
zk(Al)%7KKgpg6jz$o5Up9JJQ)Byxcykb#3*q1gmEF3LsR8PWr67zhwObit%S+plB>
z0<2MZwjx=Hm&=zW3wxqmX2D-$0z+%!&T1w_NRWu8o5VUkh=@;t#444hlz*hAKh6`~
z#_3g>BD792Np_Jjn42kzn&;t+l|eWuM_MMnSqhvGfIS_LEHKFp60rITmj<F=lSpL@
zaC{5yxTf6MGMwL>hCvm(f);nT%Gm{_;^^M6)VdriK~_oc-iD3c##AIDV_)d={}C2^
zMc`DCtMvk(CP)1q|3v#AOl}mR5HDjIq;w&G6Axu?{bl&!WaP18Xwzojyw_jrOe&?R
zDwjVaA;@JU$w4htyLEa>)?Yq?0<Oc;OqPWb`;WBNC|uMNDp=^oHD+frR#FtMon4aJ
zX{r`AVCb1}r#l2&+~LmxNI5G`pA>eb5;tEKs-8pI+pw8VP!W<Ut*-n!KlbItb2um{
zHt*4!1P;Nb%)!UX+8&6M9DZ$4FK`Oao|Tzgd>Kkw5LH7ePJ-g*ZtYScqyF$&Uk#XO
zw1oE}F5rbDNf>+-i*T>$f&UA=M*MLctzynhQ>aNGOb(=gvgTk<-6Tsau5Y+%$$%ck
zMM5}$)NW{zc~OriXPwOxSBNOp=F61hYn5s@>4j{}0H^{NT{+81M>c|<_56vpa9$xY
z^gJLgHnb2_=W<=P_i>CinxH+EUsrbHw+2#g?sA>C&mUWhhnCMw>th9u>#MuUd9IMR
zKG$wujnubfcyY}Wli>bakHhjz?#?bQ;w#QGJyCv%6Ng=R(xZW-P=*SPm~-p1tQq@V
zSP||}o)e1@;6Md@DG-Vg0$8|8&7@LrX%RArjnk$!p>$ybNgLIel!qY3R@MQE;Tdbq
zn$k=9Zh4kiki;Vh&7b<rNR0(dT8+Dx1Z%P$H#_PTQe=J+D{w?fxEj=1?t=!Py{eKx
zbBBBah<gY46{4>G`XMZVK`)}tQ#o&BE>eaxWA`8)vk5RX)Vd|JvmZFymTYEcE6Bop
zRes$``vh~g-?G-XC41SH;Mm2ZB#~n1#kaL+<nW7G{=lumPhu|msS%vwHyi&uz#%a7
zQZaUs*8k+SfVwQUENhvn6P`7BL^~utghlS0BYTgj#pUwyIxXKUFc9Lk?emK>F#1UE
znfXsmHaU}ib4I=cDC9v?MZn1F;iT01xIfHLE|(m|7N=!mXr>1;xGO37=95;zX3vl<
zsIpAE^uleyO)%_L@v{9V<MZN9n_Ja^>Mp;N%&OQ-KXj*9cwhhbJGr(**OS-6u5#`0
zAR8PPW+y=xA8Ip#;87i0aDMGn8Co@Aotd>0lja>~@c0@^*o%srk@H`DG7oVNRpY#T
z`MVK!e$TiEi!`@Rgn=H=p!LOXKoxLo1JbZjWW*sjEDQOtOlcUF+SZT&N*Zf*1<yib
zR{1J1<}B`)s2Bh%ICc$v%+b{-07demv8G16y-^7)YGjo~qMkCL@kL$$pRG^&AT<jQ
z;ch<tP*dFplPHa(Ybto$#OloR@7mA%OF>=V_Mvx!SoOA{@NtCyTiMqe@z(DJNx}N~
zjy5h<X_oHB@zMA6`M7=Gzvs3LP3xHA;^r<vI$a7&NU3w!Xy(0txZfXg?MmmZZxe|a
z%0t9pxKX!1QJ4}iNZeLB?+0=92Omh+{xfUDxM^d!gstDT#>F&83jT`f#bl8sK_wA_
z_=#l44!~4+yteHABS%U$8{#3ae?g&D$;g#t2$3aUBVJ!4P^q6RHrx`GbZ;!xI<gw~
z9Ft%hTd4+e^5>MUt*iNJ$;Egq+$J2DsKGwI#7pDBN@%)|K-MdBP>`z7;s9xnxB)q(
z)I1Lr%0wK#E(!q|p~@yg%?u$NYv_oGJHHCB$<dLpT?52NRQ1T<Ttyp9vHD{vFFTzB
zE~R#7p#i6qmD(%?KOv-*#T`cbufze!B2p?54fB1*>T_d$3A`~b!-pZ6lXl6ta}qH!
zMv1hSkJ36fGb1j~D*x-7^$dW$FT3^@_RI#XT8;&mOa@CJ@#CW|#0j_>@L3MMEX=Q>
ziLP#goWDQ5D;%dTZ>}))GURW#58d1Rxv^+SdQ;kZ^8_S3<J*x54dImqAYL^PH^qx;
zcbs#3l}3>Gn)d$N@1PnFZbiQFtpC0R3f2t9a_1!7ptb=IZt1{RsAy@51Yl)>TUdo2
zaX}iiP^P6)(*U60-wYfQ2|+moNKE1b@V^IT1=f(JsutWshafy434s#4LGZ;&(mhcZ
zM|wdRu?U{HY68N8qZ)iU++J;N6%?GIpU`V*R}xXw<o`6GiC(;c`H867D`}+CpGXSB
zK)Jbe(Wx+MRbKO&c^AhxU(95VQ#jb1vObd2gPVsHG<=i;rP?@<I;huD=L?&nevKXt
zKl;ICZ-;o81nOn&b7pTd`vC5uM=mU)u|(V9dBb`ePa+C=c&EGmX5b`BLqdWlg8hlu
zw_ePYo$v#!In{?h#47sT>^w_szH<mA&nqLI|3m!cNruK1rBWtGVUwkZa%rR4!j6Ag
z;DeR#t)BBZvc*nF**@gRRxIeX_v7P^_%i3Kykty)pgPi~EgN@*p@6|=y5c!rDVkTv
z>*XW*={cA_%HY%;mcNiP&9h&;u(|Xzb6EOitfnn&*i@^RC#>RYgF6_1_!yVKpq#A~
zkd-IJ@nc764hzR9V_N8Rk_d}o<}wrle(`5?^e;=BIox{D#=(Zb3$6VJo*lnZ7sKff
zgfzJEb-$yAzz$cY1Xz;{J08V#8E6_6GW--b7rG#sm@*!$L#9g;w%&gshdELO8nn&0
zo;C&*r8;qip7;C~W;5(iw4>*~YXt|V(^78LnY)?a7X!|UpGe*krG(RP#qKiJ+y2@b
zjNqLA5x!kLk>KL+pE|?z-xq6J8aLBgn&fEwgx^ZN8&DZJzA-~{?U%?)ok%a6_ulLR
zP&uqMQuRCxXI~q%wY8L(#jhnSk*EYEy9f^wIK2ycG<w(D{`m>W@9}^C!h?zBl>{FB
z7Pe-kIQ(l$rz`GK748B)FS@|KFB7`Le#ju%+%7pBFRfonJN3Lz$7RvY{Ljnzm4pep
zq~H8aBBK}GkSGDiuXqS`m2T9{X2^pgbnX6ffZ{Y@UKnZHWO-kC*Y`QnCri}rJq>|q
zP2W;=_m^ZVNt0184H_CoXO?qmdfEmRpgMQqVhpf4xscSn!eOJkJjr8?Y%Tl}o8c;Z
zYuV-RwpQKL4tCuu-4x_Qyc8dZh_}pPR(vP(<DivpX3%`o@$g#gi&_l4gDp_WJR%XT
zG7S-JCbH!}L$R_Ov(^sZsi|YobVHh{y6u|DhtCTC4kVmc%kXAfvRSLOlGsXpFOODc
zF}4<kFv)}eYs=Y9oCPy{w8X13?Ju_8`i~}wBk>eevzpRN_)w8HKuDOSQZWTP;xV*&
z5~vD=_TVtyKPW=PLATc2=r?7Cf!O}SnCHOo#=&Z#PGR<MutM@s`ATyqe$$@osm3SO
ze-T^&xvC_}F|1UKqY8PgHRmzMe|=h+)2(w~YR4$nly;3p2%84R(D$}xZ)+TDVlGjN
z5ba0z>j%u78%q!2>!Z+{^!MY%<+-hoRjq&C#>TAvhf+dOuO&S-0tSKisbKPs?GGo~
z=AF@<-^d3hPiL21n;F9ncDWxn^ETh!*TGCbB3J_Hhz7w|aD;rtULxuOIgn1$|L+*_
z0Ud@5CAM^Nk(9%?BKm#JUPE`S&Mafm2Pe_5_ohP$=IChG;wF~l%^liZ`%1u%(eU@A
ztizvOB_RwCEUqukO3h14>#gI2s>Ot$&*Z(+eXo-uGGm_5LAMm5>xu2`(BOESL0_fd
zWyj;RJmm4aL-^sE>;6ZCn{mUBxD1h4R6WAeQ)M2{)1d+o{eDfzl{k+*&S7E)S*Q<`
zPk*?2nU1nCuHa89?^+7y`0NDKZY8rUtwo6O#vvbdaZ|@V!&XHBOI?gN=%2TLsDC@0
ze4lYMD@|nb{}XVZbD(4B>X*fL9LY2aiqf-k8p$0<9_*McrsPOPXq_az=tpX^BSxLL
z5|Vb_Z*chSBM6SFY#Qp^cfN<12=kovE}LZu2fXo_$@K7p1gWIZoSmbZgzuw}{;nx-
zhZ5;<I!r+Umfg&B;K9imp`6BF{en|u*P$8i@4z$8U`bo28VH6H@k+h<Hy;-#z><ND
z`*w8LC;biqVIuo9)-)q~dcamPe%^jBINMyA#^3g}9Y9a-U-M-$#S*cm9_2-#KXkm{
z&8Yj=un(>C4g^YAz_#7-%GpwIBzS)pvsH^*ttabOaPZ;e{~{KD#cev90}A3E$j1#F
zGXFO~gh+>gVL!Uk({sm<N0{al0Q~sjyEkTafc4V%@lJAET3&onnQ-{yWy0u5U$?h<
z#CByruK2&G=&F(d2I-JnFaO<4`&ZPUiChmhBKPy8?e3b^kP%*`x0spd(cXRyVaa}@
znNHyKpBy&RF<`&Hxaxw3=yctLRwr!B(hqSsYH~qQ+MpupNSdseVwpbwRIeb$4jDxu
z9<=}hgLb2;B!9q}zYS6KCI)0bx<N`fhR_%!@zYvs*gI>GOdzsx=Td}1(i^?qJt5B5
zwDCxfO(yoIZWJhp2Qu`WZg^Rew*{MQg(3#ee?opA{HG{_h{=T7l0K0mQx*PzwVESV
zcg{M`p+f0VL*^6|+l#|WfHkl4(o&fiDOv2i)YFu6PLZw7NJA|C{m;hmeGhrb{jaf`
ziRWT<IT#x5D$acDhleN$f*}Yds4}%PhViB*<A-@;0RS8RR|rbf30de%b}OR-4ji`7
z5jN`nL_9<^gl%J<_ra?Btk2=it2L}qPuu3I00-UWANkd@iQPp8t<o8z&k9yJ49wW3
zjKzV%^tE)ofM+`(my=`*hvK_;=TrBsCE)Xm15W?iK$qZt12-}<m?r#6)AwZY{pA@4
zyj$Pn_)m8cY&XDJx5ZUrW3tbG&nqKVASFV@QSQHf%6fC&`XeAQ4T;wzB1hoog)hfz
z%QwD^<yu9B8_?%rU>9%C{dCuL((xeLd$`?~@44^cI}a4SALHpUa?thT=M7S!>b+El
zcN8s>#u!$W95atlP?YW4f6~$=G<iPE{nE^8jwt-dw3rcp74Hp|6L8$kr-Cdc%obC%
zIPq#*tH`dttM6dLo8UkH9#xTsZ?I*dWfO`VA*-f%T5fo;p`y^YQPr@bWSFel;Hu{w
zdcF&N7J{!CZBF9c_n>c^#uOdLP0Sw-hj%6)o!$BC0DFCR_~h2U!K(Ie<-_9)8PFg3
zrody#`JK&kT4(DPTJ@?;W%AF?sd*fpuEqztJuJUFPdZNXa2|XHc*;ry-l)U!_e^*5
z#~kFXGiX{g_jaS^WVy8xYDaS#8k(q5bz5%2zFnf&OR@RtLCmU}7lSANjCrv}h4t;g
z$*6*Tli~ZD-Ts51wn{Wuj$?#>Jv~*$)CRHP#YG)oI^pTJ-xXkj_KyTx&yROs;hSXu
zg1}#cuNTIUbAIbbpD*M4)nOZUA6<afVV=Kb^Yc-%lE&rjcli#@u%!=4?1?+0=e4OV
z1*6vouery-7mk2<h2zP2MFf(O4X+NS|8$JDcY~s>VZfRAVn-^mCaj+5`?YD0_1od5
z{y@tY`=ND5VdPgS(h&N}cV3K7%!|bf63;5yJKeU{tROpG(mc8>*_4Ib8V2t__+BXJ
zInyo_%yI#H=#TsNf^H&vQGZ-MmUd$O<p3YzIS`x=ib$w>?!mG2vPN@ftiw&iRsa_y
zrbctWH@wj|DGdWuZ%<qUJ9w#w$5ute8ImEg$|yt+8Y?q7lI#piO2snJx6Qm^$qbt4
zx@oRB;f&B~<5x;mfy?xFvFb^DaAn^>az|SLu-rj$1PuCZMc9{Q*w0vNB+Lt%S;F>v
zZgOR7$srC&Sc9+!nPxn2dJ#E_!(Z)9Kly1=77?r9lneXhmXgd(9Ory)tU%@4G`raV
z$Q4bFUYv)LR&hZTj@Ck_7g4ihFmObKok=X?SH}Vtf8dQa6k1vZOmW90KOuKh)JVg7
zd`6wv)xbT#02%tH|9D!|Q3w&tTf|nUpkF4i%3(V<2XmAqXj)*`y_u2(<rAHAkEQ!f
zwsUv$wf$1c(pD%+(n80}Yop0_swdtb#;2C`{hX3~NtjWzyX*Qr^Ly4!!1?-?LXY@_
zuF_sD8#XDi&p(;TDou%E=BXg4&n%>q$&UrMnXp~IbJKNsvwg+?dZv(Iv`fE{9&{cB
zN!8$`7#F4~eu>1qM|~AxR%$EDj=o~{x=h$fGG~XWbag!;2WCqMR=y3kJ*H}1jU}D@
zUviutE%11H4pf$odz6@Lp}@WuQfa1%t|2Soofa82`j0Hs-&nMm`DLTizn~mu=<Zs*
z+J-x#uK$(J0W0w83y+h*yYS=LAZ`2+PxKJJktmd&Amv4qQf4GhK9ViAc@*0SONH@V
zA>D_0hSsV|1jDpet(KF@M?n8a@aE<T%8k3S#iyKTOVs(x_Uhn~VepIH76MX|FWF9U
zaq?|Jnx|}3cG}~xM3a27!Fq&Oy2hHg7<MKaI_yC73P;$RJA=VYKbWLZ;Ik0%(0;;`
z>hXT;N|2qfb}e)A2W>x$E6$s!9(V?Z{z^nI;Z!wt#KVuTHA?ftPtciS-*)XFi(~WO
z-D7MuCa=-@f0|8Ha`M(+Y#56@H$Wio`A$z<ckSou(?0v+?QpkchLtJ4Aumc3x~!Nq
z;HIv7ANIuZG=FBaMsLL)B7pQ$tf@AfjPOVQShec4-9sN;g=T#}g=L4NjMEAHE%g46
zq*yA?k5AzK%q;`y3&?WpQ2XC)vSd2=i(o-awrkg~_yH5ZWZ=4~a~x_Gmv!BDMfMJz
zC<TnJ^dHlvsgw5!z&<K>sJAqV5YqZXsg+}EL|U|Z&|CU8(CM2W9QQ=rAqC{Xrw%rr
zKe5;eHQ4Hh_unF&-mW2AsN&4{w`sqf8>8U~Y1m{MpR}haVvx_?ocTL&>Pb`CGt1Xn
z)r}7FX8cHc0gHD*#V9s%49x1r9n=@Mw^1(3>|zLFlW?bg76?1m5aZ~B8<#Lx<EY5b
z!VD`)v-r!l#8%6o(*~YD{2gZOUIkJIcZDxwExm=N1s(n${5Du9?di%QCT+DGS^l3~
zSqQLY$`CELC$*zbayUcAWmq28qE}<isKRCVLuVk)m1I9@J?;3!PgA2SQoJbJ1YSl6
z{`;aHHu_Xj(VA#s%em*Dh9DarK@2E(;T!EKk~x%iYZLt;o<VLjJUJxU?8K(>L#Hh$
z_)`mZBk9Zy2Djb*%>z=3X;p`TT1YrHt;ijpo|&)I@IUuyyu3(ho$#f1B5Ktn>K>~?
zD((|a6x}$@fT2|ve)zL(vy3gL(lxK-k5KSfH6>o&$r|S0%U9h^)6y&r-Vaq>7jPW4
z`gQ-u0Bc;~nx8<4J$v-^;~dVMWuyR?W_o|;?#m8%IsIm(*fSlyqqp@s03$pi59W!P
z>zOLrUTZ+4_awSn>)CQN(f5P(W7;?D_Pvngj2KZ_jjk!Ji`>ba{+%Gpmk+|d{N8#f
zr^0ILW+7BLy9JQuAUdCqS{GE?za`H;7?-=WurPK)RdtE9fJJ#UVgjs!-yYPPh}rYV
z3IsHA_MtCGFS$T*`Uwf6ZT1oCO@q3^4vLwQz~OCq(7e3As<rB)YpFa10|4bPl{Tuo
zbe1$-^Xaz3ig7Z;NtFqZUS^|kD?ESQp8%T?kz&JFbe?7gP911~-d~=06bkd#+6NLB
zYKP*7r!>Iio8(Pn+-nH<fLNSui7~fKek4~=cK_%DQCUCPWNC&km@)i$f&}97X?H7{
zE64>n5W(EbAtLqdJcAb0P`<Bh_irgY8zAC`0OKIJ6A*MZARaD1Q(Q5>@u$Mi!AkG6
z`)Zs)<+tyB4u?{_KZU;kekn^a!i)e7VSk{{*L9)K|K4s#tG(&rAO7Q&-`?fLmjp}z
zNchJasU4J{$oJ9y<)jhfa2ovmyrxrm=-HoE!uQK{<3Ihfvml5sRUiNL*@L-K*)PAr
zYLi#=m9=#tH$|&H6yHon_b@{kkD?^$%1)2asp9cB<B}F@z)~$;4g-ohak)f>>WT%G
zAi^R>b+;0|vzQ4%9IvMN^}&j-B47ZT=5iLb_nrawk^~u}s-Slk$QJ5ZOoTzR#ya1=
zhV*N!%)o~V(>)&r5=o)DRB{+VC9Yc}LNrQn2cdF0BA{*Lq=-Ds9yA91u8^cg-Ap$m
zL9i1#ow3k}`?Aay55~BmMZ3IVYwZ?jO@70~<jkHIK9ZFgtD!IjEK~Kq{4Gy63VzMu
zjF6Gv!((aG#N0x_x5;pX!X`DKdll{jC)@Z?gl1eMDWl9m-ES7h1fT{B+dm&P8`)_v
zQ&TJ88kRN%#~s_zmZEVRsUf9pPkOl;nGLPZ^aO9`@fx4i4@$_O>E_9>kqK@6i0mc(
zL6PHXdi`;g%Wk@GF~_6ZA+Hndy71zGHZ^{$l$?L{oNkEB)b+S`kwhIquv8bnq5c_S
z)`5I_fRl(k@VV<OdQUvs{F?JEp+j}LX0QJ%<n#4;B=&#lNGZ}G3S3AW>mx!Jm0)7J
zD=^ZleK1Kc%5TQtHdyp2UBkUXfMV}(>@avEv=#?BoQl8Zv&Pj|g;t?%_XfU*?CB{O
z|KsH(QANOy(>^W>(hY~63>@sh+xeE${}<dAiij^s>2^7i3lgr=8CuI(Qd-qlMatlH
zDbWu;sK7@9<Ix1Uh2f?g2J@7HTtO>XUV2$}Wl`m1(jdd70@Wle&_^pMq1qz@94HYY
zY>P{sUMm)<MPPb?7#unnz*S@G$V`k8RtIfSkuLJXQdP3_#QZEAy%j}<Nj#$!lxE^e
zv<7J`j3kq+i>L}&8Fd7`r)`Gl5XaDKs|{>oXNg$;?)(!++S=73?`G|i`tz&g51e5D
z_%LJJLnTxw%-T;hRK#A-xCs!ZyE{8?e_7_Fb7L0SbQDV6kw2BcwnoQmGH=!L&JxPn
z*1Xtt2(gco1NQu_m+sD+dwPg~Jw3F(mWrQqhl0tFoFCYTpRezF0mMEskV-9<I!*Lx
zF#qc)Yl{-nQvQ?x8k5w1df%eVNf!QiBK?dHyeV-%_OX&WSh=WGBo9x}cOYhm+%bu8
zG(ww01ErpHEGXQO5<zhg{ee#wjk#m>>}dd(>s!|Ml<5XAKkxaJw>Pxa!TTo0T#+sX
znfDS&ue%J+%UVLG5ANtR5mM-C$OLQuL~S~whhz|UOAcoxbQM~Tb=9bN;aAtDyAM!R
zU4ASQ96Z1ly}hICC0lwlZD|bp_vv)WUv%18`z>(xTNgf+ta@8z4Rck;XxzUQO?$bt
zU7kP5@`C2nc1u}Kw=TWY>2!8W<~8O?*D+=^F8XF8o_pkv%o9n~hBdNSKcR@c`-k2F
zUb`XqLrCF50B{Lb$taJ1eU<(`_WC}4o-KuNiN;DwKf)8BWsacs(#=TuHzw{54TzJC
zqyWKvj6@7R<B3BZ2E<~o-N9JFrDuO3UrM$mL6X`+EJ!|mQ<gKnqx<>ia)wt?f$pr#
z$Ci5Wh->=vD^_;hl?t9R8^Dfcw$Yp&-0NVwMg>|~-|o1NaB+Xk5f;%zEcDSu$xoKV
zQ#$B5Br=0_nf_r{92n)n0mr&Ag8(^i8D_P0q>V6p_5bnol|gNF(b~bS#oZkWE$(hb
zibHWP6oR`GtY~p4795HdFJ2sq7KcD_C=jePK(L@+-h1b}vuDnqlgW>L=B)jkwI1=Q
zyU+(-c@}>H>WPbHlnJ_z{O$RvX)SWDv016iC3R3MIK;6l-QFuWA@e-14$1Bf^+z4(
zJVP~WU5_!J5t5Th)&wNOr#CG(E|yOIG=?(U3uXJfo*Czwkv*VoA*wWN-wYw?Yj0t+
zq4M?>vxJ#!3Ff+pI%{3<w?mV^cZbr&>}4+c90uQx{M&#yYwkqs+9CACbi^7Nv9yGm
z=P_~{KNGaA``lYhpkS5H6-eR-g;sL}x99uyf0M}B3L5f-K(eg{x#mZVQE+wo71t+Q
zQlt<*SJ+(t=o)&dD|DZCW}c8!S2w-}LeOz@QG=z^9vXjES3G83hMdVist?W2LlnKF
z1RzDu-rW(J4{v{Y*Gqtooc)dbyJ7y!8J(wjuaEnl_N|6RZv0DKRZd#JwfpiA?V1)o
zYY}U`1DgtedMO`ukNd+3^@XT&j!Cqw(X&ZMih-Zej&znIb#s~Hu@j~N1I-Sf47FHh
z&C5ez4nKr%Zj(ia`u}?Y5C_;zMU{RZM_~a^&ju(p%%f{0MgMPx!~*tsosY{1Tm!^W
z*Pwj!$0IrHlR{fs!(tmYpHH|bT4#m_y$MU&WT_tB<^k&GvCbVSPKTe|DR}RM1}+)_
zO0ixTJenOYI#Y3J<zI;=OD{8tLaOS1tJf3ttvT!wlx)zWHxu<onqsdl&V}GnD}xjk
zTPQ6OG?l96Zxjic2^L*^+FA^{w%3C4Hyi~<h~_#dJuE2?M}%HJ>e<OG2X3euMW$}#
zw7c4SDm;;zN?;WR{fVzKmpz>*7oC6K>^ll-^54Xu5TdJ7xDCC3>Ao>)vbL35Qt^Cm
z_3U~z%Frp;f)_Y9HZd-$wpAWKwsMJd;4&bPy}M-+#t=0R7RGt1&s#S7EL4s-<(!E{
zni}!vPvk|174*6Pm=3}jd%hUJ8WK8YxwYLpKWuIy)A1VK&;DoITO)FIURhM#8?MZv
z&!Kzl7CgFEldu!rXHX`A8~5Vn^As$!NPKFaZQOUlZ(ocsf#cWZ&w%r2ItbMo{%ct6
z`s-`=K#IN;&5!3k$QUP<d(VFV%MV+*h>b=o7C8QGtyOD3OA3TSoq+Kx1w_OPyQYkt
z;HNo8iMwFMo7{_c#+x%{(Utb58sc<!lX52s8nVRgPTCjh)8tn;uY5v0tVd~rYkbA5
zVS%8=<Gn&+4{SnqzYJhHVK(>!#GL<N5NRRAK83=ai;I<GQ|;+-1$xonuSu-X*Gv1R
z_BWT@q1fj1quqr(j43*l1aN$BgJ8t<%V68s3l{{kEQV-!!&?1rqA=~2x^cI44`Qu(
z@T$o)RS#l9M<htK{xY>LRJ!`4^Zsl)ik(+n!+D(sp=H|i_;IIqxIhF|$G%A(PdvcA
ziYx@Yd<cD*!tP!)1SaqI_Hok<U#&FudQ;1J%<lv%TKOk>5_-Q{OGW40c&JB54L);j
zN=61`jFJQ7E6%Ax>L~R^ck!{j`x8Tq)|#+nlTmJAuo?bqRDstGINAH7)uQ1IiMll*
zrnq84eYnpABdY`rtS3Hs9$#l|yczq<4XFcY55-Tzdv5UkWU@{E?%_k023AQZZ|f;J
z^dQD8cO`gcNA(B(Kf7$7B&!0=;WOgoEVhVg=Uq{m2|+o4K%coc&HnxILHJB<h<5GM
zTow_JD3(^#NzvBJ*gpmRwvaxM;w?wo_E@)F<mt-+Zf&n70_gCUg>`4%@dEHTg{q~2
ztIWTfjao*R*WB3fmGAb~WM^1`JMAcLWX16_pECkJd$D=VE>_aL44<oGZK{oIj%*^z
z`$~?!A~IpDRbsJA2a>*A<|{R@NF5kR_LU|=`Q<0MjN&yI#)_lYMnfHs8Ng3%nDZ7w
zSaO9H9~uC{kvWT>eR#KB+E}X0ifK<ZprI~8v+G<|f92X2qu((fj7hX~<G&Oy8@*MK
z5z7383_=9tEc)rN=DMXpHrfx-y)#6ec-h5Y#Wg;m_b~_)S+v-E{l?c&+UMD?W5)eZ
ztP!Riyy!{W)tJ`Kur_a}?A>4~#A18K-j8FNavt+)nJ`?9khOu`l<!qpp>IKg3@tnM
zi)OA7j>EHc<zyI!*wSj8$J@^%s1n^IhN5)G;-D&ctY8<5<FDD1Q1Fo|c{yzYy9Jwx
zwV=xK(+#S`(vz@MI@OnxKHeKc>9gPd9W~w!;mMFd3w+^kIEKAlV8h-F*|%ZuprhZ)
z*tO`;fW3`e^Xn?7F$&(VEtqcoc18JP2x0smAE|T4D^0GJqrTM-z3NAamqvf~GrXC?
z(`z6x%?^t!$Rb~>@LUDkNa7iiZab>TCaDIE8vC$n(^Z;$EF##6WB(vTPSh+VN%FF!
z#CsBZZz25J^giGH3z*=bxR}8dzl909n^RmM0X0F5!GVGugTO@kma>E_@2We~e(+1u
zuthGhifhwWGWq%wq;R~!2C3L__s58hRvAi^n1vQ+KhOcN2NWM{Xq|-6c`_pD2WU*`
zrur^-2wym1R<x;(29$k75F~udQCV8o7##A1(%7wCL}^{~@wO+&GEtaf;@go_e*`+|
zYEmYK_b(=Jnc<S*lm5@DwZyMI;y|GL?L^~ZWlf-DKt6&;lFht1hxXn%ahvjIeukCe
zqMM+^fB}zW3VA#HQBV_a=E;^x#`>(Gd^$QYiDs!35e4{bArc@xX`3SUXq5U(HBpn{
zE-b!E1z?+{>ii-<OD9pIR3RtMJwLq=-J2ZJk7v3?s_dFjjCiFepATU%=!TzW=VwRQ
zX1~ZqZ9y=BdhCSpoCbfL8|tr%N&4`Gr6MUl$IQy&xF;&Hd89K$V&5cLmm4Z>Gg6)q
zch71qS&P}TSUN62Vk?H$sne5Tk+hP(6;*BE-eNqX0;(UBM|UXexT`bMmYt*|fLk`I
zk;EsI70b4WAZ)EmxW_|!i6116)HIm(cT%oJ$za6)uGX1|w~kWyQ1sB(CbQAaj>sG3
zUe?XjI907^BrJeWa>Xx8dMYbLae8pu3sq*5lzg#A#6_d96|B%PETX;lc9F(&-Gd?8
z{_d>(CxZ?O-i#<JGb2yNbw)Y=*qBN($?hDCH(;A~zt^)oR{tI3+k@;`D@{LP-y~SS
z!Ss58d*~!#3ZT=^*n5HUl~UX}W;YEj$`tJHOUT~gYXPo&X=0QV5ze$fs5-?x>96bd
z?kdWAfvG3gI(i%-RU{zC0VuQWv=u_H?<JcSyd-LixMu!DhZnuF2kbF*`1vh))+DOa
zQ1&!g7+|&3R;$R(9JZK#>z;^MiM8gJY{v7uq#2)!UO#@=WGku?jcWV3&H|1+>Tny|
zq<>JreN%u2FX+$SoZo>gn%C>>^m}=3m(xk&kPA|5cy+G6Ijz;iKO<KzS~V^)#nA81
zp>Ez5uA>m+r1)iaTm1JGM+ygi$7)aNw&T-_sQ`^1m05$6o8X%kfQw`t?tVIgHPSoh
z7C?<o?L;iJ5s#!5Fczm45TB&EH12SeGz<}H9Y`yE(3_<&H+HkT0Yc_jHEfPZ$FY)B
zg<BkiggC(VxoPIIgls`?s=i9dyxm!lmvj<Cr&I3w#=c-|+bE^<Mo>dwNJ8Y$<b*j6
z*Hu-qro^K`d`XNN^I$!tXcW76R1RMO`+#C6=&0xjb&R?~2S6ARrCZ5*@YV5))H#Zb
z(>(w;up;n_)Fe30=HZPHXkvhRWc=9XOD*}a^90+`FZs0XW@`m%bh~l?`^)@Ql7Z;e
zWR2v3-Xk7>mC}e)i(_uRrQA4I(7>K218j}u>JuEPDc<eZLNye^`j{iB-}@#8&4T8R
zT)*~<xV%{7>L$*p;9<ZHBKz0E13JPY({V83Umyw&IJcCfeZCg8iOCxx#@W5(@r|4>
zz+5i?KMV!kNtg<Zas=Ntk&mr8(@mbh{b`GRtXr>Z=^DGuLS1R?wk*+RQ}jf%9;L<T
zd*YZ3XRsz_x6T@R@yq6@+bsr8C5#Uyrn-99NBE$gSpP9Nd7z2I;dbPA+BpZaX|^@)
z0`~~b%6Q&=i`lWbTbT+DUw(E;l&uv1wh%!e#v1YMcDEvFzLNOr>?PO80FLnM>EAMt
zQl*>^$*QR7Ki#lUQ;yc$Bk^0QlLy*-(zC-ZCQ?S}KG}g}lh^7FT$C8~4iCMH2?$Mr
z2(j5%au|?MHg`82Q|Qy1_ekXKu_)e%uW4(dEOG3@OGnR*5B%=Us8^*~DiBJGJuHab
zirbDm-Sh!w617Zdji~RhC9KQ+VlVx!ee74Wop|(SZ5=jB2NQZMg3#4rVw;6QceWkI
zDEe-iI8Th8?$u8Y54+w}a=A<!za&4h(<4}aZz%H|u`4?8!}|1s0KaHBQO3IipcPY?
zjEVNQAtdEKCN_Wfuh?$D@!Afiwq9#hv%86q;@O&4Anm_{Kk+Ho!(s?-m=wia)2&-k
zZeQWeJVI@hxbZCgHeT!ZPqCF5dlOnNA5w8^{49lEqY=?U?Z1}q%+;Twl2x4WxxA#N
z8z!qgtXsXTr7KQ9QaQC+76{arjiAtVU$q_%v@0;gb$^#Vnc=^`h9JNO{!v>&+ziJp
zwiV?|imob2e9iS!C$daON^Vb|aUR$v1Nogu-{iL?buY%}W`<XkH4)dQN+rqFiF_^C
z3&-8xpjj@z%U@gxp66>z#laQM76>vZSV!4eh<IZ>eaduOrE32B!-W(I_InO4Lp+2q
zD*tRiI3aYdy%(!a#jC?mj6m+~mP9D`BAyxX**j}CsP^M5>NwU&ot1J<iKn?`)GO*;
z8iPFMxm5Ts#<%^pK05%+d~&`-8`JV7H7xI`y1f#qG5hzZ%z?Wdkp%fdG?Ip<J7Y56
z#HnpH;|VjW4R{Y!MulE35DYgB>(J`{U$y?PX502hR0yxy&feoJmHTx@8!}*_)7c-;
z)37gyC^Rm~H~pMuzX)o~K#A?e)(6CpM&MRD9Uq~kvGBgV;KOiSo%d%RTw(s0m!a_T
z?O*+{bWZ8(a_cv#R2(aXpd32T<j-L2W0K&Q8c*~K>bX+oSoUwLvz*i~`d?moU&v>W
zvZJ<alWr|7Pykqk7@U-5qbR8opa3#E1r^B;yI^x1LiD%)B(u(*txv=>V#KCtX`9pP
zg@*OQ{kzqW%J%rPVetb}>%_p{x@Sl_Wk(tV@l*`Cl4YP}VJ!P60Viox?52WJP+zcR
zap{V|yz<VVg6vZ3n_t*|tlrVTv!2A12Z=qR8e+pge_wHD?Oh0ErA!3xYwyVY-|C;+
zlbZ{med|Wy&|fz8cj&JZElPJ$m6n^}D{JjCX};Rp$i|xUbf^R>I)?od7TXl&c4S!G
z$j;aLGfA~hH`=x&;__LSymTA?a>75=wp+Lx<I53=f3Rl@#IhW=Fn(QE73kcuvvlr>
zD?+Jz-l<$**U&`D$`H7mMCC2j;vLb+(`hCR%53|`04<5Gjcn*%DohSzu26(%lz1Li
z@;WX#M0!_9&8@#XO*X~LBah43bAL~U3HJ+Px603DdPly$`2Nt3sR$jbt&Tm8B;LP;
z=*#bUV4jRb#;Yue&wQsJ6rd-mD!Dm&Kug+B+e({b+VX{@4;8kpn)Kd%8D1M~-wo_{
zKD(sc#J%PoY?%ilH>Up^wEx{(1;IBz<H<`}fl`K7b(Wa^En>xlrxTrEf@nKU5NRX)
zxXTRm0-dhVe=FsQ5x2K~+L`OC`zP+dD7dK|5PD#iT__%#?}^{;?;N<%?^rOJBAm-w
z#fKx>6YD1j?c7SM^u3#dfmK}um^}JfYnATzZyI(3!!)4=r;}s$T6b9{|Jb?IR7#KZ
zsF-9k6`#9F*5dS9vmqg_8wWdHxNQfFmQ5J{x&|sYOSVd!f`zRXO4AtO@zc&JHan6B
z#faNBOQnsfQ=YR-G4zapkhvjvtepuN{*4Qe%op5Zi-_JG-_1<95dI{;mc8e|-VkrC
zI9me!35zx?L>u*1%E~;jJlawE6F%rt(mD{4iQ5-a8iC*7Hy2Qw6x^G5!RaBlQ0nHT
z!CDk!XZW@DTdK2ylo2fRlZAm1qamli6&T=hm|7U+b*;jVz2W2~!_D|rba0kSEQMwJ
zM15}QjG&r@G4?uYgmlbT6*dI315)hxh+Z0Id}6r`e(@fer#ox&+iD=1YFrZ^SctC(
z#pM!s`$s<NR3uq*cYXG=cZ0pYlk9P#uM=wXn>*+7P`v(9E!4NE0DDUyHD#W6vL}rG
zTTr_Tb|H4?AnS`_*0@~e<{#c)c<+Y7Nl~T71!eV>FH4y$w2(8q{B!;^YgWKlJd{r$
z$l|I;J&9(llrWa+)VYF_(<S+Nh*+$oi0Z73(raNR+i+_6GSd_ds>I*zj8k>GT|pCG
z>h`)T<Ves=tB7Cwg)jI;<@+Bzv|_ASP}WRE@;4ZUk!UK^TcK<tKBcWxy=1sU$d)*6
z*Y*&Zm(f1iPv^fhfoe>~oy#8I?|dg2AC8|HkHqLBg$ElW{jqyVm$X?W^YO%7+UY7U
z?Cdkc6S=RzOr77i=m$f3jxl^xBk@nJ(i8b;Ay|3}{ieZT`8O$n5|@vSyB@3u#$v_6
z__HM|nfM2F67^-TQqg$}0B<m!e|VRLO$}bcJRY7LPmXNp6Y=_4;tokQgrkC|S{LVQ
zRZhNVpW~3{>vM#1xIqA&zHZ<soE&uBJgp7mBtWl=HZh@C+=v@0f8ln^mGJXtSQ88<
zZV;7zO;v{WF1B%|X_m(%<~;T~{`g|h^Ctb~tIq$`8sw|hv?GT?KN<sT?s2ml7GK!;
zeaCiAQ8LAL%{IGp*%}31exEsOPFSeMw5077Za++K9WhC54VY^^eyd0{J%ioSbmwNr
z9c?8O84rc{QEl_f$6DK=4w!{!JPHgtZot8`O$C@Qd{C50zhE$cMIG-lX}`I?&Xv&h
zSz713Au>ZhY1_Q?7Wt2N9r;zHUkxY$pp(dNr(9KARFbR%5FG3zpDIW3MEPe>`k@X+
zVJsw8#(e5p`MsZD{R^mfp{CEt;**rlt0W^9I$EoiDHOk?0xlf6MEVmHv7#FJYlkO`
z$zx4#w6PAD(7+FmU2y7B!9)>7QW=Y;lX{CJVTbSQ0}TWY=D#>!OU%~ZJK0{g=ay>S
zc2A1v_5Sad*Z*E=HStNeJ%u_!alQDXOYYU-ECf3qcN4zJ%XP*51~-Hrim4yVX-kY2
z=%TmAtmXvtlMdyQ&L*A-f2i;Y&sUr40hTUDz3aQx#}hdb5&>F*CHR21c$TZU>nm12
ze}tmu>+qD?GxhubxpfUD?imqr2CTn(GHArFJX0q3l>U`uS<VueV<j#sL>GxKn16CF
zA8Qco@1Ruw`NvS%Tp=}n7d3^D&WJ}(0nMFlVI*V31nUEF6s6v8PY16DC-KBVNLPL3
zw<J9V5<bDY8qc^UfZpn=hnss&_0d9FD`}e@_T~$rONVNR5JZ9IZI@v9P(2SwdUjK+
z%|&MTVZV}<_YR;oFeB!sYYOeQH%DdpZCd3~{C{hk(!*L82@(#H$*4?WSqp|=H-=o_
zM^rLx(YZnNo;tPxjB@gQOov>*okZBev6GZ{pAJ=a`Nt-D*ky3XocLDBey$j8YuQVW
z7xk0*eZjXWgsi5uFh$}IC<E1vyNzy_-jCn&YSaP5Bgus6H-H3_h4ni0V9cq4%M&^0
zx!q&wD+QMRBxf4jYq~wmcp7mxpb!qSf0|}ag6w)IETqv*n;|TV6vklhcVpVVV!_`;
zd29Ax@;ff43<<VWhc_%K+sLvoDK1ZN>^~CTiI-<|ck_$JF_!mW3GxVWKuP9kr@X}f
zHYWf1=MTVK^r}*u=ijou@BzCjq3#BI%~nd^W}ooyQWoWnrA213Mf8(G-Zk2ofh)>g
zhXz8XZ=@i+VFU?i+S2DgZ8EJ~cHlf6)=h8KJB24An8>9uSG{gw!`gq48A~;*cR@A!
zCV*G}!RN+nah;{7j+h35q}oSr6Z=Lxd{%E3&D^uH{zOi;YW13XRBQ~lkN>2Bq~Euc
zfqaDDg@x0F)Gl(Y6&h0&yN=(>V5-i2eU}uz$JGbI*7#m+*g`|digUqoS-yl(E#ATk
z{WS|yBB>+NC~{Gkvwl2dunBj<9K?!-=>zF52avI?Liq9%$NB0q3@xW$3GaOMis%_q
zLKrfJByt5A<Ov^>GSq+Ch+_$}v^5~~a`bhr^f@!0by>&wu_-Si0Mf}BL4_GEW*(UV
z03Cpk*cYiv^zHQc%(=e3?5~d5Q1j+h`s3e*Q8VQz1IcOie@6wJG`f*)Yy+|6_NY}&
zK|k>4fd4Z9@OUrQc(Y&r)<fd)G6U~WjLWFXM^DVdN34-ENrnF1V+%KYnPs)@Uk5Di
znAd+l1M4oUYkYo_8hKb&D>N!Rh($V^W{QTPZxI<cAEkUGO?|`Bxcwfx)mI~2=$9U)
z5uuSWzSYYC)I$XulhY`00xka1m?0{r-38;?OB;)lqEN|dMSY0a6$WqmVDo<BAG>EY
z@AU6+G|)D~GSsiLuJ%5oxi`M7%auq(Ax!7Gw=(<pU97iDacTae8BhLVpuiyw@i^9L
zHZksH465JXbU2ZYW~t{mQ8_wgdZ)zV&%F?~i{W#c@9hP%oh8xXR^y&hj#-f<8af9a
z192~Ma{%kJu|k*(i;H`yof5n1_F>Bp>ZzOoO}+XzIbQ-?o}1L)*9-;yEq}M~{vTK(
z^4v~ztr;3bw!3t>DE>PL`LQNwZUL|VG78LyI}~)ztWa16by8bE|L$_Qujm!{Utb?X
z3M6YBLgBQLABo9tjynF<a$=FrIwQZ(hSID)W%@RJ8og@H(FNM0hno8qImt5?sUatn
z;qxtHKhm$z>4+!LWxy(IaIRLSjBvDX3QqLsQn%5K939vyPFyPvIizz9)}saw^ihse
zX{6AMcrt3z2R@%WEen=K(GV%x6n5pMmso+Ql$1U1ia(5P^EG&RgySoo;?jTUCAp2_
z{32LP4n4{@<zMy+qb@EJUzshO1K>p_pZBmt{c)}Nf&EBBO*o+~N{oY9>;X;<FE8t1
zX_0@EoLxWU9^;+y$rkeC4N}L<ImvLq@v(3M^m2Z2lT~MI{6VrrTbOQTQFGSjiEy-*
zq}ya$a`u6B)Mtc)*%1R>FPutQ7KzRKKZtLGatbjM>U1ame?Lyh+u8snrQQ6Y!U;Zd
zHMJ|b?SDGR0J<bXk7xZ#+qg}vO-X$Dqo%&9&2>Y98vr`?wyFw6dU@cr1#GT380Cq&
znxCyJV8MdVma(JtAr^i2y-q`SS4@<TR~)hyr|YfGC4csR<9f;2D9B6P3)pQJAAK48
zQ3J!h!5t(a17o(8xB}39m}a#-Ec;7S6qkUkqlX}vre+Q62^ZB`N+^!9>G9hPzR%;t
zF$#O=YEL*No!i|4ZK&REYoO72MdOW-*CAYljdWGJp3cf%<5F~LS`O<W4(K-jaqW=6
zv+U*CiD5dv@>5Fgd)|KR>3u;7XTjUeh<c|lF+b*vU!2m{NyHU?ltsb(jq7uEd8U(U
zX46v4h$0}+ywF#n@07OqUUkKN=mM2kSO)yTWX+jS_ZzS(;fHNF@uRzok;lj23po<5
zlBm(!$hH``#CLt!6qP>GUx>*n??!av)ccx@w!)xyt*4g@6twHSp}t{gsyqQfZwcr%
zXLkmd)n01|FR%_!(+Aw9Jwmr2D<(aN`Gn5p$BoIkvJ_d}m-x0kog+tyb(IE{MQwc}
zUkz&vSjx&Eg*~4jH4S~eBMv4pR$?9>JpOx#X7yd)f)+X*jCgx5oLcj+#_uAYjy34T
z<~K`rjuF}ZjB!5i4+=Y9MwHbB9UoN=ih$U7G6|ad?ki`|UTj(wgiWu+{vC{cdV5f0
z)blh#b58h|O=9+_vVi((F}=HGm&_2fwwts1Ma6TTmiC)L*UjIT+cvr9E(u{{o9`m_
z_vFMx^U%=d9Tb@?29~i^?>IR-xVArgPVJWvy5?TTC^ISS>WQWwHW^eI%FP{#KM=T3
ze%FjXuF_t!Nq-xb4p2&pW#yY2FpUTn(nuDfVCK?SjH_qL$4g)Qp_=R(LVDIO5`V`O
z9P2?pC^e2UwXk#Zs?_L6wSJpNoLoM3yG*76K~B_u7>qd;{jn7Pi5YBlis7OP-IZHa
zF_ZsU<NGpJ8~58|HOY}WG$&YzA?>d>v-ZQDZ=@SkGVc1PpCne$2bmI^cmXHBJ-l5Y
zmIr@UC}5NDG|z`$rPa$rd20fpq)?M)Jl|odwjd-URoEgBPJDm1dUf%@*%BmP%-9pS
zJ90dgZT|Byw?1=b$)PAX^a?*HjYZP`-UKr8lIEM0%$fDW6aM=;Jk5Paj-vM#Av+ik
zJS4?u{^tU)QG~*^cniYJJ#5WI-}cr;`ttW~b9bI}lX72xY<_p3w=Y#fMlJ;}Gr{eH
z1Q~r`O83)9{qf0GZ{@{pS~~0@;_2zws4o=Q=lu2Wd6qh_O`69I3I%gY%d=je$94d;
z{~Id(DK=DF=HC2DR_6SOP^o^q9PYr-!*ct4>CO8-44zU%FtFpgY+<o+)w}lL%8QLR
znV@wiZCJdp9lpJGep~~2zMniu9k2M-JR{Y7_q%{0ai`~Q^#tP3d$~{eho|ST^w>Y_
zzTKXokp6~E=;ychDD>Qc!UXdogWuS-ZF87>r@=hsWY?a6NI5wh$i4hj<Y3KLMpr+#
zld2xoxy&)HY4MvYaaHacT`9R(U$VbzaXufZ)>_4<5(b&T5+&8xghT=lwa-y$wWLys
z5+9lSa5QRa7v4g><YOay2Y>9k+W1<j#HzJ$u8QCkPc9HC#nyF?y(RaZL`uW3CF~Uz
zA`DBzEG8jM23MUDg0*cU;fR8fVsY6{|BmBTK0?OfrmwrGg8gZLc>Ze7T+Hs<E?|?@
z9Yd^tWS+!moO-mNRLE|0J-Lz!2-k)=+FheXzRmwuMmi=#%A#eUj1g-rOsyAtbCQ%d
zg(Igut!0GLDvBKQj7rQJEMf6rSea8JO;M-P=ni=dyhc!P$s%iK)wINJBIVW%GH2}=
z86slpYb%|Pf+!4_se@sc7}mA-WuVQKX-k<%BPN?6j=-?Hpx{7!pXhMVPT;{mEGhF#
zu=UQnMrP@!y#>g|O$bu+N)T2!@&Z96caSxe_u*Wt7j%y$`y#cZ^P&%NXaRY};adIg
z$KxZ!3Ur}gf?;ypEg7gBn6$q6JUeu6QFgl4X3+OI?T^kWo{I5P@1Exh7fTBrzm{iv
zKoNmsDKbtUY;%3zPZXsp0Tx-13$p?eEWZA`*(EAOfECnukNUeY)S@+X;|Jj_M2a^4
zIfk1@(h~G0ss_q09Ckwozu<0&N?n|rGpVDYj9VbV_L13R*QJs47iM<y6X*=u_dHlu
zM{Q&3pG`{?ZZ~U-N5j1v`-Z~!BCv+D_*%i}(-(h5hQzPA9<Azn9m@~nXU**QG%Ae+
zz#Nf}m2RMym&;>j6z4q*4(Nnn{ka&Z8t#O>GDQ=x?JK}lJY@!=y%R)dPCzv&Z3OE#
z$v;O{s3|dgY%k`!_W_p8-=?qb#0d+t;dik)l|%g6_$-IsGsg=hRGDCW9n6nFTMz8B
zIR;l`f;WL6AgcGxi9$`uZqNnqfIB+CM%DaL=ZElPr291;HH*@{+eKjF?x>Bu@1vxT
zNA@7BtSFyJXEH+s?ru;QBQBaK!0z5P=Qp-ce+5bubWBpBRbEcF8eivn+j1ca{nJ?d
z_Sk%j0ss7Nl#*>y=VvbN;aI+L&l>f%kp3Gvryo{#I1S*b&)x3kj2Bcv5(+<VGHSLy
zDJph4K|W=<Kiv-o$(UK3IfZWGrh?d!+OdDQrwKLxz?Zjf<2T=syF`>9RcB{WG?lC^
z_=*C}(q4XIx0l7OXTKsq1?LOs>E4{d;f2gS1@p|knQp%{%>wu6Sp|HiF7&dwXa-DG
zV<Ai8alI$lSB-q%2Z0sJfcWS!q#Z|(!`3v20w#SwvlkL{cH))isL?kf##7}ep(Q<<
zJuYF!C;IPTWhZA}A1(~!ad^%%_Q>FM|GU|9c#~UUFtz6!K=-Vw`9)hAvG=WA)P!2<
z{8ErVrIxNKx!(o1_r{BNFPJ0QyN)jQ>)*3_GdKLF0D{_~<3Wd+-?x-A9n~EKn^$dl
z`X-xm(=1zIh{Ch4Qj1Zm&MX9bZ~qO!EwN@kUfc!KwYjXmcU|pZy4{KV*n8kD-adhK
zv2sU3lEHuT9k?=Casg`nA2_59I_Iu^f0inVe!7i?*m{CyKspGN#$t!MbOo)#gBGQ5
zWEMf;-3pxzVfT_QR(Y;-vAY-th5CrIvl9Msf_9P;)h>5~E!o?n8KMBFr9?9s!TCN?
z{jiB*ML%V5b5{HuR)l!1HXZa$_Qy#w$THyh9`+<d&urAUe{r03(tyYG3;CE=`wS2G
z+ts&xY-r&&uuoStNG<JLx{#oM{1#kti9!X)MQGZiO<KW*f4}+C#<YJY0Z7bbrT^-Y
z{}M&blqiIl@VI$A^NB=!`7I_7YEpm^{1=?8zWE2pS7(W@6Zh|(oLw__!Sp@RZ;NVz
zl3+|Y+lXMnJq4ksTa>baaZBpq?D=XtqZ}>&C&FT?FI3~F?~~%?7)y=BvAI@@kOpHa
zN?lIBqkm<H^w)nSmj~0bF)>~ZrsdnknUJUQLP^r*)hVaYDs&{;@N#Y$;I>gr=y!Pw
ztePJC`7TH&Qp67J14|KxQ6wIM%S{(}))Jz(P?n;SeEE^rUf@RD+SqC!TekyUH8NKt
z5`Wa@IeLj4>w)fJ6HAtGFq=hm2bv9g!mLnk|EZ8Z8WIzhKz4plf7{vxp@Tirg($a;
z_&*1P{afv+yt_Q;qa3`Ol5z(*+`2rnf#cjkFsyZFS4h&br3F$MXaHR9nYn!Blwfe(
z0J<+ceyQP#j>_<Ss4}bibt?|_>}GayJ=xIgW%T0%G@7Eu9rTO^VOT#e0mxmuuZ2CG
z%U(MBio)<)8L(I$g<iOQU?h%S(C6cw-+Lz^2o=`vIlK7P+|QyV^ms1nJ04B(9@dRe
zmeaNB7F|pRj@&lihZ;QYQQX9&as%!cQXVb8j6<F;NQx2>&pQ_WF|&m!CmEc;RMNYx
z8?$ypWEf(KE$+;PzIQ_d=%jGFW?zIkvG2eFrj+0t)Lm&2R|W5V8h-2z4r@#Ca6=61
z?CSEHwIqepr`ylp;eNb3XxK`Mhi5<-LT-}*goxX|Gy2CXmJ~xXZSnHELplcR*=_NL
z)1t74l+xdQA-iS}LygLYj=y@7uKM>Ef<l_>&d+dnnF08C)b|32=CY-}#jxbMS0n}O
zlf1@o4;Fp!MKw@c&}B4xE$Cup>wN=r<+{&CtEYRwR(Bgb>A2_#!426Q7^vE1(Cpsy
z8!`**h1BTUm1nfMB!~7`T|FY(1&!olvwqO%?1uxzmn|aT`~`c=YyW=|?>C~S+h4%q
zhUj`6hh?*=n2Y4#?=r6Ij4b>-^Eb~|r;m{dElDt3G8UGhG&{`*7=cU(M}_9vpw%CD
znJWTFN?547lyOUW&tW7yuu4h<54X)-vg~dQ%k2PTmZV0x=yAfRW)vCu9u-u8c<;!)
z=s^56-)~~P>7J2*!>F5_*{Bmd`VZ~;B7ygFX|VC>?NwRIus;K*;t<lD?L{>8^7%?7
z??Nz&OxrHr$K^t4$Z$TqgLhKbP_g%J%6%9i7zjijoP5&5Xf?BGp4O-~EK&A$I***_
zv{Z~8ux~$>2Lz^)7fn<RKR~~WVC!c2t!vwO`nqXBwVkw`YX7Vyb4SLkB*93N&zctX
ztK5b(CU)cb#y>x0*!lp82|Il4M9ph$d?0ploTMd>(BD_ihi?T-b^hhh=~I|u3?MZQ
zw%xQZ>ab8F$sle~IOM2V+lt7_j-OR+N~k)Rjrxp#a`<b<@HEPHVwdcU(t>+}agnmu
z=zAjmO~%rJXw9cysaYxRw<q(+aE>aZCt)T2bD_T$zwEK`$f5Ppf%YWCTaAx2+{d~w
zQ!ZyGeBV=STYr<^M-b{4@Vfn>-Ov#ga4eN9aymHkfelD<eUCR+PSNwqFkHHl@Wgx0
z;jn=LBot+#>5h(u`(1YQ$lRN@V4kelS9N;kNyX#4$YB0V7>u_lFA~3K1w7V^g6(N@
zkhk?1dbMlyAaK{QU$>0pK3v`1x>={e9?n!CCJu!Wemj>)BexhV`cSb=Zrl)eiRW&@
zC5vd8PY12+HqzMn|G3Sc;{4O03I}dm(~%f>IO};u50eMqtfVh>ZlB+@d_T>9e6;XI
zeC+nWHk?7V7HYo8S;f_y`<e&PEDSrnDmd?j=YW0hf7$gA^xjA_X>1Ki^^SeW(k10p
zRQwY>PZkh(P+B0h(4cf0(LJ9VHnY8X910Db!#EJ`X`p9}CBUQpg4j)2kEdFMzLUXW
zqz}70I{vKA`?8tyk#)}-YX&#^Nzj9M`QV)7`HZSXXeNs9)b~6P@sPl~&jf4t-R#we
zu&Q2IxJJTO^BS~KyV!qCYljB)VaI8GFi+Mkc^3Y+g+(94np}|Rm^E4ya+k|zkp~c~
zG9g}%#z3`w>Y53SxVcn3*IH^&{uTEZju5-*McoFS?WhdBU*$h?u+^a!w%D}VdKyRp
z1ZJSzbz^Qix9xQ7x5LwJXd1f-`EX1@TCSMp--O{ZOY2(1N*jo$n?IsrmXB||rh5pa
z2=Q7Y87Or3WBZa6qR~bp(Gvc%VV1pfum1=gW23Mf8`}R0G)Ri(@`q%}-XVDZ<Zx*l
zHM24Zw6A-BTsU;Q?n?C@kI>i}lx|CZ$+2+N%`z|$M!qnbf@ajXeKTpMHcP#{a^D~!
zaAwd5Hp$H5v8}D2AeHdKKhx3=b;a%saSsg*=Fwm9w~9J1-a4=;()jxN9|G7i=O%Q-
zJAKhf*H&#@mP*Ey@YOyKD}DM^#{q?g_cC0);z_>%>UE@Ama#&2gZOC;#)sM_i>9nB
z+<#G*o+&827g?J}jlJIS<e}6ozn%}jz`4)Ar{B<j4J8|;u{{f>opBX9)@iK1GgsVS
z1Mh{~^ENay=Z_DUA|v;#LM}ceBKJN1Yy1?bo8i0KJ90~D4&c&X!x94<$U%LA{IX?W
zbfu!)qK(OpNr}1F&cUGfOcJauw`jN5mL7xzy$!d+j|4y8-Q0v-Ebx`-!s#m#m;QvV
z5A;Gid$GtBQKMbRKW?G&98}GQ@@-MHa`&QPob)n;U2Q9#Z+dG{<(mhtd~Bv>mDhxG
z1>1JawJX3w^v)RM&N4~AfJDYq?osS-7z4Uy*)3wi${6MJetnST`96+z1m@mMoGoa<
z@!Vz%3u&_mgJeQt0s0K1eTq99{vC2maDBp$a*>F=o?;OjnQAt8kF)eRhE_xFu6OM;
z$VhQ$H&Am{s`6nwAEMNQNN>@(L)sTd+iM{HgobC|WO)aVCW%Hp*aXV^Sk_+lA~1`W
z?_qgB-tTHFN0o)E0V(fuhzhA)JH|c)!pvazDun^W{TCS~fdfrsT(rhhw8rp$7M0}m
z@rE<N*xt<pK>RuX)blq(Z!$2Q8q3I8z%h%hQ$ko4=_~DB+tw8jcfXwx^6Ceet(psF
zYw#6WkmZ;RGw7@)dkI5|wKc$+lKYc7f2L0U_5Lk)8CghxMeV%{6Eo6Zx~p$dS>$RV
zRvr`}t)>1P;MfRDP<#`+nz{OgI+K;WOOKpg>5DE)Y82egi!>A$D0xA7D(WcN!CRFr
z+qR<oyki#WS2o?Vn_&7zVT0`5YN7o}?5m15A|_7Qd(o#|61AEk1({b=(<?FH42j}U
zOxMvDB75}uT|4V_Mne%HJ_m&WY@QM4iE^s9u69xnQbITBD#99VuN#yVRlE>;9FxeI
zFl$KKWZj=fw|V*B(eXwTw7h+bsKZw_^%A$v^A+xi-8Z}4`OVAxlS>!=tB@x1B$PT8
zS?6=@b8W{5A!-WVJi5*r9D|BbfD-#urq}z|u6yvv%)VfwG7FRG{-2Pu?y9o>BY)W)
z*ep=ztcg(i)9kB9V!?*ztFR|ttuX@L_kSBxaHtENf#@P`#xtlZuS!7YSpIyP5xYy5
zYV*sCcwPSvZVFf7&20rcB}436norZ^1iR~XwrvUiV1{Q(5}vS1JrQ61wOw=CY`J0V
z^z?f$+ncNq5p03WgY$?#U5|;%tm7*p&^p;kdaDRtaNqvTCwu+ey>>=LV(j2k0byej
zaKel0y;xIXwbL9Kx_OUJ+D@WEv0Xdhn%!K)a6@R1u?%|bL{aNtgu(u96KO@~YP#A%
zlLlD#_LC80V=J5XV^<!%i^NA^eM{sJ!sap=140-_&U}!@uvs|6_!9<_?z!&<N*f-Y
zi4_;~zQX2sBR_A-Ii_rhy@|ikof8Zw|CEFKim_|Zer%2Q-%@tpVDo#H@9C)S?9Gl?
z=Uy5L{E_^FO}y&S*8nNfoJ&fOIDU0sL&o)0O+P_o6RTCsm*v5rg#~Xg!O7lNP-J#K
z%Mkyy+X^A0c|@PLqSwP;+F5amqoQZ%5mJJq8c6|?um@1;sY~ZtydASJ<Y6f(zImnZ
zwBKHiFuWb)^BK@mnYjZl-$VF1`N`oRbT*P^DOn6GxQpsQSxoBea>FZJKr|E3Z+5ld
z!PRAp$pOyY@aaG@-tZ$2p`UH${4VNSF`5U>UG*ewexA>nK*NEn-2?4rZ@+?I5&b*q
zgPw2m=7w;?i05Gnb3)U-B<%glLLa@W9)?G|HfwnBaOQ2zW{63?vl~nf7C$!<nRRpH
zmBDTs$vO&?S`3i7`IB(pO#;1H(WJr%aqPRHx0lStyvtZ(hVlB<obmS0zSi{qhWv-X
zTZO}n(&q_hv9MpFQ}L`6RzjVhAt=}0{Wuhub$yf<h|=Lm(C=$kSb&-DmD#3w&vTpq
z=-|ZTFZ642!7TF%MB#WpQxOdZL;7B@V(wm&la)G_unnrft3vmeCjFg=?sW7KaHQ*K
zt@&9=UoyKr*fZ=PR#xUvI@Odj3=hi?3yb`Q+ltQC!h;l1vOf;ZR`A}bR*p*fd;7!M
zgbj>YJ<?uhZzEe|Htrhx>nEO$rRz*(0F>NR8UN>Ju~VvZ;0wY(t;@OM0v}kmxeK^|
zC#wLK7MtXCR28&LwDCBNqYrv)%;-;sml;bnnzry5bA8zQO6nS;<1pVc5HYu(r}Im8
zJ_1kHlVHl>)D6O?CZaftz1==OWVCE>`FW2~b*XzVV|hUoD{I@9^0S6U_bhLeM77Hi
zfRKo*vR9cYn~L|pKn~2K1*WNDwXr~%u>#}PK=@qKT+@qK$u{RwrxRBIp4JL|tKNsV
zLZp!;v!4Ncw2JwaN7P~}#n1-~%w9{DH#n!-#Tfh-+1;eZRj=NK46C4z`jt71f6w0m
z5o3i7F`e|oL_+kdYDUH$IB>;0lc=^RZOIQy`f%R7r|A<%$ld*%jU?5KGy=U$zu9M2
zO^T>_!i~R!fhb5a@utN&CL=B-$M)R07eG1eM+y8zBQyZpCr{XOCiA0&57*MOLJ{dW
zfD4PIh$i<sOl7)Yq5D7w*3&uPIo$HNGZRX*nsLGxHov88{S|ZH)<3lB?&sFCtuf^-
zZAOdI?O<QX8#?t^`)B*++7+~f+dsjk87vX8yM6d1owg`a=<jo(aB<juAu!o3^B1qZ
zb59uSLvKIQF@X1_tYIsY(H}I{D){=l@QvZbx-YM2#c1(s1Gm#48CpL~G*1XDAa!;E
z3Jh9KN+BJ0w!%su1cjHchJ9M(Rp7?ZFn{`Kj+F7))ZFjV&YQezLkHo5PP|1u*b}%{
zJW{Xnsb7tA8a;-`*$B~1h!oyPC{b++y7~H51mtMo2*k|2PGtP;E=R-nxwwYU%)MMd
zp^qAS{DE|-=?zMx+)r3Co5G2nGRHYjl&^fkDt;>)D!!n{TEp~lM)+q-A(VWi4>{p1
z8Inj2=RwtC>z@aj4OjsbBLL{QP$$|+RW~m=xDTa(=zhN{@0=r4rY_0PmMAa=vWL&S
ztl?R^ky_zuk!DraQ^`Gesby`*)=$Mq5-bwvbgHIXy+@)EcRr3bK$MUlo+=KkYwp&%
z7su<au;bU#==)$PR^u;7a01;+oEBDj*{lt(1lp6wP4%61r<_w#X;_Q)-*y**`yBn)
z0^iWmVVcVk_Pr^b6iwk_f1;~h(p!}=4&Gn?BUZe93Sx2`ef~|NZQ$Mm{dRUTJ7@AT
z5=9wI*Zop3PwlMc^m*<p^Z?}DCj%=2?L>PR`t7ej9#{r6b&iE@S-2VjeOr4F(hdXA
zN9c{EC8%8^cr8C?Rv6ZE6sV>WsyQ!{<lWheFKXehDON>udJ+P#6sv;MZ?^e5hZAeA
zW@S`PQ8zqo_6l)6{SzdB7CAoM%QDEO*Fo2E!HDCge5>JLmmEe3b*6@`Ky*|-)YYC_
z0xF$@)7wO#0I=%g6mClF>nrsg&FbbhF+Qw0BSZ(;3vy<<2?5t(1Xsq~>+faYZ@}?b
z{bcm-_}*MYxb_ph%reDRt4%Bnn-0O@On>$52MXRAzV|VS6G<B|GSeQX(DMsjMprjm
zQBWB{8EU~7Q5c@EGCD!0#jVDDjY=F%fUyMtjPc=4;=I`Al@rIXir!0$35#S|9Xjql
z{kt#G`MqHlwqzF&6W^MgR``RkjckG}%XPt{u@K(H<VR;kUiIcc{J!1w*J=gz3kp+3
zWC>`SzRelLe4@>&%%>)b=4&OiCr!cDxdkSQK(+S2fN1&3tdv~?YQfah)uyU^zn8S|
z@uR|~j4NW|{<8BniWgE-*Gr6zGWo>S{ylR}C;XwP8InptN&tW5em)j5y`(c}%aWJ}
zd+H@g$q*Ss1kt%HaGG0E%+!$Pqx#}V35;Wbd|X)I{Dy(AZ0Q2|YQy4yNlqEH9p5(e
z)@<ymZULeaC~>iP611OI)6pLN;FKFY<)_fQ{rg_pv!I!I8Odgq-QIJsDAN8zn!M)U
zrhYxGvwy=UH^HO#`4qm>lJ_p;D9|XwUsOq)eOBx3b>dD>GX9HO(AOO$AtGU5ZvGqB
zt!pWn<qD`_?&9{IC)5_+UO2N#`-lgosDDAkjo=9^-`P72g$J`q9^&~-_V5=aUR=fI
zXyqELiyzNEvMDV0zWeg#mE@wsGxdQzYlXcYpJ@6LlOu!7Oy)IRZnb#LJ>ML!;C^5S
z@3wi|AF3I+yjOk<x?2UbTcBX?k5h$1QCOkBvOx93MbrTX(;car<hjbY%p1!&G!2f4
z>1S`j5<gh?BglrlYH5rS_s%~+J1{|8lw23O%zYn?R+EHQLmqHjwiPs?*8Ct5Z}0^{
z5T<>%`%71&Wmq`4I{eS6s3n81FXQu$b{PE@s4C<}MON8b{k-jE_TkXt>)jZ|FE7-x
z8FG|G<|24A+A^n*nX4gBOHHDZ^iC<_VHiIp=x3(?{VM`3wGp;GY-&jS+fdGtVX#-`
z7;pPZAbTCq%?JLK1&f_Qf#Efqt4E?~29O5ribV#ACF|QuQ@=2pK?Y!ZVK+E&ug+21
zb@qENxw4}~@lwL@tO#+3z*1`L<0Kx~mZVvy3vG%X(C+((6ew9ES%>RrV|Vh~#xmJ>
zgyQQ14!r$jq%v+>X21Jz;T%Uoqn85lhH#=QSoPn*ZNNB~R>w(x)@h(&xOU5CSqTNn
zbEpkvmoJ?&`H~d|V0=)Lpq{?Q|J}4!#2TJTAn}>TY55%Np_`NdW~2Kln3JJaazw}}
zmuwV@O)K#VNoc``I`-bCdraQV2cVtJD&=?uPKL6fp<W6ap`;WmzQpD^J&#A-pZWGD
z9DHQ2V$Q-G2ve8Q&Z7Xo5MFgTPUraJboQ`SHa)gNW#Kw%Y4|5E+Tg@V`J=J%_|H0q
z(C2%up=ogIsuXG3$`705ss@$hCA&bQ=YyLg&oBb<yG&cqs3uY^!6s>8W-o)lC*~BA
z>^e~`)J(q^%zLDY6wZhY2sLsHh0|>Y^L$&`Oj<T4fl0I27W{hG=>vNtn3j<1a{os(
z%H0F6pmAVP75>z*V7v|wqD&pGBP#G|%tGgk`2=A&&poh=vn4>wxg1XwBADm+*-0bv
zWa}T|A<IXMk@r-E*lKi}CKx^`U3fqo<Xdbc#=O-lD@T~KN0^An{y4t)HN<p(y|ziW
zT5k>|eF(m|68K|)uf@nG<Ph^+9+TK9daul9`dW~Mv=Jfo+PP)CX!HJ;SY@rXG{|GI
ze%yZ`@*;LO&5DQ7zO#QUGcT`SUuclw(c8;9{2KGN2T`{?a+M%LebaNJ<lUHW-sghE
zYAjfW2J|VDjDiDC+0T~@HF$x=x4Y>Y!~rKs%Wq{84lGv(hU$g?9!gh&DhZMo0=KO`
z6&wf~^|HCpC-Gc#HiQT(te9vC$d`;7VBPo6(=S4}ry5)IAD^^C$qCF|*`FVaLZ$AK
zK0oMc{yon29M5}|GG{9TeEPG6WGIl&<kEF*Lrq8ayrpCVKzxtmurRwKNTv6*1UJ1h
zD((sJ*HO0*mwwXyrvC=@vBACPwEq^|!mx`*gpdAp3oAYlbH8<NQcdrS!bND$GMQCB
zNyd^V?#a{Q1pC2>khBqlxPG?MEEA2j_fR*-4VCG0Zxi*_*shHQi<Jp)T8%S+7{eJ5
zvj^J$Wd%r!4qvZ)9095Lj{cKdEWFhPYL5!xlpQEjOpi-3*?f*7<OvK*dfEK$;MbQY
zm#x5@N|q4WM89F_3gqwl=<b`617-BsfyV%`<D4(?Ug$U?jD*!CANc}pys|WQYX#dK
zzXP=A!V<ivsOj#U6&pPY@kNs$2s8y35-@Jxm&Lwaj}OVi0;l1_vmG6~VKNsY;RIV#
z1b;k@NqSi=5>pz?wd+}mY|}`F#Os7jU23m~3YH(F=$EZPAFu8i?)qg+9iyLU2><-s
zMch+_U7WAzbRuc;hj1$uv!P#wz9Urp=mYTxHVxYB@nDUT^@F}XrlHO06x@9fn^>T&
zC{+`YhDG>~FCLBA&EH_nD|Q}EM5Lk5bhb!mR-b<>Yl^>x`aVT^2<Cdgq^3ExeQ*b_
z5mIYdHz!*Xk$E>;%FGIrc`t69FMc3<4$K<9TASZHO=ku(cV|OxhiapOW%?4#^99a#
z;Jn?W;w&20H?&(Asual9X!i|I-~t9veMFJ+>4;bVU%HfGwEKadfVY2dZ-e=EMqi|`
z8UvN^`Cb2*b}ROs2{L@qjG45l^R0$g9?9*5zvk*>+RpvWdh-l!V`;b>J(+iv5SG@s
zt!E8ME6z<2O^9D%^6tCMv;)_z=|Adk>d%+^Kr=1@mM5<gYs$qkf(N73HPa{KuQI#(
zsJ2H1A`Pdq%zeK2kwhGuuU=9IT8Yb)=~0CmTS^jm4`WB|MfASDkeeDxSSPty6~Xv0
zL9y0WNZ)_`>-jDT;B#H#H|m@<P_jLB;gqnK7xEr&FEB+G@wXYWFe5{;{WDrGuvRwd
zW`<-&ie~4!xo!9wj|`vI760}1#&&b>e#IK>;;{=_g;(-XFf9<N1f+e1<;SdZqRoV~
zURmI$wE57?NAuk$--vOifP&%U9nbO6iyr(fG|if&9a(D&lqhOkHNw)sOwv#)+Jg-;
z*MQ0FDqcaOE9oMQYU6knW04>FsKjauC43tRBzNe80ivLpw~%+IoGIK7CoDT+NfYIc
z8RO5jl_Ec9oAKq~HAuFV0hVqhI~Kp1wFVs;=oO3ju9t=&z^e?&CUI-?o~rq9mXCK9
zfiRHN-DO2lF1T<Kx)ULIU-l{L{G(PR0gGO)IOxL-BV&c6!g*U$zcp#yHiz$`#L<aB
zQ*hj&8Jg}4Ld&*nvppVY9_ru*{=Aa?MOY_!<NQa_?NRtytbDUA<raRuwoF6J+g?mw
zDhd}|LGCr~|Co~he>FbP&Q6RCKZ0BpvigeV|KNHc<!<ACei&dGro0ym9OzbM+^*KX
zo|MfTnqKDeMtG&xughJaYAmy{gqsOEYa?oO6KtAIKJWaw4-ZF6R<tFp*+8RTa7s*y
zJ}GCQh`7kMu!qw>3X|rant7MLcqifneJeio{{poPO7o=zI>^_oR!+tE1Ur&t?cah5
z)oO0s?P{1m34R1Mk^(+c4)ZDp)44u@hd7QfZpLwib^SXGnQM*@1%8S~KS=nDJ^6f0
z14jI71k-u|01Qw`L_t*8S$ltPM8?YIWdff0Bxq;}ht&igY}=1HJ)A=u4yo>Uz}GI#
zzyK_^Cuq{7N>iz<osyr{=>k+~il7OX{1wEhDVpM%ItBrWbCdMl=@8^_%t&%-JIw1L
zP2iwk=<qp^uH@m@JCSP?o}I{7zs$XG;Kbs9phNRYX<yH&6N?XPcDARVwq?zZeKihT
zM`L8#1T0uO@5R}hb@PaRB(P}5iO^(o=I~gL5N}!i4Gc@`Km7!x@w(<pL}@~>7Nm5y
zrk|HlkCNYor3tl*fILEreQaqy+B4!&3*(3e8h?$v)PvC7iNLHsf|?j<P*B77$u@*@
zKus24vMsaD>Vggd$)o~3<5`bjKH9KNR!|?q#A%XHvt9i}r9!B85M5{)Iy!YsPbOIQ
z)ww<avNobtO`*~0++Zwz{7TSDsObpMf!S4iXE)KT0BekG^XXe+UPFz6^{&TE3qU0p
z>fW@jsxhxRV-X1S(SO#!2p)`R)d5)RVGbhzD}k1NCyrU_Bb2V7PVI%JY67(YV`y<K
zqRE0Ft_|}!0Y)&#<A4!)GQb!emSwvzOPg!}j&4GL2|h9}qqFj7(Ht17*Q(yu{XyoU
z8(8LoagCR`&E;+`-O<b4YH4q^azG=}^afb1BP`=u?HpLE_fRwFkjDjhtoA#tjPIy$
zAOulW0FaKyLXTnczbNo_w_)Og(7XYA<_V%YsZ~&mA>gPr;>;?v3N#`cb_*~<4GQOS
zoFqf)`p`&%E-USg)puV+wl|3{MGCdcsy=;n(!S)mG0c(r0V!knT6o+DP8m>OSQW<{
zr-D735c1tOl&@QTM)smX@qAE2@-|mUIAdFhkB1}K9$Pgo61XJ^?`#a2YoG)DH#Db8
z1TO#`Q<G#R-<7lmy8$x2>r%*Pn^s#9U;r({@M~o7ZA4$0+*8AEqbI`eP7V<vYm!j8
zeE=NBJ2Fb0hNdGlM2&d}I+C^{U6V9UH6WVBB}aSb>_-YyFXk-|<eLDl#@r0aD|Mi?
z$y9Q9WKJCZR@1*6SK#n#FMo}6^_Rch(&Fmnz_9?Ehtp>P9jo`=St}zu*gdpbLCJj3
zA+HPYSSxp28Rv>|U;@D&c^V-*MNd-e5dIAasZtHihR)6c3*}kMn?IuqFyN~CMucS(
z(+#uo-=}5)A=kVhL!9ejt5#x=jCpE0hR`(VW1tz;6aXbhT+;u-<b?{-nj%diBun~@
zC@}A4E7brj186%2ap=~s#I2eZ9mWL_j(BeDjGM88a#-!R46FK#N&gQi%&wS_Lc4wt
z_G@y!Ie>zYjzhQp88+O`*6|s=O2OXp`hTeD5OgG;529<KG8rGk42wQ>Oc$h>n9%Pw
z)Pe@dWS(`hCVPBx2!Lc6+lF^Kgm*B<Dg8>AuSo!NHtMXohylf0Uy~`W9p4?3pu{=5
zKRH1=@Wlj#)(i0A_24>KwNn!-fTZ=>0EUH@&Vh}c`{qSjdP}8d;lR>3a4f*)vFLMY
z+$puz0te!iw#0!R43sRf$s?||5a_VcB;6cw^QD1f$pMjWp`~*`6C8!tB;@MAobqNU
zSWDs$O?q@fA4P}_-wVJ=idxlD?I!1iWn5h!GLjSZ1Q^J(X@$K3Xe_5$&=>$g0lr4R
zf&^(~l+}M%<q@<Mmi0e~Hfcmts5}<S*pA>}_b^V=2-G^VtC4a15pl}$RCf4|QE+2@
zH7SO#wX8z8PUwONwH!H$mF?SaM3rI;)8qiKUF(}1um%8X#}rD2=}7K5&8wtVfLR?7
zG&2P-VPBxpijY4&Hj)4cbxc4w(}<KM0Fg-w{`E<KRm|0KCBE3F{IK5`KR{-T93e5l
z#I`6q9=yY(jQ*%H8&cP?$zz@C)rJW82C1*v+c1D?gmHAC@zBphjaB}{WRAl>JGqvH
z3I~oh2mD<XDjYbP9C-PaH;!g^%huQN1Y{2cI%HM>9?N#i9J_{ABho!~+&KK66>hWU
zLCrxDhFK9uP*spu$Pdi{O$#nuJ-1vI(5)}|&<ITcCoCf#^cYnF2zV|zjC0U)oX{CM
z{1_Tq2+r6&{VI^B_Tm`(;ponG15v0HqQluc3fae8c?E96Fq*r=aT{W{r{RAwJ~YNL
zx>rIEO7&tg03^pKO(38f>b8yWmFXCMD-9=h!{j<RwG>ms!`1qAJ!n390U4e2oAP{2
zBkZ@Lwa^E|5WaL6fWz;t?Hv)>%66l-cIT3EiO_Jg9a<O8tF35pz<gIDl4IG&ARB+0
zqg*-XZnQ@-ID<N!<R$%D*7tYEeEQ&DH{+KagJxE4_31fYX$#t^LWKiIn*+s!@o00r
z^tbiY&4mfakOKz;9S*CQS{=iSQ@UOq9MEKG1oI$Gq0|O+^<pL@GEg=T*%S?gLUJc&
zJ&31jm|QjQ?xq}D%PO$yKcMG_#}xQ$YC57kAO<0~qX!Vfxp8uNn50mjj)v$RLUIo#
zTh`GuOQF22;ru*ZM+>61fPXrh*Ae2aTSp*|<Fq*{$Y4K&XTcK3W@cK$aa&nB1C3rl
zf{FB`b~P&kB&d5M;cFc?ZG^zy5Y_`TaJ-AL4WQ{5qHQ}-ThiXq6M)V3_36;O3WKz&
z70^)|9ilA&2Ee5U^R4V$9-aAzHv9T<PKV!ojc8w!YGBZISWoRm4zD{ijaCPLlxl|d
z)YwoiD$#K#V^Vw6*1;gySNo!~y5r8|WsK_w&!iB9(%0C}f6QEdtm);QdjNAo)tTG3
z9qXZ#{tE|Ilmq@Q7x)>jXrHC=!hynpmF2*pK!?2-;IXpTptQaQIH1W=m`RxwPl28{
zqB5!iu`BO6)2vaV_D+y2PhpvlN$F-V;jwP2XD*IpmZUHq<$SwB`cbG=v7Cal4&!Pe
zY}+CV`<eyWRt+$~;Si|RhA6a;VishZ89;QgVz1CXPCNQ<s5$YxG-`Hm+A-ISS<xda
zxZl-cjG~RfNnEmB$$DX4#{Lxl6bCeua{d&mItO~H3A7xch2T((mHo*^hdFvx_UTa5
zk-!7Vwnj#(5qQ=@vMuv?Xwp?DsaiW(jmK##GMUr=6hKGN&<Rh*b`MwG>-(z7hFRB?
zzAG`9R9^oNNem><44Z3M-O(^_<NA4R7+dn|0La1namL$?_w!z^wS3Rn%Ns1?LD*XD
z!gOpCNxP0EZQZh=s{9lRpq+vFe0B)J4|HyR*|NW~99ZhL$iKan9bIWXb6~00Dc3LC
z{W(y0x-U-_13IixfXBXkD|;s(2Xx-fiTf0~6~fJ`?Mk|4ONMAQ3X%|E7ECB~8}W&)
zGO<Fuxj$T4A7RD68^B<4KZRcbhs7<YCL(sIYFnCKWudy8)q9=Q6{a1NeiqhE0)GNE
zb}8wv;W$IfLE}lIkk1O7@V$sKsr?LC6#8cn-1Vc#IX1)>YA`(3jYdQ>EAx0TVjR8I
zFra;4JPp~OW7c3AfScS#%ug+aTBG5S&~yko!si3<)2eTW5u_++l0r2X(FSugdet<J
z<c?8L*4ghEge7P~;G_KsBy={9GF%sl?!q5N9~CuB(GCYw7IQ!pZ8<K_>liPX>jivd
z6Dz?+GaYNt(L9#&om+)}2cm<WJ5Zf+d}TOLuEommb-7zD*J-&qdhln|^7pp7TfF_Y
zcUPBRa`B=DS%Al)*i;G~&m0KT9VV!@LSetNgH?AxgF<V{0!~f|v<QL_1z=a&75D`v
z*3}tY;dYo6cY%gL!3b3f>SiCO<zSyH?}0mzxdU7vEYE~Mud{a@r)UyrK>C(sz%h+B
zqaB}x4uo?bbJIkDQs7t6R{oE%GbX`7kln>`oT}NjVy=bEA6kLwX@u{FLwG98{K8C0
ze-(s##xg!QR&`=l6(GZQ7?YXJYh+_@?MYxkK4|8p&xOwB=IIC_U!RYWVR+OU=2dDs
zW|-FureSqL6IU%psUlhxz+@VqjyjRPS_C76DK^TLooZe+v3CsMN8~Ax!st+n<`tTv
zk+G^B(9%pJukS&d(#7>KS4iSP&1hPTeJ1mS3i^p?$~KB4daW}7631+^M;Rta^LlrY
z&toZMt`=bDL8}@{c{P}G9dwZ8<bE7j%(dOGQaMmKP&jbZIZ!kZN1f%-;aS_xsJ?vn
zH{;Vhf&;ztS}wo*a$eDe3I_@Y3I_@Y3I_@Y3I_@Y3I_@Y3I|q>1OFc`FO{4X0oQy0
O0000<MNUMnLSTaI8^PiL

diff --git a/doc/image_user_profile.png b/doc/image_user_profile.png
index d8e515b59a4f6c80925606fcab65b0757266f764..3e4ec70e22716388b8e7803d15ccec73d2739705 100644
GIT binary patch
literal 500972
zcmb4pWmucd)@}$;TnZF-r#J<QLveR^3-0dH;_h0YxI^*c!L<aZK(XTPQuKs(@BMxI
zI_KAUawX4XX4c5PY-Y_wDl1B%Arm44001-@X>nBm;7tGk0N0H82ATtFx&Q+Jz?fN!
zi7Cs7iIFS2I9gcSfdBydL^EUKKQc`8!zL!i#=}1u8IWB(RU;x|RE+~iz`fuJ@W{7b
z<BSY_{dIiIby)RYK%q)ci!Evl`F%aha21(8cG`qVz!>f2@bVY;sAAp{9~^LMlI~a3
zq8C7&jLRn#Bv{x&m;@4KLsfw3N5CSbgoqT(qMFE&8=^0KL=K$EDNH32lFOTVR=^@~
zgdmbZ<OY^pj}A#jfFv>}TP+LK9d`IypxisZM?@DG0m9eE*C+4`QVyob=!xhdj)=tv
z5`1;uV~06(wivdgH!d@$Hyt*Ywj4I5hdJ<oi^5^sVk-hJ=AwF|j-dLW@`*IR{Q`rT
z{kfM=5cRETC>cIJ6D>@&jby~YARC(<?0xhuzNmaB%3hApn=mRRWUOec4T`b5*kYca
zzYu$ORwuw<zT=k{$ObCvQ5Gs{VC~Dxi|@nBOJF)X8{3+xG0zzqfCts=5vyyy1B!YS
zNL$82K>_d{nnnb`!4LxAp(!XM075WC|C5%4p$EYJlMe&{!mR;t|E{A5eZM{l(C;hF
zzuvG(VE_c^KP>3i>l^StwciAMgZ)n$t{GYe_@pK#BLjV_nYn;K4z5;?ZWSP30%!)3
zleCU20Dw#L`h}5Er9OkU|I1ok+f7?Rp3ltDp2fu6(G<kuW$*Ob4nV+*51O<GxtWlA
z+1oj|@_7kT{!@bwntsh@r6m8Sikqz<rM7}HxtOC1h@6{+jfIU;2$`ImT)@TLf=^Xk
z^55jpe}a@&Zf;I|tgN1%o-CdpSR7p}S=o7cd0E*wSUEVDp*5IYy&c?4yqFzass7c-
zf7%fTxth6HJGogqI*`A%Yhvo??j}e{`P$L{JpURe$jkbFdvb97_q3oBWPL4RWoKby
z{hzj>qyn$Ge9G2dAUhp#YkMeW&_0CNIoP-b{vr5(ivG9D|031?FH&x9j{he8ucH5(
zRKpeIBIalh?bJ=^e;M{~;{PuEH=zLQ>)ihfi+=(APc9T`A!Gs8|KXYtGIf2SCzO%I
z*5XR)&^J`fULT+y^oRamZ)h4;eRocj3o2wH02%R5>RvEMIS4`K5>2<Z7ZW?;NyHLh
zxTHc#taoo4?kFrUC6JJ%C*%^!$dx1{;8pT?&vJ*8$v<%*(mRiQAKS@US(I7(J9m6g
zUQ^R{>!AdZzm<P?^n2NzF}H%x{owIXo$I)Mp$oQR(81VPge%b^`B`ss*35-UB6u~|
zaBg~@(w>+2Sh_`Sy14u5R&NFKv<EUBn>8DB-C3>^AV*9>`?v66NyO-d+&W5vrEAa4
z4RB9$KJd;+;k#&;LvCAHiM%F$j`RkG_4TdHTB)C=_qFA8x|hddHqBS(w&zeb+K!s|
zcNi}vJ3FQ-Zltu4qDv)0688~Of9PP4w`q+oX57uIWC6)(=O0Y|+Vc-vbVR6CRTjw%
zQ`h$3({?<`BU0qkFqg*#sB#i-)2{5`3Q@<oJ;(%!l<<UljxLCQ70th^n!yt8O1)w%
zz#Fs1!crPTySlX{)H6JHN5qM^A_krhim4CQ%+73o*vvmDR2_1bJsl3O_toR#oOVWb
zbyQ&+p0JS*9WKZ|<=1!g+B#sDSaMb)*fL~J%&Oy5v^gOnDml28Pi}O{z8JFHR1>-V
zM36i{cPJJhH&-yOOLlAvosn#EN<pr-mf(~%Q#$C!P6SPDSJ{Zi@|W8f&;^(I%Clcz
zV7b5WV{03$?@YMxH9<99HAtpi%MHnRvP4&J*(TFY-Zn-x=(%3hXJb_3siM0@sqqoG
z9)vjDU~d|Jpi?Ki&>;g33hok``@NJEGSdZX)0?6=*IuD?&16t={x$enC~}GMvkowc
zv!yATqg$dJ7p4w5s5dA`SHZrz!pt~TcX6!(21d*fO#pFyP6<B<7bRrx^jHg<R!NyL
z-K~7*hI?*$smiCVO=GzT$wpKv`XJ3k8Fw8ahzPB0xYlAi`A37Lg68Eet44B(k~(8r
zttvKMQza<6ZFY+QTcb8e(qHu@cYDR=Td9;t5h;uab-YYYg}*3A+C8<ro?e3u?nZ;H
zP78~tktc>~Sr%K9T2nK1kfh^j5b(#L!D1n;;^8b{xFWT)qOd`sI^{}lSx-aHH~Ju{
zBXW6mslY<|2g@ky`O+5u1|OYgSq%-Nxj@SS4|9^x{(4sN3Bce<6%xSUsAn0Dn?R>%
zx3K0r(5-b~XEn*VZI|q5pQJ^$=5s%5Uj9-5Kc-RInD5L;-JRJ$SNTpgEe*)?Ce~^<
zRZeTqxKYkls#4L+#x^gD6V%pdR%?^)5nw;%!c{CL^9VCXwOt5N<|YQ^tgbS-Wt0WB
z2d1$a6wSCiZJr*`WA5vatPILkt*(;J(REO9k}~G^acra~hZ;%Zm54C!lS~vpcZmz7
ztkBL)M(~XMvL|f0=!$nh`pR68`nda$yF$p`?^OWyGWbv$8|zadquV;9)=hHgUX0P~
z=}`eP7~5j7y#mRBy0X<jWV%<)ICjj}=6Wx9*(WE~+&cYy8^3m|WRQyZ7~>Xgv5Rmv
zb1~65>#Mm;&VNQ9X63bRIVX~5<A_xppx9F6nVPkfMPK47S<a)KPXCRLyG4_@HtnsU
zLA`!8ySplLGI4Wb26a39u1>Q$097C63?>5rn?(<SuVt0%+V6&~FmnOoxUuTxr`}`3
z;p^Prt8$S?^#`b2VW1+elfPua1nTO-7s=qf`~;Aa*nei^vNkp|mCMsDU*hDMqWs&4
z$OE_2Xu2;8gS&)d3{MWb=F_GrV|WalxUvO@7s%s~7b`KT1MDLfn!G}@n7*M0Fv1ZZ
zeC1PdM{$~CC6{_ALZ4&VcatXVRIM9OEuFB&rh(9+0o{@5flduxQ^+uNX^5mVMY?=^
z8oMzvJ|EYSVSGyDmiklCKEunRmTjY8pIKDTl~fi_cP|%_k7-^Z?ls5)QjsbQ0EE+D
zsZUenk~9Q?6(UBt05*GIfjlmu2g<RJ58UshHMwFIfN&5iBmB|6&J;61bGkO0CO&{c
z>ZM|}2pucTX$5P6XAkCVh7wF5f}tre8()u&%b2U_E!vI*e4|nWS^;?YIcloG@fc1g
z2L;fpE`>VG%E8)CqYKa>3of;lFT3SbD@!uJlW=iWYB|8eEHm<gy&hu54MJMCu6RUc
z$O^ZupNGc}4;<tdWXyEbM8B#)QYhqijB&%OeiN~GfXCL(3lx}Z)FA9Ugy%`NCz7?;
zu&Ka{QN!>n!!pFI8fFV93zO{s^frvBmZ)=jPcpkD!q0Xq*~G3Z!wpsqtbsH7UQ(0h
zulKsk;uV`UKDVIT*nL~!=WF+3g#>Pov|8;Fz9qhcq}rm?c2z?2(AryXs>(;Izn$|O
zoSR8a1>RYIYc#V{C6l<>YZD8D#voEGyBTVo$^gUj0^QP*fknGfj(5R%SxIivkvW08
z{s)Krb_Vi=CQJdV7ilfCzyFj<hu&c^VE%u7*s5^8Gxujvo_(lljEj^)7XSbFpfbYC
z2IY}>{I1@?6&?ZE&Qi2Gnrqvted+hzHcZvzwRzt*l@E9dHS6Qqj5Y&jZQZdNdXc$3
zT1`teI*Vs~#ZL2k@5UZyw#mVDob-)P!1FkTRvZpnYdSga3DyOxy+avY$axmOcBwWY
zo&}pdXAN%ld{`kpxG~wyO_Z-z^d3vAlwl~BChmJEvK2-^^rjczCYnNoWqM@qyRYc9
zWZpN~SJh|4*5SgS>2%esGFtyjJ<*K&3#Jk}R=ACp^IgW<+k|G1i}e~Z;Qr=*kwD{H
z(2f-^HcNiaY&Z@Q%bJ3Q{Hn}nbJ`bVf^034p)H+=;V)F%Zl>ZpEfaa|p<A950@nCL
zzI&#MaR^%uWqUSa!P1bWN#~=Iq_e-1Uc~L!q}5rmN<LCh%e7XKuYxr+UbB;#v!IQf
zDA7`NY8HuG$1<n#ZEAjeZmwQHbLm%^WfoL8I@0!q{OcXItq%7N;(;<`@@}vr_nFre
z-F>#h_mvV%Q_^bRpw>%z#<XK+h<J)$dAV?TIqPAr;>1B=x@n>MnB8TwDaS_Vv<4(M
z+draao4J)Rll3r;%@>RK$Wydsh^f-V-*vt#tqcwV5PnUdodl-RA9DTbQ4^5WeUo1a
zu*<l3ixB%ley4^-C4(al_JE&-QT|MP<cIVIGguBB)w2f2Y2pqb83gcZc3s#6?4N~_
z6;*!15?JAQ`^6=4bCoQ1VVGZNG8mswJP~gdMTYkDLX6J(?^q>zNMNEx=v6e(o&p>8
zV9(7xOh)A{R_+hVmYC2*Hs4--hUlJ1vGdt`XRJgIhqSiDhyd2f#Skj{(KlYn>H>ca
ziu5SRyBV-Slg!!xTTXiEOwP`v-b^PyDg!^Us2yFFJ!A=zg?MR*^U5RddtssA2;vZe
z67tRD)kF+554tUbuZ)+PVujR-91vRU8)Ko(ZRThPowXs998<wQ9O*6RuIxG-bpEui
zqW}02)rXkdHk^f#dJlQnfbk`ljv3pc!?oq*$_6|8>gvz_W!Ghvt6Z~f19yMv!NhNa
z^ZXhK8S@93&-x1LW#&^O16tWbenCCU$#=6PIKC-lA3!`1KR7AZ@s0gtG6QlTRmi(u
znqpRuwm;J{LtGkEB=H}>7<s@?0(Dq`!4x%wgqj-oKGDpi=R9#P-b<3Nu#6dSa||i1
z8FhZMB9au$=ex;@l_Dt9#_EZ*kUtoY)NDc0k&JL~W$bHMnBmHy-!I4=fCZ=}D+oVe
zrOkLpzw~L*8Qj@TFM#GKk^m*u@|ru?O0cAF5t8;etN9YrM}`_fst-jSMW6I5Vp8MY
z`1dQN%ROTmmG*c7_t)UhN{hY~rSSNE!VC2lOFhWr7OYgF86<AUVG~dsv<#idY593-
zxcF(AFV169v);^0S*Y^s>N{>F1NeA9`gUHZ5_B^;T+^cEJf2-Md}wa4WpBjvr(|n{
zGykGI%Cd4)PrHd$Z1Q``aYpG!*RV(5g3(b2;f7#hGJ|yntw%TV97CeNn9(6*jfCF=
z>2k6kT{!8~dV4}G&TCYf^nI)${)f2nd-pRPhx5L#`3w0R9t}%e5c~<fL*0_pj}Ux^
z1R(d*v8O2NsDw#?P==}>5i_zn)#)}fqpLVvz=f)<T?kPN7~`iF1i1Wb&oEXin)8iZ
z;1qcj{cdbfajD#2s*VpjC^Gnejk+&mJw-69f2McqUi{J880I-1a3qN|`g7#j36=$;
zr+w>3z?@QiVS?AD&#U8L_kl6^Dk|`#0LA_NmYJnx7i<l{iyFkRmA>^pFxdq&Ov`{&
z#>lX#5cNL$k+KA4bQl2?ssaFNpBbd98Om!vWv6)cs6GDzHC`%M$z}BKcc74CI@2eg
z&j8Saa$us|maxD$9cMPaM9L$xBXM%siQy%wVxDS-2JoX4-c~N}g2*l=uc^@#`l%Bb
zSku&JSE}>GXB^5<_>vX<Bz*>j62gHa+8D7?$)9!fFPEFt*-ZX?z<|G);#ZQ5#i6~p
zKANuxlawL;FqLiTVQ9u6j=NId6;GoZliEf-y*YfnYrX<|Ey)7$v8rsWQ-N^GFnS73
za;4>S<4E7He>0@C_6!JLNN5`2Etww1foja$u$z5ReIFy`*^D$Z_f<VN>RH3F2Y)`J
zJ$}0Uk0$05ry`wYthlFHE7*o#97^r;l7PN=&tvmgPQ+;HbCm-+uq3cIA?-#^2$Mfj
z?=G93cnnXEFBfT(X)p;V=$+~x8S?3yR4xYWa-Q?*dwk<U{3u7DgVor(QIOCX6VI_k
z(xB3+`}0M6cQ-`O5W8ioeC*S2N`_<zd|*HZ?f1=04qIV0^jsc;A6u!nN#Eg5Wl4+u
zizer{68jQ8hZmoiaA@y$OGSz-19W4D63&#pbc^}_V0Y0Hy}6nQI+t$M-j$_SwiPmm
z1X^Fp(13`Kzt2sUW$_dmKy-J>3Sc^_TZ<*ki$_n;s0HDwhL>$tRR7KtA`OoIm@qm_
z#2_H#r!q<!b*Tci9#+?aQisl(M(N+k+4^;9j(OCk1hD%RiBPGKN5{|V8B}cB=KF}R
z(}`!+9j>u~A@j>8U)2owJct<QnN+{6GE7o8?14q{w0k%!cQjRKajKG0F;j+TkpQL#
z9Fq%lLGc;#&g0SoHvI`SLR9dS@CPA4ZX?@7*S5D^4f>9b@n_v3wP@m9X%aPef~w@I
z#yIrKTlT$$y30#0spJ>U`sPu=lu$CsNZ=V4jjhVbti64#pM6Reg#{6GC88{btlqaU
zT^|otB6J_z`52TJCzHGlS8~t0<bCo%$;x!?W?fLC;G=!R6Y1MKh9s&kapc*L@Ne$h
zKy;>mSNI*udHL*(MP|#GNlr@d_@M(JSN&E)4=~ys9!yKSlGA;-_W{v$tp7@^ozr+2
z=zU$k+%aH!XXmj@eSEF!%#ctek+`Tks^}(CZqBrG@AHrZOI9#G|09+}j+B8-kv>7o
zaEuZGtduv35Mx;k>T2P>F3;7dyn@VWW-~yJVv}ec{Loq9eR%!FxtDiWP}w_*3FM!)
z(~TyDS-LMe3M}CQqKYC|6M?axf`eDzRTQI%^NwFV_?Sy++FF{a)2DH~!uDGY&^13l
zZ>Wy{1J0z-+A$unGqJX{D6;z>%rk+#%eu^~CGQF~y{Q}<BHkLkW)pq^q8k-X`ui6E
zux-ixb8J!;2@6Tg7+%ZXFsyIbej)2~Y}Ow#SoVkACnD%$_}kscbz82nf^^ofqicHy
zJNL^oRok=qBhnAyD2HNw4fcFqBLc&`m;165PTmj(UiXo?@r&5l>K$vm$z97p$fc0&
zO_lL?yxc&|z=v%z2l}u`uu>w<IgRSKPxL`^N*ji#xl0#@&tN+*yc<fXjylG!&QI@g
zq6wg!W;<aZ_p1&qWLo@m;QU){o-Qrs{RJ0>dx2JwNRh~jDpXe5?6b7LO9x8_zkeNk
zC-AYcnObT^JhNK1ZO|$am<TMCg26DVY-IaArR}?3Y7~^`V35+ZU{G2>d5st}!#xU6
zTm`bGqUEYxoe#5wB<1wl)^=o(SY?Q+`{Y?xN7dbDnP9S!UTZSGA990=WFipuNf3B(
z75zDdpB<~zlW-WL3hixkaBx<Zt*r%Llt!G+>xwtV85z8wyM;eC>Fh{+IX`rsnphiq
zo{L$D$H3N1kLkVq%LbyI4IMYke71^3^wX;h8rwnDuQt?XxfY`d=Ma6Wdw@ScRK4PN
zVCL^be%@Vw7pO27JHiFX?!OI`K<1=|&q&FF+pxx(8L~J+*bY3o>fsvc8HEKt4lk6E
z+YisjhQ{IG=iv-vHB+5^5PFcPaxW4oRD)tB>DZ(F{MvUqP^c^H0(*dtWz>#k;LqtR
zjy|P9@wFPdJBW|!vUN^(HC?1!bu?>|H;Hcoc&2Mxo$^<#bROmP{L>DsZpaw%gY%ra
zql}<(6$X@-eRCJQ=JY%JO+Z2u{0qz-ffd#i|4EEv26rrcEDruD&Kc73O&x_wB-^U5
zbX8!Hh?y#QX!$5heAPmYs_PXtP9Qv`k!^@(6etEK0;enlsN=`1q#Fx(RQC;3qQT`6
z(#^cvk->k+S!0Ad+Q<2B2@0&$6>yByHH#(qlWj-8?mOIhG0Hd)i_paeJK`<vvNl0(
zY}UIG{y`+}{G-wz+|Se!iR(`;wUQ0QD)B5HK`6dkG&pNTX1OIy`D=F%qt?CIr%y?z
zF-@9kIizQFKw<8tKfXmGUsQu+$_aL8JWX(dc0x9!{f%W|UD8TdU7?$&H})IaN7BLP
zQo&3n{O_fPjPfG5punTxM0D6QHJVn5L1i^aA$WMf&ZhtN72!IOV5Y9L(m3d}w#uda
zLu3G%0$_-h_M_8u(yAvb9lP^SNyGJ0+{RFjWD7{jLt-}NnHtLHK~AV4O@yMn6Gw(=
z5$8*U**`=+HT9>IRF6500V+S_u|S<(54eEoGQ4gHD44U1@RW{~3u%$tQLkX80TXLA
z<XZ7rUrm@HL&CtqMM}GoNKhBd%ekKSpNC$ko>ywDtlZBm9oB53aGX3z65q)*FWR=j
z-qaBZ2M0^mFz{LaR90JB>LieiCS%npu!OQuB}hg_@6qbAxBp57R#+YOPAfyHW3#kw
zi8GB`)wn~AT<%vK7~*wGQTGfrXEMRV%Y+Rprsc2prWYr|k6AwZj^=el?Ej2t_Py@o
zR=(<SUkvA&nGHD6t{&dVDbHe)%M>!Vpih;d1sCwSD3S~+Z{biRv>!RSnykj8#!<P}
zVQ1bR=5^(+dJQRUa#}K3zoM@EPv=TUw6BIBR{FK?x;fL#*S@FeG5_iky8mf;<%ndF
z2(A&zT;!Lot!BKZ*43cAV!_J(_eBaP)B769dds33W!Nu|Qthqqq|EE#E|WWN@0|6h
zpW~pg;&Z&JVYd0_ME(FQa5&1~T@FCx$=`qKkL9p>9MY@rh@pnB4>^Xe*UfGFJ6`>o
zjRbTgHWDKlimzs@Q<?mt$;UcUJMMKTy0121d&IIaC=b*rNCS@h+o}IpCtDW}m)&jN
znkDG2YUAE^?=#z<miQ)0HaO3)wWT99_JsGh6|8Wf^Q>Ou-N+QXX##XkC4l6l*3EDq
zE2u@MYdtI#CcT~?J|-H2Iw(?x1R6W3bxVPpDD7S8;K5RnE&JHDu(~&|I2A*=Qk_D2
z{*R5_R0HqWmd#@k<sSJ&-YE3Wmkua->>tvFC=AKUap{6FK=XLjV=9i`4|2usD4|f7
z9KkUqtZE#i3{@Bzd1SpMU9K;0&B&fOvDF*z2&m9VD&I1&#nBt1EO*&Rtq)11?%}<<
z^i!y|RIlpMr~M<Hs>tG9nhXP`cmFa-ugpp9R^m2n`Hrpk8)|++M(3rU=X43GF%OF*
zYA3XqU0(iL{b(#)1%ypbp(8Ni1m~I99b2>t(~OzOmg;?EG=3yyP^N)$YMey)_kN$i
z%-T?~DG=a16#sRjT#3-0)cSJ#JtI!xtJ=i)BAX*Ezr7>4HRzWMqEGi8!4f=-04BD~
zshIT_yfSn)1gc7)1eN)%pYRe+y3%nr@hqyw(l`<R_R?DoRF>yl{8dke5YNY7+f}!>
z3%D#Am5>Zo4R)!eKL|P98|T(wC5_Y(0tTQkFOr5cn@PooZKnSDG&gGJjU<-2nPn~K
z3wF^W5kq7)hRnj%B0Y4zzc44i*6@P!3Xo>W-Ty3GsRSt!E^rk}JcW&XeAO?dKF;mo
zcbuajDyhQ^wK*Is{8ilDgNkJy$rv2zVlY0j9$MdmDp5&PJnwRn0z*TolHyN*|68BJ
zJ;?+h@3JGD9OqV*uJ}h@Tittq=9NmHxj!Jru&)c+O0OR2z>!RaJB=v-OYfZ`uIc*S
z%ILEpSAM3iSFgz<yX_(WG&r6?5TJ}5#R79bvuEDBaG}F~*09NPie9m$8M!T@{_9<q
z)cwWP2sYPQa&up!?s~Wa1$3v|dJp`f{6(%O%{3{paar#YO*+Q|@(ZqJX?b}r%eJa!
z-e4Ea=3+~xqPSS%`s%7|AUeG)-7r9-liM*@x2Zx3J6^_ZGnQMO$o@yy^V60Zq^vYA
zv+{oXnRr<tV=|xC&Fz;V=ya@IbI)&7NPWMjs9(kL=d{t41?7Ek<9W~#A@D8{Zw}+I
zTrE0Jli_wt23yT`ZDFA_jYW?Xb%y$#QDCS0tVOm%mG1U%JD#DJis0pVLjuL8HSf2k
z-;*m^YK?lvyd(o~sjcmW5LaM^mWyNc^s?RcCN*IjLmvy#)(tu@ly2viwg}Y-U|;?4
z5Oiz8(ow03x6Ki4+Uppm#?s*O-dka-375H{=XBlK=i+V2eCh7jt`NuoZ)u3b=C#zy
zIrn}Iaro5gd_i87^;M0;echGG&W=!`d-C5lt=oNi{@8`p<-2PKY@J4#iPaW^?y6)T
zU%3U>o7;*tTY=^!^BBznX;ZX#v*SJ=jThODiW*gV)0Oik<xiE@>CGXmy>xmKM(Y<c
z^PK*nTxf@ccO}P&!e$5kfB1%<Emo<{%;?z#lu8;krvmF}=Z@*@3%0qe8Rv^f1Qtp)
zBUhtu>}I`dbapNY)cA{Qdk!BPW5DrwEH#d}My`Di0Ux@r(A2xSE{j2rOIZ@>b1T(>
zHJvAXgTBA=ScP7aeEQmKLE~a!f8qJozNd4H7&iuyxgZ>wooR3wEqrEr2P^z=@?bcl
z3;GK&YLrL80FezggkSBWtIW&ia8hIn5bLT#qZ9mMqNvLgVXf9ZN!z;X-=D-z>5c0&
zM>huxFsJ9|OIBNlSZpAI9vYRkLd$Y*;DQzmvugv+*Sgg(1nwg4hr?HESD$KDqU>V;
zPYc=4H8n2-j*Z`qvAka%ZqyLEJjpI+9v)XQ0ty_uHyut}Yy>h>vJw@|KT9R_ut&fI
zG+xMz>O#N}Tt}ZmzLP!HeV6cOftSbVgT@AyNfbPzjHes#7QVZ0utm=!rg1mEd->sW
z?h7l*(?TBlDfnYrZj~pwh*ryP<s<EGVky6pX&a+Ca1wY|>a=ycTvBzBCS`YH9e%&M
zQPOBI#O4Dn-fl(*3Hd09R2D5+Uo$OvYi4n(gw~^RY%oz(!c<55TqIYQ)>RRLzaqa`
z>3YRjD$;qz9S1CVQM7U#6X*ojz(4({^i36Y#@V&a<~Z+MWdLXS>@<!5okpX#wK2ZO
z)o28gPXEHv;b~;a<TBu9(Db$pUbUlH$93^&c@*<hmeXSOt4?#DwwBQ{A^9M~?Qt<)
zm=0s5N5OfL^mn!R!y#1B=@8eLwcz3<a^!w47e>M5|1Fw8ude;zlhkN#)u2&8q}^5T
z>jl_o9JjrLZd+v(QoDAR)p4Pbu8@g8w%)4NsL^CFLLeZHG`~_RA>}yKxt+X8P9$GS
z2qtTw%I;-3)Yah*&XgVh$}6U)n6c)ai-O0Z2@!k|WvAD~BXVr~bndwa>gkn0mT(em
zEl<wuA3IepC%segNiwAaewLh1bQyylt)(MIdlM(G6;0e2E1UO_W}f{jj}ql}Z0mtL
z+!G7e-t^mOcv0$nvUg-jLrz%d-B(JcRa9@VgfQJ-eukff$1})}6TFw6J85)mAwQ<e
z6O#Iw{jnS0vR2Sj>*2IrYwB~^U5^g|`Z_{%9xZz%r|08$uTN*<BFA{PnJ^yyv*iO2
zS|Z0X^p{=}8`IL5d?aipS+uSa7-4N2gMjna3bjAWaUPv_Nsg74xjA~je(HXYsD~YS
zNZMwc8$Ic-ANz|Fs3``s@0uJsB#$#W#It0YN*}#Whqw~7?USC0r0i9IbF7>Ab4a;9
zE401tk4q9^YTjT)<T%<3aq8!oVEKlS<2uAg`P^UmL-3dWgvMD~EJUy?#?l&PP5Ksi
zVXYonc6W7nvuBh{-@Rz<wS{)SDX)8>r6B9DEUx5)6#SI&oHh7_M2m5JP@-R`nDGdY
z@4iYKyTEY4WDBE(G^lUCl(H%*Y|+JaY`@?VI;<cEGSFrOwdQ#E)nX&_Y<5cBC9JAz
z(hXEL;*%8E<#D$*FcCsmPO7|F{$<|;%;do8BA++#^(@8-Q9A3RA`k!=uJC4Q-lM;H
zjl)4PA3VbKjTlwZYfbFUPdyyOGh?~GQbxJn#~x=4`GV}Lcg+W-dw|2}ef%QmPf=&h
zVidx90O<vZ^`YWn%s{#2@9dTZdmX=>!9F-BWc}}Vc%NJ2{yHTN+cpu?Igs`}@i#8=
z1BL(6pI$Wi)yax+y*U%hG*KTXHtUmEFKr~*6}k+$`ElIi8%3677flBDrGwwtV;Lmn
z{wtr1kv3&_JC6UQyyp42Y2ce)hxLcWzk%jZncTh!{9WDU`BO<WPijCIu0$Q>@mH#d
zw=mVF5Ct<+gi~TGAi1&rgmyZbS0yZ7;&uarwCrZZ7m`%5;=A_cV6?JUB}mjrcxUW>
zn2jIo`En?A;F|Z>d@8HoZ}X15O4-nBgJ!chPdlf>xj}&w%xS+go8<V2kG?LY_bA<E
zk#B;M`BV^5VYxnpX>~BmZ~%L{W^~+js5n@82aQGZ)SvRF=bWxM`%lrVsl-1XI_IFv
za^m{coDES%8eC?Ym1<|*B}u}hEO77j34AWVrdZ~Io3eIMFM)B45f;5Uj7~3Q<k<*U
zqs4u<>E2gMyozj$PEKav{_6aQR3ep;XiAj!nPA?@|CZh2z_#oV3RhrBt`joY!6Ldc
z3BOTI%(ioS1DcwZ$pP_)TsBg;tR3yEML8dWz5qD}nN+k@gorx~^a7}bH`j=9@HJLx
zkHaXjw;PW$eACBDP^w^#gyb_k_J{hNcWt|v4Ox!Ya?nh%xZ4H`gDN?z2C_V6%gFr$
zv(C)4A{|F(Xhe4E^Q-(TiS>?#KiYm&!_C#{;A*VvNsELD_+>R<!SoSh(61gTpL?h^
z{=HEHl2+j@E{vsi>5m|Q0G6S3t@XDLbloDh3xByA*#kM^+w8G|$<lqq?nv<C8;QQr
zWxtI+3_u$D%A>=bK)5)#^;OF7gUR8!3|ThA?bDC8c*Mso+iK`WkS=u)Cg5VW#JN`-
zDp%DU!DM<~^s&0nwlfOcs{F^_tC=vD<}~#jYLx%n6Q@yQ2w#-x)|;1iQEksJpEAAC
zv;E06>$}<+{<V9tf_u$Rjmvlb+x0+&M&+y+hHypNh#;Ag&{I|06Pf3`&AC1lTuyod
zeJw@`?<hnZi~PZY@pri50ct4eEYJXJQ>d2_Uy88YFSo;A^xc-@88s6E*IryBtx7xX
zYY%m8mo;!tTRbgFen(#h-V3sRn`mKaRVBIJhE$@>8@22~#u^|(r4ya&9IXh$j``pz
zGlZkt+$5e?`5V%H%m~Q;h$mC}%?R~sp=BUm;aldkq-9QkY38*e;aMN$$zCD4R<?a}
zc}|M%cjH8D^PI}IEk&A5#KV)Dz6<mFsyPk4PsV|Y6RZaH=8=(DdGC~i8j@i^c0HLZ
z4qwxc_=0#&Hox~_S>%T(EgKA+n1OAvYm>Z3nZ=j``}AP;Ku_Nc-~z%VSC~wEt6+tX
zl*M&6+kEy<9a<00Z=FU!IoL~$l)q;)f5M^wgSg0fjDH+Y_hvg`^I=U6Ki1%_4^=;C
zr{bZlgl}dJF5hD=iCY{JC3IdiHBYCFAZ-`<^1;{!ON_OtE?}HITI_L{gIZ<ig$|Q|
zCz83k)z(sl^;Pjd&WJO~NRQr&E<Rz9St4p4h4D1|4b{4S{Zqfa@l$LMZW8Us1Ra$V
zD;jxNVzvS?)6C+!9tvX(XP0C^EOo;tx}|=>UX2-U^KI>jO`wDjJpVp)4|UbM;oa6D
z2fYWgdB*1H`i0?>w8hsJCu?`V9%_Lo2;w7Z?A>b|<>2kzh*<GbIDz&0P_xUnA)Pte
z@&T8+7{4LCykTIR&mqB{n!UpI0+TsVc<z18d_bS>!T3zv!aENXpC`0#&5zHgKWjV}
zPxnn3oL3j=Y#C97p3jyJ=hhtmO6mOYIY(5$ZNf@AMndiB3li~oYOM`1ptqh#NzM_X
zH}mJu6OCGGbR*hW_E>hPMt^xM9gXrCT*Mo7HilTgpncR<I__RGw8;#zQWCn3X)`*x
z)#>tU)%9;$Jm^F?s^(ttq9QpFRuU)<xE&&zd;D9QUTi7gv1-kKz1WA&E*uTn00hJg
zV6<yxY`wMl6&zBedMefV4vC5tfUukAS^+F-Fd#hppx`kGNjf_XxaFT4Byy~p@wzB)
zJ@5>?i}pD$PCqtpgl9w%euz05Q7b9KjuX~|oq+4QXkRF6KW**iG<-~r26Briz~L@v
zpLyY>P7B(Pk@PmK2lzXW8n|E$Y3TpOZ7t#|P#lH-W_5Z~>G4cF<+9*x0gcH}F}$sy
zXs9&)Xx$$GqECF{yoe)mk#@<9xz1Fh9r0uLVv=5BF<2|$w?!yd+E>Q_V{i<}w|-s9
z6{b4dys0@_D;!uB6#z|5`$=Iart(g7rWcUjgSZFGB5de`rilp9NG*ORc$LbSr{sFP
z3?YB{ne~NB4tbzXvGPZIh3EnU-~|U^oi4KDaNlN7Gu#dx{CUKkBGaqw?8HOItaZXN
z@lV*urElY%Zf?B>;h?2LDgu7vL&nHVX?7$yK7Zm&pl)!`X4PHCm84^$*@{*OSG0EF
z?3uMizVB(7?n_cBMgs9kJ*VmYtCZzXo+XFNHeq~#?g|ZHm%irUdIyp5uPU#FyA45<
zaZuCsz0xj7T-}#4cPm>VBb`E&_SP;qMJ$~?l8pf)AIfj6>B#7Y9B4Fghw|Lge6Hsf
z>i6?!QjN2Ied&SWJ8DGe-tuECLh+H}D@v!%=)XNa*pjObF(<k+$^Ti0(#)o4s51&@
z<q+o$d$r-G`yfp_4c>mIo;-Qs`NPU^9PG{h_g%+jv+Xc&H@oy}p`p*;cXY}5FTcH|
zU;=~$IlvgW!arRG*%V=gE{;yIM-i*v96ZL68SM3vogDJM{8oFHCfqCs8=#rZ?nR95
z{Hy1NPpSp}E1Du(T+6Bj(e+e3u|8@J@?%DqmhH3Zxc{AD(J+l4yNOA$pIupNJ01im
z0^CYShbSb4UQp|>Z6Y?<4ahc`g4Vle7=$gk`cUlgbhnmFI!;5ILrRIH*bP=4y5RG$
za`=AW=4n9kLR}VJe<keJ8gCrlilu!gfyOzk_*ZVlIT%rq1noe^v=EA|-Tj#&?yUd=
zX;12Hjy%&KDlyy<mg$6SmQghi{mu1A;`n#<HX%3TDm5CCO~@WIX*7O{gz7C74!@Mk
z<({rpxG6o`vO_E01XMLsm1bQxO`9%k*K)4Xek`-fte(de?g~J0)<z+}+;y%|Rbz}R
zVx@~sr-Mhrwz);VH?8?PEXCI?&`88lOY!cb!g+<Z;big1o!PdUN^4|kGU?XFV-<a@
z=+qu-&|4N|<?F0wW$wQ4sLjs#V?+ADMX`&IM6v5V+O6PpgG1D~bKUx?`of7mvr4VO
z`SnOx5U5_Df<P}7D4ZXQ*%}|!GJ5E%uxT6Oh?5yRgSf|7GX|UBEI<qncHRT=41iZ2
z2Rrl`qsat^7Lpd;4OVo%$rqYCdn*<`TRK2wj3BmCJ~?kzomx)Ch~p9lDB6irm@gi-
zZ3P=+tk|fMje@4il*fM`!vtQCTx<3oR$c8Y&1T1Dzx#PyV!X3oD8)RFP5boL$Kt$6
zU?2jlMEZVC=yju_;L?3B-Hv%Sv$9rcnE4RHqIedUSMg3O>8RB*w`gA9^}7B(*Xu&Z
zRH07%gN6s%0a7I<vI*+ocySA(_}%>=yHR4hD)+omx64(e&Hi){hgJrPeAddmK(d=0
zyirbbhF86E@M3GSlg_8u@S>mSEl4gQBtS}3PLqg~2Bm6DwWHP50IpB+HH(oaZ55&B
zbQDB)pCZ72hBQ!>YZ^CKj|vHS)^PYV&>pvnTN~*t*Gi|pa%9NNmp6BLL4GHHG76rU
z&GvOGVz?^yJL)x2ErAp@9p;Doi67TJOwj0%8xsdkXC5>y?4;i<{27V>3L__rsxTob
z@N@+|JLpr9To0m)F$Gvmd^L%x&7)XH0L|LWAcWdsBYn@uSRdQu5iK={OFl{sn0dc>
zpMo8y(NJYZGZsXa&tL{0ig&&atD8^$JV-(Q$@OP@KP5!yVKF4w6Yy)tJ%-@!U7Q4k
zces9hkgtJ|tuO->isvAjS(OegMW5>ip3f23F*H7O@7?IkHQkENH15ecfwe$FJM4Px
zq;2u+oYco9{H#7YyJ*5>VfP?}Wpu-J>{*6F2dN!o7y_AP7FN#v^s67>A4K;k_8p&V
zi^MvL^hqR&YYg5T^!i!w?~QkCxXLUpx&4)2q$WDvGqP~dXLuSTUF&Kb{A`r=eT7~&
z#Adly11{4oj50lwQ#N;cWF2k1^2^1HF(Na+j4-#0&Jr;{zT7JEejHaTD(|RZ8UkUy
zG-|Az7oIDT8Ge^4ZVg!Jn=pXGZ@^u!cXM5#S?ulhX686o7^KN%zp=;>r>6g72U#c?
zm^V9U`i}V9oOo;1Z=;%C_=V8V?1#Yh!&o|a;RhJVaZ1;lQ#!%y7g4+&jvc{0ao(W<
zBD7VDS_-=?P2lf*Oi(gxwyjm6hD%^k%i1pjoK=+wkI_N2d!A#5dn$#hUH$zv#!Rwf
zU%V=}y)XS3yqyyXU8rj&{^Hv0-IRwjYa4m_yuY%!{p1G|67!GTZTESSn#eCZOCcpA
zj&PQOZ;U{{33<c0&^cgSh0VUY5P45B({;cLp~U_Y7|5Wkah}cLmP=N0*N!ao4lf$;
zx?S0aJ8m>>PT&DtsJ4!=ak3_m+J9OcT{hSvw8myW3^Qj|i=+@bzGi)mgJy%}FLXM7
zhcx6Bay7q)5BR+)ra)5Wy7Yz_o8^#~(2u!Ne$MY!$0$*8sjWUEK89io)+=j6<_u|p
zT__yxQQB`%z=eno#5U0{$3IxA!^+m3K0@SRa>tx_S>>ujcZ0Y_XOMUFNj0WPrdL?B
z=7&(SKqCBl^%>h8ZsTW<7!|MFc#nLySmFA3-y;4szcek{xTmiINIInwmnDrpSm7T1
zZ4rUv_KjbdXC|*^!?|n*{Awm?$1P7fWLBG3gB!&zF+dkE!n3iy4DP;>zefb5v^TAc
z?%9qLDly(siefLNJeO}PF)B<?xi~4skKe{pY=X|Y#A+LNwciA!UEh?AR2O?nmQ0ya
zcgND6_5BF{gdYHApI}CfGx4BU54Q_HF8hdu0VC|XqfhVY$V4q&@rBp#86{68+aU@4
zQKXN-^uR8T%$U80x-4{U2bcL9tvp__X@S$1%;5F*QPCC=CgoD3j1ShPrHZxH)zx6I
zFQ|n!x%}v7a@eeBt$;taLe})m1S>-en+z>_IczUma^(pR)+8<yOaLNeySHYu>?7^R
zybtaE^dBrLk8-6(zJ0<euk<>6#EqC1us4DG+4?<WgljC*heCtt9SAJD_;Za>zBgA~
zR8%vG`&@H>#B4sMyuK`{0a8rT$~ZbXm=f#((qf9{&wCg*$y07WYV;qWvr;()V;s-0
z^s30oQEEGtV!9V${DFq|RUE(LSfTq{(}kd;JxWK|OhkV?mg+vjObGE)`B8&K(N`=C
z8rfWe1Fy4u*L;gC9e$rtC*v3ArEhjUC!2%}BaB#C89fkxsJ-rD{>AaFL$6{Mr;Pvo
zWPEw`+A}V6XH5f+FM@~?Q{eiP@QN`mRknMQtC*6<jPUG`0q5*6q$RTYi&>W@NXNS1
z-+SiM_|@oCC$I|#PE99qQwW2t;HH_R<_#FP%cS3)O<Sd9#Z6jztk<i0;PB7MDUAAs
z>~swz{w4JR;!;&Dp>L#_v1=P|xzp>W4eM{D<=p%k7SWTQNTwC#?rlEAJg_#?Q`rt<
z?<QELkNW&A1?ks11bPGs6c%$dzGr=kgV!aMS`#bBkj@^Y!*6NUPoq1Q-LB{C(qwEM
z*K)W+Fk5r?dCR=X5zFF0e950u0`2#HOKyu}X_kpw)Hyb=_d&t|=lL{^BzOO_KCR&r
zHfEK4SDH}p*NN>B?>pL27%xX1K5K%H!!3zB_AyfC7(2|%Q#ZqC(Ctq+!9}|{l6DC3
z_RgtXb`!#}L+x8cJYcT^VgF_><t@GT<%&RJecA{Qnl5g3(1rXR{$rg&w_e_;p$wh`
zu66s>)rg}9FvxLuUI+yTJC3v5D6;t5ByKUeXU}#%vvp!zbCo1|{$v`84xjd?QNwGW
zY0vIV79!j68~u)m`-T8BQhapF+Weo+4ndGEcg`nAkZ1k=`%0!l-dBquitKzZ&U^=!
zzHF*$F5lsI#kNHc*RToT6yBzel)a?%Ut0=u26yD{WTY>O54TZ_!JvO91>!pm`Lfrf
z(46uu^s{q7Z<>o3dg%<m4X;CAqg+ccL$l<54(q*p<b7VDo6+>kie+^k!)Qh8G|F|)
zjtp}!t(z5uOUZWH<9=_AJ;+9cvWU9X{K0<-;qjZ92t=A>dI!Ut-D-DYWT0M`wNcRj
z1B5SIA-gJa<|<Zmy6`4zV1XV#ZK(i+lIF|=E8HWCF%$mI-(^KcPUxF8G-`gw(u>Q}
zOHM5;59gdzZMz<xW8iaJk6EgKPv8Fe2z~>v`;ND1BaCS8eJE4FxXLGNarTF_mJS{J
z)@@vyST0R1wXIcjoI-iUAh>8;W>tvrOGn#o?!)W{hyY{sh)@#%R7B6=S0h0?d)(q(
zxpYIKS9!f)q@<9;C&8$L6kTu+T@RGg<@(94_%8_09oDi=%1XlNZyFao99@!fxi4My
z>rea93$(Yj&BoU}wQA1RG~DRgtaYwcuQvWIQFT~kXN}t2zIfWV+^JZ5*^s=Roq?pm
z2&V*=Akv-0Rk_dWNF6s`Dk?gRDB+*yJ>OY!IX(jg_q)svKUS`IuE^{)g=GnNCuu60
zX<Yx>d{GH}=w99mNYUbaI_fW5pdjPdKIzB=`5t=0-AF)Bk@8;tECAm(mAfw(=pA=G
z)Yue4kCEsXBisGF(|s=XEw`RiiRKzYecdq)?}mV39F&+pzw*l2w;yx!o!C@Xm+(pG
z7}J)VKrhF>Smt_5uK3>yx<DrUIy3Sy<>dipDWNwnf#H7I3P<vQ`q$7%fBl!UtN1{{
zdq;_^=1tP~fZ#cc^o5xa*v7j>NMLaqvz|kTY_3?-UC#mN*1%I;&!Mi>r%%`R{XGvS
zz)6S6ZO5>H|D;Vy@C*!F;i=KWmeKz?xO^!08v-2DjG8Ju9z3jz1J(7~4T+=H>uK0Y
zdJrmf(L*2cgMT+MP_V?I^%}j&Y3fgsPk4ZnBgD}6q-OPEhk5R3@dJE<6*ky3)JLPL
z`Xt5Yx65))(v7V9MdZ+)z_FVj81npZQv4$s37*`4(ESQ}-u%M2vi9@iAnt9r{qvy0
z9Lz(vR5eZdx7Il47UK=d<Jb$-{1c_X`)J)CPMF^%d6YzwU|Nl=vX)F9Jj*_IIe)l*
z)4FE&N4ypGlT|3}S6_pmlhUoOvcTWo@mc)0I)WDdQ6V{=t15RZVN1ved>@>LUm}y!
z4;}AP4-!2L#RgX$+K(%m69P=`ej}?4PSXZ<ozuF%z+#Nm1uEEt>2Dn%8zE}10K$@3
zM2jB`P?ku0>Xf6#!|f>@5op(n4H(V!vyc~!`c4}UQ5OX*Q!;KIjZHsM?Sc4HGMY_?
ziZ+q=`o)wbTM;p)ExY*rM~<TEXDo?T7i7nZQ+MJ-VL&LKTV%$z`u-rcZ3s`P<SOgI
zS?oaS4*ogdN3Gso3>ID)qT<-%Z~={588^B8dWT}ZRbL>r5n3nSU83~Bq_@6k?Lu()
zelW^WrIz`HTFAO#3{#50zO_m@hJvJrNLM_W`}zwvMeFp|sZeJ64}K~B^1?BcGbLOs
z8O%#OiSKAJ;9E-tW-Z-ML^$+CJ?JLp%nN_xlFS0=NxhRB^xcqB;2MjMm47SYh6epK
zG+Cx+%us4VMdUG-DCi(g4e8)d`z?LhoIr`rSRV?UM$1kf6i3>+`B<ft(f*zD(0HhY
ze%)Sb?NeOKw)WjSO$dFh|2I~X%-dKEE`K@g;DvMMV>dX;x1NMjoKgD5{3*s3E3%z?
zEzM?PZ$#p%{ism0OA${;$@@c<DAazq)z55`WV*J-m~Beq!B^@j4Psk!1V|^dJMtfN
zYkP%PI?dEeZ%p7%bK0`-*R0UqQHizM50Kw5xJRPAl=_`Hd4w*<1=LBOXjQ8#HqM-N
zs^J0b?g&7Ww~Df@Dx*%+DU4bN+~eNjdR}%Q#xs9Vb$_iUSV=VbEOFd>jC0zL*|GIK
z5*}~}pK;u_$>nOqCWuUgY^%9mC_1(3jC7Dmr2S!~g<xV9J^)#pQsTYy4PQEcm3}LV
z5KsZplK+BM;kmUV$LY*-i<<kX&3kDdY`Z9Kda3%*ertebp8Q2A4EV$KZ~Dhrf!w5c
zUcPB`bcJ7jXbM)0+-;NdP)I1UAz!YO&V-(k1T_(jhTdm+CwqA0D8&^D`hhWA;hsq3
zSVtAicqEu<^K>kbWOVJ2TMp&gkRYdXe~o1h0wDe=hIO1&4@pqtAB%C<LpV?W1ii}D
zVL!-j5sl|wK4%#f9K-{QK8JP3yOj`couikVHF=|D=-u>{qbYmuU^Rtsn@+e@QIS(G
zi?Adq-W}!fF2pQj3$>MVfh|-3j4|5s(V;n#?{58zc&0#E(MMbgZP{PFjIx*64sALw
zW?J#p#GD}#8_%H-KUMp-ecFNg4964f?H7f>cAb@rTRo5{f&O07ZhF_11dQ;!bK&0F
zx1-;lA=iBiaeVp;*v2EfxX@cd6Fd);O*SAcXG=Iy?)I+V2Ic?~M=a18Mc!QoE5q^a
zOrBxwv-cY80r<94>$q-Nk*2YZ%6-iAv8DT^;ePDpv0A)5>hqtfnK&(5B+L!h6<}13
z&pk1h{XNcgDwR+v44Ki7&;o7Y&9BjL2}E#cRxfknkzEiW-*OVq--a+k=DawEo~zz1
z_hUxSQm|SfAF<Va_1fD+PNHEND`yS7uQBEV9y;_+GA>%k_Vew8_ZwZm;M&arg(tlr
zp2sEV0>)(dMqUds<ul1jg4F>|KjXL1tGs<K<~OCDQ6mz-y(LB=Auw@wqezccEhkF{
z-^GM$0HgK9BCBZKkJW9*WS*&4fge|2L-<0zzbICHyXCb3>6IZT;eLO_4w&`qO**b;
z5FUx$w|3{2SKBmXcpeZ?rlVQA^aW`E1jxK8&WAOZZ`=pVhn9Od)(95Tjh;!alK_l7
zxHE`_zn7c)JeJEa0?$Oa{O=pb-RL3Rr2hH_uE*Clb5T9#c9a2IRAlOUj!zz!4M*OK
zEiZCE{hz(ve`|G94_T$bYKjMu8F-zy%ynoPv8cuU5cs&!LXRdZ(J31j;Eb|=LH)q*
z_GPVLuTt_(vOH2E@z4((Lqv;%XR82by?JDiKN#;(*eN!OxVN|bE?zmONe?af>%bQj
zur+)<6X|$w5PTPInL_?nW=f@gj}5%_-on3Ds8)sq7hMvSV^qUEsbgnmu|IgWMZL%X
zRi{0loJGx@^H;XD8~$!_{0WJxFfl>Gd@rU^tFMCT%G=<0uEZxKNAT#{3OJV6Pa%x)
zA1xX)6Wa0%!%^yj@60x*caZUi-7fuYjfY<-64u?$0`c287Ms=_ULUqyl4%tkMF&UP
zSna1Z83x32I-<bTX`e8|6yei`o3050vs=62zNdH$9rVJx#t=K+Nz_GP<^G*V_{R3Q
zKXit<#SJh9OL~4Gwq_Ul;@LWhx6SfNQQ|uZIV0%fg`CrKT5I;oYz=Hj!BNe4s9nfs
zL{KmRMi5t=m8OUK+iXwz(qsMBAWqahlkhQBbJ%e$l&lAwpjj0??vW)8eg-^_*h>8Q
zZ7jm{rgnon<TZPz!Bo6!Ng#q(NLY|qNzP0?N_=(7b84C~{Mzd(R@0RbfD5)S<0P~0
z(u65@LAb7li8#8+{*Cko$ic!09}v+e$&-G&UylU&a9kQpEb<NsdKa}6qiCbO7GvgC
z=Qe{K6HW$(UDP#V>vZ#YOQ!8i$y^Z-iv+}KJ^W@9lsD-Ua}U8j)D)u{Wy3+Rw{-3P
zo*dz?ijcRSf2O*ulGYfDAjFIQ97#O%SYi;=eBX$T(1;ErTp~YQc|~#1-?7=yJNo{E
zE785)MSTt<<Y9-YrFQZBa6=G!r)U&<a}*O4w$rWbR{H@?nb}=B@W=GCmfZSo$K?dh
zX7H!46eFyUx0iXsq9iWEL!iPNmJN}M?GS85T-dM$wZ4q_v*5|L`H*tGb%gwiVX(aQ
zwlyQVTu`3bR$Ly<w&$Z=>tEV~VOD>`DvBh<Ag+nL=L;?OTj5hY7@t&KwW>g&yWrdR
zOt3L#pPa7e*0kbZc4ewN*q#OI&Uj8TsK%HCbNwzDuGha{9Kt^d`JU?BJk9GlY6arh
z62eIme}1yawy&`GX=rdvt4XBI|KnsS?@5#~?&FF%)vemd>1n{J)Inl683TtI?BkCp
z>>x5GTkc;@Vs(B#$<0+xc=#=A0pax0ZL>nBi)yem+M=K0Ge20`pb+qzE$vz^yr}BE
z;0I1GKa)6ON5@>xqVwDBN(cwuZ63~@v>&UBx?-;d7&>p?&UM}jYG2tTN50q^{U4gn
zGOo$@egAYSpdcYJk#3Q$O+ckv=@bx=lI{UgBHhiTQM!AyAl=;!qiY)s_TT6GfBl~B
z$?p5XbzbLr9`ED0k>B-!?bZAJq$$*mtGDFWSIc|u5^%r0imwjUWmPL3=zGBR76AQW
z?w_jWvbXC`{a96M9`5D07E}y^yOL#D^H1zr0TC`xJ-meazUOwS*^PRzvU`HNS}*_u
zFx)cvjbnepZYWm$6|A6b@zD)-?D?m8to>KR({E0D*$q?v!&D^CO%UZn(?b}3x=Kc=
zxCy3DK@t?`g25T!SS9o3bpFqI5(t?RfSj{BSf6e@Z^s^Y+-A3Ey^(_$X>`_B0{pG$
ze)RFL?&So?8mU@b|E2ZT=Cy3;eZ5`ov{1v!1;Fe&3_P>zL#daP1D1cWN`pAzg{h{w
zreNeVKC7B>2(#*4T7Jy91$L+u>8!bj)rDC9J~3T=_rcv-$NRwgIk#7MAE^lLj%$l0
z_d9WwLq78MpFHJxQ$~W#-{^O#`@FZNf1}O2lGG0i`F~jer7HX#%ALzQY>)ifuV-~&
zWvvZ9{h~|zV2mLUj&46V{iHVi8;fdc*~}4D+G5c8We(|80i=GHB-Bv)eqQw4p9FY0
zBAwo(>4=3DuV*SNjkmgO5H}HX<n_rW(fKR_|8?<vV2s8(yI;nnNZt4VE($?ye3pm*
z8l(-)%SYwnLpa;|^M?K-xOMzX!?CCUu`g(h_=TE_&WX*uXsjM{KMm=?;ZByI!v$7x
z*L}I-d!?4a_U6ZV+4_Obm(~foetm*Y2&bi~-edCDw)x`DvAOvSl>*%LheTYBa>p9C
zA8|pGjz^mBkzZTmzIS^L$COG}P1*LsF%x}mzSG69*36_>y6cbz+qZIE#WmgGC1mK|
z-?1GO<ms{3@ecUmg=!NSqsm!=5@s>=H~#?45`PI@K1Gw7+qP0@ynN*HqqM0Hf?Nd-
zj^LXvpZ3ajNtt*YE~(=ljcREHu2;;EyU@D4{~^|l%#@aM_yn!R`o~lHMG`mU84HJi
zKC%n&b?4|dAJxE@E$3=snuB|9GM?Z<zBm8c?Kc&I`25~5*}VXnkt@#!$-TKcFVEd3
z4Je~Roc|LxHSNiQRqIZhw-6SbU$dC$101R*w|uY=iDSq>a!Pf0Yd3)yV642zEbOn)
zZCPD3Uy*7vgb7GUJYVo>3{;UbT5vkp^S9&a!$G3FKmGH`KmcQoNP#WrH;qCbOLz<a
z$XDJ*e%VPlBk?KDXUEliq!fnQ<@%4l{ZC=VJY98{WTS8vcSDdlaqCHo%C}mvIo)Nq
zsS%g%dD2fw-%_&->*E^`3$9}_+g9aW(L?(r6aApOBV+f|b~r4tcNz8P-Xyhi%q|b{
zDBWpMb&>!abN`pi#Y)=+mB38!^V|~Z5wiH{awIDC7G{Pn0wZA{Do?d@i^V#w!zi5?
z0WDqhiD!UjK*x>;A0lIbHzW3bG2Z3TH2Yr_5WxRRFtl)3dm($EV)A;{L?PP?r57ql
z_MM1VyrTWa=Fkl7Q^(3LcY$Fa^Du;y3flupAA=n5Px(o3pFr;bPfH=cUFv3+l&Ky<
z?ruJUG-E+G`gCXcG`cT^TU%Mv)|+VdLMj95hrb5r*BicLe%ykXg4`SP;+nFN9t^`u
zHT*X7wR3t^)qMv6gI*XF_;Bk_6o7=DTGA;1mN{b)?my{811U0P2qHYL3Y-60%BK=Y
zf@9``oO!PL>IwF~ICRKErFZk-Cr^d8KpV=?*#@ZlsvqE)6ew7@KE~I$T+}7DdB=ZV
zdJ_Y{iqPrswok7g{DuW`3RudVd18jDTmAiI{|!*zSNh?UqmHKkhyTe`c*$|!Vy_3<
zdsE3hJ`m)8Zw%$WY~u9$rYi0o<CDp%ki>|-c4=94yDZ1SdQip<DS*#5+84%18xPJ)
za?)Zo<pfMTyVaNLrd3!&E$We_ZXGIjNp}aF5`whgL}&f;8?>$DjDEj-_D-ZyuPG-n
zD`586m>4p;H8Z`^N@&LUq{gI3XDkxS1k15$>@$so;qkaU@DuPE#M%=zP@$1zO7KHo
z(=Z_TvGAMEQ&?f-kXO=yZTbTNo0fn{1#W7gIGPxzw!_x7R3Qr$Y=zA_54S%$>XPan
zjj7y};tx+i>`~$PM4!)<0pGfY>77uex0$#obB`Fg0?y@UB$Q7RHpUvxzNX5+RvN|6
z`*JIc$yWL#dgP=U?H#jnu|NfV)tsC{`sVv9_1aj}IPU`MC<yHE=MM7ArKP-x6?`})
zjZ^EtOrVyq^A+<|sQQMa&10M<M)StXgN<U|-)H~Fn-|}vTpMf`U`sR@f8P6gqFi<F
zCH5LeA^6L<5w)*A^)FRU$@8r=Mm_Mv4n=4IjHqASa(SK~4^_nHA(WizK>+pSlAkqs
z$^QJ3BIVhzgy?`hv~}#iXT#S3w57ExUCzFXc`4qTKT`CvQ597|jVTcdr!vpw^Z-r8
zdjiE|(AT0r*Ay>B*Sx}YffCKS!jJw@1HZ)!$_%+mYkVQQ9tllaDlU7(HAzH%Hgzfi
zo=pWSb{S})Q>5h`uoc^S7werajo&<9Br(-{$w9nAo63;rC|dt5snYJ>b}9LV-lu=Y
z+`)1*#`T@C<WVnW1e|hLWM0Q<cq=`h5tkt!XzQA4{LL<`5n}bZ@loy#!)Uej;JVKI
z<%krAoZXmT<_x*2sdo;`2*#K3d-7H8jj9puhYZ3Ek+DR|NBaQ{q@(UnhY3r<{{AHY
zCQBWJP6=vH5PdCMY#$~JGxj%&axY`GuEYMI^Z9lRb~e$uOC9qgb5oz~X`}tI9Gc3<
zJX<LPt7CwuO>a{US@^Fpxxgtr6rN9N>56TqoK~x1wyo<aL}nYZE1G$_<c+j)@I1-&
zC)RTz&tiSL>0EGX1Piy)lu_`WEh^ApJbqkJ!1!DU$TO6I>q_<OJ^z261$G>ecgcsl
zSM_*c)q`<jfcN&s*Y@i{jdwmL&3iNHHaVriH<2gFf?WZS+i8wxa}w*XM4L8$^2|(W
zfbs+8ZZ%ABKmtFX0<ik_G-(3~gt!(v$09hgD@?)l(^MpHg7Vw!cVM0sw_w?H`$B$O
z|68>GhvOC%7x?4;P80uY_;yTEGn*81<C2G+J7N!H{Z5s=m{b6j0`2c#y}o}(+}Iw{
zxS}G^UcldaBxM-yfYN$Z*XUUiISP_@7w;#a2y1;i+PbbC`V4E=NP*#j&cE?zga)C+
zeHrP8#=R1;ZUmkW>PZdgVS;?n*-`4Uw%20hTUR5r+!TD!_cHBmpyz=9d%*2@7n=q0
z9JcnLbh3?9Nk`Wm=*%E(Al%nH^|UY6rLj|>cPpn6$+J)S6d8Q>Ic9dL-}IZjnI2-e
z{(J7@Itpwq;8P+DB>OO(oz)buH>BW~^5R|i=KW@*0HxvRoE&cJUwb{iGx_KJe%IgX
zll8S0DrPaz!pma6%e2?9JR`?ed6Ub0*iqT_6Sw71%aeo+mc)mhHK?)-XNpAdpnX>*
zKGvY?G<}LRKoa54h8#z$OI|b7zs>o}Ov@Cote0f>+hYxvT<Kv=sO!oVL)TMaU6dra
zTx{&3Th%Pi-=|08c8VV$Di2z7qafZGRy9Q82&0aycqpa)iI899KNCWz{dQQbj>dr1
zAoI4LV^7!n5hUcYO%_q2L)G3`R`ewLUt9YD=*E`a48W49AramV2tMqTVnAb|24BC2
zr-tkic29Dxp<1A&!}mnyH^Bgu;u>8@HyW4lq>^m@RHo0o9j5>_7jXZkI<nJaHS`xA
z)+3ADt(J91R<U2FjpIMx^|J2qnwe>$_OVw28MRNxH-o;(<AW7(`x)e=UiVkgRz2gY
zc#+*pM0(&?DIT+}lEaL>mSdy(AT8hT*3>o;Gc|DfKPS~6yA`W=T`;-o4~4i$zSM>0
zqP~fZg38)B1%$PQO|`GmbmhRAK}nphVm!lhfr3-qCyI6h7=Z{I-FeZuqNH&P?Dy;a
z1j+6Z9dqspqh9IqX(6TpPmO+NCOdYZ*<=Go5>QqeuJ;qZo*<V~$VMcqe9C6lRcUB5
zADsCyhUvi!oB2sN=L(P4-FtLzm*!#&7ii`gPijMmxz^-iIpehF$)UGid!txpB#hD2
z+CHIPmWtI!+&7A;f`scKPleS)zt7n&eAvW#+H9UORwJ&DmXp`REP=|ZC(9!Ju1X4@
zj;}sD{n$Wq?ANc=mA?<*Kjn=pM$krATfZK|=x2-)4ROv}lI>HbqHF~tN{(6Mr}4XX
zxh{|T0%or!6R}kd>vxj}$dmO^=4?c<ZYCwhxzLS^^n|N*mTviVwln(FJ;Yzpb8N6&
zrlH)MDDo;fUGsjKOtMRR0$wiP4TT?0p}SAbo+sZHmUHEDc#<{q3xg1Ykh46O=e6tv
zkG7@e6gt=Z=ta43{lh)#v&o6}d|g;=k<WwFYJqmBxZy)Xy^J>bOD(li^VjJ!>*;-S
z19?^LEv#Z&3Xu06$Mm$zye3bp1_2Hu_g|Ck@&1-$yaykPUN_`_B*DNB1nXnx{Q!<)
z%@NLs?_?)OwR<Idre3m=p^sGQT#XN07Hv5Wc%IKU`5zRn>1s%yTPd{}K9pc4R3wdA
z{EEuX9w75)?Io#o&|_4V&!WL$bLw^pTW-GdvaHLWqo~G}wO+{$j=B39w;;|o$By<J
zP2bb@RQ;CWxi{S`u_%?(2kdMy+q~gj=!>P?H#^_iv)V0f=S(~_A;7CG56%10?3a~o
z7s?X=e0deSOahUkt!mXkjBp`#%fxu;i($-ZI@z(!s0Z)G<{q{Q+&dTqZ3NtoTtbq<
zRk^3<>VGjqFnWw(XVK7G2_v5rj?_s^+o$E?XLP_SqDP^6Lbs_Xu%wnFtwgoR5rH&p
zekDi09B|s+$MX1Q-cqdo6zU7xC*7{O1DM{qUZ1f9&uS4y;N1BWvaYEyZDZDH#Nyj0
zdF%R3D*Vz$0DtT5OFUK_-U~j4FdP&E@r%Z5qvNu#;Qq19MxR9ee!uFHGG-xmQeNAr
z`dBV3^Cg4g>|eS@4h`p%W36cK%9dMS(0FI8R>C$m7|TRGvAS;7&@QRKv)=@+=u>D;
zfbDmTPjRD$9nsYBPyD1~V_tGes{$i%=f%3$-g~e4R(k%41tVU}NzH7AcLLA@gwv*{
ztNE-<$dEt)j?0G)Swcqrr{Mt8^?s82g(7;p^&!rg;bMA6L~@$_stx$C_@FT`N#Awh
z<XyAS{-jF9ILX^3K-ckr0HES%i)#e#d(y0Nf02qVFJIGiF+U!dND}A${&3b$f88YH
z;ItX)v5M!(g58VP>7Dp_mU6Y%k+ZW<I!16PkF{>R;@MfEZpu$t2Boq|@qyJz5fSlS
zqaWcRJ8_$>v;BDm6E>_J*8dsKepiUd^>bSvYl^z0XT#z=v<Ah`j?HH}W7$&EQTB8L
zi1WG~<|T;d+Nd1L4jUCmMxqx8@R-YBgfD`2$d#A-x{V0MY?27b`10hHk0<Am7-sZr
z%MKS|_9gYq3VVuOvOl&cbWCsSmPBN-t=!O>og`n`SWAm@!fzJe7S&CN6~}ZRa?@z?
zgC=BmRMQIp_%#aMBJlsIJQy+=+tN};g8dHalQ8;?yIt<5sKD7fi(VvE9OCXvScfz8
zHk2x=_nc9lNVE4mse-}oV&Far`~@IHm%$_V%R|wz;gUSMylh0rF`#<ts-31k<yx>{
z+2+3E5lql|06~d&l^b*NWp>bJXmD7$<<A_Z%zS+Innk-rMjlJ$y-U|v87?<A*dA+2
zUAWGG#mYN%M&P(^1#XA~sxK<$visrmnB&ay;zovUuQ>=XQaKnKql&OVN)H5nR)rl*
zq_Z5?Xc7R<C~NphL&-;MbCJ*P!<}SKmma<Kp4w$;1+v`d{AY(Pp5O(i+_1j*gfx-#
z{%QoW2qQ)0hGp8<RPPUWXnO$eF-zHkQ4soNy&uwL;#36Y3nK?2UX;tDH^?hLdM-Q9
zRZ$FJMY|KDhW$5o)_2-9<#}*>Lomaey41bB2>hKy%J3&g*w6}daW?AuDGm-bY7D)~
z3%5W)mY=<Ry9Y|QXkHi6YI@9W&O<8>JgRZZ&M((F!h0O|hIdq6RQ1u0K+2<vyJM$-
z%7^n<smN8^PVy9m;v7MRwV_Whzx9(~b(U6+=#k5yN~UFFY@XT(=^A+zs*jUyjCxET
zo1FTjE{;h^M!yO`U84V(J-|;@CWW0r@MaCzs6F`dZEgTcY>cz|wiAShGEe??(@gUl
z6ZaDtnTwHn%_dhAjkZu*CwJE}`~^Yz{!p$ICjIbv$7+52Y}ZkSc37{>=nru$_qyO?
zYF_<cyTjN8%BqgXq4MpxAP8cfaHe&wKGTFz&6h@15L^7@Bsw#B7G3XsIjvNj#dZH7
z#&_yV?IXI?P^7R(s?q24x+b1IHwpd`e4&btjk&}3bQ%>)w&4yldq5rVFvQnyEn*Gv
zHQky*_bo?`O9OPNU(J){0Cz1({O|Iz-&hTT?#VhKv&!uXA=$D<nZ2^C@_zKet_sfX
z{3sFdbS1#)ItYI16p)ZD(f)NbH1!V~DNPdMa_g+8<CnO~w7cr_Sect9XaOa&{cKtK
zr-Qm+eOCSt<Y<^+y_{E$Z~AN7U-r|peCpq)!|`)ra>`_4#=B0rbg6I`DkUZ?A}IdY
z6`xBf4)~{b^+)YRE%0PN#WzYAV3)XzGDD^8g&Oav`b2lXau6W#^Oq-mpO$xx*rl!I
zxsF=;T-H7`GJwQqjsHh?9vVs)ycuYj>Kw54sNhx9(qD8*zc=X^YWft#oX=_*eV#gC
z&5K1B+)%sodTFiEZY_D}=&UXLuYStKq#6fIY|t!5DJT_#K?(4qXycX?8Z)p#a=QD&
zCnK^<sb|d|3qiYkY@^P8x*PI?6LFhL3dye#oibEN2DUBNVM=H-&bwFsmb^`vP6TN8
zd_Rr&#^q6JZG&Py+Kx2!5EyMNXyh^4dciu&liU3BVgGyaRU$GnWtRXf_cK54^>{y_
z+@*x_>jz|(?jM#n`s%^PSo1NL*er(C92WBCRf!>Jjx9!DsrB=1{~nv<!?-EDWvK>5
z*0%!Ck8(oYTQ9nXyPee{G%jeo$<inHG&Ng2nGQS%&nzDc;VG<zqBSdo$Db^?%uZi0
zp4d98R?AuPkP4F+kfgh<6w?_s1P~!)!P7&a-(qcR?osl8p^k~U@lC7;n9{<QKDVz>
ze{5iTBdme}ddhKYF7qC@!`L=(fp*)&tshAI*$wEc-@j~p>-a+EblEtxPvg(Bxn5N_
zoT4PohF33b#{63@UH))m(RK)hC_%G&aK*{4vQjUHzy#EA)PwRLQtHw64@^d0lzc+6
z<L9%L`?r#1nf?_YD|pUE6dD|-Xq@U$H5lC5MdbNeVzGKO#b@y#z#+QA^EelHJ3*fO
zuLZt_{zV(@r*{Hhkc$bu%L9%a{-u8he;6{ObkH}!OiE)wbPe=M^n)u4?Na9$olRDL
zj{in-0)8izgq-cD1V<`k?%S3~8UbC;HUHQa^aEiX{&%}_Ia;HxsvYXgp?jdhy2<9l
zRdGiNW;LvvUFrV>A!*kko9%okHu(ImVaZVs^E>3HZ2>kYa_1o{KUT_s{rIS_Oy;0<
zc;Z2KSl<+59K?uhr?$QaZ~*Ge<`82?xxxXR`mSO3xe^`!;DC7K_{eIi9$UG5rTi;n
zw=3*E8DXXu__?8H6Sr<kPasg}4TI4}mzMt;)J=UM-udq$w_C%F4Y}0Y<F+Xek6Fnv
z30MAUhZCm9+8EeNp%ELkKn5#GRF}~UoV>_?(s5BZmTVC3=3cJZ&lr$k$mJ=qp-+_e
zVx(ou2E$NjLlU+qOyJq4ruh%vWA_z}%a^r<w}+`6n!95rZ_vI|YxSW1aTBjqTkF(c
z?PuL8hr$xO<W;@ySadeo9&O7=q1<hq`pz)}oHzUwJfM(|p7)7!n}oT<>o|RQI77uC
z{}P`qdNDpOvqsM9{mY@0$4n6lSh%R87oO*S8?NCgy&L`HmF?;>-0<rrYpG2Yv!pGQ
z@))_bbhhsaev4DBxsCHtOWZgtZ0Pn_h6lu>z390?cRWducAxL&2D4Lmd@NT+fMABa
z%-&|C_!Hb0S(jh>tzl>#N=|sTwM2Zno4M!7$A^8fEv9YpKm6ds+*4+z_?g%!<K10J
zjimNI!Oy@!eff04HLGR4Jn4p=*NF@oYhKJ|M*OusmrM!Kek5jMjv^f!V#Ac)fhQE>
z@BI!BTgq+(H@JG%{n$v>g+JT9e3oN7nD#XY0NDvobf|!Q<{x?1G>RK7>@-E=-&mid
z#pYnj^2#!&8m7av6d_1PdG-=JNQ)tzrS<sROxu{zfcl=c1&45jynn${ihQ;G-Y4<$
z?#%D<V`KB?FAYR@xa40#aE@3nZ1H>AsSr<~7L`NXx#?USZ3J)fpDqmcx<l%#0y8*C
z^vXt^%VhNM1Eruno-3HpOj1TAC5C0i*mBmnZ6_i`bx&Gu58^2K*43DDz4W}XhwR$F
z0J*3+1dB9t@V%W>zO>BJ#X*ILRu}AS+)-~L%PJ*zw#oluWBIZ9^#Tv0y*$=P7_U=&
zwzDeNCGO?uGv!>YN3HI=3M!T@ZLzAA5vD|##j?tNr>RZw$K)6nBrjMFhQd_vwJcpV
zE!g$d&;K$#D*n%6<G3FFG+<W<GRZ-7?$~PN*sSKze!Wy5-z!fplUET}Dn7o&{Pl5h
zsm}Ty8#cI11<Xj27BPnpu0m9MigsoJ*-~py*_B~&mU{P;8o<Jnwk^l&?L>`3lN<Ij
znI}icQIi>F06Gr&!4LR=2$;q!5dU!0B!Bw%(o@8Cj*I#&^v?C99S(NydECMKR|6NE
zLxYdtDu(Di%Dnj`;AXteQ73=A;@fT%)8S)dP@`vz^mjk?C-+Hyc@uB%(R(S-Vk~om
zp!6}q{Vx0y9}8j`dv8d0CggeIdf4q`!oR!YqV<bK1`!Cc)hYg?!Y?^As@~zsuvS0!
zph|kgum2Y7B!9#7`J8j(xQ>QQzJO!BXVi?)87@L(FH2`yRYGjiW2pDTS;Nxl#fhIU
z%&75$lfIoAMnzjQliH<sInt}iOcR_GuzWGJxyJ(V=~&#8IptQ01bD9aS7N+jXZxa*
z=DTHd)}F@a$Nz%_Mou5fg0WYPh4-#%wh#~Z5RPlDV0OMc3HL!N_nUN%KmQVyW68^?
z5nXvP&ID&Yv^AbQ);OT<Y;2M$YT^4dMi~W0_X1$)6hE8<d$DPJ1iv=w<))eo>(U&l
zLe0YY!&xy}OALLyTf-P@%+Ofj`k7}oAW)@J!1!gqIO|@^qV4%f>o{cqf*rl%u1^_y
zVlkBYo|6GZ!t_0_dr$LX7=8pUCqPUHm3_!;nqqfPsq6Shzj!rh^~mYF3DV<VD78w#
zlg$oW=N~kSY@u;xc|pwIe&!3j=!_Lf{om%ag_``Fh<|c~CAGOZg=^pH5glDu`a?ZV
z>ca}pFVI#@^)7#kL-X+HkeI3FD@@6r;S@NQE(Oy($NDC;*VMY8mXVvUfEmqx)#Y&e
z#h9QgY7a&|0qAm(Gx@u_B%^QQYkyxf4&R)<Q`x2+XUHr#62HC$PPXwC2oGB>sL;PW
zol5m53$#41;ZFzI&h!cv*}^Z*$=@WkPmB}<4{R=Nf2s`sX|dBcVAKFOO+E{vI1Aof
z#{aeKceFN++cK}L3B60QiW<SsVL3j!@;g2I$u&?s>cye4KV03FM9*Y31Z6LO8VS(<
zYfh>ARw$9FGj?opD5*bW$nVf2+(S3$HH!px0T8-!sa`S+F>t{ge7@FBjJe=jxXacr
zW^|pg1fI(xGHg{L10yX9QlWF}Y}k-tyMM>ejewCh7o~0vgzMwhX7w+&@FX7Pe4TAB
zV)9uY%p#3FO$69)p8pfN#zTc5#N5+4X7fd3g|5qXiO1f}UKmt#x*BN6i7{k;Hr~Sp
z%lXT}Mx<x+9W!#*lHA3EAaTjYg%{5=GkxB|HJ$!jf^uQ<nJzSc-&nlIK^*MwYpx;w
z3R+7_>13vwfekA7;&rOthWArNm55|HZU+Kd*|mjEzy_`(Ged`ABJV}0LB)}MW5)fP
zg!q4t(x+x<pl1fnnFSxPO;f79+E5ofw-2cp-mu{XQN%Uz@P|Q|)X#sZ?^D(=MWqaU
zI|YmE!U$EJ5w)YOO|+uNeWv?Oo#gEU9P5};;M@Di0bXMzfS7HZ$&bA^(zrT$s9%YZ
za;#iMKfo7EZ|@U^pruTH2s5@w>=QhUI)U|ekMwMdCy^XsV0pc8*jB5-HAJb?46E#r
z)%XQm-|K{*l&+$o{zIGhh<?z=!1vKs=Tpq2Fw#+(@*{9z&7if{i}ir({YsixE%|vm
z+PM7H54T!Q;rAV}#DOiA?++IsXjE$EG)4t3CnI|0HIXGm<zCC^Ddi8FEVZxPybB>_
zoTO3L4Ge0a=RV7yN^t&;RTS$_70bI$h@EiR<*<XVt9s=Pmus{hN+CUSjWMyhccgm4
zVCgkUtdQ#8fkZp%bf9LO29vmf>o)%q*bci%1+*%MjcBD^SnD!DYPov4C5GOwI26&B
zlE}ZGEPwKf`bkZyP(2OB`-1~!<v;HKjH7qnHLuT~x>ym1|N6A)>zk~vn;7@jckF;E
z8!x~Z|NpXAVph9og)7GU<>9SX21y926jq6~Y6LO13TJ+k=dY3%DK*~Kcv`}=&BIjL
z#ZGOn95<c~JOkPcsolUc3{bSU)L-oOx}@4{@&cC=86+Z|I0^mW`WRz)2sNBDtcw5p
zLl_|Z_RXTSmP#aGP&g@H<(C;t>wg06-G-d<dw_%dAR}bQa9~pb;Jhzk_&3!o!|aQ3
z`0l777RaD~(|d?C;G0))Y^*LW>V@ZUS^|UR=e_8fXf>(Dmclr=;>vkGA9(u*NH54(
z2~-Pvo#bhR{itIiwXc{}QOB9MmhY5ux$$#-BHhJ&LX`jI3va~Vlx$DTWyqC$?A3D(
zob+=LKYUU{``aB*1VM_v3%MZ1)`r$=v$>s@yGD?IkwxcAhlPg@YHD<nyJgGxD=&x_
zspletoSsSi>w!uQ$B?cM8e7aJ4Ok~Hp9|^of=fD1({wX+qc8p&$)_#La6p$<J%e9s
z{uP_9C}M%8{r|(2IkLWG(aI$`8tyGwr}`_%@N(QmCD9A6hZAG>Jy&%qJmOTu1;0K;
zJkfFXcs5TBI$;F)#2Ynfy8dT+Z$DizbEuz5sk!`pRY9LRu8+0iwPiMg(G3k<ri})C
zLh>`4A_-G5_M4#8fLPE&)9r{8*lbwzF~VCr5K@cTy(l_DLq~vjZ3^v|*{w7~pG-$3
z;YNqQiy_xIpd|-5Ss&YniXP?t2SO(91rZ7Y@B>Td*fUd$S&p|I8;|XN*J^;eMzrZ*
zzsK|T1MgS!$$<NTHC}=5;H!6mIorpKme=;@Qw!<({cav9Oy_cI@*9xds%;91x5kMC
z!Wy70W4dsjx?~Eu@jzZYL4MWNt-Z6c0A+{9<5}wkUc)+lY9P|nd>I8_x@Od7?bV8+
z$qw*Y=(G7R@1*7GXX(&IH||jJW#5Hum?bia0M$uK7JnCy$w{eTvz@&8yV?I>9uqKt
z1CH9sWo$GEglu&<mt}~SQ>7WUxc!!dmASxVSiU$`J+vH7C&tt!PM)GERC?{ee%TEV
zRWw7)7=%Y+O<Q0rQ7VYR`n!^N-IsrcZ!!?_{*%;6LD?3Z&-Kc#ATY-ZVlK*|PKO9h
zzB5Ygf?E$MJ>1O?l;OAG)LQZ;19Xd`w?h9rULK^aE_8x&KmYJDZRntIOuyiRg8&OI
z=_Zmg;jYq!oa;Z;ciAD`JXYtFi|1dx>;+9=Yy~fvxX*gd_0ITX9`K~qR|+|=t@MB!
z4*lJ<p~T8cOQDF5)URm%lvbK3JVl(>3%Uy36FiSn5Swg^c0SMn%A(r_=>TWSOURfb
z3%>1I)VT6Up^MW2#gfeQoP1knBQWm>2&Bjwvo>iRZchw)%k5T=1Ln4kjmHIt6#Jea
zWFw-Cg{D0!L#7-2l1b|*fsER_{0YFwXKiCMR`zNK(+cmK5AOb^Fk&<_QIW5gqsybt
zp%tYedM(cW9R`_tm`>Hp)Nt);2X>2`ZLYGJbl-XItGaM7-x+&6cBcrqexCzmiVR43
z^e#=Bj0dp4D>>U1nH~L@`C%R32KRL}k4-D`^Zl%ggu=JU175w-{M90Sgx-QJmhNiw
zRR&w-&hH_8d`#$ozi?06O9Fsn2cY7@Z<7-6F)L#S!~Of3s>6^ehbK3X>8&up;b}ZJ
zI3%^IB(Ujod^!P2T0o(J`A2A$x(YS7=$Y%_Z<{}RW;uFP3`_IJtevsWyuuChn`&S0
z7#sACDinekTmnJ*q0+bWTYcGur&8ul?&CpssqZ#Q1j`h@0j&!nUVR+3+SBKpi>5lu
z_Y823DEt^mappRjrLv7^EpG1|Ba`b0CcDr_{m6TmPyp~q+)0um&QFZv@B{7Ig?w*#
zt34!UxV<lK1F_WROyI8VLLNc-pPNY%;2lmN_jO-5S0ua#U|PzAoXhWP5WzmALv(#O
z`+liDV;;dJFlMs{tW;0E`!>mtsGZ0F8}DsyyU+-W8CMPasrUy6{-Z`MX67Fj80f#>
zKU2|qq?z#ed`x=p<@IpwSTbVoK3*S{#ZL_?R0xqq4Oo6UcP$i4_o&86XPGsdD%RjZ
zGU0%mnlemc37NbC7p+q#Spf2#<iS=V3s35{VwRN5Kvk>&u<^h4_c#n;quij|w8}ps
zx@P69s~0;ahfzYK{O4QYT5s)Z&chBf`vc04?$pHA^1vP|mD^uh$sDjPaY6ziLZewM
zt-XE5GV4bsF&-3^akhjdT~gg_-Bl7KFZ6OIsc#dE2&>FEmA^Ad6!m&w$kyDxpH5v=
zA8rT2!p6^iW_oi?bMdvxsX0WCpP$S8@6^379h`UBdHpHm)^4`2<2`?^tNXL&3gge+
zKHfn|V}vA_zm2=7F>F>nwGzQETPKcvckq>(()%yZ<+SSF#hIZI3K1yrM}$1E4?Fe!
zBSbvPzeZ45*sD!~q!8yKHsV))*OH$Co7}K4k3i<U?60n3DjU<(>mR<2G_R$V<rV`t
z`3vwwE>bc**gDBwqE6E!sn7$D<lTXVR$p9=I0ddefq`sE>OV|vGgUDIqlvoN{kl6;
zB#W^!<y1dQu<=_#;jV~E=$!aPfj|d^+103-E)I)iLzy2Y(CSHkSS1vwdg1MDos;(}
zj0YRsoq3y}Lwd1&5GLy+I%JtkgY#A1?za!7;oB0&b8B+@cgwh041EYXQjlsL9nvY#
zP^DRzlnJ9@+FfGZE~2}YS;pkvF+oK<N}{fhAeiID<@>)ju;tZ|ukUoh>}0e%d*MbM
zLP>OG=-0FGjei(oI`qZOw|+*?-*y$q>e-=;n#Xgl7!=nxJj4JG;=M@xHM@R}yL2u~
zg64Osska)RL5eZ(l%Oiqkscv&UNAK?R0%bGL}vMR-i47`LFn`7tn$(E>bXa#2t{Hh
z%+!15qx(c7Ki`S&<CXKvHS|3_#_58do_JicQxy8ryNDB_Rnp{G5C|9y#9`=*bjH-|
z?#Pb(hwJ?<({-E16%VbuQ!}vNA`%mElmos!NKgr%3WSgfJa|O7$6*qUH^c9Y`8&3a
zDGLyO9|jTm0F%;yJ8Dj-udU{dYyVgW`trlV%krEc$T%aY5-O>Y?Y7v==Be|H_=6W0
zHW;Iz?#oW}EM`C<dohsz(B=BmRo@n+(nx+}OyHN;ROH?{5$3ru0+u$02K+4+5kV<6
z(`=Jn&;;wpcpUpVS*j327=5yv=%)3g#*zZjyax<~TnHU%DT9#%EYr82x-6wm5E6Gs
z5;gLFrn~z%Up6dnMjy^s`u~f`TMIyAA|xsU?l@!|j~U-$n#1!=w!;FQu@sYHFeaS8
zN)nRps~9K1Ls3?vTsUl1eVg^=`0HNDlXA{~ieoE$Z@=+lG`&^KQe-VJOPZP_?TDL6
zzV$d=@9SffIkO7LE!$IeonRAAu{fA&+>ZR8wxi@E2r(|Q{K6D(KV7Y(&jo+G6F+n$
z?r8S-pEZ<mZdiviJRWIJgSkW+J|tL8ZP)w>1Slm&t{q#lOnC+WrtzaQ4^v5uCY=%L
z4HkzI`-Cqa(d{_=@%+ip+d(B>rwQlcR|PY}hU>(8A7d{YTni#&-T}p)xAE1fNWi7}
zNVsl`U~+@maiKXHDV|f)9Tue63(V62(QP|LdQX0I=|6*7iB_f634hr!<jaV?5h`59
zyW~oi4~)((yN@v@ryw`0&Q7@$hb}w6SNCYEf2K}E@%zVT-R|55;%iEtuNN7z9W`f6
z<eA~SDgJEPZ_%CS-i09}U+?0XjEV-IAOw?2j;BsQ9)sTAg^Yv}Q779mLa6@)9q&<f
zkXYIVon8ZUL|q8a<*qFFhn`bsY>|V7P>YEG7S%3fe4^81ezSGFT>dKhM&?g$^GGA{
z@5zz0-24QZ2(~8*g+-DY^prUs?;nISlU1MRLManQ<nLb!o7RXXrn^kHMy0Y<YQ=(|
z=vo!flx{=cV)CC9kHJe=QJII|a6s$=6&V{bfEe%AQ|O&YQ|*Z84hN@+=gQ+lr+_;U
zW%>3$#;<%a%UVx=eFw1Mpyu~mnqMFgThP4KrUiEEckSB=Jkz8C&=Sw?yAU5E=T5dm
z2HY^Ii;}JF?ps+NW|l9`b@csmBWZ&2GRSS6uj%?v(<5QxqISiMtGTDA&at2W3o^`h
zzwv&@lgGDxalIv-TDj)21e!h#Kr5IKtWCV}d&3-o)ANiPDf+21cH26p>ZO@65dwJv
z2bZ0+!I%WjaQAT$TP-@HY5Atls)h-cCIcaneR(~_(^uG_x(zm(oR1zIBbQ?xzbcwu
zRzEuDA_k(<HU2Pq&Nmd4F!XzlO1Mt*RPu4wa+lhyhaYUdttRSv-Rprxf)Q3?HV5kC
zraEI)rA9{hd_3tBMaj;>;SbPNE9I{@ESn*Hi-e^B%%yvY)<s{=p<UZu?{|R=yv@s&
zOQt>t&)g5w_k6p`41{b-gm&gTMu{nvFNm-~pHp26JINz=vpu?7Mx4$maljNP&}}^V
z@8ia$q5{8C-SUJxPk!A%p$4I!Ke7uTVD;Nl7$!<9c81Y&H9r|~L$-iv5K#YY{v>nF
zCKLb+x9?&^oQHZUZXGwGPhR#t-n%>IC%NJ;WW!n4PX*p>ri%)%8S~(5#hxj*V&YCr
zox+`i*r1`*Sj?s0g?Ajk$Wmtrs4k>!RU2M&rRirib*4u2CpImA43(4N%VI2^Sj8lz
z5upkiOEma}1pP?oC@5qZFchxsnOm|q=4$LYa~!hXH5$0!{6O}Xj>JGulq|ZIx4G>3
zdU%_OwHi}jo<{Ee_s?Pc$9cv!bHy+};kq6KZ%tcg>@>4mf!6aMI)Z1qoLOUw5gN=u
zr@08xELQkK^}%4HX;qKt<Xy>`s4X008mlL7Q?m1vpNlizxWJ+WPRGjMul8<5)Rd`h
z_g9>JT=gIiB7MJFGqu^SO~FF7YH#m0)fmv@nV#tIF86+g5cS187e}Q855Z;T>JtR8
z^#sJdOP_or?kQt?80U_nbqMF>Tu2cgDVE?Cz}kJ1CXt(8t_FalQ&+HBz{97;FbP|X
zMxN^g0!TGuEOo7`7Jm1r5A>CBN`;m~Nqjh<u%{wmW#1NhX;m_BuR!7<l!*LEg!Re~
zvK|kh!S_=iQ{&BQ-bYChBNfDH@vo~``K1%xquQBD@esLfKpy!)V<$|SEXj&z>1n#<
zmN!ZBkIL-nbW5Lfp<|Zl;z4!=-m3qtV0MT`f|GIHdC;Z#rr|EL+=gM2Uw{|cao?e_
z(0@Ks+v8elDZa0o(L#L2oDYf!n}2rRgA_p!=?{PW(>4ge>_f@afk3;8<}e?vs%Qxp
zMqFv+yk!SBeeWF4Zwft6W509j!w8NX-t)@q*bPE(*1rKF1qk{Gts~`#j;Xm#%b${q
z#CPSzGw=LeoKw(r#JB^0qTrt&qQ&x}Lc+Jj2YpBW!O3^Yi0bDY=FBPb>DRsWC^w`m
zOL7W!pVL-gI>#5Wn;rh--PE{UR=ZnbW9Ij+SOu7B=9fd1!`Jt!tqrlMa(FS+kx&10
zW4SoaGQP7w*|L?)M_8nB@cbi(9gOA9_njw7zxCj8^l_b+=Q9?}<t^56S9^PWiax>z
z1FwH;{1I!ny-X<38*C0)6F*m-nW<M5crn~ByuQa>-Li4vF;`vtn7J$4Emrd4^3cI)
z<qxZu-W-X)AGMo%GM=jNn)HU<*R985oDr9|uu51^<GAIb*;pre@|nS_jev7&z2D(X
z7;kMQrZ6yE&vKKpvm3H|Qvt<2ysrLDgyE_&;Oeb1v3H3Pee&P6?N?duV*KaV3V#Tn
zezu0pA*S}5WP&~fpqv$mvD%!Kq+{MmZEV+0PZsV-#bZO&QGfM@-b_=RSyHPeV}agR
zErRckrwM_YWGG{Ij8kg`zN}ZYp8s?le!OI#4`lWvSs?2?|5N&)GLn6lb!=j+?;6G0
zue2D$!l5Cv<#(NsCHu2B`7!2_pV;g>3`rc-=Gjq?5T-Ho5(-;yWfCn6;%Lt|3R#YO
zlaa!)=GTw4U#FeqapnyFzS;L_VtT~{T1$+rMVdM}XCI!IX8E@5Iw6pmExxqVd>O0v
zoAeLZ<d+2A92{lZCq5>oHpLD+dk$#f7A^lbfW^{;cO}_imNpyWB}+yGF(2*Jn}JWQ
z%7E%;G!?_^^q##V2}3^*h9b$&UYI0BRq){26!`g^8wW)NvTFqo#YCmBr&jw;?iu?a
zy1nihS~6Z4Y>mqlg*HEB3nDicyG&GCk{0RA&Z|1(N{WEx)W$?fO<CG1&e!>bylJm#
zU}Ih?tgCmDkBuNPvj!Kp-Rd_nD2SvtT|Y4Jt^Q?DURp}A8UOh(;J42mWiR>MVCZH`
zrdNoz64)8!?K4qxamS+Caz=5nnf7Ux1nRv)XiN#JYwk-aR|QJTTPs+oEUYk24g^FQ
zd8wx--r~@9&iSw|=LKHOde-q>L1G_msWkHDTWt~USdQ380RpKOqt&D#d6NY+HD^Ov
z0U=~SpqkXnrShjk+K_zyI;jg9ten)p&>t}nsm5lP9LlyMn=h*)+O)Cz^V~^`*?Owh
zuf1!C^|$uEoYz0sljlg?*TL&_o29?96n*(->zBQN<u8*8`aqgjvB<xwHI4&L^$NLi
znF4k7q8zeb^H2R~`y}r(bVxZPooz{c;ffCwpcm!BSQ2#)%84SX$gryra$=ChR08Zv
zYoT1EFnEaJ)yG+b-f)_c#w%UsQieu#H%0)YNLrTEH^Or0r0B@-bCLvpbL+4YLq$f~
zOO7J9q2DZQlA6RvhXiV<IQ&>sJ0Y~vTD=KHQqr4TwZz3m2ypKpbKK^ug^=<W{)S6H
zIh+^zzWMpWW^rqWp8Jm+AKsP?Jg|PIA96@x-|jQU5zgPTNBXQdV&n5jMxnZR&|RlW
zWx0OSw|H8PEP+fU&13rJa2iCsY-uQK`iCeXx!DS-Nc^s1pkrI)nD5)%+ekB%vXB6Q
zk5*G+UEO#YKDbbnhBT{(OzK0w@3vX&zj&R9op2?hFz0pbQ!PU1Dc11}kMaml1tR9Z
zEVTMp<gQlQne!7Bd={9LH3fh29}#10DmD;*sBzMRkMsF`vVgz(f6=I-vOmX$_!Hgm
z|MhfnrjoqC4(#d<iAZC_zB_J=ZNY4`W=bpBqa9xgb(JMZF^imSF_<OOBjQ^YrG)D#
zx+(%v<CVpIyD?n>!=*>O{<bP*Zf_1%bL7r`e#c9Asj67t+!zfWO7eJ^6i;<ugT1=j
z3hT^u+1X4-+5rwhOaF$#@9o@>UpSVSg^_R4vpuLHe`ur{zI`=ef@weu1=U|qXDWq!
z_4H8JOf~+N^3_S;;}PoqwDHdx>W19s32d@>H3*_6wxH<0H7_v4YC-~hxXihmN!CMa
z3XD1!W_Yp}V#4Y49BQv;4;wq~f}^<7zw5stInPB0qDEV`!t;04QMX!Odmhh;{X4Wt
z4Xh|878i7SSQm_B!`@cNBzg}(j>{g7iLDlx2KdYV+Y2W5sgyMDXjeud1oyg!X137a
zyx{<}KU^Dt>hv8#K>{}V3dogB>}^6OFb#qh_b<f!=gh`Z@_;6)PUd3>>WJpBDAP&J
z<00bJPRq?`#9;wtOk!)K^4%%`jj(Ya;%iLCn<K@;7K}k2KKZ$|YA*(FrkSR*F)LXj
z4KNb7d~INpU4Oj3cu3mGONwCDR-AwTO&)RjY(3Mtx9QWxca4nfPUt%l+s_h^??KX4
zo!dBWp3)hiP>1t8nzCGzTy6c#j)fg&rk&3~S^akwxRrlHsZ#tDo?t^{7hb)?)AXQg
zm5vlPUZ?BGOLRd}8T%1-8Zft}s5Pf0Dks@Jo!dR`6W$4xwBR1|npNLyn9{8Y6jD(8
zSl-zn^k)kG|FapHcRueTl7P<=hK`DE&WW~*gqw$ZZJ6L(PIOWy@t=DwSa1$nT<D~`
z7@Z1^6Q3yd_gkL9|5+!#dvAX9YIhmmo*Tm4MYWete{I?SG7|X|$Z=&pT+<|C9mr9G
zo9%HWvA-czkyxdFZ?4oiLBLm&9$btaD>q2AVirf!*R{5<H)4+$?`ND!{4KJtkUsiG
zjvqHI@4}ubs5rwVp3og0bl7{G&>ZBK%WOMkS|{ZC)D_Qy`p3=}v#=dbr0`FAwN(N&
ztsm!g*^Z>?rMiQlS+U6^wu(>d3ZGzVoHbi3$9BWvEdnZ#+{%5ARNy4nn^y|BhmTcW
zdthTNyRedYpb4VmK4O~u31zNyOMm`4E?lOtM>9QP+!2Zkxz@#3dZMjbbK_4WhnE(g
zhbhLCemLJ*`XZlT*v#XgNIkI2DL|*=NWWk8+Hr)K4%NU|^V%gEMq7MV`B6K7VJo}Z
zAW5e;wrP3_f7ERmhrG653X1^4;mEGNImjwXnxVUUP%Lx@CHv6p{f?rlar@-=El+76
zArcj5E#mcz*62O8+nt+;Wh6?@A8WPsjVP!ENns<t^5vp_ql88`a$kuSR3X3n=HKXI
zATTnRyl&9z7l@vYF<@<;EXd=kHV)PEi_E}bQoz6lTip)IYF+dwif8tXI89ooz%<-V
z-ym4i_3G`Fh?^;s_Lmp-eS;U4Ux-NoRkpG@O?4v<N<VVZp%6`w*@Hc;F_)byg7%am
z`fQ&d4pb#A>^^gmSD!20OOFQm?qR=Ba4tlPS?Dd6e0(oQN1?TymEW26?XPIp-y1r}
z$U=Z?jE!r9pgJK`XwPf($x&?=L9m*zd7==}+nA@aocUFTwo}6=X#W|VS{WIcW}6}#
zIM2AcXP;ijVMRs7TqVgI&Mp0w!~R|8@MyJn<%_yRz2?t$S+M2^VNLyJh>SDi)P{iw
z*%;GE=UBJkT<xQ|^0oe>W-3tgro3a-g;5ud1#NKA9rT0MFnm*{4~pKq{en9IJ0`id
z>l;vrgnZ(}2NWCr1V~ek0`FE_j;a+h1&?9_P_EvnbB@>CRj>*vQeFPb(E})^WBiJ|
z#PwIWKohaEB)(SX)<*pjdaIB)lpA~nm7W*)At|}IleyNUn2eLIqWdckghp0w&*+&6
z3DD^45@i!hYA;QEiV!_I#hv16GJDuXZ?E1l4L;op28bKQxH;^{DCwy}H@k=zCL;1L
z-xarGM&$v*RZn%v*yKMiZXe(QCF_ao4CJ|?QpYdQ4+pWCLH3e)OmAk*9xmo?3llEQ
z<sUxF$zCcHN4CHFg<00D5^n#ccXZ<r=^tsapXsCV?^3O5!PWcy`HYn2zIchpKT4en
z7gt=mZh2OS6M#$h=;Zcl&Y{JQ-EaDqh(C3+*vyOsT0B_RI9n-negt_J>5&SD!n4X`
z$!?gDm+tEA%ScamwYV|=6Cnc{k@<(YMyMshTg{)|jHX%RC%e`l5%aH;;YVy{ci&ah
zUrVq^E?$kznL1f@uj9{=EZ&UElO=ez!x-tDogGX(7Cd*7=X9-9CFd0(w~d$98s+(O
z{>KUE_?;oMOP}H~=U{o}Et8gUef#|cYYg5!mR*O?x9r<`+<Fl^H@N25+5>Z_XdMF|
zTGo_u%dC*?p!SP9Qu-lYgU8vx9H2vj@mX~&ZDfOhKN41>CGxLFvbvHTII8WG0gdY;
ze=T(shl@Wbc=`gKtB^7v0KS%5QztpT8THJ1$&vFd3eyz3y9}*8#HWxFUb+cn2^>E4
z(By4ixhDsHy69EuaMyfcdIf;Y-hnf&mC?|8i+p3Z^TNrcy_#M5l2wkhUN7|BD$Ctb
zt)3)Q2Jt`+#zgL+^DH9oEA0WnXW5>noO{`DY<iLVyYZgbe;QzL^FiXTFJ~8>NU1d>
znZ$bF@Xo3^BcccJ@b*qsVI%O7Al6aVI2#T4MCbPD$@18lLVJK~U$RT%4V`Ux%NK82
z3KOpln_8->lDiHw#IvBj53<GHJGXgEw$1=qf~)wG+gu43s2dfxutqv9#TMpk+Z;&|
zU`jL2=ArkXXn{^<dHz&8ELgM?mB>w`k!y5akr{Z|xeMFDsv|HgZ<OxJ#<Wr~I6}?u
zIs!a&CwF{8vx4-E9ns5g(XCwyQm9C!>=jD<X86mqhWdV;pxn}clGPDg&d2ymmCW;J
zAWf@$4Iy|q6U-u|)XF#|-XP}g*&OEJ3>Wp?A8^*ynq58Nn9Jc;++$3YCH|ReS=rM>
zc^~8(Wxlx`M`tmx7BlqfL>?XHJ`w#?*6dX~*-&NY6%!>pZ6X!yOgqXZX3Yxcmgq_*
z4&YC+X?`)#z+%<A9i?@<?~HjVxiR`RPItGLSp0KClHEt*Ok3_(o|(y{w+vS@j)cfg
z#`_Pmc*XBWL=Lt`ar?oPhvwLkU{IuTQY3lgxAXktD|Fo@OlT_Tv5)2{^zm`bc~Ewe
zi{)FRb?_VCb>xqR&RW9LFJI^@Yft@51AHApF|Ht(1MP`cag_Z|+1=uMKh>}e9-rkg
zkD7s9%Yfr+Ql#L76=Vuiutb3;3qa|)74_1eZe9&Gb3Rs&LhnXL@Eb*LlwH?A=Ic@3
z88s&oMy%rNtMceb@Gz%#ya8FbFP$#<q{iWcYlofDhG~y;V~pMK{fRB;EWGCVz06-5
zI(43Gn0#iGm(OM+8gG&KJ`Lix?s)q(nGm(?H?ilN+INw%C)0l&1-CXH)2wS{^vZK^
zmf6atb2xTyYM21YMugf_L53AAy)9OM>7_5ghO*7-Px+=PM}F*)CPnA4t3`f?k&lTJ
z*;cxOyr+n9#7jIosNX!Rsp$-Nro6HwzKe-_mK#V|A;YP)KCZh}KQbO*z*cSMvPAO>
zaCxILdX2A=G7MYZO@#Sld*7;EY_*IPM!iSwH%`Q^(*hb*&vdr)0TVtM=dm{|qkcK7
zkv^c80if`Tk*M~o=Y52{n%(OUB#S}K_1s(>PFE268Mzsmu`n{8Ient6V=1=5zxs5K
zmR-FtUkz@iT|@BM{@f$W0S)xPGzs)K48MW&wD_qw<LC~?_0}x``BWGG2$6XvYUs*i
zbCB||SW4ZwZ+-iCcq?O$lGi9g1~O4MN;^iU(F6;Xc6QKh4R9GU9CXeFm#4LrAh^P`
z&z%1hA|y_lETc9FpLSdfzDDkf2@E}Da(Kuk_tzwwKz^<4F$L+FB8&fAc|YR<w|%E}
zJpmJoZy@q_vV_*+Cj^<vc~?BfWZ?gT$$@1mx9g;#1rfm!6UU7ZuHqb@nv-LpDXMVm
zIR4pt>sR9!EtauB<5l>o(A4Sx>?hB|*wMA6j9}bsb&lWW)*gKy8EV49)c4qIJ2#xE
zdIFOaWZ<)Yzn8<+&*1L0xW>e5)L`t`azJg+>XChL^oll3&@xQx_NZa0;{5}?6ZGQ=
zCX75vmUmMAz-`uGAq8Kv)9>9g!qU02`o4mIiS#>%)%2J+$tr5y*|qkZ7RV-{dp}53
zdn<3L$vHSDa@JZ(0=^6%U|q57J8p1XVmG_{YhB~9;?r*gV0uOWH}G;j_&#+|&LW3&
z^8c~`?!Y_rYYEq1*}hvpUM6SvT-AqKQbZ@ZL{cAfLl6ObHk_1%AFHnuArIO)UWXEw
zZ@O@)26HdR&@?8Jl~e2@9gWK_nV8^eu|k09|0C=zqv8s(cHtnwg1bX-cL~~9aJS&@
z8Z1D7#vxd6cZc8(!2+bw-~<mYA!y^$z&)9{-#0VwUF)5->IaL{-KT2TezsTblW&z-
zdqtf7431Q<zjr|QR(7%+U59-UTtFTocUZ>Y>T%1Zu;#$7Jx<vrxH1=H#{@N#_dQ3A
z7pg9<9<U=1&f%M<n|&K!Dc6l7ecoM<FIEl<D*daN%FphP>^81tCb@6=5vZb*LcQuT
zfs3pazs}{<`;8IaHL`$Pogj_xSg>Sfu1P8Tzp*@Px-z|5xEz`QagP1L@MWK-U$33r
zg<Gf$C!c`QZ>oy0CEyWJCD(4(aXw3z(MVr>kC__+-4|Aix=ota{5b8|2Ya}BtNc!B
zYN`9?9$43N9ic(oQAdW(1ujsU5No)!|C9yvi1!>`;J7zRQ=q5$<-Lwz6VKuLU{-+R
z?4Y}Qncjq-uHGy;jr!J5lroWIw(Lu6=w4pKOUudn_<$oA`Y!tD?aI1KI&h3$M2tFw
z0BD5Z*Q&v@)O;JTw?wql3A(M^%dUhDNjOityxOIux=TpQt?>`qrDdazabLc^U4Dfi
zZ^Ri+&H&t^sZm?{2CY<ZlRhnq&+e6SL{?PNj}@j)`Ek%f&<X_!XV*!oZtPMI*y}VG
zzutc8!>s|aX-y#p7xUge*T%IW7%5aw*+I%aeR30ml3Z&N8T=u;6_0^pb)j1@Pi$!C
zr3n*PBDaBYVc)f#HcUD9Qz-~-y{fdegJRvPbhhY>ZK&m)GF9*)=1o+9<d0D-%B*q<
z+MO?GgVq_Vz_HEGIGWpyvF?0<{_2``TZqZIzsI;xgE?_LG_D@d1d2uT@mA|9ir@Jj
z<NfrpzbP66(Z@;Aadv7W0ZSRY?Uv{qJIf{?B6?cGYIGu1M%%3Ns(Q~GElc+M9ukJw
zqj@bn96qvPbhmQb!N$oY9!^il5aw=7CV<5?P=azVblF)v5?<swO9nqFt;`|GtWhDh
z3Y(_9)g(;@Iwo~XTu9i5KF}}F6^OVti=+s&N37yWLRh2|iO%)f?3IKAI3V%tM3(t&
z0terd7no^Rw-;@{O!3E;Y=kt#-fUx^ibt$jbnTl(i?0|5x{~KecZ67*KBiQn3@xJF
zh=4uV{la`sM(lG8BYW`hx;xI>pfG$8iuA|TQ@dxQ$GUL@=ZJkNAwMzdz)Pj@ob`n>
z`Q))t#&_U}H(XbFyF``IH}g9jdY#q=hj&pM*V4VGZQs9&sA3Nl{$2>F_{4~H8Ar@W
z@W2ti(=L1F|3U4~_T537T4j0rCG$C#-&jdN4#IG#IL0YUM@q^GUHdcl8c#rr2(?$N
z{}gDS6u}}7=@I3A#DCOGgc6QXPd0ta{diJWky2<o>p|ne&G^1;(U#(#wyUlXBqicQ
zp|^YQ1MikdN%buV_1>OI21dSB7}xEA|C<Xcj`|t0db%bM;eR(cZ=&zC(g(`l8Y&y`
ztG;{F_pt=5!BJc4Kgy7esLf4yJg<|&gYPIHB@cde<%o?pI`%`2m?zv79TML?akWK8
z&tV6_+T)8CJ%3+(*^nlR1%g#@+d*lhv-4rs$-A-MIeQp9`#^r?U0+wrgQeU0C;E?R
z)we?Ti^)!7>3S)66co9~bWqLs043EoCrrF>N8|Nsy=JE;Cxq2;k*G5h6PURfr}A<G
z5!cU;pl8<lSg3~$&vT4$=K5M5W<=48v61dZr;e=Rn-RKS3&><ut{@Lt;fa}2wUOQ4
zN9n<mDq~G#rN{soLGyL*e6avVv)11`!fb)do)(l3@cNNr5k1gx*D)lW4+Yx;5Vu?N
zNaDAIY4K!rx*)L2@$%LA6m5Lwqw1RW7!K3Rjib=+DLQOn#|y9Yaii0BqMp!#H@nA-
zBU^Z%TVC%GWlw^N{6C9aGabLjYqY5;_FQ(k_vFny{Tw@T?O4|->C%38-7LuA$+T)V
z(y-p+m-#^n`75Swoe3+};f(j=@?e#Y@6F6uUAt7S^q43`L1TXHSV*q>=n4-*cj1(u
zUb={3i$k1pE{_7{Vjxvtgu(n~@ZFSj{AgY6ZNOj{K7N!OCAeNWXOSM6vc9ftyqTf#
zz-`t%={a?`PEjhetHGkQ>uhUGwvxwof=x(%_9KCX0K2kFI-5Srn2tUiJrcfM=PBSu
z5SsU98J+B=r;jD8foS9Hms1Tm+Xk*etZ}<XslEp{)_3~F@Kx%&%8AV1Qr>0l0rzpF
zm_GeM^LX4G!#a%Cfl-T}^u>dB@@GfMvys7WVjK6tKV{3;%1=rsHlACQIFd2Q8W(E9
zUQX`MC1C<bix>Bi{p2OxURvYUKIcDeq1}z9KZ&t|U0w583ex_#f3DhX?Omt$JL-vH
zn<_oPoI5nXzqDVg9v7;O1LA=bdY<MZqI#I-{qP88q9xDky?M5^mh0a+CbL^SdamYs
z%T1s9Eat11&ik<PSn+<Ij|<=S7AP&6lgRlWblb)M(uoqAxuISIO4wae^vFDG)Y?R}
z*=y^qsQzs0I4t|9pNO4p6p#8A8+>q1lcVyZ@%v^+TMp*_H1Gn~^pjAiiLFlvE8@Au
zw@;`7Fb`vaJR?Psx0y^>roS!DaRZ7d@C#MWXriOkFD3V?z9){gyq8>7Gn*j_wZHVZ
zh9ED9-b*#JBbAUm(zv%Aq`YKsOTe_provOvje^%V^xF6`-9)~%O|mcYx)_x!F<L=|
zmFP?OeF*nG8rMXd6%K=JZk?{2RTwt0_Era)zKY>NT^+hUu3)X)KUKg6eccpqZw{eX
z{1|Lz|A_W-x#dQ_a)E;3kxM`1f$Ao-1FN~W!89W4@ZBkihzT^A)iHJ>I&4Jz5+Or*
zu-oZ@cQ~r9VU{z=0ExlH&GbP?S%9!O7OM0roabJ5*o+b131oXuTT|C<JB^&;K3Sxq
z99R^<#9-oPyeMn|KL}=b7+TR4kEJm83ReiY^hiK<n_?`=Z-N46as6TBc~G!Y8#Tup
zsagiAjvVmBL7;~L%@_Z6267i#FX^?3Cb%DF&<(wofxzYt^(*IP%61UIO__0)Mvm7V
z(YEm)5cWtS8hG*+Zn3ez^g`^&){74>?^b#yR#^{h+YGvI*$yM)j=1xf1~-M>SwC|T
zN_TXSpP*hCP2#}ar1)|;Qm#SKt%e?x4#6^M>TE%d4VachH(_o)o_pZA2MW^<hP(KK
zUK5Yv8LcQ0Fk;?2p;(nONO!(8RV3D@GxLmt$YxN(c4E=s=gJRqcXrPiRlkmT`(l^0
z74rFQF7Sqju&xF@N)iz8da<(K|FSG<)I;uY>&EXw{_@_{ydSjC#Fh@*yVni3VxG)5
z^4-0_(&<f8o8lRm!rBm*Fb&xkTZRkB_-MO|*PegxCAIQ^F7uJeA;#;sVSTj)N94Ji
zTCr*KDc>i@A)puXIDg$jKw+z#Nqc0@!|FlRRQxv@?6JUQ$6qPsS2rHSewzqAnx~9A
zI(D&t-kFSm?Lfz@Tp#*wQM|4LFztcvEBHt8d?2G&#E}M1d%gveLzi+mEOg5hej9Jz
zDVEKJ-{Y3N^caqP^4R$}<{t>uA*~Q^%lS`P=Qm`Q+T<^_N~4T^7pbQ}up0UxK2d-A
z1H8EbC(usFMdfhO--IyfEz0|v_a5A~3ikcEQm5XFbeHrobYm%POaq!6qwMt<stIE=
zYEK4#>rQr&Zfe?CJU)XHMAz3nFuHMWdo+F+HMhhMIVQyvyEOZi_1*L!M)YrnLhaD#
z40tTv=5|(Xx5H5qww*)(dZ6(1tQ@%O+A5e`3heFP&3*QaR?Zb(@_gZoXGhI|kpFQ{
z|Ci&oLyP=K4*69X<lOKX_S9tk6`X2X5juc)0o?xT$VS8!oHu69Low#bt{X$I97g26
z_`K8hd?V-k=OB?2TIhK@(biLoR@86$dI@&ww_)A{OBJiTB=RdYo5TmI2(qqXevbaW
zwT$O`UX4cZ?t%+A=($Vl9W!-phbUvNa4BCGo7{0&ku5x}`Q8M$-?`Q`mUjfUOn%Rr
z={)}0)McEiJF7#YQ=#G(uv!sa`S*I-j-}O1>DRZ%t^w@QQ9GEEUa*_MRrY7^U4Hil
z7OK53TjhE;cbs6A)L7)vgYrCWC)~TAu3%2U%`nsTXPRGooR}RyclQa}^H7%&KiiN#
z9|Aot=}j#~<$98`Ed{@pYi=hZW8zcOdXl5?m7jl7IS|!C{=HGa)Cu!`XODdSMet34
zxc7w9UHZkEBYq(`UUugK@B~ID1FsB=jqbm7*p5ugQKYG!ic74<2cAncG2*j@MSG<`
zEqL!j($lSF@<n}%^w)46<;0O+QM37GkZzV2t7Z$loe%=<!j-Oh;xe1qpQ>E1Q_Obs
zJdj$AG`VSs;l$VTCpb(y#7$XmzZ4<4v2C#LzG4zKrpIv}K)w&t2u^Sn@=eRG)kBip
zrP3CMj8Y`|gd?pKYe4aEFFzAYns`g@?IKq$-#w&ej@217YO}5zT`|prP&yKjjf5WC
zlN$@41Z38AV+2sQmMh<lMQU6d4CBj?NbYFVeZVR*nxk?_x$?%Ct`ph`eIzS86hhGt
z8K%|i|2fWHagUU$R=7O>B}d9V1s5}OleB)4b*@f`VC=b#8y|FcIQ6Mek{LW9w6835
zy;%H;=5`^lTY?~4q2qF<&eqvDi4z)51ZH-Ozx0P*g#R>#T80lazFq91G(DJH21l?2
zml3lP+cgxj4=lhOJxN-uzb=d>G3L!#r|K@1ybLUi)oUJdxp#eMxr8tVuZLw8LVVYM
zzeAsU79hi}Zf34kJs|$p@3bS5g@*7?*cG~^HM}pBAGtbF_jlNh*4LvM;4)E5Hty9s
zslt$MwN+i5llQ&b@6x&%ZmP~1nfA+XgO5zV-r_b{EfoFo{yFj}ZQU2*yXVLVBqRc^
zM?Hk$0?Q5RZdG=QDKzT#aC`n{?W{urw9teDxE57r`mp@!kD6yOdHk-Oii3i5^m+|+
z`=pMp1a*gjssyXFr;s0DOz3)PqwFgt)h>HepH1s--MoZkHDMF>fW{i_F6eki<AA1J
zy~L8ov`dc$O!Rc2+vRUvVJ~$tUHDJzTE|t9dbZ$x3{&NdLxXgbu-&Poq4eQV0@p7V
z$kE9Bf<Nn6&q{H6Zokzv+;TrW;+Nkx^efL>KVNG?MqR${W3xFd|LjLG6_R8|iNR~`
z1E}Gc#(u$}Z*%efxk#*@k0OPi_b1Pnx9LIOG(#NctCNvEmmu`<pi)Is+uSRmrnUt~
z-nOm%_^<ic*K}DrAw@f`ZEM_(R)tFSs@;?EVtew&XzsxW6^3)dScLxBbJ3_uDqF&a
z_dC2Z#Y(^LgqRrjC$lZ9yXw`5YcqA7%S$*KA|HL!L3#;dwPSNNhUuNpLc2Q<hh07k
zv7L8mE3S|C817>=&%>HNU$@UP5J^)d3i|{IKyk~BDcOjF0>S<PWIg0M9{i6X&A0Jh
zg90k^d5H4V=!oZ6jtTLWtz6Bx*3@vgOyLs$h{+6o-Lf4w%(lL_a+);f4}z#yiK?nc
z=Re-83p$P=fexO|BCCFUK`YM(N9&VRCqP5Sx4$0NF5F2y3(1m|FRo89f!$HjnbfBc
zkJ*k2E^5=dBp0ae?wG;fwu|bF9MZ-bt&3FEnTVU#f|SRd*61I<;)6~+x!MAhZV*B;
zou-ZtILdK8Kh}aoQ)mVj8I{;WHOD-em|%4usVkz+Ou3R#z`N;V?89B7sePQr*zPYA
zBlDduX@eW23Z~4mlTj>p70tvH%sg?-)&;&UC%3$Y!<|}C2_TAl>p4=K@vLb^-UMX<
z{%OYtAAa9r@rQSw07P46vZ)Cxv5||h?=RYFzmU$(_FXE{L?U9ojLzTqGMdwc2>5)T
zE`GWiF#v&tuYw;RCtopvwf!XbrC(r`HgCzL`W&>>X?9KBp{JWex3%!Q<%1CA3m6d5
zFj|G8)RS$@Ueigy!{M2&6LP8B$idU8d&3_s{8>Z92R=q^_K%rF7jKmlEA|bcj&7Bc
zj{u>uNYYtt8S3<{{lj!vyuM~#(T}PGV&NDCyyZp#Nr>zjnZV5(LN*<yw&|wt(#$ot
zlGaBIjiVT!qp3oJ1A+IGe)=aPV9aIUrvk502?{VIeTfA;kwt~Ob}TM=t&f3A9n2Rj
z^=Ke$k%zznKL1w9#Z=juG4wr<^U@t3<WiNZC<9Cj_oK2QVX&+g(A6ALQq`@avdvQ!
zLq8^P<uv<fnv@AA%8PiW9!;Dqh1EW1kfjYflls2|Mq1H^@fpNun&48yVVkWF9s<iX
zYsN9_6<9V%xdJ}KG*?F;{VDpNIVEz4n!2Vs6c1lvPm3=YE5|R5JYddTANsuC2}&!^
z_vIMt<)HjhMZAEY?!(5yzMb_`4^q|jzJyfsFm>$-P5~2RL_+GSQ=#_3X5|$YeerD+
z#8YeI^<fQbj}YmZ`7fpnrS3ShO&rt&iCa{BQYS4a-jhH{7_c)-h>LC3yoFf66H7d^
z-ucC2xdVkD#70#7KMRdoRuAgX<na<hac!8DCfNB-X>fb3G`<Xh2CyjjZ0Ba*klqxu
zw7l){*(|r(i#={+aRV+P9Ze1F(oKI+6{saQ*Jaztahj&InM~jH(bM%fx#vl;w}AT~
zGBT-d3z-mf|6I=!T5(H~oQ}k*0*|5emyx6MBb*J;RN*bNUOcpN|AJMEqQ)u$JR@t}
z)Tl|vRrYsg(ZReWRp)e(+YlLcVO1?{P}u3IAha!CjMesQTz3FZgSG#}*);hSQ*{38
z-1zv9>76W1=4H-;Vz!DbVD}>X)u7xDM${gAKa!*|mrXzDaZR?-j96(>`v2zkeS)N0
z{rcLr6Ha7$9rx~)n$iAo&9;}XV4xBeh{Wt}z<OPFc%yKK?JCV+8<#Ckb5i3-NU0k#
z_oBj{4k!LE_fvmIJTen0N{iL_Y38@^t8|O;dz{=f@WdUrKERM|4}|lrhn-OvDd7cb
zSjvlsmIMKG2o1my+5eD3vG!0c^inh6a&58Aa#g$Z@9#AQs`aUq%joyu-(uRZ(^RIW
zD=M34c(Q<5`WJknKp&ovcY_{pm<`U#^T4V-kKI&66!|y<asWQrAn_mGd_G48)!!Cw
zG5WL1Mg5%x4in$v-i{rI`z0eBE($N)wX?Yk#i%TRO-ukCyXuyt064ht{&Fa-2muJ9
zjwY^JtOrBX_h12slq|jQWf$KhUgC}{jnBWW>xGuBzMriDOI|u+!}~CiK-9F^!Dc%M
z+sq3*U;qR1(wfTu#9F8`3wWl&M*^O1N&?_IesxO?ZQw}}O1&v78Xj%p3sgSrzlh}4
zL9N=dJsxD|akh}-*0CFYbEOOWcEVUezZ&iQf(0B+&p)_jZr@@wR)ukp$K@IKPf0Ws
zovZso5;>&=62IGT3jA;cwYpus3Peqsk2ahR4)+Xz;T~MS*54<L-@rAS*$D)I^gqJ}
z*6CaHkQbo?-DY*d$?G1Gw2un6d_XznIdg#?R^ImYsA_Or?YUr5Cci*Mlr(!IRG=uD
zSYZG|w+!Wa<R2B$d$jvj&vG71$t1w>X)6O671&tq{Ge;ch6*ryu^;SKh&Mr%?(JgT
z-DV*b05o}2vmb=E>x47&g?a!4cd^ZSbuszN0HJ`w9}It5lt9tsv=}3h?uZCcyOHSi
z6cAh!-5hid;L*5=zVS$90@tglb<7LPJL;Ou2rU}<(s{WEM5JcHe`vl%C1z@(4X(!_
zR=reY%bV$vMkt}c#FwxyACi_<i280gV<HNynL1b<Zwg$)JV}ioU}W)LBJ9ziB0I4k
z*a07ude@5=o6Nv02mXqKGnHX9U^PA{_`erK8v@kGBmpA%g!<_g-BvdLvAaNF2eUwi
zFM(CKzz>XoFXAL+S_bvLV*R1COXp_(qx&@VXGkS(2zYMJ5Cq0})DxwgWVNA5#P-+6
zEBvWuqKXde3<@&9f}3{8ijwej*yI2U_BB|k0$<srR(I;vm0y-shoS#&#U96bCJ;dT
z$yvP#NTb#tj$@`gls6m!d=%JAH$ekNW$^Zalz$-=737fQg&|n1GN>v4IZ3W2;75|H
z-r9Ia+^?qrg<prr2U86LJfA)z0#J~WdEu8@7~k)n>gEX1ljxDLm}0Yy0y}3s8o(K#
zGr-|iAj4fcfA~S@%HrbedF^~<$DF1B;DjyMvv-i&uwj@u9)SA};^@J`xen?>H(lgK
z;BW?Kx2c|%r##aAgUZx(lkMGGk_!L{RwnR0(fJ6NqAAh&%BrGO2yjhu7|p~}YVCFM
zlN}B2)0{n<<T2@U4GN!hCN-l}bmeo-JuvwRtF10$pxXZUHc2I0owJ5Q>Yqm#Qzyt6
z*^@=EZn$wfClhm{?JwSPk?%B4jM{f88yFMe9%#DDd3O6$9YqbAL5n5tksF1K-haNm
zFCObDRf~-uVcNBgzrv4{w+y`E8t8yI=N{KZm2b+r(8eb7(2|=B55P(=FW-#MqJA@_
zjns-yY&{=rijCD{5!^D2lxB^vwm964wka+ToOPQO7p(|;;j1g<dRgvuqHm1X7VG62
zS+Llen^V;Pn0LCyk!mw_mueusjVu@RsBWIN%f-T|Vy-~SD$9Xnll*S>wYjqH^4GPs
zo|=YF>dT6#d^1huxcVNyL{iVIKHPsCF-AQ!QdS(Fy>>jgsWI~PdnD^NWt}@QJNO_O
zE2oloFD`Qa;gBPEsD;|>H1YN0w|SErJ3|p;c#@-r$C_{>HrhAvsh5smO}-FFeEw?O
zi5x`D=G8q57*l*lezKT_H}Ro*%0^e9_Tv3_5-OcE)IblsZ&&b;c#qeAJjE$1KH`SE
zODF5UL{6b-FFHHp$7Fmnw|+T5Y;w))OhbZYtEoOaklA~WOl2kPP|#dwgZwr$xn_h6
z%1@ERkW5;JoPf+z^dp5%HS=T0dwd(1bHO-vFSG58gXlvHqJXPRFpsVL*oE*I!*QMO
z$Mla)YzfQw&OX-ryJ0kpjBFt)Kl?b2LqoMQRDUdgkSLRNc`p&5(){BqvDyL^YulYp
zAZ@x5-PkxC+`E))G&xcP@?c$Yj3(=1g3EJmOfW;wpjH&Qnpo=st|*PXBaKks2^a(G
zP?OO<$Eg2;Xi<~Du2a1@*!I`!&kk^x;#aHjySfVjRO#7~BF0m99T&ChZuVVruMn=K
z!|U4JiG}QsW(S`LIUM;Q!$TN1YBD12vQ%1#OC}S;PEIDQ9d9D8ayv2_tmr2{`QNDO
zWfovujO13u4g8sW;PxJSvrup${->*I`w@FT^E9Eu!)(}&M%~GjXs^eSuSG87FvM1-
zXpj)}5AjQ2YfLHoLF{XmoL?eq;7(FL02v+vRxnS~SiqnvryyB`6|hSd1fa#vLCn?y
zt$Ja~Y_z(V#ceo6xGHQ3f8K`pHhi(c9Z-t2`loN1znKH^Tpsz+Id_=`Sa6i#OWX)z
zyK1Cy<_nF6H}_-Vr4W?Kb0Y+MvoPGHc*rGsnsHy5RJzCdRQjtJ<CNtM)8?8yZX_U$
z#4_Cv?HnU!a+OmiXg$B)jzt+}-<_k1raHo~LT*(xS8Y#3rIasn)7m<46Aqn|RD>q6
zj2W|8@hkE^#B|L-xrDaREBaQBwm0#2cN|AmDE>55%8pLXQQ5bQC**FTkQ6RTo3s;N
zF5rjLV4%@Wd?-E$=ji;<a`Sp}nb)-%H&^`C4Y;&+F}+c7%>L-I59&gPc@&A(rYjVi
z!*{*b(i=WTs=|%t#)>bVrB~0fi^FB<=kXABNZEhZlZCt6DcC=X$eyEia$ftQ42>3U
zpH#WqZ<}J)pZIrIITE)GhGj0Z)-GcIs<)1b$eubn+$BMaE)1XcUB>n$7QK~GbK@?w
z;&#NzoCVC|d&WkGc$Ux{O(jRSuQQ;$Q5C5}NR9qFWB7&zha3{0IW5=R%$xkE=O9*E
zSkK8V`?`RTU!9bg0t?>Bo%h>p@>~bShW^VCL*~uYfgHyGCn4<bOw5<SZIQnHOP-%;
zAVQJ7#T9PVxih;RhgBrmFwJwCxSCF_&$P{`s;Rqn5}LUB<_S|MVWHY@-6|%S%#K`V
zxJ!e(KDq8FrHW}%;qz_T&yZE9;xOc<WhZ@u&Ux^@2`QR1MnY7N<l2ei+4!*}A55!U
zI5n-l<6@0~B{O_B+bqCoWwSCj60pN*`dd~H(M={#Qk{I<w}5Im05spUr;3$y((;E{
zF~~YH-8(4t$7uAUI$;u9ygRWug~vlMM&5$uG{KfGlRq1eWiBr&-?@xAJ(bZ8K=vPW
z`Ky`iWKT_s%bc*5DrKpC(y$`Ab%v{#GxpxF2#+5I`p%hOv4bs6Q@L)8!~YDQnV}Ta
z6>KpQr^O|WwdT3lJD;Mbwrj(}z6deJ(g#IaMQjZzpT#qQusOx-<NJv4U*`!Si6pw>
zBMMk1s}qRm@u+NPznS<nuB<^_`-Oss4UCQMOTHU4O+`$P5iRW>dwb?M=Fz1I)!8(&
zmTuQjx0kZA%}rdHheL}_X*y8n#*>?EpCyC5+A4x+gLQqa@Ikfy@9oa(!SZG^8(0!0
z2%5-w`IjFrD?`o@UcTZ{A#ib@@hEnZbl<7W(^WO3!V@#j!<)>>YOcIm(uB`EN=#>^
z%E=Nw#2TF(UO?Oumc%W`Zl+);;w!5-PKpZVYnT(4Z^hgbOkI2A%`*ADdi7<2+jl<3
zzc_`=<gH8tr*8{2MGo8MuCdBzfa7P`BNLMc#W#*!HydTOVx?Bm!WWX?zJF8{bDX}l
zj{h0e?|&nkk-3xHN%_u_{)T(a4mzl<@-E8%XG)d&+5~@vYx2E`{Tc``wNXe0QYA8g
zeq(~_?DxqI3$H)Hps@huQIg`4k?}#>%PaI{ykq8W0PRcZndFe*MCszThJL^-^=m6M
zK1vjtSDR=){32GuRIyLN9EP*h5o<`}eiNC*m>O4HuK{ABiVdORZ-u<zsPUMVAkJx@
z%|IL$)=_84WF`u%CdY=;h~?>LlGU^~0uLkJy$O}AF;s-x)HQ1#8|4185oZe&h1i)G
z5S}-C@%VODhpbxOdS+|}$z<WNgqiMRVaD<*b!W!b#7Cthi4hF#gj(*RJ5uOfmt1&Q
z@fI@s&J4+yGk$W&(VgR96?8vIiS&41k4e*~9}$r*w%}fHD+y1Ah0K5{6AB8s%D?-G
zz;RW^<*?I#9b#1MkZ!D8+(VsiY~Ic4w!S4^Gxu@?)3@)f+($N3DCsFZ<D0b}iE|&M
zC2`+qD24@SCx(8YiCnh%cG@jG6aB|y(K{y5VjKs5%}Kw;WgX{|uc^Em#A)|BG;%$D
zv_#2aJ@pXFj?*^xx+;sqU?20vwpZ>O#yD(Ujh2@o4GM+OS9-p)gj??%U{ZH?;R{9R
zz3$!M3I0#sN~QK|6i}CgE=NFOU8N0OOiNO~Tjv*cc(-m)U@HBrtGu43q<P8dvyCpb
znOjxY-^6YQQ4<xW3w?YnGgz6-ORf6&_O9-9jYUg%B|4f+yM#K;{gq;x!;W+e6>q%O
zos}v-X{2vN%nQn-&k(iiD~mk@?x~Ov$=u@0Da*%omWU14&*Nde$#Mp-gZXMNe;>PH
zLgI^?rW*M<OWYOT_m4Z#@yHtN#8|d#n=oNL36Sr9N)Gc%j4HD|kjZK@e9{iCH7j?r
zLhjSNswV_7NRR(uo)nUXACS6eBtPni{=`O|b+?^TLwOd~gp<>h*gL;Yr|l{uM_Fkl
zpN`4n+kuc6m$&__P3M}s>HH@sV^CR+S9j9WaLHZSXOwEGEGL=LUPX3(C6PHbzNRhC
zkyglF*{w`mSC-5r@S^+E@*5a?-V0yCNe8FFImxaAq|3b8;tR5_hr6E|--DttkxzqI
zv^)g>j1{x|#n@Wsb=!OWVrJRbF5&%$BNm?5ZdY^$CQHObmKGf;Bh}MWAMQ>txcc6h
z$Q|!W)>wS;cYjy4=+NTba$Z9i>3@f0Oteqxxz;_V0YhY0>-oydCi+*wgh$T&PM}*5
z`)n&S6`ilg#~k$phfu2tnfzTqRcpW(yR{#LfaRsqhJ=ih8R6PaX!Pj5o6;&iLJ<7a
zgZORSh^z9|kQyGA7zlQ3hECFlHE|sOerI=Lpsb;nA@OcoIc&+J3Y*KfM0C@3^4Lw-
z#DJBDCUpE&FN65j8w=t6gvp_f?XU5(Ueol8)lvh#uiYE@-BR-nUt3_8xJWSRo?(-^
zF;NAdn3DbJc683sHg+0h9%M*W^NLjlyJ;@1Tq~w$eMxXuW6g?-gl+KjYeiw;F;1PS
z_?EA*jbp)O#kEBUh=ZHzu)LY#mFU$~@~$=6bjljA19xh@9LRKa+k2;7-_DFFMU;DW
zA>5Zp#V&mo6tzeLPIq-RJ=#Y_kqLYhWB5ag@lVAb<q$|#vzIzz_{p&FY*f+q4;QvP
zdL=<Lw-~1ysclYJbLNdYO=5!2>e3OhrKu=qO;xjUwpx#=Yv$b+A=<?l>w0%$j6$Ki
z15TdWi9rgL70H^LYJ#ueN0aME>z#K**bCeRym^!zL}cB0DIh&~)#so?%N<f^Px4Dc
z8jmWXV53LI+Ta2LkctD8S3Lxxre-l^W*;y3^VMVqOPe`9=z;Qm#$@wxh6Af^8lsfi
z3#C`+Sd2fv%B6xDd<-PvN<TM<EB<_epk2jBqRPNeIo~jAjU)A@f%M~V`~)Ca_yuwG
z(V**dsMfjy^Q4TSj->W$k7puWQMpUEsWY_dnV_EzMYV~hCu8P2B@M}B*KN(M_Lqy{
z?F4A9m-kvvqo_Wq<WB+KySpjL(y#T=!XLy8Ysdkv<iyKnK<~ZV!3_0Jc})yYi$<lF
zFS#iVQRA}W^^l|rb`8vZ^yM8Og;p>2u$rzfe!43AL5n!`Zq5+OOM?;bAX$J9QFC=y
zH?a8mZuss5+O#eyw{4sXc#Z`)D0I13V#N)_Hm9-QZynjn-bdhd7e_{FAB4^>zPX(N
zT<HV}A5^>E2qMP^FAtxW<DB}13iYr`M;t|n8atDq1<9_m&2ki1d5<>iw#Bc~Xkb0^
z_NV*Nt&U}5?P65_w5-s2H;xU%-HuDMz?Z?RRXcR|5haKk2hZiC9QBWYcj6r+gye9d
zw4}yRgqT9=n(zouxsHk_K?7sv=zIh2`sn;uwa=kB{I-l*PCFZxgj!IoDnc_MI&4HZ
zONf?P62yOA?4!scH5U&S5D&n{kq=Ngu&Z-qPuEEHL%X`TklC~k#Wph}bg6JBPE9VW
zr`wE8tf~ckMRv6C6(1OVaDyy0z*n5rYdQUFhLujTTE>bsewEOXFWMHdYVelHLB*O_
zjVF7ZE<c_tV+2!e*?CGlLhGxmTACCu4blzA&Lg}=Oq&9I*<Z1%>)vDus>J%Gi)Vc%
z5%Pgn`Xo*N4F{nGhh2DeSzc{t$QAPrHM+k4?f;%QN>ETSj)@U9;hP>~!J~n4dSfi)
zXI2CHJ{y6CtdxqT!t%LdwHgb0&*tE=rmOV5@^_v}&kj_&RGgDuP4f0-$B`$R?z@wc
z>(exnqJ=+s9G=d;18&HBQV5wiAWGI63LX}d?+V1r<tW@k%J`RK855`uSRIPB0XHTe
z?rL>b-rS8!WwbC&*T5Qcx&^T(C*E!vBa!#RsvF&Hbt>S*qV2bls&%PpRz?&%y?OHz
zKg`TMcD*X$vf*(+3F|l|#YiA4ob2b7^8Wn_-@EAqI=jALd!8fWYEs~foVzso({<#>
zF$2+$RuOLM+}Al(<a0RsT2b=~hHXwst)FW{ylSn(_R~cz{szs{=JW|s5tYsK79wK5
z%!9sU_gF}f#t3DrWc4ALWO%w#<$j*FUK@R&p*C~Nr1%?_!lw%YvlEg`7@X8BN`iY(
zD?-8_WSAIx6|LFJm8qsZKK^_NZ(hC{(}DWqdF3E_SL|A*#8>6MXzK0QN;^FD9>@2~
zPc<wP$u+$r*~~!yQOv~}StSzKyKOMPUgv*EL?~1(y8D!-8LzoT$Qs{f8gc$s?u<bQ
zpP^Ox!H{XaV2dDIjl*}e+Z#7a-PAF*O1C#jY0m=2+Fj1NYY-a?8DyT5h3sj?`J`!e
zTQ#4~-6p{;yk4g0O&DVdVyW>ighpDcy9s4U{z+qDW;N`298V{H)TMv8#MhGTIUm49
z+U3`aY&b=GRXojsobr7Hn+5#wQmnHZ(<4;<tocB^r=js<;b6N#!L7XxB)&7>>&E_?
zmbiDhefusv{SXKZbnIk&{u>p-E5Tul!T-sj?Uzx}Lt=6ENs<Mub784FQ?kV<6DO+y
z#E#B!hs$pzrEbhvTxGq`tJ6sW8gxFWap^i1=r^;GHt0GV9LN|qcu93RxI@C27<6Gh
zd7QGcHw&<}+g+|N3q(^HDBl;$br<5(Qy)aB1cxgb7WJbG{TJkWxC~ut7^QykWHywc
zZRyJEb;u6Xb=HEfnDI~22PY?U)nt3C;THE%qp3~<^5li7J`#LA`Q`hN>imcXkv7~G
zAKPc?AUy8gW{k{M=s7>cvDxKlbLzeSc)`N(Sx`l2KegFZXg4)!a5POGXc(i<W@r>D
zJDq(0wiCJZme~O5rS1kH8=8H?t}k!X!BL&yY5=7yM2*|M^AP*H8!Z#~Pai!>Yg`iY
zeRP8pR7*i*S1Um~$+czngSTr+>jdhTRnmsVA*X7wb%B(G`HEhG3=xwAcTFfBoyy)@
zcrO|=0w5Y4JP<>Y<~yL2<tDuxE>i^D!?L;YHwSTxx}yo3Q*swMP_H|$(s0ROoc*F<
znr&N(=OvIo4Ku`~6~RH`u)I2)_g{_Ojlld-{Z?HZVpzIBm;&$S$e>|wqipI0f8q%D
z*eH<yy*rCC9{+?dui#6L?gL52=^E-q6HET8j{PSV8`5UY(&(ajQOCNiA@J{@ch1$V
z8prfP7dne|zmV|0ifZ^<&2YcFcUWKO7^Pr+GC%3;UqosF*Xo1@cbC66wdfM8rh<h{
zZ<UjJxt54l6`JXHxD+dA&AcATqvW`nmnTB7rVo~V9fUf=MChSp|JfWG%BdW?C)Xk&
z9TkSI1)XIYR}ca+Eh!Ql5G<s*TX?35N37^@F;WxRE-#8I&XQvLqKTC<$=vjTVbw6n
z2m==I6eAGC!fH<${i#Y#?aZ-H#AphNiW>cHCUv70W2>LI!!mGROchq;_3SuC^{V%G
z`ZSoEH_h9@NCFujB~5DbP4D3kLoSYZDP74mzH3oHzwFuSh<7;uVtK3yb7ZzAnjI|N
zA+lO6L5tj1WIuTc>KNz$6k|w+;9l_S*kj`9zO)G(e!@vM`Dl1}{r={(g=59|w4Q34
z`I9eU?|wn{>p$j0Zy`ps=5>O*tnn}I(F|ZIL$w-OcgN5soqflZKlROd*77k_N+G>;
zQ{NDQC;_I#B|Rlxk@S6~nS+x9Qt3ebZJ!j5*O3z`G3Q$#IZ=TSYy_mp-ZJmguk0P)
z^cyOG1=i~yk@z!_Onk1*mbiE97dj^)KgJh&=q&GWyNFk;l~>(D_R?I7Fwbw<tE{Zw
zFN}=-zLplfWxP@X9kr#$ye)Kh@=bj*AW<4;q$)P@T|&hSi}XqwSkdkHWAPZXYrFl|
zg<nlG95+=q=v(p|7kah6uDQZeK^g>9SMleXeMmnzPCstht)ZGhoQ~)?_fAKqWU9+b
zKKX~m(dKj|KEdeC-H3t7P+G~Xt33B*h<VR|>D?!bbEFtOBccJ+t$vPoQ{)HlF}6Bv
z0;f;*jJ8+y{XdsnQVt-RVMGXJ)_;x~d_Ox^{pUAp#cd_Yi8}sH`E6yJnKo>$^QE3R
zDx$<U<phfv738sr7YL>_2NKe96T~7|^p&@$FL;}P@P5VX;Eya|*cbgDCE=!DxgNrO
zd1|s&(NyL-7|oRuE%}UT!-?<Q8t}|f$f}ab@7g!-&sMgF;&=6+2ubzh(0N0K>#?PG
zfq?(qx!>f7>ykfxiCz`fOoi-$VSxlHeX(^>i`#F>SkrcT-q|{c8TWpEPW+<kr@-I&
zlGr~`w)m3}#V8~@K4&LRO}ha}45cB&8gn06Abt`1*)gVtcpH&Rl63k?dRs+hkJZw)
zAF{N~_{EeQiyrv?b~;Lx8a#$vP;~wu!OWZR=6m{x;xH*ZFP}SBFQGn7Q{qGGadoE-
zU<oZ&#0>q4&d9wVwYSpa$rj}<^M7rZC%MZiExRv|sm^G}z@0DHe8~nnV(%WMVelsV
z-}rw}=|9$*cqBd{O=Zko_n0vX6_2L6Bx=58j`twe)OpCdxi<dEF*!DOYH&SB!oI&-
z&(5e)*>x6M(Xv(kS0zLDG~d=Jz}Q&~<>akjp?%9}y5$~I!oVD!G7D5k&jf^65LU2x
zM)o0}&UYtj<m4?JEzhoQi>bmMhE+zPe`Lq0JrSqs-qP->C0T4?p8l5XD7bi~SgUwJ
zQgbyr^m6jMfk6o$&9*ynCMst40esi-Ilm>-z~87I<*1@YMHa5z_a_<@iMx|HF5VGF
zZ)_unOk~}dZ{iJ}ONj3=vLX!~oa|S`ek|C0VE?i&jrrs>0?8lKJAZ3h=-M2tD^zwm
zg!*fMthTC(b*d{<{{V4IWqd}+ElK8Q5~_<bGZ^WspnOr?JDTm~w+4)9s`1Q#uKO>8
zS_Y9Obj7OOw%b!VhlzK80xNe5dF-%8?>_AHhtnCk*&P~WPH}>DzLevFO9jls=nqZT
z>>^kUK<?*nHtK&F!Roj^7j6f3_DK^8#xwIc97@$2H@DLPvaL0KT16t{1f#75>OiXK
zzYKNmVC$RV*ba}9Vn=s~{>x0sjp%$C*LZYH2wmf^6Uot$uK%*cziC}Q#E5l&sjth~
zR`=B%g^8#kw=Ce#J}V#$EQVxjRN=)YecW+l+)UBl>Puyr|28EcBgV+pwW)JJe3HiO
zv9IVsM@Z@xKmcOQb(=pc%X79yRe-4JFJ^rDZ^SPMM$v;z{_uSzMc-1lwV$E2F0{G=
zQco+;hNXnZ{Z07(?*k~7U?%|mih?_c6(o8)^bYEAwxtyrqj-!L6-`CObO~gp|F<22
z!zI>11neoQHMZO)8zs3!#Y}@21zY8WDf&o%EzbXKe5;^QSBlBk{H6ibLSrAK|8H-i
zhNI{Df)B!p`g8{>>ix>J%Jsw>Gl20b{p0qP>VG3R*D@-8@$G>ErlMc<2oXdV!*L8J
z0+Dr1kG{!fa#%i_6GYZv7nA>q#dlCpP+_c{A!^71H%{uYIGV~!9Y&Wd&66NMI_TwC
z0-9$|9Jz2Nf&KUS25Y=6Acm;X6L83mY^MimMCb37wJX~>>V&KD<szqu-BO_e&C%Lj
zTAY-d)uG!F{g(fQ*3^H05f_5x$!3@wl`}7q8pKj@Yacyi<!<X|(8pWt`1R`YT50`T
z0mrET8@CRUS8FTER$t$it)+!10VuT@Z8qtD-x45yn)L!7bUcu4eV)_7BToA6t?KoR
zvq@F%$@ie<!EM}Gf}d;tmR|FkE~&c2yMG}*vKMmrOl>>9G(s+iY2yE7sR7wGS^I!j
zl^tBYGPA-cwsL<|c^4;J#A9vI`Ta6IIv<@+6i9ae7p@Bb%T-#C$r2%t68=BE@oGH|
zu9QdVs4Cm%Z<$%y>L6--q^b6JA~Sw>1z~Z9t)z}ruJ*ak>@xyjasrSEmndLA{Oc9L
z8mcd<K!+ADJpZSt;3`AWZk#jZm)zBXnccNA;|cSOt}ice{xbCAg}t8<uGL>Q93lfU
z@#?JUAIE590C)M{e%Q_*-bXG2<Pm+u;~McJQ$S(aU?{U&0vQRnQGXq8&8STwFVHKs
ztH&l3ogpdC)O`N`fulP%;HyVDKxd1Ty2ByBB)(g6XD+8!-cGyT{h$Q>JqiHx`DHd~
zqrm5_0o{H%;PIamFoU1AlfhgL2+sL<Yr^BW{5^zUzu`l%k;jl_<4*X39fo!SA?^P_
zAs-D~)orGbtE9%3-^~Lviv~&JISaVnds5r@4|{YAnJ9-4AKl75Qb}8&#Cs<S0#F)a
zAYiNhOOTj(Vt+$6rxHGXFXTB%;YDIMv<Z}<TeDg}mQiNT{N(UEd83qN>=#Nip8#gM
z&L%ra7D?|)P=rwZ$J_xz5{n^LZ<PnkvjNg)p~T`>RB(tjKm!TDcUI$wiu?(0XU4^>
z2VZNN?oE8OBomWwRR3G;KcJ2wzcElw^;+(zn~VLfs%kBX3U+YX=(3&GyxuP~yF0tj
zmNpN~<P<UVDh*Qef577TX1zXB{kII|-v^g4AVt+02th4hyd)uNGBp7VYIf-ikE3-7
zWXcTI$D^q;RHr+duHk0fIsZlnpw=wt8%@j3-H{*RtF>3v4>M}D{C>BK9uIySCl|Zh
zr^54jWr3iZ-gjev+62lcegFNP|M<H&I7kPkoFQhoY@t$EsVND;qKlI)9y~8E?Wo@%
zo3BYzi6UeHOU-`A5BrbzjxwGdc&H|H6HO@C*84iv8NFXRo!vh{I70AT>IL$6mjm2~
z^`e(UoDU!cI=QsSR`dTyUIoiC0L~cR{*S~oJ5fFzW<{uFDMej${ycQs^Y|pA_E_v3
zl8-TT3J5)`f5;CPCQ+a!;$Q3^j6{DH!|6|Ld72mZ%WxNOy@SWr`DwSAc4s&s$9nS~
z328Syu=^e5M;rZ#+mjc11pE-o`_0U!KhX$_hIZ=dpOk)&!t1fP47apZh!_)pe(fzv
z$X^{15fQg)o-TyG%7pMA>VuC5;9C3k(luDY_Dgg}?|7;Ox(0bZS~LO>$r~T3|JYEV
zmaN-42lm@z-)7`_>GWA+{_FLPo7@|xNeb(mm%Lu~Z7#kWmG7*6l_drQtwN8ubrZ6M
ztc#A?#>khPpOnQ;cuq;jRcKAe`W#U1M4OySSFdhgH97=bH@Z2q97d{&t=br#-E7*X
zkA<p<pjB>Pum2p*VVyotJ>0$bOSgGB<Y|pKN8q()f9-kj$BuiHMTVSMd6>m`?PO`U
zYUi_DelkDVS1)gE;eR71K-%DGnF8JDd`X392g_(F!4g+WP7dF>@7>;&7ct4heKBg|
zK8$^@RFr#72F7AmHe<Td+{S0tggI(n6m1QueEM5x^RF@0j%Wrszj(e4YWU{1uytVV
zIttYF>niBB`x|PnGp>Ob^W5(f{cQCJ3aia8WO;HL+OT1M56HVWZQ^v^Rq(@62&Rcq
z6gB7FYWdqa_|O|pZ_w+7RcJ<(ZyrnR=QZ#BoJz_-sMp;DmCG1P{@fD7v->?xV<$`1
z>I0RvjAHJF*q;c${1mbi!%_NE*o$ePSmxh^2x*E$4L+!JYortCed&=swPwNQcc$6$
zp%`4c2kul%MS@IO84r>50(Q(YA`nvi*TR6(Ao{ahwLI&=dlS_3uCuR3;O&4$w=n@%
z_38bX4~r(^K=ZIYvHslOrGoQcimWR2IgW1;yi)_b^A0fup4RVwv7f^nt7f0-v4uUS
zF&htIB<>rv7vqtR2jAx4((*bFp31Fuzp9*<-L*}J{mABi+nT>j|5v7Au2~}M?7TrW
z`{~&<7#J{)<a-due{fXh*)q1jej|`6UdekL8M<#`OpgU>GK#p&J=l>ZvJPK=I%_&#
zF{Hc<C?6gE#Q5(k0sh92d=Z_0y*bX~dtyXFMWWwQu-3n?18Y72iv#*hcF>$3%b>S%
zsJiZ$u{ll&5>G{O|2Mw;{|^aE+}rQ^Lx((0nbAas$tN}4?f_Sg<;HhYi*-NFcm7`Y
zrfkY@xOQWd5I22V^4j}_@TBK*0s?)EX!z|m<R?}0zUR21U-C`0>xA&ri(((Qp&{<H
z=!etwwH%L;mP%s^gLx4|zibQF!snTG&K~iiGfB$Z#be$t)yvYUx-M;(rGh)ifw0b{
z7?h``a<wGRfHQpK*M-a!7*x}bzq3R<CyRIOL4S+0{(bOq0B~M%$piZ{WUg$@HR<{S
zw>{iT6k7h$4M06yn|D;>1vl&%zVl;c2lG^%{ipaN&Hl!5VKW?l8klXS-)e-3R`B2j
z$`2I1?`leqn*{DLwZ{TfqW@8@oGn`_(o3|oL=O=rgX?Ac))ggzo6FHa*l$wPaJ)40
zQ23!oj?nLQwI7vw?qhlt-R+oYIsF>V_O$QLBUSm8R>ZE((P)vYy;ohHjrT`ho{Dog
z6NGVnzw3{Gc4CWKb(?%HFW#x|{^031LgpZLxtsq;Uye?a==-1Yhz{EUULDrVAgGXT
z0Q3AVD}Mi_Z0esz_dXl^CTqUuHPaQ~z3QZj<7pCthF_XiI4?H-^sb}oGU6ZF)`snT
zAAj*5vcgUoexFTTT~*jBkpnSzy-#t^%^3cq?neBzYxP>fY~$cwNPvu*un(4yLw-1q
z#yX>!ibwj%FOR#pn-!)#BEJLjj+=gTP|DREl}T<Z`Y6g=a(wThp3wW7JCZedk7>2K
z<kUk!>)RE)BSA#R*X>j$7-Kos0!)t&cQ?6oot(qX*S^*egS-1W7cR5{PZ_blvUL~c
zO}=~0GeiVbaNd2n(to=Am$vwa=CEUfpWiBX-eWs(pyVzZ(Rl?b9sl-TGmEytNMZDR
z(h(X;RLA^vHTdCBob7IYMX*ANm@bxvsg~2R{yJ)iysv%EQ2+Rw%s&w{kF+QJ|J?=f
ze=&t^0Iw&0e1wOY06{*^YoUtIpNg{5JTW>i0<3~~Mc3yb(-d@2m#cEtl+7Az>lMbZ
zVYVMnKj{5=q5CRar_Jr=&c(A|OZ7)bk2H|zweG7zwOkLW2!0$`-P4fe`GQ~GH}eQ-
z3Qk?e{Ff_gUe!@^A?3D)rm{Kn(DC!o$u7C9txQhqde8f~6Ym`_k%=q@qFisqH1yuy
ztJfX7!R6i4zW>8hnKJup0Xbfi+gBw5Vb)K8A(jAF?ZeRARze_h#}oc`YFUTX+Ofo#
zCYiUlb}ap~`Tr0fn~e3R6COLtEuS)lN~@U68>*MQJ*o1pDBr4TL>VWp0UCY#5g+6L
zZF1-?z76J2^3!N_8+W^#3>1?tQ#pFSOLxBPz;^Q?V1BK%sZ1rM&i~dk;5llfTgfyn
z-`M+|tZ>dzmcccqB2C*^sdAPOa9&k|q-grbZ5!x3$H+~qOea|c^tU$lKfsLy#{?FR
z=(LXJxFT%+sY;KC2CU0h=6N7y_RX)IW(9WtA_r6}(+~A|UQ+4G-x^VGv+C;nis`aC
zfuL83PDa{)=n5&tf(~YmNRH(`y=RGrO%sZ5YCD1p&PjF8=O34Jznu?N(*gBGT3(Rg
z&+}0D!z?!g!ch_Vbf*>j>&xn+hPiCOwr_zyT@zE+%q<#g>-d#$a*6<n_pWTw!dh2a
z3s`+(_|pcy5%m3&N^oMecl$fQJMBVMOmvLO%@k<%+g1Da@O^Aq!>OtOtgk$MB)+y|
zAn!c=Kc*;OAc+b%;@<&xg`N~kg*{pmmO^t3*zGJ}MYzHC|JbytAjAT@C697_oA|u&
zFDZRsv5wWa8!_g0zn(t1p6#F>!-tQGzuF&;WyWhv)3{JcIrn4*7rarVi{vSq%oJDP
zcuJpbJnsV3)XwM%rRq@l+X)Dju*%+K4$Yx}b0X){j={kpm2%`tXdBfRzZVBd-l<=Q
z@&Z+Qyni02+6?@y*Zue5#qa{6wsK~3ki%~%ygjZz<8yO)K%t2Eh58Sf>hP_6YkyR9
zesx*FKPsF5waXIBh-PsfX$A{Ne0~OLUa!uViu`u}+MzG2Ug|~uoWWeOL>9Vnj+FOP
z+GD6t-*Ec9#sXM-vuVyD-R^r68BWP%fxm#_xz11>lvT_$0aMt<%FAfxHp7|+c;n02
z!;(NQ^Df(}kl(D#FQ;p;#AtgvcO|_qFQFpwOYc-c?)jG;V!zjx84)VGs!e;=%hS7=
zd%r%d`D+jZwe0`C!l@Sk#B4)Yz{39cpfwgJ70h`nPHGl#*UtY0<A&0UX#AtYgMBkN
z&zzG1kBKdWmqR8u2d5^f4GUP;V}PBn_M>qySaVNMaE)=Y`>U|AE;mr&tq+rku)JP*
zEABIDPO?>ix!$#9!)}=hjpRXAq+lw0x>yh#R&uQ#|KzH+Ry#H**$4EbYQ-39`<v$c
z7eQa3#<Ii9LX%gSjJ3-DL<nuZ_pvmD&M3)daDq2!zoy62>hEko%I|x}30@1Pe`JNn
zE&lf}CrE^1X8=~%&`EDh!`BYSk(l#uvvoVEtKIKDi_oO0g}%fhHxUuP>1i7|d1aI3
z@qp}iX%?7kn_ky!ve!lo!Af%^BOL|D9XPnADv5uPP1BjCh!^AR?b#In)yqd?`gnH0
z<J~#ZXmKptal0Ql(n#Lezq7o(-zEs6ZWDu5#|;8y_y6G^a7%qVf%1B-iP|R}VQu^g
zHiwW^mDuJs*&cK&QK`g8K-B(uEazWlq1h2`hIr*HptE;TS6jTrI2P7A=s2<U)?hmq
zr^$O|KfKC7509Xc<+xuA$G3Y7ch`&)DcAE*65qaMWHg%7B<KCykL^pyOtg)8?dcyc
zB045P9|x!1hl?myO&Y;k@&22PfoXtAbFO2&$6l)hdE7l=+{qH59Q(8{{(l(z>aZx+
z?d>h1V1d#|sYpq8Gaw4mC0z=_h?H~>A_&smF?2|Gw{&+6AVb&CFfcRly`Hnr`E5P>
zyUzK8iy3&YnJ3n>*1guc*Ry6^5RQjm@RO1%ljav`JQUfQ4<Jsje^+fm*RV7Wxa}4`
zD^J7#E<3=;yZg%?`uMc~C_xc#+R{KL^<SrM^GFZJ@O<A402br+9H%?y8=e@&vvG1B
zh%(ZBPYn!aJvxPNo3BvRjVDZDP_+y_=~Pu`uB3%X?TU7m^L{r{IiAYz*SJ1g4)e1&
z%dBlKNC4LVRm8L9*ooOL9Xb=@Y-(jK%n$CpaGQa6*}|!bh|NAw{ooY5J6WkI>kxL}
zEdSe~Aj59UMQS$}YEFW#Dz3c>gm%`1Fw*@pmGPjKthyJor)mt@eRGrikDzEo^yt(c
z<Hyl!Dc*aYYm}uL&v9KZqWM>wJ)Pu)G<5p}GLQ#4HyZaZKFk}hciM|kiqF{gNDqZE
z>F<(X3Y??yi^J6cRB`SmsyxU)c*QVdb?6h6y1l3U>`oQ*Lrjtz04dak<{7oD*r4@4
z=DWUeSwJ+RiZO?-F4{-kt0fKd^$wn?eo$?C0I@M^@?89GckW-4J}B4D9!g{{4P!Z}
z;uU1HX7OlK&V=W9vC#*}{;~#@kM{k`Lt-(G@rf$0(^eN86G3WXSC`dU{Crf@)O@l~
za7^&d+*m>1H~nn6#@c8}#&`3btfuFcmtHUG`f1O!8G#7MS9{3jAEL<Kp{r+o8bx+)
zt0Iu8QdEzb^iJ}FCe>&u0K%JIK|!#KFc|vNk-DDL)1jW9feDa(ZM7DU9ma~$;}TU}
z{c2IXO7(7x6a!^NVkt`<>VR}rb82zBCpzaBRVI!V{0i_Es+I=I<_iSV2(DHcs5*wA
zR!HLu{Qx(BAL{>`G>-eVl%jT@T)6{^=?g=I%=s8#t95DI>HCCCi`WuR{kcPjN(O5&
z;o#MPMC$JD>kOnxd41ixLhS7{vgv)i2vMbm_e9Krj!AUuX6Gf-W~wIPQ<&2=oRum+
z=!q*TWhjA9rRi{K?1n5uF<^7LNj=fvdGGgu{M~10SR`J|PaH<<o3&Ncu>1CBVCt@!
ze#zp<j#=c&cW)>D+dM2Mj>{fOz3SDVi0ByV-zO{G!U&0=$nu!AH}f}PyQObo>1+RZ
z>%ER>Wc&xM_9udET*js;teyz+<pXaTPwwOdfGI8|Li?*0l6jOJ5U2z#pLX(0UN_XI
z8)~l7`5Bs{cdSTgv@O9-P#=ENQz1;_CINBl^R7EnI6XAeGTMD!dACi~%8H+dhsB)F
zvCmok)ZU>gJ`0xm!|NoRpAGO6_j+PiN7fCZ*I%c&<H?e0ZS4RGvfzD*RGX^3vIy~V
z>eq|?P3dkg9mADvheS8m0f-b;U93u9&6@rGr)O~`NwEC<$sEk@!6?9S-=PNWmYd3e
z-QVp%^QH0^W?8sJ<doL~bp5Z6<dQ-M031w56#pwol482XD5xmx&MtV_GN0+-zC=pS
z=(3<37z<h~eE~S?Ck?Oh=Ix{f*L|*=Ja-GUZkW<+rUTWjDYo{FyrdC%>6j!sz)xKO
zNu_+Y5+ch=u^xOTr|-I3M91#jc$8wC$FFtRup9PLaMcN10%%gbG%hifPyWL9LVpPj
ze!^~3XaO?dylx>CU0k3K`8g{<<}y;O@3+v=<4=trPYncgNd{Bc0H={at31d&B#~~w
zb>eWXa6tD43MSIHqtF@w?$8;Mf1RH~=rB`pgFC0YLMGYw=HX}Jq506}N4mV$H5do&
znt6w-N1!v9dH%{0I0kjY1SzpL7BVdt`-VaraW6*3vlex0rmT>i)~b}Dn^5o39E<*B
zVTMCesp~^|yq;{<x!1R<BvM?@6|^r#DVk7MM7ns(zgcFVP@vahhK9%yNmS)i;M$oN
z9ZHqmT)!GTvQ_LIE^x;)m)lD%-^^YXRcSEh*ZF$ODa&u^JRtcE_lL#)GtQo6Jh_t|
z<Jy;~{mJV!8XGrK`?qiY#l-&uE(G1fGR><_8z3;Vie;(`ka(_1^lGaqYg&|tI`~yN
zB2g$8;=fW-ByyiudEd#Mhb<8YM?NVRA~2c0K6&yK#{oM2cGIG_dE9YN`3WtRZ19Gs
zNBb?8lY=e=$Hn@+n={4pGwVFIKpH4;Zem;(enGV}!{Wcj<}aVJcewia7Yp&9g9H8&
zMEy?`Q%FcJ%R)_9VD!i8ourJ*)?(Z!QOOJTq6Y+Z{nsGs)fo`lZ)C_BZpP3s&AXoC
zbZT<6(;pM@+dl3%c6(0ru~QV|d#-tKS-R&J?9S1Ldxr@3WIJ`SZwZuI20ran$cE=I
z>ID7QOxy6+pP2+1i?D1N?svW4HT?4faAE`sN9d2Xt5#Hmsnvhn?LoeRX|4VsWlOt!
z<!7i))l>DZjqy|Ju~%D;=;S-Ngba}aQ}honbV87954L%(@qst@V%7ql=;ik^V$gM6
zw|8LyG=4l+|24M)6teyd0`T>IcR9*bAm)j$;$PKtIeOEqy)~eNUu(Cec$WyK)D7g|
zR%EQ%bq~g>vve;K=||yR^8~+!UjM4#{`DG2#09~@?;|bvJtcfk6dIZ1H4m=uDseu1
zR`4<hMMQp-(Zrf)l4?dk3_7vWj`=mB_M77U`Bs@1%QO_aG%zY=7sF(n%Kn(OP+|0e
zbxcYoTW8E{VvgF7U9F|5nE-gETF(Y{sU<tu=#wPePYuMbre9^DN8vd{Ize*3Slqw)
z7GQgJRIMY|hqzG}K3RA`5U2zxL+>D&nv2Uv@VVXe`RTOt1`QAx0IE+}h%f*98fyUX
zY>dk|M&8WwRz<>eUk=kl*7x(vnz=xBbic!z*)0@$nZn97+z$kj-;Dhlll(s*`tLv`
zkm}%&Yoz0s)t!23W)%>>`!V7J6QR1~LyJ(gZ_JJ;szmGO_Z;9%8C2VTi~I(zJxP#*
zJ3!paol@e7bjtB5AR!e<o8?u`|8{O#-amcA0`cXOTQ8X5N@F?`6I&f`R60BKG6IC~
zwW_Vw?Lv3kJQ=&633W4CKWa~-nE<F)p(vMtRoM&rAtiA4^<UxlzvSYtn4C-jkP(Jj
znYy+9@DOJp^Q2Bumu381t|x$I|3rUlcO&B}&!S`K#}B?unPZuElVTZp7%kHO;d39C
ztZ8qO@ju<=HoaMdBxlMzrZwysl97h$Rsk|8@YtNMnK7Iygi-%t_Fu>I_piNfKH@<4
z&U0BJ^!%OdgRDC6XN|_Hh({^N(<hX!>%|d;|LS`Gd$a!iB9#ZLotZjzGA7R6-l3;v
zP|?i4Q?4&Hp!9k*vb5~s>QIZRe}zK$(wfHhz^_XDZ%6e06;}9y!V`?>=mM{VX@4@q
zd?LS0L!&;Bv}rmak9uqtk!u|{$DMXy^pG8X%5qfmA%LcvTO=vUK>hNm-_1Uv(4Id<
zE7cX7fa_7uM(T@<OZ>&Ow8#E<9Vbg7MXt{|!2eoiuR20T9{SubpW5qwxV~6C^gHH%
z_<at->VuaW^tjNOi1^GV#$QQUk-grk(KAFaGMe&_TH0s)y-d=BMt2I-Jp!V^1s&y<
zfreuwuJ<P>#xFv7sSB)au-lwHO4yl-c0UL37La-k(ql&__b`x<$Gmv*Hl%*M?j1}0
zwd8!WLC#P$X0~PBIqWOAg2HHy?&4jGnJ1&8^vMbXl>}*T1l85iL@db<{xzWx?2An4
zoA5@0P*ft^NO>ZPjGc8<MVd*Q=U);pUI|<}K=``eeI-=pNzG#nX<r&H=L4x3nxiy@
zkRFOY?AtL$RE+fMZY0HJr+00oUnAgj%vCqz4+wN(<0i9aSKr0-a?AZO^{A*hv#NTe
ze@l=kx@K@6zw5**(KfncHM`sbnV9&Ad(*InU1{s&9m=hutK`zpt{NnJE4^?yHdVJG
zz)|-m#D-Y2wy*zQKHg&qpnQ!l<-x53D3~#dPt)1X{?!HQLVh3Sa%B55!!sGz9w0Mj
zpN5o{j_W3qm23UEGmHu1X)@G63A$9z*FRO+@G$o6;9DP51cg<;mhHTFqr7VLxo#nH
ztagi_0?wc*D1n-iJs_-r^%wI1ddV1}l(9QvjZ|L(<MHNmFhXSW9~e)4{-cPrMm1`(
zGAz(!Jme2zP^B@IZ-i@vcf`(ScpQMG0Pt1q+AR|ZbyWt(|MQE}fa>-LzW}H*58;<T
zMb_)vKg6c)51miRjF}xnxR!lSgCcVP)>XyWd$Bg&uiN;|B?K)Gt+lr96~#=Mu(A-z
zXCWExV*%twth`6u0MjP``uXd#fD(o3XtUspy(~4A&Z6&$H!rgCYmp{b{MHMM;e0@b
zm>R8WZQWF>QMCw3kgJ2^<sq3u2#x;<TY{creW7;&h+JQS2cP|kWB^9gb7L&%Q!lr8
zf0F@Xwce5^xhF4*k33wfOzxxe%~*YUO7&2zzAlG6P?(n?Wq*kkz+`&f%ALudSu3p?
zKtnl4SH4A0c?XZx3#gAv`@+|YogWPc=j$X?eCck>Oc;0{EqY6z=xjh!v-?<(JBvL$
zYR&m|Nz`K^HXx(iLd0GIXPR^cAN~0B!nIn~4IGQud3~A76pK#8L76;{%<yRg`U=u^
zu+qNjP~t6qwWjb7qXmTQ|1p(Y4o4~ie@Xyu&?hRw7sdy5yLQSwFVg1oTkpMC&v9s*
z-Hp5w2&atDDu~z&K%cMQ9>@{T^N}e!{Cs#8N#S<qEr_~qMDh<e6j{epE5_u03hK$m
z<Tuwc!B#-bH$Vqw+}-ElPZVi&JWMIf>y>^VW6fIF)`>UYnQI0l$$}+lotf&@M@ARM
zUx!D~z0WJ@QUb4kPc0{|9?co=2i8BDZ26y_aL_|6#|Q@Mvn!p?0Z<iyMu_|8DO{mb
zelPz+VJ8Qug=0FQl*MB?y6Gnh>Fo6>fns%}ag?F;$rzL#JH22y$XhEd<Q4L?F@#@~
z-%fp7Q7?C}oa_BT^X$<@iqy8QZ^b>BY!#4}w{-i#pMIsCWe^uVSy`{s<tXzS{kj^{
zI4Mw46F*FZCNR$$Hwpy8Zr@UoOfPR*!@fO(_>LdH=NMtW?r*tZ`|s$R14CHYH5n-!
zJ2rMB`X|2d(Wf27+d5~mD*sM-ct!kFAyG+h$#2;&3~N0Q5C6p8(Pz`$o>(<1V_=;U
zy|!K$>!dnGg<zed(%3Uz4}>Idu?}mz&AmQ*cTiDQ7(9W|m2dQJia~#PZAM;2-HoGQ
zNhI-y*`B5-%g0+t|Ch}KxYK+UmMYIj>t0$u)y}_h>+bML)ac%HRZ)z5W&hN99PPvR
z;icUrKbBwD0PP0EkJzHLuas3gUmAZ&pI{NWd-z>_Wc(+Q&Y08d$+J?Ez#!<b_A!@5
z<tKICj-hNz=1iwl-FJF^ugopkJ9R!8lm87E`fJou9LqRR{N2aIyJrQlA^wCWYp`>y
zKxoY1t@OZ~em7sJ6M>{Q3001!v{63U>A$f9{szZ<!v6e-qXBNSX<->r<xiMdn_+CP
ziv|jHOitF;>B`!}B}$BccfbE;g!jd;`VirHcBb3uU_ukUry_Fj8&(J9g`VE_(l{@w
z=0x3t#s53$?A4<`u-dJ~9w%GcY{u0#R&*s`b#unCU}>T)HuY&mSy=%Z9f<5!{GET9
zkYB^3fAu<if?D={HM5MJ%jFaKJ?-}nfC8WHB9Q>e{{}qo3jisK+1Cf$<_maU>-sXj
z705gv>i<rV3X=H=;(-L0a3#`ThKJ{FF7lRwn)3aBS?|AtT|YH6zy^QA3aK+4TA~I8
zkG_wJ{BIcM6E^xMDFuLoqvY17flowxS`(PRygyi;BRc*csFN_j?b!sNE9>x=g(tPN
zlL9^g(1+K-Hh;6JzlvS35)id5$23qqS0@LkCli!JzW;5gKtU9c`-rMNdZ!2Q7{dW7
z^8dObX8;e0JxQ+71F@FBQV}1*tGbezF+{~ix)Za&J32P{1mN2B4Se4GTZ#VVkB`Q&
z3p$35U5@CA>d`(OBX9;5Rtdph9{Gelik&|4)xNU2w}D*koe?;{f`?zm_Ptis=}F&(
zg6>K%FGW@kP`2XNcWKJ1uOW)pY3|V(YTm1RtWmi5=1Dw1-u2&t9b3{ZI%xS6ZcUx^
zgNs{mrkO%y?coXO>W<)@FC-x9S0cZ$1aqmpsK5%5W1%ozN5hiF@5Yyg11I<H_Fh+N
ze4@E<T5^1fKdD&rNml3xd3`0S{j@)E4!^ib@+XI8LEz_YDSySMxinr<B}l>b#U^G?
z<RwEMGH=nOE#j9ErgC__6z-4EVo?<^dUmtG=mLj8N&+h|9#QDQB@h?JG22U7_a5i`
z2a{iqK;_rcA~*Bmg1MTPYu`>nH6Pf|wGZ`LTSHo;EaMg<@%5it3%?7ms1AwL8aJ$c
z*`d_V!*JuT^CQdnxdjC-gJME&TYpp`U25F2D5nJ@1E3{5c&mRM`J-zfEKuXV$0%3P
z40|qPvx06_k9q!L1RZ;W6^ks8mnDiWnhiHqNM!t(Kb4w)GjlXysK;(yG{?w5)!^5#
z{@t7FyR+lD?D-loKN!i|dn=}V4j}HFPdQQ20U=jyBejFbFFwC2wENG1<TDs`pBci!
z!FFOQv;S)b@A<1AjSSQUB%C(!K$6pO)MVsIxWzV&7E;{obE-3oDT_%$V%b!ZFTuv7
zUrM+xwYv^MPg>pNsbmh-W9i*;4oTlw&4TEkvFhci1un%Vht~Qm1<$3M;~zGs_+boZ
zqUU9T*5e=`Za=Jb<-E!cWD$UL%IFB+!c)WQw}AyS)4Kx!>%SdO^+&_FB%rF<*Jv$*
zUm09pVT4E^9Tg?&loU<3`mgli0ZnmVc5C{(SeLmT8`R#jBd}smz}0vk^9tCrRu`Z?
zn0}dPyDnF+e=*|J9aR((1V`v!l%aR(>VpjB&rpJAnci71mx~sLSwowVbq!Tpg+a(*
zXWM>=@b;mjAK2dj2r(o*J7PPZCg<Yq*sZIl2EvQAPnp?W|EiRM*Vsy*NI<RJCLzB}
zh-2+mk}WdDcFhLooy|vP2+XVGQV`SK2$053d&`J)nrs{OTsss-r$8Hus*%h7$(D!n
zN-5EFV~aE{Kt5*kGND_!_&LBm0t@bq9}<DIo+0e^gk{p1^4Xt3wk85t{<@`aSO`+#
zV3sPO?O$^r>hvGc*XnywN@6_uQF;M16bGV#_z{tKGwlFbkH*EZ7QHud6}Gfgp{amZ
z(mX^*BTAcJpiAwn1M%FJ)V|E<$<XUiQWK%Sf*@qd`e3`G0=Y<;j33v2F;=gWe-itk
z{DyFXLRXg#l6^D*_{8sLjiX{)*1{;`Z$aNBdgUTG4ho~UrSqhJlHelcst#`I0tooF
z`aVBN@a?GkNYY@b{srM__V*m&EOC=kEr)hLQ*x&~7~K#!!)+IPZGrAlqENa1?x=Cm
zpa`zqD_`Fa1rxGUomZ)pfCs&&-?hWQK;7Us)77&k-w{gJ`m5ucEB2F*16jZWu2}LR
z0IuP?Mw0PMv_*UxH!2?~A*!gF)#|1w!k91nGmy@9J`k*ZX%+f4r9W<#u4|_{a7W16
zR~|afG;Mv8El(C&U02=9%sknfEN2i#YG_E?llGW!4Y)OT;f~4Pm-1lEZKO@v9GV>Q
z#;!r??9jz{!1wM|^fq75gp)|jl3&)KV;rygskT5@H@AaPPDPM8S1;oWpBrjQ*3be-
zzHUs@vF}fAp5Pg`PNo|@RYqU&s_pe}!xvu=e%PxduqC(MrznI7gQk6k#l^|JvsJz4
z#eh+Lu{}M*XFy^BQeS{<q?COeJTBnUELwbO{fwsC`)~)~jny>%o-w=s$F*k=tWVu6
zzu6S7Td&iR@tLSC_b#v@)pWc2Datd-fui~yNsLfnKJP$%=$v5)=uLHnQ3rpz>yzRx
zLuY-60}xAStYoL1byMHO4vbcvzw<5bLMhhhc8ru30`axcNp{b{_=wWCQ<|7w`E~6)
zP!ez$30lmHQOJiatRZVKPSN-cqHgCR6s6X0l*J%3d%-4z=s~e1al!N9mwW=-)t(%W
zp_d5tG80`>jOBOWvYf)@#LC7&zT2#cRdI_n1|EJQ$NgnS>0Eff$ZeSTE$qM9e4utz
z`X}>Twq|E>Fbd#Zm)%B9oSZ#0r?UPdMhP`792$PzDG~ah0y0KYi#mvXcj7x>JDnoj
zYalEbBf>{e;Fwx4Gs*v^aV03@EIQ-Nj>Wu0PfvOqxX(g#gXYz%y_Q!Gk<DdB$?kjK
z<}9O<V&w3*Auq3-k6p+YDPzY%XQkJM;vy8iy;CeP04-IInWEZaVBha#G@Z({=dIf=
zqj0W6@Dva|w5v*dDm;+F=Epj#@-w4Bq5oMG4Tw!lvBwH<m3id%*XNT9Tx|I)-8W6G
z&bBEK(RAP3Gk(b!0aud6{5-tL!b$)!=?!=BBmVmFxmfFF9TKHvuq~V#MEr+|>aUiV
z&P$dU1QkMI&vb#~%TOuq+7Q=WeQkBY&HMV)RyJ#gAAtdL%eBAN=BqKSY&nPHC_UJj
zTXqC)Fq45e&d*yFAKTEmq1{!tf9QX>*O%;Z4xY+7BTvZd8wFG~_c)Sa>~y{Oj3qk%
zinDCVOy`kk-0;GaMf9|A)<DLagJ%$n1V?o}{lgKbSDJ45CtS_IMZ2(gqxfoeCgb8|
zYLLN`-w5$r4?e+j0My(cS1qnd-t9h>YmoQFt||cP78}yAt$d&sM`8SnN*8Cuef<sT
za@FG(M>p3H?Shz4x7}RjH*uD{cDSpQ6kRz!O=IPuWd-?>7gU3@3s4Og>q7?i+r>av
z2N_pG3vPeR#=Le|aA@y4Z$xC<f$9<|6P`XP?RiCj-ShQHZblGZP{>lEwY|N3E-h|0
zCx<cF4LX6Revc|5qMnu}W^ZPvZHhZJCoOvH4PK}xJ2>tNsiQJI5Ta*I>U;C$XGJC7
zwhYZCbch$|&vtf-XO1{S;On)aZY6GO8Pv)d=S89JgN>`+6Hoo9xQaa{?B-|Yw(Apt
z76f=_cRbN&l_eLgh-^b-8u*m4L^gmXN)~i9x`(LEfB%pOBrhT8I282w4K5j$@2CI!
zRpIvufvkOVX0rQ$RavOyW1D5Nq3*{*-9&g!h@S`fFVB);G2iyGAqEXa*1M)$b~C#k
z_K#RldXpqFc&vNC<WdexrhAh~Oa?u}Q0AvSMbPPwcV{R(w>t1vWh@phmxgMp4NqpK
zdz(WukV_lfQ6T)<DJxGjc~>W<!wrPEJQ$m@Uydiv>i2Z%;GWOwVavSk7kYtmt)tE$
zB@zUkVHS50wWx;f$=V1&@`@p8JE?W34&JM=ROB^6B5TvT*Cbo+b7249uT)WN`>vtd
zvSsk3Bf0{oK!xis|NQyr;5+P2b+r3X2F;I@qiLKoap(2RcEm(6Dlq{w)Nss%5Q{ZX
zE}ygeQG&v6Ry*^<85gxi=8@JGCi0D=*BuF#-U0=nQXK5{u8x*-J9{mU^>IQe)fha#
z&kbF*sLhMAiB+3-_4!BMK}0v7Hp7feh5J9KqWgyiYQL!Ed~m78|82l6_sQ_DsiTw1
zQAJh;uR&u*7F|OXQ9fDizf22w>e|n3s5Ksa)j8m_V+TUHS?^|lrQ&YBf*Qd2ii&Ai
zKh7&GIyN8Y?Ck(8TLpA>Vfu4M%uZdGk_g^1Nk(2vMu=*{0l#b#)H9wD!X`rmx|0KA
z_9X65DqaKgGC?N08|4`2ivM;=k8`Wb*=BYp2l^|e1td53^CGRrU<pBT#nDU<S58if
z-{$nD;}J&@+%IEJppBK+5VK~9h`rj;*WVtXn5tjA+9Y*qK3T02#ewA%UDiP+Q$k%^
z=3O1pl){iv%zp8e%K#0HB@;yX<HvrOwr%WCOz`%KSbC+^Zi3*s?sCfUuv>k()A$}z
z(8iY$GJUV|V{eG-cAm58+uy`I$O~W`C<LPBivrngC`{M8Qpc^7)zB4zIV#-02}KOi
z&&36uCUw>M&7X~?@MQF#&sV#bn0rp14j`JZR*vcyGnzgQ_2XY2Cc1w9Da=~6clOH0
zL{D4K>;b#5F}P4VW0N!RXiZ)PU1Gd#Z>l-=h%BDww}olB-%Xn;3~>yAuV?l<toiIM
zaCD_vpVTWJr7?ib1E4)`&7<!9D$-v5w7^=fH)A%+=R4AQWCj)?`)Y_}nsdFW3v_u^
zz2Vitx;?O-)Bcf&;qc{)nKOU*dd}j0wdkdkI=X*wum&u4cnv`wTGv)O$G`_$wCie;
zL{L=k9-LuBHmGB}sl*eF*b%_6$GI$N{;DSap)`hXW4Qw7sh(Zm(L^cfu^RgO`|0}s
zQzid+rljpYne>5o`U?KaLQ18%Yv@yC^wqi5Su}<9ZhGs8*%gKXI)#~x4fR+uN8m4>
zzQs_U3x0#I+u+ycpuBQK(-{`bn{5)^#_GkknjBCDdyd(Q2V>X6w0~7wz`K@d+`G2p
zb6#Hvt+w4UMdJc+2%>aRVo~9Mn^ARb=(i&i_Z+(uxNrOig0yz3VP8hHR_G%#&h2FT
zvNJgfWj!8+K%TcvS=lz9$6jeR2BRro9<(>|He(2f?1ZZ(oen6Y-FzU3v&F4bN&%RH
z(p&)4^Gf$>jw+VwW$6-!AN2fk&+)8j`@FRB>?~obab{2Go1nc|k^8VHtPgt7O`Wrd
zs>iSOJRdy~J=KC?;Po|EXm6!G8@DcXsDm|XAMCm_(DvJqIk0D?#Y0`|e=27}8t-%_
zNar6zpCA6FlKX;N1OlKmm<8tf_Pr@#iF44^u{B~(?08dWvW9jo7=$sZc4@Cmz$Bya
z{UbI8c<WWfq*TyJ_2!9gyZI?|g!ABHj4vv=8R9uW-7$1^b(mA%$1aj%y68Nfi0o&w
z6?l%|E%q#zlpvu%5MOj?ijC>pB8sMGi9ozOJHKltEm7QD__~PjmgNc2x7F{z%}C1%
z8<4{!&r`=UX3(JpKSRB#QG2rJ)n>5g`YDwNfiCLo3YpGO%ZzYT8GzW>j4aL^?Ka{7
zeL?UYzq71Sezzj(7@B7jl8zD_+DJosDM4m1=Wb#K<zS~_>mEdEGVnv+fsbu~X=A9v
zS-`*_76OVEHgwfGH(%(4D4r?S72b1tWI11vW;!&Onl*cHbe3M>W}aS=PuD5-e}>~v
zGm^@S{rbRI7^34nj{CbM^lf1bph29TZ<2y$=Izd0c>hp=qi2Lc@S=;IN?tdN^9r$(
z9SB_@y$wD)_q^10#KjnU_$g179v%L0Gk`l<%&aK)uJi-9JNK~(%wgJ?Q)VO)*RP%h
zK+k|X&B1^<AqJI~=R%*2DIK&wu%PUSNdUQ`)Daob<|0#%#Xh!=SDSQ7c{2x`w>t2h
zF>vOlt{$oEg!3OcMUGe7^ECN0Gk(7v%{RF6H?K7=gSWqs@-8fM$4~dY-p#G%;#)KN
zO<Br|Vkx<O$IKr=_o+lT-Or>!dkbhUrPJwR!}`v~9W#8V(3-1NQL8JowV?xmJFdo!
zi%A%R3{c<p?CPr~_QnP2owm@U(8jU(<K4!s%F}n?K_#&(f5h?T7=qyvWjF;pUcQVq
z+<?eNspOq+tQXmN9*5+snqM`_Iqsnc^9_fWPdm1oR}E04SHqRk{D<ibo;hxZ?TER`
zF;WEQ#lhkQvoy8s+5vuTY!|ab{n~bQ%_2JgYOWFImfj4NeesKUDFOH!QsOT7YCkuP
z!xuNiGWF>4_oA$b^c@kwHN;O1N@YTjNMY5~jyX(RSS3kE#Tz-f@&9HE?s}cx7e6RB
zv1u0e3=)Rmu_Q4*3&w*AZSoBik_)wFBgr3f-T56-$>U@FpY_1bgX3i3bk!E>bMGM-
zgjG`xx|v}eJybk9$1cq~5$D@ySNXyYTlV;c(^xm|-Nr+d``?pp#{~Oa!8EOH8CDr9
z1ZT9{!@)Z-E@&rMUmDr@<Sq&}gt^pResf8s1A5e(+x5<I64tbriYoRzRIF86f3H-d
zed{d$3NeEKjKUO5xdMH91%cL|QdG+{t;39|ot@4+La*?o#nfK*n3bY}L7v6+Iz)nk
zc0UaqKLlMXy0nk7wv8QTXtMX7d*qUF7h%&=YtR$fDbY~u*@rU^E~+hMxc%{bGgf3~
zemA}TxW?nME)llZbbdWVUnWnR`8L!Lb+BGjUTvKxXr^s&5?@@CVQaC-$8>$vZXmdN
z68$&A85Al`5db|<8<A@z0yW0P1^*_#922-AfQ3D7b9eO6JOAN!%=q>dI_0__)fwU=
z(RB97_DWo3u+^BagXnpkq+?pc^0%Qn&u5f-lZ4_d0tO*eJV$mX^Ylnj+Z`Wa=ao31
zWVvvz!6x~8=oV94bvIt7Q;*_PURZLwz9gf3(fu!~JDQmNhVu{)M5JMp#0E7fUH4~W
z()+zs{q1|6@u+aHXf^{2VsXK)#d5A;BxhM}YH<gEu$uKi+x5`1B{)sV=e4e;eO?vC
z2`AR@>{!v0B?MsPz?4r>kmW+Bx--bm9fWlASugb&#xJBU)E$wJAgLw&x^6<6w2bhW
zI)Yutz8+H0M~;xJ(;(tAM9wN8A~ikd=WV4iifQXWH!ehcsH-^J`vJV2h5hO)t@84}
zfs-`X)X+hpX)1q(#sAc95<fyAr&(6sj2cqBFhRROMzzb;K>n{>D-WHHk^X5i^Lu4?
z=8sbL?>imnJDxJ?s~S;eoZ>@~>26&h4~-Bv$>az9p6(-&Fk+|rS0Ef$YB5(BW}M=3
z^5A=F+&Q%h<m1YGUFDG|&)3~MWp-;nO?@#7{C&*;`3<sbwTm0=Woq=iZ!mi)b|x9N
z0mg_0#@^`0vZKvv&qJ3E0RhA=cxAd8tbhLK%wFawbuDz^@@=i!hB@zEX5hT$8BvFJ
z<oS2fVyEV-Nqk-xPX7M@S7g@!UANxG8RAOuc(zHxqD#r}YrpsJGNq3TfB}35^L4jq
z%B<J6Ni&aCq9{e&`*%V1-{(K}Q6WM+&QYeaJhrDtwmvWqw#Jru&5*EqJ;y?e`G!`S
zO<&vf{nBNGso-odZKoKkbkZuKdI`7Y&d@?Qt$u~1*+~_?XJ>L4*n&+2?B_Nrxwp*L
z(1!;&J;L5ijq9m9@Y+ESJ;y^9`;Zbhxa8iv9Z%Be+pZzaW|u?&46mG|oz;%hY&wR2
z?-+5(T-BKQ(jIQ12`s+`;SEw$VhS?(%x0&vm%R7djcMXo!<Ep5zVn_Lo2$_i{>6pm
z?<zKwCQhIkPKVPFR)eLAN7B1gVP&R%!<CCQ6CS6t)}{x0k4;Jb1LtuNVSNLlXd@j~
zN4QQ0cZp~2rvRV|TWf9NH+#;3o$%ld4&GTmOOAE`56er+j5h*>n-A&ONjiUfGS2az
zrZl@Tlf~mdzVLSq`X|;qmJlZf>9RL;TPD-&3LqqN6iJ-!><Zh!EDh(}G?{yZO74g3
zq(pZ(<*bZ<8L)lUnv{6L6yk(#Kviwer!D8e`qhfud%ES0@u(XIsAfxA2X^YrX<qfu
z{5H7%JRJX&Ca-9Jz-t%)yoMN{^-pBCO94{lf*WXNTM$kRk`6^Ubn#hmcZ8;!DgB4^
z{q09;IpA)DK#6U}=h~tBE5K(k;L+GJOzm=Y41LRQi+TLNeTwo=r}joYyIMZr=QyE7
zE|`=kv=8ideq3{MAHY*X%fCF(dGa4v=2y@7&v$PK0JH{P?})9rSAG=v)#zygV9{>z
z-Ff#LAovgL{m%dbm<x*$Qq@fbT1PaE57foyd9@LN)YOP#?*DHy(PF~gpaRuvw1+@n
z3gTL2r#6&jENI9^SJ?j!a{pdpT5^95zw0Gwy@H*&6y2Dsz}&%@P^Q%I|F*vx!ho0y
zdXk5@3h$BxcJ;->!Ie`YebC=@#Q#S)D?&a!gE$(vm(B^80}v5s8ux!2^ry!quW^$&
z1E61)yI@PS4D<b8@b#Wquu}hTYsP^cL*g~pv{O}NXg@d9+)@ZlHk_@YFfzyFR(Fi$
zz`U#%*G$>m>?JGK{>NSPZRvga=D_@F1yP0~sP;e=<@LVU4xq7hdbeocWNV}7h!J`J
zjmzAh^LK3Yt8e=6cW?BtNvtn`1l-%fU0CIq<!qUUo^7Eo^efzt$aQt7G-J?`Kw)ZD
zp^$#|>wlo?zeJ|~Ds&|)xRmqjrXe)8iPhG2>xQ<PPCr@<svWjNEV!&L)>y69Z-8rq
z%gxIWk!4A7a6fEzwiHK(P-=%9CD!hGJ8tgCp3XeAKr2V9PNs7exDe}RqB@o&z1nA4
zY9+acruAlI5k4+T_sO#E5)#mQvdOM+M}8}y&H3PfLz#nF4eN=1o2NJeNxs??+h`Ec
zhHR1?_e>demG9Y93~l&yfv4CVy2oz4b#1sH+aR6DB7AVFb?b4kT%q&Lv4cl%dzgKX
z$I;UE%%BvD{Y4wig7HwN9duaLc$4{h#yuM+LfAdMtYULI_Jpw+72&zKxW-~x><XQi
z6He-Db_rbSTNF7=!In8MQH@m_DG?GD8qaoR3#zjYoQQ(Z&Lh_(P^c;n119@~k9XQD
z58^Prhw?j-M)LH@S2AOvM_m1G%~jlmm~FFbQJ2j<WY%LogS{U!<~PtmMBMA?u+pxc
zY`#>k`Rwm|@>~-c$JH4=@$aaEPMvW%s{AvZMV`oP)W*N8E-Cq8S#1q1N~V)h9r`+S
zLcuS1xi*DuGAN;fwqRm}#M}w<e+KCuXUf~K<6=w^Vx7=W;e8C2t3B<H8vbBfs}sN?
z>5EkAd&LE&h1^UYDQMUi|E!qabSyVwmX{x>s+m`9szHwxwX|d5@riYUX?8Ci=|}$J
zLp39N+JgTLj~K=hCH7vm+3GqK-uFjwMWVZg7xirAz6_XCD`)zm`ynNa;zA<<q6>ul
zdt`0)-4BA^xe#^{kKHSZ23@gw8G*cR`aF#Yp>Wxp4vSMXz;%*E<EA`JQvx&Ch~Ydm
zK7NV?rMjZUdptta2ZQOYj2QSyh<A;@Z7X=%JbN>f*pXqRFMM4uNw7ZJ$C9n^$L=|)
zc8x84ymOFN*cag=u+}9G(>7U#1bW$zqqlm~Znz}B!;sJlf{bkyTUFAKTbS$Z`mkf}
z*Xj$IvD4Z&J#3re>6giEE-8=XQm~YE@iyDhOc)hqDhNW1(uz}MUb-~-Z>-KSL)=xL
zr8sP=nQ|&}V~xR_6u7POnJj}I+_7q!U@1%%-3_4;UNbW&b92hC2%dJvr4(`)Wu+Ed
znY5hQnjEZPKKe7FE3AseP=r*Ww~0tj>EtB9V!e^z%To~*?H@)?0+pGJg^*|3Np*|%
z^e8m*#G#yVSLokn3B8>6aUq>o)I9%;^_?QlJ^o_`vK4uYq1b3qJtp;sa((Zuy;fa6
zjs55EsFsOopButdE}ZGYhSk3iG>MqOJRY|5BkGJhby0VJkkHe9`>N`ioO#jmtp+Pl
z@i-v?q1JAW7P;;yhI2Tf@;RBnBy31P%!qJxM3)ec_LP7rGPCxYPeJ=B`C~kk{w8kJ
z>?Y&+2VLiyhucq`%XM7XRH8yZpUK{iq@dMMy@-8MnDOQX8FBxX9n+|gf(_5onxDBE
z`k1w7VtVRyz<x!dOF}@|M4EP89W^`dNEG>x<!=0NXZvn{T(P&1O9VsUO$<z~<c*ba
z=l$--90MMjY>lQ<Vpu^BvVnV{rp&UP_glnf4!DtuEKx5L*XiAkb@KzDaVnvUOi?n9
zuFJSiqxOn+H*|W;C`i_SE~o$-@Ac+K(!_&(Dql=zB|lO5<~=HZo%8uHvwT@P;yw}G
zw-xb9FV*)NWav~De_nbHG^0_v%O_=A6My}W`9xgz{hA)$>*C*Ak{*4;qXJ{JJSl#=
z-AjIp@oJf43zWSWTw^Rq$YdDTfK4k)_1<BWY+<$C^N1EfV5OqbJ*B|#M-1n+tN9tG
z@mm}K=6H(=!gFzxXTtK1Vc4x;efc7{boUz9d+OOho5QG<MB`qKixARXuMH_2PC3y*
z{D^Fs3ooV46~}u~&KoKmqw|>WyunyHqr7HgK69Tdc*!Lg>XyM_#yRxCtu@DT4z_xA
z%ybQh`5$#^`RY{<IW2?|XG4s*f-$mcb$%T!XeD;6COpSAfr<fga!sH|(Yo|oA-+cP
zxX)3Q1F=dHF&FfY5%o8z<}{YwGia3aYppn>;@~-k_q6m83QlMiO+{VcN}>2y!UZ*B
zftZ>4$6=Ch1h)e@_iG8XwH<4@SSFar?x8bE_Is(CuQTBGghi+2Zfw2$Ze1{iTmL~h
zDl;Gf?+o8dK%BHQgbUX?nM%}Im{nFB<o|}Nl}qUso+XSUtbM;Q(hydvX%cpEPs|%c
z^y;Canf-hQ`WubI*w@x9+dF1^7n?2*n&)4+KVPlSV2t=4UdRpDT;=TK=ew<Y{Z)TZ
z1WoSQ43pV^=6z4eWXS!##_SYl!IEh<TM%kCYis;sYhC_aikwraD=AKe<4I7JzQ?nf
z;j%2t{ElZHe1O2J&dF2X8u`~<06^WuN*;55W6JkOnIaO2%Wo)$YW5=8VzU8d9{0&i
zt*)!XJ!u~-3l+wbYWJCN?8U<A<Zn93<)1s?z8e9KSyZ_ws#<3UlfcJfbvwUC7a>az
z-qpB%frHC;knzb+tQ>oYKy|k{l5z2H(#{R?aPKm6P?&^S+)0!mbJln@m#k>{ctI>#
zk343J=sNi&>8e0=uhM-HluQw6kS|aAYYDd9XfC6yfG2wa*E`*AhiWYmP)<-ip(^$D
zp?&Y<rzHJ=kO)MTrMSxbWK4znWk4d&T}<I3Zzj)|5Y77!8kI+%;-sq!wMfiAPwCNp
z^&)P^o=DJ3_9LzMK4rIqCojuT*bukOL57yE(9NPB_e=Ia;rdc*x8fm9!Rzk&xK`i=
z>xr9^9uzoURSb%jV-yV5K9&zPar95{7Zijc32-ugOdq)zzZ47DHHI6Zs!x7{wv7>5
ztTaHe>(_D5d=0h8P#JY%HC=OYa$OKs(S>8K;os$izDv;z*LIB|a-q8#d%ws>KNi&`
z;fytU2lU6{jb?1mubY}@4LP?il|vogrM%FwwUjO&qqk1wNla-UZ{?XE8@<U?{3S96
zmcYBaz&Y$QXIrC?&bD^g>i&!cd586_N)%FPaFd&Ka7Ri~yiXnKtxVp@p2gU1v$)K&
zd50s}O2EB)FV(|`=B(q2!{MFIMVOVTFng`Y`p)P_OGq`9q3qsD+ol(CJX^%!o4ixQ
zyCAK5KszReP!yp9S!i{T^N>8DRir(|x>eHKv6p(>(jhg$y46BUKW^S`BHA=pKOp2!
zs#s3Wg!^@71#O6TE4HX5InhDO)vC=rs=Q1>ZfbG(lUF3gB(#S1c@MfsRk^&^u{-z2
z@HfV`#q1=$U*h{v1F_dkbQ8LDG~GJyFGJ40Mc-dfR1e-zSHa>ZBYW;5Yq^1ab`nQZ
ze0;!tBu-jymE3AK6&gCxwJT~5+uz<6<oO<AZ`Rmlw@+`?YK*%#rn39TH+V~A!hNS<
zR{K$U#o@r<F0b4hpFYWS7tbzh&XXwhd5DQ68}$&v#k`V@VL~psqW~Lsp~3L0!`;Yi
z2_jI&G4!D?gUpj#eI~imIf|^~z4rSR-Ghvq-Yq@tqRuT+(AQoz9M?316d|2iq`B(N
zDGlTK%9YP$?KV#kf_?!}o?{nbTBx3fjtQaTjmODr%`4V<J8ELg!PQY$(=>#tv*!hr
zOSqXnBUO?vpH3bLEIEC8;-5jiK~?u5VS~S;*y<p1nbmkn*e18WGu90cp5srl#HrwN
zOYGZXwt$>fo`>4J!nVI}aF<FYt$IV3uTXj~ZKtaKIb~<`G<dbxO4p;~;B}REbC!#Y
z<k|*%_ed5@SUUbY;zDfaIaJN|YX0tR_s4PBI66GfoaG(-+8L?Vvz8=F7H&vS8>jNh
zyGFCUJ<*3w=Z)CzPEF-r!MK(9E4)ULQe;6xq+vbvr*aDtkX)y%{Vi#z*~3xm6aF1L
zG9k--?BlW7ri4);`MQLSOe@ZLr05C%mJ@6}lP3x^c*2~w$k+7Xav(6mHLOWU;p}ON
z8Ld*r@V=#W)F+iVzV|OyJ`}ePQ)dt5z8pOtK7WjEf07(}f1C=W!@hh`marv(7rI(P
zk==kJW0-gx6(h!b`T8KTZoP9#?<!;IIPcic<3r<XA@^<MH!IG8Br3$i6y`fX^@#!d
z7ymI~S@#5l>a;3E?h-XaNMrn+iGOiuTNz;vR;pj_3OVP9-B}QQJsrJs-n|)VwQPZ5
zOUlS0h56339*9=e{FodX$H*0n?IS2hOEi$^^A~w`y~&OYnFUH*Qj}sllNK|u$Vr0J
zTKA>ci*}K>99G|J+if-PFOMF4QL{qLI*17kSmy`ODAgWC@;wA;UEn<~_z;2#K`j(t
z3zhQ>+c$&Y+~O{R%6y->86)_xR-!P<vkgU_ED{y|>^#{=G}%r!0IGUn6y(d)C7|bD
zHRE@!N;(&(i=Dnu?MG;d><i8&Clgz8H1@UQox9SHzd+;|?op|#DhcU@Xkqg=eMyDh
z7bn#bn;i^w@AfQK9w~_qJR1V1(y>1zyHgqaQR1<8a5i&|vA2GbUxGeE!P!e8q1JBe
z=*dAhqR{5+xseQ`MCh)1h)+8GEa4Xp^3buD1y8yk)J3YIeaN0k#6Z|rBg?e!x6(L#
zyLWx^=;(+VxSZ51PaGTv$sw|qFJ6&ehBLifoKnU=mz%R*{6vm#xxz_*N3={_O^fwr
zLF32j^66`>JeIF@_foZ~D%kD@QZ#tq7D^l)C2Fqg7~)nqkX*RYYe6AD=MN{k!#xpl
zxGc)Mk2Ls$9%DcTGGl0@BN>x1+b=|zCD=kH6fp!<QPL;*$0v#4hB7|KDJPH4w+Rcx
z{0f>oq;!OsHxvNnogYHuyBN~Gy1=`4(SJ@pQRHu`M^AU5uq5|3rX3O?B2g4#Ms3r6
zmkV(dx>S8@ioQCo^U+>yh52017jQw_%0+b2RaS-E*gU<WTzqIW!IbUN_I7HJd4*X#
z$;CBkV$esw_A1^EVXHf7F_M-%<$BY16J`a9y?y;LmC4O~n@p@8AL`UPLT9(RPeMHi
z6$P(k&Es-Lm;A}A20NDg?#a17OgVci&(^5m%w&1O{h>h?Jk4;x31Aqo)`|wu!k}!q
zi%M|kpP*d6D`ajCpUTF`07ZwtB`tQA$5z7~3h5oIP;g-F@=)m*;^DGbUJKrroA<w;
zbNhwVro}B#VW(!3$-8~En__;Ec+`T`5XU|h&5JCv%G&kb5+1WmDi^cx5)tI)>&Bvw
zq%#wnW9rNq`0X^OU!|FpvhIb<cA4R(k+lX|o+#)IFZq{5L%6_E^lv4TS`Be&Mk9T_
z*`zSbn!(xg+!==tattvWYDf8tBD@2o$x0Lz-IA<{?oM1x*IqP<gu!zZ-7472U|w%|
zuq1u3+HYv<kHoYw1V1#Hd33v+2d3)bS0`X#k+#J$l+5it9&K}hISA|Vv!WFppUciz
zIOHsLpM#M}6iMZEWkg9*w7I}L?mn?-V|M8~XU)p;QRxAhJq`-dy!dV~z^G+5jhrD#
zj$2FJ!@${=ZhB}O^kxSWXCTRmBC9Z!Hsi8P3f{*jStae$#ZoD$mOG<OnR&a6(1dMG
zU#3_Yq^0cHQ@HLf4{D+ug7Ah%PBuf9+^Kf@!LP1+e}mFGU~N5#hS+O(yEWUCM0LI8
zWFH})fG{mjxy}~vsd}2F#Bddw!7M#u4<|`5Z#P&NQi2}zxs1=r%=nkl&m>2bQ`{V?
z!YJyxb`C72d81&iD(KAm&rxi`c6<X9PHcH|Pg58(noiD#Tt<P~iHsEVYfwrsMIN7H
zN<mSOlS+`B8=cP-x;kEiSa$z7h@Sfur@`Pe$9>xTRu%@r#w43!-?Btsk(PH^k?l!T
zW+xcJd$RIc>j=Y+ri^z7^<Hb4__}n#%<rt@Au|ded+#)^Odm`bPtTXv5RpGO5};z1
zrdM~iD<37v5>>Q)iOw8tBG?>dzi28gU(R8|w~y>J=6PIE2a{U*g6l1)HfQ~Y&h;F>
zkUZyI7ry0&e228n#+eVTZV%;@8PmENWw`L$>ri13LsxXcyOJkCvhG3#Qk$W;)kMpa
zxn{H;rZG3KT^NKMQ+6p2&R+nXeuU8Cs!c;rJBCsPF6LWmi5a{Jv~FBhEClIZX<Q@f
zOSn-)6>ofE)@46erX#;ELnXjFRutw5^bvA`BW<8tc6+vi*lO09elI`S`Gio5!TBs^
z1@?jN@SSOw;hNM!wlK??8iT@sA&MeAVaUQhPV>SE_gD-(M^44S-sJq*lEQL9VU3f|
zocUzYLFA3Nok#|yiwu7s@6DdRrwoP#Tx+=n-|iHpTN_<68%DY$+QYcHZDnbJZm0+A
zIC3sg`wZw=UULlIAC6>6VO%SEF5K6=^vKtC_IEp8kK)<aLv*-&boG*=Wx!jbL43y-
zG3CSA)*@<mC#XB{?}n8a=F`r}Wgn~5V85ukceW`xekSW?TLFhfC=s;#ctC;+`od*e
z3yw#Qn<_lEw%Xc0i>&JJ*e(<ACe3T%{vaDeuD8q13((cKY$QrrFF?%ha;1d%!H{CF
zKDb+@`>V3aZ_s+TfYuk@N#{~#E4-ckNITGoDIfRdJi2W#P8$OL&XT0Je*#urUA##m
zOUoK4BDE!arAOWXxw4f?bxLURz#j9V$Y=x!MMJVB6SaP@UD%N2tWaWVYavZSFLE|i
z;KtGl@QM++v39Li3`7pBeA(^sNshxR+a>sDdW8hl*SIZ^9r(fyf}YpDH?cSrIN96q
zill4UQ7I+NWrTon@s+t+;?U}<eEC?4y;azGaTJwvm1Sx-^t3)>cnV0S@Hy{)4Sm!j
zH;irGqu8~Ru#63gVP3dpn&&XeDK#Ej74^C8W_R+Z-KUy!Ce3^g*0Yl1<}$bj%rctO
za-Bz=Dm=z@3y^}S&vvWR@G&c6Q=A@^C277A@u$Wi{&N?fb$2jdcNBX@LNF@}XMsi^
z=ITBU$?r6+|9DVs*Vj4yVZ{kC5w6Q?z9g++E&q=4p)`qN-ODK3h%<#F^M<>hT;6n?
zcX6`Lqa2`D{4_I*joYmAuYg=R8&Up&NvGvZLl5)4Z!p%ncIl8h<F57_Z=*5q7T-k3
zmdTl1xZdE7R3IFG5OG3y859}7n#fmQ$t7sU<rhlH6oV6IF<(YkaJYT{TCt?9Kfu$m
z&CZOs_=l5K#<f>lxScWSr$E79ce0&OxhK7<+C(cInxsN^Go_#*Kd@Fv&!Vm0RYEE!
zV_&-afh_v{;u5sKLI|clrPMqVi_8(Tzd`pv!oYY;;d+dW-zss1+*7TuW8YB)q$~`d
z!e5+|-iaTx&T6OgV&vj5*kmHd=a_8I_w>-<oW!ic{deI`O3W`N6^iRn>4f|tVw6)W
zvv#tCk4Jp(QDiC^ERRLh9o^$%2@`ml@3bX-K2URSdrD5&a7w`Rk$bw5pVXXbpDDT0
zy9eTtGfqFmLmx%-X_;L=;_}X9l#Nuq;J7HGw}XGb<+Q(_M$fLc9<x0)B_n9HP?N`5
z7*vXbAAKmaT$(QST1vx%N|+6URw>2Wj5f9<JCPH88)ZEPZsrhl3QK3VN~73#+$=TQ
zr(wpgU@Af1j>4yWO02L@W~ESLGyau>(Lg2X9_&qRQWwxWMfFAJUYhZ@{AH@u@3yRM
zI^|NGB|g&R#1o^_+GPwW^wp5f={lD{T8zS4SP*B|i?E13+JFoW_HQIaoK2i;n>E6U
z7YU<Zbk!2~NUQES)kVhyGzG5^nOQ^}REmk6x|$S-xCR}Rc}OnuDmJgKhhy23t-ly%
zlzp~kyg{ffJ`PDpWK=9u<nIucsh$vMLb}qXu3i+$7|oFE+RQ2FZfcp?!%t<Pr9{$=
zRIy{ylL|e_T+D?HHx>3q#?oIU4ls{K-R!Ed=^ARg2)5FN3fB3?k8_JX-%A*BAC*=U
z-eLtBq>y}8vr)?+rnuH?@s@%$ImOXAvH|C@LZZjigEub^s|t0uV-7pEavM>neTjIr
zq6RwE!>*CEu+v8~*3Td0JTA&F^|*kjY8-}qCVkFB1d<qzStm#&0&PzQeEgtR{P7v?
zGrY_mi`9Cw;T-+#tu;va<2Z(7?}WGS$M`pW5}kId>v$95!^w92c<xO<Y3cH`Qdg9h
zYrlI4D}i|HHbas_@zWd!lYlaR><95R-2McXK_XeN>&mrcU)_%)kACKe>tF5~2;Gt*
zTa1bPSYr&|AHSusNq|F_wfCKL>(2ATXF<)O{`!=ePBVAoUO&f!Di)eXXkjMn{0`wf
zavk5p&cddK9jD49z2=_@d~Ic)W_?BdAR%c5@gy|VDVwLE9)}rdmudBYfVAr&<IoS{
z_pPyKk}MIoF?-AO22d=2-8pN9?pL9Z&!aBdU)rDJIwEsN+=qtxZf{xL^xn34a6z+$
zi)p=`)%P&%0mTLXM%E^SHR^P45B|W9{3ID|d2;90%E}reK2u{7TpAN*D4U*%eA_Tn
zW8=*ZCBn}F2)tm$Eqw}EF@-xyo!1JTN<Gji+f(GR<WQzLBLfN|Q28izwe{RAAxY2~
z)%)t5<%Xz56Q}A5OKYKl!W~TnRK_`hQc4~9)zM^Tli@jtK-=j{ajE{8+eV)I+3<?x
zQ~uc<Q?s+CC}9r++4qzkO0G$-<lVG7Lhb9D*oH{gG&px977{OBz<Y18k$w!jvggcM
z4S7=c5_@-v+$ebNj%cr*4|6N$Fg*l)NZmb&`H7&TCiR5u964DR645-hS@y`u3NuLe
zQ98?e!AAKz60Zm`lN@%*^v$aB&+)%GsFih*TJ40~Gj<czYi3{L$;>e`7|%p9u__BN
zZ3=$;$aR!1pk$Vq<Iy5{*U0z(0cSy&zMmKM(k_+DQXM%sNIvniQ5M;-z@ll4ouuL?
zH^&<^?f7FXC+(CsWy~CaI!hNvA2OfoK!=H7iv4PzM-ug2pG~;NeQV{v8V{hS;{|uS
zF+t#J2tR)iywy$V@YFGeclxPJR|VOk(RAQ}v_oN#qKm^RaqOrAX&Pd`eojd`N$i+L
zaTt?~1z!RL;<AFDUJAv}5@c(rD3<yjo1qdnqUQs0SI*Y?<^`ABr?V)EpA8~&Uz410
zGf4Omla>uOl;Vep?2c{QI)oP89#IaIgk22i2s`_k0|yU^AgTS8ul?ekgnQhmCqch$
znoM|E6_pwW(Y^cjd8Q_IiwCaQe$wwSSC}>iN$Bj@zFh|9O1OvJxMq*XYgEB+IGSOz
z2oB@(Mm0@Mil5R1hgIbW7S$|O-7F~+H68IYr(kTVV<jb5OjF-k33ymKSxhXj3N=YA
z3v~*dyqhp&ImR!h1?}OC&=i3UySzAxBQ^aLKTB1HX}Tu)4W1lX1M6OyiW#mccwnYG
zPMzZO-{D;*U@|!f3W>mU!(3w!B>O&u@L0%?Dm^OqU{828#FfUQ>G6t^PFkR~C@Vge
z+c5F;7)TZrCrvMk2m19G<}mpXOa!0x>(;uBGG6EEWtuM(K=er*Dd}l;JOp&brUN1U
zwJ!s3B28X4*ipcB5=eOP<CApwcX{&7E+pIu-d5kA=0(53Zo4pKse(rN6L!&{*ojTt
z$L3b99+#aj?7SrH+qQ1;N6y6I_sh<kVHDqS*yi^cw@#E6o8>HFL9WN^po*WS{svwU
z`l+Ie$i`9pY?i)}w6|#}gX1T{xz{DSP5nA`g7RIPS%KcP<A>w^wr^E_IgS~npX|C?
za-r~y+qq-A)QH|X5vMM^qntV@eRI{XrNYcOOqd(6h^Rm9xZT7wmioQZO?6XDAFxT3
zn(!H^8AtKcRJVu+#UynlHBId=_1o5Wipd~!NoRathIHTFX!`Xz2(5Aa#JuBzC|%l8
zjt!G0O#@@#$tl=z&pd~tCI>M)x9@0uoZYi`uOGwv#h<Tta%8IXy<27M%+1%D6hFuw
zkU^w|wuaJu(uOu4-(y^wYv<At`bU&gib~E`WmlR5Rz9k?eshe<=Vac<(Siu`*UZQB
z?{GBk7E$))`oU2ZKmGche(r-Hpo$-q1-`W8*;V{V1WVlqv2&YU+ie)FeK~4vVAn<(
zCq-k@`BRw|vl|^RCzY2;&}E#1kIbbV>@c`Z8pjY*rFhiuX?@1+6Mnkwqu$oeo_5FO
z<QU8)yLbC%PSKLH^PW3Du_EG_$TrsZ+07D%)%RQX9(3pK=#?+4`*J<|<$Y4=+AsXz
z%u$g^%3Km9QD`Qi9Zf3E65VefIPA{eDvKKBU*8=+DTC1@kVa6#-7)O^S>so_)Q1ly
zQV5Wa9zC|KtD;K4v~9m!6OpETl!RSDn1Rq-hyErilXmNzSM4Q((!tAW^HI-Q*OYGD
z5LWTRWUlF)y{~TigpJd|S?%;mV@U;(d-w123I_Ev%5%LO6SPUvmYSw^#tP|eqP$Df
zGbn>k9f((=rK^0?wDkK%NXDd#;$j7TPLUPUP{C?!S9<Sv`~zjijs%3wtn53{5%%Q-
zFIu)KZk}mj)-D;yGBq+-$HHOmrW*DRlC>gOVv>`ZLCY)+x01gpo9d`H`Wwax?RVt8
z`zaGDX|JqUq->~Pp_n>K9YW$aGPuV^l_2_5H@>@3`gR{&yuwIcRT-pl;_B;-1S85*
zBd%(NPTPj|>@_Sxm^6}RRq%xT2q+zxRpEkCVtpj+q-~s##TZ^7DdTtk*fWkz;Rwp(
zAm+0H-b{L#1cM9lht+eucXHZ^K1_}S(?fhAjra*^hJd`ucSWm1VIq&5poqG;*DKEh
zVC>$2p9yABj_W$+k09|oJ49)e_9YwvrpIlj^x=Q#`?MdLM5eaLaXC3<ML)$4X-V^v
zOJ7OJKaS!Dp*uT9sjIB`4t1b6_7V?(|4`So!a1*e8%Oa&?2NN022<0d_@M@~qdHZl
z4z5I~@1hxhcRWYWAq-__U$}!R%}@NDjz_t1RQ;ED7j@^!0x$Y{%I>H?>MnL&`gdA2
zcP@2zLCLpo%b8tTgO^!vKALdDcS-TnOUY~|iGV6?+K2-_bdr7hFll7Wnj2uZNjSnY
z2;tkB7Y(`~j+OnhBCyug{F*=7&L-iNS(K8RCdCh96AGhr{1{Y;0FU2QLP$r!d?I2J
z)1>r#lkQmYL%+f?u#8P8f#NwtPhhdwFT0Y$ZmFDWY{yaGy*rI#AD*A<)JN*2@+7gJ
zX?QOktvYC#O)3vf!fUYge6Dy}^z3S-_+gyQ%^@e1PUzB(P&Re>?MX&vv`0n&p}*f1
zw8{&rsqKsm-U<#F<G~nlChquI!Npfh`XEQWsp4mi%**>73s;P&C5JoYI+tA=g&&cB
zeHB0K4vANIhRINU+jLy9HI1pz<Cq9!4<8c6LNzy6ul8JG1Q2L!+_2sYHjtzu@Ix6_
z+R3%V$@3HO9WoKu%5p%2X>G2gE8U6U>!_?0&Q023^kCwt*J=<Dj(e;_ZdEz-n-o8|
z(<V1ihBCQ8(2C#<q0Jx@mAGx@cHlsNRD%$j9fD0l#CT^;Q#_v5cA?eKX&jOrKF7H{
zEirK!BJA--aKfZ%qyM5EDCZnA?tYbvcjLl$AAF>rlE5?-w2ef3g$1H)x{4o|U|%lX
zQj*;G*;L4tcgIR<ffRth6dQM6CU{Z5JbMjgjb0=uaymOtJRK%)kPSOje4X&5LOP=A
z;Nw=xo{k7Hr=_pdbaCVxc?jvQ@<|?1_Fy!RTScfsU#7;7=?4#aVWbyaNuV{NaLcu0
z`kFhV6}`mg6+cp@02;@KAfNBBy9)uUq|2wHN|UAm0g%8dG@$1nA-(Ehz-KDR%kd^k
zDhTN7E2G1)Fx9;ttRO~!L!Cu<azcd0q_4CSksX)1s{<wY207{x1s4Jx!cST-h#-%}
zI;m+(@k5`yQ+Cg$W*o&&Oe<Yzq;BBPLW{ha@yaeex(h$0aUl*Lj}ftb^*-Zv?buO$
zcHWv5KS7ZIf2GHtvao{t3W%?tA<VWdm9nX+-1^b#bq{v~mIS=`PgkCoG3rk%?0!2W
z{7Gojz3G`2n;F>=zD?S7ydqr`M53sj#2r8MnedHWA@ef95XOq7kDO!U+BgmgI*dK^
z$Ag+l)kXN3c&FT;>6xTIuS9169?+z7eGSIlugpi(#Px>pfO*J<Xv`T@@uT$8Cw&fc
zl-zOFO^P4raCaADZkQLzN@$rE9h3Xjgab^Y>G@`Hpz=j&KiK0rV?nwX1@Bamh1ZgD
zIVO|h9J|Hu^i%X#flG>As`rE3_u6O2Yvp^ex0JUUEL5{EcqA4vre?VnKWTd6Z&myt
z$oCAcG>EXM317qYM=@P{U)=~O@9*bfD=@2*8atN@ib?cmxx6TDj6jSLOR?i8mkzV*
zkH8_hlV|f>{pj}op~^n=sJC@8rz_pL6z{YuQ)<_xQSlQd`MnHAEZoQ?=a3v{vXac^
z%&N|nl>XJ0)^sIl+lrr(aH&c0gTNDpd%Gw#bVp0Sc951!ZlD=e@srXX{5u>$gdzgd
zw&Dkxu|O62sX%4WIAh^zjt+5^ly9-XsypRl#wsRdyeJn$KB>L;EI%<VwHx+XT0MS7
zD}E5PskGk%+9@f76~3(GiQGc0XFB&AEXt12t|gARdlO!CXOG{}gT2t{6uU29Nb61$
z%A5Dp;72EPJW=*w2Oa319&#mUN`DgHjMZMll@Jei?2vGAkOw<?b|mB7UO5H@g@>qV
zyh2X%itsvj0s0;h!24DF!o(`hD3?^x;~|x}nCPI;;a})KQT(u*N1o!Q*eea<_0V_n
zP->4!7DxP)Bd|n)k;cL5*aL@9{4|@Af~AdJpv^HiyKWT4Pb`E}-%Sbx%0i0E3oq1v
z={bnv$F~y-a!5lzyjBz+xueqPgUgDa5D|hxJ+iB`aDsel()#!7qbYud%y!LtEPTYF
zsoZJ-2mFZX=f~7VS$Tf2yYjq*5rRxzoRg>1fAPwINnQePORM-HL$>2*{9>goW3)Yo
zullbFUFx>&f1NnN(dC+=_UVs8kv1Cd?C(b-UJw@hO%)Px9~%7C+fTnSFb*HjiS$Su
z`keKm_{r^%@2B{Q(>>h#jNjd<%GgmAKjmFLb^46CDen=+xc>L-#`D)49qQp;IaYMs
zMTx`VH=6I2toTvBiKi-lJh$7_A&mx;g_-~YIwzR`=`P8*f1{pY(stkVyh>i;Jg<^0
zwU!O~c-!&Q8k|#4NO%0`QEw#p_QOZr*_)~j1j~_$M#WDoc#RWxL{=|09ot}vpBOEB
z9|i#gOB}BgD+}6I{FLf-lj27Qv-m!OHiR-ut@ugvKBy9Am@<r44zjyvy<Fs(nnB7`
zk1b*nhLAY7Qn|FcFek7oF<zb>%S^;H%{f}?Ht5q7KNV8od0Ng$sLCTK3B25dNmGYG
zjzu~gp%F}Ap7G)p2HtN(OB6F+aUu1^`{AP|kg5BPk96xq(0NqhG_Gmg?Gq9#D@{dc
z9?(&nDo0E<5btl3EywYFDztxg@PIY$<TXrUgMtR-h$_0FMw#8QPtNeiV`{Jx*gRhH
z1i@gcH{wu55N^6UR@TaTulQ8oZA_3#f<I_;?;B}}O~2h7>Ng$-)Q)GVKZ^ZK@_kgr
z4`~_%Z~SmCL~7dF@q^%O$F|-bMf4f{6hHCUcKE0qN!U5}I!&v7_wU!2R`J6sbrwy;
z<<JI?q50(*3wKLnb1djBVN?kp7iCrb$y%SH@f|!Y#~#Xx`{_4z!L_D$QagU2Z;Bs9
zd1;@@$F~_P$1zT)kiBzsW_^e&X1!|DbuV#$P<N@m;!_k#i?J%-&2RKMORe}}j>4U8
zY9^uh;dosZ+`$9UFEU=GrvEY6)iH1MA<)uAd=;(cyO=9~%l|ft;)fly905J-aocU>
z7T%XJnmH*)z0GvKm4h8-2Vz@kGCI?bfX<nG-I2;-^<?_XtnCyswwdCmZB(b19=mzw
zWG=d4i<~Hs$QqV2)s2cDE+k)@xpcQ5Q)G%CiZpQB!XJ3KV_*ux?`Q=c=s`gl9t3P*
zFd2vLXsOAELg*R<ueg&o(&z!;(Z`_lVbyMWN5LRUf<B~c&ar?fc4$)ks6USWL3n~0
zTqqN8A&d_Gib?4L(9()!XpCxLvbSvBR8y)=TJgg~21QV6^ca}bUcaeFDwt(g3@aDo
zBPZ2gNzfV#th%iuR7R<zdLDA9tE~Lt2r)#`DKm!Pm|ZF5m3*2Kp5=08C1GwQ`|vw{
zBn}@{NW0Gx>G#R#=G1>p>knn)k9uR5N~cUhS~KaV15K?veJTzdakl6$Ob`8|r{g<a
zj@Jm2GwqY&9VV*qNijm1gquH>3|u8{Pp*@W)BfV~G(jnq<@wg*-on_^Nuy79mELpC
z+C?qxCXli~k>ep1Z(X0hTx55koF%V{Tly}J5aFI1b`pi-`U0td7np~m-C`VrkI*)t
zfnScv0Ow9ob<=~JrjBx}oA9mWiJMQRq*|YK|B@DQP~}qWMjya&DOGI2wV&dL{y>r_
z{iyIpa;!{hO8rslulGCPqgf_DVXWwuNN<pgO*@L8kQPr~5)x9N=|CBRVy0d;)1df?
zg-C-O6%)pYa75xzd6gskxXY<c!|Hrixa;w7N_%h~Lqn9Gw1v%boRMCDen{H#2H|I7
z#V(=2Men%^7^%`-X;;U*p&VhigVIzVJV#UfQ1=KOp0opcC4KTWnyKhA-iO+&b=rAl
zV>-r<ruYeM2Yw<fR>-H_6hEVtV%wqGxQBHbcl#(#EI4AU9>ff)_|Y*irWx;m9qc58
z&;oUgkH<Rt9MZxVo0|4_{D^W)7R(TrriD2B$vDz<@tyB0)n|3m0wXHp2ltnzrr$Bx
z@GUz#$Gousm$`C#il5MTb-G{Q=Z112l~rGvrn&!jpzEoGj=W2cw6oYDHBIBUrX<uL
z4JmQ(DuV~(hmS^nf?o}xe#d))33UY?mh#}(XBJCg7+K&Cr^V2}!Heg3;8GevYl^Z^
zO6H2QXZ}ewiTjyp%LMJFQ<Zy?R+>xocr=S50*n!>QSo!@zJu=E?OWXO#Zx^;nk~ms
zX2k+nVh8R4IxP#}W<{669Y06pd!sy;+cYSC@<*X)f}1WPTxp7*v(7%JcG;`uQJ-|!
zFk~EklQq&=9+r!1!>X^qNUrCcVG;JSa|QMTAs5H#Kxde<pkXXnp$&t_Wu1Y+P$q-p
za-<{ibWe1S=g2WRGon!0SR8N&LHfcu&Q*AX5PIfuF5j%YVKo?R!n$>9y?`)WYTQYm
zj3P{)Ib->R9HEnE9)(d7#w5P8CXI{yp`cL1;B-W^sN{VVGm7Fzw*@0y;<zLXD~ii$
zsOkYYDq{#8@^eJOo|2V<K_S2dpw4QgiXZ9$W!Pj8CcrS)jWB&u52-`c$qswm^+Mg`
zc%#_l>BK&o-{T!W2tql-nKZ-(<{7tqov7cTv0sjk+$a}RkE8e@-_%VoV6rjGWtb7>
z)Gt<Uf(4g?>x75$Viu7Kg{r(L-SvI~NnOT-az5qnAENFPo}FxQocJ45@dL9+y`}DH
zze0EnX{XE(KycrYPIqu?gIS|TsMk&?Hyi0rya5hPgFb09r@V&_RWO!R2ct3RH@0F2
zH|qU`PxbD1{fLz>FLft~8fiBe@Y6C?r(j=^_4f2L5{7cbzmg2m(~oo1+_W0$OQnvJ
zW>vf>Y&@xhXYbtEllVKO9a4Wh;@uxGcWymnkJLGoJk(X5N)tkgYt2run-@VgXYFf!
zjq5#a!`D}77mSDWhd#{|vDAy%d1;%Hj?k7u|Hb&nodpQ7h%=@~IU%T_?H)URqB9Qn
zWYdRNB^H@W7Z?yC^aEV#Tp7lH`qgj|HhFelKp*c)Ro}UB4n<jh%$GC>wkVKLY*Ri}
zzQR#>jGyeN2s<Zy;b>I;sko`-D2gAB*g;{pM*2mLS>@t;O-txo<gNrMCzSP=+<hq1
zp!i{o<h}{aNx|%-++A2Q?5^QrY49I9uRV&N8L1oG1=TBHrNpGV_`c+@T&`Ka8w}oO
z*4IPQovTmeA?~+$hc}4rN_ouL$2)#d%pmkc$eJ2mc}`!Q=8JFc{~~9~Z;Yq-kv1+g
zDG%x%WkVZRGidu(|M8QwlE1i|mELjbE9ifubo`=>lt&p{@e}$Icr4HBc^)97pRt&}
zPWvYMDcUUcAXT`<>DG50Gxgh^6#{ABSGkP$25aJrZ~!vYrIddjAlQ<5p&c=94q`@8
z{Am4zH(w)tDTe!E={KS#97Ruh=yQY*Pv%8BpJXgqE4*&*?w3^YL!Y^+&%6k}o;%iJ
z!#E4Cv|i?|^gGqK8+_4j({jzxiD{w;kpI%B_KpMemGO?Bz5Dk2Z~D5}Fy7)upOiDC
zb+g}@v?vP?5BfRbc#5AePi6id`mTfrk^0R%6}~2(7v<7cw-X-Fk1e53Pv^h%yOcM5
z;I_=Xs7dh?$K&NAl6F;q#p8zOT_o+tWxS@Z2UFTMWjl-_B#w`F_*degj)x<T18*J1
z71@C$yj(SUrJ?QnOhthHfp&n@iu&Q-hZt*y#ZCL>T)4O$=s4PM-0SB~Jj96^r+s+>
zBmNAMMo6Dr2;<ZIfQuqhx+QE=f5fGBJY|BrDiuHEbJAuh6iA2Vj-O2}-SLAG?9SUo
zcFD$GS#Y;Kv%`1Nn3d=k6+bYGxfLl&h=KdCT&$imD^2lp?s?~X!FVn&*fCi<B9r#q
z3M37~>PZA62mo>uY4r!w0b`61t;E~~zX*EQdVzONAfWW(C(IdT0z*<a%}PA{5=K{f
zYFKuAvFod3bYdJ<d-ZNxkHZB<uAXN)LeBKht@w$<_P)dP9`F<DUbp@>qbPp*Q;mHD
zK~NAfmRLZk2T})MNKps`R-mGhDt@Q~?0{)92vAYT#-?;6QS2vu4;P%*>7;&|zEi~y
zIAjGwvg(Z$_Bo>~>a|{Om7NDU#m}9xTfs|G5+a0CffY}7lI6lD!Tcu^2JmMB>MszM
zHmkJi{%5>m6ro)@ag6=ZS5nqY{L;3JJxa_b!-ch*<?fDr8tNg*H+@l)(6b(5>Ay%L
zm*5Jl((eRD5g|YJ-e5(u=`fMqw&5iS;^>3OkF*I-hF8dSXn7jJBKQ&7v0I4C6A?^1
zyMj8B9{h#LBkn=@gCD-QU#Kp)^)XG=kKgqp_1$A982J2j@{iKw?42CbRqvA+H@Mfs
z<KGEGrnsL+ivT#8<YIUafD?_{VHvQat!)&gPa2O{^ccp&Lx+z@8<4R<NI<1i6cWX(
z?>A5&4MGdzq=)*i{YuQBmjqjcYcfpOP(CB<js*)%VPZG#3`&Ja2x~U6!y{3M$Km34
zgvW<uY^RUU8O`S?iXSkekLIYGru<P)z?Ob26<AG!;wOho&W`aPo~0?Tloz|(vU$xY
zezwc1<y3JH#|_4;glB9~(~rkifAJGK@M)v=L#g72Hor|S9!-t%_qeCyPSuWr@DZU~
zc6=Y`sJFl?<%F+kn)s%_V8M;=FXX++_;FK&!s(bBhoL>80L$?Q%HI-e$*@x^gzdkI
zkFtXg23y+mpgS>2ajzD8hHdThC!HtlC)<h#p*ZuyX*TpTO?{oFy=vEL6Q1c&Z&Hng
zUfs`NU((j4J+5yQfS}ku>iO5)4*4F%kMKIVqe0{Jzzd_uj*ZSQV!M9tj0sKm|It^n
z%P$X2s`%OGPmahftjzHXA#K$$Zwu0Yz>}ylsh;-1j(QuW?}9#j&=KZEg*lt1fe~(_
zDBF>RH1TcJtD~|gl(93dpC!c)?H7Y!d653ydzK&CD7;I`KaJA&p!h*(4<0^$e8Xc)
z(m48UxrE-DS9?w;0*wEdw60HQYI{@F_So4QL$A#!?X#CkuT3b4rrxPHOdB?g5!1Tk
zXDJjbZD6>PXy_e3>tyl!x|#lSgr&+`QSpNiAPBBW3eBX>&d`aKR_w1d#m{zG1&tsE
zMMrv%2m%cR9{e%E)fKvF0Wq0lQoqq31I7wuVG0DcbZ3EvO9lICAD%&=9CjF0xPv+I
zc;-bI58VkAhlRm8aPW`_@#4uySdoZef-+4f@`R^DN)<o&W76RBojKM<C$hQ{nn`(C
z0KVwcl?tReiU6TLhOoIok$CvmHS!t+m2n=su6<u7t@w#AP>vI$E0CBqg;hZ_!uTHb
ztE^Mpdg}+P!(~;L2ydC7<#3{2q>3NfF?ALbR$W!~Nz)b!qp<VBkTH(3NfkfDyWOu4
z))>RHBrGIv-DBQla=>IGrWsQy9j&ASE9yL9Q{fNw5TzSTJbpIHO7}XZVFJqJL}|tG
z@?M%KQ(qApVdz(&3mwWYO-Db)4=W%2ZoOLnOFwWzCda};NPP&C$EsXuD;y<}3Y&d`
zrr-xW!;YUQF;VmOwq&Pk*Le{f{p`wP5$u>~ca=Kg?sourr{FF{Q8RUSCTn5RrV1!L
z6OEcUeAw-i5LAy-@<dsJLQD<$RVZnP>7705k5!RgVR$n1sbE0=PhCRc6Jz3g>M9CP
z!moUky0}e_OkrG#4fPa3?P*y7vqAc<imy;#L4bTh+9YXJ_}7Ww+y&3HUD^$Xx}Pd^
zNi#wbgsbtyfqtKzgKE;Yh&^Gc6S1MbGNvMA(LRD_+Gi6c?Mtycbf~U~WlUv3h0?+g
zG5ei82vsC{{ICi)j)Qc?O05vC{yU1|hw(iuYEk3^O|bC1hNLI#HlFp1Tj(!Sg;l99
zucxm67j-Tt$fFGU-SN{zT}za_ZuLh9<?<?ZUoh6ZBH$uD>56LX)1%%fFV9~|9pjlA
z$}Xq)L1};)v_3&+zpV6bOYy@lQ5K%0iahEuE6?F0!Jsmg@R9b#0vy`lXo{bTRz*Ju
z>DVO>FH5@o?)hY8e5h|MJkyemDTHyCalhWi@Wa>~?hN$1?4++$o;2-8`YC=WXBNlR
z@sZV}GFRZ(VEPxZN!{W34oaV3il2I!jr&DSsJH4~ii7y+4OO;b*TSf6u}rmH7X<v&
z^?s(r&!&o>esoIiVS!-p?n3$+cGJ*SbzX$vKi8+EiXTcm%!|U@wzs{ry1Q5L<CV42
zXJhj7B6f$=`xeTY_)CT{Y8b^2<$}^9HSDZmoKFqBogVW>9|BJ`s^UlI-6;<g`Yw17
zcCLh-S#?Z>2a)+0JAg|jUR)6SVeCrhMeuR(gY`H#(L?T$PIhU<w89*o6ELb455K~k
zF3gL<j`P$VK5jwk!#IX~kY45t@f|<VL;|uDO6-`g9VagqcIgnH-eA(YzH}+Dn3fv>
z#)!5RKeLMq?iBMzF_-M#?VqKerX4MEu9B0p9!f#YenbTJ@0Z2tve0nbn&!<BBa$B}
zSV8gg^hG_+LkH7dVU3*OU0AWd(iA_NyUbU8XH5Pcl4|8njJWUAg$JQwBj|}49E2+A
z<PZN$n(GRhKoKEAuW&*XCkT)_4LtY^N4UWR#U||7V1-kiam5b|pBG?73={T2uu|G$
zE_s$oG>l-c2}jbf8ku8&5C|b$O9cspLwObiCcWvxy)^NdraOK}XOk)6OMN(qLawvo
z7k}Khlfw!6y^5bIR*nV=g1X9%x18dq@{zB%D&Wc0l`8CTZ?Td_6*KJWN(E^2EqnLw
zlizxMK-r-JUp-PXrWvP3>Bn}wQ=gewg?Q_e-XOTEghov8Q22&PST&IVll@X=ClKP)
zgkK27q1~%cLRdvP^Tgnfa!&K^9p$G<@uL&Nlu@Lu#S^dGM2m9WxBtLY3MD8T6hA>}
z!<OrCPIgia6e4-X^+U@??7TK2eDmrRK<e!KH^PbNyfYKSJ>KEtcf)l2cX)E|mz_QG
z(v3evBNM$bCyZT!6Md;4_r?U?8=nd1wzKjiQOsy^s51x?_|x$$Ku=kP-9z9>c_>-J
zYW-xPg0@?w$pQ!#6ol20!S7BHG^g!>c0gMr9*T=t*sYlI9cjHraIrmeSC8J;R$8S-
zog3wjAI(>3#c-)V(r~4ILj^f;U9XOH(a=G$gy28NH{fe{#ScpPyx^aHO!*k_LRg04
zBxhEF;s-$u_Y<XKE_Iz%)?UdanG1Q{D0!_bFg4#f#gDo%PB7-A0yWR?B|CmJ4A0VH
zkv8x6A)a`w&^vfaaf1otOg#4S&LR#Jk}8~wX%dGTrLW)hr|ycb3xXLpQ^gkL!C0G?
zo5o+K%U5*=QS1hv0dG$E1z$66W5W&v>ON_S+hj#A&rgLp1H0~n!VI3-E1-OjrlC;+
ziekFMBdF~(JWKuMG^DMKG_;*R`n~BlC`a=BcAWOo{3})bfFa6~4(k+?s4cXJ9T@Lv
z!(n8vT$rmeCsgHGzj;xw;-{h&&l@V+yyB;F&yRn!P1<O_?N_1WH~89N6hH7#xt*en
z74Q*h+%Z3(CtiSQ%1aS$k0|XG(>RU1!>_T+GhLt%6i%Ez(BT>7EhmehA3Y%bqVfjN
zR6`%l_@_Lf+OtT2dA#y1_}eUUk#qrx`Y-qGWf=y_GV-8?`|H9&T(t@F!hHw8rpjaG
zI@lXw$)75IV1{WAP~^{c{0xTUndVFDT-r8P3Jz_Lohv2Zq_d5?0HMPe(YE4e?W`Qd
z-B}3PPw_(<%q$%O2j%FvMOiE|C$fyTj*x33^;yNw<DT|CiwxcvR)BDSO>SjMX#%{`
z6hE6fd4-i_MKm2V%ngj6GAaz{%IKC>E-~yQykdnBewDRA(3U$g4Pg)1s^36AKcQ5t
zH68l&g0IXGG_i14{pyp6dpe1xL-I$6NUAlTl)sviaKJNxvW*_0iPDFm>=mL59tDP0
z<0wr$ra|$edDFC(p7ot_m9LxthZUw_w^J|TO^P2cw9J(GG%J3%J7SZZbD5fyS;t+4
zs#+&$Nuwx!sI$5ot%{eUa;kQ%_9%Z)wx&u{CO3O!5*7=;2vZbxWt8D1Wgv8NiXW|O
zxnuhB!l<g>f~K0L$&>n7jwj_RI$|&6XVQe>A>D5Hk#{b~?skmbo>V;G$9qOG6DD8;
z47+{^2o3q;r$f}!#i~nC{CDCFOgMpp(Ft7$6a1;4)16&h>UUb7vm#^(U`h%#pQ26S
z9~13xKSw$BC|B57dQ|>SOF_{#zyqDNe=5cu+E{WY&n`JeF1@$nupDE^80Sp~X4>Q?
znUu$dy2?sY42mDTGHFlSZ*DbgQ0~?jC8*$rv_mg7z7NePjC7n${p+``wA4OI?e*DD
z@sr~nXb%V@NGpcq@<w^(1UvZ!DTEwe@e@|ghFvFtLio*%Kc#Vp5Thu5LLTE-QIt<+
z#}BcvGCaqJQAasGEH>m7g;=V<!B4+Bk{he&Z_*t<aeO8ID&`xs?#1DTeUHa2Rm?Cx
z<@kHr8b>6>V_(Iiv2f|2EcRI_kA)3Ky_I;x96vxFbM@3?npr%>J&fD@F7kTAKib(n
zto1J7nHT&LXMU{3Mbaay)WcmH;hWzuXJD5UcYRbx<ZY2%L8-!&c~#S~Q5Dtk-!#8X
zl<<S03kqXj_>K`AgiZ)m{%wOsnfiDx!?rK#ALY0b%)BV<!_Jm+YOA(g_me7qn#!=`
zmg8lKg*m6ru{G8(FG5MwYhuL@afQW(VUbR05lC*wPenQ3{*`|SBXQ-s>5Vo&jN*s!
z9v&yRv#EDctthi)eu5&VN}J2_2c#Lr&k5O~G|zpBNuqOeG2eHw!xJX2W3+!{;aa)~
zqO`cQG>++=aY~zKUMU`YG|gSI{i(<w4kX8%!$0kig&sU(!+6i)Chh0yuA#(JyEr`U
z-p>>0EQK}-HdXxao#Qv-@XPgHYMYZvkH*In1E#zKoAUT-hdV~UWBg+}YOi*LiQn~`
zr~2hRXEYp7OzL0VgKPN3j`Wx6PVAR=rNPq&lt-9P1~;rV-kM(=#Kdm#w|5-l$EPQv
zKL{yJU&*i1RCi6ge#cMz4uYe`sn3{(zQ=CyySnSM$qhg1u1^g&&J$YX6D9E+N2kje
zu@s6Q`t!8k8z&zYb~FSIF3717GTzK1TPz&)wj5=ZRs1~V+0R{Mf{0N8z_zX1dRAwU
zV2*=`SC%kgKDeLNDl5mR;)j(pIsx8&@kO=OBXMW2R<>+$3`=Uha@}=q&)&VY4g^0i
z1t?omA%Y&2L6@WSOH424&9n0We`-R<I@1x9*H?7u8Ne}ZjU!i<Fs)6BpIBg^@$(E!
zHQV2-JYdg?qTCLlSdgLV!9S)n@mMOw4-7R&4#3zk;v#Ig^pcCK(O1%ldv2>Z0*kMI
z<EHGz*Es<KEATGA^isb<o>kzfLGY@(yY`Es;+r?$-1S0-&MNn`K25XYhdky4&S4i=
zeEzm5F*?k8JmJg<(*UF-Xr(^y+%AG>5p>5UltumiB8r<br%mVXHA5&%TUCP~?7>%D
z(Nn*Plm0Sv<m#KFYZhhVH18PzRve@74M&88d!EueerOkS2p%(n>70;~m~xY$YJv#G
z=`rBMq|!GtX=gku6DCzbkt4Qra=7cW+XHqi;K1YX%8_>ns?w(0hzTY}cl&4>;N-os
z$ww6YRAGbvMLDW$B^btF1Q6tju?M{UjvX);;F-v$ikDozsK*=yg+hk$L<<ZE{lu3-
zln=hynG<%Y)+z0kwu6v>cxgA<&bEm#R1Lxgl(E!T%2FqV^l$7UVkIv-wTOoadwK*&
zuI`g|b`zfJOv*JbeR?HErhWAj>QmK7$0&-Qk}`q5p1b*=je)+#lNHiPrDYsV1!jaD
zcE?XHA0E4euY)$Gw6I~)iXW|uAr<*w>Ur9i#y)d-1?kf2=we<oN4?eK=@s--@5fdA
z5XZPH4`V(~6+hHC_}+A^q+P<#s-OaYde>geyT&wQ94h*;ZFtWqK%45FZ?BofE@7u!
z{o?i<ZDLj4Q`?rNs+9g5-xk8;6|%81hq|Z=vpip$Ux1=%YV(=qH%D9h9t0HhwQ8!j
zs!vk4QJ#aOWfZ#05r#IsQWQV%-+J5|_g2nFWvtlc>uET8k$F#LdP|XUlQE#zg;4p0
zNkMra<Cp}BlSA9twPR<u{Z#SFv8kd=X1uKWq{_9_M@+NghXt6qc~QE!Bj#7C#bWt>
zsA){|eSKz!ZxcrO1ry=OQToWr468&drN4aCTbM`o-iaG8bdA%a<5guB6#9$>YT`Z)
zyL=s1-{LOqKJdTQN+_+5h+E^jT!mV?H`0$8HeVrBd%ucfoF3j$c(N<1=HB@r&1$bH
z(QkcXUwT(Jo@xHEi;dFYiAnvddvFcE*h%uG{KIz|Cics_(%|U>%A=|)SXS}Y{066P
zntb@kZ^wD$$0+nPeWgg1rn)y_gWo}L)HwC2wBu*&7Qe@ScyDr3cMa2&cJ(vJlQ=b>
z&}3{g#m^wftqwoz_?eZt{Tdd3&D3R)nnUJ2lTy+hIepS?oDI5xlX8@wEdTM*W)(lr
zeBO%|brQ(A>QvU800Ii>fXm8|J$K%rd0G)qJ?br9t@3lP`H7xmX8L93zrOaj-0gSV
zne|o@PI+N~;H!q!KTJgHrdnNDw+~;ri7S-}g@G5QC4rh0KcRa}0?21B4t$kX=9DJi
z+|83yoItBI@mMOw4{3-EtF2%CV?WXzsAISE9mUTt|JLi|;!+w}eb`XJUL59u_rITe
z`HNpT5baNW_TSw9{OdP&gRexz4~fqy15;s7Jjsab%CL$bN?@l<9#a!5lhlogC!;g>
zaqSXiQEJjUP8xsx7yfh4D&90)>6>$?Z@KL@_toofaR2hDPrF;c{p}vFxfRZMiZO`;
zl?1-*R*F~g#bMfbA7%SD<B7*y5}-^GZH5UI7C}itq2M7r!V9IU7V@s2JiA6(hkPqt
zm~2lM_l^T@;WM}qpAR6G8ZJCT1ZuZZcKiTL*!2^#9PZyCubq6$gZ_DwC{+T8!M^lM
z8>J6Bc<4xguu}$JIVYH+T;d)o?(d0>7aX&SRLYyU{V{v;!s>U94rRdzieD~s)}!E{
zB`Nh}bb{pbBT6G*mcH&(<0T&1#X*})&7OUG-Gcbf9la3?QsQv^6hG=WuS6&-q6t@8
z)RFQ~8VAp?d!bU2sat7hM5!N8p|^$)PwrCdT|f|o`W*Y(Yf8uIjH3AIrA{5m@A%32
zP3?lBJD+aqP&TjrNS}O|cr1$J>Q_gyyq?tiHWWW`I!1Yq`M6jylRFj%`teHbctw4z
zaEWQgX^QE`_LN1~rB|N_^((*J?zgbOsa}4$9f0&j`%q?f`j$97r8b<U&Efz6KmbWZ
zK~z>!{FLfwj0t_7Dt^LoNrSav7KzL)x8yJeXUO+dh(BoXQTT{KX-~@EN>cnVs;=pC
z+*@h>H3)u{GNRXaIOeVQE}K#q+Rj#4IG8(<cG?v`+U9#vsFWOG&Lc-$dxosCkHg@G
z>_vIkcs{`Wg+bwx4-1c6TFii=DR<v$sVqoSY0=&hS-^HwcXN0Au}hY^7MoOw6}y$j
zopjEnVM=wh!bE+BXMXWU8J~LS{Fg5i;Zv!n@R72rDe$J^9nLk!Z{{hDZ#!y#JATYY
zVORvTEm!lJG6m{Uy%m-HE|tql9HW;Cn?d8D&sYk@&!Fisf0ISvkSKmm%VHH2KO1K2
zr;<!I=4R?y#m}=}_);%<fw{O-CQGaiDou8D)s9AkK<TzyZ`IEg@eGQeYPAZJ#3x?)
z7=JGEx#yhYF1hGJKd|RA^X=R3a5vw2n`Z+4?*F{2XS*?B7@Ri?F5wdPI+Z%ZQmv?r
zeFY}4zOyQJkD0>Qz9byVDSi;V?GZ*Z=mTRa;eFW2eaSz^Ta;GGm6*Cg6+ht1PAv@i
zx$gQKe7S@iRKF#>5Ip!J>xAiMaE2*;^5d@b_rF{sbkEw^t4OG*I<K6$;p^Xst^B+0
znu{G!`aI~0`^Q$bcXotaeeJbUpANYjzxhq~@lSrL>kebR(suli_)_`Cf=?CV(gxX0
z7+7gH^igjmA!}~O4DA!iG=gb%38jo#x*86F;PapLJ^t8*3(h;&-Q}FK$zG4)egp*D
zDABn?2mw%PnBctoeIIcD{P9oJ>S}(%l0O;)nx*!Z?p8|U&wX>3$3+*M@15x1zjnjd
zyA#ly|JaVaUbg#^?&|RyzVXeTW0YWiIO?s(NE!I@_LC_;1ch(;)%$>?a5_~&-Ia>6
z$taH4SrY8McX{_^N>C|mwZoHF`Vb5L`N6Sc)L*C%a+g_cb))*4#;5cUGQr1qWsBr_
zR<@ZlD`iEunhhH`u4rS=v7hRmKH&#A-ih*KRw7N*KS7KmQq%L==}-4?zt2J_Y|@}Y
z0m^}KO;bZTDDo=zOxs~;;-P&;nqGDTG7c!c*nL#RkEUT1#ZRe?#=`YDjl7puxau87
zssCXVKVU$)rKU--!7g4D_^Fw+;)j|-98HG0UE1+e&MWs0#yd`{7)k=mCVU7zt%@J&
z?q)guRF4*l6*KhP+htdOY8b;9pE1?(X7z<grFtFHkLi}%Ub!p?pKEhLh#K?8c?Dld
zi345QSWaMB3RAjcG3I0FBUK1osz<5dG#P$KT3<tVeoA&@tXWf6qV)|JdFb}Rk=<&2
z1+UzAFiL(Ize;nNQRq!JO#gd+O8mLbi!^=I7agnB4nDWkKbPiPaeGlZJZU@dp!6%L
zSt*Ji<rAv(S4x`DXH`m_a{n1l`2g+E_r=G&g;K~V!(sGFil1T2BBZ}Mg`vuip7v3V
zoh1c#Nr?lkQk^S>p<a-Ox(O%GFlS6nyqGt3%N;QoyEnai0lrKQ0v{}VtMmi#s+0(S
zOR4y2N@wcUTPIT=_B}}ie7%tfPDv#}rpkCveQ9evBJFLCZ)tr=zlkqr2KAQ<JFZ=u
z`5!kex$yKErufMrvpVgLo?dVZGQXT(CdJPIIcjd}eD9*zM#WD$_?A}P$Ac)V?Y?!>
zO)IJtQT)V{vv`LF1Ny68@p5;MyItOsr#*6%%<uilo4m3BM?H?M$D>%37@WEyq<rK?
zECf&K56f`n6hCxGlw%l*(iqBu<&w~<v_q_(-+DY{jO5;=&V{`75cRJsAMtSa-0yjY
zw?ts^v%mB!*@;<+`Bh;WSm>l(V}RlO`PclUS3sy;`uw*qe#yQ4o$pDld+G%Ws$cxs
zpYFN8A`bY@e|Vq!zyJ85*sA-Eh;$_?euD4~;Zl9tSSePj2}axo$#|*nDIN7z38}iU
z`InM16sC-}p1uFdE1|06E93UY`zv1hB6sdx&Z$QvUw{9ee{k>r@Kyc@ITT6{xZi!<
zqrdB6Zi~Nvpz==K?|J_R-FrXq!LDT~zwB6oVN4Cjc%bY`O{smR{^IYaJ?V+=yPxnl
zZ$-buPDT`>gP6Nse6impr|suA-|%~)xU7!{v@aMmHGv_fPKc%apOKwci~eXmv6C|C
zDr9`d@zAk@K=-{9UcQMNA#k;mM_P<H>oGP#`4Qs6p-Te#;!o2a{B=?(ZY$-eH!?9R
z0{U>gT9rS^U4U5fLY{p-<+)xa()3BG;--`H+Cz^J<p?zlka4dM$Cae_`ye<)@q=O4
z(VUbk2Em8lh18k+kQOzx!QLG|xRQ?<ABR6jAAb@K-?a5y+Dm<drzT1dc7PB*HvNS9
z>Nh8Z#;qyNO)`dnb!zgvae^-|B+ndil`gnZn!_l5N-KHOzE^4UEGY;kt@sK2N!16(
zX&U6cl-JyyQA)40I=WvSnO4OQ`HF>%b;V32nAH69^gCsw4b#r`*qpF4hof?Qk5Z!<
zhpp4ET<CXpZ0l7xHkAdT^NS(sN+|k>9o*BIwCng7wkjpzt<=U!ik}$6m=^VdJ4Z0e
zPw<ZNy;oT2WEbTze>5z9)C?=Q7}meWW6$vspo~|F;)iwtui9iv^RKj?4qN9+?WOk~
zn~*yCdd9K*ouJcw)SL2-y*O9>YpVDujZHev#ESCz&YX&7J-)KyXV|g`bah@-Uly8<
z$)$O9ZsDV*x3q&el|8Bd+-^?Xz$Ir`sKTPV)YKQ;rmoH3+P8CL86%)7eh5YHKdp)%
z(r0G*5D4>JsV_`i3TE7+d<$gFdP)$W&osr)loZAiCFAfx*-#;;l1$5vACy5ifBthm
z_aFIOw4p$E{G<Yj;=ta$9Ct>}=H2dYyz#~*g0!7zQ2fLzL(=1vp7Qv|x~D(+Nj;H$
z{)=CBZ++*xyY@{xQ%cy?F_uXycLB-7I5uJTLs{riI;tXN*p+6Ril6;*p<+%bywr*x
zF#Cxgeudndb620uwiQ3DV)zd~@=AA)%P;fvQ6QiUiVe)?>;LeNW&3-b2~8(BC;)!w
zB`@*}08BBv<WloGp>y5!*QeJ08=*!{@lz>KLNL#ib}8g13F*$7uJ<usVx>hY2#wvU
zD8fW#jogEymv5&Ei+s4UptV&0(y|~8;N*9Lh=7Jw&%HwAMR)t2y*|0D&Pm%FN|J|t
z$Ai6l{G*B=>~sebJJMeE!sqq0aRdmjed8avTW+~URs=1&ozf4Wj2_qQllFOPK&1_F
zF;V)dSN)i~@ci?9)Wa%%p7pfvc29WBqr1s2DSqM|KdPXSY>NpBI7IrPWaL-uJ+CZ3
zi`{@7%Md8UC+=ZqkS7$vbm>4(j67<)eu8aBNkm*xSRbi)=-9249X|*PW@H7w&v7@~
zs8fQ%MX{tV2P3%2JAxq;Kx_R(x$CH&x=&WBbMKEDCNjH3*qYushH`@OO--04Qv|YF
zvV$o{yLrr$bb>kgpe=HL5XZv>#Zft38jn8ddlqEZQEnn03sclTPZhSrcJaHeW*bHE
zQwqZoU<lJON}qQGY$(&%jgVriM_k4J2{Vl1XD+(akg#g%!X>p0&yvu7sTDuN3Li@f
zs}heyZmA=~?ua7X%2fQ+l|tl04GWqWpHxwr){mkvQI0ss?E)y5MY!uq<4vD2j~=_l
zyh3~&UF@f+U3+DbD56q<OiA%mN=F>uAn&@+DD^u`-A>zoN^7aV89fK@p_e*!gT+_6
z11Am5xV2IgKeUT^;jf#_po*Vi%ed5D;&vOS3*i{2Eo|OcpI_l;+IIXb$by!6c2kQX
z-HXD{<oFF*h3?Kt6+hlhj=oLr#0~!Td)`Oqn)Ug1Mbjss`%_l@6w9LFBo5LC55hu-
z*iirBv4lbP)}{JWT5Lz4HF7mT?rzAE;)nSe<HxurmGKNiFYv3vx6)C2{*_(j@Yewe
z%3u3sym-Fb@5*d$e@$sndO1H$ZunRCVaAm>9u22?$=B}?0zYcVcg`>A@pAxaFYuGq
zJS3zQH?>p0d}}_SUyH}j6{;|rUQI{8CrKV=%!<>|L>qq?tq;V(AId@UhykO3MRm6x
zc}??>etXj1Up1$j?<}lK{rZSHKbm&@l*%Xe&%4soXX>7QYaFS4`aLMUG~77f#EH@R
z6~4mnu=OAg+w`t<c``<va-;VHxp>yjo_5DOitkHy@6LHKyFLh<l2$oaJnyPc;DR%|
zeNMs?$oy$;^Jcesu6f~2fy_j#i#vWeB4wW({S_NOLCUPA3lpRRUE$`9H-0@1HudkD
zg_%&sCSA=);vfC+hq>p;(N3xPmrs4huhdujCT60f_=#6VG63WR`h68YBEVrKOlnFK
z%O=GS0=ZnLSeo3|8PSjeXmFfJZpGH5cKqD`KKFL7{Qj48C+}@5e#qGmzwD)scrYLN
z#J{>nKJ20XJ^p#C9QXJ6FI^p5xl|$e1224md&mQ?kfY4@`8%QR-ccsSfqwmW-r#P!
z<>uI``(COiDBviMyva`zmRj*s+KGb#jEN8DdgszKtm3CsUvr8d@Pe7Ae=DshyHhT?
zhE>9(N4;G6$cMYfKI#$PE&hoWKf0rkI8glD^^%KYum1f5AO0uzj`zIRzvp%y;Q-?>
zj6rBux>GAh=exe+A?`)b{oZan8b<MRm$T1yuleyGttm2_6hE=TM_>sG07(=A*F_P8
zS4Xx<*>sfQekV}y+?#yty^$8ueU~zm_n_<vN(PP`gK<`6U+ktwj_GcnDk@=<JHq+i
zh^Na2{1iJqUX7LLdh8pi#lEX_k#UDf00u19%3S~)0hilBh9BA|yIl!EyJ7N%;-|tK
zU*RYaj)oIo9qQiNPjs+h*N@^$+6a%mE_={6oYM7SVW4GK80`UtRVtJqrb#G%LJsoB
zPHCLnU6eC9g#i9y#ZMZJvdk%da(oJ5hf(~{-{y{OO9lIJnu^L0Ife~hW?WNJEK$zf
zc`&Nc{x=PORCZk)iIkXicNQ}3s9VWblM(^XxpJnRO!FN-q&uuUhmX{L6x)zrj&cfu
z*xpOMwVejv)o;w(R<^RZpv13ccEH5p%0kO(M|h||y}OFzdnS4X6T=-3m*065`bF*N
z=%e7%N(T26Cr%cRCM<<B`{_Y|HoamLKO(eeS9p^t3GSPe<4yF`ZP@!>>FeoJRN0|)
zcut$*Cr%sbrvJ&CYW_vrXYNP+QB%Smr3pt%v+yO1f$^yv6-}iRo@K>PsqZQ&;Q~Dt
z#EIf3QMix~_(1qOOnDs8BkR1k+K(=Xt69oNN%6x@(HzELGmK##h}>!vL%G3|A|qpA
z4sk1wN-A>7j#1*VF#Qpr<{@v&VLs^Y8EBM8>i%gM)Q7frtt*3u>*r5<k)_B(zqFWp
zGXnG#w&SN6qt)hm;J{uv>TR>zIJ-28pVIMeF|I|$Pl?$pty)FM!pd9({wq!Kvt2HM
zSH($=;TTl$)5JWL6hAO_2#VGSbI0mtRyv&)Ass?3HGLI7a>NstWv8Yj)M!%tC=-{`
zQl^eU3xN?#AX}*K+`Gr2WMOr;4s3)i?YJ0J@dI;59AJV%?U<}GCk{37Y5{gp{Lo8Z
z>~~A4`=k{=zyGIia<3FcP`Zu$nrp9ffAh9?)Ub$^4YU!Cd1F$8@(7{ceeZQoO?LM=
zisFYsg$sr~Zc;X^wxu40vT@xNXWdBTC8uPG-A8$kl{EAVu^==E4(oy$&j<<EvR3Nx
zoJ>yHF@?}*L4-{(*O-`QoTijMZqVBz_#%K~7ZI!DIS1dbk{88Tx(eB^0uknMQPQwW
zCKl-ghc*!uKOtrW&0B@Gnt>EQk&yMNPkOxj&Idn8gLj|TeDyl_#y|d3|D9{2Asy93
zs~fO|7cjf~WtX}ie9;Tt=FQbg%Rv=Ctc-uz3!m>UyX)>+YD$V9kGZhheiDVI8T^%l
z6$&u1Yu3(tAt9>*smCZbIr7ohiLT%7OD9z4A=Vrj7l?*<{JZ2c81aA~;>M0o>Y8_#
z4{uOL$cy}-5DAJOb?bPU!biB@C*)I=Ymx~`DGIbsUL+grpkeaP|BkajBp+m2^1E@<
z2DioUMhvv#tllZ@n)0CDK(n;tM~`{~`*^~RVvV>^_z=CO0i5SVDYcH3p#Dh2dfc?(
z!zeSONqf^BKXKm^r=NaEcQC~E{e=3GJijoai4Q8pxmmV*=Ps-sMe(CMe&V$AK8)h0
zxZ|fu86xGqOOCBf%~J39N#pPLO`o2-!=~w2yP}dy`XI*i*l4}LZc>V$IF2~(Vcyq^
zKy{Pwn3!J0j0h5QE6=5!qG&{6lT$b<Z<wZ|mwxb3Izq3hEC^Yu+wp>{QFi>$`Oz2j
z?y8f8G#$ltoVTI^O88WEU1Q+I;{GlUPdU;@q(@`Y57PazLzmqd^`0-LFx-2uqHDPD
zsc)@^q_?d!rg6mI8V~Ve7_vt-;LGsVmN_ZPJ3ofDMQ}oN=<D^U)lw8C<wGh9(DE<Q
zN$m$|FL4;!&a^0gN(&+5^y<5|`Lv#f_BY#|e^Fml@uOJMS8&=#+M?id7`sp+r{#cO
z-9@Bn2+y+Or&tyv@AzRsTrj2jQ`+%!Qs!hEWrsO)yVT_5F!;$Why%|%<V1#X`h_(9
zrf)xIkxqnp2s{FPx#xo=5bFN6(xG*uDJ|x<auA@tu-J~evXrLn@qe7O>sUSxjg=gh
zF~Sr-(_4bK+_~SKw~KMO9dr1U$S_5gIVW@fO)c*DDJgfbr;o|)E|R@RvZyxnJ+b2F
z&ENQ{d;Z0Db2A+ER#IZ5tKAhIzcNJxqO49?X^Nk7?sBd_;)Y@SPVOX-#u=M|6+a>j
zI3<kQ_dV;GUhu3m=<McY#T7zrR$jkX00=?%zV7p2hr`)t?Q)NL_;<R?cJJ=4Zi1;`
z70s8wa;^K=zkX6yqaN)M&U2T|<szh2LqVt8#BoDd%3$-CZ~0qMoE-P7hF<c#=lVMl
zdIld!<O!LyzT@5Rb$8r(r;_eI&)T)iJ>okb>h5;wUHxVDC@`Rb01rXjH@@{P_u0>X
z!T+(kGFAL+-MYo!8-lQq&M!5e{lXXBm#+Sb2u(K2u9|zvj+h6yvv%$jI_rYKUXBcc
z{`>y%!~Q&91ZX>V?r;yj;{NV&k9wr9cQH-k_~O;q)Z#)(@HRP`ucU<29a74`$MkrQ
zcl`X~um7eObf-Hv5P-ez5B|7zc6-eD#ygV`@?ZJLM~tQTVZw0X1?Rg5-1k2IcrFAO
zC?<T_NL^wz^S9*qurFV8t^4w|*SZtOgJ44Qri!0U8#lNY%1)0a^Pcy8z}<Z7t=`K8
z=bh)i`-zXwdPf;X{3y1*@a3yLV;p-_!5GS~`$<`IeBLf8>&=@sdSMA|fI75CnEC5&
zxWRqoUq0my$;2eL10h!Yuqy5f!Qp}Tzn?=H2z~H@9_4b|?RN+cUvdBZiBEI{Iw4gs
z*hpLuLgywHsp54M#ZSDl?TL?lOm?@}H@|hW`;FJX!6%Bz5cT@}^UjT-^D9h4au?k<
zZ~B&d--kZjb$`w?p6+(YQJW=m%dNM$cfarbHD}r}ZTq1QdZ6@059?86veNr%sW%AZ
z)!cF$bUxrGq~LY$d)(bU`Egfz?9`7wul?HfwH5Wy`<u7E-5rtsNGgfPow`C9$$zha
z|E}+Nu)EjYFL%^2>ZJ<MD8t+D+~e!S$D~et<EC$-p9sc3J0CI`^Tad}#DigiMtvX<
zR(Mx8+`Gm*bd@BkJVEh8I;Y*t%+C9|Oj1cNSyYS>B5<4=ZIqoI#{|<3I=(L_Zypl)
z!JdU2?EXm&?e4@$6hDFXS+X-i%SrvBoVZik@Jwfi6@Ez*bbY#{y~Xv)yNDwb=bZ%w
z?4JDxBuup^g1E|xpS^O_n+RBahBQ^-IjZ7Esf?odDXq3`I&a<!vZaktez8f9^;EZ#
zaJ+t`Z17#WTcAk*UkWqril0)to9>mNKIiVD>8Bjj@lDHwC-r^Wcl?mflJE{UOl~*W
zu!^4&uJNv%5L^8>-+sa!Kk%-!bJ|JvjvsMnPLngfFUYj9aFRjTu^5Ld=eJtfUg9C4
z!|+2N*1NzY#DY70pqcg&XdBy;G6t3u{q3iIbc#ABXrfSR`>DcVSPYBpMdr)&rKw<&
zwu6EqcLero-SGqOqj#v4$|zQF2S39re$sil##K_%YdpCl=#eeNr^1xx)s1)j$O0dD
zsnnDdKkQKMm30J}$x<m`15aI7{Dj@}k_#DstnW@s)4oA~dV_&SN&7tSU_YxK0min~
zsQ9`4u*kDTeutu?O$EnFuMh`C5wmzij(TfZ@iQ(ivEoN*%{=P`FIhyeBVCo+MNhea
zJ#(xJlQzE-L?+8CP4RQlMHhRaRY{mKsNx3!KgTNl!q5Da+nG5cL_vJhU;T|A1YY=@
zXUh(s^=jSoyzSe!yTAI|w+eH0K#n=#g597n`TiF?&kJgM0u(HM_Lu&f2+j|Cf#0wF
zm;W!rI8^~(`@3%xf$F!q77S7kdB7F!d!O;NEQ8kd*m=42wr{(?c+203kp1@BQEw=@
zUiHeCr{25YS6};8_h*0k*K(KA6aCS8sh~?ee*BZ4@|bY+9_2EOx%rk`-LL-6>swL$
z{OWJN-rf72cX!-rk{XWq`poCP(6v6?y#vcREjxsLzx3MQb}yH^e(rP6dvyKk%OE>`
z*na<vr##6k0bxA*8S3ZzKXjG*oZL;NOo%FeSUL79|M}+{N&N2r_!D>Sbzk#-?s>P%
zr96MA(f?nIfd6Cv`l)WcRnJT?+41s}Cp@-Rr~HmCVUFlW;__#5{rB(v2lqueLND!+
zRPl4s1sC`u$}SQmSU<x=<n8ZzkNcNTf5zhgKL#EserUHr<XutWI1hHJsYzFl53=JY
zo)kUnX;1OWDW#{+uL!np5R5TQdSCwH7rJ{0&L#83tFLx{_@+Pi&fxWPul`AQ{<(K4
zd3RqGCCuyp=#Oje#PJ|eo(^M_&b{t_H}~VOD7WbxO$3L(@N2J?U3`aY&Ikh^_ozp>
zr#|s<*$IEmvwWb`{`BWQ@80~5cS_$TtM7fqh-*iuJ@61~cK`VMfqXkre`%Y+zOvKZ
zG1Lox^qbU?#l^bPAx+ovf6K&ElsnXEpH+VN5W`CQoja<P?cCFINEkNt=b`HO?dC%q
z&)K!Jr%l|kXOCZ05aQuTtlm4RU~<{<vmh%wG5&}>HdTpUj$__1KJLy7N+7VNE^t?1
z+73&C*d`@RSY_v<s{Jg7;-@57OOFDL6B>3$^d7%eQqByc_yL3b9X2OrM@!Y;#px*>
z(-kXz>a8da-T%F$6re5*cGTO9U*JN2$k?VwF_rR~_U&;BOFN=gv*M?8+*`m!j(HR1
z35$1PQxxnbcl@ZZgLdteGu_4?sFnC{jVssA8MkBJnEpGzd*_ra_TU)Y(C;Wsp5&RG
z*Ms&qrNwXTwnkuyp}ta=a|XeAZpTl~ui3SWfY!mbs4`~hF0Zy|woQk6y+)32?Pp5k
z)1ZnU`dr!uhOrtR4HNUe-lFq!SOg-BW2vF-Y?4J=8)bK5YF4k}r&;M9#=rQOH;@@#
z@l(1}r%5?OJUcV|SUPsbi`(LIi~Ggcp7+wx-zCKli}|QescBj~llmR^n=!}p7gA5T
z(+U3CD}Fp-Q~ZpJ?`jRt9Fpon-6@SjCHuB98HY#jmzJf<L%(!PqdWchQYe1<DPT%F
zMy3(7m7J#ivan_5=wY``7A&lr$se`59A%MJ{5<tJ&tHVWDw*8MI~@!Ov(gklmt3;j
zXEolL;Vn8pgDQSta1d;=tK(u3L_b3AfPqm-%~c=$xK|pSEi3I2cC)Itv<?4i+4=M4
zxBtCY{=iJ1JbB99?ea_A<#*lfh3Dtp<(!nf|Gw$wTijJ2{g~hB^92zQGg(JqeLp!?
z?7VZ&b=>6vGp7b+06S{FaLqN{i^6bw%p<<beectsR$HCOs{PN3VuKwDXPvdvJ^H&I
zR-5D=mC5g){N?|4cigc@$$JKmodFkKaGn=>r^j3zIC#i$Kh8^E@O}Pxw{F1Xg;le^
z_#40Fp7x|CbdSqoReHJ#i28U?lwI*eAEgw^i_%17m>oZ_`-A`C<*BcF<q!HDHgPhq
zyZ#2*W%X8Z4?+f|3zPd_JIB1K0_(?q=m%uS#PEurOYeG#d)W(~Un@h>@U<Jh?x0Ow
zMj1l+#dQ$o<3s=aQTNXu|AbGUDt<T;4Fw;Hh6l-xnVir9?5_Fh*ZhhmgfREH*FD@l
zFBku!ILP@gDSp_U^n;>s&=s#tIzD*SN8AlJe#0MUuv2y|J^SfT^*d1%SK7%N{^(CT
z!W*>>_jp`%;RWu;e(+_rBg#0U;hJl|DvGM>y>gZNnBo<L48U)D*Sp<^Kk~7#Yef7b
zFz3js)TEOT^2)J6FmRYb6hG8a1Z0HyeyQV^?!F|AM(zvyK=;nDwKBP(tzIm7LrE4N
zzor3W#SeBU5HCFceD^46j|j3-qY9hYp8I_Cs6p`2GEqO^L|ux_^<Tfy{mEbamAl(z
z91ZsZ??m0s9b<J^6bl>=q=ub`uYJSsxkIv(QFnWQ=kuTS3}06jQ~YzB5`E>3-~5*3
zf1`WELm%ocyL7iqLo0LRH*a!p{F66zR&z56i+2;kWs6_W<*FMNk&*JC7&}chatN>L
z7L-3S(TnoZ<9P}!5O90ViljvqR+@+H+qSs%(WIHh8r;D{KZRfIuRI_Uo^U8=P{P!Y
zv|Mzz%T6Brhd<e=CC9R{lPEO^Wsl2oI4FI*30<OZ2h^>uAu=)iQU`PGuq1HLAB!0J
zhPD+y2yF!_udr4jmrr=-M-|broxYb83Gu=J^{dZxccI#sc$HxkKcJlxl858G5Hj_e
z(%2a*etJU<ZbM&?+hs#t8C3D33zO=K1j(!H>PW|_Cc5;&92uILQax0E)4by+Uh$n@
z<y~HRPFtdWQ~zT#sN$yzlaC{mMJn9XQ2Bc%f-<;%uVL<<KKLAX$8H~|WaU{BSe6vc
zL5bLVEE7D+fzBO^<Rz_Z8YO)Q$BLGCe)Tu$rymw25<m7ynI^?L>FB%kwNLF@wDhtd
zgy)mW+H~*3DCqz@j-Q=o!+4tVb-DBos`w$#o#DDhCLXI|MM7z@K&ntuVzeE0c5r$~
zYDQK3aAYI(J*f_2!!Cchzebe5vB@228N27+=^p_`70B8LaLhFGD?OHtohf?MTTF+3
zh4EO!$8LG7hFAQQ`mQ|vkb=_DoW1j+yaK&xye#o6DQ(4~q$pw$({VYXw!dLP2#T7r
z&whey#t`Nyly9o|VSYW&5z|X(`t@6-_PT^gT=J+ClYV8nlnGPbBc9_fW&BGOR|#`1
zlmExwc>qdM6perG?j7L*LCKN?b5>Lo1qBm`f{Iz;6A@8SM8%95{fnZgfDutZ{S2rm
z2FwXjf}kKt_%IP9NIK%(ySsn=s=H=;W@lz+cV>2H_if+3otf_H>gwuT)zv*LI@*{H
zea0{pKW)tWT%^g4dRr)0L#!<)v9!fyDt?~4$Fo<nkk4~7Id{q!{5UFp=#&vGSm$hc
z;aOJk6E6IS$8lvZlY1}Hj>b>t|GPh)p@K~WrYK?%PNxS2j?4M{mrslkwlX;j9RmHg
z9ehxxkoPNMb@Ruw&k=B@yGe(j?5T3x6iNXc!c(km!VB#2-)f63qSwD_f89aLH;!Fg
z-<IRjIF=2EB?MMFPugim{kEU$<hZ!cfB7q2_o*Ok*!%g<?Xq})?<^J&KBEA*Qmk`0
z0&JrVH!zg?<QTm7edLIEHyQjo=#{TfA#@1X^c_Daez^YxA^x`a+&WBJm3g21>=&YC
zqRhg<GmhGM@B{9r$HhVOXO8=Fbb%bdcHk@am1C{$orupzk;M;!ii2PMN-3Ys6Y;Rd
z`OKG&k9Ro}2+TpR++P%LokbM&?|;92vRd!Bdf%}oLm~I-{q~Kp=t;ZK;>W@eZtTC$
zOC@jjH?R1arTAfS_>MQeK0@hf?%Uu0VRYUF7wV_&e!qttarUMIU!7R|edQY`MVDUw
zcZ;S^gx3eZdjDvXjgkt`U;g&A=oi2Jo!(0TAUuAaC>O#7Fnt^rF#q`16#^lb+KcjH
z9WlvFyY4$Uk_)chiXQ}Ba^KL*jMnD^_T8sTxk(y_fAVNmvQQ?v2=vF#P!=zG?z7Wm
z$x((Ki=Qy?{`a|$)~nEu_bkPa!BCEGef?|ocs29mS@w>fJ$Bnw`;;&)@7(e8kAGhw
z_chJx;uk;apPYE9N2wuoqThS5SU-k>G8E{vp}$MnNOMu)6i4|T@UoZ2iYjp7&N=U|
z(Xl6dHSQCVeV$+8DH0S7@{O0I3u4#b21*Ygd=w9&9w@2sUnOlhiC;cj<T&OHOC^^@
z6f=<TGK2z$JJZZj&#u4mM&TRdk@K}Y^HgP!cyB0HaADc7kiSs`De8O#?&MidtOnDL
z1^Ml=A;O%3^P#+SJf+XS>#pUxOBh-ah&siOC(1|85+H=_#L}>5krF0P$dZ+Y;R~Nx
z)(3<arAelpP`XnF#3@RXJO9!<pDDMSZn;_85%EKpCrHUU>Wz3=`9Ls5e?&V?J9c7+
z%0mk;%7el5da^J?Xp*J)L8ulUE0c#S6jHH4CA3`1yJHFE<f;sXZq^1T((RSkK@>k(
z<xE=4ugq$R??D~HE{b!pSeV;JkZi?I{?79(;UHzkok47DNV`1zz-0-UsdqPvb$hr`
zz;Nc>-IFCC%)7@TZ&!F8@ADKtwBzh!nP`_nzl;&Iov_ced=Qu8OK+2n4{aVtI`oe3
zn;)7I8E8WI)b``vvJ@x@rudJnHc3#r-DmL?xKIInt61}}26Ha^l#nc?E$}Es?<Pk-
zr|;Y%Y*58d-Z78h%-T5O6vc9^Y+@r3(D^RE%@e1yX2?_gXul#G>R2PC9rSlwmUVC@
z;q&kK$x9=zEWn$+*~1;N$<g*5*RK*SkN!w4C)r4wc6swQcjsZYMW^8)?jrhL*^E}D
z?{(`lcM1Ua!ZOJ=iys*dGmjQF8m7;=#{ueuIyz2qwjR@u3H)hdj2%}tQQ{Okie}zb
z_7T;q09a%Gs(7>hR$Fg9uB@6!y6vd9B8#6<WIR*x^Mu`=A!D2Cq_T;!yD++UHW2S$
z|N3Wkw3hl}@l$lxds)SgE?h+*W|N@}H`qXh+{xY2jtPI3J8i!Ho$u*H79Rxe`|bS_
zwH8XV>C~V9D!7*<SuVRiX%|(zq~m<^<nKqnJ^c@9Z~1mRU-#N$_vm2{evpQL?WFIj
zbwsH6V$u4V{kjzZmx)EmiC_O#j2A)u+Hw!i``-SR)UNO-`oT|r7M*p@pJP~FmWlPi
z4v&6h^xQq3k?^;Vv(NpD-XB5-$-OTJ?7MfulHg1^p6i5fbkb)K?OX990{$h7x_9!L
zWlr7%KD^_{6hGWila-er{^aM;ndh8q+0r9YUi6$jqwOELo$Ly|RgQT(LI#1=stCHD
z+-H(@hkx=Iv8X~J(d`~3_p$8rg6AbB5VX6ak3BBB<kHK!B0OL2oZ4vzr{uH1>#qN=
z920iDT7p4OxTtv6(|3!W{KTCMG=63&eh7T(E>Dd1*!5{KoV}p<`NLU%l6-b7KCpB-
z_|*sKMA_huIrdA@wbxxQ10)NU&JmNZdAV3a`D5^~BKyp7$0zRQA^xMcd$?ME87`hI
zk`6oKND<Jj(4B+$QJ?qQ`^9<>65r;qa{I!GU(*F1Ry-&dLdSxL-FaytPPoYJrT9Tn
zzP{`#q(1Mt+tVT}ltKr+CrX{>xI^l*^}Qy<#oFi1uX{}xMc=a&KP+yr<a_U14@uMy
zB8*b;L;5oJQAhg{RfE2-Eyw;HddTa=s$_j1>K<77eC?#~xDUyb&YpxfSU|nypw}e&
z8u)e07rz{xFGm4(T(Y0^h?sTjD-Vg*<;+Tk06Xf<fcR$mwn}z(x6x7d?wB{1G4f~m
zBwy?}ML`of6f?JRJW;?5dnFD+i1p=2$<QGv=YAg?yeAFnEy|iuu}ZiK>_tg2J1e^*
zJ!0c4PcHXe%Gtno+$mOIa_pSMAzkW2dgp6;anAxHPv}hAjAtuE2%B}s4|H-=rce%O
z?A+!Jlv){;Lws#F+d8r%Icr=;A!A3}#7KjetR1WH1%YwcPr4}Cf)pF40<`gKo^m5w
za1^Oxoja1!E=!=UM`8)jvKFpb6yNNv9n!e^`OdqGhd!f>#ZM@_Oj{x49npnilJO;U
z^!K4vXBbu_B%?mr&h9XdQ>JB=v4aZf4&xOalM}*~d8bA*Zqa6NN#U$|=}(@LrTdNv
z7bhGGJ6O8$TiXGKI>h~OY2j{Gy#qe9RN@=w+D*5D66^1E^}_IE+jhqzCWU_VnJ8n^
z7U<wF+V!7zWRg4n=vHmZ={W;+1FJ}sWVg%lzie)py5Vvltb=u>Og*eaIa?P!As+hE
z8M#}wjLT|c#7irF@^HD$O>X|oI&b-MQ~X4#Y|c8OD{DhSUO!E`v>W<}tTCwg$i}2q
zV3>!H?v8)<v6h|i?l)UEGB4rXEI(YGbN!6lc!R}{^jjH9r7R_joyyQXC&?abfXl-V
z)!7z~S&AH>vz6Set!|V#OWLB&PBtKB=&;^r!$jyD9n3wF{WLInN7)jVw;`j*a+-dA
zsjQ<&Gj9U`xXU=&JN1vv6pX3!{I@*hV{RYy1`mV<BX_hxSw{5j28~yg5Y=K<PU>Mp
z8@$QtyNcqETUZl395){1(aX3nwfM>7&gA*qQEy{m@w3yeyVF+{y1W6N!Tsv1t}2Y%
zwm=j=d4dsWEvxte7U2|kP9X5&9P@X+>GiU^F?l8}jsJ6E#lull;qFE*rQc8Pd`UZm
z(<lG%R1uEdu1B}o1Q5X(LgY8T=2cyj`|~fnDEj>IC&t1y!^y-4({?OISnx6l`s8Pi
zRbf@QLvu4Z4(@FSzphK*{hd>OBzMGIVCdWh;bMvO1`+1pbE_?5e_psN=!+s$4;_n$
z2Z`m&-p_wt=&kQe=sz0>_Zc8L3Jt3tc!uEmcu_h5$K(qI8!-JUe)1-cCQu`-?nQf~
zN*a#w!<pg-0peR<f1n;^rvd&4tEChE_nSI7#EU`y4X-&scUW_`3`)NxBJg~P9JO`7
z42z#5kNI45xhPP{XWAX|+Sf!|Z@ERn`#;xStM|R!Bnt`zQ=z-pHusDUI$-~VvW0LI
zekh6%R8qHi!q2Deva{T^@}v-jzUL`^p76LGqUY`T%r1BoKPd80s+!wQj+3+F!OZW}
zU;HZi!<lCb-11lvO4uHvBtgl=-AsQJ1yXqT6yfY-eC7YVG}>^3j`a}!e?0q7(FtGw
zrj}t?)+n=feC+ms#E!c3Ub$;&g`~ZXm}7>{6uBJbk**I2tk#uw#4}IvbJc&YiLMvR
zH`@PFISOkdQ3QvAaTYfm<#@`eKaH-v{(4Os{#%*zP8&ric<3PqhotHIUoW~iamNor
zdIaeRzTWlbHzd*`!l)EKSQMILQpj;^Sp0DO*P|c)aH$V3u;BRlF{%HTUVgcLyC&ck
zTkzx^Z#YPfG2ASM<po|G{e>^dLM&)6VcU^-GUe7d?p59cQ2bEDh7LbNLAoxo_(?nI
zXD4Wgv&na}_HnCGNlCLKKSDbeOYAgHA6*5%xaY_9sRAmiCFO-2@wSHC3&0U>;V~$;
zi?9+!ak8E|J}nY!f^}qZZG5)8(jRf`CQ2CSm@j?d8d67T7j`@#f+A=@=>w<Pg`1^V
zabwg9MNQnr(ud;GN+SHi@r>`#<q4~4dyI+UxE#jg9e&U*<2n%lkOwEwP1dCjhzL@a
zbnohQN{aAkQc;QtKcO>p%quufiZLqf5N<HGr-kXjIpH13jWk`Zj|nZofOqu*;hp0(
z{TAaDaLjQyz9)3RE8irnipvxFk>6s)kGmTROAM>8-J5xe@_Ku0oy=H`RSbEg?q=t;
zv%x3KFUR?zbjYe3;qgad9NzPeL!v$s4}~NS1qF38tNaz9;D_#wx@T?9jkjT)v&h|b
z)bT&-7`Ay{sGC$MgQ-Y(M>ylR>yLcjDt_dbxp+8-fv1#fx?bn=vzva>qgG>SL%pD1
zXH3W1kj)-h<qUj!Q;fsIFZo`iyWu;2?hPdXN;Ml#=uatM1nai3B=4xkG_L7)y&N#b
z54~Wg)>Th~d+fRu_to)M#ejm+rY^m3(dT6;XVQwFZr-^KBJ?fh(t>_>%qjEhB-iQ7
zm>bw;3EGE^>GTEGmm6N%DdS9*%Tu1~m@a*aZOBZ=&68!2*6q??KqLJ*iY2Gug9F`;
zKGN0+^tGYjnfO?;+PIK6H(?A<ANQ8mkAf>pc@Aa8mZI5x7n?WO$l00O&eM}3(wih5
zKb>V`nhWrVjSuO*8s0F+v-{zg+fm=h1B$J5Im2^3o-ac=manW|EpxBCbgt9sd*$9K
zna84Vw=}v>a7gE(_?!GgH*+!a3BUMuwPJ27e&AfX?iX{jnNaKJo59{cR_L*9^3;{7
zo0}Y<ADk+F=1$pjf6oGut**6n-Dp8-?lVd*WLo@`QT(_avz&9CbS%cMzwX+kXUo5~
z*kVf&Dr7Bc9pPmZKTKXQZAVap5QO7%xKjavyDhYh4OkEyanxtJcF1nL(S{<_e0`U|
z_F@sv|E~z?S%|XJn#mJm2SO6cVec0_H$u1{Iwq*^mP^!`fHT=%OOE;imWlmVqBMB<
zOJ1mx6$I`G-f>K55C$*_=lG>(?Y^7(+Yb|+&wuH}2qi<u5obm8{Acf>cf)l2BgOSm
zv0O2sg$rA)M>Rc9tWeVKLJ@+W_>Gg|#WCUAKk^YO?6^{d=>HM{GBo0xfX(gvDx>%z
z-we12b=Y;CB`73+9OXp$Wre>~_{=>;I8*$<-&gGOlGtAZ<p&;qO75g#l8L*|w%bH6
ze%^DGhOd6>+j{p5{v1DMcdYPjj$^K{%3)^}<;`;~v8-a3I!AJ)j|2Gg`4>dr{lSlP
z5{(cM;hY_{z{2LRcfL&pEry@xpZ<Ja^!*?IB=pjE>id(Puw$2hp5o_;kAGaI)f0-J
zi!Qm;)hU_CbHB?Y9`c~*<tgPRb@k`J`Car|vGQO+WX|e0dCd|6ZzjjVq4-Hpq`vjN
zQ{*0_)BGZAp6>Jl7mC$q?Y^s_kDt!s=sNlCOvs$z6Jg+8DrB<-T^vbXs8cLbcb5F`
zyyN4-pq}s4<I~SLD>_};A3Nm{uu^A2B^gL)vsgd8`z>$k!g<MM{}*xm9Zm)Io__ez
zR>cprGWpS?-h|y8h4~n1<DsLTy!*o+)rDWjU)rPy>L{U?K7TK{@8?mSAbDfld9TE|
z<yLldi;EYGR;I4skjg?g>YTjuiKAQL9p15M<X!z_^=@|&nI8z+cGnM!;dxS5T;42$
zk})A#kE!ztopnXInI4bGKm7~skbZ-@PJ82c$WA}#7S+py&&m(@Cdp8bQPwEW<hfJ?
zsq{Bljxo}~Q;weT0>~o&Mm^$*g|v%Fn_n+u*FK5TPg71Tx>YzM$`&Vd>fi(m=^->(
zA#F<qyRzwl-OmVN*VN0dlkhHOS3emCsi!z=lRAb7fgE*~)gQR{ZagGC_`}nL9WLAn
zoH=4Gtxy1#av&ae0<cLVbmRvi+3c)S##k7_J@Q4lnsCFy!Y7+@7>gMnEN_IPC}!7l
zSSA*aG@*wwmdd33M6rbhBj2F|*7`Aog=HQ2q@SZqQ3hu5mj0OIB6FRP0*+ANlqE3J
zyh<A&t+Y$$m2yY$haj$3Ua1>8zmhf+(h#OW9^en<%-C#m83&WMdppG7-iw8LX?Q2R
zqims%d0?0q>I6b->ICi39Qm;$;qXiIlKzJNp{Om9U-HRZn7mp$vNWK>S-@~NSf@N3
ztfPneOubH*SH2?CVJEND>8!G#oax&a_&QCSLlKjuG=&!Bk@Sm<OWGKG2A~NGJ5%O(
z@A%=;;TRxx`X=*tgIFUf5i#`iWf&Z2N+?NvYTI2u)VmoOM~a-Sg;*E44T_z5YVmZw
zCiOqhG52&E2Rd=H`l!6k7PP6Ae$Hj>m34r&uZ)u!n4`dJ#xllM@<2a?fIn*tgKpX)
z$I0f+v2`Oy&<ADZmA)W66(bD;pTe<=ch()CrH=^5^SF<rUy(jC>{rsqucmp)7zQqI
z@}A|O1-zt_H3xxDOJoj`?z<QR7^8^CoWg~7$`$`xq-<uHr{p;55eoINPd8AVcL+Kw
zXXfq9=R&2S!5S?(FOu;obh;Tp=nwhb3HS9$JZ#=#VbqtZmdWsdy6NVjPVr+$y^&`f
zRxH8kbIJkF0qPAdoD2A%iF!0O#!dp0=>MjQpGhdm@_D#z*-g<JvKEn5Zuv+JLk_I?
zfv&8xaXISb<{NKpnCyDdZ?y3yqDb%p!wl4A6hCz8n5{GE;pj6a08Azi+<)ILK^r^Z
zKYG-s1;z~omLvj(cfRpp-5F?pJTLqE-{p8UC*TeJ?H$Dd3WT)#`gcx_PLq3tpv5jc
z24=UX>>|R}?G&8*gSby8?VkONr^``r4@`I?82rR%zMzZbM1Xwk{?sSyv0I@(lkATl
z^I6?d9{O{{+_RsNR4NgU0_sOU`-SfAED~S=Yi8XHqGc36SYO#>llX(U7wz?|*gD4)
zKTOIW@z4inS}>U<3u*27^j#H>i7pma2rf`wab%lW5rs6N_~AYy9D<44L@+@fww0sO
z_Lh5v(vyc@pLV(^JANOdV4;J8(Hx2&c87)Tnrp99ixHiS2Lb&mes<pR@zGw-+&#eH
zUQzs9{I|<ue~xjx$EF+0v1JFU5GoGvFVJ(^8D~W2{N;Sz^~(;vJRves9xc{DyFcy8
zX*6o-#!-r$MS&AAJzUC(yJXTXO(zQt%E<(_@TY?-85;o^b$^$g9-j=5udDv^A3eej
z!Iv~aRp4-UQK+<IadzHcFN}U6iZ9Zn|4cjhaKs1RllG2K{Cw?O-%0ti&|4Y3Uyfcx
z_!YV)EPm+0SQZxD@e^A7Ko9)X$)PNa-*w0vqD?p62=CZkf5Q#Y2afoJS6qT!M2M0Q
z0tS>0Pk-{x+MePF{^jGJ{=D!+>Ra5c8g=F&%Uwy+Cg3azc~{&dBRN4n*V!!?q#yyY
z0*H9R<Ic{=0B!`Y2w@k906_EQ*ro6E5tdi-h;WA|_k3`#1&h_*&ebQ5gS%4%C!S4l
zG<ZVW;MfZsI8jh@oK%_qmLn#YiNz6Pg<ZO9RGFYC-Pc*-v`+*?lvmaT%?4+i1Q15h
zmil$JDP(=^KaIL_SWuE=%`M0ieVZzidX|IR@Nv5eNxB7aX+cph%F@=&)Dik5j(p4O
z=fH{J9_uh%_8li0L~GUi0?TqL?|3?gZSdbb-Xk<asZpF)tWTXaO!iSgMi+U#NjGGq
z`&)y^dLlqfkGlvju?+3iZqkClEc_imnfuAnpLqgNg9|NCXvxqJ?mSP|83c}8z+ToN
z#IsE>={N{SX$xiehp>%f`B2OT9eo<5e#1|WNiCX7GN)xHI(>Pb#e*Tg;ieno{?z;!
zGf-a9=k+Q#;=l)Y43_K!lzwySCWiye^YW}phMi011Z*DY<?gng9I4uK{ON_IJ3RYX
zNY_@@mib3sf5h0q9W0DB<RR>X(y5@VW*wMyCtMoVzVrGacg!6dR`N$J#>rlaAKRQn
zdW<uS9mTHuKAtsXO<|VVZXK80l$v*kSzbDM<Dm6j^k2Q}fE;Hi8w8kxus*^uPULaR
z)^2%tDd#)ncwcD4F}KPakMoo@z%O1~EH(=^@hEcradRlfxvKMaH_vwKf$1GT9S+-D
zQN1vjHav9=npc4{2&|ef);A=V*CsNOY?W8l{I@DIBOnI;aev)Z@zY_x)hT;+LMG=8
z#m`!@R+4Rrt;tj+#SaU%P>9KZh>+vf98>Jp<T^#jy0W7l!Bd_qqxfMObd%q;%x-BG
zg52rxx>xSkrTB42G#%fS90Hs}-}Hvq)ZG9$j_cU3d`%{jGdel4fdio%6ImwF?~o(q
z!X28#<*1}jeePHl_OY-g9+M?3Mi7)CX#a#L7}5d=cB~w@|I6h*l;m+r2$WAh^Q>3^
zYV`5kMvi&gV)J{1Ub;yB@F$M$TJUf8u!lt4JCk<b`0mNk@BVOxPWVxvv2ek3T-PPi
zGcXCJoVZMtU6|B2@>E9g<MLnJ@R*FkOT(L{Zy!;(JkZ}cbNDBZR>cexSB~W2P8Ajp
zq2su>&wcp>^*`jmSE~?$WAZp+4S_$4#PHsrFz!b_bxic%8=L}_JhKpD!vCm;KP-Cg
zp3ewD^_`;t|8UmXvA?xN^PtSPxJR;l*ui@Er;b(y2k_JbcK>p8SwBa;Vezup9^EvZ
zAjk4uaPcKZg7yuR*Q@s3M=Yf78~dkU5b9hg@qRCL?W(J<$!(j@-1VtynU;=!*13Py
zGp^HuZ7gR{{-oVWVj*<SdFN~15d=94HHISR;7i`31c;|U<;j`+`ti?y8U5s!zf!y=
zq@oO7@SHuPt+&|1V|R}$uf96^$~V6)_c~pf@aKagaXx<7`?~z;7rr9aV^VZdb}Ka<
z+QljnKpgp@_jd`>+NAhF7q;eFYea{Mf-1c@zVfQ8qYobWDJgenaY4pipXa6rkK6uH
zqG;W#iyv4Eo_?0xDV5om`|=cFf-=m<R%Ma}F)Nq$-*GNID?9kCeUbhM&qeWIVVyX`
zev>|36?{f_K(nV)DKg^a$*WpEO1+|wML>!Y%<))qq~ppeXC0#LL3V)pq5!4#j|>um
zht%AUJjF|?(5JKBaQM9?27wX*;Vo?pYXF|mW6orufI<)bKN~67;R($aAE8gUQx*kE
z*6vsObZyg;586?<>(<ij^+a(*!i*tCYM$UKjf=EtlX;H(A-G^{NAMk&fiQ)Bhp`SE
zCTz*VV>_O+!m|7rv%+JEC{OaG))ivK!7g7FmP;W}7Z~>$f1pJRKz^C`>2BtX1i*$8
zn=&lp^2)AE+#(n@OyuptN2mzju_m!OhRvs_SBbo`yISC~oy#jq%AzqM9hZ4Bd`~-Q
zVeUg+R>eTjX58F|`owPQ?zm}O?%OFCtzNl!-6ca?Ox7`{z%^xtwE;`Rwaesu$HLsv
zH_5pcbvf&@Zi_?DLA|41vC~}B^?7IB$Gik3F}u;r=N;_M;)%LS{{bK1yPH!jPvE1E
zf%n-?);41O<yuNNgw-ed7pxzY-=57juX25Xqn|#os^W(l&|Q+<w2gz$u<~NUzP$AE
z{LAu?exs^jxolebaPD}k{P0c)Yl=?a(DASOmL6k5e6%zAUAMk-{e){HvpS~JPOL9w
zO_fyKc1no;ivECcxwmuiiuxhiE@LVEgr!ek!F*cNVjPx!D6jvaO%u=BWKkd4q1PQh
znZ!9ykL9IxlP-O8=+bhRFf8l6*Bpeti}GO1(Y}khojA=q<G;+&7<1f4IQZyXrzFfL
z`6T~Xg{~odRelOuHtB#TJ>C?}0g!Rr=@mrD6x<i}$&DQ;`G^qc)<v$KB*(WbOy&G^
z+-4k+yrB4r@A%Q^lWFl|{bm^s^{N+nIQog#4?TJ5_3PgceOb8mR0Qe^i>Nu|Mp0I6
zWv`BNQSS{I{m{S0e6eQr{o__!Z{1JSk;5{dR++1Rj+~&H3AlZgoa7^?Lail6AY*0m
zGf1NyH^t8qDU>2KwR^_^06+jqL_t(nPVp0|#ZS27Ykk>Gm37At)(FR*_|>i?^AtZ^
z`hLu@#|7~2VnIJYQ~wt{dr!H)<S_vZlb%)4dk;TC6cil`tb1>}jVekQe17=TpGQCb
z`7aW2_+UYL=$qddZFu+O-6>swB|P5q;g5;}<Awwz$G2tO@xul8+<$}9$(20aAjv?<
zu-okjFQfSBj+yTR-)!P<P_%<ry=?E;I)`JWuuL$wzZ^OC&<7?hq`-67$Bv9P5C!0?
zU;Z-9%NI`gO5%>BJjKuZq};BP`(e^<mz{Quo+H*+Y4??Houv2Rr2XxiJD9e;=hpVA
zPj;Xj@~-#E;wibXc-oV8j-L6nrzT+e$UEvS?~Whtm^$x*3%l^Z5AHj{BCg0$2e5Yg
z*2$-+^-LOeZ?Trze!Ju`VQH8$-(NlHJ96yWnYk0Ptojh*-R;RwQmdcPkN5BW=+x+E
zzy3|yTX$f;?cjsD7Odp;sLy|~OVKqWLaR@F@O@o){(8|xVzG6i_$S*23a=wS^Z^wl
zhmo34{22azEq>7HpfLU5p>ONr0~SIbI`Sw(t(HF0A-HeJ+oR;3sQq60qBz`MzVm}0
zMZfs%X`LgX1Qtgh!D~_><ic2`vr7}lZX6cD@RQ`U?~!m8csR2-a($Bc3y=zVOe>;L
z!~Kcx^|53<o1o3;*PQz7I#MMJMGu)5$K=a77PC??!YuO49ZdXlmN@P>H+OVK5XbY0
zHe%@!mv^wjcXQyfw36@e5?}}EN`tsTWDVlmXmzlY_iK3`UOtm!ZOGBd=6AARZ);u$
zr^{-P)u^y!d%as+`h)abs8`&$FA%`)Hry+A7(RI3LoCVnCU7QG;LFu5EN+dHNtikm
zKS_+!uQD89C=b_7W#aS;Giv1z%F}A?t~kfk=t24v!|LW%_Qf;VCX@Yg2**6ZwWlWY
z=Tc14l%yj2OZ&Y*@skbZe5>-X2zNbU{pY`Tx=rBBN|khQuCBQE`hIP0jUYd5@k3p%
z+P@5=kE)LwQ%rE6yHD%pi}hXXp<B?kw&{i+SD#r~+|RH{Z|b@@z!)*L_~~NH#CqJe
zQZ{<YeV>cujvr(mZE-y<er9D9n(1>So2!Z+&W?4vVqEcZG}k-dBuBTUb`e%r{B&2J
zt+(7VdZP%_(~I=9@EJ>+opyMv3iA#-;^TT0m0`092Rn~BDk?3Qr_;(iuETJIPo`a;
z_{`_j`XcntQ~aFp^>0RJo^!5(+926$f=~D`6h9CZDrwS+pY0y{kmzMEc~Mu>KIPP(
zisje((GGGa31;%|IsBt);%@v1ahe~`JjKthPue+pMut-61W}ru>6K<79(zCNfLF*p
zL)+SC{CvxM-mgc_#UAp~wBm=fHr{YUwS2kvHrw=Eq;e#{*T3`K=udw+FD8?6-*4}i
zL=SsVQrQLom%p-0J>AXbEw)Y;1<jxS{FhuhiWZ`z*RSG-IIn>A2S13{*iq-n=kMk2
zpU~x5{5ZwW2_dMykAvbz)C8*V`H(1n()}Zra)--Nn5uk`u}Av?DQ{IM$oIpAZwH8?
zHSK=zlb=PWiiOpR<zktjg&))aFEB_f1_h_C4FfO}X=sqH8<UGi6hqKsN+4)(_x&W{
z(zG@4tx76YR0-8S4{<gnL1<Jbp7^84;c0y2$?jG=g23Wx{v;}635pskhm|W)=z!Qw
zVQdE*JTsoaPc1J8=J+IjM3aX)>SvrL{z8W;lsvhbg!uRB`yICFZ&eO-%F7o(RfJ;_
zzgNFLCfVqV%(9mrTU#AO@uNH6eTOtx+i<VCRj1;oY?<~|QAP-C>yUm)pw9N{!OFn&
zn{F2TsFXjkWvnNviMqi&!8h-$tXK9cYCCNntDrkP>a8*(rU+We0iEkIhn0@13B^w<
zGd(4$G6$^h;y$xUR*bYs@iXXLqH<ve2tpbdBkETC%*>OMA^e@Hy)1rinOz>OGn03E
zK|u5zdA<6-H%FV^J$VI2ompSLT23mE>p)hDWwr7BCfR7eo}5rJO4n2IGYD~WSMjq<
z6h9p0;M_13KP<4{{)U61%{JM@lGo?Iue?g`6FNqBAchJZgsUI_^k-usUD#P;@$;@X
zzbV?_?srSzS|%oIZ-4&>6MlnyOrqk4;|4hHZINDT+2P7n|G6f@l56Rb#d3GedC`}?
z`oFON6sUd`KTmwz4sz66_kB)ZJL%hU3}5G{w+;pF?TxQ_RkV%V4-`5Ep+nyFUVn5<
zCuA6kA5!CJfk%q<29|0ZE2u{~b;wA(anua=DsgX)IqpY#bjsp~$<^VXJSzIv6<5Ye
zP!{{@ykt8qjACFG(RHSvnT0c^{Aq=%q2xJ;;%Cp@c8#90^Akce`hMBvf0yGIKcDcQ
zRK*X+OMFP~o?5(^i(osBviso2KNWjWUN&jQle?Yvd+7_~aC<rF`#*?&C3ju9OW;w?
zxLrrS*wILMP<vX;l(<Mg&ClOqBW(reXmL;HLEUlIM8iAcMHDY6e7y9<)n}I{vWU0w
z4CT)%Q`iWdF5L0cEP7Dx;IIY?`D}_J<D20o0iU!5#mXe`5Js9faKui$@Fg5&k9%=&
zz@YrGNiwlqz6hDRCOKebKG*nNBw%ck-{XXP5FQc7>XJRjDbH1DU9n=N%w;&HGJDr6
zeH>#&)uZ03(mBcT^ENCtp$N(2Bl8fBDIAWQr$?s^qrKS}-BiAc`kXe8Rod@Kn-69@
zO|fW-H^qT|>ZJ7nC}%ltuwO?Vq@AI>cN@XPDptO2Cd+#j7s_I);F-p47YEwhch$+M
zys@{=G{zMdV?^DGpK(6wtx7q{t-!6iHzzAE9Q8IUa`S~tmPTvO?|-bx!dZvkRPiHR
ztm~+^^o5qgRQymGp7eyBqG#{ETcTbt*nH^7PexaZ0^+{+x>xjCxx0r+3qsOgpZ5C%
ztbHJWd-LnK6DE1b4`<)M{R96`cGM@22D5l&pOdKgA@BRX_=VASqR7Z|$H_fS+$9sb
zeic8D5~271yzu!U3VlE6`=>~Gp5Ene9vm&U`DUBir#?Be{+)+?NblJQ1M(C<M}6Uo
z(S>sCpqVM#5l|>FP-LWC-cfI9I3~(l%N<czMLpyJ_wQ1k89c|vaa5j7ptx6v`-Reu
zd#x_M^wKWDi^WSXj#&sGqDlK3-lBylFoP(5cHZ&v(O%Eo-I9o(%a+|5y+g{xXeA&|
z@pHjN7e}A@((!Rzdtov2@x$KVm7g{p^~OSt-Kifu^sRa^y3x)PW$9rb`(%9N9qTBk
zbZ`obogTY=^pfX2Cyb)++$r^kv(Iryy6M`>DJUpccIT;PgJn5kAT(zQgiCh()B&dv
z^ea{-fiG~Rf#pxr%AbUlkMz?hP(sTdDRie)5!%e(Wmo9+`7~bIv2Ds5fk#g1*yI+r
z#wUZ1vnvQe)um7pdg9fuybzBxlvDE2QS8W~S&jn$hdNJ}Ss2k4EiUvJS_Aj7rjMF9
zKwm={l@VA_9`uL9+_7kanwqj3qv_SJ$ABYzp<J{rjHvSW%Oe*3w0Nsad1Go5EVaI(
z3EchS?8P_j|D?mEZ|m1(wFmN!>xB<ETD#5RT&&)|77s`pbLjt3aFI0lj0VFW(wTn8
zX0#WrVVpFDq?Hxl=!BkGHazHH(iu-y%KajE&dgxB+rdFvy=XGr)NP85_zYb*qt)ib
z_yNC4?r5PdH0ihxYi0lecA6&RDS~I_4yZ6Jf52LJnC~<WzKs`IVP*V^+K$bOhuN=`
zp&1zHW`1b=nH2@_!i607Sk4)Li=fNE6vel&s$~S8I&c_0{XCmaw-<iupsgvYSo}!$
zpe{WGSp6x>6Yv<0UbL6RIfYAd0KWDbACfV;zwCqr>$~b|%iSrB)@M@lRR>oYBc_U<
zIyf0cx5_bUix<eTw)LQE1j}7w@pGq~Kol)n9IYc)HW;l{EU!2|j(k^fJrzIUpi(xW
zxoYu)MND{+bCZgntO0(#b=QsF^Oi%pcCq~EXFrdA^s`??b`KC2RKDlKAJHSR^2+q(
zFMV<Jh=)8l5zK|SANkbL(Un(stS%A&B|j!r@$*nQc8nvuY+w)Bafeu}aLk)t5G;+#
z7F7Kzezw|T^XN?nzPd}P@zY=aI{N;pKM9jeynpCj?}*k}x-{X3WyKLkf4Xa7`qZ7D
z811p!(-N?Jr0*v(faiz;V;50mq+Ook2g|b!H@LeV^>z6_{+Wn|b=2b@v%TV5Tke1f
z9Y?rvXOtbM2;2*w^Q;g=-@keC_f@HBJ8o@)tt&EtNS6=cCd{yoIYe5t5Cv`!#m|nS
zym-O0yJ_cA^tZk5Fd;RjS}%U*&Yra6P9UtH(vE)T1Mhfi+B<5S;)im0^J`z-Wr=pf
zjW<Q_`{<E!0qmXe0!!PA{7_E2KlMq`9=mqmLG_WNJ`?@l-~XZc2HqA>Jh43Cl3sPn
z+75r+vdO+BqE-T{3F6PBU+S#814w8vP>0Q@hrd&9sBlTeVF5r5uPA-=C=n6L#7Y;>
z6IHN4m}T6jhXBXC%?d{qLVzL^OF8&!`7@;rC@niIUP8+}t)~*M?MDFvzDqQOWV+x9
z(qaMSlug1{=o6gKZNg#_Bgu3-O_5sa6V#H##5$V?RLLi0C~OM&Ke6meWUN=eo&cVs
zAA|6SIM#OUIoH@%#5dY3QuMGNE#gmJ+s^YZ!WD*D-$C2Q_0vrf2PH?p3M5TvL_g51
z$LtJwK%lspvo^6QBz?475o@c~7mH7R{7~POoYkqWfWX7>1YU{PqPAo6;x_gxdc<V~
zi=Q&QOE7BsF%k|?2dRHq7C*F&1tt3z`lE5{Lr22!2}!g~eOJ#un93=xWL0*+{<jzI
z1n?aA2>HR-HdXvoFZ7#N+!ZbR&kfN%w%D|<dFvo4r<FAQC|;Jy(R!=pj-LgKmqd#K
zB^8!Hi{@8T-VBl#i_=r_V}SiU=c?idix>pACM9c9@sn2`2fS>b=)qz&5xOg`a>u>B
z`^|^weKQyR?QhYizxZVpREBWjJC+n!RHPkukev3%Gt=H(-%N~bpi2i1NAa^<tdl+%
zEnmJooqW&lyoE=oq)FfL1JX6+E+>wALwS*Q-0Sp3vG_^zB*YoU$-Cocp&XBP*r9Ka
z7N?Gs%eyxUoY2FaO<ejKy3_u6hTJK0a_FV+PmnvHo+o!frQO#~`i>kYm%M)lnEm&8
zX@t@%?Xna<OBOGV{^vz6i0*sudqvk=d#xPHc66-B3FB|M`R37quh=({Pn2ixI7}?g
zR=DF7H{E1YIo|MfT>?vv)I0XXug1zT23|y)CVRF;V^~MET%^us2|x1cgV8&P;wSH}
zrfaUfF8YAny%jp>Jo1C@*Q4=5|EsULHu~Vljtsr@oqLmB|Em4F{0CC}+<dd%H)YB;
zLz-Tsq9{aRn06=@|M9PXC%oY&WrFlzuV*|xddd@bN`!L+!+Sn_MD*VqZ&ZJ$sE`N(
zFAYd7>$czS;fXl(yTn;`>up{r;rbsdV;2=ggY3wZ^i|SLVe14}iE$qQs&l(tLm`L0
z%S>1m6%tXcZyW;1ik&JmRLQietG&lSrC*>!Xp%p(HqyjghScBD(?Mu?#z(Del-csG
z$}f1}_^3~Tq1LPDU-CdUoDxU;6;8fAd+-NejNe!#aZoCmqKF8@v18thH<}IK!f0<!
z`6H3V5f=w;sIhl_8M7bHuAx#*5qlQ)ovw{ZTF?59J-!io_3PuCv-&6>gh#}&x@6D#
zC>jq;z4%t=9Y5p2k#<m7rJt;t1ID|RVgY3R?cn&_F@h)#n`&R_`CUkB)8kuF+cDlX
zmG3a0{l06zS3&X9Z+cbXN7)vt(mu-g)IsVWeYm{nQQ&;jF8CWKW%wxhIM>o{bKg~G
zS<K~rrD)uQ$BYqE#ZR5Jt=+g#{LH)K_Gt0abyd#N7HX%8A1R5tj(TI*ziasl6-?8>
zx!2_#Z#uZ^^71?(Jp%ee-}HvAJ1;Jl<F$@H_PA1Yh705O6KkDtV$80R&mQ;X=vDjv
zj|kZ2N5_BTThUp6I;Rld26ww#^iH8Ytt`0wpZ`=V1M-*W?zaB=D$M0LIjn?2$Dpvo
zqqd7)`uyjH-uiyRH@+F2`KNRBy`MqyHT&%wJ>b6gN#H{{fb!%f%<TGf<jt%_!l<E=
zCQI=HEctrOBex5q>HC-e_Z!jK=l;3NKg1aZ<|%%F*<0?m**<l*%?%<X|G*JP#!8V8
z*8}f&UoHEvP7r>~7rzu;bjjaBP<?;6Sb9;1((cT2&W%p^`Zp8aC~^)x<P8zew98Wb
zJpWmHMmz2B7zO3t9`fmwrRmsF^t?Uyh@SAc$HoEwz2U~_{U7~UT&GZiy-}>6(nk_v
zRmJg!H;K{$rxSAd2?fg3>*t8l;itd)wI~>Fmc=;2s_dgz^6G;THi+T}{=(<9`@>mh
zM_>Kcw-er!8;Uy=$)SU;LqGWcqFdy6xX|tOOgRE?$L=HWiWEP@fAi}OjIbyR-L1FX
zri!TBZbPA+@A7!?nd6R+F1YyOL`3Lb%|b~!gL`dr&*%-Jq{0d;bf5e3iE=FBUsd^F
ziU<@43)ff}9s0(DqD?m1C<KrGeEtQ|XC;pH8y09v4i;#uIl@GQC9&d1YX*=g9jGOt
z6F=__4$>AB&hdf=HA<i6jZ%ckox|n8<XvDKob&fs3LZ|m0y0(hD2ilUaBH3<Eqgbb
z$cxg8CP35fB$ZGEkuQfva)K2OiYh_tH3&*Sd65?_1K@q$Q0f3{{`w)f@WUD3d8$%K
z;6oQ%A2|}^JdmbB3Ld4EH00l)i|Jky@sOr6?_}Szrwn)@OdYg^M+;eEA&q6y00y3S
zEUpYw_tWr&X>>ys`;<MyeuY>jFPusC3&oiG3+>gf_lK<-bXlHMOQn2lE8jtQL>!xI
z*t2}f?1HYZjrEgmov^+5#!+u&Eq>~xt4>_wSWeb|)yantR~h`&4HX(ozZ=dULOk8?
zD0}AVJZX?N)s1g(A8DR$jLtw&6qW&==~dKrj9;VXK^EcK@3l<vGmdREi?{vyHJNYf
zAob5#{J04$-_(6++Hwr8&S3i`xiR=U+GHMaP0!Kt(j7<tHuqiQ$j>HeE;Eh~l0IX^
zRPi$i=Y|{SiYu;(Hs5^X%8A~*VtF+Gj@zPza@x{b^QL$FR5X2cSd;JjwuE%Iv>@Ff
z-6=Ujx?_Y$H^PvXR=Sbyl<tnvtu&)SVw3|0y!-r)@B8om*s&eYbKUoKo!1$54!IZs
z2rl83Li^y}@rG|k4#&Ud9QdHFfQk_x$KhqP-pCpr$Shs!8luJ%n~?UXGtUk9PgC(C
zE(R$Fl=Tbr5<2w0IJ-utacdao{;dCFS5?DMowWH>Fvu8wu9y%#DOG5>t(iH?=NGiw
zxkX%+FCcdT3)=8rjy8XuCs#CaZ2;cP6(vU3wp~?<#E~=GcAh5%r5O5MXi10dNqHRN
zlJ`U<z^FAwx_O^E{=4HWA1+y>NX^`rmdwg8c(BS>y5PWuZIoAj356JM16I3@Ewv|$
zH(rS026o%%c#7zw-BaHnj9!NBW`5&L1?fE(k4fZ;Ymzr_F<%gS8smgBac2F(@DM8J
zwh(su1#q3oa2CX7G?2ODy!S=V?PTm65TqJ@tC}Wmo?>QSl44f37A)O-$LGhHaCI^-
zkcW0|h9fBU4}M+5lLu>tWv7oyJ^Yx9{THAax&_N;i<zp6iO?zcYZ4pI6>U0Xk#a%r
zK0QhAeeCA6-LTDTD=6xT@e<c1ba0zV`qG@o`Om3j;~&WD6Qb*VVLq$T_eJPhPH)OP
zdEN9=if^#f&7S|wnr!{zyvL}wq2?N3rg9~rfma&Xu|uA{he?eO(6T{xONwN|%0bV+
z)w2d*;!<~f0V%deqvA2m7cVdtOxIO%G`>%S3{ked&tIPkAERFz9zDZXunT8`jLtnF
z`6Qtq-|MljUu>=^q0?N(3=RL@eq224p)h;*#QrH%i&IvD+MBfPH&Q7#hz}oWL`WT8
zg74n8=Lpi#_4<L%owz=)<(Bw50$rsDwb8e}AP`x&o(xcQxwGJzCMn~aQ|g#6<uOK%
zjg?;-__n~%DvUvi<Gs<<B=mU1%RiV&WvpaF#8~@Iq=RT#KCOH}6pqCsJ%S=*>OX;P
zL`o^a-jAEIn>+wzoWAe;2MQi%_%oK>XX4lEw}z3QKHAu1oT<O#`=@CYHC=eRHF`vQ
zfgdB0LO?y*uvt^DW~~Tf!PCRXqvGmAL0aQoDz5VMsW;Hd)|4%+4v5-o)m{2h(sN`Q
zQt(|g5|D$-&l-4>@BdI=J<1CMNXhG*X#OfcYx1wgI<C!TF#<~7QpKU@O+<=~wUvTR
z4r;Ha{FFN_jkHRi@RjZ@8wPspcyW^7#M(%kZ5g{hU%_&hd`lkZ1K29aP0V}dD?U0~
zI0hzPX&C)F_M(W%Uo)5jYg?<_Zx>&>=Ujyh1KufjbD1*4<ddb2E<%()Y4Z#N(zbKJ
zqh`XRfje!66lD4z$b7^<zm#JXxUISyj3WS}N~bbV9we$WObQ-dgmtlpx!{M`ntm+s
zIkLY%&@$jhZFS5Mhh%9W4|@Wv$lDbYVZtW4>8j^;jHck=5u-wx4l{BlYR|iRbZOpI
zU>OZ#b;Q^E*-<G%4ooqOgXnFa-Xp0smo?~>YcQb|&u9T($Ta1<P6ofkuA=eU&gX}t
zwK6zwhbkCTEd=)h^zX<u{V9Kaa4!LCVvt;O8yAfUd)bg`{mJFz_DmN<@T5wD<`@Nu
z-`&*unkoJ=5X(?)!3j9PpZ4v7-p(meeYP}*UUO_ous!Vl36AMq;Q4ai;&hXeoB#dt
zIT67C24)y_@QMZ<q0kX%2g|=j0eZ&unfcsr5kt>xl&X0){kE$Z5=_)OtDZeIZoX(g
z6o@E{-;PB|eoW<XoZy8#9sDbcwcaN=7;aG8LigI$5LV8AnupG(C8-L+5PHE8D)3&0
z4dQP1cr);n5Ek`#aZRM1w^6st=6d;HDr$a#klip0=zP0)jr^tViwljU$_Z;4a~tvS
z!x_GP>@Af7R<Al)ZOAiAm6V|fm@lEY^P)Q`275prvGWf@#9mUx>YTlv@lRt_YUL$)
z2=Cd-d5$_NzQ|3wY{jo0YpmY;Eec;-03>ohK1vsAC>~$8ozc<Exf}=2$MlPzcj=OT
z_b1Z_1cQ@{5bJ1_G+Vd=()|XaeMmUy)JzaCm0$h0Q54J5E$FV0zbSdUx-x_;9jh*R
zpEuOK*EnJ28vcHJNhivs>z<5@SVDNqEBOew4g`c4N|WgsrsqOgs`qkqEXNC(=A8a*
z39o*-{z4HDlm|NaGZ(!DZ@HtgdX*dis~k{Pq2eIkgf3hOSz@XMj#}eZwIRQPb~F(a
z;?xCd1Iu}9PaIhDP7+&T@?l~<nS`-nXu3V+$|s{FihkLU53;BQ^E`BwH-zi6kgNT9
z6g-OC^cz2BYAZj#;GPsZ+Qfc#T*#rr9k0UmJyIQ;>+g0=$+&BFx3u2bL_wQPR%aS1
zR;d|4()X7v{GzhI#0cjsO@dGsSHh%23!z+<(1ojX9BtQwB#x7nLn6AED>013Z#LfJ
zd5Vte)`p0apoi;w``{vOI*P{a<95koPO8nPO#26`MxBGB^gql46lUl>szazhV*LzB
z$bPa0a{V8?9dpDzc`4{tPR+{EH_aG>D1W<{eE(-EKTJnGlc@>$wWo%!cFvL?dJS3v
zMRfvu#>Bvz;~!J)ozFWf51c8t-07X**iRIUkp-~KG-YuSOJ8_3pB}uaNJfqwF<vDG
zPZj|WN7S33N>obX5)X3>A~WX2PscOC(e-1pv)&Jx_LY{gv1_eeqml}QS3hfu<~hw!
zH-hg^BMASU!JqP9bS@-@@k7=Z!fwHD7>Zjud2kFKqWhi++65(@8<0$wp(e+$9btVr
z!%KQ~O>q*8*7Zhpf0VNVj~${vLBSs_mzw%FmdQ^Gl)rrqTyt1MyxZ7layecx<feft
zUI@IS!uL9dkxY-ZJD}ZrC@x`lgsi69=Pm2NC8E>Z$rEu#PUJu3^C+gg?`WNUqHmbw
zLl77|d<lHV;mYR~_hA@$^4RRCkK$5xj~Qes*!HN(r_McpO6Q&~Jn5PPL+By}66krT
zw{9N=rU^I-){30!IyFzmORVa#ZY{35x<5jRMwEYeAZQtmw21e1;izCy*63T-E`|hW
z6)0E<Dycy&-*ORakzo{tLk>9K5H&}=F*GZ=pXuhcS9|Q^Kk-t*k61&vq0`VH=r2X@
z6)>&w6o=?$9&~R2j>-c#+LoWImH%uw2l+O=A!bhbj8LE`GmLo;c*ah6Zg{6D@dS^@
zHk(I0cj>0Ja+;KeIwqJ8t&Xm16b8bMv`5Q{LL;*YbiU9t|4{!s*fR8pP|l^K^`}_}
zMd+CtqD8)}p4O(f1cZTVn-7dt4?E@)=8@^Zdn<+S9ubneQY@-iI;nz3E_9mT-a|$h
zHWvsLU*f|j<&#Wlz{3wugp)AmE71oouN!aIAG@37{}j);8!W#VfIks1E%X<buLVqS
z#TI^|(APM$*h`=YKfK9J^g9YG>MoJuF<x{WT+F{#a0JuHB1ZW6BB<78RuQu)>(j8i
z#Z!Q##$oshhSj{DVk#Nr7|tYwLEX0U;R9V=$gvo+Odng<H>+?vNp4LE6?A36!;2Hb
zpOra;uUfTkcqx%A-9BtIrpEiR<v^HQbSa1ZzIui`We|mvi%1aWMks0;BVol40KV#8
z)0?MNh)^(G{;SW*3#=MYIr5N2CQ+`xF{AzcI{w%0&MQVTFoP#q>cjScEc$`nddwfa
zc^<pOW2+cGfZPn`my5;h1DNEeG)M8I0YNpAqXu`-VpFr|ldNC3%p`2xmWtM(XRfT8
zhy!vX@pI~)P0eI2fm3~fYj9s?U`R0UQv9@SiZ^i9_W}68HKH0h(1<)GyTCzJR{=ME
zPz=({9<5~P;Bs9!gn7!G{uKNGUOVWbiKXLw4w&H>&Bs2Qk|hobq-j5Ez4+uL6#%t1
zJw8vZ6B{@d_&?-RlvZRCVt-pWJSQeV^JllbqHYQ@O8Hog;eX!c{HOeS$>6~#CZBQh
zCC{OzdHDo;#$@Xs66p}l-E=Dx>ZIZa&{eF3Vi|m|3`?gN#z&nmdOlG}-Eh&xqCAPH
z54&hfZmHmzDCt7gR0WyDL<Ic!<>v23eyws+1UR?af!g5KfN&Y~B)&y#^;y~R`sd@8
z#d+o*77o=~y3v?S9ZvEwbhpK|*t-Cq+Cp4j61|IuSw#LXCx?gEG<8FL2rKR8zq{5E
zxwtM7IuXe3CtRzYe}uqkCyR!Bb|NhT_eqoTzSVnZV(i`O?i;kt>+ublb;1-MR{A{p
z(Ny<I3UDIBXfaB`%eeH=1LP$}<~`0j_0HX*nNM2u=Q~MkYO{Y+$-O)=;n&Td7xabG
zJ{a$eG<X_5hwE!2I@Usxhhv^#nx~Vdt<Gp4%D&(Qecvxv+#bP;MTN~P+%ByN<OXUq
zPqnPswOTl^eawn^Y%vnYsNR2OC2M|U;~1kZq$LN6>86#cbl}qTwa>}%MFu^e>6*(Y
zLV>cNJj1DiZ}ROAs|H;X*y>Lr=7~0U^9j2b^*ZV%*b#2N>+be9=lFEos=ea!t_=j8
z%3ggF&t>)81NI~X2q=zU4;GG5PZ_Y8H$$+$t&gc_f~d<yNfOhtXNW;ST~;~eyK=@h
zB`pS(KOO2tnbso%IO$TPhMO|GphfYa3R(sIvag}}isqx>Q6^UNVxBMtZ}1WqIS2EH
zXg4=Xh9eS9Qi<mcmXc!sVE_!T*3chSv=4=zM?ZhCiHqCa?NXk+ni5^sZ`2|=jZXL6
z%W&S2rH@1WMp>I(GDiI8Mg(dM9kjlQ6NhAWP<lN4`Jh=jR`=!|M?L4v6jq-)p^W(u
z=1p0`cI9Q~*=KjrVxX64hs8y>`oS7~=<a`0Y?azwHF(bateyS2WA`6qt=Jge)Sa`c
zX<>+eoPOM@4keC^zv_0RC2P(_EY}ph5*xcp9akohgv-ti;k<`J!}x4}N}~dFr$(32
zfL|7Z<Ur?NqHy>xj^mqyzqQ<&JzbSd0oQ{a1&NEu6Db&1t6}Ym3CvE(p@ros`Z1C>
zq2;Yfxe{C?{s*hCm20N3xEmI@Rb77a>cXoVymB3;{Qr0?5<vA<wn&^?WUsg{c)M!<
zH9-!kUOEJ`S`jkCHL>=OV#j!U`l8^^Ii_il!#^I-FAk|%DRex-(VA(D@dU<_*Ny<<
ze#Z+rw8MyWI2VMKcip&gAx~UGM3Zgi?cc74ET^kbn-_HH?dLg^J=WS;EC0EPdB$-3
zEu%h!fbZYjx>L4u;q9;o5|Vgql6d+Ssafz-@Y2<B7dCfHo7%fGp*7oN47Kvc1lcQc
zgg69nT1v3BY~n&F33@zZJQp<z?!L`|?APvlEUw_rky=~5uHzvsQBGU>U>JqCE?;B+
zo4g@1hZ`mk56X>F$0XUGKL~2$aK_cXT;R^}1p8)TS#!Y=t#{Ef_T9#5v<E}bUn=kE
zy_$8VTLRH}+^+2CRQzQ^)}!;TYwl*Be(TVp_WA*x>l<^-2e+&4c#NF}m-a(_lS)N=
zRqE=PXXa%F!?qvN>xdiIwWMEJLt^`ot?c>Zg;?V&2hi-(MVW7>mbsxb7Xh}oWdlQD
zf<?gnet_xPl(n7`A_8bTw`iyjn+HLy7<6d3L}nth+Yny&N?y<QN@dp+<#*BN#vkP~
ztg1J<qQXLh4jTozS^T)VGXiWiYl&LdOUU-B!hBY)NC}tu7nRRdI*bzz23O1I(6?xL
zmTmC^`ko3qYg&7j-4om%|B)K7*NJfdXbFR$0iDxGiI3d{sh4An&Nm1Cb0Gd<ZPrv^
z{Gg3ruC{``@zQgp+R%7!aufUlQ!dn`%IloqOEjc(X3~%HCFYQRnB7luO`zM$kb!}E
z47oSm(<fSu-~{;(>P8zT2|Hi<f-X3~RW*gTdt<%L%f`Bisq&b92$kbfhwo+txsvdP
zm)bg7ljKhk0z}_OEv5VQ!yCgdgs*f1S+(GduPCR6<GXjsKcRitgCVIziG>IZ<70*i
z_dF`~sR($@l086iU6HUI2P|l;LIom;``n;kdxs;K)u!EspUB*{kw)}<hQ|{jmsXck
z;IOtk&8(sKknykxQD8h{`psLK61*Le%WDnlYP%gO_W^#7b~j*Rd2Ggm@*Wb+Xt#TK
zm;@H7%u^x$Xf$ZIg>>_T<b8)a-P&^CjhX?X)`g#Ww`(+T8EdKq(RQ@6xj+uR7zd33
zWI{Du-(%Ym+)F>*y_Hy3XKnc6hBO&8ydy64kkxm{;JCNys&cDEy;q%joi}2<c>ISq
zHo1=`mKFmX-^M?pp#3nzbkXN?b<CmATdViI|2VXwT+Z!_U&@UQK23v%imK*oAAt!j
ze%3DIlwTVZD8r|A``=Httn!@c!sgs%PUF4sthSb_^6Q^_6GT<S4qKE|9|67Shcb1z
z5jSs~_2D>q*U*G^2IDteB=2=L&eKReMrUZw6Y?2`bvKUh2GaH6ucY<hH;`39Q&CHl
z42_Xf+t{5Z!G~dIC=Ngx8cRm|w2f)?1^4ppXPS$>;O|}JB^#*Iyxn3@7VxZJe~ezv
zQarD-U+O--I<f&p#p?^8ztGXZ7<35p9DUq{Ea)=%ZW|S2GQ8V5v@HMFSVEV6+Zl{~
zH0xWa`e!xVSju`BjklmNRN-nB#jTRy$F6S<@leCq@Rwth;Qn`5aZ5hxnZ=H(1h_aL
z!JzwuR4yr+==pK+V=oD@dYC_D9Lvv^Cc)pMzucr~w!YrPYVVh+4Ia^xtQ!7gFrN3A
zr3e=Z=uhS*hrr(Dy2)elO%|o5a>eyCX_NFOg+z(WzaCl5fsh2J7f%THs-h60xna1k
z`HY{x`a_{g;zX_olO&OI(kz~O3Yap|^9qYY8Yu!nT8yBBgoy$kLyS*qt44IGq-l|7
zx6X)2DW_NnOJWZ*e1|7!iuHBTU|v*IplO3#I5zeKHu*N`%_<GaU}4bW{>Me1NKt(~
zZFu;;RFG>LN_TjTcbO)QZLRIKX+qG!+DM}IovZhucTAlWp5Yc#v+|0f<@r*XfQ8i{
z^=Wy9lQeojDvQn9J2p^l`~9-Ru6KG<uVsMN?_!93e{QQXB4p53_Anrpc2AJj6!DWZ
z_TL%8i?0QrXt~ipf2WS%t#m-hbjbs7F~T6ZH-18y=f0cCdXb9WVCL($(nebKM|ReU
zOyI?(K_u|V4Ra)6L9dv(trgQtXHI{(Ri)o0?ZL%%@%qppq?~N9%<A9u2RSL+^wLKR
zIK+){m{E)J$wgZ7WD2v1KA^|AS%f;EEdC@3=5sTYAr{V-3#&?pobPSze1~n)iCj6}
zFZ*JuEj!muV(Md;vZ5q@A!kacn{OGVq2{!6f436i?O55ny~ryyRFV4qZ<0Yt5-Ncm
zNCKBkOdH22czW9P>SvRmSV#mJbV(wMVZ21}$vFmPykp@b``KXjd!@J}u+TIi+76+6
zb?%yHN+6BmYBx+MA%;!~O&U2Nv?N+^qb>cC{dR6dC{4L7VKnY`z^hfIND@9%)Q<O3
z*s^fYQ2Fg#3`PWT&^5t1&ceP}lpT4$(KVE_6s*hKq$Qo3i(Tyiq+MGTszxNu>*8fC
zY4oZdh*50mOD+9qZGl+EWxN{(nB1?EbvW3x@^rEpH{TUH9KYLy67vUC!Z6!^yh!~`
z@SPpj0pVp~H^?@|ZNHL&oEWFzc_Lr1=EJwFmc3KWVs9haoT@?Pf|&u<^A0mU=v$(B
z?3}m|MB?9D<;vcL9xZsAhg(Te5HSa{IV6qqA|=(^^vbKQx`mb8=xgScOwX=g*J%AT
zonYzWcNvYfS>*ms`RQ~KWqnxTW>(}VzB5$YRqne9ff(<wGlz`gdC_B(nhBvJ?MEXy
z^ZEL$;18buOGjO?y#rU__@V40n8xnUJ3h=0ZnrEl3<T(hkb7Sey*?~W5*~fZK2|yZ
zPZEhUMFhY^j4z7E!>#AxH%O6#SIX?O({u6qAlQ%ZY=!E>SuSN*23Ju+&VPB{0H;HZ
za;pX|+-+%}2vS{Y7n5aI6{J1io+__?h6+>%tj4-JA_DY)0@YpWI$|17QS3V0%&@CO
z&>!4U%?wl0t)E`SyTJYTxVrnD|A#KJ7j|MvH$qxUxNYZ_(R!Kh<!Kg0s|CZC_-AlQ
zrHj`V-P<(|DsX}N<hcIIEkgahAh*b%^xJ~L7JY30Ny<qQt%8@EZc#j5rAhDMncOP-
zqUHl29ivIwwFh?R`k2jpRAWOV{;0O=^)ST#88SRqcu#BxO_oBlu7te-z0PhT8I<+-
z?O(+oa$(QF`!{eZ^(NzE^hZ&BxM9!r3CZQRHlSEO8)~%<9p{<Km#puu-vj#r+Qj7}
zvQT>eTU>PDtAvPDuKFB(dQ3WYaU65)5h$rj4jQjW8LM0lkysgO!$Uz+c>lxsk;kT7
zEd0Vj9fr}n!P0BjATPCm_gXb7krw?{%gV?U{KG@4klo~+_9L1>`3*nZv6mM&tf39B
ztV~wIhKNgpM0x-q#T9gQ=0g{&6(b%pv>Np-h3#;SIP!g>D2a5bTbK|vZ{KcHjz1x_
z-Olc6VHjgMoig;%R?s-=4pcwMWmu)_sSUNsMmdm+E;~F1UP+Ao7S@BeJ=d4Jl#ogT
z!PZ#1xhP>xA?y|UZ-S4~f#2fo29Rp7_-u$;%V!*~4fPcMWz>UI?JyBU0tBo_Y@180
zW&$jf7|n9ROu!`%fB))~v~`ujWmU1Fx#FU*GOb(q&sB=KN#@IqQ65Shp{E&CH_zn$
z`-DMV+;WD&^Y8UQ^{L38Fe5HsIHzIPyMN`aCI+=*xK1%0$J(v~f0#T%V2jPMC`Huo
zj&l~oL*CqYv2=b`cKxIb=v1|os3^$MuQjbuZpz$_P)QKe-+{~{YeJ6p%<TRikK=|S
zw-%}0kvBOrSG2SJlP5cU7d-q|wVlICb}L!Ayd<dq&`{-VV3~t8Hh~}aWGQ~%);5tU
zD)Sxk?rL%1E7oc0rfO3OxGBYc8LWw3mimjs*I_*cylfbU{ypLEgik}k;;Vf&SR6k|
zC3#Lvr64@Betr=GmdY8SZ@5#~cibIO*(o^Gjv#O279Xg4cL5CCaV+^@>}Z=d=;2?7
zoU16<_Knvo<p;RHFs>US&N9tKX96{Nc<<rEOQl<rEQxF-Jp=vEHI|a??cpD^GVszh
zwjEsqX>_r(LDhz5L9iwDojbXwhL-S{KQ!#pp%Uh9lQSjVcmnE;p&L<aDQyCtV@mc+
zOoPpxo+4F*9d`)o3%ZVl&$MqQfFw2`4Bt$rZNyW3^BMrQYxs8FC{pTncKCHV3`nF-
zH~r4nkWBjQa580p0k5h3q7aaOc8%9j*rhqxD2kv|>%ncW?TCW3PpW&)mK~-_PRk3i
z)R@-GYf?QBuMtZ{UxlR#|6nnp?`R|NfjT|E0#cTj#BO2_E*1jIdWePa=*pc%s-zd(
zsrLkfiCEWcr8o>6P2Z~LdP@zh-T6KQ_#m7R*dy~PA~LMyQ;~IKmJi0o757PmS%ubH
z*+4KMd)~lDNb^q$X<kI%6%R<$q)7J_W=^ix-7^e`;Z-_!!}uP+kw8*Y`kh~{ugEcA
zAi4sJBgt$0jYEzQEgtpJ?8ykki$pcy4sT;9eqm4Efhp+vS2Y2^vp^d8;MV~ichJku
z)n4@xfea+Y$Ny7CV0arR?gpmgxA$p}Z}-FH#-c(7$HqUEA;sv_dAH7VKvTgAgEH=4
zFsSqg9qkI}gj{1EZaE)DHeA0;{U673wN`R8*Y|0DE7tWUsMHVN9gt{&*{=Q4nP4Q;
zx0u+YAGgdS(8w=YQ50;uHw^|!w_Up}rrslHJQ8o3EQ}hgg^b(kSNR0|?>P&_d8QYG
z+o+Gq@(#6?Xo{ctq5yQaN~eW2nj0x7j(8u>I_F_0@KV6V3Uc}M%^?=|$hQiJeQk;l
z$p#Z5p~=OJjLW>Ewr8s$u!P-E*tC(g1L))5H2v-TX&k}!Var04o?e@9JT|(L=m-7G
zz7@WP&z`sx>F@|3#?&H9_V+UK`0!W<2{QeXdpd|}lt%C4;VZZ)7d8v&zz45itaBb#
zbMKHSN5lt?F0+%8E)4(dB%{-o#Za(m(M|aW1Se(<kmb~Bio7zo`y|PjI`lw`D=-wu
zszWYM_?rVGC^So|i2z^P{oy;EeeQ1P<kl?J#`G2;MO28)bJVgJ7J5xZ7bR`JprvZe
zfVhYC{Ca|y?k;QjFWxKM*VKm8nMHw3G=zAPm)}Pi6TIGQ+{aBUJifuxy+p##;|-26
z;EdBEF_P)XC910q<IIn5)K=|(AJO^|A}YiA`zJ@xJFU!(wVzE7<`G+>#m%c<YHv@&
z7R`3G|8yvX0JuO$D4mIjbVB)FGHI*zu79#%Vf>Zi2jTG|(CUXb&+|MJ&lEJoz>m~f
zo{SxM&`d|7&3BTD&SfWtdcJQWaM%mQ!?Sw~c4g`wjBvkP{9cqjP_=^&1}-_fCAFt0
zxW?+)UF5`(6^^%vtmXjAT!PXU^OTmE|5Pj=idSEK@`$TVqzbdLj>}*Nq3v#njL5YU
zZ-7ji+{Wh}&hA|XRTi$x_ye=pLE&n;2+o%tu#H#fLxLQ|``o_6bZ<XuKsNeOJ&tn$
z0al6u78JY7)^P9%QF(`NYpX9SEP|=O=6`}R!dH9}<Yd8voH}y*(Im0#B$ooOLrY3y
z&|V7MQp&n66FheyqrssHF+_J9JK~pjQ=NhBQO{ys(}2w-qqdi2R9jV^(XtADC5N}D
ztYtaJtQwp%c=1C}e*tXP2MNnKP*JR#UK6b8TiyNLo~R%6om9LobtibIwO`ck!&lS^
zQaQeOEKX4IYtCts_Xmd$*2+BX@;jD}NqKBZ??i_+{C|LSn)lqCD2H0RbbZcy=TSc{
z_0W-hzw>%~?w9w^>9&f!_xiGRDR{N+A9tp#pWru0REsLP*2q34{QzG#20+@8bcYS=
zbhwonA&)}5fHn$wRpMLz``^<+X~7$Uz!Jg-T=8z+GtgFzh0?viBwrlkTYI|>d;|~E
zB`ZqNdDL%y<iAD}G$DoG<G6WPvc?>I@-`+7o>y!1sUmDqI`4`(S>1l;q;zVRZINe?
z&*=qdd59Hf!Jb+fi&%+(Q&yTT*al0iDMr4vu!CHZZN|xX3=MNa933pE-ITjGm=SwW
zjdhhKot7KU?<iYJ`;^*nTcA>e0@3UUrU_ZMZtt9jWlGo{FY{&5{<&?LrbXC|SkMT_
zu*?of7WR{EvtrmcVb9T+?xcq2QkPu4#iOY09j27o93sPqE->2zi5)hQne+iA$hOS#
zxUU>p1)R#FIk+zw2*C#pR@FZgEE2F=;f;POSR3KllpOv^5HS<#@P3t(Nq^fPOy^i)
z#?Z~{oHWGtO3-`M`xP?e+G>kZ1D^7UE*^HNl9>7{Fl<eH!r(6xR8Rw|#j1!VnuP!C
z-?eGdVE9apXLOU4hyL-DCZw^gk06|2!jdHV-{A(=k(u0xQ$*&o;0O|;e>)J?P^_Gp
zU2n3um^%3%V#le988Xj)9+%9R0Ve6i{Q1`?{EMO0amQM$s|G75GIhmWlf_Kac^dnv
zF&o&X-(N>%pIz3Wx>gC^MT`b%4eV8@F-Oy(`490d=pWODsngzn&0icj{V+byo7%KH
z-rug`EgfXAB-Mn8XYhtN>!&@n`@DUV9)gqt|13Tw%+<?b?6*QQxB8#%K|gCwTUu}!
zAf3B;^qa4Tt=0W!`M{L+a(X*&h%)qt6lJ-pe>0bhVWheFS@E=-gE57o2=ZUV87xxr
z;m@o^#g$y<1qss*p_N0ZSkPAvv1Y&i{4ofTDrGSP1FLMq&jG@Fq=d$i#@FP=IGJ*&
z!M#47(3}%gSGsN5x6WBom=(Xuh-BW8%kqDPXv^=9UO^a-<&R1H@+xzXA_CjsgbNYF
zDzAUMzj$7Q^|Rdb$E3W0*h}z9lnNsLe&S8gPpLiSwQsO%XIO#mJ!U-IW}`}9{<()%
zCEE?09=Pq-=bNXRy)8hYA#XX>*&=yK9QEzQ^!cm`7f`v$n@T)5_G?LlBhjk(Po&@g
zUm@;I(~rpiJ*2;>!l%E)oO=!8e4-G*DnbQ*esvnvQF<OfCmJR6ZM>Ob4C1N1hUg@L
zUIOXH$2GYu&b5~N5d<0SJMZe=a@8HP?8g>HNtv>=Jg>!Dc~;ew`U^aryL}ZJc!-g*
z9F9qyLv~d(^3h!EUz{}Qb$WB+y9vkPn2Sj2t>8F#^&bUDcd|$Ys~b3c<$xHk0i}?#
za`}3>&)4Bo4mnb5gCW|><k@|#QB=CB5BoCuZw{8&9(cmn*aN3Ouwoma{a&nkhrZ2{
ztVu#pN@B^cen0k-a21N5`3d5Rg$u@H3e}YL7bD=3zHapp0wT|fkdRW!zIxB@4xRHt
zE%K97uHO6cPek1zo|OLBmorH0giKKoAjq)oi7C2|<|rzS3>OF7lXI8{KGTo0wUB%S
z<?y2gq!{>3bPl}s>Qs!lMjuK6zUtr&q>;qFF{B?ikm*cA=O7L9t*Z@sb#c$-y4l{E
z0Z5YVY|D{w1ckhl##&WLY`PH<m1hF@A)#^?8%UzT3i5vBR!khq82a366=7F&;y5o1
zA_~c@^nYyQKjJ0hp-025LTn&YSG6tiKM8^S<_QyiGQE?N*0n|iB>zLb0oD_hUrcGq
zX|>VctRTo)Gr>+TI-Xy0IfAa+Qi4du)|P!WtRJ1f<Wc;`gJ-XC=z?rEd=fJDUaOgd
zl$<ulTt+v;_?*S%YJ@a%7AQT|6!{&_FvPo5c*m*t>5YKM(wJQ{Rbf^%6Jd^F(#~qN
zWQ0Ij4|G-YTn38|sEDuJNufQP&{QOD5ih~xY0Ti?$4bd4d{u+gf0CYraD!3qNBS7)
zp#Z6vACnF#>{?LG`<<jjv?#c*%w-Gvil&^hOztLf82t<Ik`pxyV8|PlZQGe!DZWxD
z8a+Nq|BAA(gVHlXSG9si6g*y)2w<j{CAEJx+g06+#@9dT?f#{!H;8%d_w8bSg+3m&
zOLrBM*3qtV2nv6P6MosL%gs$lj=H6XU4}v0TZT?cytN$|17+Mmm*uJ#IsM`<hyRb{
ztBIDcXc$x9{tWajb_>T<xWD03fr3`220Q)$e;Gzq+_i>5h?k_RShoZ`>=Gj+gXTKp
zNcf5<Q|84yP+%X1Y8UjAvt;c+<^wZl((T0VV&utDYs{)<wm;Fl89E(vwHQF%$IFE)
z`QtKPRb~|!Rp$_#B!5t(c*T5fwd-Pp*NK)806FN`_=`loi#n=0lW+80T$p3cH!b3{
zwD?Ktl0E03m9GwQC9&=Z&U{Q@2qiJWSAT3eIs3|o7>z5YN8S7TXrwt5p}#DlNe(Kp
z=YYy;Sb^sUY{PjPjEEun3S=5ksiuZBIk%}670ZC8XpKT6O>Hxi5UK)VU>n2?>@{*e
z`5gs1rVt7=e2L8U(GM^Z|AZK+n;4&HS|IlN%(ns#<?JG-PG?3)OS6UZJxx7+!x-|R
zWFaTWa0XQ{N=TO|jkqn^rx;qeh6?fEFb|oUCERcKQk+TiTZY>Rvo10l7xL$}H#OCp
z6ak4udkM6-7~humm`d@yA9e#XWKpW3zmdyc<&HSfQ;vtQN=B+2lb%xTM{Hv~aV2mo
zy#B#n(-azutN;-~;La$9HP@l9IOU?}t+A*E6?Z7|1uvvzj|2^a8h_$~zr)-mbn!58
z-Hr3OBa++jxO(Y*nj9se$4J7T)45{{Z$>L{Bk$`UG>kMF)mBQ_cj2t}f|CJ47>Sjv
zAXdgoE3@}=9TaB)C9@e+3+Q$<Y78nX>D&VyizCEU0x5JfT>1sGOU&smjCziYbve#?
zE6s)h)DZ%Vrv(f?6n}pTMtJwTEgBs}{#yC<IqPPkPlsdeBcL)(ek`BU^hUI*WWW}4
zc1=Pbcitefi=y>k<87NkPo!v`tlH8|dDvts#o^H)kf?gg`EymbpY#V2=d`@&yUgK<
zROGK798N(_XSsfD)j}?UKuDUOk-JQ~@CZeZw$%891fZxkdM{`(s3TR;sAsSL@IFkK
zhqLPQuOqFV{kUKwkw$pqWdb;n&pIQ9j%3o{=`qXUnoEc+cYjU6sG+Sm36l0_AF&^H
zN*R${dF7f`v}w0j-kQmzR`mXGiRGdWJPMAXqZC5fVJEca{%(dTL?0gBK5|N?=S;%A
z=dEc)djhHx*BBffAhQ?e-jOu#o`%pA4|=gJwHo62OMo#!shDv3K%8sIpt@<KRpexM
z;x0qUrAT_&V<`;3p-so&qGHt`UTTT>1cn_YjIEhRktK0h#Z&5J;6~)Dwu?7=nG4$Q
zL~1ib2NIl$05*ky%!r`ZFt<;RkS!$C{C3_v2%*@3+x&p^2vxX=soH2NyMhy{8iSvl
z?PVtX18dHh8hYKn23abD4373YD{7acVcGods2EyYH+noxP=B(P8zF-qNmN%-X+Lk?
z@88H2ymtG8!d(nz!-tMYNFr{O%yg(mrg?#v#isoYGNc8_OfSuo(Wd=N8{8)&25+63
z6`26#`PUD<t3;H1h1?uY$5Ef-Smi0rJ}kpq)Ysi3-q@}zBM1I}FMz+;q|)<)DP(T5
z%9GP~`mac!^ADMv6ktHdRY9A1;KXXid}L1v8)d)_jsDt+qOzfARyj}cjBzF?3OY5i
z$Z5swtuQ(&wfcQ-3-xTUPlHm?nI$jzc(0%J)99~nF06mc&YKyPhd-(tgJL|_Fqgzw
zZvls0kd4G)0Q$s@Gm!|rE|xd2l+^jO^MQtO#2v+7n}$m9Wtf8R^Dw@|2UUWYHhx?V
zh)iz`H;MH1c^j88YfOIg;S5Ou?I%OvF#cjG`J$1igGpJs;_gq-UpGr4i_&X4!FY4%
z!`aZ69V%}a>#15DW+0)f5?=<%oPfb#Bbc{NbI!+>GM^1rERp&9A0vHEl>|?=D_F^v
zPV@5Zq(Q$lke;Wjs!=wNeMnxX?3dH)SzmAi@o4`KA<%)K3*4E9HtJPo;b;oukSA6?
z9uDePmIYZ9Rjh~*T2cMaOR7jrT(P#t0T;aLG6!Frcy#wfPN{ppXc6CumoVAPQgP^^
z)_|NMRo1xV;*q%H1aoJXE@J@<(yT@Vwoz@(@n6&7m=|vu@z6s=rczPdX%7y2KL8?(
zCAxt`eu9$b?`n-`2^G7C5=BO&gBIZa#S6+d)MKaQRBF5t6gy?+X`C~hNDf$CEi=`8
zg+hD8p$`8^MpM4saZcbJ)0n>K2+67K1VFjepbPWWI1$xWQ`u3enl$P~7Wyd0vS~{^
zcFm*{be(5!0M01%PF^DrPKl>^_Q8N!<fZuwa(fJmMIR+e{Rpqpt=J>YLCt}GZFw%*
z%?Ub)OW!LjjMc4qeLwqd?5Jenw+gG}xV)aZ1_srqai#3XTJtx>E;F?m@nqb8xw_8v
z>|;g}&Ak@iOgF>|@H68r%-N4Eml`LVB0Ry5YQ--;=9oBld)Ks;je$Ms0c>6qT*69D
z+-CNToFGfJx=1|rRLwdZ&^+tz$q?|E4=JnHG9;@t%R+aiOH*K?x`mDMq;Eoo;hKe1
z!j4O+65)t;6Yqg6?Xqu}x}q@*NPm0AoR@`;41~Lf12HQECF*HCW*4^DsegXff^&XX
zA%`nAY==L)<ovYg6IJ|{ik+uEzutOBb0sfv%JD5s8W~UDN;6~;v+Oq}eoQ`V10#;T
z$l2O`hsD6h9}FRt-w0jjePif`1Zh<$#uvB<bLy1inDFreay*nEa`~{jG_kTBWNg0=
zosZ{x6K<^hj|B85zXngQt3Ob8b0@w#w}xtgx%fi2)6*r8T>t4%^wI>=SXNUbtHi0B
zj~BumsR;tgE(58mC2DBv-OyTpJvUH1Nmdwa^LaePN{Y^oL={pS`P<zFD?&#kIVF|i
zzJRiR1{2U$$V|50f=)kn^a$};xP}58a}Wy?M3M(rr^_Q(x_gPZS?wUa)Ph?FZaZha
zKhPFs<L8`R6#r}Iozn(-4FZ>#ve6^laOMaTJ8J0AN3V4skfm8{Z<c7z9VdRxaRJ|7
za+=unig(TVAhlo<%rAJK8X0ln{P$F@(%F3RB8PQ|uhJK0q7Pk&9r5~e8=7ByX~hP5
z5Uq7if~~c5A+o$(E)3zb2^369PRvA#&4%z4E1y)%H4(29pdW7p+e`@iZAl>epYNW|
z$oW$d&^z3T@6YlF2lIbAT!ft<z~uXC)2Oz_0p&Ac7h*OT+b3K-&Ub#wZ~Q|F*hu}d
z8iz^5#0Gs8-wc`r_a`^{#KcpK8P#C8%<fJk_otUo0UYFhgfYd4TGrnnqvfzN%au8P
zLw7?i1mkX#7$WV<qey-&w!w3aAjQ|CA4lK9IuE&_I(t368-=b~hf3KZGDBpCMY^9q
z<h?~?K_LDSf^O|TV)IggCE(_iv6BwC7BDBA?v&cM8ts!6uP>$NSHlZy@q$Q*ktoE_
zK-oxhmLf#l9_BI7tEDF2gIc>ST#M@;6%d5XCZ5~)t_rcGJH9&dEi}YB*C?<C$yLqY
zd!%@?+KL1#vUHPBRputl^E>DdY#kt__C5F2F$;e5Hzw(r)oFSOw#?7KdWH@T#(ajO
zSd^#b;Db_VSCg`;F*;J(2Qh5><g~zD>0=go!<YFm#f5F0{{UPted|Qk<hZ8U4`ayO
zO@^LOb{yd#!n6R#>=#3P=G<V<gv<)I=VK))VmX3+s<_JVxfL9wa;*UU<#!xVQ*ctd
z^lNTwM$+I|^o+-T)N5+Zab?4irNm6TD2C9#0-%fCKR+b&$}r^xElD@!Sdhh8hW7t@
z{Cw{8QgCogYX*uD(C}j!#>7dsg8_qNJ`6L6PxutK<nwlw@eOP6{hW7-lm=lT-7~_v
zSB_vbG#1wQ`|6ZOVgo!Cg;eV9cM>JzV0E<vF%G(v+N8Tl$V-sk3Vp3WS;|96Julls
zG6vyrZ?QM1m=$?IIdiba$;UsR{SY*CHM%sRDHTMNL|#*;8Url~+njfU-Fsl`R_$)T
zXG${Z5krPZ)EDbL<b;l;nBpZ{5{Ih0qdIr#QY28%@aWu<NmKs(_;VQwZ$_7(uT?;k
z&yC5)W70C;0Paj&RpsdjN0!bvJh*rliSLi!j8if)3zUT&e)^xLAHB>{!sZULhDc}0
zIte~u-{HUEQ*Xy57_La`*NZC%V145E`t9x<*?`;hhD$>tc2#r1MvrupY|bndpm$W6
z)iUggf~%w^JyYo=APIa1uaha$*qRpM`+9mA(Se06Wm1nVNW5@@E3X<WI9pR|d3!bf
znOPZYba|ExhpN=BYw&VKB=Z%z4A00Rl7#)oi#GLDc;0tku;{Vkf(l{%5PL4f=O4k%
z)qhEi;k+s%bIZ;Hp8@2M1pftDfThM)+fW*D=178Yu|7V_1bG(c`>KZ<{+ddM15>y*
zUXq69_GmMGw#zSXI;!v7@@HT#G#*_~v7$c0(lo;PIVO42USULVMn@v<hum)lU?n9~
zE_9!RH#5ysYJYe&E#F%Y@5N4;K(#=cbA7s89;o!37Oe`z(ohBe0@%LtD5L*XZv8hx
z3NYP1T1aW&9L!HeuJZTgx<Lnw5WWmx!j60eB_t^Gn<Kw=14O&Vu&LD*VO-6jMo;7`
zAiZ_>yT*ayBQJSF22IcpuPl$TGqAN9<%ScLcdqc_M8F|e4Zh$Zzwms>veqa7_Zop<
z&H8=J3mgGO&4e-0?j^WsnDxm8V)hY&)UDT=9#YQ9*E}S2b%%3l8Bf$%s#zBORthUi
z2q4vw6@h>MKSY&nZPvw}ZLaI47-&-~f7sqfcl(C~*L#;n@#$0;zuVz2>85KypKu6q
z*UiNF^Y8oU_aatBUV3Y+OVWAdwzkFPXM6E_ZyX}e^%Q)GhZei-K8220?R+l8HI}q!
zkVD*cZS=H*M%cW<`|VjFX6#MiKEKr7uA<EoZXSepS~?<=(+8cF+Ajns8m?Am<4jIL
zeVd0?Gr{7YC2iFg*%qgn7tZ$h73H`1Hs&Bw6iRFNFC{IbLhN~qJ~?Vxt=+n`vchoc
zGZ*KDi=1){*VKA`uPE4K1paQq!WZ59;)Gfk8R$YO!sFVwyK&x3h@-^WH((WFoa|es
z>0S+)6(6RwtMR|x=pXv=r3x`L2AnC*&1b%kA*Qo-D1ek{imggmRSD$ln%=rurrZL5
z6T<>TgocSc=&Nw9Pg172fGwHAZzl#wOWy}wB%)mEDulUs@JvQsR9d>dF2Ev8F#V+%
zX0;+7VjEu&%iY`YiaAIkn~Pd2yj%|O^}`@r^7_@+4$GZwgAz<0h%3}cb;x^Um)}r#
z`EtX3n}magOb*Jgz+ZBY3^u36J?Ew?8oZ!2Z6>l#!7IVH-X`Us0~lts>?28q<19Fw
zFd-?V{;A&40$n@13%&->?5WyXu0xhhILBL|^Nne+7C*k3*A*g6e;(cK6PuZIVQ-Yb
zMo@#+r<84#gD!JB&_oGVprRwLRk65@K7Y0G2N)rSYt4th1qx?yZXhOgmr}`8UoK0-
zxppThPSb&c2@&<qV~N>61eH~^*Ld@q8Z6ifJ4IbI`0%q)grtRrXC$O8d_4d7G%s0Q
zM)ptvabC*E-riHYdTylxyjg4!CNF+-7ujq<Zx~NnV2;U<GK-#9(#QHATDcG`hA+n*
zHH}S(y+AR9&k~@hF4%w9hA7|#O!(z#2yUEr33b}rGn3r(NW@~NxMskPFwpK6<t?}T
zQp=&pt<`0+N--PPN0hff8ECnOT!s@k-jIxo=SWIFMh+tawZ7xk;KFD#|Jwve<G{^6
zST4S?wdXbL@69^~5&C)YSag>na%nfnrGetNGL4?gRB0xX?+Lo^hJF~Me-lYj%!-t-
z0c;`%vc?Q_eafOCN=0+q8S8~si1QSm5#PQ0v?|Bp@75*%2i5GHDu*nB{1@2cnHCeX
z{e{n>o*g+q%u}`F`(c)}O4v<*9bn%~rYKhGNyGhXrSfgvNwwsVcOm-#U$ivd(*u2V
z^)=_+|0<RG9o;T<8Tr--w7k`w8^5T#R%+q${e7{8-H~mzb8=WIGaLK=qz^8-_Zp5B
zBS<IoyT<mCJBR0?Y0tpY1QnzAL{YTaQb;Sdz1unzTU!V*BCzdbIRZ#u#_Lawppqb>
z6WSK?rI!U`$(vq4^d!<82>L-T!l*6ywW+Yz?+`21-cS2rhaEZl&83^-mh-Ng0cFr{
zvs`=7I>&SQz(Sa<8F4H>RwO$yj30)+!*-van_85n@LkLWDEa|ZHG(-Ydy=5?^H;dG
zI)pIAzMa58-;WBORq3m-hYJGhsE}ih{F`O~r-p%wZY-IYxb^8_1A-+A_DBb;C9b|$
z_#ED4M|^)UcM5cw!vCp{;#%3fj+j;8a=#H%?{gFxU5M6;Ey@f>)j`=`nrHHe`@JIk
zZ8!9Hc=Pu}`ZuqIzo(U&_yv^w*p8XQ#L+x$Eo@SlvOr0d(#3*)Jf#XFH~vYN#5u#=
z^0@R@Cb56FhOU&?V=GdA=QWu<P|-*%mYZnqg+}qms76Y)Pvy5#X)sb&WPMemT4Hl5
zVFG5T`_YQ>6!^cV<<xmc?JdO_w2Jd&wlL6c9(YRXa-*4l>p}07m$Gl*S{Y_lWZ|<W
zvkFw~t1gVRVo7IHbT0b8#<o3^#!6=CUY=@gs(P~D-WG4O*akjZv{4-dPIgc5>8I{K
zzR8|6F<Mvjjtba;=}JhymrB#{T;D-h<KQBMs0lRAe{Cu+WBv~f8|a6|e7TO<pPNLm
zwyw!}qL=e5wg^4++O=Yb0dw@0s4dM43;o7ul`5pA_yWPelaQq-^BJ$8w{&gH<kZ}+
znO)a`bAR{d=Gh;q?Vcy*1wV=SB^l5m##ghlIM4~#yeIR;SiSt{5Q;A7N9YQJt;TK(
zDC%`A@g;uC7+o1A5s>kZHx{MUH#XnxZmyq(irVIyZbvCtrL%Jo0>dvl4t#j;c2dgV
zhHX;ATPcLbV~}WR5z^J*YfaJe{Eu;*`%!2CZM<et7V}6Aa8ZfCFC4UR*(u#zHI~|;
zgERSDY47<RH^3;dBG5mHB|{FFSv4v}=N`b_+qXj!NjH{sKAo@<yMWXi?%0Ez=V02y
zq;>oXT?L7QmZmEKozMKTUl)l!Pifmq6OK!I^fl~bev|C&Z8hcAK2N5t$ttD&3{T3h
z!>W!hPy_pB%jgUx<Gi3jjkIIn?wwhr^05SAR7>nHdrG^EEvqm4H1@kLx?YHL(kM3n
z`(3HS8M^D*(Q(wBdziCR@8pIn_3@su-;?Qo>ZE;4Q`W*x5O+N7t|<+bhLtQ`EwTO^
z-H&qx-ZGpZTbl*mb=F5nVbe}xb|?A7Qj8l;Sn#0cpwM}<m8Rf_&Fj@K$@>n@&IAE#
zM5C~gNl)r?Si2slW166Y%EHiwgfrWzAkBnb`YDYxVpuJ^JTF4X7L{Tn%PA@qp~=&q
zmwmHYJ{#RE?-(q;HMZAM7W0NUV4D?H^B_hV3;U-x%(Vk|sRdmemAc{k8e^rm1;!Pn
z6PTdN(*7dS+Yk&wLr%W;a%}W~dj*xgD&vwd@ic}4PYdKaXbvRQ%^+d>RrH`EY|`Uf
zo4L?<_U4Pt@jEvHZOJ`l_jbUceGE=NeJFvRTZRXt-v}SQVh+ODWa}pTX4v3)&v6{n
zJ;u|g#E?*Q2b4n-f8sY-#hx+WL`XjkEBhV9Pcxjj`ySrdQUF&WOO)RW?pJOXY5Pq1
z6Wxp){Epf^utkt2<)j#0Lxc<cGKS8IXzkI2^NIS1#q1xN9FamLmBrTp+dBO_|EH}n
z3@4GD(udn>Q3MG$z}Y3^?VBU?|44k?e*{o&6~5de9^>$gf`DUIyBwbS<G#yXQzlFI
zYM`%6snqTmQcW0$e;Q`Dn@hF3aZutux$<r<=RG==_cNjhQ+MKPcSqcbOqf08LksFM
z-o@s3>eq38bij$OsnLuq+9cyKbKMi<PH!}}$xCL!RfNj3d>)`y2xK77x~_FbNuIL-
zaQw8w?BJxCA2kYk6S4ghjJkQX<H7f^)iRen`orAI4s$vNVw-9;Hxyf6-JwKj2ydD~
z=1ff7-3I&-aqQLo=hkj9FkZ&Sa_Jiro&bq5?DzuaAfqO0-JRzeVpXein{60DqlTvU
zvsi^NI1--t*r?RW`$wUq1cu=Nl|RJxP#O2n0{IwU*=;bm^5IQkA84LXYree2CxD<!
zX@4CnyTYO)`tGXR@yhP@55JQ?x!d)87puN<Xkr+E1!Q;;@4AhbYMa)gltRZrOcdjf
z5|ov&8%645m^RdUQ$6+LJ&9r(^Wt&-wq~O8(8Va}Eg|r-c%d$)PCW}~by7Hd@*`F4
z&&?W+Fo$o*?W#Y5X4%99@J@AJg5&(xlbAKy>xqqUfp~h}SHK+AHXLm-Q}=P7(5Ykx
zD*EcL<$FI1M3izZK8$DQ69JXjCblNWHU;X=zODi8A<rckGxdXV9jDua^~d?81_>x)
z7FWc;5n-x}jCKJ}8eqOQH+nMq>x+%b(w8p1=rzD6qwBvn4-D~j27Z)1;}_bqI!sgb
z__d+vc|>&SE^Ahlc!iI`>e)TkIwaD`w(Sck$k?=8Z3v%@{v@V3<7?)><*nzHn`g;O
z$KLW9K}u60^~sdsWLwgB5*Q{~h@#s;-8{;aiWSwed&N028|<K+^}nf;SO>UslE3<w
zLT}APXxZm+U}lLkM+fh9Lh27R%x{=)w3az9?R{-H<UR{qrHhykzWQAKnebIaf8r~5
zRt}LQs}ff8zWvKKn4=R7!ho`&aJW~E38`Q{!aCd!<J&qiL2y3V@oC{KTT%+Qw3Am1
z5e6eE`V3W@HjDz4|FPP3N4zObdli+1f<*ewftiH5|Ngwkvt5aei27UN(!tJyiir;J
zTVeCgaOzu13--olx^!1xueU?Tt%tFU(>sNnlOZefX9jHo<ZIudH3{}*(>~xLye@x{
zn1{12k*9}rQM9^oP`5?Eg0~-q1iSn$y##~iZ}Jos`+3U%XpPG7Ofi(7M_TNBJdsZS
zx)1HD?1@hbw@i7G`3ERdB68POXwNOEojkp0Gs{EK+5WeUp0q!LWrUAEdhkAI%g}sP
z_oZLtr?(bMx68ECF3qS8m*umRH{+p9fRvsBLC|N<ZX=fc>YggD>n^<1g;r)!84EtP
zn0)4YLAEvx=mHwz5HFK&fW7Un)<h&-2z%0_@LvoqU{)Lzmes1@RcDxya^<`UypqG5
zT6X&<qaEbP|FTVMi@YVMp$q?iOnrr4R1fqek|G__LkrR&!q6=tAq~Pv3rI6`3>^|f
zhoneJcQ*(_Hv-ZOFmyKqEZ^Vmes=$anRnlP=bU@)xyv+f(*;5;DXyu$+}Oz!6Ea<L
zT@bKB87g?kF8p}$yVTp^3Q0J~ca}j~Uz*r}5AAI1OLa6DQCffVwJ!VilIY&bycIhh
z;g^Sz^9*3Wrd^XGB4t-Dqt&M)g=;d?HntT_ODn_o;`KTdb-UzovG(CIyiL(gAEk02
z*uQLuCfT3~C-j1u+%9KnB8VazEP*%9t_dEu$(n@vz3_zNd~;=PCXRl~+mjdlBo8yZ
zQ_i<n^w?jhad20#7jak4dn~Q)44JTc9_@Aekr;LI96zb1DPNcTs_wQ<zuNErj<hnM
zorIR(5&jnfc&?Q;689g#k@#Um=~vMM;Zar9(5G^zSdGqz4k^50n=Zl&<yLxLOKARW
z@48K_s>w@19<@V@YwlkirsE&AXR326?JrD=SNB&MQ`9CW+A&FVys8`;R`acxEZO$`
z$RSj<BRqHWnKkGXhJ?*_CgoZhneK3~$e+Hmi^?%A($fPNx(~-p_7r?X6!sB_F4cjb
z2}jrLn<#Px+luP1Yts+j=QYnT<{<JYpm<#Fe{3i}NjzCA4jeU-XjP^YD_?V1r0D~v
zcVtSHF&M{Hxx4ihM<N*cIbsBkdwie!?3b)5m-ECM!ti#h^_r^PTL0%q$O*X#f7hoT
zfxS-vE{vACjoVW3VwR`)_D(7~T$!2D=hF!Vz6oU{r$btHfvIfrD;|s#_&}1ycNHYm
z?3I7fOU=x-kHAa3P-5+2tDD46RFrB`Gd4ZM>eCk?$^qQm(g#^+;N1;NWp1DFmt8u=
z1?p#OC#M3QC*MAE5158qgE#1`81o0Gf7y6?4!v@^hFfpTKHwS?(!$?Fq7c#HUeSJ_
z+-(0HOM#wKZ5G#LACox8(~!K<vXdV5;N5xY@htZ!+^ev?^|{je`Y<(5aAey3toqDs
z6x}B5G#)z|w%0;_*eH}jy-B=#cBPl!7(Az*um)dnxK01)*OT+0a-UhXFWt}ER*>Aa
z1>mxbd_FqEQ@`Dx)zuql&wBR`hss{|VH0L@f5^kNa<PzscJj9}8Z({bQZ3gjLi#?M
z{qoy{Kn9}Ugl$6aIb#&OIP`jQ0|N(ZSS3`SW0P<(^t8JiAV+hb!*#gd6}?FPynJ2%
zyk#c+0iuwcN)wf1rRbA4IE#p6v5Wfqg|xhRo9Y+EP@}*PcE}z_lyjH1wsCr4=PGs^
zn{C>}^ZLxZyt_dLTpxoKw;a!sgNTw~`p;_sM?87>cT7$1GGQlY^=*vVU3sSUB(Ibo
z5UNzh%t!4Fl(y$XqB>Gl)(JJao^^I@F5PULzPs4b74>fk{jig%zG%=#GMwC+=~8u*
zU%z?#hM7jPRWEEd^*piY%@%dl-MPaH6J~qokoAj1fgndLHBZgln$Ks;&Oun(sAB4O
z(X+hyT3T>dnoopLbrI$kr)nJ{b{ZynFhESmI&mAl*LN&z4*9`tPds7##x0URVbnBX
z!%1JbRo)R#kLcOyY<bxzt|hXEYkGF8raK_AnC7>nWicOf^`k=q<I#!E?=n&HkLNUw
z6Ji#zlE2|MLtwYUf6{@a|Hn~5$&|^&(|+^!L!bJx-`4La5e`Y!i`6v?=5moS-n~$9
z-UogN@{*W^AaTxN#F4x2|E?-*H@)v$LdV(>buzI3HyUrr^dlNPP^LTlSqt%(DzAy|
zXQc@=3TLs!Tg9Ix9qm8s7;nceq!Skpj^9f$$<n7jjcofbDR#fI6ZqK{(tK~PJ7ibn
z$9~$~Y@yJOSMrQxNx>hfK8-wIjK1=GjM7_$zyOnHul46|h@4K^Y)4kIR@$s%LUEYN
z{A<Xf5}yYVX6qW&)}JZcJ^XB-2hyx3mcuDzI_^{C8zY|w1Yx1>CbR<M29{}hMT)X=
zh#wF5Mgu)MLLQi-R?DuqXmm)q@osU^%qYN6#$`6SfxU*Vihwh>(P!M4ez9eRvMQ;`
zpHVz?4w2EGNK|aoGwo1(^;_KnY-fY@+pyyGyF=`kiJ|n)A|Iq<f#Ku19um;)H0@K@
zL7WbJWwaO@Mf7jz1U;=l1#<<WyoEqLZQ5fak)z(VTjeR#4t!dIa>y0Fqs%^Fg^gpl
zbUf}WrwUytp{2Ds@FI<Pphv}Kb82HQHMA=Mdj6u`M=y3)ogAyDI+vr0AoiJ*5RjT<
zf7a_9lRtSU{5K-cL}f+)IZ{FH`BTcj4>=i|6<m7W7t==O@4v_!s>4zPTY9ek<!Kz<
zu!@7^(^b5#2f&r6?~@r@?g#x1<_m{&IRQ^O?LvpuMmEIfYHnQZI&Q0?RD*;_(-q@i
zJdY;WK`&xWDF{s-d}X_=x7*ttb!N9b8Tkh+vNt<u&PZ!p@drod!Q&<@In<dc(v%qq
ztG<OOFoH3fe;?2C4>MG?VipS<?mus(e;(JGtQFq#?w9jhjO{onsyR=;o%$-q3L8&9
zH)C561kWr=J~cj7a3~pf_=zKtw_xIXS3uQrM|-mQssq$BI|`1~H~+nBj1bRk_dXn}
zStrqO>9I7bCN1|E<!Nv4v(rdGY;3H^vq@vD0PDnWYjJ&%D{i>AP%xs-$iTeY2<yu9
z`jaRE@M{!xGX|fruTbziKQ5R;F-x!L;{xM`{u(tiZss7tUgH;l+PRu_2>R5LO)m6R
z;B@}WoHk78rxYCjhSFp#a|%(RpYoq5fB9_0Et@W%ZulBudYFLP;<6+XexFvoXskI`
zi1_$xT8=gtFK8I;gnUCv7S{v9hTB8<9EWCrGRqRz4{?5Jp{0OF|COhdX#(-XcSc}t
zff9>0yZ?1I&{yCkUGnTHGqFJu_cdmLSkK|wu*460A&|l2^4TE}EXx`1XsGX-vp8md
z0e-Ni9y_|p_2clMK^)YlwrA>GHO-^e4Z98^%WYuloN$_JRX3{`zvRWAN)FFyqvr-D
z-}%s{7Sop9k!v^c-Q_W{I`!KCNv&@@vU>;6{3HdvCQ_d3wxk$vTgHYp)7(4n(T^?L
z4IDhq@7)!Jse-h&yuYYl5VgV)ZD(~#)b2<l-+On@a|bkq_GWl^*xsrH^1V{<zxqj^
zhrP@7(fv7v2BV?VS~;vB$c~Z;>diV_p*i@Rb_o?U9Muc9<U1Ree2hE$CL6$G+~<5-
z+6i|VL4lN`<;|;&uS<-{ek0potx>5c(NYsT!7CW%=*Ug-K=L}7njFLWg;yW|hZkob
z3YC?m0H`%P$77ftE^#~8o2jFfBd2+->H_9vW39}5(Hr5uf1|3RI^k%lj^?j?0~tEv
z<W21~iwYEmUg+fkhDgV7GT-w~%6#30T?bkA@htsJ<kv2WAfC-ZLFWpZ&@q@4io?ft
zi3fHGbR_(ycaWr=A;R?hxi45yJ_D3yEq3}`d_*{owOUW|of0LOW`g7g9}a4QP=%y+
z>=|gy=);JNRDXTwr-R5)kj0xZ%P$#ipv~{Ya^yGfdnTa9!;L3*Ax*J6!HqieE!E#)
zu&sGp3Be&i;6IDt<<sedo9?kTm4tuE4X%obh}*d&R@teIGhTH67@(@i`+_VPa>vx-
zD2SN0vG6uKwcpkb+Ug8<tLk5%;CxoIpNm&(Rg&}&o#@*85G$Y0W2dzs95o){2IEz`
z+>)Y3kfyh5M-|2obG(L$)Ve{>J&EiNdk^HFf<5H1q~G1X6WC|&wyE_EXTtO59YFHs
zNh<IS<OxE<71goM!3ACnuzv*`Z=s%v>ZS91>Vf8D!mW1DdB>OmAp-;?N{!*h=1ui0
zKCBD6U(9ZeSK4OsP{p|s&ms50h2O)cX?+eEvBX+>Vv>tE9tpknoc1<aho;cU5e&Wp
zA%WhX3l*<JcTfzBDBdb=*g<oO?Iu1HakYQ)45U5_>#l52<xVqdeq-HsqseZM;uoI(
zn)elUmHfB1ZAV-k)KKzsP8q2sHfaN12`IuvK2w0rJFGzMstGfp=ImG?%l2~5m(*bE
z%EelM@lH)(R@Z{dj>TfBS~e3?b#siyf{e`xL+U-Aul%G8x7=puUI4l5pYH&hy?o7!
zHQ&T#(VOlW{OUQ`*|z{!ZnViEoR+}Y6?ByE>_Lg8Lm=ks7U}mleHT7NiUB&Y%ntjp
z$=(eC{PS^f!z7K4&dJnVJnv7Gt=H6-6mXd}q`^Np;@26}^1Sn=se%OG<6W9Z#B`mu
z?9WxS^t<mxPLo%RKHBesZt558jh_efeZ*fQ#R5O)oXui2B)PVTI;Q&N@`CEVC025^
zt4@g^TW$osLDLy~(8YYm)~V!c3ZU(^#dUpM5qQx$Iuj8s+J6Lcb{paNQN|Nt^A#ov
z2@|!s-~6a&a6eHN9JBc4W^ai;+StWqb7Rk%B&WevwF)$SH#CuaL;)24YxJ_L;Be12
zPx#CxTKs-UGHi}wNnh<Q0_={gsxe2zMpx_8Ue8U#*S8rnlajFvK%A>P&2fp-Usl^I
z>xbDczPTbNx*r~TvG`czJ&g3HPT#3bc~u*CkRtwMG-4!X!iy9~mg>Cf-o4F;-jnPi
z5C7-m{Cn_JiPi;5r*7LKs4A9zGWLdx%lx#9N7a@JS4{C2q^#Z4w0P|67|bx6cdQJJ
zX9JWiK#+_dsi=tc<<DPXNPxNxQDgWzSyy%gM!vG_5=IRLCqBkwd^~^gnJ;T>nz<J}
z#{-xC8n(A~rU)1Fhe?3+&Z`>rd0dgrW)a?q7rOD$Lg=!*9Z=&x?Cl$D69|JW=D>W-
zSr@Atn>&)Id>W>=ysyY!_Gm3iy>3W44sGK_$=qXXDam4_d7rS6*1qH67oH><iP{wS
zs%*~+_YK>;3ite6XLo)L247tTIxO~=uyjCDnw--xH=RQ>fu2KV1xvpq>cE)Dq_X+6
zH&w8E`?Y&Mhc9`P{wVi}R&SHq%HNoR)uKwb6?3nv?i}UuBm~(%w{J4P<BM++#i!NJ
zK|wT?41%bz{R(3()BPb0wJ=f-oN;@1Q9W2J!Wi_peFnW#ZJaF1B(Nt5+t}L77OhZk
z6>H?=^Bac)9fs};TWWPPe)=aaY5lW$_1UllOKZnaAx~X&OP~&Suj^V#jXTc$BW8$5
z=h~Nz`#N*AcOJ{<Z>^FK`?qK$4Ts>(iw}p}b{>PfOk)eD)62x;XlLiW59@u`Ut!~H
zoErPj&P+J9WGCcdk0G^PdW2BoBC5gjjn+^FS_7LqGOddx*u8g(q*mXYf5O)dhrlId
zJlR(=7x!<0ErQnj`pA_Bf6gU;=J3$a;D+7i$_<%R<Juf`CfhZ+bl#oaG7(`qlOF95
zi44UTr1-#72x?bQs|7cETMK(+`YT>u9WLfM%8u^ed>=yn70xf1D&D5X#$-Cz$VvT}
zblXt(ID0mEm6Q?UwNPYKAU*vJ^HBIR+vDTXGU~im3T8cC?$KR5_GY9A`=8=9Sk3LJ
z@jqI0g>X?`#c6NYL8)~(^DCJ%xEj$@1Mc{yACef=wfuoNTuvVu@|h+RI+cvR2Eq)l
z6-b0#P5^J$7lz59w3*cNT{IFYDN`IFE|~wkWtC6`w*GEaESV?%)S@!pS$7Y~L>rtv
z(3)yb>L3sX^xqc03=G>+x6zw=siqqnf}<+9%4YH_29D(uM?hOwo?5uFLRNEQB<^;j
z1wv&0zC}e&wb(3>7TVScHKYx5`?RinnF^a%Toj~2!M5enI!uoZwyrfayaOJIx@Tc*
z+jZ`H;>XNMXvWQ5`O$e>l`hM^x9ztrp|(9LP2Y6*>#4q+fbll?!ZzAC=l4sA!eqk@
z*i6iQxPk#|O=_zQdCQyWgBljHi2IE9jiB%I3$W*M;GBLJWzl|pI&rMHCrGl?cH?(*
z3l9%^m|lJSi`h}6ZWrj-c}!4ge9{~NPgw;>Ivzi!rR@V<3}E9=kN%IV>XLoy)0YuK
zWHs}HE?~cMQSPC^Y_20QF$o$7^2V#Fv(pj42X#YM!O;vV;^ig14}Y98uTFc3h>3~O
zKpsgAHrd79S@(0tM@M5s$bi4N+?S++hmiDutIB)hjkhEKrWR7nl_|*W+4hO@rZP@A
zE?|T^o_XT#5$;1ejkwxydV_`q74MzQcl<=o^m34ZGG9z0vWyWyP%X=83m_jg-ubRt
zDVNIc-yKGKCwrVaH8$Ng;V2Z(+T2CISL%z7vot)+af_L7THAPMKqE-)|Md7E<@Jj`
z-u(^Yt~37l&F!QYpZyv|G;Dpp@wj}e1%l$AYq__5_o0&&MV)%bi;`rwx{@YlKnSbK
z5d@Hd2~`^R{n9zA-fB-`y|}26T?nHSROrt^JqJ-ETU-vdbkAaHer|PH=m@i6X|fg*
z%r}bvd@A@0?kM8@@Z|mlPQM|>S;pig8!^l?{s)yVr`g}12kpdmQJicV6+!^#3=q$W
z6#P!C_9NeZD(Z+-gEh35x7tiE8vP&7kn=hxzxJRci}{S?nVxhMHSHjHT%m%BSuG22
z^b%VwF#u##Ha3_?M)#m00eoMA&z?aTr1-#Z5YjqK5(37G67kWt%EPAj)M1Uw4<xrk
zKa|n%nOGf)rRuWKNclysae~K4p%jNIIjeC?NQu6^;jG-XDfM9vWPq`N;<iu@ZN}}N
z11R<5Oj(L}az|Ygen9gI!x|#ZFM1`rLYNilVJrf&d_(tAYreHIx+vuOYAuOhWzc7@
z25Kj3hYn4B)7h}QTVtri`I%mxDTXhWC!lCc#IFLrB^JO66wkJ1jCgvWhu(5H1}dke
zDfk`><dTQaJ9JIMyhtLs>)cK~nfkXR3Wp!^M)Wou^@+$Anfqh6qTq-oJ*~)77CSQp
zVJ*q%{B{=MSIl3aLOZi%w)?7TGp6axTfkhFw(fB$MYq?3s;Sy#^is6#9Lrqs0bi11
z#lyF6$6;r?!Q_p2pE2L$xQVIC$9WHBhImW+cp7LKfW0r{bMXaJxWax!8IR;0N;q)D
zg|A)asU0P4Q(Y)p?wuca^IH`(xdKQo*ZI&JDnX!FhPLgl?_0(hc5Zji$s8iXG%$bf
z8#hl6H!Wr-yMKpUZl`Ub8L>5CKDY42b=6)CM^*29LF=Hy#1uCqvgiqETpsFB;F>!Y
z{akDw-$tBTs&IlsPg{7GPJ_+YXV+&-^`@E{2tFE|d?1sk`A1q&S>Yz6G|kM_56ivS
z+9%KACGq$0;lSClT#qr`KRWLFeztp#XxaSdDU#Ia620`%Su?qth2U#^3CJnh_YPUm
zjY+-{;8oziEK*w3OW;faxxZ6oEhxZU0@18=wvxa!Ru*HD=p-S(^u*d)TayE|g>Rv|
zYd>#pl-5F4a*kJFN?AwU?_nC^o{$Y9I|j?LkGDZ0XmXhblK$%^rI?x4$D6^%ZZBXQ
zTjz%vbDFR{fE*Cg;RiF-#RH}|tacRFd*FhGzULKJ?5UUfkN89HjcD)k`SenhhWnTd
zp?_)kCCSqzq;>t^8!V$FP*NInV4^c7Z+MToELf@?Z7vIEgvoU-mM|z<03gqcP}aA@
zss1;EMWGrFluN4{93X08)*_Q}?Ml*@q68@i#1?^L;IF)xrMH-~d^}Mo2<zIuzr#sC
zB4ijG`guxj;??jD?jI%Y*$^Pc{^{0^j5keX@w?F}KJ{`j(?l_MH59lwfJM+nYy?vn
z<nR|<4$0n3oc`ojRlF8o*>TPuNA!;j0pjft-|-wa)}u?_{SIm1@7^)bE_AOZ$SUtX
zI`sB|;#J2tOy*YWn*$u?yv5`)mj6w`K2~WBJk!Lh42hBZrc7KA{O><*^#bE8K>A+)
zXZs*-RhPM4?-EZyCH6Vn&~C9S@U=0Z_2W~&XL>bnRC6d?A;bmNU%a_^TQQAl>(JVw
zZ+GP08HH;Yo*O$K>%HjB?VkE{@3`Pg{}@o{h!oH;%!R|;oz^HK&~VNV<UHe+(S>R_
z0<u`0f3EqH7JXP&i!g=e;23PG0cZsY&F3(nESU_&Xkj<8ITFL4jhdTY4@XWT)G`Sh
zXsp?QZ|B=X;1Oz1hZ#Z&YOlaczC<Cc$x-qUnKrl#Q;&MfV$d=pF!l}uoJL=ZWnqCJ
zj0&US2*I0IA|beTQeHCw;?G}<IzN6dpU6SV=jh4sxNWZ??~oX+cPGSqE;Ct4SSH<r
z!tNbIlR*c6m3Iqj{KyuoI9@3vXnb`(^?Zrr|Je6JVJ`28BGSC{y7>jN9(cSm$mdsm
z*d?R4?8emQBvP}_g?|Ej=0ET7Id@P?{a-6&Noxpp4jvVXvdyuLr{O!%Ii<;an@JAY
ze=}c=?++5VY)FSD?zJ9h)^O$V5gZkZdeq+y9|`b*%O5pi|1#~_I;+Q<#V*eD^>p=f
zuQWGt(A#u+D91YB7G*}`FC-x(tVKLt!Qu3yv7Y3yA_Ov!Pr(HP->_kUQn#~GD=(vu
zuTGKER)N)nX<Z>k?>jyWc<pOgBrM3ha=phUIO+-@x1r*475@SdB_+kaz^FFIXnw`2
zD22HiG+2JtqqU&~{y{S|5#VO{LayZTH{R<8X1m`LNhq{c1E}^`!Q3x?n7agi%grhL
zm_NLA=c2PPvHYEKkZ?S>z=mI7jg_D-d=twn*X0G9s}m&tgMKZ%sNAj;mHM9TOlWX8
zrg+kD$%t+;V}O$J6OJrN?q!|hD#DF=`V%b2+?vDCe%D?w@A=Z$jm7#5GMv4m;L67_
zH5TkODoANq6kDIg6DF8_y2iE`cUzZ{%5ND5!3!tgLHF@5dMNypQ)Q5?cbetRiZgS`
z_gttVYz=5J4Jh_~M>mxA{i;MacelN!JI`Ganp0ROh=#XvnrQg4(DeQyRNf0J%CrgK
z=S*3IlC=zdqu97HRY1p21FoqAV!_f4y7*j*U^#tG9ko3vg+rudqmQ0XYM387a4l^P
z)*{GZ&UWjrk!U#j{(d)~IMWRN^{vNx9A9{rQKWSDhR4#q*_$WhGg^Fz88B5t4IU9j
zS2R^1R2cK>lAx35?xfQH<A2}sP3!C$j!=(R?Jln6z<(p!Qm%YUkN*?tKz@enV@}e+
z)fXh|cAcPc$7zYn8WyhY5ev&rUbKhry3H-2JjTp3A~r3sh)zOU_=C#^G6`V#qT%m+
zj3O!szMW@Kiz-h}H-COiem+0v$+IOUeh1}`GKUQtG(0L(BDLZ9#k0nlYDZ0Fv&Lt{
zRv^%C>bJAq9^c2eOsBh|>RRMWXv(?y4|u(BOw<jo$bAf3(!T7q1JRyi&1mosFrBf~
z4=7MV*2Ypl3OmFAJ+3_4whJc0!`1ud+cy-FC|}>}NDva4oRrgaDaPN<n&7ke)SRJ4
z6s4>1`FSJ-;HC}Z`@!aX3LI|K#!OUG=+g(j;=gOgrr0e=lGoZK3h&&&@ZsS4dM{0Z
zASe{ASfb5u2G#qZ4xEOBiLTFlJu4gJC8OA4y4lGyAR8GBj<C4J2|^*gIV;nch(md5
z?$IUZMpP+iwF0p^>nD~LFRwBw%$S_oj_&HSroE)=7RtFE+RCbF?Hgu}rn#r>&l(Fs
z=qt(N$)&1p=A(W(iA&&xZ$+(b7<LB^=H?xzy-|S|;2E3`Rk)ScRfhESUFnMz+J7Is
zXmKVK4xl++cCET4<}{K|XSe(7l1n-dI9fZx6Q_wDW>u|WAZq8efnrHls}J=;f21u|
zdB5ip)HFSQ?!?fF5#Qm?oyyfNoLYZmP=FNRQ@MZ4UmdLOJ6a?bUh<3UyLTm5RKq80
zdb<1MHdQtMfS@}z(B;^f>Uz3#Hh`&n=hLie-7&X1Oa;iU#d>TJ*))rl8g{|5=ilS$
z)eP^nAMjRNw(`Gb)o7Uxoyi=fi3*Dcm~LVTS|VQEZU>4*WO4oBS-%-P9T0XMY1Uu9
z3?2bqx4}I?$h*%_<xNrfW)p?xAH88Mf4Ga9%KXT~$AlgP6HMol4c#$xJQ}uQSB7`d
zw=#p@VP!m5k@h8lEIcz?3%12#>hQE=MrzaSnW=dX&9MQ8w-J?o0|ycWU;2U8J{3iY
z4Ue$7uUW5zERdciyot=y<Xbm;&x&sCbuyR+V>3cO#>WpOIHe+gDK#`4eHLxuw8WU5
zL1-j_i0#TQt3eNp3tt3|IzQ|QXuC(tuRTYLH>T~(;hB8+Ro*wVc3;7aFxa6=`yp0P
zTR0&}|5ks1+DV7yHSA~@Ajc3(OhK-h<*SLwF74TcnT;fJxt!FzT4m!}vqUy8h|Vdf
z*v!>W{t7*7nbm3dkADv*7`L8@k~ayqp1P3OgLgE?<PT@}Cod9+dLtNa`A92A^?h1>
z?Q-8Q=9iDr+1A4BCQ*~5@80`qD8u^yuG%~e0o|jN5BbL(0-w25El20OPyf4W5IPi2
zNL-0MRYAPAsK1>>HB0m`^<^j?*+Fd9yX`ck&UkZCd#J~YO|l9NL3Km2e9P;h@&hxa
zUgMgHY+6<YtLJj?SA-cobHC0+v(;Fo*rtKuXh=;WBGm6CtuS<#D8nN~wHA)>BwEeH
z+coPP49tsEivYb}?Eu%2%ck&ex_^pK2ggzVt;UO}TvfHSWOVx*2p27BOZ>yOAsRdJ
z9VA!W%Zo?~{F@{S8IW{Gw)uMHiN?Q`-WDi@7$dF@4=1_btSsKs<I8T2hb|qluR+h_
z&i2*R1|>VRF6BG4bT*Aa5}UE-@I|AU#^)U7OaWL8?fw%8iu~Bgqg^D4KV1T6+ZK87
zb0FTfWN)O9|IAH0=u%ghdUN~fPqwc}p{v}d3`T^a6GhT)0;heWFbg`rfZn!mw*xWJ
zlo#Vb-ZpF1W;tN@tIZdDQtq+~puKfXqWttvZ0<9~5K$2UzB^~uh!i@Vz7w67flNal
zW567laC)ow3YbfJ1-;C)>EQ?KIT6TO5_-3gC<BS>!i&<7*Atep9_)RjVgxh&Ln3P~
zJdGl^p=KsAWvA`an$*haizH&hB;V1E?G7Gl45&o{f0Gz=yHz~@*plr_)0nI5>HoO7
z!!m=QnMxEmwW^=LUJl^!%*?%g#tsP0Ii)IVJ5~xxCRe!i4!ft*?Q7<C;Ybk}#Sr}i
z)dNlNvL`I|+6VFbx?&oLQ&xWX%^TY+v+zGkzr*AXQ_pf6<9XbNKdfu0iMGURrF{Wy
zb~^gMDzAm+Je~PJNUC%0ct209`>Vh0Vj%)Yc_`>MBkDaA6v}1+@_*+}6OUY_b;-)7
zvxF;p>5iW!zy=RLI9fI5vo;u(U&Bw0`&c9)`KwaZc`aS*ZyBI7cb8FlM|RS~&Og46
zc+4zT!?8Mi^)M8^13buwdy`lp+y$H5R$6Czv6ZFWtt8<v)w6OcH>0YxoHbtSp~ETE
zqr3x^TszMNc!oTXiqc~J<n%^=!;(gFOrybN*g$?KfGIdk41ZU{J39p|=lwYg&Wf<k
zhnS!ZajqIfo=>{ZwS#N=X0vSM<(dW$HyDZ?ic)gl?>nykV~-`qEpx>(`o^K20jE;$
ziA>;jXWAsotFqyyy-6b*_o1+Qalm#wz-Lth@7YxTd!iy*%)!5@;M7CtxI5F#D8wPt
zgx0(F#%mWSgAa?*)*3Z$C-B0<Ee&wOJ8|bBU>MyMj%HCzf2IB9l`q%{-Eu<uWh)q+
ziZ+f7Usza?=aSv>eOntqg^*rGMC}*wH^U^kVG4GI<f((dL)(|%%dZ>RJs9ZjNG~!Q
z`}({JH)Uyo{#)2J%n3)thQ4fYh)^tsn16Q0vn1HwE{0c41pJbM{5Q7SM*r{NZwlng
zU;S7?Fz{jf=l4GA)E}U--L%FxE;YptWzCh}5`3^*55-cdQJ!&-s<G2c1-A2%-yY9-
zR*6`Kn>E`FwozA0zefHW?pQ?$3mI+mDYW}RZrqw$+11;qxaD8J+rPO*JYB8kIFUJN
zQ((NuJKi#U?<*yN-&NfND1Rg!^l0?BBf4iqeY-0b8!Nl-Gh^6>_A4LS3(xZ{Wo8}r
z5#Byf&J=LwRhSHCEOC&|JWi_mxDC89)2}{r!)1)JJ{N}I-(@juO5m&4oNqs#XhGNE
zi^{?o+RkUKSyPUK+RpecVymY@P6305#Ab0xVB{bCR*9_58fcF8`QCXFG^Z`EdWZ+I
zv8Z~YE_<iStX?U4SB1ZFnqJ|!<)c^0zZKV#ZTm6ZIz}m2uf@EwSS4P!Q8791{f7nw
zG)g>2e?4`r)pm)6YBY8;9I2-ubs~2YT=>E`AE)g&|J1WWEozWL&GLyQh7l7b|E>Gt
zz$@s+Zv+}@kL<<}h=;fUA3qU~n3kr$azeTT_{TXkSP>-vEVc$1%VNjCd-GBwUoC=M
zin(`*k(>LrU;XW(T<z>D*W1a#Z-i5q{AKhctLl+}uLmE9y_F><v;`+#%3!&CV@;jP
z`aGn=sKX<D>N%IlVd4I);r8&)6}mf9VypF9I?gQ~O8oh-=4re)=dOmpnEWb4hBvK6
zI5v#BKl2sZm|4I<<0@yA5o(?4ch$~<7&p3W5l`P6OTRB+X$CbuwD{rN7)@5LtnTXT
z_s^;jUHw0hvUI+52!SG+`zCdtkqUTt_N!k%5F+s%j#y5B(rKGk=QKXGcQdM`e8u-C
zg{cym2g={)NP^XfI=Wqu8s55vkJ-YH#p6)Av4Kadc#PtWu6*|6pX=5@;SvrnoOJo_
z&ZkQ~>0fGL>Mlvk4`{7kVIf`sN#MrHn^k|M0t;3=uRk89P~A7rM|dwKE%mfTOkp-5
zkDG?vGXR2AaEi1j;>Hwf7^V&+AmYKz)5o^M9u|ezO(uO8rBhd+`;+hRsYA2q?FPZZ
zX<{uXk!@u_t!<9S7!X(mj=*Qd1Ah3z(pnc_BgyC(@wo4X<o((GJ}$+paZVx^{80tm
zmpyO95u<xPH7HxJR}{gV(c9yhZr+p5mE$x?Msz%u_p3NA<5h*um^n=S6IOOW2#bRG
zM+mz~0H-C}4%U~~C|(nS(qa_Va5>YE@8=rTT?Eek@6;OI`=)XJ1X0#ZO3)6!4dV+n
z`L%1*{9d012vJW;euyLJZhjqTdzZLL%}>>)2__DU#d~9({Kngvz&8Rd7QXzdg*{44
zG`lMZC*+Zh0bZM8JjAsvTr^QsDbD&|OAr?dpMObn{PE%cZwWw0`|UlFXBIzRP??%g
z*ix;VLGP2eRIvzI?)V=i*W0KX7t;+v00(KrbiNn1X>>e1pER+zpQGWC&Qy08&BONl
zb9D0q)?iy8(tmm`4~=YXFN3^;yb`zdKmt!MeBQfyHp}~jAA=w6!wG7fv_Wwd;Fis2
z@Y{2la~unh^s{i9_rKpihWPw8lo}uWL|?x$vVRlkmvyK)xf3v;_h|Q9Q)W#CF3Jz&
z+X(0#rjc~E^D@3JFW_12*mKAuo+6_JTMlG?bl?1vqvpZ}8^^hYPIEXP7!AW3xV_HB
zt2N5Zy(Rygp6?E)<{eWOxWW-F=6}u7xUY@hjNd>ssEqvGU$yU)Kkd>;;`YDak*Q6V
zj6EwQ*nSZ1I3j78r-C0>vqoVZ$??Ka@G*wjJOZka#@jeAx8)+<yfwv<{19KZ8pkIP
za}JO^VR~C(WDW~qiDMb1^UI2q)1ug+CvYB{BNzM*zRi{P?p8Q?%_>m5bPuJAMeKE3
zfAEzFI{N!P3y2Z4^anlClOjc>xVAG2Ac?YW&1V%g1f7ymu<U7Fzy5=>%E&{C?<<3A
zEMc2~Ov6(Z9w)Fw`6_SKi?RaX_=;6Db1ZVbp)G0i7jG0j4y0>-(C&bDK6QBLy~6E$
zKuCQnL*bfj_`_wf{+XmK?;gw{Y+ul+mzCr1-q7Q6Y!77p&!aa9XK6-MO&42BTvsQ@
z2Ja&BgG+tk;S7gIH^MrDMnbF4P;<Xw=~hvJu38k*kdY(nzP5JL+|~|y5`qBtMr&Cl
zAW=Zf+-3kn4N{BKRoG$rH71>8b%u2`9KjESlw{rwo2{}2y$b$JHX$s?fj7D;p_Ds~
zc>e1z(Hs|@&vc=neg5ZLJUZ_GHw&O^vZwlsGi?=hBmh*bl{b6J1oJU%6xf7v$Trvu
zL;h6MybE(Q1T0=Et=^b=&^V%iKp+6%O@wNS6LPbo7Ts~Ls)GD{f+5ao764wp>?C`*
zUD2Vs%3INl5?i&xl)RyH%Kpb^t0`7+<k|WGkmvs?{P`F=%08OaTALI(KABTUn2`WQ
zw6K0e0L!Y$`5KfuJJ3x*B_3Z~>61ZHPC*l{7$u&=rn(7kb`l|@>Ce2DS7pPPr1Xz=
zZOQ;el`Ms!mw$RN1DlE2iJPjW+E$06h#@o7qOWs7d!Q)|&2Jy=O!ygF`QCdwNTveF
zFgThuzWpQ>a^dhQ9uvaA4UuL>mvI-%o0s}k1)C3;KT@Rnh=T<CejFRVAibhBhZor2
z8<FkxZw?Yt2f{4}E6pwD866Ur-y(bo#l$1sqr4qdw+j~2_^xt6g*c!^580iPc3bE%
zY=;UiIGa^}p;O%kS!HhQ9r0S$T*#ir2TwORLxd$y9-%EmtIzMXC{}}#|MS~E1J1MG
z9;(rt{;%Cz+C#oM03j8(OQU_KEjz^)w#%o%Ca1r+NSZU@u#Qsm*7q($H^1~2<*FR!
zlpPvXm!Ubv*PS!D@ril%Kb6frhvN+@p|K-FgK_Zq*o?8$#5_8oje`--hrI%tY7gHs
zP~<S53C-baU1GPMM+BYAOa^y*@`OD!rzLM?3LrV)#m3=m=z=sd2GJcIImFuaOnMo2
zKt#plR@Ktas3o-yZwCFWiboc%TKD(&dqq$67ug-KV{7+^8|9uJs3i*muUgA6D3J3*
zfk4!JX(&0-h?hCeqe!>etc1Y7Km66~K^sHp(JvvWQs1qxFNFD`MYR-lR`RELPY;k6
z#U|y4jD8eQVsH@94kbiZMZqG_@kNv!tx2RBagY2~#4(4BG>dvA{TbnfAd_{;!%lqR
zN&Me7lke@L-*x+ocO1U<m7hc^1!<o!<+{$>p!x=JDe1rx0@=z|A{9ICVP_fo*fNV`
z&4pm3N;^f+Hw&0>C^yMe@nI9@eU^8wgAFw<TkC)0irDjqmb!|f?YXaw4HsI&{uRjZ
znLw?Do<&P<*}Kg8x^D1dRk1^p>fiYafkD3#{<`z56;zSVxayK|tPMcL7ri1xi`G&L
zhKv2&lKt*qd%;lT^C}U?9It&|?EJIwXLC=`+V|ei-!tLzveCo;ch?m=0mqwVWq2Im
z1Bi3_`KGsYM{5sbC5AK7;<=$S-h{^jA$HOnmf5x!niSdhK*(OlskvGtj|>;jN2nu>
z9&2w}>`LOSf18g-9d(fOv!LZWSC~@<62U#_y-uEQcP(h;JqzVp8T=E_vQj;F5<};l
zR6YF`ulabU%e)MZW!o$Wrn|g4I$93Q;+#EUUhTYNRLj{D%2C_B@nX6f7v_ZdW$Zcq
zo_hj2c_*iMi6+9sL#^2avt%8Jb>}c`MoBQ;5(jC@$AU5`sw$UL2+4i+raCrR&um@_
zu{%1s>ZAT}_)+Qc3gNh@L>wx!K{HN+O>EVL@|JR8O%rJP9D?Os*@VfW9{HR8QMUPn
z|CckLdSBG$3#sX5wyh~vo4R5C?MT>37@Yri33H<`L|cR?pV3=Ra=vToJC<%3?Q7BA
zMLScEE8DYlT8hXlnF?~kqnD*!N@6U$5-!-B!9=r|pUHy(fv)-!5e=oplQX%Jvuyz7
zf<IuLp|r!*+4`u(xiW)buEzb*bk$Um|Dc^HVs4lDA7+7IE8h`<IvzpRHb=SYGH(8(
z@X^G2aYYUsO}4^4&0sd(d~=apJ*cK?1G*KM1vCz79xBuMrEeaU+*33uP#T|UZqaVn
zKg0hs1eEUh_$^1Y(KTn<+@oPZhwEa!j%ZfxQr+uk^xrCs+~2L98_UW;eLC4XU*A4D
za`KyBg670d4K~9OPM$pY(HCe}0^ew?Uei&NSvmgOnN*UH7@E{yHy_70M7RzfHw=pA
zPQiCG-RJg)Tg-<oJbGhue94RarB)?Z+jB-v6B7fv2e!77MRrSWofs72K(s4u5H9>^
zIDx$1{UfO3`n;X~RzY{{E9^_wG(+Pex-d)x^2&7Us~U`;W}C(oC^I>#`~IWNUPo>1
z9}P#hRf9W`UrSeWPr9*v99(7|rxo1-_E5wUGpBv9Sx0`ZE%&DN)>{9c>T^qb1@Gv|
zB5ZP>$a;G>NtBwp?nOfH*M>XHk)=y(Kf7{!ogL9tbx^!s8enFM_@S1eCUANwuk7KI
zsCna@SkzB#QyS6!dX9PAISEj*2xJ;}S4r<{jpm%9SJ)-w`-Rai1>16I+YmlM`yxBS
zD`&%}gQ47<r6$EF?GWrj#YUnj{&_rR4foGRV9_u>7pgZQk}^HiD4o#;l?v8O%OE@t
zI(t8{`CF(qV-k~`DiVw5sCiGkbGbK|X3OLCkfKM~S49;<IlZp>Hn(L?w@z)I`%P(@
zf!uE@nJo|Zh3?{`*Mf%mhaQoueV^H4xPjnafra1D#?BWpqhS{|_6-$qQ5_^Ha`Blh
zN@%AtY&pEG<brRVdD`3ifLeXOwff|Ur_wx2ndO4rWU4OXmZb=lYi)ag<piwRXwZYM
z6My%3d!Scq;VVAKMPRXCnZPWd)N?JHsm5vPS=kg;olMaKpB>5g96fTT1U8sq!^Y=e
z_q~h5@cI>fvLk;nan7s5AL9g__#S#f#V_-65Whq}^D+e3@oY|0cEGJ(rE5*>RKXFu
zC)WgY7S7?hHK9T@p;u=&Z|l@I7Fs$hj{=-%V2ynkDJ06?_`UTx&*j*Z$n^VW-j}MR
zHEC*Q8}U(?9S<f)1$n%qgvx)^nnGkkZUK@4=s=9UgLus`@pq<PT&><Wn0Tu@O0jwz
z&fMtR^#>>=+soH<lK89nPTm8DpG~VTD+w;4?M4za!fS8QWC4=Rg9hbr#Nu4<c#b5f
zd~9wA<;%hCu|>FH3%ahy#j>baD<$F|F_bP7a=`u(ZFS!iq2q()YviyG)he5lY!of=
zHex>8I^#>uKs3Ct19`t8KwJMtt-ba40@DGi`&#sAMh7N#Ri*DarG5YNnv}mz2`ZkG
zkGDoAeJ=0taL-dFNfJ(b#5*AD#=TYW!kcJEs+MYG-TVPBB$~Wa%@9B+rWL{bmdCOL
zl`!u~9AuglJsn0?9k%ZjnA|TaTqK|0+nw`?28(q(Us$3h)$w)ud59~RebC}v8{g4I
zNX@5r>7gdc5W#%5xf3R_i40QD=5{5n-B&vD@{VAd=ICq%W^=t%N|BO}O){~cMEpp5
z887uL;@%1nk|x8>xKSycKK_+)ERpi-`Fs)nwcch|{f!l!$mb{Y+^F`s`Rt3y_hEp4
z6)=JRC3VpAMHb2)$Nx-jHJfvS3t|9DY^}crdGit}PEBkT_U^V+jw;>$Me&URjm8@4
z1}N_3bizwP10h0n!xAOytqh<oTas@33Cy|o?L6&*&tz6Dr=z44snz{{Q2cc)i-^5@
zjPu~TFqg*-d%C5xF>u?pGq*jksKi9^PMEJ4KMAqZnw^)G;d->VEin0+oomiLt7e7E
zvbW%dGdb4$W<r5;{brXPe!WT%KF2dP9o=F^hYx;M!9YQmvS%?lqrwmuoHo6hs)sKf
zUS`AZ3f8BO`;yoi-u`(PwLv3!bto!o)x*hyda!@9E-L6dBY{`^j!6~q!!dPjQ+TB>
z*jxr`nOFy$5#~3eAX}Y0Zaw&=B<oN_Dj>|rvywaA`YuiyCMmbN_1#T<HRx;g-neAx
z>d7s;8r@kh7Uz{}YO6L<_95Q%HcQV9%8#j(C<6I&8V+Z}+f^$RXg?CB?Sd=b%hE`8
z3?P8b-Y=r}l7)NBv0((4rw^^I2M%KHtbMIlMzLp+S|>3_NEv^E_KV6vJ?r<o$8Cje
z{a*TY{3870JWh8e60EW+9H||!%i4exo?^WY&m%N553!TKnW@q>BLEVS=E!^8DfBB1
zcW<ZUr#;_KnVXc@O5rKOX)0g+-SMKsy1g^SFp55WKekz0j3%G~&yYF~&<B+0VQ>H5
zwjctO9S{LdiHx`!f2o+FMPPVtyew!7!{J4m7sW7A)*)a0Qq%p2`le0TK}>R{I^TpP
zxPmmGG@-ejA@xk_>o)@H=HOw)?$x!?Kn(R@#N2uO+ap+5F)Ls~wnVFQLAc<HJjA0)
z$O|Z;cek~d)Y6|>WH)C{35E-8(9+Rbju-&6`p?QvArZl-t=`L*`&4cjT7`nGss>lv
zGw>yAAMwlm^vgWZSLTj}*uE(JWQoOl^iF8Kjn?wcv|dbd9uFjO$<JmKuk-3GXI~c4
zprm*ApK{;n?XPfa<@zVd<k>?Lx9$J=%QV7yDW&_1;bQW?rqNfj_7Xas6D+Iqt{3M0
zEA@|8^>k;Gx@FHO>**oZUmRKabV>(HcYsw^qd==a8W?wy()E=R;Wtp71?L%hAUDbH
zc{W?=8h>QrsfUjnchEwXpG1fw`YLUO9S)AJ)X0Lq?0DEadFKcAAzT$+@$O!`wM{uc
zU>T&TX9;k!L!J|S$S~+M$?irp+&uQ$669?b1~s{%lsBd6j(-J&N#b5$q2E=19J0A<
zm_|AfcDx27%SjV|Egs{n-&@tmDrG4hj7?&+Bz<rqmX8uU6@uW3983KMA5_El=1<M8
z{0HW98XulCXmoao6973MzC>)LffRom0wY8jz;d*rC6K|R+VL8ka>od)6IYBR9vXV*
z0l(>zqu=jOHkgj5pjAg{Z9HchyqQ*hGRsFrEp6}qZK?5Q21r_9Q3SU1{g@PfPR{k;
z<GK_g??31op0tgMI-huUIFjDJ@b%s9XyjSjwiX^0u5eyBQkLY>f_m^o#iXA3RaOL3
z!ZOYy#cj|c2}7=rq#lSZ0=nL&t)TMYFt}2>*s8z!HCl$SZ+^>w2GmFwoLJ&XiQP*I
z)ciS~Y`*4%Nn})(fNvkqV$d?7y}~dx?aLY^A{~&5e#GG#+H@2|gu2$Pm$7q#>sb2&
zz2^_Vg;8`keb>KDlF$7g-K4F&+<hj<AIM;77`de?KY1GOmW4dStde^@Q%MYWs2f%S
zcyi5Z>6vyh%DSXv7Y4<}fF<f%ue?i^KB$@IMme_^l-u6Fu7DGiG<yOx*iF91@Uxd^
z?1HhIvqZUg7#+y1{<2+UMIojR@7$H;*x1zEB+Y80?K^razIZ(Ri*|+-+gp}$`Tdrk
zx?(Sc+cR8#{}SEd%vL006b`@F@2I{8?YIZ-awyn6w`tnd@GsH5nLi%m=<aw^vu>H~
z-71XFn_2YLhAaQ568&TqZ@~vO<#96LoHp~nKF3NHeB;4f+S}*W3ty8xaUP{wOxAZu
zbkAO$VL5kt4g)vn>My@am`hB$$N8sK9Q&C|2K2l4kYx3JPCF>z>Y>-s!0MdHmCj*5
zB*tx@oxy#$UKVs3-V^^+x8SzAaMlpDGPn5Av(MfK;ONS@`Oo*`?ln$x_<eY9LD`^7
z<+6(E!lld393<1-x-lR*b~PN<djb$t<LFrV&tXXU_uz5Y(SJIh$7N>GepJH|$V4HT
zze{Ji{OAW1-w!`Pr`q^!E$VJRn73s{owfJHp3Am`=T@z)%<d#U*s+FX?D}TuuWHR`
zeO}4rX|EbZ2SEbPft?o*>~jGcFO^Ux6Q_WWEbJG>Sm0cxu~Hz(*1J}tweAK3cg}n@
znkl>Mmo*_|aRY?&@*zw^89cG!<Bg81eMD}RH4{yzWap)+LVHFsS4*gZH~H<UBW9BV
z_^xkPEj>mEgk@B)P`X}?rq-PA%$BB>7z0#0s(Yb@0%8N#W*y(Nx`TP3G@|^+`g5dn
z*?q$C2OH87!>wuked`4~ko#?WX-pU{e$7k;u-{2r=X?&}LxZq#HPA=r9(~6ldf**-
zOTF0Q_|{Z?^~Z)~(a`g^B?b#!MQlQH)Zeo<@q6A_T#(Cv>KG-4D#0#_ghw=>$~`|@
zSa@Np!2LDZKw?-S{Mx0}wd;vNEvM4>+)|>zSi<->$<YV*+JYCNzKM>)e5Z^ii~FS<
zT#XPh(AYvqt)RlV3CMpa8R+M|pQ9{13QiCIRL-G=YfaHGmt0R%Zj(QvB@BC^TSjt~
z|FVA!jWbt`(gtm_6O%2Q!8(^6K&lx#O2jyag3IO5y<W15uh5>l229du71E!ovq$M)
zXEZW_e~QQPG=15@Ju^_dFTMD?H6GO;U5j#Sfti0Pt)bB87q7AxtjfDGn^ECqu~^B#
zSso_C$}tltzq4S3BSNl}D=Rwz*Hv8YHz|}$QF%gy$%MOTv3oD!RE9OQ4}S@(hRn8!
zJ6NQbIk4Ck<mcMPJ=e747HCY&a@3o$nQ?X|GSjzm+GD0{#dVkOGAqurTroFTy?xV=
z#UZ@V?aXSH%AcH9ec9Ad^ZrkTVdEgxrm+w8>&8mALzf-3X$M!l?JT*l!WJ#7L?wz?
z<|*cD#i>$R<MXC(4(1w|yK@xB`<_7W$-p0O2goetltOn~j<IYo_q&~b8ceI9OE-wA
zzp6dC-^;TbVEIR3aRhSz*ONZQ0V~OtygTxL*!}s2C!SLoz9_^4#~0F+>PQ8m1MVGO
z?+f)^$FxBFvDI)yA&;v?bdvs-G*~(X&f6z^_e-1jSlDO`Q)SL<dOmj2_qNi}@(2!t
zfPSw8k<YK_o4YApo=$$M(Obg;m&pzU-=@wCX;T3+rC2^*+DG&Q#P*PJ0I{~9?2r=3
z?1kr?Ty&TN&2>s5UzqGlr-W{2DQOYiYSgvv*^B$W7lzgIl_bsvuC3UxgWDl$!P|@5
zkB1d+X#~D$dDjt6+|O0|xMiD=d)9POV!8WPpAYNaifzec+VfoE1*d8{xL9H;W*QHu
zJl0qb#eZ}DO^vcVq}!tXL@|C0H$M^MZZAah94Zv)h8BKS|7OUecc9<Xof!N^-TBMo
zANOC0wAgRQC}}<b7tXrv{KhNTb_Ke^@gJ~qM_&6}8Pog*mOAZiqTX^_xK@jBNB!|n
zGRJD}&y9TqQZFcxPlxIjKk-aTcH;1C3tvf$%BBX|Rva)GKaUe>xHhhNM^HaEGQkMk
zVUOq@=_fna34FP1@ilZq_Ujf5*!s=?m^r)FUQWfQ3{nhGvxN#}56Mc=`~{GQS$3)k
zLE$T#@DERma-WV&QN^dcX0RD~rPWjA69mRQBc*A0y5f(zJ%2bPlvTa0a5+~YY}Zba
zQh=klqrJJLGMQDKa<d}zD|to!lAURXF{AF(#WFQ+@zq^qYz24l2h0Z?{;#$PQ$G)w
zr}{QkO(vBOb@k}9sK63hAt!oS;1h|0VP;;^>RMjdN=2iUxgM6hk$P3>9H_GaEW#7f
zP%}GQfv`-rtp7fo4_+B!WIH1>|D}xBOW&*%&N`3VkFhTim9?u$DpYmhdDA;)1;P3B
zT*=6cW@mrcfZOh=*>d^-G6hJ7uT_^ZK|izg^zUCqhe(c(aE75|BrGvHCKb46f6(Yt
zNAGxHW#1a<glvnu+~r%<Oc&v(c`nw*CwNZI(K}fER$6Vq!`0}1)xc7lOi$35y7SxW
z)=(dU?K;oh*z!4b7vI#wh~l6PDQ~8l8J_&qlar@Yc+rts!k#`0v&TV4v)_GT285s9
zL;=Hf-};}k{=r>w<7<E%9gJYA*7{u6{j7~tL59R$X{KxAp4^?%5^4sfsNlSVBX`ZT
zXfSu6;5V~5fm1ohTZ=N{EPM^arLKgGv7JPzf>WxkeO>3ge@@Z3)p&<=$FLPeS|gm?
zt9|y>ZzZ&|+Sg*uPht*`W?1?;C(ny<Zqdu(TrH|NZk)oMWOu;DmU)`y&Y?*cx7OU2
zy>a^a4ojV7HiY-DCo5;Dn;%DH=sK-@|AIx<dsQ?Zx)E~pUm4}9I~y-RzI4T|Q-%k-
zqL3MBV*dHnpLxPpZ8D$Seky{M41NOv)-o96u9BLn>u8+#um62;a#e^k8l5zCwrr1E
zL6O47x7xzT)O{2Hv``5AQHoq?EjvQj=m<jb^Z3N}6B+m6#M2~>j;dcPaf|hs)A^rN
zv5FLXhtZ@^!uS?@YvWJm0eN1pgUM-GnvldbF=b(uF;wl)|3%YT_%-#oZ=ddNq+2?q
z8<8$$NKab2JEcQlG$@V40O=l$#ON;RMw(F)L!bHi{+|DU*Uq^)cU;%|LLRaInp?3U
z|K*uZ+wbS61<XmCKS(k>=AibtWuni0o#87N&Jd=~%pWihp~VgR0_hAmV0vQaa?p2H
zdf<6>-i9NiufXb0`cqlR`aZ3PcSx{X$6c`dYAOGVC=3hLugqd~{qq*baHrJT36~Fd
z*>+Vjyu3O`*PMI4)NRMD@Df<hXGtVqc!>?3U-;4|@Z9?tvXIXWq9s7>{a;*vsoJ`w
zSDohce06h5HtXn$p@wUQ=iQ>zV^T2}HZoa@nV1r}GLt-oHg6xru5!FFsl84NGuR`G
zH#_lvo;8~=w#dU#%xOPWK6N|mnfWbopE=6N`qGWEsC?&h!KN2<b#<)oKfjOT_5Abu
z^ju;%fCE$zc*IW?A9CMO^|w21%5vCP2HGPMCEF2^{i|fMu4?yC?)mVqL-7{7dWJBy
zX?(<%a*Oy?7deH0&<_=kkx6_Xi8??e9i6@TJ8F1Mm}~OhFRJb_W@U2!UtfaXjavJY
zfm2Q+Q`QwH7&?+lIqjJb>qElWrQZ8VNn+b8MnoJdgG@V8TfxDtcnvUK42q4EGllMn
zz8&4At;G0q)*|^h#q+3687<jMouhas@t|Q$A5?F!t4cU<jDXxH#jPoCMQ>oYn$o9z
zE4Gp47IZ5{$2UK-z5fkZ%h8NHSs>n)@1DDvwj?e&?0Z_-si*8inDOu@eS`(vCqmx~
zg-s8*t58pSUKH_WAk2rsVg#3B4Y;F8+VL=mnAG#E*VZGN54d?mJ1&k=EBR}l|3}hM
z6IJ|%-AjuY3U6j!LVL)>(G&`XR8xq@jm67a;LSy`FP#&>Xn3E=J2aU)qGW$3Y-)x%
zM1zT!5}dq#jT<-{dYSY}N5)QyJiq#LwD1cWhz)38z?9dpJEb|GDqF=7m5nI*QXKgk
z9%@Vk)Si<r4Di;_okd{=g^!ZuefN=yzpeX-ol(0~VeGXq8%8%(sL;oo#Oy3JbGOR!
z%H9zfw^1>yk|`1$a(yUqlK@v+rRz~I3ImGzro444=(G&^!%>uK|Dsuk2z8rp$eR3=
zx2L%cHGO)?S|IJz!5c;XIXdU2e5M>UYgl)B$iS)9COHcN_LjORa&Ph#xV^e1U`Ax7
zFf-@v(hqcO&+6kmO$0eMsb#wk`yczroSCKAJ6rs>QW@|*mK2{?)>!G7aPbBe$hLe#
zZ_R9KQst}BT)*g~8*Y9F$V#fYIo_J}fe`RbzF*=8<vfPC&}(^fG>^=;Cl)QS{-$b|
zI&Vk3?#T)aR{F-~U#q64X6|O0p}YOwYJY-Cj%M}wobg>dz6eM=U7BbDnx6nOMv{Mr
zu8M;VM_~XCZlR#h-TMGc1IhwPqT{W*`&pFJ`)wMz2CnLqo_u;jG0F$ZU5zUL3w96q
zUZl0UC6K9N<JUu^K#ZW7>HAFkdrc)zE=eupi9Rdoi86vh4S)NKwV07*XC%1<9{`_L
ziqEINFSYYqlx=W+f(r$shWuYZq@B0`B@4}362|^DtML=Op~ChvdbT;2JHn?UTT6kM
zzP<Mg77Gt7gPIyI9#~#+{UQF*sd1iw)AP%65>?Sd<#kapW2x)hD{mGgMaK2()wx&$
z>YEq2l<d(w4nwZLa503$JEgxAFN7M+E4p}X(Lc<(#r3O>?-Cft091$wxc|C2ZlQM1
zg|%pWrj_=*HJ@b6&-Kl<>lMExH*h+;LsfkE-dQH+K5XY&-zPpzQYKBDv>RimixYbv
zri#6s`$+k+d<`$<__&za8(t4T!dgSF8v9Z8eL2h&-P*y3dw_<1?aqGPPTpSgs$mI>
zjBV)I?;Xas8;DC+zK&Xt_HV%cnc;dg;`m?NVfu6|+w>=4!CM_J+OZ%JQkHJvVC*jA
zy`r&ba44}P;=HN@Bmlc)fg4QHfF3p_{F`O%5n5Oq-8ZE&96<;H>y>kHUgnADy-e3n
z6k`nPMuZxX>Pw3$v@Tq9fvVCd+#4)<``i0WkyM+~c~?~JP%~Vw_1b}_mhWoMyz|U5
zT=5mVlI>j4ce;)r&MQ=OQQ#!Qo$}W9Jm#RM!~VbUp!=9lf6p(Em3~i9&@G&x-6{F>
zt$oPcWUxHAIFS?#BjeLvos|I<wE6}`{AmiiK(F78b(Nd=bzi}Cl(_K%)9Jlh>w}j|
zq9f2hPU;@w;ZTiw&-W-w7wqO@Q!+yT->Z-?aJ*7RTg|=^XxxZicbJA3;@O9R)Q3&<
z-X~XHv%x9(Z5#Dx<cLGwr8?2VNgVpmw%ZjQI+1**MbsC}LMky5OfPY_B&NbKUUPhV
z^>c0{dQQpZL6O2+AKlN+O{3x2O_6aZ6epL6M#?Po&1I@`MNU~tLSo6c?;)3{hMywi
z2XY#f$qbl!xm^50<<VQ~!q(h$1ag*TdM0(OBC}-T(|GqNee}D+X+Q4{>@k%c?yO{N
zx)-^8`<cPTm6Y|3yLGNNMJ%%Y?m}31V#<5Or%OY^*y^wJziW`u)d?H|;UjSNS^c(a
z3#4et|L(fFrU#vEnqJSVdaiZM0!ZlO@moDD>;3s}q(~p)%+J}Ajgpp{^8Z)V^;HQu
zzcxF2tO$CyAd}nE@3Wg1E^0h3+$JatJy#ca9`YqK(CxSp$93V6r&<q>X}LT2_#4nh
z_Y;S0#%^1)c!=c)LQ&|H_#7faA#TF~*+&kVE3D^q7NM3)Vf?>J<kj7`(_%f?vu@`R
z09i;XQUAyn54qu$AT~6`Ax~Lrwzz)jIZzkzCdnY7FwANjEY7EG|MPX%Ve>2zW@G?w
zBpA-GpRfi)^?Z4Bp5eyxH4{CI?Zdt&ai1#|POWBC;_&#4M^wmb)b>B)Gxm>FHdAv6
z;vTPu_SyA~u1T%8zqwg|-at~qLb@xIo=!xPlb6t+XWlcQhOuI&+D$UWv|1;5LP@5h
z`q{TU-#t1gYS;=T46}V&Zo@14behl~Vi=25U60&6@<+Iri8L>qPNt(GW{aNqgz6nw
zQRPNopq3l3O!LumY_;;ON+{dvZOheS=El2H<vJ$me?`cz5#mgCibwkI%n}mKZq%i|
z_=zrFqGW>>aKxb4f#<rZRF~-^(yVfNA9wNO=tPCMxEyyolqX1ZvO_#C4=Ge0N(Rb6
zN$mm6@1L9TMwf*{Qa*=(0{YA<|LWha?Yp$dZ{GP$I2$yUU4vVGdVCJC3E^Flm{!SM
zfqC*t*{@W05yY;;SRO7-Il?{36c|FEyb0*R-~30l5rBp6YE-EVnF5+kQ~JdW&SS#(
zO>buaXwkFrF#XX|XWgNAjx~qpr-z(&_^jZM$eC+&Xcz<CL4#k&(N%0O7hZ=Zk0RZ;
zPoF9+sfdia*4!96`Dj!Ye!qwwoIqR>Yq-?)=BVq|3S^bmt?Oe)2114iVrB&qs_rnw
zJPPvlwrF7NM33-WUa4l!vI>=uW?^XHe7PzVuE)e<E6@GRE>~IZIivUBp3R0&`NK^H
zXvevJWyJZ$J-_A(Q^0Qd74cr$f<(>nn^e1+R1}E}16O5u?pUNnB)0K8Zku}Y0NGH^
z>(LYAv?hd_)|yU(2BF?BKSAqx(R_Hl`mwbPar0h`IJAPfWQ#WAYT2)s>)<MNXijyt
zL|LD7ffq!|7WI4=XAK6g9)r`?2pY{=U6BzA+w3W^&Yrn|J}TcQuJae;=3%C*%`LLF
z7|p(^eJD+Y;6FoCD?$;wP?!Cj-yW00D{pe%<WGIujKA0<sNJKHyNbG1v(rXM_+c#i
zYaGS^vQm5dJxMijo)iNsj7u$*Y7rwl?5^zl0uy6CLc~5Zf}bMigrdTuX>#I&=LspH
zTm9h$$5+PsS8tq>B2d7fvruBSc`0*Bb_3~37N*{zhDs|8W|Fq#7#mqq#~@+gE8|&+
zI^tINGJ(y$ea@$yEAz@S`P(>-aS3rREKj=9Y>Qm?Hauw9{zkq*FJ!@}ctBktoy{D@
zwjt8dt*Ja0#O;ohA%3MvCKa^zB8bbUI0$cJRy`lZyzdp~$8&oR-YfiWG}|h|OiGYC
zSws7CMsQK*2%gN#qXOiYrA2aq9S)g}4XE~bY9}i_MJ;Ah8;B9|a2g>|5+f|^&Ja>+
z6jAU5P&NJQP#E?K$?p!D7~8^W+8abhn^ES*nNfBlC81Lihf08TPN<JXr(pbk46$=+
zAk`q9V<N>RtYb-Ms52x0Id53kc~7Dh*>U`Ly-+nm^PEkJ`0PQ&a0N11{9Kh7+>hzx
z*k|rS+)27R+JEM5o&M7?$#k}hPl#oBV64)~Nqj$b=V<Nalo}C4z3<MekotV5u-2r+
zmldOBGrGCq!MOt;p79Wu^r3L{h<N@I-9Hr^#S!oUBg_<I@1_sBCe$<o&EI)^8}@m3
zF??-#mX4Zgkq-Uf<u<RBe~zgacP{tqp=)weMB>6=SMOsuy?8dC`>{E;0j>OA4<@t~
z_>%Y?Qn)5M?>rG2bSX+Z4t4U<nIm;9)byNZTfKLI$%U{FyN;RgzSkkZ?;u~iIf;WS
za*(?!wMx>|l-;|{+ztN+c{@ouJI`QC##Z#SYD8CCdhyAC{`A0k{Ri?(`QdH91PetE
zcfmY74Dq&#HS_^VmrmJg#ccbM%g09LK%HWXe04xaKGZLyUeT$<*{l3?s#?<@+rH5x
zZx;$%@;j4i0TvT@4hv@iJAogo$l8x31}?PbptxPcQq2D}25JR9Y<R_9nltYr70k6f
z2$>Bdn$l%OPN%p6hW#dox<vmh5I5>!dBe+X7T+68aW#t8B(wr$i7i2mW$kdrjIgAh
z3kz9<t>GUDNIy>mkQq8v@JDiO7*vo&nFUCtQ*soA>Z}y{7cbfNimEpzGFjDjjqtuY
zOW!YbLE2<XdL&@&%6ChcUjcc&^t9TLYE`I#i~YQQN3!0Ra-fwRGa5ntJ+cYka>u1}
zOd(K{zm(hrRXjtHqCLn$N_+Dni8oYcFu3?m2ouz6E-lD@`yt7yk^231IwKvCZ<~B`
zoo_F1^vt6Zh;+iSZ|)^d+9mHQ#l3=>WpJ-XBxz<~FzW;|xnVxr7LxlR0s1f<bCJM%
zKL^w3=oyi(IL0$xz3)9&N>5+Dzhag3!II!!!UYqwgn}@28f|CkWjNcvegCOd`+7GV
zC{J%N{Pr7JC-J*24e*!#<hP?PO|P?ODG$F$58ckQx#4ITnA__-uVVdhXqZ+3jeMEZ
zEY~gjc+gDd|Dk-^HC!!PEoC$sasQ96gAKQ7d%Z_QBL9yf0-r>$nQ(!4tx)_hhK2*0
zf&tS4-1tv2B7bFd5e_EcT+%|bMer|YbWsB_(>tbMa{o!HfkPZ^$74qX0pcnkneq!2
z%aWHPput$Cnkjh2u+*!HmpoO74U=;CQbat2>V9;28BAjJP;zhL(4qVDx9w+aiD4&|
z1LH?-8`Yhlb?EAS&BLkNQS<leYS$4(MN)V0%P~H){JHDTrAf=FyD=V*fJO8VU3MGQ
zSB%_}2Co=w4!l2}lU!dP)n}zeFMT-yRX(vbo%aOEN_U4eR<5fzN`%ag!;~)|+Gv2l
z)2RtW7vfL-D%AB@Ge@}0ll3&hzcJ!30iE#PRoy?0_mb$o9VF<fx#wRenB+m~RWMoE
zD9;rU9`#y2`>?1u#nwIdpjgs@_6KE%^S9`!A4htlnFgKWy63m=W~tDi4P^6l`P|5o
z?xux~0fJs?uS0M1bVe~r(xC#!d``^nmBw$-q%2TnQWA$a(0&ad8`sZ)Qz>o$Y$8m`
zd}LTJFH!oey(ArSE{U$SA5c_GPkT`(U}_P?fBdZrS8sP_6|I~?^)+qno@Dhw60oj6
z^Ua?;uVmTI+JHLOkXlodmz0WL3C{Bi?EjiSM}sxbFvms{prn7(0P`(m*}t3~rlF{9
z<?QX}nk8d9dXnv|i$Gc(6&$W7hk(DAMo-h5OnE!;Spm)tKh&YhcIuHw-cf?ukL&tn
z2yv?*XW6eXhZb8UY)rC6orkh%<$$@IiYv>dXYKkS%jY??w|95vN$=y@q(byF=sE;k
z!T>(0h>sRf3L+~00`GO_FFU~}S4!Ofe(MKwSrsW2Zp>DwcO%lBCY@ZmCoKj~z!9;X
zN9_YS#a~>8(1FG;s@VtB+E{KCZ(s0e<&jFGGDrlH889zM^d(ITZET(74gDmbLFl#P
zuEXQ<w<>ZcX*3ny=Wflv#f*;o*r`X6HA6%Oee;$I97bD{Gv~$3?x9Ng_-l8P!3}F#
zFf2U(QPui^hp;cq%AGxaOYt>=P$19NJAU5Q$3C>QV+moj^xwPOU${@73IPj^p?cN8
zv80o{3NyJ$NO1(u+~;deEXYzaShDs%s*h=by-du$?#}PummV(G5Z3P^B$$fN9QPtw
zDu=WPNnTm;M)cmC#Hxq-diF|G^Upk=uYY3XHgL>xf&d2^n%^tV5yp#IADy9?lT#%t
zT!V`O4M?Sf3_MQAZr7yZt1ka{P{GxF>^AL+pgk#;3AZEr?KK?exq2(@Y3*Nq#DeI-
zgPxw<{N@@oZ2ixc#@uV@G#D;E(ba?mNp+?cZ}8b}5HCR65gLGO3?fH|n{14~-b2@}
z^#ezWzXt9V&+}~4s$-^sG?ScTxQVa64_W{Xv?98&2lky94&(bdfAli=NsqdVEE&)S
zG@aGmFlKbNCCgD9?Y=cA!~4L#oZ8xbCOPD2ToSjv=vTn~-Pt46%v1YixmknoaIRW}
zbp{5#JNp1K=4UB6)g|fqkxZ>_C!;qcjO>xQ6vAI}kF=?nDs0`-v|l6bgr}*=1){cJ
z@=<Wb+ZLl{OBtH)YZH!A-OlgpaS-!FW_}vOH1&SdXXc0T4i}Sh=p&})sixYtALtek
z5T+xD9Yr|se#|Yyc1~l4SaX&w)`BeE1R~ghOKMYs1Xm3eRQScIRzHI~S=sEDktR(-
zBmdTW8j&DTwnc93PvDR9hSv0x+)&AIJXQHNeH=YKed$D}^jk3R6j-9B_va7PSGj;f
z?IHtsS>=JF@(k0N8iR>ZMdeBxzs^;DsD{|QUvQao(p)*Nc%hZJnK?TD5~V!W3vc_;
zmyZ=`UEGPeGR#+9E*)11akKj1XS>|<-@`|PQGaH~NUF6cO$HkVn+nZr6=xqK*QmM)
zHuqtUbf!QAM{4>|ZFIv%NVcNbd(_XGP2gN6hA6A_J|xyf0$#Mwc5oqVm-WEI^vrR<
z!QDPz`eiZwcTKHXo)7%AA#f$x5wKoC(@(g)D-IaO<J~);GQDyf5y*gR2?6fP^tt)O
zv^tkEDW(*bCGK~^>z{CCTH=c^RmQcQzt3arVRlZI0<~|YptcN@M?ojl`|J8A4|+D!
zw}+!`7w-s^a)karrX>{)?s?X#{Xd6l*V*_`&)_wXR+;u>2oA0x@5oxnaa4Q0^83`*
z)p<T{rQXY+!+faZE7cz2EmX*V-Ai@;{TBjY6tWqz$w1#7Bq7}u66p6SIL+qbZk&Db
z4ME1lZO%Llz*4L{C1~!r(+aMhLfJs_NvW#tyeV3UGQ*n_hQ)4s{03rf2U9S41Zh(*
zE#2YiQezc+w#N$lY#lcN0uldwYBUuTRc6|;zUVe!=Z)bz1@T=7-q6o|M7}C>gU2dR
zV+BG_2*|tCx6Mf=(4ZW1_q^i@y{X?y;$5eb@I7r~uP9ls#*av1B9wccnCcIHVcoFt
z`hdQqDOX^*^3okwy;X?hfEOv1i~GJ&oyFSvy8BHUT;Q=A^mf=gRVo7jEAi;t^wx!v
zF*C+6hU&Pjurv7dzAeE*$h?Y@J7RIYR3k+DtVveFp}q6W`+Z#>DJ)<Po_&h<vHD^C
zTB>u_DPuN6cYBzBC;lkUszIdBYeeYq%MgyWmI`iQ%*jf(6r7?)iJF7u^_CP$_x0$m
zws+U3<PgPChXSQ3zK8S^mdyEgm0|N`LgPT6iwlNknYPq`Wf8o84l#rdAY&8K`_C5e
zY<gNJD%0MC>zMM@=d0+CymFJ&-~{}fW96FK7yXVeUHJI3_vlA$p(X|BO?ab*aqiCh
zG;&-H-cUlTEGB+e-3x5keXr$d*46KvtlXN~fUa1Qx#f!v6LUM+A=;r9AVid->=WN_
zPXra*p5rl=%?F;V3uWI3rW4g_YDv9&4+;nj{@uNp;)5gOgTO&mbzeI*_bzKUWGJA@
zZz4<MaxE=&q9*}D<BU;Z$@Anjd9(iVnlj}h`xd~j%>33=W<2_(6caP*7I%l|2&9*I
zh_HDeB#{nzV0YLpCagv!CN6U$%|r;+HxlmIcfEkM!hi3^xNh_i(ckzITgF$YLgzJ6
zE@nEd=m9xciZqQ94|}w29@lhf$}LQA<5z410_6Y`2-A2TJ*1)(bzpd(3?|x2L>d?y
zr1((3Up&Z7W#4j-JAo6o(`T{8UliL4jUUq}GV0!bq*0-QyS1owVyM;)^BCO3?W^2{
zi#6@{UFRx=lmynq4}I+159vyP-mx#SNli&gxBBGCW-9H;J|HB%e?D?Yaqt0bfnP}W
zS58~@fdg@sIRR_4Z8<*F{`@v@c|(mH=~l#3(ypmJ)<rtZq^3rwdYN|#8u4+L4Ht)Q
z#6!e^;US>Mk#bHwgW&A%lP_wo8Hy{2r}k{^R-3(V!5>zCHpu*9$wzp4_52T|fZBZw
z(5bncR*qG=j5&WnC?P4`2lnn`Kd%K-x{s4sj<JC*k1antYtl@wk@BU3+CELUx~<%-
z&OuGK6ma_AOOJw757;fdZE%{(qQ1~p$1a4dT84RT#mMZs!t2OJI;yMwz6!*Vjk8?B
zB|!1U8*W~Pq1N?}zRwfW?Eb-um(rWYiqGTMu&Z2q{w0*(z1}MN3e7oOC^LSIVv-3G
z6HR}i9KPns6pDefoN3Kj3Q#FhzY5Eel5Ts{W?;nT3Wt9wMhp1Jkreag3cm@5`c14Z
ze^z-3rZc!gloY;0{Z#&IXbO$sHzSv)3oBX4m{+cxlG_98PYx(LLu6t&FUc6Iv9(eB
zF8D3~@Yl?JRrth)Xmfnd`^t!|e;(YDf9`EM*%0SLz1Jn)HJU#sfobb@h%NDqb`hNl
zA*9$GC*V-IxZ?Q^)@STh1}x(f*znUXsyOFaH+aA*)gTZadCt9Z#c!YV{}v6(D%BR^
zu=9kETj2T>)CJ}<{muSC;`qvOwzZZ=qZRm%RC?Ho^v_~dvGVW<z6$OC1oyr8o??4*
zqeLs`dWT51J*gHo?9FE9r_tl!zCui^MW}x4*mxMf<9tOmC=<v3702udUA&+O=_zfs
zQpQj5#-^?x`3LLuQTX$KL!ETYC@Upz@RUG|{I*@PCQA<v&ebWY5L0nqzJ*ONfi^k2
zCeNkcr4so)ap*0*I_wtW^<ofCCXkBSK*rSdI)0=`+OPiea%Gy*1GcoZdo)o31_>58
zHlKZzaI@##TebO{&@n#hn`@!dc2uh{nx*mGQ)9{4)1UjCHI<>4b3vZysJu2r=lhm$
zf6?uQ5f&lls<Cf-_e?)(hs*3TrvQj{zQJIj^O@6|uV8mCv*Gn+!<<XTb?EZyt%K}=
zaqRz@>A&Yr{;fDtB57i<T;kK0g5FicZvc8V-E*hNJjC%=Y-lo;TZ0!<TOmF_Z1Q(j
z^u#jHzu(aCNn$uHRjHl(oZ2ug>Sksx1iWw&ZOP2)DJj^MayI>{@|OSp#FO<*m}I{q
zf$r&`;=Epwx6W%-L~oML_1|QQaKlsv?~81`r;E_fa9Nu>jCtyhzc4Gfc;`yy#~wE2
z%`9GaUqa{SvXvK4(|<kR20mql;%)s#zOB#z!4Oys;MZ^Gby&cEV@S2)*qoYRS+14N
zQ@P}o$|Oy2gY6psF}%;7C|t^ZAZqJffVTm@cIJ;6d38H77%viFODVB*sPPimAh<Hg
zJlL%FjJ`xCRVby7rn`+xlf#%ziUyHRR-P7z99Fs`D=Vc3(&r_WjYVj^A*aI$*tN;S
z1n*BW?O59CY;VQU3V)!)6T@Ql+ZS|E-JXqQLtFCdb+;;r5{ATz1#g>NhS$ZZH*w+0
z3XxQQ>P^4{j&XK+o|<Y<7IVwQZ2njn`+WF7VV(ba$h0E_^l;LmKGpy&ezgr#C@Z;|
zKtd5ovODhhv<bTBL+%JbOe&O}p5Y@I)rN9F9=x*>sFmxP&cNfl*29r_)05F*=5d{I
z>v5L~$&N&Oou*B)9I{<{i`0|iOi`1(8sL8Ui6dRlfBmyV>Z<o@?P@C$!nZtuaH_tm
z`H9>_0i9Y7Xn)Sl){gMPcfqo8?QlCdk$x+r3O2nmO=x|cI~T+Mig_j2eCH4A<R;UV
z1TVv1|0nWm%gNJyiG#zd{Tl71xmnvCPj>>j1@Bm`vK^obCd~zclXbGU+xiE{2nw)*
zCRID_d7Dc5)E95q-Pnl-5p^_ulj?Cc>XjM+@X-2Uw(-WF%??ov|9ixiy^}%)ri(ja
zY9Re_$DRv>joco@kRyE*E1K*EO=R#d^JixHu;Y&7*37tWC;l#h;Jv+iDuMCWbmYAU
zu42b13nJd8!ZU3mAEcRA?=aHEI?1Nmwip$nEme|*>Y8>8wj|b1Q|QD{T)lcpKQQyV
zqh;mYi_340eSA@SuO@lVS1je$_5wdPM^=06L4p8A5AO8Ol=CCj`+)4@7gIf4UYJsZ
zH-{C4=iNoSNn+K`*T_7Uw4l%|;gKFoF#T|x%2sxsa1r{2)w)2r+FfY7fuJj(8&vOW
zuu&n?v4!7v^u{+7P>j3Y{G;iopn+}Yg=2rBH^x84#qLc@;2xEU>#ARZ+QlqJ+xqeS
z{pybQ(G*+AB+FM1QK!+4vL@PU$gW*h1upzyXLU`wWBOu_!mB9jhwwrk+f$`pO^fZ`
zNAvfsLxQdAD|@SRNxT9jwMBTdwKLIX!>oI1lLakNK|a5L^Yn5d<G~wWnwxE}9Xdje
z5!jxCrVUTm-<HouEcrRRhhw|XmEEqdTB?4#zX|UA>)rMUJ9G0747_Jk@#^mArXK62
z3|VhYWcP*7dyBS&W|)n~aVzszPtvwcW?C&w{#@NPJf*4M5Y=zm(F!ff+anWo%=T;I
z6i5HdPOpejsdY0c!kZO?`0nLMcJhn+2lvY6w_un|0K2E%L86SAG=XKC_uE>w;jJ<u
zPG7Yfm~GYDnaeMuD?v$yck;SN$HcCZ?{FtEh|uLJ+4uDyBpCD`QJHc@#F>SM(J@Jc
zW(pwJ!#)*TK|d-iE(=NBv^hBub^VBTl}T@R&4`3Uv&Z;(3*L9@wS%jon><CUUD8V_
zvCqC`#|ry}uaQx!K1K9elJT34005y{g+4hbNI<w1rxp2!kUntLejx3@ij~MNw7En!
zJatgE(mgBVfAH$TJx(es;zDmpw;JuGg0wqM<8l6c6-R4a6Zw8U{5_A7Y0~cf{f4J7
z=p-OX`@<VLFs@F_nI{m?v7vHaROh@uc<hkzh?t9u&k~Za95WoA(-w^ExnH8aih_&+
zcwBPUk7rk}ULK@$FaUmg-0ZZmol;z%1j^}m)Bah8|N5p#N9dYWq!2tzr<9cj!JR4+
z(762mon|;e3(LJj*H#YQ^b!+7y;L)zW&r+U?4-|O1`JxdS|YCj=>R%G&o)TqyTNqr
z#-B<#2!@yI&N~{lgb!77<}kFb_};Z-Zq1YwgGEMC?gm?OTAezw-SYlOm=?t6LpmE^
zRTFb<;3Plk<(fxYcye0~We=g;C-~d>Pmimc(aBVay9CkyGpao2(#$&MpCJ^aA$)_d
z>&Nr;ZQUMI>6Q63qAzQRdpCJriT1Bor0s`m0`ouP?~8n>KQFP4&Dukd|L6({Ye{FO
zS{%-~gq4Lk^AYBZlG($og%0wI$-x)DpAn$NBRnS@{?^YfJv*PLIW}Sa5oaH_M+Y`B
zCh@-A@^*ydXk+)kr_w*bZ!>&jy#$g7Wn_+I>sn5<Dpv;V^cUY5alMV*LNT+0>|$m|
zqG*KTO)3f4rlfAE_@|W#s$>4tT1&<<v}VyFd(*B{H_yo;**^Kgb&qu+90~!^Gwd=Z
zWCUj|)DYm}i=nh;`=S@@u!rKJ$-lw<c0oJ)#^0wKdHI_Xrc&QQI9Qu6uJ@bwt>Dn+
z1=OT>Z2QIO8QW(eL0`mI+6R~Le<+Fo0;^dc#|$E8pYN-%-DhJp`yP`~6i4pjrm?0N
z__n4vxM9Zk^xVc=D>!3pZ>dG{!p-vogZ0uw#%aQ4e-OpA5Y<p4?U@hLZTi1<7Rm4L
z&_z(orqKTX?kxmQHCoWJZd-!qyt2D}A<+;QsSwk@Y&GJ>8;IcSGBti*IAXk<TV+~&
z<Lk6y2`Vrx7Sq3{`y0gTHhR?%jSPP{I2Fp?|2HiP5}8+~2aXxev36bf5@gPJ<d_cA
zfZilSQqif21z%e37eGh}$eD$EkZ@o-65^(&AM|Ol&V$|^60usRQ@{Q8eL9qKpnsoA
znS^x~hPzG6A^!a(rZO854(0vR82+f{!&d@LoS8ui1$>j;Ui>SH9~c3{DsJMWW^OVH
zvs`URS>Z6fUw~l>S|x3i1vvfXjuCTSOJ&k2Z5+oWUlzcZ%!2e4&o7%$yX3*GQw2JO
zu<f~BEAkHo<3_BOT<;qVuv?;~gt3hpG2GcxifVwlk$$Ktn4C}GASwM-rjhCITW!un
zw??e{?7tu(O%qg)&)ILMmJDN7_X?Q1CS3lIydFKtrT7h4U%XtK4=Hd^D4*nO8fyU6
z0&65fsEQXA3Uq=1>k{wH76#jg&o;$QQRFbkXH_^Ri~3jW9}Q*KVf}_*Wp9Lnz#eA-
zr&qj=h>O_!!*Uzi-kI$sgFD^rhE(mbMA5h`mI8$(Z)@=tRFr4YGgRnnzU_Pebn;A*
zGXX*F@xM>Cn>`GFV7Dg^)ZTlSCr7Z$=Q(iy_QdLYi4qD=`2DwK&pa$29*!L3F<uf)
z&Y+sK+|&@#>asW$QhIyO)zZYjYe$arN!=oJdjIVJ5LLhhiSq_6>iK+Fwd~gm5Si)A
zKfi~uERyA?ymNbz?$5?~&XfEDJc$~o3yrj<P*drMK=F*hP7=NPZ0}$6F29J3n8V}C
z%Y!<>APobrhj)4p$RKkN2`wc%q1$lf<55IEm9$Ob->oA)RxEVlTngfwteZUh={&(p
zeDSTvdXH#^S+pTKxoLT_XX5+ip_-W6snS1O#dvGDP#s2;T)HO=nFmjl5=XQp)pF)+
z>HeA>;nzOTVWH?U2U_&t71%59KyT1(GY@8Gvt)>S%D2M(@mJoZ2DD`Nd|wf7^Tkpt
z-JIedQeAP&G2JT0L^;5+1hg>D1z-%Ka1VZzt)=lqAiaOR18<tbw0!1-Sk>>fXAGke
zWNmG;R^Nop!)|;zEuDfBzgwdqWwF;TYc`zw{O}|Hu*4G+14v<dgc8Q7cFz8dMon~D
zyf2m8QLz=K2Pml@ywnaa1rEXban*pQX|s7J5-rcP(g^R>n!!Y0AfDEqZ1b*d0$j>P
z459N3Hk|}940X@HTM6TcL%59BPw&?X%N#n6ALB)q0hRKCcVI$Rc!*zmmc$@FQyvEH
z^!NYI0x(>sk^lo}p)${}SUrO+@K!&@$Oe1TREyJ5XCDxijFtEkBXSK4mI8B?J_Y}U
za6#45PR;gL6<qmixfvaw@`*qN3<Yv@`5)u}&?&vd4{TY#V;Y-#Ima~S26mnBLq2}~
zulkP)U$;A%+{lW)0WAv7)6Wf}$lh=BcrjoI8LFeAO8c7M#}cZ1o@_*^+k9G}lb4m>
zb7=D*)<NmUs1cO%el1`M=u^ey#9@x<e^Vb2FD&{~3^*MjC5J&!XvxUqb+eCIbP-^0
zJfO4d&#n-$#sZS2^qO~JGD@S>r)Qn8<4#2(FX4f(R#zyrA`SGKx)S%DAy0h48j3eu
zL4{+nzWMFo6!$K#|D4gTC}Kzf7fjO8HclKFflVr3S|j_T!ciXo7e2ni8cJ}Dnqr^%
zfR@v<thMwfaGfthH#oZV{O*~(fBDMGA@AvmVfOdL+5OS?e$K7c)96G#i4I}x-@lV$
zPEN*HeN2)ZU08B>AT8}$9q^a0>&N^w85T>fU(Ee~KurC0_VHl0wNWLWc4=elx+(RT
zkR3w$0SAsYH*;wG^gm2%{Rgz`9GDxp^CvDz+|4{8yMA>5(Kx-gZ?pnFwjw-qV*F*#
z81Lw61z~@*iUfr-d-uy*xgUn6_Rr3?7X|}P+b`~on1B1yck#%h9&x9G4HFQ#=gl*j
z3`71}C$Ru;%}d9fxv|12qA${#o88}O{OmVyg4W$#AXk@oT9RK!mq{t(C@#wjDdSHE
z)TZIzenj==_5XUmIN{c`<p#?W$%hu&0W%7IxTno#*qRrjf7mum4J<ah(hs<bxQ^c4
z7hX>*;{MI97@BUIimK6}^HAY3L}?1fMOPw2Imh#e0ewrR#rnS1i@x%iMx4?;Tu<D?
zPyxjtigcqS>JTNmi6|jjC7s~qAZHWl7Nd?osav`ZBgG9bPqBCyBs|Svr<|DDpLXs}
z%D!gGeul@eeJM)F+GkeNZU7&$fYq4sW`(KZCh!4^`lEMOcA(;;?0>95^;<Y11ujzq
z>;hZ#YtK1_S2yQIYn>SYjB_Hb+)nHQJ2XF0`>uLlY79;lIr0S7xcEtmGF8<BXMA*W
zz0SJI9TCa}J7nGn3r)I<`y#t5_}c0PWHp92liK=2<(YBL^Ah{i>^}7NupP06n{AQs
z4IylZ|6IKBp{qm6bZk4&5qnG`<VPs)Z@R>0DyyvG9`v$+c=r8HJ&vze`w*8Mg?_DN
z%)5^<B7gte1AQK(Fu~(bnvbRP^WL|Zc<gX0FFQG-uMg+G+w&BB<Yh?VK$sl%h%CH)
zL7>#|OBAGpbJe)LrGVk2k|UQF5dPa{<Kw2A#Lha#qq<~EWTuU{-Mxj6!zlhm7d9&H
z*5nu*YNzGMAN(=>h~faY$+atJ7qKH)p??gd`}k~9{B*Y}Jji~t#tOh!*h)wNOdlIm
z)30!@xYCKd+SxJd(QGR1@0@zdvZP}7Moo^9XK0nMNT`@{nBK4^)QnYQ<Mac2@g)_Y
zzVKv8>niy!wyjaO=VSIAcRI1E!v_ujUMI!VE}Qu3QlOcJ#LWuT;Smu{;!l_-j!AEt
zq`<xkWcBp!<)(N?dLGDrcwYhM)xtHWlT5~`;?a2NHSL4)*My`1Px>22$&1vG^}wZ~
zJF0N8EPP<*f5JOY@sOe8sI@;0)RscyKl>kyLWS(EF;)eGi1VAB&iPYw&7yTmW^$bk
zc)tB6bzckl6o5P3GyJ2E%NLH|IA`(85&j)}G7%wzc0Zp$T?el@tD%lRI<@=Gzy4kl
zOZ--KRxWn>c)35e{kj`=U88iTJcz;WG5?S2A#Fth)yYCmBlv91I&ahaQBf*0Y$ytn
zOR`&%O&O>mDVCAbqAj!p5t%u$x?LNv;#vvbTeisIoz`3Q^fxppEA(VS$AN~&i1&Cl
z(VpklXx-)N>YQ={terO}uvva@V~OTVu;dV$kW}>~2R;(`w~S=$llGL5+D<2Ychml2
zRkqy6;>P1Z>DG&h7DfE}cmN3_h^4j(C>!>cRtp!Ymtyl!LDZiw7a-sk@Sz4}Gh7xn
z+=uQ45eeBgntf0909${9L%E9{Y-P&C681$^$JW?ToiW@rES~<It@%f8HqyUXp0uxC
z5V2mqm4WS%hc%hLbw<;w`FtUOQLD{Xq-GSVkyRfKP_5-db_&nt9R5Sg^LB|ZzauPP
zqvQNRyd;>xvRDwrkmY^k$grxK_Fc=+Zld9v=ui9)O`jme8w~@>%FGwWtG%T*Iq?Mt
z6?=en4UGJ#;8E%`p?=2pvGkfZtINIE1vAb6Z3b^S`B%<w*I<f(w|Em62NnfFBFTe1
zcEL_29KB@1$97ADj@0RzR-FtN_uJ?{?#o*x;Y={_bYpyOlVW69_cY_=iLYwjwfoE$
z{|SF1Sknp*5WK~aIK)LKw>OXkj+ud`{Eye>;4^tM;Y};+x_7^eDY$h)POE5es5x>J
zwR;FV$x`ZpkupRWaGbCr0$YCAuWkG)rTBVb#KC0oF<G1HU{{iNzZCPn{zevdIhlv?
zltMT^qEguEFlE>yvZ3Qo!1L%&Kh%!Uja_9WO?b|mxFF>elg|OK{(KnSzFnxt{G>)b
zboC=2;tDr^nsz#ZE^?FZB!{NGE;G3@`H-R_SH(p=!bkfDmPdEF4oS%{va8%3&SSbv
zMxqCC(7qM7d`Wq^*F(4h#Mlx^q`~IOnmmAr1q&;Lz{SOS9D=s6akN?Tq{E@Cy7?Jo
zkyz~EqfnU!V!bjfj0bWv{N65j1^x6ZTBSF}mt-Gm^4U|~pkmV()fM+5l{Eys=p~`e
z>pQEzlSh-c;d+za1dj#Byu6%=(o&LdukJggNpE7bnBrV@x*&ZVO^bjAIT|M43JDHv
z)Q33EaA$J*_8{~X9A;VKqs0rA9<2_PowNyuCgqF9q8%%-j}#8Arw8fXdMAEHF$^U#
zU8#3X8nN~WlNgKUq|2{T=UmfI8Z<|@6V3}|F^4)drrGqngTFb~{&e!;PV+sTklk1~
zj!5zLuUB3%Mhtaq>*m#C-<CV(QM2~SvmSR|+*7Mss;q`0fL%@N2|ZQ_LwERN09%A)
z_k_qk!R%AarqBg0Kakh1xarT}RDw)-%%c_n<BX)uc4J;&&Po!uvXSchnslGp@pZ^2
zK?@({?{%S|za>eA*J2_17Ui!<P*j$;5qN|6*4CKTtfe$Z{ljH@D=#6a&YD2OFAdvc
z+-&c>*TA5h)Wa|ItFGT41>fSFpsH}arN;vQV)abB>pG>gnW<#ik#0(5jps{^OK&y(
z+JOEC?j%~~?NGW(YaKjlKWAymAQ|i&F7uP|n@TA*i+7c3FbCnXf}%N!!sYI~@TRr>
z%qsds?bcpjL}+L4mY%zG&7K|GUl!ZXqJpeA$x`B#1o*o#nwL==$X_iF6E~$~_O|jN
z$)o7Hh~S7fvkUd4UU%4O4R2q^0ob=HP?g<M25i_zP&*fhMti?_oJghgy3?(<5CK*>
zJ;8b(fDFg@J_6Lqp-74|DsDiA#6^Y}7*)hAM_L-k7OU2;eydFwz|by*rWOG45v16n
zm6c{Vd1c!{ukJPOlJ;BC2|*wf=G1B2C3t&ow30V#v^o%2(=_kWt%@OMtl-o%4=ON(
zdVUcbc+6c5Hl_3$XTrI95xta@+byWg7ku0Xc@SAhtO;J2m-H>lT72LWml}7&&8_5}
z6Rb-BaDqjnMKG~;osN~4C~2b|yJ<gC#>{bES)7a75#ZJ$WE+Z`fEj5e|I<qDf474p
zhw6YfKLZV!`6GZETpB^?d%SZY@5oqV=J15KAw7Aw%1@#oHq)M{_q+tc9zMZ*?hxrJ
zgbS0*Fs^pBtf2VT^i;ewUzV{y-_jX~42VG{K)2YNMEw+xw~;zfD~W1|)TgjRfTcUn
z*{{n)PA2IwSfpUS8@ID+byOAhbwEO{fkE7RYZxb9$}g6Mmq5GY__XLWU7Pq`chRlR
zPg=N;MoLTdT0AR<TYBLgi<pE;5)_?Oe^10{+ffK9XKrlYgW8fV=e$TmUU`?vU>g=$
z3=(cOuoNDQ{e+UjjB_8#^MPT5Y7hHoG)9&=LfEIcZ}f`ZhCHyjs}%B)E_%47_2m8c
znMn2bHIYoB$0zp{0@7(Qq8%c<E(LMe`W4DG3AMU!iASN|9UT!24hz7#drks1&3MM<
z@PvGZsFw@PL}`DjSFmgy>lCM-J#~70MN$qG-YXwV&2jAl!0Gtc*R^>=ofMbuGcQ(a
z{6#h^tdJjJ;C0w_6%HzN9vFleICQcUCbKb}9J%9wrx4Mv;7nUpf7sw~eS<*gZ#rz*
z!t3Ia8tV@4PSI4F&IaxXAf6^t@(8=j!ian>bo!px1X8<4JND(yP-f-Qv!QjHQLN~~
z1XACvY;o_(;~T^P7}uirP66WO5_Bdf7H=GH^iIrt_CNdWk6Yq@?jMEFv00_DajLb9
z>(GBeMRzz3n(0i9d3#V!i|*9h$mSUTJ1~pg$B#^#O1);fIKN`|2g0e?JB#u5FZNbR
zUvU%d2w|srVbE(!p4MZ<&SWnrVehfLLQ66R!zh;kEln8OFD%6FQ`MC_jNcbV#bjrJ
zX-65*t3jey+WWmp#uQ7mzB$mdF2$8E=`|L9DgTxG`vBHo_+E_XeMgUtX6VX&k@|K5
zBDJg%)nL%fVYd1VrWodljd+$mr3$21wqE4eT!x-@T!HGxoIuBOz(b17_lb+9VS7{C
zF;0q2s}+e2;Lgy&g&~}sI_4&LLtEbWv*a4<rt#cA@@cnzmk%Vf5$R0mES`Z{M9Jh;
z)Q#80jV(kVDnZM;|EW@>g}Z(Yu*_t;DRX{9aarF`gB>eQsml~(I{xmHtcz4vNe7x^
zgZ%oVA|*Aa<M?KKWY^h|WvSH{3FwSlnu{kXh+&b6e*Z$frzy}0EHV(3;(E9?>E^XE
z-=5dJ3m!4tz)G~g$^HelpX5*6ENHa}0}x#9B6pf3D(6g>o}8gN>N&e7fx1YaqvNvX
z0?(J6=!L8@lC@tPUDgM#_j0LdU~9o6ynt7qEL-B!`VoiiAX<$I+d73izg$V~a>71F
z2O0>Wk>MV>ADUhHE7oa71a<}snHIZ*vm!QnXQc5uCjA|YQn+;JNdxh|FqyzKGf1e^
zS)UBEni-#KRssT#`hc?cTU=JtwHQ+_Dh3Og?nhNBP>D#4PNdn5mRu8W&<n}H22C`x
znDN6g(PdRWRh4A=FVi9>1K(N3NbSh_w7u`xQ3o|$YsWH6;%MEbv}Dx${N6j7UdB3B
zr!9f1Bz0y2UzKj)S{$f+4mv{LF$&si*GH)kaEr8qj)I?ZZ^T9!_djY%NWVu`f9js+
zL;Oh-`m;mx+q7rZ0~V-P&s?;??9sdx5v|RFtXBD~hOS7xjiG5iPY>SR2t-l@53kt%
z3Y5olxwyhbbKO9T^AnHQSuoYM=7^<`u}0R{5i}hvf1uho+u(V`60r|Kr|<$1`vgsJ
z^=b#DQBckKI2Uz$@G$Pm1jF8W*v0RzOW<ZsqFV2SfK<Ai2Z>+V{yDI_#?OD^FCa{+
z-K5<BuJ6fflZYLcx60<7>w*f>4Bgo@pZ1{YZ}bH{Jd%!Bp~n=2ThQ4RSXOf=Ve*c&
zHAz`mzQOBuDkYxypR(rf3V(5wkt^v0oOtM{3p2Nr#l5^spp%O|nxw2ja~Ctrtr4=*
z+sAx4SCRW+HIj$bFW{@`7PwP2)QoM|%2=Drt*`L(@9yG0uX(QN07JK^gdSR{PEYLU
z;A{0l8vYtG(G=B)h;n2K$$mqWeZvSOt}5vl2{Ml4$qc0zEZ4nhs!2<aMBO%#X;6vE
zITiX)7IT(g9W2AUiA07ZYZ?d~Ng<aTUu`IWuy6b)gGca><$(P%_k2kA=v-M_8N96U
zr?TQ$3=*X4aIs!1U+^gSwDnDgta4VVyJyQkO?@ZNLvs?Jgv8@kEpiJ&$gQpN7_&n8
zc0gdmx>f<|nP#EcA%@X932GXfN4NhfvKez#eu5e^o6`}3fD$&<RyMA4?o`Z$n0n<+
z=r9?84yQY(W&&K?jWKLGjc+^?KKU=6bnhNQGGb~T9Uh3bn5@3lNI&+TIm<IR;>Tos
zRJbxN>wCSk1`2qBJ#Ake55DK$)r1A-bK2dI1fX4JO)(I-o^Oo(si>nD3SZPKFQ%0X
zjzp}Yf~%VQ^F2a0g<;E{A8G#Ybcd8qe@mgW=F!w*5#_BOC-Z6SyT_GB?jzbTCB@Sb
zOPi3q2LrPGv0twk<KI&nxxK8T1N&nWBG#Sf4!P6(?La|fJ>R+HAtYLPKD-~)>HdGw
z-h=ED9>_X?38iJlsIgxAe8rMV(-BKvY~`qm=0Oh_A!wfYRAz~aVr<q;WXd*x%UD8T
zE&Vj$f`BO-5=;P_+++m}wX9D9H38eHz47zQ5r<EwtXvR2E>v<BZ1q>E3)lG8AjsQb
z9F-(S%#iIN7LrP7l$SlsX)B3=OKN~c7-=i+NwQrh*`9<1uyu>pvrzS|(<IX=BJ_oi
zpnj%%H%4>?^i<vG*0p;WbZn4!9vCEsXhT=&jXmLB{n-qozH*a;6&2^^q(5FjF@~R6
ztuo8jjXZx=@vnQ*LT{S;-ult8Xvoe*pKO_ovRZK*kdsZd1M?Lh&ioE7{?Y0WEw;TK
z6{fshKn30=@SE9(cdiDA)>4eWMu-1xvN<B#nVs85!4kL0J5>~(<!+E5JQF-OV8u)!
z^%X=xl|AK7FNtrGA)7DU@MYvGNcw96hq7Qb$nwG*AEEjQQXD&VhL9dye;(ZFk$PwN
z1WhV5Q1b?3pG8Q;#j<P_!ir}L4<E2=&a1Emj!%)81QPuzL3p!lZ9D-<JVT?idq4dL
zKrC8l22`kg=65WF7%B-k<&WjB)SXZ|!7Q&7Ruz7|(WY9E!U_*MxAUN7Dv`GOinCI+
z!d)|+?pRgeUiCLhwtQ5DAv}o7E?rIi+jQqNu#k=IW!dk^c=1`D&bCy9bAQqK1g2Qy
zE<+z)ASpR}$RxZJJ-4{T$61?&!q4V;72<OCnA|Qfpp1||n11M)G@>XgTmRUKFh0zC
zoZhsXAOO5lXF@_VgtN}IrS#*@U_>lmSvME2(eT7p`4}!4=lSduVQO5EltWyIkNx_=
zK}<x9v^JewLZX_LbkC~yx#^cI+1+VOiS|?D?jz~<mk-Ni*Vz@;D@Q!G$fNTR$q)n~
z^=dyhPc6pPj4;d9mC)7TM(f*81R|HGhN?QcPAtGB4xXW0)TB>e=6Y((juX6((OunH
zxO#yAt=xL|I7b7m?Ge|}?I{^fhuxlqoAw6vF2=oMWv7#S`uqR9MQ*`Lt2tGeFwOfM
z6A9x>Dw2cpL&ZIDHiEl>g9ERv(a_<IOhvv^dY#HxE8Tr<9N)=b;jz|KH*LR_khicu
zv%owIwz%-8T-CNEd?>;W`|Lg53#B3vLi=X)?z!6*!dc~!P&o`HT;C*zN-GI$lQ9G@
zTaaDwFhl&JLDTvYbs2NmVt+K-N%joP(|gv?+;2vcW^j&rF;>`fm^0fAt@kOPvCH`H
z5ZS#+N#s;T^!db#7yZ?^rhQlq@j<6;;Za8bWQX9`RbugeuGrPjtLThno9H390_pZn
z%~9yzInN}TfPZf_Fs^BE#KiiiGU+XgD`@3tgkUM<T0X@&<}vjQxa2HTT{+AL%^88G
zBJXKr*{UIDhi5|19FNdoogleZTF&4{JvxUH?5nl(OyntLkp()2-MDxq-$iTe+?TV^
zs?&R;z_y-G>9A5fcBOgV;STmdM5-l1734Tty4_&-m*+%72n@SryAXrpc(0h7f`Nzt
z$J?0g)blSUb)i|*y!^T!7atP~Ts@ptbMm1r&V7Ht>BKS17qw;tI<=eh8)k$BF<RCA
zP6&$n+X6fSqj5F+v#CQ1WQv^Q=waDjIVFF@Q_hXqpE8U8@1aV5{8+`)8`uV|cqv7Q
zC1Fg^YW8G;6nJc|N^$Sqz)Cot;XNaN6&hEr{S9~az;DKBcDw;?il6RnnlDjuq1Rr$
zoz&}=ovrO(ttyb?HE98SJDdv}G|A6|M%&o@rF^}|(y1Ak=D}GhW=!@+mjbkdhc88#
z-cI9qr6)CoAK!;z%9H#U%#}mdu_&S@*|tG0+t#1+6Od$TX%h#!xz`b;=SxGSVrVCc
z!wgiC)?S)D%-ACLitbe$<F4+Mi}_z(!x>B7v@s!3jZq_(JT=QWBcI`@LFSj}Lai94
zk7%cbk9P_yR8B_ukH)rEpV@Z^oR_8+V8zZJx@ghM!_dbsfuALFtykIU|68alKAqsq
zP|mW?W}I2j<L-^#OXm&&N`WmQC;l06h>|_%R)<yz8BW8#@2$@gkdNXx56=ccvu>Q<
zU0xkMOql0oeVqR@+k1~2u{!IJwa6=Az)*kCGUmOB2x|_#+cx!ukJ~nfq>mBLNn*^H
zA<oY)K?OOh9UGgkkHP+OzXlQIsyh7nP?GpGi&yG~Q%a3t`3zF0g)87@1^Y?a0*r4L
zLJxYxsNtuI=*Yrs-4Zr9<#a*G<9TTHG6aG@F(wU*<x2P@0s3h9kyIPtRI~&XXB+!a
z-?Q|NcR#|rN|4ccOG#-3h;)r#Ptu9l8JrcED~HBQc-*`Tng1C%Zh;9}tHcxMO`fvR
zcaMOF5b_|)%*e?cCkij&`*&XG+5bPr-ZChzsOc6aNPyrT2tk5-g8LxBgS!WJcN-vh
z(4fKHEx5b8yH0Q)+y=Ol=Y7BL*L{*(b*iXh;LO>*d-v&Hz1H5J&Fe6_M=}eP{o-IS
z8pI6Vg*i*&y?SdMV0$rvPZDG;f}v&0U?xhuxTreQRW*q8v+*mLyN=#rCjtLqp<Y6k
z3_>IEB{gkK<!%t{&9gsiZDUT)qS#!c?mp9~G}fT#an0FX@dVj$6v3I33uJrPb)kJC
z74>iddMx%6fW+7Rpj+*L3qdLueOm1VbcIhA-G75;R5$rahOFBV$f#UV5w){`2T06i
zkRxsU4;-r;f%Y&si-6xGe;-N8*yHZ##_v2-&OaK^itF0<ww-+O=R*I87E%6}k}Zjf
z*~q7&hSysHTY1oT-UPSyo4LJ1WebY$JVjSn&>fSN-x_wJNu;#3@Sa2^B5JBWB)!}p
zfJtATysF%s-er{B(dM3wWN_0(uGaW&?HxyZmX`7h66$REF(nyz>9ioU3R{I7ql0$#
z;?s?1dUIQ-`L^kpF!!?2$FT}brw6|ZxIpf<6RpO>oTb<5$5&_u_d!u1A~YA@5Qj=*
zbj4y%!9!PjomeO1Z$;JVn>E671M^SZ2fp))8k-EzSg)`5@*C};;8=d+b=`i$<fIUW
z!k|Sig1)9m>2QdN_37BAQ;F0*oZVkkDbk5I2vb<I3HM;qN&ow1Hn#<QT|f?eNW@Gy
z_Iov~?SY+5O*al<dJuGX%h&D?_P^QnHmCH(n73a~8+ZTc(Uwb4_=q;jKDBA^EHXZa
zkCz~iVCVa@4>l!9#Dji7KCM(4>DcHSop~SoZ@~{yb}az)0eQKzcFqrHcp<#Z2BYo(
zL0i=@oBhHu+tCo{_R#T84?T@b>oLPEML61iu1j9yd$r+PFJ@=_Fr`J+T)=554AtiN
z&$#en2vvTe(pd1vhW9Nk6(+hZrAH+#&4<k`H}#1wbGi)Q(fC|d%0jcg;R?Sxc#l0D
z_i}$a;*-o}w>~0sp?87Zo@;D+y4vKl$K^)N0zF?P&3+>4el$Dn-KAdGh0O(5h#ApS
z_nD$MW$6)5R9;kvLq#vztH;xR>`e7@%>}m!pUiW*)&ny?q<-JkRRswN!)r~P!q7dp
zBsk$*^)=MBl}JJVT>r$6<JP2Odwfsn>&t<pK1W0S-G&E06F`k@Q~HsJ4#n;2b#-Mb
z*fOs5?Zkjnz(rM?&GqoZg@F$?*{gqC=8EvVw+9S6LpGFb6k^jHA=vG!`R!dC`h-ZW
zxEl*dup_?`it5OJb$J8M6W}0wY_)srvl4`fwxajHd1_?OX)e8GtSKYj@(KRJyZsSm
zV<xs@fdmLc2;sAINPw60`^Tk);a@^UDp|PZB_rkXeY<unhLh-&#1OR>`-tUxY0z6`
zHB~x`(3qSRB)$-^lUiq$!Tv7{m#SLmZUzWD^RX_KPAhaU*{gIAT&ZGX=@d$0UQZnx
zl=^&3na^&;VTYT1<@m=M4JqC7NP3La|Fp$DJQ>9KU`_@*R`t7LHvEUk6|#Ewgq9;Q
zC$AkS0j|laf@gt|LB+HrX&_qNtbTfH00m{_-(y66b#xw}VArx^+Bc@kly;eLAhW1o
zCuwMd7DmukmQ$@Yt)wz1Egg?Gk~9UqCHDWrEfHQJ0oG97P*U(M1e%5?g}<SLs1&~_
z<~)#6S&lZ<H#KoQl_y4O*OzF=x=}V>T{#b#$1%(7P2_@cSloU-{nIT%d7)pRdwEI4
z#HT%t9UHLBQc)6f)w_9-^<9w}j0+5oCag^MWUq?p&$Eh~7iVQ-g%mO^Hz`X?%RpAf
z6K3O8q@-kjsLJ3g-8p;%@gm2-{Phwl{Qv_hOnn2OqleU<V**o-3<ZKRv5W>Dbv+$x
zd+3IsXDq9#s14SitURsgF^ryXi3OQZ3c>#UDi89*16DgNi=V{StH&qTJH{%%TQ$`H
z(U6lXF7J?!ruKU~EzaAl<W-kUD`GcFgJ@z98~#0=1h3v5GSyLeK=Senq|{7JC({=-
zl^FZ_`h<7t{ko&6x~Tp-R;VR}tSF(wmD%u?8Xk@<%Ud3_81pGN5CB%Q)lhP9ET$u^
z3u-rzuciV%q%}p+kD;pMiv?NvLeII}ZLi@6{PXUkNN+(BP7q3CV`(Fi8nKoDgTPF9
zw97;qWMqGXb!)`>lSO7cJUrJKMh4Y8{}T;S(1712#y<iDnZKG;XVO-oP_@&=@52WX
z=v>j!#l|JPudboW&sVCg8FcdX5pC1;y`l8j48ky8Gj+$B`lm4^$=Z?c0yDkwfXn@=
z5~6Pa$%iSk)J)|CHQ;UJD@d{0S;m-lgQdsUMlNCneGs%kHQsG-u{N!GarZ7$35J#Z
zF5n*#Mv_bqS?KOSPl4=UsVR1QS|O!of;=!2dCun2k?Iq<qDi_X9TSsNs8HS7nxY1+
z$&N`K0*4hE9?)78*N+jIS)1PfGun^EQY*xz+yUFz_@A^RYIZSw4D(VZs;V+=lsBn=
zeSl73m~MIXvGcf`shMe<xc6QuZga|agd9L67hAcg15FQ)CR{uM8NIRknifnSvggL~
zs_tl;XxJF4E)$s#FdL0hA!vz16t86rb0(D0NgQk1pA<<lo~yp5i~rU8C##X6j;?zh
zhkJb=MsSG>>?iROaq&!bo*BqV=gZHl{p{%ENX0|<YYY5(x!NUF&UnG|6J=4-WTNT_
zRn;aq6w0mjL~#+4!Ss85bHXeo<qraj=*s8qY&<tgm+X^c(^WJUCncsZ-s-uQcn;;m
zUm>pxE(k!rmua-nWu?9Q`^sdBhOSJ<?(I9xhBispks<5paOfjaXE<BBC`As|n5?K#
z2-aH6h_lYHedASctdSJ$N_ZHFQ&uwGx-e0M&dt_2BaDFa9ARumqGbzF!cNnRyUur+
zk=$wgcNV;`$p;pW8>Dg^<m+i>L4^|=q`u1d&#nEMqb`4Sa({eD=wv0e-^a!$P-#l}
zN5PDF@MLsPR4I>ttSX|XXE0v5IVtcBODW8=vDH)1A7P+39=WlJ&Et{%noLMoeK;WY
z2@7~Uw515cYcu-MzffWU-9u?IH;!@^{L81mx~%m>UQAp{EG{K+y6=&)Qe#c;(&xC!
z+j)KtOxCYQoOVfQ+OF&#q*}|!SnDUzW)X31o-WAy51u=>d%p4upD@!17_vTa{e4mW
z&iG?mr<HzH{hl`fK8^-?2dS!#9-gJ8*jVB+Et{o!*|u-=vI5($wc720UxG+ivfXk0
zCL0*!Ng2FuPfS}5^xH3YU_E}1O+X_#UMNGH#4&?W@W}~F`~HoYu@y+}#Z|eo78>2P
znYo28s(mu_Cfc~P^&XA&l-+X0GB8GCwh({ID=FnLKAE(Ac`>gjTW_or)5GAC{j0-b
zN_M^hFb!{+m3ls^lq^Ko)y}dH4WmYhsjHNE_(^Xj$C90_><&%eZ*YczjUMZOBd$%?
zLxlYleTf`<j=#Y^JAsOih%V%5;DBMeAt}c_DjFHN0+x>Hd^Z7x)Ox(vbu%gmkGNMc
zBL&1ew|Y6?8I^4hlE3?dI}N53H0U-4$B+OL4!hu>jT4s8RAvru&6QUEwR);{=9L(G
z@au2tlhgN?DJpm8*CLd9@L3cg$5bzFS6%hm&nCLv81#Cc!-}kp`tnYrc(-y^oSzst
zs5GCQcc-Qsv?Y9B=pc8B_CbFQB3}t;z6NF19utdxU?P{I&)14o5v41?5tCvlEsLMc
zqwEsY2vUI_?2tVpJ`&@EXenbkCfEYd#J;|JBlSfxK0QUGAd)w7aZ3@=bfsZH-O6gx
zR^fSu%jrwH{O%a13`9~WDl>K(n{Jtq*pyK_4mZCb?}1?@0UCLwQ)}6WB<1ZmK?64?
zZM)e}lQ*QuvRg!e19XIqg3&%EInH(HnI}$u9aMF`_#wK3g!d~ATVFJY6e+lv_aoK?
zqbx#x^@VhCQCTeO`_;KQ>h$<Fmq@MO{N|+hv9atls$qhdCMI}FnyRX2u2X8&Hv})S
zU>G&kR9E^da<Prn;TY|aUo)-3#dLJRnkHy1^e4*fV{V!8`j)YjM;<a=@-H0Qo}X-E
ztxi20Kj?ksfY7HT>a*YfV*1<6+G9gyM{QX`tkq7cB@W{Sd{!J!o|4g-Cw#7w9%3E}
z(PR%o5C?4o#UIH(<z!`TT%n^cZ-US2M*?`=9#99oX-<`KH=1{QCI+N0n~n@gphddx
zGiuif>!n_or?GhilBvtn!sQm9d0VtLIdb&sY@EI~=_U_wUZUHN(Vw^b!IA2%_tllg
z@flL<nY)InQBasaNFH!87hvIbFxS*(Y?@arlKizdBPW(sv&uDYRI^sZbnH>_yR<Z1
zRNw0dpZXg>Jf79N7mppIp1h6Ibdw8%fc$7r+NGl)p=!$>wR_`rY!AB|4BiZ%H&#VA
z3Q;;TJ+H5RJl(A7HoFcctd^ltJPT!hL#kF3J!8^rk`v5m(EvhBME&f8E^k-Mbq9CJ
zXZfFQ2&z526bE@1U%e`jdDJ2N<c`thV(^|(F$##AlrTipR)GMh{E_U0UtG9?K5;_^
zeA@jfsv2|M{(?8WRkdI$C8gPPTyeYC*mD~v;CXENHcR-*?PS;Hq7Ra3+`qRmBDbTP
zB<%ifx}yPgsWi#?XjX06vVtP;W8i1i3W)v5h;!>^2BSv3R!I&mWNwb@l;$rPyuUg2
zuWorYS3UMr8)t23)ZSK>s|-er{&}Ev%Y&WwzPs&!vUWz=g_Re&S+TnoyLu6qnQz*W
zEE5ce_?^SeO4E9a6=Ku#u@n{xzzLr@sU^cl{j}r4c}VFKMZ9or(=+J6wEAUSM*b5C
zTlS%6!Uo6ndy4hOTeRa9*>PguP@!KXBV@1n9}a0QPNF;DRmQC2s4*btSAMK}ulW_3
zo199Hb48XaX^3-RBPUeHH3uqkPg)WrxX;ZLfNm$3UDKZ`>nDmZ9ANdh#y>?n$$OnK
zcpP2c1JCz{+4|oEwkW(NKTHYHi=Cri4y<O!YmpOl3?(ke!GMozM$H*RJ?XiMZb!Zt
z^>3IKNXjI-dDg$&lo3z6e?rcUXv97AmQj^#-8yJ68Iaq)xTqJ@<s-_j&>mP$N&^+b
zDt-4f5WcE;Enmq6c+|^K4xYNvsovX+{aZ{jl<Exjlo=XMyz1T7$dGnbQ<W4C+=EsI
zQG(Cghut~dG3leDgX{)tX?P!hSkd%v$=+2xS<}Cw#qX9EPgNmom5@K!XbDGMX|a=3
z{JJ$S;&EGwP_5FYh+nS3fQWHy)haH2x-jAE+iA46dvY<w6~42_@ouLjzcd2u)WZ{}
zvMI?)!VLCy(3Y}jLVnp{e7PNA8#^Cob@xhhIyQLq0lT}rs|HbzzNK#D$2lT~3P#m3
z;w#v$Cl+rq{nD={Xx#`EcsT?QXZRu?cQf2C*?0_QU&U5<Zz_HR8_M~2)dx}a{vZ$p
z_i;KW8u?5>&qKYS%SA9a{R+1A@Yv;IbsLhVz!x6%Bb(^~O~IY-RYtdd7+J7p9XKvt
z=UMa}6ugJU>a}q}27pJKwzqRf1zbs5)Y{6{pAHe7COXl#N^3k{X$EkOB^@SQPRfwJ
z4h6WAgxs7_wKY4(^xSSJe+*vbJZW2EJ@gaE_pwcE8mB~~3jsUo1)pDpzH!zA-ESsb
zCDlZod!dk1u9s$dmp*Ax_wm}2!Tc6HnR4RtoiX1VH(UP{@f4&({<l?wFY>=BPwEP1
zh37FHgqY`i@sRa;6WL^Fh%Po&_XrPEK~UKpePULIkfnhHs6W%_X50COBXKX9!gDL|
z^1N^{Jm8crX8P0x$N**D>4Rr1o9vb!l{AQn_0SS-)$G!tsl~flaY`N_?;-q)i<xYE
z+9>Dq7Wo%`fh{MGM-=ygyYD{+_GhOMUxzA!shiXU6zY$W#+h8Eu|$o#vjZSD%XmoA
z)Pc(pED<nsoHU~SRz=WrDLohIpQ)BRAg6m@%-MXiD}=r_?=IlvRR)LE1jkt06-#`d
z;oUYFLOXOxNKo{$!H|EE>}h9mP-9E8<2G}QX;9_{A$}Mm+zKu(ngL`nAvtpi+@x4D
zFy18XS=`l@P7qiAj%G1<>*{wT!?yRfYFK}&>YKN$8HvxY+}?&aK*_rxBdlpAHUZTr
z?yQdB{iApGb_(yZPb->V5d0hW*|{%toTq<OcLWdA1baPgYiF*=)_iVk?nrk_^UJLx
zP!09cMBpL%TP;`14$C>4I(A;uwpv(<OORwWoTd^<O%<E4o&KF7ap8~-f=bGltyL2i
z$Yb|yduKQ0m+h(a9sWx-!iKbA7wA-%yn(div@GV)u!%wqR-r72^dn)3hsrqM+5URG
z2rV<{B6IF>yI(>d4auV3E9r(_<I8-Q_r|lY;B$eQ+<RgSE9g?Y$pf)^ZwPp!%|Slf
zW_3UwcVM8h*0WWI^Pv<`8i6e43v7YgQ}tISi+(XqOU1bX!;0Xtd0*Y=nqXqULZ*PC
z)75M+agSicn0AJ{%x)=or<OToWX}|jGFAnq`X+i{BL)FbUX}X1vlJ)3e~tIkCQ!5G
zd}_gXr+Mu8$#ti7OPt7#p(WCU8&ue`+tZ}j_rZ+Rezb?B^jF@!{c5ma7K>bf#QYpa
z`&-J<EIF3<a0^eaIeTm<KX&Qx#>;3h0)H{6+^xv8Wv%W~Id{n%>Z_;Oc7v84^saxM
zHU4-^*7>3lvhTd2Y&QZiDg7ly@z=#723;%-5O1lm@gKE;&B|`<bA29f02KXE7=;f%
z+fDKct=!8yrM!-5mBp-fev@hif{n-}v|-**LCTgIWpP&QeidHpjL%FOX8LxFI%T$i
zDbWNth1;}OIkzs)b`Bak*m3TE_P%TpZhdKNC=cR`RH!l-LBAz`wE`_j$$qaob|I%Q
zP=Cx@L?QJqH<<s{7PDCg|7ygdkwE?F=pN<=(j%7m$)SFvn;ORvrap7?Q!*J#29Ed%
z3r@({I~7trT1?)w`3^Vrro)$hp78BD`Q0J*>x^o{&^4Z$tTvNXVK~{0jD~o&)1;&)
zX(#l^X|San3JoMp(Di48puJyBv(VsuPuq;KmtuKCba_m<o<xJbw^Gak0e3*d3fZ?c
z_tWhxdyoAY(gdzXKD2C54A+p&;K*f!rzlT(TySQ~NjFZq*u{aH%+3bO82kN?G(L3V
zNo~1zUNP2-)oX#3k6(`diWy{4VgDilXfFB7A&vPkq2U8#V`C{<3OOf2+hU+)t;vbJ
zj8;6fIv`<sMGO5ofn$5$U$`+e>(-8{-q(S#PCh6Dm)&oH$w1<<gR?~kD-0^ieu_k#
zESGPmLP@H0QM=Z$D+05r&|I@{8>lFxHUOUL3x$4DjguKKQMt(`Fs+2cE+bPx$N*Rs
zBQ{8UuIED%&Ec}NwFH1P90l}=L_oTSvk>9MEtxUN=MUSDRj)BhEw3j^WeCfo!Et#!
zE;0H;$pEB1?K*>S$iXwgNikd^_ipnBiP6O^w;Q8=i_|k@a%B8N4T<!fX7;-D0)SsY
z()=jb;OzM#)#w}lFy=h^I=|e}&m^1Av}T0;n|%7r3c;7KfehRFuzU?!yZnhqf2+JF
zv5_>QGmq02x9{jm7?Va{2YxSY00;bZ=4{recmo)shvBg~y0NLY*qLBby6bZJitW90
z7&$Iea_dP8GP~|CueMgzB>{ApS0oJb11GCiN1buHN0PY^b_NGnFH`$c*i2chhhOww
zRMuZSvo$M^Bt(NU=~8OU(mZ<`Pf`%|;uDBY$85IHNobX$hz^-=S`o7YGcm4GNW5&+
z6y7~~y3V%>bdiWu`aYENKaQ)vy6!qX>)_MFtRA&_SAQCV(V<6~>C>UYX~*#9;;X|#
z!^$-lkaZ<*NFYfb`Jz9$bp5kikvaY_nKjSlVHOF0R)qo!xtSbUd<ZtM{pp+lt$`Cx
z6&4kAX3KlKm?hZt?&O(35UD8qmyer1r&#s$4D4`~KhT@qt!gNsR<F%FCOYh|FiW|=
zpZpzNO#CsUMI0u7$EwbPh8<{NY3e`tXR+*;K)drZ#&tSv^-61pZh3Zi(h=%-FdRlG
z%{pXMEqs}^!;1!|gps4q>+9TYD|Cr7r#i+Tw@6H69eLiClIE!(AUB+>H~GXgU>`z6
zoiLf~cR*0@Yb_z6npw|0v0dJ-Z4!3cZ5$iPeWOQfCQm{hf;#M`eruPcsxk~aCUzO6
zH-eG!SeR2i-7#U6FW5a{YvZ&%9}X*;pDoLBMzq(&KauU>Wb3}IdOt*^7mN)Qe7ceF
zP!O$dDi^KzZm@?KZh%Ska_yZ34lt4@cJU)-^Cop;d$!rW)ND0OM?MP%yJJ#Y{MF-P
z|DbP*3S`OURI}QRsZ+6mndWZ<6Qz_<0NiIqa5wR6(SjP-D!um0QR!5U_iGafJDm>m
z4wYOgcTG17IL6T#ElsJVw>Kw;{;cu>ifQ=LcfyPHNoK&UGRt`{7mG^J%m*BxJ<J7;
zZZ1zCZvzIp{=PMF_C8RnMmWR!(n<4S<RON9gI|mgjnr*SlqDG=G|uUDIxlC-d!}Lb
zL(^AYtl*{X+(My=iQ8q=pVd|4SIFGP%WW_Od`YA|#wdymVB2GimC$>l%wpy|WPk`c
ze{O7TC;CvqELy>^t@WT6#2)Nj8jgP(_unz@UWQk100yU{TMOxUrgqk@eInp00ULA>
zBB<WBX%pU}GK0w0>pITk0^(I8`W^T50qP5wL}g}suCK-n41H^MkNm%47n6rXXg7w|
zfdksfw^s;_Fk*df_On5#a0u*f8!{vG#)mEXZ+Zz?N~6JWRol%Qa;TOC%<|=N(ZnCN
zFbHv-SKm*xGC2`cqS8h~?-6HWR{Mi4H)OaA?lQm;4#8QRYx5PnU3rq}c=5ubeM67U
z-zsQxW>{QccZ?krGf(^$-PT)x58T{w>TeAAB1pQ=U^|o}t(xR3>!siR05V(Y`!y2$
z2H^WV22Zz-wSOLf{CsKWAq|D=x-)7j(q+jZSr}4bVUC@8ZKxydL?SFabf1hDq&aj5
zFSZ!9+Bu8ld;&91)0c<GI{282B5%!D;$PZH(mDp!_@O8pA)B?|8oLaVweQQuQT60p
zVF4y^@>N8e@cr7(VD?kK?^%T~&$ll$R=}SQ#RgmLTHv^vpMMXMACMI8keyNunIO1$
zNyC5M4I?@p^4-qGiY+NxcRY@gz7!CZQuzt?>th~inzfZ3<JUT(ic-m|SNo;V>U-YL
zu`&`m;3Uha%q|ihAamQ1=2C6><dF(i5=R2qh!6e~8tAt;>bhuf5oUZ`vaK5eqpwbr
zo~`$BT~>?<G@@<hyFlS9#M<yk%@p4~@@NQO{cpP{9+f{8k!=c9V138)9^SpjSb^U~
z0X8s=rjJJgl{v=1>o%c&DVZpEvT32A-9r0^4sgOrDgUY9$A5Rs7$KPtLLi#aweJbw
znb|)x*$>6wZIn#xJJw{K0F_I$9c3GaspwiFc}-sslD6|Gp+zY)Qz6usZ1*gIoJTms
zCZ-T8ke;mOedv?J^Y7M#27G8}r<Wk*|5#+<-sgQZyo94t9Q92D1pvM}Sme~csAe<H
zW&p^UdZt7yHVE~UoaDkPr21ui+sSv6RgEsV@Vj_G0$9+|IWFYWpp0*^Z+ABBgf*mz
zy*knpdH3*&JYT_ykB4c#8$)EPZe%Q;IfdfjF+Zhcz~Q2H7MI6f$d1Q$8z}tESXH%7
z$6<qg>6b~!7E(p*-;4mk`CmY&=JMMI2add3`%}qK5gC*{Zxgwfq1L5m+dnPNbsoE;
z4XlIV`8kF^5V>TjOl?r+F1GhE0#ro~-D*BW2Y6^cv#QJ{m_(R4(IAV+_-P1O9;_~E
zix&#-#{X2WqVm8`gH9OF0u9wz@uj=r_aoc>1OOF`@W}@EGr_fOXsQ1?4dop)xZ7M*
z!=Dt<E&5KZ?c~HoqY@<!eu9_JYev=M97NtWslo+L&nWGbg@wF%=xF<;0j)ZM$jA|W
z4yVZ%kSyiwN?6fuUC+~%5sbzypbllfrv9|Sy;Q1BV1`re0Nv=!p}>Px7^%4<LqN*&
z82_g-vIt23SwJA`y>+kBov^+fDc`8O1o`3$)v%G5qJsjd<DAuD30ric?fXl}qkTsu
zaFgo#GYXUK1B26Fq08T)Fap5t)d+KBfMAzdaIo_V=B{<s#+(Mo42p+s=TU_Gj1s(A
zX1w1#3KC8ab{nkgAgPSow=p@2N_RXFWkHm!e<8hiHecjwkoDP<@z|O<TkU&30SfSq
z%q`dRu`Xz57E4U6=-(YRsk6>1A7pj_gdJ;OdveOv?V_wmy``jwVcQJT26{#;hWFtR
zM3Pp#USAv1LB0rD|2h}%@`!2JBRV;H7+a|#G>#KLW*8uC@96Lj@toIjy{%EwQUkH{
zNc#km&PuooZd*8Nd5Gm7j4aEpr1hD;|2<PqknuKNbnhOcJyBgS<(&#3v3tV)I+)jM
z8p%lxi3iCi1k+v5bzkt_aeTW6(t??CTHTTv2hAZNCVwx=6NUa$XTT1P7e7^d@n5y5
zx<h)yH;wG}8v|aKgnpW}eMma<ZW8)~TfDdyCH%qeks4lmDYQ8WKG|9b&Bbq1Hf_h}
z8h@2dH&co5P!=O>sGGvn#zb5dL2dnZ>b?8^-u8gyRCJWHRdQ1nRvx{ONv}%o`T6dh
zlIE7`_Nv)}{96~lFU0N9Yga7#_wE&~;4WhMGdv{mj&yDv^u^Mo)$FH~7WbeR?iSWO
zU%Mu7ucGTrhxcX8*>-Y6$Ccd*-u7M-C~XM(ONjmW2al^cdB}>!{h9;CPL{vVlL-ru
zTS716KGJuk={LSPK9fEJ*UMRC)=@J!kK&6h*<O_@{ddSo)q=^;u*<k%4LbSR?#BDV
zCVN)c8~>%N=x>kTVmy(`_t*o|TORxgOlm?xzkz+h=%SLGg+5K2G=BgHXcoOLe2M<_
z1lbDcMlc**eb|ohd!Qk4@FR9IKM}~hAA|1nnGg+kxET5LaLsPCU~8vPS->z`ifnEE
z_lyQ9@bXTx&7m}`cb0fz3ksRA?(QE_rjxmf!^zCv8(LHJ6W?^vTTxqKIic@-EVI*}
zVb5OwEerkxb2Y$2WOy>^^X9m~S2GD5R}^zQbe4^?MZ{bld@eB2PE$khd~65zVMvoc
z7FF>_?{XWj!%KqPW(%5)&dW;JG!&w|WyVQ4t|&em{*7m4b*f_n?X?>>o#fz9u`nnE
zc5d^x0tJ&Bdfm^y0v=F>2emx4zUy3;B2U*ieMMpoO!V@-!Jx*N;zTQa-Q)3k$BoOw
z-+GXBo6NC3#1oSN$Y;?=Gm~G+`Lk8b_7i<PxW@2=y6F^92_LGCqf*N))rVpQ{@wJb
z`K!2%`FLblG7gZ&DgNMmVqWrXK$K-Xexe-%cXu-q4m8Tt2qJUW5_mjS7Z$;~dC#2{
zjg{RwtEDJajHQ)$;GzY$0(Dw~7(1j4PI(JpVM7NJI!b6ZJWGC9&uyLTT@u*(VWej!
z;6gA68^~$Hz`kxWKIw{hQec7Q{{QU-u%-@lgGl<Of>pl#gnu=0s|G&lerlIQ__XR|
z<>HaK<f-%5nBR&8#fuG16LAybWlTKsjm;W<7^lZG6_>HE_3`_Z?EYk#v)LoK%&v*I
z`b?>vm>h5DoOohl5}R~5f~c>gOKyW5E6caSFic6qqxyznBwto3mQp5XkyWKAhTci8
zw#B<*A{mAUn4w$MyAbU@agG!HZ36UoX3=fyVw*&vA@Arh;&&{b&aR(JuTMKNsl(P%
zS}0w;&Ta*+@Js3&e0oSGW&?jefR=i0sZ{*@g36U-?y=BLKsOR>&+VG%=1!G<$C~Qr
zH$|r#LK-z^Ku8}k+@}Exq!!N8thRKqjHCawiajk%JKLGxUu+wuZ=M|v?iex%Wi4qn
z&RFl7j}y4FXg-=e;Wx>NAt53Df}(QNcB3Q$t~lEQ)rLRI?UDeHR`GWIX6eng>I&-5
z+O#0VfzcO>x>5alEy15{3Z(r?2kXV<!DFdpm~&<1jB5i<Nk3yu@E*YAnb_U&F7s5V
z-})I%ZbCgVG^dbxzIm3?vRez;9pr6wj*$6{Y1ZmMdk$@#k4SpF^Y%VC=<g{9Nb2?d
z8t@J`moYEUZl62b9wie!ku*h9&wzEEfgq<{+i#POy|28aGHu_p`zluYah_CZh^Rm#
zQU0iIwFpqJX6P6SwtS`O<(z$y6!5A)re(c7C~%5S;16lQZXN!}xrwE!qeY#*ygS$x
z@GyLuzLI)#4(CyPs)SKlo$`3H(S`TdwD;8?h7b#F4CfM$cxYG>vzI5Nw-j6>=hVj!
zhCHV9m>IQ#FK@T9OUGlJ@oaN!d>_w_OfMj(BL(n5ubeWE-r@P2raeTRgPZgO<V5xb
zPJ7_Lp@;<Xpk)s*16*pcqSJOxUA{cO3{aF2wCc+78ismSU2$dzoNlG9sM41~)8X+@
z+M|Zq-Fg2)Y<{y3b&t;0-aH$7_qcL*!d~E*`<W+}H>)RK!6H1D`vLi)E62P|!h?Qh
zV+^Ddm*FbrTf=PZpY=)gLj|oDOw?ET##mIAnj6g?i~Rz@7MQ6bVY)>a=N7ra(xP&i
z(JFb}96V^Zo8vNa$aHtyM}|*?V^3G6vZsai-O&Q*2GAo22Hv9H9ukjT!d0>8WGXg%
zL68WwoM^jEFcYMi&HiiN*@^>2a*~%k9OmZm$st1f9*}EI2fDk(masIz;vvB#25tPv
zouf=(Y20IwIFg@)0=IURZH+hhTEwFAUbCN+wMgz&Pdx%g+5tUDBbvdA_-RAo?z=-v
zep#~Dszq-7H#yjvZ5E|idF_f5(M%*fCf3?(DoB0}axd$;`J#G>S!j(F35lvm7cG#y
zTYT~^OtrD}J%oi%!j%+CsrO?aItJc!m-AeXH?9bocOS6r4cC9rU*tI+xX!7BdvqnZ
zdPl+tmi(3lTbVldt|8g#`7k7NRKTGj3<A)dTcK0(a@Zwa%4Te1RGr|Hd}Pg%SI5yp
zUzZ1O`X}K>X;zRQ;TP>;^V@BvZ<{od1iH&8adk_gk|ThS5dtKpda)j*zxO<juJvnY
z`jV3A_~R{}C{M1rE7pw+O$(MvBpoXDElapFb}^u>%UH!zMgFZ*qs98aR4yt8?P;9c
zN{rdQe!{~sO|FWd%tW#~ZiD(87k9_THt9pJvG!k1k#>dcdp71inYZ7EKl_ifquEp_
z4u}Q#=JHPZ1(e^)8x6?9$&HErjf6l}0H30LhwS3uKnGC(BfGR{$o;kT59M)TV!*#W
zWE$0VhcwrW>b(&(Aafx28#exjPbRdA85#t}q4(DIVdiD#2imsnL#apiw_Cg+@;+JB
z%>VrVBWfrM%S+hb*hC`Zv5hYDv)G;RGmic#zJ5|eg&rN0#3(WQ@<sN8zKRZ!qORJ9
zzd*zP^O3(h874|}rn!!~N+7A5#32QwF?xsdU(W%9+{^ce&UbN1#aC#f<C4%Vud{~k
z8vM6O{bT+tD$LN{EgB*oufvgV-SR$RRf7Nes!nEkGHg&G51Hbf1^OXB&A<L$D4co|
z<ytK(d;KEiiI+Z4I+jXN&2Eu?MV<N66ST=}?c0GZ9`RoPWTV6=zq(l~78@JZ^3#Ak
z47%+<7=JLxwXCq1z-9tmNj29?6o6dq)t$ri=0BN&Fvy>U{Ib_S`1R8NElK+yY9JKu
zthEFcR+ngD<KXnZ_BJYNNc>2-B=NTp|A!jsAB=6T+8uq!pTisd6}hymD5f`$Qc5J7
zD2Oz`AMiKV{}10)i|&6{98IB3jSlGMJL`~?92zf|69k^z;MKou4uc#>Gg-n9Tusu_
z)Mb)F=A2BO*c#9cS^0k$Zwiw0K`M)sre<tHJT_Mn9!@lAWofCJ15)B9W@zhSRRp|H
z7SDi;s%pY+ne=v~c^7=^sX^l>lbeH<Vzu^)HuuVU%Uj3tcKE@X_FmWW2Y382fDe%m
z>APPiMlo3+<5G@(MnUnf?_X38I1i0hRA$qQjL_~86$gl-#Jn^P<(<kjd(J<uS=^>x
zf)upXln3hVmOt*5PD?Mpix&KO34)mRv6~)R<~9c&csN)(DLu92j`Zez)?GY$1JT{?
z0NPaEG*2dGkort*ed*9$&~d_cB1)ShY7Y9?*0OBI)*f<<gw$q@L_usc5UU$3D0;Mt
z6P4=0Sb9%b<FbCguZy?7^$}v~0IZtEt^Z95%6LCQNwUEJdRY5zrykzS-s*KS&>HNz
z>fK&)pO&9ldv&Rm0F0ISc+$VsXua=p(YTLY)Y}>x^)BkizN3v)ImqpE5T!)jNO72|
zl9rO;@@kuUWl{ZDKam7X_I&JjH)&H>y)B~e7gG*Y5caF2YJrN@RwX5+*qjJj&atMp
zJVU3%JF7YoJ2l5K8WyjOR!<ASrg~1h+&7MVjw?8C53z(iN-i7y)L0qOiJhP8*5)|Z
zg5yUd%b#|hH-nxHSn+Ur^>*hH)!XLvg5B`Ts^T!OL4>F8r-(c*-@~r;aE*gc-M)XI
z4VLt$%1|xqQ9zlfT&mm=f7?}Kjqxi1yw~&;&{n{7EFx<D=H#c*FBP^uDJ6&e>0@MU
z<w?Y=xrL3C@qLH-Xf^Lo-O48Ur+!6SIH8u~!OPrn&!wNd6bpz=7WZ1zt`JQtyBbuL
z%}Wq&yJjm53>O>>>Y3uReLHu6EQb14^$)EfT5KWq3-g%b-2G8hEvRl$^Np3vwk{`j
zkzanFq@?_|`v`e6(_j-I$hNc*;Ht#4e%Dt50A-;4@ASN9P_tf;Ae%PhneL#O%t+3=
zg|Y7cIu^pupE}2pdEnYNp+MrUf%{EJ%w&%F9-$PI+fa%4%&8Oeh2D)ER9IB4!@4y%
zLq#1R_l{}y2z6&lxz>mwDAOI%iC@~)>H1DCiwCfI3FqB&rrk<>;DXT;Ne2;;djs0p
znNw@kF_Sq)^r-J`nRoW7pP-|R)k^>T(i`me${#My;;u(WjL-@Ql6WB-Xz!dL^cO*l
zs(L(WQV72BKG~wwLi8OfO=fNRb`wcv;%dRy{)8#-t!iIZk=t63f#;{NU||XvhS2=-
z@)<}KuN}M+J<3Y4dmnQ5P-N8}@C!pv#8YmC#%!%q@NSvQrEX&%w%IOuQLp!sbs9O%
zexdVLft$D82hAYR=vbqAq<Lbt*)nh|n5JHfaH_P`W$n^jKhkGF;sI5R_x}9}P`st>
z*M!{IS({|?=N3EAGRnf4{eXkMJ1pq)Pq4)6v-0V4IN`!irWqqC7@Btx-!V+<$0P@+
zDv<SQm+?Aii6SR#Bb$8392igCV4n{f!<-}7B><_9Y2`lFn{My|z!a}|j<x%Sgd{0P
zWC5VotjxqgI`&ES^s>Hv<c{hFz>~;JaPQomNBxLpzu%LSd&L^<h{<Bq^9c*z_KSgj
z5r(mzO<nC{*ib?LQxi@bgDvfZIb;gIt<-!u(eS#3mL$`HKm4KNjvwfqR(|KtbR0kk
zl2|x#ige`fab2lcfJC11iQ}OMO4Q_pdeC|!raYOi5{EDi@Gf_f9@aB4P6bzvy2))V
z!DU(2XJ{^X&OO)w^s|%cKe~uKe?&C08w=`I9;{d?eCy+y$}?UcUHIN*Qt`=XW&@E}
zrN-v5gi4CuGsyD=6uMih;S*k&eU`mzJT<##i$via25W?#%c(b@QpOX!6<iqGEnjvf
zG-5~EF#!H9{&sxcQqNFz(X4RyR*%13z5BYQZoQ+Rk$H<QTi-IQUwM;b{m=_-w`n#2
z`)cXRXZLc9Aj?+`ro=9_=nOx=Qm~C9uNdD2b={%-4$tfq;najm0C-58g!jOuR=tJd
z<3cjoO|N3P9Worazw2pHp1qDq%|ye)aeDT#)d$QaZ!4XGtF9{PX=p>+QGF;)w_c^|
z=(OBF=hCZM)UZoE6OR)<bq6!O+G^+l+}qY=oY<P<1ax{ST4<W+-F{sC+1;+-`^B>k
zkAbMAJLJjyqL&|Lx5a`fL`c_nAEe3t5)lK6;lC2pW+7U2+sY2$F~t!cquFg72)KS*
zZ3O=~6?}c#HQB|6cv+c?-HV004+!PEgX<_*+(2BRrq?<n)TgqU;rt#vH`>ydMmG6~
zWb?DRowR+po@Te(8;vk~IAh>y#BDqUju^$uFx`DYaD`Y%{O*-I$BdnswifY3cl6fp
zoaW`pTLn!`O_fKiqkbjjch3XLqHjTOy&Yb6V;B$G>t_`o>0(-OR#^gwxx8-%8hj7V
zXSg5SQ$ebW@$>Op9IJbC1+AvHO@?Cvt2H>4Upi@ZI@VB|*7)HMFCn}GO-}8vjW9bF
zj?cJSN6sJ7&vKwtzcJCgg9)unug&B1OA_u8?cwJ8?qu{>g-cYX9bUt!iA10sg-La3
zkY{+jj8(ZrH0je0Qh1+CWp;<8*3nlhw1v4eQt<sN!5fMq!3k^8*7yH;*9Q5xb9wl{
zoLe&71VW8NG(2;(zPaKjn+{iYy`H?=!P-ioGX3#=7wLl9{`vJNO+R-ZY7w=&?HCpZ
z0IVn)4*8Ap83Uqd8!f86`__vQT`2i!Ioo!HGwmDl;jBtu$)}e+!nCa2#uVMU6N4nV
zU>b9%I(+-#TweWmD>vruQGngrGkfG=aJjN_R5;tURpO0;w%$zHCDE(iP-V~1HqM%|
zCR6XBabKpJt8(w9$y>~zOfVdr@2VlAf%B=Q(Qs_BmGxZRM*YmJOyx$E>>6UmgajGX
zzPrXMnG>gI2WE|C6ne><ygk<~<ize@u8=<)9UGz^7Ro&l@1QJ%O}O59*D<`bmrkTF
ztG<{93}CcJ)Rh9MzZ*x?-b|3*z3)=6`w<a1uV@zH$wY<EHmm(b&vX1BZvK;wJKos+
zt4z{O?dQA8=8)`u+;!4^6d^#oN1IIR@^?J)%FBU6ov}GvQIFF%)+t9B?0p^-ZNBY*
zvT_ADv77=oj_&VMTwu|6?BOZF<&wrXsiO8|t`F2%J`;fFuOS-7(pES3<3WI7duDyU
zcB6BNlu88UIcJ1#QZ@oliG?I)?PDRgA>)i1Rf8R$Q+LzapW5RN<!8{x16;iDTdwd@
zaqj8&<;%65y2SHDHZVoTwZ;T%G-nv2s-EoCdU)k5mZ$Q?mBNP1*u2KwbvUmCX0-w{
zC8@iJx{(H;UHL@zA%B{Qmqm8N&YbjSktU8nD~e{^^Xw@@%yVQTohmB*=(Af@=k<}P
zZOpOyR$E`Fom0NFj<5Y>y?UpAT6FXt+X@)j6g&^KW~Wa4^!mIQ<#uURX05*C?ALoc
zJ;0BM21!Ab#VW<j=9{;BtSY`D34GC!A=kk-Gp(doY|W2silQsw2q;$wY72a*E{0QQ
z8nt2V{ow|v3U^^iEm$i+Yi6c|h<Hl6wJK3ZC$}jDJEYv;q_D?)0<^30Clf3*ai;Rf
zU5Fp)DMoZ+zk~Wo<$?ezM*-NjyYMX=a!KbZspp?*S}=zgV3)Rwm97O}TzO7-C5V+~
zTGKSce`r#_%f(ib8KXGxdRsS2T#8YEyk=VOJ;87Y^60Cv5hL0srM^WEBvuygk+aD5
z7_VbK+(L9E6n$(q2qa|!C#13+@^vlXA7<-rD0vkn9G|<}%^qJM3l-$pb!hfQr#x_!
zwg+Do9@N+j_UD&23HnmC7@~6`&LMHu`5+a+Qr|3Fxh^VeWDg*gyV8iZK@YNJ(GCYI
z-zJD)Ui*~t#5UKXM^$BBo=l&H@8zF`*?D73$w4_}XWXrWIar5G%NrR{_yPUO5ppO@
z_nu%r@!W`%4PEziuKt{%t|c`i8vnKmx>Y{JFJI5^^n^ReVFFQX!rJ-n4N~Ww;d9P!
z69@x5Gta3qE%}ppsZZQ{3H*LD!3wii@T9oWs`zRRJ55ajyK#toLl&zHlR0Yzyu8_y
z_E8vP^|`W7b1Q>_pCV}R(OSNNB|vWQ&?Ot8HM9O<>_r)>Y2j(|c=#fAp~5jXH$XWs
z0c~HU7Tofq{JRig$MGjYK!{CFU-)8nD`+zsnA!WSr)^?T>0brZD{ONRd`Y4=87@%k
zPS*HV>R&oQNti#t+Q>xGahfjK0ES50)K&;w=07XoI&5>r?ukGg`Ki^s6piX?C3<@L
zyqxN8O17x0U8vi};~xDDg!)fQu=X=76aDN8ROC_QDpFBW5j#Dtw4|w%)Ay<+Z4PpT
z3xv=%iu?<|n0$lW8#f@IV96C0`<0djs`1kVvf9;WdUoe2%azJ~k-Q>U#B+bPUTGKQ
z>Inzn0vSIs|I;X16t5T{lx8NWBQo+|qQXb6aDnLW+`hP|nHJ;~<iWc*nmJVACgCC`
z@S)H`xGn6{|7|ZUFfsrkHY1K==+F*_WtR>DZENc3dSX+`^QFee<4!JAU$8DIYM2ji
z-og!#%OX;S31WOumIDoZ{eKaNN&hdGG%JdoXQ<f0DFM1Le{fp-k(N?q)$q-&6l$~q
zy9t9b?VTH=|M{@_f10u!0DUF)S?$(954f2>c)ek4?bz6LO=VG8{2#Wyi%UYss8vvR
zGyz`Xzh@dBia2=U;z~lxHXU6h4bVWaS7}K}Z1sgB9TS7<6ymFO5p5BX9Ba)w9c&bI
z4BarVf7bevSqM1h*--Q)ltxm^vc%XpZ$uuYy)+5z^t2QQZSD%jxTH2L>;;y|{L6)e
zV8=u{x9^}~wTZ_jbqA6yrxh|FU(~a@gB-hF5NVv(F!dg}+WEQF6B!V#@0Mr*hLWMT
z&$$HiVYd56{>s+)52n%h;j0tGcyGIyH5QZ0C;U4&E;jZ9kEW)lVmpde5!k3$0j8R=
zk@7!C7Lphl-6zEd*E3q578h34KQR3tc>_HMKiz-F8VzNk)9(hAy8TFJ2rHMllTzR+
zu$GdOzG1U20~@`|fiae{m-=G`L0DvDWS3p76+8_p`VeJtjD#sfbj-4z5_ZJ37JcTJ
zLQ`*H@s@6~O^*M214oGM4|W`u{0HTj9&IZY@W~Yb#>OdAa`Cm)i|K|ZLch%h8FsR>
zeT|B)*jAl6Jt^MK7m}|R7wdI>m~$Uz8N@Nt*H#AAq?~c+tXG)wBp*x{YKM60a%$iu
zCCY?7uc&F19hqEV7ZC`b4eRo558IwNwz_dv*OrtlN7#KiJ)=+j`IBsocY6rBA+GdU
z#d~-o0VXQVVQBOzYE+rGUZn;*YF0b85%#paA<!{fIYw=|?#7`~_24eAZ565T0aFE|
zx`_N&U98X-B&c|=ux{rLs=Zl&2p+te%wd8o_HSvzc<z_5Q3^24xpEv|?7p-!Ls^Sx
z+5@Mf<K`59C~N{H*fxrTJ5cbuX|uji<kYU5FZg&)h8cR!$ZLH)X!}|&%Pygh*W>wo
zwXfRygxY^ls1)EI7e)L52c1ym1&)F{r?yOox!LHDncTk9KR;{0)^T<4>5s+)fqblc
zDe*nuPWePfl3Tk58D2sHsH8r{=T9?*i6D&(b097Fqul}k^4%jUeMjfNiZ%SraiA+%
zTCF~h<~2~5XL+wg-YfFzPAVqtm{CfxlC-c-`mSm(RN~cs?Wdjyb-9HosEDpDj*M8D
z7TUfPxj#Fq5jjDp1z2|g9p^uVypHGLDw%sUI3IS|O?}~$88$xT)8};?@D^Yb7<2!y
z`8nZGtVR-{yx?YVe=ff*V%1k(N?!KIE1%|K_0OqUN3RP-u4DYN0>XYn*)3Z<AP$Sx
z%|ecg)fWA}xn75fwPgzurp(S}gORAlYg29;C`HV4aUJ-)shjf+trzL2DB#l;m2r0f
zLWagK$~FnQ!=YDa_*UTU#?L{SIk8IGv-C;sZV`Ouc0AAEr|Y#8;|i?b+t@zizM36k
zK{qN(G#xGH&knGDDm+fmKKXH-=Jf@wbN$p$m#da637QVX@$Yr1CQuGWudJ);M1gBl
za>272UbTHW2sJ}cJ=3Gs^U(#h=;ks-_M%}x4<~})zf;P4WBs6uGa;Q|WM;PIL>vX+
z!NWuJdz6dEVuHj9G>}0<f<Sh4f>9XxLzwr|yWx+7o>g*RC1U*`Z)~i-{;bF3JgqM@
z4dhiF9f>Avk51!0gmEI#S@<Kb`<Z9Uydv60Sus8Iwc!_x;_)Gr={dF%9&!6w+_U6s
z4&k~8#m$=Navs9>q@tIr75yT_$yn3m&u$kz@VByia*5+7PO|4#nWJ%g>!~MY*jz_u
zQ2>o7?yoE}q`#gkzhh!jcg;Yn1unS<t9{mnr>%1?_?&l)QUa^XJH|w3**&VW0S3?`
z6Zcf;Ry7UlwU43|5s&F!xK6)z(4b_LdKI?WeOC1vdw)^AT}_vp?^Re|KgdX)CVZvQ
zY$+<_p=`YgTyve0dN#xGdA975$4Vz6kyU@OY{iPQ@1(g*9PH44E*dxDs~=00qPW51
zzs1b>qf(gxq%dGg{CD;vrqv&S`)Vv{b+yx@StGNd;^rL_1p4}#Lmuw@_*s?;vKX%2
zVbb!4n9z9s?Wg@d8P7o|d4U%nJ4CBK1o6$o<Jz?TF_k+rh|u9|r!XaK_C4=X<S#qh
z)B@PfQhIpy5MiA^97)JVV>BWNnzE8V1E0^SUqa%-l<yo8joz=bckn&i9~7)>$pw78
zT`aBHra*g{Pn)kaGrN!bt~t*~ml2e?AxF$i@%qfb-rzS0&Yf-e=ew!_p{r)lu@tO4
zFA^YWokO)vQmYaSC%hNzL@kpj_&ERk_)vK9d`t>uF(8939-vG^Y$j2MThh0=3Aw>H
zNo@V^x9!>lyD>hGb-bK%Q`}|yO9qj?4E-gI8WV!5KmUVU<L>KVhM(f9Cd!z??^(2h
z;|jawMK0d}@V~Yq2}z(Y$vmI!2go)E-g#zgL?f)oGlHZpz*pc6bJB_oC<T_t7X5KQ
zweK0(#f<M|MhbMxPaTx*Vps<RceXp8R)yU6EcXR~;j%YP(}hm+GGPAD+q3lcZg}V`
zA~3Kcq@_o8>I=;Vf8!mcjd3jTiuiQZ7zt%+VI;fb1IPJX^N9J2DzPjoJit_%Ef(6C
zrBJFG_Z=6}p?zL3|7Y4R8I`*wwYMI3zp~;68ez6r#1N)f(QMT9^n528jL=mvnkl?V
zocHN_D>@)v&1K7&#5br@l)k<YT0-8ehL@7geevr%&G}vu0A@Z^r4m~^F?MlFr36C!
zw%Zhk5f^afV`!s}3*4-VPf08Lt^O@4{S93uvX^=C|GZfLFC79*Uua@|MN1nqszz|g
zq%PV<#VKFZpwQsa7JUC7G3OZEARUBkqhDieVJe^cOVSBjU?_^6@A8-9W!Bq^3uTEB
zK?DKgC#VK%p5T%LqI}YDV6zsC1VAc<7k>5f@j$B}E&>S)&mxr5m}OZMXI$UDn|7#t
zh%exoUFFVy;4SsY6v3A?ntk6Ly!Nn(H|cj3^5a=hcH}+)zFA+E`||{eISPl}6d==m
z1_{I8tF@sMFXBfbwdy6hstEk#Q4xkwZG;ui&95%Ut1h^w-$cXj=pnN=42q8l)(@(|
zz_}|l8?Jd~{&OV4>5imb5MA7eN*<jWXw>E!4@h0~)-m=wIk>a=izu?~6oe%kRnL5I
zBd{tEB9xF~RHSpCYBhWGnItXv?WtJF`U3l^{0Sm_7wh}nK?C>^B@o>0n}v+HWadG|
z%fo@oL#E;8rOju^eRk3pbe2jWDW`^$pX}8Rj3`NdLx%?0ge8!EyY_ppw_!k(`p#=T
z6M8j+wBO4gY$zBgWH%4$pmYs5Il?e-z@y9T<SWNvMN!y(cPA9UjV?SnwIV|D^o8S7
z;n|0_+iQ+id2nksQ(!zJQ|ZSO|2mV^Vw<S6kCwhmEp}W(k}@0}Etr;$>#+$Lg#r5^
z4(PzUC++eV;j5mvolm*{Ee&Qz`~SOA2|=DCbHsq3evVqoA)ZbdQ1PP<Y>f7uMQR**
zoCK<?>sT37-l&^gWA35A^M;kfI2rk}Aa1O3;z13{#Zh9(2}S|@^ghl3Rg>9Xoz02y
z){T6^Y2zlKM&+^)Ow8{nOTysb;cwM$y}+oOlCs;L^((~a<~3klwZP<xaT=tX?c9O`
z?XSKz$jkgL@mTV!&a860UztpohX2zK<eorU5zq3aMpeS{kJ*`T!e!RR;BC0X@1z!_
zmj0-x$W{|)J~sIdu-T<rIVJKfU?tjh00qW+Z8j?2%en(X;eWCR3%kQ?Wvbc5%Xetg
zZ_v@O^#2T<n40u!YjKUq#ULuOZswLZnqQz*cKiM4rt&sWXou6?@VH~|wAwAxwm=pM
z*Txx}XwM&%VPMUqU8iU2^O*aG_o6>)-+(uC<1r8!@6YSeCRTa99HMu_-^*mXC_PRL
zs7I9vrRyN_4JE|!o#$NZ#^~SqY7%|9>vm=wNDyyuK@xo4<{|L|x<P`bpz@9T?uTUJ
zfper9F2@((A0;@O_$S*>?c|QXpDyliN4}h<J5fi+tnoE)?bQIJhlgdyOLLI=nWDND
zUrIg4uJ&igISmpq+S!!_=e)K|+FG1r9OzMPSDQsmS#&D4tO6V_N(eQA&7fqni~mE|
zTL#t9wcFakEl40pAh2+Ecb5<}xH|!YySs;AL4&)yyF0<%-EH9xr}OM@Ki{c3JNy0D
z)dkdKt=V(-9QPR4Z44&hGrpxif!V?~#2nN9V*86$w_bAB)Ex(7^hvr7Go;p)`r$Np
z%RXy&-WF-8AedKV;L@m6BOOREdr>9VVc}-JCBl+G#nXGxJDT6=vkcr-uUt~|gE+<S
zGCj>ZE0>3@JT9!~#}4SPnPuk@k-bG5J@MLxPI?$k*lm1+5mMS38|K-+!KMDk!BIev
zqK7_s*ZC}}8GNd#*C(!{67(U)<+T2klKreMlI{H_dF$+&dBI>^63L($UfdMstqjZ7
zT>KV3n}Z;TOECSw3V-Rdj#Jb9bC@Oygm-n_^%esj^a=gstuJQ$@%dj}U!n<b03Z3`
zdCW``uh&)b=`ZyMy1VH&8t24M^RjHUZ?>_#2y>P_UTlTZNE2DxUIS4rIH2MaWBDYy
zL7io<V>FWzy#c_=mFUC!g+gkA!NoIxMVinnJEl0Y?LktW9#h&1)Fda=(CvToTwpMX
z9NB^^AHmCOQW*S(oT2oMK8;@1q;`<!(ex2Q31Lb=H6SsYVncQV8Hl&sFy!u_Ndy-L
z|M@qa-S|2vjzs8c=luj>=1&SSs{&a+Hn=ufr3cnr2L~a=ubHlpJsmUvAB*GL@F=G;
z1mo%RL+qh-vvYun@=MalL^HY!uP|3!36bqUOLqfqug<V*<?F4i<<2byp@3pV)~RDS
z&e~e>{!K?1=rbqLyDi{F8*WG+*I?B3b)39x;s(x>MMpp*WbY#Z_OrVq8^#kF61_4i
zS%_#f&gkXdnsihfcUlT8#QLA?AbLxR<6$*FWqWf9*NfgkA~p`nL@{y*I!^u?gEnRb
zUe<`r9zLGD${Z|o<PodI%3=7@_Dz@^B4V-H+&S3D4RG{I>yW@_XTJ4Ynn7(TFi^{E
zM+_+R*0J&PUTTG8Ixq5W)mx43$8Spqn5=W1b*o;O%2-IeTtt*X%pvtI-l$5myT1ww
zLn?n}_+>S}wZJOaN7S#2{QhGeb>nc8CohR9u$u_-^=nrqIjymby?BFIav_zZkm1En
zWLxwf-!5U#($@cX?)`tcNFgYI29sm5_kDC+G=GA!Mxp$pfMj1ag%Cvx^;5pAl+@H<
zQHM2ke=^|fUsC$Bhzg~pw;zz2p{CM0;*wLoJ4Md)UUoW&&Hi2A%jD{M&o?&s4vx@T
z{+Z8fJ=5mpoMgs4%ejM!y&$+lVvY8Tka6<IEQ3*gZlQykxQCOQcD|dYV_LdP?5!t}
zn~Pc~REVQgLmE8=0v&Eye|aeD5BWqTQxkBgc38ix^Twum#lb<HIS1##XP|AdH$W_z
z+HZzUe=Q1Fomd7~E_g6)mxC%V^B;8?F-k{0?_gi9+fzw_11@(MFbpM~yLEH*G97a8
z_rx5XcUBW8sSMZy7}j1`HXntBR%~uaecuGQ2Qz9|bvM{X{_3wyEmxARiqE~>>$9hT
z+t%vjUryg!OQ$R5uwaXs1P+u&T^<Wt6&^cVjT(FACSR21@Q9yW9N%o6E$YfKV(g4Q
z=N{gJCxh*c28XWuh7q{z(fWRyp=pC$&Y@VM&uII(Me#%Uvr#{yRhZwP3Yb&Y-uEFv
z-&O;M(wAp6S2P{F=<fo1W~($Vgd<6}+3Rf$z0|Kl1T=fad;JxykI*Nbmv`x^{PF46
zu6`~v@*?~h_T=GU_o@thc${>PW>K#@*HQ-+xukP$@)SeNxu6P5ARPlv$MRbOjp3No
z5pnCaG$C)itRH+u-V8$SbZw0|*Fp}3n{`=Tfjz@?PV6|KHfnjCLG5!dO-}O;aGPvc
zD>Utwy6d!OQ(Dso)Az7Mm<*et9jB$82L`3!eJjJCdkZV;J2u<@EE+O$d)<0!jFXV?
zO^4nkL%o}6pJ&zjWcqR6h%Zu+GKlkT?p1J^M&bMnXDj;t;<xo5U_vrEp~FUFWy<-J
z+}z%k`SlR>6r)3`@_ZL(KQ?rU5Skx$Y63+@Y4_g%RDxUZ0C?;dm+Lvh(;rgW_%iVY
z{Ps8gbmmkCdfK9Aa<-m~I#<^s<C!A$bLyvgO&oKG0jWE#uMOmuAV;4QKkxhUJ%K2O
zU4~0*Lzo!%+m{Wc7{zV6)VuD0?P!q)j!$<V@;AEjeh}b}^LN>-JhgK?aYe1K*o>6-
zB}>|w^hWN{M+fJ7Z__d#fu-+;)%&4k8YF&h!DVkVw5OU1n{eU1Z0xbw4P#i}cc*u1
z!e9)$6}|D7hwLPcWt-iC+h~f)S>@XZlRTn1^;`{@wo7X^59ak9H|qYr;CZ?%WrA;C
za86M9SV^n$<pel`GH0fx>^R|X^p+{AScQ@Hqqe}7O}THkMZnEW!NiO)X7K4=vf>vQ
z_)H->#}6sr3V=vAxsF)B@@aITcFnKY-^XT1K4B|gb>=j(Nr!;4UHwvmLZ6R7UScX`
zYsa+gM|?c>icT0%J*WY+PhoDwf=WJ)10z52#*#~jK9~n`g`{v4n4pMqz$1+^n(Su4
zdlB~F^uRIr<6KNE_g>7JT*7o&dP0-c1cM{yj$DLaOFI2`2J87_oUeaXF?hDHjKl3!
z!*zm_;B;dx13Oox&~i6~w(CCGJ$2xnlu60ig1+&Ny*%52oBP$7<Y3;1<fw6vNZnOF
zp7VAfokj`W*A2Ayu%Mu}P_RlbM^fVwm+aV*z`?<L)G-Uw>D+>^2#IhWgNp`G7+CkJ
zgV7SvI>Mb3XBQN^wO+mie$0OpDDQ-Nz~+x&m!kd-|12IrVA4QcrJ5pd;cWqLa+8EN
z+^)QZbC8uk|CW(N!~5q1d`K4O_i=F}+mg%Or;&4Ry58P7drp;J-C(g{H_86llI*_?
z^p~l;LJCSa14F;3qa;GP)wPLvoKLS*#c!*aSy-1fPmBy&kOb;3E75&g0zlGDt3{bX
zDKWnLaBY>Ix7BBh^uNgL_6Z$uj)8qiaV*dg8QqJq)4ji!A~|3)uQCZVW;zMx5V-V=
zS&j!4xRNlcAF#SWK1qs@>rXD3g}?I}M?vq5#AykxmQ3|PNy!i$$o5Y0_K4!>;5c-&
z*<(3_{&=)(S^}Qp=TLkp+Y^hxsecA?%)?{m>@e>|JOl@5w$Uns1`RR3eeO0=N7|9x
zCVN^i1QLVmV)_(k6CjsgzS(RpM9=5CzJ8;WBllOu2*CiuV8>E!1U`9cK(`WtW&+*I
zrrdfm)XIfNP2?0)dZ1gRZ_$UJ|J{Vuk}>7Ga{zHP+_Ukx`0Xp#*5&=h^pCM;V>B^(
z#qtPGR*5(^>BF@w9Rc`-8Tsy<=kPu0c&(9ijG@oU9>TyH8AUtA%Z)}$TxmapF-Ye^
zXpP)e&b8!|kJN+n!!kec0@|E@1J1MgX}ZJRr^==a^@b#tM3}n?`4&~Up2dhbjl_k5
zE8Y5Iux-$-FV7s-jkd={G<iAJY$=h}s`N<uS7d|HQtTAd9x~0KWzj#@;_dz_V>JWm
z<;=|Ul*TJC4^%tPveU!7zLy+ZNV2!8FKC1Pui{B15n#>K^lHo5+x}ejT_9RR-?-6;
z3}#G7!p)<>4RGBXUDsbq#{u}F8&w$`*DI!(Ppin*Y3#T|bX<k((vvq`_zGp2#wkHf
zjSF0d*Xzv{@xWW2c}U$G;7eK8x~t>kWl9MeYzt#uPG<3KO63gbe`6Q0p8Sk_e)l0e
zbV?47<+m$0OJ^8+J&=IU*Z+CDIxL5Lp=IkXx1I<@<T&RBB~YS;U{=@?kXuO5w^#R8
zIYO5#)6tTU{J!&qJ|ZHgIsx&eg>OQ`z)vUI1cSG(zE@Cml2Y@=EYHfdx1v_Qag}4_
zLK>rRx#-C5$K_2VbZJ0F*om(18U8b6?u+e0-oUdt9uGWAza+<Qi(aEs#R>>q*$KX+
zJ+qCW_a!eybwd*&8tMacW}L&tw37!Wvw~1fuj+B=k<e<!Esv1)Lrc{%?CJVh9nDO$
zfwe>;{jv9Ptz4|cNe|zUEWT&P4?a`tk_S3;Yg+FeG#dD8r@A0R`ykEBqVFCe12k6;
zCDt-11o<voWvJ)tSeR)me!KQ(BUo1pmN)p@v=Rkk+nu)Q`BVVvX-`%zA8FbUdE+QW
zB>}wkem1p#<AHGA;4K%Ol2x^Pn!6m@%b2BevFYq=oa!**-Dned(>BPzn?GM7tu(Ci
zSrs!q<tFXBmySNP#5dWZt*c+U#dDBypz;^4rN4{h|7>Q1fY&h^2ahP9i0|t6x&3i0
zOUag?U~;Z;F79qAWyWtSQi}XfPVr+j1-S?UH@I13Z-77vqvVf4gL<Nz2|(OSG}z8+
zkb}X2#~J;|i2TkAjT(6sMHYy}52h*tR1<uRA~dh%d-nBKFu<yKTd)RHjo9OhFs>5`
z#nL`jIDgkglee0b*MQI6twGj6<%$$mHv~QrevV>r%-)8rue{8k#I57EEg}dn__Z1Z
z5g{&$f=hRYpY5inDD^UP+eZU1+s1%a5v2*825&VyfhLc(w4Ph_+5VC9k*&ho8JUEw
z&oktms4#+pP~4F^nhc=S;&SDq9U|U;B}5T^UkPj7yhP7C0Yt*k>IiMPkkfGBFM|f1
z@>lY?tlAv5<-7Q|?}g~8^NF0t0r<^2?YHrV%p_-|RlE<De)f9}cs<XeluRxoLM`Ab
zc>)u#{Dap)QHBWNZ{oSgDVwygvqWl5Bq*`1y!T$<Q`^-gQ$&FQrkjx6FLp~`L>`Af
zZ24Ul^A#Da2h%g5ou#|eC!;&8l}$eB4HWONmV3y$V6UgHvxX$YZ|t&$SX+BNw6>ft
z)ccM9nC-~NeVsw^ZVr?BOEuXS2Ms9mNYZ_|D+H>tn?KR1s5+q)_dn!{jzGo!iYB7L
zW{3jn(IkC$caA1vS{$>O<#UW*hFMi53~6J*{>#YDT6R_EiGaaTQZr!IZ)52t5iMZl
zX>IN~wDE}MNe}Yzg+>-N$w##75swWubcqo0>GGBWI$%UNHT;mwZee8%Rw*(bj4@td
z4n^E2vqwOq&C?VB8nL{9!}JoCv0zuZ^R<DB1UV<tIGm-+wNPCq+;_P&@}X+2pf3JC
zE9;8fUAA4K;U0YShz$U_gb3c*BJ!MGHr8+0z8Qn6|C5dCi|NrO2tNU^IvWo&lJ>RB
zfC??7y#9wYmw#ZWXII`KGHiz0a+m3$5`#Aw?;aQ+@<QrjDca4r5&5L$e!5CFfH79=
zxxFSN&?-E_CE~*s83;Fa81Cx_eIv)|5S8N4gJz;!9l<LhP>$dsz+MQ>K4VL!fW!Zu
z2i_uUTK||S%Hr(udAhgxylhzZ%SHqH70J*>F1*mBnU0RK4uP5r_DWG#ZUQ_!5y)?*
z#8qIZ&yojno^!1CZzM(E`9EJ8mHNWZG&MXfI(Cz=7G=4}bV-VI734|)M4&*;n(6W@
z$f<btWPnzvu1sfQ6959h^T*2p>7vK-`xzG@Q^tJ#s+Uckv}k-DsqL<qh$tfWYv#40
zV3UZ{?UHh^|6rH5wr$cS>sNdNh_nrFhd%cPw+|gWJ!ZX0<-5z?yr;C^900nPA=<Yp
zmdPA|p7c(0{W<1mibac0)c)!Uh_p5`7Yt>K(iphzy7Nz#TD^!%Sr2?@8{FcD6Ckf~
zV#h44gW{5@hvz4aJtD&L&;fBuP6kkfe||T9Sci2HT@Oh7(RaPR$Ve3Km77bB#0#L4
zW_3{3+I8jIHIl$vhUpe%s5(aA<*H-bmHxMD6#KZX0U)J^+Ff35Eqf5x2@s$j&`gKh
z?R^+w3Puz*=Nh45%>n?bb42SXLud2=Z;|c~3`5>;V=I6-5`*_Dz8v|t*mdw!U!v3U
zUa+eEuePm~*<z_+LTjPvb42I)oUX`-9W~ddr?YFhoz?|`hA#u_UVJ-~iV=v^x9DE-
z2f&dliqGv7OMY#*`f%(JLM<cBNA?mC6kO@q*Pu_noL^mLkKbPG4V;N~JvH=r8<sZ?
zQyd%rmfS$&gPh%>0P+e~xTR)PS_B!yBky8AkV{16gINrY@5Zh8-Z^+59mdg#--YI_
z8%_jZctzk;C8W3lSi~al&?rSPrC<Y_^pGJWI!5#?u;V!l46`G2o}43Vs}xaQA0-nY
zakf+v2qX@OmXSy&-JY$8E$$hKZBO{f_GsJ*nKrgRpQ@1@b28xSuzD+<h14e3{W1ul
z)r0bPVXN`BzJ9x(P1MEhwhPCM$>$}ic-QIq?s@Hdpcyqcehz^V++n+p_M%kYuHyG;
zjqhNnFP4BrwgbE|PpY;SEoTD#KCOZ2s3VW|nNhaXGk;^;YO2EYrZ?o0*hCd|<j4pa
z@^o!l%KCue$nwi{e49z0{|^Fg!5c*)$XoVUx@le|H18D1LB~T&SpIM}28e86leDpM
zs}Y;p+7p-(h>zT*M-f)i-7qQn?vNjW@b&;<0&O#sZL8Nl+U-1NCg?(>V3_7CIq$yk
zvjy8L`q=4AJ2!9`y?1U>Q7K!Zl{1}h7aN`nxccFM-zD?6)f4s)<S;u!u!!Yqcwavt
z5te;zz1vBBg8~bP`6F!rD^kUBX{OIQ(`F6Gq3XsH)7`6la7h1kI&|qiNw;6aGY}N;
z$$#(rQq<k@FO06%wN~Wvc$(8xEcs=e4LFQlcp?~x-zTj;ZA5cp3`ro9d^dW+W9@i1
z`ajuxxdO>mOR?Us3Uv)R%MSf+Pse%phF;FgpGb_j00*#{iw#l!sM~ebXJw|djMhi)
z;>2rUV=+Y|AO=cijuyp4P!g~*@3VCi)pP(<jz01k$R)9{lnh?Lb>TzMvrcOhv;$~U
zI(nB+2%ZJrah~4-OaS3n@$U0oXG6==^u*!$<cRlMQT%|wM}rr+@!21yTw)Zr;qxkr
zM@NCfJ865X;rB-FK-DKn*R<jzXNKH3OF&sh`P{3Co1kfBvbS@J!`b`0Qm7v=0f<v#
zJvZaSEXwj~c=GL3f$2UNN8-EyBq1pQ=vzjCu!=lOc)0w0+<cM!-Ot6DFnE&E6K?i%
zn1S76z|d`=6XPoIE-4*!b%x1!dd!s~<?_^W*ZC0$vRgkJ<k}0&{mliGCGy8jf$7Cu
zZSPpt&#|OPr2LcC_wsduOLM#Q*se>C76uf?2hdKrRnwSMZk(inCgL&IVX9?Q9t5c)
zJ-ov0Tq2<}s&@wkt}<4)1LhY8^#Viw3?n^Mcox7}CjzfA9SFhOj5oHcw+N?AS{I7k
z`Ld}-qyrVsxGmH*fdX%ER>QrwHFPWq?FtaVKy6e@QOS!bWANsbKoc*nA^)MGxP4dY
zDVWF-5d8-`fc4AROg&KA-MbEv0I4^~?^_|47a?I6O+Vek`Y%_2mU*#Q;_Am3Sn;rZ
z#{{<@xm@DVDBA7b$uc54DnhxC!Bx)qvb6UV)&+WFp;dz7rnQC5BKcZ-lUKGJfA1XZ
zAq3x0O@WMx7#!jj*>X`i_-qi_DBBwIa9GQUOklpZauJ}X#o9;t1U=;sc3UWvkd(M?
zsR^n?c*iNy>de?yqOT<+@8R+w@v=90YR|w#PAI!|3(+;W!3A#&|LQb1?F_@U{ny}k
zB@qB%1wx-A^f#R4HVJ56QMa2a$Wt{Ij(Q4q6d|T%b7W6H<j?$pVk*H!ENE1nG_nmh
z(V?vkcoKj@3v|m#;Rm8I{J{u7KL*na{j`I}kX@=N{q#~YIEHl1uWk;oVCEc0%KOb%
z!#kcn?#Cp+{SNWG=0EId22it-NdT0AK69!c#ok$J_34<$b@zk;YDlvPTx~6cnyLoE
zx1pQ&nM3BKNts0hLb@c|n|J<jh0Y^w5X6y|cwK|L_+l3ub79*j40{|Jf<ny)xDQ8v
z3IkC(YltILEC6Wm#vx5AL;yS0BWG}KyJVf9Jk#z;P;a%$F#mGgtz7Z$Ts+8`QMirh
zZDxXyhuA1vfbS!~81)VjFK;?!_EeUPlsz$Oc%lD;-@yFI`~;a>P2WOM?g`F~g3PL>
z|LZJ2l>2X{_@TGW+ep(IaCQVW?&rM41}3vuD#P=iTls}BAouh`bT2}mW&4HLsI}QP
z0Q(Cdz&otX=2Y*52Ff2_Y?6Wp6etr`uS-t8`Nsq7oC~(-d);Y_UP~Qe)grH$=e)~^
zxX}}+RuzK}hsc=~QS~rf?kPc0mWjNMa#88tPaEVPQ>*D`qxX98QIOKKz$kW_H6pE<
zo&i)#5`j?l&p^q~xbZuJtcX{NR6aCCnKKgHs3C5%*UW49)UvF$f6{&bmi2&N`Yd#u
zJEq{gU&l@zU^-yXrMsibz;R7}Ie%QQ2-bm2<wL_bdb~X%%a6n(w`A*?{S%mG)JF5C
zrY`#NsnO|@UA;8(spK}6uP#$6gg@b+bwA$_=H|Id`Hv(PXo-?0;H{A6-qf}7e0!~q
z%WaqWss@iXz_(TB&7Vm`mM{|(>u?rnl-wo9NY+9U`EgNqeIw*{5HH;;U;i3WIYWwF
z$0raT%1rp3m1<{y0u60L+#-6*7|-_C52(Z@@bc3g4O~X@y1zI*bTLrMZYZOfuQ^+`
zM++gcn*NuS3P?kB0WPIKRJ?e01mDY{)A%szVX~<>xPf@gKf#2oQG8|9&&G~%Mx;Tb
z=O<Qk=FJmbk*e7vbWf}a>xSynWKS6c<V#Ekr|d4RS}u=Mkk*j-z|z%OnUKyc4&!q=
z`c>NCO2Bk94VjH-2GXgCJ!8u?_myd8l;Xcx0Pu8Au>;1=GME9}msRq_-Ir(Hbmbnr
z9P4NvnlI_x>T60|UW^moH#zSPEx(|RqgwpF^JwE`+q%rKbAh`)I!Ml`=ZV(c0W?F_
zMq`T)@prSYjr|0>fKCZ)lmQ?7(RS+QOgznbqn3A^9{=RY(^8ifWU2!=ZT|@RDhbXK
z1-6)HU1Nb1`0rwq;q@o1+^*P2#yJ_h^wv)8dJha!#J&QHHB@<vZn%J&2nu;2eDh~3
zRfa}+{W(8eE4|Kn|C@M43DCZP+f<!!j4qNPbxNZcV*pCC^5bLc&C+7jOY2S-hUa}Z
zFt|U)lZ}g^`6jf)@v;QqJbUW?;(4zik{DR+#Ptin=Hx>w|H>tz0(~bWHRj#D>vX^D
zAwh{c4}Qy>X5iw@TY83S?Ix5!pcIwYHSvoZE7&Ze0w_M;ck@j_w{I=a5-~^-7CC*>
zOxA=q+*9zRu%@APQr-Joj#cnIJP_HrAd?Yk%_R>&8P-i#j4Obj78lFtQ!7UI^6(pb
zi+oU*$s~>IX+hdGCIt1E)N^$j(^>tdkzg6?o8OCfxi|IM$V;d8iUsOgeMiWL)6{6`
zn^T}1+c&kbQr@z3gU%P#@ig?9Hpfe`3b+lV0;3eIF44WBBDcp(@G)Y0miNwC=(b5&
z+q5G^Z{W{%Sg(-%-u_&0{~%h3<Y2Iz%3=u6BrlvtPPX2Pm_*)QXq_VnFcg5RS5#O4
zFO-P!kKkZLXa=qJgzM$f)a5@ALq<cEc6AH81{Zw)s-v+qO3x<PUHd=aKP7zO*+#n6
zKoEm94Ll16umWEtTtg(V*xvEmbrnEC4pnUIm<3(%UW4!Z@B7W^<CQFu+#=2PdxK<`
zlf{Ye&XWyM;-Vj~FXu#M6TP3#WYlX6N&yS&4l!ouAFhfwHt$r~!`$V|M~V69LRVv}
zMliH*#vy9}u0Kceh}aH*X4%N(+$qSs$k|<|rn=mbsd}t<d!2@GYgN_1XH*yM&LF^M
z`eRJ9O2Nwe!BUfMyGp1l(yc$m`FJ3M{gvg}@ZR)vGsA9sx>4zpuMwGd^~IpAS#5p6
zW#rcu8P#Ay2Wl2mPkPW|qCV6%;J^F5F|cI6(p?U}Dy8ET>-Hg)>6Gx5ez_db1&Rza
zf9v-BjhEoN|A(*f8Ja*KIt@rPBNw{7^EeL2a~>D`#q3tFkOI_($$v+#m?T&#DE<zh
z5SHLNK}kJMDUM0AvX_0~x()mWawIe$%tvWcK4T?qZ9kt-g}Dc{yydc@pKT#ZbP-2E
zV`$ztO0`(6R_!RG4u5oBn(|G99Yt(*wz3X*Yw}?>dTn!!VeUInh~YeI=-wuL4r*Q^
zT-GWikjiZN&Zv&M(eF8u836L&@n!(J`~a0WGKy;U4erq}8|nIkXSKQ^%})+;(@Sl{
zyx3h1>gL5y*1RSP@{jY8p@!0g?jL$+w8D6<KH9ZET}P{m1mQdasAcI@ROe)G&uvUh
zv2W!tX%HNy?Ij%>C7@8w{SwdtcS~>C^EEUJ*D3#Qgo(p30}8?8soOmNm%tQ5rE%LO
ze+Dngbg%kp-5=)uF0)Jw^6;X8W9_=+x9wox_A>1c1&}+KD8A@fAG$J27c0mHj^HwD
zzJDLSQeUK{abPCcD#>&LJTg1)G1Y+L>^8uNNivyGx0+irzhiI--1#~aq3r@*e`!3~
zCJDx2c<+2a%+em6aFw1QfqIw9a?iLR*NeIhD-yw2A2?Kh^=ma4?#e5GrN;ZQm!g<)
z2V=oPn|IF};7q+aA3Cj;ayfWD|M5>prJSNN4zIeisz7=LC7}7~&1ywP->Q4)14my^
zL1hnV{QS>E9adaJ$+>y=GXv4eOh`y9QJB~)Lj1b3SC89~)Yi-0Rx_SzmTC`<=-XA#
z`j=Ht<1>!EW&df*b$LAFIyBw#mg|l2Jv^<|&g)cx00dV3t}%axRua8ERJfv@+0cj`
ztUl;8INhQFEE0l42wIo*&vrL{<vU281aJs86*s=hamaG5>@stQZrucZ7S2=GnQi%V
z!$Djcr<g;;`-_XL%m1^{_zi-Q?3=E!@o;e|Ny}2!*#1#lV~xX$x_GMu<!LN|K=QkJ
zr27?;CaoZu7t@Ge4ajyFFi0`@GcYm8iG%^k^%-m;qnsAb3yLdp_aaq;$|ctT!1eL7
zdg1MneKLr?oe)>%*L`EnVfnzY12O_UsDg9eBv=Nf>pgiB!}F0pyVUQSs7lR2+r!ZT
zP}E+T$=Y;_6igOYPv8OfK`4KwbV=<dwy~bfC7kvgJjULDKZ4EET%oW+<9Ai74$;Rk
z<txF(jEXdkK|SJS>&7ia0W^?2y%vct6}j^B=1^8H&5pREG8>KPmC#B>mCqwBys}lt
zHYYT^^rnFU0sS`nNA^z@vT%n^-4ga#h_iJFc6)*RPt1C$B+{F08AtU$ydR-1rM$E_
z8vNZ}tMBtn8;V6z+lqcKa=gk9V`_yFLE-*}KDPnJHpgFpI=6bQRj|H(LzMc=T}yCZ
zt1;Ce95@1EkG2|XUJ9QB8O%ywTt2CldcJs!U5fiaV29YN;DFblE*-)#OWe2H*D7W)
zW)m(FPG&(hsLmP~N<AJ=aa11=1aU}I_i@#B-buZi_BBVQ>ELn9T4SKz;O}x=wV19j
zwXdjqPiyQ)YhQu&8*`0cbX<T#u)_1D^uxuE?@_}f^qn>W>kGb*<9WgJ`}KLkToqAG
zH-1}NX-w;g?+#N@_Yxf~2z}TK_m9h10^K+Bn{F_BzH$f>x@rez=j*u31x9~0EQCHH
zRCVcElu0o6X3TrqXu~ES+&V-D93ty!O@d;UA?`2W8@TG7phunYy1oGxbn+*`*yfwl
zz0Jw3UtTC_NH2&#KYACQa->h0gR@rWP7|-1C;e@_pIgWJOGg5Zkg&7V<laJfzSe65
zY;tw}t?f}19;C|M!4iklC>=*_uFkhFhysC>cR!Y_e!Rn*ok4Is*5nXqc%Ml4Bu{ad
ze>jUtOZc%vCI%V9Q9CR-)sb%90KqCY2=$c0u-@Va)zq%9GRiEqdWFj`)?!Ui1H>g9
zEwk584MP1^bngx$W-r?a-Bm-#8O%56jgPn`vy^iMWH97-7%;D!n{Zu4LUgFhZc!j&
zgjxJ9Blg9M#TxaCiQ~YpmPp?Bm$l{A1+}nR%lcKGA>XAm4>iQm1e{c`-E5+FQPl*W
zr(DC7whyBacjU`s2B2!`Z@Gz3wW##79#;3((4~9G5#zW+Ju1rzH_~gw({gE4hjER0
z)Ruw})ZpoxkIK(E@Ys0sKEeCpfD*_1pi8|c??d?NnUH$=jSx8nlgp<hO`+i}DvHj_
zFa7N+ayr;Lul?tteG!QXw2m4ffc~`I3(R6m*HsZzuh31&><j#5R9bKbPaE8pc~~k!
z5T^x<@s)MRWk<;@{K%ibZW3;$l>3pF$UH%8x!xC%Z=Ircb)}BA=$B25CP#jkEZR5j
z%fF0$)~anudWYE<1*d{x)1N5ma|bEZ!7xoqgLK<7u~o-5*Ui8!B;SkU3j-(m>pJL4
zwYcodmpRJ@eir^zYud!|*9sq#L04*zxA^lM^a!(<nps;|elVges=R^bObs~Q&l{+w
z9BGCPeS^YL4RDqB4w(Bt5y&c9njg+04>dTlp=bT)h1HM|P4{wcO_JRSgg<+xQ!h9o
zEqiwx{T|5lb=3bYH&Mx`m_CCBK5pXF??RpI=DzG+xK!qJ7oeV9=W;kOgF~3zA0q95
zL?1(H9>qIV4p}Hm=3nkEm%dlyPih`t9&<VCSTwm?FGKwydFK6<woL0CRm=?|>R)1J
zf}9XXhoyO$<*`yF=VC5p%js0q8zz~uJ!?5CIO+S)UYUWK^ju_x@OF=i<%9Otqap10
z7{%WA6Ot1iHU|NNPI!!=w@1~HRl$<=iO2l6sLs%N-4EPqHXpM>bREu3Iv9r1y*!he
z!rVAyLu{sm*SoqdPN6H^2_7qW8^BHOSz*!LNf7k;eVs$%k7uZf9YIKC-h3}b)3Zvr
zM`-XLgdpMtD!}s)@f;C8Sh4J1K9bucl<SayJnA~s`1Lk9Za7{y8cTP6i<7{@*##!r
z1|y=P$vqueH{=ROb!MHsgkB#OA#OP_NYhptx8%mOa4bLi@v&lbtnt{PrV@g3PQHsv
zI1{FI;jV`tq<@;n<^1$E`h>am!LK9Hti$NBH~(Xzij!J=M;GarebqUMHs|fM!y;?b
z3`Z35etlcWIW;uNQLe9AdbSn9T&i>@QLC`htDs!sx3jaKS=%h#Ua{yq2<>w42)R2A
zjzakAzI@4Mw~=l%-t<z>5nf<%Yj5K`=x#XQ%By``cS>3H=dr5OU*|KO(hoa&@kLaB
z^|-?9e9+K`n6+={YXG4KKI){9M#?0j$3Tw-R7WgIooyjQ^%u$T+Mg!ajkFVR5W|;b
zD<JRby4UmHJ`9UdoTPulGE@#RK@v^`8zT9ediu`#x(~ssQFtEtXumaV4>6e{U3}ST
z!!`_%Q#<c)vmEegL->#>&1>IZEJoO;8+FtmdVTTnZ<EIjns+1r%QswH%BQ=D6{ndG
zF9nH)@O<@r=ih`7`Rg*(rt+(JomvUpN3GTpUkYWj+)f)TJ}wu9%lqalzB=F9TlY%y
z8~WXV`ReK<W2~4q;Am1Sd=OOBQY_jgY(DUL;F}c`v(FNt@vE@04f8sV1xe~O@cPY&
zQA5zzgTp)<v~e@q+kyvtLLH>h$I{y8)u=M0+j|H0-yPu$Q#M57W?(E=s3f(U{<1X>
zd|=x6fiUIoyu>`pmSQ;<a|ny5B7Ao(uo(5tr@&v*OHF)TKuyXs&(e|NMUf8aKyQ#)
z#oMxdr0;0B$^YRqsMb%o5s;KJV$MOL*apEnNG0Zw4mjPfnF*`N@6`x}k2C2H*4>cI
zXQm6UAxj)p3dZ5|cS%b3v>mD7Zyl_B^eXq*li@|%GZ*RMO@~!Wndmph=^MC3I*guQ
znocF<o8!$MBq*e1Dc>|X!k%QGK07Vb?kkyo>>G>0SVQ8omJ_7(Er2?=E{G6Pb1G4q
z)2nW~o2;lWd#DKdJ>hPh@A>15jF$M+bAzt?eHJBc?v2V9j4edK*atg4n`+uDKWE=~
z*9RA?`PJ`3*$l@tGh)_<#i{l-m2%xivsc_!)%Zf=Q3_pJ=CPRDub$Bp^MMsD*L_EW
zdEX$QHu^N4*y;<eQ!j!8KlWiRzk4`a%V_KVY`TWe5f?sViPuLZIvhXiWcVCpGW?i6
z!eY+ZS(^>)=+I#)C+sU7YTq(x)mjMYPUS-QZIVNu{jYv8Qua*t>(ZZIKPp>5wtHTy
zCx5FL!^#PM{PZKZ;!Wkv3-%g2F;X9ii*Snwzna>;_S>Mitv-Jq+B?0tuTW`FnujcU
zQl4JjVkBu44gJxukSuBG6>r?6=yn)iaQ>DTw3F-50GPqOaxipW2EM@v0=<NHJvTXb
z5sVY5LpfH|e3!JLN?ms3G>Rh$rR+Hhx@A*>PoEQnxGjgSl7uckT~6jqmH2&1rov+-
za7F#loLgB}slty2TF*N-OqJIyjXAOsbZc1i6NHL}pId<qNymP@hEUIn*IIbHa-dt~
zlwlEIu@bgoX6}4Jeq>TMDzeV?9JOLp(zI6t{FxY{-y`Z3oTCr02bHs!%a~<b)v<mZ
z*Hg)$?@f_N=Js{v>S}yH{ubzh2QXSDt$pH!rDs42Os*2<;4s`zE=?S&n9R2}@y)W?
z^L8v$mbt$EGg%4h)iK@_hJtVHP$0!Ag=PU+5OPIQ<5eEqW=X}dtCc$lqFN;yQ9$K~
zFWEx+nd_<=ANirBrDL4db*hN5w#f>a+L?A%mwui~Na*1f*WXGX$M1Z=EW&~5Ne{48
zVQNpblg47OH-!jbzSvf`m^dZ*E>xIwqlB<MKzp|nTgP#Pah^ZbF2me7J*MDT2JL%|
ztp$lVuYHbKaep*r)r!@=+H(&_Dl;W@c4PCsj9gQ#!>bmmiBuGL(Ry*&tEN(y_k1mA
zHA<jk|GpcT%vYj6n_63&6=)m$mxwE&33Lfqaj*K$caN-JNPhGXm8O9(Abl;}Ic3@5
zH4LUTNDfD^55S3Ep+LS2-j20v5htu%>3vS+xdihWIXj*kcLP&BLHBh}(tY4WLvTIm
z4Bu?luu8i($%^C9oh}Ys&4QYM9E|*(-1BARgWZy1xmaYQ$X^1e@`=zT#r6EuPox-b
zZVGRQn!$>iie$x*Sda9-hX+)EIKh<)WdB*S|Hqc4_=(ff2}d`2WoJ!Y3iW)G`3x^R
z;LYIIwMNtj`b{g(JL6xR*ITCQ%LmWUUM`bfUt0;5rtX_$$KvXFKi1bp@>$m>hvSk9
zD5?57X;U?XH1Fz0B9(ak^|3)r_Nl!)P+U{RZzYTO;&y<!F*2)~Spr;O3y1K-e;V8P
zo}%Hfc~oe5JrFhZ$(hw8nx%UXy?S|MmCaOl!E@h0&)OCiQ6?~v%*DmruBivTHLb|s
z9{G?yqziJ9<mTqivi&nwrzN`@kJ}KwEc2&~eumb|OkTzpOZB#I;Y}V#2J%H)`AtQ_
z+TG!?F*i$Jz)<i3vNBiu0fQIoQfECmXG4R0f18NH#CmjI-#p7!FIL161@7A*OMpy!
z1|huQZ!YP%LoS0c%bb!LsV85UK9fuA;YrM5hmkauLZ!}pUC;0mo{Vy<N71F#*;S%)
z>0eZ@o)3;@W|m=n_9iEUK1`RX-OdTxj->=Y$na)a>l4qbEp~9t%*<mBhY%0Sw_xK9
zRKP-uyYcL<$8l5_4KHxTOt60&W)}tU1kdkJD11?6V9VWsnbJcPfU&2VM*}`z4~&k0
z=epz)hTx(%HLKiES*P_~i!eyxGct6{TT`wV$9OE;-CRgrKbhOGF?@bQ6~b<gr7c;|
zKAu%ImQ0oF41GMR%s3TU_hlQ8joclYDwm64Xw$%r{bBn*q?P_T+Q0xY@-Z7suS2|t
z;fp#fGs<$>ka&eVs*vFhaeylK9vuQ+@CfCY1`h=<C!P=a>vLMscf{Il`Qf*<yS_VP
zs#5XEp%0Bn?eTG1`6W^Gx~of&5T-rc&&J#1zEu$s=@0xfeQbmQUF5aLzL?{F<q5;Q
zm~OkYS{5aq7ptc1e?t`zT0)i;*T<cdDyXVNJAG&zh*I9j7xBK(D#FN?+8c?R?6Yk&
zWf=ai0QLWCB>(^K_yXw5E;(gSf7aUiTOeAmkOwxDwE2(u@Nqn#qi23yo382Yg+4;8
zau{fd1yLX#y8dsU<G=S?qXntE<oF+CHYh;T>i-$~_XP;vvv^#Y7FzG?%0Fj-73CBB
z!~eG~QvYPeDPbYgz*Cf%*NzI>3RX$WM5zkYWUiL{f$1%qCOiKQ=5H6hK<~}55MUQ6
zw2oqqbFaczsj2<s>1w>#yDNPs*-ZBPZ^z#M>%N#MVM5tz68B<G({wfIXE;gOfb<Ml
z2Pdbb{hvRdk~m>%!)*g42-CN6v$I7$OI4ST)*bQ5&kpx_=XriJKa2V~`7`Alz9}fQ
z@BL^XU$Wm2Vw2%G1&>vYxhn$xL*j_)N}&^{a#~@0ze8)7VAf-e^EInLYiB~@qszB`
z5s{HGl62DvI*s5t%+2OHGPjbk&z=JgNgk_Q$sJNhai@HWY~G~!Lo5fsWw+BuQEmqo
z=9t0WEx{znf?T;8*1^?cu)Ot_$?_~elhOF#IU}psGH-fFYc1K%-h>~Tp2c2NDfp^;
zg~~vQFE0Na$cNbEsd-ybX(pej`c1WQAmTFQmRbZi9BDAPuEEtt?K}VBt9*oNRoz5<
za20W-X=L(jaYeCA-P<LshL06Hi5?mWe8DYx(~cue4?YhqP0~dTyf@n4>g(`Zjgya?
zDHeE8_L!m&pS4#s%T&pq$D=`w#ms{%&%C=_KJ4p4Uj&Kfw9r?00*9azD84i!?O#+<
zt&py_reCK`JCn5OfFo*d3syE`@_EK?wAKt)L_3Z@piyp{sKT!3$G?SN$DbfSkej9`
zDypgxNlH$V{)sFeAsrjLR|QlZyUKG*@;)Daqsgr+AvO!8QKgM*dhn8*GXAMosdMW)
zGh=2#2fxw(^3J$^6*RR9if^^dTGq)_;(yB+{bMlLL0G;~$o3%iS7gn-hvdx_oJXZL
zsJ`{oE6Wz0#XASVw%mFDu8fj_<uUOS;)8r29HVfP_rs$3ejUR^Vio98=z6|t93^&7
zW`2Xx*5(LB$}HVgf6v{pWw@H-h9TK===$zc?vv<NLv5+Z(2&g;><U+?WI$Z<KEI5X
z!Rc5~E!QpEUKuR7j{|!*>s?8~R(p<1Q1WHNG-oLMiiUo_IMHhCJ4x$aT+KAqh%^Tz
z)<;UtpHv#}7BTK&yQpGBa}ib~8bgGiow{Ha<i%@DTa=`uby6&$6kM5Mha&j)cG(PK
zBCLDFj)bBM6o<uK(L$wNqYaNq%9JkkHDMtJvA!uB3hb(}9KD{jW|uNPehQlcI;X4u
z#!@y*Gqs5*#qnp?2kk8Q7*5GObKIryncKT)l0<NgKe>Jh9A5QkmU!?*U$(C#3kGp%
zBg*BND7ZS^BXv9@(cd)_C*M}Jj|iP{<_LAf!E;;7UVzX;&??S^c;lILG&PrUYg$-o
z01p0TZ)53{LE3&><!z3Pzu{x5_106|0xvuNM63R<06x_x;{|4SgQGg0l5)O@h*f#G
zJEHY#OQ<rsRe_M<Jx6v!d2518%ZzZdepE3Mju`xe8wnq)M@yhO#=g%D-FyF-GWx9F
z@8@bSk*?W!3G?D+=b2IfvfS+#(T19TJ^mcaF?aIyo=>FXc~>OdMdJNWpu0sO7D;ri
z$#~n&A$N(8c5vs<9GUcbM{0|C&J?-#8r)09gBNU*^=;fq*ay$|D~8S1^`F{{TG`~~
zZKnwF{RHS2Diey`U+K*)<Ac6`zNG)0n}a!=&{uN4HMV74{o3-{pzLi0p6SYEwitM$
zUOHDO7t$-2WrrkuckW!$+`n>PQ0sWREy7h&R^hVLx7u>uPU>D&ehTLyH#^+KOtJnt
z|MtIDo`py!ZyLX$i+lubo#eAMLIQSr#WZ$-`qpG%$zd(oOSGx;_6W`$6!##%HrjG0
z#j`}Efoch#_n|UlorO*P-H~2;<3`B6h%N1~FQjU>ee^(rZMLG<OWO0R#Dy~{LzBe}
zeGpF%LP<d*?|T0#f5gv&9F;vUFl3w+n5M$nCtbFxcx-W_J%pi^xYWMT>!F{?t+zi;
zZq36>aoQj<1t*SlMg$G?k^0u=zvH*T9hr&A^DG!A?LmROIP}^ckI<}v9kjSG)oq!c
zIPFh~jD3e`#M(}!y(OdGRJ(VKMOOh~Chbi8jsU-=2dNC`(j_*!-zR0VC6hKFP$#E<
zHvQ?GukE1F*{;f2tLpN=FI1=8frZ(#6TRaSs&iF@kIm^4rssxXqIJXMXL%j$Ome}B
z(81g;=a1uoVwydiCvUv38kv^WAN-@OA(xXIPHzZmHF@G$O3f{Qo7}@*w{n1Kfn^BI
zzRx)QO&Vq(>61}=uukX=1N@n_|Gb6`JtaT8@)_4qZm(QkY!&s>9ccLm2i5ZF1+B9U
zhsDx{XuH~59$Q$G*-V6+VZ!5@EuCS|L16uS80G1(HJ{&zrExbfrJtXI*(`hH2$U~0
zHy%Od3E!(tlY->V4igvmS##-1oTk4-<9p`~S?@L#*=1Uj&g8V_PlbsO6hpBleVBPs
zOe^Mb!qFcnZ2Nti(`|jl;@{m8pPHCYY(y+E61?15JXfAc?6vWHp;*pYa9>`7^<MS(
z*nEOqytajW-v7vNpB}SnlI+a4dAUTewV+-!_qM<8M~J*3J;6(*ne8em%iKIAsS9qu
zxEn&&t2nc08xfmr4KF;G4mbdGoK4BvOT$bIZ^mt2X}ENNK%k?}-nx*4q2nGsJN-@u
ztx$H!&qeJ@jS1ll^-%&jfUJB>$BiL;4ok7!ET`^^BSlMi6e`5bye_)CWoP9acC6V@
zsKbH=zAKx<&h@IH_AQ#tuPO^+KJio?SN5u6e1dljkvhuw*hv~oW=px=&BRPEpanN$
zGenFpS$?>SK9~AuN=7Fmoo!pZ#OHjqds|~hScXB#U$6UGoGq7|DkJsEe%0eW&&gni
z({|w3Ou^T=t-)MpH_O0^EI-`C-^^{C)^^x;eUv!O^{5*&-mIGlADi{ca_{DEKGzy#
zLv2|ap2R^5T}-88Z_WMak-pIn!DriSY}jH@LZrv(a!aOh+Pe4KF_PEi<dJ^PaEdkP
zbvstjxj$T>{r*c--F!rJ3VYidq}xl%4)Wa=Nf(uEjXN{duqftdAv`V4?l}80NLmkS
zsiCaiB}0$enyrNblW`P0wBYgSG8J*-9VWVAah@;bui-xt{go!?`HiZyD?LJ$W;l8(
zhH_FYb_a`qHWB)#aGF^T3!ME4R_v6!^*_=+GQ6baJwmL-fGS_=b{`w$PuX5Yt;gA<
zS@Ue@kNs&2N|YM9*(Aio=E9ibc+EwUq+QF9EuK8DBF$g#S6z-*fxw4HWy;FUIt!Nf
zZ_H*M4Y0TXW2W^k!xc!5w39R#8@qaD0X0RUgJp`Bhs1XEDHz)XUGzn$g{nJ`MN{r0
zwJ(n*-@%M;)zV5yK`&1qY$)0f^vtp2Pdco;Wip>Yg*Xc(M{FS!loIwhU%F$l@B7{1
zTvFN9YtyYjL{|<)uAWSRc8*fx*Aia>(w*MSEn%p33cu#7F``I!h(~INfl`EVpPs+U
zZt<dvK^`&+VPF;YMp~NZEh(k<53bq{j+!jCt)?+Bccpxpb9X;(gA2<{IBl-vIgxR&
ze$N~udUA?_x=Tk3`Sq=M8x2_acpAI#d1=B5skk?AB1vg2wXlX=b1X^ea7ZOd*vrYd
zIFgov38IC>j_!%ns&mniJ-WF%jd<dS&I{dFdM8u$6RLd;42M~h`7x<MhkYT&C*zhJ
z9St`sfJaDBo-zH^fO20HwHHRaThx&H!Tg)<BlzPugZsfWst3>DOqn@`{tE1G5OE@F
zQ*TcK4?@adF#XU)f+024g*P}%<@ZcdE#_4tV*RI3NKUo37l%Td+V+g=J;z8zs(zVt
zQ-Lq;7&}rsABXw^N@{5($sz@iAHM7Ln-qCUtia=3Vm7t;;Mve?8nSU0mSB=L$*nxN
zFL{v*+ZN3+y`urczV7FE8!2>aVUUckub!Z*v{Vs8LE-V@S#o=l6Q?6nB6h~+KBk3i
z_PU;H$B?s!|B^NfzQY(S$u6NG9923LTR^-TupbFBEf?Snd;nWWFvc+BnuyX_T8Hw_
zg|i`TvWA}c)4|=3xDzQzh?@V5s8dnYseCdF0G;T;C4QqAmpjqOvDRsy!0y@fx*7FD
zXn^?9m#l3Ka<xU-nO|?J^Y4|j#`LoAj<NJ_6IkXQ9BnphY<EKM;Dg=bejGhTf#HQr
zU8>~~a~&Hh(`?r?hB20Puyz?E<^5?l?~@^(i<VHZLS<y~lbC2dTu3d&)w@6O$g2Gv
zv8G3Yet<y=Z1{A>8GDaDUx~u=5_+S2?1*`Lt?$4~2h0+)HVO8bML@+~K<jO#Scg4f
zf!CSEQ#-r&)335brAdd}(`WbfjduFLr{lwC)f`5*koN4QR5r(|iK3xN6f>)D%wm>|
zb~y_M;c16awYpVbOz4Nf<vj0|REglZ>$GVcv~w0O5Dm3!73L_qE^Fn?0=dGsMwN?m
znlgo=vg9G@yqavuCS6^fvXg3=&C7It-rZD!6!rD)Awc5fujhrU_Cj-v6Nz1o6$0eX
z#<|Qvg48p0H_$UmODV;54YW_Z_FwWfb?JxB3Yn31P>a&Gf9>-lZ>=h(xD};Uc<mAs
zv|PV_eRwm5U>z^e#Dsyy85s)S+^acV)~skCs#X4$<xC%)m@ci563wjDb2}0_dJkyx
zp^(CkfyadK{EOHc{Aodlo?-MmM?rU8FFc<ct{5@Om}@M&ITi!49g3rQO6ejzUbNw*
zl^|wPu86>H7*(vYlIY{9Bkv?C{_hIEDhSm#F__U-NxW&r95N6#@h~q{5P~`3zsW`J
zm_CdIb;<p+#b*pxIAL^D!xJ4hv+6obI{L<2@nEdr-Ro^?2kvjt)$oKa;wObt*vS$u
ztyz41d-#YphevO9-YxhjqwfulI+D%hCbwTfiBISD_26w*u+~1Ls`eDf)v-U%?_u%B
z66p2DpE`h`^;$nH<>`{E3oP?P_o`xR?159{?yh2-+Px%G{{#XzP`C|Y+v!IH{MNv9
zK*k8CN<n&r;wdQktx<ckYnOhq-FOq>ujG~nyycU5(@!;QaVA_}87-_^?&3J|x8ujV
zCT{oVwEbGHP<&e_#@XkJ$9Nw(h#Ee3#W`iHpAAPzqtJ$FoTZu4)t^*2Ml&W=W}7FZ
zecxLB;$8bjso*uEbl^yRpKvH8<b`{drk+Ni#EnsZst|)C4>_7P+Nx)}9um!N%cHW<
z%^z$SfP2DRpjHzUX9uoHhG)=nMlGs*17K~;RbwdZ3LOj)g+9{i5@G4D#g^t1f))@e
zDx9gm?~f2I@X#;Zul~+cdkC^UUBSFJv(0RS8);1wVGKPD7dG~L_p@N-{)12n8#eTB
zoF5+Y>(8?&sIZ)~dTY+D=R8c<v~zjPNcRi1?ApN|H<)MefDKyPSNXxed69YSCwVpc
zMT~hIqCoBSvoOtAT>*whv^_CJ*Vv@1eJXXse-t<bAzLFmz!TZIS#3H+)yJFVJUMpc
zPqFL~N3zX{34hU2|1Y%$F<!9$_c^nNL-sUBfacS7_W66S3!zlUzj-_U?NIQZ4Ira*
zPjBXH%-t2zQqd%~1;7gg7q}a`aE{V;%;3>^F&p-4{Y|*|_kN%ppkTO#ha=404N5RT
z2BwQd@J+xAsQoBuZ54SnPnfM^uXfR=_5*;r9$KGg|3TsX`^*2s#kk@>!Ys1q&M+B`
zkzl*VfxLEZUQJ7W5}KM^2H1W(G#@(Yv8J|uFe&e=nVH!{7i3K0XTShm^_BE5b+dna
z;0ZGPF*IkXZSAdRN1oBr(8K`n%wAn|XM`lpB;%c;8C($U3joj*>4X1^X!wuj^AA{1
zM>L~AtM$aTqf>mMu%IBeX^~$+0}_Igy+p0{Go5#mAqk1ZdoRdJ#{adU{GZB6Mgkh;
z=M6ZZv(!+ou1M=?uf_{28^>_bOfe(_YGAq`?T`PcuI-aSe^<GKcx_tbE-Whf(fFQ5
z`N~Y?oPrW)>6q2U`5*t8RiGdG9~tV8==|BUkLCV&3+}ANxX9ahD94oLb4f7=o}#ug
zU9IQCya=GiMq|?GpLDZ%Wyv@EO<(_4pQ@NetH2UazD8Cko0Q88hz&@;xIjEmRHQq>
z$8Rc=3#+NCGkbi_k_**<o%tN<7Lfcn@u;xKStkE5;PmfY{p*FG3c~x^Ub%0ipMk03
zMGe?tOAZt#(Y?y}FKrqRe9x=EM>~bey-NSXMEiGy|8)^10XYji3v^yX-$ynn*sRDD
z)YXZ0jCYu=XM@cU1nRXNg9<lW)pZ9f<z#c-#NyiI{m0HI0~i2BV|w;>SP{vuSY~iW
z(`FVtpL@CcmGo&Io$8G`Iv^L6P$uxt#3A6VO)pc+(3K)gfk_b>pmDzy=hm=RrvJt)
zPWmDBnPY9S0#$0b&5;eEd2w}>h(?8jOu9SffBB~hF*3k0K5?~o=M<T5wkZ@o_Pbj)
zhk87YS0doKuZ6{IS0X5)=WR%$S3nO9f8I%|j)>LwN2l4Fd%Cm*tdhdXr4n^bCF;xH
zP52Zl6(t>&1DD+f3uUFBPLJ=ycD`$ZA_O{d{ywdIe*n!yC4K0m?<&>53%yx1$bP4w
z5=6DWotd4jsIjz|n_DXrU#dYuqsk#ArI}3F#u1Rpjz{)?7<=omD8p`RT)L!0LTOMy
zX%vtgP&%c%1*C=?ni)EzL|Pi@?ii$o2I=mW?(W~?8|QrA`CXp(`){7%;+lD$`<{L8
zz1G@mGwwA<k6q*|Pr^)n^fC{Lf&-#xQG3TmHY6Dj7|DwSJ{I>Bj2$yx>i1<Is^qMU
zdS7c#TaR&Q*>*poTpUMceo<BY?BckWn+tZNoXEE!du82fY_TsIS_44kS7+0b(l+w3
zV|kcQp90}R$>?u1`yc%q;LFB+<^>6$wjHR*uGD)iYrN4BUt2Bp!OfUmE3uYN*!;bQ
zt$HA4s&P2`uTH1Q@g&x+X4X$gn{l!V!&sBQ*mNCblYVK{X%01GMT5R|^r>Gp=sR_h
zy&->0`C^0OjRHW)1Y3au(%!uYlLGuj+KYMh6$b=8FIsL`NfG!(;Wg8)UVNV?S$`#T
z+E<i)-jZ<x#9P;kHxf(oal$$q%lmunw)-y{!1YSHQBR>OUpajLdQ`xH`R)I}m>nlB
zE%vRW3%(sE`~hgDPg*_*Kzf2>K1fPR%FXCvfoi_Jf9kEU05i`oVBsaf0;q;hNzW?-
zTKph3y?wIbFXc?#>M!|9hpTN<T_IoYFZ4?_3qPgp@y3qgEpP{-;4$5{(;A6AM;#D$
zPu~4`l~+eGO#+>uoYGJLed2vq*$Q_rbryPRp9Vb|e{$u1^?K`v??GT2hIFwu-BnxG
z1t>a+%CkrF@@jB>yUsw5x_%!)v{IDGbi4NDUEj<93cdgL3r!th)bcZ7xRNC~oCS*Q
zBc9$dXw;Q9rltPe>Pt*sd1#Y9g`@lQ?O=A5y{?>Yrr)QM1;jB)<|9;6)%X<WGV7oB
zU+c_3N876P8&|bi1(~|r<KknyJQNI)yXE1Jw>3!P&l4we`+zaRRHO#hgqy1`bE(KC
zt+7G&Qe;u+K0mr_7=sSVA-WO3TR$X=Ojst+&-dV<S+zB&{RqbKT6;ccoLgPcAfLPK
z^yl`a1RhAi^jbfD<BQTpUS<a2cnsoo)8np_NFq!TaQflu$kN~E89Ai^UQV>+5#Jo{
z?CTaAlXTo2B^y+;GwSQy-b_E7-fc0=;_}`xd~aO)YNEL7FZ=R>bbq*VKp~gSVj=%c
zl58AHhyJOYBCya1_IQ?fVBYgHt;;h9^4Q6}<=4L{gut_PJV2+nQ27G?-TKC&(ECY@
zFA!=pz_#y#=y*6eRWEPBMu+G8)$J!gy1qlAcb26#>wwJcNK&@OpSq6m=k4hS)RIx_
z7NM&l*@IqsB8H@m)bQF_=$Y5r_5lS0%5zJbm}%j=>$mkMiM+-u4sM42X(y_e!{>EA
zh)7w#o4*FMAur_Aj=|6QTcc;IN9ua3xro|M#S1H$Qvb+Y1)gX7u-k>to!^A;-f8MJ
z#!5lds!#K`zF&sr45gf(J8LyLWIN?4ujAJ{pZzW!q$2(DsOkVOsW%^^qk$I(@DVuw
z7rX7VC*gF{>;n1*6Xd6L*9{X0V7fK5ZrjG5&|)pN8JVEkDJUc3E48r^Rt`v`cYJXx
z^6qLm<fg~$M_`9x3ZF~H?eCi0CvY19=pyCX(tJ%=&xK24kimpUeJ+p}rPB4g|4k?{
z;fyS5b<1L96V*f<>n2XLxs8j6v{7qr@_O05Hw$C}+4=Q7458D%0iTobONSop>PPPU
zwj?53z<llKskD9r7LA4LvORSKD}zU$jBGzF7Cu4XUmsWFY|#drf67yq;8h=I@GQSA
zZngm8<Rlg(V|e2ZjJkiqqDD-NOv0^3><iJP*-dXOn-C%6$xEYX7(B1o>dsKs)1o~g
zWsQ9hJp1l>IDOl49wh!=@V$_s-=ofA1tFs=Ye6x_q9fJ7;+p#GV);_7Ad5m+KTA^X
zP%SYr8U6Z9+A-z@V1*q{8mO`*`Kq0-QJ5G`Mbb%DLR*rJi?Ig}NIvYR+%s}Y!VOKE
zuc2TCbYg>_6P-7H2i-Fg<wC9TQECZ7*@V|c9Pv!RGmIx|fgT1NRvu_(c66LMzl3s3
zupe_$apW9Q>;B4qUh(R7%A@%`tB6tT=YSD)%F-yE*W+fyV*GP1fMvY40KY(%e7zaJ
z1Wx7$Yt+BwiX!@d_(nyq$>*T{6dPqV<<v>%mjElfv+eYROFu<BV?vm<Ko>2*6R>2F
zzk1?5LJT#eeI$zn)2y=2z6BI^x5L9Gbs00@UsBV$j7gHfct;4g8o!o46=XO_dZq(@
zdE4dK9HWu*D+b@3eW+)k{e3tq1yR<IK-{>-nprUkd;>}Ji&EbAK&#bzkB)6K;)%GP
zDj@CXi#FeOXJodSNU(X0z(*~#r7Uv{(<x4PPlrMM`6kV)?L>YvwM*>cJSifkoZqN<
z;VaTex-Jcc;fiNZ@Az>?ylhxUM?nq=-$g=Wc8OwalHPB2eTMjjOYceaG!K@f!yVJY
zwE^=*4limzSwK<MM`A{q3y{Wiuiq7Gk6O<i3!OH*nC>G=ZE6x&8O|pur)WIDmF3nI
z_i!#%aVAXF;4V`OON%JC(72wPv;iVb53l?Z9KRBh5g3qlDC<zd3}&#)bAIUnDoH#m
zh;DMGCoO}2fHd{52qJDnr8n>00t15d-6e%JYRH1Rb{kWy%ZG#WlHOF2|5`{os~6-k
z`r=vbB(BNHc|`EsdB!>>Y`#E=N7E%;n?n8jPMvD$YL(W>+EuL?S(Bk_r%^-KT8F4h
z0QFcTH(a8$RzYr>`Rv*_4Dj=2U|4`XBDLS3tc_&RhAT689VY(dyHl}nMrDS%Q(R1P
z@;dKGOk1}t^FwcTAjMo0pWg-jB*U<~x;$QsMBShHHs3Z+3~hOYJTZ`D`+{r-WgL=1
zDL{~}UIga)Bc$a!zOK@eFcjChL%J~jsuq{Dzo;k|vrd?k6#1OZ!xq#j1bngwfwhgd
ztH*CbkEymg)>JoFvrSg|S?ixoyS1Zczn}SdAS3S_tI2a+{ir`4dRT}uWhOG1twTH$
z{fnFhkck8Ax}H}vQ)ng$G_F)Pqd5i~?%X;5>gjS<O@1y~x>EqPCR>YdEA?*{O(66?
zu|BF)TXjFCOu2NPq$$7R<b1r`2B?x;jYe)9CM#_VATq|xVZyJOTGZwNO)oYJM&{Q)
zii_?&#YS-ZOJ1s*d`wWn=X)Tl!eQtq3eXszSs|&0+o3S@4+5vZbww*2TvZM@Os}DH
zw6Ndi=96)jWhn5F@ed^j=GiS*Uv%3JVA-NvyH3mF7d)Dv+1P0pW{B`Z?tETUrC!**
zhF%vZpwQ7MR6F0U38~xgJx|h^S)1m~P2x7aoy0*5romi&PYXl!+B#S_0PR|wB!u-j
zD#!apODCW5vcpSemGuSEg)e+wJ^2{lcSwo0C~j{;G)%X<()tu14h)}7)%FEtN4M{?
z>6L5$Q8RJrbU9`|$}!K+?vLZIF_F~NVABnH5Wa&i<B*2j59uvqyzh!#6n^{q=hQLi
zYK6HTkQmRj6<Ko%(~42RZ`38K<kgvqdKVJzL$*Nvon=FmfV5-@Dxqpk!Py@C(g-qj
zS?ct2F^9xa#cZHjX?V|!yx=T4B0~(A^|Z)bXHG!F1-toL9`dbkofq`Nk-~`B<abV*
zC-Yl;v3XhWPlNLr=P=<>XHJqR(U9E2scRj`YS;fLVE&KGEduYuN_~?bs$=7_rm2%>
z_OmJF9g3X5x!)%A@=-SPZL-JiLd7aCPgQ>yIL*T`5c@TtmD<B=Pc`s*VB(x$;8Ctk
zFl@dEkmiqq>(<PFLg6#q=wD?hJ8;`)bu(U3Wz^VZgHAt0+fo_=MZ%kdLguO?TUdG=
zIrBO}^~d8<X;c^0$=e<KE2jlIUL6?(zJJ&-J$m27b*N`dxCuFpu~$L#KGB#lmO6eJ
zM>*2|cc}YyYekg8#Ox!hEJlVV#%ME7hMmWG@qqN`m08ryOG_k30QMU@l#9N~2(Agx
zh2TX_xwfx(OJEy%Pu*Up)Q4j2<4a^QMosf@&D`b#x!q+on8%Ai6x}$zbh-doe?ov|
zkB?Q@9>1kh^#R`EN9Ob<&H#Dk!jxuV4&hEc>cic#t-#LLXf^?PYxpIHC>KVv+s~CJ
zd~wvD+!Fu?eHA$xm1O{Dok1fgv^?|qSm}VS$&La6o&Lubh#!sF=+!9AVtv>Ck5%Wp
z1eG!hU+(D*9HBls&8U{`wXMmVo@Ebf%4iY{AaEdbyYCKC@Yks$8=q(x<PZ`lDhk<=
zzL}Ui!rO`8SadfKPHL8Vem$1=v94U-oY*M&i*Vva?5%XbTg^h$0J#9_IZ4vgOL@v0
z?qOqlki3#TKFOzk)?x|6js(Xpg8jP(K|W3}C4Efge8O3{8eLIq@+`l06X2~x)soOw
zY1%jZ`o2CTV(kzIM8=zHIzm&OHT%Lc_A_$}FsFP^F$r34Q?s#0va*eT?;2k|S84y(
zCZVq)TtMeVK7<8&>lXKq(&#es_-?Dx{--(?h(xml2s!>~_uCR%7aLSWknB&hb<ar{
zU|t`a)H4v&G&HwsIvAbuVfr4Ej>r3wj&cri-}RkF?)iQA=@@R5A?By6<on~&b2SOB
z9(F5GfnJ~O&R@88dgq}qVJf{`RJP6%rQuJ8i`u-2g2%_7+zfALIqmZ`Z>JuYy@lTg
zF-mB?-4Z)#98~;9I<EYxyK!9KQV1p99PDjX+&uSS{1!V-i8Q6k;FL@tL9<z&v0%9g
zkK_gOxJ?SV;T9@09jqlwkfD+^j~MA4zS2&bjcT49<9^fzcJ)Zb1nWDO<9mT;eR#Ek
zVWOC(1UuH==`M-dgFu_Y%X^VaSF!mnp*--oFY*L4ztgU|eXcH$Qe%~`ZQ9S)E17|$
zF3Urf9QH^T1rSscwF#<0<Je+8T?Q$Zb4x}e_m~jEjoO;`AaLD=8wAE9@O^Q9+B7sg
ztU5An-D+EA*C4XFU?-&d`2g8q88&uh-8QgUndzGILY#dc+T1BCh-p%jbJQUvXB0*4
z&|ItNR(Lmhu)7^)ZNPyr^h4JbP<VWrTbizNP6|0h;f>PB%Z@g1^^`_q*L1G8ax?n%
zLw=vT^X#hMMzNPsc$w73_wdq@5~OL+M-ALX>W7=*xG2I<LGs9AgG>8CG@%B&g@syb
z*nPH0OXs!z56O1;v7akspL!n?qO&p4=Ji*IpbS-+5NbV>I#&{lh(t?fMq(mf#)!@#
zQ20|SR---xCw;21TSeVL*7QI{!z5`I-04m*m^X1a$=o@3)7)F@hM2&+z5-Fd*}0!`
zauyH~XR1RzdEmOIj#S{V8klmLO3m=anTxAmjxb;NYObI5TO#|@k>6v2f`~as9s1`>
zWF&mB`UCMwiO0a9Ble9`nA8^NRe0}YL%xQzK@N3N4r695N;>ufEm*Ysr#rM>@S0Oo
znhSYX0qO8K)_(4%5NsZ%n-vwjubli*)+iIEh_x!nSzw5hCdn16k#9m~m>wjB4wD@0
zE#)%ZkB>cPd4(eOg(lwI2%(rj5HEG<^Ui)XQUUL&j@kCLQo><Zdp)jE2~Jn@UB;Jp
zN9u2uT=`gPJ)@5_9no0scDDVs*3<9zF+rP*`-RJXtE}+*{0n-quhyuT$z(*(<)>D|
zE#WDk7VNz*qxew{#A4RoF9)diT_p}8*{xNbfVFm<+Sf^fe9sKn#OpkK2`?ekrG+D{
z%%L&vT21`2U(JPaDFzhg{&H6X3!V=SS62fW)(2P>Di%lzRg@G_0($WTYPcTkjD}X1
z*`S|>OP0Iz(BK#uUcTLSZPa{yK_sg6O!6<ei3oBAs;_?<Hzm;X`urgKRn^zWif<t`
z(hzAoAwB}A%%`oNe9ALqww#0$_p>GZ^t-B&?dCd<u!*3CuaM(4?3B=V@ai77y(q>l
zvul07TE>!)i~J$eTUtTWRz%Y~@32Edw*UEyJvd%4u1T$B%mF%!Cv^30wO^Z-CuJU8
z{@&R4>elKYqFhe&+IIgEmrq<zZTfJ0gt+r{ceP<_Jx2!6y2c0?80JjvRwQ9lE%JG~
z(C1X;aX9*Y5f<GE6nnEdu*fFBI?z9z4fAR#vm4dBOKJN>cLt`=HI>KHj7PO7W5AoB
zK%ka`Zug?%ft`>zya~Vpg_B0)KH`0JriT}j3}l7KNsmf?<4u`5dVH=SNaa|jtV-|%
zG~TyczN-+SR81G_$Y!pNx}(rnQuv6=KrE2m24wwNP^Z?bmrtQinbbDE9Gm!dA1<z|
z6pb@EOP0S3Q5$+^ZpHuG3&8&91Sa@9lB9^)F<je!smTuIonI}eMP`3e8#;8h6>*4S
zwC=J&L2qZUfzJ!N{3fT`mjjT1H%~|m?aVjvHA`_wH0&HV&^1%^TrN&JM|#5*H7<QT
zx`2#83c^d)&`Gk7KgTGFbg^^ciTrZz^6tI3X>)_*8vc_ngwzbTsPnv1X~3?2<S}i5
zL8~R_Z91b?EBvIAXDCwPyolCFI=Fg_2AfrxH}3IYtNgEcJ~YdWOt6H8+Oo@Zu8|NN
zVDuCJ^}CcVHcoJGZ(r{!Lx^~8eSE2Aesd}~d?LWWM?+0bIv^)a;N&jz3Dlag0x-G*
zvYgbiJ)G^4-QoD$yzYu#Oe~|Vxl1U->lZ&aKP=z3>&ZpTRe%uG;sD<+CE(2ccA3$#
zv7&zP8q;^616B_L+mIAVJ6&fqJma-E$h#>qhZH|A>R7=yLZD92`6l?3bx5>@k7F~W
zb4rJvH{R|`-e(#|7{xmEz07?sOMB7o23H7Le81eaIAgFA3~>vpjxK9b_=pK?XC(VA
zd3@{%tO_JBFjo^6Y)E*O4=2Ia)t~-Y5XS-%IN|Uqe9KG2nt+a`nWCEyE|K8dA)%#2
zZR@K7!!TE8i4gd6uvQ1lj_s*xGs9NLlb5;eAr1t>uqQ(YG=xZZk1H#D2pnR+xRv=I
z1p(2*t`Oa)ui2So+na;H%xA;)L&fR=b#;78bqe{C!(kWiobBAV`XlFuUu_p(=WD@L
zL1qogtSFojx3{CjP#UJzksfL)Wu)s20#ZMMWLP0xB51DEcmGF3y4B!@%B4>*Uf)Yz
z!(YqOd>4JCXmK*G`y77<>fxx$#LyAe+J-F@06b9?L_>dUpYfyegZ8*~Mt;54%uJbt
zoCcSi!^J^pwY1EqiPKl&uWr-ZG#>V%S|G3O*v1JMZVivALTl(-;7BA-cS;3lI;)@0
zxi|8ys?_gH^P&<pfIr%P#<RbWI}~nDbc~djzid5gP}0l1np8*xz?l;96R~79(M0WM
zts_j<#Wb^-97Zb&5S)uDctG3QU1T%F+r9&Rhfa<sw;~Rk;lb!=Ra-7&fLh-rvSt1|
zxgpvT+@y`KNauJZ%k>Cn1L04tsdlH@5ubh*_;rxlu=<1)67}15)uU4QAHW*yc=HJV
z^!r}=<N$7&@AWY5MTeGtz|kHSP)bZ`H8;-N<V**-o^_o>sA!m#$GssyXCK>H#qZh!
zS`I2^<L*DNy5q3b`6736ch-WOG%oAU5k$sw$TP>4FK<O?g2+W5wD}F7jk?eR@SY(q
z$y3CH*>yIY@n)!t38+@s$Z3CFwiV+yvy{qhRe&7_+h(>0NmH5BV#{-~2tdoe1JuW_
zZsn4%%S(bzL9l37xxWVQd4`Xwh{0vX<T<PWB%-;!y}jQcOrc`o@~&)OtbSo`?&oTB
z2A;Q~!&snc0#BAmj^&LIjoe<L!DiFwMAkVOBO#pT?EYKvzOngfoX(LpvqvVu^x&2k
z0!0QKJ$M)PcEsw%(dERJjb+1StrfHCKjzoWK;0`CS0&Ax?eC(H_p?eGck1i7l*2KJ
z7={9$r42;ODhP4BCTNkT)D=f#C7@QNo-4=I;57o*UE4-l-j~iXy&1-HF$TBr3*~{B
z{DCtp(lb9$^lp1yaf6yP{#Gp|xano9*l9k-9^K9UVR(XPSqV~-mO=YDYnZJfCaCPI
zUm+XJo9UvkGV+^|pK&6rI8K=~Z&n26-N&J6SeNmEm4mH$eL3OiAQiOAp=V;5bjfW9
zr@s~9Xq{oduXg#!UP4-;<APE&bMJJ2#JSU#{%d+0{CK|`A=Y6(Ym>R-(-Hc{juikD
zG<WKp0@8rP!hBBR`sQ2`azgSOOgN#`EOe3Z?gEBG!VWaW-rS0>kB{pNIr5hHg#EXo
z{{P0oX?z}IW&!Oi+L4TYmeFQdRhQGN-hL+I0a+dG44VA5V1SgJe(K2*;N(xemOpMV
zX!4{wg9M?a;T<p(%l~XHXNBCC8EDqx?$JzPSyUpGGk|rsw45oxWxmV<AZZ!6H*HJ?
zB~B3RDq9>`wxEzhE0<8WnmF0tE;Qk1k1~PLs?DhBFQoNvP&h>TJRP3nz#6zTu*F}!
z;w@r^qHEH6jH>=tlopV-YoOxj?4X0s?QjtYti?GcyctMn-Dq>eokik#4=Lx7k*CM|
z%fwugkMWz|W$wS096ACX0@xEp4k#rc7Q-bfHm2mN=8}}Eli&8az>{wgr<{O4+a@ad
zC{2e`UMKM)I4{2-#%eqz8RnI#1-~C2M4kXui0>6#Tx&IPL+tS4d;o8Reug)*z~fhe
z6qk5G`q^NYXRy>Tt<qkhNdB^j(XUA5JIJgC<DcdpuF$qUwQmLA>8D|_H}CMI{T-?P
zN{pP34bsYI34ae@zVm9ATi+QJe^g107!yDpXfHLD0m!x+qcE*;Z2cRk<vnJzreqlc
ze<F<UCc&P|iAHw^f~JYXVo)z3JXsAb`^d-Q4KP7>C8$x2G|Lz`4G-sYKhUBV!LBwl
zEC9b5KbKc^-sT^ow%s{!pgus7{Mm*u30<eC!?(x$bTlxwFll^zY!v?W=1H~wYk;Mh
zk$0Zwau$khNb^uhO}<R+e^LwhMhUC#<Tj|(LZ9b@ccIYV%l7Lgtho<@oSbY65)=&=
z=MJ0uI@4{N=Yto5?6y1ZGGNSA9XSNP?wwHzBQMa?4-}<!A~@Ja>aXVsp&u28_%(uI
za+5#cij&#fU~kj;YEmld)dfORMD_~%{gvzIBLN#8nSo)C1!`ZhaNzmB!JE|+p9(km
z5@@ts`f%d!UqADKviWj>&|Y27?!jf?;2>D%4PzQ&5M>nrdh}B%MnP16KWKO`0X}GW
ziV4tTjFD<D`-?2+@&dXD<-_uLZlYxj5~{M(Lk!J3-JG4?sYiwBY556DE|LKq%Lne~
z@|6sdNx;+`Df93&j!<fr&yi>HT)4ow-`nQ9O;Bk_yMiuvZL;@7^Wg;!eGuNaP=y(G
ziCdq(I<(hhZ^zWh?YrZUnEm4!2V-DD=Q4S+tXchWhbS86&3qf5b-ey?3&Jt5&*TV1
z2tOX)sn%@^>On2i{ubhrB>K=1u*>RoHW{hB<O5Udk$$$mg^K`ifp1Z_T}s*hgaO>>
z%B+-ojtjBU#qJKxWV<c7zxlzVHAhl%xby@#fA}tb?Tnga0tdH(`cD2GN>T2U`VWsQ
zFST+ICqON4F6Ut<vdoBwMgKPr!X4=kTb+NXB4t!e05#M*zIhN%IiBI1Zxl(y1eris
zQxB*JLFG91YqOKL0`MBeDY^sE=37O{ODvsL^J@T`cc}#hyiDW5HbS%?t67bCO^Cf&
zJ}_8BBlhB<W?M|fgmnGpK_>w{xEzA*SA67AYK=}vSh>}OOAFDWZ4+N4vB3X4F3!Ed
zZ8PB5O*;>{J@Yxb-`qTUp!`|5+jy5VY_e+@o|aa~%<*t7ZsgfkM9~>W@fu+9KFUqM
zbG?ntcR0YY?dYIKM$DzD7Lldj*?*|fZA2?Q8hxcPcU|^N@3LCQ)al_VMDe`;(Zwg=
z@fy0J;=F*hxJ$$r=rpM65?r6ayVfmVITn`yr+@yk;5AllMBO3)OQLiJfab0b`y_TD
ztH3ixq`oQxduSBT`M!@Wp$|O!YFAo^WEHLFxq1hVnZ_rALOt8ODmYpSc1}MyN88Va
zhcItAR3j%5aWNn|ou?WJG~WGnA*5*Jk;T`yi)%G+Ppg(py2KcO0I~fCx3?i2W-U@x
z&{)O@ou~iP{8w>w{ULK>Zy>UyS%ZUbFHo;`N@{@LlAX7A7o*;tSq&&^b)AgE0v<mx
zp(7=cG$I<Y54fTGhDCJev~o$hC`1RQqv$X!rYf3_W-{DM^>dk5#+&uBs*D%}$UB_5
zMzw8J4;9(Mect??hU`vLGTr6uClxPmRn=pX&H}bA3ycfA01uRHiJczcw4&)o&Z*Cq
zhhA=YpzaXW(^^;D^n|Fa1BM>~$z}F&ipv{QUSrksQ?8TFAuI|@=a2UXZ!;PY_$&@f
zh+oalxSb%~@>Z_ghUGt8;g3N4h*#((&+OBTexV1@Lxc+Hy>_%{DJRen*CJpkjmv51
zXq3Z({Y{;C!X2zYZAR8Ih2&`&+jJT`f6tux{*{YxGVA30-A{5;AV4NFgWWXE*flBQ
zg*T~0=m73EkL+(yIv@lgfCffVv7X`TSxwi<1gT2xetRu6->z?Q6wTh+rT<gTKdsE*
z)TI2>@xd;L6~K=d?78Y-UI3QVzx{kq5K>x|!{<Cgw$pCdNCv|M9J|ojMr$l}*%|m4
zehxm2&LiWbh_#l0))a{tZs#smUbtB!o<@6K6!XR%Z1U3!N7{DIY-qgx>MvhIe+T3!
zuW!VPOqBJ=yB8%M>;a;c<0?cu(wfIoQ;yhx-QXpuzKi16mp-8o?l`FPZ=;H)lGQbI
zO+LesSpa3zg>+_gnx=Z}uLHDZGl;)6oa-iNw<epS_5jcR=;|rdS;^=41_^l`mYU4O
z{__l~t3!J(D(I60f^C)@`W++olQeRnEVY6_k{F8>DH*9L8CRsq8X8|5mLnjoS4g?!
zSd1=9r`c?1jT&%3-59@H;kK)-Mu5cLbc93?hPwg$7-CLP_-w|b(<@7-X5IkMYFw>r
zNzMT0iM}4jmc1T+54FOa;ymeoMF2<}>Txs5br;QUykc+gpxa?sWcnK)m_;o1*ZlOW
zBC;sZ<u_}xv#*ATWV{q4+Q92`uF0(dVwO>JFzm)}`gp?w4i$Zvb&QP4K)LC3dMS6=
zT6%e81%&}M8OO75X$zRQ_aj<bmKI{NV<Eny#=4LlRbbM>8lcL&N9NVJe`Mutb$?dZ
zveH5Q^<I*Znt(XWcTWH6B+G%~^{wYy$v^q2>X+c0X^$JuPF%gW?_FkQX4i9hK}8d)
zvh$xS@ktF~sv3r6nhtFrEU!?F`lu(y%k9Fo^IG8DxSHhK1<gWFR=+<c+Kx|y=DjVc
zjYt~X1?sC-*Cq*ZR7?^ureg?io<Yj(cF~V4cmYBle9oyZ{JoNcO*jR!2p7^(iyvNj
z-8!SfyU7m%=mq)wcfk&%sfY4zzr(jUjk_7hi<sk+9nxcBI|}aW2JM?~Q0Ch1WUiL9
z8<vZorS%x>!#EX0<N)WvI}D#E7X{?P>auioq>r&hSD&vCPrSl*7$mI;oNi+;R9n2%
zEEFyZM=NB0a1{}Ff(1${Aa9dJnIjWUUk7EL{V(NT4Hjy>`_52+ii|j;$qAWzz@Z8g
z;K&<cq)x2(lV5G$O?cssMAzr;cd`j8(E!*xsn%-TX5<qwh4tld&rM(456LND$2_@P
z2Thch8(l8lcRIAQV=mq54K}_~%$HilKl3Vp%tx54kfci1Cp^-hA6pqTW+;tR=T*fh
zeAxr2H)sd9U(}K>Mc{zzZ<n_wb9++0F9c@zT^A5{Eh`goB!ceu@Yg_as;`S`vw%Ad
zQIVoC${bvNlslwJ8;+20(>XgtC)MsdyZxhp3N4Lm$D}I{5g&^CLz(!Tg_F}TiD*hO
zj;X5kmJSNP274uQ@H*cd{7g*#`7pq#*76Mp+@f1Zn3VprNbcTlp&UrE<kNSY!sB`(
zq<py3maa?Z7~iEAZ7d|y)`HHyv+|gQp7QE>eo?(@f2moK{E*vVJhJlW{ZeVwU_y&a
z+xOojR|~dammj}`{exr`g`{YL>yX%PN^WP{ka<6Y-&xMpIJ8sE)+r6&k*i_4nOx@%
zugbabP;=`7Qv;LiS5S(q@9(=&>F+PQeiY;Qb^sh7*JODlT<qpxw6trEbdw!WbZj{S
zsvNa$B_m*u{C;hRcs@uOS=ZEjbD#n=L0^Fh&8RPi&b9;lVT~M6D5h*>9R=MV4pY|-
z(Zm^NLf`;BkCZeW@STFh(P%EaU+F0*++8|ewNVbD#;_r$g6=2oIoIizeLMQXaD;Xc
z1g5Z!p?OnYy%QFumidKW^{oi1gfuUw+sfb1ye5{W+y4)S>wjQK(|^nu#7usSjdu+M
zm^Gw2^)Xv6=MJG#<2+zkT2y}U&JJh3ZNVjD2*x-ojvjRycD~h~r)DUy%1`nt$u3J5
zBB{9u`oqc3I;}MQX#5Zm3Pqm$=yx-|^8%(dM%Ve<O@*=Z;kxCr4$CY#$lEcwq~7<u
ze%WstC+y=J14hg7b`iq>?~v;|ml>8B=jIM%hJrzU%fXgvO~WW(Y29EXo)+S#y6>-0
zP&qKg0cZP~s?ASTxc`>qopY?SULT%XjfN3N!}jP--5&`080+R6jwgavi@twT3oEy)
z`C_o&erLINL4AJ{)>427B-}ZT4^<TUcs(&O#bnZN4^Xz>?<MKI1dJw(?FKPKF%Xox
z5#h7WVfwYodN|S#Cwd0K+L|gD6H@nOe{4qZ73T)2i+2<t1?wYzud`Ug`A7kOuA#c)
z2}xC{V1rT922S#B{D}FkqR}-H9-!qN8HYPqj+j4=vp$lu0%*E7IYKEP!6lpJ1)Y97
zZ_^>_6N(=YKHSazsh}>))9m_%6QwXs`d~0_@Hacx4x3s`{LY=a`CRl~6XZwOM~%;9
zvJ!$y%w%=jjwo{pdJZ!_$~yDa_e8ySvu*5sR-?ip+D9$OkBv(g_nu8GT-7=pg^K0T
z|AOc@a9y3*D57hNiH2S86MA4zpgE#U{D9x5l73QCsSN3?o(H_-R#QcpC*R`C?bfeU
z>fjYsLOsrhk!v09AJJ~C_q@qg=^Cx!$Ez7WnkcY2(`lq#Oc`u4+Sh|U%ef1(x~UDb
z0gZU^+eW!fpoZJP_yigLN1^>+mp)FIs9LvK(I5?rP@q#^tkrzZ*(B`i8&5mXv}2O|
z3PYJaQnPl&dlJuLL<l+y-C}7;Wck=(gC`>EtnUDt7l&-yF5lkfy9P+D1P5@5d)wKs
zb!PcLQR<;!g7Zt7574gSv?Cvn^3)>{{jwwCK3lCn-X7B$mNv_3Mcuw|LM9{Hg&tu<
z|B{6uQ(e(2)Df)^I2aP6_C41m7iDRg;knF)X|{ZNS42U>Z|F;n=)Q=Yk#(N4ZfjvF
z6LfX#%<Z%xnL!TJ#K(BA!LaP62SQh8BZ#B<+A6+?{sH6pg)7twOGo(nDz4_42|@II
zejPE2WF6c$;G2~V3Qj?76^I}cJzuI8;An}tXo^~9w$5|ahDPtoymvgZ;S7Ed=}d2~
z`>RW4Gx;psppMc=6iyVPyRw8<O~i~e!4H}j{;`cxqmAbaELX1k@yFnHE<i6TI&0YI
zcd9bRzxIO<C$qn!8njwgZn-r~W{!$=go?nQd9+i<Q0;v^4@--qY~ths&y(r)MXE|A
zWW*WNDb(rDV-e8ZD+XxI6&i^aK7ITAeI$7c`Of?t=-!5xAKvVzZJ3gQ$nGQYLX!+8
z-efMSn&WQ#IIh~Lz$<jA9EL`9Df+bY1mgl6yFd<jVR74K@G5q|Nf|my(o_r2Qht!U
zDdTWx1Hi0bUhtf3akO??INnq+86nXHqB!y=%K>!CG;#BH?G47ZkezP{`#P*->clw|
z5&<4POBoO{v!&;lb+tCSJ1_V$cTy!^$2;^JP&jBlr{BY7CSSb<<i&Y_`N%x9Bu=7l
z-nd@zub|vBLd=RkV$S%2Zxvy^JK!9W;s)g%wm)8C*o?gS%(;qA%2h)M*7)R1aPJev
zpNy7o7(z6PEVzg!2AlZw%)IW^K05>x5ts+=Wc`FyJWJ%Lzu|>QqLJEn?T9fKx%wi^
zx6gMi^)NbZxxJ;=K6HXDNYcC7Ei;EQ=m_TI?^eq-GE2Hgs@q$yMKvWck(!SeH5ki`
z?o>)k4sPRK5mXIa@9imt;uUqBz%ULNbZjf2G@)WPUG<AwE61Zw_HJ@b1}@835^^$e
zU8zCi1S|x~Sak}spUDAlR1ApE1n<@wFX}hIo+<9(6?yQp9%L~BO;A0_3ow<YOzX-n
zp&sKV2PD!Z2EKuVQzqtq5Iwja7^)EA2qKHaKwdbzXfjk_1PB3W5eB3dPT_>&FIHFY
z*Ad42wcf%Ia;8XWb>R6<M*^3eMI+f>DhzWu=dR0yXbpDYUn;+zBVUDINL~a(6RoYu
z;o7QhJJbGY9cy^^U1RK92MksZ$-rmaZ~j5ZMgwQSt}CTg7tS#LW_1-PSe93R_DzMz
zG?I_>w=@3aQrUbhSEfOQ2Q~7c9I?%TYUtbG^d98)aTUBdt@A0f)gdeN)+3T?_iHKE
z$EAEbQuKo8u{cN@$)LGc^s1Zi>tAht-~2mUX9!nFQT`%VdxYimXTGP=4@ub%A0GBe
zV#jaAX>Ua~>-R!5Z~(|d17$z=_4vF>0aA2M|7&AI<P+wUH3LBNS$Qt!D61nQ-a=<O
zcWyFPU{?2dTzrw4xfeBhW%NvAkA0JUxq0lEU_vEc#iNj0`J-l0b5Rf(0Ru}*FVY0B
zW*>^WV8Ur26KWwUCQ+G8!l{pW`D)y7-NM&*@k*j+vvrBv`;WnUbC0YrG%L~<khZC?
z?>gGwF$~j^#cVzpCYH^{{perg7`YawtNmGa7h>t<P^qoBG(%ZA=-ip2si_USb*fhO
zU5Zlh_5c+PqsrzfwPBDOp+)ZzRB@E!bMAW)k1z}-n-r0cyz<ba#JJf^xa<sULp4S_
zrHjGBA*)07SHnXd*~?|SMtlm~P}NPy-Jzk(%`}n^UJq{8s~q<tpNphz*<Jw_Nu4>7
zEgo`zIyz2=VbVoClp{fD+hE-kHO-S}`m9}C(?=xTPwMmJt<P7xDJmUlFZt?bGRH70
z@){~R!@w6&Wu3l#f?)!h@0aTUU0qCU%y$pBt=@5DyiM{M?7l2j<j%7M+`wLvA3-?#
zDfjM+qTQ`t=vLnc&$IzaC31CMBCG6phu+7QI)01h6ki3j&g!aaEM3gNfcvv*<O`3a
z2;IrKAOatm>NJ8dJ65x8P2cs$EDLkMV*z;h^77b-hw|6^Ht(E1R;6jiTU{5qaowlj
zhmnJ_y2uVo=|A;ZsKxtVY`a^yN(5ld@)W~xIJt>4deb7i{znq^-<L~_e++!KZ@#d>
zS3oJXtG*=AFQSJ%v#}>?VYIPxZ0+%{{65Id<-8eA5+#<d6!ApR?avm*hBn478H0O&
z4Tig-6L#R*We4>*Y0f<!<Bpws<l-2=c>9%neD$W@@Ovni2Su2AFjS;<JES~C`r+4Q
zU#pN(@FY8_O|N)jqORoczxuBD$o+x$N=QJ&XJ^WK-{3}443O+-#{5G+EDeGBRLj<X
znbPDl#Q))v8CcQM$S4F430vIx7(E@kKC`goE>*B5LNHj~La*-VO(v{buV2Q`xx!kp
zafTHf{$^hF&xWh?bl}6aO3RgA?~rG`m%+)gkq$_c)Bc`+_@R}NXvWGQwNF7n3c#eE
z?_H9Qf}+R|sHue%svL)<Z?;xWcf_MT(?gpcwn2&#e>bN3caod`r)}m;ca@Q9@?cOv
z0StfW@YQc)Y#I;iA@#9he<$2&I_s)8x(NtEzAHOIAN)*AyS)QH&-{s*HwK;p?!PGi
zD!%w9haeT;4JfHgnhf5$IhrS_@_`rCt7~O!w8T$66?fO>S1>{YO^_VXpGd{F{IQUI
zK1u#WUf846xPIF<Qc6;CCDyR4a}cfTV5gY;M}0ZT(0Z*P3qV+e#&!HZd@|%reFqbC
z@3{vblG2*N*}jyOK6rP4P;~D%wR}Qw^*dtmD-b4MqIbLjtdoFyiXtvdGG39DmC4G1
zEB}#yCg<%StEsu;SVNqg|By{D73DqI8HDiv25<iBeH!9F=uS8lE3TW}SzPzGV1(jp
zjYtdZFCto6`L|0`UZ^C3FBSR>`OtULOqnO8bUYeVHPodZ0=1tB*QEi~orDSxcWv(<
zB-(r-fhv3z@2`>hpAdD{2cVaP7;>hpvi*$mtxHW|VQq!emmYU*fF+RZMfUK0)Sl+$
zav8XrI{`wuo{-V9us~QYql2db(e6Ny<68dYLNt4PkyJt|<KKp+{QK4j|1|WCujKye
zBus&LJz3uhaeNAO_vPB@e&*vq>^(uvB{D(H|1t2?_A6XiO<79`JaJ<Iwul~=?}rBt
zIwaP9&i|!kl^$1fo?IP(23WEE6=i>fRV0M!J8`KsKm6KDhWQxV2cf8cRO#>^wj7O*
z@e4ps_w+r8>Gml?&5-;*td=+GmoLEaMi?Hhtg=XBy22`1QB*B{t?pFJmllbqblHOe
z`Oh7NAIDIAS)fh2<uVibr9c}r&`&}PV*Z=2-+%FIOcMXX0ll7IORz%KU8=U8gVx>l
z@cy5zomz;X&$F8{3`Dup5Q?IDcszlB?a}$yb4Vw{)YPzz1w7rvKZU2g|Lij0pZ0!}
zV)oxb;lD%l|L>i8I*!b8rS0uW<j^!4@064j26$H1i_NN#ouL=mfMwSS@peefIo>OF
zX4bLa#7+&$&)Dm?MI<GIV&abB>tAdYW61y2|NF<UW1PrKsyf2h15Cm{s9OI<4~z+P
z*gLAM&gi{%UEO*75GxbnlprL;vuN_t-TzId`OmwO;`lCTg}Ct1O6rep+mi&}U66h`
z(>sQAii<adxi-p<c=;)U&ZHJ|?kmR@V!6UB^3Sd_v@!)gYpp%#QwE2QMY1B#7_ZuS
zbIJ7mEDo>LNt)^py^z768hI{Lt)E<ZkW^9*R!DMO0KJQND{>)2w@9Byo`Jd;t4+|$
z!>JG;WwJtVU!xE`eEWlsF6I4`gT`F9!=vJ^n&NkC<CuNjs#(&%Ccw$=@=-Zp&o255
z*4bUNyENzG1{P(-37i8npMUr(xKz5E4SGKUv(6z0Jb|a;qK)NIc}H(9bQ4Eg-xjCl
z%?o*ltE~l#Sm^|*#<=ujByi>(P@Yhx<3HAQxuL2#Xr$nNnQeT=X(mjIy`>eW+@>O@
z+I?^*rq?}Q$*h#pxUO5S?+AhkGSAG++z!w4mlPGn{LW!e=|rcO%+XU<0pf7{noECT
z6}g_+GxZWpndPn<`pV!E?TGI)DzHzdszTG9F@y0hKueWg+#XFo)2$Ld%Ps3pRldqA
z(mm}?^6xA9<he`QcPgTIBqzZ&-SsJhqu)i~o4rgY-~36){-{&m_WsS_U{k|-_UwJC
z6#l{{V=9gFpkOG$s$9{mBSXIQUUtxw?~so<lQSJYorb?Cn!Xm5lA^+90Xu_D7G|kD
z6Wg~EX5{JwS8JPwpU#_6@<Gv<GgPzGyOR7ndtwZBG!9Nxvo78mk7leEa1um^nK0?l
zG46*~OB|&Rvd1Z-(a*%uNL9==8jO3z(Q{iDN?uXCTz;YadB{tJZ(x~V<7fDmPF8r1
zy!*iH_ZL#85qpF+QM);oRk>MK`h|sj+KLv^xtj;^VM76)5~>w%WTJAW-o;GkTgbkR
zd^I|@T2173Lj1TyYTY%Cc1@c@<~-sg&1II%!cxgv@H?4N0QEOcwkmUheIwL4*?_0t
z?03OBcNc*=`WalJ<V4l8dmoxED!B(L)f5zA2F~%IPMRjC(_%~gtQ0h++w{mQcgv{_
zzbm#L4SKjol;hV$C^hNKiQs(6%uW(mI7^Y&7QR(^CfA!~3r;O_UZqs}jEI~C@AA6T
z$f`*vg5Vld8``JZ;|s4V++$Dk53Cl*0+2ZwOBQ_jHM*&-MfXf~p;&8<KcDttctT5o
zDJfL{?#ZXYd0i;b*d>XCn7Niuwn?SxKYd`0iEKaBS!i$sew(9ve5Ljd#!KNK836$=
zUmt~c#>&)4h~W9Iazbu`(g$%?v0o_i53LX7rCRd)n;=WkC&Xl`pId)Dx~FzKe}2>&
z>hMM)08=o{M!PaRnDSio%WIPvUx{=)a=N?nVFGS`rTndiNy!H4{wKtIbn9YVKQuW)
z^4}mZr5_>Lh`C9b`}Tw~vFl&R@wbsQV7?3TW^rB=yL>q6^?&&>CgfhpLMwrCjI#Xk
zd+<~98DCeMP{vP*U*wkc?4zpOvE{o6INTavul9<DnslNci(gZmD{vE^ufV=Fu6|27
z==h9XV-;{-!ZC=IJ)U$xD=0u08M|30^)Jtp2<Hd^am5yLzo?6bCP{&_NxX@XT;9b(
zaoF<l9MVl2ZA6ryL9n2Dq>ESv4#=&2bf-XnVur55EH-9xLK?_-T2I+OUOsDhXuYKB
z`##<9>M@MZa_PgMVfPhC6xLp6zYiTCUsd%s@{U@Pk8tujy_t<3#J^P)UXrIyhEC6y
zxq76|rqUApbEzED<o6+uUo(&Dr|TZXgZQzw6|%OP9;mvj@YWn>7gHQ6Vkd1B83d!t
z?fVARlPa|oGHL;LwLP%bKYpDGJm3-5Qr;X4+Y|i>^RMMo-Xz8>yzo?d=Uey5PKEL5
zsger-cpQyVVVa~cmPAP{zZ|?bBN;?i6ggUvqm$o$dEm~F<S-=ch{zXtLo<Mh`z{hE
z!W;i_DM2OYA~}yjEmN!W#9DBcxRt!NN}vXn0xjO&TXj)aJIPUx2A)W-pefMtFwf|s
z)H5a$$dU#^pZ@2y9A%V4qhlV5boqGIaC+Cf4BuEjQ|Go>li7WnDALx31qnN4eI-t*
z9p2k2Lu)1B?^zn%p%+{jW1$X>cy!dqu9HvuyXvdiUwCxA6lgvt?1M$_H5%&}gFzl&
zgIH53b8y}9Up9muPPKVy$RXK)H5}hhhLg;sZ7v+9J_KCpN{{Uu!`ZB@jFA_-@_BXE
zOnR_EbsEKwV(R>>Ppo;;l%`n<b4w&>K@IL-MB!%%7Y6V|>~}&7E_-_c0Ff$CH*;F*
z=K#UflR_eqy?k1rT9{2T-^DZjRRKm)YVpm7dmJlNKZ68B$*duh*5s9)>HQ838{|Q;
zgayL#i1{rod^&<!6UR8X;ku0Bd4%8e@2LAi?W0z}`#ds`cSBocb+z8>*UGRc(q4$M
zfHLD(z%iroX^vd%BUkd!wC>YPVnq(~$exdbriqLs>2kJg1|5%1ew~gagh1brA`|9}
z#h9D1%A@zM@<`xw#D-YNW4>Zq45p8IoJX|PukK=8<eGWJ5I~iPkW2c=kjgsDL(fX}
zHEIcqBQ@P<Bpq`vOxgz%H%CFhuK<G~MQgC@1-H%CZudSlIA?|8UPW1(s@_ch;Uh87
zc$c`bFi=P_i3>|`KM*}^fo#F|3B?#nmr<cn1VTYp`50A~(mxeGxh(j^xs}`AnJ_Kh
z_0;s`??aQ;w}qzWq3()-ySl~cH80@24jnJZ8-dlajv3^__Va$`cL^RM*A36=mmS9z
zHiCWl4PT%4nplH-u~4_Q-*o<(Bzu)(^);M(61%+FRdQ%0m!MP*JX7_vSX^eZh^J(s
z3<_>SVB1In4wzNf$(SHfYR75M&9lIG3Nl*%tLON}WQ~?hl<`3gk)PgmqRMItHD!+Q
zpSpW*`J5sjzwC?dKl-M>^W<UWSKRtWGRs2h22Y$f<%W@<iAooBz`B*KZG1(ArVq_O
z%w_N^bRR@XCxR{)NF?N3W|OQ^ud+UMISS8y1So!ee*&e6b<$r>J%bZ4im`+J2sK#Z
zG~1v7)FtALP13qvruqvr1Dw|ZQ$_}da6d14k|k=~b2ls{bPJP-t`NB|WUW%8-&v5@
zy<#j_#3+_AR&aZ3s4>r{LffJkQ8N^72Q?Kfvs@&)!fU#$3PZ$HhNq;GT{d5lcK@J@
zx(_|YbiGe<|M}shl$hfuS0mKr(-o<Q0)%gxl~scjMUN6QuWhH+>p=^>4m{LS)`@?P
z3k>KjeO9Jrgu7uw%u_6pD2POS?yA}lfl~_ltahB<(_kt}cp4G<_zk6fzLyDut*L$X
zCY#Evj=;dKJ;Q8zP1dn+PgaTCiPL;ge)zQ9jx?_UL4#9Y*_Iz*2kK&d4cMfd*YB}a
zm4CW2(&@HwoK3U_jYCnx%RiYi@<(!Rdhxn~zO~pl1*D1X^{>}k3x-w>3;)!?*?IMB
zu@LZ@>Zq-8HU6?Zqx1bAzJz|<pSd8S=)7i^HvltY4Z`ul5zT2OeD5&50N4vkx#=~I
zZQ>tdzg!c$#M;(<jo)b9BvzrN{t-qNFM#|T|C6o=t}_>%OLa%QHE4;*iz6@vCp5A&
zfV2LEVQGerWTrX6KtOnPdc~YWeHs<gv0UWjOxz4CZDrEH6s|5^a$lx1n>M{f={Ea4
zBar?llrtZ{kf5QXozIt8Yd391%9h2Y=uws%2jPgC<z$n_`^jdeCy*a<SuPzES;yP&
zpO*#%S3;%*0bTJ))F9TItH@h2MZ00-avXKGcb$$s!Q>3~S*Dv+9!0F%Exsw@Pi4ky
zs~U437fQwNR&%eAzzbWk`OasIDPLfM2xfl049*JGXJk*hC}q_kDk&y1B9R;t7>F>W
zdCYHn;*jKlyznNskc#5NTedbi2_ICC!9cJ088ecT8|J({z`;f}QIkigUVdyp)mpLw
z-a=7Yjd-@zqfOzt0K><~m0=|*b-nS5W(fkU0S}Np%Qt05J$Y#WSdv?7y2Dv`^PaUv
z_*B&cFt{04dHf|iRX5z~Cq1_GrNtz5Gi?m<SMrKfGWrQLT^IjS)!BN>4(0s(gF=6t
zjc}ch2*m=L=xo`Im#IZ8se|}Fo=q=8*HMH)8aeMQasOo~>s5sjt48q@yE1l7D`7JG
zqxO2UB4C-B_4EC_+ST)^nH=y#uSxN>r?hJVGSXq^i!WE~Tu8>Kj~?VOpKOYA48srf
za=X>akC+&hB~+_KQGE}P;fSz!nq&6f3B9UT?enRV&X*~jI+G3zR3S<=^TwpGL9rQs
zuM^L%EVkaxO8txFZ;y$1f0LykEN(l3il!##*-)PohS=IN^DO7i?S!5`yL)rMms?t6
zP%}CVwc4QVqNOrUlr|?bHW;*cmcY_A(YIRuHqiY!*5D#w%BP~qe$Ti`xj~4^r>T@b
zwORT3bhSmZ7Ww-FMP#Myqsbh-E>}L(0c5O&6UbgekncXFhctGgbf)v{xuec4DhGN$
zFi+?$cE`BRk;>kk5XZ#UOn>!RQr?BZ@|sY)K;~Efv^7?<FZ)k4^AU<^)uNYLfOgYG
zixc#}732S`)1SWe0V-uqgrcK)gKF&B1x6uX`K8aH3gD)a&W=w0_x_CkDk~G8Y$ovO
z2S$QpW&%%j8;ez4Rv_D%v$&e^dFqo^tZ#%-=H6@_s9Z{?J#+kDuz`R6sc0RQ=C)-e
zH>Zgf9`S?~I&m!8y8pj}3cx=;W&DF;NJPfL1EKKW7Qh&>XP=KM2Qwt6AW|q6CgrAj
z6KsbRbX(i3PjMCh<99ekMivD|Rt*)4EN^rZDPn0rs>WsGSYsn1B8JXYF0sVhi2m#5
z|MsHM8?a-~FDi`R9y#{iiekR~S?S<oty5bPLjgtfC@tq^Wr^X&7|X{1n8vR&lMnw9
z(>N9Z3_0LGc)o2n;R3GF35duJ<%%vO3`zgAC$CspqqYky^Ycq;HpW-X6SZ%vBqQ$x
zsxFU+mO1~^77($eBLKdtMrhg!A$0d8Q&bb~nX>-xSL(8AmlnB)Cd>cXZ@Vn+&lCZ`
z2K$wUhAlVR_IAJildhF6iICI$e5Jy)=S1aBR}L1owmbY=vp+f_@iFoL11jZ>>Q0Pg
zb)+&nL_@5s&f(=a30N3lKZO>&F3c*anaBe4iHhU_)P12k{}DT?4)Z34UcWt!b$G>^
zu#n4g8xyY~<(0yCTclaeC3g8L>iF0}>+J^7D;o+8j+9UTc_E}|lKz-YSrS9ziotQc
zy<D$YUq!(yuvO$$ayR&j87Cerq9gJyV^lA<HlNz$i5t1R9U+yGl?@mF7=SHSnWHEz
zdsz?AH0wDSoqdSX_#2sZ+W)jhzz(NE(g1+783r%r*EpF`0EDcnt);HbfO~d^e1t)(
zx4cnPIKXiQbiD`6ov(>uiv*%S{1p%XHyO>O`llJQ$o6Z&H#z_`rmzk<waJS=t~mbu
z4Up&RJVK+9&D^NLt>%6`rxjXq=`O&}|0&>m0mQje$D<?H>@n6?5i+0e1=#ror*cAV
zBX7BDWmIIXrY!ypWcaUN*)7OpgHm6-TKq_UMqY8VF<>a_wXbT_Ledw^FPXmS!G|pA
zbeT`H1^pR=>9BndRow|cZlz~R5hCSBk+|E^5-iMGA#?|-`sC?-cJI76-k1+0Ctmf-
z-M`x@X$<ftAV7XEf{BSlOM~j~P0;-ah#qUKB=Sue8T+T;XG;}K9TC*=ec~CdA=R)`
za+&<YACb||P-zq{x{SX-yS6HSAxJs*s}-I;Uvv2BJ-Wd@OulM1GJe^f>>ojC8bnZs
zf_gYlU|P={&^;hupz_1gR7hF7%o)e4*4nR2xI*Q>KKcLp8vnfTIeqzPnb9SR={ORz
zxz;}olTCYV!>WoSJMQoYsKP_^k#7=u$#*XC6Q$vwe6B)Y5tnJrY3v;avgtVSE1R2J
zj#V2&_A?%vsCxC!OBbQv*=cWoP8M^ohdut0-ts6_p;p3-5K8ynFaLAXuKFV$R!N0a
zS!Hd6^)jt0lOj)2RFS+Z2Agg~<$kw6Yg6)jF1oizvmtC>nU7|%@Yb3Igq@Ch)%XV&
z2mU|8-Z~(vZtWXZ1St`aQo4~A>5xW1q`O6=rMrf17!U*`2kGwa?(T*mX6TLq2KeTF
z&-0$|yyraL=fAxd%-(CSb*-zf-v!1q+Zp>1QVq2~^biluozAtPEA;WPz<}6dt+nZg
zo)}9@Eh5(TfBC}yv+AO~)QHbCbmD4gmbrA{056i^g{W-724Z_cDy2T^$mG?YWP#Rg
z@O{0s)EL!i*=q{*{lbZc5BVmF)@Tx-?E1geQ(SQAeSVznDoCLT0_@V{3lnKV#Vo%O
zg7ZvozD~CsN+8U!QrlJUn2%$y39Gq*sS&HmuIoexQpFVx%UhMj%Db4wpS!bE2?;De
zLnwc)FZqpcGR6d-oP<?X??6Q??hQGdK2U99h{9Siwgx|~EKM76j5BeCpKR{R4hjE*
zbpF@Ri{U_{orN(fh=IXd(&<okTh|_$7cK!0(yg_#kxOLO%OGknL%WD?C!cE?|Kumw
z#F%3pnyB5S&{QU&I-7PsWp;NJ4fy1g)9LI$N**nI5F1q~27pMcG;n$J;&UD#E4jp7
zKbFoi+!ZxC$XBANLKEpZgJv(re^M%E_8lY`I_9`1)6>518>268(62{=n`=d1{iK<^
zZQ1EdZ*mK&J&eeyx0>6@TdSR-0!Y8N5)C|9%%oRm)DNUg-?I<sNmGTdkbZDhxMQo7
z4padUD(S&j3oDLK&~KfNnEaWUX{4jhCN9L68WZFYm=FGv#DBNZ!893jF!XQWzW5&D
zrm;h|+gE~q;$9!-klR*XL+%-00g&|l@L!RBCy$%dflVm>`hk+MywJHD-$$bIkXlEo
zoWETKSQ1WdameDaFIhvSZ{WDm{p}{!GSDKKhW#07ZNRO~lZ5HJ0~KQDrMF*zp@xcF
z)Y-(2;p{}ESp+`t!|J6KAihtT$U-;e*Lv48jAcfNRy|2i?ehKuzn^X`f?VrAJ^5d}
zRY+@+R3&=sWL$aqzKrC^8#cQ*SM+&i@Z%fgSxzFhsd?K(`@QbXqt-XpIj{3dNnqdQ
zhD^dvjK!$2{)LzPJ3&x(A%vW2hMPSGdc+sdXbDAm-gQu}3=AFPMdmssN#0;B>y~PB
z`rP%^OuJdjk=5KOEa{x<>*6HsjBW)9L(QU6lkxa@2|C`SW$|KtT&rF+&}P}!V(7Kq
zV)LCIGwXd^BQ(10dg_-gKqY0;>;B$`1-`|d3mS~<#I}6rJzmuU0fa}i<)Vv1l3DC_
zUVv)1YIj%DEj0p&)mTd!)F^3_`uy@wAq-W==GJ*!K0jjZk?eNYa^Cz-h$hCdf9_Ke
zqmbH?!P7MNgYad+1$m*~<Vr9>wmd1Hy>Oa<>%@4vGEe5bQSk99DK-38SbyywPdtLp
zx*=Z>i~I)vNZn!e4g&3`zEx1<O+=gpzPG>pk)jwod><6}0mIS!G-#r3)hqKYs&T4(
z1%A~wwhF%vD@?I5T3sy+BtAq!N>?STG{19I=UfJ(iG&3{(|_QC_ej5e&+_Z5V!8<~
z74Q?msByaw_e0~Ao?>bR5$*UJi~is%b7}snU<+qY4Cu(mAHP55E#oTaUtI<AJ4H=B
zF@>1lDVbOfaeGgT9;VxKYL4438Bd!I2_S3M&p3Pxesk6F26ntFqSP6s%SI&;;<Ow(
z*w;CBP7TlAq=`|*{2x*B@7s?MGJfCi##LK_q_W#ivQ{O-13{4+FlAe`{XX|s0OY23
zz%*go)p*Df9maNUUAl$Q#Fc3YEC5ondT<K2LyOwE%x@a=GtS!$C?lNe8L~g($s@pz
z2+{3KOv8Tq@Ngj7AeK+(ME3ZzT1%m?=&*nezB5WFVrbaPL(%AQC_r~ov)=eF36u<e
zMuY|G8=DJx1zWXG{~7+)6bzLvZJG*76NLP#-O{`@y617M@@#VY(R|UXDqq888wV)^
zSASFI;uu)*(~PBo(+b;CjD+F0a|gcS)7NtVHF&Tp?7OAg%~C2^p+>55B4mm=ZQEn%
z6Juey>D}GO0T$i7)EhXfm3oHs{+g|MZ&UD3Te~qjUL~{s=r(pAK%3GRRw}TMQ`DN7
z?dsQQ*UqLeSFZHv4v`ZH{PkPO07G*t*{8qM`mkFcpR@`$KiXmU8g{vYV8Ypux&&C-
z&566$4Y`fEkFLsX8r^kLzy04G!xr?N&S++qNF?t_=l-elMH^-~t)0&MP;?itL&IAx
zzV1rq7E%!irY!c;og8V3T+~5)ZbOWghf`O*xxOl#pfLD#V-2KQEwLUrK&8`!NF4=6
z)J5j-1Uh(A;)k+R==)-94}3Hd^V=GVYeNftk~eQr9b%RGrC1*o=)UbCgMdJ=&(QKq
zRVxy=XVoXtWBV1_OA<J6B&S*W3b{nYm0@)a1yz|@h_-(NFln;~@kwzd&*1h_#GzJN
zLa{0E(o<v%)Th8{9Mx|YZ#xeCePdN?4poc3bXj46w{yP_omvNzS{?}5{S21BSyleE
zv%Do{saEm=Y=G)53|cic!eOK4Zbl@?`#&rARM4r$!z3bQkCm#l8eIOw2+2>@ty$#m
zE25kVQpHL?d)4?(A&-8ByF#?(rMGtvg`ma$s>Gt{@8yMj;k*K7?YG!sseb)-r=Lg<
z1-(=}%N`b8=dVa_wN}|u=2#TTh}CMGq~(hnig>ga4~ZQlgD3M>$yhq~fj&U<F&UBP
z#};)w&RiVQwyKG!z;p-98k4Lu<+#1=>ST%EaBnf0j;e0&&Bb*HC+cgdbJy_<%Qs8|
zY#hyFZXfyBApNpB6=2)N&YExs$-qe*mw*m}+~7Jyg_uIkjc$=vCeGaUU~d7h{-V2G
z6wztPsrUCU9s5v`Ydhi&t3Af*%NS82XKVGL$x>c;j}7B=dFTDIVIvy%)-|O_gnGW*
zb1-(n1!cb2w99S6WyLqfztqG3<xKx|+K;#qLEAI@h+LN9I7t`@H0rp!CB2PY5e%3;
zUqoQ^lJK5GLJkf#PVz$SjLArrKYzppIF-91dM|AqNqdve0i4S~By!nko)%kkma5?h
zC~@7*)~}0_4ce0BB&3;`^!9C9<q1k<eX6vYumNI1@v-XSf-t2<9Pq9Xj#=x?uB%R8
z+@z_dcxf@1>ug*1h~b<Mry}^Xfh4PC{2VEl0`%^(tc43O`{^8PV4V^Ks$&?=F<?1R
zXVUJs6Y3_+wg_-Dei>>Yu`ux?VnvU{1ulUHEy=~U&5c&Q^FY)&@=uq@>kpC}5$K+J
zjl3AIwHYE$KD!@F9y0B1qI~iMpGa0xOil8Gp$j2t@a1=DKYG`wRWrIhWmY{oXXf~9
zg{UH{jo4O2{SDIny2C7XqT;DNqJq2F%9l@V?XGJRh&A2Vg;N^?YpZ;><weK+tX?9;
z5X+ojL0}6<X2^Ym3}vn^&1F@3syj1}UeW{+MA=8FjkuY``rX~>+1K=}@6ZbC6K9>7
zl;sF3*I!j)+UX8#EM_|URQ{Szrd%%sN1dpK^`z$|f9O9BI5n8Dzi<v!b-SkYE%*_S
z($C>7SSTNSc`vOnA&3XH=rrjIo>>O6OE9d?_lqA$Ki*F)vn>4$KUq(w&p;PZ?&%ky
z^o3W_62E`niPd)?Gpuw8QWUl`R7enzPxhpP<2;rM>mu@?*LKhb5#c$2nFZ?;^=-rM
zYmeeqfWWoGQma;av%$ft!Y+m+D1pT-gB*jt@iKJN@PZv`2tEFaiZN*-_Szp0?fz2q
zwn^ELJtvENa-B{ZCYIbdC%{XO#$zd(p<QE<2YLQBmKT-VE8|6nb!H!y%g+w(iZ6$z
zggzG<lgMF(CgJ8bVaeB0P5Pfbu=zshhqa~HIc@V~-7kW6X0NjzEsq%PwQLhzI<ZR&
zq?<xmBt#KyQCs=fuXY~QIITP3aSO~AQ>D+28pSNvme56Tp094+a@&RG+xs0KH*J%=
zz3SFxB1!}G1Y;0tV9s9#3R{ZC6b>U1s^Ez)b8s&EOr1_vd4D)u97}C0=Fg+dZ#Vf0
zG^{<!8H1Gapix4E(`&mVkSxBhF|c$ZE$g;CexQ-Hi4`Uzc_oSKNI9}vDBo0V8v4yH
zoM0d^ZBL!igxQx$pLWk#syt(#@iy{-Jx!&1x_4$MHcbZUnmkntW{L?ND7BZo6>V09
zBrDe80m!#(?;12}u*R^8XTPsX`5cRUx^a+vMXDxqpNmTBi(^LlWFLHdEH{%R;pC^M
z*7Ox!!skK^tB7L_F4;n1`^BDa*Z7P3wxv*ukY#YfsIqmU^WPW%cI<zN{Ekz+^p`*i
zhT2*n#c>ctX8E|1!nR2&+Gu`f^h>G|<O~L4%_$Y}UFbxj70>Z3z`D#f&k3CaiqL$C
z=`EDUHbvH0^W@<}o*&+Iz&}SMtL3Snx5(-q3A;rGS4ALkyB<&%bYe>e@TPw+3`5&<
z({-Vkwk*o$7R^K$0|33_kIu>MQ5=vIlj<ZXTu0Grk5*@ER_D}4Z0d=iAx$n^bi5T$
z{xtosjZFuZysX>0AdWTj8sH1T=J|3w0HsGeV^+B4Ys+IW9v3Vkh!C8G*M=9xFaJ8Q
z5U^=xF&x6sb&40y&Vc1rnj_ZT<xT~liXuWfGZ>TJv+xg|wL2HdhVSv(BqSurS<lUZ
zzTalh_wO!gZFqNPi&!>()~*KXf8bbbF~P+CO{4{5>hVt?wkwY`@q02Fy@j00?F};j
z#($x*tiE>&=8|j5dp>O>ZYkg>il4-18`Y4cQ!aY;OFq28F}3$aCbndG2}=THjX_f8
z&5J&XqQG}3Avn}#-(I)ec`*;&`3Xp6X*=<IOc6Z)ji0HbOg_aXgf8M<DLqapEFH5y
zyQc4j9Y<8MUHx5frE6hE(0`~@q%FjaVcM7Hg$=gjprj*Z8p-PQ3TiaoByqwfR{&%`
zIDf`!`T~D$<~>Na_TSwJ{~|9$KhkGlige(H1*7>?jy-Y8JErRCj0S-yqypdQlfTi2
zO;F|4KDqFK(yNtf0leL=C#Psjo8l6gWt6!ytT2bVi&P8gvK?;Md#-|IzozaD1?^Wa
zlDXu`7oG$VORVstBz<L>+`FN?U--Rg#**h2&hU%m7d-^K%6L0%46>*|Dx%gv>uhad
z81juZ?|n-3mnaQeWE=c!OjR<*HtXo$(b_n23iF5qEqSoV*J16quX#-5|9Y=(3oHu|
z0&xhP9y{DH44no*`F^SP7Wm_JXY+pcf4R(zb2t2dOz2MM_yZ*F0y{;QQ&zzw!sOJ?
z?hEorU@?MYyMwD=G%Rn~X46|%w9*$Tfnv0rsH>Q01hQcXHO%8#*lV`@$Q$dF3PYwZ
z>(f^~8XKh4Bt$~9X%?tLZu#8OC2rist0y+wC#JY!RDX*#OI8~b8KgAEU#u*?+TBA~
z^gnp2a0;K3+S!Us=@t3q)Eci$twko1g9W~tL=eBc3E6*+mCCmk^B#BZ?zAWkwsg&V
zk=EjGMPE`o``JIDe{>``zz2Ga2IOvr=AUyFn|Vrj2!u9g)jv-N{FB80%T@IGFIQ0n
zFxPkd9mWO6?5Tp{OGZL~-6>T^2mVfDeCxgqLdu^M&^>ACs}XpYWCctbdd#@7;z>z~
ztBDd|zvpxQ#Z>tvLvi88Plq2O+9rJPQvB4F`z5*Hh<Hy4x|Opb0`S!}-MSzrT&3@w
zVrbFbI^phI&=e@i2FW7-fRsj|6PbP!)k`4c1dd~e+JdkF<?$sj__Lcyc&&mUx5{J3
zhmCP30Y3rk_Tl_|-={?f8ykvj<nt5X{z_T(BDPrd;=@{G@|Vd07LX$n4sCZZm2o@8
zA5yio15c=?t>2vN>IFC!A#}v)CW&Xx=GNZtF}-D;nG=~T^qf6!Dj)W{(8oz|;*Ds0
zoKcA44vAk3R{MJ2J2G7=XMJtDqVTT6{%&6O@(6m~@}_eoPr0UYx<S_Yy0+DyE2PYR
zh1?N;I4$bru+ReUQn)ME%Qin^szjb_#QQl1LxR50rYW0igty!1v8$~-ilfOC_FMu3
z?A|LvYP=q16^u{^?ApL7JRxx>&-w@~UyDsE8{bY|R&UL<P19T78y;j9w%agquM7L%
z4eM!Q-jA=5x|JTEcoKS^71v^dtRLA%xaWvkIfb&Kle!KqLz4R9t^baT*CE<8#)&lY
zHFS2qrHrs2W&3bEq5-p5Jeqi<J7I6);Bez!{q?H<;cJ!MU>%O@U|o|uY~q+P2WdQi
zHQJ&e{o?duJCb5{YQwNNuJZOFn}x!NL4w(ZE8ou$+~xd)qmOM}w7t8LSf=fH2nHzu
zL?-+MZ?vHNVdOe0nZwfUlPyu_rKRxRd3}y)|6r8hZUl<ej8U~S0~nQYHCVym+fKBa
zI$J9tA7xAe2Hw;FmU?za_Vp4T#FBAb$rcre%11FR`l(zD#6EtT(PEs1n!g)xjXU|J
zqsAe4%CEq<un(c-?c8^ej#5yh`-*grDW<izMfd!kn|C*TnMW+S1Eqx@d~ot#ef;WI
z5|AfVf^nhcT~uUxP&Z`W1@sxtYdfsfZ{#Qab{%QUtJT6pw{ZE*Ih3kJdkj3q5+aB$
z;$~0=yq{ecZvsh)Wa~tHu>UCcn48NXsZ)>3E@EjfZoS+9vf77#<7X#ExLU+Z_SLEn
z9A5z}HCaXru=llYHwB+CJ33kFrKw^iaC2>XyIX8I$*s^bfEnlsZ8F`1IrbcGWxV=u
z++5*~-G^@Pono>6Zabs##>qtdQVCeQFh#5dg$qU!7?x!+s-UCpH|CW}N?#~Zo;SWV
zz0){jG#DG&3~eQb70-E3HEuQubu>gX2*lz2<#1b|5oZk<g@Jhq0kt}daf`l&i^eJx
z+&t61oX-@3&l5MMO|L!Vv%hXqUqWEtk1Sn&AMo(1ssvK#4Oyj;=uUncF}fZi?YaLd
zJn)(jP>FEpVxeh1n-6l@G+kuz2F0CJ1@=_$vJ~d1=Tk>9!PmXk(Mv5USBL^BLKW=3
zQ6B28lH%%wmrZa2zW4>lB5Xf`Obo%jn}b;}hWMbEBds*-#$?}swI$J=<+~=6wv1zJ
z(sh!)BII6S3D}7pfg!Qr9QV`i6IST7+lN?6`1cjh(2t*W_bv;lD6zm!UJ$MBg0uED
zw$wmU1~Ak@Ve|T3-Z>{KGSkXU?okz<cg14~dX@HDo?IiRR{#w}Czb*#x<8vph`+ph
zHEs#FE989~SHigBNYHzC)r_c<1A_Xvz^Hg9nj59GyWy4JPJeTTT(3EIDt<zyARAvl
zvd&jZKe)~H<hSaf;Ih3my>@ZXjQM2Qk7-@4X1Mp7h`n!ONbd{zC`FSm-}!Bv2FH`R
zz%Y-h@)&$@9^2fS`)uyR=+9%HT$8rr@+Q)f5jNIzRX;Zb;c%y^xe)Ei!V0u9Gw8iU
ze<@3V%A?t0B33tUS!F1AEFOq;YAh0G<;he(j1HV`cGZ$sS)ntYy?Z$ckP3coHHxX}
zOMy~8k9_*n88G~V6l}H-OvYt0F$VcveV7mDM`wH(-n~uL)2$778>KKxR1bVE|0zw)
z7Uv^n%&)G2(<+a1wp*`NVn^4$!}qE0u?v=l!Ll&tPa}D4(vXT<$8gD}X*)Te6^Ki6
zLEk$6!evOD)pf_Fr6wy~s(vAFUb1L#fQiK4gb%)nBQ>6E;*p9Q0iM1I_ANP4&69PM
zhshqPr$>Eon?n;Rpn~nBrqY{yxZf}onN7(7m0@E+Z^`*Tmu5N=l_`}ePmzU)p5E4h
zEQ}T&r)-xj7Gn)r3z}o8;any56wcpaz9t7@=NfpEE&M2@sB<~JShd|w6g=@(v#+gw
z^P!8x5!tUKV_i90P20ycm4<Vf_9@0N`%bmgOm!1Edu5MG=oHstt7k&E2l4iBs76u@
z9wq-;Yi-!6y<i6Aks7;ZWMN{|hK|YtcZW1t*OZ4tASmcS?Pj*4K!s9MqcmbF_(Y-b
zbl>pKTxd@HlO<Py0WjTMZ^U69p#klTD4Oane&DXoS5KcBSFdqeAR4;#r`M4}@C?+o
zKFtW9ipb6BCmiLr(pFKa11#yRX1F74sgno%=px5e^?LIk^=e6$RF}#bBlx9si+@MN
zZ8DUgXmm;S2mg&uP#=LP)fiA^N%oSEgjERDUdUV~B>P7LB=i?<{JB&iGDqL)-AP4U
zoFb1_I#hI@*T{HOZT2ySpO!tK3eOEzdSVF@)7#6FRb|mKc4F_vwU-zH6`{olQ}&31
zx2Y~ovXY!aAxqA_w^2o*6MIsT3QJulh-vay5Jl?Lk$cyN%MQ(m0=aYIpX}1v8mOjQ
zb8;)COwOM#8`BUOGJT2<sXwN(ygx8?1IfAVgtK9S!`x=;=ICIx2=yU+<tFp}0bGC(
z|BRh24pq#gLR6F>w%vymiW&N{*))|t{V%*r5j3q=mM>aVq4YBa%G}2KBnNj%s&n$c
zT2zi91_J0JB9$kc+~}Q(Gm}#JbH<k|n|sW#UJd(urY2v`7o&`h2&$mbf{moHE{~HP
zw-=lzhQ?CPMT1)WuiSQFVggdx>OcQQW&TfpOS_CH9w0OP2Ya3eUSIT&5weogI6}nw
z3MyO*9E4}>_tBc{4jqMK8+rvgm|$^|x;#u0lxQ{mR$h3k{Wdzm<+H`4hL2WTR4%(d
zh`JA(ex75(<|TJak~>8$s*|-?4<<30NdAwh`+x#`pTkCQOe7PWGHA6(Z>>M|=sVOe
zD-hi~+nqB)Rh9togqpTr>e$J;U<OmALocdLVU2YyBg}BFl)=17rK&eNPe48l2%0^B
zabSq-*Y%9oCk@REU08mg<_hh-Iy5gJ2)YPr`KVAA>X-ifaOL+?s-R^4RMM{quPq-x
zAchUw;{2d<cL~8-4<=>(C(b4Ea9Tk>1LHOk;rgFS@y36)G>%LV&L{_>FzU9GcesFI
zY=IIok+*xoUl(635_wK~?WwJd#(faSGyQC0xFeuAc>QfoT(6lUNLxh_=)r#cVQo=G
zdWU*#jrjl;*2aHhOJ3*Y^wgt9)!319M7t&oa@O{98djh!oZMJfV)cfEaf_0LzX!pl
zZ-gU{2!t4>nI@%P+|U2DaF>hY<cDv~vXw)p%1FkV9$^bkd&^QSvPC1DsXa~o<zHM$
zG*U@L=ApEutfoZIC+!M;BzH#V;Ri{bZ)f<8)lxUdAp#xMaaP17>M6}SAu`Vs+#Mxf
z-)^puzIVqx6tEi}KdSb#kMYM|%OuX%<XNwBcRHE3sL|F{WCiyX^hPt(#ZT4k(!wu1
z3*W>uCeJxw+evkh#mC}+8R#yywEn0d_zla5y28JK)ogJeN?!!5r<a~O<{?vDT2Nc#
zEt5lD-yObREZ?MXbd0St^w=`;$oQ_@WsmcE73M}ywVk5R$Nxhq0>xry_{UE*+qdM{
zyy4nUw8~8*ARjrROKhX>HKoH}aEDXaeZ)$xZr7jsc$}gfM4`79Cr^<bjQw+heWrYw
z5KpcvkBtM(YQP?T;hl-A&SEhBj)&*Dmq6kb#Rn5}SBEsWR}`sx!^IFaU(lGY@oU31
zyH_lTP{C&z+{9FK-MUwE9}B<pPfyjP2udiCOhN4<9L<k^Z!R{O97P{azXH4)T_;+M
zRmPxUl_%EYbWyHC!pyTS?ktEqnaa_|Dm1LAHp!<lWTWOAQF(<3Yy_s0GXK#58uJFx
z3}s-58vy5<(Je1)GqHKs@(W=~KB7K(^7#0Kua@us!i65Bq5dM6x#?{?^iE`x5(vd$
za=}-|MNtdauedk<UO}56sXDENhZm)s)f?O0z3SPuDypXRR0<#2Vsh!K_3m-CA+^a2
zGP>O2ddGGgVh0^f<#&YeLJdIB=Ii^`Kn^@$Vf^bXLRP%g$L{Y$#oBaX5`De%0sUQb
z>12*eTEv8j2}32(3W+&IyuFjnvP+S~95QS|!e^^(h7}qQ@U}Ve$gj;`kB`!5899_S
zN|#85jo}Yn+ZSP3oyQf*s;bPq_t)w-Ml&9BfOAN9*VT^!<EwYJX>)^xmJPQLL?$D>
zC8hUnj+^d%GLHuO`ggoU_(QhG?iz#4g?e|Yv2j}ziHdQ>kz8{R3u|m~P3i7)HDk>K
zBZDz*h7PgWxwI+zqx+tQUJQ?(&bd|;s|R)aq7AU_F6ygC&85~dD3@k-UQRzNQ5t2M
z!+Vn&hsN7hqMX9UMdt>id(Xk+F`-Yd@Gd3R!$J%nC0CPL#%tz|ri9B)Ra>tf{ns1Z
zHha%6-7lO(u&eJzMA8#jgXl^jn^!AR56Ah>^zTj}1+Jz;R-91jgFYbmndgx0n9%!;
z`RgI)QQ?cm(S`o90CWLt#q(haC%dt#;{h{FK;#Z{0-3%;)z++G`rBWsE-boL%BA;n
zL#><TOGbC0_b&0AK%w>;lB~AJ8|<yPvc&qU6kc!F9ctcG-kF=?ByTtGyK~#^AH1du
zO_?^v2FCC=0=5Ug>fCacuVA`Yr6;v(W_YNENgRsl?yd^;`r_jLj*qBEJX!$z<=^0Y
z)6IqJ)|f<*bgR|pt;qm5@c~D58T@wo6gPz2XP+}kct0ide)&&`OQJ|if_}+8$8ymR
z&vJ5k#3nga)tM%oh3*5K`Mm8d80K^2ZaiL;)x;*p(9WpP$j0t;YAKWG(Dd31T2^;S
zhH@++L6b!(MxE=(9__(M(DCD=`cJpn+k8vtA7KkTY>qa=Wj7_X@IYh`ulmUX&bd}f
zqmhG3rk?jK?VIU5PszDQz>V?EPLUMu-Mx#`aVm{wy09`Kync}`_+lrO*{5+`OZGgJ
zKDy<mp!vpP$7u1MZjYt$EHbRJi0x?UG%Nh;+#}RPnyd2KSdbiydN?Ks`ND4={d@jJ
zUvG&NP-%<d+ZVr^GYjcJdLP&e<%%Qu8;=1dD@Ak1NAR=lOuWV1lej|a9{Dyt4G)(Q
z$Ec>}%1;k=$LEzsa=)#7BmzPEqe-tgKnM_J$K1#t0oTg|&hpI`p(m<ZDiEQr%-FNg
z2Y~wq--7Gh)6#pK9>C)m)MN#MH20vjw$5y|7<g~8k&j!ddP3*6PcxM^=D+U?#Y*pS
zhAPC9c{;g}`BC0NA6v+&7Lz{StHByt(E~w}1Qul?uAc&BP$%&9J`FJs)1_&8@jRYX
zWtrpw1naLhqY&W0uY=9fOOW=hnUYPja~ened_Ci%-&yc>Hca&%vtzgLs4{sW_x^zC
zslY0J*r)ul_AT?3@5z+o_!n6jg@lXjn28R;FXd71{9_h+nFC2_Hea(oKm6|3x0JYx
zT9%6)DwfOc=(W&ko3`n?az_Lucnk~zouu>ee0lj+YpD_ls-=X@hkCf~XJd}5(ten^
zAM#;Y?#w}gipk5T5u1ThvO6b2#k`R*VeSiWm>^y>h)vb5NB`M*o=eV>b^ZF)Y6w=S
zKCdvWP))Pxu19PAb%j2@+eZX}&>&DMzG_mq1w7c^&B#ucsMhXlS-9uux%9=Us6Y&w
zKW05`X&j&!OQ#)>wLv+vjc#E)_X)*!U=uJZJ`nMWDS9N>vur#vWZ&Ir_Jo^nK^-~)
zeAMNlt9nm~rVQBy4A!^i%WITM8?R9qi}EAuml>rG7^KPSQ>-3chi3Z?$pKx;HiyrW
zb_*7v=%ZCN*Aj07ln`DP3iedbE*}H?hGvg^Ew`4&n#v2}RkyzyhaB7)y15D-2at;J
z?LZ6nk`q~dR4Op9zxHJk>~EPmqp70~TQ^yc^9Xa!rGC)9_2?$y%e0xkLx41c48N?B
zyY3jgzg+)O+SMaqXirY6E5_laWGns6S<S6zg5^GSwMgN&+t<NgKO8!ZJew3cpogo_
zwM4}>?Y+ges>YsDCHuLD@5#?OA*>0g2Jrr7=RkwdFf_iB_4>2ks^eWKc+-99L32Ym
zyur&)AjpNHGV45&a1VabOXZ0bSH(cFwQ|T8w*L0(KeNyN^P(Oh?RL4p_^Q#-%6+n|
z%%*Fc>t89+Jxa5Oi^4cIR60BGqJ!I~*XJ8OOd_&9tgQT_vA%C9c&7zv9J8NKKy;l?
zy&evuM>D;}?330LA6EpnWk5&n?@hN9-fOLhYX$cDpRSQ>5=BZ2?ItJYlx7b2YvRfD
zp-^pIDz?6iz>?t)!nbI`TAko>G?-XBRBSzDt~_|3ky!N_3mVING|OcGLd`xExH<9O
z-kE4>Qn<0`(rA!UJ+fBrdxnJf{aF8bzNc9+a|t0bM>gb6K{dktS;R)0fJJIM-4<5p
zr#g{Sm#o0&wxBD|?54ZTF$O`*^Vgf8hr8sml<z0N4;SvFhmCoCxkb<Rt7K3KC&TrN
zI)iyf^n=*;49}ma9B(g4*wun_1vGnv(_WSSYOza#OY47oO_KTrO5qv0=^{H>?P!4e
zOrfw%<Y<fu0p9KV>^JrX>SU{!+VH?9*IuNj9WRI5Jkvcz>;UuD`%53gyM`3_#a(t{
z_bnU0kI>)DMNnWs2=qkhDJ?%wHn?Vo+rn6q-Wf|5)?_@%Q=LKKG*ZissnqDnwNj4^
z+<9Z=G+XUESG13|`}9!sQ!A>|fgOdIGsSpQN@^ChTxoby%2~uk)v*hs7+SOP6mRkv
z&>o(lh-PEw=dN!ma}&yceFeqLR%+8c@spb|BtGCT29Y|q+aQ&OT5<Jkp4FtJTAyPh
zLnFAPwh^c`7tho04Jq8auJ1aR5|00vrTWDak@J>!JpMa2CY~-S%!Wp$nY4TZJtR79
zlyz13^WAZ4>f!g%UuN-f=KGzBx`bXYw3@|n>Q7)2IZfdL6G#z|)gX`R>xaFyqh~cH
z<w{oU?~K9M`#B;5)cW~B<>l9-_cW6RtoIfH%QqKIU=)*!u$r?9&*K9YVTugy@Ji09
zL%E>3A*ZMCDik@Bc(&%YtV_l3W~hX@V;YICM(d7EQ~?#?dU81#`JZnu4>nj>mR_Qw
zkw+KGoOL9wg_H$gh3eo4QA&c#R`M)DdwfE*a3aUYh}Kw%vS%zmU+g77`AXM?#9HXl
zxasZK>mO%49yy_pIITXR-(;fa%u?h~j2k}fU(Pd*T`d=YUQG8qc-I&xgRC)6at+VJ
zb$Ifl1J2uYs>RD;(K1sNqFQIOHys>k&rBA{Z7Aw<j(orpecoJ>&~&n2B!+;tW&1DQ
z`Mb{#8$XC4!vExPFBeym<gC<V#0x9QxN^u;g58aQ;Yv7wW+10+lh#_7x>E&4piGjv
zi0ASqbx9%$D)ydyM^f<3Q_TdjeeND|Jsvp?>vGqSx{eFGl90fMYZW8pnrb)_bW?kM
zmCCn0Ny9Mm4ZqrGGM7N-(~z6SoD9hxMsyz=kaFV7NjB<|EN@W#a|m|&LU>II$`Ot3
zfw(}EoFe}+F1&g3X;ujn7uXSt?Fwq$5N?K3_1O@okdkNf?pvS$sM%O_?H98QKY9o8
zFqBPamv6m!Zrk&ZDa!M!Xs^L}^famtu`6$4b>gI&UIAu}I5hPc<eYz3w;=2BZeA@&
zk)hBy;qt7w8qIjJ+dXZ081PCewOB3xLV^Eb2G_nY*IV3@Pb38BO|FN@31uf_fp|{U
z7*$-SGb^0ykJsJ3Thc-kzJ`R5`E?$2F<YiauGwnxTVB{qGA<puA@n>F%rWL^=aa3a
zsx%an<5S##JH80;>ixtjO${+&9tmESD@E~Wl0=q^ABsDgAddNGg#Y8%E}0`+=<Zyi
ztRbDuk~m83lwr1BKF#2qa4u&cK+9>BBg9;%rL@%%?dy(2na@TxUW(T^xx+@iw--AK
z5$4{=e5QWPiDO7(C!*-wecBJ#V&yZ`4S24GR&Th%vg&TK$aO(b@a$BG-s-m6h6$NY
zoFnX&06rrNV|<<PoyKa?z4i2|=t=$kr0U+`R4Xa+V4atqo(su$<DW7qE|c}%>XK3T
z1LMy(hDr8DJ;7b<WE1GlnawX5swLXfo+Sc#?XYDy(Gr2IM{Y~t59!x+BBei24R5r_
zK00sglmk9&GZ7GMqSvFe-2$Y9^2%BN|77NW6DLkGNFQOc!-ENt-AYlQTDz=6ahTuu
zc{w4it)hSQHCX%!@pN8DuKxZe%PKU#==mry8V?Q*<yE`9@8UZdc>=Ye?j?4~6tO+p
z1qsz5BiuSUn*>!D8F%0~D&Z|S)X&N&0K*L6Cmd7ADz7hIfqY;ZZM8zhVsM8-N@YjV
z5%#n%vOzIa@kWX+`N%SWMagW^Kc`3^DR70z5|8XRwTh4*muCH1Z+FJ(%FI&^E=wYm
zo*V^hs=J2@b8LI;hzFUXs~j^U_&b8fW@oER{jIbGOzxZoLQOoDBZ^9>f=*seS$h()
z@^fRcmW1n8*r+`<7}=Ah^yP(oYpu`CW&@itLIT{($^8~kvP+Cqb$v0}WRjc54~!2V
zwDj{lh%58K$xe-+SMun?IlfbCIrh8NqkK+{MLn6dy+bPD@qXCe%LSm*CIu<Dv#JQU
z=`ne5_C46A)+g>35adqvUbZFKeRFTQ39IQ5=DSyl!0cop{R}G(czk~^e0NfGoHpiX
zZxFTW9qOj9lWSN%pE*<6Y4#o4o2Ti;K?#y}0_7!{@YJ?WG{z_cUpVD7i$Tq(EQQ5w
z4>M#?GOg~kyUCGokZF75pXd$$e&KhR=8FrcbPzWoT*QdxVTC&(cjUFAMxp3b!E_Cm
zw4!;XK@|pq`TDw~s^eNqLx3~^T+AD+H=Z#kn@~HydE|~9HLJ&8PDWS_rmCq^s}ME0
zD55zruFLCBszRJse>QE%!Vmx!!V<wUpA>-E4Be*1=3#fXah+pVl(6dVu5E4^N<svJ
z>DbiqaDt}UXmk!{?ce44`yB!ik0T3t8DE{MCyV<iuu<|fM;=`S-%H0B<0KhjM^S>I
zN!O5mW{R<;)zg!QKgvA>qD_0EuRVNg%jN3Bs*d1N;ECgYnL?~5$h%B`?!=-5wSPZW
z)EHd}X2Bws{%0gI+VT&t5TMMyc|h^49-WLd#7c|{AYZLGXTZ-8_A=YUEra1uoUIRe
z1xWDkD`O-n;TIpHKnJ7)GT)*oM$unU`tgWTa!`qlroaFA7pM+La*l{~ObUgE{vdM^
zHP9aQkB1p?w!a@EM{ec`AUJZI>e34!z3ioiYSEyEbqOs4Js)qa^W`U5Qi0l$Wm;5m
zB-(ZQ@hnH+flx-`0gs#Xf@|W_*!sh&eH*$;y9MGKua4--ntVwJs{y0>4EHL%*WauL
z!j`2&6fg<CBx73OPg|V6=M6jAY@xvrF`!2L%sQfdzQr}6iarYHt`tOP;;^kh_R0Ru
z)cf8SQCNSoRRUayXSO`Y{?Xh#s-mS%#c$nxDS-g>P!wDxMTg<N7S68|g@<t?47Kfv
z-Ky+ne~Y!RCcQ3ys%S!tlco5|UM8f+G046e44oxBsy6^jbJm9?JPl*sfrVibk;I9n
z(Bpi12EBikA5XmH_U+%;<^OZr?t)05Q#_`Uprzr}R8hONU{)?WnTqjKfmxt&r~lU$
z;o}Bq<-LlY?kc1Vf48Imc;&Q2d$z~W;I5kDP{N0Ueb~a|xP8<hSgp}Lfc7QXQl)EW
z(T{&AS^0nF`Ok4tOk8u#&awQ{I1)Pe&tiQAa9$8SA^U%y=^64{c(stD!H+*$v;U;=
z<e+vJ1~?a{l*^|-!ub7jgRBEZ*M3Xn=ug8ez8BW<bf$FuI|KW_g+Bi??)M9QTH~jT
zj)%waVGEX(rDb+u;b2pf^0BI_YEE<U@NU+l@6)KgA`XVc84XzvaR=kZ9XlwYn!1dN
z1$q$Z<@GmriGRNRf4v~{7nQ~XB&!rN78Jw|ch@4?TR08KD_CE;YigBCIXDQ0^eNME
z$FhWhuS+UUR}!nzhDLw!GI-q}3jjJ<>i^6U|9V?_8qSV+=a|dcyv`+01ZLgj=~SVj
zFVTMi<?(xhgbzS!=dyEEMl_UT$l*FUxUwe>0$rV^@_qWFZuZxT`5neQ2?9a9cW&Yd
zViPBCcNdb+*cjnlbn;hH|7CaNWqdmo-whvjwBUJNcSs{zhgX4d|Mq21Uzw{TCaOc=
z)6~ot1ai5V=QID?ANU>mi5dri_V$PiN=q$O)Cvj;R17pA7Yjb^zVwGFY24|)pc7Rk
zb*hGelElfQCU!GHULNDs)DtZFNMex&9(d_f5<06y-gTCA1T*^&XrO-^ISl!W*65h?
zR8xrUs?bFcp4r+0lI<`5q})GVWE7xHvj7a=!np|Kum0;{5k)GIX-5+oSm2LMh?x$j
zu{TTxQt)E8()=4K|L3ir2RcHVPRsF-i<#b{m6exG!1tDykr{8JuBoX}K{$WoXJ%BC
zwOQlUXD~8KS#T`j?`Cvc9R0EWIP-ttxA`J3GWxj%g8U0H{-Wd`f>=)QZ`nDYTs|h#
zGR)X27OFhWFVFrw$i$Q;rV6w8Vb&M%zw?iOGLk3g)vx|~VghMq*$L-qeVi2){FiT6
z9>RfFSY$=Z$f!K25+~In0<nW}x8+NI8e|UaRD6xj`(Gz2K<Sr@1IXm}%Or~kp3;Qz
zyDYu|H(Sd<YjZNR3au<V@3ZL{<e<wevXZO3%$91DI&BT2;d5-pg}f1%yniD~{_$o3
z!oMD$93a!uITl<J<gaG4Dx)a>W0YTeb)^y@_Zz>|;1ig^|Ap&+{eCBWq(c>0K|#q6
z=e>3GaPl-qIz|o-uh&3|YWClo48wsW|3-fQc}o^8+Jp~?S+%}T_yT5N7gT*;H-jWj
zmQf;=qoUD4EJ9{`dj|ZkEB5`bd30cj!YcOf=6U0PXDaG^s?7B-`Z0MHZr2oBGnBM9
z^1R3TZx7;MoZ_Wa#tMeWcD#CJLxvW`N9$p1S_3e&(J^nZWvqpQyF_o8$V}qDJq5u6
zOpy*XU{4b(!#@qMv&9q3-xJDpGsA_v4i^;a4B2mF$pQCx7@`%TrX>IEVSXAk<FTEJ
zIny%Xw7pJcYu@}z7QBe#Z+91HoR@}cbo|%qZ%jHcvK8S>pNA15TDp7&;rqS_OYj^)
zL^Qg)1KTmhFPK>v%dQttUD}@=CWxkwMg@96KPHVvxp+n#I;=Ae$i~M1Aap*E>C#qH
zA*{i|j{O%J|2HDwibSZ=mYh|zf{ZyvX<Qq{lpKi4l|DTZBqfAHG+Z=)SJm33$$*3z
z{%f=r_32x&nuJ+{cZ%a;%Z@!3v1BH=Z+0I5o?}%A%or1LY{QhKC*SI4XH{Nt@Rkbc
zS5B%vJ9be8nA@nC>&Sc@!kzboJ~#{vgh;onZ^Z9!<vzB4?BHRdBd7gOzVVL{zhK;V
zor)R?dBM)(pUrw$;lJ#s_hPvueZ}v&-PLP!pI9;>CR>E4XR@1WSFhS{UwvP$$-APE
zNO~V(c&MqSA!lTtRJ@YPE>f(+j&MIH*_U)K1FG7@VGZEsg2@(D;I|#5HqdcMr(%`J
z^S&0p8;)vQ?A-}|>GxApsVJUaqdX9zL4?C$dx=7G2oV5KnG(08g7ii~DQH&CDxHNM
zHn0)02t)$Ve9e&`ohs-Yd7dP%c=^aJCWkH}czw*cYjMO;yrf5I#RIv+O1t|b>1RkF
z8vRb0_2Uc-94_zN6)2D7h>!5>h7Cj=4QayO_zcU|Xur|gx{Q0WJQw&xEpvB`@<|fD
zojhVmm}{k<q5;FwP%mN`jjzU0?#W1h_bn@qTlvkZ-kuamoef61tK=RPCSYVob@DL)
ziK(<s>H<2W<+Cj6sPaDxn`864Z7Bd>zb?79lyBhmIc0L+I2<!M%uUc$>s$UMV4D5;
zHMh`L8zorSRd@2ER$l7oB$sZ0L%U4?4cqLiPQ^L}L}La_AYK^^`Nvs;be@%U{p?kV
z)EDr}KOP73S4fA0r<|7XtwpYT6n7ibMp_2C9^?!aeGy$Qve$&UDRMB5XBO;8IC>w<
z<OO7|h-b%Yv<#HhSw%|LHu!kgVQ;2UUuL{;^aRh*?LFKef+0D@ncLl<>7U;k?kD5u
zq(LoY_<--k@^vi|iFyrQi*CC!tnAB-sjsxkr3HFA1YPECCiE#O7oXan^r9LPR9hFw
zbS>8({0#l+OC}FnsE{`E`{h4hY;mAryS%6pO5#N}OcTlu6Ryaj3#35n$y1nLYb~zQ
zFIaPIdxSb77_Ct^$I%}(JwKpm#tOQKZ+eE$doFKmz#&D%lfa+0B<B&?ZX@pz<_ix5
zdUaA`*0(%sGoOG$$<l4$F^D^v=c#Wa@(p_Xy+g(QA%I=q2J7U$ry``ET;guA4*Ff>
z{D`#g2G2g@jVoo1=3~0PzGlYBX6FR0EoG^`>nQ$iDZ+)*e7S(+4r$>B>wK~^Aam?b
zcNp%V(chq2^m+e?+Hf3%JyO7W<}t+%j0Xrh+3Oe4po_rLn~s(l!Jc0K6G&`0s$*^S
zTX&o*Q;d@7n_EUSt(QF^&e&zX-}s;K<{jCF@SivP8hJ72L{<8O7(#;5=~@A^bJvNs
zZLY=i1%G5tI%wBjG!Y__DxBwR8Dn%QysM;Eupx_pGMN?U`vz5%jc76oz)cJMx+bZ$
zAbP#p5A8{vADN+7o!*hx9+{_RmF=2aI)+FKLY3j&L6~C+2#+jj>#!n8Z9n&Q8@@6p
zCKj6zpcoOAe!EOa!aQp|d&O=ZE;V;#>g_mn!vw!LsP7zml6F0}2YtuLF#(=8K)`ln
zOE&ED=6y;4NH0w+caF8Mu52ZP6lW=((-umUEZ_X_^}eUjv{f@(WERM?I#^l$GqC(i
zF@!zEj!3Eb8PB>Pdqg}Bc|K%WMEilNwebUrIzep8fuZBd1<Bz|T`#su3rgD0*wkW+
zfk`@b!+Rtts1Mu)5P>0LN@|TLOs928Z*={5EZqHK$bF^0M&`k8?OfUFDsst;=%y_T
z+bQjOL)viQ{McuAac<f6uk|pRK)F?G*gUiFzp$WRO{-EO_c+8mpS={ww#9CM6c6m(
z7A&nW0YNWW;Isv9r$u;u1L?`h6Ox{|fOl6-GpEn4Gss3*;h#y`K&0;NHVWQ#(A7tA
zAVpK%P*U(C1^m>H$fRd+<f^=0{*THk!uSw$g}Z^o{HyG+p7xL3SAFYf)<Ep{TV>S^
z4OXVr*h|IX8rQwL06pi~Y&1oDLMg*r^HK}Q9Y)A_Nt+Od*9#!9O$Z*0AOADx8pWED
z?)BscIy(A%W@ag(uU5!4frTf$qlQbo@h8vbkb46!v>6A&R1CD~W&{W$i)!htnZ(97
zmTg~mt@u9CoaV6TZ;70)1|)(6R=JU~-94Rd^Gdz1QtkkrANk_v7Ugdg?(^i6bvS0*
z>*Frnv-b><)1UQ5o4vB)ir-}5Vt?gaN-y}5ZyK#Azj*d=zHfufF>{o6ndAufeslHR
zW144io?*3!$HkFDaqdi)BhQ2q7*$`C7S(Y%3#_NCzP%6QoQt>$=g4HgK-VCE329#k
zsoA<O9O)R1Fv7pE__XY}A>vc4MZDi<ZeYQxX8p{<E$@xfDNRgNQPW++0d>C8HGnlE
z@&LzTBgmIzfT2IlXpdRuZ^v;j5z2dg3CNe?O;f=?97VrciT`8VTOth%JCF+0_=`}7
z;bMpl0T<>@9jDQ!Eairnva2(x_slCfCsCX~HBT&dM!!+CLK7tdx&8Vpae_1OWIx_j
zB+J<wqx1NlryR$<%Uu<PTNUzH{>gM$mxb<@G*hbfq4QDHndngXn=}K8CQuj;eoNse
zH7vi!b&JUI$Sad#o^2ou54g3CJb6}Ob^=`4&iD{K7(?dCcWG5F8kysS`2XDuO7jsz
z(#E#VC(4)+eP)WnZ_PtT0n?v!Y5);cQH`AWL3peNX&i<LrK>nL{uK{;Wj{)ti{VHS
z3u8VfK^TJ62<e+M8jSlsU(t17R82GJNJ6IzMWi%=$cgkcJTl~nd$^(2TR*#}@j-+g
z72}y1xw9`VJ?-Emh^HUA_VCt(ApUB%bqNtTbbMU}ReCd9mRBVc?H=4N&a3CJ#%{=P
z>~oq{JPq|X!)GO?f%{&ch_$5M2&j>LPIG)uri?Qe5L)rEr#J`^@AY(SjdN}!^QQf+
z{MF!6sRLig_07DThwI^%M|#dS0U|Zv2a8UHhVbmzR3&#xr#Knm-syeNg$<r`R1CuN
zA1F}sppP&K5}1hug_Sh&*&#%Bc85ivM{n0Rl+W&*9B=!)B}iFV%bXzl%{|Q9Sq%}i
zZTZt>5ux5I0wzO&_aXac&1J%EEYj1R*GS!MZ*-s*y~{no{97UdBM;Z2BF11~vMTIN
zgQ`fw!Ot$T-v{ahuIU=~(C{gSewzSSr<G^v`v_LLBV;`HbS!1LI1;TdYvtR~HFK)j
zul$Xxhb~lf(%6=oVB%#*9HS1$t4@mT?Pw_f8s}AmDG4=hD>L@Ib@_Qi|4=XKG&Bo~
ziK+10WpBw@2_TmPmG?U<SDuu2(3dtubS<r+#W<P`$Y<4hZbzMd_wf+Z@h+>3kRMc~
z1#y{<IHUZ7Pf0X|a@`reGuE0$!I;;Rv{+JL+xrhYt0pMNgo981Rkqojp8j;-nU#Ju
z0k|(RKL?qTMQ_!9fpJtkvw_C|Gkj=kqwRy~vE>C1VF~th&rlrpnx?=ml83h+%Dt^q
z#chYCX+Ok&BY%X13GOsH5L-nb?Lw$Q#Wzi)l`nOAoz&=d=se;yIua`wZ&K|`f)P=T
z#<FPWe%oz6vJGTdOJH|Coku40di=dRTj|{`SIzjAEuQ%ppdkfHcj&M;P7~0C=-m<c
z4>&W;s}*9>%!;;<PamH4(%25as{fo!O^K0?h{!NWzCfERM&Bc{3A=8CE2IO=6x75D
zej|$1(j5^w4>W493NmiMrr(#FWDIn>a7o7^@QwU5mvJAd(4>9my!XuXGzKz~;QAiI
zF)Dm|^}TQA=?}<pVk!$i&~d`%jW@|IEGA)$BtRWMFlJs;s{a;g4gIf78L!)s$<PK{
zP}5xbgTvL&0524msh9poAI4hPiFUpGH^+exe86AH5j5@VRd;=;2C>^#yaR3o`3fMu
zL3019`@F|i`n4h$A*+YeXnfm!;ppj(b;KQf>d_!h7WMEb`zPit@a7_g?raPye3H(#
zBDme}-bc#Pb#bm|__GM@qBF#B@ItW-9seswbl+iBRciKH$8Lem={&B>&j{C2iS@uc
z$J;#2OyTR>-oDNMG)w<6VjbYC2}8eM*PIIyN|u$2=23$ArF@Yya`Jl;aF~#|2}fFa
zLHaQ_2U91nH~NjlxRoQOC@lVwU6Gi`{su)(Q_v!Y`X!egDj^=j{sy!n<Y-%GoPEkL
zp)Jn_xt?UwiRx7+4q(NGUAQyJVaeEbP{$i9908yJn-K&YM)&9nmTTRCHP<jxKRokr
zj-$C+nyCqmLh;~pHAv#nBof~r?PiBi0V7{eOJ9))E2Xg=NqJRQ2T^3#BU)(U+WkBz
zgWl9!V~DI*g2Jg^&mo-8FY}a(NP_wgnF!O|x1Iu3dsmNFS~p6~HXQET`JE2=RjGv^
z)RWI%aHp+KSKFRumz)7nOS6X*-mu%v&OhxNVlaGM=c9hP=BoK-zA439s}*L`z?R%b
z;dY3awG4{O_8Yt$%{igDnpWLb4Po8#X<Q>EhTJ?TUw6E;0cd+)yt7Jh8*MuA4R{al
z{4Pp+t^bUh1O@mLVTIU@_zYJ_=Z-lbQw~><jHc~P_xQ<dK@B85#Jx%ne6`Sd)5)4%
z>}-2oQPn23FQ+3C8XB)PEcL5jFvO>NlZxXXVS}{Wd{y@`2sk-~fLy~r5bBo0vhTfz
zi+!pp4A>$4ZO*SnU!ObiBw&n$^rQB!1u$T{(WmI7z@Lm%nm64ia{m$iH4Z7)qo;MQ
z)%R6i97ReM@n3IRQV6?MjmBQqqa1sJC2ryYeZ+Ig-pk)5)SBQx2C}}ag^-3z-L!yT
z5iX)H`Zzn!LD>Zl&3%R`{DwW_3#~_-UcqI;`#a@UBNv#`G<v>vDP)z<cYXTbfM9~L
z`_jBM=dkfBdz*iSTKZR$yP5*2_C!@hl`$#ohysoGIt^b<osSXHJco`CSjQ0~ep`mu
zwe)n7h0fxZ5l`!_evP?U(JYL^7~R8~?di@m`kf$SR`cAwmMX{neol0A6A7IB>u63M
zDcFI~hyzw?wi;K)B8{o?IgQ(*nzR|JP-=x3PLR$O-2JVGTH6wa=nO}FU^~Y5zSG+G
zr^e$rih^!?``6v$QK898Om~*R)tl+2_pZumBz=*X*C#1c^nCun1v3L{aNj`}(h(g4
z8Jp+JeKVIJ!<J4(n6Hg!jDQ3Rx$RcbK7%6Bp{C`hh8qtN#2z_!zb$)tx$h%X!r1dw
z4lBs&3&*2^Gl_c|6699qE;w1qkX?Iw!we|x8T8g43cN6Lp_E#-j&9oywVRKYvSHC%
zidY$6Egz3Uv26=Q%=x??VZcfIR^Lf5NepB0QQ`<*?Gzg)ZM|AMS#exDrx;{$wsU({
zP(xX{h(v)*%vTnZiS1{cm`ch|M9FurokrtwvAe2m(|W?^nOH?87ex^&FbDUF6@^zn
zd!fA4n>`i~<Ni<)A{koqS1u~U{cmqJ$><`0OLZJ@cokdhA>C}v6+)e^{!yoJbV*sk
zQmQF!!G;YC9dF@Q2WKlW^%9A|2XSrf<-OQSH94@p7LxscjJ;(*oI#ca8r(u~NN^{3
zaMxhL-3bH|G&qgZ1Sd2ufnW^;*Wm6Dg1fs1cW>Z*?9S}&+qao{{~Bw)uDW&W*16}L
zGMM;u-RLX2oh?#%b!Th#kS+vL>=Kcv0-zfZqM&dmU2+F!{8~j(5eYvUd*c~nUJKrf
zwpW*Tk3#a&s<kB6HMt5+S2tK4rJUY;_Y&Vzh;{jih7T_pmI%EQJrH)Tl+Wr?R`C5%
zFUNyb(dAK(`K)i1%kt~pH$Cfj+I75<=#D_w>3`{LWE|lB1W=2YpK@{|6-L7(PGmBU
z10r`7u7ZsR+y+69{+=2VF`td<vz0K!kkvIX#N~LlbQrQ3At?+t9L_A9UTMT~ACNzE
zlE{?@8E={S%|t%72#y!*nI3V?pP-B7bGSTkgz!UJa<N#F-uo<LtP%^Gi@lEl?;rcP
zmAmao?r19!U1XL0_H%KS-77n3US)t~U>cG$kk&-idPo)WHNp`VmWfzY*r+X$hM+q4
z{}?Ard@jYt^7(E48ywsH+=%5pD?FLrK()^M7eQC;B43g@D_h<(b${Aiy&vU;rJ`c1
zJv?+7G<<391MNEQ5}!@UNHVn*RnsfPsNCnc11XiMabv@gZCYt$fnP_XJ>r$X1dr&1
zLr2q3jTQ;~8-SiA>y5n9;jtoJ+a^fttkF#<v|`<GR^Ya8H1=|JJFwiVA76>j0_Ej<
z(9zlLz^=86L|4~>pF+=zY{4euG#1ykdgR(D*3YZgdr)sym6rSZ#R(j&)ru2z70U_I
z4y-ORRNDMkNIK(_SB+xZ8VTJeeO)9D!?PV#S9SMGWys$IM9r44PN6!YgxL&I_cQ~&
ziX~@e@E2u=+Bgece?%mgRklRlb3j$yY=2CNS0YGKNl6QRE1sj|6TYtKj|p9Opm~rn
zioE3Ve+jzxeP6EPwned(N#=lX{{S7HPH`O<4VnMzCK)B}j3I{Eyq*teVQQ$PbsQN9
z*-!3H8@@anmEQYl!ZD6Wn5#yV^YP%+6#&Q``4XPF1UyvC-zm5mK2e89{MVfaH;FI8
z;?-M)&R!?}_J2|C5RUZG-X!Ld6NEx_kU*BN5YRyCFz-$SfU!{)Dv=KK#k!4prT2-x
zkNMxvgzk}AnoYtu5Cxoa!(SChDsM<SMIYm7TF~bE8aC?G3R}R<vIk8pPttKs-*!I^
zN0PIKSlMEQGRl2-E1Sa8%Pr_$!>j|Z?0$jdw|s%XF`g|=%u!w|%3w@<KWXq$&3`EL
zG7imOu8d<_GQQv%U`Ttd-6n!88_2gt04`OnowD0itwb35K{9dhY_u&BftzO)sfWz^
z$f;&EqfLOGlHV{#na+OUe&6f#2L3AP&Z;>r1?>2_j*;{$b{pYC9eooP39t$rI6uI!
zH6PC*W}07qJ6;gQ&kd;TGfzx>#HRWj^Lv1!QDdOC2)=8=3~P4$-MwXPkE=7Pm`{US
zEXfc?j&rATy&9PU_6NoD>6>uwV28OGKTnw}YhsBjjaq{V+oOMQ*mqq&&ogQ1O!ENj
zCm|9)53jWM7?S+yC}W0MOE{wJTl|4e*@Kk`)$Paks}YWO`_zwa=YFM&*i4TZ)PuVb
zR1Dt*LIu9y$H-dtN!4vp%7IT7IeTJ0&8-kHqD2Wh@OdSkHG3VjzW+v`PxSR~ZNL#Q
z(BfgBVlL_)Z<78&sCCW3KuxCMp_Pz3RG^m0&|Xj$T?&~3A%$6|Nh?ffe-WQp3Z!s`
zMi6|lYuupC`%<?v+$4H3Wud%?bJoI9=1|c#YKj7Gei4HNm{_!p!dZln0SaCQqy?s>
z&?je(qSig1&ihZYRp+jH`Y4-`{mL=i!vYlEH_gJ=ea{<7EE;)&C|D?sJ+3O1n_qCf
z3AOZ@JP8O|DXW$VeWzm{*>g7cj!dt00?E#vxG+PshA_dBsYXCKjl9gk-<i^I(R=ko
z-=>#U$|SY6(59E<wfm!mlH-;Sp#<X);KjMpAucFnzbrBA@iTX>K}K4tj1n~jBbALF
zMoM@6TpV9-T`x6c9aA`(`oK#Gi}!xqzDiU%TaKvg7@?gcYe)IsRdw~VdLTB$BmC|>
z82a-IuY>L^^|LOfP;J-GtSm9inv{})9QMzr=S1^=Y!DnLzrB{Q<v>+JJ*sRM6>b$X
zaww+BU2R62_|tAtAi9cs5D-Yw%Iy<Upc^27d~)GM0Nx))YJnc!UR0s7frl6kiIkbM
zOU`9R&ZLCi9L1&-l>XrM;Z69U)_3j*n`!>I+fMb~oZ`K`_pz32QsvcO$=w+~|A69o
zc^r&aEp{36rag&_?`4gc6B<^+jEfU>rU=CC%Q~<QPXq~6zumGtzguqgIyy*9-!Duw
zjRYomn+Z0-YM63=($W#{cdjDG@eySV*~oF4Yq#dq&Hp;5vExP#vdO9*b$@~?_OhQK
z`S5q+O8D9qyX8StMx90wbZGZwB|^s}&#q+QNnNzQ#<idB@pj1jvf8NxABq#ZAGxUN
zvq~>{Iua_F^u>;BZfj+ik~Zc2f-fCb7jOIV$l!j9KE5TOdRV6^-27z0-yv5M=IEM{
zi<YYV-c3fhWB3io^<E;)XsQ0BX$W&7=mgnN$M>PzG19q1)V$GRMWJDT;)K)5(zx*}
zP0Ri6k&4{c1bs}*5prtK_b@t#0YvGAj;IEq6{qj+ZKwhu*aObu!$CP85-7aMPf(1A
zTx_b*z=pl~m2xKv-{xYg0FNmfEW@hHPQCou3w-<Y-NHy70!UM2buJ-zW$k7hsRykT
z<>dNGzqY5&exD;PAO7NFlz7BCaOY?ycw0&k@4&g7j25N{y@(JlmGmrzi@5}32~ym?
z0HSp|Wr0C`*;$UZoy+$GlB9#(YTVoukM749*QC(JI9APQ{%?rLavX6k;n;a4u%Ozi
zpf?giBTYmVGe%$KZm<VSZqV1*)ht9nZ51vX(Xp?RUCrb0kE(x6sTvA4o-2Jq2H!sc
zcT+3VQlM9qCHRB|2jZ-RES~cq)%|sBixuQ!<X?f(?nnOaz+nFeHJqn9iI3d#9?!$=
zQu65V*sHb5UAO9S(>6+W;>Jp8bsON((x~$V*=K#_i)KTwj#Ax5Rt~y_$C!@Vfv;gj
zcs>%Z!)HnM%0A-~bpdzji>kx#^iD*mn-FP?gogX<AsT6v?6rXVaN0&HquIV)+JXIq
zPt^iLH^EpgS;N&&)Sh9B<dy?+rS;%8;P%^X<*#xDq4j5wR?)R8tKVWk62E4V%n-I2
zvt(FIyh_H_RBGV3d%1hOU)lc*s3Uy<!4Ae2F||S(joW*$d?3W`keXnfTPe+0T&j6<
z4Uc+33uw<!PSTQcT>Ex(e6rani)zB*kbTqAo~(aht<+<Y2j2}2{ZK2RdX7`(upkBO
za%Enpas*qsH>l3Wtol?%J`0i9;wNcan5ZTB5JG&pv*3vchmlS5T!q|A1R`3gMUjvT
z2=?A+6f0~cgb?iie&nJK`nRgj7CV$4hOea~@2W0e^n7Casr~ZNi&fyzubN>sj}egq
z;m_5sU0X*!-+SP+lkS=Kp_Q=quA}V0UX4$dj{tSQ4W0`zE<;?fuxL0L-|;$axRn5Q
zy4bk#GV0hld#W@kmxrb4q%2VsH`@yXJsf4d0%4v1aZA?lkb~47m`ni4>1{QBm?1Z>
zw70Ox?UiW6N^oR?r}~sK&0&W8rTdIz0uElS2N9@MKA0i!B$wV@Dd3oz3U{cCWO+52
zbyD9Ss5Bj<BDb%y6Qxs%G;TFCNZ6J`9tmNu9C$80@XC$@EZ`rA(%xuaQi%S}>$!HF
z<!PY2U38r7h2M@Mt9fRA)WbWry9nN4jJ!6Xu(4j<-e*J+kfbVwI~gi%{%A}j`5TS~
zS(ZOKn((CdN|+houpeyU%mb9Z$M-3W4qv*7d!iN%#)~7ut36DFCR94PPLWBS^Yf|!
zScK`fTNw6U)4{<N+p0xMv?awh@87_xc?N8yYpcOg``Qi_4Rp~hxCg^Dc3s4C<gnyN
z4U9-W-q?%j^;%=63X}5DNLKy3bzBe(5y?6)W2S*mG`;cnyT?rssEY=6J!#Mm+7J4C
z1UeCwqitL%i?VLccbaA_VT5_-cMSXL>t1v4Vb!+fFCt?QmZ=r}VsBzxxO`dh19n@5
zhNGlroKw}rMrJ03F#KYI{i(X8UeEzDZ$3hR=bI(eoZxU5bkF(kR~TSJGBC13bVhDG
zpVG=Vf7AMnONK#4$JSrWuoZZ0-)Nc7SSoV@`CIut0h$iw$YY6vdUXR<kqs(xBs-w~
zfm_e73wJP-7Nd@O&~zEicqdQJF#^MwF}tWJJzw$>W$nRp`ATyr9xbd5gm9%)>^K*w
zmMrP$E_LQjfX>WZM01I9kaV`l=>b!63#%95i6Ct`YhLcjzpnxq;@YLl`%RaV`bux_
zE<^v!5rP#5^Fw9+Do+~$d^)dpxIX67T39H^WC<0i+r02@1gKGOk2$cg85@SuJb!wS
zfTcNZY^q&V%@A6%9`~df_}Q=<%WfB5blzUX7jYlN@<lT}hkTE4goQ8r)}@L%#@E_d
zWRhpSQ#Cw0d)M%C*A-^cehzCDUYoEfAKn9e5gnsuu0Gp$!kzlclD)n=6l_~M`21J}
zhQJ<{wma_tQO|IbkSu@#u`jTF!{i@m+Cf;y0sAkQTs0s2d2{-{V|g2Dxy}zh=Ni$_
z$emTo-+95_aKYiFV>pkA19tF(sPHQSkkJT;6}7;3!~*`Lm?|bEp)?YzK}{+K^}YMC
zaIg4!Gp*J*d$;kd!U_lEm#ls*ATIsenA9y_gU`wg9Ak1Qjgc+b>4ug5o$Tc}^zo`;
z<%}F(i4K;Uk+p+Q(SxPFMFo2Q)=a#t^fUxY#tMLMMThj8rz18{e#V1=D{B@WaNBR@
zHa?fn!cl)qK%0P$|4HLRo0HAIXU^)j*}$x|<N2||0@H4DO_jH_s8mpF;Z5d`9M6FH
z)hXG$esuzqWG+wZc4t&BYeWa>1VV{cRid1JCQ~=VI>DJsWJo~%@>&UvX7-g^fgqUa
zTo7E|*dPvGUx)Wt?JV1nzTYdU#fe(CI6r*50W-t;TKCzfb@vJn`#4p#2vvIh02LRk
zYW%T%Y|>T3!fQ;z>-W3$%da!OJ#qH^iOXf7Gf`yC4VwZNHY=$=f7FwTB~icCyp~iV
z34Q4%dQmFghJntK^#t*Tp1Ut``%3VDbBAwbR+zprzzBcbr<07iQuz>#xzFNguA_jy
zE@_UIQ^Cayl>NdIH6U6#O+wc?tjF8qA!dTX6m4LnHCID8iS;M#Iv-E^8t}bwFthei
z3r_-W`F*|im6o}jt4arKcLPt+ag-QgD?*6Nbg0>4Qp8=BaI43|M5K9EH>ZY!l-Gu!
zWsDXcSH;d;Trj<L1XRPv8a9y_HWp6v=YpC5c6o&kPQJye&NugM902LFbg1~Ol@+~h
z`5RB!k_%B^Xp*D?i%K{)(8aPxMFCBesw&g>(f#t{tE`~G=98$rWz0OMQ*y|5Gqm#=
zWU00JPEVqbuNBpan(Cf(Y{Y|xeAOa4H|9z6BnvBiOo-a<<vJQXaU}ykMj5NbDbF6B
ze{zs9dW7nnCmz``w0$ISQa;CViVoX{9DNJ9<0b(*?Z)(x#2=P|jpt*Jxd1_zzbE7C
z-N-h{>poCQr`g54I9be-5Cmj-kn3Hq4sAxu7H_-&_FwA=DLl-9`X!z40dL17v!~p)
zil`G<FFTH+4x;*`^LBkqBXT5}>E1B!?eWc`$Z+p1Yf$NZd5cy4F_#3HeI8HS38AfF
z{(*BMShRw3%y1BBHc|5NL$QuDaAbvHabYa)S1Sqvp@(Tsq&O%;)ZWX69@CW-)xRI_
zevxyF)?ULr*MqQH@KJ{)Mm0NBqW>%=HQ1PV_jfbMPdBklA5$$nY*7qLp~A2;sbsu5
zl`pnLeqH_TV15wrW&Qe<u+-_PU~K`Mo7`$14<f~xrfkGZ+NavG39}X4YFd)Z=c%;F
zvubo<cRpphntyMC0S_*ijHdnrkZqX6icj5Mrn)DwwM06y3wLNaU);0z5kp-xfuYew
zyzkkJflORa-Wd%y?RiDX83hZnI+V|xq@%+>p5Y)y&HxgX_1_5lxkHx$WdqSzEih5j
zM>+-yHm{=vF;iay9I*O%%h9MO%Z?SuD(i5(;{<96cVjsU$SW5vi{et}ICIl<M%umW
zZ#v`>k)+ziU}ra>x9@T(&d|x2=13!CuiuSx>n4%Bn(|o_EaH});Ss#b%DI(nq$KLf
z4sg|xoo=zjec@gFfJ2EGOtK@R3iUa!bUh2|nO90Xa5HC*2Nc@0rmeGnw|dxKqh(}V
zTlbVt#d34Mn~<gO6C`|oG&>1Wj{f;Yh6ed6hFLi!f88N1tMmuxcJa$m8h?xXf)W4G
zew7d4)Dk}rbr;GNMLFNyI`+HkdEuRBbaj+Sv*44fq=6<j%ap!&k)lDfOSjsHORev-
zdgcA5>}YD3UFwaHw*3rK>)p2VC7Z>a|2kFzg3R;C8I@HAc(lJT@WdNwunSMPq@0`>
zVN$|PA!Q;$rp9kwDHG;9^^M_L!|hAj;I}6&lyLb!G~p<C(mE?&4GAb)v*Xv)c1Hsw
zJ#2@C$$xqQY^o5EGFy?4Zj|%!_)O)+>{PDQ)IF)20cupP%{I?QsmE+>sg!UBmY~p4
zQovK=wc&QTuKKB{P?`^x@p^v1oTws1ZCVbSl3k83ODuM}2;h@|v~htP@4>@sg6i+@
zH?!7}%9emy^89e>-OrgrkRoLoLArs4Osz11gXVQbx4-3u!9ewDM*BiVXW7jCKgF+$
z1G6mvr1^1bT7mj<UI^g3vhDE_pb}o)Qv*utZDW@XU8S!)ud|p@6Cm2fkQsYd@m79`
zG+bvwHoO5XA>_+L`ss8ox{i0;W=p072BH-yGM_0*sj0h^?Zd>qOYQFI#<&|2;)V%X
zCeMwIWY5i9#haRxdn(MZu86z<RcM$2J7UAJgQ#fdI5*}jgu;i;K*UZ2H4R(1`Q)i*
z^~q*Ic!d79-Z*{-mimKH86$lJcF++`XjM6)`%*Tvs0E<M@PuF^t$ylNd03v5`w{G7
zy>y&%w8VSqB|Kyd)lbYCA`Q?54<fKPNP(OOnoRfeBA1785`EQysy2N-jmgqHH!s;}
zh<OgSL<3v7Lp<yBdgS-(h{B*){+~^P|9&mrNNb|roU^qdA6DlJTS(@cpduBsl{hY^
zpq2{U8%o1fvqzvtC1evD?RK*l<Z=ebqRY}CU&K6V%bWI}6F9fT^ehz@4f=Qg{)Z^q
zRN3)8mFowp@`kxI`azYEwy4L)`%~G|lLb{jhDMrmEFH83I@kDbI(?tqgc<2R5nS3?
z2%Q%_yu<1KqZWl2iEn@X6^?o3+4Z!7{*P`Ea4LkD%2$mKnK0v^ZB5X+0N}c<><9k_
zEzC3{^7z51laDzx_YMGg{wNN{qa0mGypz%^-RCVwwH9?!^b354-ONri(;{d22ph*f
zdtJ{$ZFCl|npS3K{E{W6cO;wLJ`h{2+JXOoj5*+k$sh+7D&*_@J*r~8Br9I?bPTfD
zbxf>f(F*py@;x{Le8HCoRIQJ#KR|YMt>-^id?D{Ug<K8srv1bN=2)<un>t9g+JRPw
zyzAleI8WDU7*o$pjrRflG`Ok#UP4^hV)y0uR(GvHk`fK&+=Z-*`p-m!LkL`TaeX2e
z5Ql-T)w{#2$SQ;jVsn*01Q(f<vDX4Aq@%JLxLKmn&?FAUqXRb!@>i>x9`-|$9&%)I
zHGU}6+nC#AcqfK$(8DRL5bqInuPg(Wsf^|}$A5*0`RWv9RGZqKR|UG-OHdOocsd;1
z1VJlt4gh0@$R>dGvCQaqCux_Dh66o#;t|;1l5OgcU=9g{=3!4XU6db_`J5s3>|Lps
z?j-{Xlw|m$SD?vBpBK^TC$Vuo(uO1q$Ub5Xf}4M1rTXzIwef@1otywRQvG0fwGrUk
zrgU5m-Tj?*+XrBtg;u2tt?05T_LJQRAt+;*5>0KAZhR$~qj_HG8J8GivL7;uA&tIX
zhDLA5PY|o=N=0odY%+;jGkUWAJOST4^V&f%rh996SaFMh17AhTNoMOj)O%j%<LoDK
z_WHH+uQ(VN{8x6Yh7ULX6iR%h(@$OZNb=@J8wxO-FM#`hQjjRGq}W^YI^t||^fUP|
ziAb1ZKk2&6YHZ_L|DKB2@CPVx<@^|VEV$iARb8L0aR;79ok`0oB14lvrwPCmF*eF~
zpfVFB?f&;@av|FqA!h4Sl79VJa(DK9UobG>p=1CBZv(7!bJ-k%K$yImFSz$-bhX=!
zMVqGjJM?F;sM-WLS<v@Q)V^*(Q@ZD!vl3j*5)N$}p&d~u5z3)xm&c*RAx+&cu=5<^
za=l*_cq!|0t3iS%aYe@C^kBBa>E67@Be=XxAfQ3u?vHT$%yQC2wJ?B%<rW<?2gWdi
z(ml!au{+7ZX5Tmse_ZM8g%hl8f4JLcLD}b>!BNjGO1iD1%gxd-ylZgIm2XXt8H$h6
zq<k{?Kw_kIt{fwi7<7I9;km<4%qw9PZ+Z(W%n}{ZbY}OOBzMckO7O49iEV^(95`8=
zRB2c7g^qiVG|&7`f#t)gtz8~qqpDoIzyzEC&mufajV=qstZ~*TMWRqrOtjFDR*`i%
zoz`iauo_AriSd9%cPIs1jy~Xos4}CPo#o~u{xDAC9&$wXRO#@RO%@R27@VJK>S0*<
zOfZJE<jhdX3=2D2%Uk+h!uHd9+Y9gPndGGus3l^7QB!ue{;D-PQ8bxE6ote>iI~+)
ztE%!XMVW9BXmBa@B7)0}Lv~n;?_=(RSeuZFBe^wc<`hoA-_8{GxNmsCv|I|O>eSWG
z?uM9h=sO4keI~1UFmWSbw_MXUiT_l3zgQ{{nYt6uRDCI;Z{j!8C?0h=s;=`v?S#67
z4l&OkEt~z;b4R5@E(jqzLo$uT9=$MUx2o=J$F0-!`A`t=7imTOIx%q@{7-k;@7^|Y
zg1(vfc_yF>9{rK5YrYN(Qer7x^S*n}yGc~^yIK=HtS?m>#PVkzl0mSW?XKDVvQ=NC
zE*GtiM^&hI@0O!ykeT#dexmb3If{a|XY%YplcY)V#UO4^iAgC+yP3(+VP7GX?Ljt7
znjeQPmA%#YkYAs$ME5Agx*wGJaw#Ksz<zRcl(F&fW|bRNkiJ<DtoC)?mADjg+j%=#
zZ$K&D#>nBdTG7TV!6K-)x6&rC+VlJ{Q2kRv4v#8lLfmM`%FBQ@lsvo+9zIQei9!6<
zFU}~=2&PasI}^5aksC^hduA`YB=9Ljeo{?iCwsdxNd2d{5zByHYg@C#LPxV_XT2DR
zvK`ax3QY&sdT9n!IZ%(?zrtC_PDu#c{7_4S4*y(git6_J6|!ie!G+!$K}U?yJi7k~
z=+RRTX+{0WO4;*GC>wtdi6s`1I&i_GY}M`V9l_~J5m&09^pZv34W}%WL4ARWEc?Fx
z;?id6YJB1|(0blRb)fG~5kMvyRV<MW8YV`h3oTpv@+JHzoovF#H6L7ijaJT&GK^MT
zLaCqmV$PC)!V!=y4ibdUVV6?jt~Vl_xt#HSm=!|4^J$pcya_@-@!ApbKoArVwMnwC
z7c~AnP1MCWK`nVtF60U7ceW<-Xf-=FY5`;vXNqqa-qhUj+^^6veO#?HN4b5FLTV}_
zEQ$7jD<C+y{3`|2(hIP(J^#ZeZe8h=Z6mWGm@cx0*-bYl(%($}88}%(Y|f|G->p4x
zv@X(59k|lLipP<vy+CTj9V%XEb#}iFDj@*>x<juGD5}nvFbu`Zd^wF*6x&Q4RVj0(
z3-DwGhHzl&KumuCZ^626A?NJWW7QuNuM8daSF&1ajalr{z0Nv^I*NQ))Bp!q9_eYd
zA8rtA?tRypU+{<MO1(%ufZ@a*MktX6Ptg3wdlHv2T`3u27JikLHgQ&}R&o7mV1FGG
zaOwG!hj_W|b~wIU=*csrsR6z>p&;yD4fZ?hA?);@Gs~o~>H~mgSN>)rRfGMIxaW*;
zdE~<C4MjzHEoB&}ZyZ?KR{bqL-_~vSjcZ9kO0|U|E}$9pa3Ryz_ikEwkReQkV(x7Z
z<JAQZms=Ks1#Ne%pyhS&3T1%qxRwT<OrzMsi@>PIW?rrlJ!FY3(-P*&kcNw_i;Ig`
zC~1hkJLNp@N~FNyNb^KB-vk}62OuKf0yN3wZ<G02CtS8jQQtX4oR7JLfZ>+TT$F{6
z$7I#B&nVd5@~6P4|IA)jW6D==*^+1;ZCmVyN)R{dnni@Yb69UzBUBOp0@=Jn@LBUh
zD3w}BF~)pz*N}9-#}h=m1vxmZm=~841PT}=_$Xs>N}%vR5*og$(D+pp02~8RP0yQ$
zeO9KCuKuQEta5~pT(@X@7L)z)-4xG$4K6cLNMDm7X!2&2iVl#33Fz9Q@?k8!)@wAx
z4T;3GTM;+~c;a4{b<4q|qFnO{E9)%?qd_b9fcw#tr2R)hV1pn=l*5z>2X3oj$V`AB
z-|`OopOd{KT@8pt)vIjAk)OJlI;ezHs}ACrROf!(vM@pNR;J~Qj_<}9SlQNZ7h`m@
zw6$cjQ%;L|0?t<Xzs#ey&dI7dJ43Od$Njey+O{Lz6VOzoN!<k$b45YL*#dhNfsbWr
zq!vhEi71X0oeK;tVSflpYmMv|u3ejJB*I0gnjZqdKWnDh&A(1FyY$H%g4(;RwlX3N
zg^KW$;H%GGj8%Qa##VQEuesaYUQ(`)1nFj;+Tk)RxFRaFz#qQgxN@Z!)6s~zSmM3n
zoB+jc@Qb}U0;~7@;#1J+H%z0!9^SY7y&$f{WM9HKZsBU4yf07%-7hrTDY}Y4au0o9
zr!x?duvhbYT0y(l@cwz*2SSCmK{z+`v8?aIgrxMDq6*c+N-u=6(K;<y^-WmJxJ$@y
z12ouW7kQJ+u(gi$^s{$%T5d&eo&Ssq^t{6hzKIfiGh2B6wyvvs^-I7AmkxBZ<5LlW
z*xSpOC$Q*X`)u^@NfhqJn(A!W3g*_BwB9v?7C2VaF@z)RyLSt7+a~)Cvy(*Ftb{uY
z=g132H41}X3wNT%WQy3ISKA;AOB3f2?rbMgcG|&b-XVM5erGR$j{jM3RlGyM;CMZ5
zzqcQW4Gi4*DRio0E>j^rtYmFs#Q!b!vh1}tRkYpsux+5(_2YbxGtq~*zjkehp3pcU
zm;CJQ4W6$WJzjSt3JN&}GQn<c%}4W9>v=S&q8B1?EX5>l(mu9kB1@z`zdR2zTn9OF
zxBhTh!dRPI>yJw6@1jC<oWpYpZa>@Hf3eYsU`)4|7@g-u(wOvC{q@(D@lP_~)~0cb
z&?)8Np_uTlnQ#JDzdR*$vH5_Naa^p#>A&hI{wWl)B!~ZX689~Rxfi#YRN*YVt#z80
z4yJm1;w!hH@K-AOKcJP78o=R$;jKn9olFVoQpNfu0r*nP#qbB`RJ`31woU}pz!wy|
zaJTPu&=9QBFUtEq;xE<5Kc$Fr89Yxda6pD>^MYn=uBBrba4WKlKduB_^d$vB0dKmz
zFFswr>;6{7{MY*cs>auFz*RuzpK9-k#lRg{@s#*wH9yQ#gyZv3e(D8~H5k>IYYlWn
zEmG9}TVbGFhRBnq$m=)uJ9XgU7;^@n*H4m(*#;v%>E!>Nw3Gw)O(fJ+ZM|ps1aI9(
zaB%J21))mo=YKo6;_$Y9Pnt4LPC5pGYS!nyuiiFCLX+q8-zgr`dq70N=whQc$FKj=
z9QiM$n)n^PNg(ve8R1(OuQL}gq<_$4C+WKxFK67OR(o>W|GoMzQl@R<`d#l+fPE~O
zBw;Bi{e0>v*6ny{6$llwXkqfQxN`Z)0LV>cQ@~9W|93z1zXQ=PmliM=nvSNTex(P<
zj!z4gn5>KFc&HfHzJ(uiJt=qrPtp(g3;+EW|BTiD{T)BKr`V1Du4Nwqt(aqary>`*
zpZ)6c8eQy;LhwaoHnQx_7G^BJ%jbaM?XUlKo@htDp(PQE={-3Vya8_gQd?)>lfP!v
zYfSA>N%0&Y1+IhquR;AkUx;hq0vbDzOzsAKt#@H<hF9h>o>cJNzYj0UFTgyjS2)v3
z+Z@m$R@v+<4${M_&LfUJ|9?DmETSjYD{4TsUlIx3!Ug;9A`s~4$ZM<1cOCFTj`17`
z{;s9^hcu8^Iin^^YHVzrM7S?GnYf^^(E9iW%#fR#EAOHA-8ml+&4ZQd$_UW#iR=<;
z(svIc%eLN0n%TgrsLkFd|C<#}ECndV1Af|0Gf4cyNoIq$(=QUrd-j?i`(K8N-xRGl
z0L|_+mKi;%ypu&sN>G41$x?;ieXfzRKf%j5#uMX8|65l6&s8}B_vF<9dwYS-o1jj^
z$OLie5|1EbWc>ED9_-#;0oX6i5bqM+znog}1;F5-D`G$(Q~ObA#?`tcaB0=2I!sin
z`pJFqBUzbE<3T*3Z+Q;{;<el(nBm`t{*H+`rZ`o3Rin?ki1Z5Uu)f;AL&YX$8}zS(
zVGbXz>{#PRn0S^NcX15zVVPRykHS9d+3UfH2~9?JW@o`&F%xm!?9j<~cxCx;tMpVX
z<8?juk(nw&KPd`^ezsQSs;lG0A*!oH9L}m6vZYmyrLOa%+~HoR$l0}%qe}4eajTQm
z*1yl!+JBt8`dl-}?CD3Dun|r*VPUCwnPApPIK<34kQl#DD!UMNz~WY9c{MZQsCMnh
zwe>eC%s+zYAReHEe<gaWgaJ@a@YKUU(dBdr3D?*o!v`T5&@$9I4~TeatWLU_#)$3_
zgOqM2g*rF}5XwGEujYvs3M>E1H2TvBJ5=ksU@GXkD{#cAvH8!EI$)D<@@DS0l8m$V
zrcGh(N2QBz@oB6(XuYhQYm}>&&U*DO@+zQIEx>VLL+zOQU#}zEH1T49Zi7(JxzZ(`
z%o+D0vO{H0c$FvR6ku&PwL$CwiNT@XE59BRr)TitXZs2eIvxfc_2LOf%$>6G?PBKF
zv4{0~_+Vm;*zLdZmH#t$xRAmF9RjeM(rQnX{MixnM7)K-y!tYdfN8P~zI2cQolu^J
zGl(6to?`yw>@=B<GiTc5-Af)YXO}$kzl>49+l3E^($9QR4ZU_%y1-Ux4m=O1s=+nK
zw#GFF?e)L*+xWSN4J@a7wI^Q;YsuI+Wfsr=gZ0;`GH7(&v6%%+Q?@MeO+(*fs|P(M
zk`dSN!ML;n9BS$sadZNVYPdX(A50~)GJ+FwIVW1H)wPU%_!i}40cw966Vfb?d=CnT
zMXq=6rm#{)p9>R7MV!h;_5zl)jAi)?hv~BZ6><u8E&lohT3Robw|rXm)Bc`OGICIX
z<;*xnwk>!tbrOJdUVP5oE2p5=&cL|E0N<WjSN!D5c|nRU&Jh49_~8FBJ6BHIJUP4T
zT+up55-O|qu0Fo#b*+dnmcz~zBLjUhrd3rzhAxuwfbG8+UU6B<5uhm92v9W(d@yNZ
zq~YPw>g(@IWxK<AITCMFStf%kexH^-7JHev%c_9{-4Ew}r=Ax@_G?kYNMn6pSsg_+
z?12ESUTMFEZj&I^Skv@q&6Ng~mX7{iUh4CA5e}6vd$bn0{J5Cg&lLM6?&FL#3+)nA
z8)cs7^r|S0dimR+!EGmfH^!{TzTN$ncpyk{#!D*^{D4G)nzAI$T!kqeno+xGP5|y>
zgt?ENi2OBBstn*GfHwCyPs4k_++|Vp0U5Ii=*CJX%dj>2S*>Sf1Ktu%#m{>cRX~d@
zgq;`)ec77b2#TgcW9sU(>_Rk>w2~<$OtAREXI+-0aG|dNeHP8jP9rT<!@^SX+T`%d
zdc7SSYPFm&CIihn%kZCLbwXc-NlxXX*aVbs#YtITGbXH!o&FByI4lpMKouBjVjcIe
zL3^*qWW8l>E_JS(J<U3?mm)6T=>lr<`y_{q8)Ez_h$0}g?61pr<nSrK3q@gH1dib)
zm&J$FYt*?q#Vrbk3ItD|&RM@3Z;%12sTM>+D{u=22dsa}txvOG671GFWrZ{U5#Gnb
z2Md0(u(syA8q9h>kPD;k`HC~asNN1bLQD*rm-dUd$EJUY$q=1i%FM6ZbjTbk;sBix
z&|H}QqLHJZ{4G1TPK*{C!b^aUl)inKbF-XdwJR7tE@c)(v0y*F5WSvPl_gn2DAO=x
zN&p^6k~FnUK66$0P~Ile{l9CO|54-od5H-8y6PVJ?6)R~o~T5*9<Y5NG_aVZR}pZC
zPcdrdF!prow^4-xwOkz|!^BqMd=E)^Zq05s<SJjx3`M<}YLn#0Dxx6sv<sF_s0hg(
z&bos%suOlZ(@x(?zF*te92}974w<pMQ<~G9?kdWQDz<tmDn?`J9$W&7W*mTE`QZK{
zw1yo<%WrkIV71qT_*Vw}lVjFXa^^T_F2e^mNeMpOZOLCuy>tSejbmK7jAy4<aeFht
zAp!q;t(vhK5$X^so`dvf#|6S%^DqJ2Wo9W1n4_m!JA$t%mWYOuoXqtXldMOpu5sCY
z9h=>6R9EPTD7A4+T{Zl<{h%>?$5;BO125hz)bTVx`t6oU=iPI$iJPsEQq-?!vm^7-
zp;ey&vKf`EL1e%O8@2LW_VBH1v_S!G-m@3ugWtU(lHEYguPWh`g|7^i9dymDl4W4K
z&!jissGwH?R=^hK2GAtqTj)A@r!EPPG78?`uNe1peGAkzjUJk|<{1Yv|KUV}V=X6q
z7_3>s9GCX}a3YFvJm|y;isV+z<~HUjzKY~HY$i!uku}C?!7dfWLJ+5}5lf1K;HteR
z?<OsspixVwpGG|4;g=Czs2odcgc$PsOPjxI_1p;)QxO>hs`ZjP!gf0kt^#EQ*{_)R
zq>Tx`udXkwd)m-ns$AzRb>5`#Eb4lM@BC^im>9Tdwz`g1A0j+d6uKk0a4M|f2zVL$
zV^~OkMQ`o<oL_^0d+zM39k#qEtAK~*)Ufjg@Ni$MS?k)Y#D2&4QoJ!KdD$ZU`_-Q8
z`Oc<6b~c;=dpBDCtN7CCUrnH`)E_y4TV!)tzpxK6wmXovLs<@$y_Y}ONesMGQgTHk
z<4UB0nH(nUZNq`898|uib4UCc87>bQmJOa_bNW>{EP}K9r^exzUe5Z&4Z$$~;Egxk
z|A_tn*Rt@tlmQ0W%geMxY{``wCQhqYTTi)#w5zLXHtHF|R`1t=MVK~i&lb;ZwDD`L
zS#RU=5)<cHBmv9eykC$<D3R{OWy8mP*USTbzy-uXiq;E<oDCmUxx5Q-cZvM@CCDXP
z6GA9As9gZ^FvUcrT@>1`BiwXkOYaBR3-EU)EPiw1fPhZwO^D@}lN7Izy7IT%3;9M_
zQ+U=GRQvJP5)2;FC<_`<Y=YlxB@IVBmhP9pO+D*A6lKSFCB6d!s4hz`-lIcsUUjRx
zDOlJ}Zzuu+)sJ|>Ex+Zi#R>Ef+P9171WhXrZ+Cu2AkRei`62eR$NkmcbDi2#uDfS2
zpwkz7N4>{z6H%O1qpyy)+ZS?OZ8hW*1Zd!O^{IjLOY#|~<JCdR?%>`v+E?ZvN4NLO
zXgkTS&v}eiOy5ApF9mchhsL|DPzgs(@4vxM`*Yt)+7j_c?YH^F7q%gUAR>}Arazza
z&=~swx9o`V;Y3(G<8?{CLn<LZ5oD{5cj)ah|FkF;WXoZiQbf1d6!RB>{23u5{Ra+k
zK-8ZgXaB4abV0B%-xg6YnZQfu26TAr#ioK)#nzf*1RNCGAAt6b*WsIsv()SJmxa=1
zx4dO5U&UO-X$wZG0+a{<R!-Xm;#Dl)$~Qkel&f%F2f0~454aOt(>!AnQIBpFVXwN<
zLY^oRDXIY;A&2K|MbAZohp`v54bbH~rjMr}l1#EyA$Vx(fhOp@moQ8pQBwtG%c!8g
z>NIEptF<mBm<bu?C&H#tQXVjFTx?E5MMu5nBOF3g;>Ml!(|NTfTUXr~=<l_c$X2cW
zq@`cYt0(fN>S!BW!DI=1tNFzwfLj||W5wk7mmmT7GZh^>GnXk9DaJB3#lTv_yV*?w
zi2?t5pNNc;NrtAGNUBgJYy}79COD3E_5@<YXeEv*BhKmlLOa?r%3w1&EflprGAg*i
zsCrh|%(goclJfxI8*}fD>GYhZ=2sN(HV3)*VAxCtK@hj|uNdS@%WX{Ghv+WK=ueGZ
z)S(~#R1VR_QyNjZ=tD6>&9UPsnds>0`Qc+;1tjv#%|x<u)uawQC#cvZwYNkG`M%Rr
zwa)z4q5<$SDn}QSxSA*JP$?vS-PbSqNe%G(im`9L|3_uNhuOB>Cv$zXE;QVCeTa}s
zi1xlm#nGzX1gE8*THlkKCG{=0gSzStwY%FebsgU%Nf#AM@w_rp_`HfcAV+~M62(kQ
zLzM*ujZm>7FufK&u%{9+zMZ(VlH0JlebbuE>Of+vg`y}7w;c+F<VmHe=IgCT7#hJs
z_}3yRfkK*b=wjXe_8c9@+}8^a3`>vG8S$ej4vDK7zPEWc(GE{ze~l&$_WuFpM?=91
zRJ3>Rd4-!7%{EGgEzO}F$AnXwQz9MX+~q^K{5NPhTKtX<IEx+#H)ie1_bN*=Cf1PU
zpJ=3fpUnMb9SD7OwiUQ44b=5)Ne+ATGvd3<b;M1km<>zWkX3?)*LEH6tA4@rtNYHf
z5sTHTME{_Ya{$`ANqfiwZ?n*df5E|r{6Z!OeG?fh6%9dV>^faH46VJHHA8a(A)yIK
z0M{@+^V^eJpHK^gB7*WZ&2xlGx$8bxX-0C1wNC{?15kP-O_e_iN{gj~hnaZjx*hM6
z_EvaaOPfs_f_(H2DeFd8uFH*_b1AZs1An6R7-gcrlc3JC8rJ#i@!N9v$<l8-Z~D}X
zy<imgO?P)JV7jsMHD<O?-$}~VHajU0y5;&F5lt-81VJ`JYa8i8d1NWuw{l!r0J6KQ
zDt9@7D0v?{D9SjxV)^HA;(I%;j}hhOt5jlFN*+TjXbl*0%s;$i(Bsd+Lra(9V@pFh
zd-wBzPuZU2$V=!x)rTb%Cx{|Ny?7q&7WNhdDb(C|hR{76atweb3QvQL7b(Tsv6O4=
zJRG(sf|j>YWE=C3*Oit-hBd%y1qq@<=K;gPjYF{l9e@9be=HH_AWs5Py6+_{+%8oX
z%h7j!%#0E?u#cNA*1vZKU6ns~MA_dbHgIIVh%&R!pC<tI@38<xNB^$Sj3F{d^OPCW
z+7pA<J&xgxjWx0yUPAt3cA&9%n^#r^t{SrOr_N_H$`Rh1I=E_LJ(Efke5_P3#IG9o
z;4IN(WYTbz>l5eXn;gS41vDaB;0M0g?tK%p0ad<9ST!qiUr06?|Gp6%qE?}-qKJ*N
zskKuv4{=*|WAf)MinM9Je52l6t+1NcqG>OOf)t`qTn=2W!CT^t156YcV_~#(b=y<?
z4mf)Gg>GQ}5y?*rXkd2Dxc>$FU1cB_HOa%LkkT;kwO5zHPBYbZQ2yRXAs=O81Xb~Z
zP&JWZ4wSSzsMMpPD*^d6@v-B^!ILh359N@nNz$-OxO1R)NYPnJwu%d9<Jzwib9niM
zn8XAC+#HZ^p9Kr*rhC{MBLH%Gbe!MJ6LSau%H-s$IQME^MWGCC9QV9RTLZCnQ~5XT
z63|`>Ctz}y)5N0L-)OKbs$OC<^eIPcjFsW7%SCy6B2#{0^uov5QL3dO*zNKB_OyMK
zE5G@3lH8#vojUYFxQ)fAmWtT8965!)kAL=ZS-Z#Topa5taeGy#HFGo-q(Koh*&aQC
znsx@$L~_bkh9OmK7rXktf9zBx5qxM96SyJA>tqm$&J3SlC(h${!@ZeCw?c!|04KB-
z^+gN<q<2dK(dhXLf8v}qB`X42n8~O>EB~eGDb)vhp+Bb_N3wH8N(K;bGOR`JMT4-V
zBC<?<Cg}#|^R5GKdx?UT{j<IKOK0NMRQDi`)Nw{xBZXV()d)%*d@`~3>Z8j)>@>|i
z3*+Qj$vL<IOL}I^i?aQe)MR0Dj?v+6Jf)x2WLBexs2z0b=!6OJcVP0Yw0^U{htCo!
zpxg{)k&ukKTKXxtu+X6bY;>vpb|Z5~i#%@$e%zqNTkDIUN=d-=sK~oUX4BV!5E^~-
zY!fsKc;Rs{I`BJC^>fp>d|e25tx}QzJyyWwh~-Eb4!Gg44w~wD^eHwix0zh%AUxW?
z{2(lcfS{uKO~W&Ut7n8w&Rkpr{6=N3-?sujazEP+B1~5E41*}#O304tFYmZn9hSQt
zH)gyt+ax@k7h@a!w#j}qNFPlfr49fG{r<2fS^mHke90NeZk<JTq>UotS9b7F)NQF*
zs@#&R1Sr+Nw3)G0Zsqk!u+PzDxh1K8y1_`&KA5C&(Ipa!n0B`l`#tjoXH;LLN1(o#
zmY~=hq8PRj=2&Dju{yC<i)WcHl(`i5G)s=g^|qTA8E)tE*FJIQz4`^${#%qilwpfE
z{X^A*|IjA*?}?Kgv8h9)X+8K@W3$qZ*j~)lYyQM65PI8%dgpD&$L<FljtI++&6aa;
zz@n*qLM*hvwm~JU@Nt)`k*0Fdh>Qc<sR^fK=!qIB5ec;%LOJ3e^kP}Mo8(6;#6Z+R
zBwWLhF>GHG+I_sfFe}CDw<$!|&%}2=oWjoM{j=8MNL(Lv+lnFIU#k|BDi39+6&r;U
z(YSUz!(2lgWe2=WaD~^4PVgGS!sPXKm|k}jXjv=<h6X;y(9*S!=+#o3y^t({zS$<P
z=2xZ;1-;r~WO^p8GX2u2OyXIjk<`WesK4JCkim*)S$rH&UB9JxVID|fKUb-n#495g
zn)7&2t>la*wiFLKIp*mY4)<{G$OAq#Hrtdf+cBsuaCoi;M>ERYIaAP>36{nVP~{OP
zugj%h9l4X7in240JMztL6UQ*9*)+Hu6<=tat7sTa?*LXgWxzd_vF^JlnNae%sM;C~
zL()Gz!>jnTmY<nad!kepvokKqW>4^FkxiAQ8tjOvVS(nvyVn^qac<;E9;j5PYl>>`
zzSD7k|1oOJ!7w6b=ygpxD;eu!UbSsmyZ$%L3E*uh3m-uM*6%wR*Imt198b)VeP>K~
zzEAA5n9(xV@c;$uGx*%@kE~>hD=IMArW1=rID)!qVJoGz`3?cno43?}`&;a3aX}!3
z?989yj^Ky#A5@H^`)AGb`Rijm?96L|*l&MY+KnyRE)?!8Pb7@vk7AcD?IGuALbMea
zRez&lpTevMO-M%(12P<|6E*FVuokO^68K<y^gN2!?G^w~^LD&}PDq(tm<v_>cuD+c
z(bSq*lFj4rQsaSKtc{f?;kJg1x%R>0;#{;$Bc=E-r37&mGIqw{caa|mh;qc@p5jJ>
z38fc^p73xD(TJb(8{M?be+hKZNbVVIaOS0q4;+Yd@m*0CaSntV8XCep-VpmTyrMZO
zbUk@K+Nq~?i-Aa$FDVNyWgp&1&59c_I65N#_-Wqf?to|-|6^3RxC|vtG|q3-kBFS`
zul)Y^mu=xVGX!gSSd4EzBK3wtKJVq0uWkPs&s&3KJdjd`?>DR!7I~uv;)p#SK3*?F
zS_NCCGO7v0t}ATu=e=y5d{DzQyRha>?kA^JYzzi>tVyXp%qfT2_tC7@q#3;SVvbw0
zVCVrZ5~4?+&&!!3EQ%!k8i$do)$gSZG3={n%QWL5DpH@}W!u^v+P|R>2@&}%z@;^v
z2irTXVeV2%yDw)&rZSwdD{HV05xm7}z0dC`7_Pz^Diw?+Feu)=kuB(lIc1b}F0oG&
zIlW%YjOC9O;BaHLAe4Xwl`8mPtZ&}^?@#={#-ojm@+a`0C|E+U(R{}4UOWDp+qISR
zcy&@)F*)A~J9^A>tMMspSl(EO?%BTReNZhO?(!8mIWzQbT~~~`JiI-Xq%RttqMpDZ
z$fTt0k<?apVC$!MqeAS{p;DJb@#xqPj=M{}>eRO1S@2VQFfJ|W_=0Nk38zaUuCsJm
zwgxiK9pel3)Q7n=C##Ho?Jk->eSVWmAO^DFlE_{RZ)(_(-wi7n=}>ky8atGPP1wrL
zNWoyRn>gICnVM*)t%dxSK+^>Gu!J8Up2gfA-NQUS*|g1v$TG(D<jA%3|7F_$ISBhW
z@Xoa0$Vip2?#dLK8ZPkRr=ptvNQ3J8xPj`2_glk`Do17Av+Zizv5he$gQZbAL_{vy
zV_Y2yC;F-Qi?)Z~4T~&`vjpo}nM+GsYhsV6<5RD1h93@FJsz8jQ~g*_$(YUgix^yM
zAiWn<;9dF$iolO+(YI`+kt-%0sRRljW;#m#mOgF$;R^L_o!JW8I5YVBvOu{4Qz>DG
z+YrAZPSc&nCZE_z>Y~jKLhqHX-LyI()ze)0EeXy0@hhS7+H`LJcp9iyPrZ0CJ~~?7
z3_1>ET8g-Vdxpst1QD(6+gApH>C@buu~={6x4phX!YYse^egK)<z!(bff{yXW&led
zcd?*D;7Gx@58;YtJg;Unb$;&D*xMh;zo5b^+WQ=G%~!GZJGGQD)`f96>5ha7+;KZ4
zwbH$2(fV(Li|RjuOOdGCz|pXHi1lhTZk+|^OSorC{i~pBz1*>TwuZHBvHP-zL7})4
zvEEDd9<5W^+}tDbbcmPU;3-qwz~s9d(Jm}zHeC-jkl?E&Ge}LtK`*5sctK=0yWR<{
zgDbjK;oaTZzSQu(pVX@G58l$DLpH&=Y`ngo4I4$pZ3T7&GRei9=wDy1ag080?q=!I
z_yk&BD-9VH&ac^i-|~wlhMo&2<v<(P6<-Y{`+YHi2%%zzLUvy%$noJY{(ig|o@Pc2
z$m#AQLB$`VWLJF~O<rJV<YU`dP31Zf^Cg-In=blI2i~qcxAu=w>%y0lw)G*T*W=`k
z{O2Mf>|rDYN03?vlGcEZ7y*S?Oz+lZQOnsND@~nhwWVmENnVOyy;uL=y!6ii{m-mr
zBx4fUHC(`=@9yc+csvsZwa_bfo-^%MUi7w2;q;7q^_GvufSpxLERl!bsy92&jqlNm
z-eL=O!@dlG`hFve(ch?E>U*F{(bg8}>noaWoF`aAX<Y_4GCv$QvLkIGMk;MT3d@62
zI_`9h6<`_R#<#u+mEx2nQ;UtH^pCkAeO&E>+?;r4pFe5Q(S_a-=}F_2;+s~$PGOo@
zMPfoC?^94~&y%=ACMYr2?CF_8Jt*hh(qwySB}b8BzF-#Bv|#LXy1J&+dcMVqJ5lV@
zVK||RoZA&w(Q$L~F5Va>W@Pa2#0?LJeF(x1e17wsEE}0QP?7C9WO3T|N;?5b=loYF
zqhL}<UZN@0X78ox<RlX`DAoVx{6=yu{XBlkF#_Gq>npQ@qD0B-RU+qNelHIXf_}&v
zoWH`6X>8D(GI?Pnf7zI_B}w+zA&XJ8S2~~x+g;3$`ntTS{pZr!KPG4OCh<foW4Cqm
zCXuYn4*L^QFKBJqT_(QD_Q~3cjCeX9_Ns2YHjp*Vp;_!WKlAyQn~J!M$qUqd^b;<W
zVo`t0&zqM!uYX0=E%^F+uticF>4ZWBR>tIDNq*R3P`2UtiGo;Cdb~zXJy@4ijpBJo
zHdcjV;Nzgm=O_ZGOVrMur`-!?!kNAz`RFbeE{UOC)XxWN9cMbBA?dy*{cOE=NiR52
zvNm?tG|?J(tt99jL@^K*r?p5{ep@FroaPm`_ScwW6N&Z9m-Fk%6;@!98$iCYX3~hY
z##Mf%e37!Cx)$Q{pucNW=cA|L8L0N{vF?PFPL}t*yvM``#GTGQ;%bSS;B{i>&v|l<
zXk?)8CFiCF7{N~EUbUW9^`FQueF&+0wd=dr0>RN`%b$2=6*>~MGatO6f8BEbGqjU1
z;NON)m<N8p@>UX#A~ZO`>$Lg({SAi?N2_i630|eeJXi*n>fK;h4R6EbWq>1Q{+!t`
zDanGMcQnSJ-NF6$#)Y49B-hRGF0wY!&%#w{ZRkWvtjSo)*L}7Z>Y4JCwqYuQGhVKq
z>ZLLrSdB~kSkm;>dUTyNx}fb;?i{tn7m!Kf<s<E3kX4u%cX(53A3CMBlH=6#O~;H+
z)PX@8Q3uyaKN+IzF64bB!(4P3JHlwj{IKCV^Aabun4(2zG`1<Brd=;RB*adr`1`nN
z$@;#YCFUb&jpehYC5dID(0-$xTtfTEhvk@3B;O{(5nI782>(M;r$CLJJSXzi>gBs3
z0w>$N!pso@v6>1@e{@c?{XOmR5*yu%%+gBfl$|QoIWB`|>FJ>*<VO{(HqDOi8g2u2
zkd(52mo0pxOi)h{$OnwrYNTNMW*+;J>4$bL2U}rKqi1dgypq(0{j@FhzPI2M9hsLe
zZv;{OG^kj7$Ieo}d3}^z5?62O>#Ha<V;1;LwWg&*7SVifGWiXb`fuiQMjDaxWV~fU
z9ja8hBb;{mO^KnGspv-tkeP3tsru=172My-r)k4Jk6ge7EY^IexhFI*m4Izb;!yl#
zq8;Kaf<DOC1XMNzd1?zeB2VVtc7OApPAsxO^@+2+CKt3bOX641Gx~SS?h^H!4m@M=
zb<y`DNp7zWOtHLsejme50a<<Hs$y+>?G`d3YE<)`czQ(<_Pj08wgTMH!o=h7qVIlP
z<0jRf6W#~MWWHq?<D}BXu93G!ZDC_-ov_;&ywJ=0wP+jJb}X)w>;I@C|7W(Q!1Sx>
zib{>A+zu6TsoSnYqJLOspavhlHx4WM5f`N`T-ks0KJ|o4Nb~<;>@CBhY`b@H1Qbwd
z5D<_C5hSH!lx~nlN=mxB29OTv5QdNzhVJfe8HO4{hZq`%jsf;O&%1xe|Nmhh``zE~
zFZXe+>sse|u5+!`Y(vDs2YlGl>J?#c(OYUq(si%OOjj)xWQy$4$_Bf~1MfSS$$ec>
z8_vy|(}JV`L}_W8Y_?TSee+L{^pqT*@paaJI~8l}WfFiV<t9IHFL6md2)6}(&y|O)
zCmcjS$V(h5Q=oBh;PdDJ)esEpbJV?Z3%qp3aqU%J9yPSR&6maVv?bLdR;bBuoa_IB
zq5t#uHgfc&fXS-^yB%euZg!l<lj&7#!_L*Gy=P_pPpdnihLKr*^vdT`Y6)AY)Kj-&
z!e4o&F9xgY(q4f1>c(-4-v4MTc%D=7CbY*FXa-eptfE!O5$BXN&8a3zezx>7-f`{3
zE|07@+4KJl<p1Xg4#Y46qx1BZJP{f%o_CDXteMv;JcX1U{5?Ip`YHNwHO}o97+^?-
zh#aR2VpHYe;2>UXD9)|C66w4be_5;j67q+Y6hPerr=iomprPfh1!~jKzN#7+A>HS2
z*^CUs%^oYKol1m<6N~Pxz|Si`KNFH7u;Tx?SL`TeQ#XC`_1eq5IxOjW%l^NFDm{}%
z!h2Kr@H_OH;^)x`?pA}-(DQ{gDqsKw$ua%g=#1z*{6XaPUO|f}z%a#p_U;~8))MG5
zhPbK7Pm$1Mz{u=>%HQN<K|aGeTwb3s2%ein!2c*-WEdGzH7H6GO#c_;8(@UKul#v+
z-5ENn8BCRScHr9&4lRqH`!(R!&j<LC^^c&j^;)ZI->%V<r}YeMdk1b}E82WA5C?pi
zTJ$cjIsAk#5i(8g8zy>>@Caqdi(b{?ub4AD6q&f?{kLmM&oo0D637HxvoycexWwO!
zmyMycsQ992ruT-HNZQ%Ed9b+RY6(?;r(NbWpAwxnNof`RUy$a1j{gb+_kbEn%25}{
zW?}Ao1oy7@(M9fWKy-gOS%I6!_qhiT;l)=gd}<tnSYhbj$LOFD_^*Ne4|Tl22{C2_
zE+)k2us*oB>!BL6$#Z8N{fjlmI^xRquZ_KM+B>6jk*As+FW3fueIH%en;KAm`%dKl
zMm_&0VlA;HNif@GQ2MO_Xn@d_-C&qw?s_7CkJ((!J=l9^oO-30A_aJxd9@p97@ziE
zF8qIZu75-i7YP6q^Sjt5C5<f-G~ni=j(ubaZsri#uvX(D1-QTY=AisvBlI8I?0_E<
z$P8?HvW=rCObVbY`r$74-(J9f^V8=`qD5IaVi?|f6_DNR2h(+gQK6CkHN^u9Ldq)R
z|LHfVLI}fqmact|QC*##<gtSV-YES62m}J(MScvuM_*;$q=RM_?A$eq!STaNhx3bK
z;lzh+c!6Q(-QoH}87cCVbvBav<m~=P`xT7?5t-%kPV6HE@fB9X?9|)aH%al#ec(m!
zxxOYIr!H??PLuVR{{DX5mgcuGu5%VE5e-6u&Gq)`XBKmKJU&dcwoi|pHBF1}J<ICZ
zPus`a8Y%*;->~R5nYM*^`}S(GNz3Oj9`MXgnabY(WK=gg)GA6OV0LhyCNB|uxNVcH
zp6YgvFTbHg=`U&b-;^Y691>#AUn&!_XZkH%C`p+8o@>fGlMaz8aEv4MXCns8uUXW=
zWF#9(r?youcH5&+ayNsju8|R?b`RX#)q|s>AKBddfIl4vh>n|Fy$q@-<Yv1KjqCQC
zct6Ma1vH4hqJ)&NuGo}1n-kKK-a*>6-wXxM19u+?0^|`l+zX_gd&02M%$jdjHvPgL
zzv7~u>9XSdaOaUF-B}ekDd{WvSuZ0*t5c%#zhy-%$%p=puN(53j5SJv$gmhe2&`DT
z_7TR*C4~nb%8mC!ae9wX%-(ibi=2>;Mgc%oYpOK;C}pF;?>Nbq-$j^cbron813w-V
z{Z#1x{d;ux4*^@#k74Z152tET6dt<_g~UbyS_QdN=(QPCn*#nP!(AVO)$g)nD(`4v
z@k31;kKP$IKapZKJ4ja&iRFv=F0=@CYy}w!f2qF=*rDJbfe-afB#2Nizr8cexOo3=
zxxSGNvT$C0Xtd0=Re7$lR|J)B_ku3Tukjp`HmhdWo;TN;sqsl8Yc4~fNrIk%YS+j1
z;~Fb{9<aSEj;1rJ%qu))kKD_k`|(RSVfU@(%LQ-fTtWr{zJmnLPysw`3qh_A3y}KR
zsU{)Cs62ghGc3<zTi^YddNuAB2&U`ss{_L>t<B1t<ZP0{62b)b!^QB1<@9qY(U1+U
z!m~%2rQJ!~$H_)KI#AI;q0JMRzle{JZ<UhWn0y2=xtV_6Z7YGo&1SK4Ht#ZpEZ1P_
zYCHpxJt5Rfw3O2z$a=vM_M*QUtor*<-a3a>v#lgWDYy3sWErp{b5D7vQ7%3ut4<de
zWw8;^?!*t^j@8Q?0KD1viV>`52Y_5QB3b$1E*pdkR|VWWtU{HZvkGOGL>$lzS_>+g
z4Jz7&olw7|G0n)+H+NTmh})tfvqe}!JGsa-<I@H8(kR3f#uW&ds%Om5Q!@_WOSmP;
zE_h*U;VX;*RO>yw3{VZ4_hTA~Yi4wJ`ii28iN&L7I9$sMZwuL53w_ggCNtDwXRa(=
zTL!h~EuYy$l@~iqI=}zGmY({soG;(WlDkH<Acm@4b<pV4v>O<FvO;z~52}YMYr!&4
z7k`W%RbpnnMkZ)J2=*3x&b+LDckd0P+=2Zx7TQ%Hb^?4-%LiA+XN1#W2Z^r|zFuCI
z&D-}Ig;Hw8LqM~P6B?kp_=<wsfULT=32F2Cei^TeyEkZgh2V&fy2aUaG|giUul*k`
z;6Tc^ya3vqhL{K9aIY`m<|+>0GTC-Ej)UXDG#`YqX7Z-)80*Sb+3jYIl4zu&=cNHe
zyA(GTe`L=?H4ogtu6F%wXAk3TbM=MQ(i!hBlW`|S!A6Lx$ghhxq+c4Wgz^^-_3}Yy
zdgFyhq3$WU+J6F2$`gwr|1%4K-;DBh#}PC{)B!6=`owjh)pyIay9u=6|D>W{f07@I
zrRSfq+^CESOmr+aoYP}{R-8eS&D`w4cn?Pq=0u{3t*VIu^81w%%q}e%;W%o}^fwoF
z15shI0LxJ7OL}!BuD9)s)y&^;8SP!u1w}V1<3=NrMGEZ0Q_Fk>1JexxU$~@RF??|U
z)A>{}_+?0iCD})TK^XgPS?_?UVwmEm=+Fw(AM>zyiImi_;7GmtRKuZ#%V$k~nbPc}
z8P-LE*UC|mW@RA*g^WI>f+0bPdZ-<kqfW-C(``)j)Y^Tz_IhD{NlBB%m442%4aeKA
z(IZD#I%8lthw6Tz*g~YfDNzNA`17_tH^9$VHSIxlea`);y`*oau(V&yTm_D}d^lL=
zt`be;ky<5VV_^|d+g<7Gnp@Ux&SLqZNU@}ka=a?mNAgFk7FJ%By&UIW=`-cl=|4Sf
z!F`9{5LqNX|Jen^Jz9lzl}{O`BPc!Qxv()bR#T?7^jR|f_<Pm&6^(9&^UpT4&mHN^
z<$n$zfp3O-_^s|Ik0cjb7k>?3<hy<!dMV2N-niCwTLdVhx9@fZ9=BW65HF2cpW{}u
zlIF~*h0UTbA}5_E@L8rlAYKK&FzVdjO(-`|+ho7sxaeg2Ndj)?bbvR(y0XJY&)Ooq
z-J@P~JH^lJ`h#s|%TJd{)`*LqfSx9G<|Ev!)h<jB-^`lBTP1>HDlN9Tw@eEV^~5lz
z06Xq$6Em}ej#3#GXf%rFk({Nzc)Q=ZA0B#shLQGJyLh@+lJz3TIiUdOaoX7*-p>pZ
zPwG=RM25P_AYqFBieaNEZ-!pcc-a^sgK78>D59AM9eGL|kvTCbJQpL;bmzX7W;cyT
z*A~~?F~&REYRWFcl3v|IgK@R{l}iv0RWtdN_>w061#fa6fu+?jL{-GEH%9n$>7mBd
zLbkN2UP``n3UspTR+|;}!~K63!xvQ#n^k=U2c5@)s8>~y&BR}P8|kP5Qbm(7?fbdH
zt3-BwOjD*@jas87@>65`6y8MTg+{i+HzA&`z(pq=D49qU3E%;Be1Def7rayg*@um2
z;lD4zKoyq!YIG37|J%N||JC0CV;ZLc|ERpJEtgxPvf5(1*lu6mH$KfwD~GYd>HP~x
z+S6fetR=e~B`edX63`ooC%Mpfnu|G}9@6wWv<+Tk4Va`>(neaHHc!FI{BhDi3hrkj
zhF|=#+GcV^K|<G|&FMM8t2jsH_+}D6BXk^XpIRFxSuE)mC8$C@%%?r?G*VP~C3w5b
z35Ey-Rliv5m#e5uH{I>=*ww`fUvM?rr=aG{E~dWy>E<jp({=3oM%NdfL;ZHTQ7Us~
z!T%>NRf<m#N}mbajHs%oPMD+!`plQfpA*Nk@)JzrJlV_C(eg={(X|&pCQB+mOhG=H
z{7q9v<VtYo-#w`CKDw{;Pv>`|WH}iZSjz|O=zfCxM4mAhyD;(CZjjIU{mY$DJmd3u
zjr#HE7=nBGG~t>?7YA)I7JR0Z3NG-c`zxOS3Jn4?zX1b=twg;?8LXG$S!$4Q+VL}*
zB)!YkkORrKlBRl1=P&8ktnJoc?5y3t?yZR&b;-_%Du6Ns@73t&@t2lumU1uOLZVMe
z{V!S7%xgrc*`DiII$BQ!F3gCdkKb_KaVu*!%b)T%iajb1ak&6L`vJ==bsu)hIN*mz
zLe&QEj5nwNgwC;Wb92qtN~pM6s3A4nI^(GzOq2@~7NB@^iUC~F1VwrQ6$5>GEW;~l
zUAzRK?0x2;UQdE>o*!_uZ6`ri8yK6bS2>%@DxB;}jU6dZExr<pX6HS)6OukOCX6wR
zmJIku(L!4+HVhXCzZqDU`yQNfRu{U}wQ&J6#NUn1n`vio-4E1yml=aL5oGgWvJ_$<
z7p8CCy-_!+2Jkrj-g=tY9^w&z<#)@?+_jSDh+m`xFIj_Hs;XH(5Ll@tNctwrZOMAK
zB!{~&(Y3HWw_bNyv8O$|XSFYiUWzi>?Ni!6PQr<tRg^>)P6g))EgI|KKH_80%U)r!
zFR!H2`kaGakrMRo&6*vzAEtGJvT-<(!^r1nyRGlu?$Vd9wfS`^-3yEJED$SgdfHZV
z$TL!RjkKN_Q%b9`rsf>k#oXSm%z0JBA)*vhR&NbI(Q#WZ=Qggp-X%5KCw{l+MES71
zo2Dr}X!l{cUN{!#2UW%}_K2B3g(FNx!p4b5!knko2hk|UW~~k(in)$vsK|{#CZs#Z
zW@1j#+5UzF^7Q(;tzba%BiuTU9Khb}H=bhx|HLEh$!?l#Tu(vW(>~o*TmqdQDh}#`
zNn$=@<=d3*RGOcIvfq$i&%Nkh9Z(d&IWv|QQ$Z#2vyyG|X%q$WKI<=%D>v2A8Lyah
z2VE*qXA0WlCoQqc86SAP{iz(+!g)w$+%ViaXMmPytbR3+KpoaR1Ue4&z@W|jK176S
zPR=%+|1|0R7RH9)T_|{bZhMyo0Q?MZQ_ZT-mz+z}>3GbNdp5qk$5Z!r5ju1e7C0AO
z!@Ba5U&#kMB{G6(B;#E6m5=jjBsiJTtR<lcJ?CUk)zT<vHwp`~G1$N*Owu19V*;o8
zC}U9Do+SZFWwN314ap;xSTWT3cspT&Ibc93H1Mk}_&G+DyJboetz@-i?iNKh<#5@m
z!|e;FhY~zt^MN6+0@;TJG%8c1Kzdz#jwRy32*<&*&4P35)#C$nb$`9`U`dRs5{XWD
ziZ4(1G?6qXB6G`vYTJ^NHidun3yB^Lr9roN#(kq|9-_kZUWX<iH@yqlj?SuE;={n=
z(IX$V`f7%n$~M}0<-?5P{D@;fJ$Hv4ZR!&xz)2NtFw<00zR4#cq9H231hpYQkOOs=
zTFh18O=Q(>yn*S^Y);*=dHlv==RvzvGi&#*LqQ|1Kb*MfuC`95eQ*h!;5~q5kq^fY
zrwBOU$~wp;vmDi^!$p3a;}ei}ekQ-gdb+F9Wza!(eAzs>+vsraMYK+3N!=nFoyV(-
zS{(jOg9cMEzA+gF)MAn`Dq|UzdK2bYKKnfsOwIrZbe^+2{{0v^z$O72d9IB(<R2PV
zSf@G0$_+ft-4I~v@lRK`-s8YJEtH`Xnk;J~%z+8h56bU-KVBxgXynduN=84zdwf+F
zon_~dt#U&AGu6*VvI~1jWj!gwZFP^M|3SvORu^}zctXoz6L4PTpZu6J?p4DZuB|%M
zm+`7K-SL(fVz|^4M~UygV%8m}kx!q2fK87?pvN4HqhgG5aPv@~-|{qL%FQeTOt?(s
zeLEbf?~rZT-H>64UoJ9R>S@1U?_>QjS%ecgPo`NCUMT#0>K=L;{g%JH%dYgd{bu~q
zkC2mLTsG^XGIHNqim!@}6e8oLr;q^wbuNQQD$$%35+ZmgX^ES<5A?}5wvhSDtO~im
zOQMN7`{KuamfY92ZrYbk)hFl!_ened$^I+4KkFcOF*Q3|KDflPQ}&I<4q*l1uk)B$
zpKsZ;V3TCRo2Vwo*E%NkxgNL^t;b3k<DRPS_fPbjy^QGAh_#3Sf$qdrlDlB0VM9g?
zr2A`YRJLglu{rxW$R<;5p<OYG&g9e0eDb(H_fQp8^_UPo(U8xsV6k4KV>#kv!=fCU
zdUk1BnUC-{_0tko8>e#--3V3Bn3}dO_7d=uh!1PVlJP<;shSHV4T9lXLM|7};Qo5Q
z68SOliAtcT%*whWtm*JfnzO~(u2g3!(0@(NATHe7_?9I>C6|x=Wt-@=2J2Q}1)ZI}
zbVc?%Q}fR(hSaL@yUUcej3JbApk&tgPUXWk_NlIL=YJCFFAO7NV$ApV{9hwD9Hoah
zFrS|@Jr}A_$+d>`OB`kj6FqrB5#(+re!3CUtvtFa_Es4$7O|<~#%_N<K+ko-(TklQ
zMmK#DDcU<=YoI;?ekb1NDq6PZv}%e+?~wNXqg6=i>qqJ*2N3#?vrwp%TN?dcp<vko
z-_6t3!Zl^h2zkT1!r%fN{}!dEo8-fr)yZGPg}ECCE5$R?q&>K}NJv<yiUyLzVdyD0
zT5cLHS=2~=4@UKI%~~zW`TKW`Ikjs#_Ht)`ZR*-1+m>YWm4wJ(2M4s{^p&9v7VFcg
zca&GyhyE33mmjN-xh~s)ldZ=u>(#<7_VnpUPVFrcI@P?()C;P=^CS7sm)<8F$1bPw
zsCs_369YNeKHWg&Hn+ouD%Y~$3FLx!$WWxy=N!{G3-Zmb-PLVnw-F4<fYC$4%mYoJ
zxz!??*np`UhCRN83`#EKxYYfKDc!nddVX(lUUJ+cRALVTmYw{XQ)}PmPE$><g6jCM
zMcONMfPIRFy32z`+#Bg$7L3EEM7qCNWZ3`1Nsl$#y%gP*mTH$zSy$q&YCB%q_7!!n
z_P^rXha`WdC*V_?mnbWu_+p+DMf6%3;gL2S;vlTossdA-6>13zk(^Ts;DGTt7VyxO
z146woRpQOnzeh+~fF56gD`SxE<u(u>W#0gv6<K4}u!Nm{uA{F5EzS3nsrl2Rt>9Sw
zMhJU-OF1d`w}aU_d9T#^lOAy$?jWAfW0SWyt<5L{zoMi1s)*F<28`A^xWGihQ>Sb7
zm_u1m-OQYq?pUGX=!?FrgSD?B^-6pdh<d(?NZwC4fo!s{mEud<rQ)E1<GuZep<~cd
z)o|fD_GRliB7$Q}Nz9l^Q2%$Z{na7h{p6=PJ>TZ9R(AdiBrkV}n<#E#L`VS^Ir#60
z`f|SApWT1rbaIyDLh>UX!KUWRB?JRN=PZQ3z1z)seS9;PBAcvHiH3WItHBQCH$BW`
z>!glPwk%ZjRX?6+{SlEhaeKK-y!d53Pg3Y~;p!t27#NsldO|nZ9o%GNme-Z_5B}VL
zQD2i*B=qMA(w#moB$?Gbp9Ck3H!3`wisx&-AHG%j<p}(Wqf~Pz5OR>WE162EqWcRl
z^3c{vlhkUR{dFUN+BiS7RM<GJzY<js97IL$z(x6nJi>q@a1ZjhXBx(m&y;eB66h~+
zY%`i(TDfw54KvRp=1Hee2Hor)7%dk8f8u8wvTP<doQ=1=y}*e4^ylI8kWyP1zYQq@
z{`OcJ#ObK)96ylo#e4vu>Nvco9eo$&<)O1`>8)pM0PMoGp*_pNIOpj;-Zs>AuX8}7
z%J0J8?xaI_9n~vAUFK5*goPx42I_3jp{zIt5DkYp7r}=0`ny3`aZNkO*^udfYB2&W
z$rV?4Yz+PW%J={(-F=I(M~2_$^(*SO@{eNh1AeG>(f}s>a~#`uoVQvgGb~^A9Gg!X
zEKkT^qQ%v5nuAN@#jc%UAIbF4)O{M~`0YQ^#uF_;c7aIk?CNuCRX(loZ)W3^qLs&p
zOIP<>fd$DfZ~Ws_i~6#?hP2`FaeLPQxI%VQ;Nz5Ic{i7UH#qSsi(P5TuQY96=@F9N
zDFw?0KfB#)(hUYX8aBpymS;^%2ClHKn3!4_#2v51**pRc8~$w^pdV7eL_te81)rND
z_BmOroP=1fq_h9hO)WP|-elyF2)eyWd`3Z4);wWFlE6c;KFrc1#)eQNIT+^hcrq15
zk9u3aCclZ-d=-D-YHKG>x8xvAb;-!^(4kts$kwI3nU<Vjb<3*nC>=Uoy}LQVd|r==
z^#clHcV(3bHK<@3qC)Fof8Gaj-$fTTnOD^>EIB8r%|8|~aO8Fa%Tp^F9aJ{u*GNWO
zz6wu!=JLp2Nho4!rVGs>VjL+sChu6qL7IwplD10r6n%YuBx}cC4^;&<aVm2lq`yK8
zs2lAw{N`lm53=}k15Yfjml5D|<Bn%L$-U8+p70}IL84?+-K`G(S4;USnv9+sf8Mkh
z?~5Kc%?(Wd8da?I@zaF-BNO{s^*!jDpsaC*Wqq6yR@b{b$uc6rVLb_h?#Sya?G3<F
z2KTklMAzL{N^R}P&^DgwEaXTY#dYgkw1&RpYTm<1=T-l4*H7V_gjYw+a;Ukb?69jc
ztyedOrPka-Je=#+*yXqm^<D1*eY6iJrcY()|G(sCpWC{I58nQYNTpNPDuI@~@Sv?}
z`jx3vQkLVT$|P`9cJ4VjO?l$!b8KNEjncf3j-|L3pBW9v<ot46O6!JmazwgeB|Ds-
zXbq+d4dh|`H_u7CF!OqnI;oKGQ)8Y9viGgvPUpLm#$E}VKl<9&f(_!;On%>C0PQn)
zhwAeh94(A@{*1w_0y(=(O=t`{6>`I*T`ai|F`{H+mZ15w0ZH}|j^*(7tHzGaL|_U~
z`xvzC)K>NmVm5{#irp{maPIAj4!I-AqM0_n(TjT|bA=_EGk(bb_HtOh`Adb&_YCRD
zPxr_(-&*nCjT4c?;$CkDiQPZ4Oq&_0=+<wkpUa2SwVF1yDN#*KI@+DO*9*ZQgpU@S
zuYg*6`JhFwbJzIuI#Vw9in)J*NzOtbEg|7xw6}~M9C3Em;Nusd4T-y=va?37y9{xx
zx+)u&C4raPl8innqbq}evBb8W^Y~AxhlVP0rkwwMEK4~MyqdEjU*9^P&(E>NCn1by
zMKUhZ#%*#NzkmWC#EGdE$`fL|%Yjh|MmNc!8>g<%>>Lhb<?kNbDcqm++DwoG+Gnku
z`|#uXD%UD((g2A#RexVhI8!5k@cC#gq`v+=U-^__B8o0wUwbYJo0SRAeISkvs4WyE
z6B%afN1_}}(+Wo@(2C#j>wR>zOt4@j^5P>gA`u)}dh-&gTk1+YJF59h-DF1b(^Yg&
zn^Q)H6%CuZ__qnunqUuEy67tuQ%GDyY_sYZ-$!tr$y6~diNFhW^4Iq<Id)9ZEGBAw
zcb<OqKeB0Yr1}h~C@)MjWwR@%HYs&Eb!Sb(QmP3zvzDi_*L76AjxCUfN~YN_{SVPb
zkYtW6r@&Zat<vjXXQLcuR>h2+H*l+u+1V|Q`Q?8C|MJp$9TpZUXNz1K8=urFdYKOO
zAM?H1mZ$jR<D<1E6b{j|gj<s;-I8}T52%C8Z(iq^P;*LUhIe>+&>?2{*Y}6Xc^oY0
z8`ZI{5=4a)(1Ht{W$3j&Bz8Au)T-MHcjA%yjd8kEkhTy4m|vFT^ro7}RP?d_b?^$=
z15*THuYU$o2}&pd7!!O}O40s&aP|Tb3Ew<upz%KMjS?1gs^8}ex$8E>#p?wZ`we~m
z?mm@4lGX1&nx1~cm0bDJ5+slf+BdE8!i&b*5>%oJ+K^xnxti9!OYCzyk4q8Y`&!39
zYyS}c&1G6CeMD?FTsbO8uRuM8F*jmkW5dOscuuca50#zCl~802-<&?+JY@qa8-Cmv
zVL;$0ZtBiwcJNPlfFDtAI&Dh%CPs|;LN-6h5ReC{<^}kX|3svd>%<V_UHfDS7HG~B
zP&-Hyc}_W5o4zMeL={&khoQiy?`=q4nD&kEuSd2Y1$R0MX5G+$!R`loXh3m(0+YY(
zuQks=eZ;BchA&?PE(msVITpvRwsY?LDKFd|*-37%coz?N<PG;ryP|h64NPm0AxIzO
zn%vqiU>k!K@aKhSRM!E8p_x@JCirGIIN)-49qQnCo|)lGTZGn3wjZH%!LrE%7XN@c
z2Oc_Q@e&SJp*ilY;O;U<=#q#IoseLzQpq?Dl{>qt6GMu_@O^eDYin8G>mVZ-9tqp(
z9rv!w(7-5Cyt&8@EvUG|N@?E}|GL=dRFog9A-mp>xg3|8f`~|#2YQ!ItbcFQJd;-W
zy>A)V|2t?xoOz+sXVyI-z9_)ek>@Glai9V2?jOB3L_vDPw~LYk0)OP_7$@Rfu#adT
z|M_g#@+;T1qx~SzVKY+l)Gd)(-uw#6(t&taU<xgK$&BaQO_l#uPvO<>2y88(fN)iq
z$%m$F!|^XS^C{DjTgN72mZRDbRZ&V|f!3W2j-E0Wp8Q$rP`gTO_svf!^&h_(rX1#|
z1H0Q>l!Fo?vHE3ayVXLNEE9X9^ZagIzwSIcWWrAywsd}`x1c8)ST1)jY(#uYe}e(n
zicbw~tuw+SRc9&S<Y}=p8kILek_Iz2M<8i#(`qKQpzJV_Z`{(%2*G$fN6>BOd7xME
zN~Pp7^yIz7k5cQc;N$!Gq06gVuK9w)El<!j+Ci&9YjJ7pZUecMeD!LXXP3h({ObDD
z84%NbzuQK6QoOMJJckHkx+U)c$jb{i5%VSd<IZ5s9QYqES?uLF7!FB-$=wQxB<z;f
zA8zxG>N-0reK}C$ZZNaKq1~&y<@LxqGVVH2tiPd3Wfjtz&nO@KgVFU;a5EF@r#kA<
zoTuOR-;=6(#|+fJk%0>)Sbp2hYy2X;dX?Cw&wUJd!PML+Qtk84ncExg@#U5SOPbp8
znp1s*iuogJYEW4>`AtnH#FsZ*DJ?*ab-^{Gt+g$U#X_vi8u1Im@{ve8$XJ@bWDj($
z*z4Qq`8mja-OAU^u$8x!`$6`}NSHK=r<M?p(D%ob<z3IGR%}*6i82i*JXR!;Kmfx!
zJ(utX&M^O<c%xqVO_rZ4f@*MT2Bxu(4FPgAfpAB!q%I03?d*}*$P#WmMUUv?gLj~J
z*#XxrT)DT;YV)7!q_qnfPst%I#>i*}L>LFQaDt5Vx!>l=t^Ewn=-2&*g<-;Qqasvy
z!+DiNxLLbuc3nlu2=<)=ZyC=K4`<N2YL@|Mex?%gyoS^YdnmwD#goB%q|`JKT)+0H
z058o59~Q-`%S?6SSdMjvx3Q&DR^99PVSLXY9WE5P!aNTyS9$E+4_fJY5M5w=xsGy_
zJy<tjeYn4o%9-lDX)W4%J<geCK8<!6?}ps=>hCTgBYq_k^okCF>U})H&9}`N@DrDF
zPH5TTx-#6_j>z2)4e<XZ+K>!H>dSLOzj14nO6Iz;vOU+~zRsI5Nr%EZ`>pQLwbZ~M
zP(qsOMcALR^YFH2&u%>DCmSlv$1rTm7@*BMz~lC_3=Qae<9^fLMbqqR7>HQf<#E8S
z)CTTkzFvK%b#|c-&2I1Q*fsZ->eT2{-`^YG0mwAm6Vo|4b9!8}f=#v0p?W`drIe^b
z@{3aV74H<LqX-S51=tbH7`^$}1Q$v#tt4hVe#>_-dT=^_G$^sY+}ov_vATbe&@l2+
z6shIknf=s0VujWwrigUbzx7;@G2bf&clSMtGXINNkr^MeoFCY1+`0x@LxM%I%`#_#
zKU`FvqCykpCXf4pNcZu@Woy}?ekc1rB<{0Koxj|dFIrC+KL=vy8C$&811EUVewT^w
ziu@N;JihEoh>U~}h?(a_<qcoTx&nU#nFYx}fictNxz!GtZy$7`^F%h?`QVX+ISJ@>
z0|en6v4Re#EFw^h+|J(#=NoRP-0WVp&zr6I^I$Z;sqY!d3=j13YkOPqw=D$Kb?6Q#
z5_qYzP&Y&Nvm##>wSAn%`d6C#V3x=7c<M`+{Fmu&vz(1xykiaaK}7u^JXKZ=*9vW`
zx-$jckVVTA$;9{UH~E1NGl!wXt_QUCdX)-vJA~_E^7qfP^qV{f2?J9+)Xv!F=e2cn
zCD`)m;u~7YPa<4g1JqkRrE{O3_%Yj3n0N)`{`euG=fPxCM8x5BU__n9q!P#BxFH9*
zz$4C7R6ikdu>Jy0ZXitJVX`NZpT|wns4joe$hzOhT(#irIz!y1Gb^Wa`#IGW9MoZo
zfzkYg&wGlK8^I-K9)3M@QqVK-v3as<#%WWp4@)HZ8;R72PsLk_6P5dm{A$mrhFjPi
zVqi_@upyosYI%88h9E1EdcXhUm}dtoYt7r^@iO!s<*?7U`<$xUZ~ndpBU22%b&dGC
z?#Y?%>uFC*1L^)jeO><?P`=NB8_IBrwztW-K468@BnEa&IJn;*=z6XPAh%|V9K2o4
z$wfjuVPX)GAx|b`KV|dbGxje(2Ac_^5^fkM`@g*DR>gY8Q7C!QK(#ox`>i;Z&f<-v
zpNr=?%L_dw<}3ao44{fX?2E60L8Gb#ms>5L!%H9H7hwBkFXjp*#^RAu;-C`nepuz;
zhA86Dx%@gNP4#<5*->S+?R6*ACcUvzq~TOf+WOT`D>&P}`C%>i8rrrX0L(Z!V1z;A
zdAC3}gS?Q{rX#YunXaIn3p9M0_c|1nHLzc2$nk9i5U$sEuu+-LtCko7B<I!y2hbu1
z-j|~IV(Wo7q#>JF=EK?JdIH`ro(}$DC6cCn@iHR6cMC#H!SFfq<g+D-tFop7V=y%~
z;b!CzvgG3stM#rcltCQ1a^x;pw1C0l5D<`7Y$Q{?RKgr3?(T^lr&LC8C;9s$jq5~7
zvIg-Z%bkU6{G0ZGBokI2jD8lWlJze2ByU#<cD)jHBZ%^tiBQ+?Kajc>qh8tb+WzdF
zkA_OVrJ>z1bk}UBCy}wrlhjPpp65DWFQx>}McjdUI)|@BFTQK_w%xSYeO#=ritzAH
zE(h4eml@ji7dH~j>KA|D80OMb@L5ScZvP{)Tm9iyAY)5_MfX@J&EB}>TCbYfvTcE6
z^$}s9d`!-*u6!TxGx~{?22&kJwUPpaVi9mZ(<x5cPoE_=BAKTTC!66XH_E+>eq2sW
z$ds`e5yC~yE!Xw)#n|rw@pqVOUjoYrevfivnG(2eGo>DPs1gLY+<;Y{MAfw+c6Jmt
zhMxQaovBhudMyo|kjz)=`{;GsP$`K7cjC`V=gc*?Ju*<yqZESP3A(6T`BO+R<eIg@
z0;8}lRl>Dl*`uG@7)z{_@!CgeK-z}s&AV{wZy&BeG@3y_95q>OHUX@9GwmK##mR!M
zu8UHqgoC(3{w-~vA}Uhcc~^!j!(lS<TTpo9bdhyizyDbwf5XyI?YLR(O4hi8JkAs_
zXHr+N1kk}4`s#qKs@Q6w{E&58Ag-U^&ov!z*?@`#Ol>Mmqstpj@)U)u>unyeL6Z!n
zN@t1iFLYqK{5cKKa)<p}{8?+O9wWj^0@o!Ex(jJm>i&g3^95tnOe3!%Cm?Y8sunn`
z>Qn^6-T(BeR=n}NaJ>lY)L0j0^RZQMAi=}s<w7UjPOl>Dx|QwH;c}jxsqmP`Hnk0t
zWmdkBQe-UjWL}V~e3Z%ct(C`Vm-b2DFy1r`pA>c=SJgr|XO3WQ!Q>55BTk^CD#YIH
zXi|zS14oQPAz+THs$8?G$!!ExV;CNgbp@8h97uSV)4BF4^)@>l-8+>epV6(?>E^t9
zgR#1&ie!itCVoG(*~)uxw&$7k>cr{K%N(_LE>0c0WQ>vgDsO3X4+CSPI@z!UIGUwl
z$~FbXI?yG{$kAobLE(oPGP)%FZU?`LT<ABbsuQ(dJ<RN?YCV>1STJiTLiXHNQ&Jdg
zc%BS<wxqBAsUcEwzWoI3Y(Ljq7X*>C76W+HbO<yGq?DPAG6ll2$Gh!LoAPdd(=3Z4
zx;Pr#NA9@dBS9yIw1h&VO?YSzi8s7dRdSo2G?~M_hOMF3!9r&SZQXPAIWKDwK4?~g
zcx2wA^@LtU5oPnP)vAwhMBc+{-kRggf~+wl_J+yB&$uQlrGgss$+{G6oc)_q0=s#c
zW&=|PjP5UU6XFr6u!*|lm27<~L3ELGPm<N5P4lG#-OFkz8Je+3co*T&@KC+uQU~Ye
zV^Ij<X_N1wd~1o<vOss)y88bX7}$}o`nzg9N_HZ>cBL<6n)cee#JPcYu7=NS&!w9Q
z<q}_AY{ayA+f_co!rX4~MEoN(nOZI5V`BP6#4>TDgLN^a`YhKd3)XNSo#RVRajFz#
zFo)(rJf}^nW2@g}B1thrk@3pjnXKu*W!_`vsIeK6h$aha#z}lT=<$Mla4^cuRq|sT
zdR)2rnEQaGa^~$_x+^{wi(<#&KrM>oo(oX(NwngfJy@@(Bah{1(32(owE%qdMjjd)
z2k#H-65pv`$?|jB!XaPGzqoQ4)I4gGpy!zS3+eDPN^4=HJ1JlwILY3hvzo1qt*9@8
znXZKh+D<$bU2gmSeL|p-mgCtVuR)A0nZQyGYj}kf0d7kd<XpMx7{-OP_<pdlICmnr
zorYe8eyL->mwj9HsZb32@0xFh>*RiW9C3UWVfDjZ#y))e9xW>Q?k4t~6+n@HHJ<s%
zqbc$0n0E3!exM0ASL`(ZCs}e9;Qht%!)7FS7i0cWCEZ8~Buk+9eQ@9QkJF8Bh2Gb=
zNv0$GAw9v?$hMhy|G4#cj;AcG99TY*ZC9_e&1aGea#hQi{!x&|o}BVhTeO+GN#9A1
z^#r~V7fM4r7w%QPp5r<5kCLY3SA9h60=BoamTmP6Eqss5nNrm&x#w25B#UWzJ3RT^
zfI5ZGWT>sQ&ZOJlqq4_*F-)ttEYs$b%uf*A0pP)sn|v4!lRT;ho{X1f#bVCrFQR2)
zFJ<w5%y0esOizz$OKo{XW}u_Ls4Lec{r3IBT<<fKwr)VGLSj6KLqjwmBckBh*C7`K
z2<!lRA&>dX9y|1DJ3#5jd#}Y94akH(orE1>&;U|}E`-0mW#Mda-GPDDU4|vz7<sv;
zV#%_ysV#T^no8<k-0OvtU$#V+aMl)`?Kl+saPM^Avcu*zw~4UsK%M#1;mjnfcj7Z!
z63^*6kml5_-Ug1FC|{H1Dn<rYlI^(`H=O_tCZQsX!a70#HsiQu<2QRdvEN^@V6CmJ
z?T9b-O{D<LWx?WmSDh?_D%e%N^aIW2cLvO}*IH++Xm62}c*IIGE#bo-Z&QIx#9%>-
zzt5Xjc?d-e9dr_^Vc<HBLU#+ux=JD~5U9_qkXw6KFBk2aPY8vnTdF#A?R!N{3dOUN
zvQV^4M_uDdMJ5>X*^iPs`IGQW0}njXUsG^*|D{~k1t#9H&@P@MjG6iwA>`??MI8v0
z<sbqO<F4Z3mI>dgd!1{V8w-MoA(^)ROwF!uE6+ap`$d^kPt}a~`atL{rf!5;`9C!^
z{dJiDkI{u&r67_$RnG_FJu=@^1F}r|m!VuH+|f0Fl{CfvbEC7%T2Uw>=aP|QMlF8T
z(lzK=+pq(g!ZmhwD7;cz-s-KfQ?i)yd|k1x>iaXIvA12hYcjscb$JY(&2NkX<v+_S
zX#ZFrO57Y}4awXz!<Gs3O`+><xDID(EAc92W6S+(aA-fw2<-WzdLN>b5=ue>{IJ#x
zVg@zDQKT10>h9}8KaKD{`f=9eSGuXb%OOM_<&iB_JP<@>oX21gkF8HV{JJ&t=R!zO
znJKDb)dHtPm^!;=2npQzFkcyr{F2hN5KB9t6X<4I!p4}7&(;3sAnY%ZSqg<~qw(qM
z50u4V|CPg=`eLd9_kompy-Jdb)P`QE&V@MT=losTTtZVKVl09uQNGk~yRuX_GC<@2
zgw7jw$IIO;lgIvV*MAyTW%C)_nPmjgD}F5I3aI{Q>Z(O7Q?xxf<8i2-=S3O|L3<~r
zKG8NeAY4|@a^CYAv^qi7YF-H}5p#gPcL+Ylqs(d+4cI%VJWW_?8Lf6S`L%Jwwy@qR
z2=W?Kj|qS_{!$h8z&}1QChl*knh;tdhBUHIHT5G_yzzTkE2@p7>69~^Ck<JGb0_Uz
zETHZ1uhw1+#>PoFo{w(a76=1&njbICJw7wu&Z2r)yq1Q(oiV>?NQxd*-m)H;hBh*Z
zYVFdNTk{}EyFO<TP$qc#nXTkg?UDcvezpzlzk#vQJ|P*dD=7Jf1>rhyrC1H9=^QH>
znx!QJ1g2q?9~8}7y+__{%N^`YFKQ9G{27+1=n^hNJ_rW-#O3*^^%Cufy)Kg69qwvi
zxqV;c^pCklWq!8VzELs`>in&E;pEq)=v<W@K!1?5z1&uN<E*ih-9%uLb*ahBV#fh=
zq3@=ZDr@7fC5FGXyi$7qo(3zH*Q2R-$WjyTDB5n){L9JwPD8I92Hf$2ipkuo;O6av
zPvHaAf6MliR8b8q0p(tQ)g;MI5H=;)8qhVaY0zFe2_VjAD`-e0F<NPCM8YxO13Sju
zf`N_DX)nOYf1Wpv7mM+CR3xAFu#nOK{l@uA*ARvm>$>xwqizgDC3Ba|%@*M(^gx$2
zR7<+yV3NA{KsY#bf`B?SGfrZEI7IK~#qQ$N5<=7S+5^dB6aXE+W(wuPUpULbH#7Ri
z%qLlpESg1ybVX)2q?}ro?`WdeR|Uz<g7%s67pTIyGPr$NB@l;<^p%42M;_|3JBepB
zMp`h}t^F}Kf~E+;cT2QmTesIt?_tFc`c2Ellb>)JSKGW)S`7hK3&kE?x45|a6l<vq
zwYOD*=#MiPD-L}Yby3=ofTT*e{-d+M?+7t%=%Vt3s6NxQ^=gnGRRRD{i9Od&4?vaF
z>wOqCK`qxN=$buy{1@#fiqUzjG%NYlr<66S0>5TUnIuI;xw?S&E7(?47p&dc4~H;s
zU)JKuqq8#=R_rRlXE6hiQ!Q8jVWn5@ID^yZ5Gfzmuz{|wtnH2E>8xL<^JWhL6$K44
zaQ>?9OaS6;u#x@g)WYD}e{1|7{2L_)_Q0&~{!VM(&A#Wyk*|hF_q{2BB6_3Mihg`s
zz5O)hc(PVPiB43YtNaNrB0zp8tVKQA+w@ion!~jVAskhf3X+{?F?kKUkQR#LpTFB$
zvUT-z)pgp{%}{#-M;Pl~a(=4#TUOtXGI-<>pn$(fN8vvl-Wq!KtKQP*bo3!kf*wd|
z%JC?yCNk{8b|5;LC4@2OyUf%Rkq_So+2Rgi9{YTXdTavGjJD}C(m(pxz*$O-Ti?Ei
zcap?pC-GR(X(lVhTZWMF8tCji{Fc87pb(|q%fI*mnA^u?)Fd_LD}uOxlL;{|85+Uz
zBl<8V!BuS@y2laC!TpmW=(ymfTc6FZhL(EVJv7nNtlX#=Q!IdHTz%pd-Jscyj)-*5
z-i+E?>g$wfI_YobUR$!Q|JFk4>-ByVDOxti2cJLfA#Ly@pA%KME1b>;{Dp~<2oL*}
z$squRj2e|pwF5&(?aPZq^&G{fFr*7D(xQn|{E<hu)xRMjGWU#Qfo$Plt*GYuc^TPm
z(afWZOShrv>vL-cpsu%xLVkC3eD8odEX;8<2*@kQH1tA&0zq-}wr4D_uz%ap4v4#Z
z!l?Xvxn*BHF9wnV`SdY@m)mXerIjoW!8K`{e!xhn^Lo$-rGj@m{vgh3!d<chAD3Y8
z`oickD~*>3=l`$Z_1r;IT{FSV;%Uxr5syBBa9X;@$CSEdr&Ab_BAE=NcxSc3C6ecF
z=oVVnbzaNeLFTyQv;Mw76TuYTSe5Q(VOY)DOcPz(`OAn%que}^zS6CWO!g=`1$03!
zV=RAHU-y2(^R~I_dga|0+-SO!s?#MmEY&_?8ojmCo%|s}L<vo2=}pCrEl4-;${w3`
z-Nhu4_<exA%_lAqXdk6QP!XFe)oh`Bvkxg%j(rArpL>AhIzIeok1z_iIZbRc(w~HT
zOFLxH^$%lyyfZ;IvC*O#vSmSImMAA)-_^FNbmQO|UR`2TFLu4@H|Y;^NmR9ud>0I&
zQVlq~M(jj$qmIs6QIv~z6kZEE-WIAKOyP#PJfZ#&u;T|0OE`{YHorJ6a9b>4zeEr~
z$Wc!&P5gtvE^o6swH@80K%mMrybi$rN9vbqKlJe-MOs{N7CF{p?%cb2a?SP)cM-4v
z9X!mKaBiORDDA4LtH+iO%j;YZn*tTf+FJYr*{kBcLM;`ukTPRVn{jzqA{z)zbkn5;
zRA*J!JB0`k2V!{Q;p`zdyVNKZY>It<wxulcGfLLa515`N(p>w49Por05ST<RhJUrb
zx=xXxUWH{4eLz=(IFV{15=oDLe^x`ja@TKRr7V0co11!>i1$7j_OBdiQQU0^tmUf!
zYu&hAgkI@${Rg;4G=LafBbQeNfKk?l?d?S5j-E|v`&r$RCyP=G4$i~h9d)|*+y#vL
z9KY5q(}l846(d+dXG_l$ao`($Eh<mD#;S)i_4zHL4`)Bze3pHcH<QG5B2&HWY-ZDs
ztcqEBPe~F|<}+%*;XvKEXOFC$3ipj31s6mBYCWh2<O$2o6~6Lzvr6w!j8bgWv#MnE
zWt}zSgO?@b{L-K3eQ8=}nV!p>p?fzmeMDS(w2@vZi)!*2noeU2i2p<)Y1e^Tl0ofI
zi8rm*438`ei>|G2&D=K5TQ5f%t3OEu=tlH&Dv-i%K;H%T8bQcNN*9u4`(JFc&2AXU
z!p2KdkycW0y<~PxYun~6!aW&+uZ-ei)fLqtCYBo3mZ3}`e+@5oSqtT~?-lCyCYY@W
zpBjUE)`tvUqycFF%8<a&#m(LFj?q+XjZE%RtUJ3<p@AemE|%l1HEQjM7$&1bT^5p`
zjQeiMU!R5f_Vot;(S9UE$&iP%qc$TTtEY|rn`@oDWKH<R-!|h3J>sN%a#_B-YoTLD
zm&>BL9_<|S9EavtT6PQmVr{o)+6vB*e!u^D$;K9Z$_yN$1WXp!3>}3Ijg;R1kRqF|
zJyuFvpBqS68YHx$4?bdL4UQlIeCam!l|LXsT6j$(KgruojOsE0rz{7jM+Z27S%0I^
zX6@_sJMd)})?^!j?-W*(h)!zx+q^QKd!}rt0F>XW=X}9=o%z!N)|WDT-K(3u6u*Hp
zqQ8g32?8Y6W%<`rnwRV0lYMWc&`kIQqf7SDq!IJ~erV7V%>w0;vnZ2e+*wnx`y26o
z)Gt8s9!RS58W{clY23Ab<KthRhpw6MGw<6g-c__OKjE(jJDUD319wWs$RYa<2rp09
zHo+GTpdnUxB)|7T_UIeAGQAsZ$C3m|ku%amXb=9hyB$bprMx7A0~i<f7xQMvR9mj8
zwQ_P(5w%Ooz~p)*z4D#^;Y16=H*mj|2qOP-Tj34O?GID7pUgd11Zip3!mnBGBzVIC
zo8iecCTx81ac-NB`NB377DnQke)Tpp`u4543=)>BRLe<4wG{kN`e_r8nxbVK((ne+
zm(ld|C+4Pu<{istuDQE37jj+^a0>W-M{lC%(Z_qA;8vTpyZZE;6)>P~?fP?|)Dw9V
z^kfb?<@rEE`!I<cG<@AgRoT_~C&VE7tp{(E+cl4tn;ugw+k3T#^vrF9qszQyN@$SL
zau}0WZaiVUCLzv&rycKIE`^tOi*6b|w!4W@K{0dw$OL~45J@HnP#AtfYT>A-UR}ew
zA4n~Zi@4@@-&(>#Hh%ZB*nSl&IMDHnVtOzm<G(&F-G|E%><>yvN{}P0t?kVnrxt4+
zm#=;Y`1^Z;z20O2Z|_z64zNl|^iLIJ&3^VPrxKzkgKS0GsHW<!)!xj$F~`go4_3!8
z$047~AN=}OK_9P%MtGr9{3E{NR|jKA^uKQY3CT8a$*_lx)rq@5_A-1)`#ybHo?)D`
zo9@r@oSM*9w`$H(Q6YQ4Ucw-&p=LOKuUHyBYzERqXIEEw80nM}FKE^QO8uLpBl$<H
zPIaYxf(VY-`gZ<)qb*SQ)LLuRoQ8`jp4=%UNj<8PSimn08Qr8sdm3iJOc82EYN^;7
zXs98>;TKT8!^nzk`Ezf40R@PsR~8HOT#uVhTL{mFBW6Aj?DkR8MOu68U$~ejthBvj
zCSk^1#}zh&TFNAbwfQg>l4O~Z8($Y*#3bxO!2=A`O2WMK1Eh3$LRCu4*YrYD?oDNG
zX_u}>dwZSQJr+OW41IMcjJ*P0<-{#TkUe5;HV<uvjB=eiVSi63Uyl&vqP<-$xtRK`
z{!hnLd->=QRZo_xLRj#5{~3RS>e%Y?Ph(P+H0VErZ;skEdsn=~S2C`6#jagrDlzzs
z-g0g=awD~@u>xgTfqOHyi1mi#;+zQE;v;l20q<8sSlbvHU&0Zl;{=pWJxT#DUtnps
zZm171z&MYAkB5Zy<tK@0K^@t^*aV5HUZJI;e)s#IcatX|h<sn>@o5Y`s+4tZsaDb1
z-A4h1YH4McG_AL@ChxaiIyx*dYZv0FcU)T>`{o`Rr}9+mSQ-_Eyj`!(c)Lk9HP+sJ
zbi{j#(Ixi9J(ie5nwoxKkIalW=!6_1AP2m3)f9<<X!{J!<}N#HWpY_E#zoh6aX8#=
zTc;QYo?m<$hNWEo1QEI)&$eACymVyi-A<qGx{9Mhm+olzR~`48fHn%Wwzv5(q3Vji
zUpxtm65^6?$_=6yGOm2n_yezm!uQ*^s3qj7pTREZ0#EHcIbP)5$XT;VbSL|$n7`lt
zLn=B3Q*7n<b#?D@-m}#Zx6zL)OGl71vj6lIa@f|85aUy|mKxJ-lC)~8LI8p@XfhuA
zk5O;ZyNUeilYU<*;U>pI?qSnW!oYNlRP^e=N6c<s>e+5a5!veFBgC^ho#q2>R;r($
zzetw*#?%#&A@wCu_7ij~ZZK*q0p{>nvrKum&L374uVWCC*0tY{66xpkjU6T*hNhX>
zL_P!UuQLReZIzWBPFx#9Et|LAntD@&WKdl99*ia(OBbsu^?Tk;d+DL#&NrI59DNTi
zcaP??insa^?=#xA{T95$#Z#|pQKu&17;R}m7fv-bWDAK`inSUcbf}IabVnpR&q`Cg
zBZ+LYo7`;g-KrB2;9)!Hs`+@>&8L`p=LG1_{o=&FrW8>M-`zDpWnfw8{>UB$G`kJs
zYuLuc4uW$jPEn|Wllj-aAK_Q;ug-RrjxB-`yF1$<KP*JFBI>Q;6XMThV@nB&iitn?
ziH~$89Mtcvgmh#l#KxZL_?rJ^=*tu)^dm~&>Xiq#(mYjk^Yhcpj=`5du@^o-0NfWx
zgO9C{jMsrj4LUDha0+A&LG952$ZFVT{0xajkl+7Afx2J#`G(}m_F;RXe8yK*LpXxQ
z0CwD2z6e-f1>io3|LJOiGw5lt^kS=3#(~|<*R@u&Nmn;PbN!X4`m~poj)jT!v3PKT
z^b2sc*{DPLxNGIc)zH|?a^(33v$3#2R8sWr(~Xk0V=*6sah48ei;T4OR4XC!T(kRj
zILXkVpYzrttE#0h%lsd1sAhlr@{H~)YFE$Q_8Hp;!A7nAdCqc4P`9yAIA|ri+U(Wx
zT$!qG*`7@w%)=mi=EbvpzsmVm4Q~s>#4yb{2%iL1b1ONCTJ^Odu-pr7nud;Sdr_E4
z?S2ZNztl0i$t@5KsADpk$#$CSlgrLGat*s0mCcOU*IaQlBRr#cAH5$o0jfsP2TBOU
z#gcq*vpD=P%7~PIh1Zu67B4ra(bsoO3g1W+pojqF87D1x$NOO0Cy%`wza-wnTq^d0
zSc<q#r<W{O`MUf`rz_$im#})dH*8Qat7aw8vg@Lm9Q41)dds*dyS86g1O=27k?wAh
z4jBohL20B*B!+I7p&RM00qGjLOS+`HVd!ocV1PZ>b-&O1+xN4-@@ZJ-vDOj)IE3kp
z;R>ujij9hfU;&_Xqs>#gk7q1A)B@|RxmgSJ*fNy31^K|m!xXX3cy_ur%)yTva;A1a
z-|MLL4i!IrFiPkt<nLAv6PsrcjA~chknDoml(r3%^Uk*_HoNd0Rghp0NV0yM8L7d!
zi;VRs1hv+kF$jruuT*Z3Qo2xxwDR9ozA#_T_gpA%|K}Z~S#}Xk4})VK1wCtu;pxNP
zd0%=0l0j(Q+vb!&f#aG!-{krb$8jSTL_JjAw+wmvU-YiKvhKdwm^fj3#o((hR@-K-
z8iY@xx`i$~h*ZON+VAV1!mqllFaYiS`>{IG{S8N5OZ=u*ek&W|8p%tqxedD`u;*|#
zHB|xL!0(UU^L(q}s`J~2lhb)WOS~CcUhSQ3Y-{NMDVKojGRscUKha{JY+wJ7M7O67
zc-LOjGUe;3_bB20^5zB?S9Tzb^~`Y}_-b;CG8W62b6ad+Ud((#XlxI0{or|aJ6SN(
z9o6$|kFWj2fqzT_M~GjV<Oy-b<Qbtal!S2}TBQe8Dhrn={8ra7JJ<Bp7}AKdcA(D~
zTD>Zp(Ta7nz;YhaS^1#`#TOaL=l>JQ8YlP0<6A@;Mya65*Uk#=tAwL0!21vj>^;ri
zLQlmbr^PLvFW`a^)!LSwh@0fO+<z;ty5ZnAV(II2xuaT8+%ujnOR6E)Rdv8SK+83#
zK1jG}$+lPr58#Cw81U?_%lueUBI{WH==@g`s|kP3P}u}G6*O^SVpQY&^uaVce-bLw
zT+aQ-=}AyMdkGzECeEL9!dpCUl=#DpNeBB|R!r5tw~>P0D!gc|BXm`#E+y&~OnLTh
zLSA%J91{3xg7CA#l-x5Dv59P4dfkl2km#wulCfKVC}fxJ<Tj~u=9)%iSk8NS&|>J0
zy}Z;?bc)re6jeJp%H0whLP2%k2Y-79Sg9^@BMy4^7rP6PD%J%vb<WbD#TxU$n!f(#
z8Sz<*pHJeAI-r_^c=x{+ivEAbDxZ+pf*M901#Xl$LW}Zvvkz8(6U7E_sjPq^U(`KP
z0Y+tmO0I5Yc5m)>JqBxI_q~mJp6*fad)p`T`LS$p^0UScPu$oSeLnTFnZW+cO)t`=
z^B!h4N7GvLlAoNPM<)z=E=QLiTA?p#tW`-zxk?+{)l?+WcP-=c8`OxG?Da=7*gJ|H
zrK^g!hjtEXy;=tcna)Scdqx5Sgha}~S?z4xFU1o()kCR#p7w^!zqNiBlb09?66SOs
z++w1c`!jY<zK;cr0ABU(|2ARCQ5`T}9)z%-`2+D3yG{7CN^P82bToc0520wQHX`Gw
zVDL#Smw>G+-<Qa!QVfILxT*Kg&)q)`SZTduLWG$nb8Y3q?pCF3-<o!K1929f*&>$`
zdmb8V?_;iltaFUolS#`t&G>4q!@z|{!N>V66qTJj=PeEK+*@=11kkQE)ogmRxjcAX
zv!&AHlgt0IMq<G??2%#pCw(!VnYcH4$A5lR^*qELY=BJbMfXbGiBVq=mpxzum^eaL
z*L$2wf0#jeoCo~a(|%WuB;O1Tz5a;%V_8d}=Yp|4qN%xg7PtntYi((X=V+cuGX6H+
z{s4zs8pjFQBt+81XiPd7L|8Mttn4bMneY6i{)v01KU?D3>#h5|aioD>r8)_{)HKa~
zS}Ai?g3R>kMfC+t?wmjv=JjMxfNVy-dJ(<};&GJd+b8Oe7K1iLnaWI5Op_&hH$JlG
z%9%CSry|lb(+BIU(qF6f52lbU$8t`QED+Oy6W3#BM`LGYZlMH#A+4jKYoxDHZmFqC
zW{dyRTh5+<v9~TZbibO27Ut@l6a_Sb*3rzTR}>Ix_#`Jn8vGl1iG0d|K4gSg%4J0v
zIJ063lC$8ST9%XD{4KY)BPG`3Lt@ud9eMv=-<kD+hTHELkZ?>4s_JZ!Ge<R1-1+UY
zn+iqRwqDIj9sTqFWv7xQ^88*7LHq{S0j&>L<JlFp&)pk9qRUknb_%@ccUAl74N~K&
z_pS|T^p!cy?~KGnZJ!>M?}~cqpNLWpqsrzCGe%wK{j6wo>{B*G$wQ)@Z*O2+?e;H+
z9b`*;);(x<cD&l!a9RgFHD*Oc#o|X1lns$SpO_g(2Q&5?1^cVsg!1?qM>+<!sac2N
z3u|g>_PCo(<D&Hl%&QCwr_#iIrja%2R*s0L#jUUM8_>|u&a^J!$Dj-09ORg?aL`Q)
zg-C?HpMW+o!kE<OAEV1|^#c%OUEk*CrI#{P2@Te{bFum=gbqrjeX3$XqVdc)dDebB
zU&Kaz;kV4~b;{i9q={Cm_lm$&pw?zR>bzpeGqYHS9%akPo628753qdx7}8PTzWg}u
z6$2RVJ5?}L4Wsq8!JzsdFM#u$+)C~}ImFwl@@t9fXS+vkR5#AHAH<Se-T!y6@8Qk4
z!up~;q8nGzn)@V1HF-y|XOZ45|8bsz!$Ps`HrFUI)apGv@jmRr?3~xX;MV&8=6!=?
zYLszyKpY*k#Gl9DH3_SrIQU$AH;#jr>eFNt`+vGs$H=iHB_eW&j9UG{8R6do#WL#U
zv@5+^FU>4&CX(0_56NqAG)L%SjN7|yMwS~-j}LR~46`ri%^B;gaDb3TkJ=<PSbH8s
zFbkxUBcu&{TlJ-)3=-U>EmE@B5(6&esWB%wBN(;HE=lX;0ZU}y*+d5)MDnClqiPNn
z+XRjYXV*iy6O_X+v(@TMKR)1vBxFz5YEki`!*+5MntBKG4Ono)Yop@eU3e+dhxsl|
zz}J0Ho)&C$g~}U0^AWe3!LLIyNP-o#do=Wo*v`30_7TX<nE|{K2w#tYtk^h*%ql9b
z9G8*lYYYFEAJA(wu~%m%UquIPz5<RZzvF+pmq8HY+zMkqSBIIWs->yU%OEP2wXGAR
zm!-TYgr3QE3-MOU&I-f&@JXWOOE{O>*s+S@%5|KL%~t}1XHK@Vls*co;TRLf!j2TL
zDR6ZuzMbivlR<-C{dQ`?mL;VFfBo}{<-B%(>DsyUxFb+7knM*=h7p1#luDW$Y8&iI
zy9hTjm{xQj`R6c-A8y^hqaj6FydAPyrM8{>$xv<wp6lFmV*Z<P<m;Mdea|el2%RoL
zo=A~`qnj(580!l180xxlcj&4*cUpd=!H)B00KH2107M15?afSkh2!7<a$3Si6C*t{
z!4n<Cz4fkmE;9lwcSPP?Nyd8tnp1ci^9b%zab^ZRct7OVqR70)p*!g5sA-z7G+tbc
zzwx^?VhW$f`CPdw{zRkBwxo$p+qv`FPCfYJrb$Rim$@};=kQR0@eOoFp}o1CixNUP
z4X7I?Lvf@A?^l%T&`4$WFTUmEr>U`uD^OkI-QoTBKNUiUA-`yT4TlCl%ayra^RG{;
zB1fs!Q*1x4&pt-KIe~6fE8f0+_=+sxk)RS6M3CpgE=_x;#x)#{A7`-a=5p}I4E4XP
zfDbXBQzgK&;g5UwWykvJn(;^jv8EHLHX-;AnZ3v@@*5(+&i%g=R*#t`2?sNqpp||X
zkJ+k>6{LW~{?g7sSI;|RNowNJ%ZpUebt4QSOuG||C&X3dSFpd25h~WLyv!Pf{F^xI
z6DWXu?Jyf+YE@HQLPu5XVBPlIsOOzk-Cb}smEh2>n%J&(o{F;x0RSC6xJTK*WU+Pr
zo!95aVI4lmKc_OJF%o*$qOY;Y2(GIXroPHGtBpLqaI{t_xA1es=~kW<{voo9H5Xvc
zrcTYC2Anwq=C=~Ac`^z!D19$@xQ0|xg*pvO)GDe&n)g-j=_jDIjrm^uSK*3hE+5$s
zgYUNUIL=J&Moq)(xil9Ud|bleG4GNz5n5LRv;P$pvUu>2?oq&l;iCr+$|N$EvL!1Z
zrf7JePgZ9d<b=y$r@nFSzh{fJzXES;xq7zNWjloAh^p4mZtN~v8(Q=sCAK-XA%Eja
zg@rFlB8xp<;sFFE?DUK8!s3yK<>FA=J(`0D=!B~48NkpTwsbXS(g*wQVh)Fu7=9k~
zGJ-2MzSl|lyM1@AT3)=Nl2e;!Ent-u;6Ec!f=-vMxkq0l*ql(kn*VIu<9k6K3L&Ip
z@<?{*bpFk*n5gjoAC{(_i}M@d|L>TVYfI}{0RJ45G(9*7R(Ks6J7{H9xKP$>yH8Ya
zIoBT={P1-(w+WfOaLf^2jZITjZ&Fs(W;2kkV0Ccd1NskDqlC$U|J!t}A_~Kc9=5|J
z0T->O_9{(x^@V36GEI-kZelbLS|qiNZI~d(xwDdG3<&21)u;pGUQv4J{x;^)swV@u
z%|)SA5qdYI@C9Ml`?uP<=iNP_?eYwgJOI|}x!Bv63khYF?-v4`^hjI>n!=v__Ip??
zu|Z1r57p{C3CK#0mHNv~D#PtY?r)J$pHRA{wOLAR1829g3~VZ6_423>cb<u7BvSw%
z5N|sTl7jUm1_o(q+RyLo1e&H7TU^Dp;PlX@PgM^rxI&a_acMc+9*WzIDEB|K3+b0p
zrT}d}_=0yATW<OEO3u<un*-E`K6J<N+8fPlI@5j2_*uR{I;hrFe8>OMwda=sPDXlZ
zJ8Gv>F>!#=I*4ddkxA3>OK#s~a?3z36liLE_BX&%sK+!7@rRN^?X|`%jx#rT(0;nA
z*SRzqm0!G((o?XQuBvCc>SObqO4*_jVCu6n^CEI=ML}d<&||KRGL;9ZYXBi3bqMNh
zw7WZ%O4r?Y)~l~8L(;)B&6l%0&VQSl%$!fv+bD=Hce!3!I6mjXND{gd&?T;4b4#Go
zf-vjvv)MZ@(Ha$IFQ3PRuguO0{PO?NuXm?vsz9{LEiD~<_tBrfbIca`jYQ)(kB_GF
z%%r%%c}muC=F<i%ArMlI$JoSQ<?nrWy6DR100gc*$}cJh@;Hnwxo$i6?mJ2<yw*=9
z94t)Oj2LW4mAsO;unEXGly}jH(&zo!RN9x(+|`VRTiXc~Ry*{a)^*=HAtl2cF|qGm
zBiZeFZR0EYbyh2+UWeLK;JIDQzHvwvvTd%vu~u#2`^-n_3`DnQV=9s7W37YzT0T+H
zpLQ)MQH%$>&IS}1t>_)paO)03_=g#nB;*$c%X;liIFFcqj;b6(TP@k3K!NN?q19O<
zm7P_7F3FAHy=2bjAS2+(eRr)OXu(Nv#KtOTK)dN$^t#s81K8K)4PUO&F?hI>jARUD
zSDOpi_zEisqdUPs!co_uaP9uxtA?Jzbh7zgELXki1Z_F5``u14>wCt&0rCOb;EPct
zLy)RSeY%E22bhb3;vWK$4k4FK7f;5v+}A*SLfE~G!ZwynQzZaQ_s$@R&un3Bx&~Hl
z5+XF;j*pK6xqsxv9oz+Ho~cr_s6!-Ux9e?J<NGUp<85vjb>`I3JEBz%8KMBm-b9D?
z{M5h7-?EzKjS0{ox0IU%7F90MiyrP=ZJCLKOOETQR=09a)tP=>cBx|VQUv>LDaTwS
z9OrS1Vw_28uPd~1oT+k{6fg`M+@EH#m;Q+J?ut_DRRAb*PFP@Gd@qBfsbFr*45TVG
z_ih>+ZEGr`KB3<B%|r+y8?c=5yOQ_Von?Q^_S2M{&bIVFrgco<Tm{^ubxLYC*Yh4V
z9wfj8t>m+@(X?08n9utra5p}8zx4F2by6}eP^bzq`eCbXKOLD|daG_$(que}giOm>
zsp858bKRiwfWYIgePal<1A<Xr!oi92uSr3H!Pj%GAo4wR#?Dfksl6&U=h-GVMgFP1
z&UPgO)pYJI>fxvtxx&p0o0IQ%fWfa<oz+`+>C$h0Cn&Oe11HLQ58m`07WWp3#Rc7}
zACw*hCpu$kPY!{0i|si|kAk5Cvw8*eK1eIkwd;{EvZI0{O>1W%;=U)ik1YR~?auyS
zk=~(`3V&vZg=vJMyh-dF20mj=ifArw;O;XQX^DQ*sylaW(hRqFv|qb_)*o2ABVq|D
zIqBZ3Nj;vn@l8n4;Yabh<7MQU@Eb<u8KGfx<3^r@z1MvhlMKWpKjQ<Q(Lv{f^x#~R
zw6XG3v|*1gbnH?^l6q9o=kR>$Cr=K!ReTzTY-!)RdouV<9w6wc_y7JN155xcr7$Z*
zg{`~s)QYOfupBXqBHT-;--#4l)8uOzQqfbmz33*m`h=mmkNUt7>G1LgbbJ4fy|bk1
z9eVY1!tLdKVR}wXLjumf{ulP%X$v?3ELU#YZAO1DZ=jsDlUAmjtD!Vs%43c8JqS;l
zsw|SeC%(e3;Q!hh^kLf5)PCI+q|SJ);GG33dWy`pT_XV^4vM)FzhpqR$~i^z+P5a1
zuFJ}yts<ozEP)4B^2v2Z^PiJ?){7s$7JfY*@408wK)?J*`=RK4A^r00`1u4sEki@m
z<w9g|0WJ9MjabmYS6JWWT!O2PCCq$w1m2hc91|P372v+1?H-kV4cloZ=jnpe@SSzN
z18VZs^&pu_{_yOK_I|A=_+GYKdjmTV<fikMD0mp~)~2(+Q=+P{WSQpdqWy98$@~oP
z82n^!e_HST#55}}58WSy@6}B&a+bkwovvgxhDxczoE({k&;n&Jm%qZ8rU;v(@B!IN
zR|U$*oZM8#*flRVnoylAk-^wU2$N=#lo(xWv*{Ui)jq54X9GXay<2qwk=!U-h;cV#
zwMdktRfK^=7Xl=&a$-gzaqYxLig6a{zkU%dAqW5gq5HUyv~z;NqdSlmKj?gkNr~+5
zVP3ud>lu?Mvn~(0J+J(_MrDs~d#5tqMk%V$J-l>Cz8JBlqY**A@E-Fo>IrO(9D2_m
zZxC9Hp3P0r+sLa~WD?;6#n~d~qUUMQq#g5~ouP|w2Vfl-x(;?u4;h>6!evC7!<<<9
zloC$L)@ob$G=#APt3LVfQ0GI#D_F3s<n8LOkDHbBnz{JUSgc5P@fqn!VPhbXN?-#X
zlVF-Dt25rfL-`(yJsmj0y@D|B%^@QfLv>;hYaAe&_y-#}pya}DgcBn|MmL7#h~9wK
z<lHQwQ!-sXI9DO4h8-M@654;_q07!U9K&qMG&)5#6hm@wsG%TLAOK3?bhQkYj{=8I
ztGeNvvnUt;RyBJ7SzwC4W1KacVD-)deCbq$g_3(vpfK+J6za#l{?lI|J*Ug~s^_P(
z@j~HA2on6g`Wb2B&)AIJ?6nL+EA*dX>Ijs#;#g#E6<@p-b`eP>o4BUSYn)S;?#HPt
zFH%VHHe7d#cuN|)oeukFe5Hu+YWHM5rh>ebtbW}+Dpm$JYMFMHuj>dD{r!9d;ojQH
zhGS+9eH);{swxTF@xFh0EQcw3Ur_y$M11-rI#5+G=fiOo24NLk)<BTI1q`r~-q$AW
z!OWgYN*l&en%xJlV6fi}5v6`9#w7xOv&Icxe<O5s<~N5i1tngXtZgUSxI!X|XHzBf
zzffP}&s>iG^uJBc;MU`AJe<Yx#T&KD`uovHkI}DL=~?i*p!T1VxQ?GsY|F{&fv=st
z0`6&uZ#oK)8tiF-+9SOqGFAw!H<BWq3Srs!r_giymzU<vfHrm|gZ7mQ7My^tYJ1~C
z&jy0mhua{LL1&+|otwC)jym%uVMc11{&sszX*|)$0{Z~u>Qt#97wZzo&6ybVOev&b
z`U^kttq-_Q#jPbHcz+D<iT!&9H>0BP5C7e-oe~q>g4SbBQZpDh4`;sutiW#T2fyMN
zD2AwRl(=a9VnT{vPrQ#y*f{Hea&<AE*@a@nR0${cy~CE2c?6gy;Mbos(KkdWs7_<J
z>9jf|8SnMw^smVQYR7Jc{OJ{5s0DcW20OYC;#X@<h^mxIeAM@f;zD{VByi!4;>IyH
z-!wFc>CWi3OF=g8(6(QfiY;7q`Nj%%alO995rM9fvV3;19tY9)@s7nQjA#7g8{^9)
zjC_EQyiYL3x?_^hDYxi{GZUFjD=IW<@XE<?SjoYo4w;MI@MP~uG@SePn+9SDUbbS)
z)hmP#n7f%I5(+6gXLL)z?66&JC;PIMz~dq)U$CD~3x;Q`C>zWk^-$KiHz%mesUr=L
zS!qi~-Ji9G^>K>z3|3gInIYB0`v1^M=F2Ibvz#e#pHX}t7R9>IaCBC{9iD5Uui&el
zgkt7Gss10*I}-K39`2fc!R9klKYLUX<%R@6V2Z8#J5G!xpngwHq7ohcTIDC~Gdf=X
z>n-7HLq7wAV$h`o)OR!{Rt>W=nlAq<6RGyrX+#S%uikB_@G5_<b}XQ{J5QMy5HeP?
zNeLqUp;K$G4ruB5aT9@eEew+iC8V-(iLbMIxrHm&Ppx!3WOt+W(n8Ri{Z;*WY_C$?
zr%7wikNG%fVRwX}FUML<*)2UA#e(=DahJ=iv^(XPYri=BP&Elvv>{x__Qg}HwDS6X
zf_NX_tsY~#E)uG~#4#u^dV1<zh}`lRIVomr9)A4iD%U{b_y^Qy2?wW<P_HXTlH=1c
z3F>Ik#XrIY!Zu_9R}>1Nx<NK2SO)8=Jv0OWb*XblGPJp{<M6}oo-6nn81HP7u@(A(
z;0>U;5*AG4>IvJ)g3|``Qw$05s|ygGNj3rN@?SHRen1vTWm`N$-exlObucTIn3L^*
zPg>=}oCxF)=Wo04>8p+NA$dj#W1C`Jq<3C2#XW5Hg>wE^SRQPw(6~+(Z1A2uHHurz
zL&eAdSRi@6zz+ceZA)xJhV@|i6$P(XrZ0<+%EON&O-o!Z$EWkCiYHgECF?oCB2WIX
zIJ{4#&*avT<3v^H-Fj})hUE7xo2LAx&uK>22GXf&ruL<r>mLim_cvr;h$aZy#f8g=
z#6Gi?LD3kvdmn1e2L5V9k1ZUt5jE&}F3n4$@dn@QPU>_dhF+9ib4!a@lkWW(nN*xp
zw<IC#<T3?Iv-UG5DI=v>Cb560=(pl?M#-kS*nYIG_g9iK1P)a=bhbInvtAZWi0=?W
zz)Dm=kG;KFXEC|!hzg0%gXzdnar<lCl|)zC;1<<`W6X_t^Jqe8+p4X+n<+vW%W{mp
zeX7v*(M41g>0P9BzlsxOzeP80`J#$fhX)Pih%NVM;>&zfxUJ%yNUGB!a3v@gvT@A#
z{2FJ%b>naKn5*vUzwlLKqc^)v!f7Ru^>@QXJb+#Wi~{8g%d!~oYUV}KzpBRyv7438
z(hAp*&BNQ(Atikv`Oek*h?0dfT;yXiI$ea&Qfe~^fV;Q16n~s8R1_{U^FZpas=W5J
z>ugGzs;jGEai&%KHYX=xeSOj(ag8p)oS<sU;^Xzb%7*s#a*rXGXMH`hqmWqyRZpf!
zWy$`2oyW!@SzVTnGkVV~7sM>|VgG}w{&o967}fnWk-d_V(kq$x<a-(q1H`;A7ne0v
zO)=LyrtO>CyYGoDP5u>H_T7Uy#o_BahY$Hwyy~}K&51P%R!yO~x}$ksC7ge)_ztB8
zuN`x(sUW$fxk_rw<Kr>$M@hx5NBenv9cYy9&rUuadOk|{(Gr)^BDq6Uzxb1*7|16H
z<A%UTc*F#woX&|t^`uZ+4mw7wSw3Hi)53)nvp>RS*Z+trM0!U|2q&S}!{TBLD~|oj
zI%G&qI}qTsxG+WE7l48`RuFE4_7yp!SkyPm5BD}{(qjVs`JuY(VfK4fneu!cgYM6(
zcMWKi&a~es-8i*TnF^^=;U6YtDx#oO3ZioCL<&0rUCPJ5*w-o5g;DZiK1U@>0(>IU
zvSI}}?LP|0jAEUSrW%us?p0Ci717ET<2*SqaL6z(_hQFM5xa*F$`g^KzmZ+Bj#G^d
z`gA4`q2?QvON1%ObVHf}*uAufnKet6D-^Y~52|l=9p!ABZ0`Qzy-{xemW@Ta+E!Lf
z=z-PLo7euV-b`-0Tyt;*#n3DV9ztXhjsJUbrbdBr_njS^@KV>@iHJ%Ezr(Jg3nmj{
z5y^^IvRqb7{_zUwFHtZCasG>nvY&NDv_J+oFHJLPO-Hw!4{HMvfyz8ZB7%*QZeDhU
zHMO-P$&Acrj`wRAR9#N&rVGgEoY<zQ1R@>6J2O<DZW|u0;hgS+Rl^ud%U*1eflPuT
zyL~9URwqOYX>j+qD%+eAgH!l`^b<$AU~+CBJ?C?7PiK;4PT*M+bK6q`zI{_$UI`r-
zX^taOM+K0wQ*Voe8#TYE2Q3yOrnmC&$PcJyZ#vP$Fx&%0Vp_!kpr#3;R;;0EF2uzt
zb8vVl;(fix3Z9$)eUnMEfKZYQfK~Tr`V_6Tx_Mzq<ljlH-)Yl|>Jky}r5nV@%3w0z
zafY0$e$tp-eN@qK?v7ftv~9JP1=J3q!ly!TW@mZlLM);Zva#P7`#X~Tt)C{0TQqNn
z!Uc?w(gDWX{0bTuZr{pBv0`Q5A!gKF{0`9vw43~;tfV4&-O>#Hp5^I^c&+R4=_#&b
z)_mZh(l~DI!x+#+S%+ld#U?KOS1N%3I87^ey_X3+sH#h+McbWsnqz?(K)nI-+puJF
z?o=0+i=d`^-hT?M68th&`24o3I%ueMZ!A-dmbvJUI-0Imxq>NQB(d7`fvN1$9;4U1
zGW*f{&gYHMHESg176cH4VVaOM0&b3l-)C@d#%q5`@u{unOSOf&-a@uLs@GTP1tX7l
zxY~~DfFayHp@~4=M1HQm&`jbuhpD4^WK%b%?pPIJz853Oq?{ZWXO%8G^wjiR=HkHX
z=&vLlxVSQX>Jb676w!DWzqo{a=p3t^4UYCc_%0Kn9-t-z94HDe3ERYwg2LQscM$hd
zpj*)>N!88wur_&p+}h0{t-14+`KbeB8G6G2)>`gLw6R4Pn?A3P>K;X_0Kw_CFl*Kx
zsE<$A+uT|k*n8<0s4tp!%W1%mwKetvEusJb7qolV`373@{+$B`S&4uGGFWI3d$*tG
zl{QJ#9Qko}?$ZYcQ-p7-++q@VUMn69m%9|vfya!_bB#(I_d}?d+VLK^ej5lAurX^N
zb%p)bLdQly8Fys9|3HVU-5Hix%--JQ>EXed`MJIfQ8Z+-i|gJs@zG;WReP_inAQ<X
zP_Q1IEOYZ2&%}FI26OBTbaU+G{&}%cS}UBte{MH78+64@ja>40KvLzosG%<tbCpRl
z1_mi)QiKa`-?}TR{{2l!KD6#t`lZ;Bt7;SfFw}%C1^Kg_x(Dl&Zt$K&eW?6aF6~SF
zADWKt_Qgn-U0c(%LbM?Vl2{3kV8@Q?(KYR)<C%x0A^^S64<Dae)fJmNxkT4&aec6%
zeAcb{qIZ^@7+#`7_EY+C+99M<sjFEyD24V2m>=_-%JM#2<ck;!SDkmW@X42F*bLzQ
zp@Xzi#JnK(7{{yvYx?fBO$}n%Md20YfPh3sH6h?r{|%and+VTKx5f~&J53ww0pZ!Q
z>8=k^<n?Q#4#D0#>3jY^qcQ|wq93E2`TYmQWHIlA_6a_GN&I=&7S~fZhSOP*UQ>yg
zHZ6UOSta(F=?f;Fko0qRV{{>WdOhq!cz{*vZ<JWcuB}DwtD)*x?o3gM#FpZ8)%8gy
zes`{<d`9Y-m$$BUT5U-#Kkrh}gE<8`LwP3$rwh8!-iA%1)V;iwh6sV)fS8F2ZeyvR
zw2ap}qwQTnNi5qFSk~`hsVnY^4^`78J+QJHPEr%2Inx_0iyy5P_&<x25@WCHC)^tM
zr!sZRet~Pa<Bej=)9@?WhW7i#!1Bltb!7k_o9e%r2o`gZLq@2wW~`xHiz`c+Cygs7
zZd!e6H4iuvDn__NW?Jyq6UnXQdfaM{5gX*I@PPHDU}~Q()}NfOy==Vdj*2s<emJnL
zx|Ed(;EFj~*#AKFhC~s`>@erUWDBxW>{eEw1YqFSmHMxR3WdDVc;6~VYOSg%ZTHF2
z&{JoY(RSB(QHZY74W~Y1<Hg0|%`|X@KY>nI0v+FlEmvdtRy^C<?P=rs3*fnl&RL+t
z`p3O!N@ys&Qg9XApXe%~@N4V*8JOL)V3LfjYyS^hO;G6T<#KJY-P<+(f(w#~zUeDe
z#-x2Y52L%H%Mn@SQfrk3qAois^m!KUG=EOE`ihLyRr15GpvHNV1}voU#p31gL+bC3
z1^*(;k)cd+N)TbZPe`W!d7Q&yy`5)0jkCK#dh@PLvCihG2BQC~9_F)NA}=y=o3-HA
z1DIE*^#1th`#Vp}H1zlhE%+!C(%h0Hbg^ocUl>72{rEdWA*Fet!UZNIQFnQDDSE$o
z>4RnZV&q4NVOFVNax((G0N9B&gST;Tnh=K~-DvhPVRNT%Oy4i8Sa}TU^jOK=XQuY4
z%YSSxKBd=0<5HaW?@pr|=Bx)M^-Px7%<sHtETl+0J9STqgu-K-_}cpYqxLoNu~g|c
zI}>fDUJEb+c~~g4R90ep@R~jt<_jRZu0fhoRhFu8qQ&<|jLnjbUI6UpQ=X}@w!hdy
zaV!X?#I35G3|x9~F|}H!9vKo(!w8$7DxwL~`y_f#s>oD`if?W9rF-|Ai}@8E$hRAj
z%w3MNYL3pn_e;_+a~h{Bdwj@jo!80Y)u{KaK4RVVKsqEMV^o?ZSl{f?+a9d#Wub+}
zT}Pq*@Gr~`ZiEG2m^r<W<IDHX=W*0Y-7B0f!@<<Z)Z!H*C(d$gZTBAQc<OO)MvWiq
zctt7mj=aQ}lN8s$BZ*_)FqlH<&w57Euk*luKu8NX@FyvPW~C948Vg)@3g>Zer@3Hf
zOX?*2tVk>__Hi$)PzAVwP31#WHOMNatNrDLBi&B=iMYMHdp5CTm|A^+Rh4~_v%54p
z4v9%gRU%D^s|IxYpsv$1NwkJb=af6vMq6X50PaJpp}}IX;{MDwFXfRG=gpWtNGZX%
z6f5neKC-CGqe2#S^6Rr(5viPe)AjkB-zWN&Le1lP()KCW6+8l4zYc=)2kN3o^ONuY
zE=oI5v6rSFE=WtWkK_vzg!8zz=Y#E!>*{N{G;IEXNU55DN9449b1{UUlFYA4O2;8#
zaB-p>4~?hN$aGAL6D3o>G}&X8rPpjPBKw|p|8e5~d>b82mHNl4A7jCsg7#C!4Ub=k
z-T32IDK)vrc(l;&jgOyGizP3!9p|CTJOQtj9ujoznt0uxf`xPG<qU~7qsx#vUtpr>
zO$P7Z-r^&<<A!dvNF7LScmXc69lsDw<Bs>O=Jqi9hQ($qS;@B`;v-pt<9^6!@Ec+M
zdVMc1eAQsc{OGTha4KRli*8G1T^XDNhuFmVLJ6g>k4=Ij%D+0^$GdUAcC8cfF99p5
z9VMO%UT<v^!gUuXANA6ZeTNd+kBb3sWaCl!LFdC;l-Z{Su!J`|bB5n9^&?^4U*z6T
zcv&lB2N%%ZX;9i+8VU(`P3p$h{xq%HN`!Jn3L!ecsk~}y7G64-emLd6Dh*>|2=acs
zl0FND01DI^tDB73iu5;!cE#VZwixIhJwLl(0vIc5zT`E`kh;jw7arE0OvadMP4878
zSVp?6?4Tg#@<iGmE3k$}BFbIt|Kqlj7}#XIK3`7uyxGrO@YzXX7yCyv-+0hd-HL(a
zIly@--c1>JAAR|sN<DaPo{ux`27)B67Si~rASDZAXn5vN@h9J3|7@VF{4KuoDY5c8
zuNC>o;IX7ywFM^&#?Mm(F?rEhcse;s`V}u44~M+%rXeTpB91}ru7iD^c8~P++$;qx
z^3hwRBTX$B+`DoJF}G#24HxAaLb4!m7ni8Wl9radb>8Oo{v@1cDK0QrE5FXN^W95E
zH4J;k`W9ic&4FJ#&!7LIo)i!G&^9`y5*pOl(o`|=PF80~;`#K;$v*b<R|5$;=W9po
z<^9EaA;R}7&d*=#$OJb?C3;?+gSAX(muMMt7o|mba*K}Qor&#!Y$N@DQ;O!(Q840x
z-xtg2mVm}9SD#VicM3hjKSA1tS`(J6jEH<JL@1$j;B5&w@k|PwoBq5O_kZ>bed|9q
z(v%)D(Yf1|BPsaNql`>IiI4T}+{fR|ld3)i?S}z<-`v+)u~W2F!l5aHFCZdTve)ji
zP?o*D81&PRzGO7tSKFxBB`4R#=5Dj}%QmMxZ`BMg>JI~&{@O+yOdXYT;sWo(o8W1P
zj<_Qw4S6Hg(#!S7W{x7`ebP7&@_x~zbA{9noWVOpll~&>3p(h6k_f$CSfpOkP=E{k
zN2x1EUVlK1qV7l;yUOJOvJb2O_HmAJ|6)2Y{h{HmgZt%K^<OGBp8)r0bsSp+xv<?a
z1)GV&S&7bi7yL@*Wp)WSDg$=?RK3`lL_*QsVN+8ec&YgixdsV`ESA|ltLqJ>>(RIO
zr>{S|amv0>{tz#PH`vd6){x*yXW|qiixJ|zzP{DgC*da{+7n10x&Ow0FL?STEo|@J
zd^1CiXoSb*dxNslhF~oEi+tlEO|MJuVSg!FOx@@Ru@di;wy2F4lG@=>v}zXmA8m?d
z9<?(6jBCL0@=wUi7fky-3s(ouTvNOo9Y6R$`Q@lX@6TNX*eLZsapvKz3v$S+BgZk4
zccV1+8Hz*gMGO5Z8~b{Ci{gW|*ak`^<8_ZyrZ+RchF;Q3{<#W-j;coL2?ry^8~(1g
zgn9eTZbS(u4Jp~vE~2^&;7qgN?2~%js*6!KX~9chRssO%t327xqvFi4OvY1uvY@+{
z!P_15Ve?C@;Q@hpzJtxZY{7=6*b5(x_O76P$~WFf(-q4<bAj;rcp3qAeK~bV63wtH
zXoJrRh+{_E{SuRLzBiKpLw=MMvjIh6u+ckoz+eb2%{`)`BHv0uGtP@RNbmZ2Nq)*c
z14;hubta0a!Sb9&Epl@U%T`c)N4$%RV#&#4N4ZL+sTQYm3QK-o&e6BbJNx#_nL1<S
z2+&>mN?a3I;2WA@yH~$wZhXsa20Bn&bxVX=8Gt=WzxC>V+f{+#?E{;blpDfWMyk6E
zE_XY+2&|c=I!ZI}UWfpLf0TI(VX?$bIm_NhV;{Y`7#PB9#`eo+vk2G-ZQE14f&xk9
zm3%b`hu%JTEaEO(E~tuWVrJQgQuW429NyOrQ5P5K2W=_o<cS6A=<rRaum)Aa)l;CO
z*&sa_|1^ga($Aus7^tjyj7H7kN-aqgO&}~TL)**V9djmc=5-7k5_<nzy6=QNn-;}!
z_l0JtC%qBvwqIk%?Rh10OZda7=H%z{{JksczA_NoPtobE7gCTiKXK5WRJ7<C-t7IS
zg1a~Xmx;hZ0RCBwqhG4`IRmkf_qbFt4^}O#zkW#CA`MqS$5@ktC=}?v^_^i&t@X)7
zDuyvi!hU2ljs7g7x&-|bevfjdUfmS<vSL_2G%zH<gn_E-Ai`jwKY%-JA#D#9_FCcB
ziU`Y%K!nv<=jmx%<q=v#ko~sZs_6~+Vfeh^)|=PzThgTbO7m*hz=>Rvz@sch^0<z7
zOL!u?@;~Er)gi>DpcjvlP9g^j4-7{|6PD)R5-pLIP}pW)`ovIY^B;dc^lm3!BWq-`
zGt~79NeoC*3cgt4GLkSu0BxcB!<;9QSZ$)8NvXc|AJVmc(=K7b!3)n6K@a+_wb{Nt
z*x~0#J@3rvMme%jNWR?Ci;on%@+UFBCrcL8KsHvIt8T!nzD;9`UPQ6IV*XK9-)4!2
zTV&w+P9IO1$gQv*yxu<70(xhRbS_r`48912(kE}3?OT<K7StI<TrO;vEHKt^4!w4n
z4S~cH@857g0R1OeZLoi{i35ks7*tL_-CKS?<5BaDRy)v{KzbN%(epnYg&+JFzUMG9
z%i^M~+6P3d#LKcGH^=^6Fr9qs+|WH_ZWE{IOss$mvCdqU(vjorM;69?<aipVmhUs-
z3DPPsACo(OjG$!;Cauap(f3oV!aRNMfh8W`LYvAVzJT%blLua_@AdaxlJ5(aIUrj!
zKKfW6Y7~x7U;Ud*tGP~YrsET|h4v7Uw}8&h3FI^->@>uwWbnOfHP3wId}JC)ct4;O
zRDHKxP+q$^5o$w4;eHQoTCD1i+4fv;u~hiZX7?s$LjHHP$?981ux$=eSMR8z2i-67
zhZjfDVrh&zTU&aOQT#3XDn%t2CzDskXOw2X9mFUcIzFWl)F=an#kK~fqJHrguOP*?
ze6?Z%C2*8J7^`SD85S{rM(xcNEz|GM_MJ(UjFGQ77}d6Z2e1^LudjcA$zz1mkxT1V
zziyIL`br=5ab9PQ%&mJ=ndRjQj4au9VB1ft0j;>di|f?yM?yVica$!I{J=O)u)nQB
zxJ0OWr`(N%gxNalv+E+4Zzy-_8ewo0fo&kcJHQ2Jf!*z^(OX4lbc(gyIP%A~&eY9g
zHr3F!yNQ>!mUl3kFRVQGu|n~git|Tw!3%s5+S!6ZQHSY?)1%xITC}sA=TNE07x@aP
zj#kZbgShYG&r~F>_T1pO-X86Dw?p;QN6RfP@LZvk$lyu%l4qR$9$ha#l2NQ%YV-Gx
zPn!QaQ$gI$JI1ak7eq^gSF^{1@xt<FJ*Op0_?|>5X5#ZvLtrFo9b;os!^2NQh^vGm
zx6f+SI|$dTE+s#93vcqM1-mpGx&SY`-X*D=EflW0?}#I8eP`(cb^k!r*xz2ED=-4X
zX?_wEwqzFwKIhZ;E>%#Z_x3OMok@*k9>KEE9&XO_a1Eg(C8zsc4Dw$Wm0Y`p?d0ty
z*j5u{52%Zll*6q4G1OxE*r2aen3KjuSVWu20F$vSnh#7O&+cZ3zlf+#?eDdI&vQ~1
z(R}{vl?ujKZTlD}T9cou&N|g+l4yo5vs0MqN#};1=;=pt+UL5HOm-VHg4Am=S?hwS
z?HYR#PS1$XcQ+(f>^m`Sed9B?!XbuiE+hQAQ`C~b(=!hiz1hj>JvH=0@;}49`%o!}
z4KoJ56Oh??SDn(o;wQFJPjdMPozpa!RdZE)%#D)@^#lQ9J<I(#{z^~#B#E;YOq<`;
z6I`fMZbc2A%7h$MJygwT44+un2<+5Zu`?t@1gG^vOa{2$WjKjqE$2Hd&6rO}vj;&c
z9n6J=mxg(I#~o&IT}$q54K#1cS4vkFM@kzTGDyJJjnM!iGb%rFAhsqaHpniEPV|=h
zNaX$vb>?Nay()kU6<)+A%K4xxoj%mn<V)l))xmS4lmVMKtz327BTkdXY}Ey^cd!ML
zs6teSzNh9B+i>0z55;>@N{(nfNiym5cfFx87&^_qBaYkEYD>pSd_Tk0-)MgclDQ+#
zp<qqfrHobj_m=wRG_gHJJG~DoRrJVFJQo3*!nt!^l&Gmm&JY}=NAjq)oU|0Jb@V)=
z@yF*&WxiDQY}!g!#t3E{u?UoY(%bkn8zQ-CLGZdUgzJNCn<H+8Pd_#7hFp3bHPI(R
z?%nyEt~s6toQ*R=Vr%SY$%dOKp@N1F+|kw2D{NEF>25C<Xfk;*BhfQ^Tq&5I(WMSD
zi0L?}2XUIlI|_Q2zuDu(_du;?s43`xiSr(1rlgx(&c>RDM0LyXC0G@UI3AazZf449
zzFxNH9m}%`y#5mOV<Bp6H87QYp@8K;diL$);aCs5jSi~bxffaWV_6Y!(}c`+xK**d
zAln50fM%P_@o-~3JGCGxIcI(6VmWI1zWSD!F=*eN`BCc*zC3+1U2>l`9d)5uV$kZr
zr&c<CD$R%B<;%Sz4M#c8Zq9e2lvrrOLTmaHnR3mI9q2l3c93_@feKeq946vS3RaB!
zGDzl1v#A@Yfn$ToHu;DAY9J4Gw*1D7MLVl^<{5~sHrkon;C<=4{yQbzzT@`Ze1W=0
z6%PiJgnSq7Jk9eJ-rE=LxzOw#&wKa=0#OrS#+A&ykZl6ly&jM0wK0{OJ$uZaE^qRi
z&9C<)i;LAD`E2TDk1DS2OL=T9KHf`T3}N9Ci}1t2@k*6LxCsyiy%I1?#w`!NH>5xN
zF?}lPZ7ugQeX=Mrj9}EQX&dOE)mV&0ot^Pg*+InHP5V#k%U|_0<z1}5(YiE(h<loG
zzY%>H!(<M^dz21>WU;&X?woD=?pp{ja-yHBSxDjPoD;_)_a>3w^*np$ROf524Y$8N
z^*g3-f9)%}rYVX&_=LWIA|^_fF$dr9rD4xyTH;%9o-G#qWMPedB-}(#bcJKD=*%>8
zMcfk9p913_$ml%Y2iXWOs5kX(6O_~ZG|H@!(feagggSCi3g7^YenVBf;;?B>b#k<N
zc)ayp!S_Ud(0nhp(42ohX@S(W-XDZpi-BIcg1`=&VwdY;5%N-6lV4rK?ksCPK#fTh
z0B4b;FWiVj6Z>plREUVKsx+_pa=c5dYHqgaGk{e2S$8Op(&*^*>X>47pGxZPn}R@}
ziO`&d3|=k_^Lvm-pR!(x!y^_9c$g`Yn4c>e@D}U-vyHA&kWu_jU@IGFgl*bvq8lhl
z(t9R8Jv}Lu%L}yHrk?ytvR!(TZDG|(l|vzX{ZGPl=0+6a+71>-Mp#Xfy(C-kXhIl<
z7><7x(U!4)&ublli86A8J%b^X+4|Mh-}lc??FKS7_P^3^t_fhMACK=*zVzl}pJ0E)
z66j|nevjD|=;CKyaBQ^B#2<#H)!DN+`Bl6bkwquq#HQj!yz7NgwX=4}fIuhD&7z+$
zk0P!y?Jvd?C8rcLe2$yP6#$&*y8l9j!j`MKuGBZ`_+d^CTG4RuTpeNfQ`GufAJ2G9
z#POa51|2@eMrc`pYNv@&N4ReP?5}n7YiE-UUQU^v{cz^U3k+bMF?racTI#+Mvwxwx
z9A@1X3&<+AIllM#tTL*_E!UBz7xZUdBUFDYH)g%h^-^2Nv`<;HOPSR+a#YJmF2~yo
z2>SE(VyI{_pwfw(FJtag@_DM`V#46U&BOKeNF<;}&$eI(B>VRj*D?w~P4F&6k{dB~
z<Z-)vb5a6udstAnw`v3(Lo8noFD0*Un;AD0@rZ3Y-S$Lkc?$SfNwSEAPb9?E(jSH&
z<$9bgV*zYAHIsN`kZT!wPLX0lvAL0m0hij10;^zFw_E$~i90UXSpw&YTob+I;e8pV
zvn?zXm(zKi0zNlR-`9Wm6x>?oCj)3w;vS;WW#yurbd%n#*~N#x`S@ZN+#ZW*oe6xU
zZv_`;$DZgJZjY|T;J{!eU!Cl5XRfoHFmu40WCle}a`WTtDP63Ov^{QN#7s>G7TAGS
zH}3Krrjlm-G@&SiL}s8_k+i0GD?=dbQK+$f0Zu)q==2u~T|KujNZZc-w^!9eH+4=m
zJ#<z}vvmXd66L?X$a;}M>fp4L{Fw3p9OjO2qfV(2&cctLB<NJf)rxAwLm3)WIo&{K
zi{DMe;Dz%_4zYwHC^+%(gS<u0iQd*H+o&1v*1P%Li=T5LeMVpZgy3gB2oURIVd4Fu
z=sgc?8f<nsb1TOw4m^`g8^}Q?$*RvDGe0$cs>fm=^nO1p^Vr!;;AD$h2l`7|B_lR=
ziN!U1LoL}r7E5h`LTDkNg2E)MgsMxH1h1we4eMCTI-oz)<MmCbe4U)kbs}iEtd2Xj
z!QKDPiB&S`p#05O_vflf)UkS}gl0+DE1&K-Ib4j@6%Y@;@}=LT0@u+3Td!1JXk4%?
zp5+MzXWbm3w*qsi=iioB>Z+zq=9QDmXQ~J*IwVa9@;5knDa%q-7K(nGO*Z4s%X7y!
z`MH2eidT<$u$)_&tnW-SJsB6khqknmcod;@0Hl!J!UyG-A55WqSucYAE)<W7HN%S`
z-J}sls<Vho?PHQMUiKF`&Hr(zVP5zIwGf;^`uTnS_VW1TI4E|$v%R4~YgyO8a}M&b
zEudpBslSf}5K7fEbcDa!K;k@8pUZRLhbxVItx1$-u2*)=cUx3q<&T?D;3I@A8T9R{
zZwbWqM?WkQBE5Kak-b9!fY^9julF`ZLQL9#qebgG29Jy0)xx(I?Qr>p#cvreH-Wz>
z734(<zr7%ru^z;+%=%gV*>-<TPGo@PO+D7yQ0+o|@bBNf^t0(nYvD>6kenD~y&+3?
zZehgGNB5uVFex&dH|^AQdL)03b=o{^Ha=0zzL^q!6Jw7fXtH0#q8!>4xb^X(B(Mo@
z;iNv)fGW&nk}bx_GSm6UIMg~AOLYN<FAsltd1bD9jn)fg7R?jS9?(|KsSk|M0y&{Y
z?n4nxH76cR?5JO7`OB;*tS8875Rp_?dwwV#3r&^+I6?hBB^vJ|6qiB_<>JTvZY+<+
zL7o>G!ZxLa3*QDA@5=Q-!W<ya45pC)UmSp#ST)tJ*Gu(<$kybAqCksU5E)&-Ut)dH
zrIqZxmn_LXy6#F4BFf^^IXj851uaui+V}qGqb+huhLR}@N@@M&*g9XN1iM3nCpelY
zjVf|m-wY$N2uwGrPC5#!fnE-oy8&2y=%SG~S#fe-x?J0EzC0JL?yz&~D$GR9U>%#h
z?TidZh0>N4$|5F&X)GKivcj^fdYzf_c08o$-o~ahY8|WS^~&!06KvdoR)|*H&X38I
zMZnDAQ!l(CsznQ^2ow7~tnZW$tV3~--2uxHD;H?6U+XTVsWhfHhYYv9=Im5v85`7{
zeF)#v+3x+Yq6=6l)6Yo1A~bnjUhZoeE)zxMf=Hi>gv_3Zl`8eIN`2gv`KHo^dqzM4
z?liSS_5^AyL!ON<QyEdV)G{R|eAlkO>BCSEssoA9jTY1ou+R?0q?{FIi_KeyzEK;8
zx*9Ma_V2!Er+x8L;tyg1q?_)^5+?;mwv#v9i7BA#THsyghqf9D%eeL9nYR7kqe5w;
zTN%I)J20vfC@6Sv{&qV#QcQS4TI6)D56+RqA87w};PzH9PImn7Ewbt?o%88(K^Q*6
z2zpy)lEW2&wJ;t>8f2{B@g83yf*gCNH|S5!xX_khO@$GO2F*$m^HGZCf{wvd1=aoN
z@AwlR#*tIhr!^6Yw>Xq(*&Dak5xAncSK|=3M)Gc7gR>~msWDT|SMryfO)86or|TLd
z`G>N1H1iHtbc%so^U-QGKF)UuFHC?-RdyuK&-LG{pp1&(WD1_05j`-kdl=kT2GC$j
z#fV`6y0={-%f^qy@4~xqt{RJunWMI6pjiiujktVV?;0VNhKr|?&EpS+^#y84(I&G#
zCvU5zMk<d;Rd4!4DJhGY?m|rX5q1UG#~VG{8<p9@1Y*tR+fKTvo0jJ>?#@?0=d<I7
znHV=k!RC-5Qt)HW(<Y>cZv5M(q2b?WV0(A$tka$RjC@`zyM)=9nrsEm$X3TM{)%@H
zxh`(yOnb>8kOI$FQ(6Qs?g3-p8mhq!L*yW>6}gvpW5Y3Nh;!0V8hMXy+xFUBkm6^6
zl9j=H{+~3~R+GVEXQ_mkY+zS%mG}XklUC9;<9yypW8!iXFE*xTstM^X#%%2pcZ%nx
zTLF<vIeJMi`lgcgdf2_JmZ_M?B7fMX(^Mj_!P<L@_@CJ|Eqn>f{O;HItGNPz7k_au
zi>+VW<JqpeKO@b044^PfK#95`>0d73Oq_9yMK)dgy{F9@@|+)QJhiZPqNOD-W8`&M
zT4@^cMA2po$gex_0k^-`Z;x7a`DRPc+jCb5%6e8IFXN*m>33CD)}<FU>POxOiAqjP
z1TB7Ulcd7EsQoMZLv$%Wr_Sx%+q!GC5mb1!!APw8xCeQh<j2S1f4p09d6r=_s=S?;
z%*J!sQrYaaxjMhJe|&epe>q(;?e1_h&S{!N&Qkt#Z~Hjt|Je7m=dCq5sf0|6HO$|y
z+*v^0Op)cZ#BX~zZJwzjbm>_1&p)$)(Q)gCSpv=$fB49e35+M6@G8t{T??`+1!t=2
zu07zP5UAMyp{)^qxzhE;I1Bj4wn@GVLHyi*;;qOx+fOwH>)8Z(xZaAB&rEkiEUZr>
z>K}BvL=t?sZjD}w!p|||2*|<v(XV%^Vw52G=qdD?K6AU_H+3y<l{6uubMaZxulobo
zW2HUhCUd+9l^KZKI)o`usg>N*6Pea$Zzs(4S~~OSyaqK{An{_P9YtQO91NF#0s+GP
z9Z6$hTkG0ptAT?L#1Y4~c%ZkGJ2V#S%UnGC^I5F7@mw!q4qd;CGG*K5(4{ZkG<5I&
zis0xAUcS9xP`IlvEDg8ld-h8Cj1Z?ZbF0rt6cCd)5jyzpeP#H1M)^aQ@-%jv2ThFJ
z(3iTg5oX)gk<V1GnMJp^!iq3LY4<(_3znBZQnr~YS)ir=LK7eHX~x8Mz%NF`L(`M;
z&_rP}&4Ws7dgE0gb_n^9s682*t>(7`2Yo}Vq@~Z-bIeVnt;tdZZu8yX7)&y?zA;Pv
z<O}W58kyOcHe<uRMA%-z4k;2|<b6DQzqYgv`gdV&H>Pt@-h4+~@64b64*>W;2ft&z
z#(L8;_)qcyn$WQS(!0{0!U{j8M_A5<b}Z;ox=HbZzBbs1A>Injg&~({xvnbNg_HUY
zV}ZhHibq!RHs%35a)i7iK3pW8(urjSU*#yJmA<ERy1t8^{I2+I@T1c;-O#q7uh{Wg
zRoghY^K%zEZLqP3)DOyrw^t?=dg+qKz{`Q>V|^%4eba(3z{HR9B9%RTPf$mBjrqp}
z>%vJgNmC!Fcj2aA|3e4e;N#SadK)FA#a}cKX}~{}wDM9FnWzz=Bvr0@5RAe@DT`2I
zc#9-{t7>|H1y+caw}yc}@bV$&JncujoCphq{VBXwc;Ur<jmAcL?|Ihm;mu%-rW?GV
z4MOOloCI;SDdu^<9s=8c>*{fgsk$78wLODViD;l9ufq5xV8&yQ7<QveOQ9vv;|#^~
zyp3rRue8ZeqHAB(7BPP4iVM*u9P)>B;doZw){RwCc;N;L;Wy<coVs*)+iHX=F`S6Q
znmlnJ5aK6Vp^R<DV_R}Cf2lHEGB^R!X#wDRzh{hjSDGcie9=tb2}eW}=e|AW@(A8Q
zh(0Jd;V9j4xz6zlSm8Iu&A}=R-qM-#Ox|OfJPS|Z;G;sxsPIxhrpL4Eyh4_2p>N&f
zZ7J<8aIwDX@Q}}4m1gBvoQ68wI9%x>tSheQ#CS11#SitjXxd9W9<Gu<ltm-{7(8gd
z*LAp&TYElu&ihZOtB@P}j-ONeSnKXbxQFJ$AN*?dy2UthWA_cdv=7wqX+JLP`Rixf
zm2&yY4J)^g_q2$sGj^o+ba(b))0uq&@xy3cL<SGuq3wL5aAd)q*-^(j5=~)}mG%Z)
zP87iX`1XDE>fIafU#L_3-B}#)*G7-Fg=f;k5znz+kne}xMa3yU++)VQLL5=|{kz?9
zTHS9-wmsPWX{YC>-ant6F+V<WnxsR0_T8PO)#AKv89myaJELRU?d@YgOaJNP>c@R(
z=`LPhxV*^k>EXv5v?zD#KSJAOBrJ>}2KDI743WrrYhR9`Ki==`nX%F<OQ)yF<i)Xe
zc4zlKy?pz|-}$*PdAREco%}!l=l|j1|J{H8->&|bzyEiuiwmRtAKH!|ww8OssmbQF
z)m%rNAPmgO2prEiy^|hLNq{HH;&ooJ1KN+J$(<jQVVEfmy~A5D;7A7fabw1F+zE)6
zVDC+El$mp`V8c%@za|_fG~V?rx^Y;54!<dG9Xu@ZQpMD0XaaBf+p*hkb-bi6(eQjS
zRMMm{QGy0em#8Y;f|D<JhYcue9p&OSVUM=(9=>R+(7X)!7pTUJyo{aF6o-xd?46bI
zRhcJ^ltmhrZE4y$$NE9`2P&#nSRB;iGTu%t4}sn!SE?UfNJ+WJX(E!#YX_L%2hNl^
z=`&>!L<?<+vO7*kjIVIzNvuQV4*3Ti7BZz@;vTFK<|1IqRNM}5R2=Z$=vt%b%Jgy%
z5&O@xDXj7=D71~FL-ds<1NU-*oRTfb71|}>VqGXbgbRmf=n9{5=i*qnrwagcF>YGU
zz$K3bgAPd_wCWt42%Sl+8(_f&t13bSZP6AFkFcM>E2KgELPu7~WclL>G1#CRFvVp<
zn>L3Il_+B435?5Qq39*(B&3UHg-K^nmoC<MVVvW<Q#<3hAee(Ezoa2fBYJ?3_cgAW
zKp$x72-@HUsW#zhy^HDcELWw*UN?D`=5bt~)ni@&002M$Nkl<ZVmgumUO2_Y_#x)$
zFA^X|j0;qb7j$7EsW+(wgYXa|olB1?TnzhQgE4eV-6g+I$^R&?gqzZaM(9UE+e$pJ
z7<JEmPzFTEH`1qg`ATnq3Tcua(3{}!angr$Ko~KEr~21BFgSPti|zt`s29Y=jyN28
zOUs(#{Ck>^WQU7orX38%@4(ILAiCwB`J(CVrSb8Br*jN2JSyGN4{1SeC=#di#&Ez0
z9vW?qv`C-O#VH-(?dh+0#JEqPHX$jT>(7XA@w#*YwB*=e4tqjQob(|%((nKk+6lob
zOg^O0ed<J0;i9c{#c5KUVDXiYJVQ(6!{fZI=7RYk8?cl|^2g)tUF8A1P?oCFU`YH6
zrDf9}g|NUk<a20?Ls>F<Mn6txn!dpK>2KO^2|{T`D!EuEgaJbuL;6FV3*i9^iK<ks
z^Aa5D6!QtvfxdB0!veJPOStHbNg~dQo$JwGGCz`sD9Mj}a%mVTBRa+dHiaXsuZwBN
z&ndlJTpSiZr?~NZDO`+`dlnt}IT*s);_L9xeJIBp2xyTXDibN~_}#jwHQa_mC|fpz
zU)MDUf)S27Vt*P}3CP=U>TCt0aBJvn-oY5S!uJ9TnXpjd;~x{4U5o1l2o^CA)SS}c
zXwKJoxGaXGNQVU8WTUX!W021y98BVekf-60vj5&L6U!LD3Gr*l$m#XXpkNW}F@6TO
z)YY{wgv+T9j@u3Rd3&ddX?k~gO{boIL-w?wFIPqgfGjAQ_rilL^Ym4;z{4RyXm5io
z@!Eybt9DQF(%wiUXlM>7c*0Ux;*e44i*+Jc#fgq^peLRR6D+u<4PR6PMow)h`K@sX
zajvHd2pZ@rz=S~8n+1v%@8IM3Rk#OKA&MdF-sXT1rQc_4Bss(vp2!Rx!s2u&%y;%8
zl)=H8w*2wY-aW|DOMt)Hk2&Q<sz>6`7fvwYO=$u@^tYrDP~?&R1snUWZ)7tUepjno
z8&jWGi<v45F;4tkiL=H{d8O|$ZqqYzNtaD<C2`0Q=S^Rk#4E)|2hyEj<g09?`9-|w
zNWQo%BCnSlqkLlRu$;=2@|EzaXnC0sP3e#K;8ZU39{ikKF@(@VovYdw13)72kgs$o
z-y{q4fWb*_@s;jk8j6D_ZJFZ07ftzu#ZKY!i31<KZRl?54{$ieg}2Rb%v;a$V;&^0
zszz%WH&iw<Z0cv=DgMABLui{qeruj}Z#!4UbZ6%mcCXU4ll&)t_NRU?&R5@gg6U;b
z*5Z3^Zp?7;GBUjU_TT@qy8QTh^~#J$XO?gb1V8NW{n>@n#i(#)ry87F+xGg!(_J_4
zBMv;W;GOMy!Noj*p}YVix{Algy@9_;jZ>KVUc!WDW*0YH5%%Uq;B{+Z*CyjTlbIvd
zg6;#&(~2hWa>(=6<ivkvMv?E<H(Y(aS^bj_Zw0dY*>+N~2F1AD*AXm(-<!VU^xj&W
zKG?Xx7-Q+XzP?-C*|E|O*FQ|B_p9&Mc9q2C<?4medUN^G#|^q#0y#<7r&}{RS)oBD
zo<A6$I85@ch08n$57;lCzOVlHSMS$QIw}vn{`PPF+Qa{w|M_38{_>Ch1Ap#($4^{g
z)?BIw3IPKXx=|m%)Ku6MujySCHHOipt#Q~fKR4-o?&*&6Pn|(3v(Ab42}u;Wi*(aK
zhXu317@oMg&LqTk`;B$b4YM2c6#5cP`ArfjoxsL}wI-aQao*rF$Y1~-xw<lijA0OO
zlS&(uZMw=ZFk$kY(u&9qdZcK@{Q44W%Zsu2l*rd6OIp85XwDU;Ev<8vyNg_PZFCnJ
zrLgFemcnP$ELIg(4lRD9e8$cnwPz`xeZ6;;CkM}h6HTB7ulq{NuKCg*F!-j(F-YIV
z-7iz5v8jwV#YvVXn3A?yRhEe|&;D^sRdHE@*201K4fUJk9ZGvFe3h+~p4zE4z2@p^
zN^6YwUAFCG$U6p&vC~YM_x7`}uOba4`AfUZhYR>?r`0IJtbYtOj*3^ocqO!2=YEW5
zu#OYMR#XAoqRT6Z;V;k-`3l3sf$?40`jE}QsrxLddpPkGBII}sAD#u8;A1EuE*P5v
zb&j5VU?JUHFd!L>wI_V!fIKXYr1L^z_U$~cOlM~p*hp(g!oJ0tjQ^a{*i@h5`cC^x
zvF_3*rL=t>S(R}w$<_QYlxiby1-dlrccDWTCRwz6?rs;lZqhO3ka97{wyR0*DdUtU
z@~7mrPugmO8262R)Z4|y#p=d(ci{f)H@|Ve>(AAL9Z^;pP)cxRu*tz<B$Dwa7e7`Q
z_X$mLm(UpUVR+#_qTl}A?^c(uZ9Me)a`o}$%i?}v3>khl{rnW&wRH1%XU4u@w9DS{
z^XGr@m!UtaxVLp;ZIzaIU4U28=&d`@B=zgc13d<Sj2Jt*Hd*1_ooc-hlDQtDUrbv%
zVyf}L1w33lC~cx8+!TiJ@Fh~9CAdKS>2Hi*0*j##mzI~9&QmV!Atm`s<7_jHVjaO%
zX_7BVQhvl^g;&u{)05(h&KQ2m3z}{GVjP~Scl>Aa#XBx8OCe<QtMpQ-s0*sIgyVRZ
z3~?B7&__=dsPhpl@jS;vY?qA)97u-DKjd>+J<P>-zJw`GM^RGuku1g|ZeyKqPH#-J
zNiT?4>f$u#aHI#NMSSD%99|z;p-H_*PjxOQZ_sJL(4toymZODyX*-{zo#MtYZ9Gz3
z-i0q*y$e=)<Gm_oimLTvoE%Wf0~l~{%FpoD{+sGeyi*?WyI@l4qD8p!oiOsBy2+#x
zV<{n$vyLu!a=-$IqYdDR_94_*+ZoBTFKo4^A3Cb9;w@Hjo1?U-4RYqykMMXEE~!c#
zaSDf?(g97qr#M63o9M=AhmT<KamthMwuz6^i;h%~$>Z3?aD=Dtf~^ZzeC#*DA`7C&
z{X&d8<XO5Ajd*cBCsPV9yb?c0uLQ3TOuDcm-c=Fs^4gRZz5=GIsSd?5sI88}covU1
zTw%U2R@G_ZDN|m?9GJQu>Z@?mIuvk1Y+R0&2JsO->7>-cBQ3vz!3Qt&TzNR?P!(`-
zzatqG9%wk&x_+gy<;IYr@99U@<inIevHcfbv3x<UfVW+Hn`!0fG-Gs+P4lXWLJCV>
zRN3KrmX4tjKR3}#<Aev@oUVw+rR`=2t^XC>7(!*9GO4sdJ}zT|&-EYFBMc}ml8_uB
zlAK46vj^j2{~aeJE`oDBEB^r4pLX(WGz{#vRsfA*gYkv7BG45L=V_?xct>=p)vc`4
z41XDeTqIaX*_IBH;;;7OX%A!qj735keWvq5(ULEH>wRoLo8Q&EQM{Baz2lP{r7=e%
zhMoGkPDg#@B2A<%O*1@bFVpZg9>OEPRga*fb}z(-vcNMg9@h!qBkV`~;VT-713So{
zmsN1`W%!|u_Z@TQyFn8=_Sa|Zg1oW(Aw^a<#I%Pmn(|{3b^34Po6<_(V;*`gfe~z-
z-$nzur7OYW<D}R0ojAerT;gJQo+*>?rJUkoy{1M+jNFwyyTS^1k~>hy|3ViRmtB%4
zz9F{$H<nvaxiG~w`LV83xE#$iOgbW5ZG*y6dbxMO%Fn@+gvD}67Vyw9PUR+--Wb<`
zlOM?jocC824;|W!v5$Jo8GcB^H``gn?vb3+?vFUGWJgI1uV)#7uFbIU%P-$om+xMB
znlXIjT+HW-?LK*FyKOiQ?)$^tdN<LPJ-<nI76Hp{9y#&lJ(nKB;YI(!j+&S+aXA+o
zt+XpD=99z8g|YJn_tHL!`SJD3>h=2%tMkmRqa1E73>j!AF3+G6T8uWi&?mdAD1Ryk
zI(o0s@RJjuPZ))tp0#AameOylU@80@-_;0=Dci2}GKK6Ehqew&lY_Paew$89c%?LQ
ziydR@VMzYf$<imSY1*6K#dnD8P4vf%%jIERnIc^>{HU{!Wy#Ts<EMT`-?6bU$4|$!
zd0MQ~&!r=Ww~w%W>vJDzUU-&ifbz@q9>=sm6ZRZ-Njp>mlcQZ1Ha5LFK5_WeJZTrv
zIODT@IHk|?z7k;<l581%jCp*F&OY?emuH*7qc2_X`n1(D{D7YuzbU_oPaS2Tsm`bH
zUEjx|-$tiRZ*B2i(=~VaHhs-q{(Z<tyS(1<L)**%f9>KHQsD~I`x@9Dh_i8|EwWJd
z0<91h;(#?g?*l}qdL_!jgQ3TIWuC>t9(;!n{ss*Uhv9fRhxK@#t*@zYf#(_DjYjqi
z59+^{7kdaty@hya8U^pCa_tK7`*gH>B&Kcbz0`$$=F)K9K996~*0_m6%-AUu>*g_Q
zDR;_<*Nakw2DqiXH7I9?QC@D0-yGYnsnpOk)|-mwoaxf{&T+@jzI3Le$-44yGkSET
zC-|}Hez`Nl)7972`P;Xhc2r58A%1G;lF<j6J2CvgNRz>RH+oGC(G(pdg8|M53|~c~
ziTuEzEAJ^M`$#F*RW9Mx`OssTscmq)&rb5WP0H)?>NYw4X6&+1$9IhITpq5Em0(1j
zIB%cq3*4tM{T9Cx^3V2__K&tIuJ5_L8B0MrN@py3L+Ld2H$!Q6#pgIwKPDwXoT6U$
zrHQ_DC`Fs$CrL*3wXYmmViuU@XO2E)KG%-a(cE{q7BKc@_}N$bY4b~Wa_&pU627@I
zSb|pq)2FT0SReXB$NN*Jtt9<@VcPW87GJ{mP;^6kh?}}iq;1w0#DU}Pesg(|_P1U8
z1=?lqfzd}vVP(0}CSzEO4_d9!uq&J&iwD+l*ku-sDTJXP|FJlZg5xYTgRjQN>m591
zfYI(AFhZTK3^?eVakIoRdVv*mfHvoHn(sioRy+y9%2T}uyZ{F5`aU9PRE|^S;6Joo
z5xp^e+CAi|>qnJpsih0=bllMecM8~Dtv@u$o~v_B@QNF%OE15a&&tq8`IX^Ec|Vkg
zWUcc{=#xiYSL%o?;XBnpy8b|W(5Effsn7?;tfXG-BS~uexx0Lyh98bcdwqVo`q}P(
z`J<2T_MwVd@f_RGlO_(s*2(v8X83ux-p!Q4?BjiBh95e~xenN-RJ<zuAvExP2U1=2
zr`-unc$ysSV}Kt9bm8JhCLf%Sn|3JDm3-2~Bqm<qZc5)0H?0P83oCVHiaca;&9;5!
zp{dK!N@2!+&y_cmQq%M&<xKGDU{c{F@~JG}-r8lLHaM8d?L(harR9+5lD{@zt(|O)
z*;mmXp4uu+raEn@Oyx5!w{`UtM8R(nX?Tg@V;(}GtaITa3x?FLw)R`}<crQB`{*0p
zXwhQow5XjKn@?N$PWi=#pRx0EY<j8eyL#`_R>y`Pj*;_AQLDu<<c{-S&#6B~`qDg5
z{TDCELg|073$V}5&z^jwP$x%9TeLKONcAS3^5e269+FS*$KsRblD2i1>D|Shu;eGM
ztq%-PPk(8b`H;o?LwjCRf}p=Y#qkb)%{XwT&tDsEFicn>rF^dGZIKCvoHcwE=QNCN
zn8WiiRA8cTkqERmafBmkn8$^1i4HFE#=Hn4xFR2t6VQBq#jI?u>j>0o1g;xHkPKZC
zZI1`-(&I(@CWd@gn;Fxn3-1aO!;kgs^^t-a3bxlJjh(1fb@@LQ|57>{gCk~6$E9?d
z;Bx(74Br+X!{(l8r*UUfB_PLiqA)KvMtdJ?T4|EdD(&o{xW4Vp(DE1SCr2w67US@Y
zt}(*0mdJ+-`!xKdyFg_4NW;@K?BkuI<6c#F(!l?>|M2tbr$2b-YXBTwhU~PvI>s!O
zk$4=(yp8Fkvl5zq<lEJ^)yd`M=yH_suglle@bfSK?Y~)_=#HNf!oZX?B7GaZs&6e7
zWq})610{Lem1VdWVzbB|5AwvXZ+S^&iD{>Cb?@<{UHBypJoE%#N4rn>6g2zDp}gsX
zyK9^{55>EwO!?P1Hn>a&L*)C3{XJF_x_#wE8(-y3X*rcna{d@Ja&#4z%NxZl{<a3%
z@_ie6K5u=?5O~`0GsXmIZ2gT*Zw@&|yGZ-NRB#{jYh&*Ve7n}gb8LsyPn7DqkB-1m
z-L>A+%T(Wb(-p64`bzbX!nMiOrnk2ElwbP3t8}J(m-lfr&{lrnLc8quMEIx_1CR$+
z0`O8fzYA|uBKQE!jVCY>NHwtadTGy2%i=WNb<zgUdNw~8WBl1(f%PNXB6^S|92ofd
zqLv-&#ttQ%cJkn}k%+W7EVL~?ER@?~nDc>%&7BMvK0&w;=EU^lq{HqCwJ(aAgHN5d
zgg5q~sBj;B?AUzApf?7MrhE@2(Ti3SSm!vOM^c!rOk?Q5AKUPQ?8ae`lYZq7BDN>w
zImtdWPrTfi9_QG0#o5cpIvDWBXs3qiOmg)t2Tk+Pp|W+A34Z(D@gsc>_3Z$&L3h@O
z|KI)o^XiX(^<J-yetZW|d~8`G(w-x^duU~bpEvJ5u5KUi{Vt<jiWKX~Zj4;t50n~y
zkg2obrz9b}X!PQW7(bWDHziC{<1;1SxzaAt$1ciCoqmQ+%IoNk89ubr8dH`z(o!DA
z$JI5JDd7FcicDfeD-*Q@_0Yg`(C3``pB`DXH<Uv&{N(bv?Q8Oq{%@BVK)XnPqQ#3{
z^4=$}xE;#n<345U!h1^DV;z;s6z}+HTjWUV!jx%glHc1MTg)4}*d`uzdQ<x9;)TEX
ziw0<y-Ht;0Y=4z?-V{S?ow&8}$HIa2_przqaD9o|!^4_2G$`$X7WLH82;$iaY6z?1
zSp@_D7YI~70f82L6h){UVYndHhR%ju0B|`C&KDf*2SbnyNP@Lf<-DJ2Yge4%j(@1r
zDBuwKQME&TRTQ_|JkxTHu~6Jao}p>z!YhXLa%%ZZ9btjIOMe&BS)QMl2GqXIvF)0B
zPK-T1+gEz&BlfxDXJ6?agKw9eHf{X=_HX}w_4W^cT7A5D*#<f&o*h3gs-xasy*L|0
zYfR*w8h#ia<_1M!*f?ggo-eJjbvZUYawZ<I#m~w$cxsX{btmb(j9oQ1aS_r<c}_-&
zs7Z|@+RqgF67604xwH`G6F5^CI?KUKA!k=UX&xd|)4<EmL$(ybC!N4#X;q1g%S{!u
zSFfv{a>CY?_Z)1OFv*tFXBRpP?0hs(QdX4jNF6_?sSqvdDR7Jxn&4CX_+5DM0)^rg
zr+4{v;flwXjB!}{TtX+s>H8k@;#sasyX5*dWER=F;E7ikzeIms=na95+o2M<VqK?v
zo8o|Ni%;pby~p&LWEWf$3~`KOa(N%?s7@yFP<*@)2H_lC!KS!`$6@iaa8j5!eClW9
zVf>>pg2qCroYGsWcfr-g50M|*xF&rEh~xZCb={Oo<wFTf7uY6xUE^q%G5qkAusdb&
z(-#XN*S7#drCyf?2w4yXmwwf03TL~-gk3seQCocjENLSs@8Mav;oa+j(T;ZtRe)<G
z!8oSVSre=90!`<G+=k#d;j!wUOfK3KfM(`Yr<=s;QbwET42i3>?z^+j*kLu$L9dC5
z;)kYPa!Xuxi(z<a;w@PDbuvgs!8Q5RiBN|s-?p@AU8_&I?83WAt~hQAnlXJn?~0H3
z9kL&GH&S%El-TVQ*OlLGsOSB*%QD$DKj<UoWcc}F$DzF&`6x7I@7-GE-Ido}%f>FH
zhi_k3m$u^vSHeHunG<WK1UH6ndftiQhswbSj`l9s=<TwiHYwu};`B9fU4t{WE}`cW
zhW>rk6Sc;&=j07M{Wn`)o)8>4pkx3)GRuP@N3LixHdc{{%R`+aa^<8BmhdzWo7$id
z$JctBV;W+aSjf^P;Nlrdr=cid&%WgCs;6CwE!on%*`*8%uia}v`3~p!IVTC_7;_JB
zV;Fu^FJfMGGSr2~xH)=p{H~u<nR&-`X_tI)4;&WM9Q!bTy5>#XuGitJuR|lLu464V
z#kJwabef)H8DoC&bBf;-Cs<%`xx6oFtJ90VW5W+qKp9B_m+jv+`Rd~9!ndi9|5nF0
z4X@Li;L>=7cab<Qo8UnMm&?Zxv;0Z}aa~ea>MKv=r}6Q72}}&1dyezd^U?x1I?WfF
z*9aaH-k4Aaz;&!ovezIys5IC!?YoU}gV86BtH2y@6D|Xg<E@Q4);lApyWl&`byyWu
zdGslQcX15`9~Y-LgoH5K3-U)!A0!aOvCF>LFCqe(flr-|=MXaxDJw#x+I8v30u}~(
zz7)a^0gL61>8OOqVRg@S^}b6QygXB`A%<TWeu8|>{asG3I+=2?rD1V;rl2o5Vm=Dn
zx36|Z{=VgbzZaFNht92m3I@MvyhJ>1iagcYV&IVfZPV!#a^)qd4^^2&798zEuDeLn
zBKQ*hJK+vLqD@<Ie|OgbvmNx};$n4keeHmM_M6|dBSa8Gj0yL4+}o)g9u^-b2I-Wh
z;3BqLj~GYJY%A~i`jGdn==BT7rd`_IK0{n8+YnkMxIf?AtWIyPR%dVDuHL*j_ZFZ_
ze--l>Q}!amk1nofSs<<ys^s!xA>(J_)j8HRdNKU5J?FkUKDIRWO;26f*-RGtR+GLG
zZ5s{h<V&ejM_yRt71xgxW_cg3fztd)7gpfhRHj(0Vgj4ypWjDi$4d~G8$1k8(etB_
z%Nw*%nzzqSW6vkWzNNoPZrh^qKD07bUp^Jx5*^VlGDp%^t0g$c028-EOWTvr(O5@$
zh^L{k;paJy_t8k3-rC}yYPEX^Z3-^EY<!HiW=MUh!yQAe?pm5J2JP~e;fDcem8UGP
zU}Uh>XG;Kcq>$H}=Vabg4PM=tV{;XpKSQP>8w@=m4mt{VSjepd0zY5)t|BqU=!<rO
z5HrvVa`8-{hl5_h;L}Es6pnVoah#j;GH;VS8x2%6=Dg#Fgrsu_(v&*x(HLX5O-i@)
z{SfL}T{}^l#yYs6X{TONX1%<7x!>&gnd6wXP7k{dAFB7R@=jHLy2}ynG7IfiXz#kx
zSm4Vvu<J5bCRKo<cRQ9le!8UlIl|gLpVgTevOeA3tWLgPtxhk^@MA`+YUr~Um9V=D
z#}2O|!_OE}HZ8i2LBEtnQ^SyU6u{!riBl;r!Er(7X*C*5c?d>+lYD~d;zKVr*mc35
z8n*5$gMF5l_)B2Yj}C3wUvrVERkC;bDl;rjoSiZYo7UCUi6M0%x3JPwrov)XFz&>c
z8zLLCPmFqt8%G1Vyvb2+%Ku!MA?)53HBRmLiav|jIhr}SO2Sl!N)`c1=#2rRww842
z>ybIhM;g7yt@_<D7LB{+X;Ya)zBIloeQ3<h(JrEL7Vn;uD{Y77u9LJ(jwx>+9X%Jr
z4{cRljj7X{+KamQG4pOrT9R8drz~^gF@?WNxrg$iT@E&yjlpDNY(YC#`eYfQAcoE}
zs7*#1_X8;L++!KW!J+g9Fm}<r@P+I^1Vc>N(LlT$eS{lnU>jbHJ&WMb-@t<_BaY}m
zgZ@Q*4LBhVpJW)F3Xh>S<oUyi2E1RKz>m3DBv=63`!DStdKkK1u=j*40yJtsWuG>j
z{HCC4zQJmX730A!ZEuGV3^lzUg`sU&Kd#-d4namz6xv;i%+mrw8g{)-cPM)6RvNlP
zx9OqkhHA%pT{zN5>|)1H)9x5{==}cE)#{HvzUv}-F>Ja6s2H&YaLhY?T3C%1{mYM=
z)r;$^)ysG9S1(UaN25EG&?ttVwu*z0O^x2rL@fA+tAV<*v(NH!Xp{28&Hd>l&*foL
znc}z_8<J%Q?KxW0N#2igT|MeYZCb2D(C4%oSg6-&fO0Go>ll;hF(=T-JT|>LJ?t`^
zc9GUkix<1(y-!|oJ9LcsI)-jz9hK61E{2~bebniVVFv3yb@5F!m$-2>kRDAq4%P9g
z&@N;6d3?C9h_hfA#;G3Qi_YxOv2c$fE_MXss&HR$v@i;xTY-3XfKVtsT<-)`$HEz&
zj4vR#@rVY~9lnj5G3wZp3_)Jb2{Sn1;=`MTPsfh<=@Ai$eRSTy1}^&L6<)c;Q^D-Z
zr%pebz>DE$UpcTFzSK)E!{*zQ73s~6pE;Jx-DqXu(x!o?^@ldxCH^=xKp!!uJAMvX
ztJjZrt5+vyvrNKQJC^RP?eM|y!;YU1AAXvZoP!NNEYv*M==A*T^vTz-aY~m)FV@EL
zEDh8xxa_mM^qrKmv+LpheszAnX^iU9r}QuWG#bd|VN;o6u_|}e8`9-yHA#I)Mwq5_
zOcmEvpMB)mXT5_rWsON7YaNR#*FbFe8DkPXc3m8sUb$wbgQTpo5cix7KhtUVuqpRx
ztGgI}O6n73v$PrY1?3L#Q{_&b-qbJE#dncMw7TGo1smtvKIIrIJ^QAUE4Q@E;i$K2
zX8}T5m})~Xz+e=z+Cc?Tl*8D&9M%(OU}F+?{McC61}BQG@btbWvR>Q<(2YP&H;ik1
z;oHJ7?BKLpl;f}DWDJ6`j>SR_jl&gBnw%c=O)l>75t+5oh$L7v<Eh2)BMtZW$EdTn
zxA&`4TlZ-i<M!dx2dBR#o?^D~_wsAyR7OqOa>_5|+)8qzKAv)0LZ;F{TRu+3w~nsb
zy)ouBbv&L+_>jB>m+tTG_RiPpVp{RIW-MJ+-I|lz)sCOBWG|(&4<k&|uBKACr-u7<
zYsafyeOsMeUasET5o>dCgAG3<B(6Ah2|LC{Z;D2uPJhoOTxDsh@*?_8`s_O?xBkTH
zf*tEeud=ve4NOswa(UQPrhLL``Iei+<Y*NqZUKDKKRgG0mb5b^MPn>F@EVg=iR^t~
zkg@t6*eBDaH0s)rvC~|~r&P{4!qG1JqFSc$OZAt&&oMovGR5uC82RdYGzC5Q@k^yr
z>>rLLtJ70Eh9EO+iFH&eX&1v!m-1ASN8Gh|YRA)Nl2R&uFF3}2G0!@^DZO>^y_oMA
zDL%@2&!QYGd&&<kE<>~@Zukk-5{vXR(aZ-*kQF8kxQE3)K?WDcISp}($ZXNG7NP13
zw7Bg)q$t4E{3dwpc-YXBp~fgj{dd3-4dOs~VMp?EB%DNadiF#AFWVGwzNqfWU|D@J
z{Gh)%IJISSkflk6wBHyiH*GftR0=Ws?eO_%hFb94KAvgPdsAGRCrz+@^Q#MwCjMi^
zshun}{Pa!tSn_tIL;0tDmg_Pe4(+@!_v<%}hlff=+cEWyA9kCN50|#%WO4h|0LH2N
zZ(mpEZ{Mw6zu0UD=pxnC!`<qQ;f`tOYBl^Ys8>UkSr>5mOB$%Ijb0bU%pEq@0p-Hb
zBj4UO84iReUuiGN1HtC}lwJ;Iim<e^5RZM9mp+s7ed$m7iv64{@x3)M?6E%Q^02N<
zQ3eGc4_u@h#p%7&A6h3%`5X$I%4J`D#_6ont$4_farE32CzyTvb?Mu;TvPKWZ*<g~
z#vMa=4$)(-oD88iBpp-Y`UK4w_yXIMjCM@BEwAEss13GF^t;<z-}!Ezqx&&MZSZr7
z=L>ri-WFnCtj^BEWRVY8C=cuo=z70h7sF3Z4@2`JM|0O<aoNi8PUC57Sa7JDW94Wl
zY~L5fOOAc}`z$wId6&S`E;Il39Xcp?ZTJa2B!&{IF0)2ob#|EW>H!-73uI(#Wz2W{
zIFcK8;1P#82@0Wwi}q%>53%$U;m3_Y{+JklqIiiQ9MRRWZ_tCF!=aS$S}`#(2(w7>
zV!fTk^)bc<&TPb=(C&qMLPsRKkeZnWd>(0KkTyKjjpuvRTW*XzB)dx5#gu*y2e}^Z
z*pX&etkWLNz<IebIq0v?&os;FCfL~e?#8_xTJjy+c(^^~We~I(QyYHDB-tO-A>P%~
zS6g51!j2#A_&K*berD1Xty&E~eG{DGsnMI`fGw4EdEcD|)KR1ycUfMTZp0Nbos_4t
zE${Q!KrRopAeC!__I<}sZtyyPEq9YpY0TwKH%be19}P(VeQ?LbscZA$L7b*UEQJo8
zw2htSw67RbFZ-fJUrYPvUthEMKUC=r2|<6xYH+?buqcj-+o2)lgh?*t_T=nUG4ELd
z-d+2ms*FF#80)B1X7#hc=la?<I&-gMCGyqjjk3=ox8h4^wdre#pT7qBv@f*FdBaaI
zd{CC$nB({1R7bwKDyl*j04n7$!%0fpI!Lo0Cs=_Q59<dWZ3kW0{j)LdP^UFN^TvG3
z5EFQZ-94g<2*Hr#?Sb<RykZeK!i}BxzL|m<8Ie>LiDmDnwo3@~KJ5xpOzEi8%N;-U
z8tzjE>Z$E7hJ2Q>^sy?-+UcWm*)+yYC7)U^Q`1k=Gu1oI0O~Gvw4Ht|WQEWrPfB^D
zaCN%&Cah}gR7O(VlJ{Mtp+t@*m}4FfH>JBIj<(~u8-7akl)_*vdSgahGvltK#@qq3
zCOY!NkH`Df^_S19$Jdum`NR1<j(b}SKO<-K*yQRulCk;Bf!>`ZR*O!WSPbbT`=&{g
za`2y%{`=}{O1x^$*J;4#`DPt#`lWJHnPTBngZDH&ad~jsCe(y$qbd8Ndr$RQn!rP<
zi*Z|4n!W`tXh8LpMV~Gcy<O;UR}-&YmUZ+U+SztP+pf|d=jBwo=-Up(vhn=m`ewDd
zc%23x=KGKDtMiu^>xET}IH#<{c5vub<F+x3w;fo|2DGhg)x|fBLuQk>MUdVapq(x4
z(`c97@S{O6Tz&Ddq9Q^bZ$In^q9u12MS_HLAS@<Td<YG{7N&6X{q7y7L2zZ{iQ)tO
zO&p+`=)hNfxQsc(2U;;-YaJ}Y@Bc9$9FBjp0N*tUCEX2l2s?hL(~IFJ&fnvFrktu?
z#7Di6TEgYV7`ISsHxxzrdHRe!X=}7bbZA;on;4goK4}=ET)}rNI%yh?c|4q^wd*^5
z#A5jAs<p#~*Bw6}KKwMRL@b7%S!Kr}j5IJ-Cn<Ve{B%-2b_aQggE?fzu`tne6k&jx
z?zc2A(mX_>rZU9@ldrc(TgpSU7<e*&X5iLj_yJ7gfArR6;GXg>)ns41Qu<@UNuOiV
z>`QAZy-h{8Z#oYNuc@APO~=^BAYss>Ztt@V?5cw{c=YW$_javIYt=@1-tloe)YqVO
ze}C`yaI9u@apxed-`%*OhfBM0Lk&M#<mjT`u6^d9^iG5vm#rL&TzpfyyXY^+Z|tz8
ze;Zq)PwCMn+lCvr=RDl;U@#h7s<6m+bOZ4C3hpV%FfavXJu2lHw`upo4xcL2`ULe2
zKrl|m(Fd+Yu^`_a<b5MhVS$#q53bWQU+_)<D8mazCPD(7!*EBM8L`Am@cMcm-fCAA
zHwT|O%}x+Bu(*u|C^6{dmZP`qE;;5-hunU6nU3m(x8)Q{(pmZbvcsRb^$$Bt`oXaB
zcGRTZR~gxbqK<t$JXD8_r|<9XhLYLBzqq(q-CSRL*q{C8H(PPX{Bj68evt9b?)dre
z{uvm4(g7ZY?>RJj$6U#Fowj|JmvmCjz;$ec$1+fxwnU|Q2)Cv(wf?e@$!HkEuLVqL
z9nz#_j<TluTr2uO9yF0X5OEJmOv<Y*7rD%1OpwNIXXjX-hGsDBqL*n~v`Cu1p`>Sx
zKSn#m*L(N&#lXptT%`JF6P42w+j(urD&YF)2prX2`Liwv6A0I)x3>5RL|RB5m<DK<
z>2G8JVNF@}$;Yd{$XppzG-x$=@QBhpJlvC0C7wW9`HGgmMtX7j;2L8mZ$H{?6Tgdo
zgqnSnL5N5hfE=)*%MKr8aASaX{`5f>;LMN%JQp|gQB{=mkY&23z|ihJpF4hzaV%6)
zbbKDE3Shps<yCo&%Vky5N=Pm+&p-oSZY-T~8QE8|r>eWhJ{}&LU$hm=j-R1gT&-^G
z2sPGZU%v?Rl;IS|rSR3}QEzQd`2E4=hi3SB^X`KkQfC9JnYd-g&rCAnIf+rS7+2Rx
zmf#N6<*`o6r|Rs7>U#?OTpreyDfrWgFd)k{%{f{nGVh|SmdG(Tn7;ZX9e4H#?3vv|
zvbb^$99lVSs;A|t4v=HFdFYZ>LhEc@<(VTq`eu=$?x@Q-$~=ZlaXZwvoE-Z^H`Y<9
z&X4_$pAtQ^>8&lkgvSDN+!~-=ULW;FzhmQ`M_V|;4B|Uws^l|feXyN}ZXVHsbTE@t
z5O}okIH2Qaf~-fg4vqVK+*N~M#>2+8TBrnv_kbJX^c>IvtWn>>332eixj~7}Q*_{K
z@G>p|<0nhP+Ztu!+&EM{GxYOudPNWTxj5?dp&a!_xk81{t<N8$tWXCigWk_sVfD7W
zDvyLOZqYza1997#i$8|2V;>I>%_G{1#qcv!i-&{H-9)RaZ>#gS@7zemOq)^DE{s4j
zKIODEis7ei(oDP!<DCD{j<JnbE|JF;h9>#q>-3mbpOQ@4XD7K#%9G4Ez}H+J)|DyR
zlt=rtkUN2L!}c7l160XuL|*hco2;0nAa(V$kFaAKy^eAGfxa12%wZ8Ij1n0yPVf30
zW0p#F>=WHSZS|OjpOX4E$BzNp^wt)imb3Jodg6=XyQnBB^W}Xf8lYVch94iG`Y~c|
zAXzUmR|XTx0EK~8lQ_1?s0H;n#8o_xJj($f7r|H%b9I5PckWn#@7$^AGj7&J_iGr(
z4ftVaQboU#TSgwid$7wL<YKMhfpeG5#dicD3sYV{@~xs(QJ#fIojw@Dk16w*j;rlk
z04P&ki;|B|4M}ad`xTL)>sx)(vEa>L19kHn%~5V|?^dU_)B2E?^^Sc!JhYV2RxE~}
zp;}z6FsQsgfaA5f<LBi2#_%&d<>OEoWcbmfObu5olHv#>Uf*ah9b>Gee$U)9@MvFn
zy|8oGHF!hca$|NJxx5vhG5olk#PsSq$rxs;XAS5?AJxXwBZV*id^M2fA(GaWDX%-$
zM(K{9G*vlTDa_RMB;}WTKa_G>qR(9Vmaya3z}RDjy8flFCda1NR}SHgAx~YqC?3b^
z$A+K2?f0=ybo;c`so^K5+cC<m%4Lqe;`Zg&=}qZYyxzUtv&BP=GvhSH&!mTU9QVGw
zI14kY4Q5yzcYq!J9Ll(3N_8k7$K*k~e1C7_Qa|s^N``0qM%!*Usmj9ILgOXKU^pL5
z1|SNR(AG4Nw2ENS+QSJ0mv?+<;9Hrptu_xP3pCx(L%99-AA_sby$zf|=mxhY<s?I#
zuL1Cm(#(Z50eh~9oEQ%?FKv<pj7@5R2ZRqE=eUkQoknM4&c9tj6a1uYN*jyqTy29c
zi%(_1fKR^i7wfRDZpGm#ttJU#*!WpEZE2;9BcsZ?8$WqB0$o0yiRBmX#V^x9F5k*9
z_Xi~-cwO7I>+;}D`=+TLMKQ-d9!{0iHP7fHp1a}a`tfe{>SWjvZ5*LChhyIAcGbu*
z<#+si{kl4R`__+sOUuu`-bXR~ln5U>Ne%K!Fcz5Stbu)&7fs6h=)Y~y*oWqV(_I6(
zJZvgc9Q9_c)7FGTTywM(_gwwDa$YhJTJ-r`OHD|3QyHOLX~m~*g49KOO}yIZ<Zx*x
zx=ildcnpoNtA}0a(k}A%+%Be$+o3u=(k<AUPOUq{i7w?g^mk8_HO3u3sp01-m*WqK
z%T_M6ZSgsrG%VLXr7-(=?;}?qoPDHimpo~cgRO<uscNCt#vB&N)m2&??G(llz#KM6
z>I>8;x5AeJCB!;#c=TTyIwus~Sj;wl?1Xm;6`d0dLKYHug*m<VA8UDnEC8%wkj?Nz
zTdD@N;=f6V!@saQ#thPB;Ac(4f(UKe@w3lzM4dhq!_Qi|NAl1m-@Cx()I(WU<)eNr
z{asE=OPDl3dGmI!YRJZz*NFC9eATH#91l0?FdYv!!Is3)M?81KPl=w61?KAEc6Dcl
zpZD*7npKhp8GgE?#5+qvIL1b=iyju(G7apryo@m^FH7YDHI)W(d01Db7_?|&9Divh
z*>bcL_gsBcwwaoD)54<9Zu)vo&^}4>Wu|e)3S)m6I5xdG<Sp$Y^?6PPtzGCPZinjH
zGfimg<T<2??ztT!^6t&6)$c!ltxJyWB)_#?<d3T}gT=)<szpj7z|lBgegbmzuWfp3
zi$8h=EeUMZ0PQmEkGIdEV+_MoTI*_;O;}3=h@i^wQ!OwuK*LDmJBBR20t$Q`#secc
z7|{_HMs{l>QLxs8IK$%u0bthS>By)fdE7YU<N^<cgSRGU4;X{i_Z0;oDEm=wv>osO
zzKZe>vziP)CfME_b$T)U>`nbUW9;R=DwLE_8Gb0~&ut8~Go>E^qjh^p**Q{IIv)Om
z5ARlAuC7-%x3{y_L^>WGDlC1(91K5qE4wRseGXUmRNXw@uU?*<c1JV}j3IAcocF>#
zG$eK=J$ws0eunOH8YAuh{_EB1^3D0`-~QEKIk|uJ*Z+D=_T<n1;xE$?;X1PGgBD7j
zL!(!s8|kA2Hiqeo7t<L2e6l({JzcSYz<}rO{%*y&r1$n!7)Q==1RZ^vj-M;z7cMGI
zVf`2YBL+PN{XTg~fhKxiCgsy8mbq)rurdX=Tpl*HL3{yCna&N#a<rz9c?z2IXg9V~
zcF6i9CA%!g+nPK6H?N0PGO3S4-Pq-yyvOaT9Shv1{iR)`k7YsaP#kw(487xUjCQDW
zqPtJMUYwt;K3u+Do!Qt8_jmvJORYHRM7NKQv=9c9G3s!wurrQm(_35ojFYyAwj&MD
zF4NA%Miy_VtA3HXr7^7xHnia!DU@|I@{Dl}=FyJT3N8cdJ9<JatkgGzexfq`SRmoz
z%Tpy*5Yc$hn*<pAzy}z$2?QxT#s^PH5P#xv*qm5g7mW}){qJLS=TOW^=?yvvM)K#j
zYc|%$b9tuBs_xm7m0{ZzR~3d`m&@m(^J7S-+PY&dBc<cvKmPSc&(a@vcdKvTZ&ueo
z=4k**$HPO*A>+O|7=Ci;_`BbKUj6Z}-shs0u$~&w9X~^M{OH)Yo+4z?cCzD##TUL2
z!X0y?7g9K%Uk1-(uI0RX`O2(?ugS%AB7fNBRbRe-v-dEtUq=)^yu5h1dV6`<2>rvp
z3i$H%OC!vfY0XW_$E4qwkdY&&mb(l6u@_;RRF%uax`7vI2<7+qj(Y2oKTY}^GcK1m
zW703BHAmf)%6_bH<cUqCJ$X#3hxT7%=keI|#>hwYyqjLR8&%px7U5VB>`S8?$KlYV
zd8n)yC{F$B^>A-Ww+~iA8cHX+!&BUy!o~S{W$@AHJJ2{J!_QPlD|Q`=bglPweVodz
z;=9Vxy%a1Mho%9Q3EE}MGqlbAAsqc8FSyD$<AxCmCqZ@(kHdLm9=DF*IfwySxYHm2
znAopi6b=ukMY!<7qwgeQhl}B#m_gYMNn#oC_xpe<`2sB&Z4A%)8^1RQ^5%GVzrt6)
z<bl->;_vE?pL;X*JlLO97t0Ypa}eICtLNU*VD}xxku}fC<2br&8x!Ns`B^1wY#DPp
z9Ggdup3ftzdgSI)E3b>;C#R_;EY^UR8w)t*HPF)Wuna%a&+Xm)>Z==oekf=b{&O<?
z)ZJIpH6w=ZChE#>&hTIC1cgW2@pJz6-7M}TlCE}Q_<>p5#by%d*!|nv`_=jR_7=H#
z07IvsV>a4!K}XWXdfbIJ`E_qQdN%H4_0z`>Ho)0lL?HgtmoFwqKa2gv#f2G@-gswN
zx9n76_}R!g{7I8?QHtg<3L5Okc|Mfbu}&CD^^!|}nupM;8+efy)s!>4(43=HD%-Kb
z4Jp5etj{6oKGw1@X1d1S@xw0Ev!T|bD1o`QCW_w6BfD<TY*$Cgf>TaF+C}f<4#@D6
zOGnqRxE<<SuS+MoU1b{r|I!RT?=D}h&Z`|hDXpmuKj>A98}d^bO5eHg`^*-!56#78
zD<^wfd=6&`8?J#lwBPh`$kEm>(Jr&&Cm4M!?JPE9hNxE3sav$$?zsEM;{(X1GpKL~
zFyLT_2?m>hsQB=<{i`_hp}k>dU{zeo)8(mzvN)HK#BPLk7QXAq!q?~yVm}r@cU@8N
zchc|eh_ehkJ%_d{+v(hXD>e9RV}&=4$>eWw$EyMACFRd|wwj^O+w!U)PgyRHSEJ7(
zWp(Qh`$#=a@`%%-brMR)!>Qp%1a9x|eFxC@n|T<3o|EB6c`(%9h2iI$8B{J$=YL$>
za>q|KXxw+lIQNO<>FL?(<NLDhb`Lh#|M{0sR!_?Ynisaa=3UF~8ddt6A2+M7-`L3`
zce47`Pakc#&**#GeX+}{*!eTWae?UNo68knsitn9h95e}d;1z8HP)pvc6}dvL9#3V
zzVPt^S^V6W-WYhfK}>F8rD@>BWSm-*`|3MKYYe%cD~-A6v#l?nJUngfLaq^aclTh&
zULM|}<ebYLby?roHjNcZ+n%;Op0h4?B^&J`<yl{??uzfe`NTVZ`pQx|(d|pW3*MWT
z_7TnNm+K?l>Tstp{KS3mm`A<EbnDWi@Gh{znaXeL<J#g&<VxGXU5_bK{v|T!V8)`8
zqn(D$sf?turSG)M;ixxkF59CK?T!Iwv&)A!tR{4#VdUZl7Bil{U^h?|{^&b(B3OT{
z@V4fX8vVouBZUQ4@S~qzAH|KBauy!?@bF=Q3F?L?$tHuAr^RIB9XXL7GA?esI%xrL
zpX!buj&r*!*x{q()%lEHF#g1%*THrTALEGAt~}=kzwbIsJ)&ImajX>=FQ^(jHH}&4
zrpM<d&lLU4cB|a1(a?s|G)HY{Gs?lmmX3$(3_nWnU1b2e+Oq*@G5jb;7Qc@MFi_pu
z`uK+rKW(A^-9LU>U4D4IdVhAY6~3#N!8?9RJ7G-*=icabX~^SDD%FDzW*gdl3?Q#x
zz4oy&JAY1JZ0`<=dGh?(3e|0)(dFISw`N$WFK7(6f7n@@bWk+l7=L1y!^ima%cswq
zfq#t9@wv|EwNJwjNhr<RF3FvOW*7d)1j`ri=k~>W>~ae$b-NB35Hic;_=?<QEk`Tn
zweROcDNAbmy6Q7-lQnp6N^?_OoL8bHKaS(M$!LH~I1Gk4V_}G0e!FUbyel#M)X`1z
zva~GJ(M!X}rkB#+$2<MKmtz}yJSW$+9m?sZbfTNX9V_h3tCzvxQ<#5K7=GfkXsl1z
z+{gQ|+(Vw@J_<T*WveZIh%5_ky#}Uezo(!{3TT(_@0pc_g=WT2ImbGp8&NQ{*Z@|B
z9d`FHt_4qPje&**c!346A{vM!<e8|3_{~ls3vyTsuZU8|1Al=#ahC;{VBoPZP|UF7
zR4cg%WZaFfG`wKPjLUAJkg8<hIUMr_Yp8T{P5Fe;=hj#8yV0zJM!GXg$BF%=W0E?Y
zl5o_~J>?~o%V5S#OBu&JR8wB&F~$q^d8E~|2kyyY!e|9^QyzyOPduKafs*n}eiMDK
zZU(T=GS(#DA;)Q4v?(k5P0gM4EoYojIif2a54Rb9B+{LIq`)0OSJ&HXk&2(%UkpD}
zYxub7f4Tdydaw-?C$`~YUq+ssh6fvdazt}sb&IYk?9?548-0B>#~_ojWgC1d|CsL#
zh=2O<AsSRXrGE3{hwaSy`ZOX5p|knvgYDR97%+dkx>{Y|Tql^DSLAu~+Kv!8vt2nx
zhoj)GzF#Lt!foJ7o}}jhd$RiVPurGJ!uM_XDH$-8q;-LrMgzJ0VPS<y^XXYIoO-dR
zu$jx%0UwmvWHTKGLHor1FvylTINT@4(@fQdV3<IzoUy7z)_uWf(F?t>sIf1@$#&bb
zFKsmQovQ3KbKPGo^-VP`S5Lc2Hn#q9ZB191ri3TG>`;ASJJCw(6-FPNzFLjtn|pd=
zS-U=~9Rja!IVqj!c9pFJ{<1Ro=$@hy9AI{3_@Pv#WvWfj#M4)+ALf0ZeMG5FNvmWA
zCoWq#8g218oF(k|G|*HZWcUe-%cQ+B;<#~##bzsW)w47c#x#_~P?0dEXh)4;VxK;i
zx%t{veL;Ay!VLyE8D@xczV=l%rQQup&T?bCK{gk#qL-7)=~QsRI7CQTODD~`|L27*
zPV$V;B0Fu$7-NILjr-kfE<2*pM|AN|PwXB!(zv*z(E#;Q^@e)jZ8`PBQ`m0!iN{pz
z;DVNsX>`zT0dWJP)r>}3P3=vi088Xx4Nz`yW42v`g&LVa>3F!*@FVe9pTPk1ZN~<n
z#qgsX7|;Lx&5auzV?#*E(QRYhK{Q@U+9`dqO&Mp`SF1C-<LBkc_9l+cW)zBb&`xbU
zh@BXI#$L=BFZoX?m2|LAv8D@o|L(mRTEcDycGskL?9iF~{+C}&fY7PbNy1$^?=IhZ
zbdC-YzkL2Y$cR(NyF7fJhM&2Wi6(s>lB`YX7#f#bSkb`ChT2<$<1p`Kd1?4jZ>oiD
zgY=>4J7oM^-W+mz7TL$s0CiDQXr6olX#n}LC%R+P8<4atD(w&R%eIAt7+LM&F;7V2
z^y+zxcBpisE1pAr7Irw!?eg`-ld)ha?OhpukmtFTDdj<1*=mayoy9c0cRbtg`#r8^
zYj)bB+M@Q}BT7|KR8f0ZQDTdkpp>>&6}9(Xu~%$C?7d@ev0?|oH?Pn8_xn5fBadh9
z=XGD#Ip;dZzXgZAi1~&x7&o!3BRXuCHYNa9d@R=j!*3KgzC90DU^Tya<Dp4i(B9Lb
zonb>2eXo#oc#xwwRnq=m3f)Xx@r<P67I5Y9$VS`!u4uP=a-xQhR;0f>q+Oeb&~liy
ze`oz<F-<iP(pcenmGhW&<C5$^$8f&FHo!-vb#smVq?xyjbc^x2u_4MJD%ZF=cTF<9
zN(})wp#Y5W+Q%=o+&RDes9&KqAaZMoNqko}{DX_QbP^}l-Dr%eQa#on2jqCAM=%qp
zQ_$Pw-r!p-Cg-A?U>BQnJh^42Eq>p1T@jIyc5v5Qw7KV0xjlyq!2ha?L{J4yX|`pD
zna#u}u7af=HNR|R4A`XpoHX|w)s@M7CRRVPmmtsQ1mfxFaLl8py?z;8aSZ_>avE`%
zO`m0jEscShbgg?0=X2b!d6frq9_35jCT%Gj{{EN2&mn$z#4l+M9b{UA$KL7$qk6{A
zXVf54-L>WXuY866XKO8s2GEeSy1!z@W+C3d=;7g?U%BRKWQ^VM&J#vn)E_b)QJ2oX
zvD&pIj<U$AoYSeNY-O$eMZ-!5N*OL3tyloihNJ}EdxHQ93oQrZy&q%uO1FAr`6ygI
zbD><nfLNo=@FfqU*kTk8p*n9T*V+|PKfw#FqS1Cse0SdPX<<uddMjGD9jEI&Q@%;T
zHlTpf_Wtx}SF63KhVa<P&}(`~tnUE4n&7>U?tB97vD0ZMhbq~JGfAcc5yRPZkTGe<
zdFd~D_%Ogf$0f@NZuf`!og1seHO<UdFWsTr2-6IHzwY?JuL}W1-)->h?vN{W+NY<y
zuU~HGsB#OThm{Iq);M$dSaGVSlX(8(i^G{3ZR-z8+o4Ps=eaBO;_i7?ixPbizCm9;
z{I;IOImj*O9w)DUpAv(SDL1q2G&Y-nFR$3BYmU6Pi4q!9Vxt5fH8+bYE?%CCb-W`w
z`>8_--1<PJAg$1g(>*YB%71rFR5Od5iu|iJ5Q6RPHT_^78W5T>%v5A$@yyQL0j`Pc
z0C!4V|CFgFoqIzd3oFa?02UY<Sdihn`H8lyjqPo%ciV>IrWx>AAEguM!u+)NLSumw
zAP(1@7YETfZJ<;OYj^wp*-mxEkC`n;nh{j@GwOY?v5GpVR_nfge*HkXK#ysR{yw2I
z{@>xcqC8nzt!=nK87(={7$dMX(l4mixSs!K0Nz%I4&Gm)8QLY6h&lwrToN-?^oPGr
zyka+0UfunkSe={beuv?|;1VVPZWV7+>L+BV_!^RxvEh<I{5o1oi|KtS<3fLFob{Y@
zq`{e0$3ByH2qadEuM6MXmj0t}o#~xaHR019BOHG44HNyk<hdo@)i?jKvej-hj!X9B
zmD5j?X%=AxDGihU|6TwiEA?WL&;4&KuQd)1j427$P%Bf1xD9X|+i^R7cWfq;mRk9T
z5vI#Ek(#>Yd$RV`!|ep#0e}8yX`0`x)O3z7R(MI(wA~1snkpQH^Yf!ZqP`p!#eO?Y
z=xhP|+PQd>VD%cGKJfh7s9kTgxpz81QWC7Svg0M_<5%a#0@qXPOu;A}mYD(JLG23+
z^S@4Rl6)N$=-X^v`3K&+y12B!1nBDd331K{60R>0`wOMH7l>`;T-@SB`u_Vm&##y_
zR27QE;bLPhj0iWs0VIbobF<(APV{Yh&xlDher)o(uJ|yIV8|Ehxg9Md2>_h<s*JEA
z#bh-dIG$W8q+w5wvMJKs`xP_l4y7R|ZP7bgUXcN_F&DN0-jaz0;gCpMKqL^M8LtF&
zsag8K8!tXcJKp)(o_Sa425P)nv;^I;4O;SG$b#_rg*FMos(V#rnx(QI(mCi%(eew>
zT7NWjy@=SB7X~%REY(0Py}P~61;Zm)VT;d%i4R*uP{;XXvT+p7u9t=i#qL=~3fv(p
zxb%ogCfs5d?J$lZ?>LwhP|c&<LNR!R_!M2y?gHZLT>P-}IaJ(Z$x1d}v%WR|u!Ip3
zlIDbAn#ky}F~$H1#=~z`GM(K|U^EZCiQpvcHr?{p{2F8*Er>UkeXkS47=5+dC~1rR
zL@P-5g{;Plu)5t;D>FJ1DRI(qWgq;WZJf15$TF9=KagcV>%@typU3L)M+j6@Htp-E
zQfRn|Za>~D!0HG@mPz2^u1>tG1AAuHPgCD*AMW>$i}u@*12~FVs2G^-zdn)ugQ}uq
zy8Qg<mi<$gd8OnmuImbiwQbsZflqyG&a$|cX#HNF9ee6kx<2N!aGQ`{xPmiU?{jPK
zczN-_HF-FQGx$MU^;d8S=-~-&{0{dB9ZHQr0psdEZ&>Z+6tc-x{p1Hl4Tl}-71MwQ
z`x3gLhg6<H+t!{~EY00&YaUfUjHT9Nr+?b~&jUoLtgH~vJFH`OI=3)+j{rP?3h`pM
z%ptmU(ZzExfAR;D+xhu11#L~n#aDL>cB7bq#kAFD|6LY42H7aR^k!^cJ|*Ez7(ee4
zCZqnJ_tW;yL)7)fWR02p=i-k0**AisSYsWb+yvTQfq{*ABW`#8{YVz}WZZ>R?Q?ey
ztXNU5mxPgCfAJ((SXkOQ`{nO)<a&0lUd1e!q6L*}@~xBd$X*K>z|H;JmX4Z}#z05Q
zQku^X9Za5p$JCG~)w~uTyKkVgN?gU_hMkEf3Pyw|;ohB~@fY2ntEro7@#EuzCghu(
zCEwv^<g;2q)25CxBJ*pK1d3$H8mt-$cf0jnL78xgpF}%~LRfbdGbV?5$}?UB&e&Os
zZp>L6;~W}A2hMYC6u@rxZ}E9}6`dXgIU3ixVv(zE61f(LnRjt{HwnwO@v}!&%&k7F
z3Ys)1U4i?L1`g;C_o|<(X9R@|X*Rm8OFPuq?g!kG|4m_SjPCZ6(mFIbazTAuLyPdp
zf3xsoFe$$mQeCpOiYs|kzjWWfRpL3_=%fPbff|s?9Z?m#HUqHI`4~T|ORGMKe;=}h
zLmW&$7sJ-B#|4g0935>NVpGEW{pde3cn(<SxXZ}=SW#x%OfGGOG{ep%BRa0GhI`R5
z<D>fB8I1c|4{=!gMN0SCrt8`5UY1y|$(hWA*FRA#b_;hpR=;%_vI;sryy*_-kqF6C
zN2j+!&K3Y7UhaR~TTWwaYGZ;IUXApMhAcs$13#av&Uu+Lnuk7mM-fbhejl%?Eb!js
zUipRytyGPPD9d82?_a1d^73?&$17afEb-B1E}dxYLN<ed<A6M|4fRuJ!;e$aSC%Ev
zz@NLF%cn#+9ZR!J0NPgPJ6mapbL%ap9a1;`<IO<~U2Am@n;)h;jQHm7X4w+dSef?w
zr>#@q!<WW>)Imc)#5gyO^dlpTFdIOhz{$lWa>vc~NEkt!xf?;)%ourD+z<b4BDWp>
zBkJUfvbKPj;f5f`EVc`>O2i=BPb1UxN{^9){<xQOSlDQL`hW`K{Z;SkA4Ey{xbu*!
z7!&Kc5o>-VdK6_H!^qZR+f{C-e+4~#!0|6pAXoipSz9xjWxnV&QrL-sW8v}N<KR%@
zuH}4dQypWy?qA1I-JJxBu_ikYeVX1V4;NBKx3_+(7J8Qdt7~u>U$pCyD#Xij`eE`F
zr{UYKW4#O`(-cTAhrwu&`}3Tq>>7=y^y5#{{JOm9!zCoL9+1NV^NzK=3aOadA9E}a
z8a9Q3k|k^}6$Gj1NHlnSKTNlmMvT&=rm>yFzZ2d$kI@X1Qph}*?csz8${YzYojz7m
zk#gyvaAH;OdEB8`a3c0)=ZW0%%P{jejnRA59L7hK=<lR{Z9mfH@JESAO(l*fomhFt
zewY)6vW!KNCm(Y+{L*v@;3#>F9BZW@h&<MAasTr#uuZfrDE~B5+xX8Q3pk&{lf2~8
z2%QI^FL^?yDlfOAd-aYumkknW{Yoy9ECtTLA~CuwNulNv%^&>ZT%|IAL7GOE^yf~_
zv6A`i`zCjMX4n&!MVB$<5lVqK(=K$1pFz|zlgXoX_=G%s+dXuNW8yoc)nY*L^1uJt
zBr_ek{(U*}*%tjEh|<gj(d~XmKgI(weL+r4pQQ=YRnB-p&t)b^;;wMhBl(w6xk6Z}
z8!ByupXSc%_eCB4G}PbIw+^O^8sDVKuuPd3c5MQmd^_vzj@52_3?C+oXfLWYlqwh}
zfw2uYb<PD{(C8{0u;j^x)LK?~g%MLz@X)^=m!E$WYvGX5l!|Pk2FT78(t=L=SO5D+
zUqpb0!uiEVnOd8Ye9!I;+BrHm;g)_;FnxZ|i8+VtEXrWl%S#m=b?V~*On&M>@~3S7
zb$K>Fe7LYwTR4Ag{=?O;Et?mEerzCjQ9{91`G3Af7mVZ4{^F3|li$w9({EL>16dLA
z3oq-~X{^rccux`-cHxc=6pZ>C*U2<n+~hQ6dqFqcY79HzY4df=m)$0oA-D{OO<O&r
z$*5UsHsu>z_Wg2KCmz5uTd0<jGim?m=M=w{KhSuN{gYTm^>wh_ab+Wi=)psf<@xt(
zJ*O$VTZ<-0Ld&Qqwjqg-BNd&prZ|&Zn(tqyPQ87n-tOW}o5mL-^JB>AZnMmy>PpZ5
zrfgwrYaMWFbC<|+cD^}ipG><IyV5no)d5E((><0Tj{LRDfSi03`O|3pn1mYRe~Il)
zcOy<KSL2|Kt`!VYkWeLuQ<4!e(y%M`h?QtcHjjXX;ZKzZG8rSe`spMK$^pp87yEBj
z?V2lJ?(y-;RuF8BZ}E@1IC2z6u`B{cA2piI5DAB<d15n8s16q!M|twh;?u~A&G!a6
zT$CS5+gm0HMr=w$cuEg>OM4v|(Yv15^)q-pRr>xAb03lQB8{wr9!=%ueB~NPejHng
z3c-;=<}#|Ke)PL=X#{nBa8P>oGep!;Nno(lxS;J&+&ew+*t@+?{FyQi#A4dpM^nT-
zTVK(Ox!K_xeHVPzuUB;SG}WB76Ds<LW==nOH#Wtjhr7_)#!Hbi(gs_Gw(PfRqF2$1
zJ_j-UCdkYbIn`F@ftAPIGh3;2ex`?Q%P;9v$pL<`KX$WUE_ma~t=+I(%)ym={=XAf
zZbRYV8MqK^ku}m!w+g=1i8NiO&QT)x@34|j;DrQL7~@(Anh=)Uoko-PqeLe8+L^`4
zUUc@6vmW*-Gu_B&0-qQbJL?1yX5TTbo{RLkPAwh23j}U^Q69{Ne-p1gxLrN2oS_w~
zF%>wyY4tyAl}_e`52xRCnO%*9DBV)k&ZL_e3cUQu_L{GAe(gK!Iwy7w#p1t7Ne}uY
z!{k!Iz23}p6z8_bcLf1ThfhHl@$S8+OiPvCX_jb%RPgHK-X^zs?#>FUiheqGK3+vL
zwnm6R2i8)#s4#P<>NRUiMRbTXL;$-U{;AEsxpHlBuO($>r$7uX;BLVjX4hRl7?K{x
z(6O8flY(nHxc_!(NhCU*<C)+QYZx6l2(`Soz7|SV`{&F(+MF^Zoz<H*|7IkON_3%t
zs2E0MJ*Mx8pmN#6{A$%;;ihVi7Pb8LD0CY}skefDoL;ehsKya54z}a<+MsXC_>pNo
z6nWg&^2FG&WJDJf=YMH(q>YKyxJv67-j}t(W}D<VjF130XzCQfClxqU5*0_4v)JO@
z;3K=5BfX(@(LxJBiBtvi<0=K3M?~NYZBC#si<0b&PtYI-b)BvGXj@jGr0wzy4jMS}
zG}^s6-k$^M$JVumttiiNUr5%3Uc1xSB)@R*9LH^e7Yif(+J_2EV0pd)mCDV%=%7Ch
z3v9t|nv~M-1zvSRhYH$o-B)iz7h_2wT7`)?b}702^~TX$$q%WoeBBQkO$7vi{hd&j
z1C+`_99?XSn}f;J4{3-g&QbOU9sBF<cDzTCj890FM=PxCkclfCb957)TKAGF<g76k
zkLYo`>gHDBRGv)0Q;*e)*O@2w!uN46>POXaf}(UWxTaj`87^7D{kJ!|#2wAZ|F#0o
zWuu=Iglkn$f0C<MH~ut9y`teqsT{!ipMi_nfFa-q>BvXX*=$KxSK8}Ty_Tiv$Bj{y
zWJ{_w_PTK>H&SrP4~f-KSW_f&u7UmTfcd5UWR*cXli}B0_fhd1yTlJm*JFb#g$G=G
z2yPm7%%|2tv@mhhCj-y}J)9kL@5Ch9?SMF4@4RBfk>j2VJvCgjBx1w+zb9^!AIM5T
zM*IxC;5u()Hpu&bh5kdvJ#qk>74~t$Say_W!y1R~Y)vZh*(LJg1;I!{2+L{cC}5E%
zx^HR=?1*C@axC~H7zd}^KbYuA-r*8qio-&aC)JP-tL)*LF(suT>98(LzO+4lveX*h
zyB7!Fb}TUuLde6_kWfbs6(+xh$gCOK%0X60)k}Z<pOvYvTjY<;w@%+d4-#PI45%IN
z7PbhRw{rFv)-<M1nuqfb5YrWnBs9CfvL_w)>t&1`J{$+2Y;g9knk;g1vz&o7H}RS)
z(5wXU;ff{~V+Tw_vv0Y(=C!pe?0M&RPPIQ5wXY?`4h5;%)o9$^5ZF$i#S1q(Db@Ax
z6K7dXE4E4FIbto2T@V(JWqh@HA|f&EQ<_w*hO{)SRy@})r0r~S>tPyOz1xKDuX&(A
ztYRNQms!4FmvxMd@zAykhVKl&OwbeCx$}xwk?D*KsG@i6sI9UTboR?;E5AA7OWNJL
zEy+WimouL#1z_H~moi;%5*z`(j0mj50gfZ93ReI9EkToQE#!c3=6`F8l(jQF-}&h!
z#3*EN=>>KzCP<%c8xY>CE}Mtr2RN0C{<Q=#d^jJ;&=|96ea=Vk=v`@q+|^0yDkPFa
z-#-{Pm<<WT$-_REksYkh>be?H06an>%9i%@D_D?-->mKn2TA;9tx|rlZBmVe$aE?A
z2PSL;aO3(LqAvGfg0E-;k~F~z*r8oTood@)cKW!*#yX{~ap;>J*@Jk^9&lr&)r2Ed
z)>ztc)oStQPrGiyZV~LC;npjR_J_rEK{<=yzhN|+sAEns?A~g9Rs-J$H5)LPnkRI?
z^Xhnk#55(8hkY%HI(C+#bBdNXoFp90F#a1JD)(|~p>*NQGOQK|1s=&ll}GTI6*m&;
zRPqN-#2Q?8l2p?rP1R;r?Bu0EZdN~xzl~K=h4<xR8BNm#P!91|?)GQb69=w+HQZoK
z{IW={<X(Z#$}wIjUH`b%u;21=$~^7<N1PBRm(EYFX=7;M>cJ?~9Z<7E(p+?~%I}HY
z<FcWV`|l!oYG?jl5AodmXCu|g3F|BB`TyN7;hBM3-!N}HTdSDpQcQKXa9KU*m1{6T
ztCVC?NB+#wlg%HO8h;NC@2-)pW;ZBJNa{k@*jJ2QBS8UYX(+X_Ve4!SKmT4qooD_=
zzKpp5j-wJRiVlB?X_+lF2jJ`pj7)BB4^84`r6E5Hh1dPW6o*RpmsLzs-(*FtY&9+r
zxLIGc=Rk+#(F^3vdS5YZU12QqvZpaj5&vMWa4KoY%%v3>d#yyr0gG=9QOpx$uB#(i
zou;t+H#n6L_EgYwJ~?vPs5Hu3ZM=aDP6flQeW5-XtYTn(zbOEMSe;=W<C3DrkhEFD
z86B<#$ptShCt1y6keIC&7xiz-M5(y`!-UV_!lwqV8;#SM6I)?`IJvFl)f=^U2HBXc
zzKh#j<iCvK6&nWMYhGp&2gIo_L$WD=v%5=S$yDsCzC_-Hr-^i|@T*M`%K2<Cuh*UJ
z%l3{pH)_(W`|u42g**=Z*jB4qa~H&}s}4P`{qKG-6&3h0QX1%6Mo4>Hxt1Tl`M<lq
zd%Dd{1dBZo-km-pX0h{jfK$0eO73GWLH>;}o|MKC4@;`nnY>@=e;fCoEq%p2ns-Cj
ztZm&~=O;fukAr-<+F!!fR2^DiW*M2NIWiW}CtV<{yELlwK%&7At{IV8FqAR!CHT2I
z4`FL%swj<^0idR8R`4_yhGWQkj)rQZEWGAGxFy#7VF>>yNf;vc9n6FlqA@trE&H|*
zE^%>&ob<UTBqv1zzC*2PTWL995?xu}S9X&rXdi{`99Yskv9oIVWcocf!uzsf@s)vN
zCf^YJOeSJd-K}tiBuNTI!D~0FdTM1z%FrJ*j$`c^SXRjNTF|F*`Ay2{LM6x6zaj_!
z4irZ1wFm%z=<TRmT(#t1(J27BBhr(<Jh@{}kaLl|l{6)dNne|M8Xr_f*5CWK<@ga%
zBxbpT?<pkN;Srnb!U6y~q=B$MTmEIwRn_1g;vR7qCgImR{z`K|lp~qN!af+IZ$+-2
z$vVo}`*QtB%I^TQ)hq(iEuYCGvAPtejMFs6^5MjG^?p_HCZE9+fQHI!ci@Ki7fdGQ
zlI)&;e9XEr_J8kAiFAQn(MOI+vgz&?DosGbzT&rb1CjFo{^O4nG8{PQIjMb1Ij~(>
zDsE-#M=JfZVWr(oGIgsjw#<EzEV->w!+!25iC<(djFQjN#2GngtTxHD=Coec^ELkY
zWIXya!uI=R=O(Q*YJUQR{Rru<-y)fl01=8x_WtR3i0ebXMfzCL3?uv1`b*q#fpRFL
z{gx_zS%bFD=be}Pg-Dz{AlJY+TF_(3A$}sUSIN(3fT7$YBz+0FARWciH&qY2PL>1T
zI$X`vrp425)kZLL6r0_hP6*L6K)$_Y5cTT&ajQkN|DMM*W#^~V{iagkr$;G%Nmin~
zgn6j~ha8g%?A1!}Qj+uA2{fGMFr;dY37~qRFXy8{!8xX`QD=3uuWG}0rUu*L@+UVl
zthGwnXfPnKI>0p4kxNA=?+un1Ss`8vmwsR;?1|I6!*t9621FIzC-agS4<E$LsY;1q
zPvc6<R+KxMA>I`}PSOEmQ4~)BKbEG^+FWX|wfv`=&fx2!$4y9SX~W4#)0DZIGZ)}d
zp&!sA48=|0^V|3SA#g`JSR|a+N}wk>wT(Xi$&~=9#7mhA0`cxji#@|;hWYGvvhIFD
z0Ns4Ro42Ac=s^g@A@$IFe7XE988Wel<gc)LFHhe~gg(3J^i1T^S61q7<9-$APG=F{
zyUKN39kVhK!S(!a#_7>0$aZmVNp8MyXa7_w?f`t>IY-8G7J|V{^kq&9|Bim9bGDRZ
zbk3y7Xm`%jdu;g_)!f%3$d&6ZgGIDLt^zM&uP`?@jY&DMwI(taI!L|g7ji_slLoFY
z$>S7r%@F}vU{t(^&Y0lCvuBg2QK+>y(#1wsI|&sor7rOsS-2AFYXV{HiDU_<W?5|R
z=Xr?q2G<%9DF&aNLlNMGBLX+-BFPTE{r3W@M{L|H4Zcyh6c4M#Bkj61j^;W}=JJYc
zl%4^6IO>@)5zZHlss~HX<UT4Pykmin&&!*KXj1dW3k{;i%!%fCuF^_>Tki4Iqv@X{
z(z8=!&((F^A#=^GQtkFpc)<8Xc+;uy3FV5aFuLNU@es)f^fiM#GArEuZPPZHiC0_j
zz4`(|{|Ngh-O7sWOsj+}+g)c}v-CwEY>ul!&F%~JKkI1!H9tmS(cG&NZZoUC%8{ga
z;@56P6)v*Cm`W(cx<+!<_~}}N*tfGI4?mP3h~~-kG@nN(=qW~W{wTZFaQ`NDE)2JO
zbyDM=h5Q$3LLac{h_M@={DY~FtRDT<|CUB(d+N+3OPgkiecX68F0!FVn@c$*KUd=V
zd$!5D;p$)Xjl>2WGw~HAbthuBY;s!yw7D*sW};pkA1&U5c~cRVl5}r4%Ua52feIe3
zw|1yo2c~uGCEkevJQ`Bcl^S_)EW)Olxt1nlCO)e}!3yGii6f9E>(we)@8C3eq2YSD
zsrV@~fCa_o@$*RWbzM7J5lWeeU}@b<Gb7z`a&Dp2lC2xGZ_8vHonlJDDFY7s(Bsp~
zFPr&8SnRR}@~h3xB;=ssS1WJ8f+jMMYnW8sc%5l0EyC+3bqiqrEbn6S%N4TqmRoO=
z5;qNgd-s#!P+)uyo^-Xca4j-(aB<mk1-ZFV@!D~G%nFi4?eA|GA$d>S8767^<-_$V
zkohYll8dv{m0>6Om8$Q^*#UnW6L2T6uyt)Kl<AB*T%w0ovr{$FM;n%RI<Df)vli=@
zE)0|+Z;^>_w=3WqlL<~^_&qR8(X>2B=!RRWEVMJ>bIRMO{|I<RAu>4*-J$gjban7C
z85bs!xyrl5#RA&=KAjx>FXb(Dd#d+ONUQf54(%K)c3_F?I0yR=;bh}dyTA4HY7a{8
zNmjjiXX2+PM}|H6763}T?GDXL8yL|DT>ymZ1!*^ti(+`_A5&6vHzMV{C?8Qni^coC
zIEQG8c%82hW@r#tq)Bve8EX0kFiGBDGk-?0q~7WFg9>hoh&mLj%m{5Puh&>QHLqNb
zvcNR>3i$Bx3mtSvV^3(tjvT%CjJv1tJbJl*UZy2)_AkdI8%6TBD;zx(UEb&>1Z)a@
zlES%Zp-*;jiSm<YgUgncv=CUaGY)GWojqGyKrK!!^Dpg#LXq#jW5@5#T5rmjD`+88
zSCU62BD8##%Go#9W{`@EZ_9>CEsz_Zv(3p}+KF^emkEY&iS(0Hx*}Jo#wPAk-kyh?
z9qm*A>pbg<hs|0_x*{UDJcye&3hHWL|3Vcg)^RK`vFXwD5@#4=>C>;!LzkPRuyCBP
zu3;OUuwh%2k4QNH=Td+hM`h_w_7s^Tp!%<NQ84<A|51o0<F0nZ8J<t*0`kPL8G5{q
zJ%>rRilI*2M!+LC%{xZKGq-62y-S?@W<OG@=GSs-`8A5CUMR<$r{YH^>aEza)taU}
zp#-g~m~uDKZ;eoWCx?BKB?9r`?-dEWG<za}8Z*sL$lu;G%g`7xMZ3KDho=AX7~D%i
z9=2V*F{FqW7=Vbt^)Uz+r+&wDt|wIUCq7?)j3d-@y85<p1iNM)h>7Gt)wPf8AOZ0f
zCt58e2c!=!p8yGiqWW-)XVL6^H@>@Y$vPXC;%A7W=JH~_#YhLtr9-mhqPPRbcKpT?
zqjG@B)Vke)BRC>XZs0elc+K>*?HFlLy&8u{qYxGioYrJX?+cS@uv)y26WWo;Z3~KK
z2(E0rmu$eiPLp~$CCC8TlzOjqyVt=2`}qexjAPm;ntBM*%MBOms7TZ68B00jrSETR
zf4*A&kQ&>t?S^t6inwL$SI|)QI1?d3Qj39-aJv!gcYLd(;%=Xyvb6RHYqN$?elken
z*Lwb5!zNr}9xm*#^daw~__Eo1r>F2N>gw+8F7$Xon$`<s-)AH2YXzRSl!iDWw>p&8
z!MQr@R=Z~U_CU}AV<-zdBKcUeJ7DMi&qc|r`cec->*(ATd|_^kADJQQ$A(~k-ouud
z@-tkz>8Cs+b-;@(BXSvPc%(`V_-$%$7LoQr=6F@l*^y`S?v~~3{y0bS*U&YO-oDyt
z{&!rp+Ugrxx6p&X$*Hc718s!_OsMGex=*e#k`^@(-arF_Lv%Y*Z!KS_QgKS!N*BlE
zBGD(8edKaWp@(RdXEn5tPZ7ZvdD34G99Y(8zxl4cP3t=fbuF1wjLKe|;XFmX@)u#e
z;09maOiv=C;+rD?%@hDyF~_2povZI|Hu~d>>p;<3C<)iShmk(nT72TzloGine;d#H
zGJ_d7?Z}8*eD+SjQrt?u1~h#z;M_b!WbkG648(3@GcGQDf~ji71{xGALPfbRi?iJj
zK^sQ;aQCkujT#xR{bn|18k#uQTI6L({wq|l)A9Iuq?W*zjYV5;<5cRlS@$=ah&}Uc
z!pg$oVzpsD_$URppI?qek<@zx$*Xy-_KNm(-(L*>SI7SKXj8dW`q&dc2jNAmP~)sa
z3?1@DKiAH!CKqZl7){HKo{#D*;}-fIE`EWh>2$?vs$l7D$Y(*|auaTzyQ8~_181GN
zz|i>fYV;!`yT*d^W@yoL?!8%;_d21wcu($PPnGu+;_tl$&%RrDugT?X@yIYF_2Jjd
z*V=TSiH>2E;+8OJTZ&CY>1e2`Vj#(73B=%3e2-g>@9!7<-*^qLj32lJ2Dp47q1;fA
z{*KQKexk`i4Cszzq20#?fR3d%|Mb93Y7okg1QP@(0%8NXf0>b^o6(KOYk&O%R8n$#
zwl5B*u9$WeaZdKxBm)?Py&N#Lp?)|ex#u+bbQpZ7#LGk#WsSLX^=Q6A5vOM$E6ybU
zttjuZZ&Ncx>|o4qH?QVYZvTf{{+N=dFehxhJoY}#UTm4s$nvx~(`%~!wSYT2U7$W}
zxwP*mJsZrW4=!>D_(OPl+eMkeb28`PdhiT|LUB=;PBJ|J0gwIEK<EHt93JSoWd3Ik
z^nZYI8BPw#?>IJVRlfY#tKsCtO+0H`_8h>lfLwP>m1sG=ylM2TP4GBa-jlH0(n7gj
zksaS2yO%rm2o5HEW&*cQ(0xUX`8Zx)$;FiEft<!Y8(SabKcK(3#-6Sy*80saES1UO
ztR9i7t`Yn8mh}DNVo^U9tLcB9d`uB>QAa*!Uzx8nyy)_c{AW9~pm&oLlY*<nk~>GV
z)|pKqoTm~`aw_`oW==Q%gjR<M`JITK$))Bw0Y9x%Yc%2<oG3T8<F+&Zqt%NkLLg)W
zBgknu`G0ezemUeod8yNyJHG5ck(=a?xT$f;UO(H)5k9vr%6GKB?ld6y!>1kP=?_Dr
zvLik&)@vc=-2Bi}%FXxmS4h4f$b-GyNxyov&vf{WZ+>;usq{{0mUtUb$UJS~gD4Aj
zZEug(>HI}J9PL*Ad^h{(^!mq?*S*l}gZc6TLgj5R^C56#cq!IV@ajY1f@Agd`H3J+
zy7Iec|7n?J^4$7qoxGW$f!9U6$+g{-@xscPjyJ3%z6BPzf}I(XUbYm38;nv_&8%>u
zM;2h0AT4C3_xN0-M2#C5cBexi9(7T?JDje^0TP<Iz88`dd%Q>vnEo1j`A@KKf(|mX
zf3mS#cZTN?^Due9w569h9^~n7Lv}k)@`Q?haLh^CL9WHWl~qf20ois+qwPv+W`nW#
zP9N!n%=5q8jM$g)mTG7bvg*3%tMqX+`-jtc><|E3bkr6+g&pw#1HdKm;Fi&AZQK>a
z;vnBNYZ@>9EwDjdI4@3kGn4L9s71(vdf^*#thqB>zOGbRz)d;}({YHb;|BbWbebF5
z?|%dObyUOWed|s#nZ2<~bDiaDtR`8xWh}*c2+N4go)vdUZ@TW2A!!n!idG2<o?lK;
z5sX<&=^yQDXHuL=1*iIcjoIZ~a2wl-^^y5Jr+!v#pM-D(;BE|2S?#uLCvre&_1e~u
z&}vCuGvB1vc(Y@!2u}Qyd#bwp{oZyQL!c^7M%#ss`St~SQ7Z@Kg^>SpJQS0Dj>xDR
zq17Z{yScwFl$<;7T-Aa^d@=V5dDBIm>^-67{Fkrfjr7O+Mbx>8X>Z(a=+&n?99gS`
z9uY6}nG^Dqd=J7qYP!_?>Hc~5e%QFM+|a#_wW~T(t0CCehsz7^e+YhMqWs%M$8#L+
zF_HW>@wi9)lXnZ_5xcJDrubxjUcDUt52-K2x_$Z2-h=4vH8@*?`CKcgUEfRVsg6B^
z|9OYvIm;M48_a(?tWSMpA7iuU#n0apqVR&o%v0VuDeb=dc@*yWw1lvLE}V0ail_H(
zV^t+ZaG0t86-P4#xjG^&B6Dr&7Cu}IxP<3lv_!q}{YI}2yA0*mD+c&o<b4iyGK0|J
zd=_}CI({ZG0q*7-U)eVKoPG1Sl!`A<Hmh%K6b_rBkoLEA)GL<4rO$~292e#u$OK6r
zbX!#1b%2H@8jy`fNJ%%EaKcH)c-&V+2&t&{hYE$Ugc#{;fBVx?!3s)75}A<P>Ol5L
zCUNpQgID*MtsXQQe$tmZO-^Rvj^x~}JRw=5FuK@o;B&~QYW%&vEm2=bLg%=LFYMK0
zlB9H`>pnd~Bdv!HZSTtzKJ|k@Y8T(+NL-5KO^=Am(T<L=VSBUVYkjvS$VEJc#_iDP
zSmwa4pK!`-9YCk|8c^jv*#SqISh!KE(96$r9t79$J}Nt#t(jQ9wAkzK`eRyA?bm;#
zF9gQ`q7gXP&tk`ACEgd2LNM7^WC@anS(2XqksEt|&Ll6ej(?pZty$yZ&p&$zOxxt3
zfrnOpD|9CbRg~9#?~pck;0jGx)VN>C@c+WKIo;piEMw~WanbE{JJH!H$Jk0!gIC4x
z<rZXx8};wq`2AV<r;)U{WAX-epH?1*QhYxG_AtGFO?T@0)O}1zp5@gCs|-`~3If{z
zB0^hYG7rG3urJMDE4~1P>+TP~j+zed2_JCh;Fj%J7?_{Ovd!PH9Jm}Ed(G#a<`F~6
zi|b4-l22#d=V!nQs&}=pN$T{!iEzq({5gKLf9ZrbN1XR>L?S`C2`Ul+9f;b2+}O@*
z_spsaUMu57`v(D?qrN`#$vyXUey7&zG2;`4l{JQlxIB>mR|}klul~NDIrzikb%Iwy
zPid}p-lJ8$$JI>_*M-_oJP=W+uQHTxA1N~<4)oGm`iIgj*B^09VCjX^c8(m^SsTF>
z0W}=<desm1`)eX%J-p41lTHNFh?@Kv8V~6;;b&|fOjH*{kd|K9s|+e2Fvf8@Al+{Q
z$!@P)SB(cTHS6sTe}te>U(!cq=|09>Qr+zU;^q42^<~jks|J7WFu$zo9h|kPu(Yz~
zT`T}dtO;{|CBym#G*v2BQiFUXDA4Q)dJUUyy+zfi?5y6V`8!Pjj|!A2P@pb}cON8B
zLH5H6CE>`sfv13H0fX^*EB^E6Kx}j)hK>R*G2Ukhb+xYUk_QCA;AbC;52R%^^v=p!
zAzIIQDYpt>YWMO`4i9*()qEzhVNMSl1^^IjFF(f-_|s5R(BV4)gVj<-Gkv$ECpdG9
zFj6s~i|WZ*d4gwL(H)bWKMj~pA46B_VtC|Xwu2U$&c2D*pC>+uVqNMN_@^t5eqnZM
z6#%usGFQ#EwdAqrv!<Z)2E~}oO8woiJ+?&-^X@>8iIbgE@C}#$ncTnWr!?ca&|Z?E
z<bJ8j4`XOEO=(Eg06aw1B)*l;WG*X;Eg)cO->G}5YOV!!Ecn&Sgc;Rwou}SfZ;u07
zHU*WIs@L+1fsGpO+B;Ftx<{4cW$`l9?*XPr45?-&B9k2!o{Imwx$4`=($-#h(VS;v
zuF7)NH8JznTE*DGY?_LN2H|qJ!c5ci+veE{aA^GJ&L;~O><rE+a(EhAnLp-2+Hl&x
zJ{rj4EA~&)I_1y<cM3=F!nZ3NEg4J2b?|t<+*JI28q5m!Y<v>TC{MjdT5}-w*LRYw
zJf6Poxo1(ZDIruo^Ox`3$Rj2));1G5$K_|wpI9*SRsW@gqXnaUP9}~JGUvQ3htx)(
zt6k1Hv(%}%cUd4svxBh9gXtp;COO3A;)M4*x#r*t_T&;W@bM<teOl`AmSUrVN;uSS
z2)gT4&yB9)Z9@Q!Nw7LwxG-!<e#q(RnE4f{gGe57!wFmB`(-u(l%ha=Dc#a+kYZxW
zI~3)xM!0lipekE0AvgJ|wWq>cPd0LJ4O6e}eVcMWHr7@1)9(6iGJn>x?=yGj<*nA6
zrQ5Bp^t;|DRo~NMvv_HUe)TldYyM7sywpS{yRW*&o}q80HC4rSlj!;u=jTz&r)OQe
zLa_SZjp!F2J)d1_nemLYKvGvdu52X#Qn=5wgu(@_o#=OH6#R>01mgp!mVPv;_f|Oi
zbSBRH4e1|>@lw``iNH%g@KDy_X}G0#Ua#4!<YKY1GjupuPSIlIU2O8^?c#Of>vc>;
zMLVlYU0KCcR<O4Ez15WBXk%H5i<?@teyzOHu%Ks|n^pO82aEmCmloBbS~O;aqGX<>
z>w<D0va9QXq?${bMsgx)jaBRilf6D^0KaIJ?1J{7(!ovtPTJb23sazp5n3}Qntt=f
zXCnK3OZDWk)wmrh1&V!H(Q?&Z*F>u;$6Zz-q!)5EG)FQv|C^?})lp{ug%XNb-vj2=
zvQU=IAniB46H)5V;%&A0%z%foLT=rOzifP~6V^5Da*OJ)qHjG=+?7F|7fZF)(^&1U
zF7+lHz5+^5@9!XPK2Eal&%Yb>$U^EaO1XzK;U>#X-7B_N@v_g*YhYo}Sr^5H?0&3S
z4fw33Z|&$nZw+~HsUbgJmB&@*6Xkcs(GB$AgBo!r)nd4!c9$Wd5TlD-)Ef3+dLzio
zIH)(mT&-38O1w+7u;1&MDG>zEEN2nb#7V2@x&QG|7nfY^q|H}jNzuhmzk*~<=4>8n
zJ?@RGGLYH%y|K|6C}_RlnV54}eV?l=`>}SWW5l0dS#SH%_!zCLzh)i}!@|K6$=36b
z#CuiOHmMhV!|Oqr%6=m5ibK40nd=}qI)4p{!JoA?BM%GtOW=6#K2c)o^MvevYj5^x
znq86lwxk9Uy;eNbDP>6x{Pd(?#6xDD7(3t3shQro_};8ZL$e7&yc>|73H<A_#2CoZ
z_H&&6xdoV;xi+}0a&7Av1)qv|Sav9lK0z+sK}XaA(wA%jy20VjD+J@U8dR^Ip;>G2
z7>Y+v(NOzX;FA{pTd8RmX3H=2d@>sSCmYpAgo4&jw&4lJU7c8yHd2S&HT$D<!aYez
z+0i#-O;Bl}LJ?i>$?7$~YBOQ^RrR$`wfQ7YCg3A%O+SmqBV-d7k<pI|6ValxmLlnk
z8LYm;L2YS_*#=PEf$+O}F6Y-j)S9vattuY4sEoWWHqo1q>2V#pE2BS?a__Bsm%&DD
zn#PCy^W=^6+EW4)s0QP@Z?hi{D4|m7y1(OnMg1Hv8nd?M_cpIS-Zmi=Qb&5Ljoo*g
zcK4X4^QGvi5Qw9cjj=p5y$)6a_()(?sl;Bd0xB%6t*L2-FMf2Aqv4rWlVm37&NXb}
z;R1v510<TS6;IOXaqcihEcuTP`wFVm<63=|JhF^7l8>43wo~!s%k1$o3IkZ`ctK}G
z<*2G!-la)bdG;ZMvhj3{`YNpc8@-^WNy-*)9uO9}`|srMGxz+K&*_2cm9Fy`>H_Yq
zBW-%VXvl2q(G#^_QFh}tF9D}3pBhS<ce+SF5+-EvdHKv6)rSn5JY=g2pdlu;qYoyO
z%8q7erjgoll7TkKXFcrc*WtFX#;2xAl+q#wzkPT^vyttWR{uyPC4%hgoI9|&3c*FV
zbUVreT2}B_eu0->UxsQ+Qh8S{KYlULbXUr*$aHdPDGJ>&oo3H5<+}en9S@R(x$XqQ
z(vQDxEhw>=2m`B#P*c@O4Iu7XTydquB$6$!eJ&$_$lz8na0q|fFibfl_juS7X*Tp$
zt&ekzCwF)kRlK+9{dc@iV_m1bzr1+s;JR{VU-&Ija8F!%+2$v!(69vc%<1!Nm0h6R
zD{)EAs3@~<mCBL3Hi0^HVw0TI-)f1lr~L-PHJYwP5aH=f%hBD}9}YruP?xGlVqLl0
zZZlY(HsPLlqi+5AU%Q>O!*%+L*b?y%FRzq{Obv7y+J8yxsD(bTT)8LyXCZ+BvL^MA
ze@MJs<ef+Awis#R>*Ke5Ckx2w%5fN0<CRLp1FV;C#;r{6h1_Mxx>Vcn+q{|jnF-9x
z&sn^3d4toTcK5EVuM|!7KPl&{ZgFW`>wmV{_-F20s1o4oOr0P4=f{q7*|7GZzOMX(
zQ(yWyCwnRM=P!j4gtR@cXk6_`G2&h#<hdWzwTJHW4$MRhz_X?K?iK##d2V`DArys=
zuZG2~Ua3}!*759xf!WS4XWy}ZRp@{2b2LGS&Q#AX2wD`yzhL*`{NXP3Bskz|OgB!h
zy0Bx+J14d(Ui^fvEb1U1Blhg8sCLn6-s6EC_e%NkD@Bfb(()4?GR8y@=5?a&VKb+V
z6Go8gxN#-yu#F1DOz@tzu}R{^zSg3TWX!OGajICl`NMo!L=!q9eXTA&Y<k7EPJan#
zVzX>x^lx}a%+NfUs#k{7i6^P_-IaoN^5a4whazqDumDa2;LaAthwO{*6p|fL-nJUX
zwTm-LIj1q%Z#l?%Gwp6EbsV7LV37?GKOOl&yIwU0dZFCR^p_!<1tm4QkY`eiH<Hj6
zwZ(1f&;f11+iV$U`n~kIZf=#TOG7gKjCxt}edB6@c#E*AIK>FwpX}*l`JZo?NWed}
z(Qmhxs}^Ks)MnM;TFbQ`Bd}f<f9w_LCDxZezhNRPCMz3*&KHk$6uJgEo1~rF#{gmR
zJKJ%ut--z3%FVi;_C1Pkn4jzNLq9axQ;!@KioLa0zVxoMXRBud?`km4Mv9H&yqab9
z+Nc7CYw+ZvgbDoSr9$zrUZz!m2qHYl;?QzIorGr<0i;{Mz=AfB#LPv|v(d2<se@Rb
z<jsj_z)bUAu<6QhDg*M&tQ?j!_YVp;{W)VQwG&IM>(Yo;h1^xtJxF~_<3BWu{`<Qj
zRJF8(uYNvo*4irD=&Ey_1ocbKS-U($!)e313*|ghu{uDBZZB&laQAYq*%t(Hc3z%S
zx=8LbWNwtYrCfYE>50Te<A?-s58)>`!xD-(1x}lAz<{n|kEqHo;2X(ZEOP#(>QJds
zAgXvbI0u9!uW>5k6OGW`b*Gn>$)&OLl2GQ)3(%QvE;BJXn7O3oa&o+y*s#(&{jlS7
z<2o}4Z_^2kdAt`GdS;Ur(_)oR=%f0JxokOF<n0qUqvAI!{kG5F8X*6Cj3lC7Qrr^}
z;UndaW}Q~AV|cd5E*#t*ntg93Dz~(bz9;TKp+iquXbiKD0|=oR?Xyb+yfQeJ>^80t
zt@u!G8K_?CHkVt+Sv*8s;0k`-*!(#>-RT?e!|8t16?{Zc3ur0)a~tQ`G0w8cD+jr(
z^@=rFbpuTvp<%G&@V0kWu#=OMu4X?Yri*?Se;xm108qIZs9mR?f1`xsllYXIoBi(;
zqfc3!4yghDV6FL!=4FfoKn#j&L&_+y?)yzescGDQZZ7qoL*4_8ih)z{EiuPHBv=03
z82<kqd<Q`}Il11pVKwbY=D%FYqfdJAJ*V*1_p6=qR&5NYD^#`VE1KA0Ccoi}u#db;
zNo4(Q4BGz5=Zyb5^$BmJUv;j=-UIXhDt@xSD)t@m#^}N<uX>3U2(k}ba@t*=_o%?T
z5gO)Mnq2czX;-sp{+;+#OG%k|^;y{wF1+BEwqbr@9;;NM!>?uf>GP3*5-w#rvs;<|
zXgif<<?%S0DnjL~yCkdDzsS@6<9Woxxm<j}0hzvPYS&<e>46p9o@H8h%JhKfok=%b
zeQmg5qgaaBR1c8$=$aB+VxJ9OTT0AcyzNcD`$2X6j*}qx5?%=mmr(s~(9_M+hL6vo
zBy2#t+LYC`XKGe^cK(66t(8Zm?-gHm{8PB1yM}lrT8GYuMQ1Yh^in>bVlL=~SEj3>
z@l5FQN<~W`^v|w{Qh;&iX|On*NMvziq_xwwlwGSnU>Xje=a?_+nvTjmCbr8cOUmQw
zK0L}V;I6>>$wG;+NP>wwm@B(d8JVeO6IArbeN<#R{wsIGLb|m=hhmC9Y96g2pAkx@
zly7EMUEQJ^Ou_WpUA2_uk@qIWaqP=tmnm(LU~s%RHIC-MINLBSK!bPR>bXLAo<E27
zzVq<@?L84``5vV1yYRbMz6J~%B+;pn^7z~l4Ty8l{3%j?kEJjwHAF3>A_zdRhJo2T
z1c&W^Wxc*>{8gD$MdVHVE!CjUser(3%Jb<$4l*;%KDA88%f1dzv3UPmth@5Vj+CIS
z`Apv1D&q3e1{e*)S?(tjnQvXwXQX|$Y}HE<{pJpe2Jt6~^o8$h?{0#>qMGa4VD{FH
z=FeMB>Z8_HVg#5?^U1fFXldjk{7M#PXJ=PeZcTNsNpLxpy9E-YPBK$0O3Rd_6)STm
zt8e<QFK-=P{9Gy|r01|<aUOZso_FU<wiB&by}K+gYUUf%bgN+VZas0Z@Xs!$xeVKY
zLZL8y^W-x&BKmcOSzfZ7zVbqiF2t>lThd?M+}?`?-sN+~Yg*_5{RicSK_>F?jc-+d
zljrQr)K^X9uJxC$=3eRoDsbUFEkspnk&iZ9FTpM#sSi$vp;NtX@#2ir(XDoiD27Ny
z2g0bRlZ7Q6xq>IVJgA6Vo8*PLRm9lK{#g?;j2pT!B?K-?hWwDN=QnKn6iN{g@?67i
z-!<3dc|OXRZ^d}+^Y(o;^c%gLwIhY%bx7G;53B-Vs(?&COEL#`8>S%8@%&Nx)<oX*
zGqvW+I?d!r1l!;zHwPLr8K@xT_6tB>MeY4%tF~dWbk7RYOP@t*^sQeccv}A-@&Z>u
zKv`uy(vYc|)lCdbS~~R>{!Gc*u{fmD=X!ppR9jN$u!yMk^3R@dfZ3#v|DoS)_gjH8
zrj&*a8Ql;M57nq~oTtns10eY|LTA?D#X?HAqkaZ7m2y<h)cjXjXwP>WayrFv=Dv7*
zUrU3TWG75&pSU@w_vpi?He6J^FV_;`x#zQ*LtIe+i^|9idA1hIPG=@RDL1mID{tW?
zL-KnPmd0ERH2PC#m~Dnh?`bZ-)__UQ#RR=j7J4{sLehS5cwNsHyJ&~Mgz|)t&v332
zIx8VC9X5>~K_bDws!S-m5wh(gda2A{r4$)GOQvTO-)9&U_DFP8qI}lg7}aMa%=g<!
zc;aL1d^f2L7@RA@tp41a-Rl~gT(p|eK;&nx8u`u-oj~IQg<P_Uni#&lG8A?l@E%Mm
z`q15SxVGX!fk;`i8*V+Siq-13JIY=3;Qf4T#?L~1s4?6hKw%tlB%4|hnC`KUhs%-K
zT`E3<TwgCe`;7B6C)H-0QkrfbPxxHO@Dq5#D!G}{<S4Hde<%p_?Wz1^t*@8=`ei<U
zenRe<43xZAZ&%-($#)YZJb#QVH(3(#CYHpG*o@X6%svafm}IlI`?CDS#37q+NO>vP
zY7sNHF~VQj<H^%D98$Hx<<J#%oe6_!(HxMzwCW}KrfKFq;&?5@xxBUmsaA5xhSCKg
zkkD%rZ^aWXltobsC@wCG!N`=0q64+Jc~;2R@;Qz6>o`~=l~2;okj#rI)2z0lFo%Ol
z81J|AR4Q;(lJdml{s_s(pOmu3a55^_d%p4x>wr<z=^S@rFwL4Z)e|n+ZY^uts2Znr
zLfHv2g@$*a6&f0!&b*@pW-i|CH?B##@dj^`7I*hBF36&uXn))BVawzv;<iCp&NJKh
zpuYu@>$&!=i*eB2kA_Zm1b<yX1xzZklOVPaNV#LE${}N&EZN7ykOBMJ{G;;N9fAvQ
z8JPhtw)j~a1LAOJ4+x|M2Nr9qv|^7Ck6wXjTrQe$!N{plfgbITC?BtcvQEauL7GHO
zrlm!mMN=>>?}=hyy6<fBsVaycNRF|n&ws+?tv@I4BErC!lHozZ2PUxt?_5*AH-Y%Q
z&?@DZ^Vm+-VDTv}mIm#qr(=O(wNSoCLr?4agA0ofCWT&6&=qm}7HLR=Q7#)UCgadC
z+_Jf<cvo|N^Dof2?w!XzU4Beo{FY>1|8%28U<Jo}tLYyV8-tlTUQ=@oXJVoeQ<s|=
zi(jNVZ$1{Wu&&e|cY~rXsKUA~cY9Cg3zkEXU6)D^EW5-|MI<t};U2BrhkWB*9=U^#
zz$0Dh&dOq!O9+5heo3KqUD|IQ)b&i-qlih&PoWqXbq}~buV3&LL3MB)^S~|q#lQWG
zhb`lr+L*xn9GXNSy(~GJ*HL*}FYOv0>@dV|cGgWA6Z_!5zBG)+2mI8w@$p%o_r#X7
z6&f<8#msodbXqj}$a<J|?_Hy&7_Md(yH$y?A8qXN@#JPKSOU^v$2mA-lu?AK)rh0;
zz95Q4sDSnQlQOE$ogV=&`y#U!RRs(8B;S&3D9v8%mZ&XJ6Dw{T&fde5;x-@Vdwd?c
zi5y>Bz1&2u4o+Sh1X}o?O#A1)OLIpI&xA={Yx9O|<s7WY7_WjPc8M)z<eicF65G<8
zQM`JCRLa|E;(e$zcZ9`otA|(e-jv95wKxltRE0?Y3}IazP6KtHeZzU34Q-oHaGi}w
zi<?&?k>7WbDRG+uWaH9xEfUB?nr2Yau|Fpcnd%j8aslXo9WSDr{rrsE{M~&<LxG^`
z;eKiG<*JSx^)6p+bZE^b8dE=bBbL^H7-#7U^O?L-0Kk)>%9UteuCQGX(0D}%&mH<7
z#u)Gka8W7o^oF)<Ze{`m>nW95y41!490%Z<{`(n6T?QR{^<oCD<O3Xr;*Y*qjKGY8
z?R_@_>AsW(KjI;6@Y{0BxNkVA&Qh=ZskM~pA$0i<qn@5zc-+qd`B05x4uv@Wd9J|Q
zhvrwQ<-(J3BO~6}iJD7q+xVI`{KY%y$=}TXN7Y|2MBxWp!0^%tNQX#=bV>J;0@BhA
z(%s!%(%m2(BHb)VxP;Q(DP2p$0`Ky__rA~be1%`koO5R8Ov{JLz&l;$%EpygzuVEJ
zZsN|AmiVlBJEFOf?-m`z*D{Kw4|e2|{MU5Nyc06KU_VvhzXi|hVHuJW(S*ol$Qt1|
z*H@Tgk>g-GsO&zoto>mBWo!=|IXHRJI=HwyyLEkWI98!}sH$j7(~A}T;Qi}y?Lgyg
z*U_!1@OieKN}h_)4`uZhRU-&sP2ObI=(9P$f#+2tlS@OVI`Q1wu6GQan`V=AY2B$c
z!}#lU#*k+N(QiIJ9$&_W>E}!eP8R3g)D*rv9CFb69B;}{%f>3SfKx&4<CPufsox|~
z<rhOOIoJ#rL{arPaIl1Y)jSSN5$HPCbryt0C$|G0&K+`YWI>Mf2yP;W?8yETwz5Vi
z<2rXaza1AyiGZ$^vF8fSPAv`sp8iSn=6ntEYNqWR2tb%IWBD)mKfx70XFmKcn;~uI
zI1m3M8UDRUrP57YmCe0e;SB(1cdVy#yVEO7+CX8O64T`3Hb@^CbWwk<_;pHCnn*+P
zG#{Z&lEedB_7QN(Glt7*o*D9qUB=?@-3sK#Ju{cwRB{VxDX-~9`H`w)b{4eJ8QtrJ
zBKup}nTbIfV-Ur=%K0fw((@wTp9oCuvsBY3`AC@7Z`~-s{oJ0v9gdnD^LJel10u*N
z(rv<(7pxj^{FnysBjBJcX5ewyZn`Z<iC>Y!@_YK=f<VY#ph<CkQj@9IP+@{x%iB9c
zG`e3LnEu3Bk~dTEhQnJf^WL#L=|5hckvT7sUQXRIe5RzP@%KJi7?-CIU{%~1nDL|E
zUxX|4K|u)^8O=YRv0=n$p2+UFWl)JxY_h+wc<dsQl<-qS2=f1V0UYmV<eN_sEA=s$
z;fxP-REmmmM#NrINUrIdmk<vAtFoQ5Z6x$crm?b%lh@5>gLju-e+*^1CS6j`bI>q|
zT)9np`(xU7J#zeO3yTpxrECqD^qA;~>t8bvD^WmG)H^-wHVCZl{zZpnnF;X8yrK;@
zP+YO3_uxrh@iFn;a)w~kbcc*@Mx=xACzxXM;-e>?qW!{_JsU@rRJr|Dhb7@;TH2y4
z>nGB~TmqYbiF3A3Sb~}gGO$VQ?d{P-d|T<gjxrYO^`ApW$`4<xu-S+N&(qL9K*GXg
z-qX@PG~&{mDA8^DREFW!DoN;z!z@jrzT;ZRg?sHqo1<64DJR3-dyU{ws6>-K9bP7}
zFx{JR#4OK!xc6i|xZM&MXmJx9aO-U8DW_jdV`R7VA}BrHxaJ&47@e^v9-VI~zSbPc
z@nLfR@&T1gyMm&8pVQ#eNF<(jjvxVTT5GDI>zmjU@*{O3^65V(57FT)mH41q9=@lt
z3Cs20{YMK?E#O8cyNx{skA2|EdbQ2VKJZTF6;`v4TmQpX>m(8Qbn<t|cw_;*>F=LN
z1iwoHOo*Kx9AEFrTcXX|g$)Zl_oUtgJ{UcNzn8@SMh=h_Ur)vk#6z#hE`~PZ5u&<=
z5Oq9+$)GSzMXESL<Dx@ksO<q@M5$+j+ohhv0bL;~Di?3jjqUs~x!n|>{c<P+!q*d4
zat2L<l}gy8ZspFHj`U58gMyFjI3AYU!@pc+S#E`28q0Cu+3-gaS?9iN%cL{8P7Gy)
zp2WQlxy(yzVc=E=3#z!JjcWQgfrk&oJHDDl;oy_Qk}57Y+`<bVG7)28wNvyDdr@Ge
zbulUU=nBh@tbElqQ@VE8^2=!5_LI-8eQrT~i%KZk`y!3PnlPYKc@&tv<x|ypxnsKM
z3tOBW<$bCrDrvVPe0r<sO0mCe=wKCF(D~M7UBLg%WPsCPW|w4g4ns!ebfNvw@u=p`
zwlG%<(4tmlj7anQuK1aM{1Thb`BCEfpGp=FqcIBM5gqk-tKjMOs4=+`7F6R0?02d8
zwlp6!$y`%Yr-{E!_-1w3SsC0Wk0XLNFWeQ%A7qnDjxLEabf$(iRAR4Ug=L+F;uYi6
zfuqHXZ~U2WnI%|(C9!HbP9}%0rSm>V7SY?^`i5xr<#xdKR=k|f*jcDKG=wqo{M2t1
zw3sDzCtOa7hIW@Ou_6#YnCNHQq<_W%L<^>TzqG|M_Rzo1<9u9%uSc@#<ac~@)(44V
zpsS#iAqG7EDtJ0DdFLYIGF<ODx(r*Resmusnt8hZv>aGW2+epoG}@<$&nY%~df=Or
zw4UJyx}IL`C&d05;a|)P$3I6RnX%K%`*?gnIbKGQtPVYcXif|d4_m~w>{Xh~bEM`A
zmzNC2e+2#p<JBC6R{Z4M&tMdn7&b5}xqyxjj^B`K;+kGn_kSTDo1Ht|7Pjt8qbS?p
z&r{z%S``jhf^cE6D#q4!%w5O2j*Q}W>;zP-R{clAmRB5=t$P@jj^_y1E#hkJI~6f6
zOh*or*NP_HVEz*}L=WcvjRwf2kyjXYHY5vS1|;7|X6}OwPQ9I#j8@P=#D7^(bbir%
z^Dvt8BeNaycpE=6y;p=?ExZ%KJ+fyeyDT40o60s6i#95be4SL;fFVqYQ5Pm4n7c32
zl+kiqJa8``wKLumER`9;I>r#TZ`D~w`<CGITbD>DwK3t^xG@NpBxSIeBkP|OF(GQS
z2g>na>?zNuE*6!6yYTtv$PE7eI4SD!f6yUMvMbK>&sE8u%>_wTv?i#te#Nn1osevU
z#_~n|(D43wB1$pOW`rK<c)itjnuX%swZ2vIfkxX<?s_(`N0s4F8gQrmy^CI(J=u{R
zA=Au4Y^=eb6odw8gCqqzQ-23`rrvmq5K(_vgn&Z*V4Kv+3Ga6kfS3&!GRfUQA&P;V
ztV7mwb8DIIJfm*CF2U^aY!^sHt0Y3eVXQsSw1o$7V5*3$!5=7lvuO~SuFiGd7?0FC
zE>C;|7ckM5$xl1um>oqo<f;;j6tOVv6Y63(pDY!>W`b^4eRlS>!M<BzQo0piF~ib7
zr#q_-!vBvSH)GxmWy!{7v;W#Q)>BY<^sG8S&o2$+11zDa8d<n~d=-|!_NQl^c~$qe
zje2^18f&Sy@&Y0M+&+OV6|?5sN0hG1?Vlj}mkqUsiuMROU4Q^LCx)QVu9|XxETJd&
z=XU@&ln|+p1~WuHNHRNX!T38&@`aQ1@320?hgTM+6lSXYimLNrR(B%urzol4uAm3P
z$VDVd5N+EW1QSb10N+T*I)^{Lm5WT?U+GHU=1W@Om*641b;kx_wAIfMQsh?1n8wu5
zqWY>hjBYA1xtaN6nVad}5v&89?^3BQBTk5V&wy|ivNw3}vm{3F1XeT@6`z|X2<LG>
ztwDt}_7zA0@ntso7gn-~!46wPKNsJ$6PEc5VYQUnbtxJ9V%qU&KF&(}whx4np8V^@
za%mnbQji6g?5*1cYoJ>_rWZ<*;r-BsIf3mn|3ify!#*b2w8#HYAqUpjb&#KKSNB6*
z^Of2g%ME(-G}0D#bh{&};k%CrKt1;7X8b}7*fplVp*MHPUYP#?y=&9Y&CiP%)+W=y
zJhKTFV85}4U7k&+f3c9dFV;ci&5z?{2-m=;^SiBxNBe_gJ-=7w<MuDP0Pu-*<B3y#
zq5(O8qqfrC7?1FvU$E!LX|*{bEr1k`AxYZHU_k3?<3rm9s!WuoLv1x%I$XjuHm30=
zCWnl9wC2<`u`C`phLhME7E~&K>OcUFOQ~oBfR{APY4OzdJ<br06S91URHD{CZAW_p
zY`y~_PhHhqWwK2EO<{$hL^0qk{yquBwq|D*5|6pbIrQZxYV5WFGJb%f@PUArB<VfN
zFXm4Aw+K-z-S32Br*Fg3(vfKzE+#Y&6zq9q{zQ^H_@a1C3>xTv@?qUqMh;jJT!fuK
z1FWe^NUSyW`5+Zpz_j&r)U2{*lGWCTK1q0)^E{<tDj>q&r_G3Tm@i<h=fy;m_)qV6
zP0Th&Y7tG%(M$LC>YQY9JykE3W5VA@9QSwb=CW*z{TNV;`V|YWY`UAANy>80bfoMD
zNRc({n^07p)n2xDoBMpnOEnl*)_QUU&Nx)UenBLXNbcha;|Y=Uf9E~=UVu-MW#b9L
z)=X8<^LF^>C3o{%yDtDB#+ADc&}nMVEAs{@JPu$Qd5%H~{1EHccJ0N9%(WHcCH)rQ
zhguHAqY!=LxdG6PPn}y8_TkPXBSWrGb64Du!2V&)T<u-}hnVCSWp}sQ9^ZlR30_0y
zjoF&1Trq>h27wX2X5RgzQ%2KaPB)-jcx8%B`~wyisLcARUQs^?ju#dEKx?tyyiQ29
zme~h@>fNqQQU}e6Pq#!FAXRYjQ8XH=M|aLHa2k(0C}sRg-L0iJ(yUVa<6&PhDx*AG
zI;2ozvC2}XM?rbVg0B?m_r7J>4Eu-W2b@YpU1BwS0lRh9-bPjg;G^M)`7^A%xI+eP
zpJuL>0@*b5wvZ=t&q%IcT(&_${6cRZH*?Ue%bZg;J9?G73Rk&t@+#k0{LpOKtBJKd
z8iBp*b)Vm@CA~#ULbsA91P)Pg8yB$J`?A^+dkk-9^Waa@IIrV|c+x$Ck&w$ua^_vC
z#q}^GmmcWU$bwWbg$%Ko>KftRps9G;uO2}*?Lm44OT^8!IZ75(6-vM9NbMqNP=hSZ
z7Y=cqU2w{;j2sD;)~z;nS$r!hmEstRAQB`D34H5&PuWH5kW(1Ugy;a?n`kBN4ocaO
zCnBMW7s~()X`_hVb(0eS&^xO7#Km#KSP;12fg_#I(h_3S)OlvY3a4^L2zm4a0?{G}
zDl82@SJHh_%G%$l)N?CD^LN7^CkbPdm-qYY#P^Zlw=~Jpq%z-SPGys)#vLI5%aPvm
zP={cij{t)Py0*c`Tp32y2CkEIKO*N%Y_Y7fNjugFjaM>)3!d1v`8}`XiSKY|?7mMg
z&q*3rxe+gLC@mB4MUb^zCraH)q6B!9=JOn6>1tb7sqwy%<!eZ-DQj}oZ$xSk$vJV|
zuU9;|ny}}B86C{7wNu#qRQz87_5aD-v+q0a{<+R~kqPGAj<WDIwgfP|29Xqc4^Wgi
z$D2!@)!h(hzwb48u;ils@qIwQ@cUkXF-o%2k0c+GP&_8t;u?xrab%OX;URQE+z(kS
zi`m<qlE!WmV&Owr&S6Qshg)qHp0N?m>@Q!5Khpq%^}nvBxiQHH)D2G7`$@g0HsL57
z=D!z>n`flGOS1mVM4JzSmvWKhzmIoUT}2yA`YN}X@z5Ue+d4U_wTtA3s;{AG&wIz*
z5>xYSK-w_#bolu0z&tn^1!)fTn?6TUc%c$1LoYr2PLE~lj1fTSm>Ornaz5WPe|8Ap
z*HZv2?6WL1^Ko)bqXCW>CfD|k1Nu)G(4Qt&f(u@@AI=D0o&s-s*y01OkJby%D%d=I
zinCqM!tJX;Di^SL{W#0a;|;4`4$QtC`qJRBF0ni+-m-UN*MrW;*F0@75jgLbo6_yS
zG5r3|3>)lm6Ow<W@q+3c2KauyEP01k)3o@5H6VX(NY{P}UJZ8M5sg72eDgEBb;Bv7
zgRI%HOOU2>6yz*QW?Z8V2FD$|9U8y0-wrgoZ=+5%oHmFO0=_jZQXy7SbpCEe5DQ_m
z>gpq_XBcowNJz0gBO>5PUKK{OkV_qvU>p1O9>J-N9m#eH5^6~35PklGW?Py95C4)1
zohVRG<{;!`C2WF`5~(|?1uF0JJrqnuuhf51#Ho)hC|D=GCi#%6Ar(@wZ%9en8f-qg
zB`4&EGSm}k=2qk<D<77OeC%MEIUM^tiOej9W-Vbdc3Q|JGA*eF>LUt?FCb{#$-yKn
zJl)(9TnuJSIZi_IYJsgRNI+})z}$4#jsR3fCnky>?M=w0da>(=7oA*#h3bgx-~#(p
zXq$BzU4o+f2pytH8C8{+h*WiIjdQaIi}hlDq{f=R>++1!<H*R>2_e6_?_NMfw*y8!
zjxeZg6L!Fl(vudjWjap{pT_=TqwfUVm(JEj*Q^r-g~7isr1@C&rS4<C<tKX3&qG(D
zpvZT<6LobQv3s{yMlB@ei1=tTlqCp1$hVOTDNR~NqnN+62p}0JZYv$?IjJzOpo+Mi
zs*~M2`xjEr^`cCJtu^@hvXy(kCDy{Txv;*ZDfi7<1mn0UmK$FKE+#+Ws!z!IS!o><
zCAB9Xd-M^AeJd$)!>s`%p>7)5(*~ivG&zv?I-iZB9sjUK*1P*?_jxQKN4C}=Jss~L
z%xX$au=HYd)-Alxj+r}K*~Fz=-61J)2O%$QZHgl&@#pIDg(RqPxeE!i8Mxk6x|c(}
z4Z?u_Y4!U(_s#Nm#!*77=UKozy(?YF^S_@jpYH=-WpWqJbG1xrBfC-fWgpoo71&$;
z@r8nYb*AsLS|0TnJ-G(!mxe0tdr&7Va&#^qVl(C}9h4paat&$@m2`5gyX8I3eJt{*
z#)2L9GmCqVNNMs{iY3F>oG5R*Q!Uvn)@&4S^l8a5=QSVpoxTG#_MMVoBZ(ZwrTMlv
zFreb=feW_M48x5>R)X>QIVCfTxq^R#kjwxnp-hEH=`3A;C}x<&h=}4S5Z)m3FD5w!
zy+3Ay`_pg^;J^k@0F-3GzmWTQ3^a%ZpLdXvaL3HBagIn(O=ysV4>6!Y;A`g%^nkue
zV$1cVdb3UT2Ope$?$XcVYzK0M<)7Y9SiR%qHkg*#MK7AFAfr`H9Yk8P5j5+Ej3UX<
z_EFwz(xqsiiGMqx3{lQQa}q^|&}R6#$~mY={}L&ZK65Ydq}n|`X-DdcBRDY|zsza8
zor1q2Aindq&hUS`S<iV(26TADfVO>T=;3(<d2*6}d&lQu{F}nmQov<`()z1GEFWyS
zrLk00TK3=weG*t;!GioNr?m_^X;n82X&8z}ETEDHrgyC8PX(y{7bFU5Hf^gwK^@B8
zIh^e6Mgi6ac3*q3D1*svf9rNrQQeKH>B|Cm5_~TZu$=~4y-ErozHw^eiNqrQM1=^q
zh~7;q6|wzoEkG|gUKW1<#Z=2o`){3<5e?KSZaG<5Q<Z*?Uexk^wf!QZF)MH<oX<X=
zJ*}e>3cJg-5MS`hQF;|#00zg$9~<Gyp6gD>Sq_Jc)1SyMj&I-!Hz^>K;{Ym%@y(_1
zW(vD7>HQ_YR}4qekqZYw2Y-S-H8nJqm&t<!=S@9_bwKKjS^p=w3gLDSG4!swDqu&}
z>P7=(0D0j=>WU`bHXL4#nPFd4f6*L$fbg6*<4D#nt-r(KHU0H7y5?)WCchn}Gbg4e
zL4j99_ZlYLKRb|jRfea%o~*%oYBEoP{%INEVt%KNL(Wu3WuC+5NXgH8Dkpi5MK`*l
zuaLQM=IJK&CieSZ2loq344mbY3|80@!Hjhx%kmcfZlm@bHon&**0cNGs~qK07vhA(
zbhYr&hBOX!J#UqwAlIq_Uhus}{A2UUaEK`45Z59c10_Epeb<`O!S=BiXJ1!5!pBs$
zxd>a&34q|n#<}fjlFlKO8?~-GY7B!DQjt_)>^}+?=8BD8oI=zFvbG@UVDyW6$y}_f
zNsju^2zdY~u`(SkJ&eC;&Z%9@aZ06^7T!6=aTgV6)9KKsEBHohb4vi&RKy^G&ufuE
z#Hgc-idweOFci6VjyD`G6B-r_aWP0jQTgqEoJ&rn&s^~VQ8swV6Pn1dyD=_sYpQcE
zs0TowNOx|JuOP;RM2fN?03+3|n(i>5Q8>T%uQ1I=A2+!ZGXDQZoU!{+V@TaT3#&JO
zh?d<$(fu>M_8%wQT8AX$F9mER0(gBciAo!p{N{W>^gl(Su;&aSqye_u^XkP(zcrYs
zIC#l!_vCA-G+JNPkE_1NCi%6!?RN{FwRxWUPcPt$Ji1zDS!)=|UhQJx_~+_0lFM2O
zpBJwkoV|q1>%0*u?Tg@40zdjf1BgHG&f@G@<3vsRY$&|VvO;v;pKS*qG{a=@{_^EM
zTAaG1S8afXXMgi*qODjb33ns-G(K1pR|tnV41pgz(g?s7Oh~*Gh};}&z&F)nqo29X
zP%x2*a^xtj>86;uJKpEtil$qM{J9$UI-l7+o5~DLL(J=Zv-OTM*LR<^ZF4P{?l^SX
z=-ydH`lxlq7KEw=pYn^x&xYjhw-{8HZ`e7TEI*T0upN?ycpiT@eWw1JJAci)s=$UT
zS3JR|@A{`Oy?}#82zWL#bzeRR-#Jy#YDOdT(**7z#veIgpS?$fl_-PWGB)RIbTDUo
znx}?G<S6BMQ8<acO>x&N*=$4Rn=wL4<)7T$(^w=J=qo9j65y&Eb?{_O%oc_<ylue@
zj#F7$F|NbktX3)rHXd07mq#xXxvCfBe8;hL<<d#$@38jqQlHt|-F9*IxN+&_C6Lu?
zc2N`hG4=29n$@Y|#^G>&zV$Fb_TB$rmFEEMVqdoZu1Tcynz28-mc_|}j~;!3{Zb(9
zqj;LY?oXwRvX&y=Uksd?u7zQ<kB|jk8`vhl%yy@q<jk;OmLUF(YOZzhw?TP*ANc_4
zXix9I|1g?Hx96KUy2JH;Sl`XYCmawKfB;|xE4PW%cN)Z>&)DA}eCr+TbX7-n`=e~&
znPe#_&dL!^d-6~%m`vEv7Vo~dL}vXibJr(Fp>I+yAzJNV6YP!9UY_uCL&Z&)6v`xN
zHxN$|gb&dwVHcKeLwxX(VhmpsCvG;{xb13tcFv17myC>^b~F60?wE8C?>pSUA6IG?
z@;6uFYuFH+<fvA_TH0LJJi#teRqq_ZrH^9X@F@qY7_fP~cibf1#=q7P3F>nPF~4|z
z!n&5yi@MK&4Vf!mxAwdkB>6o5RyK+8t7rX>;fqn7-7RWXCS*7W<2Y3+y2rG2!0CKW
zAN>;$2y1n_X<ejp=7#+jj|F_~YuecNxlf>22L<1=FVEo9$U2%>!*!jaB`Wau%+q*9
z_4QXBrl9onEK%T-YV&IevX~(${J80r4K>`=y&2cNz!xs+;X#4fWfjn0?DF#+gbEAm
zBPt6DJP26oMdK5LVm}!Yy3~Ou;Z8<V=Q9mZh~OxLFrQaRP#pYZRKz3E0IYBe7hXtS
zQV}Br@1a}B7U)4b4_X&HGIAkr(Zu?_MSg88`!q_$^v9p$Sp+F#vpesb-iUWhpmKao
zXwq0c))U24c*FXVL~^GUOncSbWEhYT$-Vi3Y}@mE$c@KvYm!_*-xUu*cOsdWyChEJ
znr_0T3sIKsTLA!PhZ3_XKLR4xB$wjEiPHWx+ytzt5<({Soa<^7eW~0cW)>`YG(LO=
zii<4XB6S;XV@CjnDzr8#%+FI0ovpzzr6oECGf3dSV?Sp@0Om07ef;91kV{eUA0dLF
z^d+^1-xH52<_h-XimW#jHus8oXJsCO?n{vjFDfS@7QiMyzf|YY8npVmA*MM@0YVmO
z?|X*Zi><3;JL${+J^i1kB0)K?%#<Q1+J#V$rS5M4F-x%frF~mjORysU;a|dotRKW)
zMF8xFsSJUb!V!|U^0)s|W!bj>3_456k;dNpVMU1=coC?h6(YKDXHkaO<}S&5wI__8
z>O5$oq+-W0UB&tgRIMPe-8oldYS#pP%+I7lYy<!Sq<u#lI0r%-Umvo`YZ=z-f7RA9
zH=#AON!Va%DD)@ojlq%c{D#_mxk`X+1pQ@F31j2!>60P%VV?L9ohqCTT9l!dVw*3K
zd%smV-Br=W`J%xweECTO^=qnD@b^Sdcy-TH>9)Wl@U-g$ieL~0B<#6s!?dME*hDLE
ztQ||)<5jTU5~ad)?EE(nV<4nzW(4lHH=iX`HcARE;AKA1g6@q|cshcyu=L1t4f258
zd245p9zJa43F}&Wq77Up7A$qc54d|#9Vx*D3)GiAO3k-3JN1=HI9hyG%EL<~8eLfS
z@+rfD_r~!3VHUMxCs|Cb?T4gCgCylM1X(ZUe8EWBOV;Z<9W8kvm)qz7*iT6|=%Uco
zvZQMd`YTgTbQH71^l}MtaHf9J#dv<D12B9VSWUbvvGrd^T*!#PaF&j7?Gt2XyQR%u
zeY%hCIgTP`^nnP0wv8^oo*n%~pTD<=`B4`nThqaj82XVPMRjYHl#YLkzMF#(@x96o
zWxqHAAao%qBvMSWYto1OTefp-Mqs<IT*Hn`*PGx|^*2yV(t7Ps;T|wuMVN43%m5GZ
z{J_?#CKb9xk8v`51rDMAPdH7*Q7MUkTLfZBB&S8cK5EYSI!L_BV15|b{A>YN8kZ`W
zh4z<51RLYW2)<@(^#`jk8?}LSGff93!Shj`u35KO-{GjUu|Ck_a~C?DCi-r{CgJ&w
zoxi51fZ~l~%u#+{Tk4}Qx#9emD}qNf=n#vQGo8_&VTW1x|J5Q3gM}rn8c8oZ6f+Wh
zmdmdB5_z}R%aBG|?%q~pDtz6S<YcUop^QuB^Yjj7=v-|U6ZXE+=YGvF$YzW>Y+R2|
ztcfR!>lTrHMMpe>+{we9_3$5GR>qy-=VombpT24lcgpf<0<WArd$e<V)}-;eX6`pV
zoNo@htbAW(2iS<35#9f{PqA13akp~s+pB);l=@zXd3LxP%ASi!yaNV-X*$_?IEC|K
zZRxR40j+R<&+|l(q_qo*C`^gslc;;T=ERbIl;Gq|eNTj<s6_TB1asvFsmek}#>lKh
z)HL$?9ViGsg;38}$L0y()^>iClG#FytgQl8d2{L`SOBV=e@sN4rcaXWy7z*o@T_RF
zsfAQRgm-1wXYj~kXy&oy3!_=4>GVl?KeaqaSoOxiX%1a+&tSQsrh?-RL*8b!Ha7ih
zV&bFS-wb0!EwgsHo$_9)V0tXk+$rK;2Y47<UY?ddwUBXw)hKFKwM8gw3cxZt`oNsq
ztp`|{KQcY%a#HJ<`Fz`Zr6sW2PrFM<WWD8mwrE1<k-^FZSNb&!mHnPFr`X{*wS#x_
z5J@nd$)Mu>Ui6O)*lFlng0TxXO5$8^`>I5+eknB@bh&f~zGj!k`65*q8k1}sq=WcY
z+>jSj_#``i-k+4@%Cf34FiOAesBm|&iA8_C5TN|3k={=-yop!ZOg8GE=?VWIZC0&+
zvk|?8!U7tl#ZnRT&%$P=EQV8w+J0}di6A&(dvDNv;~R$xRHAT5<3UqvA6V>y>ptG+
zObuq^FS?`|K)uN0s@O6^n;MGSlMYUBa(qZ`vm*V+RU!?UATU-+wA@G-=wuCI$xTPR
zX9rQVS_tK^80AN5mz6RfICU91u`MTPsToJNQR;0Y2xK39u#RQRZ%O77Cz44$`Z51q
zNU%-6PahK`2%#lHQz!7|$w@R)RN88fm<T34^PCLjt}}tNqS7WV{o{6s+9-fT($C|J
z33b9Q3|N-KqwY?oPjvj--}N>7*y1bPA?esr)T%1qE%@Zo%WF0H`wOfww?lJ<EMVzQ
zv1!|qEt?)@MZOJu`Fpv5_06W7j|jm`<acwgwxh0HHE2RbO(+0DX|v+P+00F`0X}&B
zEPi5ymri;A+w=(gFRPl7vWqiHSd-*xdy6NmA{*kY_}?b6K1Tccqz@LS%sOr>7+9zc
z2L|}0woVrim6w-i54^(V@h`XKQ=YXh%dqkqkk-iio8W-W<?cnm{lt=r+Hx<Q-Q6ZC
zZ@PkxrKO05{Xnx6<Ok=|S~goB!q8Cnm0X0F${J{AN&TjJv>w?*wui2vMn5Ku$NuVL
z5XyuY0P`7+B`h2t$#20MkF<%D_#0D&HD#E7|Is}<`XD9VM54?T-GVsLz(M?bB8)s}
z<CSz5HVt0DJc>bEt5}Lw9bP;hd^g7JehH_newMs9d(e;bw*~6&_rG6Jmm_8>03j&7
zA&y=9j>ARB(otW|k5oNee$0x6{7~4D$X%QW;x|)8^ue!QmNx6?r(2lv1UUcPrRO$`
zpMWhL3h>(oTPX-E8$?9x+kn<~3haW1;0<MN+vATPz4)2N#505=InTL<+=PY)iY+XI
z3SP^ilf}-YzGpt=;CFFwQXV<=y)B361V9)JQy|8YD`l?B{0}N4Kc|YwaH<=8)=Z;Y
z_iPnwvGmmY8{A-<Ln5h_-^il3apN7r9r5FLvUpF_ppMESLF=j4H>`i0`PHXrCpj<v
z1^XT+)s;X)-+VKdBX{fh<?{*9&sR+5e`Hd&4n)86O1nOtX>-#PxEZ{woN|N}JqLm*
zo_@a_XCq<D6s*mC*nKBqUm3P0Kd4am*^85DVsJ0p`tt{%KPjYne4Ci*g)A}X?L=>W
zUD@|g)bn$co>0n9zZ?_U^WllNgv2%hm<RVt_*>sbkHvFCh({c^SVWTc?rHgHw~eE3
z=Bs+7`Bg$$(ugOB_+$tsTvjIrw}s#0Zg;@;onbi#idMXhN_}BlKKWF4$P%FXiIy^+
z+v!}gQ34kdy*AXFgtd<kO@<4C7r~({KVa_WKnx%nVO?x>4o%DIx-9xc`<B$FX1}oL
zo?b7==4;x-5wyiFgs_5h(B~g0A#$FT>-TSEn9*+L>HiiusRQq~6HH^q8s?e+_Ho5)
z@#`Q1{WveCfe!HE2T!QkhUMQW%;SX7v)D#8lEgTg3#XF*6HZR&8~+a>P|6Q4`J>yj
zqb%kxH$JhItjs8RioY04I}@FBvz<#4$4z3LJW6FmGl7?`oEK(cK>2oa9cAgi`R9gT
z@7zT{c*fKj+!6l5zpqUZKonaU3}|P2@$bulk<3X0DwDl9HscXA3yfU;DDmtXup~{-
zW*_1HC@S?EHELA(@;{rJvNh}oQXJ=i_He)U$?XxTbk>URDfTI<8&pFH;uN&Se12$B
zZM8Ty>O-D96Ermof*ca+R9aIPwkQGnpkV{J<X-lNj^`cupGD}MHpYM;sueg2l%*N{
z?6P43>qA+DXMPbRZ`u(4u=dl<IjJQ2`@QS#RLA7u$g_dq@om^A5T}U-1P#wteMO30
zASPpZWJ>80n%IFvX@#@i58m!hZ)wMiYN(+5MBk7B&(+2>`J{NvKYfAWwraxjhwgEA
zvvqKP^@pXGcn~~E+yU(8l<|NwuI;tredOVQM<MSbCl2C_%sb>)O65#4^kF(ThuGg<
z<qX_xK1CbIf_iyWHA!Nr*<YdC%Q?1X*R@@)2oFiZ4&l`@k3(+zXOIZiGNe-p`tHV%
z{>lEL<yz9<@EoM<wrrQmB=n!~9pg~Y|H>tM##@HHk1Sy0Q1J*tZ#xA#oKHkG)Rt9p
zhCFAre^?f+lRbSC1-}EM*ZsQCi^0Z6zX*bZ(BI08Fkg%~nlSgdSsT$PM$p4c2ICI#
z(6SksAs`1m0P1p4GmF;)@RGgw;$zUhhYDsl9UQ$$JK&FMhC%HOaeP3SaZT~N@aMK~
z8igpcg{~xRXF)LuaO(6?p_j)=^nVM&goZ4fJ&uV<XEb&g01@81OV^Qemq+7>8Zut^
zHA%AjdA~=7A-n4r7qZI+l4}Ws$18(r9jW)+{-tb)LR4`b8mU?5Nr5A122N<ba?k!K
z2@zSF-1!t+q<2);?E%~6D*Q%&`ry}ocRi0wR3@b@yf%ve`@clP`kaRxFvoLUCe&Rj
zT%%kJ_ow5*^7*f>v8=xiQ7&yH20(57Zqe@4)#gZ-^8ZzNn43LIyvG{O|3iFfT47}G
zXj5+Rfo$utz*4K-Y<l)fix`5t!OOhCl!I#gB$kM<Fk8O)N2d(XMX{q(V_&)xtjqjO
zF$StlHO$s(mbSYoqUo%v#=rXve&+Pc8rTa{DQUfEJgc_jRm<HH3Lp(~Y8URg@@3}S
zm^zL7=nwd7#rLrr9vAJstOIG9lQ=WTsg_&N!LGJ$Dtsgk0SMvN_QO(<z{qSPT^IQJ
z&`Si8D>*>+57WMsBv%Qha6ZTH)a5RLiWi6>w=T-TqlW%OG-3JXqHf?9R7KnXr6TCA
zYd({uHW>Nz_PZJK3<E<r3xU-Ee^3aiLNKP3WEeSr`KI@dyMe!3y4NVzIy`5gye9_^
zTj0=_EGG3+Vk_Isfo!kK7`+s^_C$x{iS@7bDJrr}q2gf(q?_tSJHr&6#*+kOn->sg
zV0yt%a4Ue4=7bgh`tFSa<>b-bjJsJ*r#;vCw7&B`Gt69}%X7K3Wzw#BHSP&|SI6Jo
zsRJwQb5twr6t7umkm?Y(_Lh++=jT7VWfR{y+aA}BBcw;mA7>w?Vr`lF^wf@gjo)<r
zP(B>siLQRKn8&3?9s*-D2W<ZW50T^^XqNDqVU`6^lk?a0x1!hnrdN`q;N7ig^Wv07
zki|Ox)8fIw*5Tj5t@5Lv&DNG1<#XKK)Jun&dVV|~f0vGe%Lf-1<#?FZ&nl4~u10gX
z;oR$A)XF*|ULp0qr@gj65t!P)ZRRKq5N(;1=H*6*`OtR|w?}t_j@B9Fc|S4Fb(dgP
z39Z9c{Q9vs46tAH3r;1m<|JlwsTTo(O&^5{A9<Tm9A_^{Hfx#kZ4rP5P65ZGmMkh5
zzV7>mBxN)WiNi&G^T<cl@6}0SmMt_A9+!wPV})D<Rx}Ud=rA49j(T~S?ZUTLn8NJ=
zRU|DK)-)u({QDu?+(_q>83YpqtKsCjw?%3)65>Q%UcsxsNXpf6sKlOWNWeeObQ=tG
z=_b6%IH&`O?^ILlPxe%Q8<bm?8?NY07c2L<o{T^46veeLpI^q_YQ?hidB#41!MsU&
zaX;fD=QSdpJ#04i${!yGv3?q!s4w3IzC2^a7CbGYtkyWgeBVE^DJL?5E*c#~V4VXU
zgt<X4s$;1lDO~1Qi}wF%_%hum0e|Hc$#V|zoiC)QTEsL6!RVw*7bNu}bKSK|t_I#&
zfjuk}GfJI1>W@x@*nMFj^l_@`8rw62=GAvkD)Z5&y!LKv=(K0+pmfXhIL4RfgaVt{
zN+q|*|55vmayIfm{ipUT|4nY3?T3FAp;_$UQH#)D8sN`(se6Nv4tz@reCM_|v`1Q{
z{nKcN20l7Vp!ZM_kSmLYs-=o=&3fuiAqL|({p<8YiLl+OrhX#7=tJWZN5)3dslrA)
z7uxS<Z9wossD!2k#a#4N$qgyjGb7c>J;Ui*bqz3TZ$oI;+LCnihvQb6+CI!PY55_g
z0<8O*4TrE6cO&Zw+1uJ`F_ZZhw%uxh<`vdfOS_ZSwEI5!vOdrL6pFG|B4_EK6B`$U
z6&bdwm__`VU&yC*(PyMnZnS;D4C)sg`)~FrF#VtIU=^wgIf$|{x|b2pZsyRTU+35H
zam}BmpY4lP^`f^s3_Z!PUiu4`F{4oK*Z#IHKMv*{=N1%K!)|gm$$5PSc13~~&lfDc
zDJUB->tBlBc!D7%7|?IV11w+t(CxOUvZk^lvOS?CERA|5|63ebh0V5AA8pvQva|@;
zSKqAYk34;W-`M`!lOeei;KB)GOfVsiEqT8`WAqFHL5$^GY}PS0@NFHYcn{?175xO9
zasMtfgM4(JSCEI)YuQW&gDqs$9^3Y>adY*{u==AN+*!t=?8owqcU~YR0Xmyk$md}m
zg$WlmAxODt!>+3I3z=fPF}u<^*svB3$1S+P*3*mubnA5e!R14L0G?i-M=66?(av$%
z$2`q=3p~QBneEkqdB;bo1176MM9Ey%fy2^&a^mRGmv3cPgeo=xNVgX}_q42i?R1RC
z#SOblz2yM9aJzOv=8b4XPT+UDD9<`6&l4%Yo5GkQ#awbM6oi+HK>>E5u+((7I7I4M
zEWvS<`F#o@H6Hi~T3i)WrmMLUiGW-zZw;Xe|L+D<=hh(8)sP?wqoia{Xq!gk`YgQR
z%h4Qh#|ZJ~TUz6B&wD+4Irh>?y$Ra*_$3A-hzA*#(D^p5*Wyz(yciZaH~Py&a6#wZ
z?$*}68=-l9?B|J)8^XH-msGg|B`ni>gf@AT`O@*b=*T-Z0dll*7g`yjNX5rOu)i4>
zrJs+khpH)d9PrCwO!*H`320MA3#)?(wa0-G3ms%|e-r1RwWO%#UH)1lhUuzrJT&HN
z7s?&GbE%D*T;nBUT6eWh7PVmOX7M}9rxw%dDv!mzA$A+~e|Y(%Q1TxYE}f4Q64M*@
zd+RMF-6jUG_RA}cRxG(-gbX&`?JyPp6Y(6jn7f`IUC6KMHQL%FU7Vd*l{{o1N6d7W
zU0@1Kv~CZ2CiWgYUIqJGfm@eOKku~fqOY0m_C4-)I3H*qT&49-w?7@MYuU)H!|M9c
zr+09KQK@y#p+kwWrXXn=SN_}(NeOL$NN{9*cm$&cj|}hg1j7qeJ$0Kzg2I$WkTiRm
zybc;Uk(ua5<Y|m5w|J2Eco0XWg9{e$;qjE%YeFYDM+JTXO`<f7!!A^<SK6vRV~KJI
zj6i9EA=9=@35r-3$5+JK{6zpMOmdpix~EzTrt~vUp)jlHDh%EInOLFY>8t(4Y-Qeo
zMH!P4=L~>}s~;$p;gQ1qA;>q4XvR~tmEPF~OS3)c1Lqc0u(vD;g3xmc#pz74{M(5w
z(wL%vmn^Ei-}T}SIzB5vf-Wx6Th7RlUc02!GH<BOzF|#3f(jhK0btHPTEP^`kE$Lv
z4G8-;0px&*A_Slz8zV4Og0Z>_<FRX#45$pj7u2Z%9c#@po!wRHmbkE+5*|RAg7Mb!
zb932XYqjPX!_PQ3?(Pm6CzcRO<6!n9X)LH+_B7LXkCBgq^N#C+lO9xuwqSb^+Iln5
z!yod|V>xvHAtg~&*<Gd&%+eYCQ-tSh!=00M<4t)UR86|+YWpxEA@_5%alm>-`9Q!7
zGDpjv_M0uNdC`{$!XX9z_)t^}8>4%;-u1pb^t-sZuHKSu6@12k&gfWlIUN#$#qPO&
zGI@XN&`KK?d2%(#W_%#!HA465@w18@e$Bk7WrfvOn3NTH+Ul%KApL!R9OK(5K|7ha
zCPf`T_X*?DSq5~db{+}Eg^F_3EtNtj+1<zNy5H#Ha7_$62e_4p`z-!a0AqZy$qIf&
zS^f@%yE~aw=mtW>1}Yd%VZ(1e4uLd?2~)T3Ukzs=*8JT+QVyNck58z2v*_divSbZC
z<X}+F@ZkHL^N28Ek-cA9Dx-tw2-bbd*wAx^ik*XS4le^%Bp+&ww-&u5wC!;d#Ct8O
zIivnXq4=Hfrqrn=8}Z(XDD@(g6h1Lum*$DQFZKI#3|hpj;T(Z8S&IDeu6iB|o6}|J
zLP1Ii*3l=g-Z_|e7x?LLfR05@Kx!}rd(tshb8lI20h^D-MtN)+uWWze!S$dm*?!&S
z-cH-S9&p~MhmD+~vuIl@BO~^*@p5T}qt`n%;l~9pHW~lMc*_c`nK@sX-0^=vRl$D@
zo?@WfB{w<k`>4q(3rt)2s}%Wpc1EeyJ_Q46SB@AF<Yu1@!@IDr+;Y&M1kFMBF53GW
z&W7oc#6`bh8RK&DHMG+D6+A&_9{x&;J~3X#8~DlR(||jU{m5)=12IP+kWWUUq*Z6y
zey`VP?3qW|xWyh@;h~?+vd}mk$?qP75&<tpD1^jJE9QxGPhEp6&Jf!WCV*$#6!1fW
zDYiwTS}d4F-RA+#4^agFn~HO7^&J{hQfiPWF3R187@}P7w)t5V0CML*UQXp3^xowz
z8=eR`J+(8Suph2!Ll(jNUmG^an8XVXXs6nNLGKu5qE$O7z6)C~qU${r2_=9eDy1`q
zz)Lm`2>*~JR%ZkdT}cu_6xM@Cr~leX3XffH>Bj2%DXRIoarveS4}IDA<f-(_{Ncw!
zKGY`c)2Q#X#X7w36wAbEsjy#b8m)~7^EBrvG>271x6p1{w|^w+O)bEmeAkgBvP5n1
ze%b@7H-F(oa=z`uUK&@Au>a3N$To@i?jm_8U>1IUG;*+WdiMGXxU$9;aPi<S{%~^O
zT%XGG0F5hf*-?e<mn~s+qAf0iZC7JQCSa_3j6_+?y5u2f#{J#96M0MV-rT>lT?ghO
z`QzOF$RY)6m_Om$5$8Fco*CcI|LL5_#agQtP#Uh9+Ic)B9J)DnsFp0Tvnh|XN~pZ_
z`%m74IYH!o$5=<(a9u<=cP9IX2yIiUA168L+60Xjs=&OejAV9_W=D=2kZf2IA9ph%
ze$god-ZhB*jD0k*R|_$RVs=BlGFhH&QE#QtlBU83{kPf031cz-P~w~#yOg{1#px4K
zMii1X?InIKLcU$fcw_vPNj7T8bPoq~3x|2gFDscG;P=j3Gci=mIiIXMNeD$I(FlN#
zTS<xH{Xrskz8tqMNh}%<cUAg_I;5B~7BN}hI8CfpUhlv?ORw9-$8F+DqAM~|ECp|i
zTOmN-(QmyTF?x?AdS{ImOhoueU!0N;=9^Wl-Lgc9N>xyngLq!xL@?02zJW(dy|eUq
z`dxovl^A0e#JT&QRJQh9R!c*A@_QZsv<3-!AV_|S2K6Eb=)eVm?B4-KdJ{6lZX$Iz
z9v-&0Ry!#o8kGAoJfTrK-X+S@inT@-XSHj?GF2xzK++SBKg`@o+a{3da%J2vm<r$b
zw;)ice!O?*%wsAkwWxv7XrWKX*#rKF8_e-UvY`r1^G~nxEK{dvfvy*<8}WZz<i@F2
zdGb4e!rZ<dt$t-Q<?kAuo!w9B6Vl3RNB!Lv>Ff`4ylR!zdiJ>=_vf!A{SOsjuvz=Y
z6eo6jg>Bc->M>L3oGblD3uAe$L%Wjwg;K^-e{<07C@1Ia;C;LHcjXs;slC@S93Wsq
z+;a(V8g>!l76sL^Y`N6&udE+L0U(He_$V3_M1&4kN63umLvvfNKPf3j`tRGmq>heo
zlfF=be!V0j(ZBP335v0TkdO7+?Z|g<h*W_cu|t=YJ~?9$W=1hQDH-9RLCluR(p2-*
z{mnNc%@869d}uHxuJW?ysw{&lQfK`L`$6<uIjlSpqr19Zq#y8#3xbG!+nrHcE|;a4
z%tUQ}34$_ZP5F}ksZLQ6OC4g;(9<!nWL1!!hEIqAyHyQL)C2A$zAVg<@%}SBD#NzX
zA@csyo3PCIcuiQ>VSoKI-=;egwx(IVxkx&fEHa?xJ>#tbxZfX^V`G{}9bw+f@QZ4_
zrezoNNC6V;S`0ET!EwLcYEIVH7ttSIhjk6VXEYdHHSPLxbR$}95fKi_bVtj%VW0PL
z3lxsQpRxTUw1m_fuB<*iZEej~6O|a5F)vDvIy(mZAsM>AJyV9K%lf0e03I)JGscUw
zKk~uD8t&*iKq2kFA9m*!@Do8l7c=|9HH$-0Ic%zy2N{-R(qhzM>>!={3%{7ZdhAxA
z4(mo$vF9`g!MaglL>J~Ta+FT7ZGq%xmNDL8x{m>^(R+UawRf*(1&M?!)SnR!JwZls
zf71bggW1B>*SK^qvv8CP@IArgh=OAHvOiwJ6g@UGNz=Jj;q-^b$dIflA<|<S@5?oT
zs4z9GHkOe9pG$&J?CSG1Er|<yVPR}$VuTrK34@s&zA&;V0FkO}<_%@2Urb+%uMArf
zT;MPfggP3RXogJ;BMsH-gK)kmcQ!kGQZ&(ha&E-qDYc@q*?ukAEOKQq{S=C*l?GkM
z<Ap6XiOl;`VMQ;m8x!VMVvy{Vp!0}opu{(AQcf0fG&xE5ba@X89r8G|PU3>kxE77Y
z7hX#j&Q1%Wf!pAkjKEGkW4x#PZO?W01%scU)Do53zXi20pV;1nqt8OOWMU3l2_(}-
zM0b^keXo{D${Ia>KoF7#>fOENxUjLe)0UOCFu)tXjI8PJ)f#`KvaTZNvN&PbZi-R(
z?gPci)j@7rkp)Jh+z_+ixw1yZba5)j4ij=fZI7Q$SLfGxC5BAObMbiJ$0?^365h-I
zHanjo3uNLS%~0H3aC!n;xRN86&7NQ-+?w_4l&)^~K22!Y^6s__?&6|X;<t<DzbZWZ
zrjI*UX34F@YUs=`T+V-{SM^Em{~fg3RU8`DZEeb3>%pQ2n>Q8$()01Ot_n5Kzf1lj
zwz(Pq1GxBCMdy#t10EmrJ?`|cOl(W?gwBgi5i0}L0)VBHVl+~`v#bzs2q$_xtXphG
zCn)1;Kh$6UbN9t_d%PvyrrOcujE9%zMI6P2n?F2-<)uqb;9)bA-^=8J<!1TadV%Li
z1<<?=mPm0*-XuopbDhjxNqjE_#A%7a;U=)LfBW*_GOuWP{$Q$_m|PZ$8BQ+cHN~cN
zcB1pQjU_SFjqb|=2JNKRrYc4(cOIA-50PJE>}q2!;q&(U)d@MzkzRUXpFeqZUu4|L
zq?|t39r=}v^TQKOCt@@lBgx}$DSLfnd)Z9#+MATU|5%kwc;B|wDd~f<$`m(}#y>!<
z_KS+i*@+%rF!-`4V{d$#SFVhCC!6o18EhWHXZeCr^e1R}vM{h}?aXWWl)1<rlXn-#
zQN!RU-tP!zze059>~FOU5Hs2xKRI$Xy(7LO?vE+OxDSDD)tE$1_FqnKK^eMlbU!cV
zRgc>ZwcWRntl!=5z1$m#7993@dIgQ(xiZFX-eTRxKKc%$DqFhCbRg`e9ZPK5qX%+N
zzsne<e=_{j-(MUiFZ&MQRSxdIhd_Qr%Rwbxu9x}Z1kzR+9nH3UeO#&#km2P*;iwD2
zKXM=MNV~IKJv=(R{Odo%h$tc`+O>LlOBgG9pt#T!PHt>`oN(_uLhQ``83OvO<dswc
zc(-`xR($H6Cy5aT)Gw80wHER27iN9GW)QE#Hh=ulU1sYzJb5?HG6qUOJmFpg!kd8j
zLQ}{{(C{d^5SQ0?kP7cYPF<=HFtFL{L(9owO>RLm8~25Z%(92r>_xFPj30~%cH-q{
z3;R5ti!zM8b*^->`qJ=q&2M!a7Neey-K}O6PuvXj?ox(6@g99%#(HLi?Z!ig)fr%*
z@}?PKyGnj>wV$n^eiha)UVC54$~DfP;(WiNueWLwTc8)vAez(qUvgS)mmYJknT0XH
zH@(W4ZnZ%|$L~!nBmT<Ra-$|8G0;LV=o4{*4I`7_j^0ch22qvBbm7`=#3I`5q5i@>
zYe)slo^*;zz3J`8BpP_{4)?%tUdZJ24NeoYEVBfX(TqAL;lzoj6$~I5Gx#!+Rwezt
zkA@e~j9Mg@9HsmKEi0@jCq#t_S8$n8ZNfB1%xIrdc%Zl%UUVO4zfe44J%!G?;<9N6
zSro-UqGpkCU;N$uHMc)~JRyP0Ky9{ik)%k$^q??#g0UOmPSwCT&G&9V%}BVL!&eSO
z8|b$oiU739f=>6tuNC;_9sgPizOuSF&&%#42HTujZw<A=;#Yl80e>s(Jvy9~yq)#|
zJHs9%I6jia+|mOIwG6yBnrk5?oayo75&IpFWoy{hlSW3z$JffKox<g`dvKh$melnG
zX~d#<e~7SNdtXD@zq`yQp|xH~_4qKV<;?(mvvDa1K)Ak2aQVHJ@zY3GFsgX4T5FH<
zA^^ti3Tr-nY-?-pHdr}88M0AaptJY@LWAa_<Y`9jASYrNv+c6&E{p!$A(cfoqyPpQ
zc6D}<B0@RpO8->%sxc{I>7ilYgf}n-2_)#q3UsRm)nZtK9;m&|(4RZ2t3OGMY}}l-
zB$crxf0@o>xUXmx#5<AlvVmBIdK}1QevpbkGve@1j@p@YWz@lWWXSj)BKVM48Q()9
zJ&6<c&3e34;YvB{*u0g9P<v#Qn1x^E`>cl`ZD>>V-CwaobUgo^)^72P&p(;ir5kNc
zL{$%M9b4CHm(0EeR2)gQ+UBYB%Jk;9!apeNb^w52hzrh%MO56?+vAD-jyXp1K*esa
zs`z@qA9=y9eICAfGrpy`KuSLcA4~s#rm)%zl7Q3)-+CAFj?$&A%xs=!vft-uOhei*
z*)`&_=6ccDgg5b~x#w%ePpJu=u2apV_L6HsmVc3hFNruAl_5pR?PrZO?A!vsSp`#6
z+LahbE{yJ3$&ABYS}}w)F;KsveeHcCLKWzN)xD>TP)va$#w~ltLs5mKPe2bxPKW^9
z4{_Dwfdkrz4&j*D)m2Fnp*Vbnc}`&f_xA*C@^?hp9Cz`oYvr-YlKVjE2;;2?z#v#U
z17E9L?|`j&^hbVA5e<I_2;%o*iI=bjriQ{Bxsc&%3Gha8Z$w`0Sdqf5bm<eT#R%71
z4sy#5*Ont?qPrd3mYpW}R2j%SxJPz~1e`4FI8clm+srG^U!3E^vh20uoKer(%A#u@
z0?GsInOe2|DI;uq1nI8?fDCg%ox1vY+@>FLxEI%<*GASZJx@<(F=h$R(>+l3{2X!v
zmYQi#xP}d{cYKqs7{r}+UpFGl8HGMANMC;1DG}^<<l}?K)YjK~kCkxT)~=?4u=pv#
z<3K6>Y}YS^RonekiVk#ET32<y^^OBF4%n3t%hYPHHJ=q>=HPV$4_DQM@03G-!<NBu
zK(Wr&-z?gIK8uYZe_}&<3d9b2`{<Kznxz!E?lH5Rw!a^a?e>3O0DTSx8Ym0y<wE)V
z5~n9Q8DhkpMSy>H+8Qpdx$oEC!>0<Sb*Su-m)e~nISp7AB8zlpRsic=J9htnRDESw
zlu_I5&>$@W3Im8B9fC;15Go)o9YZ51-92<mOQ%xOokQ2q-QC?ezyN1_zZ37dp8wB}
zz4sk^?RBrU@zuA7zx3e4^gx@UhQ&}KM}}N|LU&FX_vxR(<ilMZ8wKDkvzD@B%i(ul
z%aIIBq1W~dcEZhe1zhcod193y0U8=Do{SZ1Bsp#u0cPx4=Hw2;wGiVlDtq{!8qX_v
zj(N6~8s1I$N=4xad4xr&DUr0(rlWO`@D|-r9>)Y5Naywc{Z;|&t8pp#-$brT{&hvn
zzBGL8L+O@ZqibY(vL(TokRWZ23qxNd+-P)rb+CS3_&si5+bSM8RGr-lSG@dN$7iEA
zuz*_-NUS4jNDiyB;d3uWdM5Ez?%GIom3#{rUY)oy!@tBalrw}n;vq?nR7`LNp&)>>
zD`hO*qMjdN&z8XmAIdj=BZ90wiCvUug#`0amDzuV-&n^Zo8&;hmv&6TnPsgXVGUI5
zqS)9fHN;R|QGI<_v45FyH0CN+YHO?e4JuQ?z{u@f!|}K{N((VMHp#$GQVUE-q+_sp
z4FWlvC%_CPh7mi+$_O1rzF5t+R5^}2{A#43amfE8(z9O2EHP@JE&m&h;OT~qJ<NHo
z=3vq<p3_$S;GG*C2dRSY7qYcE{3@~9op(^GXBsL(M;88Gy914z0$0u^Gfh$7PdFNT
zc)(6S-Fh04d?4h7xnv^Q?jjx_TE+)}*ApNC7Fm=3Ui8;G2Epv*7^_$o=^Bw9$o66x
z{d6POjWrWe{|##t0ANA`S$BVbXuZDyu`8@JY3sI#=r;+JT5}}v9u?$z-(60ucLrz5
zr^|8P%B=3tJi(q+^&7UN%Lr+f2RyP0^tQCx(kzw&*O*a#;K$Di-#@vX1%n>x!}nR7
zKYTeJMo)vEMuOQmsh|*RYnw1^GBZ$<Hz#0X-<puoRh+Yt)yvcc`*Oy9aiRGTAtT^}
zF;OrmluofKeaL@@S5D{5mID}9ZGUEt#~ogfT882_yX1Gjx1spv0(xBC=y`vBFMU|t
zIB`NA&pnsxK)Yh_XX1WuLJGVZX`1jW7s1cg(i3?&8oFeLatYX4ROSFlr!%h<)xPYH
zt_%fFk(FA=bvo6^;TW3{OLeaSyi%NPn3bb4UtlXu)dbhe3C3Wg^wVnvbtugpW~bbC
zFghi`3w0CK@H3foJUFM1WIm@_&cB#3$q^LourrNXPHR!N*{S0o^s0F!tH84KyV6*+
z)b3ZPUR7>Yebu$J=V96#f+Fa#qz)q;0ga9xn}^QKrcB-VmmTuD2&=uS0!Qezr2Jhu
znl1sOF7RrxX9V|aBKMeEx+`-5b2HcaybsN7tS`{pn9~y}Xq3V$EG*qD4-(0T?Sr>W
z8!dhR|1SWemd6&PCEUlaXhu}1icF9UzBF~yC7w2jfjS|u<W)H@%_<``VcnR<M$yHm
zo}O8AIoL5m>LKCI=qx;Q#KNpE={D`l^w<7P9yYu?<y<@MDelYqN!jp#hK;XorgYd2
zUB&{Dm(}(;pVNOfj)+X|t$%ku>pNL_I$B}b^S;o3(h$_x*xuD>eU$(0^}_|{qbFl9
z6}@3Y&}%1os0{U%WHEp*zUknYbi6>&=TD|6YW0hKroKQR{*8L1Pm$ol)wxW21b-wL
z<XM!zZ@nq>t`3zVC*5mfr^ZL{#mfg40E~RDmQ*P6&jsB{zpYq>tC`ZP4GnOVVd&A`
zKgUgUY!z^xn_G@{gDm56GRc6W*C|R3b*@UGx?r)QKR=1mPRSvv%Bd8dpS)=QZJ?)4
zU%D4`@9n7rC)I=oN}f?Z>@FnQ)efb+<k`M*sV8FyVomdxE2raT>v0H_KfPtf$D%2e
zQmD*#)9tS$W;%-3prHM<j5p;T!BkyD)Wyhy$7`;Vd9D6vhnl{!#HF}PP#Ok!X|8kO
zeDr%hO}tR$m9Ab)o7OPDGcm6~R83_qVZ6jhGfbfMK#ly^@mOO|Sa=IoxpDBKBzS6!
z`1rd}#D#19ipv?w?`F#^m*iQ{u?Kacea}TfjpX>=U3Bqy1A=#{&@J?0sYgi76Os0J
z+oF)ixZ#J;t&EySf4FPE+}<Vy<z3AHE!!~GjIH|jJ<80jTgm%shn?<33mVS0#PMux
z>zsq?`MenNUv^9N`ZF07#jhbegiE&W!qyR!s&~}toqwD|22cepyZl3+uX)~;mZqq?
z43{#?Nd+HYN>1%~>V!=RPF&$jC3@ufjC8NBdtPBsM6Cp@U?S2^&-PFMyjW4qDeG~&
z&FgPAx6$n_y)?TGV$9Ube%Q`l=2b2fi1_bQ>|f<WILw=Nwr^(>b)Jl&t}?F|Fpqh<
zk=GD6kyYAywsd<<x7v6NFBmQmE}1n;=?;}wtaKL5Gd6DXa$S2$1U>$}II%(11b^nt
z_QJmjXpOhMvDu>JDT{)6a2;&l8y`fbTija0gys$_O7AN@HFeWV`(=jY=f`i^?waCH
z0y3nVAD&tk*&a9dxAMX?gZOHGXph!|Tf?lQ&fDu2GF#a6GDSG46*f03_;wl}J?uCb
z6ERY~Nt8maU^icbGEL0pgZujkXZ!@3bZ(o%N=&ofuWj`QeNP$>eK|Atmkbv<X5ISq
z_%wO<hllI5b2w~2k}PCDa{3Dzt@Gf^@XM)B8YaG)c|J5~+<-)Nu&1XMP(_MR>Fl^r
zm~7C^=v`N*-d*J>eK;+E0`P*X7sYCc%V|g|wOKBggC6rc5bm3RP<2`9{izdGmhI5q
z?QG;iveMSx&Hdpuf`{fx{p;5`K&EEk`^&~+>t81$o2e30h2OH~1}fQd=SiVGFD8iv
zD(Dn}RZ_1`kKXPkR7$%(Y-x8I%{!H;AfIr#-#@vM0ZzZ`WSJ-{9+>}Pa|%vLLA3$6
z7Q0jjZo|bhVo5`X5=D@cA0it4vSG1T2mEZgMmnBhjHKJE5-c3SS5*Mq1A>P9o@Wm^
zru!U7DZ?(RlxEm&a3d<nGb)-&l$_Nc&XN0GRGxg*Y?lwRt0Evemsv4cfP5;}57IKK
z75lHREj)e+KRH0Hv1+&e-=)P06=a<vq(#{L4NIw$Ag#15YQXX$*})*23rq@srE>kd
zuaT^68;l{R4a0OnEne!+{{n%p?Hq)>N}ut~TzPzS2|SFa0y#k~S|Ine_1;WTG8dn!
z8<#!+2nj0>=xXYfun;*uugHS*`ZfhHUgSE*M;DPq7Rv(P20v~ZQ`NDr?!*IEoNu?^
za`#o3yWF^><POv&-g(i9k+hhhzYEYIhbz;bZ<rq6z<-?}#iqY1jfaO{bbAO$ls%(9
zJEA8m05k!y3sLPVI3&wZAb>kf;nJ4118x<!oBEc=q{_y;^gD)+8p{o%%Dm4cN{s>v
z^?oRqefOOh0GfwT;ps7?1m&n;#4KVe?A#o{dpmOV8VQvJTS%O8AlmouBY6nzz+&8q
zYY=ey4czHLb?!1=Yl@2lt2Hn3fS`<GZI-80aCm3a+N`zgVlLT{K`5>e;dE)|`hms{
zI3sDIZkltvFfM;~iBq&cBr?2z^aJ{de7>T+@@ry*>ap8eNB>_I&gf!Su(?stZ{F9e
zK3-?X)taUHP(+KnfOr4NSH}B14Xls@g{)DDCdD7IN4N5O@|r5b_Iu>!W%c#oA8qKM
z|3MW6zcp{BChI{NdzGYOtdJ5Ac9k1aFwid~Hd|wbjZ~NrB>T)TK-T+T-^(Aq@_0!W
zj^V8r6HdaySgXN@h_&~OI8$g`R&9zCX$+<u+rU?F8Le*)hVrWl3j|CnV<Z+9!Cnh{
z5*(CIa@+ERm6%ubnOy>2PsbQpj&xb0sl+Zi-GVh>7XV}&mIEG$F^j77&)?q%;6zHJ
zneU+?7W=Pj;P;{#>QvDVtUfULEg|kx;KdQG07c1uVOtap`(-7eDN4U{1N@jqCt5HT
z{K4!z;!h&XS+02l%fQcOuiStPdXC)^i2809Jha1)#EQ{;x(>5>iJNDy)SSRB2>um|
zn6uB3m$Fc^xjhx?hlIRW)JTp5UfmKeze;#a!KY@9wkS^9o?Gd@u#F3+W(+1u7wd(B
zqdC$^PfyR-mW3A6H9>+il`ET&v_a!lAI=D5;LUyB(8(&$FV5@v7>wVYpSCg8K8AoQ
zI}`D8_x*?UQd;!*CDPU7$r%bzqCD+Xwwa;=f$#WzBz$D_-u-)laJ@$A6a3f5P}!D%
zIY$o-7lJ(64&6D4tZw8^Vj}p#8kqu8n2<(oi<>?E*b)MG!N3xYwKaKbv|^#$-ZTKx
zpWSJM0YiuEN5&OlKz{iADQzRX4PmVJVI{Au+SX+_@YKkAR@rr1s;7vLh+dxBibqu}
z`i_LO&zGn~sXcR{-!_w)pGorp;R3ON<0=A~k5lWa2;&MxjTv&dAy*g}%&b)M%*1Kx
zH!Je4(Ww2S#Syt?c^C@o`(!C9l_~b3CW-QKO<v&#79%aUQTL}UP00`_S=J#UT-@o%
z4oWc$a%#zSb)KghwcID)$y`{YK)z32VL)_sTFx#Wk0{-kzhAt+rbS*um%3r4%g$#P
z$~NHQgzx{oI+Ymi+6EV$(rVG9Z4*?2`$tizaA8M)=OXg9itOCILy&^v8Cwh2G@=Mg
z?0Xa8SMUOXef+ky^P>gy`gRvc#??KyoGnTKe{1tTFAAdRIRKNY=y>5M0u|Jy5#{FZ
zjD6aPo>2|AR+_4?y}W#o8R_4QP;9F#q(nt1VHVIm@IT+Z8%P|BahMT6cK)Ha_iQgN
zDujJ{T@T~5CoQyB=_9jqzaH2!YdKU8e*I~e(%&}tE4?G*EWu)>k0f|#5M3*s0(UkD
z<$;|n%0k+qT+`sE;PJzs8*k&@MfJ5);oW(RaYC85BzN$>F~tic@|}OKExLWH;0!3q
z_8-{auR{f$)9dzfxH&uF?-jiQxj}${{lbNo(V98-$oyt5>+Rj%4~6S3N!B~(?Gs@F
zh6^r?_umG4a%^*9xn5gWzgM>3zqY1SORos^j$P32GTDsWSpIx=e80Bc)mQa35-dO<
z6*|7Wo1LbI(Hpgp<{TotwdVWgDKq!y+5~&quvlt59XMJs=TkN<Z63LD?t3j3s$}B6
zm>e|nTRzCfo3v^-8L=x8>N~YkyJ4xM<Hk)|rr})~$&ABI+MMo(ipwK}e+H9ctY=j=
zx&8!Gl5|M@sUL9W&4`t>dtURLk4GX}=Cah~XI37L6=}H^rT|?>{||plUYrU<^~mHb
zeANGg?f$!!dp$h-mWkcu4dbx?0bx!53yP7<b4lRyEx1aC!w)8$U;?SW8Y-0HfZLr_
zHnH@M2oVN%h^V%3{|fJ7X!Y+n;{ozLB*)6;?r&XY>DC1O8s(1H;V<7z(dqrq_a|Gf
z%^;7FeBmHynPOD<KwKiD6@DT={tmgo`H?cO5#n&CHHpF3_v`!9H~?f;`SNjp{}UE|
z2=a_YyzYrIvOYMhy^5ge*NO6oqdP1ljLr2d{^^UeR&YXn^$)7m@DbhN6sKwF>>6eh
z-L()bxrv}5HHsyp+0^YK(ejZNz=UOP9)48h`}Cd%Ih!oWUPmQL<J={giDZI*U;U(8
zL*-`uwNjYl|IHw^`^BBml~_c1S+-MP<5JE|{U;~w%s;%mquqyn*TlO7UTIw(3XVC=
zIXyC<WT~$*+D+=DFBjMz?pw_~yjT(qt`XUoUq{+)WC}J}YX=W!Qg}xilAI*&)F#P6
z@;!5Sk>ppSYhHZ3!HNS{yJhk_k<YgQV}u0;XHQV+JC2oK80hWY2#VLBE~v4U{)}~}
zZsIAO+t~!_8o3)=8wGL2xyJc|h;GaqCFbye@8zC`@>`v<Y38KsYL;1yi~GU1^CLg(
zF~2d5(p3|@M3EEp$q|bg?Qx@HPaAkE|6CnXK?r$uBA8kx{HMu`63~P4S2OZl7HF7r
zB<MlW<w4!RbnXKV*-*R`YdQbYtCH)fYqk}rc!A<wp164M;Vz7A{Nm)9yN6`$Tv5)|
z<(M*c6b820-rNPj0}$ZH0k!m<MYVef>x9)q5W|VD_)hrAl0YuV_rz3GND!)MMkZ1^
z$t&zyh=cyzt=2G`i|U~A#(tjk_+Ntzc}$MNfZT1+xTAr*pI{^5-tdc?@B%<79)jK{
z6WOwaYv4q0gj>$JyJ|j^j`8}^6!O9va-G*)YjzgC5XReZAyeR?c{3|<-)X+%K>pVH
z#T7XYlMn<`9&~>y-l6{e{s=Ch-2`p?m9A>W?Sp7p)lsy$f{GV$(<DaSm1Ocie1)zd
zwE<pPLmO=W35K9?V&2{LXud&X*21X-PL848TaQR;IU2LS##Af3p(^TpHaFWg5_akt
zS1qQ}nq)E?*Y-bnR4Idp!`Y9s<Doz~pax73d<z{l{l{0?|K50c@}HMT(JSElAdI96
zxtWfhQKfGmD*d`D+AlHAtlwl>4S}D2tDFCz{6T_YVFC+0T(x`WU#;1enCq!EnphcM
zC*|qe3V0RJ<4Y4|F)`Lo+`p<tcH~%E;8i6Z-A=gbdk?Dw6w<RwY>b~<y!d^&{=C}$
z+n5qqQzYPLb#msaY7_vk06S)F#OAFJX7lEj;z%(nm-vX^XO6HOKg)GFh;(>+x~|dt
zSb8dx>>5<MLZDq9^fPwz1aT?1fBI$KIohuJ{OlXa=}(bcpGe=1y-r2^-8@B07g7y7
z*|4o@=%ptT92@yAqHjI%mEc2v+VLYX=6T}$85)(h8%*(?=CwN8>d`M=-dY_<qO-{d
z56iKd^WIMgFB)uUWN<zvMARD!UR+$7*<b!(gnc?epK7Lq2}xnA*0<9)F&qDig!)gb
zGDld9kxCgT2p~Ho8q5Q$G-hyioCkT`(Tj`XmCw$gN;fsQ-qPmD1FwQGvzj$fRdaRI
zhrN6=!^3e=^p=~X*B6F{){mE?zeSmlkB^8pB`!<_>$t4$n4hVOS~{KWORx>aSnKCz
zd+Ih33!zo2QDu!r0Im4)!hJJy<+vLmz3y^UmBm+V$Z>h3`mJY0!|_o6ui~rfR7O7M
zHZEczA@I>t&m>&wP#|xJ;1ihzhM{&@IMKm2rCYJSiH41DBcvk37-ZCgykaOg-fJJ0
z0K_y%P{dTwKtD4PLp#i50(bNFl$TactfV6wE`A?=TbX#jm0y0$*BE&?4b#%L*-gB1
zV;bQEJF2tJ<2f>~Ft55ATS9GchYD)fdO8>;86Itas!^R*M7wVHNlS>~gkq~H$@p*p
zB5cpBY4D?IR*u?5&T)xe131vlg{H7&6n$f&BTrw7q1Oq~_|#9IVoQ@goAIT)QPUdF
zYLC`8i^JW*MZsvM->5E4^=C%<kP;kbHR{O*pztZsLV6Peqo2yrdaA?&0$ydpcQK-Z
zQuE#7WO3_+nL-h(MPSL`)#_i_*Q9Tv*xcv3vPg(C$m~5~Q3JJ;k{Oc_Cc`5sUx@~F
zXsG9mu_=tOx<jm)ixFiVp?5MWZ^HVLQDXM@))7f*Lp3aS@@(Jw+t47vgA!r2C99x|
z+wGr@%Buxt*xUmDaY^F>78<e~DPkaWasL@ux4~Ek8r!07-b@<bMbkC(wngc<JUyhI
zM1|igUSL5UgP(Su^pM)FGp%QLjem76TOaRC&yx7xpT&EwY@H7coD7#=Y(B2FxSGR0
zL0DFLWxNIg>AI{>tc|}m*HST^=mpI90$}#KFAWr?eviyNJMGL`M^cdVcKi4&MZF$%
zNq?_2ujvS?)iJLIiysnJcJ^RNR!|ohFNfHmLZ}58O3!n=?zbaQxhgD6RiL5!{dugX
zXseR$^=bw(k*QEQZ)2t!5?Ks9gj)xu<=Tj~**|@T{~S2Cg*L+UV<5x%0bU_gj`yo+
zRTU{at7suc6Z;#UD>=#!6(Z!*x$H)MjS#z$Av#>XN&^+Q_I#TRq#56RztdgaKM(Bl
zK=LS>b)=<E$kg3g^;AhM98zGq?^TU-`hcTPRJg)MM7KW-tNfb`tmQYvJ^z_lK`7>j
zhQTqw#Iou$n7DV2g14&D_VHyF7=_r6h`4Mc+AR5rHU3?@6n~_JOB=>deLp{HAfm1O
z`Hw-VLsdWG06MZ#?N|ONqJv8P=(1m3-!k-$*~aV$3vb{JKF3!Rk1G%|Q)P#IEJZ|Z
zM11i>19Ugfq1jK+<!Ok0n)pO?&5lC1MItbI&p|`A638PDd>wXXrzm{0dn|_gohVp%
zGORM&HzhJ)qrQjK9m6drikkYwC$68A#6uh2Pct5FK`j=wpnAYs86k+a_#@cu$&rdO
zDjwIn;f0&ZkOeQ&og0z#Xg-A`nMUCwbXTxf9iWc{nNy3>G;f4>0ri^M)SY7*NWM0n
z>~Rp@L(fS52$K50uJ|q2cKM711=P69;-7QeHZr#U0A+P6S~txvM{fN5<c_)EZmJKB
zU|yjI9n?#2pfs}(IVa^~ka#YVRd|4ku&rK*2cBUjVVDPcVj3U28nk%doI^o)1Qj=N
zKoP#6fwqyC>Wig6CpTl<{bcc*LgJXE>_;<Vr{21Fp}eTp;m9<p8<HK3!BI#Gc9e0S
z=8-&VP~=T=La|bAke7?DO==#Nxj7ON#8n{DUXglwE@DaIcu&StO%rxB>hy>5;<8m#
zTlX2Diy(_@aT4jqbY)h2Z@eo~gO|~2T;ua^@D1doSB1FM@quXZ<w@rX$G6jL(0neK
zBEQ(qjDt}8Yihox|7Ov9V?c26*(-r7fm|m&z~*Lm7*ARJ{?BOJ?++FfG!#ZUD`2Cn
zkUp*eG4wkDJBfo;K$M&h>BDri?tLU_di|dOQI)r-?P`p6d2zc#W0nQWSRa-On(P@x
z|KNNDhh;C(0|dUXCKlELWbIEbbM|KJ`1@S>MHRL*e4ck03z{35i4lLi1LeriiJ9_#
z!jer5@JkOR7~X<8PcwzSBUp>YRK{>h16`*Fy_<@_Mf#IXK9<H+bD5x`bItr|F0gi8
zM?jT4$Gfi9;}n{@ZR`W2cCVYg3c0>LR`s1ZkgHf4=~eZCLBK*c_hNjLxA(}t#ux>p
zZ+%RrPkQQBFc4;hX7ja;5LQOuMeP5tYV~&{?NjoaK3ix;h2L_#?5rpb)lbi`(MV>x
zYHu4M?;G?#Oe$z>qyp*Nw6Gs<YI1ms^>}(_2iZ)iE|mc3oatbp4cXr>M#~n|>>2Mo
z!+n4|MGI%0bv6KU^`)`(&U>EvQ)OqwY%qDm#^Iry_%AgG6iU_*6>oEpvpe6h3<mGj
zS!Wr;h=rmGIZQ#AYRHG_@phzLSHdgUR9}?KTG!>uX2^&BYDdh$GzRB^YOknZD&Acu
zAm$9Rmj1QbR6meSAF9Azp$r?ypDbEY^PJQH4a+sl;JSK8e|i7GEk}N}&?!tQGhno}
zOJQTvu=zN3tzxi2k8jRFE?vcR;?v{Hy9nCs@Y3YGOpGSo1iZp#;vg9aNG7?l3}gSU
zyT#r)0%8t&u~T*Mk#GBz`;Y@JVE6pW9iW5>8GqX*L4zB){E$=77&tw}ezGIHd;*m@
zQk>F&=-)k395KRPIA`nUDZHUv&1c6K3wCPr2!c&MSqaT!Le-Q4g*SBBEsYFK*^tN3
z4+_4QXi}zd0gwma?~3ITmE81{P153tI|rn8^(~G&k4PuwDAy1h7QT1y&3vXt)9`ln
zxgT4FSm-<W)vMy~<N|5}DP(8Q^3d)ypK%WKA1dSD+>`3>$M%ydGYs+Yf`@{33scaY
zet~k1DOh~5l^#AlcdijM*Sh41A<ah<A<*Vndr#gUl8Q9A1?z8o_<l!;9;e=pW`pgr
zM5N_(3OH?(_OfM{Q({&+_|gSL0|7*kG9BW!r%1#=Jjx)JDgL4H^8D-%vsJUlEt^G;
z?9##dp^lY$RnueZrXFZhCR+M2&yjf)=39$I9G!eK%jz)U=B)yT-;nsIVBc99Qj$=R
zOzuLIV>58Cdw@J|VlovM1Mc>mZMDSn(M08L3hCT)``r#=!b^dNIJF{=T<>tp-f1L&
zuW8nB>bJh7w&hOty8jj<$YI4S#IHHXnXyj)x|9HJnBS9sVDUTazJ`z2wR{Ez#CE!H
zX7#yD?p(#CI+;quObnK6`2hWu4cqM+@C}#M+$DAzE!$CV<*RnT)M88CSl)zw9HW%^
zu7sPRmR#{m@ZHEEY|Oog@I_mK#oYI7os4dR*`s(*hu*{1FZbYO*N$mBsiB;p^WVZK
zk8p!Gk-pP3-gkdBrh-ktS{*m_x-8Lk9PrKs9mp}K{XaFkCTpHt;Ej-)Aa%LtVKh#a
zP?^V+;HRb@8LI8W%l868BF?>n=&SjC8UMzwsDF&Wwr=Ka2%QQz5B5b!K?YK~iuqs6
zChUXd%&%mMt3EPxnl3C=AK$0x)s*mDg?m^96W?7SY5iPB(og^b1)X9YIos!aMG3kT
zF}~ur8mX^^({5?QCUzQ-?u35?7X}5=AaiDU2LM5bVdF1#|0t%Lzj&?kMMcX$c~~^?
zx@=I2I_J!WA^!RX(|*=uA;NiX2IXfFatMibtDz?BF<lf7@DSpSR0G7I`34mEb%aN@
zY?yx{w^j}Hn@#>wX)a2WZ^0t-L%4S;883J2QWDakkWQK#ICpC}1%AKDOrMnzB5AhW
zCSmmh^Z9N%4X?ZwhbMm500AiP;sff$*h`!u>XBfCYw4#uev>)XVQ~uCCCj2`N!+7Y
zl;uz?mTCt9cUI6NK7GG%snlmHUF_;fG;+QaHXDS4XT~I{qU?K2sk_OL*R&o+zKtgg
zP`4dRddJI?A~*KE!m=8a!k>0u)Cip#ivp}2Rj6-+w#`E6mW4TyiB!_}KOW?i2cfC9
zui$e5@i%-CBbB!s(BhaUWWR627N=Nw?)K(hB~=&y^kO7O_D{IZBxEDlHF9I!{14Q_
z`SGO})v57iRG72G2pV~La23gj>5g2waei+$TTkrtP5DAD0bD$2A}$CGU}b=^!h)JK
zj@8iX^XLy`-llRarZm9n#KB}7MN4QR8U6ERAbRcb(XgG^%**)^u$+dLAx-GPA(dIt
z*-hoIcj!ZKbB<Ews(G$n8w>upLQx30!AvzYBdZbk%D%NLcM{FI2l!(s>jGR!K;6me
zCV+8EcX9BY%QIM>X}3(NI04)U38PTM=7!5O%7(X+nn=bD5P=T>JEp|oX#IVhTGzML
z@+&Li<YJtbitduS%V+PAIt}Y%g0Fl&&irf)E2>rN2$2mWNsRsE>fFipZZhMS-b(xR
zJN)-EyzKu*2lq|cMP?E?t|>Z~!Q;u4L!%I;VX2u=>%sTaA4XUsy*)g}2mvV+@b=Pw
zvNpxSip*w%JF@?wn{j8#SHh~rRRje}mr$2ERT!YC2so;N$MHV%KNYohYD9Nr1LHx#
zxA&hzKO@^bA!yhM1`)Oc>KiRt$5#{zX@f8nqQyrBm=F3NQcvO@I<5Cel=Zcg#0V;<
z1!3)HgOn8AF0DJ0ENOpAr>?+*=+Ypjk|V$8439+_2!311Bs7h|?(`lMg02bjY^iqv
zRNx?CoHyRirdilNcu$Ys*ULFFv0&tgf8k)UTfI~8%n)zNnHQ$jNBn8+kk8KH)88gz
zzm1bmX$$q4=cOlTM;{*FFGHOysJW}Y?6?p-9%C!gbchJQ5q{&-Z{Ozo9*qFY7S(o3
zAbL%Z##||3O(KzdCiMJQ59VlF?Jrt7gu^_};>BgQtgS9X2!oz`qSE&Gt2fY~CIX^9
z34V_s4O;O&sQh}z-@$mx_!5WR6}WwNKU=#4+VCh&-v(M9*K>PaN@A@}DzvfwUP5)a
zBtH7(k~uI;WkWc9%yA^Jz?0=r-PJgBG)_>rQVh<!h8Fykw5>+4f|lhKKSqcXY^6p9
z!vrCsvyJo$G|Tw+c-?1JiWZp+Wk?Q6nO=)$4bzr373gVV<;lM}djrM8WX0L-$BOIT
zGk#pX1z7l{`=K)e9iwq*Mepet_tCS=bRt@a`wj^DbJOGeXuj4uayEgfNUacLrkHWJ
zD*w*y=S3iuoTo<65(_`)JZdDa$qJ9<AIfTeXNAaj#}{;?FAY9KxltBMnQx*<&E6l*
z3S$Z%+CF0h<AvC}?$ZPhOD*_?#%<?PEliU}N1)6Qa_Twzs^0ecuz^=z6#^(jXVX7e
z!PM*!JU`t6kRvuynZJ@aC=PdiC{HI4%2Ad8>vCXPS+fpl&BHh;`uP@uQ&+;*bH<5r
z1)?o5-RT)l>MA|S%8@ZbE5R|USkqK{veb+STL39+gm-!w&*iTbe!NYODaB+GmDmLT
zr7=~CcdO$7bFTik-!X4%2KEsXTfKFxGItsHmVa_v``ru}#yjZ3&kj}3jJwLyt|D4g
zaeqF*&e}u>h||y;CfZO~a<in%8%oENqBl^H^9>5G%jkl6TExC$Kj6AJqg#I_G%<gU
zwm{1Aa^i(qB&T~28%Hcy7ouMzV=t&rDK5So=qt)k4BaT`=a%J}g*IoQw8Ue02Cmve
zm%j;F%G;)D4vM&1Yp|j!*ayj=X56*6$6=L`a}p$H`A(uI61#PcWMJ%%9L!m++vlD0
z$8y5kgq#U6<n6c0iS1_Iis-a+){yT2$!p&COMX0pb2TOu*dRiX^rhHt)Od_*k>QMs
zS-!<OtCh=wU$UX{NEl8KLa=ojqvs$9d3iIcVUR6=Mv2jqqiBsN0zKL(j0a99khwZM
zm^xYHCxGWdbw@bo`E)J&jGns}ZS2@Ql`fxk9uuwjj*C;nMi_Mt70W*AU)n6ZdcJk`
z^#OhsM@U8ua$gj`vgQ?XwXGzYCB1;#(QqhWH+7zg03m@FIcOQfN$tao3k4lM?+I5$
zD)m+Dix&d%*I%*$I!og{rIa%xY(>4!sFVe^j(O1HVcV7t-<Axv4Q@@)GIW<ciWn8C
zQ$s8r%*fRrkEk5*@F$JqNYhCO62G;zfATQMSHL^n?X|>;-P1fFTA$_H5B@#A<a3DE
ze|L~+I4P1Jj=z?A0%??W5(x%_4qs9{vo%~b-xy>uXl((1B3*sOkP>%&U?HS)@x#r`
zqE?<7im5k-xvrsWr8fV$2)}gG>YH!DwPkJdP`#5RS5$Su!tVP9D04!r6}uRf%?5lf
z;YncK<tjYiI=tXP$vRP<u~A^Sbo@V_;9bsNZvw$WkV{;1lce>gqd2cp?jFw;retCh
zU^G5jQ7rRG=<nJOs(M_Yadh4R*yPu2R{gEFrBMp``q%nNsqAGOSL%!DfeByvyhef+
z0?hc5+tcxRc;-`+LyTuOHUFmZ<Y`V3qWmbge=~~pms2T&q-c;WKYUCYpJMqmz7s7u
zt)+W*^o%wo$1y)bKoH4fk0_Cyo%egolH`;z>JvT|?X#HA=6u=PC-H6cy}%sUU$Waa
znsJkxMTUWs5XGX>)~|Gon02iiLCqD*F$_v$G9dE7@q}`{+l+lsQxFY|F4zwc9(YZt
z#TD$#D6sMoOK#pDsq-f#vKdKZ;&K<X!OQDjIFLr;Hb!7@{DpEUTH^xHlR0ECPV3bn
zX4%Q5Lzg&J7@^_M@%IObKt#h1yTDPbC^hb=HTD?h4IN)e!V>8A%bMn}MoZk;vBTJB
zvx&3Q=#8E5^z+N@2dNeLnI1EJH{`jFP8QAPhg_?91n+dZI4v)j$!0vle_Xxu(%(7#
zTB1%Bnd#j+ry?rhFX`w`*TtF@$3}!!D)SdjI@pra6gDEZ1r)bi*sLg%SDyax+9KNt
zok6oWq&p+LB;^zK^i3LvL>mdoy6%yP6sUUdqf4edCe|pf13={17H~9w<`2H&P-{|r
zNNltCmx?&WgWq%-vS&f&JYO;1;<s@eT#Hc7((89k{%I;{zSkDDC^`-~<!3+kFytrW
z7(b4Lb40bLycZ|Hy5R^R+Vo7szd*G<M3ui?VT*|KSMtj5Nvay%7jj^FcD;@FZpLS+
zNR+lK>WeX!;;5~J1$p$RuJuJ_lgfI3G>Fftn0>n+89@dUw%;*2m;LYrk*&gOT(9?B
zJik86D>7=an1y!P(PbH*qW+{S8~mVB;i%G}8;VT~Cu{5E+Vq?dx$UL*@UxLP-QfXa
zn)KKiSD<c9Qn|%SHOesNA2hx}1%<_<ar{+1S7N-5qM(mvs9stSgs0#S+z-pEaiEcn
zUU-La5tpReJWVQEwU$>u-cOHaLl<9I+-EN*te?g>g>6VBFrBI54th;4?40lwN!?7h
zN{Xu2uESqhRG_w!;{wm$Y5_qW1?%bD-&2LJYnb~RF-@$hommi}B%Ig6XU~=;M5xA4
zLrZt;aL+Mnx>3CzjBw$7CPQO{>|<xy&(<~iI5R|RZ9t_oTuMMt<>>MHzYuE?>U*|j
zT;X%KwGJ<^z)M|)kG&DWeX5wvOvCyfv+_p2apt8ThUYo|{=fwMi=|VXa9x$c$f?}R
zt_~3V*8rEYkQ?fVo2+meEzU9(vA9GoN6mX49HSfg<BtBe!W?2Ed^#!=>T^*of@kIe
zGwG5HMJL!Dx6}j)0p>U2gmXEBru+&9B+w8E{XKTw_S@!*OJE&-BoNh$0(Ibo-u+hr
z`~8^fc<vuG?cUW-oB1BL>5)LtUjx(4g5^(?S8Z`5*t#y<hMDk-JF}n6aVT-ADo0_S
z7$9nnzApmoOo=GXOSM5+n)Pov$f7a!oxezmsIK1Dn)T0m*Eu|6-y3P6#?GM4_|~Gq
z`_@vB+WHuN`-2g$+3l%Mbc8scM}DmtH*Fn07pOK8ydAu7Fhzd<VmJ!)(8QFhgKRT8
zTkrK-^x-;=qFOL#cc);UOgv*I_ABfnE<xb?!`W-@PcxMFS5cy)F2-nm1D6(m*5S!U
zLwo$t*=H-Y<>NDXOa~Ozbz>=|R_%*{A{S;2ns8V4RSU-N>dRNckz>~6UIQ;5VSoi>
zo%ctmn5JdZ%sc)OQtJ^oO(Doa{;RX_fhTBwTFDJ`_0^q#OsREH=<Lmq{IK{*iy2yK
zSF8#DvzIq(av@uA_lj{U3Fb^NwY}rJyw^uUAj=pP-s`E+LA17WDeVK7IDag*rmP4_
z@nNhf93j`}EPG*|8Ku{~DP^#CfH-02-GGUJG*u#*R>9{*70O-g6b`uTk8#GZGCKcA
z$D;m;@lxlWk1g5g8K(8t-x6)7KPGqKEyR!yzaOK;(q7}|)=kWHZtEl9io|^-4hXh2
ztBahQmRX&wsK%P4Q`W}=v1>vfaRzPEILL9VKV@Ly-zy5*%18C@1feklHEe%Qe+mOk
z^iEN62^@>@{29$|{=#rB%S2{g<#R_dABO66&+K%B><qWy?Q0j)K+zdo2l!vezlh?G
zf=}!IA)hk5KysxQ^dCKek(sRzQ2WURVKu$Kay4QLvi2+`6|UW9@28Ch{B8axu+CAm
zN|h-~nwtJ^_l4hhA|S%2b11puP;&!(7|kPD@|xTrA|zwg7xAdPyGXrB&l(CnenIJ*
z?NYF=5X7kc&ZO@5ImriEWs3leNltQ@q7{#22bWcG%f@KJhEyc=Ypu<Zbb?Cb;<})?
zPG;2J^%n9jBg;#T%^8$;X*9pqUi$D3`(2swpBJ#^<7hc|ysWH}*-Q%(GNTV3`!H`O
zoCbdrK8`b%U2TxY4#eV2%B}1of7Ms$s?!nW;~;pz9kb<*M)QpS1$ku81zWbx0+&f5
znzRCi?V6M*j(LBK>XuNM?VT7X?KAx<_@X!6Ei!fQP4lKFSHiGj4BUa{GJAP&+{8_c
z16A-lEY_@k2?GmnFGL<{dzTD37Or-mtb;5<#EPi>G#kipRmplYGg&UWo+J0*>2$w)
z{|ZWY4L&tJI{mIL<siJsM&mIeghn~^3ch`pIC3U`<9!7K4a+o+rnIA$=|9!FU+u$Q
zF%3!mae|F_-`G4p-Y?R`=^Ewd&)j)#v;E4`%sSGTLlz|DF6ELX23CnO<Idg9zc2;+
zU-N1Yj9=I1FkcIo-p$uldyVY1m<nJ)c7B`)-2Fzj9I`*8ZpS{zH>8j(G4rBIoqX+<
z7tM!&{%wxeZ2qA4b_KUZZ?RJpPU@`QgT26hNi7?1sBxI3fGGR2%04ba9jhX)v)BUn
zlqH|=nGSc4nM*MHnh-~g81_!MZPUp46(sHLcGD_<gJph{GtpY~>vBVMvllB}@NY22
z24mDawba)mIE}^u;WV&sLNS1Ast<_uCE^H!^`ovC4$^j?ZdzJeF?-LrIiSM07rw|+
z7Xh`U;VWkp-rs`g-#4A|Blbp`PlhWr4^33sJM^ID)4cDd7)ql%$($_~{an^#_9q6w
zL1K0<<0LC^L*0gsd*iV-3rz7ugbDR9`(=2G`NU+`N;iU^)0y~!<sMsjTo1R~Uqqbn
z8uDeEod_LW##p0hmUg4oU`%EfMU0}Vgctfo3fJ_x*4rQ&c&<{Be7!B>MGYWmLHo@0
zbVb)2ywu{Q<Ba$_A(T<b8cW@QI+VC6YFb|-4iB_S<y>=RAN}8nZ9jG1oT|)}fYYsF
zXYIhC@A;)jzHGw1rm2vI$G-~;_n1hY^H|!hLIp9udm@V{1LYv?LG)5k|4#JpQGRRO
zzxuT?d*-llE{o+GO9K)h=++P9g?DQ|oFx?gKwXHgfVuKD_E`vCZ$Ao`Bcavz6>+27
zwH$2CGt^}l@A?f`KO3D3ix`h{I=`DsUfZAxni@~Lor^l4fr#E>=i`eU)~%`7igxV5
zGh2qW#jw>%sQqdeFj(hsYc-+N@OdPuxfGu1twf0>z<Vqj5<<)PiaQij)t9(*-07{B
zTQ&g@i<-#q-4_2}mK|JvJCn@LMw$BJ&QH$$E)oP^r}%gZWc_LaRuji#%-l7t^)R`G
zyd=CW)s|gm3j-jyqk7o+!VH2}XSUw>go)!CFJL3(=?WO@e_PHR*rb}zvhgF(6moxC
z{=sV1y#3Zn*4QWv-sz&%k41K53?;U6@2uirExftno~i0q)$^-gmcNnlGkRnk2>qx^
z7$umxNvk#fwP?j=TY8<5O(S^qFE7RGi&gEuYJGO%&Jo-XJ^k`{t10%tZ5)E21X(fC
zr^TxMy9D}$2L!tM31JF<%2-M*(bO{)Ux{Ltg+%Ya<kI)V@H0@gmCn&|0ccVL2d1I$
zbDvYuo`wt82X2Tfyq<nzi>L=kGw_PVTHS<^M028M$9X3*9s!<p;k{BOwa-`hqeV<6
z7q(?kkp7ldrTG+y8YtNl(y&UrQ>k&6?2oQVI6lz^wka=RI)tpW{(j+;SA_ylAj0GE
z-OAuz5s#abJM}evcOA9Fx5{tbL4w~W4aOP1?U;D9kcw2c)$VfZT_D+eHjIHM?R7JM
zFo3Qq=MgVYx>GINsEJhNM=to{X9X|eL(=;0z?OAYDv*s$OR1i&XuWZ|g@>{J5Mn7!
zNRBM+Y<uWnY9O(U?5!cvz!bI)*ZTD~{9TpyYBXmbvf$RR5K0tVozKqX@r2lr)JCXN
zE|#7vSpD%f_c$)hc(&xmHhPmb-1%FnkNv+F-#5D(zuV12DERW<0Uv4q{0Rm@9`J3R
zB3bpB5&BDH7mN;z@*{iP((RqzTlx(Ewme&19EfIzuT-f}m9u#LqFLz~CY=5j%GRT`
zN`U(NYt9f4&T!?v4;`HwTUutkSZeW)n{dv@_1!*5H4f<Dy@+dr>8w6G%b+%D0ESKX
zY};)kUSD#+@-LjplViO<nF^l36@x=EkKj1WG&us|Oa9gU;(6vb3O@!Gt0cob$VE`v
zGKI~h-Da>}$zhF05&<GH%;mH5H0Ij-{WW4a=dgB&u2O`q>RqO7U9{2yJ7csR@RH@m
zQD)CuaCPD>S}`v1x~irLJ&JSlMX8r-^>$emrp~HLtQ1I}>RJ&Ez$@o|Bu$8SQF7Pm
zZz8><j-XkseWGm>nfbp0T#)2wzS~;rLRsPz;&x(fndhAxNt2X#{%sO&aQv1gZ%KK!
zPQ)eWGcA-37nxOrjNGP8My8e>E+t3n*U)Mb2^F*P=gO{U!L~8Zaf-<NLb1atioao5
z+Wl|&`F2!aD^8&X?<dwYt%y=AH`m-FkABCEqfz{F;$tU`fft!53faYyjDgYmyT4?Q
zC5n|m){T%-?O<D`$w(AZ_g2$OA1984*AXNE-iIuGo<yX0mThXU2r}o1Q8r%hOEteu
zY4d*Sks4n6*oz@2fga2h#NVNcBD`T7S$yKj$~0EcdPU;-mgMgn;=5pLHR?@A<7(-C
zySE-4oa87k;e|u?`Oz1D&z}>z>}zUQXPxjmxQopY(Dt#<Em|iq>t?auwh`8(*Kbf*
zxopqj*Yn>;FB~w=p;JuSP<`zBa9w#)>s1WAQrG#L!&~K(fj_@ofO&VULQQC+V`G?^
zYSmY~268<$xZ1if8@bo;Mr>$nH3&3e^heSZ`AfRE^YIc1<?Aa_wah8TttxTK1i7vB
z^Fahz*;l-bqGD|>PS8aS#Q*n=aEh;HOfc(*)QCIt_29z!i)`y~P~eWhDsofCGh1MJ
zh{p=OEh_F~hXju1bW5WX$$4jurk&m3@};Z9vZ5(}81%VAt79ufTf|jWs3LC`8`vyM
zdLkM6HbN@skXKc-Tkrm{y?6Y+hM8CQ`S9;M(`8d&v+2R6ONUZAVf+g1Z0s1aHpJ;r
zJCZ4~%S*=ynQp>yf+d(%oi~g5aF2l~nbpjiLczU7>3F}wOK%LLxru<ddfR#!iapI^
zAL(5X^o>s6b2?B^msS-ys*CPfKqdueM;RrbWr77kCU)&?7cByiaw_+a8U1K<?~@0f
zoh<gfxwfCAFV0EaOT6?`pz31XGr|~#9ilY*;G^#qQ1Q)*`p`*G3mCU`_XQ;%jc`Cd
zoJP+bw1+jfursW*A4;%f(aFCu{=^C6VpO4c&AlsS?Ku!^yTZgorSRI&qhi2sLyFPg
zz~M|I?X|4qI$7_+Ty)`&CuH&(J=^T*B(xF+E)6J=5++P&{GtN+e-mW(sk6~9(`(WM
z?H|!RoqxOHgt)uC6B$sp!FumyxM=1JZl-rrCcIjs7`^J#=&!8;7(=#|QYT?A6#i30
zK55PP6h|bONI-0zf~wDNhzEl1S3fXDiF)wfvWdgMHDh_D&=#}fv~5^=@o2pMB(*(E
zYsF!n*;T6j`F?2$e%+SS-*V+^-plRko6j9c`ivZClJR`#>M9Il!~;nrYa`RYe%P)O
z`pQ{W%A%G3KYOf!P8n)7Z{`1M#<!(&lx^YW;a_Clu-mNF3FD@Dv@TvB_b_Zu*hK%e
zGP%G>ReI-a__I?pc=dZ4ZGw{i$HQ(4qA{2JK5Wy&?J+JJHpE4-AT2fna0uLcPK4fh
z7I&79_v-M+U;F-#dOg%W^!ogAx&q${QBkw@;J&uwap*AEZH(P19q6V1T-K3Rg-{T^
zu%|lE&Ps(`Sl~;SbeI3<?#?h)yEHwP8>Zcl3ULabIr}ITg#2qg^7rDYjMSZy2<Wln
zly;Q!*K^X<hu=@yr><V&m=V;`;@*^2(Z&TaTKT-z-0goA(2R^LKE;Ng<zg2N6u$06
zN+P_*$(iS_uEp#rr^W2c%W2n<G<RxOsqznX4XriQ@OA7~(#dW3Tn~Z6Dj949sDAjH
zYp1I=c&Lw*>OTkoY$a`;O#NT`o1)57W0O4<Akg(N={rYjGUOInu~V_1bRer8jx&U)
zy?5|7Lx!T&HH%wIo}vHI_cbI?lSe~Q$riU{<C&J)&8Hq2$+T#EHOP>2bL2X*xaa>&
z#S%SoIX7%!(6{O*#`H#|dMbm<#lK*LBpUT;6=L)>Ex>lOAjzgLfcQJl|1}LANzqT}
zY3K=CzUlO<_KM_(LSHlkD%0PgpYK{_t-ebL_6G<{D3u!SXUM!I*yhQqIR^TCd-shz
z%o$TW6VQZ3&UPmCSxL2FQB&GP_%ll)_Uy+2->k&d2kO6zXB#)p#p!FRm*O$P_!*~S
zkUr*DSwqL1huVQT{yu^J()`QosbWk)qam-ZvE_vaWEw0n*H5sc*`@WSRHzu|8Td{O
zEzGepN@L%e2V&=?yq|)G%)bDav4UF45^#0nS&z8(+c@{KoxdD%YmL!D03nw7WOsLy
z7DC!LEL5L7<lH?{Kln}&)Wv|ozSJ95=0~VEs<6(5`LLEhhw!=SFj$~I5Tvf};OGuh
zJNh)wOqO)<Rv^LK4l(<B0kban@rq4K@U+>NS=|2|Sce4BdM5MacX+i0mSU;Re2qrz
zt_AsaTXrrNEh`AUkR=Sd8*FW*Y~+<}OY@<gD(e3Fy*U}yLeKWUT`^}~R_NI1q^?|0
zx~2WFd=q0AJn8hzhHQy}Kvt@P_2C@sU`R5(Z1--pe9in_8XLc+BTtf$4xWPStv#aL
zfRhos#d?mM1DW62q#>x4OT@k#lC&-qtH2OISyr$tt@ZR?5zPxl0VdtSwo}oGfxcqt
zxb2=t3mD<eFgIMVacijxl12skoedP|vY$^v8%qNU5^z?0QhEfI+2gf$Xh|kI`c?2W
z)~9t1j`7D1p^jik|5Cx?!1~C>Pk_^3P-bbSK8nQ<+_`3Qd26<=!jNL?-|FltuKqiA
zh<jCKtQ5eu@;LUyF!psRm|&Rm<Fy$uU6-R%aZc|;3mqP~xV-7aWbyfpR$JyPw5*{t
zS^k={)4Z!`r%OJs?#0N=`$UbRg))8J>hp+9c?)xITgE;3T6gb2rByo*-%*kRr+51{
zPvdy6!3n%ZjJlN~>0Bi9Et`88bJf~VMW4I~Ut!6(<KJe-DYq&rRdgx(;l8t~N#2R8
z^}(VTUyqrWg;s{S{|5z)AFyNDt2O&U-gwlJA!`MLmanvYSExBsZED$ZEIz*fF=4!=
z{Tz{Ytrv@Yo{>8W&Z|A3bxY7}35{^=c?sJC<Nb@V#J>w1rWF6SGq3XM%7)+8v3e}Y
z?kws0$jK&t<wGRRxa2u2wI3+o{?4>WqV8ytMvQ$*&&#U#WX?8~KD)e}2Kw2S=;KqT
z$OY;RS(Y#%Am)S8c?}dRm>9XQ%Ez_^%D7SwDRi^lR9Kj4J8{7i7}MzH4I6KW<0S(t
z7is-OFJmAdDFaJ>AM8Enqdj}g@JUD6HsEC$a!BMe%m8ORps*DaWGblJQ3R9<jo5gJ
z-^L32xwJZ~9Ni+$*sE^D3ojCl_AiXn9sLRxdQVq&o@YhHG3MkID@-fJ9TV^#b2AvT
z*hH)00T+zQ)c>>ffi8kh9-6f{*CTSxb&4Dp3KD7N_m+xyP_K7kUY4jdJG_)r<)J7C
zoYXc@fe5LEHN9>Ta{@Hs8;tJryB1h-&X?V-4_hNeNRETs)$z6a72=QRXAg}quSZCg
zyW>e^q9<f>!Q1^wvFK@gPQUE<;b`UI{;p&xyv2>yA7A1_`7E$TEaNu@!JMA8jg6Hb
z5ezNu_nxA}R&AfLeri}B(L^Hjcf>UKZ){ZoMwkd+&7@n$l}P^1_n0N5Ing!8!>BDZ
z#fZ+9XIMjcFxcEf=Kp8`>{Y-1Pma`{W}26)5Zg;U(Z)|*tsILaJ}_StYCA{8g|96A
z?$q|*+7OnOysJj6-n8LlKFWCgTF_Ae>}n-*DrpIx&Y5;pnDy<^X@>4aX7)O|eQlVq
zqo|34r4t>06jR5I6zhntmc&|9slYv!U_#TxCCF~-8jX?8_cSO}%J2KkD_|*sIl#kH
zBQk1xEiG)2+I|!%U*5WaNsV7m&*->Gbiq?yL)&Mo+x-*e9BIDx9L!x`C4I#toX;*z
zPJ23v<+>|?<RWTBid{a_m3k%>`Bwhz%*RV{DXBf8Yt|E?{cPmbbvOB(ca~85i{d2e
z&wFHyTwHG-?00YZ7xFg11Rn2N5si0iHV`e5y2=%+y5|pj@y$9kD<>|RssWKqw6Kxb
zkHb`;*dlwx?ER3wC(A4YNBXkk!3S?DQpx1K%Se<9r6Y`mae<4o9%jQ~vv1Fy{7A^m
zbu{-ZR8kO&x+}${r5q2lsF)7{yxPiNOI6rjdvGVOyTzsFj|!5eHQeDJ${5tKlI+Vy
zUY8HPU3i&1k-c46xhRoj?8j4gtg`oEWvz5Ik|VV@sdH4l+&i(^VU{Y?lc;h(n4kk8
zo%9cj_#qn>^>Jyg3-~WEE&X0dNLirt6ls&t2r6%w$6FFN{foZYQ=fYK)!EgaDE=yp
z+#fq~^Lw3KCO}-6p9r~;<W*9jH(A=G)Y|45w+S~tYe7yn+rw=K>5a2kUQ|^F4gidN
zghtyXijD~^eHlfI#gnaMADw?I)oYU$r@#@SQu*SqGi{hC@AlvLVTGA5B!s&^hUtPy
zbD}j6e(Jp{2eblfuvuctbA0~Y<V!s4ueNEwg2>+f2_*<wuLW@aFRs4wFRJ(Jng#($
zC1hYI38lM3QfWcDI|PRA0i*>6q?HnplJ0I{=<cD1hM_z39{m2EH_zw%2j^U8Uu&<m
z_S%$~0Kr85!wy1U)KqVo-*#b$S0X79Vsk(x?^Z^k&E0FsonQr!0Ut<A^4)6@FrFBi
zFAl+CSduoDQF*qlqy{-E@FHDZNsBFBPYTgq^JR#|Cgo?j&~vETWKs40uX|Hb>#RdF
z=un2Q|Gj5xlh0r9#d=S%aKDpE@$5-8|3!?{?m3S|(#Cg8f@x!l#c^RZk9+)Ao){>~
z?<ov{Zr=&csvE~2OKERViDPhG><l-|n!GAb55RG`DEAZqVN$>)2Jxso5p%;}cS8Ps
znmc#}h<WXKVQ6#J)r6z|ZO+<@$DgpKA;&vy(oPzG-m%bdh9w|{%>`U}Yn37CE0`7X
z{2!gcz$}kn6~5$yMNSoR&GkFw(1oi6b%7#V4DX(W=_dSMqJ5Q!en*!VflJ4kf%g+1
zh#eZNJu&z65Qi{|^E4cwZS51Bh?}!UfI%v+_dQXY&0J)5=%^8=FJWk!_OR1@$uFZS
zSPeHZ79bRe)2J+De(G;Fizf0(G_XTvWRX}_yGN3a;oHu8<0#z1{n5qZe7tX>Uke!q
zQxo6$Zv?sO=XW=8?4LX6*Orx)&xZ&|J*c<;fs<QS&=$m5=JhwgIg;0$5{5r{`K*}L
zbGu2cj<Od@77QnWvdC7KJYPYw^QKp0AC$Ld1ASX>k*m_|YD+cuB>!#)8b1sM-sn^p
zdM6-*?WcJ1CSg<WyFvIFo)?F@1F~oFh|R?~-{!WPx~8U#I=EfK#Ub{Of2;;lkBF}w
zUx*_s=+B!w<J<S(D80>^Z1jF6Qj0k<y_^U91(VXT{~X&f-mK+9{-8MweEds_7V>s#
z8<8wGtZxb&Su4C;Df1}oVYK6K(INy+`2(sa0z|W5xEr*l(m<V<m4Xlc?5DgDz36MA
zW|5@;T}%**#Z_d6oa(1(N2Pe}4vp0YqA*COPyl+o1-kCU_YaWguclGZna+uo+258D
z@*QyeaL)^T8n8B;F>KCVP8&-YBW;u@e%d*D;li&hu-WV%U%Fa?-<Z?iqxUW@(k!$4
zE`TXDm(p0syUX}pv#P4(1~T~2UhhMBBddC>l(=?taYAeVm8ZFUQ2C^wrIP;Y3x1R7
z+--SQD>}jo*0hgpnymf0%U(RnQFm-aj$8{SKtonhBfPOk;%;}wIo;ZMxy88EU(fTt
zW<G!8IbMk#k}rCTb%)QPBxH<Nqe63sMC<s+us2kt0u2fEUmx$5>dD48k9UiGqR}4m
zZpo?!>JJcFZ#gml4}ou|IP9K3UMg+w51;*;>mM(b1&Z`VZOHR0A&mIgr6Vh-+u&}{
z<d22U5Z(}%Hc<>V$G<phkf_bSbv=O~z=U~(s;0~^*{9FX=~CqguzsRy*?Gh2EtZ63
zm`?LuH%?YVW8~LKzuruv1fh!cTn){46T3B)Q{bv7yO#?g{*0HJdz~_3t3}!kF=hcI
zRY~zJpDI`3oL{xcVDl*ztZ!hTvAA{=d!m}9#}u4q($u*mVR7a@RhxEfv9VuB3cG&`
zaSy?u)A__dDTku~<S!IZhX#iLz-ie%Jtqv!)Jzh+C)#4OMeNkQAH(*8**hA_g}?-c
zPyDQrklOm#Tl#4JG;C>H?;i;fvNAGvVgYxiT3NXCk{?k3{I3Nv3ecn~We^zdl*=|B
zb-25n$P4h>Z7|sER55u{v0)T&<2FD)8b&7nYFAKgEZZ@-O*oe!d<Y>f7fa=5Vob9}
z_%i=oM)SaJNa%mBJ#HRL)AHY`ewKx4BS<DRC29!+`v}g=iBCU@YRBDTDci#aL|s2W
zfIaWkIkv+X^9WD<SFelJzar#lDujSI2V&zt+9Bv?{(xtUG1$$)Y0ShiZ<3vjOnQJK
zLQIQqQIvli3DN4sPZoHZdY2h}s;fMu6}qxk-}C*3uR0mzPWUMvjo*k=rXYI61^D2$
zr6mnzEI8wH?TZ_JZto|r`daig?)~@Hts(B2RdvX-lym&?lUutK^eS!1E~UyzoLZ1r
zRQ8$Tz3GO1fggh4^Ci)l<ShF<O~O!!nZu7;?p2QRCj%PC_L`dk5qXzXTFe#=PJ3)k
zZEdwj_X~%eU(pd^M5BvCzvF@y^DiaQKFCMyDC8JGUF<S(0_0roV|Z4^(aE84AJ!T@
zeU<E6p<}4Py_fyPH|Y;N`|IX^$p0oRbQBFv2>2X0XukMeKc9!BNe%m9Lui@@lrq+S
zfggL#`KFShe%mr(n*Ag2sEAplk}(B*W{+UYN<K5lFOqCF%E;X#ziOLNeyu$8p-A{q
zulcP0HX{sk;r4gH)u0l$`${!(E#PMZRkYN3&+>{fHIrP6HdiD6N=d8GYB~UAhm2Z4
z(!_5OPi6m&-=-M9Tu(qYD1ul30@he<2OQ9zl!$y)ADLVuK+2jcKRtsc2}3n0f?z8t
zIFg)k@f#PVo+5+ofB~8>`09@+fGP-QN>5i<`qaIu=r|#kw*>&T)1i#2<Nj5ZVPKe~
zU_|;&VJnPU#du&ccHbgj+IV8DV@a0EEr3lXJ-bhhKdl6Kb^HUrEa@x)f68v^v7qOY
z+_Ha{ka|#&bY~Tp!y7w7))KJ8WjEqtZ?|U}{MCB#GaMSDA<|kRmt<LhNpO)+!)@ep
z?>;kd^6Y;V<=^3lSN=A8x#{!SPM=+&(#A1bra+fvEj3OF7b%4$Y=}MABXB%Y$<S=B
zxX5@I%W?1EaE$JCn|JuHhMyRo0`M~Fnx7^Eqq7P)3P!OHP+``yyINF5Cf3zGc?cT4
zceYfM!>k66Z~GNAF3v!=K@i3TzUuK#WPB3YvlD2kfc}<%{#Lm;N-`vh`_)6CfpFi4
zmFTF`4Gb=dAStiQw=OH<<_}p9a$xPqN$HW#@t7P$Ri+@<AJ+kipdd@oTNFcr1Cd}e
z8j|_XVB-`+a`_>E!P;Cfb!X3}IVDM_#YZJjM@62f`g^k5hNVQvqK-EN-PH(tPMUz~
z9nJ{^M_P(FrVYxdNe8wPwwb6f`mxDlrW5k^RakECV9po!DnH2-C&Z3KcQM6_AWQ`Q
z`<PEwP_?-&b)Bpa{KEp2;R7&`=U3>pI*?GPo_l;^&ph=)Og7!iz~vrak9A;W+25$O
z!M}p?_)U&?ach9X%C(OMa-+rAt5@di{b=|LICtQ;U9|_-E|b8Ey#KTTJzLGDe}}%=
zA()ElqOwPruqUgL|6P+^aNBmx6NAJ0`i{nh=5v{DN1MkF(V%j|RWSL9;_))I(}R20
zM(HWCArn5MbHJ4IGYZ;wDImf7M?~{p6kL-uAU6I7u2qpc9jVuRWkksazY9;B>k0X-
z-!Xa#YM`GxCA*i69s(^)`2*tb!`)Ey{T-kc=#0ZA4_M3?#n}#YhuuChBAlbP6lGUA
zmXs=eTM)93c^xNiLq<lZ+eeD8R4X@6Qg^mgDbAT>n|Zh6=f_|2&IdD_aD!0G{axRG
zr#-p)X{8i+_)>C=z?$3VO|fZzv8g(sse?x$?skgUXUx^-H6`6%;IlA~Qe}nnPC;9R
ztb(Vw0j(V9-IJz?CYUnDzu*OIRXVjbE^p$PmU_$3#MbSq(mV~j*4ZlJfX?G4<j~Re
zUtk-wlV$I-JLh5`_}axrET8C4;NQu=o3_X?wX;B-ovS~0foDfE%Wfa~1)i{^{Bfef
z%OXESh2f#AWs7B`_K+8|qq|7@I(PpG$?xV|quAKaai%YQt3pujz9Or|6%l#Gtv(-S
ztW>+r;m*#7G+HXh_s%;I?6iSh2yT_~G>}O1+t>dOY2MeA5HGOZG`ngZv=%!{E0SFd
zvLS$Xf`WOIOfLWZB!q*fhg^|JbI(c62SW=jQOebCOi2O}g#f%Zl;SAsp_}z;=?&tp
zqBU0Sutd1jO7ogfL#^rN<@VprMUY9@<~U)(B{h?@{ct6{h({o0O-6lZBCw%<JYgeB
z;1gg19Y2&-wKF=oGZr)SEr_G|xnGKrIk`s1mk8BRL*~Q&(LrjaPKQ|nA0K@_Gy%)P
z9a!o4x$ErsdjhgIQxzx%dIKQ6SLvaKV^UJs(LZNf&-2piKxP+dT70?_Gh~3lf#L)Z
z#p&={>*U77qL!nI4W90c*nskFS4!w*v1EDniPgeK%_48>IpG@wQ^x2O=aCv%ZQSwf
z`V#YciK6^9ocAT$nmgX@XQW>C-VYjl78mGKeM>IcYDju%3-P^k{=3~i=?78?cGYB4
zM?MwKc#9TVe-Sfr%ndCsB#IM9_5lFNyRwMJ>kE~G?RAYEKsn}C2JJ8ckeR+rVsEGF
z$0?nyV#V?>R+0Cp8L&IC&}_Nu6Ulf|y^>^ic;=Y@C~Xdo$wZrL?x^*c`HNS`3ntgV
zFaF7F(VXgBoP8k@Gv+=*j58THo`emYa|Vsz#ch{Yqy&s1?zXsUwi;z3CsZix@keNs
z0E%FxsN}rQq>;G6MYr-h>*`#(Ya*Q8%qDc#%wW&%fHt@sNSe`-u=7_!QsK`^zX`pd
zLI(|Kr{G5-SE|rKU}eST>X(rO@nadM-5B>*<PwztndbxoZv_fZ71LrGOFNR4hieX_
zp8R>px74<6M(6yB_u>F2M568sT!Db@_(+e~)mOAu0IJHln@7<=@)^?=ukc8yR-h~K
z&1w;A&I;N&CDH`##BgW6vbQd4Fbdm8N-G6~#pF=X6s-ME{on64KCR&cMwL>?rz3|A
zbT*H@nJyH#z!VpN&Z%;^pyiM|ja45iwxwuS)`;PvZzaU!SNBSt$IBjBt=tP@1Q|zG
zcaoTdmFlm;r+HqK6ayz#>-C_+zN2fWRbqjPg#c$9?M>ka*6SbDZx^j{B@7>Xvj)OH
z&j!N}X0+S(o5bWq)uwkTKDOGm`{Oso6cyCj$Xw5b!sq;@B}{I-7(#g&mq*N=uLe*{
z5~K7`lTva-|3%g0yvA-|OP?HA!6hoD=89<!F1Zn$;3778?W)HXEkV*^2;pxeH4eu$
zBb<<{`89XDieaYFN7)q?O2<I7sKne!3@hIwmGSD;{bsTIDu*;To-*S3=DzBSop*5<
z1Tuq8U2>j%Uk2gdzdCZIbfS@@N-3WweEqc~^sLEIj)abF1KQdMhh_@VN%FZe$1wzj
zaO`3!jAkGg7MYJQCBOVzwjOX|H;{dnY2IFYImg(R>z9SuRTLX_+BElNS2djNs7^)*
zd|iYQkk*A5xaAtuD7jZzXB7%S>Rs(d;tlobMm}zX3k3=Wl%u^c!OPg}RE6W%?DcEA
z;;sZ8<bMGI;rTz*H2Geef{v$<VCseU4QL3<$_TwmKtFvh*aG$PLc`SE)UZD^%>-3-
ziJ&#g7}lj+yPfYZn2GZdx1M#yajmf1*H&9|aPz!w0aeGMq+l-ZO<!Dezg?LO;mT%%
z&kF(Nj61Yr;*vc5Z*Zk4?b_$r+etiN=_bd%O;`%qbzff>X*<buJ(_FtNT-DfUm-5r
z;xwld19Z`IMLvaNu??(P*YA?<vq-ua&U%U8%x9LI4Xt^JhKNk6znR*&J4ikAK%37a
z<V1%F&D-R1(68<o{c6Cg`?^wKK<YUZFxh4SD(E0UHt_zqqpU)1^{K2p2e<_)^>r;q
zCo3ssw^j>I2XAx>6eX6zmeU)Ze0-?~+bk@opL#p;rB~dv+D@o#9{Fsdzt(eYsoebs
z&;NK}#&A^;#Yec;?uSZD^K4KQ1u5$O5h+P!ir)#M!&5;N5(o6~ho96`qJXaDj_Xm}
z3RXf&h%ZF=QB3b8PNZqb2ae#nf=b$`zGZ~oWI^RVp%NR=GP*<Li+@G6x=|=&MB}46
zW`(Oj*f<n5H~!{heP6sS25>`B!1MS!u&T6MQ~7vFK)$=D+HYDdOr~eSL@3Er5B8=B
z6O!omC)fZPjG2%+i6f#R{OT&TY#*nrc>7+ti8jkfBik_J4N7yYa-6O6Nn5U1U_P1z
z!~2A4U8&(T-=D@fZVNANYfu+t&+clDfA|kfF=qAKPL2sfrNoJ~oL#&mbe0|OT<<ps
zo*!hv46i;zr=%cfC$QnY{)a*C9lx+)tEJcQ-v>*vi0hkcmy7eBL4GKV&a&U94!J;2
z2<%Fl?EY9F%#j76C(zM@8|37SlzE+3pC_i@PwP4?$`^AWRF~#(DjiqnVk-llB;J;x
z?q5!A=WtDDoq^hwtoEe-aO4fU)~!LM!f&a0PZ}!aq6W8o&eNlsG+Ep`PZC8HDq~f;
zGo4#2xQ${O<E+oN2WJS=j(JaHwc-k*y+O)k`|4H%en7*o#Atvf;J0Q+h<bp<7^^^2
zZGS0QM{gZ#)_vszjA=kB)w7M08SzieAGdqLWwF5JF(%Qqkj><H>U1WtoVOA4w$?es
z6UHREg50?lpxnIND?EAWzjZk{sp&ZRpP#7h+8LihlBi)@+)&isUIFE3$NvCm@vvjl
z6-8uWJ^u4(LUg=nk}i*kAau37Zb&!CS*`B`oy7+OxMDcRS&Z-&NgY4^F7HCJv=5OW
z6<Aa3yODm0wmC7`E|-K6AA*i9E8<R0<LT1q?97hGxMSn;1CADdbmbu|)+3W22^K%V
z0(x)UhbUxX<9DvnQwo+P@s?Odv|k5Hv3j207GZ*3{n&;G`Vm|!JTzfOC9=9jklLZ?
z{3VXt#SVzw?GiyvWD)jXm6B`&M#pHFOa|_#6LUmq<JDaFpd^gCy{Mh~-1NcPUZ8qn
z^zM6Us3^}y3&F0A&*}TCop14Z+t+<sRO}o`tAH?A9#4>@GxJ6G+5Ks!;*iI&%)x+<
z1Ho5vgMrpExvM2ueK}H1UfN=P*@O1@Jdj<3{TxYO8uYat{7rte0z96P&Py{(s!}kk
zS^;0U-PxYo$YsWC0#uDX``PzcwVecg3-VZ4$@tF|EVo0SfQNNf7_!Ujt`$Z0csuhS
z#D@MyptnWwd!9HghdKw(FQc*DZbzFqKHE)oxvyvu#(;$fmiPpWVhNm3OY%g{k3$bV
zN7v~HP)uZN@bJz;?KQrHu2qli+>9w+6|tOt#j?C6aJcH&fG~}>D?+FNl2a+B_Z<DZ
zkeGK{AoKVLw#I7aRGI7NP9zx;)6{ftM^*=FJ>!KSox4kiVeea9@8i>B)=-9{xw&2(
zUUL@^E3{&YQz*o@{C1-4w{2QBv!NUyH5RfY`SnH02~FV3!RqhU=N-)B{nYPQH&6kA
zR}}u6eTj40@{RRWgXklh+-O2mLYpJ5833fN^BV(kuwE#b;qlvd1vLjKCRAcoIMB$s
z*24b%(!b2&$649KSQgTy1dz;=5UiGRx&LTgfcD`ZS{kzt)&$*W+?l(t{BEnAP}=H5
z?tVv`UhGPKGO<X8D44Z6as*Q@L%#_0yctL{DT-dsKyV*LO!#W*6@JOl%zzbYO!+hq
zL$XzWC}{uls7iQj`(xc+CslRpO2NC!z&Bj)5pfQWP<AhO*}T(Ol_&$<*!su7S9<J_
zpW9hPkl@Hh?5x$nyInMVQ~;U7WN9--e77A%*5oq+RkT*+j#^hmrLxaiRsxw?W5g<H
zA1<6Le(bi;PT<s*V5K|BIzQD4vi#u+22%1j@S;a{9Fk=K%)(fHZQ`a?Bp1x4l@D%A
z#l%boxuzHo*6`FfqY5ckg#aWD`OwH;MA?=>WKkf2ZP9~4Z%Ka{>3UKTQ19_r#@?1?
zO85i2y<Q9P&YLftpsO6ed$b2jJs744Ow`0J`waUpX_j&EIqRNYM?SW8g-oLiU`?do
zj^?>h4I3E?*<389YUfEc)!>v?s@SpgKwDHmsodLPtKhu)i2N+2EL!Sx_VhZ(W|OK`
z*hc2-a^&$%S)7t@<4T?3{S3FcrminCif|~D%J_x3lR5~=zF6pC^0e2-T_YOHJi@I^
zpF>O_p`+m~>zk6%0y(f4@0zfBFavbG9W~@9W<ck(tbZ!@sGI7f#;c}yoL+j;5j>12
zBl^Qo#W2~mp&v3&#jtDKRZ{PVawb3oFFh4^U2g)io)}S-Zg!Rvi@jcr*Av_H*T7E^
z6-Tj(Pr@uGhohH>HZO8SNrJS)DOPowM9X6o6B1c=D8*K^CJ4$ck?z&VcD%Sg=~>h6
zC(rB^GSzj@q&+=BK3~Z-D9&5NX{8c^uK+7jJYhzc7@LqbDu<sz#>O+*TFwTGPJ@Yj
z0TwE2o0YI${K(w}Q1HtG-!(&?w{b?l8^L$0sry`EK%q9eF&sInVOAO&k@67ul&NS@
z!Op1P4<T?}^&Q1Wh&#guwCMrqoG!+WE4z=pA=i0;P7R+tW$?Tg^Z#RhalG6^G46$n
z3~{U0JR@6wEIn@erl}C1HTFE<J4rL!gjRmqON|5JXi7CmB$L(!{ae_iw{GoAxLICW
z>RkdtygYw@>YZebQ(Rs1qxUzO>NCtX+jaTz+EAMpkgFeN`wpg*gpQdbte<TMo8Ne+
zI&VFKh{c7^2KwD)&$=jS3#603_MUYNqzD1ysptI3RRO|ec<>yftwN$t#Gl9-48_QJ
z6>42`4<XEOnGDfflu{S_0#cs^mmk#p!p5;7G(R-pqLR7NCdEr?9jxDZBeS-F*zZ1R
zNwFZe2fkr_OI9yG>$6J4of!{3BLV;H?oEm+F^vrIObUa`*TwL3ek!i1j;OcN<~zuO
zPt_PGoTBVEd@UPELBm`7{=O-N_~%@|<q?}RRqv<+HzQ;w;b8re`-SIg>iA+r05%pX
z0{ItopnU))2@aWmtFA(M<nJ;*B$p2$1Cc+h$id8m&4a(PgWf@a?DSAeX8b(XUM6ce
zIWxwV2=v)4(<St$!WR!gGNgR=ySg%ZdtEoXvF+Qp?&W~<ZMO7k<Maph#JXE^b22K<
zQfYp(ttm$*eYd_}M-(w%N*(Dw;<p`0VA<GJ4A=EEXZ0+-zQYvC(UXt^#{%<L$uc$#
zR?}`bg-hojuVb=@6N%pqjWqYYTX_58y1IF!bn5uDr~r|kXEokHsY<TFH*GtsIICY@
zPE}wz-gJr*SFr#D#b!r(@q$nZDL0&p_NGpWRkBc%BUZ=CH)S(XlBz=uBKMXjIw;;*
z7GT(h#-!|LgEY-#Nq*8NRIxX0695<EI3md##7QDWO)B16!?ZT)=$|UhC%BKjfr7r<
zMI?4xxmVO;Pts+uttlsl3=Z-Hs3#kA`WjJ+_eW3SKq2WEmnmi5pjocW$x)7}1+QYZ
z3^^34x}MDlg0E5Us=Cp9^X<}KW)xt0*CPiqeWoAoyxPmN#Q&OUyi28)jQ$0oM@q$&
zQrVcEyVqt?Gj(Ybg2aeKj}e}&H{Z~Q-(C^bq?nG%>==1FSqBR<D>{yzS--E58CEIJ
z90a><*Ia>Lu0UY<5Ao%Ml*L7p7mg?Oldhr%Dvn8RsE=p;?9@%s0$n@)V#KK~nAKAp
zyZJe_1Kv)bthh`8A67|iOZuLlfVa+CDD(Atz?)h9Uu1{rF6YauVVaoUoec%F;~Lh7
zLqrI`$lxR%Dl^Imi$5yhAuPtXKm<}bE$;F5wyRB)!3Mh$jR04epD`PBIFUd=7wH|T
z{X3^xIv>{1Zqv2iQK5t;_~LE0NF4nsLbBqJnAe+<)9h9}9y0n;So^)tcV@9QK67DP
z<YFOij=IUG7zZ@X)o%d8i7X!5kbMiBF77%HiQh8GzcSbc&z!Gm>&LzRq`gGuWD_3q
zN~Vc|D+E1quc%psQr-yx@s4-)s3m{aW~p6l%N*z3pR-AsXaFQVT3w$Rfbk8CTwfFk
z&U#KQLJ5mgIZfE#<si+yS{~2LK4X5yyk!f5L-W~+qX(73cL{!3aGlEO8CTqN;RPXE
z<frTDEjd1;f9^P2zJ*~o0C~qUUM-jToU0rZ3>Y4(LeY>+t1dgVZj<i1C64m-;QKY?
z$9eDZF|qhu#!*{uH2r@RQs&<7iWxdtNO>8DL2{<~EK^QS+q>i0*?K0Z>j-(O;mgDN
zpj#`d{*>J1grbVK3!&XX%XeQZ64@JICPJ1~e8p8*I+zA&<oMpMLf(>A-=BZf?_eIJ
z=8jTTeaGh_YYQbb#<=>zScjYgL^Zs@Tved-aexk?uCC`-ZQy)L1(4)8gS>plsz#*v
z&^eBC|HjsAc#!ct>GGA*#C+cS(0WHMpM8_Gq`C2(zr|Z?4h!x>a3tuYv#VA@d$*r8
z=kbdtCUTOF!S<~VV#u{g@1fxjf5T*MxfbYwUFH<!NYp1r!l&(ZA4yJX#cCFa%~~Q7
zFjWeuspvy4O;0K%R^y3Z=5vn~l*9XqGZc(@$KDU&s28knv%upzcsZxoGriqljoP0)
z*~>BPzl-KP+`9JC=!>Edh^W=t7z|9PG096^_bjw)><eFPDds%caU}Y7Y5zm2y2xOa
zf)m}Tu9)c$>&m2nJm25ebaQ2@^dETeHVe0V|9{?)@&G`yBW7k39QzLGn}pRRd@^}#
zv1iEUE+E+g@R3TzN%<?ov-~-?p`Zgpgzqa~2Q!N(Qbv`%XW5^IQ$HRVEqEcX5+$D$
zMZZVJwz}9mstNI)sXmyUz?0ooxpJRlQ(SNXIu@7;zHuu4&E%gBrN+9kuP$BQ(?buZ
zC)@029HtlVykn^9uNOOXIcPY;1T=g3hhS3kbrwl{ck4W{A*w@pinw+$PB9V&mH@u7
z_4-w${q!tth|ylX-jZ^?XLO*Qn%z+%rJfbArSySDyH23O1>V)H!bPLsf2BfkM)@9o
z<kIs_?0vZBEy!W$mHy%jBQIPZ4YrK#m5T^5JB%74JM;F7X!7~IGN3M_-qSZCDA>lT
zN?I(WfKybQjI$gzRD0&+!IabAsRjj^L1BqedlVbF(6FPl_jjyEqJpZ8=lVgOGU`-^
zm>glz^(fPCy?B|p4n+l^qZALg2S#{F&i9M8Z~Ix1a0&^i=Fltvk^*e?sJ;)@hg9A@
zKYxY~W4?;!Yviv}BlmN1?$*ply5(u~oo6(vI5x-Dhq?m$FSF!23FssCB5Km6{?s9T
zCsOl|!W^04Goa#qNG0_nL*9wO8AEOnoOd~RZqj_v{7>nJW25ec2{u<s0+7pIa9mZ1
zl#D^Uapsi-+n+8plL`pF`b!SH7WU{PD2V%lJ)(#m+u$Yd$z3z|To#Q!Qeoi@J;FVg
zvxHg_l~|{rRsrL>Oom>4TSgc16c@hi*xB)WE!nw#M_EGmum|_k$x@c0Uk+nw)i2p+
zZv5CpH9x89-ivHR=q&Klg#kFP{U$gLZM(BRannz5cI@+fl#+U3suj4XNl%hevLk0S
z?W>tj>y!DN7$sY*XCReo++L}$mlNP^yMq{rkB7oE=m8#vIZ3jWK?Xqzqx!W4N>sut
z^3wlm*+by1XSJ_?pkQ;9Jx7EJG6y@NHgL`O@hESQQNXXwNJu8b)%8hZ>w?MK`v&WG
zx2^ZmBey^i%Dg{^Jj8#p$+X0EX32;ooi6EpKN6-MyhzK6u522{lq_i7POt;IUrrP9
z!)UMk3)Xqq?GZAwMPmP4*%r^$HYKs*dXn?f-20qlHZ9I2u;#XSLHImCeff*Q1|IjH
z6FF83*nc14)0o%on^y4y{XVaPpe4VyCq+hLF;TgQ$^(16i)1vHQAf8g*?!k!UN;=X
z2D^k}<ArcqY;3&nvn9|c<Bhxo7O%7<zE2%Jm4@N(wm0d#?*xO`@E&AG=^7Yc_9=|$
z&*~c=tcu2#aE3?j$izwAe3R)rZDH&D(Tp4{U}*@=pf5mW&!DY(UZruUu#(4c97GqH
z4W|k$7>#`bA9-;)L}2Lx^rd1SKp*-rD;joN^7~9b{7IFr<zE)BHr7(Ho+@Il4s*^^
zMRb{WEdZuu)SuLnQuLv#&(PZkgRB*GyH0dFV9il}kpk;#pNzZ(-U(&=cfqe6v_IW;
zZed_LZ#r83)8_kUJ;G6U6yg`}d=HY?o8Am-7ehQ8avB&PW+eN%oedkGpA=`#*Dj$;
zi(KSDGCO0jUiuLgRj`l=QR2lQ6Cv6MCk2mxl<YSm(jz%~DwHH}ONHMGWBGC98K=@h
z+hE!&ZoamD6%yTPdF%%;RYxA>euJvkf}LvC5xP4%vNVVpiNs@jZS+g1Jh9CNr(x@k
z9-5gja<jS1S|iY7w}9vcQwQ4N<=_sxZl*S@pux^7?|$8Zp9*k(Xm8<zj+Wh(J$}v#
zc!b0!slW81jnzMqJD><sm{lt}1}mkusdisv|DjlVwy4qBy1C!4Qoly@?rX;`l_VtF
zHX3q7P?`R@o*csj;?JN?Mz>iY9WI)c07+2eK3=1JG8FWZ*FAc%L4-iebo(9SS}Ttx
z&G%Q#2x-^dEM2L(nJ*jXz7-^pfa1O1H#k`h1brbHvVBYVW67<uqjH$z^9EF0e92EF
zSE6z~i9x7z-lsTDlDRqv<2OtNz|G?~?`nk#th|K4@pqz|DNl?FQrANRQi!L&e(TZO
zTgm4;Dy<<evw!Uff~M6Ei{8ZF3cz+H&R3VT0g`7^x0MHEzGqt(dz*tgXDs5R$*m&b
zvV)L=(QGk8@e0DM-q(N2$XZGy0~(E6AFhjDj4g%fJB^$%$LWmHNFF>tzG#js9yMvd
z>`Ll-gxAevERgJG0pYrgwy1~7-S-=rD`?*;9C*YfBX*x%aSNw$Eb$?&YJT73)6J5>
zwsWeF^)Ab(r#mABZ`_b}PIe-H>Dz;RwLj@M%&it=$Kpc{0nW5q6#2HO6o?U8G(1iZ
z{aiS)z@2kso63!2Z*n_A-F*Kd1B9bqb;yMhQR^$UX$xW*9P<hkGlAelXS(+&`s>{@
zQxl*Tq~Y8#yK;S7ipG|F2Zxs79JqG#VRlB!AuyyRLsHprCdgt4OJ3Vxk-&iNMjXw=
zPY2x_(NHK+Q6U(?V)4n)0B{D?sbQ$$x<)5<*yp5`@Q-6x&giHPs6MNxR5_Y9LOL5m
zIkfJeB>jkn3%|m~!e9PrG7BDrQUwY;C#SAc41=rJ;KcUs3b#+=1We@t)sz?89%ekk
zCFhDR>aS<bgBQ?h$h%l$qU=lG*~eh+yhe7~F+C9A<q~YR4!*5SEg1)t&~5d<^)8*X
zupzWLHC{Lb2Ig3M+^?ObPQwVZt1{q<68r^(kZd7=&t`o~6(6B${w*gt!n<?HJ+x7O
zUud13o@VN4JI@LVw#g{9hc_16A@8rWS+qOPdoC@DPZNvceC67MKf*^4;<I9|o+<}l
z)Ae+!$res$P{_d54h9^C4;t%5W!p%B$}zBC=v`$PdR*ld$+-g7o9gn&#o=<Jhq0|^
zDd<seDU{BXHYy#|0!I5D=Y)|XVS|-ZU`EQvI}QAEF-FPcZZ4+Foh8cyNHUkl&3Bx)
zWWDUJI@k@E<Q!!WFv%3mOmjOs^b;`uLpkxK>{dkSSR&Ghv(ofv)AHTdn$Jg$Sh{T*
z7Mo+VqqA1xU-+tO09YOUU)J5s&{5q>HJZfMu7mSEtF$1|vSbp;Zgi|f9+_#0Hc}mO
zE`1GHq$IOkSUYo1I4gf45^YT5HYi+ZoN~F|9~Tbt9|_7tzWE$Wg^!rnsyQLiVc22f
zr(WLg;VqQh8^HYK{zZAt_=vD0*NG0*jCs@Zlg<VvwNqyJ2E$gM(ZMEN;P+4PPE-3+
zru2vIs+$)PN5_>{C23mC88K3`E^AS~QDnBL{CSPC&6S=ov0lMti*c0c{Oxs}9Y7qi
zv{Hw_BmO|B`GFEf<zSy%O9u47hE`+o)&~2qhtmg7Bu^mcFh4a^1-JRNsa@Q9aNXF_
zp!;xts&_%6U}$_92tEUQxcN2v-tMgJg$ak99&g@SU%+5jJ~JA5KtrQaWAVU`=%|jz
zB|RoT;1Qj}xe+?W%yBdg-OS-%YKM$iGo+9)fktbElPG0)bsOY(A02{`q$x!qH*kG6
z4F7Jl|Ab+wzO<O*9?2MytD1mQTQ+#3EJo_j$MU*$8~2|a=^}SJo|PEQ$rN1m0Z4x~
zGJ7-|pNq({E7x8bI$jcqbh{mbchr494%Q!0&QHk;KLwp1yjs@owdnrmL#o_0c&r(F
zsg%JyWJ@-E+1pGjMpt)_Oz2vnAeKjr2TsgM1JT_`3}1L;rTPZ?cC&(z{_yc)9mEzh
z-hvv~LyIpiLncun<P6CE929ydU1ELunOo>q^my^Dv|dW;<ORbr3BYcHsV|D;O%Q<!
z=P!=2D=@#TOIjatbR!DU&Vy*nC;e^s_p|wT7rb{w6*q(<PY>C6Kf=SZwo^Z+&d<4D
zP>W#|Q%kLL-n@2+Q7*MnfoL7awJ^~%h3OOFhzDsIkBvg(f-nuI@VtAw_c-bm+GJ()
zz(+MQfO_dMr88-F{@V!e6~ry2@~dpQP<A`y5P)Yc&ly%J#xIq_aLt@1OQhKj5CXU9
z7Z4MeAGs{7PaVxF$(n;?i}}|s(Fbxv34<ZWgyqIs?I|6w>#YZ7(Njc2oQ$6tw5<}o
zkW#?o{u+|bhubL~o@hlLoO=aoG5Q+9K{$opP9Mezo4Da;<>@(o9bq_sm~G%T6(F>C
z#9Qj1b(mrsY$OEsU$t>;b&#47BsTgMTc?+XmAaa^?S6U9A@Qn((@&tixT3gep?9k}
zp(y&0TbZJkW&A?SbGI(;-lAefSCnmj*}{Voy}cZ48>0;CIQ#*8bgw>sV)=55-~SLT
z_uk@5w-?&i>-yx%{p}9HW#IYt2nPt7XY2+zqWPI;V)ee4=_4b%s@Bo?#&kSIAlyy9
zlbTa|b#fbn5dUhKhW#l!fM}J}Um;GE3yn^4Pes)Il_`jnp8L#l1!Fxc?7ISY5@3j=
zmX{^5dtm<Mx-b8E@MmwW;t3A_omF?#s`x5+Z`o?mL!VvVn_1zthNojIt5&Vf`ARsB
zRr6vf1#zNa)3l@e6F<$8(<EIHoix6mrI{pMJ{aDhUx%a?sluJ8GQw<((JKYFu7<Bm
zOikm_>OeGEIP0p=(;{?4N5b$uOnsH2Z6Vaw?D>rJGh2pxp(ig&qJP(e*8|Aa?_4G(
zw3I`aK>-2HO-=eHle==dx^(cb&X>c;4II%Y!);pfmu^T<1hYN^mkf8d*Iul@HYmv8
z{rGy8LT`1m^L7-OlVZ^b4-fEr@VI!GjF^URb7~V>FQ3%JT75&Vyg-I_1Qz@Ky<Pbe
z@@-e3hB}peo3lLY4v&X9Tyy*C3aB|F?j|O7iaHkiR74`M^)gdR_?|m|Ui_@eDt(a&
zlAXV_9B|v>`}gGRz}kDgxs*naeL5EhOC(#u<c@;|pz33KH8W@vr$cza(qL`6J(^DV
zYtq`Tk&|q08Lb&~Pn)bp@96ox(_MmPCq2?Bcbq?DLFp-WeNIi)p9#p2GHD5dC9e@y
zC+mCcWi`q)=sy@>oc$@?fgTmebW}raoWVB4CEZffRO<J^rB=Emd2;1f>GaE&VVE_h
z!O8h^kwCg*bj&vft$v;1soaj12F6sM`9#wWjM#aGN6fxrTul>t*saVMe?zFCpVWb6
z-P45zvKeSB7+OPgi6*}YD%ooK1O~<2Cj`7}=jqLQbW#Aj_YF;b$ZoUH=X3Ub)bE}x
zPFUPp34GeodogLPciy^Mco^QfaGO!-nHC~H!V4_;vU2sYK}8vbHx}Dutp3^6f^bM;
zxbVjzyL~G^Qw&0$&BF@rZ7l0kVW}T-!so9jco|!EMeRRJS)q1QpnDxtNMAR0M}|HS
ze-FlA^N$oOTUXInr4p7*!X{Sya8=~A^n_#{V!&N3wwg;wun3A|T$NSxUlj1X$r1)*
ze8r*cTwv9DVj*?)(Wql{EQON%nVE`WDsQNV9vkb|&F@cyQOh%hlMjT&>=>dSaF7DN
z&Z&WTB*%G>EQuEhW!8!TmWTBBtmiCwV02tDw{wc8z2~ak95pD;*eTnRJ+*V0UaFS&
zX&)Y*KWJK!J9H1jyBOj5l{=oDcVaK+aRbr=<j84m_%1lSlkKnr$usG0zIl&*2;dt`
z0jn*ob1|}+DNk8(E_grixl5OB7t=%`MsMQ`{QewZuiM$lNYZ?0)`hnDrFOHj(+Im8
z&>V-B)lufvnE2K6sPWm5xSzKWu008~A8K|)PKHR%H{EmA>A`ioYA?EUk_|lc-n-o|
z%?di}l!Nc>G(PcDOX3I?!z0KyOy-@ktc7NrulX)^nr%`^n=`+ekwOWZL15pLk~d3X
zaO?uY6{7)k&Y7Gg?Un}c%~}~RA@g`PI50I4IoKMk$6oOos;1nx5Fr*oI<A)t`;>;I
zXb?X~wZ}c@=G@B?bEqptdiqi={tzeNH>L8Yotu$iq)GgRQ_UlYT}W6rQ!T*ce)Mim
zR>_ZBf5tqgv2@Xkwn?85&Ob5LEO=}$zEkb{P184)&U`^mh6ug)R~~&wb0lbr|71Ie
z?COkKVs*V5g?WV-O-xZ1<2vC|_v<U#S+jk3dyrc}(D+Nl8(l`9KKedZja|!{m*waU
zv?$>mjR(d~<f)=dLamG@W72`CQmZsFGlPkl@g@+Uuo~L%^?U08KgVF%0c~$1U^&GG
z?`)qSnClxx{6X?X!k-QU+W{5hFmt<pXI9SdLx_#jQ1Ol;N9>iK#hpT>;d_fBNyIj>
zxZc>6N`FS^1?|>0YrfoFp+RA(#M%1+%)~M)@uL{8*JRngZ*w}EF#lNi_Co01JO5&_
z<T=NN{OCfP)z5O>Cb$jV2fk*C|56Cbvc&{M6R|o4k(jm(aqaVG{X_0XMZKukeT57L
zvox>&^|{r>CA`4;^6iqE*Zrcw!};m%LfJRul+?l8%}LK>tu#Szp!3D4Q;4sf+l$hp
zb&;Z9Yy)h+(JZJXZ^{Giu#AyL+@g6wV#g3pp9eCQ%Y*vS#?(vkf{41O4$n)zWW3j{
zYW1?)7QUma#Upb}(-MH;_%z!JtT(?FRY4BL=r5{^$+sEJ+sEL;y8ih!4E|uYub@H4
zdZ#$OguyjZi<g7jU_`3X00+QxfM{n$w?StmN~Iz~RDz9^iU=e&wv85t9u%fP*RP7&
z2u>zNmO+bdzoYN_1n9v9Sm-7Qj2}4O0MUDIdLNe1O0sV3)@L=Zqx@1k!)mU=mpb3M
zUaP#%YkZCK5Koo)ylc&Z&ywPGVFu4nUW)+@dc%)QG#vqgoY*cuW1sl!579bx{m{`>
zuUWJU<#<al*Yx>w5pZy4AdyWPo2)h=i`g)jR0P|mpgx71XG36&0BHI|kz=n<ul~t9
z`a%g7856{Hr-1K@v`q6^sNl<R)nOmpu!l8ub}g$X-DHK?VUyf*Q*}*t&0pUIHVCxb
z<tRBk{o`o!`oICEX_;HDn*fgh58u^jzs_*lf;tNpGhZEL>GMbQHiqnG3a1|i+&1`L
z-bV;4j>=nfoOCO#7{A%wJ%$YGHyeLSSM2=qTCzQ3@PzcVLx<<2SDwb>S9)I!XBsO6
zNYk~gMzeGyFxCRe_AAZ~^!D6Nmst5Ib+*5MWj*K{v|v@7FB7*U;SaYN@``Nf&I#Zi
zdP~i9L4nTGH0^`#D37ovl@Jy~3}Ig}TI27Ob+Z37Dx;wxI3mn@Qi*yWUpN-sejx%{
z5O>8NU;1#P3snQhgx=!R1{=x`ESv@l*U5cT!T!+r$H#XT?+Mg;bWz}CtKVY)gGJ-$
zQc8V5UGIP-7Wmgkw*|kLmWc4o2||x~-J2LA3OaWc3QRBCCBaR}etBeukC1hm`K-!*
zNyk%zAR9gv;qYKWj2!USJjz?r+Z4X%UQq;j-DNlbMo9krS>d?((JqQt)iT#le1=@F
zc_Kb1vF-9gJ!*AR+dRV2NB%JA2~90&h;RIv)b$}zmrf60@=sW36PA{8u?6_Ge}n+Y
zt0Blbhc&yfe;yMf3Ar?UU;|as%V|utPp84H94q-a`0R$zd(71zXE7SS=k4NCF^5ad
zgv>&UP_HuRps4E98lh_bFa{Xh*I+Ci+h=-$f0Pk1_TWS0Jct3v-2B*y<!k+0&BcjI
zc*$}bv>>cB*JWs%{{e?p*I6=(6OC37F}*3s*8iTEB`Q-lNWaLzj+xDx`Bz<$jj1P=
zja0l7i-)K}5qm&S0zgv1?sphorYM-`;-PsZuKVOn^3Oxy^~jZ-kW<W&{7#-txxT^N
z-?m%dfa+V@m7N>=ZJfke{5V8pwE|(8Si0(=*`PwB(><#0Uoieih!^`xWn<M6YaETC
zNG2Xk`uAXC3M-c9D)`3Ub>7PwWLgszp`-Kqd=J-ssJ-2W{QXYS{&;vdZIH$IRF0ro
zuMqeU#%XgMxl|1$ob|koWV&@xwWXQMbaMoXQ&APB>x^bEOH;Ys(_G*_6{+H!p}FSJ
zOr>^}_s1vxQ^TI|2|wwSoQ@AfGhSzFVRRdOCNs$GyWV4ZsB#<rtR+hvZ0VVY_zenf
zf|Kj^#ZtS?+cA%3iMd15R_#dV{!WtLT@oK(mb15bpElgq>Q@6|ynqrPfMLHouf`Vm
zJ%DssoCi+`_Y6@LsE!FsVyBKTTZs;w5r0|((Csw~s6;fdkFB?7GetTCE;B6r+Ac!H
ze3=+WbJi;%*@mjqE=C0&tX7UUQs*kN4!AFC|2OdYXElZI`=OuemCYAVRi!G;=1{ns
zkIt#~!QS6q%>!r}jB<9gw7KS7#w@)>!gd8irR*`%p!Q{tw8u5>-Ancz<z$;jhO|57
z6;Xjx@Rr~4vlVeikGV&Y^Fkw~#*~*kde<gRehqLE7BF~jBY31d(G56;G|kLkSD_w1
zf%p(>0cf<z-fOJuwe#kdrO7N_`B+_8NOT9Q-rJ77r63OJHZv2aKF9e}N;i1s_e62%
z9d^3CV!pplKSxQhl#VLJ=_U%=j0+nfcFdrSYez?}?BM6bT;|R>@lE2s>b8vBF?=<|
zxHEtjRfGX!NJj4WN$1lxVdbBh*=a8_5}Wb1gelWt6`t>{T+30XT~U-ecIAy677(6O
zqIjSVo@rSELqaCneTt#_sCu83s14ShVel)UNk&4Jg~#|AuctC!idygr67<?vpFc+$
zQv)6Vq+$OuDuzH)RFdy*>=g<nzL^!g--l%XJqG@*KWf2`K#SfIlK2KmA-`NoE4X|}
zt6VE)=li=ke20(73Sihi!NnfPUz=IpX>Y&V87X2?g`HRm<5QnW{=J+TqCyYdu8WLg
z$bXG;weNg%xWjp^LM@4p|JQUy`o-OwxeWv9gqLxFHO|<z!tlAuv&~x(>`eZL9n=*u
z-vjiuVVpoUh0%aEbjTUJ9Ncn;U_t%m1aj)AE*FQJjyvVqY#I7Ess3n_62&y;Jbh6Y
z!2H3eO}`a<)k&=g*i01#xl&Bv@7O`7upx!OEXmzkst|R{flN7-Z;SAqU^4U>6?BE`
z#t>d{^7M5CTA2{o(`@=-;vjua?aNQ$@w7MaF`-$(%iylT;q3m#&82B6ug@8KI?y&k
znD=n0&hpIZQmA7~+)G07>$VeJzwJ69G9x$P?ZXJyg*)ZZCFCfFV66rR`s3KD+qdob
zf9U|ibA>x|&ZxcTNE}Tq`7n0(ddZ_5iO#tM8u+(W+r-TD4lzuFvw8v!Yhkk5<B&Me
zCtZ#~Dy*G`wkcvrobHC29e_zbPVAQ3>z8H4SUAV#4brrPn-s>z3O&rB9Ek`D5%6tf
zlUJK$JJr&jtaiO)P!zy8u7NL411V2wGoSlWN~I9{(9)*&JV`YYzdBIp@NZ}xcosui
z=y7S#zTv)CBifUpvqOHYVcc(d>J9ngA^NikW#*z%X~$Xp2^Ta^nXZV;8xq$V!ol?`
zx(OmfCOPty8a~=Domy0V#KcP<thwyvl)=zCre(`8b<2T{NRvuuT@z#PXf723t2Q4x
ze_W|}$Y^_iI-EX>RD0uAsU-jI2i~vbM(~VN`d3AVFA=EG%yXa`gpPFAKE(jd6l&(i
z^td5P)-HxuYbk?f>+F+_U>t;Gz0W~3YqIuO0M!H49CR!4-^qyZKJ@6Vz}qDN50f8_
z?g7|Vlp54m0sBcA&x*urgP)e@<A4hU&keYov8b4a*O3bgvJs++F);53_jx?9kEz!L
zd38@(Q1u#NdDe>0&G)-<b^$7xW<oO9U$m9p{p8Dw%Wj+UVg>7Mfrz!EIyda?Zr_IO
zW|Knl3^zy4Nx{12t*~0cfeNEUFV#2!Inlo-J*0y);YSP9a?L-MXh~gogCd4H!jz>J
z10?!-1ca`PUn<1zwJQ;}*}zB5b2Fu)zk7PlgKm#oDkcg10iO;$?N4<F1RFoYnfFf9
z$^%I6@{REaa!o#_B;9G&2KACS2HY0y`75?o2r@+1xj^YQ7x<63AI8e8^j<wY`f5hY
z*7Nr2UpM|6R(#zEi9Hl<|9e4WN#-}s6CojMbeJoF)aY7^y-oT%=_hS-mW^#@{+1Xh
zK!}GO_)<xpt&TBn?B3Hfg>Lzg=pPcHjBm{fJ6^k}4VBS&`Jz}?sOZ`2HvD#?*_EdB
z{9-;5XW~Fhs_`m;AZa_tvu>(!To60MJ@;8w(<rq8mm2_`DB!T2AvMAyiz96um-4M_
z5YzRC138bzh1PxJlV5h#=x^`kCaHa!P^bEY2_qxj-w!x`G0kJfZ7$0KddOGlDHwmO
zsGMYmz;cbxvosGtqHT=)UPD&%8wCrdzz1C?;p_2VICmcz8u4ESc>gy8x`xKJ5c&VI
z0QSvNN$zI@ZwL40M@MxyIV`#dutoO}yWIZB*he5)mYyOL(=o~YZe9RY{u-Z|Mpn;F
z9m>HJQYA`CgT#<Jdh*@xdU;~XF%PWYa&Gb2VL{sLMFj|iGJryKhxn@pVOgSJ*+c}a
zy~Wo$<Y*yUffBt**z-P;kXse?<08IW02~{-6vuf;b0$yy_x9E?5%KZr;IcD#96zw`
zG8wonZ(vtfPB$;q`xJY1iAV+nP!j5+C|R4b72ymkKov!7Ra5Ww%C7Iiv5TQhe!N^l
z4;LO+PMi?t`^G$)NtQY4RH1J;0M@zMYUM<%EN;k8hGS(hL2+;T^ff`Z3{PS7=B(s=
zi&%#9cveEAktu*Ra5q2{eEsrF{O+jzcp>7GO-HV?`HB|M#<E({wkzpwvPw$)q#f2C
zJI%Ke#YJ{yxQg^oaL<2KmULJ49cJDCA6f1;;ErGOFJ%E=gpAV50rl3(x^dWN&#93n
z%lDBStI3g%{@WIx-McfSsk;wl-EjDGI~`9+Vvw*M^-Nl31=i{F!;4Ovc4e31X<x;p
zqAL0h#cRT;xHJB)lIOJT*n{K&o#DJ)8yz>OZG%)wUB=co$&m!6S1FxN448+{WmCl_
zj8mL`@r0^{trHJ92=|SKT)~dH#R?LS<)FCtanYjYQiL5D$_xk$vprXl(RW{eXX8}V
zHagBt`%wBDDfl#5Z;E;ojk*t96w;EY&C`E<eUNpn#0G&627XuetB6&T0Ll{VXs-B<
zvX*%#%T`KN1uZz*F>J#ak@0EsCuh4zGxrI;!WU{cyqjx;@5KiAW4w>sgNMh_WP~2X
zH+_<+RG}3a$uQ%oTp>e8$R@J#Ab}sfDipcroPc9P1tpBWVcXkYE0c+^n8;cYZu*Sb
zE!pmqnp>6!1K_(4IYNL$Fo=SMF8CuPyP%FV>}ZUyJHGYXvBeQAgTRp>vkZPQ*E#JG
zqnFByN4Wy75y${AEuKyE&`s7+MYv{2)KrSNBZKM=`8;uJ!^$#0bbRVorNOGg<s6=$
z@c3y)<v;84*+}%uDk;IJ4NWwsV_V}SUV$3yD)ILGd8-lB^?tE^6(-AHzlHXn^ZdVT
zD(&IMpS}8~G2+EOrr3$rtzJRN2MkqGkLoACYGo_?{Bm>2V?7dZ-C@seEvuue*IE2u
z)TiI`rZCF=Z#UjKn3h^I>XQHbt{Pn#T*vX5ZqwUk@Qh2k2_;0Tm*e&$6|ow0;*ukx
zl|ES+85o?4wf3=+M#!sdE?@~;kMi}c_7-=ty`9tZj^Gjid(iClr_W*=j6oqX9fa?t
zQe;P|sgnGRE#IiM?4J_#+gVjC^h|`hy-<9@j+dn4o+!~(8u~>AMcLc9H=s)|F>z!<
zjzAP#of~OUFuatz+!rw<2piXm7%Ciz_oFa4ko;adSN$n)3`=tL(5(dU`_(+#+oh%-
ztW`Bf0=Um2S_?D<w4U^mdRCdy+ga`jtTV=9vE1-(n~QwfBsB;kxM&HbmZZ~YqzS?F
zj@z|?J3jPt?^GyJ`dm)rwkio=7YvKSkv14?!Q(EBj+5%!>$Nn9FE2BW=!=O}`OBY^
zxBEboy}Hjw9BB&U0uW%h>1ejZDK!DUC51ps<q{Q;YNSwrnP7#wD{AsNxaw#T6k|Z+
zy^G`l7CeE}x$uI)%G1qnM>88MIs@#yYA+(tAk32XL$63}CT|5Z|BtD!3XAf4qXj`i
z8UzLyT0puR0YwQ(0SV~_VTSGw32EtWknZjpnxPwq?(P`yjKA|g&vP#3;=7oeZ-0Bg
z?_O)YYian*=q#W6_;N3Hq(*@JJ@UoV{0)qrQoU>Cwn4a`hTEP&^Kodt0jEn}W@dM~
zH-*<*6+xYI*3$*@=$7@~p2gNOldGqu?{jd^ilb!eL<|E0)!65oM`Gu8Fe;c&nHUD3
z1;@V+un=hq`wIdH2A!Rn3RL&tM(wFBj$zUx)hTKLFoAMoNjMH`SyGPwr`{^oVl~7u
zso)Z9Z{AA%Ak#8xb;|4jd}V<+g5Hva+^L2{NVz({p821>FYUXF3%xc+hlfYm#zmZV
z3?~`Tw!WJ&sli!I_1O%EmiJ6pN`YF2S6iM>%C=qRX<J#GB+Ej<>XK!2EJ(%WP@=_U
z!6Z(H$j|4;F0;YJeB-yS?R#TjIPakAfjZ8_Qf@)Y?VxbaJ7l*Pjc*P!#9}cy4OzrS
z7@OFlY<>(TthKR~>(B_&jE`_c$YGBLUK(Zm=!q|CskCC!9r&F=f={d`-0}kT8$K2&
zo8BTZNiEm^OYZv)%ci$-`b_lQK<&DxLXuJB`4~c^FN5pcLo@V3)YwMuC4!%a?Ct}a
zcVpBuuVTHk^YGn&Q{z@g1Z{>FU9^icG!Xc!Pw=KOkWW;2aO>#2lKD#G-ROBRaUzD+
z5Oi55Igd#^*=I#tS6)_~S(GxaVf{ybck9mG^Riy#qO5#MVnx3)Q@_ktybvd<&ES|~
z$9GVvlstF!lSLd{XL;BCOY?WZ?!935$D8Z(qdTkRNZQ%w0~wOK?anA&l5On?a4-lc
z!C&rqgN21=!|m2x9ETnyZS4~nGIS>Kw5^ESkSI<(5RE2oJ$$|5bj%i}$8dd{;2~?H
z*N3Ex8@t1i$#jFz0kFfI50*hw+kE#9-p5LDg<Z$3a~h=_yP8S#`L?b&*Hb%)K~wc`
z$01jYC%0P)S?@E@05RlC=&I<dDb^kL8Ux>9_GMGIXlu?#kgvTx{0KVNy1{2Y8ID}L
zQ`GvaXhJXX<;ADyg4w_6Si17n9^li0!8$x)$d%M%j78XgVw!*Big^?x^aCAyJpZ6k
z?i-e>A}d5S6ZW34vpVh3({4Xq>|zvh*rU;EvcO45wliw|n9TW=EHNZca-hDN2T$Dn
ziql{7?;6%Xs3-|d;@bBCRfh&Z?#{}|Z!RHWa7=;7L1fx%7BV8=8fTz(1gN)L<x%4U
zqozD5(+l(CJ<R(dtdtc-IrOHHNqQM#k@8JhQVhU$@N*_6?SK@1r<}LwO}6td3r)2?
z&{yu?>nKQQ`S1<x_Y!A%T=U$MS4UcxVYV0y9%%IXenf`*r({~$fy8d_9J3b6yN$0T
zUqzzETTVs=MMsn|Fk4!Xiu(9*gSl`r(ej;fx|7@lNvgh!m$=g+$h~AGGNNP1&mRW@
zB*{}~?u{97n=<c;u7Bx-=!tlw_B_o?bZS41L0tu#_|}6@U@4o2tq5@Rd}9`WPk%%-
z+T(CpxJ%GERmRL&_~(v2{AxL0$mh!KdO2*W;j}SEV(V_*jDP0`ap3sa_6M|{^r$S3
z0fnTzy*zK3UB!{cl9nYqQZWz1Qh@>GB$Gm{j>F~`nY%H_mQ*ckj>$Pr_V6;D-?nru
zxbEbxt@<nACC6c;13tEPAp6n*T;;kaD3xo+n|JQl^4^?zX*Im@N0u6z0UNKsfOg$p
zZQ5}5ytK7aN@G3s-I$LhZmJ$1cvL8<dU*b2&7jw!rNXLzo`vRlM||FHvdORJ^0pcs
z(u|x7j<#Q&VptLG0&kueg*@<v@AIC&@`VcK(BBDoB@o~xU?tAil1wECTCDuT6X4<w
zs&A#KR)pI;Wm}1Q2xF-a;Rh72>1z?KQvsjsx+F&1Fq!GuMBH;9sw=#URM77!8&^bW
z#IK$6vk)?+CbZF%jv&RyB&jb;VQ8D@W9()VP#qwaZEMxt;^okXy1kT>aPsAa(2Kep
zWTTqwf>pKGP3*`D?{6>=`Cl)@;l-lz@B7TKUJ~WKqNZ4kH0=3odPgm`zV4vKmU<R5
z8uLLK$ek&d!%H8vrY%h|Kzc|}VlWxz5Lh2{FmW9EjxNm_pLgju8jknoz`RiU!ulUq
zxhBHng%@Jq=h)_Y7CMri{kt3;hI7U=<Y!zJp2ip?ycY>6B~C%RIO~hSt+p*kIh10Z
zzE8$!+$WVjkDLhKn=?U)N6-7GLp+2>m`_Jb1O0n<GzWblXN>6MH(8}ybWfHrERU&+
z_UuBG-|jGm<&H~OiegsUyZTKFIR_UJA8?ufh5uv2*;edVMUw$M&OYCV;}hJpicgLm
ztv$Dz0(j>yEOOFFg*m@JDC<uGspoW&b{hc})tduXGV_X3^NmI=F0G;9c_rEe!Oe!v
z(eKH>MGqqWI=#?Oa&fnmiLItDA}d5l6dvFG&R&{jgz2Pt%}+6Mo*3M@=BM%~BK#eM
z)vMp9(eFOYb$m6GspdyHe%_@%@hP>8*hJ4fgc4obLDO4qp9b3*p&fb0sa35lu-T}r
zWUe_Bo&vX;<h8TIj_UWy1(uQh=v3rA7E>m(YCqb_h84mnJey0hY!Y!+q3`Uhhlkb%
zpf8ikAZGy5Pa#KI!vi{g4T8I<#fvu(`MRhaA39I$dE_Zdtfvn7=6WaXl<R=O7>%@{
zRddG+aD}aAk|ya_-v+Y!)%TNliu?1fOp_OHN2Ec{sq$$_o_{m-)gzd2$sC$*m_v@{
z`qj2WepSj>aNh%)KT*qMt?Is}kSfA+33buWEP6!jHf&+{XzWP$y(IsF-XoD=`?wqW
z<mRL4MngvZ<pnLMGzB7^h*Cmd>2FWh!VB+v5c%z|mt`)k6IlJ;LgrP>;qb1ZVq;<P
zzsMD@lq4XlxLCAUSJtDg=XyeT1OCa|Oir$J0I{ki@6D#0Dk1H2Wj-RKq;Q8-UQEO0
zepimz(?MERE)LsawUm)ms?S@4fs#XUfea7(+1j@-3|mR*(S1a~mDh_zr->g_w5x>_
zK*=EWP{Oaj_k-K~g!5>A%_V+cRBgBvBva%Rw*1ieahZ;jo<v7J26x%ai??|q<nJDR
zhGdVIN`q=rx&?`k!pjsvQ8BXw#Wjj5e@*6<Gyux-Crey%y}G32PId3rEjk0d(51oG
ztuB*UALNBvvzeC|^c#`94-sF!?EMgsJz59Ien@@=)5g=jPLg~c@{#=;?%mnOn(ndO
zOJxg#ceJ?O*o;linninWyNdWgu+F@%quFIjo!M(_5AKXCCgWuAMrrNLu%_D7H~U52
zP@_qp4?trmQ@8bCrRz3R5}nP5hWO9n2tiwPIh8TnBEo7T19{TcR$@ueYp!#s7wa_4
zdv|xTw{$jP_Ar@r?KR^3%*!oRJn285wtqv{VCf8Zn!yU3L{JuIRT40r9$Zl8s78k1
zCWoiMx{G^f*yg~lumHf8<G@M<Hg{5&&AK?-<Ic^Fa0Oa?ecR6WI2;5cHtFQEIC?>r
z0E_?84Ry6X)e9%f3ng8Pt3d*pMPdGm;BnOnrgHSW#@m>VNCS0fcjP}JY#V;2>$P~U
zl6TWCE3ON_EVA{tf7Pzm7aQ4fP^V~NcNO(@1<iHXy+$V~iiO%_o&ym`4=ccRyoH4{
zB;2_!vOzUFM^!)1yRn6k9-Ffzi-#x@6`gieD)p7(`TkZwe2G8-ZkW<FVXRu9>qX1v
zz?*6X-+8-gopeqgV0wR7h_^|1r^}x=hQ0Z~rly&x!ToIDj>@{-3%<XFD!HIT66q2L
z4LhRoaGJ<5meld-wNJp0SQ81CgJz%I2a_cUfgTJk%r5t8g?CMfWE_0UdxQ3jVmluO
zr~319{bktR)L0iN4%2^l@e*^?I{*Z)657feJ>Z+g<oQ%3^pj_?=jQoYu1ptQStjK`
zw`EYt-qvf%_9<Uf?04YGInlMlt#?wN+{LwwiDB#8vG;l5E}G&yrw2(bc00ooK^<v>
ze;7#bJ-XTy!5_+A7KS6N{vg_p(`A`OkqkmgUCbUAV4adpi>PmgPlD~v89<qgcb#4n
zau1E11{xCLUJKejzXJ|u%a6rc7E<3q=R_YN&`VN@vPCoN9v#9xCpcHOA-B^s(pJ{d
zGQ-$J{a{DfCOT7pEfXGMS_slF8NQ0&7P?5=!^w4lq+k!<A_=?7L&wuVKxEfl;0Cof
zE?aQJI}k3)(x~mJc-h?_HC{U}fZLeI3*0|Ie+XwYAi4IihzubQT%~iWK(L-ix*pdX
zVWO*Z#fyG(Xa|CVH!$>@hV7{(IbLA+CVDf!`zU%*o$`m}uJD=$ueDF%PMFI&h@>@7
zicFp#2uIclk5Y**UY71VYn`T^&>REONQP^vwv>y4Dvp%f*sKwr|0}I^qjsSFB?IQ&
z`p>(C)>)hIRPDfGVCm}<k34h8H$zI+OW@f*sg_Zu8*C@+1fyC>4eLS3N2}Hle9$0I
zSwfPlMP!LKTmnC$t+-I5%BM?ND}@7LJgcmeTZFgI-hgz;x!;>IR>T%@bbI&%(G!G`
zeFOGt^pdtf8Po;-k@!<27o(*-+gJYEcmDdqb~GPJTu{XEW~*)WsTek*IZ)HzHtP=K
zmA)J&H$n{|_(P0PJNV}12tb!rX!s-F)Ww#L5FWE_WDuFgj@f;vHngxx`YO_botR?0
zM_3j7ePH~Jwzcq)$HfCL&@E&0k)y0G-MVCd8)JqpZ0QDQl566F$e=rNuUE2iydsuI
z_&O@v9UIoq>~9YzT-lJg{XMyzt=*LvmA1wGkP(%oeA6BW)9G;KN!(W-P@Zqp7h*Xa
zdsMcj&X?G=g-4X*0Xco$u2&}Js9p3wcP_;mJ1I!5Z%;FIV3lIdUuV#`J=?(YtVF^A
zYpO|Eosj#eiNlgx#2inf0LnCX6g3)!7}HVRFiM~fP=s1yQD$)J>;}+z9``qOa+i)g
z-D3C#r|qA`-L>LJ-V282oPR!Sd^0Hl2p_h4M<0yK^x=GlM!l?d$0qN>%44l2R#6gg
zQVwF$b1vhbWQ*hf8&p=dPB+eKP<}lAnS{WhOpl`eG-NjP!xAzTTmDn#g}x6zVSC;k
z$1#W>Aw9=I9zf-?rZs3EnW~S2+?}PVvZ{Gc6rCGf7W~W%D&e8)){UetVF8qg;sRlN
zAHp4g)UvzAS|4yJ?^@LQwp=^NsD<0Jaj=CZ(BI%~k$o%(gt_Z?N#E-M(%+yKq3hX%
z1_q05$!#OamWjA15~AQ|n<D>GnmFZ>Y`J{_-FmyH`)5Mp59^0=P>s^~HZRQ(^859e
zjw%63oltxM`=>a9&|F5s&`%}S7l>Kw!uRberMGN_`WQc-3Uz<J<iXl{l1bG2CG*{%
z+41!s5Mq^&maYssU=>t<mXF22^CC7<S+|f7SKFw8a|FGz6m|Ge6*(RyI(F^|%^(_E
zrFBXNjfZ&)`S^I9AB`)$X1Sh+jcRXnYBgK4MDcGf3yqyUtPb#7$msuR-h1A^JZNX1
zu%#AKJvwPL%9i(5JC$1B&kJGgAfCad7GHKv6JL1(tIfOq_yahS=^c_nH0r0ihav&d
z$a?@pDbia$D67T!_PE(97lM8N#WrF!#kGzdt?|DAyNPppY_A#uz5@C>vV{T8t9Va%
zwM~-sRAN#RYddrLTAIDiDxVSL&t>{q1+$S4mO}X{u=1D;6!#jKNPHA~`7&4_S!!on
zgi<w1*nOHawfPjzWuio7Vs+d*Zt_U`l<HN+F@P;GC%rd}3nuB(dL=slqe{GWvZ3Cs
zNpTr4UG&{s10J)U=(>|QKI<)3^y_}bI8qaYda$`t_<Xfky|Tha)&WCvO@45}_lDvk
z%8hi{R_(%g>&S`yi@S1uSyv3LOi^aLs~Ca!VGVzcP1P8ZlWtnc-dH#X$xS<|xiqFQ
zFj!P_J#DW^EU_|!zk_IKm=$P;lk90|FFj;L)JI1>w}DJkC8zMS#Pb3(S5Cxr{3F?X
z7j1><*tU|@Wa*l0K0Z&-(Qu$fWT4(CGo6g)SAYwzY!1J11XHrQsWx$Wnb6himUSUV
zzoE-;(w&B$D6mO*yPjRvRl(@tDt?Iv->aE4?bZHYGydUEcJM(0vr8#EWd56cYi8Z}
zPX@V4r0g3TLCaUGtlj0s+RHyrPpA%)Ez8EI-bGgjz5_E6R2&DTmyTRtx6t;x{ks!d
z3{n;8Ki@0~@OwQ}wRqfk9iB3Qj7y;~Ta8l*gGwJCxZCwQ!VLIoqlCEJaR6#A<6A$a
zJA%mVYOW(a;jGjzn_MrnmKj1AtO>~|vDGvlCklxVK)=skXJiR5b!B|6ir`>+p@i%b
z&4&W3Sl3ZJ?2g)aJTAvc<jem##ZjcXh_KgbN3g~Sy89tE^iW*`fh63zcJyC$MJ+uJ
z0W-$zB4paPMShpp<Ft9HCPH5m(otJk+`*jS=~GCIjg#V22vUXP`HFCcAg)9hEnCm)
z^Yc1;j6{*V=XfF`6j-0I>={Z0r8=czv>gD4Ge1qlct@3c{pS4d)QS4)W#h7_p}(U%
zMqRPZSG2%Cz_~yDZ&XlYKh4@K<j9jOoPRr|R^Z<t)J7Dw$?un5)D)@j5mtkp^=iVZ
zqhKpNrD=uWyU%en<$CNV7*(VcE!`m<BZoAL)Z(wMAEAc79J{K{FoedM01VHB%KZ~G
z8R+SRuk;g<RtStnLop|_=4xmqqOiJRbqD(1bHeUbr*Sh@a)P3Aas1x|97YA;Gv)If
zdrg5;UI}J@p9s>(5utqD@M1V!za3?m;E9my9!!lsw%@Fa#3Tb~dCb^Oj(ih$Rr;0d
zEh4{o$l1@7rxJOl!C5#5N6#(}2p3lo_4$q0e)<-i`Q2EX;Wbv%%QnvHndYdlLOvA4
zBtDcOIH%e%ZS)_#JZ+w`IpS#z<Mf!ke4?|un(H{f+~{Xr<)Axpg43*4-i%Q(axDij
zM`f`-QjJ1g*uL`Ci86|~pf7f1pCY1fu-38Hp06I-_q|#nsX6`BK3l(QoKHE<5JoCa
z!e8f5SI^Mi`BZQINez6B7U}D;+8!vkg~*?MJ{U+UZ+pIP6r|Jc*J};@xmm;$%VZg6
z;)3{_O!4Us@AHC7$<suQ+A)w)0?>+p5%gI@TGMzigyZYfj)wAwnd;|Gl0@+PG6&T0
zLQ4M9cF3pF@H0^jsCOzZwjPPk{gJOn_U=m--bnWOE$2x@fz<2iq2N1P#(PHy*&_yA
z>JBW(0=p+Sf{hbEgo2=UOeJasuo}r}wnU-QgQaT%4#HHYE&QZ-d)Jvaf0+@gOC8DG
z_ON$Us_?|aYZv7qnl$kbLWuGR$5+B${3L2&hx9(s;$1p*KoxdvUfR^J>P`C9H23{d
z%+W*PzQFwsAde!`=QGJDJAt6Qmlr|?gFA$=TNl`^T`)@Y_&{5382by-r?&=ilt?2q
z(LAjp>85lcVnc^q$be75MGadq`iY?hJpQflp^g(VWTQyen3=CU^EIC^Z#lRc@o%xg
zZ>y>jcgbAj%vOwv0|~sLaxJ1Iy@22$7S%zl(1Y3cgR*UU(uQo&4|KuGV8IZ%d;F0M
zIk4>j7gL_Q@0t`92uC>M+rNTk-iV~=IDdMw_(KWx8oQ+Y=qI$$snI*@Clu&(u&bm_
zPbKfk1{Z0xcU5zn)MQ&dzx?C!Q|xeonP>dNOLD9>_e+oKi|uuUaIV_3Jt|8v9S0$w
zSsXci&Y*8q-)?n*pS^gN8abgsQ}>RdjxyYQ+hD%dD(wY(i-%WHVT0Z`EjBWJmb`pE
zZB-d%GIN4aZv>ehaM4|k8;iCc89JFd?^PmlN$2}lwseAMPbz>CdgG0V9;Px-t=Xi7
zcslXPaz?Nk7LEF2<QF^FE~?e=S{qa#O+oDTN$h&4Rc<l;=1&kz)$p?fpDxk=h9mJD
zMZlrvod|xTtu0K{G_3ke?`kJ^`jc<c{jS{+-ID{`k7!4#)2Wlk><kGCMV|#!rlF_0
zRwmA*-l7A?scaX}7jA5F%lPH?wQH*U7_m-eE^?ar#;~T|3LEZdlk^S7-2pMh`nODk
zSDeS8*)2ZoQh{jUau4J2P70K5Ie~f#RY+WKft0ynZ9-gKff7bz=>FnjXCyt^1a}qD
zw0?2BSaa)&QSd)u3;ox8WN#fQM5^CLnoMj6)eYuM1O$x`OLy<I2c{`;xZZ4T=EvX3
zaR{5~CMqy?=^g&8Tx6(C+bk~2rQ8t0JBm0~1G4mim5vYc^>vk4ZJv`!?i{fk+cgCW
zYMIjy{BL5u<<$s8xu1SqC{vaeunp{+B;3?qOg0gvVT#rLCi+K!hWcq*_^C=#AZ;W#
zw^2^PpxS9bIg6^(`@ZJke7#pZMm14;E#vsr!_9gx9Z(`gt5kmg4}~Zypru(4hD`tB
zX!L#(NY_;@`kP&huLzTC%+K*9lf2HJ!;V*TNA>AMn}O8*-$l<js*yx?=%HJa=hbTa
z>D)-6WCAQ#l>s@BrAC#75I>Oc!zh5uv@hrZDWVvY8X1#22-C!z<~JMWSGfcX`~sUl
zt*6p3w<a0X!^Xs?+XfShTZ^o5Zyh(_p@zkXhtBd1STx_9>oF}@GYqqN%<a724L;rp
zdW|9$Cw?JFM!z=cWb#35NxyjR3PiE4yCQ1vbk(c2*&_aXpi%#==ODw~{`2BNh($Do
z!QX${(%BGYKxe9Z{eH!m%H4Fh1r7|B?K5XfTEBXg8=;6Moba^%^r%nLzuo!Q$*eMS
z8IDMiu|Gzr>`Miy1(vs(XNM|}gVpz!^xt6sAu4tP2a^v|&`MZBz87~F>$&A)b8B_j
zMk9_lR!d|swWm6ciK({h%Gx@k^1hG7Q<wlJ`rj;tsKS>==+jJ~%@~fq5alGGVwwzW
zkW7_rZ@zlS`ZIb-*qq1>Aw{15r!12?=Uc<phkY$j#Omv)0$+-?-I-hFeUou}OEAi&
z%Db8mItI+R{XjIqPcmzk90?K;%g6o)(>eyv_=+nFu6!IvNM81v{8fiOy!yF{*pVZ^
zcThV>BV;WZ$3o8=CByLU{6QWa&H@m$7P_-I-8NAv<hzmgAqqM%9*mIRAQBux1o9$}
z*1{Lw55*uc&k<&3f>Eq&Nv9+Qkz^~U-T)V+C8a|gBSi8SeTNDkAONE<G`yiMf7=z7
z;WHpah@n|mXs^5g#o+Qs=Lj;k+Zi2F@0-Z2Dpc2=;tjN)&p=%B9d&`po--6-Y^Ob8
zKj4&G>U#_`MxR3;KWXiNhFvBC!5yM4yKnE)9D-zNvyz~-G=}jp%=?;+K5cA?yX0Jc
z2(6gCsywe;r>1m+V{g)*nW#5OUKS1TBL$RV;iHIW4vbO*c95xBtDr!kK|Z%V@2!wp
zVcmO3j|mUlaH3<E9%g@VH&gHcKQ?#1Tz@*)uT7Hq8n4ug_v=C(Mft&Tu-rXW$_V#w
zmRQ><7815pCrD`2oW=Y6>dWRB`Q9`9p`E>OD%Vhv-J-Pf`eGmAYLx8d%bhoxbXY0g
z>}ayP7#9*Xv{CG|o#T=TMHZ|SweHC)gIoCdT&+3u51g#8qCSqa;+3*GsR<eSdKsb7
z_p-0fUSA)~E=McXj-uc-e?~#FYdgT*L?yxq*ZB<JE)=ihw7TxUmt4n#Qk8R!Cq!Lz
zHhOg4MDsy%<+wnwbt_)Dm^a++`sm9{oaLHYrYkifJcjhf$OZcg+i>dDig~=hQrZ`k
zRrtvmdZ1QP=h&I!v2=>!=C`zU^t@5%O-Y>=kWTJ+Nk0j~h|;Tk@4>(HEp9!erO+1C
zm6n)_Eiw3^lO?hq1mQj)%S)|3mqK+KnzmGQRw4g%ICa8VXWNU1?qPA0^(-*w|2QU|
z?F9zu1lKW#|ETu5#e`;td(QxhaBlI_;gbA#c-?x2N@^$rZSOD6yAUom9qh_Ok(0Ne
zgFx@iiTLW{iFaVjClvwQ!IogKgmUe*FuHS>`wHBljj3H+wt*PqlA&*Gbyg~>q2Wix
zgn>>6D`+j+5(ia6uBuf9((rS#X%yG}q}iOJ9UQo<F(-!bGR-FPolT461m3k)hg6SZ
zaLt`XQiGqX@TNe3ZZ4}w9K07dA+#9iRJ~u@=tgV&Mr@nkX3Yl{bmRH><M^y4e`HqO
z6-L3(kUVtgWrMo76XKEa^-UvFA_b!=Q0=9Tn31wh4ar#KB9C!>iwR#L2zkXfp<GUS
zK{@I>w`b17(tg8CUoODvs#rMpWBwZ?J0OgQ;FHEa?Owll8B27QG7=_bsmvPIuUuMc
z%XytK>AmrA2O)Mi5RzLVamj7j9jaO*;q)$!YLMjHO0K^%j-{ApmL9ZyxmrrBujec)
zxLR&FD*(FUq7a-V_fmp3AGWl(6Bt5TD&vA{d&;=nx~1^RHhSx;t^2wsj>HRd1TU{A
z+-zSOuuSX#AQTeYNgVA(%L`Z&O5a&a`GNLV+y~-ESnIOfEo}FUEVCRzhu-KXo{H>m
zdlMhk>i{P4ZfW@=gn@q3Pvp6pMppR%L_-FP@_FB)_7Xd&k+|7gZzFFvXs#I<S~>z4
zDmqrQ?mGb|E$4?RWS^VZXLctjD<zl5GZtB4R25VbJoNg_!`wab;XedWNTPnM_usqc
z86B(^LBVVeT}MnE%vUo9*0q~M+5S1<H$RN<P*kk$a{m;|MA*cq1-fDT-f}`Ia5RmT
zAG-|SuZC(OK4)OJiRG^^HDhNeZ3pLhlSP!%nc05_4>~~snysXqdq~TsceKi9S>^D$
zPeA9{e-Fpcwe^*UrU2B;Q#fC}+?sa$+11<j)9Y4c@C>MBJNP6IS`WjJ3f61Z1NbC&
zfm$cN8R$_=b~r#R(VmVe9l*6Ji~0fXZxv>DLCk;E*>T^^WPJWR;ilVe0V~ta1;K&z
zyDK6TAEL!2WPGXB#zBUs2f<DULUoZ9$~dk+qq@Gp5+v$YLZ~*^q2G8VEUBk<M=t93
z6}PQNIQQAkTT784W;=w>yT|sS(lg^rhK(tC50;JRz|=ppf)inj2Ji_=50Yzs6r-D<
zbR2b)D8-&KrVVqM(LtN{^-$-hfGiRq*pLf#ZsBOVM`ca-JELkkcW~DDM`PTt<}+l;
z7;jC2C~!VRuqawEa|lK-t7JDTYsWz2!o|qyOaZZzXcCrMkK7KQnQ*+-U)5%hn4Zy6
zNXNyHesn`0$)<<P(rbwB{o;TbKWlOZsW$*aYr2=mn50K9fXLLO+rI<C&Qsd`QN;CS
zlmtu}z5wolTX#3~_a6p7{vjmNmRH&ii6T%H%_aS;+``bOpW@An2mNNgMlLs7<i_!A
z#8-95{rz*Xam4TC+1KQa#1_NkhZ8NyQxHHBNiQ_V8jMgr>MslLBTQk@p_!Fb>GjJf
zH%wpbzVpH1o`3Ho?tQF69=)UdkJ`!fn#g{M>POBwg}(`5?q$n%(@RH_x#0E@tGAuB
ze8FK4-x(G4<WDLm@_E|4oRhna)P535=A?8oxM-=R+MHqC?0y4n3|eAzxoA$IO_A3J
zofm2*(<XK5SFAUvDDB3PMe3_PlK~;=MbXqv97|WLeZdb8#*O4$@rLn6u`^pu_g6hm
z-bUfuj3Li&v4Z0pPC5Na8<qaF`X4j#nqoWO3kI}cOXS{Orn#E)VbaSdtuuuT_|4Bs
zqq0&5GE_eG=3EN&Mz3uWdl?y;g)N)k5qMdJ<^ch-sSy%u6X>v1F~HCg*co{j)V^Iv
zbl>XK$_1z4JZ>sq6$E`OQg~>c8DX9TPwnF0lV;#es`8F9?<s}*mLo4cm^fnI#)*84
z7U@`s{<!QY%;;@>osE)yEI+SgbRaaNbbhm~Fi9YE(0Vaoxij;ithV@YHJ1eY=qT5!
zLq5?oS}hbzJr-?gG$`g&vUG4jv4`3lT})5rR1>H!zxk35Ai|Zu#>HmJ^w{C++Ro@D
z_)805k>6_<qq6p;KwCT*t=rS1XH&s%9-5JH?w5Z)52BSvf0KkRZWmjPi6O`auq7h}
z(oK%v$7Btxo_&<!l1W)RTt}7BGE)&Ee38j#isfzShxu)a?4PGaNHtI^iFUfQh$4H7
zl+tn@l9F9ukTi1FZw*thDsHp)iUEjnk@XQFiG135!%Cx%mVLoDi@YDq0lXue192tw
zUQAM2P4vhyZF#PId=n`|vnRe0Z_f=6N=$@t;JBsnAeAGW`ih&KbJ*QQ=Oo1|$H#ug
zVkHF=P_u$-*F$WV#9H?{6$rjHQ*@rf+QlO^+n>&};E%B8`ir&aAMGWh`f8?ey4V_Q
zHXV;R#JFqj4_Q)cQ|IJO)u)-}4iG<ea$GK1JB%?Ly3SZbJ7CozW3=MHct<#s;3ucc
zocIKep~DrbCEp0WZ!29WQfO%OI17W_{1=SqqmR*=rH_TYPJ%P_Ly&-&W%?Q9MuWPE
zHM>j-nzj?%`_wCmcvm8z*KbF#XOp;%xv>#zyNZ&&J!F7wlG$O;T!E-v42OIVf9X4Q
ze29bBTCn1G`IXdMrBoo6hdx1wQfBfOyWhQBKmOQB*)^{-tU9P+=4mBuLMX75!PZ~%
zyhsDKb40;jKSX&~T<Dp&t^^-4*0_Iu&q*bm7X!x|0rgyq^$J8LHgp@^g7-oN&c>Ho
z9{z|ttwN&u&uW)d!K$6?;UA@~GtW%YwGm}b7v=kAs;$XbtW*RZwEcHdgx;?SNd3fV
zfM{Zy@O3*V#hAa5`pKD+3#=2Bb0@Vlt@5qaFD;jzGKdx$FWz?e%zK+Wk$)@Re!VR?
zsXNs$-<AB_aOkq2nd+c>64%bz84au6a=pfJ$|XX_(g=A14u*91e2!J@KrNOrq+}eK
z(MItE0quM4j@-L3ZJr46n1qa**U?`^S@V6~4Sp#oOJlwD0lm3LlCX_(cB)URL-uFQ
zaA2Et-6!bH$&s9Jtzi~BLk!wUYF-;3jbn~XUMfJiV>`!lojJOxXRU<*LMZDfl8Cli
z@3b4~Q#|w`=BJjilF#_Z7hR*eLsAf7xR`{K=EzIUZX~9Aua3`jUp>&ffc96iv)gB>
zr)#pSFowqHW+&2zeD|Z}JsuY^*r#UF)W=y#fuF)Ake|yock5{ac&9iLf1ly2YtZ1{
z)5e>IT^Ci@-TkJ-J#YQ|v3!S$7GIU6;bfNrc~fa{7HD?B&xcLtS4E_^rf_sIww?or
z`?6K`3JpjKbRxhTq~A<N6;P{4wrnRXwa#DeO|6)r1_RXqO&(_K_W&a`tSN<i7(g<w
zt<%pPCI!nyjU8|kc1`B_97uy|)G}bc2?z;V*6P{eaYnrK4{5!hmP;cU$gRJJ`b{H?
zAHpyKSy^I9Vm@P~b|vD9O{(d4k#P`#ozJqiK1+@X)dRDK4c}-r(SMmxu8Li=+Pbs{
z+u0La(<MsRL~y+B@2BTv)mVx7Tor~(Bh;0zcF;!#ASfbAQGyv|+ge9n+nx2v{-V%M
zZ+d#W8u^yQ)k{!gs7}ThV_Ptg4S+`aOU>g~?})Y49*tQu8x^UP*T_|fB@VI)*}cba
zlEqmL^&-32XPiZo_q~Q01_%}M#T4m!>&^tOAnai-&Ez_u8mC=#w&>_U!1I7Xc>9}H
zQigzh`_YKuh>%Abj>O@Tp0VXx5E#xDoK-NyL9WGT=S8lCEis_UWWsQ;v=CiT-vtxL
z8Et<58P7+_qotB>E!%2u-;U<*$%`rnpy8ghHBn?C-}?A5-ugC$wTakmPCx!1$E3z^
zD(57L?dAw6y``UR%pxyjKyp43MawTI|GoU;B}2dlm!o}ArN0A}Si`xKCi%&~v$@u<
z@$wB4FQ+4ipqd+fkpA*^NdIELgWc!NpAp2|O7G=3L<k()AJoSk9d%JSoW(COT6lTR
zOrsKvP#*okaHKeDD>WBo`3yi{7`R?Gro>Z6(o6kW=uM;ZXAe*cHpgtTq-0I8jl$2t
z5Zy30s$hw$ckDl;r2jetFWz<@u2?p$;0I<F#GOtg7P=gte}wis0GM8P`KR~b9>18Y
zie>1H<^96wm3$Vb#`A?NeVD6;UdLJLmDW@~*Wi~om+eZb?!~<I==T#H;T*-5m7gRM
z?7tj8S2V`93q(<wM<f{>$KmeYB+>@I55`8MF+U+Q{fZ!N96GyvKCFj=pto?3#%NfD
zX7nk4Cov13k65QItmE4|-)E^xN!w$7_s1!dO+lYT3W;{-IwZD#ONE333}4?>CrOhk
zv4n+RS42MHyPvX*b{y`jcw{;T@->bJ+h7--1#vumK+9up$?72$cRcV6U6FDU3L0<V
za*U8-!4e0PrVV7Tz(woF(~3cbN{DJIvJ%rNckfbkwO`KJIisZZPOxCx?R%R0(#pE`
zS+b7#-t|71RBT{srfGCY1YO3lLO8k^ZU!4%28eK{NguN*kZ)o`UW2P-ZS&8ijYgvQ
zHVS1Y<a)JLi4%>`^rQmQXa+YVaY6fx*ujCrjO*Q8M}~Arb5=}fJWC(RKlWg3z>r_r
z(HBr~r})3g5dTT-0C5e<CO~^t<e%}D2z}@5!hU|LU4f=m9-ByEC9|3j=%BM0LuzRm
zgxjVbD=m^>#DPtXOb;YW0n$8miR@h{rYDo;`V7imAj!Lp#U`;pts@kYMyE3>^4^HG
zm8A3MkmJ_1*g&`znBG;Mlm5CVmdoQLn8bk44*O?X(cgtuRi>dhW1gD7r6wx0Tr)=(
zjC<G$by4OmpyaHW>1uHPHbp^xn`83!EYU84xCnfAZwSd`+HDnVS-=3%KbsqL?_ECZ
z-j-TOh}WND9_vSeO5kJHNwbSZc-Fzxk3z%4;@P!0i{<vDz0f%pU5tNATgCZrX)O%4
zSJAV)Ih%$fpTvjVDBO-(GjR)Qg#~5BCTHu&vPwn}MqMkiqhDkY%c|o(K?!zjrrvja
z{*!2LDVvsF()e%Vg-~1+T}XUw`&Vv~w!!3TF50IIe;eH5bU{^$`vWS$jLlKBk@mun
zUM{V={StO>O|!j{OJUP}#EuM_Pg+7!%j#RUdEWy5&gCjAeakx`v>UznqaL|}j-loa
z;f@Uio=lCyM1Uqf4i3a!v+sTB5xXOVTqjMf75Y#f*B9N(KCH`a2XOn?OtV%N*EV$N
zMOuX{jx9b2c6!T~+m9E7l1Czl6p5vc%!2aE-@_P5ts+}<R}?SsT<pwW&R)u;!=i9B
z()Iz>IObO5B_HXlKc9I!9o~>|EQIO}+Av@W(;ejFL~fu?rv{r=PNU49Ph$Ae8Cr^E
z=jkXWydRlLn@i<SVyI)qd0k$ZGzQ#M=ZwRqNflI`y2^c;N|%j2p5zVB@*dHyI*s@&
zaG%k|6faySHe*mS++ARjF2UFvy*jV0N40<v%Zisb7|DFoz1aN2Ci4n2JNV8NhVOy+
zk$Y^)obms73f_aA9z&qD$bvpM%!|f%i$x8j`r&9KI5q|Th(;pefjl;Z!-+hW>5U8M
zNZ!FweEZv!{>HMkA?bA<PJ@8#vaE2Um(=4cEOAQZQTASe2g{yMuxjLlK1G#BIVs!~
z@i%0=xIffgPygS=Ox^LZ>~9W_qW|M}5ux3d_u*-|^CYTF8Kbn%?(;+Qa!Fv(^%3_f
z5S5iYEXq^#W<t0s<AbY`CQOp)EGfMAwu*F*SieOPMZuZNLMWSHAYy}8KT(?T-XWnp
zC@4M$4!5HgF-m-oxGdm=1PJYLLnQUz@1D#4-a7mENDDLDF8Y8bWO{QtneX#~WxHZ{
zzwKzi&Y6CIU^*InBdM{@BGKE8xP>igy@q>7B4Gvk`&N~???7lAHXGvi8S}Ky>m*;t
zd4aE0OtIPT@CEgnXW8$(|2`07^BhTx_fRn?Nse#e7f%ht{jl%@dCkZ1u)|fA|71?o
zj^w3j6*f9f$p)JPd_L@KL!w!?qMn^1e{46EHRSPsAw8k|x?9fb5<zku+V>y$&`jI`
zolh8Jf8^L}ucweG-GD$j?jj1B0X~rLHvR^PW@F9pU)@7jd$NgW6}aJQ?&#3g)78Jb
zDc2Q)G6oSW1q)z_WCF%IFL{b=TvIeU-XAOhowq5&pDmdXG8gsPPmi=`Ic8(m?Y7zt
z`+>{DOM<u{`M&~d%Ktl1@$`Uk1UbsK-uG%(^`{*(0?*^j1KhLw@gi0|?`;9t4|BNW
zN!(;k-Q&f!TcA?Sf)QJmWuVnRSTqQeVYBaz_3j106iyO-D2Th93ZW*m&i*8BpH2}Z
zv@n2Scuu235#~UZa6Y*9y<;uFibGaft1q;`Zd}lBL$EqJwA(}XO^y|2zENmQoAD1I
zOIma{xJyM?5=T~WO7iE=J~ApcT}BH_P?!FXPw5+LQc(J+fNsdtT~y2v@=q`zEOQGL
zhJU>qOdLYXzrDVTCU)mfm&)<0!29o<AyPop5Z$60!MjM2?9t}xaIYk7Gnw#(%+TI>
zBo>o|el=AFEKEzXypE;l!=<Un9veZQY_lm~`Q}%v-9Kp28hz~;((l%TXTHw<wNwvp
zjUzWKYZ?pA64frI`~kstIh`Juowned&%P6wnts>DVH;^hqKz$)!jrg`>;DO5pxl4u
zjHvRvVRw6Tn0df<aqWsTds|&7be$NSThP|*Ex;%>|3mO5eoX*|nxb=)bM<gFWQ3Pi
zZ=z2Xt$l_!JA!aH-#aKWaqjcUyo{bwe8*;;r#YpW?&4q)JSxv?N)^sJ!d2AK$i7+e
z_*LHpLd*N5j3bqNurFZiis{#a=r>J+J<3v0&x85p#CE?sgP@R)n+5Yl>8L2--aqJ@
z@{5`+ck7Zl=S3KbrEWu>WmfN!Q*}5uQ;~c7E;?z;1Dc3Ylwrov_>N1brAcwwZL@G*
zpbwc)-7nKl8v001%r1VBZ?<vL&GxB;eOQ1=S_3qa9*32?oC-hZmPF(ZJCjbriO-0B
zlztG<+D%W?wU3GK5p>VJF)6Lvlxn+yhq+~!kzz1phCJyMh`z=4{#emOz}_iz94@WO
zLE%n@OXr^QGSF+sulqNBQ42DO#iAUSG3^^UVO%cUQ*0!ncOys(n*_KiY&g4TUo@1J
ztzv=O&d-?fHoAacbU<Q2Fc&m1^69p{fZ*rFYWaQC0o=pU1dT84Qjt_odOyE^vE9on
zHRCP{Bk{>v#)hOurVM$fq5sUWkC?C465ul~?xUrDEl{gZ>hSNVT%pC;3Z@1HU4F{$
zK@brn`ijcRHIWnMBXiXhxZlJ`AkyD|#-&TydCQwfEX80Hu_SHt4<S=+tk+f#2^Mz6
zMOm{1*r6C-YTheu$LOYMr7;iwK!Pb|6?_2`iKs#*Z1e1srla5cDjpY>D($S$)SqXT
zkggA%vV>efx?UAQPjeR&(Dkk@&Lv^__Faxiy3@eeB#C)7Dzw6g>2V=|Y1#4gUvfNp
zK8@qE+zNG_v%ugD;U`+7Gb`khzv)U@QK@@@z5W0tA0k~H@|ON>hmXS#r+{~rCQ8vb
z+lvK)mUiFXIpG3Xj#U<er-)VnEw9Tf|9m9v#PdP;8-R7Rt{(0h)=zm?@jCTV$Tdxy
zFqrp@0YqG0x@RmU1Qo?_kRyYs<DFP~?NJYO{h>$CTgKYWQpMQxJdFvv?U@uWAT~=>
zT0UzVJBKRH_zw{fM%3jLIK9_`OFKBINwrx>;atW62xZk*DuNhMbk|z?v@)=7Fr!@f
z{fvu{&d*hwL_rH0B~biL+{W}qsGTmVfsP)pZ3QWAFDAs1>_`T!R<|;WwF}Ah%Nl#j
z+E*h(BIU_I|4XiP{)K!jD4*D*tW6J*v3`?S`fv;r=q)S3Yuk@ygN1GU?OLg&Vv{Zl
zL0z}`lI9X&$`nW%j&Fz5R&}2&S^kQP&C9)5(wIxvpf=d<C3;snoZxdLZi31)zX|0E
zYn(mj;+W3iD^e<&=rGTmLWDG#%r}jK1`O{eSr?OSm5Q1E%c6F3seUjNG|$9t;-<f=
zz80B}Opm@%%sq5H)!@{JPq6f;lj&gM4r+nn*eX*mYX03X?Ja2p;ZE#AsRGd;^x=+}
zx{Z4F!Bk7i5CJ8%pT!SHC8yIfednLw4XAY}UeCsBe9tM*O0<*O0Kp5vg@qe9vDYeG
zYJ&M8zg#&w)CO{z@3LNG{A;=#i{TA5iBEBn;+4`{gY2E3W`eVdlIkwnyh9!l09B{0
zd%vv?<sZoTmzKpw2`4Hs#tbN}{J+Ayf*}Zn07?mP{l*5gx^kX0eP4cEf#-AxZMh0z
z?MmaT_?vkJn44kCN@MKB?nGpp*a2*=e)7v-=y2WT|A_6Qn=rWxK*n9doNu!uSzG%C
zyCyhAJvm!LsYv7O85=o1lT+f9sMMnecX5;sae5iblb<ChbrD6l1##vST;eNu<%`-~
z@-abja^G<kxg51{Wloe|xSOTA#SJt@5h|1?1+9h3QHd#&RDV;NPrp#hIai;zhV$O{
zLBHI|`SE*|OZ}-+ITYFveUs9A78hUZhS=|^W&G1$l%D)tgk}`idE;+hH2UIjAuI;h
zEfj`A$h9F8Vgd^kyz2&Kb1;0qkO;5M-PYB+eIEE|H&YRJpJhG3$03q^{S!(!TGTg`
z=Jg`^iNZ0H`JNF$E*vZ#ff|b&-;*N8R@2U7cSUE%2s86Z#W+BRf&@f0JySvN>FVk;
zZ**Vl3eumi-dV#!e9Bm<(q;%~dvZM(ZndIJTKcaj7wKGlx@hL>YO-m-lrVIX^6$qx
z|5i{Vrsp^rX$F2g7_I2=S6YMSv~tehSMxoVcC7X!(+8r5+zTQ0SAQ?hIDm?;RO&CP
z0U3ROm)VLxN#nb#?*bbX!PI$SSW>dDiujDrEF|6wn>a5TsxxolQ*EQd&SP-Q@N27*
z_|NU1Uh@TooJxm}s^s${1v#XS5A3m0Q>}?A;9qfxoL`vU=`-U7s?gB3$){N^l*@SP
z0qTYvk_;!5dp<i`aijA*GoskyU>JW;xU<gW=6w!DVhMh~{}Lrm6x}`ts3C^kTO%h3
z!<3{)P$1Uxush84MA&jZa(z~73!}(fOZr3A3n^Q2<95pk+z8HHcjGUXkU=$G`0Gzv
zxhQ=2yf6J=nzpdEKc(vyer~1xXS^Tfu}1$DUzr?#1n;oYv{?|Y`F-nUd&8+?RdF+-
zco$=3=_>23JNVDkS99)TC$R6vhRA#TgubNr@P5Yk@#Z`E{G3jaFVKz$(kvL)o+vO0
z;IVe_Dq-WYVTA#V;A{Q(J3il>aTg)+CTvH^_LkVkCY-mwYZj!Pk5+Dpff8q%qc$gh
zc%&bWjb;T0s4SB}@=kBQq*PuBboOy>g52_>{a4X5IH-zJ1w*1n>;vsn>YPrFN>WD7
zHbc~<eX-vSF?AMKOrTUmR_qq?AP<Zg8yGBe(NgmGEM!rrfBv>*pU2pg3_gm%d20GH
zfI}v-7HvObTl=kp!pQGOw(wbZcRzsGs1L9ALm$1wl&R2gB%@Jwl?Tr^04hEsHdD8p
zHQC;pv5*2!8r?cEBzREypwv*03Z>YT=f%WCIUSjmWXyXO%Wp5}0RHKWc0$7uHXPAJ
zZ=CheiK*p7f;-XJ>7`XJ@<I;K;%IZ}_pIQqd_lOuT#k$5HnyM0OyhG5E#6M8wT-VR
zyM^7pi%<Z7+q{$phx(NhEWb=A?33306fe#nItz>dWc|jv`niBv7-dSfwdm_~{6u6Z
zW!!=C^^$6+)#Z2cPHoflgdgO^b(oZV`T5O+12FZk`n!0yzRp!Iy~j*zawV_P?DVon
zrS3Bmdhxv7oYKOCzf7J#mhwCYE>~A{Rlu^7B$;|Eja&U^afjxZoc}JbmWFj`4!z34
z=R19=xEd`a`qh|A>3hb1d7jT$lgui(!b24?cXK-^L=<@hXUD)zS&Xv|)o!T;$sEp_
z5*bR)-8R!*db77Yxb^j+BHsVos;L{R7}p3y;;(w0I+eIf`H)h9!T?<n#7uMp%%3ku
zu898sya0gVWsbv>2m%VT(xVgyWn+}q7vN|b!KETdW$)Mjo&-X6(G(lhEvF=n3}M+D
z!IP4a9C|BBZf)1`@=Y&+ppU80+N}8?o^5kB5rjxkBpEKOj2qJBY7w!<fKIs{vtvfT
zM%JD5R!J}#2zeQL;g&g6e)BQiW~f|9yzQ$`fqZ74YT3|u4@SGbG;91&IRB_a&4v4i
z{3Xt7l9x9+V)2c9Xqa;&ruWjdEa9bNzTa!~rSjd4cpkp|-3V+OM<ku9Tu`JQGEHk6
zACnD^l>X60Q;TYx&(AkWKYf6hmN0(x=Q}1Jecj?*wN=gEP4p&PK%Qd1G_GT#U~Uz?
zp%LEju-cH``qs~hX%QkmGtCfZ+0f0Uksh#Kr1;F>osOpE+UEJ_aSeBW6!d|7Nk{i+
zKIe_5B4zp4e;rAo2pV`fI=H(}-q(3f?{FKqZRgfl-j6f(3xMulXmKgl#WyNv;n>de
zZ+H*a;8dz+A);+TZb06X$akPYNHuhozg?EvU1C)O-5bmPylq7M!JL=xcDe+ZxHqu}
zxYmi18hE>zW0snmx9o&)k9}UgDm*V^?+i0o1~7h86p~-^;h)fu0qrB^MV*J6BHUl7
zNlmxZMM8>;5Rmvh=Akq3lv5DFC)mxoru$-dyBF^MeaK~9-4R`+(3M|UGxUn`6`oX7
zfk{=DBwg^R#mtw@wnRoTF|vS_J(D)MOKK^n)QOvqdD~twTx1y-sYE!*wA3C1CY%Ju
z4DMl$*=(!pGGdk=maRVDM3{$yUZ+!?q2cJV`0jCCG0`Uy@k*M;qd@gOqvL%jPUeUn
z)?Ei+Th6@1CI2P5k;62VLE07krT;6pQb8gEQUJ|Wu1)oTN#!nP<}oVD%A^`4zLMD2
zzhBd8*gEwR4gxj(d+8I_l&-(3^>3hkmjuC<Hzm{6P$@)rK{^oc%VlYO$Tk8vh(uFH
zxt*PzJ(%3SKlqtefJiNTgk~OJ1y9ACK5V7xU`sR$B%b~6GQ^psaZ4?}m+*xm8(3w%
z{z-p1pe#&7>2{{~sCgmpcx2VZ`$RsvF+o{Vf>LLvS@7lK39HG7=u00Z?{J9_nl|&I
z;{=HS(R|-DL_4$e^qSH?uUXwk=N%04{89T~?*c2VWOzEUl-#T-SGa8v7p)6!{da{K
z?Xs|0PcT5Rxs>IrcXB`0Mh(VQu$jpCWkaIeaeqy{42PyB*LouXApu5WucRVOk>(7T
zEoM$q@WmGL?4kGC9oTPp3I~&sd?Q4jQ|m@mHXN3fqtK~`a{R!QVHLfA3}AE>w!q-F
zSup;#V1X>>TCn^}C9*-nE9|7YQFmcApP4{2at!D7y^riMvJ4W~f;=~p<Mu?rnqN#=
zUaoYxO{9LIrk{V5T!f`LnlS<{;a}~gVR7uDgN!D>WQE7FnHmBDv%qjS3LrPES!@KF
zG2{0T1ql`{4x(q{Gr3*CYp*AcYX)({>#TNf3ObvgYg%sidk+>JHg|8no~Y*jYfoG%
z{Z9q6o@mZDE^3mrr+2EY*Zb9my{c-(xJ=@#TB(+Uu3!^;#NP*y3XY@}75gbZcLrIP
zq!CKnUqOA`<}0G{#th2<-h8flg?ba5wyejw$6HT)^Xsqm)$CfX1QPX&D>;@zhA#uj
zO#al}GcP|)T&USLd0(xr@jbha3h`w=9Csz>(%U&T(f)O@G;w2oxYF3!-CVKb2niAU
z+Ppgh6sgKPE{|N<UMO)_3C`m29p?cOmDkvA)#MO?I~*}MP0*O`z$<GB)`|KQEWh4&
zW5jR(_7=H2DM+)?Dbo*^W5R@SdSi3Yb;+0puPB7@`O!r{U5YG;yQH62>?A8D<)ve|
zGGokz<TOwNE3k!Pt+nv$#kBBADtZQ)#Jo%)L&+QWgBv`*bSMmf_(-Y*E5I+3dPSMW
zRy$JB%k03YS99yJcrqR_zqdY8ZNFzQ^^?hO6_4O>!?hOY+Cndz#T0Yr&&m->f8_d*
zH6>%rhHS`!O4>yviJ7!^vMA`?rk`Wwn6L?m!B%u{4G^}R3Xc<9<c3>9f8XO;rOv5@
zJ!V{5e*6B)k<<rwX(lOEI?2m1veistZ$ntXc39k`!XYxD9@8i4!6w(!#wvo?b@l9W
zvIBbQHgbA-Mw!91`?Tv9QsIj&!O9D76MESTg*`p`4ktG?{abw!9H1_W#ZkUZk0~!#
z-(FysV%k8(k0b#q#UG&DS=7dSoawGysF6oYsfyl{Q<+kfl_RgljdwS+_|$ty3_Z&X
zcrd?#*l=o;7xt=i>)^xS&VQ{bC5z7gq3JE$ntb2?VGsck7%E7^L>eTeyE_yRknZjt
zF`9vNH%N<g_bBP^?(T*G1D?G<zwh%0Y{!oKymwva`Kq(}G!%_6x$ga*$UWX}HSdG-
zf0Ke4jzdvwPu^|!h5(?<Fn^qIJ1mRb9G)VZuiYmTBr`X)W_XT~hXRW!v~dzDl?8<F
z_eMKnFNrtPzZR&pF<d7Ki8|k)l}=@2?f>)rKI9mNwr|QV8id-`uK*6{{6i%kx207d
zLFE5oM}wR_NJmzRon8DrP8&sb_s^`Wn-nnDltjxdw;_YgyeK!lCZ85#R^lUn#RwbE
z;D<}XTofW`qT>pEr@`+EPc{#|18gzEG=Y-MlbG{ChIICIDx8f%5@81&EMW_CM@8aq
z?eu*GH1pIaZdQJdN9817JM$5)<X?6DUI<5<>@w`$>t!8iHEYAH7@MXXDydEuY?LN)
z^l#){EeE~*F1h_%Oi3nzXpKSe0NRWxPAabjOw3Ni>)^eqN<IEOLJ*QKg>5rM<Kgov
z-JcDUhGzaUNolStyp`)gv6dLKi3%`@D2%98bB)M<g;AMrUm9miPfyp&Hh5g;E`mM|
zIxE}{@l5NJsJgt4+Eux?%_z5rmNT=AE>$aYwf$ElTJ3Ymb17^E(C2i=)QjxEy+V{7
zf9*plkXs^8pZClRcRsH734K??lYi#kBh&jeM%Z^>vwqnS%%Pf<ARm5uBFL{EWV4Ug
zkre2(D8C};Bh2>*)m5M<L>AMMx|A8{>%Z%0iY~66sEG`k>~mh9$wWyDHRc+7;G>Ze
zb+Mv4LgXqrg_UK5*JX5wevl41xL(P?bxB`jbBPox&-In@zHrQ1XiKnrrv8}Srge<l
zpcyp6V<6_VvrWh;y-+0Yz7_lJhZj{M{yF^~@a<oi3flQVUmkN;3Z8cyIKH(KlTwv|
zpC_r0+vn0A{)yn}>+eSF-R<P;e1G2FAYD4=HK!e))XR(j3SR=x_92ecC4>a#S-V63
z`Q}XB7xvhm7JRXH#$5t2CkV9gfxERegYUpMn>{#z>vyQru*huRN9C9Af>YG8GUF~<
zqt=5*ADT(8o#rWEyaF3S8BloK>bo@RQ9Sb8iQuKgxuERsJ;P4G1t-QJFklMx>koj%
zXB18aeq+&k`y$5W$6YT4cztoZF`YEx2zJJ*?(x#rw$63MG8OnibQ&k5+y*4ejdZ=4
z(#g_k0P#Rlt%Yz$xP%7nFN@ZX<CJ8*LZOf(WS*wtuMHLoJp?a8zN3zRl!a~YL7wEN
zae@%aVG5JAh6JaKI5&;e|GOXuoUK~rXNx)9{pbE;ICcA7mMl*5m^!+f=R^U{oXj=2
zMkE~SQ9k>$_NuBd&J7m?)BFnXa&PE8ap!2Ym0~>{$<M6cE8+ZWOS{l=wqHWQ&|4?<
z6paM%tO_+Sk;x5E(+CqGP4l3Nkm;8vVkPY_w!?K33s#v1^)O>#3*pp(9vwcdW39e9
zKtybPm1SDnQS#;_kGJ==>;EQ1?e=;ry}$WOF5(^k63?-_5VZ@B@01Yu%#DHlUgsCX
zC!&8WCK%u5esp!hx)sS-G)yX<nj?5m1aw4+2~ZL)hCnTv-)a3odZ=sfkOndkw<J@O
zjUlXf$pGb)NkNRH*v#JWV3$#4`3;um(UY+dd7n&iu@6T9S>AC=OxS;y=CN2!pKQEF
zTc+=Gqvyl%jtR2+gg!As1CT@n9_%FDby7rN3<f*aU3SFRmjOE)Z|7c7KhZRjW;ulA
z!MWPQ&2UFGVHr|k>3vrYQSp#Hq45tVP5S`)K#yG^4}~}_eD1nrQm~CjsE#fP&IO5>
zbfq}F=~^gg?p;dL5I48rpx>zGIrq&~A>T1+VChI*T^)^<xQhn^tN5YWo8j@xRX3kq
z&lFXA-jyZN)59WAlk>{%2C3O_<DwL05%+5v2(18WQ9c`OW}I6*^hi9`Po3!D8Y4qz
zs<E_{S)aqn`{JyNB%Rxg^l)n}aZQbUAZ=^MaEJSzT8GHKn?i7|c)j_&Z(ruRV?u{P
zAGGwKTWu%rr3C+vlpCWTVz$hb^HG@aAEfbp8)@sn!R*{v>U>rBywq!~;nXl;D?hfw
zXgJ7I`5znL*DbS~6Y7x$Sg{U$y>)(JoK|YoL7%n7YD`-VV8l}1lEw&&Tf41vX-^F>
z?3|R>RN-V+36Vk(mrB1+PN$5;+llcnOiRb5+!X0ONuWvq)e_l@TL~a*Ubyjn)L92<
zwPK9mW(d33Y~loByt!RtKWy!w_Q3|QQfiJMuWwL626iq%Mu`V(=AVT=djIBY`^Si^
zVfy*B!6PBofs$BY6z)}4&OtvQ8J8Bfsbe=T9xX+2A;7;yBE0uw;jbs>yE@S-VX3tv
zTbj<@e9m|oY1v+F?U|@NGV+4oUoq=yqg{j9)e&+8*UaVJ`K5H0k~83A`^A)N_<o6C
zmy=#*=n~xQ?ULJTOAa{43T&a?iK9~^+|L$+8OA@+QKcg+XyUc77nuzWtj$!rWTIFs
zXhDEQ8tI=aY;CM^UFFsGd-f-{m<K^iH%*wCnXg8g7QZ7};Sxxb$%i%LR3NqTsj}|7
z(c&|mEMuH#A_&|t;n2ha8aJ{cv={${E^;E)BM{LMLocHyvKQ@}2PgmgyE8BLB{Uq}
zwpc&(<qxUgM-J7q3-(9msbu-|{MQ_ZCD#XYc=Jo6UwK{;pvBPS0SHM`SZR82Vd;Ie
z`RBwuI(B%UhetkJT)mWVDv({G`)=){w@7t9zfqsY&?NoaJfT7i_M?qX7ilK@HgG^i
z{WXMb%eEqY3n-9-DQ+Rw%}bk=_>NWP+_1y3w4=(Pe<y-8SVuXpb)PZvK(vQ_<iPDc
zl(e~@%tnwYVcT=yirP8Y+@DUo>#Jq|(t7OrA55LhY!k;`v(8H=`NM?0xg@IK0(#z+
zuD>u6+duJ{ML+Y|;p6zA8xmL4jKv^S0{4H!JpcIgw(jXm!=maA39VF;h1K8l?--Hl
zVhLw5H2+8)9tEzpEgT|fHoWS(4M>#C#`Rx}GUWY?xA9|;d(CN5tML{#LuMt=Kbg8{
zr3QK_;Iq(L)2LWj@DFBO!$nL?F8uEh3l>7KrtM$Fv4smj)q0BRv+lSU?U$+s%eev_
z1UX=<H^?QK-f&p2INoOAOvJ`zxQhQoU-f0{U%;WYqS0B4BKXPb0ct+RI(|R0^CYaY
z$*n96))5Qq_Xj&S0jRplOc2{&tn}?}`IJ76Fe+NmvPpeACZzp?V)l0TO`ZN(NpUcY
z)aYGn2<o@Zm<X^)+J@HKCvq*Ai>n86psLQbY_xc4hdDS#Coa%Yp%sIl%!8=ZLAGQF
zkJTX<91MIP&B!wN0fO~?7E6ebnwX~O<1g{p2N9~o@4=|68NgRRiOnfm<xrb?`DBzj
zNruerU_L_Y{>N6*&dJ7ho=q8wR~A6wvTP5B)(mU;<qVBf_rVJl`$I~YzHOW6+OYne
z%U@H)ZZ&G?;7@8BqQ<%t<BpVp`H>QCXj3P}?dT2<qtDQq%1Yb*c`TzTMVow7j%%?a
zk=BmRLu6fECgnfTR+^G`znUWKFUK2rOt4YVIc!L)UxxQyryQl4`v|XBCPm>!p_YK=
zzkVpMVGA2BI{QrG!~<NnSqmLLrZk~@N4hX7j1jv;o{Pt(LFq2U978LfaK*k63uKQd
zqd?ifbAND<XAV`h(Z_rf{d(U}B(6Jjw>gl(v<{f4h{i?xnrYUukbmMM_O_;TKOLEA
zbrb2n(2#1q1y968UdBBi+Cv9jd)HS_n&ydrq+4%aeQ@~DT>(~Y?xX+W=#BZ^_oJt-
z)LvLC+M(_E1NF^EzLKNhhoNHriJiUQ#3`4Nbxo+yDp0nYn`7r-|3ge2TiWsFX$Tsj
zE@3*r5_W%ngUIrOPkwy<FW_&Y&2a7SfOV+$mob*pJz47wlJqL(Y7z$P)UQ%8FN@a;
z)I;Cg`r!++_{SxW`*n#x&~#?((Nnr|35(e^=ovn6ICZmdI1sI@k<7*IUFEpGyqjW9
z@ExZWaN4u;lnBpZw?}EH+d$vFN1Cl|q|BC_<;6rb8$oF#(f*@j5}tt)>hR&qRdWWz
zIVp!G95{L2O+l!Hdo_+gCXGn@wBm96|MpCAJLoavoIDp53!i2k@uABpP81B?EZ<z@
z4X1x70(=ufgq<5-t%u(lVO@H%ArG`y;wY9tJ4P(u+cAodN^cmCx0&$f>mNoot+&h|
zKXq^`N&3jl>iDW1jNFVI!Z*7flOw9BZ5`4lB~-ym+Lr4R?vmOE#@-WTpSua0A%21q
zMB~8G)j~tzs?^cPJo6}+A%!%5`cN^rWxr`(7wPLxC*-oYEo9efQ+!AXXfn8oWURz3
zq2YW<B3~kq0SxygSTEYQh}hVxKdhWie|Tlg5H%-Obw7kK>b3hskHvAyRl@D^{3)7E
zbau@!Ad-l!62GRp(TJ~{iR99BPXf*QD}@V3!m_+QtfRiCZ8quH8V)e(bSC;hTsqn7
z=NVS#E&_^3(>6_I&Yi*Oqu!AN+acY?znozyM==dubuQc?DJrDhmpan-P5$Q7`JWim
zkXmf&Pmpkm+Q`3Wn@5&SaTJOSHMRbWXZ1KHrOtMX?L;o^pw)Gu<Q$abkrl7pOq`M>
z>iv+RMQrkuTVck@qsfXbGbqf!w^#sgV6&r%S<(H+f*+Tjm9xP1mNg*Eh6gJB6-go>
z<S}ssZ8OR2o?W0ADm%xxT6>1sCJ}}KIsV8-=l_Gq^S5Y*%UqZ#e-O3*nG|d2JGU*$
z+T3mR7@2I?IaY>!4dWxM-t;7QE^+Q(GI5dC40F0cVLk=q;rAnJHCn;boBKuz?;&<-
z6;T7}c4BAe%YM=9e0>hy#3oDd;2xTVZ_9msbF0AGe1$yAxHNairc)Z|-kJBRlCl6O
z^T|CKSE2ZbCWA&u(Xt;ycZSV*i2_?#(R4vWE4|Z%;f3Nlq(fn6E%<IU!niRS2rXr{
z7pf>9J*^-Y_S#QYHu0bY-dg|(!M6uyY;YHh($S88101RDtIPsB*atxYy6~kuWc}Ko
zjsQctZ8_iP>5p2viyLmX)@u*eivsm^^d?hWa=5i}^5pykwloPs$z!noSyAu1GTOl^
z`z%G*sUl?>>Azy0hk^Q+${Hx(dBaVh(Xp^@nPPH=Bin=e*r)|xc;3GZb^w{Fs2g2^
zOTH`oo%;HpGsJA*Udn2^z-dldu}k?7?HN6bm60ViR?VAK1V7F}-Pl(Tr2Qh?pogM5
ztGW2qxpAa#JGw`aeu!{^b2otpidI|Kh|R*LZ>VOVYw?P0)S&d7M0j9+9o@%8&ft?-
zhY<NsVh?)8b5-dO{3vm7aG3G$@iJVCZ`^ioknla3124E28YL-pU>qwC7?iKho7s$l
z4487hgyI5O!f9W(BfVZD)G|vv)4V{je>=Du%=vM2<I8@j&M95`EE3>fbf>(O$>k+s
z_25VNmq6jiGdj}qlsxhP?y7an7?<rg;|&}Kc>Uuf6z9%uuS=(bG~Oa@DG9yDs39-y
zB<0!wF<dTUlO4`seP?FTlcMzxHiz|xO^|2uQBTrki*7?6ie<}kxFz2eN9y9jbrtfo
z0W*752hvCAbsE6HD8+u1e#$mdzne}?cD=K11O{T4EpXa2;%rNMne5e=TEl4ia(Qoc
zv(=5>^11!v%2?qGr0``<zg*osTd15|$QSDAfRA&Q5Ok%jX}ZdF!HlpUX1CwrJm-v=
zFR`w41RH66ve{<BK%eu&2VUZ^#K=1QzTL!}c1mQ+c_NQw#?ebXd1Y|T(|}_-syp}p
zU=(3Xvy!tyR1e&;+Yd~<KEx@x%FRHMvML+^f(o;3fH1T4={wuyHRO-wA@?ef;SX1d
zrg=F#4w8)56C3GBobc6J1R^%MsP^V86$U~t|Iaq?2+Pm{FEw_-*7<7cte-32RHWT<
zljjF0O&sBo)eZ0l&(cTwI4609mb-OEJEqpS%=YiNnS?5|1f<rGaZDgA1>1&4=T8Cn
zN5lHFXC5s9YoBF9fMs`2u4YNf87R-op&Px}ltqI2gtS;RonnNG5yn}&UXubgps>R5
zEe8s{WBNnyLli!{W7%E`i@uVscdz`m(;@Ms<G0$(_%g}zq+IRypS#2MNSIHgaM;A8
zw{@B0Q@Mhn4&!&(uY<^S?2{rQ;4<j=;T%`gUW7uWZBF5RgZNT-pD3*VohWlij`@sL
z9PHd|^e;s3QBw)@t1z>qbjS8HzQ>J2z4XL>_ck!zUi8b4-Q`B}ibqk+0nz5e2v*c0
zZ##u65P(r**!^3EF}5pK1@%=mvv7@th_g5+xf-?${v~EC?j+LHoDi09c6<89#xchj
zcgX7G#_3rVRIP>s-s*ADy{dg+@&6#Qw2@z4$?k@Y|8eeQ6G*^b`143sNF6OP3-eZH
zvNK+r<<)-rYsyoS*oZJ;Y;#I-%omzDi4u%cwJEw6Qj(;c9bn!V+&~0#KgZh+b6IF;
z<lovdCRZIVnj3D5U_n`Sz3~^`C5umUcNxZ=LQ5wJmx$KAaJ(Ix=YNQe=Dlrmf8QNf
zim}2O$)j9bqEF7Sa4&ljJ7;35^V}sVCeNgHk}CCr>;=tFn)hkQVO+yi5A`?&3ARCw
zM2iajOOwpuO^WT-f1M{-hJGVBMhs|)y@N*@bCLvQ-bD`OBL+$Ysw_JX(5p5@mX$-j
zPD4*LY0T%xJTWuODX}w049hs?a-oIQo>+Z`C`ehawVWl-oZXPeK{X;T$Y1K(Xjo^@
zzslcNsJ<|Ydh4xGa4au0cK`B|uxZ77`4mdgU+L%$L1e~wD$JJeeSU%jRC0%4hH?We
zu@Jg%j!hc{lX-ho8z0ULENdy>HIj2S!kj@H=|>ROijivtl<ix<&U&5wO@n8EU;A6;
zw;R_-K$A|%g<SvF*=h=FB$Ky($Y~k@UduaW*urmGh8%I|i<;G%^;~`}sD%_l7v+!f
zk^U@S_TkHV)av$@H}zdsT@GVc+<x|@d#XC8);*|G5Z>kCZfjp%9%k5ldc4!n$QI^D
zDCMMmIT%lD!>~Cw6*g+*7ge|E9iI=A8%l=Ilb`=KcqwO>XBF5pCOVjUYe?~%In2Im
z>?INQh12@>l2-j6pn5Ega!anjMd(cAi;!sCm`P$jHCx<-<SMED2}?pO-f17B9o8&0
zJRZspfnxLN46<gUO)D?>+22k(de^V$L~OaGjZeCh{*fR$4Oq2%Y!s5V&OqSf(@EQ7
z6hk^3yB15!>rEDXWb$H3o{TsvgYc%odV9X12GzYOj|FC;WMfKl$bczHTJ)9UcdH|k
z{XU#~A6)}$6m3!ihpKj-5_>u{#!#(rr0I*$8OK@ToGm_V!u@cJ5G*W4WLairEN(kz
zkw4pmF?MSlf;zFl!EHmC#7ptyqD<!os7f`88NZ<t!B0{J&zSOM0%S^=Wo!W=T_W2c
zMh9<`KQ5;ILBTBK9kCl1jfBRA+Am9u7$>Y|sAPMl&4B<zUcrdH25NQzgWuuKiSlQo
zW5t?$H?Lp*P3dg_8dw3br>R+H!n>tvx#AIGm&Se_a&~bZH=o&k&`}?f-tf9qBA##)
zT5UVD<mgz3ppS6106B~ex#+uplP!=N9B^d$3~g!B6Kt^nzsUfdRlGbT&j6pFE0l5_
z&f4{^a<c!^<VDM^B7|IOGgWVDEY<<m>qE3PF>7~{*Y#Q!ivnCst^fMC(V4<;X|%dp
zu%Y;cj|@Y$b4AzvxyqPA1x(;}wB`p(;ak3c%i*4*RD7debE(-fHcCEuz>o>3n=h8<
zm;V^3w=NDX-qlk$xeo~f;*X5396?k(?67`dKm$vPwGe~%dq_abCPG-X*k=oZ8I}vx
zlz3X?d0}lun^;ff^qLA%uhKY^K=-rL)@Ix{lTiEaZrAYQLbO<N+y4$t!p*`8`rmC-
z0H^X#oK@M63E%H;4}pokMO061_t{qwzH>N-C<{Ha_Zm)zX_TUpdUVL3?GLVRZRJgb
zZtzcxU>Y?Jy7JLrI+{4GrCLW}GZ+$LWcAZ;g}G1&K%DUg{oZ3P#E*Uf2vm&aT=&c7
z1ZJVKt!jq!#pBQ~7Eu1f`HY?g3YZh{%j&4ekc^p`%3$R&#0wzJUxCE%ZWLN1;(N!H
zuKx^(uVHQ7QjZDbk3DpCQKGZZ4nh7M=yBsw-Yti?+E*JdjMrY(%iTQ$L3Wrzvh%md
z?Uw{;ZIyJJA&KV6&fuk#g(RUMX<)=d*262svHKs!qQGG4mub!pO-Gq-Wl(y4gnh}{
z<&6=cP?p=W3;jmJj;{^OzM@@i4aQq|MvwSTV{7_wfpE0bzi4*l-V{D|bNdah3Y9x$
zA8hkWf|J=M$f>yA{wi~}&#eQn38*t`4#6Dcu0&6@LwIj^v+<+iC!t5Xgtgypr)*8c
zd`VM=UEi$T<-c%MlfQLW@&W$EQia_{iaVxoWK{!TPQ$7veLj4HLx^16G~*q{Gdwsw
zM40qprv##6Pd&o!TdKDR7?tDgMI>eV{&Dl6gBst|V7K;~B9KI`TI>_!-jhQT(k!PT
zF(ou?`@5u3WWghFuR#VDD*0`~n<^TaPpDXUTO#%Bq-q|LNFv686o1$)s6s>wLNs*6
zNTKX+kT(qF{}t&{rbn=FqKa(?T{&rgfyMmVpm6FT$zpq!YJOS75z1|mH$AUqVB4V<
z*?znAt^ZVVIP_Z<r`iP7N>`S>6eZERh^{l>Tk>mUakp;n6_p~4e?ur>y97rqq{gp&
zvV@DjGO!HR?n9KjxqV(Sceact#l4Xv>JJadQANP3nOgJuj2fLZyGE^)*}1MnLR2T3
zx3VbKH$S8JeOmi2TN%$IWQ$QjE5)G8!-}f4^dG*BjLj`fJ0^&HijBR58;|1(H|xn0
zmfN?`Ar?f2QdR?^mJ&_)-O8%AOqTy+`gKv(P7eI8zIp`@_S^ys5Y4#N)r<?_5BoT-
z@D&MbTq{|kzlhKWaZXq6^?94>*a@lnlI*CN{%YBix%rHT(>*tVcL-bjq}yik*(LHX
zD6vEZqFi?y%k}VTTU*jBPJtAsC6~&f%M6vZFGt`%dW6ZQ*T{RCgRR!QgV6XR%*=Vz
zkM542fO}Vo-#agfoc(IrulcGPZB@wziu|zNf&=^LK15$Vb|g<AdyWnnEH}vcLnAO^
zlN^{{=}*RB^pz18Y+w?oO52qVbffJw!s+lJy|Ecv>VA?Utd|tq)x@Nh-|7wju<IOX
zOMbwg?<j6(+D0Rw{eZ?b_=<i&DwUqa<;SR21;y{9nel{EDoyxLKh`L|&iAP0rIhP3
zQ%r9+nucj2uR<iKUb6(^bSsYo%<yk7$EV`1cnlk9=M#dZxz=DCUS#)?+5JmFW8ae3
z$!7f)|ISK`lv$?)?2-oV<#Qs;t*D7d$xSEBMBoQI_;E8eLSWPKAF6r#w_$J0ci#28
z^n^Z^aNcws(*LwGjM&G6=Xqgs-1Ic9_<A5-%>>Ju*dpB1#n7Zm`Br@t)f?VxBvZS_
z=MPNpoUQ!8gHw(+e;ZuIm%6?Q5zB{kouSxir#AQY`1A{h^kv-ex{?x8vQ#e3j%k8b
z*YA<;L3}nt7e)J_C~R3D`Z!1Gc%O{lC>Li}lR2HI{5qCoxKZy)zVG{uZbCQY_{C|2
zx-X$QEBsEAPY8!5zlvM^aj<D5XKUUmBKYi(q;$%43!=@|dIh9JJ0K*rJLZgP?Pav@
zi;zgruw)PNZm0o2d@;t8BDh?sjD4Mi+#LxpNt^O0EV@g?cc$wFwqO{&m}s(L^-V>a
zQ0R3_YO=!omEZgcr7RizO``bksW(O(HDf;nC-~Nd2k8@;$1*rSk9{?Pnb2kZ(jbPy
z#&y$+VY_fOBYHyV3d?eIj{TBjFDmy~ILn_%XZ+JAK_2~3V!T#a98U}(8LE-j;NT+<
zqz+M!0a(0+r_>Fu_p5p?ILtk$l%iPo+ZFbdg`_FQu?EUcYm5Hr(XvrQXuT9*<O^rJ
zYX`NPHD||iy?ytX{i=`7)5F4+cF9zEz<*4Zr+3A@m#D6X^Nu0xrw@D{v}Mk$XS|O9
zu-O*13<$8l5YMNgxtYRZJ9fI{5553+jb=A>4=#ry_vxAE4{gKcTzT53Y2zxJw7C=@
zCGp0a%d^VSqP@B{RFL|o`tEz%8>iui)p{FUnBvwq0=Vs{wRNDXi_36GcD>W6ZbQN9
zpkv-}cKVB0#ZMGd!R2}EV9c+0q<K?~-gAKX|G?*sI^$Avw)1b!RrNo}W*K$$+IyMs
zMg2Y!eqZ{x0yDr=p1o%OEmHh^s@PH35D22;39b^wiJ(XZZZ6`y{2n<Wn4H%@>T~b0
z@EwGtL}+yO7~nF4PpuX(f>A=)H|dI;Bq)C$>MOhRiVN+$(Gxn@?>1F-QrjtQgp>lT
zEg2x!2M2_+*T+#4;&zvLO5=_Xt2hzc>()HHnd8R$B&^p5tg=!D@9VonJ)9W=O|^I?
zWd!mR)mZbz#|*})b9pQRAm29D$Bz0pAn)VO>MYOb+C_fmx_?9A81hyw;cBdnZJ8n@
zG`rl0Y_eSo$!^Iz*<w6k$v(OIr~%Y4S4QkB)?XGTMn-t6Ce(U9_viYzl=-W&!Iw|8
z7NOU=B#$S>*<8W(_S_k6#l5n^duYW1`y{jjG%ii4OF@L$PWxwQIrdS!F7n3S7)rRt
zM36&Igo5(_6`=WF3wEMsp08<Ndc!UL#q!pPo%KN@kjGGmYsnu-8rgPHW8HQV2VQV@
z(n~AZS9WUZ6x+Ma7loiaHLopfK%a11y|FM2CH_)$&g5k3xehSlkHn56;}CEM{Gr7{
zGCZKeX17)$OmAevgNqfKgv)$>BV=aQ8x^B#+KclIEY3Wjy@DpkO6=8DanP7E^Z~U3
zMQ-A~vUl>FCjFSE*3Y?{FonTBbhH&7A52yEX5?*-_^8hK=Lvpn^1NT|pOraA`e54v
zxMb#?|9&*`Xhyqm?<2)qs5=HZC)d(-M;AeqFsnZbcutnA!o5^zY8Fw=8e66{(p_uw
zbCJ_lj9@<(#mbR;8+=H>fD3G*#CsNBaGU4D<wa*Cx$jL2cZo1;1~aUV8@O`yx1pvF
zVP~_K(<fy5&$!4(%oDUh%N19ZccCa7*ql;YcW>P<bf53!-+c56p-|#Fh_*GeVM}*9
z;;_v)N@)4mbzY#MG;(1=3IQx8jO7gI-X-Rom_${Fq8Wd(1NemR(!<;xNr|;t{u}oe
z<KP)*$t<@Ifw)e0XVJbY71P?hb_w^i`BB=`RzF!F%sR|CTkQ3y0Gz#yZnvMkyt_;P
z$$NQpqU+RbJvs(EG`FPC#>TGb9EVGrC{T&>cP&jddvWO2@d(OkiQh*c3q@;n7%C`5
z`t(Z(9mJssT;$rxH)WPTy^b>CRLcB7G2ymKkNfbxD536{z9pB5HzBM0mz@R9y?Dr*
zo(kHytU;YgOyZ+zM#UW|^!<KU)3}XzG$u?>U`NB3H$R|7KNBRl4`ZZ1H_oQYP@bX@
zPhJTS@>R)%{MwW5@a6VZP|7=46TbwF+G#D6tcuP4t|tU@)z#<!v7O?ubWab-VdGOF
zAPlVO08EJpq2qKD0($;CyOh&i8gj5N80yTWUE&L;Z%Rr^W_FU5?O-4$=!?c^-GO=2
zh3f3|hyS^uJ`tx_!7S4{5K~uv)u_2@CruBxmE{HyHR|NeHS4RD7WY#yZdoAReiCb&
z{DaB+0lw@x7u`w>U7aGdeXm7t>%i%-@pz0jog>S&wZNIIwjo7P{<)Rie>|G@GVVhs
z9)GgCcsTOgu>Dj2-q7<Mwp=(9U<(&K7wy;OSA~1<>e&r#jW-ZIZm*fH9c-;6XGah9
zQJJqwxm`R={8y`S$u_S30z;5;2CuK}>>afvZq)jUi~(lde>4kTX64f-6o;<hWD*7$
zSxEE8Qd{2Z;;xw5B*s1nNtC``I(}6Q^1~`V){mV4Y-^$cF`eEI`*dH{3>p851mOzy
z#L8O6DTfpbWLA27aGR&9b=hYp*<b)998tO;0X&P;JolqGoPInDL_~o38MUP!nLces
z83@y;jb9slgn%91ej^H2I~%8wZmfG9tNNm4d7?6R4{ALeNsRH?k~S^da0l`gRvF$E
zkc}d@jSDsM;xm<#Xq@DcSb3hB`TpKnQ>4}1+HxyHzreJEJeu0^Bh(a+6>}v;OucWQ
znSBBrvn%t)L<AwcGyms%VEedo;k3~1{nFg*>jsS#3i5ZTi0BQJFn*&4xY~WNa=o~p
zIvl7;23>Y`BOJr)KPun+XRk;4Uy&#EpdgcryPG3Lo2)cH{;B!(jD?wJO#LnQ0Or&{
zj+nWTPeHioY1WQj1Y91i456l!d|Mrc+T0DJyM-D_(I{y|zjWhHB9ciamDY*aA+3P7
zf?L&Jvt&{(?#d&@Rn-t-VBs4Ol%iAZE}wf-UF?Ie)T0TojIAY|t5t-Ef>>#aQIC_^
zxqF9s<@5d_Ndyxi&gd~KcQ&JIZTN-=x;iP+Mc+fFOUC+}BNCCpcs4(Qw&zIJsKn64
z8oi5NffH`IFVP$?E(+b-WCA^?R!7(U8s%M8mu!J?uKlc9L854luHslKmYGx-k#DPi
z0-l-*Z7ugd%Zn$ma-e1s*C%0s31Ne|J37?vdbFSv?}%>{8sjuT!*kF~x(=Z}RyTft
z(j2mCxxeho`a6KJd~TqB-@0u4cbOY464!qNFk%6Rjp4iBo18ji>|qLMK3lZsE%uGG
zjq(=R#J}+^o%`6`tOTNOo0S!pmGoM)a8G_Q*L5trAsU3G1%{{+wc#R!v|cY$*k{qo
zgg@vXqCD@k^gaHw;#cPQH(?9g5Zgh)b&*KgdnaU=^{hxU=?N#rvv~8q$D6;MHC(I*
z`huIeZnx-|#v8t)INQZc-|nnQ`Pga~BBChDzI2_n(6dVWFmJ`wuAC}QOc=q!u))us
zngCoLksYjfUhpryJ&UUoPBLC&96=X}=Jb!>vuj(1+M&;fkd_wBXjfSuWAp`rgLgVx
z)nvwJCc5*~ebcQ~Cq-e(|7p-lcX`jpB2jtCW}gUM;eIp3H0U>t`(AaT-I1J{*2fM~
zke812Mrs3dm&A!+4sJ0d>sH6sQ_Q}Ns3CC<wm<;-D7z3*^_?#_So1ZhyGK7ETeKR8
z98<H%v=WqsCGuO2D>z3(@ZRSA1vM~vN)(TqIdxq2FH*Js{^nm&6bEB9vp*F2aQeFt
zsARGLk@Ig0#|XCb!ty^~XNnKepaCdoi?zs;@EJR-U*~r&I8w_NyJ$cn4?h$h?pFt{
z4&yFeN)-E^l`v5NmT<?1!nVz=Z34QyJmcO6w#qzIUdJHb-aML**9RH!3U2NauUcGZ
zp*!u*BPQislT>qY^=)WwGiEUnRYq)Z8-3}={BhT>f0Y4ttUf+3&-?ap_*nasue$kv
zh}Nvo(M_Vuz<)<80s(!EJ!}dyG<I;d<@P>aV<w%0-?ukf67Mh0E9W&Dtk7WZcUUic
zIw?{=`wLBW{L}8Eh%k{NGH#reNGd#AJW|YoLf%&(#gLZk$ZTD5&r*8Vuxl%kE%uNo
zV~|yw#y$~Jb1$h&*GasCrzGJ8kH_3RO2*R<Db(9nh9;Hv)%iaDSUeHmN&Ds2);FAt
z{Ufjj!|hvNf9l0@&mwZ_-(Q!C(>tQz2`k3^<*_c-3f;lo!7Dv?x(|He6Qm`UO;QB?
zh#IeR@GHS4%E4X71rL_o7!lFPEqlIARnzm%@YP1g)Y>(92jOPE&W34`DwtbF=+6Fy
z1{PxGzcM+e3N8&Xw+i2T%B`0iJ3_6(EuOYnzM?A7P@&5GNh82~EYi+wedlfdnfVl8
z<!2_ZeNk_|GY1n(@0~fB$r~CY8%Hu;HG(yve9Zg>8`>+GxlfKRAsF-J`76f44UYQD
zhi{0m5$a+$ZuH?VKXc3=ah~ImC{J#mrVDc3-Ym@tSkfBe+Y{{FGEchRj+V%lQ;2PQ
zeECo(^7>AVAnZ7qPF_#BYGK;0N%t#suKF15V*}+#%2$a!aT6E9$M33S0bh;T+~gSM
zzLS%8sd=9FW?H4zQwV|{8~Sq&+l#S4K-)p#9*&11dpX59>!<0-C)u9T69E0%yc=uK
zF0;X<qebbk@rFBvFRtSD<{!q2KC?!{EBW%g!g?0L(2A?iMs2e=@=Q8|PCh<u%#*Ps
zU<9$@lkicH&CK$drrcRV>gyrxn7&5uvymbqf6m<sSmf9eA#L}<Vl}1nO-#Aq=VU6S
z5adg46-_j~2>QNbBe`-^MW>5IZAiiza&zoVu~(2{VhP{sr%FjNdi&n;X5HAMl0yUq
zMmiUat4<21RKh5#)tM&dQ(+_xN9);d=LHx(Jh*GOc)x^TwT#ORIEDeGP`I))wdD;A
zs+Cl3aW@I;jUpOmYa~_Y^V@{XahTf2q-px`bFoZauN@{2%6FkOsA>X4z*78CFYHc5
zwb?0-=CUF{wEx&LJLF4u)2gn_^~me4C3EPaw#BXsLA_8LwAm@H{m!dhPj}&-z~>1X
zDL!{#%l$B_NU}E8le$n)H*qw$4%HH5)e^vS`7XFqDI3NPX5+i}%yx|7S$h&??t+qY
z&CBN%&sE1krcjvAzY{w*^E0Xq_q{hgzX8o%7y&}K);D86F{Z*s+3n*Ebnw6U_(74%
zSh6cL2yhe@@u2P?XmcwoZ`j&=q*}K)akTG4b+bFs8s%mogLgZ)>Ak!@&Nm#l<e5Y7
zI#1Yf)t;_L6GmLirIS$YS(qyNJ_ty8^YX;<ti1Q6{Ic}2=<{E4M<PQ~$A3Eg$4Ngs
zxI&v^_xebK_w63*b4C&E>TTamOh(GJQS)L#?(opo+pQYWSykV3wL`p(X?pqnY0j8A
zdPAMntYN!!e2uDij8|la|CC4VUu!Y$FJEgB6+~p{lb?254-sJBciewo_}~%eu9SS0
z5^qOP=2h=^>P+}m6Z%zuSa(-wwJ1rSyz6#AIrA&DrpAbJW;$1C;#|DR!4za6>lq}<
zYJ<c}UI;I!p&=j<8?v_Y8lV5iP>hIs`;GiSZX3pxgVJ1o`@N96XJTQz{>NsZJTZXy
z?Nk=0?e;-H_n9+wML8QSx5VAVio4JihE#Uy>X$bA3-qhJtg&OizhgZDOX#jz_ZePU
znLh;|my>4Strj2meKg*Bi2Yc6tS?QI72XYwNw2$8=oDxKJ?#wjZk(KMNBIcgc(9tU
z%@cLqBwLegaSVoib#`+<=}j5=YcMqd94+<69twDya_RnxFn_k<<Kzb(BJ6RYUFd~)
z%TpYdaQpL;Z)I~nw(SXk+iLvGTc%8%`2A40OFYejqPSYkvOd&T_lBrlbH<wsb3~kc
zpL{(A`I)fgl;BZmk~P5xwzG2B8F)2wv(R(cX9DxTo@pTYdZ_~=%+pm#;2(UL(VO;!
z=B2(CdYKb`ac0?gITwDu`yorvtz7Vn`hSN$l*4ahj~qqMGg3<^XwPt);#AmDX}$8`
z8wvLI3`O*|%e}(C5qiwdZVYwDd&NfG%5#fBmZ`x;SRxohqgNgV<$*adGZqn9(9la?
zqQpb4#L5b2jjS!P#dhyOJUR9!brWjM@VMB?3wc^yj1ahhP8XwI7oFY%Z(9`zp<L*n
z!o%0?L5;4s+yPk|)?t5Yagzyo4b-7k%#&VwT`0>;8v~YLQO3=h{3l<V=I}Z^yPJ_Y
zgJuy#e*Ozn!!kF$&)+r|154@t{SY`a<tZgA%^;*O3?0lYP0oB3y47sZ-z^J((?|=x
z9QFaR%dtOKg26Z8KxhpH<C11n>ea}->i`2!X97n8K!<$v0`K&)Ae<(pyp<D1gFK7$
zCGLQ(h*7ZETaU1>Ooi(q^Bi>dl*WHaBJ4e9U&@G6QD8L}7468;HlE8t^CZ*2tGvO;
z)c1W+2(A*9BT%%SXga*5x{sac`bCF1*?~ISd-Aah1nn+z+vHHyre6wmINxz1EZDP^
z;|Un^=`2~pR6fYn-7wh_PfU8%MV)r_rF%QP{86f=%>GmIWk-7V&Sk=N^QGSm>E|Xg
z;hB#cbxxq&T^v@P3x=ouNc|n`;Ipogt2gq1E?(efQ%h^3PMi0|XXV8=zV(6q4QFd0
zPbF@2JRI+g^#t<b8+%pH)WSwMF~69eK7Kwnay&n*qTr*%UwEnhVIo+R_S2#TN{!Uo
zCMY$qch|nh{j0U6oH(z@7gHiB8}J|bbbWmR-%}~z_deLw!^qOZ3E4f5>cJYv*5Co7
zY(St#u>9+7ggK}iQev*0t>!rCL3Wbvex@MJd4hfiEY%b6l5#6~gdJ=hKJ`VrLa&@D
zRMX0o0&TJ5gTWpblkJol)Mcoa?G<|M_1c(@CB{t7B2@NoLuaXnkemS3Qd266zmmJd
zZpS$~@9q_^9~=^d(cHg$5orCGGHKp^Vf=(HPJddY-{z7Hxo=ElpVnQ%O-?>Ych#+U
zv>j?;0~4UYu!E1{NQimt(0iDuzRnhM$dzh~Zw5!xV>y3e|1h@kWx<QLB4#I2l;9vF
z`h0@Aar{PGjEJm2G|-3E{UM<@@Y9-2MYlM={=M5)pocS3Yq~;rzN&eDXE#Rxg)Yxq
z#Hv~Oy{1y7ZW90#s$D-G&H>V+hVKPIjOJ7F%ZzXRgGq0>bcW$G|I5Dc1Na9ts<Sf*
z$l+iq$9Gl*PfpAI&^nPac@HbY|3tm~0a%2LkIiwY%foj;@Tfk#d7Fp40vo{5TBOlL
z^M3iw05q)Yd}v_`0_+f;pP53$g*LR}Io_{WzH#G`Z?BA;dM8~2Ra#>(YVD_WLJXCb
z4iI$5tY?GCclIz5+UR&Rfm)vcw$HGAyyQDtmGzlzKSqLAT;{yCCS|6j7m<n)K1pN3
z8Q0?TS1a6$+-{sHJu-d^+hn-b3H%lx)QLLAEZ-B9?qB|!-RYp^>*(vmcyd-;>T_VE
zdXZT2pNyWQ6Sf+<c^f(?!ZY{PktO!2(cY1=3K8x0`z}L6CAJe?d4GT&n{@kC5(nw!
zKy1QX-Qp6qMn+Bay&RfF!UKL8Gcf6Vvwj(Lo3F6X^LoqD<9g1mf6It=!Qa|RV@_X3
zdQg?0{wT;hF|4v-NkMt?y^UtVcb1$U*Qa6@nYHn**eG0_%beJuNb&*N3fYoF<?T%#
z?+Y$7k2q>4F5e2<i-Ts@GHIb!Dgvv27NfRA_$A4=&z?kU^1SkgrMRk`9!V|=plNXb
zw-yy5=giL}#6PlCp0;6rZ={_V_2Co$a9B2xa>`ad4YAZa-f|MAn=?B1ld^|$b4UU+
z`l<15ikl-6ZuLzfA97ZL@)IPV3iEMqfY)kzbp!{je}bTX(<z&m>b3r2hp^!2KA7y!
zIMHVJ_02sj!;@mJPL-rWR-E+-^Cy&*TL})HbL1?i4t$kHlA~nn(|&AWfiMNKKeoSz
zyaCEHS|107EQK%r^7Bi`dC_0@isr0&)PWRF{5{1v|LX5LEQK^Ze`09+fmPixqS(r%
z{O}_kS%Q6tKx=deR8-Bm@rD2~&qO1Bs)L0U{2SWFFB&RYbSV<Ej-A}I82cBjHv%qy
zKzXufro^=wkd4QA+xOSQFVmC^f*4A0$U_{ax3BFRfBYh78juhX4MImR^l)D3%3#lu
z)Wa8g`@vy<5@BSl(&1)Xm1@dovG~ZN#19To9v3mO5Z5>v2NBYNtR8dGnOi|WhfF<J
z!onqt!YPH_o6>?vpaYqH!b|>19-Zl+K^zsE)~5bhDqh`SOW_1jSpME;y6nrzqA>2r
zKG_6Los0E4w!|zOp$qN&gTT{(&SX9Dv~kYIWM@Vc(0P6P$BJB{qv`Cl^GkuhAyCy-
zEk71xyi#Y62}(u2Pmvlr!oCmFecXxCm6sKztE;{qXV)^SODYF+EQSY49<k)kA(Pw5
z)jKzmntbYHDtd!7iLb6kZt<4Yjck&R`_TaB{jY(%4$eaPYgEJ$Tcx`2T^i^7VArL@
z8(bGhhP(%}cxJy?J2@Y5aUT^^O6KNm^Ut{0=3&!Ib!BV3!(S}lDD33vTC5ZacpKNL
zzkt$aeLH?L&K5ZJCp3^0s6rcWNK8E*whs>dN2&@QuMr5!R*|yxz`9(;->(;XC^n&f
zCG{d)J5opO4;uvq_N#Fo)*;zP9PO{hTH)6N@%ipE_`G$Gu=^S!u#g&tYnc{V^(u!@
zRYP%Kqt^3RZjl`ENZ*e+c)ugtz>l-`)X4D-C|y=SYI4WH@c7jfoS_SvPq{#z^HAV^
zcYs@1R!-tj!Vp8XmDjK6i=hY7z$G+BP_u<<dA=NrIZ2s-ZZo>$R%a2ZgILPVL1G>M
zvhvIBnx0@~6-1+CV|A%}T6ye9=pbxitR-DDy!~;f`w$2G6c{Z^J({?Z1ahdqrwMKd
zd4zv{uqA&s5Xh#=nyas_$^7g5I%KXzcd3!SRRIrY_k3`x9?<)%&rKW+GCYl=bdn5$
z<&3ms!xe^zv!5ma#`Z+Ue3kvd-u;~?<U7t;FaLJy+Tlal8Z6+mf_9o{K5_NzDOG(h
z!{*&Yn|0q>pPR2oC1<t~zgen_f<_}VrNQ1$j;?3VJ%_tbkeMw=ziYkcb@x;Xqr&MU
z7x~o7Z}9~WrnPh&4LAP%uD;>JjL!YZSR?4K=b+&T#?|anh4GGw3gM<4dXshjASh5i
zaMhSU4{qX2;0g16S$NstdFgpU?C`r6XykjudER=teq4iJ(ijb$g`nvP__TV$4V-fq
zk!)mtZ<d!$7^=Zw_u<?r_~X<~(kQ}0zSW)vxR=b!kuwwa`t974mI(`}@xf*V{)P5%
znJX@M884b(1^!Tq*JoX04j!YkQ$-S})kH|dut>xT)+`sRYLxzyfpWpu-#8BTHnBx0
zHs5MJ;`F>#WDZDkYeHQ^1Z&$Da=OmIpq=;#ksKzX#E=Rji#RH#HJ5kE;>?LNQ>|%6
zxkjA;AKNG$bb5VQ`Ny|b2R4^=qRz|t@}0k+fdnr7@lmmT1jtFfniuc|XX(t#KvIQ5
zn>Xdh-9F$yS%ZmFL$06Nid(I43IMt<t6~vxjJbt$#QEDE>Nf6vs>-si)Xm-7mHs~(
zXg9O$CC&@@mCX|oQ6t)0#sfbxJt>j-NGqF$^BEL{(*N|~0I4HDg4?-1(t^{?-(UQ-
z$&*f-^JBm<`sElE3PbmPdp8*xYLj`3bgWIM8QK8*dzMn{H6YKKPi4_N;R5X1<j(US
zuQo@AUNOw!u;{}!qD*fZ&q}iOQ={tMa$6K5$R)M8+_GC-P%Y5;HqT80Z|`O|GAi`R
z{oik8j>DM`OrNvD3i^Nol|?++=SrrD(*XldqjvCdh~=Q5+5evfAZkW=+A27<*x+9R
zRZ|Xg4MlS9!)yi^^TvrbDgk65I2m&I;*pmWUW(FBT0FT2H(NS^41_rvRK<PjP6s(4
z;>ysNA`}8ZzW|Ft;t>>slI%S}CMV%%(f*<E*CM_8`yqJLYY8}sWK9T&6L=_bBl^8}
zbYqyH7Taar)jpl&?5sTHw;~tB-gx(1zx3s%NySa`ucE(vrb>M^Q#`ueU%6Zo<}jSl
zmN<eifU$7X^DDj`WaCh)mqF`8`PzBN?e@)87Hx%O-T(yc37EHFp56_r8&i`8<q-sv
zGfGkgOx8R!ZLZ(@vpJ~5;3*ac=V(Wk6}Hkm`I8&_<5F0q_%9K-89e$&J8UyX^LWSd
z=TE{u7zz8>cx^~uoO*5y?{r@fq<6O){ovZG{+S{a@N4m!;(0Q3v2fBEm{M3zn^KL<
zpbOPVl#B_!`8$*d(DM7fq6SS25UvZf!jciAQQj_T`XguY25TaCyz%=t^3y!?S-J0d
z$O{rR)3i6|#|K~ekJnn?gKl=LZcd*b|7OYQM3dm&U=GafB(v|h=2bBk&kZC9$Zk7C
zg&aS=EvZ#m+XP-pwjo}!o8ymmugZo@d;-~jK6z>1kZ;XoiUma)tIw_H`S!^i#A(|~
z*rI<n<aqZuuHR912fD$&4saHv)!X5b7cwB!?~N!q-=QF$gIsCeE3&@aGD&EYvOjf|
z9Z@ZX@39ciRcDtEhZ><5@eHB}?0idYJ&UodE@_p3@K^+qv%=^7-BFP3N*%9g-7TiD
zv&XW^1#4+iqRmP@Uvd`Cy8{D1sNHPX8j6GYslYIUn}5CP;_oc)COS*F?QNc`d527r
z067hSrjYvgIi!l7@;^3>KG=Per)2Ud(VI@ngcQh^(|EcH+QeigjQ=yH%gL#(Jtz-T
z?wITzDadZJhgS;Cl&pf+pcjGNngJ9B`$t`?KFwN!KG(nhGFqHRW}+_yrsfa0w*87Q
z`a6m$T!e!F9U=BxjarOo?;pDsVlhO$`t`ee8RzL#qq?o<{6u%}fQmTLK(*;NBA7+?
zG!sxzCONx>LIA-DL}0S~UVXfKXeRQ$wP(Mi?bYqY6vPg1bMuWgv7Vr`;0Ebfe%E?;
zBUNezlkniXHS5h8C&E03eiU+MmS|c$`T)x4Gohy=-+LC)<Mt|hxCpBI?b$Qr1@_ny
zeU#)jwy=SN@1DicGocuDU)V$W^<lZ82M~XbQggT?H}d1(hE=1VMVb!X(O$_BV$Vc&
z2?j@z;~!RK$6`@~{K;)F9cU5u3|r2^msQR{4iq||18G=zrlEwU{Fyc_jVgWGa~=46
z(NHm4@fgt1xMCr~4IYe?C1^-@p)ABS(+3gi!9(P5!~w0AmI*bMQwe!jgRiqPNH)nA
z6K{&5b5>4S)hf>ooE~BXQ>UZY^a1Fh3CUGKBk#GRF2T*%w4RO%`YURUinrr{V4`rJ
zh_9(>I)ZL}DIN2b1D!M;Q+wzwi^6waUg;*pX>NAP<@-_$6$&X{IlO97LXEVuIOIQd
zr<r^0dL7(%Epc_bCFm-EDufMJ_ldVpOZpxK$-Q^ms<B~qgrACi<tv3(tTr)#1pfD*
z7*<pHm;kP%Yw_x9C&jwj#gak#-I8X+d1K<Y;yMl@#35sx+nBg`>X?HKzlCXcHnnq$
z_>IPWY<zT;&sN**mEWi*bm+K~+B}~o3ihLj8*elCaZ5P%zMd>KZ^XWvv>Ta@L#{l9
zlO4~V=14C|&-LwlkSR`4{9~!mzbD*QcFOT^P7paf55j$nh1cddNvNLmYfPx=72pz(
z83@?9mI%e{g`q*u6Y-VB(n*(9WWsn)i+#5CmbqppBL`u_yFahl-2g>kX!?XdTh0NJ
zT{^G8T-A2Q(i|y3S1OWs?>wh^HIY~6DT)M*jUQ`f+*6o5>W4TYG%>mi$VxC%p3==E
zj$X>C5R}4gv>V2=GqT5asn=E=xMKyh)vBzcM1vT+4Abr_t{L-j5gl-|oW!&ZwABdq
z{&-MA_*b-vAyEQBDtvbER{07~#w!5P9c?j@SHrsiM9!w^muty{I%WM=1*nONUfR4C
zKdfuY=px0<%C7fg%NS4SZ{ajx+29+hM>3a=l99`swLD0)t<ttSCO7ijTmjUs79!+l
zzVauryBj?kPBX%nea2|F>`iZvUV$?5q1$7?dfBNj<%&2@o$Ge{9_RGRv#|&i2Wcu$
zJ1zN<4XvpBWt(;<FBTEVVne1^u0OG2jU`-Pp}l-y>GzQSSc-@wgKxB7K*|$Rq>Y6J
z_oh{ht*`96NNYTi+M~h8$hElh3BOlCw&i@Y2aN4W&4t7iJEcX8KQa!Nb%=leCDF%E
znou|()O}OZ8nl%?a?h3bE{W3Hbk8Omu`6RgJrcwP74ULda|~wDgtO*TXfSEp=>+OU
z5jd&`o6)7Qra$!vznG6cuOpIZtZ+wb`l|d)DxXLm>EdY6lFwy)8}GZZe@_~6z(z_M
z=y}n+u(YF(1%2UTjD(-o_1j-O@X%PWsN0_}e$I^Rwb#&wgwq05IsvCohdrv?T@?Lw
zwlS;TFcXv#GMgpgrnScl$t{z|Ik%EBz1nmCcCV5>W%3^(`90^9*r$Cg2;*m8qDE^S
z)9{0Xd*WEmV~S)8;zJH2x|iRDvOI^EYA)rn@2J+=ZQ8^7Kc|sMK%4}Kh?UtKXP5Pv
zZ`M5-j+1nTUWq0;77UqpUQfEHNA*Nt@R6b=UBoQu;3zbZ4j+g}xp>4KJwc+){yzY)
zKu^EGkL%V|x&?;F-Y2K%4e^mJf2vdOoG#_adFEkw5XT9p%qLGUga^BV&@&7_>BbKm
znfLaDC<lpEpSH44KTU1=&D%FNSH^3OSP-$%eDvs%zpop4=r6y+@WZa_mb<)PH5C<I
zuvu)|aJYu-hjFWq7WMS{^_#kmniw8Q8XNkD57syIh{lbd`tU}r{78g8GBM~s<#^X8
zqWT<oO}%g9hVGHy_&K)&(>6!z88V0a?L)Mo&SrQ`dcAFPc$?nlO7EJ(UzvfIH9U31
zA+68hw${>vAGr)|b9fH3uW-{G9tWlSl3+L=9%5RX!|iEzZuG@NbwQ`CO;4G_r;JC#
zOS2uMk9#)e9qWve*Hu~4_@>GDy#a=A)|=&6#sh18h98E)ed|iYFKeBtPV-wx?-bWa
z-8<6*ziY`S6QM4jKYMA0o6Wr>Yx7sHejc>0erRj^r%%Eth|XTW`CaovXjT^wA3XNG
z6?-k#n4isna#if7cBMcb!g&1XiS09Nt!QQcYBxbYezI|lhnrg5h3_4I+{pd(`NQf%
z<Aa*qRI-AO7$3NORT`wwBj5O03_lyHKZykDZ|(;%G3b}y_!+B=LyX0eaV+n#@Z!_w
zLl?u3zMInrKO~=5(Aj|EpMn@-@Et;P4w(P%U;Qg5^H2WCe_oM0`{O_PGjy;=ceeV;
z4)As3hsots3_m<s{p%k#pLkg}{H&{zSG4@_?D2*%!Vjz7eDJ5p*D>}Rdu-QCUzh<z
z?wAcfKfU0pdE6BH`L(^1)vT<g8)o<kEDjKV{nCtHHbK)4e`5y7A&2fI&m}%y1G@3U
zL?*{T%z$a0`AWrbz_$W8_}%kRRg2IV;yW1^sXZdQwt<ozX|rPZV)sdURR-;?>2?}d
z&NrTtPTbK0wBhtqdd*uK9bWnLIlT*3czGdsH!1C3vgUaTgNJK_QJKlld4R?tzMf!0
z#^3pC0xX9*HoiD?lUVoyDV`e7HAcQ59L0a6XCk^$q%`0at}$p=_iWQ@M&ZO;c`VL;
zd81btaCjN~putA>g+1wozwW-T-nsGT%HffiF@$!T<NJXKiO@175t1&YJNP@U<-*g*
zxQi#ofDgK)OEc&iKKJcuy)b44&L@B8bp5eW=|Vod;fv$>{0IfEV9KX#NyjBLmw~X$
zV?Vq`ho}B<QC?tqpMCu(Bs6(XJRerYy?OV3g{*c($LN?OCKp_OwAYU1?+tfg`+5D&
zKK#0|F$+Uc)J?n+f%%~8z1O1G<>&1fT}l6{w-Oh*RWIb__$6#p9C-8tmHayf@K#ww
zleE0f+xkAQpx||H;Nc{D#H~gx;BEW)Gw-8qbTGE{9q0sy_Fu`KLs&7YcAlqi6K;EV
zJg+C=pqDk8ZpLlHXAEvr*~j2-14~<BJfZHp<RFj!hvW86IX*_t`Y$*d7p(DG1B=@}
zG&gwSl>apuY(0^mBRBl;4)rKUNUXBb7t!fco4+5M(HYm)Z}?C&uY2Q(a|TZITedyM
zJ;pmU)<%bK9`IB#WN_-o5!>~zJ%yac2RI?)D>DL9vxjoblichp*G<ef7`u_NbUd8H
zLkCTBI1g~}nyN8Q?SWQaZtn9jJExhagRgod%>FR{h&V<c#&5K&{tG@mO#M{t0Umw*
zJC~y^EP9eFhs))CY_B8ItzBMm^vMqVOB@a`o96KH8ydlbHpV)U^;OK_F&D-<#pfww
zKtJL^y<NM!o|<^+8$Xnp13h$;w<}M#y3$xd8&N&7Hln?h`fB2{eVRQGWMd2cMC;3S
znmp{pgOHp3WgRYJY9G`SH-t6D3s&FwQM}3T(`^$@Us~QaL(l6N<sK)87bI!)E9yt<
zE#=|+H-4<`adQ)0`s>$B1MW<J=l9@z+KyX4FJEjLHW*KTdG*^tLw;`fx%%>X_3qv6
z`Fw=mo_q1DjY@7a{IK%+_RZRet&8+sk`3e{p=tf<^O}F-$G(4?np}a_8O{4#Q`_%E
z9*{Yf=1};#;V8;bnq9@t_1Ol;Xuw$V&Azq8H{H4nifuHy((`Zp{GC7gqt)l{H>-zd
z^%+_huJV4iH+=o{_|b~*@GvKJeHRFvkM?>reRD@NcID^a56XGyTSW5nJi>en=vn#(
z5%DHEO0&mJ{kuqi`SY9AAO5s{E0%8@-PoQ+-z|a`?xWr0u_++t3K~;#%m)u`Ui{6^
zc6My*A-m{f-vzSo;{Ei@=YCc#{z^B4^D`ZbgF_f!x`+Oay;0=S=9)+8`$n)p9*k4E
zkwn^9_b>nV|J70cmp}jCD%8Cjet31-%k0(+lgp_XeppET-jCx9KWLg)#r^6<yz;5i
zyahOn`1y@}6p_vI0UR5tzh#Dxd{cVVh98sF-<fFoYI`l)7{d=LmobUK`ZvG4wa!|1
z>H}0sCrj#!f5<v$HP%V(u$fJUA4x>%@^uc2Z$6fIe2!O$*^Ry=Y%X(?tUP_HTQEE&
z*Yt!s2NC&ugP~LEgC~RH6>>Jo7(DO6FC7O6+)62NOSrM|HhN0OM?MZ)<c1s?312w#
z`y3iQx3O;Bh|_gwgbWz>H~?!pY{E!4_rh*?xp`8L2Po=qo?%Sk*3UL>NId|9>uc`;
z6~EfpFuuNi^Va(_>nF@l3}-BY#vK~1tS>!DeJBo%^u%}0cWAh&+)Zgi>21PsaEB2O
zs{HSEO}t6Nm?~o*xBA$WmTi+mDLx9@#0UP+;Wdmw=d5)M8FHZ0U5fqXs(qClW%yx@
zHN>b<;?EIps>?a(&P`*ijwqMoSi8p2zDrlWH~$n2KV=ydr@S0|51qq-$2A$$4mpR1
zRNcS#C(Pm96sPCb{pU32@LK+IEw?#`W9-B5v)PlT>u(CCWR?#Njj_xiZ)^^4dJ=u?
z9L}xNcp#)U{D>!qBR-_q&s_Qt?I}88wW2vZ${r8&Ory0sEN)Hf)l{^n$M%Mma<!)Y
z!5Ar63r{utY^V1zQr56@@YDevKI=<2-q~6e+FyC~ubp3=v8S*Gnooafv9a2+^{r*b
zzTI1q;X@(9_^ivkZ58PbW3+wQ2%P;BZTOK+T0DA@@-0K~CaUFbI`V~&Y`KxYhPFOw
z>qy;NKz?8==e&as4|eIczJD*LeBh%(qRh=QWgvrtlBO)<6dpOP96w)k=Yp}{;sMJw
zhaiTmby3rrXbnC^G`8k>zBe`fALrRe)S+~i%e{m{1GIs;&j(7w&lFwT9o^OCgja@5
z;WZadKJGQq?8|S$iC3qMYr#1m!Z_YJh5O9IAa(=rtKDdPX>Xs%Dd%{=mT}VBoL&Rh
zC2tqKEe$^`dKl!RIDUX(<yGO^gwE43fc)^B+b{LgJJSt6ArJH=6Z(&Xji`v6_Z;D=
zh97wITCpc~yQXv>Kieyr?8&pC&L@59fo2s|5oPuw4=*=5!vOctz8R3asfHiPBOjVx
z7bE4$WhuiLsd(y$*S!7whSvjby0M0zmv$TL5l?yAqULYxb$#BhH_3X6$|j%0Im!q<
z>VvMug3M{A!2}){=ANb3(y{Qmu>t1ohtI34&c#)<5APe$<ll^Sfn3|T((vPlz^Oea
zuOk2D%kkazk#PQvm4Qzq=7yj5_RSqOyiJA(4lAj#Hpgp?cv3X3$zN3mQ91mSjh&Aw
z-fW`ns4aHMlJg|Y(__z`KG_)Kr#&s~J&gh0`vy<ku)(SOXuFO|xhUHd!w+;G*d*Tc
zJte;R!;`GT^%Lcgo}oQO$KdnSxo*%v2RGgDGml(6+0ypTdk!Ml7*cn<zJT%QIz1Id
ze`h=U(hO|5A<GM8Kg~}I+aavIULu~puMdrcv+0j7BV!U9=WYkxMBCuCxmTOu$$k!A
zSg^BxU^75Hm9Gly<`KEbQyXmTX&^KD(gQXPzjiS9){K&T6t-~cI~=&7JL&;L^Rp+L
z?=T-@_^F?U3x!oWG5&3B+;2n}xSo)1@@w(y`XepMLFvQQMnI*6Ki=?@>JxO6txfiI
zF1I7*hZV(*$i9;4Fbp_fYZM5lX85_Ze(Gf@%Ax3(Ih?*lZ$PhhI%uK#;Owr>;pnmM
zQy$}IpgWB@obs`j?{>{|xNd7|JeA)~i`*C=nIkj@*LtqU$$mQP5cwRK&*9wQ)?AX$
zk%C#L(msdR{b_R!m*J-#SL!)H`kkYUp}R!eR5fG{XMKHT4_>rabIc>T9?Cf@&zm!x
z*8v}g=OMB3`=T63zOkFsW84_S&+v9TMuKUuc^eK2RyPE+ZnI$Bj>xF{3-cTMZ`va8
z+>&As4g|N|S>JuH9!i8NIjL1-je|A}k6!0MLF3*9H?}HjPB%CNA(zSct<#RY=rdMM
zn|8jLfr<yKaO$&O7Rr>|Q@a76hY6L2Pdn%B1}1*U0vXC;xJpAmKmGWF#krklZ+Cdb
zcNHyP14Ak}`QU^LLzvSsxZLOan=W(aOAkow#-~K9N*52y=cX6DomVPi=qMs<vL$I$
z9`uw*c@AN|5BsCtG%ekG)6EAntV(u@#=}MG=WdL%MAR%*lr$8d&+tRJYds^s<PopM
zuQf2FZffY^AzB&_{2Zg<r>PWeH))z+PbG~y`1tjD_4qu_xT~ROM+HEKH+CL$X}6xs
z8$;W${LA0GT0Q;Y$?AE->p1)9BGbSA^Z3Tks2e{Vv_F1$-wtgq{JOObDTx<PKz#S_
z)O@`9Vm?pgqv~}BJg<G>rq7e~!J{04&C%7D^#ckHw%>hR|As=~oS7WwtA}>zdH;TV
z)SAJ)mWe@@&i2ioVzqa7nq1XHmouFh-p265<k<9@ILgoM6(-C!$7J%z3^ddMuR3Gm
zV{n&)28)m{W)Q)U#DQe9<NF9_{#~1=j~=)YrG&$%Qp)nyzNJ&9fk&Az96hkQpq*^U
zWwQ==SN1K3PgWPu-p2iuZYNz?-SUm3CO6jb!}q41m@a51zqQxLmF<N(e`w!ndeEbL
z$`<{gP4}g=c16(iM!%p=(Pg;kS1ewrOAZY%PSyDVg<gMS?I!PYv<cb?@_K!+_(}aW
zZzG?K?w8g!lEm1tR~mlW*ij#CuRd$T*=IWXy)2BUZT#y^u3b&F!&b({$LYH#InF!#
z{z==1E~(=mU%oW(Ge1D4edr0~oW>Z#53j4>_7R8f>#tjj{>zOkZYaefW^;^37HB-M
zL$ilw+*X^DJhZ*C4ta`<rx7=u>C+w%;3gFdZpJ=e=o?iz#(3~~71foG@$F-$aFB7T
z;fM0Pu!Vow2kCon{aZEh{h69q9b#jMk&g24T0q8=eC+=J*?Y5HTXG~#Z|$?sJfDcj
zEOkLb=3A(tiv|Qqtp<dG1PBmxp%>Hx=mmtnQ=_}8bfE?Ww4f3?t0FVv%<~XO|7Yg5
zoX$o1u$HfBZ)e2W?u&!bXwJdNpGHE1wv3EfXm>f4WlTCOJnl)xC4dAVFm~q!2OmE2
z-8T6%zEp3a<5nXoWj3Zzw~Q}+dmmF;zL-!X(IdO+lZg%AlSvon{o_!&HO;bjmPn@p
zKLAM&9{8Pm+-GGAtVS*VNMBH$G~GpVE6ZN;7eMln@yUSnwv6>jf!9$0LB6Ti3$LGE
zH#HXXE!D>|nWisJ9qj=ibnH<pk(R}j^r$z0AL`6w0l~Sefq*`~d0Sv7BRtYqgcVUr
zXA1B`eY&Hu4F0%ZBW<&kmv2K}wAbiILeg!^rcShr<Iao5Ny(YMg+7Kl8uZu@K)=L}
z$ad#=T7s31wDF_f$iHs!<NcTSfvP9!m)kxp+NbIWZAtx3I_7qBJDp$1HkIob&iZ9k
zra+qLH{uyFX+K}JINEV1e$*Sl5A&MZ5iRK$&gKp3RBa6ZsJ6+1PcB#Fw2a})7=G+A
zywff#Khl1rG=}HNrds?UgNNEEgv-Zp#wywh)@NhJaK@|go5Xe;!^In*N_PxreS-P0
z#wx}@^$jtO#F&MC#u(^#7RCseOL6<(jKduBM!o^=a93p<F`ZI|f5y(bo|DFAQG1-l
z5BcRlQ<qZvi!qgBF;j<LVkarclkvSPV=TS3d3ZfvPE-0jYau?jNad^VE&Y@4!{#(}
zHrHR<QF1-6kY4eLd`)k00E5MSu5G!ex`>{GQx#~QNj{L{r=A=o8Al;M6wmnK_%_h?
zr~9$_CUo}z5|HCv)g_xZ6zURpWFjBdL!eE%TsJ|N0Yph1ZHl&KAO`_@X9MC-Klr>d
zjf=ixI+Q$-m+{u5w~}btf{jO5{NO(5l!@1G-`Ts%SHOwBXB>|TpC%uR;eX%`eHopV
z?}}{JPOqj{C>^pu=da2``X%zPK0xuK?J*}sFAW#6kKa52aVRU}b>t4Kh?Kfq27caY
zjpn0t1>NV(<}UH(pu5W#dXx*ETB_u>>2?EtI3`;2Z|3mLoH@!)S<yeDu6P}b5^nOb
z_{=8(<bI04&nxNVPgGaq=(9OWyp0tRX>dc<NASsfGLA*7@tc>2TK=a${bhCMDQgcm
zROYVX#U~w;_gQ`8qoi-hX%7KnY4{7x$1gv=UtMagV@}}bwZ7Ljpih)#qQ#HQ(C^<>
zhkbk9fd-+{6QJeQTh62IF2Z3|!s7N6$D;JFIsz?qbp9m^l{8NN&RKmg-*Au4Zh%A3
z-39o8#*ez&2&<>uz1EqF?>>~zng&>TA{qKZN}jUxTRV6~s|1`eIAAq1)d7h3f=8Q@
zZ_3?3HYm5=)i(y9x9@d~)rS(GS4}oKU@Gv_m%$w=iBBHhYZmtUVzyy!!Nr1heQ>27
zJ$-afU{b4$v=lbqkR9MC%4*zk!~$7*G#wKo>d8&T2}e3yC;{OQZ*LCpE!zAx-^Kzz
z-_&+XM@no`Jf07elrilfUo=7~c=~vACsE&mYkEWs6J-|dw+DWF*J%l?DM=GBipfkq
z5o2*1zNR>=5#T>{$Pij6&|@8S<BcX7EZ+CK?SUWY)+`G8z7-Q#7BzCKu0B4U9uuc-
z|AA~u7wh)XB8&-Tevg4Iwn)x>ljoeSA3+G%-sLNg*Z-TT6wYY?WJGVZ4e;Z%x-{KS
zn-VnEVa398tONJ~kj?9NH~odRx6(0<-FWS<0L|KkzRlguZS4LxEDQsr;hVfud$fBm
zl71fwv<r^Z3t0zMCj8V#7Ka!a0Da&uXhuAqeRaEyr;Q|<kl)viQO30~j(!Bbb6t_N
z=}uwlC%czvM>PkHw`%_^w3QaOhJl~>p7c3Umuk7Wl@&5`0IeSH97A$|VoVtTtwT30
zje+N8@dHrY=e8NcbGa^Kcvr&97~Yjf{_B*-@SNAFz>l|$7^Azo1^qvp8cuNxXKrP{
zkNQ6xV=B5{cdYUmn2##mv1J~8Dwn<%$CINdWAj(4Go&d1N4P^Q9dpz0{CzC&lUf|*
zH1gko=YTA!OVdXg(s1TXeRaa;7VV};z%WiBX!Wm@TKSy?ayH;Wi*b!{PA{uR98K8V
z@m5aR`eOOQf;)@xKQ?!=wxQiO?p%t^ejI22M8{I6b;)Xynn!7j<vWH|j#~k>w9WEG
zKf}sJMQ8vbJ(K2!fBBBh@3P=|WpbcB0%YN^%CV(BrF*I!?&yFxj)nu+Bpe4&#K{B!
z44G%)k6F)X9nADoz$&G4U`qF1XpW06pKhS4uWiUmN%ElHgsk}Xn^y5%CW3D^W2T#r
z$cOJx0)Duw6uEJ%dF0il>c6N%Zxpwr??(J>fgkc^{_pa4nFr0y=XZfnpI7hEUz<7D
zW}L-mE(x3}@H1EW&Vf!22JM^Y*RNj~@Pn1nP;1;bFF}6-@S{a0SsCQVfHC;7u#Gjs
z{kvE}bQi9<Tp7S7ofLD#**#ex;OL|_7okDbyI>L4LoWo_r0x->sItt+9aa`tlQ8*w
zB%mrC9Ec0}^Gm^!Bd0Uw_pZQCtqfSiJb&E<y5NU}{@*=&v>`LT14n()v0&X%IzsN1
zEMkyN>X41!HH0r0O>S0AH*;?_0+fU2uitfmAL#rf6gB^rfI|RI4BGc(C3e36_-%WX
zTw0F4_ql){zV+{(J=|E#`lh=2MzTi%x+75TA-8!+5~)K5b%0B|Js)eKT*_eT-f8F4
zhfiPh%>}rs&G;^Ik3swR<)zM!lw6TJVQ-ZOX>ce;KZ<53AGE(W0@whHT#>po5H~<0
zrz(Z+)w`ov{)`2FSd2~k6X*T3{xTtA7t)YRCuK3=DNSNf%AUU{ANLlMBHBI^PZprJ
z1Af$BN+Hfc7Wx(@Vl2k6K*bKZ)X~OY>xhW-n~;Ajn0cSlZ4ac+`*<u#@{8kZyY_uo
zn~v2mC6~6*l{LI_wAIjp{k9fA@Sop#fFG{5z)s)D<SpkFxiL{=GQdQXonm=E$}wO}
zTvA8bxd=JN&}i#v9O_N_T?m6ty7<Z@m__~^^x^x!7ZY9<3OM>9@7Gu)AkJJafFGNH
zs(e!|e(HRb&n!niCVK%LOy&S9a@qHT9Za(B+2Tr?HxqU&s~Ao>elFeZN}jJZ=>z0-
zQ1g}nD*k=jaW<t_dsp{wv;(B^#iMuPDegK-Q_8Z+^6l=XrGCI8Z5H0Rs{zJ?W=f%7
z7Q~c=sp{{2P3+UULLRhP%1k|Fhity+Lc2jO_4CWUzUTTJxsuj)(rO%J0f@TGVncrP
zA#wYGAC0vFmST4?@o}^P%1$3^yC}2GI_yw}_Z(ES#3%A6UA!2}66MYNg`u>J7okU%
z`lPh)F<K{X@e|`1ywW=Jk*qX^`vvt~eU7#jV|Xst>pX_Dn8)t#d|cdh41fMg`jQPK
z{-7_;-C2y`Dg*zeZksWjJ}vK;p*i;$p7P22cEC?*!@y~e;pk}ZZ4)CFs@hH5(W!j=
zW05uP_X>-ER9-O$Nd0zwPx~T{)D6A+>6526Z%pYk&b@4`uT%fp`&i%yiyCwwHk6e&
zgM@zY%{HB+Isoz6$P!~oX{@DgqrWjW=ksK<%9Y*`n{6mdZDShNDOiUv<~>lKl#hMr
z=xm;$FJYYHg*7kq4BPCtF*>;9V!pyW%KLt-G0=gq)|bAYbWZepSh}D;eiPs5Ge7Fs
zv{r844aguI5Cvc&^vLu(ai=4eKh!tCo~A1_-s7G4mNkFqXVlF!uSN>pHocoQ=mVs%
z;SZP2FUSX*w)U7etb?#3(`^(9ZRW<v3)$VZHKEPUIe;JKVDho_>K6FU&zgCU<*)E}
z0*KJhJTAxJZysg%Qb#VWfi%@e^hol~^-yTyUZ@=a((y^tW9BF$n`)mGHfYi>Gl#%2
zpQYTCH6Pn7Ty=>#Dm<6mHoy<{h$Cau`G)6({_6FGHi!&UhqWP~W{sCL?()991gH<E
zXz{asiI$Xo68I<!qEEDO&Bw9O+3e-|QtvNpH1GqEz~Xc282A8nYyeaM3no~wbQhC#
zrk<1?K!=k1=?R@wchL^T<mc!04l(GI(tcJ){i{HU=)mZpe|)@oR{bj-RrbDFBoBJq
zJQ<At>W9Z0Ns-B)bwA9I!%7CLrQ9)?{?l_m{w56rczgD!y(B%t_zvk&T_xEJ@dPM?
zg-Gtk05hd>Fj@Ik-~`~GJ1GR1*lEB`9Sg@;5bev&1^m?N5`ZMuZQ=MnIY3IC<`}kA
zNs9QGD+8*gj=OpQ*TQ+;eD}8N0NmtwI}}}%qEP6G1L#AVa3$c0x|;)8bG3`ecgnK_
z{Gbv=iCN^7&ed78o<jeyaXF5n3y_rZFD)9i_mp%~ywN3b`Jx$$-ZLHZM&HY%g`GPo
z&6EO4Y2Q5}rQ;u{Qz)nHM65jdjs@p5KYuUBhwhj=e$q*6&`u|QY&f{5$sUUfxf?1k
zHES2(r?zX7vB!_h=x$y>`DX%~yPVGlnzZD*(dajrkfS)39Ev=$%3bYvuQXmQEtW2y
zHCYH(n~2yhSAEYp9Bo`vR41jn{Y2ljG%>5)@sslq{RiL{yBFKu#A5>bu<#X*1xVUo
z*CBfyy{d^?(2H>D0GHsVU=b7d)Ut-f1HjyLa?b(+?Fv8rU;VC;Pdp?4MndB0yKOcI
zkcM{1cSB~#E_WpWkwV38Qs?Wn0ewL;r7;Bj02%`xr*cLQU{dMF<^?sk<cnRQ;hVPY
z@QY`yU-UX27PW3|^w5s`i;*KJ__GfDl<(N!ds5$`?NOKb=J{lPOk3-6a@b$Ca8edY
zl@ZFEFMd@cHvy!Hx*pFoO+JRLP8|5*sJ9SX=pbo%((yJHZd1QBZj9lnT#?ar9>b9j
z{ngxK_?U5N8N;QmsACu-tF#{H?=ddr{`{T`_?c@A&v~3{47W`U8tdsR@{JchZno<Z
z&cmi|3_$Qs7B6$TRObZvV1rgTHhR!c<qp6AJs|qLJUsn97Wj!YCH;UfUwAIB0r*C?
zcFhBQ0{yC;lLU3YPxQ$n)_a`$*KglsdIxdKdOx55Ck3q6j@0j$?m1+08vKUEjP>Q}
z!;vKbOzr^Fe);vexDKtYkt~Xz+8+8+Y}8|31W-a69Ka59l+>D>qr}js7|+8&3a};B
zNgSNPS{>j7AXOYFjajgWE@>Eh!eYn3AqF@({7GjUgLT|ns^1dMnhn-VfcEI1I2-Fl
zlVhhJ%NhsA{bX;NwLj<)9|s?NA~uMitKBQq!?-U8vB5sFfazz_z0%Ez=<shpbk`SI
z+jy28#SK09Svw==b%CSja)-qa`Y2_Gcld>Sy%}&)viN}p-|(gGn<g#%S{c>WN}G^k
z?gvem3>4he=3BJSsQZY->Y);|LAcbWePd?32@=^nF6giMTqyXsZqL%Um`euq71sgy
z`Klw^qCQ)DS)b#fzC&R;Ji91T`bgsYC|ZoN7f*7{`?y^*qMYx9M_hc+d|Y7cy&Ic-
z%&vKClm659Z>#_3-~C%l+Iz<j9r|BA^PR<MM%?of76##8TEwQS{PPO{q5uFQwg#&m
zfFD4L+=3+Mi2?8TvXF@#h><f3^|=L6giU$w3jEZ-1bF!QxjOi4L5+d<cLGn+VWI@+
z)GUcgN3=zGcL9FbVepecztoig9l*of8i__Xbf-~Dck26Gzz+Zm$KsiS(=cJjs|nyU
z#h~#kfuscG;9L{kfV><i0?+~2Ngek=rMsDs+n)uH=5>UH8tQ!7kv$hUW#|5;1ofVN
z59nd><2$3WhzIyt1b#FMsQXzv<lJ3ATnC-pi|qhQnJ;jdE)d0{5tAvD*4+8Tq-zD3
z?pWnwR+NgAW=bQJ%+xR5=_koEt=ojXlS2Jb6OGid1A22p<dd<@K7Mm@3oYvG7~rP_
zg5>3ZcP2;l#i=XFD-EX)-39ou1$|_u|8mEXbxfYaVadX|&pDse=N!;UziFiB>30Wp
z`-BLufj46oqmf&>qY|3%l`jUdoARM{uH>#fxz1ya`~6ij8Rmva6jf|ai=WMJ-2ytn
z-P8rh5>T`WuCNPFN4jcvW9r6&S81C3U7yL0b_|dP;OCVV1umO5kuC&~mWc=7)u*pc
zje6|;jgEP1+Iww5v>*84dufcyHg?p~*rHLgbJMvgA4N!w;3&Ghjr$Je&ExCxg}0#1
zvj+Us7O_glDwX=!Qd>^*J@)(9<9kv*Q-C7%?|Vld4ShQwPe^xez)x!NQz~~_7h(*j
z+&?^hwh=wXq*ShHf4_|3^Nr#8__gsd<S}IoCw)oRD7E|ao2K~&@R$0<du<G7vGT)!
zRqZK_;ZdG=&ISCG^|jL(!)enO2K;P}RkcOcT&F6HH6h1so~eEkeN)iE3a7L=1f7vZ
z@@9=bq8FdFF|#)2+czxe2z|iM<42DL40jjJ@;N{-m+}*vYf@b&;)R7}{N|Z&bgJnm
z0nRb^XOR*V!^b=36Hg2*!8;uF;xE7c+F2t&4#5|^1n>h`<KUC0Cf4tN_oHI-=8k#5
z2LXftA&fKhKY%t5WI2=qPXGj1nBmsbP`&UTAd0;+$9B>NB1C+2OZ=ZbecS<B2&d%F
zb==!;T0~Uw)@6f_7MRf-LOgutzV0SZKqr%$Xd^r7*ez`q;%K<sxy6s_q1se9%F7XX
zxjd+E&DXG~@QqB{yY|#in8go3j^}L*=m&Q#wecdSuEmeZf`oD`$#GO)4fvrR0nQ=Y
zxJ!xp0Wg=A1zPzgG5o@f0e&Jb&k}%*KOiA`N9szOs2s3jEo+`ld})5myGww+;5q<5
z+kWpU_X2(VO>12z0;bf;`$l&W{igolQhof;eMB`HbNh4D-Jkyc?~H6O@KavM{=WKI
z77SG6)S-ZuEMQ6&2c+8t_#v-H5AJHmE0;9;Q1JZC`_2M4zCURZJiX_He6d2|xVK$h
zt-K{Lrj!@fFF#f8*ui25I=RET=7Em5vodLL;l3h3h18Y5nZO5Gk)ULi+%CWm^x%u5
z_Y9m6UszwIRx8ALsAFI_elK^wi2r>74s!uN@Qi7)A0H@XL}kR{fFsu$Rl<r@vVJ0<
z1m@sRd!Zhst|X5iWVx0f2lqntoeP6gR|4SxXaPlYcVD}6_-5?-Ax>(!l*iro@2dv<
zFmV_<?ya<Fgq%y3E-1Df)qtWOE+#N3kdo4jcsy&9CBn*Sq}&%SoM-XFFQO)8>&lbc
z-GCqR;<$kT(bVK@V$@Y+%K#3*^Q+!tE6V9?A%{zD2NYm<D}iN5qy~mC*=8~hfNv86
zP2K>0;+wKKdsh>o;_l}H2p7_?EgFphexL#1!4U-j1@OhVlUl71$9HRX!F~Mm1qJ-J
zb?jgZi=TOcAD<XemylJ={_@2^K#S5HTQxbB7PU(9C>>c?qd7%($cz3y1%9|lmkITK
zb{Q6SQs#hDSo+bw09XR1Yy<qzM!A0>clE_i9jm6@xzxSXl@@Gjw99>te``QP|IjJ*
zI`ES&l0^#IWnM=JOLtO7T%LV>Y#Q%e-qG*##fl-`7vRT_^TCS8`kF*wF^1E|Laxgg
zj)2Nz@h)Tda{%rmjNz1~x`AMeF?<a06L}tD3}?(?d>QLXV<c(waVY5Z^*QL&SO}cr
z7!Kb6Kg{KFy|8A<iav<N2WdH)6i^|ZONLCgdukD0K!o=*C0%|Kpduf5`Bldxc4wqz
z4pN{cjZp`Fq?d3fjvWQtrXbIJLkgjcvsfdi3;p=l<eBOTePc1-2>pl!e82>%^D6-Z
z>7p$D06F>G-1BbMzC?$Ga@L=im9d}>kdV8V>Wlo_<{`RAM?i*KuzX+r^8979<^wfZ
zi>RUf?uQ>biJ<dZr#}R!5P?_)J=SR>j4khVr!)E`4t?JKh$Q-q{kD+!Lj;XX3a8BN
zyI%F1;~`7Oy@dshrGQ;I>e`Q-K%sQ>-M-1jfhjvKP34I^h)=THY6GFa;up^D<xziW
zk(3Em05=32{h%WXQ-?eNTX2l8d`+IngKxpw2D12=oL&%d;GO!&v3|J&IQxv%j$9n|
zMtmIlaS8Y#-(PhkdyMaZgNT&;;3cp3(8@On;)k0G{FJPH9L#CwmrmI&mP#i?%ndw8
zezoV+v7|=lT>R<h(mrMmzS)@cscGwf_~RevAi2;OA%QV>1-<>Iqc+ty-P2|Vteof{
zaP$Yqa-)Ht(t>jzuu)oQuiaU*3-A+yLx>EJ48Gjo6N_jbu7Oh)GAy*Eu8;pT-q1Tr
zK?&#@OW!ScHWy3Cx6e9<a0Xe*jAM{Z9SizTvZMJ3&yOYRq}k!9x0Fus#dkmssROWj
z+?;`$`U4PgyxBfnsXmeqN9lFz%J|&*?qRl^%f5CTU+EY%>Mr%im7z9&j)lvuCrU@i
zl>k+aPNZFR-FKNof3knumbpuB{$Bz7T(BtKbc_s?H|=sopKQYWxmf&AhFb&2y`_N5
z8y)fR*(QDhVx(-BE+$VEGw45Sc0Z+72Dhc~J}Ovd^i70HlXIVpaj#JB$`i(z5Orhk
z2K=<N=;NjOtG~m{y<3ONi$=bi3hRcNE2o1t6SXnGPj0~zIWdVu9<|{e(8&7_j!K~K
zR=pO04^YUWMrk6{cNRZsN{W0PT!6D?wiZn5+49A`CiLBHoODv1Bw1UDq<-7JTW*p+
zjw#%H7C$XVRz{nRwX%;;7BbxWC4BtJ<dP#;x>ghAx?KV~N;~ZO7Du|d_y0;kn)JxU
zTAhfpI90$GKNCC!TvD28?tMI$7IaW_xicsq%YtStO@5RMmnWBYW%^=GNpHa7!WSSS
zuXvU%e#+m{Wzrsx6Ex_Do@oIhAGiDD6KT)s$rwPtJH?SE<Py&`{dKVTK?W==*2Zx9
zavRSoV|c0;q_Qi4$`QVd;UNi)9W#vKwapumOFLF>jw3e(#y^pzLhjCK4DXJ0c}d3q
zF4ap?Zv*^<-p4p?b6@qtp+5yjLt}Wv^Jlr=-)0PtzLI*813%=&*zjHp*~DvIXo#(z
z)rRf%mV%Epi<+8cZ0+0R9Bl*q)GUok3*aTa2_DcZavcHxR-a(QMtdX{VP)6OEAo28
z@?6Qv7@Yu6fV*?PtNvZ6uZeLEo%N?@&(^;qPriNWgad%BpMLzo;`5E**-EGC08Q%n
z?-ZUpEK1(#jvdxl;)pgyB`N=}o$+P<Mo@v0TpB?!i4*NAXsrq81uLCWnUI6YnDQ$g
zWbN2(?QX{fa6^8^oBY7x)jZaYbkqm#?8zP0=g=TMa$r7kTbsOd;D>a`gH1qD4(bQ<
zV=t=zB_4huA?M4F_bmZG?NM(6ei|O20c|#0mG1aKCbr>#h;mVOEUfGZyu!w)+Pr)k
z@Kc`4=EplaSIBeoln27PbAyPzzvo<I%>Fo>k21B8KI5tYKksF6bN>JqH!lU)KC$DL
zX}hfhR5=wZnwG20w>Pg}u5PMNJe8x~;BeF(Kc$6p?vu%_Orppfhcy;xnFIKVe9O=C
z1%7H4Kjm_EeUAzJbUEq%x+=g=w^ZF<Q-B{*mfBYd^eX{BuXU;6I{`YWV^FuF-a^|-
z0#Ov-f7Hd}+Fh1|a!ItKyaw>Yq&Zx!7@*jl3-E)QIdsoYy1<hz>aZxp!eZ_yI|~xH
z^vJbz0W_tVrmrcO@L5_UMM2@{nN3$#GMUi6;6mk&D^ILb3#Cr<t$qLyY#uR1{RY4Z
zMSqiYxg6B7U4fsWMWan#S}r+nx(wpzUb@PWFSSMf8t{`&_DZ~z{L+PlG(3(6OZ~Qe
zr@hC#kg40Qz|YhjWkZWbDDSEISz6S}j|@Q;C2NutB*^=@vGSvQxy6qk^#+&a`kl5t
zFE5ks3#(gk-_dm7rv;Ew-f2rEi<yX%A9WM{>B1Yb;V6*wT=$3<AWr;FpHFnJQm%&*
z9%*8s4nJG0(@}E1Fo2J{IGQVdw@-kd*iGH~)_|fm#(BM^G8)S{X>7qs3F4Wiy)M8H
zdQ5E$$I6p2ei_4~W}MF$uI^~PkQv<P&-!>=AH#2G+%GMFp4u2bru?ZqN@FD9HE^M?
z&xwWIQoj;;4vpavFP<g-wi&})zRTk-`N2P%GdO~46KDvI=6Q1L*PYvUj995W1k`jo
ztRvnvRup~QYy<qDYm@+$P3zrqpAh_P&S%>AJ`OWaLSx(3mHC9W13%JGwc(h>T+^qO
zja~)#KJ47dL+*eDzx?__-@HBk8GrOb>L1RIa1($*`w9ni@w=aXw1xWcizj;A2TtkG
zrX*iX-UM*q4`)Za=^4OIEDd;DZZujw{P;LM#RD7D;vOUPulznn$^`g^Jh<lvDo=GR
z+#Q`*Kp3D9VK~btp5&u<cpwdRu(Z)FfS<6oArbulq$5;I?LU<VZ4n;U&N3uFB<i40
zSmUrcqYV7KvQumPgSc#OT=I0CyJYc0yLt8ct$|y?x62qSo690-RRde506*vwc3ihf
zbX{r#o8QVhEO+R;rA>?F&6|1Z`F?LZ@0pK-!&0JNUKQZyP?C#woX^eP|Mo?|Q}?up
zU3vN^odWisR8HX9hBo)<eAF7+g*IY8zIl5a60@fnW${CY!u>wE!?Nb%r_xSK4Q6+5
zZ46W>JXl!dZZ6=5O2I(IF=IR#82lf{i{07ahebhvm0jJg7C*N&Is7rZ8z*@BS^Uva
z)E#BS9Qua+i}Ld1NU;3g80Z|r;>U(p@x@Ve)F0oOCEp&t&duTnSyCRaZ(c^e1%6oE
z<abXIzhv=)Ki{8vkh))`JaANA>Pl^;_C2xo0{p8vI#cMs`|w$oR^*c?Nwq|K_AdcH
z`NA#z$hiPNRu{5+J#rDIAWaJV5U+H$HQy<YUn?CEL%NhEzr$@#lstq!jALnX$1eTP
zx(lu3?#McjbgARpJb(Ej_0s#Cz>oLw-lw?r4|*<g#)LlSXWzgN`Q>-xMB(GEhWNet
zWW>_sO1tTx!-U5Eqfe0^-!ONF*e%TWq90q!xWm`w#)K9g_XYg$T}z8H<XKxZB5c?-
zvTw$xRQBi&>@v+=ZE{xHF<WY9rFiK=$Ct0aY24_xr<{k{?-Zm-Eq;jK?tty4j{Bm*
zvXJrM)oWRqHWT%fW|}eut5SdJ&E_3S?3C=k%Zu~fA2&NhQ+lav@UI<x`CP|p)W)uI
z|5*lpYKw&|&ee{_<tP#AXzGTo4Aa=B|K2WUj^~v$A(wcjsjmz0LmyhMgE|i8xyDIV
zon;J{5cps1=-kI}22SQWl;t{);b~jved-wA*XN*P;#jz(=QTY=?d%lBaO?NrMSRtC
zaCi^LqWDKG+|wV%;#P$1{%kjMLAyRqw(-(DU6-$qtr2%S;HP|ym0KsYV2y*Peq1XX
zF!GHbzDV63v(7QS=m}_dx&9FRmY>Wk+O-NBv-A<s%juKyF^unqbyC#Rbi8A=hU4J<
z?j&kL>)6bKl}_s7m^bA8%ZnG?^&a&j05OkrmvzZ<w58*7dCH8g0$3AkU~aWfdx_XS
zhYHXG-=;}LgtZ6$03E76I)OqbwLI87<~gi-Add`f1EJ)Pj8e-Ni1S_2dhS}dKoqT2
zJuO&A2Jn-ot=-8(Ja>|ZHZ9V&K5@5AVQB+CB7eZQGVt@-PHhOP>9c0>^WoD+J9aKD
zYrJ3TBm5|9FeH(>+UBAuz)$UnYQDplY#oWbsSkj{n{|8Wc=VL>-QV-|o!u#(HU;{M
z1^Aig8+{h2a{HotlGi+xH>r#{)@?JD`@DVga&=Grn{1S8Q4Z=4VD=Ax`?r>7^^PAF
zwz*V(clUkukH5|j{Llg4zkAz&4tO2wP6ha>S%)xM{o}80_M4Z9&Kh9h?wwmJ?g+{|
z?>v6`y9N{xXV*J^P?P}u?#e8W<GH5rbuPdUgZMqww|hFmi({D#)Jg^1^IijfN>)%<
z@nFpp4%y~bJScB~moXWNxIB^lPXd%mJD$^cymQ~x>$g%Gr7RuOT{Yl`Z+25kZJ$dg
zh&k|+hNUz--e)@-n);Y8O4rhr7Nntz!d}{$oaQ<1o%4<727WHA4&TwlgPmbh`}loY
z{P;Z|ZY`0^2P+4Tq;rR&AF~gE^6DS?$BytRedN5rPtLLU(wJg=Y3K12zD^nV=}s)%
z0xx7w`yA_{jpaMkO2^-A?wV_NqPF=t(89!QSC@h`06#BYz0ty>FG8kaX+91Ry>Axf
zZrV|?EM`PKt*86ED(@f!_}M%UpNY<<6=2IrZO3uG%bk0dnB1mrXmRS1j@7tb0NzU$
zKeh6v`HgvZpo;~wtKnjCYf5R*twS#=^&{5__^FM>ZqfE-8N&xE*1k7yoWmH-9Y60h
z7to%B0XOd${!N4DWg$~EnL$<(PUi%EnDYR*?=prbtl8^aW4HkT(xqKC7Y(R8yd#-(
z=up1ND)hvh%W1z^)WzZl$Cy;xz(qKx5|nhI(xvdkPu;e_PipBH^jM64sXImC7Z-OF
zvYFzI?k#+xJNuaLgtPdjXM6m%uN58XGHy{7vGb?q71F<{OROO=2O8@F_(A_*d`rhW
z+6*8h$ACdMbu9S)0^r}w1<+@H()~TqK)+<;4fh4fLvi@0x}g7|hjj_{bjVTW*Rs?y
zFfpl0kOasH=UF(ggdcgBBMxy@4#1-*BlE73l~w>eShw8Q{mrq7$L{~(bb=tuC(^s+
zt5<ljApJoHELl3zCf>Y#@8DCUQGDtuCjkHq4mpnDxB0S8Ik}KdoYtQJzv>o0y5mPc
zn#qpOqACK<@D5m2I_eGZ?ag}!_dG6g^t(Oi8Gia<6ZqMjFL<e@06(?k6NluJhx^8*
zb{>$2$M5dEDtxCuzpCEc1YA4;m(T6yW(k?8VJ*PVOoIKht_S)3V;|#YBDohbuV24d
zeZTi$^+cN-Pr?1i|M=(Co1gz*W2FlGuyD-LZMkDWpyM3ti20X-g-ywdfr0<$=k3mq
zavs_Z{NS#EK51u8TzJDG>8tKcfX92bJH6YvoC@&sP=n>8`<uJFS&aWhW@ovxod%EY
z7Z*S&69XM)x9m8>>L+*m0)B3@YePFzHWy~O=SNl=;tS;fK!~`xI|ty0w(v}{FM%oF
ztV928tpt{=t!lsziXUw(l^yl=g*vv+s>9UW$7aFP?Ks|i_$Z~uFL2D|$!`=6EPqP8
z<nd9W-@Mbg!P+HrELQ@4=wd=ygfH6mxmf%#DFPhqbM0hhWjlpFYvDE=yBdGc@k5P@
z9NxAIiHX9aW(QVo{gBdJC(=s|`0>RL^65Hc&A+iOCSu+AyIA~OtRCn-FeYcY49JV!
zY)o{REHRPHC$IZnG-3xLJGOJj_{9+jHi@%9S}+rUOT~XsoUo9FvcNOr84DY^+ZXV|
z1eCgzFRYXQnlH(kiD16#k~nh$KPboS3&gP}Z`Y&VLYb#6A$um7xifviWbZ)T-R5|d
zG+qEdsh_`NmGm&9tNYZnvlHtW9QBurS5sHp<(I}8^G^BF%W-pyAEfElGTnZN@A~@9
z+x21!cMkxzgsTA|xy4T_-!kwsWC2r~u)4gNKzm&PRHC0s9c3t;n;&sX6KBFoz)y{j
zseK?kIe->ioX#`8)fA4#0zV8BF;3HmT^Hbo+IdgbnbE#u^G3+6I)-aOiDO|*8u*d`
z06+jqL_t*2F`SE}W0tav;a*F38N+SIx!NPHR`<_g3<vo6P&wi?7T_+_H$IUrGUpf`
zwEau8f2s3PPg~#m+g!j8Z3!@=1m>T@7#?lDm4r5EN`K|vlQ9y@$Gn!#^tYPKr%)Qw
zQD{>fON^y&z6O))3FP}#73yXKFxcX~#`Jt+Kzk%zdz5P2Us|%BOy@LFX4XK`O)Q&b
z*1Y1uecPa#e;4WzItz0P^mmS;&C3SZVMo0wu4yNF#&}1Yp>d!?1_+Yo&&IfzzXEJU
z95$$N{GT}+NJYqqJ3Z*5zhZ%u^X!Szzu%>kvo`SF031RL*ny;YMdmzPf984ib3KOJ
zi#*r<`fl_uJ{R-2{6zxKU%b+{{bt||Z4<v=Wr>Y0WO4Mcd_{wPBl3@L&m6e~qF_zL
zowm3&FvSn?%Gw5&KX<jZH%=aY#AZ`Ws94z$bshM5rMdQ3`~IxQG*y@E4b&^~%w0q1
z0O9QDJSF7G`c4`6`6^v}^Yo#T{)!x_d+nw|&#44n6GHs>*1=QXle*8mz=Yfru{JjJ
zBq;`cRXbx%@VYq0h{fV(+Sh)kyL=v8ZqD1MQ8`ZRSF$*~`0&o|_?i1yx@}AJPcPrD
zUjB#w%aT`tA3Cd&#SeFcyc8P0tJ@a%c_^^tkwBB&ebmuJuit&J1vC#k?K^%-_qM<{
z3-FX9bqsa@1fgU^7=scP8y}j5Cj1Y1#}7aV)+I5Jw|r?4M11j0!j-IB2s?G)2SxR#
zr<z#l8>Egw^|ink?#+q8vS!6q1Ad@~41W-M-K8%8vsrM>7qz|pQ9n{u21=U_ey9((
zZ^=5$RIklA_@J=8c&!CXElM22m4KfTA750T;PuoUi>CmveJ)+#X(v5xikQr_KD-J1
z5+Dc&Tw450X{PDYNmA-hzf!XJDNWwebmQK)2YzZ7yRr+5vb4%?$)ZAwJyNpvY4KCz
zue5`UcK2AvP~@O7_4tD(X};+0b$Xlo0qSYE^t3)svzWq8%6w5BzTRt>F^dt=$JDGd
zP8s;|2}%i^q4rAlny4ZF=uZIWxVt5Hb>N4IEJr$FUD@5xz$A+APWwt7lSy{dw!mN-
zI{iHbX*Pi$Ellg^I*zl*9re_KFh7sIZBJb2s|B6Hg}b#rfpVI~PgYhYI>)p4(GjYJ
zT~<ZlN98TvtEHnm-ZwiqbNO9|-<yk>!=<hPKe1SFoh*Kk)-r}~_C1Fl!=vuywYRTt
zWJ68rjy;Cgz^hXm!&8~@UatT3<66so&gqZgWxoCLdcdDsnuk$``dm#%?@N=1xjye3
z#+aX?Zwxss7;@gwiC(;X*}HJOt;G-gl<Gz3S*flGhy&PCn%B6l{npk*;@IWB4N~L_
zNDr;tq1!Ol2Veo9P_uG@CmZLa@7Q-Yu!fCN=z|Wdu*jd!+p!4oennP1&z>5HiJth&
zug@iCaReO$OroAZ5dF~tZ9^EAHCSoo-zzT<_r7umkdY+0oLljTnq}tM>2DLKiKKW0
zN4i+oVKW*Vv2uql%bLNDDi1sXY`u_tj7H14ZTvM_P0?7#V(p5f@k+X8SkJ%*#kcr+
zN$<|#bQ!>~QvL1nDGPh}Y}Ux&lll#5jzgYu)Ej95*h_c(e9(LeYhGNM2lWEbvvkxO
z@``(VEn)L#67Zu9ZM3J{^=%r+<F>yWfwvyZ+e{LmuULSenS^-;WZrxQ_<6s&qqeE@
zYEFXVS6@D@{_+3(d*h`F{GddX?)agjXR&@)w+-;aVEVgfn*ay{@n2*m6N8%bcFMrd
z#p<tqc)S7Rz;_7<!Rq5d;og=Kpm0dw=ZP$RN_W}-3<6N3?p%SNnq^kW>Si0@2N_YX
zv0}qQt>jQt0D?FME|+U5e(uMjGshy{(-DEWLK<=1(rkFgN5r4=IU4xUnYD8*n$F38
zPK#YQp)}`V@jIQ&#BSjT<Cz}Wkow1aOqAj`&y;5BUz=1?7fN^h)D+T`=J5M=zz>TV
zEWqdTLvixmcD|d;^FIdgqjG65=eUU6m4J}icPxFsI*3!h7`mROANZ-&mHenJ%9u{Z
z(=wOJn}$z$XK^cF625(DaR7R`m0MpPwR>brJ6g8^emrj-0jdf4{lfh~9`2T}pENN{
zLB5pl+QMEMr}UnJG$r7NNvcn*n;CaT1WPl(nY68?G}Guc;D>20dE`6JOTctWueRf!
zNk0oXIOM`4I|tTS4Pa-ftqNGo?w`jx>aBD<N(uO>efN5KyLp|%uiD~wAMdrr%)T@=
z;3v{tC*TK}v*;VgE`<!zXK4(m^~=i`o=T2)0B7|(nKdADKsJ4}i@13@PrHoaE|)pS
z@LbYU$M900a}Hy8iSJYvj4>>pbEIxwUn3s%mSci<nct*fW=S82ywWHA6FO?{N^_Ek
z!(0>#kFgd%kuIKbzhM048P+-u&fGC8c=Ub7cx3h@vo2yz%<*s>sU6mn0TiWl<Glt{
zp#w!e5tqKD2K?9@M4dZ83g10+CyuV-UL3%Q$B!Ot=!1Yf-0#M^8Xy4a91JSWxm{no
z(B?rbMsQfZ;ESGx^$q{Jn?)^}{eiDv0PgfuwK7NgH7z6WAeHm)ks>Y$*tkfYB;-PW
z49v<l_h7loQG#K;#0EyJw3;j}Uw&}+5Fk!iABD{NEPm*h0_br<d=;0AbLGD1hFJd~
zkJ{P>vgdx>aDYzm(`)g={Wyw)Oo$uKPTtXRez`+m1aPBlwhpN@0RZuxetgPNZ(WBz
z6kt&rG6nc?9e<Obl5Uaav5$9~!>Rm<xb~6F>@;gfo6eaLy79LWK>U<EIql?Ab6Ue6
z{`ki!Nhig9E$fxX*+#nezkOZZzp=Rn@>W(l4`rqELE+Q~+?!V)wBG$_^<E%8VaIa2
z+}Bg$S!Y>k%_5F?-Hh={N4;I@M3qvR$MEt-vV8mU{pz3nxY@opa~t*{&p-YB-&w>e
z@I&Zd{lMAn-DXS$=6;b~-8R4voh}w8xqB}F|3Co0kl|Hkg+%v*rAEHfdM^GMWB_k+
zH->%<AVB#6z{(xgK6yEa^H0xF6g$JLS$J^FSMKIN>P^6!K)9QkG5|k67w(^_0VpM)
zE9ZUN-=+Est&-eodHnd8-NTz5+@D|QLTo7%sbd!kN5bW<1kj~%0QInBqp1Pz*m=Q<
zz?VgdL(!&gjCCj)C~lv#%l%MF0GA%#m(r;En>xN5?HZ-%(5?jhpzlgo?Kl)yc6yz%
zqu*hN-;hfecuJEj13+}#TTP8(r%)Un5P79E!;fce(FoPlF6GSnc<Nl=ZvB*Y{e3&&
zr*`yGS&=O-{?G^S%i;$`J1&w<%g#d6D;@KeFF2Kyc+xRR_?p$_TG{({GHwg})b2rH
zV*4$-f2OwhGUuWZ?VY`NIQprni$<0(`2)R29P3fo;pz8mbT=ljxWLZQ+(o^kTg;k?
zMFqfMv4ROb6Jd6BMVqBA#Ib}mkOCrKzWyd~t=k{xe3X8tAWeCpr`!gE7P|u3=^7Jp
z{8O3{k7pkc5*O8{@flD07P2@><>CF!bDGop1+Z@9$jg%T4*e7TRLS`fcbj8&3HYf2
z&2gT8iT=6yZ38SKTk7-_N532Mg|toujq3#bQ08R}k1y}f=u5l4%NTwxV|Z;0?0(Po
z*VHk*)aTT!G`1PTTh6*~yN=;qzT=L)%f{lybsEpn?;6Akw*&z70T*44%fHh5@LtQx
z;R_es54hS<Q6=m3Aob<THv_S@1%A*mYWMbVtQ+&!T<3iv|9rC*bpQYg;EK68$1K<M
ztl9>wJRi!4{`ORQM-C*ck9w1GnFBxAkN`*kR$zsK-V%CX$nAyr1MGOz+`(kWze&e&
zOP`449H4A6{k@@iJZnE#N?QA4QmiUP{(y-biCwBo+$qGGfrBlL@(KZSKC#Z%jsln=
zgLr2gM~8hP%XU9*t|6fZGRLZn<L#jJi#Fp(0>&S@vgM{fJUz$ez4?(IRoqhbBcwU*
zZS6$Bb#J$#_q7dQ2S7@!S#T#A-xv#?O}#}Mg#?75pG17<^#MPuop_#WE-Y;gtYS0S
zD;<>v$VFl8h_?njtKAD!+GIy=r8y(=Wp0*Q=M$H*n#CKLCa$mjCedbnWu1y|)8%9{
z3#wr*+s#;q^Wl@_YfW?KZpXP#Si>8W|Dk}N;N`2@wr)j>pS{l%*a$GFH^GF!l=g&V
z>!`b?4?-O0_DJJTsm3F#cduVnfuEpp=;y!xxBrJ_*$ez|1Qq6(xnqHtPIMPn27a{o
z{9LA@K4ff=uK^}+bmZ1s9ov#R2JM=4&>X<kF2K)E0&;F?Fwb4dQieDs0E2-503&yg
zr3heMle?{epMvF1&}9Gx_{t~v)QggQpJ)*dtCQTFI`H$mrySGN?d+(ZKT}T`!4g-q
zYU%@i9_f2MY`zzu&ubmCw5{WtJbB33(zm1xt?WDGJVM)cpa?+a8ICdmX4MwcBHlip
zCE%x&58@s(nJJ|`lpnwV|At%&(v*N7(w7!g@#mYei0NuHKJHSQ;m5NM{E$caB6JpQ
zb0ssSIs9H;bY?NU?|zQjL^QPoh`eiyw)8<c@Ds|NE2kXmwcT#v+78J$k34vbXX$)$
zz8jOkT)tB#r$K+)MWZL$wU%!tpby~N<hxF*^-8<)J|_3W1W(&OqpxC2nYw6{=gDIC
zQ(0z}K!ef(ORe6O7b`tC|6Tz7e#`}nEqR;F`I_^03euE;pZalc@X16Hpo&F{nE0eL
zQ(pUkpLC~jn+E#Y9Fc*D5za9i`9cEqkHwp47ktBqidQ<Wu5S5N2D8bRKGXD8^+E08
zK)IiyF7^ReQI4zn>@H@GD<x?{AB<<3@~Q(rZ*>tl3m<LWTE_77`|mo2r?lvEx3Ty^
zV`t%&zB_lz7@kXHx8GC8@KT>s(}7QE47Ywyb1CLuSpPDv@%~Z^KP>p=j=mASn~kig
ztB+}Etnz!R2iE2X<&<UX7S~ak18lDI0Q}t69MdfGRnFA)=?QJTa@%E|0+2!5&Nm>H
ztb}N<=xqQWwYD31mUWbs-OGYCcIs-@K`HQKE7#H^cp?ETe(3uFNm$fIHu;@Y^nZMF
zcm^=BjVJ~AQhuu=vK)%K^cNiom#(=5pyL)FYz*$!B`l(Cvnb3-m(l-5)AUmqbPs!y
z7e5ILZ=;Ww9DbDLS#|_4`WkhOwJ?Awj(Lj&0UQ|+BG343@nspMq(-xJ)SKP$(;Tyl
zUg~o<wHfu>Sogqc=yFIOKKv_q88_#44JajE`H#&h@KHVLO>^Hb=&i~q)niCov-kl7
z18fVqAu|)Rs>pq<?}SyF3&{U+j}dx%>d<+gsBg&cdWGK-@S|L+7WduzX471zy~o_S
zFBN?Mrn(O>MwL4!2XGi0miEOrZTDO)e)c7kL(}a8ptw)yuKLiqYA>MBN_T!;>VCA6
z#m}M39{G&|ey|RBs^h$J2gm~0(&qqam;k-6TpItQ7M^csiVqF&FDyPYpcWh;6kt^D
zO2ALeI*0~X0w!{rL%+8Lez5M~=(gN_w1KmHgq|M@*DlhN7FqpV0Dfu~Dt*9D?K^YV
z5O)rh90Tp|3&+)!<tyu}CtR@IRZgIL`oPbowG|3L3HW&`YoB}Cg_FA)@I(FoQF3AM
z#?|6k;D|qI-GawobeUJpQYeiJX%vf+rPhdScaAa^_#vY&Cw|-;yD-MMbTPdK{M^>!
zMC=YrX_gjwQM?}LxG8o>=B^L;*#>L~`Yfu`$1~|Fxf19>J?mRgtu4Tp7LC%$Zkj)R
zP!9a0;ifEa0sK(jhq`DrCK*AK$vqZ4IbcF~dGVtR{M2^0mUbmZ-nE5#tWVxG7o<kq
z^jWfspwB1)KQ*vxPKzJvZ;WtEw#srFf7Bb36^@|DCs4?sv}jaY7zp4epZw&YW5i2k
zyz4tMxn%ZOap|{D`-C)|TQXsp+@(XE4XeI%a+}gjv+Y~Zsae6jlj4h9HeGz*u0FtG
zh@XSc)ZQoj=zP?L+||x7XIC)0gG<g9TBKX%3yh^jCDQlxn?axc6K+3e6Zi-}g=b!O
z$MBTPXl$COOYuz8T?ODLZCj-=JeS*)F&tT?G*j92)x~8DZzW|5WIac{oz)nwzLP!_
zm$qU08+!9o8^hr<byLUiyzM}%wEG?ZZN_lYrEab<9KJ842eUZNF{9y9>(%g!XX*Id
z8c-Bz`ktxo*f&;}XgrYREk|YLuBP8%p?pW{JKI|PkWXpj2aD(EE6iE>zP9q-Fol19
z+?&fCZTzBWLEj{`M{SHR&An=KkP`62I7fT*_UYC!_}$eeR~GP7N4eN+NjS%%(Wje}
zp4W8_AaWdKI6xIncvu4|bO7Y7jPGepDUC~-zX*f~(9w9z@=#LESDZTUxcKeA{4YQT
zd9X14;}1{m`=sIV&ZZ|USS-Hi8n`0K7&a!w9@@|$-15_J+Di@iq3)&bo-Fs<xv=Ib
zCXe<wHUJ_kj{X3awQ=$Hqk&mhq^|HASGV}_4N=62xV)3LX7NM#b6KGS{H2a{IyRt{
z<{`Dsm8Cwf4*aORvT`aNHC+OhQr>t+-=G~~v9zoEs_%>CRa2dM$tkMpNN<#b1#Wxd
z!W0G<;HN9CxvXxwRQrD&GEeUK(Qz892ijm;vIy$ubbq0vfFHWL-|IN9)G`2;=!AZG
zRbNzQfggZ|cr}panRYV(4(0Aw-5<p6fs(r=G#N{w-1&v~-GQRE2u}k9TuM9ioZgt<
z$mI9WHWwJmz=u<GfW>#Tq<N5rMfw`xRkDsEJ>Wx{rUrHqw`NU5SPfA7sPkI^jLOcB
z(#tKE{F`^dJB~^NkSe)8K&ljfykGquCF&$HALDBElq304P=(#|97BkbmAeu!Nu1n5
z2qgg%<CJdDQjiVMU!+)km3ern{^nLyZ)HLBreZ;qgMEI-&*m|Mk^erQQ-B}pSAN9V
zK6#r$o1J1PhGSeh5qYZt;!_bk@ryunRjPE11B%prDZEACr`Iw=JN~pgE#0LI|H%2(
z8;<?z?pQ=+riQb?%5F>?7r~ZXZMSCM&c^z}^QRSxZ0j!U{)Cc3O`q6jRYV)L3+W5T
z9oFs#V4@D_1K^)J<V1Ox6qa0VQk}~Hx=d_y`q1p#%~;=^xH)-F`72E>2+tQ(SkQPT
z<)zfG)fV5=e8#@h_eIb7UEk+f;b;$NmFj*e4;E-22?Pbu$sLpKbfKQIVF8uvIi!n4
zlzgI+FAmTj14?^)p_Q~wtXO-ZU&jK5Ix~k$fzQ&tMV?0)WC2WM;Q>z4-EJw(l*c-K
zVwKxs4&($q+IW7{3t{<pahaF<r}EJ*Wj-R`*hQQ#FtPy1B39n^=to}3YAGy`f=)cC
z1MK8Pmw2V!-KXmK#?iLM@|eqLOgdYLYxJK|e{mInpHvR%7#@qHn{pdJhC{mw{LDCp
zM;Vqe+$C|kWB6C^e>7h3U%m-o>=<r3f_Ced+T@fmJnctP05s}T_o*>e3!eG$T*Rr3
z;hSv-fIuf0%ru5~`P1En8BnwB7;gQZ^duIiQ~jHD48R|jk8|haENdCwGgAlIh_Q`s
zkONqKAaTwIzuBx-vPdY8u};DD^p^jiJNACi67Ur?N}Ie0D*>JOBYzgrOL8V02jCm?
z75DMKo5yXTFFn*I)y=gjbUV`_1W++wvijPLcQW`fKm%|C$DFMOlu*yG4l!MEqiVRv
zLw1TMoA^FxeuyptWj|(4V2oDIsV|fbE0=J#o+Jw)DKsV$;sk)xu&Z>=uQN@`#zT)<
z9ek|$Ps#e4_`ke(Ve%j!j=RJ937tNifj+P!FALK*1E7iffg&DHz6pJV@uuIjkcW0+
zOsmTy5A;RYwmXEBZ<Oth<dFhl)Z_M!A6cd-4RId{@Pvolz0xu2A9a01fLCca^UWO4
zi}2tn=Ih9XO;OxanfJZS7yWy-G736nz}zi<^7%MvYW+L);I-!P)YDC;V<*-Bv8j)F
z3GuG0;|oy7!sT<l@ZaUhH{f)IIbC|x+q3`luUC(6j@xuRCl3p)vIGErE~K;He{`0>
zPk;h`oJm?lIu#bQ0S8R4QOlvDyszB{5478Xj*bqcbY~58YInX^C-m;4rG*BIF`qe(
zM4c$Vx$h=lpkaW)a)dMC6~&Iu;`l6Y&}ox~)qYEew2Xc4wPWXyOAB`S5Wr$EGGVbj
zckujN;7Yl3_SiL9VU6=fvclO;1X-M*48*|Wsr~Q!k2-_&T^AgSa27nce5>32xvcoc
zpFVLM3~0Vz_y<UUs~szcMcCW-zSH;CE!~|UtFaniyB=?5j1>MD5TiSuO0t*&Hc_8T
zcP0TgVxfZ~!%=YD?Sn#7%C82X)$Rl;6sO}w9Yb#K^i93-d2r4v?SitF?)m|kO)c)Y
zBa6GTj^U;NKkx_eRa=ZZ2LF*KccIY#jB%+Ig)gcCYHTWjOvq9zBKkpgu9xCIz!bi4
zgh?!3*ZA@YFd$Bsc@{q32{`0<J?c53sF0kad{^y)y6%&q@}g(Wih?vOl<|bd4JpUm
zcN2X)XO)+LC<j>R<FzAB`J=Sg`goM;+A&kLg=g|-*QNKxE(_W;JF26+sSIg{Z#6;T
zSb(nN+E}#-CyN3AS>#E-N_}w9aFbGP0o`*cADclK))xGOPrHjva*KO~YP-Z)oMjgz
z-<z@<Uv_k1fiIQ8CTN9v+e%>E8~Ls4w8kUONY5AIiBVe=AS`t(OtGP%Wc|UU>$T3~
zM)r(L+%>~F-0r&NjKz;{Y76PYn;)^Uc_&cHL+PD1^-29Kzzg}~ww-9C_vG-60qsgw
zXPb-30(I3sI0qaKQsAdnFVM^KUAV@4^M}bNU{3zcd=Uy=z{^V8WdSB%Sn2Bn(*^c4
zPx>|PC^FDcbZMjPe1zB1SnP{DA8hQ%`<Gg|`Tp<e1Fe7ZK9F<w*&&Vm>A#L~o_Z>W
zq4(iM+D!`dfw_z}V|9I~gHf0G&f%hdUp3&TJBIrrFM3HXw^JL#0rRQ9F{T}R4EK7{
z))#&=-#FwLUfZ|<FQ+hu+gL@L(!Wz3!;w|E8kkOaZ458<IW-H;ZN~5zgM!Do#&Gzi
zO);jW3(n-@yY9R5G)55?vKe#2^=<BnIOXSsflaB!PJqzm_+4N6X40k>5^0Pwh~yS2
zB`brW%@Qr``Vn&+QN_5MyV`tW$hsXm=p}X+N-{3t55LSS{D?NKUop2?uhV3GlJvC~
znjf%$OCN9bwV}sRZ|IV#MGoTwbEVvU(tPU0OD(iF`rkU7>`YDiZr`MP=cFHl`$_9<
zL|FqldS}y0x&e-jgm1Oh13#{HHFS->de+7h*5h3<j3!SK(cjn@-ta_R+9DuE?$E<t
za^it1lZhiGZ0`*{_#uv-D81;5X}t})j7Qwblgg4K6d!9Y9L~lE>G0&erl~+IP5_hW
zv%?_|j*p`#{Lv>j*)Yezk}P-3nK;40WGq>PPQy`VPqgtopJStcqIWZodn;X$^j<f#
znNJ&@b6uDGYSw^e*;WJTRR(le`#>4fn>xNnpdB%X_xZJoukxd-BlkP1=jgBfo4VSJ
z=qvbiV_l!#*%vmJektBm9-%Qd%~YOKy3~icfuEbI<6oK$!+-eWA4}qx)vxcUv{^F}
zv3*$Qq%=9&?ZM?HxIb4pIO^@j2S4iV{*7`zC3$LPQ^(D9Hy6*v6Lg9eKmX<5{Ey#R
z(9RtT%8wuTxwUy1ogV+_9HWDbINVpl0(HsJNdn@e6U0(HI(Y8=ArD+`B@t<e$KcI?
z3?CmfAbi$Yw53IR76v|O00j)8L5_7;t^g=NpZjVkL+;=MFb8Lgn(8Tw@l+lb+5uK#
z0Vt;xzfq!p|Lmdi-OP{4HjX@y5dfHh5E*aq_2JVO4U}I6)@^>{nq^Dy5YP0OKl0?h
zqWrrPx67Ay<h*{bo#X77?~WWTUtb(VjEuH;@@@D=3{HS-{4P0ovu`hXQ5SMxBuzv8
z#FD2(3p#u|fFC^svb>i9eYbFblVq>@hV0Y!Mg8E&LoC_0anzqzZ@Cl5+eYboAYaN#
zopfF=?GB){JPu3&3Ob!_c-bXQ3HS+4y0yh~p`L%OSOxCF$1Z7Un>W{s$7M>fP{^XB
zE4*##00K}9c&30wzUYvHIP?uDa7<zW5>2z+$o{+SU_zm$4&>(w6WtZZO>0ELNW%n)
zwvhKhwMjhbYCDk#OC5DIJz}i3&`Ug&Z~Az6K^Dj?-DMecQFsaG{BBdirR=i9iF5#I
zY5k8h)PMQ|+OncqT<^Qv@8of1nWy&5WP*GEBI|9qzVj8CB3tsHzQ5CCjc-|+C{Qoh
zP06GpZ>O~haa%5b>niwIxp05kX5W+Z#oZN5yx66hJ7~fiU`EbY9=G?m>ao{f?O@dR
zDanBJ$SWpR@EMEAd^2^Y>u<+z#u~Re+f=-)jMB_m#OLcWIbvd)PN2yT$HG3}yDR&C
z>k>~{{|lJ+Mmq~B&lE?0VD(>v9DPXI9$5^7&v;UYan=sCIKX&=qt3?p-{cMr%1s|a
z{Yz<Dg60q38mlH8eMrtb-)z$=uCZcOdEp2C$tQI#3xK$qE>^Z$f(AVBE#aBA%{u@U
zeL_0<kNlz?QI@Xm2bg?U3qR}tZ3GvI3i%{K&Ej5J&rRbU<2`EEW{LW_@ji1?-|S08
z8omK8o@xAOt82fTo4ak-VaIs++C8!~-Ea}?42|LR(bVT<443tT?&r$gDURX8n~0V%
zJpYcR-;7Jv=UP91Yw{Rg?sH}w!_lAY=<>E#bnv!aBrcV!<`-kQZD2wt)qfmam5*=e
zp|)`VeFGVpe$(nwEr?Yc|28=Yy`M1`y*72Jj>K4BIw1j_AoO7X;M$mvP6yZ!bCcUT
zrRA3NK%S*>5<Zw0)7QHmvboj}yp}B1NK><fK(6Slp~nQBngv;e$Fp?29ATzIi07DB
zgnn7FZn5<O={{J5z-z2$d~3cfb-@);?mM})`tjM*jj<3=!4~kFY8U`R@Ch;j-5-Da
zL34vtSJFCy>MuIXckzonm@B16=o4l~X|w9AcE~s!dK%?J-Q>4{FY&H;_`9EaXr%?n
zw0KQ=!Wh|qc=~i>&OmuwH;xi0uF(`N(Z%r_XFN;Cq@T#^t;g`c$Qhc*_><PHm^0^_
z!5B-3W5>Vi9a*6hr*j%WwpRj(fWTk{YoO5c(|Bp#yqm78aU6M5cIGC?xujFR7Z~$F
zb3tSi@;05c=@W9=K8Npa^z=EY&p-H1W!JPtlS!_BJ7~#<Nc->^0juMQp4<k1Z!GQW
z?X-=1HQ%TgP4(Pc0U6uo27ahBp9~0GtquYF>`gcaCj3?4idKn7=c*OCm2Q$79iNC=
z*;&1j1@8@k{|{w}weR}H8Bk;HCn^Fz|NZ~`zkDy<8$&_ZJzHMXCbaA5WZtU7d^>Ot
zP#>Ta6rqI$O#nYMQ9vkwms|meG-i#V1#xyf^nptpSBLU8);*Es%6$!5+dF|%8dL!Q
zu-0L48|xT!IKIwyZWj`vY`qllfOS+KpMV^{zQ+8#>lkb}jtvl_&*jIKF?c_Hc+VCu
zOX=VXKm<A0j^86)%_^ouv*wrITT_dd`uABvMLZM)c+Zcui*V{D7BiIL(fzw-%@$$l
z(~nc@n*O73KYpkSLDh+G<8tdIfG>`G8~TpqMccw^h<55~(Ka0Ho~8u+^igH!g*t@J
z;!0oqLxhzUiSm4Zc>2r~!`wgpy}84#ds%tDQwT*G3z@MFg|v1w8<syzj`|eh$cMVa
z<fXO)DdMqE#v*L3uO=R}y7&G_iPa^qDDky)zRS;dwNd`kLL;=9l(5?gg}rpt198SW
zWXB|f-I=KpH?$LoNj5twnQ)YxwL|SnV==w9bBlB+kL4pR>c>;0`7xRCqetAAy5kl-
zb-u9vVG`d5E(MUFFKL9CbFvBJ*3JI0B<1bW#<%vkiZrbD?u!KdM-;F?@<>NiRDmS%
z40!obb(uC?I`V}4y^iKbu{1kGHQB1w(;DyiPdR2M=x~Ps?X=JN#4;V@yY+)M`(74z
zlnKWeWJe!XC!Be1-d?GXxnuH8dv^<}es!1psq<LWl*S+GkyJQyd`IeP-(msv5Rj63
z(f2JO_*$Ub=7JcYu=+q*4be7+9Q^YQyNiiKspbn$<&#rQ!_xOWUi>DGFTDB&E`J~2
zexZY|7Ykvf{`e{YKXZ)X92Z0#rTJ06uj?2d3t-ICjy;A)JE%Rc(-_XE(BE^HrNkIp
zx{s|K+kMk9yw>NO`WVh#im_PB_{Zjl(k7WrIru^@I)8oB!)7XcsE-`cJmfZY0Z&o8
zd#sNhy5kfbI{mgKx+N>yerqK*D{<#;tFNMyb3AkCGNpW4KUqg<Zd~6W=T`E34}gXe
z$P~1=&xJWajCXO=+i$k%m)1ioTCJW53`jQ<!7s<W{T2X`H8gK+VWIO~x-;on^RdlY
zm3f0@(EKqF#Ma#WB!wn>4{g{nV8>dvw{T>(&eL=`MPyJ4lt!r|9D2k@mrkviD9^9Y
zb+?VuJJ{oC?D(}Po8|9x?`cj~I^HcUOZv_{@8v5$f*#$Bw4>&yV)%njo?&wy{R(C0
z2)d^Qi=8yT{GGdm*qGPHGbeDczJ+dDvU+c6(%!Y+VkhtL4Z>~K>j<emITn&~b=`&4
zH8$yW0mE*r_{FzFMh@Dg{6vi9IhO?ZR_U6-9CR1pXAUXuhXxiuEAD61X(Erc_GDe*
z3|XaCOWTku0zd!F|L{M3uiXjanW$sY!t$NlLcTQ^F(~*1PrEKPbNs3S3t)i-a0Y@g
z4myB1bmliMeffk*C<{keDSXlnhLS}FgCglzjK5Sm#w&mn7V<v|Sc!p`W76QW<gi8>
z%O~lvBB}4}F6H2ZC9*Lq1(k<8T7tK7in=ed2HbGVBcGH*mmN7Q;Kx9KZ0_ADFAgDB
zK&O&p0qFic?Zi-efCKVlu=j7vxA1S9D_KRpRA%adTh@48fbY`1PVhqAd0*kH#B0f~
zkA5jWeqUr2hIQDGqh?aysQadPs7fK1&kcwd^@uRyl^pT0m<orYa9`hvBc5&<;8w1u
z5x?{-0cP+DpIpERT_z>#c1k;;9!gQc0xbA0(LR*lRNzNqv;`B}=_BRkxXu>~$M68C
z;@_bj8^th=@`zY~0lp9^piYaVY5&2bxU@4Z(t6ysUeR`0^sT!KTUfR}M*RTvV;(?S
zcC;b)%^mY~iU1yFouGOc3##6)pbTMg=Q|41azTv=9=j1qLml?~bySun16Vdh+1RNc
z6Yrc0j?|)G<ajrDnBwSLnUJ#sjzwj7&3S}g)W?3>YLm9pO?yXF9_w3`kamdT9XrG+
z2Ovwy!OMH=Khyv=oLhj<r<fH>^U>&Y_+}`b$a!;jv!?@0u;a1l^Nlax9ZR`!EHsqP
z!zK**4gr9~>vQZdgg!OZ<=NKRS~_?Guo-r4<wbwUgosHXWn^K3MSGq&zKOS`&Ah$c
zMB9@UufCz}E>z!b-*XGq4*s#cDb*=ssPHH43u?=Ad~4cB>iE8~TM^)rIzgGJNj~mN
z_G;_noP%YgFFU^km%2#XZ^s{%l=`T0dmn-o?3XG)xCy9w-^6#MZ?J_7vA`Hce}vUp
zETGwVE%3M`BlyM=FzWVZ8B|}yJ2GQzMn3U*z1}c>x)oUDw@Kr;KHi3$hfO&S^G2Jl
z^{KVge1}*FD~)qk0r(jj!>L<7o^|gjjdok=!-vQ4uEfdXSYvpM#c%{4)RRMw;f$f)
zJ8$;UT`Bo5<8j0oUP`;^cfc`R<DoCq`aI(t#&E_~ED)x+bPVt7bNG&IA~MU=6C1;y
z>wc4NU&%OybpSfmP2UvRR96+lbTA*IvF<PHltVIPOyW}aAy?7`NVAPTjGj70hfcq3
ziBIqMx)!!rPOu<dbLcSX#uwy=J~PDuP*Rt?FY`4Hr-MavJFmd{H`R&Ime4DB8lWIO
zE518Ny`evUV~)VL=0i)LL|#~dg+ouqLW%n9ak@H}>87j^p}#kBpw9y&V_hfJ?bo`9
zcy>LOye)Ij4i?aL$t!wIf)$6fIM&9PBLu&mty=^&MX(@?tUs)O0W3H@<`#Y&IN_F8
zL)X}FK@ha2L7TXzYZ%;9f8nQb&L=<GGO(rxIMMpCwxRE*O<Ja#(HAr{NY8g<z0fRn
z^h6J#e5@@}9`3<qjmAsgWDIrDFfk<|q-DOvlkezUk&;@V`244zeq^`P@OF9EHp+8^
zIdsY2+u`QiJ~*J?90S%wd^W4qbx&WDpq^LtSJfHxlsb*b>FPhX(T}G{#hQfLGId~1
zw*WtLN^@RX-0|~Wz|Vz%pF>#u<lJkeM^;MRv4uEA;OC$J%fHF{!<^pody&8*ft@9w
z9geb+O7DC^iNF%fj$})e53)kdKtTgE^P{I)22hOt^0I41)P|nN+f?8uE!DBXA0{EC
zMXr>`eZTKpaiIpH=I-izV$V+cZD`HI!>(mGpY-P?FO*}Z)HtTs>h5-S^>#5V8~7>n
z;75Wn`St0aa}0ooVh>oDJ0^oH1f@GwEtTT3TeWYYBc~O=>-<HG;b%;=>A&E529ft<
zIxO=%asB&6%$DX9dGswX2Ib9Xo7ZO8Y5O2Y0GTP@cN-a92jFMuOApCyD13<i_Th8Z
z-)(7an{NA3?>psiA2pqWwqIuAzB!&ur+6MI`r1Y-ljy~EsY~;Y;V)jk81a>Y@9XnD
z&!g6juSZq~(UqD_5!BbvoxU}|kAVTgZCFqT6ar|#QaRjn>3iskIKUTx0sEsd8r5+P
z{H$wK&Y%AVkg}SjeEe88BkBBzPzu7Zb`b4{ydq!nTLV9yeEm24t$ff!e^AzII*Muh
zuKD*c@`W+t8Q6ki);QwhEDlfo!^>pAdlR=qSN@R^;Z_#>aEj>iA}{V-;x}~BXB&Bv
ziCp;E4`mwVvfm!TUxVEyoipg^i2z-Df=xDYOr9!tlmp8m-aQYWN5pSSe3U;h8-+Ux
z1AEVOA7k-cOw(5^z|UOLJP$hW-n?90>Rz%tH?&FdL|9&cpA-2I3mex@0vvIe-aUmf
z!5v6gxAi#`363sI!5<b|{`>+{=<^-arvg8eH0?-_1^(D>KC{&7`)b2s=)2`b=KZD}
zrao_szKQ+3IW5%YmcW;<(~fSNn!oj4&CHN~ZdaF20;aaHJbuul%|F#~BWVw`c|2o@
z`|fdT#Gv2D;%CzetCW6zKPKnMJ*+h5&tW%;SOJeYvUF-L*R3DIs`*0W)XfQl%r-Jw
zfFE}E_FGB~g-@-^=NO+phpv7?b=<m`=!Smm<kln(o-h_ywS0IG86CWUl??mZM#y9i
z-GIASo;>2B_*T!@Wek7uLPx;bNLne_;(a9~V0BQsX#hXzco!Bw>y1BjaEkk5awiT8
zzW_g`=Bnxi&=DXDLUJGjD=<Ys54OnknBl0i{7#;UzU=M8;~NX=WY@riQaLyhbRpy6
zdRqfB79aA;gEF_|lRT;MR>TDG0C=%<B=tNzju*p0A>ugQ;1B&1zShl=JQ*D1Mg9?w
zm$dyypFns3LY6IjD18d{$bQnKX6>}$n{Q?8>JMNICkI8D;Rg#Y;)H&TbrXKjz!;1C
zBaTVSi1F4sWNDjj6qfSd#<(7D68ZY`UXK!YOUc-_06$%sT;*RJ$#x-&pBn;xF4ZaB
z75H-sZUKHy;VUdWTuTW&d-R}l<Q;&|^VeMVH|+9zlCTi^r8xt4IQ419m<s%qO7T?_
z{B3s5muL<9`B4tLD01BOb=j<Z#G>+WzULjIZ(?66*>zSHRXvWn=!5KWG4Zj<RbgV|
zv3CDU8(@Dk`K7>iN_k2<fevLr>DtESbuaSfD7f?}J|_QI?bs}JAfx=ABqolW)g2Sq
zf#jXwIqFB`JEk9t@^_fn;>Q#^06$EuxBG&1=BD;TbxOk_`oy*dG8}V%F}^lq%@jF~
zkr9B>rS9>HzT!fKjk*-iv{YB&sJFB{c^^U8klcpCQ~JyMI0?+r7mmy6FtKX=@|3>w
zFr3btOJCcl$z&T2<{ra;eev8n`6>PDHhk|RO<t$cwv_j8gyEWLi~qO+f73Ge6^A;=
zxJEs-K|LAojD`dQIN)Qx9}D^P5u1eEz0qAz@0+8E1N0%pKn^`qKTjcnH!}C%{0$Fk
z8}E9qb=>AR(agnf>DID73jVDvHEANx;DKZSTa*(&%3|P!-doVfffnRIIL^{HKX?Q{
zLKx|T7Qb=UR+0w8D`KTgE@Sa*K#Y7u4|dl;j{SD<h!n(kZNH6Uag~rTXyW+0reW{+
zdfH&cW43vEjELiTI!!{L87Z8$`N%7Hf_K^r^wV^fc=>Y@65tEQ_aDO$ZJqBvWo}7Z
znH%`Isnce@==e?CAO849d)EHE)=}8pHM6+GNC4pHhOSh&^XPH$$Thgw^}e2MO0oby
z+emFmvq)eD31GEj_XIVxxMS$$o4$Kj+Dv~^TJE(9$9+w&-*NZE`MNoPA4<<z^ZVXC
z&#_H6b!=1W#!ej4G*gw0H0`;*_f722wM;{N=ifP>Xs79ki<_c^dWzltmh#1qNy|6w
zQl!4`n<wNFC(@q61c~yeWsdLm5Yug)FCLCF#b3S#_!)K_`N3K3;kYN}^&7vqHvA6x
z>|0EW@&-To@9^K}6JX)yEuE7-b0*iVA2N#{tv*b#o*E;gyp511+HO44Pj}UTpR{c8
zJ|wrH@St<upY%7Zsm11)Q;^3Tec%`x^vkrZ*lB(WWj_QDeQl#}teVQ(++#Q!KR9lS
zacY~nz*PSC8Gjf4^YLJxd<N}Z_zfCkpVUFdwWz15#gFhFvpNCz!C`^&Lt^bi++TGU
z6f#-^Au!)K=sJD(FhGV1Wl`FA5nqhBf{urQ8y;fyP(g9rm<KmJW5L2PWFQgY=xapY
zEM`i1!B@klgFBwkt&|erM<fm0QC#Q+@Z)?4Zh_s%Q&0Lf@8#SRjsSk5o#UGWKhU+J
zLP!aq4Id`E4t0W^zNI00mLK%VlfUpmd^U?2efh5QvKX7RNQvu~!6_O*#%sePmPGb5
zh+E3T{uwjw9G0Ww=eDg(q8|I#9#;<fjJbiI<&K|mvOiX=w>s*r5BNC-UkmVa45=@9
zE)tLget4_}K%w>FxTohA-P4mQ5fPR?xft<>$M;Q1!JqiQD%}^QKW8@=@PiCkh(+<*
z$EAgH9qpEXt8EpKoX)BLjtSGYd@}z?=PcXOS<;=B1gQVGQv!Z!65?|Ad&vdXOjH4W
zsM~#xi(4P-h&X`baO?p6{MmPGh2J5cG0A39e2Q*g?S0pN=oDqu$I~`)+(t&}LRhIz
zZ2&)gOrJm)6YHsNNNz*nQ}bKmUp)!XFEg4<*-?BIq|nzk<{Yc$9>YI={GbKsc3eNF
zZ;{VE^55rFmVNU|9b{Z%oWm-e`WoXLW2}V<_yLS~B1@ligAlsjyZ0XqSU}x%5QTZF
zb8WzjUJVpd9yWiCK%$^G#&6WMcsED^BCf+3_(E7i*D~aX{^z*YNC&UXlX1Zd?@ktf
z02O?2@YBS_lmGRv1(!T-&?AN2@1rDM9(?3Tl+rOj3n4eNmf}~)!rqkzFby6_jw3Go
zz)=859^?RxMbaO#B7X;m^d13v254j;jc7Z_u6~zDN@)2t`RHM^h-2UoblRfu+VIlm
zLEgqMPonEj)D@D~97hJ;Y%HB>aTfpclK_3i0{k5JyB*TP=IysHtMBjLt}Y)wTHQP0
zJAVGxfB3&w7ytZU{!Q999``FPxi1p9q7t}w=l1I9!~31mz5n=m_2&IYDHfXxh$swq
zZ{J!yesHf*P&_H|-+uVmp>vKu<^+CX@`ytBRXgf;cTx(a<lB)$rpfhF+-BmmO`4pG
zIcVfG=l;#N`7Rmq9{>(YE(-iD9Tzm!mCB3K4&5QgAXkgF&AlQeUWRy>5<a$EHJb2e
z6CF(`rY<J!yKE`Xoylbyu*LhXGRw0>4!E0A;(mDgSZ{3&!gA;ByY~(Fld4c2=avBb
z2lwx3v7oioWD%UjfOe_~^~1V3<-HHi_f_{|jMxXyr$}>~wy=$iNE3@2@l55h06*+%
z?!N;lmDln<p9IGCB^--5<u0tbIK(RNmaH3br`|6|nYZazB5mo}*EZ&q$xYqU_awWg
zZ<jHgQwGiz_{qP+*mjbqyNbU<7E#Zw4r=~FJq7qlH#*oIKbptfQr*I`J3Z0kvyO7-
zF30bU$`}?j$jm_(K!K*+`^(zO#`Qx5n*Zk7fCwl`>yO0`ald1Eq^FZ}os_UNZvc@1
zKEj{);TXfC-GyJH1{>Yf&+`eN2KW&Rqz42Fkcjle4gCPl?6;oA3wi1tevplUEs{5p
z@Jo4)+?ewjokk}IoN*wBji4MC>@<nb+>`HTnJS*~ljIX+#-BNV(BkQlU8c}7nKgXD
ztMRA2B9HK+eDD)xChg*`mjtMD3-EK?Zw?sq@Uq=-zfW%Bj-OqCpSggl7dj5`*_}If
zwA`Qm{_m{V|N8I#jXnRjpZ~)G{Oq$LEVLI1v=X=w@T0pT1OTND<$~kNzDbco&p=r~
z(Lsp}MTYn&Q*S?fx)Pv|l}8=&<%K5PH@iourLr0Oo=+b6?f*~>`8UQ%No$nB%j8(S
zF)6o+6BCPV($>=KgJvzSvHtnDAszJ`#rUeaQ~VC6C~M@z^o%~RcE`^YzGKopB~2PP
z|8DK?3$ysSfA6lMw-uNDhu`d|N%c2+XYv<-acbp5AN29lXYKyvXwR|zQOcPnv&d+M
zMiZ0zQ=O2`RnD&T&U5j-t$%rGycTh19FA%cbp!c@3%~e1^epw&Lut~5uu`2`fS=Q=
zdr?E$o^0qVFX8)^z})@vzWUL*<xuK#sGro$d1W&17+&IgZrSXM-XZEn-`KUUa_xgI
zb&zq*#yNo~eD`}=P-9~cx*mXH4ghkR$II8RZC)Qdh3W>aplS776&--YTrzgIdSdj_
zO{;*=8uV$xcr+yJ#p1N_BkvA*5?B=I>BUGny>;?nzIn1P5j3K0rcp>2IsoBD)=8R@
zVxWl<FwYL23>?z$)DbV_6aW(8_!)0{w>a>lxDm%Jj=*<>gQaw#?f{cfY2?p)Q*>k8
zX&McQgSD46@Z(XNzg-^i3|{@+JXYL&h`d4`ge~rRNr1lMDgr-qffIAqh8j;b|2gvx
zTJN-``I{_$ZtISpI~VPS&j>%(b2;kmSdw4zT_j)<xLDo4dwcaj|NQLzuEJ9i5(?+1
zFJD*h1p1&*UKO_wi=R|Bp_qnJ#vmNZANsDH&W*yyyD6To&>d<*cZkWC74FD?4kgbc
z7jwye3klE$u=`?yb)nf>Hvij#;y#mHtT=Y_p1>b~Vf7Y+nb8NatN4TNWZTsev9^#h
zb+>xjFAf&q3Prez{Sf0AZn!3TyrI9_$Y=q6SXAn_=ot!6ze}y1r*U?Dr{DeUYjc!N
z)9FqH33uh;Q0WgL&pFC+?0)%J`e><b1TXW;WWF)H#P=Na;1D!&ogw6Xh`i4&59%Ou
z$Q1Z_q&pweqkp4JoZNw~#|EFUWD%U8;1=w$!CEwNF*oJiya05Vlk!7P4<{5Szb2jk
zj=zeJ0^lUwGf@a~i=Sr7Yw-+-F^`lVC&8SUIu7yHiY3peVYL}UAhiauNU7ge6vFia
z9b8j9D6fxtvjF{Z{#;)8IhX@~$io4g@H2WIL{y{4Tp~cB;D`9;6qoOUGw?}I>thtr
zWTD@7PmwX`d>Sn!kGvwgH64W!JsjUmJ_a1|>c1njeV)s>)N;JMO_c!c=PCj}Q{`~X
z_;Ub1V`;ql@_BXR!~50EhZ-MrO310W1^79YFR}1(%_VRl^Kk7Jx_M~;5IcrY7jP&M
zC@+9K96^OO5If7Snwtvz^eIZENm$>c5hcE6VE`?TCfoNyNMCuj{nlT!n6u3hYTM>F
zCSB><j7c|loa;8ZJ;l0^e^c6krd;N&X`ul>{EEIXk3aN#m$c9x*Cy3Bu+vkPiZ|4U
zW7!jvJNg{{QRlgLj6QrH_l^z8p)b7DFYYS0zWh!l>{R-pQ;|oBk8NZ$1pJipN&T)u
z-5ZkIP<SejL%m<O@$gV`-8P@-msx7vx9(Zuu{QSMk89R!{YUe+mL`ofq=)pijgU!<
zuPNQ#-gh6v;hRPEZFQC1`0r0W^j(I+`>9{AQ0gG{&wv~?0>1nGyZ6lE$2KfUfBm94
z`j@X?S6{V36F-~8D-B>l01AKrID#ES@P-9#Ptv6GXv8Gq`j`E-B>aJx{jl&B+=(!}
zn{f2g09^FpQ$EC;NQst3ZPTvfHzX_*<RE>Q=wVG00LwaDfgva3Aj=O>g=5&<62`No
z?2*68M)b+w@_;pkTl?poiYw>x<LQ3<wxTrRinQiWdZh`VhV<lb-!;j{V>wOI0~Xnn
zIFv(A=)jNH34afHcD|$KL7xd&MXKp;&hZ(siT$&2hv0Wk`R0O!HnoMh-#7IOb2{#z
z{PfYUdicX1|2QY<g;qrZV}PHZU%g%3y|2@pKD=99YGc4$$IHc0bz^v_u)kV>pPFPA
z|3w0e1P&^JslZPkbEU~+-=r~=`4Y_%tx~*G^gD#YG_~C4EYGQBUE)uZ0QFVsu_fAe
zf|TaCt1^FT?PqAwEG_%k_r88{ERA!FcS`-xsnxUip4-TXG|_hB8T78pb4YGO;m45E
zu6^Y(c)rScF8%W6*F9yKSig6oz|W>6`+b(ape~=7D~83-ca3*!_F*$7N3F7HljD9J
zd;mxRjWESlL^oI4JznU%9+lj|2cT_f><2WO^w3+o#^1ccKl0&s_nrJwXROJ~NUrGy
z9glgTdnvu`0Tg+9<BhU4&gd$Aga<f+Iv?$J9TyU!!z1#CEPBuyJW_y#B>;=1_v|~M
zkJE85hB&s!&KC)KieP+^hd`W=c>q3<4`~BT3ZTdC_-S;S`1Tf$d_<45l+pNC0m6^`
z(sHX>>;(Lf=i<&=0*n!h#m{+vL%9Fb&(BwX{Z~H>7KEeTzUip98#4kwgPD$qumC?J
zgtSCiB(O+e{}MPB@Ppol_7)0jUeZG?(wyqzUoMYbe{U;Kn-oaHKF03XUCX?rA1496
zF~AR#DJEsRx_kix`qV#;)y%_6dC@ndT#QLmqI1fAF*&XEf2Yj*F1&5q5ABlY)U^2#
zJX6!}8h;xZktXU=Ja^^e+|m!ptrUJ;#{YBsmdE0kewp!ye^*)O$?JWoZNxV?<*}-d
z@8JJjp7}R97rqa{Cv}i<%`IC6_-Ge@jC05hg%SPL8ZLiYB)~z9P$=;yni8$S7)u!G
zAtKkrGotqH`8QEL4(2Vm0GTZF$VlGOcbzGJ1W9N|Eb4FT<lh94PrTcUWvoE`5AQj6
zV&%!xHu)GBr1y{^`SBh{uR%J1mjGuXZ_Bn}(DJmzo<8C@J)b)S@B}NKZ<Dx*yv~Q;
z{I7oj3dOq}^X4D2NlaAd-@cBRf-jq^kP-ar;Zqksw$8HCA0RT`qpqw8E-%+c0<^IO
z_&M(9-o;X8D(B?IQE&Hj$B%AkpPGDc@fYA{Z=ziYFA`WJFhv5#0)F5zRYY^1hbDA$
zEe2op$yStaNS?X;m*3k<V6M8kJqOcL!f&WY`<@_;kxkUuC`Bzz9xvjb!;`)bw{Ll$
znk?q%hoU~j^U!6SQt$Z6@Kia0002M$Nkl<ZtW8XD+sKGCQJ3O5CC_d6oI?K)aYJsk
zdJ*9%FT0j|N?l0vy^im@$eLaKmosci7?S(gez{i1hUlMac&Tj!uVZBr@eldz<NJ_&
z4CPg-FD1VrI#(?mHOLm}2MQK*MZg3W{Q2(fIH%uQ`VUY6QUPEXaphs@m5#g&XfQOu
z^EOpXe~cH!aS)}6tNa2W!ao2P^00K|C7Pah)5FOF-lE+Q*LXz{P@)L4{77puAdq)x
zv~FDvc^Mx`S%Occ@&v|6k{%w!bL!%lH~av8QXY)ArjG?^F&Q_%!MI%RqSur+my0Pn
zWT$Y_;|xY832FG<yc5cA9JImCJcaUAB;r>c2@Byyvmxqr!++rKzpE5S9xq)_zl5Bw
z1gOWy1b#k#|GK(+aXNR+9OFyHQE7+BdjWn9A>}2HMFNWi&Q=27g^u5pa(AhFj&^mW
zNmwYdx#DFinJIV$4Ni=dXzj{hDLr~|quZ8(xALQG<x-W&g>D{vOif>=+VveH&t)Ro
zl^MTcqOuusSReSXO2w1cDHY6<RR0=D%fJcmN_rQW(H5wyOi+W5D0*IiUGm9k_5H5p
z-<M`6tky3Mr91cVG5yfFmvLAAw~^7g-|;iY*fS)zq3~V5+bN}|&!T?}o#K>uOL;j|
zAD-sf_n!C5!5`xii<ov_L?3}O4eM(gxlDE`;@o3tAK$^}K6Lz`J?C@C(iHHMdDN-5
z={aQ*w5WrWJw56z=9FJ`j2PDB+{N4K#b}RS+dM6l<fs1~FbT>I(2$O{?#>#!zmlXO
z>U{z69dyb;m37*Nie=Wk*Q|f6Z5f*l1A2sO?+gkbe+I9{7x9};D)@1L$jk0?hNfVa
zkn)hk<p~Y?H3!3-e#oH(xs<^2QF!wj034L*6RjLZw=H#8|3o<X;@2?X#FV##TjA~D
z)bUo{_%*+s#pcV(ulc=x;GwwYM8NO&fp&N{VLgxNFH%xH(XV6>vIk6^+v?T7X)Cg+
zGM*Bj?+A#<fsn@pe%jK``}X|Pr_~R4?#xT{gsJHOx797#@C>;Cer`T|xVj_2ZG6ig
zqP$!9V)3(^{Fc;<1QrR*C4tW$R<{{EM)r<<3pAxkm??zfHCMc#gJyT3tv>lghOi3;
zY!2w%vkN6}&)E!;@006b;t>%_-zHBsmWNV2KIa78+OkH8Lou6Mxy;Ko<~O|HMm7VF
zd`13phs~vS|1pbVvUhBk*RTBd82+|R1MR;02wiqF?kc8Qzc{9BhlsOFzu1>{Oh0sp
z@*OIVK3UXw*hWUX0zWlgj_E(O4G+m}DEydZUGhGx1fpH$b&P(Q1$OVB)z52OjQ*Z9
z=r}P#Mp|@-@QYFe{rElR8T3kNOL>f;IW|sT+laA>k#E;=Z!AA!<I(T&QToR0v9yjI
zZwUA~Hh-o3qt2GX=JX4|hkD8<ryd$Gp-jG^zuwd_!UBFaomwxAvzeR(I6}O}YYi}%
z0xZw(&>;<t4i^#SVL(ZA1a{P0+wvX2$ao;FO^$pnn_9`Zzvx@zCSH>^y~8JxH%}r1
z2;rhE>D-#O>U_Yf;)r(W^idp3tUrX)w*mYFEu*W57N$R>H=2q|Y~JG@o}i1Y0FKOp
zNV17C;)i1sS2*T-MzrxT{;)i1((xN-yh|M95#QDP6dw6Tzhsd;ng0zU`epqU31KpE
z_~&&u?RR+(XEgL94Ue#R=Cr~;o{%$3CiAHGqa`{~94kf;%${kw^qqXWCC?J>>1WCV
zS;X&<VOa3uALHv3R)S6~O@#BLU7aHE^UH@1tDo-Os}*6;|B%P04Kh=R@cy^2tNS;m
zTRPP&q~8DbMf7f2PV`ySo754G`{f`0Y4zlXhpYErZmj;<k5AGlTffgC>jn7PTKr4G
zMFNWi=90kouODr~bn}*Xrn}NDO&&uzOO;ddVL(RZ%>9Fh60HcE^BD@loHS37)}iEi
zie+DT>X$%FNNiGf0I;>6o9hu;>ld$j>6d<J_zQ0%BhtiJ8qeXJ&OZiy7rT0KLvkAm
zKYwYAE5UWqdB#X1m&MXAZ%XZyWn%rFbe8TC&}JFV!N)CEu0l!Xmv;2?@x8w%<DD%c
zWAXF#n`O-S7rLy~UB(BE6D**w7te`9<ChDd!oUZM;QlEXk%K8t(&K3w>&`u!AK}SQ
z<=@a{qocPcg=bMLp)uJ4C6?N%gacS+Y13@dBKf)u&?jCLGAVE3Wm%*EO)Q=Qt_1t>
z1YilP9b=!=4bPOpWrF;StY`19diSH^Ai7SgF!P5YT*DaScYNRAgH50Q?KG|nKr)@C
z72}#OXh3GoYsgAGk&fd*$^Sa(L@wOrZ%BYX;}n6P;BR@(D}m=9Kd$b4(c0#NhpW3=
z08b@u{`AYs)nEVgv=lVjuL%77=l|8eYI?)bbe4#V1QrP_5|}K33w9Q1FuWmUc`E1=
zJ2y(1mKIiA$t^2h?BcqtEHsyBmGa!x4{L{8+AVg~-L~>f7e?aSr+&wlTkurkqnTHf
zf)>9-0s;Kk<m&_%QTWC<cJ-!pFMUsnGwW-fN3Fi)@rIb_>lcU89wK~ddB*azZ9g=Y
z{#A%KRYtT9OVg~dY_rG#4O=rgV*w7yZ7BSVi@vbFr36@PkHtYq9J61pjV<uCul^X?
zxW2Y=Oqn=$^}l_5*HfKFNdCR^->1Rl{9J4^)>s|X_y*wfopDY&VF3QthBcR^E_hvK
zv7gb*^+p@nNTNp6B(?CiUPl=f$>Mdbb^sylh%(Xh_a;0h7^a^pPk{{fYW`#&`Rm>I
zah?qP5wAAjS`_^;#+(m3;!SBSNkb-o?~-<K!+8`b_~?|?S+OjdtSl#a1n3gHBDXcz
zqe$y_SeIkz06O>swpdY-nZo>g@a8}cp2!$}NP`P2Aphln4S5>iqj=7Tp8mf*gHFgO
z@`tXuR2V^61Uyf#C)?;mYf{V0SxbO^;oO0rIV?`j`Xf$Y{^jRStILn?SGOKLI;p#r
zionl*@n8S9^q8Yvbhv#9(m2sk_r+hruDJw!%sz?`OhE|k;3ynipG<FPi;gQ}jp8&@
z(MyvsQyfv^vTw^oma~!<;AbY`50EK+n}H-t@HLSDZBbJ=JK~0ZAw4!_o5hH2(xjq>
zpP197{#WT8838xv{o<*VabMryy7fcbw1ufMLT&r)+c#U3qW)}Cr=XE)1KaRT-p6!9
zavKUiChtq$$B{r@&zXm?tjhxDO?K8F*zK#|9V$j&+o<XALo^Q&-pBVL^4}$oyzb?{
zccFdN(i%vkIb;s}uvv+9t^jA4Q!<h<1_tO5Ac#DCeN|K(ZL}=z8VK(01cC<%4#9(a
zf)m^soB$!g-7Pr5-7PqSySooQ$S@3W$$!szxexbs9(va5-MzbN@A|6dYtp@VSA>tY
zoEq-o3=~kYg%Of}SlrEpKo6erCuZpkC<0SWncsBxNml%FGU*M5ElPH@yia*+iS0VP
z3Wy%2s%A}CvVY=?p7D9aev#u!342cSGRJO2G3ii<-+-gKvQH6=Hp9Wifn7#LCu(p_
z6Bzc#7Vxg4ZWt>&Q*OdD?HJNZjK0~&cIFQH9fumfK$9g}Aawb5bL{9RfyEnlBLm<j
zhyi}(v=vd}@VRp%;IYofyLEL$%G}0zZPnW0po@0vaOb&iz96%GU$|r~vS+dzqWePW
zzhA*6qC<_eisQAQe5X7n;slx>vJR!;dF<@CiOj2l(#?2}+zcs$lmf?GSzL1=WecIj
zY?qM<9YpIR9Z6kZO-lT{QrN#0pQh)x>4(0l9M@}qq}VlbaMxU;eJqk5nN@0ni3y@l
z5v9v41}$}6pSE)scT&_<-Xwce)pVF<MV^<<E_+Gv>6nUYQFo}Uugl5lUd7~=c>w<J
zMxg!ERms9^?V0_bD9?0h<?(XDz~fZGq8a8y^T6sN6{mn|&CB$aT&yd5j~e}3Yq7|S
zbDD5CBu1;R)MA7~I}$z|-Eb6;b);qg_W@aH)mk4HOY-1<wnz-6%8LB;#~!cL{Z6zE
zM|h`9Bz|efY1l9Ox?K+SzA_9gz`AsZ>T#d?ESiskhCmHtl7LHgj?r|uCUW_S>%boV
zCfIo3%3f~M`x|(Ft3Qs>RrW_trs1^^<$dgp+~4~6Th-KV{#+HI`aqi<?`8s28JeXY
zi6ess4vrbgJ_rHpVXacCHy&@mLw$=KD?Btp6EUEv^M&atXw?9|lK>6;MIMi7f0uA`
zvlGzjb7|9DuzbKly!lA|e78wD68PlxVx<B#+w7SXDY3i!8)p`&r(`X0Y<JHJNkntN
z2WV_H<E{K+WcBasP~{gXWd-=_{A4tEWI0de-qdM!6#m$O3oVfXt6z&IUvv2>Cw|;e
zEh*Y?vWG^!ejj6_?CREu)0%Ij`!-$v(MO{A9pNlZF}X^UIZX-)i~p!Rckle(R@TX9
zahx22i}tykuBLdPMXKL{>gv$_8i4aAh8ey*q~dT%>Y3@XFYR^5(7C82uClh}jm^OK
z48tL|U2w7YB>NpviiXkzcT_5kHUnZFwPjeSFH&I&{VIuciI;wzBZU8*$nqT&TZ}%P
zt?T-yP`rru&)0m)h+%kxyG&}L_?ndLq(~1VVCD`>HItLlN|5Z=0rQj(Q?6Iu$2-}s
zEI~r42&Gs{wIm8O7BnVrBsS3B<EcC}xf{M48a1>1%qBT>+%)%=JRl=~@;5>JOIe%$
zF{s2~gVm2mE{p8)eb@`*xjT9X;K}v&=;|0wBIyfJB(+FO!|sHvbgo+Julio8Y88nb
z`A*kvR{nMzSIsB43yS;k7oN8;*<gtmD?R8BKz6=^T({*Yqk&~_vd!E4v(rZQ%lZp}
zRyTNZgz$t{n|UsG9OhsnT?IGsFlZ+>ZqjkmZFYyLvSmQ>W;pb>yI%{2>rycl&PXW(
z@l5}YoT$>(Ae6H2jVFgwqxZj7S3UjeOyF>`CJq!TdJ*@0G==VuXE=}YwsJX$bYVmu
zEnY3Lm4{@oyUrO>T)o}RqQU6nhvP=*aqjsYPev0eLR6d)MB0)0HtM@SRPiH-xUT6P
zupVCWyMokYhdLhs{ZIzmyT*JsUsGwf!n*ep;nmsO=Qjo_&Oz@uj`69w&mjEXNz|52
zNh%aaQY*VczZuxg^?OqmV(q|wPt^!yhb~`cFLuzC=rDCGUT|;*@rY5Q;XKFsU~Nf=
zr?zzRe2464q?F%TG8g?<KAwNq5v8Z#z2oNeMxuhCuRfZY+eC@+XK%&jTSvVbOG$r%
zK>(e}1CxCObr4Cq`qz1<3^^j&tLE?<1Pc=j=hzt8|GG`J?l@7}z!bW3Mt|DIdZ}q4
z#8;~JyK%Cm$SG}+b;at)Jv8h(2dtf+Je0mGC_lLmcSYf@7Crb1@xJYwKuDQ`JF?~e
zXdd~h28YKD8BJz%9Oei9asTnJ8bb5D%G-A2>#hoNMKDDIbKC#w;Xi$TY$&*U`dnKU
zL+<q7`>`J|B9_`C;J{qMb9?^}_njtLx~{L>EIM`DcDb57y=JduoSS>7gY=|(aIf^f
zpdz8yz{R0TP~?Td)7sToiIE!*{qokU_~_KBx7iU#2gkItL*j-W1H=CExeF5ba+WDy
zm+sW3uxnob)x7uNvbKHSi&BOA(s4hi#6njS?n6q+Fd`ONVBDM&yEMV3cWAZ3W{tTY
zn`<J&_`(LfmU9sI_T_s!DrA5Sxl{*XXj!KrhC(q?!shu|ApHdwC%c)&sinL)bQ0h^
z!?ozqz<%$leW3720w$%eLb!CJjIqfq`09$Zu^)cV>@i2apMeY^ot-+qbaD5Q#%5~+
zH@r0VCre1*u#Mmx_@BTimGKC8KY^53m(xUaKM4+AycWjX!qw5`(`HSzYPwVelYuj%
zL72@4-`buqDKR334dFg|e4!-KSgj7mRZ4~)y%9e2RJFJxaJoW@;c<&{i2?4O`Mvm1
zwhLpLgpO&qP@rRpw99e$kM02h-j!DuT!R-b2fr4r$5@_!OkG%b?6F4Jmb>AOIk?Dl
zBk7D4xEJT(qsa?~Moqg7_8;#?F#aA*D(BYcKUIIqnwR8pq=`ZKgGqFs`F)AE8`jc3
zL<y%@FiF;vSX1ronU#j^x$a#i!TKHLB1L<zL~0YyxF>lqlEwAi!Oc^%x3j=aA~CQ5
za8FXUiG2TtH6uB)G`^5EGF<Y9viBSLOOy*&c!6H>0M-{GRX?Tjo&#IE5Wjw=pVm{D
z-Gu59kMVg9TN<LjIEddPe<VIyq7?PPr5St@#l2=SIYGk}_=+hojz2ycX-fZiuq5K$
zN=WjJCMkc}Y2;eahJEaS+fGXL{Lj;RqiM^xYXxA4BkT;cIA?cwzuN3nE$$DqaeuxX
zk~mj(qJ3A5WYQT>+va!ae0v#s&ig4i_vLBer5Q>*`CLD5at`@5l84Oz9=+u#<IFaH
zvBs3$Cb`{bW4(O(EPkP$RJ+8Hlv;R${dhw7wXsbnKlw2A<$RYjmb`n?#oFyse7`kX
zv}DS3C#M?8{`&x7RMR@2Ba?=5nhmvJBY5+#*&6}%xz?fO*5l+Q>dc`f?;_F9-CY4Z
z(aqKOt)o22-9^-7DN6#pbShCrR}>RvR0UcRaL)P?U-3(d5s?OrW7V-mG;{tn!Q6z!
zH}?$;HPPp~Y`wLQ`)}03FJGrFB^-(3UAVN(!eA23@S*<n#1hTlZ{1`U8o2c5)CNB!
z{z73u4KPxS76H-x1T_h835amesA0qvw!pP<ak+opxWf3t^7qPVR#LBaFIYQ)u?Fp&
zG7s6rA@9XZCb~gR%F^XIdl^nIt7$TQL@<aMvMqR3^NenNoL5ozhU$vLhzb%<4FIGa
zMrLs+_wIdlt&@=~S5zou;_PLr_Q1;=Pp*bVd%97p_cP<}5WIR%J7(c>V;zruLvPb_
z;va~ZS-$T>7#)?f7tYt9g2mCg--eD0yVFhX@x7$nG|ukbTpCU58X-sK8oBB<iFN@#
z+;`n|#{;qTVQYam79M>*c>tMf;Nw5yyt$g8hXF&1?}U@!ff(Paa)%1t%n7{{`bwGa
zlH3=oZ+f7$rL;*J27kLn9^3&!#mr>X{bU*;ILLh+-O@gllOPbA$%YcDxQ>w`vcT0}
zPss~X5A4lFruI}dGBT;MWbPczVl^6%j@_33Xj1Z>JNDvJk>=5JcD=Yet5+bhgz;wi
zJ?Ir~FptP)7k8PY`lv(2lZ<B>Fh!7!zs5N6phhQD%mg$LiZ5_q5Dyi9xc7e51FaNy
z+0z#+?a4kZ6=Tmrfh$wT3oTiPw5`}-xsHm&6F7|jy~{VOKnvU1%uNs*R2`Lc@ly}(
zkVp7HI3hgP>j`cHINBgQsafwaS?sl~uy<niwjTq3rncuH_8_XM)CpwZ2dR6*%b+bS
z<u%4^;d9P1twSlTL|fdNHNphV<XUO*9f$*$3{vZVJGo@gpDlcyQvV326raM}P(Y!Y
z+Y;L12I#3Ds;@Z^ld$5pAvGQ6@tk6!Vo-^)+gaP(+oz3<L%;gnM}cEISnGxTp({o5
zhCdr^dDD$N+ez24)L{6FOXiy|Q7x}@%1V)#n&Gk-{@+xrq0P+&z$>L6luPd)QRH??
z?8ky0PQegTARCOUm~}_6g@s&OX#NVnO|vm9yR>%RRPt>YIRLmsSe@Lu)x_A!lV0LP
zm28Q+LA3BpY%bqmJkV#iglnTGS<`NB)$6oT&<rlNl07*o<r|3byzkYIJpn>-m}Lry
zt}%jAcgI7Rnm&_K<_63}cbfSi;%o3<FI?ld^u!z=x<*CTdy!)N+>~PZz2V-`>_=Y`
z#knXVJ0Ser6nWb6LL`ez0XXU9H^g%$3>gyGz%s^mNCfE>@NmpPknP^t?9>5z(=YfT
z*Jb0$f(=)7$<hRl``7n4it(;@$@+=Z+dr~durE-Wj{J?el%(~f4Uhlu5qw2Ze$a5H
zl7KLC&mbPv2ohTJG6#+_5W7TiKW(E+sP^s3i(<Ul<nV5IXF+4VDSSQGe;`6}YlUq3
zOvw1r^i?k+J54bLj$2Ty?(#FK{I>#I3w+IL)bK?OFf$`c>?NUTej2?QSIISCj33$-
z)Ip}XJRMY+J#;CT!G!1dyxo$BC?cS4adkN)n-hP~T)nB2D#-0ndQEGDX@VDrnX{9g
zf46OoFaP{Y^zj_$m*@^@A$W7U`@$M4zkMOq0n)*S0Tb}BBnO;jSgndIq;fs|P$7^p
zf3E`;3Ck+%Dtq#I1OgLhM*9X~=ezG!Bm_({itnfzcJw2?uwK_OMLwe?fBa(1YdZD-
zH0bK4@BxtLptkn5&rjU*g4eydZRsX@Swmq|RZbj@Xd}6Aa2dN9Mxgw$#NFGVO!uFL
z&3spKI=@P<H|@-Pe%+LrQiMpp=whhtniP3;HCllwmyL_A_brdhi2-Yn?ECrc&3i24
zT%ksuMs6`qj<Si)+Yi;%fR!%411zBBhyOmN08l|Pc%9hmliss&aiQ@}8&Vn~KX7nC
z{_ORb#-7;iqb$rj#U=&zX^8yLA;k54p!Jy8K}Su0;7KYt`v>Pm32U%>2gzzz1<oI?
z>rQaNYsObvb48_CX+54J#TNW1=QoC{9mPwin4<X=IzjYLHKUD(z5>hyA<C!ZXNsYz
zbU9Aw#88+!)dp$iy4tf!+Rq5HJ|~N#bO43#E08Lu-9d<vY4i4RZ(h?NG;o$qF@==)
zWtzGf8C>FBAvVyeo_`;}4xAQD=Z|pdAor5g`{jM>lwX4I;817VMc^J!A7Ju1w*G+n
zyDMmC6&GA5@qG4H@df<$XLGbFU!$pmu+Z2l55#<zn+7B-?7nNhe&wdnvzydCvg%}t
z<Fh$IzJ3)lFxg2iDA<3k+94|RcU|K3S<5dJ@3CJzMvgYl;W-#vSFvB+zSU<x{X9{g
zaQ9+;p_0O&o5Ck1)8ZRCl$37Wrm5{Q|K7RJIFW9xL4jMb%tPbbtA)jPI-^T?2$mn5
zW6REu%$=&ju;x^w+IGivw+S(6oOm8znah{CJSrlEF`S-l{F*UmIDyhXWu#arkH5dI
zo^sD31>E~;n2;0QvL5VbFcGN^U)$XNrrkU8p_=vdJ1GcIbz|`t>E@w$v{~huE4L#(
zs_vk{U+JI;OL%&UISx4aqxLpB^2yePBFS`DXif|4;6BHHIjeQRvsXQ1s0}4Kc{if2
zhG-n!=NxIPvU{BhrLI@?e8`D;q%@$O6+Hx+#8I0^La>Rw>ux7XV4^AF!)t_F2yiq3
z8}P78_-g%37ETVBM6&g^L<)h(M#(R)kHLN^DMZ-rNs>-~_O`L4k^n>;2w&u=-<r>u
zElwuC=_zFHdUH+st0NNb28qPr`_<-2Q#EecI2qF)VvgL_&jgZ)CJSaq3^3CWiD8+#
z<G(+{-O9@sc`Uqj4jXg&3A>EiV_K-ixKc)3S<G$ZA*G=ZH?xP&Pd=;hgAZ-0gIdeS
zqP8_&ZxvW#EDF-8Mx?~tZEaHoIqzq1OVQ@<@S%n4;T=<JOqa8|*`r4Re|KcJ$9F8x
zUNnhi=r=d^WochSBHG&;z@rP%APUG=ma&b){$&~^1Cn&m=%SD*)eA43xHli?As_37
z_UOVYxZ{Stjow{9Fc<4~-Sqq3WMCxIrNiT}+X?sKw+e351_JTehtJG7vi_sjcEtxo
zUM{2)kx*%Y&Emz(qxw=Oq5*n44Fe5W@h$jT`grysdV8r^bx3`P%VLX4Z?n;-#`8g$
zHKDrH;U{xHFX2A6aG#Ta+MUQm`-pIgh^)B|L7VKZ839>)+C6`aw90B(bj(gzjR3e=
z{G2%VP<OtYeM8%){8M#8Jxga9%;OfJTjhY+Vc)?Zmrh(|x8ARQQQt@Vco`<U=;EvV
zT7Cnv0@aSQ$F`F0IjI*O7m3r{RY#XR+-sv<;qU=9SVudt0B_<xM%Bg+UQUSG$7s{X
z)9W%*)TyAl=Tknop`T&trPYG?j2iY&BOlQcj-&kg(+tWWjbPfhx(-Av6<<*4`@;q<
z36SxB*bbM1OZ_;aKGC=trG8`>e&X0TjyKI755l)B*aUc^R1|JS64S@#y3Hj&YY9ag
z8Gg{f$$)ORpp3Jap>J=wCQ=1esN)u7&)kR-V$<%IVJZx`iW-M!8JTBGQ8_Ip4VMc~
zVJ&De^52ODaEyqw6^ihJqe%YFuTG^f+}EirIIfyDjBx8NjvTx4{CRTd!sCs=)3Sf5
zI=>ia8wNFgemc;2a-_aU3?cKqYj{nWbUEb%uFn3FUkT&e<a5V?5u$IIBHQ~;`BoBS
zv}{}fG<-cMW5Wmv0v2Rd@8FO9a+%PzV==g=*RL2=w%TV$9W_Sn`trpG5Hr&aMFj>T
z{{R}GQ*zXp?3VY?Pfo10hUHBjQo5mSkyFyg^$z}REV{+_Tt^oe*f$@27tgGDE(tbc
z9YF1hA80VA5;qnXeq(nEST10p3bB$AqkiesCi0JwHg^KIMK~1{c+*yli;c}PwC$fn
z7-Upe-(8Ab)bZ*bSF&|B@EU&d6JY)vw(H`eMAs?Q<RWqHl3t>2Y&_)9P(I4qE@;H;
zFqqtRtHI>8+0kCSx#+#zn6;WGPhI!|@O=cw_zE6z78X^@GMPnCrdj?%B>S41&iwAD
z1Df72{T5rymoMJ~So(zho~URjsrsv9(LU15AS`aL0Z}5UlND`$xVzldm0__4djQ&F
zENd{{z_-KIAc08kM9uJp=Ega#j3$`dep5WKOy+hg(McIirlDYojR#SbyKe3GV7K+3
zL=;RbTY8W}m~Ko&4l?j=h=bC#LO^^Q@Eyj(Rg0wM@AtE)4mZ&Hrvnzc@RUH;E5X#W
z8soX`?OTIwR(IJW2Vbs=5oRlAs_Y6KtE$f+&4Kq`o)lydo?`AJ=JL%Z?GLE~YIKpg
z`<a53a2ZX0<)hdrFv!R8tqrfO2W_wXwPp^2Cr>hxx^4@`8+2oe&D(sx#TzsIaFNX$
z@l`EUOBi^ira~=VhfiRCdPU3AsUS_&@BBEEsvW^gAeo*kw`5YNa|EDIZ6j53(c!@S
z#{cG57pM*0e<g8|Lu&e|1^ubmvi@wLxVw>6Q+}fAC?z`5WAEsLn(pn`=<>c$Qjypo
z$lkPy-kh9LWAMUQVqyC>;acmpv`>u3W`C4EU`P2IIqH_5e73xoBNjQfQ`Khj@~AAK
zSY7zI9<Y58(IB|Mt+xk}cY-HCgBv7>lM)#J#vkuegWBJt`*-cHMJHczXpQ_7K@UFY
zhc#PE#Oxs%84=djR5}6XPdTGC3;04kYyRcWe0}ukA<})rR#rG@Idy$LM;dTIw^S$F
z{BK9Mp71okI*z9F<HLh+q{XDj0Gx5Aw_v-`0rx(NKIG4|*$8my?P50l`?Fw=4|?qC
z0>gWC!YbaR-bjJ&RdyEWE|S)B1Mti-NKUw^=|U$%9(gKbgeI>Roj!6F!MTgYi%hqa
zuCdfXxt-;}J;-V0Y1rSTDVc??>n??@5^l>gwsmhQg520vabxAn)rIQn-quP{&DI05
zVY|LMaan`uP>xs~>U#}pzKe665kDU4Y5H)*t5iowMzb^cjsH<|E1~8uHf+N6l8)7D
zljLSvK0dL|h=9bK%TExkd|WEExy)=@OjEkHw+@EO5jX3~8SB8#N*9|{zIi$Mq$}$(
zYOk^hK%IcUr`9nix7de;y%C;)PAcpu52N$;+=F%UjT7^E^y}o2vum}X^+h=UxWkO*
zW<!|z`#d<rcf$pNdYMZ*-Hv3DsL5zAXDg#DxGo)q@!`VR4()v+v{x8YX#Nt&ET4nE
zM(42LT12IywMH2Y(S#mu9|gCGuImQJl!#5MAuA)VWZL4=W(FZxeJv}zyfGihmXBMv
z^v$QXP}$e<FHO+jKH5V+l0>Kaj>W}Zi$XTwr<s3*zl9nUzIA($<;g=?o@I$^q4-K5
z(ac_?#V4IeXQ(3yG}uq*GAKs37EO8}dAo3UMB~3QCU~;^R?S!M!<t|$#AR)-jwk8u
zQ`!8v0eSo-xHouyV@DY`NU*uZBY{^rZ@t$oG6jwGMNjNzoVXq_wh!W*huO3g;zj^x
zgrinrMfVM-rfCRR4#QSsX+5riMZ_jveC55Pv~Qkffjd(MWyoJd)9wu91F>^+gQ_$W
zixf16LvmRHx}Eg3gLxo`_lBo@l5QC^+^D!Las#}+nMNIjbl&jYj8G1Ph-iIteV#Ip
zIzi9V{9@IbndZ86OFJ@uE|}UJ=)q~D7DK;zMg)lTSB`%223$aE75!Tthi*Fl(EHw*
zRJ_6^AIco#y<<?o<jJdMNqO3yr^+!Hf3Kh_w8^bm9SYp6^UC9P>&{9$?Zm(T#RFns
zzvpUJSbg+bJXy1X)r*%(j^4-nzdEBs-ojCD2p+dnt$iAZb!3zhnH=%hT=;m6<R=PG
zkJ6yGaa(Q=&BFG}LydOsmbtlkhwUQy=7So5#7Tm=7ePL!CjhH<JEL&!CM<XbpN+f$
zk%s8xH^7gK823o0Iw^MqGu0iAACI|9)x*p}NLT#2#o}l4Li)@I>hM0)qpV(jkM1I|
zN+|GKNF33pR^@*Du{TP*c*<1SPzuxy6K49Ogg@^9f-bkkJJ}MZ6>mPkW)C*jtvp&Y
zz2$1S^-Puuqd~70;ikr~xo(vEwE@~Lr|n>?fG-{~laJ~MGz3)}IFrt|DVygG4=M0^
z+P5c2kh0mYyYOuevZkL~MMQiRZygL$9qv+7`GTGP+@~f;RAWa%32`l$tdv*k;7#l8
z7e=h&U5?CWYSC3rE{kq5Y@(e1k~EJqc@&EYa8fthE*>7`IQ^{|wwb|cU<^);&*CC|
z<Mn#xy0h%f*|8VdyD6qy`Pd*bNBr5a(Ts&5&|DFg@f>}z1|QcTz+?v?iqqK-_0T90
z5Q!lE>!(rpi({=BFj>dM*OcA-$w~XzV}JAJVbMyooZV3OA%a7+JzqjcBno&g5^=ay
z>>#$_W3|uLE;tCQh}hq_ZCVL)p<Q+Q(N_G5lE8NdN0Y+croSRXA}T_HiH{K{jN6to
zr7}?Cf?;-X^wC~jd)P4P?}yAAxmh3~`n`|xc)h(?X@w(fC_=R9*+}R}^!<uH<F*^K
z>3*Os8I~&VlL8z|zhsO18sI=r%JsE~KK>JlypnG|JZ1@^@spsB+=Oiz#}$%y(Vn!n
z3?USR32Y?l-Q$0l;n+j0;VEZLE})`4`u;NYd5-Kx@d2M2*e|Svrbh=bRYX?_*Ug9b
zi3u4n)@{YPoaLW1F49Fwo2%4(TeIVya|Z#e3SaJuu24`y9!#IBU&J1&yPh-mUSP*B
zN_Wp^-H-R;ZH>Rv4G(PJSr%qVCH|T(P8??1&m5lDW<R)EvkJUIH#X{KdcAWhhpW>c
z|06Nvr+n|(GvfBAM?v+O3+{>Ha+KF{3*vig4LJblTGD~uDDiQ7ssEq7jY2i8X)1FO
zsFLXe#>EhmtKhV$#_9m<WbYT)ZKFY%Aich0#fUgM(!nOWRe1rSY0|~!NZ?fI!ogjJ
zG%y8P%l)u2<=`XHv_ng2Awl%(xMAAPsH+^z)Mk^=%>HmPW&3DfDeapHFafq?&ATjj
z@>pdjjbvBux_73dkVNVkP5K_%Dz%<Nto+r-{P6_?HU@fdcLRsZkzd?(fZ<<Y_lNms
z()3??gYel9-{pnqR~0U*eL>s`zLxraKVM4WLN%pyCt@>7;wvy~3YR0LmB*T{Aw!11
zY0mOSiYw~fF<R;YePWd{6+Rk3lZ>LGYq+tPa-*4+I4Fx5FQVhM2?)QSTXs4I$WYf-
z#f5%x`Sc!;EPRwY$1J`VlDhcAZek=dQt0_=J1w4eTG{|^9Y1f|mrLLH=d47jhH>Hu
z*Y6_2r!!ZDa*01|?zEc3e1DQ=7T@nQB1$aAmRqz6zDPX*Xw*z7key_9--Qnn%aF@-
zK*a+5oHuSIt0gm6^CqLkQ*q~1{uW1~ZErUk;*w{+Q?y+@;%*>SbESeH-CX=W0y<D#
z=4^sv&PXJdcEC)ojw{4q@5bN1y3%Fd$B#b#$rbR^GFUT_^~W%NC&_VQ2fzgh@LGd%
z91t(0J#4ElzgDDTLf1Pnz?Po{lBQh)mU;b650C4ea-^rZXS~)HHoSJbgp>Tt4?`UP
zmM?h@K34?Poyqx3INVj}HH|%eOXg^p)LDc##{aX2-Wm}YIXJNMrak?8n|D$HcROn1
z$KHYfp3P>vZ*M`o6Wy8o8%8I8YRUXpz=EGCEMfX5Lre*I_uRP_DlDRtz_c+bEV{wN
zNjDE)?V{#CH;zT5-270rb&q%F%cYt8Rayil(ZBB1gIv;mgN4uSFf8w1(`B@4*D~+s
z9h$u;A3%G@1yt;uEfV5u_q-*4=&Nrf*Aw9+I?Q>?E`8_R^CiDEe##C0ba3uLLpT2H
zHx^0dEn4E8V<7%h3{(BnE{C*&(I&wTIULRa`xCmwG>Mdf2%LxmD(jm6vVkh^r6+bE
zkOQzZB!7{q)@w;${r$Q~-vfW6NYm4f%*Z+fQLu?vEkjkjn&6?2f{sB1`>;{mXWc_a
z7LI2Q2s+!}+C7v^w!mql<Azt2jBPvUevZ=7@}w-`ZQyiY{1`W@0W%QKK0ONy_{_&F
z#dUpQz~XawmQSVPF;l`}JjWa$=Po&E>y8d1gr}c)bpyR8`#S7sx@f?x(6oFp)V^xr
z>7P!K+sS_!`2L$(kzi$rAq+N4{Q^*F(2S(+#y6$EtY0{CpV32n-s1LA5MWx$`)C^C
zhKOHWLH>>AU-M5IcbXmfMztzx0C*6HmX2|sSW45<(5DV0CpfKzgSx9bMqS@(+~(gM
zhb*Sq2soH_6PHO7n+Iw>>#tj`R<epQHI4aSe_ZCMb})YJ39AVuIGe?1aq@1ZE@H9w
zRsD_A_?EDZ+uf?UW_IE78=d5hWD0`p_4yvI7@)CD`{@Qrz3S-RxfHkb%+bz62D1~f
zd$U+cop|Y6pP*c^3ZI$7bpx(o2EFSgHt~3j4JYB|0skPl?Dv8<csa?({ORs(T&Yz-
zhRfdRN%r{I^8CtN(n47QdIlaf%toY`X@vt(v~SuJL_V7=2G{PZD9HlyVOeCRAJCyZ
z_<7i@aCkk$9~q2FL2In~ZFObhv}y{HR!c*-p<%&7KV(~BT*bU450vWH0`GO$fQ=!l
z#fg09wp|ELdsulfSBZFQyIBOvOxZ*7J7U>xxbRv&I4~rf9CKg!ph4eY3qLr1ts^sv
zpO*VXHtn8oX3N~Vex3S@(YGN5oE+uZQo+SFN4~D~jP6P>&px=j_MeI9EfMnB(y`ob
zbhjxGSzEUfDXQ3bq&MSVx_LE!I)J)1D$+lwd`5v`0cJhRI#g_0O-`nqpCCM3ebPTr
z)iiAWqLQhAiLV?8E;;}-&{^=Y<k}Cjc3yFz_O5u$xAc+aZD-Qt23F_nK2yx^6i=^z
z1H1Sl2RQI9Ad9P|==w}VNH1&=;Jn!qG0`dH#EMe{+dp(5oaZmi9cr@!{Kih?1=Z*V
zo&;i<Fjk&wwOVdT0SR}s4eGzzkmyQ>9SGB*gFESLtmRMJD_II33!_W#8-@2RIOZZM
z6WJz-ye=jnaK=q)Yws^Gu17%gjM)24SPtt9?yX~i?+WMnqV#bArueX`4O`%{KN-uy
z3uj#@ut2>eI$}jIHfN}jb=NYJ0(6&LA!>X=4j{E2Jz%i(__k_&ZDR#WAJ^nN7}_6&
zB0a$LCxoA}?8tXdMS-=Mb~lkZBx+{*?Z2r?h1NT)P2W<FLf$XLYFm1;Qm*5TatIhZ
zEk8>uLeY<FwS{kt4zm`~0T1=Clc`@|djs)e{sS=6WEj^;Z<c*k_210U(-w+mFF%t7
zE0->&4&b#9H5ag6F=O)sfUizu1Xy3$(&6i4SeJCg3F5ZgkC<sIKJ*)fa@mH~@hly?
zbLQSk=70r$fW#uLyVbl;AGnE4J3_u9)A1-`u%jZdFlhOy@~6G8L4gy3qm)BGU_w&z
zQX(j?d#c#X&4VB%l?Gu9!Htj|Deh`Ek~-`e<O(^^d8ksJX}$HXyQ3ut+o2~U3|}J*
zgAe3w4fcxLXMM@pfgi0P)RFPbSy%-|#?9LrZw!-*-C$5leK8MnFEWQvThvg%w`|{#
zdJ`h?=T9(KsFbdkav!tm?@ZTnE%vVR4Zk@#Bb5>lkj|EVJo>Ehr*hilKAC^ungGJH
zNUn7Gm|%WXxqq$yK`}t$U_M+30m{}^GNPG6#o`WLU4-SGE*~Vc2YgQUUaYfF0~e%@
zAO=)X>5NPVT=HF&NIb=5Js+@x#Rq)WzGIgVp1Z$vrc*O`^x{<nqm|T$Q33<jT!(Fm
zTg`z5+m_ceJ1u|wawg(MOzj+Qz_Hy7k%HE82WOe|v+%@gAdi!`;A*$$tKCNnPagw?
z)myD_9g@fuY!t!Rdsj>1g%&b+2@fRp(OvQ`0m_?*Bb6&|L&Vjrtw$s(Y;ey${C!^7
zb~@3IKBR3+&@<y4*9E-cT*)bX=ESG#A9JDx%cxGoBA^f}Vpk6p`h^lTy7mJe0UsU(
zL4jpxLippTe;TpMf%(QWU2i$jcG1||!2r|OcIdL&(GT{fSZN3_ciL^bwJV^1r9_~i
zCfFm3VE_A)e~{OI?nCT>(pfXWW4*D(*66uzV6c?HioHY7S*@M%F*QMrhg%l$Ml#2)
zskmrBh8zfwNoX%RpC7CbJIYx6gq!rWbAHH3?Js2D@s?wqhz1U2quzZ#UDJ%U{npIq
zC^$i>13PYN8lDDGOlv|ix{<tr@aGSSn=jjzMplv9f4la4oOPYFZqr4LLZK4u@>%o3
z$Riwv8IzmMn!?+$0aGE1$t`#aSall2QXD0RN{Yp@EHpnDsw}agJW%Mi4UxR9IASmY
zAkp%vwheeBqC)xuCV9}Kb;^59`k2%m7$rO?91oPBe4PJi3hol!D%s39W%3xDpLFdE
z5U56}YAdA;nth(PtP}N397BGJVf;7W>(o-|QS(xnaAWWnZ)7@&v_E$H!=K%S0hO^K
z#~tbDU1qlPpZCgNtwgzAsVP-k?odwb&hJW*FE_L`C2lmHvQEAK@D$1Y(;FR5Xv6(<
zqiADBD^@hs&qvE$!m<9@BsU*0Az%r-u3|TzZb>U@uC!YOEW$uh(jFd5jrQihb>-#t
z^;viCtElS%qgCx`n<>-Obd~>#B|R*3#j^0x)LfNks7Bm2GX5B;mq=23tI^+H$I~h5
zLH~8+qu(CO!)c63U0`FF`b&J+4f{)wqL?EPMwKiX0Wkd1rRMw&-v7;aU^U#|FZiL6
zoQPKWsR)amh*BMbdRzTyV!_xTKHqY01eIZ2?nI5lOK1I8n`B-=n@=GOwxRrZW><5X
zDIaR~MaHGYG3_Kd=gn1^rw3-sCX8@mvA~F<s6f&~X>p|b6Sn#N9g(U<0!kp21tYG`
z=Y+-85ugR=6dL)ESKl+)&zo`io3x@+YuKGmeEa9zglX$=8j8g`gSLatA25dl8C53{
z&+1^>Lx?Cn6e2`}5|E)6=Uz1KaB0u}7`~dhx8A#5zf-=qP5N#(MM}chh-ovMLISwD
zhtPRyY{DQ<d2ux>^P{698&LwF{4+Bsmu_773P84WfF$9BAhFj+t46J9^d5V!@vDCB
z^@tx;s;Ghg4eG)Cn3t)~&)em2s-a>D6U?!7+Ou+$jeRD^6?wq#9m^8OJc%e$NSR?Q
z?X`!;MN)L6^qje&=#GnxA#YT+)4$0au^?=zFefS%`$rU0Au7=DAExX@$|i}{K{TgK
z&76SGU~74W9p$Q}=ILuH9#aB=n+~%&RZaVY5ROYPStqBM=Pb7^>&DMU7$%U}C|4W=
zIJAmjYLz&L7z&A-(iSP78Uk)L7}=BtDJmxdHr-6ZasLMt`=73tVXioFdcXMV8$`oz
z6kjJb`6_iHECnJ$l?tEPZ?#l$Ci>C3YO?}Vt=xCML@u&VY0@GehR6Jt)sLiC(TA%=
z!#NQu-M`{00B%`^e7r=1!*@3xQg`1P*2<G9sXtK58X?@H+00-byV>-ZEb_mR=Ofni
zPA0qLDFBTQ^{6d2fTr4so#_G<JnFjIM2Qcn6Z%peuQccWV3mg2PO;BS5mMWI$$VmT
zat^R1IMG|=z7nm*20{sKWz99IDpO1ARZd==1zr&sQ^rPRotF>1V_Ts9vVWUl<0iJP
zX?a(POKDqR(?2}-ir~MgstJ2*F5MQzvvq10N!Xf;_*Rmo4eoOr28@ZMpC5Ct+RdgV
zn{*t)VbfzQpz2>Py2W&U+C9wPpu0WA<cbk?83EwzSp9qnBHybOGDh@EQ?j_WJDW=x
zJq@2x)8S5{C4)o2>%E`waPEaCyzcGUJzKevN8po@>D`@GY8Qao0)Xbi^m&RM*^U`|
z2+|sX!A;tsB(h@Di^z+tUo2WVdHNHlA-=w=0|ecAChi{rvy2b3brGYNbNsvof<-P3
z1QU$m!Lnt;T;xc~=QVFNKB+kzi2g=a@``89F9e_^Aj>s*X|~DFT3qdt`E(9-<>0TP
z1guF}5gT~gfEa>u6YISSK97GPl2F6VEP;2`%18@c&Oo_=J^h{e`rlo4tPg7C4^xeD
z;GHNG#j*5{XvmlS)A5Wqdv?-+?wzoo?uBa!j@oZi@=rOoR~D%yj@5-6*@-yL<OD1d
zt;epO4p}}qWV$;1ZMQdC3q4-~dJ5x2cUC#<zkWuV&%>=VeT6vE8<M>#_W!)%YVjBY
z<<It-h_fxQqPwAr#8h{+Sj%&)mTkwmu2byQUswP0XiGa~5gejP4^w|cos`c)<e#JK
zFHsXmeEa!buPKFl!z`iS)|#z48;Us`W6pQ3tQ^L)Ab7uN501>#T6<Vw2X^Up2Fu;U
zbStB#K4ySB21W8R&bGpn^Dp8Zu@_AP4tm~SiTh!($fG5&sOE&}-2Gc+ofw2q%3yLD
z^DEFvAyOZ0WxjSYNr}Ukcu3fFArj!bJuP)DQC7EU;yip!T<_5v;);Nd&n90)x`?nL
zY11D5&Dbz^<J?pk+%0QQpA3%sgMu=hiymBa{x6W(pa&lmvB1*3#wHLWcO_fQAx1+%
zELMdA{&)vW7EtbepB+<icce<aAg=NKLY?;zdxA@s`j+~SiT?t8C*d@A#D#RK>ig9n
z!&~+3Eo%Di8FGFmh8DKMWv70z5GHtuDP<FX8ID=a4ZZIbRNj(1z!nbdYBhlkpu=uj
z$z+qO_fy2EZ|9R<Qgu0m4zS}rfxdk8p<nz7ht|!c+b1Cp%3&zZ?q+ns`z@&e*G(|g
zqylaIOMp|qWxeQb!zqANT~l6Q?>n`F_PpbATRVq70Cag<Crtr9u#)vSHuQwNxn2<~
zbic>^$TK?xFY6I4ti6A%#I5aTPWmY1^j=4EFt;{<BY{)RDRWNPkO+!IZ3j4Tm6*7F
z>qvDG=L_Cs>)bE0wsc~X^~@^M>{1~;T9`N1?qc$YvE5IOoCTklB<JW_k-L~Qcent^
zuPp#=e#?-mD)2dWiI`3{jcNKC*Na~4vM>GW;IdtLHJ!iFc)R20MNz2AC~kVHxXndG
z&<Cr0@J$D%&zh1Bf&#@7I$W5FwS!q-g@E248%J~^GfA@HF7=2>@Kw70w5CxU{CL=x
zfowU0cPI|SmW$sStgrSg7b?xYO;D+V|0LIW0|)I#>XcIW<ID#<D33z1)QT!+u8OkR
zo$!{bjO=&5+ndk~H+JRs{Zfl#jJtHZMiTjYM{+{&DPicuicJ|%f-Jp~@oJvxu6#-c
zt7QwXtBp!B4N)I6%9a<>2abA`TK2>a>!EvFA=(b2uFzwSm05jDJtdh5OViJ|;y(<n
zjg@q1zq^gHL4?^f|IX`HuBJEn;H$r)1S6dhIo@4&;BD_xP7!PIY4`6?!@0An%qaER
z8Y&_KhmKpW6`6~|>5Qy1IK8d>J=45wdW~}EeM$8R)ffyw_b4qO$i%bxEaDlN%e6)V
z={!rUrhAGiI|H9r)3xq;SA)PUmv6Nxn#aM(FqQh!H0;{b1&L(ZoEj@|_aNvUw)Xh#
zk6muf%#*LP0o2Xpbr#j;4-qpFJG*rb6YcrHdLpstGvJDVi0$XWq2DTx3IsVSU`IoN
zMGb;hRsy29nSm6hu;gZt$dEbjVHe!2dWGEbe{}q(enxBgRUY<5ljV;jfP#s`Eq?0i
z!3TAYp{c`vqZW_PO!reF6yHsk<@u#Axp$`zJ5{N%@_?y5v7O=LcjF$bKBN-J2!5)V
zLcM?p%%ylcL{9gp*)8PKk7!&wQF_DXwQ3Tz9+<VN;0KNUZEob{XYt|RSbMZB_H*2G
z?~{jSazft0N%%g#r<@Am{n!`l`5RP%#zi>%QQ^H|k+gE!d(n{{z#8<2(Ikk)V2~T_
z;+F5*aP9)W&e23K>9yE=HZ<&75+hzqpFpW$xqir-E&hRrx<oP+5BOME9TmfoYuS9?
z`Ek_)Z)X)o_qX^WHgo3k`U>_(#dvg7DAt}TW6Y3qL0aS4{=vcc{=drSb$8^T>0Noc
z)NsDNo$yl7bqZ4UO(kv`QZtMz#tl<`J^61?Dd~2Ks1?n2lQBR&xV^o?Ue5jVwYzBg
z`u{H0kpB<WU?i?}4ACSGa!VG9oA{)SErq(xWvm;ui26IKdaMyfFzKrlo`07k2v{8k
z`tw`qI&ohKs|#>hJ@zR53C!J{QkvHMhi48Q0><tNQUi?t0U53v@Mq{3;<hDOGRs|8
zYVc&2JK=qwh+A5nZ-Fwi|A=X-WBQtht@6WQH9RQ9M}vKJ7d5B{(J47V`!?nfu+H`~
z!LNtp4i2Gk>O;Ox+Y^2E<w(qqM<TL1!~T9$onY3G5|j6pU$OX=iEzL0ungI2QbZ?J
zIP5tGLl{>T7}`MQ>mZ6MKLBVTpcN>JzELH}8{9Vt;e1WYLwTD|0zjK`Ugo3&??P#*
z8FHc-bp+tVa4<;OkFL2n5l8aK85lV!unFq~XofW3wX#<9_(lV5@Q>B*5U0F5q9%8F
ztM}JDL}mQLx8@h$=VM>K4^=fqjP1@cW$G6GFj09j-^T`Mfn-tG#pb@l*s~%Rt<&9N
z?uyUBu6sMbHqcvD{>ytK_0f2`O5L|$ugmh=j42tq3B8-B(1n?4g~ou6{?x$JiMT$;
zN;+|q*-D-I+7H$I*NyRDArJT7^eC`-T!^U-dB;h!!f3<cwiS54!v)M){V67!&)jK1
zcKi2H<MkxZJi4(&Nn5dQ36B5auc3j(IknXoY5(S-w4>(#8FlHM@M*nN{nPlUL&-i-
zOV3*y`a!46MY%Mqjs5TXUBMRBKJKaO<oen6<(*(i%1Cx<;Fya8)NO6TywosFR_DIC
zxXyrj4yZ$QY3?(-Hy4yaDEv;?h~%J&jl8>jn`gZ0X^^IRg7VZmL2W;jiUciDH3%CK
zogo*VCh5xSZ%2V)cZ+8!!Q{6O&^?i8x4gbE&u7UlIAvjoS>4}v*?p_G^|_@9f3R+w
zPDjCfx@Z_^yc&AxuDAH3NMxo6ayoB3i`Vj+@#HPk$*d<&Q0<WI%qig`hY7&)s@2s%
zTs?~g$q8;PS}m?|B%%IeO)=Vz_FbR2XUpe1{*>F>sI;Km>j08=ugFQD(!5fq5>>(g
z)NTIIvL+{^q-w(o8IY4&8klU|>du8(F#90YmmF;n)-M7d(4t6V&5HfMUI1tlQHLll
zxYuqXnInW+FL!^19tZMtS(Y)?STgk94#cWZ`U6`7F0TtIgl_nSh_wrxjESxs$R>>^
zD$l`xi}%CFau<KR4#1!4ul917@|WA<i_@F@J@^Y@DuAmnQUWDErOEwoU&gkdpEq(a
zhnEU31V=s`K-KEr+64(zAp}Gy-REAvdHnLq5KG35tGg$)t;yKIIrspNbNZOWBcft~
z=<*pk`P1w@QwD2dVKU5`0_1-<tgB!?pwa9RR<4Bnj_Z+#CQ7%OF=?G#B}ZEtRIYqp
z!Qs!zVfE0^xiN6)XPgc?V4tD%pCW#NSKm2YPp#TOp!vGMM0eC&EAayFpNHzarKbo-
z=dMEd*FgCE#aUG!m5f6}LLWqOXw7|^q<lzDMcw|+jb=!QrC@(wR!Gc?DygwU2mP8h
z7C68&HwjN0IsBvugoFJo)_gZ;8$dtaMyrs@i_}Fsy{BaPF6_tHC#;@i(D$_}rwK7~
z7ghR(qWwMQ%EasyTtMZ9mUb8mafsChv5_3cwBW*(PSoF|+1wcgC!d#D?0&K19pN~g
z{m4cfg1X(@3VzKLSn<7Cib*zzL+xcroZK{=2rxQoIh!F$Z}zM&;zD*vaG5x-Ah?+M
zUFNml>k^N^Zuo>ZOdH;?v4<dvyTJDQj9p6_yVW*d5;`9HnvWCdYJA`0IzBo!Fi(AL
zV^+t4!NMmc7+waBFA_QfA1|K$Ny-d<rJD|G+%7`t(yEik%`(MB1$G7OkKdJ{;ZA=Q
z1|ttch@s&NBUM+xe+q=RYIAiS;|0m%vjZM?0S6hjJZ_h5v`Phzt+Zh5qSrIed`Q(?
zD^1@Jgk{7=y=gYnf^)f6@+zKJ70ge~{~ku(uEzcTX&v9yel%npBW-zIqt}D*xA?aL
zapnvs`=*@ymDG}5_Qii{p9+;W<=w7}JUdU;&(;m2=Oc9u#RwlGErnyiOo{x-Z0#k$
zgK_|G%4^$^SZI0e-esykDqabxzHr-Vp#I9NM>koG`0!2_8BY+F8c`Ko;~o4)1b-+T
z+$AhFBt%q{e4Wt<PKl356QvE|I|X4r$JGy!{!sh9g%4&1g9oIO*YLce`VDb0%7loJ
zUjQNh+;RxMCkSlw#{Aj<f%JyB2wC`j@{g%{8d@}4VfXEOPX)AYstXS+*94-EX&xQ%
zzN&~bTp#46(qhwk+I&O|44pL(kV<x;82bpEM|55W*aa6)hzEu3QSXm){3L%+$RWxC
zIA@7Se%c9>&(5YHu@N_Wd`q;JA$WCegc+@EGBVeD_9zgD#os6PqzNLPeMJ&_-`=L2
zuuD+BS=p1ybq9r4Z-Y3;!?xS2bF{maV*x#rFVMgj>X8@N=9AjU%Vqad>=T@Xc<DXQ
z(b*2eH)6|xg+8CIqxx>GUmJ^CMgv}-E>sZ#4tfhE3hZQ=*ESx^$Do2ao!s`C8V$ZG
z;sHV1C-IX_)Yo11=z%O%vo3oZb-m?}4yw$qwo>2amL=q;+8LV<IMI#b{=@&tNqiS$
z{f7mtGXBZ<sNL3?mVZq5@Nor}9rA!IUS)dr)u|-r{_5<I)~XGyuR+umXnK379q=}_
z$bs4FDYA|TgB(OyqFWptA<(~~TM;2Ba!Tj{?w3cTVi3Sr4a~`nkFydTOq67yLfnf~
zxhi0Yx&UTXC<tyC>EMVF4TBLk{tK05PcqkM6wj&H`OE?ml(ylM!`}!fqDAf;u@lCo
zpa{t+Hc~cwzAP?VR_~+)bh;8V9}VE$>E8%Ns~FZAg5eSmRS&#=zB{!uJm{cZHbHBS
zDY&-Uy1xohLi!H9!hFSGvmqY51Cy=`bj_eRn_MNF<SdAnhg&edZek32YY1B_@!a=)
z*$;f$EbKGonq+mI=_w3WL`+0s!d50r<9lcO<^THwi5_C$qgSBdyP<V+!NLLD72%f~
zas8}pMquZ8P_9ZN6{3Wy9YTN+!K};w_d5U2KSjwPt|A1C8bw9R@Z0eOz}(|ucjF__
zw$r`quZ*C)yCCexD^U}_b3#aKfD!3Wcdiv{&h_1W1c|^G+?S=+fyw>(!@VyqX=8_d
zkSlh(F^7_D;Kk-cHSvg309YWsnsT)v6{DW}4uKV;sW&OXKd~g=&pe_#-n4(&tTy@i
zyY0rvayw610fD{_?=dL9+k;MdQs#9O`3fbPghXc6A^ZMG^5vLUfe_~$*!lW$Mv~J?
zQ;d!CU$L0w5(z>wLP}A)ipf_J=AYCywfS<*^YA3-^yFG&-cf5Z*1+;CY2I+pIT&cv
z&CmZDeApE2F?~SsL|L{qKOVgvYn&n99M9}12xSvrj?5`Aez>TlKOAb$+=C6>kLY&4
zSeb7Qm(Lr!a@Sw=73_*sm@4OZQ~2;CpR~|9xZmi5OZGLqYrnnUjJ;^(`qSs%nlMSZ
z_TiGw3*evX=R3wSYg+nGZWEXwR$_G1;gbWhG&)-WCkorvr7?zs`%<Uj`0X@2os!14
z?Gi3H4b#^;dDiK+7x_H<zZK*1>O$RJa!x6!m||0jN3O2Q#(!5G*6%4_e3j^uP{WDA
zN%$aU7o&=kSPnF-v^+8q&stui_Cc+G+Pk(&&f+2-p?t&u&ASzKdZn4y5l^nKhuCb8
zgC5(pUEQ0M&mAusxp<Y%8^iIwdidK7J#5qIuU91u&HM8bZ}#nrwu*Ku9-z9p30ghK
zcy);?#KaKWk+S$PQ?ZY@cvsvHO)e@bD}Ap~RIlq!W5yf-6x@gnYzss~LS`*p%Y`sq
ze+dwDKbDiX_wTQ6*bfd$Z5z0a4WXWdd=RXZJMwd3=v@DKIHKr0&y>C94=sHPZ}{wz
z=c48#Br3)ma49T=-wpgIftO5?AFCRNo^2Y~rW~_L#jJZM$>k<you0UNdCJl%++9~*
zAUHJg$)}^eEtvv!q_>x&wXH>wst~jVjJA&*8EG$(HFFQ6c>Hr1awC?gn#8!QyDF}5
ze=q1&OmHJiAQl6OD=>~p@f}Hrs$>esql?buGF@D%ly6~P=$>$UY#<MbGoAf?pFeqL
zzw&H+Xk7_=&Bn3@Bho+Ze=}{SIF5bvF?rn=8&|)$1AkLC%Job#32NsO2yhxQM9Sc%
zQDR7EW}d};Q=SulR7&RhL6Idny7}^OxL5shaIbCFx#Ui*+vd0RzeC%c1hCEi{wm+Z
zZCsHiAvL$fR*9PqqD&0X;sMKJy+|C~llS|C=jvD<(F@Mp=ivn5=aH-O?<U!MGl>k6
zA16vE0i%QMTtmOV8hrESdX<l)U*Zou9}x+AU(qM|vuMi$wM~U|{%L1purzb-$d1U2
z?70<PM(Q5Exs<J|y((r0D-wBIy}Gprbwk=NM2(eHsk5j#6C)!d_r^!`uNABO-~x1@
z5J#t0Wuldi!`)~i7?J^`i+2ctsj3k|eQ_&H`~`l?jm3E{u_x+1Ja;n!ozCcu^Gn<s
zM(7ejp0le@+8P`K9E3!{EtLbgh2wxN7vZBW!Ut7k^|5}iTo67S&PROsnj>dnB`yhL
z=Gc5ph_sOCOgLT;J%Tcv8|2`M7C8gX2M392-2^kQwe$TJs(QFHvwkXqI+`x^jbw8W
zX%I^gt_|b8L9FK5x-6W@oebu(E<!IA(zQ@&T{W)n#f>Dyk9A5uvi1>Sj0y=y%;q`p
z@;p0Or#Zh?NQ5uv5)sl$urv@L9%(*MB8nB1&~-Op&6Hh<)boF6Itzy;|M%<P8{HrR
zN)1Foq$H$cNU4BW#E_B@kS>Wa7$O29T_Q0g6_A!@lyr<9-Q78A;4?qp-}67*_xp<T
zI_IJ<K;P4WqJuWJ8tGmJ1g&u&Hof`5#+D<3iHL1ASE`Pmaa1(DxktoBjP!K7Ba?kn
zeI=|!{*~oN8DH_nnERzfrF0xMRsdmnYtMmV@8q$2#7IJL_dsu`mJE%bWc_Z}%95IS
zUEE2bs3Olqxkg3m`)q`xrO@G`{}s2)#k6OqTma7Z<Le`C>xu}6BHP1EV^Y5fy%bl(
z)3Nsqr{k)LMc?gFm)TGJE!b1?Ez<^3b_5>bzil*CuXlNx+vL@Vu%KmCf;mDm>Hd~j
z$50NUpN6e9twLhe1F<_%qREx<_-|fjZec*yS8KKjgR8puJ02!`Ma%zmLg8bMA&v_I
zZWeAl>d=WLz@aizjSJSVJ+@<<;%D*(YvZgkekZfuO)lDb8@}<_n4sOj_^e#9`Dbx<
zEE03OpAGN;zJEu4xA+@jUqc1D3IA4W+I`QGn;wbu==4Pgf7f#5nnNB5fMk*K09|EQ
z?J&a@G5yNHVK|6g&mh`h0eJUSt-IDN)xkR+WHx`g9UZ`cl%xZ5P~VR}rh(GQlLOq2
zGUZWflpuiwdR)(ky9ON}o#s!7Z$IbEzd^y%&x4eHdxvQ6p?rcu3{_Zqd}GW-cfKWW
z*aTUC>B9_SVujn?m3*DY?Nk<rcg@yJ4o4M0mIlWKg(`U8lDHHDFvSw#vU=JTr$@cW
z^b_5ta}6%qW4<j&hwI~G`6zaS(XC!fj@vg!6BKPzPl!q1xGW+Wts-aGA6q9j-3q-f
zDy=A|C9G(Y6j8jUtiQI<i_Ptf2OtyVb7Y(Ri@zd=ijTPIm{klsl~qQDP(a;vU&>~_
z(7oEFH8aluh_dT>qoRS+_m)DU-IIo^^RpzfNckxK-OOakU2F-y4$g4yhxOaWgw)}_
zOuh*@M2Qr}IO&}KK0Wa<Xoyi;3nlahN|J8hjM>>>c>CsJ48>%Wu%6^V)2j+J>G_t3
z?u$BHQ}q+2+Nd<_Co4o3-=VihcV>K|Q%99Bdho`JnXmeU@=A3$>AK0>eKpwEJ{H+V
ztZIM$FnU`=`y5_jZ;G`T^fdM8$z+RE@m_yvWFBUyJw20wr8wpHWHA~!IbfFhv=(WF
z2`mXW))g+48qvoRh!^j!m3(d6?H5<2aEl&U{#bb8ky58&A;@392dw0{?^Y#MCej&r
z8GJMDhVq-cD{b44<~nKSf(!Ik!0L?>#YasKLy*}N%zOl**sO5G{f`gX{S~O$r`Aof
z#=8$JwO4HA%Q=hLo5j@2$@>xX5J7L?iyqh)@Go5c<j8NwW7Peo(^UW<q9+qbODYU}
zU9C}813Cq*N&E36+Euc0lQ>UNy=vMZ!CD*4OV24Qx>0E;kQll+6mGscsXmw%_<Kpn
zqbBq8;LY-T>)KM@5I5<u&s%jN*_f?S-ArWDT+gPbBUP$65nO3WgU3;i8G`CM3`0wi
z>YfV^Sh(Kv-6!45tjychQ9_6b5U34H4-aAb?rHRQr6YpRduofJr*w6x5s+B}=k)$z
zqmO)_(Nk}k{6-=BKFhr3y00*FL$5}0oYDDKeoo@bM(;hj%i+OBO}P4L`)4b+mQooj
zp=_Fzp9R9+8LxU7+s9S{!Ag~^H!!rd8WZ0nUzsfGww{F<yf=-=b2ZMo=>1Lc3q5eI
zZP_hQ+e>kE!)e$+V}@n1Y8_wwPZKWCIkUXV4eMU2ja`KsFAWgDwZr_JlT+b9F2Y{t
zacXdy@BN!|4;-~jz3pc-gTMFRp&vLx=jF}QyU|$Q_)J$fv2lS>s@J*No7S)Bm{*Mb
zzWSy^RvN?Gp5gK%V#?!SI6i1qCIR8$`xToFlQPv+`0V~l9VJF}R%-p&ZZeF0@|ytD
zdM{#xCA-!Hfl+K&d1UNN_ovxhziHXU$Lf;XNw7D;Q#P5s+cIQ*V{*^rq=4P0j~*L-
zNGNTRyM;!p$KMOK{@cDb&%Dj0Ir;P}GqS)hXLI`<Qc)V~SIja|e$lzGItF}r9YDo5
z(SELXft_ok(Y%J+;s#~}iLd*fyYuX+RWLS_Jp$hV1gjq8$NfQm4MN6bLz6TESomCZ
zNJE^%`9qu8pFM<<DGxB;-Q{9f_XOVoZacG8UiVUlEhr%@=^!v-{#3x`gJyNMQGD-h
zGABU;fP)`X|Iq@H!3JeYC6bMX=-Y0^x0uKUgH6*(_Q3ZS5|mebgr+Tmy^P9a%vxM#
zkzAmc1|*@l4au$Avc_#TJ;nA3Dh;5QRrMMjPFT!)xoyXQS?BF|?7hb6IglzyvBVVf
z0y-d~gyP#p%RE-ZyGGYtp5LyP=;m106&@=FpPx^P+wr~p<TnPtcyvL;v9O`-(VY6)
znTM~w^B@#=JH-Z-6<JE#e94wzv^7NRtSTM#n_l>Y99ybtm}O?<3O~GMR1a%a-y1p}
zllgOq-Wrz##}}$qC5%kFDpe^NX(Xh)>Ls!jcsR?fd*{~Z(>%Gq?jNi!Ne$tpr~eyX
z%n=U*FK5sR`TeE+<;I_FXb4x%mxjtBS(@^m>c2YU_pAG#&h5N&&RrI|;z{#oGON&`
zd*z?jeaj)yxc>LZkYv~LZ*2~lz6ajYlMI#CpwlEXl<fAm-XZ3Yj=}XsA1zEk;RaT3
zR*~Y^B3fRXZ%_s1T*(pqdE@Nb;v|8Rw*dK7jPU-OP3hSB-z~0bHS&3Lc6{p`?Ot9E
zG#cOX=YJ{E!<!*;jevj`5F9Dv>}c9;#lyJHX};M|Trq1p-P-^Z5n_#gSF=-k8-5B@
z(F1RRfYXjebuP*|N{}<@9`Fowf9=uSfRNeegPLqQu87s;kUMnTt=!UOb^uv{G2le2
z2ASzUS*#a)>qz`81sDra<VMPsPB2j5K(YyJ61iaoad&oGt(-mI?<*pCI*`5w6iL6s
zY|1L>6Qsf{kXf(Z!g>I_^w?2Im;^}4fSiGoIHYO>gI0WJR)j)qe^*t6@)QL1ET^1K
zy>oHDbC&275zDr{8<<=v*V;>RWajN&bQn8K!&xpuJO$vn<h-5Lg7e-Q9FS8Jae=?M
zacvu(ja`t>1@E1Sk3O5AdhB=-v78?+Yf8B%b8*j?-6vO;EQ)V!kTSICMX5=?HbroT
zf2L#-5|z&3myBkVk7r*g5IOst#(>zj%m0<<SySOIC8SSUYX79x;p|@0=ZQL@2H}4;
zC0{hYK;#B3A&ax-*&kI4Bi0bUQ{5bQW6d3@clyYM3cTU_JO0>njt^;-Z+ZlFUb@Jy
zW;w8(u4kemOX&VM45^~!FSjUf3|GGXX74|<yjcNNHA$5+zPs!pT=#-(!vgI0ZA#9&
zX1|K}yswrlwXcw4(ZvR933)OjDVx#5TR2mCVgxfzj>uOzCP;p|*@+|7-Q_vwn=vQ-
z<?nwLS{;tm)UZ*i776I2`u^um`pKh`EN|sn`CTghW-ck-m_gsknlHDM^UMY#hdYJY
z&L;6}Yab@kv5;~BA@}`>zqoJ_KO+IStB4ubc=iCZRw!El3BYeAR>kK=#aLk4(Q^D`
zp9&UFpRQ(ti`#ca@*vsw>*8JeTk2UICmah)IDLlGqz{)C6yVbxiuykop|sv5lplm8
zfLFz9O~<T`m@({I;j1bK9y&N~i{g>@v&)kuO2=8j;8*Pu90D9`=HcR*DL9T7qKhZa
zJn4_%WDEgp)fcs94+6t}D<GCwj9U4P#at9)M`&E9-Z<RO&t!u!fhd!78RkCUHX*O7
zW&(ojq`{+5fhjD-ZJ*+rtitu%sBEGr3VRj*s(KYk++o0|ke>VtFkGP8af*Xoc<+aJ
z0yN74q;_c|;Qk1nWGBZ!tTll}tJl`GNAlavp^V{g^?wFi{|SUH1f7fde~`)lyrp{n
z?C_m_8}Ft0BN1S2`3WYN2!^)_mSpkYY;QuMM0Lh0W+-C;QyZ=wk>sf*Q~E-NXsLi%
z%V-CiGrL#&t=64(Emb;JOmnek_VQTo@`D2h*2>39Z!G=@8#me;CYBx)>TFyT`JZbZ
zVEr#AOSd-5<uAvm{>Zkk@G03YHthP0ofg^2{Nwxg^ZARk52Zl2^1M7~hTXM`o?K-d
z3G)?cAEkQIF?UMiydc4C!mE=1Xo4<#Fka%P(~yhD50>G7-;dNl6R)@F;N_Oj-BPW0
z6Z$D8v5O}Seoy@V9w+mvQ647r>QhnGBR>*4p)a!gbQc7IHxpKqLbw++tC{^6@~64|
z<g{7~<q0V1g9T?e*>C%wdA*sV-T4FlQEGiL#NIH_jxpT3ysdd%&<<b)c%=O)o9>ZQ
zaSI0Mfn7U@x=KaN-tb0&S!`AmwOy)gEgCo*X(lzvqDZKc<D{04W}&wRK@Wt9wNeot
z=6a6|&VU?g<0X=gbb38mBug{lrg`BsAWe=`Kq~f2>f2#m?QV~ML7i^?x!R>7ybRo1
zJhf9l7yD>bt^8awHp47=vEG2(Nn+vxKNa^Sk=P_l2<d$Zxu~EQl!Cnx#B)Ec$REUp
zJpgzcf52OBj`0Kl7QoL(a#3=UL?5EU6}+-k+<%jPdNs==ybASFJ$!cdQ)=#-JLfA>
ztgnIi^E*i11CZY|#h+5?qwL0+N1mg!S)D;e3%&QAnDsZM;m%}?o<bUub2cvf!?Lzn
z7`@a1FmU>;2%y^aJI4EQ<nbo&u^n)5Y8KvP&d9aSG(@pcpU!i8%hUL%OtZ*gtdkk>
zUf^BcNW=vr#4d1q@PkW1OBIa*ny`-g<Xw&K-rw4^;QZyWK+0$Q?PTZP-eF78YN}A2
zegE~%jO3@3HTFZlpB1ew#iCE;M!$Sbltib=kA^{F7nT0*X&wa6A}qe$9dt37EFtfZ
z`*Fd1h28akdbM#f%e&{zlXy*++paCOorqOUOm_gH;=?#+ow%7a;$gbzlYkL6tsEFQ
z%RT8F*#D_p+x7b}huM2P=)LJjp+On^4Z4}ps*^CA;n06}cpnZ<5pGP>VRuFeyue2=
zxhO(+MR@6-*C&S&mx_M^51wmi(Ap6`660lsSk>>ncNt51R7-3=K(<f2|J=^dNvE&e
z8gsF9YQXh*r2xf=i*Y)hA3K@&ySI1L0YwJuI@^*$8`H$3z?ut<c-%9f={J&9Fo4-$
zbdB5$`C1g2x~ZezloxdALTpw{)*p@x;Q^_?y?MKzIK5fOv-m#Y{Qx1~jPk|Ilb_gt
zzd6I+3nW?x2ZIVu1vvpVyG>RKYi0?gHdvZnOEioPbYl-R#cie*NOBDn#z~}ldvfEN
zCTGA4CmnkoE`TLT9)?EP_~P&OuNZ{Ql}?;Fr}accYJF;d4bwI@CG~kRcI9_hWkv%D
z8TcByil{+Fbn?xSCOT~?d@MjZI5kAyVQvf?bX4rjph$hp7g<`^Y);0Z7=kt88my1d
zGdcT`wgkU!rV5nH%E@UdzS~7DR<0TDEXBP||K?7+M<QK*Z7qo8n&YaoM=*5&a+J|M
z+$dLbd+{!i|L*B^%7nknh^c2Ld8lDURDR-w-AYxHiMaN=s&LPnZJ)#Bj@gbQhl*O*
ztNY+Ku|C?~-4E_t*0$K`nBlgL$}9bhOeXQeGXW6)be%>WKO(J?vwyYF|9mL2bKI$6
z<I-yS&Kq!~4DUQX5z|?_J2Un+)L59=lI42p$ykKa`ew@R(+|RJyswkwy+z)*pgwmG
zGQ+i|1|=hhq%j2Hrfk~|VqF`Ri3w#tXgx15{E?rBiK5%T2kRPacW&7IP-s4vyXGf2
z<Th(to8FQmYXim~DM(7hEA-7C2^4#=e_q#G)nT>C%skYA5aaJ#_v@ETtknqOm+@j}
zz2k>xt-N|Nx>lQDw3Qh4asK~4%$I42xdC2(`!de%{l0=Fbef_Bhju@iCUAD`go_sO
z_>RIC*T-uI+d?KQB^H6u<Q}W5BN3(B{+Ch>Y|yU#k~y?|R8UuL7={%8ah0PHGtGK>
zwr<3%R65!uyYvd$t?`QokYuM)eI&}5e*@kRl$wRQbtpb~Vl+)s>42TkCw039r7(&B
z6#(DZJ8zMMae(3Lz#qW~x}Ij$AAzkzvLNapu4fvK!<-~n&UkhBTpi1!Qh7><Vu>{!
z5DeQT&oUsV7S&9s=j%6jcpmur0f^J?z9zNvz{<smnM`@)as+b)4SA`1wc<;ejc}?O
zHkgWnV#gDWgy$cFH}frp1;`J|is<IABs;Fsec>@gK*(WIW1GS@6PiyEDem~&VpK=(
zg_60YFxE2KEo@UVKpc2A99qLIa9qtO?U62eohwAr48wg?9Hlng8{{rxEJ#u?uO3<P
z=aQpv`5`@&%qH#wv^;kInM^_H*luUzors@)eGw;eM1FO7Maxy41`n7yM5S7{v5#BD
z%>@shy4n|l&bJAH5vBd7ZSC9^6!k<(kYg($suio)v68zMk7P`mp`*Fz3Ic&+B3}nN
zfvt)T_Mv}lJ!$Cu7X+Zj;q<W!m1D|pJ;|LDWJQC9?@I9diWgPf=oq|L>qt4%&uovj
zz54DUAFZ==i&dA(+{)FP6;H0cfk?C&OeKiSn`NDdc<f1A9sRb?;uy(Ztp4Ax?V2*C
zKT-R~OE-|Ht<y?@%W(9w6T4~x@@(T2g5JbPx3Uy#Ps`npNcnqecQ9>=(4Wjw<P0|H
zsvq)D=pV>LzkMH8Eq^|aCS(`#MVg>o821j1YKxYW9{!s4`CCoe2G^A^RWnCI1gW|n
zPVYB-QD?E$zAzZCGRlb+wscTwStbA1&(-#dX-^i2z?_br&&-D~_x#{P+^;wVox(O$
z$e1sjh<8%f>-R1iIyR2CU0j6-kI_3MjnmS~wt6}(yk$;!TNR+5e__^KQs5(O>Fste
zKZXq!KjtS6c4U`tx)uaf-5=tV-QrNKVpITkR(-Ts_?Rq-ORA2EkIDwstRY5(UeXdX
zPg5RAdwZI%?-^63WbAS54661Rp+NY%?BuTuAM*m&0Y;(#Vy_<NQnBVd^*ZODr1w8i
z$!fg=J$nnbCMPW?H*nERzQ_mC1ZPtXSqle>-_<bv3f^QZI|UitGJEhbU#b!Efk^4o
zt)EW~7?O;@eG(qjT#>-df~ev=s3}ssn-OG@*YdEDQSTM8*C6v<x-bqT>k;#7S`gl-
zsEJeHEmuNf`-hEZ=iGD-zv=I4$Z;`(09shod^i$irNOME?t2ciYs4zxtDf3854Otr
zFx83@7-+xX{J|zW&DFrDmNKd8nU59OVKV_J)54<IqJ;Eqeb#c8cvHYplR@Qh`<;2K
zwr@;LOXnzv)1IqM|F)IL&bX^Z)yxX9KXu^Yjfr%pn>96L2TMKwRO#i|vzqdnC`3r8
zx(T9!BM~b(nXsOHxSP>Ho1H1ol_9jt8#^U?vKUnuIx$*pTP=J%9p@?8fMDW->>|ey
z!B|P!k@Z<v>OcMjZ1dV-KX$qOI7cttTS8$6114*TAvt-jv!jxt^KbK++UHLAeFXoK
z_s%b_)SU}PrxxuDiu*FwTkeM6FzDN`WK=Xx?CqL<aEX0X^pnZSnBzeb%Ujz)vXaxY
z<yF&Ir-sWm`5VQ#9L!3UV7jLyc;6TS`8J4R*P$~PmnF&FTj}$hT<Uh>Mnn@+-6TnY
zGG?*_Snk?aKw9}(97!WpBhxIS4L{eXy>1;8^%#t^yES>4!HgFQ#=1OBbc5cQ%v84m
z5~G&O1pS;}lFljLue3o*XOxk80<py3ygal>VIs!HHgsXvZEBh>zvWQ@%)w?3nZJLG
z-FxiH-f^6=UC;B3-0Yp{X5dL;m0heyOkr>7+RB6=0`SKj_wdr%eIMs8oR8!4fLII9
zsKx>iLHB1{^|&3;Mym1az+$2EdJdpuRKd}4U|St0AU$yP;CGg*zQbM%YxmUT&z_RT
zn)p+*H|~7>WZk!pI{lQyIV|)MZpQxj<FWuHK&xFGM$cgrnM)~P!{~VK9nJbRjX|Uo
zrYWG7OoZ=otBAfUG-hF5)JSLA{`VAhm6C$rskQX;&kyg7$Wbn8Z}WYeRVmYiB3+%{
zeN51@hjR_+TL{4^aQs$Y0_<yT)Fi;$Rr!rryh+IL$FI^;+l;`vpvbx^UWwqYi)*03
zLJy7C2McR(y!)H|q2Du7u=_tUi@c)q@A>O{T==@h3m)C-#^z}koaHUq#m<Z_<kA*x
z8AP9a`JuA<VA(b^Q}i`nqU8Q`R_Z@A5KXabESOv`CH1ofb+wG?G`;NIHD%u$(&ID8
zGI}Rkp{YNjZDk8yNtk9+b34sBa8H^A9gmH?PQ0wvgl!Jy#h~0*{vwFhJeE1cApc$=
z3OK=gw^uf^NchP#ku<Qf8(z1f=BB0iI*k$pl&G?}X}UR(DMZ`9%dR6`?$CqhkX0?&
zID9lo&qL(HVrdHa-ZLFRP1>hKT<!v4x2Dp1UvcI$HETghpu~kUQwu?t_Rz@UKi**;
z(M#Y{Zs5N1ka28pxi!sPc(o3rSF@rF;2cSsW57UUCM_3}P0uW0aA~&hCcxD(Rr6DL
z6QBuDK|Wmy*m=Sn2FQO@h`+-gA~MG)yv}*Nk4c-Voj!b;j%>vQ;@~hrlHz&Nr{u%v
z0_z`i9&*Db57)pl&1veT&#v9KRNlSr79h8Wcs9j4ENb^1xDVpp-vJqXyaK-5MxF-I
zMFBN$_hRnQL0LvMNiBaG|3YUVkiR1*XwgkCs6OyF{!x5b*1j}7S!PQ!Tz(+y-<9T2
zTM*5=`%8N>X}Z14_RL|$x_)kd;Z_m>gDXVECl{z4QR370=qd>S)AuQq2*xf+1|9bp
zu4|qD%|JIkaws4D`Fr5r^5euX*<|+`gZF(FhFTd#US3@<w!(1<<84?55fA6S4aASl
zcoX@<{vDJR$>}e2$_5WX=Nr)E5ouE^oLP%JK`AAh>h*+-+o$!nVzaKA$J8#Rrf=6$
z-P{-qndC*EY$ib5&>1fE`D*8)*m|n*H9Ad&{2JHe3fsTt_6C;dVg7gBx1@-YyZVDn
zx^1djF*lHqG$tR9T}xk~d~w@-=~o0k^{C{8qKM$*I7=hzw-NWZD}A)(n!vG`_(GnZ
z-R^n8W@546-+vRT_4=`gu|$II{>4zk)mQw8>E+wSi*ky;8P<031!-to?_BZW5*zfy
zCx4XN_C0UKuJidJl>d{beisf0$6|s8iSY#aux9vi61Br%*OH@()0S0D4eIlXw#6@h
zYP{LmVS2sa)Ixgz#OU(|T)HleBC}uP0oWyC$??nmmTN)8Qj3oT^6d3KX>tXQY&=ho
zoNDOoBF<+AD4;ixpVTN5raFUefVyvjneQhEH-a@$X44t?tyLa$u0|>Olj$fDsd+uE
zx(^rCP}Rcu7hD|$?_Nce8sMi)jKchi`zbVVHE;a61BBO2B+2ru$7b&Wnx1jdQ`{)#
zhrKT}uz7R=vc^O&9=~-jj;WQxox3P5l_&&Q^1FE=xSwYcb=gvdD*O^soaxHb5FQzO
zO%5g9F8NEFRNR#mR=6z__u={5fVeiIHLl4Q!t?#{fzP)N7vU03kT8^)(Ctbu*T1Ac
z)D?j4;v1!Q>tB$^*LS9){O{DC9-Gs7I!(fc#F~80*Wa;sFcz(|c8;~j7<;LIF%I^=
zv`ltU+e{XI>3sYi*P1|<HiN6`l(Qs|`te}=_|Q^OL@OQD*Id@QS_1=z{QjR)?;ikB
ztxN9;$>7OBf!f)#K~3I6vgo4&!n=1g*s-*r$GhV6w9j=UMngbqk(np>Jf=GA-p&Wb
z+8@?hnV3$O9&0S7<k}22jNg>~K-X#{D=!c2aBT9YQmFo(%w(459jZpm6DlSFhl*P5
z#NwXIr-Rj|Z?&^tC^#@nUa$sBCPi5b)NCUjPpsYb1@&HmwvFnoJe<!OgEb%XA;jry
z?$$<@+Imcm9;<sl#AON$?mj>>6n|Bi!G1_-LkHIH9hcpq-i!TZYMik4y`$CLnDr&v
zmFCp<SUvDfw3>%Z(0<k+#;b@<vtrrxtUfle+oLr6q*>m3p=xX)cSo8b(aGnkh!lNQ
zVJq6*pb7cRq+7f_rmI@yD39njs=qu}-syT8qv0@vC_oT0XW#RkylDqR{Uq`^y{n>i
z=>kO4E)m5Cpphn_8(x$FUImWmtMfm&h&ww^paW0zWJm=&Xl@oZ1N?v`aBqg3x@OQs
zj0|j{=wBs)bpjMf={v<pif#E5t^0!*5@h=58dnJ;Y8K4(wp}LG{;jD_^EP@Wkie(L
zyR`amPZl)iRbV%jS@wc~Vr5W)*BFK@igK<KH!1^Tz47r_yV#FG2z1||duriDU)RG$
zevWSjVs9P<#yqOEp1l=xkxwOeo~9&T4W1eDy2Z%8KivMcjqkVEpiJRxzxerNqP+m<
z;Tlb``p0l)=yfboyUN$XH0^9Y0O9`0gZ_A0AD+}2zohnF(2sLD0TZGU<O5alC(9E@
zVoC)4smET5At5!~S4+<@XD!QmqviA?fiS1_j6Q=cs~ig{*1PZukCyr47oF&`#nH2)
z<ROvO8A;FW*?4K8H;4WTkdNnlPUUF|pMc^M5r4eDR-FI&qEeCC#KndGsfK#RQ?M4T
z#N3R{n-SdU0CWo^rM#uYm`oju+N@*YG6B4myFL}jaWhBK+M%Hib_Y?vRE9PUcJ95B
z?_)zrE#F)vs64)@otmeqoL_xGiz1T4Rykycfs#PaB;&o^R)vK8u){=IX_jdv<28F5
zR>7F5yj^zNjGx=QsFGZ#J_#8m8F*!cb>?JJPvzlxT#iF$Yj%N8gbp^Adv{j!m)D!N
zAuP2RU!6gT!Gkz>_^INT$#GNY;?b8MfktiTpLdIRPuxsq>Z&p*TL~F<>-W++ABd=}
z*QRdWSo&s$7-33$Cp$c@@_v6{QA$Qn0iM2%DE>@$Mjrf)(P*t3uY%w>QFgrT_g5FN
z+}v~J>??bmyb3$9{AF8RX)_Yv#wocU-Vkvw24m$iULz}psU8UXQE4~X%8o?Ift27A
zVbL;4LvAvQE|JpitOPrc-XB`oi?<67=dvfFZlJ$w?YqhHsTmk;p|9I#`)Be=sq#`y
zxsE;h6x}v2*S+vrI#Zpi@{X?+H9gm;lD<}o?H<)w9^0r6p?EI`yMOaDRvH_T<-$C^
zv~T3qrS*aDQdvwWr|;h)Lj*sQ&}%zMKFo(N32X&TRP;>nIZ^-6RO(9{d0hc%U%#U6
zMpv~RRkaJ3n?muoHkTKCwK1DVKMS_@#<DofZ%%I1+G`!V++B%Pz`!+n@~tb3W*#ix
zh&-6lK0LsNTLhea`u>P2*~;Rvv5yP7!Y5e00@dd5d{|`TNIX#lEhXO<i4yK}*q0z}
z`PTnlR;}Ii!R{%35!2alHc3zn%=cYP_hrDy6LcD%ul)1dFQfgQB>jH$Q3+#X89GbG
zwQRd3e)yidn>?tBMkr%mXS}TS!Z7-2ita!ix06XAXQaF7bp^hz4aUmU!@jV?@~!9#
zSHE(tnZF0zusRv-YeHM8DFEJvmZv|+bC4=ExwZGVBH>c$2dYYoUkFiCBI-A&uPTic
zB$bUaud3z3d62Y(tL=2l38jN?=+fCh;=y1h@~~lTwpBEQ9!fDs>dWuu=H@ilZ6wA)
zF8^VrH8_9i7r={S27ftl=H?>TWTUc-W1p9=lB;DT_u%4MH(^%q(C~`Vl&n1G{%fQs
zd9sotl%?_s;5Vwj*%6Zv#9TO@Cbm{B$l{Lv8&NMW;G)ww#G|Ku%hjI)w<5~o{O<3W
zIseV(i}2mEGU<4Dz0RsDh~=m}rzn;PwqV9!361;Ma&1l7hX#Ia{Zeb&0+Q0{6LNk_
zT$bee!U&D0Ph}~dd)CMyk3W6HmITN>fA(o_>B-YW)b%Z~N`i>zeUek1Unz?bx2VfT
z<-yx`;X4+nSuE2?J5%oS_omqF>Rw+fOhtjwI>(sT2$z$|BX+mzGtc!ls-`~4vwd!x
zDiGAg&M4}hNop5u*oh+2;3gP*U##2juYSIMg)SS<nZM^c%`hqE6UfTc6UY5Ux@}Sa
zpW<<)7K$p<cSSMAvLYKIDq1PY)P1#CL)hDP{{{3MJv1ChyX9SO5-VI=NZ<pt+nbD0
z<nMn?CPFdBkxbK(+8b52f4dI4gflFMJogE<!QB^^1Q{rCiSxxMK?b5ObRxIFi}{mZ
z>JpGC1o=&=`RK~3Vv72oTZ#Wj7r&&7<v#=x9kEGFAG~+e`4my##;kO#>A<(jdps5K
zXK{XG_wz2{&^X%XaMY-+CSM8(Q8d4V!N&xdOYcAT6cd1iV#o?T!M;0|^F%k65L+?z
zw!~uWT|U9i0%0tI+G7y%a#H%$7kUcBxV$&B5IQn2Qt|3Q24Gxp&=V^a4&d}BHu^Q6
zwx!2)-Q6XW6sas7z1$x()xpq1^35pr4MX6AXT={FYQE!%aWo$k9jD1x&b1(KVP?-^
zJA=ias}s5<K2+xnJustMW~Yad*HMnDxhm(1As+)Tf!9bS>;zv2?)C@rwP@gpAVLsX
zVOEk^HIk}Gn@|Y)z)4uV2AG`ZU75i3`VvZWbU9-cBtAsDhVe(NUF#=RxOqw>GC-H0
zRm@Ouw|aLR5;}VAhk7+pr!PG8j&4*7DAQv78WtY%=q40m^hYSd>N0C_eZL2~X;(6y
zV}*FObW}@26Q?j)!L7rIk+i!v9;tk=7IDY1!_y0iQibE$(^8jN%j`Az2&*cS5@BBZ
zU8i!>{_k3PscW>5T9c@WH@ZuoO@>XsaiOa}+Nmbc^r`F>aW91R3@#@M{Wk}mZp=Tu
zpkWe{SRKe40<<CNMOIoi`55`s%*D=St+coA#Un%}b&kNFCck7{lsxYJHWR3A(-2oj
z*do0!EWPk4znRZpMzObO>iym{dH(J{stzIjnD1%wyYwP!tpQ(;m)`9r=>Lmj;jHbe
zB{&+3I%?y7Y8oCXujFjK4o?i}K8$=H-ipaI3hE`f;Tr>&n5s7Rt9kl#Mi`qO*)7Tv
z0?@nj*bxwe1h#6|L*yOWH0Q#Xo=!;d#47ujW~S_oA$Kme$v@r?V~RL0N6>Oi{RkCK
zCjtgip;^lVQ@A})ofrKLFqc~6L5TC6^dM)E#t<_gc&XWV;=3TMP>JFXxPZZ{ByNOy
z*|A)M))j~-j-&~5)o@z!6OLy*c?YN=wW?$mS@%|+dX#z@U!Ya(?TI-Wd8Z~BRx|^|
zGQz0ee$OoBP25AZC))+QD6M0ebHg;>J1a37deF3S=kgkE*@x~KS*4?isi4~m_}k7P
zA{v+dvfyh}H<Gp5UPWV0oz{Gt>QxI9`x!2NHvp%7VkE}m8$;!BOJJWgW!={!Jgped
zV0Pk_m?FG<MaSata!r-UxG<^8%AD^56TM8L?%H^A5#@{AwZt6IsQ60hD|tffK`rL2
zVngo6UaWP-borqFdE#!Yr~fIAs@#`cAS$O+Z}}vV1X6WiME7dICV8iBG69k?CCPV6
z$R`eAE?k*EQTON*aU)D>&ru7Tx=55)B<iat=PwZ%M^45=`zO7iPHh{B<#9=W0up8M
z++^A>9uCIaM7H|Bd!4Ga{DBm_mznsizQ6;7*Yf!_+DMFlwQ18MbI{gWxD?*LQ6u49
ztuUBIRzUh2w&X=Y75rbOZT=?zZ1)IosB2WeZTG!+hh<^B#LRxT%>zv>Tr2f?Wp8qW
zq^GuIs<Cxd*Z@DbBj!|?vMr^uS+<thAbV+L16vP#6A4jRok_cM;=0U85S;Izmit%j
zssmaUg4O(u$~Y+dPj7eMz>RZPB^xEmAyqWqQ%HlME+YMnMy=WDJ((*l9~A~}v6DZl
zIunGd?~IUNfMuQeK7~fZN;|rEX-U5!E${kX_QDRH1HA!)q2&58y}w}6vVd!dorIt{
zaStG<ux23$r6w{bc6D|PkgoPx^rmYBdzsIXn<aRW+VKeGQvsB~i+gO3LI=mO?ZzFT
zg1Kw7&pJOGm|ZfS62zDki;*%BmRf^L`V}&8*3##2WXeT%Hf+}!aTn{LuhgKkVx7wy
z_(#HdMR8<6tLEchwEG3}wamWMS&8b43(&LgVXv3lv)RPV`BkC%9iM*>8X@V-m>TKM
z4atp4OBW)8;)|=6w080*dqTg!$dsm5t$otWlysMPONb=p9yep3onI%AE2Dg~*hjTC
z4nOar={QQ#N{TK!S8}bhg>x?EjMd%l&*g?uOpfuESba_%lpKPxTxS0wo@7r6CI%)q
ztXaBR9d3O2^P4}rF||&??+4wbkckbBg&VDiR{VQp-;geQY;94kgvV{l!sL2~-o_q2
zCHbK&=$Onpu31Jc3KL-Bie0<pWmGVh?1v^oRa0?y4c$8!t7+@&TrDCtU(8w*@|BjX
z_C5*YX}@rD8@RShB$c(LnCLKgVuY5os<R)mNG|zC@+>V|FvQU<I@8v;w#W28lcUvm
z+^A-8<HRen$-51`p6L$5Snq#ZCfWV{VOwhNql-+f9YXMMi(T^Jio-!+W^$hVQBx4E
z_+oGU<$=f2hC{X*I@>5^5Yk7YXY-9aELqp}Hj-o$a#bRATe%8K7;+NqV53oOiVC#F
zd)$-1po~Uo9Jyu#;mA)dDzI-MYa`axs+x=AWjpFXYyE-6K9QNkOLq8jN}Zf&rHdn>
z`Zbv-ll|0`;pQO~Q%)|95vWQ+eu>95t3i#bjdS%xv)JZZLYtFt6zBN;3Bj;pFHk}A
zgA}4W1|7&vGd1GgVx0)!MfKy_=*?<ir=v^b4CtVCe{S*t#MghT=ybPJmORV6@SLdM
zjF1UWgOARRge}aO(gX=?VOBe}04n4_w7atqA%sBOsKnT>P9lx={!LThUVOG6U>6J?
zawHy}`$h|_J(qns6cqJAcZtM|=m$K#?qPpGqfqsGo^(&Qfy4v_g|y`SN(W4+^I=HN
zdydm}pPmz!Db}J~h`+kK5L{;=A!hp6rLNqSgLiO6!`ey1t)b-8{_5XTjr@6~%Y78~
z!Sufwsr~+k!c0jemoStzX?^07zK`LY$nHxuM*P8x>8$g)DVpDJUzsd4o5z!$L+8pI
zI*{-4Dd3!Yv8x-slk*rQ#3Rd}vo}tE=DC@!V@pj8-#vHh;dV^+I^4gPbjZu9-K9S;
zD=~dR?Y{Q1=j7(<_-(Y!izI!%oMn~+nDL3gtoBmhVgp6JcGtJ%RFf|-F?3D+DAvU~
zIZAwbHk5EZLsea}!}9ffBE8MuY7_ff*I$*IvZ3wYjbq@CVE-^GGRD1^veq`4SnkRC
z&)6tTxTwNn{?VY7>@L*P@VcS`^(dgSQP=HbX{%e`PB3jw`ik_7^1zLGJtcSDr;Ug#
z*8kGt@#lv{ot3kGhxU;uy8znUuxgMJAd3a-GXIRH#R7^Si2}r{e0YE>FW?AftEevN
z7S5~h0BA_bDw5k9ke7o_o0i$;AL-p$EuE3MPOdyIQ2l~vSXx%*{<#5bv}an4XC^g3
zQc5pv^n+S>e24hQGv-b1MG?OgxO7w}CA44bSlqee8knQ9zloJhrsNywY~QOCjGOau
z;Dr@a%0_8EXcUGD;6Oe+_t&2|jsY*6<3FWTurNG)k{)yfkhi3%0S^pt^y%LI0_iKt
zJL_YpHRmLD#2pr2C4Vhk`2*rN+~d~ZC_O+UV;6o7;DLqoO$3Zup~ID;(!Q#nN9i1G
z3E}B9Z;b8AAanc4S7Y$#7wBFp+o{UsQ2mD<O7iq4^On_0lymJLll&?&0@iPqF#UFl
z=Z7?NM3SE>->f&l|Dm(-oOE_MgC%1*B(K2E9pPU;ZE?s?oPjMV&sc55oBm#5bvkz&
z{9<*AmA9e-4~OYkCA0zy{%80K6Y@)868T9dwv6$+IipeAx|l-<gE|sTs2krGgBu5%
z%z7V8<T#}<UQYJ|&~db{?V)*sr;_9^5>q{fF=x*q0II!q?a{=m9O157LWg7G1>!f=
z{2@+@_}ksSbTV7#kQJktMiWlFcHfQdL~YF@R>5!Mv*~Z7QQ}j`S>Cyxt`9`)_xDcl
z+=(HS;9cz936QMI5yL!qCTWUDJG3&6pJuZd?o3tO94I{yry3OPt2NaV88>mj8}H{f
z;NyHxl6MQH&-w#oHd6%+#@vn;w|~K~U%N#>8~{6#YU^*H2j+=Ch}<X6BNXQ+L(t&B
z)PQnvM>PQ^hYtuYlhO<mCE9^rQC#%Y#UE1!h&A7iq1wJf{K!IhfVdcdrjX-hSRT1^
z&8?cuT}!9cd@2rOWuSH!O#@MjhOrDZejv?HLM)dd{0FV>k(s17nJ9Dw<;y_X`XZNT
zw+)3cqwYYoYKu`3MGB}unG7jA>}CobsRx9hH*KO=oI~g%%(Hx+HE`7e0_$(@uy|H^
zz=su;Zkc<S*h0cJ8ks>#7oR6hJ6JW2$6hL2=MF>FTDthy7mlrz-u@sWM%0te_MzH4
z*0=4Y)mUOjV}%4p0iZqjA@ob=RwFGyky_jLgYP&9O=QrnaWTZHl(UY#c!{%Tdm$jQ
z&XN6dhOmIC8zb6LE-b>x8U)%HVl(jnj`xV0#<4Gc6iRSgJ9JYh-}_yy$&HbJ{*qMN
z8-=DEFIUvcprAW_GGh6LkxDc%C{h2RtUm?I`Em}vyFF~F(n$yFl0q#bO(#RHkiQfn
zeD3v_kRo3y+ChNL-*sj_$a$~A?EB{MAc;ot>g{5scWN$1_F(BawooW#4#eeu9i}~R
z7+SP3gNE%KHG8EMwpDn2{^C}T|2@7Ke1U{scz0}yr;V<RJ*bxy^Gdo0BPeqbL`ueh
z!&5hw2#5CXc{Snd<InEcrKDL0&8TZ2)vD^aB~lw)$TAyJJymY~OI8&P%2IMTVoih`
zOey*e+|py2JBxDNl@(dZIx!4UP$Vx5>}qo(gM9kocC)D9%Q(8F9+68jL9`wE^7;WG
z;N!Jqut)~hvkp^oxU`deqS{bRDw%qCUb~>9boJHuWO7!q?3AJrbL&jav6TfZ=14rl
z-l2Ari)!d)gu(d~rE}6IJAKIdJvY6ns%V;Y<f$C)Ytt<?UE)VoOs9*?(ry9QwwOLH
z_ty^=Z^23-;iqrr%dV0Cj!G754u{5P<8ojW_A(&L$4yLcI|C|4<rR&-9aEZ-%#l(U
zR23}w(E!9@Tu{LnXZ~Qp1pTw9-X_ikuBI9FO4kzgY2ZD1B{--!Dj;N8gC!%#gV5q7
zN`uRRovAfMcZ-&5m7nwM+Y1h5l~Eak3t*3)^XScJ8Wg4BtqFV$IJ<f=8Z)eqt@;9w
zqot?!bUmr?j5^q4eYW}ZVCNrw9KFUyb1$|l?8TT!2ktQ1aTnB~%HrGu2*;NpKE2{K
z(i9dD0G_x~IK^B=vSLCDT-PJBN}!Y5PxeaSTL%dW*wRv|{TGa{1?i5zgm}NTTK%@_
zj898R%zZp1IZ-8qahqJ-NPm26xT@CH9!9mzfFl&hhdyL}Q<uF_j1_rLXKR=QDBqVb
zeBTs!8&6jCXC09yb#w0XQphqU-_)d9_@bL2$6YK<L(rECnhv~MO8C>OuzG_}H&U$n
z3xdvffHHXw!X9|FJAY|%ajww(ZqolD@rpX=e=(nJH?XajwU6r$cwagk!#ub4WZnJ<
zv9RqrvVOhG(8a6ivFrx(U068qse?=!BE7vM{ul1`v52Vbpk$wG{uere?WKXO`AVZs
z-Hp_ErG%TtS9p~FqnXQNr&y_pqY<K-<sjLJ!*a>?jEZ+Ec0*#uonj!R<cdEp&oVAO
z4+xh<UrZ`l2zIcMO?(-Nqu^S=ot%cQrrF5ArxREH3nfW3V&uQJ7xn+M0H$4R)3+4T
z?a_vn%#bDUda67jX!3L0=Rz;SKEq1Z_JF}0;uE?o43~7Gl$RKB8xa4=pZfiP?tyyc
zFFM2?6pqIqwL}u!ohP-{jgkKHEk;b~ALkrLq{%e)TH#2xm7J`_E(gWc?T6`600fy@
z4ZVgtmtmj>KT|d7BI1ONUzb?m2tK0!pv#Z|bwRdMM?K`~*fxHY+_~_{%O(aw#e=)P
z;PYJq3^6O_yEQ)~axihI-q?Po>soN>55yD?!#aA|^`2}&h&ztBxISLG1Hoer!IivJ
z8D%0;E}F1g#uQ^Rvacti%Y%^o$kMPAmW1PHqbiyMc)_qcja+~Tj-vU)%2NnQQ9QXV
z?Ck3ZJ)Dff7X<i3O;#oe7Li}L##hsCe1OZGF7`R_p99iJxh-YOcs&!S&_pke`Gw(F
z*dsV&rw|9%jomvRRbb`{j}iq0ONJv1h=j7)1d~pQ(jSsZm42ZFhWrjz3@*rQbuw}w
zEjL<@R!tlJ%AB?AHk9~XYi}4mEOn_6NyVS;AF@Egu**zJJR1o&@6AfO%e<O4S5z5g
zG$G`qbOsPe*Iy@kU?sFPlR#l-wa93k_Sm=n<O5uOfG{&#ZIpBJrBzR$+v{mZB3~C+
zK6AY<7a`%Kn-d|(Jwy76BME39y=lVxKRcMBttYErPv)|%F=zW#^N_A_?WncBFlTMv
ztg5v*%^$)!u(+erJeIveD8zF8+YImV=92`9)A1}OllSA277wgGIUh^D?fvJ;N5t0b
zM}s*@YS>Ro56W-Cn-AuFZr;~)t?9Et%B=rF)5yv7{B`E(>w~jxHn#uF7TetxTuCE9
zkgf;GPBBG;^}-8c4w%JA(NYOXV-cjuc>Qz9U`&_Tkw?;$O1Iel=2If%M<<fBn)sjw
zfve?U$Y^6^tq)v&c!p4|747C{8=31-vUdXs-Tc)f2|=IYQP3-pP1W#>OfLB0hY=mX
z2GEFDwn759=3}-13Y29kk<9|An+y2L7$AxTM8O(>CUBcQ#elNEshK>fr>lc%vqP@D
zlRPpq@pFGVPkvp})MygC?5+rs`T(3k4`|N?A?;4=eXi#2_2CA%W7j5?>p-E7y831K
z5(6fZF$ONTyqYxs8nJx#H85_EtY0FQZQ+7#ga*(+hLZ?-$36A>v(1W4>9zyrO%jyc
zKtg~DohKOaqWke+&o{n!fvS(mB$XcENZMKQ@)%AU%faISD5qsbx-5PFFT&UMoRf+V
zdE4zc2RHXfUIGYSDYcQEf!q9^4ifdfql6qLE>AJWmUxF7;p*4dRa&`66-_t9GkSgd
zS$(LKRJtgOl2%+aK}yCxZF%{Uqt``96#e<65?y`}`z`b&r$l-;S3IttNN|u-j$jG)
zPmtNKJU>{v^NA2|Ur4rE)oWt@T6g3pm1fmF&heuvnxy;mjvfuytLct9eG_Q0qQB~n
zj6){X*{-i=J0WP#%tZ1CH-t@yjKK`XCqeg|YY@MZ6SwjljW=q&g(~zjc3nE`-e;<f
zw{T3$5l;Qw75~_ZkzCb<%{0tR-8R(-+x)i}IYAGttR1F`Z4OJiC&#_9zg7IBBhAKt
zR8cun3fl>}+NkWzHc{WK3O@|`TG&?2dl1Pj;x_nNNFr@L%;O=O!d{&@CfWfPlv5*E
zD>Z>;kt`v7cduyGsSNThar<KIB*S+0fbOvp8gpY+3BRW(bGGwK2xq~h3Hh_InrJ%1
zN0>m4#lui`AWp(gj5=Sqz=NC?R~7%ni;ZUAk?G@cPXG=KKBefM(HkzXiLg3z6*6mC
zfA%&3%y|OVE~W<e$PnKqw8QXxX+kD=K?c!jm0k5Aoqnh1Y+7gzKoKCnS4S!z(;B&*
z%`>{i?aqz0wl4kSEt0-&0H-&uYq;)pt-yws_)m`&6Jg9D1IOb?3XM>vT|tJF{WiNO
zl>iQufBpRv+u=5Vd7sD~bRCQb$#V;mR2Wgs!D=mH_o{STjM;@(d%D_PKOx-NUrn&=
zHFG|6>nwZHbeCV_HAzU(RQ@mG(;m8rS0dG~HHDBkVzI|>UQk{T&5K1yzzis|Kw02n
z*Kr`ve)i7DvfhR88=>og>YqA5KA|Sgkl+iMg3qciedy-sxyBF&I7<^s{OMs|UeejV
z;(Gh7eC5GNV%hODi;Y@&|LNZl$cLz>$!6j1KYWn_PZ<+DW}KYkN%qv7D}z!PNg`z>
zI&_?8On7%Q!8^(}-RXj7x-)R}`Q~gAA062zAgpmwavUXL$F%NB!r$V=VpkzQx$_%h
zfPyUl5nD~QUM=+Y&Z#n|@iG<a+mCyw)Vx5Z?_{;(fO_H%*E<rv4^o=m@Xi>GCY^Y?
zS8Ha9TNog@3|0U8f0tAyDun#bv0c*AOTT-iZ^L75IW3&T$)+Cv3=jn|SQUDm4=Zy_
z7#|OauGc=Tuk9+MT70JO;h>rO@56=R>rxc_CUB`gr{4b41sv7Ar?R`-QbafYOy5lG
zRB#N}8H~5(s=d3l`?;a+Z<O!h2%&0Xgc1Y{@o0FRj_?xe7n03v=CP8kPQwYHjefz4
zd4uRQu=UB<%V2r^(vMA_KVCXTjOA)T8J3;?EUI1%rJ%YeROebBt{af1)-jEJ4?1{V
z9o0g^CYI9ux%Gtl!&~-r5B?~tJA1(j9xHk6zW{o-z??5zFWrA4K5~5?>3vezORx}K
zQWWzUkTTGXebkfyBC`u(3YYU}_Wl^7OHBLgq4aLOyonBn%iobq)g~{;RRVN}iPD2F
z{dN??P)DGsht*7a(iv(7f%4$!C?YBjFV-LK`K@{ZvK9Jt-cH9&t46|Gz(Zg_m+T;k
zO^P)n-3}@=HQdCr^*XDopA(z_M3W=+H`B$4-BkFvHXkwTd)|u-irhB7z1NPEV0}3M
zsu8n4e~^@430acddNgO4mTW8_aanhl=F9xe{TLe3wlSE^VXE74#m`#^7xOQS722DU
z@z{^5Hnlvv#v2xdf*8_tW)vGGj-ZeU@{@?dei~O?wqU##h5J$Tsnj$)itJ@?Y52Fv
z(e%zlA_wmy<3eqYe<Q67-+AqKP#Paqwf&xg-+Psf=>%7>-}~Cj^Eh;`fp(Jo9G~9|
zj{oq6&3`43+GEV)W*0hrCaA`0HQyq(TI7<jShE;?oHS)&SBt%eIcf0K^84F6DU$bh
z6-9uUPjS;M``!#k%8J+Ur)Mt@a7Vo?jhK=GKR?2pWXXMET4lx<!`MN77l`6En|Hf~
zT*RU<H215RJ4&ODDM5RrT5Dxc`u003;5)?!k4s3)hO4XUG(k+&Ue~8JVYBo&;sbUn
z_v-6_C6weeXrYHD2}ovy!HtY<bQ*@kr>grX1VExx+g5?CRC9EC_kg<wSoR;nl~o3s
z%G3w7@zgs@&LW;P(yg2__pAfzu8jpip!CSwNQhf^jw~2oPzXZO_E=5T6xf(2snidU
zc@RZalWBdIFTcYE4EA0`c5WTr_OAL9lmf7>6AU7zej^W<pSU_zUc1kmpH=VOUsC<<
zu;x*rLOk~kv)eNS%k_#mZCFMU@KY|pTw|i89mg9pqX!-yO;8@hpaoN>C@Ew+C_~;V
z`vTu->8v|zE3Gj|REEfbO4Itu9UHESV(*`rkir$5@xaZO4c~jMb8}4#gdAVCbuRE}
zT&Q@U(a_kGd*QO5TP@J8p}i0S3y7iZ`+HAm+*&L=VtMk_cWvHp$p=2+%ViEBv|WVk
z*27GzLP(#yVbSwkZx2Uty`680i=nU6XmhF@CFXo00z&dGn`-8bP*0KxxyR@4UE9tb
zc>*e5CznnNmf0=#OS<doYwRPGlemrQ9aJD5-v7ld2&4Y(_o4Epy>N}Z*N3Mw1us_?
zQ<;Na?_~C`sH7KY5j<oCvG&yt#<N{DL{h?5u1+cyIooKOn}qOG1iii!N)g2u+}_@K
z*=l{!Pv{-E4e%)hziKWfk<<W~nwp+((?+VkGg=_nIZI(6*{mmAA_+289&QC|hR7BW
z*C!|oG-ashsi2fq@p-_SIXW?=8^Hn}03<j%d`QiRq@0O?&%l%o<hv(d+trOQ&ViPR
z4#8mJoGHp5jcVi3xc!oPzgMHcO%Qjx>niT-G6*(nCges?mnU7Qc*-iGx3cDAXq|%*
zK~W%22ht;?At<lo*JQmHPvSPo?KI8(KOR%z1y;rBO%pcUnm2Hxunq?!Akz|-RUbPC
z`l@QoQsu?C=lWywb%5c;lNMGA?<v{UdC4uzVLBpgg=#xIOCHo8V@01JZwQ}Y0hV{B
z)0JX_q66MPDMJKhv6WRd2c<P@0Yj;$fYx(tbt8YX@jWFMgUcH<ze(VLQ2DOm=t4WC
zwo>L0A!>ScERDgK-_NkS4}1GEgsz?C|IzdnY*BX6+RqFv-Q7rofHYFlN=kQ2cZ0wn
z5=uAH9g;&ggLDWAQbS62cYgDpbG~0N*EM_X{oHHau}W<W-7ZsKX(u9r2opZ~#~wSF
z?dl)&47qrGe*S>A-9zDm9R0s(G&6nEhBFV{QS9BRy!16rh_D&ib!%HXR0jB+b`W_6
zYmFRfl%s3prSbTwb+PKYx>vw|@xEuAK|fkMv#9<<G{GVn>GqR4Nb|RQ{F0o{z7<tD
zLR_UUw`AWB7a|1-R?TWjm?GuAl4L0zEFGSxV;w#^J!c(_3xkACfEgXts6l8r7s5Xg
z87iNVB#!pDVpOYLJ_iUIb<TH5f^B5`f4*HELTEn^Hj&I^^q-EbLcr)Uzl-%JIFa9Q
zx>c3zQ-*L{gY#k=bg|6{d#)d(@i-xl-IT@uwhquSX<NB^M<T_;whBCM@uL_HNlNah
z?Ivh_!F5!_ZIKWTVHiGVK<tXG8r|W`hNcq@Q9I=Nyg%XV5#Vrip*m#-S_p2)<EK^g
zfTNuZ|9tt%g1NW|DbdyO5YA-hS5~`7rp2HuvtpUMPR_4yKOU}oylM`(Yi?^7JlU3>
zCD8JF*&capXUYtNW?pX(ul@Q864m1&x3nZv^n3h=1o7ek#-y01$sf7g#L~dFfYO)m
zQi4c7w{VAdKLEP+UK>O{r*kQd+vz!WGA)-mTY>F!3qncX1x@YaXy=Wb<;w~u^*Vzd
z)ShcJ?n@<~&NKFwSDt$%!&!Po?az}An7lo{wX5XEg{?2m40-G#vPMgKZv6}o*0fD1
zQ!m*jGXTl<vZO96=ez!3lZA_y3$o2dNX9Cwqpke!4+$d4rlRCtb7^fK?Qv9aUm9~=
z`noG4PT#Nrar!tLFCxNHp<M3#fjZg_tda`<sm9OCyiqQ;{^YC;l!nyZ?%fm}suuiD
z3c}aT)L#R(B>rs<-=A2qp*!%D3j`Cx85Hz*8F+R$Yf8B~c#flM`r0oBKz{IL9HqP}
zRth1KqWH}HD~3`*it6VJD|t{<npifi#VLb&@AVx$wysS}tQekD9cr?oqVAsa*kyX9
zWo+p5#PC9~ISJ$~3%f@s3f9h7gqL$@9YzS5s;Z2jn(OIagf4z}6DN%PobqdSC2P2G
zsj<}!X|Cr5>OS2SmA{@i^Jt-;I~NZc$u&51?43Mh?8DI?*QjtY;gRpwcpWHz-$Xj2
z2cWi@v3WA?(@7C?)OaApB1`$nahZIorFsGTCY6mAA`aFC^p5(bI8gDvw6`cUV)~7>
zDQp74V~2)qF1V-X^@K@=C5mtDuNu`k)D<qJy0`Xl_FuvTS{vH&c?6Ok0Bt~28rR&k
zh>|nKDqk);G)g!`jlYTW3yw|Pu;D5BLKdsjmuz7yf#Ci9PTnB-gD*c>=!Iqgry$R(
znH)??!u?2?)#+3MgDK_rE>!v1_5XO_)q|>&O}A*we$#i~H-Z9n5&37Mh}2Iwq7EnA
zpRW-HzNEo6FC$&d)+&@T{n~?<RrW@7oi*Y~_TMC}DG|mJELp9~<2TbyY)J~Gc8lFX
z&mL!p9I2+dxl+oI<+B$bkU!}X%>BXgTfd9dDHekM$C@~C1vUMHL*wr2|4YU16tG45
zdGzy#bi#~&aA%ilmr3|5#7^s7s2Ocw(Yr-Nrh)dn#MChX7v#mv4hLdngBXQRr0EVf
z4=TWuHf7xXb-L8X-a&qM>I~v;QaDqZM}|#RFXJ1}99xH--q(yLs)QWZ1+;K-EvO<?
zRY^iDHp3I$BWnt<i9xmFq;<<fUPj12gMfa73J7@>P0E_ybyYPF69n6on&pr}$e&)u
zz-S<}YamV_05Hcv;Zz?7A?(5XRG3yE7=2|At@#Upxg~HNUF8b*=i6;{=i$7t`xder
zed&E$Xey@E5LY<h>}&-EvwD`jan`n~MnGDVi`p{;*CO;$uN4mvqhX?a0Lh^r*m+b1
zMgoB}wKN~d!Lax+b*+GHDoWwYfAyUH$)tH0*2p{x{&?wYusAl1eulCqpf|b(Ad@OC
zY#v0Vp;o+ds8@ecIku*dDtJAscMAMaYKB|-J+JZrDy4|fr~qWp-I#S_7^ct$Dy(Ok
z=GF!J?Ojis!mc#=SliI`6$sHrpxO?vp-cS5cKdUzT)n(JB#?(B>C8sPRQYsypD((-
zKN4bs&O&+gUA*t^{e1nQS0jAC{P4CEci0FaQaOwW?5h)~xj*>V+V;Cc8b@H32|o7c
z-}Qf-U*YN4+IIJ2#Pbu>p}vS$b#4ZpsVZL@g6OvGTxpDbY&q5c_v!gYB5`gjKvE_T
zQZdfO-V^wCV3n$uJ0G#!O7re!XdH9C!_V(*ZNzrqHu%rlR6nld%jd1*C+66kakLo#
zHgUKVW1^~N=;pkjw#QSkkhb?5&aXBdWwu26{XDZ`L+aIm-1<t1du-z;M7WRkfYT@<
zs;>)AP(CU*|3bf5Xp**3&({E|_h(zL=_~iQ4ThYWqPUHZWooW@%2z5KY(&!)(X{Ke
zf%!GO7w5RaN?o^^E4ob%%CEM_b8}7Y*L_+U&cc4SjJ=n5)$TKb5Hd`DMI!Km<T?%C
zz;DjnC@`)5nKn0;;#!=ZKRqJw?<k&x)w&GY6~bX9l8`do$f;1>xDus5h+b>{j}{Xj
z_)ZLb1wJ#vFwW3nu*7Yh?JqtDrG@8WAuk~mEJneJQG6hlWESAew`KB45`qh%(cQ{X
zz6sj_)gDw}NsWErUY<jXyxD{r<zvLSV&)kRH@(Ffb|jGiYh3kjODlj=1x`>;cwutZ
zO1YLQjPvsmgXbZGfTw!+T{SYpzqi-`ih9KA5nR~>_Pv1}5Fht2oxp`LOwK)TUr`wU
zP$714UjAUfffOdIMIx~E14q}D05)DMc6IKj)oyl5=v%L#4wM`TI9uM<$L$MnF#5Pn
zg6^Nq*sOrqH5<6X5L~sRd?jYj;iA}MbTHvvMVH<+GN70$h2z?z=T3DX$#lG&^8@>Q
z#YF{{6D+Z~%yf+G-5Gvzo&xwSm3qlM98?*W*mc#}7-|}5KAk#19tk!TUeC`Ob=w*{
zP&N)3{BPoC5`v~~=<7TGj@tA;-WVGdXcFoF5`P>9H)Dm<gG~AM^gb+<3<^7M0=KLv
z-+X&jyFB{pLWuF{%-6NpF@zpAks@d%E+(-6zIIQ|bo})A+<ff}t@B(6)UE0=)u|a>
zzHbj+5%E4KoWtRR761~1!r^m$5HM?kn$@sstjODr(@5WqkWwayuM9H1FC9Z{!sj)-
zMBzwQ9iqrh%6C-eqbw5JV)yNbowLQqK2P2H;<MG-?QPD+)5e`Tzp6Uh+U>`A2dT6L
z9L8^RpLu`2q&;#Z2-l{0T+TNz`5UyOeue%+C?<w=z3cr{tA7BxbK-+{eN^wvz|;Aa
z@84}1|EB`DzvDK!<X8i24qjLP+AiQ>eR|OSur~+0TO|?BKk;Dhzpd{=Q&(?sxRJk*
zZ|#mN=z7V#T3QkIu<7cG_P~h}&yUP`(YS-Fs7@@?2j+wztqRoI8VSZ3g^>fj;;KOe
zWZ<#f)^rgssJEg_t3Oh_u$T{Nol{oMzv<uQjg%I{oLMD6>Q||+;pr?)Fbtq1Y@9qi
zR*FE1h)_uJ>`%oDo`Q+44KmnUFK|#<1IPgJ3wp9+GywTx_OEG)UeAQ_7LUvqpTJt+
zowsR%gByu4Z~ltxx~x3{TV)ayN`Xv3arGK9=;P9hECvlVxx*?m$>A(aDfiK@heda~
z7E1ptezkmL?1qs8+?c8$9F69$kBq4Jyn&pn^)nCPk_mS{KII(35XqrxEe9Vzb}rtp
z4fyJB``PU=U>U72cm;@bx;BmEG5;&J6^vophMTHw2#0AMuG%IfifvPJ<#F>H7N=49
z;K76WrpV{ZY~<5v)%Bv~^SD4w?p4-6h+dcU)isjj#xTkaJ)Oz>>_f|iP9m#|Gra~B
z91&P>S++;RG_7iEo&%Y;nw#^>Z$5a`uN8G@l1`lXnyDpcDu}HnES=ty_`7Tcc&{6h
zFvJ<i6VR=dQ5xWF#tkp)b1SAis$XEI56CBDd^>%eSdpaK?{!UJ?BN<~`Q1!@o@?)u
z>ppeq4)>p?-Oh>iSA&Gx<rYqk?|JAA4YI<p<F)H78=S67#dC`!#7t5MEw&FJ?R}4;
zrEag*((tdu4(4kG*(OClFVzy4_~-?UdJI~Fj-<&Xg&w;+%JY_td??p#*h>`Euo5%p
z$6CY<U*wfxbbKF*AwTNPuvlKjapD$KsV|=sv#7g3#u`Hb{#F0t^B)QwZF~XwJZS`z
z*>7-?%0Myu2BQ(G%Kn!^`8?PnYsr`ITopQsNY;b!5R*wRg?}ad+v;i@*J{1=giHNZ
z1&IfxcZdl_{t`kfMFfz+NP5V(d^ImKy(8I>Uk|joA<6#Sq<H4S1DAn^dRxO5dlP^P
zgxIrwXd%S}uxPf&l==#)^vsdf_ssnFO=TLXPfy4rSxzCAGVgs;{&L&KgfprSZ0sH7
zuqyIeJP@Kd-3FueBLQxBrQ*|%QM1nLTwm_Xa&i+uVnH5dG8m|dm^>h+UH=oqS=v=6
z?)JZ4fo`Ni?ZmdF4!`ZRvo$-@aE5Pqzoi9V*@sl@{%v(^quZ7VRJiER+4sxM>@;X;
zN#Q*T#ovoOt*_wt$!XXsw7LzQoS8DIbRvvwI9p4--`I>fU!FQ|nMXRereBXodWe9F
zpqW&n`Sa{hIJcLk4^{VHobUx5$YyV!Sn>ajEplC1(og?0^F6GxU?_d0DJXbLdX_mB
zQHUWSA>I27om%S+{afbvFLIBCa=Y9mZdR!SZZv|q`SmD9UrLii=gM>{#O;EEuKGDz
zN$r;QpA&azFN~Pka+zI7yh;uuCU*iVt=Hf~?PSW|8+e0ar&WVG3lB)VmS5F8Jl=5D
zI&~OWz28`RMf!(Xa7!b>xbvmgX@e7e+`br|*v4ZsH<C;5gmc8Wn9UN*Lr$b+UwlM-
z$mmoa8Z=J4KNz^;h0MT7h57L54=EXAp4nn<`i;;z)QA)_8p;g^R2=KvJP$+yjHbXg
zMQ2R=%!pYwRyow|R`4@M;g|OUNj6B)03CU=OI9-7Vp#&9V0PRx3|_HdA0b{K@Npc8
z5);@U_i^L#O!iNshJJ2<uze{G_(OcJ;Jn_TQAh8JB;*Yd2DrnQ7-sP4Fq{`eEm&2E
zC<kjp@ftvNW-Pyu!CrkA-*?Wnx|J+J%uGhT^IAz2Zy?#l>58=wv&&WP*c%DJlABQx
zv+@UFr#U95PM_3s2+F+}tvnCFK8jGiNKa4}1UW8t=jWOu3m6W_8Ocdu3o`yYiun@I
zT4HgnjIu-aBF(dc9p^W}iL#*nz64KR{DJh3ow>JO|D2E4I&{I31MG3zaXvV0HEy{}
z(Y3EQ(5lXvssdI-S*Bk(J#|^Y1RA7a-8iW_CM26zs1GgCv?wZ3bFD=Z{JnvC`su7j
zs24$pU(qK#Wo??VZ|a;sKZ^!egVq;vK<o5qH&5dtWioBkpU?G|j@pLdqPrU?yWdqm
zoPB_A7nZI*c|b-<K2a-hjo3g5=W3wO6z)L_;>3tHZ4+hqEN5abU@@Iaf4Ui01i8Md
zJmLZetK*94<4C#p(Cu+H-4abf;Ai;vYv`MtW=^m0kUI5ec;d=MV7F<ZBjMi(>-j60
zAqv}4OJhuYlhuGs8acaQ5`W+Oe5XV4lJ%N~`jg?0Hs(71rBHlp&UKrT#>67d+>RVS
zOx6CB3Ev+D{giqzP7=Y!oxX8Lc@e7Z?(8uYV&6u;f_`QvPr4Wv7ZoYp0f{Cx=%t55
zj4_*T$)IJ3)B3cGAuND)J3lH=@N>=-1H3Qxebw^ZM!W|!&w-DZQ^@C3a;M`tJy_yM
zGI*@}@bcSKc%B>THC=S>Q{JnscMC24`<@$TDo`uVlk?#y!(JNw8i8keiCBcIFTM?i
zwAjEUct8Wtoq*DrRptclHmGHOVrPM@HiD|y5z+OvNF~^}los*v&d5I#07-z=EE6aL
zDp%LAQRBpI^o5sRw`wQhEnDt<Rs|$p@5&m{11x(|HhkJB6pj5q$?@V8Zetvp6XryJ
z)Ud36EIZzCvOmXDCY_TGpJ$|J?uv`;87aqzr<YJAW{%#oFXIBV2va1JsMd-S1}qeM
z7=A?kL$<`g+cbhrU1GNWWn7Mf+rSGthKd83Fvqlnr0|da{l@=|mWS9q4>6-$;0m|w
zy$QLk|9so~M(cofzs;%mI#`=`21_qYT9FUu4)Et{)NjZqQ+r)ytS#n5y7<9Z+Zaje
z6=D-bMZ&m&WqpZ-XtrHU1ymmiyRCWg9nF?$a6r<P5^fLsvX87@K%|DwY};^tVsN0~
zJKR?-#3|R%Y#SNH#sG>Pl^yk_A><pVn8{BHUMk!T!-ju<8AEb~NaK4qX!cttHfb*o
zO-{WB4$eV1dq#S69}(PwnAH!j&*bnWw)^%;H?i$UJFSO|&&+lae?<{5XC^;bm2;D(
z&$!aP@k$lvv(@GwL6>a5p^0}T!9bc;^=!_fP|+Kh$E&$ytwFx$ATB_$!|X$Guz(6c
znjwqya3Bhd-_fvskW8B<WB6LpT+kqiDc<p@p5xPw(tTdZ_-fwHu}?O_CN25nL(~5K
z%U6a>36zcZk}<~RlCc%N!hVSj#bW_vn2wY+H1e6Vja!eO!o?}&8or|Z{CN@N)Oluz
z&sFPmM9e?FN@G>+>*aBH(YKmDyh(|rJ5cEn!c=Oo{`nNC*OE;g>^-g}C8a`#Qb{94
zVsSGko0Sb+tV^IR3cJKb;HaL2AM6T25M)wT{76qY&uUD{>w|U}*a8%puJg}+tV({9
z2?@`J^GZeLfJjfiiPMa5jNw%}IPz^*@xWRQh-fyDk~av@5>}M}O4`nT_93#o&p%2K
zX)0az@H5n4egO`4@{&K}%PP}&3am!M2W+%6Ib7pD8vo2j=PZzbhj56^US@4z2(Qsu
zW4$XrmJfcEk0^n#FYZsHG>ussT7dqAW2>W`eB;|kp-bB+ljyzqH#dN|A_2c=YD28V
zcCC}uO*CC(+`^F`vTS3%ok8C9QGMo;BMg}aE`i$0sFpi}qgo=uANn5iDt_06i(+!`
zq$fTkT#UXaJ*M(p5VJ6RCe$s><2B9lx^+5RU(8$<^|97)o?54;+Rm6&y<Bv#)WavZ
zUT+RspN(PfzBBChF8#pG8{{k(br>n}{Ud{K={(Y1qFnq=!6Ey*0Kd<oF)}zO&0gbn
z3%aSJ1K?#+mE=AIfb&$F53?Q@d9h>bb7FA2n0d4F##`4#BPveG(ZMO{>h%u1C`CI}
z%o($mGRI0E^akPHw)39ok%EftPBroE%7>RdNFV=bXgf3*=fm($NkYEC_~DSLl30r%
z4H6cISy}z#W2Fa2f>?*xujBd~6O12<JfC%{mm83DMvdaUJ7}$&%g(QB-rkP7gp>ur
zbDznA7f#tv>4~;FcvMcftv2(u=IYIzKJ<BgVXYOyH4GbmxLKdb!j;7Grp21-B86fI
zDPLJi42tU|Vb+E$`ea9%x2JNP#99cijypDif=saZbBW@C-%|ugD0FDiKT(XRdaFTE
zpa@%bRJ@}fgJ6F|+ZrV}_vQ6ci-h5~OG{XK$sHJ(O{L_$ihhMf+&6lpLb<%&cXU{J
za)y;Oi^W{5HU5BF<$nE<^=$2@Ey{50$&5WqW(uD;=f-~|fkLNV_8Ek)8Zer<guDPY
zRAG%L1H(#^6c_Z$3=E*aWEA=BM6ndz+Rf?<XTC{4LQ(|?CdQ(ipV@drU;6hUYXcca
zdVrfu0U}W%0TPeV>HU619H8LT`z!OSbdh`be9^f;HkwtzgK~G(KQr0{GqP#qcL}9G
zn#cqb;X1vK85XGIf_P;FhT}klRr^b2T5-w&<AnhfDu{aHOPx3)b&rmjeLIgjBY4ln
z9~L(VtIOVR0nhmll(RmQxgJ&CzeHv~9<$ac$e*I!lpc7{a8E6WJq1IWo%fib<H3A1
z-^y<kjXP6dYI^~tzfpXoQ+g@9hd@DdeT>T6XGU(Ec%InJOPZ;tgd2O2DB*$qH*B$&
zXVXkO@C`#tw$4ZrGqkxB=9FcrMF<M@FIffm7+2=!-S!fvyuNC8+7GwJA-R@~a`U)y
zrmg&f1!IVMHhU(v%NK*cbJ>clGyHoi^fFa}CKIJ8MwhmY`e#lI;>$N4(tE)T7hv3)
zXa!FWx<QTjmvF0auk(Zka4D$`vSUF)cl(G20tP^6EvBMfSjLlBERcE>@w?0{cF<cu
zo=|Au7hI>r(JRmm{XZNWo-bDGmkkqeRH~m`2HGg4U>__qhP?h-7r>vXhGW+5lPU`)
zQda}=!Nul>n7>4NR%Eiy0zXStjcS$nw!y|fYV|Lu{g}Fhh)HduD}*rmt(SJ>&7#NG
z1#!edt*PK@S0y3pwpGRVZmwoX<am9G?v~g+W4s4cDZ9ghGSC>n*8saUaG~!?UoqAE
z5j~POnhuYzSY8*1+tB?>1EdS`ezqlTTAGVlXS!h2F49ziP#y9_XQ{N8)+Q#KZNlKZ
zoOJyU^js*>iF=2dnz(MkX3v$LSrhiUQY%_3=3q8*y7#M;ko3_{1OY=3W6`fPKic!J
zjEANdEK!1@^ItsJ(609Lb}}X!{fOZCs5ZIW;-@cZ<Pa*QDdo1_cgV4p`s=l6UPZIy
zeT=Cd%Xi;?eIqx^2Y%W^99UV@&2PJSw<G+Tm_)1-myO&@qacI&chcGC6<<+ZxqL{K
zbS{(K&+4Uq3t+XE9+*ie7SicU-+GIa^pDJI2WgCyHrh-|6tNz1U7|O^&Piv-5pCe7
z+kh_5v>ue8blDM`7BWd`w&6fL`TUhZDM`iW0%{J1lX_fK3wThO?d&VHz{$BtzIH&;
zuW67&Oi6N$i*&nv)>-<}cVaMsR7g#S6V`HhC}tYZbzLQsc$X7=H|h7hRI2oCaTA-e
zX4zU4mX>(xv9dFBE^jh%@d&TGy`29A^Skdb1WU4w*=>Id)mRZZyuj*h<Ql+rW7~Fr
zoW{RjuO|&8>_t8G;<_QMK!X(1Z;iaA1GxY0-8s*Rzi$nF?Pml7e?c$7qDue?Fd@M#
zxFNlK`exPf%wQFO+TJ}IwGa(eS;d+j1vI|ujUA6XfH@}pp>#IcK?v%rx+IX)IwuPu
zhVX9C)5J=bCtqtF<o!FG6KUTO7OFu--4P}{d5w1;@x|n?A&8UUH^>r|D7;193uFp0
zeB=NaUpAVuz9`LZTU!PA{?Me0qRPT|EYCs{{Rn?)KTyM!vfqcO)MiBz^5YkV3!c%E
zsG-cd)|H#zDGCl^uSedcxORSA!(L*QNCf$Bq{d+9w0LReq)3&dknIR+6PMyHsXWkb
zn(hZTzRYuBj#q}Pe2_9j`~zttf1cXleT@wN_Hfbyl60F<UE5oB`oTda(Szf!`{Vp=
zhpXLSPoUCui9qGCUQ2(4Hw^*D>WBX_YA%C>i)q;uUmyo{J;k`(`7bRsa*<26FVsd7
zoMqV3xKC<&kIQvTCGHY=V)q`giKXNIc&VtP-gTz;1~d68+NN|c!(qHgDrv|Grbf_C
zweZ$FvB^56ITEI(<Aj-{c_m{$<2F%@z~Zr>o3GA+cT^hQQKGqBTXh>UP^mfJ%XlZw
zklg#3<}22WispNR)b6@ENH#}Z2@1zj*YR8>KmLwKwcbpw$u+i~I_c1FB7l#u<!xQQ
zMdZrVYRh6HC|(PhlM`8o6NLsq2c={6k%0T?_382h{&<JaT7p=xJv!)bd-t(x3m5_M
zzEr#2H_lta3MNKf`lOB!5hf&gwm9<#GP)N+HV*4F*f22jSUH&p((um6H07=N6f6(a
z5#r}@+OWTfJ_oYm<)jDEdO&YLd1eQzI<ZXzxaelct>K4-cU-U0j9uB9lPeGlZ{EQ0
zS8=L06l)`e<v%Mm#*2f35K`kzAt^pM`o03g7|SX#K<KU*K3}LN#Xk9g*I)6k9faFB
zs}V+JrH#D$i#TmDD?+qbYkF%<0D=c3JRwnJvwrG-N2j8~y_gElJ|dH{XTiY3&co34
zjoE{@sw<Z<w3YLZqz1b;ONktfF7<^~<OK&5lSzt~cG~D}nTO9&o@oA+c1(90@rbC0
z(I2M;dX5V=js4^NlqG=hpJBI!n_nkWkD=-R&&e9c^*b$8UsAO_bRGEZc&A>nG)qwp
zsJWNVr}#!fU%p#qX}@RuT4cF%JasXh%odWn|GNcZcaX8M%zjSOjq-D#VsNGHkn2Q}
zl8~#ZNid38uIa1k@gW!I#D-Hy%kfOhZO0wF)q?B<sZcupGYFr;VnZ3le-3#FsSN2n
zTnj%5r51GUkd=v925_!jTR^=<0qVQ1#TK557e>$(LvRL(Xzaf58qzPVrs_L00%$<y
zoOG7Ui@!UVk#0$tR;y<*eKoCn<+F?He{pL+FEr%<*<pI3wXX-<DSiGpW#~zH+X&MU
zV#M3A<3-6X(i{C&vM!auq8mmU`YeJh^<}>)%u0VC!5rml_TR*VRv6_PCTRS^j2VLm
zeDNb{-jeWuN$<^jqXvCBtE7^V30p>D4vlbO1rCrqAhnTAh^8p$d<p-iQsi&NL5PO2
z00wqNtAIQpWMkQ>cuu1FrJbdUKC-{k^E;JFh7=KeSg6CBZ$D)-MyO3xyto7}$w*!2
z6h#{ih_{uo@(FIBx@V_>%P)|H#o9l8!Vd1^5<I8xzrnwDuhFSh-pG4*jq?!wHzp@V
z8YQ&4KU6Z@Ay-o)eBJ07fjzHzA<rF>?#nu6y@T<tmP7IGxlcu*)#h{J)@bL_vyA`U
zV&k~5)?&D}cf!c=XpV-H>g(O8#9Q@=DWC4=gSBcu{dwmix$IaQ7A)Ai_vb}Cx-LHN
z-b}s<%Ip{D9bCEXyFDqqULG<5OS+_6y+Si#M<B*7;t;8ME~BF4lA01_W9N75ch`EP
zA_a)a&y)JKl+NQ_{zb6jWjF#)d=Ptd!tfV72#-R53W?X!{`U)4WWkdBLRX8Sjozc=
z?B1cdrydMj?^IR+C<jQ`wSVRV6-Am?ogfT}uJ4^0dL@-nE6AedfOhRv3uPu!fRS#n
z)}_x=gm~13KiQBsgs9eubzzam90)hV0NDc;%5DM7bOlpRWH4|l&h5U%D49{{F*s3?
zdyG*qW{wpxKjNTVl^df9Ap)p)tjrFSJ*%g^_4xpC?bh#<9TcKo{R^NCa`XoJnQ=WQ
z6e1DgvdL}=b{e%Q5nW0Ji-){HD*0&m$Gy~Qr)=7aKy9N2B{UemXF59Fo^AF}+YdwM
zadJF4Z;)fp^b!F{QrG%BY#7>VKSUV(!@vNs?j|-X)q+9ATCA7*Cm@O@6Zyp$_+FG8
z6lsmD&V(2WX*@7oXg#ICXfYpZgUlCub{Cb1w<dp<v&~F=SZ^Wh9|iOCDLE0IeKZ>X
z?q4LyeX-{NvU#~a*!YL>(Xfk=W;SdPA<N`(O?5Y<wOI-<of)RSG3Oht4$UxO(B4V9
zGhrH;>cR@FyquFh6ns4YmgI`Sne!qp?eIa5H#-#JOPW~=TL$civ2$vq|IFIo@7@~1
zu?PHqpVOW0>q>Q3^3v(Od3!Qz9)f%^tuSP*0%PiZ<&zAAk8MQX2P`!C`PeO3%H6+_
zm2CKhI%E91={!O-2QiRlBbSWJJWXnts5?i04)9NV5-dAbW1py*kjU+=J!SW=rE8r)
z-1NgMgZUNx;P6Wk4Z&f9G~qMO+JCbFrq`;!18EeF`Qp}@4smP0XLBGnhGKIad?ynE
z2cfV1Dv1^4{sW^x9*GCE6jee1-B)O;x~r&4?=8YLdI?qPV7O9oTu6#YQ4z@XgUkTK
z+?)5qTVm1lqf`aKt68HYb#-b+I(R_h205qfPgHkWpP+esF~zaMU%`5B9<G_sl5Ah>
z`~M9ysrBZR`6?{`>3GM3TWV9jPKoD}>-|SU=RHRXOuzv>%YaQ18tghhqrpjUr3Ol7
zs+n=7_-)UPhtauZ@kAz9M3ndY+X~aJp2Lg2d7cPY>e&xVnjpQ}V6z?0VzfbE`=cmM
z0Q$A(c?jlRW&a&?xHc&AurRn$-mvXo-S|?T<vMdCGII=y9_UO(u#$EU)|uT}ISHMr
z7A~#2DqO;{JiJYpd_0aEbi{<`lst<IX1e`a==5+3ZC^>FsiSU1l5BF5qpH<k8kx8t
zK{uFFd3ESDEVKOTekY4oCgS?e{^;58|1jio4O@~B-o5j&z(P-M=|krdSTMa%Na1(m
zX%6n)cwRu>l5QLuo>e?G4V?G60Ht~vPC={kQM0rZ*4ob8{b0N)-go8)NDi8n@;|C;
zwO4zs;B>dRA*RcRWt9qU5(t_~6D%1=c_Z>IQ!GDmVc;MTs-<7{W;DIZX0T%B{Xo+l
zp<v<3eD58VLlOfoMl(p#4eM9SmA|M5cV_iMjj%P11&vW5_2}I4IPra97kMGULisew
zxlG<?Rvy*B-W$8G`crA|g#cVgLKuK9;npuW<<pz<$AtihK+%s55bTHZaxX~fCEr<0
z$LFAD(W~_9(5q~`h_PDxyF>UJnMAprL)VJ^9W@~oqSzwjfZ+(DW6)^)ddyfGCL8$p
z^!LKLZ-QP>-LQIIo)H?-$3yY09|;C-w_$F=<_&DKUs(NU0pLjC=?PJxIfBe~z4yke
zAha-oAdevCK-UOV8H^e1^5Rb(t!YSlpI^%T)TTV72Z4KH4KS3;)g#cP><i)$AEe!Q
z_(lpAW(G#s2TP8`Qc)%(G3EXtd-R>}%UIdpBtVB)QjCU`{AkjqgoA@r$PmV+gvI4w
z0~T+bJ~!-(>LT;tlSb-E?w9MCz!z~m(8JTr8O#G32~joHt#-3GNYtI;XyRx_#d?iy
z7;gx#oJT)A7E!v?RkkNz!oii2k=5lA=VKa7&9f{x4~No#ByI3R`T4(yW16|-xca*e
z?<*R2NO#Tnt=ls1y`9(p70%Pte#rSz|E~DUP^+cU_&;$xxmo1e8tA+x0Rz<Oe9})h
zxH)39fw!RnPLYOBCi!Oy`CRcY@!Tq}(SV<FjGuT}E1InBj^D26No;mJ#vg0ldRT@a
zde-RC9O5a{sFzAfSyM#ziOE{QXA1ujS{b7Ctmh9?Y2)|D62IPRu~{l$&YkXkz>H9V
zqSL&85g=HpGI_z_H-CJ-kbi!qKMMkcRcfW9<?{EpO@(8kJX#-*wu(*ijGpJ{_k8l=
zv2bZ9Pp0BIrWXFeM2@eNO(*0d!hq0XqaZJ2tWPC=Rw^)D*gL1*-{hm4de>Tc5D2;y
z&SE<kGTy3aRf!Fi8M@%a+dN!4JU7S>%4SE8R@sVCYznEds>Mq2K<^)SjY{v#<e&Ym
zs*%m!-!Gzs$72-98Z+{$1I9*DA9pV{y0C>Yyw!~%ed=^LU&$mNInHMKAo}TQNHiF~
z_e?yo4bjpZ0cIYLkG)^bhxHI9ATdA&B1V;o%{*{O1)yX2I8vFeo~f)7L9m=>7eC!$
zt2csN%+`R^lgwj?y2Cum{8R`)GF-lU|91P<XpeB#rTS=6#yVzHRKly8%X|8*89p@Z
z;BMa^Nwl2Thb{p+|JO^&21Y9uG!M9ru*NxCVDD0_ON{s3{o%C$lmC5(MGK>ylAtz4
zVpW=LB$MmYY7>@R;yAIyGJRr;M)|qy6sA$9#q=8U#6%qbMDzGEYi8IX-0~xF5m7S6
z31P(K*ioc+A+ZC)(EG1pT`hmh{@7OwgN9ks*_?Dp&h=KKdWQV=`S0}v>`-s?xw6u-
z>$kzAHx<EkYuW<;LwK20Rpo7Ewjg6Y9+lGRUwafir%a1ynuIRB3pWmRN}Phf9=Et@
z5*=N?Z<fkPf5{Uh>W@4KmSBg=%{I30U)2Pe#mM=aOuY$^j-mL~2+LDC5!6kSwx;<B
zZFu}GV;it2SOnED1vxHIlkiO;ULh|I_b%3Ux$KDoC%d~J9fI!Ge{u>&%&OtDYqloQ
zxeOsx_bsUA69Y}WpG^Gw{RV^1&IbIBcl|6@(Xzh>bPy%2Esb6rY>}=*&y^7?!KNhA
zCruVg4t*vs-0TjmoPLV$pad;mFO`FWO2996&`ea2{n7P8dL{{YPWR+YD^p?|gh1px
zZh#Mr@WLFRI7@^$3J*?{0}KSG_KJOrA^7|Y5R&&H;(#cR4Zr}1xr(8#BL9{KSC$XQ
zWRxp+Q-UqZqS}^_)9KCRVc2v^n$Q5(Le!fnyh5~8JU;}my1)=adjuuh60T~?0i$o}
z&@K&9alS9uP`64^w!$L6<VAoYlXFqq++vZxp~HR(EbEO?6F~iyjcA5}zH*F{&s9k1
zvS4&D&s$nD)o^HG8`tYH`k7zJ=YZs7RLi?Qf}gR>sd4Eah6*?#<bl2_t8+$Xsa`Ac
zBAz^SlUp9eei#@hXw5G2T^5i0<Eq(y!DTDpd;3h3WtFkq8NXTbZiKh7vdL#4PNV5T
zlMM4q=faCZt8&GUB&*rb8Pt%O`Bu&5NRN+`p(yj$S||F~w#+a!J(MdWg4MxPHpwWL
z5C)K>zWzS1*Z4h^u#H#X+7Je!FG`XhcQU-eN+ec@;8UwbnS|FP--I51@BbrniL^W{
zdALe3D^PC5(0u4OR?t}-Jxel_ahl|V6*-%6e8uG?nOvf?D1G??)ndR-fr3(FU`aL~
zg|gjJP}Mr~&C7L`5a>uHS~Cz8&g4@rDCYIBuKnE3*Pik1C0pgh=|T~DMmHIT?{32}
z^nqry@tb9^G`hiszi_m`QIA3UCRTW_*g#y&X+56r@9LQ>!`A0Xo*4+A7f6yxikwUw
zB1m%6(@9Z1yVxSmGOoTVuL+BG!)B>9XPR|T2_Y%kwcYdFn2G+YB{E4iYXjzy|Mtq#
zxE7?PrHu=jx*Y9t&yIwKR`8RW@p3Yy*eMPuZl+mG0(46uG?VoV@n61FBIf>F(3k<R
zKc@`Z?*V`t9HB_uxWHq^L`e6|>N0(_iXytO4MqtgGiu}#;@(?~o|HSowrwz{&6iwN
zQ)z;T2&7hs!E#z1Ax4C83Q|N^bc>FOoQCMP$(h4Tlw*_-YCUo75Y<Z+D*DA=As_KW
z(VE5`yyF*Wkak$d<3Jh6Ta_qBc(V~?(pQKJTV*~{GdKTg)`g|zKG{f5GSANE?6Pyo
zN!Z5N;-Td<Bzkf)SHU`@)@Om2k^$#-J$<1$>gdt%SA>5lZrmy1AS(g@_5Ts8dE=(u
z+Oa2r+8WVeN@Krm)Ucluj*r)jEj`O~kh8Yf6t+dD?{>+0-f0&ynd1WA(2F~`qd9*S
ze?dH`G{xZ*Ha#wbGR9?awr|M)IJuJ~aFQgCW*FX?cCb6<k3T(f9kex#D_*4ZyJ}9%
z#wN|AUM^^`n)8brb`n^Uv%mv)V;47o);X-;sH*HdL7w~d8(++hw$?vy&9{%p;e&yO
z$b|^{$K3KpW>+;h%qNkj6Svqf!#_%Nz}zob9ik_-80zS_TF*u5V&sGQfVW=CyrpDX
z-HGrluk9+jNs;LZVuldC55X%62hc1Y0oVyf656ce1IsHnPuER(g48@;2}<?1-CtW}
z7-(U?i~waqooF&KuMj~%gvVBgy`_A7BLW3^J)&q&u##!1y>Dp#FQ#if(RyjQo-pJc
zkAG1JnbiCKHm70g#or>@mHGLI)2~Q~$LRf<E|ZKFACkY$=jy3^1@S^c@7ozq{&V~t
zLS;28#-%yyNf<X|^_$kH4YW8n@dF;2BQTA*K4^WWl~umD+=@k?hRv@V0Mr2XW40gM
zv3y@bu*}e<2CbEbYrm~%YF5%9@fdS#;kpGNd$+<;)1?!Ugd09^awDPBYdRVK5^~bG
zy)sl&IbbLha$d3W)HJc)t6MiZwEwHNu~+3Yo6E99m$KQC=r$uUFn1*?F{AeUX)C{j
zvB`L%pi8W2w>@sX1#);XZW+L{5TP0udrD}NzxVp~AW%Uf;9U6)T~R&u9W4>E_{3d(
zV)iWe1ETw;_22FoZM?>j__jEVX53pF@XbRS^A`_2`uL2>YI^|by+nAO^lZt{;j(uw
zi@3>#Ac>riQBW3Zt-VG_ZowSed+~3`4_5Qltonm9xN?mwap!i9vg8&t@mClRpBQ}D
zkHtksd$GP+|EJ!0XnN7P{wg@9;VALJeeDTn>Sc8_--<UT4|BG?<+J|O62Lpt4mF-{
z!H1hTZok%;6>vdSn5)MZ3}0#e$A&VupU6L;ma$jj;1ua6dz#ZBh{!kwLEv75*F!<i
z@Yc1%+E->XF{hRxv}BT8Tn|VO@qjgjS6Vh>$4|Q#(cjlpRs2f8{G8`D+)^bh*H(d0
zy~sx-em=ZfFzh!GC2C1Dh*r=&w`;R&yTl?9AT#bJi}cc*3ec@2(|gU`RFEzRcHOx9
zMU|v%^aHf@mlz3M$o*5q;oC8KGRYx`lW9UbA0&DFd%7c7CZiPGhHL_$;#Z(G+eLky
zsg=q}j^=1p$p&PLMoN}y1I&;^Ose_{LaIVd1RI&^3vYxV7R|5FP;GlQsmf=ZMZFH_
z=Ph}Cv4pKXXOwm$I4xxw(Y`}O`DyTM=*53b_O!$#Rf){vxi%2Rqj*F%q)MQ>485ct
zjO^FM518oN`g8jNi@uTYBtp=lFlAAr_o+vsHdqH<siRGXX#sZYV8+Ar@%E~gTk-lc
zo7~=fY3qr5BQhf(Y1wk7#)<hxaFu2J&Y!baJWrV?=tt?&+pxlU%9GdX(YZQfWRkc$
z8EGm?b|HxB31z0iZIrpc5z7_bJ{EWL2h$}r@!CBlH~*ZmFu(l2dV2lmr$a%)J<F_g
zT@$-@PK!-QUQg9*j`Eh^d04<E4Lq(kz$@2@tF04cun(7Ro{?=~_o09*v?Td`Hkv!M
z2<;FQJ@`A69@fl7aajtFDO2*nJ2@o&#oaUE(!eGUvaeTdH%e+~Lr5zc_-Oq9BZ!m@
z25_$$+BLe#VzZk!C^pG7la~s{X)QD3G`3|f#M}-yXln>?Kbi)ut?#8Pbk*AQx+J~X
zQ)8csJ-h-vFilsD<#wX9pqW2l5w-V#B?T-v?Z)gwawoc41S$zgo+~UeqfPdF3hC*J
zU#2;@Be9(1vs^pKw7Q2<PvM<NQ!FcI9f=&jA&my@xa#w5HS9T2sOS4YL7^z#u*wiW
zO$~7f#hEa3Eiy#GY*-2*_@h$k^8~E|QhsN$yf#G@n0Jv%Omt|qsJf2KPq7=7%AaWE
zGY_2qe-=OpIpnP<E+G5)G4!Y<7%9@;XCE9s@Iy08D8yihNf2-IFLLh?hz{(TkFz((
zT%Ql;qC+`B_S?3=`v;DD@?}VKSAJpskFzh`4kMXf72o}<DL=h;904cY$fU#?CKrlt
zKHAYVP&)%&G9q;bHzV)-C2B`^MQ@7xx?lLo2O|&~2jn!8iL<zCIIP|7_^z*w0m8)4
zR-JF9)^+bt-+}Kck%yAlTl|wUB}d<^kqHu@j3nDtJYmLU$4yUX9&X#tTB=5?wO%cP
zlr-nsxymc~f={_%A1mkcj<woL!@lWu2(3me;ZvhHsJ0>1-lR4C<-x_Ia)&&pi`sar
zWxu#kIwu%iZ?k#%dWRx*0O_!cfOv(}2SOdoLXqYC9|%fG$%)g5oz1ylE_Ng(rY&3|
zuO^VREdTyq!zX6#X;qex&zrisF%wsO3LCn_LDeAUmqx}gi1pyuzMaGqboOnxc~oG2
zrn+o>If#}0*it-qW;*kpyYqOD>$uD3$F=V{Q?R^jRsj9p@;6c1(pV<HrdAA0czS~D
zk;K{0vUJ=0v%1}ZwxPQ5|7z{B*&{E7M3@rA!jH*IgQ&Sv(DXmzdX2Vsp)TS9K`O`?
z1P6P0;0hvNT!_QstPv{JkNiZ0TT!|%cDSU#CD_ixTVNWqVE|0dBS#nMHTzLtvL$?-
zihIiSPq>57ILH@Wg$Em1(UyHiMH`EB(gG@oMTG=srIk{Y+1|HkL)Ua;2J1;ty9Rt*
zzsmY{8R{THAT{R6Zua(j+McD-PzHuFgXA_SH{+YyYZNmuS;CuGe9efKw~P>qn8^ij
zgs=^n0Km!q{vf9nZr+0jW3%S%26ecq38gMpZ>+E3Tzg}JR9HlHwGCt5G-9T?X%6?G
z7!&JyC58k+J`DAAQbBOe*h~pqtUUj^SqA?)HvVmEe}F_8DH5)MdZ<0#mh{?@R{b?r
znHuFR`qc!1u9rOk{IO>U*Q(g$+{FLVnj%u|fNS_700!`##o*7jA(qw*CH;niCHZ+P
zh#Kf$am@RT&RdTWA^LyAW6QJ&Ba=T9vCdVAB3g$R7yo#J3gXSiYf1=9lR)c(lRx4f
znS&1Q>YFX|d&ZhDRDJ``Q~O=#M{=(;r=Vm#?<r}238rM%H};(dp&Iw*1&lPsdw3is
ziejU~sF_9)5&cPqc?~x1nbsO>_M%?lx!vwLpH3>(L-AY|(((Aezp?+&gdKUg6?aD?
zAo!L2SW<G5!OBL#M@)ZzW0v>GW&a1>stx@$aaBw}b<1X)xaoh4b`QT`+aKA|1vh<j
zI}+P-)W`zE&VbnedbAbeP+x+xqo^}fjI~C-XQMgj4-^SW!^IiLdxQIh`(Y#7>ek;J
zTqNDs7mLs+i>FijWK6*t<>Tv)8fbGhXIF$cWl*M-hN%u;2TU#b*P+1pkE4PSzo*^z
z+mAlE6qi49LJIb1i%75OEHwpR_g?7RDhTCjhj459#q8Kw{v&li)h49lB|iBqR(Gpd
z4M%2xse;DnB~<P^7_UbD)<T3)(_iBSY-uUywVlNx`zo97v4U;@COS;lLS}Dl3{tl(
z&DF0ef061Vg(Y8uc(ixFMDf91NM(;ROZ<KtO*SdT8;}8Z0%EpC;y%HM<|G9+Ys#sH
zcd+jj%y5rZe(w^Y5<;T)KEM1sDYAy0FjqUh_cuEMjy^QG=Y>o&OaquFBC$>qPYp7o
z3%`Ht5imp&vb*DlhFHD!OlNn-?Z(bjxcY$sWFR@Rf#5JEJ}kDz5bq2<xydj=0JVqb
z=QO1mnc&-0Hja6o&stv-zET?j!zoUDyJ}JU#rq&eMbc5O3W=2%MkYFC=am_fJi_}@
zhAZ4b_BL8qj9HQZKiz=|%Uz+$U$k%*%k}dwyszVaq5APA<pu|Bc3fYD?dsMJS8Y|=
zxL4T{uCpGCR*UCcmzsHKzTy~KCG-FdT==Q|H9ZN;#|EdU&H{HWHwP7_zbpwg6TLtF
zy!kps-syY!&h~mA_E9MrYvf)X6<^wkBhHTW8>;TKXef?Q%RdRHpM93)AxnaiKNq~!
zjx+}=_t+<Kt~hgd^M==5`CT(&gvcd5XWkz!m)l*djy9Lwx1ydHD~Kf$4BRM9r3ydV
z>dXeNxOX^Vwh7+Pu#|Q2&vGiNE)T}aP4IiK>Q~x-7E6eV_>nW{4B%zTJ12|~9o``n
z9tlUW7Q2ZhHAfEa?K1&Qg;2wcQ%b_}^@UlV=QgK$%Jr%Qk_LE`-5PKIBgNWt+aKiH
z#n;lYrgdZFvtYlAU2h!zS=B2i$OogxShkfWn+QR@%%WZ0v!)Qzo8k8Aj!iU>Zf-{h
zFU)G$i_}=+dX-3LP?^{KB8L&xv3-Gebzs`|-<trqM>P>{DZQBmr>--E2g0l=l{AHY
z!dUyoR}2u+4iVNCUk7-x1G|awJiO<pT|m4RCGo&JK&UNIj-eVCBURb-7JIC~GQ5?6
z2Mj4hjrXkZt3pg$8OX%Kxnw2ggbltl@O~cn5z+RQLknLWS=|y~+Kr;AP53xT4Ew>c
z^xJ`%XaB7eHCR-Mj(f@MhHUlA9dcsUNY$dUA1QlOfB-`91Ox0NK~sPkyidA->uqRO
zg5$>|x6zVyao(W2No=XI*c|)Ur@@>_kq<oa(Uc;)ipumsy)V`XuT(o3mx_?`NDf!8
zKWs0G_IZL8CkrY;fk0eeRtATu$}P}i<#;WYmjYn~=oY)VB4(Z}D1Qk`QN!W}B$;XO
z=n{gOi|(soagFb9{jx~V1DdIAqD&m9FUEZXy{}yYovL@K=ly|w6GXH~EW+8dKiqD}
z-1<csk}EtNJ1Fx<jrisOz`S+4z#K^$G+UloH%4z99GRFUQRIU~_z&|aGA+!4y>G7K
z#kQ%(+_mHoWSyzo031{)Lz2m1U~4j3T1$JA|3|tmc7zN?v#G(Rhg&grk0~(XjB|!x
z$#U&G%6e&J^(R})1d_AF7XO9*U()aePRvwo^(AXbBCC}9!T<Cczpw*1fB8PEi(MZz
ze`+@UkB{#*Jha}8FawF9QP}Y|C`O_BEeQUn>Q|bo*{Hgf0SBEmP4kX4V`eUfMysH~
z6<d(iNl&J<u_B5g!065G#<@c&w;2)yB|hZKD38T-12QWfW-0xN8NjF@g<8(-YF(DS
z=F}?QyOfC%gte2m0<g5aVFVGK=C%kk&GEE+LAs&u=>sW}V{n32d7^2v8Fg?sUGqP>
zl5F;O8g%~S)7$cx-K2DLH&X9MNiR#AljX6`LBHAE7lWzA9fQsHsQZx;gCMN6!84PX
z<kbBP8Du5n;eLf}wHVH`KmoVclTNM^G~P^kTnjZX(rWIF^QG!_0Y_?rAW{?;a+yGC
zfXL`s2)Sng;woXKGsX{FqqeT%O5ys06#p=%zvE`zfRFUu<8L7=YT@g7lcFg)72~Ym
zsoVj;zk%2bioQuBxlDPGw<<C!H}R{fEYkDAwlY*e^pBrZyXIQzNMq1<no<-Rq;(rp
z-75pwJqqkTgb?aM`R4f}@tumb>!pt5A+h~<uGk4uIIhgwSlLlBNtb29%>*uzxmjSv
z>*C|p=%O8g;AzXKR*+jF_ukQ{739Z9o+Ha$$C%T?usK4I<Z{@@1wrqf>c+xBzMwVk
zxA3!jdk;NvxjD9fjfSZcy7=C&SH?==8i6+;#X0f#+15(CC9EycjJPTRzjKMW_C3?0
zZK!RgL?Xi8YZ8%nw(QVL*H$fkA$7EUU{QMA!A0vimJ@aV9Z^Tk<Cddo2<DD{3q+V-
zgxL^1_|bs>bdYZJ?E8GgJ(RA!HP)*ky(n0EyYMK6^T;olTFy84&a1AwzGF*gi#HH1
z5^w#`L8X>eY9Rc^bL9RO<G*2i;uXhzEjbML{L3$cONk$!JiV;`pcw!xW4?~#c5Eh5
zKD@_bD^JzuGYZG)zO|Ps(X@a3kH{pW^=nEvwy8siTKs!E$uMTETBJkSkaoayNRgm)
z3*(NSF8WyVJPb~<Gp=Np2DeL81B<03Pzh&$B9xB6&&9My6hDta*P+%Sl9pfdfaYxV
zYK$*@hJ9ANxstwB0qGM#W`9dql9;fbFvq6;$R)WlJwR`9<GImf%bYs-_S}l}fWA78
zaX@XhHHd9h!I8uh`L_Xv5tMuGH?a;OAD7tRXXX|&2Rc(MTEvUOV4evzg66YuQd^3F
zM7PycvngADR(>@PRV7RpX2-?T#&?wug4^neva5BjcVV{Vv)um1566}IZG>3=*`?a&
z{P~~ELC)3Y?mi!8YsHiRFAR5#-dKha$?!#Gy_q_(P*7lN=ah^Ke30}pV%KsumHmw2
z!=J!OXWi`AyuAqJ`8$UKHTOtEg@DoIG4OKGOVwvQUD3>RU)s5#jV<}=<?&bRB$7Wq
zRi3{dXgqJ~NVztt@7RBlbZlZ75VkUQYi9Z2OM3b6YfGLEa{cKIcL;5_8x8EZ!zhf}
z9YV{0nwLbsxm14G{(52au6WwZu$2zNaCmf+=g@+EW6m)8C~x$cjy7Mqu7GB#!Czrm
zB_JnFQ+;7_XhECgs-c?b@%r;~vu<~uhn!vue3SgF>D9r#r5>^+?3ZjyVS>93Z_rQ2
zj4dOQq|a|Gri<;OxDeAdB;R$@mS(TwTP**h(AfTj$9L_obBDs{*i^b{N@F9ESz=;b
zFK1NA*CTGhn9TZ}6*yj8|A`TTO#LO9n)~nT=DYTw9(iAvBl>bK?i&z%=#&52#j)L;
zjUn^@_?ie#dhbwkz(CLpxFcXQHt3m)@HQa#{AEC{9Q)>|RI4UbNN?1_vUjwCWBa0m
z#mDT7DgT-}I`^N_bSn|r$!<lLiIsm6Xu~mGf}2r}lWrKL17IXuoB4l@on=(iVb|_w
zNRbvPX+cmL=^PLgP)d~U?(QB$rCYkCYv`^)x+Mn~y1QYB0p9U>-gVabcFug9wdOxx
z_WIxV-ut?KmjVF$XhoA2i|CJ@HkF{0aW!BeMjcIAjx><0F+7NvkNobtr^u&_8pANB
zN#?&;0Bm)_>|O<WY+EvabZ@>_a3=lwkJ3P<UxM_(68PX;XSP7P{*>@4bWOhJK2Lh_
z2WD_Q`N=*F%n;V}HfqGD^V?4`1-s`6Mli+KR^?9HQ1wRpKx1*63#$5T*RAc|%n;ho
zeJ>Ci`LapVsPIBP(P>@1TtoN7oDoouuaI2~VAvD%F0w~IG2BsnM+lL=Q-H>-JqI2E
zsLMABM!OGVBOCQ=;IU5YSS}#~Ct1!Od~XJV#xW@Rwm-TgkpO>mU#vFAv!8si3CVvi
z>N>*3Id%lj@*Ad#eAy1#`SJ6P)<k@!>ZivM!4t<TfC*AlnaJu+eVOCMz6F`fsV3=R
z%mCQP`1_2KGA^;kjq+o0mh;2b@{NxYd+*rRy`jFQyd9uKkv=!HZyFB44VHI%ZTwIh
zC=x!KDCR1%0RY*q#mee*nbKN25*2*W->&wx!qDs5e<&}SITUnfIG;$30QfhX7dCmU
z^$u-fuy#a?Z}_Dg8xuODMX$cE|D^LKyu7fPz~ASH6=vX{<dJBg{@MT7#o4z=VDf`8
zGNP{Lq!kX=!aqK*tKr7+47qorTxSUvT{jwh;vM~=?iZ%YtoK3Sry5Bgyz?M(ZKO{@
zRd@aKe?n~dhPj{E{sWKAee0R0AYWZ${V^t61;HJ1l-jby?TaoZ5Gj+NhihZiSE;=f
z;ab`@<VSjCOYN-RIPx%alz1vF5w+PS@0cdh=efX1eI)Gx7BPm$Xy(!pT8#Q6jhiSA
z9rX3+$h!2v{g;dPXewupwM!4bGM!JyP=E?Y<>-Oc7RJG@7#k5i9|D0>Z3ji_O?)Jk
z<HT&ieB@A|4ff2~J9;7ox_iD;FMzcim>w*|fB#c0$cES98>OjkGENadYB%ik9f9#G
zK)bdF<6D)qZcZXNwpfx9rOUD&jBepygD1b^Ho~w3xK3dhcdl54aiGUP>BR#KnRdyP
zgpp!oPj<VdB@5s;Ri%Xrf9IqC%n+gMpUjk%=74N#v=~$`4j0w!cLUZ;L8@H8H%aRh
z)6wG+p+cz%q9Z4R;QS;Sa@pfw-Rd!j6ftI#1&Wlw$s|qoL3U<w^BDUUoJ7TpSJQI!
zR!_jG@($1NUZPt9Pl}WPEVjmyf}G6qeewm6w45)oKG#3CYDm3&HAgKQAjMl;B@5i(
zotDtMpwfo=g;hOoFdgF#AYdV}ad_Qsy3kM1Ha@A(33ToXb6SUNuGfy+6YLm7XRw3?
z@WcE_NgICp2TSZNhM!0bDnn<lGOsGcEmrHg+K(x7%clw|XImF^E<DQP@d1ZC)W^U<
zL5%FX_{U6z7X2-Uj*iT{j=<0NzI3i#ZkwuhT5o#CLItocrpT$^-tU8MD?|d;WFPsr
zJFY7p(kxekaZhi^wYjHp!vpbJ&UPEyf>ODqouwsNX!I9ZE48?$WlmyxS@#JVz>9~^
zGudwDP_xA@X4(OrYMszye}v0+5aH62qMYPzbjNw<`Ks(OsXxQ{e|D?CI-RRglF#7v
zz{x4VNC!pFP?(-BYGkGVTab0l%x@3ff2sdhC-ClUSn_Mmdknoap{Lk$EISm{6bllM
zM;4<sh^#QblPv(FlAqJLFY?r4G?=d4q}pMU<$%U&KQz8?%P%DBz>>N1n$o&UPS4&>
zB;h(JKRb|q%suhc_1oTw0j}`qlLJ}gO>0M!gUEc{>q4TXE}t9!2&i)UK9N~#?Y+vc
z1;4EAB`eDuCd-$b(>)-bi%j^Hcx0p@vbL7{Dkd<??y9=)Kc>UMxlIxQ_WiG|xNYZk
zOmFYw51Gv3Hx{p!WIoI`F}dZ9-2D?CZzK8aPNz26Pjcz$V_}5ZfzX3vaq)oMb~{|r
zH5|*@eNNN%&#yPY0i7I7`-^Wn2#jD;ceXLnm*{r{ppDiO2v0lL6{S&TrLCKSf$yt^
zug9R4lg$k+UEJrXqP!fxVRe822nA0|)ZOo;9~bBxRWJVZc0TMIp&S?_Fv3a<z%!55
zJUehGhe-vEAy(QhawV#cs`~f5ByB0l@;dV#K(2FCApi22A-QT3w$tU+-X>e2m%llf
z4xIgi0tD-wh8LgjPD25d032>_Tko^7ZQa!^MM=DUOD5g8Y0OF}+P71sVHyC+0Ms>0
zTqIR3iM0%T7eX_Yh+7{D4o?LT8&BHGy!bilS+v}2PS(wE{LT>P)U^ODb1!Z7T|ZUG
zJ(^}vOEnfxAV#|;8KM2CQvUH_f+T51U*rpVRmEGhb_#MEfHayWz`wvRDdWqyC)r9E
z#rqG5Wbc4Z*Y;}!#RL<fndpFw=Cij^O+3Rd2XVyW#pM8UxJ@SLmd+`X-f^x3!eCHu
zB1)fNsy7JS7{+8Z`c*gki6hVltk{jHIyxS%8}~{UnQdToG#7rZk}J_S3ppAE{UvV0
zKet$0dfZVPFSyE|A(-N03^Lgg!P()#dpV%D=VcInIyuAde|8W$L#MwSzM&i+_UAz%
z*XQ%ltUsIb{nfyA(|io)nXU30(FL?tST&YE(xjLh|3LJ-5Azn;C`lJbcYR~C{Ghiz
z=C{i>I*@hK1UkN|rwWw{nc4cqtJ<aQZS)reo0{{&{?Ync-lPp_F4!<i360*yv^BH6
ztF5-Z$92nm5cxJZPwx7+sVsW@K`CNUQwx@sFJ_}*YG0GlkwMb-SNp*JiQFICI^Rog
zVEBH+-yt_h29aGtimf}GdsDJkp^wUYcXZ@(ngAcGZ(ebH@>658*QqMLh{{qlM4J>T
zGQYl>Wm%YXX^tjyNA4X<$fv&(>Teg#m1mNWbi+K&2hNGyAt<pq{;Z<!zW!IeJ<Z*!
z5+h7eT%kL5z-iBzd&FUsxhcB08dNVAhcqi|(_M^p4R~gm1#+ON$LqTuJ(ts|f2D!t
zj+-bx{f8ZAx%~jma?tEsdD2+E?RNdS3?gzCfY)z2lqyk0<3|Vse~;uB;0^lhCE~7>
zdRyxAv?VHke?uD@?SAe0qT)Vp5{vglVqC8fX@~D@=Tk|(D?^fza9duFRXz11%4lGZ
z?kWN!p<xWTT^5n;`(U#1it}fi#{Kr<@{t~`@6p5ON53`kp{4???WJwF&%?;E{V<lh
z<E;OxI_EQM@y`Zz=zr?xd<+=S;p2od;w1GQ)tlvFY(Wn_erL7&5%Q7S`uH|1o8`9c
z!LJZEPR9;9k7;95es{ro1`cfyF~rU33`v1<y31!f#9(FpC2h@em+SBJL%QqH`+*wU
zV{`})C))vk%DzPVD+)X5!|80ddgHTv3l)SwyqvCba-K!%)Q(kk`2Z(LO9HcoM`&k1
zv)Ts9Lf#02u7(1{o=<8@=>tc|RI{EUWcu-=>5xK*Q$@~$%-h}SWW)I6U+WWp6p!k*
z+$4!~Mot#HetrI9he<^%&zQhNVtuLB|F+s~VXzJ&B+^!J#yXnJJbDpiA5c^kA62Kk
znWoFCSZ8x46MtEf33?X(v~j7Z$bZ3$Tlke4j!G#GBuyXw4M#lQ9NF{*3v{I5$X3l`
zHt)@h0~z8}$<(+VzB0G%YDtmkT@t*KSpFI}B>R^)u2vaeb$F>p`f#eJtuU=}^4{aC
z%3P6&dOc{letpSzX?w|MVSd3U;<(@SaW~bcS4ZlGBDb;zrgz|WOn0YHJ^Sb0JWHBs
zY)T_lq=4k`>M#{S?B(fyY~X`<SmQDgdr;gwrzO8He9>W*&~*XquU;RvJ7kK6vqHKH
zKx$dwrGA2W%$-E9^NRAJfS&^Kzn9qR7p7W|NrNRfTg@tgZkk$GH!Ff(Y`h%xgZ4~k
z#W=mazI3J0HvFWPI)2f9EI;PD4H}4FoCaISCjH5S!7fMM>5I;J;IE%&SNZ)t_tY3z
ztdY(i$YMOA@)<F(@iI_Xx$oT6|8(LXKDgh8wbXoYH^?@0m*HHS?2{=GXPTvTXGLIt
zvp#K;hoD%0X<EJRm-qe)nL5|9h@gw@hEEpgxNWs9QKzI?^q~CwD0@3x10s%WKKAp|
zL*5`e?B<G;E7}Ur2;OC_EB0?bB<tGHtSaC(ddp%h<}SUxyFO}bzlZBw#;a}bdcmsh
zKO_^>ru@n?z>QDZY%02A3~z>DSn67fB?{z^pZ1VzC?P`8?DoaC+T6uNhevj!4aSc3
zm$!E;&kbxiY?>sT_H67ZrY_Axgh-^@@)3)kd6$*AO@5OU;hN$Vcd~x*LcQ*Q&r34P
zolXRBfvR@yc)VHR?v*MV!idSg@i$zdV#C3OXvXu#=fl%D3>D?4tYE5R+a+<-pGC{p
zv%AGHqrE%enU$SVq`Y7VAOB=F02-fdEa2Js@GDB*<5FJk1EEAHXi61M`APL>YXpd5
zKfbRz!F-AhSQbBZ8A=Ootqr{ujuLZ~d~652g)*<9dG+;_5%X=_+6m|k<jU%$^_t>t
z=IBqT!V;_8EO?_bak7o^@af-|JZ0CG*C0<<)x%p-n?)u{Obs;B%#c8K*a2S?XoiJ!
zyztL)Vxsew0fu1XA^g{M>K*M(rlau>clPaP$GO$e)|`ghNga#DzuKx{{3c7}5kC&+
zKh54{x@=KCO(FM;bhs0U^6INz5_4(~-hHknap1Q@L#HXMpRs6S|Jdd9sF}fYvVR(W
zixF9&3Cdeuzqh&56NMDEYP5Uk260{}x{vlWHE&EV&p*Uv>aOuykQW(@`&zfVb7oyB
z(B6KDJA_?F`Ae;f>=~OyyFH(gz0Tvh*pv1eoaeW?C=ho)++=CL${khqJ+v|9-32{{
z2h82HEk7up-&Z`Iafv*_w&~IyZZ`@8t}I~sU3O=M573n__+iawF63B9PyCGL>gtuC
zN=bu@xu~5<UU8bEKOJ!PJI8AWBs<jzilOQR2os_otIl{In{d_C5WkV_j-g1~=RBuQ
z9wSG$;n#%E|60d-MW58Jg-WtdW+rSUB9$E@zwD@|Yy)GLqRx}}D|3Yj0A!Iw$O|>n
z7w}<sHh7vYzAum-xWL2bO(Px+@aU93?}ToQba0HNVXkksK1R%^9wl88_#bZv)XXeK
z&EA$j*NATM$kUR!KxnDMU}hF*(4P&`_zGUkhhA*wtNu$%OP0i^BQ$4yOKa@MP1y*`
zUw*x4QzhPg)qJD^OVM#I!`lsWk(jFtfyH%}p6(APM{VpCDGhWZs`j+i$?G5NzJ{V>
zWmBc)X!>F$L-y_ksX0*NCxy|orllS%_SD(;(Rv%UZ2JN3ttOZ7G^)VM3?C}AhJiW6
zDUegI8}KAm`ambCy4?`Bwe=|6Zn8aXDBr7KSiDGaSQ@B7I>Ms75oVSDLsT-V36SH`
zGWS~ZM;qOx{ffLOY^i0!r}*TOB6;uUyk@WOy8l^p@Ntq;-_U~z;?=ByU@GEIBWoa?
zO>Ey~lO=27MN?~E`p*_yQS(yIyHpc<S+ZsmTwQPVZtcwSS)#S(n~TBG7SKazg`ZmX
zm`0nLi2llU5(aYC_3YBY_Z-4f^I&s0P&7vW;+g+x>4CQ4IKQD0|3~xKF&)<F95>f0
zB@YW3RlO%Ob1=wT_fNJi<Z*c!PL}YUA;m2OERs#iSKmJOLPz|-Vxqox;Mk3yh~*LD
zb!IyRX{NEcK@D;aGuJXpL#AuU8cqC)eHOVe21L|bo=Z7&h&No48*x(FM*OH0ld{{?
zVwAIr`Sd)%-@T7;qrG#V`Ontr65W+%VwP7wqQ(2Ex5ZzU<sGfYzcoaOX9qP~%^w;^
zEbrX!f9}>CE_}r;GvZXLPk_6NDH-nDnT2~(YdByqGjk5rcV9AB*-T!=!p~@joB}`_
zFwrre1p`=4!1~s=#nG8p6ydW-Ws+$8mU)^xLc3Nvgvk3v9jWd3D~_uNUmn%goY5;!
zX(u<OR0p_^XmNVk8TN>HO`P8oVfR*%DFWwS`YUgi@y(lUjnO*?Iv8_WwnbOH*Goa2
zx^EE1gP~Wob?<Ld$}!ijEX0x`60U9gH{pjPb(4o-n<P9x2i4)%6IZK$Slb2WHY0r6
zs!N>&<pqOuEb=1lc!3Ia&B^x0R8S0&@zkA<r-7_zGS$yQ<XY)9r5iI$-a8Z{M2C^A
z_J`dE4w(aO%R~G4yz*LgDc`@72c=TDPHMomJk)3vt`%*INq8ztjH6HKW=QSP@L$Q)
ze!t2WAZ^{FqIr=-UZR|dpB?@>wqYi*5KvXc%KAx+MZ#%Sun+z%Z9q8Tt|M?kCoV!q
ztV&H;;50*M<}==FT?+TwULlfe{?9b(S&0;Cxm8<q&IoKzkzYO%nnsPH-&cR$No$%a
zZm!?Gi{PI9>S;tja{4!0#y?(TTdBo(g!Nd)ZSANcM@WbJBzc)&bEXm^@R_c|Xo$m2
zafZ8o>fq=-%#+_pY}jj$maRbT(-FbqwYEr6-Xv@3XY8sXf%;9)Pj*b4i3%32GJI+X
z438+0C0(Vuxw7Qrb$WIDL;1M4HB5O^I-4b`rgnAeP^ep;7<|L+c&ohccyA<A^v1*g
zVXUGia{(@OV3P4fd)@qUH(=KfakGumes|>i4|Pa0=tl}T8nrI8Z2y*iWjA>hgZMPt
z-8J2eOhdt{xci#8CKYTOc8UKv9E;&7&*?tKarcwBy7DCxiP2AK*+FgL^``@N@HlrV
zGx%jhbzw$PH>Hdx@Yx>YT+gS{0J1Fj{92No8-k8&&4KI(XXe0WCQB5-{h7r)=`KFg
zZUF2r+0SnOzY^gB&Li+Wj_OpIb!E_Tc9=Z5ZB6ADRVZ)S=-j*P24-<1lqU7YM4Wt@
zkjAE^dP?-aeh6(myinZaUWk@HjNNpTbeR`-k@g#&sD2$(`ExJXgC>U-T;yQx_1hDY
z0*@I1)=iW)vC-s0@DsD>9~Y~@tdxcCY5+z|z`)r2B1OsTF;))T7ac4%r$<c)H#>J2
z9|(uBhoarfYrpKdi#rebU&aPS06W&s=~o>q%b?aZ;0Vyr_L(I5NzH_~deV?EdNY{4
z2&v7PCeOm!=K;w5vG#mg8IC6Fh<OO9?i6@AZS5@BNL0zBVpa>Z(MA)MBME%T>;<Ot
zM}LTMB*Y&&R4^g}7+S|w5pn<7z9jc+`g|bLm1Hcj*95KOvj^It{gO(mmyD|d=dDhw
zI8!NosY*T&p+8rS$^wzblptrs8K_1befnF^Z$DI(Z;<{0t2)mG`@GY2;CJ(9GPL#;
zb><Y<blU9kN{b=X3<`IIvLYIEALW%v(msX(kcMv;2%KpJf9$;TV^RpRc{fvwg;s|5
zCwCp7j6aZ5-~2X~s^LeYYA-%TGdmWF;`^HUgwf4IiO^1=abL5~mNls`=ZEnlrXBs#
z+=>o{ExKQSJL&b!Zr0sRot{`w`MMs(ysGlXs$JkTBW@xk8b{@j@s}lZ;PIy=e*{b(
zV5_R9g#Ilq<ANIP^{@{h3$vQmv__g{eg5oQz265~@7cT8kc0ocXI}y=712}cuymv%
z1Wn6HTX}7`QGvzo1KR;d?y>LZ@?Z=_?b$XYzs3TFbs91>4<{rl!~*n!A2MX=h#T4}
zp0{4ee{XrpHuRRSrkWg9lTXYG9aDvnQ~JUpCtoW*afPJ)6|8PkuAbm};&_U!9f|8|
zkv|D>{xEN+IgRm(ZN(z#inq$-of*rwmU@md-+BMn3R^r`xO<-_Q9DHZ!;mdjW||JC
z<A@Cbwp|zdLt2a9^@Kka{^`NsR~UX1h-#uZ*WrYMN4%2dfMHQtyyjr!O+2#xI7P@T
zZX=SBt5odYG8%6^@AzYHDLPria}x0TpfbC(dcT%o@vgmEgat~@(|1x)8-90KbLRy9
zV|u)vo^(or&^Hz9d}-qF3F>j&Ti{ol^V-c26z8EtX~x!Zv``qSsDR-U<%3eKKD-yY
z<xYMg&XKtu7t7C;*$_$=mj&FjeO7i;vhWS&W}Am8Zj7<Cq?_Vt7olaRE0_k_;XVA&
zjMDc;ZB9EJw9q)szYtW8IdT$sal0AJ;Tqu1tFpk-a9prZ_lI0z${|o1dZP7);8Vg4
zp4_*I&;?A0>#|0gT@SJ9fd!O=Ld+P4_<JpcA|1?P(?+76nw#+ito4ShBsA!wSJUr`
zdL3ip0gWF5t!gdKp>McRS#?a3jLB~*Jr>Zz<bU6_F)%*`M;lQUA2n03p6B3-bS_z)
z>XfDaSW5snf2~zgB~Qc7&LhB;UrM*WRU95~5h}B0f=<30Bjkw{+4k1ou*xk}%vbDD
zX`b<L&DjO$ss(~*=AIJa7-A8baW-XqA$NSCADw>AVlSq&2K`NPKtJJ%vPZFf*e{o4
zS>piW0kMeVpNbz=t(I?z#Mx=8pqVP=YtR8cx{QQLQYEW&dp7&7d4rNmlqo{XrC&a<
z@)QFq>?^vjzR+wG9;m`f3-&R;6py>Y_$YJnztUY{l`&3lnQWh(72h1)lOX(7IpQxp
z96+&GG>>zKsL|VSuVQ3Hjvu0ri$P`N;<$6aPhZ$cck<^5)o+NRCLLHYDJbyEXR$H6
zW6uhRmM<Z?U(93HLr=$s*yG{yTebT=hx>WG1cYweBF*AyI8EiNj3-B$c3KM-fo0@>
zeAi4KBebbSHttbd)R09e24${DBF68iw>Gr#Y5%3_Mg7={@-2;T6~P=3`6x{%XZ1)g
z`pHM)%3@nGi4SJ&BrXs?6r?121DL4LOa+!1(G0)@rAF_=OW|O;J@Lu4bWjgX`t8u+
zetQ9J;!cgR=6F<*M@vV!7qUHS6~aB_|2kUWxs*A|1W;g#6!Yy>CO%R|4f^+Z+uXKp
z7Nay+w3m%1W9<xPU%t=0)tocB)S<Ck;>$Afi>i|++kc|9vu9_JytL%IiCC`(AhOMn
zm7PYlUdv{}KSk_u6XvnCpoB<GEO|4M2ec%I-b?sT9E=c`moY}dWQ~l0c08XQDb-2j
z-fG}|U4Kxdj{xd2n-qFEnM*!Le__={UG(vN1Rd@JJLUo(M&h{-T5{g`yUY_abIF2f
z&Fo+(6D~6S$p)a7c*3Z9sZL&mvcXbBg?KiwlsBFUl7Fp50_0$l%ug_43@v9Av~p?E
zv%wYPfew`BOiI|4?j-<~MU0mL8vrCb2L1R^3bc^L_^8!;JV^;f*=7LWOVY;}i4$kO
z+XVP;S;JGn3@Ag~Yu&+Bi71RpjGCfy6F;CpEiPaP6`4v%(eO-iQ_-%+q)Oq<ti+dH
zw8m-T#%_nWa&oM%iuT(xBKolh-iv0!QbcH#XkzFgFV);Kye-JLd~UX#G{H`t?G&KT
z*@R?|V-FJs?h+5>%Y3~Z*hplG1M-hHMrQ`(si@$MMUbSqupwN&s`})!<_hYnAPn|!
zh^-Az3~uc_fMoo)Tt<n9a=gyA2)@y41>3Gy!DwKaOBvm@bNV{%Isnn3ftCQNpv4!A
z_)nGjEn9ApGdC-Xf*yO>L!6-kJTZ9T#p{cqB!!%yS))U9P{ROp$?nkA_n&_3K#S{&
zfb_z|VzMQv6Y{KKn%p&gKZT!&_^a}~iG0QTdG9MNPNj$}rdz{Yh&$a;aKm|l`N5#y
z4Qjx?D>{~5z-$Dk*mt_i5w!6Xll>?6cy&TTcpbVI-*r^0H~4daV~n&5&Ak<+x;3`U
z$K*P!#eczKk#C@xWnveHc@m${uCA24B%r)7b~fifsj4;C&D9rSP?x<#+Co1Hu|4X9
zL8FuL?y1h(bwb)mgHO-Rht9;;930x~7L)Y9D@u-aG`cVO>>~}=KikB@jSM8AP%ssB
z5KZ-zW{Q)unv>s*@g-43VTmRPEJ%uSssJdVNk5J6hA~T^7}`-r!~O45p&`?z1<0`C
zTpmF$b7wd6!SFXvpS%;J)_ljUS2WqZO`K-_W~H0?ti_ggmw|hW;O}SQ_*!D6hgCck
zQ%woLt27syCcm^1bz>YH&1aIlBu|1$n}2|5_rA%0fw5FMJ91&$Y^g&D4T<Q9&q@Hy
zgMukfyJPjcyBGwNHK2GaI6U6-VFTzo(aQi3{yp;u=JUls8LqpMmN(3ql6)CdxJW5l
z`gkLb&^J9eP*g9Ow3{(*Xbojc7F7cZ@lN}Ay%9AQ8~|jaQ=@U`Dh=_y2v0FJ1*(SL
zNQ=sK6TJMKC@@yZOOZAxsL_Lx)c@jqWU2?iw30k*jCCvH`sj^eh4M~!$3oeJmkw^1
zSE#NozP>fq6K~@7k}jH5C49w%%JmL^bu&<g?1W3TR}XNRxGoCS{j1nAgbUZ{tN+lX
zSF&UgSZCv$)2T+vB))nd_n!Nh9s2}6Pv^EUL2_vE_2CTZhl^T({_`YY?A3-Rfb|&<
zO?NG0_>&A~xC8DHHlq(PP5?D0X+zVOh=~i_8yx@AyZI5RuMdseK@i1->IYvg{@4D^
zl=OFFW~ZVAF7)BvR}73t-xR4GO_C7OnmWbsW^efPWiqAbK25_uGg(FZT?nO&-`xSC
z2Ft(o>gio3Ok5_z0Pn-7P9db_r=5t*rSx4Z7Tk3KcFNX-*JRsFqT!HXBrtDIH~0&x
z%za{FM`j8u!pTE4ci+%jP}9mrjl|c9u{9v^>cKP>kUB%t#SE-%Uq8MhQU9bSB<Fp7
zlKi9g{@wDoO#HO2d8Xagzlme#yb8x+`y6H#CaC+d-;oq95;~aN|AQQ?OtIc~d`V9g
zxh@|HXX%blWGA<eDWUZk8Te=s?~^3#DDa|Dm026TE$$P)CMTG$_kEh=j|3>KBsu!S
zj?er3yher6Cyhp4crd$k2mRMak&*;Eg6BFwYG*n%l72%lZ6G(`U8S<&N|~ZpARSe&
z5=~+Z_$kGk>qK3gYao()$bnmvX>y8D4S&6lq-ARqYl?x4aD~`5Il&bD0P1QLg2PJ&
zM#C+B>4+^(g=PUz1o-WVxR;IuHrm(NTU!|KsgO0;+x|61XR2mFi->WMWR)Zxc=uBg
z%>7fUMk`;P_)7uM(ge)t6;5OlyuRR5)3h)uFX5)u;Mo82fv75!9XI5hNs<LhwGc6e
z5mJnUwf>EIli8!^xu21t(=&u)2CoM)VhP!$g8_|hT4Mb+JnguWPL48-#GPi0*nrB-
zqAq#u=;ZJEyk|tvK5z{D;E@PjPT#&)0UhQqv6E~s5|dv@RdhoMe@qm-B%Gb7FP7U}
zTcF;dBB!6vw=>EJ6h}C}<)Lx<lwqWD@T=(V)hNbGJ1W{Czype9Ga5f+zen%};^&d_
z`sXyUJJ<i(AMAC%38Dh+(}fD7L{AF3H)JbUamQOuPvN3Huk>0M(TxN?Lf$t&bC+Fg
zAs27Mf{kj5ZcG5I5;xbUU+yF$d~d|*a1W{-pBE4#*dns3d<_vIbKD^(Bw5ogM6Mm1
zH=Ivv)YP{LFp;MD0Ws-9=r`FOGBIT);$(GAy>L#lHYNUIL{J{em6IY31iAa}+QAvx
zW#>i$XzMpGOJmO;YR=nAsIWI0p3?Y$EZ|P;S@)lpt!xm8ci~^8s#w|%D`DgQ-eZ`4
zk6%T{mZ#sXL2%sj-J&DP*q3S^EV@cfW+|4^j)|7LE+oE7m}T5(f_?s?fqb18y_SCf
zK{uAp1|^SLtpV;01`pdJw!Fr%OC%@-uF!VS!$s7o2_yteJ1fM3=9BoP=Qr^=b~hS5
z5Mf?v88O5!iPMEP&+Uw{!j8EFeak>N01&?AGmI97c7{!U(WOx}V5dqJ<8}lH<F)+I
zlt6RT-qtJUXtakZ{N)+x1cZky2Z@>L7=Aw^H~u^PE9>$J1vj$8p8??n_GJw&txu37
zOcn%iS6cK>s<Z%P3etq6@dv6QlR_X3O4ToiRJbQO>^L8Y43c(UI~Kmm9$*GjE_8Gf
zIIr(TWGGb>R~wVdGqW?C@iCdamMrB2_m>J@t{X=QiN8%IhO4G$CeDjJufb^(Qbx;0
zn*tJ*6qS<<l$PK#?Z1F<#5@%*sHVD|`4ugI&QulJNDe?=XK`(8^7nmkS7Du~d6s98
z+1J>dro%C5UHOK`|3zEPY??<LYYfJPl>5t-?GF+N>?{LOI^Po|)i#9(RH3T!7qQn)
zqe+FEJPfm&#_z}kO`zyG&Hy^s_fh+r|Lbg*C`yp3{&pt*(?m;wv$~o^D*p6zKmRWl
zel~5F+nfg~lZCK;H~ke6u?K@~8fWyXXW|uSSlikKy4I!idn*@O?>rP82~}nrMnQaU
z|81)!6f>-bx+yCP6Z8^4IRHiVBhiuFyT`4m&oSvQ_e?ub^79#H6xl<&6uHoVd7wM)
zKaW_<3C!|?=ag}!U^1iORX}dfl9@k*12PkI(VOKJlEM}$cOQ+JHSE!BSQ|giIK(-1
z{cOlrhii?QD@jwmX5p$xFOigm(t{Znx??+71ug7#X@E3g4brXY%(iqu3S8azH)<FP
zWrpJ+vC?m$v68WMYp1nYa`nlgVA1Y29%$U)>Zl!4F9BFd?`YFJEX~9Oz^|;A<od-<
z;fK?diXWDYQ_^)rX;aifKKf|-;ExyCxY>Ak4D0SnJQ0RF6lAfZPOSyAGE9;$(g239
z1&M|VIA9J;I_94It|+jT*z?@5{2s`jRp^gX)zd)0cSBm>TfGiy5j8sjgAzsVv==4V
zFef!Xv^=eoveGNG7eJZMT<lC#E!Zl30MY8z4R_l$>`s5{{x&^nDAA$Ljv!<oOHjlC
zQ2S%fRGjo~ogHvpWycPp%f$2qeQhpu&V77QHe8Y*FGaVfh=DvPv5dvGZK*G34_HR~
z98csfC~z64Jwb`x{Z&dk5ix{9r_IY#g0i(|yVw0c(LGe?c>o=`YHNo~;BLL&q)A9#
zCU-t_=N!KMkBPK8=$bx7s=D>&_mPLg(pd|nLI1lI1t>oQ;euGn%t`2?e+9X6xuk4l
zgI=!5wU^5K@y>d%ux=lOoxcLmx~p{_ofe?tmQ8vBZce&9sWg$IRCH_NjSed;c3DSl
zsNP5!PH6TSQPnSJ!Z_G?pmo@&uv3E2Jb0FG@vl3V9oR-fOlpMOetA{bKm{uBt{wXp
zhwmKp7Q(Yk`byb9XbHZYvx?VOr54z%N8~;Fy~un2Um~RcD~021_r$)aNwkF^w@Ah9
zu$r^f(@ElW3nPE>9JTjeg3sSaztkn6NYCvYu9Rgp)~IkAC=d=en%rqo!MAv$jE&Vj
z{?n3bAA1q_RSql(0M)b~N%4FGD22#N{~39at*i(pICV2zS~1511Y`WI0Sp75JUao#
z2hR!uNV{{S=sw#-?7v07ezL%zhIv6j?y`j<#*XH1H1c>GQiYC8*grK@O=k#HFKq%m
z-&Gy=3HP;*0c1f?0Ufulc=Qu61eWcq-M~9x-3GO3{|V}he3Hk|ee`~`F08-7$B7Kz
z?}S2x03T`Ux|qo^{*2UU%7VF#sXAAH!QVKh$aE|jD$9!NU%D92%2o&JjuC5w)C(ri
ztpo?ys?sUyt^~j3cA8?!EyT<O1k3IGjAB~o$M3TjUV9OX68NHU%+Cq8q8p;`4$N0t
zUHr4;M;f#1QR*VQpC8e0>226f&L=^La~m?1#?Gc64OMUJ-#mA1?MJVt_Zu>~(FTwH
ziW1tV0BKp$@PWAojn&8w;SG2WeT@~(L6DpV!ZT$B;XZna*O&uO2Q_HJEdpz=TE_{B
zT|Te1wWtx^>tX?MPs>g!oz0A}eK>N#A<n_x^@ylC#8-QIBvh-Ozi9&L=P)lK@i1D5
zlIgUiHD2R#%ueyYar+>>jYPh6x(U{V!VA6ON&lcus=dl&_rV&+Ld!MHgh6{@|1gpC
zDVtR&>bt0ZhV`>0MM{cm6*4%MbI&--e>9YVSwe~I=Nm>qQ4J=0HyzHUgF8+()a3!8
z`uAOK$H=DVoI|W<2PKOH2{cKqnQ}>Fa$qneTx;U*+KEDVH<pwNCRjWH%QmBU?e591
z@eWqMcEEFViT+pVp}0&?G@=j+36^faGqiCs9rKG4hlMW6zZU2y9YX9Dl{*1jQ)Q4N
z;)5`lQnx8~zAcfroies3J(&0_$qIUD@EO(y50CFJ;FNTz3tJ%TJB>V_kxEx{XHjg&
zci)UK8mC0|g?Rn2iL5j0Ga%4(HXhfAPa=U#U}X~?i<2c_ZqcK)A%!GBp!+JYM_QPI
z55V>%khlA(vj!23h1hf4kA`IQgk{teVuj>-T~ke7!j7!g*kvpWwX4oiG9>2KKA%o~
z9YG$Ykw2KtV{IpnEDv=kQ%f9}?}6KZ2tcbLzEf29{Iw=zi41K-+Zvtz8D44TV0<sy
zFXoAIHoeYofbV3OsKfr((kWq@rg!^Dl4#g){<Xw5RV!UJ0ckARMQ5^3U{*!IdfHJ&
zM$SOEnBh=mO8#pt`tZPB)InAmD(e*(-G%nQuce#png6SRYEc=?+7sLJ+u8JOi9M`0
z1JmW_OPgBEAGPR>sfg0@;?A@`=&*hAk%g(JQ(N8Ttku%%D1OUK_-7MUU15PIQtNB~
z#~s>^d1GTpZeKmCncPu-SHWQ0`a(7>ds|qksr6iD7bRyru#OjllNESmUM5x}Y*0l~
zo283**?m1%K{_}erw!0>0AH_%{^`J_2jB}4vzd-~pdwDtv3BpYGYJyUj3v<ke<I<7
zNHUV5xwnQgitK3#P)sQGGj^&4Nat@YrS265f=JyCtp+O>;BzM3wsI#U^M&G2;u>Ll
zNyY>lThk}2+)jRYcd}rDmENyBfeX$PG#k=5hEKpxnTa8D1tlE}l|2A?k!anu`)rFZ
zN7IQDUYx>hg@w!?loW5F$thv^-5$Clb)>o2E7;ta*%5=*Xl7dz6)fNo8D#ouoo;b4
zBz5l<7dsXDUVp)Cpv#m+Al({~o51F_@ihRcAYe;tsy-dXu=EOJX809)0f}v1@mWl&
zG45%`ItNYPMfZp5Q@_Ls%9Y?*Nq|~-kGA)Is=S~vUZg{pUK@>Ze65w{4T~VIsW|5>
z=0#4d54u0a&h&e!vZuwIYf2dl5Kp{Ld3O&@Clo`GG}#3Kg{kY$WLeE`z*bQy0cCVw
z1K_mqV|Fk!PoSl(%{xey%3WEii{L5qSex!Nm}8h<Yf46->ZpWwC<DP>c~J-7Qm{>F
zjTgMa4N8u6AB-RkFFc=V!buFtnl$lhZ-nUia-_A1-)N3j>z>7CW=*l*Ufa#T0>1ox
z5hBu|i0y-n*FdYyG)1@6O40H+=OL3bK4y0V)s~s;249&KW?Q`K{~&eKDx%tpX&0o|
z_C5hIICy(E?YgHcu<S~j;wCQTem>)LxtIS%)Z>oSy+o<b<>IggA~>|@G1U9VUixN^
z%Q?Wc&^3%JxRpz_D6O$Se0<hxetdK#`PHE(%9>pzhV1J5x*lok6|MCrlLm#6f-rvH
zz5hu(gdWO4<ho2_Y(F3#LbWz0uHm4^ffYVZw8H;Kl;z{X4$V{35l#8rkAL|;F>W*b
z0$%UpE)Y?h=xU)Ctvtmq2$dq5l1mn2E{MKG=C75+CkH$`55~C5t4`^IiEyAxs*u5)
z8ohY4CR1i5`FjqugzR_>Y<=m-PqY>T8PtGp)B;Js!j%q=!q}eQ((KXzf#&E<E-gM*
zO$LTrd>Cs0$U;-xdc&+xI=W+(0waKKeNXX7!AH3nO&bGG9;91sguXC#Il4KcV+OPg
zc{TOyq$ZuDK@J*y)}^%OBMTiGVe2_>BKl-mbaP`RjmOXw1Lw8}_}|9Cq4=3#GTlXf
z%%{Mkn7;$KNB}CE#xt7t<qY94{e2$+lY}Es_Ra~`D}+z~B$3!D+;KcipSysf?=;Jm
zxpM!*A&}^|e^djeb)>z0#$fSUTZkFoM4f;a3AgDPE#oV_q2>hgM87gNN@<#1x?W7J
zp6fZ%i_Y&=;H)y_J-HXu-4-!v;+aBw=0Ar3NF0y0?jSs94+;NKDW7kD!#bShluLva
zc<me~WL=4R^GF6ZFH9ObnSvu$G(TGHhkO}rsYwca=k|;suP3%CT5SH)t7TMFy(8Tv
z%@-PZ>=60+#8*T{<cjfA$0Pd=7UP+23#axjNqcs(vmOU+e_iI{;BndC@f{twzh+M$
zGBdHwnmi)5#qrn|lkju*P7(W`Z;tsV5M2RTM177qdQo@D&kuh$R8Tq~FW@_octC9X
z+TGvqRsE+}{%B#;_Y46CasN!F(I!NX;^c)2OJ1~H3rrCF`=e3EAm@_znT?ELm8aIa
zG-v(fWGI<9%@B$T#3|s#L!GubJHp{)=21uL%~f|YR={y<s+vgE=sta%lQ4GPMsCh_
zB*k^obqC2s(BN*tw_=RW&Lom7veF9=xLc^7D&S7O*`tPzl|um(>?~9sc0V)GngQ99
zwnCrm2<SVPw1(Q?ABX99Gu02$OSVqfbX#%w0MsRe{L~?s%SKP!mYZF*1gO(+z@H`J
zFVe0hexBxfJC0+;py~Ac?^W(sE84HH6PcOZK{u+nI0R=1S7M+PXHRRTu+ISY?9&2!
z-!N;Q8%qC{0LR8|5f&!Vte+jIbI!^EK;73V1iff33pdyiIbo*Hc#Ga>fms-vzA-ZY
zt>wDw*9U+am8hODS+Auky0zG^Za|dCe69|Lo9Yyp19kWu7y#HGo*lvD00QModeA5^
zFYp6z6fQtJyl!%_@*Dd}(5IHvP(jnMR1j~VC(}?e8WXQ1As#bn`Vf|MRg&YUU^K95
z>FQtM!=7kCd)C({DEREpHbxie76+HNoSOUhj?K9l)c`+BAD;3nD*yl&O<qRoEBlIt
z2ICxg7bkkXp4;o8!Ujn)^c%i9x-);AG|xP%p`1Z#Yr_KvhpDx^=}G_Fp~9QPwx36|
zcxV(+4b=oq%TH!0YeV)D5<VuO137oB?V9(|JX75r*y3k0rP7E9ra^W#I<e6fO`>o2
zQV6QiaSd!kUMfx2ci*!rkLh?Q+p{$sHnu%BWJHT(7G!*W8}9C31g_?H@3k6#3jO_M
z4kYCJBrWeSm>_3AXHNlj&*yXH0zOpn2EZ&%RP2;hD*|}L2JJs@$4}|ywbpm|;=vZ?
z#fK0j2?K>0=#jP77+{s>*~&w)^F-@drB{1H2qBpG9u3Krt!CP<2)d5Vp=BqI%Kh}t
zQPOQ5N|4j#phy;`Y}^IaaDrF64REX>fgGiC3`mv<etjc9+P51W*AIVnEW?vyzwrU@
z8zGLN>`oBSo|T`+8)K@B*`#Yn%ZicgXc6x^*S8x`GTp;U43%cj=7)wGBhvxx5iG{n
zQ+o7R^#tUfvwx=8?vg-FnbipgxNSE$@eOfSuvM@C-PP{%Xkq}lt@Yx8nwjtTU~Kak
zpdW@MMRsWDf}V2+WBhl8E_ziT6z35?klCc}{_Ak&kHIyps98{>k&Z3kR}?fSaj%Ic
zR2b%pLVBu#xv*x(g+0*cyLZ#;BGMWQbn902@n;B5J?x*;J@NC2d*9R~U|<xMGArHv
zsI|7HgpFm4>BiOaQ|f(Fl~_DLBGyoG6%F#n(e@4IR$TNEexM5O?$8HqRXw}3e)VqZ
z-y;^rZ}9tY3x6C<P`Q&GZM?V@Z%q+W?siYbzjg5Z&GYB#q)=|`4{fTCMo$iW#M3-q
zF+qF;H6JUJq}yA@Ju#uu-T<aeS0UM>FbTXmzAK7_ndGR(bN_d6yXRg+?kqU#XB@-s
zuo&!L<>_`B!1dey^VLN|=Yz!5M(0exL6$qohgA22bBV#YQxZNL#R>nG7^|U!NrL*(
zR#17{cIwHjo_yc4JQs4f?S8Pkm(sTjBMoYIw?e3tusaCr1WM4=?yeSBCdv&WG1F08
zfYESBT$B6O&snBXM}1+AMztf2{CJZ-NiJt4$5VCG_Wyx_=m>9auh#+Ar&Gv^y%hza
zo}YU`4?g0%q)+Fe>j}hHw%;RH+uoi%(DD&HnG|*K`*?F#13=KHnI5pf{9qKI1?BrO
zhxEyYQ3c(YpPK0W;&<|TVpxkwpCxT(5^*OOb~I-$nPX9D8e9WE2gf~E@IDxyJV~bS
zv<7B?U!loyw~*>!CKBM$MsBJ%vnXNUQEo+nzxU{J!hXGIa{+?Tz|Yr%tHbeQOib=6
zR%CAgULm1h)co*yoSB{Hu(}FgW_N)(N+!X4@7`!TPYT3R@liSSixa%VZ23U-r@N8o
z2;Ti0`1<82VYZ5fZK70wHGnqxT{p>F$g0Y(`eJ)%R;q-Ri5>b906Dc;0vRu&(6H|A
zNkLnYyPBnPkp&NaEql4DkEN3GR4$fc`zv%<4ovLk-ZLMl=RopyOT-=p-@eG8kOqKo
zfs*)D{xMxFBC&KS-4uWLb!D5UipdS}F6893@RFsM^5+#LjIU4p^6JyW%DX+f_U4Md
z**-JdYo3W1%NYN=o1vGwF8^Ta*1=qSu6ah=>;QaC9&Eplj6z6%8ne|t=6PJox_z@9
z{Mb)u<}QqaI`e)>{8n8LMTPrp0wnfJn*zN4&$L1@$P{yvExM;h(rV1poz)p4MxMh?
z)VVQ#%vsnkAgsRFdR1>H_co`_Eb8!_=So_hCFUQj2(~%0?sU1;f%uOF<G;C2hJB?Z
z93nt!Yz{{Znse>2^E?x0VX}YkorA&ao1=5Cbi$>8A)Atnj^N$4niFjL>?b@H)!dAT
zY}hmv%#VD<0yZMARlCbD0Og-QSJoGyYen5k$A)Xljn>&HKz=|1-kwk(5wZ1p8EJH|
zrZ~5j+z=2ls9PUmgpp!!lr=r*#-UD35|RvJxizpNu288mb8PAr4E&H+BAN55j3!wo
z&2(=H(L5{obFmuu`fu(lL9@x#Zm_Ug2DVBSJ<+D43zp{hQeg%V7Sc&e+?U0i*nUwr
zyD+y=l%&f22o?OB4A$js3JkhIzn=J7MmsC`#rP;BgSVC0p71ne<Bt+}5X*tb3lMH@
zO?NaC651`lFaX?$2^pD`IHpi1s&d0Bv3L`M*(-bqZKHAn#;FP7<)u(a)WjxBvkGPn
znO9>%PExDU!31SY5>np*uD!+z)1(bWKkP2jF#35nw4XA)69n)-lOzn)e7|^}*xqDF
z%;Ud>ZHlDGr(MYsgT1WgCO-}~`^0G|EtX+hy1j-vXWQEr-9;`>10MDtyMNtJFW=UQ
zQnG-MT_D?*?=3-_NfT)Km>+R}k-g?n(ax@wM1jx#pI@w3O5z}?``lRT*#G~3^xwzI
zXs^(oxU^)Njb7hG23)X>`q14oZ0p@~O?K5oL7ta(+sa}$6H1X{8~0}o0jCv0k5{nA
z+>8q@T)LV6cCxpMm&ySbr69^8*O)SB#Ac@^S^q?4z#+V2AKAp^_e#CZtYA$CAc&vk
zzm3QkYd<e9A6sY8ITej3A0@2Mq*47byx+ToT%LI}(>IdBFrOnN&66*lzShaQZ$Ce7
zKi`@agG7e>=m7~|I=((<1l_Vkaz30&JXAjJ9X~?O&jWS{MLWFjtIkcx=?Z3=`Z532
ze|x_j+c$fNfm?c0qrF;K_Lmuv?~xickW#s`f3X_y4Ts0IfNkUytfN@Ss=~x?(+UQ<
zckr9Mp$b4$Bufa6p*C!3+bP|5xjUggC$fARrlZ$E2KPAI8vdsNjt{fZ4;#7E(xms=
zr1Bk)rw;dAqoC`o2j%S<uQfi<?cb)G$Va%=AnCR-bCRJl$T8KSZ6EKU2jroxyyRB0
z$Xn>{wOY2RZFf<=>}LIxVEJ|eA++vrpJ@zibGFy_*MRhijMh{jLwLPTLl-jqj&yy;
z=$u#;n{4man!X(Ci5wn3N=zcHD;UY0O`7~%vN(*f6g%#z4a}7C#~rQ7Ek;kW9==%2
zbT}(@j77ehczB3TA7}u%-98=&W`qZ9+Bn|{cvvfsrSJA-iMG*QS35blKU}3u28?^$
z2AupT-u4MVVrOi3oG;w01WP#T`)}vFT#7CW^o>29$c|!zPBPEO9vU9|ZgN5AYTIlE
zA|C#`9S@>g)J~xiw^ej+d49G(-en0Zx8QFFJa9b{ZX?;~gdWde?J%!LdDu<cwx5Mq
zsxauJEorpl+I454{e=kV+{D1?VZg&c$l)?|l<P5a+5H0-_Nemx0iDOg1J`{sZ-Cc6
z@*)69(n|_ozB34zyfFdY1dSRBd9-f_WuA*ZF5_-*o$K^$Qi_7k)|a=_Su@8fke3D?
z-eKjJYmfJdx+l{abPt3NjpyhF5`wpW$afF~U-z|az$KmNW3Rz=$@w%_$2zjh!EX-<
zZP^|@?w}9~Ky=*SF%zg94m6$fKCWsgru6jMEh`)?KWqg&61s8dj0QhW$c`rEqaN}Y
zc24fU3Yz-_lFdpr3lK?IzchF_q7w?Z3OW`%WFO4>JvQ*@5sBMta3`1%@ECNw2%n`N
csk?aunhsDpj?K0Y15h9N_ewI5cZT2o2d-x<JOBUy

literal 80079
zcmZU41yo$kk}wk7-QC^Y-8I48A-F?uXK<I`9)ddr_rW!|4esvx^X<OfeY^j?r%zXR
zSy%NveP*hw0IJF|NbvaZU|?WKa<Y=@U|^79U|`?^u+X0~1glY6p92#c2?<p>2?<hF
z7e`AQI}0!{+C=w+ZUqPv+~Mx@a8wvBNnTFgP;?!j#{6$=G+zC*>G&hTnuh6{JwZTo
zr0&m9-Iczy_h9_*)(&#;_>|<<-)#W;^3?}-zvl`c_QoFi(+|A8yaJKcOc)QK5_lLn
z_%IL{nVDdGr!IR?(8Lhwzb8sr@QG19-rs9=hxEL8WK5w~sQ8>PxE`!Z@HD9&?j1Y&
zt)7fpOQpVsP?nE3f{}?cI%>z<U^~~yDWH)!Ef5KVvOOLPbs7?4Fth+Z^OZ?t{*=gN
zdM=$|H*HDJO^}<0^D*;+-X!HDUo)hG{lVyDGf3r9KzX@!mMVl$tag0^_-^JVYMv?R
z;e?#qI=HHA!K`(mur}G7&ksoJm8#&>n_w!-6SEOL%gkFfZOc7q)18pnh8%<6{Fxc%
zI|u)&jhL`Oy9qDd+oE2cLB`hGcWc?TJ1gE?1S#xbgmGAWDeqJg`t-S`g|?ifk`fsG
zCkzV)9$^Fa<r4z`Z1|t;b9<7)!C*dn^v@=l5Ak1ENU?m#|AN5<{vj0mDj_HL*?%>2
zv9NG(wRUvt1NdruidwMwrtPM!q{wgPXwPh7?r3Vk>}But4-1%}7yl<{Z{cP_>Sb@|
z;L7hMME)NX{GaeY(=6nq|3TtrD@3lXq)IB`=wd<2&CJToN-hjfN=ho|Vs6Q=E-C#l
z_Ro<JxwV^{6F&=!r>7^gCkL~mixmqSA0HnJD?1B2JJTlxldHFbn~4{bgDb^<3i&TN
zk`}IJE;deXHjWOY|Hw5lb#!+VA}9Yx(SJYxxlRi&oBz?|;QBAKJ`H5~2gAa~%*yiL
zvOie`|C!}iwehmB(~-2X|8&o%4q*;9F2Vnx|Nk)mqw(K3wf~2cmxJTqIsXmwe>pW=
zEnFlV?LSqz3IET|{EPYDk^f>8WckPPfAhqDy7@n5Km9BWFUa!WdnOG31ol-L3``VE
zPEzcf7x-x>Tn3|z`vF0x=jPk7;n_=9U?><9ESbr@6s}zin^}A;5;t^*;HsI=zR*k7
z)XqUX-s;rN5ckkNI3zABG{N}ygn_svN>p)KS?f*D^|$-8lHV^mn<IjmYC1X^I_rNw
zUz+o?QD7tn4f(sXDPP@QkM}_i56K!YL|h=n7VN)wwL@sF`!Jh-36ZO>e1F^%g3I)v
zIeq!}&RWzsqh+Q}DRD;>_S$G7P<vm8wvSAQ6(6<5+Z|)ES29A~MfPA;imDHQHxQeU
z*sy%1#eSvG>8#EpPEbKc_7elX&!i?A>Z6x4j?bG%;Hdug%FY5dq|T}85c@!xuxd^<
zjOU5iZcA;5V?@DKUrM)#!csvCFZM}8&V)_SA}2*Gb$WtR%0F%2VRnAJ>BC_3+D3b+
zL7q8=d*8|ddt)P(x;#(Q;1~Ycah$WxBS&2K*EW2nI;x{HgJ;u?*jE5UZ$6tJz>bEG
zr4&-+U9F`@-5;axmdB>dv`qSA$%^dW<Bs`(VZY5lk&lVntI>fa`QF=RuT(=<j);1w
zVVtR;ep6a83KaYOPI$+^ZqsFEt;RjMPoH^Ad}3nCW?he8hjY2!1EbA0uXOU=f=?$>
zFFS6j3)Rp$W1-{ss1ApLZLagUf>}#z)jByggG@vEV0MJ9o#UuA&bH;NURVxFgR4Z9
zAi<gp^g-YFZQ~yzpI7HhqfTyu@Nk*>hr*KNfW)q>hYU}<n$Zx}J6)w&k!+!nvGMr=
zR(|0EyV1uXhb5BMkKAxt4Y*N+6meO^SXd&dA}7kf%hq0#QxgDgHr)M_*+=V(JWuIR
zO>AsY!g(GVHj*P_j0$>@#BMqvUbRHk(Cu%H=NTLeO4wr>$ufC-ob;l8X}l_NpWh27
zY2fzU_si7AuZ@BgQ`6hcql?F56VI|&;?IM7;zKV@d*VNicFO6kG>+1Q2Z=){rXvIC
z#ctSrk_hS)jB+YlcB_%N=OomTtdtu$Vmz+4ts3$NnY$l}88+{DAs%I@krzd2I4Q;M
z3M0G5zq>7ejg0CT?Q3%7Fg&wPTSYmAiQu2Y&fii$2Nbq*mVY5ptn+<FYoRQ8Uqymg
z2jXL;>Z}<4EJ-TzP0dd)cQ<5%SWh$Hsc-&vrrXZhaua>isekXUF!g=Z^K%mOY{kW-
zDZ%RpomS^=Q$okBL<W8Rkhm!O(<Ta?p^!~R<y1J5M3q)2BMktZvo^3utniolsGLbk
zR+}DURc;lmUYqAwnXIGlna-^d(1*pFiSk`M1H7E|?zz!HgV3SwR418A%%C^fCC^B9
z8E5<SQ8!#IQ%^=(d0GYbHms5fKsXqyCqmXWG&?J`d>2*;9`Mg1^#Fw@Ew_v~(9z?b
zTsjp@jwFBdmoxxYiI3cpGYW=CF<Qe4MU^E44e)YC4Di#*DuQ%;4OqzV8>eB2Exr>i
zEc@_o^=9_Q#zy$R6EDEsb5LXm-M=YRM?Lo#TiKUW!8{x)eA$;o`#l1`UVEAsgXCbx
zM~JI@uj;`N&c@Or071%nQNJHts|f4W-mEC<!`W*w5oX+m3bdFQ-tO$tpQpmMS<N<^
zFw^NlS0!Gue~a$vna&vh_Lrc!oea~}B^*J7Cr)$<l{S&fiz0y8R!CwQ4-TSgg)5y4
zC*F`^T7IDe<&0}~k|nXJL@2j5S9QTYJdEgj)%AtEDuek#u)ykiYb!a|X(|csp|2tE
z@a+ipn{+;@t!p9JNMw_X6OG#!dJ5G9E}^;OdQCiGnp21%be$XFMugYwMt%q=oeJV<
zI-HO|2C-<dnyG%5jR*@djvsj9QfEtjk8^0hf&<5Nu~M>2abR03?;xZO)S$rAo9Joz
zlvq${M|D4s>i9|6R`hv1in9NG_I-|mxI#b0+Wuk{V)#FYXU`8Q(hcyxa3Mn#{rqYy
zW!?IE-28&`mx*0}vK`C+UUM*yWo8==H~FnDI6WVgg%U*&VOv^S_JyDb5ju45oI8z<
z(1dd+i*T)ZyyRT_r%U@h+mwY)Wbs(YEj3s<ww|2(G(4o%$Hd2nHB1RgRPRh)WVC<0
z!}Ra`z%otxovgx7mhC#3394>Chih5;@gh(4cgOvUCUg+jMqCx*FX9S{T4l6^dvm*D
z=Ou2vim@=6r{)LuD8*Gf6cm*3f?nUr0A&XNM9X2L=<4Np+cRo@OoeZ;mhIfYqqPJw
ziqe#M-_`BS6p{>KS3_@J`Wp;#sH6YMA67OtV&;HJDx~+9dlJ~q=X-Y~2Lx|C5+b;w
zqJAA|WTmO_x2AG+O+t4(lARyDzM{yWN_{wkQr_4Y#4u2wj1(US2Ny@yUMFWEF|VOl
z?_B@M$Ig`8hIHSA$RDokmw8voWdcO401yMHE!_xxXn44Pcemg&@8eLVKL`@0-}@jJ
zBiCoSfkW!>`C=3I{hsLJTudH&*woaQ{B^*U1<5MUPa0RQX8XM-LKwg6Hq|!eDxZ_T
z`=h-ZjoU6Gf&Ev6$djWZg3rTQ@&)ch*pW`#ZyzkKHAr$}GxF$uiR(_+I#-G5nB5VA
z*};)%Q;E5@%u_7y!87d{N50K=o8fIYq|sdncA10NaW*(k4n35nKb{+G$_VueJZ$92
zc8ZNN$|lzF-(3n|aHnbtn{C9NjLJoC=VmX+rdBrP^9%<Ok&$F|_Rd&L=xXquJ&#oE
zSXciDC0rc937vT?d*#dHr-g*W4c58Ync}KlG<8?fL`{aqxLV(Q5$AeARt_tV;$@a0
zUKteoDI*CeI%gBLryZus-Xcz{+Xx5DD58m5dYlO?F_)jWM4@eY@K%uI)ECH4?<0+*
zj=&DZAS#<RIPZiz9hcKYAmn-b?~nJqMewW_UbvC2wm2e`GJ^c)CR!{T5t?QF%3R_X
z!~4@>M|kY{ukbOY`k;SH3vdWv-V^esa3$?`Ih!g9i27KG>W#NQ-CYC3FK=pm+`+U`
zGFv|0eEKflNF3qHa(*OL=L@#kdh*_96!o94`?!Tgwu;<4=FMmnZoZuvU0A4NPuM=>
z3V{ON&IJ0W-cD7*(mM{lm0q=z)|zb7X~l-a{R&?dRa9(YfLjhlZn8co#PI0`Kjgf&
zM+`tiB=5iIe7kvL<jX=4XQ7m$lLRS`@^A?d2=<6d@c5dPGdNLZOcW&I1^>>)Av}a-
zkg5HN6gmHb(`Qu`Zjv8p`V+A-Nk6U0j3Mm<s?n_2-PexSZiTfv#CbGVko^&E7E@Vy
zA`~*Bl)#)B?-h_e&gk#b>!k!&)AS2bHLnL-F*BjytgJs9mn1Q0`IoPZz3^$WFr<a`
zI_@>gwbXj1L7#8YwBs^wbuIijzua<lY_UEq!!ZwZ3Cv!E7-v8|b0K@T$}{?7Y7>m&
zH&EGaU;OJ?UiNIgR;k~0$g6$II@WFcil?lwJ-Ru@S)^i&vdW^HV1o|p%zeJftRWTN
zEg**LXpR=c9Es-`^!l37Yzt$vnAXD@q%0(2H#RgHvO^_1L~lt4p{;Fvcrfmkyb|uN
z@g+f-lY7y3M*u{y*}<Lt!V^HofC;F-J%8}0t))V6z1%W?JcxhS^E#~??q_-Lv+Zx+
zxB)kyit=V0nmkp)lAw@z&ZJuZV<RrWZi%1429g`L3>Q0&)a>xK=X4e!joo$`M<sIK
zMKf7lsQQ_*9!$nT7Vt6oARodgpJ5l>zMr)T7jYq|h*gNES)v}IJ|7=`o5hGZ*P>==
z_zoi3z~Z*wKtK1p1hG=>^@Bx+M~1(T&#7o+ww?R%+*YAW@~Z_N>OY){WB?=GirseI
z#_g17M;Alqlabi-MW3n{2~NJE6)9g3n5U*qR(A!kkJmzMm4bsqbH1{*nmR?(*$(4F
zRP`wQ^@u4}lfm7cHa!z&r6aM&%mS`8D-m$U1BtxHbvduhE(Zc_nmtru00@(Xmn#Eb
z66vpCLJ&!C#($GQ2CJ#(Wzn$WX4|Y%y8DG}@o@+&$LnY(an_NCkF#m(%Tq<J<2Nfw
z8K>R5QYtQ39LNEB15PXxz2yj^P?9Z|N82AW$@4Mt<|KYP*rHgs*sm^$Wpav^8*>>G
zycXve$JFX+LKK>a$-$>h0X+P_^5ew@OY^DTY2%4gXL_x&a{rb`J+DNrqVA8&ty+m6
zrHnh}<Y&!P-j?OlZI>1q{(@o_Rl*7+`p~9wnqz{n#c6=1`jvVK64uK?+F?Hl&P?s)
zi$%<Vo~B&oOagxyw=)YF+PXR{il1&^nLno4?Eq553(au$!x9~s$MGM-l4YU93%|Ek
zk^_50s!)oGicIWG&2aIQn)~Yv?Bq4E<M^I`)tU|8*IHe4UG_ktSYW|g#L1T+9i=HH
zNh|$%o{}qyc4)Fc;nqpSL-C<HecgbB3wn)X8)ISMHjfiz_sYx7g_j|AFQ#Wjd?rZ{
z6KV~Fw3NkP$|>wP<+Y0_wxt(*8mO*y{O~vM$6+UubX@&IMd))nBR@6u%MC1IgdRV<
z!)`NUX=F7Vi9nL>=fri9B#}S2{R*yX{m_?+Vd4OC!S}}^^g`Vbi1!GhW+6iF0nlvb
zc(@F)ALD7deXND$SlUFCfc&b^!v;%~*o|o3HNSkbTHiw;?I?ly$yyAjKkd(Pu*8cz
zzX=t(7u*QHmj@f%zMzZ4L+zzDGnplN=lbj0B@C5cuR4Fi!;76$ueqqw|CUNbl=h(|
zp)@<6uX>WMd453ZA_F1z0JDAh-vpA7FQ-k^hc2g}>(Rr{VBEIG)AZJg0m$CuIZc$r
zedtnxMauk3QF1Uc#LI;<oCNfoZrj|D@OkWpK-%*^n>ysb;c9r9`)#U3c0v-2b^p?O
zs0ox>tdV<3GeF4_0-YHut77O#nu9ef!++XgqV5JLKXN>Nr9jan%A63EY&XWkd~b9O
z?QKlqOGLBOuQIHINV;OaG;NQ%R)3<bNqWWo3P={iV=uddAgup_*YNJ6?w?iCS!M_H
zafZTCToM+l#O1j+OCvZwt=-LXDS1;xWV+afey{7z^8%H#Cs)f4aAq0=2h%MDGhC<Q
zxT#NZ=+<eshbW>l+#*@f$EA2DQF~F_JC0-cHv%e~jb(+825AGv2>j2ktI7CZVOO0K
znd}8hVkz@zr4XZ{ZGz{*ty_kA%a5e@6lbAG%t@eRPBKEt-sq#%{%~nr=T-{OHz5Zh
zxR>%CRWy#h_G^lGlp4$K6qNg3C*Oy?&Y&D>zWuf-B7{O5#kS+nVS+QVTl~2ao{|m)
zr8`^V!u0ArmGYG$DmP`sa)d->_hzf$R|geqI|Fa7==-(@A-1A+!>a5m6Wkc3=UB7!
zBh4Mn6Q}G~;X+OwpP^8US2?YA`=R2_)jA1GgT;kE=IVuZ-&OutHjYBX05hT}xTQFl
z?7x`V?TQf?C8prv`LJy;@W@g%<|u`r&u0o2EOh`f&&?P9a@tz5N9|{r#ld$!GjyZI
zVef&Q@lqT&y9~6F9lD;<E;k-NzAiCvoMhM+zP=_^lIOh4H%Q?ge4|C1wXJz`cubEE
z(6yQ5J2<_ZtMN}oCU$<DP4BKIT2L^0^@WOXBKa7w+S8v%PswGX<RpZ>n0mcc36Ey>
zE!Cvd)Zci^eTX=C8`uLW^3AJAS)zV4*>+(GU$f+WgWAfR^7>h+`f7*2DB8OJ(~&;!
zK=5E09|vi)+yC_|=tj$5&I&7!9OlF5?zdJXy0dmh($Eh@OFP*Q&u$;T$Jc`z#Q2L9
znvl7tOQVlRls-$fZ{zyUr_p^>Q-4IYf`sofEBY7=-^FbZlPm(@-@0F~KB}yOW4iug
zgzadEJk4GP9~clrm#F==l@TWNcN>fE7kQhDz_6&-Fm@L4J&3zhaFIT$<WQNL{9;lu
zw&BJ4zFQ_zHj3M8!&vRL9DN&qqqut|+y)EsS=^i+cq8Paxh>^+SN}<ph+gpa_Hj*9
zm~x6T<@0{y_L9Aq%ApeQ-u&?!X=Zif9>?T*x|q{a0_2^PH^$Pf_<*WHj6UdO8nRU<
z;(P9@6?QVm8!z(SdTHBz&%h3pm3$f^aD+!0p~7JCJBpr1YHNLtHbTmK+%2;R&jLr4
zm#c@6r5J%XZntN#tDEmo;7uVosnyX9FrD%TQBqeqjxWu${WR$bBhll%KTl`MB#$27
zuM?N%PoUUALDF;jLY>-lrqm%x7m*IUGc4$zCbh%15L(7gq85*&F2S&v&OfRjcx(U%
zKNjMsp{>WY9X^E31jK5E*+?`4$JwmVS%H{D<RJrcN7I>|F-TFY4f6v*1=&qI@qm>j
zdr~FBOH%~Ia4`r-X!)Nc=7+UR1e{E~Ce5SY(V$n}?B-G^YwJ14Tx5hzGr!`G*bdoN
zTqf~5gyZQVT2;91Yo0dCr)HNan#cW_u=DB8?wyrPe`Vd`wj1$*FsA?+uCK`Z{5?eW
zK>XIr#wNE>9uGtap{-qI-ZSBtkZV)1Vo(W{7};Q`@Tv0!O)B!~g;|^RCf7^BS*{i#
z&FlEZO?9842?H~m_=<yEkp*oq^2<a5NxY?qZEm#H>K@A>sR@chgp>KjGH4z~<pHm@
znVvQcoEKOvuSwL`F!#k4vfB_GvyMk(NVu5HS&D_Zc+ddHYXJuLJAe=jehOFp%4A?H
zo(Afv1yml!Y(NMbv^PDzre9)+CI++AkCfj`<T!xJ1fz_plVkOMZe3?-f44%VQy{@!
z$P7P=^Y1m;?ZedsIYfI(d*W*ghnSW|qfwBCuD>cbl5I+fsc-_!`gzVMok;?U=E><A
zH@slI_!Cir++=y9QYf~Wo_?zIqqb098l1sOZ=rsIB4$7itrjwXT!z>b5Ab5=WVfhN
zZ1>|Dmp$kZu0>f<QF`=<acZxt)4lVs5TS4C<MpKRHI~DEfkx!1{Bpibm@3<O=;&CH
zWSsYsBk#$Rf<X{Nhlw|>3z^7!H!+~zLg5cSSzpG42M*Baeir8t9&&Qg_l;ndZOS@C
z!9ui0h5zdDGNmjV<NfVMOCdyaar8FN|A1-RU;J?kMuc&6ysVF>>t@~QsDT-gCfmuk
z^)5}XKe^qr!s#6p@I2vkuVV#F`!n2U^l`TtW?S=O*m}5)uvglcY+hnXzUHxdcI-LE
z)KbCTzx70`Y0+$tu;~Br0^-#%=#53db7%!g=NJ@%@@Goa*O+;#&XWH|mI?3K5#i=J
zKim9hU!f7X!z(c%ov5<2J8QqG_PH%>jv(=6)pzRUNbx_I%rTiORRWb}IizV+70ugk
zZoK^{J%E)PzIeUYoBz<;e2$~lzChS~y>|x%TL}tfd#sd$oOELUvEf@{>s6_qTh{8=
zJeRaWpELKxX!b|YrTemr*W83i(imkILGI0Iw1(O)!=aWRh#?cA-hB*PZ7sJ5MC`w-
z)_eG)>E>o-v4cJzu^#^6AmE9ou+`TPYm{S(mnqx#S`H+z*``d--OyCdD`TVu(<1Wv
zwwc5sN5eoJ@Z8)Dzu66naM3)_Da6dFM(V2)>B_ObUBelr=NO!RT^IOhsXp)6j!G{c
zRZg(C=ZR>)Xk!tiU!`z-2E~FG>=(k|FV^_pov`ByViKpP_Pm(?MYh1WDbjO9xOm?r
zffebvf~jwEh(}~B>%TF248E*S?@)c)eAx&<BzXo7OC|cV3!ZEv>^8SGxW7<c_x)bu
z>AvZ7Ix4}b@GZ0sd3vb+=n_QH3TvdU$Y@#fZQx$w%pV7v1J=*>ZH0u_FUfu%c$GCM
zEEv!>o7^fbC>Z#9XuWlWSqZY-iewibh7MS1vd>Dca;G(Hw}>Xd%I?iXGIx_^+DS!3
zHIA=e7+27gmKuVCWWg+VK2kkGwO&ZU2}<ve?0_%Uos^7tWlwL?H`0n{EB<Q#H;xWd
z7nz{(wHybLu!=~6ddS=f8YLMH6Aij2C2ur71<IayiDSCf75-A1ZM~Uq`Fvzt5K;r$
z%dK^RZO4#%X{K=m#XXFd4N(?=kja2dthk@SS}Qi>uOmgxQVh6F&b50<AJ1-(!ti`d
z_IH}aiuktg)~vKW>@*2cV#ETUz&f9Kmc(P8%HJSk9#?CO;!v#SZF5NVYLnEo?@`5g
zAIrW~Cep(SWKd$jDjQ%e=#?a~r?r3|GL|(c>Qb>GAjdf=uD~!JB=LtSnJ_s;+5iU5
zBIMw$sqq$+ZJLr&4;Hw{K4tkT$q9%*7sz1S#9NCf(v_WwIoM<+@XSh^G@M1BX&`<^
zo{1Jbx68&Ki(Xl239^OMgkhs_F{vin=dDKx!?aeFK`OfWRi>&#az#q8OO#YIm^6h6
zY+2e;((c#glu4WbCE?3v?$p0Ooci)Fbqq|T{;p-=W=JGg-gOy3Zt62<%@uny0;w2i
zQbeGk$dfG+5sqp1g6r!QMx+18cWDrZZahO#bD~R0ABUvSjk4`Q5`nCsI4;S)*W%-N
ztL*Y8iaAN;Iw#9ee=VltPN8^Z1*Q*@%wdit-u@0I_m#u(qkAe*#0wZW<9IU8qQKo&
zek-mj_E%rbs4Z6hw6pIk4GER{F@cA9-zaIC)FB-SJM3mi67sdUZU$xr{fo1q-U>p!
zrZf*vo{D<vQ`YKq;HiG%^{`zck*uXP19VBZH7A72_kD#|d-Y*!>}NsR^ztK7I<tgf
zTV)jCN7cvo2S?b0+?Q`{uyzf_?ySDgb9r!DMsFof>+XdjKs=Z0JCTor&Cm<KiyvoW
zunL*`)OEwm1QxGxWC1VtXKF758`so+niAjS2m-9Xp)rOo;&ah1Da4w%NA%K$Gby+s
zs{l#w30(H~jw`xpAy<YqL|GCLa9XADZeksM^)M4dK<7Rruq_)u5j*ny0EiAH;;4&7
z?1vU5Q2kJlYm)aWAFp*zC-$D9J{LZsnwDL^{9jd_?eDFXHMxkB&TK8gP9;d5v|BH_
zUi!N~UQ+uwmQ5p)zz<<Zh58(In|(}L2e$l39uF=7D!%+{eh)(RXXm|_-gKDb6-Mr1
zyDZDcmwPM$8eS8h@&jb2?ziGiy+-M1PDDZea1jv3&?T{F<`44xK%8{)Q}n@F0?47M
z&Qo+GmwqH3IuVDIm&qnTlH{6Ip1Y+&+7xyPCKbACDy_j*z*Of0fZVm_Pv<36dd2MU
z7(%XR=61j6RuRcdab>GuA<s%3Qk`|a_u&l9t?8E=vr^aLg;@0A_sdI)_>6{)(Bah0
zmQC+?V`ilLp;(aARJZ<Ho9;`Tz`6NPdfy$r#p5V)i`&`sE^X&B#@<*gYQ%b%y;Q9T
z6oCHPWk{A}fgh#sRkV?s2)^5@ZLFZF^?;3&9v=Q_YeA7~N%D-&??#iwT5=~ukHoJW
z{Ir-ntH<Fe?ohdIim*0oWVII3N}~vf6A2!V1T|jMQ(!^TB!+*AS%S2kzgy$PB#O9$
zS>?Fs2!tyx-D7U4<Dq(k%8AT%g&%|1h^p|+N1Vz+O2(}t5>9?YdpE@48X`hx0M_em
z^aW00Q~=qmxaxe0RtNTAaw~X@WUkHbu)^UI+su%;vit0Y<%%4n#yufoaDm^foJu+e
ztXK2APJFuGmPjMq2Jad_0|BIJ7i|1niYINUbRp72P6))6x~dlwq>$sftKZSi=~=>)
z1(Q=jl#C&2qvES{1tqV&2C=PQ<1A?zpd()_jHUfy%|c941TpP*Be^M*aaO0bZF;w?
zOuQnHpFr_$`I34sz<Z1}8{0~ydGLZJJyvUz$apv!TTi@<pOxhhmR`jm`><@SlH^9l
zg=sBT<DtMs^<2MW>Z)5SOGFHxi+enV09_fXycCuSOI!zt=6irJ)Xy~g5MU>MZ!SAe
zTj?)a8B7dwWY1kfZ#oWOR_5=gqTI~BVs@4erF)w^rNl7D*?nqf(uZP-#F}p-j&<s7
zD%1^;ApoWo_L%!#8<(9wWmXg4EN}kE;b*Yw35DB?ymQBNE|1*Yw|sOG+RA!#=zA=e
z{16P4g2JYFKfl<(J5Ob!Z>&<}c~UGz*LD8B<B$rEp2H*?CTm4Q3H6m+I~zPRK_hUn
z^PCGri5JtbcRH8%*5NQkN+>?W<1~JX4HbC9OEKpzq!J<k56vJPKZ=K?h##xp)L}FQ
z2D-oZ?2rlN&1|Y&MqVqICD>a)+aBTfahAc$!~&(~Dg{ZC3A{h<^lQ#!EbE4=OH#GK
zR~*C=|Dd%?k$!!<_{je@r0Zq+;13c&RTkQ#ESwd4^yh;t4pEic?JQg~&gWA|LDejx
z#~U-I+D26t1n~|cx=exx@jy$#O1#R9*}IB#98vqRyyL!GFIN86_QK>GZ%u)Lr;#73
z8zwGt$Szpv6llW@+1+qo?0mMTxQp?(=G_LzF+`Nq4nAPv<`n<4Cp~)HfuS<%((^&q
zOC%vIOJP-~`pVlq@ST(c#;oXxG)1V;6HqAM|IvlwtKDy(+*=f$E*sj(iouuk_=wRw
z0tWrSPQh78*sB5)gdWH`!gyI*Jvp6d9;`FezZQzqSY^p&vsQthY;`SaW-Nv`U<_ml
zpzhars)ph3kL&o1z8c${;qHwF%1y<q3c__<%Q9`7FZECHTSbu(>+8ZGE6GrpEVHND
zb<0;{P$NV0BI4rRt{B(DvskjANEiO4Kg;pLa{%JHSUi=on06F%^r^R*zUO!`mvxt<
z^WMxVsOP^)D<)9K1c^Lxvpn-R=oy5hcql2ulonh^D9NLeg-BVNz^NF}2o9RcH_}-S
ziS+Us=}fj!a~mirJ_(yx>CVN>+@F_Ig}@@DA7o}a$;Ep&dM9x!pd9h|WP4gG+Bgov
zw^>N*&;*1=cA$lOr0_!y)v23a^{B`aAg*>IHY|nIgKDgK`?*N3^1*1&ZTz@9qHo9E
zHWC)4q$jD+aY-oxhE}iT@wB_asuvEv=W}K+7uZ@<b{vg0iZG<8Nwh|){A~_$nLl%<
zEC}^X<GzYtr5dBD|DzYD*HcC-QXDJHQAVXPFLf-HYh^x8?f2~-j&{R2?|Vh8*IoC>
z&BBF^-y=w-?`7km_i~@-<9Bw`Cf>&L0?!EF1%;c|^1A2d`rIyyHe3``5ax2b&099;
zar~efxhDA}dD&&~{<i{48|_-a2jrjlcnw^|SENFt_dTQWr%14$prpXU%(PStX4{?=
zqVeGzcu^WjDu5NZl2~B)Xph(rL_IxiIZ-&%p-B0l%v7ub`+A~aF<$3PXn3Y{G~-rY
za5Jovd^<YU1`UT0Gip0JN!FnYni?->uctA?Uo%4-Hu}73me#T!qu!*5qx&SU8;;-2
zdfMIDTPsZY$rn=<{MjPDkn{&Wgd|rvszt4`*z&@ez1CHrIUUmKfOxMSC71iwHILBA
z)yQXCGlW(fZ5sGj+mH=u^Ze?e0>d$&VvrZ`75O7`R+Ct#b$xKA20$C;p*U6|xL!%Z
zmbq@+8nfe7`U4mTRdgMEutN2~@h+}fgd{jP^J_(7?M!wv@9oT80cP+b-A$E#WX#HS
zc1bT79^2u0gWBTp1eHE&q9R`GSl+=L>v4f*oRHMcRWPU$ygh=p5FQ&dh#(IotCKN^
zVTUAYO#VqNRUAb%fE@kuh#r5D0zY!{nCv*P0)Il_UKoqdvLOU9pUs)?Y7#8-j42Ig
z?bjSFejQ_e#!yA*v`J;G+{X2$7nmkA0titH`di7_*X&=}P4sHdx_!7(L4S=`2v&HQ
z{$Tr(rztJ8R@cs22?&{rD+HV_7~=q!<p)LQG*d5I{?eY(KGvaQE~b+QtSBR|tcF5@
zbgs;$yA|OI?Lab{<57c@7}-GaybKSQWes#FtI0r<1efN>%(}vZOyKWbqf^=fW5+Yy
zziRTOU$g?*hW=<@F!~MB)Qe?Ftqy)SM~~Cz!q<<Xac4><sB5t$AxI+9nvE242O^%a
z@M0<Hi$wXqckzCyPNzW9GJJv(+VuOpHyt^zU4@WQR!qYg7D8E9gCp-VrAUEsY_<@m
z{k;0TUtw(vs3CL+h|7&P5%nv~M*GoEU`Vp@H0c@f=D9J~^Z9e*EhG2N82g(W1{XzP
zj3UYR_g|@#w4AaokQJw47C!vJxUF7bC?<8tqEp272>pyb$@MiqqB%R*u&=h-mf&=?
zA6=tYTrwDH-d&oJ;vV-8N+8oa3t%k9BpJ+(Bzs>XJUdk!E>{0aj^3>~zegdJTv@|s
zw2&PS;z6eIwnT~RaGv5!HTy~CI(HmrONVC{#F%LJ#TA(b_-C2`vKzRhzA}CkidZ=0
z(2k<&fi11b6v?_Q?W;+_752~E;Gu(8q`V-HnrH!hhZlTBbz<7*1Z=Dn6xB-MwDuPO
zo4R*p*|ajvAB`!Cv^UNILtWk{V0_w@qqQlW=UDN5x3luPc-E{=X3f|Zk#4NCs+q}F
zRSRtfCZu*pk`5rqi?bQ9@>REFW=k867;<O}hs1{{U7gH^AP(Gt@YextwLCsZ)o+K?
z7T2+V^f<YnG~6!sb8cs(v60+62$EAAU8>ke!D0G!uGXkwCXQDBu%juCKdxf;WXP0t
zBokBPE{mbH9Plf!m_h7j!i7U#H+CdBa6H~v`BzS!B{i>A$Au%WnNp4RT^CPGjWSvh
zBv=0G%+Y7&G%<#vjxmTDdJb!yDlsr-z=NmO>-4e7sKrAVQRN`!Xo?D{r$$oL{S2%p
zvE%hV=G(;>`%-Eb<!8vC<jytb1gp%Lj~dUIC0<-|?Kvb?(tjyar~QFtogN?B9K5a*
zleb*p9-i&oy%k(Grjvj1z==kbN^0pKZ9c`pz>+oBFQsp?s45z~IWQ3FHp+!ZN*Nsd
zVN;e}Y^Atycm5ETScl{?{RRc%a|tBUb@<VGHtFMrkiHjLk-3Zl6tl>hG1+R(keAyT
zju<k<s9^+eG#HV21}Ch_i?R-n0vgRY#YJdP_Qj83P2*;*N9}Bd6pV+b)Y8_UKIbqw
zSy~&Pw|{{9r1I|d=N5i(rmh%gH3roOedY>iKud@G;HE(~LhR?-sD!t9Lg>uot=YGP
zWFm0;)ea^K!_7Cpma#;djC==@A?D$xXIPHb=8ilywg!^;a~2f+s&s|gY5hta{|bs%
zJjO0zh7Z<}<VWDg6fl$4siy+O8EA$6<|Lg*a{y1eB~$P)Eat_<xR{jgn`;|+HDKpW
zxNXe3J05pB`>HK4#N>AcR<RV9Nr;E^PMrlrmS-*f7COLoj)`FiS8`Pq%qfp>w1@h~
zzDS~Uxxb6zPf0li3mW$dJK~Y}{uT$WLCkDo1X=<Q+?XUsGAS_6zZ>|f)tg+UeM^I@
z8NX8zNjL4)1GN9938Y3y8#G+tTA^mbP{T|}bXcWP`1(p2-~7YED&#k(8XHQ3p2)a>
z{i67_!(9YZ$u1X1U&6E04;vV@%c|HC#-8B6Y?iB;hf%47XDT+bb1__(6Xddt_-2^)
zy*`6RctwpVW4!Jw-UG-`?YoHt8(%ojaDLq|jYT(;87zYdtUzz2<O}gA(gNdtFwtQW
z&|S}7rJ&2=<3=wv3#a6H=mpSeYT#0`-j!iUuA5gE_+bn&)t!dhNUenXmz!pUB?sw#
zWTtoQhJoh4s=x(2mwgDmE;x2l5z^%ac)y(ls7)0X(R?OGH76d>^X^kDV-wnB93r;D
zFOLIgax(}?6^N{V@iq>?e6`QJh8rd!@PAtj;1WF1!Ko1M_q1x&-}?MZG^<KW_$@js
zz^$SV=^)3mS`$^EgO@R6Tw$-dag9=ZvK?D5tDb}9%Q7_j(81$Q9M&Ld`PrJjB@VCC
zWTf83ui}N)pRwV*BtXhFyhv9C{BV2cHN_zrZ8^f036sO!^FJpJZ~oEz+I_oJ02zau
zyz#saOP1Kiwg&xqry6<|R^E?}GRv06E9D<~Z>LsAr^pu;LzfZv;%&v`+}S999{a4!
zl^@2!P;k328YL!oGeDMn9N}v9F%#S|v~}e~pzPQ#nn4+_Si}y8&a=2v0;yq=fb6;j
z_8mk_=)DWIHuGTUY+fcqbFL^g0$GQ)UTZ7)CrM~15yShqj+Df8)uKDndrFK~fJykA
z;&E}!Yz+_Y7)9KI4UauSjCf#JCSyRkM8~x`k+n1f8qWH-W2bNx5yi)qorecrSnFJz
z4xUX`Vt-AoPUaPay~B0zf>@<j&+rP{A^lX30xvX8s0=p&B6&LhfW0mQADKoP354!a
zH~<j7lGLb$ZPhSN<$|nYd2S{_6g6cw<o{-Rz|FuwGP@G$7#$kiWKs!J0KO2jq%jk-
zK3>z8%wAWosHuoQM%ZS<STm`(<I>=k+@Y!GS!hX}p3aFsYZ=wKP{5m~ag7`c<fs65
zep9<o@)n>N@aZRWy_&OqpL08e7bWyO?!qQM)nv24Xk)C9v4b<CYE1@Fi*X=%MHMZq
zx}l58;Lzm)Bb$(7&we`Vt|e>8Qj4$w>m2}$DOB~2Jbwq@N=Jw%zO89l_ELc~0~#k7
zt({l8%Mjs`!{+82Kg2BUA-*h6t;}*RJ!&`T`NXZiSUz0{AO1$ZYAnma@xP|2nI<xe
z7k;+(8WwmLF~~Ot-<xoJZ?Xdw8c-7q<=VsU<mz)d(PzD>0l&%q1`Z2fWuBgnGZ<c7
zU#=bKlSf>~r$&P-BYvnc3-i95a_qIU8_nQBZ4+QpK}zii;$^r#5(^IlWxlqI$*}Wv
zyzh`~mO=su9G?>jSGuK=A3qmCVEBHW9_|{pC{=%Ym$!F(YBhp2#y!1<4x!hyOs*we
zgg1mFjP5`bj?AwPXvwlegn?eyXPCHKXYpB%ZulEs<A0Wgo=#YuBGP<TcO10bY_Grz
zaHXdc>1jgV+u0&tt(S%AXdZdWbDb0v(!bBPveE9+>sP|IlQP-cB5Z?S8y}&``NcD}
z{K6*lnV=56mt#vhCAjl#4<Sy4?|_jad37yr24!r!XTmzK`&i-Fv)I@Ae!}rb=j`=a
zA<z3L%P5RofV42{YDfCt3<Da4=$4XnZhhvK@weLwLuLkvvR~r2FWnCWAaObv4x`6e
zEi97jXM_(B7l@FqOzUGFiyam>Sz3_K(kE><o_#wTaZ`yMDw(y|<{LAcv3nQ_sL;uD
zQt)lw2_%IBJ(_2WW@S~{Z?(4|JPt>bjhaig80)y)A*YF+SA#>YD*f?Xr(#PuUg#_a
z5rr3CFEui;8$8LGvzJ6ED<Mz-4-vZ)4p1|&XsrJReWfzrQ(xw%cnrx7MshzyVLLFj
zHL9rOP{$!*7SCP?54fxc#|7c>CPF)e*Nee&7+;Mh%=E{Li4G}jB-+@%0$_wML&EJ@
z@XM&oV?}y7m?fMVxnr!>;=rNcizj#+px`Z_*3K$nh)~!=dL4+!+#1bNawfis7eT_v
zL562~GE6CvQ@G{Q&I;P2%C2IjyP~tdY4FLJM>ZY>%6c{QDLz~KSSwVX!eNb#41H5x
zDwN|NkAHkrr8);jm#}HRZdrt|odPjkibIUN4;aIq8u<CynAsV<!Q;>B{W7>KYGSu~
zHEqWr^>Rlb?{Ix}Q{irRi69LT7k=5MAbB{+M;q<dUhV&GC5fsVS))>td!bC`69*I7
z)mWvWQ|Js+-yR@6phtP8_WOe2jqQ9qM3IK1*^!|4f=mB?IRkEipLRdL8&qE^Nawkt
zYMxf5j(gpuv7$1Y@15RY0~C`(Pa%%qeKmWoDn#$|GJzfZ>6=v$N6)<-nCEbS5x~Sy
zzDRXi?`2OID4{Qc5iI;7#cWX))P&-DVml|j?2wZN$<Xi1GdU`|M~Z)HtCtciW;D?3
zh-md4{~2l-!nQ($R@i%$v?Uv^E5*B#myBW{4BRz~cV8&k$dDgzghX$JQTRF}`|Cqw
zGUv4)(~3RW-}?pVkKKu5@d|b+;aFn6spI&|l6ULQG+#=dZYsRlS6N>vjP{V0JIKa%
zJt&c5m?JV2-y#+QD@?q*!yGRrM<LT?C<1viZzT3`h}(+p=dt=|JN9;i423yu^Ik#<
z9KX)HD#rzaUwS3gN+;8p=35;%lUXA7OBjD88VL7R>A`ZCU!v5+4WFXYN1LXsj`AsN
z1os5KxLEFvR&=3uUEevL^k7=T4l@{CC0+I%h-jt8l$T4)g2x|W4Qm;=ySMJ8nM1*D
z%r&@z$_$^_H3+#!?tt7rVb?v!Gu1SCh%5m|2>sDUui{m-X$a6O`UkJgA|FP_19xNe
zmvliT1TE2d-iBUtSsTBEKQDX6Ob$#|`<@O&l%5h%gvXx&D*n`h&%3#oqYQ?>Nk~f6
zf08lJ-c)-!V;I!W-xsd|CHmd3<%YvPE?2Z7l2wG|2<Yh<R(HGJ&Mrb(0Nb@zo8X;*
zgO9gpkvXv+`p!!bKu`Css|nk^kdhjEYirrX^lq0}TAzrq7U#+D+8QQzppWMr5d;;%
zHifb6<!jJiEyJs%&&(T>H1kgN*Bz3NVt)aEw=rhwb4TQ`aglncM=ZPR$RN%y^|W0_
z;a6=<^No+*c;JNZQ8NeK+`$;vvOdCOtn+k%Xy?uP2d@7~RX06vcD{2%$z|uq;}j^i
z`xQa1Ke;sW)d}VmM)(%ZUZ1)9boSEK#d+wL+dOVF&OfH8;uZ7?o@BfBP09XgB}xc_
z-Fy5@AUe8Rqn#Z1i22MSb@aO-z;~j^ix;^ve=PcV=BCRt$BLBB3TJRwq7#6w@_i83
z+vLo-3gPeBN=61i19;ioHbVF0SLf;AsBbwtt`zw~lGb0Sx!xXaV94ibSMAVDY89*k
z$&`M~C127w9NMEsbzk*O_GE~`n#d7^&rd{|N|%`}+Efgqe_*y4E}UH3$D!+-xU;Jw
z)8%o#j2bG)1|%(wPpv@eJMTo_j>zYA6yMG>H1n^B;^|(N%{4cN{dKT@R4xCzm%#%m
z*~MTMh1Y?*o9~NC0-XtHdHZelEFnnms$P9rhVhw3mbRR(te<8WObRo2LySnucOb48
zig+ZdYF(~nced*Sl`OF)iy#OPNK<<(a=H*m$S&#6`w{A^w=tp+EN9Y=5tQD;IjVc&
zQKd1zC6!}l3h7(_9E~gcW@Z}lXTA1K>XlYtcVZOl-FTRf&GHXl_b5-H;>M*+C9Z5;
ztV<R!nr^DqRdcX;H?uI<)C@%D*$lA^!|Z9dldZBrLl<BoY2diTPHCe-rEv|*ay_g=
zo#MT#-9LhIZNhOo_(O87eb=m<pB0A~F&D3L=!#;oc!C-;#rZaw*3H{Uf}hvQVs)i?
zlxHp~T`e#gR8mRlu#35uOT!LB!ZE%;k*7(6t4%{P$Q`vShdcXQD}m=nlb>x07d2d3
zwL!xVmGwS{q1_(nUHx1W_N19<$5S~6w?UTkhJ0{#rvm}g1r}0Q+NM(d>6U4dD<%O3
zE9#hpn$PuM>s^2O`b>YUh?V?TK<zqK1hmy3MQTN&u`v+)p5{S)<cbk+44*Efz0un*
zJY{+}e2n7@lJWs04M0gN{@DLTeMKH0bLL^Un(|PPlv`<^29^TjrAq+sc#OWu@vW`D
z0A|SCUeqChb8q<hF0}LX`_C~NxD_crcG}It<U#21x?z&^#DzWfx&ly#vL$i(zX~yi
zZyRVeA%HQ<uFL_KFx?gHqjskgkB>b5JDz=?VX=V48!^6;#?D!-kBy7Bg7eV&F51M~
z-SAm`w4eWE<)=VX1)!`2#IMvm^?Kfw%TEI(g`G4PxkV(0Vm77KG+T-6yY)-PC#>C|
zormohTk#yC-4H8(JDpt@e-ppg)4OkaK1BJ+@0*!J@RccnV8IQQVE{J7f}U!T)~?6l
zs_93r>jEAK#oPk!W{@Sopqpi2;=7%hSwetlSG*ad{&mJHRR~1$JX+biG*fL!PI&q<
zLE}<y;!LepwWTB&9VI_Jf+<u5Aku%^i^8hO<C-HN5|EdQdOZO31!h+rzN;oV8}7S^
zHfc94ECCgxJRHxVGc+(IlGh;RNK?nV9O0tV=Yb*CP??m3h(lrn1OSN@AtPTFsqyHD
zWv-A{h`3pPUYgB816Z4B0%ZAxHGgu$b@`O^O7%v~nOjI@NGG2*IXWe~C6MPhCvgT6
z$(5S|`vZ65X+Yb8oMAZlsaj+Xg|V-|h#{1VNI8k0<j>TGp>mMAzQqXq{XC}a#hSJ~
zXRpiZ$k;!ACSn=|DhglHuPK<5CxXnS1sHtbvikDA0Q#kTmD0$#e_ZG9I%kLvnP#47
zSK}KYJSn~z8>OX{x;5@VVO*tu=#qFzZ9Jx%!={&T#nvx$8Q5l_Pgu<^&}fv3uiL=R
z6%bhycmu4WY-fdoY&SepR9ighOOu1Xk8VO>kL7Uxy#H258{H0c<y+0<JaZ<<2IlNa
zbC|XrpEam0q-ny9s&61<hryJZb>}-%y?K;0lbN))Zi>I_rkZ`tVyG)!o_&nIuDE6j
z0&)m8c>HiKR}xnkn2)6b<&;jD+Y2#>S?ur)$#GjM&<m)G9?|S6>R>Rl0eqg5+M2!w
z|Jmpk4dr0JVdqHpb#G#;-{;|Mqp>*oI+N0*hnW2YJ+ZNi!Rd`krWtguW;+~6WXxBF
zNjy=+mZd3=@5RQ>U1PpAyy>q7Z|6~ii~8Xa?xmOM;g21Tg_?%?)PL!&zs#jN(vzp-
z)P*%aqFP!hIvBo;DaPP28xADY1x2cl8!>#v>4LHK&W&<&*1>xFxdCN}8fW#CWv$Dr
z%84jXk5^z%S7YIS=`h$DBV4Kk*&M@C>h8k<k#ngeq57`mLg7I7`uDN9$Z<abNm`-f
zXHIAfvguDh&O{~uv)S1=oY%S9)DsDvpU=HYH*_LZ3Uz#`2yk=bC2LfJO>&c?<*V&!
zB^-O3^v3K6ERt1<cvG5<ogON4UB?#H@XVb(d`v@eZ^1(vhl}c!6_oB`oLDzn3;U(Z
z^zmtK3Z#6!TIb!kizT}TNz6<<c<oes>9h@0=BYWzKzJav4)|zduO=bkepf2qLq_di
zFJ?Re4!JP)pCVZ|(|*B!G=u-d7mA3-jKBjj*_P856_RZWFjduRf8ZtGxcsl$I_+rK
z+Lh32ij(I~+#Lo<TG0$&x1X!?MO#_zRepM10XDJwSS7p5>H4%V?=?UdulKp!-#B1|
zR5%QbqmhKpz`V|P%i5ETIct@gr>D99TeueySOsxa+3;l+l7hg_r>6uXdM~o*EB^17
zn&{O0Gnd+2$X5>p{7a|O5r>}@$XXm!E#QNTzTyQw6?<*7g{-7lG=`ad!qX-ShOWr#
zgJ{13Ma34D`UT4LC>cWC(Pi=xzH}IzKrqr65JH4`OuS4F_8t}JQJ}WiRzhcvoJEtU
zjlxbZF@s$Bwhs-i7dy}gi2JAf&yP3AuwBtJccqiNxNh9rehY3s4;Rd`<{2HVwH)1t
zR`Gv<hOaM0xX_qVt@^X}BUlSFD7ezQy}7WYu9SkM*_PSX_D`!8r|XSc`;Zg|Eo0?c
zzWS}Sx11N375d(HcRJ<?Z|&BxpPR@nPS%>9z^CU%{hU>=9P(rYNOp<%JOY_~E!dQ?
zpPZ}YpLI%|;>%BgqHUEk)oqn+p|{vU3-$Yc){coD4nX{Nn(>U;{x;Hvtj)f>D$E?s
zbA)&suCV!3(&~NJ25wYb<9mbmU#j|ihMcX=2%El24(0?2K@yFpy2W>%GWkRhf6eU|
zEUJhi&$Zw_%Z=n&`P5KQe^J|VEuZ_yHfQ;etUv-6nWB~Jn;n95hGI@=b@sV9<%jy~
zeH7_EuJvX4?eOBue%>ZZl6GJDh@-vzwDOtY6h1(QnyieL3m7VvN!wdsgVuS&+mZ;c
z;$Y3*z+|U+(6ri{|Jh+BOsywmb{M0H8Kl3Ui?Av`vSJ!&f?!|9ZOQif!o^tSc#@>Q
z{H;bqN007XCzC_0FP9F%!b!%H>Y>YhLDkpJTVbh8x-J$!_yt46*m}#I&8GAz|H(vp
zF~}pW-)P^NxxRwZzRi&h5m9*Dsk}YD<5L$Y^wUFTWLTLhY+5+SgO!`~6d;c>A{F_I
z7rVV?_~3I}8prI?<tZmQ-U0sx+dV}|z#%9oOIFWN$7V=T5$<k-3BA%fFB#Lu8&Hf~
zvUz&x(sPYmJNQfV4LjWTY>c2*b7^GO41PnjIIl7zOsOg18PXi;c7K#Ul0EFPGfXO6
z6u=Iy?`|3=svkhn%SCaJ0%w;jg7*kDtemN!V@R1Z-LXM*v`9S=8g9;FXIsh0YtxCj
zci@y;uVtk9%cdL=rnly*x{@(4aOD{_Yz>!L+~Ro@?agydQ!NSiUD0Tkn6gcFmvXZ!
z|186+Nh|lW&V&%(o5lg*qpA6KjKsOtxGP1|I$_HEjI{0s?rG|x(G^;-=|;iDx@O!^
zW3!LK3Ky-e*;=HW&x(gLQ@T-mGl#MS-u|r!sTiKFz{>9&1_Y8DBX6G~eeAifRm9j5
zOt%KYx#fP=FPv1smbAgJp$)>gSp`LOy(V5W9on>jZJ4Rs70VOa0V-seuud8q*?_%7
zgC|B4*Y`&)1(FUQTSf=DSnXuSMC>)Ts@c>CTcu<t?eUtwiy1I%T5tPigw=Vc0v_kE
z5=UQr4Z8GFmXPKq>8KP9JlK#%9H@Qz%q)*paf$L&q4_qmNgU2%?zb`9-bYkk$IZcq
z=SexAhkfQ-k7K8IVXU5ZP62cn$Fxal9T<op1pq)2QWQ%n1+I98YD1@$vty_Njq_cb
z_KUt8V6L$_1=I5}k-134Hq`%-)0Jbv<~m6}nb&lYH%2-)G=!5%YjB1?FS2s}yNSGz
z7QpODv|OL9o-s8XT#t;_6rzi>_0{CGoF)2o#L-lTvWU}*Wo{u0lP~=)E>LL{(>WhG
zZ9#&}{Bm}*1~|MicWhJhuKxu>qmJb(|B;XN;<iU)xue3|f^1Xd7oTY?vB2w-X2e7&
zTMT%KVqg7~{h>oI`=kQJ9`rp3!-c$+2B3(=Njxk#0YflLfU%VW+A@BfrFets(@ZL9
zvsC5`xF7wxrHWB%sE)sUIE`kz^D^)MW9uCQGh3Fn;YlVF+s?$cZ9AFpj%}M0+xEmZ
z?%1}CiEVSfoU`|S-t(O2`?Y?p)!ntas;;W8zG_-S2?|Qs4G*jc)_{*e+AMwr#JfjP
z7Wbz*Hk@u$Q@G8wkl=cOg^gU3`v6VeZwC<Z0K#!E*~u9*7D=rF4K!l+!7U{@4fJ02
zc9RtbO*<l5J^jJiLC-UrkF0D_A4cBhz50t<`FvhQ5n~@lIT_-SyiJ-Y<Fg2NRqP&e
ze+ud@PTX%@&A2T32if4^?+atlUv?77sBw*B-}=pK8`3lCLZ4FGz5#hwJ<Cl6_cY@w
zxgPT>6m2W&EmP9G7OL@*Ua>$`>?e*yr_15fX(K23ikzwFgaLTH(rXU2-Q4EHrJC;h
zWzyP`8Wd{=-p38;g4l~TrnVB~@_q4afjp;eZUf7Jw4dzns9z<7abR4omiQuT4A>*j
zOL_WZOe#uwLU9uq(b_(<L$r^>HaAA*M5!(MuJsO(W*U-rsKj+Y^<8k%cA^}LE3l0~
z^8Eb}8Ulr?WXQJlcfocTTw!eL#Eh(JLcsgxc32=BrCo3fkEO%vS%gU1V}*OG2iS5$
zbGK1aGFnGs^p`Dqw~n}YA17o&;AZw=jPF83FlrFWlb;`BaFkYX@};J)ilREUReJ~6
zdJ^4Q=9CBp4$0so_yh=v0?Ev?8{$x~QF+;;yYmo3jbx)CVag|&8dKSFnIA>>$;t_h
zpF?-)4p<)mTF~jjYCY7^T@nPZJ=${2!?yY5E?_6{LC+&!Na}axI3mYoPM8n-6x!Vt
zGG7tCVRik(1t6>=lTsJz6VD<7FKyO{OQ(-&t$}fE*-zF(OIHlGm1>+F(4QPAxi?#b
zFEPdim7+j04|UnB7V3_I-<oP_*c`}ZZwcE{XFHabw;x%KR~A*5pqx&l?mc<x7JnnM
zB6*Y$KiOwsxo6oDgq%RJ70D=*jB|`qHmK{S?v_v`7Tyjgu`{+crnHRdLTH;@9u%mp
zo=QUCzS6H1KmU6v70_q>J+gn#`YaF=4@V_@BOw86--j=bY0ktm_3p6kl+zo%1_ON%
z8r{`G-l)caQUY_JkS~FEySS}|CdM8&`Cu8&vxkI<jY-GuVQC?{+sj(YM9HX|IsOMu
z#)v+B;Uq;gi6@1cCm-*-_s^fh*e`ci60=85(s69+LRM2$4b!P_2A4rMI+6Sbwmt)H
z2H3N4GAsof9;?mheH7X+C?U4u7$Kz=K$eu5j2zzGm4)SOD_uzrOzS%D0j_<c--gWf
zVhFUsJZU(WwONma4EF1TUE>pPXXel4LYDd-_w&*1fg0Lhc6Zl|!D$kdVtdRpMyUJN
zvlDIE@{r4~)m+0U4vh6jcXa81s>|wcsx}WgUUbgx*uD`MY1Q#XI!mq*{62%nq34XU
zT7$!1q#{R?zQ|)E@tckR4iQ-`gS{YQM&A@!+@z(E?-kCe@FJ<pDxlq`jN89fU6Vw-
z)eYUK*)|^>jeArzMhmeNHGFGz_s7BPA!w*~(ol{iL&-Ef<StL|=Vju+C@L;u!5YH#
zlu6Fiv=+00v)Pbe+Ix%CHc9+;5=VMq{b0V5&Pv8kqskaqIyI_6gf+LC&>%;U4r_>l
zlNfaUtxnkC^gg%;sRfR63gy8&hIbxO+oulLv!biY4f?fjSHFr5W<gPuT;AEfZz_s-
zni))USaa&>k(2ne{Dmh?<HiaxK-lVV16;t0p0y1Y#tAWyKUl6UsTM?$+Se6g!od(a
z{2ASTu@3pjo}L;cA6R^Cejli-GYBgMP&~L$aR-FfIi(X!aeOR$SFUhx*Wf<WCgX+6
zsvelHDB~ZPOy#Ky1oZ0Iq;C+?w)QH!v-ATZOL>CJss`Rv;efbK!35@!YkGY=HYDNa
zy$%JoEjuM;U(n{1H$e}6NFYNipzTU2pe3qON1go<L1JPans5z5C`r%Z$MX{<RETX9
zz1Af4^IS82q?<Dih)!*b(sfJk80UA1I3MV7JPXOv*kbAwZN2~4gd#RMHpIr6N85+T
zE~JiRVP=P#Zf50jAL7SxGH)UEqRFnip)4wmEW!bD`1GRPF2YscOJyW=ebIYRNBsfc
zKLx(_U~{MxUgyWZ-#hmvLw3eXM{yKsS_@XmQ*Vc=bTPaZk2Q7~FIodRNWrGJNEa2H
zW?-92OElStI$v`vnym$O@-5tZ)X?W=d2eA+p;!vQ_+jYthggDz*DYFpybs2Ji0IV6
z&^kvQhs~uSsK)z6@AsOJop;0u-;s^~qMe1pU4ce=*nxKH*WUo_-Y~3>34rW^lluLf
z^$M^Bpu84ObR?%NSH3lZ1tDENq+D5LrqK-wdtlH0AW`3OmHvY)KIiyg&UhDTt7IZ6
z=t_n1X1Km{)2D4LoL%b~uaFd9kVQ<DFH7^1Z7RWwqSInXR*oIgLS8Te7-Y8Pj~qv(
z_tce^Kk3+x+F13I*=E$*EB)3zCmSpfz++qmQ@?jONfn%vlCgAPV!CJBY5++Npagq5
zQ1sU0n>R}4!%7{OG^F5i3c(RlrRu=HJQ4k#-00FG06)`r>{W?kCTiPoFvVSj;o(KS
zDG58mci*4=XYe|~XI4J|2j_0cb3n4{xo=>n!E~}^1wAC{d?GO{;<|waxprV<PfZWh
zH*&_%HXKFXo@Q24<I0<D%u~kj<4OzK@HuLi(vf;l);(^Ms$rn!VKXq!U1Unt!RER*
zz@xkEM%q4Fy(BZOt7Yhuv?m$gv&k`JP$mvTUDNBp^hvXomc~n>YZ0OrC+itF+VzZ)
zV|<saI;(VG%acea`)efc`TiwMBw<W}ackg7B12!}j=FQGq2I6_u-~ChChF6UgJ2fb
z#d{rd-i`tHfjK68jn~EWFRe(e8YV9Ci)I0G_9Be<%S4*b+5oDMU4m8OT7v3LXL24n
z5NbOlq&1MKSxjiq4pa}OI)+uoad-JR|3tYT`FR%dk=Joi$iE9%!jP?3;Jz;Hksau#
z$y^#KNO4t8_P0-LQsJ6<vsr&K6wEOm{$${Jeux<u;w;Y9?7WY1<p0i1>yCw!XbWAq
zO+zC;-uJjwXy#G~HgUFwW^tpR=T|FRw|!9)nDFO;71Z+H<De7VJBw`6*g6Zn-E{4*
zX!MsYqQX6S#~(f4TEj_siNs=Hg0FjtMHS=avJNEg_Y^#ykE@v8gY&gMM)K^;=hYy{
zvfsnspeUrBTaJ3@9BZTz6{^^R-#{VXiVrMSMD-bXUY?W_ZH-AGzm?FjE5l}cXe56J
z{wJf&-~2Yy)L`Jzu-n}MTL=hSp7I<o{R!C1sw?oW5{cPR8c<BkbPwt4m})GoldN%@
z<T`ibo5qqAY|ad@xQ}<^Beeg;_3$s+oI6}U!yNpF%N0D6d|K=No@5c`dWQ*@w9*m#
zSSqp4s8Gtv)IvEyUvGEPh2-M!f6?sxdr|`Oh^Yi?xbPPnbKT_~gR=sw*P=<k0+i3`
zIa8}6idf^lB1?$H$v;Ua{+oqIj}cPF3Wl{Mva&}+b6O%IrqagdR&hC1ty7=Q0WaBh
z1QwpO_R4<ZK-&qa7J&EPuXiIIF@k#S&f2)|a(;(le$a8@#x)H#AssI{ViIGgH4O|t
zdFE@KUE2;03GQ8#Me!paotmEhrK9}xbmwRH9bs+vjFbES5ChqPjUowbV`VI39+=*2
zSoY0?BS25Zu49`nnGR_+({aPHQ8U3&3$C$QLlXB1>o0W%=wqAQ@TO(0u>~P1h6_3D
zyN%DC|1Id#FMfM%)J$EA>md~Z>z?R;uR#^sPjE(5!Dc)%Ci=KdCyur?TstnRYb8Wj
z&empxJrJ2rR?P@yq^9Cc?@ZXxSwk*8GaV^Ng5{SPf=+L*@84fpqyja~a&qFh4J7W1
zuFUAoJx;qCPI&EowCP|!<90KvG+DXb|LcEEn?jDu$cG#YL$InS-{7JLnK@WcMZzPD
zy60*<Gg$eUgtF#ozOb%t-_5cE*!>sB5BXl`8cnDLy%b&K9%kUen(D1@1D!nj%D#Nw
z<BvZ*7J(Y=;x^)WJO1nYHG_zlcv$dV&*_IDj`G}W>vn7J`cr-*SDD3G_-wYA2>iG0
zromlL-G{y2F&CPKEe~gZZ|*x{(0oI~R=DVM8>^6v%rxkX{SoM?1}}AIE25U>3BKfT
zI%Z}y1S`MtrTC=0jXa5w8^z>~l#Yj``D~i^P&Cq`ZjzVT_KV`R?_2Pzl*~WJLF;HI
zqyFN9>RuK9?M&awkm);ky`90niS==iuM2zm=Zw8i*ZYX-oVdvf7XgF+KCuzO|8V+m
z)uu-}YG9PLVTSa=T{v1E=+H(joyeak=UkCO<jhJeud{SG?Ja_uX*FVt<+aQ2>(e)o
zY4~J(K|&h;XH;y#-@Wzni<vxk{rD<gZY`&|Lw)Zt=kUAv#M`X|8yktrbn~1ZhivE6
z@FVCa;x6Yfo)7XSPP?-qk)!T&d7KmtGPSaihgr(nO-#5vMmU)M52E51%g7w{qwWjh
zL624w+yz9*9~b|UeYb~zo<vg%o}Og}?n;_j+_pz+J=|r$y%CO<HQ5&Y5S#-+%jLuL
zFR-Z4US@JUM7)VuEn1*?pSqst@FZko=vdZ2yXgyP#Y_D;TpxOKwa%YiyYDIzn-85+
z!B{KU+6fuL6L!&e_i9?$ba4xWHYgj7k9y0U#)iLnj@3F{EuXFymMn5^%El%U=sTY%
zJ=RKsHhk@<-X^-dB$G0MALF;Z!PiI~ZrekTyvQFp2s)J7KNLEvJ!eGI$tH`Uqu@it
z_pN3j`2U>S&1oZGQ~PrLMd?+c2v+o@YHnmAD!)R4V%da8ccVnTwyhA~6it#~nPG&p
zj^Cna^L=%G5YC~RdpnZBX)jLvV`R9<+Jlyjy=S{;!p<rt1HCfreY;*cq<21(hSqvn
zCvt4-pEW~*HoQ)8oPu#;v=SaI0*f<|;U)R6YAT>*C+2%IpN0+!@?C><Jyu9(vEYKJ
zySyErPGh_T&kr-bWH|1ZiF<hw_z5wn&ftKXw-Rm&ZdzV)c{IF~qq77Zw_CltZI2=3
z54ey^f1a|h+F>T)w?4MBwh2DY%gaP~-^WpUZ9^UUsb=|oE&$tEIZx>qN^YyiL#Hks
zmzxgqec_6kbe{}H|28BQAYkz4Pyn%m-Y2mZYyjC3_<qxg2InC&)uA8wGd!ZH&uqr-
zQ#PVaYKz}90}X-F0>gBe0Sn=jb_`$MhppRK#$DR6XvIT~FWpL0EDa(f^(Td5taaEl
z;>Yga3T%S(bfvO13Z{M|3r04$`7}D#`{!w5DG)>3DOxUk0Lyf29iz)ry7@Jid=EM3
z)hX#{5`sVtneKpdKQ+t3)EKgGzh}tx^i@@{E>%zXc*l0_Aj*#xUd`vwWaC6^tM3E}
zn4CwVeH*~-gmz-N)g^n(KAL=vq2=b1iTXdw{0cVsWF?Q&_haKbo`MmZSPGHIWGfzE
z%ha;5Mwea;(gCY?&~8s3(6_zDUcU9CA%+hA%;hf&?$;R9P5a1$`w_DK#dYXC#(eNd
zs+r^v!+78@#^)J&<HCfq$oA=H#q1mK(Ti%`<FfB$JlTR7XK-k|Y;=s>rd@c8lfbD*
zgWSYtxyAk>cj&QX?Xxt5*RWr@j@MIQ3dbIOc=9WrYv85K@E50uf7_Q+Fu&v8hh~E&
zmG_Z0!u#xWX|jj#1~^lHc--{a^KW14Uiy%cxVw(0oe<nS2g-++{@Hh`l%TGBSL#po
zkKSZF`-z5qV|4Cie8=i?$w^u4?K`tqYJZ%P4kFPqA|^RGc}~}rdlf^hp#b|gps$-+
z8jD?yw;S8VzB^bR@J9wK7LTNy?ER4&%%V^)y^X*lHjZ9<^duN}hT<_`_{YPeui$xs
zAA*M=e8K{8FFNq3wD`~8wzonOY{wHCx;6R$gWV=10`9W0<?Zvzl4Sw1nvF~<#sE40
zi$%eQ(_fBT{S@+7ztsvyE;R#x`>vneZ5F}_RwjGb3;CT(^QX<Hi;NuZJX`T_o4C!v
z()l=ZyuZPn1oQqhctN`{JTz|)mbwbia|xaZfdm`x3T`i0<biY1{6uH#O`K=35vZY;
zt4E71ZsdLEkGy3Vgs51n;R-Vw{|Ic64%G0tg}|YtoG+-tLDWTMt^iC0?#StHH){#w
zcm8>^^pDp&ngj9rkEgjuFHPnlsj++hPI|A8HTIyK*0<x6K`5okQ>~9X<{u!dp1&2I
zmeo2jii%A8YD0iqMT5`TxBCuXRpaP*UJL9}DFlIAf6N2~nfex1LIN?$1ySm#`c*k6
z;alwg7RMH3jrtsQf_CIIb@1fSU%o@nnW*~#n7^07Q*p4YvXYE~iaY+E7i*`&mT~}F
z4Do{#=ZAdtWc1WY;>6H~+v__>AyfcUzqPGZLIki;7C#;8k@u!x7PrySl0$G71!vaW
zRMpV+f3(9I%%d1|<Qudl4%JlFUk*+{1?<B6N%CEfjbO}YmpX^o)Qili@NnJ`0f9d5
z@~!6E`FqZLQC;8U!U`sxoLjJ#`NZfXsRQ3IYPuo(R?(qvBE|c*Cx6rZ`Q2{Je0g(q
z*ZpRy$Q1bdX-Ze|wS8DGz=(QNRGyPF)4f-5TO}ojz~|I7hX6h$<Mbce#R__`QL-fi
zux5DAN2iaT^W|!@sZNw1ccEK{d&I(y!n+??zcBw+@lTkr?0OYj(R~)5vC<S=FUKWT
zJ>auKdKf%5$<Mn^s;1j@i=w+wZ2YRDQMsKVkii~Z4(b%B6@3*Q<1IRcXq}j`?J^3N
zPr>Cvof)+Iwc^T1dL6;W_q*45@Y)2mXB}^J#)-OyLBvM=UO+{leo}b7_|{WTR;MVP
z@YrAAw?YtXk6REY_yleXDm|n8^wD3_n>_T~f%!%w_&Qg=Wu5lLC3ZPt3ksX#vi<ge
z9oN|auxEmB2G=-at*<vUVtu^+_lcC=J#n{tPLzY}J=rrp0$}R_FLM?kG&Z?!7Y9n9
zHgLAdURusUyhEl2Bo!DI<*q+gZRsa`ZoK5EY0QnRo{2|S4QLwOhhlbMJ@FeUJxjOB
zzb^g<wg~I_Eb;0O9XTz#JR#-sapMH<^_4WPU_A?*lz63jlF0l!2O{}LdAF~02R&mM
z?iG)H^VjCm+wZoQ`3R(g(}|_Q0hwN}j_n2AzhU>M3XNx<V+@3J$py<(TnWxCm(0%P
zuGP>w+BY+Co2R93R}1s2pf7Ly1JnJ`Axn^1Sf)Qx6+=2JzFE4u{9gqBiyg9INxd~f
zIO!Kii(^KHkAZT^sl-avPfOV*;^x0q<v)?XULm5{yefK1xJVnNoyZt=c&92mh3BO|
zuAIZ)p6EY`JiC>M85p=r4}FuV@U@}R6UyZdVZsrSZmhYSazRx6iyrA!ursUw659Ww
z0L22mg~)~k$oZsux#-h5o)&&aB(*Oh;@OQe!ukh(mmz+;@$|v2u)o3ne<C8Mu<rIF
zTtE{+(UShuf&*#Ep{f6%@L|kkl<hL=gA5-c{Kj)5Lyf&@DuHZO!6}ox+`p07|C&Iw
z01D_=+^#e}L^j(s>}bC00+T;vDM8#vQVI*d*Vu3|n&fct;3v-D?ng$-dNL<J?S~X%
zDl~*sFfo@8#z`G(<7j~h6u0Z=!K$B1sDK9dG!@Rx2q|l;Y)2sEpQmWz{E&tt{H~N5
z%_KRi?V)~ai=Ek5U2eK*&&l^bBo1jgDW@!fgfX4#Q!Wr%&^u2o^y{*eECA&Q4m;nH
z1?&B1Stvj(&WoerUeq+}jadw)roY|L6RZ5%n!E2Q7-gSEVVB)EQWgFxuqqJ)M*Ze2
zeX6@#VVS{29g6zjld?g?jJ4q5rvWnD=u0xPz#7)~=^GB?RN6doR(}4F3O=w3Jj{56
z-d6LNF9ZvCX8GI8;;w5m4KcYL)z|D)JEieOg;Ftv0j8g%qWE(S^^U}LUWCargYDq~
z4*ACqShr_vX9mZ#HrXZ>xbC(ps`kyQEQ7X$^1EA9G#~Z13#*|sJuM8pNnX3whciPa
zG`k2PCed*NZ=NcXNS2DDHwbBi52+@kXO$|wNfJ>*$B-4oV8MN<<EDFXj8W@)C5ywA
z#Hl>x(cS4{zn0-?$m(H@PG@cqE>o9fx0GG)wSLE3>*!r&LyQ4tv+FOogB6h;Nvvx@
zQBy3f$n#n#c6NrMb44e^<$JMm@^E(!i;zGDrcP*6AAVJ7!a5T2s~OrDPw7;ABwcIM
z0X+u1;<KyN{FA9s@MHepZe=0*tO-jys`Cq<-1}rg8u!@)2ab2dIjh-r{GL$>#0wqS
z@lXu(=tyFu8_UZY;bgvkBzcZDvSP{PM9Uszh9?Md_Kg0iA0%xxMTu^MlW4x!s3+6z
z5tq{FCqW-;hyTyk79b!(l<fxM9Su1p!<ocK1eEU3bk4juIuq&bPtKjeOi1E(ZcvHL
zOl>Nr8Q>2%^_L5x7fIA-Z&|ZWFnr0!WOmeDW>z!9K=0)F!A?!TkK5`<n*%!P4(ptI
z)`Fb4d%*Tj5dG^CG1T;9ZNh)hDLE_;G|<x?eaq~<g$t?1;n4gc=zs)XmyCN5cw~Ax
zY&za|bdFUGu7P<Ah?&aMMBe?1ulK_Vz#(_88587qE_ST$M&myGgeiE@`bi|X4$uC*
zY)iqT?{U3iTvLV~mIvWqp>6qBzea^ah#@08Mk-$itH{g2iLrftXU^AFk*!z&-nTY*
z<rNERTm=Tp_Gv%Ta^cEEVsMKNY`%eZ6H`llBO`5i;TI_f3}SX;tgtbT2Xg$q616(i
zTb;@WB6K}d=jNq-rF$6#p!(#<%iV9LWQ(?>L7TZwIzSHLGuzqlfymv#DWME+@bdlS
z(NF0eg>RNcu3jktbHCOlh6SV{)cm7x#Pw*u*QB5FoR5TRVREpreZN;a=oSINUj2>(
zS^$N!iQ79~<!)`6+?v^Lt!@!meoYUbzN5hA7QRuo3ZN7Sw8b!0`4Gdnb@(={ps{jj
zuvbL0Z;v(ffiCYgN%7$7X{gQI?h!Pmlm(?J_pL-t9W%$d1<%IJ-;x#ua&Pcb;BZr#
znS(AMnYJ#Fa%#aOc)zWs08i`Fd0T$-ThKaGbW-?fk#MVT;tE;si82sUHv^SoY+?F2
z7wSg+>*p(G(TJgGP7Y9GNYvsz4LD7xIoXPY@hAFx%Jq@jlGLRQsn|)A@LA?ITOEkk
zUqdwZ%2=a6#E{H3TxLgkX~CajtfUSFf5|36sb(7V{WcbuCXeR4R{ld*5N(Iew~B4J
zX7F2^<*CtueN}&jh=V1&KzmMnl~i_c>F6`HP(cQ!>YSBwV&)uwZfoL{q_JrZq26fC
z)PSJm+xVUf55Pe|(p)o=n>Oo^Cl5L()U9v5v<}E^xRGxSD@f95X=nhjYPS5S>`E=S
zN-;Q9Xk+F&R|xPOvCGp&d<2ky4t$m%V*jAFJd-BEO*-d^U>R_tiPzo&*L9Ubght5n
zwpMD2A1;H);;Zkkf5oEQ+h?H|jy%^RiEBnOIk8`{rG32-2Z~<gX&t$6L2<Y864W4C
zs5ezGxP3os2o%jF=QD8$A(eA0$C1aH{*}5wohd(C$y>1qNsrfDFbd5)Q2!eEDE4u~
zvIrmcn$4mknUL1nJQd3QwY0VmN6l9ZNbV$NsQ)iHuz{(1W))`}E`y4D{}>QijZY&_
zU?a5w{xak~9bWVkz%e<8xd<%uW*7+O%C7W>=`vqY6<;$MMRJ;xAGhQ-v$?u%0zp-s
z+@pOOE$lJ8KmaRfa*>fR4PT@|{+Q)x)<6O?>k{<Ml{#}E^+*gTLzC^2oO@C7FJKke
z)@S%>G$9+IM0^R$gvrs5LBcS{MH>()gCqq&Hqpje&-c&Sm<o@tU?@C{3N3bFAgu8x
zSV^alf@Q$$9}l}>_E#d2f(d{DCvvo&M5-N-OJdNE*a7#2L5SfBp9TTvdX`FI_2>Fh
zwtunc6W^4!aa)#Idk)Ur#?EOfK<ZQ<2+7eY4OvyA^VK<<4KzrM+{9%xZ~`QahBJ|x
znek@@d<-60TJoQ=yWvI7A_wy!v(w&K8Lu;B_Big<TE$81+863S=Dubq5Wymx@ei>s
zaZt#s!Ev+=aW_OJCNUJc@np(H^E1BaAwLsCauQ0w^|>>zc6UNuSQZS9WzdF@*2aWf
zZs=i@2codC9%t-%q8_N$)6L_tyIAV5imN#f3Y{qj@w$c{45-GZvPJ(4bYeWOe*3Zk
zk7B^rsM)Df$(OM<HoH2~MHAVw;6MHVWE9=BZ(RkG?kJc_kp5xOXk-_uHTV{(c7}dS
zbL+eJ(UxeWO(`gBJ~a6SG(f12hS5Yln>k1YQ&q!E4->LqY(q3G%vAh($xo030v374
z+69)z$`S@5a{@s$c<m&Vobw2_o?~#mikj!xQd_nDm*IWr0s*)*7XuRenJoBY9hcrk
zybg09eIYBT?*q7>ztw$ksUstps&`gOt40Evfwni2M(zYwl1jF_#y}NbqGW6XT+JI3
zp+N!;7Z%a>afUjniGzs*YFaMivBI)S+@dFU8wWiKNd+uIw3Jt#1Y@qjxkSa)^oRSJ
zUg+bI<O(`>I`Seax!Ba+@R%N#=z1v@p_YuY*T0xRm6O@l=L@8wJDNP2`syNmsrf`7
z-Rw$%cZJYuYfn@~EtqP+%Yunnq_rsdVy~2E>Wf|=9r_-GJ2NNMEOq7dZ(E-O6LSWB
zaMe~>`W;gim7Kw~S!GVGaOGN>^9=@?8)lh*@mZg7dtv$4F3jJE+($E(V~b~oY;O+=
zUk*GHP04+T*&2_(_PkgvbIfDL%nZoyMkyIQEO&MA3L2XkiJ<;E?nc3<;A5meV-<AZ
z*cdjqPit4l702@p*EFuOeaGBDM}fOg#22}kU$!zMWeGP^g(m668-aX1pD0mEqLL3F
zRsyL2$MoYJz!QZw3{PDR!dO#%8Uv?>k;!LzBoNn&`d)K#+Mf9JJh2K<+!}lo5hJyx
z7e#(>+DFd}!i;#l{~S)uIt6Pjd`#+YK7VAfg^9}*zzIg7{lP)r4HQ1>P!2~f9WD*O
zY?&He8=X~U%~4U(9qf{7CYrD~HQ?TF8{b+F7G0FtgoU_`ux9{1g>8I5TKb6k3<6J>
z4>VY1y5<zC;Rt5KJ9Oy@!{%1(Y|OC)Y(z>qd+XLtcJ2lVKgmn_7vWLysAvG(cQG2u
zfOo_|j?4PQ;7TeJnddrJ+0^zD(D}RugMAA4sh{GeCcBFihG*#Fdfr4e=qK$-xZzN1
zOUvnO><Ma49igI4%LZly!}aNzB<@k0hJ}=ca`=nrH|R6$(@$gRV%-rP9@mo#yU@zK
z>zeru+fplf-i0=7fOE{D<^gKq_jLNtn|7IEMXzgOA;?Yws>NlPr%dJ3RTkJAH-wGO
zI+JH59<mcjU7RbR27^t5E1qM;3oprF7u%m$fjX)?z4Uf=vt)-zHYr;&k(rl5coE@<
zv(sj^pnU{zOmy?~3hQUnUosWnc{YNDt;yb3EVZ0m37*)ZV$d`hh6@~k=`Mm^yA37P
z%ux3#cpnG(7V>_6=5EU{dlExH835~Er^Ld5X-yeWv+z?Nq(iiwmo%eR7R9ifr!}im
zH1qmIp3zG<JDgrovkh8pZ}&2~n@UKJcHTXaps9Vvr3Se`44*<c^R|}Rzk#6jo7`Ii
zEoJ0Xg*TzsIy@=t`x27+J7mKVTMCU~KZbLqG<poP3)l;uU)rkHaxDIWdUQjwwQIy~
z<rsQ(Zg@%?M(b$m#A<ClTvSF>bL6j)&^fI=n~=a<;bVq@0Kq_O;+f0a1|hoWylJ76
zbH`F&FFwf5!Hl*2A4nEb6j1REMF4#T69fF%1Q%t89lyzLyX;tlR1WMhYMbj{ltuP!
z#5LH6>eSBJw5<(r^MLZT;I2j#t0304M*DePn*?TNL}!I-+SG%s>y;wyQ_MvuTt8!h
z@0a_@giK)uExUkhEq9t|fk#Ly9POQY5^b{5GElHJRpDp4{IkzC9Hn=IB=MN8dy4V9
zJ^ak&MX~)-5VZI}ren3aee=%{0j^CJ>Y;+h{uGCKcA8OL|AFn+@Fy|p7kO*NfpW4X
zD<JdwlaYx?=ds^TpP=a^$I8OW(`kjBhcIpcVbuXRL<bqWakonIbvM-qCb7A{YXz!i
z`)8u-1>2AH$6-BSc{pKS-!h~5yQrF9!Nl)Z0}~4mUZ0ym*@Pm2>j@?q?W*poVzn#f
z$MZb2xB2YY>s7rSm#7Si4r{ZYLw^O_>MWoQdoHvOd1;&hQjj8!K}knycr#B(&jLBr
z2I59`zDIFh(+u{BcQ1-VQ?N|>&E~R6=vY6F*tv|Rd%pXvr9AOXWlkS=3-v~i0J)!B
zf_>VAQ4y-DRE(v-u{^RWGjlSNdmzG5wvf|PUdck(M_m1$vc)YzBYI@mju>Nl6ExPb
zo|C6*)?r0k&-#KY-o_=|GLXpVHt%;sC^VTMm9Qnr&1cgb(zNl3t7AyFCAz3o($yq$
zc$}yEom*`_qfY;B&9d7Dh!OVo)GG-x-qeQN{gSaVh6FPE3_(28HQWTEOf+mRB1)Ft
zqoh@tyih#hjTkkrwC`!LVF=b^(9YSXPmR{GY{Ji@{SiPosOXQ=f1g(fA^k=M<NjJe
z-0P*@tdNXf1p~JDSwyQ+eJ)dhm^5?zODGW@9A-T81R5r#cbI$Yk|AMD!K7cF&uo~+
zO3W7QX0J`40+RKrf6FY=Jc3+pWl>b}6cs<nTnZI;7VaEZ?LZ1wIZ)Q!GSR%fAgSCf
zDZ3royIpFLi1XqM@(vBuvU1Y#10smH{8CBRxsJFJD0-zaleHq7C4W-aBsh(JRGc^R
zizsHNtqCqq%Y98j#N@n_m~K(NVbmBcT~&W$<1<t@QVI*^Ws0(m3jw5n`LR}m-Gi~3
zA1i`8{nm$dXZIXYhQL3!nWov*=q)Z(Ubvdh9q*5{i`X)3o5^SROdu{&gU)~b*x?%Y
z#NkP9m`o`wmerInQ+`>SQM(Qxw+;issZ47ou?z#@de2tqh*iv1gIxLl9b>5?gQQ)$
z{zs<d6c0qrVzCX(?E}y6KIoIu|FpcF7x+))_wVyvTF~nup;yYdibOi#u^LS>PKeFU
z8Y?f}!(<Wr7iY-Nj0D8b^))1QDzKAWGJyO|%TiB|pna*TeF~NA<kIJAlQI07@zkpX
zlLwUB!$#NJM1KtzuU)AeosJ<%CIuF-+<UaEH92>eIaP8tJZPA?Hy$h#@C`UJb;zki
z;ur*3sZdU|1FN!$5>p2#n53gb?=$ZZjCpLRq!*fMXKI{9Y#z$=w4lnq#3yioO8a$N
zr9un$Ie3a$fbabz`AWq8xtJOs#LzXrAlN!||HJ`iv*bj{?M9|kqMT8Xw(q!_(Gsna
zQwojSa|)uZba|5!7@9g*0nQd-_QDmhl@^67ym}OjRl`m(fHp*j9D8nC%CY9awRk#>
z7GTVRH=dt!boa-V5hm^^+V|fAP}qs^D4$IqKF54OqOgQm^r_+viSAw%?G8S?M)ADf
zKxA<Bd^9x$^9W1Zgl9*~qqlR3-lF`>nd$jK&i6J{J-~+6G4srNaT43GJ_CzCBru_N
z^(k29Xc#OI4n7sY+h3sA<(GW?@yKUd%#RMKhBb6&ZaLMn+|}{Xlo_<>4V95bY0=1W
zH&>I^3jd;=R8Cx-Svcf}bmQr0i{`@FATGPij5gkwQZS%6(c<^FkHl+sc2BxMR()j@
zjrs~q%iDUI)wUp7zC*a~Jn<L!9Pmv>Z}3*HV&RV-%b%=TPy(qvKC5JUX%u6o++9pB
zg${roC)m*t)nEv5r-fpb2Zg``<BWB!4l?KjFIvYiSfXIF73F*lK`ScbdbMNA0r7PW
z3saS<k$(L|ItzsBzC<u-p5!NXIRImsi;w7PP^T-7IPfmWiFMejr+W4jbVt`IQ?a+K
z;VY86H1#3YOX;(1n!=;{0;FA0;X-nii%&V%T=r9^@Xw3JIaLnnJ2U;_s1}TzGB7xJ
zgi}4U!k0^fJg$y%WCbP(Mpw_K9m9}$O%)3maR^}+$2%Vb@cnO*93mlEZsXk%?9S3s
zQsA|YWfkKv9q?tSdhe2A*+SK$GT3sR)+J(`NcKGxF@h^X#t=-#DAcd&D~?*Y4QPy^
za~2K8EC*_#t#^%>_<2mg*>2M4ftq&1R!L>3wQR1qyn}ejywLrEHHp(+x90BDTOm;y
zgU;mSGfKCC03ti;V4!Osx=#U)J6nCC+t9xYttHhmH|4fz2+p$^FNu6_sRvPt{TOPy
zp3OO%Fq5!l8EGyfSd$N!_zp~`n9}nAz=pt>Yl<5{nd@w+*xnM!^7?Aw(2ZJ<OX=(;
zOs8?4$+;l?6MgcK(4102uvIgSV}z{2z!i+FtLQ^{bj5}`<pcGOs+X=a)+5*>9;mzY
z7MD4*Ba=5x*7dT)Zd^beVi+<>2SjhHj*ON$U>}R80?sB6g8tM?8Jz7)5&2*Qyt|$H
z4^&LJYcjF3+Z)jV5DhpAi~6q)%*$RyETns*1&A?B@*Mz#A!bnE9j0FciFC_gxCo^n
zjyOsj8;Pt=GP<WwH^@|;72CJ7-#>>3dAMWg+g$mJVl1SxvQvVAbjm0$p&1Piy%i$a
zDMXWnzo9yZPnE7OM4xHNV|>cmU}U8gLCQ3n2B5MqAIo?k>!+QDu4xx_WNfd!3Ey@l
z>w*F=QHw$|qJzStvr=gS_|+fEk3~5jEjN>?wQ-DqvI{r}KcMOamHA=Bka(mW(>$rW
zh^&hrRo!Y{n8XL7p=KZ3JIMh{JDcu4bc#O}j1&Uz9Pik^R{P}b-q=VBflZ$ELp<^X
zixA5pD<+$+H$TMP4U!c_olm1~?b;bX5P4sB(8zED>#mG#&fO|hyA1fwH?mc3(B{%N
zve7%AHd*S498e;ff=(Cac7UeborNsCQo59v6pS0s%3H>)F<}Ud8$k=-@qD?GB-=P9
zrn2#f1Af6|wr#Ib5+u#4ey@8n+&obIzOa<1`#BP9xfe`Y1;y^Z%?7ca@mf<wV2Spr
zF;8nu2dBtqzI0P&rG=LQCIG<LG4N>6%z;<Cab%WXpy_Ho>ZYPUOV=56Q-N$No{kDO
z+nq`+Cv8Q4!^rQ6{*Li#<AtFh0;h;WfNR<4CGG9a24n}5(B*ng2`W87zl}cM2+m2m
z{4-A~&t_A)VA87~o0vTxyVV-$6*3!Vi>qek`}ENk&j;ue1wchAE|3Q$A&iv=YvSU4
z(`s5EbfLek#xCC<0aK)QAgqCR&Wk*S{n6@VG|6?EX9G0^vXWn7ZnO|06-tzt?!Hd)
zr5q~o>yNu!hVNtU=0Sy$Mx>P2-B`HD-~(^i=uMa<*M;_2l`{pCl>E{iElob?eK>{J
zz{j#FB7W&xp=XsSSIQ#_;n5=KbmM0hnlla3C^4K{W_B3cOR@ZmAQ3v!P5Za39Y)Pr
z%wnK2$CzcC>3G}lRxjFa*xWDu0w?Q6j><ZvgKrE`X*|1Ej9EcZ5*Za)X5r6aSo20f
z?&?kBP5EKRwe*2(lge>hN*y>w_Upzr#@8rCRJ7X>Ybs-Ibc>bBob<_ml!A1lolrth
z)z`SlHcHaf@sU$D-_qN=t>v)Y$$0H7F$mTUt&N-6ONU$fn}<H@YB}$FtvdFMDb=B~
z2ixZ6cACiVR8~Rrfu_Z7t@oGx7=Oyhl83gpwFJL=U3U!E|3{y!ucV)t_D>($vM@1m
zdP=L7k6ydN&zdm_06F4WnbmE~F6U<%*&VAk6td?42w=Sb3>8l5%i3Ut((0Vq*@!zX
z_zNr=CIwUIPjnOnf{JYhJ`Z^D(0XfA<Wv|Ct7mTb47-GU=F(|YV0{lv{x*yYHZ?X7
zET%mna!ZmQ7uzJWAA}u_T3-7I2_i~*&d=6>gF4ZOP0o2ow|YUDDol7TIT%H$FuB?J
z+em7u<v9865sdVLZ~_$utr9Hgdq&@l2=P++fRlw+3Kc=<qeX9NsIbzX1pfoqTnA0j
zA7#_1bEcQltY4TvKF;t!BPb6y3-dhX#^P*tBs}{rP@2IA%Y6QrQm!>)y1IBT7aSrR
z0g=ZHaLo3NjZJs^rN%Rdktn|F8QOTFx7Vw&L^j(P;gcpC3FcEdLw$v@2oLVKZGc9E
z2$(dhu>ZZW9=Z70X7CV?dt1`!Et^sdox@wm@;JN$G9!q7Ikyi%R2)B@Vbf4KnSQQ4
zEo^p$f)E+~1tVnQ?1bLk;#VAp@t|rHg`8G?X4;S3u}x-a4GX3zG+^SkZgPo>Q`AvP
zkPC1jR3KN=IV&hB`AwCnZT9#CTLPh`4AJhHioVLm{y4dlXYr&;8c&x+vn0EOj*@#n
zcUkS$#)b#bm}We3gsP^Ix@=v$ZKz{&i2ybY%0($q&;T8^FQMiyBQa6CW?Ft#x!$(F
z71o&%z7Ur33|Nf<@)}l}$W0xQ5Hd{{9+>YZ5TN3ED@sd|&uy(k7MJtvRLh%TAwlEm
zEBYY2@<4{v1&V&`oyFpOLQ9EqF#?cW6_r`TDL$!ma$`8c-#02cslSE(Xe`jMM~~s@
zs^GTf0w#HA*s2DCFTpX2wfVpl0-u&9oo-_K$kw*|3wg&lBwDl~F86iHM;mDf4pgo5
z6un06`rzJ&dy|^RH3Vvh_aSKRN>KP$9yr}cQok57erqPJ0;^5CcPI;7^i?AVB?Gt8
zHXJjq3yNwVKFLLGuN0}8T&I;UgX`N5`)txB3yHcWV;wOEo4SsgWB)Nu?*fo&1Ad)2
zQUV{-hMy?4!}wf@x&>IBXb2n9grudOssCy$X-<Bzpmv46w*hCeA#GOWY@{bOKg`;N
zhKpR!dOEg-ea*o4o2n1<OU_@%u^8S{*NsbOrRdMOuuswmD5M{_%`(anIjMVi;=7cl
zsk4{hpOdGKQ03^qsKxsja9v+b`<HLfUS{h0mv1oES8D?><~&T+(;4%SyKbeVwQ`(}
zhn)58eQ8hQXm7b@d0d`2JAaEbsg;lD%RYZXGY`K!??r#F&`I^gqF<lS3Qt~a0r#+9
zT2;FXDEV!&J(_K7pE9H3Y&k^#5OIF)W_da*>hoJA$+gW5<3tVJH7(<soqvp6qp=CS
zO_oB?_T2XY71*9kikl>Fpgp@qb>97};p<A%390PFC<CyLdQk2;6qW4TO5qv0yNY|v
z+-HKb_g@LldPE>c_FwJ&bZ(eBrF-VwTs_ctqqht_<6@py>W{jXA9b#yo-4-B?S6>i
zw3C(ev)6AU7gOh&{Q)FMBb_*);6f{Ce}K<sT!fCe2_~dIGia0t!5tSO9uO37Bemec
zE4y>AZ-x*^xV`@0BilPxP}eh$_Ubt`Qv2d8SDzv!xUqA)`UD8!SaLGe#s`8_*v~#r
zaC1!hR?vvy2<#*JXJH%j4I)r=lnh?0sXcQ!OM^=t{~TTaN2RDb52z#1m$MCYtvM<j
zG*LT_Rw`QCPrH2N@Vn8y_(OkSDW?Y1PJXF@gD-#*x~FP*QRV&m+MZJlQMZZ%ktH6S
zJz_KGKjW#tYiI5tK)-shGOr~;Fnudg9{mA@jgLL~AdI4?&Uez{K37zu-xn;my|MKf
zKmFIN{LqB;oNh&y>`z^qzd|6efZ^iFCdZ12D?~pLF^Pw7{5`;bKBjdc=S>sF^2;`1
z#ZOeAb^;Z=_HhB*$jpd?V&Z+drNYN;{Qot|zeJ$!0oJ?`1p{y4@L3q^=Z2q|2Gz*v
zl-N^qx-H(4{ln%y69IvM0~1^{B&~d@_p1NT)TaBFWtGfFpC$`6&HK%^*gz5ip*yN@
z^CB#}2+y8=$;|d;d!1}unVqNSfj!RJU1Xf0Z#WpRY<k^yn@HaY=M^rsHIvTIy}LLc
z+gvYJJRdjPM58VPIu0{AnYYH@rdAf-F4c;Zw01SM?R#tH+)DU3mkD@Zo69S0!z2_+
z)xFY5vv7s16vCSOO7BioE>3Wi;2O1JfIbJ<j{%<94%`p;(?s0Zpus;0^&m0-<1tMN
z#!~A?MM+ER+7Sq=xT&Ca`VxbCVQan8z7mDcELQ$Z3?{Y49-(NeOEhzDI@Pwk^~fl=
zA9iP-_!~Mzr9?v=Rj6G-5&{aDEDC$a=5-8bBjw|aiQP+qe`}DhmB0_$oepC%<AN@m
zf9}3YL0bFq(GVLBM7SG~0Sxx<hhIJ%h}OoBvd@zJd4e$ywP(Y6b>E#wrOMfKk>p=Y
zjChR;i!a>Gy{xRPZG%N(Y75*q$$i*~(kAlw=OfectblbqgpOtF<Xp-8kXG|uqU01y
z*qHYAyHK0u6iYAd7T@bPtc#+ELr=6m@O$SjkG-U-lKQ6q5=FOS)X&yF#$mQ?^5#B5
z6@zYbvl2(r2}S$>CPu`g!4j2tN;)g?cf0y|>pBf8PgwV4VK$BmUWUTDS!7oUNJT@#
z1j|#}wz^T@{rwKJoD?T^tX@zi0|j5#P0G1`Dzx*t4wo@H*I#7e$_8iTkp$aA?|Z#y
zLYqV|syYEtD%i|czcM)O6Fr?*xYWEEh=OXV#LpZaKZL6a8&!~1P@QplcnlJ1qyJlU
zxR52>0Iyx;Q9wVd+@+cXGCMDViV4-6w?C0`%sicRYd;zkh=o&S-EW(CfdaFxU!h)s
zQjS^FSgS3UKikezVw^m-JFPZUho(vh-#6}{h`oVbzAtc(izgE!gzdUCBx>2b;-OmG
zyFW<JFkUwu<sD2X)6W)d*SzQBeyJr_IhV6v-jSHOI(xA3X9eK>_w;tZfizw)<V>dv
zx5y{Q)_|c&+60)4m_?0gRbq<Npa$Ld-JTmj34wk#@j4?SfB^#$!VZKjWEyzR^XCtS
zJ3HLibGhKAFgn_nOUMm|r{Nh}vUN<!;amfRha1DIbgTiYQkC+*{7}r!WEvq^?^*TW
zzjTyK-#M5{{e8GpEnkQ;UmXLa!Yh-0#?A8_{z)~Ic`~5X<UQqsu}Q89F_t7{?VVrE
zuDDYKes_Kzwh(;o8K~FvD0im3G8ToozIBg;7im5=>CLoBv9a~DY%{(lk7A~3Lgj%f
zp#nXQQ{i=IxmrF5hsw|px2+tZRoma)QJSiQ^PKAI(qL;ancC9y(O?Yb8TkF>?gK%Y
zV`Md{S^kST1d?&)c-pn|olHa4|B=`v(HB3zt5{T!?Q|J?4_T~iOUR5up`VB+P9&0I
z3YwZ!$NS-=crGH)W{FN=eKND6SZp=6z}Q@wJ~zNltj$N4tK00XT4;Bd&j+|<MoMP7
zZJoz0-Q(z*<pu;j?*s}oH2Y<R%b%M$i1n^@uSy&;Nxbiz!-re0xmUw%kE65B7a_iG
zgDFw?T+`lfW%cWOGksbKu?cmv2FKL{@5hCt7d;rskSk_!1>cz07Y?&C=l>L^Y`O%z
zJlLB~W@nNxQnOh(ZoS+aN&<Ob+7I58*9(g$?)D%y@`FE%V)n!z`A#eKeC>BZA>w9-
zuAVH%<_!E_9;)mv0Sr@F%%<VwPR(FKgH4wkmp;!|V>ZXDl|Jx?nJfz%PBUI39~cw2
z$Ly*vDF^=U<ZjfO<u$#m2G%c&)%eb=AS1n_F8dmefo`nLT!g;Oa6*9}ubqeMc`6z0
zHn0EB+Wcj7T2cI_v*45xhy-TbPkaE}t{fGjrLU!mQJCB#x{0s^d~b__ePV8)futp3
zvlET_M3Rulbt8yC+a-847PmZf@|Y!U26p%CDX8Q?X?gycELxbGfx4enP&Bl7jr`J~
zV9CgP1EiL-G*XetZdhQa{mRwVcET@eE!V^$kk1t8r-M3=Q91m3i`HwTs_{BDtIONj
zBC*{d<wK)U-0OO!)i1s;-M1GXhv=Ru+%$A5Ob5PC%V;*4>N3d{D`#PIqi}0@i}@(z
z(daiyhoQ>;X4WezS>JgFY|d%~t6mf;+b&nw2mI%k04c%HSUl=$zem7#dA%1&9^=L)
z?JNs;1J5L}JcA0ZGnkDdf9ve%06I6PFx2>Wl2Q@8OS!*Koh3hN|DH_e9Y`qgeO?K;
zc$DuTyof5a*}S_p*hoJb(xO$xUw>P)-Z-B`ulqrCt&$%MdzJi)pXavdlzx${LhJVz
zBmm*9uv|LVqW8;*<3<zj!UchCmwb)(@tAxz+aw@UC*D-w;D0$IJxC{VP;1YnHbP?Y
zSHvtfjYEvFR`nRG<2ld~Np;UFgM6CAn#kJ@?!3lxa@*&M>SPl-b8~AXM`aHKzu;i-
z+Oh^$dKm8Gad$NuVHmlr3rYWnHfgP{Vc23&`xfmA)UW-}%#J$}gu($E6(jk${ridL
zhP*xQXA8!f8*M)qAQ=@`$ZTv7!O@hIg>}A%sJl*g$!J|Qoo`+a!V~P#aT_KRQ+^ep
zv4XJgi4Q!*?LQnNXc;fh<Z?-kWS^B&f(V6$qO6$`7}asJY{a3sa75^6?pNy_1(O^l
znJeTK`R1;%6Ktm}jPGq+S}@Po)7GW!yUp`RVBp;gXS)k;?<(C+QD`vna{WH#UXEff
zru*ElvT497Iyj{MJ%wDtgl`anu+L(F<Px&qv_U-p-~HLTm5K3x{OI318xLgS=3v)1
z@XE}E<6OcpIDJ5rj!qv}!A-utuh1CT4D}n0TOUCMXl1;8ZZQ~>2(ApFe{T3UkDHK$
z)SBr(J(oESo(RJiwM@ks)t{Wp+wT~j(h~$EwL*kUtZ{i6n0PzRp_P;vv%Q@DFgo;6
z-=GKUoLw8q1?CBTB!Vyr(01%%DGh@;*|LM|&f{e4wo>5CgoV&1Q~`z2%_HVQkUQ=_
zQXfCk&AgOF*$tz|COaTVQS?MUzQ<e05|fol7{nst5zyJ`q5yj$2CNa(S}s2>B!L!l
zlg(0XZ2!I1Zy@U{y-x%?Kl4t_w8y;v5QBvo#sxm&BE!d8U?hbE5OU+v*>55@m~L%e
z>72DsoR?QcCpk~;SMKOEs;<{Vqpfeu<5lE0(fsameN?{i^*M7OnC7=$hkOKXeN@J_
z{w#GW5V#QhWmhTe5f-shnbWzNKwrgb%)TF+=cZ&$BS58J2O?1f?MO(xAtA~2W#*}0
z#k38Fo@;@ZEIAWG$8D^w$b3UaoaP6W^L^<DML9~He=G&aBZt1xhbI3?&>iw?VE+)t
zx75a#L-gvK<B?joZMvtKu_-5ZsW+r#`t_sM_;{@$eewSVm_USo-9n0-u83W$W5*Ul
z0Mfa&4!I}0NEro;!cJkIN4ipeOmxOmM#GvbS}E6(h;eKHLZN>PN*c$e|7kxR*~1qd
z!Etb?p{r{RG=9z$p6`4};BOt9N~dKMQ|`dQWt8aG)Q@{^>8EujDco@=Mz*7-y%9gC
z#`-gBvuE#jy0t{2dhB5TY*<ZZ5Il@3UKi*WFL>GI^T!Eavm*Sz0j4sf2zc=5&Rc0z
zH5QrBQEBcy%QL6_EJY5|`8^s=ymGh4^|9ro;}(YHG;UZ1q30S_eCW@9P=rIQDgTAg
zD5=_6L}0xxGM?4?Ox!-rink8Ay7}^-qYHA~6!k*U_;EddV1odgM7muu7($SF)vgp)
zh7^J1%P;>%s0~OPPq-+>R_l(O%QzZGa>@8$&m8SNTIFI1Aut-1KGKX$b7u0^oR4dF
z>7*?BRMn~fK&2qU-)+HI4nn~7BN6tu`2;6qk2mG6U-SVR<ii<Ay_mnO4&7teRFQ5R
zPtPCQ0G-cc_>CX!kC_SWg+gsZ;X<&}yYXNASSGuWnQATO;C)}oR<xYO40cs22`=1G
zF7Er@3*-H92p<mSu~6PxeD12tdY%sR|KS2~a+Jp`PT|X@a*x{=p=HAzX%}lp4nxGo
z&e({EN8KhDqCrH)DjATv%~8g0ouk!h<UEdNGrv=?@$?g6WDAV6JBte}IKDnc@UmYO
zAJv#@i3xpOqm5_sDlv*C=KC~wr|Z6?v%$jI+?M)Mr_*%*VDWz-IVNcTLrLK4#}mi*
zFof`E<VR)nO!A#T78i-#fFB6g0&fLgrFU}dcylB22jeEsE-G}_GzF=ugjvi`Xm_R$
zdyF>eB=kQ;YJLt$U9%Ik_iyR8Ij4}rxk1zEc5Re@PY&sDt?54U_Z_RED91*tLR?}Q
zhNmowGu`wWGL7BE1vZuk+RYG{xBa=8qLVf;CH65ZQZM*{Deg;XXpcqZ=vN>pnTq3y
zt?jru@`0PjD)~Spw5i_GLr&cdD(Yc{7EkzNWK#fg1SK2LFq)I&vm>G#Bda_Ek=7S<
z`4|rg+6(HQ&Zi$JA0zuU6gi`J@Ebe31eVbMN7r9RHQmPl<G3OqAtA^B=|(`1mhKqc
z9a19FIU11`q(cM-NDGdX8r|K}-937+5#QbK^ZuOQ?fpK#KX=Zy*Unzo>v}$)kGh(O
z&n;=_d-{cdl<|W4P9Cc3A8r3zp(vHHUJR%67CE*Rvv#{FJSksp>?cugt}smzX+f0K
z@K^>Md!K|kRU;%7A!*lSeV=g*ho+>Z4J)mAtX%KQ;z>;Yh`aRL_2*-VD=&(vi)aD4
z2tjZ6OWxVt+{kcUH#_%fL~LPMhDv^4Hfx?+^4ejze`y(vVb>DRU};H9@_GXN)hLMh
zoQYzg2qNB_<#$$7Bd$n!q6>VasXepXa$ZMs(RiJ6>wLdaCmqhV$g<)Wo59hAMigPT
zE%W^lN^Q={SU4VV=RN}+uL}P7;c*Asx&8{`_lhlcqDPdqBlLD?PCC}t2bU;d^$K39
zV1AlP3z1+v)i@jv4J4Ux4!SEMKm??<+)0S<n%>td+@!Ufi<B52in|XtIDKF3{RYgD
z402$McItyf{dXtG!m5Gz4ya~}m@h7WPt3><&nOV+TP|#7%i}sKAvXF-reM%>vtI|e
zSrcpZ&Xw?@z}|N{+fZPkSr$|0a-+;^BAnHqU>f)Y5dK_nJc+p;T(R04Rwm6s9iJ*w
z)kd(?OopqM-(neNQ9Uk0WUrk~%ULrjpf@Z}O7v!*WiTvRX8^N=QT7Aiv%)j=BtrZw
z^(HFE=Z-X{a&!BUPv}kIkAd}LjIx+NiYE3>`<vJqyS;R6!qn3OCAFL*Hni`&`2Jgc
zo&<xLrCyyYu_!>TR;1t|(CkBjL3MvJ!sIC)@j$_I&uv7*rKI;&lpIUO8vw@}g8&{N
zq_PJXz5l6rYD(~$<~45I+<chlTb5unbh#JZ&Nl~Yaq8Qusc!>S{`)}YJr91*;)lLP
zel*EMG^M~nYK|B&RbaprVE#{OXe0C{rHki!jF!mBBwyhj4;nYN(dPfpPmi@kH^Mog
zIUvFo{=!|6Oa3oJ;D7LeHYiqbqF4IkD)nVW4KIho*#F$icUY7%SUQ5+X$`ZAW84=S
zj>Z2Or$5ofQw2ZXYOY#&c6{B1WsNteps4$Q9wrnLA*s%g<A**@;D^pHSj0o(ZZ=_(
zN{QwvliQ-?-QqZD1moKQ5+Zvy6a58rh6@;)2#yAp>}Toc+osYa=dFio^bzvk?F!q$
z!r&>U<7!p43c(MJ9kNL2-f=ZxI9g9U0fhYf`_}{c+Sm>iR2r7mdRm`JZ<<3r0#`Rq
z6A9)k?S5Kgs2?YF1#q>{^V_W&_xY^P)BKKE(9ENsvx*Ytle6_OJ4B9cZr)dxFdr#H
zd=l;K+H}S6yuICYID~$be-+;Zj8f?Gof@F~2IMYJx&3*4nS>`qz%<SrDg{da;f%Q0
zATed)5!Wp-4gGm0G-F%ro{wh}2x(t*-@{qZ5Y+bT$-1QivRdD+&33vn{`6R&qqKqq
z)^KLILlR4~rVn<>_6yC5&0*b0?1KtT9>4IC-DYo!=FtS<FP;jQ9RfGtgKxZlX8e4I
zbXd%$HJ1SC3fQfb*twh&OYTa=u#@-_n2L=nlW?p5e$T?y#qSmdv8B<<V)Ja{+DiDv
zxz63Z1k7LDYJUJ0e!r!B*JB6U>vf@A^YQ)FQC_w8<$<cNP-qevqc(6;$I6+$-+SP(
zWwg_&DVNPKGY8h4<us=Hqu<AUv|`<!Cq|zjfN;hETu@zVaw63PrWA||VzS@R8S*c_
z<ufG9+-#)Ld40_{MFl4Pp(qF}Pz4&MuaNyNs1a?D`Y6zRil<{I;daP~Z#sy-!%x{m
zV1^$c)xogCA^$suZcoU5H<s~MZnt<@Ge~{Ae}y<u$e4K0?LnF1S*E#<I`!Luum=js
zj7N;~H9i31J6HJ3E^#7pmNxaBEUVuyE#@*=guLy=Q)zI-u2TY@@tsR?D&c_4xa&vP
zgtFJoEnW2<Wt7I&T#PFzQ@P@f>0G@r>%B7FF(3veRwXa+&z%g$ftw_%uf*%$w0bPx
zHmUs<7arAOP)A8D!EBY)CRkl7YbKYMfy}p;%av}=u*WKSwyo?kBR1d(7-jvH!H-~_
zYu+q4attf2Pv&~Sw>;d_uC<8DkG;b-is*j6$7NP36VDX`Qqs+qw@OVk;1``EAhb2y
zZ_XKcRp6(hyg%3K0EEx(e<0*2dtG>b$Y95qpJ>0ToL8Ps3KEUpq&Vq2Bj!szixlOK
z-&J_X`a|^5J=R4%=m$~5)%#bms1#Js#Qc8R3bU=>kN(8ow%H}ACGOD)v2fP3P~Y9<
zU~b)OseQKW17&{7!xz#6ylhX-3tq5QL$69a&U_sQ!9g6-or}E|Tai&8U7h_y(<{D$
zaabv0Md72)Fh=#zLM3=m4#r%#fRmw#$$$RrX<3I44l2C~fwX?|9hwH-#s%F98$+!i
zqL1rs5tTq@QU@)tNiXiJv50Jcz57st7qa<1N%Rd4B)HGIipWT5N1Zg7cjy=r&LTg`
zzD$x;nv12R69#WG4t#*rk(@`WFhPhMZ@;~r8OPQ<JF0o!Tf9)r=sM#4LodogTlLtk
zRdY)=DMz)K9hjZ(j!<BILb2TH09zfe9~-l!9_|n)5J5wrQC-C@fXL3MBmJ3AgPf~l
zgdDO<9Gf`Y2~Lf?mPpXOnQsW^9fGRQ_Ml@s;J6X#eQO`8BsRAtb-kLT&(9UzjYr3Y
zXZ$8oKyMUgdT$+<*J7suSy?%EcusMdao8Wt7hCdu4+^is^#0pJkfAAHp#5%V)jM1w
zF`>#Rn%o(IRR)tLQ$f7<!^8v#wH^L57NhpbM_BPGjk%mZJgq<5CuZ{i!zB~Z1N{r@
z8?o3&fXRfU!#Fz!%ubxXr>;TUKn>k&HrLlgb>L~h%<L@V&dcQI`ZhEFXN!J^bs>t@
z%;twX7o}yIM50#RbyMan{i!}Ko=pUd%{fFyS~wwoQx+YNkuPFWZnl*OshIt(SKEj}
z-+=(gd-Hbj|5T}ey2rPAgk#<(u{{SH`Su^}7p3ULv1E(&eLimF#=cw-`#;6%4Lbf)
zw6Bg;mr743=qZY5`}6gvrc{kSfA2n<<nU{Z=P#8qS^sY}{GaO>5%#6{Xb8CW`xV=f
zyAx1~uKtA6eAvO!H~PM*bo62ck#4_uB6T;PVjTQ`su$`GFrb0aFWF#BR!Mg{v))V%
z#7caN>+3`_<v*RLE7f1}cP(9|2{v-w&DqF@SN*USj73Kq?<rUKu~R%cU}kgt<A!%p
zu4u$1GvnWII;t*GVR;j)>t>5XKsE)XrRAVYJUpJNDKD!O5|TjeK9kNYF`pV18RcmQ
zSCHt%kA+IOB!t;#DVxu?I77h68`>Dx+NiP!pSP`qZq}qjLh;0&7&cfyjq5x?8gAEW
z>RH_Fc;p&?ATn6CjsZ!lH7ky@wY~g&RLxPWm~hy}_Gsp;6aE~|wA0+y@?1T`!Vxul
zdAxzZ3ALERCsf?);E{yRfu7_1UMggPZF=|$rWA+An0sBtYYROXk?1zZJDllVLLt-O
z<B1E3e{}^eCJY)3bL)t-_YZD28qxV<fYt`_uItFD-vUsT-${W^-QC>~b=Yz&hbtrS
zgKB<1=Rq3@CvnG;YPi8{EZNjb7HEJ^|7Co}5fl!UE1eK@p^|W`K=y6fGDlzleXoW=
zqWy{hfXhT3C``m^nz7FJd_*mY6|lh)L$S6EelwEPKRvBynC$FMvQ^-8F%}Gw=0Jb?
zaepzdoWql#^}<WuQok*JsW|t^Vi_=RsblM4Y556%7~ilFXevJVSSAg8jQ@wf{!*Gb
ziYkNKHTggn&@4U}YR$)ASFb69S)cm^*%)DF;x-E;l)Xtk;EKOJWjaPpNv5shDaZak
zw(;F(V1HztdLNhUPUM(urI#9&$YS%!^hbRLvaYSNxuh+4!Ep2K)qIUH3JFGtV~;E}
zl>4>3P&{b+Yqmo5W*T(VADOt(be3jUe}?uJ*Wf(f*JnmzGiljt2Ub*SI-`U=Bnylw
z14KB=%1Wj*mCvX|Jv7)zg!(jXW%~)ETNeUv6pLQVa8%ro3+1;`NLW@Jm_Bz5MHh-n
z%9(tx^5Q_rs%qq?s>tFfwSO*eC)`mjO`KL2Bz9_41RUdah9t|~2*ZDIEp8%ouH3##
z0|H$}wF5J)*Pwgr+Do5(CYQdF4udwr3A`kl;OT0{uUDb^VE%t54V^lhofansse*#;
zW!<xtpu2Addi1!X4@o!As0bT?_t*D3i^uyike?tf_~oc+w!Gf?<JNjIeU(ZHU(k@h
zQU1*1nbB65^i1o|I^>M`ZU1ZY)?wli3FzUo2pgB_P3~n~_k-j8i*O&4H7!tQ*!kk&
zE33r?`??wisrZ&SQ=i(PR&&EfPDlJvF0&)OvK3CRKLT~OK|9!Xp>ZF8<R8w-8g%~+
zhpk>Ky!#Y-xoalvfBNh(PMS(ecLAiVX}CKeSrpsSJjZSB6IR_@KtQI<r$(9o(hc&i
zq*jl`Ej9@a5A_<gyQ;J}R}eC6I;aP170_JWpT%!m#Wk>P87>!PjbC4)i-PwT;D>23
z-h0O%l6s%H>u38_&1k=xu9R9IL9Z6oFy4t<Y}<e1bxj-5jfFef67t(=5VBtr_U5u8
z^W7_~u!F{Md3bb$nVVG?y+_CO^X2^-9l-o(<R#mS$dzvvPnx4i-RG+xG(;$UoNP1S
z_8;Upu85Na4jFvFpPpv<)9q>TYjw_p(T0g>8K=fky_a!a!!cxm?VRAXmjQi5G?TO9
z16v%a)f@-<FY!J}EC^d?sLnHBeCccGRRa;-8jP5T$6)?pZBKA<;OoNgk7sXn!6e7;
zhdli{@Q9@^Z`XC5+Dj&vDA)z9GZ_qGP~6By;~*;R<i<hMj3cDgKyhjanCEZ)WV5-{
z`NP>9v3;n6e!u%epN%cTn`V1RM7kBZm4y~GUwRq!jTTFI)b8EN2P>}Rvt^Ce>|HQz
zs!CZ}hSwp<!MpwRu<JKw6|%6xEWf*1XM;DRy1IJx<`Rf63#^XMw*?X#u%)^_Ki$5?
zPyhGE>(WHu_N7XGaqXv)7QV~nI)0i*04_6_)jkPBE01G9+4@vx&siq;8Z$%AD2n4V
zlnoV!nlnbigYN3>cq0rty-lw=;Q+CyQnAk*+>P~MlLyZ(UUqdbVh`|b6Wd53U2nFj
z(YmB5K_Y<6R-Z#%_(URqOaH_5Y5~4-Y7fA|wBdHhP&yUA{^2gquRtwB<ftp^IM0&F
zk}gJs!YQP_{#V*H_#Y%wU5DqAhRaGV4(LD0M<?|zFc*DXf(3x?V%zS&WmvuB7Z!6#
zw1n(x{(7&}<yPhW=RLerLUd#25jk>xcm{zXhU>h!eJ}#YJ)6;Y@K}K&fvvPbum+X}
z_E9jyX64k8n85}jiWA{|p7r(k$}fBPWm=8HogRGHT=ekc!F&c$fNlHJR6$gB-SmSA
z2On`)R*W2yj!yCk@=J<lY-=J0bY0aQyu$(!9vsdPZ~rqT=zZ4NM2p`=s!GkP1BQQ#
zVZ_pk(RiSz6N8IhhG!#DoY~@zGgN!|ppefJr?OE6%)&@Yk5}3C+xb^Lf;&(Vi*eoD
zsYR_%|DyX5qV;-haI<<?tkNZng@gt`Dz!-d+=Rjz#T(z=NZjv=!_)0Y&b#<e-*(BJ
zIMNWKlOtBL{mTh&aLnO3O@6K3*EaC}(vhoCJ+r5DuP=|Ukl@SEW&K2gNC9Ene!qTX
z53;Q~Joz>8;0%h(TvTz@Tx7J~LDA@&I}r!en_#k6h=w~9DU`^cp<D15_xARf5DS|y
zsgzJz!`&Wv6r>bRKoe*$w{lPYNXx<sqr**eg)o2^@9_`+6$lmN!%3Hyq$#L<(aVO4
z<Yf4wpeTEHMBMIY+@HP|RtY6yuA~*vB6IV-jgaC_dP05PQ0GHs)G8WSI9S9uS39vW
zz~~4Qm`;AUJS}B3%kSu2@juXyYc#6t$z+y#I9DTsEgfHBjnCFw#g!6&4)w(Jx&6I$
zOd8}n*Ieb{ct(L}tv{uQ;mf080h*&j6*K+L_|zfSg%A8u%9`VHg1C+?dqJC91$F?A
zGAB~STO3G)?&r~s^4I5jcz|{D>$dLYiIZB!%<Fk|8W)k!$Sh6(&APKelNGteT9Z;F
zn)N3)bll12Q4XRQL7b8gF9u}`kjmt|`7qVW)Vu6#vE8Vb*aQY`G6IV9@xYfEMoqub
zZgrR4d!m+c+AkT1MrHj%cS?3}8F^nqd<Y*~y3u~;n2Iv~E@YAK7dZZt@wD;yv$j8g
zzl`3z`7W@bZazx_^!<QqBk=G~!Ld3K7e--vqJ`w<NVeZ+7LP7Wi`R=x$3A}D7<dO7
zCbTuM)=j>4Xx~4`*s*?G(>ik;)%N?oZ||cIC{yd1)6c))Erj+xJJVzJ@K;|XdtkJF
zKiBvge@6J^I}4y+>xOEMQ?JT73cAYc+_X^(tirV%Z~G34+)*-_oA_<qZL+WqHob`L
ziu?+CvME-=$^{4HAi>6mzaq864##7KL|HDU5<~52ICJXLq`S;X&w>*C$;XT>NA6FC
zD2_{J^y_E3*FDc_Dy5l6y)tTHo(~r*aHLyLFDd5k7L~wX)#-ymjw8%h(0cQOGuekE
zuE}T{F>RQP_h$6FC_MV@qA9;}IWT|nsf}ldL$Q&O&f$Wb&G=oj2)*+U_2n}|iA&iw
z^#qPhC93q*iu<W4`BU82RgI7o=YmQlkGf);-Ie=&X6N>P{RfeCqM}ohK#R0%W@`QV
zcyH^8rxu-#QDIv&+2)6&eH=!vf|gSLDgGg`s}qF*F3sv?%m;NxS1b4SivuBCO-!fM
z3tfLc=79u%-~pammvLlPDpkB4Rk+@61xB(+i<Hr0rT&x{Soi+TQA-M$UG0VXDt{?f
zc`4R4W2Iss=Axk8-W|tXT1v6G`IwwH`gBIW4Bj~Huw3wGsY_EPX=lmvtTBdMnQGMG
z>jkb|$UFM@HJ6|gAtLt>hvaU*-}NL_vn$N~T=Bf`Rj(lNxwu)a3D@tA;~$|9xu)C>
zus}S5SnL*xa4WGf9LkC8$<)Kkf(N??y6eHj(%~m&;8)fmAoNmdMnep^=$oyjPh}R|
zTBZ6S@yz?~G-wz%Mhu2$>u3p~jA#CYW(h^1v%L2!Q6$85{MKEofy+FbT$?hdkCo@G
zPU%iDj@oU*hQh?@OyA|eXwJP|`#;i29NdbOBMv|OYhY`u41Giy=Bchg6lhU?shAlL
zY!X?_d7N;WQ_^6lsBY?3Jw>!vqFU8;Br)qcq{!TTlUliZ$N-4*55ssU9i0B+H}6ns
zl!7R;=g5o-YSV?7!}|EyAOu5&z;Le3PaN;)yc$8cxTYvP@yi*>x@sSwaFqVbLS#;_
zF7|;sbEzEjoq=0NL?Nf+ZfKcQ{Y4-d&1Vu?a|>?&m1<|45~Jjal)guivC?3vmH;PW
z`~INi%TShY!cJm8@<28hi-t8;1(FrgKi|BUCB0CRylkO<^-y52@AX9&A0)xk87u*@
z;$@D#nA<-NJoAJR2^y7Ypjea7#dp#~KKQq8BxlsSWQ}enP=0#lA*+ZOyG+JMkBK?_
zx`GNek`U6;lYtu(MQB1YeaY9ORSDdX)bhyRFVYl=|2-JS+Mo|aYuSTU&RJ7ae3HCP
zb<w3HE!IOTzLsk>2eta=E$x#A?+u<7e^gc;Mb}yJMFxoDGf--X@8r{mqtl_<qq(9n
zGnSQ(q+AXPY&PPKPzuV=6ca&#r=9ni1c14|w!o54p5|liPo)k~;-odx^hk=}c)NiQ
ziwp4%(O30GSY-WA3H!*gi3tI%EMJghd+mlz#J4oeR3|BCx00l;YW0V?xX)@vMRi}8
zeDyJ<b^2(n$B;z1DN~7`5=;!h=a8u%2)0Tv*ycl0VspD$U|V$992-{dnz#&LuTu+k
zLMx@hY2^)~%}DIA<`&z~+I3-NLfW?t9xPk;Q@yX{Y$JA(>ZwywKcVlxO&1}1cSFfl
z>1^=YFIvuCV?6*vgrw>@*nQ#!Y#l3=271N;>Xa*LWa;WYySF9nUVQUA<?qUdAf!Do
zR@P1mCFW}+;*dI{nPK}*!pT8wA8YTWnWX}+b2d82@FU9iLz9bc>r&GA4LVP7UhF4Q
zAoR5j`>Gz*Z80ut&%@1p*Q~6m77(S&d91Q>f3$J*v-Eet(pQ7_7g01fiRk@Sz>d2;
z5h7TX>S9LA>CTKFyKI^7h_YgGBE|Z*Wb}ls<4OJOgI_P}m+sOSrQ5X5%CqCpPTbk>
z*(z-DvFQvXHw3N&qMR)f?oh)6>C}G9WK|Ure1@jJ%h@BEVj+tNRBd3H01<kyra=<J
zu$6~dE4Nh>Rh%DAPT2#RNP#aDL3eHE&fTSTP-fRhvnPRK*Xzc|Kvyom2d47?ge!;r
zd0`x!k<}k+?r;BfCHoab#p?U|mckh#Em1UBmiP-nD%7nKoG^uEABV%4V2<p6a<7c0
zNKJlFtV|U7I8i#&W1IQk7z!mWRx+j^T8iQ=fprAK*03A(@{Wm$Fhpg#aky%g-O)}N
zT7FK&?Ug8j7bE9eY8$&G71)vRe08`ZHV2lQ7;aqM73SG-&K!ko!^0=dYpKdsB^!M{
zJ-BmwIH#Qm@qjB2;jeTla01teL}*kssS;&(>+WCPj`9QWILNzv3Av<Iz8EQ-%}^UA
z(@GK2PB1<k_lJoy+gQ^_Nz*l+{Sq+&_v_F_V*#naK2|mns`iXli@uf9gvK(MypdSv
zOc42{ma13poM=+J9;3Q?S4^~oChf>4seFArE0nbLZ5D=)&C9e5B4aVyhaL0syO`IW
z^uPI0Gvjx6ubUf9$F1&=iu41r#IiGGTrAhG>Nr1U$n8N5o!rA)CkCig<?(;9sljj2
z()E_=I}`BmHSIN;vdlnEEIN)^y(s#$)Ee2imC<^bco_ld=ttG6DosFDSBM9rfzRwA
zkwuc0NwhQ_zSYF8j@zTpTe^HsvgQ5rh0^o3q0Nqp319mV!p?LI<mr2aWyWzz%zWrn
z-viU-AVv{A&(ynLcSC^`?(`P2`K5iM>F(>veC=(sZ!p;)sKCfCL@^bOAUowpQUmtr
z>%r_`rP*UQAVy$9{!<pa&LrG4WLcoWt=n^25Af?_b5RDKkud2Av|;?BjHX2IPtCc6
zmf;<?K=NbPS9S7fwRLK@Vvhz>yQmci*|6+eI_i^%GMNWA89Cl_<?u2+?of6ve?C*Y
zm5AZ`OP&5X5$$}S$>NGQ=<Y`Rw_s@63h{&^L&Bha%az(n*I%gGTh8YvrB^1WfUI;M
z>l+{0NOj~h!?1mZWz59jL4y<nSC<<L?m+^n!|x=qVN9-sY>cMwq)Inesa+e^GhAC<
zKTQ4Esb3Ml|85|QUw_Jg*<e2+B6pSMv-C^V`hwn%)RmdT!O*`7ySS_*;0&XT*j8!4
zY16@wuNrld-r*eh9pQ@lA#Np$K?Whm^P$HUF`bbC2@NV~DrWk9)Uu3PpB@D$Bw|%$
z<94ijHOGHCl^z-B{vFeOBfJOyS*?DCdybXF)*2M!6yB)I!NQi7#XSg3cKw?2Ec&Uz
zT1T({?1v<s@pyCqYQ4NL{*!RjN3Hz5Jyv*kshI{>`DeEi_HIPIrZeI?;W*Tl!I~?7
zEivGic@<GRGFspH%>9ow22J27xp)BofJX1)>?61t@ClU@?N23dP`{f3P0`4=Rc)JA
zR%~{cU<OCwz1BHfMZuh~N9(?DJD>Y5j<qU<b!@CBobI@xk*||<rV%6Ym+BaU4A|Ul
z4yQj=10`rU_;$(ZMlW#a+?oWN&XbeyO~Qsj=?tx;T~=ZJq=kKA&gm+IY>vE|SNhVV
zu+Qt~bM~XkL^3Dsjs53Ldl+-rK4}%PW}tf$P_GOZqAcX%>#9~sKeK;uz+bcL3x*tS
zDLhp?h~UDtqJsu&x@F)+piyQ#V8At_3;k5i-7-68s)MgM$!p$aSJDiL{`}zlXhn28
z(qfjSBo;Nxa-ha4xMGE`bL*^i9Nl7pV&2e?fN%~z!$YYU1IhER4e#|C)X;f8MU%Yx
z-j4=HoZ&fb#@`28X<qfwIicA<!v$(79ta*7;JRdG&@zVNDH*=2X0^#1!b=Rb2<<Ab
zU?HFguF@astb!A~ik>UNtO}0?NFfBx{aTQ#1w}V_O6{NhE_@*MZR4p1^p?Vbh&Eht
zVuMK1Q+aKsyrp+h7#30ipB*A^q9vL)i5bTv^qwisqJX<|PO~mz)kK|{c!@}yV|8mK
zt2H&m`>KuS44<J>e6G64hE>d0x#GAtlT7><8|9qv#tzreoos)YP%O=GS+AHtY?g%8
zj}^GWYX*mjeqK5o-cfv_-p^F}q9d-u8`R!1T78VY)uFVKqI1Q7vTqCk;U4TcJa0WW
zgPJ1o&>au*!GfYuH+`&ES*(6X?+mSYC*IsV3bo3+m3%{I853cR)<wI-)b-6{l5K%W
zwJtkkwa2d$tVK8%cGe<^xS}e)1TZ;AC7vdMGsN-~M~8cwKAx`moE*0>r_ZAbAnY8L
zdd019#k?*gRi=13vMh_`2dISqsq!?&$i5R-6@c-%>-y6wG%@I=kPx~59+%j0mEQkH
z8wc-&5y0?akC1dFHY&*4ynn;jDBSvxnD0aVVI8aa@%r{Q94E2)QzW+Z_m>AY**Se8
z%JAnnJI_<4B7cGu!qeeFyYlA0K8$1vIX;mufQE<RTr~L}Fzbt6N;j)e!>;#>ZiO~?
zVxOAtRt*|!xvq!GS?>a5E02fi>o(0aU6xunFAd*DOmiv{MJp@f&&?W?Sz$+VRqSP4
z(*eG7>=R5+-PHK)%06W*(*o=v38sBWc?G&TUID+>HV7%%X0qklDES9rM=Si~yL*;(
z$5d+C;9vmNzhbw1jtgX+4wMDfS6p7&nALCpndV2#KEJy%hb552Hpew_8fwTWHg$Ov
z#4$DwT!_|y#YvmDv4kGkDq`tT9f}yRUF6m(jm>+|fwG&oMHMh5%HDq8CP(YKe{03W
zF#ovy9@Ywq%T_Oz5bgaThbdWR{d6mR8h7uARaNM1Dz4gHsKGlr`Vr<NB-JJ6#jgwB
z5^aY1a6j8UN;<h1Vs@=sTYPz?iO$M%&JSf;v#D+zIu4(3EYN9sKAqfW(OW;2j@dd=
zez*~l2)vNdLSk`SW<Es+S+0-S{F~}&tNbK+@Uv*~qT#$7L2MHl%_9GyxYQG+cNiL(
z!cJnW*gnlDVJ2dC3A!|nzOBhNZyKXvp!vl~+up(~5ih3Ln0WLn@7R#+y>V0J?lYK|
zj^JTP7;C<5T1+#86b+|A`eMb{^VZ(a+V{>Zb3H2%%WyQ0p4T7aUt%ltv$nJt<nsUv
z{WuUzB}U7MTs=Pt-85STQwHMC;%i(3hJ9H44gl@9aPf0yIiQcR$z;oI!r-P4u(8u2
zNo`8rxxU{~3z$mmQkI1N?AC*E>Oj%CP{`|Ih&Ms*NP@Tp26X|0{B9c<`Mssr3#8jS
zK!ca0-ibhC^gDhYmcQi--gR_@fw`HE$ENgB28t~gS<tGzS~nS5F7%2jx|i@DUn_*O
zcwSd&B~GDLojN6l%ZlH=6*o9hQ)9gLw%WpUe&s8;Zsq>})BXj6{$Y$>Lnh>oT}-oF
z%@~WBPVwBEct1C(J`0A>OMZ7<|HYY_kKSdNVDgBRcR-Qt6ou9u&m`U>me~FQt&q_a
zfO>nIxPpGGaD-|w@NIjbZU|%e{;N6SS58uXb(S{g|GL@+Y4906cFC^f3E?0m<z$N{
z>VmQ@34+|{zBrFj$2u;7U{|LSolh@Gqcs{1cih{4C8J{;M&<b-FLWLhpvz39f>5q6
z-6=K4GW|m=m0Al|I>=CLl#=}tTd2xk-L*!2xX1Qi^P^-PUEBK-CPRkmNoapGdVydi
zyG#!9`8)S17g*6TY;AU0Q45===`Z;69hSBHCl0q8rF*EzcIo*3{}FV2q8odHMXA)K
zWUG}`&Q9rwTZNwRbENBE9oqj>elY|~>Z0lm<SI)Ld%%_}*B_I?IIB<mm8L2YC!w>f
z<;s|Mw<p;~&ZQj%uqmvD6eAz&eOxl@uuDGoGusJqEngk;k1ENbRw{p34`aCFm&!z#
zlGFy86wNK1jb@z9*aYi~;GQN%3cZR;M2{EjlL@q5Ck3kr#;dX^@=2Qi00nhuc<c4^
z9BJSZ;q6NBJE0U@utcuB=9td7VGygJSZqx~)9S7k>h-}DO^cbPrt8%746r5vvsh7z
z<kgAIE@5t0sT%m&(p<5jMc-(AG#C&C>Ezbb=hnXM5MsagAo<F!^<(RzY*aoj*@ieE
z+sExRaQ*1=yOjWTZ^AFNLbr#OH@T7?4hRv-tL##mU(t;&O%3xHOSWQop1qd`&E67g
zWY?w^yyhg$q|SNpjQ|fFE{7yoLo9jI9dSHC08vwPQ-q!8X5oFyM1(lY$Q`U=k+DXH
z$yeQHDX(?TNJ2O5dkJD*EgNbEf^&Ce34VVbY7@gdN%kVT!U+`js~X`!y*9hCBN})0
zp$MEeVwg>AWdXSjQ0Q>`Rl2qIyPOh<)tl&?@$HM6`M&#LVNOwY(SA|T1Crt07H{G4
zeZVN+14_TqFum1mIfOO(?J^DWay@txfqznVxsp=RQoFMv96-<sV%wLSA5cB^6LXV(
zxY&Ad8&IW+i;c3nz57unB3HRq+;KjBfI4@_A?;V}{@{)x><iVUZqUbJ%(HPasr1r9
z{4yGg?5(XWNN&qA=~tA<!>6>NLGx-nN&uh=z+^cM@cL|QDvswFv16*YD2GuJ)D0~z
zW*_b(eJpl*;O_yN3iqR08dZ20H)A=twG`B9y;|a9c=29~?XK`%lo^3@z|BdOk;<~L
z`<26L3IW6*zv%sr83fY;`3FT=j1}w-&$`#1B^PdzAIVUyp%R4ZvQLV5u$9t&gY>kw
zy_Q63In|F%b7;@q6|=pOaGq?vJ>dz+#CP`uy?DLfMV#}aQDFRsw~fNBL+7lr_8&*a
zoz3@!Aq652PF<c0m*ZiQwkxHZ3rV8DVfM6<lR)oSHpt2_%sf&7XB_>F9R|!P(^A6&
zX|tS~bG&cZY9VP3DRZ}2@_ty14mywDD;Ut3PB26CdrF9>=G_>zmH-0>$H(6hzjOUy
zaZ1Z#N8z_`4hfCrv!wGR9u7&$;cYGnQoLl<y{^R?XIu82V7^_1vlL7u)S=xdX})&c
zXdPTPrhMR06q}o^VfFrG5Kbr&G~Ow;KbD^VUHzh+Y%Ch1m*$6?TERDwRl<xGF6WK+
z(r*i51CwMb)w#+&FZ%Ep1F2E~xBa~G7v!WrX(yX-{zwH-Da;CJ8+x~jjBISc@Ut2`
zJSOu7XH~7NMjhOY0q!(dfvUIF(Lqm2;q8K)i?|Q>h?OKp*8U{#>dwHdqdU~FuacXj
ze`}X^Bw9W4;{KK*!ygjWN<7azk^Ykcrf|?5w`OqWNIa8pgfQ^t6u!E8px@2PLl%#6
z1>BJ|v&5Ld4ul>?B7&aOuP|l|rzO{#i-$YPyy$R)-GxkqhQe8K%9CIay+BgR_TF*|
zBKIx9<W{S!JGdBC6$)0p@QZ-|to_}8$$1QJ4s7H(>wJ%mL-Voa3I0Pu0bVQR8keHD
zuU6~FlKI`Oh?P!nni8Y1EYJ6WVV7gBM-Jz!DX^j0M7>Nx$>i?xR~sxD9|FYQ|4jQp
zqj&o8Ni*=g`e6h6)t8U#KCjrebql3rLurNY7515;5*7=ySb%&p_*;>}Oh#Z0rg=N6
zW8}Fk#l<zxE4Ar0%Bb+?S@zIfHbZ@`xNxgS)ii3BUoh~#!nq(tv>2wv{<4S^HIWc)
z(!x1M)#^LEVLb?yyF|cRJ}9<%<GGJ*<0I|bR+wvwXx@p?SzY#l{{l)M8O;?$1uH;8
zHa3J)!OCYV`Nts}kssE{L+wLP(szYMnLG+O*x6HBgZ6?RUy3Pn9eFD?+=D}5ZoS`Q
z%b{OzdmYR-J)C;NM6*UN<DE<Mf*)~DU+IVsKKohcG~m|z<v+%)$1<ym3>p;A-gPgV
zw*RU05e_)MrKa#+W{iCP0ws3s#kBmtpJ*d|6bv3wI$By|V`F3V(*S<*PGIL^Cq1_{
z04&yO{~O@@U)e}3J-QKP`_fGmi&OzqLI0}9_mtG+`W9VX2ne=(V^?YPZ?xrqCnicM
zbfq7Rhdy81pq*4dg_89ennVX?H8|Uy$6MHMs(<pyd0G6cnEtQdu}w2vl3v8bR}EmY
zM*qsK*?N-cFU0Zx{UB0wdQkm@&LKgi6w^9>(L||!FXQg?s0oS=-!J-e^lzB=|9!KC
zLQecQS{S}QXT~nSpdd1M9Y^6Fo|+1X!ehIXpkh+|#`iH2dc@`I@h{Csn??4?+OEsl
zw_GPlT{iiBN!4cK-*&xgyNGZ5Az#&ORbL(rpjkFU5%CKZ0qWB}6)T@_vo$;Jyo)=i
zC4M#fNc8~r!gj%a9rZQRJsscvqmPNDLC1;3X6wBB1sP%Z#7+r)EGWR=_;Xnk^T=eH
z_T>mzv{9jNEPq^Psm8T)K`cw2d?Q%q(g}Va;?!9yR_O?x*M+&YM-eMPmDMu#5c;~l
zri0Uto)7R?2~R`VX<<%%@oW~ieWHlX{M(tLs<iCoYUh@XK>r7K!=jf&R05b8JYL^2
zxGnN2X3OJN5~@&sBtP2?{5rP{FS?zu?jotrfK1`e1NPTZO%f=Z6Yf8}i8WQ+(%&Yn
zw=K_4I3gTeILzjhQ=Lx;{!BOa^f1p%Lsh+k(>&(3#w`VCSQ|5*=|<XooQi*ZyqHv^
zG;sXkIrROx+6+hZ+9|_R=-Gr;oum41<V^@yBMC+A4sVy*pN5A9xgS9pNECIjuY&|B
z=e@6_z29B7T}rQJff_?vb8_tBQ2chlS@TZ4omB+X^v}~JB_D7|D3<9PI6A5u^*hpo
z)3o{w@$=Vh{;K=^`_J7%76Fu?b>|sqZ;xerX@u=>qb+4CmTVz)x_Nqimz>s%=Q*`{
zm00{tSGY=>>a35RGRcU*sa)#r&Jxt1{T8lk5g2C|rwq4P$ASBvuNJoD6*(qmoLZ*;
zG34D1Kjm^Yzh4%EAl}vMJ3W3flDFUT<mRB~b(!j7Eo~}HeolYYos1`EAmD}}t)cy{
zl#&ttaJR1sDuzh)sV6`!+C|~Z?rLY#^6R4lutj~se8)3F94d7wDg&Fpp3h(_oZwIB
z3HVS&4vt}nDi-0y*z7J)L;pRW!#fEWbDjH&6rBtK==(GNNE@~nCdIW!b1;G9*@MCu
z@@+Pz;7;rf`FfN)R8`MP+#mt-^5Atp^%p8v(Tw~+0cEsT?Fu+^FIg`SUVdM!T|yS9
zZZdz<c`4Ga^#@8X=F^os?>j?gz%Gx!2P1_Wl<&5hXXBvMAVrM>UwA+}3|!FpM)^3g
zpT;(3ghJcA7E|w>ql{Y}>rhr`m4gOn7>=L{*5d#eptPT}o%D{a*_yL?ASrGJn(!wQ
zzsJ}tRmp#RVq*K_K81hv;{X(*{)Qwogu*D@0GkE3#Q>}B7;<&@fo3h?G?rSR&G>-K
z+5OE!g<0)6+}!_rZzQtS_ku+I)7u@P)*9=t(8Cx*r{Pr7BTJtOe0%^Jzk{TCd(+&}
zqTd+z@d(y2s1x(*(tVk3Gpn8sp<i;Sn+iCXnw?tT-)i0@FD|9>^trk`yFa-PE}<Vd
zUkg~D`%1yB2nJ>t*Ey!$g5^%V=i~lD^53BQ0)jW$8gv#1T%u6gGWDx(9#Ob}p8gCB
za78_1u(W=^M$>{zf|1epDg)GlX$!ay$G}HhCoRSMN;EoY%z?&#%o5};MI_{)rtLXD
z6m|S0Epr=l;0H<R<KFkK1-{!~Q~rlj?FSl#bamgR6E7-oNhmAo)l4ycc#sH0*83R~
z$z8y-_ncae{b}MlIw*g9;)@J(+?#hio+95^8p6(7yeU#~&$DjA+B}ee`E~@Tne>x|
z<f9R;*RKR@wcf9i&gfQsYFX=OI6h1&^zE3St<+4BcBoaB{vtvXIM>|l6+PoVo%!k?
z#TJ;7G_^Vidj32rF8vq+eFuq8`Owrge|x}1eHDGz+PFtAC}r~O`Skq?bmtX3IC@9<
zs`b7q2Lu+qUrQHUdNRs&<$3YsP#on7701N-<=1WL1><$zJ<9COJ#)(B^NcC4^LwD9
z`?8k`K9cr&IEI4!+-B)VWicj@34^1AhnP5p!|N}}6GS3UoT>d@vjppkUBp2bGhgKe
zOkf~blU5@5jlGnr;?hK23yv>IhuleV0<m`&t;#vwTLUk2;koe4<B?4MIV;)bA~W;7
zz|*2MO4$n<aj&p=ktj!^x$4%Qv~x8bT|Gr$28);Z!(0ePudq+p)!+4N>=k-qS@1I#
ze|J^s(BztKh&D*ruh9A*dbg0XXnNcDs<QLxP42q@!zy9=M=wm?IyX|F%;edVf`Rw(
zpcIA?bWrK2iM!LhAe=?*<(8Ar3I={!8ItiT0s~HClRHYE`j&`;b8a)NyL2g9SueVU
z2tHqR*^*sA1pVxPGBLTx0VI1Yb%TVhY>+*b{UE7J13z1^_qW7AyLh-^A@H6Ejb#He
zKYHAO;etAFEYLzsv4_&p#CTJjiJKQ4F@s%#+&_XhwBEy4>LbhpjZM&=)^jyISPE)Z
zUDAy<HE3D7gHKb*FFuNrcm}zQPC^VgSS^Gc$yV5vXKvg$B(hr&Ch(>q*bYJS>TBv`
zQ!ex`w#A<dVAV6=onI8jU8+YkHY%<D*G%T>2b;n??>Th(TAuJ*k^}1y7E*Hzo2-nI
z^$Pb#5nLqE1MlLtR<Q6z>%=DRDd{4+81^zn_OpfV2wV#oLl&n^kxMIqJDlwob;rL@
zo?)9)<)ABS2|<tB?MnTnyHU=w>rYf<^o&N=cazSHgDLMUnGn4+OYG^VR|voOUvcK1
zwSJB_zOH)<2P>+J$iJbWPjoKVgo=r@jShB>3C{VGAtMyll}p4EC;<|8aLw(xZTV^R
z&Oer;2l|!52#7sb)9?A$zP3rmm4O)#%-I+vJ&BTFta|;1^)jKr_85|#k}&qJUuE6K
zU#7O4f}S4hT9-3mx4B0zyGp#6Xty-|rR`oAdujRx(*&$I93zKSIzVnIWWc$_ankQm
zG))l6dg;^(GX#xJmP13NkEsJ)qd<ltQdOH-qOT`ze#Vn*%NTPI)eo-Sd)<YdH-$wu
zI4u8G2ipx~MYn6T00FD%Rb^|;^_qY~wwR1iLUkPG`h2?j098tUozi;T?W|Soml%W!
zJeSQamLaG_gGTEt6BZd{ks`^tB8}_z^h_qmE%B|`$*nOsKlpAks?3=;Y?e(gonN+q
zklAt|zSqon8|9ZjM?ViHT|TqPmuub66uL(xTzR=Z9{hSX#DksM{0sW??zIVCTG{ep
zAanj+Zd$7MAmF+OXwEFA_x`U2a4Z0eacNCsj@8M?)h_XE^t5b@#;qC1L?#^3W%<!<
z&t}}GxVGJK&hW@|vEn*jQ-w66#=Oq64hdfo-EVcYU5kr<rBVRGAcx@()ZHW|8d{pg
z-w~C&R&csK4SnXn6#*AEspRP<<PJ|Da=r*7FgZjb_MUJR9+^K$FH!8;46PLYIE=O7
zwkx8N`-z?E_`_g#7@l$*>S#sTUleOCB7WV7_Sq=abo%^f_DON){XNW8DZKHL(@sk%
zO=wS39!K%5nZ%JCK96KLHDoL8{&Gp(+~>-HYnki;5&F)UP0r{#8axG-SnarQ85A*`
z@Q~>fCJMOfvV2}%&|%`OxB8WOak*(TDr-NA=8x%fE$oD0P>ZXWm*d}mk`KBl_Ix$h
z?`!+=pD3GbY)ijWP`N!*-Is$YBRcl50b<-VWtCw)b(cuPhTF2aW8(Gqq4vaj&qJM0
zNQbhIt(;ukLud&%X-G|$ljBz{3*Tc|s&vS?HkpQ@_ixC{EfPqb*bv2`Ki6n~c`?$H
z<+R<EAU=Y5K;TN-xW-obb}qE+0~YT2G;MpEhO3x%6pPZB{k9?Jrg9S1hr1EV^fV>K
z&Dn|<?N#II+i{Y7orvF2*0}fYW8=Pe`IpEKX>+!}EV#3j(@azr;o!2~vWf}uSJCaA
z4<CRApEPxx9`O57J<q@>(Zx0IeJRp{>!U#CUTR>b`nP@^UjohU*d-zl6@(R^9d|z&
z1p4wQ%*t95ff(4$2n3W^6UrbWQ+8kyUJcNzCT;(+XHj@9td-QxE!Er()L9A#K0(y)
z<?YS6o2>`FtpPzC>%vZr9|C<b8t79oaR>N-M|T7?P^~Mn#tGY!PT|vh!Ul&#$4b-h
zky5!-=Xi4EwCw37OC62hfHi?eKBQa|y~#lY^_P2@$gATNUit9A#K|4e!Fsx?hS0KL
zHegm1S{4Ckgs}Lph5K)gFZTG|4}j&)`hO2iOL3AYdoa9I-7fSto*v4Znw^|_W!r@)
zzEH2X8zwx!EvZaMTQSRsQ6HE4HNB%)h97NW(x+&8_=_^J@om~F+Z9zmTCcdA0aN7H
z;KgDn@e$Co83?#GW@NVheHajO0KmP!w7lyW&@PS_Qf2a>vo=e?PLoEw6IYCeG3-YS
z`Ak_oSNm!Ed%L&=bswKNd1C?5#!4rOc*Gp*Q*DJIX>aLy&Yl;^b%bc5wuo+KI|@Y{
z?a#ZX;t49jL*8;=N_jmJ6bklLCA1+l#xc~%=UKroob&M)8@be;w)x=qc^j0tnSXl?
z+Y)4Qy8Fe8h(<ItGc(rnTQFg{r}C-U$fegSw>D{58q8>&@pCf@a#mU$7$uFV7F(K@
zvH^f8EGHnlpe)3$UR*Wx+jkAMgfff6oeiezDjGo!z;V8%d?Xd!IjOADR!HIx-}@Ns
zSVlbcrrO2cPlWc#QIEpT2j4FkaF7osug1qq*;K?o^{fujD8|_QgnnahiC$H|#u=%S
z*oFDhIA8dpxL$UqJ@ofy%1$lS{7=y@)8};xQBe^}Zv|1Fz<N4y#DJE%3)u|+4meSz
zHs^AT0Guxx(PvjiFdC!kuFD)pok&R&MWxfh9pP@mnm$jsOn@T9E@gov5VC|eX{JRF
ze*bLL0fFlSHD=z=kjgFI78oLef0+PxdC>eXvk4ym$XnzqN$QN^kch)N1!ByAv{D}A
z#Z?A}KGTr$1<pb-EJ6mL)NxqYsi`^|NP{}XzRr@#-lD#ksq8myks)s?a+pi!P-zNH
z6ac&2Aa-9eiM?Md?g$PzXa%}p9J8$wdsUSyadk<zbQLY_vK{3VNVZ|F{3CP$<~IUu
zuar~}59$m9lJZmI&uuHpsW#eR;i9IJ-6?us)O4IE?~_&M?Tryd@cBO5=bQ@EZ^yf2
z$D(!y;9O6Iq^j5%SkS)D|EQP7US@q@9WuU0ERU!MY-4JQ!Zz)I{*h^x=XzB?EN-y+
z2W9DA6gVg6cMEeTpGtcv9;iH<nshEBj8VIAEEjor5l6wz_bx&gN9!%{B0hm&&_KZF
zBE?30NeD?%2`ADMGob=0`MPq*CKHg-gC5+Yo4-g5B-}-fA;tS6;ge6naDdQ61cNgW
zP|K9~4z1nN<8wXZPZza)XDlY`FCLq`EU{<D$%jSgub@}A4K2d?J9PacMIAjg2I1b*
z+M(CCb2he-V8Fv|rIwU8t-`ygl!R}Ud~U43w|YQPgAbZoOaHN-?uw#S9oQ{@RsLU{
zYBE2Q@gO4Psmb_i)R9y5>3i0CXaG_$uHTHC%TF2TUCz6suQ(CH6?o0l$~t{zU~6!g
zh(=D-UUwtmcY@J+%IN7bZkZsYD=^um<W|d(V|XiZd!{uv<(9~y6TJx3?m@^2$wnT2
zWMm^_@x!lN!QU$dW)1y7LRbKeDVEspHgt*J6zhhBU3Q<_deh7CV4@)mPpT_-G2GQ%
z2nIUR1Ki0NtRrWM11mnb49jQNJN+_fUAjZAG_Al(&pQV?n`W?r=SbZ5I@mPXqdhqE
z>Q|*EP}w<~dQng+<ine64O%r`YZ<*Bc|h}cC+2m4Bdv?f(~C>=XIfzktAh8pzUuhD
z#k@1=)5m8ABk$Fe@^7!^uB~rTPJu0L_sOSSp95UlNQqAV(NYs@zB6q%Jz2lu4fyV;
zx|$QP`}}~++7on1=ks27nT^+WfXoMwwCF&W9ZztzF;^?1G|}!J6woa{w@^|pGM@B@
z_)fl%v6j{qGb2ShrX7hizgK+OVik_(me%L37(!*3BJjmNgRsrDQ?$KF=Y(vwypCoR
z1Ph!XmZf5WcoXs}5|27p#lHlzoZGa%S|7(8{N^1rOR#JHu#Tyo`dbY0U4OtIf<^QH
zvH+@W4&!D2hO2_FuxgG>Vc{{9M<1;TCW#DJNCrty9IIM(3^<S(wS0mN6IFf2(a*+`
z!-!H{TR^>DLCB;!|7n`&?V<EZfp^L1a+~q{zcNftr*IekG;Wk;pV`D`qolwdpdo^M
z1`ivYr=N5&GzS$tQV?CjpXG83A>$>+l$G3UuZfVvg~QDFp*7Qlm}RU6A$;scwXX5=
zFL56cO$+^_dbiTg+|s7qz%Pmp>Ps~GTEQA$L1!$Dxn1BM<Et1@Rp0x0O1Y2B$ywDh
z1-m-G;G;amclFcd`<3>W63p}VhtE7SGK{O&pGjfw$OrV0tjCJGuqTNBEFf@sGMbn;
zHJKMzf#=H;)kCU_ol6Ue`szXf{KE-}c6|)hay<|EkTB9La5~iSeeMKriHMxZceV^5
z_%@Qyy-&c*1h-N;AqwCAc|KI-O?(77o+joP2jjCv9^fpf5cyZv*ajsJp4L@w*mK_*
z^!?e-KPVth2nGnH_EpyCi015nG%X)krl#N5K&dVDSq|D;3btCPY_Q<^t-t%SCH}{}
zK1byg!Lb=cyOIhLygAKFii?f>G~MsTXv!mz%NKQ^B@j#~rI%b0dU-~!V6fAJQRIvJ
zI0GzR_d&N<iHPaYy}s}d&|J|L(8V!EFcrxZoK4w*MrE(h3SR8a!HSYGNIQ3NI`uQj
z<BCbkeI~l~@HxzjP$SR{dv<L)ynr=hGla|Cb`*R~S+$q%w5T-0NLK&0kVpJKs|c*{
z2=o?9LwtgXD3oF>tt-)Dg8H$lTHCT@ZiUc6KXzSJ82fZY-QnW@;R!9VtglL=blFa-
zlo?sGo%eAmM_sk=Kt#$HJBfsA;TY$OC}^>n2@}pwMDUDorl!~2c$=7qQjoYd>fh3r
zf6V5+SZF9>xVMiouJ6s<w@c^NJ7Xa88@12yQva*kUsdK~C|&lZZK?XeyI3uS&2fXG
zzZ;z1M`&UVdICgKAzY8RMMhuy7CrTQ^XKpPgtwhWNvhYkztSqvu`{y#Z<@+KVRW%p
z=s5p>IZ;$V-Q1g=eokp0Ja%WI<WVzepUYXi6vdfLy9bijGgg8gf9yz1$)DWEeY|IA
zu0D280<zd(r57uI7PG5?>`(Yk=^obL+Z^_Gvgg5B2bnDVeo=5u-d_gRS{2@V4hb{w
z=L1!QzA@SFiV2}9SJ|N%nCl<p=!|al*-hmrUI~ABS)+)sztTl{mSpi709~@K2J(~1
zA8cC$U&TW1?o}pYtx6AdnT$*m4!Nh4zBo2mw#X%X^eN2Oh@1FY{ZxB?L{m#`sfK4f
zrxkHeeP`{<mG~)NK0`F`VN(tc-bLp2#@0l5UYO<tzRixE2YGN!sWxl0E=R*<gw$EP
z^pRs%rvkp&dDCl6dA^HWIvzvm^Ls1tvHM*MBt~ghOJYUW-*`;f3oTaC2i6@u18=5K
zk*dq(H@Gv}tq-Q%YT3pDVq$lqx)OIG#18^HyoD!sLjqUQC`0+1TR6CO?QO@o^u_z1
z3t5diLF&6YOAWE}f(j(yI5S&6JqxIVr;fR8>`w_}<72`8`?oiWK0Xch#c{Rj<PsLr
zRVbHD5=Uef%2{4&HfB>SBFd!*1gD%cNfu8Ajh^c`{z2JG*CZsBKDk{Q_f_#<mePC}
zHXvJqt?K2_A{2tr_H92BCz1i)a~xOyQPs{I>!k@eIr+vYeDXlp=W`LKImY&g``VaP
zH)ll_8S^AQW1OaIw|Hls74d{Mh|dOXTx<%HYX6OOnVG+BCX4WPPEwIdUvhbg)emMk
zamV79P!gtowS7IEkb^FDI#}>NL7eNKwPvfkCXww_e`mBejWW(TBWifT)H1Z$ApXh1
zsW5!j>_D79K2v`rJ|{hU2za~i;zLTXi`7B|vz%H`Z|>%boKLf5>t7&BE~!lMlt3#B
zAnrk(e|^I=Sg|-_kk+B%$(!O6tx!l}2B5Nt5N5vEZwS;Txx%m*M}qs%FY~%ZKnu8y
z9<Ki%Tkjl~_rL#-ujS=h_OfkUVY$|7*~YSURm--QTeiD$E!(zj`*-bp&i8jad!KWA
z{u?g5o)6veSSx07$*D<%Qa^vQJ$koE``i9R1Qtu@sIguZed|a(S7JW69betSs7(*+
zaIT(U>C@1cX_%d&&^7@I_&AiQ^F)fG$-o&YdQsYuY8E=Vu%3>RT!az@yRQ<;H4~On
z5=_ywBlbrA&jeAEgjbWQpSodNFI1^obobp14QoT}vrp`2nBoZQ<yKeVwy*fIX3xq=
z!~l;rLaNsUV76&{&jZQgOAR&B+pSBx`_#M?ROIl|_&}!yTa?iEJg1NEed|b){AUBS
zoTB!gnz$p!SeU81ZSu7D9(PXmrHtIapb}-sJwJ3mZT#ewqJy22wnQfjrx!vEtR;16
z(I_Z?CI4ye!C_2SrC6q+L!XE%&y0RXtkt`Q;kp0rh?5kbK)^1W?k!3&=Y2Qj$7`r}
zkERr%)VGG4M9R5j*W&(^hmag-3)x;x7MjzP*=i{J@K5zu)?z<YKWwCXb8Qm5c}9*`
zj>*Cil$-^%ZKdogNv1SRVozp;Iic{au0Shc?|Fh!kJjD^U@dx^E@3W?U3>*hEgrfi
ztICgM=l-a$zp95)wpp2;+TwiqM<@d`Ii#Sz`#ICyo_;8g9GL>AC~S6HbTR@<OMA#c
z52tysdrKS%`yWpl(E#VE|BBpyC^F-XrPB{7=2BVfQePV@^Ard81DKerFuO6?rLjV1
zRYJs=|5K>)Z?DdEZjp|-tTs<~nZs6-kP$Pd+*!Jq<lxoK3$McbSq@H40*R@*qJKtr
zaOe$w$0FRQwc)rb=-ZrMKT62$*grEp4SIn^mK?7zDu{ifrQ^LV_(t)M3in@VP*wpN
z_q;IHjEJI<2fET)N4~UzH72VPhqMq{V-y~qJEgLPyr1FYf5VseqFEKLWyWzL)m=UI
z58|{IM~BO8hcAquBo>-N25nx3Ijrxqya-jRoB!+c1>4Y}`TX3?FU4a%AoxDu1+YwJ
z;Uk#M0vb;ilAhMLw@fgo*YvDhjE7nGtwKI<jx=Yisj+>lhD|X~mPl{adM2WGvv9WV
zUVjA(l_3VO<f>VAz&Xsg;^#AqBxikGIGp2dqmk0bk6$vGCvS=*Juz%6>sMQA>?FQ+
zefj&HL~=nr#yht>xF}2Y-C;(BlS|qh#*C0T3>**q@(4$C%ZZzZFDtH2`eWmV53mnD
z`lf#R{Sa@Z{Ls&Of)9TMEXQpOTsVCN3rJUx5Oy0CL_dA1efygJ)O@yja)|tkPH$}$
z^Uox{7Junzi0i(w$0m)}&=;RIi?+;lZ`JQKA_0GLFtWQRo~3I!Bu@MfrbON;9<$_9
zIm@8gFaj>(86qm3oh%Dv@sCE6ITnzPmzGL2z=HKB0Kj|PMrcuaWp?$iXK;Q0^ZJj{
z30Pwa5!680#JF1rFv%$)ep`9O;!KLQc76KbG!S90an%5#L!;~c(;rx29Kv%2mJuwU
ze2N-tPD(5_2Mi%gV=?NUfracFZLf`gRN9z?)5ndP6cjxf>}i}12KEZ_HZT4JD_{+0
zZ6uvXLUWXkS3Z8U+Oy$<yJruKDxJ}}<cF<?V~)MX6by+*_YfKniyITEn$_dc47Q$D
zu#MO9B17{E%9&@c)mlxvX5hn-gsu}p0QqZIU&Z+XVzXEg&EI!~yBKZ1|ENUS9&Z~T
zdwW4nVo=B@ejy+<Haq<Z8T3cAowTye2yjSz>YylqZ=}1lc9Cb)+R$DrcE7XNZGYh9
zs?;UE%b1d<PljTtcnqYpdEikbFVyH?PDg1!?z5ZWanN#bC^=Hrp@p*XZQeWJ(vp!G
z<gi%;ZA9{Lg)Zs1nTIs^uXwFF1kKoEdSlZL4lO4E>>FL^N{cLP|6sR`NK1G3-}N2=
z0O!0^ev`tdOybt%mB+0o0k?>w{oxy*pS*UO@ky=cb2V%$>iB;Hf?dHvHQjHu{i}H+
zBbUbB>LCD?gv241(5E7+QkR_`incH8rJOF6c?-&+G6W~A1NCwit!32hJBD5=d9p7c
zeaI&_u14>tl2qPwAqbf<B!n-Ruiv(3tyd2+fL3m-w|ii@GRVj3S|tc<g!i2brc@lG
z+X)%F2@GyDURFPsX=Jm)!B&axr+|$TK%035D5jQv{=o-P=*3>1v^@frwS4-|!D*g`
zJe{pJJ%LtjM}HLk1HS?TyKJulaLQ2UDjxlti4KNSH5P5Y54DiWf}hyEGPbVIs?^~P
z6{sJ32jq=EwdD_wIb?aOq?JhCl-lnsIK<j1gFPI{rGA*%X63KGJWi+rHG>bINUJ|x
zN71Ej&a@X#CT*PtbC?HqKP%lz-yIxNj9KS&XN{y$UkE8ysQ2qLzDGgkuULrdyT4~5
z*4_1x4ktZH5i@*;qV<?R#P&=69~C)Sm{3TVnT(<4_Z0+)5707S_4wdy>iF*H__1<R
z0xmh{7LlahtNF(X>F>4q8Q{8uNQTBAz%6zYiEhjHnsD%XyYu1H)6)VS?DsK?^O)5R
zJz!1oj_K$vR-6VY5`8d=<nWjcvG+)8a={Wo*zT06oIRH|`x;Vo0B@BD=4R~d^zt#0
zms^p?c|Zd{L>Jy=g1pYo5p&at2ePYbeQE`t{3GUXa%+p|B%9aG4dgh+=%3FCc}L&@
zL7zLqGSDu@Lt^;`RkD*lq_yd6LCiW0wMq83>zc=@6taP$j>h3FQR#)X6TjhRzL(=!
z4H%VH2~hh9Hcl%E@c+!hfnR^()b6>-;c`{C7p5#X2(=Aa?L9q4<T}xZ2d0Mfr+s5e
zzm@aeOXr`_6_I4^p+G}{vIboWZ=QbS0`#66S0U2ykYOG5-2u)7T~;1M21&@QthXXt
zPcyxqB}n-gGCs3%N@6$^71`LfW2kmoo>9&Y4edPWGL*aKiR_XjqyDVZbZN{KJojB{
zamegjr^^Jf9{Bb*hor8Iu)@gO8w-|jo;a@EKZp_vvi1Cq{5y551PPhb-xes93I!Pn
zEUFGxUHjkNXeENACRpnT><qW(rC!?#l#GoJ)<6ygYRSrAlaMyWIC_$1nRrA@1Ki)V
zS=C^!LI!0NNGt#<7=JF^JGA-Lw=vQ+b?Fq?{c_X(*8B3@iQT%iY--FhV!%(#v*T}<
zfg^+!JSL11W}r^|B)ICbp;ZAhG_3l1gkrsP`hxshC<ofWMMtArWw7|=w<nnO?`y&N
zI|Xk9Y^wkgjXgn3;q%mreHZKa&c4BNWmait9=F_H6&}Jjo<XN|z*9Yz|5<hyDa8M)
zHoJGAjRF<pkGH2FZ(IC^vzskKE`#W&>zTc#=1eWHB-me1NZE!IYR@evHsoD)VY`|`
z<v9r`-)ac;YKlA_n6xj7(>Gn!#ZS|CbNjUNofLN+y72-0>#jNV`cKwT*jPe`qB?KS
zj|f24;FrtNc2aoM*Q;`>WsgyAg=Ltf_U3E9r|k&r^uxh!$45ER72W&UiWRU~LEhQM
zDZ*L%{8ljjHu!z(G#H!K5qAX(zpV6?bIVBXYcx2`2?7Gq{GJ|M;L`t6aDi=Luzb+f
zxlH1Gxnxc>m3JPmPZ<yfb;`9kNe;CamXP-6m=3-xNe&)c`h_N8iv<3bH3Nn1pkqxZ
zrEyNFc71>_*F$@I-kkAbK}OT}K@>iel-_x*6Ui^sWk?JySW`PNw3E!OG#W8kyZ-Gq
zC9u)StE{<}Ob$`24l-zs(sjR1$g@84!E>olE3J+Ru_ZvwYM{7)6iMX3I)w|t==T-p
z-M4%%h!X7dy$*yFs~!0|Yq^k88flUI;j>u@Dmzs5MZQsz$41!<5fi-@+-#}?dLof7
z>Jy2Z--pkY%E4c?Dp3hPlsXQ>lluJL)mC^@b4gy3CvWT_&eS}vL}Wp66j4hsLqiA@
zvufm@>(kNR6}jU)Nn+8ZsZt)7JG?;Fb91*=*k&APQ!x#8ii$DFRz!3)u?ek1tbRV1
zk?}m91dGHk{pFmN{!{A(CXelZ&dRP^AT%2s!Nh6h!l{xh+ouRy_?(O|hZKQSo?p)?
z<?sHzI6B+5JxTTjHR@L^&8yhi0a{65@2lAl!8_ZMMf$aK3=7fH$>eOvsZG_|R-%x&
zi3mNZbJYA1op!4zzL##P)y>}-98J-;bVvTa4aR5WP?y)osU^U-5KSHwH8Qorn*N~Y
z_1*wFEq+ZcEwm^64My0Que|8!7^SoeMhU1?2{Wl=*!<;vNJGc&eFDKb9|rKMr6gZT
zw4^$76H$YSRgIc8YrlVsWsZb@)TQDsEM00^iev1Tp*EyAaU<436_X?^uM#6g+gd2e
z6<_xUijN&P_MEwj^z#}1rwrf^_KVgLQKr7Usv#L39i9JEV=nC+N|B38&iCt~hPIo6
z9jl$lvmkA~i~&@4T$bm|6Na@09%Oelvj(0?6!zsw^R*q=V{~BC%jF9phk^dYXDF!A
zJ@g2+Y#m80(5%uVeHsT&=B4a=o~|;)q7wNrW}g`ev90gd8d2m{LL{VR<&{0`Scs*h
z;TJ{=7_dVaRg#2_zbCNAEE35LxX<E?VoUY1<ny$#P%#gU;>HJ6=u|$zVODW`4;Yv(
z|5D0By8xGnND~?+trXtO_}!z<wu(u~LZ}cJt{7Mks$mh=;u6K>rA~m_d~TeIO)QP3
z3*halKMb5tF~vHpa`?N9C@0Ng5nzr-hhOtKuEk7xRLint5nd#Nl3cRD?L!Mm&hn>_
zOxzDBn-JB8POo`K$s!`SErDqh`_7GmlaQgm&P%G2nlaQSIi49$@8pwO8yO82WzlaD
z!Blp3PLE5U(XW#B<LW*}Rdb__o1x-Zb*V<gxh~}S&<61lmgCq`9|Y}Y7DNz3S1g9+
zn$Om*;)$2$04~}@qVT==!+F0N&x*>WVkyvBlq4@<`lXsADS1PbMLNvl$SL+qb7L;)
z$H;-ZOu-1`)(&>0V2M9`e)>Uu!rx?j0!k<n)@VXSBT-*&tU6o5GLD<VM3U+*Osd&e
z4J_i~{Qg^VYAb!*iUqm3bW*?~haKJ@T`HIff^$YRR#~`1d8HTA3MQsl{j4Q1_i0*1
z>uxvoj=(bIum0K_fpEv*R52M^9Aps-*9{*%do+la#4U+>g~tPGcd<K?B);j-g@y24
zUttI#kW>;{0(P5L$taPRi{Awqd|R}fq-NR}MB)Q3elN9{>Werpx77EPOl!itwB!t?
z(<iMy6DszugPSCxOOIrkZg4pB*``R@5J~FaM&kRgYByNZz8`sRS}iV#A~?j{gF+@g
z&p7B8NhO5sg4O#6QiS<GV4Lu4rtLeX_SV^gwES=ZgZUCFMx%?Shs7%(i_iI?1}5cu
z5`+a1nxAJnVf9x^bsuT7lkqFR%~5e*2%}_>T8x{VajfNQNVJyTOMSddX8nwN0Fmex
z3_OlwxLP2)4n^u0$>^rD4a)rT?}mW|o4O)&Qd(o=3w{z5-BweP$BFp#R46T#^Bcbd
z(>w-iK_?7S+aUQ;bi;4@!;St*3y)rbTR%UI32pg&;)r29G-MHVRsD!@vUkk$4Y=so
z`bDNZ+Jk#>_#cTr2-nA6YBf>rC&jzGFR^6%O)Q!MgK@7qE3LOORDdGA+enzW6C#y}
zvz3=z-T7)Wa5C097Ig1PIZcbpf6Rp-Zn<V}-;-+v<&mP2uvU?NqbR#puio$R?PwDQ
z+w-n&xl~gjhuQ54=maZxOTk90f9F(ZEd+aHjz_cp{PEUS?J}Zlnn`Z0jTu|--m(Ua
zdj!Lf^o@HH@2XrYNlQMcudu4qUVFLbuEJH#M_;-vpLb4rf3B27{UOv_jDdx_D+ELp
zc@O7=Y1Be)DWrCv$>UKSh0-K)D?Ql_B?_wopCV)Ic$KgNi7uKGH1#EIpNp1agi{2j
zVGpZyYplkTJ^>2#PVgnaAuyfS)AwoIW7O7=oCiXUf!nERtWsi$3Y*botR-c#=LH_(
z;uExMez<1^gYMofWvMf^OsGw-TT{40XA7>uNLZ<tp2P?H41B}3r-3<R{!uj1+dpo>
zaiof6-k&EEy`>d>=)7_v<ULFbFTG6`&MWQz4ksDHKRE0;^#ztlj&6~Q$y2fZZOvOX
z;A=@Wqsi5{S%h8R?AfOZNKg?k9V0|(8#gbJ|5-f`9p<g5GKQH1^72$fa4RMMgA;jC
z`3Kw+_}knKl;5^NU&VIAmOwn+|DpeQ=0IQ#SmwJ=^~5r2N(O(}4o2e0l%t;~Wg~VU
zJIJ8lxD6ztN6L-yI=sP*y(ycaV%9q&<J?x-5N03pyZ=L-a#k(Sy1s5BB1LTQdi;Lo
z-1n(X?E2*p+)vOWAl$XT&10TYRmbdu?YXYWcO2%lqZ1P%5KU}K=l9#_KF3^7{n?lx
zzn#2RehBI1r({c}|Ngz*{ici8p47T&k2_3?5i+3Z4)?ANVb4vmhnpmBz4EI;m_7WB
zd?J!gMeOwS_c9+#U%JE;Kx=Kc{=~Ni;r`wBx0@%=57sjy0{x3`4-2A~uelg|@MJ;U
zy}a)G@1BXE-!$%~0B!3bWaK#d$Y904VI-0hE(&8jmlW-_ek?j)^^ZCV#RR5HwQ8E9
zx2F>Psaw2B)7y5VF{;vlf0S84#QzUYc>=q&d$6;-msnoL1LG)63$w7DD<esDbs%1r
z48&+F?^@Zj3NMS8il^tzfY*5|@<OVCC}Q(-;rbd{^$wN(#7ZOf5;A8Atd10aF>Vz!
zt%Z5BnyR86P1jn+upL=TG`63N^F=agNJULgGVMksB{r7VVb(KSN5SQ?zkfH8Wn*r^
zQ5+MC^r=z*(#S_Em`SRNRW6rZe!C?sz36^V5GsT2DQR4XNCThq0(iUkYEX~GidIH8
zsMD4dD&I!iNahJ5ac`lcdsdT@i5MbLDF)>>W<!HlI`yy>7Lyz5HZD<-2g?Bum}-gY
z*c&Lduf7hncno;>&cGv};XaADw21VfB2N8d%;OH{Ga8*`OWi2>b<h&6o^~A=L=JuL
zqek*!;8x|wW4JOY|C6t4MtA~x0KKb}ZL|QsnitQqTK62+CB_&q@larCQ$U&QN?lTq
zbq4O{cY%FGa#imIL=ZjygbTdDCzDV)u`QKkS;Ik<z#G^h0ze~qnreB|vZ)T(LGLAM
z_O7T@v!;O4Xq7OcV4TIQ{U3uk?j}^T7-f4Y@hkBsJLeE6G}EiRv*j6`k^}-aI8?t}
z-=6M&3dy(^QBid>(?pAH%x_oqUwQ}`lXBnd+s_Q!W^Th$_EEm?-d+Yf27t@C9(Wdn
za%?aiypNJFwWeC(KTQY0Zi-QG>pTfEgFPi7u36V4al>WDC2zwZg1fvtbX8?_g<TA3
z|2RV4tYU4TqU+k|-@`LRpRR~^q@<1ga&6_d;VY8k8(zK=Qr_LO%NOG^5(8$IIDhwV
zfkMtux2N+bFeqP6JX4BHO1hy4<7Mq%hH}b`SZ5|!<$ZtCo4I%WBN_bbZxGNx^`DEo
zAMfspa|zx0e+H@idn_WHVrfhU_POSg;3Qil(@#A~oqeqKr5DaEX8Sv#{%hK~ZcF-A
z15e$VM3kb_t)g1c3+|ZEKbo+-n%pj9UbBA<PB56^Qay$y>`Vb;S|($6@oSwJ!Hu+m
z#|k;V%N2ZLiLBwQy@0br19Z_*2#e9A6fSxnnU3mFfD_wH$kD}M4X}IZ8^bsC{M{BN
z$@|gnCHshvtdwh6setPw3MCobt~jeu@*5C2LzHTckQebi#nznf65Jf{u~Fem4#rw?
z_ulf(N+OT~e#i2po3fIcI#wX+89sB;)RD+MKCCdaP}3U)Gq1WqIXKgxbbb=NcH-r}
zwOl9YW$#epXwsB>RqNyu*RYTeY-!W;l-(l6Jabb2`zhDJbgZb7UNG`BL$px0uhjVU
zCORpQB}P2srFPS));ixX+i7BAVwvq`g3vhB91PZ0yXyX*o$=yLjHnRU$QbSVkCE~9
z*zJd7(WJ5D-*;?F^UWGWk!1X5Us|<i@=lr^fW>TUzivkq0``9W|Gvd@DZdWKW)q*}
zd~V3VDSXJ7sVDH!UDx>8AO5vp0%ZE7nwc9c;VRTPH~)hyeU7nCl*V77uM_J>#n1gz
zeD~GRk6c{RSa9~KO>y&I4Swm;^2YNey=&3$(Wa;6_9ooN$)x1h-&d!n09{W3mOx8h
zwggk!Fhh(AjP_BEIoTVwgr1x8Hj;~`iz8VlU{}|3%%elfDoX|F=R!FzMaR#F9YAyu
z&GN#dn~%bd=v+JdZ|{bTYwavH1F5HZEn$Xo61{4ZRIC*X6F_9{jk@H+p>5YFP6Si3
za~erRn|CxvUld;4k0o7NzC1Vms&72qnM^onLnm<zx^r{RI`i2}7X4k%Q@rcMSI%^S
z^}uo&YTF)tF=4J{TJG~v1CZ1va%y_W_f5T!t%W!v$cZ&wqTTqdy_GGh{)8Igx$Wxu
z6rU9X4j$fzIe^8V#1G<QC&brow~RUaeZcgL%*o7RtEJ}2%<cQ6+L})*w0eGU!Gk2I
z(F!CuaI8>pqUf<)D^I7cr>!q9ZHsFQYYV1Pdue>FO06#PGv$x@kHyZ)o*XAR@hf5Y
ziF=jz%}F)4@u~637|Hy5^}LfV8G$NVn1FDK)xJ}{lt4(6V5UtgyVam19yJTRvxecg
ziPxjtpYO);Ca5rZav~@X{OzupX2)EoJE(-zTl!GuN4yhapK}>v4|}GrHg=Ufu9SjI
zcLFM2m{MfZ^A>9E7)F4ZoT2B1`|i9rFsnP6o<G+Zcef<aZfiY33|3JxI*!VPab=G}
zHU_42>I?A<_J&?zT0PF(fOp^N#@jjyFdKp!l=ojoF%oR2SwBwH4rc5n1aWXXnpBE0
zC#vKw0kw!SuX8~BItva&Nv2x(DT8Kb;y_aex|E+kzfP9I$idL%4GrR5F{lp=EY5#r
z+Q;cj5&LL1lDA$%Z4~(>T$#YJLssw}M^KpRGD~u|KrJx?Pn70s)T?Y?z9#e4;|L&C
z_Sd<<8?PV?f^Hl*$hJ`0yDw(6^^IJy;3D=GX#R|1<Gw-)+@l3WL_DG(;g4{>##>?*
zZXTKM+-yx<>JO=YTXcKY&j@qH?%8oud;Qt5Yj@JRhgmDEPh>JZmOfe!Z0g95T>M6i
z)h^Jt(phOw_e)UH4X=#S*!;Sg+dNfM`HqGo(;7G&l!)SDUH~D(wl?Vzor|9ng(LGs
zf)r+HMd&%>sw;3rJXBkdCiR{X=4;HCA9(T3-}p;Jcj@Gh_%?KNxctC1p+XaHQ!ba2
zAc_}ew~BP}zUw_LNO+Rpir89YD4a{>$1tjIpf0>GvYqK^6H^L*b+nc<P(=#+qRDca
zPFRkWMH;6BwT-Nr`~Lo%`k_9%4j=r5OsoD{ho<qpY0B<2Oz3Q8B|zEg5<3^OBPyWP
z^4I*RDrQ(QNtsu}yr4!}S-B^_AbX7?F{_lCDY7@{Q~u6fe<&b`;y6>FVbfN^L@d`B
zn@)PtsaSqctKOz4&NUr^T!$dSl#AHcAibD!5T!1FL0|G6k(&P!_b(eSOlzWEfQ5?v
zupANa)wTssLl7NC;)y%NNQU&m56aPsBdD^t%S==#E7@H6Me@_q5RMvcmE)swLmIph
z#@^Kgl$R2-)qTB>SX=%1ufcD)XS{B}CX~5>oi&dVLnrR8e0gbTYcj)(waR-p_ooM~
zJlJ%tORITCgzweKy1nI$t%wy?EqhyCatHA<al&RZ7t2ULo%k76T(|D7-QMe&!yI{n
zzpH|XPyDn^8Iiasl)A4An-?RXbGu(3JQf9)<6o;Sq0=zdM@EovgTg6f@T-4QU<5Ah
zCB{w3lt`O6>+>G;-^mp0nZ^<NwC`=3FW_-!gv$raRN`_?J>M>^WiCJVeA@`MwBbmE
zcYGXjNu=nxFZB8E1DWe1pZR<U5o^M%P!9GY9Y7QAas3%<QE`F2GmJvVe^*{gk)#96
zMrYJGxU^e+*YeOrMa88%cgZKbHMh#IhWtroTIhcrQKab=X#tj%!`LIi@=yT|o+2zv
z(~q`%)vdp3>Jlsm?WgL~YmY=D&=es5*J=I*p<-)^Kq#0*n!GeIgjPL7{QMn-IWP&u
zTdWPEJJN;2H<FPk!NRL&|G;tnA!Ar}EYLId>T*VT!DjASLooR>HV5keILxIm#|Dsy
zcu!c-+A#u*OI!3S^=)OAHk4Oe4CVbgNnK(?d!r#~8_9eAKrjcT>eqL;k&#hxGP8N~
z@caMuLSe-ro6u`pTIhG>oyX)%OX3_k`3$t>C@qeT5*myNlG8p5=fM2;8O2AMcjkVf
zjMj=dY3Q;Vw6aZatRx(*Z<CW5<*UZ?ui^Hz1=JIML4k0C?AJ@|X6AK!RH|ka|C;bL
zL+l}Je*i5}B4nWPzdrftj(&D#$}kSMIdi)tVSuxA*+BrDs;-y9<r3iPWS~#*DsLOp
z#$De2ZlK1^EzoQCiO}7bMqUc5;-2wG<1<*=rMsi+yiv97gH3$qYw27cyCnEz%tN-r
z`0*^C==q0^3nTA9?c|r;KEQM5W9!GR$4!9(=RDbnSL?^=5ey0JCVEaIr#N>!?SH(w
zj*W?qZ6grTk_<nWG4XZW8N?thnzcQ>*5&R79Cq1<1UzB4?WVmxt*Z`gw!rCeDTy12
zU6a@&;>WD*L0GML`e8Y$Jr%CliVPiT!OJsU$y0$=Eh-b;M?XfMn3j8Wd0EDGY8(;M
zMAbOm=ao6~lB0za3{Oc|Y6B7LYf)m97>Kuvo(CiKM3}Rdr`-Lm(OI`FT7jzz6nW~Q
zBB=+cI#9dU5Ns@0+mi|saed}dxx0}46-LETxzehK3C{y~6^nEm#^Unl&lM=`m#ia?
z?W?a3+Zo?vg5;1CR{y?QqGb^Eog>2x*Ooz=>hN|(!@6l-;}zIqff|`;>;eMXp#9C5
z;@{2Yhyq7aa}-Q%4<Q53$F=3NX&M?*8P~ydabZzRlxq)Z__2W3X3bddhk4WD?jxC}
zbt=Im;NfU<obKCzja($-@NFLPb})ib!j`Gd<?*&KcBZ7D>(k!#@$B5nxn$#t4eP_C
zsys#e4YtjfafNR|=I8CNhncZ97=-QXxKS3Noamqef!jqRQEMvTo0UtrKhfva>x?K0
z;`{r(SlzcGY>Jvo{W4*KrjGvZ?6bEglrwyP+3pa$;J*&eSI8rLLKfe7h(t6TCY+@*
z8-!)oOt(1F#;?TF$HS$-7$?PwK~9idV^DJ$o8f-OH`BN^uN?qlYu=$7mjLg>XoW|$
zWf=aHm5a^O`U&qofE5t;3?8~jHV6pZ^-9!n`u5K)KIWJIB%mY<#>^sqNGzb3<3Hcb
zn8B;Re7O(HpwWE(LLLR|-x+AOd2w9Ik@=Foli$(mU@7(NV*@NM*VC^(1~Ei-NmqTM
zb8q4H!x6?N<cqER>G(xu4$BDoQYs2W>vp3FpjI8QL84RXj{58U1>>0&LQvk@Z7m>!
zj7&@3_6Os`qmp)bPspkT%1>=2GtnHK&1q6nQxk;OT3EZ@@j5u4GYBQ6aThOb&x>C#
zRRFJqdQs^Fn7n4=t_M%YfEO9tR>b$8?}mpi2i?Iwea|A7*L~WlEl0mH6;obNO2qqr
z_;eM+nI>}MPtM8>-ByOl(^Fy(b|f>Hj_cc6M%HlS7?oRojO>z?Fj~rR8LAGj(Htch
z8Ic{(O@z@l@;J}K-k+_18cmzRB#g4y*a~wPs8iJ{j?#TSG!OeC_*ldj#}+#lXVAa=
z`cUDx#n8doFtq&%VRSGJ3)BoV4(kT>R*mq`HvPqZ&eIgV9u2E;AjHmxFW;TaW*=!`
zSb=fvt|c`Vm!*Y@iqf$Tnry#zToim&#Z;5DZRyoN-ai>FkvWErS_yAKU=vPVzQ*oD
zXNcK~Hw=1T{vwJkl}uUw6u8ahIdTIpKG8)4EGidqZ-Fi{WV7ZpFgIpch{e7F-}hwt
z(Xryv{v$#rwg(<(BgRJgYr8v7(;LVhqQG-d%Sq`qVO&{cT7>ABZp&V7he^}GR`=fD
z*YHXka;r#x`-&$cJPhLFDD=h26m5ZcG*Q0GSFiJd2l!J8oLTKpFi_akhz5uPSMVrj
z843^k1#R;jKV<Xg+P1m|ELLB$iGCSrz|4$!9~#E)tz0KVK?Y(%Am-Gt25+YgdEi;6
z8m_x-{i^r^zJU0#TdaCJrQ@YIj6k*9k+W90u_5O1j0E5NI$}>B9HwfOB;HZ1DuXK-
zz<mTOPDa)Kc9PDk1_wc#c>NF0pz-u%xDQ?{e1frBj?ck(j_WM0Z&zGxR1bUJFGkpN
z?`x{9(7GU8YAi9H?+JCk%!(l+*1X+&c>gLP_Bxlx72McRrCfVLl2?GVkW-Zan=fsr
zA!@cTlc$2VBzEJ@-g;AG`vk5(aQwi@HLPtpBFmJ1eJmp$SLXP2Ran9#-}clfFU5#W
zwm;mc%mF89h_FfgveF!D%73aq;L^`2u9*jB3LxZ0wx~_WNsN2n#u3G|>|10;4~Jw6
z6Aj#9Z!cqD+W~iFbUB>i_nN&9HQPBGwHTXluwTk-EGa9z=Oi!-5T-w!bIbvhZhoti
z9HKjI<V)yInL-K3PquB)M5*{eUwU8My?xT$FP9{GV{SAO**4wTchSA7-|osWT&)Ag
zs%(u+?zQ}fn5L@)k1^_1a6H5<FTH%p-45cdqI5fOtXB?d5%oTAav{%5g!kG1JM)gJ
zBF*=GW$@OD)+3f#NwkZRzuQ0*BnVnBwsJrLZGJ~8E>iM0`T=fZaUO{QPqC*yH@rkq
z(F{_%ZT7nLhx&G=#y?I}jf1@>*oZTMC5UNMMiiS3Z({)Rw+)w9X`~M-%wEp{W3txK
zWpZ)^4S6sA9lExTf^G?x0luR)&kw0a>%^|Wr4jkBtR>ZXr+}BvBclXEF;#Z87(LA3
zlOOHZmv040oTz=73-RDcNEMsZc38#=BAUCeXugPPYK<uKDkay0(<Utu5j(xUl)h^j
z+bVDU^sqzxn7<t|r(NXtmQWYbHWL*DxUb$WXn#?UjUod-!?)Snd#`)Cls4{neU1r;
zT|P4fynXLpP20B?2I9(_rB%5=M(e6xZ*_x53tapnFt_rDu|HPF2Ho=Lx|#I@PO)`v
z54SN52rAK?rVPua%G;hdw?z%KTEJ`28k<@B<Q(5_<X5AglK@_Rc&4aZIWUL~ZY=yP
z<k3&MJLrgv^y~cbR{H*@QLXE^#Z2LoNv}7Wv1`DdH7fb|%VxFk;QG_=EJ7fTbI%4(
z&c>~wJLd*CMdD#Iwx1O*dni=il)c3XX9$@OP#o-Q4>8`VkZ23Xn@IK3Fa2Bq?VZ;t
zt~S00^#K$Zau|Z0uHu5my-@w0o2$~t=SAXH!H1o<mqmYNN|kl5ZOsg8y@~&d?L`p(
zDkfuQBGhpmhu&$!Y$Fc3EY-0DCc|^}S=l;8LLTehZA;M3pRm@9(!$qVGyt!s22_o)
z(ia|Vk|6kC6cY^T<`8<4-)Eetm-bu25i%;?&zrUt7A^~#Kx!dYdF<r#o-ro|Gc_rP
zE7vnEd+N^V@z81(c*%#qS3y6$A8du!e0#m{Y;dz*VP@B*qSLEqIHX;0516>H0KC5E
zU}er4uA&tY4Ax8(zSiuy2Rk_R1p$gw!)p?b8s**MeUct=)WP`>UCukqQp(6F2s1ot
z4NZk<9J<5`wAT-fWO-wluzy0_q<tf=-jQ@-yB~7miX*rW*H_%X*Jhvjyi#&*iTjja
zpj}Zy)or&$$b9x(iG%%Kp2s2kePo%>{<}LLdEgf$dX*403PhM9N{(0k9$3G`87!4)
zLHa=K$P=q5DJE%VgIq)E6vI?9@p-yDf(6=so-)mD)Lg;V^@FIHN5>>f?=5tc>+A<K
z3L$*LpQ1L?HQ>UMBHKX)iDXG7NZ-YgRMUw)ZC?WE^M9OWVs>VMDVO2`<$BE26B~r9
zG>8up!D>b8e8TwE(57X(hDBu{%y))}cKJDjRp}$kyDBf)*RVtNwDnyZg5GnUfyJhb
zZPX3<cJ)CHB58H!2ihI%%vAOZL3-*F7jq_TYV8*asz=4|g7(^hHg6vi3X&qPJe73s
zOWbF7!h35LyVsizdG}m`6?zBX?d7S}>E#7tCTGTnp=CVx2lXcwxfBoYpA8QA`q)aV
zL{sa+9H#AUOuqY#5}y#v>vDpb1Q+$L$7yPPgE`O^M;)6;6E!1spbiDO7hS*E%!lnu
z`)!|hK_K-~^C5w;qk)}p=lIC=#~VG7hGhlQjlK^bkk{&+Rxk2~NgBU}(<364pnT8*
z%Xvm3?1}bU-)_}8df%DOgj%_sgAQ0!hZ5WPntC-!k`kCrV~$nYbRS;k4vLB9S{z>Y
z6nS9%o6nA`@x*Y%`x|IV1mfd%U~%euS6&~Qn|P^4>x&JNe|f72{|ftO=~Dnv2WC?8
zl<X1yHj_al#T^U4Z>sju=Bpmb^1eZN<9sMh6=v$<rCMqRr6ct5gz^UI8;{~Mc9-SS
zNg;m1jaT6bXIpx#ZJOmXGaX9dsUM8^UDaer>}_QaiJRB8{yZ!1t<Z2+<r!*afH<vQ
z=eEH{ob<c{Xw$u7&OD4bF=u*uI8(5zGQ~OHVnj8BZ25YMH*@EwU=XQHRYHi7BjX^k
zyW7ppLkD@SH*bH!0SVNSKvZkfpPsIo^o#jn`1DD0!q&DXfsrlt3y77wAyHdxA|`=P
zC2A2zc-c~Nr9<ttydnu7zDI6=Q+&`w#PK@0gqZxsRt4#+-0(}~qRD8|mhbxf3FY##
zsuJ?HUya1Mfd#=5Vq>@u0uQr$X-l}*^{;<)c#9S!+7gTKem3|s^#xNnem;T?-5e}l
z;na|(Y?j<p-}dHbP0P<rf*o5%^)hkqY4;kS2SK&dRX-I?g_N?ACd*_xgbZ!RzQ4m3
zXnEUm*>^5Jdwj&7ab4FFOcEk#zStA-0X#0b+{D}_QD0N!qv^nUKR>z*j5xiN=+<iw
znPquzS=q^Xj-82SJ_9cwUA%Hp>deGKs?W>IpyU$}&_DqZ%A9PwSrP&*(t?v=q<%5#
zj{4BHvY$TkFwIaIA#CWrOg%kyFyq<bfOC@HnCe!Lavs35i=kTJvZQHeLU464h=A<x
z+?oS}XHLT5I}ewH6pyc84~5r#&w`cKr#BwrvDsB2yL$N=rKuQY69}hSH@xsc@s4N3
zy-$U?-c*@K<zw=I>8+=Jy$kf9&Bj(sR9Tj*;cuI*ni^K+=fzy@tE{lNOpgJFSk0Z`
zU|fzg6cH}O*ac+l<z#F5tw$@BiGJ9?=z*oc0xnx_eC7ItXnVc9a36kEO+)sM_Of5V
zevA^~WjNJdl%|61=SUu1l{DB7OZ)V7i+gEE#1bzBcb@!iH%o%^ki%9k?f0!~Sukhc
zwtG8VY&naDHxD`yQHDw^nqE*n!4o65O4jQ>fn#t9)$we~gWv5&&2id7q1-_xF7_WW
z>recew#f2u_iE@>!2z^C235#=9X9AJe81@n8(ly|4%{Moh=$!KnJK{7?Jdo8-C$$?
z33)%fxdqx8B7oe4Eig628GkyPx$l8QU>e{GKE~UWVN{zwdxu+A(gW9Nw;Le(w0les
zH>NcczO}>&#K`Zj)6QeYlaI$=j2vn(1Qm9a4^ko-CMM7j1l{n6{yu!4M8Ez8F49w1
zl59KZL?w;puqk%11Hv2s+EiSnL!`l33W7P!xRVs0YswAQ8<CvAw1wL{FcrLhUIy}7
z*o&HsvAV;KXuy_T#B2UjUqgy5m^ozGHrz<xS^qnxjMsDRyuZZpX+6mLjljr*z-?1S
zkRXt94T*KG=20mJ&}QDl<z<`J+4P*V>Y#?BYw6MD_G<xAGMuNKLIGk+{;xqn$IDkp
zHtW4Pl7&pr^`wlqlXI2pk5l+^hIF}JMqqH%B=edLjXe(F?PP2<zU6Y{?1TuN$=3qC
z?{Ki)d&62A$-vkrO3mKH(9Y`%p!*@b4?zFmvj<RC`S693I3_CKP}o7e$hB09a3m{J
z^ah82to1@c+HVMEUZ>j_CR`Pl9UwERQ@Pop9_Rh`TRXHX>+v+sberT^6oXM}y%eSd
zEh>?XY>k?D4C<A5^~t6QY2>o=mPVxv=c5BI1oPT+ZC@xU`<wuZuYQjT>nH^&*#iX8
z{N0*!kdBGo^&s$m`8+siA4z}t^j5tQJ_0BpJURlbSr+PTj9*I%LsJ_usPjQ7zFbMA
z!IgB_;THb{X~pPr?=?ZlwQ4toFOI;~vjE|#@>?xHPi&WnMVi~8yHB3wTy2CT9q9FT
zWuy4E(%>a6*6ZpPsud6cVL#%>D3v%04~R)Qs%k%Tlb@)O#rcS1HWq&dU1iWmd`-<(
z->A=?#?ugMTgL8Kj%jEmRsX7R>rT-CG~5?x@DjhWcQ@zQphpRu%rijs8+!PO?Rm15
zFpJjc7E%3@8_(bbVaRgEEabJHnvrimf8s(Eh$-B5(e)k8$>;ek6(^AUIvCB5?4;Gc
zj?*-z<(FWUG(7!-+kHYTF%7JBz%OL)21`Su8H|T$Or2>2z4lK|HxG(?M2-W{*+f$O
zH}R}tjNeKA<E#UxIz|V^D6j?AeR#9m$?7{mvY=D|Bh2{aV(~Hehqw*dM~uDnC^F#X
zDYNif7vN^=1E=%L&D<G5`}2pWZ#aT4JA82$7B;6c9zEW#F&W&Wk1Zl?-jln-@+dx`
zurMI4n_&sS$rJOBnwwC%;Dt7$5)#DE3)R<FW9~V~;Vh`kyg2OkSbilzEXeoChcX#t
zbH+niQv~93nQa>@?oKRrrCPu{2b*bOK}hv|Ta)dGvo_@l?|O@3ALFcW;of36+i~|2
zEvgr2mT|9Z=#3@D?uD7Tu~wF+e#(0-qH~+f2m4Dt?}w-WNGCm8#K?9P!B0;usrQM)
zEG$I48=O2bQBn4<WoelgOLHCi%G(~>seE(&T<(-&UoL-3W2qXqYqa@vV|A`Q1)NEN
zK6f&i=qbI<+q8&m;6}Hf<3`Ec|6uTQG1L31=l_zjA719X%MIx2d0z^Zji9GK<646g
zX%lbrQGwL1l=kJE3sBbl2+_Xw9M+EAbxbru)r&-|To;w~Av^2s4k&vAr%g~o4!HeY
zVBx0linCb}ivp_7=@UZ+0T;gKK7VcSIx5gDg@Lt;r-2z5+UEPhjYwb??d<^Gfr4mR
zzG(stpRvolyXtvV5C?c@)})vABq5?Kk;LI&-tOZ#>5mG3yMA4Y<yy8c8H6jt9^yC!
z`eSb4_#K{6w8tOF_n&xkUtTs>c%|aS<Dq%~osLE^kSx6qPX?UiLBfJq?J1U4)`t>`
zBYXG-1;#$R_XeO^gyd%zc~623`8B-J#I?%_+hu(t@o{TBz(aLftnda}N<?Nt4JMGw
zETM*v)r`R&hf9m3qn@VResw<nMD<`D-!(kMa{s4--VX~u3xkJQWK&=yysmYL=F<GQ
zk9}ZFK-K^)g7|y37aa5R<m6$*vXe4fTrrGQ+?`j{2fwl9<9IxF#An<Nj|Bm&-u~(C
z-12-qTL$>&e9fwJNVh1{hFNt2NCy?`afTAclANfVqn3@J4O>~26Hb3(Iwc%9H&{Xy
z*Gf~+zK?MGnBSqUO4T^PeEg!fU;;7{6K@otUs#><*-e&r<%)DsL>+pRPZw5_wwz{X
z?|{FpKxTf9YE;!uhq?kgS05VPkC@C{Y%sLt-d!m%iNF1x;IGlSZ8G~;Q>e0va&k6@
zjX)Edb2TDQtr)a;ISYJonK$c&+w04@7AsX1ufflq>s2;m@ttb?H9MB@S96q_c!H2t
zFeOgNUFKGP?W-l#Vdq&PEWa~(TrI;O5?$n6_w1h89t<^8=q1_h5ntyTxOTS6dLH`m
z`HVla2r;vclFzIa`K<jc<-Abc89!Pau*aEsYuaxQM7H&8O1hk<ANSPBC~h%KQ=W;h
z-<i)96vlx332>M$%W*e4_(!V;0+?lynGSkXFjyoVWrvffy!YD3tN7l>BJy|3@&DQn
z#B85|D$AgRn3xxxfzOk~HV3OVFgq+-^A(PueVp)r)XD!h5}p5sRBCf~)<lb;M@?23
zZWjK}Zu8$iCWAB_CC_1k9g)>r9Qxyb0SkM6Fga+Bg7F5!qJO2wf1@piNVQEkJ|%}a
z|E$;l?;mgv<C$4PaXe%l>d^RF_&(@Jud25^JY2T|SCMc-&Cr8}j>u3fY>DdODCH$W
zlAgCOAhrgr-!>M|r+@GB3&b@>$OibMxEd)4%3<$v4fuAtyZi4X+6&s}M&{|Z6$TK0
zj}$+WFO=4T%4pT}&k|lt23`#;N22avDgh!?pM}VwuLz)$-+XgxClUJkwoS0xz^1cn
z>7PjZk%=FuS?v53S-a8KM#B9m{ke!APMUt;xx$NnUP6D5@p}%ax3P7kypCU}rXlYx
z^x@rsjl4T;ZqyM(BJ9|&4~$WB+Y3gFzIE(ZVS1N#R{EX9P8pZ1m;z)}qgRZfLt7gE
zf>BYn9L_Wya2!yL>h0-<ATo`zv@dV65uKx5OPB%?GI>)DCnIJn+rSXruV7X*|99`$
z>!dylE9LUhGc0~zJXRQ-Q=#07uo)eU*{g1XY`IA{W)`<WUE8MZqVW7Ch_%=)CofNR
zWMT2aeMec<Nr991=(2U_5+Uz}%c)6{?1Sr~ApfUJ$T;|4;XV^VqDEg?gXv!BzvB=g
zN!Lh6cIKAmxW49J2v59>tIlwFjp|x)4dVv-a6iq{37(<u7*%9F(hyW?Cm~bc?=uYG
z2=HA-P>Uh{iulMbN)(4Od&{?x$cji9+&Pp*!Qt$9ougIEm-Epd(HQ2dG%`*07+{#k
z0!5E-^O1rsl*Ym^RBx;SMD*(tIHWZ^mSO(G#d<8AgU)00{#g&S^g;~-_Sr4)xd@yS
zKg;wV;hw8w(27}^=FoxvE?tR4P*dXv$|1#a%;L-#9Yz#sKs05h87fR)yFldct0TYV
z%#8PNq`;HZ3{Pia;f5kFjg-vFrC`x!Ejc8VFU0keQVoFH7v-(#L(E*1c=ryj{?(Or
zC4u-bWOSJ}yn`%_SH+*aQq1auc!W5lfuoNyXXiW%r(QBbJ}R9q>6%rsWLt#P0<QUu
z!kG2-0Y3ghpPdH|@&hx+%yBP4Y7t?q_Gf?)-hP{g2lbTK`Eu8BdQxfW34n=BHI625
z<iWWehJx}e4n};TzMGXe2eoEHg$XUy^E9H>o1f+Ux?ENQ=3A1rt;fz@?~6x|v+6_P
z#(HSLV-ZtZcagpjDaa8Ko&x|Ssj3VGr^_dax(%Qc0i9y;)M+4Qe3VlC`8=<yvB>|S
z1y7wuC#wi6{`b-MSj&WOkJ1w4?-SpmV*o~qJi=ep3n9MnlAJR?Uh)rnEYK9<oJW>7
z!T{CT`*`Q0t6&I9FDE!R^+FtVq}T=@1*J89eYKD4JlDYuftBO>HC{0>AJ($dm)+ak
zZ=-lB=Gu%tr^GXY-Q+Ab%&0(^>XcdCkwvG48?b`JF>^(txD6X<%*(pybO6(!`f5FV
ze3o!*47AkKLvsC77Va;?l`HyVepGR(<k-%HYI^I&=@9DVo+z{H&?Cu(9(W+yO5$qs
z1(JXOA=45o22?0G>fOTM5)+zwu`VcKFFMXka$P6kINYag*?dGoU!AW`HZMEl6~_#C
z?eh6h*8DSq+S`M|99@&0L|VDcPUiGy_o$6**zR$m+pi+Ty>dPQ?51nmH2F00GV^PT
z&5oTk3Z|aP(){{O-1#J%SW;jH<k3cTw&kcq!;-PP51hI2Yd_|meg6pFj0S4iTA$<e
z`2Ia@!~{QA-GNunP-$m*yF2&tz|tu;0x3HaIRoe1*9L@aOy0RNZh<B@V#PRSlm(~4
zmhU$Z-NP^Y$|IEUcH1esH1J!kezY<R(Us-?sAp#6D{tk<6v>!ho2;A>`sF~PWglK@
zG$tz1hvmYf$T7a-<Avw<XpcGkr2)-NGD87u#H7do96mJ+c7lD%7%K;EJuQVm?g`jP
zRYwDREft4rDcjF7zWJQt&-=7U*hJC8HK^a*{MXf_*MLHbLP(2IcnYyu?8=oN!`UQo
zG;s{Imp{sU+V77^2Z&)b`l2Om7G%u^;s9=-v3dM-e2M7OT}P5Wo2dbJGT)oMB7%Bt
zAT>pcmF@A{tn|=w7=U!07C-(FNN`ht8m|lMhtI2cim>T<OBSNg-xu3DhUOZ=7hmYW
zHi&$a%m>DFm*uDRSN8X|XlqT^pVD$bO@e(_7y7@HgXCXmf76As{oLI%$4+?1e0~lL
z?ppA8N927gyI80#2zXrxx&ofOl=u#8)w<`unFoADJl@$tfA5f*I-^<EV2{m|rlldp
zsFuccolAB$NACIZj!}MTXvUx5KK%=2dx2>!Fs(p6-fJ#BQyG=RcXxM}A1Lo?Wu=`V
zp+GNSbfd2Y!%|aIGcaxB%zkh@saQ=DfCa;<vh)6QdfJgyKd6w9XOWMPTAk!(Zf<TT
z;7e>z51BpLfWO``dBIp(Lf5qBKxDMa;d@jV%}6~wqiFVhOQwX%aBDz;35ysz$CvS|
zmdi^1R5mp~QfTe@n$DRAvGdp?BK1iY2MQYLUXv_YPEjGoRACdL5=e&-zXoI_iqai3
zI0``!J5zi?5`5eC+Temo8#(sCwh@Q%Hc<c>6L+S|RM~^vJx*zXgc={NehlD9-(AdS
zTXln+Aa{WKGHsCoujm@ux+Kn7w-DYBcFP|iOd_MMt#^c9Y1`RPQUL+u>l`Bzt%(MK
z<oIsgSkCDqe|DM$Y*u)*G+)+WMK|qtS8UJJ_e^vS#-wAk*(4nr4wN}rHF7C0qT+s3
z+Ogd+_zpg>#*zfgZrGRD2A=AB8_rh?SLx9M*+oUACpW-3l^v7p99h>acFTv$T7(W~
zisE1w1uZm(?ew2%=!R`~axmVW>7$%<+UG^MY;5*5ieH%0Ha^@m2g}%{*$+<)h`!%?
zBIzh2+*;XLD8O&4*uzS?3)7>_Bi>~?8;7N<UVd(*zzBIdh|Vj`IPjoy?@Z`Kq-&mG
z*n-HUP5DtGV?*}MM((8N8VH;hWdW8IITP{wWELnZts!Ph*YhK?o2L~8&6eZI;y)l3
zMA$qi;3|HmlOP&8L_QDJhQ?5Yed1{3uaA`QH*Gt+Wfs%V43aY68n%)d+s$NHeG+_$
z#pnYnPOkA@lg|m_ScRh_%zfH<N4afD*NZI34(FjEi1;11VyGOXL1U4Dwd<SL_W^Op
zvzDxr8CPZ>F>l|&Abi)(!h(*+;a1k(@xqKXooEpWQx1<;e}j&Oz|=G!+M4$OZnYkV
z6-S};a$V_mByyB{DtbyxuLDgUHUO;~p#Z57x&QaNdw6Ks$4757=1bOV(S$xkKTUCt
zeTk(gMoF~HYVNmc4;;t!%;s*3_iky-{Uhue(+j&&?|*4GeRM|=2to0i&^!0%C1n=r
zNYd<OkmvD$nLCV68t8+ROM#_w5>xRNI>bXrjDO7d!dNt8D4#({d0}jhTuNnjzFIKs
zR=|La^uhdPYnJPM@waR!SsWec>f9`{v-b=W2(#pl<T2bL+RxWKatSUR0hoR!1iSo^
za@34boj&B=uA9AmlL_Qm{=%wQw#HwvTw&&;{ZMyZ5WA#U-rE_|2%|`7t`DgkDbXU%
z-<HdUc8al|NH#bqfMj7V3AsB71$!9j{5G#w!bh$sqqIx`Lmh(3yjGWij<g%;mXvZC
z)@-E{LY+&41K+mM;TE6kZfxaB^As30*h*Wr_3aJ7Zc>Esx$B_N`YEpU%%2(3Si3dq
z3a%A4{vq}Cyh81}K*CyXpoMNYC-t;j0&7BhVbrG;TYdpzM!%Gg=_8T=i<|7J12gT6
zkD_R4k(S_n2o3aG{u>dbkM@fJ94V#8E`}ORMNzV$?y3n5#q9Ei<F>dj)votSd36b0
zho5mX_IA2pbc1Yc!7=aS440pk2!8T-`+Cr8qBD-^8N^N6>~sc)T{mDm3}7+FkhzOp
z_03g52vZhKXA1PGiI)gX3^Px8;^zFQgid>0PF)JD9bBE<Q~j1W#~DjjvtZ3&t(|K7
zE0=L25JG+_C^f6*!*ooWIj8zKCTT_krJrS@1o56Z*2Uww$7+sKWnblyWLG;m5&;qk
zAsLe3+9CDNo$AqUHazy@gKsot2DelAapIu3wAdMfUZvqpLTxQ2X1t-AA(>E_ttQuP
zNgP+JC<@c0L?>vP^OzJ^lag<yEyX6>i1iWj6zq#u=f!#vCS&VNJKws8`fwSqJzG%r
zrizV7N2b`-tTxkrJg5`-UA1*?AE?v(e*wQKK-Tri2(shhX*CHOzdG0nzhqPZ+>}lM
zYGIVmr`Ls#FGNfu4aa=LGaoS0sQ>ud0uaeNhevw)C!RXCmUyRzfV_ODf1B-s;s;vL
z3`PR8)q$r`#DEYgT302k{6Lfm7e{<d24K)&asxhysHTO{OP6oMt#xTTq5OmF#51T0
zc+#mNr~C`@1tZEjR?~yA4_gF<Sjdp@1wOcK8Tio-j*BjcLuT?vmW*D24DlAu`|z}~
zFh38oo6qp{d}&KND;;SYC$uev_sZ(37(pAV1`pdruD0+^Xf}oE;Vp%9WMM(;`MTb;
zXC)cR&S-+eD$b^Ck)Aw0spT8!$cm2`MXPJQgN2#t8Ldmk3)<&a!T`hKjCP&hr0vMV
zP&IecVc0CoBq%FmHo16C22f0lQb&#*T~vdSK9?O8ysPHyy|Cwce@x2=y{(J>dDFea
zh@}iMkf^8T{5xcEJ$_op+m=TUjfM}JXW>^Xlh7NLJo!4|_3q{HVV|8mzE3L(BX=3C
zJbV+{216TdI<YvTCoIM7US%)JkI$#cW%78RK7gK~3dX~?4Oh`c?xxPunMoHpz@6xX
z)sr$h`Bf8-TZYp#-qZHsruaejumS{*tps5JnVgnCj=b2pAkT}ixyL{CIU4LSf-I2+
zX|rUehf2VJfWZY8ycrf5t`Eu9BN@;DZ_DYv;q;JA!n_>1Ql$-M`AZUJsTjt{taEDc
zSj7+ZC{$u!QDoXpL|_mld1J6c-drQ!M7{{qKPZek*?$fAj?zox<3pzuQvG|O4X=CE
zbVWzrD|?^7P{TufJv}9`yr=ptA>z{ak9SMM!bop?({~HP*02Yf5+TM_y0s1%OJ>$X
zaIGUy9+DP}kN~7|3z!)uW<ZHx{FgtvuMmchIb+d%PJw{OA$JJFKZt}lg~Sx&aQ%=c
zbE-3akkXWPFjax(RcCP}-#S1RpNXf?{s(N@%`kv5YT^rR!|SbAlH}S~(pb5{@9$TT
zMHr%UNKyuZ3%@BDup@IguuMkAl7)pir3LfQ{<9sEH2zAQByDbOsxgQR!3bg%Dc~qC
z**eH7Sy(lKP8mJP3r4flo8o@2H$PAc@`ZPaXLJ2>r?ly#1c)6Gz6=ZxJ-}ZlJOpw}
z-YKu|46Z+9?^u=s{+9kx_S6e@iUec1x6Do6p1LE1EI!bbGGc-i{YYWv%ZPelD{hp9
zrb}k2M_%WY32C%H=zu)-I&jlU1E!#(1L$Ma+0GYFGs8#EGUqM$p$ugUz4G@ES$Mxf
zS?=~rg8LU=+M1~I!&Bz87<I112nrqCnX_lp#*4;zTpK2~ttnyop<iNa+VZ&-X@IiX
z8X92W=I7~*Xy^hgcx+K)wz0V3sTTK~@X0#qa#%XiGiR4nW>$od)0~|NVsH1noQ1Na
zrR8&SdBT+27O|t3C@im>>udasR`jC=4ub})j*miC-r-jC08ZckX8Z-gA3!QWvNok8
zxiq>+_M*5_0zP42L{H@4h8JbVpRIs6^u2uUTzBvq;&!AhUy&7)u;py=$bt@z6#*Ri
z!=ODQ=RDbBH9srut>}#S&=0VRGM3EAxzM?}S*=%Au1vzw-gADFf$*`TN4o=Bwq&hJ
zAHqrjX*jk_om*K=j)}|`2kjHbj%xa{tk`fyw8}A%hA({=Mhx5JY!$-Jtun!u89{k!
z;i8`f=~r}ME}6l2BTwik=!prqq$fX6VRr+L)fO}2dA9zFFlV@qEX?a59=$M_M`^ku
zCAtclrLUe)IAn%*P^{=&@;_+b3;!ZV|B4}MZb|A0t81(l`SZJs5!gD25k;Dn5q`If
z1cq&Q^vyN_fv#tyU&a7BedbKakEGIfvs(pSl7`M$&83do;cw2nEiPpJKL#wu^1Rfw
zpp-CvSz2C}aZa}C$BG|g7b+@-62=f=R{iqZD=pvAj~a*?ICvTey+gHJy?b4HU}bhq
z4V%!4E*@Fb^Een7r%s<qPM$uk%h)t(q2tL&(fHCK(nzmJ;}7g<F-n+8p~G8{hI|l5
z!(Tk>m8TYDr$aj>4JR^(Mo(0dYK2ai2Gin^i3S<tX>|=FEp%u!w$U}$K#`HEhx?k7
zh>&}v0pCEl2~&Mfc~^DTOWuk6KbsMiiC7(ksJPNhXgqj2>n>^O06M&L_U_T1;5>K2
z5Mbpeto|5}E?7E~&pR4c92i_OyDj|EKWIo$CvKQ|hLjGrQkD{<2gy_MREp9c5eayu
z4J{43{H8g87K5*pMJc(%m#`yk;aQpOXL1Xjv%)f;tjZZ1wn&j<;^t+mMVQG?blY(l
z=u@6@p@a-NPYVso$+x#FVyuM0c%sISwXqmWr=+joA_}%TGr)!p_^0LUsI}u5X7jR=
zz@VBw0jD$j66Ru^tdAfMOWOBvD`*p!eW<qg=2C7k<(yE}?f{IZMKR3K$2l?5=qJe&
zjyzgEsZ;31@<SeR48()c<N7O2;b%<yRQg-CeUmTBhl)2VgZ6C2V=OtUhRKxbk^b+j
zz`{a}P3eCJsjK8kjREA8^01vE=o{lA<RRn(X~|#Ed8iA-;m9-cn;y|cw~U0&DwCvA
zKg_Y(#Mpr??Y!GKR*A?9F4UPbs&6J-kI5xS_EBKUi}ZjdmtaWwU>Gflu?1ZmhOW^H
z&%w~NGD0?tAL{z-Y-XG(@*(}H>W{0^2F-3+eWTqx=x#UkQI%)Y$X@fUlIb9(HR*Ae
z+)Z+2dEkswX1SC>Nj$SLG!{5IGdd+&Uz-68ouL^EGP9?K1D)57!UM}nk#$7SV^(YT
zov2D=?O!H`v&bZ!t@a}FOCFf5!th&>%OFArMc$chYNRHJMpvML{GQ(ekQlWO8n7V?
zzVDBC7xYgDpu9u9e>!F}<Wz1!3C?gafNa!4ilno}6bC(-y5*&lt!B(-8?q*cAH_}O
zFk>pk)o&3AE#0Q3G@z5xQy8qDKvsFBWCbexOW9c`+>>`;({wh$aVrnB^&d1{Ivo~l
z6-L6=fte5J;dWsgl;wf;drc;__`|@#QBSOIKm(zF2pZ6p3PYAOJzpyAF~$q9w5xM7
zVK56lR$1uWLV7T4Se?gFj%>*aE}Z4jnR)4tFnTbmfMI6>G|cf#-xK;Qi!JvV&|94e
zyrtwtP6p4JfmDQV{a!E<)<rJzNS=a`z{Nf@VfBU-(L-jeW%SDWD>Xu7kdLD-LK_1R
zU6J>&-Awu`S#dLCg>r%(ItbeZMTV9Gc?^UjvYNsQ6?`!sFbpz%`f^e6(w|TkK8CQ$
zL%*99k;-Rfgdv!fGln7b?UITet!v@phD7yhTi?-lFs#B>yYQy64KmxthSpdu+z_4N
zTRjP@eEMd~N)+P~ecKUNT-Xu9YMa(EFL(erbiY;oaaG!&84a54$EMvp=q|61{PUpl
zml-*xkx}u<%V%{yAAi`MLmOBwHGTJDdTByAl=LxT0yP@lvqWu<2GtymleK2dg5em?
zSf@xMpEZ^&J%=rGWWeAR=>Yt&ZNoZ0@)R0Z@*~ocsp$4i1FpYpEkjvj)0)sBMrfix
z>yOx0LUrd^--m*#PXoih)K2LR3z&@t96bAIDyu^F7XTH4(T!@<45NDWrnx{k{>y+6
zx@?8v-PQM3-(duh|JwFwP)T=?1G)4saFbw<@kgw3ksqX#7C{D@sr(XFUP?RQ^Xuyu
zI5_je)A{pjZ6Z1H6H9v5E6t{^5~K!_Z7B;$wE5FI5vKJ@BIzfrZG~Q_RZ_1+=jN<0
zv1jwm2xFkn>QUWRH^Zy*>96cIGFBGo`)n(mt$q~I;JvZ!U*tt<O&81t_2jkeO7bpn
zNkHpun#Ly|7UpHAg-j%32!@>tfgdYq!BJm%$LyU9?;CPvt41kkl}CLSI+h0vEYn97
z+HR#yh>;R<L><h=BIw(mk&&{?JAHq}D8i8^tA|zxiMJIhjLze-1#>Rl!pV(zR_A~_
z-<)s9b~A6;Td|v54nFPOGIHv2ZtqSVoKHW9W#-#gFY6%TM$0jNM$Z}!V`lqtX*9IK
zl}6SY71F%(xUCvuPh^+ERkG5c`aQ$DBtj%xqJd!smHXA$lw{>_MVsQyPo?Kwv3H!Y
z9?3l0AcVFc%uIVk0Mlohg*2$E18NZEU==``I=K-}Q^4)ECjl0ogdnFpW$>v^BE5XZ
z%hCc6gK)l65ez<VL&quQKt4neXL6?i-Ae-n=P=A>b~7?$zz>vsCzS74*JRc_d;~je
zDeJ7t4Rlf`Feo_7V~(;nHzz9;tWw~-`a{2BP?>M-waJl}A%komV%8sn4N2Iwz)BEX
ztNessA62c3L2}>E0cQhi!e#j(t;)fBRdy(_rH;I^s>CNFFw9uh2osX&B?3xFw#CWW
z8*`a{kkM<{Ac}pqKL+EM6&B9nurjouBt~U;7IQ9+ts0A%-pQ-345!V3Q6k-%5cPTI
zQ&r&c$ZF^2Ot>I}vO?dSSL3@;VYM+cyo~Po;O|vG*}UeR4_ZAx@^BE#ENjn(*Ks2w
z$Qp_&cGduSm@4I2W8~pS6?Sdgs7OXS`>28Qtvp$kn6#yrJ<w~_dsxKY!}m5ep(Gf1
z^1gdGo|%Eo+5u97v1IRh7?=nRMGV-i;OMFjLl6U0fLy6-8Zd$`gOdOrZV1ynKn`OC
zUnjsrSnU@`iNlz&K{L-Cv3!_}(PZ$%5$=kiF+58?_`Z=6V^A9mCz4qyy1y7gO{*+3
z=5}zOEoCdKvI4^4b}@GSYD%hd)$*DQGK;<!Q)zgl<uhu}&i-)2Sup0pq4+93J}*^p
z`MOmETSklR=j_D5U<@lh;bX5y7bkR!TJQEcxg<a?`>}1^o=-2FPpJxrZDANeIJPnF
zhY?gIVV0H|J6T-Q(2kxz!XUTpi?o3uJg#TXd>hh~YJ>|KpnWZ?-%&^98qhwfJh^d7
zth@cH>NA3}K}-frLu!lcbcCI5OP&o3fT6?M-sb8~21xRc0Y(=#3_^@5oup5*c>J>X
zIvB&_C9|-{(XiAeOY;>%XM06fi}SO;RkHTccM-<NMGYM+qgfPR6o&vl-iPyPO@z(k
zGf>ROHKAiLg}7QS266a({l=(!FDys-yxX(+23O^KSPPqPcty{?zWa1D8?r_?gi?Dz
zjRt++gm@30y!%Fz%ieU06R7eKE`JMcg!YKiZfM77L{qB96S|sCn<4?-LMH*}6K19!
z`U-<jKxjf{N76(vwOUj@1qxf+Qb0}CA4LpWQW6Rq+AWfF^&t=^BV9mNNLbTQc?-%l
z0QUZRUG=8y-cwWgOg8hqL{)KFS{N|T&9W^iY@eFxZiUOH=lzyCc=&HPc%Fq}(R0R+
zBG%klsbna?g_6nQ(vTIEPlwKqe7_1qh*cmAA>HC9D+I>F6S6q!2LGg1s<CYG!0L6;
zDN;{g5=7P%ea}*Xbh03Yr&bv)w$g3Omhv-aH*_l@XKzhLwwJRlE;Ddg<>ce%%`SxK
zvMUG}fAaxhmBgM+kHnLIi|@^cE2H$)0bX6IJA2IgMvr*EBSYVR$DDQmN0dJcq><g9
zpilRzYl@bhAR<48yPkcMl{bucF8^R#CKoH{n1)%9)R}DNTYTQHOV7g(`t14m@SyYz
zuZR{}$QPx-SNW#BBAQirqf9g2`1yKdX+``j&HN&mVd6$b*VFNB0QK8H=VdMWsAwK$
z&}UEEhuhKvmkq}p(jhN=AeHX};fm<`I3K?2&d(}NuMs3Q+6FF~n*MMglpAPpld|X7
zX7j@32H`b87{g^IP`?H(jS|;&+H9WfC8ojA@BQRkqLiP}Ngw+>3b+V{IR47gm5|FL
z6~Y|*xx);zP9y8q4bq?`|FwH-IumMmMKT({LHwe85r@n3%EKC*hb@B1(^29jKY4t>
zisDMp`=ULr4J-j_BU2yILl8l}aXz1>3-g;~!(cr<i!*$CHkdqL|7<4#v$sSV8Dy0R
zl{knkhK67fPJPWscxeKh5|sgp5;C9s%3!6F5|MCs(3E}@b<9tc0IUNqnm_oZofly-
zgp^oU2xUZy3_uZ$b^-S38?SMaD1;c#1nBUqNG@Fyc}~R)=?>Q7C5l+`PLzXC6n?67
z-UZslfXP?KLi3R7HxKLUsHIt)`NXrOj1>zsp|j{Rh~%1T55uYk1NE@tA_L5ot|_Uc
zxs(PvT2LC}enijEGM#30C6`tv0<Lw+JsSo89;CHGlZ+019tH-E?O^m-mZp*EkHtX`
zI96!b=I8r8J5w8A!j<2k`)zXcrA#C;kni0x7xGEJ%gtx325rffLX0VHH)Zve0lLXY
z-)nN?`*U2#%0RyEf(M^J!j?yPnd5w~#rNhTJLhHnz2O)<bpm<KW$_`#%9JP4e0tcH
z7|Mc63@i=5dnQZUjeguf9ePSRJf^=|2P9s%zsWhEGfQVxcj$9@YdJY4SFI~d4_4qo
zn^?7-tXtOvS6Sz1jOx+&K+lJVrC&yCn7owH99GsQ_{Q6OaHu~#TUZ0RAA>hPmpt@i
z9bJ>J2_t(ifTis)$2i?TTfFz{(mib3u&3?AP3V~M#z#_j>~=@V+m>8_;As)2eK{-d
z?lIlg+^YlQb=ehHYmi>2>u9)(UQ-%^?hQB6vC&}DSo1vu9EHP;YQYf>H?L3-x`$a{
z=^Vm7Yv~|m>qFSXOyy&*cxH=*gQHDeukbOml&N+czTkblKU{0u2rRwpOa15SO|Bs2
zMjYFFOec&H*Tw+q(u{$DHBi;5G$E>k^(KH3rj-Wr<)ZJLygT;p;oE>LPuLE;>xWbY
zhs7E}Os~=wJh4c^M^?qhu<1J513_Sr$S@-9tol2J6fR8%&S19FLqF6DX6<#jt%J_M
z#8C7}SmE$AUVhT35O1QT%@kYeSk>Tfb5m~Ol0U0|7+L&nb__{L9H5_U@Q6@hpc^tq
zS=s=eG}^|Z{vp7m$^cLt=fk+_8pq5!1AIF(rjhbPyUIBi9EOly?JOM}Z8Qh}?Xv}z
z?g!yvfSuzh<+LR<Gd|%8Z~7hI$0$Sn38hN;VJt!42D(i2+2DS?d%d>6Xdf9Ugs{nw
z^?T%nK7_sv-gad(+x`4X1_OF?e3xccaQL)@cU%O3&eIanZ|H={N6w>RFonUJ(W1Yy
zNlJ`Q4)&9Gw)JsFh!wdiCq>?4gkg{|E|?NL&c@(Ef1jQ0>S#WKE{!MWhP*M?TSvIW
zAiNLMG4ifur1gs}rc87MdaQ6^3?o0V7=LyvEn@}n&@=iDD{edbeW~#Fa&uv3D3R{#
zkm-T4A2$OJ1&Fh83%U=NF@$eXSsP}!_&tLtpO6b%P02;a$Ebm*ffgFDwxJ4z7Sjx%
zDmc={uD|~JoogQT*e)Cl_()y<N5Qg#k+c(`iIRpv8bD^GeB%x=8lMGe2*Ve>L)XdR
zgHA6rV(iBsPfi>^-fiMd2xgi1Fj|p=#)LFxCr%ufMvrDCYg&+<gSW^7Z8}q~_Vh5&
z#GpGXGd*--CypP}bUF|6Kx1m~TnEaB)NCLH?tVJtpd-jT%A!9uK<R)_a4Zi+_(rdc
zzNP2=B6#DMhpl^-tV!Pbp^A9)L+wsXm0sI$98@Q5cz9S})_MKeGlLNA{L{K&^}^CU
zy`nIA0>^k2!PLdkQ1Q;1Posj))hT_R!vHYI6`qA@;7zK^kk*|pvQUt^6fjeA8D)!*
zHi9ifMgvZn^P~hs>n7>MXInmnLyE*UBxEr?W|Xp!>DK%#r8mT+({4B$j2mqR$P7Te
zp2#72yzk_x<iZmt`toCKCtHLVD4OG{X4*>cx`*f`l8>q~IacIYvEaKu-oRiWydo<{
z&@JmHko454Q#xq&dWTN<xN0^$&GWZ1V6g5#&WYhe-$QxwogxO=eCG~>4I_eQ?~sA-
z_1KCF1_w05>Wo%Hp(O^AY*pj>@KA${!kEHhI0P527^nZ^T55~q+d1dtquor>p;PLd
z)mscc`fIz<j(ChC^vUjxOfNfX9Pv@`d|#b13acPerbYDuePIB=M^~eSuXpsD@HAZD
z=!Z`nJKhDuI6EWXd%`h3K%beRB;yXb;33>}Cq{_W%{f;$DHm=5wD%01g>O{Rh6Fvc
zvUpB-(w9RQ<0%-Gy>B+(9HJgyAY%{DpKmHrR^VWKLDz`uija<Tg7JnvaG8GfabJI|
zP9ua54=Yrj7Ilm^(ELH@9~AxB_R+LB_1y|3ORl8eSOQP!qxs!@OFZ?Ut0$|F;#&|>
zbq<h>cRw_uupPFE(-q(J<IpEAzEDw+Z-Pl889(<qkstmu(sr)1l4<~gV|gQLHl3&-
zyq2y8gTu51x5hSgY+;*haD}y}2M%s!g^O!sf<xX>XcW;xo(*XakT;)x$0u%6Ixl<q
ze^+0PpyQWcp$~`Ur(4-_0$aTKx#T^4+zlCuXcC9ZH^zBbgY&ROFnKykyyPd34_HxL
z33^|29+zia8=3lu9)gIXe3~xIZ;}mz_4F*x@a@@PEZ+O3Z?qq7d6R&(ZxU(g;B@Ip
zqeF37L(WQ0HoyQfzyhRj0;t7Hj}j6y<PafzfRV;o1~g&W2#M%y6hMFUQU#3&jdh3=
z7&Y=jV;|@+!|ggMk(U)3HE1Nve@VvE!5#*p@zH^o$RIIfUPa+zGrzV2Cvd~Wj7rh>
z<8|(J*atX3gq1@Ps->A<1e1?z3k&|J^OTpV=ovY2o_A4P9@p~M1U8R**q_l@`qmw~
ztXR2~39iN=`LlMV3~+#Wt!dUMr1<culc_F!08t82JRct}1a1*IN?OKtH)tB|I^THX
zo3nD{xvcJZyfPT0H;8Za%WzHJI^LmQlO0#GbryH!RadFcPu+I7y7Vrg0Ro=34>u(R
z4rhAU4%d+dib<J}V(Z(wOt8(!q#VxZ$~t~BnXb`VoTY#I){iLcEg$`$`nnIzl&dj^
z;3w4GtOA?JmJjnSe!u+PkfVrZn>hcn=kfB-P4t$>$v-!t;bA<!4||yNEnJpQ;>*&?
zbYSp#-@_YS-k@w%BS4VmjTa09bG&yVQ5fx8WcNAD1d$F3Mnn!!<2mkQh5{2(`Yy|l
z8XeG5e#}aFd{$YQ{T4$+d^JXlzszif@s40fnr&B6gC{jp)R3|?eE1=w<tL@3`BQ;X
zI)QM;Gfe~6t^l_VN%bP}LbB3Xxl{rTkXS_*BaSkyg2}VoE!aGtVV{xHw`rbURUUkr
z&S*PCia9NM7u!5IEPiuNgF)8S(ZQ(h9hNV<KCYtt-OOJRZ_2ys%`d{~lY2xFo|g{~
zLw;0rybQZR)6=gD!=Jxy`yC3gvW+yE&9CFpHoYm`Iym~LG&?E+MuIt3ML0uK<^1IV
zRd9o(dD=eQgch=Q#0Z*A7LqLtBL336SRO5Zv;u1~HkcX<=Ov4HefT|Z(Q$wpK~@o(
z2y!rBT;&ykD0Fj{7xOKCuYD@&sAVXkX>mooe4N3d2Osal`LsHnd72(JPt&JWg-1=V
ziq}C*^Lq2T@bOi08;<#gXFj-=rs0&BPbX~8cCJOi(vbyQZHNy>3IydZ9WjQ6*YN3N
zQU-YPk}4>H^o8Ym?C?MYD=b2nRUdI(RPa(lYWzroJu9I81OrG76GhB;XDdZ1p0HE{
zO;>sZPM$eXO{0kvDSB#@r5rR;QUO3t@i`lX^duo3h8@YKI|=qVulp*vKE#F*Wi&B>
zEY18TB+FoJrZG>W3_cGx%I9tkO6wrBt$x(WL%NlThjpX|5_MSofta>H(rVlS{vd%`
zd6mgzTeN(6mch1-=Uvb1PU{eDM@^yErMH!(2;S?)hl^m#;%M9R=RSem4f&zTCkB|X
z%ImdEzEQ*%=>}OWzACSlX8tJT@ieOFA|I{{rmXC2l>}UFQS+Aw5a*p=HSzJI2z%N-
z+>|!5PfUiDAo}Z~m7t>hd0vFgJ^rcB(O?*HSgiO#%`Req;3#M`Wn1*}c>X@zlxBn*
ztp=KuK@*tKN_GFGme;j^ChVV8OA=uXZ<a<g6rRlY8tuaIf~GHNew+Z!@CPOa5P*hh
zK56bP&l4fp94MCFMG17${v##TO;k)emu`AWi-A6(5O05)l%i^K3e!t1C2)W|#H@2^
zRw}`a&VkCW2~{b1bz;8k2O;Tk^7N{?m#2D~21Q&^yrmx`e?R2h&Ea&u-dq06KS;VC
zRF*pZ`Mmn@!2IX~21Sl|JsQ|+O>$B7ZoBj@*e;u*v4-3++D*3raCuxy%lm2sQJJkm
zch3OiAzM6;f<_B=dkSiL+Eb<tn3c&@7T<v6$Uyv{piCYQV~opSE&U+*qmb2Tm_Mrg
zLC@1<wp5*XKAt$6ZR*CS0kbV^wM9vgMh_;wHo%1CPib@sNF2_?gcyMpU~uF|apbux
zngvKzn1JMhA%b7!6;jk#bI*Z{L>N6y3phkUTa#qVm^P)BF3$w@mO5E3vfJvEizY~#
ztYjHk#HeycwhE>wZBV=-T2*o6qb@I&?#rW!=BU$(bWs%7mamnq;ZetpuKA={*!vdt
z>D~9s2oC6*zCF6iCQLnFpMH?SMe+|KU(>uQTwwz-2-kj`GC5dH9>#gPgzaw_nuFr%
zJX8g;FH`F)D!^-#>iE?X7!d$E*NLwjJzqAtq{7Fx@P$MMjBKXM(Mh=CZnrElL<Y=8
z58B#x??n)u^kwt8*eJMpm8%_kE_mpn+WQc;!@0OoeEobBG@MASgT5W~;nsu-5JpX}
zfW<&jWpb6p55#0ZB>!@j$pegUU}04TJD?c5iZ%*ajRwAd$%hr(=o%fiJJ5l_V@Ixz
zp5W`Psk3ceJnFWHEJ<#iau0@p-;GWK-Dj@56)y6OaiSbyD$mTm=-9#1^t_L3F_Q$)
zx%QU9sp<g|8VQ!hi0_?H7FeA6zzSS<Yn|kEjGGKJnyo8rUz7Yzj_Z}y`T<wrs?;SR
z&=LUfI7=h6k>AcTg?O&jl`D5sozx8o5UEJ=AXND;c*D9TG|GzB^KJs;;YS@V%7?4E
zBV%V}3PI1)!g)Ur=kbRTsz~3XqCJdUO>v9lZ;I>TOy54PDBKjT2)9Wb^!+<*{H$qd
zLZb-A=#b{aMX>uF=gYL)9`3hLBjAtTecp^@+wy1&Z+!O7ADx8$3hIgU5`t0ZV}3mQ
z#u2oq<%|!`=VH!%Vn-9@hf5JA)2;OuSLUk`RF>%DdHFqKpSGXjwsN(FZ<Dv(q}@i2
z`zvhJf4h>`0ptLqNc4=uXl3xvz|77G@!VtZ1W*|~NWA<DrdMVFL4%=%8WGaKU~m|H
z455IL#t5+DW5pOyHxMHH86a)20t1Ct9z2V+vn@1O$(fi8jjE<;o}fcZhfE=4hESL(
zlr;DQA?Y27<dK*V8$`3P0>+<R(*UM#YTh>DVWsPVF=pGiA{uS*_Y?Lo$ex!m|Gc02
zE8>07<tvhT(73~rSLDQvx*ayUfqMMiYYk}sFHBkKX>7dI2-&#xA9ZD_k*jfXTe;f8
zHx@(ycawG-IrbYi-%v0>-U~Ns{qK#`d!d<P9!1Ol#gM(RZ1?HvIt_FU9C;4LMv7yA
zqRe2x@SXq^(k<C@;2?h(m<B^6yjeg70}EKr4~3PUUJBuMI7!3E!Jwk<@gJGhS_$yF
zl}!{7K5{Bes=#LYNQHx=-~o$MS|OyrX8c%QV1?_ID8#k-&E#l<6=9obpr~xJ(r!Xw
z`0{PWr5K8j@ebpP@?S)6uiES)-c4mJ!rhNJd+&$ky`MTa+3RHRM;(UU_BkxR$?ScT
zpE|S~^=LO}dii6xR|BWzJ8?yBBmEF}Q5tbfrJhZ{D^v#a_Ip3CNsW!5GD0!_@N1y$
zAfPVC`G%tCwwXE_2O&M5kAoo7C}k*G{%=Cx$`gDGX9hx!0+q{GFb$^>;@KGouxv2F
zmCoC;O|Oe*2^cmKw(FkxfMve}590}=C%qL7Yek54;hiz$jVbbDS8Qj@1vQ|{c8T<w
z=~R5=gpd4Jr))xSsNX@$WhkbH6xj^aVZ=X7o#50|Sh*Y~zfExMy>0@#+c3O-_Xl?n
zbY3Te-;O!#cDP~ji|Rz3{6#qX5tnuu`@#B9@jNacJofJFnI(Okn#kJ0B3-hF<Y>)R
z;5T1%rP4Y~?Z4N>p$QiiP?hj);j3shO}oEzGEMa{2wYvc)up%PIZAqUv`Du<<h0o~
z&imM^g|4#rOi-!hkNB_x!?vYR)6{54E1h1CN-G4MdKfP9=Pd!XE;1?gYHfK04p+1y
zqyh^;Gip=@B_Tif(3HPy5Y25eM5z=aER8H-;)D;dCb+_+@uW9sMU}NuYxgrKUOO%a
z7rHk84flOl<<RPI70f|OL*81w93+8<l%+1Ob?HO0@1mFSDU0tSo=1Vh_O2XM)rX$t
zph|ae<jSuE9X+<F!g&LFW4nKKyv;=o4Z823ko(h%KFTW4wz9W{AEm63*5DebE4RAz
zwwy;vuZ|Y!_J^Fzwy7bMBAQXu9hn)yT%)1e=+ZP6aY)9lQ1eIpu(HDHiu}ncO9}zY
z;4nVGPmK_Tk!Lk@G)@x*Kq-rXlQxdFR09=~QbAKP!4P5#9!ZcOJ>`#4q$@~4P)U+B
zyP-yXq*tVrD){{_K=jAg#SDs9NB6L#*X6%1y-7a3z1;+UzhRJf6#3rod^F`>m)E-V
zrhMw)%J^72Qip#&q#w3-<$Tb2%tNLD>ubH?GOZJq84RH|7d7bFqV=`4UJCIkY9MML
zYG98v&_9o}3yeJzA2wDJR}^i@_<q7^%iY{rhv#X;aMVE5!1<>E*BqRGnTt7!8rUZd
z<hR4=9Hz)(K4NuthU<a7LA6iWjBKL@q6VS{q6VS{q6VS{q6VS{q6VS{f(BM|-w7&O
zjn&U^452Vx3A*yCtA-%4_o$uDW->Y1dw)1mtgo&lGjsD@fA1yBIF`wnL9`DQ(IgI+
z*GC@K;5=**OrDMsFZs#i17>AVf>Ifsr)M&^k*Siw0ZlB*r|H7{CZZdxr)P17Z_frp
z{MP1XGBFk2C&Ii#qjK!&WlPzdTt$6uqmzuu7PYy#nPf}q?fUjuGA-Zg;bUj`>zmth
zTbu0WmD||n=ckkPbL;ZmAo&)L;BKv~>9>ByWI1MieJh!qpB1?~BKKsnEjPt%Z)_&x
zGt<fTn!8O#wyt$H*W^1$Q^}aL7t=@Q^}+>ip4(nslP!JHq_VwD_IQlT($2997L(O;
zs|h?d<by@ybF-obp|RdcCa0yTl=dZ&gPPmI_xQq0a&l!e*^%(Vg-4R<j^s`L%~t5n
z>wMI8y6U~NG%NciW!n~Z)b*x`jKG{)IhQO>PxrKcrspahd!0b13YPTf4u%Hm>Lulh
z8)X~jb8dBuIIHt_-*aDb=_MC>u%2+8%m&QtcF`cu((8QB$A|0i>e5F^%XsYuO`~1s
z8*hB;`*{C(#+9`N9<L0hj(1hM(JzCkldlZ#Fvri0_g;i;D?mN($&mX|<4RCL4apzI
zDgcR5Q3Fu}P6KB*H@Z)}KRlZOM1#fGrM0b+hZQk~>?_asT#jJRlVb2p&q#+Px5I6%
zZzP?yO)+G)#E2PBro<581E&}+TS5Z^q_Zl9k{Ce5?T9giA%q)CZ;!zM+S(Et>nq9n
z$umi3c_rC6doFxD6$41zPHI4{3m*(CJ8QJLvMzii?%cU#{nVLca&|h|5+iFz3?9Xk
zQ^M27=9A@nmt;wPTMVb+mgSXb{As0ujpe=D%6)(77nW+SmTgr)_XFVRBTSYI?|KGy
zHRSF(6>iC@0ULSp8Q0E_oOmP{HIUOlleVZMeRSS&MRrryYi2dYI?v(q`w-9b?8fXx
z4GgV;qf^27fRKNG#5B<r5@TSqUFKW7`S&cHlVBJQOo-7jJ~5e0PfsLkr`8hY>5fi|
zQFDAc*<9TayGM)=S>4%^6{Kx3n79Z5<7Z4(24q@Kd|BP<h|wfQ5DsH#OjdlR7LN)G
zx$r<{{>DUZg_qT*BLcrZHk)jSp`$Z{GQj3Chlzy+HMq9cHj=3YS%DJ64aW_38)sHT
zh6SZv^yO*=j6b`kfeSB?vfLx*>-8S7+!r>BHCM~-i%<uc>84E7B%@nAZ<N!^PJ~;s
z8Uc?G=$LEgM@Tl3I;0vHcJ;C@+fCY{h1vdHVRjR39evXOKNL}d4&pJlu>b%707*qo
IM6N<$f*8B%#{d8T


From c677f11cf739df628c2758f84a70946013e90536 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 11:32:54 +0100
Subject: [PATCH 30/33] restore the original download times record link

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 343f95d..3cacc1e 100644
--- a/README.md
+++ b/README.md
@@ -1,4 +1,4 @@
-![GitHub all releases](https://img.shields.io/github/downloads/guanshaoheng/PretextView/total?style=plastic)
+![GitHub all releases](https://img.shields.io/github/downloads/sanger-tol/PretextView/total?style=plastic)
 # PretextView
 Paired REad TEXTure Viewer. OpenGL Powered Pretext Contact Map Viewer.<br/>
 

From 649e10334115724fe431e9722f8f4c1025cbf57c Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 13:35:06 +0100
Subject: [PATCH 31/33] fix the bug of vary length array

---
 src/copy_texture.cpp | 24 ++++++++++++++++++------
 1 file changed, 18 insertions(+), 6 deletions(-)

diff --git a/src/copy_texture.cpp b/src/copy_texture.cpp
index 81ae13e..cab6001 100644
--- a/src/copy_texture.cpp
+++ b/src/copy_texture.cpp
@@ -39,8 +39,9 @@ Show_State::Show_State()
 Show_State::~Show_State() {}
 
 
-void get_linear_mask(f32* linear_array, u32 length)
+void get_linear_mask(std::vector<f32>& linear_array)
 {   
+    u32 length = linear_array.size();
     if (length < 2)
     {
         fmt::print(stderr, "Error: link score calculation: length({}) < 2\n", length);
@@ -50,7 +51,7 @@ void get_linear_mask(f32* linear_array, u32 length)
     {
         linear_array[i] = std::sqrt(1.0f / (f32)(i+1)) ;
     }
-    f32 sum = std::accumulate(linear_array, linear_array+length, 0.0f);
+    f32 sum = std::accumulate(linear_array.data(), linear_array.data()+length, 0.0f);
     for (int i  = 0 ; i < length; i++)
     {
         linear_array[i] /= sum;
@@ -867,6 +868,13 @@ void TexturesArray4AI::cal_maximum_number_of_shift(
         norm_diag_mean.push_back(tmp_diag_mean);
         tmp_diag_mean = this->cal_diagonal_mean_within_fragments(++D, Contigs); 
     }
+
+    if (D < 2)
+    {
+        fmt::print(stderr, 
+            "Warning: D is less than 2, please check the Contigs of this sample.\n");
+        assert(0);
+    }
     return ;
 }
 
@@ -991,7 +999,8 @@ void TexturesArray4AI::cal_compressed_hic(
                     frags->length[i], 
                     frags->length[j], 
                     D,
-                    norm_diag_mean, buffer_values_on_channel);
+                    norm_diag_mean, 
+                    buffer_values_on_channel);
             for (u32 channel = 0; channel < 4 ; channel ++ )
             {
                 (*compressed_hic)(i, j, channel) = (*compressed_hic)(j, i, Switch_Channel_Symetric[channel]) = buffer_values_on_channel.c[channel];
@@ -1035,7 +1044,9 @@ f32 TexturesArray4AI::cal_diagonal_mean_within_fragments(int shift, const contig
     return (f32)sum / (f32)std::max(cnt, 1u);
 }
 
-
+/* todo 
+windows： 编译错误，尝试访问 d 
+*/
 void TexturesArray4AI::get_interaction_score(
     const u32& row, 
     const u32& column, 
@@ -1055,10 +1066,11 @@ void TexturesArray4AI::get_interaction_score(
         };
 
     u32 number_of_scores = std::min(D-1, std::max(num_row, num_column));
+    u32 len_of_weight_mask = std::min(D-1, (u32)(number_of_scores + 5) );  // 往外多算几个像素点，因此小的片段的link score分就会相对下降, 特别是很小的片段
     // f32 linear_weight[number_of_scores]; 
     // get_linear_mask(linear_weight, number_of_scores);  // assign the linear weight mask
-    f32 linear_weight[std::min(D-1, (u32)(number_of_scores +5) ) ];  // use this to consider the score caused by repeat 
-    get_linear_mask(linear_weight, std::min(D-1, (u32)(number_of_scores +5) ) );  // 往外多算几个像素点，因此小的片段的link score分就会相对下降, 特别是很小的片段
+    std::vector<f32> linear_weight(len_of_weight_mask, 0.f);
+    get_linear_mask(linear_weight ); 
 
     buffer_values_on_channel.initilize();
     for (int channel = 0; channel < 4; channel++)

From 3200506822c3339bb75004fa9110217fce096797 Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 13:50:38 +0100
Subject: [PATCH 32/33] fix bug of varying length array

---
 PretextView.cpp | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/PretextView.cpp b/PretextView.cpp
index b6b60a8..34604da 100644
--- a/PretextView.cpp
+++ b/PretextView.cpp
@@ -5510,7 +5510,7 @@ LoadFile(const char *filePath, memory_arena *arena, char **fileName, u64 *header
         if (!is_pix_density_added) // 未添加pixel density
         {
             u32 added_num = Extensions.get_num_extensions();
-            u32 graph_data[Number_of_Pixels_1D];
+            u32* graph_data = new u32[Number_of_Pixels_1D];
             f32 max_desity = 0.f;
             for (auto& i : frag_cut_cal_ptr->hic_pixel_density_origin) max_desity = std::max(max_desity, i);
             for (u32 i=0; i < Number_of_Pixels_1D; i ++)
@@ -5518,6 +5518,7 @@ LoadFile(const char *filePath, memory_arena *arena, char **fileName, u64 *header
                 graph_data[i] = (u32)(255 * (1.001f- frag_cut_cal_ptr->hic_pixel_density_origin[i] / max_desity));
             }
             add_graph_to_extensions(arena, (u32*)graph_name.c_str(), graph_data);
+            delete[] graph_data;
             
             push_extensions_to_opengl(arena, added_num, 0.05f); // push hic_pixel_density to opengl buffer
         }

From 21b1abc38a794dee751fc8420afd64511f5ebe5b Mon Sep 17 00:00:00 2001
From: shguanWHU <shaohengguan@gmail.com>
Date: Mon, 7 Apr 2025 13:55:06 +0100
Subject: [PATCH 33/33] fix bug of vary length array

---
 PretextView.cpp | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/PretextView.cpp b/PretextView.cpp
index 34604da..ea20743 100644
--- a/PretextView.cpp
+++ b/PretextView.cpp
@@ -10718,7 +10718,7 @@ GenerateAGP(char *path, u08 overwrite, u08 formatSingletons, u08 preserveOrder)
 {   
 
     // 这里粘贴所有的original contigs
-    u32 tmp_orignal_contig_ids[Number_of_Pixels_1D];
+    u32* tmp_orignal_contig_ids = new u32[Number_of_Pixels_1D];
     for (u32 i = 0 ; i< Number_of_Pixels_1D; i++ )tmp_orignal_contig_ids[i] = Map_State->originalContigIds[i];
     Map_State->restore_cutted_contigs_all(Number_of_Pixels_1D);
     UpdateContigsFromMapState(); // todo 检查粘贴后会不会影响 painted 的 scaffolds
@@ -10831,6 +10831,7 @@ GenerateAGP(char *path, u08 overwrite, u08 formatSingletons, u08 preserveOrder)
     // 把临时切开的再粘贴上
     for (u32 i = 0; i < Number_of_Pixels_1D; i++) Map_State->originalContigIds[i] = tmp_orignal_contig_ids[i]; 
     UpdateContigsFromMapState();
+    delete[] tmp_orignal_contig_ids;
     return;
 }